{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 38.61003861003861, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03861003861003861, "grad_norm": 9.624889373779297, "learning_rate": 1.8e-06, "loss": 0.931, "step": 10 }, { "epoch": 0.07722007722007722, "grad_norm": 5.079489707946777, "learning_rate": 3.8e-06, "loss": 0.8664, "step": 20 }, { "epoch": 0.11583011583011583, "grad_norm": 1.8355929851531982, "learning_rate": 5.8e-06, "loss": 0.5242, "step": 30 }, { "epoch": 0.15444015444015444, "grad_norm": 2.5522167682647705, "learning_rate": 7.8e-06, "loss": 0.3724, "step": 40 }, { "epoch": 0.19305019305019305, "grad_norm": 1.2228738069534302, "learning_rate": 9.800000000000001e-06, "loss": 0.3039, "step": 50 }, { "epoch": 0.23166023166023167, "grad_norm": 1.3149656057357788, "learning_rate": 1.18e-05, "loss": 0.2879, "step": 60 }, { "epoch": 0.2702702702702703, "grad_norm": 1.1572262048721313, "learning_rate": 1.3800000000000002e-05, "loss": 0.2425, "step": 70 }, { "epoch": 0.3088803088803089, "grad_norm": 1.2678970098495483, "learning_rate": 1.58e-05, "loss": 0.2169, "step": 80 }, { "epoch": 0.3474903474903475, "grad_norm": 1.601481556892395, "learning_rate": 1.78e-05, "loss": 0.1882, "step": 90 }, { "epoch": 0.3861003861003861, "grad_norm": 0.8347041010856628, "learning_rate": 1.9800000000000004e-05, "loss": 0.1923, "step": 100 }, { "epoch": 0.4247104247104247, "grad_norm": 0.7805326581001282, "learning_rate": 2.18e-05, "loss": 0.1731, "step": 110 }, { "epoch": 0.46332046332046334, "grad_norm": 0.7223260998725891, "learning_rate": 2.38e-05, "loss": 0.1668, "step": 120 }, { "epoch": 0.5019305019305019, "grad_norm": 1.5452854633331299, "learning_rate": 2.58e-05, "loss": 0.1628, "step": 130 }, { "epoch": 0.5405405405405406, "grad_norm": 1.3716341257095337, "learning_rate": 2.7800000000000005e-05, "loss": 0.1614, "step": 140 }, { "epoch": 0.5791505791505791, "grad_norm": 1.1876602172851562, "learning_rate": 2.98e-05, "loss": 0.1521, "step": 150 }, { "epoch": 0.6177606177606177, "grad_norm": 1.1677711009979248, "learning_rate": 3.18e-05, "loss": 0.1541, "step": 160 }, { "epoch": 0.6563706563706564, "grad_norm": 1.7922990322113037, "learning_rate": 3.38e-05, "loss": 0.148, "step": 170 }, { "epoch": 0.694980694980695, "grad_norm": 0.7983670830726624, "learning_rate": 3.58e-05, "loss": 0.1493, "step": 180 }, { "epoch": 0.7335907335907336, "grad_norm": 1.5428334474563599, "learning_rate": 3.7800000000000004e-05, "loss": 0.1316, "step": 190 }, { "epoch": 0.7722007722007722, "grad_norm": 0.9854490756988525, "learning_rate": 3.9800000000000005e-05, "loss": 0.129, "step": 200 }, { "epoch": 0.8108108108108109, "grad_norm": 1.2998489141464233, "learning_rate": 4.18e-05, "loss": 0.1384, "step": 210 }, { "epoch": 0.8494208494208494, "grad_norm": 0.8203567266464233, "learning_rate": 4.38e-05, "loss": 0.1322, "step": 220 }, { "epoch": 0.888030888030888, "grad_norm": 1.01772940158844, "learning_rate": 4.58e-05, "loss": 0.1305, "step": 230 }, { "epoch": 0.9266409266409267, "grad_norm": 1.2549607753753662, "learning_rate": 4.78e-05, "loss": 0.1331, "step": 240 }, { "epoch": 0.9652509652509652, "grad_norm": 0.8164927363395691, "learning_rate": 4.9800000000000004e-05, "loss": 0.1222, "step": 250 }, { "epoch": 1.0038610038610039, "grad_norm": 1.0856412649154663, "learning_rate": 5.1800000000000005e-05, "loss": 0.1162, "step": 260 }, { "epoch": 1.0424710424710424, "grad_norm": 1.2567695379257202, "learning_rate": 5.380000000000001e-05, "loss": 0.1395, "step": 270 }, { "epoch": 1.0810810810810811, "grad_norm": 1.3449362516403198, "learning_rate": 5.580000000000001e-05, "loss": 0.123, "step": 280 }, { "epoch": 1.1196911196911197, "grad_norm": 0.9897775053977966, "learning_rate": 5.7799999999999995e-05, "loss": 0.1122, "step": 290 }, { "epoch": 1.1583011583011582, "grad_norm": 1.4862051010131836, "learning_rate": 5.9800000000000003e-05, "loss": 0.1138, "step": 300 }, { "epoch": 1.196911196911197, "grad_norm": 0.6288107633590698, "learning_rate": 6.18e-05, "loss": 0.1152, "step": 310 }, { "epoch": 1.2355212355212355, "grad_norm": 1.0391337871551514, "learning_rate": 6.38e-05, "loss": 0.1208, "step": 320 }, { "epoch": 1.2741312741312742, "grad_norm": 1.1934906244277954, "learning_rate": 6.58e-05, "loss": 0.1239, "step": 330 }, { "epoch": 1.3127413127413128, "grad_norm": 1.023949384689331, "learning_rate": 6.780000000000001e-05, "loss": 0.1038, "step": 340 }, { "epoch": 1.3513513513513513, "grad_norm": 0.5593318939208984, "learning_rate": 6.98e-05, "loss": 0.1076, "step": 350 }, { "epoch": 1.3899613899613898, "grad_norm": 0.6812185049057007, "learning_rate": 7.18e-05, "loss": 0.1012, "step": 360 }, { "epoch": 1.4285714285714286, "grad_norm": 1.0067138671875, "learning_rate": 7.38e-05, "loss": 0.1106, "step": 370 }, { "epoch": 1.4671814671814671, "grad_norm": 0.8913452625274658, "learning_rate": 7.58e-05, "loss": 0.1036, "step": 380 }, { "epoch": 1.505791505791506, "grad_norm": 1.1695553064346313, "learning_rate": 7.780000000000001e-05, "loss": 0.1127, "step": 390 }, { "epoch": 1.5444015444015444, "grad_norm": 0.6972833871841431, "learning_rate": 7.98e-05, "loss": 0.0983, "step": 400 }, { "epoch": 1.583011583011583, "grad_norm": 0.7710580229759216, "learning_rate": 8.18e-05, "loss": 0.1106, "step": 410 }, { "epoch": 1.6216216216216215, "grad_norm": 0.8190672993659973, "learning_rate": 8.38e-05, "loss": 0.1001, "step": 420 }, { "epoch": 1.6602316602316602, "grad_norm": 1.1941280364990234, "learning_rate": 8.58e-05, "loss": 0.1126, "step": 430 }, { "epoch": 1.698841698841699, "grad_norm": 0.8625528812408447, "learning_rate": 8.78e-05, "loss": 0.1065, "step": 440 }, { "epoch": 1.7374517374517375, "grad_norm": 0.9149879217147827, "learning_rate": 8.98e-05, "loss": 0.1071, "step": 450 }, { "epoch": 1.776061776061776, "grad_norm": 0.6451426148414612, "learning_rate": 9.180000000000001e-05, "loss": 0.1022, "step": 460 }, { "epoch": 1.8146718146718146, "grad_norm": 0.7317308187484741, "learning_rate": 9.38e-05, "loss": 0.108, "step": 470 }, { "epoch": 1.8532818532818531, "grad_norm": 0.921375572681427, "learning_rate": 9.58e-05, "loss": 0.1, "step": 480 }, { "epoch": 1.8918918918918919, "grad_norm": 1.1917394399642944, "learning_rate": 9.78e-05, "loss": 0.1019, "step": 490 }, { "epoch": 1.9305019305019306, "grad_norm": 0.9788500070571899, "learning_rate": 9.98e-05, "loss": 0.0922, "step": 500 }, { "epoch": 1.9691119691119692, "grad_norm": 0.565569281578064, "learning_rate": 9.9999778549206e-05, "loss": 0.0991, "step": 510 }, { "epoch": 2.0077220077220077, "grad_norm": 0.740204930305481, "learning_rate": 9.999901304280685e-05, "loss": 0.0956, "step": 520 }, { "epoch": 2.0463320463320462, "grad_norm": 0.625038743019104, "learning_rate": 9.999770075521164e-05, "loss": 0.1083, "step": 530 }, { "epoch": 2.0849420849420848, "grad_norm": 0.57440185546875, "learning_rate": 9.99958417007713e-05, "loss": 0.106, "step": 540 }, { "epoch": 2.1235521235521237, "grad_norm": 0.5461800694465637, "learning_rate": 9.999343589981615e-05, "loss": 0.0921, "step": 550 }, { "epoch": 2.1621621621621623, "grad_norm": 0.7997955083847046, "learning_rate": 9.999048337865568e-05, "loss": 0.0935, "step": 560 }, { "epoch": 2.200772200772201, "grad_norm": 0.6534411907196045, "learning_rate": 9.998698416957815e-05, "loss": 0.0905, "step": 570 }, { "epoch": 2.2393822393822393, "grad_norm": 0.5729870796203613, "learning_rate": 9.998293831085037e-05, "loss": 0.096, "step": 580 }, { "epoch": 2.277992277992278, "grad_norm": 1.1095328330993652, "learning_rate": 9.997834584671719e-05, "loss": 0.0946, "step": 590 }, { "epoch": 2.3166023166023164, "grad_norm": 1.3313345909118652, "learning_rate": 9.997320682740107e-05, "loss": 0.0846, "step": 600 }, { "epoch": 2.3552123552123554, "grad_norm": 1.0345401763916016, "learning_rate": 9.996752130910149e-05, "loss": 0.0941, "step": 610 }, { "epoch": 2.393822393822394, "grad_norm": 1.1533430814743042, "learning_rate": 9.99612893539944e-05, "loss": 0.0868, "step": 620 }, { "epoch": 2.4324324324324325, "grad_norm": 0.7162007689476013, "learning_rate": 9.995451103023144e-05, "loss": 0.0842, "step": 630 }, { "epoch": 2.471042471042471, "grad_norm": 0.6974789500236511, "learning_rate": 9.994718641193928e-05, "loss": 0.0747, "step": 640 }, { "epoch": 2.5096525096525095, "grad_norm": 0.6345629096031189, "learning_rate": 9.993931557921874e-05, "loss": 0.0889, "step": 650 }, { "epoch": 2.5482625482625485, "grad_norm": 0.4718223512172699, "learning_rate": 9.993089861814402e-05, "loss": 0.0776, "step": 660 }, { "epoch": 2.586872586872587, "grad_norm": 0.7468153834342957, "learning_rate": 9.992193562076166e-05, "loss": 0.0843, "step": 670 }, { "epoch": 2.6254826254826256, "grad_norm": 0.875212550163269, "learning_rate": 9.991242668508954e-05, "loss": 0.085, "step": 680 }, { "epoch": 2.664092664092664, "grad_norm": 1.0050700902938843, "learning_rate": 9.990237191511587e-05, "loss": 0.0871, "step": 690 }, { "epoch": 2.7027027027027026, "grad_norm": 0.6065554618835449, "learning_rate": 9.989177142079802e-05, "loss": 0.0867, "step": 700 }, { "epoch": 2.741312741312741, "grad_norm": 0.7518108487129211, "learning_rate": 9.988062531806126e-05, "loss": 0.0875, "step": 710 }, { "epoch": 2.7799227799227797, "grad_norm": 0.861382782459259, "learning_rate": 9.986893372879762e-05, "loss": 0.0844, "step": 720 }, { "epoch": 2.8185328185328187, "grad_norm": 0.7107915282249451, "learning_rate": 9.985669678086443e-05, "loss": 0.0837, "step": 730 }, { "epoch": 2.857142857142857, "grad_norm": 0.6414670348167419, "learning_rate": 9.984391460808298e-05, "loss": 0.0792, "step": 740 }, { "epoch": 2.8957528957528957, "grad_norm": 0.5961190462112427, "learning_rate": 9.983058735023709e-05, "loss": 0.0773, "step": 750 }, { "epoch": 2.9343629343629343, "grad_norm": 0.8107931017875671, "learning_rate": 9.98167151530715e-05, "loss": 0.0737, "step": 760 }, { "epoch": 2.972972972972973, "grad_norm": 0.6621590256690979, "learning_rate": 9.980229816829034e-05, "loss": 0.0845, "step": 770 }, { "epoch": 3.011583011583012, "grad_norm": 0.5449572205543518, "learning_rate": 9.978733655355544e-05, "loss": 0.0746, "step": 780 }, { "epoch": 3.0501930501930503, "grad_norm": 0.6653662323951721, "learning_rate": 9.977183047248464e-05, "loss": 0.0707, "step": 790 }, { "epoch": 3.088803088803089, "grad_norm": 0.36069151759147644, "learning_rate": 9.975578009464992e-05, "loss": 0.0692, "step": 800 }, { "epoch": 3.1274131274131274, "grad_norm": 0.4731850326061249, "learning_rate": 9.97391855955757e-05, "loss": 0.069, "step": 810 }, { "epoch": 3.166023166023166, "grad_norm": 0.8284837007522583, "learning_rate": 9.972204715673669e-05, "loss": 0.0661, "step": 820 }, { "epoch": 3.2046332046332044, "grad_norm": 0.5856471657752991, "learning_rate": 9.970436496555617e-05, "loss": 0.0618, "step": 830 }, { "epoch": 3.2432432432432434, "grad_norm": 0.48746347427368164, "learning_rate": 9.968613921540373e-05, "loss": 0.0754, "step": 840 }, { "epoch": 3.281853281853282, "grad_norm": 0.7799074649810791, "learning_rate": 9.966737010559326e-05, "loss": 0.0755, "step": 850 }, { "epoch": 3.3204633204633205, "grad_norm": 0.6583216190338135, "learning_rate": 9.964805784138072e-05, "loss": 0.0704, "step": 860 }, { "epoch": 3.359073359073359, "grad_norm": 0.8938873410224915, "learning_rate": 9.962820263396195e-05, "loss": 0.0802, "step": 870 }, { "epoch": 3.3976833976833976, "grad_norm": 0.48943403363227844, "learning_rate": 9.960780470047033e-05, "loss": 0.076, "step": 880 }, { "epoch": 3.436293436293436, "grad_norm": 0.6848331689834595, "learning_rate": 9.958686426397437e-05, "loss": 0.0771, "step": 890 }, { "epoch": 3.474903474903475, "grad_norm": 0.7400741577148438, "learning_rate": 9.956538155347534e-05, "loss": 0.0718, "step": 900 }, { "epoch": 3.5135135135135136, "grad_norm": 0.718845546245575, "learning_rate": 9.95433568039047e-05, "loss": 0.071, "step": 910 }, { "epoch": 3.552123552123552, "grad_norm": 0.5269817113876343, "learning_rate": 9.952079025612162e-05, "loss": 0.0696, "step": 920 }, { "epoch": 3.5907335907335907, "grad_norm": 0.963441014289856, "learning_rate": 9.949768215691022e-05, "loss": 0.0702, "step": 930 }, { "epoch": 3.629343629343629, "grad_norm": 0.961230993270874, "learning_rate": 9.9474032758977e-05, "loss": 0.0753, "step": 940 }, { "epoch": 3.667953667953668, "grad_norm": 0.5776188373565674, "learning_rate": 9.944984232094794e-05, "loss": 0.0825, "step": 950 }, { "epoch": 3.7065637065637067, "grad_norm": 0.6693083047866821, "learning_rate": 9.942511110736584e-05, "loss": 0.076, "step": 960 }, { "epoch": 3.7451737451737452, "grad_norm": 0.7064602971076965, "learning_rate": 9.939983938868726e-05, "loss": 0.0742, "step": 970 }, { "epoch": 3.7837837837837838, "grad_norm": 0.7173698544502258, "learning_rate": 9.93740274412797e-05, "loss": 0.0716, "step": 980 }, { "epoch": 3.8223938223938223, "grad_norm": 0.49540087580680847, "learning_rate": 9.934767554741846e-05, "loss": 0.0753, "step": 990 }, { "epoch": 3.861003861003861, "grad_norm": 0.7561555504798889, "learning_rate": 9.932078399528361e-05, "loss": 0.0751, "step": 1000 }, { "epoch": 3.8996138996138994, "grad_norm": 0.7253299355506897, "learning_rate": 9.929335307895689e-05, "loss": 0.0667, "step": 1010 }, { "epoch": 3.9382239382239383, "grad_norm": 0.6338692307472229, "learning_rate": 9.926538309841839e-05, "loss": 0.065, "step": 1020 }, { "epoch": 3.976833976833977, "grad_norm": 0.7858821749687195, "learning_rate": 9.923687435954334e-05, "loss": 0.0613, "step": 1030 }, { "epoch": 4.015444015444015, "grad_norm": 0.9242402911186218, "learning_rate": 9.920782717409873e-05, "loss": 0.0681, "step": 1040 }, { "epoch": 4.054054054054054, "grad_norm": 0.5482421517372131, "learning_rate": 9.917824185973994e-05, "loss": 0.0643, "step": 1050 }, { "epoch": 4.0926640926640925, "grad_norm": 0.5903205275535583, "learning_rate": 9.914811874000723e-05, "loss": 0.0659, "step": 1060 }, { "epoch": 4.1312741312741315, "grad_norm": 0.6177347302436829, "learning_rate": 9.911745814432218e-05, "loss": 0.0697, "step": 1070 }, { "epoch": 4.1698841698841695, "grad_norm": 0.5361984968185425, "learning_rate": 9.90862604079842e-05, "loss": 0.0552, "step": 1080 }, { "epoch": 4.2084942084942085, "grad_norm": 0.49505937099456787, "learning_rate": 9.90545258721667e-05, "loss": 0.0557, "step": 1090 }, { "epoch": 4.2471042471042475, "grad_norm": 0.36627769470214844, "learning_rate": 9.90222548839135e-05, "loss": 0.0554, "step": 1100 }, { "epoch": 4.285714285714286, "grad_norm": 0.3721001446247101, "learning_rate": 9.898944779613495e-05, "loss": 0.0626, "step": 1110 }, { "epoch": 4.324324324324325, "grad_norm": 0.5092461109161377, "learning_rate": 9.89561049676041e-05, "loss": 0.0608, "step": 1120 }, { "epoch": 4.362934362934363, "grad_norm": 0.6615317463874817, "learning_rate": 9.89222267629528e-05, "loss": 0.0722, "step": 1130 }, { "epoch": 4.401544401544402, "grad_norm": 0.6472844481468201, "learning_rate": 9.888781355266763e-05, "loss": 0.0569, "step": 1140 }, { "epoch": 4.440154440154441, "grad_norm": 0.4949578046798706, "learning_rate": 9.885286571308598e-05, "loss": 0.0668, "step": 1150 }, { "epoch": 4.478764478764479, "grad_norm": 0.542823314666748, "learning_rate": 9.881738362639182e-05, "loss": 0.0624, "step": 1160 }, { "epoch": 4.517374517374518, "grad_norm": 0.5955585241317749, "learning_rate": 9.878136768061154e-05, "loss": 0.0626, "step": 1170 }, { "epoch": 4.555984555984556, "grad_norm": 0.5950126051902771, "learning_rate": 9.874481826960979e-05, "loss": 0.0564, "step": 1180 }, { "epoch": 4.594594594594595, "grad_norm": 0.8586946129798889, "learning_rate": 9.870773579308503e-05, "loss": 0.0636, "step": 1190 }, { "epoch": 4.633204633204633, "grad_norm": 0.653706431388855, "learning_rate": 9.867012065656533e-05, "loss": 0.0715, "step": 1200 }, { "epoch": 4.671814671814672, "grad_norm": 0.7195799946784973, "learning_rate": 9.863197327140376e-05, "loss": 0.0615, "step": 1210 }, { "epoch": 4.710424710424711, "grad_norm": 0.6180616617202759, "learning_rate": 9.859329405477403e-05, "loss": 0.0632, "step": 1220 }, { "epoch": 4.749034749034749, "grad_norm": 0.5667834281921387, "learning_rate": 9.855408342966585e-05, "loss": 0.0647, "step": 1230 }, { "epoch": 4.787644787644788, "grad_norm": 0.5417667031288147, "learning_rate": 9.851434182488033e-05, "loss": 0.0616, "step": 1240 }, { "epoch": 4.826254826254826, "grad_norm": 0.41703125834465027, "learning_rate": 9.84740696750253e-05, "loss": 0.0601, "step": 1250 }, { "epoch": 4.864864864864865, "grad_norm": 0.5548096895217896, "learning_rate": 9.843326742051055e-05, "loss": 0.0579, "step": 1260 }, { "epoch": 4.903474903474903, "grad_norm": 0.4253508448600769, "learning_rate": 9.839193550754297e-05, "loss": 0.0542, "step": 1270 }, { "epoch": 4.942084942084942, "grad_norm": 0.6980704665184021, "learning_rate": 9.835007438812177e-05, "loss": 0.0606, "step": 1280 }, { "epoch": 4.980694980694981, "grad_norm": 0.49001839756965637, "learning_rate": 9.830768452003341e-05, "loss": 0.0603, "step": 1290 }, { "epoch": 5.019305019305019, "grad_norm": 0.7339425683021545, "learning_rate": 9.826476636684671e-05, "loss": 0.0688, "step": 1300 }, { "epoch": 5.057915057915058, "grad_norm": 0.5982056856155396, "learning_rate": 9.822132039790773e-05, "loss": 0.0606, "step": 1310 }, { "epoch": 5.096525096525096, "grad_norm": 0.9272540211677551, "learning_rate": 9.817734708833461e-05, "loss": 0.0656, "step": 1320 }, { "epoch": 5.135135135135135, "grad_norm": 0.7408027052879333, "learning_rate": 9.813284691901243e-05, "loss": 0.0522, "step": 1330 }, { "epoch": 5.173745173745174, "grad_norm": 0.5034894943237305, "learning_rate": 9.808782037658792e-05, "loss": 0.06, "step": 1340 }, { "epoch": 5.212355212355212, "grad_norm": 0.5284120440483093, "learning_rate": 9.804226795346411e-05, "loss": 0.0563, "step": 1350 }, { "epoch": 5.250965250965251, "grad_norm": 0.5650848746299744, "learning_rate": 9.799619014779503e-05, "loss": 0.0564, "step": 1360 }, { "epoch": 5.289575289575289, "grad_norm": 0.5692643523216248, "learning_rate": 9.794958746348013e-05, "loss": 0.0587, "step": 1370 }, { "epoch": 5.328185328185328, "grad_norm": 0.5468897819519043, "learning_rate": 9.790246041015896e-05, "loss": 0.0605, "step": 1380 }, { "epoch": 5.366795366795367, "grad_norm": 0.6080969572067261, "learning_rate": 9.785480950320538e-05, "loss": 0.0571, "step": 1390 }, { "epoch": 5.405405405405405, "grad_norm": 0.7144999504089355, "learning_rate": 9.78066352637221e-05, "loss": 0.0616, "step": 1400 }, { "epoch": 5.444015444015444, "grad_norm": 0.5601004958152771, "learning_rate": 9.775793821853488e-05, "loss": 0.0479, "step": 1410 }, { "epoch": 5.482625482625482, "grad_norm": 0.4471278488636017, "learning_rate": 9.77087189001868e-05, "loss": 0.0612, "step": 1420 }, { "epoch": 5.521235521235521, "grad_norm": 0.4172019958496094, "learning_rate": 9.765897784693243e-05, "loss": 0.0513, "step": 1430 }, { "epoch": 5.559845559845559, "grad_norm": 0.43264463543891907, "learning_rate": 9.760871560273197e-05, "loss": 0.0567, "step": 1440 }, { "epoch": 5.598455598455598, "grad_norm": 0.4638456702232361, "learning_rate": 9.755793271724526e-05, "loss": 0.0548, "step": 1450 }, { "epoch": 5.637065637065637, "grad_norm": 0.46340683102607727, "learning_rate": 9.750662974582584e-05, "loss": 0.0537, "step": 1460 }, { "epoch": 5.675675675675675, "grad_norm": 0.5492231249809265, "learning_rate": 9.745480724951473e-05, "loss": 0.0578, "step": 1470 }, { "epoch": 5.714285714285714, "grad_norm": 0.5391159057617188, "learning_rate": 9.740246579503447e-05, "loss": 0.0601, "step": 1480 }, { "epoch": 5.752895752895753, "grad_norm": 0.46856486797332764, "learning_rate": 9.734960595478284e-05, "loss": 0.0548, "step": 1490 }, { "epoch": 5.7915057915057915, "grad_norm": 0.42841145396232605, "learning_rate": 9.729622830682657e-05, "loss": 0.0557, "step": 1500 }, { "epoch": 5.8301158301158305, "grad_norm": 0.7276766300201416, "learning_rate": 9.724233343489504e-05, "loss": 0.0592, "step": 1510 }, { "epoch": 5.8687258687258685, "grad_norm": 0.47596219182014465, "learning_rate": 9.718792192837396e-05, "loss": 0.0578, "step": 1520 }, { "epoch": 5.9073359073359075, "grad_norm": 0.6250131130218506, "learning_rate": 9.713299438229886e-05, "loss": 0.0583, "step": 1530 }, { "epoch": 5.945945945945946, "grad_norm": 0.5623660087585449, "learning_rate": 9.707755139734855e-05, "loss": 0.0542, "step": 1540 }, { "epoch": 5.984555984555985, "grad_norm": 0.5723385810852051, "learning_rate": 9.702159357983866e-05, "loss": 0.0635, "step": 1550 }, { "epoch": 6.023166023166024, "grad_norm": 0.6064860224723816, "learning_rate": 9.696512154171492e-05, "loss": 0.0542, "step": 1560 }, { "epoch": 6.061776061776062, "grad_norm": 0.480323851108551, "learning_rate": 9.690813590054645e-05, "loss": 0.0461, "step": 1570 }, { "epoch": 6.100386100386101, "grad_norm": 0.3705624043941498, "learning_rate": 9.685063727951914e-05, "loss": 0.0489, "step": 1580 }, { "epoch": 6.138996138996139, "grad_norm": 0.6532600522041321, "learning_rate": 9.679262630742865e-05, "loss": 0.0574, "step": 1590 }, { "epoch": 6.177606177606178, "grad_norm": 0.38052064180374146, "learning_rate": 9.673410361867373e-05, "loss": 0.0511, "step": 1600 }, { "epoch": 6.216216216216216, "grad_norm": 0.4939205050468445, "learning_rate": 9.667506985324909e-05, "loss": 0.0529, "step": 1610 }, { "epoch": 6.254826254826255, "grad_norm": 0.37428799271583557, "learning_rate": 9.661552565673855e-05, "loss": 0.053, "step": 1620 }, { "epoch": 6.293436293436294, "grad_norm": 0.30815252661705017, "learning_rate": 9.655547168030789e-05, "loss": 0.0492, "step": 1630 }, { "epoch": 6.332046332046332, "grad_norm": 0.45762449502944946, "learning_rate": 9.649490858069777e-05, "loss": 0.0581, "step": 1640 }, { "epoch": 6.370656370656371, "grad_norm": 0.4566146433353424, "learning_rate": 9.643383702021658e-05, "loss": 0.0604, "step": 1650 }, { "epoch": 6.409266409266409, "grad_norm": 0.4395817816257477, "learning_rate": 9.637225766673307e-05, "loss": 0.0501, "step": 1660 }, { "epoch": 6.447876447876448, "grad_norm": 0.4267268180847168, "learning_rate": 9.631017119366922e-05, "loss": 0.0563, "step": 1670 }, { "epoch": 6.486486486486487, "grad_norm": 0.5258854627609253, "learning_rate": 9.624757827999273e-05, "loss": 0.0493, "step": 1680 }, { "epoch": 6.525096525096525, "grad_norm": 0.4203065037727356, "learning_rate": 9.618447961020971e-05, "loss": 0.0566, "step": 1690 }, { "epoch": 6.563706563706564, "grad_norm": 0.6417315006256104, "learning_rate": 9.612087587435707e-05, "loss": 0.0463, "step": 1700 }, { "epoch": 6.602316602316602, "grad_norm": 0.560469925403595, "learning_rate": 9.605676776799508e-05, "loss": 0.051, "step": 1710 }, { "epoch": 6.640926640926641, "grad_norm": 0.5156351923942566, "learning_rate": 9.599215599219973e-05, "loss": 0.0577, "step": 1720 }, { "epoch": 6.67953667953668, "grad_norm": 0.35196298360824585, "learning_rate": 9.592704125355505e-05, "loss": 0.055, "step": 1730 }, { "epoch": 6.718146718146718, "grad_norm": 0.5210710763931274, "learning_rate": 9.586142426414538e-05, "loss": 0.0506, "step": 1740 }, { "epoch": 6.756756756756757, "grad_norm": 0.47786369919776917, "learning_rate": 9.57953057415476e-05, "loss": 0.0442, "step": 1750 }, { "epoch": 6.795366795366795, "grad_norm": 0.463230162858963, "learning_rate": 9.572868640882328e-05, "loss": 0.0523, "step": 1760 }, { "epoch": 6.833976833976834, "grad_norm": 0.5917426943778992, "learning_rate": 9.56615669945108e-05, "loss": 0.0554, "step": 1770 }, { "epoch": 6.872586872586872, "grad_norm": 0.43890470266342163, "learning_rate": 9.55939482326173e-05, "loss": 0.0517, "step": 1780 }, { "epoch": 6.911196911196911, "grad_norm": 0.4829425811767578, "learning_rate": 9.552583086261069e-05, "loss": 0.0494, "step": 1790 }, { "epoch": 6.94980694980695, "grad_norm": 0.4498508870601654, "learning_rate": 9.545721562941168e-05, "loss": 0.0453, "step": 1800 }, { "epoch": 6.988416988416988, "grad_norm": 0.426923006772995, "learning_rate": 9.538810328338543e-05, "loss": 0.0565, "step": 1810 }, { "epoch": 7.027027027027027, "grad_norm": 0.46216297149658203, "learning_rate": 9.531849458033349e-05, "loss": 0.0516, "step": 1820 }, { "epoch": 7.065637065637065, "grad_norm": 0.41856151819229126, "learning_rate": 9.524839028148547e-05, "loss": 0.0554, "step": 1830 }, { "epoch": 7.104247104247104, "grad_norm": 0.2807966470718384, "learning_rate": 9.517779115349077e-05, "loss": 0.0463, "step": 1840 }, { "epoch": 7.142857142857143, "grad_norm": 0.5471675992012024, "learning_rate": 9.510669796841014e-05, "loss": 0.0532, "step": 1850 }, { "epoch": 7.181467181467181, "grad_norm": 0.4649882912635803, "learning_rate": 9.503511150370727e-05, "loss": 0.0504, "step": 1860 }, { "epoch": 7.22007722007722, "grad_norm": 0.6518912315368652, "learning_rate": 9.496303254224024e-05, "loss": 0.0604, "step": 1870 }, { "epoch": 7.258687258687258, "grad_norm": 0.6045863032341003, "learning_rate": 9.489046187225306e-05, "loss": 0.0564, "step": 1880 }, { "epoch": 7.297297297297297, "grad_norm": 0.6644754409790039, "learning_rate": 9.481740028736692e-05, "loss": 0.0446, "step": 1890 }, { "epoch": 7.335907335907336, "grad_norm": 0.562257707118988, "learning_rate": 9.474384858657164e-05, "loss": 0.0574, "step": 1900 }, { "epoch": 7.374517374517374, "grad_norm": 0.5704360604286194, "learning_rate": 9.466980757421679e-05, "loss": 0.0516, "step": 1910 }, { "epoch": 7.413127413127413, "grad_norm": 0.7471466660499573, "learning_rate": 9.459527806000305e-05, "loss": 0.0462, "step": 1920 }, { "epoch": 7.4517374517374515, "grad_norm": 0.4494033455848694, "learning_rate": 9.452026085897325e-05, "loss": 0.0501, "step": 1930 }, { "epoch": 7.4903474903474905, "grad_norm": 0.6094410419464111, "learning_rate": 9.444475679150348e-05, "loss": 0.051, "step": 1940 }, { "epoch": 7.528957528957529, "grad_norm": 0.45507022738456726, "learning_rate": 9.436876668329411e-05, "loss": 0.0427, "step": 1950 }, { "epoch": 7.5675675675675675, "grad_norm": 0.2931894063949585, "learning_rate": 9.429229136536079e-05, "loss": 0.045, "step": 1960 }, { "epoch": 7.6061776061776065, "grad_norm": 0.4837070405483246, "learning_rate": 9.421533167402534e-05, "loss": 0.0509, "step": 1970 }, { "epoch": 7.644787644787645, "grad_norm": 0.47066646814346313, "learning_rate": 9.413788845090666e-05, "loss": 0.0538, "step": 1980 }, { "epoch": 7.683397683397684, "grad_norm": 0.36199939250946045, "learning_rate": 9.405996254291136e-05, "loss": 0.0549, "step": 1990 }, { "epoch": 7.722007722007722, "grad_norm": 0.38093823194503784, "learning_rate": 9.398155480222474e-05, "loss": 0.0431, "step": 2000 }, { "epoch": 7.760617760617761, "grad_norm": 0.33619001507759094, "learning_rate": 9.390266608630128e-05, "loss": 0.0452, "step": 2010 }, { "epoch": 7.799227799227799, "grad_norm": 0.5024441480636597, "learning_rate": 9.38232972578553e-05, "loss": 0.0459, "step": 2020 }, { "epoch": 7.837837837837838, "grad_norm": 0.3622889518737793, "learning_rate": 9.374344918485164e-05, "loss": 0.0468, "step": 2030 }, { "epoch": 7.876447876447877, "grad_norm": 0.38623934984207153, "learning_rate": 9.366312274049602e-05, "loss": 0.0511, "step": 2040 }, { "epoch": 7.915057915057915, "grad_norm": 0.6467400789260864, "learning_rate": 9.358231880322554e-05, "loss": 0.0433, "step": 2050 }, { "epoch": 7.953667953667954, "grad_norm": 0.3720637559890747, "learning_rate": 9.350103825669916e-05, "loss": 0.0437, "step": 2060 }, { "epoch": 7.992277992277993, "grad_norm": 0.44834697246551514, "learning_rate": 9.341928198978787e-05, "loss": 0.042, "step": 2070 }, { "epoch": 8.03088803088803, "grad_norm": 0.3857465982437134, "learning_rate": 9.333705089656512e-05, "loss": 0.0442, "step": 2080 }, { "epoch": 8.069498069498069, "grad_norm": 0.5917789936065674, "learning_rate": 9.325434587629698e-05, "loss": 0.0407, "step": 2090 }, { "epoch": 8.108108108108109, "grad_norm": 0.49043840169906616, "learning_rate": 9.31711678334323e-05, "loss": 0.0465, "step": 2100 }, { "epoch": 8.146718146718147, "grad_norm": 0.33593717217445374, "learning_rate": 9.308751767759282e-05, "loss": 0.0448, "step": 2110 }, { "epoch": 8.185328185328185, "grad_norm": 0.37338685989379883, "learning_rate": 9.300339632356325e-05, "loss": 0.0386, "step": 2120 }, { "epoch": 8.223938223938223, "grad_norm": 0.5644637942314148, "learning_rate": 9.291880469128124e-05, "loss": 0.0426, "step": 2130 }, { "epoch": 8.262548262548263, "grad_norm": 0.6692359447479248, "learning_rate": 9.283374370582732e-05, "loss": 0.0431, "step": 2140 }, { "epoch": 8.301158301158301, "grad_norm": 0.4001326262950897, "learning_rate": 9.274821429741482e-05, "loss": 0.0573, "step": 2150 }, { "epoch": 8.339768339768339, "grad_norm": 0.39123591780662537, "learning_rate": 9.266221740137961e-05, "loss": 0.0444, "step": 2160 }, { "epoch": 8.378378378378379, "grad_norm": 0.5335126519203186, "learning_rate": 9.257575395817001e-05, "loss": 0.057, "step": 2170 }, { "epoch": 8.416988416988417, "grad_norm": 0.5357592105865479, "learning_rate": 9.248882491333637e-05, "loss": 0.0451, "step": 2180 }, { "epoch": 8.455598455598455, "grad_norm": 0.43949824571609497, "learning_rate": 9.240143121752076e-05, "loss": 0.055, "step": 2190 }, { "epoch": 8.494208494208495, "grad_norm": 0.7482830286026001, "learning_rate": 9.23135738264467e-05, "loss": 0.0404, "step": 2200 }, { "epoch": 8.532818532818533, "grad_norm": 0.5364919304847717, "learning_rate": 9.222525370090849e-05, "loss": 0.0492, "step": 2210 }, { "epoch": 8.571428571428571, "grad_norm": 0.3352760076522827, "learning_rate": 9.213647180676088e-05, "loss": 0.056, "step": 2220 }, { "epoch": 8.61003861003861, "grad_norm": 0.537573516368866, "learning_rate": 9.204722911490846e-05, "loss": 0.0584, "step": 2230 }, { "epoch": 8.64864864864865, "grad_norm": 0.7373439073562622, "learning_rate": 9.1957526601295e-05, "loss": 0.0537, "step": 2240 }, { "epoch": 8.687258687258687, "grad_norm": 0.4323110282421112, "learning_rate": 9.186736524689281e-05, "loss": 0.054, "step": 2250 }, { "epoch": 8.725868725868725, "grad_norm": 0.42572522163391113, "learning_rate": 9.177674603769204e-05, "loss": 0.0581, "step": 2260 }, { "epoch": 8.764478764478765, "grad_norm": 0.539223849773407, "learning_rate": 9.168566996468983e-05, "loss": 0.0523, "step": 2270 }, { "epoch": 8.803088803088803, "grad_norm": 0.596268892288208, "learning_rate": 9.159413802387951e-05, "loss": 0.0493, "step": 2280 }, { "epoch": 8.841698841698841, "grad_norm": 0.30029815435409546, "learning_rate": 9.150215121623974e-05, "loss": 0.0435, "step": 2290 }, { "epoch": 8.880308880308881, "grad_norm": 0.5296053886413574, "learning_rate": 9.140971054772349e-05, "loss": 0.042, "step": 2300 }, { "epoch": 8.91891891891892, "grad_norm": 0.30834901332855225, "learning_rate": 9.131681702924713e-05, "loss": 0.049, "step": 2310 }, { "epoch": 8.957528957528957, "grad_norm": 0.4735316038131714, "learning_rate": 9.122347167667926e-05, "loss": 0.0459, "step": 2320 }, { "epoch": 8.996138996138995, "grad_norm": 0.48870229721069336, "learning_rate": 9.112967551082973e-05, "loss": 0.0463, "step": 2330 }, { "epoch": 9.034749034749035, "grad_norm": 0.39390528202056885, "learning_rate": 9.103542955743835e-05, "loss": 0.0515, "step": 2340 }, { "epoch": 9.073359073359073, "grad_norm": 0.561843752861023, "learning_rate": 9.094073484716381e-05, "loss": 0.048, "step": 2350 }, { "epoch": 9.111969111969112, "grad_norm": 0.3307531476020813, "learning_rate": 9.084559241557226e-05, "loss": 0.0418, "step": 2360 }, { "epoch": 9.150579150579151, "grad_norm": 0.4625450074672699, "learning_rate": 9.075000330312608e-05, "loss": 0.0437, "step": 2370 }, { "epoch": 9.18918918918919, "grad_norm": 0.41069215536117554, "learning_rate": 9.065396855517253e-05, "loss": 0.0392, "step": 2380 }, { "epoch": 9.227799227799228, "grad_norm": 0.3695158362388611, "learning_rate": 9.055748922193219e-05, "loss": 0.0412, "step": 2390 }, { "epoch": 9.266409266409266, "grad_norm": 0.49469563364982605, "learning_rate": 9.046056635848761e-05, "loss": 0.0487, "step": 2400 }, { "epoch": 9.305019305019306, "grad_norm": 0.5143020749092102, "learning_rate": 9.036320102477169e-05, "loss": 0.0432, "step": 2410 }, { "epoch": 9.343629343629344, "grad_norm": 0.6467536091804504, "learning_rate": 9.02653942855561e-05, "loss": 0.0404, "step": 2420 }, { "epoch": 9.382239382239382, "grad_norm": 0.3653818964958191, "learning_rate": 9.016714721043971e-05, "loss": 0.0454, "step": 2430 }, { "epoch": 9.420849420849422, "grad_norm": 0.39167219400405884, "learning_rate": 9.006846087383675e-05, "loss": 0.0439, "step": 2440 }, { "epoch": 9.45945945945946, "grad_norm": 0.40652039647102356, "learning_rate": 8.996933635496523e-05, "loss": 0.0395, "step": 2450 }, { "epoch": 9.498069498069498, "grad_norm": 0.5425982475280762, "learning_rate": 8.986977473783498e-05, "loss": 0.0408, "step": 2460 }, { "epoch": 9.536679536679536, "grad_norm": 0.5995064973831177, "learning_rate": 8.97697771112359e-05, "loss": 0.0418, "step": 2470 }, { "epoch": 9.575289575289576, "grad_norm": 0.5640287399291992, "learning_rate": 8.966934456872602e-05, "loss": 0.0388, "step": 2480 }, { "epoch": 9.613899613899614, "grad_norm": 0.416124165058136, "learning_rate": 8.95684782086195e-05, "loss": 0.0432, "step": 2490 }, { "epoch": 9.652509652509652, "grad_norm": 0.3699228763580322, "learning_rate": 8.946717913397476e-05, "loss": 0.0385, "step": 2500 }, { "epoch": 9.691119691119692, "grad_norm": 0.33271324634552, "learning_rate": 8.93654484525822e-05, "loss": 0.0422, "step": 2510 }, { "epoch": 9.72972972972973, "grad_norm": 0.5058172345161438, "learning_rate": 8.926328727695226e-05, "loss": 0.0422, "step": 2520 }, { "epoch": 9.768339768339768, "grad_norm": 0.5327197313308716, "learning_rate": 8.916069672430319e-05, "loss": 0.0321, "step": 2530 }, { "epoch": 9.806949806949808, "grad_norm": 0.37798357009887695, "learning_rate": 8.905767791654884e-05, "loss": 0.0444, "step": 2540 }, { "epoch": 9.845559845559846, "grad_norm": 0.4148670732975006, "learning_rate": 8.895423198028638e-05, "loss": 0.0406, "step": 2550 }, { "epoch": 9.884169884169884, "grad_norm": 0.41860342025756836, "learning_rate": 8.885036004678402e-05, "loss": 0.0467, "step": 2560 }, { "epoch": 9.922779922779922, "grad_norm": 0.3116677403450012, "learning_rate": 8.874606325196857e-05, "loss": 0.0411, "step": 2570 }, { "epoch": 9.961389961389962, "grad_norm": 0.4406808018684387, "learning_rate": 8.864134273641304e-05, "loss": 0.0444, "step": 2580 }, { "epoch": 10.0, "grad_norm": 0.3659777045249939, "learning_rate": 8.853619964532427e-05, "loss": 0.0347, "step": 2590 }, { "epoch": 10.038610038610038, "grad_norm": 0.22327017784118652, "learning_rate": 8.843063512853019e-05, "loss": 0.0396, "step": 2600 }, { "epoch": 10.077220077220078, "grad_norm": 0.38703998923301697, "learning_rate": 8.832465034046749e-05, "loss": 0.0442, "step": 2610 }, { "epoch": 10.115830115830116, "grad_norm": 0.3925214409828186, "learning_rate": 8.821824644016882e-05, "loss": 0.0395, "step": 2620 }, { "epoch": 10.154440154440154, "grad_norm": 0.38523709774017334, "learning_rate": 8.811142459125019e-05, "loss": 0.0384, "step": 2630 }, { "epoch": 10.193050193050192, "grad_norm": 0.3844083845615387, "learning_rate": 8.800418596189822e-05, "loss": 0.045, "step": 2640 }, { "epoch": 10.231660231660232, "grad_norm": 0.42650488018989563, "learning_rate": 8.789653172485737e-05, "loss": 0.034, "step": 2650 }, { "epoch": 10.27027027027027, "grad_norm": 0.3765309751033783, "learning_rate": 8.778846305741715e-05, "loss": 0.0445, "step": 2660 }, { "epoch": 10.308880308880308, "grad_norm": 0.42378807067871094, "learning_rate": 8.767998114139918e-05, "loss": 0.0421, "step": 2670 }, { "epoch": 10.347490347490348, "grad_norm": 0.39118635654449463, "learning_rate": 8.757108716314429e-05, "loss": 0.0402, "step": 2680 }, { "epoch": 10.386100386100386, "grad_norm": 0.4010487198829651, "learning_rate": 8.746178231349962e-05, "loss": 0.0456, "step": 2690 }, { "epoch": 10.424710424710424, "grad_norm": 0.40013548731803894, "learning_rate": 8.735206778780549e-05, "loss": 0.0438, "step": 2700 }, { "epoch": 10.463320463320464, "grad_norm": 0.39128270745277405, "learning_rate": 8.724194478588234e-05, "loss": 0.0405, "step": 2710 }, { "epoch": 10.501930501930502, "grad_norm": 0.5453833937644958, "learning_rate": 8.713141451201772e-05, "loss": 0.0435, "step": 2720 }, { "epoch": 10.54054054054054, "grad_norm": 0.6112650632858276, "learning_rate": 8.702047817495295e-05, "loss": 0.046, "step": 2730 }, { "epoch": 10.579150579150578, "grad_norm": 0.4512874484062195, "learning_rate": 8.69091369878701e-05, "loss": 0.0389, "step": 2740 }, { "epoch": 10.617760617760618, "grad_norm": 0.3846628963947296, "learning_rate": 8.679739216837849e-05, "loss": 0.0387, "step": 2750 }, { "epoch": 10.656370656370656, "grad_norm": 0.2874908447265625, "learning_rate": 8.66852449385016e-05, "loss": 0.0413, "step": 2760 }, { "epoch": 10.694980694980694, "grad_norm": 0.4614006280899048, "learning_rate": 8.657269652466356e-05, "loss": 0.0397, "step": 2770 }, { "epoch": 10.733590733590734, "grad_norm": 0.39504802227020264, "learning_rate": 8.645974815767577e-05, "loss": 0.0455, "step": 2780 }, { "epoch": 10.772200772200772, "grad_norm": 0.3581386208534241, "learning_rate": 8.634640107272351e-05, "loss": 0.0407, "step": 2790 }, { "epoch": 10.81081081081081, "grad_norm": 0.3735191226005554, "learning_rate": 8.623265650935234e-05, "loss": 0.0369, "step": 2800 }, { "epoch": 10.849420849420849, "grad_norm": 0.48280179500579834, "learning_rate": 8.611851571145456e-05, "loss": 0.0395, "step": 2810 }, { "epoch": 10.888030888030888, "grad_norm": 0.242467001080513, "learning_rate": 8.600397992725566e-05, "loss": 0.042, "step": 2820 }, { "epoch": 10.926640926640927, "grad_norm": 0.5253610014915466, "learning_rate": 8.588905040930061e-05, "loss": 0.0382, "step": 2830 }, { "epoch": 10.965250965250965, "grad_norm": 0.40605440735816956, "learning_rate": 8.577372841444022e-05, "loss": 0.0355, "step": 2840 }, { "epoch": 11.003861003861005, "grad_norm": 0.33260899782180786, "learning_rate": 8.565801520381736e-05, "loss": 0.0414, "step": 2850 }, { "epoch": 11.042471042471043, "grad_norm": 0.3277893364429474, "learning_rate": 8.554191204285313e-05, "loss": 0.0406, "step": 2860 }, { "epoch": 11.08108108108108, "grad_norm": 0.4931524693965912, "learning_rate": 8.542542020123315e-05, "loss": 0.042, "step": 2870 }, { "epoch": 11.11969111969112, "grad_norm": 0.36509305238723755, "learning_rate": 8.530854095289347e-05, "loss": 0.0384, "step": 2880 }, { "epoch": 11.158301158301159, "grad_norm": 0.2553371489048004, "learning_rate": 8.519127557600688e-05, "loss": 0.0338, "step": 2890 }, { "epoch": 11.196911196911197, "grad_norm": 0.42066043615341187, "learning_rate": 8.507362535296871e-05, "loss": 0.0415, "step": 2900 }, { "epoch": 11.235521235521235, "grad_norm": 0.46299177408218384, "learning_rate": 8.495559157038299e-05, "loss": 0.0467, "step": 2910 }, { "epoch": 11.274131274131275, "grad_norm": 0.36362677812576294, "learning_rate": 8.483717551904823e-05, "loss": 0.0505, "step": 2920 }, { "epoch": 11.312741312741313, "grad_norm": 0.33156195282936096, "learning_rate": 8.47183784939434e-05, "loss": 0.0378, "step": 2930 }, { "epoch": 11.35135135135135, "grad_norm": 0.5630045533180237, "learning_rate": 8.459920179421374e-05, "loss": 0.0389, "step": 2940 }, { "epoch": 11.38996138996139, "grad_norm": 0.3690023422241211, "learning_rate": 8.447964672315656e-05, "loss": 0.0436, "step": 2950 }, { "epoch": 11.428571428571429, "grad_norm": 0.5662466883659363, "learning_rate": 8.435971458820692e-05, "loss": 0.0418, "step": 2960 }, { "epoch": 11.467181467181467, "grad_norm": 0.3692759871482849, "learning_rate": 8.423940670092345e-05, "loss": 0.0411, "step": 2970 }, { "epoch": 11.505791505791505, "grad_norm": 0.5392592549324036, "learning_rate": 8.411872437697394e-05, "loss": 0.0406, "step": 2980 }, { "epoch": 11.544401544401545, "grad_norm": 0.35389575362205505, "learning_rate": 8.399766893612096e-05, "loss": 0.0365, "step": 2990 }, { "epoch": 11.583011583011583, "grad_norm": 0.37920311093330383, "learning_rate": 8.38762417022074e-05, "loss": 0.035, "step": 3000 }, { "epoch": 11.621621621621621, "grad_norm": 0.38484132289886475, "learning_rate": 8.375444400314204e-05, "loss": 0.0397, "step": 3010 }, { "epoch": 11.660231660231661, "grad_norm": 0.44138237833976746, "learning_rate": 8.3632277170885e-05, "loss": 0.0381, "step": 3020 }, { "epoch": 11.698841698841699, "grad_norm": 0.4417366087436676, "learning_rate": 8.350974254143318e-05, "loss": 0.0401, "step": 3030 }, { "epoch": 11.737451737451737, "grad_norm": 0.3455044627189636, "learning_rate": 8.338684145480566e-05, "loss": 0.038, "step": 3040 }, { "epoch": 11.776061776061777, "grad_norm": 0.4324539005756378, "learning_rate": 8.326357525502904e-05, "loss": 0.0377, "step": 3050 }, { "epoch": 11.814671814671815, "grad_norm": 0.24957092106342316, "learning_rate": 8.313994529012273e-05, "loss": 0.0385, "step": 3060 }, { "epoch": 11.853281853281853, "grad_norm": 0.5175673365592957, "learning_rate": 8.301595291208422e-05, "loss": 0.0391, "step": 3070 }, { "epoch": 11.891891891891891, "grad_norm": 0.42861613631248474, "learning_rate": 8.289159947687427e-05, "loss": 0.0391, "step": 3080 }, { "epoch": 11.930501930501931, "grad_norm": 0.2719945013523102, "learning_rate": 8.276688634440216e-05, "loss": 0.0327, "step": 3090 }, { "epoch": 11.96911196911197, "grad_norm": 0.3701878488063812, "learning_rate": 8.26418148785107e-05, "loss": 0.0366, "step": 3100 }, { "epoch": 12.007722007722007, "grad_norm": 0.23779451847076416, "learning_rate": 8.251638644696141e-05, "loss": 0.0401, "step": 3110 }, { "epoch": 12.046332046332047, "grad_norm": 0.42985931038856506, "learning_rate": 8.23906024214195e-05, "loss": 0.0458, "step": 3120 }, { "epoch": 12.084942084942085, "grad_norm": 0.4434528946876526, "learning_rate": 8.226446417743897e-05, "loss": 0.036, "step": 3130 }, { "epoch": 12.123552123552123, "grad_norm": 0.43846243619918823, "learning_rate": 8.213797309444742e-05, "loss": 0.0389, "step": 3140 }, { "epoch": 12.162162162162161, "grad_norm": 0.4050768315792084, "learning_rate": 8.201113055573105e-05, "loss": 0.039, "step": 3150 }, { "epoch": 12.200772200772201, "grad_norm": 0.44880127906799316, "learning_rate": 8.188393794841958e-05, "loss": 0.0379, "step": 3160 }, { "epoch": 12.23938223938224, "grad_norm": 0.38949722051620483, "learning_rate": 8.175639666347094e-05, "loss": 0.0381, "step": 3170 }, { "epoch": 12.277992277992277, "grad_norm": 0.3987986147403717, "learning_rate": 8.162850809565623e-05, "loss": 0.0403, "step": 3180 }, { "epoch": 12.316602316602317, "grad_norm": 0.38119158148765564, "learning_rate": 8.150027364354431e-05, "loss": 0.0358, "step": 3190 }, { "epoch": 12.355212355212355, "grad_norm": 0.2881115972995758, "learning_rate": 8.137169470948662e-05, "loss": 0.0433, "step": 3200 }, { "epoch": 12.393822393822393, "grad_norm": 0.5546713471412659, "learning_rate": 8.124277269960179e-05, "loss": 0.0411, "step": 3210 }, { "epoch": 12.432432432432432, "grad_norm": 0.4417155683040619, "learning_rate": 8.111350902376023e-05, "loss": 0.0394, "step": 3220 }, { "epoch": 12.471042471042471, "grad_norm": 0.5674039721488953, "learning_rate": 8.098390509556883e-05, "loss": 0.0375, "step": 3230 }, { "epoch": 12.50965250965251, "grad_norm": 0.4557495713233948, "learning_rate": 8.085396233235536e-05, "loss": 0.0384, "step": 3240 }, { "epoch": 12.548262548262548, "grad_norm": 0.40320858359336853, "learning_rate": 8.072368215515306e-05, "loss": 0.031, "step": 3250 }, { "epoch": 12.586872586872587, "grad_norm": 0.43268531560897827, "learning_rate": 8.059306598868506e-05, "loss": 0.0372, "step": 3260 }, { "epoch": 12.625482625482626, "grad_norm": 0.28157323598861694, "learning_rate": 8.046211526134888e-05, "loss": 0.0421, "step": 3270 }, { "epoch": 12.664092664092664, "grad_norm": 0.2906109392642975, "learning_rate": 8.033083140520065e-05, "loss": 0.0343, "step": 3280 }, { "epoch": 12.702702702702704, "grad_norm": 0.43920794129371643, "learning_rate": 8.019921585593962e-05, "loss": 0.0366, "step": 3290 }, { "epoch": 12.741312741312742, "grad_norm": 0.4903682768344879, "learning_rate": 8.006727005289232e-05, "loss": 0.0346, "step": 3300 }, { "epoch": 12.77992277992278, "grad_norm": 0.4271104037761688, "learning_rate": 7.993499543899692e-05, "loss": 0.0342, "step": 3310 }, { "epoch": 12.818532818532818, "grad_norm": 0.30340924859046936, "learning_rate": 7.980239346078742e-05, "loss": 0.0356, "step": 3320 }, { "epoch": 12.857142857142858, "grad_norm": 0.37377700209617615, "learning_rate": 7.966946556837778e-05, "loss": 0.0343, "step": 3330 }, { "epoch": 12.895752895752896, "grad_norm": 0.28563475608825684, "learning_rate": 7.953621321544616e-05, "loss": 0.0406, "step": 3340 }, { "epoch": 12.934362934362934, "grad_norm": 0.2693694233894348, "learning_rate": 7.940263785921896e-05, "loss": 0.0339, "step": 3350 }, { "epoch": 12.972972972972974, "grad_norm": 0.3787376582622528, "learning_rate": 7.926874096045482e-05, "loss": 0.0418, "step": 3360 }, { "epoch": 13.011583011583012, "grad_norm": 0.4588676393032074, "learning_rate": 7.913452398342881e-05, "loss": 0.0374, "step": 3370 }, { "epoch": 13.05019305019305, "grad_norm": 0.37496957182884216, "learning_rate": 7.89999883959163e-05, "loss": 0.0443, "step": 3380 }, { "epoch": 13.088803088803088, "grad_norm": 0.519055187702179, "learning_rate": 7.886513566917687e-05, "loss": 0.0397, "step": 3390 }, { "epoch": 13.127413127413128, "grad_norm": 0.3142504394054413, "learning_rate": 7.872996727793838e-05, "loss": 0.0352, "step": 3400 }, { "epoch": 13.166023166023166, "grad_norm": 0.32780155539512634, "learning_rate": 7.859448470038069e-05, "loss": 0.0352, "step": 3410 }, { "epoch": 13.204633204633204, "grad_norm": 0.3536975383758545, "learning_rate": 7.845868941811956e-05, "loss": 0.0403, "step": 3420 }, { "epoch": 13.243243243243244, "grad_norm": 0.33499446511268616, "learning_rate": 7.832258291619043e-05, "loss": 0.0323, "step": 3430 }, { "epoch": 13.281853281853282, "grad_norm": 0.38436001539230347, "learning_rate": 7.81861666830322e-05, "loss": 0.0305, "step": 3440 }, { "epoch": 13.32046332046332, "grad_norm": 0.4771931767463684, "learning_rate": 7.804944221047097e-05, "loss": 0.0405, "step": 3450 }, { "epoch": 13.35907335907336, "grad_norm": 0.3527798354625702, "learning_rate": 7.791241099370364e-05, "loss": 0.0357, "step": 3460 }, { "epoch": 13.397683397683398, "grad_norm": 0.32941561937332153, "learning_rate": 7.777507453128163e-05, "loss": 0.0399, "step": 3470 }, { "epoch": 13.436293436293436, "grad_norm": 0.3091055452823639, "learning_rate": 7.763743432509451e-05, "loss": 0.0369, "step": 3480 }, { "epoch": 13.474903474903474, "grad_norm": 0.31441813707351685, "learning_rate": 7.749949188035353e-05, "loss": 0.0384, "step": 3490 }, { "epoch": 13.513513513513514, "grad_norm": 0.5186731815338135, "learning_rate": 7.736124870557516e-05, "loss": 0.0453, "step": 3500 }, { "epoch": 13.552123552123552, "grad_norm": 0.3249494433403015, "learning_rate": 7.722270631256459e-05, "loss": 0.0347, "step": 3510 }, { "epoch": 13.59073359073359, "grad_norm": 0.2922497093677521, "learning_rate": 7.708386621639925e-05, "loss": 0.0409, "step": 3520 }, { "epoch": 13.62934362934363, "grad_norm": 0.31381967663764954, "learning_rate": 7.694472993541219e-05, "loss": 0.0352, "step": 3530 }, { "epoch": 13.667953667953668, "grad_norm": 0.504213809967041, "learning_rate": 7.680529899117547e-05, "loss": 0.0352, "step": 3540 }, { "epoch": 13.706563706563706, "grad_norm": 0.313487708568573, "learning_rate": 7.666557490848358e-05, "loss": 0.0397, "step": 3550 }, { "epoch": 13.745173745173744, "grad_norm": 0.5103364586830139, "learning_rate": 7.65255592153367e-05, "loss": 0.0394, "step": 3560 }, { "epoch": 13.783783783783784, "grad_norm": 0.4513188600540161, "learning_rate": 7.638525344292402e-05, "loss": 0.0386, "step": 3570 }, { "epoch": 13.822393822393822, "grad_norm": 0.3880644738674164, "learning_rate": 7.624465912560697e-05, "loss": 0.0359, "step": 3580 }, { "epoch": 13.86100386100386, "grad_norm": 0.4982539713382721, "learning_rate": 7.610377780090249e-05, "loss": 0.0412, "step": 3590 }, { "epoch": 13.8996138996139, "grad_norm": 0.28544214367866516, "learning_rate": 7.596261100946618e-05, "loss": 0.0272, "step": 3600 }, { "epoch": 13.938223938223938, "grad_norm": 0.28790661692619324, "learning_rate": 7.582116029507542e-05, "loss": 0.0329, "step": 3610 }, { "epoch": 13.976833976833976, "grad_norm": 0.46929270029067993, "learning_rate": 7.56794272046126e-05, "loss": 0.0301, "step": 3620 }, { "epoch": 14.015444015444016, "grad_norm": 0.39010658860206604, "learning_rate": 7.55374132880481e-05, "loss": 0.0384, "step": 3630 }, { "epoch": 14.054054054054054, "grad_norm": 0.31634047627449036, "learning_rate": 7.539512009842333e-05, "loss": 0.0351, "step": 3640 }, { "epoch": 14.092664092664092, "grad_norm": 0.27322113513946533, "learning_rate": 7.525254919183382e-05, "loss": 0.0354, "step": 3650 }, { "epoch": 14.13127413127413, "grad_norm": 0.3212028443813324, "learning_rate": 7.510970212741215e-05, "loss": 0.0355, "step": 3660 }, { "epoch": 14.16988416988417, "grad_norm": 0.37816670536994934, "learning_rate": 7.496658046731096e-05, "loss": 0.0303, "step": 3670 }, { "epoch": 14.208494208494209, "grad_norm": 0.3786605894565582, "learning_rate": 7.482318577668578e-05, "loss": 0.0386, "step": 3680 }, { "epoch": 14.247104247104247, "grad_norm": 0.36963769793510437, "learning_rate": 7.467951962367796e-05, "loss": 0.0423, "step": 3690 }, { "epoch": 14.285714285714286, "grad_norm": 0.4539084732532501, "learning_rate": 7.453558357939755e-05, "loss": 0.0422, "step": 3700 }, { "epoch": 14.324324324324325, "grad_norm": 0.3639368414878845, "learning_rate": 7.439137921790606e-05, "loss": 0.0336, "step": 3710 }, { "epoch": 14.362934362934363, "grad_norm": 0.3499019742012024, "learning_rate": 7.42469081161993e-05, "loss": 0.0421, "step": 3720 }, { "epoch": 14.4015444015444, "grad_norm": 0.39822128415107727, "learning_rate": 7.410217185419006e-05, "loss": 0.0308, "step": 3730 }, { "epoch": 14.44015444015444, "grad_norm": 0.3240340054035187, "learning_rate": 7.395717201469095e-05, "loss": 0.0331, "step": 3740 }, { "epoch": 14.478764478764479, "grad_norm": 0.3973540961742401, "learning_rate": 7.381191018339696e-05, "loss": 0.0438, "step": 3750 }, { "epoch": 14.517374517374517, "grad_norm": 0.3613256812095642, "learning_rate": 7.36663879488682e-05, "loss": 0.0344, "step": 3760 }, { "epoch": 14.555984555984557, "grad_norm": 0.31584304571151733, "learning_rate": 7.352060690251254e-05, "loss": 0.0321, "step": 3770 }, { "epoch": 14.594594594594595, "grad_norm": 0.4012964069843292, "learning_rate": 7.337456863856811e-05, "loss": 0.0394, "step": 3780 }, { "epoch": 14.633204633204633, "grad_norm": 0.36479297280311584, "learning_rate": 7.3228274754086e-05, "loss": 0.0313, "step": 3790 }, { "epoch": 14.671814671814673, "grad_norm": 0.3085634112358093, "learning_rate": 7.308172684891267e-05, "loss": 0.0326, "step": 3800 }, { "epoch": 14.71042471042471, "grad_norm": 0.35768958926200867, "learning_rate": 7.293492652567255e-05, "loss": 0.0347, "step": 3810 }, { "epoch": 14.749034749034749, "grad_norm": 0.3040185272693634, "learning_rate": 7.278787538975043e-05, "loss": 0.0315, "step": 3820 }, { "epoch": 14.787644787644787, "grad_norm": 0.24185171723365784, "learning_rate": 7.2640575049274e-05, "loss": 0.033, "step": 3830 }, { "epoch": 14.826254826254827, "grad_norm": 0.27290382981300354, "learning_rate": 7.249302711509616e-05, "loss": 0.0345, "step": 3840 }, { "epoch": 14.864864864864865, "grad_norm": 0.39349237084388733, "learning_rate": 7.23452332007775e-05, "loss": 0.0326, "step": 3850 }, { "epoch": 14.903474903474903, "grad_norm": 0.5151225924491882, "learning_rate": 7.219719492256858e-05, "loss": 0.0339, "step": 3860 }, { "epoch": 14.942084942084943, "grad_norm": 0.24791927635669708, "learning_rate": 7.20489138993923e-05, "loss": 0.0346, "step": 3870 }, { "epoch": 14.980694980694981, "grad_norm": 0.5492618083953857, "learning_rate": 7.190039175282614e-05, "loss": 0.0356, "step": 3880 }, { "epoch": 15.019305019305019, "grad_norm": 0.2631070017814636, "learning_rate": 7.175163010708455e-05, "loss": 0.0313, "step": 3890 }, { "epoch": 15.057915057915057, "grad_norm": 0.3902176022529602, "learning_rate": 7.1602630589001e-05, "loss": 0.0265, "step": 3900 }, { "epoch": 15.096525096525097, "grad_norm": 0.26857301592826843, "learning_rate": 7.14533948280104e-05, "loss": 0.0305, "step": 3910 }, { "epoch": 15.135135135135135, "grad_norm": 0.25791504979133606, "learning_rate": 7.130392445613109e-05, "loss": 0.0307, "step": 3920 }, { "epoch": 15.173745173745173, "grad_norm": 0.45596209168434143, "learning_rate": 7.115422110794711e-05, "loss": 0.0323, "step": 3930 }, { "epoch": 15.212355212355213, "grad_norm": 0.39027926325798035, "learning_rate": 7.100428642059033e-05, "loss": 0.0345, "step": 3940 }, { "epoch": 15.250965250965251, "grad_norm": 0.4901469647884369, "learning_rate": 7.08541220337224e-05, "loss": 0.0337, "step": 3950 }, { "epoch": 15.28957528957529, "grad_norm": 0.4429934024810791, "learning_rate": 7.070372958951706e-05, "loss": 0.0361, "step": 3960 }, { "epoch": 15.328185328185327, "grad_norm": 0.4682411551475525, "learning_rate": 7.055311073264194e-05, "loss": 0.0345, "step": 3970 }, { "epoch": 15.366795366795367, "grad_norm": 0.39673370122909546, "learning_rate": 7.040226711024077e-05, "loss": 0.0377, "step": 3980 }, { "epoch": 15.405405405405405, "grad_norm": 0.24011917412281036, "learning_rate": 7.02512003719152e-05, "loss": 0.0298, "step": 3990 }, { "epoch": 15.444015444015443, "grad_norm": 0.2510679066181183, "learning_rate": 7.00999121697069e-05, "loss": 0.0303, "step": 4000 }, { "epoch": 15.482625482625483, "grad_norm": 0.6246134042739868, "learning_rate": 6.99484041580794e-05, "loss": 0.0334, "step": 4010 }, { "epoch": 15.521235521235521, "grad_norm": 0.3347349464893341, "learning_rate": 6.979667799390004e-05, "loss": 0.0386, "step": 4020 }, { "epoch": 15.55984555984556, "grad_norm": 0.413377046585083, "learning_rate": 6.964473533642185e-05, "loss": 0.0322, "step": 4030 }, { "epoch": 15.5984555984556, "grad_norm": 0.24597837030887604, "learning_rate": 6.949257784726539e-05, "loss": 0.0336, "step": 4040 }, { "epoch": 15.637065637065637, "grad_norm": 0.3981229066848755, "learning_rate": 6.934020719040056e-05, "loss": 0.0416, "step": 4050 }, { "epoch": 15.675675675675675, "grad_norm": 0.409208744764328, "learning_rate": 6.918762503212848e-05, "loss": 0.0294, "step": 4060 }, { "epoch": 15.714285714285714, "grad_norm": 0.3906567096710205, "learning_rate": 6.903483304106319e-05, "loss": 0.0343, "step": 4070 }, { "epoch": 15.752895752895753, "grad_norm": 0.4177800714969635, "learning_rate": 6.888183288811341e-05, "loss": 0.0341, "step": 4080 }, { "epoch": 15.791505791505791, "grad_norm": 0.3486420810222626, "learning_rate": 6.87286262464643e-05, "loss": 0.0297, "step": 4090 }, { "epoch": 15.83011583011583, "grad_norm": 0.3111885190010071, "learning_rate": 6.857521479155915e-05, "loss": 0.0285, "step": 4100 }, { "epoch": 15.86872586872587, "grad_norm": 0.2644309997558594, "learning_rate": 6.842160020108104e-05, "loss": 0.0395, "step": 4110 }, { "epoch": 15.907335907335908, "grad_norm": 0.2589646577835083, "learning_rate": 6.826778415493455e-05, "loss": 0.036, "step": 4120 }, { "epoch": 15.945945945945946, "grad_norm": 0.3872307240962982, "learning_rate": 6.811376833522729e-05, "loss": 0.0347, "step": 4130 }, { "epoch": 15.984555984555985, "grad_norm": 0.3081093430519104, "learning_rate": 6.795955442625159e-05, "loss": 0.0388, "step": 4140 }, { "epoch": 16.023166023166024, "grad_norm": 0.2758740782737732, "learning_rate": 6.780514411446608e-05, "loss": 0.0335, "step": 4150 }, { "epoch": 16.06177606177606, "grad_norm": 0.2429054230451584, "learning_rate": 6.765053908847716e-05, "loss": 0.0302, "step": 4160 }, { "epoch": 16.1003861003861, "grad_norm": 0.29284581542015076, "learning_rate": 6.749574103902064e-05, "loss": 0.0249, "step": 4170 }, { "epoch": 16.138996138996138, "grad_norm": 0.30065611004829407, "learning_rate": 6.734075165894317e-05, "loss": 0.0373, "step": 4180 }, { "epoch": 16.177606177606176, "grad_norm": 0.30541205406188965, "learning_rate": 6.71855726431838e-05, "loss": 0.0302, "step": 4190 }, { "epoch": 16.216216216216218, "grad_norm": 0.20503666996955872, "learning_rate": 6.703020568875538e-05, "loss": 0.0331, "step": 4200 }, { "epoch": 16.254826254826256, "grad_norm": 0.2356152981519699, "learning_rate": 6.687465249472603e-05, "loss": 0.0281, "step": 4210 }, { "epoch": 16.293436293436294, "grad_norm": 0.3513968586921692, "learning_rate": 6.671891476220055e-05, "loss": 0.0285, "step": 4220 }, { "epoch": 16.332046332046332, "grad_norm": 0.3512837290763855, "learning_rate": 6.656299419430183e-05, "loss": 0.0287, "step": 4230 }, { "epoch": 16.37065637065637, "grad_norm": 0.3847474455833435, "learning_rate": 6.640689249615223e-05, "loss": 0.0342, "step": 4240 }, { "epoch": 16.409266409266408, "grad_norm": 0.2660544216632843, "learning_rate": 6.625061137485491e-05, "loss": 0.0337, "step": 4250 }, { "epoch": 16.447876447876446, "grad_norm": 0.2844618260860443, "learning_rate": 6.609415253947517e-05, "loss": 0.034, "step": 4260 }, { "epoch": 16.486486486486488, "grad_norm": 0.43186402320861816, "learning_rate": 6.593751770102178e-05, "loss": 0.0388, "step": 4270 }, { "epoch": 16.525096525096526, "grad_norm": 0.25738397240638733, "learning_rate": 6.578070857242823e-05, "loss": 0.0291, "step": 4280 }, { "epoch": 16.563706563706564, "grad_norm": 0.37723344564437866, "learning_rate": 6.562372686853402e-05, "loss": 0.0281, "step": 4290 }, { "epoch": 16.602316602316602, "grad_norm": 0.39555278420448303, "learning_rate": 6.546657430606593e-05, "loss": 0.0291, "step": 4300 }, { "epoch": 16.64092664092664, "grad_norm": 0.3441212475299835, "learning_rate": 6.530925260361918e-05, "loss": 0.0336, "step": 4310 }, { "epoch": 16.679536679536678, "grad_norm": 0.5214934349060059, "learning_rate": 6.515176348163871e-05, "loss": 0.0347, "step": 4320 }, { "epoch": 16.71814671814672, "grad_norm": 0.3228941857814789, "learning_rate": 6.499410866240032e-05, "loss": 0.0361, "step": 4330 }, { "epoch": 16.756756756756758, "grad_norm": 0.3555833697319031, "learning_rate": 6.48362898699919e-05, "loss": 0.0333, "step": 4340 }, { "epoch": 16.795366795366796, "grad_norm": 0.46881037950515747, "learning_rate": 6.467830883029443e-05, "loss": 0.0326, "step": 4350 }, { "epoch": 16.833976833976834, "grad_norm": 0.3284620940685272, "learning_rate": 6.452016727096326e-05, "loss": 0.0336, "step": 4360 }, { "epoch": 16.872586872586872, "grad_norm": 0.3858324885368347, "learning_rate": 6.436186692140916e-05, "loss": 0.0262, "step": 4370 }, { "epoch": 16.91119691119691, "grad_norm": 0.6424157619476318, "learning_rate": 6.420340951277938e-05, "loss": 0.041, "step": 4380 }, { "epoch": 16.94980694980695, "grad_norm": 0.28588026762008667, "learning_rate": 6.404479677793874e-05, "loss": 0.0325, "step": 4390 }, { "epoch": 16.98841698841699, "grad_norm": 0.34062522649765015, "learning_rate": 6.388603045145075e-05, "loss": 0.0323, "step": 4400 }, { "epoch": 17.027027027027028, "grad_norm": 0.4384026527404785, "learning_rate": 6.372711226955843e-05, "loss": 0.0396, "step": 4410 }, { "epoch": 17.065637065637066, "grad_norm": 0.4747617244720459, "learning_rate": 6.356804397016564e-05, "loss": 0.0344, "step": 4420 }, { "epoch": 17.104247104247104, "grad_norm": 0.4412192702293396, "learning_rate": 6.340882729281779e-05, "loss": 0.0326, "step": 4430 }, { "epoch": 17.142857142857142, "grad_norm": 0.2094019204378128, "learning_rate": 6.324946397868294e-05, "loss": 0.0281, "step": 4440 }, { "epoch": 17.18146718146718, "grad_norm": 0.3119041919708252, "learning_rate": 6.308995577053276e-05, "loss": 0.0343, "step": 4450 }, { "epoch": 17.22007722007722, "grad_norm": 0.43302884697914124, "learning_rate": 6.293030441272347e-05, "loss": 0.0276, "step": 4460 }, { "epoch": 17.25868725868726, "grad_norm": 0.333455353975296, "learning_rate": 6.277051165117677e-05, "loss": 0.0319, "step": 4470 }, { "epoch": 17.2972972972973, "grad_norm": 0.31548604369163513, "learning_rate": 6.261057923336064e-05, "loss": 0.0299, "step": 4480 }, { "epoch": 17.335907335907336, "grad_norm": 0.5287732481956482, "learning_rate": 6.245050890827042e-05, "loss": 0.0323, "step": 4490 }, { "epoch": 17.374517374517374, "grad_norm": 0.3672085106372833, "learning_rate": 6.229030242640952e-05, "loss": 0.0322, "step": 4500 }, { "epoch": 17.413127413127413, "grad_norm": 0.46073928475379944, "learning_rate": 6.212996153977037e-05, "loss": 0.0319, "step": 4510 }, { "epoch": 17.45173745173745, "grad_norm": 0.31028667092323303, "learning_rate": 6.196948800181523e-05, "loss": 0.0308, "step": 4520 }, { "epoch": 17.49034749034749, "grad_norm": 0.3594716787338257, "learning_rate": 6.180888356745695e-05, "loss": 0.0343, "step": 4530 }, { "epoch": 17.52895752895753, "grad_norm": 0.3466090261936188, "learning_rate": 6.164814999303995e-05, "loss": 0.0313, "step": 4540 }, { "epoch": 17.56756756756757, "grad_norm": 0.35559895634651184, "learning_rate": 6.148728903632081e-05, "loss": 0.0291, "step": 4550 }, { "epoch": 17.606177606177607, "grad_norm": 0.316925048828125, "learning_rate": 6.132630245644921e-05, "loss": 0.0269, "step": 4560 }, { "epoch": 17.644787644787645, "grad_norm": 0.3211555480957031, "learning_rate": 6.116519201394857e-05, "loss": 0.0289, "step": 4570 }, { "epoch": 17.683397683397683, "grad_norm": 0.23555810749530792, "learning_rate": 6.10039594706969e-05, "loss": 0.0325, "step": 4580 }, { "epoch": 17.72200772200772, "grad_norm": 0.32163241505622864, "learning_rate": 6.084260658990744e-05, "loss": 0.0256, "step": 4590 }, { "epoch": 17.760617760617762, "grad_norm": 0.2548912763595581, "learning_rate": 6.068113513610943e-05, "loss": 0.0318, "step": 4600 }, { "epoch": 17.7992277992278, "grad_norm": 0.3132483959197998, "learning_rate": 6.0519546875128876e-05, "loss": 0.0337, "step": 4610 }, { "epoch": 17.83783783783784, "grad_norm": 0.2327328324317932, "learning_rate": 6.035784357406906e-05, "loss": 0.0308, "step": 4620 }, { "epoch": 17.876447876447877, "grad_norm": 0.3992815315723419, "learning_rate": 6.01960270012914e-05, "loss": 0.0354, "step": 4630 }, { "epoch": 17.915057915057915, "grad_norm": 0.28766167163848877, "learning_rate": 6.003409892639599e-05, "loss": 0.0279, "step": 4640 }, { "epoch": 17.953667953667953, "grad_norm": 0.24114270508289337, "learning_rate": 5.9872061120202336e-05, "loss": 0.0244, "step": 4650 }, { "epoch": 17.99227799227799, "grad_norm": 0.2907788157463074, "learning_rate": 5.9709915354729914e-05, "loss": 0.0312, "step": 4660 }, { "epoch": 18.030888030888033, "grad_norm": 0.3101578652858734, "learning_rate": 5.9547663403178824e-05, "loss": 0.0345, "step": 4670 }, { "epoch": 18.06949806949807, "grad_norm": 0.58769291639328, "learning_rate": 5.9385307039910445e-05, "loss": 0.0335, "step": 4680 }, { "epoch": 18.10810810810811, "grad_norm": 0.3008973300457001, "learning_rate": 5.922284804042792e-05, "loss": 0.0281, "step": 4690 }, { "epoch": 18.146718146718147, "grad_norm": 0.2778293788433075, "learning_rate": 5.906028818135687e-05, "loss": 0.0318, "step": 4700 }, { "epoch": 18.185328185328185, "grad_norm": 0.35999932885169983, "learning_rate": 5.889762924042585e-05, "loss": 0.0353, "step": 4710 }, { "epoch": 18.223938223938223, "grad_norm": 0.3294525444507599, "learning_rate": 5.873487299644699e-05, "loss": 0.0281, "step": 4720 }, { "epoch": 18.26254826254826, "grad_norm": 0.5216575860977173, "learning_rate": 5.857202122929649e-05, "loss": 0.0368, "step": 4730 }, { "epoch": 18.301158301158303, "grad_norm": 0.4763036370277405, "learning_rate": 5.840907571989518e-05, "loss": 0.0253, "step": 4740 }, { "epoch": 18.33976833976834, "grad_norm": 0.3455185890197754, "learning_rate": 5.824603825018904e-05, "loss": 0.028, "step": 4750 }, { "epoch": 18.37837837837838, "grad_norm": 0.34274202585220337, "learning_rate": 5.808291060312975e-05, "loss": 0.0264, "step": 4760 }, { "epoch": 18.416988416988417, "grad_norm": 0.34912335872650146, "learning_rate": 5.7919694562655083e-05, "loss": 0.035, "step": 4770 }, { "epoch": 18.455598455598455, "grad_norm": 0.3614436984062195, "learning_rate": 5.775639191366954e-05, "loss": 0.0291, "step": 4780 }, { "epoch": 18.494208494208493, "grad_norm": 0.47598758339881897, "learning_rate": 5.75930044420247e-05, "loss": 0.0358, "step": 4790 }, { "epoch": 18.53281853281853, "grad_norm": 0.27316945791244507, "learning_rate": 5.74295339344998e-05, "loss": 0.0255, "step": 4800 }, { "epoch": 18.571428571428573, "grad_norm": 0.27487045526504517, "learning_rate": 5.726598217878211e-05, "loss": 0.0341, "step": 4810 }, { "epoch": 18.61003861003861, "grad_norm": 0.36344462633132935, "learning_rate": 5.71023509634474e-05, "loss": 0.0273, "step": 4820 }, { "epoch": 18.64864864864865, "grad_norm": 0.2788923382759094, "learning_rate": 5.693864207794049e-05, "loss": 0.0279, "step": 4830 }, { "epoch": 18.687258687258687, "grad_norm": 0.40637412667274475, "learning_rate": 5.677485731255545e-05, "loss": 0.0291, "step": 4840 }, { "epoch": 18.725868725868725, "grad_norm": 0.2882944345474243, "learning_rate": 5.6610998458416296e-05, "loss": 0.0357, "step": 4850 }, { "epoch": 18.764478764478763, "grad_norm": 0.3950125575065613, "learning_rate": 5.644706730745716e-05, "loss": 0.026, "step": 4860 }, { "epoch": 18.8030888030888, "grad_norm": 0.31441977620124817, "learning_rate": 5.628306565240287e-05, "loss": 0.0264, "step": 4870 }, { "epoch": 18.841698841698843, "grad_norm": 0.24917064607143402, "learning_rate": 5.611899528674923e-05, "loss": 0.0298, "step": 4880 }, { "epoch": 18.88030888030888, "grad_norm": 0.36281150579452515, "learning_rate": 5.595485800474349e-05, "loss": 0.0284, "step": 4890 }, { "epoch": 18.91891891891892, "grad_norm": 0.3348575234413147, "learning_rate": 5.579065560136467e-05, "loss": 0.0288, "step": 4900 }, { "epoch": 18.957528957528957, "grad_norm": 0.3499665856361389, "learning_rate": 5.562638987230392e-05, "loss": 0.0353, "step": 4910 }, { "epoch": 18.996138996138995, "grad_norm": 0.32521766424179077, "learning_rate": 5.546206261394498e-05, "loss": 0.0333, "step": 4920 }, { "epoch": 19.034749034749034, "grad_norm": 0.3005838692188263, "learning_rate": 5.529767562334437e-05, "loss": 0.0301, "step": 4930 }, { "epoch": 19.07335907335907, "grad_norm": 0.33923184871673584, "learning_rate": 5.5133230698211926e-05, "loss": 0.029, "step": 4940 }, { "epoch": 19.111969111969113, "grad_norm": 0.20707932114601135, "learning_rate": 5.496872963689096e-05, "loss": 0.0272, "step": 4950 }, { "epoch": 19.15057915057915, "grad_norm": 0.2848384380340576, "learning_rate": 5.4804174238338756e-05, "loss": 0.032, "step": 4960 }, { "epoch": 19.18918918918919, "grad_norm": 0.37866681814193726, "learning_rate": 5.463956630210678e-05, "loss": 0.0266, "step": 4970 }, { "epoch": 19.227799227799228, "grad_norm": 0.3295511305332184, "learning_rate": 5.4474907628321046e-05, "loss": 0.0305, "step": 4980 }, { "epoch": 19.266409266409266, "grad_norm": 0.3358299136161804, "learning_rate": 5.431020001766244e-05, "loss": 0.0369, "step": 4990 }, { "epoch": 19.305019305019304, "grad_norm": 0.2716432511806488, "learning_rate": 5.4145445271346986e-05, "loss": 0.0264, "step": 5000 }, { "epoch": 19.343629343629345, "grad_norm": 0.3606038987636566, "learning_rate": 5.398064519110622e-05, "loss": 0.0316, "step": 5010 }, { "epoch": 19.382239382239383, "grad_norm": 0.42593032121658325, "learning_rate": 5.3815801579167394e-05, "loss": 0.0287, "step": 5020 }, { "epoch": 19.42084942084942, "grad_norm": 0.25910916924476624, "learning_rate": 5.365091623823382e-05, "loss": 0.0285, "step": 5030 }, { "epoch": 19.45945945945946, "grad_norm": 0.42744606733322144, "learning_rate": 5.348599097146521e-05, "loss": 0.0316, "step": 5040 }, { "epoch": 19.498069498069498, "grad_norm": 0.4178139865398407, "learning_rate": 5.3321027582457836e-05, "loss": 0.0235, "step": 5050 }, { "epoch": 19.536679536679536, "grad_norm": 0.2788640260696411, "learning_rate": 5.315602787522491e-05, "loss": 0.0264, "step": 5060 }, { "epoch": 19.575289575289574, "grad_norm": 0.31901830434799194, "learning_rate": 5.299099365417678e-05, "loss": 0.0223, "step": 5070 }, { "epoch": 19.613899613899616, "grad_norm": 0.25880661606788635, "learning_rate": 5.2825926724101236e-05, "loss": 0.0298, "step": 5080 }, { "epoch": 19.652509652509654, "grad_norm": 0.3076111972332001, "learning_rate": 5.26608288901438e-05, "loss": 0.0261, "step": 5090 }, { "epoch": 19.69111969111969, "grad_norm": 0.419592946767807, "learning_rate": 5.24957019577879e-05, "loss": 0.0271, "step": 5100 }, { "epoch": 19.72972972972973, "grad_norm": 0.5848473310470581, "learning_rate": 5.2330547732835266e-05, "loss": 0.0338, "step": 5110 }, { "epoch": 19.768339768339768, "grad_norm": 0.2379187047481537, "learning_rate": 5.2165368021385996e-05, "loss": 0.0251, "step": 5120 }, { "epoch": 19.806949806949806, "grad_norm": 0.3098015785217285, "learning_rate": 5.200016462981897e-05, "loss": 0.0283, "step": 5130 }, { "epoch": 19.845559845559844, "grad_norm": 0.33206260204315186, "learning_rate": 5.1834939364772015e-05, "loss": 0.0251, "step": 5140 }, { "epoch": 19.884169884169886, "grad_norm": 0.4062725305557251, "learning_rate": 5.166969403312214e-05, "loss": 0.027, "step": 5150 }, { "epoch": 19.922779922779924, "grad_norm": 0.38283345103263855, "learning_rate": 5.1504430441965844e-05, "loss": 0.0322, "step": 5160 }, { "epoch": 19.961389961389962, "grad_norm": 0.3244209587574005, "learning_rate": 5.133915039859923e-05, "loss": 0.0307, "step": 5170 }, { "epoch": 20.0, "grad_norm": 0.30077436566352844, "learning_rate": 5.1173855710498444e-05, "loss": 0.0274, "step": 5180 }, { "epoch": 20.038610038610038, "grad_norm": 0.2549975514411926, "learning_rate": 5.100854818529967e-05, "loss": 0.0268, "step": 5190 }, { "epoch": 20.077220077220076, "grad_norm": 0.3340621590614319, "learning_rate": 5.084322963077951e-05, "loss": 0.0301, "step": 5200 }, { "epoch": 20.115830115830114, "grad_norm": 0.311779648065567, "learning_rate": 5.067790185483522e-05, "loss": 0.0309, "step": 5210 }, { "epoch": 20.154440154440156, "grad_norm": 0.2883121073246002, "learning_rate": 5.0512566665464844e-05, "loss": 0.0325, "step": 5220 }, { "epoch": 20.193050193050194, "grad_norm": 0.2152983546257019, "learning_rate": 5.034722587074755e-05, "loss": 0.0309, "step": 5230 }, { "epoch": 20.231660231660232, "grad_norm": 0.3085116147994995, "learning_rate": 5.018188127882375e-05, "loss": 0.0285, "step": 5240 }, { "epoch": 20.27027027027027, "grad_norm": 0.20602749288082123, "learning_rate": 5.0016534697875417e-05, "loss": 0.0264, "step": 5250 }, { "epoch": 20.30888030888031, "grad_norm": 0.3064497411251068, "learning_rate": 4.9851187936106294e-05, "loss": 0.0287, "step": 5260 }, { "epoch": 20.347490347490346, "grad_norm": 0.2662597894668579, "learning_rate": 4.968584280172206e-05, "loss": 0.0254, "step": 5270 }, { "epoch": 20.386100386100384, "grad_norm": 0.33520036935806274, "learning_rate": 4.95205011029106e-05, "loss": 0.0316, "step": 5280 }, { "epoch": 20.424710424710426, "grad_norm": 0.3352830111980438, "learning_rate": 4.935516464782227e-05, "loss": 0.0267, "step": 5290 }, { "epoch": 20.463320463320464, "grad_norm": 0.2322608232498169, "learning_rate": 4.918983524455003e-05, "loss": 0.0243, "step": 5300 }, { "epoch": 20.501930501930502, "grad_norm": 0.39726293087005615, "learning_rate": 4.9024514701109766e-05, "loss": 0.0235, "step": 5310 }, { "epoch": 20.54054054054054, "grad_norm": 0.41719067096710205, "learning_rate": 4.885920482542043e-05, "loss": 0.0314, "step": 5320 }, { "epoch": 20.57915057915058, "grad_norm": 0.23369985818862915, "learning_rate": 4.869390742528438e-05, "loss": 0.0234, "step": 5330 }, { "epoch": 20.617760617760617, "grad_norm": 0.20701904594898224, "learning_rate": 4.852862430836744e-05, "loss": 0.0229, "step": 5340 }, { "epoch": 20.656370656370655, "grad_norm": 0.2636873126029968, "learning_rate": 4.836335728217933e-05, "loss": 0.0258, "step": 5350 }, { "epoch": 20.694980694980696, "grad_norm": 0.24337151646614075, "learning_rate": 4.819810815405379e-05, "loss": 0.0257, "step": 5360 }, { "epoch": 20.733590733590734, "grad_norm": 0.2224522829055786, "learning_rate": 4.803287873112877e-05, "loss": 0.0307, "step": 5370 }, { "epoch": 20.772200772200772, "grad_norm": 0.3021717071533203, "learning_rate": 4.786767082032681e-05, "loss": 0.0299, "step": 5380 }, { "epoch": 20.81081081081081, "grad_norm": 0.3682388365268707, "learning_rate": 4.77024862283351e-05, "loss": 0.0252, "step": 5390 }, { "epoch": 20.84942084942085, "grad_norm": 0.2683052718639374, "learning_rate": 4.753732676158593e-05, "loss": 0.0255, "step": 5400 }, { "epoch": 20.888030888030887, "grad_norm": 0.41546890139579773, "learning_rate": 4.737219422623672e-05, "loss": 0.028, "step": 5410 }, { "epoch": 20.92664092664093, "grad_norm": 0.2585113048553467, "learning_rate": 4.720709042815044e-05, "loss": 0.0278, "step": 5420 }, { "epoch": 20.965250965250966, "grad_norm": 0.2921794354915619, "learning_rate": 4.704201717287578e-05, "loss": 0.0238, "step": 5430 }, { "epoch": 21.003861003861005, "grad_norm": 0.3413592576980591, "learning_rate": 4.6876976265627404e-05, "loss": 0.022, "step": 5440 }, { "epoch": 21.042471042471043, "grad_norm": 0.31393495202064514, "learning_rate": 4.671196951126626e-05, "loss": 0.0296, "step": 5450 }, { "epoch": 21.08108108108108, "grad_norm": 0.36006689071655273, "learning_rate": 4.654699871427971e-05, "loss": 0.0293, "step": 5460 }, { "epoch": 21.11969111969112, "grad_norm": 0.30264416337013245, "learning_rate": 4.6382065678762034e-05, "loss": 0.0289, "step": 5470 }, { "epoch": 21.158301158301157, "grad_norm": 0.4143695831298828, "learning_rate": 4.6217172208394424e-05, "loss": 0.0263, "step": 5480 }, { "epoch": 21.1969111969112, "grad_norm": 0.3586815595626831, "learning_rate": 4.605232010642549e-05, "loss": 0.0282, "step": 5490 }, { "epoch": 21.235521235521237, "grad_norm": 0.33841633796691895, "learning_rate": 4.588751117565142e-05, "loss": 0.0286, "step": 5500 }, { "epoch": 21.274131274131275, "grad_norm": 0.32587555050849915, "learning_rate": 4.5722747218396214e-05, "loss": 0.0254, "step": 5510 }, { "epoch": 21.312741312741313, "grad_norm": 0.2053976058959961, "learning_rate": 4.5558030036492194e-05, "loss": 0.0213, "step": 5520 }, { "epoch": 21.35135135135135, "grad_norm": 0.23872408270835876, "learning_rate": 4.539336143125999e-05, "loss": 0.0269, "step": 5530 }, { "epoch": 21.38996138996139, "grad_norm": 0.23521359264850616, "learning_rate": 4.522874320348916e-05, "loss": 0.025, "step": 5540 }, { "epoch": 21.428571428571427, "grad_norm": 0.2618771195411682, "learning_rate": 4.506417715341821e-05, "loss": 0.0293, "step": 5550 }, { "epoch": 21.46718146718147, "grad_norm": 0.32607150077819824, "learning_rate": 4.489966508071511e-05, "loss": 0.0184, "step": 5560 }, { "epoch": 21.505791505791507, "grad_norm": 0.23522065579891205, "learning_rate": 4.4735208784457575e-05, "loss": 0.0233, "step": 5570 }, { "epoch": 21.544401544401545, "grad_norm": 0.4004756808280945, "learning_rate": 4.457081006311325e-05, "loss": 0.0257, "step": 5580 }, { "epoch": 21.583011583011583, "grad_norm": 0.17714454233646393, "learning_rate": 4.440647071452027e-05, "loss": 0.0216, "step": 5590 }, { "epoch": 21.62162162162162, "grad_norm": 0.4133980870246887, "learning_rate": 4.424219253586737e-05, "loss": 0.0305, "step": 5600 }, { "epoch": 21.66023166023166, "grad_norm": 0.32950690388679504, "learning_rate": 4.407797732367443e-05, "loss": 0.0248, "step": 5610 }, { "epoch": 21.698841698841697, "grad_norm": 0.35809797048568726, "learning_rate": 4.391382687377268e-05, "loss": 0.0259, "step": 5620 }, { "epoch": 21.73745173745174, "grad_norm": 0.37596961855888367, "learning_rate": 4.374974298128512e-05, "loss": 0.0248, "step": 5630 }, { "epoch": 21.776061776061777, "grad_norm": 0.2942051887512207, "learning_rate": 4.358572744060699e-05, "loss": 0.024, "step": 5640 }, { "epoch": 21.814671814671815, "grad_norm": 0.2354700118303299, "learning_rate": 4.342178204538588e-05, "loss": 0.0229, "step": 5650 }, { "epoch": 21.853281853281853, "grad_norm": 0.36603614687919617, "learning_rate": 4.325790858850241e-05, "loss": 0.0301, "step": 5660 }, { "epoch": 21.89189189189189, "grad_norm": 0.346853643655777, "learning_rate": 4.309410886205043e-05, "loss": 0.0211, "step": 5670 }, { "epoch": 21.93050193050193, "grad_norm": 0.27295777201652527, "learning_rate": 4.293038465731752e-05, "loss": 0.0258, "step": 5680 }, { "epoch": 21.969111969111967, "grad_norm": 0.2574758231639862, "learning_rate": 4.276673776476533e-05, "loss": 0.0262, "step": 5690 }, { "epoch": 22.00772200772201, "grad_norm": 0.5941938161849976, "learning_rate": 4.260316997401007e-05, "loss": 0.0286, "step": 5700 }, { "epoch": 22.046332046332047, "grad_norm": 0.2250533103942871, "learning_rate": 4.243968307380293e-05, "loss": 0.0279, "step": 5710 }, { "epoch": 22.084942084942085, "grad_norm": 0.24899111688137054, "learning_rate": 4.22762788520104e-05, "loss": 0.0214, "step": 5720 }, { "epoch": 22.123552123552123, "grad_norm": 0.1589987874031067, "learning_rate": 4.211295909559491e-05, "loss": 0.0263, "step": 5730 }, { "epoch": 22.16216216216216, "grad_norm": 0.35328927636146545, "learning_rate": 4.194972559059511e-05, "loss": 0.0299, "step": 5740 }, { "epoch": 22.2007722007722, "grad_norm": 0.3259015679359436, "learning_rate": 4.178658012210651e-05, "loss": 0.0253, "step": 5750 }, { "epoch": 22.23938223938224, "grad_norm": 0.22607411444187164, "learning_rate": 4.162352447426177e-05, "loss": 0.028, "step": 5760 }, { "epoch": 22.27799227799228, "grad_norm": 0.2671155333518982, "learning_rate": 4.146056043021135e-05, "loss": 0.0224, "step": 5770 }, { "epoch": 22.316602316602317, "grad_norm": 0.246576189994812, "learning_rate": 4.1297689772103944e-05, "loss": 0.0214, "step": 5780 }, { "epoch": 22.355212355212355, "grad_norm": 0.3703608214855194, "learning_rate": 4.113491428106694e-05, "loss": 0.0209, "step": 5790 }, { "epoch": 22.393822393822393, "grad_norm": 0.2342519462108612, "learning_rate": 4.0972235737187055e-05, "loss": 0.0293, "step": 5800 }, { "epoch": 22.43243243243243, "grad_norm": 0.26736652851104736, "learning_rate": 4.080965591949076e-05, "loss": 0.0253, "step": 5810 }, { "epoch": 22.47104247104247, "grad_norm": 0.3427676856517792, "learning_rate": 4.0647176605924924e-05, "loss": 0.025, "step": 5820 }, { "epoch": 22.50965250965251, "grad_norm": 0.22933968901634216, "learning_rate": 4.0484799573337255e-05, "loss": 0.0279, "step": 5830 }, { "epoch": 22.54826254826255, "grad_norm": 0.24397163093090057, "learning_rate": 4.032252659745699e-05, "loss": 0.0281, "step": 5840 }, { "epoch": 22.586872586872587, "grad_norm": 0.21098196506500244, "learning_rate": 4.016035945287539e-05, "loss": 0.0235, "step": 5850 }, { "epoch": 22.625482625482626, "grad_norm": 0.2663086950778961, "learning_rate": 3.999829991302635e-05, "loss": 0.0271, "step": 5860 }, { "epoch": 22.664092664092664, "grad_norm": 0.1799061894416809, "learning_rate": 3.983634975016707e-05, "loss": 0.0284, "step": 5870 }, { "epoch": 22.7027027027027, "grad_norm": 0.34151723980903625, "learning_rate": 3.967451073535854e-05, "loss": 0.0287, "step": 5880 }, { "epoch": 22.74131274131274, "grad_norm": 0.28058362007141113, "learning_rate": 3.951278463844633e-05, "loss": 0.0294, "step": 5890 }, { "epoch": 22.77992277992278, "grad_norm": 0.27997541427612305, "learning_rate": 3.935117322804111e-05, "loss": 0.028, "step": 5900 }, { "epoch": 22.81853281853282, "grad_norm": 0.27230140566825867, "learning_rate": 3.918967827149938e-05, "loss": 0.0243, "step": 5910 }, { "epoch": 22.857142857142858, "grad_norm": 0.28498831391334534, "learning_rate": 3.9028301534904094e-05, "loss": 0.028, "step": 5920 }, { "epoch": 22.895752895752896, "grad_norm": 0.2946162521839142, "learning_rate": 3.88670447830454e-05, "loss": 0.0247, "step": 5930 }, { "epoch": 22.934362934362934, "grad_norm": 0.3814082443714142, "learning_rate": 3.870590977940132e-05, "loss": 0.0259, "step": 5940 }, { "epoch": 22.972972972972972, "grad_norm": 0.2759941816329956, "learning_rate": 3.8544898286118404e-05, "loss": 0.0235, "step": 5950 }, { "epoch": 23.01158301158301, "grad_norm": 0.2239333689212799, "learning_rate": 3.838401206399257e-05, "loss": 0.0211, "step": 5960 }, { "epoch": 23.05019305019305, "grad_norm": 0.41559910774230957, "learning_rate": 3.822325287244975e-05, "loss": 0.0235, "step": 5970 }, { "epoch": 23.08880308880309, "grad_norm": 0.3017045557498932, "learning_rate": 3.8062622469526725e-05, "loss": 0.0211, "step": 5980 }, { "epoch": 23.127413127413128, "grad_norm": 0.2614284157752991, "learning_rate": 3.790212261185183e-05, "loss": 0.0237, "step": 5990 }, { "epoch": 23.166023166023166, "grad_norm": 0.15439043939113617, "learning_rate": 3.7741755054625794e-05, "loss": 0.0204, "step": 6000 }, { "epoch": 23.204633204633204, "grad_norm": 0.23036736249923706, "learning_rate": 3.758152155160255e-05, "loss": 0.0252, "step": 6010 }, { "epoch": 23.243243243243242, "grad_norm": 0.373740017414093, "learning_rate": 3.742142385506999e-05, "loss": 0.0234, "step": 6020 }, { "epoch": 23.28185328185328, "grad_norm": 0.21924659609794617, "learning_rate": 3.72614637158309e-05, "loss": 0.0226, "step": 6030 }, { "epoch": 23.320463320463322, "grad_norm": 0.2984350919723511, "learning_rate": 3.710164288318371e-05, "loss": 0.0242, "step": 6040 }, { "epoch": 23.35907335907336, "grad_norm": 0.29278889298439026, "learning_rate": 3.694196310490345e-05, "loss": 0.0229, "step": 6050 }, { "epoch": 23.397683397683398, "grad_norm": 0.2716312110424042, "learning_rate": 3.678242612722259e-05, "loss": 0.0242, "step": 6060 }, { "epoch": 23.436293436293436, "grad_norm": 0.16538392007350922, "learning_rate": 3.6623033694811953e-05, "loss": 0.0178, "step": 6070 }, { "epoch": 23.474903474903474, "grad_norm": 0.1870395839214325, "learning_rate": 3.6463787550761665e-05, "loss": 0.0241, "step": 6080 }, { "epoch": 23.513513513513512, "grad_norm": 0.2741674482822418, "learning_rate": 3.630468943656202e-05, "loss": 0.0273, "step": 6090 }, { "epoch": 23.552123552123554, "grad_norm": 0.3386199176311493, "learning_rate": 3.6145741092084523e-05, "loss": 0.023, "step": 6100 }, { "epoch": 23.590733590733592, "grad_norm": 0.4225952625274658, "learning_rate": 3.598694425556278e-05, "loss": 0.0297, "step": 6110 }, { "epoch": 23.62934362934363, "grad_norm": 0.2701086103916168, "learning_rate": 3.58283006635736e-05, "loss": 0.0242, "step": 6120 }, { "epoch": 23.667953667953668, "grad_norm": 0.3287146985530853, "learning_rate": 3.566981205101781e-05, "loss": 0.025, "step": 6130 }, { "epoch": 23.706563706563706, "grad_norm": 0.25520870089530945, "learning_rate": 3.5511480151101556e-05, "loss": 0.0241, "step": 6140 }, { "epoch": 23.745173745173744, "grad_norm": 0.29353955388069153, "learning_rate": 3.5353306695317104e-05, "loss": 0.0253, "step": 6150 }, { "epoch": 23.783783783783782, "grad_norm": 0.22901315987110138, "learning_rate": 3.519529341342402e-05, "loss": 0.0189, "step": 6160 }, { "epoch": 23.822393822393824, "grad_norm": 0.22092147171497345, "learning_rate": 3.503744203343026e-05, "loss": 0.0237, "step": 6170 }, { "epoch": 23.861003861003862, "grad_norm": 0.321250319480896, "learning_rate": 3.487975428157318e-05, "loss": 0.0241, "step": 6180 }, { "epoch": 23.8996138996139, "grad_norm": 0.2735673487186432, "learning_rate": 3.472223188230083e-05, "loss": 0.0228, "step": 6190 }, { "epoch": 23.93822393822394, "grad_norm": 0.31167569756507874, "learning_rate": 3.4564876558252866e-05, "loss": 0.0276, "step": 6200 }, { "epoch": 23.976833976833976, "grad_norm": 0.28117433190345764, "learning_rate": 3.440769003024195e-05, "loss": 0.0285, "step": 6210 }, { "epoch": 24.015444015444015, "grad_norm": 0.3108993470668793, "learning_rate": 3.425067401723477e-05, "loss": 0.0268, "step": 6220 }, { "epoch": 24.054054054054053, "grad_norm": 0.17630359530448914, "learning_rate": 3.409383023633325e-05, "loss": 0.0229, "step": 6230 }, { "epoch": 24.092664092664094, "grad_norm": 0.32493704557418823, "learning_rate": 3.3937160402755894e-05, "loss": 0.0284, "step": 6240 }, { "epoch": 24.131274131274132, "grad_norm": 0.20445747673511505, "learning_rate": 3.378066622981885e-05, "loss": 0.0213, "step": 6250 }, { "epoch": 24.16988416988417, "grad_norm": 0.29104888439178467, "learning_rate": 3.362434942891738e-05, "loss": 0.026, "step": 6260 }, { "epoch": 24.20849420849421, "grad_norm": 0.2636573016643524, "learning_rate": 3.346821170950693e-05, "loss": 0.0274, "step": 6270 }, { "epoch": 24.247104247104247, "grad_norm": 0.31042179465293884, "learning_rate": 3.3312254779084585e-05, "loss": 0.0306, "step": 6280 }, { "epoch": 24.285714285714285, "grad_norm": 0.23835359513759613, "learning_rate": 3.315648034317039e-05, "loss": 0.0213, "step": 6290 }, { "epoch": 24.324324324324323, "grad_norm": 0.22583673894405365, "learning_rate": 3.3000890105288564e-05, "loss": 0.0281, "step": 6300 }, { "epoch": 24.362934362934364, "grad_norm": 0.2857336401939392, "learning_rate": 3.284548576694908e-05, "loss": 0.0197, "step": 6310 }, { "epoch": 24.401544401544403, "grad_norm": 0.2383348047733307, "learning_rate": 3.2690269027628815e-05, "loss": 0.0241, "step": 6320 }, { "epoch": 24.44015444015444, "grad_norm": 0.2590544819831848, "learning_rate": 3.253524158475324e-05, "loss": 0.0223, "step": 6330 }, { "epoch": 24.47876447876448, "grad_norm": 0.3190777897834778, "learning_rate": 3.238040513367757e-05, "loss": 0.023, "step": 6340 }, { "epoch": 24.517374517374517, "grad_norm": 0.25423505902290344, "learning_rate": 3.222576136766843e-05, "loss": 0.0209, "step": 6350 }, { "epoch": 24.555984555984555, "grad_norm": 0.2614216208457947, "learning_rate": 3.2071311977885324e-05, "loss": 0.0228, "step": 6360 }, { "epoch": 24.594594594594593, "grad_norm": 0.3432169556617737, "learning_rate": 3.191705865336197e-05, "loss": 0.0205, "step": 6370 }, { "epoch": 24.633204633204635, "grad_norm": 0.19699585437774658, "learning_rate": 3.1763003080988075e-05, "loss": 0.0217, "step": 6380 }, { "epoch": 24.671814671814673, "grad_norm": 0.21527795493602753, "learning_rate": 3.160914694549063e-05, "loss": 0.022, "step": 6390 }, { "epoch": 24.71042471042471, "grad_norm": 0.3318612575531006, "learning_rate": 3.145549192941573e-05, "loss": 0.0203, "step": 6400 }, { "epoch": 24.74903474903475, "grad_norm": 0.28894364833831787, "learning_rate": 3.130203971310999e-05, "loss": 0.0215, "step": 6410 }, { "epoch": 24.787644787644787, "grad_norm": 0.21171656250953674, "learning_rate": 3.114879197470225e-05, "loss": 0.0255, "step": 6420 }, { "epoch": 24.826254826254825, "grad_norm": 0.24591733515262604, "learning_rate": 3.0995750390085285e-05, "loss": 0.0234, "step": 6430 }, { "epoch": 24.864864864864863, "grad_norm": 0.24135924875736237, "learning_rate": 3.084291663289728e-05, "loss": 0.0216, "step": 6440 }, { "epoch": 24.903474903474905, "grad_norm": 0.3177614212036133, "learning_rate": 3.069029237450375e-05, "loss": 0.0202, "step": 6450 }, { "epoch": 24.942084942084943, "grad_norm": 0.2474350780248642, "learning_rate": 3.053787928397911e-05, "loss": 0.0242, "step": 6460 }, { "epoch": 24.98069498069498, "grad_norm": 0.37294405698776245, "learning_rate": 3.0385679028088526e-05, "loss": 0.0199, "step": 6470 }, { "epoch": 25.01930501930502, "grad_norm": 0.21189779043197632, "learning_rate": 3.023369327126959e-05, "loss": 0.0191, "step": 6480 }, { "epoch": 25.057915057915057, "grad_norm": 0.19981230795383453, "learning_rate": 3.0081923675614198e-05, "loss": 0.0281, "step": 6490 }, { "epoch": 25.096525096525095, "grad_norm": 0.3046863377094269, "learning_rate": 2.993037190085034e-05, "loss": 0.019, "step": 6500 }, { "epoch": 25.135135135135137, "grad_norm": 0.2717072367668152, "learning_rate": 2.977903960432392e-05, "loss": 0.0204, "step": 6510 }, { "epoch": 25.173745173745175, "grad_norm": 0.26114633679389954, "learning_rate": 2.9627928440980722e-05, "loss": 0.0168, "step": 6520 }, { "epoch": 25.212355212355213, "grad_norm": 0.2789161205291748, "learning_rate": 2.9477040063348183e-05, "loss": 0.0237, "step": 6530 }, { "epoch": 25.25096525096525, "grad_norm": 0.2511959671974182, "learning_rate": 2.9326376121517456e-05, "loss": 0.0186, "step": 6540 }, { "epoch": 25.28957528957529, "grad_norm": 0.163689523935318, "learning_rate": 2.9175938263125236e-05, "loss": 0.0185, "step": 6550 }, { "epoch": 25.328185328185327, "grad_norm": 0.23432201147079468, "learning_rate": 2.9025728133335873e-05, "loss": 0.018, "step": 6560 }, { "epoch": 25.366795366795365, "grad_norm": 0.2731872498989105, "learning_rate": 2.8875747374823288e-05, "loss": 0.0232, "step": 6570 }, { "epoch": 25.405405405405407, "grad_norm": 0.3436281085014343, "learning_rate": 2.872599762775298e-05, "loss": 0.025, "step": 6580 }, { "epoch": 25.444015444015445, "grad_norm": 0.3698206841945648, "learning_rate": 2.857648052976425e-05, "loss": 0.0296, "step": 6590 }, { "epoch": 25.482625482625483, "grad_norm": 0.26774466037750244, "learning_rate": 2.8427197715952047e-05, "loss": 0.0233, "step": 6600 }, { "epoch": 25.52123552123552, "grad_norm": 0.3142160475254059, "learning_rate": 2.8278150818849393e-05, "loss": 0.0225, "step": 6610 }, { "epoch": 25.55984555984556, "grad_norm": 0.34488993883132935, "learning_rate": 2.812934146840922e-05, "loss": 0.0208, "step": 6620 }, { "epoch": 25.598455598455597, "grad_norm": 0.22315387427806854, "learning_rate": 2.7980771291986764e-05, "loss": 0.0241, "step": 6630 }, { "epoch": 25.637065637065636, "grad_norm": 0.2939479351043701, "learning_rate": 2.783244191432167e-05, "loss": 0.0225, "step": 6640 }, { "epoch": 25.675675675675677, "grad_norm": 0.3250037431716919, "learning_rate": 2.768435495752022e-05, "loss": 0.0243, "step": 6650 }, { "epoch": 25.714285714285715, "grad_norm": 0.3358057737350464, "learning_rate": 2.753651204103771e-05, "loss": 0.024, "step": 6660 }, { "epoch": 25.752895752895753, "grad_norm": 0.2413211613893509, "learning_rate": 2.7388914781660523e-05, "loss": 0.017, "step": 6670 }, { "epoch": 25.79150579150579, "grad_norm": 0.24309025704860687, "learning_rate": 2.7241564793488693e-05, "loss": 0.021, "step": 6680 }, { "epoch": 25.83011583011583, "grad_norm": 0.2526884078979492, "learning_rate": 2.7094463687918037e-05, "loss": 0.0195, "step": 6690 }, { "epoch": 25.868725868725868, "grad_norm": 0.2860826551914215, "learning_rate": 2.694761307362268e-05, "loss": 0.0228, "step": 6700 }, { "epoch": 25.907335907335906, "grad_norm": 0.22474128007888794, "learning_rate": 2.6801014556537467e-05, "loss": 0.0244, "step": 6710 }, { "epoch": 25.945945945945947, "grad_norm": 0.180031418800354, "learning_rate": 2.6654669739840243e-05, "loss": 0.0206, "step": 6720 }, { "epoch": 25.984555984555985, "grad_norm": 0.2767072319984436, "learning_rate": 2.650858022393451e-05, "loss": 0.0217, "step": 6730 }, { "epoch": 26.023166023166024, "grad_norm": 0.23873543739318848, "learning_rate": 2.6362747606431747e-05, "loss": 0.0232, "step": 6740 }, { "epoch": 26.06177606177606, "grad_norm": 0.27126994729042053, "learning_rate": 2.6217173482134172e-05, "loss": 0.0231, "step": 6750 }, { "epoch": 26.1003861003861, "grad_norm": 0.2849481403827667, "learning_rate": 2.6071859443017044e-05, "loss": 0.0221, "step": 6760 }, { "epoch": 26.138996138996138, "grad_norm": 0.244388610124588, "learning_rate": 2.5926807078211414e-05, "loss": 0.0177, "step": 6770 }, { "epoch": 26.177606177606176, "grad_norm": 0.20935313403606415, "learning_rate": 2.5782017973986728e-05, "loss": 0.0193, "step": 6780 }, { "epoch": 26.216216216216218, "grad_norm": 0.22296316921710968, "learning_rate": 2.5637493713733374e-05, "loss": 0.0147, "step": 6790 }, { "epoch": 26.254826254826256, "grad_norm": 0.28912580013275146, "learning_rate": 2.549323587794559e-05, "loss": 0.0197, "step": 6800 }, { "epoch": 26.293436293436294, "grad_norm": 0.2849845588207245, "learning_rate": 2.5349246044203895e-05, "loss": 0.0197, "step": 6810 }, { "epoch": 26.332046332046332, "grad_norm": 0.2911399304866791, "learning_rate": 2.520552578715808e-05, "loss": 0.0267, "step": 6820 }, { "epoch": 26.37065637065637, "grad_norm": 0.30250683426856995, "learning_rate": 2.506207667850981e-05, "loss": 0.0159, "step": 6830 }, { "epoch": 26.409266409266408, "grad_norm": 0.15140190720558167, "learning_rate": 2.4918900286995555e-05, "loss": 0.0178, "step": 6840 }, { "epoch": 26.447876447876446, "grad_norm": 0.20052176713943481, "learning_rate": 2.4775998178369458e-05, "loss": 0.0169, "step": 6850 }, { "epoch": 26.486486486486488, "grad_norm": 0.28836214542388916, "learning_rate": 2.4633371915386017e-05, "loss": 0.0223, "step": 6860 }, { "epoch": 26.525096525096526, "grad_norm": 0.28271952271461487, "learning_rate": 2.4491023057783235e-05, "loss": 0.0189, "step": 6870 }, { "epoch": 26.563706563706564, "grad_norm": 0.24197126924991608, "learning_rate": 2.4348953162265375e-05, "loss": 0.0225, "step": 6880 }, { "epoch": 26.602316602316602, "grad_norm": 0.27072247862815857, "learning_rate": 2.420716378248607e-05, "loss": 0.0209, "step": 6890 }, { "epoch": 26.64092664092664, "grad_norm": 0.2721431851387024, "learning_rate": 2.4065656469031266e-05, "loss": 0.0262, "step": 6900 }, { "epoch": 26.679536679536678, "grad_norm": 0.22869786620140076, "learning_rate": 2.3924432769402268e-05, "loss": 0.0217, "step": 6910 }, { "epoch": 26.71814671814672, "grad_norm": 0.2811211347579956, "learning_rate": 2.3783494227998844e-05, "loss": 0.0219, "step": 6920 }, { "epoch": 26.756756756756758, "grad_norm": 0.30034950375556946, "learning_rate": 2.3642842386102264e-05, "loss": 0.0234, "step": 6930 }, { "epoch": 26.795366795366796, "grad_norm": 0.31990310549736023, "learning_rate": 2.3502478781858567e-05, "loss": 0.0285, "step": 6940 }, { "epoch": 26.833976833976834, "grad_norm": 0.2223096489906311, "learning_rate": 2.3362404950261628e-05, "loss": 0.0206, "step": 6950 }, { "epoch": 26.872586872586872, "grad_norm": 0.2684755027294159, "learning_rate": 2.3222622423136458e-05, "loss": 0.0207, "step": 6960 }, { "epoch": 26.91119691119691, "grad_norm": 0.19645178318023682, "learning_rate": 2.3083132729122332e-05, "loss": 0.0213, "step": 6970 }, { "epoch": 26.94980694980695, "grad_norm": 0.2216830849647522, "learning_rate": 2.294393739365621e-05, "loss": 0.0191, "step": 6980 }, { "epoch": 26.98841698841699, "grad_norm": 0.33184412121772766, "learning_rate": 2.2805037938956e-05, "loss": 0.0227, "step": 6990 }, { "epoch": 27.027027027027028, "grad_norm": 0.22032798826694489, "learning_rate": 2.266643588400386e-05, "loss": 0.0177, "step": 7000 }, { "epoch": 27.065637065637066, "grad_norm": 0.2350461781024933, "learning_rate": 2.252813274452969e-05, "loss": 0.0223, "step": 7010 }, { "epoch": 27.104247104247104, "grad_norm": 0.33744093775749207, "learning_rate": 2.2390130032994427e-05, "loss": 0.0209, "step": 7020 }, { "epoch": 27.142857142857142, "grad_norm": 0.17533046007156372, "learning_rate": 2.2252429258573633e-05, "loss": 0.0167, "step": 7030 }, { "epoch": 27.18146718146718, "grad_norm": 0.21129107475280762, "learning_rate": 2.2115031927140904e-05, "loss": 0.0179, "step": 7040 }, { "epoch": 27.22007722007722, "grad_norm": 0.30376043915748596, "learning_rate": 2.1977939541251463e-05, "loss": 0.0225, "step": 7050 }, { "epoch": 27.25868725868726, "grad_norm": 0.3463219106197357, "learning_rate": 2.1841153600125684e-05, "loss": 0.0206, "step": 7060 }, { "epoch": 27.2972972972973, "grad_norm": 0.35649845004081726, "learning_rate": 2.170467559963267e-05, "loss": 0.0246, "step": 7070 }, { "epoch": 27.335907335907336, "grad_norm": 0.4564409554004669, "learning_rate": 2.1568507032273982e-05, "loss": 0.0199, "step": 7080 }, { "epoch": 27.374517374517374, "grad_norm": 0.2515897750854492, "learning_rate": 2.1432649387167264e-05, "loss": 0.0179, "step": 7090 }, { "epoch": 27.413127413127413, "grad_norm": 0.23988781869411469, "learning_rate": 2.1297104150029973e-05, "loss": 0.0177, "step": 7100 }, { "epoch": 27.45173745173745, "grad_norm": 0.3372357487678528, "learning_rate": 2.116187280316307e-05, "loss": 0.0219, "step": 7110 }, { "epoch": 27.49034749034749, "grad_norm": 0.19336280226707458, "learning_rate": 2.1026956825434908e-05, "loss": 0.0216, "step": 7120 }, { "epoch": 27.52895752895753, "grad_norm": 0.19362467527389526, "learning_rate": 2.0892357692265017e-05, "loss": 0.0193, "step": 7130 }, { "epoch": 27.56756756756757, "grad_norm": 0.30222082138061523, "learning_rate": 2.0758076875607947e-05, "loss": 0.022, "step": 7140 }, { "epoch": 27.606177606177607, "grad_norm": 0.1858168989419937, "learning_rate": 2.0624115843937207e-05, "loss": 0.0204, "step": 7150 }, { "epoch": 27.644787644787645, "grad_norm": 0.2328641265630722, "learning_rate": 2.0490476062229157e-05, "loss": 0.0219, "step": 7160 }, { "epoch": 27.683397683397683, "grad_norm": 0.1741163283586502, "learning_rate": 2.035715899194704e-05, "loss": 0.023, "step": 7170 }, { "epoch": 27.72200772200772, "grad_norm": 0.27827292680740356, "learning_rate": 2.022416609102499e-05, "loss": 0.0199, "step": 7180 }, { "epoch": 27.760617760617762, "grad_norm": 0.1920846700668335, "learning_rate": 2.009149881385205e-05, "loss": 0.0176, "step": 7190 }, { "epoch": 27.7992277992278, "grad_norm": 0.2521073520183563, "learning_rate": 1.995915861125634e-05, "loss": 0.0229, "step": 7200 }, { "epoch": 27.83783783783784, "grad_norm": 0.23119695484638214, "learning_rate": 1.9827146930489065e-05, "loss": 0.0248, "step": 7210 }, { "epoch": 27.876447876447877, "grad_norm": 0.1516123116016388, "learning_rate": 1.9695465215208848e-05, "loss": 0.0187, "step": 7220 }, { "epoch": 27.915057915057915, "grad_norm": 0.2979636788368225, "learning_rate": 1.9564114905465813e-05, "loss": 0.015, "step": 7230 }, { "epoch": 27.953667953667953, "grad_norm": 0.19580204784870148, "learning_rate": 1.9433097437685936e-05, "loss": 0.0194, "step": 7240 }, { "epoch": 27.99227799227799, "grad_norm": 0.24929602444171906, "learning_rate": 1.930241424465521e-05, "loss": 0.0166, "step": 7250 }, { "epoch": 28.030888030888033, "grad_norm": 0.4262058436870575, "learning_rate": 1.9172066755504115e-05, "loss": 0.0249, "step": 7260 }, { "epoch": 28.06949806949807, "grad_norm": 0.1503635048866272, "learning_rate": 1.9042056395691914e-05, "loss": 0.0152, "step": 7270 }, { "epoch": 28.10810810810811, "grad_norm": 0.2532370984554291, "learning_rate": 1.8912384586991066e-05, "loss": 0.0182, "step": 7280 }, { "epoch": 28.146718146718147, "grad_norm": 0.28780001401901245, "learning_rate": 1.8783052747471717e-05, "loss": 0.0214, "step": 7290 }, { "epoch": 28.185328185328185, "grad_norm": 0.4166366755962372, "learning_rate": 1.865406229148611e-05, "loss": 0.0179, "step": 7300 }, { "epoch": 28.223938223938223, "grad_norm": 0.18886317312717438, "learning_rate": 1.8525414629653233e-05, "loss": 0.0194, "step": 7310 }, { "epoch": 28.26254826254826, "grad_norm": 0.2156158983707428, "learning_rate": 1.8397111168843255e-05, "loss": 0.0227, "step": 7320 }, { "epoch": 28.301158301158303, "grad_norm": 0.22111840546131134, "learning_rate": 1.8269153312162323e-05, "loss": 0.0224, "step": 7330 }, { "epoch": 28.33976833976834, "grad_norm": 0.20181381702423096, "learning_rate": 1.8141542458937054e-05, "loss": 0.0193, "step": 7340 }, { "epoch": 28.37837837837838, "grad_norm": 0.2345784306526184, "learning_rate": 1.8014280004699268e-05, "loss": 0.0202, "step": 7350 }, { "epoch": 28.416988416988417, "grad_norm": 0.32246366143226624, "learning_rate": 1.788736734117078e-05, "loss": 0.0254, "step": 7360 }, { "epoch": 28.455598455598455, "grad_norm": 0.2772466540336609, "learning_rate": 1.7760805856248152e-05, "loss": 0.0175, "step": 7370 }, { "epoch": 28.494208494208493, "grad_norm": 0.4173099994659424, "learning_rate": 1.7634596933987518e-05, "loss": 0.0193, "step": 7380 }, { "epoch": 28.53281853281853, "grad_norm": 0.22678948938846588, "learning_rate": 1.7508741954589404e-05, "loss": 0.0189, "step": 7390 }, { "epoch": 28.571428571428573, "grad_norm": 0.23506483435630798, "learning_rate": 1.7383242294383717e-05, "loss": 0.0183, "step": 7400 }, { "epoch": 28.61003861003861, "grad_norm": 0.12952037155628204, "learning_rate": 1.7258099325814632e-05, "loss": 0.0222, "step": 7410 }, { "epoch": 28.64864864864865, "grad_norm": 0.11573129892349243, "learning_rate": 1.7133314417425594e-05, "loss": 0.0199, "step": 7420 }, { "epoch": 28.687258687258687, "grad_norm": 0.1827162206172943, "learning_rate": 1.7008888933844408e-05, "loss": 0.0181, "step": 7430 }, { "epoch": 28.725868725868725, "grad_norm": 0.18070828914642334, "learning_rate": 1.6884824235768172e-05, "loss": 0.0243, "step": 7440 }, { "epoch": 28.764478764478763, "grad_norm": 0.3690528869628906, "learning_rate": 1.6761121679948592e-05, "loss": 0.0215, "step": 7450 }, { "epoch": 28.8030888030888, "grad_norm": 0.25985127687454224, "learning_rate": 1.663778261917695e-05, "loss": 0.0174, "step": 7460 }, { "epoch": 28.841698841698843, "grad_norm": 0.2993326485157013, "learning_rate": 1.651480840226952e-05, "loss": 0.0165, "step": 7470 }, { "epoch": 28.88030888030888, "grad_norm": 0.21274708211421967, "learning_rate": 1.639220037405258e-05, "loss": 0.0184, "step": 7480 }, { "epoch": 28.91891891891892, "grad_norm": 0.2361726462841034, "learning_rate": 1.6269959875347906e-05, "loss": 0.0187, "step": 7490 }, { "epoch": 28.957528957528957, "grad_norm": 0.3593890070915222, "learning_rate": 1.614808824295802e-05, "loss": 0.0189, "step": 7500 }, { "epoch": 28.996138996138995, "grad_norm": 0.18137873709201813, "learning_rate": 1.602658680965152e-05, "loss": 0.0199, "step": 7510 }, { "epoch": 29.034749034749034, "grad_norm": 0.2655872404575348, "learning_rate": 1.5905456904148686e-05, "loss": 0.0164, "step": 7520 }, { "epoch": 29.07335907335907, "grad_norm": 0.29168081283569336, "learning_rate": 1.57846998511067e-05, "loss": 0.0186, "step": 7530 }, { "epoch": 29.111969111969113, "grad_norm": 0.2996509075164795, "learning_rate": 1.566431697110538e-05, "loss": 0.0184, "step": 7540 }, { "epoch": 29.15057915057915, "grad_norm": 0.2727248966693878, "learning_rate": 1.554430958063259e-05, "loss": 0.0211, "step": 7550 }, { "epoch": 29.18918918918919, "grad_norm": 0.36617186665534973, "learning_rate": 1.5424678992069912e-05, "loss": 0.0189, "step": 7560 }, { "epoch": 29.227799227799228, "grad_norm": 0.19298312067985535, "learning_rate": 1.5305426513678362e-05, "loss": 0.0141, "step": 7570 }, { "epoch": 29.266409266409266, "grad_norm": 0.2666356563568115, "learning_rate": 1.518655344958388e-05, "loss": 0.018, "step": 7580 }, { "epoch": 29.305019305019304, "grad_norm": 0.35917577147483826, "learning_rate": 1.5068061099763275e-05, "loss": 0.0212, "step": 7590 }, { "epoch": 29.343629343629345, "grad_norm": 0.3404170572757721, "learning_rate": 1.494995076002988e-05, "loss": 0.021, "step": 7600 }, { "epoch": 29.382239382239383, "grad_norm": 0.25935497879981995, "learning_rate": 1.4832223722019456e-05, "loss": 0.0148, "step": 7610 }, { "epoch": 29.42084942084942, "grad_norm": 0.26347196102142334, "learning_rate": 1.4714881273176035e-05, "loss": 0.018, "step": 7620 }, { "epoch": 29.45945945945946, "grad_norm": 0.12636244297027588, "learning_rate": 1.4597924696737835e-05, "loss": 0.0127, "step": 7630 }, { "epoch": 29.498069498069498, "grad_norm": 0.21686862409114838, "learning_rate": 1.4481355271723252e-05, "loss": 0.018, "step": 7640 }, { "epoch": 29.536679536679536, "grad_norm": 0.18001554906368256, "learning_rate": 1.4365174272916809e-05, "loss": 0.0168, "step": 7650 }, { "epoch": 29.575289575289574, "grad_norm": 0.22990286350250244, "learning_rate": 1.4249382970855319e-05, "loss": 0.0157, "step": 7660 }, { "epoch": 29.613899613899616, "grad_norm": 0.2203466296195984, "learning_rate": 1.4133982631813903e-05, "loss": 0.0197, "step": 7670 }, { "epoch": 29.652509652509654, "grad_norm": 0.18794023990631104, "learning_rate": 1.4018974517792194e-05, "loss": 0.0157, "step": 7680 }, { "epoch": 29.69111969111969, "grad_norm": 0.23368437588214874, "learning_rate": 1.390435988650048e-05, "loss": 0.0177, "step": 7690 }, { "epoch": 29.72972972972973, "grad_norm": 0.19165118038654327, "learning_rate": 1.3790139991346006e-05, "loss": 0.0175, "step": 7700 }, { "epoch": 29.768339768339768, "grad_norm": 0.2268296629190445, "learning_rate": 1.367631608141926e-05, "loss": 0.0181, "step": 7710 }, { "epoch": 29.806949806949806, "grad_norm": 0.2143620252609253, "learning_rate": 1.3562889401480278e-05, "loss": 0.0165, "step": 7720 }, { "epoch": 29.845559845559844, "grad_norm": 0.18440692126750946, "learning_rate": 1.3449861191945074e-05, "loss": 0.0169, "step": 7730 }, { "epoch": 29.884169884169886, "grad_norm": 0.13429401814937592, "learning_rate": 1.3337232688872009e-05, "loss": 0.0157, "step": 7740 }, { "epoch": 29.922779922779924, "grad_norm": 0.2586115598678589, "learning_rate": 1.3225005123948364e-05, "loss": 0.0182, "step": 7750 }, { "epoch": 29.961389961389962, "grad_norm": 0.18588311970233917, "learning_rate": 1.311317972447681e-05, "loss": 0.0161, "step": 7760 }, { "epoch": 30.0, "grad_norm": 0.4120975434780121, "learning_rate": 1.3001757713361996e-05, "loss": 0.0215, "step": 7770 }, { "epoch": 30.038610038610038, "grad_norm": 0.30354854464530945, "learning_rate": 1.2890740309097204e-05, "loss": 0.0166, "step": 7780 }, { "epoch": 30.077220077220076, "grad_norm": 0.29400011897087097, "learning_rate": 1.2780128725750944e-05, "loss": 0.0154, "step": 7790 }, { "epoch": 30.115830115830114, "grad_norm": 0.1959076225757599, "learning_rate": 1.266992417295379e-05, "loss": 0.0188, "step": 7800 }, { "epoch": 30.154440154440156, "grad_norm": 0.16418759524822235, "learning_rate": 1.2560127855885073e-05, "loss": 0.0164, "step": 7810 }, { "epoch": 30.193050193050194, "grad_norm": 0.1775139570236206, "learning_rate": 1.2450740975259745e-05, "loss": 0.0163, "step": 7820 }, { "epoch": 30.231660231660232, "grad_norm": 0.24529078602790833, "learning_rate": 1.234176472731517e-05, "loss": 0.0183, "step": 7830 }, { "epoch": 30.27027027027027, "grad_norm": 0.1190020740032196, "learning_rate": 1.2233200303798158e-05, "loss": 0.0196, "step": 7840 }, { "epoch": 30.30888030888031, "grad_norm": 0.17785975337028503, "learning_rate": 1.2125048891951846e-05, "loss": 0.0267, "step": 7850 }, { "epoch": 30.347490347490346, "grad_norm": 0.16408364474773407, "learning_rate": 1.2017311674502745e-05, "loss": 0.0138, "step": 7860 }, { "epoch": 30.386100386100384, "grad_norm": 0.42137470841407776, "learning_rate": 1.1909989829647822e-05, "loss": 0.0205, "step": 7870 }, { "epoch": 30.424710424710426, "grad_norm": 0.13139937818050385, "learning_rate": 1.1803084531041553e-05, "loss": 0.0139, "step": 7880 }, { "epoch": 30.463320463320464, "grad_norm": 0.1757706254720688, "learning_rate": 1.1696596947783162e-05, "loss": 0.016, "step": 7890 }, { "epoch": 30.501930501930502, "grad_norm": 0.2886413037776947, "learning_rate": 1.1590528244403803e-05, "loss": 0.0201, "step": 7900 }, { "epoch": 30.54054054054054, "grad_norm": 0.1935712844133377, "learning_rate": 1.148487958085382e-05, "loss": 0.0174, "step": 7910 }, { "epoch": 30.57915057915058, "grad_norm": 0.15784715116024017, "learning_rate": 1.1379652112490086e-05, "loss": 0.017, "step": 7920 }, { "epoch": 30.617760617760617, "grad_norm": 0.1632302850484848, "learning_rate": 1.1274846990063315e-05, "loss": 0.0184, "step": 7930 }, { "epoch": 30.656370656370655, "grad_norm": 0.30083027482032776, "learning_rate": 1.117046535970554e-05, "loss": 0.0184, "step": 7940 }, { "epoch": 30.694980694980696, "grad_norm": 0.32924333214759827, "learning_rate": 1.106650836291755e-05, "loss": 0.0153, "step": 7950 }, { "epoch": 30.733590733590734, "grad_norm": 0.28148210048675537, "learning_rate": 1.0962977136556418e-05, "loss": 0.0178, "step": 7960 }, { "epoch": 30.772200772200772, "grad_norm": 0.1910584419965744, "learning_rate": 1.0859872812823024e-05, "loss": 0.0232, "step": 7970 }, { "epoch": 30.81081081081081, "grad_norm": 0.3276840150356293, "learning_rate": 1.0757196519249747e-05, "loss": 0.0213, "step": 7980 }, { "epoch": 30.84942084942085, "grad_norm": 0.22837305068969727, "learning_rate": 1.0654949378688077e-05, "loss": 0.0195, "step": 7990 }, { "epoch": 30.888030888030887, "grad_norm": 0.24823689460754395, "learning_rate": 1.0553132509296376e-05, "loss": 0.0217, "step": 8000 }, { "epoch": 30.92664092664093, "grad_norm": 0.1904963254928589, "learning_rate": 1.0451747024527613e-05, "loss": 0.0187, "step": 8010 }, { "epoch": 30.965250965250966, "grad_norm": 0.14187836647033691, "learning_rate": 1.0350794033117189e-05, "loss": 0.0233, "step": 8020 }, { "epoch": 31.003861003861005, "grad_norm": 0.2642929255962372, "learning_rate": 1.0250274639070856e-05, "loss": 0.0166, "step": 8030 }, { "epoch": 31.042471042471043, "grad_norm": 0.25369855761528015, "learning_rate": 1.0150189941652599e-05, "loss": 0.018, "step": 8040 }, { "epoch": 31.08108108108108, "grad_norm": 0.2564759850502014, "learning_rate": 1.0050541035372635e-05, "loss": 0.019, "step": 8050 }, { "epoch": 31.11969111969112, "grad_norm": 0.17894713580608368, "learning_rate": 9.951329009975458e-06, "loss": 0.017, "step": 8060 }, { "epoch": 31.158301158301157, "grad_norm": 0.1546516716480255, "learning_rate": 9.852554950427845e-06, "loss": 0.0116, "step": 8070 }, { "epoch": 31.1969111969112, "grad_norm": 0.4103129804134369, "learning_rate": 9.754219936907105e-06, "loss": 0.0211, "step": 8080 }, { "epoch": 31.235521235521237, "grad_norm": 0.192494735121727, "learning_rate": 9.656325044789194e-06, "loss": 0.022, "step": 8090 }, { "epoch": 31.274131274131275, "grad_norm": 0.24065561592578888, "learning_rate": 9.55887134463697e-06, "loss": 0.0188, "step": 8100 }, { "epoch": 31.312741312741313, "grad_norm": 0.29938822984695435, "learning_rate": 9.461859902188475e-06, "loss": 0.0156, "step": 8110 }, { "epoch": 31.35135135135135, "grad_norm": 0.2863124907016754, "learning_rate": 9.365291778345303e-06, "loss": 0.0143, "step": 8120 }, { "epoch": 31.38996138996139, "grad_norm": 0.40058434009552, "learning_rate": 9.269168029160991e-06, "loss": 0.0172, "step": 8130 }, { "epoch": 31.428571428571427, "grad_norm": 0.25773149728775024, "learning_rate": 9.173489705829447e-06, "loss": 0.0164, "step": 8140 }, { "epoch": 31.46718146718147, "grad_norm": 0.1472177654504776, "learning_rate": 9.078257854673516e-06, "loss": 0.0112, "step": 8150 }, { "epoch": 31.505791505791507, "grad_norm": 0.3204963803291321, "learning_rate": 8.983473517133429e-06, "loss": 0.0206, "step": 8160 }, { "epoch": 31.544401544401545, "grad_norm": 0.17740993201732635, "learning_rate": 8.889137729755537e-06, "loss": 0.0199, "step": 8170 }, { "epoch": 31.583011583011583, "grad_norm": 0.13243339955806732, "learning_rate": 8.79525152418087e-06, "loss": 0.0186, "step": 8180 }, { "epoch": 31.62162162162162, "grad_norm": 0.13363532721996307, "learning_rate": 8.701815927133961e-06, "loss": 0.0162, "step": 8190 }, { "epoch": 31.66023166023166, "grad_norm": 0.19426418840885162, "learning_rate": 8.608831960411534e-06, "loss": 0.0175, "step": 8200 }, { "epoch": 31.698841698841697, "grad_norm": 0.3544776141643524, "learning_rate": 8.516300640871321e-06, "loss": 0.016, "step": 8210 }, { "epoch": 31.73745173745174, "grad_norm": 0.11085174977779388, "learning_rate": 8.424222980421038e-06, "loss": 0.0136, "step": 8220 }, { "epoch": 31.776061776061777, "grad_norm": 0.21357914805412292, "learning_rate": 8.332599986007184e-06, "loss": 0.0168, "step": 8230 }, { "epoch": 31.814671814671815, "grad_norm": 0.10666314512491226, "learning_rate": 8.241432659604203e-06, "loss": 0.0165, "step": 8240 }, { "epoch": 31.853281853281853, "grad_norm": 0.25809913873672485, "learning_rate": 8.150721998203331e-06, "loss": 0.0225, "step": 8250 }, { "epoch": 31.89189189189189, "grad_norm": 0.23415352404117584, "learning_rate": 8.06046899380184e-06, "loss": 0.0137, "step": 8260 }, { "epoch": 31.93050193050193, "grad_norm": 0.15521034598350525, "learning_rate": 7.970674633392133e-06, "loss": 0.0149, "step": 8270 }, { "epoch": 31.969111969111967, "grad_norm": 0.18646027147769928, "learning_rate": 7.881339898950924e-06, "loss": 0.0148, "step": 8280 }, { "epoch": 32.00772200772201, "grad_norm": 0.33421990275382996, "learning_rate": 7.792465767428597e-06, "loss": 0.0219, "step": 8290 }, { "epoch": 32.04633204633205, "grad_norm": 0.16285410523414612, "learning_rate": 7.704053210738376e-06, "loss": 0.0205, "step": 8300 }, { "epoch": 32.084942084942085, "grad_norm": 0.17426452040672302, "learning_rate": 7.6161031957458494e-06, "loss": 0.0136, "step": 8310 }, { "epoch": 32.12355212355212, "grad_norm": 0.1769731640815735, "learning_rate": 7.5286166842582605e-06, "loss": 0.013, "step": 8320 }, { "epoch": 32.16216216216216, "grad_norm": 0.2457035481929779, "learning_rate": 7.4415946330140814e-06, "loss": 0.0192, "step": 8330 }, { "epoch": 32.2007722007722, "grad_norm": 0.2060357928276062, "learning_rate": 7.3550379936725644e-06, "loss": 0.0195, "step": 8340 }, { "epoch": 32.23938223938224, "grad_norm": 0.26165908575057983, "learning_rate": 7.2689477128032035e-06, "loss": 0.0196, "step": 8350 }, { "epoch": 32.277992277992276, "grad_norm": 0.20386703312397003, "learning_rate": 7.183324731875551e-06, "loss": 0.0126, "step": 8360 }, { "epoch": 32.316602316602314, "grad_norm": 0.22269505262374878, "learning_rate": 7.098169987248782e-06, "loss": 0.0166, "step": 8370 }, { "epoch": 32.35521235521235, "grad_norm": 0.2452581226825714, "learning_rate": 7.013484410161553e-06, "loss": 0.0112, "step": 8380 }, { "epoch": 32.3938223938224, "grad_norm": 0.18786628544330597, "learning_rate": 6.92926892672176e-06, "loss": 0.0159, "step": 8390 }, { "epoch": 32.432432432432435, "grad_norm": 0.2967357635498047, "learning_rate": 6.845524457896446e-06, "loss": 0.0156, "step": 8400 }, { "epoch": 32.47104247104247, "grad_norm": 0.24428898096084595, "learning_rate": 6.7622519195017165e-06, "loss": 0.0133, "step": 8410 }, { "epoch": 32.50965250965251, "grad_norm": 0.296485960483551, "learning_rate": 6.679452222192684e-06, "loss": 0.0173, "step": 8420 }, { "epoch": 32.54826254826255, "grad_norm": 0.16247673332691193, "learning_rate": 6.597126271453579e-06, "loss": 0.0198, "step": 8430 }, { "epoch": 32.58687258687259, "grad_norm": 0.2788502871990204, "learning_rate": 6.51527496758782e-06, "loss": 0.0145, "step": 8440 }, { "epoch": 32.625482625482626, "grad_norm": 0.19357165694236755, "learning_rate": 6.433899205708155e-06, "loss": 0.0149, "step": 8450 }, { "epoch": 32.664092664092664, "grad_norm": 0.17584428191184998, "learning_rate": 6.352999875726856e-06, "loss": 0.0156, "step": 8460 }, { "epoch": 32.7027027027027, "grad_norm": 0.2467210441827774, "learning_rate": 6.272577862346052e-06, "loss": 0.0159, "step": 8470 }, { "epoch": 32.74131274131274, "grad_norm": 0.36892810463905334, "learning_rate": 6.192634045047996e-06, "loss": 0.0157, "step": 8480 }, { "epoch": 32.77992277992278, "grad_norm": 0.24180449545383453, "learning_rate": 6.113169298085458e-06, "loss": 0.0121, "step": 8490 }, { "epoch": 32.818532818532816, "grad_norm": 0.2090541273355484, "learning_rate": 6.034184490472195e-06, "loss": 0.0157, "step": 8500 }, { "epoch": 32.857142857142854, "grad_norm": 0.16316676139831543, "learning_rate": 5.955680485973386e-06, "loss": 0.0254, "step": 8510 }, { "epoch": 32.89575289575289, "grad_norm": 0.14482209086418152, "learning_rate": 5.877658143096265e-06, "loss": 0.0165, "step": 8520 }, { "epoch": 32.93436293436294, "grad_norm": 0.14665259420871735, "learning_rate": 5.800118315080661e-06, "loss": 0.0152, "step": 8530 }, { "epoch": 32.972972972972975, "grad_norm": 0.3260785639286041, "learning_rate": 5.723061849889716e-06, "loss": 0.0171, "step": 8540 }, { "epoch": 33.011583011583014, "grad_norm": 0.17532190680503845, "learning_rate": 5.646489590200604e-06, "loss": 0.0135, "step": 8550 }, { "epoch": 33.05019305019305, "grad_norm": 0.11559351533651352, "learning_rate": 5.570402373395256e-06, "loss": 0.0129, "step": 8560 }, { "epoch": 33.08880308880309, "grad_norm": 0.32188138365745544, "learning_rate": 5.494801031551305e-06, "loss": 0.02, "step": 8570 }, { "epoch": 33.12741312741313, "grad_norm": 0.27871063351631165, "learning_rate": 5.41968639143291e-06, "loss": 0.0147, "step": 8580 }, { "epoch": 33.166023166023166, "grad_norm": 0.2064131200313568, "learning_rate": 5.345059274481751e-06, "loss": 0.0135, "step": 8590 }, { "epoch": 33.204633204633204, "grad_norm": 0.2294102907180786, "learning_rate": 5.270920496808002e-06, "loss": 0.0198, "step": 8600 }, { "epoch": 33.24324324324324, "grad_norm": 0.33526143431663513, "learning_rate": 5.1972708691814695e-06, "loss": 0.0158, "step": 8610 }, { "epoch": 33.28185328185328, "grad_norm": 0.257347047328949, "learning_rate": 5.124111197022674e-06, "loss": 0.0205, "step": 8620 }, { "epoch": 33.32046332046332, "grad_norm": 0.24532243609428406, "learning_rate": 5.051442280394081e-06, "loss": 0.0157, "step": 8630 }, { "epoch": 33.359073359073356, "grad_norm": 0.199360653758049, "learning_rate": 4.979264913991322e-06, "loss": 0.0179, "step": 8640 }, { "epoch": 33.397683397683394, "grad_norm": 0.1698353886604309, "learning_rate": 4.907579887134489e-06, "loss": 0.0147, "step": 8650 }, { "epoch": 33.43629343629344, "grad_norm": 0.24843882024288177, "learning_rate": 4.836387983759572e-06, "loss": 0.0172, "step": 8660 }, { "epoch": 33.47490347490348, "grad_norm": 0.20173601806163788, "learning_rate": 4.765689982409816e-06, "loss": 0.0181, "step": 8670 }, { "epoch": 33.513513513513516, "grad_norm": 0.1983555257320404, "learning_rate": 4.695486656227233e-06, "loss": 0.0153, "step": 8680 }, { "epoch": 33.552123552123554, "grad_norm": 0.23845815658569336, "learning_rate": 4.625778772944156e-06, "loss": 0.0147, "step": 8690 }, { "epoch": 33.59073359073359, "grad_norm": 0.22628502547740936, "learning_rate": 4.556567094874825e-06, "loss": 0.0146, "step": 8700 }, { "epoch": 33.62934362934363, "grad_norm": 0.34669455885887146, "learning_rate": 4.487852378907059e-06, "loss": 0.0172, "step": 8710 }, { "epoch": 33.66795366795367, "grad_norm": 0.1536358892917633, "learning_rate": 4.419635376493986e-06, "loss": 0.0143, "step": 8720 }, { "epoch": 33.706563706563706, "grad_norm": 0.18267276883125305, "learning_rate": 4.351916833645825e-06, "loss": 0.0151, "step": 8730 }, { "epoch": 33.745173745173744, "grad_norm": 0.19443340599536896, "learning_rate": 4.284697490921691e-06, "loss": 0.0158, "step": 8740 }, { "epoch": 33.78378378378378, "grad_norm": 0.23747295141220093, "learning_rate": 4.2179780834215585e-06, "loss": 0.0157, "step": 8750 }, { "epoch": 33.82239382239382, "grad_norm": 0.19157427549362183, "learning_rate": 4.151759340778178e-06, "loss": 0.0152, "step": 8760 }, { "epoch": 33.86100386100386, "grad_norm": 0.24690277874469757, "learning_rate": 4.086041987149109e-06, "loss": 0.0157, "step": 8770 }, { "epoch": 33.8996138996139, "grad_norm": 0.543507993221283, "learning_rate": 4.020826741208811e-06, "loss": 0.0178, "step": 8780 }, { "epoch": 33.938223938223935, "grad_norm": 0.15995490550994873, "learning_rate": 3.956114316140746e-06, "loss": 0.0159, "step": 8790 }, { "epoch": 33.97683397683398, "grad_norm": 0.12463945895433426, "learning_rate": 3.891905419629643e-06, "loss": 0.0146, "step": 8800 }, { "epoch": 34.01544401544402, "grad_norm": 0.23251637816429138, "learning_rate": 3.8282007538536946e-06, "loss": 0.0204, "step": 8810 }, { "epoch": 34.054054054054056, "grad_norm": 0.2535611689090729, "learning_rate": 3.7650010154769265e-06, "loss": 0.0179, "step": 8820 }, { "epoch": 34.092664092664094, "grad_norm": 0.11479510366916656, "learning_rate": 3.7023068956415608e-06, "loss": 0.0161, "step": 8830 }, { "epoch": 34.13127413127413, "grad_norm": 0.1411840319633484, "learning_rate": 3.6401190799604303e-06, "loss": 0.0112, "step": 8840 }, { "epoch": 34.16988416988417, "grad_norm": 0.14113600552082062, "learning_rate": 3.578438248509536e-06, "loss": 0.0159, "step": 8850 }, { "epoch": 34.20849420849421, "grad_norm": 0.11504536122083664, "learning_rate": 3.5172650758205583e-06, "loss": 0.0163, "step": 8860 }, { "epoch": 34.24710424710425, "grad_norm": 0.2882764935493469, "learning_rate": 3.45660023087353e-06, "loss": 0.0187, "step": 8870 }, { "epoch": 34.285714285714285, "grad_norm": 0.31600257754325867, "learning_rate": 3.3964443770894528e-06, "loss": 0.0209, "step": 8880 }, { "epoch": 34.32432432432432, "grad_norm": 0.11367549002170563, "learning_rate": 3.3367981723231245e-06, "loss": 0.0141, "step": 8890 }, { "epoch": 34.36293436293436, "grad_norm": 0.22795264422893524, "learning_rate": 3.2776622688558746e-06, "loss": 0.0172, "step": 8900 }, { "epoch": 34.4015444015444, "grad_norm": 0.16641278564929962, "learning_rate": 3.2190373133884677e-06, "loss": 0.0152, "step": 8910 }, { "epoch": 34.44015444015444, "grad_norm": 0.21200251579284668, "learning_rate": 3.1609239470340446e-06, "loss": 0.0135, "step": 8920 }, { "epoch": 34.47876447876448, "grad_norm": 0.26414111256599426, "learning_rate": 3.1033228053110373e-06, "loss": 0.0212, "step": 8930 }, { "epoch": 34.51737451737452, "grad_norm": 0.2669637203216553, "learning_rate": 3.0462345181363314e-06, "loss": 0.0161, "step": 8940 }, { "epoch": 34.55598455598456, "grad_norm": 0.23096701502799988, "learning_rate": 2.9896597098182654e-06, "loss": 0.0124, "step": 8950 }, { "epoch": 34.5945945945946, "grad_norm": 0.18518953025341034, "learning_rate": 2.933598999049891e-06, "loss": 0.0135, "step": 8960 }, { "epoch": 34.633204633204635, "grad_norm": 0.2118380069732666, "learning_rate": 2.8780529989021697e-06, "loss": 0.0163, "step": 8970 }, { "epoch": 34.67181467181467, "grad_norm": 0.23276583850383759, "learning_rate": 2.823022316817242e-06, "loss": 0.0141, "step": 8980 }, { "epoch": 34.71042471042471, "grad_norm": 0.20471914112567902, "learning_rate": 2.7685075546018456e-06, "loss": 0.0155, "step": 8990 }, { "epoch": 34.74903474903475, "grad_norm": 0.1357901245355606, "learning_rate": 2.7145093084206598e-06, "loss": 0.0156, "step": 9000 }, { "epoch": 34.78764478764479, "grad_norm": 0.25041064620018005, "learning_rate": 2.661028168789892e-06, "loss": 0.0247, "step": 9010 }, { "epoch": 34.826254826254825, "grad_norm": 0.24415242671966553, "learning_rate": 2.6080647205706855e-06, "loss": 0.0181, "step": 9020 }, { "epoch": 34.86486486486486, "grad_norm": 0.23001091182231903, "learning_rate": 2.555619542962834e-06, "loss": 0.0153, "step": 9030 }, { "epoch": 34.9034749034749, "grad_norm": 0.15836270153522491, "learning_rate": 2.503693209498409e-06, "loss": 0.0134, "step": 9040 }, { "epoch": 34.94208494208494, "grad_norm": 0.2661552131175995, "learning_rate": 2.452286288035449e-06, "loss": 0.0157, "step": 9050 }, { "epoch": 34.98069498069498, "grad_norm": 0.13560421764850616, "learning_rate": 2.4013993407518363e-06, "loss": 0.0176, "step": 9060 }, { "epoch": 35.01930501930502, "grad_norm": 0.15679091215133667, "learning_rate": 2.351032924139063e-06, "loss": 0.014, "step": 9070 }, { "epoch": 35.05791505791506, "grad_norm": 0.19196201860904694, "learning_rate": 2.30118758899619e-06, "loss": 0.0156, "step": 9080 }, { "epoch": 35.0965250965251, "grad_norm": 0.29018867015838623, "learning_rate": 2.2518638804238157e-06, "loss": 0.0121, "step": 9090 }, { "epoch": 35.13513513513514, "grad_norm": 0.24716682732105255, "learning_rate": 2.203062337818118e-06, "loss": 0.0175, "step": 9100 }, { "epoch": 35.173745173745175, "grad_norm": 0.1430145651102066, "learning_rate": 2.1547834948649483e-06, "loss": 0.0122, "step": 9110 }, { "epoch": 35.21235521235521, "grad_norm": 0.19398616254329681, "learning_rate": 2.1070278795340017e-06, "loss": 0.0192, "step": 9120 }, { "epoch": 35.25096525096525, "grad_norm": 0.16792966425418854, "learning_rate": 2.059796014073029e-06, "loss": 0.0164, "step": 9130 }, { "epoch": 35.28957528957529, "grad_norm": 0.1834522932767868, "learning_rate": 2.01308841500214e-06, "loss": 0.0211, "step": 9140 }, { "epoch": 35.32818532818533, "grad_norm": 0.192967489361763, "learning_rate": 1.9669055931081704e-06, "loss": 0.0146, "step": 9150 }, { "epoch": 35.366795366795365, "grad_norm": 0.24933141469955444, "learning_rate": 1.9212480534390507e-06, "loss": 0.0131, "step": 9160 }, { "epoch": 35.4054054054054, "grad_norm": 0.32256147265434265, "learning_rate": 1.8761162952983246e-06, "loss": 0.017, "step": 9170 }, { "epoch": 35.44401544401544, "grad_norm": 0.20442728698253632, "learning_rate": 1.8315108122396618e-06, "loss": 0.0144, "step": 9180 }, { "epoch": 35.48262548262548, "grad_norm": 0.17914916574954987, "learning_rate": 1.787432092061475e-06, "loss": 0.0137, "step": 9190 }, { "epoch": 35.52123552123552, "grad_norm": 0.20625437796115875, "learning_rate": 1.743880616801602e-06, "loss": 0.0204, "step": 9200 }, { "epoch": 35.55984555984556, "grad_norm": 0.15013466775417328, "learning_rate": 1.7008568627319865e-06, "loss": 0.0156, "step": 9210 }, { "epoch": 35.5984555984556, "grad_norm": 0.21691270172595978, "learning_rate": 1.6583613003535226e-06, "loss": 0.0139, "step": 9220 }, { "epoch": 35.63706563706564, "grad_norm": 0.304378479719162, "learning_rate": 1.6163943943908522e-06, "loss": 0.0173, "step": 9230 }, { "epoch": 35.67567567567568, "grad_norm": 0.22544310986995697, "learning_rate": 1.5749566037873476e-06, "loss": 0.0179, "step": 9240 }, { "epoch": 35.714285714285715, "grad_norm": 0.22566020488739014, "learning_rate": 1.5340483817000428e-06, "loss": 0.019, "step": 9250 }, { "epoch": 35.75289575289575, "grad_norm": 0.2109433114528656, "learning_rate": 1.4936701754947101e-06, "loss": 0.0166, "step": 9260 }, { "epoch": 35.79150579150579, "grad_norm": 0.3437354266643524, "learning_rate": 1.4538224267409361e-06, "loss": 0.0174, "step": 9270 }, { "epoch": 35.83011583011583, "grad_norm": 0.18232221901416779, "learning_rate": 1.414505571207314e-06, "loss": 0.0151, "step": 9280 }, { "epoch": 35.86872586872587, "grad_norm": 0.2294638305902481, "learning_rate": 1.3757200388566816e-06, "loss": 0.0141, "step": 9290 }, { "epoch": 35.907335907335906, "grad_norm": 0.24555842578411102, "learning_rate": 1.3374662538414074e-06, "loss": 0.0133, "step": 9300 }, { "epoch": 35.945945945945944, "grad_norm": 0.23130197823047638, "learning_rate": 1.2997446344987617e-06, "loss": 0.0166, "step": 9310 }, { "epoch": 35.98455598455598, "grad_norm": 0.26748335361480713, "learning_rate": 1.262555593346315e-06, "loss": 0.0177, "step": 9320 }, { "epoch": 36.02316602316602, "grad_norm": 0.15173888206481934, "learning_rate": 1.2258995370774685e-06, "loss": 0.0198, "step": 9330 }, { "epoch": 36.061776061776065, "grad_norm": 0.21677173674106598, "learning_rate": 1.1897768665569798e-06, "loss": 0.0178, "step": 9340 }, { "epoch": 36.1003861003861, "grad_norm": 0.23376335203647614, "learning_rate": 1.1541879768165954e-06, "loss": 0.0139, "step": 9350 }, { "epoch": 36.13899613899614, "grad_norm": 0.2168349027633667, "learning_rate": 1.1191332570507085e-06, "loss": 0.0147, "step": 9360 }, { "epoch": 36.17760617760618, "grad_norm": 0.1469596028327942, "learning_rate": 1.0846130906121132e-06, "loss": 0.0148, "step": 9370 }, { "epoch": 36.21621621621622, "grad_norm": 0.16849203407764435, "learning_rate": 1.0506278550078131e-06, "loss": 0.0148, "step": 9380 }, { "epoch": 36.254826254826256, "grad_norm": 0.17859606444835663, "learning_rate": 1.0171779218949185e-06, "loss": 0.0198, "step": 9390 }, { "epoch": 36.293436293436294, "grad_norm": 0.22008682787418365, "learning_rate": 9.842636570765174e-07, "loss": 0.0165, "step": 9400 }, { "epoch": 36.33204633204633, "grad_norm": 0.20242340862751007, "learning_rate": 9.518854204977612e-07, "loss": 0.0158, "step": 9410 }, { "epoch": 36.37065637065637, "grad_norm": 0.1525328904390335, "learning_rate": 9.200435662418349e-07, "loss": 0.0167, "step": 9420 }, { "epoch": 36.40926640926641, "grad_norm": 0.1258171945810318, "learning_rate": 8.887384425261658e-07, "loss": 0.0134, "step": 9430 }, { "epoch": 36.447876447876446, "grad_norm": 0.26265257596969604, "learning_rate": 8.579703916985648e-07, "loss": 0.0169, "step": 9440 }, { "epoch": 36.486486486486484, "grad_norm": 0.2112697958946228, "learning_rate": 8.277397502335194e-07, "loss": 0.014, "step": 9450 }, { "epoch": 36.52509652509652, "grad_norm": 0.14722761511802673, "learning_rate": 7.980468487284675e-07, "loss": 0.0125, "step": 9460 }, { "epoch": 36.56370656370656, "grad_norm": 0.1330070197582245, "learning_rate": 7.688920119002297e-07, "loss": 0.0184, "step": 9470 }, { "epoch": 36.602316602316606, "grad_norm": 0.1656855195760727, "learning_rate": 7.402755585814269e-07, "loss": 0.015, "step": 9480 }, { "epoch": 36.640926640926644, "grad_norm": 0.2136625200510025, "learning_rate": 7.121978017170073e-07, "loss": 0.0134, "step": 9490 }, { "epoch": 36.67953667953668, "grad_norm": 0.44845545291900635, "learning_rate": 6.846590483608306e-07, "loss": 0.0159, "step": 9500 }, { "epoch": 36.71814671814672, "grad_norm": 0.24025790393352509, "learning_rate": 6.576595996722834e-07, "loss": 0.0145, "step": 9510 }, { "epoch": 36.75675675675676, "grad_norm": 0.10134408622980118, "learning_rate": 6.311997509130141e-07, "loss": 0.0155, "step": 9520 }, { "epoch": 36.795366795366796, "grad_norm": 0.28141042590141296, "learning_rate": 6.052797914436803e-07, "loss": 0.0178, "step": 9530 }, { "epoch": 36.833976833976834, "grad_norm": 0.2607072591781616, "learning_rate": 5.799000047208181e-07, "loss": 0.0173, "step": 9540 }, { "epoch": 36.87258687258687, "grad_norm": 0.23714537918567657, "learning_rate": 5.550606682937054e-07, "loss": 0.0179, "step": 9550 }, { "epoch": 36.91119691119691, "grad_norm": 0.17405599355697632, "learning_rate": 5.307620538013481e-07, "loss": 0.0166, "step": 9560 }, { "epoch": 36.94980694980695, "grad_norm": 0.2704300880432129, "learning_rate": 5.070044269694874e-07, "loss": 0.0163, "step": 9570 }, { "epoch": 36.988416988416986, "grad_norm": 0.14315839111804962, "learning_rate": 4.837880476077417e-07, "loss": 0.0134, "step": 9580 }, { "epoch": 37.027027027027025, "grad_norm": 0.18686214089393616, "learning_rate": 4.6111316960670835e-07, "loss": 0.0163, "step": 9590 }, { "epoch": 37.06563706563706, "grad_norm": 0.2980360984802246, "learning_rate": 4.389800409352218e-07, "loss": 0.0212, "step": 9600 }, { "epoch": 37.1042471042471, "grad_norm": 0.29351887106895447, "learning_rate": 4.173889036376277e-07, "loss": 0.017, "step": 9610 }, { "epoch": 37.142857142857146, "grad_norm": 0.3116950988769531, "learning_rate": 3.963399938311463e-07, "loss": 0.0127, "step": 9620 }, { "epoch": 37.181467181467184, "grad_norm": 0.32205232977867126, "learning_rate": 3.7583354170328545e-07, "loss": 0.0148, "step": 9630 }, { "epoch": 37.22007722007722, "grad_norm": 0.22022823989391327, "learning_rate": 3.558697715093207e-07, "loss": 0.0248, "step": 9640 }, { "epoch": 37.25868725868726, "grad_norm": 0.18680430948734283, "learning_rate": 3.3644890156983576e-07, "loss": 0.0174, "step": 9650 }, { "epoch": 37.2972972972973, "grad_norm": 0.37667378783226013, "learning_rate": 3.175711442683638e-07, "loss": 0.0184, "step": 9660 }, { "epoch": 37.335907335907336, "grad_norm": 0.2872820198535919, "learning_rate": 2.9923670604902197e-07, "loss": 0.0152, "step": 9670 }, { "epoch": 37.374517374517374, "grad_norm": 0.3825961947441101, "learning_rate": 2.814457874143028e-07, "loss": 0.0157, "step": 9680 }, { "epoch": 37.41312741312741, "grad_norm": 0.26287728548049927, "learning_rate": 2.641985829228366e-07, "loss": 0.0173, "step": 9690 }, { "epoch": 37.45173745173745, "grad_norm": 0.20948605239391327, "learning_rate": 2.474952811872877e-07, "loss": 0.0173, "step": 9700 }, { "epoch": 37.49034749034749, "grad_norm": 0.18369565904140472, "learning_rate": 2.3133606487228397e-07, "loss": 0.0138, "step": 9710 }, { "epoch": 37.52895752895753, "grad_norm": 0.19915831089019775, "learning_rate": 2.157211106924295e-07, "loss": 0.0147, "step": 9720 }, { "epoch": 37.567567567567565, "grad_norm": 0.16339580714702606, "learning_rate": 2.006505894103672e-07, "loss": 0.0132, "step": 9730 }, { "epoch": 37.6061776061776, "grad_norm": 0.0872827097773552, "learning_rate": 1.8612466583489696e-07, "loss": 0.0201, "step": 9740 }, { "epoch": 37.64478764478765, "grad_norm": 0.13683360815048218, "learning_rate": 1.7214349881918834e-07, "loss": 0.0155, "step": 9750 }, { "epoch": 37.683397683397686, "grad_norm": 0.15519395470619202, "learning_rate": 1.5870724125904845e-07, "loss": 0.0152, "step": 9760 }, { "epoch": 37.722007722007724, "grad_norm": 0.3538365662097931, "learning_rate": 1.4581604009124006e-07, "loss": 0.017, "step": 9770 }, { "epoch": 37.76061776061776, "grad_norm": 0.2578124701976776, "learning_rate": 1.334700362918717e-07, "loss": 0.0148, "step": 9780 }, { "epoch": 37.7992277992278, "grad_norm": 0.12069237977266312, "learning_rate": 1.2166936487486015e-07, "loss": 0.0122, "step": 9790 }, { "epoch": 37.83783783783784, "grad_norm": 0.24136237800121307, "learning_rate": 1.1041415489045914e-07, "loss": 0.0158, "step": 9800 }, { "epoch": 37.87644787644788, "grad_norm": 0.16466361284255981, "learning_rate": 9.970452942384412e-08, "loss": 0.011, "step": 9810 }, { "epoch": 37.915057915057915, "grad_norm": 0.1522415727376938, "learning_rate": 8.954060559375754e-08, "loss": 0.0186, "step": 9820 }, { "epoch": 37.95366795366795, "grad_norm": 0.18273107707500458, "learning_rate": 7.99224945512489e-08, "loss": 0.014, "step": 9830 }, { "epoch": 37.99227799227799, "grad_norm": 0.23603107035160065, "learning_rate": 7.085030147843675e-08, "loss": 0.0163, "step": 9840 }, { "epoch": 38.03088803088803, "grad_norm": 0.20355328917503357, "learning_rate": 6.232412558736523e-08, "loss": 0.0164, "step": 9850 }, { "epoch": 38.06949806949807, "grad_norm": 0.18533001840114594, "learning_rate": 5.434406011893822e-08, "loss": 0.0144, "step": 9860 }, { "epoch": 38.108108108108105, "grad_norm": 0.1358797252178192, "learning_rate": 4.6910192341864664e-08, "loss": 0.018, "step": 9870 }, { "epoch": 38.14671814671814, "grad_norm": 0.19409286975860596, "learning_rate": 4.0022603551737035e-08, "loss": 0.0178, "step": 9880 }, { "epoch": 38.18532818532819, "grad_norm": 0.2586413323879242, "learning_rate": 3.3681369070120985e-08, "loss": 0.0172, "step": 9890 }, { "epoch": 38.22393822393823, "grad_norm": 0.17179569602012634, "learning_rate": 2.7886558243744866e-08, "loss": 0.019, "step": 9900 }, { "epoch": 38.262548262548265, "grad_norm": 0.2121245414018631, "learning_rate": 2.2638234443722596e-08, "loss": 0.0129, "step": 9910 }, { "epoch": 38.3011583011583, "grad_norm": 0.13148300349712372, "learning_rate": 1.7936455064887504e-08, "loss": 0.0156, "step": 9920 }, { "epoch": 38.33976833976834, "grad_norm": 0.2647940516471863, "learning_rate": 1.378127152514841e-08, "loss": 0.0172, "step": 9930 }, { "epoch": 38.37837837837838, "grad_norm": 0.2810283303260803, "learning_rate": 1.0172729264917857e-08, "loss": 0.0272, "step": 9940 }, { "epoch": 38.41698841698842, "grad_norm": 0.19779835641384125, "learning_rate": 7.1108677466458215e-09, "loss": 0.0115, "step": 9950 }, { "epoch": 38.455598455598455, "grad_norm": 0.1843729466199875, "learning_rate": 4.595720454353414e-09, "loss": 0.0149, "step": 9960 }, { "epoch": 38.49420849420849, "grad_norm": 0.21859480440616608, "learning_rate": 2.627314893294264e-09, "loss": 0.017, "step": 9970 }, { "epoch": 38.53281853281853, "grad_norm": 0.18717947602272034, "learning_rate": 1.2056725896270048e-09, "loss": 0.0175, "step": 9980 }, { "epoch": 38.57142857142857, "grad_norm": 0.2383173257112503, "learning_rate": 3.308090902098826e-10, "loss": 0.0155, "step": 9990 }, { "epoch": 38.61003861003861, "grad_norm": 0.24821463227272034, "learning_rate": 2.7339624120159555e-12, "loss": 0.0187, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 39, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }