diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,84043 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 10.907107798582077, + "eval_steps": 500, + "global_step": 120000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0009089256498818396, + "grad_norm": 5.142341136932373, + "learning_rate": 4.5e-08, + "loss": 0.5664, + "step": 10 + }, + { + "epoch": 0.0018178512997636793, + "grad_norm": 6.3508620262146, + "learning_rate": 9.5e-08, + "loss": 0.608, + "step": 20 + }, + { + "epoch": 0.002726776949645519, + "grad_norm": 5.855591297149658, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.574, + "step": 30 + }, + { + "epoch": 0.0036357025995273585, + "grad_norm": 4.943865776062012, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5451, + "step": 40 + }, + { + "epoch": 0.004544628249409198, + "grad_norm": 7.146821022033691, + "learning_rate": 2.45e-07, + "loss": 0.5847, + "step": 50 + }, + { + "epoch": 0.005453553899291038, + "grad_norm": 5.5400800704956055, + "learning_rate": 2.95e-07, + "loss": 0.5732, + "step": 60 + }, + { + "epoch": 0.006362479549172878, + "grad_norm": 5.4721856117248535, + "learning_rate": 3.45e-07, + "loss": 0.5405, + "step": 70 + }, + { + "epoch": 0.007271405199054717, + "grad_norm": 5.023528575897217, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.504, + "step": 80 + }, + { + "epoch": 0.008180330848936558, + "grad_norm": 4.953211784362793, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.469, + "step": 90 + }, + { + "epoch": 0.009089256498818397, + "grad_norm": 4.57370138168335, + "learning_rate": 4.95e-07, + "loss": 0.4192, + "step": 100 + }, + { + "epoch": 0.009998182148700236, + "grad_norm": 3.6289403438568115, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4157, + "step": 110 + }, + { + "epoch": 0.010907107798582076, + "grad_norm": 2.375800609588623, + "learning_rate": 5.95e-07, + "loss": 0.3194, + "step": 120 + }, + { + "epoch": 0.011816033448463915, + "grad_norm": 1.7480032444000244, + "learning_rate": 6.45e-07, + "loss": 0.2631, + "step": 130 + }, + { + "epoch": 0.012724959098345756, + "grad_norm": 1.2692229747772217, + "learning_rate": 6.95e-07, + "loss": 0.2641, + "step": 140 + }, + { + "epoch": 0.013633884748227595, + "grad_norm": 1.1782046556472778, + "learning_rate": 7.45e-07, + "loss": 0.2464, + "step": 150 + }, + { + "epoch": 0.014542810398109434, + "grad_norm": 0.6911550760269165, + "learning_rate": 7.95e-07, + "loss": 0.2251, + "step": 160 + }, + { + "epoch": 0.015451736047991275, + "grad_norm": 0.8450331091880798, + "learning_rate": 8.45e-07, + "loss": 0.1832, + "step": 170 + }, + { + "epoch": 0.016360661697873115, + "grad_norm": 0.5415970087051392, + "learning_rate": 8.95e-07, + "loss": 0.1786, + "step": 180 + }, + { + "epoch": 0.017269587347754953, + "grad_norm": 0.5538029670715332, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1719, + "step": 190 + }, + { + "epoch": 0.018178512997636793, + "grad_norm": 0.8243781924247742, + "learning_rate": 9.95e-07, + "loss": 0.1635, + "step": 200 + }, + { + "epoch": 0.019087438647518634, + "grad_norm": 0.38369444012641907, + "learning_rate": 1.045e-06, + "loss": 0.1628, + "step": 210 + }, + { + "epoch": 0.01999636429740047, + "grad_norm": 0.5153273940086365, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1652, + "step": 220 + }, + { + "epoch": 0.020905289947282312, + "grad_norm": 0.5912274122238159, + "learning_rate": 1.145e-06, + "loss": 0.155, + "step": 230 + }, + { + "epoch": 0.021814215597164153, + "grad_norm": 0.5956100821495056, + "learning_rate": 1.195e-06, + "loss": 0.1691, + "step": 240 + }, + { + "epoch": 0.02272314124704599, + "grad_norm": 0.5506021976470947, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1552, + "step": 250 + }, + { + "epoch": 0.02363206689692783, + "grad_norm": 0.4963970184326172, + "learning_rate": 1.295e-06, + "loss": 0.1528, + "step": 260 + }, + { + "epoch": 0.02454099254680967, + "grad_norm": 0.5694977641105652, + "learning_rate": 1.345e-06, + "loss": 0.1511, + "step": 270 + }, + { + "epoch": 0.025449918196691512, + "grad_norm": 0.457369863986969, + "learning_rate": 1.395e-06, + "loss": 0.135, + "step": 280 + }, + { + "epoch": 0.02635884384657335, + "grad_norm": 0.4367586672306061, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1381, + "step": 290 + }, + { + "epoch": 0.02726776949645519, + "grad_norm": 0.8469308018684387, + "learning_rate": 1.495e-06, + "loss": 0.1439, + "step": 300 + }, + { + "epoch": 0.02817669514633703, + "grad_norm": 0.46191591024398804, + "learning_rate": 1.545e-06, + "loss": 0.1493, + "step": 310 + }, + { + "epoch": 0.029085620796218868, + "grad_norm": 0.5551906228065491, + "learning_rate": 1.595e-06, + "loss": 0.1314, + "step": 320 + }, + { + "epoch": 0.02999454644610071, + "grad_norm": 0.5542528033256531, + "learning_rate": 1.645e-06, + "loss": 0.1421, + "step": 330 + }, + { + "epoch": 0.03090347209598255, + "grad_norm": 0.469052791595459, + "learning_rate": 1.695e-06, + "loss": 0.1337, + "step": 340 + }, + { + "epoch": 0.03181239774586439, + "grad_norm": 0.5372846126556396, + "learning_rate": 1.745e-06, + "loss": 0.1311, + "step": 350 + }, + { + "epoch": 0.03272132339574623, + "grad_norm": 0.7013001441955566, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1342, + "step": 360 + }, + { + "epoch": 0.033630249045628065, + "grad_norm": 0.6155235171318054, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.034539174695509905, + "grad_norm": 0.3972844183444977, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1255, + "step": 380 + }, + { + "epoch": 0.035448100345391746, + "grad_norm": 0.4715825617313385, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1253, + "step": 390 + }, + { + "epoch": 0.03635702599527359, + "grad_norm": 0.8323620557785034, + "learning_rate": 1.995e-06, + "loss": 0.1255, + "step": 400 + }, + { + "epoch": 0.03726595164515543, + "grad_norm": 0.47903138399124146, + "learning_rate": 2.045e-06, + "loss": 0.1237, + "step": 410 + }, + { + "epoch": 0.03817487729503727, + "grad_norm": 0.40829524397850037, + "learning_rate": 2.095e-06, + "loss": 0.1251, + "step": 420 + }, + { + "epoch": 0.03908380294491911, + "grad_norm": 0.47905880212783813, + "learning_rate": 2.145e-06, + "loss": 0.124, + "step": 430 + }, + { + "epoch": 0.03999272859480094, + "grad_norm": 0.6843248009681702, + "learning_rate": 2.195e-06, + "loss": 0.1128, + "step": 440 + }, + { + "epoch": 0.040901654244682784, + "grad_norm": 0.43051350116729736, + "learning_rate": 2.245e-06, + "loss": 0.1254, + "step": 450 + }, + { + "epoch": 0.041810579894564624, + "grad_norm": 0.5338430404663086, + "learning_rate": 2.295e-06, + "loss": 0.1168, + "step": 460 + }, + { + "epoch": 0.042719505544446465, + "grad_norm": 0.5056979060173035, + "learning_rate": 2.345e-06, + "loss": 0.1186, + "step": 470 + }, + { + "epoch": 0.043628431194328306, + "grad_norm": 0.5985885858535767, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.114, + "step": 480 + }, + { + "epoch": 0.044537356844210146, + "grad_norm": 0.6011636853218079, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1172, + "step": 490 + }, + { + "epoch": 0.04544628249409198, + "grad_norm": 0.4860469698905945, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1117, + "step": 500 + }, + { + "epoch": 0.04635520814397382, + "grad_norm": 0.7282719612121582, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1074, + "step": 510 + }, + { + "epoch": 0.04726413379385566, + "grad_norm": 0.5821688771247864, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1141, + "step": 520 + }, + { + "epoch": 0.0481730594437375, + "grad_norm": 0.48014211654663086, + "learning_rate": 2.645e-06, + "loss": 0.115, + "step": 530 + }, + { + "epoch": 0.04908198509361934, + "grad_norm": 0.48185163736343384, + "learning_rate": 2.695e-06, + "loss": 0.1084, + "step": 540 + }, + { + "epoch": 0.049990910743501184, + "grad_norm": 0.4627598226070404, + "learning_rate": 2.745e-06, + "loss": 0.1045, + "step": 550 + }, + { + "epoch": 0.050899836393383024, + "grad_norm": 0.8217755556106567, + "learning_rate": 2.795e-06, + "loss": 0.1118, + "step": 560 + }, + { + "epoch": 0.05180876204326486, + "grad_norm": 0.6816278100013733, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1115, + "step": 570 + }, + { + "epoch": 0.0527176876931467, + "grad_norm": 0.6204538345336914, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1124, + "step": 580 + }, + { + "epoch": 0.05362661334302854, + "grad_norm": 0.47710031270980835, + "learning_rate": 2.945e-06, + "loss": 0.1105, + "step": 590 + }, + { + "epoch": 0.05453553899291038, + "grad_norm": 0.5174102783203125, + "learning_rate": 2.995e-06, + "loss": 0.108, + "step": 600 + }, + { + "epoch": 0.05544446464279222, + "grad_norm": 0.7783994674682617, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.1028, + "step": 610 + }, + { + "epoch": 0.05635339029267406, + "grad_norm": 0.6555732488632202, + "learning_rate": 3.095e-06, + "loss": 0.1058, + "step": 620 + }, + { + "epoch": 0.057262315942555896, + "grad_norm": 0.40201637148857117, + "learning_rate": 3.145e-06, + "loss": 0.1069, + "step": 630 + }, + { + "epoch": 0.058171241592437736, + "grad_norm": 0.5610593557357788, + "learning_rate": 3.195e-06, + "loss": 0.1039, + "step": 640 + }, + { + "epoch": 0.05908016724231958, + "grad_norm": 0.4806128144264221, + "learning_rate": 3.245e-06, + "loss": 0.1064, + "step": 650 + }, + { + "epoch": 0.05998909289220142, + "grad_norm": 0.4764435887336731, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.1019, + "step": 660 + }, + { + "epoch": 0.06089801854208326, + "grad_norm": 0.5325391292572021, + "learning_rate": 3.345e-06, + "loss": 0.1052, + "step": 670 + }, + { + "epoch": 0.0618069441919651, + "grad_norm": 0.5369204878807068, + "learning_rate": 3.395e-06, + "loss": 0.101, + "step": 680 + }, + { + "epoch": 0.06271586984184693, + "grad_norm": 0.6801527142524719, + "learning_rate": 3.445e-06, + "loss": 0.0951, + "step": 690 + }, + { + "epoch": 0.06362479549172878, + "grad_norm": 0.4192810356616974, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0959, + "step": 700 + }, + { + "epoch": 0.06453372114161061, + "grad_norm": 0.4996216595172882, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.1009, + "step": 710 + }, + { + "epoch": 0.06544264679149246, + "grad_norm": 0.48684173822402954, + "learning_rate": 3.595e-06, + "loss": 0.1009, + "step": 720 + }, + { + "epoch": 0.0663515724413743, + "grad_norm": 0.6220418810844421, + "learning_rate": 3.645e-06, + "loss": 0.0978, + "step": 730 + }, + { + "epoch": 0.06726049809125613, + "grad_norm": 0.4554404318332672, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.1003, + "step": 740 + }, + { + "epoch": 0.06816942374113798, + "grad_norm": 0.5508406162261963, + "learning_rate": 3.745e-06, + "loss": 0.1018, + "step": 750 + }, + { + "epoch": 0.06907834939101981, + "grad_norm": 0.5557074546813965, + "learning_rate": 3.795e-06, + "loss": 0.0998, + "step": 760 + }, + { + "epoch": 0.06998727504090166, + "grad_norm": 0.5697064399719238, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1023, + "step": 770 + }, + { + "epoch": 0.07089620069078349, + "grad_norm": 0.4464508593082428, + "learning_rate": 3.895e-06, + "loss": 0.1011, + "step": 780 + }, + { + "epoch": 0.07180512634066534, + "grad_norm": 0.6867432594299316, + "learning_rate": 3.945e-06, + "loss": 0.0966, + "step": 790 + }, + { + "epoch": 0.07271405199054717, + "grad_norm": 1.0665061473846436, + "learning_rate": 3.995e-06, + "loss": 0.1038, + "step": 800 + }, + { + "epoch": 0.07362297764042901, + "grad_norm": 0.4841720461845398, + "learning_rate": 4.045e-06, + "loss": 0.0906, + "step": 810 + }, + { + "epoch": 0.07453190329031086, + "grad_norm": 0.513376772403717, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0986, + "step": 820 + }, + { + "epoch": 0.07544082894019269, + "grad_norm": 0.6003584265708923, + "learning_rate": 4.145e-06, + "loss": 0.0975, + "step": 830 + }, + { + "epoch": 0.07634975459007454, + "grad_norm": 0.5348184704780579, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0974, + "step": 840 + }, + { + "epoch": 0.07725868023995637, + "grad_norm": 0.4122712016105652, + "learning_rate": 4.245e-06, + "loss": 0.0945, + "step": 850 + }, + { + "epoch": 0.07816760588983822, + "grad_norm": 0.5381463766098022, + "learning_rate": 4.295e-06, + "loss": 0.0922, + "step": 860 + }, + { + "epoch": 0.07907653153972005, + "grad_norm": 0.8734978437423706, + "learning_rate": 4.345000000000001e-06, + "loss": 0.09, + "step": 870 + }, + { + "epoch": 0.07998545718960189, + "grad_norm": 0.5671399831771851, + "learning_rate": 4.395e-06, + "loss": 0.0975, + "step": 880 + }, + { + "epoch": 0.08089438283948373, + "grad_norm": 0.6146457195281982, + "learning_rate": 4.445e-06, + "loss": 0.0915, + "step": 890 + }, + { + "epoch": 0.08180330848936557, + "grad_norm": 0.509801983833313, + "learning_rate": 4.495e-06, + "loss": 0.0936, + "step": 900 + }, + { + "epoch": 0.08271223413924741, + "grad_norm": 0.6508568525314331, + "learning_rate": 4.545e-06, + "loss": 0.0885, + "step": 910 + }, + { + "epoch": 0.08362115978912925, + "grad_norm": 0.7533997297286987, + "learning_rate": 4.595e-06, + "loss": 0.0913, + "step": 920 + }, + { + "epoch": 0.08453008543901108, + "grad_norm": 0.7011661529541016, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0948, + "step": 930 + }, + { + "epoch": 0.08543901108889293, + "grad_norm": 0.7136930823326111, + "learning_rate": 4.695e-06, + "loss": 0.0912, + "step": 940 + }, + { + "epoch": 0.08634793673877476, + "grad_norm": 0.46569958329200745, + "learning_rate": 4.745e-06, + "loss": 0.0926, + "step": 950 + }, + { + "epoch": 0.08725686238865661, + "grad_norm": 0.4648204445838928, + "learning_rate": 4.795e-06, + "loss": 0.0905, + "step": 960 + }, + { + "epoch": 0.08816578803853845, + "grad_norm": 0.5478107929229736, + "learning_rate": 4.845e-06, + "loss": 0.0905, + "step": 970 + }, + { + "epoch": 0.08907471368842029, + "grad_norm": 0.5909463763237, + "learning_rate": 4.895e-06, + "loss": 0.0867, + "step": 980 + }, + { + "epoch": 0.08998363933830213, + "grad_norm": 0.6179943084716797, + "learning_rate": 4.945e-06, + "loss": 0.0928, + "step": 990 + }, + { + "epoch": 0.09089256498818396, + "grad_norm": 0.5512673258781433, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0843, + "step": 1000 + }, + { + "epoch": 0.09180149063806581, + "grad_norm": 0.5822701454162598, + "learning_rate": 5.045e-06, + "loss": 0.0907, + "step": 1010 + }, + { + "epoch": 0.09271041628794764, + "grad_norm": 0.48837459087371826, + "learning_rate": 5.095e-06, + "loss": 0.0967, + "step": 1020 + }, + { + "epoch": 0.09361934193782949, + "grad_norm": 0.5112066864967346, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0886, + "step": 1030 + }, + { + "epoch": 0.09452826758771132, + "grad_norm": 0.6347087025642395, + "learning_rate": 5.195e-06, + "loss": 0.0811, + "step": 1040 + }, + { + "epoch": 0.09543719323759317, + "grad_norm": 0.45846128463745117, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0853, + "step": 1050 + }, + { + "epoch": 0.096346118887475, + "grad_norm": 0.5639715790748596, + "learning_rate": 5.295e-06, + "loss": 0.0833, + "step": 1060 + }, + { + "epoch": 0.09725504453735684, + "grad_norm": 0.47275617718696594, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0858, + "step": 1070 + }, + { + "epoch": 0.09816397018723869, + "grad_norm": 0.441364586353302, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0896, + "step": 1080 + }, + { + "epoch": 0.09907289583712052, + "grad_norm": 0.5424106121063232, + "learning_rate": 5.445e-06, + "loss": 0.0859, + "step": 1090 + }, + { + "epoch": 0.09998182148700237, + "grad_norm": 0.4296520948410034, + "learning_rate": 5.495e-06, + "loss": 0.08, + "step": 1100 + }, + { + "epoch": 0.1008907471368842, + "grad_norm": 0.3924048840999603, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0811, + "step": 1110 + }, + { + "epoch": 0.10179967278676605, + "grad_norm": 0.3739417493343353, + "learning_rate": 5.595e-06, + "loss": 0.0833, + "step": 1120 + }, + { + "epoch": 0.10270859843664788, + "grad_norm": 0.41514962911605835, + "learning_rate": 5.645e-06, + "loss": 0.0869, + "step": 1130 + }, + { + "epoch": 0.10361752408652972, + "grad_norm": 0.5074795484542847, + "learning_rate": 5.695e-06, + "loss": 0.0856, + "step": 1140 + }, + { + "epoch": 0.10452644973641156, + "grad_norm": 0.37289518117904663, + "learning_rate": 5.745e-06, + "loss": 0.0808, + "step": 1150 + }, + { + "epoch": 0.1054353753862934, + "grad_norm": 0.4628245234489441, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0849, + "step": 1160 + }, + { + "epoch": 0.10634430103617525, + "grad_norm": 0.3584427535533905, + "learning_rate": 5.845e-06, + "loss": 0.0898, + "step": 1170 + }, + { + "epoch": 0.10725322668605708, + "grad_norm": 0.41567352414131165, + "learning_rate": 5.895e-06, + "loss": 0.0794, + "step": 1180 + }, + { + "epoch": 0.10816215233593893, + "grad_norm": 0.4720224440097809, + "learning_rate": 5.945e-06, + "loss": 0.0802, + "step": 1190 + }, + { + "epoch": 0.10907107798582076, + "grad_norm": 0.6269563436508179, + "learning_rate": 5.995e-06, + "loss": 0.0827, + "step": 1200 + }, + { + "epoch": 0.1099800036357026, + "grad_norm": 0.4757440388202667, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0767, + "step": 1210 + }, + { + "epoch": 0.11088892928558444, + "grad_norm": 0.4439792037010193, + "learning_rate": 6.095e-06, + "loss": 0.0804, + "step": 1220 + }, + { + "epoch": 0.11179785493546628, + "grad_norm": 0.48423057794570923, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.085, + "step": 1230 + }, + { + "epoch": 0.11270678058534812, + "grad_norm": 0.4771044850349426, + "learning_rate": 6.195e-06, + "loss": 0.0778, + "step": 1240 + }, + { + "epoch": 0.11361570623522996, + "grad_norm": 0.4573425352573395, + "learning_rate": 6.245e-06, + "loss": 0.0813, + "step": 1250 + }, + { + "epoch": 0.11452463188511179, + "grad_norm": 0.47508350014686584, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0823, + "step": 1260 + }, + { + "epoch": 0.11543355753499364, + "grad_norm": 0.6641986966133118, + "learning_rate": 6.345e-06, + "loss": 0.0768, + "step": 1270 + }, + { + "epoch": 0.11634248318487547, + "grad_norm": 0.5082783102989197, + "learning_rate": 6.395000000000001e-06, + "loss": 0.083, + "step": 1280 + }, + { + "epoch": 0.11725140883475732, + "grad_norm": 0.5429797768592834, + "learning_rate": 6.445e-06, + "loss": 0.0803, + "step": 1290 + }, + { + "epoch": 0.11816033448463915, + "grad_norm": 0.47736337780952454, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0784, + "step": 1300 + }, + { + "epoch": 0.119069260134521, + "grad_norm": 0.42957803606987, + "learning_rate": 6.545e-06, + "loss": 0.0779, + "step": 1310 + }, + { + "epoch": 0.11997818578440284, + "grad_norm": 0.5390623807907104, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0702, + "step": 1320 + }, + { + "epoch": 0.12088711143428467, + "grad_norm": 0.38133788108825684, + "learning_rate": 6.645e-06, + "loss": 0.0878, + "step": 1330 + }, + { + "epoch": 0.12179603708416652, + "grad_norm": 0.47153717279434204, + "learning_rate": 6.695e-06, + "loss": 0.0808, + "step": 1340 + }, + { + "epoch": 0.12270496273404835, + "grad_norm": 0.3647540509700775, + "learning_rate": 6.745e-06, + "loss": 0.0787, + "step": 1350 + }, + { + "epoch": 0.1236138883839302, + "grad_norm": 0.40490618348121643, + "learning_rate": 6.795e-06, + "loss": 0.074, + "step": 1360 + }, + { + "epoch": 0.12452281403381203, + "grad_norm": 0.44806215167045593, + "learning_rate": 6.845e-06, + "loss": 0.0767, + "step": 1370 + }, + { + "epoch": 0.12543173968369387, + "grad_norm": 0.54866623878479, + "learning_rate": 6.895e-06, + "loss": 0.0755, + "step": 1380 + }, + { + "epoch": 0.1263406653335757, + "grad_norm": 0.5088191032409668, + "learning_rate": 6.945e-06, + "loss": 0.072, + "step": 1390 + }, + { + "epoch": 0.12724959098345756, + "grad_norm": 0.5052351355552673, + "learning_rate": 6.995e-06, + "loss": 0.0725, + "step": 1400 + }, + { + "epoch": 0.12815851663333938, + "grad_norm": 0.4849494397640228, + "learning_rate": 7.045e-06, + "loss": 0.0737, + "step": 1410 + }, + { + "epoch": 0.12906744228322123, + "grad_norm": 0.5659723877906799, + "learning_rate": 7.095e-06, + "loss": 0.0761, + "step": 1420 + }, + { + "epoch": 0.12997636793310308, + "grad_norm": 0.5013991594314575, + "learning_rate": 7.145e-06, + "loss": 0.0746, + "step": 1430 + }, + { + "epoch": 0.13088529358298492, + "grad_norm": 0.4878064692020416, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0701, + "step": 1440 + }, + { + "epoch": 0.13179421923286674, + "grad_norm": 0.33322420716285706, + "learning_rate": 7.245e-06, + "loss": 0.0731, + "step": 1450 + }, + { + "epoch": 0.1327031448827486, + "grad_norm": 0.4975486695766449, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0812, + "step": 1460 + }, + { + "epoch": 0.13361207053263044, + "grad_norm": 0.601776659488678, + "learning_rate": 7.345e-06, + "loss": 0.0734, + "step": 1470 + }, + { + "epoch": 0.13452099618251226, + "grad_norm": 0.4338124394416809, + "learning_rate": 7.395e-06, + "loss": 0.0728, + "step": 1480 + }, + { + "epoch": 0.1354299218323941, + "grad_norm": 0.585129976272583, + "learning_rate": 7.445000000000001e-06, + "loss": 0.075, + "step": 1490 + }, + { + "epoch": 0.13633884748227595, + "grad_norm": 0.40535876154899597, + "learning_rate": 7.495e-06, + "loss": 0.0764, + "step": 1500 + }, + { + "epoch": 0.1372477731321578, + "grad_norm": 0.37111443281173706, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0796, + "step": 1510 + }, + { + "epoch": 0.13815669878203962, + "grad_norm": 0.49845826625823975, + "learning_rate": 7.595e-06, + "loss": 0.0724, + "step": 1520 + }, + { + "epoch": 0.13906562443192147, + "grad_norm": 0.4641841948032379, + "learning_rate": 7.645e-06, + "loss": 0.0789, + "step": 1530 + }, + { + "epoch": 0.13997455008180332, + "grad_norm": 0.4493994414806366, + "learning_rate": 7.695e-06, + "loss": 0.0753, + "step": 1540 + }, + { + "epoch": 0.14088347573168514, + "grad_norm": 0.4318861663341522, + "learning_rate": 7.745e-06, + "loss": 0.0742, + "step": 1550 + }, + { + "epoch": 0.14179240138156698, + "grad_norm": 0.4534268081188202, + "learning_rate": 7.795e-06, + "loss": 0.0721, + "step": 1560 + }, + { + "epoch": 0.14270132703144883, + "grad_norm": 0.4156348705291748, + "learning_rate": 7.845e-06, + "loss": 0.076, + "step": 1570 + }, + { + "epoch": 0.14361025268133068, + "grad_norm": 0.37263840436935425, + "learning_rate": 7.895e-06, + "loss": 0.0743, + "step": 1580 + }, + { + "epoch": 0.1445191783312125, + "grad_norm": 0.5091559886932373, + "learning_rate": 7.945e-06, + "loss": 0.0717, + "step": 1590 + }, + { + "epoch": 0.14542810398109435, + "grad_norm": 0.5153735280036926, + "learning_rate": 7.995e-06, + "loss": 0.074, + "step": 1600 + }, + { + "epoch": 0.1463370296309762, + "grad_norm": 0.4642679691314697, + "learning_rate": 8.045e-06, + "loss": 0.0712, + "step": 1610 + }, + { + "epoch": 0.14724595528085802, + "grad_norm": 0.40805888175964355, + "learning_rate": 8.095e-06, + "loss": 0.0725, + "step": 1620 + }, + { + "epoch": 0.14815488093073986, + "grad_norm": 0.39019790291786194, + "learning_rate": 8.145e-06, + "loss": 0.0733, + "step": 1630 + }, + { + "epoch": 0.1490638065806217, + "grad_norm": 0.31650519371032715, + "learning_rate": 8.195e-06, + "loss": 0.0685, + "step": 1640 + }, + { + "epoch": 0.14997273223050356, + "grad_norm": 0.3074605166912079, + "learning_rate": 8.245e-06, + "loss": 0.0673, + "step": 1650 + }, + { + "epoch": 0.15088165788038538, + "grad_norm": 0.43762606382369995, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0698, + "step": 1660 + }, + { + "epoch": 0.15179058353026723, + "grad_norm": 0.4818330407142639, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.15269950918014907, + "grad_norm": 0.5147614479064941, + "learning_rate": 8.395e-06, + "loss": 0.0756, + "step": 1680 + }, + { + "epoch": 0.1536084348300309, + "grad_norm": 0.4130430221557617, + "learning_rate": 8.445e-06, + "loss": 0.0721, + "step": 1690 + }, + { + "epoch": 0.15451736047991274, + "grad_norm": 0.46807846426963806, + "learning_rate": 8.495e-06, + "loss": 0.0816, + "step": 1700 + }, + { + "epoch": 0.1554262861297946, + "grad_norm": 0.5442712903022766, + "learning_rate": 8.545e-06, + "loss": 0.0725, + "step": 1710 + }, + { + "epoch": 0.15633521177967644, + "grad_norm": 0.4125683903694153, + "learning_rate": 8.595e-06, + "loss": 0.0728, + "step": 1720 + }, + { + "epoch": 0.15724413742955826, + "grad_norm": 0.4022798538208008, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0761, + "step": 1730 + }, + { + "epoch": 0.1581530630794401, + "grad_norm": 0.4019310772418976, + "learning_rate": 8.695e-06, + "loss": 0.0717, + "step": 1740 + }, + { + "epoch": 0.15906198872932195, + "grad_norm": 0.38936278223991394, + "learning_rate": 8.745e-06, + "loss": 0.0777, + "step": 1750 + }, + { + "epoch": 0.15997091437920377, + "grad_norm": 0.3076677620410919, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0665, + "step": 1760 + }, + { + "epoch": 0.16087984002908562, + "grad_norm": 0.5165063738822937, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0728, + "step": 1770 + }, + { + "epoch": 0.16178876567896747, + "grad_norm": 0.41853952407836914, + "learning_rate": 8.895e-06, + "loss": 0.0684, + "step": 1780 + }, + { + "epoch": 0.1626976913288493, + "grad_norm": 0.2919950485229492, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0748, + "step": 1790 + }, + { + "epoch": 0.16360661697873113, + "grad_norm": 0.49655160307884216, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0706, + "step": 1800 + }, + { + "epoch": 0.16451554262861298, + "grad_norm": 0.48480281233787537, + "learning_rate": 9.045e-06, + "loss": 0.0666, + "step": 1810 + }, + { + "epoch": 0.16542446827849483, + "grad_norm": 0.6117209196090698, + "learning_rate": 9.095e-06, + "loss": 0.0707, + "step": 1820 + }, + { + "epoch": 0.16633339392837665, + "grad_norm": 0.5614205002784729, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0701, + "step": 1830 + }, + { + "epoch": 0.1672423195782585, + "grad_norm": 0.41813912987709045, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0735, + "step": 1840 + }, + { + "epoch": 0.16815124522814034, + "grad_norm": 0.39823973178863525, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0777, + "step": 1850 + }, + { + "epoch": 0.16906017087802216, + "grad_norm": 0.35153281688690186, + "learning_rate": 9.295e-06, + "loss": 0.0716, + "step": 1860 + }, + { + "epoch": 0.169969096527904, + "grad_norm": 0.6184330582618713, + "learning_rate": 9.345e-06, + "loss": 0.0721, + "step": 1870 + }, + { + "epoch": 0.17087802217778586, + "grad_norm": 0.3351885676383972, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0687, + "step": 1880 + }, + { + "epoch": 0.1717869478276677, + "grad_norm": 0.3466022312641144, + "learning_rate": 9.445e-06, + "loss": 0.0688, + "step": 1890 + }, + { + "epoch": 0.17269587347754953, + "grad_norm": 0.3122941851615906, + "learning_rate": 9.495e-06, + "loss": 0.0644, + "step": 1900 + }, + { + "epoch": 0.17360479912743138, + "grad_norm": 0.445300430059433, + "learning_rate": 9.545e-06, + "loss": 0.0725, + "step": 1910 + }, + { + "epoch": 0.17451372477731322, + "grad_norm": 0.45175501704216003, + "learning_rate": 9.595e-06, + "loss": 0.0673, + "step": 1920 + }, + { + "epoch": 0.17542265042719504, + "grad_norm": 0.566214919090271, + "learning_rate": 9.645e-06, + "loss": 0.07, + "step": 1930 + }, + { + "epoch": 0.1763315760770769, + "grad_norm": 0.4020490348339081, + "learning_rate": 9.695e-06, + "loss": 0.0695, + "step": 1940 + }, + { + "epoch": 0.17724050172695874, + "grad_norm": 0.4413970112800598, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0677, + "step": 1950 + }, + { + "epoch": 0.17814942737684059, + "grad_norm": 0.3525959849357605, + "learning_rate": 9.795e-06, + "loss": 0.069, + "step": 1960 + }, + { + "epoch": 0.1790583530267224, + "grad_norm": 0.358815461397171, + "learning_rate": 9.845e-06, + "loss": 0.0708, + "step": 1970 + }, + { + "epoch": 0.17996727867660425, + "grad_norm": 0.5534330606460571, + "learning_rate": 9.895e-06, + "loss": 0.0714, + "step": 1980 + }, + { + "epoch": 0.1808762043264861, + "grad_norm": 0.32233819365501404, + "learning_rate": 9.945e-06, + "loss": 0.0654, + "step": 1990 + }, + { + "epoch": 0.18178512997636792, + "grad_norm": 0.3312453031539917, + "learning_rate": 9.995e-06, + "loss": 0.0686, + "step": 2000 + }, + { + "epoch": 0.18269405562624977, + "grad_norm": 0.34086593985557556, + "learning_rate": 1.0045e-05, + "loss": 0.0687, + "step": 2010 + }, + { + "epoch": 0.18360298127613162, + "grad_norm": 0.3363412022590637, + "learning_rate": 1.0095e-05, + "loss": 0.068, + "step": 2020 + }, + { + "epoch": 0.18451190692601346, + "grad_norm": 0.40297842025756836, + "learning_rate": 1.0145e-05, + "loss": 0.0711, + "step": 2030 + }, + { + "epoch": 0.18542083257589528, + "grad_norm": 0.3619011640548706, + "learning_rate": 1.0195e-05, + "loss": 0.0655, + "step": 2040 + }, + { + "epoch": 0.18632975822577713, + "grad_norm": 0.4580240845680237, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0709, + "step": 2050 + }, + { + "epoch": 0.18723868387565898, + "grad_norm": 0.4522128999233246, + "learning_rate": 1.0295e-05, + "loss": 0.0683, + "step": 2060 + }, + { + "epoch": 0.1881476095255408, + "grad_norm": 0.367236465215683, + "learning_rate": 1.0345e-05, + "loss": 0.0679, + "step": 2070 + }, + { + "epoch": 0.18905653517542265, + "grad_norm": 0.4386281967163086, + "learning_rate": 1.0395e-05, + "loss": 0.0631, + "step": 2080 + }, + { + "epoch": 0.1899654608253045, + "grad_norm": 0.3953603506088257, + "learning_rate": 1.0445e-05, + "loss": 0.0699, + "step": 2090 + }, + { + "epoch": 0.19087438647518634, + "grad_norm": 0.3962622582912445, + "learning_rate": 1.0495e-05, + "loss": 0.0666, + "step": 2100 + }, + { + "epoch": 0.19178331212506816, + "grad_norm": 0.2862510681152344, + "learning_rate": 1.0545e-05, + "loss": 0.0689, + "step": 2110 + }, + { + "epoch": 0.19269223777495, + "grad_norm": 0.4321883022785187, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.061, + "step": 2120 + }, + { + "epoch": 0.19360116342483186, + "grad_norm": 0.41177862882614136, + "learning_rate": 1.0645e-05, + "loss": 0.0686, + "step": 2130 + }, + { + "epoch": 0.19451008907471368, + "grad_norm": 0.4986853003501892, + "learning_rate": 1.0695e-05, + "loss": 0.066, + "step": 2140 + }, + { + "epoch": 0.19541901472459552, + "grad_norm": 0.4283527135848999, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0655, + "step": 2150 + }, + { + "epoch": 0.19632794037447737, + "grad_norm": 0.36543622612953186, + "learning_rate": 1.0795e-05, + "loss": 0.0626, + "step": 2160 + }, + { + "epoch": 0.19723686602435922, + "grad_norm": 0.3797658681869507, + "learning_rate": 1.0845e-05, + "loss": 0.0641, + "step": 2170 + }, + { + "epoch": 0.19814579167424104, + "grad_norm": 0.2987470328807831, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0652, + "step": 2180 + }, + { + "epoch": 0.1990547173241229, + "grad_norm": 0.3525169789791107, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0634, + "step": 2190 + }, + { + "epoch": 0.19996364297400473, + "grad_norm": 0.41089358925819397, + "learning_rate": 1.0995e-05, + "loss": 0.0729, + "step": 2200 + }, + { + "epoch": 0.20087256862388655, + "grad_norm": 0.4604075849056244, + "learning_rate": 1.1045e-05, + "loss": 0.0677, + "step": 2210 + }, + { + "epoch": 0.2017814942737684, + "grad_norm": 0.3177061378955841, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0665, + "step": 2220 + }, + { + "epoch": 0.20269041992365025, + "grad_norm": 0.4938071072101593, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0642, + "step": 2230 + }, + { + "epoch": 0.2035993455735321, + "grad_norm": 0.3650347888469696, + "learning_rate": 1.1195e-05, + "loss": 0.0705, + "step": 2240 + }, + { + "epoch": 0.20450827122341392, + "grad_norm": 0.33096954226493835, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0642, + "step": 2250 + }, + { + "epoch": 0.20541719687329577, + "grad_norm": 0.2533232867717743, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0667, + "step": 2260 + }, + { + "epoch": 0.2063261225231776, + "grad_norm": 0.4190867245197296, + "learning_rate": 1.1345e-05, + "loss": 0.0659, + "step": 2270 + }, + { + "epoch": 0.20723504817305943, + "grad_norm": 0.3413848280906677, + "learning_rate": 1.1395e-05, + "loss": 0.0603, + "step": 2280 + }, + { + "epoch": 0.20814397382294128, + "grad_norm": 0.3148519694805145, + "learning_rate": 1.1445e-05, + "loss": 0.0661, + "step": 2290 + }, + { + "epoch": 0.20905289947282313, + "grad_norm": 0.3443912863731384, + "learning_rate": 1.1495e-05, + "loss": 0.065, + "step": 2300 + }, + { + "epoch": 0.20996182512270498, + "grad_norm": 0.2971969246864319, + "learning_rate": 1.1545e-05, + "loss": 0.0646, + "step": 2310 + }, + { + "epoch": 0.2108707507725868, + "grad_norm": 0.36810529232025146, + "learning_rate": 1.1595e-05, + "loss": 0.066, + "step": 2320 + }, + { + "epoch": 0.21177967642246864, + "grad_norm": 0.29141274094581604, + "learning_rate": 1.1645e-05, + "loss": 0.0663, + "step": 2330 + }, + { + "epoch": 0.2126886020723505, + "grad_norm": 0.30333903431892395, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.062, + "step": 2340 + }, + { + "epoch": 0.2135975277222323, + "grad_norm": 0.3264903128147125, + "learning_rate": 1.1745e-05, + "loss": 0.0691, + "step": 2350 + }, + { + "epoch": 0.21450645337211416, + "grad_norm": 0.4041016399860382, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.215415379021996, + "grad_norm": 0.3747559189796448, + "learning_rate": 1.1845e-05, + "loss": 0.0687, + "step": 2370 + }, + { + "epoch": 0.21632430467187785, + "grad_norm": 0.3984030783176422, + "learning_rate": 1.1895e-05, + "loss": 0.0712, + "step": 2380 + }, + { + "epoch": 0.21723323032175967, + "grad_norm": 0.411191463470459, + "learning_rate": 1.1945e-05, + "loss": 0.0702, + "step": 2390 + }, + { + "epoch": 0.21814215597164152, + "grad_norm": 0.35119935870170593, + "learning_rate": 1.1995e-05, + "loss": 0.0682, + "step": 2400 + }, + { + "epoch": 0.21905108162152337, + "grad_norm": 0.45232003927230835, + "learning_rate": 1.2045e-05, + "loss": 0.0634, + "step": 2410 + }, + { + "epoch": 0.2199600072714052, + "grad_norm": 0.36612462997436523, + "learning_rate": 1.2095e-05, + "loss": 0.0695, + "step": 2420 + }, + { + "epoch": 0.22086893292128704, + "grad_norm": 0.2557939291000366, + "learning_rate": 1.2145e-05, + "loss": 0.0641, + "step": 2430 + }, + { + "epoch": 0.22177785857116888, + "grad_norm": 0.2694337069988251, + "learning_rate": 1.2195e-05, + "loss": 0.0681, + "step": 2440 + }, + { + "epoch": 0.2226867842210507, + "grad_norm": 0.33925190567970276, + "learning_rate": 1.2245e-05, + "loss": 0.0681, + "step": 2450 + }, + { + "epoch": 0.22359570987093255, + "grad_norm": 0.510261058807373, + "learning_rate": 1.2295e-05, + "loss": 0.0663, + "step": 2460 + }, + { + "epoch": 0.2245046355208144, + "grad_norm": 0.32837826013565063, + "learning_rate": 1.2345e-05, + "loss": 0.0653, + "step": 2470 + }, + { + "epoch": 0.22541356117069625, + "grad_norm": 0.3581598103046417, + "learning_rate": 1.2395e-05, + "loss": 0.0655, + "step": 2480 + }, + { + "epoch": 0.22632248682057807, + "grad_norm": 0.2914462387561798, + "learning_rate": 1.2445e-05, + "loss": 0.0705, + "step": 2490 + }, + { + "epoch": 0.22723141247045991, + "grad_norm": 0.321918249130249, + "learning_rate": 1.2495e-05, + "loss": 0.0642, + "step": 2500 + }, + { + "epoch": 0.22814033812034176, + "grad_norm": 0.40168648958206177, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0654, + "step": 2510 + }, + { + "epoch": 0.22904926377022358, + "grad_norm": 0.2909962832927704, + "learning_rate": 1.2595e-05, + "loss": 0.0623, + "step": 2520 + }, + { + "epoch": 0.22995818942010543, + "grad_norm": 0.4044923484325409, + "learning_rate": 1.2645e-05, + "loss": 0.0676, + "step": 2530 + }, + { + "epoch": 0.23086711506998728, + "grad_norm": 0.3975919187068939, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0634, + "step": 2540 + }, + { + "epoch": 0.23177604071986913, + "grad_norm": 0.31659597158432007, + "learning_rate": 1.2745e-05, + "loss": 0.0661, + "step": 2550 + }, + { + "epoch": 0.23268496636975095, + "grad_norm": 0.35358092188835144, + "learning_rate": 1.2795e-05, + "loss": 0.0684, + "step": 2560 + }, + { + "epoch": 0.2335938920196328, + "grad_norm": 0.4075811207294464, + "learning_rate": 1.2845e-05, + "loss": 0.0611, + "step": 2570 + }, + { + "epoch": 0.23450281766951464, + "grad_norm": 0.4067555367946625, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0677, + "step": 2580 + }, + { + "epoch": 0.23541174331939646, + "grad_norm": 0.35957926511764526, + "learning_rate": 1.2945e-05, + "loss": 0.061, + "step": 2590 + }, + { + "epoch": 0.2363206689692783, + "grad_norm": 0.36920198798179626, + "learning_rate": 1.2995e-05, + "loss": 0.0668, + "step": 2600 + }, + { + "epoch": 0.23722959461916016, + "grad_norm": 0.35895222425460815, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0614, + "step": 2610 + }, + { + "epoch": 0.238138520269042, + "grad_norm": 0.36858999729156494, + "learning_rate": 1.3095e-05, + "loss": 0.0622, + "step": 2620 + }, + { + "epoch": 0.23904744591892382, + "grad_norm": 0.42415329813957214, + "learning_rate": 1.3145e-05, + "loss": 0.0651, + "step": 2630 + }, + { + "epoch": 0.23995637156880567, + "grad_norm": 0.4130004048347473, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0681, + "step": 2640 + }, + { + "epoch": 0.24086529721868752, + "grad_norm": 0.3024885356426239, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0642, + "step": 2650 + }, + { + "epoch": 0.24177422286856934, + "grad_norm": 0.2532311975955963, + "learning_rate": 1.3295e-05, + "loss": 0.0619, + "step": 2660 + }, + { + "epoch": 0.24268314851845119, + "grad_norm": 0.29364368319511414, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0672, + "step": 2670 + }, + { + "epoch": 0.24359207416833303, + "grad_norm": 0.3247676491737366, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0637, + "step": 2680 + }, + { + "epoch": 0.24450099981821488, + "grad_norm": 0.41420242190361023, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0632, + "step": 2690 + }, + { + "epoch": 0.2454099254680967, + "grad_norm": 0.36173519492149353, + "learning_rate": 1.3495e-05, + "loss": 0.0611, + "step": 2700 + }, + { + "epoch": 0.24631885111797855, + "grad_norm": 0.36842525005340576, + "learning_rate": 1.3545e-05, + "loss": 0.0621, + "step": 2710 + }, + { + "epoch": 0.2472277767678604, + "grad_norm": 0.410807728767395, + "learning_rate": 1.3595e-05, + "loss": 0.0658, + "step": 2720 + }, + { + "epoch": 0.24813670241774222, + "grad_norm": 0.36475473642349243, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0658, + "step": 2730 + }, + { + "epoch": 0.24904562806762406, + "grad_norm": 0.349046915769577, + "learning_rate": 1.3695e-05, + "loss": 0.0641, + "step": 2740 + }, + { + "epoch": 0.2499545537175059, + "grad_norm": 0.4921092987060547, + "learning_rate": 1.3745e-05, + "loss": 0.0637, + "step": 2750 + }, + { + "epoch": 0.25086347936738773, + "grad_norm": 0.3486356735229492, + "learning_rate": 1.3795e-05, + "loss": 0.0575, + "step": 2760 + }, + { + "epoch": 0.2517724050172696, + "grad_norm": 0.40467703342437744, + "learning_rate": 1.3845e-05, + "loss": 0.072, + "step": 2770 + }, + { + "epoch": 0.2526813306671514, + "grad_norm": 0.37536197900772095, + "learning_rate": 1.3895e-05, + "loss": 0.0618, + "step": 2780 + }, + { + "epoch": 0.25359025631703325, + "grad_norm": 0.29910269379615784, + "learning_rate": 1.3945e-05, + "loss": 0.0624, + "step": 2790 + }, + { + "epoch": 0.2544991819669151, + "grad_norm": 0.2833632528781891, + "learning_rate": 1.3995e-05, + "loss": 0.0663, + "step": 2800 + }, + { + "epoch": 0.25540810761679694, + "grad_norm": 0.3226313591003418, + "learning_rate": 1.4045e-05, + "loss": 0.0641, + "step": 2810 + }, + { + "epoch": 0.25631703326667876, + "grad_norm": 0.39734870195388794, + "learning_rate": 1.4095e-05, + "loss": 0.0658, + "step": 2820 + }, + { + "epoch": 0.25722595891656064, + "grad_norm": 0.29596197605133057, + "learning_rate": 1.4145e-05, + "loss": 0.061, + "step": 2830 + }, + { + "epoch": 0.25813488456644246, + "grad_norm": 0.33789244294166565, + "learning_rate": 1.4195e-05, + "loss": 0.0626, + "step": 2840 + }, + { + "epoch": 0.2590438102163243, + "grad_norm": 0.4610069990158081, + "learning_rate": 1.4245e-05, + "loss": 0.0613, + "step": 2850 + }, + { + "epoch": 0.25995273586620615, + "grad_norm": 0.3358405530452728, + "learning_rate": 1.4295e-05, + "loss": 0.0663, + "step": 2860 + }, + { + "epoch": 0.260861661516088, + "grad_norm": 0.4274219274520874, + "learning_rate": 1.4345e-05, + "loss": 0.0683, + "step": 2870 + }, + { + "epoch": 0.26177058716596985, + "grad_norm": 0.30274853110313416, + "learning_rate": 1.4395e-05, + "loss": 0.0629, + "step": 2880 + }, + { + "epoch": 0.26267951281585167, + "grad_norm": 0.36349236965179443, + "learning_rate": 1.4445e-05, + "loss": 0.0666, + "step": 2890 + }, + { + "epoch": 0.2635884384657335, + "grad_norm": 0.2990400791168213, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0596, + "step": 2900 + }, + { + "epoch": 0.26449736411561536, + "grad_norm": 0.4198426306247711, + "learning_rate": 1.4545e-05, + "loss": 0.0623, + "step": 2910 + }, + { + "epoch": 0.2654062897654972, + "grad_norm": 0.38104864954948425, + "learning_rate": 1.4595e-05, + "loss": 0.0707, + "step": 2920 + }, + { + "epoch": 0.266315215415379, + "grad_norm": 0.32979616522789, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0659, + "step": 2930 + }, + { + "epoch": 0.2672241410652609, + "grad_norm": 0.48916715383529663, + "learning_rate": 1.4695e-05, + "loss": 0.0651, + "step": 2940 + }, + { + "epoch": 0.2681330667151427, + "grad_norm": 0.39585480093955994, + "learning_rate": 1.4745e-05, + "loss": 0.0617, + "step": 2950 + }, + { + "epoch": 0.2690419923650245, + "grad_norm": 0.37473630905151367, + "learning_rate": 1.4795e-05, + "loss": 0.0649, + "step": 2960 + }, + { + "epoch": 0.2699509180149064, + "grad_norm": 0.317751407623291, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0607, + "step": 2970 + }, + { + "epoch": 0.2708598436647882, + "grad_norm": 0.3780079185962677, + "learning_rate": 1.4895e-05, + "loss": 0.0643, + "step": 2980 + }, + { + "epoch": 0.27176876931467003, + "grad_norm": 0.411120742559433, + "learning_rate": 1.4945e-05, + "loss": 0.0623, + "step": 2990 + }, + { + "epoch": 0.2726776949645519, + "grad_norm": 0.2883855402469635, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0619, + "step": 3000 + }, + { + "epoch": 0.27358662061443373, + "grad_norm": 0.31510674953460693, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0572, + "step": 3010 + }, + { + "epoch": 0.2744955462643156, + "grad_norm": 0.35954052209854126, + "learning_rate": 1.5095e-05, + "loss": 0.0633, + "step": 3020 + }, + { + "epoch": 0.2754044719141974, + "grad_norm": 0.30163946747779846, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0551, + "step": 3030 + }, + { + "epoch": 0.27631339756407924, + "grad_norm": 0.37070244550704956, + "learning_rate": 1.5195e-05, + "loss": 0.0652, + "step": 3040 + }, + { + "epoch": 0.2772223232139611, + "grad_norm": 0.3196491003036499, + "learning_rate": 1.5245e-05, + "loss": 0.0586, + "step": 3050 + }, + { + "epoch": 0.27813124886384294, + "grad_norm": 0.33795931935310364, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.061, + "step": 3060 + }, + { + "epoch": 0.27904017451372476, + "grad_norm": 0.4125255048274994, + "learning_rate": 1.5345e-05, + "loss": 0.0635, + "step": 3070 + }, + { + "epoch": 0.27994910016360663, + "grad_norm": 0.3110499382019043, + "learning_rate": 1.5395e-05, + "loss": 0.0588, + "step": 3080 + }, + { + "epoch": 0.28085802581348845, + "grad_norm": 0.37663084268569946, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0639, + "step": 3090 + }, + { + "epoch": 0.2817669514633703, + "grad_norm": 0.30412793159484863, + "learning_rate": 1.5495e-05, + "loss": 0.0639, + "step": 3100 + }, + { + "epoch": 0.28267587711325215, + "grad_norm": 0.3895590305328369, + "learning_rate": 1.5545e-05, + "loss": 0.0583, + "step": 3110 + }, + { + "epoch": 0.28358480276313397, + "grad_norm": 0.3582421839237213, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0608, + "step": 3120 + }, + { + "epoch": 0.2844937284130158, + "grad_norm": 0.45156097412109375, + "learning_rate": 1.5645e-05, + "loss": 0.0644, + "step": 3130 + }, + { + "epoch": 0.28540265406289766, + "grad_norm": 0.3519822359085083, + "learning_rate": 1.5695e-05, + "loss": 0.0634, + "step": 3140 + }, + { + "epoch": 0.2863115797127795, + "grad_norm": 0.32449132204055786, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0605, + "step": 3150 + }, + { + "epoch": 0.28722050536266136, + "grad_norm": 0.27821025252342224, + "learning_rate": 1.5795e-05, + "loss": 0.0606, + "step": 3160 + }, + { + "epoch": 0.2881294310125432, + "grad_norm": 0.32676801085472107, + "learning_rate": 1.5845e-05, + "loss": 0.0574, + "step": 3170 + }, + { + "epoch": 0.289038356662425, + "grad_norm": 0.31156641244888306, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0591, + "step": 3180 + }, + { + "epoch": 0.2899472823123069, + "grad_norm": 0.3160015940666199, + "learning_rate": 1.5945e-05, + "loss": 0.0599, + "step": 3190 + }, + { + "epoch": 0.2908562079621887, + "grad_norm": 0.3506348729133606, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0589, + "step": 3200 + }, + { + "epoch": 0.2917651336120705, + "grad_norm": 0.28403621912002563, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0576, + "step": 3210 + }, + { + "epoch": 0.2926740592619524, + "grad_norm": 0.428037166595459, + "learning_rate": 1.6095e-05, + "loss": 0.0583, + "step": 3220 + }, + { + "epoch": 0.2935829849118342, + "grad_norm": 0.3317340612411499, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0584, + "step": 3230 + }, + { + "epoch": 0.29449191056171603, + "grad_norm": 0.2568194568157196, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0623, + "step": 3240 + }, + { + "epoch": 0.2954008362115979, + "grad_norm": 0.30690792202949524, + "learning_rate": 1.6245e-05, + "loss": 0.0583, + "step": 3250 + }, + { + "epoch": 0.2963097618614797, + "grad_norm": 0.2843332588672638, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0603, + "step": 3260 + }, + { + "epoch": 0.29721868751136155, + "grad_norm": 0.26969799399375916, + "learning_rate": 1.6345e-05, + "loss": 0.06, + "step": 3270 + }, + { + "epoch": 0.2981276131612434, + "grad_norm": 0.30513036251068115, + "learning_rate": 1.6395e-05, + "loss": 0.0615, + "step": 3280 + }, + { + "epoch": 0.29903653881112524, + "grad_norm": 0.33301371335983276, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0661, + "step": 3290 + }, + { + "epoch": 0.2999454644610071, + "grad_norm": 0.29709747433662415, + "learning_rate": 1.6495e-05, + "loss": 0.0633, + "step": 3300 + }, + { + "epoch": 0.30085439011088894, + "grad_norm": 0.3238356411457062, + "learning_rate": 1.6545e-05, + "loss": 0.058, + "step": 3310 + }, + { + "epoch": 0.30176331576077076, + "grad_norm": 0.3497830629348755, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0621, + "step": 3320 + }, + { + "epoch": 0.30267224141065263, + "grad_norm": 0.25603967905044556, + "learning_rate": 1.6645e-05, + "loss": 0.0605, + "step": 3330 + }, + { + "epoch": 0.30358116706053445, + "grad_norm": 0.3564808666706085, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0595, + "step": 3340 + }, + { + "epoch": 0.30449009271041627, + "grad_norm": 0.352626770734787, + "learning_rate": 1.6745e-05, + "loss": 0.0618, + "step": 3350 + }, + { + "epoch": 0.30539901836029815, + "grad_norm": 0.3339376747608185, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0578, + "step": 3360 + }, + { + "epoch": 0.30630794401017997, + "grad_norm": 0.4939483404159546, + "learning_rate": 1.6845e-05, + "loss": 0.0612, + "step": 3370 + }, + { + "epoch": 0.3072168696600618, + "grad_norm": 0.3217848837375641, + "learning_rate": 1.6895e-05, + "loss": 0.0604, + "step": 3380 + }, + { + "epoch": 0.30812579530994366, + "grad_norm": 0.33011728525161743, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0599, + "step": 3390 + }, + { + "epoch": 0.3090347209598255, + "grad_norm": 0.3215973675251007, + "learning_rate": 1.6995e-05, + "loss": 0.0663, + "step": 3400 + }, + { + "epoch": 0.3099436466097073, + "grad_norm": 0.3558996617794037, + "learning_rate": 1.7045e-05, + "loss": 0.0572, + "step": 3410 + }, + { + "epoch": 0.3108525722595892, + "grad_norm": 0.24993331730365753, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0566, + "step": 3420 + }, + { + "epoch": 0.311761497909471, + "grad_norm": 0.3437276780605316, + "learning_rate": 1.7145e-05, + "loss": 0.062, + "step": 3430 + }, + { + "epoch": 0.3126704235593529, + "grad_norm": 0.3417102098464966, + "learning_rate": 1.7195e-05, + "loss": 0.0641, + "step": 3440 + }, + { + "epoch": 0.3135793492092347, + "grad_norm": 0.27577558159828186, + "learning_rate": 1.7245e-05, + "loss": 0.059, + "step": 3450 + }, + { + "epoch": 0.3144882748591165, + "grad_norm": 0.2560344934463501, + "learning_rate": 1.7295e-05, + "loss": 0.0595, + "step": 3460 + }, + { + "epoch": 0.3153972005089984, + "grad_norm": 0.24801300466060638, + "learning_rate": 1.7345e-05, + "loss": 0.054, + "step": 3470 + }, + { + "epoch": 0.3163061261588802, + "grad_norm": 0.30988577008247375, + "learning_rate": 1.7395e-05, + "loss": 0.0617, + "step": 3480 + }, + { + "epoch": 0.317215051808762, + "grad_norm": 0.25673243403434753, + "learning_rate": 1.7445e-05, + "loss": 0.0598, + "step": 3490 + }, + { + "epoch": 0.3181239774586439, + "grad_norm": 0.35730063915252686, + "learning_rate": 1.7495e-05, + "loss": 0.0583, + "step": 3500 + }, + { + "epoch": 0.3190329031085257, + "grad_norm": 0.3536713719367981, + "learning_rate": 1.7545e-05, + "loss": 0.0546, + "step": 3510 + }, + { + "epoch": 0.31994182875840754, + "grad_norm": 0.2756310701370239, + "learning_rate": 1.7595e-05, + "loss": 0.0618, + "step": 3520 + }, + { + "epoch": 0.3208507544082894, + "grad_norm": 0.22890642285346985, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0621, + "step": 3530 + }, + { + "epoch": 0.32175968005817124, + "grad_norm": 0.33372873067855835, + "learning_rate": 1.7695e-05, + "loss": 0.06, + "step": 3540 + }, + { + "epoch": 0.32266860570805306, + "grad_norm": 0.25788500905036926, + "learning_rate": 1.7745e-05, + "loss": 0.0603, + "step": 3550 + }, + { + "epoch": 0.32357753135793493, + "grad_norm": 0.30196863412857056, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0606, + "step": 3560 + }, + { + "epoch": 0.32448645700781675, + "grad_norm": 0.28120771050453186, + "learning_rate": 1.7845e-05, + "loss": 0.0585, + "step": 3570 + }, + { + "epoch": 0.3253953826576986, + "grad_norm": 0.2888234853744507, + "learning_rate": 1.7895e-05, + "loss": 0.0614, + "step": 3580 + }, + { + "epoch": 0.32630430830758045, + "grad_norm": 0.3084333837032318, + "learning_rate": 1.7945e-05, + "loss": 0.0596, + "step": 3590 + }, + { + "epoch": 0.32721323395746227, + "grad_norm": 0.3085821568965912, + "learning_rate": 1.7995e-05, + "loss": 0.0594, + "step": 3600 + }, + { + "epoch": 0.32812215960734414, + "grad_norm": 0.3304099440574646, + "learning_rate": 1.8045e-05, + "loss": 0.0583, + "step": 3610 + }, + { + "epoch": 0.32903108525722596, + "grad_norm": 0.230202779173851, + "learning_rate": 1.8095e-05, + "loss": 0.0588, + "step": 3620 + }, + { + "epoch": 0.3299400109071078, + "grad_norm": 0.34347978234291077, + "learning_rate": 1.8145e-05, + "loss": 0.056, + "step": 3630 + }, + { + "epoch": 0.33084893655698966, + "grad_norm": 0.286900132894516, + "learning_rate": 1.8195e-05, + "loss": 0.0573, + "step": 3640 + }, + { + "epoch": 0.3317578622068715, + "grad_norm": 0.27875030040740967, + "learning_rate": 1.8245e-05, + "loss": 0.0592, + "step": 3650 + }, + { + "epoch": 0.3326667878567533, + "grad_norm": 0.27106955647468567, + "learning_rate": 1.8295e-05, + "loss": 0.0599, + "step": 3660 + }, + { + "epoch": 0.3335757135066352, + "grad_norm": 0.2745857238769531, + "learning_rate": 1.8345e-05, + "loss": 0.0598, + "step": 3670 + }, + { + "epoch": 0.334484639156517, + "grad_norm": 0.35625332593917847, + "learning_rate": 1.8395e-05, + "loss": 0.0571, + "step": 3680 + }, + { + "epoch": 0.3353935648063988, + "grad_norm": 0.2921353876590729, + "learning_rate": 1.8445e-05, + "loss": 0.0598, + "step": 3690 + }, + { + "epoch": 0.3363024904562807, + "grad_norm": 0.275747150182724, + "learning_rate": 1.8495e-05, + "loss": 0.0579, + "step": 3700 + }, + { + "epoch": 0.3372114161061625, + "grad_norm": 0.3563058376312256, + "learning_rate": 1.8545e-05, + "loss": 0.0576, + "step": 3710 + }, + { + "epoch": 0.33812034175604433, + "grad_norm": 0.5174416303634644, + "learning_rate": 1.8595e-05, + "loss": 0.0632, + "step": 3720 + }, + { + "epoch": 0.3390292674059262, + "grad_norm": 0.37203505635261536, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.059, + "step": 3730 + }, + { + "epoch": 0.339938193055808, + "grad_norm": 0.4004843533039093, + "learning_rate": 1.8695e-05, + "loss": 0.0614, + "step": 3740 + }, + { + "epoch": 0.3408471187056899, + "grad_norm": 0.38872063159942627, + "learning_rate": 1.8745e-05, + "loss": 0.0607, + "step": 3750 + }, + { + "epoch": 0.3417560443555717, + "grad_norm": 0.39403682947158813, + "learning_rate": 1.8795e-05, + "loss": 0.0606, + "step": 3760 + }, + { + "epoch": 0.34266497000545354, + "grad_norm": 0.2660045325756073, + "learning_rate": 1.8845e-05, + "loss": 0.0582, + "step": 3770 + }, + { + "epoch": 0.3435738956553354, + "grad_norm": 0.3145920932292938, + "learning_rate": 1.8895e-05, + "loss": 0.0579, + "step": 3780 + }, + { + "epoch": 0.34448282130521723, + "grad_norm": 0.26681774854660034, + "learning_rate": 1.8945e-05, + "loss": 0.0591, + "step": 3790 + }, + { + "epoch": 0.34539174695509905, + "grad_norm": 0.2681003212928772, + "learning_rate": 1.8995e-05, + "loss": 0.0584, + "step": 3800 + }, + { + "epoch": 0.34630067260498093, + "grad_norm": 0.31365811824798584, + "learning_rate": 1.9045e-05, + "loss": 0.0574, + "step": 3810 + }, + { + "epoch": 0.34720959825486275, + "grad_norm": 0.38997235894203186, + "learning_rate": 1.9095e-05, + "loss": 0.0557, + "step": 3820 + }, + { + "epoch": 0.34811852390474457, + "grad_norm": 0.3218485414981842, + "learning_rate": 1.9145e-05, + "loss": 0.0617, + "step": 3830 + }, + { + "epoch": 0.34902744955462645, + "grad_norm": 0.3144035041332245, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0566, + "step": 3840 + }, + { + "epoch": 0.34993637520450827, + "grad_norm": 0.303850382566452, + "learning_rate": 1.9245e-05, + "loss": 0.0594, + "step": 3850 + }, + { + "epoch": 0.3508453008543901, + "grad_norm": 0.3470322787761688, + "learning_rate": 1.9295e-05, + "loss": 0.0591, + "step": 3860 + }, + { + "epoch": 0.35175422650427196, + "grad_norm": 0.23239845037460327, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0585, + "step": 3870 + }, + { + "epoch": 0.3526631521541538, + "grad_norm": 0.24539801478385925, + "learning_rate": 1.9395e-05, + "loss": 0.0555, + "step": 3880 + }, + { + "epoch": 0.35357207780403566, + "grad_norm": 0.2931928336620331, + "learning_rate": 1.9445e-05, + "loss": 0.0554, + "step": 3890 + }, + { + "epoch": 0.3544810034539175, + "grad_norm": 0.3055608868598938, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0547, + "step": 3900 + }, + { + "epoch": 0.3553899291037993, + "grad_norm": 0.2967032492160797, + "learning_rate": 1.9545e-05, + "loss": 0.062, + "step": 3910 + }, + { + "epoch": 0.35629885475368117, + "grad_norm": 0.3620520532131195, + "learning_rate": 1.9595e-05, + "loss": 0.0633, + "step": 3920 + }, + { + "epoch": 0.357207780403563, + "grad_norm": 0.24999567866325378, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0548, + "step": 3930 + }, + { + "epoch": 0.3581167060534448, + "grad_norm": 0.3241874873638153, + "learning_rate": 1.9695e-05, + "loss": 0.0603, + "step": 3940 + }, + { + "epoch": 0.3590256317033267, + "grad_norm": 0.2541414201259613, + "learning_rate": 1.9745e-05, + "loss": 0.0542, + "step": 3950 + }, + { + "epoch": 0.3599345573532085, + "grad_norm": 0.2952078878879547, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0591, + "step": 3960 + }, + { + "epoch": 0.3608434830030903, + "grad_norm": 0.24288655817508698, + "learning_rate": 1.9845e-05, + "loss": 0.0562, + "step": 3970 + }, + { + "epoch": 0.3617524086529722, + "grad_norm": 0.31199172139167786, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0653, + "step": 3980 + }, + { + "epoch": 0.362661334302854, + "grad_norm": 0.2515959143638611, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0598, + "step": 3990 + }, + { + "epoch": 0.36357025995273584, + "grad_norm": 0.27723026275634766, + "learning_rate": 1.9995e-05, + "loss": 0.0555, + "step": 4000 + }, + { + "epoch": 0.3644791856026177, + "grad_norm": 0.30606982111930847, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0574, + "step": 4010 + }, + { + "epoch": 0.36538811125249954, + "grad_norm": 0.24645878374576569, + "learning_rate": 2.0095e-05, + "loss": 0.0546, + "step": 4020 + }, + { + "epoch": 0.3662970369023814, + "grad_norm": 0.36588114500045776, + "learning_rate": 2.0145e-05, + "loss": 0.0566, + "step": 4030 + }, + { + "epoch": 0.36720596255226323, + "grad_norm": 0.2959708273410797, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0584, + "step": 4040 + }, + { + "epoch": 0.36811488820214505, + "grad_norm": 0.2583031952381134, + "learning_rate": 2.0245e-05, + "loss": 0.0568, + "step": 4050 + }, + { + "epoch": 0.3690238138520269, + "grad_norm": 0.3178505003452301, + "learning_rate": 2.0295e-05, + "loss": 0.0582, + "step": 4060 + }, + { + "epoch": 0.36993273950190875, + "grad_norm": 0.291858971118927, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0575, + "step": 4070 + }, + { + "epoch": 0.37084166515179057, + "grad_norm": 0.24564874172210693, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 0.37175059080167244, + "grad_norm": 0.2714059054851532, + "learning_rate": 2.0445e-05, + "loss": 0.0567, + "step": 4090 + }, + { + "epoch": 0.37265951645155426, + "grad_norm": 0.3685589134693146, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0574, + "step": 4100 + }, + { + "epoch": 0.3735684421014361, + "grad_norm": 0.3035966753959656, + "learning_rate": 2.0545e-05, + "loss": 0.0572, + "step": 4110 + }, + { + "epoch": 0.37447736775131796, + "grad_norm": 0.356089323759079, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.057, + "step": 4120 + }, + { + "epoch": 0.3753862934011998, + "grad_norm": 0.31210723519325256, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0526, + "step": 4130 + }, + { + "epoch": 0.3762952190510816, + "grad_norm": 0.22744810581207275, + "learning_rate": 2.0695e-05, + "loss": 0.0588, + "step": 4140 + }, + { + "epoch": 0.3772041447009635, + "grad_norm": 0.2611781358718872, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0519, + "step": 4150 + }, + { + "epoch": 0.3781130703508453, + "grad_norm": 0.306612104177475, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0596, + "step": 4160 + }, + { + "epoch": 0.37902199600072717, + "grad_norm": 0.2966499924659729, + "learning_rate": 2.0845e-05, + "loss": 0.0558, + "step": 4170 + }, + { + "epoch": 0.379930921650609, + "grad_norm": 0.32641953229904175, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0577, + "step": 4180 + }, + { + "epoch": 0.3808398473004908, + "grad_norm": 0.3320779502391815, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0536, + "step": 4190 + }, + { + "epoch": 0.3817487729503727, + "grad_norm": 0.2730516493320465, + "learning_rate": 2.0995e-05, + "loss": 0.0535, + "step": 4200 + }, + { + "epoch": 0.3826576986002545, + "grad_norm": 0.24854202568531036, + "learning_rate": 2.1045e-05, + "loss": 0.0536, + "step": 4210 + }, + { + "epoch": 0.3835666242501363, + "grad_norm": 0.33025237917900085, + "learning_rate": 2.1095e-05, + "loss": 0.0535, + "step": 4220 + }, + { + "epoch": 0.3844755499000182, + "grad_norm": 0.44324585795402527, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0527, + "step": 4230 + }, + { + "epoch": 0.3853844755499, + "grad_norm": 0.2577708959579468, + "learning_rate": 2.1195e-05, + "loss": 0.0586, + "step": 4240 + }, + { + "epoch": 0.38629340119978184, + "grad_norm": 0.4535130560398102, + "learning_rate": 2.1245e-05, + "loss": 0.058, + "step": 4250 + }, + { + "epoch": 0.3872023268496637, + "grad_norm": 0.3480747640132904, + "learning_rate": 2.1295e-05, + "loss": 0.0588, + "step": 4260 + }, + { + "epoch": 0.38811125249954553, + "grad_norm": 0.23482321202754974, + "learning_rate": 2.1345e-05, + "loss": 0.053, + "step": 4270 + }, + { + "epoch": 0.38902017814942735, + "grad_norm": 0.31652891635894775, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0567, + "step": 4280 + }, + { + "epoch": 0.38992910379930923, + "grad_norm": 0.2912371754646301, + "learning_rate": 2.1445e-05, + "loss": 0.0592, + "step": 4290 + }, + { + "epoch": 0.39083802944919105, + "grad_norm": 0.30332624912261963, + "learning_rate": 2.1495e-05, + "loss": 0.0566, + "step": 4300 + }, + { + "epoch": 0.39174695509907287, + "grad_norm": 0.3086753785610199, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0572, + "step": 4310 + }, + { + "epoch": 0.39265588074895474, + "grad_norm": 0.3442608714103699, + "learning_rate": 2.1595e-05, + "loss": 0.0549, + "step": 4320 + }, + { + "epoch": 0.39356480639883656, + "grad_norm": 0.2798554599285126, + "learning_rate": 2.1645e-05, + "loss": 0.0545, + "step": 4330 + }, + { + "epoch": 0.39447373204871844, + "grad_norm": 0.3111012279987335, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0617, + "step": 4340 + }, + { + "epoch": 0.39538265769860026, + "grad_norm": 0.34918075799942017, + "learning_rate": 2.1745e-05, + "loss": 0.0555, + "step": 4350 + }, + { + "epoch": 0.3962915833484821, + "grad_norm": 0.41185083985328674, + "learning_rate": 2.1795e-05, + "loss": 0.0553, + "step": 4360 + }, + { + "epoch": 0.39720050899836395, + "grad_norm": 0.22948874533176422, + "learning_rate": 2.1845e-05, + "loss": 0.0592, + "step": 4370 + }, + { + "epoch": 0.3981094346482458, + "grad_norm": 0.2985573410987854, + "learning_rate": 2.1895e-05, + "loss": 0.0606, + "step": 4380 + }, + { + "epoch": 0.3990183602981276, + "grad_norm": 0.2512398064136505, + "learning_rate": 2.1945e-05, + "loss": 0.0574, + "step": 4390 + }, + { + "epoch": 0.39992728594800947, + "grad_norm": 0.3945372700691223, + "learning_rate": 2.1995e-05, + "loss": 0.0599, + "step": 4400 + }, + { + "epoch": 0.4008362115978913, + "grad_norm": 0.3843232989311218, + "learning_rate": 2.2045e-05, + "loss": 0.0612, + "step": 4410 + }, + { + "epoch": 0.4017451372477731, + "grad_norm": 0.28321510553359985, + "learning_rate": 2.2095e-05, + "loss": 0.0577, + "step": 4420 + }, + { + "epoch": 0.402654062897655, + "grad_norm": 0.32515454292297363, + "learning_rate": 2.2145e-05, + "loss": 0.0576, + "step": 4430 + }, + { + "epoch": 0.4035629885475368, + "grad_norm": 0.3406917154788971, + "learning_rate": 2.2195e-05, + "loss": 0.0547, + "step": 4440 + }, + { + "epoch": 0.4044719141974186, + "grad_norm": 0.36531752347946167, + "learning_rate": 2.2245e-05, + "loss": 0.0579, + "step": 4450 + }, + { + "epoch": 0.4053808398473005, + "grad_norm": 0.2893613576889038, + "learning_rate": 2.2295e-05, + "loss": 0.0513, + "step": 4460 + }, + { + "epoch": 0.4062897654971823, + "grad_norm": 0.32492032647132874, + "learning_rate": 2.2345e-05, + "loss": 0.0577, + "step": 4470 + }, + { + "epoch": 0.4071986911470642, + "grad_norm": 0.23493535816669464, + "learning_rate": 2.2395e-05, + "loss": 0.0583, + "step": 4480 + }, + { + "epoch": 0.408107616796946, + "grad_norm": 0.31516891717910767, + "learning_rate": 2.2445e-05, + "loss": 0.0543, + "step": 4490 + }, + { + "epoch": 0.40901654244682784, + "grad_norm": 0.36818304657936096, + "learning_rate": 2.2495e-05, + "loss": 0.0603, + "step": 4500 + }, + { + "epoch": 0.4099254680967097, + "grad_norm": 0.24519942700862885, + "learning_rate": 2.2545e-05, + "loss": 0.052, + "step": 4510 + }, + { + "epoch": 0.41083439374659153, + "grad_norm": 0.32441750168800354, + "learning_rate": 2.2595e-05, + "loss": 0.0543, + "step": 4520 + }, + { + "epoch": 0.41174331939647335, + "grad_norm": 0.2598000764846802, + "learning_rate": 2.2645e-05, + "loss": 0.0546, + "step": 4530 + }, + { + "epoch": 0.4126522450463552, + "grad_norm": 0.28344082832336426, + "learning_rate": 2.2695e-05, + "loss": 0.0558, + "step": 4540 + }, + { + "epoch": 0.41356117069623705, + "grad_norm": 0.33692309260368347, + "learning_rate": 2.2745e-05, + "loss": 0.06, + "step": 4550 + }, + { + "epoch": 0.41447009634611887, + "grad_norm": 0.2615864872932434, + "learning_rate": 2.2795e-05, + "loss": 0.0535, + "step": 4560 + }, + { + "epoch": 0.41537902199600074, + "grad_norm": 0.2412802129983902, + "learning_rate": 2.2845e-05, + "loss": 0.0538, + "step": 4570 + }, + { + "epoch": 0.41628794764588256, + "grad_norm": 0.3473518192768097, + "learning_rate": 2.2895e-05, + "loss": 0.0557, + "step": 4580 + }, + { + "epoch": 0.4171968732957644, + "grad_norm": 0.290528267621994, + "learning_rate": 2.2945e-05, + "loss": 0.0557, + "step": 4590 + }, + { + "epoch": 0.41810579894564626, + "grad_norm": 0.2814607620239258, + "learning_rate": 2.2995e-05, + "loss": 0.0573, + "step": 4600 + }, + { + "epoch": 0.4190147245955281, + "grad_norm": 0.33710020780563354, + "learning_rate": 2.3045e-05, + "loss": 0.059, + "step": 4610 + }, + { + "epoch": 0.41992365024540995, + "grad_norm": 0.2612219750881195, + "learning_rate": 2.3095e-05, + "loss": 0.0561, + "step": 4620 + }, + { + "epoch": 0.42083257589529177, + "grad_norm": 0.2746369242668152, + "learning_rate": 2.3145e-05, + "loss": 0.0533, + "step": 4630 + }, + { + "epoch": 0.4217415015451736, + "grad_norm": 0.2850051522254944, + "learning_rate": 2.3195e-05, + "loss": 0.0564, + "step": 4640 + }, + { + "epoch": 0.42265042719505547, + "grad_norm": 0.28482890129089355, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0575, + "step": 4650 + }, + { + "epoch": 0.4235593528449373, + "grad_norm": 0.3222864270210266, + "learning_rate": 2.3295e-05, + "loss": 0.0537, + "step": 4660 + }, + { + "epoch": 0.4244682784948191, + "grad_norm": 0.25332513451576233, + "learning_rate": 2.3345e-05, + "loss": 0.0592, + "step": 4670 + }, + { + "epoch": 0.425377204144701, + "grad_norm": 0.31597650051116943, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0569, + "step": 4680 + }, + { + "epoch": 0.4262861297945828, + "grad_norm": 0.2752333879470825, + "learning_rate": 2.3445e-05, + "loss": 0.0562, + "step": 4690 + }, + { + "epoch": 0.4271950554444646, + "grad_norm": 0.22534038126468658, + "learning_rate": 2.3495e-05, + "loss": 0.0566, + "step": 4700 + }, + { + "epoch": 0.4281039810943465, + "grad_norm": 0.3032161593437195, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0547, + "step": 4710 + }, + { + "epoch": 0.4290129067442283, + "grad_norm": 0.3661860525608063, + "learning_rate": 2.3595e-05, + "loss": 0.0544, + "step": 4720 + }, + { + "epoch": 0.42992183239411014, + "grad_norm": 0.37232881784439087, + "learning_rate": 2.3645e-05, + "loss": 0.0544, + "step": 4730 + }, + { + "epoch": 0.430830758043992, + "grad_norm": 0.2874091863632202, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0537, + "step": 4740 + }, + { + "epoch": 0.43173968369387383, + "grad_norm": 0.2836018204689026, + "learning_rate": 2.3745e-05, + "loss": 0.0593, + "step": 4750 + }, + { + "epoch": 0.4326486093437557, + "grad_norm": 0.26815205812454224, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0577, + "step": 4760 + }, + { + "epoch": 0.43355753499363753, + "grad_norm": 0.3591286838054657, + "learning_rate": 2.3845e-05, + "loss": 0.0561, + "step": 4770 + }, + { + "epoch": 0.43446646064351935, + "grad_norm": 0.2391611635684967, + "learning_rate": 2.3895e-05, + "loss": 0.0558, + "step": 4780 + }, + { + "epoch": 0.4353753862934012, + "grad_norm": 0.3105160892009735, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0585, + "step": 4790 + }, + { + "epoch": 0.43628431194328304, + "grad_norm": 0.43707650899887085, + "learning_rate": 2.3995e-05, + "loss": 0.0522, + "step": 4800 + }, + { + "epoch": 0.43719323759316486, + "grad_norm": 0.36395615339279175, + "learning_rate": 2.4045e-05, + "loss": 0.0591, + "step": 4810 + }, + { + "epoch": 0.43810216324304674, + "grad_norm": 0.23208026587963104, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0574, + "step": 4820 + }, + { + "epoch": 0.43901108889292856, + "grad_norm": 0.20566122233867645, + "learning_rate": 2.4145e-05, + "loss": 0.0568, + "step": 4830 + }, + { + "epoch": 0.4399200145428104, + "grad_norm": 0.25077319145202637, + "learning_rate": 2.4195e-05, + "loss": 0.0553, + "step": 4840 + }, + { + "epoch": 0.44082894019269225, + "grad_norm": 0.30497586727142334, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0531, + "step": 4850 + }, + { + "epoch": 0.4417378658425741, + "grad_norm": 0.29999080300331116, + "learning_rate": 2.4295e-05, + "loss": 0.0593, + "step": 4860 + }, + { + "epoch": 0.4426467914924559, + "grad_norm": 0.23176725208759308, + "learning_rate": 2.4345e-05, + "loss": 0.0555, + "step": 4870 + }, + { + "epoch": 0.44355571714233777, + "grad_norm": 0.22923621535301208, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0542, + "step": 4880 + }, + { + "epoch": 0.4444646427922196, + "grad_norm": 0.36732617020606995, + "learning_rate": 2.4445e-05, + "loss": 0.053, + "step": 4890 + }, + { + "epoch": 0.4453735684421014, + "grad_norm": 0.32465410232543945, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0533, + "step": 4900 + }, + { + "epoch": 0.4462824940919833, + "grad_norm": 0.28960302472114563, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.054, + "step": 4910 + }, + { + "epoch": 0.4471914197418651, + "grad_norm": 0.2625841200351715, + "learning_rate": 2.4595e-05, + "loss": 0.0568, + "step": 4920 + }, + { + "epoch": 0.448100345391747, + "grad_norm": 0.2876671850681305, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0563, + "step": 4930 + }, + { + "epoch": 0.4490092710416288, + "grad_norm": 0.3105892241001129, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0572, + "step": 4940 + }, + { + "epoch": 0.4499181966915106, + "grad_norm": 0.29491034150123596, + "learning_rate": 2.4745e-05, + "loss": 0.0598, + "step": 4950 + }, + { + "epoch": 0.4508271223413925, + "grad_norm": 0.311950147151947, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0565, + "step": 4960 + }, + { + "epoch": 0.4517360479912743, + "grad_norm": 0.2351275086402893, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0547, + "step": 4970 + }, + { + "epoch": 0.45264497364115613, + "grad_norm": 0.35425087809562683, + "learning_rate": 2.4895e-05, + "loss": 0.0568, + "step": 4980 + }, + { + "epoch": 0.453553899291038, + "grad_norm": 0.37492460012435913, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0539, + "step": 4990 + }, + { + "epoch": 0.45446282494091983, + "grad_norm": 0.2467612773180008, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0586, + "step": 5000 + }, + { + "epoch": 0.45537175059080165, + "grad_norm": 0.27256572246551514, + "learning_rate": 2.5045e-05, + "loss": 0.0566, + "step": 5010 + }, + { + "epoch": 0.4562806762406835, + "grad_norm": 0.313374400138855, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0564, + "step": 5020 + }, + { + "epoch": 0.45718960189056534, + "grad_norm": 0.2956707179546356, + "learning_rate": 2.5145e-05, + "loss": 0.0535, + "step": 5030 + }, + { + "epoch": 0.45809852754044716, + "grad_norm": 0.2674490213394165, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0551, + "step": 5040 + }, + { + "epoch": 0.45900745319032904, + "grad_norm": 0.3018375635147095, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0567, + "step": 5050 + }, + { + "epoch": 0.45991637884021086, + "grad_norm": 0.26326611638069153, + "learning_rate": 2.5295e-05, + "loss": 0.0559, + "step": 5060 + }, + { + "epoch": 0.46082530449009274, + "grad_norm": 0.288176029920578, + "learning_rate": 2.5345e-05, + "loss": 0.051, + "step": 5070 + }, + { + "epoch": 0.46173423013997456, + "grad_norm": 0.2510232925415039, + "learning_rate": 2.5395e-05, + "loss": 0.0521, + "step": 5080 + }, + { + "epoch": 0.4626431557898564, + "grad_norm": 0.21269504725933075, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0548, + "step": 5090 + }, + { + "epoch": 0.46355208143973825, + "grad_norm": 0.211288720369339, + "learning_rate": 2.5495e-05, + "loss": 0.0535, + "step": 5100 + }, + { + "epoch": 0.46446100708962007, + "grad_norm": 0.24636512994766235, + "learning_rate": 2.5545e-05, + "loss": 0.0549, + "step": 5110 + }, + { + "epoch": 0.4653699327395019, + "grad_norm": 0.26236575841903687, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0549, + "step": 5120 + }, + { + "epoch": 0.46627885838938377, + "grad_norm": 0.34995439648628235, + "learning_rate": 2.5645e-05, + "loss": 0.0544, + "step": 5130 + }, + { + "epoch": 0.4671877840392656, + "grad_norm": 0.28245946764945984, + "learning_rate": 2.5695e-05, + "loss": 0.058, + "step": 5140 + }, + { + "epoch": 0.4680967096891474, + "grad_norm": 0.26269882917404175, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0553, + "step": 5150 + }, + { + "epoch": 0.4690056353390293, + "grad_norm": 0.15957599878311157, + "learning_rate": 2.5795e-05, + "loss": 0.0572, + "step": 5160 + }, + { + "epoch": 0.4699145609889111, + "grad_norm": 0.21800892055034637, + "learning_rate": 2.5845e-05, + "loss": 0.0514, + "step": 5170 + }, + { + "epoch": 0.4708234866387929, + "grad_norm": 0.2567364573478699, + "learning_rate": 2.5895e-05, + "loss": 0.0487, + "step": 5180 + }, + { + "epoch": 0.4717324122886748, + "grad_norm": 0.28419357538223267, + "learning_rate": 2.5945e-05, + "loss": 0.0543, + "step": 5190 + }, + { + "epoch": 0.4726413379385566, + "grad_norm": 0.2827862501144409, + "learning_rate": 2.5995e-05, + "loss": 0.0556, + "step": 5200 + }, + { + "epoch": 0.4735502635884385, + "grad_norm": 0.29553109407424927, + "learning_rate": 2.6045e-05, + "loss": 0.0565, + "step": 5210 + }, + { + "epoch": 0.4744591892383203, + "grad_norm": 0.18970374763011932, + "learning_rate": 2.6095e-05, + "loss": 0.0529, + "step": 5220 + }, + { + "epoch": 0.47536811488820213, + "grad_norm": 0.2686721086502075, + "learning_rate": 2.6145e-05, + "loss": 0.0518, + "step": 5230 + }, + { + "epoch": 0.476277040538084, + "grad_norm": 0.30130910873413086, + "learning_rate": 2.6195e-05, + "loss": 0.0542, + "step": 5240 + }, + { + "epoch": 0.4771859661879658, + "grad_norm": 0.3451700210571289, + "learning_rate": 2.6245e-05, + "loss": 0.0551, + "step": 5250 + }, + { + "epoch": 0.47809489183784765, + "grad_norm": 0.30439358949661255, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0559, + "step": 5260 + }, + { + "epoch": 0.4790038174877295, + "grad_norm": 0.23240090906620026, + "learning_rate": 2.6345e-05, + "loss": 0.0569, + "step": 5270 + }, + { + "epoch": 0.47991274313761134, + "grad_norm": 0.2240114063024521, + "learning_rate": 2.6395e-05, + "loss": 0.052, + "step": 5280 + }, + { + "epoch": 0.48082166878749316, + "grad_norm": 0.2261715680360794, + "learning_rate": 2.6445e-05, + "loss": 0.0536, + "step": 5290 + }, + { + "epoch": 0.48173059443737504, + "grad_norm": 0.23731718957424164, + "learning_rate": 2.6495e-05, + "loss": 0.0567, + "step": 5300 + }, + { + "epoch": 0.48263952008725686, + "grad_norm": 0.28251802921295166, + "learning_rate": 2.6545e-05, + "loss": 0.0562, + "step": 5310 + }, + { + "epoch": 0.4835484457371387, + "grad_norm": 0.27070456743240356, + "learning_rate": 2.6595e-05, + "loss": 0.0544, + "step": 5320 + }, + { + "epoch": 0.48445737138702055, + "grad_norm": 0.24233168363571167, + "learning_rate": 2.6645e-05, + "loss": 0.0538, + "step": 5330 + }, + { + "epoch": 0.48536629703690237, + "grad_norm": 0.30124393105506897, + "learning_rate": 2.6695e-05, + "loss": 0.0528, + "step": 5340 + }, + { + "epoch": 0.48627522268678425, + "grad_norm": 0.31780317425727844, + "learning_rate": 2.6745e-05, + "loss": 0.0545, + "step": 5350 + }, + { + "epoch": 0.48718414833666607, + "grad_norm": 0.22772179543972015, + "learning_rate": 2.6795e-05, + "loss": 0.0514, + "step": 5360 + }, + { + "epoch": 0.4880930739865479, + "grad_norm": 0.2990436851978302, + "learning_rate": 2.6845e-05, + "loss": 0.0519, + "step": 5370 + }, + { + "epoch": 0.48900199963642976, + "grad_norm": 0.22937847673892975, + "learning_rate": 2.6895e-05, + "loss": 0.0548, + "step": 5380 + }, + { + "epoch": 0.4899109252863116, + "grad_norm": 0.30897074937820435, + "learning_rate": 2.6945e-05, + "loss": 0.0553, + "step": 5390 + }, + { + "epoch": 0.4908198509361934, + "grad_norm": 0.2935243844985962, + "learning_rate": 2.6995e-05, + "loss": 0.0547, + "step": 5400 + }, + { + "epoch": 0.4917287765860753, + "grad_norm": 0.2347809076309204, + "learning_rate": 2.7045e-05, + "loss": 0.0534, + "step": 5410 + }, + { + "epoch": 0.4926377022359571, + "grad_norm": 0.19506925344467163, + "learning_rate": 2.7095e-05, + "loss": 0.0581, + "step": 5420 + }, + { + "epoch": 0.4935466278858389, + "grad_norm": 0.27148714661598206, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0523, + "step": 5430 + }, + { + "epoch": 0.4944555535357208, + "grad_norm": 0.36605024337768555, + "learning_rate": 2.7195e-05, + "loss": 0.0533, + "step": 5440 + }, + { + "epoch": 0.4953644791856026, + "grad_norm": 0.3088148832321167, + "learning_rate": 2.7245e-05, + "loss": 0.0565, + "step": 5450 + }, + { + "epoch": 0.49627340483548443, + "grad_norm": 0.2857334613800049, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.051, + "step": 5460 + }, + { + "epoch": 0.4971823304853663, + "grad_norm": 0.32528722286224365, + "learning_rate": 2.7345e-05, + "loss": 0.0567, + "step": 5470 + }, + { + "epoch": 0.49809125613524813, + "grad_norm": 0.32941150665283203, + "learning_rate": 2.7395e-05, + "loss": 0.0554, + "step": 5480 + }, + { + "epoch": 0.49900018178513, + "grad_norm": 0.21631044149398804, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0524, + "step": 5490 + }, + { + "epoch": 0.4999091074350118, + "grad_norm": 0.28948724269866943, + "learning_rate": 2.7495e-05, + "loss": 0.0565, + "step": 5500 + }, + { + "epoch": 0.5008180330848937, + "grad_norm": 0.25311118364334106, + "learning_rate": 2.7545e-05, + "loss": 0.0526, + "step": 5510 + }, + { + "epoch": 0.5017269587347755, + "grad_norm": 0.22802378237247467, + "learning_rate": 2.7595e-05, + "loss": 0.0533, + "step": 5520 + }, + { + "epoch": 0.5026358843846573, + "grad_norm": 0.22559191286563873, + "learning_rate": 2.7645e-05, + "loss": 0.0502, + "step": 5530 + }, + { + "epoch": 0.5035448100345392, + "grad_norm": 0.26586025953292847, + "learning_rate": 2.7695e-05, + "loss": 0.0564, + "step": 5540 + }, + { + "epoch": 0.504453735684421, + "grad_norm": 0.2923668920993805, + "learning_rate": 2.7745e-05, + "loss": 0.0555, + "step": 5550 + }, + { + "epoch": 0.5053626613343029, + "grad_norm": 0.28769731521606445, + "learning_rate": 2.7795e-05, + "loss": 0.0559, + "step": 5560 + }, + { + "epoch": 0.5062715869841847, + "grad_norm": 0.29427459836006165, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0535, + "step": 5570 + }, + { + "epoch": 0.5071805126340665, + "grad_norm": 0.2913834750652313, + "learning_rate": 2.7895e-05, + "loss": 0.0595, + "step": 5580 + }, + { + "epoch": 0.5080894382839484, + "grad_norm": 0.3218977153301239, + "learning_rate": 2.7945e-05, + "loss": 0.0552, + "step": 5590 + }, + { + "epoch": 0.5089983639338302, + "grad_norm": 0.23592625558376312, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0582, + "step": 5600 + }, + { + "epoch": 0.509907289583712, + "grad_norm": 0.23905231058597565, + "learning_rate": 2.8045e-05, + "loss": 0.0599, + "step": 5610 + }, + { + "epoch": 0.5108162152335939, + "grad_norm": 0.24240872263908386, + "learning_rate": 2.8095e-05, + "loss": 0.0539, + "step": 5620 + }, + { + "epoch": 0.5117251408834758, + "grad_norm": 0.29280179738998413, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0574, + "step": 5630 + }, + { + "epoch": 0.5126340665333575, + "grad_norm": 0.25353094935417175, + "learning_rate": 2.8195e-05, + "loss": 0.0518, + "step": 5640 + }, + { + "epoch": 0.5135429921832394, + "grad_norm": 0.23786824941635132, + "learning_rate": 2.8245e-05, + "loss": 0.0544, + "step": 5650 + }, + { + "epoch": 0.5144519178331213, + "grad_norm": 0.24216972291469574, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0523, + "step": 5660 + }, + { + "epoch": 0.515360843483003, + "grad_norm": 0.23467372357845306, + "learning_rate": 2.8345e-05, + "loss": 0.0563, + "step": 5670 + }, + { + "epoch": 0.5162697691328849, + "grad_norm": 0.3132319748401642, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0531, + "step": 5680 + }, + { + "epoch": 0.5171786947827668, + "grad_norm": 0.18966040015220642, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0536, + "step": 5690 + }, + { + "epoch": 0.5180876204326486, + "grad_norm": 0.1636870801448822, + "learning_rate": 2.8495e-05, + "loss": 0.0526, + "step": 5700 + }, + { + "epoch": 0.5189965460825304, + "grad_norm": 0.24162857234477997, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0526, + "step": 5710 + }, + { + "epoch": 0.5199054717324123, + "grad_norm": 0.2182847112417221, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0546, + "step": 5720 + }, + { + "epoch": 0.5208143973822942, + "grad_norm": 0.2477782517671585, + "learning_rate": 2.8645e-05, + "loss": 0.0565, + "step": 5730 + }, + { + "epoch": 0.521723323032176, + "grad_norm": 0.23338963091373444, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0537, + "step": 5740 + }, + { + "epoch": 0.5226322486820578, + "grad_norm": 0.20637448132038116, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0511, + "step": 5750 + }, + { + "epoch": 0.5235411743319397, + "grad_norm": 0.3145206570625305, + "learning_rate": 2.8795e-05, + "loss": 0.0528, + "step": 5760 + }, + { + "epoch": 0.5244500999818215, + "grad_norm": 0.33072736859321594, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0506, + "step": 5770 + }, + { + "epoch": 0.5253590256317033, + "grad_norm": 0.22637803852558136, + "learning_rate": 2.8895e-05, + "loss": 0.0546, + "step": 5780 + }, + { + "epoch": 0.5262679512815852, + "grad_norm": 0.3604140281677246, + "learning_rate": 2.8945e-05, + "loss": 0.0511, + "step": 5790 + }, + { + "epoch": 0.527176876931467, + "grad_norm": 0.24904665350914001, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0553, + "step": 5800 + }, + { + "epoch": 0.5280858025813489, + "grad_norm": 0.22131119668483734, + "learning_rate": 2.9045e-05, + "loss": 0.0546, + "step": 5810 + }, + { + "epoch": 0.5289947282312307, + "grad_norm": 0.30118781328201294, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0548, + "step": 5820 + }, + { + "epoch": 0.5299036538811125, + "grad_norm": 0.22611406445503235, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0498, + "step": 5830 + }, + { + "epoch": 0.5308125795309944, + "grad_norm": 0.25792208313941956, + "learning_rate": 2.9195e-05, + "loss": 0.0544, + "step": 5840 + }, + { + "epoch": 0.5317215051808762, + "grad_norm": 0.24166078865528107, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.054, + "step": 5850 + }, + { + "epoch": 0.532630430830758, + "grad_norm": 0.2249297946691513, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0511, + "step": 5860 + }, + { + "epoch": 0.5335393564806399, + "grad_norm": 0.191885843873024, + "learning_rate": 2.9345e-05, + "loss": 0.049, + "step": 5870 + }, + { + "epoch": 0.5344482821305218, + "grad_norm": 0.28692665696144104, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0567, + "step": 5880 + }, + { + "epoch": 0.5353572077804035, + "grad_norm": 0.30722367763519287, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0485, + "step": 5890 + }, + { + "epoch": 0.5362661334302854, + "grad_norm": 0.27081725001335144, + "learning_rate": 2.9495e-05, + "loss": 0.0522, + "step": 5900 + }, + { + "epoch": 0.5371750590801673, + "grad_norm": 0.2730909585952759, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0557, + "step": 5910 + }, + { + "epoch": 0.538083984730049, + "grad_norm": 0.2598942518234253, + "learning_rate": 2.9595e-05, + "loss": 0.052, + "step": 5920 + }, + { + "epoch": 0.5389929103799309, + "grad_norm": 0.29466864466667175, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0593, + "step": 5930 + }, + { + "epoch": 0.5399018360298128, + "grad_norm": 0.24872232973575592, + "learning_rate": 2.9695e-05, + "loss": 0.053, + "step": 5940 + }, + { + "epoch": 0.5408107616796946, + "grad_norm": 0.24600301682949066, + "learning_rate": 2.9745e-05, + "loss": 0.0544, + "step": 5950 + }, + { + "epoch": 0.5417196873295764, + "grad_norm": 0.2332497388124466, + "learning_rate": 2.9795e-05, + "loss": 0.0523, + "step": 5960 + }, + { + "epoch": 0.5426286129794583, + "grad_norm": 0.32024553418159485, + "learning_rate": 2.9845e-05, + "loss": 0.052, + "step": 5970 + }, + { + "epoch": 0.5435375386293401, + "grad_norm": 0.31406399607658386, + "learning_rate": 2.9895e-05, + "loss": 0.0524, + "step": 5980 + }, + { + "epoch": 0.5444464642792219, + "grad_norm": 0.2713526487350464, + "learning_rate": 2.9945e-05, + "loss": 0.0528, + "step": 5990 + }, + { + "epoch": 0.5453553899291038, + "grad_norm": 0.225476935505867, + "learning_rate": 2.9995e-05, + "loss": 0.0516, + "step": 6000 + }, + { + "epoch": 0.5462643155789857, + "grad_norm": 0.3296831548213959, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0535, + "step": 6010 + }, + { + "epoch": 0.5471732412288675, + "grad_norm": 0.2923344671726227, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.0519, + "step": 6020 + }, + { + "epoch": 0.5480821668787493, + "grad_norm": 0.2960372567176819, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0547, + "step": 6030 + }, + { + "epoch": 0.5489910925286312, + "grad_norm": 0.3368293344974518, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.053, + "step": 6040 + }, + { + "epoch": 0.549900018178513, + "grad_norm": 0.3112386465072632, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0526, + "step": 6050 + }, + { + "epoch": 0.5508089438283948, + "grad_norm": 0.2775225043296814, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0521, + "step": 6060 + }, + { + "epoch": 0.5517178694782767, + "grad_norm": 0.2607252597808838, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0509, + "step": 6070 + }, + { + "epoch": 0.5526267951281585, + "grad_norm": 0.19151853024959564, + "learning_rate": 2.999996445280659e-05, + "loss": 0.055, + "step": 6080 + }, + { + "epoch": 0.5535357207780404, + "grad_norm": 0.28361013531684875, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0536, + "step": 6090 + }, + { + "epoch": 0.5544446464279222, + "grad_norm": 0.267088919878006, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0548, + "step": 6100 + }, + { + "epoch": 0.555353572077804, + "grad_norm": 0.24052470922470093, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0517, + "step": 6110 + }, + { + "epoch": 0.5562624977276859, + "grad_norm": 0.2628992199897766, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0546, + "step": 6120 + }, + { + "epoch": 0.5571714233775678, + "grad_norm": 0.28774577379226685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0511, + "step": 6130 + }, + { + "epoch": 0.5580803490274495, + "grad_norm": 0.2564772367477417, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0538, + "step": 6140 + }, + { + "epoch": 0.5589892746773314, + "grad_norm": 0.21377591788768768, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0541, + "step": 6150 + }, + { + "epoch": 0.5598982003272133, + "grad_norm": 0.211139976978302, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0519, + "step": 6160 + }, + { + "epoch": 0.560807125977095, + "grad_norm": 0.295835018157959, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0489, + "step": 6170 + }, + { + "epoch": 0.5617160516269769, + "grad_norm": 0.28520533442497253, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0534, + "step": 6180 + }, + { + "epoch": 0.5626249772768588, + "grad_norm": 0.2833225131034851, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0525, + "step": 6190 + }, + { + "epoch": 0.5635339029267405, + "grad_norm": 0.2616976797580719, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0539, + "step": 6200 + }, + { + "epoch": 0.5644428285766224, + "grad_norm": 0.2685079872608185, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0536, + "step": 6210 + }, + { + "epoch": 0.5653517542265043, + "grad_norm": 0.2503528892993927, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.053, + "step": 6220 + }, + { + "epoch": 0.5662606798763861, + "grad_norm": 0.2587029039859772, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0506, + "step": 6230 + }, + { + "epoch": 0.5671696055262679, + "grad_norm": 0.29137393832206726, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.051, + "step": 6240 + }, + { + "epoch": 0.5680785311761498, + "grad_norm": 0.23929867148399353, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0496, + "step": 6250 + }, + { + "epoch": 0.5689874568260316, + "grad_norm": 0.22135137021541595, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0488, + "step": 6260 + }, + { + "epoch": 0.5698963824759135, + "grad_norm": 0.206459179520607, + "learning_rate": 2.999958785135414e-05, + "loss": 0.047, + "step": 6270 + }, + { + "epoch": 0.5708053081257953, + "grad_norm": 0.2737051844596863, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0512, + "step": 6280 + }, + { + "epoch": 0.5717142337756771, + "grad_norm": 0.2880627512931824, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0527, + "step": 6290 + }, + { + "epoch": 0.572623159425559, + "grad_norm": 0.2576262652873993, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0521, + "step": 6300 + }, + { + "epoch": 0.5735320850754408, + "grad_norm": 0.27448371052742004, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0484, + "step": 6310 + }, + { + "epoch": 0.5744410107253227, + "grad_norm": 0.2709333598613739, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0539, + "step": 6320 + }, + { + "epoch": 0.5753499363752045, + "grad_norm": 0.36486339569091797, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0561, + "step": 6330 + }, + { + "epoch": 0.5762588620250864, + "grad_norm": 0.28731778264045715, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0557, + "step": 6340 + }, + { + "epoch": 0.5771677876749682, + "grad_norm": 0.31042540073394775, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0576, + "step": 6350 + }, + { + "epoch": 0.57807671332485, + "grad_norm": 0.33163073658943176, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0503, + "step": 6360 + }, + { + "epoch": 0.5789856389747319, + "grad_norm": 0.2830809950828552, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0529, + "step": 6370 + }, + { + "epoch": 0.5798945646246138, + "grad_norm": 0.1890731155872345, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.053, + "step": 6380 + }, + { + "epoch": 0.5808034902744955, + "grad_norm": 0.2011909782886505, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0526, + "step": 6390 + }, + { + "epoch": 0.5817124159243774, + "grad_norm": 0.2016221433877945, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0548, + "step": 6400 + }, + { + "epoch": 0.5826213415742593, + "grad_norm": 0.2815428376197815, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0549, + "step": 6410 + }, + { + "epoch": 0.583530267224141, + "grad_norm": 0.2956758737564087, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0527, + "step": 6420 + }, + { + "epoch": 0.5844391928740229, + "grad_norm": 0.30815815925598145, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0519, + "step": 6430 + }, + { + "epoch": 0.5853481185239048, + "grad_norm": 0.20665393769741058, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0508, + "step": 6440 + }, + { + "epoch": 0.5862570441737865, + "grad_norm": 0.282214879989624, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0495, + "step": 6450 + }, + { + "epoch": 0.5871659698236684, + "grad_norm": 0.27863621711730957, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0531, + "step": 6460 + }, + { + "epoch": 0.5880748954735503, + "grad_norm": 0.2622636556625366, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0523, + "step": 6470 + }, + { + "epoch": 0.5889838211234321, + "grad_norm": 0.23116838932037354, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0527, + "step": 6480 + }, + { + "epoch": 0.5898927467733139, + "grad_norm": 0.24401499330997467, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0545, + "step": 6490 + }, + { + "epoch": 0.5908016724231958, + "grad_norm": 0.22782346606254578, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0521, + "step": 6500 + }, + { + "epoch": 0.5917105980730776, + "grad_norm": 0.22244158387184143, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0532, + "step": 6510 + }, + { + "epoch": 0.5926195237229595, + "grad_norm": 0.19593307375907898, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0527, + "step": 6520 + }, + { + "epoch": 0.5935284493728413, + "grad_norm": 0.19490881264209747, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0538, + "step": 6530 + }, + { + "epoch": 0.5944373750227231, + "grad_norm": 0.22136439383029938, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0504, + "step": 6540 + }, + { + "epoch": 0.595346300672605, + "grad_norm": 0.19092866778373718, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0536, + "step": 6550 + }, + { + "epoch": 0.5962552263224868, + "grad_norm": 0.22099344432353973, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0529, + "step": 6560 + }, + { + "epoch": 0.5971641519723686, + "grad_norm": 0.24948260188102722, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0528, + "step": 6570 + }, + { + "epoch": 0.5980730776222505, + "grad_norm": 0.18987756967544556, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0512, + "step": 6580 + }, + { + "epoch": 0.5989820032721324, + "grad_norm": 0.2546764314174652, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0508, + "step": 6590 + }, + { + "epoch": 0.5998909289220142, + "grad_norm": 0.26777181029319763, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0522, + "step": 6600 + }, + { + "epoch": 0.600799854571896, + "grad_norm": 0.3361927270889282, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0552, + "step": 6610 + }, + { + "epoch": 0.6017087802217779, + "grad_norm": 0.17631341516971588, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0529, + "step": 6620 + }, + { + "epoch": 0.6026177058716597, + "grad_norm": 0.2523130178451538, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0539, + "step": 6630 + }, + { + "epoch": 0.6035266315215415, + "grad_norm": 0.23011784255504608, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0538, + "step": 6640 + }, + { + "epoch": 0.6044355571714234, + "grad_norm": 0.23234693706035614, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.057, + "step": 6650 + }, + { + "epoch": 0.6053444828213053, + "grad_norm": 0.19565199315547943, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 0.606253408471187, + "grad_norm": 0.31070753931999207, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0508, + "step": 6670 + }, + { + "epoch": 0.6071623341210689, + "grad_norm": 0.273070365190506, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0533, + "step": 6680 + }, + { + "epoch": 0.6080712597709508, + "grad_norm": 0.2648565173149109, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0486, + "step": 6690 + }, + { + "epoch": 0.6089801854208325, + "grad_norm": 0.22167012095451355, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0461, + "step": 6700 + }, + { + "epoch": 0.6098891110707144, + "grad_norm": 0.2196054458618164, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0499, + "step": 6710 + }, + { + "epoch": 0.6107980367205963, + "grad_norm": 0.24620504677295685, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0575, + "step": 6720 + }, + { + "epoch": 0.6117069623704781, + "grad_norm": 0.29269182682037354, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0541, + "step": 6730 + }, + { + "epoch": 0.6126158880203599, + "grad_norm": 0.2586071491241455, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0526, + "step": 6740 + }, + { + "epoch": 0.6135248136702418, + "grad_norm": 0.3286111056804657, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0579, + "step": 6750 + }, + { + "epoch": 0.6144337393201236, + "grad_norm": 0.2632269263267517, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0522, + "step": 6760 + }, + { + "epoch": 0.6153426649700054, + "grad_norm": 0.2108609676361084, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0541, + "step": 6770 + }, + { + "epoch": 0.6162515906198873, + "grad_norm": 0.2675747573375702, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0543, + "step": 6780 + }, + { + "epoch": 0.6171605162697691, + "grad_norm": 0.27146410942077637, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0494, + "step": 6790 + }, + { + "epoch": 0.618069441919651, + "grad_norm": 0.3198733627796173, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0511, + "step": 6800 + }, + { + "epoch": 0.6189783675695328, + "grad_norm": 0.20420151948928833, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0547, + "step": 6810 + }, + { + "epoch": 0.6198872932194146, + "grad_norm": 0.23398584127426147, + "learning_rate": 2.999617967180403e-05, + "loss": 0.0515, + "step": 6820 + }, + { + "epoch": 0.6207962188692965, + "grad_norm": 0.17402291297912598, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0515, + "step": 6830 + }, + { + "epoch": 0.6217051445191784, + "grad_norm": 0.27347826957702637, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0501, + "step": 6840 + }, + { + "epoch": 0.6226140701690601, + "grad_norm": 0.29578107595443726, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.053, + "step": 6850 + }, + { + "epoch": 0.623522995818942, + "grad_norm": 0.2776813507080078, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0516, + "step": 6860 + }, + { + "epoch": 0.6244319214688239, + "grad_norm": 0.2264457494020462, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0508, + "step": 6870 + }, + { + "epoch": 0.6253408471187057, + "grad_norm": 0.24953427910804749, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0533, + "step": 6880 + }, + { + "epoch": 0.6262497727685875, + "grad_norm": 0.2027846723794937, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0508, + "step": 6890 + }, + { + "epoch": 0.6271586984184694, + "grad_norm": 0.2090650349855423, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0539, + "step": 6900 + }, + { + "epoch": 0.6280676240683513, + "grad_norm": 0.19834332168102264, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0518, + "step": 6910 + }, + { + "epoch": 0.628976549718233, + "grad_norm": 0.23949551582336426, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0505, + "step": 6920 + }, + { + "epoch": 0.6298854753681149, + "grad_norm": 0.21028633415699005, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0489, + "step": 6930 + }, + { + "epoch": 0.6307944010179968, + "grad_norm": 0.2601897418498993, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0499, + "step": 6940 + }, + { + "epoch": 0.6317033266678785, + "grad_norm": 0.27245450019836426, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0543, + "step": 6950 + }, + { + "epoch": 0.6326122523177604, + "grad_norm": 0.3398328423500061, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.052, + "step": 6960 + }, + { + "epoch": 0.6335211779676423, + "grad_norm": 0.2867478132247925, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0481, + "step": 6970 + }, + { + "epoch": 0.634430103617524, + "grad_norm": 0.20872463285923004, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.053, + "step": 6980 + }, + { + "epoch": 0.6353390292674059, + "grad_norm": 0.23736810684204102, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0492, + "step": 6990 + }, + { + "epoch": 0.6362479549172878, + "grad_norm": 0.24111640453338623, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0509, + "step": 7000 + }, + { + "epoch": 0.6371568805671696, + "grad_norm": 0.19191798567771912, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0527, + "step": 7010 + }, + { + "epoch": 0.6380658062170514, + "grad_norm": 0.29575255513191223, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0497, + "step": 7020 + }, + { + "epoch": 0.6389747318669333, + "grad_norm": 0.2858394384384155, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0478, + "step": 7030 + }, + { + "epoch": 0.6398836575168151, + "grad_norm": 0.23822011053562164, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0517, + "step": 7040 + }, + { + "epoch": 0.640792583166697, + "grad_norm": 0.2668016254901886, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0493, + "step": 7050 + }, + { + "epoch": 0.6417015088165788, + "grad_norm": 0.2376893162727356, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.047, + "step": 7060 + }, + { + "epoch": 0.6426104344664606, + "grad_norm": 0.2323344349861145, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0539, + "step": 7070 + }, + { + "epoch": 0.6435193601163425, + "grad_norm": 0.20076020061969757, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0492, + "step": 7080 + }, + { + "epoch": 0.6444282857662244, + "grad_norm": 0.2175072431564331, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0521, + "step": 7090 + }, + { + "epoch": 0.6453372114161061, + "grad_norm": 0.21339580416679382, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0481, + "step": 7100 + }, + { + "epoch": 0.646246137065988, + "grad_norm": 0.2633015215396881, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0502, + "step": 7110 + }, + { + "epoch": 0.6471550627158699, + "grad_norm": 0.24931228160858154, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0522, + "step": 7120 + }, + { + "epoch": 0.6480639883657516, + "grad_norm": 0.18593566119670868, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0501, + "step": 7130 + }, + { + "epoch": 0.6489729140156335, + "grad_norm": 0.20997826755046844, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0511, + "step": 7140 + }, + { + "epoch": 0.6498818396655154, + "grad_norm": 0.27878162264823914, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0512, + "step": 7150 + }, + { + "epoch": 0.6507907653153971, + "grad_norm": 0.25732260942459106, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0534, + "step": 7160 + }, + { + "epoch": 0.651699690965279, + "grad_norm": 0.2165483683347702, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0507, + "step": 7170 + }, + { + "epoch": 0.6526086166151609, + "grad_norm": 0.33918997645378113, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0519, + "step": 7180 + }, + { + "epoch": 0.6535175422650428, + "grad_norm": 0.22547757625579834, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0527, + "step": 7190 + }, + { + "epoch": 0.6544264679149245, + "grad_norm": 0.22812825441360474, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0498, + "step": 7200 + }, + { + "epoch": 0.6553353935648064, + "grad_norm": 0.2200174331665039, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0492, + "step": 7210 + }, + { + "epoch": 0.6562443192146883, + "grad_norm": 0.2251828908920288, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0452, + "step": 7220 + }, + { + "epoch": 0.65715324486457, + "grad_norm": 0.21572019159793854, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0563, + "step": 7230 + }, + { + "epoch": 0.6580621705144519, + "grad_norm": 0.17474038898944855, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0521, + "step": 7240 + }, + { + "epoch": 0.6589710961643338, + "grad_norm": 0.2660854160785675, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0481, + "step": 7250 + }, + { + "epoch": 0.6598800218142156, + "grad_norm": 0.24693869054317474, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0501, + "step": 7260 + }, + { + "epoch": 0.6607889474640974, + "grad_norm": 0.21433234214782715, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0526, + "step": 7270 + }, + { + "epoch": 0.6616978731139793, + "grad_norm": 0.23004141449928284, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0507, + "step": 7280 + }, + { + "epoch": 0.6626067987638611, + "grad_norm": 0.25067004561424255, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0488, + "step": 7290 + }, + { + "epoch": 0.663515724413743, + "grad_norm": 0.18686044216156006, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0477, + "step": 7300 + }, + { + "epoch": 0.6644246500636248, + "grad_norm": 0.2626658082008362, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0494, + "step": 7310 + }, + { + "epoch": 0.6653335757135066, + "grad_norm": 0.1710175722837448, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0525, + "step": 7320 + }, + { + "epoch": 0.6662425013633885, + "grad_norm": 0.3010982871055603, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0517, + "step": 7330 + }, + { + "epoch": 0.6671514270132703, + "grad_norm": 0.25173676013946533, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0486, + "step": 7340 + }, + { + "epoch": 0.6680603526631521, + "grad_norm": 0.21248246729373932, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0499, + "step": 7350 + }, + { + "epoch": 0.668969278313034, + "grad_norm": 0.24406082928180695, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0522, + "step": 7360 + }, + { + "epoch": 0.6698782039629159, + "grad_norm": 0.2543180286884308, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0567, + "step": 7370 + }, + { + "epoch": 0.6707871296127976, + "grad_norm": 0.1851692497730255, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.052, + "step": 7380 + }, + { + "epoch": 0.6716960552626795, + "grad_norm": 0.3339602053165436, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0503, + "step": 7390 + }, + { + "epoch": 0.6726049809125614, + "grad_norm": 0.19371025264263153, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0475, + "step": 7400 + }, + { + "epoch": 0.6735139065624431, + "grad_norm": 0.26807984709739685, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0484, + "step": 7410 + }, + { + "epoch": 0.674422832212325, + "grad_norm": 0.2515440881252289, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0516, + "step": 7420 + }, + { + "epoch": 0.6753317578622069, + "grad_norm": 0.196345254778862, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0482, + "step": 7430 + }, + { + "epoch": 0.6762406835120887, + "grad_norm": 0.2170475572347641, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0508, + "step": 7440 + }, + { + "epoch": 0.6771496091619705, + "grad_norm": 0.2820887267589569, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0472, + "step": 7450 + }, + { + "epoch": 0.6780585348118524, + "grad_norm": 0.188895121216774, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.051, + "step": 7460 + }, + { + "epoch": 0.6789674604617343, + "grad_norm": 0.25495657324790955, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0533, + "step": 7470 + }, + { + "epoch": 0.679876386111616, + "grad_norm": 0.2511043846607208, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0518, + "step": 7480 + }, + { + "epoch": 0.6807853117614979, + "grad_norm": 0.2179972380399704, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0484, + "step": 7490 + }, + { + "epoch": 0.6816942374113798, + "grad_norm": 0.19715851545333862, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0497, + "step": 7500 + }, + { + "epoch": 0.6826031630612616, + "grad_norm": 0.24459494650363922, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0496, + "step": 7510 + }, + { + "epoch": 0.6835120887111434, + "grad_norm": 0.19235004484653473, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0537, + "step": 7520 + }, + { + "epoch": 0.6844210143610253, + "grad_norm": 0.20094437897205353, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0488, + "step": 7530 + }, + { + "epoch": 0.6853299400109071, + "grad_norm": 0.20099641382694244, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0482, + "step": 7540 + }, + { + "epoch": 0.686238865660789, + "grad_norm": 0.2369614839553833, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0471, + "step": 7550 + }, + { + "epoch": 0.6871477913106708, + "grad_norm": 0.2608358860015869, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0493, + "step": 7560 + }, + { + "epoch": 0.6880567169605526, + "grad_norm": 0.24400851130485535, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0477, + "step": 7570 + }, + { + "epoch": 0.6889656426104345, + "grad_norm": 0.23592688143253326, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0508, + "step": 7580 + }, + { + "epoch": 0.6898745682603163, + "grad_norm": 0.21962867677211761, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0471, + "step": 7590 + }, + { + "epoch": 0.6907834939101981, + "grad_norm": 0.20309974253177643, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.055, + "step": 7600 + }, + { + "epoch": 0.69169241956008, + "grad_norm": 0.2307950258255005, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0468, + "step": 7610 + }, + { + "epoch": 0.6926013452099619, + "grad_norm": 0.2553301751613617, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0479, + "step": 7620 + }, + { + "epoch": 0.6935102708598436, + "grad_norm": 0.2143932431936264, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0502, + "step": 7630 + }, + { + "epoch": 0.6944191965097255, + "grad_norm": 0.17689794301986694, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0501, + "step": 7640 + }, + { + "epoch": 0.6953281221596074, + "grad_norm": 0.27757132053375244, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0475, + "step": 7650 + }, + { + "epoch": 0.6962370478094891, + "grad_norm": 0.26202842593193054, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0528, + "step": 7660 + }, + { + "epoch": 0.697145973459371, + "grad_norm": 0.20695815980434418, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0532, + "step": 7670 + }, + { + "epoch": 0.6980548991092529, + "grad_norm": 0.21370723843574524, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0465, + "step": 7680 + }, + { + "epoch": 0.6989638247591347, + "grad_norm": 0.204051673412323, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0476, + "step": 7690 + }, + { + "epoch": 0.6998727504090165, + "grad_norm": 0.19897423684597015, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0518, + "step": 7700 + }, + { + "epoch": 0.7007816760588984, + "grad_norm": 0.17612123489379883, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0509, + "step": 7710 + }, + { + "epoch": 0.7016906017087802, + "grad_norm": 0.189670592546463, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0469, + "step": 7720 + }, + { + "epoch": 0.702599527358662, + "grad_norm": 0.22081419825553894, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0472, + "step": 7730 + }, + { + "epoch": 0.7035084530085439, + "grad_norm": 0.201465904712677, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0481, + "step": 7740 + }, + { + "epoch": 0.7044173786584257, + "grad_norm": 0.20923101902008057, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0487, + "step": 7750 + }, + { + "epoch": 0.7053263043083076, + "grad_norm": 0.22961045801639557, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0478, + "step": 7760 + }, + { + "epoch": 0.7062352299581894, + "grad_norm": 0.22327256202697754, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0538, + "step": 7770 + }, + { + "epoch": 0.7071441556080713, + "grad_norm": 0.2672477960586548, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0471, + "step": 7780 + }, + { + "epoch": 0.7080530812579531, + "grad_norm": 0.2685855031013489, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0479, + "step": 7790 + }, + { + "epoch": 0.708962006907835, + "grad_norm": 0.2251366823911667, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0493, + "step": 7800 + }, + { + "epoch": 0.7098709325577168, + "grad_norm": 0.22550442814826965, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0469, + "step": 7810 + }, + { + "epoch": 0.7107798582075986, + "grad_norm": 0.3008698523044586, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0452, + "step": 7820 + }, + { + "epoch": 0.7116887838574805, + "grad_norm": 0.22882160544395447, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0483, + "step": 7830 + }, + { + "epoch": 0.7125977095073623, + "grad_norm": 0.300363689661026, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0496, + "step": 7840 + }, + { + "epoch": 0.7135066351572441, + "grad_norm": 0.2283293902873993, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0495, + "step": 7850 + }, + { + "epoch": 0.714415560807126, + "grad_norm": 0.2325492948293686, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0501, + "step": 7860 + }, + { + "epoch": 0.7153244864570079, + "grad_norm": 0.19275642931461334, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.049, + "step": 7870 + }, + { + "epoch": 0.7162334121068896, + "grad_norm": 0.23403795063495636, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0456, + "step": 7880 + }, + { + "epoch": 0.7171423377567715, + "grad_norm": 0.2029605209827423, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0513, + "step": 7890 + }, + { + "epoch": 0.7180512634066534, + "grad_norm": 0.23442625999450684, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.048, + "step": 7900 + }, + { + "epoch": 0.7189601890565351, + "grad_norm": 0.2916816771030426, + "learning_rate": 2.997924785476766e-05, + "loss": 0.05, + "step": 7910 + }, + { + "epoch": 0.719869114706417, + "grad_norm": 0.2214813232421875, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0469, + "step": 7920 + }, + { + "epoch": 0.7207780403562989, + "grad_norm": 0.248961940407753, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0484, + "step": 7930 + }, + { + "epoch": 0.7216869660061807, + "grad_norm": 0.2931472361087799, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0497, + "step": 7940 + }, + { + "epoch": 0.7225958916560625, + "grad_norm": 0.18621587753295898, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0473, + "step": 7950 + }, + { + "epoch": 0.7235048173059444, + "grad_norm": 0.3402620553970337, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0505, + "step": 7960 + }, + { + "epoch": 0.7244137429558262, + "grad_norm": 0.17450954020023346, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0501, + "step": 7970 + }, + { + "epoch": 0.725322668605708, + "grad_norm": 0.2077232301235199, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0503, + "step": 7980 + }, + { + "epoch": 0.7262315942555899, + "grad_norm": 0.23998259007930756, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0487, + "step": 7990 + }, + { + "epoch": 0.7271405199054717, + "grad_norm": 0.17295002937316895, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0479, + "step": 8000 + }, + { + "epoch": 0.7280494455553536, + "grad_norm": 0.2706528604030609, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0485, + "step": 8010 + }, + { + "epoch": 0.7289583712052354, + "grad_norm": 0.17512111365795135, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0481, + "step": 8020 + }, + { + "epoch": 0.7298672968551172, + "grad_norm": 0.17323556542396545, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0478, + "step": 8030 + }, + { + "epoch": 0.7307762225049991, + "grad_norm": 0.21128128468990326, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0498, + "step": 8040 + }, + { + "epoch": 0.731685148154881, + "grad_norm": 0.18239738047122955, + "learning_rate": 2.997609328922287e-05, + "loss": 0.048, + "step": 8050 + }, + { + "epoch": 0.7325940738047628, + "grad_norm": 0.19546958804130554, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0513, + "step": 8060 + }, + { + "epoch": 0.7335029994546446, + "grad_norm": 0.35228589177131653, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0467, + "step": 8070 + }, + { + "epoch": 0.7344119251045265, + "grad_norm": 0.22077226638793945, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0494, + "step": 8080 + }, + { + "epoch": 0.7353208507544083, + "grad_norm": 0.1670738011598587, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0488, + "step": 8090 + }, + { + "epoch": 0.7362297764042901, + "grad_norm": 0.1707305610179901, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0489, + "step": 8100 + }, + { + "epoch": 0.737138702054172, + "grad_norm": 0.28278055787086487, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0477, + "step": 8110 + }, + { + "epoch": 0.7380476277040539, + "grad_norm": 0.27411752939224243, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.051, + "step": 8120 + }, + { + "epoch": 0.7389565533539356, + "grad_norm": 0.19234056770801544, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0515, + "step": 8130 + }, + { + "epoch": 0.7398654790038175, + "grad_norm": 0.1981707513332367, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0508, + "step": 8140 + }, + { + "epoch": 0.7407744046536994, + "grad_norm": 0.2401772141456604, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0525, + "step": 8150 + }, + { + "epoch": 0.7416833303035811, + "grad_norm": 0.28203579783439636, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0491, + "step": 8160 + }, + { + "epoch": 0.742592255953463, + "grad_norm": 0.25807827711105347, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0522, + "step": 8170 + }, + { + "epoch": 0.7435011816033449, + "grad_norm": 0.27139246463775635, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0507, + "step": 8180 + }, + { + "epoch": 0.7444101072532266, + "grad_norm": 0.18463386595249176, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0514, + "step": 8190 + }, + { + "epoch": 0.7453190329031085, + "grad_norm": 0.23065169155597687, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0478, + "step": 8200 + }, + { + "epoch": 0.7462279585529904, + "grad_norm": 0.25646573305130005, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0467, + "step": 8210 + }, + { + "epoch": 0.7471368842028722, + "grad_norm": 0.21268872916698456, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0472, + "step": 8220 + }, + { + "epoch": 0.748045809852754, + "grad_norm": 0.23608918488025665, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.0471, + "step": 8230 + }, + { + "epoch": 0.7489547355026359, + "grad_norm": 0.2650007903575897, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0477, + "step": 8240 + }, + { + "epoch": 0.7498636611525177, + "grad_norm": 0.2778491973876953, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0492, + "step": 8250 + }, + { + "epoch": 0.7507725868023996, + "grad_norm": 0.25588634610176086, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0477, + "step": 8260 + }, + { + "epoch": 0.7516815124522814, + "grad_norm": 0.2724026143550873, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0496, + "step": 8270 + }, + { + "epoch": 0.7525904381021632, + "grad_norm": 0.1649373322725296, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0487, + "step": 8280 + }, + { + "epoch": 0.7534993637520451, + "grad_norm": 0.20351797342300415, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0531, + "step": 8290 + }, + { + "epoch": 0.754408289401927, + "grad_norm": 0.21807296574115753, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0453, + "step": 8300 + }, + { + "epoch": 0.7553172150518087, + "grad_norm": 0.20078642666339874, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0511, + "step": 8310 + }, + { + "epoch": 0.7562261407016906, + "grad_norm": 0.25784561038017273, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0466, + "step": 8320 + }, + { + "epoch": 0.7571350663515725, + "grad_norm": 0.24861650168895721, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0466, + "step": 8330 + }, + { + "epoch": 0.7580439920014543, + "grad_norm": 0.18715667724609375, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0454, + "step": 8340 + }, + { + "epoch": 0.7589529176513361, + "grad_norm": 0.2418592870235443, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0466, + "step": 8350 + }, + { + "epoch": 0.759861843301218, + "grad_norm": 0.19177629053592682, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0464, + "step": 8360 + }, + { + "epoch": 0.7607707689510999, + "grad_norm": 0.228382408618927, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0439, + "step": 8370 + }, + { + "epoch": 0.7616796946009816, + "grad_norm": 0.20770283043384552, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0467, + "step": 8380 + }, + { + "epoch": 0.7625886202508635, + "grad_norm": 0.25008559226989746, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.046, + "step": 8390 + }, + { + "epoch": 0.7634975459007454, + "grad_norm": 0.19675347208976746, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0503, + "step": 8400 + }, + { + "epoch": 0.7644064715506271, + "grad_norm": 0.2154797613620758, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0511, + "step": 8410 + }, + { + "epoch": 0.765315397200509, + "grad_norm": 0.2747587263584137, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0468, + "step": 8420 + }, + { + "epoch": 0.7662243228503909, + "grad_norm": 0.17648565769195557, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.046, + "step": 8430 + }, + { + "epoch": 0.7671332485002726, + "grad_norm": 0.2432120442390442, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0513, + "step": 8440 + }, + { + "epoch": 0.7680421741501545, + "grad_norm": 0.20344844460487366, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0548, + "step": 8450 + }, + { + "epoch": 0.7689510998000364, + "grad_norm": 0.18204176425933838, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0476, + "step": 8460 + }, + { + "epoch": 0.7698600254499182, + "grad_norm": 0.23061425983905792, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.0489, + "step": 8470 + }, + { + "epoch": 0.7707689510998, + "grad_norm": 0.2090093046426773, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0501, + "step": 8480 + }, + { + "epoch": 0.7716778767496819, + "grad_norm": 0.22102810442447662, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0492, + "step": 8490 + }, + { + "epoch": 0.7725868023995637, + "grad_norm": 0.22209478914737701, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0515, + "step": 8500 + }, + { + "epoch": 0.7734957280494456, + "grad_norm": 0.2830013632774353, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0487, + "step": 8510 + }, + { + "epoch": 0.7744046536993274, + "grad_norm": 0.28471726179122925, + "learning_rate": 2.996387288904491e-05, + "loss": 0.048, + "step": 8520 + }, + { + "epoch": 0.7753135793492092, + "grad_norm": 0.19665536284446716, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0468, + "step": 8530 + }, + { + "epoch": 0.7762225049990911, + "grad_norm": 0.3020164370536804, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0527, + "step": 8540 + }, + { + "epoch": 0.7771314306489729, + "grad_norm": 0.1685352921485901, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0447, + "step": 8550 + }, + { + "epoch": 0.7780403562988547, + "grad_norm": 0.24027131497859955, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0508, + "step": 8560 + }, + { + "epoch": 0.7789492819487366, + "grad_norm": 0.21722637116909027, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0527, + "step": 8570 + }, + { + "epoch": 0.7798582075986185, + "grad_norm": 0.22285057604312897, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0476, + "step": 8580 + }, + { + "epoch": 0.7807671332485002, + "grad_norm": 0.1739153414964676, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0463, + "step": 8590 + }, + { + "epoch": 0.7816760588983821, + "grad_norm": 0.1855645477771759, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0457, + "step": 8600 + }, + { + "epoch": 0.782584984548264, + "grad_norm": 0.18880285322666168, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0463, + "step": 8610 + }, + { + "epoch": 0.7834939101981457, + "grad_norm": 0.25211361050605774, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0458, + "step": 8620 + }, + { + "epoch": 0.7844028358480276, + "grad_norm": 0.22930525243282318, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0486, + "step": 8630 + }, + { + "epoch": 0.7853117614979095, + "grad_norm": 0.23132479190826416, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0465, + "step": 8640 + }, + { + "epoch": 0.7862206871477914, + "grad_norm": 0.20086495578289032, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0496, + "step": 8650 + }, + { + "epoch": 0.7871296127976731, + "grad_norm": 0.18521204590797424, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0506, + "step": 8660 + }, + { + "epoch": 0.788038538447555, + "grad_norm": 0.19221846759319305, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0502, + "step": 8670 + }, + { + "epoch": 0.7889474640974369, + "grad_norm": 0.21987426280975342, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0478, + "step": 8680 + }, + { + "epoch": 0.7898563897473186, + "grad_norm": 0.21666979789733887, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0449, + "step": 8690 + }, + { + "epoch": 0.7907653153972005, + "grad_norm": 0.13873784244060516, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0457, + "step": 8700 + }, + { + "epoch": 0.7916742410470824, + "grad_norm": 0.21647080779075623, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0496, + "step": 8710 + }, + { + "epoch": 0.7925831666969642, + "grad_norm": 0.1657085120677948, + "learning_rate": 2.995791120254078e-05, + "loss": 0.047, + "step": 8720 + }, + { + "epoch": 0.793492092346846, + "grad_norm": 0.21663567423820496, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0447, + "step": 8730 + }, + { + "epoch": 0.7944010179967279, + "grad_norm": 0.16562850773334503, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0481, + "step": 8740 + }, + { + "epoch": 0.7953099436466097, + "grad_norm": 0.17849838733673096, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0443, + "step": 8750 + }, + { + "epoch": 0.7962188692964915, + "grad_norm": 0.1394987255334854, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0468, + "step": 8760 + }, + { + "epoch": 0.7971277949463734, + "grad_norm": 0.2402866780757904, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0518, + "step": 8770 + }, + { + "epoch": 0.7980367205962552, + "grad_norm": 0.2934916615486145, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0492, + "step": 8780 + }, + { + "epoch": 0.7989456462461371, + "grad_norm": 0.26684460043907166, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0479, + "step": 8790 + }, + { + "epoch": 0.7998545718960189, + "grad_norm": 0.20728403329849243, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.045, + "step": 8800 + }, + { + "epoch": 0.8007634975459007, + "grad_norm": 0.2823468744754791, + "learning_rate": 2.995508019056967e-05, + "loss": 0.045, + "step": 8810 + }, + { + "epoch": 0.8016724231957826, + "grad_norm": 0.2228637933731079, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0474, + "step": 8820 + }, + { + "epoch": 0.8025813488456645, + "grad_norm": 0.18181538581848145, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0498, + "step": 8830 + }, + { + "epoch": 0.8034902744955462, + "grad_norm": 0.23095960915088654, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0492, + "step": 8840 + }, + { + "epoch": 0.8043992001454281, + "grad_norm": 0.26155468821525574, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0509, + "step": 8850 + }, + { + "epoch": 0.80530812579531, + "grad_norm": 0.2795552611351013, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0464, + "step": 8860 + }, + { + "epoch": 0.8062170514451917, + "grad_norm": 0.21880106627941132, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0486, + "step": 8870 + }, + { + "epoch": 0.8071259770950736, + "grad_norm": 0.21262186765670776, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0451, + "step": 8880 + }, + { + "epoch": 0.8080349027449555, + "grad_norm": 0.23268140852451324, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.048, + "step": 8890 + }, + { + "epoch": 0.8089438283948372, + "grad_norm": 0.1915808916091919, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0469, + "step": 8900 + }, + { + "epoch": 0.8098527540447191, + "grad_norm": 0.18503227829933167, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0454, + "step": 8910 + }, + { + "epoch": 0.810761679694601, + "grad_norm": 0.252713680267334, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0492, + "step": 8920 + }, + { + "epoch": 0.8116706053444829, + "grad_norm": 0.23801177740097046, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0509, + "step": 8930 + }, + { + "epoch": 0.8125795309943646, + "grad_norm": 0.31841135025024414, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0512, + "step": 8940 + }, + { + "epoch": 0.8134884566442465, + "grad_norm": 0.20514997839927673, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0498, + "step": 8950 + }, + { + "epoch": 0.8143973822941284, + "grad_norm": 0.2710212469100952, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0523, + "step": 8960 + }, + { + "epoch": 0.8153063079440102, + "grad_norm": 0.21022585034370422, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0489, + "step": 8970 + }, + { + "epoch": 0.816215233593892, + "grad_norm": 0.1897699385881424, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0495, + "step": 8980 + }, + { + "epoch": 0.8171241592437739, + "grad_norm": 0.18747884035110474, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0504, + "step": 8990 + }, + { + "epoch": 0.8180330848936557, + "grad_norm": 0.20372244715690613, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0489, + "step": 9000 + }, + { + "epoch": 0.8189420105435375, + "grad_norm": 0.21477572619915009, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0485, + "step": 9010 + }, + { + "epoch": 0.8198509361934194, + "grad_norm": 0.2007143348455429, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0477, + "step": 9020 + }, + { + "epoch": 0.8207598618433012, + "grad_norm": 0.2711217999458313, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0516, + "step": 9030 + }, + { + "epoch": 0.8216687874931831, + "grad_norm": 0.23680776357650757, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0481, + "step": 9040 + }, + { + "epoch": 0.8225777131430649, + "grad_norm": 0.19621418416500092, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0464, + "step": 9050 + }, + { + "epoch": 0.8234866387929467, + "grad_norm": 0.32212772965431213, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0496, + "step": 9060 + }, + { + "epoch": 0.8243955644428286, + "grad_norm": 0.23581746220588684, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0496, + "step": 9070 + }, + { + "epoch": 0.8253044900927105, + "grad_norm": 0.2103641778230667, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0472, + "step": 9080 + }, + { + "epoch": 0.8262134157425922, + "grad_norm": 0.18274140357971191, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0498, + "step": 9090 + }, + { + "epoch": 0.8271223413924741, + "grad_norm": 0.19452445209026337, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0478, + "step": 9100 + }, + { + "epoch": 0.828031267042356, + "grad_norm": 0.18550024926662445, + "learning_rate": 2.994497917870469e-05, + "loss": 0.0469, + "step": 9110 + }, + { + "epoch": 0.8289401926922377, + "grad_norm": 0.18164171278476715, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0474, + "step": 9120 + }, + { + "epoch": 0.8298491183421196, + "grad_norm": 0.15937931835651398, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0474, + "step": 9130 + }, + { + "epoch": 0.8307580439920015, + "grad_norm": 0.15443098545074463, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0452, + "step": 9140 + }, + { + "epoch": 0.8316669696418832, + "grad_norm": 0.24238218367099762, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0492, + "step": 9150 + }, + { + "epoch": 0.8325758952917651, + "grad_norm": 0.22391052544116974, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0442, + "step": 9160 + }, + { + "epoch": 0.833484820941647, + "grad_norm": 0.1948886513710022, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0489, + "step": 9170 + }, + { + "epoch": 0.8343937465915288, + "grad_norm": 0.2669660449028015, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0458, + "step": 9180 + }, + { + "epoch": 0.8353026722414106, + "grad_norm": 0.2201799601316452, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0485, + "step": 9190 + }, + { + "epoch": 0.8362115978912925, + "grad_norm": 0.15754565596580505, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0453, + "step": 9200 + }, + { + "epoch": 0.8371205235411743, + "grad_norm": 0.18771110475063324, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0483, + "step": 9210 + }, + { + "epoch": 0.8380294491910562, + "grad_norm": 0.2588583827018738, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0489, + "step": 9220 + }, + { + "epoch": 0.838938374840938, + "grad_norm": 0.17708109319210052, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0482, + "step": 9230 + }, + { + "epoch": 0.8398473004908199, + "grad_norm": 0.26439717411994934, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0458, + "step": 9240 + }, + { + "epoch": 0.8407562261407017, + "grad_norm": 0.183246448636055, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0446, + "step": 9250 + }, + { + "epoch": 0.8416651517905835, + "grad_norm": 0.19444520771503448, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0483, + "step": 9260 + }, + { + "epoch": 0.8425740774404654, + "grad_norm": 0.24196045100688934, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0497, + "step": 9270 + }, + { + "epoch": 0.8434830030903472, + "grad_norm": 0.20626065135002136, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0489, + "step": 9280 + }, + { + "epoch": 0.8443919287402291, + "grad_norm": 0.18580278754234314, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0476, + "step": 9290 + }, + { + "epoch": 0.8453008543901109, + "grad_norm": 0.17508159577846527, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0489, + "step": 9300 + }, + { + "epoch": 0.8462097800399927, + "grad_norm": 0.18137440085411072, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0426, + "step": 9310 + }, + { + "epoch": 0.8471187056898746, + "grad_norm": 0.18413211405277252, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0448, + "step": 9320 + }, + { + "epoch": 0.8480276313397564, + "grad_norm": 0.20791150629520416, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0431, + "step": 9330 + }, + { + "epoch": 0.8489365569896382, + "grad_norm": 0.18125179409980774, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0497, + "step": 9340 + }, + { + "epoch": 0.8498454826395201, + "grad_norm": 0.268119752407074, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0492, + "step": 9350 + }, + { + "epoch": 0.850754408289402, + "grad_norm": 0.201127290725708, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0476, + "step": 9360 + }, + { + "epoch": 0.8516633339392837, + "grad_norm": 0.21040084958076477, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0478, + "step": 9370 + }, + { + "epoch": 0.8525722595891656, + "grad_norm": 0.21021278202533722, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0422, + "step": 9380 + }, + { + "epoch": 0.8534811852390475, + "grad_norm": 0.29613935947418213, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0459, + "step": 9390 + }, + { + "epoch": 0.8543901108889292, + "grad_norm": 0.23749980330467224, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0483, + "step": 9400 + }, + { + "epoch": 0.8552990365388111, + "grad_norm": 0.19182904064655304, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0463, + "step": 9410 + }, + { + "epoch": 0.856207962188693, + "grad_norm": 0.21992060542106628, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0455, + "step": 9420 + }, + { + "epoch": 0.8571168878385748, + "grad_norm": 0.2157416045665741, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0478, + "step": 9430 + }, + { + "epoch": 0.8580258134884566, + "grad_norm": 0.17094680666923523, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0481, + "step": 9440 + }, + { + "epoch": 0.8589347391383385, + "grad_norm": 0.20020833611488342, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0456, + "step": 9450 + }, + { + "epoch": 0.8598436647882203, + "grad_norm": 0.2103964388370514, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0485, + "step": 9460 + }, + { + "epoch": 0.8607525904381021, + "grad_norm": 0.19590336084365845, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0487, + "step": 9470 + }, + { + "epoch": 0.861661516087984, + "grad_norm": 0.21228598058223724, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0477, + "step": 9480 + }, + { + "epoch": 0.8625704417378658, + "grad_norm": 0.21234038472175598, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0461, + "step": 9490 + }, + { + "epoch": 0.8634793673877477, + "grad_norm": 0.20808303356170654, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0463, + "step": 9500 + }, + { + "epoch": 0.8643882930376295, + "grad_norm": 0.241983100771904, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0477, + "step": 9510 + }, + { + "epoch": 0.8652972186875114, + "grad_norm": 0.23247279226779938, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0478, + "step": 9520 + }, + { + "epoch": 0.8662061443373932, + "grad_norm": 0.19655448198318481, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0447, + "step": 9530 + }, + { + "epoch": 0.8671150699872751, + "grad_norm": 0.19026044011116028, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0483, + "step": 9540 + }, + { + "epoch": 0.8680239956371569, + "grad_norm": 0.24312818050384521, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0459, + "step": 9550 + }, + { + "epoch": 0.8689329212870387, + "grad_norm": 0.20392319560050964, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0454, + "step": 9560 + }, + { + "epoch": 0.8698418469369206, + "grad_norm": 0.20162178575992584, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0447, + "step": 9570 + }, + { + "epoch": 0.8707507725868024, + "grad_norm": 0.2064831703901291, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0491, + "step": 9580 + }, + { + "epoch": 0.8716596982366842, + "grad_norm": 0.25564029812812805, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0452, + "step": 9590 + }, + { + "epoch": 0.8725686238865661, + "grad_norm": 0.23017814755439758, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0476, + "step": 9600 + }, + { + "epoch": 0.873477549536448, + "grad_norm": 0.2335692197084427, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0437, + "step": 9610 + }, + { + "epoch": 0.8743864751863297, + "grad_norm": 0.19044488668441772, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0478, + "step": 9620 + }, + { + "epoch": 0.8752954008362116, + "grad_norm": 0.2835487127304077, + "learning_rate": 2.99250514479627e-05, + "loss": 0.0456, + "step": 9630 + }, + { + "epoch": 0.8762043264860935, + "grad_norm": 0.17049972712993622, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0461, + "step": 9640 + }, + { + "epoch": 0.8771132521359752, + "grad_norm": 0.18128736317157745, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0477, + "step": 9650 + }, + { + "epoch": 0.8780221777858571, + "grad_norm": 0.23346951603889465, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0439, + "step": 9660 + }, + { + "epoch": 0.878931103435739, + "grad_norm": 0.14180588722229004, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0447, + "step": 9670 + }, + { + "epoch": 0.8798400290856208, + "grad_norm": 0.2511714994907379, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0452, + "step": 9680 + }, + { + "epoch": 0.8807489547355026, + "grad_norm": 0.17788268625736237, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0446, + "step": 9690 + }, + { + "epoch": 0.8816578803853845, + "grad_norm": 0.24161937832832336, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0482, + "step": 9700 + }, + { + "epoch": 0.8825668060352663, + "grad_norm": 0.14893025159835815, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0459, + "step": 9710 + }, + { + "epoch": 0.8834757316851481, + "grad_norm": 0.15852811932563782, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0471, + "step": 9720 + }, + { + "epoch": 0.88438465733503, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0459, + "step": 9730 + }, + { + "epoch": 0.8852935829849118, + "grad_norm": 0.16526204347610474, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.051, + "step": 9740 + }, + { + "epoch": 0.8862025086347937, + "grad_norm": 0.18318583071231842, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0474, + "step": 9750 + }, + { + "epoch": 0.8871114342846755, + "grad_norm": 0.14728504419326782, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0446, + "step": 9760 + }, + { + "epoch": 0.8880203599345573, + "grad_norm": 0.22425974905490875, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0458, + "step": 9770 + }, + { + "epoch": 0.8889292855844392, + "grad_norm": 0.17587783932685852, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0503, + "step": 9780 + }, + { + "epoch": 0.889838211234321, + "grad_norm": 0.14566591382026672, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0496, + "step": 9790 + }, + { + "epoch": 0.8907471368842028, + "grad_norm": 0.2664507031440735, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0464, + "step": 9800 + }, + { + "epoch": 0.8916560625340847, + "grad_norm": 0.2102450132369995, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0429, + "step": 9810 + }, + { + "epoch": 0.8925649881839666, + "grad_norm": 0.27887192368507385, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0471, + "step": 9820 + }, + { + "epoch": 0.8934739138338484, + "grad_norm": 0.2712528109550476, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0467, + "step": 9830 + }, + { + "epoch": 0.8943828394837302, + "grad_norm": 0.24406598508358002, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0483, + "step": 9840 + }, + { + "epoch": 0.8952917651336121, + "grad_norm": 0.24617546796798706, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0475, + "step": 9850 + }, + { + "epoch": 0.896200690783494, + "grad_norm": 0.1893535703420639, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0446, + "step": 9860 + }, + { + "epoch": 0.8971096164333757, + "grad_norm": 0.23937766253948212, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0515, + "step": 9870 + }, + { + "epoch": 0.8980185420832576, + "grad_norm": 0.20585045218467712, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0463, + "step": 9880 + }, + { + "epoch": 0.8989274677331395, + "grad_norm": 0.21191814541816711, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0478, + "step": 9890 + }, + { + "epoch": 0.8998363933830212, + "grad_norm": 0.18468935787677765, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0461, + "step": 9900 + }, + { + "epoch": 0.9007453190329031, + "grad_norm": 0.21745678782463074, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0476, + "step": 9910 + }, + { + "epoch": 0.901654244682785, + "grad_norm": 0.24685350060462952, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0468, + "step": 9920 + }, + { + "epoch": 0.9025631703326668, + "grad_norm": 0.2032952457666397, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0517, + "step": 9930 + }, + { + "epoch": 0.9034720959825486, + "grad_norm": 0.2031661719083786, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0478, + "step": 9940 + }, + { + "epoch": 0.9043810216324305, + "grad_norm": 0.3068313002586365, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0483, + "step": 9950 + }, + { + "epoch": 0.9052899472823123, + "grad_norm": 0.2506161034107208, + "learning_rate": 2.991081506946978e-05, + "loss": 0.045, + "step": 9960 + }, + { + "epoch": 0.9061988729321941, + "grad_norm": 0.20954887568950653, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0427, + "step": 9970 + }, + { + "epoch": 0.907107798582076, + "grad_norm": 0.24890713393688202, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0446, + "step": 9980 + }, + { + "epoch": 0.9080167242319578, + "grad_norm": 0.20173248648643494, + "learning_rate": 2.990945968675917e-05, + "loss": 0.045, + "step": 9990 + }, + { + "epoch": 0.9089256498818397, + "grad_norm": 0.2182382345199585, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0448, + "step": 10000 + }, + { + "epoch": 0.9098345755317215, + "grad_norm": 0.19237685203552246, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0437, + "step": 10010 + }, + { + "epoch": 0.9107435011816033, + "grad_norm": 0.20224560797214508, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0447, + "step": 10020 + }, + { + "epoch": 0.9116524268314852, + "grad_norm": 0.2514999210834503, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0452, + "step": 10030 + }, + { + "epoch": 0.912561352481367, + "grad_norm": 0.20593084394931793, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0465, + "step": 10040 + }, + { + "epoch": 0.9134702781312488, + "grad_norm": 0.16114456951618195, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0447, + "step": 10050 + }, + { + "epoch": 0.9143792037811307, + "grad_norm": 0.22187383472919464, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0427, + "step": 10060 + }, + { + "epoch": 0.9152881294310126, + "grad_norm": 0.1887282431125641, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0433, + "step": 10070 + }, + { + "epoch": 0.9161970550808943, + "grad_norm": 0.17510150372982025, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0472, + "step": 10080 + }, + { + "epoch": 0.9171059807307762, + "grad_norm": 0.2080381065607071, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0447, + "step": 10090 + }, + { + "epoch": 0.9180149063806581, + "grad_norm": 0.24385763704776764, + "learning_rate": 2.990440277209787e-05, + "loss": 0.049, + "step": 10100 + }, + { + "epoch": 0.91892383203054, + "grad_norm": 0.19573915004730225, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0462, + "step": 10110 + }, + { + "epoch": 0.9198327576804217, + "grad_norm": 0.17442142963409424, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0423, + "step": 10120 + }, + { + "epoch": 0.9207416833303036, + "grad_norm": 0.20535624027252197, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0453, + "step": 10130 + }, + { + "epoch": 0.9216506089801855, + "grad_norm": 0.25228264927864075, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0468, + "step": 10140 + }, + { + "epoch": 0.9225595346300672, + "grad_norm": 0.26177623867988586, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0449, + "step": 10150 + }, + { + "epoch": 0.9234684602799491, + "grad_norm": 0.21755263209342957, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0428, + "step": 10160 + }, + { + "epoch": 0.924377385929831, + "grad_norm": 0.20552167296409607, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0489, + "step": 10170 + }, + { + "epoch": 0.9252863115797128, + "grad_norm": 0.2542552351951599, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.049, + "step": 10180 + }, + { + "epoch": 0.9261952372295946, + "grad_norm": 0.1901460438966751, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0419, + "step": 10190 + }, + { + "epoch": 0.9271041628794765, + "grad_norm": 0.20008572936058044, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0441, + "step": 10200 + }, + { + "epoch": 0.9280130885293583, + "grad_norm": 0.18924804031848907, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0426, + "step": 10210 + }, + { + "epoch": 0.9289220141792401, + "grad_norm": 0.1880718171596527, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0459, + "step": 10220 + }, + { + "epoch": 0.929830939829122, + "grad_norm": 0.2114868462085724, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0436, + "step": 10230 + }, + { + "epoch": 0.9307398654790038, + "grad_norm": 0.1772727370262146, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0395, + "step": 10240 + }, + { + "epoch": 0.9316487911288857, + "grad_norm": 0.2070719450712204, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0437, + "step": 10250 + }, + { + "epoch": 0.9325577167787675, + "grad_norm": 0.16719403862953186, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0455, + "step": 10260 + }, + { + "epoch": 0.9334666424286493, + "grad_norm": 0.1925038993358612, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0436, + "step": 10270 + }, + { + "epoch": 0.9343755680785312, + "grad_norm": 0.18015536665916443, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0473, + "step": 10280 + }, + { + "epoch": 0.935284493728413, + "grad_norm": 0.19839437305927277, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0481, + "step": 10290 + }, + { + "epoch": 0.9361934193782948, + "grad_norm": 0.18479220569133759, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0471, + "step": 10300 + }, + { + "epoch": 0.9371023450281767, + "grad_norm": 0.2188715636730194, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0463, + "step": 10310 + }, + { + "epoch": 0.9380112706780586, + "grad_norm": 0.2013077735900879, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0437, + "step": 10320 + }, + { + "epoch": 0.9389201963279403, + "grad_norm": 0.1927318274974823, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0477, + "step": 10330 + }, + { + "epoch": 0.9398291219778222, + "grad_norm": 0.230552077293396, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0447, + "step": 10340 + }, + { + "epoch": 0.9407380476277041, + "grad_norm": 0.2175379991531372, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0487, + "step": 10350 + }, + { + "epoch": 0.9416469732775858, + "grad_norm": 0.22253547608852386, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0459, + "step": 10360 + }, + { + "epoch": 0.9425558989274677, + "grad_norm": 0.1799159198999405, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0431, + "step": 10370 + }, + { + "epoch": 0.9434648245773496, + "grad_norm": 0.20050674676895142, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0444, + "step": 10380 + }, + { + "epoch": 0.9443737502272315, + "grad_norm": 0.23412321507930756, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0456, + "step": 10390 + }, + { + "epoch": 0.9452826758771132, + "grad_norm": 0.1978531926870346, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0492, + "step": 10400 + }, + { + "epoch": 0.9461916015269951, + "grad_norm": 0.23782093822956085, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0452, + "step": 10410 + }, + { + "epoch": 0.947100527176877, + "grad_norm": 0.2286175936460495, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0454, + "step": 10420 + }, + { + "epoch": 0.9480094528267587, + "grad_norm": 0.17889957129955292, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.045, + "step": 10430 + }, + { + "epoch": 0.9489183784766406, + "grad_norm": 0.23699882626533508, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0414, + "step": 10440 + }, + { + "epoch": 0.9498273041265225, + "grad_norm": 0.21840962767601013, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0426, + "step": 10450 + }, + { + "epoch": 0.9507362297764043, + "grad_norm": 0.21538928151130676, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0502, + "step": 10460 + }, + { + "epoch": 0.9516451554262861, + "grad_norm": 0.24738821387290955, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0449, + "step": 10470 + }, + { + "epoch": 0.952554081076168, + "grad_norm": 0.23165082931518555, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0442, + "step": 10480 + }, + { + "epoch": 0.9534630067260498, + "grad_norm": 0.24278849363327026, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0493, + "step": 10490 + }, + { + "epoch": 0.9543719323759317, + "grad_norm": 0.17292948067188263, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0499, + "step": 10500 + }, + { + "epoch": 0.9552808580258135, + "grad_norm": 0.22347436845302582, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0446, + "step": 10510 + }, + { + "epoch": 0.9561897836756953, + "grad_norm": 0.17118465900421143, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0443, + "step": 10520 + }, + { + "epoch": 0.9570987093255772, + "grad_norm": 0.24689707159996033, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0474, + "step": 10530 + }, + { + "epoch": 0.958007634975459, + "grad_norm": 0.2828623354434967, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0472, + "step": 10540 + }, + { + "epoch": 0.9589165606253408, + "grad_norm": 0.1843985617160797, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0485, + "step": 10550 + }, + { + "epoch": 0.9598254862752227, + "grad_norm": 0.2067168802022934, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0462, + "step": 10560 + }, + { + "epoch": 0.9607344119251046, + "grad_norm": 0.15282532572746277, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0462, + "step": 10570 + }, + { + "epoch": 0.9616433375749863, + "grad_norm": 0.18613895773887634, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0425, + "step": 10580 + }, + { + "epoch": 0.9625522632248682, + "grad_norm": 0.2089492380619049, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0459, + "step": 10590 + }, + { + "epoch": 0.9634611888747501, + "grad_norm": 0.27196231484413147, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.044, + "step": 10600 + }, + { + "epoch": 0.9643701145246318, + "grad_norm": 0.2824159264564514, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0452, + "step": 10610 + }, + { + "epoch": 0.9652790401745137, + "grad_norm": 0.3050622045993805, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0474, + "step": 10620 + }, + { + "epoch": 0.9661879658243956, + "grad_norm": 0.2229769080877304, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0474, + "step": 10630 + }, + { + "epoch": 0.9670968914742774, + "grad_norm": 0.23462875187397003, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0463, + "step": 10640 + }, + { + "epoch": 0.9680058171241592, + "grad_norm": 0.23895615339279175, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0448, + "step": 10650 + }, + { + "epoch": 0.9689147427740411, + "grad_norm": 0.15536090731620789, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0429, + "step": 10660 + }, + { + "epoch": 0.9698236684239229, + "grad_norm": 0.19151300191879272, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0459, + "step": 10670 + }, + { + "epoch": 0.9707325940738047, + "grad_norm": 0.26558300852775574, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0461, + "step": 10680 + }, + { + "epoch": 0.9716415197236866, + "grad_norm": 0.2432330846786499, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0445, + "step": 10690 + }, + { + "epoch": 0.9725504453735685, + "grad_norm": 0.2013140618801117, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0447, + "step": 10700 + }, + { + "epoch": 0.9734593710234503, + "grad_norm": 0.17192937433719635, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0464, + "step": 10710 + }, + { + "epoch": 0.9743682966733321, + "grad_norm": 0.1904268115758896, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0455, + "step": 10720 + }, + { + "epoch": 0.975277222323214, + "grad_norm": 0.2143162041902542, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0452, + "step": 10730 + }, + { + "epoch": 0.9761861479730958, + "grad_norm": 0.16544201970100403, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0461, + "step": 10740 + }, + { + "epoch": 0.9770950736229777, + "grad_norm": 0.16387546062469482, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0461, + "step": 10750 + }, + { + "epoch": 0.9780039992728595, + "grad_norm": 0.21320293843746185, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.044, + "step": 10760 + }, + { + "epoch": 0.9789129249227413, + "grad_norm": 0.13109347224235535, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0418, + "step": 10770 + }, + { + "epoch": 0.9798218505726232, + "grad_norm": 0.18552669882774353, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0428, + "step": 10780 + }, + { + "epoch": 0.980730776222505, + "grad_norm": 0.2472105622291565, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0431, + "step": 10790 + }, + { + "epoch": 0.9816397018723868, + "grad_norm": 0.2316143661737442, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0471, + "step": 10800 + }, + { + "epoch": 0.9825486275222687, + "grad_norm": 0.28511208295822144, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0489, + "step": 10810 + }, + { + "epoch": 0.9834575531721506, + "grad_norm": 0.21900852024555206, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0458, + "step": 10820 + }, + { + "epoch": 0.9843664788220323, + "grad_norm": 0.22348807752132416, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0429, + "step": 10830 + }, + { + "epoch": 0.9852754044719142, + "grad_norm": 0.22406747937202454, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0473, + "step": 10840 + }, + { + "epoch": 0.9861843301217961, + "grad_norm": 0.181293785572052, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0443, + "step": 10850 + }, + { + "epoch": 0.9870932557716778, + "grad_norm": 0.22763073444366455, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0446, + "step": 10860 + }, + { + "epoch": 0.9880021814215597, + "grad_norm": 0.2048628330230713, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0428, + "step": 10870 + }, + { + "epoch": 0.9889111070714416, + "grad_norm": 0.23177838325500488, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0447, + "step": 10880 + }, + { + "epoch": 0.9898200327213234, + "grad_norm": 0.17523930966854095, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0437, + "step": 10890 + }, + { + "epoch": 0.9907289583712052, + "grad_norm": 0.25708261132240295, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0428, + "step": 10900 + }, + { + "epoch": 0.9916378840210871, + "grad_norm": 0.21049687266349792, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0437, + "step": 10910 + }, + { + "epoch": 0.9925468096709689, + "grad_norm": 0.1825278103351593, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0433, + "step": 10920 + }, + { + "epoch": 0.9934557353208507, + "grad_norm": 0.20871856808662415, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0421, + "step": 10930 + }, + { + "epoch": 0.9943646609707326, + "grad_norm": 0.23490343987941742, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0472, + "step": 10940 + }, + { + "epoch": 0.9952735866206144, + "grad_norm": 0.17898419499397278, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0426, + "step": 10950 + }, + { + "epoch": 0.9961825122704963, + "grad_norm": 0.19226472079753876, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.044, + "step": 10960 + }, + { + "epoch": 0.9970914379203781, + "grad_norm": 0.2117571234703064, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0478, + "step": 10970 + }, + { + "epoch": 0.99800036357026, + "grad_norm": 0.18243448436260223, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0448, + "step": 10980 + }, + { + "epoch": 0.9989092892201418, + "grad_norm": 0.15876609086990356, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0479, + "step": 10990 + }, + { + "epoch": 0.9998182148700236, + "grad_norm": 0.21290788054466248, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0467, + "step": 11000 + }, + { + "epoch": 1.0007271405199054, + "grad_norm": 0.2566711902618408, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0426, + "step": 11010 + }, + { + "epoch": 1.0016360661697874, + "grad_norm": 0.1800299435853958, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0463, + "step": 11020 + }, + { + "epoch": 1.0025449918196692, + "grad_norm": 0.21893590688705444, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0411, + "step": 11030 + }, + { + "epoch": 1.003453917469551, + "grad_norm": 0.21869686245918274, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0449, + "step": 11040 + }, + { + "epoch": 1.004362843119433, + "grad_norm": 0.20672503113746643, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0451, + "step": 11050 + }, + { + "epoch": 1.0052717687693147, + "grad_norm": 0.21250903606414795, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0421, + "step": 11060 + }, + { + "epoch": 1.0061806944191964, + "grad_norm": 0.1933600902557373, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0448, + "step": 11070 + }, + { + "epoch": 1.0070896200690784, + "grad_norm": 0.27677804231643677, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0454, + "step": 11080 + }, + { + "epoch": 1.0079985457189602, + "grad_norm": 0.2600491940975189, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0446, + "step": 11090 + }, + { + "epoch": 1.008907471368842, + "grad_norm": 0.17764927446842194, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0426, + "step": 11100 + }, + { + "epoch": 1.009816397018724, + "grad_norm": 0.17611724138259888, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0441, + "step": 11110 + }, + { + "epoch": 1.0107253226686057, + "grad_norm": 0.1807049959897995, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0437, + "step": 11120 + }, + { + "epoch": 1.0116342483184875, + "grad_norm": 0.22016455233097076, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0455, + "step": 11130 + }, + { + "epoch": 1.0125431739683695, + "grad_norm": 0.22346694767475128, + "learning_rate": 2.984983021314809e-05, + "loss": 0.044, + "step": 11140 + }, + { + "epoch": 1.0134520996182512, + "grad_norm": 0.20554766058921814, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0417, + "step": 11150 + }, + { + "epoch": 1.014361025268133, + "grad_norm": 0.16440682113170624, + "learning_rate": 2.984866104806282e-05, + "loss": 0.044, + "step": 11160 + }, + { + "epoch": 1.015269950918015, + "grad_norm": 0.20076526701450348, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0428, + "step": 11170 + }, + { + "epoch": 1.0161788765678967, + "grad_norm": 0.20911820232868195, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0407, + "step": 11180 + }, + { + "epoch": 1.0170878022177785, + "grad_norm": 0.13757623732089996, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0474, + "step": 11190 + }, + { + "epoch": 1.0179967278676605, + "grad_norm": 0.20908738672733307, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0431, + "step": 11200 + }, + { + "epoch": 1.0189056535175423, + "grad_norm": 0.24481844902038574, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0443, + "step": 11210 + }, + { + "epoch": 1.019814579167424, + "grad_norm": 0.15947861969470978, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0407, + "step": 11220 + }, + { + "epoch": 1.020723504817306, + "grad_norm": 0.20907928049564362, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0453, + "step": 11230 + }, + { + "epoch": 1.0216324304671878, + "grad_norm": 0.19343087077140808, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0424, + "step": 11240 + }, + { + "epoch": 1.0225413561170695, + "grad_norm": 0.24570955336093903, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0463, + "step": 11250 + }, + { + "epoch": 1.0234502817669515, + "grad_norm": 0.19881205260753632, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0437, + "step": 11260 + }, + { + "epoch": 1.0243592074168333, + "grad_norm": 0.1605815291404724, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0438, + "step": 11270 + }, + { + "epoch": 1.025268133066715, + "grad_norm": 0.17117278277873993, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0418, + "step": 11280 + }, + { + "epoch": 1.026177058716597, + "grad_norm": 0.22824905812740326, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0425, + "step": 11290 + }, + { + "epoch": 1.0270859843664788, + "grad_norm": 0.1764717698097229, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0438, + "step": 11300 + }, + { + "epoch": 1.0279949100163606, + "grad_norm": 0.2728503346443176, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0443, + "step": 11310 + }, + { + "epoch": 1.0289038356662425, + "grad_norm": 0.2113536149263382, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0421, + "step": 11320 + }, + { + "epoch": 1.0298127613161243, + "grad_norm": 0.18083985149860382, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0452, + "step": 11330 + }, + { + "epoch": 1.030721686966006, + "grad_norm": 0.20122520625591278, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0446, + "step": 11340 + }, + { + "epoch": 1.031630612615888, + "grad_norm": 0.18695023655891418, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0464, + "step": 11350 + }, + { + "epoch": 1.0325395382657698, + "grad_norm": 0.1918337047100067, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0434, + "step": 11360 + }, + { + "epoch": 1.0334484639156516, + "grad_norm": 0.1699228733778, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0426, + "step": 11370 + }, + { + "epoch": 1.0343573895655336, + "grad_norm": 0.21208246052265167, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0436, + "step": 11380 + }, + { + "epoch": 1.0352663152154153, + "grad_norm": 0.2171522080898285, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0444, + "step": 11390 + }, + { + "epoch": 1.036175240865297, + "grad_norm": 0.1806924194097519, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0458, + "step": 11400 + }, + { + "epoch": 1.037084166515179, + "grad_norm": 0.15393805503845215, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0429, + "step": 11410 + }, + { + "epoch": 1.0379930921650609, + "grad_norm": 0.22041921317577362, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0401, + "step": 11420 + }, + { + "epoch": 1.0389020178149426, + "grad_norm": 0.1836814135313034, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0413, + "step": 11430 + }, + { + "epoch": 1.0398109434648246, + "grad_norm": 0.19501078128814697, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0445, + "step": 11440 + }, + { + "epoch": 1.0407198691147064, + "grad_norm": 0.2243889570236206, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0444, + "step": 11450 + }, + { + "epoch": 1.0416287947645884, + "grad_norm": 0.19588007032871246, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0446, + "step": 11460 + }, + { + "epoch": 1.0425377204144701, + "grad_norm": 0.17626333236694336, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0382, + "step": 11470 + }, + { + "epoch": 1.043446646064352, + "grad_norm": 0.13355682790279388, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0429, + "step": 11480 + }, + { + "epoch": 1.0443555717142339, + "grad_norm": 0.1756056398153305, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0435, + "step": 11490 + }, + { + "epoch": 1.0452644973641156, + "grad_norm": 0.19193421304225922, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0428, + "step": 11500 + }, + { + "epoch": 1.0461734230139974, + "grad_norm": 0.19427503645420074, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0427, + "step": 11510 + }, + { + "epoch": 1.0470823486638794, + "grad_norm": 0.20457762479782104, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0482, + "step": 11520 + }, + { + "epoch": 1.0479912743137612, + "grad_norm": 0.2177312821149826, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0476, + "step": 11530 + }, + { + "epoch": 1.048900199963643, + "grad_norm": 0.2705530822277069, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0498, + "step": 11540 + }, + { + "epoch": 1.049809125613525, + "grad_norm": 0.19399431347846985, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0469, + "step": 11550 + }, + { + "epoch": 1.0507180512634067, + "grad_norm": 0.24318693578243256, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0452, + "step": 11560 + }, + { + "epoch": 1.0516269769132884, + "grad_norm": 0.17799678444862366, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0431, + "step": 11570 + }, + { + "epoch": 1.0525359025631704, + "grad_norm": 0.16420628130435944, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0454, + "step": 11580 + }, + { + "epoch": 1.0534448282130522, + "grad_norm": 0.19751368463039398, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0457, + "step": 11590 + }, + { + "epoch": 1.054353753862934, + "grad_norm": 0.1574273705482483, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0443, + "step": 11600 + }, + { + "epoch": 1.055262679512816, + "grad_norm": 0.1651993989944458, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0488, + "step": 11610 + }, + { + "epoch": 1.0561716051626977, + "grad_norm": 0.16538509726524353, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0503, + "step": 11620 + }, + { + "epoch": 1.0570805308125795, + "grad_norm": 0.14782948791980743, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0504, + "step": 11630 + }, + { + "epoch": 1.0579894564624615, + "grad_norm": 0.14585520327091217, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0442, + "step": 11640 + }, + { + "epoch": 1.0588983821123432, + "grad_norm": 0.2442912608385086, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0437, + "step": 11650 + }, + { + "epoch": 1.059807307762225, + "grad_norm": 0.21426397562026978, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0424, + "step": 11660 + }, + { + "epoch": 1.060716233412107, + "grad_norm": 0.23297074437141418, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.046, + "step": 11670 + }, + { + "epoch": 1.0616251590619887, + "grad_norm": 0.1723955124616623, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0386, + "step": 11680 + }, + { + "epoch": 1.0625340847118705, + "grad_norm": 0.177279993891716, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0437, + "step": 11690 + }, + { + "epoch": 1.0634430103617525, + "grad_norm": 0.28594619035720825, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0447, + "step": 11700 + }, + { + "epoch": 1.0643519360116342, + "grad_norm": 0.1923300325870514, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0446, + "step": 11710 + }, + { + "epoch": 1.065260861661516, + "grad_norm": 0.2577943503856659, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0433, + "step": 11720 + }, + { + "epoch": 1.066169787311398, + "grad_norm": 0.2285197377204895, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.043, + "step": 11730 + }, + { + "epoch": 1.0670787129612798, + "grad_norm": 0.19768570363521576, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0446, + "step": 11740 + }, + { + "epoch": 1.0679876386111615, + "grad_norm": 0.24909929931163788, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0446, + "step": 11750 + }, + { + "epoch": 1.0688965642610435, + "grad_norm": 0.1871173232793808, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0437, + "step": 11760 + }, + { + "epoch": 1.0698054899109253, + "grad_norm": 0.13640782237052917, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.042, + "step": 11770 + }, + { + "epoch": 1.070714415560807, + "grad_norm": 0.2185574769973755, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0424, + "step": 11780 + }, + { + "epoch": 1.071623341210689, + "grad_norm": 0.21180127561092377, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0446, + "step": 11790 + }, + { + "epoch": 1.0725322668605708, + "grad_norm": 0.20700682699680328, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0447, + "step": 11800 + }, + { + "epoch": 1.0734411925104526, + "grad_norm": 0.2306687831878662, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0465, + "step": 11810 + }, + { + "epoch": 1.0743501181603345, + "grad_norm": 0.20409682393074036, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0452, + "step": 11820 + }, + { + "epoch": 1.0752590438102163, + "grad_norm": 0.1386854201555252, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.045, + "step": 11830 + }, + { + "epoch": 1.076167969460098, + "grad_norm": 0.1727799028158188, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0458, + "step": 11840 + }, + { + "epoch": 1.07707689510998, + "grad_norm": 0.1405276507139206, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0436, + "step": 11850 + }, + { + "epoch": 1.0779858207598618, + "grad_norm": 0.16839808225631714, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0424, + "step": 11860 + }, + { + "epoch": 1.0788947464097436, + "grad_norm": 0.14023159444332123, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0414, + "step": 11870 + }, + { + "epoch": 1.0798036720596256, + "grad_norm": 0.19566203653812408, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0435, + "step": 11880 + }, + { + "epoch": 1.0807125977095073, + "grad_norm": 0.1841118335723877, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0445, + "step": 11890 + }, + { + "epoch": 1.081621523359389, + "grad_norm": 0.26612839102745056, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0431, + "step": 11900 + }, + { + "epoch": 1.082530449009271, + "grad_norm": 0.36272507905960083, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0434, + "step": 11910 + }, + { + "epoch": 1.0834393746591529, + "grad_norm": 0.27770015597343445, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.044, + "step": 11920 + }, + { + "epoch": 1.0843483003090346, + "grad_norm": 0.16324889659881592, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0482, + "step": 11930 + }, + { + "epoch": 1.0852572259589166, + "grad_norm": 0.1590684950351715, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0483, + "step": 11940 + }, + { + "epoch": 1.0861661516087984, + "grad_norm": 0.16403186321258545, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0442, + "step": 11950 + }, + { + "epoch": 1.0870750772586804, + "grad_norm": 0.1818169355392456, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0483, + "step": 11960 + }, + { + "epoch": 1.0879840029085621, + "grad_norm": 0.20558829605579376, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0408, + "step": 11970 + }, + { + "epoch": 1.0888929285584439, + "grad_norm": 0.20835480093955994, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0449, + "step": 11980 + }, + { + "epoch": 1.0898018542083259, + "grad_norm": 0.2564977705478668, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0428, + "step": 11990 + }, + { + "epoch": 1.0907107798582076, + "grad_norm": 0.17838118970394135, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0464, + "step": 12000 + }, + { + "epoch": 1.0916197055080894, + "grad_norm": 0.19285725057125092, + "learning_rate": 2.979480675356209e-05, + "loss": 0.045, + "step": 12010 + }, + { + "epoch": 1.0925286311579714, + "grad_norm": 0.24221763014793396, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0456, + "step": 12020 + }, + { + "epoch": 1.0934375568078532, + "grad_norm": 0.27363356947898865, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.044, + "step": 12030 + }, + { + "epoch": 1.094346482457735, + "grad_norm": 0.23911097645759583, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0433, + "step": 12040 + }, + { + "epoch": 1.095255408107617, + "grad_norm": 0.22225715219974518, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0443, + "step": 12050 + }, + { + "epoch": 1.0961643337574987, + "grad_norm": 0.22100894153118134, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0444, + "step": 12060 + }, + { + "epoch": 1.0970732594073804, + "grad_norm": 0.20416633784770966, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0414, + "step": 12070 + }, + { + "epoch": 1.0979821850572624, + "grad_norm": 0.24297885596752167, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0442, + "step": 12080 + }, + { + "epoch": 1.0988911107071442, + "grad_norm": 0.17204557359218597, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0426, + "step": 12090 + }, + { + "epoch": 1.099800036357026, + "grad_norm": 0.13874132931232452, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0445, + "step": 12100 + }, + { + "epoch": 1.100708962006908, + "grad_norm": 0.1713542491197586, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0449, + "step": 12110 + }, + { + "epoch": 1.1016178876567897, + "grad_norm": 0.18085192143917084, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.045, + "step": 12120 + }, + { + "epoch": 1.1025268133066715, + "grad_norm": 0.19548489153385162, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0418, + "step": 12130 + }, + { + "epoch": 1.1034357389565534, + "grad_norm": 0.166118323802948, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0409, + "step": 12140 + }, + { + "epoch": 1.1043446646064352, + "grad_norm": 0.19054511189460754, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0422, + "step": 12150 + }, + { + "epoch": 1.105253590256317, + "grad_norm": 0.23155075311660767, + "learning_rate": 2.978445950595551e-05, + "loss": 0.045, + "step": 12160 + }, + { + "epoch": 1.106162515906199, + "grad_norm": 0.19337330758571625, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0415, + "step": 12170 + }, + { + "epoch": 1.1070714415560807, + "grad_norm": 0.18185372650623322, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0451, + "step": 12180 + }, + { + "epoch": 1.1079803672059625, + "grad_norm": 0.21236933767795563, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0406, + "step": 12190 + }, + { + "epoch": 1.1088892928558445, + "grad_norm": 0.18492180109024048, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0462, + "step": 12200 + }, + { + "epoch": 1.1097982185057262, + "grad_norm": 0.185042142868042, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0473, + "step": 12210 + }, + { + "epoch": 1.110707144155608, + "grad_norm": 0.16674759984016418, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0418, + "step": 12220 + }, + { + "epoch": 1.11161606980549, + "grad_norm": 0.16401632130146027, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0422, + "step": 12230 + }, + { + "epoch": 1.1125249954553718, + "grad_norm": 0.21320495009422302, + "learning_rate": 2.977883766666944e-05, + "loss": 0.046, + "step": 12240 + }, + { + "epoch": 1.1134339211052535, + "grad_norm": 0.16008056700229645, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0421, + "step": 12250 + }, + { + "epoch": 1.1143428467551355, + "grad_norm": 0.18554730713367462, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0427, + "step": 12260 + }, + { + "epoch": 1.1152517724050173, + "grad_norm": 0.1725737601518631, + "learning_rate": 2.977671095660518e-05, + "loss": 0.048, + "step": 12270 + }, + { + "epoch": 1.116160698054899, + "grad_norm": 0.20487797260284424, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0412, + "step": 12280 + }, + { + "epoch": 1.117069623704781, + "grad_norm": 0.176766037940979, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0399, + "step": 12290 + }, + { + "epoch": 1.1179785493546628, + "grad_norm": 0.19212746620178223, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0448, + "step": 12300 + }, + { + "epoch": 1.1188874750045446, + "grad_norm": 0.25957271456718445, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0452, + "step": 12310 + }, + { + "epoch": 1.1197964006544265, + "grad_norm": 0.2723216414451599, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.041, + "step": 12320 + }, + { + "epoch": 1.1207053263043083, + "grad_norm": 0.16132424771785736, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0424, + "step": 12330 + }, + { + "epoch": 1.12161425195419, + "grad_norm": 0.15822091698646545, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0426, + "step": 12340 + }, + { + "epoch": 1.122523177604072, + "grad_norm": 0.17679046094417572, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0443, + "step": 12350 + }, + { + "epoch": 1.1234321032539538, + "grad_norm": 0.1604025512933731, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0427, + "step": 12360 + }, + { + "epoch": 1.1243410289038356, + "grad_norm": 0.16356003284454346, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.042, + "step": 12370 + }, + { + "epoch": 1.1252499545537176, + "grad_norm": 0.15330560505390167, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0413, + "step": 12380 + }, + { + "epoch": 1.1261588802035993, + "grad_norm": 0.14955934882164001, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0413, + "step": 12390 + }, + { + "epoch": 1.127067805853481, + "grad_norm": 0.17660032212734222, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0416, + "step": 12400 + }, + { + "epoch": 1.127976731503363, + "grad_norm": 0.23531347513198853, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0385, + "step": 12410 + }, + { + "epoch": 1.1288856571532448, + "grad_norm": 0.20567849278450012, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0462, + "step": 12420 + }, + { + "epoch": 1.1297945828031266, + "grad_norm": 0.1494264304637909, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0447, + "step": 12430 + }, + { + "epoch": 1.1307035084530086, + "grad_norm": 0.1753614991903305, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0441, + "step": 12440 + }, + { + "epoch": 1.1316124341028904, + "grad_norm": 0.18637576699256897, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0418, + "step": 12450 + }, + { + "epoch": 1.1325213597527721, + "grad_norm": 0.2558162808418274, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0417, + "step": 12460 + }, + { + "epoch": 1.1334302854026541, + "grad_norm": 0.16048744320869446, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0443, + "step": 12470 + }, + { + "epoch": 1.1343392110525359, + "grad_norm": 0.15510894358158112, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0413, + "step": 12480 + }, + { + "epoch": 1.1352481367024176, + "grad_norm": 0.23833104968070984, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0453, + "step": 12490 + }, + { + "epoch": 1.1361570623522996, + "grad_norm": 0.22778794169425964, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0421, + "step": 12500 + }, + { + "epoch": 1.1370659880021814, + "grad_norm": 0.13740932941436768, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0471, + "step": 12510 + }, + { + "epoch": 1.1379749136520632, + "grad_norm": 0.18217132985591888, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0398, + "step": 12520 + }, + { + "epoch": 1.1388838393019451, + "grad_norm": 0.18237634003162384, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0453, + "step": 12530 + }, + { + "epoch": 1.139792764951827, + "grad_norm": 0.16545413434505463, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0431, + "step": 12540 + }, + { + "epoch": 1.1407016906017087, + "grad_norm": 0.19325804710388184, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0425, + "step": 12550 + }, + { + "epoch": 1.1416106162515907, + "grad_norm": 0.18488851189613342, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0462, + "step": 12560 + }, + { + "epoch": 1.1425195419014724, + "grad_norm": 0.172540083527565, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0451, + "step": 12570 + }, + { + "epoch": 1.1434284675513542, + "grad_norm": 0.19761130213737488, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0436, + "step": 12580 + }, + { + "epoch": 1.1443373932012362, + "grad_norm": 0.14343398809432983, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0419, + "step": 12590 + }, + { + "epoch": 1.145246318851118, + "grad_norm": 0.14652763307094574, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0408, + "step": 12600 + }, + { + "epoch": 1.1461552445009997, + "grad_norm": 0.19842924177646637, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0402, + "step": 12610 + }, + { + "epoch": 1.1470641701508817, + "grad_norm": 0.19749978184700012, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.043, + "step": 12620 + }, + { + "epoch": 1.1479730958007635, + "grad_norm": 0.1617807000875473, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0469, + "step": 12630 + }, + { + "epoch": 1.1488820214506452, + "grad_norm": 0.1939033567905426, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0462, + "step": 12640 + }, + { + "epoch": 1.1497909471005272, + "grad_norm": 0.13775935769081116, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0391, + "step": 12650 + }, + { + "epoch": 1.150699872750409, + "grad_norm": 0.20893600583076477, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.045, + "step": 12660 + }, + { + "epoch": 1.1516087984002907, + "grad_norm": 0.1565670669078827, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0413, + "step": 12670 + }, + { + "epoch": 1.1525177240501727, + "grad_norm": 0.15814153850078583, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0411, + "step": 12680 + }, + { + "epoch": 1.1534266497000545, + "grad_norm": 0.144985631108284, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0457, + "step": 12690 + }, + { + "epoch": 1.1543355753499365, + "grad_norm": 0.17230087518692017, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0413, + "step": 12700 + }, + { + "epoch": 1.1552445009998182, + "grad_norm": 0.2010572999715805, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0424, + "step": 12710 + }, + { + "epoch": 1.1561534266497, + "grad_norm": 0.15397875010967255, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0442, + "step": 12720 + }, + { + "epoch": 1.157062352299582, + "grad_norm": 0.16249752044677734, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0457, + "step": 12730 + }, + { + "epoch": 1.1579712779494638, + "grad_norm": 0.2378629893064499, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0427, + "step": 12740 + }, + { + "epoch": 1.1588802035993455, + "grad_norm": 0.27284225821495056, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0422, + "step": 12750 + }, + { + "epoch": 1.1597891292492275, + "grad_norm": 0.18377934396266937, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0468, + "step": 12760 + }, + { + "epoch": 1.1606980548991093, + "grad_norm": 0.22303375601768494, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0457, + "step": 12770 + }, + { + "epoch": 1.161606980548991, + "grad_norm": 0.16184881329536438, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0416, + "step": 12780 + }, + { + "epoch": 1.162515906198873, + "grad_norm": 0.18596085906028748, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0436, + "step": 12790 + }, + { + "epoch": 1.1634248318487548, + "grad_norm": 0.1546899676322937, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0414, + "step": 12800 + }, + { + "epoch": 1.1643337574986365, + "grad_norm": 0.16172322630882263, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0414, + "step": 12810 + }, + { + "epoch": 1.1652426831485185, + "grad_norm": 0.1484745591878891, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0429, + "step": 12820 + }, + { + "epoch": 1.1661516087984003, + "grad_norm": 0.2127726823091507, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0435, + "step": 12830 + }, + { + "epoch": 1.167060534448282, + "grad_norm": 0.2012225091457367, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0439, + "step": 12840 + }, + { + "epoch": 1.167969460098164, + "grad_norm": 0.15178994834423065, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0464, + "step": 12850 + }, + { + "epoch": 1.1688783857480458, + "grad_norm": 0.23344194889068604, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0449, + "step": 12860 + }, + { + "epoch": 1.1697873113979276, + "grad_norm": 0.19137711822986603, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0411, + "step": 12870 + }, + { + "epoch": 1.1706962370478096, + "grad_norm": 0.21124418079853058, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0421, + "step": 12880 + }, + { + "epoch": 1.1716051626976913, + "grad_norm": 0.1604299247264862, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0454, + "step": 12890 + }, + { + "epoch": 1.172514088347573, + "grad_norm": 0.16101138293743134, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0426, + "step": 12900 + }, + { + "epoch": 1.173423013997455, + "grad_norm": 0.15046457946300507, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0386, + "step": 12910 + }, + { + "epoch": 1.1743319396473368, + "grad_norm": 0.17534270882606506, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.042, + "step": 12920 + }, + { + "epoch": 1.1752408652972186, + "grad_norm": 0.17701199650764465, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0471, + "step": 12930 + }, + { + "epoch": 1.1761497909471006, + "grad_norm": 0.17609833180904388, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0409, + "step": 12940 + }, + { + "epoch": 1.1770587165969824, + "grad_norm": 0.1765468269586563, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0411, + "step": 12950 + }, + { + "epoch": 1.1779676422468641, + "grad_norm": 0.14186012744903564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0437, + "step": 12960 + }, + { + "epoch": 1.178876567896746, + "grad_norm": 0.1701977550983429, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0407, + "step": 12970 + }, + { + "epoch": 1.1797854935466279, + "grad_norm": 0.20137670636177063, + "learning_rate": 2.972343411911009e-05, + "loss": 0.043, + "step": 12980 + }, + { + "epoch": 1.1806944191965096, + "grad_norm": 0.1706564575433731, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0427, + "step": 12990 + }, + { + "epoch": 1.1816033448463916, + "grad_norm": 0.18291562795639038, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.043, + "step": 13000 + }, + { + "epoch": 1.1825122704962734, + "grad_norm": 0.22450344264507294, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0436, + "step": 13010 + }, + { + "epoch": 1.1834211961461552, + "grad_norm": 0.23014938831329346, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0424, + "step": 13020 + }, + { + "epoch": 1.1843301217960371, + "grad_norm": 0.26926741003990173, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0405, + "step": 13030 + }, + { + "epoch": 1.185239047445919, + "grad_norm": 0.14386652410030365, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0395, + "step": 13040 + }, + { + "epoch": 1.1861479730958007, + "grad_norm": 0.19420136511325836, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0436, + "step": 13050 + }, + { + "epoch": 1.1870568987456827, + "grad_norm": 0.1469237506389618, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0429, + "step": 13060 + }, + { + "epoch": 1.1879658243955644, + "grad_norm": 0.20024679601192474, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0401, + "step": 13070 + }, + { + "epoch": 1.1888747500454464, + "grad_norm": 0.21854285895824432, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0449, + "step": 13080 + }, + { + "epoch": 1.1897836756953282, + "grad_norm": 0.17793157696723938, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0439, + "step": 13090 + }, + { + "epoch": 1.19069260134521, + "grad_norm": 0.2737067937850952, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0438, + "step": 13100 + }, + { + "epoch": 1.191601526995092, + "grad_norm": 0.17430925369262695, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0425, + "step": 13110 + }, + { + "epoch": 1.1925104526449737, + "grad_norm": 0.17054621875286102, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0414, + "step": 13120 + }, + { + "epoch": 1.1934193782948554, + "grad_norm": 0.15750473737716675, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0439, + "step": 13130 + }, + { + "epoch": 1.1943283039447374, + "grad_norm": 0.19074410200119019, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0413, + "step": 13140 + }, + { + "epoch": 1.1952372295946192, + "grad_norm": 0.18702590465545654, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0414, + "step": 13150 + }, + { + "epoch": 1.196146155244501, + "grad_norm": 0.19683240354061127, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.04, + "step": 13160 + }, + { + "epoch": 1.197055080894383, + "grad_norm": 0.2124345600605011, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0448, + "step": 13170 + }, + { + "epoch": 1.1979640065442647, + "grad_norm": 0.14083169400691986, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0425, + "step": 13180 + }, + { + "epoch": 1.1988729321941465, + "grad_norm": 0.2408747524023056, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.047, + "step": 13190 + }, + { + "epoch": 1.1997818578440285, + "grad_norm": 0.18947604298591614, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0401, + "step": 13200 + }, + { + "epoch": 1.2006907834939102, + "grad_norm": 0.17545129358768463, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0447, + "step": 13210 + }, + { + "epoch": 1.201599709143792, + "grad_norm": 0.18383683264255524, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0408, + "step": 13220 + }, + { + "epoch": 1.202508634793674, + "grad_norm": 0.19113342463970184, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0436, + "step": 13230 + }, + { + "epoch": 1.2034175604435557, + "grad_norm": 0.15470574796199799, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0387, + "step": 13240 + }, + { + "epoch": 1.2043264860934375, + "grad_norm": 0.16139103472232819, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0431, + "step": 13250 + }, + { + "epoch": 1.2052354117433195, + "grad_norm": 0.18045440316200256, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0436, + "step": 13260 + }, + { + "epoch": 1.2061443373932013, + "grad_norm": 0.1725582778453827, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.041, + "step": 13270 + }, + { + "epoch": 1.207053263043083, + "grad_norm": 0.15615011751651764, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0408, + "step": 13280 + }, + { + "epoch": 1.207962188692965, + "grad_norm": 0.19577321410179138, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0447, + "step": 13290 + }, + { + "epoch": 1.2088711143428468, + "grad_norm": 0.21323587000370026, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0408, + "step": 13300 + }, + { + "epoch": 1.2097800399927285, + "grad_norm": 0.2154727280139923, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0409, + "step": 13310 + }, + { + "epoch": 1.2106889656426105, + "grad_norm": 0.14508432149887085, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0467, + "step": 13320 + }, + { + "epoch": 1.2115978912924923, + "grad_norm": 0.20044608414173126, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0442, + "step": 13330 + }, + { + "epoch": 1.212506816942374, + "grad_norm": 0.18292324244976044, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0403, + "step": 13340 + }, + { + "epoch": 1.213415742592256, + "grad_norm": 0.17023715376853943, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0437, + "step": 13350 + }, + { + "epoch": 1.2143246682421378, + "grad_norm": 0.1684408336877823, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0457, + "step": 13360 + }, + { + "epoch": 1.2152335938920196, + "grad_norm": 0.15124501287937164, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0393, + "step": 13370 + }, + { + "epoch": 1.2161425195419016, + "grad_norm": 0.18795016407966614, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0428, + "step": 13380 + }, + { + "epoch": 1.2170514451917833, + "grad_norm": 0.1982436627149582, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0399, + "step": 13390 + }, + { + "epoch": 1.217960370841665, + "grad_norm": 0.18964123725891113, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0439, + "step": 13400 + }, + { + "epoch": 1.218869296491547, + "grad_norm": 0.14339594542980194, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0426, + "step": 13410 + }, + { + "epoch": 1.2197782221414288, + "grad_norm": 0.1722923070192337, + "learning_rate": 2.968758727655112e-05, + "loss": 0.045, + "step": 13420 + }, + { + "epoch": 1.2206871477913106, + "grad_norm": 0.20478466153144836, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0426, + "step": 13430 + }, + { + "epoch": 1.2215960734411926, + "grad_norm": 0.2596146762371063, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0446, + "step": 13440 + }, + { + "epoch": 1.2225049990910744, + "grad_norm": 0.1824934035539627, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0412, + "step": 13450 + }, + { + "epoch": 1.2234139247409561, + "grad_norm": 0.21002498269081116, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.041, + "step": 13460 + }, + { + "epoch": 1.224322850390838, + "grad_norm": 0.18762479722499847, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0442, + "step": 13470 + }, + { + "epoch": 1.2252317760407199, + "grad_norm": 0.1963743418455124, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0432, + "step": 13480 + }, + { + "epoch": 1.2261407016906016, + "grad_norm": 0.15891602635383606, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0423, + "step": 13490 + }, + { + "epoch": 1.2270496273404836, + "grad_norm": 0.21507400274276733, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0458, + "step": 13500 + }, + { + "epoch": 1.2279585529903654, + "grad_norm": 0.21273833513259888, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0438, + "step": 13510 + }, + { + "epoch": 1.2288674786402471, + "grad_norm": 0.18319319188594818, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0418, + "step": 13520 + }, + { + "epoch": 1.2297764042901291, + "grad_norm": 0.16872328519821167, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0397, + "step": 13530 + }, + { + "epoch": 1.230685329940011, + "grad_norm": 0.18002882599830627, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0435, + "step": 13540 + }, + { + "epoch": 1.2315942555898927, + "grad_norm": 0.1607378125190735, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0427, + "step": 13550 + }, + { + "epoch": 1.2325031812397746, + "grad_norm": 0.16036030650138855, + "learning_rate": 2.967572834590274e-05, + "loss": 0.041, + "step": 13560 + }, + { + "epoch": 1.2334121068896564, + "grad_norm": 0.2018328458070755, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0436, + "step": 13570 + }, + { + "epoch": 1.2343210325395382, + "grad_norm": 0.23576276004314423, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0438, + "step": 13580 + }, + { + "epoch": 1.2352299581894202, + "grad_norm": 0.1494191735982895, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0437, + "step": 13590 + }, + { + "epoch": 1.236138883839302, + "grad_norm": 0.19222813844680786, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0439, + "step": 13600 + }, + { + "epoch": 1.2370478094891837, + "grad_norm": 0.194656640291214, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0399, + "step": 13610 + }, + { + "epoch": 1.2379567351390657, + "grad_norm": 0.1491241455078125, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0411, + "step": 13620 + }, + { + "epoch": 1.2388656607889474, + "grad_norm": 0.17013825476169586, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0424, + "step": 13630 + }, + { + "epoch": 1.2397745864388292, + "grad_norm": 0.22360572218894958, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0403, + "step": 13640 + }, + { + "epoch": 1.2406835120887112, + "grad_norm": 0.18272830545902252, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0457, + "step": 13650 + }, + { + "epoch": 1.241592437738593, + "grad_norm": 0.1681375354528427, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0409, + "step": 13660 + }, + { + "epoch": 1.2425013633884747, + "grad_norm": 0.19051140546798706, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0424, + "step": 13670 + }, + { + "epoch": 1.2434102890383567, + "grad_norm": 0.2176285833120346, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0446, + "step": 13680 + }, + { + "epoch": 1.2443192146882385, + "grad_norm": 0.17195509374141693, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0427, + "step": 13690 + }, + { + "epoch": 1.2452281403381202, + "grad_norm": 0.23669682443141937, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0431, + "step": 13700 + }, + { + "epoch": 1.2461370659880022, + "grad_norm": 0.1911829262971878, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0425, + "step": 13710 + }, + { + "epoch": 1.247045991637884, + "grad_norm": 0.14725163578987122, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0398, + "step": 13720 + }, + { + "epoch": 1.2479549172877658, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0441, + "step": 13730 + }, + { + "epoch": 1.2488638429376477, + "grad_norm": 0.23396606743335724, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0419, + "step": 13740 + }, + { + "epoch": 1.2497727685875295, + "grad_norm": 0.149898961186409, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0452, + "step": 13750 + }, + { + "epoch": 1.2506816942374113, + "grad_norm": 0.19988778233528137, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0416, + "step": 13760 + }, + { + "epoch": 1.2515906198872933, + "grad_norm": 0.20767448842525482, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0395, + "step": 13770 + }, + { + "epoch": 1.252499545537175, + "grad_norm": 0.13941749930381775, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.041, + "step": 13780 + }, + { + "epoch": 1.2534084711870568, + "grad_norm": 0.17284810543060303, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0412, + "step": 13790 + }, + { + "epoch": 1.2543173968369388, + "grad_norm": 0.1960815191268921, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0405, + "step": 13800 + }, + { + "epoch": 1.2552263224868205, + "grad_norm": 0.15711890161037445, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0403, + "step": 13810 + }, + { + "epoch": 1.2561352481367023, + "grad_norm": 0.17390109598636627, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0425, + "step": 13820 + }, + { + "epoch": 1.2570441737865843, + "grad_norm": 0.19116121530532837, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0415, + "step": 13830 + }, + { + "epoch": 1.257953099436466, + "grad_norm": 0.19422760605812073, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0413, + "step": 13840 + }, + { + "epoch": 1.2588620250863478, + "grad_norm": 0.2299289107322693, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0434, + "step": 13850 + }, + { + "epoch": 1.2597709507362298, + "grad_norm": 0.2840384244918823, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0379, + "step": 13860 + }, + { + "epoch": 1.2606798763861116, + "grad_norm": 0.18506966531276703, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0453, + "step": 13870 + }, + { + "epoch": 1.2615888020359933, + "grad_norm": 0.17552730441093445, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0465, + "step": 13880 + }, + { + "epoch": 1.2624977276858753, + "grad_norm": 0.17635229229927063, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0448, + "step": 13890 + }, + { + "epoch": 1.263406653335757, + "grad_norm": 0.17926646769046783, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0435, + "step": 13900 + }, + { + "epoch": 1.2643155789856388, + "grad_norm": 0.15634077787399292, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0445, + "step": 13910 + }, + { + "epoch": 1.2652245046355208, + "grad_norm": 0.1563355028629303, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0412, + "step": 13920 + }, + { + "epoch": 1.2661334302854026, + "grad_norm": 0.16547061502933502, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0402, + "step": 13930 + }, + { + "epoch": 1.2670423559352844, + "grad_norm": 0.2099078744649887, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0425, + "step": 13940 + }, + { + "epoch": 1.2679512815851663, + "grad_norm": 0.2158689647912979, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0396, + "step": 13950 + }, + { + "epoch": 1.268860207235048, + "grad_norm": 0.2972472608089447, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0437, + "step": 13960 + }, + { + "epoch": 1.2697691328849299, + "grad_norm": 0.23452560603618622, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0438, + "step": 13970 + }, + { + "epoch": 1.2706780585348119, + "grad_norm": 0.14379937946796417, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0405, + "step": 13980 + }, + { + "epoch": 1.2715869841846936, + "grad_norm": 0.187812939286232, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0397, + "step": 13990 + }, + { + "epoch": 1.2724959098345756, + "grad_norm": 0.21033167839050293, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.04, + "step": 14000 + }, + { + "epoch": 1.2734048354844574, + "grad_norm": 0.18420188128948212, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0426, + "step": 14010 + }, + { + "epoch": 1.2743137611343391, + "grad_norm": 0.16521772742271423, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0437, + "step": 14020 + }, + { + "epoch": 1.2752226867842211, + "grad_norm": 0.2429179549217224, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0399, + "step": 14030 + }, + { + "epoch": 1.276131612434103, + "grad_norm": 0.17190448939800262, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0415, + "step": 14040 + }, + { + "epoch": 1.2770405380839847, + "grad_norm": 0.19656823575496674, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0438, + "step": 14050 + }, + { + "epoch": 1.2779494637338666, + "grad_norm": 0.15375284850597382, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0391, + "step": 14060 + }, + { + "epoch": 1.2788583893837484, + "grad_norm": 0.1631784439086914, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0386, + "step": 14070 + }, + { + "epoch": 1.2797673150336302, + "grad_norm": 0.15656892955303192, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0378, + "step": 14080 + }, + { + "epoch": 1.2806762406835122, + "grad_norm": 0.1737211048603058, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0398, + "step": 14090 + }, + { + "epoch": 1.281585166333394, + "grad_norm": 0.14551092684268951, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0415, + "step": 14100 + }, + { + "epoch": 1.2824940919832757, + "grad_norm": 0.1491597294807434, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0404, + "step": 14110 + }, + { + "epoch": 1.2834030176331577, + "grad_norm": 0.21427276730537415, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0416, + "step": 14120 + }, + { + "epoch": 1.2843119432830394, + "grad_norm": 0.15114368498325348, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0426, + "step": 14130 + }, + { + "epoch": 1.2852208689329214, + "grad_norm": 0.14748764038085938, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.042, + "step": 14140 + }, + { + "epoch": 1.2861297945828032, + "grad_norm": 0.2111600637435913, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0413, + "step": 14150 + }, + { + "epoch": 1.287038720232685, + "grad_norm": 0.16424083709716797, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0411, + "step": 14160 + }, + { + "epoch": 1.287947645882567, + "grad_norm": 0.1768375188112259, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0402, + "step": 14170 + }, + { + "epoch": 1.2888565715324487, + "grad_norm": 0.14442791044712067, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0415, + "step": 14180 + }, + { + "epoch": 1.2897654971823305, + "grad_norm": 0.1810288429260254, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0399, + "step": 14190 + }, + { + "epoch": 1.2906744228322125, + "grad_norm": 0.1780414581298828, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0419, + "step": 14200 + }, + { + "epoch": 1.2915833484820942, + "grad_norm": 0.1974322646856308, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0424, + "step": 14210 + }, + { + "epoch": 1.292492274131976, + "grad_norm": 0.15540142357349396, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0423, + "step": 14220 + }, + { + "epoch": 1.293401199781858, + "grad_norm": 0.15884260833263397, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0431, + "step": 14230 + }, + { + "epoch": 1.2943101254317397, + "grad_norm": 0.16664133965969086, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0416, + "step": 14240 + }, + { + "epoch": 1.2952190510816215, + "grad_norm": 0.25029104948043823, + "learning_rate": 2.961409280636245e-05, + "loss": 0.045, + "step": 14250 + }, + { + "epoch": 1.2961279767315035, + "grad_norm": 0.1669912189245224, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0405, + "step": 14260 + }, + { + "epoch": 1.2970369023813852, + "grad_norm": 0.22703322768211365, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.041, + "step": 14270 + }, + { + "epoch": 1.297945828031267, + "grad_norm": 0.14368757605552673, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0361, + "step": 14280 + }, + { + "epoch": 1.298854753681149, + "grad_norm": 0.1959294080734253, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0421, + "step": 14290 + }, + { + "epoch": 1.2997636793310308, + "grad_norm": 0.2183523178100586, + "learning_rate": 2.960942086834339e-05, + "loss": 0.0389, + "step": 14300 + }, + { + "epoch": 1.3006726049809125, + "grad_norm": 0.15234805643558502, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0451, + "step": 14310 + }, + { + "epoch": 1.3015815306307945, + "grad_norm": 0.1635727733373642, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0418, + "step": 14320 + }, + { + "epoch": 1.3024904562806763, + "grad_norm": 0.14696864783763885, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0424, + "step": 14330 + }, + { + "epoch": 1.303399381930558, + "grad_norm": 0.15609794855117798, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0431, + "step": 14340 + }, + { + "epoch": 1.30430830758044, + "grad_norm": 0.14806030690670013, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0403, + "step": 14350 + }, + { + "epoch": 1.3052172332303218, + "grad_norm": 0.1902691125869751, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0388, + "step": 14360 + }, + { + "epoch": 1.3061261588802036, + "grad_norm": 0.24311301112174988, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0396, + "step": 14370 + }, + { + "epoch": 1.3070350845300855, + "grad_norm": 0.21988347172737122, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.039, + "step": 14380 + }, + { + "epoch": 1.3079440101799673, + "grad_norm": 0.2012740522623062, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0428, + "step": 14390 + }, + { + "epoch": 1.308852935829849, + "grad_norm": 0.18528451025485992, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0397, + "step": 14400 + }, + { + "epoch": 1.309761861479731, + "grad_norm": 0.2098444402217865, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0411, + "step": 14410 + }, + { + "epoch": 1.3106707871296128, + "grad_norm": 0.1484665870666504, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0364, + "step": 14420 + }, + { + "epoch": 1.3115797127794946, + "grad_norm": 0.16567005217075348, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0453, + "step": 14430 + }, + { + "epoch": 1.3124886384293766, + "grad_norm": 0.1728101372718811, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0425, + "step": 14440 + }, + { + "epoch": 1.3133975640792583, + "grad_norm": 0.1659095585346222, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0395, + "step": 14450 + }, + { + "epoch": 1.31430648972914, + "grad_norm": 0.20774003863334656, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0409, + "step": 14460 + }, + { + "epoch": 1.315215415379022, + "grad_norm": 0.19641171395778656, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0401, + "step": 14470 + }, + { + "epoch": 1.3161243410289039, + "grad_norm": 0.16423091292381287, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0405, + "step": 14480 + }, + { + "epoch": 1.3170332666787856, + "grad_norm": 0.2093425691127777, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0413, + "step": 14490 + }, + { + "epoch": 1.3179421923286676, + "grad_norm": 0.1725950688123703, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0408, + "step": 14500 + }, + { + "epoch": 1.3188511179785494, + "grad_norm": 0.14086031913757324, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0425, + "step": 14510 + }, + { + "epoch": 1.3197600436284311, + "grad_norm": 0.1942043900489807, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.0389, + "step": 14520 + }, + { + "epoch": 1.3206689692783131, + "grad_norm": 0.20726317167282104, + "learning_rate": 2.958757283645799e-05, + "loss": 0.04, + "step": 14530 + }, + { + "epoch": 1.3215778949281949, + "grad_norm": 0.21617549657821655, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0394, + "step": 14540 + }, + { + "epoch": 1.3224868205780766, + "grad_norm": 0.17343173921108246, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0417, + "step": 14550 + }, + { + "epoch": 1.3233957462279586, + "grad_norm": 0.19017352163791656, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.043, + "step": 14560 + }, + { + "epoch": 1.3243046718778404, + "grad_norm": 0.1441139429807663, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0408, + "step": 14570 + }, + { + "epoch": 1.3252135975277222, + "grad_norm": 0.2005908340215683, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0375, + "step": 14580 + }, + { + "epoch": 1.3261225231776042, + "grad_norm": 0.18918393552303314, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0434, + "step": 14590 + }, + { + "epoch": 1.327031448827486, + "grad_norm": 0.2062591165304184, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0423, + "step": 14600 + }, + { + "epoch": 1.3279403744773677, + "grad_norm": 0.18434257805347443, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.043, + "step": 14610 + }, + { + "epoch": 1.3288493001272497, + "grad_norm": 0.14959262311458588, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.041, + "step": 14620 + }, + { + "epoch": 1.3297582257771314, + "grad_norm": 0.20252379775047302, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0427, + "step": 14630 + }, + { + "epoch": 1.3306671514270132, + "grad_norm": 0.2021208256483078, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0419, + "step": 14640 + }, + { + "epoch": 1.3315760770768952, + "grad_norm": 0.2180335521697998, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0416, + "step": 14650 + }, + { + "epoch": 1.332485002726777, + "grad_norm": 0.17642995715141296, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0403, + "step": 14660 + }, + { + "epoch": 1.3333939283766587, + "grad_norm": 0.15116342902183533, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0441, + "step": 14670 + }, + { + "epoch": 1.3343028540265407, + "grad_norm": 0.19887466728687286, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0413, + "step": 14680 + }, + { + "epoch": 1.3352117796764225, + "grad_norm": 0.22436535358428955, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0399, + "step": 14690 + }, + { + "epoch": 1.3361207053263042, + "grad_norm": 0.15925700962543488, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0424, + "step": 14700 + }, + { + "epoch": 1.3370296309761862, + "grad_norm": 0.1782078742980957, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0406, + "step": 14710 + }, + { + "epoch": 1.337938556626068, + "grad_norm": 0.1558130830526352, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0395, + "step": 14720 + }, + { + "epoch": 1.3388474822759497, + "grad_norm": 0.15631066262722015, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0415, + "step": 14730 + }, + { + "epoch": 1.3397564079258317, + "grad_norm": 0.141697496175766, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0363, + "step": 14740 + }, + { + "epoch": 1.3406653335757135, + "grad_norm": 0.18126429617404938, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0413, + "step": 14750 + }, + { + "epoch": 1.3415742592255953, + "grad_norm": 0.17588360607624054, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0449, + "step": 14760 + }, + { + "epoch": 1.3424831848754772, + "grad_norm": 0.2149839699268341, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0382, + "step": 14770 + }, + { + "epoch": 1.343392110525359, + "grad_norm": 0.1910882592201233, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0402, + "step": 14780 + }, + { + "epoch": 1.3443010361752408, + "grad_norm": 0.13856671750545502, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.0382, + "step": 14790 + }, + { + "epoch": 1.3452099618251228, + "grad_norm": 0.15464571118354797, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0414, + "step": 14800 + }, + { + "epoch": 1.3461188874750045, + "grad_norm": 0.163631409406662, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0433, + "step": 14810 + }, + { + "epoch": 1.3470278131248863, + "grad_norm": 0.16562563180923462, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0391, + "step": 14820 + }, + { + "epoch": 1.3479367387747683, + "grad_norm": 0.18215061724185944, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0416, + "step": 14830 + }, + { + "epoch": 1.34884566442465, + "grad_norm": 0.165035218000412, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0399, + "step": 14840 + }, + { + "epoch": 1.3497545900745318, + "grad_norm": 0.13128957152366638, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.038, + "step": 14850 + }, + { + "epoch": 1.3506635157244138, + "grad_norm": 0.18869595229625702, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0402, + "step": 14860 + }, + { + "epoch": 1.3515724413742956, + "grad_norm": 0.15257427096366882, + "learning_rate": 2.9554202524562e-05, + "loss": 0.039, + "step": 14870 + }, + { + "epoch": 1.3524813670241773, + "grad_norm": 0.2213669866323471, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0405, + "step": 14880 + }, + { + "epoch": 1.3533902926740593, + "grad_norm": 0.1426287740468979, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0397, + "step": 14890 + }, + { + "epoch": 1.354299218323941, + "grad_norm": 0.14855048060417175, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0431, + "step": 14900 + }, + { + "epoch": 1.3552081439738228, + "grad_norm": 0.17221325635910034, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0433, + "step": 14910 + }, + { + "epoch": 1.3561170696237048, + "grad_norm": 0.14895640313625336, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0429, + "step": 14920 + }, + { + "epoch": 1.3570259952735866, + "grad_norm": 0.16796040534973145, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0441, + "step": 14930 + }, + { + "epoch": 1.3579349209234683, + "grad_norm": 0.18506425619125366, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0432, + "step": 14940 + }, + { + "epoch": 1.3588438465733503, + "grad_norm": 0.20869180560112, + "learning_rate": 2.954616488731095e-05, + "loss": 0.037, + "step": 14950 + }, + { + "epoch": 1.359752772223232, + "grad_norm": 0.18971173465251923, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0396, + "step": 14960 + }, + { + "epoch": 1.3606616978731139, + "grad_norm": 0.16643884778022766, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0414, + "step": 14970 + }, + { + "epoch": 1.3615706235229958, + "grad_norm": 0.19656245410442352, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0398, + "step": 14980 + }, + { + "epoch": 1.3624795491728776, + "grad_norm": 0.17961107194423676, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.042, + "step": 14990 + }, + { + "epoch": 1.3633884748227594, + "grad_norm": 0.17137441039085388, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0382, + "step": 15000 + }, + { + "epoch": 1.3642974004726414, + "grad_norm": 0.15696153044700623, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0425, + "step": 15010 + }, + { + "epoch": 1.3652063261225231, + "grad_norm": 0.1375163197517395, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0411, + "step": 15020 + }, + { + "epoch": 1.366115251772405, + "grad_norm": 0.2383534461259842, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0393, + "step": 15030 + }, + { + "epoch": 1.3670241774222869, + "grad_norm": 0.1845560073852539, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0378, + "step": 15040 + }, + { + "epoch": 1.3679331030721686, + "grad_norm": 0.16288304328918457, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0402, + "step": 15050 + }, + { + "epoch": 1.3688420287220504, + "grad_norm": 0.1644960641860962, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0425, + "step": 15060 + }, + { + "epoch": 1.3697509543719324, + "grad_norm": 0.19863833487033844, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0452, + "step": 15070 + }, + { + "epoch": 1.3706598800218142, + "grad_norm": 0.2139146327972412, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0421, + "step": 15080 + }, + { + "epoch": 1.371568805671696, + "grad_norm": 0.14994141459465027, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0434, + "step": 15090 + }, + { + "epoch": 1.372477731321578, + "grad_norm": 0.24475212395191193, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0428, + "step": 15100 + }, + { + "epoch": 1.3733866569714597, + "grad_norm": 0.24842676520347595, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0402, + "step": 15110 + }, + { + "epoch": 1.3742955826213414, + "grad_norm": 0.14102886617183685, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0403, + "step": 15120 + }, + { + "epoch": 1.3752045082712234, + "grad_norm": 0.15645895898342133, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0386, + "step": 15130 + }, + { + "epoch": 1.3761134339211052, + "grad_norm": 0.18005484342575073, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0434, + "step": 15140 + }, + { + "epoch": 1.3770223595709872, + "grad_norm": 0.1716563105583191, + "learning_rate": 2.952576156874546e-05, + "loss": 0.036, + "step": 15150 + }, + { + "epoch": 1.377931285220869, + "grad_norm": 0.23669283092021942, + "learning_rate": 2.952472981471207e-05, + "loss": 0.04, + "step": 15160 + }, + { + "epoch": 1.3788402108707507, + "grad_norm": 0.20746968686580658, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0392, + "step": 15170 + }, + { + "epoch": 1.3797491365206327, + "grad_norm": 0.20605704188346863, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0414, + "step": 15180 + }, + { + "epoch": 1.3806580621705145, + "grad_norm": 0.18780483305454254, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.042, + "step": 15190 + }, + { + "epoch": 1.3815669878203962, + "grad_norm": 0.16201989352703094, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0402, + "step": 15200 + }, + { + "epoch": 1.3824759134702782, + "grad_norm": 0.16080342233181, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0411, + "step": 15210 + }, + { + "epoch": 1.38338483912016, + "grad_norm": 0.144723579287529, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0424, + "step": 15220 + }, + { + "epoch": 1.3842937647700417, + "grad_norm": 0.1670612245798111, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0419, + "step": 15230 + }, + { + "epoch": 1.3852026904199237, + "grad_norm": 0.19271285831928253, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0403, + "step": 15240 + }, + { + "epoch": 1.3861116160698055, + "grad_norm": 0.21632784605026245, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0417, + "step": 15250 + }, + { + "epoch": 1.3870205417196872, + "grad_norm": 0.1630091816186905, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0397, + "step": 15260 + }, + { + "epoch": 1.3879294673695692, + "grad_norm": 0.22165243327617645, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0433, + "step": 15270 + }, + { + "epoch": 1.388838393019451, + "grad_norm": 0.16306503117084503, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0425, + "step": 15280 + }, + { + "epoch": 1.389747318669333, + "grad_norm": 0.2044500857591629, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0389, + "step": 15290 + }, + { + "epoch": 1.3906562443192148, + "grad_norm": 0.2098514586687088, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0421, + "step": 15300 + }, + { + "epoch": 1.3915651699690965, + "grad_norm": 0.2116941660642624, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.042, + "step": 15310 + }, + { + "epoch": 1.3924740956189785, + "grad_norm": 0.19745129346847534, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0377, + "step": 15320 + }, + { + "epoch": 1.3933830212688603, + "grad_norm": 0.13776959478855133, + "learning_rate": 2.950702129262842e-05, + "loss": 0.04, + "step": 15330 + }, + { + "epoch": 1.394291946918742, + "grad_norm": 0.22999173402786255, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0418, + "step": 15340 + }, + { + "epoch": 1.395200872568624, + "grad_norm": 0.1760820746421814, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0384, + "step": 15350 + }, + { + "epoch": 1.3961097982185058, + "grad_norm": 0.1694323718547821, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0438, + "step": 15360 + }, + { + "epoch": 1.3970187238683875, + "grad_norm": 0.1692339926958084, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0403, + "step": 15370 + }, + { + "epoch": 1.3979276495182695, + "grad_norm": 0.21988563239574432, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0385, + "step": 15380 + }, + { + "epoch": 1.3988365751681513, + "grad_norm": 0.14232857525348663, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.041, + "step": 15390 + }, + { + "epoch": 1.399745500818033, + "grad_norm": 0.1962636113166809, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0411, + "step": 15400 + }, + { + "epoch": 1.400654426467915, + "grad_norm": 0.17379921674728394, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0367, + "step": 15410 + }, + { + "epoch": 1.4015633521177968, + "grad_norm": 0.19331975281238556, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.0421, + "step": 15420 + }, + { + "epoch": 1.4024722777676786, + "grad_norm": 0.24675510823726654, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0384, + "step": 15430 + }, + { + "epoch": 1.4033812034175606, + "grad_norm": 0.20762276649475098, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0433, + "step": 15440 + }, + { + "epoch": 1.4042901290674423, + "grad_norm": 0.17303277552127838, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0369, + "step": 15450 + }, + { + "epoch": 1.405199054717324, + "grad_norm": 0.2005726546049118, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0375, + "step": 15460 + }, + { + "epoch": 1.406107980367206, + "grad_norm": 0.21824230253696442, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0352, + "step": 15470 + }, + { + "epoch": 1.4070169060170878, + "grad_norm": 0.19065344333648682, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0399, + "step": 15480 + }, + { + "epoch": 1.4079258316669696, + "grad_norm": 0.15611107647418976, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0408, + "step": 15490 + }, + { + "epoch": 1.4088347573168516, + "grad_norm": 0.18182595074176788, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0435, + "step": 15500 + }, + { + "epoch": 1.4097436829667334, + "grad_norm": 0.19053412973880768, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0395, + "step": 15510 + }, + { + "epoch": 1.4106526086166151, + "grad_norm": 0.2089206427335739, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0436, + "step": 15520 + }, + { + "epoch": 1.411561534266497, + "grad_norm": 0.21028822660446167, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0377, + "step": 15530 + }, + { + "epoch": 1.4124704599163789, + "grad_norm": 0.1485883593559265, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0409, + "step": 15540 + }, + { + "epoch": 1.4133793855662606, + "grad_norm": 0.17854148149490356, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0391, + "step": 15550 + }, + { + "epoch": 1.4142883112161426, + "grad_norm": 0.19274961948394775, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0419, + "step": 15560 + }, + { + "epoch": 1.4151972368660244, + "grad_norm": 0.1930166482925415, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0417, + "step": 15570 + }, + { + "epoch": 1.4161061625159062, + "grad_norm": 0.21229548752307892, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0417, + "step": 15580 + }, + { + "epoch": 1.4170150881657881, + "grad_norm": 0.16230139136314392, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0393, + "step": 15590 + }, + { + "epoch": 1.41792401381567, + "grad_norm": 0.20785777270793915, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0381, + "step": 15600 + }, + { + "epoch": 1.4188329394655517, + "grad_norm": 0.16502776741981506, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0407, + "step": 15610 + }, + { + "epoch": 1.4197418651154337, + "grad_norm": 0.19724269211292267, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0397, + "step": 15620 + }, + { + "epoch": 1.4206507907653154, + "grad_norm": 0.1495850384235382, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0393, + "step": 15630 + }, + { + "epoch": 1.4215597164151972, + "grad_norm": 0.2527087330818176, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0388, + "step": 15640 + }, + { + "epoch": 1.4224686420650792, + "grad_norm": 0.1503797024488449, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0394, + "step": 15650 + }, + { + "epoch": 1.423377567714961, + "grad_norm": 0.25641122460365295, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0401, + "step": 15660 + }, + { + "epoch": 1.4242864933648427, + "grad_norm": 0.14962758123874664, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0382, + "step": 15670 + }, + { + "epoch": 1.4251954190147247, + "grad_norm": 0.1582927256822586, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0383, + "step": 15680 + }, + { + "epoch": 1.4261043446646064, + "grad_norm": 0.15208658576011658, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0374, + "step": 15690 + }, + { + "epoch": 1.4270132703144882, + "grad_norm": 0.1848893165588379, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0388, + "step": 15700 + }, + { + "epoch": 1.4279221959643702, + "grad_norm": 0.18593640625476837, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.0415, + "step": 15710 + }, + { + "epoch": 1.428831121614252, + "grad_norm": 0.20365995168685913, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0379, + "step": 15720 + }, + { + "epoch": 1.4297400472641337, + "grad_norm": 0.14577262103557587, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.038, + "step": 15730 + }, + { + "epoch": 1.4306489729140157, + "grad_norm": 0.19851170480251312, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0397, + "step": 15740 + }, + { + "epoch": 1.4315578985638975, + "grad_norm": 0.17837651073932648, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0423, + "step": 15750 + }, + { + "epoch": 1.4324668242137792, + "grad_norm": 0.15929678082466125, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0422, + "step": 15760 + }, + { + "epoch": 1.4333757498636612, + "grad_norm": 0.16249117255210876, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0407, + "step": 15770 + }, + { + "epoch": 1.434284675513543, + "grad_norm": 0.18814004957675934, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0415, + "step": 15780 + }, + { + "epoch": 1.4351936011634248, + "grad_norm": 0.22910703718662262, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0411, + "step": 15790 + }, + { + "epoch": 1.4361025268133067, + "grad_norm": 0.17910706996917725, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0425, + "step": 15800 + }, + { + "epoch": 1.4370114524631885, + "grad_norm": 0.17388801276683807, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0397, + "step": 15810 + }, + { + "epoch": 1.4379203781130703, + "grad_norm": 0.17612694203853607, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.039, + "step": 15820 + }, + { + "epoch": 1.4388293037629523, + "grad_norm": 0.21394631266593933, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0361, + "step": 15830 + }, + { + "epoch": 1.439738229412834, + "grad_norm": 0.20800194144248962, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0404, + "step": 15840 + }, + { + "epoch": 1.4406471550627158, + "grad_norm": 0.1545456349849701, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0381, + "step": 15850 + }, + { + "epoch": 1.4415560807125978, + "grad_norm": 0.17767289280891418, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0402, + "step": 15860 + }, + { + "epoch": 1.4424650063624795, + "grad_norm": 0.2036893367767334, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0397, + "step": 15870 + }, + { + "epoch": 1.4433739320123613, + "grad_norm": 0.1683395355939865, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0403, + "step": 15880 + }, + { + "epoch": 1.4442828576622433, + "grad_norm": 0.15401670336723328, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0415, + "step": 15890 + }, + { + "epoch": 1.445191783312125, + "grad_norm": 0.2239566296339035, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.04, + "step": 15900 + }, + { + "epoch": 1.4461007089620068, + "grad_norm": 0.1588655710220337, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.041, + "step": 15910 + }, + { + "epoch": 1.4470096346118888, + "grad_norm": 0.18617916107177734, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0401, + "step": 15920 + }, + { + "epoch": 1.4479185602617706, + "grad_norm": 0.17407822608947754, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0378, + "step": 15930 + }, + { + "epoch": 1.4488274859116523, + "grad_norm": 0.17696905136108398, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0384, + "step": 15940 + }, + { + "epoch": 1.4497364115615343, + "grad_norm": 0.14923858642578125, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0394, + "step": 15950 + }, + { + "epoch": 1.450645337211416, + "grad_norm": 0.20700973272323608, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0378, + "step": 15960 + }, + { + "epoch": 1.4515542628612979, + "grad_norm": 0.18127496540546417, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0388, + "step": 15970 + }, + { + "epoch": 1.4524631885111798, + "grad_norm": 0.19819886982440948, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0419, + "step": 15980 + }, + { + "epoch": 1.4533721141610616, + "grad_norm": 0.16177622973918915, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0415, + "step": 15990 + }, + { + "epoch": 1.4542810398109434, + "grad_norm": 0.16978523135185242, + "learning_rate": 2.94341324778745e-05, + "loss": 0.037, + "step": 16000 + }, + { + "epoch": 1.4551899654608254, + "grad_norm": 0.174326553940773, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0374, + "step": 16010 + }, + { + "epoch": 1.4560988911107071, + "grad_norm": 0.19166342914104462, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0384, + "step": 16020 + }, + { + "epoch": 1.4570078167605889, + "grad_norm": 0.19166579842567444, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0367, + "step": 16030 + }, + { + "epoch": 1.4579167424104709, + "grad_norm": 0.24249394237995148, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0392, + "step": 16040 + }, + { + "epoch": 1.4588256680603526, + "grad_norm": 0.18298368155956268, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0429, + "step": 16050 + }, + { + "epoch": 1.4597345937102344, + "grad_norm": 0.19750167429447174, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0396, + "step": 16060 + }, + { + "epoch": 1.4606435193601164, + "grad_norm": 0.22862936556339264, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0407, + "step": 16070 + }, + { + "epoch": 1.4615524450099981, + "grad_norm": 0.16091904044151306, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0428, + "step": 16080 + }, + { + "epoch": 1.46246137065988, + "grad_norm": 0.33251288533210754, + "learning_rate": 2.942396599516316e-05, + "loss": 0.041, + "step": 16090 + }, + { + "epoch": 1.463370296309762, + "grad_norm": 0.18243378400802612, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0373, + "step": 16100 + }, + { + "epoch": 1.4642792219596437, + "grad_norm": 0.18044732511043549, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.041, + "step": 16110 + }, + { + "epoch": 1.4651881476095254, + "grad_norm": 0.2800557017326355, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0386, + "step": 16120 + }, + { + "epoch": 1.4660970732594074, + "grad_norm": 0.20292165875434875, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0392, + "step": 16130 + }, + { + "epoch": 1.4670059989092892, + "grad_norm": 0.1894267499446869, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0407, + "step": 16140 + }, + { + "epoch": 1.467914924559171, + "grad_norm": 0.15495117008686066, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0379, + "step": 16150 + }, + { + "epoch": 1.468823850209053, + "grad_norm": 0.19055497646331787, + "learning_rate": 2.941599738495953e-05, + "loss": 0.04, + "step": 16160 + }, + { + "epoch": 1.4697327758589347, + "grad_norm": 0.1623472273349762, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0428, + "step": 16170 + }, + { + "epoch": 1.4706417015088165, + "grad_norm": 0.18112750351428986, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0391, + "step": 16180 + }, + { + "epoch": 1.4715506271586984, + "grad_norm": 0.18392126262187958, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0415, + "step": 16190 + }, + { + "epoch": 1.4724595528085802, + "grad_norm": 0.19015300273895264, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0406, + "step": 16200 + }, + { + "epoch": 1.473368478458462, + "grad_norm": 0.19722749292850494, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0422, + "step": 16210 + }, + { + "epoch": 1.474277404108344, + "grad_norm": 0.14655472338199615, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0413, + "step": 16220 + }, + { + "epoch": 1.4751863297582257, + "grad_norm": 0.18754242360591888, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0379, + "step": 16230 + }, + { + "epoch": 1.4760952554081075, + "grad_norm": 0.16426251828670502, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0362, + "step": 16240 + }, + { + "epoch": 1.4770041810579895, + "grad_norm": 0.16547103226184845, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0393, + "step": 16250 + }, + { + "epoch": 1.4779131067078712, + "grad_norm": 0.22632566094398499, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0438, + "step": 16260 + }, + { + "epoch": 1.478822032357753, + "grad_norm": 0.21476444602012634, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0439, + "step": 16270 + }, + { + "epoch": 1.479730958007635, + "grad_norm": 0.16481386125087738, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0404, + "step": 16280 + }, + { + "epoch": 1.4806398836575168, + "grad_norm": 0.22443313896656036, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0426, + "step": 16290 + }, + { + "epoch": 1.4815488093073985, + "grad_norm": 0.1784551739692688, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0405, + "step": 16300 + }, + { + "epoch": 1.4824577349572805, + "grad_norm": 0.1569932997226715, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0395, + "step": 16310 + }, + { + "epoch": 1.4833666606071623, + "grad_norm": 0.18689769506454468, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0394, + "step": 16320 + }, + { + "epoch": 1.4842755862570443, + "grad_norm": 0.15507252514362335, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0394, + "step": 16330 + }, + { + "epoch": 1.485184511906926, + "grad_norm": 0.1740468591451645, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0403, + "step": 16340 + }, + { + "epoch": 1.4860934375568078, + "grad_norm": 0.14829982817173004, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0395, + "step": 16350 + }, + { + "epoch": 1.4870023632066898, + "grad_norm": 0.16879568994045258, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0391, + "step": 16360 + }, + { + "epoch": 1.4879112888565715, + "grad_norm": 0.17275094985961914, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0395, + "step": 16370 + }, + { + "epoch": 1.4888202145064533, + "grad_norm": 0.21093790233135223, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0375, + "step": 16380 + }, + { + "epoch": 1.4897291401563353, + "grad_norm": 0.1446392983198166, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0406, + "step": 16390 + }, + { + "epoch": 1.490638065806217, + "grad_norm": 0.15108722448349, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0391, + "step": 16400 + }, + { + "epoch": 1.4915469914560988, + "grad_norm": 0.21775616705417633, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0404, + "step": 16410 + }, + { + "epoch": 1.4924559171059808, + "grad_norm": 0.12149787694215775, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0383, + "step": 16420 + }, + { + "epoch": 1.4933648427558626, + "grad_norm": 0.17760063707828522, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0398, + "step": 16430 + }, + { + "epoch": 1.4942737684057443, + "grad_norm": 0.18369942903518677, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0379, + "step": 16440 + }, + { + "epoch": 1.4951826940556263, + "grad_norm": 0.1967269480228424, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0423, + "step": 16450 + }, + { + "epoch": 1.496091619705508, + "grad_norm": 0.1688293218612671, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0354, + "step": 16460 + }, + { + "epoch": 1.49700054535539, + "grad_norm": 0.1992066651582718, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0363, + "step": 16470 + }, + { + "epoch": 1.4979094710052718, + "grad_norm": 0.14602915942668915, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0405, + "step": 16480 + }, + { + "epoch": 1.4988183966551536, + "grad_norm": 0.22522462904453278, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0435, + "step": 16490 + }, + { + "epoch": 1.4997273223050356, + "grad_norm": 0.16902150213718414, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0384, + "step": 16500 + }, + { + "epoch": 1.5006362479549171, + "grad_norm": 0.21683278679847717, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0391, + "step": 16510 + }, + { + "epoch": 1.501545173604799, + "grad_norm": 0.12451165169477463, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0378, + "step": 16520 + }, + { + "epoch": 1.502454099254681, + "grad_norm": 0.18156500160694122, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0399, + "step": 16530 + }, + { + "epoch": 1.5033630249045629, + "grad_norm": 0.14581304788589478, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0369, + "step": 16540 + }, + { + "epoch": 1.5042719505544446, + "grad_norm": 0.15864868462085724, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0386, + "step": 16550 + }, + { + "epoch": 1.5051808762043266, + "grad_norm": 0.17344826459884644, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0377, + "step": 16560 + }, + { + "epoch": 1.5060898018542084, + "grad_norm": 0.14846351742744446, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0378, + "step": 16570 + }, + { + "epoch": 1.5069987275040901, + "grad_norm": 0.17173129320144653, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0397, + "step": 16580 + }, + { + "epoch": 1.5079076531539721, + "grad_norm": 0.1805376559495926, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0407, + "step": 16590 + }, + { + "epoch": 1.508816578803854, + "grad_norm": 0.1774982511997223, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0413, + "step": 16600 + }, + { + "epoch": 1.5097255044537357, + "grad_norm": 0.18601776659488678, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0405, + "step": 16610 + }, + { + "epoch": 1.5106344301036176, + "grad_norm": 0.12554484605789185, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0383, + "step": 16620 + }, + { + "epoch": 1.5115433557534994, + "grad_norm": 0.21339905261993408, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0376, + "step": 16630 + }, + { + "epoch": 1.5124522814033812, + "grad_norm": 0.16442753374576569, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0414, + "step": 16640 + }, + { + "epoch": 1.5133612070532632, + "grad_norm": 0.20595991611480713, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0385, + "step": 16650 + }, + { + "epoch": 1.514270132703145, + "grad_norm": 0.1540970802307129, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0374, + "step": 16660 + }, + { + "epoch": 1.5151790583530267, + "grad_norm": 0.20227743685245514, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0421, + "step": 16670 + }, + { + "epoch": 1.5160879840029087, + "grad_norm": 0.1291247308254242, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0372, + "step": 16680 + }, + { + "epoch": 1.5169969096527904, + "grad_norm": 0.18134325742721558, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0412, + "step": 16690 + }, + { + "epoch": 1.5179058353026722, + "grad_norm": 0.2508258819580078, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0404, + "step": 16700 + }, + { + "epoch": 1.5188147609525542, + "grad_norm": 0.23730653524398804, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0374, + "step": 16710 + }, + { + "epoch": 1.519723686602436, + "grad_norm": 0.20907984673976898, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0371, + "step": 16720 + }, + { + "epoch": 1.5206326122523177, + "grad_norm": 0.16148777306079865, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0413, + "step": 16730 + }, + { + "epoch": 1.5215415379021997, + "grad_norm": 0.22122643887996674, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0386, + "step": 16740 + }, + { + "epoch": 1.5224504635520815, + "grad_norm": 0.18458682298660278, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0399, + "step": 16750 + }, + { + "epoch": 1.5233593892019632, + "grad_norm": 0.19485941529273987, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0375, + "step": 16760 + }, + { + "epoch": 1.5242683148518452, + "grad_norm": 0.15232306718826294, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0404, + "step": 16770 + }, + { + "epoch": 1.525177240501727, + "grad_norm": 0.16139771044254303, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0374, + "step": 16780 + }, + { + "epoch": 1.5260861661516087, + "grad_norm": 0.12622502446174622, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0379, + "step": 16790 + }, + { + "epoch": 1.5269950918014907, + "grad_norm": 0.17319156229496002, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0401, + "step": 16800 + }, + { + "epoch": 1.5279040174513725, + "grad_norm": 0.19729571044445038, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0402, + "step": 16810 + }, + { + "epoch": 1.5288129431012543, + "grad_norm": 0.15117479860782623, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0384, + "step": 16820 + }, + { + "epoch": 1.5297218687511362, + "grad_norm": 0.18889816105365753, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0445, + "step": 16830 + }, + { + "epoch": 1.530630794401018, + "grad_norm": 0.23158949613571167, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0365, + "step": 16840 + }, + { + "epoch": 1.5315397200508998, + "grad_norm": 0.14534789323806763, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0378, + "step": 16850 + }, + { + "epoch": 1.5324486457007818, + "grad_norm": 0.22161023318767548, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0369, + "step": 16860 + }, + { + "epoch": 1.5333575713506635, + "grad_norm": 0.15021342039108276, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0374, + "step": 16870 + }, + { + "epoch": 1.5342664970005453, + "grad_norm": 0.1449742615222931, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0388, + "step": 16880 + }, + { + "epoch": 1.5351754226504273, + "grad_norm": 0.16184403002262115, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0392, + "step": 16890 + }, + { + "epoch": 1.536084348300309, + "grad_norm": 0.16897977888584137, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0399, + "step": 16900 + }, + { + "epoch": 1.5369932739501908, + "grad_norm": 0.21523526310920715, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0383, + "step": 16910 + }, + { + "epoch": 1.5379021996000728, + "grad_norm": 0.15990610420703888, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0374, + "step": 16920 + }, + { + "epoch": 1.5388111252499546, + "grad_norm": 0.16797508299350739, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0367, + "step": 16930 + }, + { + "epoch": 1.5397200508998363, + "grad_norm": 0.16012004017829895, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0382, + "step": 16940 + }, + { + "epoch": 1.5406289765497183, + "grad_norm": 0.15826022624969482, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0387, + "step": 16950 + }, + { + "epoch": 1.5415379021996, + "grad_norm": 0.16351309418678284, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0394, + "step": 16960 + }, + { + "epoch": 1.5424468278494818, + "grad_norm": 0.1205107793211937, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0345, + "step": 16970 + }, + { + "epoch": 1.5433557534993638, + "grad_norm": 0.2112058699131012, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0377, + "step": 16980 + }, + { + "epoch": 1.5442646791492456, + "grad_norm": 0.1651793122291565, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0404, + "step": 16990 + }, + { + "epoch": 1.5451736047991274, + "grad_norm": 0.17407971620559692, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0397, + "step": 17000 + }, + { + "epoch": 1.5460825304490093, + "grad_norm": 0.14657418429851532, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0406, + "step": 17010 + }, + { + "epoch": 1.546991456098891, + "grad_norm": 0.20686928927898407, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0379, + "step": 17020 + }, + { + "epoch": 1.5479003817487729, + "grad_norm": 0.3946498930454254, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0372, + "step": 17030 + }, + { + "epoch": 1.5488093073986549, + "grad_norm": 0.1475900262594223, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.038, + "step": 17040 + }, + { + "epoch": 1.5497182330485366, + "grad_norm": 0.17923687398433685, + "learning_rate": 2.931001555898993e-05, + "loss": 0.039, + "step": 17050 + }, + { + "epoch": 1.5506271586984184, + "grad_norm": 0.17597466707229614, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0405, + "step": 17060 + }, + { + "epoch": 1.5515360843483004, + "grad_norm": 0.1819058209657669, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0363, + "step": 17070 + }, + { + "epoch": 1.5524450099981821, + "grad_norm": 0.16790081560611725, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0408, + "step": 17080 + }, + { + "epoch": 1.553353935648064, + "grad_norm": 0.2149350643157959, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0394, + "step": 17090 + }, + { + "epoch": 1.5542628612979459, + "grad_norm": 0.16653192043304443, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0384, + "step": 17100 + }, + { + "epoch": 1.5551717869478277, + "grad_norm": 0.17945846915245056, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0449, + "step": 17110 + }, + { + "epoch": 1.5560807125977094, + "grad_norm": 0.18150071799755096, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0362, + "step": 17120 + }, + { + "epoch": 1.5569896382475914, + "grad_norm": 0.21757055819034576, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0395, + "step": 17130 + }, + { + "epoch": 1.5578985638974732, + "grad_norm": 0.20305049419403076, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0394, + "step": 17140 + }, + { + "epoch": 1.558807489547355, + "grad_norm": 0.16744528710842133, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0397, + "step": 17150 + }, + { + "epoch": 1.559716415197237, + "grad_norm": 0.17758522927761078, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0389, + "step": 17160 + }, + { + "epoch": 1.5606253408471187, + "grad_norm": 0.15038999915122986, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0378, + "step": 17170 + }, + { + "epoch": 1.5615342664970004, + "grad_norm": 0.19865068793296814, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.043, + "step": 17180 + }, + { + "epoch": 1.5624431921468824, + "grad_norm": 0.16853147745132446, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0373, + "step": 17190 + }, + { + "epoch": 1.5633521177967642, + "grad_norm": 0.17590591311454773, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0399, + "step": 17200 + }, + { + "epoch": 1.564261043446646, + "grad_norm": 0.15085701644420624, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0397, + "step": 17210 + }, + { + "epoch": 1.565169969096528, + "grad_norm": 0.20674899220466614, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0401, + "step": 17220 + }, + { + "epoch": 1.5660788947464097, + "grad_norm": 0.17693305015563965, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0424, + "step": 17230 + }, + { + "epoch": 1.5669878203962915, + "grad_norm": 0.17693060636520386, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0392, + "step": 17240 + }, + { + "epoch": 1.5678967460461735, + "grad_norm": 0.1551772952079773, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0421, + "step": 17250 + }, + { + "epoch": 1.5688056716960552, + "grad_norm": 0.13215573132038116, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0381, + "step": 17260 + }, + { + "epoch": 1.569714597345937, + "grad_norm": 0.16997568309307098, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0382, + "step": 17270 + }, + { + "epoch": 1.570623522995819, + "grad_norm": 0.1669899970293045, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0371, + "step": 17280 + }, + { + "epoch": 1.5715324486457007, + "grad_norm": 0.15723662078380585, + "learning_rate": 2.927995984978089e-05, + "loss": 0.036, + "step": 17290 + }, + { + "epoch": 1.5724413742955825, + "grad_norm": 0.13205552101135254, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0393, + "step": 17300 + }, + { + "epoch": 1.5733502999454645, + "grad_norm": 0.15367437899112701, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0412, + "step": 17310 + }, + { + "epoch": 1.5742592255953463, + "grad_norm": 0.20379139482975006, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0393, + "step": 17320 + }, + { + "epoch": 1.575168151245228, + "grad_norm": 0.20353220403194427, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0377, + "step": 17330 + }, + { + "epoch": 1.57607707689511, + "grad_norm": 0.17886844277381897, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0379, + "step": 17340 + }, + { + "epoch": 1.5769860025449918, + "grad_norm": 0.19211682677268982, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0362, + "step": 17350 + }, + { + "epoch": 1.5778949281948735, + "grad_norm": 0.16223496198654175, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0372, + "step": 17360 + }, + { + "epoch": 1.5788038538447555, + "grad_norm": 0.19166351854801178, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0391, + "step": 17370 + }, + { + "epoch": 1.5797127794946373, + "grad_norm": 0.20787383615970612, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0386, + "step": 17380 + }, + { + "epoch": 1.580621705144519, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0393, + "step": 17390 + }, + { + "epoch": 1.581530630794401, + "grad_norm": 0.24690423905849457, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0418, + "step": 17400 + }, + { + "epoch": 1.5824395564442828, + "grad_norm": 0.1873948574066162, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0363, + "step": 17410 + }, + { + "epoch": 1.5833484820941646, + "grad_norm": 0.17039431631565094, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0365, + "step": 17420 + }, + { + "epoch": 1.5842574077440466, + "grad_norm": 0.15311115980148315, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0371, + "step": 17430 + }, + { + "epoch": 1.5851663333939285, + "grad_norm": 0.15634793043136597, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0351, + "step": 17440 + }, + { + "epoch": 1.58607525904381, + "grad_norm": 0.15091364085674286, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0362, + "step": 17450 + }, + { + "epoch": 1.586984184693692, + "grad_norm": 0.2052987515926361, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0401, + "step": 17460 + }, + { + "epoch": 1.587893110343574, + "grad_norm": 0.1455843448638916, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0366, + "step": 17470 + }, + { + "epoch": 1.5888020359934556, + "grad_norm": 0.16488389670848846, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0389, + "step": 17480 + }, + { + "epoch": 1.5897109616433376, + "grad_norm": 0.17014259099960327, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0385, + "step": 17490 + }, + { + "epoch": 1.5906198872932196, + "grad_norm": 0.13956227898597717, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0383, + "step": 17500 + }, + { + "epoch": 1.5915288129431011, + "grad_norm": 0.15602026879787445, + "learning_rate": 2.925186001126643e-05, + "loss": 0.038, + "step": 17510 + }, + { + "epoch": 1.592437738592983, + "grad_norm": 0.13305197656154633, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0389, + "step": 17520 + }, + { + "epoch": 1.593346664242865, + "grad_norm": 0.15525171160697937, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0435, + "step": 17530 + }, + { + "epoch": 1.5942555898927466, + "grad_norm": 0.20900417864322662, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0398, + "step": 17540 + }, + { + "epoch": 1.5951645155426286, + "grad_norm": 0.15116573870182037, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0381, + "step": 17550 + }, + { + "epoch": 1.5960734411925106, + "grad_norm": 0.16280953586101532, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0382, + "step": 17560 + }, + { + "epoch": 1.5969823668423921, + "grad_norm": 0.15049807727336884, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0396, + "step": 17570 + }, + { + "epoch": 1.5978912924922741, + "grad_norm": 0.23786766827106476, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0423, + "step": 17580 + }, + { + "epoch": 1.5988002181421561, + "grad_norm": 0.16331911087036133, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0399, + "step": 17590 + }, + { + "epoch": 1.5997091437920377, + "grad_norm": 0.1609484702348709, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0403, + "step": 17600 + }, + { + "epoch": 1.6006180694419196, + "grad_norm": 0.17137806117534637, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0386, + "step": 17610 + }, + { + "epoch": 1.6015269950918016, + "grad_norm": 0.18994590640068054, + "learning_rate": 2.923761358355331e-05, + "loss": 0.036, + "step": 17620 + }, + { + "epoch": 1.6024359207416832, + "grad_norm": 0.20598889887332916, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0412, + "step": 17630 + }, + { + "epoch": 1.6033448463915652, + "grad_norm": 0.18531493842601776, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0406, + "step": 17640 + }, + { + "epoch": 1.6042537720414471, + "grad_norm": 0.15451273322105408, + "learning_rate": 2.923370548234891e-05, + "loss": 0.038, + "step": 17650 + }, + { + "epoch": 1.6051626976913287, + "grad_norm": 0.15584300458431244, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0391, + "step": 17660 + }, + { + "epoch": 1.6060716233412107, + "grad_norm": 0.26267310976982117, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0376, + "step": 17670 + }, + { + "epoch": 1.6069805489910927, + "grad_norm": 0.1872933804988861, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0376, + "step": 17680 + }, + { + "epoch": 1.6078894746409742, + "grad_norm": 0.1887940764427185, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0404, + "step": 17690 + }, + { + "epoch": 1.6087984002908562, + "grad_norm": 0.22904549539089203, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0361, + "step": 17700 + }, + { + "epoch": 1.6097073259407382, + "grad_norm": 0.16531111299991608, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0384, + "step": 17710 + }, + { + "epoch": 1.61061625159062, + "grad_norm": 0.20064957439899445, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0381, + "step": 17720 + }, + { + "epoch": 1.6115251772405017, + "grad_norm": 0.17839984595775604, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0389, + "step": 17730 + }, + { + "epoch": 1.6124341028903837, + "grad_norm": 0.18564769625663757, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0357, + "step": 17740 + }, + { + "epoch": 1.6133430285402655, + "grad_norm": 0.151804119348526, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0374, + "step": 17750 + }, + { + "epoch": 1.6142519541901472, + "grad_norm": 0.20935791730880737, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0366, + "step": 17760 + }, + { + "epoch": 1.6151608798400292, + "grad_norm": 0.14543452858924866, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0357, + "step": 17770 + }, + { + "epoch": 1.616069805489911, + "grad_norm": 0.20738843083381653, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0378, + "step": 17780 + }, + { + "epoch": 1.6169787311397927, + "grad_norm": 0.19412481784820557, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0355, + "step": 17790 + }, + { + "epoch": 1.6178876567896747, + "grad_norm": 0.2047697901725769, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0365, + "step": 17800 + }, + { + "epoch": 1.6187965824395565, + "grad_norm": 0.1818288266658783, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0398, + "step": 17810 + }, + { + "epoch": 1.6197055080894383, + "grad_norm": 0.14021702110767365, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0398, + "step": 17820 + }, + { + "epoch": 1.6206144337393202, + "grad_norm": 0.15965205430984497, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0388, + "step": 17830 + }, + { + "epoch": 1.621523359389202, + "grad_norm": 0.17519736289978027, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0385, + "step": 17840 + }, + { + "epoch": 1.6224322850390838, + "grad_norm": 0.19246874749660492, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0427, + "step": 17850 + }, + { + "epoch": 1.6233412106889658, + "grad_norm": 0.15174584090709686, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0403, + "step": 17860 + }, + { + "epoch": 1.6242501363388475, + "grad_norm": 0.18702679872512817, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0394, + "step": 17870 + }, + { + "epoch": 1.6251590619887293, + "grad_norm": 0.17292872071266174, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0398, + "step": 17880 + }, + { + "epoch": 1.6260679876386113, + "grad_norm": 0.17382989823818207, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0363, + "step": 17890 + }, + { + "epoch": 1.626976913288493, + "grad_norm": 0.14256832003593445, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0381, + "step": 17900 + }, + { + "epoch": 1.6278858389383748, + "grad_norm": 0.13661658763885498, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0395, + "step": 17910 + }, + { + "epoch": 1.6287947645882568, + "grad_norm": 0.15462905168533325, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0377, + "step": 17920 + }, + { + "epoch": 1.6297036902381385, + "grad_norm": 0.17015130817890167, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0392, + "step": 17930 + }, + { + "epoch": 1.6306126158880203, + "grad_norm": 0.16835148632526398, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0389, + "step": 17940 + }, + { + "epoch": 1.6315215415379023, + "grad_norm": 0.15207840502262115, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0379, + "step": 17950 + }, + { + "epoch": 1.632430467187784, + "grad_norm": 0.14992618560791016, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0368, + "step": 17960 + }, + { + "epoch": 1.6333393928376658, + "grad_norm": 0.2075468897819519, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0405, + "step": 17970 + }, + { + "epoch": 1.6342483184875478, + "grad_norm": 0.19228313863277435, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0355, + "step": 17980 + }, + { + "epoch": 1.6351572441374296, + "grad_norm": 0.21105322241783142, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0394, + "step": 17990 + }, + { + "epoch": 1.6360661697873113, + "grad_norm": 0.14209431409835815, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0371, + "step": 18000 + }, + { + "epoch": 1.6369750954371933, + "grad_norm": 0.1877410113811493, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0367, + "step": 18010 + }, + { + "epoch": 1.637884021087075, + "grad_norm": 0.19424079358577728, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0391, + "step": 18020 + }, + { + "epoch": 1.6387929467369569, + "grad_norm": 0.1632542908191681, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0397, + "step": 18030 + }, + { + "epoch": 1.6397018723868388, + "grad_norm": 0.1723988950252533, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0403, + "step": 18040 + }, + { + "epoch": 1.6406107980367206, + "grad_norm": 0.1944129467010498, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0393, + "step": 18050 + }, + { + "epoch": 1.6415197236866024, + "grad_norm": 0.16040103137493134, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0372, + "step": 18060 + }, + { + "epoch": 1.6424286493364844, + "grad_norm": 0.1870621144771576, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0386, + "step": 18070 + }, + { + "epoch": 1.6433375749863661, + "grad_norm": 0.1323477029800415, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0391, + "step": 18080 + }, + { + "epoch": 1.6442465006362479, + "grad_norm": 0.20910024642944336, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0379, + "step": 18090 + }, + { + "epoch": 1.6451554262861299, + "grad_norm": 0.12991172075271606, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0339, + "step": 18100 + }, + { + "epoch": 1.6460643519360116, + "grad_norm": 0.19276674091815948, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0396, + "step": 18110 + }, + { + "epoch": 1.6469732775858934, + "grad_norm": 0.17648278176784515, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0371, + "step": 18120 + }, + { + "epoch": 1.6478822032357754, + "grad_norm": 0.14244700968265533, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0361, + "step": 18130 + }, + { + "epoch": 1.6487911288856572, + "grad_norm": 0.2088344395160675, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0462, + "step": 18140 + }, + { + "epoch": 1.649700054535539, + "grad_norm": 0.16321666538715363, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0405, + "step": 18150 + }, + { + "epoch": 1.650608980185421, + "grad_norm": 0.18658830225467682, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0398, + "step": 18160 + }, + { + "epoch": 1.6515179058353027, + "grad_norm": 0.17967332899570465, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0412, + "step": 18170 + }, + { + "epoch": 1.6524268314851844, + "grad_norm": 0.154499813914299, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0391, + "step": 18180 + }, + { + "epoch": 1.6533357571350664, + "grad_norm": 0.16970930993556976, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0365, + "step": 18190 + }, + { + "epoch": 1.6542446827849482, + "grad_norm": 0.14788475632667542, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0353, + "step": 18200 + }, + { + "epoch": 1.65515360843483, + "grad_norm": 0.18028880655765533, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0374, + "step": 18210 + }, + { + "epoch": 1.656062534084712, + "grad_norm": 0.1662997603416443, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0411, + "step": 18220 + }, + { + "epoch": 1.6569714597345937, + "grad_norm": 0.20795029401779175, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0368, + "step": 18230 + }, + { + "epoch": 1.6578803853844755, + "grad_norm": 0.17507652938365936, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0386, + "step": 18240 + }, + { + "epoch": 1.6587893110343574, + "grad_norm": 0.14907115697860718, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0384, + "step": 18250 + }, + { + "epoch": 1.6596982366842392, + "grad_norm": 0.23692277073860168, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0395, + "step": 18260 + }, + { + "epoch": 1.660607162334121, + "grad_norm": 0.1721537560224533, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0374, + "step": 18270 + }, + { + "epoch": 1.661516087984003, + "grad_norm": 0.2054237723350525, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0384, + "step": 18280 + }, + { + "epoch": 1.6624250136338847, + "grad_norm": 0.14136147499084473, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0382, + "step": 18290 + }, + { + "epoch": 1.6633339392837665, + "grad_norm": 0.1611238270998001, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0386, + "step": 18300 + }, + { + "epoch": 1.6642428649336485, + "grad_norm": 0.2007000595331192, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0381, + "step": 18310 + }, + { + "epoch": 1.6651517905835302, + "grad_norm": 0.1258324682712555, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.037, + "step": 18320 + }, + { + "epoch": 1.666060716233412, + "grad_norm": 0.20020857453346252, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0393, + "step": 18330 + }, + { + "epoch": 1.666969641883294, + "grad_norm": 0.13455772399902344, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.038, + "step": 18340 + }, + { + "epoch": 1.6678785675331758, + "grad_norm": 0.1900779902935028, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0359, + "step": 18350 + }, + { + "epoch": 1.6687874931830575, + "grad_norm": 0.16852067410945892, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0406, + "step": 18360 + }, + { + "epoch": 1.6696964188329395, + "grad_norm": 0.1600334495306015, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0399, + "step": 18370 + }, + { + "epoch": 1.6706053444828213, + "grad_norm": 0.20120613276958466, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.041, + "step": 18380 + }, + { + "epoch": 1.671514270132703, + "grad_norm": 0.1877945214509964, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0371, + "step": 18390 + }, + { + "epoch": 1.672423195782585, + "grad_norm": 0.18012899160385132, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0381, + "step": 18400 + }, + { + "epoch": 1.6733321214324668, + "grad_norm": 0.22687949240207672, + "learning_rate": 2.91314625716846e-05, + "loss": 0.042, + "step": 18410 + }, + { + "epoch": 1.6742410470823486, + "grad_norm": 0.1620958000421524, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0367, + "step": 18420 + }, + { + "epoch": 1.6751499727322305, + "grad_norm": 0.22850453853607178, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0396, + "step": 18430 + }, + { + "epoch": 1.6760588983821123, + "grad_norm": 0.1552967131137848, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0349, + "step": 18440 + }, + { + "epoch": 1.676967824031994, + "grad_norm": 0.17952404916286469, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0373, + "step": 18450 + }, + { + "epoch": 1.677876749681876, + "grad_norm": 0.236776664853096, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0407, + "step": 18460 + }, + { + "epoch": 1.6787856753317578, + "grad_norm": 0.1819484382867813, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0371, + "step": 18470 + }, + { + "epoch": 1.6796946009816396, + "grad_norm": 0.2068413645029068, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0367, + "step": 18480 + }, + { + "epoch": 1.6806035266315216, + "grad_norm": 0.1641455441713333, + "learning_rate": 2.912033877795915e-05, + "loss": 0.036, + "step": 18490 + }, + { + "epoch": 1.6815124522814033, + "grad_norm": 0.19677025079727173, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0376, + "step": 18500 + }, + { + "epoch": 1.682421377931285, + "grad_norm": 0.1517343670129776, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0385, + "step": 18510 + }, + { + "epoch": 1.683330303581167, + "grad_norm": 0.1730227768421173, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0396, + "step": 18520 + }, + { + "epoch": 1.6842392292310489, + "grad_norm": 0.21976521611213684, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0371, + "step": 18530 + }, + { + "epoch": 1.6851481548809306, + "grad_norm": 0.16427528858184814, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0378, + "step": 18540 + }, + { + "epoch": 1.6860570805308126, + "grad_norm": 0.19392907619476318, + "learning_rate": 2.91119508911589e-05, + "loss": 0.036, + "step": 18550 + }, + { + "epoch": 1.6869660061806944, + "grad_norm": 0.14583352208137512, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0363, + "step": 18560 + }, + { + "epoch": 1.6878749318305761, + "grad_norm": 0.11771427094936371, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0393, + "step": 18570 + }, + { + "epoch": 1.6887838574804581, + "grad_norm": 0.19352379441261292, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 1.68969278313034, + "grad_norm": 0.18184368312358856, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0382, + "step": 18590 + }, + { + "epoch": 1.6906017087802216, + "grad_norm": 0.19588245451450348, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0408, + "step": 18600 + }, + { + "epoch": 1.6915106344301036, + "grad_norm": 0.14529134333133698, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0376, + "step": 18610 + }, + { + "epoch": 1.6924195600799856, + "grad_norm": 0.16566011309623718, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0362, + "step": 18620 + }, + { + "epoch": 1.6933284857298672, + "grad_norm": 0.16586361825466156, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0374, + "step": 18630 + }, + { + "epoch": 1.6942374113797491, + "grad_norm": 0.17962464690208435, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0374, + "step": 18640 + }, + { + "epoch": 1.6951463370296311, + "grad_norm": 0.17814703285694122, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0353, + "step": 18650 + }, + { + "epoch": 1.6960552626795127, + "grad_norm": 0.16261005401611328, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0379, + "step": 18660 + }, + { + "epoch": 1.6969641883293947, + "grad_norm": 0.15648753941059113, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0365, + "step": 18670 + }, + { + "epoch": 1.6978731139792766, + "grad_norm": 0.18658742308616638, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0396, + "step": 18680 + }, + { + "epoch": 1.6987820396291582, + "grad_norm": 0.20040805637836456, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.0382, + "step": 18690 + }, + { + "epoch": 1.6996909652790402, + "grad_norm": 0.17315340042114258, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0389, + "step": 18700 + }, + { + "epoch": 1.7005998909289222, + "grad_norm": 0.165126770734787, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0385, + "step": 18710 + }, + { + "epoch": 1.7015088165788037, + "grad_norm": 0.19653916358947754, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0371, + "step": 18720 + }, + { + "epoch": 1.7024177422286857, + "grad_norm": 0.18804271519184113, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.036, + "step": 18730 + }, + { + "epoch": 1.7033266678785677, + "grad_norm": 0.21867698431015015, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0396, + "step": 18740 + }, + { + "epoch": 1.7042355935284492, + "grad_norm": 0.151628777384758, + "learning_rate": 2.908371275367924e-05, + "loss": 0.041, + "step": 18750 + }, + { + "epoch": 1.7051445191783312, + "grad_norm": 0.16727173328399658, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0368, + "step": 18760 + }, + { + "epoch": 1.7060534448282132, + "grad_norm": 0.2066250741481781, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0381, + "step": 18770 + }, + { + "epoch": 1.7069623704780947, + "grad_norm": 0.18661518394947052, + "learning_rate": 2.907944011202334e-05, + "loss": 0.035, + "step": 18780 + }, + { + "epoch": 1.7078712961279767, + "grad_norm": 0.156740203499794, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0371, + "step": 18790 + }, + { + "epoch": 1.7087802217778587, + "grad_norm": 0.20875664055347443, + "learning_rate": 2.907658633786968e-05, + "loss": 0.041, + "step": 18800 + }, + { + "epoch": 1.7096891474277403, + "grad_norm": 0.18405385315418243, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0363, + "step": 18810 + }, + { + "epoch": 1.7105980730776222, + "grad_norm": 0.13181810081005096, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0385, + "step": 18820 + }, + { + "epoch": 1.7115069987275042, + "grad_norm": 0.19153928756713867, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0347, + "step": 18830 + }, + { + "epoch": 1.7124159243773858, + "grad_norm": 0.1934434175491333, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.037, + "step": 18840 + }, + { + "epoch": 1.7133248500272678, + "grad_norm": 0.17574919760227203, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0399, + "step": 18850 + }, + { + "epoch": 1.7142337756771497, + "grad_norm": 0.13189968466758728, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0385, + "step": 18860 + }, + { + "epoch": 1.7151427013270313, + "grad_norm": 0.1279727667570114, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0362, + "step": 18870 + }, + { + "epoch": 1.7160516269769133, + "grad_norm": 0.14184506237506866, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0365, + "step": 18880 + }, + { + "epoch": 1.7169605526267953, + "grad_norm": 0.2010291963815689, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0372, + "step": 18890 + }, + { + "epoch": 1.717869478276677, + "grad_norm": 0.18556733429431915, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0374, + "step": 18900 + }, + { + "epoch": 1.7187784039265588, + "grad_norm": 0.2656380534172058, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0385, + "step": 18910 + }, + { + "epoch": 1.7196873295764408, + "grad_norm": 0.14113198220729828, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0387, + "step": 18920 + }, + { + "epoch": 1.7205962552263225, + "grad_norm": 0.1614321917295456, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0367, + "step": 18930 + }, + { + "epoch": 1.7215051808762043, + "grad_norm": 0.15641333162784576, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0355, + "step": 18940 + }, + { + "epoch": 1.7224141065260863, + "grad_norm": 0.18698807060718536, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0375, + "step": 18950 + }, + { + "epoch": 1.723323032175968, + "grad_norm": 0.16441506147384644, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0365, + "step": 18960 + }, + { + "epoch": 1.7242319578258498, + "grad_norm": 0.19547513127326965, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0389, + "step": 18970 + }, + { + "epoch": 1.7251408834757318, + "grad_norm": 0.16130085289478302, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0385, + "step": 18980 + }, + { + "epoch": 1.7260498091256136, + "grad_norm": 0.15705044567584991, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0379, + "step": 18990 + }, + { + "epoch": 1.7269587347754953, + "grad_norm": 0.1605977863073349, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0376, + "step": 19000 + }, + { + "epoch": 1.7278676604253773, + "grad_norm": 0.16556136310100555, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0367, + "step": 19010 + }, + { + "epoch": 1.728776586075259, + "grad_norm": 0.1839601993560791, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0355, + "step": 19020 + }, + { + "epoch": 1.7296855117251408, + "grad_norm": 0.1491796374320984, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.038, + "step": 19030 + }, + { + "epoch": 1.7305944373750228, + "grad_norm": 0.19307903945446014, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0409, + "step": 19040 + }, + { + "epoch": 1.7315033630249046, + "grad_norm": 0.2098098248243332, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0389, + "step": 19050 + }, + { + "epoch": 1.7324122886747864, + "grad_norm": 0.20307697355747223, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0401, + "step": 19060 + }, + { + "epoch": 1.7333212143246683, + "grad_norm": 0.16241510212421417, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0396, + "step": 19070 + }, + { + "epoch": 1.7342301399745501, + "grad_norm": 0.12922362983226776, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0366, + "step": 19080 + }, + { + "epoch": 1.7351390656244319, + "grad_norm": 0.13932615518569946, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0363, + "step": 19090 + }, + { + "epoch": 1.7360479912743139, + "grad_norm": 0.16777463257312775, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0411, + "step": 19100 + }, + { + "epoch": 1.7369569169241956, + "grad_norm": 0.2092953324317932, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0408, + "step": 19110 + }, + { + "epoch": 1.7378658425740774, + "grad_norm": 0.2010180801153183, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0378, + "step": 19120 + }, + { + "epoch": 1.7387747682239594, + "grad_norm": 0.16410458087921143, + "learning_rate": 2.90288823542141e-05, + "loss": 0.037, + "step": 19130 + }, + { + "epoch": 1.7396836938738411, + "grad_norm": 0.21531733870506287, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0383, + "step": 19140 + }, + { + "epoch": 1.740592619523723, + "grad_norm": 0.2002219259738922, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0377, + "step": 19150 + }, + { + "epoch": 1.741501545173605, + "grad_norm": 0.18431171774864197, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0384, + "step": 19160 + }, + { + "epoch": 1.7424104708234867, + "grad_norm": 0.14828172326087952, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0351, + "step": 19170 + }, + { + "epoch": 1.7433193964733684, + "grad_norm": 0.1445445418357849, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0377, + "step": 19180 + }, + { + "epoch": 1.7442283221232504, + "grad_norm": 0.1845911741256714, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0383, + "step": 19190 + }, + { + "epoch": 1.7451372477731322, + "grad_norm": 0.1779165416955948, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0408, + "step": 19200 + }, + { + "epoch": 1.746046173423014, + "grad_norm": 0.1918453574180603, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0374, + "step": 19210 + }, + { + "epoch": 1.746955099072896, + "grad_norm": 0.1753605306148529, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0367, + "step": 19220 + }, + { + "epoch": 1.7478640247227777, + "grad_norm": 0.16597093641757965, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0381, + "step": 19230 + }, + { + "epoch": 1.7487729503726595, + "grad_norm": 0.17037779092788696, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0356, + "step": 19240 + }, + { + "epoch": 1.7496818760225414, + "grad_norm": 0.21845535933971405, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0387, + "step": 19250 + }, + { + "epoch": 1.7505908016724232, + "grad_norm": 0.12781749665737152, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0383, + "step": 19260 + }, + { + "epoch": 1.751499727322305, + "grad_norm": 0.18625755608081818, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0353, + "step": 19270 + }, + { + "epoch": 1.752408652972187, + "grad_norm": 0.1777392029762268, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0402, + "step": 19280 + }, + { + "epoch": 1.7533175786220687, + "grad_norm": 0.18146635591983795, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0388, + "step": 19290 + }, + { + "epoch": 1.7542265042719505, + "grad_norm": 0.22970373928546906, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.039, + "step": 19300 + }, + { + "epoch": 1.7551354299218325, + "grad_norm": 0.19050046801567078, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0369, + "step": 19310 + }, + { + "epoch": 1.7560443555717142, + "grad_norm": 0.15904827415943146, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0388, + "step": 19320 + }, + { + "epoch": 1.756953281221596, + "grad_norm": 0.16599079966545105, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0363, + "step": 19330 + }, + { + "epoch": 1.757862206871478, + "grad_norm": 0.15301232039928436, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0388, + "step": 19340 + }, + { + "epoch": 1.7587711325213597, + "grad_norm": 0.1578368842601776, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0365, + "step": 19350 + }, + { + "epoch": 1.7596800581712415, + "grad_norm": 0.18845772743225098, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0339, + "step": 19360 + }, + { + "epoch": 1.7605889838211235, + "grad_norm": 0.14664173126220703, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0385, + "step": 19370 + }, + { + "epoch": 1.7614979094710053, + "grad_norm": 0.16458362340927124, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0354, + "step": 19380 + }, + { + "epoch": 1.762406835120887, + "grad_norm": 0.1902923732995987, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0366, + "step": 19390 + }, + { + "epoch": 1.763315760770769, + "grad_norm": 0.174685537815094, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0379, + "step": 19400 + }, + { + "epoch": 1.7642246864206508, + "grad_norm": 0.14696656167507172, + "learning_rate": 2.898749624168966e-05, + "loss": 0.038, + "step": 19410 + }, + { + "epoch": 1.7651336120705325, + "grad_norm": 0.15568022429943085, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0376, + "step": 19420 + }, + { + "epoch": 1.7660425377204145, + "grad_norm": 0.2163154035806656, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0371, + "step": 19430 + }, + { + "epoch": 1.7669514633702963, + "grad_norm": 0.16677317023277283, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0394, + "step": 19440 + }, + { + "epoch": 1.767860389020178, + "grad_norm": 0.12284742295742035, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0346, + "step": 19450 + }, + { + "epoch": 1.76876931467006, + "grad_norm": 0.14316865801811218, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0335, + "step": 19460 + }, + { + "epoch": 1.7696782403199418, + "grad_norm": 0.15991491079330444, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0347, + "step": 19470 + }, + { + "epoch": 1.7705871659698236, + "grad_norm": 0.15489616990089417, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0354, + "step": 19480 + }, + { + "epoch": 1.7714960916197056, + "grad_norm": 0.17272238433361053, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.035, + "step": 19490 + }, + { + "epoch": 1.7724050172695873, + "grad_norm": 0.13315513730049133, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0341, + "step": 19500 + }, + { + "epoch": 1.773313942919469, + "grad_norm": 0.17086821794509888, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0404, + "step": 19510 + }, + { + "epoch": 1.774222868569351, + "grad_norm": 0.2021840512752533, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0356, + "step": 19520 + }, + { + "epoch": 1.7751317942192328, + "grad_norm": 0.13640274107456207, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0397, + "step": 19530 + }, + { + "epoch": 1.7760407198691146, + "grad_norm": 0.19203563034534454, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0348, + "step": 19540 + }, + { + "epoch": 1.7769496455189966, + "grad_norm": 0.22474288940429688, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0378, + "step": 19550 + }, + { + "epoch": 1.7778585711688784, + "grad_norm": 0.166643425822258, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0398, + "step": 19560 + }, + { + "epoch": 1.7787674968187601, + "grad_norm": 0.1704113930463791, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0395, + "step": 19570 + }, + { + "epoch": 1.779676422468642, + "grad_norm": 0.20513354241847992, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0379, + "step": 19580 + }, + { + "epoch": 1.7805853481185239, + "grad_norm": 0.2448979914188385, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0354, + "step": 19590 + }, + { + "epoch": 1.7814942737684056, + "grad_norm": 0.14778435230255127, + "learning_rate": 2.895893835994484e-05, + "loss": 0.0365, + "step": 19600 + }, + { + "epoch": 1.7824031994182876, + "grad_norm": 0.15804095566272736, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0369, + "step": 19610 + }, + { + "epoch": 1.7833121250681694, + "grad_norm": 0.18581096827983856, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0357, + "step": 19620 + }, + { + "epoch": 1.7842210507180511, + "grad_norm": 0.12598079442977905, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0374, + "step": 19630 + }, + { + "epoch": 1.7851299763679331, + "grad_norm": 0.12576445937156677, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0362, + "step": 19640 + }, + { + "epoch": 1.786038902017815, + "grad_norm": 0.17113761603832245, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0396, + "step": 19650 + }, + { + "epoch": 1.7869478276676967, + "grad_norm": 0.1763591170310974, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0392, + "step": 19660 + }, + { + "epoch": 1.7878567533175787, + "grad_norm": 0.191595658659935, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.039, + "step": 19670 + }, + { + "epoch": 1.7887656789674604, + "grad_norm": 0.16918224096298218, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0348, + "step": 19680 + }, + { + "epoch": 1.7896746046173422, + "grad_norm": 0.231667622923851, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0365, + "step": 19690 + }, + { + "epoch": 1.7905835302672242, + "grad_norm": 0.13641966879367828, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0391, + "step": 19700 + }, + { + "epoch": 1.791492455917106, + "grad_norm": 0.13582776486873627, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0319, + "step": 19710 + }, + { + "epoch": 1.7924013815669877, + "grad_norm": 0.1977994740009308, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0362, + "step": 19720 + }, + { + "epoch": 1.7933103072168697, + "grad_norm": 0.21458618342876434, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0372, + "step": 19730 + }, + { + "epoch": 1.7942192328667514, + "grad_norm": 0.14885760843753815, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.035, + "step": 19740 + }, + { + "epoch": 1.7951281585166332, + "grad_norm": 0.15464016795158386, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0372, + "step": 19750 + }, + { + "epoch": 1.7960370841665152, + "grad_norm": 0.13073775172233582, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0381, + "step": 19760 + }, + { + "epoch": 1.7969460098163972, + "grad_norm": 0.20851127803325653, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0379, + "step": 19770 + }, + { + "epoch": 1.7978549354662787, + "grad_norm": 0.14128346741199493, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0345, + "step": 19780 + }, + { + "epoch": 1.7987638611161607, + "grad_norm": 0.1834966540336609, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0378, + "step": 19790 + }, + { + "epoch": 1.7996727867660427, + "grad_norm": 0.17173343896865845, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0392, + "step": 19800 + }, + { + "epoch": 1.8005817124159242, + "grad_norm": 0.1496376395225525, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0366, + "step": 19810 + }, + { + "epoch": 1.8014906380658062, + "grad_norm": 0.18145428597927094, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0337, + "step": 19820 + }, + { + "epoch": 1.8023995637156882, + "grad_norm": 0.14671723544597626, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0366, + "step": 19830 + }, + { + "epoch": 1.8033084893655698, + "grad_norm": 0.15412470698356628, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0372, + "step": 19840 + }, + { + "epoch": 1.8042174150154517, + "grad_norm": 0.15045300126075745, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0372, + "step": 19850 + }, + { + "epoch": 1.8051263406653337, + "grad_norm": 0.17146240174770355, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0407, + "step": 19860 + }, + { + "epoch": 1.8060352663152153, + "grad_norm": 0.17537301778793335, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0373, + "step": 19870 + }, + { + "epoch": 1.8069441919650973, + "grad_norm": 0.2008477747440338, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0347, + "step": 19880 + }, + { + "epoch": 1.8078531176149792, + "grad_norm": 0.19546300172805786, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0373, + "step": 19890 + }, + { + "epoch": 1.8087620432648608, + "grad_norm": 0.17281341552734375, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0346, + "step": 19900 + }, + { + "epoch": 1.8096709689147428, + "grad_norm": 0.17090924084186554, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0399, + "step": 19910 + }, + { + "epoch": 1.8105798945646248, + "grad_norm": 0.1414732187986374, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0341, + "step": 19920 + }, + { + "epoch": 1.8114888202145063, + "grad_norm": 0.16682983934879303, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0412, + "step": 19930 + }, + { + "epoch": 1.8123977458643883, + "grad_norm": 0.1648736149072647, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0363, + "step": 19940 + }, + { + "epoch": 1.8133066715142703, + "grad_norm": 0.17931930720806122, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0354, + "step": 19950 + }, + { + "epoch": 1.8142155971641518, + "grad_norm": 0.14836636185646057, + "learning_rate": 2.89037798520224e-05, + "loss": 0.039, + "step": 19960 + }, + { + "epoch": 1.8151245228140338, + "grad_norm": 0.1760379523038864, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0362, + "step": 19970 + }, + { + "epoch": 1.8160334484639158, + "grad_norm": 0.15899164974689484, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0375, + "step": 19980 + }, + { + "epoch": 1.8169423741137973, + "grad_norm": 0.1795760691165924, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0321, + "step": 19990 + }, + { + "epoch": 1.8178512997636793, + "grad_norm": 0.15427102148532867, + "learning_rate": 2.889756655699208e-05, + "loss": 0.038, + "step": 20000 + }, + { + "epoch": 1.8187602254135613, + "grad_norm": 0.22279855608940125, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0339, + "step": 20010 + }, + { + "epoch": 1.8196691510634428, + "grad_norm": 0.17210164666175842, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0357, + "step": 20020 + }, + { + "epoch": 1.8205780767133248, + "grad_norm": 0.1745513379573822, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0399, + "step": 20030 + }, + { + "epoch": 1.8214870023632068, + "grad_norm": 0.16305135190486908, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0361, + "step": 20040 + }, + { + "epoch": 1.8223959280130886, + "grad_norm": 0.1543959677219391, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0328, + "step": 20050 + }, + { + "epoch": 1.8233048536629703, + "grad_norm": 0.18696659803390503, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0364, + "step": 20060 + }, + { + "epoch": 1.8242137793128523, + "grad_norm": 0.16789674758911133, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0359, + "step": 20070 + }, + { + "epoch": 1.825122704962734, + "grad_norm": 0.18416273593902588, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0371, + "step": 20080 + }, + { + "epoch": 1.8260316306126159, + "grad_norm": 0.1565874218940735, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0353, + "step": 20090 + }, + { + "epoch": 1.8269405562624979, + "grad_norm": 0.15700258314609528, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0368, + "step": 20100 + }, + { + "epoch": 1.8278494819123796, + "grad_norm": 0.15921398997306824, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0364, + "step": 20110 + }, + { + "epoch": 1.8287584075622614, + "grad_norm": 0.1776263415813446, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0366, + "step": 20120 + }, + { + "epoch": 1.8296673332121434, + "grad_norm": 0.1830349564552307, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0372, + "step": 20130 + }, + { + "epoch": 1.8305762588620251, + "grad_norm": 0.19790445268154144, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0357, + "step": 20140 + }, + { + "epoch": 1.831485184511907, + "grad_norm": 0.16671213507652283, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0369, + "step": 20150 + }, + { + "epoch": 1.8323941101617889, + "grad_norm": 0.1625368595123291, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0349, + "step": 20160 + }, + { + "epoch": 1.8333030358116706, + "grad_norm": 0.21950876712799072, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0372, + "step": 20170 + }, + { + "epoch": 1.8342119614615524, + "grad_norm": 0.1345209926366806, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0377, + "step": 20180 + }, + { + "epoch": 1.8351208871114344, + "grad_norm": 0.16477446258068085, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0365, + "step": 20190 + }, + { + "epoch": 1.8360298127613162, + "grad_norm": 0.19538794457912445, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0361, + "step": 20200 + }, + { + "epoch": 1.836938738411198, + "grad_norm": 0.15050725638866425, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0401, + "step": 20210 + }, + { + "epoch": 1.83784766406108, + "grad_norm": 0.20276984572410583, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0391, + "step": 20220 + }, + { + "epoch": 1.8387565897109617, + "grad_norm": 0.1354019045829773, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0368, + "step": 20230 + }, + { + "epoch": 1.8396655153608434, + "grad_norm": 0.10410725325345993, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0365, + "step": 20240 + }, + { + "epoch": 1.8405744410107254, + "grad_norm": 0.17227144539356232, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0345, + "step": 20250 + }, + { + "epoch": 1.8414833666606072, + "grad_norm": 0.1584329605102539, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0392, + "step": 20260 + }, + { + "epoch": 1.842392292310489, + "grad_norm": 0.1574667990207672, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0365, + "step": 20270 + }, + { + "epoch": 1.843301217960371, + "grad_norm": 0.16814886033535004, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0355, + "step": 20280 + }, + { + "epoch": 1.8442101436102527, + "grad_norm": 0.16504709422588348, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.035, + "step": 20290 + }, + { + "epoch": 1.8451190692601345, + "grad_norm": 0.15797556936740875, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0363, + "step": 20300 + }, + { + "epoch": 1.8460279949100165, + "grad_norm": 0.14572837948799133, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0369, + "step": 20310 + }, + { + "epoch": 1.8469369205598982, + "grad_norm": 0.17166292667388916, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0364, + "step": 20320 + }, + { + "epoch": 1.84784584620978, + "grad_norm": 0.1312624216079712, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0375, + "step": 20330 + }, + { + "epoch": 1.848754771859662, + "grad_norm": 0.14530129730701447, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0338, + "step": 20340 + }, + { + "epoch": 1.8496636975095437, + "grad_norm": 0.17239950597286224, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.037, + "step": 20350 + }, + { + "epoch": 1.8505726231594255, + "grad_norm": 0.14343880116939545, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0375, + "step": 20360 + }, + { + "epoch": 1.8514815488093075, + "grad_norm": 0.16703525185585022, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0361, + "step": 20370 + }, + { + "epoch": 1.8523904744591893, + "grad_norm": 0.16017450392246246, + "learning_rate": 2.8837699097573e-05, + "loss": 0.041, + "step": 20380 + }, + { + "epoch": 1.853299400109071, + "grad_norm": 0.1531965434551239, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0335, + "step": 20390 + }, + { + "epoch": 1.854208325758953, + "grad_norm": 0.17072531580924988, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0365, + "step": 20400 + }, + { + "epoch": 1.8551172514088348, + "grad_norm": 0.1558724343776703, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0346, + "step": 20410 + }, + { + "epoch": 1.8560261770587165, + "grad_norm": 0.15449529886245728, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0355, + "step": 20420 + }, + { + "epoch": 1.8569351027085985, + "grad_norm": 0.18946866691112518, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0369, + "step": 20430 + }, + { + "epoch": 1.8578440283584803, + "grad_norm": 0.24861611425876617, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0359, + "step": 20440 + }, + { + "epoch": 1.858752954008362, + "grad_norm": 0.22610847651958466, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.036, + "step": 20450 + }, + { + "epoch": 1.859661879658244, + "grad_norm": 0.18469324707984924, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0361, + "step": 20460 + }, + { + "epoch": 1.8605708053081258, + "grad_norm": 0.15959666669368744, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0371, + "step": 20470 + }, + { + "epoch": 1.8614797309580076, + "grad_norm": 0.16275286674499512, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0392, + "step": 20480 + }, + { + "epoch": 1.8623886566078895, + "grad_norm": 0.17713631689548492, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0388, + "step": 20490 + }, + { + "epoch": 1.8632975822577713, + "grad_norm": 0.17384174466133118, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0335, + "step": 20500 + }, + { + "epoch": 1.864206507907653, + "grad_norm": 0.16565603017807007, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0373, + "step": 20510 + }, + { + "epoch": 1.865115433557535, + "grad_norm": 0.17847464978694916, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0348, + "step": 20520 + }, + { + "epoch": 1.8660243592074168, + "grad_norm": 0.1608586609363556, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0348, + "step": 20530 + }, + { + "epoch": 1.8669332848572986, + "grad_norm": 0.17415191233158112, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0373, + "step": 20540 + }, + { + "epoch": 1.8678422105071806, + "grad_norm": 0.15377484261989594, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0363, + "step": 20550 + }, + { + "epoch": 1.8687511361570623, + "grad_norm": 0.16340234875679016, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0368, + "step": 20560 + }, + { + "epoch": 1.869660061806944, + "grad_norm": 0.25783345103263855, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0359, + "step": 20570 + }, + { + "epoch": 1.870568987456826, + "grad_norm": 0.14801256358623505, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0361, + "step": 20580 + }, + { + "epoch": 1.8714779131067079, + "grad_norm": 0.16208654642105103, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0366, + "step": 20590 + }, + { + "epoch": 1.8723868387565896, + "grad_norm": 0.1767996847629547, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0379, + "step": 20600 + }, + { + "epoch": 1.8732957644064716, + "grad_norm": 0.1903439462184906, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0373, + "step": 20610 + }, + { + "epoch": 1.8742046900563534, + "grad_norm": 0.1528051197528839, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0341, + "step": 20620 + }, + { + "epoch": 1.8751136157062351, + "grad_norm": 0.16984589397907257, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0372, + "step": 20630 + }, + { + "epoch": 1.8760225413561171, + "grad_norm": 0.18681307137012482, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0349, + "step": 20640 + }, + { + "epoch": 1.8769314670059989, + "grad_norm": 0.19160854816436768, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0356, + "step": 20650 + }, + { + "epoch": 1.8778403926558807, + "grad_norm": 0.1614740788936615, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0356, + "step": 20660 + }, + { + "epoch": 1.8787493183057626, + "grad_norm": 0.13602229952812195, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0341, + "step": 20670 + }, + { + "epoch": 1.8796582439556444, + "grad_norm": 0.1979421079158783, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0372, + "step": 20680 + }, + { + "epoch": 1.8805671696055262, + "grad_norm": 0.13565583527088165, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0364, + "step": 20690 + }, + { + "epoch": 1.8814760952554082, + "grad_norm": 0.1763925403356552, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0364, + "step": 20700 + }, + { + "epoch": 1.88238502090529, + "grad_norm": 0.16318942606449127, + "learning_rate": 2.87844776006562e-05, + "loss": 0.035, + "step": 20710 + }, + { + "epoch": 1.8832939465551717, + "grad_norm": 0.17972271144390106, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0407, + "step": 20720 + }, + { + "epoch": 1.8842028722050537, + "grad_norm": 0.14144594967365265, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0357, + "step": 20730 + }, + { + "epoch": 1.8851117978549354, + "grad_norm": 0.1604517102241516, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0366, + "step": 20740 + }, + { + "epoch": 1.8860207235048172, + "grad_norm": 0.19542934000492096, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0372, + "step": 20750 + }, + { + "epoch": 1.8869296491546992, + "grad_norm": 0.16911107301712036, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0337, + "step": 20760 + }, + { + "epoch": 1.887838574804581, + "grad_norm": 0.14185817539691925, + "learning_rate": 2.877467835492537e-05, + "loss": 0.035, + "step": 20770 + }, + { + "epoch": 1.8887475004544627, + "grad_norm": 0.17984262108802795, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0359, + "step": 20780 + }, + { + "epoch": 1.8896564261043447, + "grad_norm": 0.179992213845253, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0335, + "step": 20790 + }, + { + "epoch": 1.8905653517542265, + "grad_norm": 0.16679055988788605, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0328, + "step": 20800 + }, + { + "epoch": 1.8914742774041082, + "grad_norm": 0.1635269820690155, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0388, + "step": 20810 + }, + { + "epoch": 1.8923832030539902, + "grad_norm": 0.1288490742444992, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0327, + "step": 20820 + }, + { + "epoch": 1.893292128703872, + "grad_norm": 0.15894314646720886, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0352, + "step": 20830 + }, + { + "epoch": 1.8942010543537537, + "grad_norm": 0.14257533848285675, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.0324, + "step": 20840 + }, + { + "epoch": 1.8951099800036357, + "grad_norm": 0.17389044165611267, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0344, + "step": 20850 + }, + { + "epoch": 1.8960189056535175, + "grad_norm": 0.13561011850833893, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0345, + "step": 20860 + }, + { + "epoch": 1.8969278313033993, + "grad_norm": 0.16809093952178955, + "learning_rate": 2.875826260432918e-05, + "loss": 0.037, + "step": 20870 + }, + { + "epoch": 1.8978367569532812, + "grad_norm": 0.14069151878356934, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0359, + "step": 20880 + }, + { + "epoch": 1.898745682603163, + "grad_norm": 0.170017272233963, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0344, + "step": 20890 + }, + { + "epoch": 1.8996546082530448, + "grad_norm": 0.15128527581691742, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0381, + "step": 20900 + }, + { + "epoch": 1.9005635339029268, + "grad_norm": 0.1513081043958664, + "learning_rate": 2.875166704132208e-05, + "loss": 0.036, + "step": 20910 + }, + { + "epoch": 1.9014724595528085, + "grad_norm": 0.17864766716957092, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0392, + "step": 20920 + }, + { + "epoch": 1.9023813852026903, + "grad_norm": 0.14738181233406067, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0364, + "step": 20930 + }, + { + "epoch": 1.9032903108525723, + "grad_norm": 0.20803159475326538, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0355, + "step": 20940 + }, + { + "epoch": 1.9041992365024543, + "grad_norm": 0.1349308341741562, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0376, + "step": 20950 + }, + { + "epoch": 1.9051081621523358, + "grad_norm": 0.14553409814834595, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0381, + "step": 20960 + }, + { + "epoch": 1.9060170878022178, + "grad_norm": 0.13183756172657013, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0348, + "step": 20970 + }, + { + "epoch": 1.9069260134520998, + "grad_norm": 0.11552274227142334, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0382, + "step": 20980 + }, + { + "epoch": 1.9078349391019813, + "grad_norm": 0.1365584284067154, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0351, + "step": 20990 + }, + { + "epoch": 1.9087438647518633, + "grad_norm": 0.13688084483146667, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0347, + "step": 21000 + }, + { + "epoch": 1.9096527904017453, + "grad_norm": 0.15719814598560333, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0393, + "step": 21010 + }, + { + "epoch": 1.9105617160516268, + "grad_norm": 0.1489710658788681, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.038, + "step": 21020 + }, + { + "epoch": 1.9114706417015088, + "grad_norm": 0.18657103180885315, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0352, + "step": 21030 + }, + { + "epoch": 1.9123795673513908, + "grad_norm": 0.21824254095554352, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0376, + "step": 21040 + }, + { + "epoch": 1.9132884930012724, + "grad_norm": 0.1674203723669052, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0361, + "step": 21050 + }, + { + "epoch": 1.9141974186511543, + "grad_norm": 0.20556220412254333, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0343, + "step": 21060 + }, + { + "epoch": 1.9151063443010363, + "grad_norm": 0.16456326842308044, + "learning_rate": 2.872511777401895e-05, + "loss": 0.033, + "step": 21070 + }, + { + "epoch": 1.9160152699509179, + "grad_norm": 0.16485995054244995, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0361, + "step": 21080 + }, + { + "epoch": 1.9169241956007999, + "grad_norm": 0.14961227774620056, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0335, + "step": 21090 + }, + { + "epoch": 1.9178331212506818, + "grad_norm": 0.16391538083553314, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0336, + "step": 21100 + }, + { + "epoch": 1.9187420469005634, + "grad_norm": 0.16004227101802826, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0327, + "step": 21110 + }, + { + "epoch": 1.9196509725504454, + "grad_norm": 0.19343344867229462, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0347, + "step": 21120 + }, + { + "epoch": 1.9205598982003274, + "grad_norm": 0.1408742517232895, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.0373, + "step": 21130 + }, + { + "epoch": 1.921468823850209, + "grad_norm": 0.1901286095380783, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0361, + "step": 21140 + }, + { + "epoch": 1.9223777495000909, + "grad_norm": 0.17459644377231598, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0355, + "step": 21150 + }, + { + "epoch": 1.9232866751499729, + "grad_norm": 0.1861971914768219, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0334, + "step": 21160 + }, + { + "epoch": 1.9241956007998544, + "grad_norm": 0.23452194035053253, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0386, + "step": 21170 + }, + { + "epoch": 1.9251045264497364, + "grad_norm": 0.12701556086540222, + "learning_rate": 2.870671033527906e-05, + "loss": 0.036, + "step": 21180 + }, + { + "epoch": 1.9260134520996184, + "grad_norm": 0.16400332748889923, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0368, + "step": 21190 + }, + { + "epoch": 1.9269223777495, + "grad_norm": 0.15591835975646973, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0354, + "step": 21200 + }, + { + "epoch": 1.927831303399382, + "grad_norm": 0.14587388932704926, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0347, + "step": 21210 + }, + { + "epoch": 1.928740229049264, + "grad_norm": 0.21282705664634705, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0378, + "step": 21220 + }, + { + "epoch": 1.9296491546991457, + "grad_norm": 0.18534594774246216, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.036, + "step": 21230 + }, + { + "epoch": 1.9305580803490274, + "grad_norm": 0.16576212644577026, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0368, + "step": 21240 + }, + { + "epoch": 1.9314670059989094, + "grad_norm": 0.15385235846042633, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0325, + "step": 21250 + }, + { + "epoch": 1.9323759316487912, + "grad_norm": 0.1463586539030075, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0329, + "step": 21260 + }, + { + "epoch": 1.933284857298673, + "grad_norm": 0.15962658822536469, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0352, + "step": 21270 + }, + { + "epoch": 1.934193782948555, + "grad_norm": 0.19544944167137146, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0343, + "step": 21280 + }, + { + "epoch": 1.9351027085984367, + "grad_norm": 0.19503796100616455, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.038, + "step": 21290 + }, + { + "epoch": 1.9360116342483185, + "grad_norm": 0.16592387855052948, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0366, + "step": 21300 + }, + { + "epoch": 1.9369205598982004, + "grad_norm": 0.17534780502319336, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0403, + "step": 21310 + }, + { + "epoch": 1.9378294855480822, + "grad_norm": 0.16248175501823425, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0345, + "step": 21320 + }, + { + "epoch": 1.938738411197964, + "grad_norm": 0.1589919477701187, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0358, + "step": 21330 + }, + { + "epoch": 1.939647336847846, + "grad_norm": 0.17478728294372559, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0391, + "step": 21340 + }, + { + "epoch": 1.9405562624977277, + "grad_norm": 0.15958760678768158, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0359, + "step": 21350 + }, + { + "epoch": 1.9414651881476095, + "grad_norm": 0.17371612787246704, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0394, + "step": 21360 + }, + { + "epoch": 1.9423741137974915, + "grad_norm": 0.2328067272901535, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0353, + "step": 21370 + }, + { + "epoch": 1.9432830394473732, + "grad_norm": 0.14491865038871765, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0353, + "step": 21380 + }, + { + "epoch": 1.944191965097255, + "grad_norm": 0.17347192764282227, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0363, + "step": 21390 + }, + { + "epoch": 1.945100890747137, + "grad_norm": 0.19151349365711212, + "learning_rate": 2.866951776928e-05, + "loss": 0.0356, + "step": 21400 + }, + { + "epoch": 1.9460098163970188, + "grad_norm": 0.12516546249389648, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0367, + "step": 21410 + }, + { + "epoch": 1.9469187420469005, + "grad_norm": 0.17927761375904083, + "learning_rate": 2.866611169167349e-05, + "loss": 0.037, + "step": 21420 + }, + { + "epoch": 1.9478276676967825, + "grad_norm": 0.15957070887088776, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0344, + "step": 21430 + }, + { + "epoch": 1.9487365933466643, + "grad_norm": 0.12543538212776184, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0341, + "step": 21440 + }, + { + "epoch": 1.949645518996546, + "grad_norm": 0.11130441725254059, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0322, + "step": 21450 + }, + { + "epoch": 1.950554444646428, + "grad_norm": 0.14047236740589142, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0345, + "step": 21460 + }, + { + "epoch": 1.9514633702963098, + "grad_norm": 0.1433563083410263, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0373, + "step": 21470 + }, + { + "epoch": 1.9523722959461915, + "grad_norm": 0.14976835250854492, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0332, + "step": 21480 + }, + { + "epoch": 1.9532812215960735, + "grad_norm": 0.16067835688591003, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.037, + "step": 21490 + }, + { + "epoch": 1.9541901472459553, + "grad_norm": 0.19288639724254608, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0354, + "step": 21500 + }, + { + "epoch": 1.955099072895837, + "grad_norm": 0.1743946522474289, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0339, + "step": 21510 + }, + { + "epoch": 1.956007998545719, + "grad_norm": 0.17157430946826935, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0389, + "step": 21520 + }, + { + "epoch": 1.9569169241956008, + "grad_norm": 0.13914166390895844, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0345, + "step": 21530 + }, + { + "epoch": 1.9578258498454826, + "grad_norm": 0.12898056209087372, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0335, + "step": 21540 + }, + { + "epoch": 1.9587347754953646, + "grad_norm": 0.14682252705097198, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0374, + "step": 21550 + }, + { + "epoch": 1.9596437011452463, + "grad_norm": 0.16849438846111298, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0352, + "step": 21560 + }, + { + "epoch": 1.960552626795128, + "grad_norm": 0.20567123591899872, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0354, + "step": 21570 + }, + { + "epoch": 1.96146155244501, + "grad_norm": 0.1588432937860489, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.035, + "step": 21580 + }, + { + "epoch": 1.9623704780948918, + "grad_norm": 0.22252850234508514, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0353, + "step": 21590 + }, + { + "epoch": 1.9632794037447736, + "grad_norm": 0.18251360952854156, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0328, + "step": 21600 + }, + { + "epoch": 1.9641883293946556, + "grad_norm": 0.16760528087615967, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0379, + "step": 21610 + }, + { + "epoch": 1.9650972550445374, + "grad_norm": 0.11655689030885696, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0342, + "step": 21620 + }, + { + "epoch": 1.9660061806944191, + "grad_norm": 0.12147042900323868, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0336, + "step": 21630 + }, + { + "epoch": 1.9669151063443011, + "grad_norm": 0.199283629655838, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0344, + "step": 21640 + }, + { + "epoch": 1.9678240319941829, + "grad_norm": 0.16806209087371826, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0347, + "step": 21650 + }, + { + "epoch": 1.9687329576440646, + "grad_norm": 0.16708964109420776, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0325, + "step": 21660 + }, + { + "epoch": 1.9696418832939466, + "grad_norm": 0.13458232581615448, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0365, + "step": 21670 + }, + { + "epoch": 1.9705508089438284, + "grad_norm": 0.14415410161018372, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0346, + "step": 21680 + }, + { + "epoch": 1.9714597345937102, + "grad_norm": 0.14495527744293213, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0328, + "step": 21690 + }, + { + "epoch": 1.9723686602435921, + "grad_norm": 0.13152088224887848, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0356, + "step": 21700 + }, + { + "epoch": 1.973277585893474, + "grad_norm": 0.15159200131893158, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0338, + "step": 21710 + }, + { + "epoch": 1.9741865115433557, + "grad_norm": 0.21733863651752472, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0381, + "step": 21720 + }, + { + "epoch": 1.9750954371932377, + "grad_norm": 0.19984252750873566, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0368, + "step": 21730 + }, + { + "epoch": 1.9760043628431194, + "grad_norm": 0.171487495303154, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0343, + "step": 21740 + }, + { + "epoch": 1.9769132884930012, + "grad_norm": 0.1703348010778427, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0414, + "step": 21750 + }, + { + "epoch": 1.9778222141428832, + "grad_norm": 0.15242350101470947, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0341, + "step": 21760 + }, + { + "epoch": 1.978731139792765, + "grad_norm": 0.13623583316802979, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0362, + "step": 21770 + }, + { + "epoch": 1.9796400654426467, + "grad_norm": 0.14999905228614807, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0368, + "step": 21780 + }, + { + "epoch": 1.9805489910925287, + "grad_norm": 0.1413697749376297, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0333, + "step": 21790 + }, + { + "epoch": 1.9814579167424105, + "grad_norm": 0.16693642735481262, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.036, + "step": 21800 + }, + { + "epoch": 1.9823668423922922, + "grad_norm": 0.15130840241909027, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0383, + "step": 21810 + }, + { + "epoch": 1.9832757680421742, + "grad_norm": 0.17985931038856506, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0353, + "step": 21820 + }, + { + "epoch": 1.984184693692056, + "grad_norm": 0.12120140343904495, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0374, + "step": 21830 + }, + { + "epoch": 1.9850936193419377, + "grad_norm": 0.16647368669509888, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0364, + "step": 21840 + }, + { + "epoch": 1.9860025449918197, + "grad_norm": 0.19934925436973572, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0358, + "step": 21850 + }, + { + "epoch": 1.9869114706417015, + "grad_norm": 0.17463895678520203, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0366, + "step": 21860 + }, + { + "epoch": 1.9878203962915832, + "grad_norm": 0.19852592051029205, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0362, + "step": 21870 + }, + { + "epoch": 1.9887293219414652, + "grad_norm": 0.19287249445915222, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.033, + "step": 21880 + }, + { + "epoch": 1.989638247591347, + "grad_norm": 0.16515107452869415, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0362, + "step": 21890 + }, + { + "epoch": 1.9905471732412288, + "grad_norm": 0.17242729663848877, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0348, + "step": 21900 + }, + { + "epoch": 1.9914560988911107, + "grad_norm": 0.15482620894908905, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.037, + "step": 21910 + }, + { + "epoch": 1.9923650245409925, + "grad_norm": 0.17318843305110931, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0353, + "step": 21920 + }, + { + "epoch": 1.9932739501908743, + "grad_norm": 0.12899388372898102, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0399, + "step": 21930 + }, + { + "epoch": 1.9941828758407563, + "grad_norm": 0.12032776325941086, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0357, + "step": 21940 + }, + { + "epoch": 1.995091801490638, + "grad_norm": 0.13882209360599518, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0356, + "step": 21950 + }, + { + "epoch": 1.9960007271405198, + "grad_norm": 0.16642652451992035, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0335, + "step": 21960 + }, + { + "epoch": 1.9969096527904018, + "grad_norm": 0.18331089615821838, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0363, + "step": 21970 + }, + { + "epoch": 1.9978185784402835, + "grad_norm": 0.13419640064239502, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0357, + "step": 21980 + }, + { + "epoch": 1.9987275040901653, + "grad_norm": 0.1346755027770996, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0372, + "step": 21990 + }, + { + "epoch": 1.9996364297400473, + "grad_norm": 0.133191779255867, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0378, + "step": 22000 + }, + { + "epoch": 2.0005453553899293, + "grad_norm": 0.21110059320926666, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0351, + "step": 22010 + }, + { + "epoch": 2.001454281039811, + "grad_norm": 0.1401866376399994, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0342, + "step": 22020 + }, + { + "epoch": 2.002363206689693, + "grad_norm": 0.16833296418190002, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0345, + "step": 22030 + }, + { + "epoch": 2.003272132339575, + "grad_norm": 0.1503724306821823, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0343, + "step": 22040 + }, + { + "epoch": 2.0041810579894563, + "grad_norm": 0.1735742837190628, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0355, + "step": 22050 + }, + { + "epoch": 2.0050899836393383, + "grad_norm": 0.16211095452308655, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.034, + "step": 22060 + }, + { + "epoch": 2.0059989092892203, + "grad_norm": 0.1958262324333191, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0339, + "step": 22070 + }, + { + "epoch": 2.006907834939102, + "grad_norm": 0.17298831045627594, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0356, + "step": 22080 + }, + { + "epoch": 2.007816760588984, + "grad_norm": 0.19028714299201965, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0341, + "step": 22090 + }, + { + "epoch": 2.008725686238866, + "grad_norm": 0.2104586809873581, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0343, + "step": 22100 + }, + { + "epoch": 2.0096346118887474, + "grad_norm": 0.12308922410011292, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0393, + "step": 22110 + }, + { + "epoch": 2.0105435375386294, + "grad_norm": 0.1781570166349411, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0339, + "step": 22120 + }, + { + "epoch": 2.0114524631885113, + "grad_norm": 0.138077050447464, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0371, + "step": 22130 + }, + { + "epoch": 2.012361388838393, + "grad_norm": 0.15490984916687012, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0358, + "step": 22140 + }, + { + "epoch": 2.013270314488275, + "grad_norm": 0.1453007310628891, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0345, + "step": 22150 + }, + { + "epoch": 2.014179240138157, + "grad_norm": 0.21000412106513977, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0386, + "step": 22160 + }, + { + "epoch": 2.0150881657880384, + "grad_norm": 0.19581927359104156, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0374, + "step": 22170 + }, + { + "epoch": 2.0159970914379204, + "grad_norm": 0.1527836173772812, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0363, + "step": 22180 + }, + { + "epoch": 2.0169060170878024, + "grad_norm": 0.16688573360443115, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0333, + "step": 22190 + }, + { + "epoch": 2.017814942737684, + "grad_norm": 0.15053318440914154, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0335, + "step": 22200 + }, + { + "epoch": 2.018723868387566, + "grad_norm": 0.13803325593471527, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0368, + "step": 22210 + }, + { + "epoch": 2.019632794037448, + "grad_norm": 0.19816003739833832, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0342, + "step": 22220 + }, + { + "epoch": 2.0205417196873294, + "grad_norm": 0.11883584409952164, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.034, + "step": 22230 + }, + { + "epoch": 2.0214506453372114, + "grad_norm": 0.13014286756515503, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0329, + "step": 22240 + }, + { + "epoch": 2.0223595709870934, + "grad_norm": 0.132943794131279, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0335, + "step": 22250 + }, + { + "epoch": 2.023268496636975, + "grad_norm": 0.15541242063045502, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.034, + "step": 22260 + }, + { + "epoch": 2.024177422286857, + "grad_norm": 0.1626824289560318, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0371, + "step": 22270 + }, + { + "epoch": 2.025086347936739, + "grad_norm": 0.1588718146085739, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0379, + "step": 22280 + }, + { + "epoch": 2.0259952735866205, + "grad_norm": 0.15847913920879364, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0342, + "step": 22290 + }, + { + "epoch": 2.0269041992365024, + "grad_norm": 0.1374506950378418, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0361, + "step": 22300 + }, + { + "epoch": 2.0278131248863844, + "grad_norm": 0.1401481330394745, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0349, + "step": 22310 + }, + { + "epoch": 2.028722050536266, + "grad_norm": 0.162989541888237, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0347, + "step": 22320 + }, + { + "epoch": 2.029630976186148, + "grad_norm": 0.12294286489486694, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0374, + "step": 22330 + }, + { + "epoch": 2.03053990183603, + "grad_norm": 0.16091389954090118, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0365, + "step": 22340 + }, + { + "epoch": 2.0314488274859115, + "grad_norm": 0.14344646036624908, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0372, + "step": 22350 + }, + { + "epoch": 2.0323577531357935, + "grad_norm": 0.1584004908800125, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0348, + "step": 22360 + }, + { + "epoch": 2.0332666787856755, + "grad_norm": 0.13778647780418396, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0378, + "step": 22370 + }, + { + "epoch": 2.034175604435557, + "grad_norm": 0.22977513074874878, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0329, + "step": 22380 + }, + { + "epoch": 2.035084530085439, + "grad_norm": 0.14240632951259613, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0388, + "step": 22390 + }, + { + "epoch": 2.035993455735321, + "grad_norm": 0.1546754091978073, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0345, + "step": 22400 + }, + { + "epoch": 2.0369023813852025, + "grad_norm": 0.1649809628725052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0379, + "step": 22410 + }, + { + "epoch": 2.0378113070350845, + "grad_norm": 0.21715450286865234, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0373, + "step": 22420 + }, + { + "epoch": 2.0387202326849665, + "grad_norm": 0.24871611595153809, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0356, + "step": 22430 + }, + { + "epoch": 2.039629158334848, + "grad_norm": 0.16928130388259888, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0343, + "step": 22440 + }, + { + "epoch": 2.04053808398473, + "grad_norm": 0.14051032066345215, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0363, + "step": 22450 + }, + { + "epoch": 2.041447009634612, + "grad_norm": 0.12724238634109497, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0351, + "step": 22460 + }, + { + "epoch": 2.0423559352844936, + "grad_norm": 0.132507786154747, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0341, + "step": 22470 + }, + { + "epoch": 2.0432648609343755, + "grad_norm": 0.2127864807844162, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0344, + "step": 22480 + }, + { + "epoch": 2.0441737865842575, + "grad_norm": 0.19707432389259338, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.033, + "step": 22490 + }, + { + "epoch": 2.045082712234139, + "grad_norm": 0.1231202557682991, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0353, + "step": 22500 + }, + { + "epoch": 2.045991637884021, + "grad_norm": 0.16637536883354187, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0342, + "step": 22510 + }, + { + "epoch": 2.046900563533903, + "grad_norm": 0.1784791201353073, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0368, + "step": 22520 + }, + { + "epoch": 2.0478094891837846, + "grad_norm": 0.2469760775566101, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0332, + "step": 22530 + }, + { + "epoch": 2.0487184148336666, + "grad_norm": 0.16270138323307037, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0361, + "step": 22540 + }, + { + "epoch": 2.0496273404835486, + "grad_norm": 0.14544764161109924, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0361, + "step": 22550 + }, + { + "epoch": 2.05053626613343, + "grad_norm": 0.17918522655963898, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0326, + "step": 22560 + }, + { + "epoch": 2.051445191783312, + "grad_norm": 0.14032869040966034, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0351, + "step": 22570 + }, + { + "epoch": 2.052354117433194, + "grad_norm": 0.13234272599220276, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.037, + "step": 22580 + }, + { + "epoch": 2.0532630430830756, + "grad_norm": 0.16780512034893036, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0364, + "step": 22590 + }, + { + "epoch": 2.0541719687329576, + "grad_norm": 0.1880781054496765, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0324, + "step": 22600 + }, + { + "epoch": 2.0550808943828396, + "grad_norm": 0.21121688187122345, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0394, + "step": 22610 + }, + { + "epoch": 2.055989820032721, + "grad_norm": 0.18611368536949158, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0361, + "step": 22620 + }, + { + "epoch": 2.056898745682603, + "grad_norm": 0.17370958626270294, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0374, + "step": 22630 + }, + { + "epoch": 2.057807671332485, + "grad_norm": 0.20324639976024628, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0358, + "step": 22640 + }, + { + "epoch": 2.0587165969823666, + "grad_norm": 0.15969744324684143, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0359, + "step": 22650 + }, + { + "epoch": 2.0596255226322486, + "grad_norm": 0.12697312235832214, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.035, + "step": 22660 + }, + { + "epoch": 2.0605344482821306, + "grad_norm": 0.1440528929233551, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0368, + "step": 22670 + }, + { + "epoch": 2.061443373932012, + "grad_norm": 0.1498168408870697, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0365, + "step": 22680 + }, + { + "epoch": 2.062352299581894, + "grad_norm": 0.11599958688020706, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0363, + "step": 22690 + }, + { + "epoch": 2.063261225231776, + "grad_norm": 0.17546436190605164, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0365, + "step": 22700 + }, + { + "epoch": 2.0641701508816577, + "grad_norm": 0.1966724544763565, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0348, + "step": 22710 + }, + { + "epoch": 2.0650790765315397, + "grad_norm": 0.12227731198072433, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0349, + "step": 22720 + }, + { + "epoch": 2.0659880021814216, + "grad_norm": 0.18815574049949646, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0344, + "step": 22730 + }, + { + "epoch": 2.066896927831303, + "grad_norm": 0.1753205955028534, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0354, + "step": 22740 + }, + { + "epoch": 2.067805853481185, + "grad_norm": 0.13510173559188843, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0335, + "step": 22750 + }, + { + "epoch": 2.068714779131067, + "grad_norm": 0.1771705597639084, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0351, + "step": 22760 + }, + { + "epoch": 2.069623704780949, + "grad_norm": 0.17252275347709656, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0334, + "step": 22770 + }, + { + "epoch": 2.0705326304308307, + "grad_norm": 0.17365899682044983, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0356, + "step": 22780 + }, + { + "epoch": 2.0714415560807127, + "grad_norm": 0.19908206164836884, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0342, + "step": 22790 + }, + { + "epoch": 2.072350481730594, + "grad_norm": 0.14257706701755524, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0337, + "step": 22800 + }, + { + "epoch": 2.073259407380476, + "grad_norm": 0.16498896479606628, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0343, + "step": 22810 + }, + { + "epoch": 2.074168333030358, + "grad_norm": 0.1910673975944519, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0333, + "step": 22820 + }, + { + "epoch": 2.07507725868024, + "grad_norm": 0.17894478142261505, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0325, + "step": 22830 + }, + { + "epoch": 2.0759861843301217, + "grad_norm": 0.19651781022548676, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0349, + "step": 22840 + }, + { + "epoch": 2.0768951099800037, + "grad_norm": 0.17259517312049866, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0335, + "step": 22850 + }, + { + "epoch": 2.0778040356298852, + "grad_norm": 0.1826646625995636, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0338, + "step": 22860 + }, + { + "epoch": 2.0787129612797672, + "grad_norm": 0.14523331820964813, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0353, + "step": 22870 + }, + { + "epoch": 2.079621886929649, + "grad_norm": 0.1136784553527832, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0355, + "step": 22880 + }, + { + "epoch": 2.080530812579531, + "grad_norm": 0.2193610519170761, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0349, + "step": 22890 + }, + { + "epoch": 2.0814397382294128, + "grad_norm": 0.21939079463481903, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0348, + "step": 22900 + }, + { + "epoch": 2.0823486638792947, + "grad_norm": 0.15418125689029694, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0354, + "step": 22910 + }, + { + "epoch": 2.0832575895291767, + "grad_norm": 0.12587285041809082, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0356, + "step": 22920 + }, + { + "epoch": 2.0841665151790583, + "grad_norm": 0.3014485239982605, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0342, + "step": 22930 + }, + { + "epoch": 2.0850754408289403, + "grad_norm": 0.19326502084732056, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.036, + "step": 22940 + }, + { + "epoch": 2.0859843664788222, + "grad_norm": 0.18584474921226501, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0329, + "step": 22950 + }, + { + "epoch": 2.086893292128704, + "grad_norm": 0.18337363004684448, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.035, + "step": 22960 + }, + { + "epoch": 2.0878022177785858, + "grad_norm": 0.15845392644405365, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0342, + "step": 22970 + }, + { + "epoch": 2.0887111434284678, + "grad_norm": 0.12274560332298279, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0345, + "step": 22980 + }, + { + "epoch": 2.0896200690783493, + "grad_norm": 0.14729911088943481, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0362, + "step": 22990 + }, + { + "epoch": 2.0905289947282313, + "grad_norm": 0.22667869925498962, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0348, + "step": 23000 + }, + { + "epoch": 2.0914379203781133, + "grad_norm": 0.17735038697719574, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.036, + "step": 23010 + }, + { + "epoch": 2.092346846027995, + "grad_norm": 0.18216878175735474, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0325, + "step": 23020 + }, + { + "epoch": 2.093255771677877, + "grad_norm": 0.15885376930236816, + "learning_rate": 2.83783964791861e-05, + "loss": 0.037, + "step": 23030 + }, + { + "epoch": 2.094164697327759, + "grad_norm": 0.14920319616794586, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0344, + "step": 23040 + }, + { + "epoch": 2.0950736229776403, + "grad_norm": 0.1489100158214569, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0357, + "step": 23050 + }, + { + "epoch": 2.0959825486275223, + "grad_norm": 0.15178553760051727, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0353, + "step": 23060 + }, + { + "epoch": 2.0968914742774043, + "grad_norm": 0.18347544968128204, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0343, + "step": 23070 + }, + { + "epoch": 2.097800399927286, + "grad_norm": 0.14704790711402893, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.036, + "step": 23080 + }, + { + "epoch": 2.098709325577168, + "grad_norm": 0.13624207675457, + "learning_rate": 2.836716156259365e-05, + "loss": 0.034, + "step": 23090 + }, + { + "epoch": 2.09961825122705, + "grad_norm": 0.10985977947711945, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.036, + "step": 23100 + }, + { + "epoch": 2.1005271768769314, + "grad_norm": 0.13247843086719513, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0381, + "step": 23110 + }, + { + "epoch": 2.1014361025268133, + "grad_norm": 0.16457203030586243, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0346, + "step": 23120 + }, + { + "epoch": 2.1023450281766953, + "grad_norm": 0.14583000540733337, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0366, + "step": 23130 + }, + { + "epoch": 2.103253953826577, + "grad_norm": 0.20507733523845673, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0339, + "step": 23140 + }, + { + "epoch": 2.104162879476459, + "grad_norm": 0.19334937632083893, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0367, + "step": 23150 + }, + { + "epoch": 2.105071805126341, + "grad_norm": 0.3277796506881714, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.038, + "step": 23160 + }, + { + "epoch": 2.1059807307762224, + "grad_norm": 0.1911814659833908, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0335, + "step": 23170 + }, + { + "epoch": 2.1068896564261044, + "grad_norm": 0.15557736158370972, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0361, + "step": 23180 + }, + { + "epoch": 2.1077985820759864, + "grad_norm": 0.17675404250621796, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0358, + "step": 23190 + }, + { + "epoch": 2.108707507725868, + "grad_norm": 0.14476734399795532, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0337, + "step": 23200 + }, + { + "epoch": 2.10961643337575, + "grad_norm": 0.16623413562774658, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0318, + "step": 23210 + }, + { + "epoch": 2.110525359025632, + "grad_norm": 0.2075507789850235, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0353, + "step": 23220 + }, + { + "epoch": 2.1114342846755134, + "grad_norm": 0.1533510982990265, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0341, + "step": 23230 + }, + { + "epoch": 2.1123432103253954, + "grad_norm": 0.20513810217380524, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0358, + "step": 23240 + }, + { + "epoch": 2.1132521359752774, + "grad_norm": 0.14066235721111298, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0329, + "step": 23250 + }, + { + "epoch": 2.114161061625159, + "grad_norm": 0.14175069332122803, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0355, + "step": 23260 + }, + { + "epoch": 2.115069987275041, + "grad_norm": 0.16217584908008575, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0369, + "step": 23270 + }, + { + "epoch": 2.115978912924923, + "grad_norm": 0.14876151084899902, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0356, + "step": 23280 + }, + { + "epoch": 2.1168878385748044, + "grad_norm": 0.18083563446998596, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.034, + "step": 23290 + }, + { + "epoch": 2.1177967642246864, + "grad_norm": 0.16635651886463165, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0343, + "step": 23300 + }, + { + "epoch": 2.1187056898745684, + "grad_norm": 0.1964752972126007, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.034, + "step": 23310 + }, + { + "epoch": 2.11961461552445, + "grad_norm": 0.21349099278450012, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0318, + "step": 23320 + }, + { + "epoch": 2.120523541174332, + "grad_norm": 0.18733945488929749, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0341, + "step": 23330 + }, + { + "epoch": 2.121432466824214, + "grad_norm": 0.19528190791606903, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0377, + "step": 23340 + }, + { + "epoch": 2.1223413924740955, + "grad_norm": 0.19702598452568054, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0334, + "step": 23350 + }, + { + "epoch": 2.1232503181239775, + "grad_norm": 0.14029772579669952, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0342, + "step": 23360 + }, + { + "epoch": 2.1241592437738595, + "grad_norm": 0.19629959762096405, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0361, + "step": 23370 + }, + { + "epoch": 2.125068169423741, + "grad_norm": 0.21334053575992584, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0347, + "step": 23380 + }, + { + "epoch": 2.125977095073623, + "grad_norm": 0.18955376744270325, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0382, + "step": 23390 + }, + { + "epoch": 2.126886020723505, + "grad_norm": 0.1557779163122177, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0341, + "step": 23400 + }, + { + "epoch": 2.1277949463733865, + "grad_norm": 0.11952902376651764, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.035, + "step": 23410 + }, + { + "epoch": 2.1287038720232685, + "grad_norm": 0.15596242249011993, + "learning_rate": 2.830471710251995e-05, + "loss": 0.036, + "step": 23420 + }, + { + "epoch": 2.1296127976731505, + "grad_norm": 0.11824432015419006, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0365, + "step": 23430 + }, + { + "epoch": 2.130521723323032, + "grad_norm": 0.16434453427791595, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0334, + "step": 23440 + }, + { + "epoch": 2.131430648972914, + "grad_norm": 0.15980450809001923, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0335, + "step": 23450 + }, + { + "epoch": 2.132339574622796, + "grad_norm": 0.14575710892677307, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0362, + "step": 23460 + }, + { + "epoch": 2.1332485002726775, + "grad_norm": 0.16184058785438538, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0357, + "step": 23470 + }, + { + "epoch": 2.1341574259225595, + "grad_norm": 0.14947448670864105, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0358, + "step": 23480 + }, + { + "epoch": 2.1350663515724415, + "grad_norm": 0.1514957845211029, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0335, + "step": 23490 + }, + { + "epoch": 2.135975277222323, + "grad_norm": 0.14937008917331696, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0354, + "step": 23500 + }, + { + "epoch": 2.136884202872205, + "grad_norm": 0.19262175261974335, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0349, + "step": 23510 + }, + { + "epoch": 2.137793128522087, + "grad_norm": 0.21608002483844757, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0351, + "step": 23520 + }, + { + "epoch": 2.1387020541719686, + "grad_norm": 0.16381889581680298, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0334, + "step": 23530 + }, + { + "epoch": 2.1396109798218506, + "grad_norm": 0.19987569749355316, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0393, + "step": 23540 + }, + { + "epoch": 2.1405199054717325, + "grad_norm": 0.15193146467208862, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0345, + "step": 23550 + }, + { + "epoch": 2.141428831121614, + "grad_norm": 0.2133064568042755, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.035, + "step": 23560 + }, + { + "epoch": 2.142337756771496, + "grad_norm": 0.20385774970054626, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0334, + "step": 23570 + }, + { + "epoch": 2.143246682421378, + "grad_norm": 0.17909514904022217, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0341, + "step": 23580 + }, + { + "epoch": 2.1441556080712596, + "grad_norm": 0.16325385868549347, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0329, + "step": 23590 + }, + { + "epoch": 2.1450645337211416, + "grad_norm": 0.1509324312210083, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0343, + "step": 23600 + }, + { + "epoch": 2.1459734593710236, + "grad_norm": 0.1564115583896637, + "learning_rate": 2.82682647596595e-05, + "loss": 0.033, + "step": 23610 + }, + { + "epoch": 2.146882385020905, + "grad_norm": 0.16515925526618958, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0369, + "step": 23620 + }, + { + "epoch": 2.147791310670787, + "grad_norm": 0.16819676756858826, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0343, + "step": 23630 + }, + { + "epoch": 2.148700236320669, + "grad_norm": 0.19503946602344513, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0331, + "step": 23640 + }, + { + "epoch": 2.1496091619705506, + "grad_norm": 0.1586313098669052, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0361, + "step": 23650 + }, + { + "epoch": 2.1505180876204326, + "grad_norm": 0.13291682302951813, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0395, + "step": 23660 + }, + { + "epoch": 2.1514270132703146, + "grad_norm": 0.215993270277977, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0369, + "step": 23670 + }, + { + "epoch": 2.152335938920196, + "grad_norm": 0.1707286387681961, + "learning_rate": 2.825474320101202e-05, + "loss": 0.033, + "step": 23680 + }, + { + "epoch": 2.153244864570078, + "grad_norm": 0.12792202830314636, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0315, + "step": 23690 + }, + { + "epoch": 2.15415379021996, + "grad_norm": 0.18120065331459045, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0329, + "step": 23700 + }, + { + "epoch": 2.1550627158698417, + "grad_norm": 0.13897357881069183, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0364, + "step": 23710 + }, + { + "epoch": 2.1559716415197236, + "grad_norm": 0.14385102689266205, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0343, + "step": 23720 + }, + { + "epoch": 2.1568805671696056, + "grad_norm": 0.14585982263088226, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0347, + "step": 23730 + }, + { + "epoch": 2.157789492819487, + "grad_norm": 0.16131573915481567, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0342, + "step": 23740 + }, + { + "epoch": 2.158698418469369, + "grad_norm": 0.14613521099090576, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0348, + "step": 23750 + }, + { + "epoch": 2.159607344119251, + "grad_norm": 0.1816590577363968, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0368, + "step": 23760 + }, + { + "epoch": 2.1605162697691327, + "grad_norm": 0.165024533867836, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0348, + "step": 23770 + }, + { + "epoch": 2.1614251954190147, + "grad_norm": 0.12038061022758484, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0339, + "step": 23780 + }, + { + "epoch": 2.1623341210688967, + "grad_norm": 0.18788835406303406, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0351, + "step": 23790 + }, + { + "epoch": 2.163243046718778, + "grad_norm": 0.12421922385692596, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0355, + "step": 23800 + }, + { + "epoch": 2.16415197236866, + "grad_norm": 0.12620848417282104, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0351, + "step": 23810 + }, + { + "epoch": 2.165060898018542, + "grad_norm": 0.1363711655139923, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0348, + "step": 23820 + }, + { + "epoch": 2.1659698236684237, + "grad_norm": 0.15121369063854218, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0346, + "step": 23830 + }, + { + "epoch": 2.1668787493183057, + "grad_norm": 0.17635416984558105, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0334, + "step": 23840 + }, + { + "epoch": 2.1677876749681877, + "grad_norm": 0.16352105140686035, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0376, + "step": 23850 + }, + { + "epoch": 2.1686966006180692, + "grad_norm": 0.1263897866010666, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0337, + "step": 23860 + }, + { + "epoch": 2.1696055262679512, + "grad_norm": 0.1539987325668335, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0352, + "step": 23870 + }, + { + "epoch": 2.170514451917833, + "grad_norm": 0.14957861602306366, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0314, + "step": 23880 + }, + { + "epoch": 2.1714233775677148, + "grad_norm": 0.16712726652622223, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0318, + "step": 23890 + }, + { + "epoch": 2.1723323032175967, + "grad_norm": 0.14287491142749786, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0346, + "step": 23900 + }, + { + "epoch": 2.1732412288674787, + "grad_norm": 0.1685696244239807, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0343, + "step": 23910 + }, + { + "epoch": 2.1741501545173607, + "grad_norm": 0.17902123928070068, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0337, + "step": 23920 + }, + { + "epoch": 2.1750590801672423, + "grad_norm": 0.17377091944217682, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0332, + "step": 23930 + }, + { + "epoch": 2.1759680058171242, + "grad_norm": 0.17717155814170837, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0355, + "step": 23940 + }, + { + "epoch": 2.176876931467006, + "grad_norm": 0.14229369163513184, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0339, + "step": 23950 + }, + { + "epoch": 2.1777858571168878, + "grad_norm": 0.1859610229730606, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0367, + "step": 23960 + }, + { + "epoch": 2.1786947827667698, + "grad_norm": 0.139067605137825, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0381, + "step": 23970 + }, + { + "epoch": 2.1796037084166517, + "grad_norm": 0.14744195342063904, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.0343, + "step": 23980 + }, + { + "epoch": 2.1805126340665333, + "grad_norm": 0.14136101305484772, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0323, + "step": 23990 + }, + { + "epoch": 2.1814215597164153, + "grad_norm": 0.14083971083164215, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0338, + "step": 24000 + }, + { + "epoch": 2.182330485366297, + "grad_norm": 0.17239534854888916, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0354, + "step": 24010 + }, + { + "epoch": 2.183239411016179, + "grad_norm": 0.11582590639591217, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0333, + "step": 24020 + }, + { + "epoch": 2.184148336666061, + "grad_norm": 0.11977744847536087, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0318, + "step": 24030 + }, + { + "epoch": 2.1850572623159428, + "grad_norm": 0.14146319031715393, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0307, + "step": 24040 + }, + { + "epoch": 2.1859661879658243, + "grad_norm": 0.17784899473190308, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0339, + "step": 24050 + }, + { + "epoch": 2.1868751136157063, + "grad_norm": 0.17177951335906982, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0355, + "step": 24060 + }, + { + "epoch": 2.187784039265588, + "grad_norm": 0.1111745685338974, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0318, + "step": 24070 + }, + { + "epoch": 2.18869296491547, + "grad_norm": 0.11921021342277527, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0331, + "step": 24080 + }, + { + "epoch": 2.189601890565352, + "grad_norm": 0.17614205181598663, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0323, + "step": 24090 + }, + { + "epoch": 2.190510816215234, + "grad_norm": 0.2279019057750702, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0364, + "step": 24100 + }, + { + "epoch": 2.1914197418651153, + "grad_norm": 0.17467322945594788, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0354, + "step": 24110 + }, + { + "epoch": 2.1923286675149973, + "grad_norm": 0.14741715788841248, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0329, + "step": 24120 + }, + { + "epoch": 2.1932375931648793, + "grad_norm": 0.1672716587781906, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0338, + "step": 24130 + }, + { + "epoch": 2.194146518814761, + "grad_norm": 0.16663245856761932, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0323, + "step": 24140 + }, + { + "epoch": 2.195055444464643, + "grad_norm": 0.15424536168575287, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0315, + "step": 24150 + }, + { + "epoch": 2.195964370114525, + "grad_norm": 0.1372607946395874, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.0358, + "step": 24160 + }, + { + "epoch": 2.1968732957644064, + "grad_norm": 0.18870234489440918, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0362, + "step": 24170 + }, + { + "epoch": 2.1977822214142884, + "grad_norm": 0.13517074286937714, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0375, + "step": 24180 + }, + { + "epoch": 2.1986911470641703, + "grad_norm": 0.17887406051158905, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0314, + "step": 24190 + }, + { + "epoch": 2.199600072714052, + "grad_norm": 0.20612122118473053, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0371, + "step": 24200 + }, + { + "epoch": 2.200508998363934, + "grad_norm": 0.2151893675327301, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0331, + "step": 24210 + }, + { + "epoch": 2.201417924013816, + "grad_norm": 0.1654006689786911, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0346, + "step": 24220 + }, + { + "epoch": 2.2023268496636974, + "grad_norm": 0.16367658972740173, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0329, + "step": 24230 + }, + { + "epoch": 2.2032357753135794, + "grad_norm": 0.1794711798429489, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0317, + "step": 24240 + }, + { + "epoch": 2.2041447009634614, + "grad_norm": 0.1752845197916031, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0342, + "step": 24250 + }, + { + "epoch": 2.205053626613343, + "grad_norm": 0.15124720335006714, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0337, + "step": 24260 + }, + { + "epoch": 2.205962552263225, + "grad_norm": 0.13884073495864868, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0323, + "step": 24270 + }, + { + "epoch": 2.206871477913107, + "grad_norm": 0.12932395935058594, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0339, + "step": 24280 + }, + { + "epoch": 2.2077804035629884, + "grad_norm": 0.14618000388145447, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0362, + "step": 24290 + }, + { + "epoch": 2.2086893292128704, + "grad_norm": 0.15860095620155334, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0377, + "step": 24300 + }, + { + "epoch": 2.2095982548627524, + "grad_norm": 0.15432479977607727, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0352, + "step": 24310 + }, + { + "epoch": 2.210507180512634, + "grad_norm": 0.1345023214817047, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.0346, + "step": 24320 + }, + { + "epoch": 2.211416106162516, + "grad_norm": 0.17498451471328735, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0308, + "step": 24330 + }, + { + "epoch": 2.212325031812398, + "grad_norm": 0.1339508444070816, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0351, + "step": 24340 + }, + { + "epoch": 2.2132339574622795, + "grad_norm": 0.13514141738414764, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0341, + "step": 24350 + }, + { + "epoch": 2.2141428831121615, + "grad_norm": 0.14570850133895874, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0373, + "step": 24360 + }, + { + "epoch": 2.2150518087620434, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0322, + "step": 24370 + }, + { + "epoch": 2.215960734411925, + "grad_norm": 0.12058911472558975, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.032, + "step": 24380 + }, + { + "epoch": 2.216869660061807, + "grad_norm": 0.1782122254371643, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0342, + "step": 24390 + }, + { + "epoch": 2.217778585711689, + "grad_norm": 0.1807096004486084, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0357, + "step": 24400 + }, + { + "epoch": 2.2186875113615705, + "grad_norm": 0.2152816504240036, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.036, + "step": 24410 + }, + { + "epoch": 2.2195964370114525, + "grad_norm": 0.14051620662212372, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0326, + "step": 24420 + }, + { + "epoch": 2.2205053626613345, + "grad_norm": 0.17378471791744232, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0347, + "step": 24430 + }, + { + "epoch": 2.221414288311216, + "grad_norm": 0.14577393233776093, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0367, + "step": 24440 + }, + { + "epoch": 2.222323213961098, + "grad_norm": 0.14644794166088104, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0365, + "step": 24450 + }, + { + "epoch": 2.22323213961098, + "grad_norm": 0.15704382956027985, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0323, + "step": 24460 + }, + { + "epoch": 2.2241410652608615, + "grad_norm": 0.19725443422794342, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0317, + "step": 24470 + }, + { + "epoch": 2.2250499909107435, + "grad_norm": 0.1381104588508606, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0351, + "step": 24480 + }, + { + "epoch": 2.2259589165606255, + "grad_norm": 0.18695984780788422, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0313, + "step": 24490 + }, + { + "epoch": 2.226867842210507, + "grad_norm": 0.14611804485321045, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0336, + "step": 24500 + }, + { + "epoch": 2.227776767860389, + "grad_norm": 0.15197867155075073, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0332, + "step": 24510 + }, + { + "epoch": 2.228685693510271, + "grad_norm": 0.18415333330631256, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0335, + "step": 24520 + }, + { + "epoch": 2.2295946191601526, + "grad_norm": 0.17294619977474213, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0345, + "step": 24530 + }, + { + "epoch": 2.2305035448100345, + "grad_norm": 0.12429935485124588, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0343, + "step": 24540 + }, + { + "epoch": 2.2314124704599165, + "grad_norm": 0.11881612241268158, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0332, + "step": 24550 + }, + { + "epoch": 2.232321396109798, + "grad_norm": 0.14975786209106445, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0342, + "step": 24560 + }, + { + "epoch": 2.23323032175968, + "grad_norm": 0.16111701726913452, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0332, + "step": 24570 + }, + { + "epoch": 2.234139247409562, + "grad_norm": 0.17143002152442932, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0349, + "step": 24580 + }, + { + "epoch": 2.2350481730594436, + "grad_norm": 0.1182711198925972, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0328, + "step": 24590 + }, + { + "epoch": 2.2359570987093256, + "grad_norm": 0.13149790465831757, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.034, + "step": 24600 + }, + { + "epoch": 2.2368660243592076, + "grad_norm": 0.16568699479103088, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0334, + "step": 24610 + }, + { + "epoch": 2.237774950009089, + "grad_norm": 0.20225627720355988, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0335, + "step": 24620 + }, + { + "epoch": 2.238683875658971, + "grad_norm": 0.19510960578918457, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0334, + "step": 24630 + }, + { + "epoch": 2.239592801308853, + "grad_norm": 0.17292174696922302, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0321, + "step": 24640 + }, + { + "epoch": 2.2405017269587346, + "grad_norm": 0.20075462758541107, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0312, + "step": 24650 + }, + { + "epoch": 2.2414106526086166, + "grad_norm": 0.22460708022117615, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.035, + "step": 24660 + }, + { + "epoch": 2.2423195782584986, + "grad_norm": 0.16554118692874908, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0317, + "step": 24670 + }, + { + "epoch": 2.24322850390838, + "grad_norm": 0.17846789956092834, + "learning_rate": 2.805621740412736e-05, + "loss": 0.033, + "step": 24680 + }, + { + "epoch": 2.244137429558262, + "grad_norm": 0.16171950101852417, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0337, + "step": 24690 + }, + { + "epoch": 2.245046355208144, + "grad_norm": 0.12096307426691055, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0338, + "step": 24700 + }, + { + "epoch": 2.2459552808580256, + "grad_norm": 0.1484556943178177, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0331, + "step": 24710 + }, + { + "epoch": 2.2468642065079076, + "grad_norm": 0.1466083824634552, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0307, + "step": 24720 + }, + { + "epoch": 2.2477731321577896, + "grad_norm": 0.1873260736465454, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0321, + "step": 24730 + }, + { + "epoch": 2.248682057807671, + "grad_norm": 0.1290832757949829, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0348, + "step": 24740 + }, + { + "epoch": 2.249590983457553, + "grad_norm": 0.13401605188846588, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0309, + "step": 24750 + }, + { + "epoch": 2.250499909107435, + "grad_norm": 0.1482284516096115, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0328, + "step": 24760 + }, + { + "epoch": 2.2514088347573167, + "grad_norm": 0.1930164396762848, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0348, + "step": 24770 + }, + { + "epoch": 2.2523177604071987, + "grad_norm": 0.12778505682945251, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0324, + "step": 24780 + }, + { + "epoch": 2.2532266860570807, + "grad_norm": 0.14444303512573242, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0357, + "step": 24790 + }, + { + "epoch": 2.254135611706962, + "grad_norm": 0.12906020879745483, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0328, + "step": 24800 + }, + { + "epoch": 2.255044537356844, + "grad_norm": 0.1524304747581482, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0323, + "step": 24810 + }, + { + "epoch": 2.255953463006726, + "grad_norm": 0.13719342648983002, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0346, + "step": 24820 + }, + { + "epoch": 2.2568623886566077, + "grad_norm": 0.17767463624477386, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0352, + "step": 24830 + }, + { + "epoch": 2.2577713143064897, + "grad_norm": 0.14223742485046387, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0328, + "step": 24840 + }, + { + "epoch": 2.2586802399563717, + "grad_norm": 0.14129288494586945, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0339, + "step": 24850 + }, + { + "epoch": 2.2595891656062532, + "grad_norm": 0.14904361963272095, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0354, + "step": 24860 + }, + { + "epoch": 2.260498091256135, + "grad_norm": 0.13365837931632996, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0331, + "step": 24870 + }, + { + "epoch": 2.261407016906017, + "grad_norm": 0.19269660115242004, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0355, + "step": 24880 + }, + { + "epoch": 2.2623159425558987, + "grad_norm": 0.13244526088237762, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0337, + "step": 24890 + }, + { + "epoch": 2.2632248682057807, + "grad_norm": 0.11207720637321472, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0317, + "step": 24900 + }, + { + "epoch": 2.2641337938556627, + "grad_norm": 0.23352758586406708, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0378, + "step": 24910 + }, + { + "epoch": 2.2650427195055443, + "grad_norm": 0.17066404223442078, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0343, + "step": 24920 + }, + { + "epoch": 2.2659516451554262, + "grad_norm": 0.14155130088329315, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0345, + "step": 24930 + }, + { + "epoch": 2.2668605708053082, + "grad_norm": 0.1658909171819687, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0367, + "step": 24940 + }, + { + "epoch": 2.26776949645519, + "grad_norm": 0.14387215673923492, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0306, + "step": 24950 + }, + { + "epoch": 2.2686784221050718, + "grad_norm": 0.13560348749160767, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0363, + "step": 24960 + }, + { + "epoch": 2.2695873477549537, + "grad_norm": 0.14641359448432922, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0341, + "step": 24970 + }, + { + "epoch": 2.2704962734048353, + "grad_norm": 0.13989782333374023, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0339, + "step": 24980 + }, + { + "epoch": 2.2714051990547173, + "grad_norm": 0.11538433283567429, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0322, + "step": 24990 + }, + { + "epoch": 2.2723141247045993, + "grad_norm": 0.1490464210510254, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0368, + "step": 25000 + }, + { + "epoch": 2.2732230503544812, + "grad_norm": 0.16454437375068665, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0336, + "step": 25010 + }, + { + "epoch": 2.274131976004363, + "grad_norm": 0.21664154529571533, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0348, + "step": 25020 + }, + { + "epoch": 2.2750409016542448, + "grad_norm": 0.1526898443698883, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0339, + "step": 25030 + }, + { + "epoch": 2.2759498273041263, + "grad_norm": 0.13281138241291046, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0347, + "step": 25040 + }, + { + "epoch": 2.2768587529540083, + "grad_norm": 0.13153065741062164, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0318, + "step": 25050 + }, + { + "epoch": 2.2777676786038903, + "grad_norm": 0.1307174116373062, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0389, + "step": 25060 + }, + { + "epoch": 2.2786766042537723, + "grad_norm": 0.14495964348316193, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0348, + "step": 25070 + }, + { + "epoch": 2.279585529903654, + "grad_norm": 0.20408621430397034, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0354, + "step": 25080 + }, + { + "epoch": 2.280494455553536, + "grad_norm": 0.12892967462539673, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0341, + "step": 25090 + }, + { + "epoch": 2.2814033812034173, + "grad_norm": 0.15494148433208466, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0325, + "step": 25100 + }, + { + "epoch": 2.2823123068532993, + "grad_norm": 0.1286173313856125, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0364, + "step": 25110 + }, + { + "epoch": 2.2832212325031813, + "grad_norm": 0.20810382068157196, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0308, + "step": 25120 + }, + { + "epoch": 2.2841301581530633, + "grad_norm": 0.1756974160671234, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0353, + "step": 25130 + }, + { + "epoch": 2.285039083802945, + "grad_norm": 0.15091076493263245, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0322, + "step": 25140 + }, + { + "epoch": 2.285948009452827, + "grad_norm": 0.14837786555290222, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0323, + "step": 25150 + }, + { + "epoch": 2.2868569351027084, + "grad_norm": 0.16038651764392853, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0331, + "step": 25160 + }, + { + "epoch": 2.2877658607525904, + "grad_norm": 0.11152328550815582, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0355, + "step": 25170 + }, + { + "epoch": 2.2886747864024723, + "grad_norm": 0.2487601488828659, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0361, + "step": 25180 + }, + { + "epoch": 2.2895837120523543, + "grad_norm": 0.2027425765991211, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0326, + "step": 25190 + }, + { + "epoch": 2.290492637702236, + "grad_norm": 0.18161940574645996, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0338, + "step": 25200 + }, + { + "epoch": 2.291401563352118, + "grad_norm": 0.1356310248374939, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0365, + "step": 25210 + }, + { + "epoch": 2.2923104890019994, + "grad_norm": 0.19586680829524994, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0323, + "step": 25220 + }, + { + "epoch": 2.2932194146518814, + "grad_norm": 0.13667763769626617, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0326, + "step": 25230 + }, + { + "epoch": 2.2941283403017634, + "grad_norm": 0.19646073877811432, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0324, + "step": 25240 + }, + { + "epoch": 2.2950372659516454, + "grad_norm": 0.1827101707458496, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0361, + "step": 25250 + }, + { + "epoch": 2.295946191601527, + "grad_norm": 0.22372350096702576, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0325, + "step": 25260 + }, + { + "epoch": 2.296855117251409, + "grad_norm": 0.1569242775440216, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0336, + "step": 25270 + }, + { + "epoch": 2.2977640429012904, + "grad_norm": 0.15961501002311707, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0331, + "step": 25280 + }, + { + "epoch": 2.2986729685511724, + "grad_norm": 0.1434725821018219, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0356, + "step": 25290 + }, + { + "epoch": 2.2995818942010544, + "grad_norm": 0.16500534117221832, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0317, + "step": 25300 + }, + { + "epoch": 2.3004908198509364, + "grad_norm": 0.17559613287448883, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0335, + "step": 25310 + }, + { + "epoch": 2.301399745500818, + "grad_norm": 0.1354442685842514, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0325, + "step": 25320 + }, + { + "epoch": 2.3023086711507, + "grad_norm": 0.16235969960689545, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0338, + "step": 25330 + }, + { + "epoch": 2.3032175968005815, + "grad_norm": 0.17332574725151062, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0367, + "step": 25340 + }, + { + "epoch": 2.3041265224504635, + "grad_norm": 0.12563961744308472, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0325, + "step": 25350 + }, + { + "epoch": 2.3050354481003454, + "grad_norm": 0.12744230031967163, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0321, + "step": 25360 + }, + { + "epoch": 2.3059443737502274, + "grad_norm": 0.17766962945461273, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0349, + "step": 25370 + }, + { + "epoch": 2.306853299400109, + "grad_norm": 0.13534484803676605, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0338, + "step": 25380 + }, + { + "epoch": 2.307762225049991, + "grad_norm": 0.14469285309314728, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0336, + "step": 25390 + }, + { + "epoch": 2.308671150699873, + "grad_norm": 0.1499573439359665, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0307, + "step": 25400 + }, + { + "epoch": 2.3095800763497545, + "grad_norm": 0.21517696976661682, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0374, + "step": 25410 + }, + { + "epoch": 2.3104890019996365, + "grad_norm": 0.20611676573753357, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0374, + "step": 25420 + }, + { + "epoch": 2.3113979276495185, + "grad_norm": 0.15239006280899048, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0313, + "step": 25430 + }, + { + "epoch": 2.3123068532994, + "grad_norm": 0.17100869119167328, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0314, + "step": 25440 + }, + { + "epoch": 2.313215778949282, + "grad_norm": 0.12796221673488617, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0356, + "step": 25450 + }, + { + "epoch": 2.314124704599164, + "grad_norm": 0.12295611202716827, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0319, + "step": 25460 + }, + { + "epoch": 2.3150336302490455, + "grad_norm": 0.15505139529705048, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0355, + "step": 25470 + }, + { + "epoch": 2.3159425558989275, + "grad_norm": 0.15435868501663208, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0354, + "step": 25480 + }, + { + "epoch": 2.3168514815488095, + "grad_norm": 0.14466767013072968, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0327, + "step": 25490 + }, + { + "epoch": 2.317760407198691, + "grad_norm": 0.14676661789417267, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0347, + "step": 25500 + }, + { + "epoch": 2.318669332848573, + "grad_norm": 0.2019292563199997, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.033, + "step": 25510 + }, + { + "epoch": 2.319578258498455, + "grad_norm": 0.21594174206256866, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0338, + "step": 25520 + }, + { + "epoch": 2.3204871841483365, + "grad_norm": 0.19752857089042664, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0338, + "step": 25530 + }, + { + "epoch": 2.3213961097982185, + "grad_norm": 0.22113604843616486, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0338, + "step": 25540 + }, + { + "epoch": 2.3223050354481005, + "grad_norm": 0.11263782531023026, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0334, + "step": 25550 + }, + { + "epoch": 2.323213961097982, + "grad_norm": 0.14741253852844238, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0333, + "step": 25560 + }, + { + "epoch": 2.324122886747864, + "grad_norm": 0.1414068043231964, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0354, + "step": 25570 + }, + { + "epoch": 2.325031812397746, + "grad_norm": 0.1873990148305893, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0318, + "step": 25580 + }, + { + "epoch": 2.3259407380476276, + "grad_norm": 0.14797523617744446, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0353, + "step": 25590 + }, + { + "epoch": 2.3268496636975096, + "grad_norm": 0.16598115861415863, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0326, + "step": 25600 + }, + { + "epoch": 2.3277585893473915, + "grad_norm": 0.1547769159078598, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0332, + "step": 25610 + }, + { + "epoch": 2.328667514997273, + "grad_norm": 0.16734863817691803, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0322, + "step": 25620 + }, + { + "epoch": 2.329576440647155, + "grad_norm": 0.18915362656116486, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0351, + "step": 25630 + }, + { + "epoch": 2.330485366297037, + "grad_norm": 0.18157050013542175, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0319, + "step": 25640 + }, + { + "epoch": 2.3313942919469186, + "grad_norm": 0.15408003330230713, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0345, + "step": 25650 + }, + { + "epoch": 2.3323032175968006, + "grad_norm": 0.15781241655349731, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0287, + "step": 25660 + }, + { + "epoch": 2.3332121432466826, + "grad_norm": 0.17689508199691772, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0381, + "step": 25670 + }, + { + "epoch": 2.334121068896564, + "grad_norm": 0.16804984211921692, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0322, + "step": 25680 + }, + { + "epoch": 2.335029994546446, + "grad_norm": 0.15865635871887207, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0323, + "step": 25690 + }, + { + "epoch": 2.335938920196328, + "grad_norm": 0.11461278796195984, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0347, + "step": 25700 + }, + { + "epoch": 2.3368478458462096, + "grad_norm": 0.18055297434329987, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0321, + "step": 25710 + }, + { + "epoch": 2.3377567714960916, + "grad_norm": 0.15852698683738708, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0338, + "step": 25720 + }, + { + "epoch": 2.3386656971459736, + "grad_norm": 0.12436269968748093, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0318, + "step": 25730 + }, + { + "epoch": 2.339574622795855, + "grad_norm": 0.13437382876873016, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0344, + "step": 25740 + }, + { + "epoch": 2.340483548445737, + "grad_norm": 0.12187877297401428, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0319, + "step": 25750 + }, + { + "epoch": 2.341392474095619, + "grad_norm": 0.172430157661438, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0323, + "step": 25760 + }, + { + "epoch": 2.3423013997455007, + "grad_norm": 0.13950073719024658, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0337, + "step": 25770 + }, + { + "epoch": 2.3432103253953827, + "grad_norm": 0.20705333352088928, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0329, + "step": 25780 + }, + { + "epoch": 2.3441192510452646, + "grad_norm": 0.158171609044075, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0362, + "step": 25790 + }, + { + "epoch": 2.345028176695146, + "grad_norm": 0.1404315084218979, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0325, + "step": 25800 + }, + { + "epoch": 2.345937102345028, + "grad_norm": 0.1824568808078766, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0322, + "step": 25810 + }, + { + "epoch": 2.34684602799491, + "grad_norm": 0.17829887568950653, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0332, + "step": 25820 + }, + { + "epoch": 2.3477549536447917, + "grad_norm": 0.15108005702495575, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0299, + "step": 25830 + }, + { + "epoch": 2.3486638792946737, + "grad_norm": 0.12933197617530823, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0348, + "step": 25840 + }, + { + "epoch": 2.3495728049445557, + "grad_norm": 0.14609909057617188, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0316, + "step": 25850 + }, + { + "epoch": 2.350481730594437, + "grad_norm": 0.16911636292934418, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0345, + "step": 25860 + }, + { + "epoch": 2.351390656244319, + "grad_norm": 0.19833095371723175, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0387, + "step": 25870 + }, + { + "epoch": 2.352299581894201, + "grad_norm": 0.1749381721019745, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0332, + "step": 25880 + }, + { + "epoch": 2.3532085075440827, + "grad_norm": 0.1642262190580368, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0333, + "step": 25890 + }, + { + "epoch": 2.3541174331939647, + "grad_norm": 0.18784858286380768, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0357, + "step": 25900 + }, + { + "epoch": 2.3550263588438467, + "grad_norm": 0.18121536076068878, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0318, + "step": 25910 + }, + { + "epoch": 2.3559352844937282, + "grad_norm": 0.1629398912191391, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0302, + "step": 25920 + }, + { + "epoch": 2.3568442101436102, + "grad_norm": 0.17479686439037323, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0337, + "step": 25930 + }, + { + "epoch": 2.357753135793492, + "grad_norm": 0.13278532028198242, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0317, + "step": 25940 + }, + { + "epoch": 2.3586620614433738, + "grad_norm": 0.1546788215637207, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0326, + "step": 25950 + }, + { + "epoch": 2.3595709870932557, + "grad_norm": 0.2420499175786972, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0316, + "step": 25960 + }, + { + "epoch": 2.3604799127431377, + "grad_norm": 0.12131790071725845, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0325, + "step": 25970 + }, + { + "epoch": 2.3613888383930193, + "grad_norm": 0.1788730025291443, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0372, + "step": 25980 + }, + { + "epoch": 2.3622977640429013, + "grad_norm": 0.10886150598526001, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0337, + "step": 25990 + }, + { + "epoch": 2.3632066896927832, + "grad_norm": 0.18961258232593536, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0329, + "step": 26000 + }, + { + "epoch": 2.364115615342665, + "grad_norm": 0.15669813752174377, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0324, + "step": 26010 + }, + { + "epoch": 2.3650245409925468, + "grad_norm": 0.14186154305934906, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0334, + "step": 26020 + }, + { + "epoch": 2.3659334666424288, + "grad_norm": 0.12576361000537872, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0311, + "step": 26030 + }, + { + "epoch": 2.3668423922923103, + "grad_norm": 0.17558914422988892, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.033, + "step": 26040 + }, + { + "epoch": 2.3677513179421923, + "grad_norm": 0.15792331099510193, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0316, + "step": 26050 + }, + { + "epoch": 2.3686602435920743, + "grad_norm": 0.16193841397762299, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.0374, + "step": 26060 + }, + { + "epoch": 2.369569169241956, + "grad_norm": 0.12543265521526337, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0319, + "step": 26070 + }, + { + "epoch": 2.370478094891838, + "grad_norm": 0.17739617824554443, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0321, + "step": 26080 + }, + { + "epoch": 2.37138702054172, + "grad_norm": 0.14417538046836853, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0325, + "step": 26090 + }, + { + "epoch": 2.3722959461916013, + "grad_norm": 0.14808978140354156, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0316, + "step": 26100 + }, + { + "epoch": 2.3732048718414833, + "grad_norm": 0.1371920108795166, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0351, + "step": 26110 + }, + { + "epoch": 2.3741137974913653, + "grad_norm": 0.15588054060935974, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.036, + "step": 26120 + }, + { + "epoch": 2.375022723141247, + "grad_norm": 0.145936518907547, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0328, + "step": 26130 + }, + { + "epoch": 2.375931648791129, + "grad_norm": 0.1652826964855194, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0319, + "step": 26140 + }, + { + "epoch": 2.376840574441011, + "grad_norm": 0.16869927942752838, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0313, + "step": 26150 + }, + { + "epoch": 2.377749500090893, + "grad_norm": 0.13716866075992584, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0332, + "step": 26160 + }, + { + "epoch": 2.3786584257407744, + "grad_norm": 0.17480172216892242, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0346, + "step": 26170 + }, + { + "epoch": 2.3795673513906563, + "grad_norm": 0.1411801129579544, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0331, + "step": 26180 + }, + { + "epoch": 2.380476277040538, + "grad_norm": 0.13307742774486542, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0313, + "step": 26190 + }, + { + "epoch": 2.38138520269042, + "grad_norm": 0.14322148263454437, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.038, + "step": 26200 + }, + { + "epoch": 2.382294128340302, + "grad_norm": 0.22250258922576904, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0337, + "step": 26210 + }, + { + "epoch": 2.383203053990184, + "grad_norm": 0.1645289957523346, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0337, + "step": 26220 + }, + { + "epoch": 2.3841119796400654, + "grad_norm": 0.11326144635677338, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0321, + "step": 26230 + }, + { + "epoch": 2.3850209052899474, + "grad_norm": 0.1590202897787094, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0335, + "step": 26240 + }, + { + "epoch": 2.385929830939829, + "grad_norm": 0.18294939398765564, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0325, + "step": 26250 + }, + { + "epoch": 2.386838756589711, + "grad_norm": 0.13846619427204132, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0365, + "step": 26260 + }, + { + "epoch": 2.387747682239593, + "grad_norm": 0.18725250661373138, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.033, + "step": 26270 + }, + { + "epoch": 2.388656607889475, + "grad_norm": 0.12946827709674835, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.034, + "step": 26280 + }, + { + "epoch": 2.3895655335393564, + "grad_norm": 0.1472131460905075, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0334, + "step": 26290 + }, + { + "epoch": 2.3904744591892384, + "grad_norm": 0.13550294935703278, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.031, + "step": 26300 + }, + { + "epoch": 2.39138338483912, + "grad_norm": 0.14005765318870544, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0358, + "step": 26310 + }, + { + "epoch": 2.392292310489002, + "grad_norm": 0.1775454878807068, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.033, + "step": 26320 + }, + { + "epoch": 2.393201236138884, + "grad_norm": 0.15063601732254028, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0325, + "step": 26330 + }, + { + "epoch": 2.394110161788766, + "grad_norm": 0.1624850332736969, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0371, + "step": 26340 + }, + { + "epoch": 2.3950190874386474, + "grad_norm": 0.14907152950763702, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.033, + "step": 26350 + }, + { + "epoch": 2.3959280130885294, + "grad_norm": 0.16778139770030975, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0305, + "step": 26360 + }, + { + "epoch": 2.396836938738411, + "grad_norm": 0.18856392800807953, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0331, + "step": 26370 + }, + { + "epoch": 2.397745864388293, + "grad_norm": 0.16435189545154572, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.034, + "step": 26380 + }, + { + "epoch": 2.398654790038175, + "grad_norm": 0.19101689755916595, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0354, + "step": 26390 + }, + { + "epoch": 2.399563715688057, + "grad_norm": 0.12628066539764404, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0339, + "step": 26400 + }, + { + "epoch": 2.4004726413379385, + "grad_norm": 0.18366508185863495, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0334, + "step": 26410 + }, + { + "epoch": 2.4013815669878205, + "grad_norm": 0.14487522840499878, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0323, + "step": 26420 + }, + { + "epoch": 2.402290492637702, + "grad_norm": 0.17358407378196716, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0334, + "step": 26430 + }, + { + "epoch": 2.403199418287584, + "grad_norm": 0.17748914659023285, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0349, + "step": 26440 + }, + { + "epoch": 2.404108343937466, + "grad_norm": 0.13347984850406647, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0317, + "step": 26450 + }, + { + "epoch": 2.405017269587348, + "grad_norm": 0.16281896829605103, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0322, + "step": 26460 + }, + { + "epoch": 2.4059261952372295, + "grad_norm": 0.14611801505088806, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.032, + "step": 26470 + }, + { + "epoch": 2.4068351208871115, + "grad_norm": 0.15810316801071167, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0336, + "step": 26480 + }, + { + "epoch": 2.407744046536993, + "grad_norm": 0.139622762799263, + "learning_rate": 2.767178055427033e-05, + "loss": 0.033, + "step": 26490 + }, + { + "epoch": 2.408652972186875, + "grad_norm": 0.133967325091362, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0348, + "step": 26500 + }, + { + "epoch": 2.409561897836757, + "grad_norm": 0.17810052633285522, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0363, + "step": 26510 + }, + { + "epoch": 2.410470823486639, + "grad_norm": 0.20053236186504364, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0318, + "step": 26520 + }, + { + "epoch": 2.4113797491365205, + "grad_norm": 0.18349161744117737, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0331, + "step": 26530 + }, + { + "epoch": 2.4122886747864025, + "grad_norm": 0.14123651385307312, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0317, + "step": 26540 + }, + { + "epoch": 2.413197600436284, + "grad_norm": 0.23438133299350739, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.0345, + "step": 26550 + }, + { + "epoch": 2.414106526086166, + "grad_norm": 0.14156576991081238, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0315, + "step": 26560 + }, + { + "epoch": 2.415015451736048, + "grad_norm": 0.15003304183483124, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0375, + "step": 26570 + }, + { + "epoch": 2.41592437738593, + "grad_norm": 0.1592501550912857, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0368, + "step": 26580 + }, + { + "epoch": 2.4168333030358116, + "grad_norm": 0.19662776589393616, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0319, + "step": 26590 + }, + { + "epoch": 2.4177422286856936, + "grad_norm": 0.1755184531211853, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0369, + "step": 26600 + }, + { + "epoch": 2.4186511543355755, + "grad_norm": 0.18557681143283844, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0319, + "step": 26610 + }, + { + "epoch": 2.419560079985457, + "grad_norm": 0.16313788294792175, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0329, + "step": 26620 + }, + { + "epoch": 2.420469005635339, + "grad_norm": 0.17769283056259155, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0353, + "step": 26630 + }, + { + "epoch": 2.421377931285221, + "grad_norm": 0.20588253438472748, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.036, + "step": 26640 + }, + { + "epoch": 2.4222868569351026, + "grad_norm": 0.19055846333503723, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0325, + "step": 26650 + }, + { + "epoch": 2.4231957825849846, + "grad_norm": 0.1475432813167572, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0327, + "step": 26660 + }, + { + "epoch": 2.4241047082348666, + "grad_norm": 0.19939357042312622, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0335, + "step": 26670 + }, + { + "epoch": 2.425013633884748, + "grad_norm": 0.147215336561203, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0326, + "step": 26680 + }, + { + "epoch": 2.42592255953463, + "grad_norm": 0.14440524578094482, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0347, + "step": 26690 + }, + { + "epoch": 2.426831485184512, + "grad_norm": 0.13242407143115997, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0324, + "step": 26700 + }, + { + "epoch": 2.4277404108343936, + "grad_norm": 0.16356824338436127, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0324, + "step": 26710 + }, + { + "epoch": 2.4286493364842756, + "grad_norm": 0.15098844468593597, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0324, + "step": 26720 + }, + { + "epoch": 2.4295582621341576, + "grad_norm": 0.19436295330524445, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0354, + "step": 26730 + }, + { + "epoch": 2.430467187784039, + "grad_norm": 0.16093489527702332, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0299, + "step": 26740 + }, + { + "epoch": 2.431376113433921, + "grad_norm": 0.15131326019763947, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0317, + "step": 26750 + }, + { + "epoch": 2.432285039083803, + "grad_norm": 0.17794589698314667, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0342, + "step": 26760 + }, + { + "epoch": 2.4331939647336847, + "grad_norm": 0.14589503407478333, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0337, + "step": 26770 + }, + { + "epoch": 2.4341028903835666, + "grad_norm": 0.17109166085720062, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0323, + "step": 26780 + }, + { + "epoch": 2.4350118160334486, + "grad_norm": 0.15193383395671844, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0341, + "step": 26790 + }, + { + "epoch": 2.43592074168333, + "grad_norm": 0.16468095779418945, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0358, + "step": 26800 + }, + { + "epoch": 2.436829667333212, + "grad_norm": 0.19232936203479767, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0323, + "step": 26810 + }, + { + "epoch": 2.437738592983094, + "grad_norm": 0.20423170924186707, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0309, + "step": 26820 + }, + { + "epoch": 2.4386475186329757, + "grad_norm": 0.17231513559818268, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0329, + "step": 26830 + }, + { + "epoch": 2.4395564442828577, + "grad_norm": 0.20656634867191315, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0313, + "step": 26840 + }, + { + "epoch": 2.4404653699327397, + "grad_norm": 0.20403745770454407, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0334, + "step": 26850 + }, + { + "epoch": 2.441374295582621, + "grad_norm": 0.10177117586135864, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0349, + "step": 26860 + }, + { + "epoch": 2.442283221232503, + "grad_norm": 0.1855831891298294, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0336, + "step": 26870 + }, + { + "epoch": 2.443192146882385, + "grad_norm": 0.19658896327018738, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0312, + "step": 26880 + }, + { + "epoch": 2.4441010725322667, + "grad_norm": 0.1642596274614334, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0311, + "step": 26890 + }, + { + "epoch": 2.4450099981821487, + "grad_norm": 0.12628965079784393, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0344, + "step": 26900 + }, + { + "epoch": 2.4459189238320307, + "grad_norm": 0.16898013651371002, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0322, + "step": 26910 + }, + { + "epoch": 2.4468278494819122, + "grad_norm": 0.12113088369369507, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0306, + "step": 26920 + }, + { + "epoch": 2.447736775131794, + "grad_norm": 0.17149172723293304, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0333, + "step": 26930 + }, + { + "epoch": 2.448645700781676, + "grad_norm": 0.1793266236782074, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.032, + "step": 26940 + }, + { + "epoch": 2.4495546264315577, + "grad_norm": 0.17687232792377472, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0316, + "step": 26950 + }, + { + "epoch": 2.4504635520814397, + "grad_norm": 0.15994255244731903, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0326, + "step": 26960 + }, + { + "epoch": 2.4513724777313217, + "grad_norm": 0.12573835253715515, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0298, + "step": 26970 + }, + { + "epoch": 2.4522814033812033, + "grad_norm": 0.14864389598369598, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0327, + "step": 26980 + }, + { + "epoch": 2.4531903290310852, + "grad_norm": 0.14193812012672424, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0319, + "step": 26990 + }, + { + "epoch": 2.4540992546809672, + "grad_norm": 0.14076431095600128, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0326, + "step": 27000 + }, + { + "epoch": 2.4550081803308488, + "grad_norm": 0.15489982068538666, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0312, + "step": 27010 + }, + { + "epoch": 2.4559171059807308, + "grad_norm": 0.15466968715190887, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.033, + "step": 27020 + }, + { + "epoch": 2.4568260316306127, + "grad_norm": 0.14559857547283173, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0336, + "step": 27030 + }, + { + "epoch": 2.4577349572804943, + "grad_norm": 0.19880737364292145, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0337, + "step": 27040 + }, + { + "epoch": 2.4586438829303763, + "grad_norm": 0.18202580511569977, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0311, + "step": 27050 + }, + { + "epoch": 2.4595528085802583, + "grad_norm": 0.14301614463329315, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0345, + "step": 27060 + }, + { + "epoch": 2.46046173423014, + "grad_norm": 0.1252974420785904, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0342, + "step": 27070 + }, + { + "epoch": 2.461370659880022, + "grad_norm": 0.13793779909610748, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0335, + "step": 27080 + }, + { + "epoch": 2.462279585529904, + "grad_norm": 0.15571148693561554, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0326, + "step": 27090 + }, + { + "epoch": 2.4631885111797853, + "grad_norm": 0.18771469593048096, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0297, + "step": 27100 + }, + { + "epoch": 2.4640974368296673, + "grad_norm": 0.1843256801366806, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0308, + "step": 27110 + }, + { + "epoch": 2.4650063624795493, + "grad_norm": 0.17226742208003998, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0317, + "step": 27120 + }, + { + "epoch": 2.465915288129431, + "grad_norm": 0.1687847226858139, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0343, + "step": 27130 + }, + { + "epoch": 2.466824213779313, + "grad_norm": 0.15288695693016052, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0305, + "step": 27140 + }, + { + "epoch": 2.467733139429195, + "grad_norm": 0.14716671407222748, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0323, + "step": 27150 + }, + { + "epoch": 2.4686420650790764, + "grad_norm": 0.14054331183433533, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0304, + "step": 27160 + }, + { + "epoch": 2.4695509907289583, + "grad_norm": 0.14266155660152435, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0303, + "step": 27170 + }, + { + "epoch": 2.4704599163788403, + "grad_norm": 0.16496002674102783, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0326, + "step": 27180 + }, + { + "epoch": 2.471368842028722, + "grad_norm": 0.15299643576145172, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0324, + "step": 27190 + }, + { + "epoch": 2.472277767678604, + "grad_norm": 0.13618402183055878, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.033, + "step": 27200 + }, + { + "epoch": 2.473186693328486, + "grad_norm": 0.2055921107530594, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0344, + "step": 27210 + }, + { + "epoch": 2.4740956189783674, + "grad_norm": 0.149324432015419, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0305, + "step": 27220 + }, + { + "epoch": 2.4750045446282494, + "grad_norm": 0.16091959178447723, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0292, + "step": 27230 + }, + { + "epoch": 2.4759134702781314, + "grad_norm": 0.18193399906158447, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0309, + "step": 27240 + }, + { + "epoch": 2.476822395928013, + "grad_norm": 0.15368357300758362, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0347, + "step": 27250 + }, + { + "epoch": 2.477731321577895, + "grad_norm": 0.13040907680988312, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0327, + "step": 27260 + }, + { + "epoch": 2.478640247227777, + "grad_norm": 0.1395898163318634, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0302, + "step": 27270 + }, + { + "epoch": 2.4795491728776584, + "grad_norm": 0.1501321643590927, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0315, + "step": 27280 + }, + { + "epoch": 2.4804580985275404, + "grad_norm": 0.14476990699768066, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0329, + "step": 27290 + }, + { + "epoch": 2.4813670241774224, + "grad_norm": 0.16348396241664886, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0343, + "step": 27300 + }, + { + "epoch": 2.4822759498273044, + "grad_norm": 0.12994851171970367, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0319, + "step": 27310 + }, + { + "epoch": 2.483184875477186, + "grad_norm": 0.17474451661109924, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0318, + "step": 27320 + }, + { + "epoch": 2.484093801127068, + "grad_norm": 0.1324346512556076, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0315, + "step": 27330 + }, + { + "epoch": 2.4850027267769494, + "grad_norm": 0.11593184620141983, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0331, + "step": 27340 + }, + { + "epoch": 2.4859116524268314, + "grad_norm": 0.1315821260213852, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0302, + "step": 27350 + }, + { + "epoch": 2.4868205780767134, + "grad_norm": 0.1476755440235138, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0307, + "step": 27360 + }, + { + "epoch": 2.4877295037265954, + "grad_norm": 0.14154019951820374, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0292, + "step": 27370 + }, + { + "epoch": 2.488638429376477, + "grad_norm": 0.19213183224201202, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0325, + "step": 27380 + }, + { + "epoch": 2.489547355026359, + "grad_norm": 0.1910087615251541, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0322, + "step": 27390 + }, + { + "epoch": 2.4904562806762405, + "grad_norm": 0.15041813254356384, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0327, + "step": 27400 + }, + { + "epoch": 2.4913652063261225, + "grad_norm": 0.17691348493099213, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.032, + "step": 27410 + }, + { + "epoch": 2.4922741319760044, + "grad_norm": 0.14823873341083527, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0312, + "step": 27420 + }, + { + "epoch": 2.4931830576258864, + "grad_norm": 0.15979892015457153, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0332, + "step": 27430 + }, + { + "epoch": 2.494091983275768, + "grad_norm": 0.15497387945652008, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.034, + "step": 27440 + }, + { + "epoch": 2.49500090892565, + "grad_norm": 0.1492585688829422, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0356, + "step": 27450 + }, + { + "epoch": 2.4959098345755315, + "grad_norm": 0.15691104531288147, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0331, + "step": 27460 + }, + { + "epoch": 2.4968187602254135, + "grad_norm": 0.16026780009269714, + "learning_rate": 2.745041482537885e-05, + "loss": 0.031, + "step": 27470 + }, + { + "epoch": 2.4977276858752955, + "grad_norm": 0.1662590205669403, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0316, + "step": 27480 + }, + { + "epoch": 2.4986366115251775, + "grad_norm": 0.14840897917747498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0312, + "step": 27490 + }, + { + "epoch": 2.499545537175059, + "grad_norm": 0.1851847618818283, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0299, + "step": 27500 + }, + { + "epoch": 2.500454462824941, + "grad_norm": 0.19740021228790283, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.033, + "step": 27510 + }, + { + "epoch": 2.5013633884748225, + "grad_norm": 0.15178567171096802, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0333, + "step": 27520 + }, + { + "epoch": 2.5022723141247045, + "grad_norm": 0.190213143825531, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0318, + "step": 27530 + }, + { + "epoch": 2.5031812397745865, + "grad_norm": 0.19656972587108612, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0337, + "step": 27540 + }, + { + "epoch": 2.5040901654244685, + "grad_norm": 0.16388234496116638, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0333, + "step": 27550 + }, + { + "epoch": 2.50499909107435, + "grad_norm": 0.13188935816287994, + "learning_rate": 2.742962757296782e-05, + "loss": 0.0345, + "step": 27560 + }, + { + "epoch": 2.505908016724232, + "grad_norm": 0.22707797586917877, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0304, + "step": 27570 + }, + { + "epoch": 2.5068169423741136, + "grad_norm": 0.15608231723308563, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0304, + "step": 27580 + }, + { + "epoch": 2.5077258680239956, + "grad_norm": 0.14437256753444672, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0314, + "step": 27590 + }, + { + "epoch": 2.5086347936738775, + "grad_norm": 0.16778786480426788, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0313, + "step": 27600 + }, + { + "epoch": 2.5095437193237595, + "grad_norm": 0.20312818884849548, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0332, + "step": 27610 + }, + { + "epoch": 2.510452644973641, + "grad_norm": 0.1498051881790161, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0331, + "step": 27620 + }, + { + "epoch": 2.511361570623523, + "grad_norm": 0.17630966007709503, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0341, + "step": 27630 + }, + { + "epoch": 2.5122704962734046, + "grad_norm": 0.19833719730377197, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0319, + "step": 27640 + }, + { + "epoch": 2.5131794219232866, + "grad_norm": 0.14097176492214203, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0337, + "step": 27650 + }, + { + "epoch": 2.5140883475731686, + "grad_norm": 0.14631834626197815, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0334, + "step": 27660 + }, + { + "epoch": 2.5149972732230506, + "grad_norm": 0.12588630616664886, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0329, + "step": 27670 + }, + { + "epoch": 2.515906198872932, + "grad_norm": 0.1339040994644165, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0337, + "step": 27680 + }, + { + "epoch": 2.516815124522814, + "grad_norm": 0.13559606671333313, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0334, + "step": 27690 + }, + { + "epoch": 2.5177240501726956, + "grad_norm": 0.16942362487316132, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0318, + "step": 27700 + }, + { + "epoch": 2.5186329758225776, + "grad_norm": 0.20392154157161713, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0358, + "step": 27710 + }, + { + "epoch": 2.5195419014724596, + "grad_norm": 0.18161971867084503, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0366, + "step": 27720 + }, + { + "epoch": 2.5204508271223416, + "grad_norm": 0.16783392429351807, + "learning_rate": 2.739015425415693e-05, + "loss": 0.034, + "step": 27730 + }, + { + "epoch": 2.521359752772223, + "grad_norm": 0.18349863588809967, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0331, + "step": 27740 + }, + { + "epoch": 2.522268678422105, + "grad_norm": 0.15218691527843475, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0374, + "step": 27750 + }, + { + "epoch": 2.5231776040719867, + "grad_norm": 0.16958382725715637, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0311, + "step": 27760 + }, + { + "epoch": 2.5240865297218686, + "grad_norm": 0.13147015869617462, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0297, + "step": 27770 + }, + { + "epoch": 2.5249954553717506, + "grad_norm": 0.13189253211021423, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0344, + "step": 27780 + }, + { + "epoch": 2.5259043810216326, + "grad_norm": 0.13725551962852478, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0326, + "step": 27790 + }, + { + "epoch": 2.526813306671514, + "grad_norm": 0.14106380939483643, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0358, + "step": 27800 + }, + { + "epoch": 2.527722232321396, + "grad_norm": 0.1209939569234848, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0315, + "step": 27810 + }, + { + "epoch": 2.5286311579712777, + "grad_norm": 0.15641506016254425, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0342, + "step": 27820 + }, + { + "epoch": 2.5295400836211597, + "grad_norm": 0.1287931501865387, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0314, + "step": 27830 + }, + { + "epoch": 2.5304490092710417, + "grad_norm": 0.1690453290939331, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0292, + "step": 27840 + }, + { + "epoch": 2.5313579349209236, + "grad_norm": 0.14175349473953247, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0321, + "step": 27850 + }, + { + "epoch": 2.532266860570805, + "grad_norm": 0.17326878011226654, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0304, + "step": 27860 + }, + { + "epoch": 2.533175786220687, + "grad_norm": 0.17744505405426025, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0325, + "step": 27870 + }, + { + "epoch": 2.5340847118705687, + "grad_norm": 0.15505000948905945, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0332, + "step": 27880 + }, + { + "epoch": 2.5349936375204507, + "grad_norm": 0.15451034903526306, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0324, + "step": 27890 + }, + { + "epoch": 2.5359025631703327, + "grad_norm": 0.16743066906929016, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0339, + "step": 27900 + }, + { + "epoch": 2.5368114888202147, + "grad_norm": 0.13954298198223114, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0348, + "step": 27910 + }, + { + "epoch": 2.537720414470096, + "grad_norm": 0.1632082462310791, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0369, + "step": 27920 + }, + { + "epoch": 2.538629340119978, + "grad_norm": 0.13393829762935638, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0299, + "step": 27930 + }, + { + "epoch": 2.5395382657698597, + "grad_norm": 0.13054826855659485, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0305, + "step": 27940 + }, + { + "epoch": 2.5404471914197417, + "grad_norm": 0.15264561772346497, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0322, + "step": 27950 + }, + { + "epoch": 2.5413561170696237, + "grad_norm": 0.1493452489376068, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0354, + "step": 27960 + }, + { + "epoch": 2.5422650427195057, + "grad_norm": 0.19372697174549103, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0308, + "step": 27970 + }, + { + "epoch": 2.5431739683693873, + "grad_norm": 0.15836451947689056, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0305, + "step": 27980 + }, + { + "epoch": 2.5440828940192692, + "grad_norm": 0.11787498742341995, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0336, + "step": 27990 + }, + { + "epoch": 2.5449918196691512, + "grad_norm": 0.1428709328174591, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0329, + "step": 28000 + }, + { + "epoch": 2.5459007453190328, + "grad_norm": 0.17337153851985931, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0336, + "step": 28010 + }, + { + "epoch": 2.5468096709689148, + "grad_norm": 0.16365543007850647, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0324, + "step": 28020 + }, + { + "epoch": 2.5477185966187967, + "grad_norm": 0.12393038719892502, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0301, + "step": 28030 + }, + { + "epoch": 2.5486275222686783, + "grad_norm": 0.16883614659309387, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.034, + "step": 28040 + }, + { + "epoch": 2.5495364479185603, + "grad_norm": 0.2201436161994934, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0341, + "step": 28050 + }, + { + "epoch": 2.5504453735684423, + "grad_norm": 0.1296076625585556, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0355, + "step": 28060 + }, + { + "epoch": 2.551354299218324, + "grad_norm": 0.13386565446853638, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0294, + "step": 28070 + }, + { + "epoch": 2.552263224868206, + "grad_norm": 0.17629146575927734, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0341, + "step": 28080 + }, + { + "epoch": 2.5531721505180878, + "grad_norm": 0.1630316823720932, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0351, + "step": 28090 + }, + { + "epoch": 2.5540810761679693, + "grad_norm": 0.14526061713695526, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0306, + "step": 28100 + }, + { + "epoch": 2.5549900018178513, + "grad_norm": 0.1635238081216812, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0336, + "step": 28110 + }, + { + "epoch": 2.5558989274677333, + "grad_norm": 0.18598434329032898, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0323, + "step": 28120 + }, + { + "epoch": 2.556807853117615, + "grad_norm": 0.1723354160785675, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0321, + "step": 28130 + }, + { + "epoch": 2.557716778767497, + "grad_norm": 0.14925561845302582, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0328, + "step": 28140 + }, + { + "epoch": 2.558625704417379, + "grad_norm": 0.1375485211610794, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0324, + "step": 28150 + }, + { + "epoch": 2.5595346300672603, + "grad_norm": 0.16682077944278717, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0328, + "step": 28160 + }, + { + "epoch": 2.5604435557171423, + "grad_norm": 0.12893377244472504, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0327, + "step": 28170 + }, + { + "epoch": 2.5613524813670243, + "grad_norm": 0.1455254703760147, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0328, + "step": 28180 + }, + { + "epoch": 2.562261407016906, + "grad_norm": 0.1590094417333603, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0309, + "step": 28190 + }, + { + "epoch": 2.563170332666788, + "grad_norm": 0.12314071506261826, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0288, + "step": 28200 + }, + { + "epoch": 2.56407925831667, + "grad_norm": 0.12654146552085876, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0298, + "step": 28210 + }, + { + "epoch": 2.5649881839665514, + "grad_norm": 0.15538500249385834, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0309, + "step": 28220 + }, + { + "epoch": 2.5658971096164334, + "grad_norm": 0.18536637723445892, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0289, + "step": 28230 + }, + { + "epoch": 2.5668060352663153, + "grad_norm": 0.16231849789619446, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0322, + "step": 28240 + }, + { + "epoch": 2.567714960916197, + "grad_norm": 0.18383623659610748, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0339, + "step": 28250 + }, + { + "epoch": 2.568623886566079, + "grad_norm": 0.14533472061157227, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0357, + "step": 28260 + }, + { + "epoch": 2.569532812215961, + "grad_norm": 0.1522463709115982, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0349, + "step": 28270 + }, + { + "epoch": 2.570441737865843, + "grad_norm": 0.15043863654136658, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0309, + "step": 28280 + }, + { + "epoch": 2.5713506635157244, + "grad_norm": 0.2114235907793045, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0317, + "step": 28290 + }, + { + "epoch": 2.5722595891656064, + "grad_norm": 0.16916878521442413, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0299, + "step": 28300 + }, + { + "epoch": 2.573168514815488, + "grad_norm": 0.13089819252490997, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0344, + "step": 28310 + }, + { + "epoch": 2.57407744046537, + "grad_norm": 0.16062401235103607, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0315, + "step": 28320 + }, + { + "epoch": 2.574986366115252, + "grad_norm": 0.14596132934093475, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0317, + "step": 28330 + }, + { + "epoch": 2.575895291765134, + "grad_norm": 0.11819732934236526, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0316, + "step": 28340 + }, + { + "epoch": 2.5768042174150154, + "grad_norm": 0.14181244373321533, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0319, + "step": 28350 + }, + { + "epoch": 2.5777131430648974, + "grad_norm": 0.16020874679088593, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0332, + "step": 28360 + }, + { + "epoch": 2.578622068714779, + "grad_norm": 0.1741514801979065, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0307, + "step": 28370 + }, + { + "epoch": 2.579530994364661, + "grad_norm": 0.14541201293468475, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0317, + "step": 28380 + }, + { + "epoch": 2.580439920014543, + "grad_norm": 0.13735759258270264, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0338, + "step": 28390 + }, + { + "epoch": 2.581348845664425, + "grad_norm": 0.14192919433116913, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0302, + "step": 28400 + }, + { + "epoch": 2.5822577713143064, + "grad_norm": 0.11899036914110184, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0309, + "step": 28410 + }, + { + "epoch": 2.5831666969641884, + "grad_norm": 0.1424725204706192, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0276, + "step": 28420 + }, + { + "epoch": 2.58407562261407, + "grad_norm": 0.1467887908220291, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0303, + "step": 28430 + }, + { + "epoch": 2.584984548263952, + "grad_norm": 0.18879210948944092, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0324, + "step": 28440 + }, + { + "epoch": 2.585893473913834, + "grad_norm": 0.14174194633960724, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0295, + "step": 28450 + }, + { + "epoch": 2.586802399563716, + "grad_norm": 0.1584935039281845, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0302, + "step": 28460 + }, + { + "epoch": 2.5877113252135975, + "grad_norm": 0.1384730488061905, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0298, + "step": 28470 + }, + { + "epoch": 2.5886202508634795, + "grad_norm": 0.17897197604179382, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.033, + "step": 28480 + }, + { + "epoch": 2.589529176513361, + "grad_norm": 0.14458610117435455, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0306, + "step": 28490 + }, + { + "epoch": 2.590438102163243, + "grad_norm": 0.17092257738113403, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0349, + "step": 28500 + }, + { + "epoch": 2.591347027813125, + "grad_norm": 0.17378823459148407, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0336, + "step": 28510 + }, + { + "epoch": 2.592255953463007, + "grad_norm": 0.14339832961559296, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0307, + "step": 28520 + }, + { + "epoch": 2.5931648791128885, + "grad_norm": 0.15657833218574524, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0307, + "step": 28530 + }, + { + "epoch": 2.5940738047627705, + "grad_norm": 0.20132626593112946, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0329, + "step": 28540 + }, + { + "epoch": 2.594982730412652, + "grad_norm": 0.18215787410736084, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0277, + "step": 28550 + }, + { + "epoch": 2.595891656062534, + "grad_norm": 0.14266397058963776, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0307, + "step": 28560 + }, + { + "epoch": 2.596800581712416, + "grad_norm": 0.14305613934993744, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0327, + "step": 28570 + }, + { + "epoch": 2.597709507362298, + "grad_norm": 0.1090536043047905, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0344, + "step": 28580 + }, + { + "epoch": 2.5986184330121795, + "grad_norm": 0.14907072484493256, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0312, + "step": 28590 + }, + { + "epoch": 2.5995273586620615, + "grad_norm": 0.14900802075862885, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0315, + "step": 28600 + }, + { + "epoch": 2.600436284311943, + "grad_norm": 0.14915376901626587, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0303, + "step": 28610 + }, + { + "epoch": 2.601345209961825, + "grad_norm": 0.1710962951183319, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0323, + "step": 28620 + }, + { + "epoch": 2.602254135611707, + "grad_norm": 0.1605297327041626, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0308, + "step": 28630 + }, + { + "epoch": 2.603163061261589, + "grad_norm": 0.21248827874660492, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0326, + "step": 28640 + }, + { + "epoch": 2.6040719869114706, + "grad_norm": 0.2563180923461914, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0311, + "step": 28650 + }, + { + "epoch": 2.6049809125613526, + "grad_norm": 0.17104025185108185, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0314, + "step": 28660 + }, + { + "epoch": 2.605889838211234, + "grad_norm": 0.146572545170784, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0326, + "step": 28670 + }, + { + "epoch": 2.606798763861116, + "grad_norm": 0.1429024338722229, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0294, + "step": 28680 + }, + { + "epoch": 2.607707689510998, + "grad_norm": 0.173868790268898, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0335, + "step": 28690 + }, + { + "epoch": 2.60861661516088, + "grad_norm": 0.15156859159469604, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0308, + "step": 28700 + }, + { + "epoch": 2.6095255408107616, + "grad_norm": 0.13047891855239868, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0317, + "step": 28710 + }, + { + "epoch": 2.6104344664606436, + "grad_norm": 0.13153310120105743, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0337, + "step": 28720 + }, + { + "epoch": 2.611343392110525, + "grad_norm": 0.178600013256073, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.031, + "step": 28730 + }, + { + "epoch": 2.612252317760407, + "grad_norm": 0.1607854962348938, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0301, + "step": 28740 + }, + { + "epoch": 2.613161243410289, + "grad_norm": 0.21229690313339233, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0323, + "step": 28750 + }, + { + "epoch": 2.614070169060171, + "grad_norm": 0.22143439948558807, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0337, + "step": 28760 + }, + { + "epoch": 2.6149790947100526, + "grad_norm": 0.1913987696170807, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0318, + "step": 28770 + }, + { + "epoch": 2.6158880203599346, + "grad_norm": 0.21662089228630066, + "learning_rate": 2.714035528039113e-05, + "loss": 0.034, + "step": 28780 + }, + { + "epoch": 2.616796946009816, + "grad_norm": 0.16957996785640717, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.031, + "step": 28790 + }, + { + "epoch": 2.617705871659698, + "grad_norm": 0.22695434093475342, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0299, + "step": 28800 + }, + { + "epoch": 2.61861479730958, + "grad_norm": 0.1979096382856369, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0338, + "step": 28810 + }, + { + "epoch": 2.619523722959462, + "grad_norm": 0.19463883340358734, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0322, + "step": 28820 + }, + { + "epoch": 2.6204326486093437, + "grad_norm": 0.16699108481407166, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0325, + "step": 28830 + }, + { + "epoch": 2.6213415742592256, + "grad_norm": 0.15734542906284332, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0318, + "step": 28840 + }, + { + "epoch": 2.622250499909107, + "grad_norm": 0.13387058675289154, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.033, + "step": 28850 + }, + { + "epoch": 2.623159425558989, + "grad_norm": 0.1359146237373352, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0317, + "step": 28860 + }, + { + "epoch": 2.624068351208871, + "grad_norm": 0.15266868472099304, + "learning_rate": 2.71184679679729e-05, + "loss": 0.036, + "step": 28870 + }, + { + "epoch": 2.624977276858753, + "grad_norm": 0.1471046358346939, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0346, + "step": 28880 + }, + { + "epoch": 2.6258862025086347, + "grad_norm": 0.14974592626094818, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0311, + "step": 28890 + }, + { + "epoch": 2.6267951281585167, + "grad_norm": 0.14572028815746307, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0318, + "step": 28900 + }, + { + "epoch": 2.627704053808398, + "grad_norm": 0.13511954247951508, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0339, + "step": 28910 + }, + { + "epoch": 2.62861297945828, + "grad_norm": 0.19174571335315704, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0319, + "step": 28920 + }, + { + "epoch": 2.629521905108162, + "grad_norm": 0.11193472892045975, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0319, + "step": 28930 + }, + { + "epoch": 2.630430830758044, + "grad_norm": 0.18421944975852966, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0328, + "step": 28940 + }, + { + "epoch": 2.6313397564079257, + "grad_norm": 0.1561426818370819, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0316, + "step": 28950 + }, + { + "epoch": 2.6322486820578077, + "grad_norm": 0.1457977145910263, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0341, + "step": 28960 + }, + { + "epoch": 2.6331576077076893, + "grad_norm": 0.13319474458694458, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0311, + "step": 28970 + }, + { + "epoch": 2.6340665333575712, + "grad_norm": 0.1713014841079712, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0329, + "step": 28980 + }, + { + "epoch": 2.6349754590074532, + "grad_norm": 0.2044817954301834, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0306, + "step": 28990 + }, + { + "epoch": 2.635884384657335, + "grad_norm": 0.14710377156734467, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0319, + "step": 29000 + }, + { + "epoch": 2.6367933103072168, + "grad_norm": 0.12927214801311493, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0311, + "step": 29010 + }, + { + "epoch": 2.6377022359570987, + "grad_norm": 0.1363520622253418, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0331, + "step": 29020 + }, + { + "epoch": 2.6386111616069803, + "grad_norm": 0.13269765675067902, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0304, + "step": 29030 + }, + { + "epoch": 2.6395200872568623, + "grad_norm": 0.15315485000610352, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0326, + "step": 29040 + }, + { + "epoch": 2.6404290129067443, + "grad_norm": 0.17547115683555603, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0328, + "step": 29050 + }, + { + "epoch": 2.6413379385566262, + "grad_norm": 0.137929305434227, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0321, + "step": 29060 + }, + { + "epoch": 2.642246864206508, + "grad_norm": 0.1490202099084854, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0313, + "step": 29070 + }, + { + "epoch": 2.6431557898563898, + "grad_norm": 0.12684305012226105, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0339, + "step": 29080 + }, + { + "epoch": 2.6440647155062713, + "grad_norm": 0.11788252741098404, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0316, + "step": 29090 + }, + { + "epoch": 2.6449736411561533, + "grad_norm": 0.12164406478404999, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0349, + "step": 29100 + }, + { + "epoch": 2.6458825668060353, + "grad_norm": 0.153250053524971, + "learning_rate": 2.705973772304842e-05, + "loss": 0.0319, + "step": 29110 + }, + { + "epoch": 2.6467914924559173, + "grad_norm": 0.1762060523033142, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0334, + "step": 29120 + }, + { + "epoch": 2.647700418105799, + "grad_norm": 0.11491841822862625, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0337, + "step": 29130 + }, + { + "epoch": 2.648609343755681, + "grad_norm": 0.14058469235897064, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0348, + "step": 29140 + }, + { + "epoch": 2.649518269405563, + "grad_norm": 0.1553429812192917, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0353, + "step": 29150 + }, + { + "epoch": 2.6504271950554443, + "grad_norm": 0.12161046266555786, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0321, + "step": 29160 + }, + { + "epoch": 2.6513361207053263, + "grad_norm": 0.1392403542995453, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0311, + "step": 29170 + }, + { + "epoch": 2.6522450463552083, + "grad_norm": 0.14240571856498718, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0326, + "step": 29180 + }, + { + "epoch": 2.65315397200509, + "grad_norm": 0.13554590940475464, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0302, + "step": 29190 + }, + { + "epoch": 2.654062897654972, + "grad_norm": 0.12577129900455475, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0286, + "step": 29200 + }, + { + "epoch": 2.654971823304854, + "grad_norm": 0.19563496112823486, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0329, + "step": 29210 + }, + { + "epoch": 2.6558807489547354, + "grad_norm": 0.16490554809570312, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0297, + "step": 29220 + }, + { + "epoch": 2.6567896746046173, + "grad_norm": 0.1568729132413864, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0327, + "step": 29230 + }, + { + "epoch": 2.6576986002544993, + "grad_norm": 0.14630043506622314, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0356, + "step": 29240 + }, + { + "epoch": 2.658607525904381, + "grad_norm": 0.1410026252269745, + "learning_rate": 2.702523463056731e-05, + "loss": 0.03, + "step": 29250 + }, + { + "epoch": 2.659516451554263, + "grad_norm": 0.20436449348926544, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0342, + "step": 29260 + }, + { + "epoch": 2.660425377204145, + "grad_norm": 0.13378705084323883, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0316, + "step": 29270 + }, + { + "epoch": 2.6613343028540264, + "grad_norm": 0.14612115919589996, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0319, + "step": 29280 + }, + { + "epoch": 2.6622432285039084, + "grad_norm": 0.1284344345331192, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0325, + "step": 29290 + }, + { + "epoch": 2.6631521541537904, + "grad_norm": 0.12650351226329803, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0322, + "step": 29300 + }, + { + "epoch": 2.664061079803672, + "grad_norm": 0.1807493418455124, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0295, + "step": 29310 + }, + { + "epoch": 2.664970005453554, + "grad_norm": 0.14492884278297424, + "learning_rate": 2.700791592927203e-05, + "loss": 0.034, + "step": 29320 + }, + { + "epoch": 2.665878931103436, + "grad_norm": 0.12300306558609009, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0308, + "step": 29330 + }, + { + "epoch": 2.6667878567533174, + "grad_norm": 0.14803357422351837, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0333, + "step": 29340 + }, + { + "epoch": 2.6676967824031994, + "grad_norm": 0.16484776139259338, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0333, + "step": 29350 + }, + { + "epoch": 2.6686057080530814, + "grad_norm": 0.1561662256717682, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0321, + "step": 29360 + }, + { + "epoch": 2.669514633702963, + "grad_norm": 0.13944213092327118, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0304, + "step": 29370 + }, + { + "epoch": 2.670423559352845, + "grad_norm": 0.12666095793247223, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0293, + "step": 29380 + }, + { + "epoch": 2.671332485002727, + "grad_norm": 0.15350337326526642, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0323, + "step": 29390 + }, + { + "epoch": 2.6722414106526085, + "grad_norm": 0.1701808124780655, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0306, + "step": 29400 + }, + { + "epoch": 2.6731503363024904, + "grad_norm": 0.19572556018829346, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.03, + "step": 29410 + }, + { + "epoch": 2.6740592619523724, + "grad_norm": 0.1769261211156845, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0345, + "step": 29420 + }, + { + "epoch": 2.6749681876022544, + "grad_norm": 0.14372074604034424, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0375, + "step": 29430 + }, + { + "epoch": 2.675877113252136, + "grad_norm": 0.13742008805274963, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0287, + "step": 29440 + }, + { + "epoch": 2.676786038902018, + "grad_norm": 0.20102860033512115, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0297, + "step": 29450 + }, + { + "epoch": 2.6776949645518995, + "grad_norm": 0.24746869504451752, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0319, + "step": 29460 + }, + { + "epoch": 2.6786038902017815, + "grad_norm": 0.13874007761478424, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0309, + "step": 29470 + }, + { + "epoch": 2.6795128158516635, + "grad_norm": 0.2117520421743393, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0309, + "step": 29480 + }, + { + "epoch": 2.6804217415015454, + "grad_norm": 0.15622322261333466, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0343, + "step": 29490 + }, + { + "epoch": 2.681330667151427, + "grad_norm": 0.17840346693992615, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0322, + "step": 29500 + }, + { + "epoch": 2.682239592801309, + "grad_norm": 0.25253826379776, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.0303, + "step": 29510 + }, + { + "epoch": 2.6831485184511905, + "grad_norm": 0.14672136306762695, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0312, + "step": 29520 + }, + { + "epoch": 2.6840574441010725, + "grad_norm": 0.12735195457935333, + "learning_rate": 2.695569197902694e-05, + "loss": 0.03, + "step": 29530 + }, + { + "epoch": 2.6849663697509545, + "grad_norm": 0.15241436660289764, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0312, + "step": 29540 + }, + { + "epoch": 2.6858752954008365, + "grad_norm": 0.1896163523197174, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0312, + "step": 29550 + }, + { + "epoch": 2.686784221050718, + "grad_norm": 0.13706065714359283, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0304, + "step": 29560 + }, + { + "epoch": 2.6876931467006, + "grad_norm": 0.19005009531974792, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0311, + "step": 29570 + }, + { + "epoch": 2.6886020723504815, + "grad_norm": 0.16778485476970673, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0349, + "step": 29580 + }, + { + "epoch": 2.6895109980003635, + "grad_norm": 0.1751796156167984, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0301, + "step": 29590 + }, + { + "epoch": 2.6904199236502455, + "grad_norm": 0.1917346715927124, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.033, + "step": 29600 + }, + { + "epoch": 2.6913288493001275, + "grad_norm": 0.1667744517326355, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0308, + "step": 29610 + }, + { + "epoch": 2.692237774950009, + "grad_norm": 0.11286471784114838, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0305, + "step": 29620 + }, + { + "epoch": 2.693146700599891, + "grad_norm": 0.1818513721227646, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0322, + "step": 29630 + }, + { + "epoch": 2.6940556262497726, + "grad_norm": 0.21561549603939056, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0331, + "step": 29640 + }, + { + "epoch": 2.6949645518996546, + "grad_norm": 0.14269347488880157, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0321, + "step": 29650 + }, + { + "epoch": 2.6958734775495365, + "grad_norm": 0.12441715598106384, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0308, + "step": 29660 + }, + { + "epoch": 2.6967824031994185, + "grad_norm": 0.19229398667812347, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0314, + "step": 29670 + }, + { + "epoch": 2.6976913288493, + "grad_norm": 0.17466047406196594, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0292, + "step": 29680 + }, + { + "epoch": 2.698600254499182, + "grad_norm": 0.18871143460273743, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0315, + "step": 29690 + }, + { + "epoch": 2.6995091801490636, + "grad_norm": 0.14990587532520294, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0321, + "step": 29700 + }, + { + "epoch": 2.7004181057989456, + "grad_norm": 0.11922817677259445, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0295, + "step": 29710 + }, + { + "epoch": 2.7013270314488276, + "grad_norm": 0.20468170940876007, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0303, + "step": 29720 + }, + { + "epoch": 2.7022359570987096, + "grad_norm": 0.14460216462612152, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0303, + "step": 29730 + }, + { + "epoch": 2.703144882748591, + "grad_norm": 0.17615117132663727, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.03, + "step": 29740 + }, + { + "epoch": 2.704053808398473, + "grad_norm": 0.17269878089427948, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0318, + "step": 29750 + }, + { + "epoch": 2.7049627340483546, + "grad_norm": 0.11961200088262558, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0323, + "step": 29760 + }, + { + "epoch": 2.7058716596982366, + "grad_norm": 0.11338043957948685, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0336, + "step": 29770 + }, + { + "epoch": 2.7067805853481186, + "grad_norm": 0.12741407752037048, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0328, + "step": 29780 + }, + { + "epoch": 2.7076895109980006, + "grad_norm": 0.12508165836334229, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0297, + "step": 29790 + }, + { + "epoch": 2.708598436647882, + "grad_norm": 0.18153119087219238, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0336, + "step": 29800 + }, + { + "epoch": 2.709507362297764, + "grad_norm": 0.19494600594043732, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.033, + "step": 29810 + }, + { + "epoch": 2.7104162879476457, + "grad_norm": 0.14633987843990326, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0296, + "step": 29820 + }, + { + "epoch": 2.7113252135975277, + "grad_norm": 0.1551511138677597, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0321, + "step": 29830 + }, + { + "epoch": 2.7122341392474096, + "grad_norm": 0.20861518383026123, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0319, + "step": 29840 + }, + { + "epoch": 2.7131430648972916, + "grad_norm": 0.16290758550167084, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0307, + "step": 29850 + }, + { + "epoch": 2.714051990547173, + "grad_norm": 0.1537388563156128, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0323, + "step": 29860 + }, + { + "epoch": 2.714960916197055, + "grad_norm": 0.16967332363128662, + "learning_rate": 2.687029075082441e-05, + "loss": 0.03, + "step": 29870 + }, + { + "epoch": 2.7158698418469367, + "grad_norm": 0.15837326645851135, + "learning_rate": 2.686776313743295e-05, + "loss": 0.0314, + "step": 29880 + }, + { + "epoch": 2.7167787674968187, + "grad_norm": 0.1819511502981186, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0324, + "step": 29890 + }, + { + "epoch": 2.7176876931467007, + "grad_norm": 0.11532271653413773, + "learning_rate": 2.686270520700719e-05, + "loss": 0.0298, + "step": 29900 + }, + { + "epoch": 2.7185966187965827, + "grad_norm": 0.14462041854858398, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0313, + "step": 29910 + }, + { + "epoch": 2.719505544446464, + "grad_norm": 0.14045827090740204, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0339, + "step": 29920 + }, + { + "epoch": 2.720414470096346, + "grad_norm": 0.13998600840568542, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0323, + "step": 29930 + }, + { + "epoch": 2.7213233957462277, + "grad_norm": 0.1339585781097412, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0334, + "step": 29940 + }, + { + "epoch": 2.7222323213961097, + "grad_norm": 0.1146501749753952, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0337, + "step": 29950 + }, + { + "epoch": 2.7231412470459917, + "grad_norm": 0.12666364014148712, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0314, + "step": 29960 + }, + { + "epoch": 2.7240501726958737, + "grad_norm": 0.12790675461292267, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0322, + "step": 29970 + }, + { + "epoch": 2.7249590983457552, + "grad_norm": 0.23798829317092896, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0313, + "step": 29980 + }, + { + "epoch": 2.725868023995637, + "grad_norm": 0.1689607948064804, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0287, + "step": 29990 + }, + { + "epoch": 2.7267769496455188, + "grad_norm": 0.130876362323761, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0287, + "step": 30000 + }, + { + "epoch": 2.7276858752954007, + "grad_norm": 0.16387943923473358, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0328, + "step": 30010 + }, + { + "epoch": 2.7285948009452827, + "grad_norm": 0.15148302912712097, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0306, + "step": 30020 + }, + { + "epoch": 2.7295037265951647, + "grad_norm": 0.17840145528316498, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0301, + "step": 30030 + }, + { + "epoch": 2.7304126522450463, + "grad_norm": 0.1653406172990799, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0331, + "step": 30040 + }, + { + "epoch": 2.7313215778949282, + "grad_norm": 0.14868462085723877, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.032, + "step": 30050 + }, + { + "epoch": 2.73223050354481, + "grad_norm": 0.14724138379096985, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0332, + "step": 30060 + }, + { + "epoch": 2.7331394291946918, + "grad_norm": 0.136787548661232, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0325, + "step": 30070 + }, + { + "epoch": 2.7340483548445738, + "grad_norm": 0.1243073120713234, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0292, + "step": 30080 + }, + { + "epoch": 2.7349572804944557, + "grad_norm": 0.1824236512184143, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.034, + "step": 30090 + }, + { + "epoch": 2.7358662061443373, + "grad_norm": 0.14192496240139008, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0354, + "step": 30100 + }, + { + "epoch": 2.7367751317942193, + "grad_norm": 0.1608245074748993, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0307, + "step": 30110 + }, + { + "epoch": 2.737684057444101, + "grad_norm": 0.16748805344104767, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0297, + "step": 30120 + }, + { + "epoch": 2.738592983093983, + "grad_norm": 0.18298916518688202, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0333, + "step": 30130 + }, + { + "epoch": 2.739501908743865, + "grad_norm": 0.15964213013648987, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0323, + "step": 30140 + }, + { + "epoch": 2.7404108343937468, + "grad_norm": 0.14229585230350494, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.031, + "step": 30150 + }, + { + "epoch": 2.7413197600436283, + "grad_norm": 0.1366785764694214, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0305, + "step": 30160 + }, + { + "epoch": 2.7422286856935103, + "grad_norm": 0.15059910714626312, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0291, + "step": 30170 + }, + { + "epoch": 2.743137611343392, + "grad_norm": 0.15916801989078522, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0317, + "step": 30180 + }, + { + "epoch": 2.744046536993274, + "grad_norm": 0.11928148567676544, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0276, + "step": 30190 + }, + { + "epoch": 2.744955462643156, + "grad_norm": 0.13773319125175476, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0321, + "step": 30200 + }, + { + "epoch": 2.745864388293038, + "grad_norm": 0.18688197433948517, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0315, + "step": 30210 + }, + { + "epoch": 2.7467733139429193, + "grad_norm": 0.14684376120567322, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0302, + "step": 30220 + }, + { + "epoch": 2.7476822395928013, + "grad_norm": 0.1399802416563034, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0318, + "step": 30230 + }, + { + "epoch": 2.748591165242683, + "grad_norm": 0.1398986428976059, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0323, + "step": 30240 + }, + { + "epoch": 2.749500090892565, + "grad_norm": 0.12950368225574493, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0317, + "step": 30250 + }, + { + "epoch": 2.750409016542447, + "grad_norm": 0.12679634988307953, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.03, + "step": 30260 + }, + { + "epoch": 2.751317942192329, + "grad_norm": 0.1378110647201538, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0341, + "step": 30270 + }, + { + "epoch": 2.7522268678422104, + "grad_norm": 0.20497845113277435, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0293, + "step": 30280 + }, + { + "epoch": 2.7531357934920924, + "grad_norm": 0.19678737223148346, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0309, + "step": 30290 + }, + { + "epoch": 2.7540447191419744, + "grad_norm": 0.1386503130197525, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0324, + "step": 30300 + }, + { + "epoch": 2.754953644791856, + "grad_norm": 0.19575095176696777, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0323, + "step": 30310 + }, + { + "epoch": 2.755862570441738, + "grad_norm": 0.1651073694229126, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0298, + "step": 30320 + }, + { + "epoch": 2.75677149609162, + "grad_norm": 0.17502693831920624, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0337, + "step": 30330 + }, + { + "epoch": 2.7576804217415014, + "grad_norm": 0.12534594535827637, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0305, + "step": 30340 + }, + { + "epoch": 2.7585893473913834, + "grad_norm": 0.13419444859027863, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0329, + "step": 30350 + }, + { + "epoch": 2.7594982730412654, + "grad_norm": 0.12875548005104065, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0309, + "step": 30360 + }, + { + "epoch": 2.760407198691147, + "grad_norm": 0.14464712142944336, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0318, + "step": 30370 + }, + { + "epoch": 2.761316124341029, + "grad_norm": 0.13750147819519043, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0302, + "step": 30380 + }, + { + "epoch": 2.762225049990911, + "grad_norm": 0.13865141570568085, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0317, + "step": 30390 + }, + { + "epoch": 2.7631339756407924, + "grad_norm": 0.15426377952098846, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0347, + "step": 30400 + }, + { + "epoch": 2.7640429012906744, + "grad_norm": 0.12642855942249298, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0335, + "step": 30410 + }, + { + "epoch": 2.7649518269405564, + "grad_norm": 0.12521088123321533, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0314, + "step": 30420 + }, + { + "epoch": 2.765860752590438, + "grad_norm": 0.17702017724514008, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0299, + "step": 30430 + }, + { + "epoch": 2.76676967824032, + "grad_norm": 0.13361677527427673, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0273, + "step": 30440 + }, + { + "epoch": 2.767678603890202, + "grad_norm": 0.14606201648712158, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0309, + "step": 30450 + }, + { + "epoch": 2.7685875295400835, + "grad_norm": 0.14452295005321503, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0321, + "step": 30460 + }, + { + "epoch": 2.7694964551899655, + "grad_norm": 0.14978957176208496, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0305, + "step": 30470 + }, + { + "epoch": 2.7704053808398474, + "grad_norm": 0.1457444429397583, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.031, + "step": 30480 + }, + { + "epoch": 2.771314306489729, + "grad_norm": 0.1741725653409958, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0304, + "step": 30490 + }, + { + "epoch": 2.772223232139611, + "grad_norm": 0.13309229910373688, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0278, + "step": 30500 + }, + { + "epoch": 2.773132157789493, + "grad_norm": 0.13325431942939758, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0278, + "step": 30510 + }, + { + "epoch": 2.7740410834393745, + "grad_norm": 0.1267113834619522, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0315, + "step": 30520 + }, + { + "epoch": 2.7749500090892565, + "grad_norm": 0.1582508087158203, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0318, + "step": 30530 + }, + { + "epoch": 2.7758589347391385, + "grad_norm": 0.1465083360671997, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0316, + "step": 30540 + }, + { + "epoch": 2.77676786038902, + "grad_norm": 0.14446674287319183, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0311, + "step": 30550 + }, + { + "epoch": 2.777676786038902, + "grad_norm": 0.13963867723941803, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.03, + "step": 30560 + }, + { + "epoch": 2.778585711688784, + "grad_norm": 0.1818837970495224, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0327, + "step": 30570 + }, + { + "epoch": 2.779494637338666, + "grad_norm": 0.1775745004415512, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0283, + "step": 30580 + }, + { + "epoch": 2.7804035629885475, + "grad_norm": 0.16260139644145966, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0303, + "step": 30590 + }, + { + "epoch": 2.7813124886384295, + "grad_norm": 0.15280579030513763, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0319, + "step": 30600 + }, + { + "epoch": 2.782221414288311, + "grad_norm": 0.16807132959365845, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0338, + "step": 30610 + }, + { + "epoch": 2.783130339938193, + "grad_norm": 0.18383705615997314, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0304, + "step": 30620 + }, + { + "epoch": 2.784039265588075, + "grad_norm": 0.1746561974287033, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0315, + "step": 30630 + }, + { + "epoch": 2.784948191237957, + "grad_norm": 0.1482304334640503, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0321, + "step": 30640 + }, + { + "epoch": 2.7858571168878385, + "grad_norm": 0.12026659399271011, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0283, + "step": 30650 + }, + { + "epoch": 2.7867660425377205, + "grad_norm": 0.136226624250412, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0308, + "step": 30660 + }, + { + "epoch": 2.787674968187602, + "grad_norm": 0.17553462088108063, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0294, + "step": 30670 + }, + { + "epoch": 2.788583893837484, + "grad_norm": 0.16678215563297272, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0333, + "step": 30680 + }, + { + "epoch": 2.789492819487366, + "grad_norm": 0.1300824135541916, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.03, + "step": 30690 + }, + { + "epoch": 2.790401745137248, + "grad_norm": 0.1492035686969757, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0323, + "step": 30700 + }, + { + "epoch": 2.7913106707871296, + "grad_norm": 0.1558852642774582, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0328, + "step": 30710 + }, + { + "epoch": 2.7922195964370116, + "grad_norm": 0.17817093431949615, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0301, + "step": 30720 + }, + { + "epoch": 2.793128522086893, + "grad_norm": 0.19296404719352722, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0351, + "step": 30730 + }, + { + "epoch": 2.794037447736775, + "grad_norm": 0.1394781619310379, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.031, + "step": 30740 + }, + { + "epoch": 2.794946373386657, + "grad_norm": 0.16695450246334076, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0341, + "step": 30750 + }, + { + "epoch": 2.795855299036539, + "grad_norm": 0.15328814089298248, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0326, + "step": 30760 + }, + { + "epoch": 2.7967642246864206, + "grad_norm": 0.13398675620555878, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0313, + "step": 30770 + }, + { + "epoch": 2.7976731503363026, + "grad_norm": 0.12902505695819855, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0302, + "step": 30780 + }, + { + "epoch": 2.798582075986184, + "grad_norm": 0.12705540657043457, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0295, + "step": 30790 + }, + { + "epoch": 2.799491001636066, + "grad_norm": 0.14218777418136597, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0331, + "step": 30800 + }, + { + "epoch": 2.800399927285948, + "grad_norm": 0.2055811882019043, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0303, + "step": 30810 + }, + { + "epoch": 2.80130885293583, + "grad_norm": 0.14213085174560547, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0308, + "step": 30820 + }, + { + "epoch": 2.8022177785857116, + "grad_norm": 0.15348291397094727, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0321, + "step": 30830 + }, + { + "epoch": 2.8031267042355936, + "grad_norm": 0.15103738009929657, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0283, + "step": 30840 + }, + { + "epoch": 2.804035629885475, + "grad_norm": 0.1756531000137329, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0307, + "step": 30850 + }, + { + "epoch": 2.804944555535357, + "grad_norm": 0.1400781273841858, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0306, + "step": 30860 + }, + { + "epoch": 2.805853481185239, + "grad_norm": 0.14777547121047974, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0266, + "step": 30870 + }, + { + "epoch": 2.806762406835121, + "grad_norm": 0.11170168220996857, + "learning_rate": 2.661048262196233e-05, + "loss": 0.03, + "step": 30880 + }, + { + "epoch": 2.8076713324850027, + "grad_norm": 0.12479893863201141, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0315, + "step": 30890 + }, + { + "epoch": 2.8085802581348847, + "grad_norm": 0.1630609780550003, + "learning_rate": 2.660524641755484e-05, + "loss": 0.034, + "step": 30900 + }, + { + "epoch": 2.809489183784766, + "grad_norm": 0.13932955265045166, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0293, + "step": 30910 + }, + { + "epoch": 2.810398109434648, + "grad_norm": 0.14820118248462677, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0289, + "step": 30920 + }, + { + "epoch": 2.81130703508453, + "grad_norm": 0.15590697526931763, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0311, + "step": 30930 + }, + { + "epoch": 2.812215960734412, + "grad_norm": 0.17380648851394653, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0297, + "step": 30940 + }, + { + "epoch": 2.8131248863842937, + "grad_norm": 0.21934081614017487, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0315, + "step": 30950 + }, + { + "epoch": 2.8140338120341757, + "grad_norm": 0.15111571550369263, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0328, + "step": 30960 + }, + { + "epoch": 2.8149427376840572, + "grad_norm": 0.10755475610494614, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0301, + "step": 30970 + }, + { + "epoch": 2.815851663333939, + "grad_norm": 0.11361601203680038, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0309, + "step": 30980 + }, + { + "epoch": 2.816760588983821, + "grad_norm": 0.1379053294658661, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0313, + "step": 30990 + }, + { + "epoch": 2.817669514633703, + "grad_norm": 0.1524042934179306, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0322, + "step": 31000 + }, + { + "epoch": 2.8185784402835847, + "grad_norm": 0.15289293229579926, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.031, + "step": 31010 + }, + { + "epoch": 2.8194873659334667, + "grad_norm": 0.12142954766750336, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0307, + "step": 31020 + }, + { + "epoch": 2.8203962915833483, + "grad_norm": 0.225092813372612, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.032, + "step": 31030 + }, + { + "epoch": 2.8213052172332302, + "grad_norm": 0.1395580768585205, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.032, + "step": 31040 + }, + { + "epoch": 2.8222141428831122, + "grad_norm": 0.13548967242240906, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0321, + "step": 31050 + }, + { + "epoch": 2.823123068532994, + "grad_norm": 0.1736127734184265, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0322, + "step": 31060 + }, + { + "epoch": 2.8240319941828758, + "grad_norm": 0.13610070943832397, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0295, + "step": 31070 + }, + { + "epoch": 2.8249409198327577, + "grad_norm": 0.1676744818687439, + "learning_rate": 2.655796212758018e-05, + "loss": 0.032, + "step": 31080 + }, + { + "epoch": 2.8258498454826393, + "grad_norm": 0.12642014026641846, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0279, + "step": 31090 + }, + { + "epoch": 2.8267587711325213, + "grad_norm": 0.15970604121685028, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0322, + "step": 31100 + }, + { + "epoch": 2.8276676967824033, + "grad_norm": 0.13978463411331177, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0326, + "step": 31110 + }, + { + "epoch": 2.8285766224322852, + "grad_norm": 0.16898775100708008, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0348, + "step": 31120 + }, + { + "epoch": 2.829485548082167, + "grad_norm": 0.19531728327274323, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.032, + "step": 31130 + }, + { + "epoch": 2.8303944737320488, + "grad_norm": 0.14376655220985413, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0304, + "step": 31140 + }, + { + "epoch": 2.8313033993819303, + "grad_norm": 0.13302360475063324, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0287, + "step": 31150 + }, + { + "epoch": 2.8322123250318123, + "grad_norm": 0.19368626177310944, + "learning_rate": 2.653685553226646e-05, + "loss": 0.03, + "step": 31160 + }, + { + "epoch": 2.8331212506816943, + "grad_norm": 0.1681632399559021, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0311, + "step": 31170 + }, + { + "epoch": 2.8340301763315763, + "grad_norm": 0.14594611525535583, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.033, + "step": 31180 + }, + { + "epoch": 2.834939101981458, + "grad_norm": 0.14614757895469666, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0326, + "step": 31190 + }, + { + "epoch": 2.83584802763134, + "grad_norm": 0.13346980512142181, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0296, + "step": 31200 + }, + { + "epoch": 2.8367569532812213, + "grad_norm": 0.12056049704551697, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0295, + "step": 31210 + }, + { + "epoch": 2.8376658789311033, + "grad_norm": 0.113380566239357, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0301, + "step": 31220 + }, + { + "epoch": 2.8385748045809853, + "grad_norm": 0.1676401048898697, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0299, + "step": 31230 + }, + { + "epoch": 2.8394837302308673, + "grad_norm": 0.13822272419929504, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0299, + "step": 31240 + }, + { + "epoch": 2.840392655880749, + "grad_norm": 0.12134786695241928, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0318, + "step": 31250 + }, + { + "epoch": 2.841301581530631, + "grad_norm": 0.16223718225955963, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0302, + "step": 31260 + }, + { + "epoch": 2.8422105071805124, + "grad_norm": 0.19224898517131805, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0321, + "step": 31270 + }, + { + "epoch": 2.8431194328303944, + "grad_norm": 0.18413971364498138, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0292, + "step": 31280 + }, + { + "epoch": 2.8440283584802764, + "grad_norm": 0.14999234676361084, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0335, + "step": 31290 + }, + { + "epoch": 2.8449372841301583, + "grad_norm": 0.16004398465156555, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0295, + "step": 31300 + }, + { + "epoch": 2.84584620978004, + "grad_norm": 0.172446608543396, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0305, + "step": 31310 + }, + { + "epoch": 2.846755135429922, + "grad_norm": 0.13977979123592377, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0319, + "step": 31320 + }, + { + "epoch": 2.8476640610798034, + "grad_norm": 0.19043248891830444, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0321, + "step": 31330 + }, + { + "epoch": 2.8485729867296854, + "grad_norm": 0.17781297862529755, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0319, + "step": 31340 + }, + { + "epoch": 2.8494819123795674, + "grad_norm": 0.1423209309577942, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0297, + "step": 31350 + }, + { + "epoch": 2.8503908380294494, + "grad_norm": 0.12612034380435944, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0301, + "step": 31360 + }, + { + "epoch": 2.851299763679331, + "grad_norm": 0.1684209108352661, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0328, + "step": 31370 + }, + { + "epoch": 2.852208689329213, + "grad_norm": 0.1479019671678543, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0279, + "step": 31380 + }, + { + "epoch": 2.8531176149790944, + "grad_norm": 0.20995713770389557, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0315, + "step": 31390 + }, + { + "epoch": 2.8540265406289764, + "grad_norm": 0.2076563984155655, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0308, + "step": 31400 + }, + { + "epoch": 2.8549354662788584, + "grad_norm": 0.14742836356163025, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.0342, + "step": 31410 + }, + { + "epoch": 2.8558443919287404, + "grad_norm": 0.13883426785469055, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0285, + "step": 31420 + }, + { + "epoch": 2.856753317578622, + "grad_norm": 0.1341303288936615, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0299, + "step": 31430 + }, + { + "epoch": 2.857662243228504, + "grad_norm": 0.13285775482654572, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0294, + "step": 31440 + }, + { + "epoch": 2.8585711688783855, + "grad_norm": 0.14730298519134521, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0315, + "step": 31450 + }, + { + "epoch": 2.8594800945282675, + "grad_norm": 0.12057898193597794, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.027, + "step": 31460 + }, + { + "epoch": 2.8603890201781494, + "grad_norm": 0.16059695184230804, + "learning_rate": 2.645453877635046e-05, + "loss": 0.03, + "step": 31470 + }, + { + "epoch": 2.8612979458280314, + "grad_norm": 0.1723269820213318, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0307, + "step": 31480 + }, + { + "epoch": 2.862206871477913, + "grad_norm": 0.14946328103542328, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0299, + "step": 31490 + }, + { + "epoch": 2.863115797127795, + "grad_norm": 0.18318508565425873, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0316, + "step": 31500 + }, + { + "epoch": 2.864024722777677, + "grad_norm": 0.11655567586421967, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0285, + "step": 31510 + }, + { + "epoch": 2.8649336484275585, + "grad_norm": 0.16908925771713257, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0298, + "step": 31520 + }, + { + "epoch": 2.8658425740774405, + "grad_norm": 0.14461879432201385, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.031, + "step": 31530 + }, + { + "epoch": 2.8667514997273225, + "grad_norm": 0.17127078771591187, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0296, + "step": 31540 + }, + { + "epoch": 2.867660425377204, + "grad_norm": 0.1990242600440979, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0313, + "step": 31550 + }, + { + "epoch": 2.868569351027086, + "grad_norm": 0.1672896295785904, + "learning_rate": 2.643048351516179e-05, + "loss": 0.032, + "step": 31560 + }, + { + "epoch": 2.869478276676968, + "grad_norm": 0.14292682707309723, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0314, + "step": 31570 + }, + { + "epoch": 2.8703872023268495, + "grad_norm": 0.11605484038591385, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0291, + "step": 31580 + }, + { + "epoch": 2.8712961279767315, + "grad_norm": 0.1169004738330841, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0278, + "step": 31590 + }, + { + "epoch": 2.8722050536266135, + "grad_norm": 0.15918979048728943, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0294, + "step": 31600 + }, + { + "epoch": 2.873113979276495, + "grad_norm": 0.1854044795036316, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0312, + "step": 31610 + }, + { + "epoch": 2.874022904926377, + "grad_norm": 0.14705878496170044, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0319, + "step": 31620 + }, + { + "epoch": 2.874931830576259, + "grad_norm": 0.14595726132392883, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.0309, + "step": 31630 + }, + { + "epoch": 2.8758407562261405, + "grad_norm": 0.13591863214969635, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0285, + "step": 31640 + }, + { + "epoch": 2.8767496818760225, + "grad_norm": 0.13952696323394775, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0317, + "step": 31650 + }, + { + "epoch": 2.8776586075259045, + "grad_norm": 0.12432755529880524, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0291, + "step": 31660 + }, + { + "epoch": 2.878567533175786, + "grad_norm": 0.143733412027359, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0299, + "step": 31670 + }, + { + "epoch": 2.879476458825668, + "grad_norm": 0.15009461343288422, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0298, + "step": 31680 + }, + { + "epoch": 2.88038538447555, + "grad_norm": 0.11566843092441559, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0277, + "step": 31690 + }, + { + "epoch": 2.8812943101254316, + "grad_norm": 0.14148753881454468, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0325, + "step": 31700 + }, + { + "epoch": 2.8822032357753136, + "grad_norm": 0.14153897762298584, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0288, + "step": 31710 + }, + { + "epoch": 2.8831121614251956, + "grad_norm": 0.12382383644580841, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0292, + "step": 31720 + }, + { + "epoch": 2.884021087075077, + "grad_norm": 0.1329651027917862, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0295, + "step": 31730 + }, + { + "epoch": 2.884930012724959, + "grad_norm": 0.19482187926769257, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.032, + "step": 31740 + }, + { + "epoch": 2.885838938374841, + "grad_norm": 0.1531609296798706, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0303, + "step": 31750 + }, + { + "epoch": 2.8867478640247226, + "grad_norm": 0.13232266902923584, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0279, + "step": 31760 + }, + { + "epoch": 2.8876567896746046, + "grad_norm": 0.14230208098888397, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0307, + "step": 31770 + }, + { + "epoch": 2.8885657153244866, + "grad_norm": 0.14207783341407776, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0315, + "step": 31780 + }, + { + "epoch": 2.8894746409743686, + "grad_norm": 0.17197157442569733, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0325, + "step": 31790 + }, + { + "epoch": 2.89038356662425, + "grad_norm": 0.11061709374189377, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.0279, + "step": 31800 + }, + { + "epoch": 2.891292492274132, + "grad_norm": 0.1436084657907486, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0324, + "step": 31810 + }, + { + "epoch": 2.8922014179240136, + "grad_norm": 0.16952642798423767, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0313, + "step": 31820 + }, + { + "epoch": 2.8931103435738956, + "grad_norm": 0.14962232112884521, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0298, + "step": 31830 + }, + { + "epoch": 2.8940192692237776, + "grad_norm": 0.14000669121742249, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0319, + "step": 31840 + }, + { + "epoch": 2.8949281948736596, + "grad_norm": 0.1607324182987213, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.035, + "step": 31850 + }, + { + "epoch": 2.895837120523541, + "grad_norm": 0.1511267125606537, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0346, + "step": 31860 + }, + { + "epoch": 2.896746046173423, + "grad_norm": 0.1431674361228943, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0299, + "step": 31870 + }, + { + "epoch": 2.8976549718233047, + "grad_norm": 0.14791177213191986, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0316, + "step": 31880 + }, + { + "epoch": 2.8985638974731867, + "grad_norm": 0.16497613489627838, + "learning_rate": 2.634168044873305e-05, + "loss": 0.034, + "step": 31890 + }, + { + "epoch": 2.8994728231230686, + "grad_norm": 0.13891412317752838, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0321, + "step": 31900 + }, + { + "epoch": 2.9003817487729506, + "grad_norm": 0.13929802179336548, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0295, + "step": 31910 + }, + { + "epoch": 2.901290674422832, + "grad_norm": 0.1440490335226059, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0282, + "step": 31920 + }, + { + "epoch": 2.902199600072714, + "grad_norm": 0.170762836933136, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.03, + "step": 31930 + }, + { + "epoch": 2.9031085257225957, + "grad_norm": 0.12683051824569702, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0313, + "step": 31940 + }, + { + "epoch": 2.9040174513724777, + "grad_norm": 0.1610388159751892, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0351, + "step": 31950 + }, + { + "epoch": 2.9049263770223597, + "grad_norm": 0.14567877352237701, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0279, + "step": 31960 + }, + { + "epoch": 2.9058353026722417, + "grad_norm": 0.15191060304641724, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0331, + "step": 31970 + }, + { + "epoch": 2.906744228322123, + "grad_norm": 0.14441156387329102, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0321, + "step": 31980 + }, + { + "epoch": 2.907653153972005, + "grad_norm": 0.12322787940502167, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0305, + "step": 31990 + }, + { + "epoch": 2.9085620796218867, + "grad_norm": 0.16533412039279938, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0294, + "step": 32000 + }, + { + "epoch": 2.9094710052717687, + "grad_norm": 0.15060535073280334, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0305, + "step": 32010 + }, + { + "epoch": 2.9103799309216507, + "grad_norm": 0.1424243301153183, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0318, + "step": 32020 + }, + { + "epoch": 2.9112888565715327, + "grad_norm": 0.15310092270374298, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0311, + "step": 32030 + }, + { + "epoch": 2.9121977822214142, + "grad_norm": 0.15730004012584686, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0319, + "step": 32040 + }, + { + "epoch": 2.913106707871296, + "grad_norm": 0.1700444370508194, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0322, + "step": 32050 + }, + { + "epoch": 2.9140156335211778, + "grad_norm": 0.16288109123706818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0276, + "step": 32060 + }, + { + "epoch": 2.9149245591710597, + "grad_norm": 0.20477785170078278, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0311, + "step": 32070 + }, + { + "epoch": 2.9158334848209417, + "grad_norm": 0.14812704920768738, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0296, + "step": 32080 + }, + { + "epoch": 2.9167424104708237, + "grad_norm": 0.1797877699136734, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0329, + "step": 32090 + }, + { + "epoch": 2.9176513361207053, + "grad_norm": 0.14459554851055145, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0347, + "step": 32100 + }, + { + "epoch": 2.9185602617705872, + "grad_norm": 0.15531472861766815, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0297, + "step": 32110 + }, + { + "epoch": 2.919469187420469, + "grad_norm": 0.14968480169773102, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0307, + "step": 32120 + }, + { + "epoch": 2.920378113070351, + "grad_norm": 0.1478489488363266, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0303, + "step": 32130 + }, + { + "epoch": 2.9212870387202328, + "grad_norm": 0.14865200221538544, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0328, + "step": 32140 + }, + { + "epoch": 2.9221959643701148, + "grad_norm": 0.1452280879020691, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0325, + "step": 32150 + }, + { + "epoch": 2.9231048900199963, + "grad_norm": 0.13320516049861908, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0315, + "step": 32160 + }, + { + "epoch": 2.9240138156698783, + "grad_norm": 0.14542579650878906, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0281, + "step": 32170 + }, + { + "epoch": 2.92492274131976, + "grad_norm": 0.13369297981262207, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.03, + "step": 32180 + }, + { + "epoch": 2.925831666969642, + "grad_norm": 0.12691278755664825, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0295, + "step": 32190 + }, + { + "epoch": 2.926740592619524, + "grad_norm": 0.1422904133796692, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0299, + "step": 32200 + }, + { + "epoch": 2.927649518269406, + "grad_norm": 0.20425385236740112, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0292, + "step": 32210 + }, + { + "epoch": 2.9285584439192873, + "grad_norm": 0.13142549991607666, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0309, + "step": 32220 + }, + { + "epoch": 2.9294673695691693, + "grad_norm": 0.163040891289711, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0309, + "step": 32230 + }, + { + "epoch": 2.930376295219051, + "grad_norm": 0.17021650075912476, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0295, + "step": 32240 + }, + { + "epoch": 2.931285220868933, + "grad_norm": 0.1518167108297348, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0315, + "step": 32250 + }, + { + "epoch": 2.932194146518815, + "grad_norm": 0.1352534145116806, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0293, + "step": 32260 + }, + { + "epoch": 2.933103072168697, + "grad_norm": 0.13885299861431122, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0299, + "step": 32270 + }, + { + "epoch": 2.9340119978185784, + "grad_norm": 0.18758520483970642, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0297, + "step": 32280 + }, + { + "epoch": 2.9349209234684603, + "grad_norm": 0.12550200521945953, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0285, + "step": 32290 + }, + { + "epoch": 2.935829849118342, + "grad_norm": 0.12800316512584686, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0309, + "step": 32300 + }, + { + "epoch": 2.936738774768224, + "grad_norm": 0.1363116055727005, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0335, + "step": 32310 + }, + { + "epoch": 2.937647700418106, + "grad_norm": 0.13640958070755005, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0304, + "step": 32320 + }, + { + "epoch": 2.938556626067988, + "grad_norm": 0.16100554168224335, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0278, + "step": 32330 + }, + { + "epoch": 2.9394655517178694, + "grad_norm": 0.12750770151615143, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0322, + "step": 32340 + }, + { + "epoch": 2.9403744773677514, + "grad_norm": 0.13702838122844696, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0323, + "step": 32350 + }, + { + "epoch": 2.941283403017633, + "grad_norm": 0.14516977965831757, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0322, + "step": 32360 + }, + { + "epoch": 2.942192328667515, + "grad_norm": 0.18312013149261475, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0296, + "step": 32370 + }, + { + "epoch": 2.943101254317397, + "grad_norm": 0.13234896957874298, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0323, + "step": 32380 + }, + { + "epoch": 2.944010179967279, + "grad_norm": 0.17623069882392883, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0289, + "step": 32390 + }, + { + "epoch": 2.9449191056171604, + "grad_norm": 0.1615341305732727, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.03, + "step": 32400 + }, + { + "epoch": 2.9458280312670424, + "grad_norm": 0.1501336544752121, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0273, + "step": 32410 + }, + { + "epoch": 2.946736956916924, + "grad_norm": 0.1633436679840088, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0335, + "step": 32420 + }, + { + "epoch": 2.947645882566806, + "grad_norm": 0.11848217248916626, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.031, + "step": 32430 + }, + { + "epoch": 2.948554808216688, + "grad_norm": 0.17862960696220398, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.032, + "step": 32440 + }, + { + "epoch": 2.94946373386657, + "grad_norm": 0.16005802154541016, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0291, + "step": 32450 + }, + { + "epoch": 2.9503726595164514, + "grad_norm": 0.13045911490917206, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0279, + "step": 32460 + }, + { + "epoch": 2.9512815851663334, + "grad_norm": 0.17794157564640045, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0303, + "step": 32470 + }, + { + "epoch": 2.952190510816215, + "grad_norm": 0.12584827840328217, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0288, + "step": 32480 + }, + { + "epoch": 2.953099436466097, + "grad_norm": 0.15469233691692352, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0306, + "step": 32490 + }, + { + "epoch": 2.954008362115979, + "grad_norm": 0.17964816093444824, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0325, + "step": 32500 + }, + { + "epoch": 2.954917287765861, + "grad_norm": 0.13775910437107086, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0311, + "step": 32510 + }, + { + "epoch": 2.9558262134157425, + "grad_norm": 0.13837912678718567, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0299, + "step": 32520 + }, + { + "epoch": 2.9567351390656245, + "grad_norm": 0.19169574975967407, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0328, + "step": 32530 + }, + { + "epoch": 2.957644064715506, + "grad_norm": 0.17710642516613007, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0296, + "step": 32540 + }, + { + "epoch": 2.958552990365388, + "grad_norm": 0.13566479086875916, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0297, + "step": 32550 + }, + { + "epoch": 2.95946191601527, + "grad_norm": 0.16637323796749115, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0317, + "step": 32560 + }, + { + "epoch": 2.960370841665152, + "grad_norm": 0.1506177932024002, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0323, + "step": 32570 + }, + { + "epoch": 2.9612797673150335, + "grad_norm": 0.12719877064228058, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0306, + "step": 32580 + }, + { + "epoch": 2.9621886929649155, + "grad_norm": 0.15392877161502838, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0297, + "step": 32590 + }, + { + "epoch": 2.963097618614797, + "grad_norm": 0.11697126179933548, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0297, + "step": 32600 + }, + { + "epoch": 2.964006544264679, + "grad_norm": 0.17634592950344086, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0351, + "step": 32610 + }, + { + "epoch": 2.964915469914561, + "grad_norm": 0.17040549218654633, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0293, + "step": 32620 + }, + { + "epoch": 2.965824395564443, + "grad_norm": 0.13326941430568695, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0309, + "step": 32630 + }, + { + "epoch": 2.9667333212143245, + "grad_norm": 0.1659083515405655, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0281, + "step": 32640 + }, + { + "epoch": 2.9676422468642065, + "grad_norm": 0.14018382132053375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0303, + "step": 32650 + }, + { + "epoch": 2.9685511725140885, + "grad_norm": 0.11960762739181519, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0288, + "step": 32660 + }, + { + "epoch": 2.96946009816397, + "grad_norm": 0.14998720586299896, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0309, + "step": 32670 + }, + { + "epoch": 2.970369023813852, + "grad_norm": 0.1565612256526947, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0292, + "step": 32680 + }, + { + "epoch": 2.971277949463734, + "grad_norm": 0.16833044588565826, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0307, + "step": 32690 + }, + { + "epoch": 2.9721868751136156, + "grad_norm": 0.13893653452396393, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0313, + "step": 32700 + }, + { + "epoch": 2.9730958007634976, + "grad_norm": 0.19096510112285614, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0331, + "step": 32710 + }, + { + "epoch": 2.9740047264133795, + "grad_norm": 0.17378269135951996, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0294, + "step": 32720 + }, + { + "epoch": 2.974913652063261, + "grad_norm": 0.14959178864955902, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0315, + "step": 32730 + }, + { + "epoch": 2.975822577713143, + "grad_norm": 0.13975603878498077, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0303, + "step": 32740 + }, + { + "epoch": 2.976731503363025, + "grad_norm": 0.12368189543485641, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0314, + "step": 32750 + }, + { + "epoch": 2.9776404290129066, + "grad_norm": 0.19842824339866638, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0322, + "step": 32760 + }, + { + "epoch": 2.9785493546627886, + "grad_norm": 0.27907004952430725, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0335, + "step": 32770 + }, + { + "epoch": 2.9794582803126706, + "grad_norm": 0.13934575021266937, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0286, + "step": 32780 + }, + { + "epoch": 2.980367205962552, + "grad_norm": 0.14683467149734497, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0312, + "step": 32790 + }, + { + "epoch": 2.981276131612434, + "grad_norm": 0.12826360762119293, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.028, + "step": 32800 + }, + { + "epoch": 2.982185057262316, + "grad_norm": 0.18100851774215698, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.0309, + "step": 32810 + }, + { + "epoch": 2.9830939829121976, + "grad_norm": 0.14578503370285034, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0298, + "step": 32820 + }, + { + "epoch": 2.9840029085620796, + "grad_norm": 0.18544016778469086, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.027, + "step": 32830 + }, + { + "epoch": 2.9849118342119616, + "grad_norm": 0.16878563165664673, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.032, + "step": 32840 + }, + { + "epoch": 2.985820759861843, + "grad_norm": 0.15695378184318542, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0303, + "step": 32850 + }, + { + "epoch": 2.986729685511725, + "grad_norm": 0.16257241368293762, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0272, + "step": 32860 + }, + { + "epoch": 2.987638611161607, + "grad_norm": 0.23503166437149048, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0311, + "step": 32870 + }, + { + "epoch": 2.9885475368114887, + "grad_norm": 0.13640575110912323, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0298, + "step": 32880 + }, + { + "epoch": 2.9894564624613706, + "grad_norm": 0.1486557126045227, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0304, + "step": 32890 + }, + { + "epoch": 2.9903653881112526, + "grad_norm": 0.12453336268663406, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0304, + "step": 32900 + }, + { + "epoch": 2.991274313761134, + "grad_norm": 0.1419551819562912, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0303, + "step": 32910 + }, + { + "epoch": 2.992183239411016, + "grad_norm": 0.14127813279628754, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.028, + "step": 32920 + }, + { + "epoch": 2.993092165060898, + "grad_norm": 0.12124686688184738, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0308, + "step": 32930 + }, + { + "epoch": 2.99400109071078, + "grad_norm": 0.13692444562911987, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0288, + "step": 32940 + }, + { + "epoch": 2.9949100163606617, + "grad_norm": 0.12293092161417007, + "learning_rate": 2.605012603001793e-05, + "loss": 0.032, + "step": 32950 + }, + { + "epoch": 2.9958189420105437, + "grad_norm": 0.13846617937088013, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0311, + "step": 32960 + }, + { + "epoch": 2.996727867660425, + "grad_norm": 0.1358357071876526, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0321, + "step": 32970 + }, + { + "epoch": 2.997636793310307, + "grad_norm": 0.18062706291675568, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0255, + "step": 32980 + }, + { + "epoch": 2.998545718960189, + "grad_norm": 0.14824077486991882, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0294, + "step": 32990 + }, + { + "epoch": 2.999454644610071, + "grad_norm": 0.13459840416908264, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0299, + "step": 33000 + }, + { + "epoch": 3.0003635702599527, + "grad_norm": 0.16913282871246338, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0307, + "step": 33010 + }, + { + "epoch": 3.0012724959098347, + "grad_norm": 0.18331322073936462, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0275, + "step": 33020 + }, + { + "epoch": 3.0021814215597162, + "grad_norm": 0.17031128704547882, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0305, + "step": 33030 + }, + { + "epoch": 3.003090347209598, + "grad_norm": 0.15672330558300018, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0295, + "step": 33040 + }, + { + "epoch": 3.00399927285948, + "grad_norm": 0.28101181983947754, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0297, + "step": 33050 + }, + { + "epoch": 3.0049081985093617, + "grad_norm": 0.15551745891571045, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0282, + "step": 33060 + }, + { + "epoch": 3.0058171241592437, + "grad_norm": 0.17307831346988678, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0287, + "step": 33070 + }, + { + "epoch": 3.0067260498091257, + "grad_norm": 0.13855434954166412, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0303, + "step": 33080 + }, + { + "epoch": 3.0076349754590073, + "grad_norm": 0.12340788543224335, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0305, + "step": 33090 + }, + { + "epoch": 3.0085439011088893, + "grad_norm": 0.149302676320076, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0307, + "step": 33100 + }, + { + "epoch": 3.0094528267587712, + "grad_norm": 0.14386704564094543, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0292, + "step": 33110 + }, + { + "epoch": 3.010361752408653, + "grad_norm": 0.14223332703113556, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0259, + "step": 33120 + }, + { + "epoch": 3.0112706780585348, + "grad_norm": 0.11758385598659515, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0296, + "step": 33130 + }, + { + "epoch": 3.0121796037084168, + "grad_norm": 0.10358021408319473, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0269, + "step": 33140 + }, + { + "epoch": 3.0130885293582983, + "grad_norm": 0.13128355145454407, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0303, + "step": 33150 + }, + { + "epoch": 3.0139974550081803, + "grad_norm": 0.1059117391705513, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0262, + "step": 33160 + }, + { + "epoch": 3.0149063806580623, + "grad_norm": 0.20521818101406097, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0292, + "step": 33170 + }, + { + "epoch": 3.015815306307944, + "grad_norm": 0.17147202789783478, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0288, + "step": 33180 + }, + { + "epoch": 3.016724231957826, + "grad_norm": 0.26754122972488403, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0306, + "step": 33190 + }, + { + "epoch": 3.017633157607708, + "grad_norm": 0.11486552655696869, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0286, + "step": 33200 + }, + { + "epoch": 3.0185420832575893, + "grad_norm": 0.16891682147979736, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0323, + "step": 33210 + }, + { + "epoch": 3.0194510089074713, + "grad_norm": 0.1480226367712021, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0304, + "step": 33220 + }, + { + "epoch": 3.0203599345573533, + "grad_norm": 0.11404646188020706, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0325, + "step": 33230 + }, + { + "epoch": 3.021268860207235, + "grad_norm": 0.1444234699010849, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0335, + "step": 33240 + }, + { + "epoch": 3.022177785857117, + "grad_norm": 0.12192333489656448, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0316, + "step": 33250 + }, + { + "epoch": 3.023086711506999, + "grad_norm": 0.14024248719215393, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0316, + "step": 33260 + }, + { + "epoch": 3.0239956371568804, + "grad_norm": 0.16889376938343048, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0291, + "step": 33270 + }, + { + "epoch": 3.0249045628067623, + "grad_norm": 0.14161574840545654, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0267, + "step": 33280 + }, + { + "epoch": 3.0258134884566443, + "grad_norm": 0.17235377430915833, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0306, + "step": 33290 + }, + { + "epoch": 3.026722414106526, + "grad_norm": 0.14365911483764648, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.029, + "step": 33300 + }, + { + "epoch": 3.027631339756408, + "grad_norm": 0.1652815192937851, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0306, + "step": 33310 + }, + { + "epoch": 3.02854026540629, + "grad_norm": 0.13731621205806732, + "learning_rate": 2.594612417857538e-05, + "loss": 0.03, + "step": 33320 + }, + { + "epoch": 3.0294491910561714, + "grad_norm": 0.13884128630161285, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0282, + "step": 33330 + }, + { + "epoch": 3.0303581167060534, + "grad_norm": 0.12765926122665405, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0282, + "step": 33340 + }, + { + "epoch": 3.0312670423559354, + "grad_norm": 0.13806882500648499, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0344, + "step": 33350 + }, + { + "epoch": 3.0321759680058173, + "grad_norm": 0.16734150052070618, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0294, + "step": 33360 + }, + { + "epoch": 3.033084893655699, + "grad_norm": 0.10798624902963638, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0276, + "step": 33370 + }, + { + "epoch": 3.033993819305581, + "grad_norm": 0.1621035784482956, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0283, + "step": 33380 + }, + { + "epoch": 3.034902744955463, + "grad_norm": 0.12044645100831985, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0278, + "step": 33390 + }, + { + "epoch": 3.0358116706053444, + "grad_norm": 0.12238098680973053, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0286, + "step": 33400 + }, + { + "epoch": 3.0367205962552264, + "grad_norm": 0.19144657254219055, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0344, + "step": 33410 + }, + { + "epoch": 3.0376295219051084, + "grad_norm": 0.14648255705833435, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0303, + "step": 33420 + }, + { + "epoch": 3.03853844755499, + "grad_norm": 0.14034220576286316, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0283, + "step": 33430 + }, + { + "epoch": 3.039447373204872, + "grad_norm": 0.16292691230773926, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0343, + "step": 33440 + }, + { + "epoch": 3.040356298854754, + "grad_norm": 0.14737877249717712, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0307, + "step": 33450 + }, + { + "epoch": 3.0412652245046354, + "grad_norm": 0.1358979046344757, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0283, + "step": 33460 + }, + { + "epoch": 3.0421741501545174, + "grad_norm": 0.14161434769630432, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.031, + "step": 33470 + }, + { + "epoch": 3.0430830758043994, + "grad_norm": 0.1562948077917099, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0333, + "step": 33480 + }, + { + "epoch": 3.043992001454281, + "grad_norm": 0.15541906654834747, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0316, + "step": 33490 + }, + { + "epoch": 3.044900927104163, + "grad_norm": 0.10353478044271469, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0313, + "step": 33500 + }, + { + "epoch": 3.045809852754045, + "grad_norm": 0.17537818849086761, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0332, + "step": 33510 + }, + { + "epoch": 3.0467187784039265, + "grad_norm": 0.13814190030097961, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0312, + "step": 33520 + }, + { + "epoch": 3.0476277040538085, + "grad_norm": 0.2126212865114212, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0323, + "step": 33530 + }, + { + "epoch": 3.0485366297036904, + "grad_norm": 0.14325568079948425, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0328, + "step": 33540 + }, + { + "epoch": 3.049445555353572, + "grad_norm": 0.13847313821315765, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0311, + "step": 33550 + }, + { + "epoch": 3.050354481003454, + "grad_norm": 0.1872251033782959, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0302, + "step": 33560 + }, + { + "epoch": 3.051263406653336, + "grad_norm": 0.1615745574235916, + "learning_rate": 2.587520774974495e-05, + "loss": 0.031, + "step": 33570 + }, + { + "epoch": 3.0521723323032175, + "grad_norm": 0.15097764134407043, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.033, + "step": 33580 + }, + { + "epoch": 3.0530812579530995, + "grad_norm": 0.16257570683956146, + "learning_rate": 2.586951209145751e-05, + "loss": 0.031, + "step": 33590 + }, + { + "epoch": 3.0539901836029815, + "grad_norm": 0.15522105991840363, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0282, + "step": 33600 + }, + { + "epoch": 3.054899109252863, + "grad_norm": 0.14627303183078766, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0283, + "step": 33610 + }, + { + "epoch": 3.055808034902745, + "grad_norm": 0.14067082107067108, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0296, + "step": 33620 + }, + { + "epoch": 3.056716960552627, + "grad_norm": 0.13189920783042908, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0324, + "step": 33630 + }, + { + "epoch": 3.0576258862025085, + "grad_norm": 0.15227264165878296, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0311, + "step": 33640 + }, + { + "epoch": 3.0585348118523905, + "grad_norm": 0.18272271752357483, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0341, + "step": 33650 + }, + { + "epoch": 3.0594437375022725, + "grad_norm": 0.15465031564235687, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.0286, + "step": 33660 + }, + { + "epoch": 3.060352663152154, + "grad_norm": 0.14646399021148682, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0321, + "step": 33670 + }, + { + "epoch": 3.061261588802036, + "grad_norm": 0.15463997423648834, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0309, + "step": 33680 + }, + { + "epoch": 3.062170514451918, + "grad_norm": 0.0965825691819191, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0304, + "step": 33690 + }, + { + "epoch": 3.0630794401017996, + "grad_norm": 0.1776861548423767, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0336, + "step": 33700 + }, + { + "epoch": 3.0639883657516815, + "grad_norm": 0.157448872923851, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0274, + "step": 33710 + }, + { + "epoch": 3.0648972914015635, + "grad_norm": 0.14861589670181274, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0284, + "step": 33720 + }, + { + "epoch": 3.065806217051445, + "grad_norm": 0.15384656190872192, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0286, + "step": 33730 + }, + { + "epoch": 3.066715142701327, + "grad_norm": 0.15324550867080688, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0283, + "step": 33740 + }, + { + "epoch": 3.067624068351209, + "grad_norm": 0.1631994992494583, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0294, + "step": 33750 + }, + { + "epoch": 3.0685329940010906, + "grad_norm": 0.16479787230491638, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0306, + "step": 33760 + }, + { + "epoch": 3.0694419196509726, + "grad_norm": 0.14924417436122894, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0281, + "step": 33770 + }, + { + "epoch": 3.0703508453008546, + "grad_norm": 0.14943107962608337, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0313, + "step": 33780 + }, + { + "epoch": 3.071259770950736, + "grad_norm": 0.12099345773458481, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0306, + "step": 33790 + }, + { + "epoch": 3.072168696600618, + "grad_norm": 0.13436409831047058, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0305, + "step": 33800 + }, + { + "epoch": 3.0730776222505, + "grad_norm": 0.11773775517940521, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.03, + "step": 33810 + }, + { + "epoch": 3.0739865479003816, + "grad_norm": 0.16571898758411407, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0291, + "step": 33820 + }, + { + "epoch": 3.0748954735502636, + "grad_norm": 0.18985968828201294, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0306, + "step": 33830 + }, + { + "epoch": 3.0758043992001456, + "grad_norm": 0.176985502243042, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0307, + "step": 33840 + }, + { + "epoch": 3.076713324850027, + "grad_norm": 0.1653837412595749, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0296, + "step": 33850 + }, + { + "epoch": 3.077622250499909, + "grad_norm": 0.09935039281845093, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0307, + "step": 33860 + }, + { + "epoch": 3.078531176149791, + "grad_norm": 0.14128489792346954, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0321, + "step": 33870 + }, + { + "epoch": 3.0794401017996726, + "grad_norm": 0.12325926870107651, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0294, + "step": 33880 + }, + { + "epoch": 3.0803490274495546, + "grad_norm": 0.15179768204689026, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0322, + "step": 33890 + }, + { + "epoch": 3.0812579530994366, + "grad_norm": 0.14796137809753418, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0334, + "step": 33900 + }, + { + "epoch": 3.082166878749318, + "grad_norm": 0.08995889127254486, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0303, + "step": 33910 + }, + { + "epoch": 3.0830758043992, + "grad_norm": 0.13888351619243622, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0284, + "step": 33920 + }, + { + "epoch": 3.083984730049082, + "grad_norm": 0.15156415104866028, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0295, + "step": 33930 + }, + { + "epoch": 3.0848936556989637, + "grad_norm": 0.15616574883460999, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0312, + "step": 33940 + }, + { + "epoch": 3.0858025813488457, + "grad_norm": 0.19692642986774445, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0295, + "step": 33950 + }, + { + "epoch": 3.0867115069987276, + "grad_norm": 0.1550079584121704, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0311, + "step": 33960 + }, + { + "epoch": 3.087620432648609, + "grad_norm": 0.12710034847259521, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0276, + "step": 33970 + }, + { + "epoch": 3.088529358298491, + "grad_norm": 0.15800830721855164, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0302, + "step": 33980 + }, + { + "epoch": 3.089438283948373, + "grad_norm": 0.1732550710439682, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0281, + "step": 33990 + }, + { + "epoch": 3.0903472095982547, + "grad_norm": 0.1194148063659668, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0277, + "step": 34000 + }, + { + "epoch": 3.0912561352481367, + "grad_norm": 0.11852142959833145, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0305, + "step": 34010 + }, + { + "epoch": 3.0921650608980187, + "grad_norm": 0.10677532851696014, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0309, + "step": 34020 + }, + { + "epoch": 3.0930739865479002, + "grad_norm": 0.1697574108839035, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0301, + "step": 34030 + }, + { + "epoch": 3.093982912197782, + "grad_norm": 0.13659337162971497, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0297, + "step": 34040 + }, + { + "epoch": 3.094891837847664, + "grad_norm": 0.17157532274723053, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0331, + "step": 34050 + }, + { + "epoch": 3.0958007634975457, + "grad_norm": 0.12204690277576447, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0299, + "step": 34060 + }, + { + "epoch": 3.0967096891474277, + "grad_norm": 0.14131978154182434, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0266, + "step": 34070 + }, + { + "epoch": 3.0976186147973097, + "grad_norm": 0.16598674654960632, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0282, + "step": 34080 + }, + { + "epoch": 3.0985275404471913, + "grad_norm": 0.15625670552253723, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0316, + "step": 34090 + }, + { + "epoch": 3.0994364660970732, + "grad_norm": 0.14761652052402496, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0283, + "step": 34100 + }, + { + "epoch": 3.1003453917469552, + "grad_norm": 0.2016219198703766, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.031, + "step": 34110 + }, + { + "epoch": 3.1012543173968368, + "grad_norm": 0.15050537884235382, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0267, + "step": 34120 + }, + { + "epoch": 3.1021632430467188, + "grad_norm": 0.15027175843715668, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0286, + "step": 34130 + }, + { + "epoch": 3.1030721686966007, + "grad_norm": 0.16769960522651672, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.033, + "step": 34140 + }, + { + "epoch": 3.1039810943464823, + "grad_norm": 0.13749700784683228, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0303, + "step": 34150 + }, + { + "epoch": 3.1048900199963643, + "grad_norm": 0.1280163824558258, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0289, + "step": 34160 + }, + { + "epoch": 3.1057989456462463, + "grad_norm": 0.21350166201591492, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0307, + "step": 34170 + }, + { + "epoch": 3.106707871296128, + "grad_norm": 0.13044677674770355, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0286, + "step": 34180 + }, + { + "epoch": 3.10761679694601, + "grad_norm": 0.15386231243610382, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0291, + "step": 34190 + }, + { + "epoch": 3.1085257225958918, + "grad_norm": 0.12489689886569977, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0314, + "step": 34200 + }, + { + "epoch": 3.1094346482457733, + "grad_norm": 0.14790964126586914, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0309, + "step": 34210 + }, + { + "epoch": 3.1103435738956553, + "grad_norm": 0.1452818363904953, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0289, + "step": 34220 + }, + { + "epoch": 3.1112524995455373, + "grad_norm": 0.1410800665616989, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0301, + "step": 34230 + }, + { + "epoch": 3.112161425195419, + "grad_norm": 0.17457665503025055, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0279, + "step": 34240 + }, + { + "epoch": 3.113070350845301, + "grad_norm": 0.15784338116645813, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0306, + "step": 34250 + }, + { + "epoch": 3.113979276495183, + "grad_norm": 0.15388409793376923, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0302, + "step": 34260 + }, + { + "epoch": 3.1148882021450643, + "grad_norm": 0.20771926641464233, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0285, + "step": 34270 + }, + { + "epoch": 3.1157971277949463, + "grad_norm": 0.15381674468517303, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0303, + "step": 34280 + }, + { + "epoch": 3.1167060534448283, + "grad_norm": 0.12173410505056381, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.027, + "step": 34290 + }, + { + "epoch": 3.11761497909471, + "grad_norm": 0.16884729266166687, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.031, + "step": 34300 + }, + { + "epoch": 3.118523904744592, + "grad_norm": 0.1374787837266922, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0317, + "step": 34310 + }, + { + "epoch": 3.119432830394474, + "grad_norm": 0.13595391809940338, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0303, + "step": 34320 + }, + { + "epoch": 3.1203417560443554, + "grad_norm": 0.16683916747570038, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0316, + "step": 34330 + }, + { + "epoch": 3.1212506816942374, + "grad_norm": 0.12968122959136963, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0329, + "step": 34340 + }, + { + "epoch": 3.1221596073441193, + "grad_norm": 0.15933911502361298, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0326, + "step": 34350 + }, + { + "epoch": 3.123068532994001, + "grad_norm": 0.13337862491607666, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0302, + "step": 34360 + }, + { + "epoch": 3.123977458643883, + "grad_norm": 0.1472354382276535, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0295, + "step": 34370 + }, + { + "epoch": 3.124886384293765, + "grad_norm": 0.17934538424015045, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0305, + "step": 34380 + }, + { + "epoch": 3.1257953099436464, + "grad_norm": 0.14972898364067078, + "learning_rate": 2.563899678157095e-05, + "loss": 0.029, + "step": 34390 + }, + { + "epoch": 3.1267042355935284, + "grad_norm": 0.17093640565872192, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.029, + "step": 34400 + }, + { + "epoch": 3.1276131612434104, + "grad_norm": 0.154061421751976, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0328, + "step": 34410 + }, + { + "epoch": 3.128522086893292, + "grad_norm": 0.12065830081701279, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0283, + "step": 34420 + }, + { + "epoch": 3.129431012543174, + "grad_norm": 0.11827414482831955, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0264, + "step": 34430 + }, + { + "epoch": 3.130339938193056, + "grad_norm": 0.15986812114715576, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0304, + "step": 34440 + }, + { + "epoch": 3.131248863842938, + "grad_norm": 0.13503094017505646, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0295, + "step": 34450 + }, + { + "epoch": 3.1321577894928194, + "grad_norm": 0.1679171919822693, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0293, + "step": 34460 + }, + { + "epoch": 3.1330667151427014, + "grad_norm": 0.147881880402565, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0295, + "step": 34470 + }, + { + "epoch": 3.133975640792583, + "grad_norm": 0.1583574414253235, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0304, + "step": 34480 + }, + { + "epoch": 3.134884566442465, + "grad_norm": 0.17375950515270233, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0305, + "step": 34490 + }, + { + "epoch": 3.135793492092347, + "grad_norm": 0.11121977865695953, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0322, + "step": 34500 + }, + { + "epoch": 3.136702417742229, + "grad_norm": 0.12462025135755539, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0307, + "step": 34510 + }, + { + "epoch": 3.1376113433921105, + "grad_norm": 0.15829677879810333, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.027, + "step": 34520 + }, + { + "epoch": 3.1385202690419924, + "grad_norm": 0.20119445025920868, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0306, + "step": 34530 + }, + { + "epoch": 3.139429194691874, + "grad_norm": 0.18415844440460205, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0326, + "step": 34540 + }, + { + "epoch": 3.140338120341756, + "grad_norm": 0.1472216248512268, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.031, + "step": 34550 + }, + { + "epoch": 3.141247045991638, + "grad_norm": 0.19620612263679504, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0316, + "step": 34560 + }, + { + "epoch": 3.14215597164152, + "grad_norm": 0.15598170459270477, + "learning_rate": 2.558641420481684e-05, + "loss": 0.0275, + "step": 34570 + }, + { + "epoch": 3.1430648972914015, + "grad_norm": 0.2005077749490738, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.03, + "step": 34580 + }, + { + "epoch": 3.1439738229412835, + "grad_norm": 0.14712753891944885, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0342, + "step": 34590 + }, + { + "epoch": 3.144882748591165, + "grad_norm": 0.13273009657859802, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0306, + "step": 34600 + }, + { + "epoch": 3.145791674241047, + "grad_norm": 0.15374626219272614, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0295, + "step": 34610 + }, + { + "epoch": 3.146700599890929, + "grad_norm": 0.16916382312774658, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0274, + "step": 34620 + }, + { + "epoch": 3.147609525540811, + "grad_norm": 0.13451850414276123, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0284, + "step": 34630 + }, + { + "epoch": 3.1485184511906925, + "grad_norm": 0.13241657614707947, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0295, + "step": 34640 + }, + { + "epoch": 3.1494273768405745, + "grad_norm": 0.16144941747188568, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0311, + "step": 34650 + }, + { + "epoch": 3.1503363024904565, + "grad_norm": 0.15937744081020355, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0276, + "step": 34660 + }, + { + "epoch": 3.151245228140338, + "grad_norm": 0.15714886784553528, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0289, + "step": 34670 + }, + { + "epoch": 3.15215415379022, + "grad_norm": 0.12757952511310577, + "learning_rate": 2.555415209239832e-05, + "loss": 0.029, + "step": 34680 + }, + { + "epoch": 3.153063079440102, + "grad_norm": 0.10685999691486359, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0295, + "step": 34690 + }, + { + "epoch": 3.1539720050899835, + "grad_norm": 0.1322133094072342, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0297, + "step": 34700 + }, + { + "epoch": 3.1548809307398655, + "grad_norm": 0.15169647336006165, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0275, + "step": 34710 + }, + { + "epoch": 3.1557898563897475, + "grad_norm": 0.17334288358688354, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0289, + "step": 34720 + }, + { + "epoch": 3.156698782039629, + "grad_norm": 0.11596804112195969, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0294, + "step": 34730 + }, + { + "epoch": 3.157607707689511, + "grad_norm": 0.14687132835388184, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0304, + "step": 34740 + }, + { + "epoch": 3.158516633339393, + "grad_norm": 0.16898059844970703, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0307, + "step": 34750 + }, + { + "epoch": 3.1594255589892746, + "grad_norm": 0.14569151401519775, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0325, + "step": 34760 + }, + { + "epoch": 3.1603344846391566, + "grad_norm": 0.13289977610111237, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0306, + "step": 34770 + }, + { + "epoch": 3.1612434102890385, + "grad_norm": 0.13568797707557678, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0318, + "step": 34780 + }, + { + "epoch": 3.16215233593892, + "grad_norm": 0.20224016904830933, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0311, + "step": 34790 + }, + { + "epoch": 3.163061261588802, + "grad_norm": 0.15673159062862396, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0284, + "step": 34800 + }, + { + "epoch": 3.163970187238684, + "grad_norm": 0.14222203195095062, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0316, + "step": 34810 + }, + { + "epoch": 3.1648791128885656, + "grad_norm": 0.2195242941379547, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0287, + "step": 34820 + }, + { + "epoch": 3.1657880385384476, + "grad_norm": 0.19941943883895874, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0284, + "step": 34830 + }, + { + "epoch": 3.1666969641883296, + "grad_norm": 0.1527346670627594, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0279, + "step": 34840 + }, + { + "epoch": 3.167605889838211, + "grad_norm": 0.1658296436071396, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0301, + "step": 34850 + }, + { + "epoch": 3.168514815488093, + "grad_norm": 0.12076205015182495, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0297, + "step": 34860 + }, + { + "epoch": 3.169423741137975, + "grad_norm": 0.1260419487953186, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0292, + "step": 34870 + }, + { + "epoch": 3.1703326667878566, + "grad_norm": 0.16197842359542847, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0299, + "step": 34880 + }, + { + "epoch": 3.1712415924377386, + "grad_norm": 0.13209685683250427, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0301, + "step": 34890 + }, + { + "epoch": 3.1721505180876206, + "grad_norm": 0.1390591561794281, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.029, + "step": 34900 + }, + { + "epoch": 3.173059443737502, + "grad_norm": 0.14219768345355988, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0298, + "step": 34910 + }, + { + "epoch": 3.173968369387384, + "grad_norm": 0.1266009658575058, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0278, + "step": 34920 + }, + { + "epoch": 3.174877295037266, + "grad_norm": 0.1307084709405899, + "learning_rate": 2.548046889595952e-05, + "loss": 0.03, + "step": 34930 + }, + { + "epoch": 3.1757862206871477, + "grad_norm": 0.11636979877948761, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.03, + "step": 34940 + }, + { + "epoch": 3.1766951463370297, + "grad_norm": 0.11533499509096146, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0301, + "step": 34950 + }, + { + "epoch": 3.1776040719869116, + "grad_norm": 0.13479353487491608, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0282, + "step": 34960 + }, + { + "epoch": 3.178512997636793, + "grad_norm": 0.14280883967876434, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0306, + "step": 34970 + }, + { + "epoch": 3.179421923286675, + "grad_norm": 0.15173950791358948, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0282, + "step": 34980 + }, + { + "epoch": 3.180330848936557, + "grad_norm": 0.11474179476499557, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.03, + "step": 34990 + }, + { + "epoch": 3.1812397745864387, + "grad_norm": 0.1365828514099121, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0301, + "step": 35000 + }, + { + "epoch": 3.1821487002363207, + "grad_norm": 0.12417672574520111, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0261, + "step": 35010 + }, + { + "epoch": 3.1830576258862027, + "grad_norm": 0.14581607282161713, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0281, + "step": 35020 + }, + { + "epoch": 3.183966551536084, + "grad_norm": 0.17289747297763824, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0319, + "step": 35030 + }, + { + "epoch": 3.184875477185966, + "grad_norm": 0.12714646756649017, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0246, + "step": 35040 + }, + { + "epoch": 3.185784402835848, + "grad_norm": 0.13046695291996002, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0299, + "step": 35050 + }, + { + "epoch": 3.1866933284857297, + "grad_norm": 0.1732548624277115, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0294, + "step": 35060 + }, + { + "epoch": 3.1876022541356117, + "grad_norm": 0.14522883296012878, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0299, + "step": 35070 + }, + { + "epoch": 3.1885111797854937, + "grad_norm": 0.13864849507808685, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0309, + "step": 35080 + }, + { + "epoch": 3.1894201054353752, + "grad_norm": 0.11911217123270035, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0284, + "step": 35090 + }, + { + "epoch": 3.1903290310852572, + "grad_norm": 0.14447356760501862, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0289, + "step": 35100 + }, + { + "epoch": 3.191237956735139, + "grad_norm": 0.11324875056743622, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0289, + "step": 35110 + }, + { + "epoch": 3.1921468823850208, + "grad_norm": 0.14829932153224945, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0285, + "step": 35120 + }, + { + "epoch": 3.1930558080349027, + "grad_norm": 0.17186035215854645, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0272, + "step": 35130 + }, + { + "epoch": 3.1939647336847847, + "grad_norm": 0.12889574468135834, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0296, + "step": 35140 + }, + { + "epoch": 3.1948736593346663, + "grad_norm": 0.1594291627407074, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0293, + "step": 35150 + }, + { + "epoch": 3.1957825849845483, + "grad_norm": 0.15994685888290405, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0302, + "step": 35160 + }, + { + "epoch": 3.1966915106344302, + "grad_norm": 0.15517640113830566, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0262, + "step": 35170 + }, + { + "epoch": 3.197600436284312, + "grad_norm": 0.12336135655641556, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0271, + "step": 35180 + }, + { + "epoch": 3.1985093619341938, + "grad_norm": 0.1171584501862526, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0259, + "step": 35190 + }, + { + "epoch": 3.1994182875840758, + "grad_norm": 0.1153886690735817, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0294, + "step": 35200 + }, + { + "epoch": 3.2003272132339573, + "grad_norm": 0.15669551491737366, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0331, + "step": 35210 + }, + { + "epoch": 3.2012361388838393, + "grad_norm": 0.1629725843667984, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0305, + "step": 35220 + }, + { + "epoch": 3.2021450645337213, + "grad_norm": 0.1425437480211258, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0307, + "step": 35230 + }, + { + "epoch": 3.203053990183603, + "grad_norm": 0.18351098895072937, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0315, + "step": 35240 + }, + { + "epoch": 3.203962915833485, + "grad_norm": 0.1625269055366516, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0284, + "step": 35250 + }, + { + "epoch": 3.204871841483367, + "grad_norm": 0.14416393637657166, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0284, + "step": 35260 + }, + { + "epoch": 3.2057807671332483, + "grad_norm": 0.1754370629787445, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0273, + "step": 35270 + }, + { + "epoch": 3.2066896927831303, + "grad_norm": 0.15456630289554596, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0315, + "step": 35280 + }, + { + "epoch": 3.2075986184330123, + "grad_norm": 0.1595781445503235, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0302, + "step": 35290 + }, + { + "epoch": 3.208507544082894, + "grad_norm": 0.1638202965259552, + "learning_rate": 2.537050587875249e-05, + "loss": 0.03, + "step": 35300 + }, + { + "epoch": 3.209416469732776, + "grad_norm": 0.1507560759782791, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0263, + "step": 35310 + }, + { + "epoch": 3.210325395382658, + "grad_norm": 0.13901354372501373, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0291, + "step": 35320 + }, + { + "epoch": 3.2112343210325394, + "grad_norm": 0.15057802200317383, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0321, + "step": 35330 + }, + { + "epoch": 3.2121432466824213, + "grad_norm": 0.17928546667099, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0285, + "step": 35340 + }, + { + "epoch": 3.2130521723323033, + "grad_norm": 0.14172734320163727, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0304, + "step": 35350 + }, + { + "epoch": 3.213961097982185, + "grad_norm": 0.14630746841430664, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0295, + "step": 35360 + }, + { + "epoch": 3.214870023632067, + "grad_norm": 0.14330226182937622, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0298, + "step": 35370 + }, + { + "epoch": 3.215778949281949, + "grad_norm": 0.11680649220943451, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.027, + "step": 35380 + }, + { + "epoch": 3.2166878749318304, + "grad_norm": 0.1629396677017212, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.03, + "step": 35390 + }, + { + "epoch": 3.2175968005817124, + "grad_norm": 0.12460025399923325, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0319, + "step": 35400 + }, + { + "epoch": 3.2185057262315944, + "grad_norm": 0.15280790627002716, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0312, + "step": 35410 + }, + { + "epoch": 3.219414651881476, + "grad_norm": 0.24013751745224, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0329, + "step": 35420 + }, + { + "epoch": 3.220323577531358, + "grad_norm": 0.16501598060131073, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0291, + "step": 35430 + }, + { + "epoch": 3.22123250318124, + "grad_norm": 0.18799938261508942, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0293, + "step": 35440 + }, + { + "epoch": 3.2221414288311214, + "grad_norm": 0.14870667457580566, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0265, + "step": 35450 + }, + { + "epoch": 3.2230503544810034, + "grad_norm": 0.14456361532211304, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0264, + "step": 35460 + }, + { + "epoch": 3.2239592801308854, + "grad_norm": 0.11334704607725143, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0295, + "step": 35470 + }, + { + "epoch": 3.224868205780767, + "grad_norm": 0.14839884638786316, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0269, + "step": 35480 + }, + { + "epoch": 3.225777131430649, + "grad_norm": 0.15064729750156403, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0285, + "step": 35490 + }, + { + "epoch": 3.226686057080531, + "grad_norm": 0.1281290203332901, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0276, + "step": 35500 + }, + { + "epoch": 3.2275949827304125, + "grad_norm": 0.15926511585712433, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0286, + "step": 35510 + }, + { + "epoch": 3.2285039083802944, + "grad_norm": 0.16102661192417145, + "learning_rate": 2.530461059658129e-05, + "loss": 0.032, + "step": 35520 + }, + { + "epoch": 3.2294128340301764, + "grad_norm": 0.17410169541835785, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.029, + "step": 35530 + }, + { + "epoch": 3.230321759680058, + "grad_norm": 0.18165507912635803, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0286, + "step": 35540 + }, + { + "epoch": 3.23123068532994, + "grad_norm": 0.14935070276260376, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0294, + "step": 35550 + }, + { + "epoch": 3.232139610979822, + "grad_norm": 0.1409977823495865, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0334, + "step": 35560 + }, + { + "epoch": 3.2330485366297035, + "grad_norm": 0.1305711567401886, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0298, + "step": 35570 + }, + { + "epoch": 3.2339574622795855, + "grad_norm": 0.15965636074543, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0311, + "step": 35580 + }, + { + "epoch": 3.2348663879294675, + "grad_norm": 0.13062027096748352, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0268, + "step": 35590 + }, + { + "epoch": 3.2357753135793494, + "grad_norm": 0.23420006036758423, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.03, + "step": 35600 + }, + { + "epoch": 3.236684239229231, + "grad_norm": 0.12744209170341492, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0304, + "step": 35610 + }, + { + "epoch": 3.237593164879113, + "grad_norm": 0.15093468129634857, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0279, + "step": 35620 + }, + { + "epoch": 3.2385020905289945, + "grad_norm": 0.1705586016178131, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0286, + "step": 35630 + }, + { + "epoch": 3.2394110161788765, + "grad_norm": 0.1593538075685501, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0279, + "step": 35640 + }, + { + "epoch": 3.2403199418287585, + "grad_norm": 0.16656652092933655, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0297, + "step": 35650 + }, + { + "epoch": 3.2412288674786405, + "grad_norm": 0.12227635830640793, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0315, + "step": 35660 + }, + { + "epoch": 3.242137793128522, + "grad_norm": 0.13777220249176025, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.029, + "step": 35670 + }, + { + "epoch": 3.243046718778404, + "grad_norm": 0.17615863680839539, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0297, + "step": 35680 + }, + { + "epoch": 3.2439556444282855, + "grad_norm": 0.14822502434253693, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0305, + "step": 35690 + }, + { + "epoch": 3.2448645700781675, + "grad_norm": 0.1366381198167801, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.0279, + "step": 35700 + }, + { + "epoch": 3.2457734957280495, + "grad_norm": 0.15755406022071838, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0294, + "step": 35710 + }, + { + "epoch": 3.2466824213779315, + "grad_norm": 0.18377180397510529, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.028, + "step": 35720 + }, + { + "epoch": 3.247591347027813, + "grad_norm": 0.13352207839488983, + "learning_rate": 2.524135719213194e-05, + "loss": 0.0292, + "step": 35730 + }, + { + "epoch": 3.248500272677695, + "grad_norm": 0.12717708945274353, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0283, + "step": 35740 + }, + { + "epoch": 3.2494091983275766, + "grad_norm": 0.1508179008960724, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0274, + "step": 35750 + }, + { + "epoch": 3.2503181239774586, + "grad_norm": 0.14521059393882751, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0281, + "step": 35760 + }, + { + "epoch": 3.2512270496273405, + "grad_norm": 0.1605907827615738, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0289, + "step": 35770 + }, + { + "epoch": 3.2521359752772225, + "grad_norm": 0.15790635347366333, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0253, + "step": 35780 + }, + { + "epoch": 3.253044900927104, + "grad_norm": 0.1797802597284317, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0253, + "step": 35790 + }, + { + "epoch": 3.253953826576986, + "grad_norm": 0.1320931315422058, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0271, + "step": 35800 + }, + { + "epoch": 3.2548627522268676, + "grad_norm": 0.20099574327468872, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0272, + "step": 35810 + }, + { + "epoch": 3.2557716778767496, + "grad_norm": 0.18667416274547577, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0301, + "step": 35820 + }, + { + "epoch": 3.2566806035266316, + "grad_norm": 0.1187349334359169, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0294, + "step": 35830 + }, + { + "epoch": 3.2575895291765136, + "grad_norm": 0.23372915387153625, + "learning_rate": 2.520808743581702e-05, + "loss": 0.027, + "step": 35840 + }, + { + "epoch": 3.258498454826395, + "grad_norm": 0.14508765935897827, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0291, + "step": 35850 + }, + { + "epoch": 3.259407380476277, + "grad_norm": 0.11868996173143387, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0277, + "step": 35860 + }, + { + "epoch": 3.2603163061261586, + "grad_norm": 0.14636372029781342, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0307, + "step": 35870 + }, + { + "epoch": 3.2612252317760406, + "grad_norm": 0.15156316757202148, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.026, + "step": 35880 + }, + { + "epoch": 3.2621341574259226, + "grad_norm": 0.14559686183929443, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0286, + "step": 35890 + }, + { + "epoch": 3.2630430830758046, + "grad_norm": 0.14735911786556244, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0279, + "step": 35900 + }, + { + "epoch": 3.263952008725686, + "grad_norm": 0.15665121376514435, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0294, + "step": 35910 + }, + { + "epoch": 3.264860934375568, + "grad_norm": 0.11485953629016876, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0308, + "step": 35920 + }, + { + "epoch": 3.26576986002545, + "grad_norm": 0.14822711050510406, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0268, + "step": 35930 + }, + { + "epoch": 3.2666787856753317, + "grad_norm": 0.1786378026008606, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.027, + "step": 35940 + }, + { + "epoch": 3.2675877113252136, + "grad_norm": 0.1485966295003891, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0303, + "step": 35950 + }, + { + "epoch": 3.2684966369750956, + "grad_norm": 0.13004252314567566, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0268, + "step": 35960 + }, + { + "epoch": 3.269405562624977, + "grad_norm": 0.1471886932849884, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0308, + "step": 35970 + }, + { + "epoch": 3.270314488274859, + "grad_norm": 0.1788729429244995, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0327, + "step": 35980 + }, + { + "epoch": 3.271223413924741, + "grad_norm": 0.12974034249782562, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0286, + "step": 35990 + }, + { + "epoch": 3.2721323395746227, + "grad_norm": 0.13675330579280853, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0325, + "step": 36000 + }, + { + "epoch": 3.2730412652245047, + "grad_norm": 0.12441583722829819, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0279, + "step": 36010 + }, + { + "epoch": 3.2739501908743867, + "grad_norm": 0.13953478634357452, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0289, + "step": 36020 + }, + { + "epoch": 3.274859116524268, + "grad_norm": 0.11579938977956772, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0273, + "step": 36030 + }, + { + "epoch": 3.27576804217415, + "grad_norm": 0.14750830829143524, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0289, + "step": 36040 + }, + { + "epoch": 3.276676967824032, + "grad_norm": 0.1295635849237442, + "learning_rate": 2.514431222506134e-05, + "loss": 0.0275, + "step": 36050 + }, + { + "epoch": 3.2775858934739137, + "grad_norm": 0.13667091727256775, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0279, + "step": 36060 + }, + { + "epoch": 3.2784948191237957, + "grad_norm": 0.19664324820041656, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0271, + "step": 36070 + }, + { + "epoch": 3.2794037447736777, + "grad_norm": 0.14735440909862518, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0307, + "step": 36080 + }, + { + "epoch": 3.2803126704235592, + "grad_norm": 0.16360604763031006, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0289, + "step": 36090 + }, + { + "epoch": 3.281221596073441, + "grad_norm": 0.13143572211265564, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0315, + "step": 36100 + }, + { + "epoch": 3.282130521723323, + "grad_norm": 0.18506500124931335, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0281, + "step": 36110 + }, + { + "epoch": 3.2830394473732047, + "grad_norm": 0.18036872148513794, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0279, + "step": 36120 + }, + { + "epoch": 3.2839483730230867, + "grad_norm": 0.14158405363559723, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0277, + "step": 36130 + }, + { + "epoch": 3.2848572986729687, + "grad_norm": 0.1166732907295227, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0271, + "step": 36140 + }, + { + "epoch": 3.2857662243228503, + "grad_norm": 0.1453104466199875, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0283, + "step": 36150 + }, + { + "epoch": 3.2866751499727322, + "grad_norm": 0.13124483823776245, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0282, + "step": 36160 + }, + { + "epoch": 3.2875840756226142, + "grad_norm": 0.11665186285972595, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.029, + "step": 36170 + }, + { + "epoch": 3.2884930012724958, + "grad_norm": 0.1330624669790268, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0284, + "step": 36180 + }, + { + "epoch": 3.2894019269223778, + "grad_norm": 0.12592630088329315, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.029, + "step": 36190 + }, + { + "epoch": 3.2903108525722597, + "grad_norm": 0.13292637467384338, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0278, + "step": 36200 + }, + { + "epoch": 3.2912197782221413, + "grad_norm": 0.1584780514240265, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0296, + "step": 36210 + }, + { + "epoch": 3.2921287038720233, + "grad_norm": 0.11009152978658676, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0291, + "step": 36220 + }, + { + "epoch": 3.2930376295219053, + "grad_norm": 0.1445978283882141, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0286, + "step": 36230 + }, + { + "epoch": 3.293946555171787, + "grad_norm": 0.1290206015110016, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0301, + "step": 36240 + }, + { + "epoch": 3.294855480821669, + "grad_norm": 0.12613633275032043, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0295, + "step": 36250 + }, + { + "epoch": 3.295764406471551, + "grad_norm": 0.11504855006933212, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0281, + "step": 36260 + }, + { + "epoch": 3.2966733321214323, + "grad_norm": 0.17869766056537628, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0311, + "step": 36270 + }, + { + "epoch": 3.2975822577713143, + "grad_norm": 0.16764505207538605, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0281, + "step": 36280 + }, + { + "epoch": 3.2984911834211963, + "grad_norm": 0.14624696969985962, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0291, + "step": 36290 + }, + { + "epoch": 3.299400109071078, + "grad_norm": 0.17885886132717133, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0302, + "step": 36300 + }, + { + "epoch": 3.30030903472096, + "grad_norm": 0.16810478270053864, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0302, + "step": 36310 + }, + { + "epoch": 3.301217960370842, + "grad_norm": 0.2578586935997009, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0279, + "step": 36320 + }, + { + "epoch": 3.3021268860207234, + "grad_norm": 0.1740657091140747, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0289, + "step": 36330 + }, + { + "epoch": 3.3030358116706053, + "grad_norm": 0.18638809025287628, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0278, + "step": 36340 + }, + { + "epoch": 3.3039447373204873, + "grad_norm": 0.11479452252388, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0309, + "step": 36350 + }, + { + "epoch": 3.304853662970369, + "grad_norm": 0.13260523974895477, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0273, + "step": 36360 + }, + { + "epoch": 3.305762588620251, + "grad_norm": 0.16043967008590698, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0306, + "step": 36370 + }, + { + "epoch": 3.306671514270133, + "grad_norm": 0.14604245126247406, + "learning_rate": 2.504340844386208e-05, + "loss": 0.031, + "step": 36380 + }, + { + "epoch": 3.3075804399200144, + "grad_norm": 0.1400817185640335, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.032, + "step": 36390 + }, + { + "epoch": 3.3084893655698964, + "grad_norm": 0.1447252333164215, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0322, + "step": 36400 + }, + { + "epoch": 3.3093982912197784, + "grad_norm": 0.14052926003932953, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0285, + "step": 36410 + }, + { + "epoch": 3.31030721686966, + "grad_norm": 0.1997174173593521, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0308, + "step": 36420 + }, + { + "epoch": 3.311216142519542, + "grad_norm": 0.18654510378837585, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0276, + "step": 36430 + }, + { + "epoch": 3.312125068169424, + "grad_norm": 0.18250498175621033, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0299, + "step": 36440 + }, + { + "epoch": 3.3130339938193054, + "grad_norm": 0.1780354231595993, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.029, + "step": 36450 + }, + { + "epoch": 3.3139429194691874, + "grad_norm": 0.15666581690311432, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0298, + "step": 36460 + }, + { + "epoch": 3.3148518451190694, + "grad_norm": 0.1559508889913559, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0278, + "step": 36470 + }, + { + "epoch": 3.315760770768951, + "grad_norm": 0.14496704936027527, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0286, + "step": 36480 + }, + { + "epoch": 3.316669696418833, + "grad_norm": 0.11769497394561768, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0316, + "step": 36490 + }, + { + "epoch": 3.317578622068715, + "grad_norm": 0.18593809008598328, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.029, + "step": 36500 + }, + { + "epoch": 3.3184875477185964, + "grad_norm": 0.11218901723623276, + "learning_rate": 2.500342994941466e-05, + "loss": 0.028, + "step": 36510 + }, + { + "epoch": 3.3193964733684784, + "grad_norm": 0.14741478860378265, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0291, + "step": 36520 + }, + { + "epoch": 3.3203053990183604, + "grad_norm": 0.1749597042798996, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0294, + "step": 36530 + }, + { + "epoch": 3.321214324668242, + "grad_norm": 0.11940786987543106, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0255, + "step": 36540 + }, + { + "epoch": 3.322123250318124, + "grad_norm": 0.1292211264371872, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0297, + "step": 36550 + }, + { + "epoch": 3.323032175968006, + "grad_norm": 0.14195680618286133, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0273, + "step": 36560 + }, + { + "epoch": 3.3239411016178875, + "grad_norm": 0.1389155089855194, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0286, + "step": 36570 + }, + { + "epoch": 3.3248500272677695, + "grad_norm": 0.16506196558475494, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0308, + "step": 36580 + }, + { + "epoch": 3.3257589529176514, + "grad_norm": 0.12435410171747208, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0321, + "step": 36590 + }, + { + "epoch": 3.326667878567533, + "grad_norm": 0.13418394327163696, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0273, + "step": 36600 + }, + { + "epoch": 3.327576804217415, + "grad_norm": 0.1022360622882843, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0249, + "step": 36610 + }, + { + "epoch": 3.328485729867297, + "grad_norm": 0.13355223834514618, + "learning_rate": 2.496950169240971e-05, + "loss": 0.031, + "step": 36620 + }, + { + "epoch": 3.3293946555171785, + "grad_norm": 0.11618020385503769, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0297, + "step": 36630 + }, + { + "epoch": 3.3303035811670605, + "grad_norm": 0.19005578756332397, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0311, + "step": 36640 + }, + { + "epoch": 3.3312125068169425, + "grad_norm": 0.1487089991569519, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0302, + "step": 36650 + }, + { + "epoch": 3.332121432466824, + "grad_norm": 0.15482819080352783, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0323, + "step": 36660 + }, + { + "epoch": 3.333030358116706, + "grad_norm": 0.1386234611272812, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0269, + "step": 36670 + }, + { + "epoch": 3.333939283766588, + "grad_norm": 0.149174302816391, + "learning_rate": 2.495095673748091e-05, + "loss": 0.029, + "step": 36680 + }, + { + "epoch": 3.33484820941647, + "grad_norm": 0.10890396684408188, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0272, + "step": 36690 + }, + { + "epoch": 3.3357571350663515, + "grad_norm": 0.1593647301197052, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0311, + "step": 36700 + }, + { + "epoch": 3.3366660607162335, + "grad_norm": 0.1222846731543541, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0303, + "step": 36710 + }, + { + "epoch": 3.337574986366115, + "grad_norm": 0.1485145539045334, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0268, + "step": 36720 + }, + { + "epoch": 3.338483912015997, + "grad_norm": 0.14770229160785675, + "learning_rate": 2.49354818242132e-05, + "loss": 0.03, + "step": 36730 + }, + { + "epoch": 3.339392837665879, + "grad_norm": 0.11509238928556442, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0276, + "step": 36740 + }, + { + "epoch": 3.340301763315761, + "grad_norm": 0.170248880982399, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0291, + "step": 36750 + }, + { + "epoch": 3.3412106889656426, + "grad_norm": 0.1646675318479538, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0281, + "step": 36760 + }, + { + "epoch": 3.3421196146155245, + "grad_norm": 0.19755393266677856, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0303, + "step": 36770 + }, + { + "epoch": 3.343028540265406, + "grad_norm": 0.12961789965629578, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.026, + "step": 36780 + }, + { + "epoch": 3.343937465915288, + "grad_norm": 0.11762965470552444, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0286, + "step": 36790 + }, + { + "epoch": 3.34484639156517, + "grad_norm": 0.14506837725639343, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0272, + "step": 36800 + }, + { + "epoch": 3.345755317215052, + "grad_norm": 0.14947384595870972, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0314, + "step": 36810 + }, + { + "epoch": 3.3466642428649336, + "grad_norm": 0.11291320621967316, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0268, + "step": 36820 + }, + { + "epoch": 3.3475731685148156, + "grad_norm": 0.15183861553668976, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0281, + "step": 36830 + }, + { + "epoch": 3.348482094164697, + "grad_norm": 0.14899419248104095, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0272, + "step": 36840 + }, + { + "epoch": 3.349391019814579, + "grad_norm": 0.13253004848957062, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.029, + "step": 36850 + }, + { + "epoch": 3.350299945464461, + "grad_norm": 0.11556883901357651, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0308, + "step": 36860 + }, + { + "epoch": 3.351208871114343, + "grad_norm": 0.17016896605491638, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0305, + "step": 36870 + }, + { + "epoch": 3.3521177967642246, + "grad_norm": 0.17199604213237762, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0298, + "step": 36880 + }, + { + "epoch": 3.3530267224141066, + "grad_norm": 0.13833731412887573, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0285, + "step": 36890 + }, + { + "epoch": 3.353935648063988, + "grad_norm": 0.1564026176929474, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0308, + "step": 36900 + }, + { + "epoch": 3.35484457371387, + "grad_norm": 0.14113390445709229, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0297, + "step": 36910 + }, + { + "epoch": 3.355753499363752, + "grad_norm": 0.15954609215259552, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0283, + "step": 36920 + }, + { + "epoch": 3.356662425013634, + "grad_norm": 0.14668065309524536, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0296, + "step": 36930 + }, + { + "epoch": 3.3575713506635156, + "grad_norm": 0.16166028380393982, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0281, + "step": 36940 + }, + { + "epoch": 3.3584802763133976, + "grad_norm": 0.17834435403347015, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0302, + "step": 36950 + }, + { + "epoch": 3.359389201963279, + "grad_norm": 0.1913929134607315, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0279, + "step": 36960 + }, + { + "epoch": 3.360298127613161, + "grad_norm": 0.11254964023828506, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0261, + "step": 36970 + }, + { + "epoch": 3.361207053263043, + "grad_norm": 0.11465303599834442, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0301, + "step": 36980 + }, + { + "epoch": 3.362115978912925, + "grad_norm": 0.10390942543745041, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0253, + "step": 36990 + }, + { + "epoch": 3.3630249045628067, + "grad_norm": 0.13481643795967102, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0279, + "step": 37000 + }, + { + "epoch": 3.3639338302126887, + "grad_norm": 0.11577078700065613, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0283, + "step": 37010 + }, + { + "epoch": 3.36484275586257, + "grad_norm": 0.1672864556312561, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0271, + "step": 37020 + }, + { + "epoch": 3.365751681512452, + "grad_norm": 0.14636249840259552, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.028, + "step": 37030 + }, + { + "epoch": 3.366660607162334, + "grad_norm": 0.1327391266822815, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0326, + "step": 37040 + }, + { + "epoch": 3.367569532812216, + "grad_norm": 0.12846198678016663, + "learning_rate": 2.483599695058918e-05, + "loss": 0.031, + "step": 37050 + }, + { + "epoch": 3.3684784584620977, + "grad_norm": 0.17908437550067902, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0316, + "step": 37060 + }, + { + "epoch": 3.3693873841119797, + "grad_norm": 0.12656597793102264, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.03, + "step": 37070 + }, + { + "epoch": 3.3702963097618612, + "grad_norm": 0.12668505311012268, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0283, + "step": 37080 + }, + { + "epoch": 3.371205235411743, + "grad_norm": 0.1374061405658722, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.028, + "step": 37090 + }, + { + "epoch": 3.372114161061625, + "grad_norm": 0.17191414535045624, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0305, + "step": 37100 + }, + { + "epoch": 3.373023086711507, + "grad_norm": 0.10738180577754974, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.026, + "step": 37110 + }, + { + "epoch": 3.3739320123613887, + "grad_norm": 0.18391864001750946, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0289, + "step": 37120 + }, + { + "epoch": 3.3748409380112707, + "grad_norm": 0.13624274730682373, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0308, + "step": 37130 + }, + { + "epoch": 3.3757498636611527, + "grad_norm": 0.19017599523067474, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0263, + "step": 37140 + }, + { + "epoch": 3.3766587893110342, + "grad_norm": 0.11768215894699097, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0284, + "step": 37150 + }, + { + "epoch": 3.3775677149609162, + "grad_norm": 0.1406077891588211, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0287, + "step": 37160 + }, + { + "epoch": 3.378476640610798, + "grad_norm": 0.15494418144226074, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0267, + "step": 37170 + }, + { + "epoch": 3.3793855662606798, + "grad_norm": 0.2165760099887848, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0281, + "step": 37180 + }, + { + "epoch": 3.3802944919105617, + "grad_norm": 0.16072309017181396, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0295, + "step": 37190 + }, + { + "epoch": 3.3812034175604437, + "grad_norm": 0.14300517737865448, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0276, + "step": 37200 + }, + { + "epoch": 3.3821123432103253, + "grad_norm": 0.14764849841594696, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0269, + "step": 37210 + }, + { + "epoch": 3.3830212688602073, + "grad_norm": 0.1552264392375946, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0285, + "step": 37220 + }, + { + "epoch": 3.3839301945100893, + "grad_norm": 0.14408521354198456, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0292, + "step": 37230 + }, + { + "epoch": 3.384839120159971, + "grad_norm": 0.1570548266172409, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0263, + "step": 37240 + }, + { + "epoch": 3.385748045809853, + "grad_norm": 0.14971497654914856, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0287, + "step": 37250 + }, + { + "epoch": 3.3866569714597348, + "grad_norm": 0.11661026626825333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0302, + "step": 37260 + }, + { + "epoch": 3.3875658971096163, + "grad_norm": 0.2178274244070053, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0298, + "step": 37270 + }, + { + "epoch": 3.3884748227594983, + "grad_norm": 0.1348731517791748, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0304, + "step": 37280 + }, + { + "epoch": 3.3893837484093803, + "grad_norm": 0.15704186260700226, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0324, + "step": 37290 + }, + { + "epoch": 3.390292674059262, + "grad_norm": 0.12964993715286255, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0299, + "step": 37300 + }, + { + "epoch": 3.391201599709144, + "grad_norm": 0.1610429286956787, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0281, + "step": 37310 + }, + { + "epoch": 3.392110525359026, + "grad_norm": 0.12610790133476257, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0283, + "step": 37320 + }, + { + "epoch": 3.3930194510089073, + "grad_norm": 0.14628875255584717, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0295, + "step": 37330 + }, + { + "epoch": 3.3939283766587893, + "grad_norm": 0.16357119381427765, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0286, + "step": 37340 + }, + { + "epoch": 3.3948373023086713, + "grad_norm": 0.13412722945213318, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0286, + "step": 37350 + }, + { + "epoch": 3.395746227958553, + "grad_norm": 0.1608828753232956, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0279, + "step": 37360 + }, + { + "epoch": 3.396655153608435, + "grad_norm": 0.15071529150009155, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0284, + "step": 37370 + }, + { + "epoch": 3.397564079258317, + "grad_norm": 0.13727906346321106, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0271, + "step": 37380 + }, + { + "epoch": 3.3984730049081984, + "grad_norm": 0.1200380027294159, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0261, + "step": 37390 + }, + { + "epoch": 3.3993819305580804, + "grad_norm": 0.1176733672618866, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0306, + "step": 37400 + }, + { + "epoch": 3.4002908562079623, + "grad_norm": 0.1625390201807022, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0297, + "step": 37410 + }, + { + "epoch": 3.401199781857844, + "grad_norm": 0.15554864704608917, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0304, + "step": 37420 + }, + { + "epoch": 3.402108707507726, + "grad_norm": 0.149881049990654, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0275, + "step": 37430 + }, + { + "epoch": 3.403017633157608, + "grad_norm": 0.14639341831207275, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0275, + "step": 37440 + }, + { + "epoch": 3.4039265588074894, + "grad_norm": 0.15311181545257568, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0314, + "step": 37450 + }, + { + "epoch": 3.4048354844573714, + "grad_norm": 0.1346728354692459, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0293, + "step": 37460 + }, + { + "epoch": 3.4057444101072534, + "grad_norm": 0.1595831960439682, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0269, + "step": 37470 + }, + { + "epoch": 3.406653335757135, + "grad_norm": 0.11618897318840027, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0248, + "step": 37480 + }, + { + "epoch": 3.407562261407017, + "grad_norm": 0.11456581950187683, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0272, + "step": 37490 + }, + { + "epoch": 3.408471187056899, + "grad_norm": 0.156296968460083, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0308, + "step": 37500 + }, + { + "epoch": 3.4093801127067804, + "grad_norm": 0.19409160315990448, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0324, + "step": 37510 + }, + { + "epoch": 3.4102890383566624, + "grad_norm": 0.14649035036563873, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0302, + "step": 37520 + }, + { + "epoch": 3.4111979640065444, + "grad_norm": 0.13230587542057037, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.03, + "step": 37530 + }, + { + "epoch": 3.412106889656426, + "grad_norm": 0.16300919651985168, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0294, + "step": 37540 + }, + { + "epoch": 3.413015815306308, + "grad_norm": 0.1509818583726883, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0266, + "step": 37550 + }, + { + "epoch": 3.41392474095619, + "grad_norm": 0.16483908891677856, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0296, + "step": 37560 + }, + { + "epoch": 3.4148336666060715, + "grad_norm": 0.13623462617397308, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0322, + "step": 37570 + }, + { + "epoch": 3.4157425922559534, + "grad_norm": 0.15584617853164673, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0291, + "step": 37580 + }, + { + "epoch": 3.4166515179058354, + "grad_norm": 0.17480607330799103, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0288, + "step": 37590 + }, + { + "epoch": 3.417560443555717, + "grad_norm": 0.1609143316745758, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0336, + "step": 37600 + }, + { + "epoch": 3.418469369205599, + "grad_norm": 0.1611342430114746, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0293, + "step": 37610 + }, + { + "epoch": 3.419378294855481, + "grad_norm": 0.1436658799648285, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0275, + "step": 37620 + }, + { + "epoch": 3.4202872205053625, + "grad_norm": 0.1649564802646637, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0298, + "step": 37630 + }, + { + "epoch": 3.4211961461552445, + "grad_norm": 0.11594291776418686, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0272, + "step": 37640 + }, + { + "epoch": 3.4221050718051265, + "grad_norm": 0.14792592823505402, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0299, + "step": 37650 + }, + { + "epoch": 3.423013997455008, + "grad_norm": 0.10570305585861206, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0291, + "step": 37660 + }, + { + "epoch": 3.42392292310489, + "grad_norm": 0.1315293163061142, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0276, + "step": 37670 + }, + { + "epoch": 3.424831848754772, + "grad_norm": 0.11728784441947937, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0267, + "step": 37680 + }, + { + "epoch": 3.4257407744046535, + "grad_norm": 0.12435539811849594, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0286, + "step": 37690 + }, + { + "epoch": 3.4266497000545355, + "grad_norm": 0.14280778169631958, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0276, + "step": 37700 + }, + { + "epoch": 3.4275586257044175, + "grad_norm": 0.12278527021408081, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0292, + "step": 37710 + }, + { + "epoch": 3.428467551354299, + "grad_norm": 0.1250564306974411, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.0269, + "step": 37720 + }, + { + "epoch": 3.429376477004181, + "grad_norm": 0.12305837869644165, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0292, + "step": 37730 + }, + { + "epoch": 3.430285402654063, + "grad_norm": 0.14355631172657013, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.032, + "step": 37740 + }, + { + "epoch": 3.4311943283039446, + "grad_norm": 0.13376787304878235, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0301, + "step": 37750 + }, + { + "epoch": 3.4321032539538265, + "grad_norm": 0.18440204858779907, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0283, + "step": 37760 + }, + { + "epoch": 3.4330121796037085, + "grad_norm": 0.16727544367313385, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0299, + "step": 37770 + }, + { + "epoch": 3.43392110525359, + "grad_norm": 0.16631074249744415, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0298, + "step": 37780 + }, + { + "epoch": 3.434830030903472, + "grad_norm": 0.16499099135398865, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0294, + "step": 37790 + }, + { + "epoch": 3.435738956553354, + "grad_norm": 0.15249110758304596, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0279, + "step": 37800 + }, + { + "epoch": 3.4366478822032356, + "grad_norm": 0.14218971133232117, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0276, + "step": 37810 + }, + { + "epoch": 3.4375568078531176, + "grad_norm": 0.16782966256141663, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0275, + "step": 37820 + }, + { + "epoch": 3.4384657335029996, + "grad_norm": 0.15302573144435883, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0301, + "step": 37830 + }, + { + "epoch": 3.4393746591528815, + "grad_norm": 0.2440381497144699, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0289, + "step": 37840 + }, + { + "epoch": 3.440283584802763, + "grad_norm": 0.21838818490505219, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.029, + "step": 37850 + }, + { + "epoch": 3.441192510452645, + "grad_norm": 0.13871778547763824, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0278, + "step": 37860 + }, + { + "epoch": 3.4421014361025266, + "grad_norm": 0.16760866343975067, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.028, + "step": 37870 + }, + { + "epoch": 3.4430103617524086, + "grad_norm": 0.16557441651821136, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0311, + "step": 37880 + }, + { + "epoch": 3.4439192874022906, + "grad_norm": 0.11778439581394196, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0266, + "step": 37890 + }, + { + "epoch": 3.4448282130521726, + "grad_norm": 0.13897882401943207, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0256, + "step": 37900 + }, + { + "epoch": 3.445737138702054, + "grad_norm": 0.1344902664422989, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0278, + "step": 37910 + }, + { + "epoch": 3.446646064351936, + "grad_norm": 0.16043514013290405, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0317, + "step": 37920 + }, + { + "epoch": 3.4475549900018176, + "grad_norm": 0.17417262494564056, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0287, + "step": 37930 + }, + { + "epoch": 3.4484639156516996, + "grad_norm": 0.14420801401138306, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0268, + "step": 37940 + }, + { + "epoch": 3.4493728413015816, + "grad_norm": 0.10738685727119446, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0286, + "step": 37950 + }, + { + "epoch": 3.4502817669514636, + "grad_norm": 0.11127075552940369, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.029, + "step": 37960 + }, + { + "epoch": 3.451190692601345, + "grad_norm": 0.12835286557674408, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0299, + "step": 37970 + }, + { + "epoch": 3.452099618251227, + "grad_norm": 0.11357732117176056, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.028, + "step": 37980 + }, + { + "epoch": 3.4530085439011087, + "grad_norm": 0.15414491295814514, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0268, + "step": 37990 + }, + { + "epoch": 3.4539174695509907, + "grad_norm": 0.10737892240285873, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.028, + "step": 38000 + }, + { + "epoch": 3.4548263952008726, + "grad_norm": 0.12717336416244507, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0257, + "step": 38010 + }, + { + "epoch": 3.4557353208507546, + "grad_norm": 0.13277605175971985, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0269, + "step": 38020 + }, + { + "epoch": 3.456644246500636, + "grad_norm": 0.14063513278961182, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0259, + "step": 38030 + }, + { + "epoch": 3.457553172150518, + "grad_norm": 0.11657757312059402, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0292, + "step": 38040 + }, + { + "epoch": 3.4584620978003997, + "grad_norm": 0.15525534749031067, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0291, + "step": 38050 + }, + { + "epoch": 3.4593710234502817, + "grad_norm": 0.13337725400924683, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0299, + "step": 38060 + }, + { + "epoch": 3.4602799491001637, + "grad_norm": 0.15917134284973145, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0282, + "step": 38070 + }, + { + "epoch": 3.4611888747500457, + "grad_norm": 0.13261741399765015, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.03, + "step": 38080 + }, + { + "epoch": 3.462097800399927, + "grad_norm": 0.16914653778076172, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0293, + "step": 38090 + }, + { + "epoch": 3.463006726049809, + "grad_norm": 0.15235339105129242, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0255, + "step": 38100 + }, + { + "epoch": 3.4639156516996907, + "grad_norm": 0.10917831212282181, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0279, + "step": 38110 + }, + { + "epoch": 3.4648245773495727, + "grad_norm": 0.14207439124584198, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0278, + "step": 38120 + }, + { + "epoch": 3.4657335029994547, + "grad_norm": 0.12297812104225159, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0296, + "step": 38130 + }, + { + "epoch": 3.4666424286493367, + "grad_norm": 0.14574764668941498, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0271, + "step": 38140 + }, + { + "epoch": 3.4675513542992182, + "grad_norm": 0.15744063258171082, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0336, + "step": 38150 + }, + { + "epoch": 3.4684602799491002, + "grad_norm": 0.13023822009563446, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0288, + "step": 38160 + }, + { + "epoch": 3.4693692055989818, + "grad_norm": 0.16444134712219238, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0281, + "step": 38170 + }, + { + "epoch": 3.4702781312488638, + "grad_norm": 0.15962521731853485, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0284, + "step": 38180 + }, + { + "epoch": 3.4711870568987457, + "grad_norm": 0.1608612984418869, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0286, + "step": 38190 + }, + { + "epoch": 3.4720959825486277, + "grad_norm": 0.16976694762706757, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.027, + "step": 38200 + }, + { + "epoch": 3.4730049081985093, + "grad_norm": 0.17507602274417877, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0305, + "step": 38210 + }, + { + "epoch": 3.4739138338483913, + "grad_norm": 0.14726074039936066, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0257, + "step": 38220 + }, + { + "epoch": 3.474822759498273, + "grad_norm": 0.1614016443490982, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0266, + "step": 38230 + }, + { + "epoch": 3.475731685148155, + "grad_norm": 0.10734491795301437, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0258, + "step": 38240 + }, + { + "epoch": 3.4766406107980368, + "grad_norm": 0.13153865933418274, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0263, + "step": 38250 + }, + { + "epoch": 3.4775495364479188, + "grad_norm": 0.18092632293701172, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0262, + "step": 38260 + }, + { + "epoch": 3.4784584620978003, + "grad_norm": 0.17797453701496124, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0307, + "step": 38270 + }, + { + "epoch": 3.4793673877476823, + "grad_norm": 0.11217096447944641, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0269, + "step": 38280 + }, + { + "epoch": 3.4802763133975643, + "grad_norm": 0.13350187242031097, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0288, + "step": 38290 + }, + { + "epoch": 3.481185239047446, + "grad_norm": 0.15860393643379211, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0309, + "step": 38300 + }, + { + "epoch": 3.482094164697328, + "grad_norm": 0.1593407541513443, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0293, + "step": 38310 + }, + { + "epoch": 3.48300309034721, + "grad_norm": 0.14919887483119965, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0295, + "step": 38320 + }, + { + "epoch": 3.4839120159970913, + "grad_norm": 0.13106662034988403, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0269, + "step": 38330 + }, + { + "epoch": 3.4848209416469733, + "grad_norm": 0.148404061794281, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0305, + "step": 38340 + }, + { + "epoch": 3.4857298672968553, + "grad_norm": 0.18516255915164948, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0272, + "step": 38350 + }, + { + "epoch": 3.486638792946737, + "grad_norm": 0.17089109122753143, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0266, + "step": 38360 + }, + { + "epoch": 3.487547718596619, + "grad_norm": 0.1639484316110611, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0281, + "step": 38370 + }, + { + "epoch": 3.488456644246501, + "grad_norm": 0.1829783320426941, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0255, + "step": 38380 + }, + { + "epoch": 3.4893655698963824, + "grad_norm": 0.1394299417734146, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0287, + "step": 38390 + }, + { + "epoch": 3.4902744955462643, + "grad_norm": 0.14270208775997162, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0276, + "step": 38400 + }, + { + "epoch": 3.4911834211961463, + "grad_norm": 0.12443443387746811, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0263, + "step": 38410 + }, + { + "epoch": 3.492092346846028, + "grad_norm": 0.13598056137561798, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0298, + "step": 38420 + }, + { + "epoch": 3.49300127249591, + "grad_norm": 0.1441357582807541, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.029, + "step": 38430 + }, + { + "epoch": 3.493910198145792, + "grad_norm": 0.20425249636173248, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0293, + "step": 38440 + }, + { + "epoch": 3.4948191237956734, + "grad_norm": 0.14542034268379211, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0335, + "step": 38450 + }, + { + "epoch": 3.4957280494455554, + "grad_norm": 0.1471671462059021, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0287, + "step": 38460 + }, + { + "epoch": 3.4966369750954374, + "grad_norm": 0.16513797640800476, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0286, + "step": 38470 + }, + { + "epoch": 3.497545900745319, + "grad_norm": 0.1411045342683792, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0279, + "step": 38480 + }, + { + "epoch": 3.498454826395201, + "grad_norm": 0.16809913516044617, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0279, + "step": 38490 + }, + { + "epoch": 3.499363752045083, + "grad_norm": 0.1301134079694748, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0289, + "step": 38500 + }, + { + "epoch": 3.5002726776949644, + "grad_norm": 0.14681623876094818, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0287, + "step": 38510 + }, + { + "epoch": 3.5011816033448464, + "grad_norm": 0.11739318817853928, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0282, + "step": 38520 + }, + { + "epoch": 3.5020905289947284, + "grad_norm": 0.14634555578231812, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0305, + "step": 38530 + }, + { + "epoch": 3.50299945464461, + "grad_norm": 0.12115152180194855, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0257, + "step": 38540 + }, + { + "epoch": 3.503908380294492, + "grad_norm": 0.13926760852336884, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0287, + "step": 38550 + }, + { + "epoch": 3.504817305944374, + "grad_norm": 0.13851642608642578, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0293, + "step": 38560 + }, + { + "epoch": 3.5057262315942554, + "grad_norm": 0.13527579605579376, + "learning_rate": 2.435313188161029e-05, + "loss": 0.027, + "step": 38570 + }, + { + "epoch": 3.5066351572441374, + "grad_norm": 0.12978793680667877, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0278, + "step": 38580 + }, + { + "epoch": 3.5075440828940194, + "grad_norm": 0.15728804469108582, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0312, + "step": 38590 + }, + { + "epoch": 3.508453008543901, + "grad_norm": 0.1363018900156021, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0303, + "step": 38600 + }, + { + "epoch": 3.509361934193783, + "grad_norm": 0.1595749706029892, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0319, + "step": 38610 + }, + { + "epoch": 3.510270859843665, + "grad_norm": 0.16915017366409302, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0278, + "step": 38620 + }, + { + "epoch": 3.5111797854935465, + "grad_norm": 0.1417984664440155, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0273, + "step": 38630 + }, + { + "epoch": 3.5120887111434285, + "grad_norm": 0.14680352807044983, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.027, + "step": 38640 + }, + { + "epoch": 3.5129976367933105, + "grad_norm": 0.16989652812480927, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0281, + "step": 38650 + }, + { + "epoch": 3.513906562443192, + "grad_norm": 0.10527398437261581, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0254, + "step": 38660 + }, + { + "epoch": 3.514815488093074, + "grad_norm": 0.1584756225347519, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0281, + "step": 38670 + }, + { + "epoch": 3.515724413742956, + "grad_norm": 0.11017578840255737, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0297, + "step": 38680 + }, + { + "epoch": 3.5166333393928375, + "grad_norm": 0.19322960078716278, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0256, + "step": 38690 + }, + { + "epoch": 3.5175422650427195, + "grad_norm": 0.18182964622974396, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0303, + "step": 38700 + }, + { + "epoch": 3.5184511906926015, + "grad_norm": 0.1208077073097229, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0265, + "step": 38710 + }, + { + "epoch": 3.519360116342483, + "grad_norm": 0.1161116510629654, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0275, + "step": 38720 + }, + { + "epoch": 3.520269041992365, + "grad_norm": 0.15851573646068573, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0274, + "step": 38730 + }, + { + "epoch": 3.521177967642247, + "grad_norm": 0.1793825477361679, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0278, + "step": 38740 + }, + { + "epoch": 3.5220868932921285, + "grad_norm": 0.17432865500450134, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0281, + "step": 38750 + }, + { + "epoch": 3.5229958189420105, + "grad_norm": 0.13506847620010376, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0297, + "step": 38760 + }, + { + "epoch": 3.5239047445918925, + "grad_norm": 0.15209358930587769, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0275, + "step": 38770 + }, + { + "epoch": 3.524813670241774, + "grad_norm": 0.14445482194423676, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0269, + "step": 38780 + }, + { + "epoch": 3.525722595891656, + "grad_norm": 0.14536955952644348, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0289, + "step": 38790 + }, + { + "epoch": 3.526631521541538, + "grad_norm": 0.12323416769504547, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0255, + "step": 38800 + }, + { + "epoch": 3.52754044719142, + "grad_norm": 0.17338065803050995, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0265, + "step": 38810 + }, + { + "epoch": 3.5284493728413016, + "grad_norm": 0.15460923314094543, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0279, + "step": 38820 + }, + { + "epoch": 3.5293582984911835, + "grad_norm": 0.13843902945518494, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0304, + "step": 38830 + }, + { + "epoch": 3.530267224141065, + "grad_norm": 0.1570543497800827, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0273, + "step": 38840 + }, + { + "epoch": 3.531176149790947, + "grad_norm": 0.13217693567276, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0274, + "step": 38850 + }, + { + "epoch": 3.532085075440829, + "grad_norm": 0.11859201639890671, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0252, + "step": 38860 + }, + { + "epoch": 3.532994001090711, + "grad_norm": 0.14789028465747833, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0286, + "step": 38870 + }, + { + "epoch": 3.5339029267405926, + "grad_norm": 0.13406743109226227, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0284, + "step": 38880 + }, + { + "epoch": 3.5348118523904746, + "grad_norm": 0.11682151257991791, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0266, + "step": 38890 + }, + { + "epoch": 3.535720778040356, + "grad_norm": 0.13615073263645172, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0279, + "step": 38900 + }, + { + "epoch": 3.536629703690238, + "grad_norm": 0.11958518624305725, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0276, + "step": 38910 + }, + { + "epoch": 3.53753862934012, + "grad_norm": 0.11851118505001068, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0255, + "step": 38920 + }, + { + "epoch": 3.538447554990002, + "grad_norm": 0.12140911817550659, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0292, + "step": 38930 + }, + { + "epoch": 3.5393564806398836, + "grad_norm": 0.11916932463645935, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0267, + "step": 38940 + }, + { + "epoch": 3.5402654062897656, + "grad_norm": 0.130393847823143, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0277, + "step": 38950 + }, + { + "epoch": 3.541174331939647, + "grad_norm": 0.19491241872310638, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0283, + "step": 38960 + }, + { + "epoch": 3.542083257589529, + "grad_norm": 0.15372179448604584, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0252, + "step": 38970 + }, + { + "epoch": 3.542992183239411, + "grad_norm": 0.11123041063547134, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0269, + "step": 38980 + }, + { + "epoch": 3.543901108889293, + "grad_norm": 0.12423209100961685, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0308, + "step": 38990 + }, + { + "epoch": 3.5448100345391746, + "grad_norm": 0.12862753868103027, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0289, + "step": 39000 + }, + { + "epoch": 3.5457189601890566, + "grad_norm": 0.17585065960884094, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0281, + "step": 39010 + }, + { + "epoch": 3.546627885838938, + "grad_norm": 0.12105239182710648, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0275, + "step": 39020 + }, + { + "epoch": 3.54753681148882, + "grad_norm": 0.1338099241256714, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.03, + "step": 39030 + }, + { + "epoch": 3.548445737138702, + "grad_norm": 0.12285393476486206, + "learning_rate": 2.420046337658205e-05, + "loss": 0.028, + "step": 39040 + }, + { + "epoch": 3.549354662788584, + "grad_norm": 0.1599685698747635, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0274, + "step": 39050 + }, + { + "epoch": 3.5502635884384657, + "grad_norm": 0.10990972816944122, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0286, + "step": 39060 + }, + { + "epoch": 3.5511725140883477, + "grad_norm": 0.13761457800865173, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0275, + "step": 39070 + }, + { + "epoch": 3.552081439738229, + "grad_norm": 0.1659054011106491, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0259, + "step": 39080 + }, + { + "epoch": 3.552990365388111, + "grad_norm": 0.14508742094039917, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0276, + "step": 39090 + }, + { + "epoch": 3.553899291037993, + "grad_norm": 0.1399577558040619, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0254, + "step": 39100 + }, + { + "epoch": 3.554808216687875, + "grad_norm": 0.1473022848367691, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0282, + "step": 39110 + }, + { + "epoch": 3.5557171423377567, + "grad_norm": 0.1415456086397171, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0291, + "step": 39120 + }, + { + "epoch": 3.5566260679876387, + "grad_norm": 0.10876928269863129, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0274, + "step": 39130 + }, + { + "epoch": 3.5575349936375202, + "grad_norm": 0.13043063879013062, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0285, + "step": 39140 + }, + { + "epoch": 3.5584439192874022, + "grad_norm": 0.13827864825725555, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0293, + "step": 39150 + }, + { + "epoch": 3.559352844937284, + "grad_norm": 0.1408395767211914, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.027, + "step": 39160 + }, + { + "epoch": 3.560261770587166, + "grad_norm": 0.13573981821537018, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0296, + "step": 39170 + }, + { + "epoch": 3.5611706962370477, + "grad_norm": 0.18252608180046082, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0297, + "step": 39180 + }, + { + "epoch": 3.5620796218869297, + "grad_norm": 0.13623209297657013, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0276, + "step": 39190 + }, + { + "epoch": 3.5629885475368113, + "grad_norm": 0.12429483234882355, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0279, + "step": 39200 + }, + { + "epoch": 3.5638974731866933, + "grad_norm": 0.13393056392669678, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0288, + "step": 39210 + }, + { + "epoch": 3.5648063988365752, + "grad_norm": 0.11603590846061707, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0287, + "step": 39220 + }, + { + "epoch": 3.5657153244864572, + "grad_norm": 0.14166757464408875, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0304, + "step": 39230 + }, + { + "epoch": 3.5666242501363388, + "grad_norm": 0.14068962633609772, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0291, + "step": 39240 + }, + { + "epoch": 3.5675331757862208, + "grad_norm": 0.1492970585823059, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0276, + "step": 39250 + }, + { + "epoch": 3.5684421014361023, + "grad_norm": 0.15161293745040894, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0312, + "step": 39260 + }, + { + "epoch": 3.5693510270859843, + "grad_norm": 0.1574881672859192, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0289, + "step": 39270 + }, + { + "epoch": 3.5702599527358663, + "grad_norm": 0.1346648931503296, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0279, + "step": 39280 + }, + { + "epoch": 3.5711688783857483, + "grad_norm": 0.1216152012348175, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0266, + "step": 39290 + }, + { + "epoch": 3.57207780403563, + "grad_norm": 0.11860740929841995, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0264, + "step": 39300 + }, + { + "epoch": 3.572986729685512, + "grad_norm": 0.10847210139036179, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0304, + "step": 39310 + }, + { + "epoch": 3.5738956553353933, + "grad_norm": 0.13037437200546265, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.032, + "step": 39320 + }, + { + "epoch": 3.5748045809852753, + "grad_norm": 0.14538182318210602, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0268, + "step": 39330 + }, + { + "epoch": 3.5757135066351573, + "grad_norm": 0.12694279849529266, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0292, + "step": 39340 + }, + { + "epoch": 3.5766224322850393, + "grad_norm": 0.15376010537147522, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0296, + "step": 39350 + }, + { + "epoch": 3.577531357934921, + "grad_norm": 0.1259845346212387, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.029, + "step": 39360 + }, + { + "epoch": 3.578440283584803, + "grad_norm": 0.14743369817733765, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0287, + "step": 39370 + }, + { + "epoch": 3.5793492092346844, + "grad_norm": 0.19423452019691467, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0284, + "step": 39380 + }, + { + "epoch": 3.5802581348845663, + "grad_norm": 0.12347894161939621, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0316, + "step": 39390 + }, + { + "epoch": 3.5811670605344483, + "grad_norm": 0.141344353556633, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.026, + "step": 39400 + }, + { + "epoch": 3.5820759861843303, + "grad_norm": 0.18373706936836243, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0308, + "step": 39410 + }, + { + "epoch": 3.582984911834212, + "grad_norm": 0.15940868854522705, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0289, + "step": 39420 + }, + { + "epoch": 3.583893837484094, + "grad_norm": 0.14903829991817474, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0276, + "step": 39430 + }, + { + "epoch": 3.5848027631339754, + "grad_norm": 0.11909395456314087, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0276, + "step": 39440 + }, + { + "epoch": 3.5857116887838574, + "grad_norm": 0.11607155203819275, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0286, + "step": 39450 + }, + { + "epoch": 3.5866206144337394, + "grad_norm": 0.12419041246175766, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0291, + "step": 39460 + }, + { + "epoch": 3.5875295400836213, + "grad_norm": 0.15012645721435547, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0284, + "step": 39470 + }, + { + "epoch": 3.588438465733503, + "grad_norm": 0.11934713274240494, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0245, + "step": 39480 + }, + { + "epoch": 3.589347391383385, + "grad_norm": 0.1413666158914566, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0253, + "step": 39490 + }, + { + "epoch": 3.5902563170332664, + "grad_norm": 0.12734445929527283, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.0247, + "step": 39500 + }, + { + "epoch": 3.5911652426831484, + "grad_norm": 0.13282664120197296, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0291, + "step": 39510 + }, + { + "epoch": 3.5920741683330304, + "grad_norm": 0.12596207857131958, + "learning_rate": 2.404295374765938e-05, + "loss": 0.025, + "step": 39520 + }, + { + "epoch": 3.5929830939829124, + "grad_norm": 0.14804257452487946, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0281, + "step": 39530 + }, + { + "epoch": 3.593892019632794, + "grad_norm": 0.14551973342895508, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0301, + "step": 39540 + }, + { + "epoch": 3.594800945282676, + "grad_norm": 0.17820389568805695, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0266, + "step": 39550 + }, + { + "epoch": 3.5957098709325575, + "grad_norm": 0.14789901673793793, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.029, + "step": 39560 + }, + { + "epoch": 3.5966187965824394, + "grad_norm": 0.19211125373840332, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0328, + "step": 39570 + }, + { + "epoch": 3.5975277222323214, + "grad_norm": 0.16240429878234863, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0266, + "step": 39580 + }, + { + "epoch": 3.5984366478822034, + "grad_norm": 0.15394212305545807, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0255, + "step": 39590 + }, + { + "epoch": 3.599345573532085, + "grad_norm": 0.16021698713302612, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0286, + "step": 39600 + }, + { + "epoch": 3.600254499181967, + "grad_norm": 0.1664794683456421, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0292, + "step": 39610 + }, + { + "epoch": 3.6011634248318485, + "grad_norm": 0.17323166131973267, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0309, + "step": 39620 + }, + { + "epoch": 3.6020723504817305, + "grad_norm": 0.14941754937171936, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0265, + "step": 39630 + }, + { + "epoch": 3.6029812761316125, + "grad_norm": 0.13309089839458466, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0277, + "step": 39640 + }, + { + "epoch": 3.6038902017814944, + "grad_norm": 0.1708577573299408, + "learning_rate": 2.400002143698994e-05, + "loss": 0.029, + "step": 39650 + }, + { + "epoch": 3.604799127431376, + "grad_norm": 0.13084794580936432, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0299, + "step": 39660 + }, + { + "epoch": 3.605708053081258, + "grad_norm": 0.13116832077503204, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0255, + "step": 39670 + }, + { + "epoch": 3.60661697873114, + "grad_norm": 0.3361426889896393, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0272, + "step": 39680 + }, + { + "epoch": 3.6075259043810215, + "grad_norm": 0.16729164123535156, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0316, + "step": 39690 + }, + { + "epoch": 3.6084348300309035, + "grad_norm": 0.13700374960899353, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0292, + "step": 39700 + }, + { + "epoch": 3.6093437556807855, + "grad_norm": 0.2462116926908493, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0281, + "step": 39710 + }, + { + "epoch": 3.610252681330667, + "grad_norm": 0.17136697471141815, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0255, + "step": 39720 + }, + { + "epoch": 3.611161606980549, + "grad_norm": 0.13743440806865692, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0262, + "step": 39730 + }, + { + "epoch": 3.612070532630431, + "grad_norm": 0.13883920013904572, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0293, + "step": 39740 + }, + { + "epoch": 3.6129794582803125, + "grad_norm": 0.1542753428220749, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0267, + "step": 39750 + }, + { + "epoch": 3.6138883839301945, + "grad_norm": 0.1479600965976715, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0269, + "step": 39760 + }, + { + "epoch": 3.6147973095800765, + "grad_norm": 0.16537567973136902, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0287, + "step": 39770 + }, + { + "epoch": 3.615706235229958, + "grad_norm": 0.15847907960414886, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.025, + "step": 39780 + }, + { + "epoch": 3.61661516087984, + "grad_norm": 0.15733537077903748, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0289, + "step": 39790 + }, + { + "epoch": 3.617524086529722, + "grad_norm": 0.17272816598415375, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.029, + "step": 39800 + }, + { + "epoch": 3.6184330121796036, + "grad_norm": 0.14946813881397247, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0267, + "step": 39810 + }, + { + "epoch": 3.6193419378294855, + "grad_norm": 0.12948711216449738, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.032, + "step": 39820 + }, + { + "epoch": 3.6202508634793675, + "grad_norm": 0.15288402140140533, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0282, + "step": 39830 + }, + { + "epoch": 3.621159789129249, + "grad_norm": 0.12943552434444427, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0276, + "step": 39840 + }, + { + "epoch": 3.622068714779131, + "grad_norm": 0.11214439570903778, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0297, + "step": 39850 + }, + { + "epoch": 3.622977640429013, + "grad_norm": 0.13189436495304108, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0274, + "step": 39860 + }, + { + "epoch": 3.6238865660788946, + "grad_norm": 0.11950542032718658, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0267, + "step": 39870 + }, + { + "epoch": 3.6247954917287766, + "grad_norm": 0.1294006109237671, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0286, + "step": 39880 + }, + { + "epoch": 3.6257044173786586, + "grad_norm": 0.14559121429920197, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0271, + "step": 39890 + }, + { + "epoch": 3.62661334302854, + "grad_norm": 0.17395134270191193, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0295, + "step": 39900 + }, + { + "epoch": 3.627522268678422, + "grad_norm": 0.14212758839130402, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0297, + "step": 39910 + }, + { + "epoch": 3.628431194328304, + "grad_norm": 0.1869899034500122, + "learning_rate": 2.391048587855479e-05, + "loss": 0.028, + "step": 39920 + }, + { + "epoch": 3.6293401199781856, + "grad_norm": 0.12245742976665497, + "learning_rate": 2.390716024246585e-05, + "loss": 0.028, + "step": 39930 + }, + { + "epoch": 3.6302490456280676, + "grad_norm": 0.14054062962532043, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0243, + "step": 39940 + }, + { + "epoch": 3.6311579712779496, + "grad_norm": 0.1441534161567688, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0281, + "step": 39950 + }, + { + "epoch": 3.6320668969278316, + "grad_norm": 0.15512555837631226, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0268, + "step": 39960 + }, + { + "epoch": 3.632975822577713, + "grad_norm": 0.1294974982738495, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0278, + "step": 39970 + }, + { + "epoch": 3.633884748227595, + "grad_norm": 0.15016469359397888, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0282, + "step": 39980 + }, + { + "epoch": 3.6347936738774766, + "grad_norm": 0.12587836384773254, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0272, + "step": 39990 + }, + { + "epoch": 3.6357025995273586, + "grad_norm": 0.1533900797367096, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0314, + "step": 40000 + }, + { + "epoch": 3.6366115251772406, + "grad_norm": 0.13991421461105347, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0273, + "step": 40010 + }, + { + "epoch": 3.6375204508271226, + "grad_norm": 0.13677667081356049, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0277, + "step": 40020 + }, + { + "epoch": 3.638429376477004, + "grad_norm": 0.1332559436559677, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0268, + "step": 40030 + }, + { + "epoch": 3.639338302126886, + "grad_norm": 0.12659761309623718, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0265, + "step": 40040 + }, + { + "epoch": 3.6402472277767677, + "grad_norm": 0.12645314633846283, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0282, + "step": 40050 + }, + { + "epoch": 3.6411561534266497, + "grad_norm": 0.16264456510543823, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0261, + "step": 40060 + }, + { + "epoch": 3.6420650790765317, + "grad_norm": 0.1414022445678711, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0286, + "step": 40070 + }, + { + "epoch": 3.6429740047264136, + "grad_norm": 0.1620587259531021, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0328, + "step": 40080 + }, + { + "epoch": 3.643882930376295, + "grad_norm": 0.13953226804733276, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0277, + "step": 40090 + }, + { + "epoch": 3.644791856026177, + "grad_norm": 0.17347168922424316, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0293, + "step": 40100 + }, + { + "epoch": 3.6457007816760587, + "grad_norm": 0.10265520960092545, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0261, + "step": 40110 + }, + { + "epoch": 3.6466097073259407, + "grad_norm": 0.1108442097902298, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0255, + "step": 40120 + }, + { + "epoch": 3.6475186329758227, + "grad_norm": 0.14928823709487915, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0279, + "step": 40130 + }, + { + "epoch": 3.6484275586257047, + "grad_norm": 0.11974356323480606, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0277, + "step": 40140 + }, + { + "epoch": 3.649336484275586, + "grad_norm": 0.1560451090335846, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0273, + "step": 40150 + }, + { + "epoch": 3.650245409925468, + "grad_norm": 0.11242460459470749, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0285, + "step": 40160 + }, + { + "epoch": 3.6511543355753497, + "grad_norm": 0.14120955765247345, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0293, + "step": 40170 + }, + { + "epoch": 3.6520632612252317, + "grad_norm": 0.13555069267749786, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0251, + "step": 40180 + }, + { + "epoch": 3.6529721868751137, + "grad_norm": 0.11787337064743042, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0298, + "step": 40190 + }, + { + "epoch": 3.6538811125249957, + "grad_norm": 0.1142067164182663, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0249, + "step": 40200 + }, + { + "epoch": 3.6547900381748772, + "grad_norm": 0.13205809891223907, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0274, + "step": 40210 + }, + { + "epoch": 3.6556989638247592, + "grad_norm": 0.12771107256412506, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0277, + "step": 40220 + }, + { + "epoch": 3.6566078894746408, + "grad_norm": 0.13035845756530762, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0262, + "step": 40230 + }, + { + "epoch": 3.6575168151245228, + "grad_norm": 0.6248196363449097, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.031, + "step": 40240 + }, + { + "epoch": 3.6584257407744047, + "grad_norm": 0.14339782297611237, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0304, + "step": 40250 + }, + { + "epoch": 3.6593346664242867, + "grad_norm": 0.1542704999446869, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0282, + "step": 40260 + }, + { + "epoch": 3.6602435920741683, + "grad_norm": 0.15138593316078186, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0266, + "step": 40270 + }, + { + "epoch": 3.6611525177240503, + "grad_norm": 0.14509163796901703, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0262, + "step": 40280 + }, + { + "epoch": 3.662061443373932, + "grad_norm": 0.14742866158485413, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0277, + "step": 40290 + }, + { + "epoch": 3.662970369023814, + "grad_norm": 0.11565099656581879, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0299, + "step": 40300 + }, + { + "epoch": 3.6638792946736958, + "grad_norm": 0.2079046219587326, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0289, + "step": 40310 + }, + { + "epoch": 3.6647882203235778, + "grad_norm": 0.15395961701869965, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.026, + "step": 40320 + }, + { + "epoch": 3.6656971459734593, + "grad_norm": 0.13826815783977509, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0287, + "step": 40330 + }, + { + "epoch": 3.6666060716233413, + "grad_norm": 0.16627639532089233, + "learning_rate": 2.377022965365761e-05, + "loss": 0.027, + "step": 40340 + }, + { + "epoch": 3.667514997273223, + "grad_norm": 0.1230824664235115, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0286, + "step": 40350 + }, + { + "epoch": 3.668423922923105, + "grad_norm": 0.15636637806892395, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0287, + "step": 40360 + }, + { + "epoch": 3.669332848572987, + "grad_norm": 0.16088896989822388, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0308, + "step": 40370 + }, + { + "epoch": 3.670241774222869, + "grad_norm": 0.14193986356258392, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0289, + "step": 40380 + }, + { + "epoch": 3.6711506998727503, + "grad_norm": 0.13185980916023254, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0258, + "step": 40390 + }, + { + "epoch": 3.6720596255226323, + "grad_norm": 0.13658864796161652, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0278, + "step": 40400 + }, + { + "epoch": 3.672968551172514, + "grad_norm": 0.15165109932422638, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.029, + "step": 40410 + }, + { + "epoch": 3.673877476822396, + "grad_norm": 0.14393538236618042, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0276, + "step": 40420 + }, + { + "epoch": 3.674786402472278, + "grad_norm": 0.13160082697868347, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0279, + "step": 40430 + }, + { + "epoch": 3.67569532812216, + "grad_norm": 0.22056329250335693, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0308, + "step": 40440 + }, + { + "epoch": 3.6766042537720414, + "grad_norm": 0.13721595704555511, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0253, + "step": 40450 + }, + { + "epoch": 3.6775131794219234, + "grad_norm": 0.18860892951488495, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0279, + "step": 40460 + }, + { + "epoch": 3.678422105071805, + "grad_norm": 0.1436915546655655, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0279, + "step": 40470 + }, + { + "epoch": 3.679331030721687, + "grad_norm": 0.1673458218574524, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0276, + "step": 40480 + }, + { + "epoch": 3.680239956371569, + "grad_norm": 0.10937394201755524, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0262, + "step": 40490 + }, + { + "epoch": 3.681148882021451, + "grad_norm": 0.13297715783119202, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0258, + "step": 40500 + }, + { + "epoch": 3.6820578076713324, + "grad_norm": 0.15519167482852936, + "learning_rate": 2.371312419582388e-05, + "loss": 0.027, + "step": 40510 + }, + { + "epoch": 3.6829667333212144, + "grad_norm": 0.14324982464313507, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0258, + "step": 40520 + }, + { + "epoch": 3.683875658971096, + "grad_norm": 0.1703028529882431, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0237, + "step": 40530 + }, + { + "epoch": 3.684784584620978, + "grad_norm": 0.15399497747421265, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0284, + "step": 40540 + }, + { + "epoch": 3.68569351027086, + "grad_norm": 0.1787704974412918, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0273, + "step": 40550 + }, + { + "epoch": 3.686602435920742, + "grad_norm": 0.12312978506088257, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0294, + "step": 40560 + }, + { + "epoch": 3.6875113615706234, + "grad_norm": 0.13837918639183044, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0284, + "step": 40570 + }, + { + "epoch": 3.6884202872205054, + "grad_norm": 0.15840347111225128, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0271, + "step": 40580 + }, + { + "epoch": 3.689329212870387, + "grad_norm": 0.1623680144548416, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0257, + "step": 40590 + }, + { + "epoch": 3.690238138520269, + "grad_norm": 0.12414079159498215, + "learning_rate": 2.368281439559657e-05, + "loss": 0.03, + "step": 40600 + }, + { + "epoch": 3.691147064170151, + "grad_norm": 0.1538112610578537, + "learning_rate": 2.367944333959188e-05, + "loss": 0.029, + "step": 40610 + }, + { + "epoch": 3.692055989820033, + "grad_norm": 0.19994176924228668, + "learning_rate": 2.367607162444074e-05, + "loss": 0.028, + "step": 40620 + }, + { + "epoch": 3.6929649154699145, + "grad_norm": 0.16593891382217407, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0279, + "step": 40630 + }, + { + "epoch": 3.6938738411197964, + "grad_norm": 0.14864720404148102, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.028, + "step": 40640 + }, + { + "epoch": 3.694782766769678, + "grad_norm": 0.1285313367843628, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0263, + "step": 40650 + }, + { + "epoch": 3.69569169241956, + "grad_norm": 0.14182300865650177, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0311, + "step": 40660 + }, + { + "epoch": 3.696600618069442, + "grad_norm": 0.13082534074783325, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0261, + "step": 40670 + }, + { + "epoch": 3.697509543719324, + "grad_norm": 0.130438894033432, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0289, + "step": 40680 + }, + { + "epoch": 3.6984184693692055, + "grad_norm": 0.1577245444059372, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0286, + "step": 40690 + }, + { + "epoch": 3.6993273950190875, + "grad_norm": 0.1086277887225151, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0284, + "step": 40700 + }, + { + "epoch": 3.700236320668969, + "grad_norm": 0.17116737365722656, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0269, + "step": 40710 + }, + { + "epoch": 3.701145246318851, + "grad_norm": 0.13738079369068146, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0289, + "step": 40720 + }, + { + "epoch": 3.702054171968733, + "grad_norm": 0.12071609497070312, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0268, + "step": 40730 + }, + { + "epoch": 3.702963097618615, + "grad_norm": 0.20122705399990082, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0289, + "step": 40740 + }, + { + "epoch": 3.7038720232684965, + "grad_norm": 0.15469475090503693, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0265, + "step": 40750 + }, + { + "epoch": 3.7047809489183785, + "grad_norm": 0.1901894509792328, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0315, + "step": 40760 + }, + { + "epoch": 3.70568987456826, + "grad_norm": 0.126474529504776, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0281, + "step": 40770 + }, + { + "epoch": 3.706598800218142, + "grad_norm": 0.12078749388456345, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0287, + "step": 40780 + }, + { + "epoch": 3.707507725868024, + "grad_norm": 0.13409623503684998, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0256, + "step": 40790 + }, + { + "epoch": 3.708416651517906, + "grad_norm": 0.14311067759990692, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0257, + "step": 40800 + }, + { + "epoch": 3.7093255771677875, + "grad_norm": 0.14879633486270905, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0268, + "step": 40810 + }, + { + "epoch": 3.7102345028176695, + "grad_norm": 0.14450840651988983, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0288, + "step": 40820 + }, + { + "epoch": 3.7111434284675515, + "grad_norm": 0.13446173071861267, + "learning_rate": 2.360511379736256e-05, + "loss": 0.029, + "step": 40830 + }, + { + "epoch": 3.712052354117433, + "grad_norm": 0.13360346853733063, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0259, + "step": 40840 + }, + { + "epoch": 3.712961279767315, + "grad_norm": 0.13387440145015717, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0277, + "step": 40850 + }, + { + "epoch": 3.713870205417197, + "grad_norm": 0.1527571678161621, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0275, + "step": 40860 + }, + { + "epoch": 3.7147791310670786, + "grad_norm": 0.13624019920825958, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0272, + "step": 40870 + }, + { + "epoch": 3.7156880567169606, + "grad_norm": 0.11691347509622574, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0281, + "step": 40880 + }, + { + "epoch": 3.7165969823668425, + "grad_norm": 0.13619515299797058, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0278, + "step": 40890 + }, + { + "epoch": 3.717505908016724, + "grad_norm": 0.20115584135055542, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0266, + "step": 40900 + }, + { + "epoch": 3.718414833666606, + "grad_norm": 0.19840243458747864, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0278, + "step": 40910 + }, + { + "epoch": 3.719323759316488, + "grad_norm": 0.1510859727859497, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0272, + "step": 40920 + }, + { + "epoch": 3.7202326849663696, + "grad_norm": 0.15583762526512146, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0287, + "step": 40930 + }, + { + "epoch": 3.7211416106162516, + "grad_norm": 0.15777157247066498, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0276, + "step": 40940 + }, + { + "epoch": 3.7220505362661336, + "grad_norm": 0.1556655615568161, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0288, + "step": 40950 + }, + { + "epoch": 3.722959461916015, + "grad_norm": 0.1269613355398178, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0282, + "step": 40960 + }, + { + "epoch": 3.723868387565897, + "grad_norm": 0.14936824142932892, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0296, + "step": 40970 + }, + { + "epoch": 3.724777313215779, + "grad_norm": 0.13304229080677032, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0285, + "step": 40980 + }, + { + "epoch": 3.7256862388656606, + "grad_norm": 0.15071475505828857, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0264, + "step": 40990 + }, + { + "epoch": 3.7265951645155426, + "grad_norm": 0.13745757937431335, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0275, + "step": 41000 + }, + { + "epoch": 3.7275040901654246, + "grad_norm": 0.14260682463645935, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0273, + "step": 41010 + }, + { + "epoch": 3.728413015815306, + "grad_norm": 0.11041214317083359, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0258, + "step": 41020 + }, + { + "epoch": 3.729321941465188, + "grad_norm": 0.1573951691389084, + "learning_rate": 2.353726694434432e-05, + "loss": 0.027, + "step": 41030 + }, + { + "epoch": 3.73023086711507, + "grad_norm": 0.11698150634765625, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.028, + "step": 41040 + }, + { + "epoch": 3.7311397927649517, + "grad_norm": 0.12878747284412384, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0284, + "step": 41050 + }, + { + "epoch": 3.7320487184148337, + "grad_norm": 0.15278799831867218, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.027, + "step": 41060 + }, + { + "epoch": 3.7329576440647156, + "grad_norm": 0.09313280880451202, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0292, + "step": 41070 + }, + { + "epoch": 3.733866569714597, + "grad_norm": 0.10817251354455948, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0246, + "step": 41080 + }, + { + "epoch": 3.734775495364479, + "grad_norm": 0.20619326829910278, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0274, + "step": 41090 + }, + { + "epoch": 3.735684421014361, + "grad_norm": 0.12001074105501175, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0263, + "step": 41100 + }, + { + "epoch": 3.7365933466642427, + "grad_norm": 0.15438875555992126, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0288, + "step": 41110 + }, + { + "epoch": 3.7375022723141247, + "grad_norm": 0.138901948928833, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.029, + "step": 41120 + }, + { + "epoch": 3.7384111979640067, + "grad_norm": 0.14516116678714752, + "learning_rate": 2.350324613659711e-05, + "loss": 0.028, + "step": 41130 + }, + { + "epoch": 3.739320123613888, + "grad_norm": 0.14791959524154663, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0283, + "step": 41140 + }, + { + "epoch": 3.74022904926377, + "grad_norm": 0.12917818129062653, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0259, + "step": 41150 + }, + { + "epoch": 3.741137974913652, + "grad_norm": 0.23610560595989227, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0279, + "step": 41160 + }, + { + "epoch": 3.742046900563534, + "grad_norm": 0.134189173579216, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0253, + "step": 41170 + }, + { + "epoch": 3.7429558262134157, + "grad_norm": 0.17175817489624023, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0301, + "step": 41180 + }, + { + "epoch": 3.7438647518632977, + "grad_norm": 0.10512970387935638, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0272, + "step": 41190 + }, + { + "epoch": 3.7447736775131792, + "grad_norm": 0.12799392640590668, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0259, + "step": 41200 + }, + { + "epoch": 3.7456826031630612, + "grad_norm": 0.14360283315181732, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0268, + "step": 41210 + }, + { + "epoch": 3.746591528812943, + "grad_norm": 0.13469554483890533, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0291, + "step": 41220 + }, + { + "epoch": 3.747500454462825, + "grad_norm": 0.10995388776063919, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0273, + "step": 41230 + }, + { + "epoch": 3.7484093801127067, + "grad_norm": 0.13215364515781403, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.025, + "step": 41240 + }, + { + "epoch": 3.7493183057625887, + "grad_norm": 0.1374896913766861, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0275, + "step": 41250 + }, + { + "epoch": 3.7502272314124703, + "grad_norm": 0.1454111486673355, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0258, + "step": 41260 + }, + { + "epoch": 3.7511361570623523, + "grad_norm": 0.16365788877010345, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0282, + "step": 41270 + }, + { + "epoch": 3.7520450827122342, + "grad_norm": 0.14891397953033447, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0278, + "step": 41280 + }, + { + "epoch": 3.7529540083621162, + "grad_norm": 0.15253649652004242, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0268, + "step": 41290 + }, + { + "epoch": 3.7538629340119978, + "grad_norm": 0.15039588510990143, + "learning_rate": 2.344526269895174e-05, + "loss": 0.029, + "step": 41300 + }, + { + "epoch": 3.7547718596618798, + "grad_norm": 0.14731895923614502, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0263, + "step": 41310 + }, + { + "epoch": 3.7556807853117613, + "grad_norm": 0.18362154066562653, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0257, + "step": 41320 + }, + { + "epoch": 3.7565897109616433, + "grad_norm": 0.12622793018817902, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0291, + "step": 41330 + }, + { + "epoch": 3.7574986366115253, + "grad_norm": 0.1464890092611313, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0266, + "step": 41340 + }, + { + "epoch": 3.7584075622614073, + "grad_norm": 0.16916793584823608, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0281, + "step": 41350 + }, + { + "epoch": 3.759316487911289, + "grad_norm": 0.13707515597343445, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0271, + "step": 41360 + }, + { + "epoch": 3.760225413561171, + "grad_norm": 0.14633367955684662, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0278, + "step": 41370 + }, + { + "epoch": 3.7611343392110523, + "grad_norm": 0.11950068175792694, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0274, + "step": 41380 + }, + { + "epoch": 3.7620432648609343, + "grad_norm": 0.10404278337955475, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0283, + "step": 41390 + }, + { + "epoch": 3.7629521905108163, + "grad_norm": 0.13233700394630432, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0296, + "step": 41400 + }, + { + "epoch": 3.7638611161606983, + "grad_norm": 0.16562408208847046, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0277, + "step": 41410 + }, + { + "epoch": 3.76477004181058, + "grad_norm": 0.14466042816638947, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.025, + "step": 41420 + }, + { + "epoch": 3.765678967460462, + "grad_norm": 0.12380529940128326, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0283, + "step": 41430 + }, + { + "epoch": 3.7665878931103434, + "grad_norm": 0.1809171736240387, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0287, + "step": 41440 + }, + { + "epoch": 3.7674968187602254, + "grad_norm": 0.12599726021289825, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0284, + "step": 41450 + }, + { + "epoch": 3.7684057444101073, + "grad_norm": 0.12617944180965424, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0242, + "step": 41460 + }, + { + "epoch": 3.7693146700599893, + "grad_norm": 0.14903753995895386, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0316, + "step": 41470 + }, + { + "epoch": 3.770223595709871, + "grad_norm": 0.1204192042350769, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0268, + "step": 41480 + }, + { + "epoch": 3.771132521359753, + "grad_norm": 0.15835562348365784, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0283, + "step": 41490 + }, + { + "epoch": 3.7720414470096344, + "grad_norm": 0.16681641340255737, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.026, + "step": 41500 + }, + { + "epoch": 3.7729503726595164, + "grad_norm": 0.16023904085159302, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0289, + "step": 41510 + }, + { + "epoch": 3.7738592983093984, + "grad_norm": 0.1432938426733017, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0275, + "step": 41520 + }, + { + "epoch": 3.7747682239592804, + "grad_norm": 0.11220147460699081, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0273, + "step": 41530 + }, + { + "epoch": 3.775677149609162, + "grad_norm": 0.14988286793231964, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0253, + "step": 41540 + }, + { + "epoch": 3.776586075259044, + "grad_norm": 0.11027409881353378, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0267, + "step": 41550 + }, + { + "epoch": 3.7774950009089254, + "grad_norm": 0.18963520228862762, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0282, + "step": 41560 + }, + { + "epoch": 3.7784039265588074, + "grad_norm": 0.15563338994979858, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.028, + "step": 41570 + }, + { + "epoch": 3.7793128522086894, + "grad_norm": 0.13193535804748535, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0274, + "step": 41580 + }, + { + "epoch": 3.7802217778585714, + "grad_norm": 0.1355358064174652, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0269, + "step": 41590 + }, + { + "epoch": 3.781130703508453, + "grad_norm": 0.12078381329774857, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0253, + "step": 41600 + }, + { + "epoch": 3.782039629158335, + "grad_norm": 0.11047215014696121, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0243, + "step": 41610 + }, + { + "epoch": 3.7829485548082165, + "grad_norm": 0.16160964965820312, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0276, + "step": 41620 + }, + { + "epoch": 3.7838574804580984, + "grad_norm": 0.15577028691768646, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0292, + "step": 41630 + }, + { + "epoch": 3.7847664061079804, + "grad_norm": 0.1718808114528656, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0288, + "step": 41640 + }, + { + "epoch": 3.7856753317578624, + "grad_norm": 0.1552131175994873, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0249, + "step": 41650 + }, + { + "epoch": 3.786584257407744, + "grad_norm": 0.1181882992386818, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0285, + "step": 41660 + }, + { + "epoch": 3.787493183057626, + "grad_norm": 0.15517167747020721, + "learning_rate": 2.331842446374684e-05, + "loss": 0.026, + "step": 41670 + }, + { + "epoch": 3.7884021087075075, + "grad_norm": 0.15386171638965607, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0255, + "step": 41680 + }, + { + "epoch": 3.7893110343573895, + "grad_norm": 0.12192405015230179, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.028, + "step": 41690 + }, + { + "epoch": 3.7902199600072715, + "grad_norm": 0.22323113679885864, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0286, + "step": 41700 + }, + { + "epoch": 3.7911288856571534, + "grad_norm": 0.1355631947517395, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0255, + "step": 41710 + }, + { + "epoch": 3.792037811307035, + "grad_norm": 0.1653657704591751, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0274, + "step": 41720 + }, + { + "epoch": 3.792946736956917, + "grad_norm": 0.18613950908184052, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0288, + "step": 41730 + }, + { + "epoch": 3.7938556626067985, + "grad_norm": 0.11704002320766449, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0261, + "step": 41740 + }, + { + "epoch": 3.7947645882566805, + "grad_norm": 0.1760173887014389, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.026, + "step": 41750 + }, + { + "epoch": 3.7956735139065625, + "grad_norm": 0.1369471251964569, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0274, + "step": 41760 + }, + { + "epoch": 3.7965824395564445, + "grad_norm": 0.13598307967185974, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0266, + "step": 41770 + }, + { + "epoch": 3.797491365206326, + "grad_norm": 0.13864155113697052, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0291, + "step": 41780 + }, + { + "epoch": 3.798400290856208, + "grad_norm": 0.1551838368177414, + "learning_rate": 2.327710137160376e-05, + "loss": 0.03, + "step": 41790 + }, + { + "epoch": 3.7993092165060895, + "grad_norm": 0.15064217150211334, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0271, + "step": 41800 + }, + { + "epoch": 3.8002181421559715, + "grad_norm": 0.13434943556785583, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0242, + "step": 41810 + }, + { + "epoch": 3.8011270678058535, + "grad_norm": 0.15828941762447357, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0279, + "step": 41820 + }, + { + "epoch": 3.8020359934557355, + "grad_norm": 0.1360432356595993, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0266, + "step": 41830 + }, + { + "epoch": 3.802944919105617, + "grad_norm": 0.15012191236019135, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0274, + "step": 41840 + }, + { + "epoch": 3.803853844755499, + "grad_norm": 0.12168065458536148, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0264, + "step": 41850 + }, + { + "epoch": 3.8047627704053806, + "grad_norm": 0.13788586854934692, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0254, + "step": 41860 + }, + { + "epoch": 3.8056716960552626, + "grad_norm": 0.2006262093782425, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0278, + "step": 41870 + }, + { + "epoch": 3.8065806217051446, + "grad_norm": 0.14484499394893646, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0284, + "step": 41880 + }, + { + "epoch": 3.8074895473550265, + "grad_norm": 0.1317932903766632, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0283, + "step": 41890 + }, + { + "epoch": 3.808398473004908, + "grad_norm": 0.17657479643821716, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0279, + "step": 41900 + }, + { + "epoch": 3.80930739865479, + "grad_norm": 0.14232291281223297, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0271, + "step": 41910 + }, + { + "epoch": 3.8102163243046716, + "grad_norm": 0.12227810174226761, + "learning_rate": 2.323223255665661e-05, + "loss": 0.025, + "step": 41920 + }, + { + "epoch": 3.8111252499545536, + "grad_norm": 0.16387973725795746, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0288, + "step": 41930 + }, + { + "epoch": 3.8120341756044356, + "grad_norm": 0.17048858106136322, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0272, + "step": 41940 + }, + { + "epoch": 3.8129431012543176, + "grad_norm": 0.16584512591362, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0269, + "step": 41950 + }, + { + "epoch": 3.813852026904199, + "grad_norm": 0.14159977436065674, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0283, + "step": 41960 + }, + { + "epoch": 3.814760952554081, + "grad_norm": 0.12633098661899567, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0288, + "step": 41970 + }, + { + "epoch": 3.8156698782039626, + "grad_norm": 0.10959349572658539, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.029, + "step": 41980 + }, + { + "epoch": 3.8165788038538446, + "grad_norm": 0.13654735684394836, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0246, + "step": 41990 + }, + { + "epoch": 3.8174877295037266, + "grad_norm": 0.1349276453256607, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0254, + "step": 42000 + }, + { + "epoch": 3.8183966551536086, + "grad_norm": 0.09352292120456696, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0275, + "step": 42010 + }, + { + "epoch": 3.81930558080349, + "grad_norm": 0.16513222455978394, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0258, + "step": 42020 + }, + { + "epoch": 3.820214506453372, + "grad_norm": 0.145066037774086, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0247, + "step": 42030 + }, + { + "epoch": 3.821123432103254, + "grad_norm": 0.13713084161281586, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0264, + "step": 42040 + }, + { + "epoch": 3.8220323577531357, + "grad_norm": 0.14165395498275757, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0276, + "step": 42050 + }, + { + "epoch": 3.8229412834030176, + "grad_norm": 0.15306995809078217, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0274, + "step": 42060 + }, + { + "epoch": 3.8238502090528996, + "grad_norm": 0.14311917126178741, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0335, + "step": 42070 + }, + { + "epoch": 3.824759134702781, + "grad_norm": 0.16318954527378082, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0283, + "step": 42080 + }, + { + "epoch": 3.825668060352663, + "grad_norm": 0.13629160821437836, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0244, + "step": 42090 + }, + { + "epoch": 3.826576986002545, + "grad_norm": 0.1264946311712265, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0258, + "step": 42100 + }, + { + "epoch": 3.8274859116524267, + "grad_norm": 0.1263478398323059, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0282, + "step": 42110 + }, + { + "epoch": 3.8283948373023087, + "grad_norm": 0.14044813811779022, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0242, + "step": 42120 + }, + { + "epoch": 3.8293037629521907, + "grad_norm": 0.1768442690372467, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0243, + "step": 42130 + }, + { + "epoch": 3.830212688602072, + "grad_norm": 0.24207669496536255, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0277, + "step": 42140 + }, + { + "epoch": 3.831121614251954, + "grad_norm": 0.16992738842964172, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0268, + "step": 42150 + }, + { + "epoch": 3.832030539901836, + "grad_norm": 0.15240778028964996, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0294, + "step": 42160 + }, + { + "epoch": 3.8329394655517177, + "grad_norm": 0.13550607860088348, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0283, + "step": 42170 + }, + { + "epoch": 3.8338483912015997, + "grad_norm": 0.10267576575279236, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0252, + "step": 42180 + }, + { + "epoch": 3.8347573168514817, + "grad_norm": 0.15995003283023834, + "learning_rate": 2.313870654271428e-05, + "loss": 0.026, + "step": 42190 + }, + { + "epoch": 3.8356662425013632, + "grad_norm": 0.14163810014724731, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0277, + "step": 42200 + }, + { + "epoch": 3.836575168151245, + "grad_norm": 0.1205059364438057, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0241, + "step": 42210 + }, + { + "epoch": 3.837484093801127, + "grad_norm": 0.16949039697647095, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0256, + "step": 42220 + }, + { + "epoch": 3.8383930194510087, + "grad_norm": 0.13764864206314087, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0264, + "step": 42230 + }, + { + "epoch": 3.8393019451008907, + "grad_norm": 0.1294698715209961, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0262, + "step": 42240 + }, + { + "epoch": 3.8402108707507727, + "grad_norm": 0.12605774402618408, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0287, + "step": 42250 + }, + { + "epoch": 3.8411197964006543, + "grad_norm": 0.15280885994434357, + "learning_rate": 2.311438529704657e-05, + "loss": 0.032, + "step": 42260 + }, + { + "epoch": 3.8420287220505362, + "grad_norm": 0.15035058557987213, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0283, + "step": 42270 + }, + { + "epoch": 3.8429376477004182, + "grad_norm": 0.11026208847761154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0259, + "step": 42280 + }, + { + "epoch": 3.8438465733502998, + "grad_norm": 0.12977896630764008, + "learning_rate": 2.310395265725751e-05, + "loss": 0.028, + "step": 42290 + }, + { + "epoch": 3.8447554990001818, + "grad_norm": 0.20201386511325836, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0263, + "step": 42300 + }, + { + "epoch": 3.8456644246500638, + "grad_norm": 0.15406785905361176, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0268, + "step": 42310 + }, + { + "epoch": 3.8465733502999457, + "grad_norm": 0.15820272266864777, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0255, + "step": 42320 + }, + { + "epoch": 3.8474822759498273, + "grad_norm": 0.13107752799987793, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0267, + "step": 42330 + }, + { + "epoch": 3.8483912015997093, + "grad_norm": 0.13565194606781006, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0258, + "step": 42340 + }, + { + "epoch": 3.849300127249591, + "grad_norm": 0.16144689917564392, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0256, + "step": 42350 + }, + { + "epoch": 3.850209052899473, + "grad_norm": 0.11686809360980988, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0242, + "step": 42360 + }, + { + "epoch": 3.851117978549355, + "grad_norm": 0.1366736888885498, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0328, + "step": 42370 + }, + { + "epoch": 3.8520269041992368, + "grad_norm": 0.17195585370063782, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0293, + "step": 42380 + }, + { + "epoch": 3.8529358298491183, + "grad_norm": 0.17222458124160767, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0279, + "step": 42390 + }, + { + "epoch": 3.8538447554990003, + "grad_norm": 0.149159774184227, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0246, + "step": 42400 + }, + { + "epoch": 3.854753681148882, + "grad_norm": 0.1608327031135559, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0253, + "step": 42410 + }, + { + "epoch": 3.855662606798764, + "grad_norm": 0.1363644301891327, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0295, + "step": 42420 + }, + { + "epoch": 3.856571532448646, + "grad_norm": 0.15621614456176758, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0284, + "step": 42430 + }, + { + "epoch": 3.857480458098528, + "grad_norm": 0.12482628971338272, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0295, + "step": 42440 + }, + { + "epoch": 3.8583893837484093, + "grad_norm": 0.15926453471183777, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0282, + "step": 42450 + }, + { + "epoch": 3.8592983093982913, + "grad_norm": 0.16044007241725922, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0286, + "step": 42460 + }, + { + "epoch": 3.860207235048173, + "grad_norm": 0.1822386384010315, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0291, + "step": 42470 + }, + { + "epoch": 3.861116160698055, + "grad_norm": 0.10121404379606247, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0248, + "step": 42480 + }, + { + "epoch": 3.862025086347937, + "grad_norm": 0.13921326398849487, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0282, + "step": 42490 + }, + { + "epoch": 3.862934011997819, + "grad_norm": 0.11962001770734787, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0276, + "step": 42500 + }, + { + "epoch": 3.8638429376477004, + "grad_norm": 0.11136838793754578, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0265, + "step": 42510 + }, + { + "epoch": 3.8647518632975824, + "grad_norm": 0.11763390153646469, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0275, + "step": 42520 + }, + { + "epoch": 3.865660788947464, + "grad_norm": 0.11132112890481949, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0283, + "step": 42530 + }, + { + "epoch": 3.866569714597346, + "grad_norm": 0.13176578283309937, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0295, + "step": 42540 + }, + { + "epoch": 3.867478640247228, + "grad_norm": 0.09530481696128845, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0278, + "step": 42550 + }, + { + "epoch": 3.86838756589711, + "grad_norm": 0.160922572016716, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0282, + "step": 42560 + }, + { + "epoch": 3.8692964915469914, + "grad_norm": 0.13783100247383118, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0272, + "step": 42570 + }, + { + "epoch": 3.8702054171968734, + "grad_norm": 0.1387844830751419, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0289, + "step": 42580 + }, + { + "epoch": 3.871114342846755, + "grad_norm": 0.12711885571479797, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0274, + "step": 42590 + }, + { + "epoch": 3.872023268496637, + "grad_norm": 0.12908776104450226, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0268, + "step": 42600 + }, + { + "epoch": 3.872932194146519, + "grad_norm": 0.14787153899669647, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0258, + "step": 42610 + }, + { + "epoch": 3.873841119796401, + "grad_norm": 0.13165831565856934, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0258, + "step": 42620 + }, + { + "epoch": 3.8747500454462824, + "grad_norm": 0.12310001254081726, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0287, + "step": 42630 + }, + { + "epoch": 3.8756589710961644, + "grad_norm": 0.1762830913066864, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0284, + "step": 42640 + }, + { + "epoch": 3.876567896746046, + "grad_norm": 0.12024302035570145, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0253, + "step": 42650 + }, + { + "epoch": 3.877476822395928, + "grad_norm": 0.151494562625885, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0261, + "step": 42660 + }, + { + "epoch": 3.87838574804581, + "grad_norm": 0.14494360983371735, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0274, + "step": 42670 + }, + { + "epoch": 3.879294673695692, + "grad_norm": 0.10937272012233734, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0259, + "step": 42680 + }, + { + "epoch": 3.8802035993455735, + "grad_norm": 0.12404001504182816, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0259, + "step": 42690 + }, + { + "epoch": 3.8811125249954554, + "grad_norm": 0.13292115926742554, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0273, + "step": 42700 + }, + { + "epoch": 3.882021450645337, + "grad_norm": 0.13353636860847473, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0258, + "step": 42710 + }, + { + "epoch": 3.882930376295219, + "grad_norm": 0.16566789150238037, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0265, + "step": 42720 + }, + { + "epoch": 3.883839301945101, + "grad_norm": 0.1724608987569809, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.028, + "step": 42730 + }, + { + "epoch": 3.884748227594983, + "grad_norm": 0.14513692259788513, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0282, + "step": 42740 + }, + { + "epoch": 3.8856571532448645, + "grad_norm": 0.1350686252117157, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0253, + "step": 42750 + }, + { + "epoch": 3.8865660788947465, + "grad_norm": 0.20253950357437134, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0236, + "step": 42760 + }, + { + "epoch": 3.887475004544628, + "grad_norm": 0.12726160883903503, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0274, + "step": 42770 + }, + { + "epoch": 3.88838393019451, + "grad_norm": 0.19159065186977386, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0277, + "step": 42780 + }, + { + "epoch": 3.889292855844392, + "grad_norm": 0.1445232480764389, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0274, + "step": 42790 + }, + { + "epoch": 3.890201781494274, + "grad_norm": 0.1259576678276062, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0252, + "step": 42800 + }, + { + "epoch": 3.8911107071441555, + "grad_norm": 0.11373864859342575, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0245, + "step": 42810 + }, + { + "epoch": 3.8920196327940375, + "grad_norm": 0.1267562210559845, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0274, + "step": 42820 + }, + { + "epoch": 3.892928558443919, + "grad_norm": 0.1607062667608261, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0252, + "step": 42830 + }, + { + "epoch": 3.893837484093801, + "grad_norm": 0.1272663176059723, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0265, + "step": 42840 + }, + { + "epoch": 3.894746409743683, + "grad_norm": 0.21656577289104462, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0278, + "step": 42850 + }, + { + "epoch": 3.895655335393565, + "grad_norm": 0.1392773985862732, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0239, + "step": 42860 + }, + { + "epoch": 3.8965642610434466, + "grad_norm": 0.11294852197170258, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0251, + "step": 42870 + }, + { + "epoch": 3.8974731866933285, + "grad_norm": 0.21716858446598053, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0286, + "step": 42880 + }, + { + "epoch": 3.89838211234321, + "grad_norm": 0.18818466365337372, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0258, + "step": 42890 + }, + { + "epoch": 3.899291037993092, + "grad_norm": 0.14022672176361084, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0284, + "step": 42900 + }, + { + "epoch": 3.900199963642974, + "grad_norm": 0.1245957612991333, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0243, + "step": 42910 + }, + { + "epoch": 3.901108889292856, + "grad_norm": 0.20438127219676971, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0294, + "step": 42920 + }, + { + "epoch": 3.9020178149427376, + "grad_norm": 0.12497731298208237, + "learning_rate": 2.288008171679078e-05, + "loss": 0.0271, + "step": 42930 + }, + { + "epoch": 3.9029267405926196, + "grad_norm": 0.13625502586364746, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0267, + "step": 42940 + }, + { + "epoch": 3.903835666242501, + "grad_norm": 0.11935462057590485, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0269, + "step": 42950 + }, + { + "epoch": 3.904744591892383, + "grad_norm": 0.16179701685905457, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0254, + "step": 42960 + }, + { + "epoch": 3.905653517542265, + "grad_norm": 0.128456711769104, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0294, + "step": 42970 + }, + { + "epoch": 3.906562443192147, + "grad_norm": 0.11728266626596451, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0241, + "step": 42980 + }, + { + "epoch": 3.9074713688420286, + "grad_norm": 0.17803549766540527, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0264, + "step": 42990 + }, + { + "epoch": 3.9083802944919106, + "grad_norm": 0.11284089088439941, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0275, + "step": 43000 + }, + { + "epoch": 3.909289220141792, + "grad_norm": 0.15495921671390533, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0299, + "step": 43010 + }, + { + "epoch": 3.910198145791674, + "grad_norm": 0.1882009506225586, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0299, + "step": 43020 + }, + { + "epoch": 3.911107071441556, + "grad_norm": 0.15633678436279297, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0282, + "step": 43030 + }, + { + "epoch": 3.912015997091438, + "grad_norm": 0.11802812665700912, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0281, + "step": 43040 + }, + { + "epoch": 3.9129249227413196, + "grad_norm": 0.14000460505485535, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0277, + "step": 43050 + }, + { + "epoch": 3.9138338483912016, + "grad_norm": 0.15038225054740906, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.027, + "step": 43060 + }, + { + "epoch": 3.914742774041083, + "grad_norm": 0.1474815309047699, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0276, + "step": 43070 + }, + { + "epoch": 3.915651699690965, + "grad_norm": 0.11290230602025986, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0265, + "step": 43080 + }, + { + "epoch": 3.916560625340847, + "grad_norm": 0.10909006744623184, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0272, + "step": 43090 + }, + { + "epoch": 3.917469550990729, + "grad_norm": 0.13664332032203674, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0283, + "step": 43100 + }, + { + "epoch": 3.9183784766406107, + "grad_norm": 0.10853813588619232, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0248, + "step": 43110 + }, + { + "epoch": 3.9192874022904927, + "grad_norm": 0.11896227300167084, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0253, + "step": 43120 + }, + { + "epoch": 3.920196327940374, + "grad_norm": 0.15054559707641602, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0282, + "step": 43130 + }, + { + "epoch": 3.921105253590256, + "grad_norm": 0.16059096157550812, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0277, + "step": 43140 + }, + { + "epoch": 3.922014179240138, + "grad_norm": 0.15700112283229828, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0265, + "step": 43150 + }, + { + "epoch": 3.92292310489002, + "grad_norm": 0.13222353160381317, + "learning_rate": 2.27990257177228e-05, + "loss": 0.028, + "step": 43160 + }, + { + "epoch": 3.9238320305399017, + "grad_norm": 0.13032881915569305, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0297, + "step": 43170 + }, + { + "epoch": 3.9247409561897837, + "grad_norm": 0.17035125195980072, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0281, + "step": 43180 + }, + { + "epoch": 3.9256498818396657, + "grad_norm": 0.12085377424955368, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0251, + "step": 43190 + }, + { + "epoch": 3.926558807489547, + "grad_norm": 0.16110822558403015, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0286, + "step": 43200 + }, + { + "epoch": 3.927467733139429, + "grad_norm": 0.14582443237304688, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0244, + "step": 43210 + }, + { + "epoch": 3.928376658789311, + "grad_norm": 0.1449993997812271, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.032, + "step": 43220 + }, + { + "epoch": 3.9292855844391927, + "grad_norm": 0.1457662135362625, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0276, + "step": 43230 + }, + { + "epoch": 3.9301945100890747, + "grad_norm": 0.18728898465633392, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0253, + "step": 43240 + }, + { + "epoch": 3.9311034357389567, + "grad_norm": 0.1336599439382553, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0268, + "step": 43250 + }, + { + "epoch": 3.9320123613888383, + "grad_norm": 0.20802932977676392, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.027, + "step": 43260 + }, + { + "epoch": 3.9329212870387202, + "grad_norm": 0.12513644993305206, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0302, + "step": 43270 + }, + { + "epoch": 3.9338302126886022, + "grad_norm": 0.14574569463729858, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0273, + "step": 43280 + }, + { + "epoch": 3.9347391383384838, + "grad_norm": 0.11747688055038452, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0259, + "step": 43290 + }, + { + "epoch": 3.9356480639883658, + "grad_norm": 0.12489935010671616, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0247, + "step": 43300 + }, + { + "epoch": 3.9365569896382477, + "grad_norm": 0.13282886147499084, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0248, + "step": 43310 + }, + { + "epoch": 3.9374659152881293, + "grad_norm": 0.146922305226326, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0267, + "step": 43320 + }, + { + "epoch": 3.9383748409380113, + "grad_norm": 0.15413887798786163, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0264, + "step": 43330 + }, + { + "epoch": 3.9392837665878933, + "grad_norm": 0.10832811892032623, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0272, + "step": 43340 + }, + { + "epoch": 3.940192692237775, + "grad_norm": 0.1379302740097046, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0272, + "step": 43350 + }, + { + "epoch": 3.941101617887657, + "grad_norm": 0.14759130775928497, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0256, + "step": 43360 + }, + { + "epoch": 3.9420105435375388, + "grad_norm": 0.121646948158741, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0271, + "step": 43370 + }, + { + "epoch": 3.9429194691874203, + "grad_norm": 0.21311326324939728, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0285, + "step": 43380 + }, + { + "epoch": 3.9438283948373023, + "grad_norm": 0.1165025383234024, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.025, + "step": 43390 + }, + { + "epoch": 3.9447373204871843, + "grad_norm": 0.17252206802368164, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0262, + "step": 43400 + }, + { + "epoch": 3.945646246137066, + "grad_norm": 0.17383162677288055, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0273, + "step": 43410 + }, + { + "epoch": 3.946555171786948, + "grad_norm": 0.14667174220085144, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0245, + "step": 43420 + }, + { + "epoch": 3.94746409743683, + "grad_norm": 0.1493997871875763, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0252, + "step": 43430 + }, + { + "epoch": 3.9483730230867113, + "grad_norm": 0.13705803453922272, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0249, + "step": 43440 + }, + { + "epoch": 3.9492819487365933, + "grad_norm": 0.2384645938873291, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0234, + "step": 43450 + }, + { + "epoch": 3.9501908743864753, + "grad_norm": 0.13732290267944336, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0279, + "step": 43460 + }, + { + "epoch": 3.9510998000363573, + "grad_norm": 0.1533399373292923, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0257, + "step": 43470 + }, + { + "epoch": 3.952008725686239, + "grad_norm": 0.16905531287193298, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0267, + "step": 43480 + }, + { + "epoch": 3.952917651336121, + "grad_norm": 0.13507786393165588, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0267, + "step": 43490 + }, + { + "epoch": 3.9538265769860024, + "grad_norm": 0.1527792364358902, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0272, + "step": 43500 + }, + { + "epoch": 3.9547355026358844, + "grad_norm": 0.15040715038776398, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0275, + "step": 43510 + }, + { + "epoch": 3.9556444282857663, + "grad_norm": 0.1570412814617157, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0285, + "step": 43520 + }, + { + "epoch": 3.9565533539356483, + "grad_norm": 0.11572767049074173, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0264, + "step": 43530 + }, + { + "epoch": 3.95746227958553, + "grad_norm": 0.15863031148910522, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0273, + "step": 43540 + }, + { + "epoch": 3.958371205235412, + "grad_norm": 0.13130788505077362, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0273, + "step": 43550 + }, + { + "epoch": 3.9592801308852934, + "grad_norm": 0.1376325488090515, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0288, + "step": 43560 + }, + { + "epoch": 3.9601890565351754, + "grad_norm": 0.1702488511800766, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0269, + "step": 43570 + }, + { + "epoch": 3.9610979821850574, + "grad_norm": 0.13915656507015228, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0263, + "step": 43580 + }, + { + "epoch": 3.9620069078349394, + "grad_norm": 0.1659315824508667, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0292, + "step": 43590 + }, + { + "epoch": 3.962915833484821, + "grad_norm": 0.1122099906206131, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0277, + "step": 43600 + }, + { + "epoch": 3.963824759134703, + "grad_norm": 0.15412883460521698, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0266, + "step": 43610 + }, + { + "epoch": 3.9647336847845844, + "grad_norm": 0.11599072068929672, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0256, + "step": 43620 + }, + { + "epoch": 3.9656426104344664, + "grad_norm": 0.1571917086839676, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0302, + "step": 43630 + }, + { + "epoch": 3.9665515360843484, + "grad_norm": 0.1228957548737526, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0266, + "step": 43640 + }, + { + "epoch": 3.9674604617342304, + "grad_norm": 0.19058915972709656, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0281, + "step": 43650 + }, + { + "epoch": 3.968369387384112, + "grad_norm": 0.19505806267261505, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.027, + "step": 43660 + }, + { + "epoch": 3.969278313033994, + "grad_norm": 0.14038203656673431, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0238, + "step": 43670 + }, + { + "epoch": 3.9701872386838755, + "grad_norm": 0.11949088424444199, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0261, + "step": 43680 + }, + { + "epoch": 3.9710961643337575, + "grad_norm": 0.17699217796325684, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0293, + "step": 43690 + }, + { + "epoch": 3.9720050899836394, + "grad_norm": 0.1074586734175682, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0261, + "step": 43700 + }, + { + "epoch": 3.9729140156335214, + "grad_norm": 0.13453702628612518, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0273, + "step": 43710 + }, + { + "epoch": 3.973822941283403, + "grad_norm": 0.17986445128917694, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0257, + "step": 43720 + }, + { + "epoch": 3.974731866933285, + "grad_norm": 0.1432991772890091, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0264, + "step": 43730 + }, + { + "epoch": 3.9756407925831665, + "grad_norm": 0.1423931121826172, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0272, + "step": 43740 + }, + { + "epoch": 3.9765497182330485, + "grad_norm": 0.14869634807109833, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0268, + "step": 43750 + }, + { + "epoch": 3.9774586438829305, + "grad_norm": 0.19271422922611237, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0286, + "step": 43760 + }, + { + "epoch": 3.9783675695328125, + "grad_norm": 0.1479814648628235, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0273, + "step": 43770 + }, + { + "epoch": 3.979276495182694, + "grad_norm": 0.14447036385536194, + "learning_rate": 2.257897548253514e-05, + "loss": 0.027, + "step": 43780 + }, + { + "epoch": 3.980185420832576, + "grad_norm": 0.12995681166648865, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0263, + "step": 43790 + }, + { + "epoch": 3.9810943464824575, + "grad_norm": 0.13797220587730408, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0311, + "step": 43800 + }, + { + "epoch": 3.9820032721323395, + "grad_norm": 0.13113370537757874, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0264, + "step": 43810 + }, + { + "epoch": 3.9829121977822215, + "grad_norm": 0.12080135196447372, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0273, + "step": 43820 + }, + { + "epoch": 3.9838211234321035, + "grad_norm": 0.10060163587331772, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0248, + "step": 43830 + }, + { + "epoch": 3.984730049081985, + "grad_norm": 0.10419678688049316, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0287, + "step": 43840 + }, + { + "epoch": 3.985638974731867, + "grad_norm": 0.10087472945451736, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.027, + "step": 43850 + }, + { + "epoch": 3.9865479003817486, + "grad_norm": 0.15018832683563232, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0266, + "step": 43860 + }, + { + "epoch": 3.9874568260316305, + "grad_norm": 0.13440482318401337, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0268, + "step": 43870 + }, + { + "epoch": 3.9883657516815125, + "grad_norm": 0.14031872153282166, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0266, + "step": 43880 + }, + { + "epoch": 3.9892746773313945, + "grad_norm": 0.1512184888124466, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0285, + "step": 43890 + }, + { + "epoch": 3.990183602981276, + "grad_norm": 0.13491588830947876, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0267, + "step": 43900 + }, + { + "epoch": 3.991092528631158, + "grad_norm": 0.18680404126644135, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0294, + "step": 43910 + }, + { + "epoch": 3.9920014542810396, + "grad_norm": 0.13106416165828705, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0284, + "step": 43920 + }, + { + "epoch": 3.9929103799309216, + "grad_norm": 0.13720248639583588, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0262, + "step": 43930 + }, + { + "epoch": 3.9938193055808036, + "grad_norm": 0.13438192009925842, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0252, + "step": 43940 + }, + { + "epoch": 3.9947282312306855, + "grad_norm": 0.1364498883485794, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0285, + "step": 43950 + }, + { + "epoch": 3.995637156880567, + "grad_norm": 0.14517997205257416, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.026, + "step": 43960 + }, + { + "epoch": 3.996546082530449, + "grad_norm": 0.13500259816646576, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0281, + "step": 43970 + }, + { + "epoch": 3.9974550081803306, + "grad_norm": 0.17011599242687225, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0294, + "step": 43980 + }, + { + "epoch": 3.9983639338302126, + "grad_norm": 0.18532992899417877, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0282, + "step": 43990 + }, + { + "epoch": 3.9992728594800946, + "grad_norm": 0.18699239194393158, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0257, + "step": 44000 + }, + { + "epoch": 4.000181785129977, + "grad_norm": 0.15372714400291443, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0238, + "step": 44010 + }, + { + "epoch": 4.001090710779859, + "grad_norm": 0.1796838343143463, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0251, + "step": 44020 + }, + { + "epoch": 4.00199963642974, + "grad_norm": 0.11628758907318115, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0264, + "step": 44030 + }, + { + "epoch": 4.002908562079622, + "grad_norm": 0.14788252115249634, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0268, + "step": 44040 + }, + { + "epoch": 4.003817487729504, + "grad_norm": 0.12092521786689758, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0265, + "step": 44050 + }, + { + "epoch": 4.004726413379386, + "grad_norm": 0.14072316884994507, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0284, + "step": 44060 + }, + { + "epoch": 4.005635339029268, + "grad_norm": 0.12703250348567963, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0261, + "step": 44070 + }, + { + "epoch": 4.00654426467915, + "grad_norm": 0.13166096806526184, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.026, + "step": 44080 + }, + { + "epoch": 4.007453190329031, + "grad_norm": 0.21356548368930817, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0276, + "step": 44090 + }, + { + "epoch": 4.008362115978913, + "grad_norm": 0.13660360872745514, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0261, + "step": 44100 + }, + { + "epoch": 4.009271041628795, + "grad_norm": 0.184359610080719, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.029, + "step": 44110 + }, + { + "epoch": 4.010179967278677, + "grad_norm": 0.12281668931245804, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0265, + "step": 44120 + }, + { + "epoch": 4.011088892928559, + "grad_norm": 0.18271246552467346, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0274, + "step": 44130 + }, + { + "epoch": 4.011997818578441, + "grad_norm": 0.1632125973701477, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0253, + "step": 44140 + }, + { + "epoch": 4.012906744228322, + "grad_norm": 0.17776349186897278, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0283, + "step": 44150 + }, + { + "epoch": 4.013815669878204, + "grad_norm": 0.1335531324148178, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0287, + "step": 44160 + }, + { + "epoch": 4.014724595528086, + "grad_norm": 0.12688252329826355, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0288, + "step": 44170 + }, + { + "epoch": 4.015633521177968, + "grad_norm": 0.1284332424402237, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0257, + "step": 44180 + }, + { + "epoch": 4.01654244682785, + "grad_norm": 0.13166747987270355, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0252, + "step": 44190 + }, + { + "epoch": 4.017451372477732, + "grad_norm": 0.14059440791606903, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0255, + "step": 44200 + }, + { + "epoch": 4.018360298127613, + "grad_norm": 0.14129337668418884, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0255, + "step": 44210 + }, + { + "epoch": 4.019269223777495, + "grad_norm": 0.12093551456928253, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0275, + "step": 44220 + }, + { + "epoch": 4.020178149427377, + "grad_norm": 0.14262054860591888, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0252, + "step": 44230 + }, + { + "epoch": 4.021087075077259, + "grad_norm": 0.14918287098407745, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0289, + "step": 44240 + }, + { + "epoch": 4.021996000727141, + "grad_norm": 0.1310160607099533, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0283, + "step": 44250 + }, + { + "epoch": 4.022904926377023, + "grad_norm": 0.14735300838947296, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0274, + "step": 44260 + }, + { + "epoch": 4.023813852026904, + "grad_norm": 0.1291436403989792, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0254, + "step": 44270 + }, + { + "epoch": 4.024722777676786, + "grad_norm": 0.20697437226772308, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0282, + "step": 44280 + }, + { + "epoch": 4.025631703326668, + "grad_norm": 0.14001218974590302, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0275, + "step": 44290 + }, + { + "epoch": 4.02654062897655, + "grad_norm": 0.12951034307479858, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.025, + "step": 44300 + }, + { + "epoch": 4.027449554626432, + "grad_norm": 0.12512974441051483, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0233, + "step": 44310 + }, + { + "epoch": 4.028358480276314, + "grad_norm": 0.13512425124645233, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.03, + "step": 44320 + }, + { + "epoch": 4.029267405926195, + "grad_norm": 0.13545191287994385, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0266, + "step": 44330 + }, + { + "epoch": 4.030176331576077, + "grad_norm": 0.1552325040102005, + "learning_rate": 2.237831698974614e-05, + "loss": 0.025, + "step": 44340 + }, + { + "epoch": 4.031085257225959, + "grad_norm": 0.16164591908454895, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0259, + "step": 44350 + }, + { + "epoch": 4.031994182875841, + "grad_norm": 0.13946840167045593, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0275, + "step": 44360 + }, + { + "epoch": 4.032903108525723, + "grad_norm": 0.16019003093242645, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0263, + "step": 44370 + }, + { + "epoch": 4.033812034175605, + "grad_norm": 0.17749890685081482, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0275, + "step": 44380 + }, + { + "epoch": 4.034720959825486, + "grad_norm": 0.12562337517738342, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0242, + "step": 44390 + }, + { + "epoch": 4.035629885475368, + "grad_norm": 0.1314900517463684, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0283, + "step": 44400 + }, + { + "epoch": 4.03653881112525, + "grad_norm": 0.17160025238990784, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0276, + "step": 44410 + }, + { + "epoch": 4.037447736775132, + "grad_norm": 0.14989930391311646, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0288, + "step": 44420 + }, + { + "epoch": 4.038356662425014, + "grad_norm": 0.1861855536699295, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0261, + "step": 44430 + }, + { + "epoch": 4.039265588074896, + "grad_norm": 0.17626342177391052, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0263, + "step": 44440 + }, + { + "epoch": 4.040174513724777, + "grad_norm": 0.14545413851737976, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0272, + "step": 44450 + }, + { + "epoch": 4.041083439374659, + "grad_norm": 0.125377357006073, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0263, + "step": 44460 + }, + { + "epoch": 4.041992365024541, + "grad_norm": 0.12247419357299805, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0276, + "step": 44470 + }, + { + "epoch": 4.042901290674423, + "grad_norm": 0.15176501870155334, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0251, + "step": 44480 + }, + { + "epoch": 4.043810216324305, + "grad_norm": 0.18067090213298798, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0248, + "step": 44490 + }, + { + "epoch": 4.044719141974187, + "grad_norm": 0.12708473205566406, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0281, + "step": 44500 + }, + { + "epoch": 4.045628067624069, + "grad_norm": 0.14948686957359314, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0247, + "step": 44510 + }, + { + "epoch": 4.04653699327395, + "grad_norm": 0.14721429347991943, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0245, + "step": 44520 + }, + { + "epoch": 4.047445918923832, + "grad_norm": 0.15841318666934967, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0292, + "step": 44530 + }, + { + "epoch": 4.048354844573714, + "grad_norm": 0.1813308745622635, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0284, + "step": 44540 + }, + { + "epoch": 4.049263770223596, + "grad_norm": 0.16217462718486786, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0276, + "step": 44550 + }, + { + "epoch": 4.050172695873478, + "grad_norm": 0.13965144753456116, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0275, + "step": 44560 + }, + { + "epoch": 4.05108162152336, + "grad_norm": 0.14742229878902435, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0295, + "step": 44570 + }, + { + "epoch": 4.051990547173241, + "grad_norm": 0.16086451709270477, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0255, + "step": 44580 + }, + { + "epoch": 4.052899472823123, + "grad_norm": 0.11854033917188644, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0252, + "step": 44590 + }, + { + "epoch": 4.053808398473005, + "grad_norm": 0.12566900253295898, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0238, + "step": 44600 + }, + { + "epoch": 4.054717324122887, + "grad_norm": 0.12489896267652512, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.0251, + "step": 44610 + }, + { + "epoch": 4.055626249772769, + "grad_norm": 0.09959734231233597, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0267, + "step": 44620 + }, + { + "epoch": 4.056535175422651, + "grad_norm": 0.12790058553218842, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0272, + "step": 44630 + }, + { + "epoch": 4.057444101072532, + "grad_norm": 0.17297552525997162, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0241, + "step": 44640 + }, + { + "epoch": 4.058353026722414, + "grad_norm": 0.14175081253051758, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0279, + "step": 44650 + }, + { + "epoch": 4.059261952372296, + "grad_norm": 0.10205555707216263, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0241, + "step": 44660 + }, + { + "epoch": 4.060170878022178, + "grad_norm": 0.13349147140979767, + "learning_rate": 2.225924589053685e-05, + "loss": 0.028, + "step": 44670 + }, + { + "epoch": 4.06107980367206, + "grad_norm": 0.1916884183883667, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0308, + "step": 44680 + }, + { + "epoch": 4.061988729321942, + "grad_norm": 0.12500861287117004, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.029, + "step": 44690 + }, + { + "epoch": 4.062897654971823, + "grad_norm": 0.18462827801704407, + "learning_rate": 2.224839132723868e-05, + "loss": 0.028, + "step": 44700 + }, + { + "epoch": 4.063806580621705, + "grad_norm": 0.1554311215877533, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0256, + "step": 44710 + }, + { + "epoch": 4.064715506271587, + "grad_norm": 0.13268442451953888, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0309, + "step": 44720 + }, + { + "epoch": 4.065624431921469, + "grad_norm": 0.12585653364658356, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0272, + "step": 44730 + }, + { + "epoch": 4.066533357571351, + "grad_norm": 0.11416815966367722, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0263, + "step": 44740 + }, + { + "epoch": 4.067442283221233, + "grad_norm": 0.16098552942276, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0244, + "step": 44750 + }, + { + "epoch": 4.068351208871114, + "grad_norm": 0.14432840049266815, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0246, + "step": 44760 + }, + { + "epoch": 4.069260134520996, + "grad_norm": 0.1530766785144806, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0251, + "step": 44770 + }, + { + "epoch": 4.070169060170878, + "grad_norm": 0.14362990856170654, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0254, + "step": 44780 + }, + { + "epoch": 4.07107798582076, + "grad_norm": 0.16490259766578674, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0259, + "step": 44790 + }, + { + "epoch": 4.071986911470642, + "grad_norm": 0.13982972502708435, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0249, + "step": 44800 + }, + { + "epoch": 4.072895837120524, + "grad_norm": 0.1906231939792633, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0285, + "step": 44810 + }, + { + "epoch": 4.073804762770405, + "grad_norm": 0.13522249460220337, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0252, + "step": 44820 + }, + { + "epoch": 4.074713688420287, + "grad_norm": 0.14539359509944916, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0269, + "step": 44830 + }, + { + "epoch": 4.075622614070169, + "grad_norm": 0.12148576974868774, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0251, + "step": 44840 + }, + { + "epoch": 4.076531539720051, + "grad_norm": 0.11599826067686081, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0264, + "step": 44850 + }, + { + "epoch": 4.077440465369933, + "grad_norm": 0.1408497840166092, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0248, + "step": 44860 + }, + { + "epoch": 4.078349391019815, + "grad_norm": 0.1269199252128601, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0251, + "step": 44870 + }, + { + "epoch": 4.079258316669696, + "grad_norm": 0.13460765779018402, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0267, + "step": 44880 + }, + { + "epoch": 4.080167242319578, + "grad_norm": 0.14702807366847992, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0269, + "step": 44890 + }, + { + "epoch": 4.08107616796946, + "grad_norm": 0.1372494399547577, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0276, + "step": 44900 + }, + { + "epoch": 4.081985093619342, + "grad_norm": 0.1998899281024933, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.027, + "step": 44910 + }, + { + "epoch": 4.082894019269224, + "grad_norm": 0.1583947390317917, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0264, + "step": 44920 + }, + { + "epoch": 4.083802944919106, + "grad_norm": 0.16781561076641083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0261, + "step": 44930 + }, + { + "epoch": 4.084711870568987, + "grad_norm": 0.17467816174030304, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.027, + "step": 44940 + }, + { + "epoch": 4.085620796218869, + "grad_norm": 0.14679527282714844, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0261, + "step": 44950 + }, + { + "epoch": 4.086529721868751, + "grad_norm": 0.10288897901773453, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0255, + "step": 44960 + }, + { + "epoch": 4.087438647518633, + "grad_norm": 0.12804001569747925, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0241, + "step": 44970 + }, + { + "epoch": 4.088347573168515, + "grad_norm": 0.167776420712471, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0241, + "step": 44980 + }, + { + "epoch": 4.089256498818397, + "grad_norm": 0.12680503726005554, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0265, + "step": 44990 + }, + { + "epoch": 4.090165424468278, + "grad_norm": 0.13705958425998688, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.029, + "step": 45000 + }, + { + "epoch": 4.09107435011816, + "grad_norm": 0.12949027121067047, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0286, + "step": 45010 + }, + { + "epoch": 4.091983275768042, + "grad_norm": 0.135112002491951, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0285, + "step": 45020 + }, + { + "epoch": 4.092892201417924, + "grad_norm": 0.13345032930374146, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0267, + "step": 45030 + }, + { + "epoch": 4.093801127067806, + "grad_norm": 0.16293010115623474, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.031, + "step": 45040 + }, + { + "epoch": 4.094710052717688, + "grad_norm": 0.15452122688293457, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.026, + "step": 45050 + }, + { + "epoch": 4.095618978367569, + "grad_norm": 0.1044573113322258, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0258, + "step": 45060 + }, + { + "epoch": 4.096527904017451, + "grad_norm": 0.13904882967472076, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0292, + "step": 45070 + }, + { + "epoch": 4.097436829667333, + "grad_norm": 0.11362932622432709, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0241, + "step": 45080 + }, + { + "epoch": 4.098345755317215, + "grad_norm": 0.14900213479995728, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0263, + "step": 45090 + }, + { + "epoch": 4.099254680967097, + "grad_norm": 0.13509003818035126, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0247, + "step": 45100 + }, + { + "epoch": 4.100163606616979, + "grad_norm": 0.14763566851615906, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.028, + "step": 45110 + }, + { + "epoch": 4.10107253226686, + "grad_norm": 0.15205834805965424, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0248, + "step": 45120 + }, + { + "epoch": 4.101981457916742, + "grad_norm": 0.12246311455965042, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0257, + "step": 45130 + }, + { + "epoch": 4.102890383566624, + "grad_norm": 0.14252328872680664, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0281, + "step": 45140 + }, + { + "epoch": 4.103799309216506, + "grad_norm": 0.14453886449337006, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0254, + "step": 45150 + }, + { + "epoch": 4.104708234866388, + "grad_norm": 0.11444026976823807, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0253, + "step": 45160 + }, + { + "epoch": 4.10561716051627, + "grad_norm": 0.10880998522043228, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0252, + "step": 45170 + }, + { + "epoch": 4.106526086166151, + "grad_norm": 0.13941386342048645, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0234, + "step": 45180 + }, + { + "epoch": 4.107435011816033, + "grad_norm": 0.18937304615974426, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0251, + "step": 45190 + }, + { + "epoch": 4.108343937465915, + "grad_norm": 0.14025649428367615, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0272, + "step": 45200 + }, + { + "epoch": 4.109252863115797, + "grad_norm": 0.15390728414058685, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0278, + "step": 45210 + }, + { + "epoch": 4.110161788765679, + "grad_norm": 0.1370895802974701, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0252, + "step": 45220 + }, + { + "epoch": 4.111070714415561, + "grad_norm": 0.17259719967842102, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0245, + "step": 45230 + }, + { + "epoch": 4.111979640065442, + "grad_norm": 0.12258259207010269, + "learning_rate": 2.205216922378357e-05, + "loss": 0.026, + "step": 45240 + }, + { + "epoch": 4.112888565715324, + "grad_norm": 0.11297283321619034, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0281, + "step": 45250 + }, + { + "epoch": 4.113797491365206, + "grad_norm": 0.1480085849761963, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0277, + "step": 45260 + }, + { + "epoch": 4.114706417015088, + "grad_norm": 0.15909701585769653, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0279, + "step": 45270 + }, + { + "epoch": 4.11561534266497, + "grad_norm": 0.11828441917896271, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0248, + "step": 45280 + }, + { + "epoch": 4.116524268314852, + "grad_norm": 0.1281217634677887, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0272, + "step": 45290 + }, + { + "epoch": 4.117433193964733, + "grad_norm": 0.1744980365037918, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0263, + "step": 45300 + }, + { + "epoch": 4.118342119614615, + "grad_norm": 0.15832506120204926, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0292, + "step": 45310 + }, + { + "epoch": 4.119251045264497, + "grad_norm": 0.13767790794372559, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0237, + "step": 45320 + }, + { + "epoch": 4.120159970914379, + "grad_norm": 0.1477966457605362, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0251, + "step": 45330 + }, + { + "epoch": 4.121068896564261, + "grad_norm": 0.15505757927894592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.027, + "step": 45340 + }, + { + "epoch": 4.121977822214143, + "grad_norm": 0.12948273122310638, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0265, + "step": 45350 + }, + { + "epoch": 4.122886747864024, + "grad_norm": 0.1512318104505539, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.024, + "step": 45360 + }, + { + "epoch": 4.123795673513906, + "grad_norm": 0.14807875454425812, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0276, + "step": 45370 + }, + { + "epoch": 4.124704599163788, + "grad_norm": 0.13582700490951538, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0255, + "step": 45380 + }, + { + "epoch": 4.12561352481367, + "grad_norm": 0.11363188922405243, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0271, + "step": 45390 + }, + { + "epoch": 4.126522450463552, + "grad_norm": 0.11392319202423096, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0252, + "step": 45400 + }, + { + "epoch": 4.127431376113434, + "grad_norm": 0.14232124388217926, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.023, + "step": 45410 + }, + { + "epoch": 4.128340301763315, + "grad_norm": 0.13991425931453705, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0258, + "step": 45420 + }, + { + "epoch": 4.129249227413197, + "grad_norm": 0.15905271470546722, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0258, + "step": 45430 + }, + { + "epoch": 4.130158153063079, + "grad_norm": 0.11344780772924423, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0236, + "step": 45440 + }, + { + "epoch": 4.131067078712961, + "grad_norm": 0.13451622426509857, + "learning_rate": 2.197543647586431e-05, + "loss": 0.027, + "step": 45450 + }, + { + "epoch": 4.131976004362843, + "grad_norm": 0.15434546768665314, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0295, + "step": 45460 + }, + { + "epoch": 4.132884930012725, + "grad_norm": 0.1398824155330658, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0261, + "step": 45470 + }, + { + "epoch": 4.133793855662606, + "grad_norm": 0.14598722755908966, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0263, + "step": 45480 + }, + { + "epoch": 4.134702781312488, + "grad_norm": 0.12265633791685104, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0255, + "step": 45490 + }, + { + "epoch": 4.13561170696237, + "grad_norm": 0.18668554723262787, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0246, + "step": 45500 + }, + { + "epoch": 4.136520632612252, + "grad_norm": 0.14463771879673004, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0252, + "step": 45510 + }, + { + "epoch": 4.137429558262134, + "grad_norm": 0.1678943783044815, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0293, + "step": 45520 + }, + { + "epoch": 4.138338483912016, + "grad_norm": 0.15413318574428558, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0241, + "step": 45530 + }, + { + "epoch": 4.139247409561898, + "grad_norm": 0.12812310457229614, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0258, + "step": 45540 + }, + { + "epoch": 4.140156335211779, + "grad_norm": 0.10812357068061829, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0277, + "step": 45550 + }, + { + "epoch": 4.141065260861661, + "grad_norm": 0.1306268274784088, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0271, + "step": 45560 + }, + { + "epoch": 4.141974186511543, + "grad_norm": 0.1395256370306015, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0284, + "step": 45570 + }, + { + "epoch": 4.142883112161425, + "grad_norm": 0.15395382046699524, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0256, + "step": 45580 + }, + { + "epoch": 4.143792037811307, + "grad_norm": 0.13091745972633362, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0263, + "step": 45590 + }, + { + "epoch": 4.144700963461188, + "grad_norm": 0.16177460551261902, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0273, + "step": 45600 + }, + { + "epoch": 4.14560988911107, + "grad_norm": 0.13999834656715393, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.027, + "step": 45610 + }, + { + "epoch": 4.146518814760952, + "grad_norm": 0.11858700215816498, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0238, + "step": 45620 + }, + { + "epoch": 4.147427740410834, + "grad_norm": 0.14288249611854553, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0258, + "step": 45630 + }, + { + "epoch": 4.148336666060716, + "grad_norm": 0.12174994498491287, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0271, + "step": 45640 + }, + { + "epoch": 4.149245591710598, + "grad_norm": 0.11106406897306442, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0255, + "step": 45650 + }, + { + "epoch": 4.15015451736048, + "grad_norm": 0.16769035160541534, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0278, + "step": 45660 + }, + { + "epoch": 4.1510634430103615, + "grad_norm": 0.1562432050704956, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0247, + "step": 45670 + }, + { + "epoch": 4.151972368660243, + "grad_norm": 0.12205801904201508, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0256, + "step": 45680 + }, + { + "epoch": 4.152881294310125, + "grad_norm": 0.1429969072341919, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0231, + "step": 45690 + }, + { + "epoch": 4.153790219960007, + "grad_norm": 0.1359611302614212, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0253, + "step": 45700 + }, + { + "epoch": 4.154699145609889, + "grad_norm": 0.17137545347213745, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0292, + "step": 45710 + }, + { + "epoch": 4.1556080712597705, + "grad_norm": 0.1434371918439865, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0248, + "step": 45720 + }, + { + "epoch": 4.1565169969096525, + "grad_norm": 0.17371854186058044, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0233, + "step": 45730 + }, + { + "epoch": 4.1574259225595345, + "grad_norm": 0.11565165221691132, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0249, + "step": 45740 + }, + { + "epoch": 4.1583348482094165, + "grad_norm": 0.12144811451435089, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0276, + "step": 45750 + }, + { + "epoch": 4.159243773859298, + "grad_norm": 0.12803246080875397, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0285, + "step": 45760 + }, + { + "epoch": 4.16015269950918, + "grad_norm": 0.13871850073337555, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0229, + "step": 45770 + }, + { + "epoch": 4.161061625159062, + "grad_norm": 0.12765651941299438, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0265, + "step": 45780 + }, + { + "epoch": 4.1619705508089435, + "grad_norm": 0.11126983165740967, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0251, + "step": 45790 + }, + { + "epoch": 4.1628794764588255, + "grad_norm": 0.14251497387886047, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0252, + "step": 45800 + }, + { + "epoch": 4.1637884021087075, + "grad_norm": 0.12540282309055328, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0255, + "step": 45810 + }, + { + "epoch": 4.1646973277585895, + "grad_norm": 0.12606380879878998, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0261, + "step": 45820 + }, + { + "epoch": 4.1656062534084715, + "grad_norm": 0.12443399429321289, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0249, + "step": 45830 + }, + { + "epoch": 4.1665151790583534, + "grad_norm": 0.255077600479126, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0261, + "step": 45840 + }, + { + "epoch": 4.1674241047082345, + "grad_norm": 0.14124029874801636, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0244, + "step": 45850 + }, + { + "epoch": 4.1683330303581165, + "grad_norm": 0.13939158618450165, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0246, + "step": 45860 + }, + { + "epoch": 4.1692419560079985, + "grad_norm": 0.11690129339694977, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0252, + "step": 45870 + }, + { + "epoch": 4.1701508816578805, + "grad_norm": 0.1483093798160553, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0254, + "step": 45880 + }, + { + "epoch": 4.1710598073077625, + "grad_norm": 0.14484979212284088, + "learning_rate": 2.181390862723362e-05, + "loss": 0.026, + "step": 45890 + }, + { + "epoch": 4.1719687329576445, + "grad_norm": 0.17399899661540985, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0275, + "step": 45900 + }, + { + "epoch": 4.172877658607526, + "grad_norm": 0.11937721073627472, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0251, + "step": 45910 + }, + { + "epoch": 4.173786584257408, + "grad_norm": 0.12316877394914627, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0268, + "step": 45920 + }, + { + "epoch": 4.1746955099072895, + "grad_norm": 0.1700224131345749, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0264, + "step": 45930 + }, + { + "epoch": 4.1756044355571715, + "grad_norm": 0.11408409476280212, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0258, + "step": 45940 + }, + { + "epoch": 4.1765133612070535, + "grad_norm": 0.12062083929777145, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0241, + "step": 45950 + }, + { + "epoch": 4.1774222868569355, + "grad_norm": 0.12455981224775314, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0278, + "step": 45960 + }, + { + "epoch": 4.178331212506817, + "grad_norm": 0.11336000263690948, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0258, + "step": 45970 + }, + { + "epoch": 4.179240138156699, + "grad_norm": 0.1751512885093689, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0302, + "step": 45980 + }, + { + "epoch": 4.180149063806581, + "grad_norm": 0.11169908940792084, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0224, + "step": 45990 + }, + { + "epoch": 4.181057989456463, + "grad_norm": 0.11132428050041199, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.028, + "step": 46000 + }, + { + "epoch": 4.1819669151063446, + "grad_norm": 0.1379801332950592, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0243, + "step": 46010 + }, + { + "epoch": 4.1828758407562265, + "grad_norm": 0.13539282977581024, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0253, + "step": 46020 + }, + { + "epoch": 4.183784766406108, + "grad_norm": 0.13076196610927582, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.025, + "step": 46030 + }, + { + "epoch": 4.18469369205599, + "grad_norm": 0.14952963590621948, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0276, + "step": 46040 + }, + { + "epoch": 4.185602617705872, + "grad_norm": 0.11460956186056137, + "learning_rate": 2.175492159090619e-05, + "loss": 0.0266, + "step": 46050 + }, + { + "epoch": 4.186511543355754, + "grad_norm": 0.13483422994613647, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0252, + "step": 46060 + }, + { + "epoch": 4.187420469005636, + "grad_norm": 0.1361168473958969, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0272, + "step": 46070 + }, + { + "epoch": 4.188329394655518, + "grad_norm": 0.11320122331380844, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0243, + "step": 46080 + }, + { + "epoch": 4.189238320305399, + "grad_norm": 0.12508752942085266, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0229, + "step": 46090 + }, + { + "epoch": 4.190147245955281, + "grad_norm": 0.15086929500102997, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0243, + "step": 46100 + }, + { + "epoch": 4.191056171605163, + "grad_norm": 0.17090089619159698, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0272, + "step": 46110 + }, + { + "epoch": 4.191965097255045, + "grad_norm": 0.14506708085536957, + "learning_rate": 2.172907339903944e-05, + "loss": 0.027, + "step": 46120 + }, + { + "epoch": 4.192874022904927, + "grad_norm": 0.12666864693164825, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0234, + "step": 46130 + }, + { + "epoch": 4.193782948554809, + "grad_norm": 0.13621409237384796, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0243, + "step": 46140 + }, + { + "epoch": 4.19469187420469, + "grad_norm": 0.16440102458000183, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0269, + "step": 46150 + }, + { + "epoch": 4.195600799854572, + "grad_norm": 0.13968904316425323, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0258, + "step": 46160 + }, + { + "epoch": 4.196509725504454, + "grad_norm": 0.14871719479560852, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0267, + "step": 46170 + }, + { + "epoch": 4.197418651154336, + "grad_norm": 0.19163751602172852, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0291, + "step": 46180 + }, + { + "epoch": 4.198327576804218, + "grad_norm": 0.12853702902793884, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0223, + "step": 46190 + }, + { + "epoch": 4.1992365024541, + "grad_norm": 0.15203161537647247, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.026, + "step": 46200 + }, + { + "epoch": 4.200145428103981, + "grad_norm": 0.1500093936920166, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0248, + "step": 46210 + }, + { + "epoch": 4.201054353753863, + "grad_norm": 0.16525736451148987, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0247, + "step": 46220 + }, + { + "epoch": 4.201963279403745, + "grad_norm": 0.1528998166322708, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0258, + "step": 46230 + }, + { + "epoch": 4.202872205053627, + "grad_norm": 0.1472453624010086, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0276, + "step": 46240 + }, + { + "epoch": 4.203781130703509, + "grad_norm": 0.1203448623418808, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0229, + "step": 46250 + }, + { + "epoch": 4.204690056353391, + "grad_norm": 0.17540965974330902, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0264, + "step": 46260 + }, + { + "epoch": 4.205598982003272, + "grad_norm": 0.1451127976179123, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0261, + "step": 46270 + }, + { + "epoch": 4.206507907653154, + "grad_norm": 0.1312674731016159, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0252, + "step": 46280 + }, + { + "epoch": 4.207416833303036, + "grad_norm": 0.14758551120758057, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0229, + "step": 46290 + }, + { + "epoch": 4.208325758952918, + "grad_norm": 0.12517240643501282, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0282, + "step": 46300 + }, + { + "epoch": 4.2092346846028, + "grad_norm": 0.13322308659553528, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0253, + "step": 46310 + }, + { + "epoch": 4.210143610252682, + "grad_norm": 0.12882770597934723, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.027, + "step": 46320 + }, + { + "epoch": 4.211052535902563, + "grad_norm": 0.1268293708562851, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0246, + "step": 46330 + }, + { + "epoch": 4.211961461552445, + "grad_norm": 0.1421269327402115, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0282, + "step": 46340 + }, + { + "epoch": 4.212870387202327, + "grad_norm": 0.1365530639886856, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0239, + "step": 46350 + }, + { + "epoch": 4.213779312852209, + "grad_norm": 0.12586946785449982, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0258, + "step": 46360 + }, + { + "epoch": 4.214688238502091, + "grad_norm": 0.14051774144172668, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0264, + "step": 46370 + }, + { + "epoch": 4.215597164151973, + "grad_norm": 0.1684417873620987, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0267, + "step": 46380 + }, + { + "epoch": 4.216506089801854, + "grad_norm": 0.15807045996189117, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0259, + "step": 46390 + }, + { + "epoch": 4.217415015451736, + "grad_norm": 0.1614532768726349, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0261, + "step": 46400 + }, + { + "epoch": 4.218323941101618, + "grad_norm": 0.1274830549955368, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0264, + "step": 46410 + }, + { + "epoch": 4.2192328667515, + "grad_norm": 0.11421353369951248, + "learning_rate": 2.161801300808089e-05, + "loss": 0.024, + "step": 46420 + }, + { + "epoch": 4.220141792401382, + "grad_norm": 0.15279626846313477, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0265, + "step": 46430 + }, + { + "epoch": 4.221050718051264, + "grad_norm": 0.23922449350357056, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0266, + "step": 46440 + }, + { + "epoch": 4.221959643701145, + "grad_norm": 0.13655468821525574, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0267, + "step": 46450 + }, + { + "epoch": 4.222868569351027, + "grad_norm": 0.1552736610174179, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0255, + "step": 46460 + }, + { + "epoch": 4.223777495000909, + "grad_norm": 0.12029021978378296, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0287, + "step": 46470 + }, + { + "epoch": 4.224686420650791, + "grad_norm": 0.14615042507648468, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0292, + "step": 46480 + }, + { + "epoch": 4.225595346300673, + "grad_norm": 0.13133762776851654, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0242, + "step": 46490 + }, + { + "epoch": 4.226504271950555, + "grad_norm": 0.13962969183921814, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0256, + "step": 46500 + }, + { + "epoch": 4.227413197600436, + "grad_norm": 0.11827094107866287, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.024, + "step": 46510 + }, + { + "epoch": 4.228322123250318, + "grad_norm": 0.13072067499160767, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.026, + "step": 46520 + }, + { + "epoch": 4.2292310489002, + "grad_norm": 0.13815298676490784, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0249, + "step": 46530 + }, + { + "epoch": 4.230139974550082, + "grad_norm": 0.13036300241947174, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0294, + "step": 46540 + }, + { + "epoch": 4.231048900199964, + "grad_norm": 0.1296500265598297, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0242, + "step": 46550 + }, + { + "epoch": 4.231957825849846, + "grad_norm": 0.10897675901651382, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0256, + "step": 46560 + }, + { + "epoch": 4.232866751499727, + "grad_norm": 0.1317068338394165, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0243, + "step": 46570 + }, + { + "epoch": 4.233775677149609, + "grad_norm": 0.11576242744922638, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0252, + "step": 46580 + }, + { + "epoch": 4.234684602799491, + "grad_norm": 0.18941359221935272, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0257, + "step": 46590 + }, + { + "epoch": 4.235593528449373, + "grad_norm": 0.15053609013557434, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0267, + "step": 46600 + }, + { + "epoch": 4.236502454099255, + "grad_norm": 0.13006021082401276, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0238, + "step": 46610 + }, + { + "epoch": 4.237411379749137, + "grad_norm": 0.11921507120132446, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0275, + "step": 46620 + }, + { + "epoch": 4.238320305399018, + "grad_norm": 0.1216982752084732, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0245, + "step": 46630 + }, + { + "epoch": 4.2392292310489, + "grad_norm": 0.12324576824903488, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0259, + "step": 46640 + }, + { + "epoch": 4.240138156698782, + "grad_norm": 0.17033430933952332, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0258, + "step": 46650 + }, + { + "epoch": 4.241047082348664, + "grad_norm": 0.09964343160390854, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0257, + "step": 46660 + }, + { + "epoch": 4.241956007998546, + "grad_norm": 0.11091228574514389, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0238, + "step": 46670 + }, + { + "epoch": 4.242864933648428, + "grad_norm": 0.12458638846874237, + "learning_rate": 2.152139450890464e-05, + "loss": 0.028, + "step": 46680 + }, + { + "epoch": 4.243773859298309, + "grad_norm": 0.13776181638240814, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0235, + "step": 46690 + }, + { + "epoch": 4.244682784948191, + "grad_norm": 0.15018677711486816, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0252, + "step": 46700 + }, + { + "epoch": 4.245591710598073, + "grad_norm": 0.1261228621006012, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0229, + "step": 46710 + }, + { + "epoch": 4.246500636247955, + "grad_norm": 0.13367345929145813, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0252, + "step": 46720 + }, + { + "epoch": 4.247409561897837, + "grad_norm": 0.1288018822669983, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0267, + "step": 46730 + }, + { + "epoch": 4.248318487547719, + "grad_norm": 0.1316433846950531, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0216, + "step": 46740 + }, + { + "epoch": 4.2492274131976, + "grad_norm": 0.13376715779304504, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0258, + "step": 46750 + }, + { + "epoch": 4.250136338847482, + "grad_norm": 0.1421327292919159, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0264, + "step": 46760 + }, + { + "epoch": 4.251045264497364, + "grad_norm": 0.1606261432170868, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0253, + "step": 46770 + }, + { + "epoch": 4.251954190147246, + "grad_norm": 0.16401013731956482, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0281, + "step": 46780 + }, + { + "epoch": 4.252863115797128, + "grad_norm": 0.12766647338867188, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0254, + "step": 46790 + }, + { + "epoch": 4.25377204144701, + "grad_norm": 0.14802826941013336, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0261, + "step": 46800 + }, + { + "epoch": 4.254680967096892, + "grad_norm": 0.13182318210601807, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0258, + "step": 46810 + }, + { + "epoch": 4.255589892746773, + "grad_norm": 0.12417402863502502, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0253, + "step": 46820 + }, + { + "epoch": 4.256498818396655, + "grad_norm": 0.12458524852991104, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0234, + "step": 46830 + }, + { + "epoch": 4.257407744046537, + "grad_norm": 0.13293060660362244, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0242, + "step": 46840 + }, + { + "epoch": 4.258316669696419, + "grad_norm": 0.1382237821817398, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0255, + "step": 46850 + }, + { + "epoch": 4.259225595346301, + "grad_norm": 0.12429898232221603, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0227, + "step": 46860 + }, + { + "epoch": 4.260134520996182, + "grad_norm": 0.1351989358663559, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0239, + "step": 46870 + }, + { + "epoch": 4.261043446646064, + "grad_norm": 0.12582556903362274, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.027, + "step": 46880 + }, + { + "epoch": 4.261952372295946, + "grad_norm": 0.11865396052598953, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0219, + "step": 46890 + }, + { + "epoch": 4.262861297945828, + "grad_norm": 0.12114950269460678, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0259, + "step": 46900 + }, + { + "epoch": 4.26377022359571, + "grad_norm": 0.15644614398479462, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0264, + "step": 46910 + }, + { + "epoch": 4.264679149245592, + "grad_norm": 0.1303543597459793, + "learning_rate": 2.143191093594534e-05, + "loss": 0.029, + "step": 46920 + }, + { + "epoch": 4.265588074895474, + "grad_norm": 0.1599053144454956, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0258, + "step": 46930 + }, + { + "epoch": 4.266497000545355, + "grad_norm": 0.13948315382003784, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0272, + "step": 46940 + }, + { + "epoch": 4.267405926195237, + "grad_norm": 0.1493384838104248, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0284, + "step": 46950 + }, + { + "epoch": 4.268314851845119, + "grad_norm": 0.15027913451194763, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0247, + "step": 46960 + }, + { + "epoch": 4.269223777495001, + "grad_norm": 0.13114354014396667, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0263, + "step": 46970 + }, + { + "epoch": 4.270132703144883, + "grad_norm": 0.1524362862110138, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0252, + "step": 46980 + }, + { + "epoch": 4.271041628794764, + "grad_norm": 0.11537697911262512, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0282, + "step": 46990 + }, + { + "epoch": 4.271950554444646, + "grad_norm": 0.12456978112459183, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0265, + "step": 47000 + }, + { + "epoch": 4.272859480094528, + "grad_norm": 0.12450860440731049, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0261, + "step": 47010 + }, + { + "epoch": 4.27376840574441, + "grad_norm": 0.1886328011751175, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0268, + "step": 47020 + }, + { + "epoch": 4.274677331394292, + "grad_norm": 0.14849378168582916, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.026, + "step": 47030 + }, + { + "epoch": 4.275586257044174, + "grad_norm": 0.11552634090185165, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0224, + "step": 47040 + }, + { + "epoch": 4.276495182694056, + "grad_norm": 0.20063725113868713, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0234, + "step": 47050 + }, + { + "epoch": 4.277404108343937, + "grad_norm": 0.19120638072490692, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0256, + "step": 47060 + }, + { + "epoch": 4.278313033993819, + "grad_norm": 0.10534434020519257, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0253, + "step": 47070 + }, + { + "epoch": 4.279221959643701, + "grad_norm": 0.1486610323190689, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0263, + "step": 47080 + }, + { + "epoch": 4.280130885293583, + "grad_norm": 0.12368809431791306, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0266, + "step": 47090 + }, + { + "epoch": 4.281039810943465, + "grad_norm": 0.14548777043819427, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.024, + "step": 47100 + }, + { + "epoch": 4.281948736593347, + "grad_norm": 0.13676616549491882, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0249, + "step": 47110 + }, + { + "epoch": 4.282857662243228, + "grad_norm": 0.12044837325811386, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0273, + "step": 47120 + }, + { + "epoch": 4.28376658789311, + "grad_norm": 0.1707943230867386, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0245, + "step": 47130 + }, + { + "epoch": 4.284675513542992, + "grad_norm": 0.13000284135341644, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0251, + "step": 47140 + }, + { + "epoch": 4.285584439192874, + "grad_norm": 0.11191849410533905, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0236, + "step": 47150 + }, + { + "epoch": 4.286493364842756, + "grad_norm": 0.13084420561790466, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0255, + "step": 47160 + }, + { + "epoch": 4.287402290492638, + "grad_norm": 0.13144071400165558, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0276, + "step": 47170 + }, + { + "epoch": 4.288311216142519, + "grad_norm": 0.12190592288970947, + "learning_rate": 2.133465302295309e-05, + "loss": 0.026, + "step": 47180 + }, + { + "epoch": 4.289220141792401, + "grad_norm": 0.11433727294206619, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0242, + "step": 47190 + }, + { + "epoch": 4.290129067442283, + "grad_norm": 0.12004344910383224, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0265, + "step": 47200 + }, + { + "epoch": 4.291037993092165, + "grad_norm": 0.11114230751991272, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0269, + "step": 47210 + }, + { + "epoch": 4.291946918742047, + "grad_norm": 0.1363285928964615, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0245, + "step": 47220 + }, + { + "epoch": 4.292855844391929, + "grad_norm": 0.1257597953081131, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0277, + "step": 47230 + }, + { + "epoch": 4.29376477004181, + "grad_norm": 0.13090316951274872, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0257, + "step": 47240 + }, + { + "epoch": 4.294673695691692, + "grad_norm": 0.13761575520038605, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0252, + "step": 47250 + }, + { + "epoch": 4.295582621341574, + "grad_norm": 0.20934122800827026, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0234, + "step": 47260 + }, + { + "epoch": 4.296491546991456, + "grad_norm": 0.1403317153453827, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0288, + "step": 47270 + }, + { + "epoch": 4.297400472641338, + "grad_norm": 0.162592813372612, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0241, + "step": 47280 + }, + { + "epoch": 4.29830939829122, + "grad_norm": 0.1521233320236206, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0279, + "step": 47290 + }, + { + "epoch": 4.299218323941101, + "grad_norm": 0.13204772770404816, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0248, + "step": 47300 + }, + { + "epoch": 4.300127249590983, + "grad_norm": 0.17916327714920044, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0238, + "step": 47310 + }, + { + "epoch": 4.301036175240865, + "grad_norm": 0.14137819409370422, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0246, + "step": 47320 + }, + { + "epoch": 4.301945100890747, + "grad_norm": 0.12448273599147797, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0264, + "step": 47330 + }, + { + "epoch": 4.302854026540629, + "grad_norm": 0.12149522453546524, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0257, + "step": 47340 + }, + { + "epoch": 4.303762952190511, + "grad_norm": 0.1581735461950302, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0242, + "step": 47350 + }, + { + "epoch": 4.304671877840392, + "grad_norm": 0.14179064333438873, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0244, + "step": 47360 + }, + { + "epoch": 4.305580803490274, + "grad_norm": 0.13870321214199066, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0284, + "step": 47370 + }, + { + "epoch": 4.306489729140156, + "grad_norm": 0.12070396542549133, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.027, + "step": 47380 + }, + { + "epoch": 4.307398654790038, + "grad_norm": 0.12985719740390778, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0261, + "step": 47390 + }, + { + "epoch": 4.30830758043992, + "grad_norm": 0.13090045750141144, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0259, + "step": 47400 + }, + { + "epoch": 4.309216506089802, + "grad_norm": 0.11991458386182785, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0242, + "step": 47410 + }, + { + "epoch": 4.310125431739683, + "grad_norm": 0.12301605939865112, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0239, + "step": 47420 + }, + { + "epoch": 4.311034357389565, + "grad_norm": 0.14624829590320587, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0264, + "step": 47430 + }, + { + "epoch": 4.311943283039447, + "grad_norm": 0.14215262234210968, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0264, + "step": 47440 + }, + { + "epoch": 4.312852208689329, + "grad_norm": 0.15972506999969482, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0245, + "step": 47450 + }, + { + "epoch": 4.313761134339211, + "grad_norm": 0.1288360208272934, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0248, + "step": 47460 + }, + { + "epoch": 4.314670059989093, + "grad_norm": 0.14486686885356903, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0257, + "step": 47470 + }, + { + "epoch": 4.315578985638974, + "grad_norm": 0.14535577595233917, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0248, + "step": 47480 + }, + { + "epoch": 4.316487911288856, + "grad_norm": 0.19745296239852905, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0283, + "step": 47490 + }, + { + "epoch": 4.317396836938738, + "grad_norm": 0.12632600963115692, + "learning_rate": 2.121450506909385e-05, + "loss": 0.026, + "step": 47500 + }, + { + "epoch": 4.31830576258862, + "grad_norm": 0.11418093740940094, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0255, + "step": 47510 + }, + { + "epoch": 4.319214688238502, + "grad_norm": 0.1334766149520874, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0261, + "step": 47520 + }, + { + "epoch": 4.320123613888384, + "grad_norm": 0.17258794605731964, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0289, + "step": 47530 + }, + { + "epoch": 4.321032539538265, + "grad_norm": 0.11742721498012543, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0233, + "step": 47540 + }, + { + "epoch": 4.321941465188147, + "grad_norm": 0.1539246290922165, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0284, + "step": 47550 + }, + { + "epoch": 4.322850390838029, + "grad_norm": 0.1305869072675705, + "learning_rate": 2.119192324338028e-05, + "loss": 0.025, + "step": 47560 + }, + { + "epoch": 4.323759316487911, + "grad_norm": 0.16699592769145966, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0256, + "step": 47570 + }, + { + "epoch": 4.324668242137793, + "grad_norm": 0.13519717752933502, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0272, + "step": 47580 + }, + { + "epoch": 4.325577167787675, + "grad_norm": 0.14174094796180725, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0258, + "step": 47590 + }, + { + "epoch": 4.326486093437556, + "grad_norm": 0.12980663776397705, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0247, + "step": 47600 + }, + { + "epoch": 4.327395019087438, + "grad_norm": 0.15883761644363403, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0265, + "step": 47610 + }, + { + "epoch": 4.32830394473732, + "grad_norm": 0.1779603809118271, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0262, + "step": 47620 + }, + { + "epoch": 4.329212870387202, + "grad_norm": 0.15498323738574982, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0305, + "step": 47630 + }, + { + "epoch": 4.330121796037084, + "grad_norm": 0.14595022797584534, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0271, + "step": 47640 + }, + { + "epoch": 4.331030721686966, + "grad_norm": 0.13782988488674164, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0233, + "step": 47650 + }, + { + "epoch": 4.331939647336847, + "grad_norm": 0.16581402719020844, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0251, + "step": 47660 + }, + { + "epoch": 4.332848572986729, + "grad_norm": 0.1204456090927124, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0233, + "step": 47670 + }, + { + "epoch": 4.333757498636611, + "grad_norm": 0.11702153086662292, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0245, + "step": 47680 + }, + { + "epoch": 4.334666424286493, + "grad_norm": 0.11227535456418991, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0248, + "step": 47690 + }, + { + "epoch": 4.335575349936375, + "grad_norm": 0.13528965413570404, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0249, + "step": 47700 + }, + { + "epoch": 4.336484275586257, + "grad_norm": 0.15301711857318878, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0252, + "step": 47710 + }, + { + "epoch": 4.3373932012361385, + "grad_norm": 0.1522151380777359, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0258, + "step": 47720 + }, + { + "epoch": 4.3383021268860205, + "grad_norm": 0.1684333235025406, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0268, + "step": 47730 + }, + { + "epoch": 4.3392110525359024, + "grad_norm": 0.14310622215270996, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.025, + "step": 47740 + }, + { + "epoch": 4.340119978185784, + "grad_norm": 0.13315895199775696, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.024, + "step": 47750 + }, + { + "epoch": 4.341028903835666, + "grad_norm": 0.18223845958709717, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0243, + "step": 47760 + }, + { + "epoch": 4.341937829485548, + "grad_norm": 0.21289944648742676, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0279, + "step": 47770 + }, + { + "epoch": 4.3428467551354295, + "grad_norm": 0.1375504583120346, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0273, + "step": 47780 + }, + { + "epoch": 4.3437556807853115, + "grad_norm": 0.15064597129821777, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0265, + "step": 47790 + }, + { + "epoch": 4.3446646064351935, + "grad_norm": 0.1446169763803482, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0226, + "step": 47800 + }, + { + "epoch": 4.3455735320850755, + "grad_norm": 0.16055257618427277, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0263, + "step": 47810 + }, + { + "epoch": 4.3464824577349574, + "grad_norm": 0.19352476298809052, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0245, + "step": 47820 + }, + { + "epoch": 4.347391383384839, + "grad_norm": 0.135198712348938, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.023, + "step": 47830 + }, + { + "epoch": 4.348300309034721, + "grad_norm": 0.12767736613750458, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0262, + "step": 47840 + }, + { + "epoch": 4.3492092346846025, + "grad_norm": 0.129098579287529, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0255, + "step": 47850 + }, + { + "epoch": 4.3501181603344845, + "grad_norm": 0.14212702214717865, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0256, + "step": 47860 + }, + { + "epoch": 4.3510270859843665, + "grad_norm": 0.15505081415176392, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0259, + "step": 47870 + }, + { + "epoch": 4.3519360116342485, + "grad_norm": 0.12155203521251678, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0254, + "step": 47880 + }, + { + "epoch": 4.3528449372841305, + "grad_norm": 0.1160605326294899, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.026, + "step": 47890 + }, + { + "epoch": 4.353753862934012, + "grad_norm": 0.1059550866484642, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0263, + "step": 47900 + }, + { + "epoch": 4.3546627885838936, + "grad_norm": 0.15096868574619293, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0255, + "step": 47910 + }, + { + "epoch": 4.3555717142337755, + "grad_norm": 0.15770401060581207, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.026, + "step": 47920 + }, + { + "epoch": 4.3564806398836575, + "grad_norm": 0.14588642120361328, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0252, + "step": 47930 + }, + { + "epoch": 4.3573895655335395, + "grad_norm": 0.1243036612868309, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0277, + "step": 47940 + }, + { + "epoch": 4.3582984911834215, + "grad_norm": 0.14397716522216797, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0274, + "step": 47950 + }, + { + "epoch": 4.3592074168333035, + "grad_norm": 0.14989137649536133, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0262, + "step": 47960 + }, + { + "epoch": 4.360116342483185, + "grad_norm": 0.12487185746431351, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0247, + "step": 47970 + }, + { + "epoch": 4.361025268133067, + "grad_norm": 0.1699877381324768, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0245, + "step": 47980 + }, + { + "epoch": 4.3619341937829486, + "grad_norm": 0.1300087422132492, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0245, + "step": 47990 + }, + { + "epoch": 4.3628431194328305, + "grad_norm": 0.13175107538700104, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0249, + "step": 48000 + }, + { + "epoch": 4.3637520450827125, + "grad_norm": 0.1279250979423523, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0239, + "step": 48010 + }, + { + "epoch": 4.364660970732594, + "grad_norm": 0.1471509337425232, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0251, + "step": 48020 + }, + { + "epoch": 4.365569896382476, + "grad_norm": 0.10717494785785675, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0231, + "step": 48030 + }, + { + "epoch": 4.366478822032358, + "grad_norm": 0.12701493501663208, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.025, + "step": 48040 + }, + { + "epoch": 4.36738774768224, + "grad_norm": 0.11959902197122574, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0226, + "step": 48050 + }, + { + "epoch": 4.368296673332122, + "grad_norm": 0.14782333374023438, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0271, + "step": 48060 + }, + { + "epoch": 4.369205598982004, + "grad_norm": 0.15100917220115662, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0239, + "step": 48070 + }, + { + "epoch": 4.3701145246318855, + "grad_norm": 0.13313692808151245, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0251, + "step": 48080 + }, + { + "epoch": 4.371023450281767, + "grad_norm": 0.1472422033548355, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0227, + "step": 48090 + }, + { + "epoch": 4.371932375931649, + "grad_norm": 0.13886374235153198, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0249, + "step": 48100 + }, + { + "epoch": 4.372841301581531, + "grad_norm": 0.12911468744277954, + "learning_rate": 2.098414220174082e-05, + "loss": 0.0276, + "step": 48110 + }, + { + "epoch": 4.373750227231413, + "grad_norm": 0.14271533489227295, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0244, + "step": 48120 + }, + { + "epoch": 4.374659152881295, + "grad_norm": 0.12723658978939056, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0243, + "step": 48130 + }, + { + "epoch": 4.375568078531176, + "grad_norm": 0.1261950135231018, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0257, + "step": 48140 + }, + { + "epoch": 4.376477004181058, + "grad_norm": 0.13982336223125458, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0278, + "step": 48150 + }, + { + "epoch": 4.37738592983094, + "grad_norm": 0.09517780691385269, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0228, + "step": 48160 + }, + { + "epoch": 4.378294855480822, + "grad_norm": 0.10842353850603104, + "learning_rate": 2.09613911545763e-05, + "loss": 0.026, + "step": 48170 + }, + { + "epoch": 4.379203781130704, + "grad_norm": 0.1190728172659874, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0247, + "step": 48180 + }, + { + "epoch": 4.380112706780586, + "grad_norm": 0.11121375858783722, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0247, + "step": 48190 + }, + { + "epoch": 4.381021632430468, + "grad_norm": 0.1559593677520752, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0237, + "step": 48200 + }, + { + "epoch": 4.381930558080349, + "grad_norm": 0.13944461941719055, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0239, + "step": 48210 + }, + { + "epoch": 4.382839483730231, + "grad_norm": 0.1339556872844696, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0253, + "step": 48220 + }, + { + "epoch": 4.383748409380113, + "grad_norm": 0.13477963209152222, + "learning_rate": 2.0938623809196e-05, + "loss": 0.026, + "step": 48230 + }, + { + "epoch": 4.384657335029995, + "grad_norm": 0.15790638327598572, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0255, + "step": 48240 + }, + { + "epoch": 4.385566260679877, + "grad_norm": 0.15596048533916473, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.025, + "step": 48250 + }, + { + "epoch": 4.386475186329759, + "grad_norm": 0.14024533331394196, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0248, + "step": 48260 + }, + { + "epoch": 4.38738411197964, + "grad_norm": 0.1651417464017868, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.024, + "step": 48270 + }, + { + "epoch": 4.388293037629522, + "grad_norm": 0.13296650350093842, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0234, + "step": 48280 + }, + { + "epoch": 4.389201963279404, + "grad_norm": 0.1447208970785141, + "learning_rate": 2.091584022784496e-05, + "loss": 0.027, + "step": 48290 + }, + { + "epoch": 4.390110888929286, + "grad_norm": 0.12069131433963776, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0236, + "step": 48300 + }, + { + "epoch": 4.391019814579168, + "grad_norm": 0.1384306699037552, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0272, + "step": 48310 + }, + { + "epoch": 4.39192874022905, + "grad_norm": 0.1440284699201584, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0268, + "step": 48320 + }, + { + "epoch": 4.392837665878931, + "grad_norm": 0.13577872514724731, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0235, + "step": 48330 + }, + { + "epoch": 4.393746591528813, + "grad_norm": 0.11138196289539337, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0252, + "step": 48340 + }, + { + "epoch": 4.394655517178695, + "grad_norm": 0.1556771993637085, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0252, + "step": 48350 + }, + { + "epoch": 4.395564442828577, + "grad_norm": 0.16206397116184235, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0242, + "step": 48360 + }, + { + "epoch": 4.396473368478459, + "grad_norm": 0.13319604098796844, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0261, + "step": 48370 + }, + { + "epoch": 4.397382294128341, + "grad_norm": 0.13370326161384583, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0241, + "step": 48380 + }, + { + "epoch": 4.398291219778222, + "grad_norm": 0.14299175143241882, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0238, + "step": 48390 + }, + { + "epoch": 4.399200145428104, + "grad_norm": 0.13358712196350098, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0222, + "step": 48400 + }, + { + "epoch": 4.400109071077986, + "grad_norm": 0.1601892113685608, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0277, + "step": 48410 + }, + { + "epoch": 4.401017996727868, + "grad_norm": 0.11332689225673676, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0225, + "step": 48420 + }, + { + "epoch": 4.40192692237775, + "grad_norm": 0.15544554591178894, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0263, + "step": 48430 + }, + { + "epoch": 4.402835848027632, + "grad_norm": 0.10244132578372955, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0257, + "step": 48440 + }, + { + "epoch": 4.403744773677513, + "grad_norm": 0.12560975551605225, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0301, + "step": 48450 + }, + { + "epoch": 4.404653699327395, + "grad_norm": 0.13590273261070251, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.022, + "step": 48460 + }, + { + "epoch": 4.405562624977277, + "grad_norm": 0.13504673540592194, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.024, + "step": 48470 + }, + { + "epoch": 4.406471550627159, + "grad_norm": 0.13456374406814575, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0259, + "step": 48480 + }, + { + "epoch": 4.407380476277041, + "grad_norm": 0.15908464789390564, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0237, + "step": 48490 + }, + { + "epoch": 4.408289401926923, + "grad_norm": 0.13857096433639526, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0258, + "step": 48500 + }, + { + "epoch": 4.409198327576804, + "grad_norm": 0.14483606815338135, + "learning_rate": 2.083216252891219e-05, + "loss": 0.025, + "step": 48510 + }, + { + "epoch": 4.410107253226686, + "grad_norm": 0.15627074241638184, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0252, + "step": 48520 + }, + { + "epoch": 4.411016178876568, + "grad_norm": 0.13894854485988617, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.025, + "step": 48530 + }, + { + "epoch": 4.41192510452645, + "grad_norm": 0.12062759697437286, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0249, + "step": 48540 + }, + { + "epoch": 4.412834030176332, + "grad_norm": 0.11821646243333817, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0277, + "step": 48550 + }, + { + "epoch": 4.413742955826214, + "grad_norm": 0.12909336388111115, + "learning_rate": 2.08131148627986e-05, + "loss": 0.029, + "step": 48560 + }, + { + "epoch": 4.414651881476095, + "grad_norm": 0.14585328102111816, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0242, + "step": 48570 + }, + { + "epoch": 4.415560807125977, + "grad_norm": 0.1135367751121521, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0269, + "step": 48580 + }, + { + "epoch": 4.416469732775859, + "grad_norm": 0.13140273094177246, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0249, + "step": 48590 + }, + { + "epoch": 4.417378658425741, + "grad_norm": 0.1548326462507248, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0287, + "step": 48600 + }, + { + "epoch": 4.418287584075623, + "grad_norm": 0.14713945984840393, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0255, + "step": 48610 + }, + { + "epoch": 4.419196509725505, + "grad_norm": 0.15273189544677734, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0281, + "step": 48620 + }, + { + "epoch": 4.420105435375386, + "grad_norm": 0.12385056912899017, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0252, + "step": 48630 + }, + { + "epoch": 4.421014361025268, + "grad_norm": 0.15977442264556885, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0251, + "step": 48640 + }, + { + "epoch": 4.42192328667515, + "grad_norm": 0.12563160061836243, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0282, + "step": 48650 + }, + { + "epoch": 4.422832212325032, + "grad_norm": 0.12273812294006348, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.025, + "step": 48660 + }, + { + "epoch": 4.423741137974914, + "grad_norm": 0.11545190215110779, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0257, + "step": 48670 + }, + { + "epoch": 4.424650063624796, + "grad_norm": 0.12153569608926773, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0239, + "step": 48680 + }, + { + "epoch": 4.425558989274677, + "grad_norm": 0.13067267835140228, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0249, + "step": 48690 + }, + { + "epoch": 4.426467914924559, + "grad_norm": 0.13678297400474548, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0264, + "step": 48700 + }, + { + "epoch": 4.427376840574441, + "grad_norm": 0.12929639220237732, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0256, + "step": 48710 + }, + { + "epoch": 4.428285766224323, + "grad_norm": 0.13609319925308228, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0246, + "step": 48720 + }, + { + "epoch": 4.429194691874205, + "grad_norm": 0.16059210896492004, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0268, + "step": 48730 + }, + { + "epoch": 4.430103617524087, + "grad_norm": 0.13401958346366882, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0229, + "step": 48740 + }, + { + "epoch": 4.431012543173968, + "grad_norm": 0.1609012931585312, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0241, + "step": 48750 + }, + { + "epoch": 4.43192146882385, + "grad_norm": 0.13369721174240112, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0257, + "step": 48760 + }, + { + "epoch": 4.432830394473732, + "grad_norm": 0.1323770135641098, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0246, + "step": 48770 + }, + { + "epoch": 4.433739320123614, + "grad_norm": 0.13121189177036285, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0253, + "step": 48780 + }, + { + "epoch": 4.434648245773496, + "grad_norm": 0.1839457005262375, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0257, + "step": 48790 + }, + { + "epoch": 4.435557171423378, + "grad_norm": 0.12470711022615433, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0275, + "step": 48800 + }, + { + "epoch": 4.436466097073259, + "grad_norm": 0.16189637780189514, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.026, + "step": 48810 + }, + { + "epoch": 4.437375022723141, + "grad_norm": 0.11395128071308136, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0239, + "step": 48820 + }, + { + "epoch": 4.438283948373023, + "grad_norm": 0.1333724707365036, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0257, + "step": 48830 + }, + { + "epoch": 4.439192874022905, + "grad_norm": 0.20669488608837128, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.027, + "step": 48840 + }, + { + "epoch": 4.440101799672787, + "grad_norm": 0.14715269207954407, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0236, + "step": 48850 + }, + { + "epoch": 4.441010725322669, + "grad_norm": 0.12933285534381866, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0253, + "step": 48860 + }, + { + "epoch": 4.44191965097255, + "grad_norm": 0.12035852670669556, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0223, + "step": 48870 + }, + { + "epoch": 4.442828576622432, + "grad_norm": 0.18521460890769958, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0283, + "step": 48880 + }, + { + "epoch": 4.443737502272314, + "grad_norm": 0.10275856405496597, + "learning_rate": 2.068712516137462e-05, + "loss": 0.025, + "step": 48890 + }, + { + "epoch": 4.444646427922196, + "grad_norm": 0.14503492414951324, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0275, + "step": 48900 + }, + { + "epoch": 4.445555353572078, + "grad_norm": 0.15704649686813354, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0228, + "step": 48910 + }, + { + "epoch": 4.44646427922196, + "grad_norm": 0.1526537388563156, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0248, + "step": 48920 + }, + { + "epoch": 4.447373204871841, + "grad_norm": 0.14608994126319885, + "learning_rate": 2.067182151424423e-05, + "loss": 0.027, + "step": 48930 + }, + { + "epoch": 4.448282130521723, + "grad_norm": 0.14266537129878998, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0251, + "step": 48940 + }, + { + "epoch": 4.449191056171605, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0244, + "step": 48950 + }, + { + "epoch": 4.450099981821487, + "grad_norm": 0.16274724900722504, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0273, + "step": 48960 + }, + { + "epoch": 4.451008907471369, + "grad_norm": 0.16608506441116333, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0242, + "step": 48970 + }, + { + "epoch": 4.451917833121251, + "grad_norm": 0.08526156842708588, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0265, + "step": 48980 + }, + { + "epoch": 4.452826758771133, + "grad_norm": 0.11740469187498093, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0241, + "step": 48990 + }, + { + "epoch": 4.453735684421014, + "grad_norm": 0.284183531999588, + "learning_rate": 2.064502355956822e-05, + "loss": 0.024, + "step": 49000 + }, + { + "epoch": 4.454644610070896, + "grad_norm": 0.0929589569568634, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0238, + "step": 49010 + }, + { + "epoch": 4.455553535720778, + "grad_norm": 0.12262474000453949, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0313, + "step": 49020 + }, + { + "epoch": 4.45646246137066, + "grad_norm": 0.14438839256763458, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0238, + "step": 49030 + }, + { + "epoch": 4.457371387020542, + "grad_norm": 0.13508501648902893, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0229, + "step": 49040 + }, + { + "epoch": 4.458280312670423, + "grad_norm": 0.15355284512043, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0254, + "step": 49050 + }, + { + "epoch": 4.459189238320305, + "grad_norm": 0.11139452457427979, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0251, + "step": 49060 + }, + { + "epoch": 4.460098163970187, + "grad_norm": 0.12605150043964386, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0243, + "step": 49070 + }, + { + "epoch": 4.461007089620069, + "grad_norm": 0.11716443300247192, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0251, + "step": 49080 + }, + { + "epoch": 4.461916015269951, + "grad_norm": 0.20713534951210022, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0249, + "step": 49090 + }, + { + "epoch": 4.462824940919833, + "grad_norm": 0.16831539571285248, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0265, + "step": 49100 + }, + { + "epoch": 4.463733866569715, + "grad_norm": 0.1373203992843628, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0261, + "step": 49110 + }, + { + "epoch": 4.464642792219596, + "grad_norm": 0.1744706928730011, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0268, + "step": 49120 + }, + { + "epoch": 4.465551717869478, + "grad_norm": 0.15564732253551483, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0235, + "step": 49130 + }, + { + "epoch": 4.46646064351936, + "grad_norm": 0.1268274486064911, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0247, + "step": 49140 + }, + { + "epoch": 4.467369569169242, + "grad_norm": 0.14417417347431183, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0263, + "step": 49150 + }, + { + "epoch": 4.468278494819124, + "grad_norm": 0.12481914460659027, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0276, + "step": 49160 + }, + { + "epoch": 4.469187420469005, + "grad_norm": 0.14536622166633606, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0254, + "step": 49170 + }, + { + "epoch": 4.470096346118887, + "grad_norm": 0.12197145074605942, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0256, + "step": 49180 + }, + { + "epoch": 4.471005271768769, + "grad_norm": 0.12699414789676666, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0245, + "step": 49190 + }, + { + "epoch": 4.471914197418651, + "grad_norm": 0.12703067064285278, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0277, + "step": 49200 + }, + { + "epoch": 4.472823123068533, + "grad_norm": 0.14432598650455475, + "learning_rate": 2.056450405638196e-05, + "loss": 0.027, + "step": 49210 + }, + { + "epoch": 4.473732048718415, + "grad_norm": 0.15335708856582642, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0244, + "step": 49220 + }, + { + "epoch": 4.474640974368297, + "grad_norm": 0.14211226999759674, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0262, + "step": 49230 + }, + { + "epoch": 4.475549900018178, + "grad_norm": 0.13359731435775757, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0248, + "step": 49240 + }, + { + "epoch": 4.47645882566806, + "grad_norm": 0.13114818930625916, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0234, + "step": 49250 + }, + { + "epoch": 4.477367751317942, + "grad_norm": 0.10966645926237106, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0248, + "step": 49260 + }, + { + "epoch": 4.478276676967824, + "grad_norm": 0.12071753293275833, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0287, + "step": 49270 + }, + { + "epoch": 4.479185602617706, + "grad_norm": 0.11678314954042435, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0246, + "step": 49280 + }, + { + "epoch": 4.480094528267587, + "grad_norm": 0.1576126217842102, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0269, + "step": 49290 + }, + { + "epoch": 4.481003453917469, + "grad_norm": 0.18966343998908997, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0258, + "step": 49300 + }, + { + "epoch": 4.481912379567351, + "grad_norm": 0.15244199335575104, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0226, + "step": 49310 + }, + { + "epoch": 4.482821305217233, + "grad_norm": 0.14164960384368896, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0245, + "step": 49320 + }, + { + "epoch": 4.483730230867115, + "grad_norm": 0.15170182287693024, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0275, + "step": 49330 + }, + { + "epoch": 4.484639156516997, + "grad_norm": 0.1172860711812973, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0261, + "step": 49340 + }, + { + "epoch": 4.485548082166879, + "grad_norm": 0.16536729037761688, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0247, + "step": 49350 + }, + { + "epoch": 4.48645700781676, + "grad_norm": 0.13928677141666412, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0251, + "step": 49360 + }, + { + "epoch": 4.487365933466642, + "grad_norm": 0.10256270319223404, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0236, + "step": 49370 + }, + { + "epoch": 4.488274859116524, + "grad_norm": 0.10876251012086868, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0243, + "step": 49380 + }, + { + "epoch": 4.489183784766406, + "grad_norm": 0.2004053294658661, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0253, + "step": 49390 + }, + { + "epoch": 4.490092710416288, + "grad_norm": 0.18924495577812195, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0252, + "step": 49400 + }, + { + "epoch": 4.49100163606617, + "grad_norm": 0.16157762706279755, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0273, + "step": 49410 + }, + { + "epoch": 4.491910561716051, + "grad_norm": 0.1351243257522583, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0246, + "step": 49420 + }, + { + "epoch": 4.492819487365933, + "grad_norm": 0.17627368867397308, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0265, + "step": 49430 + }, + { + "epoch": 4.493728413015815, + "grad_norm": 0.14508841931819916, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0249, + "step": 49440 + }, + { + "epoch": 4.494637338665697, + "grad_norm": 0.1398521363735199, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0248, + "step": 49450 + }, + { + "epoch": 4.495546264315579, + "grad_norm": 0.1448148936033249, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0226, + "step": 49460 + }, + { + "epoch": 4.496455189965461, + "grad_norm": 0.14566168189048767, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0225, + "step": 49470 + }, + { + "epoch": 4.497364115615342, + "grad_norm": 0.15453700721263885, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0262, + "step": 49480 + }, + { + "epoch": 4.498273041265224, + "grad_norm": 0.14914663136005402, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0271, + "step": 49490 + }, + { + "epoch": 4.499181966915106, + "grad_norm": 0.11807655543088913, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0234, + "step": 49500 + }, + { + "epoch": 4.500090892564988, + "grad_norm": 0.1394844502210617, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0228, + "step": 49510 + }, + { + "epoch": 4.50099981821487, + "grad_norm": 0.12289104610681534, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0246, + "step": 49520 + }, + { + "epoch": 4.501908743864751, + "grad_norm": 0.12453174591064453, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0242, + "step": 49530 + }, + { + "epoch": 4.502817669514633, + "grad_norm": 0.11199140548706055, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0258, + "step": 49540 + }, + { + "epoch": 4.503726595164515, + "grad_norm": 0.1366352140903473, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0257, + "step": 49550 + }, + { + "epoch": 4.504635520814397, + "grad_norm": 0.124833844602108, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0245, + "step": 49560 + }, + { + "epoch": 4.505544446464279, + "grad_norm": 0.12729164958000183, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0259, + "step": 49570 + }, + { + "epoch": 4.506453372114161, + "grad_norm": 0.10522066801786423, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0253, + "step": 49580 + }, + { + "epoch": 4.507362297764043, + "grad_norm": 0.1478882133960724, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0275, + "step": 49590 + }, + { + "epoch": 4.508271223413924, + "grad_norm": 0.1591508835554123, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0258, + "step": 49600 + }, + { + "epoch": 4.509180149063806, + "grad_norm": 0.15357109904289246, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0254, + "step": 49610 + }, + { + "epoch": 4.510089074713688, + "grad_norm": 0.13343359529972076, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0244, + "step": 49620 + }, + { + "epoch": 4.51099800036357, + "grad_norm": 0.2610541582107544, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0242, + "step": 49630 + }, + { + "epoch": 4.511906926013452, + "grad_norm": 0.13728895783424377, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0263, + "step": 49640 + }, + { + "epoch": 4.512815851663334, + "grad_norm": 0.12231455743312836, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0245, + "step": 49650 + }, + { + "epoch": 4.513724777313215, + "grad_norm": 0.11939503252506256, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0242, + "step": 49660 + }, + { + "epoch": 4.514633702963097, + "grad_norm": 0.12913228571414948, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0246, + "step": 49670 + }, + { + "epoch": 4.515542628612979, + "grad_norm": 0.11394840478897095, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0251, + "step": 49680 + }, + { + "epoch": 4.516451554262861, + "grad_norm": 0.12705554068088531, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0253, + "step": 49690 + }, + { + "epoch": 4.517360479912743, + "grad_norm": 0.1591845452785492, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0258, + "step": 49700 + }, + { + "epoch": 4.518269405562625, + "grad_norm": 0.1333991289138794, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0234, + "step": 49710 + }, + { + "epoch": 4.5191783312125064, + "grad_norm": 0.3527941107749939, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0237, + "step": 49720 + }, + { + "epoch": 4.520087256862388, + "grad_norm": 0.12559181451797485, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0254, + "step": 49730 + }, + { + "epoch": 4.52099618251227, + "grad_norm": 0.14051347970962524, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0261, + "step": 49740 + }, + { + "epoch": 4.521905108162152, + "grad_norm": 0.15906789898872375, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0268, + "step": 49750 + }, + { + "epoch": 4.522814033812034, + "grad_norm": 0.12939953804016113, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0245, + "step": 49760 + }, + { + "epoch": 4.523722959461916, + "grad_norm": 0.11832413822412491, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0223, + "step": 49770 + }, + { + "epoch": 4.5246318851117975, + "grad_norm": 0.1122160330414772, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0262, + "step": 49780 + }, + { + "epoch": 4.5255408107616795, + "grad_norm": 0.13017095625400543, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0233, + "step": 49790 + }, + { + "epoch": 4.5264497364115615, + "grad_norm": 0.1640831083059311, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0261, + "step": 49800 + }, + { + "epoch": 4.527358662061443, + "grad_norm": 0.11941882967948914, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0238, + "step": 49810 + }, + { + "epoch": 4.528267587711325, + "grad_norm": 0.15270482003688812, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0257, + "step": 49820 + }, + { + "epoch": 4.529176513361207, + "grad_norm": 0.16140183806419373, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0257, + "step": 49830 + }, + { + "epoch": 4.5300854390110885, + "grad_norm": 0.12936659157276154, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0262, + "step": 49840 + }, + { + "epoch": 4.5309943646609705, + "grad_norm": 0.16151078045368195, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0254, + "step": 49850 + }, + { + "epoch": 4.5319032903108525, + "grad_norm": 0.14126461744308472, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0282, + "step": 49860 + }, + { + "epoch": 4.5328122159607345, + "grad_norm": 0.16030572354793549, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0256, + "step": 49870 + }, + { + "epoch": 4.5337211416106165, + "grad_norm": 0.12062334269285202, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0243, + "step": 49880 + }, + { + "epoch": 4.534630067260498, + "grad_norm": 0.1483590453863144, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0252, + "step": 49890 + }, + { + "epoch": 4.53553899291038, + "grad_norm": 0.13356344401836395, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0248, + "step": 49900 + }, + { + "epoch": 4.5364479185602615, + "grad_norm": 0.11817321926355362, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0237, + "step": 49910 + }, + { + "epoch": 4.5373568442101435, + "grad_norm": 0.1515001803636551, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0289, + "step": 49920 + }, + { + "epoch": 4.5382657698600255, + "grad_norm": 0.12455679476261139, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0234, + "step": 49930 + }, + { + "epoch": 4.5391746955099075, + "grad_norm": 0.17079903185367584, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0277, + "step": 49940 + }, + { + "epoch": 4.5400836211597895, + "grad_norm": 0.12888818979263306, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0283, + "step": 49950 + }, + { + "epoch": 4.540992546809671, + "grad_norm": 0.11430853605270386, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0221, + "step": 49960 + }, + { + "epoch": 4.541901472459553, + "grad_norm": 0.12094715237617493, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0237, + "step": 49970 + }, + { + "epoch": 4.5428103981094345, + "grad_norm": 0.14185146987438202, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0248, + "step": 49980 + }, + { + "epoch": 4.5437193237593165, + "grad_norm": 0.1614653468132019, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0251, + "step": 49990 + }, + { + "epoch": 4.5446282494091985, + "grad_norm": 0.1368187516927719, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0243, + "step": 50000 + }, + { + "epoch": 4.5455371750590805, + "grad_norm": 0.11148001253604889, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0251, + "step": 50010 + }, + { + "epoch": 4.5464461007089625, + "grad_norm": 0.13500970602035522, + "learning_rate": 2.025220726516947e-05, + "loss": 0.026, + "step": 50020 + }, + { + "epoch": 4.547355026358844, + "grad_norm": 0.1859329491853714, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0229, + "step": 50030 + }, + { + "epoch": 4.548263952008726, + "grad_norm": 0.11787818372249603, + "learning_rate": 2.024446248569512e-05, + "loss": 0.025, + "step": 50040 + }, + { + "epoch": 4.549172877658608, + "grad_norm": 0.11515838652849197, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0247, + "step": 50050 + }, + { + "epoch": 4.5500818033084895, + "grad_norm": 0.12265542894601822, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0267, + "step": 50060 + }, + { + "epoch": 4.5509907289583715, + "grad_norm": 0.14039433002471924, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0277, + "step": 50070 + }, + { + "epoch": 4.551899654608253, + "grad_norm": 0.15345749258995056, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0253, + "step": 50080 + }, + { + "epoch": 4.552808580258135, + "grad_norm": 0.1420241892337799, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0239, + "step": 50090 + }, + { + "epoch": 4.553717505908017, + "grad_norm": 0.15385717153549194, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0257, + "step": 50100 + }, + { + "epoch": 4.554626431557899, + "grad_norm": 0.17918166518211365, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0237, + "step": 50110 + }, + { + "epoch": 4.555535357207781, + "grad_norm": 0.13548889756202698, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0234, + "step": 50120 + }, + { + "epoch": 4.556444282857663, + "grad_norm": 0.1314959079027176, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0255, + "step": 50130 + }, + { + "epoch": 4.5573532085075446, + "grad_norm": 0.14543621242046356, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.026, + "step": 50140 + }, + { + "epoch": 4.558262134157426, + "grad_norm": 0.16185447573661804, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0263, + "step": 50150 + }, + { + "epoch": 4.559171059807308, + "grad_norm": 0.12656423449516296, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.023, + "step": 50160 + }, + { + "epoch": 4.56007998545719, + "grad_norm": 0.1469605267047882, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0268, + "step": 50170 + }, + { + "epoch": 4.560988911107072, + "grad_norm": 0.14286768436431885, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0262, + "step": 50180 + }, + { + "epoch": 4.561897836756954, + "grad_norm": 0.14047357439994812, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0245, + "step": 50190 + }, + { + "epoch": 4.562806762406835, + "grad_norm": 0.12708625197410583, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0286, + "step": 50200 + }, + { + "epoch": 4.563715688056717, + "grad_norm": 0.11900753527879715, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0242, + "step": 50210 + }, + { + "epoch": 4.564624613706599, + "grad_norm": 0.11053812503814697, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0252, + "step": 50220 + }, + { + "epoch": 4.565533539356481, + "grad_norm": 0.14729081094264984, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0267, + "step": 50230 + }, + { + "epoch": 4.566442465006363, + "grad_norm": 0.11269795149564743, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0214, + "step": 50240 + }, + { + "epoch": 4.567351390656245, + "grad_norm": 0.12765882909297943, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.022, + "step": 50250 + }, + { + "epoch": 4.568260316306127, + "grad_norm": 0.1350378841161728, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0252, + "step": 50260 + }, + { + "epoch": 4.569169241956008, + "grad_norm": 0.14029058814048767, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0249, + "step": 50270 + }, + { + "epoch": 4.57007816760589, + "grad_norm": 0.14792858064174652, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0247, + "step": 50280 + }, + { + "epoch": 4.570987093255772, + "grad_norm": 0.13791590929031372, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0244, + "step": 50290 + }, + { + "epoch": 4.571896018905654, + "grad_norm": 0.10856311023235321, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0254, + "step": 50300 + }, + { + "epoch": 4.572804944555536, + "grad_norm": 0.1375037133693695, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0229, + "step": 50310 + }, + { + "epoch": 4.573713870205417, + "grad_norm": 0.11917924135923386, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0242, + "step": 50320 + }, + { + "epoch": 4.574622795855299, + "grad_norm": 0.16110765933990479, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0257, + "step": 50330 + }, + { + "epoch": 4.575531721505181, + "grad_norm": 0.1180584505200386, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0244, + "step": 50340 + }, + { + "epoch": 4.576440647155063, + "grad_norm": 0.11827713251113892, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0227, + "step": 50350 + }, + { + "epoch": 4.577349572804945, + "grad_norm": 0.13947029411792755, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.023, + "step": 50360 + }, + { + "epoch": 4.578258498454827, + "grad_norm": 0.15345968306064606, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0263, + "step": 50370 + }, + { + "epoch": 4.579167424104709, + "grad_norm": 0.11501242965459824, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0255, + "step": 50380 + }, + { + "epoch": 4.58007634975459, + "grad_norm": 0.10795928537845612, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0251, + "step": 50390 + }, + { + "epoch": 4.580985275404472, + "grad_norm": 0.1288185864686966, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0234, + "step": 50400 + }, + { + "epoch": 4.581894201054354, + "grad_norm": 0.1433289498090744, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0263, + "step": 50410 + }, + { + "epoch": 4.582803126704236, + "grad_norm": 0.130781888961792, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0249, + "step": 50420 + }, + { + "epoch": 4.583712052354118, + "grad_norm": 0.13909125328063965, + "learning_rate": 2.00931237622297e-05, + "loss": 0.025, + "step": 50430 + }, + { + "epoch": 4.584620978003999, + "grad_norm": 0.12178529798984528, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0232, + "step": 50440 + }, + { + "epoch": 4.585529903653881, + "grad_norm": 0.1470785140991211, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0248, + "step": 50450 + }, + { + "epoch": 4.586438829303763, + "grad_norm": 0.13811326026916504, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0259, + "step": 50460 + }, + { + "epoch": 4.587347754953645, + "grad_norm": 0.10473068058490753, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0278, + "step": 50470 + }, + { + "epoch": 4.588256680603527, + "grad_norm": 0.15423554182052612, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0257, + "step": 50480 + }, + { + "epoch": 4.589165606253409, + "grad_norm": 0.13285355269908905, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.025, + "step": 50490 + }, + { + "epoch": 4.590074531903291, + "grad_norm": 0.12636490166187286, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0239, + "step": 50500 + }, + { + "epoch": 4.590983457553172, + "grad_norm": 0.29408133029937744, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0255, + "step": 50510 + }, + { + "epoch": 4.591892383203054, + "grad_norm": 0.11724980175495148, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0231, + "step": 50520 + }, + { + "epoch": 4.592801308852936, + "grad_norm": 0.14209330081939697, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0228, + "step": 50530 + }, + { + "epoch": 4.593710234502818, + "grad_norm": 0.11872461438179016, + "learning_rate": 2.005033135718655e-05, + "loss": 0.023, + "step": 50540 + }, + { + "epoch": 4.5946191601527, + "grad_norm": 0.10841648280620575, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0235, + "step": 50550 + }, + { + "epoch": 4.595528085802581, + "grad_norm": 0.13506048917770386, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0273, + "step": 50560 + }, + { + "epoch": 4.596437011452463, + "grad_norm": 0.1431642472743988, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0229, + "step": 50570 + }, + { + "epoch": 4.597345937102345, + "grad_norm": 0.15333303809165955, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0255, + "step": 50580 + }, + { + "epoch": 4.598254862752227, + "grad_norm": 0.12402600049972534, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0247, + "step": 50590 + }, + { + "epoch": 4.599163788402109, + "grad_norm": 0.16422836482524872, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0253, + "step": 50600 + }, + { + "epoch": 4.600072714051991, + "grad_norm": 0.15417316555976868, + "learning_rate": 2.002307563891023e-05, + "loss": 0.025, + "step": 50610 + }, + { + "epoch": 4.600981639701873, + "grad_norm": 0.13789574801921844, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0263, + "step": 50620 + }, + { + "epoch": 4.601890565351754, + "grad_norm": 0.18182629346847534, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0258, + "step": 50630 + }, + { + "epoch": 4.602799491001636, + "grad_norm": 0.20037801563739777, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0257, + "step": 50640 + }, + { + "epoch": 4.603708416651518, + "grad_norm": 0.12526194751262665, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0277, + "step": 50650 + }, + { + "epoch": 4.6046173423014, + "grad_norm": 0.11948402225971222, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0216, + "step": 50660 + }, + { + "epoch": 4.605526267951282, + "grad_norm": 0.14691320061683655, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0237, + "step": 50670 + }, + { + "epoch": 4.606435193601163, + "grad_norm": 0.15108369290828705, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0248, + "step": 50680 + }, + { + "epoch": 4.607344119251045, + "grad_norm": 0.1315719187259674, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0244, + "step": 50690 + }, + { + "epoch": 4.608253044900927, + "grad_norm": 0.11711347103118896, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0253, + "step": 50700 + }, + { + "epoch": 4.609161970550809, + "grad_norm": 0.12633033096790314, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0248, + "step": 50710 + }, + { + "epoch": 4.610070896200691, + "grad_norm": 0.11325907707214355, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0263, + "step": 50720 + }, + { + "epoch": 4.610979821850573, + "grad_norm": 0.13371653854846954, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0228, + "step": 50730 + }, + { + "epoch": 4.611888747500455, + "grad_norm": 0.13473255932331085, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0237, + "step": 50740 + }, + { + "epoch": 4.612797673150336, + "grad_norm": 0.11030949652194977, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0245, + "step": 50750 + }, + { + "epoch": 4.613706598800218, + "grad_norm": 0.11542923003435135, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0237, + "step": 50760 + }, + { + "epoch": 4.6146155244501, + "grad_norm": 0.11353649944067001, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.025, + "step": 50770 + }, + { + "epoch": 4.615524450099982, + "grad_norm": 0.1359304040670395, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0257, + "step": 50780 + }, + { + "epoch": 4.616433375749864, + "grad_norm": 0.16056665778160095, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.0251, + "step": 50790 + }, + { + "epoch": 4.617342301399746, + "grad_norm": 0.16865567862987518, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0272, + "step": 50800 + }, + { + "epoch": 4.618251227049627, + "grad_norm": 0.11811242252588272, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0212, + "step": 50810 + }, + { + "epoch": 4.619160152699509, + "grad_norm": 0.1354888677597046, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0248, + "step": 50820 + }, + { + "epoch": 4.620069078349391, + "grad_norm": 0.10865435004234314, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0234, + "step": 50830 + }, + { + "epoch": 4.620978003999273, + "grad_norm": 0.1506468951702118, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.024, + "step": 50840 + }, + { + "epoch": 4.621886929649155, + "grad_norm": 0.14260752499103546, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0232, + "step": 50850 + }, + { + "epoch": 4.622795855299037, + "grad_norm": 0.15531635284423828, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0252, + "step": 50860 + }, + { + "epoch": 4.623704780948918, + "grad_norm": 0.1424817442893982, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0229, + "step": 50870 + }, + { + "epoch": 4.6246137065988, + "grad_norm": 0.14257057011127472, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0244, + "step": 50880 + }, + { + "epoch": 4.625522632248682, + "grad_norm": 0.13310253620147705, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0248, + "step": 50890 + }, + { + "epoch": 4.626431557898564, + "grad_norm": 0.14342308044433594, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0225, + "step": 50900 + }, + { + "epoch": 4.627340483548446, + "grad_norm": 0.12010304629802704, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0255, + "step": 50910 + }, + { + "epoch": 4.628249409198328, + "grad_norm": 0.11262910068035126, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0253, + "step": 50920 + }, + { + "epoch": 4.629158334848209, + "grad_norm": 0.1453857719898224, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0258, + "step": 50930 + }, + { + "epoch": 4.630067260498091, + "grad_norm": 0.11575829982757568, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0232, + "step": 50940 + }, + { + "epoch": 4.630976186147973, + "grad_norm": 0.12462399899959564, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0232, + "step": 50950 + }, + { + "epoch": 4.631885111797855, + "grad_norm": 0.12621262669563293, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0231, + "step": 50960 + }, + { + "epoch": 4.632794037447737, + "grad_norm": 0.11606328189373016, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0248, + "step": 50970 + }, + { + "epoch": 4.633702963097619, + "grad_norm": 0.11133481562137604, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0244, + "step": 50980 + }, + { + "epoch": 4.6346118887475, + "grad_norm": 0.12698446214199066, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0258, + "step": 50990 + }, + { + "epoch": 4.635520814397382, + "grad_norm": 0.1454915702342987, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0242, + "step": 51000 + }, + { + "epoch": 4.636429740047264, + "grad_norm": 0.10565044730901718, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0216, + "step": 51010 + }, + { + "epoch": 4.637338665697146, + "grad_norm": 0.15763920545578003, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0226, + "step": 51020 + }, + { + "epoch": 4.638247591347028, + "grad_norm": 0.13070033490657806, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0255, + "step": 51030 + }, + { + "epoch": 4.63915651699691, + "grad_norm": 0.19895152747631073, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0253, + "step": 51040 + }, + { + "epoch": 4.640065442646791, + "grad_norm": 0.1322072297334671, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0239, + "step": 51050 + }, + { + "epoch": 4.640974368296673, + "grad_norm": 0.17951379716396332, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0257, + "step": 51060 + }, + { + "epoch": 4.641883293946555, + "grad_norm": 0.11645688116550446, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0241, + "step": 51070 + }, + { + "epoch": 4.642792219596437, + "grad_norm": 0.16651551425457, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0248, + "step": 51080 + }, + { + "epoch": 4.643701145246319, + "grad_norm": 0.23474693298339844, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0234, + "step": 51090 + }, + { + "epoch": 4.644610070896201, + "grad_norm": 0.14789126813411713, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0252, + "step": 51100 + }, + { + "epoch": 4.645518996546082, + "grad_norm": 0.15070612728595734, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0241, + "step": 51110 + }, + { + "epoch": 4.646427922195964, + "grad_norm": 0.13529083132743835, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0226, + "step": 51120 + }, + { + "epoch": 4.647336847845846, + "grad_norm": 0.14237231016159058, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0261, + "step": 51130 + }, + { + "epoch": 4.648245773495728, + "grad_norm": 0.11998802423477173, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0234, + "step": 51140 + }, + { + "epoch": 4.64915469914561, + "grad_norm": 0.12309221178293228, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0244, + "step": 51150 + }, + { + "epoch": 4.650063624795492, + "grad_norm": 0.12630870938301086, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0237, + "step": 51160 + }, + { + "epoch": 4.650972550445374, + "grad_norm": 0.15902093052864075, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.027, + "step": 51170 + }, + { + "epoch": 4.651881476095255, + "grad_norm": 0.1329156458377838, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0248, + "step": 51180 + }, + { + "epoch": 4.652790401745137, + "grad_norm": 0.21017101407051086, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0237, + "step": 51190 + }, + { + "epoch": 4.653699327395019, + "grad_norm": 0.1391133964061737, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0246, + "step": 51200 + }, + { + "epoch": 4.654608253044901, + "grad_norm": 0.11877711862325668, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0258, + "step": 51210 + }, + { + "epoch": 4.655517178694783, + "grad_norm": 0.12359856814146042, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0252, + "step": 51220 + }, + { + "epoch": 4.656426104344664, + "grad_norm": 0.1124296635389328, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0241, + "step": 51230 + }, + { + "epoch": 4.657335029994546, + "grad_norm": 0.15335527062416077, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0237, + "step": 51240 + }, + { + "epoch": 4.658243955644428, + "grad_norm": 0.1555967926979065, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0244, + "step": 51250 + }, + { + "epoch": 4.65915288129431, + "grad_norm": 0.1330394148826599, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0257, + "step": 51260 + }, + { + "epoch": 4.660061806944192, + "grad_norm": 0.13677559792995453, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0224, + "step": 51270 + }, + { + "epoch": 4.660970732594074, + "grad_norm": 0.14281609654426575, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0248, + "step": 51280 + }, + { + "epoch": 4.661879658243956, + "grad_norm": 0.12219922244548798, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0244, + "step": 51290 + }, + { + "epoch": 4.662788583893837, + "grad_norm": 0.15584298968315125, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0263, + "step": 51300 + }, + { + "epoch": 4.663697509543719, + "grad_norm": 0.15429547429084778, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0251, + "step": 51310 + }, + { + "epoch": 4.664606435193601, + "grad_norm": 0.16330213844776154, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.024, + "step": 51320 + }, + { + "epoch": 4.665515360843483, + "grad_norm": 0.1398254930973053, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0237, + "step": 51330 + }, + { + "epoch": 4.666424286493365, + "grad_norm": 0.19237367808818817, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0248, + "step": 51340 + }, + { + "epoch": 4.667333212143246, + "grad_norm": 0.14274436235427856, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0235, + "step": 51350 + }, + { + "epoch": 4.668242137793128, + "grad_norm": 0.1896711140871048, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0243, + "step": 51360 + }, + { + "epoch": 4.66915106344301, + "grad_norm": 0.16650810837745667, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0244, + "step": 51370 + }, + { + "epoch": 4.670059989092892, + "grad_norm": 0.129511296749115, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0269, + "step": 51380 + }, + { + "epoch": 4.670968914742774, + "grad_norm": 0.0974595919251442, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0228, + "step": 51390 + }, + { + "epoch": 4.671877840392656, + "grad_norm": 0.13551273941993713, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0246, + "step": 51400 + }, + { + "epoch": 4.672786766042538, + "grad_norm": 0.17306391894817352, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.025, + "step": 51410 + }, + { + "epoch": 4.673695691692419, + "grad_norm": 0.1415431648492813, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0253, + "step": 51420 + }, + { + "epoch": 4.674604617342301, + "grad_norm": 0.1263875961303711, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0252, + "step": 51430 + }, + { + "epoch": 4.675513542992183, + "grad_norm": 0.12641514837741852, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0227, + "step": 51440 + }, + { + "epoch": 4.676422468642065, + "grad_norm": 0.14105232059955597, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0244, + "step": 51450 + }, + { + "epoch": 4.677331394291947, + "grad_norm": 0.18388181924819946, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0237, + "step": 51460 + }, + { + "epoch": 4.678240319941828, + "grad_norm": 0.13447511196136475, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0223, + "step": 51470 + }, + { + "epoch": 4.67914924559171, + "grad_norm": 0.1420726776123047, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0253, + "step": 51480 + }, + { + "epoch": 4.680058171241592, + "grad_norm": 0.11414962261915207, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0241, + "step": 51490 + }, + { + "epoch": 4.680967096891474, + "grad_norm": 0.1263291835784912, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0249, + "step": 51500 + }, + { + "epoch": 4.681876022541356, + "grad_norm": 0.13190536201000214, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0244, + "step": 51510 + }, + { + "epoch": 4.682784948191238, + "grad_norm": 0.09813898801803589, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0237, + "step": 51520 + }, + { + "epoch": 4.68369387384112, + "grad_norm": 0.11106111109256744, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0233, + "step": 51530 + }, + { + "epoch": 4.684602799491001, + "grad_norm": 0.19118979573249817, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0267, + "step": 51540 + }, + { + "epoch": 4.685511725140883, + "grad_norm": 0.1442815512418747, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0249, + "step": 51550 + }, + { + "epoch": 4.686420650790765, + "grad_norm": 0.11950038373470306, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0255, + "step": 51560 + }, + { + "epoch": 4.687329576440647, + "grad_norm": 0.18408769369125366, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0234, + "step": 51570 + }, + { + "epoch": 4.688238502090529, + "grad_norm": 0.13303713500499725, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0231, + "step": 51580 + }, + { + "epoch": 4.68914742774041, + "grad_norm": 0.18642428517341614, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0244, + "step": 51590 + }, + { + "epoch": 4.690056353390292, + "grad_norm": 0.14259572327136993, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0251, + "step": 51600 + }, + { + "epoch": 4.690965279040174, + "grad_norm": 0.1555519849061966, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0264, + "step": 51610 + }, + { + "epoch": 4.691874204690056, + "grad_norm": 0.13562752306461334, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0264, + "step": 51620 + }, + { + "epoch": 4.692783130339938, + "grad_norm": 0.1310662031173706, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0253, + "step": 51630 + }, + { + "epoch": 4.69369205598982, + "grad_norm": 0.16971954703330994, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0267, + "step": 51640 + }, + { + "epoch": 4.694600981639702, + "grad_norm": 0.17151017487049103, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0254, + "step": 51650 + }, + { + "epoch": 4.695509907289583, + "grad_norm": 0.12441258877515793, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0221, + "step": 51660 + }, + { + "epoch": 4.696418832939465, + "grad_norm": 0.12087757140398026, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0234, + "step": 51670 + }, + { + "epoch": 4.697327758589347, + "grad_norm": 0.1450270116329193, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0231, + "step": 51680 + }, + { + "epoch": 4.698236684239229, + "grad_norm": 0.1436799019575119, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0257, + "step": 51690 + }, + { + "epoch": 4.699145609889111, + "grad_norm": 0.12740552425384521, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0257, + "step": 51700 + }, + { + "epoch": 4.700054535538992, + "grad_norm": 0.11706358939409256, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0249, + "step": 51710 + }, + { + "epoch": 4.700963461188874, + "grad_norm": 0.13540096580982208, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0263, + "step": 51720 + }, + { + "epoch": 4.701872386838756, + "grad_norm": 0.1306244134902954, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0232, + "step": 51730 + }, + { + "epoch": 4.702781312488638, + "grad_norm": 0.1283494383096695, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0274, + "step": 51740 + }, + { + "epoch": 4.70369023813852, + "grad_norm": 0.1400115042924881, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0234, + "step": 51750 + }, + { + "epoch": 4.704599163788402, + "grad_norm": 0.15728367865085602, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.024, + "step": 51760 + }, + { + "epoch": 4.705508089438284, + "grad_norm": 0.4618186056613922, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0252, + "step": 51770 + }, + { + "epoch": 4.7064170150881655, + "grad_norm": 0.11064144968986511, + "learning_rate": 1.95648274960463e-05, + "loss": 0.021, + "step": 51780 + }, + { + "epoch": 4.707325940738047, + "grad_norm": 0.14891542494297028, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0268, + "step": 51790 + }, + { + "epoch": 4.708234866387929, + "grad_norm": 0.11116912215948105, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0205, + "step": 51800 + }, + { + "epoch": 4.709143792037811, + "grad_norm": 0.18109308183193207, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0239, + "step": 51810 + }, + { + "epoch": 4.710052717687693, + "grad_norm": 0.1359344720840454, + "learning_rate": 1.954907428044873e-05, + "loss": 0.024, + "step": 51820 + }, + { + "epoch": 4.7109616433375745, + "grad_norm": 0.1413196325302124, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0254, + "step": 51830 + }, + { + "epoch": 4.7118705689874565, + "grad_norm": 0.1386408805847168, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0253, + "step": 51840 + }, + { + "epoch": 4.7127794946373385, + "grad_norm": 0.14032694697380066, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0262, + "step": 51850 + }, + { + "epoch": 4.7136884202872205, + "grad_norm": 0.1114039346575737, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0244, + "step": 51860 + }, + { + "epoch": 4.7145973459371024, + "grad_norm": 0.13083432614803314, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0262, + "step": 51870 + }, + { + "epoch": 4.715506271586984, + "grad_norm": 0.1173759326338768, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.024, + "step": 51880 + }, + { + "epoch": 4.716415197236866, + "grad_norm": 0.10658050328493118, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0221, + "step": 51890 + }, + { + "epoch": 4.7173241228867475, + "grad_norm": 0.11368253082036972, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0241, + "step": 51900 + }, + { + "epoch": 4.7182330485366295, + "grad_norm": 0.15465250611305237, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.023, + "step": 51910 + }, + { + "epoch": 4.7191419741865115, + "grad_norm": 0.1373397707939148, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0224, + "step": 51920 + }, + { + "epoch": 4.7200508998363935, + "grad_norm": 0.13235002756118774, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0251, + "step": 51930 + }, + { + "epoch": 4.7209598254862755, + "grad_norm": 0.12452556192874908, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0244, + "step": 51940 + }, + { + "epoch": 4.7218687511361574, + "grad_norm": 0.09304965287446976, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0222, + "step": 51950 + }, + { + "epoch": 4.7227776767860385, + "grad_norm": 0.14506672322750092, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0248, + "step": 51960 + }, + { + "epoch": 4.7236866024359205, + "grad_norm": 0.14252431690692902, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0241, + "step": 51970 + }, + { + "epoch": 4.7245955280858025, + "grad_norm": 0.12282931804656982, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0255, + "step": 51980 + }, + { + "epoch": 4.7255044537356845, + "grad_norm": 0.14722289144992828, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0227, + "step": 51990 + }, + { + "epoch": 4.7264133793855665, + "grad_norm": 0.12194109708070755, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0266, + "step": 52000 + }, + { + "epoch": 4.7273223050354485, + "grad_norm": 0.1466827690601349, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0232, + "step": 52010 + }, + { + "epoch": 4.72823123068533, + "grad_norm": 0.12571385502815247, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0225, + "step": 52020 + }, + { + "epoch": 4.729140156335212, + "grad_norm": 0.12452968955039978, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0244, + "step": 52030 + }, + { + "epoch": 4.7300490819850936, + "grad_norm": 0.10884170234203339, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0263, + "step": 52040 + }, + { + "epoch": 4.7309580076349755, + "grad_norm": 0.19711889326572418, + "learning_rate": 1.94583866102197e-05, + "loss": 0.023, + "step": 52050 + }, + { + "epoch": 4.7318669332848575, + "grad_norm": 0.14209550619125366, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0244, + "step": 52060 + }, + { + "epoch": 4.7327758589347395, + "grad_norm": 0.13203857839107513, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0236, + "step": 52070 + }, + { + "epoch": 4.733684784584621, + "grad_norm": 0.15927459299564362, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0209, + "step": 52080 + }, + { + "epoch": 4.734593710234503, + "grad_norm": 0.09473446011543274, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0267, + "step": 52090 + }, + { + "epoch": 4.735502635884385, + "grad_norm": 0.1462664008140564, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0249, + "step": 52100 + }, + { + "epoch": 4.736411561534267, + "grad_norm": 0.14198748767375946, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0236, + "step": 52110 + }, + { + "epoch": 4.7373204871841486, + "grad_norm": 0.11909914761781693, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0254, + "step": 52120 + }, + { + "epoch": 4.7382294128340305, + "grad_norm": 0.11344292014837265, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0242, + "step": 52130 + }, + { + "epoch": 4.739138338483912, + "grad_norm": 0.1324811577796936, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0226, + "step": 52140 + }, + { + "epoch": 4.740047264133794, + "grad_norm": 0.12755931913852692, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0247, + "step": 52150 + }, + { + "epoch": 4.740956189783676, + "grad_norm": 0.11923783272504807, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0223, + "step": 52160 + }, + { + "epoch": 4.741865115433558, + "grad_norm": 0.13348907232284546, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0217, + "step": 52170 + }, + { + "epoch": 4.74277404108344, + "grad_norm": 0.15115724503993988, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0215, + "step": 52180 + }, + { + "epoch": 4.743682966733322, + "grad_norm": 0.14367766678333282, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0228, + "step": 52190 + }, + { + "epoch": 4.744591892383203, + "grad_norm": 0.12987282872200012, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0245, + "step": 52200 + }, + { + "epoch": 4.745500818033085, + "grad_norm": 0.11566807329654694, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0232, + "step": 52210 + }, + { + "epoch": 4.746409743682967, + "grad_norm": 0.09370800107717514, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0259, + "step": 52220 + }, + { + "epoch": 4.747318669332849, + "grad_norm": 0.13527382910251617, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0266, + "step": 52230 + }, + { + "epoch": 4.748227594982731, + "grad_norm": 0.14369480311870575, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0211, + "step": 52240 + }, + { + "epoch": 4.749136520632613, + "grad_norm": 0.17993122339248657, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0246, + "step": 52250 + }, + { + "epoch": 4.750045446282494, + "grad_norm": 0.13731321692466736, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0231, + "step": 52260 + }, + { + "epoch": 4.750954371932376, + "grad_norm": 0.22577515244483948, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0241, + "step": 52270 + }, + { + "epoch": 4.751863297582258, + "grad_norm": 0.12246169149875641, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0233, + "step": 52280 + }, + { + "epoch": 4.75277222323214, + "grad_norm": 0.19063393771648407, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0238, + "step": 52290 + }, + { + "epoch": 4.753681148882022, + "grad_norm": 0.11078287661075592, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0233, + "step": 52300 + }, + { + "epoch": 4.754590074531904, + "grad_norm": 0.13032814860343933, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0232, + "step": 52310 + }, + { + "epoch": 4.755499000181786, + "grad_norm": 0.13232187926769257, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0238, + "step": 52320 + }, + { + "epoch": 4.756407925831667, + "grad_norm": 0.14614321291446686, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0249, + "step": 52330 + }, + { + "epoch": 4.757316851481549, + "grad_norm": 0.11521051079034805, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0223, + "step": 52340 + }, + { + "epoch": 4.758225777131431, + "grad_norm": 0.1258455216884613, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0213, + "step": 52350 + }, + { + "epoch": 4.759134702781313, + "grad_norm": 0.1499365121126175, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0233, + "step": 52360 + }, + { + "epoch": 4.760043628431195, + "grad_norm": 0.2311684638261795, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0262, + "step": 52370 + }, + { + "epoch": 4.760952554081076, + "grad_norm": 0.12582753598690033, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0252, + "step": 52380 + }, + { + "epoch": 4.761861479730958, + "grad_norm": 0.13626770675182343, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0246, + "step": 52390 + }, + { + "epoch": 4.76277040538084, + "grad_norm": 0.13145852088928223, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0252, + "step": 52400 + }, + { + "epoch": 4.763679331030722, + "grad_norm": 0.14538905024528503, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0261, + "step": 52410 + }, + { + "epoch": 4.764588256680604, + "grad_norm": 0.15955302119255066, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0243, + "step": 52420 + }, + { + "epoch": 4.765497182330486, + "grad_norm": 0.14390826225280762, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0242, + "step": 52430 + }, + { + "epoch": 4.766406107980368, + "grad_norm": 0.12768442928791046, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0254, + "step": 52440 + }, + { + "epoch": 4.767315033630249, + "grad_norm": 0.1351068764925003, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0274, + "step": 52450 + }, + { + "epoch": 4.768223959280131, + "grad_norm": 0.13002586364746094, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0222, + "step": 52460 + }, + { + "epoch": 4.769132884930013, + "grad_norm": 0.1375267505645752, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0234, + "step": 52470 + }, + { + "epoch": 4.770041810579895, + "grad_norm": 0.11005139350891113, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0203, + "step": 52480 + }, + { + "epoch": 4.770950736229777, + "grad_norm": 0.13954752683639526, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0244, + "step": 52490 + }, + { + "epoch": 4.771859661879658, + "grad_norm": 0.21409738063812256, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0243, + "step": 52500 + }, + { + "epoch": 4.77276858752954, + "grad_norm": 0.12512779235839844, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0224, + "step": 52510 + }, + { + "epoch": 4.773677513179422, + "grad_norm": 0.15089385211467743, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0246, + "step": 52520 + }, + { + "epoch": 4.774586438829304, + "grad_norm": 0.12105084955692291, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0218, + "step": 52530 + }, + { + "epoch": 4.775495364479186, + "grad_norm": 0.13498850166797638, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0237, + "step": 52540 + }, + { + "epoch": 4.776404290129068, + "grad_norm": 0.13137538731098175, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0214, + "step": 52550 + }, + { + "epoch": 4.77731321577895, + "grad_norm": 0.13377325236797333, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0231, + "step": 52560 + }, + { + "epoch": 4.778222141428831, + "grad_norm": 0.1360025703907013, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0269, + "step": 52570 + }, + { + "epoch": 4.779131067078713, + "grad_norm": 0.13029442727565765, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0234, + "step": 52580 + }, + { + "epoch": 4.780039992728595, + "grad_norm": 0.1473338007926941, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0248, + "step": 52590 + }, + { + "epoch": 4.780948918378477, + "grad_norm": 0.1379014402627945, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0236, + "step": 52600 + }, + { + "epoch": 4.781857844028359, + "grad_norm": 0.15159830451011658, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0217, + "step": 52610 + }, + { + "epoch": 4.78276676967824, + "grad_norm": 0.13151605427265167, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0254, + "step": 52620 + }, + { + "epoch": 4.783675695328122, + "grad_norm": 0.14145302772521973, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0234, + "step": 52630 + }, + { + "epoch": 4.784584620978004, + "grad_norm": 0.13597945868968964, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0236, + "step": 52640 + }, + { + "epoch": 4.785493546627886, + "grad_norm": 0.2029675394296646, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0233, + "step": 52650 + }, + { + "epoch": 4.786402472277768, + "grad_norm": 0.13387733697891235, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0266, + "step": 52660 + }, + { + "epoch": 4.78731139792765, + "grad_norm": 0.26934346556663513, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0262, + "step": 52670 + }, + { + "epoch": 4.788220323577532, + "grad_norm": 0.1536525934934616, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0251, + "step": 52680 + }, + { + "epoch": 4.789129249227413, + "grad_norm": 0.10180053114891052, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0234, + "step": 52690 + }, + { + "epoch": 4.790038174877295, + "grad_norm": 0.1387081742286682, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0227, + "step": 52700 + }, + { + "epoch": 4.790947100527177, + "grad_norm": 0.17202486097812653, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0241, + "step": 52710 + }, + { + "epoch": 4.791856026177059, + "grad_norm": 0.13013288378715515, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0265, + "step": 52720 + }, + { + "epoch": 4.792764951826941, + "grad_norm": 0.10806698352098465, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0248, + "step": 52730 + }, + { + "epoch": 4.793673877476822, + "grad_norm": 0.12617100775241852, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0238, + "step": 52740 + }, + { + "epoch": 4.794582803126704, + "grad_norm": 0.1442950963973999, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0247, + "step": 52750 + }, + { + "epoch": 4.795491728776586, + "grad_norm": 0.12793275713920593, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0248, + "step": 52760 + }, + { + "epoch": 4.796400654426468, + "grad_norm": 0.19131556153297424, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0227, + "step": 52770 + }, + { + "epoch": 4.79730958007635, + "grad_norm": 0.12674179673194885, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0243, + "step": 52780 + }, + { + "epoch": 4.798218505726232, + "grad_norm": 0.12965336441993713, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0228, + "step": 52790 + }, + { + "epoch": 4.799127431376114, + "grad_norm": 0.13986195623874664, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0243, + "step": 52800 + }, + { + "epoch": 4.800036357025995, + "grad_norm": 0.11962283402681351, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0214, + "step": 52810 + }, + { + "epoch": 4.800945282675877, + "grad_norm": 0.10844763368368149, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0259, + "step": 52820 + }, + { + "epoch": 4.801854208325759, + "grad_norm": 0.1322653740644455, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0237, + "step": 52830 + }, + { + "epoch": 4.802763133975641, + "grad_norm": 0.11107612401247025, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0241, + "step": 52840 + }, + { + "epoch": 4.803672059625523, + "grad_norm": 0.13137075304985046, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0248, + "step": 52850 + }, + { + "epoch": 4.804580985275404, + "grad_norm": 0.17486003041267395, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0251, + "step": 52860 + }, + { + "epoch": 4.805489910925286, + "grad_norm": 0.12437976896762848, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0217, + "step": 52870 + }, + { + "epoch": 4.806398836575168, + "grad_norm": 0.36367136240005493, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0244, + "step": 52880 + }, + { + "epoch": 4.80730776222505, + "grad_norm": 0.12560150027275085, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0237, + "step": 52890 + }, + { + "epoch": 4.808216687874932, + "grad_norm": 0.20448465645313263, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0271, + "step": 52900 + }, + { + "epoch": 4.809125613524814, + "grad_norm": 0.13161104917526245, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0221, + "step": 52910 + }, + { + "epoch": 4.810034539174696, + "grad_norm": 0.13243693113327026, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0218, + "step": 52920 + }, + { + "epoch": 4.810943464824577, + "grad_norm": 0.11344243586063385, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0222, + "step": 52930 + }, + { + "epoch": 4.811852390474459, + "grad_norm": 0.18068456649780273, + "learning_rate": 1.910581018101795e-05, + "loss": 0.028, + "step": 52940 + }, + { + "epoch": 4.812761316124341, + "grad_norm": 0.11625206470489502, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.023, + "step": 52950 + }, + { + "epoch": 4.813670241774223, + "grad_norm": 0.13782428205013275, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0249, + "step": 52960 + }, + { + "epoch": 4.814579167424105, + "grad_norm": 0.17705221474170685, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0251, + "step": 52970 + }, + { + "epoch": 4.815488093073986, + "grad_norm": 0.13054704666137695, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0241, + "step": 52980 + }, + { + "epoch": 4.816397018723868, + "grad_norm": 0.16558201611042023, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0231, + "step": 52990 + }, + { + "epoch": 4.81730594437375, + "grad_norm": 0.13435380160808563, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0231, + "step": 53000 + }, + { + "epoch": 4.818214870023632, + "grad_norm": 0.11085676401853561, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0221, + "step": 53010 + }, + { + "epoch": 4.819123795673514, + "grad_norm": 0.15756680071353912, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0232, + "step": 53020 + }, + { + "epoch": 4.820032721323396, + "grad_norm": 0.11439207196235657, + "learning_rate": 1.907001533255806e-05, + "loss": 0.024, + "step": 53030 + }, + { + "epoch": 4.820941646973278, + "grad_norm": 0.15811273455619812, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0262, + "step": 53040 + }, + { + "epoch": 4.821850572623159, + "grad_norm": 0.12771280109882355, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0233, + "step": 53050 + }, + { + "epoch": 4.822759498273041, + "grad_norm": 0.10056714713573456, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0243, + "step": 53060 + }, + { + "epoch": 4.823668423922923, + "grad_norm": 0.12942013144493103, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0245, + "step": 53070 + }, + { + "epoch": 4.824577349572805, + "grad_norm": 0.1641487032175064, + "learning_rate": 1.905011847336159e-05, + "loss": 0.025, + "step": 53080 + }, + { + "epoch": 4.825486275222687, + "grad_norm": 0.12476909160614014, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0232, + "step": 53090 + }, + { + "epoch": 4.826395200872568, + "grad_norm": 0.1334802508354187, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0254, + "step": 53100 + }, + { + "epoch": 4.82730412652245, + "grad_norm": 0.11978837102651596, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.024, + "step": 53110 + }, + { + "epoch": 4.828213052172332, + "grad_norm": 0.10437297075986862, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0257, + "step": 53120 + }, + { + "epoch": 4.829121977822214, + "grad_norm": 0.12260022759437561, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0239, + "step": 53130 + }, + { + "epoch": 4.830030903472096, + "grad_norm": 0.1318908929824829, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0229, + "step": 53140 + }, + { + "epoch": 4.830939829121978, + "grad_norm": 0.21947146952152252, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0224, + "step": 53150 + }, + { + "epoch": 4.83184875477186, + "grad_norm": 0.16212305426597595, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.023, + "step": 53160 + }, + { + "epoch": 4.832757680421741, + "grad_norm": 0.09540637582540512, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0232, + "step": 53170 + }, + { + "epoch": 4.833666606071623, + "grad_norm": 0.10799350589513779, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0239, + "step": 53180 + }, + { + "epoch": 4.834575531721505, + "grad_norm": 0.10602448880672455, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0234, + "step": 53190 + }, + { + "epoch": 4.835484457371387, + "grad_norm": 0.14362657070159912, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0259, + "step": 53200 + }, + { + "epoch": 4.836393383021269, + "grad_norm": 0.14219294488430023, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0227, + "step": 53210 + }, + { + "epoch": 4.837302308671151, + "grad_norm": 0.10701682418584824, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0218, + "step": 53220 + }, + { + "epoch": 4.838211234321032, + "grad_norm": 0.12057019025087357, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0258, + "step": 53230 + }, + { + "epoch": 4.839120159970914, + "grad_norm": 0.1345278024673462, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0282, + "step": 53240 + }, + { + "epoch": 4.840029085620796, + "grad_norm": 0.13276906311511993, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0254, + "step": 53250 + }, + { + "epoch": 4.840938011270678, + "grad_norm": 0.14383836090564728, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0235, + "step": 53260 + }, + { + "epoch": 4.84184693692056, + "grad_norm": 0.1282692402601242, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0267, + "step": 53270 + }, + { + "epoch": 4.842755862570442, + "grad_norm": 0.1471211463212967, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0232, + "step": 53280 + }, + { + "epoch": 4.843664788220323, + "grad_norm": 0.12765805423259735, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0255, + "step": 53290 + }, + { + "epoch": 4.844573713870205, + "grad_norm": 0.15810756385326385, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0231, + "step": 53300 + }, + { + "epoch": 4.845482639520087, + "grad_norm": 0.1459038108587265, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0262, + "step": 53310 + }, + { + "epoch": 4.846391565169969, + "grad_norm": 0.15422175824642181, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0279, + "step": 53320 + }, + { + "epoch": 4.847300490819851, + "grad_norm": 0.13557446002960205, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0239, + "step": 53330 + }, + { + "epoch": 4.848209416469733, + "grad_norm": 0.13339951634407043, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0237, + "step": 53340 + }, + { + "epoch": 4.849118342119614, + "grad_norm": 0.17318473756313324, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0241, + "step": 53350 + }, + { + "epoch": 4.850027267769496, + "grad_norm": 0.14252035319805145, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0242, + "step": 53360 + }, + { + "epoch": 4.850936193419378, + "grad_norm": 0.13668350875377655, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0239, + "step": 53370 + }, + { + "epoch": 4.85184511906926, + "grad_norm": 0.1629025787115097, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0216, + "step": 53380 + }, + { + "epoch": 4.852754044719142, + "grad_norm": 0.16981454193592072, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0217, + "step": 53390 + }, + { + "epoch": 4.853662970369024, + "grad_norm": 0.1508556306362152, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0236, + "step": 53400 + }, + { + "epoch": 4.854571896018905, + "grad_norm": 0.1268685758113861, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0237, + "step": 53410 + }, + { + "epoch": 4.855480821668787, + "grad_norm": 0.10936010628938675, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0235, + "step": 53420 + }, + { + "epoch": 4.856389747318669, + "grad_norm": 0.1118282601237297, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0203, + "step": 53430 + }, + { + "epoch": 4.857298672968551, + "grad_norm": 0.11031316965818405, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0233, + "step": 53440 + }, + { + "epoch": 4.858207598618433, + "grad_norm": 0.18954239785671234, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0242, + "step": 53450 + }, + { + "epoch": 4.859116524268315, + "grad_norm": 0.17847318947315216, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0232, + "step": 53460 + }, + { + "epoch": 4.860025449918197, + "grad_norm": 0.17213088274002075, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0244, + "step": 53470 + }, + { + "epoch": 4.860934375568078, + "grad_norm": 0.12986642122268677, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0229, + "step": 53480 + }, + { + "epoch": 4.86184330121796, + "grad_norm": 0.14563430845737457, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0219, + "step": 53490 + }, + { + "epoch": 4.862752226867842, + "grad_norm": 0.13790802657604218, + "learning_rate": 1.888268495736356e-05, + "loss": 0.023, + "step": 53500 + }, + { + "epoch": 4.863661152517724, + "grad_norm": 0.16341626644134521, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0265, + "step": 53510 + }, + { + "epoch": 4.864570078167606, + "grad_norm": 0.1525239497423172, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0258, + "step": 53520 + }, + { + "epoch": 4.865479003817487, + "grad_norm": 0.1437717080116272, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0231, + "step": 53530 + }, + { + "epoch": 4.866387929467369, + "grad_norm": 0.13522560894489288, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0237, + "step": 53540 + }, + { + "epoch": 4.867296855117251, + "grad_norm": 0.13465483486652374, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0235, + "step": 53550 + }, + { + "epoch": 4.868205780767133, + "grad_norm": 0.16589243710041046, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0245, + "step": 53560 + }, + { + "epoch": 4.869114706417015, + "grad_norm": 0.12082068622112274, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0225, + "step": 53570 + }, + { + "epoch": 4.870023632066897, + "grad_norm": 0.09672912210226059, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0225, + "step": 53580 + }, + { + "epoch": 4.870932557716779, + "grad_norm": 0.1376856416463852, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0243, + "step": 53590 + }, + { + "epoch": 4.87184148336666, + "grad_norm": 0.20045335590839386, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0242, + "step": 53600 + }, + { + "epoch": 4.872750409016542, + "grad_norm": 0.1610792875289917, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0237, + "step": 53610 + }, + { + "epoch": 4.873659334666424, + "grad_norm": 0.14330686628818512, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0244, + "step": 53620 + }, + { + "epoch": 4.874568260316306, + "grad_norm": 0.1595146507024765, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0251, + "step": 53630 + }, + { + "epoch": 4.875477185966188, + "grad_norm": 0.1453242003917694, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0245, + "step": 53640 + }, + { + "epoch": 4.876386111616069, + "grad_norm": 0.12654611468315125, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0236, + "step": 53650 + }, + { + "epoch": 4.877295037265951, + "grad_norm": 0.10844719409942627, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0252, + "step": 53660 + }, + { + "epoch": 4.878203962915833, + "grad_norm": 0.1317293643951416, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0248, + "step": 53670 + }, + { + "epoch": 4.879112888565715, + "grad_norm": 0.1335703432559967, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0258, + "step": 53680 + }, + { + "epoch": 4.880021814215597, + "grad_norm": 0.14011944830417633, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0249, + "step": 53690 + }, + { + "epoch": 4.880930739865479, + "grad_norm": 0.15435843169689178, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0229, + "step": 53700 + }, + { + "epoch": 4.881839665515361, + "grad_norm": 0.12472515553236008, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0256, + "step": 53710 + }, + { + "epoch": 4.882748591165242, + "grad_norm": 0.13074728846549988, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0226, + "step": 53720 + }, + { + "epoch": 4.883657516815124, + "grad_norm": 0.10420998185873032, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0222, + "step": 53730 + }, + { + "epoch": 4.884566442465006, + "grad_norm": 0.149525448679924, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0239, + "step": 53740 + }, + { + "epoch": 4.885475368114888, + "grad_norm": 0.14659546315670013, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0224, + "step": 53750 + }, + { + "epoch": 4.88638429376477, + "grad_norm": 0.14900192618370056, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0255, + "step": 53760 + }, + { + "epoch": 4.8872932194146514, + "grad_norm": 0.14164456725120544, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0269, + "step": 53770 + }, + { + "epoch": 4.888202145064533, + "grad_norm": 0.11813420057296753, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0228, + "step": 53780 + }, + { + "epoch": 4.889111070714415, + "grad_norm": 0.13613301515579224, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0226, + "step": 53790 + }, + { + "epoch": 4.890019996364297, + "grad_norm": 0.11336851865053177, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0236, + "step": 53800 + }, + { + "epoch": 4.890928922014179, + "grad_norm": 0.1434420347213745, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0251, + "step": 53810 + }, + { + "epoch": 4.891837847664061, + "grad_norm": 0.10399433225393295, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0215, + "step": 53820 + }, + { + "epoch": 4.892746773313943, + "grad_norm": 0.22117406129837036, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0248, + "step": 53830 + }, + { + "epoch": 4.8936556989638245, + "grad_norm": 0.11450938880443573, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0234, + "step": 53840 + }, + { + "epoch": 4.8945646246137064, + "grad_norm": 0.13410167396068573, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0249, + "step": 53850 + }, + { + "epoch": 4.895473550263588, + "grad_norm": 0.13712742924690247, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0234, + "step": 53860 + }, + { + "epoch": 4.89638247591347, + "grad_norm": 0.16750529408454895, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0242, + "step": 53870 + }, + { + "epoch": 4.897291401563352, + "grad_norm": 0.16165457665920258, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0239, + "step": 53880 + }, + { + "epoch": 4.8982003272132335, + "grad_norm": 0.12560635805130005, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0234, + "step": 53890 + }, + { + "epoch": 4.8991092528631155, + "grad_norm": 0.1490941196680069, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0247, + "step": 53900 + }, + { + "epoch": 4.9000181785129975, + "grad_norm": 0.12130533158779144, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0253, + "step": 53910 + }, + { + "epoch": 4.9009271041628795, + "grad_norm": 0.13375219702720642, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.026, + "step": 53920 + }, + { + "epoch": 4.9018360298127615, + "grad_norm": 0.11845127493143082, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0236, + "step": 53930 + }, + { + "epoch": 4.902744955462643, + "grad_norm": 0.14414797723293304, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0231, + "step": 53940 + }, + { + "epoch": 4.903653881112525, + "grad_norm": 0.11794774234294891, + "learning_rate": 1.870271530813327e-05, + "loss": 0.024, + "step": 53950 + }, + { + "epoch": 4.9045628067624065, + "grad_norm": 0.14072896540164948, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0224, + "step": 53960 + }, + { + "epoch": 4.9054717324122885, + "grad_norm": 0.14453095197677612, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0256, + "step": 53970 + }, + { + "epoch": 4.9063806580621705, + "grad_norm": 0.10873862355947495, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0246, + "step": 53980 + }, + { + "epoch": 4.9072895837120525, + "grad_norm": 0.1352875530719757, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0247, + "step": 53990 + }, + { + "epoch": 4.9081985093619345, + "grad_norm": 0.13899485766887665, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0231, + "step": 54000 + }, + { + "epoch": 4.909107435011816, + "grad_norm": 0.16527903079986572, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0244, + "step": 54010 + }, + { + "epoch": 4.9100163606616976, + "grad_norm": 0.11811397224664688, + "learning_rate": 1.867466815081593e-05, + "loss": 0.0231, + "step": 54020 + }, + { + "epoch": 4.9109252863115795, + "grad_norm": 0.1621156632900238, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0259, + "step": 54030 + }, + { + "epoch": 4.9118342119614615, + "grad_norm": 0.12015821039676666, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0257, + "step": 54040 + }, + { + "epoch": 4.9127431376113435, + "grad_norm": 0.12786319851875305, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0238, + "step": 54050 + }, + { + "epoch": 4.9136520632612255, + "grad_norm": 0.1247427761554718, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0244, + "step": 54060 + }, + { + "epoch": 4.9145609889111075, + "grad_norm": 0.1433185189962387, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0228, + "step": 54070 + }, + { + "epoch": 4.915469914560989, + "grad_norm": 0.10236307978630066, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0216, + "step": 54080 + }, + { + "epoch": 4.916378840210871, + "grad_norm": 0.18566037714481354, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0245, + "step": 54090 + }, + { + "epoch": 4.917287765860753, + "grad_norm": 0.15132257342338562, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0252, + "step": 54100 + }, + { + "epoch": 4.9181966915106345, + "grad_norm": 0.14335240423679352, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0242, + "step": 54110 + }, + { + "epoch": 4.9191056171605165, + "grad_norm": 0.10394521057605743, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0229, + "step": 54120 + }, + { + "epoch": 4.920014542810398, + "grad_norm": 0.11861991882324219, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0252, + "step": 54130 + }, + { + "epoch": 4.92092346846028, + "grad_norm": 0.12377887964248657, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0229, + "step": 54140 + }, + { + "epoch": 4.921832394110162, + "grad_norm": 0.1337338387966156, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0237, + "step": 54150 + }, + { + "epoch": 4.922741319760044, + "grad_norm": 0.11461767554283142, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0237, + "step": 54160 + }, + { + "epoch": 4.923650245409926, + "grad_norm": 0.12114110589027405, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0233, + "step": 54170 + }, + { + "epoch": 4.924559171059808, + "grad_norm": 0.14482811093330383, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0218, + "step": 54180 + }, + { + "epoch": 4.9254680967096895, + "grad_norm": 0.1234435960650444, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0223, + "step": 54190 + }, + { + "epoch": 4.926377022359571, + "grad_norm": 0.1272391378879547, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0253, + "step": 54200 + }, + { + "epoch": 4.927285948009453, + "grad_norm": 0.14849472045898438, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0229, + "step": 54210 + }, + { + "epoch": 4.928194873659335, + "grad_norm": 0.1442444771528244, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0236, + "step": 54220 + }, + { + "epoch": 4.929103799309217, + "grad_norm": 0.12769447267055511, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0232, + "step": 54230 + }, + { + "epoch": 4.930012724959099, + "grad_norm": 0.15273632109165192, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0255, + "step": 54240 + }, + { + "epoch": 4.93092165060898, + "grad_norm": 0.12974566221237183, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0254, + "step": 54250 + }, + { + "epoch": 4.931830576258862, + "grad_norm": 0.10433877259492874, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0228, + "step": 54260 + }, + { + "epoch": 4.932739501908744, + "grad_norm": 0.1323772519826889, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0248, + "step": 54270 + }, + { + "epoch": 4.933648427558626, + "grad_norm": 0.13917921483516693, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0239, + "step": 54280 + }, + { + "epoch": 4.934557353208508, + "grad_norm": 0.09969920665025711, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0231, + "step": 54290 + }, + { + "epoch": 4.93546627885839, + "grad_norm": 0.13286766409873962, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0255, + "step": 54300 + }, + { + "epoch": 4.936375204508272, + "grad_norm": 0.15688550472259521, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0214, + "step": 54310 + }, + { + "epoch": 4.937284130158153, + "grad_norm": 0.15561066567897797, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0273, + "step": 54320 + }, + { + "epoch": 4.938193055808035, + "grad_norm": 0.1646977961063385, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.025, + "step": 54330 + }, + { + "epoch": 4.939101981457917, + "grad_norm": 0.12025749683380127, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0235, + "step": 54340 + }, + { + "epoch": 4.940010907107799, + "grad_norm": 0.16624440252780914, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0258, + "step": 54350 + }, + { + "epoch": 4.940919832757681, + "grad_norm": 0.12232169508934021, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0226, + "step": 54360 + }, + { + "epoch": 4.941828758407563, + "grad_norm": 0.15257371962070465, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0221, + "step": 54370 + }, + { + "epoch": 4.942737684057444, + "grad_norm": 0.12355367094278336, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0229, + "step": 54380 + }, + { + "epoch": 4.943646609707326, + "grad_norm": 0.11617542058229446, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0237, + "step": 54390 + }, + { + "epoch": 4.944555535357208, + "grad_norm": 0.13095255196094513, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0259, + "step": 54400 + }, + { + "epoch": 4.94546446100709, + "grad_norm": 0.19156891107559204, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0232, + "step": 54410 + }, + { + "epoch": 4.946373386656972, + "grad_norm": 0.16366049647331238, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0276, + "step": 54420 + }, + { + "epoch": 4.947282312306854, + "grad_norm": 0.1431087702512741, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0234, + "step": 54430 + }, + { + "epoch": 4.948191237956735, + "grad_norm": 0.13642485439777374, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0236, + "step": 54440 + }, + { + "epoch": 4.949100163606617, + "grad_norm": 0.13272343575954437, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0242, + "step": 54450 + }, + { + "epoch": 4.950009089256499, + "grad_norm": 0.14243482053279877, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0231, + "step": 54460 + }, + { + "epoch": 4.950918014906381, + "grad_norm": 0.12995608150959015, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0233, + "step": 54470 + }, + { + "epoch": 4.951826940556263, + "grad_norm": 0.13532640039920807, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0254, + "step": 54480 + }, + { + "epoch": 4.952735866206145, + "grad_norm": 0.10627775639295578, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0219, + "step": 54490 + }, + { + "epoch": 4.953644791856026, + "grad_norm": 0.11866504698991776, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0227, + "step": 54500 + }, + { + "epoch": 4.954553717505908, + "grad_norm": 0.08844336122274399, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0244, + "step": 54510 + }, + { + "epoch": 4.95546264315579, + "grad_norm": 0.1524464339017868, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0218, + "step": 54520 + }, + { + "epoch": 4.956371568805672, + "grad_norm": 0.12575656175613403, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0215, + "step": 54530 + }, + { + "epoch": 4.957280494455554, + "grad_norm": 0.16725167632102966, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0243, + "step": 54540 + }, + { + "epoch": 4.958189420105436, + "grad_norm": 0.13790234923362732, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0252, + "step": 54550 + }, + { + "epoch": 4.959098345755317, + "grad_norm": 0.11341936886310577, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.024, + "step": 54560 + }, + { + "epoch": 4.960007271405199, + "grad_norm": 0.13721714913845062, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0238, + "step": 54570 + }, + { + "epoch": 4.960916197055081, + "grad_norm": 0.14207401871681213, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0233, + "step": 54580 + }, + { + "epoch": 4.961825122704963, + "grad_norm": 0.13693100214004517, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.023, + "step": 54590 + }, + { + "epoch": 4.962734048354845, + "grad_norm": 0.11506187915802002, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.024, + "step": 54600 + }, + { + "epoch": 4.963642974004727, + "grad_norm": 0.13385780155658722, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0217, + "step": 54610 + }, + { + "epoch": 4.964551899654609, + "grad_norm": 0.12284614145755768, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0246, + "step": 54620 + }, + { + "epoch": 4.96546082530449, + "grad_norm": 0.1209849864244461, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0219, + "step": 54630 + }, + { + "epoch": 4.966369750954372, + "grad_norm": 0.12145671248435974, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0253, + "step": 54640 + }, + { + "epoch": 4.967278676604254, + "grad_norm": 0.21449938416481018, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0234, + "step": 54650 + }, + { + "epoch": 4.968187602254136, + "grad_norm": 0.13746468722820282, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0249, + "step": 54660 + }, + { + "epoch": 4.969096527904018, + "grad_norm": 0.13327519595623016, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0216, + "step": 54670 + }, + { + "epoch": 4.970005453553899, + "grad_norm": 0.1246887594461441, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.023, + "step": 54680 + }, + { + "epoch": 4.970914379203781, + "grad_norm": 0.11989650130271912, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.024, + "step": 54690 + }, + { + "epoch": 4.971823304853663, + "grad_norm": 0.11002272367477417, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0203, + "step": 54700 + }, + { + "epoch": 4.972732230503545, + "grad_norm": 0.12434858083724976, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0219, + "step": 54710 + }, + { + "epoch": 4.973641156153427, + "grad_norm": 0.1287437528371811, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0217, + "step": 54720 + }, + { + "epoch": 4.974550081803309, + "grad_norm": 0.1310037225484848, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0249, + "step": 54730 + }, + { + "epoch": 4.975459007453191, + "grad_norm": 0.1577829271554947, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0224, + "step": 54740 + }, + { + "epoch": 4.976367933103072, + "grad_norm": 0.11097663640975952, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0237, + "step": 54750 + }, + { + "epoch": 4.977276858752954, + "grad_norm": 0.12622982263565063, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0235, + "step": 54760 + }, + { + "epoch": 4.978185784402836, + "grad_norm": 0.13819244503974915, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0245, + "step": 54770 + }, + { + "epoch": 4.979094710052718, + "grad_norm": 0.14119091629981995, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0247, + "step": 54780 + }, + { + "epoch": 4.9800036357026, + "grad_norm": 0.12194442003965378, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0208, + "step": 54790 + }, + { + "epoch": 4.980912561352481, + "grad_norm": 0.1938020884990692, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0219, + "step": 54800 + }, + { + "epoch": 4.981821487002363, + "grad_norm": 0.1373331993818283, + "learning_rate": 1.835721280454979e-05, + "loss": 0.025, + "step": 54810 + }, + { + "epoch": 4.982730412652245, + "grad_norm": 0.1699085384607315, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0225, + "step": 54820 + }, + { + "epoch": 4.983639338302127, + "grad_norm": 0.11497226357460022, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0241, + "step": 54830 + }, + { + "epoch": 4.984548263952009, + "grad_norm": 0.12857207655906677, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.024, + "step": 54840 + }, + { + "epoch": 4.985457189601891, + "grad_norm": 0.17119407653808594, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0233, + "step": 54850 + }, + { + "epoch": 4.986366115251773, + "grad_norm": 0.15860430896282196, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0225, + "step": 54860 + }, + { + "epoch": 4.987275040901654, + "grad_norm": 0.15683166682720184, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0244, + "step": 54870 + }, + { + "epoch": 4.988183966551536, + "grad_norm": 0.14106744527816772, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0234, + "step": 54880 + }, + { + "epoch": 4.989092892201418, + "grad_norm": 0.1266927570104599, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0257, + "step": 54890 + }, + { + "epoch": 4.9900018178513, + "grad_norm": 0.14704281091690063, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.026, + "step": 54900 + }, + { + "epoch": 4.990910743501182, + "grad_norm": 0.10480725765228271, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0197, + "step": 54910 + }, + { + "epoch": 4.991819669151063, + "grad_norm": 0.14096154272556305, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0267, + "step": 54920 + }, + { + "epoch": 4.992728594800945, + "grad_norm": 0.12237652391195297, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.023, + "step": 54930 + }, + { + "epoch": 4.993637520450827, + "grad_norm": 0.1255248486995697, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0214, + "step": 54940 + }, + { + "epoch": 4.994546446100709, + "grad_norm": 0.11650476604700089, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0252, + "step": 54950 + }, + { + "epoch": 4.995455371750591, + "grad_norm": 0.13353388011455536, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0212, + "step": 54960 + }, + { + "epoch": 4.996364297400473, + "grad_norm": 0.15503941476345062, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0229, + "step": 54970 + }, + { + "epoch": 4.997273223050355, + "grad_norm": 0.12025678157806396, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0227, + "step": 54980 + }, + { + "epoch": 4.998182148700236, + "grad_norm": 0.1041373685002327, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0213, + "step": 54990 + }, + { + "epoch": 4.999091074350118, + "grad_norm": 0.12685514986515045, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.025, + "step": 55000 + }, + { + "epoch": 5.0, + "grad_norm": 0.1491420716047287, + "learning_rate": 1.827658600504252e-05, + "loss": 0.021, + "step": 55010 + }, + { + "epoch": 5.000908925649882, + "grad_norm": 0.204736590385437, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0246, + "step": 55020 + }, + { + "epoch": 5.001817851299764, + "grad_norm": 0.15914255380630493, + "learning_rate": 1.826851781029516e-05, + "loss": 0.022, + "step": 55030 + }, + { + "epoch": 5.002726776949646, + "grad_norm": 0.13781923055648804, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0246, + "step": 55040 + }, + { + "epoch": 5.003635702599527, + "grad_norm": 0.21346953511238098, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0224, + "step": 55050 + }, + { + "epoch": 5.004544628249409, + "grad_norm": 0.17048822343349457, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0215, + "step": 55060 + }, + { + "epoch": 5.005453553899291, + "grad_norm": 0.12075828015804291, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0248, + "step": 55070 + }, + { + "epoch": 5.006362479549173, + "grad_norm": 0.13483139872550964, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0224, + "step": 55080 + }, + { + "epoch": 5.007271405199055, + "grad_norm": 0.16250592470169067, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0246, + "step": 55090 + }, + { + "epoch": 5.008180330848937, + "grad_norm": 0.17182649672031403, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0219, + "step": 55100 + }, + { + "epoch": 5.009089256498818, + "grad_norm": 0.19181513786315918, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.023, + "step": 55110 + }, + { + "epoch": 5.0099981821487, + "grad_norm": 0.1277533918619156, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0267, + "step": 55120 + }, + { + "epoch": 5.010907107798582, + "grad_norm": 0.15059830248355865, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0208, + "step": 55130 + }, + { + "epoch": 5.011816033448464, + "grad_norm": 0.12652045488357544, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0252, + "step": 55140 + }, + { + "epoch": 5.012724959098346, + "grad_norm": 0.15666364133358002, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0239, + "step": 55150 + }, + { + "epoch": 5.013633884748228, + "grad_norm": 0.12688077986240387, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0212, + "step": 55160 + }, + { + "epoch": 5.014542810398109, + "grad_norm": 0.14667941629886627, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0256, + "step": 55170 + }, + { + "epoch": 5.015451736047991, + "grad_norm": 0.12432928383350372, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0247, + "step": 55180 + }, + { + "epoch": 5.016360661697873, + "grad_norm": 0.10767863690853119, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.025, + "step": 55190 + }, + { + "epoch": 5.017269587347755, + "grad_norm": 0.14241160452365875, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0246, + "step": 55200 + }, + { + "epoch": 5.018178512997637, + "grad_norm": 0.16349929571151733, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0209, + "step": 55210 + }, + { + "epoch": 5.019087438647519, + "grad_norm": 0.14928802847862244, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0248, + "step": 55220 + }, + { + "epoch": 5.0199963642974, + "grad_norm": 0.14552126824855804, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0229, + "step": 55230 + }, + { + "epoch": 5.020905289947282, + "grad_norm": 0.14747950434684753, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0213, + "step": 55240 + }, + { + "epoch": 5.021814215597164, + "grad_norm": 0.12726446986198425, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0243, + "step": 55250 + }, + { + "epoch": 5.022723141247046, + "grad_norm": 0.1227441132068634, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0215, + "step": 55260 + }, + { + "epoch": 5.023632066896928, + "grad_norm": 0.12037404626607895, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0229, + "step": 55270 + }, + { + "epoch": 5.02454099254681, + "grad_norm": 0.11283867806196213, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0256, + "step": 55280 + }, + { + "epoch": 5.025449918196691, + "grad_norm": 0.22008545696735382, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.023, + "step": 55290 + }, + { + "epoch": 5.026358843846573, + "grad_norm": 0.13816508650779724, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0257, + "step": 55300 + }, + { + "epoch": 5.027267769496455, + "grad_norm": 0.13091395795345306, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0246, + "step": 55310 + }, + { + "epoch": 5.028176695146337, + "grad_norm": 0.11931006610393524, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0218, + "step": 55320 + }, + { + "epoch": 5.029085620796219, + "grad_norm": 0.13975518941879272, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0229, + "step": 55330 + }, + { + "epoch": 5.029994546446101, + "grad_norm": 0.1146119013428688, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0233, + "step": 55340 + }, + { + "epoch": 5.030903472095982, + "grad_norm": 0.1220293790102005, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0237, + "step": 55350 + }, + { + "epoch": 5.031812397745864, + "grad_norm": 0.15533186495304108, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0222, + "step": 55360 + }, + { + "epoch": 5.032721323395746, + "grad_norm": 0.13209351897239685, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0238, + "step": 55370 + }, + { + "epoch": 5.033630249045628, + "grad_norm": 0.13511036336421967, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0237, + "step": 55380 + }, + { + "epoch": 5.03453917469551, + "grad_norm": 0.14686793088912964, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0246, + "step": 55390 + }, + { + "epoch": 5.035448100345392, + "grad_norm": 0.1304035186767578, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0264, + "step": 55400 + }, + { + "epoch": 5.036357025995273, + "grad_norm": 0.13835187256336212, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.024, + "step": 55410 + }, + { + "epoch": 5.037265951645155, + "grad_norm": 0.1462281048297882, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0234, + "step": 55420 + }, + { + "epoch": 5.038174877295037, + "grad_norm": 0.14237026870250702, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0226, + "step": 55430 + }, + { + "epoch": 5.039083802944919, + "grad_norm": 0.1148211881518364, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.022, + "step": 55440 + }, + { + "epoch": 5.039992728594801, + "grad_norm": 0.12634624540805817, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0216, + "step": 55450 + }, + { + "epoch": 5.040901654244683, + "grad_norm": 0.16992712020874023, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0244, + "step": 55460 + }, + { + "epoch": 5.041810579894564, + "grad_norm": 0.14038948714733124, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0232, + "step": 55470 + }, + { + "epoch": 5.042719505544446, + "grad_norm": 0.12481627613306046, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0218, + "step": 55480 + }, + { + "epoch": 5.043628431194328, + "grad_norm": 0.11030663549900055, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0231, + "step": 55490 + }, + { + "epoch": 5.04453735684421, + "grad_norm": 0.15871581435203552, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0259, + "step": 55500 + }, + { + "epoch": 5.045446282494092, + "grad_norm": 0.15290452539920807, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0222, + "step": 55510 + }, + { + "epoch": 5.046355208143974, + "grad_norm": 0.13210417330265045, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0217, + "step": 55520 + }, + { + "epoch": 5.047264133793855, + "grad_norm": 0.1237512081861496, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.023, + "step": 55530 + }, + { + "epoch": 5.048173059443737, + "grad_norm": 0.18406306207180023, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0244, + "step": 55540 + }, + { + "epoch": 5.049081985093619, + "grad_norm": 0.13324588537216187, + "learning_rate": 1.805840341767881e-05, + "loss": 0.022, + "step": 55550 + }, + { + "epoch": 5.049990910743501, + "grad_norm": 0.13313832879066467, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0227, + "step": 55560 + }, + { + "epoch": 5.050899836393383, + "grad_norm": 0.11709155887365341, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.025, + "step": 55570 + }, + { + "epoch": 5.051808762043265, + "grad_norm": 0.19984668493270874, + "learning_rate": 1.804626185776716e-05, + "loss": 0.024, + "step": 55580 + }, + { + "epoch": 5.052717687693146, + "grad_norm": 0.12069413810968399, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0238, + "step": 55590 + }, + { + "epoch": 5.053626613343028, + "grad_norm": 0.12325676530599594, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0245, + "step": 55600 + }, + { + "epoch": 5.05453553899291, + "grad_norm": 0.13144204020500183, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0259, + "step": 55610 + }, + { + "epoch": 5.055444464642792, + "grad_norm": 0.11871471256017685, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0248, + "step": 55620 + }, + { + "epoch": 5.056353390292674, + "grad_norm": 0.12040061503648758, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0211, + "step": 55630 + }, + { + "epoch": 5.057262315942556, + "grad_norm": 0.1267864853143692, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0213, + "step": 55640 + }, + { + "epoch": 5.058171241592437, + "grad_norm": 0.12122853845357895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0238, + "step": 55650 + }, + { + "epoch": 5.059080167242319, + "grad_norm": 0.09494191408157349, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0222, + "step": 55660 + }, + { + "epoch": 5.059989092892201, + "grad_norm": 0.10644174367189407, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0228, + "step": 55670 + }, + { + "epoch": 5.060898018542083, + "grad_norm": 0.11932067573070526, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0245, + "step": 55680 + }, + { + "epoch": 5.061806944191965, + "grad_norm": 0.13809025287628174, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0237, + "step": 55690 + }, + { + "epoch": 5.062715869841847, + "grad_norm": 0.10556891560554504, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0232, + "step": 55700 + }, + { + "epoch": 5.063624795491728, + "grad_norm": 0.15592911839485168, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0251, + "step": 55710 + }, + { + "epoch": 5.06453372114161, + "grad_norm": 0.12554338574409485, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0239, + "step": 55720 + }, + { + "epoch": 5.065442646791492, + "grad_norm": 0.11372953653335571, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0257, + "step": 55730 + }, + { + "epoch": 5.066351572441374, + "grad_norm": 0.15076474845409393, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0246, + "step": 55740 + }, + { + "epoch": 5.067260498091256, + "grad_norm": 0.1418251395225525, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0236, + "step": 55750 + }, + { + "epoch": 5.068169423741138, + "grad_norm": 0.13051463663578033, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0222, + "step": 55760 + }, + { + "epoch": 5.069078349391019, + "grad_norm": 0.10453107208013535, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0243, + "step": 55770 + }, + { + "epoch": 5.069987275040901, + "grad_norm": 0.11594042181968689, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0235, + "step": 55780 + }, + { + "epoch": 5.070896200690783, + "grad_norm": 0.12995189428329468, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0236, + "step": 55790 + }, + { + "epoch": 5.071805126340665, + "grad_norm": 0.127841055393219, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0246, + "step": 55800 + }, + { + "epoch": 5.072714051990547, + "grad_norm": 0.1341555416584015, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0222, + "step": 55810 + }, + { + "epoch": 5.073622977640429, + "grad_norm": 0.15184558928012848, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0239, + "step": 55820 + }, + { + "epoch": 5.0745319032903105, + "grad_norm": 0.1223970577120781, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0251, + "step": 55830 + }, + { + "epoch": 5.075440828940192, + "grad_norm": 0.1364685297012329, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0238, + "step": 55840 + }, + { + "epoch": 5.076349754590074, + "grad_norm": 0.13520747423171997, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0213, + "step": 55850 + }, + { + "epoch": 5.077258680239956, + "grad_norm": 0.14691925048828125, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0223, + "step": 55860 + }, + { + "epoch": 5.078167605889838, + "grad_norm": 0.10533205419778824, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0231, + "step": 55870 + }, + { + "epoch": 5.07907653153972, + "grad_norm": 0.14246556162834167, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0232, + "step": 55880 + }, + { + "epoch": 5.0799854571896015, + "grad_norm": 0.1421019732952118, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0238, + "step": 55890 + }, + { + "epoch": 5.0808943828394835, + "grad_norm": 0.2174583524465561, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0233, + "step": 55900 + }, + { + "epoch": 5.0818033084893655, + "grad_norm": 0.11446370929479599, + "learning_rate": 1.791256910742756e-05, + "loss": 0.0231, + "step": 55910 + }, + { + "epoch": 5.082712234139247, + "grad_norm": 0.17657975852489471, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0249, + "step": 55920 + }, + { + "epoch": 5.083621159789129, + "grad_norm": 0.13085032999515533, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0236, + "step": 55930 + }, + { + "epoch": 5.084530085439011, + "grad_norm": 0.17390844225883484, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0234, + "step": 55940 + }, + { + "epoch": 5.0854390110888925, + "grad_norm": 0.15001879632472992, + "learning_rate": 1.789634733729274e-05, + "loss": 0.024, + "step": 55950 + }, + { + "epoch": 5.0863479367387745, + "grad_norm": 0.14609140157699585, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0248, + "step": 55960 + }, + { + "epoch": 5.0872568623886565, + "grad_norm": 0.11031022667884827, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0214, + "step": 55970 + }, + { + "epoch": 5.0881657880385385, + "grad_norm": 0.151853546500206, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0255, + "step": 55980 + }, + { + "epoch": 5.0890747136884205, + "grad_norm": 0.15943428874015808, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0253, + "step": 55990 + }, + { + "epoch": 5.0899836393383024, + "grad_norm": 0.11414501070976257, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0221, + "step": 56000 + }, + { + "epoch": 5.0908925649881835, + "grad_norm": 0.14242912828922272, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0246, + "step": 56010 + }, + { + "epoch": 5.0918014906380655, + "grad_norm": 0.11065492779016495, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0225, + "step": 56020 + }, + { + "epoch": 5.0927104162879475, + "grad_norm": 0.15706956386566162, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0214, + "step": 56030 + }, + { + "epoch": 5.0936193419378295, + "grad_norm": 0.12574589252471924, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0277, + "step": 56040 + }, + { + "epoch": 5.0945282675877115, + "grad_norm": 0.13260768353939056, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0243, + "step": 56050 + }, + { + "epoch": 5.0954371932375935, + "grad_norm": 0.1290639489889145, + "learning_rate": 1.785171938215246e-05, + "loss": 0.024, + "step": 56060 + }, + { + "epoch": 5.096346118887475, + "grad_norm": 0.11673673242330551, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0273, + "step": 56070 + }, + { + "epoch": 5.097255044537357, + "grad_norm": 0.11380094289779663, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0205, + "step": 56080 + }, + { + "epoch": 5.0981639701872385, + "grad_norm": 0.12627218663692474, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0258, + "step": 56090 + }, + { + "epoch": 5.0990728958371205, + "grad_norm": 0.11874411255121231, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0249, + "step": 56100 + }, + { + "epoch": 5.0999818214870025, + "grad_norm": 0.1476520448923111, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0242, + "step": 56110 + }, + { + "epoch": 5.1008907471368845, + "grad_norm": 0.1471448689699173, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0227, + "step": 56120 + }, + { + "epoch": 5.1017996727867665, + "grad_norm": 0.17307911813259125, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0249, + "step": 56130 + }, + { + "epoch": 5.102708598436648, + "grad_norm": 0.1585296094417572, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0241, + "step": 56140 + }, + { + "epoch": 5.10361752408653, + "grad_norm": 0.13113029301166534, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0256, + "step": 56150 + }, + { + "epoch": 5.104526449736412, + "grad_norm": 0.13553199172019958, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0234, + "step": 56160 + }, + { + "epoch": 5.1054353753862936, + "grad_norm": 0.16008883714675903, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0232, + "step": 56170 + }, + { + "epoch": 5.1063443010361755, + "grad_norm": 0.12604300677776337, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0244, + "step": 56180 + }, + { + "epoch": 5.107253226686057, + "grad_norm": 0.17562244832515717, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0221, + "step": 56190 + }, + { + "epoch": 5.108162152335939, + "grad_norm": 0.10947135835886002, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0232, + "step": 56200 + }, + { + "epoch": 5.109071077985821, + "grad_norm": 0.13614079356193542, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0246, + "step": 56210 + }, + { + "epoch": 5.109980003635703, + "grad_norm": 0.11000818014144897, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0251, + "step": 56220 + }, + { + "epoch": 5.110888929285585, + "grad_norm": 0.14499443769454956, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0227, + "step": 56230 + }, + { + "epoch": 5.111797854935467, + "grad_norm": 0.10847815871238708, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0207, + "step": 56240 + }, + { + "epoch": 5.1127067805853486, + "grad_norm": 0.09997531771659851, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0224, + "step": 56250 + }, + { + "epoch": 5.11361570623523, + "grad_norm": 0.10126059502363205, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0226, + "step": 56260 + }, + { + "epoch": 5.114524631885112, + "grad_norm": 0.10840486735105515, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0259, + "step": 56270 + }, + { + "epoch": 5.115433557534994, + "grad_norm": 0.11693456768989563, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0215, + "step": 56280 + }, + { + "epoch": 5.116342483184876, + "grad_norm": 0.11027580499649048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0218, + "step": 56290 + }, + { + "epoch": 5.117251408834758, + "grad_norm": 0.11018101871013641, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0233, + "step": 56300 + }, + { + "epoch": 5.11816033448464, + "grad_norm": 0.14273205399513245, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.025, + "step": 56310 + }, + { + "epoch": 5.119069260134521, + "grad_norm": 0.1434493064880371, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0243, + "step": 56320 + }, + { + "epoch": 5.119978185784403, + "grad_norm": 0.1383335292339325, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0231, + "step": 56330 + }, + { + "epoch": 5.120887111434285, + "grad_norm": 0.14351756870746613, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0233, + "step": 56340 + }, + { + "epoch": 5.121796037084167, + "grad_norm": 0.12898074090480804, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0252, + "step": 56350 + }, + { + "epoch": 5.122704962734049, + "grad_norm": 0.1343463957309723, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0252, + "step": 56360 + }, + { + "epoch": 5.123613888383931, + "grad_norm": 0.1732936054468155, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0238, + "step": 56370 + }, + { + "epoch": 5.124522814033812, + "grad_norm": 0.1447392702102661, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0235, + "step": 56380 + }, + { + "epoch": 5.125431739683694, + "grad_norm": 0.18303944170475006, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0246, + "step": 56390 + }, + { + "epoch": 5.126340665333576, + "grad_norm": 0.10783504694700241, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0253, + "step": 56400 + }, + { + "epoch": 5.127249590983458, + "grad_norm": 0.16109517216682434, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0234, + "step": 56410 + }, + { + "epoch": 5.12815851663334, + "grad_norm": 0.13928014039993286, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0235, + "step": 56420 + }, + { + "epoch": 5.129067442283222, + "grad_norm": 0.1538034826517105, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0246, + "step": 56430 + }, + { + "epoch": 5.129976367933103, + "grad_norm": 0.1336008757352829, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0237, + "step": 56440 + }, + { + "epoch": 5.130885293582985, + "grad_norm": 0.12488743662834167, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0246, + "step": 56450 + }, + { + "epoch": 5.131794219232867, + "grad_norm": 0.11674907058477402, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0232, + "step": 56460 + }, + { + "epoch": 5.132703144882749, + "grad_norm": 0.11685045808553696, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0225, + "step": 56470 + }, + { + "epoch": 5.133612070532631, + "grad_norm": 0.16523437201976776, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0219, + "step": 56480 + }, + { + "epoch": 5.134520996182513, + "grad_norm": 0.12999723851680756, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0235, + "step": 56490 + }, + { + "epoch": 5.135429921832394, + "grad_norm": 0.15183143317699432, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0208, + "step": 56500 + }, + { + "epoch": 5.136338847482276, + "grad_norm": 0.1528136283159256, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0209, + "step": 56510 + }, + { + "epoch": 5.137247773132158, + "grad_norm": 0.17041757702827454, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0273, + "step": 56520 + }, + { + "epoch": 5.13815669878204, + "grad_norm": 0.16388514637947083, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0233, + "step": 56530 + }, + { + "epoch": 5.139065624431922, + "grad_norm": 0.13290198147296906, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0251, + "step": 56540 + }, + { + "epoch": 5.139974550081804, + "grad_norm": 0.12539373338222504, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0226, + "step": 56550 + }, + { + "epoch": 5.140883475731685, + "grad_norm": 0.14635150134563446, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0208, + "step": 56560 + }, + { + "epoch": 5.141792401381567, + "grad_norm": 0.12424039095640182, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0217, + "step": 56570 + }, + { + "epoch": 5.142701327031449, + "grad_norm": 0.11512432247400284, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0211, + "step": 56580 + }, + { + "epoch": 5.143610252681331, + "grad_norm": 0.16243059933185577, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0242, + "step": 56590 + }, + { + "epoch": 5.144519178331213, + "grad_norm": 0.11651045829057693, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0219, + "step": 56600 + }, + { + "epoch": 5.145428103981095, + "grad_norm": 0.12301761656999588, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0214, + "step": 56610 + }, + { + "epoch": 5.146337029630976, + "grad_norm": 0.12844879925251007, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0244, + "step": 56620 + }, + { + "epoch": 5.147245955280858, + "grad_norm": 0.17176318168640137, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0221, + "step": 56630 + }, + { + "epoch": 5.14815488093074, + "grad_norm": 0.13641437888145447, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0227, + "step": 56640 + }, + { + "epoch": 5.149063806580622, + "grad_norm": 0.14837805926799774, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0217, + "step": 56650 + }, + { + "epoch": 5.149972732230504, + "grad_norm": 0.17458097636699677, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0244, + "step": 56660 + }, + { + "epoch": 5.150881657880386, + "grad_norm": 0.13446441292762756, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0254, + "step": 56670 + }, + { + "epoch": 5.151790583530267, + "grad_norm": 0.1471051275730133, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0233, + "step": 56680 + }, + { + "epoch": 5.152699509180149, + "grad_norm": 0.1431380659341812, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0214, + "step": 56690 + }, + { + "epoch": 5.153608434830031, + "grad_norm": 0.13457417488098145, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0213, + "step": 56700 + }, + { + "epoch": 5.154517360479913, + "grad_norm": 0.12400725483894348, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0224, + "step": 56710 + }, + { + "epoch": 5.155426286129795, + "grad_norm": 0.1400112807750702, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0228, + "step": 56720 + }, + { + "epoch": 5.156335211779677, + "grad_norm": 0.1507963240146637, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.022, + "step": 56730 + }, + { + "epoch": 5.157244137429558, + "grad_norm": 0.13885658979415894, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0225, + "step": 56740 + }, + { + "epoch": 5.15815306307944, + "grad_norm": 0.21753084659576416, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0245, + "step": 56750 + }, + { + "epoch": 5.159061988729322, + "grad_norm": 0.1371556967496872, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0211, + "step": 56760 + }, + { + "epoch": 5.159970914379204, + "grad_norm": 0.10847539454698563, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0215, + "step": 56770 + }, + { + "epoch": 5.160879840029086, + "grad_norm": 0.1215808317065239, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0222, + "step": 56780 + }, + { + "epoch": 5.161788765678968, + "grad_norm": 0.14400199055671692, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0218, + "step": 56790 + }, + { + "epoch": 5.162697691328849, + "grad_norm": 0.13418276607990265, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0229, + "step": 56800 + }, + { + "epoch": 5.163606616978731, + "grad_norm": 0.13051718473434448, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.024, + "step": 56810 + }, + { + "epoch": 5.164515542628613, + "grad_norm": 0.14188188314437866, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0204, + "step": 56820 + }, + { + "epoch": 5.165424468278495, + "grad_norm": 0.11975192278623581, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0231, + "step": 56830 + }, + { + "epoch": 5.166333393928377, + "grad_norm": 0.13909178972244263, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0267, + "step": 56840 + }, + { + "epoch": 5.167242319578259, + "grad_norm": 0.1752159297466278, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0243, + "step": 56850 + }, + { + "epoch": 5.16815124522814, + "grad_norm": 0.11043014377355576, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0221, + "step": 56860 + }, + { + "epoch": 5.169060170878022, + "grad_norm": 0.12694723904132843, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0226, + "step": 56870 + }, + { + "epoch": 5.169969096527904, + "grad_norm": 0.12445582449436188, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0253, + "step": 56880 + }, + { + "epoch": 5.170878022177786, + "grad_norm": 0.11933369189500809, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0231, + "step": 56890 + }, + { + "epoch": 5.171786947827668, + "grad_norm": 0.10746628791093826, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0237, + "step": 56900 + }, + { + "epoch": 5.17269587347755, + "grad_norm": 0.16682831943035126, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0214, + "step": 56910 + }, + { + "epoch": 5.173604799127431, + "grad_norm": 0.12672467529773712, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.023, + "step": 56920 + }, + { + "epoch": 5.174513724777313, + "grad_norm": 0.11362654715776443, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0221, + "step": 56930 + }, + { + "epoch": 5.175422650427195, + "grad_norm": 0.15926796197891235, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0245, + "step": 56940 + }, + { + "epoch": 5.176331576077077, + "grad_norm": 0.13457393646240234, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.024, + "step": 56950 + }, + { + "epoch": 5.177240501726959, + "grad_norm": 0.13304638862609863, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0223, + "step": 56960 + }, + { + "epoch": 5.178149427376841, + "grad_norm": 0.11213219165802002, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0225, + "step": 56970 + }, + { + "epoch": 5.179058353026722, + "grad_norm": 0.12427347153425217, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0233, + "step": 56980 + }, + { + "epoch": 5.179967278676604, + "grad_norm": 0.1358128935098648, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0217, + "step": 56990 + }, + { + "epoch": 5.180876204326486, + "grad_norm": 0.11378312855958939, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0241, + "step": 57000 + }, + { + "epoch": 5.181785129976368, + "grad_norm": 0.09957848489284515, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0209, + "step": 57010 + }, + { + "epoch": 5.18269405562625, + "grad_norm": 0.08449706435203552, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0217, + "step": 57020 + }, + { + "epoch": 5.183602981276132, + "grad_norm": 0.10604853928089142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0215, + "step": 57030 + }, + { + "epoch": 5.184511906926013, + "grad_norm": 0.16208556294441223, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0263, + "step": 57040 + }, + { + "epoch": 5.185420832575895, + "grad_norm": 0.11290372908115387, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0213, + "step": 57050 + }, + { + "epoch": 5.186329758225777, + "grad_norm": 0.14501704275608063, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.021, + "step": 57060 + }, + { + "epoch": 5.187238683875659, + "grad_norm": 0.13852928578853607, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0234, + "step": 57070 + }, + { + "epoch": 5.188147609525541, + "grad_norm": 0.15155471861362457, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0215, + "step": 57080 + }, + { + "epoch": 5.189056535175423, + "grad_norm": 0.12027377635240555, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0241, + "step": 57090 + }, + { + "epoch": 5.189965460825304, + "grad_norm": 0.11472401767969131, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0228, + "step": 57100 + }, + { + "epoch": 5.190874386475186, + "grad_norm": 0.12479842454195023, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0226, + "step": 57110 + }, + { + "epoch": 5.191783312125068, + "grad_norm": 0.17310675978660583, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0226, + "step": 57120 + }, + { + "epoch": 5.19269223777495, + "grad_norm": 0.11856698244810104, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0225, + "step": 57130 + }, + { + "epoch": 5.193601163424832, + "grad_norm": 0.12671782076358795, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.022, + "step": 57140 + }, + { + "epoch": 5.194510089074714, + "grad_norm": 0.19499114155769348, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0227, + "step": 57150 + }, + { + "epoch": 5.195419014724595, + "grad_norm": 0.15714263916015625, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0255, + "step": 57160 + }, + { + "epoch": 5.196327940374477, + "grad_norm": 0.11425133049488068, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0238, + "step": 57170 + }, + { + "epoch": 5.197236866024359, + "grad_norm": 0.13953539729118347, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.023, + "step": 57180 + }, + { + "epoch": 5.198145791674241, + "grad_norm": 0.1642838716506958, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0232, + "step": 57190 + }, + { + "epoch": 5.199054717324123, + "grad_norm": 0.16859537363052368, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0253, + "step": 57200 + }, + { + "epoch": 5.199963642974005, + "grad_norm": 0.14119184017181396, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0253, + "step": 57210 + }, + { + "epoch": 5.200872568623886, + "grad_norm": 0.13571083545684814, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0244, + "step": 57220 + }, + { + "epoch": 5.201781494273768, + "grad_norm": 0.12355265766382217, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0258, + "step": 57230 + }, + { + "epoch": 5.20269041992365, + "grad_norm": 0.1592511385679245, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0216, + "step": 57240 + }, + { + "epoch": 5.203599345573532, + "grad_norm": 0.10288811475038528, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0249, + "step": 57250 + }, + { + "epoch": 5.204508271223414, + "grad_norm": 0.10730777680873871, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0241, + "step": 57260 + }, + { + "epoch": 5.205417196873296, + "grad_norm": 0.12930697202682495, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0223, + "step": 57270 + }, + { + "epoch": 5.206326122523178, + "grad_norm": 0.1258140504360199, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0256, + "step": 57280 + }, + { + "epoch": 5.207235048173059, + "grad_norm": 0.10564292222261429, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.022, + "step": 57290 + }, + { + "epoch": 5.208143973822941, + "grad_norm": 0.14918892085552216, + "learning_rate": 1.734692525292751e-05, + "loss": 0.022, + "step": 57300 + }, + { + "epoch": 5.209052899472823, + "grad_norm": 0.131225124001503, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.024, + "step": 57310 + }, + { + "epoch": 5.209961825122705, + "grad_norm": 0.13321223855018616, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0242, + "step": 57320 + }, + { + "epoch": 5.210870750772587, + "grad_norm": 0.11405815929174423, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0217, + "step": 57330 + }, + { + "epoch": 5.211779676422468, + "grad_norm": 0.12762369215488434, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0235, + "step": 57340 + }, + { + "epoch": 5.21268860207235, + "grad_norm": 0.13802139461040497, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0227, + "step": 57350 + }, + { + "epoch": 5.213597527722232, + "grad_norm": 0.19514873623847961, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0262, + "step": 57360 + }, + { + "epoch": 5.214506453372114, + "grad_norm": 0.19027923047542572, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0251, + "step": 57370 + }, + { + "epoch": 5.215415379021996, + "grad_norm": 0.11935082823038101, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0221, + "step": 57380 + }, + { + "epoch": 5.216324304671878, + "grad_norm": 0.1414990872144699, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0205, + "step": 57390 + }, + { + "epoch": 5.21723323032176, + "grad_norm": 0.13389523327350616, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0247, + "step": 57400 + }, + { + "epoch": 5.218142155971641, + "grad_norm": 0.15456382930278778, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0249, + "step": 57410 + }, + { + "epoch": 5.219051081621523, + "grad_norm": 0.13648319244384766, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0238, + "step": 57420 + }, + { + "epoch": 5.219960007271405, + "grad_norm": 0.11395590007305145, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0226, + "step": 57430 + }, + { + "epoch": 5.220868932921287, + "grad_norm": 0.11067435145378113, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0217, + "step": 57440 + }, + { + "epoch": 5.221777858571169, + "grad_norm": 0.12919221818447113, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0243, + "step": 57450 + }, + { + "epoch": 5.22268678422105, + "grad_norm": 0.15301954746246338, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0231, + "step": 57460 + }, + { + "epoch": 5.223595709870932, + "grad_norm": 0.11258959770202637, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0213, + "step": 57470 + }, + { + "epoch": 5.224504635520814, + "grad_norm": 0.14580893516540527, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.025, + "step": 57480 + }, + { + "epoch": 5.225413561170696, + "grad_norm": 0.13352857530117035, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0203, + "step": 57490 + }, + { + "epoch": 5.226322486820578, + "grad_norm": 0.1420784890651703, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0222, + "step": 57500 + }, + { + "epoch": 5.22723141247046, + "grad_norm": 0.1880870759487152, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0222, + "step": 57510 + }, + { + "epoch": 5.228140338120342, + "grad_norm": 0.1183379665017128, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0228, + "step": 57520 + }, + { + "epoch": 5.229049263770223, + "grad_norm": 0.1066339984536171, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0227, + "step": 57530 + }, + { + "epoch": 5.229958189420105, + "grad_norm": 0.13167518377304077, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0217, + "step": 57540 + }, + { + "epoch": 5.230867115069987, + "grad_norm": 0.15246644616127014, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0221, + "step": 57550 + }, + { + "epoch": 5.231776040719869, + "grad_norm": 0.12967514991760254, + "learning_rate": 1.72407140459686e-05, + "loss": 0.023, + "step": 57560 + }, + { + "epoch": 5.232684966369751, + "grad_norm": 0.13349492847919464, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0227, + "step": 57570 + }, + { + "epoch": 5.233593892019633, + "grad_norm": 0.12917035818099976, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0222, + "step": 57580 + }, + { + "epoch": 5.234502817669514, + "grad_norm": 0.11816420406103134, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0248, + "step": 57590 + }, + { + "epoch": 5.235411743319396, + "grad_norm": 0.1425902545452118, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0217, + "step": 57600 + }, + { + "epoch": 5.236320668969278, + "grad_norm": 0.33865126967430115, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0227, + "step": 57610 + }, + { + "epoch": 5.23722959461916, + "grad_norm": 0.1376231461763382, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0222, + "step": 57620 + }, + { + "epoch": 5.238138520269042, + "grad_norm": 0.14961141347885132, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0213, + "step": 57630 + }, + { + "epoch": 5.239047445918924, + "grad_norm": 0.15407879650592804, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.024, + "step": 57640 + }, + { + "epoch": 5.239956371568805, + "grad_norm": 0.16634002327919006, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0226, + "step": 57650 + }, + { + "epoch": 5.240865297218687, + "grad_norm": 0.16021938621997833, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0215, + "step": 57660 + }, + { + "epoch": 5.241774222868569, + "grad_norm": 0.12387116998434067, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0228, + "step": 57670 + }, + { + "epoch": 5.242683148518451, + "grad_norm": 0.16716749966144562, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0212, + "step": 57680 + }, + { + "epoch": 5.243592074168333, + "grad_norm": 0.16057637333869934, + "learning_rate": 1.718756496420356e-05, + "loss": 0.022, + "step": 57690 + }, + { + "epoch": 5.244500999818215, + "grad_norm": 0.1636778712272644, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0244, + "step": 57700 + }, + { + "epoch": 5.245409925468096, + "grad_norm": 0.12338866293430328, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0201, + "step": 57710 + }, + { + "epoch": 5.246318851117978, + "grad_norm": 0.11110564321279526, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0223, + "step": 57720 + }, + { + "epoch": 5.24722777676786, + "grad_norm": 0.14659273624420166, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0218, + "step": 57730 + }, + { + "epoch": 5.248136702417742, + "grad_norm": 0.1127043142914772, + "learning_rate": 1.716711549651584e-05, + "loss": 0.023, + "step": 57740 + }, + { + "epoch": 5.249045628067624, + "grad_norm": 0.13668079674243927, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0237, + "step": 57750 + }, + { + "epoch": 5.249954553717506, + "grad_norm": 0.13991941511631012, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0232, + "step": 57760 + }, + { + "epoch": 5.250863479367387, + "grad_norm": 0.10941290855407715, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0204, + "step": 57770 + }, + { + "epoch": 5.251772405017269, + "grad_norm": 0.13875101506710052, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0217, + "step": 57780 + }, + { + "epoch": 5.252681330667151, + "grad_norm": 0.12238472700119019, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0238, + "step": 57790 + }, + { + "epoch": 5.253590256317033, + "grad_norm": 0.15922610461711884, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0247, + "step": 57800 + }, + { + "epoch": 5.254499181966915, + "grad_norm": 0.11495624482631683, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0224, + "step": 57810 + }, + { + "epoch": 5.255408107616797, + "grad_norm": 0.13477064669132233, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0194, + "step": 57820 + }, + { + "epoch": 5.256317033266678, + "grad_norm": 0.15940824151039124, + "learning_rate": 1.713029611235338e-05, + "loss": 0.022, + "step": 57830 + }, + { + "epoch": 5.25722595891656, + "grad_norm": 0.1193910464644432, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0226, + "step": 57840 + }, + { + "epoch": 5.258134884566442, + "grad_norm": 0.11409645527601242, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0228, + "step": 57850 + }, + { + "epoch": 5.259043810216324, + "grad_norm": 0.14184844493865967, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0213, + "step": 57860 + }, + { + "epoch": 5.259952735866206, + "grad_norm": 0.12105752527713776, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0212, + "step": 57870 + }, + { + "epoch": 5.260861661516088, + "grad_norm": 0.14995349943637848, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0196, + "step": 57880 + }, + { + "epoch": 5.2617705871659695, + "grad_norm": 0.12429267168045044, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0243, + "step": 57890 + }, + { + "epoch": 5.2626795128158514, + "grad_norm": 0.14729337394237518, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0244, + "step": 57900 + }, + { + "epoch": 5.263588438465733, + "grad_norm": 0.12568365037441254, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0215, + "step": 57910 + }, + { + "epoch": 5.264497364115615, + "grad_norm": 0.1339317262172699, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0233, + "step": 57920 + }, + { + "epoch": 5.265406289765497, + "grad_norm": 0.11310476809740067, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0236, + "step": 57930 + }, + { + "epoch": 5.266315215415379, + "grad_norm": 0.08920560777187347, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0199, + "step": 57940 + }, + { + "epoch": 5.2672241410652605, + "grad_norm": 0.11753471940755844, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0213, + "step": 57950 + }, + { + "epoch": 5.2681330667151425, + "grad_norm": 0.1465528905391693, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0223, + "step": 57960 + }, + { + "epoch": 5.2690419923650245, + "grad_norm": 0.12999673187732697, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0232, + "step": 57970 + }, + { + "epoch": 5.2699509180149064, + "grad_norm": 0.12163761258125305, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0208, + "step": 57980 + }, + { + "epoch": 5.270859843664788, + "grad_norm": 0.12030892819166183, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0227, + "step": 57990 + }, + { + "epoch": 5.27176876931467, + "grad_norm": 0.15046241879463196, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0224, + "step": 58000 + }, + { + "epoch": 5.2726776949645515, + "grad_norm": 0.11773907393217087, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0202, + "step": 58010 + }, + { + "epoch": 5.2735866206144335, + "grad_norm": 0.12249220162630081, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0245, + "step": 58020 + }, + { + "epoch": 5.2744955462643155, + "grad_norm": 0.14014528691768646, + "learning_rate": 1.70484286724112e-05, + "loss": 0.022, + "step": 58030 + }, + { + "epoch": 5.2754044719141975, + "grad_norm": 0.13742280006408691, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0232, + "step": 58040 + }, + { + "epoch": 5.2763133975640795, + "grad_norm": 0.12285032123327255, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.023, + "step": 58050 + }, + { + "epoch": 5.2772223232139615, + "grad_norm": 0.13292327523231506, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0234, + "step": 58060 + }, + { + "epoch": 5.2781312488638426, + "grad_norm": 0.1403045356273651, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0249, + "step": 58070 + }, + { + "epoch": 5.2790401745137245, + "grad_norm": 0.12912780046463013, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0219, + "step": 58080 + }, + { + "epoch": 5.2799491001636065, + "grad_norm": 0.10972423106431961, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.022, + "step": 58090 + }, + { + "epoch": 5.2808580258134885, + "grad_norm": 0.10014953464269638, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0236, + "step": 58100 + }, + { + "epoch": 5.2817669514633705, + "grad_norm": 0.09534083306789398, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0226, + "step": 58110 + }, + { + "epoch": 5.2826758771132525, + "grad_norm": 0.13475048542022705, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0241, + "step": 58120 + }, + { + "epoch": 5.283584802763134, + "grad_norm": 0.17582017183303833, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0237, + "step": 58130 + }, + { + "epoch": 5.284493728413016, + "grad_norm": 0.11976093053817749, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0218, + "step": 58140 + }, + { + "epoch": 5.2854026540628976, + "grad_norm": 0.16841977834701538, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0206, + "step": 58150 + }, + { + "epoch": 5.2863115797127795, + "grad_norm": 0.12374503165483475, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0254, + "step": 58160 + }, + { + "epoch": 5.2872205053626615, + "grad_norm": 0.13539628684520721, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0224, + "step": 58170 + }, + { + "epoch": 5.2881294310125435, + "grad_norm": 0.12888680398464203, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0214, + "step": 58180 + }, + { + "epoch": 5.289038356662425, + "grad_norm": 0.15535278618335724, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.025, + "step": 58190 + }, + { + "epoch": 5.289947282312307, + "grad_norm": 0.139631450176239, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0248, + "step": 58200 + }, + { + "epoch": 5.290856207962189, + "grad_norm": 0.17433786392211914, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0251, + "step": 58210 + }, + { + "epoch": 5.291765133612071, + "grad_norm": 0.16461971402168274, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.023, + "step": 58220 + }, + { + "epoch": 5.292674059261953, + "grad_norm": 0.1163829118013382, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0231, + "step": 58230 + }, + { + "epoch": 5.2935829849118345, + "grad_norm": 0.12525297701358795, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0213, + "step": 58240 + }, + { + "epoch": 5.294491910561716, + "grad_norm": 0.1220097541809082, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0229, + "step": 58250 + }, + { + "epoch": 5.295400836211598, + "grad_norm": 0.14983989298343658, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0215, + "step": 58260 + }, + { + "epoch": 5.29630976186148, + "grad_norm": 0.16856689751148224, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.021, + "step": 58270 + }, + { + "epoch": 5.297218687511362, + "grad_norm": 0.16502653062343597, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0207, + "step": 58280 + }, + { + "epoch": 5.298127613161244, + "grad_norm": 0.18343788385391235, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0224, + "step": 58290 + }, + { + "epoch": 5.299036538811126, + "grad_norm": 0.14453081786632538, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0235, + "step": 58300 + }, + { + "epoch": 5.299945464461008, + "grad_norm": 0.14478428661823273, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0244, + "step": 58310 + }, + { + "epoch": 5.300854390110889, + "grad_norm": 0.1704445332288742, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0248, + "step": 58320 + }, + { + "epoch": 5.301763315760771, + "grad_norm": 0.15178026258945465, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0236, + "step": 58330 + }, + { + "epoch": 5.302672241410653, + "grad_norm": 0.1341109275817871, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0235, + "step": 58340 + }, + { + "epoch": 5.303581167060535, + "grad_norm": 0.10304109752178192, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0225, + "step": 58350 + }, + { + "epoch": 5.304490092710417, + "grad_norm": 0.12205192446708679, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0222, + "step": 58360 + }, + { + "epoch": 5.305399018360298, + "grad_norm": 0.16887266933918, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0219, + "step": 58370 + }, + { + "epoch": 5.30630794401018, + "grad_norm": 0.20658540725708008, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0225, + "step": 58380 + }, + { + "epoch": 5.307216869660062, + "grad_norm": 0.16493725776672363, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0215, + "step": 58390 + }, + { + "epoch": 5.308125795309944, + "grad_norm": 0.18184597790241241, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0241, + "step": 58400 + }, + { + "epoch": 5.309034720959826, + "grad_norm": 0.1842091828584671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0222, + "step": 58410 + }, + { + "epoch": 5.309943646609708, + "grad_norm": 0.14861765503883362, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0225, + "step": 58420 + }, + { + "epoch": 5.31085257225959, + "grad_norm": 0.12004588544368744, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0213, + "step": 58430 + }, + { + "epoch": 5.311761497909471, + "grad_norm": 0.139350026845932, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0219, + "step": 58440 + }, + { + "epoch": 5.312670423559353, + "grad_norm": 0.13429737091064453, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0202, + "step": 58450 + }, + { + "epoch": 5.313579349209235, + "grad_norm": 0.17436687648296356, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0249, + "step": 58460 + }, + { + "epoch": 5.314488274859117, + "grad_norm": 0.15689419209957123, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0208, + "step": 58470 + }, + { + "epoch": 5.315397200508999, + "grad_norm": 0.12998653948307037, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0233, + "step": 58480 + }, + { + "epoch": 5.31630612615888, + "grad_norm": 0.12556706368923187, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0227, + "step": 58490 + }, + { + "epoch": 5.317215051808762, + "grad_norm": 0.11061535030603409, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0228, + "step": 58500 + }, + { + "epoch": 5.318123977458644, + "grad_norm": 0.12490083277225494, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0243, + "step": 58510 + }, + { + "epoch": 5.319032903108526, + "grad_norm": 0.1282125562429428, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0201, + "step": 58520 + }, + { + "epoch": 5.319941828758408, + "grad_norm": 0.17079247534275055, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0256, + "step": 58530 + }, + { + "epoch": 5.32085075440829, + "grad_norm": 0.10223444551229477, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0235, + "step": 58540 + }, + { + "epoch": 5.321759680058172, + "grad_norm": 0.14105358719825745, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0229, + "step": 58550 + }, + { + "epoch": 5.322668605708053, + "grad_norm": 0.12736719846725464, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0215, + "step": 58560 + }, + { + "epoch": 5.323577531357935, + "grad_norm": 0.15196196734905243, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0244, + "step": 58570 + }, + { + "epoch": 5.324486457007817, + "grad_norm": 0.13288821280002594, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0233, + "step": 58580 + }, + { + "epoch": 5.325395382657699, + "grad_norm": 0.1147209107875824, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0225, + "step": 58590 + }, + { + "epoch": 5.326304308307581, + "grad_norm": 0.12412390857934952, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0212, + "step": 58600 + }, + { + "epoch": 5.327213233957462, + "grad_norm": 0.12346234172582626, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.023, + "step": 58610 + }, + { + "epoch": 5.328122159607344, + "grad_norm": 0.12353117763996124, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0224, + "step": 58620 + }, + { + "epoch": 5.329031085257226, + "grad_norm": 0.13518260419368744, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0233, + "step": 58630 + }, + { + "epoch": 5.329940010907108, + "grad_norm": 0.1230369582772255, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0221, + "step": 58640 + }, + { + "epoch": 5.33084893655699, + "grad_norm": 0.13352274894714355, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0212, + "step": 58650 + }, + { + "epoch": 5.331757862206872, + "grad_norm": 0.12144359201192856, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.025, + "step": 58660 + }, + { + "epoch": 5.332666787856754, + "grad_norm": 0.1265411376953125, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0263, + "step": 58670 + }, + { + "epoch": 5.333575713506635, + "grad_norm": 0.14007128775119781, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0253, + "step": 58680 + }, + { + "epoch": 5.334484639156517, + "grad_norm": 0.1360645443201065, + "learning_rate": 1.677783817030201e-05, + "loss": 0.023, + "step": 58690 + }, + { + "epoch": 5.335393564806399, + "grad_norm": 0.14288076758384705, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0216, + "step": 58700 + }, + { + "epoch": 5.336302490456281, + "grad_norm": 0.10392934828996658, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0223, + "step": 58710 + }, + { + "epoch": 5.337211416106163, + "grad_norm": 0.13166217505931854, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0215, + "step": 58720 + }, + { + "epoch": 5.338120341756044, + "grad_norm": 0.12720638513565063, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0201, + "step": 58730 + }, + { + "epoch": 5.339029267405926, + "grad_norm": 0.17845873534679413, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0234, + "step": 58740 + }, + { + "epoch": 5.339938193055808, + "grad_norm": 0.12525436282157898, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0218, + "step": 58750 + }, + { + "epoch": 5.34084711870569, + "grad_norm": 0.13890981674194336, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0248, + "step": 58760 + }, + { + "epoch": 5.341756044355572, + "grad_norm": 0.12298829108476639, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.024, + "step": 58770 + }, + { + "epoch": 5.342664970005454, + "grad_norm": 0.12218773365020752, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0222, + "step": 58780 + }, + { + "epoch": 5.343573895655336, + "grad_norm": 0.13724501430988312, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0222, + "step": 58790 + }, + { + "epoch": 5.344482821305217, + "grad_norm": 0.1556575894355774, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0268, + "step": 58800 + }, + { + "epoch": 5.345391746955099, + "grad_norm": 0.13097979128360748, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0249, + "step": 58810 + }, + { + "epoch": 5.346300672604981, + "grad_norm": 0.140180766582489, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0239, + "step": 58820 + }, + { + "epoch": 5.347209598254863, + "grad_norm": 0.14139190316200256, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0212, + "step": 58830 + }, + { + "epoch": 5.348118523904745, + "grad_norm": 0.1173475831747055, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0235, + "step": 58840 + }, + { + "epoch": 5.349027449554627, + "grad_norm": 0.1269337236881256, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.023, + "step": 58850 + }, + { + "epoch": 5.349936375204508, + "grad_norm": 0.1634581983089447, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0207, + "step": 58860 + }, + { + "epoch": 5.35084530085439, + "grad_norm": 0.1589055061340332, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0196, + "step": 58870 + }, + { + "epoch": 5.351754226504272, + "grad_norm": 0.14143170416355133, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0228, + "step": 58880 + }, + { + "epoch": 5.352663152154154, + "grad_norm": 0.17290791869163513, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0265, + "step": 58890 + }, + { + "epoch": 5.353572077804036, + "grad_norm": 0.15018856525421143, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0227, + "step": 58900 + }, + { + "epoch": 5.354481003453918, + "grad_norm": 0.16527631878852844, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0231, + "step": 58910 + }, + { + "epoch": 5.355389929103799, + "grad_norm": 0.12711060047149658, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0225, + "step": 58920 + }, + { + "epoch": 5.356298854753681, + "grad_norm": 0.1317061334848404, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0226, + "step": 58930 + }, + { + "epoch": 5.357207780403563, + "grad_norm": 0.1491503268480301, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0217, + "step": 58940 + }, + { + "epoch": 5.358116706053445, + "grad_norm": 0.15086808800697327, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0231, + "step": 58950 + }, + { + "epoch": 5.359025631703327, + "grad_norm": 0.13663461804389954, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0218, + "step": 58960 + }, + { + "epoch": 5.359934557353209, + "grad_norm": 0.14250332117080688, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0206, + "step": 58970 + }, + { + "epoch": 5.36084348300309, + "grad_norm": 0.2505422532558441, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0208, + "step": 58980 + }, + { + "epoch": 5.361752408652972, + "grad_norm": 0.1364266276359558, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0222, + "step": 58990 + }, + { + "epoch": 5.362661334302854, + "grad_norm": 0.16143296658992767, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0211, + "step": 59000 + }, + { + "epoch": 5.363570259952736, + "grad_norm": 0.1212884858250618, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0222, + "step": 59010 + }, + { + "epoch": 5.364479185602618, + "grad_norm": 0.10913871228694916, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0209, + "step": 59020 + }, + { + "epoch": 5.3653881112525, + "grad_norm": 0.1334683746099472, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0215, + "step": 59030 + }, + { + "epoch": 5.366297036902381, + "grad_norm": 0.13704738020896912, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0242, + "step": 59040 + }, + { + "epoch": 5.367205962552263, + "grad_norm": 0.125392347574234, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0229, + "step": 59050 + }, + { + "epoch": 5.368114888202145, + "grad_norm": 0.11253941804170609, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0233, + "step": 59060 + }, + { + "epoch": 5.369023813852027, + "grad_norm": 0.12533500790596008, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0248, + "step": 59070 + }, + { + "epoch": 5.369932739501909, + "grad_norm": 0.12109420448541641, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0253, + "step": 59080 + }, + { + "epoch": 5.370841665151791, + "grad_norm": 0.11720914393663406, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0216, + "step": 59090 + }, + { + "epoch": 5.371750590801672, + "grad_norm": 0.12336961925029755, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.021, + "step": 59100 + }, + { + "epoch": 5.372659516451554, + "grad_norm": 0.117982879281044, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0203, + "step": 59110 + }, + { + "epoch": 5.373568442101436, + "grad_norm": 0.13815629482269287, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0217, + "step": 59120 + }, + { + "epoch": 5.374477367751318, + "grad_norm": 0.12920044362545013, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0218, + "step": 59130 + }, + { + "epoch": 5.3753862934012, + "grad_norm": 0.15780216455459595, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0219, + "step": 59140 + }, + { + "epoch": 5.376295219051082, + "grad_norm": 0.1992519646883011, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0221, + "step": 59150 + }, + { + "epoch": 5.377204144700963, + "grad_norm": 0.11334620416164398, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0239, + "step": 59160 + }, + { + "epoch": 5.378113070350845, + "grad_norm": 0.11106134206056595, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0262, + "step": 59170 + }, + { + "epoch": 5.379021996000727, + "grad_norm": 0.11803869903087616, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0222, + "step": 59180 + }, + { + "epoch": 5.379930921650609, + "grad_norm": 0.12132298946380615, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0212, + "step": 59190 + }, + { + "epoch": 5.380839847300491, + "grad_norm": 0.1629326194524765, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0277, + "step": 59200 + }, + { + "epoch": 5.381748772950373, + "grad_norm": 0.14594267308712006, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0228, + "step": 59210 + }, + { + "epoch": 5.382657698600254, + "grad_norm": 0.11754072457551956, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.021, + "step": 59220 + }, + { + "epoch": 5.383566624250136, + "grad_norm": 0.11448181420564651, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0222, + "step": 59230 + }, + { + "epoch": 5.384475549900018, + "grad_norm": 0.13215360045433044, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0204, + "step": 59240 + }, + { + "epoch": 5.3853844755499, + "grad_norm": 0.10473521798849106, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0213, + "step": 59250 + }, + { + "epoch": 5.386293401199782, + "grad_norm": 0.11523794382810593, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0217, + "step": 59260 + }, + { + "epoch": 5.387202326849664, + "grad_norm": 0.10844828188419342, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0222, + "step": 59270 + }, + { + "epoch": 5.388111252499545, + "grad_norm": 0.12870357930660248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0202, + "step": 59280 + }, + { + "epoch": 5.389020178149427, + "grad_norm": 0.14345362782478333, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0253, + "step": 59290 + }, + { + "epoch": 5.389929103799309, + "grad_norm": 0.10753655433654785, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0207, + "step": 59300 + }, + { + "epoch": 5.390838029449191, + "grad_norm": 0.12245530635118484, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0225, + "step": 59310 + }, + { + "epoch": 5.391746955099073, + "grad_norm": 0.14013776183128357, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0225, + "step": 59320 + }, + { + "epoch": 5.392655880748955, + "grad_norm": 0.16984960436820984, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0227, + "step": 59330 + }, + { + "epoch": 5.393564806398836, + "grad_norm": 0.13285958766937256, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0213, + "step": 59340 + }, + { + "epoch": 5.394473732048718, + "grad_norm": 0.11383504420518875, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0231, + "step": 59350 + }, + { + "epoch": 5.3953826576986, + "grad_norm": 0.1800517439842224, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0232, + "step": 59360 + }, + { + "epoch": 5.396291583348482, + "grad_norm": 0.12804758548736572, + "learning_rate": 1.64984337915455e-05, + "loss": 0.022, + "step": 59370 + }, + { + "epoch": 5.397200508998364, + "grad_norm": 0.1213470920920372, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0206, + "step": 59380 + }, + { + "epoch": 5.398109434648246, + "grad_norm": 0.16079141199588776, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0206, + "step": 59390 + }, + { + "epoch": 5.399018360298127, + "grad_norm": 0.09861018508672714, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0212, + "step": 59400 + }, + { + "epoch": 5.399927285948009, + "grad_norm": 0.11961232870817184, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0253, + "step": 59410 + }, + { + "epoch": 5.400836211597891, + "grad_norm": 0.11466231197118759, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0236, + "step": 59420 + }, + { + "epoch": 5.401745137247773, + "grad_norm": 0.0923413559794426, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0196, + "step": 59430 + }, + { + "epoch": 5.402654062897655, + "grad_norm": 0.12438227981328964, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0219, + "step": 59440 + }, + { + "epoch": 5.403562988547537, + "grad_norm": 0.13833226263523102, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.022, + "step": 59450 + }, + { + "epoch": 5.404471914197418, + "grad_norm": 0.08913487941026688, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0203, + "step": 59460 + }, + { + "epoch": 5.4053808398473, + "grad_norm": 0.1500292420387268, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0236, + "step": 59470 + }, + { + "epoch": 5.406289765497182, + "grad_norm": 0.12841403484344482, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0231, + "step": 59480 + }, + { + "epoch": 5.407198691147064, + "grad_norm": 0.18428634107112885, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0228, + "step": 59490 + }, + { + "epoch": 5.408107616796946, + "grad_norm": 0.13195529580116272, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0224, + "step": 59500 + }, + { + "epoch": 5.409016542446828, + "grad_norm": 0.1316898912191391, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0204, + "step": 59510 + }, + { + "epoch": 5.409925468096709, + "grad_norm": 0.17609994113445282, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0238, + "step": 59520 + }, + { + "epoch": 5.410834393746591, + "grad_norm": 0.15698014199733734, + "learning_rate": 1.643261147735204e-05, + "loss": 0.026, + "step": 59530 + }, + { + "epoch": 5.411743319396473, + "grad_norm": 0.15855275094509125, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0224, + "step": 59540 + }, + { + "epoch": 5.412652245046355, + "grad_norm": 0.14540642499923706, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0231, + "step": 59550 + }, + { + "epoch": 5.413561170696237, + "grad_norm": 0.17948301136493683, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0243, + "step": 59560 + }, + { + "epoch": 5.414470096346119, + "grad_norm": 0.1373734176158905, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0254, + "step": 59570 + }, + { + "epoch": 5.415379021996001, + "grad_norm": 0.10774601995944977, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0211, + "step": 59580 + }, + { + "epoch": 5.416287947645882, + "grad_norm": 0.13267992436885834, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.024, + "step": 59590 + }, + { + "epoch": 5.417196873295764, + "grad_norm": 0.13554780185222626, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0231, + "step": 59600 + }, + { + "epoch": 5.418105798945646, + "grad_norm": 0.11787440627813339, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.024, + "step": 59610 + }, + { + "epoch": 5.419014724595528, + "grad_norm": 0.12507905066013336, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0235, + "step": 59620 + }, + { + "epoch": 5.41992365024541, + "grad_norm": 0.11820895969867706, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0253, + "step": 59630 + }, + { + "epoch": 5.420832575895291, + "grad_norm": 0.11944480985403061, + "learning_rate": 1.638734240261125e-05, + "loss": 0.022, + "step": 59640 + }, + { + "epoch": 5.421741501545173, + "grad_norm": 0.10303112864494324, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0222, + "step": 59650 + }, + { + "epoch": 5.422650427195055, + "grad_norm": 0.12916937470436096, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0215, + "step": 59660 + }, + { + "epoch": 5.423559352844937, + "grad_norm": 0.14244721829891205, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.022, + "step": 59670 + }, + { + "epoch": 5.424468278494819, + "grad_norm": 0.12955956161022186, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.0225, + "step": 59680 + }, + { + "epoch": 5.425377204144701, + "grad_norm": 0.1176242008805275, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.021, + "step": 59690 + }, + { + "epoch": 5.426286129794583, + "grad_norm": 0.1487249881029129, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0214, + "step": 59700 + }, + { + "epoch": 5.427195055444464, + "grad_norm": 0.13332930207252502, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.022, + "step": 59710 + }, + { + "epoch": 5.428103981094346, + "grad_norm": 0.13529184460639954, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0207, + "step": 59720 + }, + { + "epoch": 5.429012906744228, + "grad_norm": 0.13570883870124817, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0222, + "step": 59730 + }, + { + "epoch": 5.42992183239411, + "grad_norm": 0.15231868624687195, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0216, + "step": 59740 + }, + { + "epoch": 5.430830758043992, + "grad_norm": 0.13415047526359558, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0208, + "step": 59750 + }, + { + "epoch": 5.431739683693873, + "grad_norm": 0.13275669515132904, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0241, + "step": 59760 + }, + { + "epoch": 5.432648609343755, + "grad_norm": 0.11362104117870331, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0248, + "step": 59770 + }, + { + "epoch": 5.433557534993637, + "grad_norm": 0.11826013028621674, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0234, + "step": 59780 + }, + { + "epoch": 5.434466460643519, + "grad_norm": 0.154282346367836, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0213, + "step": 59790 + }, + { + "epoch": 5.435375386293401, + "grad_norm": 0.13126695156097412, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0212, + "step": 59800 + }, + { + "epoch": 5.436284311943283, + "grad_norm": 0.16148504614830017, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0224, + "step": 59810 + }, + { + "epoch": 5.437193237593165, + "grad_norm": 0.11842388659715652, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0214, + "step": 59820 + }, + { + "epoch": 5.438102163243046, + "grad_norm": 0.1059088259935379, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0241, + "step": 59830 + }, + { + "epoch": 5.439011088892928, + "grad_norm": 0.13336306810379028, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.023, + "step": 59840 + }, + { + "epoch": 5.43992001454281, + "grad_norm": 0.11639715731143951, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0207, + "step": 59850 + }, + { + "epoch": 5.440828940192692, + "grad_norm": 0.1089014858007431, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.024, + "step": 59860 + }, + { + "epoch": 5.441737865842574, + "grad_norm": 0.137302428483963, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0227, + "step": 59870 + }, + { + "epoch": 5.4426467914924554, + "grad_norm": 0.12189992517232895, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0218, + "step": 59880 + }, + { + "epoch": 5.443555717142337, + "grad_norm": 0.10532815754413605, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0228, + "step": 59890 + }, + { + "epoch": 5.444464642792219, + "grad_norm": 0.14612829685211182, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0219, + "step": 59900 + }, + { + "epoch": 5.445373568442101, + "grad_norm": 0.15644952654838562, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.023, + "step": 59910 + }, + { + "epoch": 5.446282494091983, + "grad_norm": 0.10982740670442581, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0244, + "step": 59920 + }, + { + "epoch": 5.447191419741865, + "grad_norm": 0.14261402189731598, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0226, + "step": 59930 + }, + { + "epoch": 5.448100345391747, + "grad_norm": 0.1435793936252594, + "learning_rate": 1.626381771136563e-05, + "loss": 0.021, + "step": 59940 + }, + { + "epoch": 5.4490092710416285, + "grad_norm": 0.17523527145385742, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0225, + "step": 59950 + }, + { + "epoch": 5.4499181966915105, + "grad_norm": 0.1000848039984703, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0211, + "step": 59960 + }, + { + "epoch": 5.450827122341392, + "grad_norm": 0.17113669216632843, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0258, + "step": 59970 + }, + { + "epoch": 5.451736047991274, + "grad_norm": 0.1552135944366455, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0215, + "step": 59980 + }, + { + "epoch": 5.452644973641156, + "grad_norm": 0.24297897517681122, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.023, + "step": 59990 + }, + { + "epoch": 5.453553899291038, + "grad_norm": 0.13112226128578186, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0222, + "step": 60000 + }, + { + "epoch": 5.4544628249409195, + "grad_norm": 0.12999454140663147, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0243, + "step": 60010 + }, + { + "epoch": 5.4553717505908015, + "grad_norm": 0.11432912945747375, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0229, + "step": 60020 + }, + { + "epoch": 5.4562806762406835, + "grad_norm": 0.16527113318443298, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.0244, + "step": 60030 + }, + { + "epoch": 5.4571896018905655, + "grad_norm": 0.14243416488170624, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0224, + "step": 60040 + }, + { + "epoch": 5.458098527540447, + "grad_norm": 0.13010017573833466, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0206, + "step": 60050 + }, + { + "epoch": 5.459007453190329, + "grad_norm": 0.17769905924797058, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.0235, + "step": 60060 + }, + { + "epoch": 5.4599163788402105, + "grad_norm": 0.11141417175531387, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0266, + "step": 60070 + }, + { + "epoch": 5.4608253044900925, + "grad_norm": 0.12387698143720627, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.0226, + "step": 60080 + }, + { + "epoch": 5.4617342301399745, + "grad_norm": 0.12868265807628632, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0214, + "step": 60090 + }, + { + "epoch": 5.4626431557898565, + "grad_norm": 0.13497279584407806, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.021, + "step": 60100 + }, + { + "epoch": 5.4635520814397385, + "grad_norm": 0.1852896809577942, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0236, + "step": 60110 + }, + { + "epoch": 5.4644610070896205, + "grad_norm": 0.1724509745836258, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.021, + "step": 60120 + }, + { + "epoch": 5.465369932739502, + "grad_norm": 0.14020735025405884, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0205, + "step": 60130 + }, + { + "epoch": 5.4662788583893835, + "grad_norm": 0.13195893168449402, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0233, + "step": 60140 + }, + { + "epoch": 5.4671877840392655, + "grad_norm": 0.12138587236404419, + "learning_rate": 1.617729852865125e-05, + "loss": 0.0223, + "step": 60150 + }, + { + "epoch": 5.4680967096891475, + "grad_norm": 0.1414712816476822, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0208, + "step": 60160 + }, + { + "epoch": 5.4690056353390295, + "grad_norm": 0.10857667028903961, + "learning_rate": 1.616905650452066e-05, + "loss": 0.0207, + "step": 60170 + }, + { + "epoch": 5.4699145609889115, + "grad_norm": 0.11504074931144714, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0227, + "step": 60180 + }, + { + "epoch": 5.470823486638793, + "grad_norm": 0.19089938700199127, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0195, + "step": 60190 + }, + { + "epoch": 5.471732412288675, + "grad_norm": 0.22515586018562317, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0205, + "step": 60200 + }, + { + "epoch": 5.472641337938557, + "grad_norm": 0.11363942176103592, + "learning_rate": 1.61525713933778e-05, + "loss": 0.0235, + "step": 60210 + }, + { + "epoch": 5.4735502635884385, + "grad_norm": 0.13733012974262238, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0241, + "step": 60220 + }, + { + "epoch": 5.4744591892383205, + "grad_norm": 0.15266728401184082, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0202, + "step": 60230 + }, + { + "epoch": 5.4753681148882025, + "grad_norm": 0.20376789569854736, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0257, + "step": 60240 + }, + { + "epoch": 5.476277040538084, + "grad_norm": 0.15351834893226624, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0241, + "step": 60250 + }, + { + "epoch": 5.477185966187966, + "grad_norm": 0.11839159578084946, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0227, + "step": 60260 + }, + { + "epoch": 5.478094891837848, + "grad_norm": 0.13525480031967163, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0217, + "step": 60270 + }, + { + "epoch": 5.47900381748773, + "grad_norm": 0.1418711096048355, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.02, + "step": 60280 + }, + { + "epoch": 5.479912743137612, + "grad_norm": 0.15879319608211517, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0211, + "step": 60290 + }, + { + "epoch": 5.4808216687874936, + "grad_norm": 0.15339554846286774, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.0231, + "step": 60300 + }, + { + "epoch": 5.481730594437375, + "grad_norm": 0.10218580812215805, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0223, + "step": 60310 + }, + { + "epoch": 5.482639520087257, + "grad_norm": 0.12794508039951324, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.0212, + "step": 60320 + }, + { + "epoch": 5.483548445737139, + "grad_norm": 0.1011282280087471, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0224, + "step": 60330 + }, + { + "epoch": 5.484457371387021, + "grad_norm": 0.10190997272729874, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.0212, + "step": 60340 + }, + { + "epoch": 5.485366297036903, + "grad_norm": 0.09043766558170319, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0214, + "step": 60350 + }, + { + "epoch": 5.486275222686785, + "grad_norm": 0.14976610243320465, + "learning_rate": 1.609073991709496e-05, + "loss": 0.0249, + "step": 60360 + }, + { + "epoch": 5.487184148336666, + "grad_norm": 0.13984698057174683, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0232, + "step": 60370 + }, + { + "epoch": 5.488093073986548, + "grad_norm": 0.11210978776216507, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0223, + "step": 60380 + }, + { + "epoch": 5.48900199963643, + "grad_norm": 0.125299870967865, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0224, + "step": 60390 + }, + { + "epoch": 5.489910925286312, + "grad_norm": 0.16451174020767212, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0226, + "step": 60400 + }, + { + "epoch": 5.490819850936194, + "grad_norm": 0.14239828288555145, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0223, + "step": 60410 + }, + { + "epoch": 5.491728776586076, + "grad_norm": 0.14591294527053833, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0226, + "step": 60420 + }, + { + "epoch": 5.492637702235957, + "grad_norm": 0.11605766415596008, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0254, + "step": 60430 + }, + { + "epoch": 5.493546627885839, + "grad_norm": 0.12974192202091217, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0201, + "step": 60440 + }, + { + "epoch": 5.494455553535721, + "grad_norm": 0.12425573915243149, + "learning_rate": 1.605363201758077e-05, + "loss": 0.0219, + "step": 60450 + }, + { + "epoch": 5.495364479185603, + "grad_norm": 0.10735001415014267, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.021, + "step": 60460 + }, + { + "epoch": 5.496273404835485, + "grad_norm": 0.14535404741764069, + "learning_rate": 1.60453849294712e-05, + "loss": 0.0253, + "step": 60470 + }, + { + "epoch": 5.497182330485367, + "grad_norm": 0.12177706509828568, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0192, + "step": 60480 + }, + { + "epoch": 5.498091256135248, + "grad_norm": 0.10458561033010483, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0228, + "step": 60490 + }, + { + "epoch": 5.49900018178513, + "grad_norm": 0.12133309245109558, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0211, + "step": 60500 + }, + { + "epoch": 5.499909107435012, + "grad_norm": 0.15785706043243408, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.0247, + "step": 60510 + }, + { + "epoch": 5.500818033084894, + "grad_norm": 0.17735907435417175, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0222, + "step": 60520 + }, + { + "epoch": 5.501726958734776, + "grad_norm": 0.1363718956708908, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0214, + "step": 60530 + }, + { + "epoch": 5.502635884384658, + "grad_norm": 0.11407589912414551, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.0215, + "step": 60540 + }, + { + "epoch": 5.503544810034539, + "grad_norm": 0.12615904211997986, + "learning_rate": 1.601239342648345e-05, + "loss": 0.019, + "step": 60550 + }, + { + "epoch": 5.504453735684421, + "grad_norm": 0.11098602414131165, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.024, + "step": 60560 + }, + { + "epoch": 5.505362661334303, + "grad_norm": 0.12378008663654327, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0245, + "step": 60570 + }, + { + "epoch": 5.506271586984185, + "grad_norm": 0.12369867414236069, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0224, + "step": 60580 + }, + { + "epoch": 5.507180512634067, + "grad_norm": 0.13011902570724487, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0228, + "step": 60590 + }, + { + "epoch": 5.508089438283949, + "grad_norm": 0.11863945424556732, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.0214, + "step": 60600 + }, + { + "epoch": 5.508998363933831, + "grad_norm": 0.14192727208137512, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0207, + "step": 60610 + }, + { + "epoch": 5.509907289583712, + "grad_norm": 0.1271606981754303, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0253, + "step": 60620 + }, + { + "epoch": 5.510816215233594, + "grad_norm": 0.1269829422235489, + "learning_rate": 1.597939700288329e-05, + "loss": 0.022, + "step": 60630 + }, + { + "epoch": 5.511725140883476, + "grad_norm": 0.1577921211719513, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0223, + "step": 60640 + }, + { + "epoch": 5.512634066533358, + "grad_norm": 0.17512670159339905, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0257, + "step": 60650 + }, + { + "epoch": 5.51354299218324, + "grad_norm": 0.1202736422419548, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0186, + "step": 60660 + }, + { + "epoch": 5.514451917833121, + "grad_norm": 0.12869566679000854, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0233, + "step": 60670 + }, + { + "epoch": 5.515360843483003, + "grad_norm": 0.125078946352005, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0204, + "step": 60680 + }, + { + "epoch": 5.516269769132885, + "grad_norm": 0.11901658028364182, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0213, + "step": 60690 + }, + { + "epoch": 5.517178694782767, + "grad_norm": 0.1254761815071106, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0222, + "step": 60700 + }, + { + "epoch": 5.518087620432649, + "grad_norm": 0.1552484780550003, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.022, + "step": 60710 + }, + { + "epoch": 5.518996546082531, + "grad_norm": 0.12050727009773254, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.0225, + "step": 60720 + }, + { + "epoch": 5.519905471732413, + "grad_norm": 0.15650801360607147, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0217, + "step": 60730 + }, + { + "epoch": 5.520814397382294, + "grad_norm": 0.12868818640708923, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0233, + "step": 60740 + }, + { + "epoch": 5.521723323032176, + "grad_norm": 0.14403335750102997, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0245, + "step": 60750 + }, + { + "epoch": 5.522632248682058, + "grad_norm": 0.18283621966838837, + "learning_rate": 1.592576773310894e-05, + "loss": 0.0233, + "step": 60760 + }, + { + "epoch": 5.52354117433194, + "grad_norm": 0.16688795387744904, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0202, + "step": 60770 + }, + { + "epoch": 5.524450099981822, + "grad_norm": 0.12525120377540588, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0249, + "step": 60780 + }, + { + "epoch": 5.525359025631703, + "grad_norm": 0.14895208179950714, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0251, + "step": 60790 + }, + { + "epoch": 5.526267951281585, + "grad_norm": 0.15096963942050934, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0214, + "step": 60800 + }, + { + "epoch": 5.527176876931467, + "grad_norm": 0.12673532962799072, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0218, + "step": 60810 + }, + { + "epoch": 5.528085802581349, + "grad_norm": 0.16417910158634186, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0247, + "step": 60820 + }, + { + "epoch": 5.528994728231231, + "grad_norm": 0.17245228588581085, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.0256, + "step": 60830 + }, + { + "epoch": 5.529903653881113, + "grad_norm": 0.11559180915355682, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0217, + "step": 60840 + }, + { + "epoch": 5.530812579530995, + "grad_norm": 0.11849625408649445, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0219, + "step": 60850 + }, + { + "epoch": 5.531721505180876, + "grad_norm": 0.140710711479187, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0214, + "step": 60860 + }, + { + "epoch": 5.532630430830758, + "grad_norm": 0.12349947541952133, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0193, + "step": 60870 + }, + { + "epoch": 5.53353935648064, + "grad_norm": 0.11803016066551208, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0241, + "step": 60880 + }, + { + "epoch": 5.534448282130522, + "grad_norm": 0.15463905036449432, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0221, + "step": 60890 + }, + { + "epoch": 5.535357207780404, + "grad_norm": 0.12798573076725006, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0217, + "step": 60900 + }, + { + "epoch": 5.536266133430285, + "grad_norm": 0.2103399783372879, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0232, + "step": 60910 + }, + { + "epoch": 5.537175059080167, + "grad_norm": 0.1856740415096283, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0225, + "step": 60920 + }, + { + "epoch": 5.538083984730049, + "grad_norm": 0.13202139735221863, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0212, + "step": 60930 + }, + { + "epoch": 5.538992910379931, + "grad_norm": 0.13626644015312195, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0217, + "step": 60940 + }, + { + "epoch": 5.539901836029813, + "grad_norm": 0.13620105385780334, + "learning_rate": 1.584736531020125e-05, + "loss": 0.0244, + "step": 60950 + }, + { + "epoch": 5.540810761679695, + "grad_norm": 0.15215308964252472, + "learning_rate": 1.584323820456826e-05, + "loss": 0.0223, + "step": 60960 + }, + { + "epoch": 5.541719687329577, + "grad_norm": 0.1787509173154831, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0241, + "step": 60970 + }, + { + "epoch": 5.542628612979458, + "grad_norm": 0.11176944524049759, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.0207, + "step": 60980 + }, + { + "epoch": 5.54353753862934, + "grad_norm": 0.14973673224449158, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0212, + "step": 60990 + }, + { + "epoch": 5.544446464279222, + "grad_norm": 0.11471419036388397, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0215, + "step": 61000 + }, + { + "epoch": 5.545355389929104, + "grad_norm": 0.12783099710941315, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0238, + "step": 61010 + }, + { + "epoch": 5.546264315578986, + "grad_norm": 0.12502047419548035, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0247, + "step": 61020 + }, + { + "epoch": 5.547173241228867, + "grad_norm": 0.14410874247550964, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0233, + "step": 61030 + }, + { + "epoch": 5.548082166878749, + "grad_norm": 0.12375950813293457, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0235, + "step": 61040 + }, + { + "epoch": 5.548991092528631, + "grad_norm": 0.1213027760386467, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0226, + "step": 61050 + }, + { + "epoch": 5.549900018178513, + "grad_norm": 0.14608293771743774, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0216, + "step": 61060 + }, + { + "epoch": 5.550808943828395, + "grad_norm": 0.09387248754501343, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0203, + "step": 61070 + }, + { + "epoch": 5.551717869478277, + "grad_norm": 0.13580554723739624, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0235, + "step": 61080 + }, + { + "epoch": 5.552626795128159, + "grad_norm": 0.14263974130153656, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0227, + "step": 61090 + }, + { + "epoch": 5.55353572077804, + "grad_norm": 0.14622057974338531, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.021, + "step": 61100 + }, + { + "epoch": 5.554444646427922, + "grad_norm": 0.163898766040802, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0187, + "step": 61110 + }, + { + "epoch": 5.555353572077804, + "grad_norm": 0.10329697281122208, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0228, + "step": 61120 + }, + { + "epoch": 5.556262497727686, + "grad_norm": 0.11197671294212341, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0236, + "step": 61130 + }, + { + "epoch": 5.557171423377568, + "grad_norm": 0.19062696397304535, + "learning_rate": 1.576893965634554e-05, + "loss": 0.023, + "step": 61140 + }, + { + "epoch": 5.558080349027449, + "grad_norm": 0.13889528810977936, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.0214, + "step": 61150 + }, + { + "epoch": 5.558989274677331, + "grad_norm": 0.17355138063430786, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0233, + "step": 61160 + }, + { + "epoch": 5.559898200327213, + "grad_norm": 0.12203329056501389, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0201, + "step": 61170 + }, + { + "epoch": 5.560807125977095, + "grad_norm": 0.1549149602651596, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0232, + "step": 61180 + }, + { + "epoch": 5.561716051626977, + "grad_norm": 0.1204608604311943, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0214, + "step": 61190 + }, + { + "epoch": 5.562624977276859, + "grad_norm": 0.12121281772851944, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.0225, + "step": 61200 + }, + { + "epoch": 5.563533902926741, + "grad_norm": 0.1346883922815323, + "learning_rate": 1.574004056626338e-05, + "loss": 0.025, + "step": 61210 + }, + { + "epoch": 5.564442828576622, + "grad_norm": 0.12003662437200546, + "learning_rate": 1.573591189751e-05, + "loss": 0.021, + "step": 61220 + }, + { + "epoch": 5.565351754226504, + "grad_norm": 0.10309959203004837, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0237, + "step": 61230 + }, + { + "epoch": 5.566260679876386, + "grad_norm": 0.12907704710960388, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0201, + "step": 61240 + }, + { + "epoch": 5.567169605526268, + "grad_norm": 0.16952712833881378, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0221, + "step": 61250 + }, + { + "epoch": 5.56807853117615, + "grad_norm": 0.13621579110622406, + "learning_rate": 1.571939666675547e-05, + "loss": 0.025, + "step": 61260 + }, + { + "epoch": 5.568987456826031, + "grad_norm": 0.12940339744091034, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0231, + "step": 61270 + }, + { + "epoch": 5.569896382475913, + "grad_norm": 0.12737077474594116, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0207, + "step": 61280 + }, + { + "epoch": 5.570805308125795, + "grad_norm": 0.1585296392440796, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0249, + "step": 61290 + }, + { + "epoch": 5.571714233775677, + "grad_norm": 0.12818138301372528, + "learning_rate": 1.570288056186616e-05, + "loss": 0.025, + "step": 61300 + }, + { + "epoch": 5.572623159425559, + "grad_norm": 0.13042202591896057, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0198, + "step": 61310 + }, + { + "epoch": 5.573532085075441, + "grad_norm": 0.10053355991840363, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0208, + "step": 61320 + }, + { + "epoch": 5.574441010725323, + "grad_norm": 0.11785581707954407, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0222, + "step": 61330 + }, + { + "epoch": 5.575349936375204, + "grad_norm": 0.12248669564723969, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0234, + "step": 61340 + }, + { + "epoch": 5.576258862025086, + "grad_norm": 0.1171405017375946, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0234, + "step": 61350 + }, + { + "epoch": 5.577167787674968, + "grad_norm": 0.13539552688598633, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.0197, + "step": 61360 + }, + { + "epoch": 5.57807671332485, + "grad_norm": 0.1216900497674942, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.0214, + "step": 61370 + }, + { + "epoch": 5.578985638974732, + "grad_norm": 0.1180599182844162, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.0209, + "step": 61380 + }, + { + "epoch": 5.579894564624614, + "grad_norm": 0.1038932204246521, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0201, + "step": 61390 + }, + { + "epoch": 5.580803490274495, + "grad_norm": 0.12125887721776962, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0249, + "step": 61400 + }, + { + "epoch": 5.581712415924377, + "grad_norm": 0.12257237732410431, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0204, + "step": 61410 + }, + { + "epoch": 5.582621341574259, + "grad_norm": 0.11557234823703766, + "learning_rate": 1.565332720308093e-05, + "loss": 0.0209, + "step": 61420 + }, + { + "epoch": 5.583530267224141, + "grad_norm": 0.13548816740512848, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0209, + "step": 61430 + }, + { + "epoch": 5.584439192874023, + "grad_norm": 0.12581110000610352, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0218, + "step": 61440 + }, + { + "epoch": 5.585348118523905, + "grad_norm": 0.1141590103507042, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0214, + "step": 61450 + }, + { + "epoch": 5.586257044173786, + "grad_norm": 0.13780000805854797, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0227, + "step": 61460 + }, + { + "epoch": 5.587165969823668, + "grad_norm": 0.1699170172214508, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0232, + "step": 61470 + }, + { + "epoch": 5.58807489547355, + "grad_norm": 0.1312125027179718, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0225, + "step": 61480 + }, + { + "epoch": 5.588983821123432, + "grad_norm": 0.10460589081048965, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0225, + "step": 61490 + }, + { + "epoch": 5.589892746773314, + "grad_norm": 0.10274886339902878, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0214, + "step": 61500 + }, + { + "epoch": 5.590801672423196, + "grad_norm": 0.13499124348163605, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0221, + "step": 61510 + }, + { + "epoch": 5.591710598073077, + "grad_norm": 0.13355886936187744, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0208, + "step": 61520 + }, + { + "epoch": 5.592619523722959, + "grad_norm": 0.18196064233779907, + "learning_rate": 1.560789700123175e-05, + "loss": 0.0233, + "step": 61530 + }, + { + "epoch": 5.593528449372841, + "grad_norm": 0.1063167080283165, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0211, + "step": 61540 + }, + { + "epoch": 5.594437375022723, + "grad_norm": 0.113223135471344, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0219, + "step": 61550 + }, + { + "epoch": 5.595346300672605, + "grad_norm": 0.1229001134634018, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0207, + "step": 61560 + }, + { + "epoch": 5.596255226322487, + "grad_norm": 0.12838929891586304, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0203, + "step": 61570 + }, + { + "epoch": 5.597164151972368, + "grad_norm": 0.11306323111057281, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.0232, + "step": 61580 + }, + { + "epoch": 5.59807307762225, + "grad_norm": 0.12362082302570343, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0217, + "step": 61590 + }, + { + "epoch": 5.598982003272132, + "grad_norm": 0.11961408704519272, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0203, + "step": 61600 + }, + { + "epoch": 5.599890928922014, + "grad_norm": 0.1372358500957489, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.0213, + "step": 61610 + }, + { + "epoch": 5.600799854571896, + "grad_norm": 0.15659523010253906, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0219, + "step": 61620 + }, + { + "epoch": 5.601708780221778, + "grad_norm": 0.1729724258184433, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0214, + "step": 61630 + }, + { + "epoch": 5.60261770587166, + "grad_norm": 0.10322710871696472, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.0211, + "step": 61640 + }, + { + "epoch": 5.603526631521541, + "grad_norm": 0.11671203374862671, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.022, + "step": 61650 + }, + { + "epoch": 5.604435557171423, + "grad_norm": 0.11812612414360046, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.0203, + "step": 61660 + }, + { + "epoch": 5.605344482821305, + "grad_norm": 0.1195341944694519, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0212, + "step": 61670 + }, + { + "epoch": 5.606253408471187, + "grad_norm": 0.1378045529127121, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0236, + "step": 61680 + }, + { + "epoch": 5.607162334121069, + "grad_norm": 0.12592288851737976, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0199, + "step": 61690 + }, + { + "epoch": 5.60807125977095, + "grad_norm": 0.14632733166217804, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0205, + "step": 61700 + }, + { + "epoch": 5.608980185420832, + "grad_norm": 0.12256485223770142, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.0202, + "step": 61710 + }, + { + "epoch": 5.609889111070714, + "grad_norm": 0.13091091811656952, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0217, + "step": 61720 + }, + { + "epoch": 5.610798036720596, + "grad_norm": 0.13185341656208038, + "learning_rate": 1.552528261458322e-05, + "loss": 0.0227, + "step": 61730 + }, + { + "epoch": 5.611706962370478, + "grad_norm": 0.1257806420326233, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0199, + "step": 61740 + }, + { + "epoch": 5.61261588802036, + "grad_norm": 0.10723815858364105, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0238, + "step": 61750 + }, + { + "epoch": 5.613524813670242, + "grad_norm": 0.15427975356578827, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0227, + "step": 61760 + }, + { + "epoch": 5.614433739320123, + "grad_norm": 0.12001065909862518, + "learning_rate": 1.550875774213821e-05, + "loss": 0.0226, + "step": 61770 + }, + { + "epoch": 5.615342664970005, + "grad_norm": 0.12660329043865204, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0211, + "step": 61780 + }, + { + "epoch": 5.616251590619887, + "grad_norm": 0.13262300193309784, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.022, + "step": 61790 + }, + { + "epoch": 5.617160516269769, + "grad_norm": 0.14083871245384216, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0217, + "step": 61800 + }, + { + "epoch": 5.618069441919651, + "grad_norm": 0.1111614927649498, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0238, + "step": 61810 + }, + { + "epoch": 5.618978367569532, + "grad_norm": 0.10933896154165268, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.0228, + "step": 61820 + }, + { + "epoch": 5.619887293219414, + "grad_norm": 0.22090622782707214, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0259, + "step": 61830 + }, + { + "epoch": 5.620796218869296, + "grad_norm": 0.1218695193529129, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.023, + "step": 61840 + }, + { + "epoch": 5.621705144519178, + "grad_norm": 0.31269705295562744, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0231, + "step": 61850 + }, + { + "epoch": 5.62261407016906, + "grad_norm": 0.1214980036020279, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0212, + "step": 61860 + }, + { + "epoch": 5.623522995818942, + "grad_norm": 0.17026852071285248, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0226, + "step": 61870 + }, + { + "epoch": 5.624431921468824, + "grad_norm": 0.12943395972251892, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.021, + "step": 61880 + }, + { + "epoch": 5.625340847118705, + "grad_norm": 0.11600268632173538, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0201, + "step": 61890 + }, + { + "epoch": 5.626249772768587, + "grad_norm": 0.12162382155656815, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0214, + "step": 61900 + }, + { + "epoch": 5.627158698418469, + "grad_norm": 0.2105938196182251, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.0226, + "step": 61910 + }, + { + "epoch": 5.628067624068351, + "grad_norm": 0.13019892573356628, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.0198, + "step": 61920 + }, + { + "epoch": 5.628976549718233, + "grad_norm": 0.12002481520175934, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0213, + "step": 61930 + }, + { + "epoch": 5.6298854753681145, + "grad_norm": 0.1993207037448883, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0247, + "step": 61940 + }, + { + "epoch": 5.630794401017996, + "grad_norm": 0.15358640253543854, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0239, + "step": 61950 + }, + { + "epoch": 5.631703326667878, + "grad_norm": 0.2904185354709625, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0229, + "step": 61960 + }, + { + "epoch": 5.63261225231776, + "grad_norm": 0.12021315097808838, + "learning_rate": 1.542612450869506e-05, + "loss": 0.0209, + "step": 61970 + }, + { + "epoch": 5.633521177967642, + "grad_norm": 0.15095995366573334, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.0225, + "step": 61980 + }, + { + "epoch": 5.634430103617524, + "grad_norm": 0.12501651048660278, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0207, + "step": 61990 + }, + { + "epoch": 5.635339029267406, + "grad_norm": 0.13531078398227692, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0222, + "step": 62000 + }, + { + "epoch": 5.6362479549172875, + "grad_norm": 0.15524742007255554, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.021, + "step": 62010 + }, + { + "epoch": 5.6371568805671695, + "grad_norm": 0.09712246060371399, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0211, + "step": 62020 + }, + { + "epoch": 5.6380658062170514, + "grad_norm": 0.1460903137922287, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0227, + "step": 62030 + }, + { + "epoch": 5.638974731866933, + "grad_norm": 0.16429750621318817, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.021, + "step": 62040 + }, + { + "epoch": 5.639883657516815, + "grad_norm": 0.1237972229719162, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0217, + "step": 62050 + }, + { + "epoch": 5.6407925831666965, + "grad_norm": 0.12351322919130325, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0195, + "step": 62060 + }, + { + "epoch": 5.6417015088165785, + "grad_norm": 0.17693351209163666, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0228, + "step": 62070 + }, + { + "epoch": 5.6426104344664605, + "grad_norm": 0.1239989623427391, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0213, + "step": 62080 + }, + { + "epoch": 5.6435193601163425, + "grad_norm": 0.11825328320264816, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0202, + "step": 62090 + }, + { + "epoch": 5.6444282857662245, + "grad_norm": 0.2122933566570282, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.021, + "step": 62100 + }, + { + "epoch": 5.6453372114161064, + "grad_norm": 0.11755210161209106, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0202, + "step": 62110 + }, + { + "epoch": 5.646246137065988, + "grad_norm": 0.11829824000597, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.0208, + "step": 62120 + }, + { + "epoch": 5.6471550627158695, + "grad_norm": 0.13352367281913757, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0202, + "step": 62130 + }, + { + "epoch": 5.6480639883657515, + "grad_norm": 0.1220584362745285, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.021, + "step": 62140 + }, + { + "epoch": 5.6489729140156335, + "grad_norm": 0.13071666657924652, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0208, + "step": 62150 + }, + { + "epoch": 5.6498818396655155, + "grad_norm": 0.1173131912946701, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0217, + "step": 62160 + }, + { + "epoch": 5.6507907653153975, + "grad_norm": 0.12834985554218292, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0219, + "step": 62170 + }, + { + "epoch": 5.651699690965279, + "grad_norm": 0.10922470688819885, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0226, + "step": 62180 + }, + { + "epoch": 5.652608616615161, + "grad_norm": 0.16949132084846497, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0231, + "step": 62190 + }, + { + "epoch": 5.6535175422650426, + "grad_norm": 0.12742875516414642, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0228, + "step": 62200 + }, + { + "epoch": 5.6544264679149245, + "grad_norm": 0.1285697966814041, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0212, + "step": 62210 + }, + { + "epoch": 5.6553353935648065, + "grad_norm": 0.17238982021808624, + "learning_rate": 1.532281505788039e-05, + "loss": 0.0217, + "step": 62220 + }, + { + "epoch": 5.6562443192146885, + "grad_norm": 0.1261514127254486, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0216, + "step": 62230 + }, + { + "epoch": 5.6571532448645705, + "grad_norm": 0.11106337606906891, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0216, + "step": 62240 + }, + { + "epoch": 5.658062170514452, + "grad_norm": 0.15617838501930237, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0234, + "step": 62250 + }, + { + "epoch": 5.658971096164334, + "grad_norm": 0.1421806514263153, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0218, + "step": 62260 + }, + { + "epoch": 5.659880021814216, + "grad_norm": 0.16013966500759125, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0236, + "step": 62270 + }, + { + "epoch": 5.6607889474640976, + "grad_norm": 0.1332966536283493, + "learning_rate": 1.52980183248759e-05, + "loss": 0.0223, + "step": 62280 + }, + { + "epoch": 5.6616978731139795, + "grad_norm": 0.12243349850177765, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.0216, + "step": 62290 + }, + { + "epoch": 5.662606798763861, + "grad_norm": 0.1233859434723854, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0198, + "step": 62300 + }, + { + "epoch": 5.663515724413743, + "grad_norm": 0.1400901973247528, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0199, + "step": 62310 + }, + { + "epoch": 5.664424650063625, + "grad_norm": 0.127544105052948, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0209, + "step": 62320 + }, + { + "epoch": 5.665333575713507, + "grad_norm": 0.14599132537841797, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.0237, + "step": 62330 + }, + { + "epoch": 5.666242501363389, + "grad_norm": 0.15811268985271454, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0229, + "step": 62340 + }, + { + "epoch": 5.667151427013271, + "grad_norm": 0.1733967810869217, + "learning_rate": 1.526908777801797e-05, + "loss": 0.0226, + "step": 62350 + }, + { + "epoch": 5.668060352663153, + "grad_norm": 0.18460465967655182, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0227, + "step": 62360 + }, + { + "epoch": 5.668969278313034, + "grad_norm": 0.13349837064743042, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.0227, + "step": 62370 + }, + { + "epoch": 5.669878203962916, + "grad_norm": 0.1498951017856598, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0229, + "step": 62380 + }, + { + "epoch": 5.670787129612798, + "grad_norm": 0.17375046014785767, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0228, + "step": 62390 + }, + { + "epoch": 5.67169605526268, + "grad_norm": 0.12974432110786438, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0231, + "step": 62400 + }, + { + "epoch": 5.672604980912562, + "grad_norm": 0.1220737174153328, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0202, + "step": 62410 + }, + { + "epoch": 5.673513906562443, + "grad_norm": 0.09045973420143127, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0218, + "step": 62420 + }, + { + "epoch": 5.674422832212325, + "grad_norm": 0.12633644044399261, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0231, + "step": 62430 + }, + { + "epoch": 5.675331757862207, + "grad_norm": 0.14674486219882965, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0231, + "step": 62440 + }, + { + "epoch": 5.676240683512089, + "grad_norm": 0.11394625157117844, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0245, + "step": 62450 + }, + { + "epoch": 5.677149609161971, + "grad_norm": 0.13444028794765472, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.0215, + "step": 62460 + }, + { + "epoch": 5.678058534811853, + "grad_norm": 0.13024979829788208, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0204, + "step": 62470 + }, + { + "epoch": 5.678967460461735, + "grad_norm": 0.11887981742620468, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0196, + "step": 62480 + }, + { + "epoch": 5.679876386111616, + "grad_norm": 0.12195833772420883, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0229, + "step": 62490 + }, + { + "epoch": 5.680785311761498, + "grad_norm": 0.13829363882541656, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0215, + "step": 62500 + }, + { + "epoch": 5.68169423741138, + "grad_norm": 0.11276445537805557, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0207, + "step": 62510 + }, + { + "epoch": 5.682603163061262, + "grad_norm": 0.10502763092517853, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0212, + "step": 62520 + }, + { + "epoch": 5.683512088711144, + "grad_norm": 0.12301613390445709, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0195, + "step": 62530 + }, + { + "epoch": 5.684421014361026, + "grad_norm": 0.15128369629383087, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.0219, + "step": 62540 + }, + { + "epoch": 5.685329940010907, + "grad_norm": 0.11366292834281921, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0236, + "step": 62550 + }, + { + "epoch": 5.686238865660789, + "grad_norm": 0.13584202527999878, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0236, + "step": 62560 + }, + { + "epoch": 5.687147791310671, + "grad_norm": 0.11674688011407852, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0201, + "step": 62570 + }, + { + "epoch": 5.688056716960553, + "grad_norm": 0.13053034245967865, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.022, + "step": 62580 + }, + { + "epoch": 5.688965642610435, + "grad_norm": 0.11744866520166397, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0221, + "step": 62590 + }, + { + "epoch": 5.689874568260317, + "grad_norm": 0.10714341700077057, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0196, + "step": 62600 + }, + { + "epoch": 5.690783493910198, + "grad_norm": 0.12626175582408905, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0232, + "step": 62610 + }, + { + "epoch": 5.69169241956008, + "grad_norm": 0.16950684785842896, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0234, + "step": 62620 + }, + { + "epoch": 5.692601345209962, + "grad_norm": 0.1042758971452713, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.021, + "step": 62630 + }, + { + "epoch": 5.693510270859844, + "grad_norm": 0.13660304248332977, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0227, + "step": 62640 + }, + { + "epoch": 5.694419196509726, + "grad_norm": 0.14983384311199188, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0216, + "step": 62650 + }, + { + "epoch": 5.695328122159608, + "grad_norm": 0.1455707997083664, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.0206, + "step": 62660 + }, + { + "epoch": 5.696237047809489, + "grad_norm": 0.12906597554683685, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0232, + "step": 62670 + }, + { + "epoch": 5.697145973459371, + "grad_norm": 0.13721542060375214, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0199, + "step": 62680 + }, + { + "epoch": 5.698054899109253, + "grad_norm": 0.13303618133068085, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0207, + "step": 62690 + }, + { + "epoch": 5.698963824759135, + "grad_norm": 0.12046858668327332, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0233, + "step": 62700 + }, + { + "epoch": 5.699872750409017, + "grad_norm": 0.1405707150697708, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0243, + "step": 62710 + }, + { + "epoch": 5.700781676058899, + "grad_norm": 0.15058749914169312, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.0225, + "step": 62720 + }, + { + "epoch": 5.70169060170878, + "grad_norm": 0.09799517691135406, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0233, + "step": 62730 + }, + { + "epoch": 5.702599527358662, + "grad_norm": 0.11600363999605179, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0212, + "step": 62740 + }, + { + "epoch": 5.703508453008544, + "grad_norm": 0.1532885581254959, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0274, + "step": 62750 + }, + { + "epoch": 5.704417378658426, + "grad_norm": 0.1117679700255394, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0199, + "step": 62760 + }, + { + "epoch": 5.705326304308308, + "grad_norm": 0.1404341757297516, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0237, + "step": 62770 + }, + { + "epoch": 5.70623522995819, + "grad_norm": 0.17424467206001282, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0202, + "step": 62780 + }, + { + "epoch": 5.707144155608072, + "grad_norm": 0.12912039458751678, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.0246, + "step": 62790 + }, + { + "epoch": 5.708053081257953, + "grad_norm": 0.14501920342445374, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.0228, + "step": 62800 + }, + { + "epoch": 5.708962006907835, + "grad_norm": 0.10727924108505249, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0202, + "step": 62810 + }, + { + "epoch": 5.709870932557717, + "grad_norm": 0.12085720151662827, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0225, + "step": 62820 + }, + { + "epoch": 5.710779858207599, + "grad_norm": 0.1776050329208374, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0208, + "step": 62830 + }, + { + "epoch": 5.711688783857481, + "grad_norm": 0.13133293390274048, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.023, + "step": 62840 + }, + { + "epoch": 5.712597709507362, + "grad_norm": 0.09626714885234833, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0203, + "step": 62850 + }, + { + "epoch": 5.713506635157244, + "grad_norm": 0.1997305452823639, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0202, + "step": 62860 + }, + { + "epoch": 5.714415560807126, + "grad_norm": 0.1233481764793396, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.0206, + "step": 62870 + }, + { + "epoch": 5.715324486457008, + "grad_norm": 0.12447388470172882, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0228, + "step": 62880 + }, + { + "epoch": 5.71623341210689, + "grad_norm": 0.0897996798157692, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0196, + "step": 62890 + }, + { + "epoch": 5.717142337756772, + "grad_norm": 0.11249232292175293, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0208, + "step": 62900 + }, + { + "epoch": 5.718051263406654, + "grad_norm": 0.1611606627702713, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0207, + "step": 62910 + }, + { + "epoch": 5.718960189056535, + "grad_norm": 0.1470264047384262, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0258, + "step": 62920 + }, + { + "epoch": 5.719869114706417, + "grad_norm": 0.16259229183197021, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0199, + "step": 62930 + }, + { + "epoch": 5.720778040356299, + "grad_norm": 0.1107429563999176, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0211, + "step": 62940 + }, + { + "epoch": 5.721686966006181, + "grad_norm": 0.1477135717868805, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0235, + "step": 62950 + }, + { + "epoch": 5.722595891656063, + "grad_norm": 0.16947215795516968, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0219, + "step": 62960 + }, + { + "epoch": 5.723504817305944, + "grad_norm": 0.137398824095726, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0216, + "step": 62970 + }, + { + "epoch": 5.724413742955826, + "grad_norm": 0.14629578590393066, + "learning_rate": 1.500868071605827e-05, + "loss": 0.023, + "step": 62980 + }, + { + "epoch": 5.725322668605708, + "grad_norm": 0.1516667902469635, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0219, + "step": 62990 + }, + { + "epoch": 5.72623159425559, + "grad_norm": 0.13369229435920715, + "learning_rate": 1.500041336745437e-05, + "loss": 0.024, + "step": 63000 + }, + { + "epoch": 5.727140519905472, + "grad_norm": 0.13643647730350494, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0197, + "step": 63010 + }, + { + "epoch": 5.728049445555354, + "grad_norm": 0.14707903563976288, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0235, + "step": 63020 + }, + { + "epoch": 5.728958371205236, + "grad_norm": 0.11915574222803116, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0214, + "step": 63030 + }, + { + "epoch": 5.729867296855117, + "grad_norm": 0.12842348217964172, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0246, + "step": 63040 + }, + { + "epoch": 5.730776222504999, + "grad_norm": 0.09842664748430252, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0202, + "step": 63050 + }, + { + "epoch": 5.731685148154881, + "grad_norm": 0.18858252465724945, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0219, + "step": 63060 + }, + { + "epoch": 5.732594073804763, + "grad_norm": 0.11990848928689957, + "learning_rate": 1.497147766283292e-05, + "loss": 0.0223, + "step": 63070 + }, + { + "epoch": 5.733502999454645, + "grad_norm": 0.16628439724445343, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0215, + "step": 63080 + }, + { + "epoch": 5.734411925104526, + "grad_norm": 0.14865873754024506, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0217, + "step": 63090 + }, + { + "epoch": 5.735320850754408, + "grad_norm": 0.15404647588729858, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0225, + "step": 63100 + }, + { + "epoch": 5.73622977640429, + "grad_norm": 0.14738093316555023, + "learning_rate": 1.495494301522528e-05, + "loss": 0.0213, + "step": 63110 + }, + { + "epoch": 5.737138702054172, + "grad_norm": 0.13883183896541595, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0197, + "step": 63120 + }, + { + "epoch": 5.738047627704054, + "grad_norm": 0.1218172088265419, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0217, + "step": 63130 + }, + { + "epoch": 5.738956553353936, + "grad_norm": 0.11556920409202576, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0244, + "step": 63140 + }, + { + "epoch": 5.739865479003818, + "grad_norm": 0.11764740198850632, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.0226, + "step": 63150 + }, + { + "epoch": 5.740774404653699, + "grad_norm": 0.11068661510944366, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0218, + "step": 63160 + }, + { + "epoch": 5.741683330303581, + "grad_norm": 0.15774233639240265, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0202, + "step": 63170 + }, + { + "epoch": 5.742592255953463, + "grad_norm": 0.21172484755516052, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0243, + "step": 63180 + }, + { + "epoch": 5.743501181603345, + "grad_norm": 0.10455655306577682, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0215, + "step": 63190 + }, + { + "epoch": 5.744410107253227, + "grad_norm": 0.122980035841465, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.0203, + "step": 63200 + }, + { + "epoch": 5.745319032903108, + "grad_norm": 0.12794144451618195, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0207, + "step": 63210 + }, + { + "epoch": 5.74622795855299, + "grad_norm": 0.14211232960224152, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0207, + "step": 63220 + }, + { + "epoch": 5.747136884202872, + "grad_norm": 0.14727377891540527, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.023, + "step": 63230 + }, + { + "epoch": 5.748045809852754, + "grad_norm": 0.14540989696979523, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0225, + "step": 63240 + }, + { + "epoch": 5.748954735502636, + "grad_norm": 0.2327090948820114, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0204, + "step": 63250 + }, + { + "epoch": 5.749863661152518, + "grad_norm": 0.14601722359657288, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0228, + "step": 63260 + }, + { + "epoch": 5.7507725868024, + "grad_norm": 0.13574428856372833, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0197, + "step": 63270 + }, + { + "epoch": 5.751681512452281, + "grad_norm": 0.15116772055625916, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0201, + "step": 63280 + }, + { + "epoch": 5.752590438102163, + "grad_norm": 0.11747930943965912, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.0214, + "step": 63290 + }, + { + "epoch": 5.753499363752045, + "grad_norm": 0.14693602919578552, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0228, + "step": 63300 + }, + { + "epoch": 5.754408289401927, + "grad_norm": 0.22771194577217102, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0211, + "step": 63310 + }, + { + "epoch": 5.755317215051809, + "grad_norm": 0.12644892930984497, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0199, + "step": 63320 + }, + { + "epoch": 5.75622614070169, + "grad_norm": 0.14594446122646332, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.0215, + "step": 63330 + }, + { + "epoch": 5.757135066351572, + "grad_norm": 0.15179063379764557, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.0208, + "step": 63340 + }, + { + "epoch": 5.758043992001454, + "grad_norm": 0.15861453115940094, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.0252, + "step": 63350 + }, + { + "epoch": 5.758952917651336, + "grad_norm": 0.10464303195476532, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0216, + "step": 63360 + }, + { + "epoch": 5.759861843301218, + "grad_norm": 0.14729301631450653, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0228, + "step": 63370 + }, + { + "epoch": 5.7607707689511, + "grad_norm": 0.1367502212524414, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0215, + "step": 63380 + }, + { + "epoch": 5.761679694600982, + "grad_norm": 0.1395420879125595, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.0241, + "step": 63390 + }, + { + "epoch": 5.762588620250863, + "grad_norm": 0.1471455693244934, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0219, + "step": 63400 + }, + { + "epoch": 5.763497545900745, + "grad_norm": 0.12700217962265015, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.022, + "step": 63410 + }, + { + "epoch": 5.764406471550627, + "grad_norm": 0.12480642646551132, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0202, + "step": 63420 + }, + { + "epoch": 5.765315397200509, + "grad_norm": 0.10144317895174026, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0217, + "step": 63430 + }, + { + "epoch": 5.766224322850391, + "grad_norm": 0.10454452037811279, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0201, + "step": 63440 + }, + { + "epoch": 5.767133248500272, + "grad_norm": 0.15774521231651306, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0224, + "step": 63450 + }, + { + "epoch": 5.768042174150154, + "grad_norm": 0.1194932833313942, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0227, + "step": 63460 + }, + { + "epoch": 5.768951099800036, + "grad_norm": 0.10814372450113297, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.0201, + "step": 63470 + }, + { + "epoch": 5.769860025449918, + "grad_norm": 0.18899008631706238, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0227, + "step": 63480 + }, + { + "epoch": 5.7707689510998, + "grad_norm": 0.126585453748703, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0207, + "step": 63490 + }, + { + "epoch": 5.771677876749682, + "grad_norm": 0.12500523030757904, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0199, + "step": 63500 + }, + { + "epoch": 5.772586802399564, + "grad_norm": 0.12180227041244507, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.0205, + "step": 63510 + }, + { + "epoch": 5.773495728049445, + "grad_norm": 0.13992370665073395, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0214, + "step": 63520 + }, + { + "epoch": 5.774404653699327, + "grad_norm": 0.1721871793270111, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0214, + "step": 63530 + }, + { + "epoch": 5.775313579349209, + "grad_norm": 0.12143829464912415, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.0206, + "step": 63540 + }, + { + "epoch": 5.776222504999091, + "grad_norm": 0.1177392452955246, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0225, + "step": 63550 + }, + { + "epoch": 5.777131430648973, + "grad_norm": 0.17918084561824799, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0219, + "step": 63560 + }, + { + "epoch": 5.778040356298854, + "grad_norm": 0.10086822509765625, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0207, + "step": 63570 + }, + { + "epoch": 5.778949281948736, + "grad_norm": 0.16201893985271454, + "learning_rate": 1.476067039948277e-05, + "loss": 0.0224, + "step": 63580 + }, + { + "epoch": 5.779858207598618, + "grad_norm": 0.15112917125225067, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0219, + "step": 63590 + }, + { + "epoch": 5.7807671332485, + "grad_norm": 0.17188678681850433, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0207, + "step": 63600 + }, + { + "epoch": 5.781676058898382, + "grad_norm": 0.09240343421697617, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0197, + "step": 63610 + }, + { + "epoch": 5.782584984548264, + "grad_norm": 0.10867595672607422, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0212, + "step": 63620 + }, + { + "epoch": 5.783493910198146, + "grad_norm": 0.16851018369197845, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0211, + "step": 63630 + }, + { + "epoch": 5.784402835848027, + "grad_norm": 0.14330069720745087, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0209, + "step": 63640 + }, + { + "epoch": 5.785311761497909, + "grad_norm": 0.10915612429380417, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.024, + "step": 63650 + }, + { + "epoch": 5.786220687147791, + "grad_norm": 0.11276938021183014, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0235, + "step": 63660 + }, + { + "epoch": 5.787129612797673, + "grad_norm": 0.12429101765155792, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0208, + "step": 63670 + }, + { + "epoch": 5.788038538447555, + "grad_norm": 0.13463327288627625, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0229, + "step": 63680 + }, + { + "epoch": 5.788947464097437, + "grad_norm": 0.11121329665184021, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0244, + "step": 63690 + }, + { + "epoch": 5.789856389747318, + "grad_norm": 0.08693620562553406, + "learning_rate": 1.471107401829042e-05, + "loss": 0.0214, + "step": 63700 + }, + { + "epoch": 5.7907653153972, + "grad_norm": 0.11216181516647339, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0226, + "step": 63710 + }, + { + "epoch": 5.791674241047082, + "grad_norm": 0.18162983655929565, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0218, + "step": 63720 + }, + { + "epoch": 5.792583166696964, + "grad_norm": 0.12108059972524643, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0232, + "step": 63730 + }, + { + "epoch": 5.793492092346846, + "grad_norm": 0.1582793891429901, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0205, + "step": 63740 + }, + { + "epoch": 5.794401017996728, + "grad_norm": 0.12790381908416748, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0222, + "step": 63750 + }, + { + "epoch": 5.795309943646609, + "grad_norm": 0.17755411565303802, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0218, + "step": 63760 + }, + { + "epoch": 5.796218869296491, + "grad_norm": 0.1885816752910614, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0221, + "step": 63770 + }, + { + "epoch": 5.797127794946373, + "grad_norm": 0.11810880154371262, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.0204, + "step": 63780 + }, + { + "epoch": 5.798036720596255, + "grad_norm": 0.13578404486179352, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0201, + "step": 63790 + }, + { + "epoch": 5.798945646246137, + "grad_norm": 0.1396212875843048, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0237, + "step": 63800 + }, + { + "epoch": 5.799854571896019, + "grad_norm": 0.15289117395877838, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0233, + "step": 63810 + }, + { + "epoch": 5.8007634975459, + "grad_norm": 0.17644478380680084, + "learning_rate": 1.466148079674631e-05, + "loss": 0.0231, + "step": 63820 + }, + { + "epoch": 5.801672423195782, + "grad_norm": 0.12061529606580734, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0225, + "step": 63830 + }, + { + "epoch": 5.802581348845664, + "grad_norm": 0.13348384201526642, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0198, + "step": 63840 + }, + { + "epoch": 5.803490274495546, + "grad_norm": 0.1387958973646164, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.023, + "step": 63850 + }, + { + "epoch": 5.804399200145428, + "grad_norm": 0.10198728740215302, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.019, + "step": 63860 + }, + { + "epoch": 5.80530812579531, + "grad_norm": 0.1363973170518875, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0211, + "step": 63870 + }, + { + "epoch": 5.806217051445191, + "grad_norm": 0.13420183956623077, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0205, + "step": 63880 + }, + { + "epoch": 5.807125977095073, + "grad_norm": 0.14188003540039062, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0205, + "step": 63890 + }, + { + "epoch": 5.808034902744955, + "grad_norm": 0.220962792634964, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0211, + "step": 63900 + }, + { + "epoch": 5.808943828394837, + "grad_norm": 0.15297424793243408, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.0244, + "step": 63910 + }, + { + "epoch": 5.809852754044719, + "grad_norm": 0.13037456572055817, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0235, + "step": 63920 + }, + { + "epoch": 5.810761679694601, + "grad_norm": 0.17904780805110931, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0213, + "step": 63930 + }, + { + "epoch": 5.811670605344483, + "grad_norm": 0.13070574402809143, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0228, + "step": 63940 + }, + { + "epoch": 5.812579530994364, + "grad_norm": 0.13137637078762054, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0193, + "step": 63950 + }, + { + "epoch": 5.813488456644246, + "grad_norm": 0.13670198619365692, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0214, + "step": 63960 + }, + { + "epoch": 5.814397382294128, + "grad_norm": 0.221156045794487, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0208, + "step": 63970 + }, + { + "epoch": 5.81530630794401, + "grad_norm": 0.14967377483844757, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0229, + "step": 63980 + }, + { + "epoch": 5.816215233593892, + "grad_norm": 0.11387895792722702, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0225, + "step": 63990 + }, + { + "epoch": 5.8171241592437735, + "grad_norm": 0.16624294221401215, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0207, + "step": 64000 + }, + { + "epoch": 5.8180330848936554, + "grad_norm": 0.13684140145778656, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0202, + "step": 64010 + }, + { + "epoch": 5.818942010543537, + "grad_norm": 0.1335516721010208, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0213, + "step": 64020 + }, + { + "epoch": 5.819850936193419, + "grad_norm": 0.19859696924686432, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.0201, + "step": 64030 + }, + { + "epoch": 5.820759861843301, + "grad_norm": 0.1332544982433319, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.0215, + "step": 64040 + }, + { + "epoch": 5.821668787493183, + "grad_norm": 0.15242928266525269, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0246, + "step": 64050 + }, + { + "epoch": 5.822577713143065, + "grad_norm": 0.18050701916217804, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0207, + "step": 64060 + }, + { + "epoch": 5.8234866387929465, + "grad_norm": 0.1340499073266983, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.0196, + "step": 64070 + }, + { + "epoch": 5.8243955644428285, + "grad_norm": 0.16313786804676056, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0219, + "step": 64080 + }, + { + "epoch": 5.8253044900927105, + "grad_norm": 0.13172419369220734, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0227, + "step": 64090 + }, + { + "epoch": 5.826213415742592, + "grad_norm": 0.19241823256015778, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0217, + "step": 64100 + }, + { + "epoch": 5.827122341392474, + "grad_norm": 0.16914640367031097, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0222, + "step": 64110 + }, + { + "epoch": 5.8280312670423555, + "grad_norm": 0.18843314051628113, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0225, + "step": 64120 + }, + { + "epoch": 5.8289401926922375, + "grad_norm": 0.13107599318027496, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.021, + "step": 64130 + }, + { + "epoch": 5.8298491183421195, + "grad_norm": 0.11208147555589676, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0203, + "step": 64140 + }, + { + "epoch": 5.8307580439920015, + "grad_norm": 0.1209508553147316, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0207, + "step": 64150 + }, + { + "epoch": 5.8316669696418835, + "grad_norm": 0.13558514416217804, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.024, + "step": 64160 + }, + { + "epoch": 5.8325758952917655, + "grad_norm": 0.12834659218788147, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.021, + "step": 64170 + }, + { + "epoch": 5.833484820941647, + "grad_norm": 0.12649011611938477, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0217, + "step": 64180 + }, + { + "epoch": 5.8343937465915285, + "grad_norm": 0.12860268354415894, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0191, + "step": 64190 + }, + { + "epoch": 5.8353026722414105, + "grad_norm": 0.10226568579673767, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.0226, + "step": 64200 + }, + { + "epoch": 5.8362115978912925, + "grad_norm": 0.0991983488202095, + "learning_rate": 1.45003312024933e-05, + "loss": 0.021, + "step": 64210 + }, + { + "epoch": 5.8371205235411745, + "grad_norm": 0.11824476718902588, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0202, + "step": 64220 + }, + { + "epoch": 5.8380294491910565, + "grad_norm": 0.1049150824546814, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0213, + "step": 64230 + }, + { + "epoch": 5.838938374840938, + "grad_norm": 0.13103412091732025, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0221, + "step": 64240 + }, + { + "epoch": 5.83984730049082, + "grad_norm": 0.12858577072620392, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0199, + "step": 64250 + }, + { + "epoch": 5.840756226140702, + "grad_norm": 0.14973358809947968, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.019, + "step": 64260 + }, + { + "epoch": 5.8416651517905835, + "grad_norm": 0.1105445846915245, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.0211, + "step": 64270 + }, + { + "epoch": 5.8425740774404655, + "grad_norm": 0.10453902184963226, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0209, + "step": 64280 + }, + { + "epoch": 5.8434830030903475, + "grad_norm": 0.11347455531358719, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0228, + "step": 64290 + }, + { + "epoch": 5.8443919287402295, + "grad_norm": 0.13951076567173004, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0215, + "step": 64300 + }, + { + "epoch": 5.845300854390111, + "grad_norm": 0.13672737777233124, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0234, + "step": 64310 + }, + { + "epoch": 5.846209780039993, + "grad_norm": 0.1772775799036026, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0207, + "step": 64320 + }, + { + "epoch": 5.847118705689875, + "grad_norm": 0.1275787651538849, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.0197, + "step": 64330 + }, + { + "epoch": 5.848027631339757, + "grad_norm": 0.1903161257505417, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0203, + "step": 64340 + }, + { + "epoch": 5.8489365569896385, + "grad_norm": 0.11829141527414322, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0213, + "step": 64350 + }, + { + "epoch": 5.84984548263952, + "grad_norm": 0.14357785880565643, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.0233, + "step": 64360 + }, + { + "epoch": 5.850754408289402, + "grad_norm": 0.12142477929592133, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0222, + "step": 64370 + }, + { + "epoch": 5.851663333939284, + "grad_norm": 0.1580205112695694, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0223, + "step": 64380 + }, + { + "epoch": 5.852572259589166, + "grad_norm": 0.13819383084774017, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0217, + "step": 64390 + }, + { + "epoch": 5.853481185239048, + "grad_norm": 0.1339603215456009, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0219, + "step": 64400 + }, + { + "epoch": 5.85439011088893, + "grad_norm": 0.26006051898002625, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0212, + "step": 64410 + }, + { + "epoch": 5.855299036538812, + "grad_norm": 0.15512599050998688, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0204, + "step": 64420 + }, + { + "epoch": 5.856207962188693, + "grad_norm": 0.13984617590904236, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0214, + "step": 64430 + }, + { + "epoch": 5.857116887838575, + "grad_norm": 0.13886681199073792, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0189, + "step": 64440 + }, + { + "epoch": 5.858025813488457, + "grad_norm": 0.14081181585788727, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0214, + "step": 64450 + }, + { + "epoch": 5.858934739138339, + "grad_norm": 0.15211378037929535, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0188, + "step": 64460 + }, + { + "epoch": 5.859843664788221, + "grad_norm": 0.15918166935443878, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.0231, + "step": 64470 + }, + { + "epoch": 5.860752590438102, + "grad_norm": 0.25695666670799255, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0196, + "step": 64480 + }, + { + "epoch": 5.861661516087984, + "grad_norm": 0.16193890571594238, + "learning_rate": 1.438466857240496e-05, + "loss": 0.023, + "step": 64490 + }, + { + "epoch": 5.862570441737866, + "grad_norm": 0.11946829408407211, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0191, + "step": 64500 + }, + { + "epoch": 5.863479367387748, + "grad_norm": 0.18847812712192535, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0202, + "step": 64510 + }, + { + "epoch": 5.86438829303763, + "grad_norm": 0.14434808492660522, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0221, + "step": 64520 + }, + { + "epoch": 5.865297218687512, + "grad_norm": 0.117594875395298, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0201, + "step": 64530 + }, + { + "epoch": 5.866206144337394, + "grad_norm": 0.1522115021944046, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0232, + "step": 64540 + }, + { + "epoch": 5.867115069987275, + "grad_norm": 0.09546950459480286, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0236, + "step": 64550 + }, + { + "epoch": 5.868023995637157, + "grad_norm": 0.10919345915317535, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0193, + "step": 64560 + }, + { + "epoch": 5.868932921287039, + "grad_norm": 0.15389734506607056, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.019, + "step": 64570 + }, + { + "epoch": 5.869841846936921, + "grad_norm": 0.10999342054128647, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0197, + "step": 64580 + }, + { + "epoch": 5.870750772586803, + "grad_norm": 0.09734686464071274, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0212, + "step": 64590 + }, + { + "epoch": 5.871659698236684, + "grad_norm": 0.1384236365556717, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.0223, + "step": 64600 + }, + { + "epoch": 5.872568623886566, + "grad_norm": 0.11853879690170288, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.0195, + "step": 64610 + }, + { + "epoch": 5.873477549536448, + "grad_norm": 0.13491326570510864, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.021, + "step": 64620 + }, + { + "epoch": 5.87438647518633, + "grad_norm": 0.12478939443826675, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0185, + "step": 64630 + }, + { + "epoch": 5.875295400836212, + "grad_norm": 0.11240430921316147, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0206, + "step": 64640 + }, + { + "epoch": 5.876204326486094, + "grad_norm": 0.12055804580450058, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.0199, + "step": 64650 + }, + { + "epoch": 5.877113252135976, + "grad_norm": 0.1302010864019394, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0205, + "step": 64660 + }, + { + "epoch": 5.878022177785857, + "grad_norm": 0.13988469541072845, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0222, + "step": 64670 + }, + { + "epoch": 5.878931103435739, + "grad_norm": 0.09388969838619232, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0207, + "step": 64680 + }, + { + "epoch": 5.879840029085621, + "grad_norm": 0.10823766887187958, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0207, + "step": 64690 + }, + { + "epoch": 5.880748954735503, + "grad_norm": 0.11501678079366684, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0217, + "step": 64700 + }, + { + "epoch": 5.881657880385385, + "grad_norm": 0.13015246391296387, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0228, + "step": 64710 + }, + { + "epoch": 5.882566806035266, + "grad_norm": 0.12572596967220306, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0186, + "step": 64720 + }, + { + "epoch": 5.883475731685148, + "grad_norm": 0.13809330761432648, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0223, + "step": 64730 + }, + { + "epoch": 5.88438465733503, + "grad_norm": 0.11217989027500153, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0188, + "step": 64740 + }, + { + "epoch": 5.885293582984912, + "grad_norm": 0.10632485896348953, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0218, + "step": 64750 + }, + { + "epoch": 5.886202508634794, + "grad_norm": 0.12263345718383789, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.019, + "step": 64760 + }, + { + "epoch": 5.887111434284676, + "grad_norm": 0.14015476405620575, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0228, + "step": 64770 + }, + { + "epoch": 5.888020359934558, + "grad_norm": 0.1539931297302246, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0209, + "step": 64780 + }, + { + "epoch": 5.888929285584439, + "grad_norm": 0.10549871623516083, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0203, + "step": 64790 + }, + { + "epoch": 5.889838211234321, + "grad_norm": 0.12801459431648254, + "learning_rate": 1.425665653918373e-05, + "loss": 0.022, + "step": 64800 + }, + { + "epoch": 5.890747136884203, + "grad_norm": 0.12482264637947083, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0206, + "step": 64810 + }, + { + "epoch": 5.891656062534085, + "grad_norm": 0.15592476725578308, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0225, + "step": 64820 + }, + { + "epoch": 5.892564988183967, + "grad_norm": 0.11758241802453995, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.0225, + "step": 64830 + }, + { + "epoch": 5.893473913833849, + "grad_norm": 0.11288212984800339, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0236, + "step": 64840 + }, + { + "epoch": 5.89438283948373, + "grad_norm": 0.12963546812534332, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.0219, + "step": 64850 + }, + { + "epoch": 5.895291765133612, + "grad_norm": 0.11326508224010468, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0216, + "step": 64860 + }, + { + "epoch": 5.896200690783494, + "grad_norm": 0.14607684314250946, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0197, + "step": 64870 + }, + { + "epoch": 5.897109616433376, + "grad_norm": 0.14066998660564423, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.0201, + "step": 64880 + }, + { + "epoch": 5.898018542083258, + "grad_norm": 0.10024883598089218, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0186, + "step": 64890 + }, + { + "epoch": 5.89892746773314, + "grad_norm": 0.1207190752029419, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0206, + "step": 64900 + }, + { + "epoch": 5.899836393383021, + "grad_norm": 0.11742702126502991, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0184, + "step": 64910 + }, + { + "epoch": 5.900745319032903, + "grad_norm": 0.1516062319278717, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0192, + "step": 64920 + }, + { + "epoch": 5.901654244682785, + "grad_norm": 0.1385277956724167, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.0222, + "step": 64930 + }, + { + "epoch": 5.902563170332667, + "grad_norm": 0.12367808818817139, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.0204, + "step": 64940 + }, + { + "epoch": 5.903472095982549, + "grad_norm": 0.0853579044342041, + "learning_rate": 1.419473413173778e-05, + "loss": 0.019, + "step": 64950 + }, + { + "epoch": 5.904381021632431, + "grad_norm": 0.1780293583869934, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0192, + "step": 64960 + }, + { + "epoch": 5.905289947282312, + "grad_norm": 0.14199037849903107, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.0205, + "step": 64970 + }, + { + "epoch": 5.906198872932194, + "grad_norm": 0.12573038041591644, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.0228, + "step": 64980 + }, + { + "epoch": 5.907107798582076, + "grad_norm": 0.12632614374160767, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.024, + "step": 64990 + }, + { + "epoch": 5.908016724231958, + "grad_norm": 0.09957747906446457, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0205, + "step": 65000 + }, + { + "epoch": 5.90892564988184, + "grad_norm": 0.181473970413208, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0223, + "step": 65010 + }, + { + "epoch": 5.909834575531722, + "grad_norm": 0.11158038675785065, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0216, + "step": 65020 + }, + { + "epoch": 5.910743501181603, + "grad_norm": 0.20908083021640778, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0244, + "step": 65030 + }, + { + "epoch": 5.911652426831485, + "grad_norm": 0.13316725194454193, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0203, + "step": 65040 + }, + { + "epoch": 5.912561352481367, + "grad_norm": 0.14183486998081207, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0204, + "step": 65050 + }, + { + "epoch": 5.913470278131249, + "grad_norm": 0.1340533345937729, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0188, + "step": 65060 + }, + { + "epoch": 5.914379203781131, + "grad_norm": 0.13865289092063904, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0215, + "step": 65070 + }, + { + "epoch": 5.915288129431013, + "grad_norm": 0.12434660643339157, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.021, + "step": 65080 + }, + { + "epoch": 5.916197055080894, + "grad_norm": 0.1345825493335724, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0219, + "step": 65090 + }, + { + "epoch": 5.917105980730776, + "grad_norm": 0.10993675887584686, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.02, + "step": 65100 + }, + { + "epoch": 5.918014906380658, + "grad_norm": 0.14297935366630554, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0201, + "step": 65110 + }, + { + "epoch": 5.91892383203054, + "grad_norm": 0.10317785292863846, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0197, + "step": 65120 + }, + { + "epoch": 5.919832757680422, + "grad_norm": 0.14995267987251282, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0207, + "step": 65130 + }, + { + "epoch": 5.920741683330304, + "grad_norm": 0.13142341375350952, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0216, + "step": 65140 + }, + { + "epoch": 5.921650608980185, + "grad_norm": 0.11931964010000229, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0201, + "step": 65150 + }, + { + "epoch": 5.922559534630067, + "grad_norm": 0.1387496143579483, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0203, + "step": 65160 + }, + { + "epoch": 5.923468460279949, + "grad_norm": 0.11029352992773056, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.0213, + "step": 65170 + }, + { + "epoch": 5.924377385929831, + "grad_norm": 0.12477531284093857, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0198, + "step": 65180 + }, + { + "epoch": 5.925286311579713, + "grad_norm": 0.130434051156044, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.0221, + "step": 65190 + }, + { + "epoch": 5.926195237229595, + "grad_norm": 0.1543947458267212, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.0232, + "step": 65200 + }, + { + "epoch": 5.927104162879477, + "grad_norm": 0.1661003828048706, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0233, + "step": 65210 + }, + { + "epoch": 5.928013088529358, + "grad_norm": 0.1351844221353531, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0227, + "step": 65220 + }, + { + "epoch": 5.92892201417924, + "grad_norm": 0.12059105187654495, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0211, + "step": 65230 + }, + { + "epoch": 5.929830939829122, + "grad_norm": 0.13989023864269257, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0218, + "step": 65240 + }, + { + "epoch": 5.930739865479004, + "grad_norm": 0.17058546841144562, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0213, + "step": 65250 + }, + { + "epoch": 5.931648791128886, + "grad_norm": 0.1352432370185852, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0253, + "step": 65260 + }, + { + "epoch": 5.932557716778767, + "grad_norm": 0.10167931020259857, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0209, + "step": 65270 + }, + { + "epoch": 5.933466642428649, + "grad_norm": 0.1138206347823143, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.0214, + "step": 65280 + }, + { + "epoch": 5.934375568078531, + "grad_norm": 0.13831187784671783, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0221, + "step": 65290 + }, + { + "epoch": 5.935284493728413, + "grad_norm": 0.17513339221477509, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0226, + "step": 65300 + }, + { + "epoch": 5.936193419378295, + "grad_norm": 0.12518273293972015, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0226, + "step": 65310 + }, + { + "epoch": 5.937102345028177, + "grad_norm": 0.11904432624578476, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.0223, + "step": 65320 + }, + { + "epoch": 5.938011270678059, + "grad_norm": 0.17064324021339417, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0195, + "step": 65330 + }, + { + "epoch": 5.93892019632794, + "grad_norm": 0.11757323890924454, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0201, + "step": 65340 + }, + { + "epoch": 5.939829121977822, + "grad_norm": 0.13611063361167908, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0227, + "step": 65350 + }, + { + "epoch": 5.940738047627704, + "grad_norm": 0.12288966774940491, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.0195, + "step": 65360 + }, + { + "epoch": 5.941646973277586, + "grad_norm": 0.14582739770412445, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0206, + "step": 65370 + }, + { + "epoch": 5.942555898927468, + "grad_norm": 0.11768075078725815, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0212, + "step": 65380 + }, + { + "epoch": 5.943464824577349, + "grad_norm": 0.14046548306941986, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0217, + "step": 65390 + }, + { + "epoch": 5.944373750227231, + "grad_norm": 0.11593221127986908, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0202, + "step": 65400 + }, + { + "epoch": 5.945282675877113, + "grad_norm": 0.1224864050745964, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0224, + "step": 65410 + }, + { + "epoch": 5.946191601526995, + "grad_norm": 0.10342998057603836, + "learning_rate": 1.40008045614636e-05, + "loss": 0.023, + "step": 65420 + }, + { + "epoch": 5.947100527176877, + "grad_norm": 0.16631482541561127, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0209, + "step": 65430 + }, + { + "epoch": 5.948009452826759, + "grad_norm": 0.11337587237358093, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.0189, + "step": 65440 + }, + { + "epoch": 5.948918378476641, + "grad_norm": 0.12590457499027252, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0218, + "step": 65450 + }, + { + "epoch": 5.949827304126522, + "grad_norm": 0.12225430458784103, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0215, + "step": 65460 + }, + { + "epoch": 5.950736229776404, + "grad_norm": 0.11554975807666779, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0212, + "step": 65470 + }, + { + "epoch": 5.951645155426286, + "grad_norm": 0.12281348556280136, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0216, + "step": 65480 + }, + { + "epoch": 5.952554081076168, + "grad_norm": 0.15579742193222046, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.0207, + "step": 65490 + }, + { + "epoch": 5.95346300672605, + "grad_norm": 0.13517074286937714, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0184, + "step": 65500 + }, + { + "epoch": 5.954371932375931, + "grad_norm": 0.11369123309850693, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0203, + "step": 65510 + }, + { + "epoch": 5.955280858025813, + "grad_norm": 0.14608646929264069, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0203, + "step": 65520 + }, + { + "epoch": 5.956189783675695, + "grad_norm": 0.11452297121286392, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0207, + "step": 65530 + }, + { + "epoch": 5.957098709325577, + "grad_norm": 0.12714950740337372, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.02, + "step": 65540 + }, + { + "epoch": 5.958007634975459, + "grad_norm": 0.1342388540506363, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0179, + "step": 65550 + }, + { + "epoch": 5.958916560625341, + "grad_norm": 0.14718078076839447, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0205, + "step": 65560 + }, + { + "epoch": 5.959825486275223, + "grad_norm": 0.11478712409734726, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.0203, + "step": 65570 + }, + { + "epoch": 5.960734411925104, + "grad_norm": 0.10867433995008469, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0216, + "step": 65580 + }, + { + "epoch": 5.961643337574986, + "grad_norm": 0.10328013449907303, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0206, + "step": 65590 + }, + { + "epoch": 5.962552263224868, + "grad_norm": 0.12641915678977966, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0252, + "step": 65600 + }, + { + "epoch": 5.96346118887475, + "grad_norm": 0.18434621393680573, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0211, + "step": 65610 + }, + { + "epoch": 5.964370114524632, + "grad_norm": 0.18146741390228271, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0208, + "step": 65620 + }, + { + "epoch": 5.965279040174513, + "grad_norm": 0.10107483714818954, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0226, + "step": 65630 + }, + { + "epoch": 5.966187965824395, + "grad_norm": 0.13728588819503784, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.021, + "step": 65640 + }, + { + "epoch": 5.967096891474277, + "grad_norm": 0.11863550543785095, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0196, + "step": 65650 + }, + { + "epoch": 5.968005817124159, + "grad_norm": 0.11305871605873108, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0222, + "step": 65660 + }, + { + "epoch": 5.968914742774041, + "grad_norm": 0.15162956714630127, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0228, + "step": 65670 + }, + { + "epoch": 5.969823668423923, + "grad_norm": 0.1427958756685257, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0211, + "step": 65680 + }, + { + "epoch": 5.970732594073805, + "grad_norm": 0.13878147304058075, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0218, + "step": 65690 + }, + { + "epoch": 5.971641519723686, + "grad_norm": 0.11496412754058838, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0213, + "step": 65700 + }, + { + "epoch": 5.972550445373568, + "grad_norm": 0.1183004155755043, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.0191, + "step": 65710 + }, + { + "epoch": 5.97345937102345, + "grad_norm": 0.09684962779283524, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0231, + "step": 65720 + }, + { + "epoch": 5.974368296673332, + "grad_norm": 0.13810107111930847, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0209, + "step": 65730 + }, + { + "epoch": 5.975277222323214, + "grad_norm": 0.17982253432273865, + "learning_rate": 1.386886134183505e-05, + "loss": 0.02, + "step": 65740 + }, + { + "epoch": 5.976186147973095, + "grad_norm": 0.1182318925857544, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.023, + "step": 65750 + }, + { + "epoch": 5.977095073622977, + "grad_norm": 0.12293178588151932, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0231, + "step": 65760 + }, + { + "epoch": 5.978003999272859, + "grad_norm": 0.15025313198566437, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0219, + "step": 65770 + }, + { + "epoch": 5.978912924922741, + "grad_norm": 0.1346287727355957, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0251, + "step": 65780 + }, + { + "epoch": 5.979821850572623, + "grad_norm": 0.14633409678936005, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.0211, + "step": 65790 + }, + { + "epoch": 5.980730776222505, + "grad_norm": 0.1292431652545929, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0205, + "step": 65800 + }, + { + "epoch": 5.981639701872387, + "grad_norm": 0.11134259402751923, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0204, + "step": 65810 + }, + { + "epoch": 5.982548627522268, + "grad_norm": 0.11681841313838959, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0208, + "step": 65820 + }, + { + "epoch": 5.98345755317215, + "grad_norm": 0.15376321971416473, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0228, + "step": 65830 + }, + { + "epoch": 5.984366478822032, + "grad_norm": 0.11827770620584488, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0195, + "step": 65840 + }, + { + "epoch": 5.985275404471914, + "grad_norm": 0.11523868888616562, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0193, + "step": 65850 + }, + { + "epoch": 5.986184330121796, + "grad_norm": 0.1753588318824768, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0198, + "step": 65860 + }, + { + "epoch": 5.987093255771677, + "grad_norm": 0.10733908414840698, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0206, + "step": 65870 + }, + { + "epoch": 5.988002181421559, + "grad_norm": 0.13290837407112122, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0222, + "step": 65880 + }, + { + "epoch": 5.988911107071441, + "grad_norm": 0.175453320145607, + "learning_rate": 1.380704261284659e-05, + "loss": 0.0235, + "step": 65890 + }, + { + "epoch": 5.989820032721323, + "grad_norm": 0.13441681861877441, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0204, + "step": 65900 + }, + { + "epoch": 5.990728958371205, + "grad_norm": 0.13934755325317383, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0199, + "step": 65910 + }, + { + "epoch": 5.991637884021087, + "grad_norm": 0.132072314620018, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.0233, + "step": 65920 + }, + { + "epoch": 5.992546809670969, + "grad_norm": 0.11176648736000061, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.0211, + "step": 65930 + }, + { + "epoch": 5.99345573532085, + "grad_norm": 0.17440631985664368, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0196, + "step": 65940 + }, + { + "epoch": 5.994364660970732, + "grad_norm": 0.15932227671146393, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0204, + "step": 65950 + }, + { + "epoch": 5.995273586620614, + "grad_norm": 0.14837689697742462, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0201, + "step": 65960 + }, + { + "epoch": 5.996182512270496, + "grad_norm": 0.1716843545436859, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.0232, + "step": 65970 + }, + { + "epoch": 5.997091437920378, + "grad_norm": 0.13304685056209564, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.0215, + "step": 65980 + }, + { + "epoch": 5.99800036357026, + "grad_norm": 0.1419852375984192, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0198, + "step": 65990 + }, + { + "epoch": 5.998909289220141, + "grad_norm": 0.12523029744625092, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0236, + "step": 66000 + }, + { + "epoch": 5.999818214870023, + "grad_norm": 0.22943644225597382, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.02, + "step": 66010 + }, + { + "epoch": 6.000727140519905, + "grad_norm": 0.12997502088546753, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0197, + "step": 66020 + }, + { + "epoch": 6.001636066169787, + "grad_norm": 0.12054087221622467, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.022, + "step": 66030 + }, + { + "epoch": 6.002544991819669, + "grad_norm": 0.12051748484373093, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0218, + "step": 66040 + }, + { + "epoch": 6.003453917469551, + "grad_norm": 0.11189880222082138, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0228, + "step": 66050 + }, + { + "epoch": 6.0043628431194325, + "grad_norm": 0.14749890565872192, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0204, + "step": 66060 + }, + { + "epoch": 6.0052717687693145, + "grad_norm": 0.22542409598827362, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0201, + "step": 66070 + }, + { + "epoch": 6.006180694419196, + "grad_norm": 0.12463533878326416, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.0212, + "step": 66080 + }, + { + "epoch": 6.007089620069078, + "grad_norm": 0.13958412408828735, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0223, + "step": 66090 + }, + { + "epoch": 6.00799854571896, + "grad_norm": 0.16365166008472443, + "learning_rate": 1.372053087451907e-05, + "loss": 0.0195, + "step": 66100 + }, + { + "epoch": 6.008907471368842, + "grad_norm": 0.11874357610940933, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0216, + "step": 66110 + }, + { + "epoch": 6.0098163970187235, + "grad_norm": 0.13471567630767822, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0218, + "step": 66120 + }, + { + "epoch": 6.0107253226686055, + "grad_norm": 0.14838698506355286, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0192, + "step": 66130 + }, + { + "epoch": 6.0116342483184875, + "grad_norm": 0.14540007710456848, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0217, + "step": 66140 + }, + { + "epoch": 6.0125431739683695, + "grad_norm": 0.0982920378446579, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0192, + "step": 66150 + }, + { + "epoch": 6.0134520996182514, + "grad_norm": 0.126852884888649, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0202, + "step": 66160 + }, + { + "epoch": 6.014361025268133, + "grad_norm": 0.18747268617153168, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0199, + "step": 66170 + }, + { + "epoch": 6.0152699509180145, + "grad_norm": 0.11709772050380707, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0225, + "step": 66180 + }, + { + "epoch": 6.0161788765678965, + "grad_norm": 0.1012762188911438, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0201, + "step": 66190 + }, + { + "epoch": 6.0170878022177785, + "grad_norm": 0.13829369843006134, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0222, + "step": 66200 + }, + { + "epoch": 6.0179967278676605, + "grad_norm": 0.1330641359090805, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0216, + "step": 66210 + }, + { + "epoch": 6.0189056535175425, + "grad_norm": 0.15241113305091858, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0205, + "step": 66220 + }, + { + "epoch": 6.0198145791674245, + "grad_norm": 0.10594239085912704, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.0189, + "step": 66230 + }, + { + "epoch": 6.020723504817306, + "grad_norm": 0.12887455523014069, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0224, + "step": 66240 + }, + { + "epoch": 6.0216324304671875, + "grad_norm": 0.1420157253742218, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0225, + "step": 66250 + }, + { + "epoch": 6.0225413561170695, + "grad_norm": 0.30135396122932434, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0208, + "step": 66260 + }, + { + "epoch": 6.0234502817669515, + "grad_norm": 0.13622628152370453, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.022, + "step": 66270 + }, + { + "epoch": 6.0243592074168335, + "grad_norm": 0.24657347798347473, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0214, + "step": 66280 + }, + { + "epoch": 6.0252681330667155, + "grad_norm": 0.11687003821134567, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0202, + "step": 66290 + }, + { + "epoch": 6.026177058716597, + "grad_norm": 0.10470312088727951, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.0234, + "step": 66300 + }, + { + "epoch": 6.027085984366479, + "grad_norm": 0.11603283137083054, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0215, + "step": 66310 + }, + { + "epoch": 6.027994910016361, + "grad_norm": 0.11739940196275711, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0203, + "step": 66320 + }, + { + "epoch": 6.0289038356662425, + "grad_norm": 0.13245020806789398, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0227, + "step": 66330 + }, + { + "epoch": 6.0298127613161245, + "grad_norm": 0.15400277078151703, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0219, + "step": 66340 + }, + { + "epoch": 6.0307216869660065, + "grad_norm": 0.14227987825870514, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0216, + "step": 66350 + }, + { + "epoch": 6.031630612615888, + "grad_norm": 0.11940189450979233, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.0222, + "step": 66360 + }, + { + "epoch": 6.03253953826577, + "grad_norm": 0.10841377824544907, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0191, + "step": 66370 + }, + { + "epoch": 6.033448463915652, + "grad_norm": 0.13640952110290527, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0204, + "step": 66380 + }, + { + "epoch": 6.034357389565534, + "grad_norm": 0.12874901294708252, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0213, + "step": 66390 + }, + { + "epoch": 6.035266315215416, + "grad_norm": 0.1371314376592636, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.026, + "step": 66400 + }, + { + "epoch": 6.0361752408652976, + "grad_norm": 0.10526842623949051, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.0191, + "step": 66410 + }, + { + "epoch": 6.037084166515179, + "grad_norm": 0.13233253359794617, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0227, + "step": 66420 + }, + { + "epoch": 6.037993092165061, + "grad_norm": 0.14666876196861267, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0208, + "step": 66430 + }, + { + "epoch": 6.038902017814943, + "grad_norm": 0.12449424713850021, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0187, + "step": 66440 + }, + { + "epoch": 6.039810943464825, + "grad_norm": 0.13492783904075623, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0217, + "step": 66450 + }, + { + "epoch": 6.040719869114707, + "grad_norm": 0.15320970118045807, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.02, + "step": 66460 + }, + { + "epoch": 6.041628794764589, + "grad_norm": 0.10614907741546631, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0205, + "step": 66470 + }, + { + "epoch": 6.04253772041447, + "grad_norm": 0.14197486639022827, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0235, + "step": 66480 + }, + { + "epoch": 6.043446646064352, + "grad_norm": 0.13804957270622253, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.0206, + "step": 66490 + }, + { + "epoch": 6.044355571714234, + "grad_norm": 0.12420278042554855, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0182, + "step": 66500 + }, + { + "epoch": 6.045264497364116, + "grad_norm": 0.11751344799995422, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.023, + "step": 66510 + }, + { + "epoch": 6.046173423013998, + "grad_norm": 0.11045676469802856, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.02, + "step": 66520 + }, + { + "epoch": 6.04708234866388, + "grad_norm": 0.1116158664226532, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.0211, + "step": 66530 + }, + { + "epoch": 6.047991274313761, + "grad_norm": 0.15888576209545135, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0202, + "step": 66540 + }, + { + "epoch": 6.048900199963643, + "grad_norm": 0.11052310466766357, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.022, + "step": 66550 + }, + { + "epoch": 6.049809125613525, + "grad_norm": 0.10773984342813492, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0225, + "step": 66560 + }, + { + "epoch": 6.050718051263407, + "grad_norm": 0.11604680866003036, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0199, + "step": 66570 + }, + { + "epoch": 6.051626976913289, + "grad_norm": 0.1106758564710617, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0221, + "step": 66580 + }, + { + "epoch": 6.052535902563171, + "grad_norm": 0.13921047747135162, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0217, + "step": 66590 + }, + { + "epoch": 6.053444828213052, + "grad_norm": 0.15144585072994232, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0189, + "step": 66600 + }, + { + "epoch": 6.054353753862934, + "grad_norm": 0.0991760715842247, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.0199, + "step": 66610 + }, + { + "epoch": 6.055262679512816, + "grad_norm": 0.15066342055797577, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0202, + "step": 66620 + }, + { + "epoch": 6.056171605162698, + "grad_norm": 0.13350127637386322, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.023, + "step": 66630 + }, + { + "epoch": 6.05708053081258, + "grad_norm": 0.11389582604169846, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0216, + "step": 66640 + }, + { + "epoch": 6.057989456462462, + "grad_norm": 0.10690376162528992, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.021, + "step": 66650 + }, + { + "epoch": 6.058898382112343, + "grad_norm": 0.13790231943130493, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0203, + "step": 66660 + }, + { + "epoch": 6.059807307762225, + "grad_norm": 0.10691428184509277, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0216, + "step": 66670 + }, + { + "epoch": 6.060716233412107, + "grad_norm": 0.09528546035289764, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0196, + "step": 66680 + }, + { + "epoch": 6.061625159061989, + "grad_norm": 0.19917331635951996, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.0209, + "step": 66690 + }, + { + "epoch": 6.062534084711871, + "grad_norm": 0.10739839822053909, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0213, + "step": 66700 + }, + { + "epoch": 6.063443010361753, + "grad_norm": 0.1359909474849701, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.021, + "step": 66710 + }, + { + "epoch": 6.064351936011635, + "grad_norm": 0.118263840675354, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.022, + "step": 66720 + }, + { + "epoch": 6.065260861661516, + "grad_norm": 0.11487162113189697, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.0223, + "step": 66730 + }, + { + "epoch": 6.066169787311398, + "grad_norm": 0.10567227005958557, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0208, + "step": 66740 + }, + { + "epoch": 6.06707871296128, + "grad_norm": 0.13023823499679565, + "learning_rate": 1.345304084483299e-05, + "loss": 0.022, + "step": 66750 + }, + { + "epoch": 6.067987638611162, + "grad_norm": 0.15413573384284973, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0212, + "step": 66760 + }, + { + "epoch": 6.068896564261044, + "grad_norm": 0.18177881836891174, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0195, + "step": 66770 + }, + { + "epoch": 6.069805489910926, + "grad_norm": 0.10933264344930649, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.0207, + "step": 66780 + }, + { + "epoch": 6.070714415560807, + "grad_norm": 0.1407882124185562, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0221, + "step": 66790 + }, + { + "epoch": 6.071623341210689, + "grad_norm": 0.1538967788219452, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0206, + "step": 66800 + }, + { + "epoch": 6.072532266860571, + "grad_norm": 0.11526171863079071, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0215, + "step": 66810 + }, + { + "epoch": 6.073441192510453, + "grad_norm": 0.14051520824432373, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0209, + "step": 66820 + }, + { + "epoch": 6.074350118160335, + "grad_norm": 0.12025295943021774, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0199, + "step": 66830 + }, + { + "epoch": 6.075259043810217, + "grad_norm": 0.11440834403038025, + "learning_rate": 1.341604094324139e-05, + "loss": 0.0242, + "step": 66840 + }, + { + "epoch": 6.076167969460098, + "grad_norm": 0.15291395783424377, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0213, + "step": 66850 + }, + { + "epoch": 6.07707689510998, + "grad_norm": 0.12495966255664825, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.0217, + "step": 66860 + }, + { + "epoch": 6.077985820759862, + "grad_norm": 0.12057577073574066, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0207, + "step": 66870 + }, + { + "epoch": 6.078894746409744, + "grad_norm": 0.09548119455575943, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.0201, + "step": 66880 + }, + { + "epoch": 6.079803672059626, + "grad_norm": 0.10858970135450363, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0213, + "step": 66890 + }, + { + "epoch": 6.080712597709508, + "grad_norm": 0.13036054372787476, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0198, + "step": 66900 + }, + { + "epoch": 6.081621523359389, + "grad_norm": 0.1280144900083542, + "learning_rate": 1.33872699666556e-05, + "loss": 0.0189, + "step": 66910 + }, + { + "epoch": 6.082530449009271, + "grad_norm": 0.16507679224014282, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0237, + "step": 66920 + }, + { + "epoch": 6.083439374659153, + "grad_norm": 0.1559533178806305, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0224, + "step": 66930 + }, + { + "epoch": 6.084348300309035, + "grad_norm": 0.12821760773658752, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0209, + "step": 66940 + }, + { + "epoch": 6.085257225958917, + "grad_norm": 0.11759389191865921, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0187, + "step": 66950 + }, + { + "epoch": 6.086166151608799, + "grad_norm": 0.10815928131341934, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0195, + "step": 66960 + }, + { + "epoch": 6.08707507725868, + "grad_norm": 0.11270449310541153, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0208, + "step": 66970 + }, + { + "epoch": 6.087984002908562, + "grad_norm": 0.11213963478803635, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0223, + "step": 66980 + }, + { + "epoch": 6.088892928558444, + "grad_norm": 0.11266970634460449, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.0202, + "step": 66990 + }, + { + "epoch": 6.089801854208326, + "grad_norm": 0.1668974608182907, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.0229, + "step": 67000 + }, + { + "epoch": 6.090710779858208, + "grad_norm": 0.11976339668035507, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0193, + "step": 67010 + }, + { + "epoch": 6.09161970550809, + "grad_norm": 0.13607153296470642, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0204, + "step": 67020 + }, + { + "epoch": 6.092528631157971, + "grad_norm": 0.1424032300710678, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0166, + "step": 67030 + }, + { + "epoch": 6.093437556807853, + "grad_norm": 0.12608003616333008, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.0221, + "step": 67040 + }, + { + "epoch": 6.094346482457735, + "grad_norm": 0.14737218618392944, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0202, + "step": 67050 + }, + { + "epoch": 6.095255408107617, + "grad_norm": 0.12941674888134003, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0212, + "step": 67060 + }, + { + "epoch": 6.096164333757499, + "grad_norm": 0.15289537608623505, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0227, + "step": 67070 + }, + { + "epoch": 6.097073259407381, + "grad_norm": 0.12037497013807297, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0195, + "step": 67080 + }, + { + "epoch": 6.097982185057262, + "grad_norm": 0.10884302109479904, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.02, + "step": 67090 + }, + { + "epoch": 6.098891110707144, + "grad_norm": 0.14363667368888855, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.02, + "step": 67100 + }, + { + "epoch": 6.099800036357026, + "grad_norm": 0.1289491355419159, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0226, + "step": 67110 + }, + { + "epoch": 6.100708962006908, + "grad_norm": 0.1158825233578682, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0179, + "step": 67120 + }, + { + "epoch": 6.10161788765679, + "grad_norm": 0.10788667947053909, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0201, + "step": 67130 + }, + { + "epoch": 6.102526813306672, + "grad_norm": 0.12228542566299438, + "learning_rate": 1.329277958632556e-05, + "loss": 0.0217, + "step": 67140 + }, + { + "epoch": 6.103435738956553, + "grad_norm": 0.13157320022583008, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0217, + "step": 67150 + }, + { + "epoch": 6.104344664606435, + "grad_norm": 0.14500153064727783, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0212, + "step": 67160 + }, + { + "epoch": 6.105253590256317, + "grad_norm": 0.10665042698383331, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.023, + "step": 67170 + }, + { + "epoch": 6.106162515906199, + "grad_norm": 0.10706500709056854, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.021, + "step": 67180 + }, + { + "epoch": 6.107071441556081, + "grad_norm": 0.14311550557613373, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0226, + "step": 67190 + }, + { + "epoch": 6.107980367205963, + "grad_norm": 0.16550609469413757, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0193, + "step": 67200 + }, + { + "epoch": 6.108889292855844, + "grad_norm": 0.11550316214561462, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0212, + "step": 67210 + }, + { + "epoch": 6.109798218505726, + "grad_norm": 0.13610568642616272, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.0229, + "step": 67220 + }, + { + "epoch": 6.110707144155608, + "grad_norm": 0.12751412391662598, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.0209, + "step": 67230 + }, + { + "epoch": 6.11161606980549, + "grad_norm": 0.1527644246816635, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0202, + "step": 67240 + }, + { + "epoch": 6.112524995455372, + "grad_norm": 0.20488125085830688, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0207, + "step": 67250 + }, + { + "epoch": 6.113433921105254, + "grad_norm": 0.08788349479436874, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.0211, + "step": 67260 + }, + { + "epoch": 6.114342846755135, + "grad_norm": 0.11959360539913177, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0202, + "step": 67270 + }, + { + "epoch": 6.115251772405017, + "grad_norm": 0.15818294882774353, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0233, + "step": 67280 + }, + { + "epoch": 6.116160698054899, + "grad_norm": 0.12220501899719238, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0189, + "step": 67290 + }, + { + "epoch": 6.117069623704781, + "grad_norm": 0.13518989086151123, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0201, + "step": 67300 + }, + { + "epoch": 6.117978549354663, + "grad_norm": 0.1252623051404953, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.0202, + "step": 67310 + }, + { + "epoch": 6.118887475004545, + "grad_norm": 0.14989754557609558, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0231, + "step": 67320 + }, + { + "epoch": 6.119796400654426, + "grad_norm": 0.11310790479183197, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0197, + "step": 67330 + }, + { + "epoch": 6.120705326304308, + "grad_norm": 0.1112234964966774, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.0195, + "step": 67340 + }, + { + "epoch": 6.12161425195419, + "grad_norm": 0.12446058541536331, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0222, + "step": 67350 + }, + { + "epoch": 6.122523177604072, + "grad_norm": 0.13238878548145294, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0195, + "step": 67360 + }, + { + "epoch": 6.123432103253954, + "grad_norm": 0.16855299472808838, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.018, + "step": 67370 + }, + { + "epoch": 6.124341028903836, + "grad_norm": 0.1606617122888565, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.0217, + "step": 67380 + }, + { + "epoch": 6.125249954553717, + "grad_norm": 0.13273735344409943, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.0193, + "step": 67390 + }, + { + "epoch": 6.126158880203599, + "grad_norm": 0.13254916667938232, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.0204, + "step": 67400 + }, + { + "epoch": 6.127067805853481, + "grad_norm": 0.14320404827594757, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0191, + "step": 67410 + }, + { + "epoch": 6.127976731503363, + "grad_norm": 0.1413758248090744, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0205, + "step": 67420 + }, + { + "epoch": 6.128885657153245, + "grad_norm": 0.13199560344219208, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0227, + "step": 67430 + }, + { + "epoch": 6.129794582803127, + "grad_norm": 0.16364918649196625, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0229, + "step": 67440 + }, + { + "epoch": 6.130703508453008, + "grad_norm": 0.14522214233875275, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0215, + "step": 67450 + }, + { + "epoch": 6.13161243410289, + "grad_norm": 0.12993797659873962, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0218, + "step": 67460 + }, + { + "epoch": 6.132521359752772, + "grad_norm": 0.1420457512140274, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0191, + "step": 67470 + }, + { + "epoch": 6.133430285402654, + "grad_norm": 0.1472363919019699, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0215, + "step": 67480 + }, + { + "epoch": 6.134339211052536, + "grad_norm": 0.09552846848964691, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.0193, + "step": 67490 + }, + { + "epoch": 6.135248136702418, + "grad_norm": 0.10186707973480225, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0204, + "step": 67500 + }, + { + "epoch": 6.136157062352299, + "grad_norm": 0.10084915161132812, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0206, + "step": 67510 + }, + { + "epoch": 6.137065988002181, + "grad_norm": 0.09853579103946686, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0218, + "step": 67520 + }, + { + "epoch": 6.137974913652063, + "grad_norm": 0.11622382700443268, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0209, + "step": 67530 + }, + { + "epoch": 6.138883839301945, + "grad_norm": 0.11889436095952988, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0205, + "step": 67540 + }, + { + "epoch": 6.139792764951827, + "grad_norm": 0.11039933562278748, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0213, + "step": 67550 + }, + { + "epoch": 6.140701690601709, + "grad_norm": 0.1435967981815338, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0195, + "step": 67560 + }, + { + "epoch": 6.14161061625159, + "grad_norm": 0.13323169946670532, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.02, + "step": 67570 + }, + { + "epoch": 6.142519541901472, + "grad_norm": 0.12116946280002594, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0223, + "step": 67580 + }, + { + "epoch": 6.143428467551354, + "grad_norm": 0.10598652809858322, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0202, + "step": 67590 + }, + { + "epoch": 6.144337393201236, + "grad_norm": 0.12015562504529953, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0204, + "step": 67600 + }, + { + "epoch": 6.145246318851118, + "grad_norm": 0.1448618769645691, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0229, + "step": 67610 + }, + { + "epoch": 6.146155244501, + "grad_norm": 0.1672632247209549, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0236, + "step": 67620 + }, + { + "epoch": 6.147064170150881, + "grad_norm": 0.1312318742275238, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0187, + "step": 67630 + }, + { + "epoch": 6.147973095800763, + "grad_norm": 0.13794586062431335, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0204, + "step": 67640 + }, + { + "epoch": 6.148882021450645, + "grad_norm": 0.12438039481639862, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0185, + "step": 67650 + }, + { + "epoch": 6.149790947100527, + "grad_norm": 0.09912765771150589, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0217, + "step": 67660 + }, + { + "epoch": 6.150699872750409, + "grad_norm": 0.14032094180583954, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0183, + "step": 67670 + }, + { + "epoch": 6.151608798400291, + "grad_norm": 0.13369540870189667, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0226, + "step": 67680 + }, + { + "epoch": 6.152517724050172, + "grad_norm": 0.14997300505638123, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.0176, + "step": 67690 + }, + { + "epoch": 6.153426649700054, + "grad_norm": 0.1704021692276001, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.022, + "step": 67700 + }, + { + "epoch": 6.154335575349936, + "grad_norm": 0.10800627619028091, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0203, + "step": 67710 + }, + { + "epoch": 6.155244500999818, + "grad_norm": 0.13871894776821136, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.02, + "step": 67720 + }, + { + "epoch": 6.1561534266497, + "grad_norm": 0.1312067210674286, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0202, + "step": 67730 + }, + { + "epoch": 6.157062352299582, + "grad_norm": 0.13203808665275574, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0243, + "step": 67740 + }, + { + "epoch": 6.157971277949463, + "grad_norm": 0.1353537142276764, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0209, + "step": 67750 + }, + { + "epoch": 6.158880203599345, + "grad_norm": 0.13394282758235931, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.0222, + "step": 67760 + }, + { + "epoch": 6.159789129249227, + "grad_norm": 0.15400980412960052, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0205, + "step": 67770 + }, + { + "epoch": 6.160698054899109, + "grad_norm": 0.10980257391929626, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0202, + "step": 67780 + }, + { + "epoch": 6.161606980548991, + "grad_norm": 0.1128997728228569, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0197, + "step": 67790 + }, + { + "epoch": 6.162515906198873, + "grad_norm": 0.12137371301651001, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.0226, + "step": 67800 + }, + { + "epoch": 6.163424831848754, + "grad_norm": 0.11677610874176025, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0199, + "step": 67810 + }, + { + "epoch": 6.164333757498636, + "grad_norm": 0.12623587250709534, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0211, + "step": 67820 + }, + { + "epoch": 6.165242683148518, + "grad_norm": 0.10896242409944534, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0212, + "step": 67830 + }, + { + "epoch": 6.1661516087984, + "grad_norm": 0.14300811290740967, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0212, + "step": 67840 + }, + { + "epoch": 6.167060534448282, + "grad_norm": 0.1127895936369896, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0199, + "step": 67850 + }, + { + "epoch": 6.167969460098164, + "grad_norm": 0.14932946860790253, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0207, + "step": 67860 + }, + { + "epoch": 6.168878385748046, + "grad_norm": 0.11584021896123886, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.018, + "step": 67870 + }, + { + "epoch": 6.169787311397927, + "grad_norm": 0.1346767097711563, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0209, + "step": 67880 + }, + { + "epoch": 6.170696237047809, + "grad_norm": 0.1305159330368042, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0215, + "step": 67890 + }, + { + "epoch": 6.171605162697691, + "grad_norm": 0.10967609286308289, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.0208, + "step": 67900 + }, + { + "epoch": 6.172514088347573, + "grad_norm": 0.12154966592788696, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0208, + "step": 67910 + }, + { + "epoch": 6.173423013997455, + "grad_norm": 0.12729185819625854, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0215, + "step": 67920 + }, + { + "epoch": 6.174331939647336, + "grad_norm": 0.15403839945793152, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.0218, + "step": 67930 + }, + { + "epoch": 6.175240865297218, + "grad_norm": 0.13537482917308807, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0225, + "step": 67940 + }, + { + "epoch": 6.1761497909471, + "grad_norm": 0.1364797204732895, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.0219, + "step": 67950 + }, + { + "epoch": 6.177058716596982, + "grad_norm": 0.147129088640213, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0234, + "step": 67960 + }, + { + "epoch": 6.177967642246864, + "grad_norm": 0.10705330967903137, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0197, + "step": 67970 + }, + { + "epoch": 6.178876567896746, + "grad_norm": 0.1263802945613861, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0201, + "step": 67980 + }, + { + "epoch": 6.179785493546628, + "grad_norm": 0.12250542640686035, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0215, + "step": 67990 + }, + { + "epoch": 6.180694419196509, + "grad_norm": 0.12662173807621002, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0199, + "step": 68000 + }, + { + "epoch": 6.181603344846391, + "grad_norm": 0.10846677422523499, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0213, + "step": 68010 + }, + { + "epoch": 6.182512270496273, + "grad_norm": 0.13957272469997406, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0203, + "step": 68020 + }, + { + "epoch": 6.183421196146155, + "grad_norm": 0.13657744228839874, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0195, + "step": 68030 + }, + { + "epoch": 6.184330121796037, + "grad_norm": 0.12035361677408218, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0206, + "step": 68040 + }, + { + "epoch": 6.185239047445919, + "grad_norm": 0.15147264301776886, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0218, + "step": 68050 + }, + { + "epoch": 6.1861479730958004, + "grad_norm": 0.14438261091709137, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0245, + "step": 68060 + }, + { + "epoch": 6.187056898745682, + "grad_norm": 0.129057839512825, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.0214, + "step": 68070 + }, + { + "epoch": 6.187965824395564, + "grad_norm": 0.1357303112745285, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0191, + "step": 68080 + }, + { + "epoch": 6.188874750045446, + "grad_norm": 0.12321256846189499, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0187, + "step": 68090 + }, + { + "epoch": 6.189783675695328, + "grad_norm": 0.10966606438159943, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.023, + "step": 68100 + }, + { + "epoch": 6.19069260134521, + "grad_norm": 0.11228107661008835, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0231, + "step": 68110 + }, + { + "epoch": 6.1916015269950915, + "grad_norm": 0.13364169001579285, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0212, + "step": 68120 + }, + { + "epoch": 6.1925104526449735, + "grad_norm": 0.16952086985111237, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0212, + "step": 68130 + }, + { + "epoch": 6.1934193782948554, + "grad_norm": 0.1294461041688919, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0178, + "step": 68140 + }, + { + "epoch": 6.194328303944737, + "grad_norm": 0.11013849079608917, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0188, + "step": 68150 + }, + { + "epoch": 6.195237229594619, + "grad_norm": 0.13345523178577423, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0213, + "step": 68160 + }, + { + "epoch": 6.196146155244501, + "grad_norm": 0.12096432596445084, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0228, + "step": 68170 + }, + { + "epoch": 6.1970550808943825, + "grad_norm": 0.22291962802410126, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0239, + "step": 68180 + }, + { + "epoch": 6.1979640065442645, + "grad_norm": 0.13251501321792603, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0221, + "step": 68190 + }, + { + "epoch": 6.1988729321941465, + "grad_norm": 0.12832915782928467, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0244, + "step": 68200 + }, + { + "epoch": 6.1997818578440285, + "grad_norm": 0.13067927956581116, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.021, + "step": 68210 + }, + { + "epoch": 6.2006907834939105, + "grad_norm": 0.11869189143180847, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0188, + "step": 68220 + }, + { + "epoch": 6.201599709143792, + "grad_norm": 0.12013164907693863, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0207, + "step": 68230 + }, + { + "epoch": 6.2025086347936735, + "grad_norm": 0.18515366315841675, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0188, + "step": 68240 + }, + { + "epoch": 6.2034175604435555, + "grad_norm": 0.13810913264751434, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0195, + "step": 68250 + }, + { + "epoch": 6.2043264860934375, + "grad_norm": 0.12951168417930603, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.0197, + "step": 68260 + }, + { + "epoch": 6.2052354117433195, + "grad_norm": 0.11013428866863251, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0214, + "step": 68270 + }, + { + "epoch": 6.2061443373932015, + "grad_norm": 0.11957786232233047, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0192, + "step": 68280 + }, + { + "epoch": 6.2070532630430835, + "grad_norm": 0.12294145673513412, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.0177, + "step": 68290 + }, + { + "epoch": 6.207962188692965, + "grad_norm": 0.1211753860116005, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0218, + "step": 68300 + }, + { + "epoch": 6.2088711143428466, + "grad_norm": 0.14338137209415436, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0209, + "step": 68310 + }, + { + "epoch": 6.2097800399927285, + "grad_norm": 0.12051370739936829, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0204, + "step": 68320 + }, + { + "epoch": 6.2106889656426105, + "grad_norm": 0.15300290286540985, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0224, + "step": 68330 + }, + { + "epoch": 6.2115978912924925, + "grad_norm": 0.12655456364154816, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0193, + "step": 68340 + }, + { + "epoch": 6.2125068169423745, + "grad_norm": 0.09539680927991867, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0194, + "step": 68350 + }, + { + "epoch": 6.213415742592256, + "grad_norm": 0.12666474282741547, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0201, + "step": 68360 + }, + { + "epoch": 6.214324668242138, + "grad_norm": 0.1034204289317131, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0191, + "step": 68370 + }, + { + "epoch": 6.21523359389202, + "grad_norm": 0.1167985200881958, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0203, + "step": 68380 + }, + { + "epoch": 6.216142519541902, + "grad_norm": 0.12622520327568054, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.0201, + "step": 68390 + }, + { + "epoch": 6.2170514451917835, + "grad_norm": 0.12165524065494537, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0203, + "step": 68400 + }, + { + "epoch": 6.2179603708416655, + "grad_norm": 0.16061833500862122, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.021, + "step": 68410 + }, + { + "epoch": 6.218869296491547, + "grad_norm": 0.13492445647716522, + "learning_rate": 1.276827841111951e-05, + "loss": 0.0221, + "step": 68420 + }, + { + "epoch": 6.219778222141429, + "grad_norm": 0.183275043964386, + "learning_rate": 1.276419082873553e-05, + "loss": 0.0206, + "step": 68430 + }, + { + "epoch": 6.220687147791311, + "grad_norm": 0.11312754452228546, + "learning_rate": 1.27601034161465e-05, + "loss": 0.0209, + "step": 68440 + }, + { + "epoch": 6.221596073441193, + "grad_norm": 0.11588999629020691, + "learning_rate": 1.275601617366283e-05, + "loss": 0.02, + "step": 68450 + }, + { + "epoch": 6.222504999091075, + "grad_norm": 0.1324116289615631, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0207, + "step": 68460 + }, + { + "epoch": 6.223413924740957, + "grad_norm": 0.10099894553422928, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0187, + "step": 68470 + }, + { + "epoch": 6.224322850390838, + "grad_norm": 0.12350031733512878, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0192, + "step": 68480 + }, + { + "epoch": 6.22523177604072, + "grad_norm": 0.1410062164068222, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0204, + "step": 68490 + }, + { + "epoch": 6.226140701690602, + "grad_norm": 0.13619624078273773, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.02, + "step": 68500 + }, + { + "epoch": 6.227049627340484, + "grad_norm": 0.10202237218618393, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0202, + "step": 68510 + }, + { + "epoch": 6.227958552990366, + "grad_norm": 0.1267390102148056, + "learning_rate": 1.272741026529923e-05, + "loss": 0.023, + "step": 68520 + }, + { + "epoch": 6.228867478640248, + "grad_norm": 0.084900863468647, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0208, + "step": 68530 + }, + { + "epoch": 6.229776404290129, + "grad_norm": 0.12132617086172104, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0208, + "step": 68540 + }, + { + "epoch": 6.230685329940011, + "grad_norm": 0.19339051842689514, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0206, + "step": 68550 + }, + { + "epoch": 6.231594255589893, + "grad_norm": 0.1618817150592804, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0196, + "step": 68560 + }, + { + "epoch": 6.232503181239775, + "grad_norm": 0.1318468302488327, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0188, + "step": 68570 + }, + { + "epoch": 6.233412106889657, + "grad_norm": 0.16861936450004578, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0197, + "step": 68580 + }, + { + "epoch": 6.234321032539539, + "grad_norm": 0.12337591499090195, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0219, + "step": 68590 + }, + { + "epoch": 6.23522995818942, + "grad_norm": 0.12191134691238403, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0206, + "step": 68600 + }, + { + "epoch": 6.236138883839302, + "grad_norm": 0.1390169858932495, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0234, + "step": 68610 + }, + { + "epoch": 6.237047809489184, + "grad_norm": 0.12398050725460052, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.0194, + "step": 68620 + }, + { + "epoch": 6.237956735139066, + "grad_norm": 0.13971494138240814, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0194, + "step": 68630 + }, + { + "epoch": 6.238865660788948, + "grad_norm": 0.1040286123752594, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.0209, + "step": 68640 + }, + { + "epoch": 6.23977458643883, + "grad_norm": 0.09771417081356049, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0212, + "step": 68650 + }, + { + "epoch": 6.240683512088711, + "grad_norm": 0.16073593497276306, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.0196, + "step": 68660 + }, + { + "epoch": 6.241592437738593, + "grad_norm": 0.12295322865247726, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.0208, + "step": 68670 + }, + { + "epoch": 6.242501363388475, + "grad_norm": 0.1254149228334427, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0178, + "step": 68680 + }, + { + "epoch": 6.243410289038357, + "grad_norm": 0.18104980885982513, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0202, + "step": 68690 + }, + { + "epoch": 6.244319214688239, + "grad_norm": 0.1562449038028717, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0215, + "step": 68700 + }, + { + "epoch": 6.245228140338121, + "grad_norm": 0.13909868896007538, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.0186, + "step": 68710 + }, + { + "epoch": 6.246137065988002, + "grad_norm": 0.12304549664258957, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0221, + "step": 68720 + }, + { + "epoch": 6.247045991637884, + "grad_norm": 0.11796580255031586, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0229, + "step": 68730 + }, + { + "epoch": 6.247954917287766, + "grad_norm": 0.13726823031902313, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.0208, + "step": 68740 + }, + { + "epoch": 6.248863842937648, + "grad_norm": 0.1397128403186798, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0217, + "step": 68750 + }, + { + "epoch": 6.24977276858753, + "grad_norm": 0.14672577381134033, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0212, + "step": 68760 + }, + { + "epoch": 6.250681694237412, + "grad_norm": 0.11866188794374466, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.021, + "step": 68770 + }, + { + "epoch": 6.251590619887293, + "grad_norm": 0.17486295104026794, + "learning_rate": 1.262123463168752e-05, + "loss": 0.021, + "step": 68780 + }, + { + "epoch": 6.252499545537175, + "grad_norm": 0.09444549679756165, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0199, + "step": 68790 + }, + { + "epoch": 6.253408471187057, + "grad_norm": 0.139812633395195, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0206, + "step": 68800 + }, + { + "epoch": 6.254317396836939, + "grad_norm": 0.11188624054193497, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0195, + "step": 68810 + }, + { + "epoch": 6.255226322486821, + "grad_norm": 0.14105598628520966, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0235, + "step": 68820 + }, + { + "epoch": 6.256135248136703, + "grad_norm": 0.1347184181213379, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0211, + "step": 68830 + }, + { + "epoch": 6.257044173786584, + "grad_norm": 0.14356908202171326, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0212, + "step": 68840 + }, + { + "epoch": 6.257953099436466, + "grad_norm": 0.12938852608203888, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0208, + "step": 68850 + }, + { + "epoch": 6.258862025086348, + "grad_norm": 0.12379366159439087, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0199, + "step": 68860 + }, + { + "epoch": 6.25977095073623, + "grad_norm": 0.1229868158698082, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0215, + "step": 68870 + }, + { + "epoch": 6.260679876386112, + "grad_norm": 0.12863197922706604, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0222, + "step": 68880 + }, + { + "epoch": 6.261588802035994, + "grad_norm": 0.11849579215049744, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0191, + "step": 68890 + }, + { + "epoch": 6.262497727685876, + "grad_norm": 0.11266331374645233, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0209, + "step": 68900 + }, + { + "epoch": 6.263406653335757, + "grad_norm": 0.12296075373888016, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0223, + "step": 68910 + }, + { + "epoch": 6.264315578985639, + "grad_norm": 0.13457942008972168, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0205, + "step": 68920 + }, + { + "epoch": 6.265224504635521, + "grad_norm": 0.12040028721094131, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0206, + "step": 68930 + }, + { + "epoch": 6.266133430285403, + "grad_norm": 0.15157680213451385, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0211, + "step": 68940 + }, + { + "epoch": 6.267042355935285, + "grad_norm": 0.10160267353057861, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0206, + "step": 68950 + }, + { + "epoch": 6.267951281585166, + "grad_norm": 0.1968524008989334, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0197, + "step": 68960 + }, + { + "epoch": 6.268860207235048, + "grad_norm": 0.13598959147930145, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0201, + "step": 68970 + }, + { + "epoch": 6.26976913288493, + "grad_norm": 0.1499781310558319, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.02, + "step": 68980 + }, + { + "epoch": 6.270678058534812, + "grad_norm": 0.12176140397787094, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0227, + "step": 68990 + }, + { + "epoch": 6.271586984184694, + "grad_norm": 0.12302673608064651, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.0227, + "step": 69000 + }, + { + "epoch": 6.272495909834576, + "grad_norm": 0.12387610226869583, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.022, + "step": 69010 + }, + { + "epoch": 6.273404835484458, + "grad_norm": 0.0969795435667038, + "learning_rate": 1.252333462101563e-05, + "loss": 0.0194, + "step": 69020 + }, + { + "epoch": 6.274313761134339, + "grad_norm": 0.11775080859661102, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0183, + "step": 69030 + }, + { + "epoch": 6.275222686784221, + "grad_norm": 0.1625920981168747, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0213, + "step": 69040 + }, + { + "epoch": 6.276131612434103, + "grad_norm": 0.1104586124420166, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.0201, + "step": 69050 + }, + { + "epoch": 6.277040538083985, + "grad_norm": 0.11251900345087051, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0221, + "step": 69060 + }, + { + "epoch": 6.277949463733867, + "grad_norm": 0.12210677564144135, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0209, + "step": 69070 + }, + { + "epoch": 6.278858389383748, + "grad_norm": 0.12683208286762238, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0189, + "step": 69080 + }, + { + "epoch": 6.27976731503363, + "grad_norm": 0.100763700902462, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0184, + "step": 69090 + }, + { + "epoch": 6.280676240683512, + "grad_norm": 0.14936745166778564, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0205, + "step": 69100 + }, + { + "epoch": 6.281585166333394, + "grad_norm": 0.1109953373670578, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0209, + "step": 69110 + }, + { + "epoch": 6.282494091983276, + "grad_norm": 0.1481008231639862, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.019, + "step": 69120 + }, + { + "epoch": 6.283403017633158, + "grad_norm": 0.1750049740076065, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.023, + "step": 69130 + }, + { + "epoch": 6.28431194328304, + "grad_norm": 0.11596126109361649, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0206, + "step": 69140 + }, + { + "epoch": 6.285220868932921, + "grad_norm": 0.09996895492076874, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0222, + "step": 69150 + }, + { + "epoch": 6.286129794582803, + "grad_norm": 0.10348265618085861, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.02, + "step": 69160 + }, + { + "epoch": 6.287038720232685, + "grad_norm": 0.16748392581939697, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0231, + "step": 69170 + }, + { + "epoch": 6.287947645882567, + "grad_norm": 0.14982108771800995, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.0205, + "step": 69180 + }, + { + "epoch": 6.288856571532449, + "grad_norm": 0.1456674039363861, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0211, + "step": 69190 + }, + { + "epoch": 6.28976549718233, + "grad_norm": 0.16920721530914307, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0193, + "step": 69200 + }, + { + "epoch": 6.290674422832212, + "grad_norm": 0.15300829708576202, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.019, + "step": 69210 + }, + { + "epoch": 6.291583348482094, + "grad_norm": 0.1397537738084793, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0227, + "step": 69220 + }, + { + "epoch": 6.292492274131976, + "grad_norm": 0.18503761291503906, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0208, + "step": 69230 + }, + { + "epoch": 6.293401199781858, + "grad_norm": 0.15779376029968262, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0207, + "step": 69240 + }, + { + "epoch": 6.29431012543174, + "grad_norm": 0.11754079908132553, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0199, + "step": 69250 + }, + { + "epoch": 6.295219051081622, + "grad_norm": 0.12775878608226776, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0213, + "step": 69260 + }, + { + "epoch": 6.296127976731503, + "grad_norm": 0.14001233875751495, + "learning_rate": 1.242147070892255e-05, + "loss": 0.021, + "step": 69270 + }, + { + "epoch": 6.297036902381385, + "grad_norm": 0.12655501067638397, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0187, + "step": 69280 + }, + { + "epoch": 6.297945828031267, + "grad_norm": 0.131971538066864, + "learning_rate": 1.241332681912234e-05, + "loss": 0.0209, + "step": 69290 + }, + { + "epoch": 6.298854753681149, + "grad_norm": 0.09996717423200607, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0178, + "step": 69300 + }, + { + "epoch": 6.299763679331031, + "grad_norm": 0.16707108914852142, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0183, + "step": 69310 + }, + { + "epoch": 6.300672604980913, + "grad_norm": 0.1304529905319214, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.0201, + "step": 69320 + }, + { + "epoch": 6.301581530630794, + "grad_norm": 0.1611240953207016, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0202, + "step": 69330 + }, + { + "epoch": 6.302490456280676, + "grad_norm": 0.1258333921432495, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0242, + "step": 69340 + }, + { + "epoch": 6.303399381930558, + "grad_norm": 0.10306375473737717, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0194, + "step": 69350 + }, + { + "epoch": 6.30430830758044, + "grad_norm": 0.13593845069408417, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0196, + "step": 69360 + }, + { + "epoch": 6.305217233230322, + "grad_norm": 0.1532493382692337, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0206, + "step": 69370 + }, + { + "epoch": 6.306126158880204, + "grad_norm": 0.13016942143440247, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0217, + "step": 69380 + }, + { + "epoch": 6.307035084530085, + "grad_norm": 0.12390958517789841, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0209, + "step": 69390 + }, + { + "epoch": 6.307944010179967, + "grad_norm": 0.11321163177490234, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0211, + "step": 69400 + }, + { + "epoch": 6.308852935829849, + "grad_norm": 0.09830807149410248, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0186, + "step": 69410 + }, + { + "epoch": 6.309761861479731, + "grad_norm": 0.11001650989055634, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0227, + "step": 69420 + }, + { + "epoch": 6.310670787129613, + "grad_norm": 0.12307223677635193, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0223, + "step": 69430 + }, + { + "epoch": 6.311579712779495, + "grad_norm": 0.13363473117351532, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.0197, + "step": 69440 + }, + { + "epoch": 6.312488638429376, + "grad_norm": 0.11105284839868546, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.02, + "step": 69450 + }, + { + "epoch": 6.313397564079258, + "grad_norm": 0.11953122913837433, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0209, + "step": 69460 + }, + { + "epoch": 6.31430648972914, + "grad_norm": 0.16651363670825958, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0225, + "step": 69470 + }, + { + "epoch": 6.315215415379022, + "grad_norm": 0.13623295724391937, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0193, + "step": 69480 + }, + { + "epoch": 6.316124341028904, + "grad_norm": 0.38360151648521423, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0205, + "step": 69490 + }, + { + "epoch": 6.317033266678786, + "grad_norm": 0.1730136126279831, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.021, + "step": 69500 + }, + { + "epoch": 6.317942192328667, + "grad_norm": 0.1844894140958786, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0214, + "step": 69510 + }, + { + "epoch": 6.318851117978549, + "grad_norm": 0.11166511476039886, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0208, + "step": 69520 + }, + { + "epoch": 6.319760043628431, + "grad_norm": 0.15125153958797455, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0223, + "step": 69530 + }, + { + "epoch": 6.320668969278313, + "grad_norm": 0.11431486159563065, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0212, + "step": 69540 + }, + { + "epoch": 6.321577894928195, + "grad_norm": 0.14425115287303925, + "learning_rate": 1.230752865632506e-05, + "loss": 0.0213, + "step": 69550 + }, + { + "epoch": 6.322486820578077, + "grad_norm": 0.15888987481594086, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0199, + "step": 69560 + }, + { + "epoch": 6.323395746227958, + "grad_norm": 0.14971937239170074, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0222, + "step": 69570 + }, + { + "epoch": 6.32430467187784, + "grad_norm": 0.11117704212665558, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0218, + "step": 69580 + }, + { + "epoch": 6.325213597527722, + "grad_norm": 0.10699430108070374, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0204, + "step": 69590 + }, + { + "epoch": 6.326122523177604, + "grad_norm": 0.1656959354877472, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0186, + "step": 69600 + }, + { + "epoch": 6.327031448827486, + "grad_norm": 0.13471291959285736, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0241, + "step": 69610 + }, + { + "epoch": 6.327940374477368, + "grad_norm": 0.13692882657051086, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0203, + "step": 69620 + }, + { + "epoch": 6.328849300127249, + "grad_norm": 0.11197061836719513, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.0188, + "step": 69630 + }, + { + "epoch": 6.329758225777131, + "grad_norm": 0.13470381498336792, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0216, + "step": 69640 + }, + { + "epoch": 6.330667151427013, + "grad_norm": 0.13851383328437805, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0224, + "step": 69650 + }, + { + "epoch": 6.331576077076895, + "grad_norm": 0.12657508254051208, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0195, + "step": 69660 + }, + { + "epoch": 6.332485002726777, + "grad_norm": 0.12418321520090103, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0194, + "step": 69670 + }, + { + "epoch": 6.333393928376659, + "grad_norm": 0.10296697914600372, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0173, + "step": 69680 + }, + { + "epoch": 6.33430285402654, + "grad_norm": 0.13198597729206085, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0213, + "step": 69690 + }, + { + "epoch": 6.335211779676422, + "grad_norm": 0.1491917222738266, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0208, + "step": 69700 + }, + { + "epoch": 6.336120705326304, + "grad_norm": 0.15172336995601654, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0198, + "step": 69710 + }, + { + "epoch": 6.337029630976186, + "grad_norm": 0.14818531274795532, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0215, + "step": 69720 + }, + { + "epoch": 6.337938556626068, + "grad_norm": 0.12528015673160553, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0206, + "step": 69730 + }, + { + "epoch": 6.33884748227595, + "grad_norm": 0.11349184066057205, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.0223, + "step": 69740 + }, + { + "epoch": 6.339756407925831, + "grad_norm": 0.1404275745153427, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0199, + "step": 69750 + }, + { + "epoch": 6.340665333575713, + "grad_norm": 0.13331501185894012, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0199, + "step": 69760 + }, + { + "epoch": 6.341574259225595, + "grad_norm": 0.1889750212430954, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.02, + "step": 69770 + }, + { + "epoch": 6.342483184875477, + "grad_norm": 0.11427148431539536, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.0184, + "step": 69780 + }, + { + "epoch": 6.343392110525359, + "grad_norm": 0.21467618644237518, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0186, + "step": 69790 + }, + { + "epoch": 6.344301036175241, + "grad_norm": 0.1332533061504364, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0221, + "step": 69800 + }, + { + "epoch": 6.345209961825122, + "grad_norm": 0.1386847347021103, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0194, + "step": 69810 + }, + { + "epoch": 6.346118887475004, + "grad_norm": 0.12924832105636597, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0211, + "step": 69820 + }, + { + "epoch": 6.347027813124886, + "grad_norm": 0.1963062435388565, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0209, + "step": 69830 + }, + { + "epoch": 6.347936738774768, + "grad_norm": 0.09422840178012848, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0203, + "step": 69840 + }, + { + "epoch": 6.34884566442465, + "grad_norm": 0.17295657098293304, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.0214, + "step": 69850 + }, + { + "epoch": 6.349754590074532, + "grad_norm": 0.16703441739082336, + "learning_rate": 1.218156580082751e-05, + "loss": 0.023, + "step": 69860 + }, + { + "epoch": 6.350663515724413, + "grad_norm": 0.13216781616210938, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0231, + "step": 69870 + }, + { + "epoch": 6.351572441374295, + "grad_norm": 0.1399165540933609, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0212, + "step": 69880 + }, + { + "epoch": 6.352481367024177, + "grad_norm": 0.11200222373008728, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0212, + "step": 69890 + }, + { + "epoch": 6.353390292674059, + "grad_norm": 0.16613644361495972, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0219, + "step": 69900 + }, + { + "epoch": 6.354299218323941, + "grad_norm": 0.10852695256471634, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0198, + "step": 69910 + }, + { + "epoch": 6.355208143973823, + "grad_norm": 0.10220488160848618, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0219, + "step": 69920 + }, + { + "epoch": 6.356117069623704, + "grad_norm": 0.10986530780792236, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0202, + "step": 69930 + }, + { + "epoch": 6.357025995273586, + "grad_norm": 0.09787842631340027, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0187, + "step": 69940 + }, + { + "epoch": 6.357934920923468, + "grad_norm": 0.11035080999135971, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0194, + "step": 69950 + }, + { + "epoch": 6.35884384657335, + "grad_norm": 0.1837126910686493, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0196, + "step": 69960 + }, + { + "epoch": 6.359752772223232, + "grad_norm": 0.13032032549381256, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0189, + "step": 69970 + }, + { + "epoch": 6.360661697873114, + "grad_norm": 0.12276769429445267, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.0215, + "step": 69980 + }, + { + "epoch": 6.361570623522995, + "grad_norm": 0.13861910998821259, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0219, + "step": 69990 + }, + { + "epoch": 6.362479549172877, + "grad_norm": 0.14570844173431396, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0213, + "step": 70000 + }, + { + "epoch": 6.363388474822759, + "grad_norm": 0.12666384875774384, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0196, + "step": 70010 + }, + { + "epoch": 6.364297400472641, + "grad_norm": 0.11008736491203308, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0181, + "step": 70020 + }, + { + "epoch": 6.365206326122523, + "grad_norm": 0.11335267871618271, + "learning_rate": 1.211257613765845e-05, + "loss": 0.02, + "step": 70030 + }, + { + "epoch": 6.366115251772405, + "grad_norm": 0.09896586835384369, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.019, + "step": 70040 + }, + { + "epoch": 6.367024177422287, + "grad_norm": 0.12618756294250488, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0197, + "step": 70050 + }, + { + "epoch": 6.367933103072168, + "grad_norm": 0.129560187458992, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0193, + "step": 70060 + }, + { + "epoch": 6.36884202872205, + "grad_norm": 0.12756092846393585, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.0206, + "step": 70070 + }, + { + "epoch": 6.369750954371932, + "grad_norm": 0.10914931446313858, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.0178, + "step": 70080 + }, + { + "epoch": 6.370659880021814, + "grad_norm": 0.15155945718288422, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.0198, + "step": 70090 + }, + { + "epoch": 6.371568805671696, + "grad_norm": 0.18577535450458527, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0203, + "step": 70100 + }, + { + "epoch": 6.3724777313215775, + "grad_norm": 0.11670383810997009, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0187, + "step": 70110 + }, + { + "epoch": 6.3733866569714595, + "grad_norm": 0.13275857269763947, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0201, + "step": 70120 + }, + { + "epoch": 6.374295582621341, + "grad_norm": 0.12730687856674194, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.018, + "step": 70130 + }, + { + "epoch": 6.375204508271223, + "grad_norm": 0.15976189076900482, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0203, + "step": 70140 + }, + { + "epoch": 6.376113433921105, + "grad_norm": 0.17956268787384033, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0226, + "step": 70150 + }, + { + "epoch": 6.377022359570987, + "grad_norm": 0.14166395366191864, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.021, + "step": 70160 + }, + { + "epoch": 6.377931285220869, + "grad_norm": 0.1402294784784317, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0212, + "step": 70170 + }, + { + "epoch": 6.3788402108707505, + "grad_norm": 0.13302768766880035, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.019, + "step": 70180 + }, + { + "epoch": 6.3797491365206325, + "grad_norm": 0.12037330865859985, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0194, + "step": 70190 + }, + { + "epoch": 6.3806580621705145, + "grad_norm": 0.10496922582387924, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0188, + "step": 70200 + }, + { + "epoch": 6.381566987820396, + "grad_norm": 0.11777421087026596, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.022, + "step": 70210 + }, + { + "epoch": 6.382475913470278, + "grad_norm": 0.138751819729805, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.021, + "step": 70220 + }, + { + "epoch": 6.3833848391201595, + "grad_norm": 0.11233678460121155, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0195, + "step": 70230 + }, + { + "epoch": 6.3842937647700415, + "grad_norm": 0.14240026473999023, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0222, + "step": 70240 + }, + { + "epoch": 6.3852026904199235, + "grad_norm": 0.12347698956727982, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0189, + "step": 70250 + }, + { + "epoch": 6.3861116160698055, + "grad_norm": 0.11018184572458267, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0191, + "step": 70260 + }, + { + "epoch": 6.3870205417196875, + "grad_norm": 0.1708378791809082, + "learning_rate": 1.201528720560666e-05, + "loss": 0.0219, + "step": 70270 + }, + { + "epoch": 6.3879294673695695, + "grad_norm": 0.11050254106521606, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0213, + "step": 70280 + }, + { + "epoch": 6.3888383930194514, + "grad_norm": 0.0993940606713295, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0201, + "step": 70290 + }, + { + "epoch": 6.3897473186693325, + "grad_norm": 0.13868844509124756, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0199, + "step": 70300 + }, + { + "epoch": 6.3906562443192145, + "grad_norm": 0.11424790322780609, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0203, + "step": 70310 + }, + { + "epoch": 6.3915651699690965, + "grad_norm": 0.12551219761371613, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0192, + "step": 70320 + }, + { + "epoch": 6.3924740956189785, + "grad_norm": 0.11739020049571991, + "learning_rate": 1.19909852066355e-05, + "loss": 0.0229, + "step": 70330 + }, + { + "epoch": 6.3933830212688605, + "grad_norm": 0.11925806850194931, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0183, + "step": 70340 + }, + { + "epoch": 6.394291946918742, + "grad_norm": 0.13729475438594818, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.0206, + "step": 70350 + }, + { + "epoch": 6.395200872568624, + "grad_norm": 0.14600370824337006, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0211, + "step": 70360 + }, + { + "epoch": 6.396109798218506, + "grad_norm": 0.15168283879756927, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0203, + "step": 70370 + }, + { + "epoch": 6.3970187238683875, + "grad_norm": 0.10075273364782333, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0195, + "step": 70380 + }, + { + "epoch": 6.3979276495182695, + "grad_norm": 0.13843445479869843, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0206, + "step": 70390 + }, + { + "epoch": 6.3988365751681515, + "grad_norm": 0.11226920038461685, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0228, + "step": 70400 + }, + { + "epoch": 6.3997455008180335, + "grad_norm": 0.14957235753536224, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0193, + "step": 70410 + }, + { + "epoch": 6.400654426467915, + "grad_norm": 0.13442043960094452, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0182, + "step": 70420 + }, + { + "epoch": 6.401563352117797, + "grad_norm": 0.12190115451812744, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0212, + "step": 70430 + }, + { + "epoch": 6.402472277767679, + "grad_norm": 0.11772610992193222, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.0235, + "step": 70440 + }, + { + "epoch": 6.403381203417561, + "grad_norm": 0.12886592745780945, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0193, + "step": 70450 + }, + { + "epoch": 6.4042901290674425, + "grad_norm": 0.10872625559568405, + "learning_rate": 1.193835918571248e-05, + "loss": 0.0217, + "step": 70460 + }, + { + "epoch": 6.4051990547173245, + "grad_norm": 0.1304250955581665, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.0213, + "step": 70470 + }, + { + "epoch": 6.406107980367206, + "grad_norm": 0.20432715117931366, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0205, + "step": 70480 + }, + { + "epoch": 6.407016906017088, + "grad_norm": 0.14862202107906342, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0179, + "step": 70490 + }, + { + "epoch": 6.40792583166697, + "grad_norm": 0.1292875111103058, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0193, + "step": 70500 + }, + { + "epoch": 6.408834757316852, + "grad_norm": 0.16367100179195404, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.0196, + "step": 70510 + }, + { + "epoch": 6.409743682966734, + "grad_norm": 0.11654018610715866, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0172, + "step": 70520 + }, + { + "epoch": 6.410652608616616, + "grad_norm": 0.10060899704694748, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.0203, + "step": 70530 + }, + { + "epoch": 6.411561534266497, + "grad_norm": 0.20108620822429657, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0206, + "step": 70540 + }, + { + "epoch": 6.412470459916379, + "grad_norm": 0.10315526276826859, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.0204, + "step": 70550 + }, + { + "epoch": 6.413379385566261, + "grad_norm": 0.17797188460826874, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0224, + "step": 70560 + }, + { + "epoch": 6.414288311216143, + "grad_norm": 0.16577164828777313, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0211, + "step": 70570 + }, + { + "epoch": 6.415197236866025, + "grad_norm": 0.1328299641609192, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0201, + "step": 70580 + }, + { + "epoch": 6.416106162515907, + "grad_norm": 0.10828804969787598, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0206, + "step": 70590 + }, + { + "epoch": 6.417015088165788, + "grad_norm": 0.14322030544281006, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.02, + "step": 70600 + }, + { + "epoch": 6.41792401381567, + "grad_norm": 0.1256575733423233, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0223, + "step": 70610 + }, + { + "epoch": 6.418832939465552, + "grad_norm": 0.11467579007148743, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0211, + "step": 70620 + }, + { + "epoch": 6.419741865115434, + "grad_norm": 0.11810275912284851, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0186, + "step": 70630 + }, + { + "epoch": 6.420650790765316, + "grad_norm": 0.17864470183849335, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0198, + "step": 70640 + }, + { + "epoch": 6.421559716415198, + "grad_norm": 0.13611853122711182, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.0178, + "step": 70650 + }, + { + "epoch": 6.422468642065079, + "grad_norm": 0.1595030277967453, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.024, + "step": 70660 + }, + { + "epoch": 6.423377567714961, + "grad_norm": 0.11042408645153046, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0213, + "step": 70670 + }, + { + "epoch": 6.424286493364843, + "grad_norm": 0.12194974720478058, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.02, + "step": 70680 + }, + { + "epoch": 6.425195419014725, + "grad_norm": 0.12507879734039307, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0197, + "step": 70690 + }, + { + "epoch": 6.426104344664607, + "grad_norm": 0.14448769390583038, + "learning_rate": 1.184130718987356e-05, + "loss": 0.0185, + "step": 70700 + }, + { + "epoch": 6.427013270314489, + "grad_norm": 0.15389801561832428, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0217, + "step": 70710 + }, + { + "epoch": 6.42792219596437, + "grad_norm": 0.1246585100889206, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0222, + "step": 70720 + }, + { + "epoch": 6.428831121614252, + "grad_norm": 0.12177033722400665, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0203, + "step": 70730 + }, + { + "epoch": 6.429740047264134, + "grad_norm": 0.1402529776096344, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0209, + "step": 70740 + }, + { + "epoch": 6.430648972914016, + "grad_norm": 0.13204823434352875, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0185, + "step": 70750 + }, + { + "epoch": 6.431557898563898, + "grad_norm": 0.1593298763036728, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0187, + "step": 70760 + }, + { + "epoch": 6.43246682421378, + "grad_norm": 0.11077041178941727, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0194, + "step": 70770 + }, + { + "epoch": 6.433375749863661, + "grad_norm": 0.1488598734140396, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0201, + "step": 70780 + }, + { + "epoch": 6.434284675513543, + "grad_norm": 0.1077876165509224, + "learning_rate": 1.180494808593063e-05, + "loss": 0.0195, + "step": 70790 + }, + { + "epoch": 6.435193601163425, + "grad_norm": 0.19947104156017303, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0227, + "step": 70800 + }, + { + "epoch": 6.436102526813307, + "grad_norm": 0.1604989618062973, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0222, + "step": 70810 + }, + { + "epoch": 6.437011452463189, + "grad_norm": 0.12624061107635498, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0189, + "step": 70820 + }, + { + "epoch": 6.437920378113071, + "grad_norm": 0.1639634072780609, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0191, + "step": 70830 + }, + { + "epoch": 6.438829303762952, + "grad_norm": 0.1252899467945099, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0209, + "step": 70840 + }, + { + "epoch": 6.439738229412834, + "grad_norm": 0.11543361097574234, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0199, + "step": 70850 + }, + { + "epoch": 6.440647155062716, + "grad_norm": 0.09710332006216049, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0197, + "step": 70860 + }, + { + "epoch": 6.441556080712598, + "grad_norm": 0.17471802234649658, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.0216, + "step": 70870 + }, + { + "epoch": 6.44246500636248, + "grad_norm": 0.14141270518302917, + "learning_rate": 1.176860863606944e-05, + "loss": 0.0191, + "step": 70880 + }, + { + "epoch": 6.443373932012362, + "grad_norm": 0.09763350337743759, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.0188, + "step": 70890 + }, + { + "epoch": 6.444282857662243, + "grad_norm": 0.22211231291294098, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0197, + "step": 70900 + }, + { + "epoch": 6.445191783312125, + "grad_norm": 0.10212617367506027, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0207, + "step": 70910 + }, + { + "epoch": 6.446100708962007, + "grad_norm": 0.12891335785388947, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0205, + "step": 70920 + }, + { + "epoch": 6.447009634611889, + "grad_norm": 0.1321186125278473, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0204, + "step": 70930 + }, + { + "epoch": 6.447918560261771, + "grad_norm": 0.10025005042552948, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0208, + "step": 70940 + }, + { + "epoch": 6.448827485911653, + "grad_norm": 0.12685364484786987, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.0195, + "step": 70950 + }, + { + "epoch": 6.449736411561534, + "grad_norm": 0.1399812549352646, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.022, + "step": 70960 + }, + { + "epoch": 6.450645337211416, + "grad_norm": 0.13960866630077362, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0189, + "step": 70970 + }, + { + "epoch": 6.451554262861298, + "grad_norm": 0.12892703711986542, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.022, + "step": 70980 + }, + { + "epoch": 6.45246318851118, + "grad_norm": 0.20723572373390198, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0182, + "step": 70990 + }, + { + "epoch": 6.453372114161062, + "grad_norm": 0.12039665132761002, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0219, + "step": 71000 + }, + { + "epoch": 6.454281039810944, + "grad_norm": 0.15634070336818695, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0193, + "step": 71010 + }, + { + "epoch": 6.455189965460825, + "grad_norm": 0.13735158741474152, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0193, + "step": 71020 + }, + { + "epoch": 6.456098891110707, + "grad_norm": 0.10271202027797699, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0199, + "step": 71030 + }, + { + "epoch": 6.457007816760589, + "grad_norm": 0.13363510370254517, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.0195, + "step": 71040 + }, + { + "epoch": 6.457916742410471, + "grad_norm": 0.20417095720767975, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0207, + "step": 71050 + }, + { + "epoch": 6.458825668060353, + "grad_norm": 0.09495747089385986, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0198, + "step": 71060 + }, + { + "epoch": 6.459734593710235, + "grad_norm": 0.15136562287807465, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0199, + "step": 71070 + }, + { + "epoch": 6.460643519360116, + "grad_norm": 0.13565844297409058, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.0202, + "step": 71080 + }, + { + "epoch": 6.461552445009998, + "grad_norm": 0.13949233293533325, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0213, + "step": 71090 + }, + { + "epoch": 6.46246137065988, + "grad_norm": 0.22225582599639893, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0217, + "step": 71100 + }, + { + "epoch": 6.463370296309762, + "grad_norm": 0.13055028021335602, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0223, + "step": 71110 + }, + { + "epoch": 6.464279221959644, + "grad_norm": 0.15382327139377594, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.0197, + "step": 71120 + }, + { + "epoch": 6.465188147609526, + "grad_norm": 0.1394142061471939, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0215, + "step": 71130 + }, + { + "epoch": 6.466097073259407, + "grad_norm": 0.07938014715909958, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.0193, + "step": 71140 + }, + { + "epoch": 6.467005998909289, + "grad_norm": 0.1253560185432434, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0205, + "step": 71150 + }, + { + "epoch": 6.467914924559171, + "grad_norm": 0.13531331717967987, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0191, + "step": 71160 + }, + { + "epoch": 6.468823850209053, + "grad_norm": 0.11469744890928268, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0202, + "step": 71170 + }, + { + "epoch": 6.469732775858935, + "grad_norm": 0.13650397956371307, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.0228, + "step": 71180 + }, + { + "epoch": 6.470641701508817, + "grad_norm": 0.258407324552536, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0201, + "step": 71190 + }, + { + "epoch": 6.471550627158699, + "grad_norm": 0.12628839910030365, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0194, + "step": 71200 + }, + { + "epoch": 6.47245955280858, + "grad_norm": 0.13268326222896576, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0212, + "step": 71210 + }, + { + "epoch": 6.473368478458462, + "grad_norm": 0.2074037492275238, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0196, + "step": 71220 + }, + { + "epoch": 6.474277404108344, + "grad_norm": 0.11643961817026138, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0213, + "step": 71230 + }, + { + "epoch": 6.475186329758226, + "grad_norm": 0.13763481378555298, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.019, + "step": 71240 + }, + { + "epoch": 6.476095255408108, + "grad_norm": 0.14689359068870544, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.0207, + "step": 71250 + }, + { + "epoch": 6.477004181057989, + "grad_norm": 0.12405047565698624, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0214, + "step": 71260 + }, + { + "epoch": 6.477913106707871, + "grad_norm": 0.20865102112293243, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0195, + "step": 71270 + }, + { + "epoch": 6.478822032357753, + "grad_norm": 0.1835489571094513, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0187, + "step": 71280 + }, + { + "epoch": 6.479730958007635, + "grad_norm": 0.12661632895469666, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0193, + "step": 71290 + }, + { + "epoch": 6.480639883657517, + "grad_norm": 0.1235903725028038, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0186, + "step": 71300 + }, + { + "epoch": 6.481548809307399, + "grad_norm": 0.14632189273834229, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0208, + "step": 71310 + }, + { + "epoch": 6.482457734957281, + "grad_norm": 0.10155174136161804, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0198, + "step": 71320 + }, + { + "epoch": 6.483366660607162, + "grad_norm": 0.12327612936496735, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0174, + "step": 71330 + }, + { + "epoch": 6.484275586257044, + "grad_norm": 0.11827461421489716, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0195, + "step": 71340 + }, + { + "epoch": 6.485184511906926, + "grad_norm": 0.15844738483428955, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0207, + "step": 71350 + }, + { + "epoch": 6.486093437556808, + "grad_norm": 0.15186160802841187, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0198, + "step": 71360 + }, + { + "epoch": 6.48700236320669, + "grad_norm": 0.2491445243358612, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0211, + "step": 71370 + }, + { + "epoch": 6.487911288856571, + "grad_norm": 0.1287083476781845, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.0211, + "step": 71380 + }, + { + "epoch": 6.488820214506453, + "grad_norm": 0.10030622035264969, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.0189, + "step": 71390 + }, + { + "epoch": 6.489729140156335, + "grad_norm": 0.1681828498840332, + "learning_rate": 1.155904354214324e-05, + "loss": 0.0222, + "step": 71400 + }, + { + "epoch": 6.490638065806217, + "grad_norm": 0.15804734826087952, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0222, + "step": 71410 + }, + { + "epoch": 6.491546991456099, + "grad_norm": 0.11835003644227982, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0198, + "step": 71420 + }, + { + "epoch": 6.492455917105981, + "grad_norm": 0.12002670764923096, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0207, + "step": 71430 + }, + { + "epoch": 6.493364842755863, + "grad_norm": 0.1365475356578827, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0167, + "step": 71440 + }, + { + "epoch": 6.494273768405744, + "grad_norm": 0.12774688005447388, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0213, + "step": 71450 + }, + { + "epoch": 6.495182694055626, + "grad_norm": 0.1701701134443283, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0215, + "step": 71460 + }, + { + "epoch": 6.496091619705508, + "grad_norm": 0.10470354557037354, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.0172, + "step": 71470 + }, + { + "epoch": 6.49700054535539, + "grad_norm": 0.2494218796491623, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0211, + "step": 71480 + }, + { + "epoch": 6.497909471005272, + "grad_norm": 0.19739560782909393, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.022, + "step": 71490 + }, + { + "epoch": 6.498818396655153, + "grad_norm": 0.2221386432647705, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0179, + "step": 71500 + }, + { + "epoch": 6.499727322305035, + "grad_norm": 0.16874925792217255, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0197, + "step": 71510 + }, + { + "epoch": 6.500636247954917, + "grad_norm": 0.15616829693317413, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0195, + "step": 71520 + }, + { + "epoch": 6.501545173604799, + "grad_norm": 0.09581371396780014, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0191, + "step": 71530 + }, + { + "epoch": 6.502454099254681, + "grad_norm": 0.13659682869911194, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0197, + "step": 71540 + }, + { + "epoch": 6.503363024904563, + "grad_norm": 0.11519958078861237, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0183, + "step": 71550 + }, + { + "epoch": 6.504271950554445, + "grad_norm": 0.12910477817058563, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0207, + "step": 71560 + }, + { + "epoch": 6.505180876204326, + "grad_norm": 0.11649836599826813, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0175, + "step": 71570 + }, + { + "epoch": 6.506089801854208, + "grad_norm": 0.11885973066091537, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0214, + "step": 71580 + }, + { + "epoch": 6.50699872750409, + "grad_norm": 0.12154700607061386, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0197, + "step": 71590 + }, + { + "epoch": 6.507907653153972, + "grad_norm": 0.13076943159103394, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.018, + "step": 71600 + }, + { + "epoch": 6.508816578803854, + "grad_norm": 0.13344629108905792, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0181, + "step": 71610 + }, + { + "epoch": 6.509725504453735, + "grad_norm": 0.1442292332649231, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0186, + "step": 71620 + }, + { + "epoch": 6.510634430103617, + "grad_norm": 0.12070095539093018, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0213, + "step": 71630 + }, + { + "epoch": 6.511543355753499, + "grad_norm": 0.12823504209518433, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0203, + "step": 71640 + }, + { + "epoch": 6.512452281403381, + "grad_norm": 0.1121855154633522, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0187, + "step": 71650 + }, + { + "epoch": 6.513361207053263, + "grad_norm": 0.1388688087463379, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0214, + "step": 71660 + }, + { + "epoch": 6.514270132703145, + "grad_norm": 0.15535464882850647, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0192, + "step": 71670 + }, + { + "epoch": 6.515179058353027, + "grad_norm": 0.1041565015912056, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.0195, + "step": 71680 + }, + { + "epoch": 6.516087984002908, + "grad_norm": 0.15023162961006165, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.019, + "step": 71690 + }, + { + "epoch": 6.51699690965279, + "grad_norm": 0.1252824068069458, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0189, + "step": 71700 + }, + { + "epoch": 6.517905835302672, + "grad_norm": 0.1423046588897705, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0225, + "step": 71710 + }, + { + "epoch": 6.518814760952554, + "grad_norm": 0.11688801646232605, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0193, + "step": 71720 + }, + { + "epoch": 6.519723686602436, + "grad_norm": 0.10624101012945175, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0223, + "step": 71730 + }, + { + "epoch": 6.520632612252317, + "grad_norm": 0.13785739243030548, + "learning_rate": 1.142239957298505e-05, + "loss": 0.02, + "step": 71740 + }, + { + "epoch": 6.521541537902199, + "grad_norm": 0.09934176504611969, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0174, + "step": 71750 + }, + { + "epoch": 6.522450463552081, + "grad_norm": 0.11639048159122467, + "learning_rate": 1.141437135633602e-05, + "loss": 0.0179, + "step": 71760 + }, + { + "epoch": 6.523359389201963, + "grad_norm": 0.1263504922389984, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0195, + "step": 71770 + }, + { + "epoch": 6.524268314851845, + "grad_norm": 0.126254141330719, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0221, + "step": 71780 + }, + { + "epoch": 6.525177240501727, + "grad_norm": 0.11752812564373016, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0184, + "step": 71790 + }, + { + "epoch": 6.526086166151609, + "grad_norm": 0.11944970488548279, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0183, + "step": 71800 + }, + { + "epoch": 6.52699509180149, + "grad_norm": 0.1695803552865982, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0186, + "step": 71810 + }, + { + "epoch": 6.527904017451372, + "grad_norm": 0.1067604273557663, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0199, + "step": 71820 + }, + { + "epoch": 6.528812943101254, + "grad_norm": 0.18000949919223785, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0183, + "step": 71830 + }, + { + "epoch": 6.529721868751136, + "grad_norm": 0.10036353766918182, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0236, + "step": 71840 + }, + { + "epoch": 6.530630794401018, + "grad_norm": 0.16118408739566803, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.021, + "step": 71850 + }, + { + "epoch": 6.5315397200509, + "grad_norm": 0.1329103410243988, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.02, + "step": 71860 + }, + { + "epoch": 6.532448645700781, + "grad_norm": 0.15258827805519104, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.021, + "step": 71870 + }, + { + "epoch": 6.533357571350663, + "grad_norm": 0.166982039809227, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0182, + "step": 71880 + }, + { + "epoch": 6.534266497000545, + "grad_norm": 0.1192001923918724, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.0212, + "step": 71890 + }, + { + "epoch": 6.535175422650427, + "grad_norm": 0.11678697168827057, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0202, + "step": 71900 + }, + { + "epoch": 6.536084348300309, + "grad_norm": 0.13319146633148193, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0191, + "step": 71910 + }, + { + "epoch": 6.536993273950191, + "grad_norm": 0.10711164772510529, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0191, + "step": 71920 + }, + { + "epoch": 6.537902199600072, + "grad_norm": 0.10983732342720032, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0192, + "step": 71930 + }, + { + "epoch": 6.538811125249954, + "grad_norm": 0.16548162698745728, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0201, + "step": 71940 + }, + { + "epoch": 6.539720050899836, + "grad_norm": 0.10528641194105148, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0201, + "step": 71950 + }, + { + "epoch": 6.540628976549718, + "grad_norm": 0.11872503906488419, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0217, + "step": 71960 + }, + { + "epoch": 6.5415379021996, + "grad_norm": 0.1187700480222702, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.022, + "step": 71970 + }, + { + "epoch": 6.542446827849482, + "grad_norm": 0.1223922148346901, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.018, + "step": 71980 + }, + { + "epoch": 6.543355753499363, + "grad_norm": 0.1788889616727829, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0201, + "step": 71990 + }, + { + "epoch": 6.544264679149245, + "grad_norm": 0.11709113419055939, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0196, + "step": 72000 + }, + { + "epoch": 6.545173604799127, + "grad_norm": 0.1044902354478836, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0202, + "step": 72010 + }, + { + "epoch": 6.546082530449009, + "grad_norm": 0.12856434285640717, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.023, + "step": 72020 + }, + { + "epoch": 6.546991456098891, + "grad_norm": 0.12855538725852966, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0207, + "step": 72030 + }, + { + "epoch": 6.547900381748773, + "grad_norm": 0.12207399308681488, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0182, + "step": 72040 + }, + { + "epoch": 6.548809307398654, + "grad_norm": 0.1398884803056717, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.02, + "step": 72050 + }, + { + "epoch": 6.549718233048536, + "grad_norm": 0.14823074638843536, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0192, + "step": 72060 + }, + { + "epoch": 6.550627158698418, + "grad_norm": 0.10407661646604538, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0206, + "step": 72070 + }, + { + "epoch": 6.5515360843483, + "grad_norm": 0.1861853003501892, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0204, + "step": 72080 + }, + { + "epoch": 6.552445009998182, + "grad_norm": 0.10222629457712173, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0202, + "step": 72090 + }, + { + "epoch": 6.553353935648064, + "grad_norm": 0.1400003433227539, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0199, + "step": 72100 + }, + { + "epoch": 6.554262861297945, + "grad_norm": 0.11973198503255844, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0195, + "step": 72110 + }, + { + "epoch": 6.555171786947827, + "grad_norm": 0.11436495929956436, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0212, + "step": 72120 + }, + { + "epoch": 6.556080712597709, + "grad_norm": 0.14369286596775055, + "learning_rate": 1.126604838049735e-05, + "loss": 0.0182, + "step": 72130 + }, + { + "epoch": 6.556989638247591, + "grad_norm": 0.13774728775024414, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.0185, + "step": 72140 + }, + { + "epoch": 6.557898563897473, + "grad_norm": 0.15860389173030853, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0185, + "step": 72150 + }, + { + "epoch": 6.558807489547355, + "grad_norm": 0.14893585443496704, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0193, + "step": 72160 + }, + { + "epoch": 6.5597164151972365, + "grad_norm": 0.1921052485704422, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0205, + "step": 72170 + }, + { + "epoch": 6.5606253408471185, + "grad_norm": 0.13306833803653717, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0183, + "step": 72180 + }, + { + "epoch": 6.5615342664970004, + "grad_norm": 0.15573857724666595, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0202, + "step": 72190 + }, + { + "epoch": 6.562443192146882, + "grad_norm": 0.1632489264011383, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0199, + "step": 72200 + }, + { + "epoch": 6.563352117796764, + "grad_norm": 0.11996224522590637, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0203, + "step": 72210 + }, + { + "epoch": 6.564261043446646, + "grad_norm": 0.11258892714977264, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0197, + "step": 72220 + }, + { + "epoch": 6.565169969096528, + "grad_norm": 0.13435542583465576, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.0197, + "step": 72230 + }, + { + "epoch": 6.5660788947464095, + "grad_norm": 0.1467272937297821, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0227, + "step": 72240 + }, + { + "epoch": 6.5669878203962915, + "grad_norm": 0.113522469997406, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0211, + "step": 72250 + }, + { + "epoch": 6.5678967460461735, + "grad_norm": 0.12018831819295883, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0206, + "step": 72260 + }, + { + "epoch": 6.5688056716960554, + "grad_norm": 0.11001142114400864, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0188, + "step": 72270 + }, + { + "epoch": 6.569714597345937, + "grad_norm": 0.11027465015649796, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0203, + "step": 72280 + }, + { + "epoch": 6.5706235229958185, + "grad_norm": 0.1375676542520523, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0189, + "step": 72290 + }, + { + "epoch": 6.5715324486457005, + "grad_norm": 0.10294167697429657, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0177, + "step": 72300 + }, + { + "epoch": 6.5724413742955825, + "grad_norm": 0.14933232963085175, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0241, + "step": 72310 + }, + { + "epoch": 6.5733502999454645, + "grad_norm": 0.13056018948554993, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0199, + "step": 72320 + }, + { + "epoch": 6.5742592255953465, + "grad_norm": 0.23163259029388428, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0188, + "step": 72330 + }, + { + "epoch": 6.5751681512452285, + "grad_norm": 0.16423869132995605, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0182, + "step": 72340 + }, + { + "epoch": 6.5760770768951105, + "grad_norm": 0.1409207582473755, + "learning_rate": 1.117803938925701e-05, + "loss": 0.0189, + "step": 72350 + }, + { + "epoch": 6.5769860025449915, + "grad_norm": 0.12712320685386658, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0183, + "step": 72360 + }, + { + "epoch": 6.5778949281948735, + "grad_norm": 0.12061209976673126, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0199, + "step": 72370 + }, + { + "epoch": 6.5788038538447555, + "grad_norm": 0.10084524005651474, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0202, + "step": 72380 + }, + { + "epoch": 6.5797127794946375, + "grad_norm": 0.1259927600622177, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0207, + "step": 72390 + }, + { + "epoch": 6.5806217051445195, + "grad_norm": 0.1286095678806305, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0201, + "step": 72400 + }, + { + "epoch": 6.581530630794401, + "grad_norm": 0.1721486747264862, + "learning_rate": 1.115406118227002e-05, + "loss": 0.021, + "step": 72410 + }, + { + "epoch": 6.582439556444283, + "grad_norm": 0.1611398458480835, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0213, + "step": 72420 + }, + { + "epoch": 6.583348482094165, + "grad_norm": 0.1189362108707428, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0206, + "step": 72430 + }, + { + "epoch": 6.5842574077440466, + "grad_norm": 0.14343567192554474, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0209, + "step": 72440 + }, + { + "epoch": 6.5851663333939285, + "grad_norm": 0.13516312837600708, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0187, + "step": 72450 + }, + { + "epoch": 6.5860752590438105, + "grad_norm": 0.1379394680261612, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.0214, + "step": 72460 + }, + { + "epoch": 6.5869841846936925, + "grad_norm": 0.12145747989416122, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.02, + "step": 72470 + }, + { + "epoch": 6.587893110343574, + "grad_norm": 0.14239822328090668, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0186, + "step": 72480 + }, + { + "epoch": 6.588802035993456, + "grad_norm": 0.20239846408367157, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0195, + "step": 72490 + }, + { + "epoch": 6.589710961643338, + "grad_norm": 0.12058307975530624, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0203, + "step": 72500 + }, + { + "epoch": 6.59061988729322, + "grad_norm": 0.14761073887348175, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0184, + "step": 72510 + }, + { + "epoch": 6.591528812943102, + "grad_norm": 0.2146056741476059, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0224, + "step": 72520 + }, + { + "epoch": 6.592437738592983, + "grad_norm": 0.12930914759635925, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.018, + "step": 72530 + }, + { + "epoch": 6.593346664242865, + "grad_norm": 0.10016843676567078, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0185, + "step": 72540 + }, + { + "epoch": 6.594255589892747, + "grad_norm": 0.1393231302499771, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0207, + "step": 72550 + }, + { + "epoch": 6.595164515542629, + "grad_norm": 0.16316024959087372, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.0211, + "step": 72560 + }, + { + "epoch": 6.596073441192511, + "grad_norm": 0.14304150640964508, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0192, + "step": 72570 + }, + { + "epoch": 6.596982366842393, + "grad_norm": 0.12843850255012512, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0221, + "step": 72580 + }, + { + "epoch": 6.597891292492275, + "grad_norm": 0.12041808664798737, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0187, + "step": 72590 + }, + { + "epoch": 6.598800218142156, + "grad_norm": 0.14779675006866455, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.0228, + "step": 72600 + }, + { + "epoch": 6.599709143792038, + "grad_norm": 0.14649741351604462, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0194, + "step": 72610 + }, + { + "epoch": 6.60061806944192, + "grad_norm": 0.14191533625125885, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0185, + "step": 72620 + }, + { + "epoch": 6.601526995091802, + "grad_norm": 0.14886978268623352, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0209, + "step": 72630 + }, + { + "epoch": 6.602435920741684, + "grad_norm": 0.18997620046138763, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0229, + "step": 72640 + }, + { + "epoch": 6.603344846391565, + "grad_norm": 0.13536331057548523, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0204, + "step": 72650 + }, + { + "epoch": 6.604253772041447, + "grad_norm": 0.17086653411388397, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0221, + "step": 72660 + }, + { + "epoch": 6.605162697691329, + "grad_norm": 0.1683153659105301, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0211, + "step": 72670 + }, + { + "epoch": 6.606071623341211, + "grad_norm": 0.12886951863765717, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0191, + "step": 72680 + }, + { + "epoch": 6.606980548991093, + "grad_norm": 0.09595250338315964, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0205, + "step": 72690 + }, + { + "epoch": 6.607889474640975, + "grad_norm": 0.10812563449144363, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0179, + "step": 72700 + }, + { + "epoch": 6.608798400290857, + "grad_norm": 0.12983591854572296, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0172, + "step": 72710 + }, + { + "epoch": 6.609707325940738, + "grad_norm": 0.12148313969373703, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0215, + "step": 72720 + }, + { + "epoch": 6.61061625159062, + "grad_norm": 0.12590688467025757, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0184, + "step": 72730 + }, + { + "epoch": 6.611525177240502, + "grad_norm": 0.1254136860370636, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0184, + "step": 72740 + }, + { + "epoch": 6.612434102890384, + "grad_norm": 0.14205779135227203, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0208, + "step": 72750 + }, + { + "epoch": 6.613343028540266, + "grad_norm": 0.13341659307479858, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0186, + "step": 72760 + }, + { + "epoch": 6.614251954190147, + "grad_norm": 0.12585406005382538, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0216, + "step": 72770 + }, + { + "epoch": 6.615160879840029, + "grad_norm": 0.1310882270336151, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0218, + "step": 72780 + }, + { + "epoch": 6.616069805489911, + "grad_norm": 0.10205511003732681, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0221, + "step": 72790 + }, + { + "epoch": 6.616978731139793, + "grad_norm": 0.09610167145729065, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0189, + "step": 72800 + }, + { + "epoch": 6.617887656789675, + "grad_norm": 0.12108419835567474, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0219, + "step": 72810 + }, + { + "epoch": 6.618796582439557, + "grad_norm": 0.15246470272541046, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0198, + "step": 72820 + }, + { + "epoch": 6.619705508089439, + "grad_norm": 0.14432945847511292, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0193, + "step": 72830 + }, + { + "epoch": 6.62061443373932, + "grad_norm": 0.14923956990242004, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.0186, + "step": 72840 + }, + { + "epoch": 6.621523359389202, + "grad_norm": 0.1410226672887802, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0191, + "step": 72850 + }, + { + "epoch": 6.622432285039084, + "grad_norm": 0.12279423326253891, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0201, + "step": 72860 + }, + { + "epoch": 6.623341210688966, + "grad_norm": 0.10747867822647095, + "learning_rate": 1.097058241668254e-05, + "loss": 0.0182, + "step": 72870 + }, + { + "epoch": 6.624250136338848, + "grad_norm": 0.14662481844425201, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0197, + "step": 72880 + }, + { + "epoch": 6.625159061988729, + "grad_norm": 0.13544924557209015, + "learning_rate": 1.09626195543497e-05, + "loss": 0.0206, + "step": 72890 + }, + { + "epoch": 6.626067987638611, + "grad_norm": 0.11711770296096802, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0202, + "step": 72900 + }, + { + "epoch": 6.626976913288493, + "grad_norm": 0.1233973503112793, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0196, + "step": 72910 + }, + { + "epoch": 6.627885838938375, + "grad_norm": 0.18857519328594208, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0223, + "step": 72920 + }, + { + "epoch": 6.628794764588257, + "grad_norm": 0.11345981806516647, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0187, + "step": 72930 + }, + { + "epoch": 6.629703690238139, + "grad_norm": 0.10657567530870438, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0203, + "step": 72940 + }, + { + "epoch": 6.630612615888021, + "grad_norm": 0.13364429771900177, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.0163, + "step": 72950 + }, + { + "epoch": 6.631521541537902, + "grad_norm": 0.15502865612506866, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0222, + "step": 72960 + }, + { + "epoch": 6.632430467187784, + "grad_norm": 0.12607106566429138, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0201, + "step": 72970 + }, + { + "epoch": 6.633339392837666, + "grad_norm": 0.13356105983257294, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0201, + "step": 72980 + }, + { + "epoch": 6.634248318487548, + "grad_norm": 0.11795342713594437, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0208, + "step": 72990 + }, + { + "epoch": 6.63515724413743, + "grad_norm": 0.10950755327939987, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.0201, + "step": 73000 + }, + { + "epoch": 6.636066169787312, + "grad_norm": 0.11981919407844543, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.0191, + "step": 73010 + }, + { + "epoch": 6.636975095437193, + "grad_norm": 0.16036377847194672, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0185, + "step": 73020 + }, + { + "epoch": 6.637884021087075, + "grad_norm": 0.115991972386837, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0189, + "step": 73030 + }, + { + "epoch": 6.638792946736957, + "grad_norm": 0.14146728813648224, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0197, + "step": 73040 + }, + { + "epoch": 6.639701872386839, + "grad_norm": 0.13047021627426147, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0203, + "step": 73050 + }, + { + "epoch": 6.640610798036721, + "grad_norm": 0.1799040585756302, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0219, + "step": 73060 + }, + { + "epoch": 6.641519723686603, + "grad_norm": 0.1189151331782341, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0194, + "step": 73070 + }, + { + "epoch": 6.642428649336484, + "grad_norm": 0.18343161046504974, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0203, + "step": 73080 + }, + { + "epoch": 6.643337574986366, + "grad_norm": 0.14137178659439087, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0203, + "step": 73090 + }, + { + "epoch": 6.644246500636248, + "grad_norm": 0.13205718994140625, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0199, + "step": 73100 + }, + { + "epoch": 6.64515542628613, + "grad_norm": 0.14472486078739166, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.0209, + "step": 73110 + }, + { + "epoch": 6.646064351936012, + "grad_norm": 0.21381309628486633, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0262, + "step": 73120 + }, + { + "epoch": 6.646973277585894, + "grad_norm": 0.1395048052072525, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0179, + "step": 73130 + }, + { + "epoch": 6.647882203235775, + "grad_norm": 0.1584019809961319, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.0231, + "step": 73140 + }, + { + "epoch": 6.648791128885657, + "grad_norm": 0.1762332320213318, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0197, + "step": 73150 + }, + { + "epoch": 6.649700054535539, + "grad_norm": 0.08315537124872208, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.0199, + "step": 73160 + }, + { + "epoch": 6.650608980185421, + "grad_norm": 0.11477385461330414, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0241, + "step": 73170 + }, + { + "epoch": 6.651517905835303, + "grad_norm": 0.12834836542606354, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0171, + "step": 73180 + }, + { + "epoch": 6.652426831485185, + "grad_norm": 0.11934128403663635, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0206, + "step": 73190 + }, + { + "epoch": 6.653335757135066, + "grad_norm": 0.1571062058210373, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0199, + "step": 73200 + }, + { + "epoch": 6.654244682784948, + "grad_norm": 0.1395409256219864, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0217, + "step": 73210 + }, + { + "epoch": 6.65515360843483, + "grad_norm": 0.11618360131978989, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0211, + "step": 73220 + }, + { + "epoch": 6.656062534084712, + "grad_norm": 0.13292306661605835, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.0193, + "step": 73230 + }, + { + "epoch": 6.656971459734594, + "grad_norm": 0.15760572254657745, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0186, + "step": 73240 + }, + { + "epoch": 6.657880385384476, + "grad_norm": 0.19795596599578857, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0174, + "step": 73250 + }, + { + "epoch": 6.658789311034357, + "grad_norm": 0.12501060962677002, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0173, + "step": 73260 + }, + { + "epoch": 6.659698236684239, + "grad_norm": 0.12711039185523987, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.02, + "step": 73270 + }, + { + "epoch": 6.660607162334121, + "grad_norm": 0.12401945888996124, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0214, + "step": 73280 + }, + { + "epoch": 6.661516087984003, + "grad_norm": 0.13853077590465546, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0182, + "step": 73290 + }, + { + "epoch": 6.662425013633885, + "grad_norm": 0.1554444134235382, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0187, + "step": 73300 + }, + { + "epoch": 6.663333939283767, + "grad_norm": 0.12478482723236084, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0187, + "step": 73310 + }, + { + "epoch": 6.664242864933648, + "grad_norm": 0.11671537160873413, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0199, + "step": 73320 + }, + { + "epoch": 6.66515179058353, + "grad_norm": 0.14248310029506683, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.0203, + "step": 73330 + }, + { + "epoch": 6.666060716233412, + "grad_norm": 0.11422029882669449, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.019, + "step": 73340 + }, + { + "epoch": 6.666969641883294, + "grad_norm": 0.11338848620653152, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0182, + "step": 73350 + }, + { + "epoch": 6.667878567533176, + "grad_norm": 0.1288190484046936, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0209, + "step": 73360 + }, + { + "epoch": 6.668787493183058, + "grad_norm": 0.12209183722734451, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0212, + "step": 73370 + }, + { + "epoch": 6.66969641883294, + "grad_norm": 0.11719200760126114, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.0199, + "step": 73380 + }, + { + "epoch": 6.670605344482821, + "grad_norm": 0.12732775509357452, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0194, + "step": 73390 + }, + { + "epoch": 6.671514270132703, + "grad_norm": 0.12015826255083084, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0194, + "step": 73400 + }, + { + "epoch": 6.672423195782585, + "grad_norm": 0.13336607813835144, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0189, + "step": 73410 + }, + { + "epoch": 6.673332121432467, + "grad_norm": 0.1370534747838974, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.018, + "step": 73420 + }, + { + "epoch": 6.674241047082349, + "grad_norm": 0.1722218245267868, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0203, + "step": 73430 + }, + { + "epoch": 6.67514997273223, + "grad_norm": 0.15177467465400696, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0217, + "step": 73440 + }, + { + "epoch": 6.676058898382112, + "grad_norm": 0.18902705609798431, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0201, + "step": 73450 + }, + { + "epoch": 6.676967824031994, + "grad_norm": 0.16931378841400146, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0198, + "step": 73460 + }, + { + "epoch": 6.677876749681876, + "grad_norm": 0.11203522980213165, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.018, + "step": 73470 + }, + { + "epoch": 6.678785675331758, + "grad_norm": 0.15245391428470612, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0194, + "step": 73480 + }, + { + "epoch": 6.67969460098164, + "grad_norm": 0.1824895441532135, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.021, + "step": 73490 + }, + { + "epoch": 6.680603526631522, + "grad_norm": 0.10753290355205536, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0186, + "step": 73500 + }, + { + "epoch": 6.681512452281403, + "grad_norm": 0.2063184529542923, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0201, + "step": 73510 + }, + { + "epoch": 6.682421377931285, + "grad_norm": 0.22240163385868073, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0222, + "step": 73520 + }, + { + "epoch": 6.683330303581167, + "grad_norm": 0.10735537111759186, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0199, + "step": 73530 + }, + { + "epoch": 6.684239229231049, + "grad_norm": 0.09936030209064484, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0192, + "step": 73540 + }, + { + "epoch": 6.685148154880931, + "grad_norm": 0.19182012975215912, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0173, + "step": 73550 + }, + { + "epoch": 6.686057080530812, + "grad_norm": 0.13663600385189056, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0204, + "step": 73560 + }, + { + "epoch": 6.686966006180694, + "grad_norm": 0.11726027727127075, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0196, + "step": 73570 + }, + { + "epoch": 6.687874931830576, + "grad_norm": 0.11964274197816849, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.019, + "step": 73580 + }, + { + "epoch": 6.688783857480458, + "grad_norm": 0.1692415475845337, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0197, + "step": 73590 + }, + { + "epoch": 6.68969278313034, + "grad_norm": 0.15981055796146393, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0201, + "step": 73600 + }, + { + "epoch": 6.690601708780222, + "grad_norm": 0.09772444516420364, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.0189, + "step": 73610 + }, + { + "epoch": 6.691510634430104, + "grad_norm": 0.1839902400970459, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.021, + "step": 73620 + }, + { + "epoch": 6.692419560079985, + "grad_norm": 0.14726226031780243, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0198, + "step": 73630 + }, + { + "epoch": 6.693328485729867, + "grad_norm": 0.15981507301330566, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.0231, + "step": 73640 + }, + { + "epoch": 6.694237411379749, + "grad_norm": 0.11945784091949463, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0202, + "step": 73650 + }, + { + "epoch": 6.695146337029631, + "grad_norm": 0.12980161607265472, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0186, + "step": 73660 + }, + { + "epoch": 6.696055262679513, + "grad_norm": 0.12364175915718079, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0183, + "step": 73670 + }, + { + "epoch": 6.696964188329394, + "grad_norm": 0.12958186864852905, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0199, + "step": 73680 + }, + { + "epoch": 6.697873113979276, + "grad_norm": 0.15633848309516907, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0188, + "step": 73690 + }, + { + "epoch": 6.698782039629158, + "grad_norm": 0.10736706852912903, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0198, + "step": 73700 + }, + { + "epoch": 6.69969096527904, + "grad_norm": 0.11934308707714081, + "learning_rate": 1.063722589416218e-05, + "loss": 0.0206, + "step": 73710 + }, + { + "epoch": 6.700599890928922, + "grad_norm": 0.18432244658470154, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0186, + "step": 73720 + }, + { + "epoch": 6.701508816578804, + "grad_norm": 0.09507063031196594, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.019, + "step": 73730 + }, + { + "epoch": 6.702417742228686, + "grad_norm": 0.11731616407632828, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0206, + "step": 73740 + }, + { + "epoch": 6.703326667878567, + "grad_norm": 0.09649772942066193, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0186, + "step": 73750 + }, + { + "epoch": 6.704235593528449, + "grad_norm": 0.12094147503376007, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0178, + "step": 73760 + }, + { + "epoch": 6.705144519178331, + "grad_norm": 0.10822650045156479, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0192, + "step": 73770 + }, + { + "epoch": 6.706053444828213, + "grad_norm": 0.19321750104427338, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.02, + "step": 73780 + }, + { + "epoch": 6.706962370478095, + "grad_norm": 0.1271815001964569, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0212, + "step": 73790 + }, + { + "epoch": 6.707871296127976, + "grad_norm": 0.1281682401895523, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.0195, + "step": 73800 + }, + { + "epoch": 6.708780221777858, + "grad_norm": 0.10566423088312149, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0199, + "step": 73810 + }, + { + "epoch": 6.70968914742774, + "grad_norm": 0.12378326058387756, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0165, + "step": 73820 + }, + { + "epoch": 6.710598073077622, + "grad_norm": 0.13678255677223206, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0201, + "step": 73830 + }, + { + "epoch": 6.711506998727504, + "grad_norm": 0.1465856432914734, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.0175, + "step": 73840 + }, + { + "epoch": 6.712415924377386, + "grad_norm": 0.15662764012813568, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0177, + "step": 73850 + }, + { + "epoch": 6.713324850027268, + "grad_norm": 0.11864914000034332, + "learning_rate": 1.05779388131875e-05, + "loss": 0.017, + "step": 73860 + }, + { + "epoch": 6.714233775677149, + "grad_norm": 0.12443345785140991, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0172, + "step": 73870 + }, + { + "epoch": 6.715142701327031, + "grad_norm": 0.13679206371307373, + "learning_rate": 1.057003955600053e-05, + "loss": 0.0179, + "step": 73880 + }, + { + "epoch": 6.716051626976913, + "grad_norm": 0.08679886162281036, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0183, + "step": 73890 + }, + { + "epoch": 6.716960552626795, + "grad_norm": 0.109830342233181, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0207, + "step": 73900 + }, + { + "epoch": 6.717869478276677, + "grad_norm": 0.10891077667474747, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0195, + "step": 73910 + }, + { + "epoch": 6.718778403926558, + "grad_norm": 0.13199929893016815, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0181, + "step": 73920 + }, + { + "epoch": 6.71968732957644, + "grad_norm": 0.12819363176822662, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0186, + "step": 73930 + }, + { + "epoch": 6.720596255226322, + "grad_norm": 0.11910054832696915, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0171, + "step": 73940 + }, + { + "epoch": 6.721505180876204, + "grad_norm": 0.11535884439945221, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0193, + "step": 73950 + }, + { + "epoch": 6.722414106526086, + "grad_norm": 0.20901337265968323, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0206, + "step": 73960 + }, + { + "epoch": 6.723323032175968, + "grad_norm": 0.13545066118240356, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0199, + "step": 73970 + }, + { + "epoch": 6.72423195782585, + "grad_norm": 0.13708874583244324, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0205, + "step": 73980 + }, + { + "epoch": 6.725140883475731, + "grad_norm": 0.12062951177358627, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0181, + "step": 73990 + }, + { + "epoch": 6.726049809125613, + "grad_norm": 0.12907247245311737, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0212, + "step": 74000 + }, + { + "epoch": 6.726958734775495, + "grad_norm": 0.1946709156036377, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0177, + "step": 74010 + }, + { + "epoch": 6.727867660425377, + "grad_norm": 0.15729939937591553, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0208, + "step": 74020 + }, + { + "epoch": 6.728776586075259, + "grad_norm": 0.09701287001371384, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0192, + "step": 74030 + }, + { + "epoch": 6.72968551172514, + "grad_norm": 0.09201730787754059, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0178, + "step": 74040 + }, + { + "epoch": 6.730594437375022, + "grad_norm": 0.10190533101558685, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0211, + "step": 74050 + }, + { + "epoch": 6.731503363024904, + "grad_norm": 0.1323106735944748, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0199, + "step": 74060 + }, + { + "epoch": 6.732412288674786, + "grad_norm": 0.18964038789272308, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0227, + "step": 74070 + }, + { + "epoch": 6.733321214324668, + "grad_norm": 0.15217505395412445, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0176, + "step": 74080 + }, + { + "epoch": 6.73423013997455, + "grad_norm": 0.10583418607711792, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0175, + "step": 74090 + }, + { + "epoch": 6.735139065624432, + "grad_norm": 0.14794324338436127, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.021, + "step": 74100 + }, + { + "epoch": 6.736047991274313, + "grad_norm": 0.14959470927715302, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0203, + "step": 74110 + }, + { + "epoch": 6.736956916924195, + "grad_norm": 0.12276310473680496, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0188, + "step": 74120 + }, + { + "epoch": 6.737865842574077, + "grad_norm": 0.15854650735855103, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0169, + "step": 74130 + }, + { + "epoch": 6.738774768223959, + "grad_norm": 0.21739649772644043, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0194, + "step": 74140 + }, + { + "epoch": 6.739683693873841, + "grad_norm": 0.13968394696712494, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.0221, + "step": 74150 + }, + { + "epoch": 6.7405926195237225, + "grad_norm": 0.13480103015899658, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0188, + "step": 74160 + }, + { + "epoch": 6.7415015451736044, + "grad_norm": 0.17404451966285706, + "learning_rate": 1.045565276297473e-05, + "loss": 0.02, + "step": 74170 + }, + { + "epoch": 6.742410470823486, + "grad_norm": 0.12005995213985443, + "learning_rate": 1.045171352558859e-05, + "loss": 0.0197, + "step": 74180 + }, + { + "epoch": 6.743319396473368, + "grad_norm": 0.13531318306922913, + "learning_rate": 1.044777463361479e-05, + "loss": 0.0225, + "step": 74190 + }, + { + "epoch": 6.74422832212325, + "grad_norm": 0.10698812454938889, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0179, + "step": 74200 + }, + { + "epoch": 6.745137247773132, + "grad_norm": 0.13382528722286224, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0195, + "step": 74210 + }, + { + "epoch": 6.746046173423014, + "grad_norm": 0.12391777336597443, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0194, + "step": 74220 + }, + { + "epoch": 6.7469550990728955, + "grad_norm": 0.10341914743185043, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0226, + "step": 74230 + }, + { + "epoch": 6.7478640247227775, + "grad_norm": 0.15766578912734985, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0208, + "step": 74240 + }, + { + "epoch": 6.7487729503726595, + "grad_norm": 0.11082720011472702, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.0168, + "step": 74250 + }, + { + "epoch": 6.749681876022541, + "grad_norm": 0.16736958920955658, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0194, + "step": 74260 + }, + { + "epoch": 6.750590801672423, + "grad_norm": 0.15668606758117676, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.0225, + "step": 74270 + }, + { + "epoch": 6.751499727322305, + "grad_norm": 0.10480800271034241, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.0198, + "step": 74280 + }, + { + "epoch": 6.7524086529721865, + "grad_norm": 0.11379745602607727, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0195, + "step": 74290 + }, + { + "epoch": 6.7533175786220685, + "grad_norm": 0.13427381217479706, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0197, + "step": 74300 + }, + { + "epoch": 6.7542265042719505, + "grad_norm": 0.35401347279548645, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.0209, + "step": 74310 + }, + { + "epoch": 6.7551354299218325, + "grad_norm": 0.1643257588148117, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0188, + "step": 74320 + }, + { + "epoch": 6.7560443555717145, + "grad_norm": 0.11075789481401443, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0193, + "step": 74330 + }, + { + "epoch": 6.756953281221596, + "grad_norm": 0.104257732629776, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0192, + "step": 74340 + }, + { + "epoch": 6.7578622068714775, + "grad_norm": 0.1805899739265442, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.02, + "step": 74350 + }, + { + "epoch": 6.7587711325213595, + "grad_norm": 0.14668236672878265, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.017, + "step": 74360 + }, + { + "epoch": 6.7596800581712415, + "grad_norm": 0.10782982409000397, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0183, + "step": 74370 + }, + { + "epoch": 6.7605889838211235, + "grad_norm": 0.1178225576877594, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0194, + "step": 74380 + }, + { + "epoch": 6.7614979094710055, + "grad_norm": 0.15171021223068237, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0228, + "step": 74390 + }, + { + "epoch": 6.7624068351208875, + "grad_norm": 0.12556377053260803, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0176, + "step": 74400 + }, + { + "epoch": 6.763315760770769, + "grad_norm": 0.126704603433609, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0198, + "step": 74410 + }, + { + "epoch": 6.764224686420651, + "grad_norm": 0.11852239072322845, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.019, + "step": 74420 + }, + { + "epoch": 6.7651336120705325, + "grad_norm": 0.11181540042161942, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0213, + "step": 74430 + }, + { + "epoch": 6.7660425377204145, + "grad_norm": 0.14452748000621796, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0193, + "step": 74440 + }, + { + "epoch": 6.7669514633702965, + "grad_norm": 0.12226708233356476, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.0208, + "step": 74450 + }, + { + "epoch": 6.7678603890201785, + "grad_norm": 0.20203132927417755, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0194, + "step": 74460 + }, + { + "epoch": 6.76876931467006, + "grad_norm": 0.09897518903017044, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0175, + "step": 74470 + }, + { + "epoch": 6.769678240319942, + "grad_norm": 0.13270772993564606, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0176, + "step": 74480 + }, + { + "epoch": 6.770587165969824, + "grad_norm": 0.11923039704561234, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0182, + "step": 74490 + }, + { + "epoch": 6.771496091619706, + "grad_norm": 0.10625828057527542, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0182, + "step": 74500 + }, + { + "epoch": 6.7724050172695875, + "grad_norm": 0.10556481778621674, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0209, + "step": 74510 + }, + { + "epoch": 6.7733139429194695, + "grad_norm": 0.14889012277126312, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0199, + "step": 74520 + }, + { + "epoch": 6.7742228685693515, + "grad_norm": 0.11385795474052429, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.0203, + "step": 74530 + }, + { + "epoch": 6.775131794219233, + "grad_norm": 0.10920251905918121, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.0185, + "step": 74540 + }, + { + "epoch": 6.776040719869115, + "grad_norm": 0.12318619340658188, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0198, + "step": 74550 + }, + { + "epoch": 6.776949645518997, + "grad_norm": 0.11403863877058029, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.0219, + "step": 74560 + }, + { + "epoch": 6.777858571168879, + "grad_norm": 0.10553590208292007, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0215, + "step": 74570 + }, + { + "epoch": 6.778767496818761, + "grad_norm": 0.09190023690462112, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0201, + "step": 74580 + }, + { + "epoch": 6.779676422468642, + "grad_norm": 0.11469640582799911, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0197, + "step": 74590 + }, + { + "epoch": 6.780585348118524, + "grad_norm": 0.1221131831407547, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0219, + "step": 74600 + }, + { + "epoch": 6.781494273768406, + "grad_norm": 0.1304321140050888, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0212, + "step": 74610 + }, + { + "epoch": 6.782403199418288, + "grad_norm": 0.09541915357112885, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0181, + "step": 74620 + }, + { + "epoch": 6.78331212506817, + "grad_norm": 0.1355053186416626, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0208, + "step": 74630 + }, + { + "epoch": 6.784221050718052, + "grad_norm": 0.10846066474914551, + "learning_rate": 1.027088684228894e-05, + "loss": 0.017, + "step": 74640 + }, + { + "epoch": 6.785129976367934, + "grad_norm": 0.09933800250291824, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0226, + "step": 74650 + }, + { + "epoch": 6.786038902017815, + "grad_norm": 0.12518176436424255, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.021, + "step": 74660 + }, + { + "epoch": 6.786947827667697, + "grad_norm": 0.27415764331817627, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0203, + "step": 74670 + }, + { + "epoch": 6.787856753317579, + "grad_norm": 0.09875127673149109, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0187, + "step": 74680 + }, + { + "epoch": 6.788765678967461, + "grad_norm": 0.12640273571014404, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0186, + "step": 74690 + }, + { + "epoch": 6.789674604617343, + "grad_norm": 0.10341892391443253, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0202, + "step": 74700 + }, + { + "epoch": 6.790583530267224, + "grad_norm": 0.1575084924697876, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0197, + "step": 74710 + }, + { + "epoch": 6.791492455917106, + "grad_norm": 0.15862911939620972, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.0236, + "step": 74720 + }, + { + "epoch": 6.792401381566988, + "grad_norm": 0.17164847254753113, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0176, + "step": 74730 + }, + { + "epoch": 6.79331030721687, + "grad_norm": 0.10483714938163757, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0201, + "step": 74740 + }, + { + "epoch": 6.794219232866752, + "grad_norm": 0.15803003311157227, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0183, + "step": 74750 + }, + { + "epoch": 6.795128158516634, + "grad_norm": 0.15735650062561035, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0181, + "step": 74760 + }, + { + "epoch": 6.796037084166516, + "grad_norm": 0.1352885365486145, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0167, + "step": 74770 + }, + { + "epoch": 6.796946009816397, + "grad_norm": 0.12857605516910553, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.0208, + "step": 74780 + }, + { + "epoch": 6.797854935466279, + "grad_norm": 0.12621444463729858, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0178, + "step": 74790 + }, + { + "epoch": 6.798763861116161, + "grad_norm": 0.13088686764240265, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0183, + "step": 74800 + }, + { + "epoch": 6.799672786766043, + "grad_norm": 0.15522651374340057, + "learning_rate": 1.020425036775911e-05, + "loss": 0.0181, + "step": 74810 + }, + { + "epoch": 6.800581712415925, + "grad_norm": 0.13402844965457916, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0191, + "step": 74820 + }, + { + "epoch": 6.801490638065806, + "grad_norm": 0.16143232583999634, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.019, + "step": 74830 + }, + { + "epoch": 6.802399563715688, + "grad_norm": 0.12822340428829193, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0201, + "step": 74840 + }, + { + "epoch": 6.80330848936557, + "grad_norm": 0.1805252581834793, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0183, + "step": 74850 + }, + { + "epoch": 6.804217415015452, + "grad_norm": 0.12336011976003647, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.019, + "step": 74860 + }, + { + "epoch": 6.805126340665334, + "grad_norm": 0.1309773325920105, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0177, + "step": 74870 + }, + { + "epoch": 6.806035266315216, + "grad_norm": 0.08830394595861435, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.017, + "step": 74880 + }, + { + "epoch": 6.806944191965098, + "grad_norm": 0.12521570920944214, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0197, + "step": 74890 + }, + { + "epoch": 6.807853117614979, + "grad_norm": 0.1836264729499817, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0206, + "step": 74900 + }, + { + "epoch": 6.808762043264861, + "grad_norm": 0.12677538394927979, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0189, + "step": 74910 + }, + { + "epoch": 6.809670968914743, + "grad_norm": 0.10097337514162064, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.0204, + "step": 74920 + }, + { + "epoch": 6.810579894564625, + "grad_norm": 0.12919895350933075, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.018, + "step": 74930 + }, + { + "epoch": 6.811488820214507, + "grad_norm": 0.22059398889541626, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0188, + "step": 74940 + }, + { + "epoch": 6.812397745864388, + "grad_norm": 0.09590866416692734, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.019, + "step": 74950 + }, + { + "epoch": 6.81330667151427, + "grad_norm": 0.16778987646102905, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0192, + "step": 74960 + }, + { + "epoch": 6.814215597164152, + "grad_norm": 0.12207747250795364, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.017, + "step": 74970 + }, + { + "epoch": 6.815124522814034, + "grad_norm": 0.15609411895275116, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0216, + "step": 74980 + }, + { + "epoch": 6.816033448463916, + "grad_norm": 0.10601604729890823, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.0197, + "step": 74990 + }, + { + "epoch": 6.816942374113798, + "grad_norm": 0.10586821287870407, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0195, + "step": 75000 + }, + { + "epoch": 6.81785129976368, + "grad_norm": 0.12892140448093414, + "learning_rate": 1.012598938117799e-05, + "loss": 0.0171, + "step": 75010 + }, + { + "epoch": 6.818760225413561, + "grad_norm": 0.11003713309764862, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0193, + "step": 75020 + }, + { + "epoch": 6.819669151063443, + "grad_norm": 0.12286687642335892, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0198, + "step": 75030 + }, + { + "epoch": 6.820578076713325, + "grad_norm": 0.16829073429107666, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0189, + "step": 75040 + }, + { + "epoch": 6.821487002363207, + "grad_norm": 0.13193367421627045, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0177, + "step": 75050 + }, + { + "epoch": 6.822395928013089, + "grad_norm": 0.12781497836112976, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0202, + "step": 75060 + }, + { + "epoch": 6.82330485366297, + "grad_norm": 0.13784146308898926, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.0185, + "step": 75070 + }, + { + "epoch": 6.824213779312852, + "grad_norm": 0.12911368906497955, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0196, + "step": 75080 + }, + { + "epoch": 6.825122704962734, + "grad_norm": 0.1553933322429657, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.0214, + "step": 75090 + }, + { + "epoch": 6.826031630612616, + "grad_norm": 0.11133915930986404, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0212, + "step": 75100 + }, + { + "epoch": 6.826940556262498, + "grad_norm": 0.24928219616413116, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0215, + "step": 75110 + }, + { + "epoch": 6.82784948191238, + "grad_norm": 0.11342552304267883, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0202, + "step": 75120 + }, + { + "epoch": 6.828758407562262, + "grad_norm": 0.19822299480438232, + "learning_rate": 1.007910370568197e-05, + "loss": 0.0184, + "step": 75130 + }, + { + "epoch": 6.829667333212143, + "grad_norm": 0.18622736632823944, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0193, + "step": 75140 + }, + { + "epoch": 6.830576258862025, + "grad_norm": 0.1443137526512146, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.0208, + "step": 75150 + }, + { + "epoch": 6.831485184511907, + "grad_norm": 0.1305931955575943, + "learning_rate": 1.006739067525053e-05, + "loss": 0.018, + "step": 75160 + }, + { + "epoch": 6.832394110161789, + "grad_norm": 0.19558173418045044, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.0204, + "step": 75170 + }, + { + "epoch": 6.833303035811671, + "grad_norm": 0.10500591993331909, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.02, + "step": 75180 + }, + { + "epoch": 6.834211961461552, + "grad_norm": 0.1170150488615036, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0173, + "step": 75190 + }, + { + "epoch": 6.835120887111434, + "grad_norm": 0.10670311003923416, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0184, + "step": 75200 + }, + { + "epoch": 6.836029812761316, + "grad_norm": 0.1481781005859375, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0185, + "step": 75210 + }, + { + "epoch": 6.836938738411198, + "grad_norm": 0.10029014199972153, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0202, + "step": 75220 + }, + { + "epoch": 6.83784766406108, + "grad_norm": 0.11422175168991089, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0182, + "step": 75230 + }, + { + "epoch": 6.838756589710962, + "grad_norm": 0.15767832100391388, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0198, + "step": 75240 + }, + { + "epoch": 6.839665515360844, + "grad_norm": 0.14270660281181335, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0196, + "step": 75250 + }, + { + "epoch": 6.840574441010725, + "grad_norm": 0.1262889951467514, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0189, + "step": 75260 + }, + { + "epoch": 6.841483366660607, + "grad_norm": 0.11171120405197144, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0174, + "step": 75270 + }, + { + "epoch": 6.842392292310489, + "grad_norm": 0.16581390798091888, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0181, + "step": 75280 + }, + { + "epoch": 6.843301217960371, + "grad_norm": 0.14161936938762665, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0198, + "step": 75290 + }, + { + "epoch": 6.844210143610253, + "grad_norm": 0.15115441381931305, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0181, + "step": 75300 + }, + { + "epoch": 6.845119069260134, + "grad_norm": 0.12003160268068314, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0216, + "step": 75310 + }, + { + "epoch": 6.846027994910016, + "grad_norm": 0.12347909063100815, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0175, + "step": 75320 + }, + { + "epoch": 6.846936920559898, + "grad_norm": 0.11265985667705536, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.017, + "step": 75330 + }, + { + "epoch": 6.84784584620978, + "grad_norm": 0.11277476698160172, + "learning_rate": 9.997183571932562e-06, + "loss": 0.017, + "step": 75340 + }, + { + "epoch": 6.848754771859662, + "grad_norm": 0.1521582305431366, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0195, + "step": 75350 + }, + { + "epoch": 6.849663697509544, + "grad_norm": 0.10784672200679779, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0168, + "step": 75360 + }, + { + "epoch": 6.850572623159426, + "grad_norm": 0.10957935452461243, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0201, + "step": 75370 + }, + { + "epoch": 6.851481548809307, + "grad_norm": 0.09756813198328018, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0178, + "step": 75380 + }, + { + "epoch": 6.852390474459189, + "grad_norm": 0.1397620439529419, + "learning_rate": 9.977703371038337e-06, + "loss": 0.0178, + "step": 75390 + }, + { + "epoch": 6.853299400109071, + "grad_norm": 0.13741156458854675, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0184, + "step": 75400 + }, + { + "epoch": 6.854208325758953, + "grad_norm": 0.14371468126773834, + "learning_rate": 9.969913958481688e-06, + "loss": 0.0188, + "step": 75410 + }, + { + "epoch": 6.855117251408835, + "grad_norm": 0.15638262033462524, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0224, + "step": 75420 + }, + { + "epoch": 6.856026177058717, + "grad_norm": 0.13910508155822754, + "learning_rate": 9.962126073932371e-06, + "loss": 0.0207, + "step": 75430 + }, + { + "epoch": 6.856935102708598, + "grad_norm": 0.12442003190517426, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0199, + "step": 75440 + }, + { + "epoch": 6.85784402835848, + "grad_norm": 0.10244827717542648, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0193, + "step": 75450 + }, + { + "epoch": 6.858752954008362, + "grad_norm": 0.14449335634708405, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0201, + "step": 75460 + }, + { + "epoch": 6.859661879658244, + "grad_norm": 0.10747303813695908, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0183, + "step": 75470 + }, + { + "epoch": 6.860570805308126, + "grad_norm": 0.11883260309696198, + "learning_rate": 9.942663063115248e-06, + "loss": 0.0184, + "step": 75480 + }, + { + "epoch": 6.861479730958008, + "grad_norm": 0.1391931176185608, + "learning_rate": 9.938771611983594e-06, + "loss": 0.0227, + "step": 75490 + }, + { + "epoch": 6.862388656607889, + "grad_norm": 0.1690395623445511, + "learning_rate": 9.934880545218832e-06, + "loss": 0.02, + "step": 75500 + }, + { + "epoch": 6.863297582257771, + "grad_norm": 0.15482449531555176, + "learning_rate": 9.93098986311647e-06, + "loss": 0.0197, + "step": 75510 + }, + { + "epoch": 6.864206507907653, + "grad_norm": 0.1229301169514656, + "learning_rate": 9.927099565971976e-06, + "loss": 0.02, + "step": 75520 + }, + { + "epoch": 6.865115433557535, + "grad_norm": 0.23053446412086487, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0217, + "step": 75530 + }, + { + "epoch": 6.866024359207417, + "grad_norm": 0.20812451839447021, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0201, + "step": 75540 + }, + { + "epoch": 6.866933284857299, + "grad_norm": 0.1148795560002327, + "learning_rate": 9.915430987239967e-06, + "loss": 0.0175, + "step": 75550 + }, + { + "epoch": 6.86784221050718, + "grad_norm": 0.10725054889917374, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0175, + "step": 75560 + }, + { + "epoch": 6.868751136157062, + "grad_norm": 0.11162581294775009, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0208, + "step": 75570 + }, + { + "epoch": 6.869660061806944, + "grad_norm": 0.12434538453817368, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0186, + "step": 75580 + }, + { + "epoch": 6.870568987456826, + "grad_norm": 0.10502729564905167, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0172, + "step": 75590 + }, + { + "epoch": 6.871477913106708, + "grad_norm": 0.1180121973156929, + "learning_rate": 9.895991082746168e-06, + "loss": 0.0209, + "step": 75600 + }, + { + "epoch": 6.87238683875659, + "grad_norm": 0.11425621807575226, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0188, + "step": 75610 + }, + { + "epoch": 6.873295764406471, + "grad_norm": 0.11868014186620712, + "learning_rate": 9.888217832192542e-06, + "loss": 0.0199, + "step": 75620 + }, + { + "epoch": 6.874204690056353, + "grad_norm": 0.12453455477952957, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0208, + "step": 75630 + }, + { + "epoch": 6.875113615706235, + "grad_norm": 0.12271934002637863, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0175, + "step": 75640 + }, + { + "epoch": 6.876022541356117, + "grad_norm": 0.10625763237476349, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0211, + "step": 75650 + }, + { + "epoch": 6.876931467005999, + "grad_norm": 0.13777130842208862, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0206, + "step": 75660 + }, + { + "epoch": 6.877840392655881, + "grad_norm": 0.12625257670879364, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0192, + "step": 75670 + }, + { + "epoch": 6.878749318305763, + "grad_norm": 0.2872496545314789, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0203, + "step": 75680 + }, + { + "epoch": 6.879658243955644, + "grad_norm": 0.13877297937870026, + "learning_rate": 9.861023699239261e-06, + "loss": 0.019, + "step": 75690 + }, + { + "epoch": 6.880567169605526, + "grad_norm": 0.12643243372440338, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0189, + "step": 75700 + }, + { + "epoch": 6.881476095255408, + "grad_norm": 0.12845920026302338, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0213, + "step": 75710 + }, + { + "epoch": 6.88238502090529, + "grad_norm": 0.13367782533168793, + "learning_rate": 9.84937491900087e-06, + "loss": 0.0187, + "step": 75720 + }, + { + "epoch": 6.883293946555172, + "grad_norm": 0.12424041330814362, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0191, + "step": 75730 + }, + { + "epoch": 6.884202872205053, + "grad_norm": 0.09825724363327026, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0195, + "step": 75740 + }, + { + "epoch": 6.885111797854935, + "grad_norm": 0.12389619648456573, + "learning_rate": 9.837729659166156e-06, + "loss": 0.0183, + "step": 75750 + }, + { + "epoch": 6.886020723504817, + "grad_norm": 0.12038883566856384, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0185, + "step": 75760 + }, + { + "epoch": 6.886929649154699, + "grad_norm": 0.15471769869327545, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0188, + "step": 75770 + }, + { + "epoch": 6.887838574804581, + "grad_norm": 0.11872291564941406, + "learning_rate": 9.826087927694543e-06, + "loss": 0.0196, + "step": 75780 + }, + { + "epoch": 6.888747500454463, + "grad_norm": 0.12718594074249268, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0192, + "step": 75790 + }, + { + "epoch": 6.889656426104345, + "grad_norm": 0.37225043773651123, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0196, + "step": 75800 + }, + { + "epoch": 6.890565351754226, + "grad_norm": 0.08644232153892517, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0188, + "step": 75810 + }, + { + "epoch": 6.891474277404108, + "grad_norm": 0.1395767629146576, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0206, + "step": 75820 + }, + { + "epoch": 6.89238320305399, + "grad_norm": 0.1310458779335022, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0176, + "step": 75830 + }, + { + "epoch": 6.893292128703872, + "grad_norm": 0.1389191895723343, + "learning_rate": 9.802815081666256e-06, + "loss": 0.02, + "step": 75840 + }, + { + "epoch": 6.894201054353754, + "grad_norm": 0.154864102602005, + "learning_rate": 9.79893765369845e-06, + "loss": 0.02, + "step": 75850 + }, + { + "epoch": 6.895109980003635, + "grad_norm": 0.13339027762413025, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0178, + "step": 75860 + }, + { + "epoch": 6.896018905653517, + "grad_norm": 0.1159692332148552, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0179, + "step": 75870 + }, + { + "epoch": 6.896927831303399, + "grad_norm": 0.11226107180118561, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0178, + "step": 75880 + }, + { + "epoch": 6.897836756953281, + "grad_norm": 0.10970805585384369, + "learning_rate": 9.783431894635035e-06, + "loss": 0.0175, + "step": 75890 + }, + { + "epoch": 6.898745682603163, + "grad_norm": 0.10377451032400131, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0199, + "step": 75900 + }, + { + "epoch": 6.899654608253045, + "grad_norm": 0.13896580040454865, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0184, + "step": 75910 + }, + { + "epoch": 6.900563533902927, + "grad_norm": 0.11185972392559052, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0191, + "step": 75920 + }, + { + "epoch": 6.901472459552808, + "grad_norm": 0.1421777755022049, + "learning_rate": 9.767932474193758e-06, + "loss": 0.019, + "step": 75930 + }, + { + "epoch": 6.90238138520269, + "grad_norm": 0.11710323393344879, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0174, + "step": 75940 + }, + { + "epoch": 6.903290310852572, + "grad_norm": 0.11829771101474762, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0199, + "step": 75950 + }, + { + "epoch": 6.904199236502454, + "grad_norm": 0.1684776246547699, + "learning_rate": 9.756312079913248e-06, + "loss": 0.0193, + "step": 75960 + }, + { + "epoch": 6.905108162152336, + "grad_norm": 0.11165254563093185, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0194, + "step": 75970 + }, + { + "epoch": 6.906017087802217, + "grad_norm": 0.12534183263778687, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0216, + "step": 75980 + }, + { + "epoch": 6.906926013452099, + "grad_norm": 0.13918207585811615, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0195, + "step": 75990 + }, + { + "epoch": 6.907834939101981, + "grad_norm": 0.11153416335582733, + "learning_rate": 9.740823797373594e-06, + "loss": 0.0187, + "step": 76000 + }, + { + "epoch": 6.908743864751863, + "grad_norm": 0.19452553987503052, + "learning_rate": 9.73695272450292e-06, + "loss": 0.017, + "step": 76010 + }, + { + "epoch": 6.909652790401745, + "grad_norm": 0.1552513986825943, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0219, + "step": 76020 + }, + { + "epoch": 6.910561716051627, + "grad_norm": 0.21026277542114258, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0189, + "step": 76030 + }, + { + "epoch": 6.911470641701509, + "grad_norm": 0.12332921475172043, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0199, + "step": 76040 + }, + { + "epoch": 6.91237956735139, + "grad_norm": 0.10093947499990463, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0178, + "step": 76050 + }, + { + "epoch": 6.913288493001272, + "grad_norm": 0.1508329063653946, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0193, + "step": 76060 + }, + { + "epoch": 6.914197418651154, + "grad_norm": 0.08968227356672287, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0186, + "step": 76070 + }, + { + "epoch": 6.915106344301036, + "grad_norm": 0.17621690034866333, + "learning_rate": 9.709866422291219e-06, + "loss": 0.0196, + "step": 76080 + }, + { + "epoch": 6.916015269950918, + "grad_norm": 0.09569847583770752, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0218, + "step": 76090 + }, + { + "epoch": 6.916924195600799, + "grad_norm": 0.10603482276201248, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0201, + "step": 76100 + }, + { + "epoch": 6.917833121250681, + "grad_norm": 0.12931670248508453, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0193, + "step": 76110 + }, + { + "epoch": 6.918742046900563, + "grad_norm": 0.12765870988368988, + "learning_rate": 9.694397367364635e-06, + "loss": 0.0199, + "step": 76120 + }, + { + "epoch": 6.919650972550445, + "grad_norm": 0.1094433143734932, + "learning_rate": 9.690531110212598e-06, + "loss": 0.021, + "step": 76130 + }, + { + "epoch": 6.920559898200327, + "grad_norm": 0.1523047387599945, + "learning_rate": 9.68666525627968e-06, + "loss": 0.0182, + "step": 76140 + }, + { + "epoch": 6.921468823850209, + "grad_norm": 0.1362660527229309, + "learning_rate": 9.682799805859474e-06, + "loss": 0.019, + "step": 76150 + }, + { + "epoch": 6.922377749500091, + "grad_norm": 0.15126129984855652, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0202, + "step": 76160 + }, + { + "epoch": 6.923286675149972, + "grad_norm": 0.1657763421535492, + "learning_rate": 9.675070116731382e-06, + "loss": 0.0175, + "step": 76170 + }, + { + "epoch": 6.924195600799854, + "grad_norm": 0.1937551349401474, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0171, + "step": 76180 + }, + { + "epoch": 6.925104526449736, + "grad_norm": 0.129297137260437, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0183, + "step": 76190 + }, + { + "epoch": 6.926013452099618, + "grad_norm": 0.18151889741420746, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0184, + "step": 76200 + }, + { + "epoch": 6.9269223777495, + "grad_norm": 0.13781942427158356, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0192, + "step": 76210 + }, + { + "epoch": 6.9278313033993815, + "grad_norm": 0.1409931480884552, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0183, + "step": 76220 + }, + { + "epoch": 6.9287402290492635, + "grad_norm": 0.12988345324993134, + "learning_rate": 9.651890764175553e-06, + "loss": 0.018, + "step": 76230 + }, + { + "epoch": 6.929649154699145, + "grad_norm": 0.09752220660448074, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0203, + "step": 76240 + }, + { + "epoch": 6.930558080349027, + "grad_norm": 0.11664316058158875, + "learning_rate": 9.6441675594234e-06, + "loss": 0.0209, + "step": 76250 + }, + { + "epoch": 6.931467005998909, + "grad_norm": 0.13943059742450714, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0179, + "step": 76260 + }, + { + "epoch": 6.932375931648791, + "grad_norm": 0.1704849898815155, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0183, + "step": 76270 + }, + { + "epoch": 6.933284857298673, + "grad_norm": 0.14522092044353485, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0197, + "step": 76280 + }, + { + "epoch": 6.9341937829485545, + "grad_norm": 0.10299304872751236, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0187, + "step": 76290 + }, + { + "epoch": 6.9351027085984365, + "grad_norm": 0.13026317954063416, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0172, + "step": 76300 + }, + { + "epoch": 6.9360116342483185, + "grad_norm": 0.11189598590135574, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0182, + "step": 76310 + }, + { + "epoch": 6.9369205598982004, + "grad_norm": 0.11546012759208679, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0169, + "step": 76320 + }, + { + "epoch": 6.937829485548082, + "grad_norm": 0.11448334902524948, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0194, + "step": 76330 + }, + { + "epoch": 6.9387384111979635, + "grad_norm": 0.13220174610614777, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0192, + "step": 76340 + }, + { + "epoch": 6.9396473368478455, + "grad_norm": 0.13098673522472382, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0197, + "step": 76350 + }, + { + "epoch": 6.9405562624977275, + "grad_norm": 0.12062887102365494, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0201, + "step": 76360 + }, + { + "epoch": 6.9414651881476095, + "grad_norm": 0.12283825129270554, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0167, + "step": 76370 + }, + { + "epoch": 6.9423741137974915, + "grad_norm": 0.14506949484348297, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0175, + "step": 76380 + }, + { + "epoch": 6.9432830394473735, + "grad_norm": 0.12904275953769684, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0188, + "step": 76390 + }, + { + "epoch": 6.9441919650972554, + "grad_norm": 0.11720488220453262, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0217, + "step": 76400 + }, + { + "epoch": 6.9451008907471365, + "grad_norm": 0.1351051777601242, + "learning_rate": 9.582440688952473e-06, + "loss": 0.0178, + "step": 76410 + }, + { + "epoch": 6.9460098163970185, + "grad_norm": 0.10498631745576859, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0163, + "step": 76420 + }, + { + "epoch": 6.9469187420469005, + "grad_norm": 0.12487290054559708, + "learning_rate": 9.57473221124502e-06, + "loss": 0.0184, + "step": 76430 + }, + { + "epoch": 6.9478276676967825, + "grad_norm": 0.12212062627077103, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0203, + "step": 76440 + }, + { + "epoch": 6.9487365933466645, + "grad_norm": 0.11256805062294006, + "learning_rate": 9.567025381590669e-06, + "loss": 0.0181, + "step": 76450 + }, + { + "epoch": 6.949645518996546, + "grad_norm": 0.1318061649799347, + "learning_rate": 9.563172585515032e-06, + "loss": 0.0176, + "step": 76460 + }, + { + "epoch": 6.950554444646428, + "grad_norm": 0.101763516664505, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0176, + "step": 76470 + }, + { + "epoch": 6.95146337029631, + "grad_norm": 0.17350411415100098, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0184, + "step": 76480 + }, + { + "epoch": 6.9523722959461915, + "grad_norm": 0.12280374020338058, + "learning_rate": 9.551616675805319e-06, + "loss": 0.0205, + "step": 76490 + }, + { + "epoch": 6.9532812215960735, + "grad_norm": 0.12530115246772766, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0238, + "step": 76500 + }, + { + "epoch": 6.9541901472459555, + "grad_norm": 0.15633869171142578, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0206, + "step": 76510 + }, + { + "epoch": 6.9550990728958375, + "grad_norm": 0.10112718492746353, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0182, + "step": 76520 + }, + { + "epoch": 6.956007998545719, + "grad_norm": 0.12082356959581375, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0201, + "step": 76530 + }, + { + "epoch": 6.956916924195601, + "grad_norm": 0.10326100140810013, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0177, + "step": 76540 + }, + { + "epoch": 6.957825849845483, + "grad_norm": 0.12021225690841675, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0185, + "step": 76550 + }, + { + "epoch": 6.958734775495365, + "grad_norm": 0.2221907079219818, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0178, + "step": 76560 + }, + { + "epoch": 6.9596437011452466, + "grad_norm": 0.12741313874721527, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0182, + "step": 76570 + }, + { + "epoch": 6.9605526267951285, + "grad_norm": 0.10802628844976425, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0173, + "step": 76580 + }, + { + "epoch": 6.96146155244501, + "grad_norm": 0.2149023860692978, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0199, + "step": 76590 + }, + { + "epoch": 6.962370478094892, + "grad_norm": 0.2029014676809311, + "learning_rate": 9.509276927101462e-06, + "loss": 0.0194, + "step": 76600 + }, + { + "epoch": 6.963279403744774, + "grad_norm": 0.11060931533575058, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0178, + "step": 76610 + }, + { + "epoch": 6.964188329394656, + "grad_norm": 0.23470014333724976, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0206, + "step": 76620 + }, + { + "epoch": 6.965097255044538, + "grad_norm": 0.12279000133275986, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0192, + "step": 76630 + }, + { + "epoch": 6.96600618069442, + "grad_norm": 0.10575298219919205, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0196, + "step": 76640 + }, + { + "epoch": 6.966915106344301, + "grad_norm": 0.1648329198360443, + "learning_rate": 9.490048242768298e-06, + "loss": 0.019, + "step": 76650 + }, + { + "epoch": 6.967824031994183, + "grad_norm": 0.1389508992433548, + "learning_rate": 9.486203760067013e-06, + "loss": 0.0178, + "step": 76660 + }, + { + "epoch": 6.968732957644065, + "grad_norm": 0.14014053344726562, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0192, + "step": 76670 + }, + { + "epoch": 6.969641883293947, + "grad_norm": 0.12770994007587433, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0199, + "step": 76680 + }, + { + "epoch": 6.970550808943829, + "grad_norm": 0.11592839658260345, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0166, + "step": 76690 + }, + { + "epoch": 6.971459734593711, + "grad_norm": 0.10830975323915482, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0176, + "step": 76700 + }, + { + "epoch": 6.972368660243592, + "grad_norm": 0.12899735569953918, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0203, + "step": 76710 + }, + { + "epoch": 6.973277585893474, + "grad_norm": 0.1158519834280014, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0187, + "step": 76720 + }, + { + "epoch": 6.974186511543356, + "grad_norm": 0.12648889422416687, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0196, + "step": 76730 + }, + { + "epoch": 6.975095437193238, + "grad_norm": 0.11662494391202927, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0191, + "step": 76740 + }, + { + "epoch": 6.97600436284312, + "grad_norm": 0.14690612256526947, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0174, + "step": 76750 + }, + { + "epoch": 6.976913288493002, + "grad_norm": 0.1074044331908226, + "learning_rate": 9.447782011716307e-06, + "loss": 0.0174, + "step": 76760 + }, + { + "epoch": 6.977822214142883, + "grad_norm": 0.27308449149131775, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0197, + "step": 76770 + }, + { + "epoch": 6.978731139792765, + "grad_norm": 0.20789533853530884, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0188, + "step": 76780 + }, + { + "epoch": 6.979640065442647, + "grad_norm": 0.13026973605155945, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0214, + "step": 76790 + }, + { + "epoch": 6.980548991092529, + "grad_norm": 0.1654299944639206, + "learning_rate": 9.432425102360072e-06, + "loss": 0.0182, + "step": 76800 + }, + { + "epoch": 6.981457916742411, + "grad_norm": 0.12466361373662949, + "learning_rate": 9.428586931343613e-06, + "loss": 0.0183, + "step": 76810 + }, + { + "epoch": 6.982366842392293, + "grad_norm": 0.10712969303131104, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0179, + "step": 76820 + }, + { + "epoch": 6.983275768042175, + "grad_norm": 0.1275935173034668, + "learning_rate": 9.4209118589383e-06, + "loss": 0.019, + "step": 76830 + }, + { + "epoch": 6.984184693692056, + "grad_norm": 0.12307653576135635, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0204, + "step": 76840 + }, + { + "epoch": 6.985093619341938, + "grad_norm": 0.1430194228887558, + "learning_rate": 9.413238481312658e-06, + "loss": 0.018, + "step": 76850 + }, + { + "epoch": 6.98600254499182, + "grad_norm": 0.1411808878183365, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0204, + "step": 76860 + }, + { + "epoch": 6.986911470641702, + "grad_norm": 0.14245152473449707, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0175, + "step": 76870 + }, + { + "epoch": 6.987820396291584, + "grad_norm": 0.15200234949588776, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0178, + "step": 76880 + }, + { + "epoch": 6.988729321941465, + "grad_norm": 0.143358513712883, + "learning_rate": 9.39789681972375e-06, + "loss": 0.0169, + "step": 76890 + }, + { + "epoch": 6.989638247591347, + "grad_norm": 0.13928672671318054, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0189, + "step": 76900 + }, + { + "epoch": 6.990547173241229, + "grad_norm": 0.19751930236816406, + "learning_rate": 9.390228540420876e-06, + "loss": 0.0217, + "step": 76910 + }, + { + "epoch": 6.991456098891111, + "grad_norm": 0.13326922059059143, + "learning_rate": 9.386395039661539e-06, + "loss": 0.0181, + "step": 76920 + }, + { + "epoch": 6.992365024540993, + "grad_norm": 0.12241346389055252, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0183, + "step": 76930 + }, + { + "epoch": 6.993273950190875, + "grad_norm": 0.11461411416530609, + "learning_rate": 9.378729317382715e-06, + "loss": 0.0181, + "step": 76940 + }, + { + "epoch": 6.994182875840757, + "grad_norm": 0.13770568370819092, + "learning_rate": 9.37489709644538e-06, + "loss": 0.0187, + "step": 76950 + }, + { + "epoch": 6.995091801490638, + "grad_norm": 0.1200706884264946, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0175, + "step": 76960 + }, + { + "epoch": 6.99600072714052, + "grad_norm": 0.12658725678920746, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0194, + "step": 76970 + }, + { + "epoch": 6.996909652790402, + "grad_norm": 0.1660231500864029, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0215, + "step": 76980 + }, + { + "epoch": 6.997818578440284, + "grad_norm": 0.1267361044883728, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0195, + "step": 76990 + }, + { + "epoch": 6.998727504090166, + "grad_norm": 0.11137457937002182, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0213, + "step": 77000 + }, + { + "epoch": 6.999636429740047, + "grad_norm": 0.17442482709884644, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0198, + "step": 77010 + }, + { + "epoch": 7.000545355389929, + "grad_norm": 0.1634000837802887, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0192, + "step": 77020 + }, + { + "epoch": 7.001454281039811, + "grad_norm": 0.11592484265565872, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0181, + "step": 77030 + }, + { + "epoch": 7.002363206689693, + "grad_norm": 0.1322048455476761, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0227, + "step": 77040 + }, + { + "epoch": 7.003272132339575, + "grad_norm": 0.12444586306810379, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0154, + "step": 77050 + }, + { + "epoch": 7.004181057989457, + "grad_norm": 0.14580394327640533, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0195, + "step": 77060 + }, + { + "epoch": 7.005089983639338, + "grad_norm": 0.12282763421535492, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0209, + "step": 77070 + }, + { + "epoch": 7.00599890928922, + "grad_norm": 0.13541163504123688, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0183, + "step": 77080 + }, + { + "epoch": 7.006907834939102, + "grad_norm": 0.12913544476032257, + "learning_rate": 9.321290990574668e-06, + "loss": 0.0192, + "step": 77090 + }, + { + "epoch": 7.007816760588984, + "grad_norm": 0.10224539041519165, + "learning_rate": 9.317465208019126e-06, + "loss": 0.019, + "step": 77100 + }, + { + "epoch": 7.008725686238866, + "grad_norm": 0.13279639184474945, + "learning_rate": 9.313639857014607e-06, + "loss": 0.0172, + "step": 77110 + }, + { + "epoch": 7.009634611888748, + "grad_norm": 0.09379728138446808, + "learning_rate": 9.3098149378516e-06, + "loss": 0.0194, + "step": 77120 + }, + { + "epoch": 7.010543537538629, + "grad_norm": 0.12797032296657562, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0187, + "step": 77130 + }, + { + "epoch": 7.011452463188511, + "grad_norm": 0.13018636405467987, + "learning_rate": 9.302166396212059e-06, + "loss": 0.0196, + "step": 77140 + }, + { + "epoch": 7.012361388838393, + "grad_norm": 0.09987868368625641, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0174, + "step": 77150 + }, + { + "epoch": 7.013270314488275, + "grad_norm": 0.1725594699382782, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0187, + "step": 77160 + }, + { + "epoch": 7.014179240138157, + "grad_norm": 0.1255631446838379, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0183, + "step": 77170 + }, + { + "epoch": 7.015088165788039, + "grad_norm": 0.14383262395858765, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0187, + "step": 77180 + }, + { + "epoch": 7.01599709143792, + "grad_norm": 0.1256657838821411, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0185, + "step": 77190 + }, + { + "epoch": 7.016906017087802, + "grad_norm": 0.1529524177312851, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0186, + "step": 77200 + }, + { + "epoch": 7.017814942737684, + "grad_norm": 0.15191863477230072, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0215, + "step": 77210 + }, + { + "epoch": 7.018723868387566, + "grad_norm": 0.14339576661586761, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0177, + "step": 77220 + }, + { + "epoch": 7.019632794037448, + "grad_norm": 0.17484284937381744, + "learning_rate": 9.26776941165236e-06, + "loss": 0.021, + "step": 77230 + }, + { + "epoch": 7.02054171968733, + "grad_norm": 0.12303241342306137, + "learning_rate": 9.263949697235511e-06, + "loss": 0.018, + "step": 77240 + }, + { + "epoch": 7.021450645337211, + "grad_norm": 0.14260423183441162, + "learning_rate": 9.260130418433819e-06, + "loss": 0.017, + "step": 77250 + }, + { + "epoch": 7.022359570987093, + "grad_norm": 0.1254025250673294, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0163, + "step": 77260 + }, + { + "epoch": 7.023268496636975, + "grad_norm": 0.16907739639282227, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0192, + "step": 77270 + }, + { + "epoch": 7.024177422286857, + "grad_norm": 0.26880526542663574, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0206, + "step": 77280 + }, + { + "epoch": 7.025086347936739, + "grad_norm": 0.1347511112689972, + "learning_rate": 9.244857665179213e-06, + "loss": 0.0174, + "step": 77290 + }, + { + "epoch": 7.025995273586621, + "grad_norm": 0.19067814946174622, + "learning_rate": 9.241040568803432e-06, + "loss": 0.0196, + "step": 77300 + }, + { + "epoch": 7.026904199236503, + "grad_norm": 0.10957163572311401, + "learning_rate": 9.23722390978261e-06, + "loss": 0.0206, + "step": 77310 + }, + { + "epoch": 7.027813124886384, + "grad_norm": 0.12554237246513367, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0178, + "step": 77320 + }, + { + "epoch": 7.028722050536266, + "grad_norm": 0.13079698383808136, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0191, + "step": 77330 + }, + { + "epoch": 7.029630976186148, + "grad_norm": 0.2425452172756195, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0178, + "step": 77340 + }, + { + "epoch": 7.03053990183603, + "grad_norm": 0.1302425116300583, + "learning_rate": 9.221961653045449e-06, + "loss": 0.0174, + "step": 77350 + }, + { + "epoch": 7.031448827485912, + "grad_norm": 0.1503162384033203, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0194, + "step": 77360 + }, + { + "epoch": 7.032357753135794, + "grad_norm": 0.1582588404417038, + "learning_rate": 9.214333156341175e-06, + "loss": 0.0217, + "step": 77370 + }, + { + "epoch": 7.033266678785675, + "grad_norm": 0.28107693791389465, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0194, + "step": 77380 + }, + { + "epoch": 7.034175604435557, + "grad_norm": 0.11089035123586655, + "learning_rate": 9.206706417169721e-06, + "loss": 0.02, + "step": 77390 + }, + { + "epoch": 7.035084530085439, + "grad_norm": 0.13428843021392822, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0174, + "step": 77400 + }, + { + "epoch": 7.035993455735321, + "grad_norm": 0.16799437999725342, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0192, + "step": 77410 + }, + { + "epoch": 7.036902381385203, + "grad_norm": 0.13980303704738617, + "learning_rate": 9.195269608854445e-06, + "loss": 0.0175, + "step": 77420 + }, + { + "epoch": 7.037811307035085, + "grad_norm": 0.1334458887577057, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0182, + "step": 77430 + }, + { + "epoch": 7.038720232684966, + "grad_norm": 0.12316840887069702, + "learning_rate": 9.1876472736499e-06, + "loss": 0.018, + "step": 77440 + }, + { + "epoch": 7.039629158334848, + "grad_norm": 0.1280287206172943, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0171, + "step": 77450 + }, + { + "epoch": 7.04053808398473, + "grad_norm": 0.1322208046913147, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0192, + "step": 77460 + }, + { + "epoch": 7.041447009634612, + "grad_norm": 0.18123506009578705, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0174, + "step": 77470 + }, + { + "epoch": 7.042355935284494, + "grad_norm": 0.14648349583148956, + "learning_rate": 9.172407902473605e-06, + "loss": 0.0181, + "step": 77480 + }, + { + "epoch": 7.043264860934376, + "grad_norm": 0.1089586690068245, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0187, + "step": 77490 + }, + { + "epoch": 7.044173786584257, + "grad_norm": 0.11750108003616333, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0185, + "step": 77500 + }, + { + "epoch": 7.045082712234139, + "grad_norm": 0.11793916672468185, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0194, + "step": 77510 + }, + { + "epoch": 7.045991637884021, + "grad_norm": 0.15043742954730988, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0194, + "step": 77520 + }, + { + "epoch": 7.046900563533903, + "grad_norm": 0.10959699749946594, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0177, + "step": 77530 + }, + { + "epoch": 7.047809489183785, + "grad_norm": 0.14522448182106018, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0202, + "step": 77540 + }, + { + "epoch": 7.048718414833667, + "grad_norm": 0.14884227514266968, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0216, + "step": 77550 + }, + { + "epoch": 7.049627340483548, + "grad_norm": 0.17718836665153503, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0184, + "step": 77560 + }, + { + "epoch": 7.05053626613343, + "grad_norm": 0.13033628463745117, + "learning_rate": 9.138145235699377e-06, + "loss": 0.0203, + "step": 77570 + }, + { + "epoch": 7.051445191783312, + "grad_norm": 0.21669000387191772, + "learning_rate": 9.134340494716546e-06, + "loss": 0.0192, + "step": 77580 + }, + { + "epoch": 7.052354117433194, + "grad_norm": 0.15010862052440643, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0168, + "step": 77590 + }, + { + "epoch": 7.053263043083076, + "grad_norm": 0.12390005588531494, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0207, + "step": 77600 + }, + { + "epoch": 7.054171968732958, + "grad_norm": 0.1237475797533989, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0178, + "step": 77610 + }, + { + "epoch": 7.055080894382839, + "grad_norm": 0.10291589796543121, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0179, + "step": 77620 + }, + { + "epoch": 7.055989820032721, + "grad_norm": 0.1119919866323471, + "learning_rate": 9.115323477452056e-06, + "loss": 0.017, + "step": 77630 + }, + { + "epoch": 7.056898745682603, + "grad_norm": 0.10825745016336441, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0197, + "step": 77640 + }, + { + "epoch": 7.057807671332485, + "grad_norm": 0.10461170971393585, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0201, + "step": 77650 + }, + { + "epoch": 7.058716596982367, + "grad_norm": 0.12371239066123962, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0199, + "step": 77660 + }, + { + "epoch": 7.059625522632249, + "grad_norm": 0.13159939646720886, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0202, + "step": 77670 + }, + { + "epoch": 7.06053444828213, + "grad_norm": 0.09988030791282654, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0178, + "step": 77680 + }, + { + "epoch": 7.061443373932012, + "grad_norm": 0.16139762103557587, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0174, + "step": 77690 + }, + { + "epoch": 7.062352299581894, + "grad_norm": 0.1467158943414688, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0187, + "step": 77700 + }, + { + "epoch": 7.063261225231776, + "grad_norm": 0.14380308985710144, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0192, + "step": 77710 + }, + { + "epoch": 7.064170150881658, + "grad_norm": 0.13265752792358398, + "learning_rate": 9.081121025468816e-06, + "loss": 0.0179, + "step": 77720 + }, + { + "epoch": 7.06507907653154, + "grad_norm": 0.11968034505844116, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0174, + "step": 77730 + }, + { + "epoch": 7.065988002181421, + "grad_norm": 0.11637157201766968, + "learning_rate": 9.073525417675752e-06, + "loss": 0.0182, + "step": 77740 + }, + { + "epoch": 7.066896927831303, + "grad_norm": 0.1387191265821457, + "learning_rate": 9.069728288749965e-06, + "loss": 0.018, + "step": 77750 + }, + { + "epoch": 7.067805853481185, + "grad_norm": 0.14401575922966003, + "learning_rate": 9.065931610189176e-06, + "loss": 0.019, + "step": 77760 + }, + { + "epoch": 7.068714779131067, + "grad_norm": 0.1183251217007637, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0178, + "step": 77770 + }, + { + "epoch": 7.069623704780949, + "grad_norm": 0.11798009276390076, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0176, + "step": 77780 + }, + { + "epoch": 7.070532630430831, + "grad_norm": 0.10916748642921448, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0216, + "step": 77790 + }, + { + "epoch": 7.071441556080712, + "grad_norm": 0.10437203198671341, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0185, + "step": 77800 + }, + { + "epoch": 7.072350481730594, + "grad_norm": 0.12213707715272903, + "learning_rate": 9.046954982950699e-06, + "loss": 0.019, + "step": 77810 + }, + { + "epoch": 7.073259407380476, + "grad_norm": 0.11493085324764252, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0179, + "step": 77820 + }, + { + "epoch": 7.074168333030358, + "grad_norm": 0.10958172380924225, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0183, + "step": 77830 + }, + { + "epoch": 7.07507725868024, + "grad_norm": 0.12034790217876434, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0201, + "step": 77840 + }, + { + "epoch": 7.075986184330122, + "grad_norm": 0.14495916664600372, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0171, + "step": 77850 + }, + { + "epoch": 7.076895109980003, + "grad_norm": 0.19801822304725647, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0198, + "step": 77860 + }, + { + "epoch": 7.077804035629885, + "grad_norm": 0.11847519874572754, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0183, + "step": 77870 + }, + { + "epoch": 7.078712961279767, + "grad_norm": 0.1280861347913742, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0183, + "step": 77880 + }, + { + "epoch": 7.079621886929649, + "grad_norm": 0.1560956835746765, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0182, + "step": 77890 + }, + { + "epoch": 7.080530812579531, + "grad_norm": 0.08338062465190887, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0164, + "step": 77900 + }, + { + "epoch": 7.081439738229413, + "grad_norm": 0.10499729961156845, + "learning_rate": 9.009035671562044e-06, + "loss": 0.019, + "step": 77910 + }, + { + "epoch": 7.082348663879294, + "grad_norm": 0.09771978110074997, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0164, + "step": 77920 + }, + { + "epoch": 7.083257589529176, + "grad_norm": 0.13994759321212769, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0204, + "step": 77930 + }, + { + "epoch": 7.084166515179058, + "grad_norm": 0.16656963527202606, + "learning_rate": 8.997668737043157e-06, + "loss": 0.02, + "step": 77940 + }, + { + "epoch": 7.08507544082894, + "grad_norm": 0.12395050376653671, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0195, + "step": 77950 + }, + { + "epoch": 7.085984366478822, + "grad_norm": 0.13308121263980865, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0194, + "step": 77960 + }, + { + "epoch": 7.086893292128704, + "grad_norm": 0.1935187727212906, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0181, + "step": 77970 + }, + { + "epoch": 7.087802217778585, + "grad_norm": 0.18389296531677246, + "learning_rate": 8.982519207417674e-06, + "loss": 0.022, + "step": 77980 + }, + { + "epoch": 7.088711143428467, + "grad_norm": 0.12573814392089844, + "learning_rate": 8.978732966762205e-06, + "loss": 0.0176, + "step": 77990 + }, + { + "epoch": 7.089620069078349, + "grad_norm": 0.12354150414466858, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0187, + "step": 78000 + }, + { + "epoch": 7.090528994728231, + "grad_norm": 0.20452147722244263, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0223, + "step": 78010 + }, + { + "epoch": 7.091437920378113, + "grad_norm": 0.12068686634302139, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0189, + "step": 78020 + }, + { + "epoch": 7.092346846027995, + "grad_norm": 0.18526309728622437, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0191, + "step": 78030 + }, + { + "epoch": 7.093255771677876, + "grad_norm": 0.0983976498246193, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0164, + "step": 78040 + }, + { + "epoch": 7.094164697327758, + "grad_norm": 0.10615134984254837, + "learning_rate": 8.95602513567626e-06, + "loss": 0.0174, + "step": 78050 + }, + { + "epoch": 7.09507362297764, + "grad_norm": 0.14461901783943176, + "learning_rate": 8.952242101985645e-06, + "loss": 0.0203, + "step": 78060 + }, + { + "epoch": 7.095982548627522, + "grad_norm": 0.16714239120483398, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0188, + "step": 78070 + }, + { + "epoch": 7.096891474277404, + "grad_norm": 0.1314818114042282, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0171, + "step": 78080 + }, + { + "epoch": 7.097800399927286, + "grad_norm": 0.14036335051059723, + "learning_rate": 8.940895757786624e-06, + "loss": 0.0161, + "step": 78090 + }, + { + "epoch": 7.098709325577167, + "grad_norm": 0.11554978787899017, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0174, + "step": 78100 + }, + { + "epoch": 7.099618251227049, + "grad_norm": 0.18113431334495544, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0209, + "step": 78110 + }, + { + "epoch": 7.100527176876931, + "grad_norm": 0.10341209918260574, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0203, + "step": 78120 + }, + { + "epoch": 7.101436102526813, + "grad_norm": 0.11861375719308853, + "learning_rate": 8.92577374227998e-06, + "loss": 0.019, + "step": 78130 + }, + { + "epoch": 7.102345028176695, + "grad_norm": 0.11991365253925323, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0192, + "step": 78140 + }, + { + "epoch": 7.103253953826577, + "grad_norm": 0.17455218732357025, + "learning_rate": 8.918215501162718e-06, + "loss": 0.019, + "step": 78150 + }, + { + "epoch": 7.104162879476458, + "grad_norm": 0.18922822177410126, + "learning_rate": 8.914437073267688e-06, + "loss": 0.02, + "step": 78160 + }, + { + "epoch": 7.10507180512634, + "grad_norm": 0.1353844255208969, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0207, + "step": 78170 + }, + { + "epoch": 7.105980730776222, + "grad_norm": 0.11598841100931168, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0182, + "step": 78180 + }, + { + "epoch": 7.106889656426104, + "grad_norm": 0.11317233741283417, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0186, + "step": 78190 + }, + { + "epoch": 7.107798582075986, + "grad_norm": 0.13064812123775482, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0204, + "step": 78200 + }, + { + "epoch": 7.108707507725868, + "grad_norm": 0.13019873201847076, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0207, + "step": 78210 + }, + { + "epoch": 7.1096164333757494, + "grad_norm": 0.1485842615365982, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0182, + "step": 78220 + }, + { + "epoch": 7.110525359025631, + "grad_norm": 0.1324986070394516, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0213, + "step": 78230 + }, + { + "epoch": 7.111434284675513, + "grad_norm": 0.11041749268770218, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0203, + "step": 78240 + }, + { + "epoch": 7.112343210325395, + "grad_norm": 0.15133784711360931, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0194, + "step": 78250 + }, + { + "epoch": 7.113252135975277, + "grad_norm": 0.1658971607685089, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0186, + "step": 78260 + }, + { + "epoch": 7.114161061625159, + "grad_norm": 0.11263780295848846, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0196, + "step": 78270 + }, + { + "epoch": 7.1150699872750405, + "grad_norm": 0.12900041043758392, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0186, + "step": 78280 + }, + { + "epoch": 7.1159789129249225, + "grad_norm": 0.14590568840503693, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0197, + "step": 78290 + }, + { + "epoch": 7.1168878385748044, + "grad_norm": 0.12384076416492462, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0167, + "step": 78300 + }, + { + "epoch": 7.117796764224686, + "grad_norm": 0.13901162147521973, + "learning_rate": 8.857816274449894e-06, + "loss": 0.019, + "step": 78310 + }, + { + "epoch": 7.118705689874568, + "grad_norm": 0.12321698665618896, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0189, + "step": 78320 + }, + { + "epoch": 7.11961461552445, + "grad_norm": 0.1275184601545334, + "learning_rate": 8.850274743290876e-06, + "loss": 0.0216, + "step": 78330 + }, + { + "epoch": 7.1205235411743315, + "grad_norm": 0.14856955409049988, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0157, + "step": 78340 + }, + { + "epoch": 7.1214324668242135, + "grad_norm": 0.12067095190286636, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0187, + "step": 78350 + }, + { + "epoch": 7.1223413924740955, + "grad_norm": 0.10213353484869003, + "learning_rate": 8.838965950000907e-06, + "loss": 0.0189, + "step": 78360 + }, + { + "epoch": 7.1232503181239775, + "grad_norm": 0.1465008556842804, + "learning_rate": 8.835197287635673e-06, + "loss": 0.0208, + "step": 78370 + }, + { + "epoch": 7.1241592437738595, + "grad_norm": 0.25042924284935, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0187, + "step": 78380 + }, + { + "epoch": 7.125068169423741, + "grad_norm": 0.11412517726421356, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0174, + "step": 78390 + }, + { + "epoch": 7.1259770950736225, + "grad_norm": 0.20794197916984558, + "learning_rate": 8.823894110741103e-06, + "loss": 0.0196, + "step": 78400 + }, + { + "epoch": 7.1268860207235045, + "grad_norm": 0.16461652517318726, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0172, + "step": 78410 + }, + { + "epoch": 7.1277949463733865, + "grad_norm": 0.19854192435741425, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0181, + "step": 78420 + }, + { + "epoch": 7.1287038720232685, + "grad_norm": 0.13517490029335022, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0188, + "step": 78430 + }, + { + "epoch": 7.1296127976731505, + "grad_norm": 0.1466912478208542, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0183, + "step": 78440 + }, + { + "epoch": 7.1305217233230325, + "grad_norm": 0.15711773931980133, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0196, + "step": 78450 + }, + { + "epoch": 7.1314306489729145, + "grad_norm": 0.11919526755809784, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0175, + "step": 78460 + }, + { + "epoch": 7.1323395746227956, + "grad_norm": 0.1537659913301468, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0193, + "step": 78470 + }, + { + "epoch": 7.1332485002726775, + "grad_norm": 0.15718503296375275, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0185, + "step": 78480 + }, + { + "epoch": 7.1341574259225595, + "grad_norm": 0.12700966000556946, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0207, + "step": 78490 + }, + { + "epoch": 7.1350663515724415, + "grad_norm": 0.14999349415302277, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0171, + "step": 78500 + }, + { + "epoch": 7.1359752772223235, + "grad_norm": 0.15508244931697845, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0181, + "step": 78510 + }, + { + "epoch": 7.136884202872205, + "grad_norm": 0.11297991126775742, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0182, + "step": 78520 + }, + { + "epoch": 7.137793128522087, + "grad_norm": 0.1122405081987381, + "learning_rate": 8.774962556249487e-06, + "loss": 0.0192, + "step": 78530 + }, + { + "epoch": 7.138702054171969, + "grad_norm": 0.1201505959033966, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0191, + "step": 78540 + }, + { + "epoch": 7.139610979821851, + "grad_norm": 0.15185053646564484, + "learning_rate": 8.767441700331522e-06, + "loss": 0.02, + "step": 78550 + }, + { + "epoch": 7.1405199054717325, + "grad_norm": 0.13434569537639618, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0189, + "step": 78560 + }, + { + "epoch": 7.1414288311216145, + "grad_norm": 0.19131779670715332, + "learning_rate": 8.759922737700205e-06, + "loss": 0.02, + "step": 78570 + }, + { + "epoch": 7.1423377567714965, + "grad_norm": 0.1373991072177887, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0198, + "step": 78580 + }, + { + "epoch": 7.143246682421378, + "grad_norm": 0.16663509607315063, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0207, + "step": 78590 + }, + { + "epoch": 7.14415560807126, + "grad_norm": 0.13965411484241486, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0191, + "step": 78600 + }, + { + "epoch": 7.145064533721142, + "grad_norm": 0.10311023890972137, + "learning_rate": 8.744890501433196e-06, + "loss": 0.0196, + "step": 78610 + }, + { + "epoch": 7.145973459371024, + "grad_norm": 0.33378610014915466, + "learning_rate": 8.741133629238754e-06, + "loss": 0.018, + "step": 78620 + }, + { + "epoch": 7.146882385020906, + "grad_norm": 0.09281880408525467, + "learning_rate": 8.737377232363901e-06, + "loss": 0.0161, + "step": 78630 + }, + { + "epoch": 7.1477913106707875, + "grad_norm": 0.1661580204963684, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0193, + "step": 78640 + }, + { + "epoch": 7.148700236320669, + "grad_norm": 0.1282464563846588, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0166, + "step": 78650 + }, + { + "epoch": 7.149609161970551, + "grad_norm": 0.13683979213237762, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0185, + "step": 78660 + }, + { + "epoch": 7.150518087620433, + "grad_norm": 0.11079786717891693, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0183, + "step": 78670 + }, + { + "epoch": 7.151427013270315, + "grad_norm": 0.14626489579677582, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0178, + "step": 78680 + }, + { + "epoch": 7.152335938920197, + "grad_norm": 0.12316171079874039, + "learning_rate": 8.71484884879911e-06, + "loss": 0.0194, + "step": 78690 + }, + { + "epoch": 7.153244864570079, + "grad_norm": 0.14027179777622223, + "learning_rate": 8.711095787147064e-06, + "loss": 0.02, + "step": 78700 + }, + { + "epoch": 7.15415379021996, + "grad_norm": 0.08659862726926804, + "learning_rate": 8.707343203095797e-06, + "loss": 0.0172, + "step": 78710 + }, + { + "epoch": 7.155062715869842, + "grad_norm": 0.11985059827566147, + "learning_rate": 8.703591096930281e-06, + "loss": 0.019, + "step": 78720 + }, + { + "epoch": 7.155971641519724, + "grad_norm": 0.09465597569942474, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0192, + "step": 78730 + }, + { + "epoch": 7.156880567169606, + "grad_norm": 0.13074319064617157, + "learning_rate": 8.696088319396272e-06, + "loss": 0.0185, + "step": 78740 + }, + { + "epoch": 7.157789492819488, + "grad_norm": 0.11848706007003784, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0197, + "step": 78750 + }, + { + "epoch": 7.15869841846937, + "grad_norm": 0.16223862767219543, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0196, + "step": 78760 + }, + { + "epoch": 7.159607344119251, + "grad_norm": 0.16299103200435638, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0169, + "step": 78770 + }, + { + "epoch": 7.160516269769133, + "grad_norm": 0.12261346727609634, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0185, + "step": 78780 + }, + { + "epoch": 7.161425195419015, + "grad_norm": 0.09184900671243668, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0192, + "step": 78790 + }, + { + "epoch": 7.162334121068897, + "grad_norm": 0.1300220936536789, + "learning_rate": 8.673591485679457e-06, + "loss": 0.016, + "step": 78800 + }, + { + "epoch": 7.163243046718779, + "grad_norm": 0.127990260720253, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0184, + "step": 78810 + }, + { + "epoch": 7.164151972368661, + "grad_norm": 0.11717256158590317, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0177, + "step": 78820 + }, + { + "epoch": 7.165060898018542, + "grad_norm": 0.1385333389043808, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0177, + "step": 78830 + }, + { + "epoch": 7.165969823668424, + "grad_norm": 0.11075383424758911, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0192, + "step": 78840 + }, + { + "epoch": 7.166878749318306, + "grad_norm": 0.12499619275331497, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0196, + "step": 78850 + }, + { + "epoch": 7.167787674968188, + "grad_norm": 0.09537047147750854, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0197, + "step": 78860 + }, + { + "epoch": 7.16869660061807, + "grad_norm": 0.09871439635753632, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0166, + "step": 78870 + }, + { + "epoch": 7.169605526267952, + "grad_norm": 0.08876600116491318, + "learning_rate": 8.643622623150848e-06, + "loss": 0.02, + "step": 78880 + }, + { + "epoch": 7.170514451917833, + "grad_norm": 0.11860169470310211, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0181, + "step": 78890 + }, + { + "epoch": 7.171423377567715, + "grad_norm": 0.17516906559467316, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0169, + "step": 78900 + }, + { + "epoch": 7.172332303217597, + "grad_norm": 0.10391317307949066, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0172, + "step": 78910 + }, + { + "epoch": 7.173241228867479, + "grad_norm": 0.13411325216293335, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0173, + "step": 78920 + }, + { + "epoch": 7.174150154517361, + "grad_norm": 0.1298002153635025, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0193, + "step": 78930 + }, + { + "epoch": 7.175059080167243, + "grad_norm": 0.1060541495680809, + "learning_rate": 8.62116624273252e-06, + "loss": 0.017, + "step": 78940 + }, + { + "epoch": 7.175968005817124, + "grad_norm": 0.11573199182748795, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0169, + "step": 78950 + }, + { + "epoch": 7.176876931467006, + "grad_norm": 0.146213561296463, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0202, + "step": 78960 + }, + { + "epoch": 7.177785857116888, + "grad_norm": 0.13387207686901093, + "learning_rate": 8.60994458849601e-06, + "loss": 0.0178, + "step": 78970 + }, + { + "epoch": 7.17869478276677, + "grad_norm": 0.1235366240143776, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0173, + "step": 78980 + }, + { + "epoch": 7.179603708416652, + "grad_norm": 0.1617838740348816, + "learning_rate": 8.602465911610909e-06, + "loss": 0.0214, + "step": 78990 + }, + { + "epoch": 7.180512634066534, + "grad_norm": 0.11354222893714905, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0212, + "step": 79000 + }, + { + "epoch": 7.181421559716415, + "grad_norm": 0.1578620821237564, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0171, + "step": 79010 + }, + { + "epoch": 7.182330485366297, + "grad_norm": 0.1377275139093399, + "learning_rate": 8.591251540871672e-06, + "loss": 0.018, + "step": 79020 + }, + { + "epoch": 7.183239411016179, + "grad_norm": 0.12315567582845688, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0165, + "step": 79030 + }, + { + "epoch": 7.184148336666061, + "grad_norm": 0.14339235424995422, + "learning_rate": 8.583777726749796e-06, + "loss": 0.019, + "step": 79040 + }, + { + "epoch": 7.185057262315943, + "grad_norm": 0.17506644129753113, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0185, + "step": 79050 + }, + { + "epoch": 7.185966187965825, + "grad_norm": 0.47398310899734497, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0195, + "step": 79060 + }, + { + "epoch": 7.186875113615706, + "grad_norm": 0.10367919504642487, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0188, + "step": 79070 + }, + { + "epoch": 7.187784039265588, + "grad_norm": 0.15586717426776886, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0179, + "step": 79080 + }, + { + "epoch": 7.18869296491547, + "grad_norm": 0.11122006177902222, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0176, + "step": 79090 + }, + { + "epoch": 7.189601890565352, + "grad_norm": 0.11155057698488235, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0172, + "step": 79100 + }, + { + "epoch": 7.190510816215234, + "grad_norm": 0.1171698123216629, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0156, + "step": 79110 + }, + { + "epoch": 7.191419741865116, + "grad_norm": 0.09071232378482819, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0161, + "step": 79120 + }, + { + "epoch": 7.192328667514997, + "grad_norm": 0.1073535606265068, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0178, + "step": 79130 + }, + { + "epoch": 7.193237593164879, + "grad_norm": 0.127632275223732, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0175, + "step": 79140 + }, + { + "epoch": 7.194146518814761, + "grad_norm": 0.13739551603794098, + "learning_rate": 8.542706649709004e-06, + "loss": 0.02, + "step": 79150 + }, + { + "epoch": 7.195055444464643, + "grad_norm": 0.13340748846530914, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0201, + "step": 79160 + }, + { + "epoch": 7.195964370114525, + "grad_norm": 0.1285991072654724, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0175, + "step": 79170 + }, + { + "epoch": 7.196873295764407, + "grad_norm": 0.1483502835035324, + "learning_rate": 8.53151572917108e-06, + "loss": 0.0202, + "step": 79180 + }, + { + "epoch": 7.197782221414288, + "grad_norm": 0.16345621645450592, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0202, + "step": 79190 + }, + { + "epoch": 7.19869114706417, + "grad_norm": 0.1199190616607666, + "learning_rate": 8.524057571200438e-06, + "loss": 0.0206, + "step": 79200 + }, + { + "epoch": 7.199600072714052, + "grad_norm": 0.12712980806827545, + "learning_rate": 8.520329229781158e-06, + "loss": 0.0167, + "step": 79210 + }, + { + "epoch": 7.200508998363934, + "grad_norm": 0.11986567080020905, + "learning_rate": 8.516601380450119e-06, + "loss": 0.0177, + "step": 79220 + }, + { + "epoch": 7.201417924013816, + "grad_norm": 0.1462279111146927, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0186, + "step": 79230 + }, + { + "epoch": 7.202326849663698, + "grad_norm": 0.12003909796476364, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0172, + "step": 79240 + }, + { + "epoch": 7.203235775313579, + "grad_norm": 0.1386869251728058, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0178, + "step": 79250 + }, + { + "epoch": 7.204144700963461, + "grad_norm": 0.1072721928358078, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0191, + "step": 79260 + }, + { + "epoch": 7.205053626613343, + "grad_norm": 0.1902099847793579, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0172, + "step": 79270 + }, + { + "epoch": 7.205962552263225, + "grad_norm": 0.14328354597091675, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0223, + "step": 79280 + }, + { + "epoch": 7.206871477913107, + "grad_norm": 0.1906253546476364, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0185, + "step": 79290 + }, + { + "epoch": 7.207780403562989, + "grad_norm": 0.21308284997940063, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0202, + "step": 79300 + }, + { + "epoch": 7.20868932921287, + "grad_norm": 0.1744888871908188, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0208, + "step": 79310 + }, + { + "epoch": 7.209598254862752, + "grad_norm": 0.12099598348140717, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0191, + "step": 79320 + }, + { + "epoch": 7.210507180512634, + "grad_norm": 0.12621437013149261, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0183, + "step": 79330 + }, + { + "epoch": 7.211416106162516, + "grad_norm": 0.18795891106128693, + "learning_rate": 8.471905674372712e-06, + "loss": 0.0194, + "step": 79340 + }, + { + "epoch": 7.212325031812398, + "grad_norm": 0.16049747169017792, + "learning_rate": 8.468184247937668e-06, + "loss": 0.019, + "step": 79350 + }, + { + "epoch": 7.21323395746228, + "grad_norm": 0.12475612014532089, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0179, + "step": 79360 + }, + { + "epoch": 7.214142883112161, + "grad_norm": 0.13221728801727295, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0198, + "step": 79370 + }, + { + "epoch": 7.215051808762043, + "grad_norm": 0.09393446147441864, + "learning_rate": 8.457022946052628e-06, + "loss": 0.0203, + "step": 79380 + }, + { + "epoch": 7.215960734411925, + "grad_norm": 0.13562211394309998, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0182, + "step": 79390 + }, + { + "epoch": 7.216869660061807, + "grad_norm": 0.09792531281709671, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0175, + "step": 79400 + }, + { + "epoch": 7.217778585711689, + "grad_norm": 0.09248189628124237, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0184, + "step": 79410 + }, + { + "epoch": 7.218687511361571, + "grad_norm": 0.12413708120584488, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0172, + "step": 79420 + }, + { + "epoch": 7.219596437011452, + "grad_norm": 0.1261642724275589, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0186, + "step": 79430 + }, + { + "epoch": 7.220505362661334, + "grad_norm": 0.12448976188898087, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0189, + "step": 79440 + }, + { + "epoch": 7.221414288311216, + "grad_norm": 0.12195989489555359, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0191, + "step": 79450 + }, + { + "epoch": 7.222323213961098, + "grad_norm": 0.11750604212284088, + "learning_rate": 8.42728135848896e-06, + "loss": 0.0189, + "step": 79460 + }, + { + "epoch": 7.22323213961098, + "grad_norm": 0.14827343821525574, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0182, + "step": 79470 + }, + { + "epoch": 7.224141065260862, + "grad_norm": 0.1138945072889328, + "learning_rate": 8.419850947498814e-06, + "loss": 0.017, + "step": 79480 + }, + { + "epoch": 7.225049990910744, + "grad_norm": 0.119106724858284, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0203, + "step": 79490 + }, + { + "epoch": 7.225958916560625, + "grad_norm": 0.11239474266767502, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0179, + "step": 79500 + }, + { + "epoch": 7.226867842210507, + "grad_norm": 0.10385091602802277, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0185, + "step": 79510 + }, + { + "epoch": 7.227776767860389, + "grad_norm": 0.12989427149295807, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0186, + "step": 79520 + }, + { + "epoch": 7.228685693510271, + "grad_norm": 0.09781546890735626, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0179, + "step": 79530 + }, + { + "epoch": 7.229594619160153, + "grad_norm": 0.1362602710723877, + "learning_rate": 8.39757171680725e-06, + "loss": 0.0183, + "step": 79540 + }, + { + "epoch": 7.230503544810034, + "grad_norm": 0.14794597029685974, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0225, + "step": 79550 + }, + { + "epoch": 7.231412470459916, + "grad_norm": 0.10173824429512024, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0171, + "step": 79560 + }, + { + "epoch": 7.232321396109798, + "grad_norm": 0.1062910407781601, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0206, + "step": 79570 + }, + { + "epoch": 7.23323032175968, + "grad_norm": 0.13609609007835388, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0194, + "step": 79580 + }, + { + "epoch": 7.234139247409562, + "grad_norm": 0.11150293052196503, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0184, + "step": 79590 + }, + { + "epoch": 7.235048173059444, + "grad_norm": 0.13293065130710602, + "learning_rate": 8.375310536902511e-06, + "loss": 0.0197, + "step": 79600 + }, + { + "epoch": 7.235957098709326, + "grad_norm": 0.172196164727211, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0194, + "step": 79610 + }, + { + "epoch": 7.236866024359207, + "grad_norm": 0.12894299626350403, + "learning_rate": 8.36789416540753e-06, + "loss": 0.0199, + "step": 79620 + }, + { + "epoch": 7.237774950009089, + "grad_norm": 0.14866860210895538, + "learning_rate": 8.364186735016254e-06, + "loss": 0.0191, + "step": 79630 + }, + { + "epoch": 7.238683875658971, + "grad_norm": 0.17843899130821228, + "learning_rate": 8.360479808571197e-06, + "loss": 0.02, + "step": 79640 + }, + { + "epoch": 7.239592801308853, + "grad_norm": 0.13773773610591888, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0185, + "step": 79650 + }, + { + "epoch": 7.240501726958735, + "grad_norm": 0.13404721021652222, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0203, + "step": 79660 + }, + { + "epoch": 7.241410652608616, + "grad_norm": 0.16300086677074432, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0197, + "step": 79670 + }, + { + "epoch": 7.242319578258498, + "grad_norm": 0.1000521332025528, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0182, + "step": 79680 + }, + { + "epoch": 7.24322850390838, + "grad_norm": 0.14664024114608765, + "learning_rate": 8.341952745391034e-06, + "loss": 0.02, + "step": 79690 + }, + { + "epoch": 7.244137429558262, + "grad_norm": 0.10114600509405136, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0182, + "step": 79700 + }, + { + "epoch": 7.245046355208144, + "grad_norm": 0.24546724557876587, + "learning_rate": 8.334545457592674e-06, + "loss": 0.0194, + "step": 79710 + }, + { + "epoch": 7.245955280858026, + "grad_norm": 0.20475803315639496, + "learning_rate": 8.330842572848802e-06, + "loss": 0.017, + "step": 79720 + }, + { + "epoch": 7.246864206507908, + "grad_norm": 0.1420096755027771, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0194, + "step": 79730 + }, + { + "epoch": 7.247773132157789, + "grad_norm": 0.12240656465291977, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0194, + "step": 79740 + }, + { + "epoch": 7.248682057807671, + "grad_norm": 0.10722228139638901, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0203, + "step": 79750 + }, + { + "epoch": 7.249590983457553, + "grad_norm": 0.1181342676281929, + "learning_rate": 8.316036101469748e-06, + "loss": 0.0169, + "step": 79760 + }, + { + "epoch": 7.250499909107435, + "grad_norm": 0.11948064714670181, + "learning_rate": 8.312335751929666e-06, + "loss": 0.0191, + "step": 79770 + }, + { + "epoch": 7.251408834757317, + "grad_norm": 0.14629380404949188, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0161, + "step": 79780 + }, + { + "epoch": 7.252317760407198, + "grad_norm": 0.12336622923612595, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0163, + "step": 79790 + }, + { + "epoch": 7.25322668605708, + "grad_norm": 0.1268613189458847, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0184, + "step": 79800 + }, + { + "epoch": 7.254135611706962, + "grad_norm": 0.1175103709101677, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0173, + "step": 79810 + }, + { + "epoch": 7.255044537356844, + "grad_norm": 0.11624196171760559, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0167, + "step": 79820 + }, + { + "epoch": 7.255953463006726, + "grad_norm": 0.16055426001548767, + "learning_rate": 8.290144330085352e-06, + "loss": 0.0185, + "step": 79830 + }, + { + "epoch": 7.256862388656608, + "grad_norm": 0.11263524740934372, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0188, + "step": 79840 + }, + { + "epoch": 7.25777131430649, + "grad_norm": 0.12243378162384033, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0213, + "step": 79850 + }, + { + "epoch": 7.258680239956371, + "grad_norm": 0.11008350551128387, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0211, + "step": 79860 + }, + { + "epoch": 7.259589165606253, + "grad_norm": 0.09777598083019257, + "learning_rate": 8.275360236495652e-06, + "loss": 0.0176, + "step": 79870 + }, + { + "epoch": 7.260498091256135, + "grad_norm": 0.11690101027488708, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0218, + "step": 79880 + }, + { + "epoch": 7.261407016906017, + "grad_norm": 0.15320102870464325, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0189, + "step": 79890 + }, + { + "epoch": 7.262315942555899, + "grad_norm": 0.12162932753562927, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0159, + "step": 79900 + }, + { + "epoch": 7.263224868205781, + "grad_norm": 0.1726498305797577, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0188, + "step": 79910 + }, + { + "epoch": 7.264133793855662, + "grad_norm": 0.12582793831825256, + "learning_rate": 8.256891612181913e-06, + "loss": 0.018, + "step": 79920 + }, + { + "epoch": 7.265042719505544, + "grad_norm": 0.14230899512767792, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0174, + "step": 79930 + }, + { + "epoch": 7.265951645155426, + "grad_norm": 0.15926700830459595, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0207, + "step": 79940 + }, + { + "epoch": 7.266860570805308, + "grad_norm": 0.4182406961917877, + "learning_rate": 8.24581658048461e-06, + "loss": 0.015, + "step": 79950 + }, + { + "epoch": 7.26776949645519, + "grad_norm": 0.09850619733333588, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0171, + "step": 79960 + }, + { + "epoch": 7.268678422105072, + "grad_norm": 0.14649108052253723, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0198, + "step": 79970 + }, + { + "epoch": 7.269587347754953, + "grad_norm": 0.14292094111442566, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0186, + "step": 79980 + }, + { + "epoch": 7.270496273404835, + "grad_norm": 0.13145840167999268, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0189, + "step": 79990 + }, + { + "epoch": 7.271405199054717, + "grad_norm": 0.08916206657886505, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0186, + "step": 80000 + }, + { + "epoch": 7.272314124704599, + "grad_norm": 0.13626989722251892, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0195, + "step": 80010 + }, + { + "epoch": 7.273223050354481, + "grad_norm": 0.12904825806617737, + "learning_rate": 8.219992805682155e-06, + "loss": 0.02, + "step": 80020 + }, + { + "epoch": 7.274131976004363, + "grad_norm": 0.13143837451934814, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0205, + "step": 80030 + }, + { + "epoch": 7.275040901654244, + "grad_norm": 0.16759337484836578, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0185, + "step": 80040 + }, + { + "epoch": 7.275949827304126, + "grad_norm": 0.11724188178777695, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0182, + "step": 80050 + }, + { + "epoch": 7.276858752954008, + "grad_norm": 0.13814930617809296, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0182, + "step": 80060 + }, + { + "epoch": 7.27776767860389, + "grad_norm": 0.09655523300170898, + "learning_rate": 8.20156269355164e-06, + "loss": 0.0184, + "step": 80070 + }, + { + "epoch": 7.278676604253772, + "grad_norm": 0.1098516434431076, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0168, + "step": 80080 + }, + { + "epoch": 7.279585529903654, + "grad_norm": 0.17857146263122559, + "learning_rate": 8.194194260815338e-06, + "loss": 0.018, + "step": 80090 + }, + { + "epoch": 7.280494455553535, + "grad_norm": 0.15816272795200348, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0199, + "step": 80100 + }, + { + "epoch": 7.281403381203417, + "grad_norm": 0.12267372757196426, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0194, + "step": 80110 + }, + { + "epoch": 7.282312306853299, + "grad_norm": 0.14203505218029022, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0165, + "step": 80120 + }, + { + "epoch": 7.283221232503181, + "grad_norm": 0.12358052283525467, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0168, + "step": 80130 + }, + { + "epoch": 7.284130158153063, + "grad_norm": 0.09433343261480331, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0179, + "step": 80140 + }, + { + "epoch": 7.285039083802945, + "grad_norm": 0.14448687434196472, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0189, + "step": 80150 + }, + { + "epoch": 7.285948009452826, + "grad_norm": 0.1060294583439827, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0172, + "step": 80160 + }, + { + "epoch": 7.286856935102708, + "grad_norm": 0.12430369853973389, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0169, + "step": 80170 + }, + { + "epoch": 7.28776586075259, + "grad_norm": 0.10736590623855591, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0185, + "step": 80180 + }, + { + "epoch": 7.288674786402472, + "grad_norm": 0.10252165049314499, + "learning_rate": 8.15738315323984e-06, + "loss": 0.0171, + "step": 80190 + }, + { + "epoch": 7.289583712052354, + "grad_norm": 0.08228420466184616, + "learning_rate": 8.153704895955e-06, + "loss": 0.0178, + "step": 80200 + }, + { + "epoch": 7.290492637702236, + "grad_norm": 0.08103550225496292, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0167, + "step": 80210 + }, + { + "epoch": 7.291401563352117, + "grad_norm": 0.13566935062408447, + "learning_rate": 8.146349941457401e-06, + "loss": 0.0194, + "step": 80220 + }, + { + "epoch": 7.292310489001999, + "grad_norm": 0.12559391558170319, + "learning_rate": 8.142673244803215e-06, + "loss": 0.016, + "step": 80230 + }, + { + "epoch": 7.293219414651881, + "grad_norm": 0.10522197932004929, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0184, + "step": 80240 + }, + { + "epoch": 7.294128340301763, + "grad_norm": 0.10060597956180573, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0172, + "step": 80250 + }, + { + "epoch": 7.295037265951645, + "grad_norm": 0.30376216769218445, + "learning_rate": 8.131646280569605e-06, + "loss": 0.018, + "step": 80260 + }, + { + "epoch": 7.295946191601527, + "grad_norm": 0.11012967675924301, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0187, + "step": 80270 + }, + { + "epoch": 7.2968551172514085, + "grad_norm": 0.13917572796344757, + "learning_rate": 8.124297578645989e-06, + "loss": 0.018, + "step": 80280 + }, + { + "epoch": 7.29776404290129, + "grad_norm": 0.14058053493499756, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0211, + "step": 80290 + }, + { + "epoch": 7.298672968551172, + "grad_norm": 0.1308629959821701, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0161, + "step": 80300 + }, + { + "epoch": 7.299581894201054, + "grad_norm": 0.08782170712947845, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0166, + "step": 80310 + }, + { + "epoch": 7.300490819850936, + "grad_norm": 0.09543052315711975, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0183, + "step": 80320 + }, + { + "epoch": 7.301399745500818, + "grad_norm": 0.1351492702960968, + "learning_rate": 8.10593496659236e-06, + "loss": 0.02, + "step": 80330 + }, + { + "epoch": 7.3023086711506995, + "grad_norm": 0.13093966245651245, + "learning_rate": 8.102264013742442e-06, + "loss": 0.017, + "step": 80340 + }, + { + "epoch": 7.3032175968005815, + "grad_norm": 0.1177465170621872, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0183, + "step": 80350 + }, + { + "epoch": 7.3041265224504635, + "grad_norm": 0.10087449848651886, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0185, + "step": 80360 + }, + { + "epoch": 7.305035448100345, + "grad_norm": 0.11721821129322052, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0198, + "step": 80370 + }, + { + "epoch": 7.305944373750227, + "grad_norm": 0.1318911463022232, + "learning_rate": 8.087585443505305e-06, + "loss": 0.0191, + "step": 80380 + }, + { + "epoch": 7.306853299400109, + "grad_norm": 0.1055622324347496, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0191, + "step": 80390 + }, + { + "epoch": 7.3077622250499905, + "grad_norm": 0.14380206167697906, + "learning_rate": 8.0802493069857e-06, + "loss": 0.0177, + "step": 80400 + }, + { + "epoch": 7.3086711506998725, + "grad_norm": 0.12740807235240936, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0195, + "step": 80410 + }, + { + "epoch": 7.3095800763497545, + "grad_norm": 0.10318997502326965, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0188, + "step": 80420 + }, + { + "epoch": 7.3104890019996365, + "grad_norm": 0.13479141891002655, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0205, + "step": 80430 + }, + { + "epoch": 7.3113979276495185, + "grad_norm": 0.09513901174068451, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0195, + "step": 80440 + }, + { + "epoch": 7.3123068532994004, + "grad_norm": 0.1247163861989975, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0183, + "step": 80450 + }, + { + "epoch": 7.3132157789492815, + "grad_norm": 0.09524582326412201, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0188, + "step": 80460 + }, + { + "epoch": 7.3141247045991635, + "grad_norm": 0.12162724882364273, + "learning_rate": 8.054589397331757e-06, + "loss": 0.017, + "step": 80470 + }, + { + "epoch": 7.3150336302490455, + "grad_norm": 0.14451755583286285, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0202, + "step": 80480 + }, + { + "epoch": 7.3159425558989275, + "grad_norm": 0.11329764872789383, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0176, + "step": 80490 + }, + { + "epoch": 7.3168514815488095, + "grad_norm": 0.13617119193077087, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0181, + "step": 80500 + }, + { + "epoch": 7.3177604071986915, + "grad_norm": 0.10638419538736343, + "learning_rate": 8.03993818976936e-06, + "loss": 0.019, + "step": 80510 + }, + { + "epoch": 7.318669332848573, + "grad_norm": 0.15733586251735687, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0167, + "step": 80520 + }, + { + "epoch": 7.319578258498455, + "grad_norm": 0.0918661579489708, + "learning_rate": 8.03261575630017e-06, + "loss": 0.0159, + "step": 80530 + }, + { + "epoch": 7.3204871841483365, + "grad_norm": 0.14125384390354156, + "learning_rate": 8.028955333116846e-06, + "loss": 0.0188, + "step": 80540 + }, + { + "epoch": 7.3213961097982185, + "grad_norm": 0.12613317370414734, + "learning_rate": 8.025295439338347e-06, + "loss": 0.0172, + "step": 80550 + }, + { + "epoch": 7.3223050354481005, + "grad_norm": 0.11595331132411957, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0179, + "step": 80560 + }, + { + "epoch": 7.3232139610979825, + "grad_norm": 0.13974516093730927, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0176, + "step": 80570 + }, + { + "epoch": 7.324122886747864, + "grad_norm": 0.13257832825183868, + "learning_rate": 8.01431893721115e-06, + "loss": 0.0178, + "step": 80580 + }, + { + "epoch": 7.325031812397746, + "grad_norm": 0.11115431785583496, + "learning_rate": 8.01066116383105e-06, + "loss": 0.0182, + "step": 80590 + }, + { + "epoch": 7.325940738047628, + "grad_norm": 0.12945111095905304, + "learning_rate": 8.00700392124512e-06, + "loss": 0.0178, + "step": 80600 + }, + { + "epoch": 7.32684966369751, + "grad_norm": 0.1716475933790207, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0171, + "step": 80610 + }, + { + "epoch": 7.3277585893473915, + "grad_norm": 0.10857147723436356, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0184, + "step": 80620 + }, + { + "epoch": 7.3286675149972735, + "grad_norm": 0.10974922776222229, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0201, + "step": 80630 + }, + { + "epoch": 7.3295764406471555, + "grad_norm": 0.15332840383052826, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0186, + "step": 80640 + }, + { + "epoch": 7.330485366297037, + "grad_norm": 0.14058507978916168, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0188, + "step": 80650 + }, + { + "epoch": 7.331394291946919, + "grad_norm": 0.14213313162326813, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0182, + "step": 80660 + }, + { + "epoch": 7.332303217596801, + "grad_norm": 0.16807588934898376, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0212, + "step": 80670 + }, + { + "epoch": 7.333212143246683, + "grad_norm": 0.11778382211923599, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0183, + "step": 80680 + }, + { + "epoch": 7.334121068896565, + "grad_norm": 0.12198170274496078, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0225, + "step": 80690 + }, + { + "epoch": 7.335029994546446, + "grad_norm": 0.13260310888290405, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0177, + "step": 80700 + }, + { + "epoch": 7.335938920196328, + "grad_norm": 0.1273895800113678, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0191, + "step": 80710 + }, + { + "epoch": 7.33684784584621, + "grad_norm": 0.10857585072517395, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0156, + "step": 80720 + }, + { + "epoch": 7.337756771496092, + "grad_norm": 0.109218068420887, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0183, + "step": 80730 + }, + { + "epoch": 7.338665697145974, + "grad_norm": 0.14948055148124695, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0186, + "step": 80740 + }, + { + "epoch": 7.339574622795856, + "grad_norm": 0.14253520965576172, + "learning_rate": 7.952209166524571e-06, + "loss": 0.0187, + "step": 80750 + }, + { + "epoch": 7.340483548445738, + "grad_norm": 0.2969891428947449, + "learning_rate": 7.948560454390814e-06, + "loss": 0.0156, + "step": 80760 + }, + { + "epoch": 7.341392474095619, + "grad_norm": 0.12475724518299103, + "learning_rate": 7.94491227776735e-06, + "loss": 0.0166, + "step": 80770 + }, + { + "epoch": 7.342301399745501, + "grad_norm": 0.12525981664657593, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0189, + "step": 80780 + }, + { + "epoch": 7.343210325395383, + "grad_norm": 0.1587962508201599, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0175, + "step": 80790 + }, + { + "epoch": 7.344119251045265, + "grad_norm": 0.11824716627597809, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0184, + "step": 80800 + }, + { + "epoch": 7.345028176695147, + "grad_norm": 0.14412260055541992, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0213, + "step": 80810 + }, + { + "epoch": 7.345937102345028, + "grad_norm": 0.11425802111625671, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0188, + "step": 80820 + }, + { + "epoch": 7.34684602799491, + "grad_norm": 0.13297384977340698, + "learning_rate": 7.923034479255065e-06, + "loss": 0.0183, + "step": 80830 + }, + { + "epoch": 7.347754953644792, + "grad_norm": 0.10304471105337143, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0177, + "step": 80840 + }, + { + "epoch": 7.348663879294674, + "grad_norm": 0.10145886242389679, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0189, + "step": 80850 + }, + { + "epoch": 7.349572804944556, + "grad_norm": 0.11915354430675507, + "learning_rate": 7.912102831820231e-06, + "loss": 0.022, + "step": 80860 + }, + { + "epoch": 7.350481730594438, + "grad_norm": 0.12636497616767883, + "learning_rate": 7.908460025531051e-06, + "loss": 0.0199, + "step": 80870 + }, + { + "epoch": 7.35139065624432, + "grad_norm": 0.14849723875522614, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0182, + "step": 80880 + }, + { + "epoch": 7.352299581894201, + "grad_norm": 0.09997159242630005, + "learning_rate": 7.901176028896274e-06, + "loss": 0.018, + "step": 80890 + }, + { + "epoch": 7.353208507544083, + "grad_norm": 0.11821995675563812, + "learning_rate": 7.897534839103856e-06, + "loss": 0.018, + "step": 80900 + }, + { + "epoch": 7.354117433193965, + "grad_norm": 0.11198028177022934, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0199, + "step": 80910 + }, + { + "epoch": 7.355026358843847, + "grad_norm": 0.13425785303115845, + "learning_rate": 7.89025407795157e-06, + "loss": 0.0182, + "step": 80920 + }, + { + "epoch": 7.355935284493729, + "grad_norm": 0.1277696043252945, + "learning_rate": 7.886614507144628e-06, + "loss": 0.0187, + "step": 80930 + }, + { + "epoch": 7.35684421014361, + "grad_norm": 0.14091743528842926, + "learning_rate": 7.882975476552372e-06, + "loss": 0.0197, + "step": 80940 + }, + { + "epoch": 7.357753135793492, + "grad_norm": 0.14284229278564453, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0151, + "step": 80950 + }, + { + "epoch": 7.358662061443374, + "grad_norm": 0.11236590892076492, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0176, + "step": 80960 + }, + { + "epoch": 7.359570987093256, + "grad_norm": 0.13829205930233002, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0183, + "step": 80970 + }, + { + "epoch": 7.360479912743138, + "grad_norm": 0.14597192406654358, + "learning_rate": 7.86842476185677e-06, + "loss": 0.015, + "step": 80980 + }, + { + "epoch": 7.36138883839302, + "grad_norm": 0.10719352960586548, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0161, + "step": 80990 + }, + { + "epoch": 7.362297764042902, + "grad_norm": 0.15136538445949554, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0181, + "step": 81000 + }, + { + "epoch": 7.363206689692783, + "grad_norm": 0.21677738428115845, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0156, + "step": 81010 + }, + { + "epoch": 7.364115615342665, + "grad_norm": 0.10918114334344864, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0205, + "step": 81020 + }, + { + "epoch": 7.365024540992547, + "grad_norm": 0.12600086629390717, + "learning_rate": 7.850248556468741e-06, + "loss": 0.017, + "step": 81030 + }, + { + "epoch": 7.365933466642429, + "grad_norm": 0.20611310005187988, + "learning_rate": 7.846614943216361e-06, + "loss": 0.019, + "step": 81040 + }, + { + "epoch": 7.366842392292311, + "grad_norm": 0.13398697972297668, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0192, + "step": 81050 + }, + { + "epoch": 7.367751317942193, + "grad_norm": 0.1216828003525734, + "learning_rate": 7.839349346744675e-06, + "loss": 0.0196, + "step": 81060 + }, + { + "epoch": 7.368660243592074, + "grad_norm": 0.1182580515742302, + "learning_rate": 7.835717364077138e-06, + "loss": 0.0195, + "step": 81070 + }, + { + "epoch": 7.369569169241956, + "grad_norm": 0.13920296728610992, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0166, + "step": 81080 + }, + { + "epoch": 7.370478094891838, + "grad_norm": 0.11404740810394287, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0184, + "step": 81090 + }, + { + "epoch": 7.37138702054172, + "grad_norm": 0.11923746764659882, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0169, + "step": 81100 + }, + { + "epoch": 7.372295946191602, + "grad_norm": 0.10484892874956131, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0178, + "step": 81110 + }, + { + "epoch": 7.373204871841484, + "grad_norm": 0.14065168797969818, + "learning_rate": 7.817565617454305e-06, + "loss": 0.0185, + "step": 81120 + }, + { + "epoch": 7.374113797491365, + "grad_norm": 0.10378692299127579, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0181, + "step": 81130 + }, + { + "epoch": 7.375022723141247, + "grad_norm": 0.11013565957546234, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0194, + "step": 81140 + }, + { + "epoch": 7.375931648791129, + "grad_norm": 0.12722913920879364, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0175, + "step": 81150 + }, + { + "epoch": 7.376840574441011, + "grad_norm": 0.10543165355920792, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0181, + "step": 81160 + }, + { + "epoch": 7.377749500090893, + "grad_norm": 0.12515391409397125, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0176, + "step": 81170 + }, + { + "epoch": 7.378658425740775, + "grad_norm": 0.08778976649045944, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0162, + "step": 81180 + }, + { + "epoch": 7.379567351390656, + "grad_norm": 0.13668833673000336, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0184, + "step": 81190 + }, + { + "epoch": 7.380476277040538, + "grad_norm": 0.19681404531002045, + "learning_rate": 7.788551201021787e-06, + "loss": 0.0177, + "step": 81200 + }, + { + "epoch": 7.38138520269042, + "grad_norm": 0.13456031680107117, + "learning_rate": 7.784926860555315e-06, + "loss": 0.0197, + "step": 81210 + }, + { + "epoch": 7.382294128340302, + "grad_norm": 0.15017536282539368, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0168, + "step": 81220 + }, + { + "epoch": 7.383203053990184, + "grad_norm": 0.11664550006389618, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0205, + "step": 81230 + }, + { + "epoch": 7.384111979640066, + "grad_norm": 0.08820391446352005, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0178, + "step": 81240 + }, + { + "epoch": 7.385020905289947, + "grad_norm": 0.1305295079946518, + "learning_rate": 7.770434980813133e-06, + "loss": 0.018, + "step": 81250 + }, + { + "epoch": 7.385929830939829, + "grad_norm": 0.10517986118793488, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0173, + "step": 81260 + }, + { + "epoch": 7.386838756589711, + "grad_norm": 0.13327820599079132, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0169, + "step": 81270 + }, + { + "epoch": 7.387747682239593, + "grad_norm": 0.14537282288074493, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0179, + "step": 81280 + }, + { + "epoch": 7.388656607889475, + "grad_norm": 0.15470406413078308, + "learning_rate": 7.75595188567416e-06, + "loss": 0.018, + "step": 81290 + }, + { + "epoch": 7.389565533539357, + "grad_norm": 0.14080463349819183, + "learning_rate": 7.7523324865462e-06, + "loss": 0.0168, + "step": 81300 + }, + { + "epoch": 7.390474459189238, + "grad_norm": 0.1053541824221611, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0176, + "step": 81310 + }, + { + "epoch": 7.39138338483912, + "grad_norm": 0.25327399373054504, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0168, + "step": 81320 + }, + { + "epoch": 7.392292310489002, + "grad_norm": 0.1260490119457245, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0182, + "step": 81330 + }, + { + "epoch": 7.393201236138884, + "grad_norm": 0.13882111012935638, + "learning_rate": 7.737860396907623e-06, + "loss": 0.0192, + "step": 81340 + }, + { + "epoch": 7.394110161788766, + "grad_norm": 0.12275698781013489, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0173, + "step": 81350 + }, + { + "epoch": 7.395019087438648, + "grad_norm": 0.13745521008968353, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0152, + "step": 81360 + }, + { + "epoch": 7.395928013088529, + "grad_norm": 0.12935079634189606, + "learning_rate": 7.72701211958842e-06, + "loss": 0.0185, + "step": 81370 + }, + { + "epoch": 7.396836938738411, + "grad_norm": 0.10068066418170929, + "learning_rate": 7.723397131453438e-06, + "loss": 0.017, + "step": 81380 + }, + { + "epoch": 7.397745864388293, + "grad_norm": 0.13958625495433807, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0188, + "step": 81390 + }, + { + "epoch": 7.398654790038175, + "grad_norm": 0.11333927512168884, + "learning_rate": 7.716168813287858e-06, + "loss": 0.019, + "step": 81400 + }, + { + "epoch": 7.399563715688057, + "grad_norm": 0.09801704436540604, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0213, + "step": 81410 + }, + { + "epoch": 7.400472641337939, + "grad_norm": 0.16883249580860138, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0171, + "step": 81420 + }, + { + "epoch": 7.40138156698782, + "grad_norm": 0.1008836105465889, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0169, + "step": 81430 + }, + { + "epoch": 7.402290492637702, + "grad_norm": 0.13500092923641205, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0189, + "step": 81440 + }, + { + "epoch": 7.403199418287584, + "grad_norm": 0.1663246601819992, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0194, + "step": 81450 + }, + { + "epoch": 7.404108343937466, + "grad_norm": 0.14442679286003113, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0171, + "step": 81460 + }, + { + "epoch": 7.405017269587348, + "grad_norm": 0.14449600875377655, + "learning_rate": 7.690887138617335e-06, + "loss": 0.0201, + "step": 81470 + }, + { + "epoch": 7.40592619523723, + "grad_norm": 0.10595856606960297, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0186, + "step": 81480 + }, + { + "epoch": 7.406835120887111, + "grad_norm": 0.09751391410827637, + "learning_rate": 7.68366879459393e-06, + "loss": 0.018, + "step": 81490 + }, + { + "epoch": 7.407744046536993, + "grad_norm": 0.14745530486106873, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0173, + "step": 81500 + }, + { + "epoch": 7.408652972186875, + "grad_norm": 0.11560803651809692, + "learning_rate": 7.676452673078779e-06, + "loss": 0.017, + "step": 81510 + }, + { + "epoch": 7.409561897836757, + "grad_norm": 0.15374800562858582, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0185, + "step": 81520 + }, + { + "epoch": 7.410470823486639, + "grad_norm": 0.1547999382019043, + "learning_rate": 7.669238776263927e-06, + "loss": 0.02, + "step": 81530 + }, + { + "epoch": 7.411379749136521, + "grad_norm": 0.10602646321058273, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0183, + "step": 81540 + }, + { + "epoch": 7.412288674786402, + "grad_norm": 0.12639141082763672, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0196, + "step": 81550 + }, + { + "epoch": 7.413197600436284, + "grad_norm": 0.15046745538711548, + "learning_rate": 7.658422107148222e-06, + "loss": 0.017, + "step": 81560 + }, + { + "epoch": 7.414106526086166, + "grad_norm": 0.10328101366758347, + "learning_rate": 7.654817665500058e-06, + "loss": 0.0196, + "step": 81570 + }, + { + "epoch": 7.415015451736048, + "grad_norm": 0.15389306843280792, + "learning_rate": 7.651213781670016e-06, + "loss": 0.0186, + "step": 81580 + }, + { + "epoch": 7.41592437738593, + "grad_norm": 0.14724692702293396, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0184, + "step": 81590 + }, + { + "epoch": 7.416833303035812, + "grad_norm": 0.13612814247608185, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0176, + "step": 81600 + }, + { + "epoch": 7.417742228685693, + "grad_norm": 0.12994544208049774, + "learning_rate": 7.64040547982533e-06, + "loss": 0.019, + "step": 81610 + }, + { + "epoch": 7.418651154335575, + "grad_norm": 0.11307881772518158, + "learning_rate": 7.636803830004267e-06, + "loss": 0.018, + "step": 81620 + }, + { + "epoch": 7.419560079985457, + "grad_norm": 0.14617414772510529, + "learning_rate": 7.633202739369368e-06, + "loss": 0.018, + "step": 81630 + }, + { + "epoch": 7.420469005635339, + "grad_norm": 0.14355044066905975, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0176, + "step": 81640 + }, + { + "epoch": 7.421377931285221, + "grad_norm": 0.1425418257713318, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0181, + "step": 81650 + }, + { + "epoch": 7.422286856935103, + "grad_norm": 0.10133646428585052, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0183, + "step": 81660 + }, + { + "epoch": 7.423195782584984, + "grad_norm": 0.10788065195083618, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0186, + "step": 81670 + }, + { + "epoch": 7.424104708234866, + "grad_norm": 0.12331493943929672, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0161, + "step": 81680 + }, + { + "epoch": 7.425013633884748, + "grad_norm": 0.11082510650157928, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0186, + "step": 81690 + }, + { + "epoch": 7.42592255953463, + "grad_norm": 0.1544552445411682, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0184, + "step": 81700 + }, + { + "epoch": 7.426831485184512, + "grad_norm": 0.16079558432102203, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.0205, + "step": 81710 + }, + { + "epoch": 7.427740410834394, + "grad_norm": 0.20018747448921204, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0194, + "step": 81720 + }, + { + "epoch": 7.428649336484275, + "grad_norm": 0.1273151934146881, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0179, + "step": 81730 + }, + { + "epoch": 7.429558262134157, + "grad_norm": 0.0937107652425766, + "learning_rate": 7.593627726856365e-06, + "loss": 0.0168, + "step": 81740 + }, + { + "epoch": 7.430467187784039, + "grad_norm": 0.12482091039419174, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0162, + "step": 81750 + }, + { + "epoch": 7.431376113433921, + "grad_norm": 0.15417899191379547, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0183, + "step": 81760 + }, + { + "epoch": 7.432285039083803, + "grad_norm": 0.1331845372915268, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.0205, + "step": 81770 + }, + { + "epoch": 7.433193964733685, + "grad_norm": 0.13778990507125854, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0201, + "step": 81780 + }, + { + "epoch": 7.434102890383567, + "grad_norm": 0.1130346804857254, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0162, + "step": 81790 + }, + { + "epoch": 7.435011816033448, + "grad_norm": 0.14437714219093323, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0191, + "step": 81800 + }, + { + "epoch": 7.43592074168333, + "grad_norm": 0.13782574236392975, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.0193, + "step": 81810 + }, + { + "epoch": 7.436829667333212, + "grad_norm": 0.1489807367324829, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0199, + "step": 81820 + }, + { + "epoch": 7.437738592983094, + "grad_norm": 0.11189691722393036, + "learning_rate": 7.561298776610552e-06, + "loss": 0.017, + "step": 81830 + }, + { + "epoch": 7.438647518632976, + "grad_norm": 0.1358034461736679, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0186, + "step": 81840 + }, + { + "epoch": 7.439556444282857, + "grad_norm": 0.18755601346492767, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0207, + "step": 81850 + }, + { + "epoch": 7.440465369932739, + "grad_norm": 0.15341152250766754, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0172, + "step": 81860 + }, + { + "epoch": 7.441374295582621, + "grad_norm": 0.11405839025974274, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0196, + "step": 81870 + }, + { + "epoch": 7.442283221232503, + "grad_norm": 0.10833624005317688, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0191, + "step": 81880 + }, + { + "epoch": 7.443192146882385, + "grad_norm": 0.13533326983451843, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0169, + "step": 81890 + }, + { + "epoch": 7.444101072532267, + "grad_norm": 0.12281890213489532, + "learning_rate": 7.536185662874997e-06, + "loss": 0.0206, + "step": 81900 + }, + { + "epoch": 7.445009998182149, + "grad_norm": 0.1210232600569725, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0162, + "step": 81910 + }, + { + "epoch": 7.44591892383203, + "grad_norm": 0.21473628282546997, + "learning_rate": 7.529015584883188e-06, + "loss": 0.018, + "step": 81920 + }, + { + "epoch": 7.446827849481912, + "grad_norm": 0.10545940697193146, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.018, + "step": 81930 + }, + { + "epoch": 7.447736775131794, + "grad_norm": 0.12792055308818817, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0195, + "step": 81940 + }, + { + "epoch": 7.448645700781676, + "grad_norm": 0.13892045617103577, + "learning_rate": 7.518264723865572e-06, + "loss": 0.0171, + "step": 81950 + }, + { + "epoch": 7.449554626431558, + "grad_norm": 0.20662783086299896, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0184, + "step": 81960 + }, + { + "epoch": 7.450463552081439, + "grad_norm": 0.15256436169147491, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0201, + "step": 81970 + }, + { + "epoch": 7.451372477731321, + "grad_norm": 0.12232830375432968, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0181, + "step": 81980 + }, + { + "epoch": 7.452281403381203, + "grad_norm": 0.1277163177728653, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0197, + "step": 81990 + }, + { + "epoch": 7.453190329031085, + "grad_norm": 0.14586523175239563, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.0213, + "step": 82000 + }, + { + "epoch": 7.454099254680967, + "grad_norm": 0.12568719685077667, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0197, + "step": 82010 + }, + { + "epoch": 7.455008180330849, + "grad_norm": 0.14303293824195862, + "learning_rate": 7.493199280778683e-06, + "loss": 0.0179, + "step": 82020 + }, + { + "epoch": 7.455917105980731, + "grad_norm": 0.16327112913131714, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0193, + "step": 82030 + }, + { + "epoch": 7.456826031630612, + "grad_norm": 0.1240883469581604, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0174, + "step": 82040 + }, + { + "epoch": 7.457734957280494, + "grad_norm": 0.1607934981584549, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0197, + "step": 82050 + }, + { + "epoch": 7.458643882930376, + "grad_norm": 0.10701407492160797, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.017, + "step": 82060 + }, + { + "epoch": 7.459552808580258, + "grad_norm": 0.11370110511779785, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.0181, + "step": 82070 + }, + { + "epoch": 7.46046173423014, + "grad_norm": 0.09943472594022751, + "learning_rate": 7.471736845323507e-06, + "loss": 0.018, + "step": 82080 + }, + { + "epoch": 7.461370659880021, + "grad_norm": 0.11093783378601074, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.023, + "step": 82090 + }, + { + "epoch": 7.462279585529903, + "grad_norm": 0.14366194605827332, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0163, + "step": 82100 + }, + { + "epoch": 7.463188511179785, + "grad_norm": 0.16006876528263092, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0189, + "step": 82110 + }, + { + "epoch": 7.464097436829667, + "grad_norm": 0.13170969486236572, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0183, + "step": 82120 + }, + { + "epoch": 7.465006362479549, + "grad_norm": 0.10152481496334076, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0188, + "step": 82130 + }, + { + "epoch": 7.465915288129431, + "grad_norm": 0.11366113275289536, + "learning_rate": 7.450294991852334e-06, + "loss": 0.0197, + "step": 82140 + }, + { + "epoch": 7.466824213779313, + "grad_norm": 0.11895178258419037, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0172, + "step": 82150 + }, + { + "epoch": 7.467733139429194, + "grad_norm": 0.21491791307926178, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0169, + "step": 82160 + }, + { + "epoch": 7.468642065079076, + "grad_norm": 0.12497447431087494, + "learning_rate": 7.439581801682075e-06, + "loss": 0.02, + "step": 82170 + }, + { + "epoch": 7.469550990728958, + "grad_norm": 0.0959862470626831, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0161, + "step": 82180 + }, + { + "epoch": 7.47045991637884, + "grad_norm": 0.12703315913677216, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0181, + "step": 82190 + }, + { + "epoch": 7.471368842028722, + "grad_norm": 0.13342733681201935, + "learning_rate": 7.428873778986364e-06, + "loss": 0.019, + "step": 82200 + }, + { + "epoch": 7.472277767678604, + "grad_norm": 0.1553453952074051, + "learning_rate": 7.42530558768056e-06, + "loss": 0.0175, + "step": 82210 + }, + { + "epoch": 7.473186693328485, + "grad_norm": 0.09673822671175003, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0173, + "step": 82220 + }, + { + "epoch": 7.474095618978367, + "grad_norm": 0.1357039511203766, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0166, + "step": 82230 + }, + { + "epoch": 7.475004544628249, + "grad_norm": 0.11456421762704849, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0192, + "step": 82240 + }, + { + "epoch": 7.475913470278131, + "grad_norm": 0.1139589324593544, + "learning_rate": 7.411038577647002e-06, + "loss": 0.0172, + "step": 82250 + }, + { + "epoch": 7.476822395928013, + "grad_norm": 0.12520207464694977, + "learning_rate": 7.407473265290385e-06, + "loss": 0.02, + "step": 82260 + }, + { + "epoch": 7.477731321577895, + "grad_norm": 0.13245108723640442, + "learning_rate": 7.403908529536066e-06, + "loss": 0.0169, + "step": 82270 + }, + { + "epoch": 7.478640247227776, + "grad_norm": 0.1630336046218872, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0179, + "step": 82280 + }, + { + "epoch": 7.479549172877658, + "grad_norm": 0.13107821345329285, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0178, + "step": 82290 + }, + { + "epoch": 7.48045809852754, + "grad_norm": 0.13286180794239044, + "learning_rate": 7.393217784593878e-06, + "loss": 0.018, + "step": 82300 + }, + { + "epoch": 7.481367024177422, + "grad_norm": 0.17827114462852478, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0183, + "step": 82310 + }, + { + "epoch": 7.482275949827304, + "grad_norm": 0.1889968067407608, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0197, + "step": 82320 + }, + { + "epoch": 7.483184875477186, + "grad_norm": 0.11355014145374298, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.0174, + "step": 82330 + }, + { + "epoch": 7.4840938011270675, + "grad_norm": 0.1714038848876953, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0211, + "step": 82340 + }, + { + "epoch": 7.4850027267769494, + "grad_norm": 0.12752191722393036, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0185, + "step": 82350 + }, + { + "epoch": 7.485911652426831, + "grad_norm": 0.3123430013656616, + "learning_rate": 7.371851899504712e-06, + "loss": 0.0183, + "step": 82360 + }, + { + "epoch": 7.486820578076713, + "grad_norm": 0.15780462324619293, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0187, + "step": 82370 + }, + { + "epoch": 7.487729503726595, + "grad_norm": 0.11498873680830002, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0177, + "step": 82380 + }, + { + "epoch": 7.488638429376477, + "grad_norm": 0.10612058639526367, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0159, + "step": 82390 + }, + { + "epoch": 7.4895473550263585, + "grad_norm": 0.23079395294189453, + "learning_rate": 7.357619558656079e-06, + "loss": 0.0165, + "step": 82400 + }, + { + "epoch": 7.4904562806762405, + "grad_norm": 0.20600786805152893, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0205, + "step": 82410 + }, + { + "epoch": 7.4913652063261225, + "grad_norm": 0.5720297694206238, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0172, + "step": 82420 + }, + { + "epoch": 7.4922741319760044, + "grad_norm": 0.14850282669067383, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0182, + "step": 82430 + }, + { + "epoch": 7.493183057625886, + "grad_norm": 0.14684654772281647, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0171, + "step": 82440 + }, + { + "epoch": 7.494091983275768, + "grad_norm": 0.19622130692005157, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0201, + "step": 82450 + }, + { + "epoch": 7.4950009089256495, + "grad_norm": 0.12710566818714142, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0164, + "step": 82460 + }, + { + "epoch": 7.4959098345755315, + "grad_norm": 0.09306564927101135, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0166, + "step": 82470 + }, + { + "epoch": 7.4968187602254135, + "grad_norm": 0.08970934897661209, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0179, + "step": 82480 + }, + { + "epoch": 7.4977276858752955, + "grad_norm": 0.15040263533592224, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0182, + "step": 82490 + }, + { + "epoch": 7.4986366115251775, + "grad_norm": 0.13827607035636902, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0181, + "step": 82500 + }, + { + "epoch": 7.4995455371750595, + "grad_norm": 0.11032095551490784, + "learning_rate": 7.31852855533313e-06, + "loss": 0.019, + "step": 82510 + }, + { + "epoch": 7.5004544628249405, + "grad_norm": 0.11033838242292404, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0173, + "step": 82520 + }, + { + "epoch": 7.5013633884748225, + "grad_norm": 0.13364094495773315, + "learning_rate": 7.311428673304513e-06, + "loss": 0.0166, + "step": 82530 + }, + { + "epoch": 7.5022723141247045, + "grad_norm": 0.18228574097156525, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0177, + "step": 82540 + }, + { + "epoch": 7.5031812397745865, + "grad_norm": 0.10887867212295532, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0171, + "step": 82550 + }, + { + "epoch": 7.5040901654244685, + "grad_norm": 0.10406485199928284, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0174, + "step": 82560 + }, + { + "epoch": 7.5049990910743505, + "grad_norm": 0.18005818128585815, + "learning_rate": 7.297235918158213e-06, + "loss": 0.017, + "step": 82570 + }, + { + "epoch": 7.505908016724232, + "grad_norm": 0.12391812354326248, + "learning_rate": 7.293689191133345e-06, + "loss": 0.0189, + "step": 82580 + }, + { + "epoch": 7.506816942374114, + "grad_norm": 0.1380435973405838, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0177, + "step": 82590 + }, + { + "epoch": 7.5077258680239956, + "grad_norm": 0.13655413687229156, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0192, + "step": 82600 + }, + { + "epoch": 7.5086347936738775, + "grad_norm": 0.14547550678253174, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0183, + "step": 82610 + }, + { + "epoch": 7.5095437193237595, + "grad_norm": 0.1152597963809967, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0184, + "step": 82620 + }, + { + "epoch": 7.5104526449736415, + "grad_norm": 0.14976604282855988, + "learning_rate": 7.275964340046174e-06, + "loss": 0.0192, + "step": 82630 + }, + { + "epoch": 7.511361570623523, + "grad_norm": 0.10800162702798843, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0166, + "step": 82640 + }, + { + "epoch": 7.512270496273405, + "grad_norm": 0.1521609127521515, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0185, + "step": 82650 + }, + { + "epoch": 7.513179421923287, + "grad_norm": 0.14706335961818695, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0179, + "step": 82660 + }, + { + "epoch": 7.514088347573169, + "grad_norm": 0.12763918936252594, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0164, + "step": 82670 + }, + { + "epoch": 7.5149972732230506, + "grad_norm": 0.16779197752475739, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0177, + "step": 82680 + }, + { + "epoch": 7.5159061988729325, + "grad_norm": 0.1475307196378708, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0222, + "step": 82690 + }, + { + "epoch": 7.5168151245228145, + "grad_norm": 0.09389976412057877, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.0185, + "step": 82700 + }, + { + "epoch": 7.517724050172696, + "grad_norm": 0.11036904156208038, + "learning_rate": 7.24763509494671e-06, + "loss": 0.0191, + "step": 82710 + }, + { + "epoch": 7.518632975822578, + "grad_norm": 0.19237945973873138, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0142, + "step": 82720 + }, + { + "epoch": 7.51954190147246, + "grad_norm": 0.1356520652770996, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0156, + "step": 82730 + }, + { + "epoch": 7.520450827122342, + "grad_norm": 0.10431323945522308, + "learning_rate": 7.237021334866599e-06, + "loss": 0.0176, + "step": 82740 + }, + { + "epoch": 7.521359752772224, + "grad_norm": 0.1265321671962738, + "learning_rate": 7.2334845935757e-06, + "loss": 0.0199, + "step": 82750 + }, + { + "epoch": 7.522268678422105, + "grad_norm": 0.11413589864969254, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0179, + "step": 82760 + }, + { + "epoch": 7.523177604071987, + "grad_norm": 0.09951414912939072, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.017, + "step": 82770 + }, + { + "epoch": 7.524086529721869, + "grad_norm": 0.10784230381250381, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0164, + "step": 82780 + }, + { + "epoch": 7.524995455371751, + "grad_norm": 0.151466965675354, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0179, + "step": 82790 + }, + { + "epoch": 7.525904381021633, + "grad_norm": 0.13656306266784668, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0193, + "step": 82800 + }, + { + "epoch": 7.526813306671515, + "grad_norm": 0.13687840104103088, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0218, + "step": 82810 + }, + { + "epoch": 7.527722232321397, + "grad_norm": 0.16196642816066742, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0189, + "step": 82820 + }, + { + "epoch": 7.528631157971278, + "grad_norm": 0.17591720819473267, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0176, + "step": 82830 + }, + { + "epoch": 7.52954008362116, + "grad_norm": 0.13382133841514587, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0179, + "step": 82840 + }, + { + "epoch": 7.530449009271042, + "grad_norm": 0.12309568375349045, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0162, + "step": 82850 + }, + { + "epoch": 7.531357934920924, + "grad_norm": 0.12575829029083252, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0171, + "step": 82860 + }, + { + "epoch": 7.532266860570806, + "grad_norm": 0.17101998627185822, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0187, + "step": 82870 + }, + { + "epoch": 7.533175786220687, + "grad_norm": 0.1753026843070984, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0189, + "step": 82880 + }, + { + "epoch": 7.534084711870569, + "grad_norm": 0.1368037760257721, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0225, + "step": 82890 + }, + { + "epoch": 7.534993637520451, + "grad_norm": 0.12006287276744843, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0176, + "step": 82900 + }, + { + "epoch": 7.535902563170333, + "grad_norm": 0.12927888333797455, + "learning_rate": 7.176977130346983e-06, + "loss": 0.017, + "step": 82910 + }, + { + "epoch": 7.536811488820215, + "grad_norm": 0.12649324536323547, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0182, + "step": 82920 + }, + { + "epoch": 7.537720414470097, + "grad_norm": 0.1253509223461151, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0189, + "step": 82930 + }, + { + "epoch": 7.538629340119979, + "grad_norm": 0.10133042931556702, + "learning_rate": 7.166398879937302e-06, + "loss": 0.0184, + "step": 82940 + }, + { + "epoch": 7.53953826576986, + "grad_norm": 0.19001446664333344, + "learning_rate": 7.162873985931023e-06, + "loss": 0.0188, + "step": 82950 + }, + { + "epoch": 7.540447191419742, + "grad_norm": 0.11389331519603729, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0168, + "step": 82960 + }, + { + "epoch": 7.541356117069624, + "grad_norm": 0.13858793675899506, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0171, + "step": 82970 + }, + { + "epoch": 7.542265042719506, + "grad_norm": 0.15800423920154572, + "learning_rate": 7.152302876050733e-06, + "loss": 0.019, + "step": 82980 + }, + { + "epoch": 7.543173968369388, + "grad_norm": 0.1828904002904892, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0198, + "step": 82990 + }, + { + "epoch": 7.544082894019269, + "grad_norm": 0.1577499955892563, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0193, + "step": 83000 + }, + { + "epoch": 7.544991819669151, + "grad_norm": 0.12125390022993088, + "learning_rate": 7.141737129997433e-06, + "loss": 0.018, + "step": 83010 + }, + { + "epoch": 7.545900745319033, + "grad_norm": 0.09374529868364334, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0166, + "step": 83020 + }, + { + "epoch": 7.546809670968915, + "grad_norm": 0.10587669163942337, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0176, + "step": 83030 + }, + { + "epoch": 7.547718596618797, + "grad_norm": 0.10580530017614365, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0161, + "step": 83040 + }, + { + "epoch": 7.548627522268679, + "grad_norm": 0.10150222480297089, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0192, + "step": 83050 + }, + { + "epoch": 7.549536447918561, + "grad_norm": 0.2596376836299896, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0178, + "step": 83060 + }, + { + "epoch": 7.550445373568442, + "grad_norm": 0.15499675273895264, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0172, + "step": 83070 + }, + { + "epoch": 7.551354299218324, + "grad_norm": 0.1210860013961792, + "learning_rate": 7.117104622425724e-06, + "loss": 0.0187, + "step": 83080 + }, + { + "epoch": 7.552263224868206, + "grad_norm": 0.12537750601768494, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0191, + "step": 83090 + }, + { + "epoch": 7.553172150518088, + "grad_norm": 0.12795700132846832, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.0182, + "step": 83100 + }, + { + "epoch": 7.55408107616797, + "grad_norm": 0.1424245834350586, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0223, + "step": 83110 + }, + { + "epoch": 7.554990001817851, + "grad_norm": 0.10852713882923126, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0172, + "step": 83120 + }, + { + "epoch": 7.555898927467733, + "grad_norm": 0.11502302438020706, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0192, + "step": 83130 + }, + { + "epoch": 7.556807853117615, + "grad_norm": 0.09847883135080338, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0165, + "step": 83140 + }, + { + "epoch": 7.557716778767497, + "grad_norm": 0.17534378170967102, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0184, + "step": 83150 + }, + { + "epoch": 7.558625704417379, + "grad_norm": 0.1504884511232376, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0173, + "step": 83160 + }, + { + "epoch": 7.559534630067261, + "grad_norm": 0.14362972974777222, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0182, + "step": 83170 + }, + { + "epoch": 7.560443555717143, + "grad_norm": 0.11945541203022003, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0181, + "step": 83180 + }, + { + "epoch": 7.561352481367024, + "grad_norm": 0.12525597214698792, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0167, + "step": 83190 + }, + { + "epoch": 7.562261407016906, + "grad_norm": 0.11260256171226501, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0189, + "step": 83200 + }, + { + "epoch": 7.563170332666788, + "grad_norm": 0.13834881782531738, + "learning_rate": 7.071436431318798e-06, + "loss": 0.0183, + "step": 83210 + }, + { + "epoch": 7.56407925831667, + "grad_norm": 0.14227959513664246, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0189, + "step": 83220 + }, + { + "epoch": 7.564988183966552, + "grad_norm": 0.23721221089363098, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0168, + "step": 83230 + }, + { + "epoch": 7.565897109616433, + "grad_norm": 0.16164079308509827, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0182, + "step": 83240 + }, + { + "epoch": 7.566806035266315, + "grad_norm": 0.10480359941720963, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0193, + "step": 83250 + }, + { + "epoch": 7.567714960916197, + "grad_norm": 0.13968272507190704, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0168, + "step": 83260 + }, + { + "epoch": 7.568623886566079, + "grad_norm": 0.18842485547065735, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0183, + "step": 83270 + }, + { + "epoch": 7.569532812215961, + "grad_norm": 0.1553032249212265, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0144, + "step": 83280 + }, + { + "epoch": 7.570441737865843, + "grad_norm": 0.11681567132472992, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0182, + "step": 83290 + }, + { + "epoch": 7.571350663515725, + "grad_norm": 0.12283319979906082, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0206, + "step": 83300 + }, + { + "epoch": 7.572259589165606, + "grad_norm": 0.135886549949646, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0212, + "step": 83310 + }, + { + "epoch": 7.573168514815488, + "grad_norm": 0.12524376809597015, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0197, + "step": 83320 + }, + { + "epoch": 7.57407744046537, + "grad_norm": 0.1440458595752716, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0184, + "step": 83330 + }, + { + "epoch": 7.574986366115252, + "grad_norm": 0.17547553777694702, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0166, + "step": 83340 + }, + { + "epoch": 7.575895291765134, + "grad_norm": 0.13237664103507996, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0181, + "step": 83350 + }, + { + "epoch": 7.576804217415015, + "grad_norm": 0.10238441079854965, + "learning_rate": 7.018868847443365e-06, + "loss": 0.017, + "step": 83360 + }, + { + "epoch": 7.577713143064897, + "grad_norm": 0.1179216131567955, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.0161, + "step": 83370 + }, + { + "epoch": 7.578622068714779, + "grad_norm": 0.14966729283332825, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0204, + "step": 83380 + }, + { + "epoch": 7.579530994364661, + "grad_norm": 0.11965446919202805, + "learning_rate": 7.008371667026958e-06, + "loss": 0.0181, + "step": 83390 + }, + { + "epoch": 7.580439920014543, + "grad_norm": 0.1216856837272644, + "learning_rate": 7.004873820356754e-06, + "loss": 0.0193, + "step": 83400 + }, + { + "epoch": 7.581348845664425, + "grad_norm": 0.1045636385679245, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0178, + "step": 83410 + }, + { + "epoch": 7.582257771314307, + "grad_norm": 0.13987840712070465, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0167, + "step": 83420 + }, + { + "epoch": 7.583166696964188, + "grad_norm": 0.09323287010192871, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0177, + "step": 83430 + }, + { + "epoch": 7.58407562261407, + "grad_norm": 0.15510432422161102, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0162, + "step": 83440 + }, + { + "epoch": 7.584984548263952, + "grad_norm": 0.12978990375995636, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0184, + "step": 83450 + }, + { + "epoch": 7.585893473913834, + "grad_norm": 0.11081790924072266, + "learning_rate": 6.98389950034815e-06, + "loss": 0.0146, + "step": 83460 + }, + { + "epoch": 7.586802399563716, + "grad_norm": 0.11793234944343567, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0188, + "step": 83470 + }, + { + "epoch": 7.587711325213597, + "grad_norm": 0.10237415879964828, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0169, + "step": 83480 + }, + { + "epoch": 7.588620250863479, + "grad_norm": 0.11477809399366379, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0171, + "step": 83490 + }, + { + "epoch": 7.589529176513361, + "grad_norm": 0.15956519544124603, + "learning_rate": 6.969928792201108e-06, + "loss": 0.019, + "step": 83500 + }, + { + "epoch": 7.590438102163243, + "grad_norm": 0.11693339794874191, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0195, + "step": 83510 + }, + { + "epoch": 7.591347027813125, + "grad_norm": 0.10333875566720963, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.0165, + "step": 83520 + }, + { + "epoch": 7.592255953463007, + "grad_norm": 0.11818675696849823, + "learning_rate": 6.95945716338691e-06, + "loss": 0.017, + "step": 83530 + }, + { + "epoch": 7.593164879112889, + "grad_norm": 0.10551503300666809, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0185, + "step": 83540 + }, + { + "epoch": 7.59407380476277, + "grad_norm": 0.12713153660297394, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.0166, + "step": 83550 + }, + { + "epoch": 7.594982730412652, + "grad_norm": 0.10348194092512131, + "learning_rate": 6.948991030207918e-06, + "loss": 0.0187, + "step": 83560 + }, + { + "epoch": 7.595891656062534, + "grad_norm": 0.12290100008249283, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0164, + "step": 83570 + }, + { + "epoch": 7.596800581712416, + "grad_norm": 0.12493783235549927, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0156, + "step": 83580 + }, + { + "epoch": 7.597709507362298, + "grad_norm": 0.10721595585346222, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0174, + "step": 83590 + }, + { + "epoch": 7.59861843301218, + "grad_norm": 0.14207206666469574, + "learning_rate": 6.935044747098684e-06, + "loss": 0.0194, + "step": 83600 + }, + { + "epoch": 7.599527358662061, + "grad_norm": 0.11483742296695709, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0177, + "step": 83610 + }, + { + "epoch": 7.600436284311943, + "grad_norm": 0.12348119169473648, + "learning_rate": 6.928075279365821e-06, + "loss": 0.0168, + "step": 83620 + }, + { + "epoch": 7.601345209961825, + "grad_norm": 0.14868809282779694, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0188, + "step": 83630 + }, + { + "epoch": 7.602254135611707, + "grad_norm": 0.1805180013179779, + "learning_rate": 6.921108263670561e-06, + "loss": 0.0167, + "step": 83640 + }, + { + "epoch": 7.603163061261589, + "grad_norm": 0.12746453285217285, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0173, + "step": 83650 + }, + { + "epoch": 7.604071986911471, + "grad_norm": 0.12191548943519592, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0165, + "step": 83660 + }, + { + "epoch": 7.604980912561352, + "grad_norm": 0.12334369868040085, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0185, + "step": 83670 + }, + { + "epoch": 7.605889838211234, + "grad_norm": 0.17860476672649384, + "learning_rate": 6.907181596857688e-06, + "loss": 0.018, + "step": 83680 + }, + { + "epoch": 7.606798763861116, + "grad_norm": 0.12756314873695374, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0177, + "step": 83690 + }, + { + "epoch": 7.607707689510998, + "grad_norm": 0.11788446456193924, + "learning_rate": 6.900221949970629e-06, + "loss": 0.017, + "step": 83700 + }, + { + "epoch": 7.60861661516088, + "grad_norm": 0.11324987560510635, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0181, + "step": 83710 + }, + { + "epoch": 7.609525540810762, + "grad_norm": 0.11803227663040161, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0186, + "step": 83720 + }, + { + "epoch": 7.610434466460643, + "grad_norm": 0.10348538309335709, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0176, + "step": 83730 + }, + { + "epoch": 7.611343392110525, + "grad_norm": 0.09438060969114304, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0166, + "step": 83740 + }, + { + "epoch": 7.612252317760407, + "grad_norm": 0.12273791432380676, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0161, + "step": 83750 + }, + { + "epoch": 7.613161243410289, + "grad_norm": 0.14796994626522064, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.016, + "step": 83760 + }, + { + "epoch": 7.614070169060171, + "grad_norm": 0.12360876053571701, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0187, + "step": 83770 + }, + { + "epoch": 7.614979094710053, + "grad_norm": 0.13234741985797882, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0167, + "step": 83780 + }, + { + "epoch": 7.615888020359934, + "grad_norm": 0.09994760900735855, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.0193, + "step": 83790 + }, + { + "epoch": 7.616796946009816, + "grad_norm": 0.13070033490657806, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0165, + "step": 83800 + }, + { + "epoch": 7.617705871659698, + "grad_norm": 0.1331276148557663, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0163, + "step": 83810 + }, + { + "epoch": 7.61861479730958, + "grad_norm": 0.1335492730140686, + "learning_rate": 6.8585158130643e-06, + "loss": 0.017, + "step": 83820 + }, + { + "epoch": 7.619523722959462, + "grad_norm": 0.14997722208499908, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0184, + "step": 83830 + }, + { + "epoch": 7.620432648609344, + "grad_norm": 0.22960437834262848, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0168, + "step": 83840 + }, + { + "epoch": 7.621341574259226, + "grad_norm": 0.10860521346330643, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0176, + "step": 83850 + }, + { + "epoch": 7.622250499909107, + "grad_norm": 0.1370566338300705, + "learning_rate": 6.844633530259747e-06, + "loss": 0.018, + "step": 83860 + }, + { + "epoch": 7.623159425558989, + "grad_norm": 0.1049443930387497, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0187, + "step": 83870 + }, + { + "epoch": 7.624068351208871, + "grad_norm": 0.2065502405166626, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0193, + "step": 83880 + }, + { + "epoch": 7.624977276858753, + "grad_norm": 0.13267849385738373, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0184, + "step": 83890 + }, + { + "epoch": 7.625886202508635, + "grad_norm": 0.13598255813121796, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0176, + "step": 83900 + }, + { + "epoch": 7.626795128158516, + "grad_norm": 0.09693173319101334, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0162, + "step": 83910 + }, + { + "epoch": 7.627704053808398, + "grad_norm": 0.12448810786008835, + "learning_rate": 6.82382869170707e-06, + "loss": 0.017, + "step": 83920 + }, + { + "epoch": 7.62861297945828, + "grad_norm": 0.17824597656726837, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0167, + "step": 83930 + }, + { + "epoch": 7.629521905108162, + "grad_norm": 0.10380107909440994, + "learning_rate": 6.816898710124658e-06, + "loss": 0.018, + "step": 83940 + }, + { + "epoch": 7.630430830758044, + "grad_norm": 0.20309260487556458, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0185, + "step": 83950 + }, + { + "epoch": 7.631339756407926, + "grad_norm": 0.1113014668226242, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.0202, + "step": 83960 + }, + { + "epoch": 7.632248682057808, + "grad_norm": 0.1242624893784523, + "learning_rate": 6.806508399302644e-06, + "loss": 0.017, + "step": 83970 + }, + { + "epoch": 7.633157607707689, + "grad_norm": 0.11247134953737259, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0176, + "step": 83980 + }, + { + "epoch": 7.634066533357571, + "grad_norm": 0.11517330259084702, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0184, + "step": 83990 + }, + { + "epoch": 7.634975459007453, + "grad_norm": 0.14144548773765564, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0168, + "step": 84000 + }, + { + "epoch": 7.635884384657335, + "grad_norm": 0.11541802436113358, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0172, + "step": 84010 + }, + { + "epoch": 7.636793310307217, + "grad_norm": 0.11959941685199738, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0197, + "step": 84020 + }, + { + "epoch": 7.637702235957098, + "grad_norm": 0.12379425019025803, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0181, + "step": 84030 + }, + { + "epoch": 7.63861116160698, + "grad_norm": 0.1148868203163147, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0174, + "step": 84040 + }, + { + "epoch": 7.639520087256862, + "grad_norm": 0.1195734441280365, + "learning_rate": 6.778828301683322e-06, + "loss": 0.02, + "step": 84050 + }, + { + "epoch": 7.640429012906744, + "grad_norm": 0.1452808827161789, + "learning_rate": 6.775371096269681e-06, + "loss": 0.02, + "step": 84060 + }, + { + "epoch": 7.641337938556626, + "grad_norm": 0.1436878889799118, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0163, + "step": 84070 + }, + { + "epoch": 7.642246864206508, + "grad_norm": 0.11058573424816132, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0158, + "step": 84080 + }, + { + "epoch": 7.64315578985639, + "grad_norm": 0.11815822124481201, + "learning_rate": 6.765003228716437e-06, + "loss": 0.0166, + "step": 84090 + }, + { + "epoch": 7.644064715506271, + "grad_norm": 0.1289963722229004, + "learning_rate": 6.761548523302776e-06, + "loss": 0.0177, + "step": 84100 + }, + { + "epoch": 7.644973641156153, + "grad_norm": 0.11246370524168015, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0162, + "step": 84110 + }, + { + "epoch": 7.645882566806035, + "grad_norm": 0.1363385170698166, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0183, + "step": 84120 + }, + { + "epoch": 7.646791492455917, + "grad_norm": 0.1759594827890396, + "learning_rate": 6.75118816204714e-06, + "loss": 0.018, + "step": 84130 + }, + { + "epoch": 7.647700418105799, + "grad_norm": 0.12025181204080582, + "learning_rate": 6.747735960831258e-06, + "loss": 0.0168, + "step": 84140 + }, + { + "epoch": 7.64860934375568, + "grad_norm": 0.0991235300898552, + "learning_rate": 6.74428438632037e-06, + "loss": 0.018, + "step": 84150 + }, + { + "epoch": 7.649518269405562, + "grad_norm": 0.12449491024017334, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0177, + "step": 84160 + }, + { + "epoch": 7.650427195055444, + "grad_norm": 0.13073593378067017, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0178, + "step": 84170 + }, + { + "epoch": 7.651336120705326, + "grad_norm": 0.10608493536710739, + "learning_rate": 6.733933425638679e-06, + "loss": 0.0196, + "step": 84180 + }, + { + "epoch": 7.652245046355208, + "grad_norm": 0.14080114662647247, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0195, + "step": 84190 + }, + { + "epoch": 7.65315397200509, + "grad_norm": 0.1371307522058487, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0162, + "step": 84200 + }, + { + "epoch": 7.654062897654972, + "grad_norm": 0.11431460827589035, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0164, + "step": 84210 + }, + { + "epoch": 7.654971823304853, + "grad_norm": 0.09466101974248886, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0174, + "step": 84220 + }, + { + "epoch": 7.655880748954735, + "grad_norm": 0.13107869029045105, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0168, + "step": 84230 + }, + { + "epoch": 7.656789674604617, + "grad_norm": 0.14452296495437622, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0183, + "step": 84240 + }, + { + "epoch": 7.657698600254499, + "grad_norm": 0.12272067368030548, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0168, + "step": 84250 + }, + { + "epoch": 7.658607525904381, + "grad_norm": 0.14856524765491486, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0202, + "step": 84260 + }, + { + "epoch": 7.659516451554262, + "grad_norm": 0.1148906722664833, + "learning_rate": 6.702914470544795e-06, + "loss": 0.018, + "step": 84270 + }, + { + "epoch": 7.660425377204144, + "grad_norm": 0.11932060122489929, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0173, + "step": 84280 + }, + { + "epoch": 7.661334302854026, + "grad_norm": 0.11122201383113861, + "learning_rate": 6.696028293895167e-06, + "loss": 0.0184, + "step": 84290 + }, + { + "epoch": 7.662243228503908, + "grad_norm": 0.11852619796991348, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.017, + "step": 84300 + }, + { + "epoch": 7.66315215415379, + "grad_norm": 0.13069391250610352, + "learning_rate": 6.689144639772895e-06, + "loss": 0.0189, + "step": 84310 + }, + { + "epoch": 7.664061079803672, + "grad_norm": 0.11921925097703934, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0196, + "step": 84320 + }, + { + "epoch": 7.664970005453554, + "grad_norm": 0.11662229895591736, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0172, + "step": 84330 + }, + { + "epoch": 7.665878931103435, + "grad_norm": 0.15106506645679474, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.0184, + "step": 84340 + }, + { + "epoch": 7.666787856753317, + "grad_norm": 0.16974180936813354, + "learning_rate": 6.675384907473927e-06, + "loss": 0.015, + "step": 84350 + }, + { + "epoch": 7.667696782403199, + "grad_norm": 0.11794698238372803, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0174, + "step": 84360 + }, + { + "epoch": 7.668605708053081, + "grad_norm": 0.1484239548444748, + "learning_rate": 6.668508833477072e-06, + "loss": 0.0197, + "step": 84370 + }, + { + "epoch": 7.669514633702963, + "grad_norm": 0.12106671929359436, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0164, + "step": 84380 + }, + { + "epoch": 7.6704235593528445, + "grad_norm": 0.1380266398191452, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0157, + "step": 84390 + }, + { + "epoch": 7.6713324850027265, + "grad_norm": 0.12471135705709457, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0165, + "step": 84400 + }, + { + "epoch": 7.6722414106526085, + "grad_norm": 0.09447970986366272, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0165, + "step": 84410 + }, + { + "epoch": 7.67315033630249, + "grad_norm": 0.1856907606124878, + "learning_rate": 6.651329725683009e-06, + "loss": 0.015, + "step": 84420 + }, + { + "epoch": 7.674059261952372, + "grad_norm": 0.09413423389196396, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0154, + "step": 84430 + }, + { + "epoch": 7.674968187602254, + "grad_norm": 0.19108878076076508, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0194, + "step": 84440 + }, + { + "epoch": 7.675877113252136, + "grad_norm": 0.10243616998195648, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0188, + "step": 84450 + }, + { + "epoch": 7.6767860389020175, + "grad_norm": 0.11840160191059113, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0172, + "step": 84460 + }, + { + "epoch": 7.6776949645518995, + "grad_norm": 0.13741803169250488, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0175, + "step": 84470 + }, + { + "epoch": 7.6786038902017815, + "grad_norm": 0.18375667929649353, + "learning_rate": 6.630735722066602e-06, + "loss": 0.0183, + "step": 84480 + }, + { + "epoch": 7.6795128158516635, + "grad_norm": 0.1147468313574791, + "learning_rate": 6.627305611176788e-06, + "loss": 0.0179, + "step": 84490 + }, + { + "epoch": 7.680421741501545, + "grad_norm": 0.12655441462993622, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0186, + "step": 84500 + }, + { + "epoch": 7.6813306671514265, + "grad_norm": 0.11541386693716049, + "learning_rate": 6.620447297210253e-06, + "loss": 0.0171, + "step": 84510 + }, + { + "epoch": 7.6822395928013085, + "grad_norm": 0.10650799423456192, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0177, + "step": 84520 + }, + { + "epoch": 7.6831485184511905, + "grad_norm": 0.11874610185623169, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0191, + "step": 84530 + }, + { + "epoch": 7.6840574441010725, + "grad_norm": 0.1285473108291626, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0178, + "step": 84540 + }, + { + "epoch": 7.6849663697509545, + "grad_norm": 0.1518418788909912, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0175, + "step": 84550 + }, + { + "epoch": 7.6858752954008365, + "grad_norm": 0.16266778111457825, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0174, + "step": 84560 + }, + { + "epoch": 7.6867842210507185, + "grad_norm": 0.15311697125434875, + "learning_rate": 6.599887636561452e-06, + "loss": 0.0166, + "step": 84570 + }, + { + "epoch": 7.6876931467005996, + "grad_norm": 0.13930350542068481, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0164, + "step": 84580 + }, + { + "epoch": 7.6886020723504815, + "grad_norm": 0.1479211151599884, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0211, + "step": 84590 + }, + { + "epoch": 7.6895109980003635, + "grad_norm": 0.13343772292137146, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0165, + "step": 84600 + }, + { + "epoch": 7.6904199236502455, + "grad_norm": 0.13320523500442505, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0167, + "step": 84610 + }, + { + "epoch": 7.6913288493001275, + "grad_norm": 0.11799948662519455, + "learning_rate": 6.582772126793552e-06, + "loss": 0.0183, + "step": 84620 + }, + { + "epoch": 7.692237774950009, + "grad_norm": 0.16523052752017975, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0174, + "step": 84630 + }, + { + "epoch": 7.693146700599891, + "grad_norm": 0.13278405368328094, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0187, + "step": 84640 + }, + { + "epoch": 7.694055626249773, + "grad_norm": 0.12326890230178833, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0193, + "step": 84650 + }, + { + "epoch": 7.694964551899655, + "grad_norm": 0.1130458414554596, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0183, + "step": 84660 + }, + { + "epoch": 7.6958734775495365, + "grad_norm": 0.12464206665754318, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0179, + "step": 84670 + }, + { + "epoch": 7.6967824031994185, + "grad_norm": 0.11942902952432632, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0153, + "step": 84680 + }, + { + "epoch": 7.6976913288493005, + "grad_norm": 0.11969815939664841, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0178, + "step": 84690 + }, + { + "epoch": 7.698600254499182, + "grad_norm": 0.12430941313505173, + "learning_rate": 6.555420564777748e-06, + "loss": 0.0216, + "step": 84700 + }, + { + "epoch": 7.699509180149064, + "grad_norm": 0.11280257254838943, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0187, + "step": 84710 + }, + { + "epoch": 7.700418105798946, + "grad_norm": 0.1263544261455536, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0177, + "step": 84720 + }, + { + "epoch": 7.701327031448828, + "grad_norm": 0.1068965494632721, + "learning_rate": 6.545174303499989e-06, + "loss": 0.0194, + "step": 84730 + }, + { + "epoch": 7.70223595709871, + "grad_norm": 0.08305899798870087, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0155, + "step": 84740 + }, + { + "epoch": 7.7031448827485915, + "grad_norm": 0.11690539121627808, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0168, + "step": 84750 + }, + { + "epoch": 7.704053808398473, + "grad_norm": 0.156155526638031, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0185, + "step": 84760 + }, + { + "epoch": 7.704962734048355, + "grad_norm": 0.1269388198852539, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.017, + "step": 84770 + }, + { + "epoch": 7.705871659698237, + "grad_norm": 0.13501039147377014, + "learning_rate": 6.528110046590617e-06, + "loss": 0.0185, + "step": 84780 + }, + { + "epoch": 7.706780585348119, + "grad_norm": 0.14655452966690063, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.02, + "step": 84790 + }, + { + "epoch": 7.707689510998001, + "grad_norm": 0.09860209375619888, + "learning_rate": 6.52128884570186e-06, + "loss": 0.017, + "step": 84800 + }, + { + "epoch": 7.708598436647883, + "grad_norm": 0.09783412516117096, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0189, + "step": 84810 + }, + { + "epoch": 7.709507362297764, + "grad_norm": 0.11710499227046967, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0184, + "step": 84820 + }, + { + "epoch": 7.710416287947646, + "grad_norm": 0.12217194586992264, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0178, + "step": 84830 + }, + { + "epoch": 7.711325213597528, + "grad_norm": 0.14594826102256775, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0176, + "step": 84840 + }, + { + "epoch": 7.71223413924741, + "grad_norm": 0.12999464571475983, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0187, + "step": 84850 + }, + { + "epoch": 7.713143064897292, + "grad_norm": 0.09236236661672592, + "learning_rate": 6.500840704971546e-06, + "loss": 0.0166, + "step": 84860 + }, + { + "epoch": 7.714051990547174, + "grad_norm": 0.09880439192056656, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0171, + "step": 84870 + }, + { + "epoch": 7.714960916197055, + "grad_norm": 0.14031565189361572, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0153, + "step": 84880 + }, + { + "epoch": 7.715869841846937, + "grad_norm": 0.15138772130012512, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0165, + "step": 84890 + }, + { + "epoch": 7.716778767496819, + "grad_norm": 0.1402880698442459, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0165, + "step": 84900 + }, + { + "epoch": 7.717687693146701, + "grad_norm": 0.1446343958377838, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0184, + "step": 84910 + }, + { + "epoch": 7.718596618796583, + "grad_norm": 0.11218110471963882, + "learning_rate": 6.480415800618211e-06, + "loss": 0.0171, + "step": 84920 + }, + { + "epoch": 7.719505544446465, + "grad_norm": 0.12314970791339874, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0171, + "step": 84930 + }, + { + "epoch": 7.720414470096346, + "grad_norm": 0.10896799713373184, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0167, + "step": 84940 + }, + { + "epoch": 7.721323395746228, + "grad_norm": 0.17147572338581085, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0174, + "step": 84950 + }, + { + "epoch": 7.72223232139611, + "grad_norm": 0.1206645742058754, + "learning_rate": 6.466812134392675e-06, + "loss": 0.0177, + "step": 84960 + }, + { + "epoch": 7.723141247045992, + "grad_norm": 0.16372065246105194, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0193, + "step": 84970 + }, + { + "epoch": 7.724050172695874, + "grad_norm": 0.14224299788475037, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0167, + "step": 84980 + }, + { + "epoch": 7.724959098345756, + "grad_norm": 0.12015493214130402, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0166, + "step": 84990 + }, + { + "epoch": 7.725868023995637, + "grad_norm": 0.1280585676431656, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0169, + "step": 85000 + }, + { + "epoch": 7.726776949645519, + "grad_norm": 0.12502743303775787, + "learning_rate": 6.449822134429607e-06, + "loss": 0.0175, + "step": 85010 + }, + { + "epoch": 7.727685875295401, + "grad_norm": 0.15167103707790375, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0173, + "step": 85020 + }, + { + "epoch": 7.728594800945283, + "grad_norm": 0.13290344178676605, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0208, + "step": 85030 + }, + { + "epoch": 7.729503726595165, + "grad_norm": 0.12918488681316376, + "learning_rate": 6.439635924342405e-06, + "loss": 0.019, + "step": 85040 + }, + { + "epoch": 7.730412652245047, + "grad_norm": 0.10920868068933487, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0187, + "step": 85050 + }, + { + "epoch": 7.731321577894928, + "grad_norm": 0.1112840548157692, + "learning_rate": 6.432848367703634e-06, + "loss": 0.0167, + "step": 85060 + }, + { + "epoch": 7.73223050354481, + "grad_norm": 0.17763465642929077, + "learning_rate": 6.429455565183332e-06, + "loss": 0.0193, + "step": 85070 + }, + { + "epoch": 7.733139429194692, + "grad_norm": 0.14096485078334808, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0181, + "step": 85080 + }, + { + "epoch": 7.734048354844574, + "grad_norm": 0.14268991351127625, + "learning_rate": 6.422671913029213e-06, + "loss": 0.0184, + "step": 85090 + }, + { + "epoch": 7.734957280494456, + "grad_norm": 0.11841094493865967, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0162, + "step": 85100 + }, + { + "epoch": 7.735866206144338, + "grad_norm": 0.1625455915927887, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0159, + "step": 85110 + }, + { + "epoch": 7.73677513179422, + "grad_norm": 0.17363907396793365, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0158, + "step": 85120 + }, + { + "epoch": 7.737684057444101, + "grad_norm": 0.1244877278804779, + "learning_rate": 6.409112427478281e-06, + "loss": 0.0169, + "step": 85130 + }, + { + "epoch": 7.738592983093983, + "grad_norm": 0.14741729199886322, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.0162, + "step": 85140 + }, + { + "epoch": 7.739501908743865, + "grad_norm": 0.0973203256726265, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0168, + "step": 85150 + }, + { + "epoch": 7.740410834393747, + "grad_norm": 0.12725237011909485, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0177, + "step": 85160 + }, + { + "epoch": 7.741319760043629, + "grad_norm": 0.10726003348827362, + "learning_rate": 6.395563380665818e-06, + "loss": 0.0155, + "step": 85170 + }, + { + "epoch": 7.74222868569351, + "grad_norm": 0.16676528751850128, + "learning_rate": 6.392177751945165e-06, + "loss": 0.0155, + "step": 85180 + }, + { + "epoch": 7.743137611343392, + "grad_norm": 0.09221618622541428, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0182, + "step": 85190 + }, + { + "epoch": 7.744046536993274, + "grad_norm": 0.27530068159103394, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0184, + "step": 85200 + }, + { + "epoch": 7.744955462643156, + "grad_norm": 0.17428860068321228, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0169, + "step": 85210 + }, + { + "epoch": 7.745864388293038, + "grad_norm": 0.15791970491409302, + "learning_rate": 6.378641776705912e-06, + "loss": 0.0164, + "step": 85220 + }, + { + "epoch": 7.74677331394292, + "grad_norm": 0.21972310543060303, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0165, + "step": 85230 + }, + { + "epoch": 7.747682239592802, + "grad_norm": 0.11542212218046188, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0165, + "step": 85240 + }, + { + "epoch": 7.748591165242683, + "grad_norm": 0.1755380630493164, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0184, + "step": 85250 + }, + { + "epoch": 7.749500090892565, + "grad_norm": 0.13007083535194397, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0172, + "step": 85260 + }, + { + "epoch": 7.750409016542447, + "grad_norm": 0.12990841269493103, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0194, + "step": 85270 + }, + { + "epoch": 7.751317942192329, + "grad_norm": 0.11556356400251389, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0164, + "step": 85280 + }, + { + "epoch": 7.752226867842211, + "grad_norm": 0.12304751574993134, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0186, + "step": 85290 + }, + { + "epoch": 7.753135793492092, + "grad_norm": 0.15282540023326874, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0194, + "step": 85300 + }, + { + "epoch": 7.754044719141974, + "grad_norm": 0.18787141144275665, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0202, + "step": 85310 + }, + { + "epoch": 7.754953644791856, + "grad_norm": 0.21037839353084564, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0175, + "step": 85320 + }, + { + "epoch": 7.755862570441738, + "grad_norm": 0.11931221932172775, + "learning_rate": 6.341471909877353e-06, + "loss": 0.018, + "step": 85330 + }, + { + "epoch": 7.75677149609162, + "grad_norm": 0.14128361642360687, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0179, + "step": 85340 + }, + { + "epoch": 7.757680421741502, + "grad_norm": 0.14190416038036346, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0168, + "step": 85350 + }, + { + "epoch": 7.758589347391384, + "grad_norm": 0.13293610513210297, + "learning_rate": 6.331348467825299e-06, + "loss": 0.0166, + "step": 85360 + }, + { + "epoch": 7.759498273041265, + "grad_norm": 0.1361767202615738, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0194, + "step": 85370 + }, + { + "epoch": 7.760407198691147, + "grad_norm": 0.14220529794692993, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0167, + "step": 85380 + }, + { + "epoch": 7.761316124341029, + "grad_norm": 0.13142889738082886, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0176, + "step": 85390 + }, + { + "epoch": 7.762225049990911, + "grad_norm": 0.13947774469852448, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0162, + "step": 85400 + }, + { + "epoch": 7.763133975640793, + "grad_norm": 0.22825872898101807, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0176, + "step": 85410 + }, + { + "epoch": 7.764042901290674, + "grad_norm": 0.13643942773342133, + "learning_rate": 6.311119365461163e-06, + "loss": 0.0165, + "step": 85420 + }, + { + "epoch": 7.764951826940556, + "grad_norm": 0.21035940945148468, + "learning_rate": 6.307750156428322e-06, + "loss": 0.0182, + "step": 85430 + }, + { + "epoch": 7.765860752590438, + "grad_norm": 0.13723362982273102, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0177, + "step": 85440 + }, + { + "epoch": 7.76676967824032, + "grad_norm": 0.10036035627126694, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0194, + "step": 85450 + }, + { + "epoch": 7.767678603890202, + "grad_norm": 0.15874667465686798, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0207, + "step": 85460 + }, + { + "epoch": 7.768587529540084, + "grad_norm": 0.10848287492990494, + "learning_rate": 6.294279924045003e-06, + "loss": 0.0184, + "step": 85470 + }, + { + "epoch": 7.769496455189966, + "grad_norm": 0.17067326605319977, + "learning_rate": 6.290914018164905e-06, + "loss": 0.0183, + "step": 85480 + }, + { + "epoch": 7.770405380839847, + "grad_norm": 0.12085124105215073, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0175, + "step": 85490 + }, + { + "epoch": 7.771314306489729, + "grad_norm": 0.11380152404308319, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.0169, + "step": 85500 + }, + { + "epoch": 7.772223232139611, + "grad_norm": 0.1358112096786499, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0161, + "step": 85510 + }, + { + "epoch": 7.773132157789493, + "grad_norm": 0.21045054495334625, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0172, + "step": 85520 + }, + { + "epoch": 7.774041083439375, + "grad_norm": 0.080381378531456, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0169, + "step": 85530 + }, + { + "epoch": 7.774950009089256, + "grad_norm": 0.21625541150569916, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0184, + "step": 85540 + }, + { + "epoch": 7.775858934739138, + "grad_norm": 0.1376381665468216, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0169, + "step": 85550 + }, + { + "epoch": 7.77676786038902, + "grad_norm": 0.15394926071166992, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0163, + "step": 85560 + }, + { + "epoch": 7.777676786038902, + "grad_norm": 0.1707555651664734, + "learning_rate": 6.260650658793892e-06, + "loss": 0.0179, + "step": 85570 + }, + { + "epoch": 7.778585711688784, + "grad_norm": 0.12340693175792694, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.0196, + "step": 85580 + }, + { + "epoch": 7.779494637338666, + "grad_norm": 0.10914401710033417, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0181, + "step": 85590 + }, + { + "epoch": 7.780403562988548, + "grad_norm": 0.1341734528541565, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0158, + "step": 85600 + }, + { + "epoch": 7.781312488638429, + "grad_norm": 0.11829273402690887, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0195, + "step": 85610 + }, + { + "epoch": 7.782221414288311, + "grad_norm": 0.11657477915287018, + "learning_rate": 6.243860898806956e-06, + "loss": 0.0187, + "step": 85620 + }, + { + "epoch": 7.783130339938193, + "grad_norm": 0.1320897787809372, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0185, + "step": 85630 + }, + { + "epoch": 7.784039265588075, + "grad_norm": 0.12660036981105804, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0168, + "step": 85640 + }, + { + "epoch": 7.784948191237957, + "grad_norm": 0.1351269781589508, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.017, + "step": 85650 + }, + { + "epoch": 7.785857116887838, + "grad_norm": 0.10397011786699295, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0192, + "step": 85660 + }, + { + "epoch": 7.78676604253772, + "grad_norm": 0.14594240486621857, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0168, + "step": 85670 + }, + { + "epoch": 7.787674968187602, + "grad_norm": 0.15589913725852966, + "learning_rate": 6.223735133664553e-06, + "loss": 0.0181, + "step": 85680 + }, + { + "epoch": 7.788583893837484, + "grad_norm": 0.13947832584381104, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0144, + "step": 85690 + }, + { + "epoch": 7.789492819487366, + "grad_norm": 0.16834691166877747, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0182, + "step": 85700 + }, + { + "epoch": 7.790401745137248, + "grad_norm": 0.21554683148860931, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0166, + "step": 85710 + }, + { + "epoch": 7.79131067078713, + "grad_norm": 0.12760065495967865, + "learning_rate": 6.210331283494173e-06, + "loss": 0.0159, + "step": 85720 + }, + { + "epoch": 7.792219596437011, + "grad_norm": 0.0885249525308609, + "learning_rate": 6.206981989108829e-06, + "loss": 0.016, + "step": 85730 + }, + { + "epoch": 7.793128522086893, + "grad_norm": 0.1166095957159996, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0176, + "step": 85740 + }, + { + "epoch": 7.794037447736775, + "grad_norm": 0.17501933872699738, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.0161, + "step": 85750 + }, + { + "epoch": 7.794946373386657, + "grad_norm": 0.13488911092281342, + "learning_rate": 6.196938113600098e-06, + "loss": 0.0198, + "step": 85760 + }, + { + "epoch": 7.795855299036539, + "grad_norm": 0.15505006909370422, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0163, + "step": 85770 + }, + { + "epoch": 7.79676422468642, + "grad_norm": 0.08874861150979996, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0162, + "step": 85780 + }, + { + "epoch": 7.797673150336302, + "grad_norm": 0.19939792156219482, + "learning_rate": 6.186900254901158e-06, + "loss": 0.0194, + "step": 85790 + }, + { + "epoch": 7.798582075986184, + "grad_norm": 0.14665263891220093, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0185, + "step": 85800 + }, + { + "epoch": 7.799491001636066, + "grad_norm": 0.13801270723342896, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0182, + "step": 85810 + }, + { + "epoch": 7.800399927285948, + "grad_norm": 0.15136507153511047, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0205, + "step": 85820 + }, + { + "epoch": 7.80130885293583, + "grad_norm": 0.11929619312286377, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0195, + "step": 85830 + }, + { + "epoch": 7.802217778585712, + "grad_norm": 0.15760090947151184, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.0179, + "step": 85840 + }, + { + "epoch": 7.803126704235593, + "grad_norm": 0.13094176352024078, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0169, + "step": 85850 + }, + { + "epoch": 7.804035629885475, + "grad_norm": 0.10088188946247101, + "learning_rate": 6.163502021839628e-06, + "loss": 0.0152, + "step": 85860 + }, + { + "epoch": 7.804944555535357, + "grad_norm": 0.1056891530752182, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0173, + "step": 85870 + }, + { + "epoch": 7.805853481185239, + "grad_norm": 0.08508507907390594, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0149, + "step": 85880 + }, + { + "epoch": 7.806762406835121, + "grad_norm": 0.21249297261238098, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0172, + "step": 85890 + }, + { + "epoch": 7.807671332485003, + "grad_norm": 0.09230610728263855, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0168, + "step": 85900 + }, + { + "epoch": 7.808580258134884, + "grad_norm": 0.11827390640974045, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0159, + "step": 85910 + }, + { + "epoch": 7.809489183784766, + "grad_norm": 0.13512219488620758, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0179, + "step": 85920 + }, + { + "epoch": 7.810398109434648, + "grad_norm": 0.11854810267686844, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0172, + "step": 85930 + }, + { + "epoch": 7.81130703508453, + "grad_norm": 0.19189304113388062, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0181, + "step": 85940 + }, + { + "epoch": 7.812215960734412, + "grad_norm": 0.11307535320520401, + "learning_rate": 6.133466908794758e-06, + "loss": 0.0163, + "step": 85950 + }, + { + "epoch": 7.813124886384294, + "grad_norm": 0.08864537626504898, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0165, + "step": 85960 + }, + { + "epoch": 7.814033812034175, + "grad_norm": 0.14873762428760529, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0179, + "step": 85970 + }, + { + "epoch": 7.814942737684057, + "grad_norm": 0.2083912193775177, + "learning_rate": 6.123467315710249e-06, + "loss": 0.016, + "step": 85980 + }, + { + "epoch": 7.815851663333939, + "grad_norm": 0.13386699557304382, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.0174, + "step": 85990 + }, + { + "epoch": 7.816760588983821, + "grad_norm": 0.14684490859508514, + "learning_rate": 6.116804290469075e-06, + "loss": 0.019, + "step": 86000 + }, + { + "epoch": 7.817669514633703, + "grad_norm": 0.15239454805850983, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0171, + "step": 86010 + }, + { + "epoch": 7.818578440283585, + "grad_norm": 0.17329421639442444, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0161, + "step": 86020 + }, + { + "epoch": 7.819487365933466, + "grad_norm": 0.12311708182096481, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0166, + "step": 86030 + }, + { + "epoch": 7.820396291583348, + "grad_norm": 0.11502500623464584, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0177, + "step": 86040 + }, + { + "epoch": 7.82130521723323, + "grad_norm": 0.20937737822532654, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0166, + "step": 86050 + }, + { + "epoch": 7.822214142883112, + "grad_norm": 0.1676219254732132, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0173, + "step": 86060 + }, + { + "epoch": 7.823123068532994, + "grad_norm": 0.1174788847565651, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0167, + "step": 86070 + }, + { + "epoch": 7.824031994182876, + "grad_norm": 0.10254926979541779, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0178, + "step": 86080 + }, + { + "epoch": 7.824940919832757, + "grad_norm": 0.13054049015045166, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0196, + "step": 86090 + }, + { + "epoch": 7.825849845482639, + "grad_norm": 0.12035272270441055, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0177, + "step": 86100 + }, + { + "epoch": 7.826758771132521, + "grad_norm": 0.17599746584892273, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0185, + "step": 86110 + }, + { + "epoch": 7.827667696782403, + "grad_norm": 0.19532109797000885, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0178, + "step": 86120 + }, + { + "epoch": 7.828576622432285, + "grad_norm": 0.1331012099981308, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0169, + "step": 86130 + }, + { + "epoch": 7.829485548082167, + "grad_norm": 0.15607163310050964, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0157, + "step": 86140 + }, + { + "epoch": 7.830394473732048, + "grad_norm": 0.16254977881908417, + "learning_rate": 6.066917754871587e-06, + "loss": 0.018, + "step": 86150 + }, + { + "epoch": 7.83130339938193, + "grad_norm": 0.16932183504104614, + "learning_rate": 6.063597403678799e-06, + "loss": 0.018, + "step": 86160 + }, + { + "epoch": 7.832212325031812, + "grad_norm": 0.14028474688529968, + "learning_rate": 6.060277731146825e-06, + "loss": 0.0198, + "step": 86170 + }, + { + "epoch": 7.833121250681694, + "grad_norm": 0.14520959556102753, + "learning_rate": 6.056958737527759e-06, + "loss": 0.0177, + "step": 86180 + }, + { + "epoch": 7.834030176331576, + "grad_norm": 0.12663264572620392, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0161, + "step": 86190 + }, + { + "epoch": 7.834939101981458, + "grad_norm": 0.1256934106349945, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0157, + "step": 86200 + }, + { + "epoch": 7.835848027631339, + "grad_norm": 0.1237650141119957, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0193, + "step": 86210 + }, + { + "epoch": 7.836756953281221, + "grad_norm": 0.14821398258209229, + "learning_rate": 6.043689557220918e-06, + "loss": 0.016, + "step": 86220 + }, + { + "epoch": 7.837665878931103, + "grad_norm": 0.12680159509181976, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0188, + "step": 86230 + }, + { + "epoch": 7.838574804580985, + "grad_norm": 0.10803283751010895, + "learning_rate": 6.037059047095944e-06, + "loss": 0.0186, + "step": 86240 + }, + { + "epoch": 7.839483730230867, + "grad_norm": 0.17082823812961578, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0168, + "step": 86250 + }, + { + "epoch": 7.840392655880749, + "grad_norm": 0.1629055291414261, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0193, + "step": 86260 + }, + { + "epoch": 7.841301581530631, + "grad_norm": 0.10556209087371826, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0175, + "step": 86270 + }, + { + "epoch": 7.842210507180512, + "grad_norm": 0.09921140968799591, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0186, + "step": 86280 + }, + { + "epoch": 7.843119432830394, + "grad_norm": 0.10677261650562286, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0166, + "step": 86290 + }, + { + "epoch": 7.844028358480276, + "grad_norm": 0.15362854301929474, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0169, + "step": 86300 + }, + { + "epoch": 7.844937284130158, + "grad_norm": 0.1385800689458847, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0175, + "step": 86310 + }, + { + "epoch": 7.84584620978004, + "grad_norm": 0.14261727035045624, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0169, + "step": 86320 + }, + { + "epoch": 7.846755135429921, + "grad_norm": 0.11230266839265823, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0189, + "step": 86330 + }, + { + "epoch": 7.847664061079803, + "grad_norm": 0.11406083405017853, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0163, + "step": 86340 + }, + { + "epoch": 7.848572986729685, + "grad_norm": 0.1824183613061905, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0182, + "step": 86350 + }, + { + "epoch": 7.849481912379567, + "grad_norm": 0.14147046208381653, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0185, + "step": 86360 + }, + { + "epoch": 7.850390838029449, + "grad_norm": 0.15624786913394928, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0163, + "step": 86370 + }, + { + "epoch": 7.851299763679331, + "grad_norm": 0.1213158443570137, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0166, + "step": 86380 + }, + { + "epoch": 7.852208689329213, + "grad_norm": 0.13649308681488037, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0163, + "step": 86390 + }, + { + "epoch": 7.853117614979094, + "grad_norm": 0.11076326668262482, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0164, + "step": 86400 + }, + { + "epoch": 7.854026540628976, + "grad_norm": 0.17869864404201508, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0184, + "step": 86410 + }, + { + "epoch": 7.854935466278858, + "grad_norm": 0.18529050052165985, + "learning_rate": 5.977507219115265e-06, + "loss": 0.0167, + "step": 86420 + }, + { + "epoch": 7.85584439192874, + "grad_norm": 0.11921916157007217, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.0167, + "step": 86430 + }, + { + "epoch": 7.856753317578622, + "grad_norm": 0.11886696517467499, + "learning_rate": 5.970904026539538e-06, + "loss": 0.0173, + "step": 86440 + }, + { + "epoch": 7.8576622432285035, + "grad_norm": 0.11417366564273834, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0168, + "step": 86450 + }, + { + "epoch": 7.8585711688783855, + "grad_norm": 0.11901182681322098, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0164, + "step": 86460 + }, + { + "epoch": 7.8594800945282675, + "grad_norm": 0.13270032405853271, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0211, + "step": 86470 + }, + { + "epoch": 7.860389020178149, + "grad_norm": 0.14052152633666992, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0205, + "step": 86480 + }, + { + "epoch": 7.861297945828031, + "grad_norm": 0.11091160029172897, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0176, + "step": 86490 + }, + { + "epoch": 7.862206871477913, + "grad_norm": 0.2138950228691101, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0165, + "step": 86500 + }, + { + "epoch": 7.863115797127795, + "grad_norm": 0.1299852877855301, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0184, + "step": 86510 + }, + { + "epoch": 7.8640247227776765, + "grad_norm": 0.1778668612241745, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.0166, + "step": 86520 + }, + { + "epoch": 7.8649336484275585, + "grad_norm": 0.1704927682876587, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0187, + "step": 86530 + }, + { + "epoch": 7.8658425740774405, + "grad_norm": 0.10643947124481201, + "learning_rate": 5.937929245763599e-06, + "loss": 0.0176, + "step": 86540 + }, + { + "epoch": 7.8667514997273225, + "grad_norm": 0.15470664203166962, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0203, + "step": 86550 + }, + { + "epoch": 7.8676604253772044, + "grad_norm": 0.09271415323019028, + "learning_rate": 5.931342540052567e-06, + "loss": 0.0163, + "step": 86560 + }, + { + "epoch": 7.8685693510270855, + "grad_norm": 0.49640241265296936, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0183, + "step": 86570 + }, + { + "epoch": 7.8694782766769675, + "grad_norm": 0.1627887785434723, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0197, + "step": 86580 + }, + { + "epoch": 7.8703872023268495, + "grad_norm": 0.13050848245620728, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0173, + "step": 86590 + }, + { + "epoch": 7.8712961279767315, + "grad_norm": 0.10283033549785614, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0177, + "step": 86600 + }, + { + "epoch": 7.8722050536266135, + "grad_norm": 0.1283063441514969, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0175, + "step": 86610 + }, + { + "epoch": 7.8731139792764955, + "grad_norm": 0.12254087626934052, + "learning_rate": 5.91159895983091e-06, + "loss": 0.017, + "step": 86620 + }, + { + "epoch": 7.8740229049263775, + "grad_norm": 0.08373022824525833, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0158, + "step": 86630 + }, + { + "epoch": 7.874931830576259, + "grad_norm": 0.10995616018772125, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0184, + "step": 86640 + }, + { + "epoch": 7.8758407562261405, + "grad_norm": 0.16771914064884186, + "learning_rate": 5.901736484102119e-06, + "loss": 0.0181, + "step": 86650 + }, + { + "epoch": 7.8767496818760225, + "grad_norm": 0.21201634407043457, + "learning_rate": 5.898450373765829e-06, + "loss": 0.0167, + "step": 86660 + }, + { + "epoch": 7.8776586075259045, + "grad_norm": 0.11044290661811829, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0158, + "step": 86670 + }, + { + "epoch": 7.8785675331757865, + "grad_norm": 0.26563188433647156, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0162, + "step": 86680 + }, + { + "epoch": 7.879476458825668, + "grad_norm": 0.16880540549755096, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0166, + "step": 86690 + }, + { + "epoch": 7.88038538447555, + "grad_norm": 0.1310718059539795, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0191, + "step": 86700 + }, + { + "epoch": 7.881294310125432, + "grad_norm": 0.12287979573011398, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0166, + "step": 86710 + }, + { + "epoch": 7.882203235775314, + "grad_norm": 0.1954355090856552, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0169, + "step": 86720 + }, + { + "epoch": 7.8831121614251956, + "grad_norm": 0.10314248502254486, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0181, + "step": 86730 + }, + { + "epoch": 7.8840210870750775, + "grad_norm": 0.1212855875492096, + "learning_rate": 5.872186395322063e-06, + "loss": 0.0163, + "step": 86740 + }, + { + "epoch": 7.8849300127249595, + "grad_norm": 0.13279138505458832, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0165, + "step": 86750 + }, + { + "epoch": 7.885838938374841, + "grad_norm": 0.19175146520137787, + "learning_rate": 5.865627327698698e-06, + "loss": 0.017, + "step": 86760 + }, + { + "epoch": 7.886747864024723, + "grad_norm": 0.12022826820611954, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0183, + "step": 86770 + }, + { + "epoch": 7.887656789674605, + "grad_norm": 0.14838610589504242, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0186, + "step": 86780 + }, + { + "epoch": 7.888565715324487, + "grad_norm": 0.1460009217262268, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0175, + "step": 86790 + }, + { + "epoch": 7.889474640974369, + "grad_norm": 0.10871396958827972, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0171, + "step": 86800 + }, + { + "epoch": 7.89038356662425, + "grad_norm": 0.14225958287715912, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0176, + "step": 86810 + }, + { + "epoch": 7.891292492274132, + "grad_norm": 0.19180047512054443, + "learning_rate": 5.845966781481725e-06, + "loss": 0.0174, + "step": 86820 + }, + { + "epoch": 7.892201417924014, + "grad_norm": 0.11609794199466705, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0173, + "step": 86830 + }, + { + "epoch": 7.893110343573896, + "grad_norm": 0.09436925500631332, + "learning_rate": 5.839418824929713e-06, + "loss": 0.0172, + "step": 86840 + }, + { + "epoch": 7.894019269223778, + "grad_norm": 0.13791809976100922, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0172, + "step": 86850 + }, + { + "epoch": 7.89492819487366, + "grad_norm": 0.13415205478668213, + "learning_rate": 5.832873651120394e-06, + "loss": 0.017, + "step": 86860 + }, + { + "epoch": 7.895837120523542, + "grad_norm": 0.19303971529006958, + "learning_rate": 5.829602108365602e-06, + "loss": 0.019, + "step": 86870 + }, + { + "epoch": 7.896746046173423, + "grad_norm": 0.10071898251771927, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0157, + "step": 86880 + }, + { + "epoch": 7.897654971823305, + "grad_norm": 0.11375239491462708, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0169, + "step": 86890 + }, + { + "epoch": 7.898563897473187, + "grad_norm": 0.13710755109786987, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0172, + "step": 86900 + }, + { + "epoch": 7.899472823123069, + "grad_norm": 0.12831903994083405, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0188, + "step": 86910 + }, + { + "epoch": 7.900381748772951, + "grad_norm": 0.21256671845912933, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.0173, + "step": 86920 + }, + { + "epoch": 7.901290674422832, + "grad_norm": 0.16200537979602814, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0161, + "step": 86930 + }, + { + "epoch": 7.902199600072714, + "grad_norm": 0.08667983114719391, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0188, + "step": 86940 + }, + { + "epoch": 7.903108525722596, + "grad_norm": 0.14155158400535583, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0168, + "step": 86950 + }, + { + "epoch": 7.904017451372478, + "grad_norm": 0.13036614656448364, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0178, + "step": 86960 + }, + { + "epoch": 7.90492637702236, + "grad_norm": 0.17273752391338348, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0178, + "step": 86970 + }, + { + "epoch": 7.905835302672242, + "grad_norm": 0.10066892951726913, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0174, + "step": 86980 + }, + { + "epoch": 7.906744228322124, + "grad_norm": 0.18580415844917297, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0183, + "step": 86990 + }, + { + "epoch": 7.907653153972005, + "grad_norm": 0.1242523267865181, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0176, + "step": 87000 + }, + { + "epoch": 7.908562079621887, + "grad_norm": 0.13548694550991058, + "learning_rate": 5.783873748023966e-06, + "loss": 0.019, + "step": 87010 + }, + { + "epoch": 7.909471005271769, + "grad_norm": 0.11392626911401749, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0162, + "step": 87020 + }, + { + "epoch": 7.910379930921651, + "grad_norm": 0.12088858336210251, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0164, + "step": 87030 + }, + { + "epoch": 7.911288856571533, + "grad_norm": 0.1471061408519745, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0188, + "step": 87040 + }, + { + "epoch": 7.912197782221414, + "grad_norm": 0.10701092332601547, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0158, + "step": 87050 + }, + { + "epoch": 7.913106707871296, + "grad_norm": 0.11496774852275848, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0178, + "step": 87060 + }, + { + "epoch": 7.914015633521178, + "grad_norm": 0.21875597536563873, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0169, + "step": 87070 + }, + { + "epoch": 7.91492455917106, + "grad_norm": 0.1322181522846222, + "learning_rate": 5.761060968269583e-06, + "loss": 0.019, + "step": 87080 + }, + { + "epoch": 7.915833484820942, + "grad_norm": 0.12126797437667847, + "learning_rate": 5.757804804299801e-06, + "loss": 0.016, + "step": 87090 + }, + { + "epoch": 7.916742410470824, + "grad_norm": 0.11832000315189362, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0191, + "step": 87100 + }, + { + "epoch": 7.917651336120706, + "grad_norm": 0.17458027601242065, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0184, + "step": 87110 + }, + { + "epoch": 7.918560261770587, + "grad_norm": 0.13406716287136078, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0184, + "step": 87120 + }, + { + "epoch": 7.919469187420469, + "grad_norm": 0.13942115008831024, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0152, + "step": 87130 + }, + { + "epoch": 7.920378113070351, + "grad_norm": 0.10232435911893845, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0148, + "step": 87140 + }, + { + "epoch": 7.921287038720233, + "grad_norm": 0.1461653858423233, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0183, + "step": 87150 + }, + { + "epoch": 7.922195964370115, + "grad_norm": 0.15649966895580292, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0159, + "step": 87160 + }, + { + "epoch": 7.923104890019997, + "grad_norm": 0.10013316571712494, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0165, + "step": 87170 + }, + { + "epoch": 7.924013815669878, + "grad_norm": 0.1581231951713562, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0179, + "step": 87180 + }, + { + "epoch": 7.92492274131976, + "grad_norm": 0.1434575319290161, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0187, + "step": 87190 + }, + { + "epoch": 7.925831666969642, + "grad_norm": 0.13869459927082062, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0172, + "step": 87200 + }, + { + "epoch": 7.926740592619524, + "grad_norm": 0.11970139294862747, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0175, + "step": 87210 + }, + { + "epoch": 7.927649518269406, + "grad_norm": 0.18489240109920502, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0158, + "step": 87220 + }, + { + "epoch": 7.928558443919288, + "grad_norm": 0.13821768760681152, + "learning_rate": 5.712292318931916e-06, + "loss": 0.017, + "step": 87230 + }, + { + "epoch": 7.929467369569169, + "grad_norm": 0.15655328333377838, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0187, + "step": 87240 + }, + { + "epoch": 7.930376295219051, + "grad_norm": 0.20709487795829773, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0177, + "step": 87250 + }, + { + "epoch": 7.931285220868933, + "grad_norm": 0.16638681292533875, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0183, + "step": 87260 + }, + { + "epoch": 7.932194146518815, + "grad_norm": 0.17324088513851166, + "learning_rate": 5.699314114321709e-06, + "loss": 0.0185, + "step": 87270 + }, + { + "epoch": 7.933103072168697, + "grad_norm": 0.18146531283855438, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0192, + "step": 87280 + }, + { + "epoch": 7.934011997818579, + "grad_norm": 0.12436339259147644, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0186, + "step": 87290 + }, + { + "epoch": 7.93492092346846, + "grad_norm": 0.1119014322757721, + "learning_rate": 5.689587876421451e-06, + "loss": 0.016, + "step": 87300 + }, + { + "epoch": 7.935829849118342, + "grad_norm": 0.1389857828617096, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0201, + "step": 87310 + }, + { + "epoch": 7.936738774768224, + "grad_norm": 0.11437205970287323, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0172, + "step": 87320 + }, + { + "epoch": 7.937647700418106, + "grad_norm": 0.12493535131216049, + "learning_rate": 5.679868002100092e-06, + "loss": 0.02, + "step": 87330 + }, + { + "epoch": 7.938556626067988, + "grad_norm": 0.15534503757953644, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0162, + "step": 87340 + }, + { + "epoch": 7.93946555171787, + "grad_norm": 0.11864790320396423, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0187, + "step": 87350 + }, + { + "epoch": 7.940374477367751, + "grad_norm": 0.1946849524974823, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0184, + "step": 87360 + }, + { + "epoch": 7.941283403017633, + "grad_norm": 0.12659162282943726, + "learning_rate": 5.666918080053974e-06, + "loss": 0.017, + "step": 87370 + }, + { + "epoch": 7.942192328667515, + "grad_norm": 0.09818227589130402, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0151, + "step": 87380 + }, + { + "epoch": 7.943101254317397, + "grad_norm": 0.09599742293357849, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0174, + "step": 87390 + }, + { + "epoch": 7.944010179967279, + "grad_norm": 0.1557120531797409, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0181, + "step": 87400 + }, + { + "epoch": 7.944919105617161, + "grad_norm": 0.17986387014389038, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0186, + "step": 87410 + }, + { + "epoch": 7.945828031267043, + "grad_norm": 0.12115055322647095, + "learning_rate": 5.650746627022022e-06, + "loss": 0.016, + "step": 87420 + }, + { + "epoch": 7.946736956916924, + "grad_norm": 0.14316539466381073, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0185, + "step": 87430 + }, + { + "epoch": 7.947645882566806, + "grad_norm": 0.14315567910671234, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0177, + "step": 87440 + }, + { + "epoch": 7.948554808216688, + "grad_norm": 0.11281447112560272, + "learning_rate": 5.641052273407098e-06, + "loss": 0.018, + "step": 87450 + }, + { + "epoch": 7.94946373386657, + "grad_norm": 0.16859467327594757, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0224, + "step": 87460 + }, + { + "epoch": 7.950372659516452, + "grad_norm": 0.10088154673576355, + "learning_rate": 5.634592924339714e-06, + "loss": 0.0157, + "step": 87470 + }, + { + "epoch": 7.951281585166333, + "grad_norm": 0.1322157233953476, + "learning_rate": 5.631364316544696e-06, + "loss": 0.0182, + "step": 87480 + }, + { + "epoch": 7.952190510816215, + "grad_norm": 0.145071879029274, + "learning_rate": 5.628136420235731e-06, + "loss": 0.0161, + "step": 87490 + }, + { + "epoch": 7.953099436466097, + "grad_norm": 0.10583119094371796, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0177, + "step": 87500 + }, + { + "epoch": 7.954008362115979, + "grad_norm": 0.11223862320184708, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0171, + "step": 87510 + }, + { + "epoch": 7.954917287765861, + "grad_norm": 0.14515049755573273, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0198, + "step": 87520 + }, + { + "epoch": 7.955826213415743, + "grad_norm": 0.11315302550792694, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0181, + "step": 87530 + }, + { + "epoch": 7.956735139065625, + "grad_norm": 0.13392256200313568, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0169, + "step": 87540 + }, + { + "epoch": 7.957644064715506, + "grad_norm": 0.12547647953033447, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0169, + "step": 87550 + }, + { + "epoch": 7.958552990365388, + "grad_norm": 0.12392263859510422, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0165, + "step": 87560 + }, + { + "epoch": 7.95946191601527, + "grad_norm": 0.1275874525308609, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0165, + "step": 87570 + }, + { + "epoch": 7.960370841665152, + "grad_norm": 0.1764792501926422, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0168, + "step": 87580 + }, + { + "epoch": 7.961279767315034, + "grad_norm": 0.12516367435455322, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0174, + "step": 87590 + }, + { + "epoch": 7.962188692964915, + "grad_norm": 0.15870048105716705, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0174, + "step": 87600 + }, + { + "epoch": 7.963097618614797, + "grad_norm": 0.1074099913239479, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0156, + "step": 87610 + }, + { + "epoch": 7.964006544264679, + "grad_norm": 0.1612565666437149, + "learning_rate": 5.58623862491361e-06, + "loss": 0.019, + "step": 87620 + }, + { + "epoch": 7.964915469914561, + "grad_norm": 0.1251436024904251, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0184, + "step": 87630 + }, + { + "epoch": 7.965824395564443, + "grad_norm": 0.13714945316314697, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0175, + "step": 87640 + }, + { + "epoch": 7.966733321214325, + "grad_norm": 0.17029768228530884, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0178, + "step": 87650 + }, + { + "epoch": 7.967642246864207, + "grad_norm": 0.1321033239364624, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0202, + "step": 87660 + }, + { + "epoch": 7.968551172514088, + "grad_norm": 0.1526738554239273, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.017, + "step": 87670 + }, + { + "epoch": 7.96946009816397, + "grad_norm": 0.13799948990345, + "learning_rate": 5.566941898424208e-06, + "loss": 0.0153, + "step": 87680 + }, + { + "epoch": 7.970369023813852, + "grad_norm": 0.12433820217847824, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0165, + "step": 87690 + }, + { + "epoch": 7.971277949463734, + "grad_norm": 0.1135748028755188, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0181, + "step": 87700 + }, + { + "epoch": 7.972186875113616, + "grad_norm": 0.12294171005487442, + "learning_rate": 5.557303202992719e-06, + "loss": 0.0163, + "step": 87710 + }, + { + "epoch": 7.973095800763497, + "grad_norm": 0.12157972902059555, + "learning_rate": 5.554091738411292e-06, + "loss": 0.0166, + "step": 87720 + }, + { + "epoch": 7.974004726413379, + "grad_norm": 0.17868563532829285, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0162, + "step": 87730 + }, + { + "epoch": 7.974913652063261, + "grad_norm": 0.12474051117897034, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0183, + "step": 87740 + }, + { + "epoch": 7.975822577713143, + "grad_norm": 0.11406674981117249, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0155, + "step": 87750 + }, + { + "epoch": 7.976731503363025, + "grad_norm": 0.1386190503835678, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.017, + "step": 87760 + }, + { + "epoch": 7.977640429012907, + "grad_norm": 0.16638609766960144, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0171, + "step": 87770 + }, + { + "epoch": 7.978549354662789, + "grad_norm": 0.11938568949699402, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0184, + "step": 87780 + }, + { + "epoch": 7.97945828031267, + "grad_norm": 0.16774483025074005, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0155, + "step": 87790 + }, + { + "epoch": 7.980367205962552, + "grad_norm": 0.11827901005744934, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0187, + "step": 87800 + }, + { + "epoch": 7.981276131612434, + "grad_norm": 0.12898637354373932, + "learning_rate": 5.525220867374957e-06, + "loss": 0.019, + "step": 87810 + }, + { + "epoch": 7.982185057262316, + "grad_norm": 0.1218302845954895, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.017, + "step": 87820 + }, + { + "epoch": 7.983093982912198, + "grad_norm": 0.15415963530540466, + "learning_rate": 5.518813027022363e-06, + "loss": 0.0182, + "step": 87830 + }, + { + "epoch": 7.984002908562079, + "grad_norm": 0.10622833669185638, + "learning_rate": 5.515610186774787e-06, + "loss": 0.016, + "step": 87840 + }, + { + "epoch": 7.984911834211961, + "grad_norm": 0.12137220054864883, + "learning_rate": 5.512408066804025e-06, + "loss": 0.0181, + "step": 87850 + }, + { + "epoch": 7.985820759861843, + "grad_norm": 0.12635624408721924, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0161, + "step": 87860 + }, + { + "epoch": 7.986729685511725, + "grad_norm": 0.11967110633850098, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0179, + "step": 87870 + }, + { + "epoch": 7.987638611161607, + "grad_norm": 0.14369527995586395, + "learning_rate": 5.502806030984159e-06, + "loss": 0.0175, + "step": 87880 + }, + { + "epoch": 7.988547536811489, + "grad_norm": 0.1265021711587906, + "learning_rate": 5.499606794551913e-06, + "loss": 0.0171, + "step": 87890 + }, + { + "epoch": 7.989456462461371, + "grad_norm": 0.12520307302474976, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0175, + "step": 87900 + }, + { + "epoch": 7.990365388111252, + "grad_norm": 0.1021786779165268, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0163, + "step": 87910 + }, + { + "epoch": 7.991274313761134, + "grad_norm": 0.13399212062358856, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0156, + "step": 87920 + }, + { + "epoch": 7.992183239411016, + "grad_norm": 0.24197299778461456, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0156, + "step": 87930 + }, + { + "epoch": 7.993092165060898, + "grad_norm": 0.14635512232780457, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0178, + "step": 87940 + }, + { + "epoch": 7.99400109071078, + "grad_norm": 0.11638598889112473, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0171, + "step": 87950 + }, + { + "epoch": 7.994910016360661, + "grad_norm": 0.09378086775541306, + "learning_rate": 5.477232354905746e-06, + "loss": 0.0157, + "step": 87960 + }, + { + "epoch": 7.995818942010543, + "grad_norm": 0.13029512763023376, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0168, + "step": 87970 + }, + { + "epoch": 7.996727867660425, + "grad_norm": 0.149397611618042, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0176, + "step": 87980 + }, + { + "epoch": 7.997636793310307, + "grad_norm": 0.12488090991973877, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0161, + "step": 87990 + }, + { + "epoch": 7.998545718960189, + "grad_norm": 0.16336293518543243, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0166, + "step": 88000 + }, + { + "epoch": 7.999454644610071, + "grad_norm": 0.1246187761425972, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0196, + "step": 88010 + }, + { + "epoch": 8.000363570259953, + "grad_norm": 0.11960507184267044, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0176, + "step": 88020 + }, + { + "epoch": 8.001272495909834, + "grad_norm": 0.11622167378664017, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0169, + "step": 88030 + }, + { + "epoch": 8.002181421559717, + "grad_norm": 0.1862000972032547, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.0174, + "step": 88040 + }, + { + "epoch": 8.003090347209598, + "grad_norm": 0.14012783765792847, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0161, + "step": 88050 + }, + { + "epoch": 8.00399927285948, + "grad_norm": 0.12748371064662933, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0185, + "step": 88060 + }, + { + "epoch": 8.004908198509362, + "grad_norm": 0.11745677888393402, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0178, + "step": 88070 + }, + { + "epoch": 8.005817124159243, + "grad_norm": 0.16359636187553406, + "learning_rate": 5.438958662497988e-06, + "loss": 0.0187, + "step": 88080 + }, + { + "epoch": 8.006726049809126, + "grad_norm": 0.14438574016094208, + "learning_rate": 5.435773901998663e-06, + "loss": 0.0161, + "step": 88090 + }, + { + "epoch": 8.007634975459007, + "grad_norm": 0.12426578253507614, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0182, + "step": 88100 + }, + { + "epoch": 8.00854390110889, + "grad_norm": 0.14302968978881836, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0188, + "step": 88110 + }, + { + "epoch": 8.009452826758771, + "grad_norm": 0.11283335089683533, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0175, + "step": 88120 + }, + { + "epoch": 8.010361752408652, + "grad_norm": 0.1077876091003418, + "learning_rate": 5.423042125817681e-06, + "loss": 0.0165, + "step": 88130 + }, + { + "epoch": 8.011270678058535, + "grad_norm": 0.12204215675592422, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0167, + "step": 88140 + }, + { + "epoch": 8.012179603708416, + "grad_norm": 0.11705002188682556, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0162, + "step": 88150 + }, + { + "epoch": 8.0130885293583, + "grad_norm": 0.12886397540569305, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0157, + "step": 88160 + }, + { + "epoch": 8.01399745500818, + "grad_norm": 0.13290512561798096, + "learning_rate": 5.410321986543565e-06, + "loss": 0.0182, + "step": 88170 + }, + { + "epoch": 8.014906380658061, + "grad_norm": 0.11079583317041397, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0147, + "step": 88180 + }, + { + "epoch": 8.015815306307944, + "grad_norm": 0.15742529928684235, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0177, + "step": 88190 + }, + { + "epoch": 8.016724231957825, + "grad_norm": 0.20450527966022491, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0195, + "step": 88200 + }, + { + "epoch": 8.017633157607708, + "grad_norm": 0.1125643402338028, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0171, + "step": 88210 + }, + { + "epoch": 8.01854208325759, + "grad_norm": 0.11720845848321915, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0171, + "step": 88220 + }, + { + "epoch": 8.019451008907472, + "grad_norm": 0.1283055990934372, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0196, + "step": 88230 + }, + { + "epoch": 8.020359934557353, + "grad_norm": 0.15129031240940094, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0164, + "step": 88240 + }, + { + "epoch": 8.021268860207234, + "grad_norm": 0.16832856833934784, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0167, + "step": 88250 + }, + { + "epoch": 8.022177785857117, + "grad_norm": 0.1070980355143547, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0153, + "step": 88260 + }, + { + "epoch": 8.023086711506998, + "grad_norm": 0.11879854649305344, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0163, + "step": 88270 + }, + { + "epoch": 8.023995637156881, + "grad_norm": 0.10684356093406677, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0166, + "step": 88280 + }, + { + "epoch": 8.024904562806762, + "grad_norm": 0.1189311295747757, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0187, + "step": 88290 + }, + { + "epoch": 8.025813488456643, + "grad_norm": 0.13955660164356232, + "learning_rate": 5.369062087996716e-06, + "loss": 0.0178, + "step": 88300 + }, + { + "epoch": 8.026722414106526, + "grad_norm": 0.13197879493236542, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.016, + "step": 88310 + }, + { + "epoch": 8.027631339756407, + "grad_norm": 0.11741073429584503, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0199, + "step": 88320 + }, + { + "epoch": 8.02854026540629, + "grad_norm": 0.12207334488630295, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0163, + "step": 88330 + }, + { + "epoch": 8.029449191056171, + "grad_norm": 0.1194332167506218, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0182, + "step": 88340 + }, + { + "epoch": 8.030358116706054, + "grad_norm": 0.14852367341518402, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0184, + "step": 88350 + }, + { + "epoch": 8.031267042355935, + "grad_norm": 0.1926189661026001, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0182, + "step": 88360 + }, + { + "epoch": 8.032175968005816, + "grad_norm": 0.09735077619552612, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0172, + "step": 88370 + }, + { + "epoch": 8.0330848936557, + "grad_norm": 0.11712775379419327, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0163, + "step": 88380 + }, + { + "epoch": 8.03399381930558, + "grad_norm": 0.18699301779270172, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0184, + "step": 88390 + }, + { + "epoch": 8.034902744955463, + "grad_norm": 0.11312249302864075, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0185, + "step": 88400 + }, + { + "epoch": 8.035811670605344, + "grad_norm": 0.15726235508918762, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0168, + "step": 88410 + }, + { + "epoch": 8.036720596255226, + "grad_norm": 0.14567366242408752, + "learning_rate": 5.331085726584114e-06, + "loss": 0.016, + "step": 88420 + }, + { + "epoch": 8.037629521905108, + "grad_norm": 0.19629421830177307, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0167, + "step": 88430 + }, + { + "epoch": 8.03853844755499, + "grad_norm": 0.15742109715938568, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0182, + "step": 88440 + }, + { + "epoch": 8.039447373204872, + "grad_norm": 0.08938880264759064, + "learning_rate": 5.321608141548763e-06, + "loss": 0.0164, + "step": 88450 + }, + { + "epoch": 8.040356298854753, + "grad_norm": 0.12420102208852768, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0166, + "step": 88460 + }, + { + "epoch": 8.041265224504636, + "grad_norm": 0.10563952475786209, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0165, + "step": 88470 + }, + { + "epoch": 8.042174150154517, + "grad_norm": 0.12194638699293137, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0182, + "step": 88480 + }, + { + "epoch": 8.043083075804399, + "grad_norm": 0.16041532158851624, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0168, + "step": 88490 + }, + { + "epoch": 8.043992001454281, + "grad_norm": 0.17581085860729218, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0156, + "step": 88500 + }, + { + "epoch": 8.044900927104162, + "grad_norm": 0.12863574922084808, + "learning_rate": 5.302672823220107e-06, + "loss": 0.018, + "step": 88510 + }, + { + "epoch": 8.045809852754045, + "grad_norm": 0.11486513912677765, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0176, + "step": 88520 + }, + { + "epoch": 8.046718778403926, + "grad_norm": 0.14763370156288147, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0187, + "step": 88530 + }, + { + "epoch": 8.047627704053808, + "grad_norm": 0.15798059105873108, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0202, + "step": 88540 + }, + { + "epoch": 8.04853662970369, + "grad_norm": 0.14648722112178802, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0168, + "step": 88550 + }, + { + "epoch": 8.049445555353572, + "grad_norm": 0.15691418945789337, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0177, + "step": 88560 + }, + { + "epoch": 8.050354481003454, + "grad_norm": 0.16431930661201477, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.0166, + "step": 88570 + }, + { + "epoch": 8.051263406653336, + "grad_norm": 0.14576475322246552, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0162, + "step": 88580 + }, + { + "epoch": 8.052172332303218, + "grad_norm": 0.13223417103290558, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0152, + "step": 88590 + }, + { + "epoch": 8.0530812579531, + "grad_norm": 0.12490204721689224, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0186, + "step": 88600 + }, + { + "epoch": 8.05399018360298, + "grad_norm": 0.10603321343660355, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0177, + "step": 88610 + }, + { + "epoch": 8.054899109252863, + "grad_norm": 0.10656800866127014, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0176, + "step": 88620 + }, + { + "epoch": 8.055808034902745, + "grad_norm": 0.09632960706949234, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0161, + "step": 88630 + }, + { + "epoch": 8.056716960552627, + "grad_norm": 0.1110493391752243, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0159, + "step": 88640 + }, + { + "epoch": 8.057625886202509, + "grad_norm": 0.10901741683483124, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0149, + "step": 88650 + }, + { + "epoch": 8.05853481185239, + "grad_norm": 0.12408575415611267, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0162, + "step": 88660 + }, + { + "epoch": 8.059443737502272, + "grad_norm": 0.13078229129314423, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0171, + "step": 88670 + }, + { + "epoch": 8.060352663152154, + "grad_norm": 0.08969324082136154, + "learning_rate": 5.249166901686548e-06, + "loss": 0.016, + "step": 88680 + }, + { + "epoch": 8.061261588802036, + "grad_norm": 0.1300882250070572, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0188, + "step": 88690 + }, + { + "epoch": 8.062170514451918, + "grad_norm": 0.12496684491634369, + "learning_rate": 5.242886134420562e-06, + "loss": 0.018, + "step": 88700 + }, + { + "epoch": 8.0630794401018, + "grad_norm": 0.1546122282743454, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0162, + "step": 88710 + }, + { + "epoch": 8.063988365751682, + "grad_norm": 0.12064792960882187, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0166, + "step": 88720 + }, + { + "epoch": 8.064897291401563, + "grad_norm": 0.12977264821529388, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0181, + "step": 88730 + }, + { + "epoch": 8.065806217051446, + "grad_norm": 0.15909534692764282, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0176, + "step": 88740 + }, + { + "epoch": 8.066715142701327, + "grad_norm": 0.10233619064092636, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0164, + "step": 88750 + }, + { + "epoch": 8.06762406835121, + "grad_norm": 0.10326539725065231, + "learning_rate": 5.22406162397125e-06, + "loss": 0.0176, + "step": 88760 + }, + { + "epoch": 8.06853299400109, + "grad_norm": 0.12455225735902786, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0173, + "step": 88770 + }, + { + "epoch": 8.069441919650972, + "grad_norm": 0.13753126561641693, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0178, + "step": 88780 + }, + { + "epoch": 8.070350845300855, + "grad_norm": 0.10488937050104141, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0151, + "step": 88790 + }, + { + "epoch": 8.071259770950736, + "grad_norm": 0.14594152569770813, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0181, + "step": 88800 + }, + { + "epoch": 8.072168696600619, + "grad_norm": 0.1395723670721054, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.017, + "step": 88810 + }, + { + "epoch": 8.0730776222505, + "grad_norm": 0.10660012066364288, + "learning_rate": 5.205263840564161e-06, + "loss": 0.0163, + "step": 88820 + }, + { + "epoch": 8.073986547900383, + "grad_norm": 0.19100792706012726, + "learning_rate": 5.202133478735065e-06, + "loss": 0.0144, + "step": 88830 + }, + { + "epoch": 8.074895473550264, + "grad_norm": 0.1701461672782898, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0186, + "step": 88840 + }, + { + "epoch": 8.075804399200145, + "grad_norm": 0.1408599466085434, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0175, + "step": 88850 + }, + { + "epoch": 8.076713324850028, + "grad_norm": 0.11358015984296799, + "learning_rate": 5.192746858698216e-06, + "loss": 0.017, + "step": 88860 + }, + { + "epoch": 8.077622250499909, + "grad_norm": 0.10604238510131836, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0178, + "step": 88870 + }, + { + "epoch": 8.078531176149792, + "grad_norm": 0.10874045640230179, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0194, + "step": 88880 + }, + { + "epoch": 8.079440101799673, + "grad_norm": 0.12650857865810394, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0216, + "step": 88890 + }, + { + "epoch": 8.080349027449554, + "grad_norm": 0.13920475542545319, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0166, + "step": 88900 + }, + { + "epoch": 8.081257953099437, + "grad_norm": 0.1228310838341713, + "learning_rate": 5.177117391058615e-06, + "loss": 0.0182, + "step": 88910 + }, + { + "epoch": 8.082166878749318, + "grad_norm": 0.12791907787322998, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0186, + "step": 88920 + }, + { + "epoch": 8.0830758043992, + "grad_norm": 0.201776385307312, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0179, + "step": 88930 + }, + { + "epoch": 8.083984730049082, + "grad_norm": 0.10891737043857574, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0181, + "step": 88940 + }, + { + "epoch": 8.084893655698965, + "grad_norm": 0.1967482566833496, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0194, + "step": 88950 + }, + { + "epoch": 8.085802581348846, + "grad_norm": 0.13946330547332764, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0188, + "step": 88960 + }, + { + "epoch": 8.086711506998727, + "grad_norm": 0.19710589945316315, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0184, + "step": 88970 + }, + { + "epoch": 8.08762043264861, + "grad_norm": 0.10150504112243652, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0156, + "step": 88980 + }, + { + "epoch": 8.08852935829849, + "grad_norm": 0.09663446247577667, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0155, + "step": 88990 + }, + { + "epoch": 8.089438283948374, + "grad_norm": 0.15276113152503967, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0163, + "step": 89000 + }, + { + "epoch": 8.090347209598255, + "grad_norm": 0.1557604968547821, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0184, + "step": 89010 + }, + { + "epoch": 8.091256135248138, + "grad_norm": 0.11251068860292435, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0177, + "step": 89020 + }, + { + "epoch": 8.092165060898019, + "grad_norm": 0.11069308966398239, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0193, + "step": 89030 + }, + { + "epoch": 8.0930739865479, + "grad_norm": 0.17758604884147644, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0171, + "step": 89040 + }, + { + "epoch": 8.093982912197783, + "grad_norm": 0.1837311089038849, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.0189, + "step": 89050 + }, + { + "epoch": 8.094891837847664, + "grad_norm": 0.1934235394001007, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0155, + "step": 89060 + }, + { + "epoch": 8.095800763497547, + "grad_norm": 0.11384861171245575, + "learning_rate": 5.12722856576104e-06, + "loss": 0.02, + "step": 89070 + }, + { + "epoch": 8.096709689147428, + "grad_norm": 0.09689206629991531, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0172, + "step": 89080 + }, + { + "epoch": 8.097618614797309, + "grad_norm": 0.1178361177444458, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0201, + "step": 89090 + }, + { + "epoch": 8.098527540447192, + "grad_norm": 0.1473473310470581, + "learning_rate": 5.117895750297807e-06, + "loss": 0.019, + "step": 89100 + }, + { + "epoch": 8.099436466097073, + "grad_norm": 0.15765003859996796, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0167, + "step": 89110 + }, + { + "epoch": 8.100345391746956, + "grad_norm": 0.13937141001224518, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0198, + "step": 89120 + }, + { + "epoch": 8.101254317396837, + "grad_norm": 0.19260820746421814, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0199, + "step": 89130 + }, + { + "epoch": 8.10216324304672, + "grad_norm": 0.12680870294570923, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0155, + "step": 89140 + }, + { + "epoch": 8.1030721686966, + "grad_norm": 0.13494329154491425, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0161, + "step": 89150 + }, + { + "epoch": 8.103981094346482, + "grad_norm": 0.10179483890533447, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0168, + "step": 89160 + }, + { + "epoch": 8.104890019996365, + "grad_norm": 0.11090024560689926, + "learning_rate": 5.096145458719924e-06, + "loss": 0.0194, + "step": 89170 + }, + { + "epoch": 8.105798945646246, + "grad_norm": 0.10273569077253342, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0167, + "step": 89180 + }, + { + "epoch": 8.106707871296129, + "grad_norm": 0.10472586005926132, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0197, + "step": 89190 + }, + { + "epoch": 8.10761679694601, + "grad_norm": 0.12022123485803604, + "learning_rate": 5.086835182454458e-06, + "loss": 0.0154, + "step": 89200 + }, + { + "epoch": 8.10852572259589, + "grad_norm": 0.15342597663402557, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0168, + "step": 89210 + }, + { + "epoch": 8.109434648245774, + "grad_norm": 0.11723445355892181, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0179, + "step": 89220 + }, + { + "epoch": 8.110343573895655, + "grad_norm": 0.13599728047847748, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0166, + "step": 89230 + }, + { + "epoch": 8.111252499545538, + "grad_norm": 0.12338422983884811, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0167, + "step": 89240 + }, + { + "epoch": 8.112161425195419, + "grad_norm": 0.09063027054071426, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0188, + "step": 89250 + }, + { + "epoch": 8.113070350845302, + "grad_norm": 0.08855520933866501, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0174, + "step": 89260 + }, + { + "epoch": 8.113979276495183, + "grad_norm": 0.10985587537288666, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0162, + "step": 89270 + }, + { + "epoch": 8.114888202145064, + "grad_norm": 0.0969245657324791, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0165, + "step": 89280 + }, + { + "epoch": 8.115797127794947, + "grad_norm": 0.15847943723201752, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0155, + "step": 89290 + }, + { + "epoch": 8.116706053444828, + "grad_norm": 0.21233099699020386, + "learning_rate": 5.05584989850861e-06, + "loss": 0.017, + "step": 89300 + }, + { + "epoch": 8.11761497909471, + "grad_norm": 0.10578477382659912, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0145, + "step": 89310 + }, + { + "epoch": 8.118523904744592, + "grad_norm": 0.1344679445028305, + "learning_rate": 5.049661896504736e-06, + "loss": 0.0167, + "step": 89320 + }, + { + "epoch": 8.119432830394473, + "grad_norm": 0.12817969918251038, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0161, + "step": 89330 + }, + { + "epoch": 8.120341756044356, + "grad_norm": 0.12262337654829025, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0157, + "step": 89340 + }, + { + "epoch": 8.121250681694237, + "grad_norm": 0.12067801505327225, + "learning_rate": 5.040385561554829e-06, + "loss": 0.0147, + "step": 89350 + }, + { + "epoch": 8.12215960734412, + "grad_norm": 0.11118044704198837, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0129, + "step": 89360 + }, + { + "epoch": 8.123068532994001, + "grad_norm": 0.11999958008527756, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0174, + "step": 89370 + }, + { + "epoch": 8.123977458643884, + "grad_norm": 0.12480052560567856, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0187, + "step": 89380 + }, + { + "epoch": 8.124886384293765, + "grad_norm": 0.1284300833940506, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0166, + "step": 89390 + }, + { + "epoch": 8.125795309943646, + "grad_norm": 0.1476324200630188, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0173, + "step": 89400 + }, + { + "epoch": 8.126704235593529, + "grad_norm": 0.1742507368326187, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0184, + "step": 89410 + }, + { + "epoch": 8.12761316124341, + "grad_norm": 0.21142393350601196, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0182, + "step": 89420 + }, + { + "epoch": 8.128522086893293, + "grad_norm": 0.2010597139596939, + "learning_rate": 5.0156819657067e-06, + "loss": 0.019, + "step": 89430 + }, + { + "epoch": 8.129431012543174, + "grad_norm": 0.13807940483093262, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0198, + "step": 89440 + }, + { + "epoch": 8.130339938193055, + "grad_norm": 0.15614421665668488, + "learning_rate": 5.009513644487276e-06, + "loss": 0.0157, + "step": 89450 + }, + { + "epoch": 8.131248863842938, + "grad_norm": 0.11784721165895462, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0162, + "step": 89460 + }, + { + "epoch": 8.132157789492819, + "grad_norm": 0.1025172621011734, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0151, + "step": 89470 + }, + { + "epoch": 8.133066715142702, + "grad_norm": 0.1283038705587387, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0163, + "step": 89480 + }, + { + "epoch": 8.133975640792583, + "grad_norm": 0.19410385191440582, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0173, + "step": 89490 + }, + { + "epoch": 8.134884566442466, + "grad_norm": 0.10014474391937256, + "learning_rate": 4.994106122986448e-06, + "loss": 0.0166, + "step": 89500 + }, + { + "epoch": 8.135793492092347, + "grad_norm": 0.11583993583917618, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0148, + "step": 89510 + }, + { + "epoch": 8.136702417742228, + "grad_norm": 0.08462053537368774, + "learning_rate": 4.98794843191963e-06, + "loss": 0.0145, + "step": 89520 + }, + { + "epoch": 8.137611343392111, + "grad_norm": 0.08099109679460526, + "learning_rate": 4.984870726793188e-06, + "loss": 0.016, + "step": 89530 + }, + { + "epoch": 8.138520269041992, + "grad_norm": 0.12948212027549744, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0199, + "step": 89540 + }, + { + "epoch": 8.139429194691875, + "grad_norm": 0.17171449959278107, + "learning_rate": 4.978717598522905e-06, + "loss": 0.0163, + "step": 89550 + }, + { + "epoch": 8.140338120341756, + "grad_norm": 0.12870457768440247, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0161, + "step": 89560 + }, + { + "epoch": 8.141247045991637, + "grad_norm": 0.11948030441999435, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0175, + "step": 89570 + }, + { + "epoch": 8.14215597164152, + "grad_norm": 0.13623479008674622, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0186, + "step": 89580 + }, + { + "epoch": 8.143064897291401, + "grad_norm": 0.11145632714033127, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0174, + "step": 89590 + }, + { + "epoch": 8.143973822941284, + "grad_norm": 0.10654862970113754, + "learning_rate": 4.963348100313064e-06, + "loss": 0.0191, + "step": 89600 + }, + { + "epoch": 8.144882748591165, + "grad_norm": 0.15263207256793976, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0177, + "step": 89610 + }, + { + "epoch": 8.145791674241048, + "grad_norm": 0.10806041955947876, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0162, + "step": 89620 + }, + { + "epoch": 8.146700599890929, + "grad_norm": 0.09847992658615112, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0171, + "step": 89630 + }, + { + "epoch": 8.14760952554081, + "grad_norm": 0.14573878049850464, + "learning_rate": 4.951066220258231e-06, + "loss": 0.0165, + "step": 89640 + }, + { + "epoch": 8.148518451190693, + "grad_norm": 0.1206236407160759, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0162, + "step": 89650 + }, + { + "epoch": 8.149427376840574, + "grad_norm": 0.13636384904384613, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0178, + "step": 89660 + }, + { + "epoch": 8.150336302490457, + "grad_norm": 0.12429415434598923, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0177, + "step": 89670 + }, + { + "epoch": 8.151245228140338, + "grad_norm": 0.12695737183094025, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0161, + "step": 89680 + }, + { + "epoch": 8.15215415379022, + "grad_norm": 0.12487169355154037, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0141, + "step": 89690 + }, + { + "epoch": 8.153063079440102, + "grad_norm": 0.11824563890695572, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0152, + "step": 89700 + }, + { + "epoch": 8.153972005089983, + "grad_norm": 0.09119894355535507, + "learning_rate": 4.929602320411451e-06, + "loss": 0.0159, + "step": 89710 + }, + { + "epoch": 8.154880930739866, + "grad_norm": 0.13754718005657196, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0175, + "step": 89720 + }, + { + "epoch": 8.155789856389747, + "grad_norm": 0.1210225373506546, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0172, + "step": 89730 + }, + { + "epoch": 8.15669878203963, + "grad_norm": 0.13163743913173676, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0175, + "step": 89740 + }, + { + "epoch": 8.157607707689511, + "grad_norm": 0.11161406338214874, + "learning_rate": 4.917354054827757e-06, + "loss": 0.016, + "step": 89750 + }, + { + "epoch": 8.158516633339392, + "grad_norm": 0.14797712862491608, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0183, + "step": 89760 + }, + { + "epoch": 8.159425558989275, + "grad_norm": 0.16949792206287384, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0177, + "step": 89770 + }, + { + "epoch": 8.160334484639156, + "grad_norm": 0.16994641721248627, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0187, + "step": 89780 + }, + { + "epoch": 8.161243410289039, + "grad_norm": 0.11966695636510849, + "learning_rate": 4.905118040609615e-06, + "loss": 0.0169, + "step": 89790 + }, + { + "epoch": 8.16215233593892, + "grad_norm": 0.10213067382574081, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.0176, + "step": 89800 + }, + { + "epoch": 8.163061261588801, + "grad_norm": 0.15563593804836273, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0177, + "step": 89810 + }, + { + "epoch": 8.163970187238684, + "grad_norm": 0.12915877997875214, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0153, + "step": 89820 + }, + { + "epoch": 8.164879112888565, + "grad_norm": 0.1537480503320694, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0168, + "step": 89830 + }, + { + "epoch": 8.165788038538448, + "grad_norm": 0.1161641776561737, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.0163, + "step": 89840 + }, + { + "epoch": 8.16669696418833, + "grad_norm": 0.21154922246932983, + "learning_rate": 4.886787023112308e-06, + "loss": 0.0169, + "step": 89850 + }, + { + "epoch": 8.167605889838212, + "grad_norm": 0.1653720885515213, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0165, + "step": 89860 + }, + { + "epoch": 8.168514815488093, + "grad_norm": 0.12915337085723877, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0193, + "step": 89870 + }, + { + "epoch": 8.169423741137974, + "grad_norm": 0.14591346681118011, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0174, + "step": 89880 + }, + { + "epoch": 8.170332666787857, + "grad_norm": 0.16512338817119598, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0166, + "step": 89890 + }, + { + "epoch": 8.171241592437738, + "grad_norm": 0.11468423902988434, + "learning_rate": 4.871532293942683e-06, + "loss": 0.0161, + "step": 89900 + }, + { + "epoch": 8.172150518087621, + "grad_norm": 0.13888069987297058, + "learning_rate": 4.868483654753097e-06, + "loss": 0.0156, + "step": 89910 + }, + { + "epoch": 8.173059443737502, + "grad_norm": 0.2208104431629181, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0173, + "step": 89920 + }, + { + "epoch": 8.173968369387383, + "grad_norm": 0.13245511054992676, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0175, + "step": 89930 + }, + { + "epoch": 8.174877295037266, + "grad_norm": 0.15549983084201813, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0153, + "step": 89940 + }, + { + "epoch": 8.175786220687147, + "grad_norm": 0.20897771418094635, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0173, + "step": 89950 + }, + { + "epoch": 8.17669514633703, + "grad_norm": 0.1311911642551422, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0179, + "step": 89960 + }, + { + "epoch": 8.177604071986911, + "grad_norm": 0.11216913908720016, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0158, + "step": 89970 + }, + { + "epoch": 8.178512997636794, + "grad_norm": 0.11827854812145233, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0146, + "step": 89980 + }, + { + "epoch": 8.179421923286675, + "grad_norm": 0.12410406768321991, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0187, + "step": 89990 + }, + { + "epoch": 8.180330848936556, + "grad_norm": 0.10701246559619904, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0164, + "step": 90000 + }, + { + "epoch": 8.18123977458644, + "grad_norm": 0.11192013323307037, + "learning_rate": 4.838039619228974e-06, + "loss": 0.0164, + "step": 90010 + }, + { + "epoch": 8.18214870023632, + "grad_norm": 0.1657409369945526, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0161, + "step": 90020 + }, + { + "epoch": 8.183057625886203, + "grad_norm": 0.13312587141990662, + "learning_rate": 4.831960065536115e-06, + "loss": 0.016, + "step": 90030 + }, + { + "epoch": 8.183966551536084, + "grad_norm": 0.11652741581201553, + "learning_rate": 4.828921446867572e-06, + "loss": 0.0176, + "step": 90040 + }, + { + "epoch": 8.184875477185965, + "grad_norm": 0.15335039794445038, + "learning_rate": 4.825883600625316e-06, + "loss": 0.017, + "step": 90050 + }, + { + "epoch": 8.185784402835848, + "grad_norm": 0.12856994569301605, + "learning_rate": 4.822846527040058e-06, + "loss": 0.016, + "step": 90060 + }, + { + "epoch": 8.18669332848573, + "grad_norm": 0.33788079023361206, + "learning_rate": 4.819810226342449e-06, + "loss": 0.0171, + "step": 90070 + }, + { + "epoch": 8.187602254135612, + "grad_norm": 0.1254245936870575, + "learning_rate": 4.816774698763067e-06, + "loss": 0.015, + "step": 90080 + }, + { + "epoch": 8.188511179785493, + "grad_norm": 0.13167496025562286, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0159, + "step": 90090 + }, + { + "epoch": 8.189420105435376, + "grad_norm": 0.19020242989063263, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0173, + "step": 90100 + }, + { + "epoch": 8.190329031085257, + "grad_norm": 0.12121471762657166, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0213, + "step": 90110 + }, + { + "epoch": 8.191237956735138, + "grad_norm": 0.15080100297927856, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0187, + "step": 90120 + }, + { + "epoch": 8.192146882385021, + "grad_norm": 0.09888605028390884, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0162, + "step": 90130 + }, + { + "epoch": 8.193055808034902, + "grad_norm": 0.16634970903396606, + "learning_rate": 4.798577781675066e-06, + "loss": 0.0173, + "step": 90140 + }, + { + "epoch": 8.193964733684785, + "grad_norm": 0.1597321480512619, + "learning_rate": 4.795547672374798e-06, + "loss": 0.017, + "step": 90150 + }, + { + "epoch": 8.194873659334666, + "grad_norm": 0.10088713467121124, + "learning_rate": 4.792518338035334e-06, + "loss": 0.0186, + "step": 90160 + }, + { + "epoch": 8.19578258498455, + "grad_norm": 0.15661533176898956, + "learning_rate": 4.78948977888674e-06, + "loss": 0.017, + "step": 90170 + }, + { + "epoch": 8.19669151063443, + "grad_norm": 0.14302270114421844, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0142, + "step": 90180 + }, + { + "epoch": 8.197600436284311, + "grad_norm": 0.15392182767391205, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0186, + "step": 90190 + }, + { + "epoch": 8.198509361934194, + "grad_norm": 0.1507337987422943, + "learning_rate": 4.780408754885869e-06, + "loss": 0.0163, + "step": 90200 + }, + { + "epoch": 8.199418287584075, + "grad_norm": 0.16228097677230835, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0167, + "step": 90210 + }, + { + "epoch": 8.200327213233958, + "grad_norm": 0.1264745593070984, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0174, + "step": 90220 + }, + { + "epoch": 8.20123613888384, + "grad_norm": 0.2071687877178192, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0182, + "step": 90230 + }, + { + "epoch": 8.20214506453372, + "grad_norm": 0.17323604226112366, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0178, + "step": 90240 + }, + { + "epoch": 8.203053990183603, + "grad_norm": 0.11810877919197083, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0161, + "step": 90250 + }, + { + "epoch": 8.203962915833484, + "grad_norm": 0.09818526357412338, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0167, + "step": 90260 + }, + { + "epoch": 8.204871841483367, + "grad_norm": 0.11411062628030777, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0171, + "step": 90270 + }, + { + "epoch": 8.205780767133248, + "grad_norm": 0.11772608011960983, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0174, + "step": 90280 + }, + { + "epoch": 8.206689692783131, + "grad_norm": 0.20864525437355042, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0196, + "step": 90290 + }, + { + "epoch": 8.207598618433012, + "grad_norm": 0.1506885290145874, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0172, + "step": 90300 + }, + { + "epoch": 8.208507544082893, + "grad_norm": 0.2158721387386322, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0162, + "step": 90310 + }, + { + "epoch": 8.209416469732776, + "grad_norm": 0.12072665989398956, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0174, + "step": 90320 + }, + { + "epoch": 8.210325395382657, + "grad_norm": 0.11029162257909775, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0167, + "step": 90330 + }, + { + "epoch": 8.21123432103254, + "grad_norm": 0.14688177406787872, + "learning_rate": 4.738123100204238e-06, + "loss": 0.0185, + "step": 90340 + }, + { + "epoch": 8.212143246682421, + "grad_norm": 0.20390009880065918, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0169, + "step": 90350 + }, + { + "epoch": 8.213052172332302, + "grad_norm": 0.11298779398202896, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0181, + "step": 90360 + }, + { + "epoch": 8.213961097982185, + "grad_norm": 0.12290438264608383, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0164, + "step": 90370 + }, + { + "epoch": 8.214870023632066, + "grad_norm": 0.12745791673660278, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0178, + "step": 90380 + }, + { + "epoch": 8.21577894928195, + "grad_norm": 0.1023053228855133, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0179, + "step": 90390 + }, + { + "epoch": 8.21668787493183, + "grad_norm": 0.09325513988733292, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0165, + "step": 90400 + }, + { + "epoch": 8.217596800581713, + "grad_norm": 0.1138942763209343, + "learning_rate": 4.717037513697221e-06, + "loss": 0.0185, + "step": 90410 + }, + { + "epoch": 8.218505726231594, + "grad_norm": 0.0984894186258316, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0171, + "step": 90420 + }, + { + "epoch": 8.219414651881475, + "grad_norm": 0.16726386547088623, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0168, + "step": 90430 + }, + { + "epoch": 8.220323577531358, + "grad_norm": 0.15950843691825867, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0158, + "step": 90440 + }, + { + "epoch": 8.22123250318124, + "grad_norm": 0.10042889416217804, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0167, + "step": 90450 + }, + { + "epoch": 8.222141428831122, + "grad_norm": 0.22714778780937195, + "learning_rate": 4.701999803595437e-06, + "loss": 0.0178, + "step": 90460 + }, + { + "epoch": 8.223050354481003, + "grad_norm": 0.10671450197696686, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0182, + "step": 90470 + }, + { + "epoch": 8.223959280130885, + "grad_norm": 0.09040413051843643, + "learning_rate": 4.695990192412397e-06, + "loss": 0.018, + "step": 90480 + }, + { + "epoch": 8.224868205780767, + "grad_norm": 0.11082901060581207, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0154, + "step": 90490 + }, + { + "epoch": 8.225777131430648, + "grad_norm": 0.15843036770820618, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0168, + "step": 90500 + }, + { + "epoch": 8.226686057080531, + "grad_norm": 0.135324165225029, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.0163, + "step": 90510 + }, + { + "epoch": 8.227594982730412, + "grad_norm": 0.09501850605010986, + "learning_rate": 4.683980362129283e-06, + "loss": 0.016, + "step": 90520 + }, + { + "epoch": 8.228503908380295, + "grad_norm": 0.12232866883277893, + "learning_rate": 4.680979862572639e-06, + "loss": 0.0181, + "step": 90530 + }, + { + "epoch": 8.229412834030176, + "grad_norm": 0.1135229840874672, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0173, + "step": 90540 + }, + { + "epoch": 8.230321759680058, + "grad_norm": 0.17171628773212433, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0195, + "step": 90550 + }, + { + "epoch": 8.23123068532994, + "grad_norm": 0.1856953650712967, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0185, + "step": 90560 + }, + { + "epoch": 8.232139610979821, + "grad_norm": 0.08041879534721375, + "learning_rate": 4.668985703238654e-06, + "loss": 0.0152, + "step": 90570 + }, + { + "epoch": 8.233048536629704, + "grad_norm": 0.154694601893425, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0147, + "step": 90580 + }, + { + "epoch": 8.233957462279585, + "grad_norm": 0.1174919456243515, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0165, + "step": 90590 + }, + { + "epoch": 8.234866387929467, + "grad_norm": 0.11854851990938187, + "learning_rate": 4.659998320950685e-06, + "loss": 0.018, + "step": 90600 + }, + { + "epoch": 8.23577531357935, + "grad_norm": 0.13835008442401886, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0181, + "step": 90610 + }, + { + "epoch": 8.23668423922923, + "grad_norm": 0.1875542551279068, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.0165, + "step": 90620 + }, + { + "epoch": 8.237593164879113, + "grad_norm": 0.13255932927131653, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0175, + "step": 90630 + }, + { + "epoch": 8.238502090528995, + "grad_norm": 0.1252383440732956, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0182, + "step": 90640 + }, + { + "epoch": 8.239411016178877, + "grad_norm": 0.09430450201034546, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0156, + "step": 90650 + }, + { + "epoch": 8.240319941828758, + "grad_norm": 0.13394315540790558, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0192, + "step": 90660 + }, + { + "epoch": 8.24122886747864, + "grad_norm": 0.18144264817237854, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0173, + "step": 90670 + }, + { + "epoch": 8.242137793128522, + "grad_norm": 0.11231903731822968, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0157, + "step": 90680 + }, + { + "epoch": 8.243046718778404, + "grad_norm": 0.09922700375318527, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.02, + "step": 90690 + }, + { + "epoch": 8.243955644428286, + "grad_norm": 0.1280343383550644, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0171, + "step": 90700 + }, + { + "epoch": 8.244864570078168, + "grad_norm": 0.12390051037073135, + "learning_rate": 4.627105097181728e-06, + "loss": 0.0151, + "step": 90710 + }, + { + "epoch": 8.245773495728049, + "grad_norm": 0.12216304987668991, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0161, + "step": 90720 + }, + { + "epoch": 8.246682421377932, + "grad_norm": 0.11554165184497833, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0166, + "step": 90730 + }, + { + "epoch": 8.247591347027813, + "grad_norm": 0.08542884886264801, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0182, + "step": 90740 + }, + { + "epoch": 8.248500272677695, + "grad_norm": 0.1434752643108368, + "learning_rate": 4.615167541602233e-06, + "loss": 0.0155, + "step": 90750 + }, + { + "epoch": 8.249409198327577, + "grad_norm": 0.13462334871292114, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0158, + "step": 90760 + }, + { + "epoch": 8.25031812397746, + "grad_norm": 0.1525416076183319, + "learning_rate": 4.609203494863e-06, + "loss": 0.0174, + "step": 90770 + }, + { + "epoch": 8.25122704962734, + "grad_norm": 0.12247847020626068, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0154, + "step": 90780 + }, + { + "epoch": 8.252135975277222, + "grad_norm": 0.11446035653352737, + "learning_rate": 4.603242604573403e-06, + "loss": 0.015, + "step": 90790 + }, + { + "epoch": 8.253044900927105, + "grad_norm": 0.11713676899671555, + "learning_rate": 4.600263343663104e-06, + "loss": 0.0158, + "step": 90800 + }, + { + "epoch": 8.253953826576986, + "grad_norm": 0.14768370985984802, + "learning_rate": 4.597284872544181e-06, + "loss": 0.0182, + "step": 90810 + }, + { + "epoch": 8.254862752226868, + "grad_norm": 0.10037726163864136, + "learning_rate": 4.59430719144282e-06, + "loss": 0.0148, + "step": 90820 + }, + { + "epoch": 8.25577167787675, + "grad_norm": 0.12016214430332184, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0189, + "step": 90830 + }, + { + "epoch": 8.25668060352663, + "grad_norm": 0.12801522016525269, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0169, + "step": 90840 + }, + { + "epoch": 8.257589529176514, + "grad_norm": 0.08690780401229858, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0154, + "step": 90850 + }, + { + "epoch": 8.258498454826395, + "grad_norm": 0.1683654487133026, + "learning_rate": 4.582404371734787e-06, + "loss": 0.018, + "step": 90860 + }, + { + "epoch": 8.259407380476278, + "grad_norm": 0.11903467029333115, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.0162, + "step": 90870 + }, + { + "epoch": 8.260316306126159, + "grad_norm": 0.13126207888126373, + "learning_rate": 4.57645770786281e-06, + "loss": 0.017, + "step": 90880 + }, + { + "epoch": 8.261225231776042, + "grad_norm": 0.14021581411361694, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0175, + "step": 90890 + }, + { + "epoch": 8.262134157425923, + "grad_norm": 0.2637922167778015, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0168, + "step": 90900 + }, + { + "epoch": 8.263043083075804, + "grad_norm": 0.1413276046514511, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0169, + "step": 90910 + }, + { + "epoch": 8.263952008725687, + "grad_norm": 0.19451701641082764, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0163, + "step": 90920 + }, + { + "epoch": 8.264860934375568, + "grad_norm": 0.20301175117492676, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0185, + "step": 90930 + }, + { + "epoch": 8.26576986002545, + "grad_norm": 0.1245378851890564, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0186, + "step": 90940 + }, + { + "epoch": 8.266678785675332, + "grad_norm": 0.12682873010635376, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0206, + "step": 90950 + }, + { + "epoch": 8.267587711325213, + "grad_norm": 0.18156035244464874, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0185, + "step": 90960 + }, + { + "epoch": 8.268496636975096, + "grad_norm": 0.13812077045440674, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0166, + "step": 90970 + }, + { + "epoch": 8.269405562624977, + "grad_norm": 0.15530087053775787, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.0166, + "step": 90980 + }, + { + "epoch": 8.27031448827486, + "grad_norm": 0.10187815129756927, + "learning_rate": 4.543807704298849e-06, + "loss": 0.0179, + "step": 90990 + }, + { + "epoch": 8.27122341392474, + "grad_norm": 0.1914108395576477, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0163, + "step": 91000 + }, + { + "epoch": 8.272132339574624, + "grad_norm": 0.11026380211114883, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0171, + "step": 91010 + }, + { + "epoch": 8.273041265224505, + "grad_norm": 0.13893669843673706, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0176, + "step": 91020 + }, + { + "epoch": 8.273950190874386, + "grad_norm": 0.16849997639656067, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0203, + "step": 91030 + }, + { + "epoch": 8.274859116524269, + "grad_norm": 0.12290173023939133, + "learning_rate": 4.528998538715274e-06, + "loss": 0.016, + "step": 91040 + }, + { + "epoch": 8.27576804217415, + "grad_norm": 0.13138112425804138, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0168, + "step": 91050 + }, + { + "epoch": 8.276676967824033, + "grad_norm": 0.1133439764380455, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0192, + "step": 91060 + }, + { + "epoch": 8.277585893473914, + "grad_norm": 0.20777902007102966, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0184, + "step": 91070 + }, + { + "epoch": 8.278494819123797, + "grad_norm": 0.16769301891326904, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0178, + "step": 91080 + }, + { + "epoch": 8.279403744773678, + "grad_norm": 0.14324890077114105, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0151, + "step": 91090 + }, + { + "epoch": 8.280312670423559, + "grad_norm": 0.12207535654306412, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0164, + "step": 91100 + }, + { + "epoch": 8.281221596073442, + "grad_norm": 0.13703523576259613, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0153, + "step": 91110 + }, + { + "epoch": 8.282130521723323, + "grad_norm": 0.1390964686870575, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0159, + "step": 91120 + }, + { + "epoch": 8.283039447373206, + "grad_norm": 0.14959841966629028, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0158, + "step": 91130 + }, + { + "epoch": 8.283948373023087, + "grad_norm": 0.16387932002544403, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0161, + "step": 91140 + }, + { + "epoch": 8.284857298672968, + "grad_norm": 0.26732921600341797, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0204, + "step": 91150 + }, + { + "epoch": 8.28576622432285, + "grad_norm": 0.11891240626573563, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0161, + "step": 91160 + }, + { + "epoch": 8.286675149972732, + "grad_norm": 0.14962878823280334, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0161, + "step": 91170 + }, + { + "epoch": 8.287584075622615, + "grad_norm": 0.1257023960351944, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0175, + "step": 91180 + }, + { + "epoch": 8.288493001272496, + "grad_norm": 0.12309221178293228, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0154, + "step": 91190 + }, + { + "epoch": 8.289401926922377, + "grad_norm": 0.19105888903141022, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0159, + "step": 91200 + }, + { + "epoch": 8.29031085257226, + "grad_norm": 0.14213503897190094, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0171, + "step": 91210 + }, + { + "epoch": 8.29121977822214, + "grad_norm": 0.11691510677337646, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0143, + "step": 91220 + }, + { + "epoch": 8.292128703872024, + "grad_norm": 0.13078463077545166, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0173, + "step": 91230 + }, + { + "epoch": 8.293037629521905, + "grad_norm": 0.11580034345388412, + "learning_rate": 4.469960957813453e-06, + "loss": 0.017, + "step": 91240 + }, + { + "epoch": 8.293946555171788, + "grad_norm": 0.16377098858356476, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0162, + "step": 91250 + }, + { + "epoch": 8.294855480821669, + "grad_norm": 0.134069561958313, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0156, + "step": 91260 + }, + { + "epoch": 8.29576440647155, + "grad_norm": 0.14216166734695435, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0135, + "step": 91270 + }, + { + "epoch": 8.296673332121433, + "grad_norm": 0.16636334359645844, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0182, + "step": 91280 + }, + { + "epoch": 8.297582257771314, + "grad_norm": 0.11518138647079468, + "learning_rate": 4.455251473055955e-06, + "loss": 0.0166, + "step": 91290 + }, + { + "epoch": 8.298491183421197, + "grad_norm": 0.21436423063278198, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0191, + "step": 91300 + }, + { + "epoch": 8.299400109071078, + "grad_norm": 0.10737928748130798, + "learning_rate": 4.449373283217887e-06, + "loss": 0.0161, + "step": 91310 + }, + { + "epoch": 8.30030903472096, + "grad_norm": 0.1103641614317894, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0154, + "step": 91320 + }, + { + "epoch": 8.301217960370842, + "grad_norm": 0.14566974341869354, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0192, + "step": 91330 + }, + { + "epoch": 8.302126886020723, + "grad_norm": 0.1218113899230957, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0153, + "step": 91340 + }, + { + "epoch": 8.303035811670606, + "grad_norm": 0.12734538316726685, + "learning_rate": 4.437626520331889e-06, + "loss": 0.0161, + "step": 91350 + }, + { + "epoch": 8.303944737320487, + "grad_norm": 0.12188763171434402, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0177, + "step": 91360 + }, + { + "epoch": 8.30485366297037, + "grad_norm": 0.18599626421928406, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0139, + "step": 91370 + }, + { + "epoch": 8.30576258862025, + "grad_norm": 0.15331774950027466, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0175, + "step": 91380 + }, + { + "epoch": 8.306671514270132, + "grad_norm": 0.1262824684381485, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0166, + "step": 91390 + }, + { + "epoch": 8.307580439920015, + "grad_norm": 0.13098442554473877, + "learning_rate": 4.422961116601457e-06, + "loss": 0.0183, + "step": 91400 + }, + { + "epoch": 8.308489365569896, + "grad_norm": 0.2536260485649109, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0183, + "step": 91410 + }, + { + "epoch": 8.309398291219779, + "grad_norm": 0.18868441879749298, + "learning_rate": 4.417100576344586e-06, + "loss": 0.0164, + "step": 91420 + }, + { + "epoch": 8.31030721686966, + "grad_norm": 0.121228888630867, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0161, + "step": 91430 + }, + { + "epoch": 8.311216142519541, + "grad_norm": 0.14118549227714539, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0213, + "step": 91440 + }, + { + "epoch": 8.312125068169424, + "grad_norm": 0.16709722578525543, + "learning_rate": 4.408315794275502e-06, + "loss": 0.0157, + "step": 91450 + }, + { + "epoch": 8.313033993819305, + "grad_norm": 0.17746609449386597, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0159, + "step": 91460 + }, + { + "epoch": 8.313942919469188, + "grad_norm": 0.10284269601106644, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0168, + "step": 91470 + }, + { + "epoch": 8.314851845119069, + "grad_norm": 0.10640730708837509, + "learning_rate": 4.399538251522701e-06, + "loss": 0.0166, + "step": 91480 + }, + { + "epoch": 8.315760770768952, + "grad_norm": 0.1119438111782074, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0165, + "step": 91490 + }, + { + "epoch": 8.316669696418833, + "grad_norm": 0.1379261016845703, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0166, + "step": 91500 + }, + { + "epoch": 8.317578622068714, + "grad_norm": 0.15774144232273102, + "learning_rate": 4.390767954085548e-06, + "loss": 0.0203, + "step": 91510 + }, + { + "epoch": 8.318487547718597, + "grad_norm": 0.17979544401168823, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0189, + "step": 91520 + }, + { + "epoch": 8.319396473368478, + "grad_norm": 0.11294276267290115, + "learning_rate": 4.38492511726322e-06, + "loss": 0.015, + "step": 91530 + }, + { + "epoch": 8.32030539901836, + "grad_norm": 0.22718213498592377, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0155, + "step": 91540 + }, + { + "epoch": 8.321214324668242, + "grad_norm": 0.33582401275634766, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0162, + "step": 91550 + }, + { + "epoch": 8.322123250318125, + "grad_norm": 0.14826594293117523, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0185, + "step": 91560 + }, + { + "epoch": 8.323032175968006, + "grad_norm": 0.16776259243488312, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0159, + "step": 91570 + }, + { + "epoch": 8.323941101617887, + "grad_norm": 0.12554211914539337, + "learning_rate": 4.370332136622768e-06, + "loss": 0.0159, + "step": 91580 + }, + { + "epoch": 8.32485002726777, + "grad_norm": 0.12364080548286438, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0181, + "step": 91590 + }, + { + "epoch": 8.325758952917651, + "grad_norm": 0.11655868589878082, + "learning_rate": 4.364500593587355e-06, + "loss": 0.017, + "step": 91600 + }, + { + "epoch": 8.326667878567534, + "grad_norm": 0.11854947358369827, + "learning_rate": 4.361586033502936e-06, + "loss": 0.0163, + "step": 91610 + }, + { + "epoch": 8.327576804217415, + "grad_norm": 0.13648191094398499, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0153, + "step": 91620 + }, + { + "epoch": 8.328485729867296, + "grad_norm": 0.2789325416088104, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0165, + "step": 91630 + }, + { + "epoch": 8.329394655517179, + "grad_norm": 0.22210922837257385, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0171, + "step": 91640 + }, + { + "epoch": 8.33030358116706, + "grad_norm": 0.14135366678237915, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0198, + "step": 91650 + }, + { + "epoch": 8.331212506816943, + "grad_norm": 0.13678297400474548, + "learning_rate": 4.347025356265514e-06, + "loss": 0.017, + "step": 91660 + }, + { + "epoch": 8.332121432466824, + "grad_norm": 0.17274288833141327, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0174, + "step": 91670 + }, + { + "epoch": 8.333030358116707, + "grad_norm": 0.13798144459724426, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0155, + "step": 91680 + }, + { + "epoch": 8.333939283766588, + "grad_norm": 0.09295523911714554, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0155, + "step": 91690 + }, + { + "epoch": 8.334848209416469, + "grad_norm": 0.146989107131958, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0161, + "step": 91700 + }, + { + "epoch": 8.335757135066352, + "grad_norm": 0.10209048539400101, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0173, + "step": 91710 + }, + { + "epoch": 8.336666060716233, + "grad_norm": 0.11344137042760849, + "learning_rate": 4.329579243769564e-06, + "loss": 0.0177, + "step": 91720 + }, + { + "epoch": 8.337574986366116, + "grad_norm": 0.12404001504182816, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0182, + "step": 91730 + }, + { + "epoch": 8.338483912015997, + "grad_norm": 0.14726704359054565, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0179, + "step": 91740 + }, + { + "epoch": 8.339392837665878, + "grad_norm": 0.17033155262470245, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0174, + "step": 91750 + }, + { + "epoch": 8.340301763315761, + "grad_norm": 0.1587667614221573, + "learning_rate": 4.317964706121786e-06, + "loss": 0.016, + "step": 91760 + }, + { + "epoch": 8.341210688965642, + "grad_norm": 0.15003442764282227, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0169, + "step": 91770 + }, + { + "epoch": 8.342119614615525, + "grad_norm": 0.11773456633090973, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0178, + "step": 91780 + }, + { + "epoch": 8.343028540265406, + "grad_norm": 0.2021755576133728, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0203, + "step": 91790 + }, + { + "epoch": 8.343937465915289, + "grad_norm": 0.15864154696464539, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0166, + "step": 91800 + }, + { + "epoch": 8.34484639156517, + "grad_norm": 0.1078331470489502, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0163, + "step": 91810 + }, + { + "epoch": 8.345755317215051, + "grad_norm": 0.1239318773150444, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0196, + "step": 91820 + }, + { + "epoch": 8.346664242864934, + "grad_norm": 0.11340132355690002, + "learning_rate": 4.297670506074299e-06, + "loss": 0.0164, + "step": 91830 + }, + { + "epoch": 8.347573168514815, + "grad_norm": 0.10717250406742096, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0156, + "step": 91840 + }, + { + "epoch": 8.348482094164698, + "grad_norm": 0.12958912551403046, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0157, + "step": 91850 + }, + { + "epoch": 8.349391019814579, + "grad_norm": 0.10355935990810394, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0177, + "step": 91860 + }, + { + "epoch": 8.35029994546446, + "grad_norm": 0.11355459690093994, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0143, + "step": 91870 + }, + { + "epoch": 8.351208871114343, + "grad_norm": 0.15172229707241058, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0187, + "step": 91880 + }, + { + "epoch": 8.352117796764224, + "grad_norm": 0.15705133974552155, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0176, + "step": 91890 + }, + { + "epoch": 8.353026722414107, + "grad_norm": 0.11566857993602753, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0153, + "step": 91900 + }, + { + "epoch": 8.353935648063988, + "grad_norm": 0.13481056690216064, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0168, + "step": 91910 + }, + { + "epoch": 8.354844573713871, + "grad_norm": 0.11939645558595657, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0167, + "step": 91920 + }, + { + "epoch": 8.355753499363752, + "grad_norm": 0.13517427444458008, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0163, + "step": 91930 + }, + { + "epoch": 8.356662425013633, + "grad_norm": 0.2000468224287033, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0153, + "step": 91940 + }, + { + "epoch": 8.357571350663516, + "grad_norm": 0.1050291433930397, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0168, + "step": 91950 + }, + { + "epoch": 8.358480276313397, + "grad_norm": 0.2240365892648697, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0163, + "step": 91960 + }, + { + "epoch": 8.35938920196328, + "grad_norm": 0.10960175096988678, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0172, + "step": 91970 + }, + { + "epoch": 8.360298127613161, + "grad_norm": 0.17765924334526062, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0213, + "step": 91980 + }, + { + "epoch": 8.361207053263042, + "grad_norm": 0.12403330951929092, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0166, + "step": 91990 + }, + { + "epoch": 8.362115978912925, + "grad_norm": 0.13441476225852966, + "learning_rate": 4.248550546359458e-06, + "loss": 0.0168, + "step": 92000 + }, + { + "epoch": 8.363024904562806, + "grad_norm": 0.13002803921699524, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0191, + "step": 92010 + }, + { + "epoch": 8.363933830212689, + "grad_norm": 0.11945252120494843, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0194, + "step": 92020 + }, + { + "epoch": 8.36484275586257, + "grad_norm": 0.14082024991512299, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0166, + "step": 92030 + }, + { + "epoch": 8.365751681512453, + "grad_norm": 0.11489105969667435, + "learning_rate": 4.237027161956122e-06, + "loss": 0.0176, + "step": 92040 + }, + { + "epoch": 8.366660607162334, + "grad_norm": 0.1822463721036911, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0171, + "step": 92050 + }, + { + "epoch": 8.367569532812215, + "grad_norm": 0.12149617075920105, + "learning_rate": 4.231270373140037e-06, + "loss": 0.0172, + "step": 92060 + }, + { + "epoch": 8.368478458462098, + "grad_norm": 0.14826031029224396, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0195, + "step": 92070 + }, + { + "epoch": 8.36938738411198, + "grad_norm": 0.13740599155426025, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.0151, + "step": 92080 + }, + { + "epoch": 8.370296309761862, + "grad_norm": 0.12499119341373444, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.0171, + "step": 92090 + }, + { + "epoch": 8.371205235411743, + "grad_norm": 0.12352289259433746, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0174, + "step": 92100 + }, + { + "epoch": 8.372114161061624, + "grad_norm": 0.120528943836689, + "learning_rate": 4.216892716666588e-06, + "loss": 0.0164, + "step": 92110 + }, + { + "epoch": 8.373023086711507, + "grad_norm": 0.1281387060880661, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0189, + "step": 92120 + }, + { + "epoch": 8.373932012361388, + "grad_norm": 0.140353262424469, + "learning_rate": 4.211147384889858e-06, + "loss": 0.015, + "step": 92130 + }, + { + "epoch": 8.374840938011271, + "grad_norm": 0.14632318913936615, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0188, + "step": 92140 + }, + { + "epoch": 8.375749863661152, + "grad_norm": 0.10941895842552185, + "learning_rate": 4.205405330481681e-06, + "loss": 0.019, + "step": 92150 + }, + { + "epoch": 8.376658789311035, + "grad_norm": 0.13972483575344086, + "learning_rate": 4.202535532835934e-06, + "loss": 0.0172, + "step": 92160 + }, + { + "epoch": 8.377567714960916, + "grad_norm": 0.14827357232570648, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0188, + "step": 92170 + }, + { + "epoch": 8.378476640610797, + "grad_norm": 0.1404334157705307, + "learning_rate": 4.196798397750821e-06, + "loss": 0.018, + "step": 92180 + }, + { + "epoch": 8.37938556626068, + "grad_norm": 0.26327088475227356, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0174, + "step": 92190 + }, + { + "epoch": 8.380294491910561, + "grad_norm": 0.11751948297023773, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0208, + "step": 92200 + }, + { + "epoch": 8.381203417560444, + "grad_norm": 0.11794888228178024, + "learning_rate": 4.188198848906393e-06, + "loss": 0.0157, + "step": 92210 + }, + { + "epoch": 8.382112343210325, + "grad_norm": 0.1339098960161209, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0173, + "step": 92220 + }, + { + "epoch": 8.383021268860206, + "grad_norm": 0.10651978850364685, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0143, + "step": 92230 + }, + { + "epoch": 8.38393019451009, + "grad_norm": 0.12990114092826843, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0165, + "step": 92240 + }, + { + "epoch": 8.38483912015997, + "grad_norm": 0.1554722636938095, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0176, + "step": 92250 + }, + { + "epoch": 8.385748045809853, + "grad_norm": 0.10410550236701965, + "learning_rate": 4.173882692097379e-06, + "loss": 0.015, + "step": 92260 + }, + { + "epoch": 8.386656971459734, + "grad_norm": 0.13498540222644806, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0146, + "step": 92270 + }, + { + "epoch": 8.387565897109617, + "grad_norm": 0.2001132369041443, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.017, + "step": 92280 + }, + { + "epoch": 8.388474822759498, + "grad_norm": 0.0924672931432724, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0159, + "step": 92290 + }, + { + "epoch": 8.38938374840938, + "grad_norm": 0.15618646144866943, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0199, + "step": 92300 + }, + { + "epoch": 8.390292674059262, + "grad_norm": 0.1328132450580597, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0164, + "step": 92310 + }, + { + "epoch": 8.391201599709143, + "grad_norm": 0.10120869427919388, + "learning_rate": 4.156730437992965e-06, + "loss": 0.016, + "step": 92320 + }, + { + "epoch": 8.392110525359026, + "grad_norm": 0.11099019646644592, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0158, + "step": 92330 + }, + { + "epoch": 8.393019451008907, + "grad_norm": 0.12299312651157379, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0167, + "step": 92340 + }, + { + "epoch": 8.393928376658788, + "grad_norm": 0.27450448274612427, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0181, + "step": 92350 + }, + { + "epoch": 8.394837302308671, + "grad_norm": 0.14710424840450287, + "learning_rate": 4.145312068523652e-06, + "loss": 0.0174, + "step": 92360 + }, + { + "epoch": 8.395746227958552, + "grad_norm": 0.24401989579200745, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0173, + "step": 92370 + }, + { + "epoch": 8.396655153608435, + "grad_norm": 0.12435044348239899, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0178, + "step": 92380 + }, + { + "epoch": 8.397564079258316, + "grad_norm": 0.1278161108493805, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0167, + "step": 92390 + }, + { + "epoch": 8.3984730049082, + "grad_norm": 0.1632792055606842, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0151, + "step": 92400 + }, + { + "epoch": 8.39938193055808, + "grad_norm": 0.12411876022815704, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0186, + "step": 92410 + }, + { + "epoch": 8.400290856207961, + "grad_norm": 0.1945306807756424, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0179, + "step": 92420 + }, + { + "epoch": 8.401199781857844, + "grad_norm": 0.15629842877388, + "learning_rate": 4.125361667450295e-06, + "loss": 0.0185, + "step": 92430 + }, + { + "epoch": 8.402108707507725, + "grad_norm": 0.2520545423030853, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0185, + "step": 92440 + }, + { + "epoch": 8.403017633157608, + "grad_norm": 0.1322925090789795, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0168, + "step": 92450 + }, + { + "epoch": 8.40392655880749, + "grad_norm": 0.1994166374206543, + "learning_rate": 4.116823879089642e-06, + "loss": 0.015, + "step": 92460 + }, + { + "epoch": 8.404835484457372, + "grad_norm": 0.15061107277870178, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0166, + "step": 92470 + }, + { + "epoch": 8.405744410107253, + "grad_norm": 0.1429567039012909, + "learning_rate": 4.11113615234934e-06, + "loss": 0.0178, + "step": 92480 + }, + { + "epoch": 8.406653335757134, + "grad_norm": 0.18608148396015167, + "learning_rate": 4.108293529277245e-06, + "loss": 0.016, + "step": 92490 + }, + { + "epoch": 8.407562261407017, + "grad_norm": 0.12807758152484894, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0156, + "step": 92500 + }, + { + "epoch": 8.408471187056898, + "grad_norm": 0.14647023379802704, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0158, + "step": 92510 + }, + { + "epoch": 8.409380112706781, + "grad_norm": 0.2094535380601883, + "learning_rate": 4.099770623843519e-06, + "loss": 0.0166, + "step": 92520 + }, + { + "epoch": 8.410289038356662, + "grad_norm": 0.09375967085361481, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0168, + "step": 92530 + }, + { + "epoch": 8.411197964006544, + "grad_norm": 0.10138744115829468, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0175, + "step": 92540 + }, + { + "epoch": 8.412106889656426, + "grad_norm": 0.12363128364086151, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0179, + "step": 92550 + }, + { + "epoch": 8.413015815306307, + "grad_norm": 0.10653761029243469, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0175, + "step": 92560 + }, + { + "epoch": 8.41392474095619, + "grad_norm": 0.12216643244028091, + "learning_rate": 4.08558234033726e-06, + "loss": 0.0174, + "step": 92570 + }, + { + "epoch": 8.414833666606071, + "grad_norm": 0.12463018298149109, + "learning_rate": 4.082747169408292e-06, + "loss": 0.0164, + "step": 92580 + }, + { + "epoch": 8.415742592255953, + "grad_norm": 0.12398459017276764, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0192, + "step": 92590 + }, + { + "epoch": 8.416651517905835, + "grad_norm": 0.1508534550666809, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0143, + "step": 92600 + }, + { + "epoch": 8.417560443555717, + "grad_norm": 0.12492948025465012, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.0174, + "step": 92610 + }, + { + "epoch": 8.4184693692056, + "grad_norm": 0.12513990700244904, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0161, + "step": 92620 + }, + { + "epoch": 8.41937829485548, + "grad_norm": 0.11442876607179642, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0161, + "step": 92630 + }, + { + "epoch": 8.420287220505363, + "grad_norm": 0.08925323188304901, + "learning_rate": 4.065753562325428e-06, + "loss": 0.0167, + "step": 92640 + }, + { + "epoch": 8.421196146155244, + "grad_norm": 0.11119996756315231, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0177, + "step": 92650 + }, + { + "epoch": 8.422105071805126, + "grad_norm": 0.12440408021211624, + "learning_rate": 4.060095667408566e-06, + "loss": 0.016, + "step": 92660 + }, + { + "epoch": 8.423013997455008, + "grad_norm": 0.11550350487232208, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0189, + "step": 92670 + }, + { + "epoch": 8.42392292310489, + "grad_norm": 0.1473119854927063, + "learning_rate": 4.054441095745782e-06, + "loss": 0.0175, + "step": 92680 + }, + { + "epoch": 8.424831848754772, + "grad_norm": 0.17853844165802002, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.0173, + "step": 92690 + }, + { + "epoch": 8.425740774404654, + "grad_norm": 0.10838153958320618, + "learning_rate": 4.048789849054798e-06, + "loss": 0.016, + "step": 92700 + }, + { + "epoch": 8.426649700054536, + "grad_norm": 0.1457313746213913, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0172, + "step": 92710 + }, + { + "epoch": 8.427558625704417, + "grad_norm": 0.14925089478492737, + "learning_rate": 4.043141929052309e-06, + "loss": 0.018, + "step": 92720 + }, + { + "epoch": 8.428467551354299, + "grad_norm": 0.192050963640213, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0179, + "step": 92730 + }, + { + "epoch": 8.429376477004181, + "grad_norm": 0.1611781269311905, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0173, + "step": 92740 + }, + { + "epoch": 8.430285402654063, + "grad_norm": 0.12292326241731644, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0161, + "step": 92750 + }, + { + "epoch": 8.431194328303945, + "grad_norm": 0.1192937046289444, + "learning_rate": 4.031856075974555e-06, + "loss": 0.016, + "step": 92760 + }, + { + "epoch": 8.432103253953827, + "grad_norm": 0.15653927624225616, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0182, + "step": 92770 + }, + { + "epoch": 8.433012179603708, + "grad_norm": 0.10996157675981522, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0148, + "step": 92780 + }, + { + "epoch": 8.43392110525359, + "grad_norm": 0.30498865246772766, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0172, + "step": 92790 + }, + { + "epoch": 8.434830030903472, + "grad_norm": 0.08565637469291687, + "learning_rate": 4.020583550225896e-06, + "loss": 0.019, + "step": 92800 + }, + { + "epoch": 8.435738956553354, + "grad_norm": 0.13675443828105927, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0187, + "step": 92810 + }, + { + "epoch": 8.436647882203236, + "grad_norm": 0.15525534749031067, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0163, + "step": 92820 + }, + { + "epoch": 8.437556807853118, + "grad_norm": 0.1171882227063179, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0158, + "step": 92830 + }, + { + "epoch": 8.438465733503, + "grad_norm": 0.1072222962975502, + "learning_rate": 4.009324365503516e-06, + "loss": 0.015, + "step": 92840 + }, + { + "epoch": 8.43937465915288, + "grad_norm": 0.146460622549057, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0174, + "step": 92850 + }, + { + "epoch": 8.440283584802764, + "grad_norm": 0.16638222336769104, + "learning_rate": 4.003699780303114e-06, + "loss": 0.0179, + "step": 92860 + }, + { + "epoch": 8.441192510452645, + "grad_norm": 0.09876806288957596, + "learning_rate": 4.000888740240793e-06, + "loss": 0.015, + "step": 92870 + }, + { + "epoch": 8.442101436102528, + "grad_norm": 0.11462154984474182, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0166, + "step": 92880 + }, + { + "epoch": 8.443010361752409, + "grad_norm": 0.1392362266778946, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0167, + "step": 92890 + }, + { + "epoch": 8.44391928740229, + "grad_norm": 0.18891297280788422, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0166, + "step": 92900 + }, + { + "epoch": 8.444828213052173, + "grad_norm": 0.15150266885757446, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0189, + "step": 92910 + }, + { + "epoch": 8.445737138702054, + "grad_norm": 0.1584743708372116, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0192, + "step": 92920 + }, + { + "epoch": 8.446646064351937, + "grad_norm": 0.12691007554531097, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0183, + "step": 92930 + }, + { + "epoch": 8.447554990001818, + "grad_norm": 0.10278772562742233, + "learning_rate": 3.981234860428963e-06, + "loss": 0.015, + "step": 92940 + }, + { + "epoch": 8.4484639156517, + "grad_norm": 0.13202564418315887, + "learning_rate": 3.978430508818044e-06, + "loss": 0.018, + "step": 92950 + }, + { + "epoch": 8.449372841301582, + "grad_norm": 0.0950646623969078, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0171, + "step": 92960 + }, + { + "epoch": 8.450281766951463, + "grad_norm": 0.0973958671092987, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0149, + "step": 92970 + }, + { + "epoch": 8.451190692601346, + "grad_norm": 0.1669062227010727, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0161, + "step": 92980 + }, + { + "epoch": 8.452099618251227, + "grad_norm": 0.16957035660743713, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0169, + "step": 92990 + }, + { + "epoch": 8.45300854390111, + "grad_norm": 0.1356400102376938, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0178, + "step": 93000 + }, + { + "epoch": 8.45391746955099, + "grad_norm": 0.1760982722043991, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0163, + "step": 93010 + }, + { + "epoch": 8.454826395200872, + "grad_norm": 0.12555915117263794, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0166, + "step": 93020 + }, + { + "epoch": 8.455735320850755, + "grad_norm": 0.13776378333568573, + "learning_rate": 3.95602584636382e-06, + "loss": 0.015, + "step": 93030 + }, + { + "epoch": 8.456644246500636, + "grad_norm": 0.19024114310741425, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0162, + "step": 93040 + }, + { + "epoch": 8.457553172150519, + "grad_norm": 0.1458582878112793, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0159, + "step": 93050 + }, + { + "epoch": 8.4584620978004, + "grad_norm": 0.15888652205467224, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0171, + "step": 93060 + }, + { + "epoch": 8.459371023450283, + "grad_norm": 0.12202146649360657, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0164, + "step": 93070 + }, + { + "epoch": 8.460279949100164, + "grad_norm": 0.1330171525478363, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0184, + "step": 93080 + }, + { + "epoch": 8.461188874750045, + "grad_norm": 0.12338284403085709, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0174, + "step": 93090 + }, + { + "epoch": 8.462097800399928, + "grad_norm": 0.13952846825122833, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0155, + "step": 93100 + }, + { + "epoch": 8.463006726049809, + "grad_norm": 0.14659687876701355, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0175, + "step": 93110 + }, + { + "epoch": 8.463915651699692, + "grad_norm": 0.20078054070472717, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0183, + "step": 93120 + }, + { + "epoch": 8.464824577349573, + "grad_norm": 0.11685682833194733, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0158, + "step": 93130 + }, + { + "epoch": 8.465733502999454, + "grad_norm": 0.209571972489357, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0148, + "step": 93140 + }, + { + "epoch": 8.466642428649337, + "grad_norm": 0.13519275188446045, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0163, + "step": 93150 + }, + { + "epoch": 8.467551354299218, + "grad_norm": 0.1367112249135971, + "learning_rate": 3.919732802975102e-06, + "loss": 0.0168, + "step": 93160 + }, + { + "epoch": 8.4684602799491, + "grad_norm": 0.13337358832359314, + "learning_rate": 3.916946914788272e-06, + "loss": 0.0186, + "step": 93170 + }, + { + "epoch": 8.469369205598982, + "grad_norm": 0.13519172370433807, + "learning_rate": 3.914161868286177e-06, + "loss": 0.0165, + "step": 93180 + }, + { + "epoch": 8.470278131248865, + "grad_norm": 0.1500076949596405, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0158, + "step": 93190 + }, + { + "epoch": 8.471187056898746, + "grad_norm": 0.14609608054161072, + "learning_rate": 3.908594301182172e-06, + "loss": 0.0153, + "step": 93200 + }, + { + "epoch": 8.472095982548627, + "grad_norm": 0.10335072129964828, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0158, + "step": 93210 + }, + { + "epoch": 8.47300490819851, + "grad_norm": 0.14444321393966675, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0144, + "step": 93220 + }, + { + "epoch": 8.47391383384839, + "grad_norm": 0.13064302504062653, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0163, + "step": 93230 + }, + { + "epoch": 8.474822759498274, + "grad_norm": 0.1297224462032318, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0168, + "step": 93240 + }, + { + "epoch": 8.475731685148155, + "grad_norm": 0.14502041041851044, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0168, + "step": 93250 + }, + { + "epoch": 8.476640610798036, + "grad_norm": 0.11411932110786438, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0164, + "step": 93260 + }, + { + "epoch": 8.477549536447919, + "grad_norm": 0.1596653014421463, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0146, + "step": 93270 + }, + { + "epoch": 8.4784584620978, + "grad_norm": 0.08680915832519531, + "learning_rate": 3.88635774242559e-06, + "loss": 0.0192, + "step": 93280 + }, + { + "epoch": 8.479367387747683, + "grad_norm": 0.13215214014053345, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0163, + "step": 93290 + }, + { + "epoch": 8.480276313397564, + "grad_norm": 0.09768591076135635, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0149, + "step": 93300 + }, + { + "epoch": 8.481185239047447, + "grad_norm": 0.1540188044309616, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0171, + "step": 93310 + }, + { + "epoch": 8.482094164697328, + "grad_norm": 0.10672154277563095, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0203, + "step": 93320 + }, + { + "epoch": 8.483003090347209, + "grad_norm": 0.14534281194210052, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0179, + "step": 93330 + }, + { + "epoch": 8.483912015997092, + "grad_norm": 0.16496168076992035, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0168, + "step": 93340 + }, + { + "epoch": 8.484820941646973, + "grad_norm": 0.1347239464521408, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0158, + "step": 93350 + }, + { + "epoch": 8.485729867296856, + "grad_norm": 0.10787416249513626, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0207, + "step": 93360 + }, + { + "epoch": 8.486638792946737, + "grad_norm": 0.08567582815885544, + "learning_rate": 3.861406185766894e-06, + "loss": 0.016, + "step": 93370 + }, + { + "epoch": 8.487547718596618, + "grad_norm": 0.14290070533752441, + "learning_rate": 3.858638017290704e-06, + "loss": 0.0161, + "step": 93380 + }, + { + "epoch": 8.4884566442465, + "grad_norm": 0.20339441299438477, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0165, + "step": 93390 + }, + { + "epoch": 8.489365569896382, + "grad_norm": 0.12452522665262222, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0167, + "step": 93400 + }, + { + "epoch": 8.490274495546265, + "grad_norm": 0.1606743186712265, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0172, + "step": 93410 + }, + { + "epoch": 8.491183421196146, + "grad_norm": 0.19377531111240387, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0177, + "step": 93420 + }, + { + "epoch": 8.492092346846029, + "grad_norm": 0.13714270293712616, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0189, + "step": 93430 + }, + { + "epoch": 8.49300127249591, + "grad_norm": 0.13105879724025726, + "learning_rate": 3.842046782158189e-06, + "loss": 0.0176, + "step": 93440 + }, + { + "epoch": 8.493910198145791, + "grad_norm": 0.1564486026763916, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0188, + "step": 93450 + }, + { + "epoch": 8.494819123795674, + "grad_norm": 0.11519697308540344, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0159, + "step": 93460 + }, + { + "epoch": 8.495728049445555, + "grad_norm": 0.1454581916332245, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0173, + "step": 93470 + }, + { + "epoch": 8.496636975095438, + "grad_norm": 0.1668495386838913, + "learning_rate": 3.831002903394414e-06, + "loss": 0.017, + "step": 93480 + }, + { + "epoch": 8.497545900745319, + "grad_norm": 0.14300832152366638, + "learning_rate": 3.828244053708484e-06, + "loss": 0.0167, + "step": 93490 + }, + { + "epoch": 8.4984548263952, + "grad_norm": 0.14484398066997528, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0151, + "step": 93500 + }, + { + "epoch": 8.499363752045083, + "grad_norm": 0.11601753532886505, + "learning_rate": 3.822728899809501e-06, + "loss": 0.0183, + "step": 93510 + }, + { + "epoch": 8.500272677694964, + "grad_norm": 0.1055472269654274, + "learning_rate": 3.819972596015276e-06, + "loss": 0.015, + "step": 93520 + }, + { + "epoch": 8.501181603344847, + "grad_norm": 0.14512179791927338, + "learning_rate": 3.817217141270351e-06, + "loss": 0.0145, + "step": 93530 + }, + { + "epoch": 8.502090528994728, + "grad_norm": 0.09479429572820663, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0166, + "step": 93540 + }, + { + "epoch": 8.50299945464461, + "grad_norm": 0.134456068277359, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0182, + "step": 93550 + }, + { + "epoch": 8.503908380294492, + "grad_norm": 0.13373787701129913, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0176, + "step": 93560 + }, + { + "epoch": 8.504817305944373, + "grad_norm": 0.15424147248268127, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0179, + "step": 93570 + }, + { + "epoch": 8.505726231594256, + "grad_norm": 0.14062239229679108, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0163, + "step": 93580 + }, + { + "epoch": 8.506635157244137, + "grad_norm": 0.11532151699066162, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0172, + "step": 93590 + }, + { + "epoch": 8.50754408289402, + "grad_norm": 0.18750540912151337, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.016, + "step": 93600 + }, + { + "epoch": 8.508453008543901, + "grad_norm": 0.09889768064022064, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0136, + "step": 93610 + }, + { + "epoch": 8.509361934193784, + "grad_norm": 0.16066184639930725, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.0175, + "step": 93620 + }, + { + "epoch": 8.510270859843665, + "grad_norm": 0.15706460177898407, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0156, + "step": 93630 + }, + { + "epoch": 8.511179785493546, + "grad_norm": 0.12536625564098358, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.017, + "step": 93640 + }, + { + "epoch": 8.512088711143429, + "grad_norm": 0.15323100984096527, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0153, + "step": 93650 + }, + { + "epoch": 8.51299763679331, + "grad_norm": 0.11472265422344208, + "learning_rate": 3.781473588197192e-06, + "loss": 0.0151, + "step": 93660 + }, + { + "epoch": 8.513906562443193, + "grad_norm": 0.1568039506673813, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0167, + "step": 93670 + }, + { + "epoch": 8.514815488093074, + "grad_norm": 0.0950729101896286, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0151, + "step": 93680 + }, + { + "epoch": 8.515724413742955, + "grad_norm": 0.1849900335073471, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0194, + "step": 93690 + }, + { + "epoch": 8.516633339392838, + "grad_norm": 0.11267752200365067, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0173, + "step": 93700 + }, + { + "epoch": 8.517542265042719, + "grad_norm": 0.14316076040267944, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0169, + "step": 93710 + }, + { + "epoch": 8.518451190692602, + "grad_norm": 0.09519457817077637, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0174, + "step": 93720 + }, + { + "epoch": 8.519360116342483, + "grad_norm": 0.09757114201784134, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0169, + "step": 93730 + }, + { + "epoch": 8.520269041992364, + "grad_norm": 0.1294654905796051, + "learning_rate": 3.759549092040512e-06, + "loss": 0.0169, + "step": 93740 + }, + { + "epoch": 8.521177967642247, + "grad_norm": 0.10607796907424927, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0177, + "step": 93750 + }, + { + "epoch": 8.522086893292128, + "grad_norm": 0.15554311871528625, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0175, + "step": 93760 + }, + { + "epoch": 8.522995818942011, + "grad_norm": 0.12415748834609985, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.0161, + "step": 93770 + }, + { + "epoch": 8.523904744591892, + "grad_norm": 0.17992688715457916, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0163, + "step": 93780 + }, + { + "epoch": 8.524813670241775, + "grad_norm": 0.12585999071598053, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.017, + "step": 93790 + }, + { + "epoch": 8.525722595891656, + "grad_norm": 0.11834968626499176, + "learning_rate": 3.74314156688988e-06, + "loss": 0.0147, + "step": 93800 + }, + { + "epoch": 8.526631521541537, + "grad_norm": 0.10613841563463211, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.0148, + "step": 93810 + }, + { + "epoch": 8.52754044719142, + "grad_norm": 0.1271161437034607, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0161, + "step": 93820 + }, + { + "epoch": 8.528449372841301, + "grad_norm": 0.13236504793167114, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.0188, + "step": 93830 + }, + { + "epoch": 8.529358298491184, + "grad_norm": 0.1673162281513214, + "learning_rate": 3.732220311702337e-06, + "loss": 0.0159, + "step": 93840 + }, + { + "epoch": 8.530267224141065, + "grad_norm": 0.09672768414020538, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0165, + "step": 93850 + }, + { + "epoch": 8.531176149790948, + "grad_norm": 0.10666189342737198, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0155, + "step": 93860 + }, + { + "epoch": 8.532085075440829, + "grad_norm": 0.13124829530715942, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0148, + "step": 93870 + }, + { + "epoch": 8.53299400109071, + "grad_norm": 0.11451859027147293, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0161, + "step": 93880 + }, + { + "epoch": 8.533902926740593, + "grad_norm": 0.12012698501348495, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0167, + "step": 93890 + }, + { + "epoch": 8.534811852390474, + "grad_norm": 0.12485793977975845, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0189, + "step": 93900 + }, + { + "epoch": 8.535720778040357, + "grad_norm": 0.1342533528804779, + "learning_rate": 3.713141068064569e-06, + "loss": 0.0189, + "step": 93910 + }, + { + "epoch": 8.536629703690238, + "grad_norm": 0.15044686198234558, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0169, + "step": 93920 + }, + { + "epoch": 8.53753862934012, + "grad_norm": 0.13628539443016052, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0166, + "step": 93930 + }, + { + "epoch": 8.538447554990002, + "grad_norm": 0.14362971484661102, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.0184, + "step": 93940 + }, + { + "epoch": 8.539356480639883, + "grad_norm": 0.12440098822116852, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0169, + "step": 93950 + }, + { + "epoch": 8.540265406289766, + "grad_norm": 0.2073969542980194, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.0153, + "step": 93960 + }, + { + "epoch": 8.541174331939647, + "grad_norm": 0.09335912019014359, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0155, + "step": 93970 + }, + { + "epoch": 8.542083257589528, + "grad_norm": 0.12416976690292358, + "learning_rate": 3.694103825374046e-06, + "loss": 0.016, + "step": 93980 + }, + { + "epoch": 8.542992183239411, + "grad_norm": 0.13658885657787323, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0173, + "step": 93990 + }, + { + "epoch": 8.543901108889292, + "grad_norm": 0.11859430372714996, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0161, + "step": 94000 + }, + { + "epoch": 8.544810034539175, + "grad_norm": 0.206221342086792, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0158, + "step": 94010 + }, + { + "epoch": 8.545718960189056, + "grad_norm": 0.162737175822258, + "learning_rate": 3.683244285816826e-06, + "loss": 0.0162, + "step": 94020 + }, + { + "epoch": 8.546627885838939, + "grad_norm": 0.17623217403888702, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.0147, + "step": 94030 + }, + { + "epoch": 8.54753681148882, + "grad_norm": 0.09875698387622833, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0151, + "step": 94040 + }, + { + "epoch": 8.548445737138701, + "grad_norm": 0.1285347044467926, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0206, + "step": 94050 + }, + { + "epoch": 8.549354662788584, + "grad_norm": 0.12319866567850113, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0186, + "step": 94060 + }, + { + "epoch": 8.550263588438465, + "grad_norm": 0.17441469430923462, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0171, + "step": 94070 + }, + { + "epoch": 8.551172514088348, + "grad_norm": 0.2884029150009155, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0173, + "step": 94080 + }, + { + "epoch": 8.55208143973823, + "grad_norm": 0.13081729412078857, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0173, + "step": 94090 + }, + { + "epoch": 8.552990365388112, + "grad_norm": 0.11741422861814499, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0185, + "step": 94100 + }, + { + "epoch": 8.553899291037993, + "grad_norm": 0.14766336977481842, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0146, + "step": 94110 + }, + { + "epoch": 8.554808216687874, + "grad_norm": 0.10176141560077667, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0147, + "step": 94120 + }, + { + "epoch": 8.555717142337757, + "grad_norm": 0.12701047956943512, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0162, + "step": 94130 + }, + { + "epoch": 8.556626067987638, + "grad_norm": 0.11491619795560837, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0158, + "step": 94140 + }, + { + "epoch": 8.557534993637521, + "grad_norm": 0.12220693379640579, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0186, + "step": 94150 + }, + { + "epoch": 8.558443919287402, + "grad_norm": 0.10093224048614502, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.0162, + "step": 94160 + }, + { + "epoch": 8.559352844937283, + "grad_norm": 0.11895972490310669, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.015, + "step": 94170 + }, + { + "epoch": 8.560261770587166, + "grad_norm": 0.10534460842609406, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0182, + "step": 94180 + }, + { + "epoch": 8.561170696237047, + "grad_norm": 0.1824204921722412, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.017, + "step": 94190 + }, + { + "epoch": 8.56207962188693, + "grad_norm": 0.14764632284641266, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0166, + "step": 94200 + }, + { + "epoch": 8.562988547536811, + "grad_norm": 0.14313170313835144, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0172, + "step": 94210 + }, + { + "epoch": 8.563897473186694, + "grad_norm": 0.08593371510505676, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0144, + "step": 94220 + }, + { + "epoch": 8.564806398836575, + "grad_norm": 0.1154162734746933, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.0164, + "step": 94230 + }, + { + "epoch": 8.565715324486456, + "grad_norm": 0.11071813851594925, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0171, + "step": 94240 + }, + { + "epoch": 8.56662425013634, + "grad_norm": 0.15662303566932678, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0194, + "step": 94250 + }, + { + "epoch": 8.56753317578622, + "grad_norm": 0.11335616558790207, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0162, + "step": 94260 + }, + { + "epoch": 8.568442101436103, + "grad_norm": 0.11360324919223785, + "learning_rate": 3.615684229793514e-06, + "loss": 0.0181, + "step": 94270 + }, + { + "epoch": 8.569351027085984, + "grad_norm": 0.13422967493534088, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0169, + "step": 94280 + }, + { + "epoch": 8.570259952735865, + "grad_norm": 0.11749468743801117, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.0173, + "step": 94290 + }, + { + "epoch": 8.571168878385748, + "grad_norm": 0.13579072058200836, + "learning_rate": 3.607613271657616e-06, + "loss": 0.0175, + "step": 94300 + }, + { + "epoch": 8.57207780403563, + "grad_norm": 0.15625441074371338, + "learning_rate": 3.60492468235981e-06, + "loss": 0.016, + "step": 94310 + }, + { + "epoch": 8.572986729685512, + "grad_norm": 0.11681080609560013, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.0185, + "step": 94320 + }, + { + "epoch": 8.573895655335393, + "grad_norm": 0.11393929272890091, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.017, + "step": 94330 + }, + { + "epoch": 8.574804580985276, + "grad_norm": 0.09335270524024963, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0162, + "step": 94340 + }, + { + "epoch": 8.575713506635157, + "grad_norm": 0.1458883136510849, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0165, + "step": 94350 + }, + { + "epoch": 8.576622432285038, + "grad_norm": 0.13377699255943298, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0164, + "step": 94360 + }, + { + "epoch": 8.577531357934921, + "grad_norm": 0.128787100315094, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0166, + "step": 94370 + }, + { + "epoch": 8.578440283584802, + "grad_norm": 0.16164955496788025, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0179, + "step": 94380 + }, + { + "epoch": 8.579349209234685, + "grad_norm": 0.11583004891872406, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.017, + "step": 94390 + }, + { + "epoch": 8.580258134884566, + "grad_norm": 0.0999496728181839, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0174, + "step": 94400 + }, + { + "epoch": 8.581167060534447, + "grad_norm": 0.15650653839111328, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.015, + "step": 94410 + }, + { + "epoch": 8.58207598618433, + "grad_norm": 0.12949688732624054, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0183, + "step": 94420 + }, + { + "epoch": 8.582984911834211, + "grad_norm": 0.1099485456943512, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0161, + "step": 94430 + }, + { + "epoch": 8.583893837484094, + "grad_norm": 0.1316312849521637, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.017, + "step": 94440 + }, + { + "epoch": 8.584802763133975, + "grad_norm": 0.17374171316623688, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0193, + "step": 94450 + }, + { + "epoch": 8.585711688783858, + "grad_norm": 0.11755461245775223, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.0152, + "step": 94460 + }, + { + "epoch": 8.58662061443374, + "grad_norm": 0.1159849613904953, + "learning_rate": 3.562025084021153e-06, + "loss": 0.0171, + "step": 94470 + }, + { + "epoch": 8.58752954008362, + "grad_norm": 0.11671266704797745, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0181, + "step": 94480 + }, + { + "epoch": 8.588438465733503, + "grad_norm": 0.16863900423049927, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0193, + "step": 94490 + }, + { + "epoch": 8.589347391383384, + "grad_norm": 0.22849741578102112, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0177, + "step": 94500 + }, + { + "epoch": 8.590256317033267, + "grad_norm": 0.12120340764522552, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0161, + "step": 94510 + }, + { + "epoch": 8.591165242683148, + "grad_norm": 0.14553585648536682, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.016, + "step": 94520 + }, + { + "epoch": 8.592074168333031, + "grad_norm": 0.12810906767845154, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0166, + "step": 94530 + }, + { + "epoch": 8.592983093982912, + "grad_norm": 0.12547418475151062, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0178, + "step": 94540 + }, + { + "epoch": 8.593892019632793, + "grad_norm": 0.10590460896492004, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0156, + "step": 94550 + }, + { + "epoch": 8.594800945282676, + "grad_norm": 0.1908303052186966, + "learning_rate": 3.53799172837971e-06, + "loss": 0.017, + "step": 94560 + }, + { + "epoch": 8.595709870932557, + "grad_norm": 0.11911322921514511, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0171, + "step": 94570 + }, + { + "epoch": 8.59661879658244, + "grad_norm": 0.12305350601673126, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0163, + "step": 94580 + }, + { + "epoch": 8.597527722232321, + "grad_norm": 0.17082899808883667, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0164, + "step": 94590 + }, + { + "epoch": 8.598436647882203, + "grad_norm": 0.18755914270877838, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0185, + "step": 94600 + }, + { + "epoch": 8.599345573532085, + "grad_norm": 0.10682795196771622, + "learning_rate": 3.524670320908895e-06, + "loss": 0.0152, + "step": 94610 + }, + { + "epoch": 8.600254499181966, + "grad_norm": 0.12740455567836761, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.017, + "step": 94620 + }, + { + "epoch": 8.60116342483185, + "grad_norm": 0.1425522118806839, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0171, + "step": 94630 + }, + { + "epoch": 8.60207235048173, + "grad_norm": 0.11399439722299576, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0153, + "step": 94640 + }, + { + "epoch": 8.602981276131612, + "grad_norm": 0.18157954514026642, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.016, + "step": 94650 + }, + { + "epoch": 8.603890201781494, + "grad_norm": 0.13515900075435638, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.0164, + "step": 94660 + }, + { + "epoch": 8.604799127431376, + "grad_norm": 0.12005835026502609, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.0162, + "step": 94670 + }, + { + "epoch": 8.605708053081258, + "grad_norm": 0.1254250705242157, + "learning_rate": 3.506056958075152e-06, + "loss": 0.0164, + "step": 94680 + }, + { + "epoch": 8.60661697873114, + "grad_norm": 0.16753797233104706, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.018, + "step": 94690 + }, + { + "epoch": 8.607525904381022, + "grad_norm": 0.13286881148815155, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0159, + "step": 94700 + }, + { + "epoch": 8.608434830030903, + "grad_norm": 0.15480801463127136, + "learning_rate": 3.49809289187469e-06, + "loss": 0.0155, + "step": 94710 + }, + { + "epoch": 8.609343755680785, + "grad_norm": 0.16933202743530273, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.0163, + "step": 94720 + }, + { + "epoch": 8.610252681330667, + "grad_norm": 0.13193607330322266, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0164, + "step": 94730 + }, + { + "epoch": 8.611161606980549, + "grad_norm": 0.12568892538547516, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0163, + "step": 94740 + }, + { + "epoch": 8.612070532630431, + "grad_norm": 0.11776060611009598, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0165, + "step": 94750 + }, + { + "epoch": 8.612979458280313, + "grad_norm": 0.11748785525560379, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0166, + "step": 94760 + }, + { + "epoch": 8.613888383930195, + "grad_norm": 0.15930213034152985, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0173, + "step": 94770 + }, + { + "epoch": 8.614797309580077, + "grad_norm": 0.13722805678844452, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.0176, + "step": 94780 + }, + { + "epoch": 8.615706235229958, + "grad_norm": 0.12888120114803314, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0171, + "step": 94790 + }, + { + "epoch": 8.61661516087984, + "grad_norm": 0.15752054750919342, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.0164, + "step": 94800 + }, + { + "epoch": 8.617524086529722, + "grad_norm": 0.17290052771568298, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0164, + "step": 94810 + }, + { + "epoch": 8.618433012179604, + "grad_norm": 0.14463335275650024, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0151, + "step": 94820 + }, + { + "epoch": 8.619341937829486, + "grad_norm": 0.1360713094472885, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0151, + "step": 94830 + }, + { + "epoch": 8.620250863479367, + "grad_norm": 0.10974536836147308, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.0156, + "step": 94840 + }, + { + "epoch": 8.62115978912925, + "grad_norm": 0.14339230954647064, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.0148, + "step": 94850 + }, + { + "epoch": 8.62206871477913, + "grad_norm": 0.27014490962028503, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0146, + "step": 94860 + }, + { + "epoch": 8.622977640429013, + "grad_norm": 0.2232498675584793, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0174, + "step": 94870 + }, + { + "epoch": 8.623886566078895, + "grad_norm": 0.23107688128948212, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0167, + "step": 94880 + }, + { + "epoch": 8.624795491728776, + "grad_norm": 0.14428015053272247, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0171, + "step": 94890 + }, + { + "epoch": 8.625704417378659, + "grad_norm": 0.14124202728271484, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.0164, + "step": 94900 + }, + { + "epoch": 8.62661334302854, + "grad_norm": 0.1424626111984253, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0201, + "step": 94910 + }, + { + "epoch": 8.627522268678423, + "grad_norm": 0.13860632479190826, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.015, + "step": 94920 + }, + { + "epoch": 8.628431194328304, + "grad_norm": 0.1617194265127182, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0183, + "step": 94930 + }, + { + "epoch": 8.629340119978187, + "grad_norm": 0.11077827215194702, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0166, + "step": 94940 + }, + { + "epoch": 8.630249045628068, + "grad_norm": 0.15892431139945984, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0157, + "step": 94950 + }, + { + "epoch": 8.631157971277949, + "grad_norm": 0.20668919384479523, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0154, + "step": 94960 + }, + { + "epoch": 8.632066896927832, + "grad_norm": 0.12088894098997116, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0178, + "step": 94970 + }, + { + "epoch": 8.632975822577713, + "grad_norm": 0.14379288256168365, + "learning_rate": 3.426770712526696e-06, + "loss": 0.0171, + "step": 94980 + }, + { + "epoch": 8.633884748227596, + "grad_norm": 0.25844746828079224, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0176, + "step": 94990 + }, + { + "epoch": 8.634793673877477, + "grad_norm": 0.19391223788261414, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0174, + "step": 95000 + }, + { + "epoch": 8.63570259952736, + "grad_norm": 0.18115395307540894, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0164, + "step": 95010 + }, + { + "epoch": 8.63661152517724, + "grad_norm": 0.12263385206460953, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0146, + "step": 95020 + }, + { + "epoch": 8.637520450827122, + "grad_norm": 0.10504930466413498, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0172, + "step": 95030 + }, + { + "epoch": 8.638429376477005, + "grad_norm": 0.1483638733625412, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0178, + "step": 95040 + }, + { + "epoch": 8.639338302126886, + "grad_norm": 0.1318831592798233, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0167, + "step": 95050 + }, + { + "epoch": 8.640247227776769, + "grad_norm": 0.12577320635318756, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0141, + "step": 95060 + }, + { + "epoch": 8.64115615342665, + "grad_norm": 0.2646089494228363, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0161, + "step": 95070 + }, + { + "epoch": 8.64206507907653, + "grad_norm": 0.10147023946046829, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0186, + "step": 95080 + }, + { + "epoch": 8.642974004726414, + "grad_norm": 0.13219532370567322, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0167, + "step": 95090 + }, + { + "epoch": 8.643882930376295, + "grad_norm": 0.15632258355617523, + "learning_rate": 3.395277396807372e-06, + "loss": 0.0188, + "step": 95100 + }, + { + "epoch": 8.644791856026178, + "grad_norm": 0.15704867243766785, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0165, + "step": 95110 + }, + { + "epoch": 8.645700781676059, + "grad_norm": 0.18304739892482758, + "learning_rate": 3.39004083979155e-06, + "loss": 0.0165, + "step": 95120 + }, + { + "epoch": 8.64660970732594, + "grad_norm": 0.16711987555027008, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0173, + "step": 95130 + }, + { + "epoch": 8.647518632975823, + "grad_norm": 0.19437254965305328, + "learning_rate": 3.384807809574769e-06, + "loss": 0.0157, + "step": 95140 + }, + { + "epoch": 8.648427558625704, + "grad_norm": 0.13868612051010132, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0173, + "step": 95150 + }, + { + "epoch": 8.649336484275587, + "grad_norm": 0.12856148183345795, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0189, + "step": 95160 + }, + { + "epoch": 8.650245409925468, + "grad_norm": 0.10153938084840775, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0162, + "step": 95170 + }, + { + "epoch": 8.65115433557535, + "grad_norm": 0.12806443870067596, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0158, + "step": 95180 + }, + { + "epoch": 8.652063261225232, + "grad_norm": 0.10511744767427444, + "learning_rate": 3.371740674208153e-06, + "loss": 0.016, + "step": 95190 + }, + { + "epoch": 8.652972186875113, + "grad_norm": 0.237268328666687, + "learning_rate": 3.36912989560994e-06, + "loss": 0.0185, + "step": 95200 + }, + { + "epoch": 8.653881112524996, + "grad_norm": 0.12173093110322952, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0183, + "step": 95210 + }, + { + "epoch": 8.654790038174877, + "grad_norm": 0.13030187785625458, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0162, + "step": 95220 + }, + { + "epoch": 8.65569896382476, + "grad_norm": 0.11541852355003357, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0174, + "step": 95230 + }, + { + "epoch": 8.65660788947464, + "grad_norm": 0.09606580436229706, + "learning_rate": 3.358695616077078e-06, + "loss": 0.0161, + "step": 95240 + }, + { + "epoch": 8.657516815124524, + "grad_norm": 0.12026909738779068, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0177, + "step": 95250 + }, + { + "epoch": 8.658425740774405, + "grad_norm": 0.14811144769191742, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0146, + "step": 95260 + }, + { + "epoch": 8.659334666424286, + "grad_norm": 0.1776551902294159, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0174, + "step": 95270 + }, + { + "epoch": 8.660243592074169, + "grad_norm": 0.17353858053684235, + "learning_rate": 3.348275481826746e-06, + "loss": 0.0177, + "step": 95280 + }, + { + "epoch": 8.66115251772405, + "grad_norm": 0.09601887315511703, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0156, + "step": 95290 + }, + { + "epoch": 8.662061443373933, + "grad_norm": 0.1033150851726532, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.018, + "step": 95300 + }, + { + "epoch": 8.662970369023814, + "grad_norm": 0.11162235587835312, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0155, + "step": 95310 + }, + { + "epoch": 8.663879294673695, + "grad_norm": 0.17102769017219543, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0168, + "step": 95320 + }, + { + "epoch": 8.664788220323578, + "grad_norm": 0.12551026046276093, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0148, + "step": 95330 + }, + { + "epoch": 8.665697145973459, + "grad_norm": 0.1902606338262558, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0175, + "step": 95340 + }, + { + "epoch": 8.666606071623342, + "grad_norm": 0.10697111487388611, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.018, + "step": 95350 + }, + { + "epoch": 8.667514997273223, + "grad_norm": 0.10946494340896606, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0143, + "step": 95360 + }, + { + "epoch": 8.668423922923106, + "grad_norm": 0.14258235692977905, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.0161, + "step": 95370 + }, + { + "epoch": 8.669332848572987, + "grad_norm": 0.1266772598028183, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0175, + "step": 95380 + }, + { + "epoch": 8.670241774222868, + "grad_norm": 0.151971697807312, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0169, + "step": 95390 + }, + { + "epoch": 8.67115069987275, + "grad_norm": 0.10457606613636017, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0144, + "step": 95400 + }, + { + "epoch": 8.672059625522632, + "grad_norm": 0.1227976456284523, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0171, + "step": 95410 + }, + { + "epoch": 8.672968551172515, + "grad_norm": 0.10328377783298492, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0155, + "step": 95420 + }, + { + "epoch": 8.673877476822396, + "grad_norm": 0.09980896860361099, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0139, + "step": 95430 + }, + { + "epoch": 8.674786402472277, + "grad_norm": 0.09207793325185776, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0158, + "step": 95440 + }, + { + "epoch": 8.67569532812216, + "grad_norm": 0.12353163212537766, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0173, + "step": 95450 + }, + { + "epoch": 8.676604253772041, + "grad_norm": 0.12814171612262726, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.017, + "step": 95460 + }, + { + "epoch": 8.677513179421924, + "grad_norm": 0.1517598032951355, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.015, + "step": 95470 + }, + { + "epoch": 8.678422105071805, + "grad_norm": 0.10917135328054428, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0174, + "step": 95480 + }, + { + "epoch": 8.679331030721688, + "grad_norm": 0.18033742904663086, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0176, + "step": 95490 + }, + { + "epoch": 8.680239956371569, + "grad_norm": 0.11357691884040833, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0137, + "step": 95500 + }, + { + "epoch": 8.68114888202145, + "grad_norm": 0.1746942549943924, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0184, + "step": 95510 + }, + { + "epoch": 8.682057807671333, + "grad_norm": 0.10202144831418991, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0154, + "step": 95520 + }, + { + "epoch": 8.682966733321214, + "grad_norm": 0.16160525381565094, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0194, + "step": 95530 + }, + { + "epoch": 8.683875658971097, + "grad_norm": 0.11324625462293625, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0177, + "step": 95540 + }, + { + "epoch": 8.684784584620978, + "grad_norm": 0.15678557753562927, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0164, + "step": 95550 + }, + { + "epoch": 8.685693510270859, + "grad_norm": 0.18678127229213715, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0149, + "step": 95560 + }, + { + "epoch": 8.686602435920742, + "grad_norm": 0.12024681270122528, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0149, + "step": 95570 + }, + { + "epoch": 8.687511361570623, + "grad_norm": 0.16543078422546387, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0163, + "step": 95580 + }, + { + "epoch": 8.688420287220506, + "grad_norm": 0.21325869858264923, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.017, + "step": 95590 + }, + { + "epoch": 8.689329212870387, + "grad_norm": 0.093158058822155, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0151, + "step": 95600 + }, + { + "epoch": 8.69023813852027, + "grad_norm": 0.11089558899402618, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.017, + "step": 95610 + }, + { + "epoch": 8.691147064170151, + "grad_norm": 0.14263780415058136, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0183, + "step": 95620 + }, + { + "epoch": 8.692055989820032, + "grad_norm": 0.11590591818094254, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0149, + "step": 95630 + }, + { + "epoch": 8.692964915469915, + "grad_norm": 0.10562557727098465, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0169, + "step": 95640 + }, + { + "epoch": 8.693873841119796, + "grad_norm": 0.13804931938648224, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0176, + "step": 95650 + }, + { + "epoch": 8.694782766769679, + "grad_norm": 0.130571186542511, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0167, + "step": 95660 + }, + { + "epoch": 8.69569169241956, + "grad_norm": 0.1759788691997528, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0177, + "step": 95670 + }, + { + "epoch": 8.696600618069443, + "grad_norm": 0.10672000795602798, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0167, + "step": 95680 + }, + { + "epoch": 8.697509543719324, + "grad_norm": 0.18545593321323395, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0172, + "step": 95690 + }, + { + "epoch": 8.698418469369205, + "grad_norm": 0.19054657220840454, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.0168, + "step": 95700 + }, + { + "epoch": 8.699327395019088, + "grad_norm": 0.1351829171180725, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0144, + "step": 95710 + }, + { + "epoch": 8.700236320668969, + "grad_norm": 0.1592693328857422, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0184, + "step": 95720 + }, + { + "epoch": 8.701145246318852, + "grad_norm": 0.10842593014240265, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0164, + "step": 95730 + }, + { + "epoch": 8.702054171968733, + "grad_norm": 0.155960813164711, + "learning_rate": 3.229464710685389e-06, + "loss": 0.016, + "step": 95740 + }, + { + "epoch": 8.702963097618614, + "grad_norm": 0.1251726895570755, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.0169, + "step": 95750 + }, + { + "epoch": 8.703872023268497, + "grad_norm": 0.09370702505111694, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.015, + "step": 95760 + }, + { + "epoch": 8.704780948918378, + "grad_norm": 0.31253230571746826, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0156, + "step": 95770 + }, + { + "epoch": 8.705689874568261, + "grad_norm": 0.17737233638763428, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.0158, + "step": 95780 + }, + { + "epoch": 8.706598800218142, + "grad_norm": 0.12172812223434448, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.0173, + "step": 95790 + }, + { + "epoch": 8.707507725868023, + "grad_norm": 0.0840122327208519, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0161, + "step": 95800 + }, + { + "epoch": 8.708416651517906, + "grad_norm": 0.18188560009002686, + "learning_rate": 3.211550156962556e-06, + "loss": 0.0164, + "step": 95810 + }, + { + "epoch": 8.709325577167787, + "grad_norm": 0.1454169601202011, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0141, + "step": 95820 + }, + { + "epoch": 8.71023450281767, + "grad_norm": 0.1527523249387741, + "learning_rate": 3.206439767422759e-06, + "loss": 0.0177, + "step": 95830 + }, + { + "epoch": 8.711143428467551, + "grad_norm": 0.17927908897399902, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0183, + "step": 95840 + }, + { + "epoch": 8.712052354117434, + "grad_norm": 0.12469925731420517, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0182, + "step": 95850 + }, + { + "epoch": 8.712961279767315, + "grad_norm": 0.13644202053546906, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0184, + "step": 95860 + }, + { + "epoch": 8.713870205417196, + "grad_norm": 0.144242063164711, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0171, + "step": 95870 + }, + { + "epoch": 8.714779131067079, + "grad_norm": 0.07121098041534424, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0145, + "step": 95880 + }, + { + "epoch": 8.71568805671696, + "grad_norm": 0.12774458527565002, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0168, + "step": 95890 + }, + { + "epoch": 8.716596982366843, + "grad_norm": 0.21995416283607483, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0176, + "step": 95900 + }, + { + "epoch": 8.717505908016724, + "grad_norm": 0.14646220207214355, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0166, + "step": 95910 + }, + { + "epoch": 8.718414833666607, + "grad_norm": 0.25171899795532227, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0176, + "step": 95920 + }, + { + "epoch": 8.719323759316488, + "grad_norm": 0.13454918563365936, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.016, + "step": 95930 + }, + { + "epoch": 8.72023268496637, + "grad_norm": 0.09443696588277817, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.0157, + "step": 95940 + }, + { + "epoch": 8.721141610616252, + "grad_norm": 0.1693974882364273, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0168, + "step": 95950 + }, + { + "epoch": 8.722050536266133, + "grad_norm": 0.17973701655864716, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0185, + "step": 95960 + }, + { + "epoch": 8.722959461916016, + "grad_norm": 0.1397119164466858, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0157, + "step": 95970 + }, + { + "epoch": 8.723868387565897, + "grad_norm": 0.15216709673404694, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.0162, + "step": 95980 + }, + { + "epoch": 8.724777313215778, + "grad_norm": 0.2197987288236618, + "learning_rate": 3.165685753876885e-06, + "loss": 0.016, + "step": 95990 + }, + { + "epoch": 8.725686238865661, + "grad_norm": 0.11338487267494202, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.0148, + "step": 96000 + }, + { + "epoch": 8.726595164515542, + "grad_norm": 0.15955358743667603, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0148, + "step": 96010 + }, + { + "epoch": 8.727504090165425, + "grad_norm": 0.1375647634267807, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0171, + "step": 96020 + }, + { + "epoch": 8.728413015815306, + "grad_norm": 0.10902199149131775, + "learning_rate": 3.155533169103323e-06, + "loss": 0.0157, + "step": 96030 + }, + { + "epoch": 8.729321941465187, + "grad_norm": 0.1350325345993042, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0146, + "step": 96040 + }, + { + "epoch": 8.73023086711507, + "grad_norm": 0.1087348610162735, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0156, + "step": 96050 + }, + { + "epoch": 8.731139792764951, + "grad_norm": 0.12798254191875458, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0176, + "step": 96060 + }, + { + "epoch": 8.732048718414834, + "grad_norm": 0.14449401199817657, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0159, + "step": 96070 + }, + { + "epoch": 8.732957644064715, + "grad_norm": 0.1566208004951477, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0141, + "step": 96080 + }, + { + "epoch": 8.733866569714598, + "grad_norm": 0.17786802351474762, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0169, + "step": 96090 + }, + { + "epoch": 8.73477549536448, + "grad_norm": 0.15196539461612701, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0174, + "step": 96100 + }, + { + "epoch": 8.73568442101436, + "grad_norm": 0.15531817078590393, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0151, + "step": 96110 + }, + { + "epoch": 8.736593346664243, + "grad_norm": 0.22573982179164886, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0147, + "step": 96120 + }, + { + "epoch": 8.737502272314124, + "grad_norm": 0.12717966735363007, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0158, + "step": 96130 + }, + { + "epoch": 8.738411197964007, + "grad_norm": 0.14574125409126282, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0155, + "step": 96140 + }, + { + "epoch": 8.739320123613888, + "grad_norm": 0.34176063537597656, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0153, + "step": 96150 + }, + { + "epoch": 8.740229049263771, + "grad_norm": 0.12283363938331604, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0172, + "step": 96160 + }, + { + "epoch": 8.741137974913652, + "grad_norm": 0.3056349456310272, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0158, + "step": 96170 + }, + { + "epoch": 8.742046900563533, + "grad_norm": 0.17706100642681122, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0153, + "step": 96180 + }, + { + "epoch": 8.742955826213416, + "grad_norm": 0.11904853582382202, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0161, + "step": 96190 + }, + { + "epoch": 8.743864751863297, + "grad_norm": 0.08872325718402863, + "learning_rate": 3.112545394938825e-06, + "loss": 0.016, + "step": 96200 + }, + { + "epoch": 8.74477367751318, + "grad_norm": 0.1273043006658554, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0152, + "step": 96210 + }, + { + "epoch": 8.745682603163061, + "grad_norm": 0.12290317565202713, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0171, + "step": 96220 + }, + { + "epoch": 8.746591528812942, + "grad_norm": 0.14558911323547363, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0158, + "step": 96230 + }, + { + "epoch": 8.747500454462825, + "grad_norm": 0.22591666877269745, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0152, + "step": 96240 + }, + { + "epoch": 8.748409380112706, + "grad_norm": 0.13767266273498535, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0154, + "step": 96250 + }, + { + "epoch": 8.74931830576259, + "grad_norm": 0.13815701007843018, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.0159, + "step": 96260 + }, + { + "epoch": 8.75022723141247, + "grad_norm": 0.19697710871696472, + "learning_rate": 3.094920326034417e-06, + "loss": 0.0163, + "step": 96270 + }, + { + "epoch": 8.751136157062351, + "grad_norm": 0.1071443259716034, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0168, + "step": 96280 + }, + { + "epoch": 8.752045082712234, + "grad_norm": 0.25269630551338196, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0149, + "step": 96290 + }, + { + "epoch": 8.752954008362115, + "grad_norm": 0.158115953207016, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0158, + "step": 96300 + }, + { + "epoch": 8.753862934011998, + "grad_norm": 0.18007059395313263, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.0155, + "step": 96310 + }, + { + "epoch": 8.75477185966188, + "grad_norm": 0.2025955766439438, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0178, + "step": 96320 + }, + { + "epoch": 8.755680785311762, + "grad_norm": 0.09037807583808899, + "learning_rate": 3.079848382001451e-06, + "loss": 0.0149, + "step": 96330 + }, + { + "epoch": 8.756589710961643, + "grad_norm": 0.15030816197395325, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0154, + "step": 96340 + }, + { + "epoch": 8.757498636611524, + "grad_norm": 0.0876535028219223, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0151, + "step": 96350 + }, + { + "epoch": 8.758407562261407, + "grad_norm": 0.0961214154958725, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0142, + "step": 96360 + }, + { + "epoch": 8.759316487911288, + "grad_norm": 0.12568135559558868, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0151, + "step": 96370 + }, + { + "epoch": 8.760225413561171, + "grad_norm": 0.09829648584127426, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.016, + "step": 96380 + }, + { + "epoch": 8.761134339211052, + "grad_norm": 0.1158696711063385, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0163, + "step": 96390 + }, + { + "epoch": 8.762043264860935, + "grad_norm": 0.11596585065126419, + "learning_rate": 3.062305639362225e-06, + "loss": 0.015, + "step": 96400 + }, + { + "epoch": 8.762952190510816, + "grad_norm": 0.12758296728134155, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0164, + "step": 96410 + }, + { + "epoch": 8.763861116160697, + "grad_norm": 0.13385708630084991, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0168, + "step": 96420 + }, + { + "epoch": 8.76477004181058, + "grad_norm": 0.13931730389595032, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0172, + "step": 96430 + }, + { + "epoch": 8.765678967460461, + "grad_norm": 0.1374617964029312, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0155, + "step": 96440 + }, + { + "epoch": 8.766587893110344, + "grad_norm": 0.12206440418958664, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.0168, + "step": 96450 + }, + { + "epoch": 8.767496818760225, + "grad_norm": 0.11449534446001053, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0165, + "step": 96460 + }, + { + "epoch": 8.768405744410106, + "grad_norm": 0.1405477672815323, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0164, + "step": 96470 + }, + { + "epoch": 8.76931467005999, + "grad_norm": 0.0902453064918518, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0176, + "step": 96480 + }, + { + "epoch": 8.77022359570987, + "grad_norm": 0.12784619629383087, + "learning_rate": 3.039815968052534e-06, + "loss": 0.014, + "step": 96490 + }, + { + "epoch": 8.771132521359753, + "grad_norm": 0.13699759542942047, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0163, + "step": 96500 + }, + { + "epoch": 8.772041447009634, + "grad_norm": 0.09034095704555511, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0167, + "step": 96510 + }, + { + "epoch": 8.772950372659517, + "grad_norm": 0.14064420759677887, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0143, + "step": 96520 + }, + { + "epoch": 8.773859298309398, + "grad_norm": 0.15700407326221466, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0153, + "step": 96530 + }, + { + "epoch": 8.77476822395928, + "grad_norm": 0.1624026894569397, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0166, + "step": 96540 + }, + { + "epoch": 8.775677149609162, + "grad_norm": 0.1435837745666504, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.016, + "step": 96550 + }, + { + "epoch": 8.776586075259043, + "grad_norm": 0.12553557753562927, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0148, + "step": 96560 + }, + { + "epoch": 8.777495000908926, + "grad_norm": 0.10371417552232742, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0147, + "step": 96570 + }, + { + "epoch": 8.778403926558807, + "grad_norm": 0.16228830814361572, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.0169, + "step": 96580 + }, + { + "epoch": 8.779312852208689, + "grad_norm": 0.12138967961072922, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0166, + "step": 96590 + }, + { + "epoch": 8.780221777858571, + "grad_norm": 0.16392873227596283, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.0177, + "step": 96600 + }, + { + "epoch": 8.781130703508452, + "grad_norm": 0.19882692396640778, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0176, + "step": 96610 + }, + { + "epoch": 8.782039629158335, + "grad_norm": 0.21789823472499847, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0179, + "step": 96620 + }, + { + "epoch": 8.782948554808216, + "grad_norm": 0.10489680618047714, + "learning_rate": 3.004978321428544e-06, + "loss": 0.0179, + "step": 96630 + }, + { + "epoch": 8.7838574804581, + "grad_norm": 0.14059798419475555, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0185, + "step": 96640 + }, + { + "epoch": 8.78476640610798, + "grad_norm": 0.10768094658851624, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0137, + "step": 96650 + }, + { + "epoch": 8.785675331757862, + "grad_norm": 0.1516115367412567, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0158, + "step": 96660 + }, + { + "epoch": 8.786584257407744, + "grad_norm": 0.2078978568315506, + "learning_rate": 2.995057479394476e-06, + "loss": 0.014, + "step": 96670 + }, + { + "epoch": 8.787493183057626, + "grad_norm": 0.10726914554834366, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.0172, + "step": 96680 + }, + { + "epoch": 8.788402108707508, + "grad_norm": 0.11632814258337021, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0151, + "step": 96690 + }, + { + "epoch": 8.78931103435739, + "grad_norm": 0.10527195781469345, + "learning_rate": 2.987626420019107e-06, + "loss": 0.016, + "step": 96700 + }, + { + "epoch": 8.79021996000727, + "grad_norm": 0.11842034757137299, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0161, + "step": 96710 + }, + { + "epoch": 8.791128885657153, + "grad_norm": 0.1401272714138031, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0181, + "step": 96720 + }, + { + "epoch": 8.792037811307035, + "grad_norm": 0.0960172712802887, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0182, + "step": 96730 + }, + { + "epoch": 8.792946736956917, + "grad_norm": 0.17541097104549408, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0161, + "step": 96740 + }, + { + "epoch": 8.793855662606799, + "grad_norm": 0.1146048828959465, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0158, + "step": 96750 + }, + { + "epoch": 8.794764588256681, + "grad_norm": 0.10335710644721985, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0159, + "step": 96760 + }, + { + "epoch": 8.795673513906562, + "grad_norm": 0.18504859507083893, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.016, + "step": 96770 + }, + { + "epoch": 8.796582439556444, + "grad_norm": 0.10724784433841705, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0169, + "step": 96780 + }, + { + "epoch": 8.797491365206326, + "grad_norm": 0.11481451988220215, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0163, + "step": 96790 + }, + { + "epoch": 8.798400290856208, + "grad_norm": 0.12083131074905396, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0158, + "step": 96800 + }, + { + "epoch": 8.79930921650609, + "grad_norm": 0.10841618478298187, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0195, + "step": 96810 + }, + { + "epoch": 8.800218142155972, + "grad_norm": 0.19413091242313385, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0168, + "step": 96820 + }, + { + "epoch": 8.801127067805854, + "grad_norm": 0.16889087855815887, + "learning_rate": 2.955520102911375e-06, + "loss": 0.018, + "step": 96830 + }, + { + "epoch": 8.802035993455736, + "grad_norm": 0.3720013499259949, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0191, + "step": 96840 + }, + { + "epoch": 8.802944919105617, + "grad_norm": 0.11845823377370834, + "learning_rate": 2.950594379688023e-06, + "loss": 0.0172, + "step": 96850 + }, + { + "epoch": 8.8038538447555, + "grad_norm": 0.11967938393354416, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0149, + "step": 96860 + }, + { + "epoch": 8.80476277040538, + "grad_norm": 0.15420232713222504, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0175, + "step": 96870 + }, + { + "epoch": 8.805671696055263, + "grad_norm": 0.14913785457611084, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0155, + "step": 96880 + }, + { + "epoch": 8.806580621705145, + "grad_norm": 0.18041209876537323, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0167, + "step": 96890 + }, + { + "epoch": 8.807489547355026, + "grad_norm": 0.1924706995487213, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.019, + "step": 96900 + }, + { + "epoch": 8.808398473004909, + "grad_norm": 0.12516486644744873, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0152, + "step": 96910 + }, + { + "epoch": 8.80930739865479, + "grad_norm": 0.13032342493534088, + "learning_rate": 2.933383183009594e-06, + "loss": 0.0175, + "step": 96920 + }, + { + "epoch": 8.810216324304672, + "grad_norm": 0.1265336126089096, + "learning_rate": 2.930928104653789e-06, + "loss": 0.016, + "step": 96930 + }, + { + "epoch": 8.811125249954554, + "grad_norm": 0.1391792744398117, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0153, + "step": 96940 + }, + { + "epoch": 8.812034175604435, + "grad_norm": 0.1353389471769333, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.0204, + "step": 96950 + }, + { + "epoch": 8.812943101254318, + "grad_norm": 0.09235768765211105, + "learning_rate": 2.923568369729992e-06, + "loss": 0.0162, + "step": 96960 + }, + { + "epoch": 8.813852026904199, + "grad_norm": 0.12913909554481506, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0144, + "step": 96970 + }, + { + "epoch": 8.814760952554082, + "grad_norm": 0.09778057038784027, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0173, + "step": 96980 + }, + { + "epoch": 8.815669878203963, + "grad_norm": 0.1669250726699829, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0165, + "step": 96990 + }, + { + "epoch": 8.816578803853846, + "grad_norm": 0.139292374253273, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.0182, + "step": 97000 + }, + { + "epoch": 8.817487729503727, + "grad_norm": 0.10354138910770416, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0189, + "step": 97010 + }, + { + "epoch": 8.818396655153608, + "grad_norm": 0.08785372972488403, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0155, + "step": 97020 + }, + { + "epoch": 8.81930558080349, + "grad_norm": 0.13613344728946686, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0176, + "step": 97030 + }, + { + "epoch": 8.820214506453372, + "grad_norm": 0.12736830115318298, + "learning_rate": 2.903982777088368e-06, + "loss": 0.016, + "step": 97040 + }, + { + "epoch": 8.821123432103255, + "grad_norm": 0.09103083610534668, + "learning_rate": 2.901538709814486e-06, + "loss": 0.015, + "step": 97050 + }, + { + "epoch": 8.822032357753136, + "grad_norm": 0.11347278952598572, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0177, + "step": 97060 + }, + { + "epoch": 8.822941283403019, + "grad_norm": 0.12815602123737335, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0166, + "step": 97070 + }, + { + "epoch": 8.8238502090529, + "grad_norm": 0.12449793517589569, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0152, + "step": 97080 + }, + { + "epoch": 8.82475913470278, + "grad_norm": 0.14904972910881042, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0194, + "step": 97090 + }, + { + "epoch": 8.825668060352664, + "grad_norm": 0.16990944743156433, + "learning_rate": 2.889332159129238e-06, + "loss": 0.015, + "step": 97100 + }, + { + "epoch": 8.826576986002545, + "grad_norm": 0.15614762902259827, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.016, + "step": 97110 + }, + { + "epoch": 8.827485911652428, + "grad_norm": 0.081304632127285, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0138, + "step": 97120 + }, + { + "epoch": 8.828394837302309, + "grad_norm": 0.15691038966178894, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0151, + "step": 97130 + }, + { + "epoch": 8.82930376295219, + "grad_norm": 0.10920993983745575, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0145, + "step": 97140 + }, + { + "epoch": 8.830212688602073, + "grad_norm": 0.21341818571090698, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0158, + "step": 97150 + }, + { + "epoch": 8.831121614251954, + "grad_norm": 0.11443641781806946, + "learning_rate": 2.874714651273537e-06, + "loss": 0.0182, + "step": 97160 + }, + { + "epoch": 8.832030539901837, + "grad_norm": 0.29945871233940125, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.019, + "step": 97170 + }, + { + "epoch": 8.832939465551718, + "grad_norm": 0.11538063734769821, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0178, + "step": 97180 + }, + { + "epoch": 8.833848391201599, + "grad_norm": 0.1391768902540207, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0168, + "step": 97190 + }, + { + "epoch": 8.834757316851482, + "grad_norm": 0.15407133102416992, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0165, + "step": 97200 + }, + { + "epoch": 8.835666242501363, + "grad_norm": 0.17176426947116852, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.0169, + "step": 97210 + }, + { + "epoch": 8.836575168151246, + "grad_norm": 0.16716159880161285, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0151, + "step": 97220 + }, + { + "epoch": 8.837484093801127, + "grad_norm": 0.1681155264377594, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0164, + "step": 97230 + }, + { + "epoch": 8.83839301945101, + "grad_norm": 0.1395251452922821, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0147, + "step": 97240 + }, + { + "epoch": 8.83930194510089, + "grad_norm": 0.12001735717058182, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0168, + "step": 97250 + }, + { + "epoch": 8.840210870750772, + "grad_norm": 0.11302712559700012, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.0177, + "step": 97260 + }, + { + "epoch": 8.841119796400655, + "grad_norm": 0.11167166382074356, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.0166, + "step": 97270 + }, + { + "epoch": 8.842028722050536, + "grad_norm": 0.12604665756225586, + "learning_rate": 2.845579125636639e-06, + "loss": 0.019, + "step": 97280 + }, + { + "epoch": 8.842937647700419, + "grad_norm": 0.15117871761322021, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0137, + "step": 97290 + }, + { + "epoch": 8.8438465733503, + "grad_norm": 0.151219442486763, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0182, + "step": 97300 + }, + { + "epoch": 8.844755499000183, + "grad_norm": 0.11779429018497467, + "learning_rate": 2.838316000375582e-06, + "loss": 0.015, + "step": 97310 + }, + { + "epoch": 8.845664424650064, + "grad_norm": 0.11051474511623383, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0171, + "step": 97320 + }, + { + "epoch": 8.846573350299945, + "grad_norm": 0.10346323996782303, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0168, + "step": 97330 + }, + { + "epoch": 8.847482275949828, + "grad_norm": 0.14479444921016693, + "learning_rate": 2.831061187510856e-06, + "loss": 0.016, + "step": 97340 + }, + { + "epoch": 8.848391201599709, + "grad_norm": 0.36754709482192993, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0164, + "step": 97350 + }, + { + "epoch": 8.849300127249592, + "grad_norm": 0.17761175334453583, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0194, + "step": 97360 + }, + { + "epoch": 8.850209052899473, + "grad_norm": 0.15608882904052734, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.016, + "step": 97370 + }, + { + "epoch": 8.851117978549354, + "grad_norm": 0.1506519913673401, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0167, + "step": 97380 + }, + { + "epoch": 8.852026904199237, + "grad_norm": 0.21595938503742218, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.017, + "step": 97390 + }, + { + "epoch": 8.852935829849118, + "grad_norm": 0.12239808589220047, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0143, + "step": 97400 + }, + { + "epoch": 8.853844755499, + "grad_norm": 0.14480572938919067, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0149, + "step": 97410 + }, + { + "epoch": 8.854753681148882, + "grad_norm": 0.1005377247929573, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0157, + "step": 97420 + }, + { + "epoch": 8.855662606798763, + "grad_norm": 0.09643490612506866, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0147, + "step": 97430 + }, + { + "epoch": 8.856571532448646, + "grad_norm": 0.1230727881193161, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.017, + "step": 97440 + }, + { + "epoch": 8.857480458098527, + "grad_norm": 0.12451667338609695, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0162, + "step": 97450 + }, + { + "epoch": 8.85838938374841, + "grad_norm": 0.10325875133275986, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.015, + "step": 97460 + }, + { + "epoch": 8.85929830939829, + "grad_norm": 0.12422624975442886, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0166, + "step": 97470 + }, + { + "epoch": 8.860207235048174, + "grad_norm": 0.1957891881465912, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.017, + "step": 97480 + }, + { + "epoch": 8.861116160698055, + "grad_norm": 0.12344527244567871, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0148, + "step": 97490 + }, + { + "epoch": 8.862025086347936, + "grad_norm": 0.190042644739151, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0169, + "step": 97500 + }, + { + "epoch": 8.862934011997819, + "grad_norm": 0.14952139556407928, + "learning_rate": 2.790107832247314e-06, + "loss": 0.0163, + "step": 97510 + }, + { + "epoch": 8.8638429376477, + "grad_norm": 0.09478028863668442, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0157, + "step": 97520 + }, + { + "epoch": 8.864751863297583, + "grad_norm": 0.17811912298202515, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0161, + "step": 97530 + }, + { + "epoch": 8.865660788947464, + "grad_norm": 0.128569558262825, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0153, + "step": 97540 + }, + { + "epoch": 8.866569714597347, + "grad_norm": 0.1256781965494156, + "learning_rate": 2.780510660360556e-06, + "loss": 0.0159, + "step": 97550 + }, + { + "epoch": 8.867478640247228, + "grad_norm": 0.1488211750984192, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0155, + "step": 97560 + }, + { + "epoch": 8.868387565897109, + "grad_norm": 0.1659182608127594, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0165, + "step": 97570 + }, + { + "epoch": 8.869296491546992, + "grad_norm": 0.14195013046264648, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0149, + "step": 97580 + }, + { + "epoch": 8.870205417196873, + "grad_norm": 0.10922771692276001, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0156, + "step": 97590 + }, + { + "epoch": 8.871114342846756, + "grad_norm": 0.14557404816150665, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0143, + "step": 97600 + }, + { + "epoch": 8.872023268496637, + "grad_norm": 0.14228060841560364, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0168, + "step": 97610 + }, + { + "epoch": 8.872932194146518, + "grad_norm": 0.16019794344902039, + "learning_rate": 2.763751344158856e-06, + "loss": 0.0149, + "step": 97620 + }, + { + "epoch": 8.873841119796401, + "grad_norm": 0.2007058709859848, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0138, + "step": 97630 + }, + { + "epoch": 8.874750045446282, + "grad_norm": 0.21823804080486298, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0154, + "step": 97640 + }, + { + "epoch": 8.875658971096165, + "grad_norm": 0.12463022768497467, + "learning_rate": 2.756582715372719e-06, + "loss": 0.0175, + "step": 97650 + }, + { + "epoch": 8.876567896746046, + "grad_norm": 0.2099458873271942, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0161, + "step": 97660 + }, + { + "epoch": 8.877476822395929, + "grad_norm": 0.21198266744613647, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.0171, + "step": 97670 + }, + { + "epoch": 8.87838574804581, + "grad_norm": 0.1351025402545929, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0158, + "step": 97680 + }, + { + "epoch": 8.879294673695691, + "grad_norm": 0.267978698015213, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0143, + "step": 97690 + }, + { + "epoch": 8.880203599345574, + "grad_norm": 0.1329433172941208, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0159, + "step": 97700 + }, + { + "epoch": 8.881112524995455, + "grad_norm": 0.20742154121398926, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0184, + "step": 97710 + }, + { + "epoch": 8.882021450645338, + "grad_norm": 0.1379358023405075, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0148, + "step": 97720 + }, + { + "epoch": 8.882930376295219, + "grad_norm": 0.11649248749017715, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0152, + "step": 97730 + }, + { + "epoch": 8.8838393019451, + "grad_norm": 0.15533508360385895, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0149, + "step": 97740 + }, + { + "epoch": 8.884748227594983, + "grad_norm": 0.09617424011230469, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0166, + "step": 97750 + }, + { + "epoch": 8.885657153244864, + "grad_norm": 0.10617100447416306, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0152, + "step": 97760 + }, + { + "epoch": 8.886566078894747, + "grad_norm": 0.0850234255194664, + "learning_rate": 2.727991931741916e-06, + "loss": 0.0125, + "step": 97770 + }, + { + "epoch": 8.887475004544628, + "grad_norm": 0.20162494480609894, + "learning_rate": 2.725615419989159e-06, + "loss": 0.017, + "step": 97780 + }, + { + "epoch": 8.888383930194511, + "grad_norm": 0.11881451308727264, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.0186, + "step": 97790 + }, + { + "epoch": 8.889292855844392, + "grad_norm": 0.17121316492557526, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0155, + "step": 97800 + }, + { + "epoch": 8.890201781494273, + "grad_norm": 0.09280572831630707, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0147, + "step": 97810 + }, + { + "epoch": 8.891110707144156, + "grad_norm": 0.09339626133441925, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0177, + "step": 97820 + }, + { + "epoch": 8.892019632794037, + "grad_norm": 0.11818967759609222, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0182, + "step": 97830 + }, + { + "epoch": 8.89292855844392, + "grad_norm": 0.12801875174045563, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0163, + "step": 97840 + }, + { + "epoch": 8.893837484093801, + "grad_norm": 0.19563867151737213, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0135, + "step": 97850 + }, + { + "epoch": 8.894746409743682, + "grad_norm": 0.11191780865192413, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0159, + "step": 97860 + }, + { + "epoch": 8.895655335393565, + "grad_norm": 0.0877610519528389, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0177, + "step": 97870 + }, + { + "epoch": 8.896564261043446, + "grad_norm": 0.08608563989400864, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0138, + "step": 97880 + }, + { + "epoch": 8.897473186693329, + "grad_norm": 0.18136103451251984, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0175, + "step": 97890 + }, + { + "epoch": 8.89838211234321, + "grad_norm": 0.13490010797977448, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.019, + "step": 97900 + }, + { + "epoch": 8.899291037993093, + "grad_norm": 0.15200133621692657, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0163, + "step": 97910 + }, + { + "epoch": 8.900199963642974, + "grad_norm": 0.10464310646057129, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0153, + "step": 97920 + }, + { + "epoch": 8.901108889292855, + "grad_norm": 0.13862594962120056, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0156, + "step": 97930 + }, + { + "epoch": 8.902017814942738, + "grad_norm": 0.20806123316287994, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0159, + "step": 97940 + }, + { + "epoch": 8.90292674059262, + "grad_norm": 0.1111646443605423, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0142, + "step": 97950 + }, + { + "epoch": 8.903835666242502, + "grad_norm": 0.14023420214653015, + "learning_rate": 2.682997782086152e-06, + "loss": 0.0159, + "step": 97960 + }, + { + "epoch": 8.904744591892383, + "grad_norm": 0.16405649483203888, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0159, + "step": 97970 + }, + { + "epoch": 8.905653517542266, + "grad_norm": 0.12769241631031036, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0172, + "step": 97980 + }, + { + "epoch": 8.906562443192147, + "grad_norm": 0.15966804325580597, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0157, + "step": 97990 + }, + { + "epoch": 8.907471368842028, + "grad_norm": 0.1038496345281601, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0162, + "step": 98000 + }, + { + "epoch": 8.908380294491911, + "grad_norm": 0.09879356622695923, + "learning_rate": 2.671213289819149e-06, + "loss": 0.017, + "step": 98010 + }, + { + "epoch": 8.909289220141792, + "grad_norm": 0.11854495108127594, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0176, + "step": 98020 + }, + { + "epoch": 8.910198145791675, + "grad_norm": 0.14551903307437897, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0168, + "step": 98030 + }, + { + "epoch": 8.911107071441556, + "grad_norm": 0.14257672429084778, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0155, + "step": 98040 + }, + { + "epoch": 8.912015997091437, + "grad_norm": 0.135885551571846, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0165, + "step": 98050 + }, + { + "epoch": 8.91292492274132, + "grad_norm": 0.09859802573919296, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0168, + "step": 98060 + }, + { + "epoch": 8.913833848391201, + "grad_norm": 0.1658245474100113, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0149, + "step": 98070 + }, + { + "epoch": 8.914742774041084, + "grad_norm": 0.1513161063194275, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0151, + "step": 98080 + }, + { + "epoch": 8.915651699690965, + "grad_norm": 0.13101214170455933, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0162, + "step": 98090 + }, + { + "epoch": 8.916560625340846, + "grad_norm": 0.1605924814939499, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0144, + "step": 98100 + }, + { + "epoch": 8.91746955099073, + "grad_norm": 0.11093144863843918, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.0175, + "step": 98110 + }, + { + "epoch": 8.91837847664061, + "grad_norm": 0.11853229999542236, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0154, + "step": 98120 + }, + { + "epoch": 8.919287402290493, + "grad_norm": 0.20218230783939362, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0165, + "step": 98130 + }, + { + "epoch": 8.920196327940374, + "grad_norm": 0.13443094491958618, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0139, + "step": 98140 + }, + { + "epoch": 8.921105253590257, + "grad_norm": 0.11654748767614365, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0145, + "step": 98150 + }, + { + "epoch": 8.922014179240138, + "grad_norm": 0.09884502738714218, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0165, + "step": 98160 + }, + { + "epoch": 8.92292310489002, + "grad_norm": 0.1264423429965973, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0151, + "step": 98170 + }, + { + "epoch": 8.923832030539902, + "grad_norm": 0.08882208913564682, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0176, + "step": 98180 + }, + { + "epoch": 8.924740956189783, + "grad_norm": 0.1774922013282776, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0176, + "step": 98190 + }, + { + "epoch": 8.925649881839666, + "grad_norm": 0.15769042074680328, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0145, + "step": 98200 + }, + { + "epoch": 8.926558807489547, + "grad_norm": 0.10998213291168213, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0151, + "step": 98210 + }, + { + "epoch": 8.92746773313943, + "grad_norm": 0.11185592412948608, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0145, + "step": 98220 + }, + { + "epoch": 8.928376658789311, + "grad_norm": 0.151886448264122, + "learning_rate": 2.619639902212251e-06, + "loss": 0.0144, + "step": 98230 + }, + { + "epoch": 8.929285584439192, + "grad_norm": 0.09129831939935684, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0153, + "step": 98240 + }, + { + "epoch": 8.930194510089075, + "grad_norm": 0.10350368171930313, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0155, + "step": 98250 + }, + { + "epoch": 8.931103435738956, + "grad_norm": 0.13779699802398682, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0168, + "step": 98260 + }, + { + "epoch": 8.93201236138884, + "grad_norm": 0.12466688454151154, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0171, + "step": 98270 + }, + { + "epoch": 8.93292128703872, + "grad_norm": 0.10915543884038925, + "learning_rate": 2.607982073748596e-06, + "loss": 0.015, + "step": 98280 + }, + { + "epoch": 8.933830212688601, + "grad_norm": 0.17507022619247437, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0171, + "step": 98290 + }, + { + "epoch": 8.934739138338484, + "grad_norm": 0.1374911069869995, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0176, + "step": 98300 + }, + { + "epoch": 8.935648063988365, + "grad_norm": 0.14206671714782715, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0165, + "step": 98310 + }, + { + "epoch": 8.936556989638248, + "grad_norm": 0.14426127076148987, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0163, + "step": 98320 + }, + { + "epoch": 8.93746591528813, + "grad_norm": 0.1412079930305481, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0176, + "step": 98330 + }, + { + "epoch": 8.93837484093801, + "grad_norm": 0.13858330249786377, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0161, + "step": 98340 + }, + { + "epoch": 8.939283766587893, + "grad_norm": 0.14497266709804535, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.0194, + "step": 98350 + }, + { + "epoch": 8.940192692237774, + "grad_norm": 0.12991683185100555, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.0167, + "step": 98360 + }, + { + "epoch": 8.941101617887657, + "grad_norm": 0.11626679450273514, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0152, + "step": 98370 + }, + { + "epoch": 8.942010543537538, + "grad_norm": 0.10889684408903122, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.014, + "step": 98380 + }, + { + "epoch": 8.942919469187421, + "grad_norm": 0.14572688937187195, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0156, + "step": 98390 + }, + { + "epoch": 8.943828394837302, + "grad_norm": 0.10992126911878586, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0163, + "step": 98400 + }, + { + "epoch": 8.944737320487183, + "grad_norm": 0.1335267424583435, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0196, + "step": 98410 + }, + { + "epoch": 8.945646246137066, + "grad_norm": 0.3122636377811432, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0189, + "step": 98420 + }, + { + "epoch": 8.946555171786947, + "grad_norm": 0.11699274927377701, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0163, + "step": 98430 + }, + { + "epoch": 8.94746409743683, + "grad_norm": 0.1664879471063614, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0153, + "step": 98440 + }, + { + "epoch": 8.948373023086711, + "grad_norm": 0.10325510054826736, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0164, + "step": 98450 + }, + { + "epoch": 8.949281948736594, + "grad_norm": 0.11390969157218933, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0138, + "step": 98460 + }, + { + "epoch": 8.950190874386475, + "grad_norm": 0.13162565231323242, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.0157, + "step": 98470 + }, + { + "epoch": 8.951099800036356, + "grad_norm": 0.11871901154518127, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0165, + "step": 98480 + }, + { + "epoch": 8.95200872568624, + "grad_norm": 0.12708115577697754, + "learning_rate": 2.559276369275866e-06, + "loss": 0.018, + "step": 98490 + }, + { + "epoch": 8.95291765133612, + "grad_norm": 0.12880846858024597, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.0158, + "step": 98500 + }, + { + "epoch": 8.953826576986003, + "grad_norm": 0.1395493596792221, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0161, + "step": 98510 + }, + { + "epoch": 8.954735502635884, + "grad_norm": 0.2310246080160141, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0166, + "step": 98520 + }, + { + "epoch": 8.955644428285765, + "grad_norm": 0.17009828984737396, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0179, + "step": 98530 + }, + { + "epoch": 8.956553353935648, + "grad_norm": 0.09740030765533447, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.0182, + "step": 98540 + }, + { + "epoch": 8.95746227958553, + "grad_norm": 0.20096561312675476, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0155, + "step": 98550 + }, + { + "epoch": 8.958371205235412, + "grad_norm": 0.13015854358673096, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0168, + "step": 98560 + }, + { + "epoch": 8.959280130885293, + "grad_norm": 0.1310141682624817, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.0163, + "step": 98570 + }, + { + "epoch": 8.960189056535175, + "grad_norm": 0.20144149661064148, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0168, + "step": 98580 + }, + { + "epoch": 8.961097982185057, + "grad_norm": 0.21358610689640045, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0176, + "step": 98590 + }, + { + "epoch": 8.962006907834938, + "grad_norm": 0.11460746824741364, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0157, + "step": 98600 + }, + { + "epoch": 8.962915833484821, + "grad_norm": 0.12207245826721191, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0163, + "step": 98610 + }, + { + "epoch": 8.963824759134702, + "grad_norm": 0.17652519047260284, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0159, + "step": 98620 + }, + { + "epoch": 8.964733684784585, + "grad_norm": 0.2640214264392853, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0159, + "step": 98630 + }, + { + "epoch": 8.965642610434466, + "grad_norm": 0.1255936324596405, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.015, + "step": 98640 + }, + { + "epoch": 8.966551536084348, + "grad_norm": 0.14878198504447937, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0172, + "step": 98650 + }, + { + "epoch": 8.96746046173423, + "grad_norm": 0.12608405947685242, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0177, + "step": 98660 + }, + { + "epoch": 8.968369387384111, + "grad_norm": 0.1461082547903061, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0173, + "step": 98670 + }, + { + "epoch": 8.969278313033994, + "grad_norm": 0.37517842650413513, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0161, + "step": 98680 + }, + { + "epoch": 8.970187238683875, + "grad_norm": 0.1607171595096588, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.017, + "step": 98690 + }, + { + "epoch": 8.971096164333758, + "grad_norm": 0.10467534512281418, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0182, + "step": 98700 + }, + { + "epoch": 8.97200508998364, + "grad_norm": 0.22804926335811615, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0162, + "step": 98710 + }, + { + "epoch": 8.97291401563352, + "grad_norm": 0.19161999225616455, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0156, + "step": 98720 + }, + { + "epoch": 8.973822941283403, + "grad_norm": 0.15382540225982666, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.0176, + "step": 98730 + }, + { + "epoch": 8.974731866933285, + "grad_norm": 0.12061220407485962, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0168, + "step": 98740 + }, + { + "epoch": 8.975640792583167, + "grad_norm": 0.13241004943847656, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0161, + "step": 98750 + }, + { + "epoch": 8.976549718233048, + "grad_norm": 0.1322048008441925, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0183, + "step": 98760 + }, + { + "epoch": 8.97745864388293, + "grad_norm": 0.11854259669780731, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0144, + "step": 98770 + }, + { + "epoch": 8.978367569532812, + "grad_norm": 0.12317338585853577, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0165, + "step": 98780 + }, + { + "epoch": 8.979276495182694, + "grad_norm": 0.13109543919563293, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0174, + "step": 98790 + }, + { + "epoch": 8.980185420832576, + "grad_norm": 0.11365839838981628, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0157, + "step": 98800 + }, + { + "epoch": 8.981094346482458, + "grad_norm": 0.11080987006425858, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0164, + "step": 98810 + }, + { + "epoch": 8.98200327213234, + "grad_norm": 0.13904716074466705, + "learning_rate": 2.483581262943742e-06, + "loss": 0.0167, + "step": 98820 + }, + { + "epoch": 8.982912197782221, + "grad_norm": 0.12096700072288513, + "learning_rate": 2.481303599532329e-06, + "loss": 0.019, + "step": 98830 + }, + { + "epoch": 8.983821123432103, + "grad_norm": 0.11136981844902039, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0157, + "step": 98840 + }, + { + "epoch": 8.984730049081985, + "grad_norm": 0.14740408957004547, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0158, + "step": 98850 + }, + { + "epoch": 8.985638974731867, + "grad_norm": 0.11586789041757584, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.0163, + "step": 98860 + }, + { + "epoch": 8.98654790038175, + "grad_norm": 0.13925069570541382, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0187, + "step": 98870 + }, + { + "epoch": 8.98745682603163, + "grad_norm": 0.08381679654121399, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.0156, + "step": 98880 + }, + { + "epoch": 8.988365751681512, + "grad_norm": 0.5503324866294861, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0168, + "step": 98890 + }, + { + "epoch": 8.989274677331395, + "grad_norm": 0.12335320562124252, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0162, + "step": 98900 + }, + { + "epoch": 8.990183602981276, + "grad_norm": 0.1643621027469635, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0194, + "step": 98910 + }, + { + "epoch": 8.991092528631158, + "grad_norm": 0.15831197798252106, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0161, + "step": 98920 + }, + { + "epoch": 8.99200145428104, + "grad_norm": 0.12451887875795364, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.016, + "step": 98930 + }, + { + "epoch": 8.992910379930922, + "grad_norm": 0.19515325129032135, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0183, + "step": 98940 + }, + { + "epoch": 8.993819305580804, + "grad_norm": 0.20505401492118835, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0169, + "step": 98950 + }, + { + "epoch": 8.994728231230685, + "grad_norm": 0.12636880576610565, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0165, + "step": 98960 + }, + { + "epoch": 8.995637156880568, + "grad_norm": 0.1134340837597847, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0161, + "step": 98970 + }, + { + "epoch": 8.996546082530449, + "grad_norm": 0.12121676653623581, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0163, + "step": 98980 + }, + { + "epoch": 8.997455008180332, + "grad_norm": 0.08541832864284515, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0138, + "step": 98990 + }, + { + "epoch": 8.998363933830213, + "grad_norm": 0.1276867389678955, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0164, + "step": 99000 + }, + { + "epoch": 8.999272859480094, + "grad_norm": 0.12570033967494965, + "learning_rate": 2.440468397204442e-06, + "loss": 0.0177, + "step": 99010 + }, + { + "epoch": 9.000181785129977, + "grad_norm": 0.2600955069065094, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0174, + "step": 99020 + }, + { + "epoch": 9.001090710779858, + "grad_norm": 0.12272357195615768, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0168, + "step": 99030 + }, + { + "epoch": 9.00199963642974, + "grad_norm": 0.11143393069505692, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0168, + "step": 99040 + }, + { + "epoch": 9.002908562079622, + "grad_norm": 0.09021732956171036, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0141, + "step": 99050 + }, + { + "epoch": 9.003817487729505, + "grad_norm": 0.21472206711769104, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.0172, + "step": 99060 + }, + { + "epoch": 9.004726413379386, + "grad_norm": 0.14799776673316956, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0163, + "step": 99070 + }, + { + "epoch": 9.005635339029267, + "grad_norm": 0.14362521469593048, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0165, + "step": 99080 + }, + { + "epoch": 9.00654426467915, + "grad_norm": 0.0860215499997139, + "learning_rate": 2.422418555347329e-06, + "loss": 0.0139, + "step": 99090 + }, + { + "epoch": 9.00745319032903, + "grad_norm": 0.15823349356651306, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.0164, + "step": 99100 + }, + { + "epoch": 9.008362115978914, + "grad_norm": 0.12281614542007446, + "learning_rate": 2.41791564330191e-06, + "loss": 0.0185, + "step": 99110 + }, + { + "epoch": 9.009271041628795, + "grad_norm": 0.1436079442501068, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0157, + "step": 99120 + }, + { + "epoch": 9.010179967278676, + "grad_norm": 0.08203576505184174, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0158, + "step": 99130 + }, + { + "epoch": 9.011088892928559, + "grad_norm": 0.1754666566848755, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0167, + "step": 99140 + }, + { + "epoch": 9.01199781857844, + "grad_norm": 0.12316351383924484, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0178, + "step": 99150 + }, + { + "epoch": 9.012906744228323, + "grad_norm": 0.09121730178594589, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0157, + "step": 99160 + }, + { + "epoch": 9.013815669878204, + "grad_norm": 0.11675070226192474, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0158, + "step": 99170 + }, + { + "epoch": 9.014724595528087, + "grad_norm": 0.10711871832609177, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0144, + "step": 99180 + }, + { + "epoch": 9.015633521177968, + "grad_norm": 0.10628274083137512, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0149, + "step": 99190 + }, + { + "epoch": 9.016542446827849, + "grad_norm": 0.12919752299785614, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0178, + "step": 99200 + }, + { + "epoch": 9.017451372477732, + "grad_norm": 0.1471094936132431, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0147, + "step": 99210 + }, + { + "epoch": 9.018360298127613, + "grad_norm": 0.1331847459077835, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0173, + "step": 99220 + }, + { + "epoch": 9.019269223777496, + "grad_norm": 0.14023356139659882, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.018, + "step": 99230 + }, + { + "epoch": 9.020178149427377, + "grad_norm": 0.14276345074176788, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0176, + "step": 99240 + }, + { + "epoch": 9.021087075077258, + "grad_norm": 0.1567142754793167, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0175, + "step": 99250 + }, + { + "epoch": 9.02199600072714, + "grad_norm": 0.1257413625717163, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0171, + "step": 99260 + }, + { + "epoch": 9.022904926377022, + "grad_norm": 0.12874403595924377, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0153, + "step": 99270 + }, + { + "epoch": 9.023813852026905, + "grad_norm": 0.19608740508556366, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0152, + "step": 99280 + }, + { + "epoch": 9.024722777676786, + "grad_norm": 0.15780891478061676, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.0151, + "step": 99290 + }, + { + "epoch": 9.025631703326669, + "grad_norm": 0.18198493123054504, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0182, + "step": 99300 + }, + { + "epoch": 9.02654062897655, + "grad_norm": 0.15387405455112457, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.019, + "step": 99310 + }, + { + "epoch": 9.02744955462643, + "grad_norm": 0.15839281678199768, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0157, + "step": 99320 + }, + { + "epoch": 9.028358480276314, + "grad_norm": 0.17610657215118408, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0154, + "step": 99330 + }, + { + "epoch": 9.029267405926195, + "grad_norm": 0.1694067120552063, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.016, + "step": 99340 + }, + { + "epoch": 9.030176331576078, + "grad_norm": 0.12351836264133453, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0165, + "step": 99350 + }, + { + "epoch": 9.031085257225959, + "grad_norm": 0.13763342797756195, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0169, + "step": 99360 + }, + { + "epoch": 9.03199418287584, + "grad_norm": 0.11839578300714493, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.0142, + "step": 99370 + }, + { + "epoch": 9.032903108525723, + "grad_norm": 0.17182733118534088, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0162, + "step": 99380 + }, + { + "epoch": 9.033812034175604, + "grad_norm": 0.1736196130514145, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0184, + "step": 99390 + }, + { + "epoch": 9.034720959825487, + "grad_norm": 0.11522328108549118, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0158, + "step": 99400 + }, + { + "epoch": 9.035629885475368, + "grad_norm": 0.13674917817115784, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0177, + "step": 99410 + }, + { + "epoch": 9.03653881112525, + "grad_norm": 0.15106512606143951, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0177, + "step": 99420 + }, + { + "epoch": 9.037447736775132, + "grad_norm": 0.09011288732290268, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0155, + "step": 99430 + }, + { + "epoch": 9.038356662425013, + "grad_norm": 0.1175341084599495, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0151, + "step": 99440 + }, + { + "epoch": 9.039265588074896, + "grad_norm": 0.14935338497161865, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0202, + "step": 99450 + }, + { + "epoch": 9.040174513724777, + "grad_norm": 0.10013817995786667, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0175, + "step": 99460 + }, + { + "epoch": 9.04108343937466, + "grad_norm": 0.11351916193962097, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0167, + "step": 99470 + }, + { + "epoch": 9.04199236502454, + "grad_norm": 0.13984976708889008, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.0148, + "step": 99480 + }, + { + "epoch": 9.042901290674422, + "grad_norm": 0.11117847263813019, + "learning_rate": 2.333088066801849e-06, + "loss": 0.016, + "step": 99490 + }, + { + "epoch": 9.043810216324305, + "grad_norm": 0.08397920429706573, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0161, + "step": 99500 + }, + { + "epoch": 9.044719141974186, + "grad_norm": 0.0947093665599823, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0155, + "step": 99510 + }, + { + "epoch": 9.045628067624069, + "grad_norm": 0.1440289318561554, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0163, + "step": 99520 + }, + { + "epoch": 9.04653699327395, + "grad_norm": 0.13957324624061584, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0176, + "step": 99530 + }, + { + "epoch": 9.047445918923833, + "grad_norm": 0.12697674334049225, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0175, + "step": 99540 + }, + { + "epoch": 9.048354844573714, + "grad_norm": 0.12342600524425507, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.018, + "step": 99550 + }, + { + "epoch": 9.049263770223595, + "grad_norm": 0.13643433153629303, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.0165, + "step": 99560 + }, + { + "epoch": 9.050172695873478, + "grad_norm": 0.12484727799892426, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0179, + "step": 99570 + }, + { + "epoch": 9.051081621523359, + "grad_norm": 0.09114673733711243, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0172, + "step": 99580 + }, + { + "epoch": 9.051990547173242, + "grad_norm": 0.11401379108428955, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0148, + "step": 99590 + }, + { + "epoch": 9.052899472823123, + "grad_norm": 0.12364993989467621, + "learning_rate": 2.308791556473921e-06, + "loss": 0.016, + "step": 99600 + }, + { + "epoch": 9.053808398473006, + "grad_norm": 0.12991543114185333, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0182, + "step": 99610 + }, + { + "epoch": 9.054717324122887, + "grad_norm": 0.11419755220413208, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0191, + "step": 99620 + }, + { + "epoch": 9.055626249772768, + "grad_norm": 0.14538128674030304, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0189, + "step": 99630 + }, + { + "epoch": 9.05653517542265, + "grad_norm": 0.11249418556690216, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0148, + "step": 99640 + }, + { + "epoch": 9.057444101072532, + "grad_norm": 0.19335083663463593, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0152, + "step": 99650 + }, + { + "epoch": 9.058353026722415, + "grad_norm": 0.08708131313323975, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0156, + "step": 99660 + }, + { + "epoch": 9.059261952372296, + "grad_norm": 0.13812142610549927, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0161, + "step": 99670 + }, + { + "epoch": 9.060170878022177, + "grad_norm": 0.13183172047138214, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0164, + "step": 99680 + }, + { + "epoch": 9.06107980367206, + "grad_norm": 0.0881526917219162, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0131, + "step": 99690 + }, + { + "epoch": 9.061988729321941, + "grad_norm": 0.14308716356754303, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0179, + "step": 99700 + }, + { + "epoch": 9.062897654971824, + "grad_norm": 0.17435674369335175, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0161, + "step": 99710 + }, + { + "epoch": 9.063806580621705, + "grad_norm": 0.10393442213535309, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0147, + "step": 99720 + }, + { + "epoch": 9.064715506271588, + "grad_norm": 0.11529184132814407, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0174, + "step": 99730 + }, + { + "epoch": 9.065624431921469, + "grad_norm": 0.1351209282875061, + "learning_rate": 2.27803742162607e-06, + "loss": 0.0159, + "step": 99740 + }, + { + "epoch": 9.06653335757135, + "grad_norm": 0.1557433158159256, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.015, + "step": 99750 + }, + { + "epoch": 9.067442283221233, + "grad_norm": 0.09308037161827087, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0135, + "step": 99760 + }, + { + "epoch": 9.068351208871114, + "grad_norm": 0.09848591685295105, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0164, + "step": 99770 + }, + { + "epoch": 9.069260134520997, + "grad_norm": 0.12649913132190704, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0161, + "step": 99780 + }, + { + "epoch": 9.070169060170878, + "grad_norm": 0.09877830743789673, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0152, + "step": 99790 + }, + { + "epoch": 9.071077985820759, + "grad_norm": 0.09804488718509674, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0166, + "step": 99800 + }, + { + "epoch": 9.071986911470642, + "grad_norm": 0.12504999339580536, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0171, + "step": 99810 + }, + { + "epoch": 9.072895837120523, + "grad_norm": 0.12810733914375305, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0155, + "step": 99820 + }, + { + "epoch": 9.073804762770406, + "grad_norm": 0.2031307816505432, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0155, + "step": 99830 + }, + { + "epoch": 9.074713688420287, + "grad_norm": 0.09308414906263351, + "learning_rate": 2.256186093617204e-06, + "loss": 0.0171, + "step": 99840 + }, + { + "epoch": 9.07562261407017, + "grad_norm": 0.09118961542844772, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0144, + "step": 99850 + }, + { + "epoch": 9.076531539720051, + "grad_norm": 0.10635729879140854, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0161, + "step": 99860 + }, + { + "epoch": 9.077440465369932, + "grad_norm": 0.202515110373497, + "learning_rate": 2.249649559938383e-06, + "loss": 0.0163, + "step": 99870 + }, + { + "epoch": 9.078349391019815, + "grad_norm": 0.13376553356647491, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0136, + "step": 99880 + }, + { + "epoch": 9.079258316669696, + "grad_norm": 0.1378868818283081, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0139, + "step": 99890 + }, + { + "epoch": 9.080167242319579, + "grad_norm": 0.15803766250610352, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0161, + "step": 99900 + }, + { + "epoch": 9.08107616796946, + "grad_norm": 0.1237478107213974, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0151, + "step": 99910 + }, + { + "epoch": 9.081985093619341, + "grad_norm": 0.09852864593267441, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0162, + "step": 99920 + }, + { + "epoch": 9.082894019269224, + "grad_norm": 0.1284945160150528, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0139, + "step": 99930 + }, + { + "epoch": 9.083802944919105, + "grad_norm": 0.12166869640350342, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0148, + "step": 99940 + }, + { + "epoch": 9.084711870568988, + "grad_norm": 0.09491570293903351, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0165, + "step": 99950 + }, + { + "epoch": 9.085620796218869, + "grad_norm": 0.15039318799972534, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0144, + "step": 99960 + }, + { + "epoch": 9.086529721868752, + "grad_norm": 0.23428748548030853, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0171, + "step": 99970 + }, + { + "epoch": 9.087438647518633, + "grad_norm": 0.10473526269197464, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0149, + "step": 99980 + }, + { + "epoch": 9.088347573168514, + "grad_norm": 0.1304069310426712, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0176, + "step": 99990 + }, + { + "epoch": 9.089256498818397, + "grad_norm": 0.24283957481384277, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0159, + "step": 100000 + }, + { + "epoch": 9.090165424468278, + "grad_norm": 0.12468770146369934, + "learning_rate": 2.2192610364286997e-06, + "loss": 0.014, + "step": 100010 + }, + { + "epoch": 9.091074350118161, + "grad_norm": 0.07775309681892395, + "learning_rate": 2.2170977018533865e-06, + "loss": 0.0136, + "step": 100020 + }, + { + "epoch": 9.091983275768042, + "grad_norm": 0.2090291678905487, + "learning_rate": 2.214935338055141e-06, + "loss": 0.0151, + "step": 100030 + }, + { + "epoch": 9.092892201417923, + "grad_norm": 0.13976676762104034, + "learning_rate": 2.2127739451981803e-06, + "loss": 0.0149, + "step": 100040 + }, + { + "epoch": 9.093801127067806, + "grad_norm": 0.0876021459698677, + "learning_rate": 2.2106135234466545e-06, + "loss": 0.0147, + "step": 100050 + }, + { + "epoch": 9.094710052717687, + "grad_norm": 0.24948908388614655, + "learning_rate": 2.2084540729646286e-06, + "loss": 0.0162, + "step": 100060 + }, + { + "epoch": 9.09561897836757, + "grad_norm": 0.12660890817642212, + "learning_rate": 2.206295593916096e-06, + "loss": 0.015, + "step": 100070 + }, + { + "epoch": 9.096527904017451, + "grad_norm": 0.4317459464073181, + "learning_rate": 2.2041380864649814e-06, + "loss": 0.0177, + "step": 100080 + }, + { + "epoch": 9.097436829667334, + "grad_norm": 0.11081935465335846, + "learning_rate": 2.2019815507751377e-06, + "loss": 0.0162, + "step": 100090 + }, + { + "epoch": 9.098345755317215, + "grad_norm": 0.11082558333873749, + "learning_rate": 2.1998259870103347e-06, + "loss": 0.0156, + "step": 100100 + }, + { + "epoch": 9.099254680967096, + "grad_norm": 0.15149283409118652, + "learning_rate": 2.1976713953342683e-06, + "loss": 0.0155, + "step": 100110 + }, + { + "epoch": 9.100163606616979, + "grad_norm": 0.15038351714611053, + "learning_rate": 2.1955177759105784e-06, + "loss": 0.0151, + "step": 100120 + }, + { + "epoch": 9.10107253226686, + "grad_norm": 0.14488960802555084, + "learning_rate": 2.193365128902811e-06, + "loss": 0.0142, + "step": 100130 + }, + { + "epoch": 9.101981457916743, + "grad_norm": 0.10400039702653885, + "learning_rate": 2.191213454474441e-06, + "loss": 0.0158, + "step": 100140 + }, + { + "epoch": 9.102890383566624, + "grad_norm": 0.12885814905166626, + "learning_rate": 2.1890627527888807e-06, + "loss": 0.0172, + "step": 100150 + }, + { + "epoch": 9.103799309216505, + "grad_norm": 0.10238434374332428, + "learning_rate": 2.186913024009461e-06, + "loss": 0.0165, + "step": 100160 + }, + { + "epoch": 9.104708234866388, + "grad_norm": 0.13220064342021942, + "learning_rate": 2.184764268299436e-06, + "loss": 0.0159, + "step": 100170 + }, + { + "epoch": 9.10561716051627, + "grad_norm": 0.1555153876543045, + "learning_rate": 2.1826164858219943e-06, + "loss": 0.0141, + "step": 100180 + }, + { + "epoch": 9.106526086166152, + "grad_norm": 0.12758073210716248, + "learning_rate": 2.1804696767402397e-06, + "loss": 0.0154, + "step": 100190 + }, + { + "epoch": 9.107435011816033, + "grad_norm": 0.11393333971500397, + "learning_rate": 2.1783238412172156e-06, + "loss": 0.0167, + "step": 100200 + }, + { + "epoch": 9.108343937465916, + "grad_norm": 0.14742153882980347, + "learning_rate": 2.1761789794158763e-06, + "loss": 0.0217, + "step": 100210 + }, + { + "epoch": 9.109252863115797, + "grad_norm": 0.09975715726613998, + "learning_rate": 2.1740350914991135e-06, + "loss": 0.0156, + "step": 100220 + }, + { + "epoch": 9.110161788765678, + "grad_norm": 0.14597854018211365, + "learning_rate": 2.171892177629745e-06, + "loss": 0.0156, + "step": 100230 + }, + { + "epoch": 9.111070714415561, + "grad_norm": 0.10305413603782654, + "learning_rate": 2.1697502379705043e-06, + "loss": 0.0151, + "step": 100240 + }, + { + "epoch": 9.111979640065442, + "grad_norm": 0.13867847621440887, + "learning_rate": 2.167609272684064e-06, + "loss": 0.0151, + "step": 100250 + }, + { + "epoch": 9.112888565715325, + "grad_norm": 0.11857102066278458, + "learning_rate": 2.1654692819330095e-06, + "loss": 0.0173, + "step": 100260 + }, + { + "epoch": 9.113797491365206, + "grad_norm": 0.12438924610614777, + "learning_rate": 2.1633302658798628e-06, + "loss": 0.0141, + "step": 100270 + }, + { + "epoch": 9.114706417015087, + "grad_norm": 0.10506460815668106, + "learning_rate": 2.16119222468707e-06, + "loss": 0.0164, + "step": 100280 + }, + { + "epoch": 9.11561534266497, + "grad_norm": 0.1163957417011261, + "learning_rate": 2.1590551585169992e-06, + "loss": 0.0134, + "step": 100290 + }, + { + "epoch": 9.116524268314851, + "grad_norm": 0.14385159313678741, + "learning_rate": 2.156919067531945e-06, + "loss": 0.0164, + "step": 100300 + }, + { + "epoch": 9.117433193964734, + "grad_norm": 0.15229631960391998, + "learning_rate": 2.154783951894129e-06, + "loss": 0.0152, + "step": 100310 + }, + { + "epoch": 9.118342119614615, + "grad_norm": 0.17126934230327606, + "learning_rate": 2.152649811765704e-06, + "loss": 0.0174, + "step": 100320 + }, + { + "epoch": 9.119251045264498, + "grad_norm": 0.11640433967113495, + "learning_rate": 2.150516647308738e-06, + "loss": 0.0157, + "step": 100330 + }, + { + "epoch": 9.12015997091438, + "grad_norm": 0.24817313253879547, + "learning_rate": 2.1483844586852325e-06, + "loss": 0.0155, + "step": 100340 + }, + { + "epoch": 9.12106889656426, + "grad_norm": 0.14169898629188538, + "learning_rate": 2.146253246057118e-06, + "loss": 0.0146, + "step": 100350 + }, + { + "epoch": 9.121977822214143, + "grad_norm": 0.12542259693145752, + "learning_rate": 2.1441230095862424e-06, + "loss": 0.0161, + "step": 100360 + }, + { + "epoch": 9.122886747864024, + "grad_norm": 0.1337292641401291, + "learning_rate": 2.141993749434379e-06, + "loss": 0.0133, + "step": 100370 + }, + { + "epoch": 9.123795673513907, + "grad_norm": 0.15602943301200867, + "learning_rate": 2.1398654657632347e-06, + "loss": 0.0156, + "step": 100380 + }, + { + "epoch": 9.124704599163788, + "grad_norm": 0.1253301203250885, + "learning_rate": 2.137738158734443e-06, + "loss": 0.0146, + "step": 100390 + }, + { + "epoch": 9.12561352481367, + "grad_norm": 0.18248580396175385, + "learning_rate": 2.135611828509554e-06, + "loss": 0.018, + "step": 100400 + }, + { + "epoch": 9.126522450463552, + "grad_norm": 0.15587863326072693, + "learning_rate": 2.1334864752500434e-06, + "loss": 0.017, + "step": 100410 + }, + { + "epoch": 9.127431376113433, + "grad_norm": 0.13812553882598877, + "learning_rate": 2.1313620991173288e-06, + "loss": 0.0176, + "step": 100420 + }, + { + "epoch": 9.128340301763316, + "grad_norm": 0.13010160624980927, + "learning_rate": 2.1292387002727383e-06, + "loss": 0.0163, + "step": 100430 + }, + { + "epoch": 9.129249227413197, + "grad_norm": 0.14874184131622314, + "learning_rate": 2.127116278877524e-06, + "loss": 0.018, + "step": 100440 + }, + { + "epoch": 9.13015815306308, + "grad_norm": 0.22512081265449524, + "learning_rate": 2.1249948350928755e-06, + "loss": 0.0176, + "step": 100450 + }, + { + "epoch": 9.131067078712961, + "grad_norm": 0.1256115585565567, + "learning_rate": 2.122874369079904e-06, + "loss": 0.0161, + "step": 100460 + }, + { + "epoch": 9.131976004362842, + "grad_norm": 0.09660790115594864, + "learning_rate": 2.120754880999644e-06, + "loss": 0.0159, + "step": 100470 + }, + { + "epoch": 9.132884930012725, + "grad_norm": 0.12971767783164978, + "learning_rate": 2.1186363710130518e-06, + "loss": 0.0159, + "step": 100480 + }, + { + "epoch": 9.133793855662606, + "grad_norm": 0.16039296984672546, + "learning_rate": 2.116518839281017e-06, + "loss": 0.0164, + "step": 100490 + }, + { + "epoch": 9.13470278131249, + "grad_norm": 0.09003417193889618, + "learning_rate": 2.1144022859643557e-06, + "loss": 0.0148, + "step": 100500 + }, + { + "epoch": 9.13561170696237, + "grad_norm": 0.12444009631872177, + "learning_rate": 2.1122867112238016e-06, + "loss": 0.0163, + "step": 100510 + }, + { + "epoch": 9.136520632612251, + "grad_norm": 0.18040265142917633, + "learning_rate": 2.110172115220019e-06, + "loss": 0.0166, + "step": 100520 + }, + { + "epoch": 9.137429558262134, + "grad_norm": 0.1127294972538948, + "learning_rate": 2.108058498113601e-06, + "loss": 0.0148, + "step": 100530 + }, + { + "epoch": 9.138338483912015, + "grad_norm": 0.18617232143878937, + "learning_rate": 2.1059458600650593e-06, + "loss": 0.0173, + "step": 100540 + }, + { + "epoch": 9.139247409561898, + "grad_norm": 0.10808227211236954, + "learning_rate": 2.1038342012348394e-06, + "loss": 0.0155, + "step": 100550 + }, + { + "epoch": 9.14015633521178, + "grad_norm": 0.11133318394422531, + "learning_rate": 2.101723521783299e-06, + "loss": 0.0177, + "step": 100560 + }, + { + "epoch": 9.141065260861662, + "grad_norm": 0.12626822292804718, + "learning_rate": 2.099613821870742e-06, + "loss": 0.0141, + "step": 100570 + }, + { + "epoch": 9.141974186511543, + "grad_norm": 0.10933918505907059, + "learning_rate": 2.0975051016573755e-06, + "loss": 0.0173, + "step": 100580 + }, + { + "epoch": 9.142883112161424, + "grad_norm": 0.1604907214641571, + "learning_rate": 2.0953973613033504e-06, + "loss": 0.0191, + "step": 100590 + }, + { + "epoch": 9.143792037811307, + "grad_norm": 0.195362851023674, + "learning_rate": 2.0932906009687297e-06, + "loss": 0.0145, + "step": 100600 + }, + { + "epoch": 9.144700963461188, + "grad_norm": 0.10199140757322311, + "learning_rate": 2.0911848208135116e-06, + "loss": 0.0143, + "step": 100610 + }, + { + "epoch": 9.145609889111071, + "grad_norm": 0.14475271105766296, + "learning_rate": 2.0890800209976197e-06, + "loss": 0.0177, + "step": 100620 + }, + { + "epoch": 9.146518814760952, + "grad_norm": 0.12359105795621872, + "learning_rate": 2.0869762016808896e-06, + "loss": 0.0159, + "step": 100630 + }, + { + "epoch": 9.147427740410834, + "grad_norm": 0.13137653470039368, + "learning_rate": 2.0848733630231005e-06, + "loss": 0.0147, + "step": 100640 + }, + { + "epoch": 9.148336666060716, + "grad_norm": 0.12119317799806595, + "learning_rate": 2.082771505183949e-06, + "loss": 0.0138, + "step": 100650 + }, + { + "epoch": 9.149245591710597, + "grad_norm": 0.1713821440935135, + "learning_rate": 2.0806706283230563e-06, + "loss": 0.0162, + "step": 100660 + }, + { + "epoch": 9.15015451736048, + "grad_norm": 0.10992203652858734, + "learning_rate": 2.0785707325999666e-06, + "loss": 0.0132, + "step": 100670 + }, + { + "epoch": 9.151063443010361, + "grad_norm": 0.1580643206834793, + "learning_rate": 2.076471818174155e-06, + "loss": 0.0154, + "step": 100680 + }, + { + "epoch": 9.151972368660244, + "grad_norm": 0.11746852099895477, + "learning_rate": 2.0743738852050243e-06, + "loss": 0.0152, + "step": 100690 + }, + { + "epoch": 9.152881294310125, + "grad_norm": 0.17482030391693115, + "learning_rate": 2.0722769338518955e-06, + "loss": 0.0151, + "step": 100700 + }, + { + "epoch": 9.153790219960007, + "grad_norm": 0.1133376806974411, + "learning_rate": 2.070180964274011e-06, + "loss": 0.0156, + "step": 100710 + }, + { + "epoch": 9.15469914560989, + "grad_norm": 0.15447424352169037, + "learning_rate": 2.068085976630559e-06, + "loss": 0.0174, + "step": 100720 + }, + { + "epoch": 9.15560807125977, + "grad_norm": 0.1338167041540146, + "learning_rate": 2.0659919710806346e-06, + "loss": 0.0143, + "step": 100730 + }, + { + "epoch": 9.156516996909653, + "grad_norm": 0.11009819805622101, + "learning_rate": 2.0638989477832598e-06, + "loss": 0.0144, + "step": 100740 + }, + { + "epoch": 9.157425922559534, + "grad_norm": 0.2177838236093521, + "learning_rate": 2.061806906897389e-06, + "loss": 0.0171, + "step": 100750 + }, + { + "epoch": 9.158334848209417, + "grad_norm": 0.1193116307258606, + "learning_rate": 2.0597158485819013e-06, + "loss": 0.0155, + "step": 100760 + }, + { + "epoch": 9.159243773859298, + "grad_norm": 0.09486235678195953, + "learning_rate": 2.057625772995598e-06, + "loss": 0.0168, + "step": 100770 + }, + { + "epoch": 9.16015269950918, + "grad_norm": 0.15646454691886902, + "learning_rate": 2.0555366802972015e-06, + "loss": 0.0158, + "step": 100780 + }, + { + "epoch": 9.161061625159062, + "grad_norm": 0.31658706068992615, + "learning_rate": 2.053448570645367e-06, + "loss": 0.0158, + "step": 100790 + }, + { + "epoch": 9.161970550808944, + "grad_norm": 0.14627327024936676, + "learning_rate": 2.0513614441986765e-06, + "loss": 0.0166, + "step": 100800 + }, + { + "epoch": 9.162879476458826, + "grad_norm": 0.12719707190990448, + "learning_rate": 2.0492753011156303e-06, + "loss": 0.0157, + "step": 100810 + }, + { + "epoch": 9.163788402108707, + "grad_norm": 0.25771644711494446, + "learning_rate": 2.0471901415546556e-06, + "loss": 0.0155, + "step": 100820 + }, + { + "epoch": 9.164697327758589, + "grad_norm": 0.18493403494358063, + "learning_rate": 2.0451059656741116e-06, + "loss": 0.0146, + "step": 100830 + }, + { + "epoch": 9.165606253408471, + "grad_norm": 0.13729895651340485, + "learning_rate": 2.043022773632275e-06, + "loss": 0.0154, + "step": 100840 + }, + { + "epoch": 9.166515179058353, + "grad_norm": 0.16733697056770325, + "learning_rate": 2.040940565587347e-06, + "loss": 0.0153, + "step": 100850 + }, + { + "epoch": 9.167424104708235, + "grad_norm": 0.09600909054279327, + "learning_rate": 2.0388593416974616e-06, + "loss": 0.0146, + "step": 100860 + }, + { + "epoch": 9.168333030358117, + "grad_norm": 0.10944870859384537, + "learning_rate": 2.0367791021206754e-06, + "loss": 0.0162, + "step": 100870 + }, + { + "epoch": 9.169241956008, + "grad_norm": 0.13304011523723602, + "learning_rate": 2.0346998470149626e-06, + "loss": 0.0173, + "step": 100880 + }, + { + "epoch": 9.17015088165788, + "grad_norm": 0.15082287788391113, + "learning_rate": 2.032621576538238e-06, + "loss": 0.0137, + "step": 100890 + }, + { + "epoch": 9.171059807307762, + "grad_norm": 0.1537683755159378, + "learning_rate": 2.030544290848322e-06, + "loss": 0.0182, + "step": 100900 + }, + { + "epoch": 9.171968732957644, + "grad_norm": 0.17702338099479675, + "learning_rate": 2.0284679901029766e-06, + "loss": 0.0168, + "step": 100910 + }, + { + "epoch": 9.172877658607526, + "grad_norm": 0.11681672930717468, + "learning_rate": 2.026392674459888e-06, + "loss": 0.0158, + "step": 100920 + }, + { + "epoch": 9.173786584257408, + "grad_norm": 0.1472407877445221, + "learning_rate": 2.0243183440766517e-06, + "loss": 0.0138, + "step": 100930 + }, + { + "epoch": 9.17469550990729, + "grad_norm": 0.16805589199066162, + "learning_rate": 2.0222449991108094e-06, + "loss": 0.0168, + "step": 100940 + }, + { + "epoch": 9.17560443555717, + "grad_norm": 0.16392965614795685, + "learning_rate": 2.0201726397198104e-06, + "loss": 0.0135, + "step": 100950 + }, + { + "epoch": 9.176513361207054, + "grad_norm": 0.13363254070281982, + "learning_rate": 2.0181012660610443e-06, + "loss": 0.0176, + "step": 100960 + }, + { + "epoch": 9.177422286856935, + "grad_norm": 0.1538945883512497, + "learning_rate": 2.0160308782918095e-06, + "loss": 0.0183, + "step": 100970 + }, + { + "epoch": 9.178331212506817, + "grad_norm": 0.11625348776578903, + "learning_rate": 2.0139614765693426e-06, + "loss": 0.0147, + "step": 100980 + }, + { + "epoch": 9.179240138156699, + "grad_norm": 0.16569016873836517, + "learning_rate": 2.0118930610508058e-06, + "loss": 0.0176, + "step": 100990 + }, + { + "epoch": 9.180149063806581, + "grad_norm": 0.1113751232624054, + "learning_rate": 2.009825631893275e-06, + "loss": 0.0148, + "step": 101000 + }, + { + "epoch": 9.181057989456463, + "grad_norm": 0.10713560879230499, + "learning_rate": 2.007759189253755e-06, + "loss": 0.0172, + "step": 101010 + }, + { + "epoch": 9.181966915106344, + "grad_norm": 0.09689561277627945, + "learning_rate": 2.0056937332891882e-06, + "loss": 0.0148, + "step": 101020 + }, + { + "epoch": 9.182875840756227, + "grad_norm": 0.13115926086902618, + "learning_rate": 2.003629264156428e-06, + "loss": 0.0157, + "step": 101030 + }, + { + "epoch": 9.183784766406108, + "grad_norm": 0.12510579824447632, + "learning_rate": 2.0015657820122534e-06, + "loss": 0.0142, + "step": 101040 + }, + { + "epoch": 9.18469369205599, + "grad_norm": 0.1513090431690216, + "learning_rate": 1.999503287013376e-06, + "loss": 0.0143, + "step": 101050 + }, + { + "epoch": 9.185602617705872, + "grad_norm": 0.16189692914485931, + "learning_rate": 1.9974417793164305e-06, + "loss": 0.0162, + "step": 101060 + }, + { + "epoch": 9.186511543355753, + "grad_norm": 0.14454568922519684, + "learning_rate": 1.9953812590779735e-06, + "loss": 0.0195, + "step": 101070 + }, + { + "epoch": 9.187420469005636, + "grad_norm": 0.13460738956928253, + "learning_rate": 1.993321726454483e-06, + "loss": 0.0143, + "step": 101080 + }, + { + "epoch": 9.188329394655517, + "grad_norm": 0.29291826486587524, + "learning_rate": 1.9912631816023718e-06, + "loss": 0.015, + "step": 101090 + }, + { + "epoch": 9.1892383203054, + "grad_norm": 0.1378566175699234, + "learning_rate": 1.9892056246779737e-06, + "loss": 0.0166, + "step": 101100 + }, + { + "epoch": 9.19014724595528, + "grad_norm": 0.14038793742656708, + "learning_rate": 1.9871490558375404e-06, + "loss": 0.016, + "step": 101110 + }, + { + "epoch": 9.191056171605164, + "grad_norm": 0.2164400964975357, + "learning_rate": 1.985093475237262e-06, + "loss": 0.0166, + "step": 101120 + }, + { + "epoch": 9.191965097255045, + "grad_norm": 0.14279945194721222, + "learning_rate": 1.983038883033245e-06, + "loss": 0.0153, + "step": 101130 + }, + { + "epoch": 9.192874022904926, + "grad_norm": 0.11092863976955414, + "learning_rate": 1.980985279381519e-06, + "loss": 0.0145, + "step": 101140 + }, + { + "epoch": 9.193782948554809, + "grad_norm": 0.1255856305360794, + "learning_rate": 1.9789326644380407e-06, + "loss": 0.0142, + "step": 101150 + }, + { + "epoch": 9.19469187420469, + "grad_norm": 0.16085490584373474, + "learning_rate": 1.976881038358696e-06, + "loss": 0.0152, + "step": 101160 + }, + { + "epoch": 9.195600799854573, + "grad_norm": 0.09465985745191574, + "learning_rate": 1.9748304012992925e-06, + "loss": 0.0146, + "step": 101170 + }, + { + "epoch": 9.196509725504454, + "grad_norm": 0.10691724717617035, + "learning_rate": 1.9727807534155602e-06, + "loss": 0.0157, + "step": 101180 + }, + { + "epoch": 9.197418651154335, + "grad_norm": 0.17819687724113464, + "learning_rate": 1.97073209486316e-06, + "loss": 0.0157, + "step": 101190 + }, + { + "epoch": 9.198327576804218, + "grad_norm": 0.10285012423992157, + "learning_rate": 1.9686844257976683e-06, + "loss": 0.0149, + "step": 101200 + }, + { + "epoch": 9.199236502454099, + "grad_norm": 0.12233617156744003, + "learning_rate": 1.966637746374598e-06, + "loss": 0.0174, + "step": 101210 + }, + { + "epoch": 9.200145428103982, + "grad_norm": 0.1343059092760086, + "learning_rate": 1.9645920567493753e-06, + "loss": 0.0155, + "step": 101220 + }, + { + "epoch": 9.201054353753863, + "grad_norm": 0.1435554474592209, + "learning_rate": 1.96254735707736e-06, + "loss": 0.0141, + "step": 101230 + }, + { + "epoch": 9.201963279403746, + "grad_norm": 0.13635686039924622, + "learning_rate": 1.9605036475138368e-06, + "loss": 0.0145, + "step": 101240 + }, + { + "epoch": 9.202872205053627, + "grad_norm": 0.11400561034679413, + "learning_rate": 1.9584609282140043e-06, + "loss": 0.016, + "step": 101250 + }, + { + "epoch": 9.203781130703508, + "grad_norm": 0.20712392032146454, + "learning_rate": 1.9564191993330003e-06, + "loss": 0.0157, + "step": 101260 + }, + { + "epoch": 9.20469005635339, + "grad_norm": 0.17191749811172485, + "learning_rate": 1.9543784610258764e-06, + "loss": 0.0141, + "step": 101270 + }, + { + "epoch": 9.205598982003272, + "grad_norm": 0.1274002492427826, + "learning_rate": 1.952338713447613e-06, + "loss": 0.0154, + "step": 101280 + }, + { + "epoch": 9.206507907653155, + "grad_norm": 0.1327614039182663, + "learning_rate": 1.950299956753121e-06, + "loss": 0.0172, + "step": 101290 + }, + { + "epoch": 9.207416833303036, + "grad_norm": 0.12308287620544434, + "learning_rate": 1.948262191097227e-06, + "loss": 0.0167, + "step": 101300 + }, + { + "epoch": 9.208325758952917, + "grad_norm": 0.25413277745246887, + "learning_rate": 1.946225416634684e-06, + "loss": 0.0157, + "step": 101310 + }, + { + "epoch": 9.2092346846028, + "grad_norm": 0.09484699368476868, + "learning_rate": 1.944189633520171e-06, + "loss": 0.0166, + "step": 101320 + }, + { + "epoch": 9.21014361025268, + "grad_norm": 0.12064298242330551, + "learning_rate": 1.9421548419082978e-06, + "loss": 0.0164, + "step": 101330 + }, + { + "epoch": 9.211052535902564, + "grad_norm": 0.14305953681468964, + "learning_rate": 1.9401210419535883e-06, + "loss": 0.0161, + "step": 101340 + }, + { + "epoch": 9.211961461552445, + "grad_norm": 0.10353401303291321, + "learning_rate": 1.938088233810496e-06, + "loss": 0.0133, + "step": 101350 + }, + { + "epoch": 9.212870387202328, + "grad_norm": 0.1206827238202095, + "learning_rate": 1.9360564176334068e-06, + "loss": 0.0155, + "step": 101360 + }, + { + "epoch": 9.213779312852209, + "grad_norm": 0.11973452568054199, + "learning_rate": 1.9340255935766164e-06, + "loss": 0.0171, + "step": 101370 + }, + { + "epoch": 9.21468823850209, + "grad_norm": 0.13149400055408478, + "learning_rate": 1.9319957617943505e-06, + "loss": 0.0146, + "step": 101380 + }, + { + "epoch": 9.215597164151973, + "grad_norm": 0.17424385249614716, + "learning_rate": 1.9299669224407674e-06, + "loss": 0.0131, + "step": 101390 + }, + { + "epoch": 9.216506089801854, + "grad_norm": 0.12932215631008148, + "learning_rate": 1.9279390756699435e-06, + "loss": 0.0174, + "step": 101400 + }, + { + "epoch": 9.217415015451737, + "grad_norm": 0.14248088002204895, + "learning_rate": 1.925912221635879e-06, + "loss": 0.0155, + "step": 101410 + }, + { + "epoch": 9.218323941101618, + "grad_norm": 0.11421097069978714, + "learning_rate": 1.9238863604924937e-06, + "loss": 0.0142, + "step": 101420 + }, + { + "epoch": 9.219232866751499, + "grad_norm": 0.13639505207538605, + "learning_rate": 1.9218614923936508e-06, + "loss": 0.0153, + "step": 101430 + }, + { + "epoch": 9.220141792401382, + "grad_norm": 0.12037285417318344, + "learning_rate": 1.919837617493119e-06, + "loss": 0.0163, + "step": 101440 + }, + { + "epoch": 9.221050718051263, + "grad_norm": 0.30286577343940735, + "learning_rate": 1.9178147359445953e-06, + "loss": 0.0161, + "step": 101450 + }, + { + "epoch": 9.221959643701146, + "grad_norm": 0.14265869557857513, + "learning_rate": 1.915792847901707e-06, + "loss": 0.0153, + "step": 101460 + }, + { + "epoch": 9.222868569351027, + "grad_norm": 0.10573901981115341, + "learning_rate": 1.9137719535180077e-06, + "loss": 0.0158, + "step": 101470 + }, + { + "epoch": 9.22377749500091, + "grad_norm": 0.1286465972661972, + "learning_rate": 1.9117520529469652e-06, + "loss": 0.0164, + "step": 101480 + }, + { + "epoch": 9.22468642065079, + "grad_norm": 0.1292559802532196, + "learning_rate": 1.9097331463419805e-06, + "loss": 0.0148, + "step": 101490 + }, + { + "epoch": 9.225595346300672, + "grad_norm": 0.10659051686525345, + "learning_rate": 1.9077152338563724e-06, + "loss": 0.0164, + "step": 101500 + }, + { + "epoch": 9.226504271950555, + "grad_norm": 0.2398325800895691, + "learning_rate": 1.9056983156433939e-06, + "loss": 0.0162, + "step": 101510 + }, + { + "epoch": 9.227413197600436, + "grad_norm": 0.14623603224754333, + "learning_rate": 1.9036823918562102e-06, + "loss": 0.0184, + "step": 101520 + }, + { + "epoch": 9.228322123250319, + "grad_norm": 0.09766720980405807, + "learning_rate": 1.9016674626479208e-06, + "loss": 0.0167, + "step": 101530 + }, + { + "epoch": 9.2292310489002, + "grad_norm": 0.10634519159793854, + "learning_rate": 1.8996535281715483e-06, + "loss": 0.0152, + "step": 101540 + }, + { + "epoch": 9.230139974550081, + "grad_norm": 0.23467694222927094, + "learning_rate": 1.8976405885800335e-06, + "loss": 0.0167, + "step": 101550 + }, + { + "epoch": 9.231048900199964, + "grad_norm": 0.13136878609657288, + "learning_rate": 1.895628644026251e-06, + "loss": 0.016, + "step": 101560 + }, + { + "epoch": 9.231957825849845, + "grad_norm": 0.12666216492652893, + "learning_rate": 1.8936176946629886e-06, + "loss": 0.0161, + "step": 101570 + }, + { + "epoch": 9.232866751499728, + "grad_norm": 0.21113060414791107, + "learning_rate": 1.8916077406429706e-06, + "loss": 0.0188, + "step": 101580 + }, + { + "epoch": 9.233775677149609, + "grad_norm": 0.13610997796058655, + "learning_rate": 1.8895987821188348e-06, + "loss": 0.0166, + "step": 101590 + }, + { + "epoch": 9.234684602799492, + "grad_norm": 0.07917439937591553, + "learning_rate": 1.8875908192431512e-06, + "loss": 0.0157, + "step": 101600 + }, + { + "epoch": 9.235593528449373, + "grad_norm": 0.08753815293312073, + "learning_rate": 1.8855838521684093e-06, + "loss": 0.0155, + "step": 101610 + }, + { + "epoch": 9.236502454099254, + "grad_norm": 0.10839856415987015, + "learning_rate": 1.8835778810470255e-06, + "loss": 0.0152, + "step": 101620 + }, + { + "epoch": 9.237411379749137, + "grad_norm": 0.1380133330821991, + "learning_rate": 1.8815729060313448e-06, + "loss": 0.015, + "step": 101630 + }, + { + "epoch": 9.238320305399018, + "grad_norm": 0.1546178162097931, + "learning_rate": 1.8795689272736232e-06, + "loss": 0.0177, + "step": 101640 + }, + { + "epoch": 9.2392292310489, + "grad_norm": 0.10869043320417404, + "learning_rate": 1.8775659449260545e-06, + "loss": 0.019, + "step": 101650 + }, + { + "epoch": 9.240138156698782, + "grad_norm": 0.14574943482875824, + "learning_rate": 1.8755639591407554e-06, + "loss": 0.0154, + "step": 101660 + }, + { + "epoch": 9.241047082348663, + "grad_norm": 0.13536983728408813, + "learning_rate": 1.8735629700697604e-06, + "loss": 0.0141, + "step": 101670 + }, + { + "epoch": 9.241956007998546, + "grad_norm": 0.13279567658901215, + "learning_rate": 1.8715629778650267e-06, + "loss": 0.0161, + "step": 101680 + }, + { + "epoch": 9.242864933648427, + "grad_norm": 0.2135138213634491, + "learning_rate": 1.8695639826784461e-06, + "loss": 0.0171, + "step": 101690 + }, + { + "epoch": 9.24377385929831, + "grad_norm": 0.16047167778015137, + "learning_rate": 1.8675659846618287e-06, + "loss": 0.0156, + "step": 101700 + }, + { + "epoch": 9.244682784948191, + "grad_norm": 0.10387624055147171, + "learning_rate": 1.86556898396691e-06, + "loss": 0.0149, + "step": 101710 + }, + { + "epoch": 9.245591710598074, + "grad_norm": 0.149133563041687, + "learning_rate": 1.86357298074534e-06, + "loss": 0.018, + "step": 101720 + }, + { + "epoch": 9.246500636247955, + "grad_norm": 0.1698974370956421, + "learning_rate": 1.861577975148716e-06, + "loss": 0.0154, + "step": 101730 + }, + { + "epoch": 9.247409561897836, + "grad_norm": 0.09381745010614395, + "learning_rate": 1.8595839673285386e-06, + "loss": 0.0162, + "step": 101740 + }, + { + "epoch": 9.248318487547719, + "grad_norm": 0.13626410067081451, + "learning_rate": 1.8575909574362365e-06, + "loss": 0.0168, + "step": 101750 + }, + { + "epoch": 9.2492274131976, + "grad_norm": 0.13364355266094208, + "learning_rate": 1.8555989456231687e-06, + "loss": 0.0172, + "step": 101760 + }, + { + "epoch": 9.250136338847483, + "grad_norm": 0.13544948399066925, + "learning_rate": 1.8536079320406191e-06, + "loss": 0.0168, + "step": 101770 + }, + { + "epoch": 9.251045264497364, + "grad_norm": 0.11217780411243439, + "learning_rate": 1.8516179168397885e-06, + "loss": 0.0169, + "step": 101780 + }, + { + "epoch": 9.251954190147245, + "grad_norm": 0.12050264328718185, + "learning_rate": 1.8496289001718025e-06, + "loss": 0.0154, + "step": 101790 + }, + { + "epoch": 9.252863115797128, + "grad_norm": 0.1350516676902771, + "learning_rate": 1.8476408821877156e-06, + "loss": 0.0197, + "step": 101800 + }, + { + "epoch": 9.253772041447009, + "grad_norm": 0.10374617576599121, + "learning_rate": 1.8456538630385084e-06, + "loss": 0.0144, + "step": 101810 + }, + { + "epoch": 9.254680967096892, + "grad_norm": 0.10869985818862915, + "learning_rate": 1.843667842875077e-06, + "loss": 0.0166, + "step": 101820 + }, + { + "epoch": 9.255589892746773, + "grad_norm": 0.1525564044713974, + "learning_rate": 1.841682821848249e-06, + "loss": 0.017, + "step": 101830 + }, + { + "epoch": 9.256498818396656, + "grad_norm": 0.12146679311990738, + "learning_rate": 1.8396988001087756e-06, + "loss": 0.0169, + "step": 101840 + }, + { + "epoch": 9.257407744046537, + "grad_norm": 0.15457308292388916, + "learning_rate": 1.8377157778073261e-06, + "loss": 0.0141, + "step": 101850 + }, + { + "epoch": 9.258316669696418, + "grad_norm": 0.13292798399925232, + "learning_rate": 1.835733755094502e-06, + "loss": 0.0147, + "step": 101860 + }, + { + "epoch": 9.259225595346301, + "grad_norm": 0.14519420266151428, + "learning_rate": 1.8337527321208193e-06, + "loss": 0.0178, + "step": 101870 + }, + { + "epoch": 9.260134520996182, + "grad_norm": 0.11717405915260315, + "learning_rate": 1.8317727090367292e-06, + "loss": 0.0146, + "step": 101880 + }, + { + "epoch": 9.261043446646065, + "grad_norm": 0.11169310659170151, + "learning_rate": 1.8297936859925984e-06, + "loss": 0.015, + "step": 101890 + }, + { + "epoch": 9.261952372295946, + "grad_norm": 0.1292433887720108, + "learning_rate": 1.8278156631387216e-06, + "loss": 0.0172, + "step": 101900 + }, + { + "epoch": 9.262861297945829, + "grad_norm": 0.17100276052951813, + "learning_rate": 1.8258386406253152e-06, + "loss": 0.0146, + "step": 101910 + }, + { + "epoch": 9.26377022359571, + "grad_norm": 0.1168254092335701, + "learning_rate": 1.8238626186025226e-06, + "loss": 0.0163, + "step": 101920 + }, + { + "epoch": 9.264679149245591, + "grad_norm": 0.12617364525794983, + "learning_rate": 1.8218875972204101e-06, + "loss": 0.016, + "step": 101930 + }, + { + "epoch": 9.265588074895474, + "grad_norm": 0.1393294781446457, + "learning_rate": 1.8199135766289648e-06, + "loss": 0.0149, + "step": 101940 + }, + { + "epoch": 9.266497000545355, + "grad_norm": 0.10853135585784912, + "learning_rate": 1.8179405569781032e-06, + "loss": 0.015, + "step": 101950 + }, + { + "epoch": 9.267405926195238, + "grad_norm": 0.1165776327252388, + "learning_rate": 1.8159685384176638e-06, + "loss": 0.0155, + "step": 101960 + }, + { + "epoch": 9.268314851845119, + "grad_norm": 0.1251818984746933, + "learning_rate": 1.813997521097407e-06, + "loss": 0.0164, + "step": 101970 + }, + { + "epoch": 9.269223777495, + "grad_norm": 0.1540914624929428, + "learning_rate": 1.812027505167016e-06, + "loss": 0.016, + "step": 101980 + }, + { + "epoch": 9.270132703144883, + "grad_norm": 0.16660626232624054, + "learning_rate": 1.810058490776103e-06, + "loss": 0.0158, + "step": 101990 + }, + { + "epoch": 9.271041628794764, + "grad_norm": 0.1166481077671051, + "learning_rate": 1.8080904780742052e-06, + "loss": 0.0153, + "step": 102000 + }, + { + "epoch": 9.271950554444647, + "grad_norm": 0.15226687490940094, + "learning_rate": 1.8061234672107747e-06, + "loss": 0.017, + "step": 102010 + }, + { + "epoch": 9.272859480094528, + "grad_norm": 0.14561963081359863, + "learning_rate": 1.80415745833519e-06, + "loss": 0.0167, + "step": 102020 + }, + { + "epoch": 9.27376840574441, + "grad_norm": 0.1320546269416809, + "learning_rate": 1.802192451596767e-06, + "loss": 0.0177, + "step": 102030 + }, + { + "epoch": 9.274677331394292, + "grad_norm": 0.11319838464260101, + "learning_rate": 1.8002284471447311e-06, + "loss": 0.0186, + "step": 102040 + }, + { + "epoch": 9.275586257044173, + "grad_norm": 0.13547033071517944, + "learning_rate": 1.7982654451282282e-06, + "loss": 0.0148, + "step": 102050 + }, + { + "epoch": 9.276495182694056, + "grad_norm": 0.11675385385751724, + "learning_rate": 1.7963034456963424e-06, + "loss": 0.0162, + "step": 102060 + }, + { + "epoch": 9.277404108343937, + "grad_norm": 0.11893518269062042, + "learning_rate": 1.7943424489980758e-06, + "loss": 0.018, + "step": 102070 + }, + { + "epoch": 9.27831303399382, + "grad_norm": 0.1620035618543625, + "learning_rate": 1.7923824551823499e-06, + "loss": 0.0148, + "step": 102080 + }, + { + "epoch": 9.279221959643701, + "grad_norm": 0.16650737822055817, + "learning_rate": 1.7904234643980117e-06, + "loss": 0.0174, + "step": 102090 + }, + { + "epoch": 9.280130885293582, + "grad_norm": 0.11555292457342148, + "learning_rate": 1.788465476793834e-06, + "loss": 0.016, + "step": 102100 + }, + { + "epoch": 9.281039810943465, + "grad_norm": 0.12679874897003174, + "learning_rate": 1.78650849251852e-06, + "loss": 0.0171, + "step": 102110 + }, + { + "epoch": 9.281948736593346, + "grad_norm": 0.25357338786125183, + "learning_rate": 1.7845525117206802e-06, + "loss": 0.0151, + "step": 102120 + }, + { + "epoch": 9.282857662243229, + "grad_norm": 0.11334002763032913, + "learning_rate": 1.7825975345488626e-06, + "loss": 0.0145, + "step": 102130 + }, + { + "epoch": 9.28376658789311, + "grad_norm": 0.11606865376234055, + "learning_rate": 1.7806435611515387e-06, + "loss": 0.0163, + "step": 102140 + }, + { + "epoch": 9.284675513542993, + "grad_norm": 0.15607528388500214, + "learning_rate": 1.7786905916770962e-06, + "loss": 0.0143, + "step": 102150 + }, + { + "epoch": 9.285584439192874, + "grad_norm": 0.11806595325469971, + "learning_rate": 1.7767386262738483e-06, + "loss": 0.0157, + "step": 102160 + }, + { + "epoch": 9.286493364842755, + "grad_norm": 0.1312676966190338, + "learning_rate": 1.774787665090033e-06, + "loss": 0.0152, + "step": 102170 + }, + { + "epoch": 9.287402290492638, + "grad_norm": 0.10163141041994095, + "learning_rate": 1.7728377082738218e-06, + "loss": 0.016, + "step": 102180 + }, + { + "epoch": 9.28831121614252, + "grad_norm": 0.1305139809846878, + "learning_rate": 1.7708887559732894e-06, + "loss": 0.0145, + "step": 102190 + }, + { + "epoch": 9.289220141792402, + "grad_norm": 0.11341360211372375, + "learning_rate": 1.768940808336456e-06, + "loss": 0.015, + "step": 102200 + }, + { + "epoch": 9.290129067442283, + "grad_norm": 0.1313796192407608, + "learning_rate": 1.7669938655112467e-06, + "loss": 0.0156, + "step": 102210 + }, + { + "epoch": 9.291037993092164, + "grad_norm": 0.2250227928161621, + "learning_rate": 1.7650479276455244e-06, + "loss": 0.016, + "step": 102220 + }, + { + "epoch": 9.291946918742047, + "grad_norm": 0.13094903528690338, + "learning_rate": 1.7631029948870713e-06, + "loss": 0.0164, + "step": 102230 + }, + { + "epoch": 9.292855844391928, + "grad_norm": 0.16390368342399597, + "learning_rate": 1.7611590673835859e-06, + "loss": 0.0133, + "step": 102240 + }, + { + "epoch": 9.293764770041811, + "grad_norm": 0.14693693816661835, + "learning_rate": 1.759216145282705e-06, + "loss": 0.0175, + "step": 102250 + }, + { + "epoch": 9.294673695691692, + "grad_norm": 0.11033441871404648, + "learning_rate": 1.757274228731972e-06, + "loss": 0.015, + "step": 102260 + }, + { + "epoch": 9.295582621341575, + "grad_norm": 0.20225705206394196, + "learning_rate": 1.755333317878871e-06, + "loss": 0.0154, + "step": 102270 + }, + { + "epoch": 9.296491546991456, + "grad_norm": 0.11019633710384369, + "learning_rate": 1.753393412870794e-06, + "loss": 0.0168, + "step": 102280 + }, + { + "epoch": 9.297400472641337, + "grad_norm": 0.12198768556118011, + "learning_rate": 1.751454513855068e-06, + "loss": 0.0151, + "step": 102290 + }, + { + "epoch": 9.29830939829122, + "grad_norm": 0.16472388803958893, + "learning_rate": 1.7495166209789404e-06, + "loss": 0.0164, + "step": 102300 + }, + { + "epoch": 9.299218323941101, + "grad_norm": 0.1536923348903656, + "learning_rate": 1.74757973438958e-06, + "loss": 0.0158, + "step": 102310 + }, + { + "epoch": 9.300127249590984, + "grad_norm": 0.20444998145103455, + "learning_rate": 1.7456438542340758e-06, + "loss": 0.0162, + "step": 102320 + }, + { + "epoch": 9.301036175240865, + "grad_norm": 0.14016038179397583, + "learning_rate": 1.7437089806594553e-06, + "loss": 0.0157, + "step": 102330 + }, + { + "epoch": 9.301945100890746, + "grad_norm": 0.1129390150308609, + "learning_rate": 1.7417751138126525e-06, + "loss": 0.0169, + "step": 102340 + }, + { + "epoch": 9.30285402654063, + "grad_norm": 0.17588624358177185, + "learning_rate": 1.7398422538405317e-06, + "loss": 0.0141, + "step": 102350 + }, + { + "epoch": 9.30376295219051, + "grad_norm": 0.12415069341659546, + "learning_rate": 1.7379104008898805e-06, + "loss": 0.0163, + "step": 102360 + }, + { + "epoch": 9.304671877840393, + "grad_norm": 0.12741300463676453, + "learning_rate": 1.7359795551074165e-06, + "loss": 0.0147, + "step": 102370 + }, + { + "epoch": 9.305580803490274, + "grad_norm": 0.18073055148124695, + "learning_rate": 1.7340497166397707e-06, + "loss": 0.0175, + "step": 102380 + }, + { + "epoch": 9.306489729140157, + "grad_norm": 0.09954237192869186, + "learning_rate": 1.7321208856334974e-06, + "loss": 0.016, + "step": 102390 + }, + { + "epoch": 9.307398654790038, + "grad_norm": 0.13726669549942017, + "learning_rate": 1.7301930622350832e-06, + "loss": 0.0153, + "step": 102400 + }, + { + "epoch": 9.30830758043992, + "grad_norm": 0.17953653633594513, + "learning_rate": 1.7282662465909355e-06, + "loss": 0.0152, + "step": 102410 + }, + { + "epoch": 9.309216506089802, + "grad_norm": 0.11855386942625046, + "learning_rate": 1.7263404388473807e-06, + "loss": 0.0163, + "step": 102420 + }, + { + "epoch": 9.310125431739683, + "grad_norm": 0.14540497958660126, + "learning_rate": 1.7244156391506654e-06, + "loss": 0.0157, + "step": 102430 + }, + { + "epoch": 9.311034357389566, + "grad_norm": 0.20411837100982666, + "learning_rate": 1.7224918476469758e-06, + "loss": 0.0165, + "step": 102440 + }, + { + "epoch": 9.311943283039447, + "grad_norm": 0.16975170373916626, + "learning_rate": 1.7205690644824068e-06, + "loss": 0.0174, + "step": 102450 + }, + { + "epoch": 9.312852208689328, + "grad_norm": 0.13764485716819763, + "learning_rate": 1.7186472898029782e-06, + "loss": 0.0148, + "step": 102460 + }, + { + "epoch": 9.313761134339211, + "grad_norm": 0.17357346415519714, + "learning_rate": 1.7167265237546382e-06, + "loss": 0.0178, + "step": 102470 + }, + { + "epoch": 9.314670059989092, + "grad_norm": 0.11333439499139786, + "learning_rate": 1.71480676648326e-06, + "loss": 0.0166, + "step": 102480 + }, + { + "epoch": 9.315578985638975, + "grad_norm": 0.11041215062141418, + "learning_rate": 1.7128880181346307e-06, + "loss": 0.0206, + "step": 102490 + }, + { + "epoch": 9.316487911288856, + "grad_norm": 0.10474342852830887, + "learning_rate": 1.7109702788544718e-06, + "loss": 0.0154, + "step": 102500 + }, + { + "epoch": 9.31739683693874, + "grad_norm": 0.17614604532718658, + "learning_rate": 1.7090535487884167e-06, + "loss": 0.0171, + "step": 102510 + }, + { + "epoch": 9.31830576258862, + "grad_norm": 0.13928036391735077, + "learning_rate": 1.707137828082036e-06, + "loss": 0.0172, + "step": 102520 + }, + { + "epoch": 9.319214688238501, + "grad_norm": 0.12312206625938416, + "learning_rate": 1.705223116880808e-06, + "loss": 0.0149, + "step": 102530 + }, + { + "epoch": 9.320123613888384, + "grad_norm": 0.18080799281597137, + "learning_rate": 1.7033094153301466e-06, + "loss": 0.0166, + "step": 102540 + }, + { + "epoch": 9.321032539538265, + "grad_norm": 0.17279241979122162, + "learning_rate": 1.7013967235753886e-06, + "loss": 0.0172, + "step": 102550 + }, + { + "epoch": 9.321941465188148, + "grad_norm": 0.10368528217077255, + "learning_rate": 1.699485041761783e-06, + "loss": 0.0149, + "step": 102560 + }, + { + "epoch": 9.32285039083803, + "grad_norm": 0.2331407219171524, + "learning_rate": 1.6975743700345154e-06, + "loss": 0.0143, + "step": 102570 + }, + { + "epoch": 9.32375931648791, + "grad_norm": 0.08382339030504227, + "learning_rate": 1.695664708538683e-06, + "loss": 0.0152, + "step": 102580 + }, + { + "epoch": 9.324668242137793, + "grad_norm": 0.11677534878253937, + "learning_rate": 1.6937560574193145e-06, + "loss": 0.0165, + "step": 102590 + }, + { + "epoch": 9.325577167787674, + "grad_norm": 0.09809934347867966, + "learning_rate": 1.6918484168213627e-06, + "loss": 0.0181, + "step": 102600 + }, + { + "epoch": 9.326486093437557, + "grad_norm": 0.18006743490695953, + "learning_rate": 1.6899417868896983e-06, + "loss": 0.0192, + "step": 102610 + }, + { + "epoch": 9.327395019087438, + "grad_norm": 0.13793188333511353, + "learning_rate": 1.688036167769112e-06, + "loss": 0.015, + "step": 102620 + }, + { + "epoch": 9.328303944737321, + "grad_norm": 0.1829003095626831, + "learning_rate": 1.6861315596043297e-06, + "loss": 0.0162, + "step": 102630 + }, + { + "epoch": 9.329212870387202, + "grad_norm": 0.20014312863349915, + "learning_rate": 1.6842279625399925e-06, + "loss": 0.0163, + "step": 102640 + }, + { + "epoch": 9.330121796037083, + "grad_norm": 0.13265419006347656, + "learning_rate": 1.6823253767206632e-06, + "loss": 0.0168, + "step": 102650 + }, + { + "epoch": 9.331030721686966, + "grad_norm": 0.13645115494728088, + "learning_rate": 1.6804238022908326e-06, + "loss": 0.0176, + "step": 102660 + }, + { + "epoch": 9.331939647336847, + "grad_norm": 0.13760216534137726, + "learning_rate": 1.6785232393949152e-06, + "loss": 0.0146, + "step": 102670 + }, + { + "epoch": 9.33284857298673, + "grad_norm": 0.15008816123008728, + "learning_rate": 1.6766236881772439e-06, + "loss": 0.0147, + "step": 102680 + }, + { + "epoch": 9.333757498636611, + "grad_norm": 0.11106467992067337, + "learning_rate": 1.6747251487820736e-06, + "loss": 0.0147, + "step": 102690 + }, + { + "epoch": 9.334666424286493, + "grad_norm": 0.19915129244327545, + "learning_rate": 1.6728276213535903e-06, + "loss": 0.0177, + "step": 102700 + }, + { + "epoch": 9.335575349936375, + "grad_norm": 0.13646793365478516, + "learning_rate": 1.6709311060358989e-06, + "loss": 0.015, + "step": 102710 + }, + { + "epoch": 9.336484275586256, + "grad_norm": 0.08726493269205093, + "learning_rate": 1.6690356029730274e-06, + "loss": 0.0131, + "step": 102720 + }, + { + "epoch": 9.33739320123614, + "grad_norm": 0.14277102053165436, + "learning_rate": 1.6671411123089175e-06, + "loss": 0.0161, + "step": 102730 + }, + { + "epoch": 9.33830212688602, + "grad_norm": 0.102022185921669, + "learning_rate": 1.6652476341874572e-06, + "loss": 0.0175, + "step": 102740 + }, + { + "epoch": 9.339211052535903, + "grad_norm": 0.24796079099178314, + "learning_rate": 1.6633551687524385e-06, + "loss": 0.0159, + "step": 102750 + }, + { + "epoch": 9.340119978185784, + "grad_norm": 0.14480523765087128, + "learning_rate": 1.661463716147576e-06, + "loss": 0.0162, + "step": 102760 + }, + { + "epoch": 9.341028903835666, + "grad_norm": 0.12386812269687653, + "learning_rate": 1.6595732765165184e-06, + "loss": 0.0188, + "step": 102770 + }, + { + "epoch": 9.341937829485548, + "grad_norm": 0.09113601595163345, + "learning_rate": 1.657683850002834e-06, + "loss": 0.016, + "step": 102780 + }, + { + "epoch": 9.34284675513543, + "grad_norm": 0.14431236684322357, + "learning_rate": 1.65579543675001e-06, + "loss": 0.0164, + "step": 102790 + }, + { + "epoch": 9.343755680785312, + "grad_norm": 0.15139298141002655, + "learning_rate": 1.6539080369014564e-06, + "loss": 0.0171, + "step": 102800 + }, + { + "epoch": 9.344664606435193, + "grad_norm": 0.11223012208938599, + "learning_rate": 1.6520216506005087e-06, + "loss": 0.0133, + "step": 102810 + }, + { + "epoch": 9.345573532085076, + "grad_norm": 0.10917547345161438, + "learning_rate": 1.650136277990432e-06, + "loss": 0.0159, + "step": 102820 + }, + { + "epoch": 9.346482457734957, + "grad_norm": 0.09646528959274292, + "learning_rate": 1.6482519192144007e-06, + "loss": 0.0133, + "step": 102830 + }, + { + "epoch": 9.347391383384839, + "grad_norm": 0.15132661163806915, + "learning_rate": 1.6463685744155215e-06, + "loss": 0.0144, + "step": 102840 + }, + { + "epoch": 9.348300309034721, + "grad_norm": 0.14201460778713226, + "learning_rate": 1.6444862437368268e-06, + "loss": 0.0159, + "step": 102850 + }, + { + "epoch": 9.349209234684603, + "grad_norm": 0.17186735570430756, + "learning_rate": 1.642604927321259e-06, + "loss": 0.0149, + "step": 102860 + }, + { + "epoch": 9.350118160334485, + "grad_norm": 0.17039459943771362, + "learning_rate": 1.6407246253116987e-06, + "loss": 0.0175, + "step": 102870 + }, + { + "epoch": 9.351027085984366, + "grad_norm": 0.1389954537153244, + "learning_rate": 1.6388453378509383e-06, + "loss": 0.0154, + "step": 102880 + }, + { + "epoch": 9.351936011634248, + "grad_norm": 0.1379847228527069, + "learning_rate": 1.6369670650817004e-06, + "loss": 0.0143, + "step": 102890 + }, + { + "epoch": 9.35284493728413, + "grad_norm": 0.17963139712810516, + "learning_rate": 1.6350898071466225e-06, + "loss": 0.0145, + "step": 102900 + }, + { + "epoch": 9.353753862934012, + "grad_norm": 0.11506941169500351, + "learning_rate": 1.6332135641882757e-06, + "loss": 0.0174, + "step": 102910 + }, + { + "epoch": 9.354662788583894, + "grad_norm": 0.1379062533378601, + "learning_rate": 1.6313383363491424e-06, + "loss": 0.0183, + "step": 102920 + }, + { + "epoch": 9.355571714233776, + "grad_norm": 0.1174943745136261, + "learning_rate": 1.6294641237716373e-06, + "loss": 0.0147, + "step": 102930 + }, + { + "epoch": 9.356480639883657, + "grad_norm": 0.12372636049985886, + "learning_rate": 1.6275909265980982e-06, + "loss": 0.0151, + "step": 102940 + }, + { + "epoch": 9.35738956553354, + "grad_norm": 0.11241048574447632, + "learning_rate": 1.6257187449707745e-06, + "loss": 0.0131, + "step": 102950 + }, + { + "epoch": 9.35829849118342, + "grad_norm": 0.11756923794746399, + "learning_rate": 1.6238475790318492e-06, + "loss": 0.0176, + "step": 102960 + }, + { + "epoch": 9.359207416833303, + "grad_norm": 0.13009941577911377, + "learning_rate": 1.6219774289234285e-06, + "loss": 0.0151, + "step": 102970 + }, + { + "epoch": 9.360116342483185, + "grad_norm": 0.09908907860517502, + "learning_rate": 1.620108294787534e-06, + "loss": 0.0142, + "step": 102980 + }, + { + "epoch": 9.361025268133067, + "grad_norm": 0.11186620593070984, + "learning_rate": 1.618240176766112e-06, + "loss": 0.0142, + "step": 102990 + }, + { + "epoch": 9.361934193782949, + "grad_norm": 0.21217821538448334, + "learning_rate": 1.6163730750010376e-06, + "loss": 0.0153, + "step": 103000 + }, + { + "epoch": 9.36284311943283, + "grad_norm": 0.21259628236293793, + "learning_rate": 1.6145069896341069e-06, + "loss": 0.015, + "step": 103010 + }, + { + "epoch": 9.363752045082713, + "grad_norm": 0.1388179212808609, + "learning_rate": 1.6126419208070336e-06, + "loss": 0.0159, + "step": 103020 + }, + { + "epoch": 9.364660970732594, + "grad_norm": 0.11203529685735703, + "learning_rate": 1.610777868661451e-06, + "loss": 0.0175, + "step": 103030 + }, + { + "epoch": 9.365569896382477, + "grad_norm": 0.14190183579921722, + "learning_rate": 1.6089148333389342e-06, + "loss": 0.0156, + "step": 103040 + }, + { + "epoch": 9.366478822032358, + "grad_norm": 0.12107337266206741, + "learning_rate": 1.6070528149809632e-06, + "loss": 0.019, + "step": 103050 + }, + { + "epoch": 9.36738774768224, + "grad_norm": 0.282999724149704, + "learning_rate": 1.6051918137289417e-06, + "loss": 0.0167, + "step": 103060 + }, + { + "epoch": 9.368296673332122, + "grad_norm": 0.1695752590894699, + "learning_rate": 1.6033318297242033e-06, + "loss": 0.0175, + "step": 103070 + }, + { + "epoch": 9.369205598982003, + "grad_norm": 0.16158685088157654, + "learning_rate": 1.601472863108005e-06, + "loss": 0.0151, + "step": 103080 + }, + { + "epoch": 9.370114524631886, + "grad_norm": 0.1202201098203659, + "learning_rate": 1.5996149140215206e-06, + "loss": 0.0163, + "step": 103090 + }, + { + "epoch": 9.371023450281767, + "grad_norm": 0.11231476068496704, + "learning_rate": 1.597757982605847e-06, + "loss": 0.016, + "step": 103100 + }, + { + "epoch": 9.37193237593165, + "grad_norm": 0.09401696920394897, + "learning_rate": 1.595902069002005e-06, + "loss": 0.0171, + "step": 103110 + }, + { + "epoch": 9.37284130158153, + "grad_norm": 0.18930889666080475, + "learning_rate": 1.5940471733509465e-06, + "loss": 0.0133, + "step": 103120 + }, + { + "epoch": 9.373750227231412, + "grad_norm": 0.09401583671569824, + "learning_rate": 1.5921932957935292e-06, + "loss": 0.0152, + "step": 103130 + }, + { + "epoch": 9.374659152881295, + "grad_norm": 0.11119768768548965, + "learning_rate": 1.5903404364705488e-06, + "loss": 0.015, + "step": 103140 + }, + { + "epoch": 9.375568078531176, + "grad_norm": 0.167364701628685, + "learning_rate": 1.5884885955227173e-06, + "loss": 0.0155, + "step": 103150 + }, + { + "epoch": 9.376477004181059, + "grad_norm": 0.1718003749847412, + "learning_rate": 1.5866377730906678e-06, + "loss": 0.0169, + "step": 103160 + }, + { + "epoch": 9.37738592983094, + "grad_norm": 0.1236250102519989, + "learning_rate": 1.5847879693149609e-06, + "loss": 0.0151, + "step": 103170 + }, + { + "epoch": 9.37829485548082, + "grad_norm": 0.10674560070037842, + "learning_rate": 1.5829391843360707e-06, + "loss": 0.0147, + "step": 103180 + }, + { + "epoch": 9.379203781130704, + "grad_norm": 0.1566414088010788, + "learning_rate": 1.5810914182944085e-06, + "loss": 0.0194, + "step": 103190 + }, + { + "epoch": 9.380112706780585, + "grad_norm": 0.132230743765831, + "learning_rate": 1.5792446713302937e-06, + "loss": 0.0157, + "step": 103200 + }, + { + "epoch": 9.381021632430468, + "grad_norm": 0.12096261233091354, + "learning_rate": 1.5773989435839792e-06, + "loss": 0.0167, + "step": 103210 + }, + { + "epoch": 9.381930558080349, + "grad_norm": 0.17302018404006958, + "learning_rate": 1.5755542351956308e-06, + "loss": 0.0161, + "step": 103220 + }, + { + "epoch": 9.382839483730232, + "grad_norm": 0.10714767873287201, + "learning_rate": 1.5737105463053452e-06, + "loss": 0.0165, + "step": 103230 + }, + { + "epoch": 9.383748409380113, + "grad_norm": 0.1559232622385025, + "learning_rate": 1.5718678770531403e-06, + "loss": 0.0147, + "step": 103240 + }, + { + "epoch": 9.384657335029994, + "grad_norm": 0.12750175595283508, + "learning_rate": 1.5700262275789489e-06, + "loss": 0.0171, + "step": 103250 + }, + { + "epoch": 9.385566260679877, + "grad_norm": 0.12564626336097717, + "learning_rate": 1.5681855980226377e-06, + "loss": 0.016, + "step": 103260 + }, + { + "epoch": 9.386475186329758, + "grad_norm": 0.09702452272176743, + "learning_rate": 1.5663459885239868e-06, + "loss": 0.0161, + "step": 103270 + }, + { + "epoch": 9.38738411197964, + "grad_norm": 0.12287244200706482, + "learning_rate": 1.5645073992227056e-06, + "loss": 0.0164, + "step": 103280 + }, + { + "epoch": 9.388293037629522, + "grad_norm": 0.13550761342048645, + "learning_rate": 1.5626698302584174e-06, + "loss": 0.0169, + "step": 103290 + }, + { + "epoch": 9.389201963279405, + "grad_norm": 0.20618319511413574, + "learning_rate": 1.5608332817706777e-06, + "loss": 0.0141, + "step": 103300 + }, + { + "epoch": 9.390110888929286, + "grad_norm": 0.14924342930316925, + "learning_rate": 1.558997753898963e-06, + "loss": 0.016, + "step": 103310 + }, + { + "epoch": 9.391019814579167, + "grad_norm": 0.18232031166553497, + "learning_rate": 1.5571632467826652e-06, + "loss": 0.0144, + "step": 103320 + }, + { + "epoch": 9.39192874022905, + "grad_norm": 0.17804668843746185, + "learning_rate": 1.5553297605610978e-06, + "loss": 0.0166, + "step": 103330 + }, + { + "epoch": 9.39283766587893, + "grad_norm": 0.13166095316410065, + "learning_rate": 1.5534972953735143e-06, + "loss": 0.0176, + "step": 103340 + }, + { + "epoch": 9.393746591528814, + "grad_norm": 0.14352580904960632, + "learning_rate": 1.551665851359072e-06, + "loss": 0.0174, + "step": 103350 + }, + { + "epoch": 9.394655517178695, + "grad_norm": 0.13180042803287506, + "learning_rate": 1.5498354286568545e-06, + "loss": 0.0143, + "step": 103360 + }, + { + "epoch": 9.395564442828576, + "grad_norm": 0.13848282396793365, + "learning_rate": 1.5480060274058721e-06, + "loss": 0.0179, + "step": 103370 + }, + { + "epoch": 9.396473368478459, + "grad_norm": 0.10693332552909851, + "learning_rate": 1.546177647745059e-06, + "loss": 0.0155, + "step": 103380 + }, + { + "epoch": 9.39738229412834, + "grad_norm": 0.10540972650051117, + "learning_rate": 1.5443502898132672e-06, + "loss": 0.0155, + "step": 103390 + }, + { + "epoch": 9.398291219778223, + "grad_norm": 0.1706380993127823, + "learning_rate": 1.5425239537492674e-06, + "loss": 0.0153, + "step": 103400 + }, + { + "epoch": 9.399200145428104, + "grad_norm": 0.12080085277557373, + "learning_rate": 1.5406986396917622e-06, + "loss": 0.0162, + "step": 103410 + }, + { + "epoch": 9.400109071077987, + "grad_norm": 0.12537258863449097, + "learning_rate": 1.5388743477793737e-06, + "loss": 0.0165, + "step": 103420 + }, + { + "epoch": 9.401017996727868, + "grad_norm": 0.13123078644275665, + "learning_rate": 1.5370510781506415e-06, + "loss": 0.0163, + "step": 103430 + }, + { + "epoch": 9.401926922377749, + "grad_norm": 0.13414214551448822, + "learning_rate": 1.535228830944031e-06, + "loss": 0.0161, + "step": 103440 + }, + { + "epoch": 9.402835848027632, + "grad_norm": 0.10792528092861176, + "learning_rate": 1.5334076062979335e-06, + "loss": 0.0169, + "step": 103450 + }, + { + "epoch": 9.403744773677513, + "grad_norm": 0.1685778796672821, + "learning_rate": 1.5315874043506584e-06, + "loss": 0.0171, + "step": 103460 + }, + { + "epoch": 9.404653699327396, + "grad_norm": 0.1470065414905548, + "learning_rate": 1.5297682252404316e-06, + "loss": 0.0173, + "step": 103470 + }, + { + "epoch": 9.405562624977277, + "grad_norm": 0.14849591255187988, + "learning_rate": 1.5279500691054127e-06, + "loss": 0.0144, + "step": 103480 + }, + { + "epoch": 9.406471550627158, + "grad_norm": 0.15112371742725372, + "learning_rate": 1.5261329360836813e-06, + "loss": 0.0153, + "step": 103490 + }, + { + "epoch": 9.40738047627704, + "grad_norm": 0.18348096311092377, + "learning_rate": 1.52431682631323e-06, + "loss": 0.0175, + "step": 103500 + }, + { + "epoch": 9.408289401926922, + "grad_norm": 0.1282050758600235, + "learning_rate": 1.522501739931989e-06, + "loss": 0.0154, + "step": 103510 + }, + { + "epoch": 9.409198327576805, + "grad_norm": 0.12533338367938995, + "learning_rate": 1.5206876770777928e-06, + "loss": 0.0144, + "step": 103520 + }, + { + "epoch": 9.410107253226686, + "grad_norm": 0.13370224833488464, + "learning_rate": 1.5188746378884111e-06, + "loss": 0.0145, + "step": 103530 + }, + { + "epoch": 9.411016178876569, + "grad_norm": 0.12504078447818756, + "learning_rate": 1.517062622501537e-06, + "loss": 0.0159, + "step": 103540 + }, + { + "epoch": 9.41192510452645, + "grad_norm": 0.13513946533203125, + "learning_rate": 1.5152516310547727e-06, + "loss": 0.018, + "step": 103550 + }, + { + "epoch": 9.412834030176331, + "grad_norm": 0.12656861543655396, + "learning_rate": 1.5134416636856607e-06, + "loss": 0.0156, + "step": 103560 + }, + { + "epoch": 9.413742955826214, + "grad_norm": 0.10080912709236145, + "learning_rate": 1.5116327205316465e-06, + "loss": 0.015, + "step": 103570 + }, + { + "epoch": 9.414651881476095, + "grad_norm": 0.13591808080673218, + "learning_rate": 1.5098248017301136e-06, + "loss": 0.0171, + "step": 103580 + }, + { + "epoch": 9.415560807125978, + "grad_norm": 0.14449210464954376, + "learning_rate": 1.5080179074183587e-06, + "loss": 0.0156, + "step": 103590 + }, + { + "epoch": 9.416469732775859, + "grad_norm": 0.10185254365205765, + "learning_rate": 1.5062120377336036e-06, + "loss": 0.014, + "step": 103600 + }, + { + "epoch": 9.41737865842574, + "grad_norm": 0.17457516491413116, + "learning_rate": 1.5044071928129953e-06, + "loss": 0.0154, + "step": 103610 + }, + { + "epoch": 9.418287584075623, + "grad_norm": 0.10957769304513931, + "learning_rate": 1.5026033727935978e-06, + "loss": 0.0156, + "step": 103620 + }, + { + "epoch": 9.419196509725504, + "grad_norm": 0.19568973779678345, + "learning_rate": 1.5008005778123963e-06, + "loss": 0.0161, + "step": 103630 + }, + { + "epoch": 9.420105435375387, + "grad_norm": 0.1913815140724182, + "learning_rate": 1.4989988080063043e-06, + "loss": 0.0164, + "step": 103640 + }, + { + "epoch": 9.421014361025268, + "grad_norm": 0.13730937242507935, + "learning_rate": 1.4971980635121562e-06, + "loss": 0.0162, + "step": 103650 + }, + { + "epoch": 9.42192328667515, + "grad_norm": 0.26684433221817017, + "learning_rate": 1.4953983444667025e-06, + "loss": 0.0137, + "step": 103660 + }, + { + "epoch": 9.422832212325032, + "grad_norm": 0.12091740220785141, + "learning_rate": 1.4935996510066203e-06, + "loss": 0.0168, + "step": 103670 + }, + { + "epoch": 9.423741137974913, + "grad_norm": 0.130889430642128, + "learning_rate": 1.4918019832685136e-06, + "loss": 0.0177, + "step": 103680 + }, + { + "epoch": 9.424650063624796, + "grad_norm": 0.1329699009656906, + "learning_rate": 1.490005341388902e-06, + "loss": 0.0166, + "step": 103690 + }, + { + "epoch": 9.425558989274677, + "grad_norm": 0.15828004479408264, + "learning_rate": 1.488209725504221e-06, + "loss": 0.014, + "step": 103700 + }, + { + "epoch": 9.42646791492456, + "grad_norm": 0.14999423921108246, + "learning_rate": 1.4864151357508433e-06, + "loss": 0.0147, + "step": 103710 + }, + { + "epoch": 9.427376840574441, + "grad_norm": 0.08984575420618057, + "learning_rate": 1.4846215722650547e-06, + "loss": 0.0184, + "step": 103720 + }, + { + "epoch": 9.428285766224322, + "grad_norm": 0.11003144830465317, + "learning_rate": 1.4828290351830665e-06, + "loss": 0.0161, + "step": 103730 + }, + { + "epoch": 9.429194691874205, + "grad_norm": 0.14026454091072083, + "learning_rate": 1.4810375246409996e-06, + "loss": 0.0169, + "step": 103740 + }, + { + "epoch": 9.430103617524086, + "grad_norm": 0.09430599957704544, + "learning_rate": 1.479247040774922e-06, + "loss": 0.0159, + "step": 103750 + }, + { + "epoch": 9.431012543173969, + "grad_norm": 0.17432019114494324, + "learning_rate": 1.4774575837208032e-06, + "loss": 0.0156, + "step": 103760 + }, + { + "epoch": 9.43192146882385, + "grad_norm": 0.12239013612270355, + "learning_rate": 1.475669153614536e-06, + "loss": 0.015, + "step": 103770 + }, + { + "epoch": 9.432830394473733, + "grad_norm": 0.12483002245426178, + "learning_rate": 1.473881750591944e-06, + "loss": 0.0147, + "step": 103780 + }, + { + "epoch": 9.433739320123614, + "grad_norm": 0.10648220777511597, + "learning_rate": 1.472095374788771e-06, + "loss": 0.0147, + "step": 103790 + }, + { + "epoch": 9.434648245773495, + "grad_norm": 0.13482308387756348, + "learning_rate": 1.470310026340676e-06, + "loss": 0.0194, + "step": 103800 + }, + { + "epoch": 9.435557171423378, + "grad_norm": 0.2045266181230545, + "learning_rate": 1.4685257053832485e-06, + "loss": 0.0155, + "step": 103810 + }, + { + "epoch": 9.436466097073259, + "grad_norm": 0.19423820078372955, + "learning_rate": 1.46674241205199e-06, + "loss": 0.0157, + "step": 103820 + }, + { + "epoch": 9.437375022723142, + "grad_norm": 0.16856049001216888, + "learning_rate": 1.4649601464823376e-06, + "loss": 0.0149, + "step": 103830 + }, + { + "epoch": 9.438283948373023, + "grad_norm": 0.1282351016998291, + "learning_rate": 1.463178908809636e-06, + "loss": 0.0154, + "step": 103840 + }, + { + "epoch": 9.439192874022904, + "grad_norm": 0.20958386361598969, + "learning_rate": 1.4613986991691603e-06, + "loss": 0.0147, + "step": 103850 + }, + { + "epoch": 9.440101799672787, + "grad_norm": 0.12357739359140396, + "learning_rate": 1.4596195176961092e-06, + "loss": 0.0177, + "step": 103860 + }, + { + "epoch": 9.441010725322668, + "grad_norm": 0.11934009194374084, + "learning_rate": 1.4578413645255945e-06, + "loss": 0.0154, + "step": 103870 + }, + { + "epoch": 9.441919650972551, + "grad_norm": 0.14560960233211517, + "learning_rate": 1.4560642397926599e-06, + "loss": 0.0141, + "step": 103880 + }, + { + "epoch": 9.442828576622432, + "grad_norm": 0.12442056834697723, + "learning_rate": 1.4542881436322625e-06, + "loss": 0.0158, + "step": 103890 + }, + { + "epoch": 9.443737502272315, + "grad_norm": 0.11504605412483215, + "learning_rate": 1.452513076179286e-06, + "loss": 0.0148, + "step": 103900 + }, + { + "epoch": 9.444646427922196, + "grad_norm": 0.14904776215553284, + "learning_rate": 1.4507390375685376e-06, + "loss": 0.0143, + "step": 103910 + }, + { + "epoch": 9.445555353572077, + "grad_norm": 0.1030309721827507, + "learning_rate": 1.4489660279347427e-06, + "loss": 0.0157, + "step": 103920 + }, + { + "epoch": 9.44646427922196, + "grad_norm": 0.13855351507663727, + "learning_rate": 1.4471940474125472e-06, + "loss": 0.0156, + "step": 103930 + }, + { + "epoch": 9.447373204871841, + "grad_norm": 0.18820635974407196, + "learning_rate": 1.4454230961365217e-06, + "loss": 0.017, + "step": 103940 + }, + { + "epoch": 9.448282130521724, + "grad_norm": 0.15189211070537567, + "learning_rate": 1.4436531742411617e-06, + "loss": 0.016, + "step": 103950 + }, + { + "epoch": 9.449191056171605, + "grad_norm": 0.105583056807518, + "learning_rate": 1.4418842818608785e-06, + "loss": 0.0162, + "step": 103960 + }, + { + "epoch": 9.450099981821488, + "grad_norm": 0.1440572887659073, + "learning_rate": 1.4401164191300076e-06, + "loss": 0.0157, + "step": 103970 + }, + { + "epoch": 9.451008907471369, + "grad_norm": 0.22985070943832397, + "learning_rate": 1.4383495861828083e-06, + "loss": 0.0153, + "step": 103980 + }, + { + "epoch": 9.45191783312125, + "grad_norm": 0.12938760221004486, + "learning_rate": 1.43658378315346e-06, + "loss": 0.0152, + "step": 103990 + }, + { + "epoch": 9.452826758771133, + "grad_norm": 0.146579772233963, + "learning_rate": 1.4348190101760606e-06, + "loss": 0.0148, + "step": 104000 + }, + { + "epoch": 9.453735684421014, + "grad_norm": 0.14186160266399384, + "learning_rate": 1.4330552673846342e-06, + "loss": 0.0158, + "step": 104010 + }, + { + "epoch": 9.454644610070897, + "grad_norm": 0.2026309072971344, + "learning_rate": 1.4312925549131289e-06, + "loss": 0.0158, + "step": 104020 + }, + { + "epoch": 9.455553535720778, + "grad_norm": 0.14876534044742584, + "learning_rate": 1.4295308728954093e-06, + "loss": 0.0171, + "step": 104030 + }, + { + "epoch": 9.45646246137066, + "grad_norm": 0.11100955307483673, + "learning_rate": 1.4277702214652566e-06, + "loss": 0.0159, + "step": 104040 + }, + { + "epoch": 9.457371387020542, + "grad_norm": 0.1632111519575119, + "learning_rate": 1.4260106007563906e-06, + "loss": 0.0143, + "step": 104050 + }, + { + "epoch": 9.458280312670423, + "grad_norm": 0.11087119579315186, + "learning_rate": 1.424252010902441e-06, + "loss": 0.0145, + "step": 104060 + }, + { + "epoch": 9.459189238320306, + "grad_norm": 0.11249266564846039, + "learning_rate": 1.422494452036956e-06, + "loss": 0.0138, + "step": 104070 + }, + { + "epoch": 9.460098163970187, + "grad_norm": 0.11591168493032455, + "learning_rate": 1.4207379242934121e-06, + "loss": 0.0146, + "step": 104080 + }, + { + "epoch": 9.461007089620068, + "grad_norm": 0.10039780288934708, + "learning_rate": 1.418982427805211e-06, + "loss": 0.0154, + "step": 104090 + }, + { + "epoch": 9.461916015269951, + "grad_norm": 0.14178112149238586, + "learning_rate": 1.4172279627056695e-06, + "loss": 0.0168, + "step": 104100 + }, + { + "epoch": 9.462824940919832, + "grad_norm": 0.15028049051761627, + "learning_rate": 1.4154745291280209e-06, + "loss": 0.0156, + "step": 104110 + }, + { + "epoch": 9.463733866569715, + "grad_norm": 0.11999195069074631, + "learning_rate": 1.4137221272054318e-06, + "loss": 0.0151, + "step": 104120 + }, + { + "epoch": 9.464642792219596, + "grad_norm": 0.0844268724322319, + "learning_rate": 1.4119707570709895e-06, + "loss": 0.0151, + "step": 104130 + }, + { + "epoch": 9.465551717869479, + "grad_norm": 0.14869116246700287, + "learning_rate": 1.4102204188576905e-06, + "loss": 0.0165, + "step": 104140 + }, + { + "epoch": 9.46646064351936, + "grad_norm": 0.16838042438030243, + "learning_rate": 1.4084711126984674e-06, + "loss": 0.0147, + "step": 104150 + }, + { + "epoch": 9.467369569169241, + "grad_norm": 0.10788612812757492, + "learning_rate": 1.4067228387261682e-06, + "loss": 0.0154, + "step": 104160 + }, + { + "epoch": 9.468278494819124, + "grad_norm": 0.13670162856578827, + "learning_rate": 1.4049755970735606e-06, + "loss": 0.0161, + "step": 104170 + }, + { + "epoch": 9.469187420469005, + "grad_norm": 0.15691348910331726, + "learning_rate": 1.4032293878733398e-06, + "loss": 0.0166, + "step": 104180 + }, + { + "epoch": 9.470096346118888, + "grad_norm": 0.12865404784679413, + "learning_rate": 1.4014842112581117e-06, + "loss": 0.0154, + "step": 104190 + }, + { + "epoch": 9.47100527176877, + "grad_norm": 0.1316698044538498, + "learning_rate": 1.39974006736042e-06, + "loss": 0.0144, + "step": 104200 + }, + { + "epoch": 9.471914197418652, + "grad_norm": 0.11985179781913757, + "learning_rate": 1.397996956312712e-06, + "loss": 0.0164, + "step": 104210 + }, + { + "epoch": 9.472823123068533, + "grad_norm": 0.12528827786445618, + "learning_rate": 1.3962548782473723e-06, + "loss": 0.0157, + "step": 104220 + }, + { + "epoch": 9.473732048718414, + "grad_norm": 0.16165241599082947, + "learning_rate": 1.3945138332966962e-06, + "loss": 0.015, + "step": 104230 + }, + { + "epoch": 9.474640974368297, + "grad_norm": 0.10385619103908539, + "learning_rate": 1.3927738215929064e-06, + "loss": 0.0136, + "step": 104240 + }, + { + "epoch": 9.475549900018178, + "grad_norm": 0.16712045669555664, + "learning_rate": 1.3910348432681458e-06, + "loss": 0.0155, + "step": 104250 + }, + { + "epoch": 9.476458825668061, + "grad_norm": 0.1597047597169876, + "learning_rate": 1.3892968984544767e-06, + "loss": 0.016, + "step": 104260 + }, + { + "epoch": 9.477367751317942, + "grad_norm": 0.1283843070268631, + "learning_rate": 1.387559987283884e-06, + "loss": 0.0168, + "step": 104270 + }, + { + "epoch": 9.478276676967823, + "grad_norm": 0.10232612490653992, + "learning_rate": 1.38582410988828e-06, + "loss": 0.0168, + "step": 104280 + }, + { + "epoch": 9.479185602617706, + "grad_norm": 0.12802189588546753, + "learning_rate": 1.384089266399488e-06, + "loss": 0.0158, + "step": 104290 + }, + { + "epoch": 9.480094528267587, + "grad_norm": 0.14002543687820435, + "learning_rate": 1.3823554569492557e-06, + "loss": 0.0173, + "step": 104300 + }, + { + "epoch": 9.48100345391747, + "grad_norm": 0.11892646551132202, + "learning_rate": 1.3806226816692595e-06, + "loss": 0.0151, + "step": 104310 + }, + { + "epoch": 9.481912379567351, + "grad_norm": 0.11528339236974716, + "learning_rate": 1.3788909406910927e-06, + "loss": 0.0161, + "step": 104320 + }, + { + "epoch": 9.482821305217232, + "grad_norm": 0.25585293769836426, + "learning_rate": 1.3771602341462664e-06, + "loss": 0.0158, + "step": 104330 + }, + { + "epoch": 9.483730230867115, + "grad_norm": 0.13151803612709045, + "learning_rate": 1.3754305621662138e-06, + "loss": 0.015, + "step": 104340 + }, + { + "epoch": 9.484639156516996, + "grad_norm": 0.15723367035388947, + "learning_rate": 1.3737019248823e-06, + "loss": 0.015, + "step": 104350 + }, + { + "epoch": 9.48554808216688, + "grad_norm": 0.14019989967346191, + "learning_rate": 1.3719743224257986e-06, + "loss": 0.0151, + "step": 104360 + }, + { + "epoch": 9.48645700781676, + "grad_norm": 0.12575535476207733, + "learning_rate": 1.3702477549279092e-06, + "loss": 0.0143, + "step": 104370 + }, + { + "epoch": 9.487365933466643, + "grad_norm": 0.18247447907924652, + "learning_rate": 1.3685222225197536e-06, + "loss": 0.0157, + "step": 104380 + }, + { + "epoch": 9.488274859116524, + "grad_norm": 0.14610250294208527, + "learning_rate": 1.366797725332377e-06, + "loss": 0.0156, + "step": 104390 + }, + { + "epoch": 9.489183784766405, + "grad_norm": 0.1520216017961502, + "learning_rate": 1.3650742634967433e-06, + "loss": 0.0143, + "step": 104400 + }, + { + "epoch": 9.490092710416288, + "grad_norm": 0.3091738820075989, + "learning_rate": 1.3633518371437326e-06, + "loss": 0.0191, + "step": 104410 + }, + { + "epoch": 9.49100163606617, + "grad_norm": 0.10835396498441696, + "learning_rate": 1.3616304464041568e-06, + "loss": 0.0154, + "step": 104420 + }, + { + "epoch": 9.491910561716052, + "grad_norm": 0.143049418926239, + "learning_rate": 1.3599100914087465e-06, + "loss": 0.0143, + "step": 104430 + }, + { + "epoch": 9.492819487365933, + "grad_norm": 0.10528197139501572, + "learning_rate": 1.3581907722881436e-06, + "loss": 0.0156, + "step": 104440 + }, + { + "epoch": 9.493728413015816, + "grad_norm": 0.18500782549381256, + "learning_rate": 1.356472489172924e-06, + "loss": 0.0168, + "step": 104450 + }, + { + "epoch": 9.494637338665697, + "grad_norm": 0.11781930923461914, + "learning_rate": 1.3547552421935832e-06, + "loss": 0.0164, + "step": 104460 + }, + { + "epoch": 9.495546264315578, + "grad_norm": 0.12556098401546478, + "learning_rate": 1.3530390314805303e-06, + "loss": 0.0149, + "step": 104470 + }, + { + "epoch": 9.496455189965461, + "grad_norm": 0.12349027395248413, + "learning_rate": 1.3513238571640973e-06, + "loss": 0.0146, + "step": 104480 + }, + { + "epoch": 9.497364115615342, + "grad_norm": 0.13872532546520233, + "learning_rate": 1.3496097193745437e-06, + "loss": 0.0156, + "step": 104490 + }, + { + "epoch": 9.498273041265225, + "grad_norm": 0.12654247879981995, + "learning_rate": 1.3478966182420522e-06, + "loss": 0.0161, + "step": 104500 + }, + { + "epoch": 9.499181966915106, + "grad_norm": 0.12787099182605743, + "learning_rate": 1.3461845538967115e-06, + "loss": 0.0145, + "step": 104510 + }, + { + "epoch": 9.500090892564987, + "grad_norm": 0.13689064979553223, + "learning_rate": 1.3444735264685497e-06, + "loss": 0.0157, + "step": 104520 + }, + { + "epoch": 9.50099981821487, + "grad_norm": 0.13680610060691833, + "learning_rate": 1.342763536087503e-06, + "loss": 0.015, + "step": 104530 + }, + { + "epoch": 9.501908743864751, + "grad_norm": 0.13927681744098663, + "learning_rate": 1.3410545828834352e-06, + "loss": 0.0155, + "step": 104540 + }, + { + "epoch": 9.502817669514634, + "grad_norm": 0.11028414219617844, + "learning_rate": 1.339346666986135e-06, + "loss": 0.0164, + "step": 104550 + }, + { + "epoch": 9.503726595164515, + "grad_norm": 0.1299813985824585, + "learning_rate": 1.3376397885252977e-06, + "loss": 0.0169, + "step": 104560 + }, + { + "epoch": 9.504635520814396, + "grad_norm": 0.12530705332756042, + "learning_rate": 1.3359339476305589e-06, + "loss": 0.016, + "step": 104570 + }, + { + "epoch": 9.50554444646428, + "grad_norm": 0.18403035402297974, + "learning_rate": 1.3342291444314592e-06, + "loss": 0.0144, + "step": 104580 + }, + { + "epoch": 9.50645337211416, + "grad_norm": 0.15165682137012482, + "learning_rate": 1.332525379057472e-06, + "loss": 0.015, + "step": 104590 + }, + { + "epoch": 9.507362297764043, + "grad_norm": 0.16214048862457275, + "learning_rate": 1.3308226516379824e-06, + "loss": 0.0152, + "step": 104600 + }, + { + "epoch": 9.508271223413924, + "grad_norm": 0.1186642050743103, + "learning_rate": 1.3291209623023032e-06, + "loss": 0.0146, + "step": 104610 + }, + { + "epoch": 9.509180149063807, + "grad_norm": 0.15547211468219757, + "learning_rate": 1.3274203111796695e-06, + "loss": 0.0115, + "step": 104620 + }, + { + "epoch": 9.510089074713688, + "grad_norm": 0.11494596302509308, + "learning_rate": 1.325720698399233e-06, + "loss": 0.0146, + "step": 104630 + }, + { + "epoch": 9.51099800036357, + "grad_norm": 0.12165243923664093, + "learning_rate": 1.3240221240900607e-06, + "loss": 0.0151, + "step": 104640 + }, + { + "epoch": 9.511906926013452, + "grad_norm": 0.11491052806377411, + "learning_rate": 1.3223245883811592e-06, + "loss": 0.0185, + "step": 104650 + }, + { + "epoch": 9.512815851663333, + "grad_norm": 0.13613663613796234, + "learning_rate": 1.3206280914014418e-06, + "loss": 0.0149, + "step": 104660 + }, + { + "epoch": 9.513724777313216, + "grad_norm": 0.13132457435131073, + "learning_rate": 1.3189326332797425e-06, + "loss": 0.017, + "step": 104670 + }, + { + "epoch": 9.514633702963097, + "grad_norm": 0.12359238415956497, + "learning_rate": 1.3172382141448214e-06, + "loss": 0.0145, + "step": 104680 + }, + { + "epoch": 9.51554262861298, + "grad_norm": 0.14382949471473694, + "learning_rate": 1.3155448341253623e-06, + "loss": 0.0155, + "step": 104690 + }, + { + "epoch": 9.516451554262861, + "grad_norm": 0.14982637763023376, + "learning_rate": 1.3138524933499641e-06, + "loss": 0.0166, + "step": 104700 + }, + { + "epoch": 9.517360479912742, + "grad_norm": 0.17779502272605896, + "learning_rate": 1.3121611919471455e-06, + "loss": 0.0171, + "step": 104710 + }, + { + "epoch": 9.518269405562625, + "grad_norm": 0.10722720623016357, + "learning_rate": 1.310470930045352e-06, + "loss": 0.0162, + "step": 104720 + }, + { + "epoch": 9.519178331212506, + "grad_norm": 0.1333388090133667, + "learning_rate": 1.3087817077729513e-06, + "loss": 0.0189, + "step": 104730 + }, + { + "epoch": 9.52008725686239, + "grad_norm": 0.07923126220703125, + "learning_rate": 1.307093525258224e-06, + "loss": 0.0174, + "step": 104740 + }, + { + "epoch": 9.52099618251227, + "grad_norm": 0.152533158659935, + "learning_rate": 1.3054063826293771e-06, + "loss": 0.0154, + "step": 104750 + }, + { + "epoch": 9.521905108162152, + "grad_norm": 0.12900637090206146, + "learning_rate": 1.303720280014542e-06, + "loss": 0.0179, + "step": 104760 + }, + { + "epoch": 9.522814033812034, + "grad_norm": 0.15320125222206116, + "learning_rate": 1.3020352175417626e-06, + "loss": 0.0157, + "step": 104770 + }, + { + "epoch": 9.523722959461915, + "grad_norm": 0.1076178103685379, + "learning_rate": 1.3003511953390084e-06, + "loss": 0.0118, + "step": 104780 + }, + { + "epoch": 9.524631885111798, + "grad_norm": 0.3168320953845978, + "learning_rate": 1.2986682135341722e-06, + "loss": 0.0145, + "step": 104790 + }, + { + "epoch": 9.52554081076168, + "grad_norm": 0.10529151558876038, + "learning_rate": 1.2969862722550647e-06, + "loss": 0.015, + "step": 104800 + }, + { + "epoch": 9.526449736411562, + "grad_norm": 0.1277824342250824, + "learning_rate": 1.2953053716294173e-06, + "loss": 0.0158, + "step": 104810 + }, + { + "epoch": 9.527358662061443, + "grad_norm": 0.09962745010852814, + "learning_rate": 1.2936255117848843e-06, + "loss": 0.0164, + "step": 104820 + }, + { + "epoch": 9.528267587711325, + "grad_norm": 0.10852918028831482, + "learning_rate": 1.2919466928490392e-06, + "loss": 0.0155, + "step": 104830 + }, + { + "epoch": 9.529176513361207, + "grad_norm": 0.0921255499124527, + "learning_rate": 1.2902689149493779e-06, + "loss": 0.0166, + "step": 104840 + }, + { + "epoch": 9.530085439011089, + "grad_norm": 0.2424924373626709, + "learning_rate": 1.2885921782133153e-06, + "loss": 0.0157, + "step": 104850 + }, + { + "epoch": 9.530994364660971, + "grad_norm": 0.11232401430606842, + "learning_rate": 1.286916482768188e-06, + "loss": 0.0166, + "step": 104860 + }, + { + "epoch": 9.531903290310852, + "grad_norm": 0.17637497186660767, + "learning_rate": 1.2852418287412588e-06, + "loss": 0.0148, + "step": 104870 + }, + { + "epoch": 9.532812215960734, + "grad_norm": 0.1332460194826126, + "learning_rate": 1.2835682162596997e-06, + "loss": 0.0163, + "step": 104880 + }, + { + "epoch": 9.533721141610616, + "grad_norm": 0.11751876026391983, + "learning_rate": 1.281895645450617e-06, + "loss": 0.0157, + "step": 104890 + }, + { + "epoch": 9.534630067260498, + "grad_norm": 0.1270950436592102, + "learning_rate": 1.2802241164410262e-06, + "loss": 0.0151, + "step": 104900 + }, + { + "epoch": 9.53553899291038, + "grad_norm": 0.1326606422662735, + "learning_rate": 1.2785536293578703e-06, + "loss": 0.0158, + "step": 104910 + }, + { + "epoch": 9.536447918560262, + "grad_norm": 0.20185348391532898, + "learning_rate": 1.2768841843280165e-06, + "loss": 0.0152, + "step": 104920 + }, + { + "epoch": 9.537356844210144, + "grad_norm": 0.1392907351255417, + "learning_rate": 1.2752157814782433e-06, + "loss": 0.0158, + "step": 104930 + }, + { + "epoch": 9.538265769860026, + "grad_norm": 0.12384951114654541, + "learning_rate": 1.2735484209352528e-06, + "loss": 0.0168, + "step": 104940 + }, + { + "epoch": 9.539174695509907, + "grad_norm": 0.12213040143251419, + "learning_rate": 1.271882102825675e-06, + "loss": 0.0149, + "step": 104950 + }, + { + "epoch": 9.54008362115979, + "grad_norm": 0.1122923493385315, + "learning_rate": 1.2702168272760538e-06, + "loss": 0.0145, + "step": 104960 + }, + { + "epoch": 9.54099254680967, + "grad_norm": 0.09271710366010666, + "learning_rate": 1.268552594412855e-06, + "loss": 0.0155, + "step": 104970 + }, + { + "epoch": 9.541901472459553, + "grad_norm": 0.2014773041009903, + "learning_rate": 1.2668894043624657e-06, + "loss": 0.017, + "step": 104980 + }, + { + "epoch": 9.542810398109435, + "grad_norm": 0.14827512204647064, + "learning_rate": 1.2652272572511998e-06, + "loss": 0.018, + "step": 104990 + }, + { + "epoch": 9.543719323759316, + "grad_norm": 0.13233615458011627, + "learning_rate": 1.2635661532052812e-06, + "loss": 0.0156, + "step": 105000 + }, + { + "epoch": 9.544628249409199, + "grad_norm": 0.11764826625585556, + "learning_rate": 1.2619060923508574e-06, + "loss": 0.0149, + "step": 105010 + }, + { + "epoch": 9.54553717505908, + "grad_norm": 0.17087431252002716, + "learning_rate": 1.2602470748140028e-06, + "loss": 0.0166, + "step": 105020 + }, + { + "epoch": 9.546446100708962, + "grad_norm": 0.19391170144081116, + "learning_rate": 1.2585891007207113e-06, + "loss": 0.0147, + "step": 105030 + }, + { + "epoch": 9.547355026358844, + "grad_norm": 0.13550807535648346, + "learning_rate": 1.2569321701968922e-06, + "loss": 0.015, + "step": 105040 + }, + { + "epoch": 9.548263952008726, + "grad_norm": 0.1347244828939438, + "learning_rate": 1.2552762833683735e-06, + "loss": 0.0167, + "step": 105050 + }, + { + "epoch": 9.549172877658608, + "grad_norm": 0.17809554934501648, + "learning_rate": 1.2536214403609193e-06, + "loss": 0.0135, + "step": 105060 + }, + { + "epoch": 9.550081803308489, + "grad_norm": 0.15714490413665771, + "learning_rate": 1.2519676413001974e-06, + "loss": 0.0161, + "step": 105070 + }, + { + "epoch": 9.550990728958372, + "grad_norm": 0.18172587454319, + "learning_rate": 1.2503148863118008e-06, + "loss": 0.0165, + "step": 105080 + }, + { + "epoch": 9.551899654608253, + "grad_norm": 0.09045309573411942, + "learning_rate": 1.2486631755212507e-06, + "loss": 0.0158, + "step": 105090 + }, + { + "epoch": 9.552808580258136, + "grad_norm": 0.0995364561676979, + "learning_rate": 1.2470125090539819e-06, + "loss": 0.0163, + "step": 105100 + }, + { + "epoch": 9.553717505908017, + "grad_norm": 0.15495195984840393, + "learning_rate": 1.2453628870353506e-06, + "loss": 0.0134, + "step": 105110 + }, + { + "epoch": 9.5546264315579, + "grad_norm": 0.11570623517036438, + "learning_rate": 1.2437143095906367e-06, + "loss": 0.0183, + "step": 105120 + }, + { + "epoch": 9.55553535720778, + "grad_norm": 0.09955990314483643, + "learning_rate": 1.242066776845035e-06, + "loss": 0.0153, + "step": 105130 + }, + { + "epoch": 9.556444282857662, + "grad_norm": 0.22112999856472015, + "learning_rate": 1.2404202889236705e-06, + "loss": 0.0181, + "step": 105140 + }, + { + "epoch": 9.557353208507545, + "grad_norm": 0.10205277800559998, + "learning_rate": 1.2387748459515763e-06, + "loss": 0.0147, + "step": 105150 + }, + { + "epoch": 9.558262134157426, + "grad_norm": 0.13398446142673492, + "learning_rate": 1.237130448053716e-06, + "loss": 0.0147, + "step": 105160 + }, + { + "epoch": 9.559171059807309, + "grad_norm": 0.11489690840244293, + "learning_rate": 1.2354870953549745e-06, + "loss": 0.0158, + "step": 105170 + }, + { + "epoch": 9.56007998545719, + "grad_norm": 0.2341056764125824, + "learning_rate": 1.2338447879801472e-06, + "loss": 0.0178, + "step": 105180 + }, + { + "epoch": 9.56098891110707, + "grad_norm": 0.14820359647274017, + "learning_rate": 1.2322035260539627e-06, + "loss": 0.019, + "step": 105190 + }, + { + "epoch": 9.561897836756954, + "grad_norm": 0.18144091963768005, + "learning_rate": 1.2305633097010577e-06, + "loss": 0.0153, + "step": 105200 + }, + { + "epoch": 9.562806762406835, + "grad_norm": 0.1510148048400879, + "learning_rate": 1.2289241390459994e-06, + "loss": 0.0168, + "step": 105210 + }, + { + "epoch": 9.563715688056718, + "grad_norm": 0.09487173706293106, + "learning_rate": 1.2272860142132731e-06, + "loss": 0.0162, + "step": 105220 + }, + { + "epoch": 9.564624613706599, + "grad_norm": 0.18148870766162872, + "learning_rate": 1.2256489353272843e-06, + "loss": 0.0154, + "step": 105230 + }, + { + "epoch": 9.56553353935648, + "grad_norm": 0.12755368649959564, + "learning_rate": 1.224012902512352e-06, + "loss": 0.0162, + "step": 105240 + }, + { + "epoch": 9.566442465006363, + "grad_norm": 0.12869642674922943, + "learning_rate": 1.222377915892725e-06, + "loss": 0.015, + "step": 105250 + }, + { + "epoch": 9.567351390656244, + "grad_norm": 0.24813459813594818, + "learning_rate": 1.2207439755925743e-06, + "loss": 0.0174, + "step": 105260 + }, + { + "epoch": 9.568260316306127, + "grad_norm": 0.13818947970867157, + "learning_rate": 1.2191110817359818e-06, + "loss": 0.0161, + "step": 105270 + }, + { + "epoch": 9.569169241956008, + "grad_norm": 0.09789356589317322, + "learning_rate": 1.2174792344469554e-06, + "loss": 0.0159, + "step": 105280 + }, + { + "epoch": 9.57007816760589, + "grad_norm": 0.24246609210968018, + "learning_rate": 1.215848433849429e-06, + "loss": 0.0145, + "step": 105290 + }, + { + "epoch": 9.570987093255772, + "grad_norm": 0.15601055324077606, + "learning_rate": 1.214218680067244e-06, + "loss": 0.015, + "step": 105300 + }, + { + "epoch": 9.571896018905653, + "grad_norm": 0.1732558012008667, + "learning_rate": 1.2125899732241709e-06, + "loss": 0.0168, + "step": 105310 + }, + { + "epoch": 9.572804944555536, + "grad_norm": 0.12177739292383194, + "learning_rate": 1.2109623134439007e-06, + "loss": 0.016, + "step": 105320 + }, + { + "epoch": 9.573713870205417, + "grad_norm": 0.12071380764245987, + "learning_rate": 1.209335700850045e-06, + "loss": 0.0163, + "step": 105330 + }, + { + "epoch": 9.5746227958553, + "grad_norm": 0.1567932367324829, + "learning_rate": 1.207710135566133e-06, + "loss": 0.0178, + "step": 105340 + }, + { + "epoch": 9.57553172150518, + "grad_norm": 0.2874896824359894, + "learning_rate": 1.2060856177156092e-06, + "loss": 0.0135, + "step": 105350 + }, + { + "epoch": 9.576440647155064, + "grad_norm": 0.12771624326705933, + "learning_rate": 1.2044621474218553e-06, + "loss": 0.0173, + "step": 105360 + }, + { + "epoch": 9.577349572804945, + "grad_norm": 0.17526663839817047, + "learning_rate": 1.202839724808159e-06, + "loss": 0.0153, + "step": 105370 + }, + { + "epoch": 9.578258498454826, + "grad_norm": 0.13392728567123413, + "learning_rate": 1.2012183499977303e-06, + "loss": 0.0133, + "step": 105380 + }, + { + "epoch": 9.579167424104709, + "grad_norm": 0.10362666100263596, + "learning_rate": 1.1995980231137038e-06, + "loss": 0.0179, + "step": 105390 + }, + { + "epoch": 9.58007634975459, + "grad_norm": 0.12544362246990204, + "learning_rate": 1.1979787442791363e-06, + "loss": 0.0172, + "step": 105400 + }, + { + "epoch": 9.580985275404473, + "grad_norm": 0.10952313244342804, + "learning_rate": 1.1963605136169959e-06, + "loss": 0.0161, + "step": 105410 + }, + { + "epoch": 9.581894201054354, + "grad_norm": 0.10160112380981445, + "learning_rate": 1.194743331250176e-06, + "loss": 0.0151, + "step": 105420 + }, + { + "epoch": 9.582803126704235, + "grad_norm": 0.23684340715408325, + "learning_rate": 1.1931271973014952e-06, + "loss": 0.0175, + "step": 105430 + }, + { + "epoch": 9.583712052354118, + "grad_norm": 0.09160847961902618, + "learning_rate": 1.1915121118936866e-06, + "loss": 0.0146, + "step": 105440 + }, + { + "epoch": 9.584620978003999, + "grad_norm": 0.1214049756526947, + "learning_rate": 1.1898980751494043e-06, + "loss": 0.0159, + "step": 105450 + }, + { + "epoch": 9.585529903653882, + "grad_norm": 0.1567125916481018, + "learning_rate": 1.1882850871912231e-06, + "loss": 0.0163, + "step": 105460 + }, + { + "epoch": 9.586438829303763, + "grad_norm": 0.11644519120454788, + "learning_rate": 1.1866731481416438e-06, + "loss": 0.0136, + "step": 105470 + }, + { + "epoch": 9.587347754953644, + "grad_norm": 0.11580824106931686, + "learning_rate": 1.1850622581230747e-06, + "loss": 0.0143, + "step": 105480 + }, + { + "epoch": 9.588256680603527, + "grad_norm": 0.13010047376155853, + "learning_rate": 1.1834524172578598e-06, + "loss": 0.0154, + "step": 105490 + }, + { + "epoch": 9.589165606253408, + "grad_norm": 0.10645253211259842, + "learning_rate": 1.1818436256682514e-06, + "loss": 0.0157, + "step": 105500 + }, + { + "epoch": 9.59007453190329, + "grad_norm": 0.11172500252723694, + "learning_rate": 1.1802358834764283e-06, + "loss": 0.0157, + "step": 105510 + }, + { + "epoch": 9.590983457553172, + "grad_norm": 0.15345001220703125, + "learning_rate": 1.178629190804486e-06, + "loss": 0.017, + "step": 105520 + }, + { + "epoch": 9.591892383203055, + "grad_norm": 0.12633194029331207, + "learning_rate": 1.1770235477744457e-06, + "loss": 0.017, + "step": 105530 + }, + { + "epoch": 9.592801308852936, + "grad_norm": 0.1406298577785492, + "learning_rate": 1.1754189545082394e-06, + "loss": 0.016, + "step": 105540 + }, + { + "epoch": 9.593710234502817, + "grad_norm": 0.16997899115085602, + "learning_rate": 1.1738154111277317e-06, + "loss": 0.0139, + "step": 105550 + }, + { + "epoch": 9.5946191601527, + "grad_norm": 0.1824723780155182, + "learning_rate": 1.1722129177546997e-06, + "loss": 0.0145, + "step": 105560 + }, + { + "epoch": 9.59552808580258, + "grad_norm": 0.11348533630371094, + "learning_rate": 1.1706114745108381e-06, + "loss": 0.0169, + "step": 105570 + }, + { + "epoch": 9.596437011452464, + "grad_norm": 0.09722401946783066, + "learning_rate": 1.1690110815177697e-06, + "loss": 0.014, + "step": 105580 + }, + { + "epoch": 9.597345937102345, + "grad_norm": 0.10945852845907211, + "learning_rate": 1.1674117388970367e-06, + "loss": 0.0146, + "step": 105590 + }, + { + "epoch": 9.598254862752228, + "grad_norm": 0.13692031800746918, + "learning_rate": 1.1658134467700927e-06, + "loss": 0.0166, + "step": 105600 + }, + { + "epoch": 9.599163788402109, + "grad_norm": 0.1167033314704895, + "learning_rate": 1.1642162052583182e-06, + "loss": 0.0143, + "step": 105610 + }, + { + "epoch": 9.60007271405199, + "grad_norm": 0.12701207399368286, + "learning_rate": 1.162620014483015e-06, + "loss": 0.0165, + "step": 105620 + }, + { + "epoch": 9.600981639701873, + "grad_norm": 0.12289734184741974, + "learning_rate": 1.1610248745654061e-06, + "loss": 0.0161, + "step": 105630 + }, + { + "epoch": 9.601890565351754, + "grad_norm": 0.1483953893184662, + "learning_rate": 1.1594307856266278e-06, + "loss": 0.0142, + "step": 105640 + }, + { + "epoch": 9.602799491001637, + "grad_norm": 0.1521347016096115, + "learning_rate": 1.1578377477877366e-06, + "loss": 0.0153, + "step": 105650 + }, + { + "epoch": 9.603708416651518, + "grad_norm": 0.14078208804130554, + "learning_rate": 1.1562457611697241e-06, + "loss": 0.0163, + "step": 105660 + }, + { + "epoch": 9.604617342301399, + "grad_norm": 0.18143713474273682, + "learning_rate": 1.1546548258934837e-06, + "loss": 0.0154, + "step": 105670 + }, + { + "epoch": 9.605526267951282, + "grad_norm": 0.12412941455841064, + "learning_rate": 1.1530649420798368e-06, + "loss": 0.0161, + "step": 105680 + }, + { + "epoch": 9.606435193601163, + "grad_norm": 0.17644470930099487, + "learning_rate": 1.1514761098495236e-06, + "loss": 0.0149, + "step": 105690 + }, + { + "epoch": 9.607344119251046, + "grad_norm": 0.31373825669288635, + "learning_rate": 1.1498883293232126e-06, + "loss": 0.0181, + "step": 105700 + }, + { + "epoch": 9.608253044900927, + "grad_norm": 0.166509747505188, + "learning_rate": 1.1483016006214793e-06, + "loss": 0.0143, + "step": 105710 + }, + { + "epoch": 9.609161970550808, + "grad_norm": 0.1387593001127243, + "learning_rate": 1.14671592386482e-06, + "loss": 0.0136, + "step": 105720 + }, + { + "epoch": 9.61007089620069, + "grad_norm": 0.09133870899677277, + "learning_rate": 1.145131299173669e-06, + "loss": 0.0153, + "step": 105730 + }, + { + "epoch": 9.610979821850572, + "grad_norm": 0.10934294015169144, + "learning_rate": 1.1435477266683614e-06, + "loss": 0.0168, + "step": 105740 + }, + { + "epoch": 9.611888747500455, + "grad_norm": 0.1770629733800888, + "learning_rate": 1.1419652064691561e-06, + "loss": 0.0154, + "step": 105750 + }, + { + "epoch": 9.612797673150336, + "grad_norm": 0.14387297630310059, + "learning_rate": 1.1403837386962384e-06, + "loss": 0.0149, + "step": 105760 + }, + { + "epoch": 9.613706598800219, + "grad_norm": 0.08129538595676422, + "learning_rate": 1.1388033234697127e-06, + "loss": 0.0148, + "step": 105770 + }, + { + "epoch": 9.6146155244501, + "grad_norm": 0.10060110688209534, + "learning_rate": 1.1372239609095992e-06, + "loss": 0.0177, + "step": 105780 + }, + { + "epoch": 9.615524450099981, + "grad_norm": 0.253196656703949, + "learning_rate": 1.135645651135837e-06, + "loss": 0.0159, + "step": 105790 + }, + { + "epoch": 9.616433375749864, + "grad_norm": 0.16125406324863434, + "learning_rate": 1.1340683942682905e-06, + "loss": 0.0147, + "step": 105800 + }, + { + "epoch": 9.617342301399745, + "grad_norm": 0.139334037899971, + "learning_rate": 1.1324921904267454e-06, + "loss": 0.0133, + "step": 105810 + }, + { + "epoch": 9.618251227049628, + "grad_norm": 0.08501077443361282, + "learning_rate": 1.1309170397308977e-06, + "loss": 0.0149, + "step": 105820 + }, + { + "epoch": 9.619160152699509, + "grad_norm": 0.10605644434690475, + "learning_rate": 1.1293429423003732e-06, + "loss": 0.0166, + "step": 105830 + }, + { + "epoch": 9.620069078349392, + "grad_norm": 0.1489717811346054, + "learning_rate": 1.1277698982547162e-06, + "loss": 0.0123, + "step": 105840 + }, + { + "epoch": 9.620978003999273, + "grad_norm": 0.10284051299095154, + "learning_rate": 1.126197907713385e-06, + "loss": 0.0172, + "step": 105850 + }, + { + "epoch": 9.621886929649154, + "grad_norm": 0.11237294971942902, + "learning_rate": 1.1246269707957652e-06, + "loss": 0.0163, + "step": 105860 + }, + { + "epoch": 9.622795855299037, + "grad_norm": 0.15513154864311218, + "learning_rate": 1.1230570876211549e-06, + "loss": 0.0166, + "step": 105870 + }, + { + "epoch": 9.623704780948918, + "grad_norm": 0.12819693982601166, + "learning_rate": 1.1214882583087805e-06, + "loss": 0.0162, + "step": 105880 + }, + { + "epoch": 9.624613706598801, + "grad_norm": 0.09546062350273132, + "learning_rate": 1.1199204829777814e-06, + "loss": 0.0163, + "step": 105890 + }, + { + "epoch": 9.625522632248682, + "grad_norm": 0.13336928188800812, + "learning_rate": 1.1183537617472246e-06, + "loss": 0.0138, + "step": 105900 + }, + { + "epoch": 9.626431557898563, + "grad_norm": 0.10220038145780563, + "learning_rate": 1.1167880947360842e-06, + "loss": 0.0157, + "step": 105910 + }, + { + "epoch": 9.627340483548446, + "grad_norm": 0.14956142008304596, + "learning_rate": 1.1152234820632672e-06, + "loss": 0.0158, + "step": 105920 + }, + { + "epoch": 9.628249409198327, + "grad_norm": 0.12149976193904877, + "learning_rate": 1.1136599238475986e-06, + "loss": 0.0156, + "step": 105930 + }, + { + "epoch": 9.62915833484821, + "grad_norm": 0.12690788507461548, + "learning_rate": 1.1120974202078133e-06, + "loss": 0.0148, + "step": 105940 + }, + { + "epoch": 9.630067260498091, + "grad_norm": 0.3099822402000427, + "learning_rate": 1.1105359712625778e-06, + "loss": 0.0157, + "step": 105950 + }, + { + "epoch": 9.630976186147974, + "grad_norm": 0.13324446976184845, + "learning_rate": 1.1089755771304743e-06, + "loss": 0.0142, + "step": 105960 + }, + { + "epoch": 9.631885111797855, + "grad_norm": 0.4547131359577179, + "learning_rate": 1.1074162379300046e-06, + "loss": 0.0159, + "step": 105970 + }, + { + "epoch": 9.632794037447736, + "grad_norm": 0.15906760096549988, + "learning_rate": 1.1058579537795838e-06, + "loss": 0.0156, + "step": 105980 + }, + { + "epoch": 9.633702963097619, + "grad_norm": 0.15031173825263977, + "learning_rate": 1.1043007247975605e-06, + "loss": 0.0139, + "step": 105990 + }, + { + "epoch": 9.6346118887475, + "grad_norm": 0.1260749101638794, + "learning_rate": 1.1027445511021939e-06, + "loss": 0.015, + "step": 106000 + }, + { + "epoch": 9.635520814397383, + "grad_norm": 0.11845077574253082, + "learning_rate": 1.1011894328116656e-06, + "loss": 0.0173, + "step": 106010 + }, + { + "epoch": 9.636429740047264, + "grad_norm": 0.15122708678245544, + "learning_rate": 1.0996353700440713e-06, + "loss": 0.017, + "step": 106020 + }, + { + "epoch": 9.637338665697145, + "grad_norm": 0.1575605273246765, + "learning_rate": 1.0980823629174403e-06, + "loss": 0.0165, + "step": 106030 + }, + { + "epoch": 9.638247591347028, + "grad_norm": 0.13738583028316498, + "learning_rate": 1.0965304115497094e-06, + "loss": 0.0163, + "step": 106040 + }, + { + "epoch": 9.63915651699691, + "grad_norm": 0.11599688977003098, + "learning_rate": 1.0949795160587361e-06, + "loss": 0.0157, + "step": 106050 + }, + { + "epoch": 9.640065442646792, + "grad_norm": 0.12168373167514801, + "learning_rate": 1.093429676562303e-06, + "loss": 0.0138, + "step": 106060 + }, + { + "epoch": 9.640974368296673, + "grad_norm": 0.19683484733104706, + "learning_rate": 1.0918808931781144e-06, + "loss": 0.0166, + "step": 106070 + }, + { + "epoch": 9.641883293946556, + "grad_norm": 0.17453569173812866, + "learning_rate": 1.090333166023786e-06, + "loss": 0.016, + "step": 106080 + }, + { + "epoch": 9.642792219596437, + "grad_norm": 0.15181927382946014, + "learning_rate": 1.0887864952168554e-06, + "loss": 0.0174, + "step": 106090 + }, + { + "epoch": 9.643701145246318, + "grad_norm": 0.1156793087720871, + "learning_rate": 1.0872408808747837e-06, + "loss": 0.0162, + "step": 106100 + }, + { + "epoch": 9.644610070896201, + "grad_norm": 0.1415800005197525, + "learning_rate": 1.0856963231149535e-06, + "loss": 0.0156, + "step": 106110 + }, + { + "epoch": 9.645518996546082, + "grad_norm": 0.10446072369813919, + "learning_rate": 1.0841528220546598e-06, + "loss": 0.0165, + "step": 106120 + }, + { + "epoch": 9.646427922195965, + "grad_norm": 0.14768481254577637, + "learning_rate": 1.0826103778111218e-06, + "loss": 0.0164, + "step": 106130 + }, + { + "epoch": 9.647336847845846, + "grad_norm": 0.0990675762295723, + "learning_rate": 1.0810689905014808e-06, + "loss": 0.0152, + "step": 106140 + }, + { + "epoch": 9.648245773495727, + "grad_norm": 0.11635320633649826, + "learning_rate": 1.0795286602427933e-06, + "loss": 0.0164, + "step": 106150 + }, + { + "epoch": 9.64915469914561, + "grad_norm": 0.13685888051986694, + "learning_rate": 1.0779893871520342e-06, + "loss": 0.0168, + "step": 106160 + }, + { + "epoch": 9.650063624795491, + "grad_norm": 0.11781377345323563, + "learning_rate": 1.0764511713461044e-06, + "loss": 0.0142, + "step": 106170 + }, + { + "epoch": 9.650972550445374, + "grad_norm": 0.13150209188461304, + "learning_rate": 1.0749140129418228e-06, + "loss": 0.0184, + "step": 106180 + }, + { + "epoch": 9.651881476095255, + "grad_norm": 0.13874156773090363, + "learning_rate": 1.0733779120559223e-06, + "loss": 0.0131, + "step": 106190 + }, + { + "epoch": 9.652790401745138, + "grad_norm": 0.2495000809431076, + "learning_rate": 1.0718428688050647e-06, + "loss": 0.017, + "step": 106200 + }, + { + "epoch": 9.65369932739502, + "grad_norm": 0.10976362228393555, + "learning_rate": 1.07030888330582e-06, + "loss": 0.0156, + "step": 106210 + }, + { + "epoch": 9.6546082530449, + "grad_norm": 0.11916086077690125, + "learning_rate": 1.0687759556746884e-06, + "loss": 0.0147, + "step": 106220 + }, + { + "epoch": 9.655517178694783, + "grad_norm": 0.11007575690746307, + "learning_rate": 1.067244086028087e-06, + "loss": 0.0153, + "step": 106230 + }, + { + "epoch": 9.656426104344664, + "grad_norm": 0.09778062999248505, + "learning_rate": 1.0657132744823456e-06, + "loss": 0.0135, + "step": 106240 + }, + { + "epoch": 9.657335029994547, + "grad_norm": 0.11074025928974152, + "learning_rate": 1.0641835211537265e-06, + "loss": 0.0128, + "step": 106250 + }, + { + "epoch": 9.658243955644428, + "grad_norm": 0.18292337656021118, + "learning_rate": 1.0626548261583986e-06, + "loss": 0.0148, + "step": 106260 + }, + { + "epoch": 9.659152881294311, + "grad_norm": 0.11325354129076004, + "learning_rate": 1.0611271896124585e-06, + "loss": 0.0128, + "step": 106270 + }, + { + "epoch": 9.660061806944192, + "grad_norm": 0.13119681179523468, + "learning_rate": 1.059600611631919e-06, + "loss": 0.0143, + "step": 106280 + }, + { + "epoch": 9.660970732594073, + "grad_norm": 0.13411200046539307, + "learning_rate": 1.0580750923327136e-06, + "loss": 0.0167, + "step": 106290 + }, + { + "epoch": 9.661879658243956, + "grad_norm": 0.1713848114013672, + "learning_rate": 1.0565506318307e-06, + "loss": 0.0141, + "step": 106300 + }, + { + "epoch": 9.662788583893837, + "grad_norm": 0.08443088829517365, + "learning_rate": 1.0550272302416453e-06, + "loss": 0.0138, + "step": 106310 + }, + { + "epoch": 9.66369750954372, + "grad_norm": 0.11465699225664139, + "learning_rate": 1.0535048876812402e-06, + "loss": 0.0153, + "step": 106320 + }, + { + "epoch": 9.664606435193601, + "grad_norm": 0.11600404977798462, + "learning_rate": 1.0519836042651042e-06, + "loss": 0.0158, + "step": 106330 + }, + { + "epoch": 9.665515360843482, + "grad_norm": 0.13891524076461792, + "learning_rate": 1.0504633801087648e-06, + "loss": 0.0156, + "step": 106340 + }, + { + "epoch": 9.666424286493365, + "grad_norm": 0.15379321575164795, + "learning_rate": 1.0489442153276696e-06, + "loss": 0.014, + "step": 106350 + }, + { + "epoch": 9.667333212143246, + "grad_norm": 0.19760844111442566, + "learning_rate": 1.047426110037193e-06, + "loss": 0.0142, + "step": 106360 + }, + { + "epoch": 9.66824213779313, + "grad_norm": 0.1525457799434662, + "learning_rate": 1.0459090643526275e-06, + "loss": 0.0165, + "step": 106370 + }, + { + "epoch": 9.66915106344301, + "grad_norm": 0.14710547029972076, + "learning_rate": 1.0443930783891782e-06, + "loss": 0.017, + "step": 106380 + }, + { + "epoch": 9.670059989092891, + "grad_norm": 0.24668508768081665, + "learning_rate": 1.0428781522619741e-06, + "loss": 0.0153, + "step": 106390 + }, + { + "epoch": 9.670968914742774, + "grad_norm": 0.13647273182868958, + "learning_rate": 1.041364286086065e-06, + "loss": 0.0136, + "step": 106400 + }, + { + "epoch": 9.671877840392655, + "grad_norm": 0.1056378185749054, + "learning_rate": 1.0398514799764226e-06, + "loss": 0.014, + "step": 106410 + }, + { + "epoch": 9.672786766042538, + "grad_norm": 0.1465841829776764, + "learning_rate": 1.0383397340479279e-06, + "loss": 0.0163, + "step": 106420 + }, + { + "epoch": 9.67369569169242, + "grad_norm": 0.16153758764266968, + "learning_rate": 1.0368290484153908e-06, + "loss": 0.0153, + "step": 106430 + }, + { + "epoch": 9.674604617342302, + "grad_norm": 0.09474332630634308, + "learning_rate": 1.0353194231935431e-06, + "loss": 0.0165, + "step": 106440 + }, + { + "epoch": 9.675513542992183, + "grad_norm": 0.17951413989067078, + "learning_rate": 1.0338108584970247e-06, + "loss": 0.0157, + "step": 106450 + }, + { + "epoch": 9.676422468642064, + "grad_norm": 0.46756601333618164, + "learning_rate": 1.0323033544404003e-06, + "loss": 0.0155, + "step": 106460 + }, + { + "epoch": 9.677331394291947, + "grad_norm": 0.21722114086151123, + "learning_rate": 1.0307969111381588e-06, + "loss": 0.0169, + "step": 106470 + }, + { + "epoch": 9.678240319941828, + "grad_norm": 0.11066920310258865, + "learning_rate": 1.0292915287047033e-06, + "loss": 0.0154, + "step": 106480 + }, + { + "epoch": 9.679149245591711, + "grad_norm": 0.14325040578842163, + "learning_rate": 1.0277872072543576e-06, + "loss": 0.0194, + "step": 106490 + }, + { + "epoch": 9.680058171241592, + "grad_norm": 0.17328155040740967, + "learning_rate": 1.0262839469013653e-06, + "loss": 0.0141, + "step": 106500 + }, + { + "epoch": 9.680967096891475, + "grad_norm": 0.12062229961156845, + "learning_rate": 1.0247817477598882e-06, + "loss": 0.0159, + "step": 106510 + }, + { + "epoch": 9.681876022541356, + "grad_norm": 0.12550586462020874, + "learning_rate": 1.02328060994401e-06, + "loss": 0.0191, + "step": 106520 + }, + { + "epoch": 9.682784948191237, + "grad_norm": 0.14420431852340698, + "learning_rate": 1.0217805335677282e-06, + "loss": 0.017, + "step": 106530 + }, + { + "epoch": 9.68369387384112, + "grad_norm": 0.13772691786289215, + "learning_rate": 1.0202815187449683e-06, + "loss": 0.016, + "step": 106540 + }, + { + "epoch": 9.684602799491001, + "grad_norm": 0.16707201302051544, + "learning_rate": 1.0187835655895706e-06, + "loss": 0.016, + "step": 106550 + }, + { + "epoch": 9.685511725140884, + "grad_norm": 0.1441461741924286, + "learning_rate": 1.017286674215291e-06, + "loss": 0.0152, + "step": 106560 + }, + { + "epoch": 9.686420650790765, + "grad_norm": 0.11600767076015472, + "learning_rate": 1.0157908447358117e-06, + "loss": 0.0147, + "step": 106570 + }, + { + "epoch": 9.687329576440646, + "grad_norm": 0.11910948902368546, + "learning_rate": 1.0142960772647285e-06, + "loss": 0.0139, + "step": 106580 + }, + { + "epoch": 9.68823850209053, + "grad_norm": 0.15990087389945984, + "learning_rate": 1.012802371915561e-06, + "loss": 0.0138, + "step": 106590 + }, + { + "epoch": 9.68914742774041, + "grad_norm": 0.15804678201675415, + "learning_rate": 1.0113097288017481e-06, + "loss": 0.0147, + "step": 106600 + }, + { + "epoch": 9.690056353390293, + "grad_norm": 0.13056595623493195, + "learning_rate": 1.0098181480366442e-06, + "loss": 0.0144, + "step": 106610 + }, + { + "epoch": 9.690965279040174, + "grad_norm": 0.12441833317279816, + "learning_rate": 1.008327629733522e-06, + "loss": 0.0146, + "step": 106620 + }, + { + "epoch": 9.691874204690055, + "grad_norm": 0.16555477678775787, + "learning_rate": 1.0068381740055792e-06, + "loss": 0.0162, + "step": 106630 + }, + { + "epoch": 9.692783130339938, + "grad_norm": 0.10302383452653885, + "learning_rate": 1.005349780965934e-06, + "loss": 0.0167, + "step": 106640 + }, + { + "epoch": 9.69369205598982, + "grad_norm": 0.1177801564335823, + "learning_rate": 1.0038624507276123e-06, + "loss": 0.0175, + "step": 106650 + }, + { + "epoch": 9.694600981639702, + "grad_norm": 0.13086062669754028, + "learning_rate": 1.0023761834035706e-06, + "loss": 0.0162, + "step": 106660 + }, + { + "epoch": 9.695509907289583, + "grad_norm": 0.11778031289577484, + "learning_rate": 1.000890979106685e-06, + "loss": 0.017, + "step": 106670 + }, + { + "epoch": 9.696418832939466, + "grad_norm": 0.15217848122119904, + "learning_rate": 9.99406837949744e-07, + "loss": 0.0148, + "step": 106680 + }, + { + "epoch": 9.697327758589347, + "grad_norm": 0.10914202034473419, + "learning_rate": 9.97923760045454e-07, + "loss": 0.0158, + "step": 106690 + }, + { + "epoch": 9.698236684239228, + "grad_norm": 0.13553187251091003, + "learning_rate": 9.964417455064495e-07, + "loss": 0.0168, + "step": 106700 + }, + { + "epoch": 9.699145609889111, + "grad_norm": 0.12656980752944946, + "learning_rate": 9.94960794445281e-07, + "loss": 0.0161, + "step": 106710 + }, + { + "epoch": 9.700054535538992, + "grad_norm": 0.1853780448436737, + "learning_rate": 9.93480906974415e-07, + "loss": 0.0163, + "step": 106720 + }, + { + "epoch": 9.700963461188875, + "grad_norm": 0.180792897939682, + "learning_rate": 9.920020832062348e-07, + "loss": 0.0171, + "step": 106730 + }, + { + "epoch": 9.701872386838756, + "grad_norm": 0.09601389616727829, + "learning_rate": 9.905243232530559e-07, + "loss": 0.0159, + "step": 106740 + }, + { + "epoch": 9.70278131248864, + "grad_norm": 0.12148711830377579, + "learning_rate": 9.890476272270998e-07, + "loss": 0.0144, + "step": 106750 + }, + { + "epoch": 9.70369023813852, + "grad_norm": 0.1359223872423172, + "learning_rate": 9.875719952405104e-07, + "loss": 0.0153, + "step": 106760 + }, + { + "epoch": 9.704599163788401, + "grad_norm": 0.1572616696357727, + "learning_rate": 9.860974274053547e-07, + "loss": 0.0161, + "step": 106770 + }, + { + "epoch": 9.705508089438284, + "grad_norm": 0.12250623106956482, + "learning_rate": 9.846239238336163e-07, + "loss": 0.016, + "step": 106780 + }, + { + "epoch": 9.706417015088165, + "grad_norm": 0.12190848588943481, + "learning_rate": 9.831514846371974e-07, + "loss": 0.0147, + "step": 106790 + }, + { + "epoch": 9.707325940738048, + "grad_norm": 0.6115052103996277, + "learning_rate": 9.816801099279204e-07, + "loss": 0.0152, + "step": 106800 + }, + { + "epoch": 9.70823486638793, + "grad_norm": 0.188559889793396, + "learning_rate": 9.80209799817524e-07, + "loss": 0.0145, + "step": 106810 + }, + { + "epoch": 9.70914379203781, + "grad_norm": 0.12240263819694519, + "learning_rate": 9.787405544176743e-07, + "loss": 0.0137, + "step": 106820 + }, + { + "epoch": 9.710052717687693, + "grad_norm": 0.15485142171382904, + "learning_rate": 9.772723738399448e-07, + "loss": 0.0129, + "step": 106830 + }, + { + "epoch": 9.710961643337575, + "grad_norm": 0.11722812056541443, + "learning_rate": 9.75805258195835e-07, + "loss": 0.0156, + "step": 106840 + }, + { + "epoch": 9.711870568987457, + "grad_norm": 0.1435403972864151, + "learning_rate": 9.743392075967688e-07, + "loss": 0.017, + "step": 106850 + }, + { + "epoch": 9.712779494637338, + "grad_norm": 0.1006111204624176, + "learning_rate": 9.728742221540744e-07, + "loss": 0.015, + "step": 106860 + }, + { + "epoch": 9.71368842028722, + "grad_norm": 0.20340311527252197, + "learning_rate": 9.714103019790137e-07, + "loss": 0.016, + "step": 106870 + }, + { + "epoch": 9.714597345937102, + "grad_norm": 0.10983390361070633, + "learning_rate": 9.699474471827585e-07, + "loss": 0.017, + "step": 106880 + }, + { + "epoch": 9.715506271586984, + "grad_norm": 0.11332561075687408, + "learning_rate": 9.68485657876408e-07, + "loss": 0.0159, + "step": 106890 + }, + { + "epoch": 9.716415197236866, + "grad_norm": 0.1134650781750679, + "learning_rate": 9.670249341709681e-07, + "loss": 0.0167, + "step": 106900 + }, + { + "epoch": 9.717324122886748, + "grad_norm": 0.13176238536834717, + "learning_rate": 9.65565276177377e-07, + "loss": 0.017, + "step": 106910 + }, + { + "epoch": 9.71823304853663, + "grad_norm": 0.11189384013414383, + "learning_rate": 9.641066840064827e-07, + "loss": 0.0158, + "step": 106920 + }, + { + "epoch": 9.719141974186511, + "grad_norm": 0.12082542479038239, + "learning_rate": 9.62649157769055e-07, + "loss": 0.0164, + "step": 106930 + }, + { + "epoch": 9.720050899836393, + "grad_norm": 0.14808009564876556, + "learning_rate": 9.6119269757579e-07, + "loss": 0.0148, + "step": 106940 + }, + { + "epoch": 9.720959825486275, + "grad_norm": 0.1376855969429016, + "learning_rate": 9.59737303537288e-07, + "loss": 0.0173, + "step": 106950 + }, + { + "epoch": 9.721868751136157, + "grad_norm": 0.09972478449344635, + "learning_rate": 9.582829757640804e-07, + "loss": 0.0134, + "step": 106960 + }, + { + "epoch": 9.72277767678604, + "grad_norm": 0.11360286921262741, + "learning_rate": 9.568297143666154e-07, + "loss": 0.0151, + "step": 106970 + }, + { + "epoch": 9.72368660243592, + "grad_norm": 0.11907684057950974, + "learning_rate": 9.553775194552566e-07, + "loss": 0.0167, + "step": 106980 + }, + { + "epoch": 9.724595528085803, + "grad_norm": 0.14899399876594543, + "learning_rate": 9.539263911402873e-07, + "loss": 0.0174, + "step": 106990 + }, + { + "epoch": 9.725504453735685, + "grad_norm": 0.11206747591495514, + "learning_rate": 9.524763295319127e-07, + "loss": 0.0137, + "step": 107000 + }, + { + "epoch": 9.726413379385566, + "grad_norm": 0.10514821857213974, + "learning_rate": 9.510273347402565e-07, + "loss": 0.0184, + "step": 107010 + }, + { + "epoch": 9.727322305035448, + "grad_norm": 0.15156719088554382, + "learning_rate": 9.495794068753588e-07, + "loss": 0.0175, + "step": 107020 + }, + { + "epoch": 9.72823123068533, + "grad_norm": 0.15495868027210236, + "learning_rate": 9.481325460471769e-07, + "loss": 0.0153, + "step": 107030 + }, + { + "epoch": 9.729140156335212, + "grad_norm": 0.10423485934734344, + "learning_rate": 9.466867523655992e-07, + "loss": 0.0172, + "step": 107040 + }, + { + "epoch": 9.730049081985094, + "grad_norm": 0.1008576974272728, + "learning_rate": 9.452420259404166e-07, + "loss": 0.0163, + "step": 107050 + }, + { + "epoch": 9.730958007634975, + "grad_norm": 0.1150919497013092, + "learning_rate": 9.437983668813477e-07, + "loss": 0.0151, + "step": 107060 + }, + { + "epoch": 9.731866933284858, + "grad_norm": 0.09917355328798294, + "learning_rate": 9.4235577529803e-07, + "loss": 0.0149, + "step": 107070 + }, + { + "epoch": 9.732775858934739, + "grad_norm": 0.13812652230262756, + "learning_rate": 9.409142513000224e-07, + "loss": 0.0134, + "step": 107080 + }, + { + "epoch": 9.733684784584621, + "grad_norm": 0.1268615573644638, + "learning_rate": 9.394737949967941e-07, + "loss": 0.0184, + "step": 107090 + }, + { + "epoch": 9.734593710234503, + "grad_norm": 0.113217793405056, + "learning_rate": 9.380344064977376e-07, + "loss": 0.0153, + "step": 107100 + }, + { + "epoch": 9.735502635884385, + "grad_norm": 0.12588347494602203, + "learning_rate": 9.36596085912167e-07, + "loss": 0.0138, + "step": 107110 + }, + { + "epoch": 9.736411561534267, + "grad_norm": 0.14211180806159973, + "learning_rate": 9.351588333493149e-07, + "loss": 0.0147, + "step": 107120 + }, + { + "epoch": 9.737320487184148, + "grad_norm": 0.1761690080165863, + "learning_rate": 9.337226489183275e-07, + "loss": 0.0149, + "step": 107130 + }, + { + "epoch": 9.73822941283403, + "grad_norm": 0.12007205933332443, + "learning_rate": 9.322875327282759e-07, + "loss": 0.0174, + "step": 107140 + }, + { + "epoch": 9.739138338483912, + "grad_norm": 0.1666489541530609, + "learning_rate": 9.308534848881495e-07, + "loss": 0.0162, + "step": 107150 + }, + { + "epoch": 9.740047264133795, + "grad_norm": 0.1409251093864441, + "learning_rate": 9.294205055068511e-07, + "loss": 0.0154, + "step": 107160 + }, + { + "epoch": 9.740956189783676, + "grad_norm": 0.12152803689241409, + "learning_rate": 9.279885946932088e-07, + "loss": 0.015, + "step": 107170 + }, + { + "epoch": 9.741865115433557, + "grad_norm": 0.11815223097801208, + "learning_rate": 9.265577525559637e-07, + "loss": 0.0156, + "step": 107180 + }, + { + "epoch": 9.74277404108344, + "grad_norm": 0.09894535690546036, + "learning_rate": 9.251279792037843e-07, + "loss": 0.0155, + "step": 107190 + }, + { + "epoch": 9.74368296673332, + "grad_norm": 0.12643976509571075, + "learning_rate": 9.23699274745245e-07, + "loss": 0.0156, + "step": 107200 + }, + { + "epoch": 9.744591892383204, + "grad_norm": 0.11812417209148407, + "learning_rate": 9.222716392888525e-07, + "loss": 0.015, + "step": 107210 + }, + { + "epoch": 9.745500818033085, + "grad_norm": 0.09725745022296906, + "learning_rate": 9.208450729430217e-07, + "loss": 0.0174, + "step": 107220 + }, + { + "epoch": 9.746409743682968, + "grad_norm": 0.09238822758197784, + "learning_rate": 9.194195758160944e-07, + "loss": 0.0167, + "step": 107230 + }, + { + "epoch": 9.747318669332849, + "grad_norm": 0.21702080965042114, + "learning_rate": 9.179951480163273e-07, + "loss": 0.0143, + "step": 107240 + }, + { + "epoch": 9.74822759498273, + "grad_norm": 0.11312313377857208, + "learning_rate": 9.165717896518938e-07, + "loss": 0.0137, + "step": 107250 + }, + { + "epoch": 9.749136520632613, + "grad_norm": 0.1430116891860962, + "learning_rate": 9.151495008308941e-07, + "loss": 0.0153, + "step": 107260 + }, + { + "epoch": 9.750045446282494, + "grad_norm": 0.16336293518543243, + "learning_rate": 9.137282816613335e-07, + "loss": 0.0168, + "step": 107270 + }, + { + "epoch": 9.750954371932377, + "grad_norm": 0.1690066009759903, + "learning_rate": 9.123081322511523e-07, + "loss": 0.0161, + "step": 107280 + }, + { + "epoch": 9.751863297582258, + "grad_norm": 0.08760269731283188, + "learning_rate": 9.108890527081942e-07, + "loss": 0.0147, + "step": 107290 + }, + { + "epoch": 9.752772223232139, + "grad_norm": 0.13052916526794434, + "learning_rate": 9.094710431402331e-07, + "loss": 0.0156, + "step": 107300 + }, + { + "epoch": 9.753681148882022, + "grad_norm": 0.1359696388244629, + "learning_rate": 9.080541036549594e-07, + "loss": 0.0168, + "step": 107310 + }, + { + "epoch": 9.754590074531903, + "grad_norm": 0.18936212360858917, + "learning_rate": 9.066382343599789e-07, + "loss": 0.016, + "step": 107320 + }, + { + "epoch": 9.755499000181786, + "grad_norm": 0.16427527368068695, + "learning_rate": 9.052234353628103e-07, + "loss": 0.016, + "step": 107330 + }, + { + "epoch": 9.756407925831667, + "grad_norm": 0.14062105119228363, + "learning_rate": 9.038097067709094e-07, + "loss": 0.0148, + "step": 107340 + }, + { + "epoch": 9.75731685148155, + "grad_norm": 0.1975901871919632, + "learning_rate": 9.023970486916339e-07, + "loss": 0.016, + "step": 107350 + }, + { + "epoch": 9.75822577713143, + "grad_norm": 0.25842225551605225, + "learning_rate": 9.009854612322643e-07, + "loss": 0.0176, + "step": 107360 + }, + { + "epoch": 9.759134702781312, + "grad_norm": 0.2773764133453369, + "learning_rate": 8.995749445000051e-07, + "loss": 0.015, + "step": 107370 + }, + { + "epoch": 9.760043628431195, + "grad_norm": 0.10930922627449036, + "learning_rate": 8.981654986019739e-07, + "loss": 0.0143, + "step": 107380 + }, + { + "epoch": 9.760952554081076, + "grad_norm": 0.1181912049651146, + "learning_rate": 8.967571236452116e-07, + "loss": 0.0149, + "step": 107390 + }, + { + "epoch": 9.761861479730959, + "grad_norm": 0.08469244092702866, + "learning_rate": 8.953498197366677e-07, + "loss": 0.0147, + "step": 107400 + }, + { + "epoch": 9.76277040538084, + "grad_norm": 0.07842153310775757, + "learning_rate": 8.939435869832235e-07, + "loss": 0.015, + "step": 107410 + }, + { + "epoch": 9.763679331030723, + "grad_norm": 0.24712087213993073, + "learning_rate": 8.925384254916752e-07, + "loss": 0.018, + "step": 107420 + }, + { + "epoch": 9.764588256680604, + "grad_norm": 0.14189596474170685, + "learning_rate": 8.91134335368729e-07, + "loss": 0.0159, + "step": 107430 + }, + { + "epoch": 9.765497182330485, + "grad_norm": 0.14197272062301636, + "learning_rate": 8.897313167210197e-07, + "loss": 0.0146, + "step": 107440 + }, + { + "epoch": 9.766406107980368, + "grad_norm": 0.21626050770282745, + "learning_rate": 8.883293696551003e-07, + "loss": 0.0142, + "step": 107450 + }, + { + "epoch": 9.767315033630249, + "grad_norm": 0.10903684794902802, + "learning_rate": 8.869284942774359e-07, + "loss": 0.0154, + "step": 107460 + }, + { + "epoch": 9.768223959280132, + "grad_norm": 0.12122239172458649, + "learning_rate": 8.85528690694411e-07, + "loss": 0.0154, + "step": 107470 + }, + { + "epoch": 9.769132884930013, + "grad_norm": 0.12386701256036758, + "learning_rate": 8.841299590123358e-07, + "loss": 0.0162, + "step": 107480 + }, + { + "epoch": 9.770041810579894, + "grad_norm": 0.12198945879936218, + "learning_rate": 8.827322993374354e-07, + "loss": 0.0141, + "step": 107490 + }, + { + "epoch": 9.770950736229777, + "grad_norm": 0.09083325415849686, + "learning_rate": 8.813357117758497e-07, + "loss": 0.015, + "step": 107500 + }, + { + "epoch": 9.771859661879658, + "grad_norm": 0.1349671185016632, + "learning_rate": 8.799401964336423e-07, + "loss": 0.0137, + "step": 107510 + }, + { + "epoch": 9.77276858752954, + "grad_norm": 0.1621108502149582, + "learning_rate": 8.785457534167918e-07, + "loss": 0.0153, + "step": 107520 + }, + { + "epoch": 9.773677513179422, + "grad_norm": 0.14598970115184784, + "learning_rate": 8.771523828311967e-07, + "loss": 0.0146, + "step": 107530 + }, + { + "epoch": 9.774586438829303, + "grad_norm": 0.14159941673278809, + "learning_rate": 8.757600847826775e-07, + "loss": 0.0166, + "step": 107540 + }, + { + "epoch": 9.775495364479186, + "grad_norm": 0.1286296248435974, + "learning_rate": 8.743688593769661e-07, + "loss": 0.0143, + "step": 107550 + }, + { + "epoch": 9.776404290129067, + "grad_norm": 0.12239359319210052, + "learning_rate": 8.729787067197215e-07, + "loss": 0.0159, + "step": 107560 + }, + { + "epoch": 9.77731321577895, + "grad_norm": 0.11289620399475098, + "learning_rate": 8.715896269165124e-07, + "loss": 0.0151, + "step": 107570 + }, + { + "epoch": 9.77822214142883, + "grad_norm": 0.1337796300649643, + "learning_rate": 8.70201620072833e-07, + "loss": 0.0178, + "step": 107580 + }, + { + "epoch": 9.779131067078714, + "grad_norm": 0.15860764682292938, + "learning_rate": 8.688146862940887e-07, + "loss": 0.016, + "step": 107590 + }, + { + "epoch": 9.780039992728595, + "grad_norm": 0.10672745108604431, + "learning_rate": 8.674288256856139e-07, + "loss": 0.017, + "step": 107600 + }, + { + "epoch": 9.780948918378476, + "grad_norm": 0.13271507620811462, + "learning_rate": 8.660440383526525e-07, + "loss": 0.0153, + "step": 107610 + }, + { + "epoch": 9.781857844028359, + "grad_norm": 0.3079392611980438, + "learning_rate": 8.646603244003721e-07, + "loss": 0.0178, + "step": 107620 + }, + { + "epoch": 9.78276676967824, + "grad_norm": 0.14007265865802765, + "learning_rate": 8.632776839338502e-07, + "loss": 0.013, + "step": 107630 + }, + { + "epoch": 9.783675695328123, + "grad_norm": 0.09699360281229019, + "learning_rate": 8.61896117058098e-07, + "loss": 0.0153, + "step": 107640 + }, + { + "epoch": 9.784584620978004, + "grad_norm": 0.150844007730484, + "learning_rate": 8.605156238780315e-07, + "loss": 0.0171, + "step": 107650 + }, + { + "epoch": 9.785493546627887, + "grad_norm": 0.12541338801383972, + "learning_rate": 8.591362044984902e-07, + "loss": 0.0152, + "step": 107660 + }, + { + "epoch": 9.786402472277768, + "grad_norm": 0.1310167759656906, + "learning_rate": 8.577578590242319e-07, + "loss": 0.0147, + "step": 107670 + }, + { + "epoch": 9.787311397927649, + "grad_norm": 0.14394362270832062, + "learning_rate": 8.563805875599362e-07, + "loss": 0.0142, + "step": 107680 + }, + { + "epoch": 9.788220323577532, + "grad_norm": 0.16632495820522308, + "learning_rate": 8.550043902101945e-07, + "loss": 0.0142, + "step": 107690 + }, + { + "epoch": 9.789129249227413, + "grad_norm": 0.09534835815429688, + "learning_rate": 8.536292670795181e-07, + "loss": 0.014, + "step": 107700 + }, + { + "epoch": 9.790038174877296, + "grad_norm": 0.12274777889251709, + "learning_rate": 8.52255218272342e-07, + "loss": 0.018, + "step": 107710 + }, + { + "epoch": 9.790947100527177, + "grad_norm": 0.14244601130485535, + "learning_rate": 8.508822438930175e-07, + "loss": 0.017, + "step": 107720 + }, + { + "epoch": 9.791856026177058, + "grad_norm": 0.11677169799804688, + "learning_rate": 8.495103440458097e-07, + "loss": 0.0149, + "step": 107730 + }, + { + "epoch": 9.79276495182694, + "grad_norm": 0.11100242286920547, + "learning_rate": 8.481395188349017e-07, + "loss": 0.0143, + "step": 107740 + }, + { + "epoch": 9.793673877476822, + "grad_norm": 0.13087549805641174, + "learning_rate": 8.467697683644071e-07, + "loss": 0.0179, + "step": 107750 + }, + { + "epoch": 9.794582803126705, + "grad_norm": 0.20357738435268402, + "learning_rate": 8.454010927383459e-07, + "loss": 0.0156, + "step": 107760 + }, + { + "epoch": 9.795491728776586, + "grad_norm": 0.15758952498435974, + "learning_rate": 8.440334920606568e-07, + "loss": 0.0144, + "step": 107770 + }, + { + "epoch": 9.796400654426467, + "grad_norm": 0.1831786334514618, + "learning_rate": 8.426669664352032e-07, + "loss": 0.0152, + "step": 107780 + }, + { + "epoch": 9.79730958007635, + "grad_norm": 0.21602743864059448, + "learning_rate": 8.413015159657656e-07, + "loss": 0.0156, + "step": 107790 + }, + { + "epoch": 9.798218505726231, + "grad_norm": 0.12267104536294937, + "learning_rate": 8.399371407560358e-07, + "loss": 0.0161, + "step": 107800 + }, + { + "epoch": 9.799127431376114, + "grad_norm": 0.09761404991149902, + "learning_rate": 8.385738409096344e-07, + "loss": 0.0149, + "step": 107810 + }, + { + "epoch": 9.800036357025995, + "grad_norm": 0.10920865833759308, + "learning_rate": 8.37211616530092e-07, + "loss": 0.0151, + "step": 107820 + }, + { + "epoch": 9.800945282675878, + "grad_norm": 0.11294656991958618, + "learning_rate": 8.358504677208606e-07, + "loss": 0.0162, + "step": 107830 + }, + { + "epoch": 9.801854208325759, + "grad_norm": 0.09512767940759659, + "learning_rate": 8.344903945853094e-07, + "loss": 0.0146, + "step": 107840 + }, + { + "epoch": 9.80276313397564, + "grad_norm": 0.13988234102725983, + "learning_rate": 8.33131397226729e-07, + "loss": 0.0155, + "step": 107850 + }, + { + "epoch": 9.803672059625523, + "grad_norm": 0.10258805751800537, + "learning_rate": 8.317734757483287e-07, + "loss": 0.0157, + "step": 107860 + }, + { + "epoch": 9.804580985275404, + "grad_norm": 0.11177822202444077, + "learning_rate": 8.304166302532274e-07, + "loss": 0.0152, + "step": 107870 + }, + { + "epoch": 9.805489910925287, + "grad_norm": 0.1966686099767685, + "learning_rate": 8.290608608444744e-07, + "loss": 0.0164, + "step": 107880 + }, + { + "epoch": 9.806398836575168, + "grad_norm": 0.24703511595726013, + "learning_rate": 8.277061676250275e-07, + "loss": 0.0149, + "step": 107890 + }, + { + "epoch": 9.80730776222505, + "grad_norm": 0.13351532816886902, + "learning_rate": 8.263525506977659e-07, + "loss": 0.0152, + "step": 107900 + }, + { + "epoch": 9.808216687874932, + "grad_norm": 0.10189245641231537, + "learning_rate": 8.250000101654942e-07, + "loss": 0.0148, + "step": 107910 + }, + { + "epoch": 9.809125613524813, + "grad_norm": 0.10659068822860718, + "learning_rate": 8.236485461309251e-07, + "loss": 0.018, + "step": 107920 + }, + { + "epoch": 9.810034539174696, + "grad_norm": 0.1957988739013672, + "learning_rate": 8.222981586966899e-07, + "loss": 0.0147, + "step": 107930 + }, + { + "epoch": 9.810943464824577, + "grad_norm": 0.11962583661079407, + "learning_rate": 8.209488479653465e-07, + "loss": 0.0147, + "step": 107940 + }, + { + "epoch": 9.81185239047446, + "grad_norm": 0.12524816393852234, + "learning_rate": 8.196006140393648e-07, + "loss": 0.0161, + "step": 107950 + }, + { + "epoch": 9.812761316124341, + "grad_norm": 0.13429392874240875, + "learning_rate": 8.182534570211325e-07, + "loss": 0.0136, + "step": 107960 + }, + { + "epoch": 9.813670241774222, + "grad_norm": 0.10515061020851135, + "learning_rate": 8.169073770129582e-07, + "loss": 0.0146, + "step": 107970 + }, + { + "epoch": 9.814579167424105, + "grad_norm": 0.17562802135944366, + "learning_rate": 8.155623741170715e-07, + "loss": 0.0173, + "step": 107980 + }, + { + "epoch": 9.815488093073986, + "grad_norm": 0.12399420142173767, + "learning_rate": 8.142184484356141e-07, + "loss": 0.0172, + "step": 107990 + }, + { + "epoch": 9.816397018723869, + "grad_norm": 0.13450278341770172, + "learning_rate": 8.128756000706427e-07, + "loss": 0.0149, + "step": 108000 + }, + { + "epoch": 9.81730594437375, + "grad_norm": 0.10553817451000214, + "learning_rate": 8.115338291241442e-07, + "loss": 0.0163, + "step": 108010 + }, + { + "epoch": 9.818214870023631, + "grad_norm": 0.12137605249881744, + "learning_rate": 8.101931356980185e-07, + "loss": 0.0156, + "step": 108020 + }, + { + "epoch": 9.819123795673514, + "grad_norm": 0.11667807400226593, + "learning_rate": 8.088535198940794e-07, + "loss": 0.0157, + "step": 108030 + }, + { + "epoch": 9.820032721323395, + "grad_norm": 0.10744304955005646, + "learning_rate": 8.07514981814057e-07, + "loss": 0.0144, + "step": 108040 + }, + { + "epoch": 9.820941646973278, + "grad_norm": 0.12919816374778748, + "learning_rate": 8.061775215596135e-07, + "loss": 0.0191, + "step": 108050 + }, + { + "epoch": 9.821850572623159, + "grad_norm": 0.18251541256904602, + "learning_rate": 8.048411392323157e-07, + "loss": 0.0146, + "step": 108060 + }, + { + "epoch": 9.822759498273042, + "grad_norm": 0.132619246840477, + "learning_rate": 8.035058349336527e-07, + "loss": 0.0188, + "step": 108070 + }, + { + "epoch": 9.823668423922923, + "grad_norm": 0.1026909202337265, + "learning_rate": 8.021716087650332e-07, + "loss": 0.0148, + "step": 108080 + }, + { + "epoch": 9.824577349572804, + "grad_norm": 0.12386191636323929, + "learning_rate": 8.008384608277831e-07, + "loss": 0.0151, + "step": 108090 + }, + { + "epoch": 9.825486275222687, + "grad_norm": 0.13031171262264252, + "learning_rate": 7.99506391223146e-07, + "loss": 0.0162, + "step": 108100 + }, + { + "epoch": 9.826395200872568, + "grad_norm": 0.15337757766246796, + "learning_rate": 7.981754000522812e-07, + "loss": 0.0127, + "step": 108110 + }, + { + "epoch": 9.827304126522451, + "grad_norm": 0.1437917798757553, + "learning_rate": 7.968454874162711e-07, + "loss": 0.015, + "step": 108120 + }, + { + "epoch": 9.828213052172332, + "grad_norm": 0.14445635676383972, + "learning_rate": 7.955166534161168e-07, + "loss": 0.0152, + "step": 108130 + }, + { + "epoch": 9.829121977822215, + "grad_norm": 0.10881959646940231, + "learning_rate": 7.941888981527274e-07, + "loss": 0.0142, + "step": 108140 + }, + { + "epoch": 9.830030903472096, + "grad_norm": 0.12947200238704681, + "learning_rate": 7.928622217269421e-07, + "loss": 0.0163, + "step": 108150 + }, + { + "epoch": 9.830939829121977, + "grad_norm": 0.14757710695266724, + "learning_rate": 7.915366242395139e-07, + "loss": 0.0149, + "step": 108160 + }, + { + "epoch": 9.83184875477186, + "grad_norm": 0.19604669511318207, + "learning_rate": 7.902121057911088e-07, + "loss": 0.0178, + "step": 108170 + }, + { + "epoch": 9.832757680421741, + "grad_norm": 0.10849522054195404, + "learning_rate": 7.888886664823214e-07, + "loss": 0.0155, + "step": 108180 + }, + { + "epoch": 9.833666606071624, + "grad_norm": 0.1036875918507576, + "learning_rate": 7.875663064136512e-07, + "loss": 0.0159, + "step": 108190 + }, + { + "epoch": 9.834575531721505, + "grad_norm": 0.13932585716247559, + "learning_rate": 7.862450256855281e-07, + "loss": 0.0149, + "step": 108200 + }, + { + "epoch": 9.835484457371386, + "grad_norm": 0.12718819081783295, + "learning_rate": 7.849248243982915e-07, + "loss": 0.0154, + "step": 108210 + }, + { + "epoch": 9.836393383021269, + "grad_norm": 0.1149863451719284, + "learning_rate": 7.836057026522048e-07, + "loss": 0.0175, + "step": 108220 + }, + { + "epoch": 9.83730230867115, + "grad_norm": 0.10768893361091614, + "learning_rate": 7.822876605474427e-07, + "loss": 0.0149, + "step": 108230 + }, + { + "epoch": 9.838211234321033, + "grad_norm": 0.12737016379833221, + "learning_rate": 7.809706981841036e-07, + "loss": 0.0146, + "step": 108240 + }, + { + "epoch": 9.839120159970914, + "grad_norm": 0.16615337133407593, + "learning_rate": 7.796548156622057e-07, + "loss": 0.0174, + "step": 108250 + }, + { + "epoch": 9.840029085620797, + "grad_norm": 0.12139099836349487, + "learning_rate": 7.783400130816759e-07, + "loss": 0.0169, + "step": 108260 + }, + { + "epoch": 9.840938011270678, + "grad_norm": 0.12073002010583878, + "learning_rate": 7.770262905423659e-07, + "loss": 0.0152, + "step": 108270 + }, + { + "epoch": 9.84184693692056, + "grad_norm": 0.1788141280412674, + "learning_rate": 7.757136481440491e-07, + "loss": 0.0141, + "step": 108280 + }, + { + "epoch": 9.842755862570442, + "grad_norm": 0.17272402346134186, + "learning_rate": 7.744020859864092e-07, + "loss": 0.015, + "step": 108290 + }, + { + "epoch": 9.843664788220323, + "grad_norm": 0.18175746500492096, + "learning_rate": 7.730916041690467e-07, + "loss": 0.0155, + "step": 108300 + }, + { + "epoch": 9.844573713870206, + "grad_norm": 0.14978887140750885, + "learning_rate": 7.717822027914883e-07, + "loss": 0.0186, + "step": 108310 + }, + { + "epoch": 9.845482639520087, + "grad_norm": 0.10429130494594574, + "learning_rate": 7.704738819531764e-07, + "loss": 0.015, + "step": 108320 + }, + { + "epoch": 9.846391565169968, + "grad_norm": 0.13430358469486237, + "learning_rate": 7.691666417534648e-07, + "loss": 0.0156, + "step": 108330 + }, + { + "epoch": 9.847300490819851, + "grad_norm": 0.10564488172531128, + "learning_rate": 7.678604822916274e-07, + "loss": 0.0139, + "step": 108340 + }, + { + "epoch": 9.848209416469732, + "grad_norm": 0.10218102484941483, + "learning_rate": 7.665554036668682e-07, + "loss": 0.0146, + "step": 108350 + }, + { + "epoch": 9.849118342119615, + "grad_norm": 0.29032424092292786, + "learning_rate": 7.652514059782911e-07, + "loss": 0.0162, + "step": 108360 + }, + { + "epoch": 9.850027267769496, + "grad_norm": 0.09200204908847809, + "learning_rate": 7.639484893249288e-07, + "loss": 0.0155, + "step": 108370 + }, + { + "epoch": 9.850936193419379, + "grad_norm": 0.10152026265859604, + "learning_rate": 7.626466538057286e-07, + "loss": 0.0157, + "step": 108380 + }, + { + "epoch": 9.85184511906926, + "grad_norm": 0.11392038315534592, + "learning_rate": 7.613458995195582e-07, + "loss": 0.016, + "step": 108390 + }, + { + "epoch": 9.852754044719141, + "grad_norm": 0.1312577873468399, + "learning_rate": 7.600462265651986e-07, + "loss": 0.0162, + "step": 108400 + }, + { + "epoch": 9.853662970369024, + "grad_norm": 0.1735302358865738, + "learning_rate": 7.587476350413525e-07, + "loss": 0.0164, + "step": 108410 + }, + { + "epoch": 9.854571896018905, + "grad_norm": 0.15589793026447296, + "learning_rate": 7.574501250466376e-07, + "loss": 0.0142, + "step": 108420 + }, + { + "epoch": 9.855480821668788, + "grad_norm": 0.16643518209457397, + "learning_rate": 7.561536966795968e-07, + "loss": 0.0137, + "step": 108430 + }, + { + "epoch": 9.85638974731867, + "grad_norm": 0.097032330930233, + "learning_rate": 7.548583500386779e-07, + "loss": 0.0138, + "step": 108440 + }, + { + "epoch": 9.85729867296855, + "grad_norm": 0.10721535980701447, + "learning_rate": 7.535640852222591e-07, + "loss": 0.0172, + "step": 108450 + }, + { + "epoch": 9.858207598618433, + "grad_norm": 0.1353415846824646, + "learning_rate": 7.522709023286317e-07, + "loss": 0.0153, + "step": 108460 + }, + { + "epoch": 9.859116524268314, + "grad_norm": 0.11404073238372803, + "learning_rate": 7.509788014559988e-07, + "loss": 0.017, + "step": 108470 + }, + { + "epoch": 9.860025449918197, + "grad_norm": 0.11448348313570023, + "learning_rate": 7.496877827024934e-07, + "loss": 0.013, + "step": 108480 + }, + { + "epoch": 9.860934375568078, + "grad_norm": 0.12742295861244202, + "learning_rate": 7.483978461661556e-07, + "loss": 0.017, + "step": 108490 + }, + { + "epoch": 9.861843301217961, + "grad_norm": 0.13736529648303986, + "learning_rate": 7.471089919449503e-07, + "loss": 0.0161, + "step": 108500 + }, + { + "epoch": 9.862752226867842, + "grad_norm": 0.13791367411613464, + "learning_rate": 7.458212201367559e-07, + "loss": 0.0161, + "step": 108510 + }, + { + "epoch": 9.863661152517723, + "grad_norm": 0.1493535190820694, + "learning_rate": 7.445345308393708e-07, + "loss": 0.0161, + "step": 108520 + }, + { + "epoch": 9.864570078167606, + "grad_norm": 0.16558398306369781, + "learning_rate": 7.432489241505086e-07, + "loss": 0.017, + "step": 108530 + }, + { + "epoch": 9.865479003817487, + "grad_norm": 0.13635492324829102, + "learning_rate": 7.419644001678027e-07, + "loss": 0.0158, + "step": 108540 + }, + { + "epoch": 9.86638792946737, + "grad_norm": 0.12353188544511795, + "learning_rate": 7.406809589888103e-07, + "loss": 0.0164, + "step": 108550 + }, + { + "epoch": 9.867296855117251, + "grad_norm": 0.15450628101825714, + "learning_rate": 7.393986007109915e-07, + "loss": 0.0158, + "step": 108560 + }, + { + "epoch": 9.868205780767134, + "grad_norm": 0.12912650406360626, + "learning_rate": 7.381173254317402e-07, + "loss": 0.0167, + "step": 108570 + }, + { + "epoch": 9.869114706417015, + "grad_norm": 0.10811211913824081, + "learning_rate": 7.36837133248357e-07, + "loss": 0.0164, + "step": 108580 + }, + { + "epoch": 9.870023632066896, + "grad_norm": 0.10787873715162277, + "learning_rate": 7.355580242580656e-07, + "loss": 0.0157, + "step": 108590 + }, + { + "epoch": 9.87093255771678, + "grad_norm": 0.11019565910100937, + "learning_rate": 7.342799985580035e-07, + "loss": 0.0162, + "step": 108600 + }, + { + "epoch": 9.87184148336666, + "grad_norm": 0.15011635422706604, + "learning_rate": 7.330030562452295e-07, + "loss": 0.0152, + "step": 108610 + }, + { + "epoch": 9.872750409016543, + "grad_norm": 0.23538829386234283, + "learning_rate": 7.317271974167211e-07, + "loss": 0.0134, + "step": 108620 + }, + { + "epoch": 9.873659334666424, + "grad_norm": 0.22473874688148499, + "learning_rate": 7.304524221693709e-07, + "loss": 0.0151, + "step": 108630 + }, + { + "epoch": 9.874568260316305, + "grad_norm": 0.1303132176399231, + "learning_rate": 7.29178730599983e-07, + "loss": 0.0141, + "step": 108640 + }, + { + "epoch": 9.875477185966188, + "grad_norm": 0.24328182637691498, + "learning_rate": 7.279061228052952e-07, + "loss": 0.015, + "step": 108650 + }, + { + "epoch": 9.87638611161607, + "grad_norm": 0.10776006430387497, + "learning_rate": 7.266345988819518e-07, + "loss": 0.0172, + "step": 108660 + }, + { + "epoch": 9.877295037265952, + "grad_norm": 0.10243411362171173, + "learning_rate": 7.253641589265103e-07, + "loss": 0.0157, + "step": 108670 + }, + { + "epoch": 9.878203962915833, + "grad_norm": 0.18612992763519287, + "learning_rate": 7.240948030354572e-07, + "loss": 0.0148, + "step": 108680 + }, + { + "epoch": 9.879112888565714, + "grad_norm": 0.09836644679307938, + "learning_rate": 7.228265313051935e-07, + "loss": 0.0164, + "step": 108690 + }, + { + "epoch": 9.880021814215597, + "grad_norm": 0.10983976721763611, + "learning_rate": 7.215593438320339e-07, + "loss": 0.0156, + "step": 108700 + }, + { + "epoch": 9.880930739865478, + "grad_norm": 0.15287642180919647, + "learning_rate": 7.202932407122098e-07, + "loss": 0.0146, + "step": 108710 + }, + { + "epoch": 9.881839665515361, + "grad_norm": 0.1373572200536728, + "learning_rate": 7.190282220418775e-07, + "loss": 0.0152, + "step": 108720 + }, + { + "epoch": 9.882748591165242, + "grad_norm": 0.18950189650058746, + "learning_rate": 7.177642879171053e-07, + "loss": 0.0174, + "step": 108730 + }, + { + "epoch": 9.883657516815125, + "grad_norm": 0.13492318987846375, + "learning_rate": 7.165014384338814e-07, + "loss": 0.0152, + "step": 108740 + }, + { + "epoch": 9.884566442465006, + "grad_norm": 0.2485324591398239, + "learning_rate": 7.152396736881089e-07, + "loss": 0.0154, + "step": 108750 + }, + { + "epoch": 9.885475368114887, + "grad_norm": 0.1361803263425827, + "learning_rate": 7.139789937756147e-07, + "loss": 0.0145, + "step": 108760 + }, + { + "epoch": 9.88638429376477, + "grad_norm": 0.10058701038360596, + "learning_rate": 7.127193987921354e-07, + "loss": 0.0169, + "step": 108770 + }, + { + "epoch": 9.887293219414651, + "grad_norm": 0.08539588749408722, + "learning_rate": 7.114608888333279e-07, + "loss": 0.0133, + "step": 108780 + }, + { + "epoch": 9.888202145064534, + "grad_norm": 0.09301537275314331, + "learning_rate": 7.102034639947708e-07, + "loss": 0.0172, + "step": 108790 + }, + { + "epoch": 9.889111070714415, + "grad_norm": 0.10039901733398438, + "learning_rate": 7.089471243719575e-07, + "loss": 0.0141, + "step": 108800 + }, + { + "epoch": 9.890019996364298, + "grad_norm": 0.12621648609638214, + "learning_rate": 7.076918700602953e-07, + "loss": 0.0159, + "step": 108810 + }, + { + "epoch": 9.89092892201418, + "grad_norm": 0.09055831283330917, + "learning_rate": 7.06437701155116e-07, + "loss": 0.0152, + "step": 108820 + }, + { + "epoch": 9.89183784766406, + "grad_norm": 0.11985063552856445, + "learning_rate": 7.051846177516635e-07, + "loss": 0.014, + "step": 108830 + }, + { + "epoch": 9.892746773313943, + "grad_norm": 0.16213713586330414, + "learning_rate": 7.039326199451002e-07, + "loss": 0.0164, + "step": 108840 + }, + { + "epoch": 9.893655698963824, + "grad_norm": 0.1444852501153946, + "learning_rate": 7.026817078305115e-07, + "loss": 0.0147, + "step": 108850 + }, + { + "epoch": 9.894564624613707, + "grad_norm": 0.18784382939338684, + "learning_rate": 7.014318815028914e-07, + "loss": 0.0165, + "step": 108860 + }, + { + "epoch": 9.895473550263588, + "grad_norm": 0.1428319364786148, + "learning_rate": 7.001831410571591e-07, + "loss": 0.0155, + "step": 108870 + }, + { + "epoch": 9.89638247591347, + "grad_norm": 0.14414218068122864, + "learning_rate": 6.989354865881436e-07, + "loss": 0.0163, + "step": 108880 + }, + { + "epoch": 9.897291401563352, + "grad_norm": 0.07567166537046432, + "learning_rate": 6.976889181906027e-07, + "loss": 0.015, + "step": 108890 + }, + { + "epoch": 9.898200327213234, + "grad_norm": 0.1124953180551529, + "learning_rate": 6.964434359591987e-07, + "loss": 0.0166, + "step": 108900 + }, + { + "epoch": 9.899109252863116, + "grad_norm": 0.14526240527629852, + "learning_rate": 6.95199039988521e-07, + "loss": 0.0159, + "step": 108910 + }, + { + "epoch": 9.900018178512997, + "grad_norm": 0.22187553346157074, + "learning_rate": 6.939557303730743e-07, + "loss": 0.014, + "step": 108920 + }, + { + "epoch": 9.900927104162879, + "grad_norm": 0.15727545320987701, + "learning_rate": 6.927135072072777e-07, + "loss": 0.0173, + "step": 108930 + }, + { + "epoch": 9.901836029812761, + "grad_norm": 0.08765674382448196, + "learning_rate": 6.914723705854692e-07, + "loss": 0.0147, + "step": 108940 + }, + { + "epoch": 9.902744955462643, + "grad_norm": 0.12126309424638748, + "learning_rate": 6.902323206019051e-07, + "loss": 0.0134, + "step": 108950 + }, + { + "epoch": 9.903653881112525, + "grad_norm": 0.124839648604393, + "learning_rate": 6.889933573507618e-07, + "loss": 0.0158, + "step": 108960 + }, + { + "epoch": 9.904562806762407, + "grad_norm": 0.12157502770423889, + "learning_rate": 6.877554809261289e-07, + "loss": 0.0152, + "step": 108970 + }, + { + "epoch": 9.90547173241229, + "grad_norm": 0.22843997180461884, + "learning_rate": 6.865186914220128e-07, + "loss": 0.0149, + "step": 108980 + }, + { + "epoch": 9.90638065806217, + "grad_norm": 0.14854562282562256, + "learning_rate": 6.852829889323436e-07, + "loss": 0.0161, + "step": 108990 + }, + { + "epoch": 9.907289583712052, + "grad_norm": 0.12231859564781189, + "learning_rate": 6.840483735509611e-07, + "loss": 0.0171, + "step": 109000 + }, + { + "epoch": 9.908198509361934, + "grad_norm": 0.09190637618303299, + "learning_rate": 6.82814845371627e-07, + "loss": 0.0149, + "step": 109010 + }, + { + "epoch": 9.909107435011816, + "grad_norm": 0.1762196570634842, + "learning_rate": 6.815824044880214e-07, + "loss": 0.0152, + "step": 109020 + }, + { + "epoch": 9.910016360661698, + "grad_norm": 0.13324327766895294, + "learning_rate": 6.803510509937394e-07, + "loss": 0.0178, + "step": 109030 + }, + { + "epoch": 9.91092528631158, + "grad_norm": 0.103465236723423, + "learning_rate": 6.791207849822945e-07, + "loss": 0.0172, + "step": 109040 + }, + { + "epoch": 9.911834211961462, + "grad_norm": 0.1273583471775055, + "learning_rate": 6.778916065471103e-07, + "loss": 0.0186, + "step": 109050 + }, + { + "epoch": 9.912743137611344, + "grad_norm": 0.1677308827638626, + "learning_rate": 6.76663515781547e-07, + "loss": 0.0141, + "step": 109060 + }, + { + "epoch": 9.913652063261225, + "grad_norm": 0.14425109326839447, + "learning_rate": 6.754365127788636e-07, + "loss": 0.0154, + "step": 109070 + }, + { + "epoch": 9.914560988911107, + "grad_norm": 0.19348131120204926, + "learning_rate": 6.742105976322405e-07, + "loss": 0.0151, + "step": 109080 + }, + { + "epoch": 9.915469914560989, + "grad_norm": 0.1514805108308792, + "learning_rate": 6.729857704347797e-07, + "loss": 0.0162, + "step": 109090 + }, + { + "epoch": 9.916378840210871, + "grad_norm": 0.14145661890506744, + "learning_rate": 6.717620312795036e-07, + "loss": 0.0162, + "step": 109100 + }, + { + "epoch": 9.917287765860753, + "grad_norm": 0.1443990170955658, + "learning_rate": 6.705393802593396e-07, + "loss": 0.015, + "step": 109110 + }, + { + "epoch": 9.918196691510634, + "grad_norm": 0.1363602727651596, + "learning_rate": 6.693178174671466e-07, + "loss": 0.0142, + "step": 109120 + }, + { + "epoch": 9.919105617160517, + "grad_norm": 0.12176219373941422, + "learning_rate": 6.680973429956888e-07, + "loss": 0.0149, + "step": 109130 + }, + { + "epoch": 9.920014542810398, + "grad_norm": 0.2221270203590393, + "learning_rate": 6.668779569376571e-07, + "loss": 0.0157, + "step": 109140 + }, + { + "epoch": 9.92092346846028, + "grad_norm": 0.2769482433795929, + "learning_rate": 6.656596593856523e-07, + "loss": 0.0155, + "step": 109150 + }, + { + "epoch": 9.921832394110162, + "grad_norm": 0.1261701136827469, + "learning_rate": 6.644424504321972e-07, + "loss": 0.0147, + "step": 109160 + }, + { + "epoch": 9.922741319760043, + "grad_norm": 0.13108070194721222, + "learning_rate": 6.632263301697345e-07, + "loss": 0.0146, + "step": 109170 + }, + { + "epoch": 9.923650245409926, + "grad_norm": 0.13398970663547516, + "learning_rate": 6.620112986906152e-07, + "loss": 0.0168, + "step": 109180 + }, + { + "epoch": 9.924559171059807, + "grad_norm": 0.14296092092990875, + "learning_rate": 6.607973560871173e-07, + "loss": 0.0146, + "step": 109190 + }, + { + "epoch": 9.92546809670969, + "grad_norm": 0.10513138771057129, + "learning_rate": 6.595845024514285e-07, + "loss": 0.016, + "step": 109200 + }, + { + "epoch": 9.92637702235957, + "grad_norm": 0.10080163180828094, + "learning_rate": 6.583727378756571e-07, + "loss": 0.0159, + "step": 109210 + }, + { + "epoch": 9.927285948009454, + "grad_norm": 0.1149703785777092, + "learning_rate": 6.571620624518326e-07, + "loss": 0.0182, + "step": 109220 + }, + { + "epoch": 9.928194873659335, + "grad_norm": 0.23035919666290283, + "learning_rate": 6.559524762718933e-07, + "loss": 0.0149, + "step": 109230 + }, + { + "epoch": 9.929103799309216, + "grad_norm": 0.11821715533733368, + "learning_rate": 6.547439794277005e-07, + "loss": 0.0149, + "step": 109240 + }, + { + "epoch": 9.930012724959099, + "grad_norm": 0.1071641743183136, + "learning_rate": 6.535365720110326e-07, + "loss": 0.0139, + "step": 109250 + }, + { + "epoch": 9.93092165060898, + "grad_norm": 0.12492687255144119, + "learning_rate": 6.523302541135845e-07, + "loss": 0.0158, + "step": 109260 + }, + { + "epoch": 9.931830576258863, + "grad_norm": 0.12962663173675537, + "learning_rate": 6.511250258269647e-07, + "loss": 0.013, + "step": 109270 + }, + { + "epoch": 9.932739501908744, + "grad_norm": 0.20213989913463593, + "learning_rate": 6.499208872427065e-07, + "loss": 0.0149, + "step": 109280 + }, + { + "epoch": 9.933648427558627, + "grad_norm": 0.12607437372207642, + "learning_rate": 6.487178384522552e-07, + "loss": 0.013, + "step": 109290 + }, + { + "epoch": 9.934557353208508, + "grad_norm": 0.10089826583862305, + "learning_rate": 6.475158795469743e-07, + "loss": 0.0145, + "step": 109300 + }, + { + "epoch": 9.935466278858389, + "grad_norm": 0.11967259645462036, + "learning_rate": 6.463150106181426e-07, + "loss": 0.0148, + "step": 109310 + }, + { + "epoch": 9.936375204508272, + "grad_norm": 0.13890525698661804, + "learning_rate": 6.451152317569587e-07, + "loss": 0.0172, + "step": 109320 + }, + { + "epoch": 9.937284130158153, + "grad_norm": 0.14383992552757263, + "learning_rate": 6.439165430545413e-07, + "loss": 0.0166, + "step": 109330 + }, + { + "epoch": 9.938193055808036, + "grad_norm": 0.0911528468132019, + "learning_rate": 6.42718944601921e-07, + "loss": 0.0134, + "step": 109340 + }, + { + "epoch": 9.939101981457917, + "grad_norm": 0.08777453750371933, + "learning_rate": 6.415224364900435e-07, + "loss": 0.0168, + "step": 109350 + }, + { + "epoch": 9.940010907107798, + "grad_norm": 0.1230335682630539, + "learning_rate": 6.403270188097827e-07, + "loss": 0.0163, + "step": 109360 + }, + { + "epoch": 9.94091983275768, + "grad_norm": 0.10322709381580353, + "learning_rate": 6.391326916519192e-07, + "loss": 0.0168, + "step": 109370 + }, + { + "epoch": 9.941828758407562, + "grad_norm": 0.14063064754009247, + "learning_rate": 6.379394551071521e-07, + "loss": 0.0145, + "step": 109380 + }, + { + "epoch": 9.942737684057445, + "grad_norm": 0.11402522027492523, + "learning_rate": 6.367473092661025e-07, + "loss": 0.0176, + "step": 109390 + }, + { + "epoch": 9.943646609707326, + "grad_norm": 0.2374735176563263, + "learning_rate": 6.355562542193078e-07, + "loss": 0.0156, + "step": 109400 + }, + { + "epoch": 9.944555535357209, + "grad_norm": 0.09941313415765762, + "learning_rate": 6.343662900572172e-07, + "loss": 0.016, + "step": 109410 + }, + { + "epoch": 9.94546446100709, + "grad_norm": 0.12252918630838394, + "learning_rate": 6.33177416870202e-07, + "loss": 0.0139, + "step": 109420 + }, + { + "epoch": 9.94637338665697, + "grad_norm": 0.16747024655342102, + "learning_rate": 6.319896347485466e-07, + "loss": 0.0162, + "step": 109430 + }, + { + "epoch": 9.947282312306854, + "grad_norm": 0.16157600283622742, + "learning_rate": 6.30802943782462e-07, + "loss": 0.0151, + "step": 109440 + }, + { + "epoch": 9.948191237956735, + "grad_norm": 0.12835335731506348, + "learning_rate": 6.296173440620628e-07, + "loss": 0.0132, + "step": 109450 + }, + { + "epoch": 9.949100163606618, + "grad_norm": 0.12106890976428986, + "learning_rate": 6.284328356773889e-07, + "loss": 0.0154, + "step": 109460 + }, + { + "epoch": 9.950009089256499, + "grad_norm": 0.14509069919586182, + "learning_rate": 6.272494187183997e-07, + "loss": 0.0145, + "step": 109470 + }, + { + "epoch": 9.95091801490638, + "grad_norm": 0.13404656946659088, + "learning_rate": 6.260670932749635e-07, + "loss": 0.0161, + "step": 109480 + }, + { + "epoch": 9.951826940556263, + "grad_norm": 0.24480381608009338, + "learning_rate": 6.24885859436875e-07, + "loss": 0.016, + "step": 109490 + }, + { + "epoch": 9.952735866206144, + "grad_norm": 0.12314805388450623, + "learning_rate": 6.237057172938343e-07, + "loss": 0.0138, + "step": 109500 + }, + { + "epoch": 9.953644791856027, + "grad_norm": 0.11412493139505386, + "learning_rate": 6.225266669354713e-07, + "loss": 0.0144, + "step": 109510 + }, + { + "epoch": 9.954553717505908, + "grad_norm": 0.18111459910869598, + "learning_rate": 6.213487084513242e-07, + "loss": 0.0169, + "step": 109520 + }, + { + "epoch": 9.95546264315579, + "grad_norm": 0.10326660424470901, + "learning_rate": 6.201718419308517e-07, + "loss": 0.0153, + "step": 109530 + }, + { + "epoch": 9.956371568805672, + "grad_norm": 0.3007878065109253, + "learning_rate": 6.189960674634287e-07, + "loss": 0.016, + "step": 109540 + }, + { + "epoch": 9.957280494455553, + "grad_norm": 0.12306605279445648, + "learning_rate": 6.178213851383474e-07, + "loss": 0.0159, + "step": 109550 + }, + { + "epoch": 9.958189420105436, + "grad_norm": 0.11754478514194489, + "learning_rate": 6.166477950448213e-07, + "loss": 0.0139, + "step": 109560 + }, + { + "epoch": 9.959098345755317, + "grad_norm": 0.11528030037879944, + "learning_rate": 6.154752972719691e-07, + "loss": 0.0157, + "step": 109570 + }, + { + "epoch": 9.9600072714052, + "grad_norm": 0.1868639886379242, + "learning_rate": 6.143038919088395e-07, + "loss": 0.014, + "step": 109580 + }, + { + "epoch": 9.96091619705508, + "grad_norm": 0.14212295413017273, + "learning_rate": 6.131335790443932e-07, + "loss": 0.0159, + "step": 109590 + }, + { + "epoch": 9.961825122704962, + "grad_norm": 0.1285598874092102, + "learning_rate": 6.119643587675073e-07, + "loss": 0.0173, + "step": 109600 + }, + { + "epoch": 9.962734048354845, + "grad_norm": 0.13449811935424805, + "learning_rate": 6.10796231166974e-07, + "loss": 0.0149, + "step": 109610 + }, + { + "epoch": 9.963642974004726, + "grad_norm": 0.0948534831404686, + "learning_rate": 6.096291963315059e-07, + "loss": 0.0165, + "step": 109620 + }, + { + "epoch": 9.964551899654609, + "grad_norm": 0.22172744572162628, + "learning_rate": 6.084632543497337e-07, + "loss": 0.0146, + "step": 109630 + }, + { + "epoch": 9.96546082530449, + "grad_norm": 0.09595538675785065, + "learning_rate": 6.072984053102032e-07, + "loss": 0.0156, + "step": 109640 + }, + { + "epoch": 9.966369750954373, + "grad_norm": 0.11885353922843933, + "learning_rate": 6.061346493013686e-07, + "loss": 0.0152, + "step": 109650 + }, + { + "epoch": 9.967278676604254, + "grad_norm": 0.14061102271080017, + "learning_rate": 6.049719864116227e-07, + "loss": 0.0182, + "step": 109660 + }, + { + "epoch": 9.968187602254135, + "grad_norm": 0.10416347533464432, + "learning_rate": 6.038104167292546e-07, + "loss": 0.0148, + "step": 109670 + }, + { + "epoch": 9.969096527904018, + "grad_norm": 0.1486363410949707, + "learning_rate": 6.026499403424757e-07, + "loss": 0.0171, + "step": 109680 + }, + { + "epoch": 9.970005453553899, + "grad_norm": 0.20869936048984528, + "learning_rate": 6.014905573394202e-07, + "loss": 0.0155, + "step": 109690 + }, + { + "epoch": 9.970914379203782, + "grad_norm": 0.12153011560440063, + "learning_rate": 6.00332267808138e-07, + "loss": 0.0155, + "step": 109700 + }, + { + "epoch": 9.971823304853663, + "grad_norm": 0.12160089612007141, + "learning_rate": 5.991750718365885e-07, + "loss": 0.015, + "step": 109710 + }, + { + "epoch": 9.972732230503546, + "grad_norm": 0.13230392336845398, + "learning_rate": 5.980189695126547e-07, + "loss": 0.0143, + "step": 109720 + }, + { + "epoch": 9.973641156153427, + "grad_norm": 0.08389860391616821, + "learning_rate": 5.968639609241349e-07, + "loss": 0.0161, + "step": 109730 + }, + { + "epoch": 9.974550081803308, + "grad_norm": 0.14116990566253662, + "learning_rate": 5.957100461587472e-07, + "loss": 0.0163, + "step": 109740 + }, + { + "epoch": 9.97545900745319, + "grad_norm": 0.12802593410015106, + "learning_rate": 5.945572253041198e-07, + "loss": 0.0152, + "step": 109750 + }, + { + "epoch": 9.976367933103072, + "grad_norm": 0.12463568896055222, + "learning_rate": 5.934054984478026e-07, + "loss": 0.0158, + "step": 109760 + }, + { + "epoch": 9.977276858752955, + "grad_norm": 0.09878823161125183, + "learning_rate": 5.922548656772641e-07, + "loss": 0.0139, + "step": 109770 + }, + { + "epoch": 9.978185784402836, + "grad_norm": 0.3033805787563324, + "learning_rate": 5.911053270798861e-07, + "loss": 0.0152, + "step": 109780 + }, + { + "epoch": 9.979094710052717, + "grad_norm": 0.2019449770450592, + "learning_rate": 5.899568827429668e-07, + "loss": 0.015, + "step": 109790 + }, + { + "epoch": 9.9800036357026, + "grad_norm": 0.14677973091602325, + "learning_rate": 5.888095327537235e-07, + "loss": 0.0157, + "step": 109800 + }, + { + "epoch": 9.980912561352481, + "grad_norm": 0.13796226680278778, + "learning_rate": 5.876632771992929e-07, + "loss": 0.0149, + "step": 109810 + }, + { + "epoch": 9.981821487002364, + "grad_norm": 0.14000539481639862, + "learning_rate": 5.865181161667221e-07, + "loss": 0.0135, + "step": 109820 + }, + { + "epoch": 9.982730412652245, + "grad_norm": 0.11949416995048523, + "learning_rate": 5.8537404974298e-07, + "loss": 0.0148, + "step": 109830 + }, + { + "epoch": 9.983639338302126, + "grad_norm": 0.12480267137289047, + "learning_rate": 5.842310780149502e-07, + "loss": 0.0163, + "step": 109840 + }, + { + "epoch": 9.984548263952009, + "grad_norm": 0.12724147737026215, + "learning_rate": 5.830892010694333e-07, + "loss": 0.0161, + "step": 109850 + }, + { + "epoch": 9.98545718960189, + "grad_norm": 0.21492348611354828, + "learning_rate": 5.819484189931501e-07, + "loss": 0.0157, + "step": 109860 + }, + { + "epoch": 9.986366115251773, + "grad_norm": 0.12101984024047852, + "learning_rate": 5.808087318727329e-07, + "loss": 0.0152, + "step": 109870 + }, + { + "epoch": 9.987275040901654, + "grad_norm": 0.17520006000995636, + "learning_rate": 5.796701397947341e-07, + "loss": 0.0164, + "step": 109880 + }, + { + "epoch": 9.988183966551537, + "grad_norm": 0.14006361365318298, + "learning_rate": 5.785326428456211e-07, + "loss": 0.0133, + "step": 109890 + }, + { + "epoch": 9.989092892201418, + "grad_norm": 0.12152432650327682, + "learning_rate": 5.773962411117834e-07, + "loss": 0.0169, + "step": 109900 + }, + { + "epoch": 9.990001817851299, + "grad_norm": 0.1277284026145935, + "learning_rate": 5.762609346795184e-07, + "loss": 0.0189, + "step": 109910 + }, + { + "epoch": 9.990910743501182, + "grad_norm": 0.21031466126441956, + "learning_rate": 5.751267236350454e-07, + "loss": 0.0148, + "step": 109920 + }, + { + "epoch": 9.991819669151063, + "grad_norm": 0.13327054679393768, + "learning_rate": 5.739936080645041e-07, + "loss": 0.0152, + "step": 109930 + }, + { + "epoch": 9.992728594800946, + "grad_norm": 0.08479277044534683, + "learning_rate": 5.728615880539456e-07, + "loss": 0.0148, + "step": 109940 + }, + { + "epoch": 9.993637520450827, + "grad_norm": 0.09562544524669647, + "learning_rate": 5.717306636893344e-07, + "loss": 0.0154, + "step": 109950 + }, + { + "epoch": 9.99454644610071, + "grad_norm": 0.10863112658262253, + "learning_rate": 5.70600835056565e-07, + "loss": 0.0134, + "step": 109960 + }, + { + "epoch": 9.995455371750591, + "grad_norm": 0.11801169067621231, + "learning_rate": 5.694721022414373e-07, + "loss": 0.0162, + "step": 109970 + }, + { + "epoch": 9.996364297400472, + "grad_norm": 0.10204499214887619, + "learning_rate": 5.683444653296676e-07, + "loss": 0.0119, + "step": 109980 + }, + { + "epoch": 9.997273223050355, + "grad_norm": 0.1950603574514389, + "learning_rate": 5.672179244068959e-07, + "loss": 0.0127, + "step": 109990 + }, + { + "epoch": 9.998182148700236, + "grad_norm": 0.16242119669914246, + "learning_rate": 5.660924795586753e-07, + "loss": 0.0177, + "step": 110000 + }, + { + "epoch": 9.999091074350119, + "grad_norm": 0.14274771511554718, + "learning_rate": 5.649681308704758e-07, + "loss": 0.0152, + "step": 110010 + }, + { + "epoch": 10.0, + "grad_norm": 0.11295881867408752, + "learning_rate": 5.638448784276822e-07, + "loss": 0.0148, + "step": 110020 + }, + { + "epoch": 10.000908925649881, + "grad_norm": 0.12912070751190186, + "learning_rate": 5.627227223156001e-07, + "loss": 0.0157, + "step": 110030 + }, + { + "epoch": 10.001817851299764, + "grad_norm": 0.21822357177734375, + "learning_rate": 5.616016626194492e-07, + "loss": 0.0153, + "step": 110040 + }, + { + "epoch": 10.002726776949645, + "grad_norm": 0.1774781495332718, + "learning_rate": 5.604816994243683e-07, + "loss": 0.0143, + "step": 110050 + }, + { + "epoch": 10.003635702599528, + "grad_norm": 0.11822286248207092, + "learning_rate": 5.593628328154077e-07, + "loss": 0.0126, + "step": 110060 + }, + { + "epoch": 10.004544628249409, + "grad_norm": 0.0958256721496582, + "learning_rate": 5.582450628775426e-07, + "loss": 0.0159, + "step": 110070 + }, + { + "epoch": 10.005453553899292, + "grad_norm": 0.08919943869113922, + "learning_rate": 5.571283896956569e-07, + "loss": 0.0159, + "step": 110080 + }, + { + "epoch": 10.006362479549173, + "grad_norm": 0.13989394903182983, + "learning_rate": 5.560128133545545e-07, + "loss": 0.0139, + "step": 110090 + }, + { + "epoch": 10.007271405199054, + "grad_norm": 0.13167250156402588, + "learning_rate": 5.548983339389557e-07, + "loss": 0.0138, + "step": 110100 + }, + { + "epoch": 10.008180330848937, + "grad_norm": 0.22635072469711304, + "learning_rate": 5.537849515335014e-07, + "loss": 0.0155, + "step": 110110 + }, + { + "epoch": 10.009089256498818, + "grad_norm": 0.12653467059135437, + "learning_rate": 5.526726662227421e-07, + "loss": 0.0165, + "step": 110120 + }, + { + "epoch": 10.009998182148701, + "grad_norm": 0.15685315430164337, + "learning_rate": 5.515614780911521e-07, + "loss": 0.0155, + "step": 110130 + }, + { + "epoch": 10.010907107798582, + "grad_norm": 0.10688085108995438, + "learning_rate": 5.504513872231137e-07, + "loss": 0.0161, + "step": 110140 + }, + { + "epoch": 10.011816033448463, + "grad_norm": 0.1440022885799408, + "learning_rate": 5.493423937029346e-07, + "loss": 0.0144, + "step": 110150 + }, + { + "epoch": 10.012724959098346, + "grad_norm": 0.13394683599472046, + "learning_rate": 5.48234497614834e-07, + "loss": 0.0172, + "step": 110160 + }, + { + "epoch": 10.013633884748227, + "grad_norm": 0.1325039565563202, + "learning_rate": 5.471276990429497e-07, + "loss": 0.0139, + "step": 110170 + }, + { + "epoch": 10.01454281039811, + "grad_norm": 0.11773300915956497, + "learning_rate": 5.460219980713377e-07, + "loss": 0.0148, + "step": 110180 + }, + { + "epoch": 10.015451736047991, + "grad_norm": 0.1681821346282959, + "learning_rate": 5.449173947839659e-07, + "loss": 0.015, + "step": 110190 + }, + { + "epoch": 10.016360661697874, + "grad_norm": 0.11935656517744064, + "learning_rate": 5.438138892647237e-07, + "loss": 0.0148, + "step": 110200 + }, + { + "epoch": 10.017269587347755, + "grad_norm": 0.1260422319173813, + "learning_rate": 5.427114815974127e-07, + "loss": 0.0144, + "step": 110210 + }, + { + "epoch": 10.018178512997636, + "grad_norm": 0.16554903984069824, + "learning_rate": 5.416101718657556e-07, + "loss": 0.0154, + "step": 110220 + }, + { + "epoch": 10.019087438647519, + "grad_norm": 0.10388590395450592, + "learning_rate": 5.405099601533892e-07, + "loss": 0.0157, + "step": 110230 + }, + { + "epoch": 10.0199963642974, + "grad_norm": 0.10503006726503372, + "learning_rate": 5.394108465438679e-07, + "loss": 0.0158, + "step": 110240 + }, + { + "epoch": 10.020905289947283, + "grad_norm": 0.13697420060634613, + "learning_rate": 5.383128311206586e-07, + "loss": 0.0172, + "step": 110250 + }, + { + "epoch": 10.021814215597164, + "grad_norm": 0.18799316883087158, + "learning_rate": 5.372159139671512e-07, + "loss": 0.0148, + "step": 110260 + }, + { + "epoch": 10.022723141247045, + "grad_norm": 0.07890411466360092, + "learning_rate": 5.361200951666522e-07, + "loss": 0.0145, + "step": 110270 + }, + { + "epoch": 10.023632066896928, + "grad_norm": 0.12525559961795807, + "learning_rate": 5.350253748023754e-07, + "loss": 0.0163, + "step": 110280 + }, + { + "epoch": 10.02454099254681, + "grad_norm": 0.1215386688709259, + "learning_rate": 5.339317529574606e-07, + "loss": 0.0137, + "step": 110290 + }, + { + "epoch": 10.025449918196692, + "grad_norm": 0.11549922078847885, + "learning_rate": 5.328392297149647e-07, + "loss": 0.0153, + "step": 110300 + }, + { + "epoch": 10.026358843846573, + "grad_norm": 0.16826725006103516, + "learning_rate": 5.317478051578534e-07, + "loss": 0.0161, + "step": 110310 + }, + { + "epoch": 10.027267769496456, + "grad_norm": 0.17871445417404175, + "learning_rate": 5.306574793690117e-07, + "loss": 0.0152, + "step": 110320 + }, + { + "epoch": 10.028176695146337, + "grad_norm": 0.11324118077754974, + "learning_rate": 5.295682524312467e-07, + "loss": 0.0154, + "step": 110330 + }, + { + "epoch": 10.029085620796218, + "grad_norm": 0.13134709000587463, + "learning_rate": 5.284801244272774e-07, + "loss": 0.0157, + "step": 110340 + }, + { + "epoch": 10.029994546446101, + "grad_norm": 0.18276609480381012, + "learning_rate": 5.273930954397393e-07, + "loss": 0.0159, + "step": 110350 + }, + { + "epoch": 10.030903472095982, + "grad_norm": 0.14355400204658508, + "learning_rate": 5.263071655511798e-07, + "loss": 0.0149, + "step": 110360 + }, + { + "epoch": 10.031812397745865, + "grad_norm": 0.12736253440380096, + "learning_rate": 5.252223348440776e-07, + "loss": 0.0154, + "step": 110370 + }, + { + "epoch": 10.032721323395746, + "grad_norm": 0.10229997336864471, + "learning_rate": 5.241386034008139e-07, + "loss": 0.015, + "step": 110380 + }, + { + "epoch": 10.033630249045627, + "grad_norm": 0.13256724178791046, + "learning_rate": 5.230559713036892e-07, + "loss": 0.0161, + "step": 110390 + }, + { + "epoch": 10.03453917469551, + "grad_norm": 0.13982948660850525, + "learning_rate": 5.219744386349229e-07, + "loss": 0.0177, + "step": 110400 + }, + { + "epoch": 10.035448100345391, + "grad_norm": 0.11940015107393265, + "learning_rate": 5.208940054766543e-07, + "loss": 0.0142, + "step": 110410 + }, + { + "epoch": 10.036357025995274, + "grad_norm": 0.13073018193244934, + "learning_rate": 5.198146719109309e-07, + "loss": 0.0153, + "step": 110420 + }, + { + "epoch": 10.037265951645155, + "grad_norm": 0.22407591342926025, + "learning_rate": 5.187364380197224e-07, + "loss": 0.0161, + "step": 110430 + }, + { + "epoch": 10.038174877295038, + "grad_norm": 0.12440039217472076, + "learning_rate": 5.176593038849131e-07, + "loss": 0.0155, + "step": 110440 + }, + { + "epoch": 10.03908380294492, + "grad_norm": 0.1299843043088913, + "learning_rate": 5.165832695883043e-07, + "loss": 0.0169, + "step": 110450 + }, + { + "epoch": 10.0399927285948, + "grad_norm": 0.14166903495788574, + "learning_rate": 5.155083352116141e-07, + "loss": 0.0154, + "step": 110460 + }, + { + "epoch": 10.040901654244683, + "grad_norm": 0.13733135163784027, + "learning_rate": 5.144345008364753e-07, + "loss": 0.0153, + "step": 110470 + }, + { + "epoch": 10.041810579894564, + "grad_norm": 0.13315901160240173, + "learning_rate": 5.133617665444429e-07, + "loss": 0.0154, + "step": 110480 + }, + { + "epoch": 10.042719505544447, + "grad_norm": 0.14719361066818237, + "learning_rate": 5.122901324169782e-07, + "loss": 0.0136, + "step": 110490 + }, + { + "epoch": 10.043628431194328, + "grad_norm": 0.17136090993881226, + "learning_rate": 5.112195985354695e-07, + "loss": 0.0169, + "step": 110500 + }, + { + "epoch": 10.04453735684421, + "grad_norm": 0.16063961386680603, + "learning_rate": 5.101501649812118e-07, + "loss": 0.0131, + "step": 110510 + }, + { + "epoch": 10.045446282494092, + "grad_norm": 0.12967924773693085, + "learning_rate": 5.090818318354285e-07, + "loss": 0.0166, + "step": 110520 + }, + { + "epoch": 10.046355208143973, + "grad_norm": 0.19971469044685364, + "learning_rate": 5.080145991792445e-07, + "loss": 0.0195, + "step": 110530 + }, + { + "epoch": 10.047264133793856, + "grad_norm": 0.21245121955871582, + "learning_rate": 5.069484670937152e-07, + "loss": 0.0179, + "step": 110540 + }, + { + "epoch": 10.048173059443737, + "grad_norm": 0.1455296277999878, + "learning_rate": 5.058834356598024e-07, + "loss": 0.0175, + "step": 110550 + }, + { + "epoch": 10.04908198509362, + "grad_norm": 0.11522455513477325, + "learning_rate": 5.048195049583898e-07, + "loss": 0.0163, + "step": 110560 + }, + { + "epoch": 10.049990910743501, + "grad_norm": 0.14035557210445404, + "learning_rate": 5.037566750702777e-07, + "loss": 0.0152, + "step": 110570 + }, + { + "epoch": 10.050899836393382, + "grad_norm": 0.154846653342247, + "learning_rate": 5.026949460761765e-07, + "loss": 0.0154, + "step": 110580 + }, + { + "epoch": 10.051808762043265, + "grad_norm": 0.09690159559249878, + "learning_rate": 5.016343180567217e-07, + "loss": 0.0159, + "step": 110590 + }, + { + "epoch": 10.052717687693146, + "grad_norm": 0.1315806359052658, + "learning_rate": 5.005747910924607e-07, + "loss": 0.0146, + "step": 110600 + }, + { + "epoch": 10.05362661334303, + "grad_norm": 0.1054791733622551, + "learning_rate": 4.995163652638556e-07, + "loss": 0.0155, + "step": 110610 + }, + { + "epoch": 10.05453553899291, + "grad_norm": 0.10360413789749146, + "learning_rate": 4.984590406512874e-07, + "loss": 0.015, + "step": 110620 + }, + { + "epoch": 10.055444464642791, + "grad_norm": 0.13724088668823242, + "learning_rate": 4.974028173350531e-07, + "loss": 0.0184, + "step": 110630 + }, + { + "epoch": 10.056353390292674, + "grad_norm": 0.1030259057879448, + "learning_rate": 4.963476953953672e-07, + "loss": 0.0141, + "step": 110640 + }, + { + "epoch": 10.057262315942555, + "grad_norm": 0.10093062371015549, + "learning_rate": 4.952936749123572e-07, + "loss": 0.0168, + "step": 110650 + }, + { + "epoch": 10.058171241592438, + "grad_norm": 0.17525330185890198, + "learning_rate": 4.942407559660689e-07, + "loss": 0.0171, + "step": 110660 + }, + { + "epoch": 10.05908016724232, + "grad_norm": 0.09782544523477554, + "learning_rate": 4.931889386364669e-07, + "loss": 0.014, + "step": 110670 + }, + { + "epoch": 10.059989092892202, + "grad_norm": 0.2359483689069748, + "learning_rate": 4.921382230034288e-07, + "loss": 0.0163, + "step": 110680 + }, + { + "epoch": 10.060898018542083, + "grad_norm": 0.13705803453922272, + "learning_rate": 4.910886091467476e-07, + "loss": 0.0167, + "step": 110690 + }, + { + "epoch": 10.061806944191964, + "grad_norm": 0.17035700380802155, + "learning_rate": 4.900400971461361e-07, + "loss": 0.0173, + "step": 110700 + }, + { + "epoch": 10.062715869841847, + "grad_norm": 0.15179839730262756, + "learning_rate": 4.889926870812222e-07, + "loss": 0.0158, + "step": 110710 + }, + { + "epoch": 10.063624795491728, + "grad_norm": 0.11714909970760345, + "learning_rate": 4.879463790315508e-07, + "loss": 0.0167, + "step": 110720 + }, + { + "epoch": 10.064533721141611, + "grad_norm": 0.21969150006771088, + "learning_rate": 4.869011730765799e-07, + "loss": 0.0148, + "step": 110730 + }, + { + "epoch": 10.065442646791492, + "grad_norm": 0.1334993541240692, + "learning_rate": 4.858570692956843e-07, + "loss": 0.0143, + "step": 110740 + }, + { + "epoch": 10.066351572441373, + "grad_norm": 0.1143890991806984, + "learning_rate": 4.848140677681623e-07, + "loss": 0.0171, + "step": 110750 + }, + { + "epoch": 10.067260498091256, + "grad_norm": 0.17195071280002594, + "learning_rate": 4.837721685732188e-07, + "loss": 0.0165, + "step": 110760 + }, + { + "epoch": 10.068169423741137, + "grad_norm": 0.09876701235771179, + "learning_rate": 4.827313717899807e-07, + "loss": 0.015, + "step": 110770 + }, + { + "epoch": 10.06907834939102, + "grad_norm": 0.11741326004266739, + "learning_rate": 4.816916774974894e-07, + "loss": 0.0164, + "step": 110780 + }, + { + "epoch": 10.069987275040901, + "grad_norm": 0.10248801112174988, + "learning_rate": 4.806530857747021e-07, + "loss": 0.0189, + "step": 110790 + }, + { + "epoch": 10.070896200690784, + "grad_norm": 0.28164374828338623, + "learning_rate": 4.796155967004939e-07, + "loss": 0.0159, + "step": 110800 + }, + { + "epoch": 10.071805126340665, + "grad_norm": 0.14257696270942688, + "learning_rate": 4.785792103536551e-07, + "loss": 0.0178, + "step": 110810 + }, + { + "epoch": 10.072714051990546, + "grad_norm": 0.11889950186014175, + "learning_rate": 4.775439268128928e-07, + "loss": 0.0169, + "step": 110820 + }, + { + "epoch": 10.07362297764043, + "grad_norm": 0.08176963776350021, + "learning_rate": 4.765097461568274e-07, + "loss": 0.0168, + "step": 110830 + }, + { + "epoch": 10.07453190329031, + "grad_norm": 0.16385029256343842, + "learning_rate": 4.7547666846400275e-07, + "loss": 0.016, + "step": 110840 + }, + { + "epoch": 10.075440828940193, + "grad_norm": 0.20534151792526245, + "learning_rate": 4.744446938128694e-07, + "loss": 0.0174, + "step": 110850 + }, + { + "epoch": 10.076349754590074, + "grad_norm": 0.1366693675518036, + "learning_rate": 4.734138222818013e-07, + "loss": 0.0127, + "step": 110860 + }, + { + "epoch": 10.077258680239956, + "grad_norm": 0.20487241446971893, + "learning_rate": 4.723840539490876e-07, + "loss": 0.0172, + "step": 110870 + }, + { + "epoch": 10.078167605889838, + "grad_norm": 0.24748608469963074, + "learning_rate": 4.713553888929306e-07, + "loss": 0.0148, + "step": 110880 + }, + { + "epoch": 10.07907653153972, + "grad_norm": 0.1626555323600769, + "learning_rate": 4.703278271914496e-07, + "loss": 0.0168, + "step": 110890 + }, + { + "epoch": 10.079985457189602, + "grad_norm": 0.18633733689785004, + "learning_rate": 4.6930136892268547e-07, + "loss": 0.0155, + "step": 110900 + }, + { + "epoch": 10.080894382839483, + "grad_norm": 0.09779210388660431, + "learning_rate": 4.6827601416458756e-07, + "loss": 0.0154, + "step": 110910 + }, + { + "epoch": 10.081803308489366, + "grad_norm": 0.10074475407600403, + "learning_rate": 4.6725176299502346e-07, + "loss": 0.0128, + "step": 110920 + }, + { + "epoch": 10.082712234139247, + "grad_norm": 0.12209907919168472, + "learning_rate": 4.662286154917794e-07, + "loss": 0.0138, + "step": 110930 + }, + { + "epoch": 10.083621159789129, + "grad_norm": 0.1168660819530487, + "learning_rate": 4.6520657173256155e-07, + "loss": 0.0165, + "step": 110940 + }, + { + "epoch": 10.084530085439011, + "grad_norm": 0.25441160798072815, + "learning_rate": 4.641856317949811e-07, + "loss": 0.0152, + "step": 110950 + }, + { + "epoch": 10.085439011088893, + "grad_norm": 0.19108550250530243, + "learning_rate": 4.63165795756571e-07, + "loss": 0.0182, + "step": 110960 + }, + { + "epoch": 10.086347936738775, + "grad_norm": 0.10147732496261597, + "learning_rate": 4.6214706369478775e-07, + "loss": 0.0152, + "step": 110970 + }, + { + "epoch": 10.087256862388656, + "grad_norm": 0.11112668365240097, + "learning_rate": 4.611294356869944e-07, + "loss": 0.016, + "step": 110980 + }, + { + "epoch": 10.088165788038538, + "grad_norm": 0.1513560563325882, + "learning_rate": 4.601129118104691e-07, + "loss": 0.0161, + "step": 110990 + }, + { + "epoch": 10.08907471368842, + "grad_norm": 0.09238162636756897, + "learning_rate": 4.590974921424135e-07, + "loss": 0.0168, + "step": 111000 + }, + { + "epoch": 10.089983639338302, + "grad_norm": 0.11373761296272278, + "learning_rate": 4.5808317675994416e-07, + "loss": 0.0143, + "step": 111010 + }, + { + "epoch": 10.090892564988184, + "grad_norm": 0.13560743629932404, + "learning_rate": 4.5706996574008787e-07, + "loss": 0.0149, + "step": 111020 + }, + { + "epoch": 10.091801490638066, + "grad_norm": 0.12230279296636581, + "learning_rate": 4.560578591597914e-07, + "loss": 0.0188, + "step": 111030 + }, + { + "epoch": 10.092710416287948, + "grad_norm": 0.14035280048847198, + "learning_rate": 4.550468570959199e-07, + "loss": 0.0177, + "step": 111040 + }, + { + "epoch": 10.09361934193783, + "grad_norm": 0.12397449463605881, + "learning_rate": 4.540369596252519e-07, + "loss": 0.0158, + "step": 111050 + }, + { + "epoch": 10.09452826758771, + "grad_norm": 0.1123838946223259, + "learning_rate": 4.530281668244812e-07, + "loss": 0.0152, + "step": 111060 + }, + { + "epoch": 10.095437193237593, + "grad_norm": 0.10053113847970963, + "learning_rate": 4.520204787702181e-07, + "loss": 0.0172, + "step": 111070 + }, + { + "epoch": 10.096346118887475, + "grad_norm": 0.1595657616853714, + "learning_rate": 4.5101389553899476e-07, + "loss": 0.0179, + "step": 111080 + }, + { + "epoch": 10.097255044537357, + "grad_norm": 0.12836909294128418, + "learning_rate": 4.500084172072499e-07, + "loss": 0.0157, + "step": 111090 + }, + { + "epoch": 10.098163970187239, + "grad_norm": 0.1276714950799942, + "learning_rate": 4.4900404385134266e-07, + "loss": 0.0148, + "step": 111100 + }, + { + "epoch": 10.09907289583712, + "grad_norm": 0.197264164686203, + "learning_rate": 4.4800077554755015e-07, + "loss": 0.0151, + "step": 111110 + }, + { + "epoch": 10.099981821487003, + "grad_norm": 0.14396154880523682, + "learning_rate": 4.4699861237206654e-07, + "loss": 0.0157, + "step": 111120 + }, + { + "epoch": 10.100890747136884, + "grad_norm": 0.1492292433977127, + "learning_rate": 4.459975544009942e-07, + "loss": 0.0179, + "step": 111130 + }, + { + "epoch": 10.101799672786766, + "grad_norm": 0.14409780502319336, + "learning_rate": 4.4499760171036075e-07, + "loss": 0.0161, + "step": 111140 + }, + { + "epoch": 10.102708598436648, + "grad_norm": 0.1681881844997406, + "learning_rate": 4.4399875437610374e-07, + "loss": 0.0162, + "step": 111150 + }, + { + "epoch": 10.10361752408653, + "grad_norm": 0.09873469918966293, + "learning_rate": 4.4300101247408074e-07, + "loss": 0.0141, + "step": 111160 + }, + { + "epoch": 10.104526449736412, + "grad_norm": 0.11134332418441772, + "learning_rate": 4.4200437608006294e-07, + "loss": 0.0151, + "step": 111170 + }, + { + "epoch": 10.105435375386293, + "grad_norm": 0.10257423669099808, + "learning_rate": 4.4100884526973806e-07, + "loss": 0.015, + "step": 111180 + }, + { + "epoch": 10.106344301036176, + "grad_norm": 0.12372908741235733, + "learning_rate": 4.400144201187123e-07, + "loss": 0.0162, + "step": 111190 + }, + { + "epoch": 10.107253226686057, + "grad_norm": 0.18046972155570984, + "learning_rate": 4.390211007025002e-07, + "loss": 0.0157, + "step": 111200 + }, + { + "epoch": 10.10816215233594, + "grad_norm": 0.12690523266792297, + "learning_rate": 4.3802888709654487e-07, + "loss": 0.0133, + "step": 111210 + }, + { + "epoch": 10.10907107798582, + "grad_norm": 0.12162168323993683, + "learning_rate": 4.3703777937619263e-07, + "loss": 0.0156, + "step": 111220 + }, + { + "epoch": 10.109980003635702, + "grad_norm": 0.1074204221367836, + "learning_rate": 4.3604777761671487e-07, + "loss": 0.015, + "step": 111230 + }, + { + "epoch": 10.110888929285585, + "grad_norm": 0.12462299317121506, + "learning_rate": 4.350588818932949e-07, + "loss": 0.0149, + "step": 111240 + }, + { + "epoch": 10.111797854935466, + "grad_norm": 0.12108778208494186, + "learning_rate": 4.340710922810326e-07, + "loss": 0.0143, + "step": 111250 + }, + { + "epoch": 10.112706780585349, + "grad_norm": 0.12906624376773834, + "learning_rate": 4.330844088549413e-07, + "loss": 0.016, + "step": 111260 + }, + { + "epoch": 10.11361570623523, + "grad_norm": 0.12482018023729324, + "learning_rate": 4.320988316899593e-07, + "loss": 0.0155, + "step": 111270 + }, + { + "epoch": 10.114524631885113, + "grad_norm": 0.11265060305595398, + "learning_rate": 4.311143608609319e-07, + "loss": 0.0175, + "step": 111280 + }, + { + "epoch": 10.115433557534994, + "grad_norm": 0.12699250876903534, + "learning_rate": 4.301309964426192e-07, + "loss": 0.0145, + "step": 111290 + }, + { + "epoch": 10.116342483184875, + "grad_norm": 0.1866028755903244, + "learning_rate": 4.291487385097065e-07, + "loss": 0.0146, + "step": 111300 + }, + { + "epoch": 10.117251408834758, + "grad_norm": 0.17269659042358398, + "learning_rate": 4.2816758713678903e-07, + "loss": 0.0162, + "step": 111310 + }, + { + "epoch": 10.118160334484639, + "grad_norm": 0.20160387456417084, + "learning_rate": 4.2718754239837567e-07, + "loss": 0.016, + "step": 111320 + }, + { + "epoch": 10.119069260134522, + "grad_norm": 0.09748171269893646, + "learning_rate": 4.2620860436889674e-07, + "loss": 0.0155, + "step": 111330 + }, + { + "epoch": 10.119978185784403, + "grad_norm": 0.1380191147327423, + "learning_rate": 4.252307731226945e-07, + "loss": 0.017, + "step": 111340 + }, + { + "epoch": 10.120887111434286, + "grad_norm": 0.13008339703083038, + "learning_rate": 4.2425404873403295e-07, + "loss": 0.016, + "step": 111350 + }, + { + "epoch": 10.121796037084167, + "grad_norm": 0.12592199444770813, + "learning_rate": 4.232784312770827e-07, + "loss": 0.0138, + "step": 111360 + }, + { + "epoch": 10.122704962734048, + "grad_norm": 0.14188408851623535, + "learning_rate": 4.2230392082593605e-07, + "loss": 0.0154, + "step": 111370 + }, + { + "epoch": 10.12361388838393, + "grad_norm": 0.13345034420490265, + "learning_rate": 4.213305174546056e-07, + "loss": 0.0144, + "step": 111380 + }, + { + "epoch": 10.124522814033812, + "grad_norm": 0.16862627863883972, + "learning_rate": 4.203582212370105e-07, + "loss": 0.0164, + "step": 111390 + }, + { + "epoch": 10.125431739683695, + "grad_norm": 0.10818278044462204, + "learning_rate": 4.1938703224699e-07, + "loss": 0.0152, + "step": 111400 + }, + { + "epoch": 10.126340665333576, + "grad_norm": 0.14406552910804749, + "learning_rate": 4.184169505583002e-07, + "loss": 0.0175, + "step": 111410 + }, + { + "epoch": 10.127249590983457, + "grad_norm": 0.12232468277215958, + "learning_rate": 4.1744797624461537e-07, + "loss": 0.0177, + "step": 111420 + }, + { + "epoch": 10.12815851663334, + "grad_norm": 0.1852586269378662, + "learning_rate": 4.164801093795184e-07, + "loss": 0.014, + "step": 111430 + }, + { + "epoch": 10.12906744228322, + "grad_norm": 0.12282033264636993, + "learning_rate": 4.1551335003651545e-07, + "loss": 0.0164, + "step": 111440 + }, + { + "epoch": 10.129976367933104, + "grad_norm": 0.12854520976543427, + "learning_rate": 4.145476982890228e-07, + "loss": 0.0137, + "step": 111450 + }, + { + "epoch": 10.130885293582985, + "grad_norm": 0.09208843111991882, + "learning_rate": 4.135831542103785e-07, + "loss": 0.0146, + "step": 111460 + }, + { + "epoch": 10.131794219232868, + "grad_norm": 0.14616888761520386, + "learning_rate": 4.1261971787383056e-07, + "loss": 0.0161, + "step": 111470 + }, + { + "epoch": 10.132703144882749, + "grad_norm": 0.14403896033763885, + "learning_rate": 4.1165738935254716e-07, + "loss": 0.0161, + "step": 111480 + }, + { + "epoch": 10.13361207053263, + "grad_norm": 0.10469312965869904, + "learning_rate": 4.1069616871961147e-07, + "loss": 0.0166, + "step": 111490 + }, + { + "epoch": 10.134520996182513, + "grad_norm": 0.20760414004325867, + "learning_rate": 4.0973605604801846e-07, + "loss": 0.0135, + "step": 111500 + }, + { + "epoch": 10.135429921832394, + "grad_norm": 0.13380694389343262, + "learning_rate": 4.0877705141068645e-07, + "loss": 0.0155, + "step": 111510 + }, + { + "epoch": 10.136338847482277, + "grad_norm": 0.1401909738779068, + "learning_rate": 4.078191548804422e-07, + "loss": 0.015, + "step": 111520 + }, + { + "epoch": 10.137247773132158, + "grad_norm": 0.11293312907218933, + "learning_rate": 4.068623665300342e-07, + "loss": 0.0151, + "step": 111530 + }, + { + "epoch": 10.138156698782039, + "grad_norm": 0.15003977715969086, + "learning_rate": 4.059066864321226e-07, + "loss": 0.0157, + "step": 111540 + }, + { + "epoch": 10.139065624431922, + "grad_norm": 0.12952542304992676, + "learning_rate": 4.049521146592877e-07, + "loss": 0.0167, + "step": 111550 + }, + { + "epoch": 10.139974550081803, + "grad_norm": 0.2505146265029907, + "learning_rate": 4.0399865128401824e-07, + "loss": 0.0168, + "step": 111560 + }, + { + "epoch": 10.140883475731686, + "grad_norm": 0.1208658292889595, + "learning_rate": 4.030462963787246e-07, + "loss": 0.0164, + "step": 111570 + }, + { + "epoch": 10.141792401381567, + "grad_norm": 0.11503562331199646, + "learning_rate": 4.020950500157372e-07, + "loss": 0.015, + "step": 111580 + }, + { + "epoch": 10.14270132703145, + "grad_norm": 0.08380598574876785, + "learning_rate": 4.0114491226729e-07, + "loss": 0.0147, + "step": 111590 + }, + { + "epoch": 10.14361025268133, + "grad_norm": 0.1106099858880043, + "learning_rate": 4.0019588320554355e-07, + "loss": 0.0174, + "step": 111600 + }, + { + "epoch": 10.144519178331212, + "grad_norm": 0.2898823320865631, + "learning_rate": 3.9924796290257025e-07, + "loss": 0.015, + "step": 111610 + }, + { + "epoch": 10.145428103981095, + "grad_norm": 0.23676754534244537, + "learning_rate": 3.983011514303575e-07, + "loss": 0.0142, + "step": 111620 + }, + { + "epoch": 10.146337029630976, + "grad_norm": 0.16997632384300232, + "learning_rate": 3.973554488608078e-07, + "loss": 0.0144, + "step": 111630 + }, + { + "epoch": 10.147245955280859, + "grad_norm": 0.12401235103607178, + "learning_rate": 3.9641085526574364e-07, + "loss": 0.0158, + "step": 111640 + }, + { + "epoch": 10.14815488093074, + "grad_norm": 0.11687759310007095, + "learning_rate": 3.9546737071690097e-07, + "loss": 0.0151, + "step": 111650 + }, + { + "epoch": 10.149063806580621, + "grad_norm": 0.11752137541770935, + "learning_rate": 3.9452499528592924e-07, + "loss": 0.0131, + "step": 111660 + }, + { + "epoch": 10.149972732230504, + "grad_norm": 0.16281744837760925, + "learning_rate": 3.9358372904439277e-07, + "loss": 0.0165, + "step": 111670 + }, + { + "epoch": 10.150881657880385, + "grad_norm": 0.12486811727285385, + "learning_rate": 3.926435720637828e-07, + "loss": 0.0154, + "step": 111680 + }, + { + "epoch": 10.151790583530268, + "grad_norm": 0.10298359394073486, + "learning_rate": 3.917045244154921e-07, + "loss": 0.0141, + "step": 111690 + }, + { + "epoch": 10.152699509180149, + "grad_norm": 0.13686995208263397, + "learning_rate": 3.907665861708354e-07, + "loss": 0.0148, + "step": 111700 + }, + { + "epoch": 10.153608434830032, + "grad_norm": 0.17924384772777557, + "learning_rate": 3.89829757401044e-07, + "loss": 0.0163, + "step": 111710 + }, + { + "epoch": 10.154517360479913, + "grad_norm": 0.10997623205184937, + "learning_rate": 3.8889403817726596e-07, + "loss": 0.0164, + "step": 111720 + }, + { + "epoch": 10.155426286129794, + "grad_norm": 0.09653464704751968, + "learning_rate": 3.8795942857055957e-07, + "loss": 0.0151, + "step": 111730 + }, + { + "epoch": 10.156335211779677, + "grad_norm": 0.26526379585266113, + "learning_rate": 3.870259286519029e-07, + "loss": 0.0161, + "step": 111740 + }, + { + "epoch": 10.157244137429558, + "grad_norm": 0.11296549439430237, + "learning_rate": 3.860935384921893e-07, + "loss": 0.0133, + "step": 111750 + }, + { + "epoch": 10.15815306307944, + "grad_norm": 0.13057975471019745, + "learning_rate": 3.851622581622305e-07, + "loss": 0.0166, + "step": 111760 + }, + { + "epoch": 10.159061988729322, + "grad_norm": 0.09319497644901276, + "learning_rate": 3.8423208773274654e-07, + "loss": 0.0144, + "step": 111770 + }, + { + "epoch": 10.159970914379203, + "grad_norm": 0.12033887207508087, + "learning_rate": 3.8330302727437924e-07, + "loss": 0.014, + "step": 111780 + }, + { + "epoch": 10.160879840029086, + "grad_norm": 0.14004598557949066, + "learning_rate": 3.8237507685768724e-07, + "loss": 0.013, + "step": 111790 + }, + { + "epoch": 10.161788765678967, + "grad_norm": 0.13238553702831268, + "learning_rate": 3.814482365531391e-07, + "loss": 0.015, + "step": 111800 + }, + { + "epoch": 10.16269769132885, + "grad_norm": 0.1665482372045517, + "learning_rate": 3.805225064311235e-07, + "loss": 0.0171, + "step": 111810 + }, + { + "epoch": 10.163606616978731, + "grad_norm": 0.10498996824026108, + "learning_rate": 3.795978865619426e-07, + "loss": 0.0146, + "step": 111820 + }, + { + "epoch": 10.164515542628614, + "grad_norm": 0.13292941451072693, + "learning_rate": 3.7867437701581854e-07, + "loss": 0.0146, + "step": 111830 + }, + { + "epoch": 10.165424468278495, + "grad_norm": 0.11673679947853088, + "learning_rate": 3.7775197786288017e-07, + "loss": 0.014, + "step": 111840 + }, + { + "epoch": 10.166333393928376, + "grad_norm": 0.09617530554533005, + "learning_rate": 3.7683068917318317e-07, + "loss": 0.0141, + "step": 111850 + }, + { + "epoch": 10.167242319578259, + "grad_norm": 0.10345597565174103, + "learning_rate": 3.759105110166883e-07, + "loss": 0.0147, + "step": 111860 + }, + { + "epoch": 10.16815124522814, + "grad_norm": 0.14910490810871124, + "learning_rate": 3.7499144346328117e-07, + "loss": 0.0147, + "step": 111870 + }, + { + "epoch": 10.169060170878023, + "grad_norm": 0.09337771683931351, + "learning_rate": 3.740734865827561e-07, + "loss": 0.0157, + "step": 111880 + }, + { + "epoch": 10.169969096527904, + "grad_norm": 0.13580575585365295, + "learning_rate": 3.731566404448272e-07, + "loss": 0.0151, + "step": 111890 + }, + { + "epoch": 10.170878022177785, + "grad_norm": 0.12250814586877823, + "learning_rate": 3.722409051191239e-07, + "loss": 0.014, + "step": 111900 + }, + { + "epoch": 10.171786947827668, + "grad_norm": 0.13107533752918243, + "learning_rate": 3.713262806751888e-07, + "loss": 0.0146, + "step": 111910 + }, + { + "epoch": 10.172695873477549, + "grad_norm": 0.11155318468809128, + "learning_rate": 3.704127671824814e-07, + "loss": 0.0173, + "step": 111920 + }, + { + "epoch": 10.173604799127432, + "grad_norm": 0.15661349892616272, + "learning_rate": 3.6950036471037786e-07, + "loss": 0.0168, + "step": 111930 + }, + { + "epoch": 10.174513724777313, + "grad_norm": 0.18195568025112152, + "learning_rate": 3.685890733281677e-07, + "loss": 0.0145, + "step": 111940 + }, + { + "epoch": 10.175422650427196, + "grad_norm": 0.08724375069141388, + "learning_rate": 3.6767889310505895e-07, + "loss": 0.0161, + "step": 111950 + }, + { + "epoch": 10.176331576077077, + "grad_norm": 0.12061804533004761, + "learning_rate": 3.667698241101747e-07, + "loss": 0.0162, + "step": 111960 + }, + { + "epoch": 10.177240501726958, + "grad_norm": 0.17254824936389923, + "learning_rate": 3.6586186641254794e-07, + "loss": 0.0167, + "step": 111970 + }, + { + "epoch": 10.178149427376841, + "grad_norm": 0.093277707695961, + "learning_rate": 3.6495502008113855e-07, + "loss": 0.0126, + "step": 111980 + }, + { + "epoch": 10.179058353026722, + "grad_norm": 0.1678168922662735, + "learning_rate": 3.640492851848132e-07, + "loss": 0.0189, + "step": 111990 + }, + { + "epoch": 10.179967278676605, + "grad_norm": 0.1307184249162674, + "learning_rate": 3.631446617923534e-07, + "loss": 0.0146, + "step": 112000 + }, + { + "epoch": 10.180876204326486, + "grad_norm": 0.1871691793203354, + "learning_rate": 3.622411499724626e-07, + "loss": 0.0151, + "step": 112010 + }, + { + "epoch": 10.181785129976367, + "grad_norm": 0.18779674172401428, + "learning_rate": 3.613387497937559e-07, + "loss": 0.0165, + "step": 112020 + }, + { + "epoch": 10.18269405562625, + "grad_norm": 0.14625751972198486, + "learning_rate": 3.604374613247652e-07, + "loss": 0.0154, + "step": 112030 + }, + { + "epoch": 10.183602981276131, + "grad_norm": 0.11290446668863297, + "learning_rate": 3.5953728463393565e-07, + "loss": 0.0168, + "step": 112040 + }, + { + "epoch": 10.184511906926014, + "grad_norm": 0.12440186738967896, + "learning_rate": 3.586382197896293e-07, + "loss": 0.0171, + "step": 112050 + }, + { + "epoch": 10.185420832575895, + "grad_norm": 0.24616266787052155, + "learning_rate": 3.5774026686012817e-07, + "loss": 0.0146, + "step": 112060 + }, + { + "epoch": 10.186329758225778, + "grad_norm": 0.09613075852394104, + "learning_rate": 3.5684342591362106e-07, + "loss": 0.0147, + "step": 112070 + }, + { + "epoch": 10.187238683875659, + "grad_norm": 0.11300881206989288, + "learning_rate": 3.559476970182185e-07, + "loss": 0.0162, + "step": 112080 + }, + { + "epoch": 10.18814760952554, + "grad_norm": 0.13411976397037506, + "learning_rate": 3.550530802419494e-07, + "loss": 0.0172, + "step": 112090 + }, + { + "epoch": 10.189056535175423, + "grad_norm": 0.09411630034446716, + "learning_rate": 3.541595756527477e-07, + "loss": 0.0145, + "step": 112100 + }, + { + "epoch": 10.189965460825304, + "grad_norm": 0.17583884298801422, + "learning_rate": 3.532671833184742e-07, + "loss": 0.0144, + "step": 112110 + }, + { + "epoch": 10.190874386475187, + "grad_norm": 0.096446193754673, + "learning_rate": 3.5237590330689793e-07, + "loss": 0.0157, + "step": 112120 + }, + { + "epoch": 10.191783312125068, + "grad_norm": 0.175863116979599, + "learning_rate": 3.5148573568570485e-07, + "loss": 0.0168, + "step": 112130 + }, + { + "epoch": 10.19269223777495, + "grad_norm": 0.13619348406791687, + "learning_rate": 3.5059668052249916e-07, + "loss": 0.0149, + "step": 112140 + }, + { + "epoch": 10.193601163424832, + "grad_norm": 0.11375832557678223, + "learning_rate": 3.497087378847985e-07, + "loss": 0.0158, + "step": 112150 + }, + { + "epoch": 10.194510089074713, + "grad_norm": 0.12485463172197342, + "learning_rate": 3.4882190784003397e-07, + "loss": 0.0143, + "step": 112160 + }, + { + "epoch": 10.195419014724596, + "grad_norm": 0.1123829185962677, + "learning_rate": 3.4793619045555505e-07, + "loss": 0.0138, + "step": 112170 + }, + { + "epoch": 10.196327940374477, + "grad_norm": 0.12210646271705627, + "learning_rate": 3.4705158579863116e-07, + "loss": 0.016, + "step": 112180 + }, + { + "epoch": 10.19723686602436, + "grad_norm": 0.1552659124135971, + "learning_rate": 3.461680939364353e-07, + "loss": 0.0158, + "step": 112190 + }, + { + "epoch": 10.198145791674241, + "grad_norm": 0.11104105412960052, + "learning_rate": 3.452857149360672e-07, + "loss": 0.0133, + "step": 112200 + }, + { + "epoch": 10.199054717324122, + "grad_norm": 0.12907813489437103, + "learning_rate": 3.444044488645348e-07, + "loss": 0.015, + "step": 112210 + }, + { + "epoch": 10.199963642974005, + "grad_norm": 0.10081078857183456, + "learning_rate": 3.43524295788768e-07, + "loss": 0.0154, + "step": 112220 + }, + { + "epoch": 10.200872568623886, + "grad_norm": 0.13115404546260834, + "learning_rate": 3.426452557756066e-07, + "loss": 0.0141, + "step": 112230 + }, + { + "epoch": 10.201781494273769, + "grad_norm": 0.13551121950149536, + "learning_rate": 3.417673288918055e-07, + "loss": 0.0153, + "step": 112240 + }, + { + "epoch": 10.20269041992365, + "grad_norm": 0.1365627646446228, + "learning_rate": 3.4089051520404144e-07, + "loss": 0.0152, + "step": 112250 + }, + { + "epoch": 10.203599345573533, + "grad_norm": 0.17622223496437073, + "learning_rate": 3.4001481477890284e-07, + "loss": 0.0164, + "step": 112260 + }, + { + "epoch": 10.204508271223414, + "grad_norm": 0.13157102465629578, + "learning_rate": 3.391402276828881e-07, + "loss": 0.0145, + "step": 112270 + }, + { + "epoch": 10.205417196873295, + "grad_norm": 0.13990549743175507, + "learning_rate": 3.382667539824225e-07, + "loss": 0.0155, + "step": 112280 + }, + { + "epoch": 10.206326122523178, + "grad_norm": 0.20427589118480682, + "learning_rate": 3.373943937438362e-07, + "loss": 0.0149, + "step": 112290 + }, + { + "epoch": 10.20723504817306, + "grad_norm": 0.1353849619626999, + "learning_rate": 3.365231470333813e-07, + "loss": 0.0151, + "step": 112300 + }, + { + "epoch": 10.208143973822942, + "grad_norm": 0.12074586004018784, + "learning_rate": 3.356530139172231e-07, + "loss": 0.0152, + "step": 112310 + }, + { + "epoch": 10.209052899472823, + "grad_norm": 0.12823139131069183, + "learning_rate": 3.3478399446144223e-07, + "loss": 0.0139, + "step": 112320 + }, + { + "epoch": 10.209961825122704, + "grad_norm": 0.09043359756469727, + "learning_rate": 3.3391608873203585e-07, + "loss": 0.0142, + "step": 112330 + }, + { + "epoch": 10.210870750772587, + "grad_norm": 0.12176848948001862, + "learning_rate": 3.3304929679491445e-07, + "loss": 0.0155, + "step": 112340 + }, + { + "epoch": 10.211779676422468, + "grad_norm": 0.38112062215805054, + "learning_rate": 3.321836187159039e-07, + "loss": 0.0155, + "step": 112350 + }, + { + "epoch": 10.212688602072351, + "grad_norm": 0.11917633563280106, + "learning_rate": 3.313190545607497e-07, + "loss": 0.0149, + "step": 112360 + }, + { + "epoch": 10.213597527722232, + "grad_norm": 0.1331474483013153, + "learning_rate": 3.3045560439510794e-07, + "loss": 0.0127, + "step": 112370 + }, + { + "epoch": 10.214506453372113, + "grad_norm": 0.08815555274486542, + "learning_rate": 3.295932682845526e-07, + "loss": 0.0156, + "step": 112380 + }, + { + "epoch": 10.215415379021996, + "grad_norm": 0.135409876704216, + "learning_rate": 3.287320462945731e-07, + "loss": 0.0136, + "step": 112390 + }, + { + "epoch": 10.216324304671877, + "grad_norm": 0.09773589670658112, + "learning_rate": 3.278719384905721e-07, + "loss": 0.0169, + "step": 112400 + }, + { + "epoch": 10.21723323032176, + "grad_norm": 0.1359173059463501, + "learning_rate": 3.270129449378689e-07, + "loss": 0.0145, + "step": 112410 + }, + { + "epoch": 10.218142155971641, + "grad_norm": 0.11299053579568863, + "learning_rate": 3.261550657016998e-07, + "loss": 0.0141, + "step": 112420 + }, + { + "epoch": 10.219051081621524, + "grad_norm": 0.1328568458557129, + "learning_rate": 3.25298300847216e-07, + "loss": 0.0146, + "step": 112430 + }, + { + "epoch": 10.219960007271405, + "grad_norm": 0.1490432471036911, + "learning_rate": 3.2444265043947875e-07, + "loss": 0.0153, + "step": 112440 + }, + { + "epoch": 10.220868932921286, + "grad_norm": 0.12904977798461914, + "learning_rate": 3.235881145434744e-07, + "loss": 0.0154, + "step": 112450 + }, + { + "epoch": 10.22177785857117, + "grad_norm": 0.12767533957958221, + "learning_rate": 3.2273469322409597e-07, + "loss": 0.0162, + "step": 112460 + }, + { + "epoch": 10.22268678422105, + "grad_norm": 0.124334916472435, + "learning_rate": 3.218823865461551e-07, + "loss": 0.0145, + "step": 112470 + }, + { + "epoch": 10.223595709870933, + "grad_norm": 0.16460014879703522, + "learning_rate": 3.2103119457438156e-07, + "loss": 0.0145, + "step": 112480 + }, + { + "epoch": 10.224504635520814, + "grad_norm": 0.15669132769107819, + "learning_rate": 3.201811173734154e-07, + "loss": 0.0152, + "step": 112490 + }, + { + "epoch": 10.225413561170697, + "grad_norm": 0.1009189561009407, + "learning_rate": 3.1933215500781486e-07, + "loss": 0.0152, + "step": 112500 + }, + { + "epoch": 10.226322486820578, + "grad_norm": 0.1928064376115799, + "learning_rate": 3.184843075420535e-07, + "loss": 0.0123, + "step": 112510 + }, + { + "epoch": 10.22723141247046, + "grad_norm": 0.17756769061088562, + "learning_rate": 3.176375750405197e-07, + "loss": 0.0143, + "step": 112520 + }, + { + "epoch": 10.228140338120342, + "grad_norm": 0.10260095447301865, + "learning_rate": 3.1679195756751543e-07, + "loss": 0.0141, + "step": 112530 + }, + { + "epoch": 10.229049263770223, + "grad_norm": 0.14011219143867493, + "learning_rate": 3.15947455187261e-07, + "loss": 0.0164, + "step": 112540 + }, + { + "epoch": 10.229958189420106, + "grad_norm": 0.10978474467992783, + "learning_rate": 3.1510406796389336e-07, + "loss": 0.014, + "step": 112550 + }, + { + "epoch": 10.230867115069987, + "grad_norm": 0.17393937706947327, + "learning_rate": 3.1426179596145964e-07, + "loss": 0.0171, + "step": 112560 + }, + { + "epoch": 10.231776040719868, + "grad_norm": 0.16542242467403412, + "learning_rate": 3.1342063924392373e-07, + "loss": 0.0177, + "step": 112570 + }, + { + "epoch": 10.232684966369751, + "grad_norm": 0.11290992796421051, + "learning_rate": 3.125805978751678e-07, + "loss": 0.0145, + "step": 112580 + }, + { + "epoch": 10.233593892019632, + "grad_norm": 0.13874569535255432, + "learning_rate": 3.117416719189875e-07, + "loss": 0.0138, + "step": 112590 + }, + { + "epoch": 10.234502817669515, + "grad_norm": 0.13126111030578613, + "learning_rate": 3.1090386143909186e-07, + "loss": 0.0145, + "step": 112600 + }, + { + "epoch": 10.235411743319396, + "grad_norm": 0.09378574788570404, + "learning_rate": 3.1006716649910835e-07, + "loss": 0.0151, + "step": 112610 + }, + { + "epoch": 10.23632066896928, + "grad_norm": 0.15656942129135132, + "learning_rate": 3.092315871625795e-07, + "loss": 0.0162, + "step": 112620 + }, + { + "epoch": 10.23722959461916, + "grad_norm": 0.12224388122558594, + "learning_rate": 3.0839712349296113e-07, + "loss": 0.0128, + "step": 112630 + }, + { + "epoch": 10.238138520269041, + "grad_norm": 0.0846441239118576, + "learning_rate": 3.0756377555362425e-07, + "loss": 0.0132, + "step": 112640 + }, + { + "epoch": 10.239047445918924, + "grad_norm": 0.13450269401073456, + "learning_rate": 3.0673154340785656e-07, + "loss": 0.0152, + "step": 112650 + }, + { + "epoch": 10.239956371568805, + "grad_norm": 0.12899534404277802, + "learning_rate": 3.0590042711886257e-07, + "loss": 0.014, + "step": 112660 + }, + { + "epoch": 10.240865297218688, + "grad_norm": 0.3573146164417267, + "learning_rate": 3.050704267497584e-07, + "loss": 0.0155, + "step": 112670 + }, + { + "epoch": 10.24177422286857, + "grad_norm": 0.08689213544130325, + "learning_rate": 3.042415423635736e-07, + "loss": 0.0158, + "step": 112680 + }, + { + "epoch": 10.24268314851845, + "grad_norm": 0.13023178279399872, + "learning_rate": 3.034137740232645e-07, + "loss": 0.0153, + "step": 112690 + }, + { + "epoch": 10.243592074168333, + "grad_norm": 0.12613630294799805, + "learning_rate": 3.0258712179168924e-07, + "loss": 0.0139, + "step": 112700 + }, + { + "epoch": 10.244500999818214, + "grad_norm": 0.23104147613048553, + "learning_rate": 3.017615857316258e-07, + "loss": 0.0168, + "step": 112710 + }, + { + "epoch": 10.245409925468097, + "grad_norm": 0.12071772664785385, + "learning_rate": 3.009371659057708e-07, + "loss": 0.0157, + "step": 112720 + }, + { + "epoch": 10.246318851117978, + "grad_norm": 0.10431080311536789, + "learning_rate": 3.001138623767341e-07, + "loss": 0.0165, + "step": 112730 + }, + { + "epoch": 10.247227776767861, + "grad_norm": 0.1410875916481018, + "learning_rate": 2.9929167520703737e-07, + "loss": 0.0147, + "step": 112740 + }, + { + "epoch": 10.248136702417742, + "grad_norm": 0.10777048766613007, + "learning_rate": 2.984706044591223e-07, + "loss": 0.0169, + "step": 112750 + }, + { + "epoch": 10.249045628067623, + "grad_norm": 0.21279114484786987, + "learning_rate": 2.976506501953424e-07, + "loss": 0.0153, + "step": 112760 + }, + { + "epoch": 10.249954553717506, + "grad_norm": 0.133260115981102, + "learning_rate": 2.968318124779695e-07, + "loss": 0.0147, + "step": 112770 + }, + { + "epoch": 10.250863479367387, + "grad_norm": 0.16977877914905548, + "learning_rate": 2.9601409136918723e-07, + "loss": 0.0164, + "step": 112780 + }, + { + "epoch": 10.25177240501727, + "grad_norm": 0.12054866552352905, + "learning_rate": 2.951974869310958e-07, + "loss": 0.0145, + "step": 112790 + }, + { + "epoch": 10.252681330667151, + "grad_norm": 0.22378870844841003, + "learning_rate": 2.943819992257141e-07, + "loss": 0.0142, + "step": 112800 + }, + { + "epoch": 10.253590256317032, + "grad_norm": 0.14056552946567535, + "learning_rate": 2.9356762831497076e-07, + "loss": 0.0137, + "step": 112810 + }, + { + "epoch": 10.254499181966915, + "grad_norm": 0.17358244955539703, + "learning_rate": 2.927543742607114e-07, + "loss": 0.0147, + "step": 112820 + }, + { + "epoch": 10.255408107616796, + "grad_norm": 0.13369444012641907, + "learning_rate": 2.9194223712469825e-07, + "loss": 0.0153, + "step": 112830 + }, + { + "epoch": 10.25631703326668, + "grad_norm": 0.10378063470125198, + "learning_rate": 2.9113121696860537e-07, + "loss": 0.0166, + "step": 112840 + }, + { + "epoch": 10.25722595891656, + "grad_norm": 0.13244017958641052, + "learning_rate": 2.9032131385403004e-07, + "loss": 0.0168, + "step": 112850 + }, + { + "epoch": 10.258134884566443, + "grad_norm": 0.18916039168834686, + "learning_rate": 2.8951252784247317e-07, + "loss": 0.0157, + "step": 112860 + }, + { + "epoch": 10.259043810216324, + "grad_norm": 0.12602803111076355, + "learning_rate": 2.8870485899535893e-07, + "loss": 0.0169, + "step": 112870 + }, + { + "epoch": 10.259952735866205, + "grad_norm": 0.15098409354686737, + "learning_rate": 2.878983073740249e-07, + "loss": 0.0146, + "step": 112880 + }, + { + "epoch": 10.260861661516088, + "grad_norm": 0.09524811059236526, + "learning_rate": 2.8709287303972373e-07, + "loss": 0.0143, + "step": 112890 + }, + { + "epoch": 10.26177058716597, + "grad_norm": 0.14898033440113068, + "learning_rate": 2.862885560536216e-07, + "loss": 0.0152, + "step": 112900 + }, + { + "epoch": 10.262679512815852, + "grad_norm": 0.26956674456596375, + "learning_rate": 2.854853564768012e-07, + "loss": 0.0146, + "step": 112910 + }, + { + "epoch": 10.263588438465733, + "grad_norm": 0.17394526302814484, + "learning_rate": 2.846832743702621e-07, + "loss": 0.0149, + "step": 112920 + }, + { + "epoch": 10.264497364115615, + "grad_norm": 0.11143089085817337, + "learning_rate": 2.838823097949156e-07, + "loss": 0.0153, + "step": 112930 + }, + { + "epoch": 10.265406289765497, + "grad_norm": 0.20572707056999207, + "learning_rate": 2.830824628115863e-07, + "loss": 0.015, + "step": 112940 + }, + { + "epoch": 10.266315215415379, + "grad_norm": 0.12121173739433289, + "learning_rate": 2.822837334810241e-07, + "loss": 0.0154, + "step": 112950 + }, + { + "epoch": 10.267224141065261, + "grad_norm": 0.10563413798809052, + "learning_rate": 2.814861218638837e-07, + "loss": 0.0163, + "step": 112960 + }, + { + "epoch": 10.268133066715142, + "grad_norm": 0.11753861606121063, + "learning_rate": 2.8068962802073995e-07, + "loss": 0.0127, + "step": 112970 + }, + { + "epoch": 10.269041992365025, + "grad_norm": 0.14372016489505768, + "learning_rate": 2.7989425201207453e-07, + "loss": 0.0174, + "step": 112980 + }, + { + "epoch": 10.269950918014906, + "grad_norm": 0.13874410092830658, + "learning_rate": 2.7909999389830076e-07, + "loss": 0.0141, + "step": 112990 + }, + { + "epoch": 10.270859843664788, + "grad_norm": 0.11403944343328476, + "learning_rate": 2.7830685373973375e-07, + "loss": 0.0159, + "step": 113000 + }, + { + "epoch": 10.27176876931467, + "grad_norm": 0.10212542116641998, + "learning_rate": 2.7751483159660364e-07, + "loss": 0.0142, + "step": 113010 + }, + { + "epoch": 10.272677694964552, + "grad_norm": 0.08554700762033463, + "learning_rate": 2.767239275290623e-07, + "loss": 0.0155, + "step": 113020 + }, + { + "epoch": 10.273586620614434, + "grad_norm": 0.10893561691045761, + "learning_rate": 2.75934141597175e-07, + "loss": 0.0151, + "step": 113030 + }, + { + "epoch": 10.274495546264315, + "grad_norm": 0.12172801047563553, + "learning_rate": 2.751454738609205e-07, + "loss": 0.0164, + "step": 113040 + }, + { + "epoch": 10.275404471914197, + "grad_norm": 0.1076214388012886, + "learning_rate": 2.7435792438018747e-07, + "loss": 0.0136, + "step": 113050 + }, + { + "epoch": 10.27631339756408, + "grad_norm": 0.1324530690908432, + "learning_rate": 2.7357149321479315e-07, + "loss": 0.0148, + "step": 113060 + }, + { + "epoch": 10.27722232321396, + "grad_norm": 0.17789654433727264, + "learning_rate": 2.7278618042445637e-07, + "loss": 0.0182, + "step": 113070 + }, + { + "epoch": 10.278131248863843, + "grad_norm": 0.1671445071697235, + "learning_rate": 2.720019860688178e-07, + "loss": 0.0128, + "step": 113080 + }, + { + "epoch": 10.279040174513725, + "grad_norm": 0.14789758622646332, + "learning_rate": 2.7121891020743305e-07, + "loss": 0.0151, + "step": 113090 + }, + { + "epoch": 10.279949100163607, + "grad_norm": 0.17574456334114075, + "learning_rate": 2.704369528997697e-07, + "loss": 0.0162, + "step": 113100 + }, + { + "epoch": 10.280858025813489, + "grad_norm": 0.09854944795370102, + "learning_rate": 2.696561142052134e-07, + "loss": 0.015, + "step": 113110 + }, + { + "epoch": 10.28176695146337, + "grad_norm": 0.11344318091869354, + "learning_rate": 2.688763941830635e-07, + "loss": 0.0167, + "step": 113120 + }, + { + "epoch": 10.282675877113252, + "grad_norm": 0.17969389259815216, + "learning_rate": 2.6809779289253435e-07, + "loss": 0.0149, + "step": 113130 + }, + { + "epoch": 10.283584802763134, + "grad_norm": 0.1922486275434494, + "learning_rate": 2.6732031039275684e-07, + "loss": 0.018, + "step": 113140 + }, + { + "epoch": 10.284493728413016, + "grad_norm": 0.17404958605766296, + "learning_rate": 2.6654394674277216e-07, + "loss": 0.0178, + "step": 113150 + }, + { + "epoch": 10.285402654062898, + "grad_norm": 0.1344565451145172, + "learning_rate": 2.657687020015431e-07, + "loss": 0.0182, + "step": 113160 + }, + { + "epoch": 10.286311579712779, + "grad_norm": 0.15490224957466125, + "learning_rate": 2.6499457622794596e-07, + "loss": 0.0135, + "step": 113170 + }, + { + "epoch": 10.287220505362662, + "grad_norm": 0.1634392887353897, + "learning_rate": 2.6422156948076536e-07, + "loss": 0.014, + "step": 113180 + }, + { + "epoch": 10.288129431012543, + "grad_norm": 0.16052484512329102, + "learning_rate": 2.6344968181871097e-07, + "loss": 0.016, + "step": 113190 + }, + { + "epoch": 10.289038356662425, + "grad_norm": 0.109339639544487, + "learning_rate": 2.626789133003993e-07, + "loss": 0.0153, + "step": 113200 + }, + { + "epoch": 10.289947282312307, + "grad_norm": 0.4054649770259857, + "learning_rate": 2.619092639843651e-07, + "loss": 0.0153, + "step": 113210 + }, + { + "epoch": 10.29085620796219, + "grad_norm": 0.09604236483573914, + "learning_rate": 2.6114073392906166e-07, + "loss": 0.0157, + "step": 113220 + }, + { + "epoch": 10.29176513361207, + "grad_norm": 0.09759057313203812, + "learning_rate": 2.603733231928507e-07, + "loss": 0.0124, + "step": 113230 + }, + { + "epoch": 10.292674059261952, + "grad_norm": 0.18537752330303192, + "learning_rate": 2.5960703183401215e-07, + "loss": 0.0165, + "step": 113240 + }, + { + "epoch": 10.293582984911835, + "grad_norm": 0.18031734228134155, + "learning_rate": 2.588418599107428e-07, + "loss": 0.0136, + "step": 113250 + }, + { + "epoch": 10.294491910561716, + "grad_norm": 0.1272757351398468, + "learning_rate": 2.5807780748114963e-07, + "loss": 0.0147, + "step": 113260 + }, + { + "epoch": 10.295400836211599, + "grad_norm": 0.10984521359205246, + "learning_rate": 2.573148746032594e-07, + "loss": 0.0159, + "step": 113270 + }, + { + "epoch": 10.29630976186148, + "grad_norm": 0.09989926964044571, + "learning_rate": 2.565530613350109e-07, + "loss": 0.0158, + "step": 113280 + }, + { + "epoch": 10.29721868751136, + "grad_norm": 0.1208203136920929, + "learning_rate": 2.557923677342594e-07, + "loss": 0.0156, + "step": 113290 + }, + { + "epoch": 10.298127613161244, + "grad_norm": 0.34701138734817505, + "learning_rate": 2.550327938587754e-07, + "loss": 0.0133, + "step": 113300 + }, + { + "epoch": 10.299036538811125, + "grad_norm": 0.1097029522061348, + "learning_rate": 2.542743397662395e-07, + "loss": 0.0139, + "step": 113310 + }, + { + "epoch": 10.299945464461008, + "grad_norm": 0.09447923302650452, + "learning_rate": 2.535170055142555e-07, + "loss": 0.0147, + "step": 113320 + }, + { + "epoch": 10.300854390110889, + "grad_norm": 0.13431289792060852, + "learning_rate": 2.5276079116033747e-07, + "loss": 0.0177, + "step": 113330 + }, + { + "epoch": 10.301763315760772, + "grad_norm": 0.10926789045333862, + "learning_rate": 2.5200569676191444e-07, + "loss": 0.0145, + "step": 113340 + }, + { + "epoch": 10.302672241410653, + "grad_norm": 0.33949464559555054, + "learning_rate": 2.512517223763272e-07, + "loss": 0.0167, + "step": 113350 + }, + { + "epoch": 10.303581167060534, + "grad_norm": 0.10265246033668518, + "learning_rate": 2.5049886806084163e-07, + "loss": 0.0147, + "step": 113360 + }, + { + "epoch": 10.304490092710417, + "grad_norm": 0.11953078955411911, + "learning_rate": 2.497471338726287e-07, + "loss": 0.0148, + "step": 113370 + }, + { + "epoch": 10.305399018360298, + "grad_norm": 0.15425068140029907, + "learning_rate": 2.48996519868776e-07, + "loss": 0.0174, + "step": 113380 + }, + { + "epoch": 10.30630794401018, + "grad_norm": 0.12710587680339813, + "learning_rate": 2.482470261062897e-07, + "loss": 0.017, + "step": 113390 + }, + { + "epoch": 10.307216869660062, + "grad_norm": 0.1056370809674263, + "learning_rate": 2.4749865264209083e-07, + "loss": 0.0146, + "step": 113400 + }, + { + "epoch": 10.308125795309945, + "grad_norm": 0.07690790295600891, + "learning_rate": 2.467513995330106e-07, + "loss": 0.0169, + "step": 113410 + }, + { + "epoch": 10.309034720959826, + "grad_norm": 0.10131268948316574, + "learning_rate": 2.4600526683579705e-07, + "loss": 0.015, + "step": 113420 + }, + { + "epoch": 10.309943646609707, + "grad_norm": 0.13260875642299652, + "learning_rate": 2.4526025460711476e-07, + "loss": 0.0161, + "step": 113430 + }, + { + "epoch": 10.31085257225959, + "grad_norm": 0.1503085494041443, + "learning_rate": 2.4451636290354683e-07, + "loss": 0.0175, + "step": 113440 + }, + { + "epoch": 10.31176149790947, + "grad_norm": 0.13587023317813873, + "learning_rate": 2.4377359178157977e-07, + "loss": 0.0147, + "step": 113450 + }, + { + "epoch": 10.312670423559354, + "grad_norm": 0.09987065196037292, + "learning_rate": 2.4303194129762676e-07, + "loss": 0.0133, + "step": 113460 + }, + { + "epoch": 10.313579349209235, + "grad_norm": 0.20044320821762085, + "learning_rate": 2.422914115080127e-07, + "loss": 0.013, + "step": 113470 + }, + { + "epoch": 10.314488274859116, + "grad_norm": 0.27238157391548157, + "learning_rate": 2.41552002468971e-07, + "loss": 0.0152, + "step": 113480 + }, + { + "epoch": 10.315397200508999, + "grad_norm": 0.12486064434051514, + "learning_rate": 2.4081371423665843e-07, + "loss": 0.017, + "step": 113490 + }, + { + "epoch": 10.31630612615888, + "grad_norm": 0.0902237519621849, + "learning_rate": 2.4007654686714175e-07, + "loss": 0.0163, + "step": 113500 + }, + { + "epoch": 10.317215051808763, + "grad_norm": 0.1470591425895691, + "learning_rate": 2.3934050041640456e-07, + "loss": 0.0168, + "step": 113510 + }, + { + "epoch": 10.318123977458644, + "grad_norm": 0.13617801666259766, + "learning_rate": 2.386055749403454e-07, + "loss": 0.017, + "step": 113520 + }, + { + "epoch": 10.319032903108525, + "grad_norm": 0.12078254669904709, + "learning_rate": 2.3787177049477472e-07, + "loss": 0.015, + "step": 113530 + }, + { + "epoch": 10.319941828758408, + "grad_norm": 0.12485625594854355, + "learning_rate": 2.371390871354212e-07, + "loss": 0.0159, + "step": 113540 + }, + { + "epoch": 10.320850754408289, + "grad_norm": 0.15081200003623962, + "learning_rate": 2.364075249179287e-07, + "loss": 0.0171, + "step": 113550 + }, + { + "epoch": 10.321759680058172, + "grad_norm": 0.14242887496948242, + "learning_rate": 2.3567708389785282e-07, + "loss": 0.014, + "step": 113560 + }, + { + "epoch": 10.322668605708053, + "grad_norm": 0.1469068080186844, + "learning_rate": 2.3494776413066744e-07, + "loss": 0.0156, + "step": 113570 + }, + { + "epoch": 10.323577531357936, + "grad_norm": 0.11380726844072342, + "learning_rate": 2.3421956567175663e-07, + "loss": 0.0134, + "step": 113580 + }, + { + "epoch": 10.324486457007817, + "grad_norm": 0.11045274138450623, + "learning_rate": 2.3349248857642612e-07, + "loss": 0.0155, + "step": 113590 + }, + { + "epoch": 10.325395382657698, + "grad_norm": 0.13309629261493683, + "learning_rate": 2.3276653289989014e-07, + "loss": 0.0162, + "step": 113600 + }, + { + "epoch": 10.32630430830758, + "grad_norm": 0.17609445750713348, + "learning_rate": 2.3204169869727953e-07, + "loss": 0.019, + "step": 113610 + }, + { + "epoch": 10.327213233957462, + "grad_norm": 0.1561780571937561, + "learning_rate": 2.3131798602364197e-07, + "loss": 0.0142, + "step": 113620 + }, + { + "epoch": 10.328122159607345, + "grad_norm": 0.14250533282756805, + "learning_rate": 2.305953949339401e-07, + "loss": 0.0178, + "step": 113630 + }, + { + "epoch": 10.329031085257226, + "grad_norm": 0.12184563279151917, + "learning_rate": 2.2987392548304841e-07, + "loss": 0.0153, + "step": 113640 + }, + { + "epoch": 10.329940010907109, + "grad_norm": 0.1382737159729004, + "learning_rate": 2.2915357772575473e-07, + "loss": 0.0166, + "step": 113650 + }, + { + "epoch": 10.33084893655699, + "grad_norm": 0.10328449308872223, + "learning_rate": 2.2843435171677029e-07, + "loss": 0.0146, + "step": 113660 + }, + { + "epoch": 10.33175786220687, + "grad_norm": 0.1734289675951004, + "learning_rate": 2.2771624751071308e-07, + "loss": 0.0148, + "step": 113670 + }, + { + "epoch": 10.332666787856754, + "grad_norm": 0.20845825970172882, + "learning_rate": 2.2699926516211778e-07, + "loss": 0.0186, + "step": 113680 + }, + { + "epoch": 10.333575713506635, + "grad_norm": 0.10756650567054749, + "learning_rate": 2.262834047254342e-07, + "loss": 0.015, + "step": 113690 + }, + { + "epoch": 10.334484639156518, + "grad_norm": 0.09609803557395935, + "learning_rate": 2.2556866625502881e-07, + "loss": 0.0153, + "step": 113700 + }, + { + "epoch": 10.335393564806399, + "grad_norm": 0.11918995529413223, + "learning_rate": 2.2485504980518156e-07, + "loss": 0.0143, + "step": 113710 + }, + { + "epoch": 10.33630249045628, + "grad_norm": 0.15405312180519104, + "learning_rate": 2.241425554300841e-07, + "loss": 0.0153, + "step": 113720 + }, + { + "epoch": 10.337211416106163, + "grad_norm": 0.13111542165279388, + "learning_rate": 2.2343118318384648e-07, + "loss": 0.0144, + "step": 113730 + }, + { + "epoch": 10.338120341756044, + "grad_norm": 0.1262841820716858, + "learning_rate": 2.2272093312049546e-07, + "loss": 0.0156, + "step": 113740 + }, + { + "epoch": 10.339029267405927, + "grad_norm": 0.31191498041152954, + "learning_rate": 2.2201180529396792e-07, + "loss": 0.0125, + "step": 113750 + }, + { + "epoch": 10.339938193055808, + "grad_norm": 0.12115083634853363, + "learning_rate": 2.2130379975811576e-07, + "loss": 0.0155, + "step": 113760 + }, + { + "epoch": 10.34084711870569, + "grad_norm": 0.12097570300102234, + "learning_rate": 2.20596916566711e-07, + "loss": 0.0163, + "step": 113770 + }, + { + "epoch": 10.341756044355572, + "grad_norm": 0.1316337138414383, + "learning_rate": 2.1989115577343566e-07, + "loss": 0.0169, + "step": 113780 + }, + { + "epoch": 10.342664970005453, + "grad_norm": 0.17059341073036194, + "learning_rate": 2.1918651743188522e-07, + "loss": 0.0153, + "step": 113790 + }, + { + "epoch": 10.343573895655336, + "grad_norm": 0.135531947016716, + "learning_rate": 2.1848300159557355e-07, + "loss": 0.0158, + "step": 113800 + }, + { + "epoch": 10.344482821305217, + "grad_norm": 0.19643577933311462, + "learning_rate": 2.1778060831792957e-07, + "loss": 0.0141, + "step": 113810 + }, + { + "epoch": 10.3453917469551, + "grad_norm": 0.17848902940750122, + "learning_rate": 2.1707933765229392e-07, + "loss": 0.0176, + "step": 113820 + }, + { + "epoch": 10.34630067260498, + "grad_norm": 0.1021571159362793, + "learning_rate": 2.1637918965192405e-07, + "loss": 0.0143, + "step": 113830 + }, + { + "epoch": 10.347209598254862, + "grad_norm": 0.21090616285800934, + "learning_rate": 2.1568016436998904e-07, + "loss": 0.0151, + "step": 113840 + }, + { + "epoch": 10.348118523904745, + "grad_norm": 0.10060630738735199, + "learning_rate": 2.1498226185957981e-07, + "loss": 0.0143, + "step": 113850 + }, + { + "epoch": 10.349027449554626, + "grad_norm": 0.18158085644245148, + "learning_rate": 2.1428548217369558e-07, + "loss": 0.014, + "step": 113860 + }, + { + "epoch": 10.349936375204509, + "grad_norm": 0.3616820275783539, + "learning_rate": 2.1358982536524907e-07, + "loss": 0.0148, + "step": 113870 + }, + { + "epoch": 10.35084530085439, + "grad_norm": 0.17830069363117218, + "learning_rate": 2.1289529148707632e-07, + "loss": 0.0157, + "step": 113880 + }, + { + "epoch": 10.351754226504273, + "grad_norm": 0.1581626832485199, + "learning_rate": 2.1220188059191848e-07, + "loss": 0.0145, + "step": 113890 + }, + { + "epoch": 10.352663152154154, + "grad_norm": 0.09079354256391525, + "learning_rate": 2.1150959273243674e-07, + "loss": 0.016, + "step": 113900 + }, + { + "epoch": 10.353572077804035, + "grad_norm": 0.11152864992618561, + "learning_rate": 2.1081842796120575e-07, + "loss": 0.0168, + "step": 113910 + }, + { + "epoch": 10.354481003453918, + "grad_norm": 0.12405280768871307, + "learning_rate": 2.1012838633071518e-07, + "loss": 0.016, + "step": 113920 + }, + { + "epoch": 10.355389929103799, + "grad_norm": 0.1575600802898407, + "learning_rate": 2.0943946789336975e-07, + "loss": 0.016, + "step": 113930 + }, + { + "epoch": 10.356298854753682, + "grad_norm": 0.14011356234550476, + "learning_rate": 2.087516727014893e-07, + "loss": 0.0163, + "step": 113940 + }, + { + "epoch": 10.357207780403563, + "grad_norm": 0.11777205765247345, + "learning_rate": 2.080650008073004e-07, + "loss": 0.0167, + "step": 113950 + }, + { + "epoch": 10.358116706053444, + "grad_norm": 0.1792873591184616, + "learning_rate": 2.073794522629613e-07, + "loss": 0.0145, + "step": 113960 + }, + { + "epoch": 10.359025631703327, + "grad_norm": 0.26915860176086426, + "learning_rate": 2.066950271205287e-07, + "loss": 0.0167, + "step": 113970 + }, + { + "epoch": 10.359934557353208, + "grad_norm": 0.11314482986927032, + "learning_rate": 2.0601172543197933e-07, + "loss": 0.0178, + "step": 113980 + }, + { + "epoch": 10.36084348300309, + "grad_norm": 0.10359898209571838, + "learning_rate": 2.053295472492084e-07, + "loss": 0.0152, + "step": 113990 + }, + { + "epoch": 10.361752408652972, + "grad_norm": 0.16212515532970428, + "learning_rate": 2.046484926240244e-07, + "loss": 0.0149, + "step": 114000 + }, + { + "epoch": 10.362661334302855, + "grad_norm": 0.12192656099796295, + "learning_rate": 2.0396856160814603e-07, + "loss": 0.0148, + "step": 114010 + }, + { + "epoch": 10.363570259952736, + "grad_norm": 0.17725327610969543, + "learning_rate": 2.0328975425320694e-07, + "loss": 0.0158, + "step": 114020 + }, + { + "epoch": 10.364479185602617, + "grad_norm": 0.21590383350849152, + "learning_rate": 2.0261207061076425e-07, + "loss": 0.0155, + "step": 114030 + }, + { + "epoch": 10.3653881112525, + "grad_norm": 0.11927013099193573, + "learning_rate": 2.019355107322801e-07, + "loss": 0.0162, + "step": 114040 + }, + { + "epoch": 10.366297036902381, + "grad_norm": 0.0922645851969719, + "learning_rate": 2.0126007466913344e-07, + "loss": 0.014, + "step": 114050 + }, + { + "epoch": 10.367205962552264, + "grad_norm": 0.11998223513364792, + "learning_rate": 2.005857624726215e-07, + "loss": 0.0172, + "step": 114060 + }, + { + "epoch": 10.368114888202145, + "grad_norm": 0.133022278547287, + "learning_rate": 1.9991257419395502e-07, + "loss": 0.0162, + "step": 114070 + }, + { + "epoch": 10.369023813852026, + "grad_norm": 0.1393030434846878, + "learning_rate": 1.992405098842548e-07, + "loss": 0.0168, + "step": 114080 + }, + { + "epoch": 10.369932739501909, + "grad_norm": 0.13888117671012878, + "learning_rate": 1.9856956959456163e-07, + "loss": 0.0152, + "step": 114090 + }, + { + "epoch": 10.37084166515179, + "grad_norm": 0.1286603808403015, + "learning_rate": 1.9789975337582812e-07, + "loss": 0.0158, + "step": 114100 + }, + { + "epoch": 10.371750590801673, + "grad_norm": 0.2128531038761139, + "learning_rate": 1.972310612789252e-07, + "loss": 0.0173, + "step": 114110 + }, + { + "epoch": 10.372659516451554, + "grad_norm": 0.14821292459964752, + "learning_rate": 1.965634933546323e-07, + "loss": 0.0154, + "step": 114120 + }, + { + "epoch": 10.373568442101437, + "grad_norm": 0.13708114624023438, + "learning_rate": 1.958970496536472e-07, + "loss": 0.0157, + "step": 114130 + }, + { + "epoch": 10.374477367751318, + "grad_norm": 0.08564601838588715, + "learning_rate": 1.952317302265827e-07, + "loss": 0.0153, + "step": 114140 + }, + { + "epoch": 10.375386293401199, + "grad_norm": 0.12756483256816864, + "learning_rate": 1.9456753512396508e-07, + "loss": 0.0138, + "step": 114150 + }, + { + "epoch": 10.376295219051082, + "grad_norm": 0.09548357874155045, + "learning_rate": 1.9390446439623733e-07, + "loss": 0.0152, + "step": 114160 + }, + { + "epoch": 10.377204144700963, + "grad_norm": 0.11810117959976196, + "learning_rate": 1.932425180937525e-07, + "loss": 0.0158, + "step": 114170 + }, + { + "epoch": 10.378113070350846, + "grad_norm": 0.2089371681213379, + "learning_rate": 1.92581696266782e-07, + "loss": 0.0129, + "step": 114180 + }, + { + "epoch": 10.379021996000727, + "grad_norm": 0.15051235258579254, + "learning_rate": 1.9192199896551243e-07, + "loss": 0.0153, + "step": 114190 + }, + { + "epoch": 10.379930921650608, + "grad_norm": 0.13004393875598907, + "learning_rate": 1.91263426240042e-07, + "loss": 0.017, + "step": 114200 + }, + { + "epoch": 10.380839847300491, + "grad_norm": 0.15114866197109222, + "learning_rate": 1.9060597814038404e-07, + "loss": 0.0147, + "step": 114210 + }, + { + "epoch": 10.381748772950372, + "grad_norm": 0.1186424195766449, + "learning_rate": 1.8994965471646863e-07, + "loss": 0.0163, + "step": 114220 + }, + { + "epoch": 10.382657698600255, + "grad_norm": 0.1048760861158371, + "learning_rate": 1.892944560181409e-07, + "loss": 0.0151, + "step": 114230 + }, + { + "epoch": 10.383566624250136, + "grad_norm": 0.1142694428563118, + "learning_rate": 1.8864038209515598e-07, + "loss": 0.0148, + "step": 114240 + }, + { + "epoch": 10.384475549900019, + "grad_norm": 0.1416153460741043, + "learning_rate": 1.8798743299718757e-07, + "loss": 0.0149, + "step": 114250 + }, + { + "epoch": 10.3853844755499, + "grad_norm": 0.150302916765213, + "learning_rate": 1.873356087738226e-07, + "loss": 0.0134, + "step": 114260 + }, + { + "epoch": 10.386293401199781, + "grad_norm": 0.1376061588525772, + "learning_rate": 1.8668490947456317e-07, + "loss": 0.0147, + "step": 114270 + }, + { + "epoch": 10.387202326849664, + "grad_norm": 0.22180555760860443, + "learning_rate": 1.8603533514882642e-07, + "loss": 0.0169, + "step": 114280 + }, + { + "epoch": 10.388111252499545, + "grad_norm": 0.2364526093006134, + "learning_rate": 1.8538688584594123e-07, + "loss": 0.0146, + "step": 114290 + }, + { + "epoch": 10.389020178149428, + "grad_norm": 0.16479724645614624, + "learning_rate": 1.8473956161515481e-07, + "loss": 0.0138, + "step": 114300 + }, + { + "epoch": 10.38992910379931, + "grad_norm": 0.17166376113891602, + "learning_rate": 1.8409336250562626e-07, + "loss": 0.0167, + "step": 114310 + }, + { + "epoch": 10.39083802944919, + "grad_norm": 0.09566178172826767, + "learning_rate": 1.834482885664296e-07, + "loss": 0.0145, + "step": 114320 + }, + { + "epoch": 10.391746955099073, + "grad_norm": 0.11355727910995483, + "learning_rate": 1.82804339846554e-07, + "loss": 0.0166, + "step": 114330 + }, + { + "epoch": 10.392655880748954, + "grad_norm": 0.14151526987552643, + "learning_rate": 1.8216151639490696e-07, + "loss": 0.0151, + "step": 114340 + }, + { + "epoch": 10.393564806398837, + "grad_norm": 0.12875929474830627, + "learning_rate": 1.8151981826030117e-07, + "loss": 0.0146, + "step": 114350 + }, + { + "epoch": 10.394473732048718, + "grad_norm": 0.2098059207201004, + "learning_rate": 1.8087924549147094e-07, + "loss": 0.0143, + "step": 114360 + }, + { + "epoch": 10.395382657698601, + "grad_norm": 0.12076090276241302, + "learning_rate": 1.802397981370657e-07, + "loss": 0.0139, + "step": 114370 + }, + { + "epoch": 10.396291583348482, + "grad_norm": 0.0995701476931572, + "learning_rate": 1.796014762456466e-07, + "loss": 0.0146, + "step": 114380 + }, + { + "epoch": 10.397200508998363, + "grad_norm": 0.13375812768936157, + "learning_rate": 1.789642798656882e-07, + "loss": 0.0164, + "step": 114390 + }, + { + "epoch": 10.398109434648246, + "grad_norm": 0.17989672720432281, + "learning_rate": 1.7832820904558178e-07, + "loss": 0.0154, + "step": 114400 + }, + { + "epoch": 10.399018360298127, + "grad_norm": 0.1272241622209549, + "learning_rate": 1.7769326383363538e-07, + "loss": 0.0174, + "step": 114410 + }, + { + "epoch": 10.39992728594801, + "grad_norm": 0.11659332364797592, + "learning_rate": 1.7705944427806543e-07, + "loss": 0.0153, + "step": 114420 + }, + { + "epoch": 10.400836211597891, + "grad_norm": 0.1367245614528656, + "learning_rate": 1.764267504270084e-07, + "loss": 0.0144, + "step": 114430 + }, + { + "epoch": 10.401745137247772, + "grad_norm": 0.15860189497470856, + "learning_rate": 1.7579518232851256e-07, + "loss": 0.0175, + "step": 114440 + }, + { + "epoch": 10.402654062897655, + "grad_norm": 0.1507052332162857, + "learning_rate": 1.7516474003054118e-07, + "loss": 0.0183, + "step": 114450 + }, + { + "epoch": 10.403562988547536, + "grad_norm": 0.17245164513587952, + "learning_rate": 1.7453542358097097e-07, + "loss": 0.015, + "step": 114460 + }, + { + "epoch": 10.40447191419742, + "grad_norm": 0.16734114289283752, + "learning_rate": 1.7390723302759704e-07, + "loss": 0.016, + "step": 114470 + }, + { + "epoch": 10.4053808398473, + "grad_norm": 0.12479735910892487, + "learning_rate": 1.7328016841812622e-07, + "loss": 0.0175, + "step": 114480 + }, + { + "epoch": 10.406289765497183, + "grad_norm": 0.09744474291801453, + "learning_rate": 1.7265422980017709e-07, + "loss": 0.0149, + "step": 114490 + }, + { + "epoch": 10.407198691147064, + "grad_norm": 0.1513664275407791, + "learning_rate": 1.720294172212883e-07, + "loss": 0.0159, + "step": 114500 + }, + { + "epoch": 10.408107616796945, + "grad_norm": 0.09689345955848694, + "learning_rate": 1.7140573072890853e-07, + "loss": 0.0141, + "step": 114510 + }, + { + "epoch": 10.409016542446828, + "grad_norm": 0.36329397559165955, + "learning_rate": 1.7078317037040326e-07, + "loss": 0.017, + "step": 114520 + }, + { + "epoch": 10.40992546809671, + "grad_norm": 0.12569990754127502, + "learning_rate": 1.7016173619305298e-07, + "loss": 0.0158, + "step": 114530 + }, + { + "epoch": 10.410834393746592, + "grad_norm": 0.10888071358203888, + "learning_rate": 1.6954142824404995e-07, + "loss": 0.0151, + "step": 114540 + }, + { + "epoch": 10.411743319396473, + "grad_norm": 0.16615810990333557, + "learning_rate": 1.689222465705015e-07, + "loss": 0.0146, + "step": 114550 + }, + { + "epoch": 10.412652245046356, + "grad_norm": 0.15643242001533508, + "learning_rate": 1.683041912194333e-07, + "loss": 0.0177, + "step": 114560 + }, + { + "epoch": 10.413561170696237, + "grad_norm": 0.13787256181240082, + "learning_rate": 1.676872622377812e-07, + "loss": 0.0154, + "step": 114570 + }, + { + "epoch": 10.414470096346118, + "grad_norm": 0.15943218767642975, + "learning_rate": 1.67071459672396e-07, + "loss": 0.0157, + "step": 114580 + }, + { + "epoch": 10.415379021996001, + "grad_norm": 0.11149334907531738, + "learning_rate": 1.6645678357004367e-07, + "loss": 0.015, + "step": 114590 + }, + { + "epoch": 10.416287947645882, + "grad_norm": 0.1625005602836609, + "learning_rate": 1.6584323397740853e-07, + "loss": 0.0163, + "step": 114600 + }, + { + "epoch": 10.417196873295765, + "grad_norm": 0.20830769836902618, + "learning_rate": 1.6523081094108161e-07, + "loss": 0.0157, + "step": 114610 + }, + { + "epoch": 10.418105798945646, + "grad_norm": 0.09301119297742844, + "learning_rate": 1.646195145075724e-07, + "loss": 0.0136, + "step": 114620 + }, + { + "epoch": 10.419014724595527, + "grad_norm": 0.16995027661323547, + "learning_rate": 1.6400934472330708e-07, + "loss": 0.0164, + "step": 114630 + }, + { + "epoch": 10.41992365024541, + "grad_norm": 0.20355349779129028, + "learning_rate": 1.634003016346236e-07, + "loss": 0.0174, + "step": 114640 + }, + { + "epoch": 10.420832575895291, + "grad_norm": 0.14259430766105652, + "learning_rate": 1.627923852877733e-07, + "loss": 0.0144, + "step": 114650 + }, + { + "epoch": 10.421741501545174, + "grad_norm": 0.12968365848064423, + "learning_rate": 1.6218559572892422e-07, + "loss": 0.0166, + "step": 114660 + }, + { + "epoch": 10.422650427195055, + "grad_norm": 0.09160219877958298, + "learning_rate": 1.6157993300415786e-07, + "loss": 0.0154, + "step": 114670 + }, + { + "epoch": 10.423559352844936, + "grad_norm": 0.12693138420581818, + "learning_rate": 1.609753971594724e-07, + "loss": 0.0154, + "step": 114680 + }, + { + "epoch": 10.42446827849482, + "grad_norm": 0.16666744649410248, + "learning_rate": 1.6037198824077447e-07, + "loss": 0.0163, + "step": 114690 + }, + { + "epoch": 10.4253772041447, + "grad_norm": 0.2860628068447113, + "learning_rate": 1.5976970629389076e-07, + "loss": 0.0164, + "step": 114700 + }, + { + "epoch": 10.426286129794583, + "grad_norm": 0.11998523026704788, + "learning_rate": 1.5916855136456298e-07, + "loss": 0.0158, + "step": 114710 + }, + { + "epoch": 10.427195055444464, + "grad_norm": 0.11910942941904068, + "learning_rate": 1.5856852349844131e-07, + "loss": 0.0137, + "step": 114720 + }, + { + "epoch": 10.428103981094347, + "grad_norm": 0.22264865040779114, + "learning_rate": 1.5796962274109595e-07, + "loss": 0.0161, + "step": 114730 + }, + { + "epoch": 10.429012906744228, + "grad_norm": 0.09525347501039505, + "learning_rate": 1.5737184913800718e-07, + "loss": 0.0148, + "step": 114740 + }, + { + "epoch": 10.42992183239411, + "grad_norm": 0.14111442863941193, + "learning_rate": 1.5677520273457534e-07, + "loss": 0.0181, + "step": 114750 + }, + { + "epoch": 10.430830758043992, + "grad_norm": 0.1327064484357834, + "learning_rate": 1.5617968357610922e-07, + "loss": 0.017, + "step": 114760 + }, + { + "epoch": 10.431739683693873, + "grad_norm": 0.16051997244358063, + "learning_rate": 1.5558529170783597e-07, + "loss": 0.0165, + "step": 114770 + }, + { + "epoch": 10.432648609343756, + "grad_norm": 0.12215515971183777, + "learning_rate": 1.5499202717489614e-07, + "loss": 0.0161, + "step": 114780 + }, + { + "epoch": 10.433557534993637, + "grad_norm": 0.14940603077411652, + "learning_rate": 1.54399890022342e-07, + "loss": 0.0152, + "step": 114790 + }, + { + "epoch": 10.43446646064352, + "grad_norm": 0.10071360319852829, + "learning_rate": 1.5380888029514428e-07, + "loss": 0.015, + "step": 114800 + }, + { + "epoch": 10.435375386293401, + "grad_norm": 0.19960148632526398, + "learning_rate": 1.532189980381854e-07, + "loss": 0.0166, + "step": 114810 + }, + { + "epoch": 10.436284311943282, + "grad_norm": 0.11250408738851547, + "learning_rate": 1.526302432962645e-07, + "loss": 0.0143, + "step": 114820 + }, + { + "epoch": 10.437193237593165, + "grad_norm": 0.1600852906703949, + "learning_rate": 1.5204261611409088e-07, + "loss": 0.0168, + "step": 114830 + }, + { + "epoch": 10.438102163243046, + "grad_norm": 0.13271978497505188, + "learning_rate": 1.5145611653629378e-07, + "loss": 0.0165, + "step": 114840 + }, + { + "epoch": 10.43901108889293, + "grad_norm": 0.10504041612148285, + "learning_rate": 1.5087074460741257e-07, + "loss": 0.0149, + "step": 114850 + }, + { + "epoch": 10.43992001454281, + "grad_norm": 0.13329315185546875, + "learning_rate": 1.5028650037190171e-07, + "loss": 0.0159, + "step": 114860 + }, + { + "epoch": 10.440828940192691, + "grad_norm": 0.11520635336637497, + "learning_rate": 1.497033838741324e-07, + "loss": 0.0166, + "step": 114870 + }, + { + "epoch": 10.441737865842574, + "grad_norm": 0.22434186935424805, + "learning_rate": 1.4912139515838752e-07, + "loss": 0.0167, + "step": 114880 + }, + { + "epoch": 10.442646791492455, + "grad_norm": 0.17436109483242035, + "learning_rate": 1.4854053426886505e-07, + "loss": 0.0143, + "step": 114890 + }, + { + "epoch": 10.443555717142338, + "grad_norm": 0.10580995678901672, + "learning_rate": 1.4796080124967803e-07, + "loss": 0.0162, + "step": 114900 + }, + { + "epoch": 10.44446464279222, + "grad_norm": 0.1531774252653122, + "learning_rate": 1.473821961448546e-07, + "loss": 0.015, + "step": 114910 + }, + { + "epoch": 10.4453735684421, + "grad_norm": 0.17876358330249786, + "learning_rate": 1.4680471899833292e-07, + "loss": 0.0136, + "step": 114920 + }, + { + "epoch": 10.446282494091983, + "grad_norm": 0.13139499723911285, + "learning_rate": 1.4622836985396958e-07, + "loss": 0.0149, + "step": 114930 + }, + { + "epoch": 10.447191419741864, + "grad_norm": 0.13558685779571533, + "learning_rate": 1.4565314875553626e-07, + "loss": 0.0165, + "step": 114940 + }, + { + "epoch": 10.448100345391747, + "grad_norm": 0.26996123790740967, + "learning_rate": 1.4507905574671797e-07, + "loss": 0.0161, + "step": 114950 + }, + { + "epoch": 10.449009271041628, + "grad_norm": 0.1308196634054184, + "learning_rate": 1.445060908711082e-07, + "loss": 0.014, + "step": 114960 + }, + { + "epoch": 10.449918196691511, + "grad_norm": 0.23821324110031128, + "learning_rate": 1.439342541722255e-07, + "loss": 0.0156, + "step": 114970 + }, + { + "epoch": 10.450827122341392, + "grad_norm": 0.11854006350040436, + "learning_rate": 1.433635456934934e-07, + "loss": 0.014, + "step": 114980 + }, + { + "epoch": 10.451736047991274, + "grad_norm": 0.22151508927345276, + "learning_rate": 1.4279396547825564e-07, + "loss": 0.0172, + "step": 114990 + }, + { + "epoch": 10.452644973641156, + "grad_norm": 0.18446847796440125, + "learning_rate": 1.4222551356976587e-07, + "loss": 0.0151, + "step": 115000 + }, + { + "epoch": 10.453553899291038, + "grad_norm": 0.10813744366168976, + "learning_rate": 1.4165819001119628e-07, + "loss": 0.0139, + "step": 115010 + }, + { + "epoch": 10.45446282494092, + "grad_norm": 0.18539683520793915, + "learning_rate": 1.4109199484563075e-07, + "loss": 0.0154, + "step": 115020 + }, + { + "epoch": 10.455371750590801, + "grad_norm": 0.15080676972866058, + "learning_rate": 1.4052692811606816e-07, + "loss": 0.0152, + "step": 115030 + }, + { + "epoch": 10.456280676240684, + "grad_norm": 0.12360849231481552, + "learning_rate": 1.3996298986542087e-07, + "loss": 0.0143, + "step": 115040 + }, + { + "epoch": 10.457189601890565, + "grad_norm": 0.09323947876691818, + "learning_rate": 1.39400180136518e-07, + "loss": 0.0162, + "step": 115050 + }, + { + "epoch": 10.458098527540447, + "grad_norm": 0.13798823952674866, + "learning_rate": 1.3883849897209865e-07, + "loss": 0.0145, + "step": 115060 + }, + { + "epoch": 10.45900745319033, + "grad_norm": 0.13039414584636688, + "learning_rate": 1.382779464148204e-07, + "loss": 0.0167, + "step": 115070 + }, + { + "epoch": 10.45991637884021, + "grad_norm": 0.1501094251871109, + "learning_rate": 1.3771852250725413e-07, + "loss": 0.0165, + "step": 115080 + }, + { + "epoch": 10.460825304490093, + "grad_norm": 0.12889614701271057, + "learning_rate": 1.371602272918826e-07, + "loss": 0.0164, + "step": 115090 + }, + { + "epoch": 10.461734230139974, + "grad_norm": 0.117621049284935, + "learning_rate": 1.3660306081110517e-07, + "loss": 0.018, + "step": 115100 + }, + { + "epoch": 10.462643155789856, + "grad_norm": 0.13026390969753265, + "learning_rate": 1.360470231072364e-07, + "loss": 0.0143, + "step": 115110 + }, + { + "epoch": 10.463552081439738, + "grad_norm": 0.1170746311545372, + "learning_rate": 1.354921142225024e-07, + "loss": 0.0142, + "step": 115120 + }, + { + "epoch": 10.46446100708962, + "grad_norm": 0.0914081260561943, + "learning_rate": 1.3493833419904454e-07, + "loss": 0.0149, + "step": 115130 + }, + { + "epoch": 10.465369932739502, + "grad_norm": 0.2234908789396286, + "learning_rate": 1.3438568307892085e-07, + "loss": 0.0164, + "step": 115140 + }, + { + "epoch": 10.466278858389384, + "grad_norm": 0.09521325677633286, + "learning_rate": 1.338341609040994e-07, + "loss": 0.0154, + "step": 115150 + }, + { + "epoch": 10.467187784039266, + "grad_norm": 0.29966065287590027, + "learning_rate": 1.3328376771646334e-07, + "loss": 0.0143, + "step": 115160 + }, + { + "epoch": 10.468096709689148, + "grad_norm": 0.21447107195854187, + "learning_rate": 1.327345035578159e-07, + "loss": 0.0145, + "step": 115170 + }, + { + "epoch": 10.469005635339029, + "grad_norm": 0.09567426890134811, + "learning_rate": 1.3218636846986543e-07, + "loss": 0.0162, + "step": 115180 + }, + { + "epoch": 10.469914560988911, + "grad_norm": 0.1277104914188385, + "learning_rate": 1.316393624942436e-07, + "loss": 0.0165, + "step": 115190 + }, + { + "epoch": 10.470823486638793, + "grad_norm": 0.12626196444034576, + "learning_rate": 1.3109348567248712e-07, + "loss": 0.0157, + "step": 115200 + }, + { + "epoch": 10.471732412288675, + "grad_norm": 0.1715201735496521, + "learning_rate": 1.3054873804605627e-07, + "loss": 0.0126, + "step": 115210 + }, + { + "epoch": 10.472641337938557, + "grad_norm": 0.15489286184310913, + "learning_rate": 1.300051196563179e-07, + "loss": 0.0161, + "step": 115220 + }, + { + "epoch": 10.473550263588438, + "grad_norm": 0.13081902265548706, + "learning_rate": 1.294626305445573e-07, + "loss": 0.0145, + "step": 115230 + }, + { + "epoch": 10.47445918923832, + "grad_norm": 0.46765318512916565, + "learning_rate": 1.2892127075197324e-07, + "loss": 0.017, + "step": 115240 + }, + { + "epoch": 10.475368114888202, + "grad_norm": 0.1934202015399933, + "learning_rate": 1.2838104031967778e-07, + "loss": 0.0164, + "step": 115250 + }, + { + "epoch": 10.476277040538085, + "grad_norm": 0.13029925525188446, + "learning_rate": 1.2784193928869647e-07, + "loss": 0.015, + "step": 115260 + }, + { + "epoch": 10.477185966187966, + "grad_norm": 0.12757116556167603, + "learning_rate": 1.2730396769997487e-07, + "loss": 0.0166, + "step": 115270 + }, + { + "epoch": 10.478094891837848, + "grad_norm": 0.11862832307815552, + "learning_rate": 1.267671255943653e-07, + "loss": 0.0164, + "step": 115280 + }, + { + "epoch": 10.47900381748773, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.2623141301263686e-07, + "loss": 0.0171, + "step": 115290 + }, + { + "epoch": 10.47991274313761, + "grad_norm": 0.1002020314335823, + "learning_rate": 1.256968299954736e-07, + "loss": 0.0132, + "step": 115300 + }, + { + "epoch": 10.480821668787494, + "grad_norm": 0.184060201048851, + "learning_rate": 1.2516337658347643e-07, + "loss": 0.016, + "step": 115310 + }, + { + "epoch": 10.481730594437375, + "grad_norm": 0.1497005671262741, + "learning_rate": 1.2463105281715458e-07, + "loss": 0.0147, + "step": 115320 + }, + { + "epoch": 10.482639520087258, + "grad_norm": 0.10722093284130096, + "learning_rate": 1.240998587369341e-07, + "loss": 0.0161, + "step": 115330 + }, + { + "epoch": 10.483548445737139, + "grad_norm": 0.1505749374628067, + "learning_rate": 1.2356979438315764e-07, + "loss": 0.0154, + "step": 115340 + }, + { + "epoch": 10.48445737138702, + "grad_norm": 0.1621452420949936, + "learning_rate": 1.230408597960797e-07, + "loss": 0.0155, + "step": 115350 + }, + { + "epoch": 10.485366297036903, + "grad_norm": 0.136884406208992, + "learning_rate": 1.2251305501586818e-07, + "loss": 0.016, + "step": 115360 + }, + { + "epoch": 10.486275222686784, + "grad_norm": 0.13945822417736053, + "learning_rate": 1.2198638008260765e-07, + "loss": 0.0156, + "step": 115370 + }, + { + "epoch": 10.487184148336667, + "grad_norm": 0.13799221813678741, + "learning_rate": 1.2146083503629612e-07, + "loss": 0.0162, + "step": 115380 + }, + { + "epoch": 10.488093073986548, + "grad_norm": 0.11947191506624222, + "learning_rate": 1.20936419916845e-07, + "loss": 0.0162, + "step": 115390 + }, + { + "epoch": 10.48900199963643, + "grad_norm": 0.12429262697696686, + "learning_rate": 1.2041313476407744e-07, + "loss": 0.0161, + "step": 115400 + }, + { + "epoch": 10.489910925286312, + "grad_norm": 0.18892206251621246, + "learning_rate": 1.1989097961773666e-07, + "loss": 0.0156, + "step": 115410 + }, + { + "epoch": 10.490819850936193, + "grad_norm": 0.11997628957033157, + "learning_rate": 1.193699545174759e-07, + "loss": 0.0165, + "step": 115420 + }, + { + "epoch": 10.491728776586076, + "grad_norm": 0.11792425811290741, + "learning_rate": 1.1885005950286354e-07, + "loss": 0.0165, + "step": 115430 + }, + { + "epoch": 10.492637702235957, + "grad_norm": 0.1805015504360199, + "learning_rate": 1.1833129461338133e-07, + "loss": 0.0158, + "step": 115440 + }, + { + "epoch": 10.49354662788584, + "grad_norm": 0.19086433947086334, + "learning_rate": 1.1781365988842774e-07, + "loss": 0.0163, + "step": 115450 + }, + { + "epoch": 10.49445555353572, + "grad_norm": 0.1020715981721878, + "learning_rate": 1.1729715536731134e-07, + "loss": 0.0149, + "step": 115460 + }, + { + "epoch": 10.495364479185602, + "grad_norm": 0.11111854016780853, + "learning_rate": 1.1678178108926075e-07, + "loss": 0.016, + "step": 115470 + }, + { + "epoch": 10.496273404835485, + "grad_norm": 0.15802699327468872, + "learning_rate": 1.1626753709341132e-07, + "loss": 0.0152, + "step": 115480 + }, + { + "epoch": 10.497182330485366, + "grad_norm": 0.14620518684387207, + "learning_rate": 1.1575442341882014e-07, + "loss": 0.0164, + "step": 115490 + }, + { + "epoch": 10.498091256135249, + "grad_norm": 0.15051205456256866, + "learning_rate": 1.1524244010445106e-07, + "loss": 0.015, + "step": 115500 + }, + { + "epoch": 10.49900018178513, + "grad_norm": 0.12720468640327454, + "learning_rate": 1.1473158718918963e-07, + "loss": 0.0157, + "step": 115510 + }, + { + "epoch": 10.499909107435013, + "grad_norm": 0.13448214530944824, + "learning_rate": 1.1422186471182816e-07, + "loss": 0.0173, + "step": 115520 + }, + { + "epoch": 10.500818033084894, + "grad_norm": 0.13915368914604187, + "learning_rate": 1.1371327271107901e-07, + "loss": 0.0174, + "step": 115530 + }, + { + "epoch": 10.501726958734775, + "grad_norm": 0.11144927144050598, + "learning_rate": 1.1320581122556628e-07, + "loss": 0.0147, + "step": 115540 + }, + { + "epoch": 10.502635884384658, + "grad_norm": 0.14391553401947021, + "learning_rate": 1.1269948029382915e-07, + "loss": 0.0153, + "step": 115550 + }, + { + "epoch": 10.503544810034539, + "grad_norm": 0.09980274736881256, + "learning_rate": 1.1219427995431685e-07, + "loss": 0.0154, + "step": 115560 + }, + { + "epoch": 10.504453735684422, + "grad_norm": 0.11312098056077957, + "learning_rate": 1.1169021024539872e-07, + "loss": 0.0152, + "step": 115570 + }, + { + "epoch": 10.505362661334303, + "grad_norm": 0.21414095163345337, + "learning_rate": 1.111872712053541e-07, + "loss": 0.0161, + "step": 115580 + }, + { + "epoch": 10.506271586984184, + "grad_norm": 0.0781923234462738, + "learning_rate": 1.1068546287237913e-07, + "loss": 0.0151, + "step": 115590 + }, + { + "epoch": 10.507180512634067, + "grad_norm": 0.17330068349838257, + "learning_rate": 1.1018478528458332e-07, + "loss": 0.0167, + "step": 115600 + }, + { + "epoch": 10.508089438283948, + "grad_norm": 0.11999895423650742, + "learning_rate": 1.096852384799879e-07, + "loss": 0.0159, + "step": 115610 + }, + { + "epoch": 10.50899836393383, + "grad_norm": 0.1293952763080597, + "learning_rate": 1.0918682249653255e-07, + "loss": 0.0143, + "step": 115620 + }, + { + "epoch": 10.509907289583712, + "grad_norm": 0.10967563092708588, + "learning_rate": 1.0868953737206533e-07, + "loss": 0.0151, + "step": 115630 + }, + { + "epoch": 10.510816215233595, + "grad_norm": 0.12251003086566925, + "learning_rate": 1.0819338314435267e-07, + "loss": 0.0143, + "step": 115640 + }, + { + "epoch": 10.511725140883476, + "grad_norm": 0.13524764776229858, + "learning_rate": 1.0769835985107779e-07, + "loss": 0.0143, + "step": 115650 + }, + { + "epoch": 10.512634066533357, + "grad_norm": 0.15589477121829987, + "learning_rate": 1.072044675298306e-07, + "loss": 0.0147, + "step": 115660 + }, + { + "epoch": 10.51354299218324, + "grad_norm": 0.10970349609851837, + "learning_rate": 1.0671170621811943e-07, + "loss": 0.0166, + "step": 115670 + }, + { + "epoch": 10.51445191783312, + "grad_norm": 0.26494932174682617, + "learning_rate": 1.0622007595336936e-07, + "loss": 0.0164, + "step": 115680 + }, + { + "epoch": 10.515360843483004, + "grad_norm": 0.2228367179632187, + "learning_rate": 1.0572957677291217e-07, + "loss": 0.0167, + "step": 115690 + }, + { + "epoch": 10.516269769132885, + "grad_norm": 0.15754787623882294, + "learning_rate": 1.0524020871400141e-07, + "loss": 0.0134, + "step": 115700 + }, + { + "epoch": 10.517178694782768, + "grad_norm": 0.3034052848815918, + "learning_rate": 1.0475197181379903e-07, + "loss": 0.016, + "step": 115710 + }, + { + "epoch": 10.518087620432649, + "grad_norm": 0.12560440599918365, + "learning_rate": 1.0426486610938368e-07, + "loss": 0.0146, + "step": 115720 + }, + { + "epoch": 10.51899654608253, + "grad_norm": 0.11570723354816437, + "learning_rate": 1.0377889163774911e-07, + "loss": 0.017, + "step": 115730 + }, + { + "epoch": 10.519905471732413, + "grad_norm": 0.16776461899280548, + "learning_rate": 1.0329404843580248e-07, + "loss": 0.0171, + "step": 115740 + }, + { + "epoch": 10.520814397382294, + "grad_norm": 0.3749101161956787, + "learning_rate": 1.0281033654036098e-07, + "loss": 0.015, + "step": 115750 + }, + { + "epoch": 10.521723323032177, + "grad_norm": 0.1422327607870102, + "learning_rate": 1.0232775598816358e-07, + "loss": 0.0159, + "step": 115760 + }, + { + "epoch": 10.522632248682058, + "grad_norm": 0.12746556103229523, + "learning_rate": 1.0184630681585427e-07, + "loss": 0.0148, + "step": 115770 + }, + { + "epoch": 10.523541174331939, + "grad_norm": 0.1587931215763092, + "learning_rate": 1.013659890600005e-07, + "loss": 0.0164, + "step": 115780 + }, + { + "epoch": 10.524450099981822, + "grad_norm": 0.16218426823616028, + "learning_rate": 1.0088680275707807e-07, + "loss": 0.0178, + "step": 115790 + }, + { + "epoch": 10.525359025631703, + "grad_norm": 0.13527677953243256, + "learning_rate": 1.004087479434762e-07, + "loss": 0.0181, + "step": 115800 + }, + { + "epoch": 10.526267951281586, + "grad_norm": 0.12002506107091904, + "learning_rate": 9.993182465550254e-08, + "loss": 0.0151, + "step": 115810 + }, + { + "epoch": 10.527176876931467, + "grad_norm": 0.10555748641490936, + "learning_rate": 9.945603292937311e-08, + "loss": 0.0146, + "step": 115820 + }, + { + "epoch": 10.528085802581348, + "grad_norm": 0.524749219417572, + "learning_rate": 9.8981372801224e-08, + "loss": 0.0136, + "step": 115830 + }, + { + "epoch": 10.52899472823123, + "grad_norm": 0.16430965065956116, + "learning_rate": 9.850784430710135e-08, + "loss": 0.0157, + "step": 115840 + }, + { + "epoch": 10.529903653881112, + "grad_norm": 0.1791956126689911, + "learning_rate": 9.803544748296644e-08, + "loss": 0.0175, + "step": 115850 + }, + { + "epoch": 10.530812579530995, + "grad_norm": 0.19844257831573486, + "learning_rate": 9.756418236469555e-08, + "loss": 0.0158, + "step": 115860 + }, + { + "epoch": 10.531721505180876, + "grad_norm": 0.13097289204597473, + "learning_rate": 9.709404898807673e-08, + "loss": 0.0164, + "step": 115870 + }, + { + "epoch": 10.532630430830759, + "grad_norm": 0.19170716404914856, + "learning_rate": 9.662504738881473e-08, + "loss": 0.0164, + "step": 115880 + }, + { + "epoch": 10.53353935648064, + "grad_norm": 0.09133142232894897, + "learning_rate": 9.615717760252607e-08, + "loss": 0.0169, + "step": 115890 + }, + { + "epoch": 10.534448282130521, + "grad_norm": 0.13085073232650757, + "learning_rate": 9.569043966474233e-08, + "loss": 0.0171, + "step": 115900 + }, + { + "epoch": 10.535357207780404, + "grad_norm": 0.1253242790699005, + "learning_rate": 9.522483361091183e-08, + "loss": 0.0146, + "step": 115910 + }, + { + "epoch": 10.536266133430285, + "grad_norm": 0.12498795241117477, + "learning_rate": 9.476035947639128e-08, + "loss": 0.0153, + "step": 115920 + }, + { + "epoch": 10.537175059080168, + "grad_norm": 0.1403212994337082, + "learning_rate": 9.429701729645413e-08, + "loss": 0.013, + "step": 115930 + }, + { + "epoch": 10.538083984730049, + "grad_norm": 0.17945507168769836, + "learning_rate": 9.383480710629056e-08, + "loss": 0.0137, + "step": 115940 + }, + { + "epoch": 10.538992910379932, + "grad_norm": 0.14287039637565613, + "learning_rate": 9.337372894100248e-08, + "loss": 0.0154, + "step": 115950 + }, + { + "epoch": 10.539901836029813, + "grad_norm": 0.09980426728725433, + "learning_rate": 9.291378283560358e-08, + "loss": 0.0149, + "step": 115960 + }, + { + "epoch": 10.540810761679694, + "grad_norm": 0.14271576702594757, + "learning_rate": 9.245496882502258e-08, + "loss": 0.0154, + "step": 115970 + }, + { + "epoch": 10.541719687329577, + "grad_norm": 0.2334717959165573, + "learning_rate": 9.199728694410825e-08, + "loss": 0.0163, + "step": 115980 + }, + { + "epoch": 10.542628612979458, + "grad_norm": 0.22226200997829437, + "learning_rate": 9.154073722761613e-08, + "loss": 0.0151, + "step": 115990 + }, + { + "epoch": 10.54353753862934, + "grad_norm": 0.13959093391895294, + "learning_rate": 9.108531971021516e-08, + "loss": 0.0155, + "step": 116000 + }, + { + "epoch": 10.544446464279222, + "grad_norm": 0.21370665729045868, + "learning_rate": 9.063103442649434e-08, + "loss": 0.0158, + "step": 116010 + }, + { + "epoch": 10.545355389929103, + "grad_norm": 0.1639741212129593, + "learning_rate": 9.01778814109544e-08, + "loss": 0.0161, + "step": 116020 + }, + { + "epoch": 10.546264315578986, + "grad_norm": 0.20991575717926025, + "learning_rate": 8.972586069800948e-08, + "loss": 0.0135, + "step": 116030 + }, + { + "epoch": 10.547173241228867, + "grad_norm": 0.10206429660320282, + "learning_rate": 8.927497232198379e-08, + "loss": 0.0174, + "step": 116040 + }, + { + "epoch": 10.54808216687875, + "grad_norm": 0.2186601161956787, + "learning_rate": 8.88252163171216e-08, + "loss": 0.0161, + "step": 116050 + }, + { + "epoch": 10.548991092528631, + "grad_norm": 0.1523003876209259, + "learning_rate": 8.837659271758058e-08, + "loss": 0.0142, + "step": 116060 + }, + { + "epoch": 10.549900018178512, + "grad_norm": 0.15309551358222961, + "learning_rate": 8.79291015574285e-08, + "loss": 0.0169, + "step": 116070 + }, + { + "epoch": 10.550808943828395, + "grad_norm": 0.13983574509620667, + "learning_rate": 8.748274287064817e-08, + "loss": 0.0154, + "step": 116080 + }, + { + "epoch": 10.551717869478276, + "grad_norm": 0.15079091489315033, + "learning_rate": 8.703751669114246e-08, + "loss": 0.0153, + "step": 116090 + }, + { + "epoch": 10.552626795128159, + "grad_norm": 0.13633878529071808, + "learning_rate": 8.659342305271933e-08, + "loss": 0.0156, + "step": 116100 + }, + { + "epoch": 10.55353572077804, + "grad_norm": 0.12712135910987854, + "learning_rate": 8.615046198910514e-08, + "loss": 0.015, + "step": 116110 + }, + { + "epoch": 10.554444646427923, + "grad_norm": 0.1660947948694229, + "learning_rate": 8.570863353394132e-08, + "loss": 0.0146, + "step": 116120 + }, + { + "epoch": 10.555353572077804, + "grad_norm": 0.1264820247888565, + "learning_rate": 8.526793772078101e-08, + "loss": 0.0144, + "step": 116130 + }, + { + "epoch": 10.556262497727685, + "grad_norm": 0.13879480957984924, + "learning_rate": 8.482837458309079e-08, + "loss": 0.0155, + "step": 116140 + }, + { + "epoch": 10.557171423377568, + "grad_norm": 0.16341786086559296, + "learning_rate": 8.43899441542556e-08, + "loss": 0.0164, + "step": 116150 + }, + { + "epoch": 10.558080349027449, + "grad_norm": 0.1403639167547226, + "learning_rate": 8.395264646756884e-08, + "loss": 0.0151, + "step": 116160 + }, + { + "epoch": 10.558989274677332, + "grad_norm": 0.08128473907709122, + "learning_rate": 8.351648155624225e-08, + "loss": 0.0153, + "step": 116170 + }, + { + "epoch": 10.559898200327213, + "grad_norm": 0.09851615130901337, + "learning_rate": 8.308144945339768e-08, + "loss": 0.016, + "step": 116180 + }, + { + "epoch": 10.560807125977096, + "grad_norm": 0.20171834528446198, + "learning_rate": 8.26475501920737e-08, + "loss": 0.0161, + "step": 116190 + }, + { + "epoch": 10.561716051626977, + "grad_norm": 0.12324430048465729, + "learning_rate": 8.22147838052223e-08, + "loss": 0.0158, + "step": 116200 + }, + { + "epoch": 10.562624977276858, + "grad_norm": 0.14031630754470825, + "learning_rate": 8.178315032571049e-08, + "loss": 0.0189, + "step": 116210 + }, + { + "epoch": 10.563533902926741, + "grad_norm": 0.1154051274061203, + "learning_rate": 8.135264978631707e-08, + "loss": 0.015, + "step": 116220 + }, + { + "epoch": 10.564442828576622, + "grad_norm": 0.12048709392547607, + "learning_rate": 8.092328221973256e-08, + "loss": 0.0151, + "step": 116230 + }, + { + "epoch": 10.565351754226505, + "grad_norm": 0.1575593501329422, + "learning_rate": 8.04950476585692e-08, + "loss": 0.0139, + "step": 116240 + }, + { + "epoch": 10.566260679876386, + "grad_norm": 0.16641348600387573, + "learning_rate": 8.006794613534762e-08, + "loss": 0.0145, + "step": 116250 + }, + { + "epoch": 10.567169605526267, + "grad_norm": 0.10569335520267487, + "learning_rate": 7.964197768250192e-08, + "loss": 0.0129, + "step": 116260 + }, + { + "epoch": 10.56807853117615, + "grad_norm": 0.12484464049339294, + "learning_rate": 7.92171423323812e-08, + "loss": 0.0149, + "step": 116270 + }, + { + "epoch": 10.568987456826031, + "grad_norm": 0.18100564181804657, + "learning_rate": 7.879344011725132e-08, + "loss": 0.0156, + "step": 116280 + }, + { + "epoch": 10.569896382475914, + "grad_norm": 0.10476602613925934, + "learning_rate": 7.83708710692882e-08, + "loss": 0.0139, + "step": 116290 + }, + { + "epoch": 10.570805308125795, + "grad_norm": 0.16341976821422577, + "learning_rate": 7.794943522058285e-08, + "loss": 0.015, + "step": 116300 + }, + { + "epoch": 10.571714233775676, + "grad_norm": 0.15150199830532074, + "learning_rate": 7.752913260313965e-08, + "loss": 0.0148, + "step": 116310 + }, + { + "epoch": 10.572623159425559, + "grad_norm": 0.11433573812246323, + "learning_rate": 7.710996324888142e-08, + "loss": 0.0144, + "step": 116320 + }, + { + "epoch": 10.57353208507544, + "grad_norm": 0.14015497267246246, + "learning_rate": 7.669192718963769e-08, + "loss": 0.0134, + "step": 116330 + }, + { + "epoch": 10.574441010725323, + "grad_norm": 0.16761039197444916, + "learning_rate": 7.62750244571564e-08, + "loss": 0.015, + "step": 116340 + }, + { + "epoch": 10.575349936375204, + "grad_norm": 0.1750180870294571, + "learning_rate": 7.585925508309887e-08, + "loss": 0.0148, + "step": 116350 + }, + { + "epoch": 10.576258862025087, + "grad_norm": 0.303083598613739, + "learning_rate": 7.544461909904154e-08, + "loss": 0.0152, + "step": 116360 + }, + { + "epoch": 10.577167787674968, + "grad_norm": 0.1969505101442337, + "learning_rate": 7.503111653647087e-08, + "loss": 0.0145, + "step": 116370 + }, + { + "epoch": 10.57807671332485, + "grad_norm": 0.14347274601459503, + "learning_rate": 7.46187474267901e-08, + "loss": 0.016, + "step": 116380 + }, + { + "epoch": 10.578985638974732, + "grad_norm": 0.14064644277095795, + "learning_rate": 7.420751180131913e-08, + "loss": 0.0161, + "step": 116390 + }, + { + "epoch": 10.579894564624613, + "grad_norm": 0.14013908803462982, + "learning_rate": 7.379740969128468e-08, + "loss": 0.0138, + "step": 116400 + }, + { + "epoch": 10.580803490274496, + "grad_norm": 0.11593649536371231, + "learning_rate": 7.338844112783182e-08, + "loss": 0.0157, + "step": 116410 + }, + { + "epoch": 10.581712415924377, + "grad_norm": 0.09408050775527954, + "learning_rate": 7.298060614201907e-08, + "loss": 0.0147, + "step": 116420 + }, + { + "epoch": 10.58262134157426, + "grad_norm": 0.13929229974746704, + "learning_rate": 7.257390476482162e-08, + "loss": 0.0174, + "step": 116430 + }, + { + "epoch": 10.583530267224141, + "grad_norm": 0.1692131757736206, + "learning_rate": 7.216833702712145e-08, + "loss": 0.0163, + "step": 116440 + }, + { + "epoch": 10.584439192874022, + "grad_norm": 0.1461973488330841, + "learning_rate": 7.176390295972223e-08, + "loss": 0.0182, + "step": 116450 + }, + { + "epoch": 10.585348118523905, + "grad_norm": 0.1854832023382187, + "learning_rate": 7.136060259333776e-08, + "loss": 0.0155, + "step": 116460 + }, + { + "epoch": 10.586257044173786, + "grad_norm": 0.14324449002742767, + "learning_rate": 7.095843595859352e-08, + "loss": 0.0154, + "step": 116470 + }, + { + "epoch": 10.587165969823669, + "grad_norm": 0.14392699301242828, + "learning_rate": 7.055740308603342e-08, + "loss": 0.0147, + "step": 116480 + }, + { + "epoch": 10.58807489547355, + "grad_norm": 0.2318502962589264, + "learning_rate": 7.015750400611309e-08, + "loss": 0.0143, + "step": 116490 + }, + { + "epoch": 10.588983821123431, + "grad_norm": 0.09355450421571732, + "learning_rate": 6.975873874920158e-08, + "loss": 0.0165, + "step": 116500 + }, + { + "epoch": 10.589892746773314, + "grad_norm": 0.1695730835199356, + "learning_rate": 6.936110734558298e-08, + "loss": 0.0137, + "step": 116510 + }, + { + "epoch": 10.590801672423195, + "grad_norm": 0.12234403938055038, + "learning_rate": 6.896460982545483e-08, + "loss": 0.0155, + "step": 116520 + }, + { + "epoch": 10.591710598073078, + "grad_norm": 0.18566973507404327, + "learning_rate": 6.856924621892802e-08, + "loss": 0.0164, + "step": 116530 + }, + { + "epoch": 10.59261952372296, + "grad_norm": 0.1397365778684616, + "learning_rate": 6.817501655602687e-08, + "loss": 0.015, + "step": 116540 + }, + { + "epoch": 10.593528449372842, + "grad_norm": 0.1071656346321106, + "learning_rate": 6.77819208666941e-08, + "loss": 0.0161, + "step": 116550 + }, + { + "epoch": 10.594437375022723, + "grad_norm": 0.329717218875885, + "learning_rate": 6.738995918077916e-08, + "loss": 0.0139, + "step": 116560 + }, + { + "epoch": 10.595346300672604, + "grad_norm": 0.10330381244421005, + "learning_rate": 6.699913152804992e-08, + "loss": 0.0169, + "step": 116570 + }, + { + "epoch": 10.596255226322487, + "grad_norm": 0.17268650233745575, + "learning_rate": 6.660943793818764e-08, + "loss": 0.014, + "step": 116580 + }, + { + "epoch": 10.597164151972368, + "grad_norm": 0.1166575625538826, + "learning_rate": 6.622087844078862e-08, + "loss": 0.017, + "step": 116590 + }, + { + "epoch": 10.598073077622251, + "grad_norm": 0.1239418089389801, + "learning_rate": 6.58334530653576e-08, + "loss": 0.016, + "step": 116600 + }, + { + "epoch": 10.598982003272132, + "grad_norm": 0.10924037545919418, + "learning_rate": 6.544716184131772e-08, + "loss": 0.0124, + "step": 116610 + }, + { + "epoch": 10.599890928922015, + "grad_norm": 0.15094009041786194, + "learning_rate": 6.506200479800883e-08, + "loss": 0.0149, + "step": 116620 + }, + { + "epoch": 10.600799854571896, + "grad_norm": 0.11728499084711075, + "learning_rate": 6.467798196467755e-08, + "loss": 0.0149, + "step": 116630 + }, + { + "epoch": 10.601708780221777, + "grad_norm": 0.1102827787399292, + "learning_rate": 6.429509337048889e-08, + "loss": 0.0136, + "step": 116640 + }, + { + "epoch": 10.60261770587166, + "grad_norm": 0.1404903084039688, + "learning_rate": 6.391333904452124e-08, + "loss": 0.0154, + "step": 116650 + }, + { + "epoch": 10.603526631521541, + "grad_norm": 0.19185182452201843, + "learning_rate": 6.353271901576641e-08, + "loss": 0.0156, + "step": 116660 + }, + { + "epoch": 10.604435557171424, + "grad_norm": 0.09924865514039993, + "learning_rate": 6.315323331312794e-08, + "loss": 0.0172, + "step": 116670 + }, + { + "epoch": 10.605344482821305, + "grad_norm": 0.07162746787071228, + "learning_rate": 6.27748819654278e-08, + "loss": 0.0135, + "step": 116680 + }, + { + "epoch": 10.606253408471186, + "grad_norm": 0.11963045597076416, + "learning_rate": 6.239766500139799e-08, + "loss": 0.0147, + "step": 116690 + }, + { + "epoch": 10.60716233412107, + "grad_norm": 0.1859874129295349, + "learning_rate": 6.202158244968559e-08, + "loss": 0.0188, + "step": 116700 + }, + { + "epoch": 10.60807125977095, + "grad_norm": 0.14746375381946564, + "learning_rate": 6.164663433885276e-08, + "loss": 0.0149, + "step": 116710 + }, + { + "epoch": 10.608980185420833, + "grad_norm": 0.15713828802108765, + "learning_rate": 6.127282069737172e-08, + "loss": 0.0132, + "step": 116720 + }, + { + "epoch": 10.609889111070714, + "grad_norm": 0.14726081490516663, + "learning_rate": 6.090014155363477e-08, + "loss": 0.0151, + "step": 116730 + }, + { + "epoch": 10.610798036720595, + "grad_norm": 0.2516188323497772, + "learning_rate": 6.05285969359426e-08, + "loss": 0.0154, + "step": 116740 + }, + { + "epoch": 10.611706962370478, + "grad_norm": 0.12859608232975006, + "learning_rate": 6.015818687251096e-08, + "loss": 0.0151, + "step": 116750 + }, + { + "epoch": 10.61261588802036, + "grad_norm": 0.11967010796070099, + "learning_rate": 5.978891139147069e-08, + "loss": 0.0149, + "step": 116760 + }, + { + "epoch": 10.613524813670242, + "grad_norm": 0.14030654728412628, + "learning_rate": 5.942077052086603e-08, + "loss": 0.0155, + "step": 116770 + }, + { + "epoch": 10.614433739320123, + "grad_norm": 0.19334566593170166, + "learning_rate": 5.905376428865461e-08, + "loss": 0.0134, + "step": 116780 + }, + { + "epoch": 10.615342664970006, + "grad_norm": 0.1335577815771103, + "learning_rate": 5.868789272270747e-08, + "loss": 0.0156, + "step": 116790 + }, + { + "epoch": 10.616251590619887, + "grad_norm": 0.13052761554718018, + "learning_rate": 5.832315585081072e-08, + "loss": 0.0148, + "step": 116800 + }, + { + "epoch": 10.617160516269768, + "grad_norm": 0.24409820139408112, + "learning_rate": 5.7959553700665526e-08, + "loss": 0.0158, + "step": 116810 + }, + { + "epoch": 10.618069441919651, + "grad_norm": 0.148975670337677, + "learning_rate": 5.759708629988148e-08, + "loss": 0.0159, + "step": 116820 + }, + { + "epoch": 10.618978367569532, + "grad_norm": 0.1374858170747757, + "learning_rate": 5.723575367598821e-08, + "loss": 0.0139, + "step": 116830 + }, + { + "epoch": 10.619887293219415, + "grad_norm": 0.10817321389913559, + "learning_rate": 5.687555585642546e-08, + "loss": 0.0131, + "step": 116840 + }, + { + "epoch": 10.620796218869296, + "grad_norm": 0.15364232659339905, + "learning_rate": 5.651649286854799e-08, + "loss": 0.014, + "step": 116850 + }, + { + "epoch": 10.62170514451918, + "grad_norm": 0.09213026612997055, + "learning_rate": 5.615856473962566e-08, + "loss": 0.0145, + "step": 116860 + }, + { + "epoch": 10.62261407016906, + "grad_norm": 0.12705083191394806, + "learning_rate": 5.580177149683841e-08, + "loss": 0.016, + "step": 116870 + }, + { + "epoch": 10.623522995818941, + "grad_norm": 0.1057448461651802, + "learning_rate": 5.544611316728454e-08, + "loss": 0.0151, + "step": 116880 + }, + { + "epoch": 10.624431921468824, + "grad_norm": 0.13460320234298706, + "learning_rate": 5.5091589777972464e-08, + "loss": 0.0161, + "step": 116890 + }, + { + "epoch": 10.625340847118705, + "grad_norm": 0.12863457202911377, + "learning_rate": 5.4738201355825635e-08, + "loss": 0.014, + "step": 116900 + }, + { + "epoch": 10.626249772768588, + "grad_norm": 0.11340285837650299, + "learning_rate": 5.438594792768259e-08, + "loss": 0.0143, + "step": 116910 + }, + { + "epoch": 10.62715869841847, + "grad_norm": 0.13584545254707336, + "learning_rate": 5.403482952029526e-08, + "loss": 0.0168, + "step": 116920 + }, + { + "epoch": 10.62806762406835, + "grad_norm": 0.1160559132695198, + "learning_rate": 5.3684846160327314e-08, + "loss": 0.0131, + "step": 116930 + }, + { + "epoch": 10.628976549718233, + "grad_norm": 0.13537581264972687, + "learning_rate": 5.33359978743575e-08, + "loss": 0.0146, + "step": 116940 + }, + { + "epoch": 10.629885475368114, + "grad_norm": 0.12100183963775635, + "learning_rate": 5.2988284688879616e-08, + "loss": 0.016, + "step": 116950 + }, + { + "epoch": 10.630794401017997, + "grad_norm": 0.1378898024559021, + "learning_rate": 5.2641706630299214e-08, + "loss": 0.0151, + "step": 116960 + }, + { + "epoch": 10.631703326667878, + "grad_norm": 0.09339484572410583, + "learning_rate": 5.229626372493856e-08, + "loss": 0.0149, + "step": 116970 + }, + { + "epoch": 10.63261225231776, + "grad_norm": 0.11117950081825256, + "learning_rate": 5.1951955999028354e-08, + "loss": 0.0162, + "step": 116980 + }, + { + "epoch": 10.633521177967642, + "grad_norm": 0.16742952167987823, + "learning_rate": 5.1608783478721e-08, + "loss": 0.0154, + "step": 116990 + }, + { + "epoch": 10.634430103617523, + "grad_norm": 0.14477165043354034, + "learning_rate": 5.126674619007565e-08, + "loss": 0.0146, + "step": 117000 + }, + { + "epoch": 10.635339029267406, + "grad_norm": 0.11790766566991806, + "learning_rate": 5.092584415906654e-08, + "loss": 0.0144, + "step": 117010 + }, + { + "epoch": 10.636247954917287, + "grad_norm": 0.17633572220802307, + "learning_rate": 5.058607741158461e-08, + "loss": 0.016, + "step": 117020 + }, + { + "epoch": 10.63715688056717, + "grad_norm": 0.16550585627555847, + "learning_rate": 5.0247445973432557e-08, + "loss": 0.0149, + "step": 117030 + }, + { + "epoch": 10.638065806217051, + "grad_norm": 0.13694334030151367, + "learning_rate": 4.990994987032649e-08, + "loss": 0.0165, + "step": 117040 + }, + { + "epoch": 10.638974731866933, + "grad_norm": 0.1677303910255432, + "learning_rate": 4.9573589127899225e-08, + "loss": 0.0153, + "step": 117050 + }, + { + "epoch": 10.639883657516815, + "grad_norm": 0.15477727353572845, + "learning_rate": 4.9238363771691994e-08, + "loss": 0.0169, + "step": 117060 + }, + { + "epoch": 10.640792583166697, + "grad_norm": 0.1158832535147667, + "learning_rate": 4.890427382716611e-08, + "loss": 0.0131, + "step": 117070 + }, + { + "epoch": 10.64170150881658, + "grad_norm": 0.13209748268127441, + "learning_rate": 4.8571319319691276e-08, + "loss": 0.0156, + "step": 117080 + }, + { + "epoch": 10.64261043446646, + "grad_norm": 0.15046443045139313, + "learning_rate": 4.8239500274553924e-08, + "loss": 0.0133, + "step": 117090 + }, + { + "epoch": 10.643519360116343, + "grad_norm": 0.18214501440525055, + "learning_rate": 4.790881671695391e-08, + "loss": 0.0145, + "step": 117100 + }, + { + "epoch": 10.644428285766224, + "grad_norm": 0.10677812993526459, + "learning_rate": 4.75792686720028e-08, + "loss": 0.0146, + "step": 117110 + }, + { + "epoch": 10.645337211416106, + "grad_norm": 0.17077256739139557, + "learning_rate": 4.725085616473057e-08, + "loss": 0.0158, + "step": 117120 + }, + { + "epoch": 10.646246137065988, + "grad_norm": 0.11755786091089249, + "learning_rate": 4.6923579220075616e-08, + "loss": 0.0163, + "step": 117130 + }, + { + "epoch": 10.64715506271587, + "grad_norm": 0.1168491393327713, + "learning_rate": 4.659743786289306e-08, + "loss": 0.019, + "step": 117140 + }, + { + "epoch": 10.648063988365752, + "grad_norm": 0.18194228410720825, + "learning_rate": 4.627243211794974e-08, + "loss": 0.0164, + "step": 117150 + }, + { + "epoch": 10.648972914015634, + "grad_norm": 0.11531153321266174, + "learning_rate": 4.5948562009930915e-08, + "loss": 0.0142, + "step": 117160 + }, + { + "epoch": 10.649881839665515, + "grad_norm": 0.14606709778308868, + "learning_rate": 4.562582756343026e-08, + "loss": 0.0154, + "step": 117170 + }, + { + "epoch": 10.650790765315397, + "grad_norm": 0.1556229591369629, + "learning_rate": 4.530422880295815e-08, + "loss": 0.0159, + "step": 117180 + }, + { + "epoch": 10.651699690965279, + "grad_norm": 0.1086379885673523, + "learning_rate": 4.498376575293672e-08, + "loss": 0.0154, + "step": 117190 + }, + { + "epoch": 10.652608616615161, + "grad_norm": 0.15304772555828094, + "learning_rate": 4.466443843770485e-08, + "loss": 0.0144, + "step": 117200 + }, + { + "epoch": 10.653517542265043, + "grad_norm": 0.10914810746908188, + "learning_rate": 4.434624688150979e-08, + "loss": 0.0139, + "step": 117210 + }, + { + "epoch": 10.654426467914924, + "grad_norm": 0.1537492871284485, + "learning_rate": 4.402919110852055e-08, + "loss": 0.0157, + "step": 117220 + }, + { + "epoch": 10.655335393564807, + "grad_norm": 0.132416233420372, + "learning_rate": 4.371327114281454e-08, + "loss": 0.0149, + "step": 117230 + }, + { + "epoch": 10.656244319214688, + "grad_norm": 0.1698627471923828, + "learning_rate": 4.33984870083809e-08, + "loss": 0.0143, + "step": 117240 + }, + { + "epoch": 10.65715324486457, + "grad_norm": 0.11798369139432907, + "learning_rate": 4.3084838729127165e-08, + "loss": 0.0148, + "step": 117250 + }, + { + "epoch": 10.658062170514452, + "grad_norm": 0.12901954352855682, + "learning_rate": 4.2772326328872624e-08, + "loss": 0.0152, + "step": 117260 + }, + { + "epoch": 10.658971096164334, + "grad_norm": 0.12010299414396286, + "learning_rate": 4.246094983135163e-08, + "loss": 0.0164, + "step": 117270 + }, + { + "epoch": 10.659880021814216, + "grad_norm": 0.1289399117231369, + "learning_rate": 4.215070926021025e-08, + "loss": 0.0159, + "step": 117280 + }, + { + "epoch": 10.660788947464097, + "grad_norm": 0.12634992599487305, + "learning_rate": 4.1841604639009654e-08, + "loss": 0.0139, + "step": 117290 + }, + { + "epoch": 10.66169787311398, + "grad_norm": 0.13452400267124176, + "learning_rate": 4.153363599122273e-08, + "loss": 0.0185, + "step": 117300 + }, + { + "epoch": 10.66260679876386, + "grad_norm": 0.11372703313827515, + "learning_rate": 4.122680334024076e-08, + "loss": 0.0178, + "step": 117310 + }, + { + "epoch": 10.663515724413744, + "grad_norm": 0.13634739816188812, + "learning_rate": 4.092110670936178e-08, + "loss": 0.0157, + "step": 117320 + }, + { + "epoch": 10.664424650063625, + "grad_norm": 0.08816560357809067, + "learning_rate": 4.0616546121803874e-08, + "loss": 0.0151, + "step": 117330 + }, + { + "epoch": 10.665333575713507, + "grad_norm": 0.15019099414348602, + "learning_rate": 4.031312160069689e-08, + "loss": 0.0178, + "step": 117340 + }, + { + "epoch": 10.666242501363389, + "grad_norm": 0.18542854487895966, + "learning_rate": 4.0010833169082384e-08, + "loss": 0.0157, + "step": 117350 + }, + { + "epoch": 10.66715142701327, + "grad_norm": 0.17203901708126068, + "learning_rate": 3.9709680849916995e-08, + "loss": 0.0176, + "step": 117360 + }, + { + "epoch": 10.668060352663153, + "grad_norm": 0.23545393347740173, + "learning_rate": 3.94096646660741e-08, + "loss": 0.0146, + "step": 117370 + }, + { + "epoch": 10.668969278313034, + "grad_norm": 0.12844081223011017, + "learning_rate": 3.9110784640333796e-08, + "loss": 0.0149, + "step": 117380 + }, + { + "epoch": 10.669878203962917, + "grad_norm": 0.22382205724716187, + "learning_rate": 3.881304079539627e-08, + "loss": 0.016, + "step": 117390 + }, + { + "epoch": 10.670787129612798, + "grad_norm": 0.13726432621479034, + "learning_rate": 3.851643315387343e-08, + "loss": 0.0138, + "step": 117400 + }, + { + "epoch": 10.671696055262679, + "grad_norm": 0.23564179241657257, + "learning_rate": 3.822096173829059e-08, + "loss": 0.0148, + "step": 117410 + }, + { + "epoch": 10.672604980912562, + "grad_norm": 0.15721382200717926, + "learning_rate": 3.7926626571086474e-08, + "loss": 0.0152, + "step": 117420 + }, + { + "epoch": 10.673513906562443, + "grad_norm": 0.15414534509181976, + "learning_rate": 3.763342767461486e-08, + "loss": 0.0168, + "step": 117430 + }, + { + "epoch": 10.674422832212326, + "grad_norm": 0.11212006956338882, + "learning_rate": 3.734136507113961e-08, + "loss": 0.0154, + "step": 117440 + }, + { + "epoch": 10.675331757862207, + "grad_norm": 0.13570338487625122, + "learning_rate": 3.705043878284464e-08, + "loss": 0.0169, + "step": 117450 + }, + { + "epoch": 10.676240683512088, + "grad_norm": 0.23574671149253845, + "learning_rate": 3.676064883182062e-08, + "loss": 0.0144, + "step": 117460 + }, + { + "epoch": 10.67714960916197, + "grad_norm": 0.16503477096557617, + "learning_rate": 3.6471995240076605e-08, + "loss": 0.0153, + "step": 117470 + }, + { + "epoch": 10.678058534811852, + "grad_norm": 0.11990775913000107, + "learning_rate": 3.6184478029533394e-08, + "loss": 0.0133, + "step": 117480 + }, + { + "epoch": 10.678967460461735, + "grad_norm": 0.14900237321853638, + "learning_rate": 3.589809722202853e-08, + "loss": 0.0142, + "step": 117490 + }, + { + "epoch": 10.679876386111616, + "grad_norm": 0.11182516068220139, + "learning_rate": 3.561285283930626e-08, + "loss": 0.0159, + "step": 117500 + }, + { + "epoch": 10.680785311761499, + "grad_norm": 0.12805165350437164, + "learning_rate": 3.5328744903032614e-08, + "loss": 0.0146, + "step": 117510 + }, + { + "epoch": 10.68169423741138, + "grad_norm": 0.24836018681526184, + "learning_rate": 3.504577343478366e-08, + "loss": 0.0154, + "step": 117520 + }, + { + "epoch": 10.68260316306126, + "grad_norm": 0.1237228587269783, + "learning_rate": 3.47639384560472e-08, + "loss": 0.0171, + "step": 117530 + }, + { + "epoch": 10.683512088711144, + "grad_norm": 0.16414684057235718, + "learning_rate": 3.448323998822611e-08, + "loss": 0.0155, + "step": 117540 + }, + { + "epoch": 10.684421014361025, + "grad_norm": 0.1218625083565712, + "learning_rate": 3.420367805264002e-08, + "loss": 0.0148, + "step": 117550 + }, + { + "epoch": 10.685329940010908, + "grad_norm": 0.15386350452899933, + "learning_rate": 3.392525267052027e-08, + "loss": 0.0139, + "step": 117560 + }, + { + "epoch": 10.686238865660789, + "grad_norm": 0.2749406695365906, + "learning_rate": 3.3647963863009943e-08, + "loss": 0.016, + "step": 117570 + }, + { + "epoch": 10.687147791310672, + "grad_norm": 0.10001368075609207, + "learning_rate": 3.3371811651165543e-08, + "loss": 0.017, + "step": 117580 + }, + { + "epoch": 10.688056716960553, + "grad_norm": 0.10538922250270844, + "learning_rate": 3.309679605596194e-08, + "loss": 0.0157, + "step": 117590 + }, + { + "epoch": 10.688965642610434, + "grad_norm": 0.14606650173664093, + "learning_rate": 3.282291709828411e-08, + "loss": 0.0155, + "step": 117600 + }, + { + "epoch": 10.689874568260317, + "grad_norm": 0.13958720862865448, + "learning_rate": 3.255017479893041e-08, + "loss": 0.0142, + "step": 117610 + }, + { + "epoch": 10.690783493910198, + "grad_norm": 0.11254779994487762, + "learning_rate": 3.2278569178614273e-08, + "loss": 0.0148, + "step": 117620 + }, + { + "epoch": 10.69169241956008, + "grad_norm": 0.12928491830825806, + "learning_rate": 3.2008100257962524e-08, + "loss": 0.0171, + "step": 117630 + }, + { + "epoch": 10.692601345209962, + "grad_norm": 0.11208337545394897, + "learning_rate": 3.1738768057517073e-08, + "loss": 0.0156, + "step": 117640 + }, + { + "epoch": 10.693510270859843, + "grad_norm": 0.11948471516370773, + "learning_rate": 3.147057259772823e-08, + "loss": 0.0164, + "step": 117650 + }, + { + "epoch": 10.694419196509726, + "grad_norm": 0.16345296800136566, + "learning_rate": 3.120351389896636e-08, + "loss": 0.0164, + "step": 117660 + }, + { + "epoch": 10.695328122159607, + "grad_norm": 0.13255134224891663, + "learning_rate": 3.0937591981511916e-08, + "loss": 0.0159, + "step": 117670 + }, + { + "epoch": 10.69623704780949, + "grad_norm": 0.09886055439710617, + "learning_rate": 3.06728068655604e-08, + "loss": 0.0153, + "step": 117680 + }, + { + "epoch": 10.69714597345937, + "grad_norm": 0.17253190279006958, + "learning_rate": 3.0409158571219063e-08, + "loss": 0.0134, + "step": 117690 + }, + { + "epoch": 10.698054899109254, + "grad_norm": 0.1375715434551239, + "learning_rate": 3.0146647118513556e-08, + "loss": 0.0153, + "step": 117700 + }, + { + "epoch": 10.698963824759135, + "grad_norm": 0.10064633190631866, + "learning_rate": 2.988527252737627e-08, + "loss": 0.0139, + "step": 117710 + }, + { + "epoch": 10.699872750409016, + "grad_norm": 0.1011563241481781, + "learning_rate": 2.9625034817657993e-08, + "loss": 0.0156, + "step": 117720 + }, + { + "epoch": 10.700781676058899, + "grad_norm": 0.08418525010347366, + "learning_rate": 2.936593400912457e-08, + "loss": 0.0143, + "step": 117730 + }, + { + "epoch": 10.70169060170878, + "grad_norm": 0.11103270947933197, + "learning_rate": 2.9107970121448612e-08, + "loss": 0.0143, + "step": 117740 + }, + { + "epoch": 10.702599527358663, + "grad_norm": 0.1426171362400055, + "learning_rate": 2.8851143174224437e-08, + "loss": 0.0137, + "step": 117750 + }, + { + "epoch": 10.703508453008544, + "grad_norm": 0.2344736009836197, + "learning_rate": 2.8595453186954777e-08, + "loss": 0.0163, + "step": 117760 + }, + { + "epoch": 10.704417378658425, + "grad_norm": 0.14876066148281097, + "learning_rate": 2.8340900179057438e-08, + "loss": 0.0168, + "step": 117770 + }, + { + "epoch": 10.705326304308308, + "grad_norm": 0.1607694774866104, + "learning_rate": 2.808748416986362e-08, + "loss": 0.0162, + "step": 117780 + }, + { + "epoch": 10.706235229958189, + "grad_norm": 0.12169775366783142, + "learning_rate": 2.783520517862126e-08, + "loss": 0.0157, + "step": 117790 + }, + { + "epoch": 10.707144155608072, + "grad_norm": 0.12050489336252213, + "learning_rate": 2.7584063224485034e-08, + "loss": 0.0156, + "step": 117800 + }, + { + "epoch": 10.708053081257953, + "grad_norm": 0.11776214838027954, + "learning_rate": 2.7334058326531354e-08, + "loss": 0.0156, + "step": 117810 + }, + { + "epoch": 10.708962006907836, + "grad_norm": 0.13920967280864716, + "learning_rate": 2.7085190503745027e-08, + "loss": 0.0149, + "step": 117820 + }, + { + "epoch": 10.709870932557717, + "grad_norm": 0.11310988664627075, + "learning_rate": 2.683745977502594e-08, + "loss": 0.0143, + "step": 117830 + }, + { + "epoch": 10.710779858207598, + "grad_norm": 0.1689884215593338, + "learning_rate": 2.659086615918571e-08, + "loss": 0.0135, + "step": 117840 + }, + { + "epoch": 10.71168878385748, + "grad_norm": 0.12853948771953583, + "learning_rate": 2.6345409674956024e-08, + "loss": 0.0153, + "step": 117850 + }, + { + "epoch": 10.712597709507362, + "grad_norm": 0.11099285632371902, + "learning_rate": 2.610109034097363e-08, + "loss": 0.0146, + "step": 117860 + }, + { + "epoch": 10.713506635157245, + "grad_norm": 0.1265709102153778, + "learning_rate": 2.5857908175793697e-08, + "loss": 0.0166, + "step": 117870 + }, + { + "epoch": 10.714415560807126, + "grad_norm": 0.17007219791412354, + "learning_rate": 2.5615863197884782e-08, + "loss": 0.015, + "step": 117880 + }, + { + "epoch": 10.715324486457007, + "grad_norm": 0.26896312832832336, + "learning_rate": 2.537495542563051e-08, + "loss": 0.0151, + "step": 117890 + }, + { + "epoch": 10.71623341210689, + "grad_norm": 0.13580460846424103, + "learning_rate": 2.5135184877322914e-08, + "loss": 0.0146, + "step": 117900 + }, + { + "epoch": 10.717142337756771, + "grad_norm": 0.11525449901819229, + "learning_rate": 2.4896551571172432e-08, + "loss": 0.0179, + "step": 117910 + }, + { + "epoch": 10.718051263406654, + "grad_norm": 0.11696870625019073, + "learning_rate": 2.4659055525302898e-08, + "loss": 0.0144, + "step": 117920 + }, + { + "epoch": 10.718960189056535, + "grad_norm": 0.12734751403331757, + "learning_rate": 2.442269675774822e-08, + "loss": 0.0163, + "step": 117930 + }, + { + "epoch": 10.719869114706418, + "grad_norm": 0.23203711211681366, + "learning_rate": 2.4187475286460702e-08, + "loss": 0.0154, + "step": 117940 + }, + { + "epoch": 10.720778040356299, + "grad_norm": 0.13575899600982666, + "learning_rate": 2.395339112930106e-08, + "loss": 0.0129, + "step": 117950 + }, + { + "epoch": 10.72168696600618, + "grad_norm": 0.2158169448375702, + "learning_rate": 2.372044430405007e-08, + "loss": 0.0166, + "step": 117960 + }, + { + "epoch": 10.722595891656063, + "grad_norm": 0.11915555596351624, + "learning_rate": 2.348863482839525e-08, + "loss": 0.0156, + "step": 117970 + }, + { + "epoch": 10.723504817305944, + "grad_norm": 0.14982236921787262, + "learning_rate": 2.3257962719942516e-08, + "loss": 0.0144, + "step": 117980 + }, + { + "epoch": 10.724413742955827, + "grad_norm": 0.16873042285442352, + "learning_rate": 2.302842799620952e-08, + "loss": 0.016, + "step": 117990 + }, + { + "epoch": 10.725322668605708, + "grad_norm": 0.17973820865154266, + "learning_rate": 2.2800030674627327e-08, + "loss": 0.0141, + "step": 118000 + }, + { + "epoch": 10.72623159425559, + "grad_norm": 0.16980043053627014, + "learning_rate": 2.2572770772543718e-08, + "loss": 0.0167, + "step": 118010 + }, + { + "epoch": 10.727140519905472, + "grad_norm": 0.11162397265434265, + "learning_rate": 2.234664830721489e-08, + "loss": 0.0152, + "step": 118020 + }, + { + "epoch": 10.728049445555353, + "grad_norm": 0.08425960689783096, + "learning_rate": 2.2121663295813776e-08, + "loss": 0.0157, + "step": 118030 + }, + { + "epoch": 10.728958371205236, + "grad_norm": 0.1590510904788971, + "learning_rate": 2.189781575542671e-08, + "loss": 0.0148, + "step": 118040 + }, + { + "epoch": 10.729867296855117, + "grad_norm": 0.0939348116517067, + "learning_rate": 2.167510570305342e-08, + "loss": 0.0177, + "step": 118050 + }, + { + "epoch": 10.730776222505, + "grad_norm": 0.1515420377254486, + "learning_rate": 2.145353315560872e-08, + "loss": 0.0151, + "step": 118060 + }, + { + "epoch": 10.731685148154881, + "grad_norm": 0.18417468667030334, + "learning_rate": 2.123309812991747e-08, + "loss": 0.0146, + "step": 118070 + }, + { + "epoch": 10.732594073804762, + "grad_norm": 0.14469967782497406, + "learning_rate": 2.1013800642722957e-08, + "loss": 0.016, + "step": 118080 + }, + { + "epoch": 10.733502999454645, + "grad_norm": 0.16235080361366272, + "learning_rate": 2.079564071067519e-08, + "loss": 0.015, + "step": 118090 + }, + { + "epoch": 10.734411925104526, + "grad_norm": 0.16947554051876068, + "learning_rate": 2.0578618350345912e-08, + "loss": 0.0157, + "step": 118100 + }, + { + "epoch": 10.735320850754409, + "grad_norm": 0.11546468734741211, + "learning_rate": 2.036273357821361e-08, + "loss": 0.0136, + "step": 118110 + }, + { + "epoch": 10.73622977640429, + "grad_norm": 0.24204279482364655, + "learning_rate": 2.0147986410675168e-08, + "loss": 0.0166, + "step": 118120 + }, + { + "epoch": 10.737138702054171, + "grad_norm": 0.15731032192707062, + "learning_rate": 1.9934376864039204e-08, + "loss": 0.0151, + "step": 118130 + }, + { + "epoch": 10.738047627704054, + "grad_norm": 0.18542906641960144, + "learning_rate": 1.9721904954527747e-08, + "loss": 0.0161, + "step": 118140 + }, + { + "epoch": 10.738956553353935, + "grad_norm": 0.13873739540576935, + "learning_rate": 1.951057069827622e-08, + "loss": 0.0153, + "step": 118150 + }, + { + "epoch": 10.739865479003818, + "grad_norm": 0.08732368797063828, + "learning_rate": 1.9300374111333452e-08, + "loss": 0.015, + "step": 118160 + }, + { + "epoch": 10.740774404653699, + "grad_norm": 0.13508310914039612, + "learning_rate": 1.909131520966334e-08, + "loss": 0.018, + "step": 118170 + }, + { + "epoch": 10.741683330303582, + "grad_norm": 0.1258886754512787, + "learning_rate": 1.8883394009143186e-08, + "loss": 0.0143, + "step": 118180 + }, + { + "epoch": 10.742592255953463, + "grad_norm": 0.1705961674451828, + "learning_rate": 1.867661052556202e-08, + "loss": 0.0166, + "step": 118190 + }, + { + "epoch": 10.743501181603344, + "grad_norm": 0.14960865676403046, + "learning_rate": 1.8470964774623954e-08, + "loss": 0.0168, + "step": 118200 + }, + { + "epoch": 10.744410107253227, + "grad_norm": 0.10661343485116959, + "learning_rate": 1.8266456771948158e-08, + "loss": 0.0158, + "step": 118210 + }, + { + "epoch": 10.745319032903108, + "grad_norm": 0.12774141132831573, + "learning_rate": 1.806308653306221e-08, + "loss": 0.0159, + "step": 118220 + }, + { + "epoch": 10.746227958552991, + "grad_norm": 0.11034590750932693, + "learning_rate": 1.786085407341376e-08, + "loss": 0.0146, + "step": 118230 + }, + { + "epoch": 10.747136884202872, + "grad_norm": 0.15547287464141846, + "learning_rate": 1.7659759408358845e-08, + "loss": 0.0142, + "step": 118240 + }, + { + "epoch": 10.748045809852755, + "grad_norm": 0.09741663187742233, + "learning_rate": 1.745980255317192e-08, + "loss": 0.0183, + "step": 118250 + }, + { + "epoch": 10.748954735502636, + "grad_norm": 0.12270299345254898, + "learning_rate": 1.7260983523035846e-08, + "loss": 0.0165, + "step": 118260 + }, + { + "epoch": 10.749863661152517, + "grad_norm": 0.1196635439991951, + "learning_rate": 1.70633023330502e-08, + "loss": 0.0152, + "step": 118270 + }, + { + "epoch": 10.7507725868024, + "grad_norm": 0.1065274178981781, + "learning_rate": 1.686675899822965e-08, + "loss": 0.0161, + "step": 118280 + }, + { + "epoch": 10.751681512452281, + "grad_norm": 0.11327704787254333, + "learning_rate": 1.6671353533497247e-08, + "loss": 0.0142, + "step": 118290 + }, + { + "epoch": 10.752590438102164, + "grad_norm": 0.12687520682811737, + "learning_rate": 1.647708595369446e-08, + "loss": 0.0163, + "step": 118300 + }, + { + "epoch": 10.753499363752045, + "grad_norm": 0.18002845346927643, + "learning_rate": 1.6283956273574484e-08, + "loss": 0.015, + "step": 118310 + }, + { + "epoch": 10.754408289401926, + "grad_norm": 0.18937845528125763, + "learning_rate": 1.6091964507803926e-08, + "loss": 0.0172, + "step": 118320 + }, + { + "epoch": 10.755317215051809, + "grad_norm": 0.11552153527736664, + "learning_rate": 1.5901110670962783e-08, + "loss": 0.0158, + "step": 118330 + }, + { + "epoch": 10.75622614070169, + "grad_norm": 0.11890138685703278, + "learning_rate": 1.5711394777546127e-08, + "loss": 0.0129, + "step": 118340 + }, + { + "epoch": 10.757135066351573, + "grad_norm": 0.14689825475215912, + "learning_rate": 1.552281684196244e-08, + "loss": 0.017, + "step": 118350 + }, + { + "epoch": 10.758043992001454, + "grad_norm": 0.108400858938694, + "learning_rate": 1.5335376878531927e-08, + "loss": 0.0161, + "step": 118360 + }, + { + "epoch": 10.758952917651335, + "grad_norm": 0.1434568464756012, + "learning_rate": 1.5149074901488202e-08, + "loss": 0.0144, + "step": 118370 + }, + { + "epoch": 10.759861843301218, + "grad_norm": 0.11728490144014359, + "learning_rate": 1.4963910924981616e-08, + "loss": 0.0158, + "step": 118380 + }, + { + "epoch": 10.7607707689511, + "grad_norm": 0.11118271201848984, + "learning_rate": 1.4779884963072587e-08, + "loss": 0.0148, + "step": 118390 + }, + { + "epoch": 10.761679694600982, + "grad_norm": 0.11372431367635727, + "learning_rate": 1.4596997029736603e-08, + "loss": 0.0143, + "step": 118400 + }, + { + "epoch": 10.762588620250863, + "grad_norm": 0.12700681388378143, + "learning_rate": 1.4415247138865883e-08, + "loss": 0.0174, + "step": 118410 + }, + { + "epoch": 10.763497545900746, + "grad_norm": 0.13466398417949677, + "learning_rate": 1.423463530425939e-08, + "loss": 0.0144, + "step": 118420 + }, + { + "epoch": 10.764406471550627, + "grad_norm": 0.13140025734901428, + "learning_rate": 1.4055161539636152e-08, + "loss": 0.0145, + "step": 118430 + }, + { + "epoch": 10.765315397200508, + "grad_norm": 0.16604061424732208, + "learning_rate": 1.3876825858623598e-08, + "loss": 0.0157, + "step": 118440 + }, + { + "epoch": 10.766224322850391, + "grad_norm": 0.1281573623418808, + "learning_rate": 1.3699628274767561e-08, + "loss": 0.0188, + "step": 118450 + }, + { + "epoch": 10.767133248500272, + "grad_norm": 0.0944821909070015, + "learning_rate": 1.352356880152228e-08, + "loss": 0.0146, + "step": 118460 + }, + { + "epoch": 10.768042174150155, + "grad_norm": 0.14113584160804749, + "learning_rate": 1.3348647452260387e-08, + "loss": 0.015, + "step": 118470 + }, + { + "epoch": 10.768951099800036, + "grad_norm": 0.09508536756038666, + "learning_rate": 1.317486424026626e-08, + "loss": 0.0159, + "step": 118480 + }, + { + "epoch": 10.769860025449919, + "grad_norm": 0.10452780872583389, + "learning_rate": 1.3002219178736007e-08, + "loss": 0.0175, + "step": 118490 + }, + { + "epoch": 10.7707689510998, + "grad_norm": 0.13348081707954407, + "learning_rate": 1.283071228078081e-08, + "loss": 0.0145, + "step": 118500 + }, + { + "epoch": 10.771677876749681, + "grad_norm": 0.14879164099693298, + "learning_rate": 1.2660343559426913e-08, + "loss": 0.0146, + "step": 118510 + }, + { + "epoch": 10.772586802399564, + "grad_norm": 0.1047331690788269, + "learning_rate": 1.24911130276123e-08, + "loss": 0.0158, + "step": 118520 + }, + { + "epoch": 10.773495728049445, + "grad_norm": 0.13582324981689453, + "learning_rate": 1.2323020698190023e-08, + "loss": 0.0155, + "step": 118530 + }, + { + "epoch": 10.774404653699328, + "grad_norm": 0.13367359340190887, + "learning_rate": 1.2156066583923208e-08, + "loss": 0.015, + "step": 118540 + }, + { + "epoch": 10.77531357934921, + "grad_norm": 0.11243392527103424, + "learning_rate": 1.1990250697490047e-08, + "loss": 0.0158, + "step": 118550 + }, + { + "epoch": 10.77622250499909, + "grad_norm": 0.10125166922807693, + "learning_rate": 1.1825573051487127e-08, + "loss": 0.0148, + "step": 118560 + }, + { + "epoch": 10.777131430648973, + "grad_norm": 0.11922574788331985, + "learning_rate": 1.1662033658419447e-08, + "loss": 0.0168, + "step": 118570 + }, + { + "epoch": 10.778040356298854, + "grad_norm": 0.12048574537038803, + "learning_rate": 1.149963253070374e-08, + "loss": 0.0143, + "step": 118580 + }, + { + "epoch": 10.778949281948737, + "grad_norm": 0.11338308453559875, + "learning_rate": 1.1338369680676809e-08, + "loss": 0.0163, + "step": 118590 + }, + { + "epoch": 10.779858207598618, + "grad_norm": 0.15135347843170166, + "learning_rate": 1.1178245120583851e-08, + "loss": 0.0175, + "step": 118600 + }, + { + "epoch": 10.7807671332485, + "grad_norm": 0.1709575206041336, + "learning_rate": 1.1019258862585146e-08, + "loss": 0.0152, + "step": 118610 + }, + { + "epoch": 10.781676058898382, + "grad_norm": 0.11269596964120865, + "learning_rate": 1.0861410918756032e-08, + "loss": 0.0157, + "step": 118620 + }, + { + "epoch": 10.782584984548263, + "grad_norm": 0.10537561774253845, + "learning_rate": 1.070470130108192e-08, + "loss": 0.0145, + "step": 118630 + }, + { + "epoch": 10.783493910198146, + "grad_norm": 0.14072760939598083, + "learning_rate": 1.0549130021464959e-08, + "loss": 0.015, + "step": 118640 + }, + { + "epoch": 10.784402835848027, + "grad_norm": 0.13263830542564392, + "learning_rate": 1.0394697091720695e-08, + "loss": 0.0156, + "step": 118650 + }, + { + "epoch": 10.78531176149791, + "grad_norm": 0.14002971351146698, + "learning_rate": 1.0241402523576416e-08, + "loss": 0.0147, + "step": 118660 + }, + { + "epoch": 10.786220687147791, + "grad_norm": 0.18648749589920044, + "learning_rate": 1.0089246328672807e-08, + "loss": 0.0131, + "step": 118670 + }, + { + "epoch": 10.787129612797672, + "grad_norm": 0.13879691064357758, + "learning_rate": 9.938228518567294e-09, + "loss": 0.0142, + "step": 118680 + }, + { + "epoch": 10.788038538447555, + "grad_norm": 0.13290034234523773, + "learning_rate": 9.788349104725703e-09, + "loss": 0.0156, + "step": 118690 + }, + { + "epoch": 10.788947464097436, + "grad_norm": 0.13057780265808105, + "learning_rate": 9.639608098533925e-09, + "loss": 0.0141, + "step": 118700 + }, + { + "epoch": 10.78985638974732, + "grad_norm": 0.14648118615150452, + "learning_rate": 9.49200551128626e-09, + "loss": 0.0162, + "step": 118710 + }, + { + "epoch": 10.7907653153972, + "grad_norm": 0.16569465398788452, + "learning_rate": 9.345541354190413e-09, + "loss": 0.0159, + "step": 118720 + }, + { + "epoch": 10.791674241047083, + "grad_norm": 0.1572066843509674, + "learning_rate": 9.200215638372477e-09, + "loss": 0.0169, + "step": 118730 + }, + { + "epoch": 10.792583166696964, + "grad_norm": 0.09375607967376709, + "learning_rate": 9.056028374866965e-09, + "loss": 0.0154, + "step": 118740 + }, + { + "epoch": 10.793492092346845, + "grad_norm": 0.13765814900398254, + "learning_rate": 8.912979574625113e-09, + "loss": 0.0146, + "step": 118750 + }, + { + "epoch": 10.794401017996728, + "grad_norm": 0.177315816283226, + "learning_rate": 8.771069248508234e-09, + "loss": 0.0165, + "step": 118760 + }, + { + "epoch": 10.79530994364661, + "grad_norm": 0.21588468551635742, + "learning_rate": 8.630297407297704e-09, + "loss": 0.0181, + "step": 118770 + }, + { + "epoch": 10.796218869296492, + "grad_norm": 0.14581574499607086, + "learning_rate": 8.490664061679977e-09, + "loss": 0.0168, + "step": 118780 + }, + { + "epoch": 10.797127794946373, + "grad_norm": 0.12350354343652725, + "learning_rate": 8.352169222259897e-09, + "loss": 0.014, + "step": 118790 + }, + { + "epoch": 10.798036720596254, + "grad_norm": 0.12523217499256134, + "learning_rate": 8.214812899559054e-09, + "loss": 0.0146, + "step": 118800 + }, + { + "epoch": 10.798945646246137, + "grad_norm": 0.07638468593358994, + "learning_rate": 8.078595104004106e-09, + "loss": 0.0147, + "step": 118810 + }, + { + "epoch": 10.799854571896018, + "grad_norm": 0.09242305159568787, + "learning_rate": 7.943515845941773e-09, + "loss": 0.0174, + "step": 118820 + }, + { + "epoch": 10.800763497545901, + "grad_norm": 0.283251017332077, + "learning_rate": 7.80957513563052e-09, + "loss": 0.0156, + "step": 118830 + }, + { + "epoch": 10.801672423195782, + "grad_norm": 0.1317639797925949, + "learning_rate": 7.676772983243874e-09, + "loss": 0.0167, + "step": 118840 + }, + { + "epoch": 10.802581348845665, + "grad_norm": 0.13859498500823975, + "learning_rate": 7.545109398863769e-09, + "loss": 0.0144, + "step": 118850 + }, + { + "epoch": 10.803490274495546, + "grad_norm": 0.11470242589712143, + "learning_rate": 7.414584392490542e-09, + "loss": 0.0135, + "step": 118860 + }, + { + "epoch": 10.804399200145427, + "grad_norm": 0.0984066054224968, + "learning_rate": 7.285197974039592e-09, + "loss": 0.0137, + "step": 118870 + }, + { + "epoch": 10.80530812579531, + "grad_norm": 0.1387663632631302, + "learning_rate": 7.156950153333064e-09, + "loss": 0.0159, + "step": 118880 + }, + { + "epoch": 10.806217051445191, + "grad_norm": 0.16234391927719116, + "learning_rate": 7.029840940111498e-09, + "loss": 0.0145, + "step": 118890 + }, + { + "epoch": 10.807125977095074, + "grad_norm": 0.12270663678646088, + "learning_rate": 6.90387034402884e-09, + "loss": 0.0148, + "step": 118900 + }, + { + "epoch": 10.808034902744955, + "grad_norm": 0.1705525815486908, + "learning_rate": 6.779038374652435e-09, + "loss": 0.0157, + "step": 118910 + }, + { + "epoch": 10.808943828394836, + "grad_norm": 0.18018457293510437, + "learning_rate": 6.655345041459704e-09, + "loss": 0.0163, + "step": 118920 + }, + { + "epoch": 10.80985275404472, + "grad_norm": 0.1169220507144928, + "learning_rate": 6.5327903538464625e-09, + "loss": 0.0153, + "step": 118930 + }, + { + "epoch": 10.8107616796946, + "grad_norm": 0.14491626620292664, + "learning_rate": 6.411374321120267e-09, + "loss": 0.0159, + "step": 118940 + }, + { + "epoch": 10.811670605344483, + "grad_norm": 0.09560241550207138, + "learning_rate": 6.2910969525020735e-09, + "loss": 0.0147, + "step": 118950 + }, + { + "epoch": 10.812579530994364, + "grad_norm": 0.09364883601665497, + "learning_rate": 6.1719582571229115e-09, + "loss": 0.0138, + "step": 118960 + }, + { + "epoch": 10.813488456644247, + "grad_norm": 0.11985420435667038, + "learning_rate": 6.0539582440338745e-09, + "loss": 0.0151, + "step": 118970 + }, + { + "epoch": 10.814397382294128, + "grad_norm": 0.1563190221786499, + "learning_rate": 5.937096922196128e-09, + "loss": 0.0174, + "step": 118980 + }, + { + "epoch": 10.81530630794401, + "grad_norm": 0.0999121144413948, + "learning_rate": 5.821374300482574e-09, + "loss": 0.0142, + "step": 118990 + }, + { + "epoch": 10.816215233593892, + "grad_norm": 0.25255531072616577, + "learning_rate": 5.706790387682848e-09, + "loss": 0.0146, + "step": 119000 + }, + { + "epoch": 10.817124159243773, + "grad_norm": 0.1459117829799652, + "learning_rate": 5.593345192498322e-09, + "loss": 0.0157, + "step": 119010 + }, + { + "epoch": 10.818033084893656, + "grad_norm": 0.22566264867782593, + "learning_rate": 5.481038723545439e-09, + "loss": 0.0141, + "step": 119020 + }, + { + "epoch": 10.818942010543537, + "grad_norm": 0.2038138508796692, + "learning_rate": 5.369870989352377e-09, + "loss": 0.0148, + "step": 119030 + }, + { + "epoch": 10.819850936193419, + "grad_norm": 0.16371239721775055, + "learning_rate": 5.259841998362381e-09, + "loss": 0.0138, + "step": 119040 + }, + { + "epoch": 10.820759861843301, + "grad_norm": 0.12477941811084747, + "learning_rate": 5.150951758928768e-09, + "loss": 0.0144, + "step": 119050 + }, + { + "epoch": 10.821668787493183, + "grad_norm": 0.1656823456287384, + "learning_rate": 5.043200279324922e-09, + "loss": 0.0146, + "step": 119060 + }, + { + "epoch": 10.822577713143065, + "grad_norm": 0.15750186145305634, + "learning_rate": 4.936587567730965e-09, + "loss": 0.0151, + "step": 119070 + }, + { + "epoch": 10.823486638792946, + "grad_norm": 0.10934082418680191, + "learning_rate": 4.831113632243755e-09, + "loss": 0.0144, + "step": 119080 + }, + { + "epoch": 10.82439556444283, + "grad_norm": 0.10666850954294205, + "learning_rate": 4.726778480875216e-09, + "loss": 0.0161, + "step": 119090 + }, + { + "epoch": 10.82530449009271, + "grad_norm": 0.1453956961631775, + "learning_rate": 4.6235821215473425e-09, + "loss": 0.0168, + "step": 119100 + }, + { + "epoch": 10.826213415742592, + "grad_norm": 0.1129472553730011, + "learning_rate": 4.521524562097201e-09, + "loss": 0.0141, + "step": 119110 + }, + { + "epoch": 10.827122341392474, + "grad_norm": 0.1395692527294159, + "learning_rate": 4.420605810275258e-09, + "loss": 0.014, + "step": 119120 + }, + { + "epoch": 10.828031267042356, + "grad_norm": 0.09927228093147278, + "learning_rate": 4.320825873747048e-09, + "loss": 0.017, + "step": 119130 + }, + { + "epoch": 10.828940192692238, + "grad_norm": 0.10919781029224396, + "learning_rate": 4.222184760088177e-09, + "loss": 0.0178, + "step": 119140 + }, + { + "epoch": 10.82984911834212, + "grad_norm": 0.11208412051200867, + "learning_rate": 4.124682476792652e-09, + "loss": 0.0159, + "step": 119150 + }, + { + "epoch": 10.830758043992002, + "grad_norm": 0.14614376425743103, + "learning_rate": 4.028319031261218e-09, + "loss": 0.0143, + "step": 119160 + }, + { + "epoch": 10.831666969641883, + "grad_norm": 0.15783600509166718, + "learning_rate": 3.933094430816353e-09, + "loss": 0.0151, + "step": 119170 + }, + { + "epoch": 10.832575895291765, + "grad_norm": 0.13652391731739044, + "learning_rate": 3.8390086826856076e-09, + "loss": 0.0172, + "step": 119180 + }, + { + "epoch": 10.833484820941647, + "grad_norm": 0.15412983298301697, + "learning_rate": 3.746061794016597e-09, + "loss": 0.0165, + "step": 119190 + }, + { + "epoch": 10.834393746591529, + "grad_norm": 0.14792321622371674, + "learning_rate": 3.654253771868676e-09, + "loss": 0.016, + "step": 119200 + }, + { + "epoch": 10.835302672241411, + "grad_norm": 0.10714013874530792, + "learning_rate": 3.563584623211269e-09, + "loss": 0.0159, + "step": 119210 + }, + { + "epoch": 10.836211597891293, + "grad_norm": 0.27099645137786865, + "learning_rate": 3.474054354932199e-09, + "loss": 0.0135, + "step": 119220 + }, + { + "epoch": 10.837120523541174, + "grad_norm": 0.09451621770858765, + "learning_rate": 3.3856629738310273e-09, + "loss": 0.0153, + "step": 119230 + }, + { + "epoch": 10.838029449191056, + "grad_norm": 0.20360639691352844, + "learning_rate": 3.298410486619052e-09, + "loss": 0.0156, + "step": 119240 + }, + { + "epoch": 10.838938374840938, + "grad_norm": 0.10495386272668839, + "learning_rate": 3.212296899924305e-09, + "loss": 0.0144, + "step": 119250 + }, + { + "epoch": 10.83984730049082, + "grad_norm": 0.11185069382190704, + "learning_rate": 3.127322220283224e-09, + "loss": 0.0169, + "step": 119260 + }, + { + "epoch": 10.840756226140702, + "grad_norm": 0.1237892135977745, + "learning_rate": 3.043486454152311e-09, + "loss": 0.0163, + "step": 119270 + }, + { + "epoch": 10.841665151790583, + "grad_norm": 0.1300337016582489, + "learning_rate": 2.96078960789814e-09, + "loss": 0.0153, + "step": 119280 + }, + { + "epoch": 10.842574077440466, + "grad_norm": 0.18933990597724915, + "learning_rate": 2.879231687799022e-09, + "loss": 0.0163, + "step": 119290 + }, + { + "epoch": 10.843483003090347, + "grad_norm": 0.13044436275959015, + "learning_rate": 2.798812700050002e-09, + "loss": 0.0165, + "step": 119300 + }, + { + "epoch": 10.84439192874023, + "grad_norm": 0.10959141701459885, + "learning_rate": 2.7195326507595264e-09, + "loss": 0.0148, + "step": 119310 + }, + { + "epoch": 10.84530085439011, + "grad_norm": 0.11992815881967545, + "learning_rate": 2.6413915459461146e-09, + "loss": 0.0141, + "step": 119320 + }, + { + "epoch": 10.846209780039993, + "grad_norm": 0.11396076530218124, + "learning_rate": 2.5643893915450187e-09, + "loss": 0.0159, + "step": 119330 + }, + { + "epoch": 10.847118705689875, + "grad_norm": 0.14800572395324707, + "learning_rate": 2.4885261934032287e-09, + "loss": 0.0154, + "step": 119340 + }, + { + "epoch": 10.848027631339756, + "grad_norm": 0.12188857793807983, + "learning_rate": 2.413801957284467e-09, + "loss": 0.0132, + "step": 119350 + }, + { + "epoch": 10.848936556989639, + "grad_norm": 0.14252059161663055, + "learning_rate": 2.3402166888608635e-09, + "loss": 0.0159, + "step": 119360 + }, + { + "epoch": 10.84984548263952, + "grad_norm": 0.1193389743566513, + "learning_rate": 2.2677703937229453e-09, + "loss": 0.016, + "step": 119370 + }, + { + "epoch": 10.850754408289403, + "grad_norm": 0.11633165925741196, + "learning_rate": 2.1964630773713136e-09, + "loss": 0.018, + "step": 119380 + }, + { + "epoch": 10.851663333939284, + "grad_norm": 0.21071641147136688, + "learning_rate": 2.1262947452199697e-09, + "loss": 0.0159, + "step": 119390 + }, + { + "epoch": 10.852572259589166, + "grad_norm": 0.12620195746421814, + "learning_rate": 2.057265402601316e-09, + "loss": 0.0161, + "step": 119400 + }, + { + "epoch": 10.853481185239048, + "grad_norm": 0.1386329084634781, + "learning_rate": 1.989375054754494e-09, + "loss": 0.0145, + "step": 119410 + }, + { + "epoch": 10.854390110888929, + "grad_norm": 0.1636934131383896, + "learning_rate": 1.92262370683538e-09, + "loss": 0.0179, + "step": 119420 + }, + { + "epoch": 10.855299036538812, + "grad_norm": 0.12873812019824982, + "learning_rate": 1.8570113639149177e-09, + "loss": 0.0145, + "step": 119430 + }, + { + "epoch": 10.856207962188693, + "grad_norm": 0.2061549872159958, + "learning_rate": 1.7925380309757878e-09, + "loss": 0.0142, + "step": 119440 + }, + { + "epoch": 10.857116887838576, + "grad_norm": 0.10992180556058884, + "learning_rate": 1.7292037129140737e-09, + "loss": 0.0165, + "step": 119450 + }, + { + "epoch": 10.858025813488457, + "grad_norm": 0.14655892550945282, + "learning_rate": 1.6670084145392617e-09, + "loss": 0.0157, + "step": 119460 + }, + { + "epoch": 10.858934739138338, + "grad_norm": 0.14699913561344147, + "learning_rate": 1.6059521405742406e-09, + "loss": 0.0165, + "step": 119470 + }, + { + "epoch": 10.85984366478822, + "grad_norm": 0.11298591643571854, + "learning_rate": 1.546034895656967e-09, + "loss": 0.0156, + "step": 119480 + }, + { + "epoch": 10.860752590438102, + "grad_norm": 0.1015361025929451, + "learning_rate": 1.487256684337135e-09, + "loss": 0.0144, + "step": 119490 + }, + { + "epoch": 10.861661516087985, + "grad_norm": 0.16271591186523438, + "learning_rate": 1.4296175110778409e-09, + "loss": 0.0161, + "step": 119500 + }, + { + "epoch": 10.862570441737866, + "grad_norm": 0.147472083568573, + "learning_rate": 1.3731173802572495e-09, + "loss": 0.0156, + "step": 119510 + }, + { + "epoch": 10.863479367387747, + "grad_norm": 0.1664656698703766, + "learning_rate": 1.3177562961669276e-09, + "loss": 0.0151, + "step": 119520 + }, + { + "epoch": 10.86438829303763, + "grad_norm": 0.21658477187156677, + "learning_rate": 1.2635342630085144e-09, + "loss": 0.0138, + "step": 119530 + }, + { + "epoch": 10.86529721868751, + "grad_norm": 0.09075450152158737, + "learning_rate": 1.210451284903713e-09, + "loss": 0.0171, + "step": 119540 + }, + { + "epoch": 10.866206144337394, + "grad_norm": 0.12158998101949692, + "learning_rate": 1.1585073658809675e-09, + "loss": 0.0152, + "step": 119550 + }, + { + "epoch": 10.867115069987275, + "grad_norm": 0.13746514916419983, + "learning_rate": 1.1077025098854554e-09, + "loss": 0.0144, + "step": 119560 + }, + { + "epoch": 10.868023995637158, + "grad_norm": 0.162228062748909, + "learning_rate": 1.058036720775757e-09, + "loss": 0.0168, + "step": 119570 + }, + { + "epoch": 10.868932921287039, + "grad_norm": 0.19419138133525848, + "learning_rate": 1.00951000232552e-09, + "loss": 0.0149, + "step": 119580 + }, + { + "epoch": 10.86984184693692, + "grad_norm": 0.12627479434013367, + "learning_rate": 9.621223582167993e-10, + "loss": 0.0157, + "step": 119590 + }, + { + "epoch": 10.870750772586803, + "grad_norm": 0.10194044560194016, + "learning_rate": 9.158737920500481e-10, + "loss": 0.0133, + "step": 119600 + }, + { + "epoch": 10.871659698236684, + "grad_norm": 0.11894699931144714, + "learning_rate": 8.707643073357918e-10, + "loss": 0.0179, + "step": 119610 + }, + { + "epoch": 10.872568623886567, + "grad_norm": 0.13653908669948578, + "learning_rate": 8.26793907502954e-10, + "loss": 0.0176, + "step": 119620 + }, + { + "epoch": 10.873477549536448, + "grad_norm": 0.1275501400232315, + "learning_rate": 7.839625958905306e-10, + "loss": 0.0142, + "step": 119630 + }, + { + "epoch": 10.87438647518633, + "grad_norm": 0.14450962841510773, + "learning_rate": 7.422703757475891e-10, + "loss": 0.0141, + "step": 119640 + }, + { + "epoch": 10.875295400836212, + "grad_norm": 0.14513729512691498, + "learning_rate": 7.017172502432612e-10, + "loss": 0.0147, + "step": 119650 + }, + { + "epoch": 10.876204326486093, + "grad_norm": 0.11480561643838882, + "learning_rate": 6.623032224584158e-10, + "loss": 0.0147, + "step": 119660 + }, + { + "epoch": 10.877113252135976, + "grad_norm": 0.1219574585556984, + "learning_rate": 6.240282953823284e-10, + "loss": 0.0182, + "step": 119670 + }, + { + "epoch": 10.878022177785857, + "grad_norm": 0.169297456741333, + "learning_rate": 5.868924719243385e-10, + "loss": 0.017, + "step": 119680 + }, + { + "epoch": 10.87893110343574, + "grad_norm": 0.0894232988357544, + "learning_rate": 5.508957549055227e-10, + "loss": 0.0145, + "step": 119690 + }, + { + "epoch": 10.87984002908562, + "grad_norm": 0.13386619091033936, + "learning_rate": 5.16038147058695e-10, + "loss": 0.0144, + "step": 119700 + }, + { + "epoch": 10.880748954735502, + "grad_norm": 0.125949427485466, + "learning_rate": 4.82319651030072e-10, + "loss": 0.0159, + "step": 119710 + }, + { + "epoch": 10.881657880385385, + "grad_norm": 0.1125035211443901, + "learning_rate": 4.4974026938093824e-10, + "loss": 0.0168, + "step": 119720 + }, + { + "epoch": 10.882566806035266, + "grad_norm": 0.1671532243490219, + "learning_rate": 4.183000045859808e-10, + "loss": 0.0155, + "step": 119730 + }, + { + "epoch": 10.883475731685149, + "grad_norm": 0.14129582047462463, + "learning_rate": 3.8799885903162415e-10, + "loss": 0.017, + "step": 119740 + }, + { + "epoch": 10.88438465733503, + "grad_norm": 0.10958769172430038, + "learning_rate": 3.5883683502102583e-10, + "loss": 0.016, + "step": 119750 + }, + { + "epoch": 10.885293582984911, + "grad_norm": 0.13929007947444916, + "learning_rate": 3.308139347674155e-10, + "loss": 0.0154, + "step": 119760 + }, + { + "epoch": 10.886202508634794, + "grad_norm": 0.16126461327075958, + "learning_rate": 3.039301603990907e-10, + "loss": 0.0163, + "step": 119770 + }, + { + "epoch": 10.887111434284675, + "grad_norm": 0.11960944533348083, + "learning_rate": 2.7818551395775163e-10, + "loss": 0.0139, + "step": 119780 + }, + { + "epoch": 10.888020359934558, + "grad_norm": 0.1546565443277359, + "learning_rate": 2.535799973985009e-10, + "loss": 0.0137, + "step": 119790 + }, + { + "epoch": 10.888929285584439, + "grad_norm": 0.14052769541740417, + "learning_rate": 2.3011361258984397e-10, + "loss": 0.0144, + "step": 119800 + }, + { + "epoch": 10.889838211234322, + "grad_norm": 0.0875767171382904, + "learning_rate": 2.077863613153541e-10, + "loss": 0.0154, + "step": 119810 + }, + { + "epoch": 10.890747136884203, + "grad_norm": 0.1376085877418518, + "learning_rate": 1.8659824526867653e-10, + "loss": 0.0146, + "step": 119820 + }, + { + "epoch": 10.891656062534084, + "grad_norm": 0.16995321214199066, + "learning_rate": 1.6654926606018973e-10, + "loss": 0.0126, + "step": 119830 + }, + { + "epoch": 10.892564988183967, + "grad_norm": 0.14020726084709167, + "learning_rate": 1.4763942521200946e-10, + "loss": 0.0158, + "step": 119840 + }, + { + "epoch": 10.893473913833848, + "grad_norm": 0.1792147010564804, + "learning_rate": 1.2986872415965413e-10, + "loss": 0.0158, + "step": 119850 + }, + { + "epoch": 10.89438283948373, + "grad_norm": 0.11279813200235367, + "learning_rate": 1.1323716425371e-10, + "loss": 0.0161, + "step": 119860 + }, + { + "epoch": 10.895291765133612, + "grad_norm": 0.1617463380098343, + "learning_rate": 9.77447467565007e-11, + "loss": 0.0142, + "step": 119870 + }, + { + "epoch": 10.896200690783495, + "grad_norm": 0.1999717801809311, + "learning_rate": 8.33914728454177e-11, + "loss": 0.0155, + "step": 119880 + }, + { + "epoch": 10.897109616433376, + "grad_norm": 0.09909720718860626, + "learning_rate": 7.017734360958983e-11, + "loss": 0.0162, + "step": 119890 + }, + { + "epoch": 10.898018542083257, + "grad_norm": 0.19188310205936432, + "learning_rate": 5.810236005154845e-11, + "loss": 0.0146, + "step": 119900 + }, + { + "epoch": 10.89892746773314, + "grad_norm": 0.09167955070734024, + "learning_rate": 4.716652309222358e-11, + "loss": 0.0158, + "step": 119910 + }, + { + "epoch": 10.899836393383021, + "grad_norm": 0.1647862046957016, + "learning_rate": 3.7369833557621137e-11, + "loss": 0.0163, + "step": 119920 + }, + { + "epoch": 10.900745319032904, + "grad_norm": 0.19313175976276398, + "learning_rate": 2.8712292193810995e-11, + "loss": 0.0182, + "step": 119930 + }, + { + "epoch": 10.901654244682785, + "grad_norm": 0.12509065866470337, + "learning_rate": 2.11938996586003e-11, + "loss": 0.0143, + "step": 119940 + }, + { + "epoch": 10.902563170332666, + "grad_norm": 0.11799655109643936, + "learning_rate": 1.4814656521533464e-11, + "loss": 0.0147, + "step": 119950 + }, + { + "epoch": 10.903472095982549, + "grad_norm": 0.14873692393302917, + "learning_rate": 9.57456326888817e-12, + "loss": 0.0146, + "step": 119960 + }, + { + "epoch": 10.90438102163243, + "grad_norm": 0.14650768041610718, + "learning_rate": 5.4736202986793714e-12, + "loss": 0.0142, + "step": 119970 + }, + { + "epoch": 10.905289947282313, + "grad_norm": 0.14756593108177185, + "learning_rate": 2.5118279189939585e-12, + "loss": 0.0143, + "step": 119980 + }, + { + "epoch": 10.906198872932194, + "grad_norm": 0.1653110384941101, + "learning_rate": 6.891863596480974e-13, + "loss": 0.014, + "step": 119990 + }, + { + "epoch": 10.907107798582077, + "grad_norm": 0.09866949170827866, + "learning_rate": 5.69575553388546e-15, + "loss": 0.0158, + "step": 120000 + }, + { + "epoch": 10.907107798582077, + "step": 120000, + "total_flos": 0.0, + "train_loss": 0.008849201625026762, + "train_runtime": 113170.8482, + "train_samples_per_second": 135.724, + "train_steps_per_second": 1.06 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 11, + "save_steps": 20000, + "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": 64, + "trial_name": null, + "trial_params": null +}