diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,70043 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 19.413766322023204, + "eval_steps": 500, + "global_step": 100000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0019416533178001068, + "grad_norm": 0.21218517422676086, + "learning_rate": 1.8e-07, + "loss": 0.9167, + "step": 10 + }, + { + "epoch": 0.0038833066356002135, + "grad_norm": 0.4294266700744629, + "learning_rate": 3.8e-07, + "loss": 0.9866, + "step": 20 + }, + { + "epoch": 0.005824959953400321, + "grad_norm": 0.35965779423713684, + "learning_rate": 5.8e-07, + "loss": 1.0808, + "step": 30 + }, + { + "epoch": 0.007766613271200427, + "grad_norm": 0.3115023672580719, + "learning_rate": 7.8e-07, + "loss": 1.0433, + "step": 40 + }, + { + "epoch": 0.009708266589000534, + "grad_norm": 0.6420189738273621, + "learning_rate": 9.8e-07, + "loss": 1.0237, + "step": 50 + }, + { + "epoch": 0.011649919906800641, + "grad_norm": 0.3880619704723358, + "learning_rate": 1.18e-06, + "loss": 1.0953, + "step": 60 + }, + { + "epoch": 0.013591573224600747, + "grad_norm": 0.4199892580509186, + "learning_rate": 1.3800000000000001e-06, + "loss": 0.9873, + "step": 70 + }, + { + "epoch": 0.015533226542400854, + "grad_norm": 0.29853570461273193, + "learning_rate": 1.5800000000000003e-06, + "loss": 1.0828, + "step": 80 + }, + { + "epoch": 0.01747487986020096, + "grad_norm": 0.31146228313446045, + "learning_rate": 1.7800000000000001e-06, + "loss": 0.9347, + "step": 90 + }, + { + "epoch": 0.01941653317800107, + "grad_norm": 0.26538991928100586, + "learning_rate": 1.98e-06, + "loss": 0.9492, + "step": 100 + }, + { + "epoch": 0.021358186495801176, + "grad_norm": 0.47638407349586487, + "learning_rate": 2.1800000000000003e-06, + "loss": 1.0107, + "step": 110 + }, + { + "epoch": 0.023299839813601283, + "grad_norm": 0.3185844123363495, + "learning_rate": 2.38e-06, + "loss": 0.9913, + "step": 120 + }, + { + "epoch": 0.025241493131401387, + "grad_norm": 0.41236910223960876, + "learning_rate": 2.5800000000000003e-06, + "loss": 0.9782, + "step": 130 + }, + { + "epoch": 0.027183146449201494, + "grad_norm": 0.6183874607086182, + "learning_rate": 2.78e-06, + "loss": 1.0143, + "step": 140 + }, + { + "epoch": 0.0291247997670016, + "grad_norm": 0.6396399140357971, + "learning_rate": 2.9800000000000003e-06, + "loss": 0.9768, + "step": 150 + }, + { + "epoch": 0.031066453084801708, + "grad_norm": 0.32260116934776306, + "learning_rate": 3.1800000000000005e-06, + "loss": 0.9302, + "step": 160 + }, + { + "epoch": 0.033008106402601815, + "grad_norm": 0.4774690568447113, + "learning_rate": 3.38e-06, + "loss": 0.9661, + "step": 170 + }, + { + "epoch": 0.03494975972040192, + "grad_norm": 0.27719658613204956, + "learning_rate": 3.58e-06, + "loss": 0.9309, + "step": 180 + }, + { + "epoch": 0.03689141303820203, + "grad_norm": 0.4081399440765381, + "learning_rate": 3.7800000000000002e-06, + "loss": 0.9399, + "step": 190 + }, + { + "epoch": 0.03883306635600214, + "grad_norm": 0.8970211148262024, + "learning_rate": 3.98e-06, + "loss": 0.9814, + "step": 200 + }, + { + "epoch": 0.040774719673802244, + "grad_norm": 0.6562154293060303, + "learning_rate": 4.18e-06, + "loss": 0.919, + "step": 210 + }, + { + "epoch": 0.04271637299160235, + "grad_norm": 0.5051197409629822, + "learning_rate": 4.38e-06, + "loss": 0.8638, + "step": 220 + }, + { + "epoch": 0.04465802630940246, + "grad_norm": 0.7981653213500977, + "learning_rate": 4.58e-06, + "loss": 0.7955, + "step": 230 + }, + { + "epoch": 0.046599679627202566, + "grad_norm": 0.8666023015975952, + "learning_rate": 4.780000000000001e-06, + "loss": 0.7865, + "step": 240 + }, + { + "epoch": 0.04854133294500267, + "grad_norm": 1.1471010446548462, + "learning_rate": 4.98e-06, + "loss": 0.767, + "step": 250 + }, + { + "epoch": 0.05048298626280277, + "grad_norm": 0.6557904481887817, + "learning_rate": 5.18e-06, + "loss": 0.7069, + "step": 260 + }, + { + "epoch": 0.05242463958060288, + "grad_norm": 0.769278347492218, + "learning_rate": 5.38e-06, + "loss": 0.6552, + "step": 270 + }, + { + "epoch": 0.05436629289840299, + "grad_norm": 0.7451953291893005, + "learning_rate": 5.580000000000001e-06, + "loss": 0.627, + "step": 280 + }, + { + "epoch": 0.056307946216203095, + "grad_norm": 0.7728846669197083, + "learning_rate": 5.78e-06, + "loss": 0.5696, + "step": 290 + }, + { + "epoch": 0.0582495995340032, + "grad_norm": 0.6863597631454468, + "learning_rate": 5.98e-06, + "loss": 0.5765, + "step": 300 + }, + { + "epoch": 0.06019125285180331, + "grad_norm": 0.6269662976264954, + "learning_rate": 6.18e-06, + "loss": 0.5416, + "step": 310 + }, + { + "epoch": 0.062132906169603416, + "grad_norm": 0.513298511505127, + "learning_rate": 6.38e-06, + "loss": 0.4664, + "step": 320 + }, + { + "epoch": 0.06407455948740352, + "grad_norm": 0.5509614944458008, + "learning_rate": 6.58e-06, + "loss": 0.4477, + "step": 330 + }, + { + "epoch": 0.06601621280520363, + "grad_norm": 0.44493043422698975, + "learning_rate": 6.78e-06, + "loss": 0.3985, + "step": 340 + }, + { + "epoch": 0.06795786612300374, + "grad_norm": 0.41506287455558777, + "learning_rate": 6.98e-06, + "loss": 0.4134, + "step": 350 + }, + { + "epoch": 0.06989951944080385, + "grad_norm": 0.44425109028816223, + "learning_rate": 7.180000000000001e-06, + "loss": 0.3824, + "step": 360 + }, + { + "epoch": 0.07184117275860395, + "grad_norm": 0.48225346207618713, + "learning_rate": 7.3800000000000005e-06, + "loss": 0.3308, + "step": 370 + }, + { + "epoch": 0.07378282607640406, + "grad_norm": 0.3093814551830292, + "learning_rate": 7.580000000000001e-06, + "loss": 0.3351, + "step": 380 + }, + { + "epoch": 0.07572447939420417, + "grad_norm": 0.37395501136779785, + "learning_rate": 7.78e-06, + "loss": 0.3334, + "step": 390 + }, + { + "epoch": 0.07766613271200427, + "grad_norm": 0.27279290556907654, + "learning_rate": 7.98e-06, + "loss": 0.3481, + "step": 400 + }, + { + "epoch": 0.07960778602980438, + "grad_norm": 0.35626593232154846, + "learning_rate": 8.18e-06, + "loss": 0.3053, + "step": 410 + }, + { + "epoch": 0.08154943934760449, + "grad_norm": 0.2599826455116272, + "learning_rate": 8.380000000000001e-06, + "loss": 0.3083, + "step": 420 + }, + { + "epoch": 0.0834910926654046, + "grad_norm": 0.32523685693740845, + "learning_rate": 8.580000000000001e-06, + "loss": 0.2806, + "step": 430 + }, + { + "epoch": 0.0854327459832047, + "grad_norm": 0.29902803897857666, + "learning_rate": 8.78e-06, + "loss": 0.293, + "step": 440 + }, + { + "epoch": 0.08737439930100481, + "grad_norm": 0.2582097351551056, + "learning_rate": 8.98e-06, + "loss": 0.2777, + "step": 450 + }, + { + "epoch": 0.08931605261880492, + "grad_norm": 0.2508487403392792, + "learning_rate": 9.180000000000002e-06, + "loss": 0.2602, + "step": 460 + }, + { + "epoch": 0.09125770593660502, + "grad_norm": 0.27472224831581116, + "learning_rate": 9.38e-06, + "loss": 0.2528, + "step": 470 + }, + { + "epoch": 0.09319935925440513, + "grad_norm": 0.3565694987773895, + "learning_rate": 9.58e-06, + "loss": 0.2608, + "step": 480 + }, + { + "epoch": 0.09514101257220524, + "grad_norm": 0.3092890977859497, + "learning_rate": 9.78e-06, + "loss": 0.2718, + "step": 490 + }, + { + "epoch": 0.09708266589000535, + "grad_norm": 0.3551526665687561, + "learning_rate": 9.980000000000001e-06, + "loss": 0.2532, + "step": 500 + }, + { + "epoch": 0.09902431920780545, + "grad_norm": 0.3190300166606903, + "learning_rate": 1.018e-05, + "loss": 0.2585, + "step": 510 + }, + { + "epoch": 0.10096597252560555, + "grad_norm": 0.26600393652915955, + "learning_rate": 1.038e-05, + "loss": 0.2562, + "step": 520 + }, + { + "epoch": 0.10290762584340565, + "grad_norm": 0.2362293154001236, + "learning_rate": 1.058e-05, + "loss": 0.2546, + "step": 530 + }, + { + "epoch": 0.10484927916120576, + "grad_norm": 0.26931825280189514, + "learning_rate": 1.0780000000000002e-05, + "loss": 0.2686, + "step": 540 + }, + { + "epoch": 0.10679093247900587, + "grad_norm": 0.301399827003479, + "learning_rate": 1.098e-05, + "loss": 0.2609, + "step": 550 + }, + { + "epoch": 0.10873258579680598, + "grad_norm": 0.3594169318675995, + "learning_rate": 1.118e-05, + "loss": 0.251, + "step": 560 + }, + { + "epoch": 0.11067423911460608, + "grad_norm": 0.3587716817855835, + "learning_rate": 1.1380000000000001e-05, + "loss": 0.2579, + "step": 570 + }, + { + "epoch": 0.11261589243240619, + "grad_norm": 0.324817955493927, + "learning_rate": 1.1580000000000001e-05, + "loss": 0.2586, + "step": 580 + }, + { + "epoch": 0.1145575457502063, + "grad_norm": 0.2199435830116272, + "learning_rate": 1.178e-05, + "loss": 0.2203, + "step": 590 + }, + { + "epoch": 0.1164991990680064, + "grad_norm": 0.31064555048942566, + "learning_rate": 1.198e-05, + "loss": 0.2279, + "step": 600 + }, + { + "epoch": 0.11844085238580651, + "grad_norm": 0.17017029225826263, + "learning_rate": 1.2180000000000002e-05, + "loss": 0.2257, + "step": 610 + }, + { + "epoch": 0.12038250570360662, + "grad_norm": 0.19394166767597198, + "learning_rate": 1.238e-05, + "loss": 0.2157, + "step": 620 + }, + { + "epoch": 0.12232415902140673, + "grad_norm": 0.264403760433197, + "learning_rate": 1.258e-05, + "loss": 0.2114, + "step": 630 + }, + { + "epoch": 0.12426581233920683, + "grad_norm": 0.26269105076789856, + "learning_rate": 1.278e-05, + "loss": 0.2275, + "step": 640 + }, + { + "epoch": 0.12620746565700694, + "grad_norm": 0.3848176598548889, + "learning_rate": 1.2980000000000001e-05, + "loss": 0.2368, + "step": 650 + }, + { + "epoch": 0.12814911897480705, + "grad_norm": 0.2568192183971405, + "learning_rate": 1.3180000000000001e-05, + "loss": 0.2214, + "step": 660 + }, + { + "epoch": 0.13009077229260715, + "grad_norm": 0.29272446036338806, + "learning_rate": 1.338e-05, + "loss": 0.2042, + "step": 670 + }, + { + "epoch": 0.13203242561040726, + "grad_norm": 0.43549418449401855, + "learning_rate": 1.358e-05, + "loss": 0.2323, + "step": 680 + }, + { + "epoch": 0.13397407892820737, + "grad_norm": 0.21792808175086975, + "learning_rate": 1.3780000000000002e-05, + "loss": 0.1883, + "step": 690 + }, + { + "epoch": 0.13591573224600748, + "grad_norm": 0.25664177536964417, + "learning_rate": 1.3980000000000002e-05, + "loss": 0.2134, + "step": 700 + }, + { + "epoch": 0.13785738556380758, + "grad_norm": 0.3094502091407776, + "learning_rate": 1.4180000000000001e-05, + "loss": 0.2187, + "step": 710 + }, + { + "epoch": 0.1397990388816077, + "grad_norm": 0.22395721077919006, + "learning_rate": 1.4380000000000001e-05, + "loss": 0.1844, + "step": 720 + }, + { + "epoch": 0.1417406921994078, + "grad_norm": 0.22183483839035034, + "learning_rate": 1.4580000000000003e-05, + "loss": 0.1946, + "step": 730 + }, + { + "epoch": 0.1436823455172079, + "grad_norm": 0.22427183389663696, + "learning_rate": 1.4779999999999999e-05, + "loss": 0.1961, + "step": 740 + }, + { + "epoch": 0.145623998835008, + "grad_norm": 0.20909243822097778, + "learning_rate": 1.4979999999999999e-05, + "loss": 0.18, + "step": 750 + }, + { + "epoch": 0.14756565215280812, + "grad_norm": 0.23044180870056152, + "learning_rate": 1.518e-05, + "loss": 0.2036, + "step": 760 + }, + { + "epoch": 0.14950730547060823, + "grad_norm": 0.1749858558177948, + "learning_rate": 1.538e-05, + "loss": 0.1803, + "step": 770 + }, + { + "epoch": 0.15144895878840833, + "grad_norm": 0.27577605843544006, + "learning_rate": 1.558e-05, + "loss": 0.1891, + "step": 780 + }, + { + "epoch": 0.15339061210620844, + "grad_norm": 0.2773001194000244, + "learning_rate": 1.578e-05, + "loss": 0.1647, + "step": 790 + }, + { + "epoch": 0.15533226542400855, + "grad_norm": 0.34949949383735657, + "learning_rate": 1.598e-05, + "loss": 0.2027, + "step": 800 + }, + { + "epoch": 0.15727391874180865, + "grad_norm": 0.25597110390663147, + "learning_rate": 1.618e-05, + "loss": 0.1652, + "step": 810 + }, + { + "epoch": 0.15921557205960876, + "grad_norm": 0.2991543412208557, + "learning_rate": 1.6380000000000002e-05, + "loss": 0.1918, + "step": 820 + }, + { + "epoch": 0.16115722537740887, + "grad_norm": 0.20948243141174316, + "learning_rate": 1.658e-05, + "loss": 0.1995, + "step": 830 + }, + { + "epoch": 0.16309887869520898, + "grad_norm": 0.18522851169109344, + "learning_rate": 1.6780000000000002e-05, + "loss": 0.1929, + "step": 840 + }, + { + "epoch": 0.16504053201300908, + "grad_norm": 0.34683772921562195, + "learning_rate": 1.698e-05, + "loss": 0.1876, + "step": 850 + }, + { + "epoch": 0.1669821853308092, + "grad_norm": 0.3188510239124298, + "learning_rate": 1.718e-05, + "loss": 0.1952, + "step": 860 + }, + { + "epoch": 0.1689238386486093, + "grad_norm": 0.34033751487731934, + "learning_rate": 1.7380000000000003e-05, + "loss": 0.1832, + "step": 870 + }, + { + "epoch": 0.1708654919664094, + "grad_norm": 0.35232973098754883, + "learning_rate": 1.758e-05, + "loss": 0.187, + "step": 880 + }, + { + "epoch": 0.1728071452842095, + "grad_norm": 0.4431001543998718, + "learning_rate": 1.7780000000000003e-05, + "loss": 0.1782, + "step": 890 + }, + { + "epoch": 0.17474879860200962, + "grad_norm": 0.37776461243629456, + "learning_rate": 1.798e-05, + "loss": 0.1577, + "step": 900 + }, + { + "epoch": 0.17669045191980973, + "grad_norm": 0.3473679721355438, + "learning_rate": 1.818e-05, + "loss": 0.1693, + "step": 910 + }, + { + "epoch": 0.17863210523760983, + "grad_norm": 0.3796978294849396, + "learning_rate": 1.838e-05, + "loss": 0.1857, + "step": 920 + }, + { + "epoch": 0.18057375855540994, + "grad_norm": 0.3505909740924835, + "learning_rate": 1.858e-05, + "loss": 0.1908, + "step": 930 + }, + { + "epoch": 0.18251541187321005, + "grad_norm": 0.29971903562545776, + "learning_rate": 1.878e-05, + "loss": 0.1681, + "step": 940 + }, + { + "epoch": 0.18445706519101016, + "grad_norm": 0.34310805797576904, + "learning_rate": 1.898e-05, + "loss": 0.1736, + "step": 950 + }, + { + "epoch": 0.18639871850881026, + "grad_norm": 0.2994224429130554, + "learning_rate": 1.918e-05, + "loss": 0.1882, + "step": 960 + }, + { + "epoch": 0.18834037182661037, + "grad_norm": 0.38159388303756714, + "learning_rate": 1.938e-05, + "loss": 0.1725, + "step": 970 + }, + { + "epoch": 0.19028202514441048, + "grad_norm": 0.4154254198074341, + "learning_rate": 1.9580000000000002e-05, + "loss": 0.1617, + "step": 980 + }, + { + "epoch": 0.19222367846221058, + "grad_norm": 0.2736925184726715, + "learning_rate": 1.978e-05, + "loss": 0.1585, + "step": 990 + }, + { + "epoch": 0.1941653317800107, + "grad_norm": 0.273093044757843, + "learning_rate": 1.9980000000000002e-05, + "loss": 0.168, + "step": 1000 + }, + { + "epoch": 0.1961069850978108, + "grad_norm": 0.3772861063480377, + "learning_rate": 2.0180000000000003e-05, + "loss": 0.1882, + "step": 1010 + }, + { + "epoch": 0.1980486384156109, + "grad_norm": 0.3541392982006073, + "learning_rate": 2.038e-05, + "loss": 0.158, + "step": 1020 + }, + { + "epoch": 0.199990291733411, + "grad_norm": 0.25439202785491943, + "learning_rate": 2.0580000000000003e-05, + "loss": 0.145, + "step": 1030 + }, + { + "epoch": 0.2019319450512111, + "grad_norm": 0.381062775850296, + "learning_rate": 2.078e-05, + "loss": 0.1673, + "step": 1040 + }, + { + "epoch": 0.2038735983690112, + "grad_norm": 0.29503053426742554, + "learning_rate": 2.098e-05, + "loss": 0.1509, + "step": 1050 + }, + { + "epoch": 0.2058152516868113, + "grad_norm": 0.3622860610485077, + "learning_rate": 2.118e-05, + "loss": 0.1644, + "step": 1060 + }, + { + "epoch": 0.20775690500461141, + "grad_norm": 0.24583682417869568, + "learning_rate": 2.138e-05, + "loss": 0.1589, + "step": 1070 + }, + { + "epoch": 0.20969855832241152, + "grad_norm": 0.3752485513687134, + "learning_rate": 2.158e-05, + "loss": 0.1773, + "step": 1080 + }, + { + "epoch": 0.21164021164021163, + "grad_norm": 0.3834071159362793, + "learning_rate": 2.178e-05, + "loss": 0.1606, + "step": 1090 + }, + { + "epoch": 0.21358186495801174, + "grad_norm": 0.28428196907043457, + "learning_rate": 2.198e-05, + "loss": 0.1495, + "step": 1100 + }, + { + "epoch": 0.21552351827581184, + "grad_norm": 0.5008877515792847, + "learning_rate": 2.218e-05, + "loss": 0.156, + "step": 1110 + }, + { + "epoch": 0.21746517159361195, + "grad_norm": 0.3540477752685547, + "learning_rate": 2.2380000000000003e-05, + "loss": 0.1537, + "step": 1120 + }, + { + "epoch": 0.21940682491141206, + "grad_norm": 0.516070544719696, + "learning_rate": 2.258e-05, + "loss": 0.1582, + "step": 1130 + }, + { + "epoch": 0.22134847822921216, + "grad_norm": 0.3449985980987549, + "learning_rate": 2.2780000000000002e-05, + "loss": 0.1488, + "step": 1140 + }, + { + "epoch": 0.22329013154701227, + "grad_norm": 0.23199807107448578, + "learning_rate": 2.298e-05, + "loss": 0.1714, + "step": 1150 + }, + { + "epoch": 0.22523178486481238, + "grad_norm": 0.29003822803497314, + "learning_rate": 2.318e-05, + "loss": 0.1432, + "step": 1160 + }, + { + "epoch": 0.22717343818261249, + "grad_norm": 0.32432663440704346, + "learning_rate": 2.3380000000000003e-05, + "loss": 0.1535, + "step": 1170 + }, + { + "epoch": 0.2291150915004126, + "grad_norm": 0.3305611312389374, + "learning_rate": 2.358e-05, + "loss": 0.1574, + "step": 1180 + }, + { + "epoch": 0.2310567448182127, + "grad_norm": 0.3399295210838318, + "learning_rate": 2.3780000000000003e-05, + "loss": 0.154, + "step": 1190 + }, + { + "epoch": 0.2329983981360128, + "grad_norm": 0.393882155418396, + "learning_rate": 2.398e-05, + "loss": 0.1527, + "step": 1200 + }, + { + "epoch": 0.23494005145381291, + "grad_norm": 0.5982651710510254, + "learning_rate": 2.418e-05, + "loss": 0.1785, + "step": 1210 + }, + { + "epoch": 0.23688170477161302, + "grad_norm": 0.36082449555397034, + "learning_rate": 2.438e-05, + "loss": 0.1445, + "step": 1220 + }, + { + "epoch": 0.23882335808941313, + "grad_norm": 0.3866993188858032, + "learning_rate": 2.4580000000000002e-05, + "loss": 0.1533, + "step": 1230 + }, + { + "epoch": 0.24076501140721324, + "grad_norm": 0.3766174018383026, + "learning_rate": 2.478e-05, + "loss": 0.1611, + "step": 1240 + }, + { + "epoch": 0.24270666472501334, + "grad_norm": 0.5024001002311707, + "learning_rate": 2.498e-05, + "loss": 0.1451, + "step": 1250 + }, + { + "epoch": 0.24464831804281345, + "grad_norm": 0.4443887174129486, + "learning_rate": 2.5180000000000003e-05, + "loss": 0.1395, + "step": 1260 + }, + { + "epoch": 0.24658997136061356, + "grad_norm": 0.5494886636734009, + "learning_rate": 2.5380000000000004e-05, + "loss": 0.1415, + "step": 1270 + }, + { + "epoch": 0.24853162467841367, + "grad_norm": 0.3426859676837921, + "learning_rate": 2.5580000000000002e-05, + "loss": 0.1624, + "step": 1280 + }, + { + "epoch": 0.2504732779962138, + "grad_norm": 0.4617508053779602, + "learning_rate": 2.5779999999999997e-05, + "loss": 0.1555, + "step": 1290 + }, + { + "epoch": 0.2524149313140139, + "grad_norm": 0.3064330816268921, + "learning_rate": 2.598e-05, + "loss": 0.1436, + "step": 1300 + }, + { + "epoch": 0.254356584631814, + "grad_norm": 0.4819096028804779, + "learning_rate": 2.618e-05, + "loss": 0.1423, + "step": 1310 + }, + { + "epoch": 0.2562982379496141, + "grad_norm": 0.3557845652103424, + "learning_rate": 2.6379999999999998e-05, + "loss": 0.1579, + "step": 1320 + }, + { + "epoch": 0.2582398912674142, + "grad_norm": 0.3883956968784332, + "learning_rate": 2.658e-05, + "loss": 0.1471, + "step": 1330 + }, + { + "epoch": 0.2601815445852143, + "grad_norm": 0.38636600971221924, + "learning_rate": 2.678e-05, + "loss": 0.1567, + "step": 1340 + }, + { + "epoch": 0.2621231979030144, + "grad_norm": 0.2659381330013275, + "learning_rate": 2.698e-05, + "loss": 0.1297, + "step": 1350 + }, + { + "epoch": 0.2640648512208145, + "grad_norm": 0.5242550373077393, + "learning_rate": 2.718e-05, + "loss": 0.1524, + "step": 1360 + }, + { + "epoch": 0.26600650453861463, + "grad_norm": 0.3424222469329834, + "learning_rate": 2.738e-05, + "loss": 0.1499, + "step": 1370 + }, + { + "epoch": 0.26794815785641474, + "grad_norm": 0.399684339761734, + "learning_rate": 2.758e-05, + "loss": 0.1387, + "step": 1380 + }, + { + "epoch": 0.26988981117421484, + "grad_norm": 0.3608193099498749, + "learning_rate": 2.778e-05, + "loss": 0.1405, + "step": 1390 + }, + { + "epoch": 0.27183146449201495, + "grad_norm": 0.7548706531524658, + "learning_rate": 2.798e-05, + "loss": 0.1553, + "step": 1400 + }, + { + "epoch": 0.27377311780981506, + "grad_norm": 0.3982442617416382, + "learning_rate": 2.818e-05, + "loss": 0.153, + "step": 1410 + }, + { + "epoch": 0.27571477112761517, + "grad_norm": 0.329367995262146, + "learning_rate": 2.8380000000000003e-05, + "loss": 0.1398, + "step": 1420 + }, + { + "epoch": 0.2776564244454153, + "grad_norm": 0.4286341369152069, + "learning_rate": 2.858e-05, + "loss": 0.1501, + "step": 1430 + }, + { + "epoch": 0.2795980777632154, + "grad_norm": 0.3483797609806061, + "learning_rate": 2.8780000000000002e-05, + "loss": 0.135, + "step": 1440 + }, + { + "epoch": 0.2815397310810155, + "grad_norm": 0.43133267760276794, + "learning_rate": 2.898e-05, + "loss": 0.1393, + "step": 1450 + }, + { + "epoch": 0.2834813843988156, + "grad_norm": 0.3093181848526001, + "learning_rate": 2.9180000000000002e-05, + "loss": 0.1337, + "step": 1460 + }, + { + "epoch": 0.2854230377166157, + "grad_norm": 0.297151654958725, + "learning_rate": 2.9380000000000003e-05, + "loss": 0.1096, + "step": 1470 + }, + { + "epoch": 0.2873646910344158, + "grad_norm": 0.3105931580066681, + "learning_rate": 2.958e-05, + "loss": 0.1244, + "step": 1480 + }, + { + "epoch": 0.2893063443522159, + "grad_norm": 0.3255787193775177, + "learning_rate": 2.9780000000000003e-05, + "loss": 0.1314, + "step": 1490 + }, + { + "epoch": 0.291247997670016, + "grad_norm": 0.30105817317962646, + "learning_rate": 2.998e-05, + "loss": 0.14, + "step": 1500 + }, + { + "epoch": 0.29318965098781613, + "grad_norm": 0.6042082905769348, + "learning_rate": 3.0180000000000002e-05, + "loss": 0.1294, + "step": 1510 + }, + { + "epoch": 0.29513130430561624, + "grad_norm": 0.604836642742157, + "learning_rate": 3.0380000000000004e-05, + "loss": 0.1452, + "step": 1520 + }, + { + "epoch": 0.29707295762341634, + "grad_norm": 0.6539593935012817, + "learning_rate": 3.058e-05, + "loss": 0.1481, + "step": 1530 + }, + { + "epoch": 0.29901461094121645, + "grad_norm": 0.6290391087532043, + "learning_rate": 3.078e-05, + "loss": 0.121, + "step": 1540 + }, + { + "epoch": 0.30095626425901656, + "grad_norm": 0.4327686131000519, + "learning_rate": 3.0980000000000005e-05, + "loss": 0.1312, + "step": 1550 + }, + { + "epoch": 0.30289791757681667, + "grad_norm": 0.4042474329471588, + "learning_rate": 3.118e-05, + "loss": 0.1332, + "step": 1560 + }, + { + "epoch": 0.3048395708946168, + "grad_norm": 0.424498587846756, + "learning_rate": 3.138e-05, + "loss": 0.1382, + "step": 1570 + }, + { + "epoch": 0.3067812242124169, + "grad_norm": 0.270175576210022, + "learning_rate": 3.1580000000000006e-05, + "loss": 0.1391, + "step": 1580 + }, + { + "epoch": 0.308722877530217, + "grad_norm": 0.408345490694046, + "learning_rate": 3.1780000000000004e-05, + "loss": 0.1478, + "step": 1590 + }, + { + "epoch": 0.3106645308480171, + "grad_norm": 0.35831472277641296, + "learning_rate": 3.198e-05, + "loss": 0.1432, + "step": 1600 + }, + { + "epoch": 0.3126061841658172, + "grad_norm": 0.30652421712875366, + "learning_rate": 3.218e-05, + "loss": 0.1431, + "step": 1610 + }, + { + "epoch": 0.3145478374836173, + "grad_norm": 0.36688724160194397, + "learning_rate": 3.238e-05, + "loss": 0.1186, + "step": 1620 + }, + { + "epoch": 0.3164894908014174, + "grad_norm": 0.3252390921115875, + "learning_rate": 3.2579999999999996e-05, + "loss": 0.1228, + "step": 1630 + }, + { + "epoch": 0.3184311441192175, + "grad_norm": 0.4773605763912201, + "learning_rate": 3.278e-05, + "loss": 0.1338, + "step": 1640 + }, + { + "epoch": 0.32037279743701763, + "grad_norm": 0.2803518772125244, + "learning_rate": 3.298e-05, + "loss": 0.1247, + "step": 1650 + }, + { + "epoch": 0.32231445075481774, + "grad_norm": 0.49088436365127563, + "learning_rate": 3.318e-05, + "loss": 0.1228, + "step": 1660 + }, + { + "epoch": 0.32425610407261785, + "grad_norm": 0.39276036620140076, + "learning_rate": 3.338e-05, + "loss": 0.1266, + "step": 1670 + }, + { + "epoch": 0.32619775739041795, + "grad_norm": 0.5256606936454773, + "learning_rate": 3.358e-05, + "loss": 0.128, + "step": 1680 + }, + { + "epoch": 0.32813941070821806, + "grad_norm": 0.37073323130607605, + "learning_rate": 3.378e-05, + "loss": 0.1446, + "step": 1690 + }, + { + "epoch": 0.33008106402601817, + "grad_norm": 0.4407821297645569, + "learning_rate": 3.398e-05, + "loss": 0.1306, + "step": 1700 + }, + { + "epoch": 0.3320227173438183, + "grad_norm": 0.33624300360679626, + "learning_rate": 3.418e-05, + "loss": 0.1243, + "step": 1710 + }, + { + "epoch": 0.3339643706616184, + "grad_norm": 0.6156623959541321, + "learning_rate": 3.438e-05, + "loss": 0.1391, + "step": 1720 + }, + { + "epoch": 0.3359060239794185, + "grad_norm": 0.36948248744010925, + "learning_rate": 3.4580000000000004e-05, + "loss": 0.1193, + "step": 1730 + }, + { + "epoch": 0.3378476772972186, + "grad_norm": 0.3581908345222473, + "learning_rate": 3.478e-05, + "loss": 0.1323, + "step": 1740 + }, + { + "epoch": 0.3397893306150187, + "grad_norm": 0.3734908401966095, + "learning_rate": 3.498e-05, + "loss": 0.1211, + "step": 1750 + }, + { + "epoch": 0.3417309839328188, + "grad_norm": 0.40379348397254944, + "learning_rate": 3.518e-05, + "loss": 0.1305, + "step": 1760 + }, + { + "epoch": 0.3436726372506189, + "grad_norm": 0.3234756588935852, + "learning_rate": 3.5380000000000003e-05, + "loss": 0.1221, + "step": 1770 + }, + { + "epoch": 0.345614290568419, + "grad_norm": 0.29760029911994934, + "learning_rate": 3.558e-05, + "loss": 0.1153, + "step": 1780 + }, + { + "epoch": 0.34755594388621913, + "grad_norm": 0.6381915211677551, + "learning_rate": 3.578e-05, + "loss": 0.1345, + "step": 1790 + }, + { + "epoch": 0.34949759720401924, + "grad_norm": 0.46883684396743774, + "learning_rate": 3.5980000000000004e-05, + "loss": 0.1417, + "step": 1800 + }, + { + "epoch": 0.35143925052181935, + "grad_norm": 0.3787420392036438, + "learning_rate": 3.618e-05, + "loss": 0.1187, + "step": 1810 + }, + { + "epoch": 0.35338090383961945, + "grad_norm": 0.4227631092071533, + "learning_rate": 3.638e-05, + "loss": 0.1198, + "step": 1820 + }, + { + "epoch": 0.35532255715741956, + "grad_norm": 0.4520735740661621, + "learning_rate": 3.6580000000000006e-05, + "loss": 0.1227, + "step": 1830 + }, + { + "epoch": 0.35726421047521967, + "grad_norm": 0.402316153049469, + "learning_rate": 3.6780000000000004e-05, + "loss": 0.1195, + "step": 1840 + }, + { + "epoch": 0.3592058637930198, + "grad_norm": 0.39427292346954346, + "learning_rate": 3.698e-05, + "loss": 0.1313, + "step": 1850 + }, + { + "epoch": 0.3611475171108199, + "grad_norm": 0.6389641761779785, + "learning_rate": 3.7180000000000007e-05, + "loss": 0.1357, + "step": 1860 + }, + { + "epoch": 0.36308917042862, + "grad_norm": 0.5538950562477112, + "learning_rate": 3.7380000000000005e-05, + "loss": 0.1292, + "step": 1870 + }, + { + "epoch": 0.3650308237464201, + "grad_norm": 0.3233535587787628, + "learning_rate": 3.758e-05, + "loss": 0.1309, + "step": 1880 + }, + { + "epoch": 0.3669724770642202, + "grad_norm": 0.4324755072593689, + "learning_rate": 3.778000000000001e-05, + "loss": 0.1545, + "step": 1890 + }, + { + "epoch": 0.3689141303820203, + "grad_norm": 0.48314452171325684, + "learning_rate": 3.7980000000000006e-05, + "loss": 0.1142, + "step": 1900 + }, + { + "epoch": 0.3708557836998204, + "grad_norm": 0.3538840711116791, + "learning_rate": 3.818e-05, + "loss": 0.12, + "step": 1910 + }, + { + "epoch": 0.3727974370176205, + "grad_norm": 0.5219427347183228, + "learning_rate": 3.838e-05, + "loss": 0.1376, + "step": 1920 + }, + { + "epoch": 0.37473909033542063, + "grad_norm": 0.4419742524623871, + "learning_rate": 3.858e-05, + "loss": 0.1402, + "step": 1930 + }, + { + "epoch": 0.37668074365322074, + "grad_norm": 0.4564119875431061, + "learning_rate": 3.878e-05, + "loss": 0.1076, + "step": 1940 + }, + { + "epoch": 0.37862239697102085, + "grad_norm": 0.5966014266014099, + "learning_rate": 3.898e-05, + "loss": 0.1307, + "step": 1950 + }, + { + "epoch": 0.38056405028882095, + "grad_norm": 0.49646493792533875, + "learning_rate": 3.918e-05, + "loss": 0.1316, + "step": 1960 + }, + { + "epoch": 0.38250570360662106, + "grad_norm": 0.42884162068367004, + "learning_rate": 3.938e-05, + "loss": 0.1251, + "step": 1970 + }, + { + "epoch": 0.38444735692442117, + "grad_norm": 0.5742794275283813, + "learning_rate": 3.958e-05, + "loss": 0.1193, + "step": 1980 + }, + { + "epoch": 0.3863890102422213, + "grad_norm": 0.337443470954895, + "learning_rate": 3.978e-05, + "loss": 0.1163, + "step": 1990 + }, + { + "epoch": 0.3883306635600214, + "grad_norm": 0.44600456953048706, + "learning_rate": 3.998e-05, + "loss": 0.116, + "step": 2000 + }, + { + "epoch": 0.3902723168778215, + "grad_norm": 0.24732014536857605, + "learning_rate": 4.018e-05, + "loss": 0.1215, + "step": 2010 + }, + { + "epoch": 0.3922139701956216, + "grad_norm": 0.25291571021080017, + "learning_rate": 4.038e-05, + "loss": 0.1149, + "step": 2020 + }, + { + "epoch": 0.3941556235134217, + "grad_norm": 0.2905019223690033, + "learning_rate": 4.058e-05, + "loss": 0.1259, + "step": 2030 + }, + { + "epoch": 0.3960972768312218, + "grad_norm": 0.455221563577652, + "learning_rate": 4.078e-05, + "loss": 0.1225, + "step": 2040 + }, + { + "epoch": 0.3980389301490219, + "grad_norm": 0.5624095797538757, + "learning_rate": 4.0980000000000004e-05, + "loss": 0.1112, + "step": 2050 + }, + { + "epoch": 0.399980583466822, + "grad_norm": 0.3693099617958069, + "learning_rate": 4.118e-05, + "loss": 0.1198, + "step": 2060 + }, + { + "epoch": 0.40192223678462213, + "grad_norm": 0.5624871253967285, + "learning_rate": 4.138e-05, + "loss": 0.12, + "step": 2070 + }, + { + "epoch": 0.4038638901024222, + "grad_norm": 0.3715735971927643, + "learning_rate": 4.1580000000000005e-05, + "loss": 0.1305, + "step": 2080 + }, + { + "epoch": 0.4058055434202223, + "grad_norm": 0.39136505126953125, + "learning_rate": 4.178e-05, + "loss": 0.1192, + "step": 2090 + }, + { + "epoch": 0.4077471967380224, + "grad_norm": 0.3375052213668823, + "learning_rate": 4.198e-05, + "loss": 0.1049, + "step": 2100 + }, + { + "epoch": 0.4096888500558225, + "grad_norm": 0.45576292276382446, + "learning_rate": 4.2180000000000006e-05, + "loss": 0.1302, + "step": 2110 + }, + { + "epoch": 0.4116305033736226, + "grad_norm": 0.36734798550605774, + "learning_rate": 4.2380000000000004e-05, + "loss": 0.1334, + "step": 2120 + }, + { + "epoch": 0.4135721566914227, + "grad_norm": 0.3638569414615631, + "learning_rate": 4.258e-05, + "loss": 0.1158, + "step": 2130 + }, + { + "epoch": 0.41551381000922283, + "grad_norm": 0.42031002044677734, + "learning_rate": 4.278e-05, + "loss": 0.1195, + "step": 2140 + }, + { + "epoch": 0.41745546332702294, + "grad_norm": 0.3131773769855499, + "learning_rate": 4.2980000000000005e-05, + "loss": 0.1192, + "step": 2150 + }, + { + "epoch": 0.41939711664482304, + "grad_norm": 0.3451506793498993, + "learning_rate": 4.318e-05, + "loss": 0.1229, + "step": 2160 + }, + { + "epoch": 0.42133876996262315, + "grad_norm": 0.4586566090583801, + "learning_rate": 4.338e-05, + "loss": 0.1017, + "step": 2170 + }, + { + "epoch": 0.42328042328042326, + "grad_norm": 0.4611234962940216, + "learning_rate": 4.3580000000000006e-05, + "loss": 0.1026, + "step": 2180 + }, + { + "epoch": 0.42522207659822336, + "grad_norm": 0.5162106156349182, + "learning_rate": 4.3780000000000004e-05, + "loss": 0.1128, + "step": 2190 + }, + { + "epoch": 0.42716372991602347, + "grad_norm": 0.4410744607448578, + "learning_rate": 4.398e-05, + "loss": 0.1171, + "step": 2200 + }, + { + "epoch": 0.4291053832338236, + "grad_norm": 0.4229501783847809, + "learning_rate": 4.418000000000001e-05, + "loss": 0.1342, + "step": 2210 + }, + { + "epoch": 0.4310470365516237, + "grad_norm": 0.7384501695632935, + "learning_rate": 4.438e-05, + "loss": 0.1189, + "step": 2220 + }, + { + "epoch": 0.4329886898694238, + "grad_norm": 0.5036812424659729, + "learning_rate": 4.458e-05, + "loss": 0.1244, + "step": 2230 + }, + { + "epoch": 0.4349303431872239, + "grad_norm": 0.4508838355541229, + "learning_rate": 4.478e-05, + "loss": 0.1117, + "step": 2240 + }, + { + "epoch": 0.436871996505024, + "grad_norm": 0.38880112767219543, + "learning_rate": 4.498e-05, + "loss": 0.1251, + "step": 2250 + }, + { + "epoch": 0.4388136498228241, + "grad_norm": 0.3350354731082916, + "learning_rate": 4.518e-05, + "loss": 0.1384, + "step": 2260 + }, + { + "epoch": 0.4407553031406242, + "grad_norm": 0.5432226657867432, + "learning_rate": 4.538e-05, + "loss": 0.1236, + "step": 2270 + }, + { + "epoch": 0.44269695645842433, + "grad_norm": 0.35084637999534607, + "learning_rate": 4.558e-05, + "loss": 0.1196, + "step": 2280 + }, + { + "epoch": 0.44463860977622444, + "grad_norm": 0.44737955927848816, + "learning_rate": 4.578e-05, + "loss": 0.1142, + "step": 2290 + }, + { + "epoch": 0.44658026309402454, + "grad_norm": 0.6129721403121948, + "learning_rate": 4.5980000000000004e-05, + "loss": 0.1221, + "step": 2300 + }, + { + "epoch": 0.44852191641182465, + "grad_norm": 0.35767707228660583, + "learning_rate": 4.618e-05, + "loss": 0.112, + "step": 2310 + }, + { + "epoch": 0.45046356972962476, + "grad_norm": 0.3986961543560028, + "learning_rate": 4.638e-05, + "loss": 0.1191, + "step": 2320 + }, + { + "epoch": 0.45240522304742486, + "grad_norm": 0.44816187024116516, + "learning_rate": 4.6580000000000005e-05, + "loss": 0.1119, + "step": 2330 + }, + { + "epoch": 0.45434687636522497, + "grad_norm": 0.3941743075847626, + "learning_rate": 4.678e-05, + "loss": 0.1189, + "step": 2340 + }, + { + "epoch": 0.4562885296830251, + "grad_norm": 0.24701175093650818, + "learning_rate": 4.698e-05, + "loss": 0.1293, + "step": 2350 + }, + { + "epoch": 0.4582301830008252, + "grad_norm": 0.3553253412246704, + "learning_rate": 4.718e-05, + "loss": 0.1117, + "step": 2360 + }, + { + "epoch": 0.4601718363186253, + "grad_norm": 0.27699634432792664, + "learning_rate": 4.7380000000000004e-05, + "loss": 0.1142, + "step": 2370 + }, + { + "epoch": 0.4621134896364254, + "grad_norm": 0.28968513011932373, + "learning_rate": 4.758e-05, + "loss": 0.1103, + "step": 2380 + }, + { + "epoch": 0.4640551429542255, + "grad_norm": 0.3162919580936432, + "learning_rate": 4.778e-05, + "loss": 0.1345, + "step": 2390 + }, + { + "epoch": 0.4659967962720256, + "grad_norm": 0.3461274206638336, + "learning_rate": 4.7980000000000005e-05, + "loss": 0.1134, + "step": 2400 + }, + { + "epoch": 0.4679384495898257, + "grad_norm": 0.33303558826446533, + "learning_rate": 4.818e-05, + "loss": 0.1171, + "step": 2410 + }, + { + "epoch": 0.46988010290762583, + "grad_norm": 0.2543747127056122, + "learning_rate": 4.838e-05, + "loss": 0.1063, + "step": 2420 + }, + { + "epoch": 0.47182175622542594, + "grad_norm": 0.4373631179332733, + "learning_rate": 4.8580000000000006e-05, + "loss": 0.1265, + "step": 2430 + }, + { + "epoch": 0.47376340954322604, + "grad_norm": 0.5122209191322327, + "learning_rate": 4.8780000000000004e-05, + "loss": 0.1039, + "step": 2440 + }, + { + "epoch": 0.47570506286102615, + "grad_norm": 0.40580785274505615, + "learning_rate": 4.898e-05, + "loss": 0.1193, + "step": 2450 + }, + { + "epoch": 0.47764671617882626, + "grad_norm": 0.5385798811912537, + "learning_rate": 4.918000000000001e-05, + "loss": 0.1358, + "step": 2460 + }, + { + "epoch": 0.47958836949662637, + "grad_norm": 0.32101917266845703, + "learning_rate": 4.9380000000000005e-05, + "loss": 0.1127, + "step": 2470 + }, + { + "epoch": 0.4815300228144265, + "grad_norm": 0.34603774547576904, + "learning_rate": 4.958e-05, + "loss": 0.1157, + "step": 2480 + }, + { + "epoch": 0.4834716761322266, + "grad_norm": 0.4038693606853485, + "learning_rate": 4.978e-05, + "loss": 0.1084, + "step": 2490 + }, + { + "epoch": 0.4854133294500267, + "grad_norm": 0.36467260122299194, + "learning_rate": 4.9980000000000006e-05, + "loss": 0.125, + "step": 2500 + }, + { + "epoch": 0.4873549827678268, + "grad_norm": 0.27007102966308594, + "learning_rate": 5.0180000000000004e-05, + "loss": 0.1293, + "step": 2510 + }, + { + "epoch": 0.4892966360856269, + "grad_norm": 0.43191587924957275, + "learning_rate": 5.038e-05, + "loss": 0.1071, + "step": 2520 + }, + { + "epoch": 0.491238289403427, + "grad_norm": 0.42346036434173584, + "learning_rate": 5.058000000000001e-05, + "loss": 0.1053, + "step": 2530 + }, + { + "epoch": 0.4931799427212271, + "grad_norm": 0.37822139263153076, + "learning_rate": 5.0780000000000005e-05, + "loss": 0.131, + "step": 2540 + }, + { + "epoch": 0.4951215960390272, + "grad_norm": 0.3773422837257385, + "learning_rate": 5.098e-05, + "loss": 0.1159, + "step": 2550 + }, + { + "epoch": 0.49706324935682733, + "grad_norm": 0.5148995518684387, + "learning_rate": 5.118000000000001e-05, + "loss": 0.1069, + "step": 2560 + }, + { + "epoch": 0.49900490267462744, + "grad_norm": 0.42886555194854736, + "learning_rate": 5.1380000000000006e-05, + "loss": 0.1261, + "step": 2570 + }, + { + "epoch": 0.5009465559924275, + "grad_norm": 0.3292868733406067, + "learning_rate": 5.1580000000000004e-05, + "loss": 0.1181, + "step": 2580 + }, + { + "epoch": 0.5028882093102277, + "grad_norm": 0.30323532223701477, + "learning_rate": 5.178000000000001e-05, + "loss": 0.1042, + "step": 2590 + }, + { + "epoch": 0.5048298626280278, + "grad_norm": 0.3875257670879364, + "learning_rate": 5.198000000000001e-05, + "loss": 0.1244, + "step": 2600 + }, + { + "epoch": 0.5067715159458279, + "grad_norm": 0.3368631601333618, + "learning_rate": 5.2180000000000005e-05, + "loss": 0.1125, + "step": 2610 + }, + { + "epoch": 0.508713169263628, + "grad_norm": 0.38184675574302673, + "learning_rate": 5.238000000000001e-05, + "loss": 0.1188, + "step": 2620 + }, + { + "epoch": 0.5106548225814281, + "grad_norm": 0.5940945148468018, + "learning_rate": 5.258000000000001e-05, + "loss": 0.1181, + "step": 2630 + }, + { + "epoch": 0.5125964758992282, + "grad_norm": 0.39295968413352966, + "learning_rate": 5.2780000000000006e-05, + "loss": 0.1191, + "step": 2640 + }, + { + "epoch": 0.5145381292170284, + "grad_norm": 0.3633505702018738, + "learning_rate": 5.2980000000000004e-05, + "loss": 0.1244, + "step": 2650 + }, + { + "epoch": 0.5164797825348284, + "grad_norm": 0.5199979543685913, + "learning_rate": 5.318000000000001e-05, + "loss": 0.1091, + "step": 2660 + }, + { + "epoch": 0.5184214358526286, + "grad_norm": 0.3843442499637604, + "learning_rate": 5.338000000000001e-05, + "loss": 0.1351, + "step": 2670 + }, + { + "epoch": 0.5203630891704286, + "grad_norm": 0.4536723792552948, + "learning_rate": 5.3580000000000005e-05, + "loss": 0.1112, + "step": 2680 + }, + { + "epoch": 0.5223047424882288, + "grad_norm": 0.43102288246154785, + "learning_rate": 5.378e-05, + "loss": 0.1101, + "step": 2690 + }, + { + "epoch": 0.5242463958060288, + "grad_norm": 0.5252617597579956, + "learning_rate": 5.3979999999999995e-05, + "loss": 0.1194, + "step": 2700 + }, + { + "epoch": 0.526188049123829, + "grad_norm": 0.3620033264160156, + "learning_rate": 5.418e-05, + "loss": 0.1297, + "step": 2710 + }, + { + "epoch": 0.528129702441629, + "grad_norm": 0.34628045558929443, + "learning_rate": 5.438e-05, + "loss": 0.1284, + "step": 2720 + }, + { + "epoch": 0.5300713557594292, + "grad_norm": 0.48661062121391296, + "learning_rate": 5.4579999999999996e-05, + "loss": 0.1176, + "step": 2730 + }, + { + "epoch": 0.5320130090772293, + "grad_norm": 0.547987699508667, + "learning_rate": 5.478e-05, + "loss": 0.1128, + "step": 2740 + }, + { + "epoch": 0.5339546623950294, + "grad_norm": 0.33957749605178833, + "learning_rate": 5.498e-05, + "loss": 0.1083, + "step": 2750 + }, + { + "epoch": 0.5358963157128295, + "grad_norm": 0.4933692514896393, + "learning_rate": 5.518e-05, + "loss": 0.1174, + "step": 2760 + }, + { + "epoch": 0.5378379690306295, + "grad_norm": 0.3567765951156616, + "learning_rate": 5.538e-05, + "loss": 0.1152, + "step": 2770 + }, + { + "epoch": 0.5397796223484297, + "grad_norm": 0.4545426368713379, + "learning_rate": 5.558e-05, + "loss": 0.094, + "step": 2780 + }, + { + "epoch": 0.5417212756662297, + "grad_norm": 0.3717484474182129, + "learning_rate": 5.578e-05, + "loss": 0.1161, + "step": 2790 + }, + { + "epoch": 0.5436629289840299, + "grad_norm": 0.42190054059028625, + "learning_rate": 5.5979999999999996e-05, + "loss": 0.1043, + "step": 2800 + }, + { + "epoch": 0.54560458230183, + "grad_norm": 0.348673939704895, + "learning_rate": 5.618e-05, + "loss": 0.1034, + "step": 2810 + }, + { + "epoch": 0.5475462356196301, + "grad_norm": 0.3480939269065857, + "learning_rate": 5.638e-05, + "loss": 0.1077, + "step": 2820 + }, + { + "epoch": 0.5494878889374302, + "grad_norm": 0.3090517520904541, + "learning_rate": 5.658e-05, + "loss": 0.1192, + "step": 2830 + }, + { + "epoch": 0.5514295422552303, + "grad_norm": 0.38090044260025024, + "learning_rate": 5.678e-05, + "loss": 0.1037, + "step": 2840 + }, + { + "epoch": 0.5533711955730304, + "grad_norm": 0.28798919916152954, + "learning_rate": 5.698e-05, + "loss": 0.1022, + "step": 2850 + }, + { + "epoch": 0.5553128488908305, + "grad_norm": 0.48720329999923706, + "learning_rate": 5.718e-05, + "loss": 0.1088, + "step": 2860 + }, + { + "epoch": 0.5572545022086306, + "grad_norm": 0.978462815284729, + "learning_rate": 5.738e-05, + "loss": 0.106, + "step": 2870 + }, + { + "epoch": 0.5591961555264308, + "grad_norm": 0.34166720509529114, + "learning_rate": 5.758e-05, + "loss": 0.1011, + "step": 2880 + }, + { + "epoch": 0.5611378088442308, + "grad_norm": 0.34180137515068054, + "learning_rate": 5.778e-05, + "loss": 0.12, + "step": 2890 + }, + { + "epoch": 0.563079462162031, + "grad_norm": 0.29501649737358093, + "learning_rate": 5.7980000000000004e-05, + "loss": 0.1167, + "step": 2900 + }, + { + "epoch": 0.565021115479831, + "grad_norm": 0.48099297285079956, + "learning_rate": 5.818e-05, + "loss": 0.1084, + "step": 2910 + }, + { + "epoch": 0.5669627687976312, + "grad_norm": 0.4643601179122925, + "learning_rate": 5.838e-05, + "loss": 0.1125, + "step": 2920 + }, + { + "epoch": 0.5689044221154312, + "grad_norm": 0.48665329813957214, + "learning_rate": 5.858e-05, + "loss": 0.1081, + "step": 2930 + }, + { + "epoch": 0.5708460754332314, + "grad_norm": 0.38237273693084717, + "learning_rate": 5.878e-05, + "loss": 0.1265, + "step": 2940 + }, + { + "epoch": 0.5727877287510315, + "grad_norm": 0.6060514450073242, + "learning_rate": 5.898e-05, + "loss": 0.1084, + "step": 2950 + }, + { + "epoch": 0.5747293820688316, + "grad_norm": 0.4419492185115814, + "learning_rate": 5.918e-05, + "loss": 0.1259, + "step": 2960 + }, + { + "epoch": 0.5766710353866317, + "grad_norm": 0.42942875623703003, + "learning_rate": 5.9380000000000004e-05, + "loss": 0.1035, + "step": 2970 + }, + { + "epoch": 0.5786126887044318, + "grad_norm": 0.33616387844085693, + "learning_rate": 5.958e-05, + "loss": 0.105, + "step": 2980 + }, + { + "epoch": 0.5805543420222319, + "grad_norm": 0.47214198112487793, + "learning_rate": 5.978e-05, + "loss": 0.1082, + "step": 2990 + }, + { + "epoch": 0.582495995340032, + "grad_norm": 0.4012189507484436, + "learning_rate": 5.9980000000000005e-05, + "loss": 0.1211, + "step": 3000 + }, + { + "epoch": 0.5844376486578321, + "grad_norm": 0.3659912049770355, + "learning_rate": 6.018e-05, + "loss": 0.1076, + "step": 3010 + }, + { + "epoch": 0.5863793019756323, + "grad_norm": 0.3279411494731903, + "learning_rate": 6.038e-05, + "loss": 0.1028, + "step": 3020 + }, + { + "epoch": 0.5883209552934323, + "grad_norm": 0.24905721843242645, + "learning_rate": 6.0580000000000006e-05, + "loss": 0.102, + "step": 3030 + }, + { + "epoch": 0.5902626086112325, + "grad_norm": 0.389285147190094, + "learning_rate": 6.0780000000000004e-05, + "loss": 0.1137, + "step": 3040 + }, + { + "epoch": 0.5922042619290325, + "grad_norm": 0.5488048791885376, + "learning_rate": 6.098e-05, + "loss": 0.1269, + "step": 3050 + }, + { + "epoch": 0.5941459152468327, + "grad_norm": 0.3863506019115448, + "learning_rate": 6.118000000000001e-05, + "loss": 0.0994, + "step": 3060 + }, + { + "epoch": 0.5960875685646327, + "grad_norm": 0.37363219261169434, + "learning_rate": 6.138e-05, + "loss": 0.1128, + "step": 3070 + }, + { + "epoch": 0.5980292218824329, + "grad_norm": 0.4092799425125122, + "learning_rate": 6.158e-05, + "loss": 0.1275, + "step": 3080 + }, + { + "epoch": 0.599970875200233, + "grad_norm": 0.3347376883029938, + "learning_rate": 6.178000000000001e-05, + "loss": 0.1086, + "step": 3090 + }, + { + "epoch": 0.6019125285180331, + "grad_norm": 0.43428969383239746, + "learning_rate": 6.198e-05, + "loss": 0.1078, + "step": 3100 + }, + { + "epoch": 0.6038541818358332, + "grad_norm": 0.48623523116111755, + "learning_rate": 6.218e-05, + "loss": 0.1125, + "step": 3110 + }, + { + "epoch": 0.6057958351536333, + "grad_norm": 0.3270756006240845, + "learning_rate": 6.238000000000001e-05, + "loss": 0.1027, + "step": 3120 + }, + { + "epoch": 0.6077374884714334, + "grad_norm": 0.3803662657737732, + "learning_rate": 6.258e-05, + "loss": 0.1115, + "step": 3130 + }, + { + "epoch": 0.6096791417892335, + "grad_norm": 0.3472345173358917, + "learning_rate": 6.278e-05, + "loss": 0.1104, + "step": 3140 + }, + { + "epoch": 0.6116207951070336, + "grad_norm": 0.4020943343639374, + "learning_rate": 6.298000000000001e-05, + "loss": 0.1091, + "step": 3150 + }, + { + "epoch": 0.6135624484248338, + "grad_norm": 0.38898685574531555, + "learning_rate": 6.318e-05, + "loss": 0.1055, + "step": 3160 + }, + { + "epoch": 0.6155041017426338, + "grad_norm": 0.2819019854068756, + "learning_rate": 6.338e-05, + "loss": 0.12, + "step": 3170 + }, + { + "epoch": 0.617445755060434, + "grad_norm": 0.37089774012565613, + "learning_rate": 6.358000000000001e-05, + "loss": 0.1141, + "step": 3180 + }, + { + "epoch": 0.619387408378234, + "grad_norm": 0.5597869753837585, + "learning_rate": 6.378e-05, + "loss": 0.1101, + "step": 3190 + }, + { + "epoch": 0.6213290616960342, + "grad_norm": 0.5341514348983765, + "learning_rate": 6.398000000000001e-05, + "loss": 0.1244, + "step": 3200 + }, + { + "epoch": 0.6232707150138342, + "grad_norm": 0.5046377778053284, + "learning_rate": 6.418000000000001e-05, + "loss": 0.0938, + "step": 3210 + }, + { + "epoch": 0.6252123683316344, + "grad_norm": 0.25269511342048645, + "learning_rate": 6.438e-05, + "loss": 0.1087, + "step": 3220 + }, + { + "epoch": 0.6271540216494345, + "grad_norm": 0.5119882225990295, + "learning_rate": 6.458000000000001e-05, + "loss": 0.1131, + "step": 3230 + }, + { + "epoch": 0.6290956749672346, + "grad_norm": 0.3512127697467804, + "learning_rate": 6.478000000000001e-05, + "loss": 0.1004, + "step": 3240 + }, + { + "epoch": 0.6310373282850347, + "grad_norm": 0.44413095712661743, + "learning_rate": 6.498e-05, + "loss": 0.1036, + "step": 3250 + }, + { + "epoch": 0.6329789816028348, + "grad_norm": 0.4074602723121643, + "learning_rate": 6.518000000000001e-05, + "loss": 0.1097, + "step": 3260 + }, + { + "epoch": 0.6349206349206349, + "grad_norm": 0.3497657775878906, + "learning_rate": 6.538000000000001e-05, + "loss": 0.0997, + "step": 3270 + }, + { + "epoch": 0.636862288238435, + "grad_norm": 0.3762771189212799, + "learning_rate": 6.558e-05, + "loss": 0.1069, + "step": 3280 + }, + { + "epoch": 0.6388039415562351, + "grad_norm": 0.3820244073867798, + "learning_rate": 6.578000000000001e-05, + "loss": 0.1008, + "step": 3290 + }, + { + "epoch": 0.6407455948740353, + "grad_norm": 0.3316919505596161, + "learning_rate": 6.598e-05, + "loss": 0.1169, + "step": 3300 + }, + { + "epoch": 0.6426872481918353, + "grad_norm": 0.31327325105667114, + "learning_rate": 6.618e-05, + "loss": 0.1081, + "step": 3310 + }, + { + "epoch": 0.6446289015096355, + "grad_norm": 0.3468875586986542, + "learning_rate": 6.638e-05, + "loss": 0.1009, + "step": 3320 + }, + { + "epoch": 0.6465705548274355, + "grad_norm": 0.3441944420337677, + "learning_rate": 6.658e-05, + "loss": 0.0987, + "step": 3330 + }, + { + "epoch": 0.6485122081452357, + "grad_norm": 0.2765507996082306, + "learning_rate": 6.678e-05, + "loss": 0.1065, + "step": 3340 + }, + { + "epoch": 0.6504538614630357, + "grad_norm": 0.4345357120037079, + "learning_rate": 6.698e-05, + "loss": 0.1037, + "step": 3350 + }, + { + "epoch": 0.6523955147808359, + "grad_norm": 0.5665486454963684, + "learning_rate": 6.718e-05, + "loss": 0.1034, + "step": 3360 + }, + { + "epoch": 0.654337168098636, + "grad_norm": 0.5142514109611511, + "learning_rate": 6.738e-05, + "loss": 0.1118, + "step": 3370 + }, + { + "epoch": 0.6562788214164361, + "grad_norm": 0.3032837212085724, + "learning_rate": 6.758e-05, + "loss": 0.1032, + "step": 3380 + }, + { + "epoch": 0.6582204747342362, + "grad_norm": 0.4939611554145813, + "learning_rate": 6.778e-05, + "loss": 0.1154, + "step": 3390 + }, + { + "epoch": 0.6601621280520363, + "grad_norm": 0.3742654621601105, + "learning_rate": 6.798e-05, + "loss": 0.0989, + "step": 3400 + }, + { + "epoch": 0.6621037813698364, + "grad_norm": 0.38113874197006226, + "learning_rate": 6.818e-05, + "loss": 0.1045, + "step": 3410 + }, + { + "epoch": 0.6640454346876365, + "grad_norm": 0.4506068527698517, + "learning_rate": 6.838e-05, + "loss": 0.1006, + "step": 3420 + }, + { + "epoch": 0.6659870880054366, + "grad_norm": 0.4376467168331146, + "learning_rate": 6.858e-05, + "loss": 0.1124, + "step": 3430 + }, + { + "epoch": 0.6679287413232368, + "grad_norm": 0.5707558989524841, + "learning_rate": 6.878e-05, + "loss": 0.121, + "step": 3440 + }, + { + "epoch": 0.6698703946410368, + "grad_norm": 0.3640248477458954, + "learning_rate": 6.898e-05, + "loss": 0.107, + "step": 3450 + }, + { + "epoch": 0.671812047958837, + "grad_norm": 0.4440123736858368, + "learning_rate": 6.918e-05, + "loss": 0.1062, + "step": 3460 + }, + { + "epoch": 0.673753701276637, + "grad_norm": 0.5093589425086975, + "learning_rate": 6.938e-05, + "loss": 0.0984, + "step": 3470 + }, + { + "epoch": 0.6756953545944372, + "grad_norm": 0.4809255599975586, + "learning_rate": 6.958e-05, + "loss": 0.1078, + "step": 3480 + }, + { + "epoch": 0.6776370079122372, + "grad_norm": 0.6277799606323242, + "learning_rate": 6.978e-05, + "loss": 0.1065, + "step": 3490 + }, + { + "epoch": 0.6795786612300374, + "grad_norm": 0.41659480333328247, + "learning_rate": 6.998e-05, + "loss": 0.1, + "step": 3500 + }, + { + "epoch": 0.6815203145478375, + "grad_norm": 0.5918940901756287, + "learning_rate": 7.018e-05, + "loss": 0.1123, + "step": 3510 + }, + { + "epoch": 0.6834619678656376, + "grad_norm": 0.43870028853416443, + "learning_rate": 7.038e-05, + "loss": 0.109, + "step": 3520 + }, + { + "epoch": 0.6854036211834377, + "grad_norm": 0.356687992811203, + "learning_rate": 7.058e-05, + "loss": 0.0918, + "step": 3530 + }, + { + "epoch": 0.6873452745012378, + "grad_norm": 0.33280789852142334, + "learning_rate": 7.078e-05, + "loss": 0.1153, + "step": 3540 + }, + { + "epoch": 0.6892869278190379, + "grad_norm": 0.5337461829185486, + "learning_rate": 7.098e-05, + "loss": 0.1031, + "step": 3550 + }, + { + "epoch": 0.691228581136838, + "grad_norm": 0.3312370777130127, + "learning_rate": 7.118e-05, + "loss": 0.0944, + "step": 3560 + }, + { + "epoch": 0.6931702344546381, + "grad_norm": 0.3104833960533142, + "learning_rate": 7.138e-05, + "loss": 0.109, + "step": 3570 + }, + { + "epoch": 0.6951118877724383, + "grad_norm": 0.46791139245033264, + "learning_rate": 7.158e-05, + "loss": 0.1138, + "step": 3580 + }, + { + "epoch": 0.6970535410902383, + "grad_norm": 0.350525438785553, + "learning_rate": 7.178000000000001e-05, + "loss": 0.0867, + "step": 3590 + }, + { + "epoch": 0.6989951944080385, + "grad_norm": 0.321768194437027, + "learning_rate": 7.198e-05, + "loss": 0.1166, + "step": 3600 + }, + { + "epoch": 0.7009368477258385, + "grad_norm": 0.4125913083553314, + "learning_rate": 7.218e-05, + "loss": 0.0985, + "step": 3610 + }, + { + "epoch": 0.7028785010436387, + "grad_norm": 0.3677160441875458, + "learning_rate": 7.238000000000001e-05, + "loss": 0.1028, + "step": 3620 + }, + { + "epoch": 0.7048201543614387, + "grad_norm": 0.37941667437553406, + "learning_rate": 7.258e-05, + "loss": 0.1069, + "step": 3630 + }, + { + "epoch": 0.7067618076792389, + "grad_norm": 0.33407992124557495, + "learning_rate": 7.278e-05, + "loss": 0.0976, + "step": 3640 + }, + { + "epoch": 0.708703460997039, + "grad_norm": 0.5812451839447021, + "learning_rate": 7.298000000000001e-05, + "loss": 0.1092, + "step": 3650 + }, + { + "epoch": 0.7106451143148391, + "grad_norm": 0.4467465281486511, + "learning_rate": 7.318e-05, + "loss": 0.0908, + "step": 3660 + }, + { + "epoch": 0.7125867676326392, + "grad_norm": 0.346441388130188, + "learning_rate": 7.338e-05, + "loss": 0.0954, + "step": 3670 + }, + { + "epoch": 0.7145284209504393, + "grad_norm": 0.3157282769680023, + "learning_rate": 7.358000000000001e-05, + "loss": 0.0998, + "step": 3680 + }, + { + "epoch": 0.7164700742682394, + "grad_norm": 0.302165687084198, + "learning_rate": 7.378e-05, + "loss": 0.0897, + "step": 3690 + }, + { + "epoch": 0.7184117275860396, + "grad_norm": 0.3712438642978668, + "learning_rate": 7.398e-05, + "loss": 0.0836, + "step": 3700 + }, + { + "epoch": 0.7203533809038396, + "grad_norm": 0.30088311433792114, + "learning_rate": 7.418000000000001e-05, + "loss": 0.0921, + "step": 3710 + }, + { + "epoch": 0.7222950342216398, + "grad_norm": 0.48612216114997864, + "learning_rate": 7.438e-05, + "loss": 0.1003, + "step": 3720 + }, + { + "epoch": 0.7242366875394398, + "grad_norm": 0.43690434098243713, + "learning_rate": 7.458000000000001e-05, + "loss": 0.1054, + "step": 3730 + }, + { + "epoch": 0.72617834085724, + "grad_norm": 0.456929475069046, + "learning_rate": 7.478e-05, + "loss": 0.0969, + "step": 3740 + }, + { + "epoch": 0.72811999417504, + "grad_norm": 0.3730933666229248, + "learning_rate": 7.498e-05, + "loss": 0.0875, + "step": 3750 + }, + { + "epoch": 0.7300616474928402, + "grad_norm": 0.28361719846725464, + "learning_rate": 7.518000000000001e-05, + "loss": 0.0828, + "step": 3760 + }, + { + "epoch": 0.7320033008106402, + "grad_norm": 0.3812446594238281, + "learning_rate": 7.538e-05, + "loss": 0.0797, + "step": 3770 + }, + { + "epoch": 0.7339449541284404, + "grad_norm": 0.42226195335388184, + "learning_rate": 7.558e-05, + "loss": 0.0969, + "step": 3780 + }, + { + "epoch": 0.7358866074462405, + "grad_norm": 0.2924696207046509, + "learning_rate": 7.578000000000001e-05, + "loss": 0.0944, + "step": 3790 + }, + { + "epoch": 0.7378282607640406, + "grad_norm": 0.540582001209259, + "learning_rate": 7.598e-05, + "loss": 0.1023, + "step": 3800 + }, + { + "epoch": 0.7397699140818407, + "grad_norm": 0.37405362725257874, + "learning_rate": 7.618e-05, + "loss": 0.0887, + "step": 3810 + }, + { + "epoch": 0.7417115673996408, + "grad_norm": 0.3975103795528412, + "learning_rate": 7.638000000000001e-05, + "loss": 0.0939, + "step": 3820 + }, + { + "epoch": 0.7436532207174409, + "grad_norm": 0.45785877108573914, + "learning_rate": 7.658e-05, + "loss": 0.0891, + "step": 3830 + }, + { + "epoch": 0.745594874035241, + "grad_norm": 0.27734100818634033, + "learning_rate": 7.678000000000001e-05, + "loss": 0.0909, + "step": 3840 + }, + { + "epoch": 0.7475365273530411, + "grad_norm": 0.6380115747451782, + "learning_rate": 7.698000000000001e-05, + "loss": 0.0979, + "step": 3850 + }, + { + "epoch": 0.7494781806708413, + "grad_norm": 0.5374745726585388, + "learning_rate": 7.718e-05, + "loss": 0.1046, + "step": 3860 + }, + { + "epoch": 0.7514198339886413, + "grad_norm": 0.394864946603775, + "learning_rate": 7.738000000000001e-05, + "loss": 0.0941, + "step": 3870 + }, + { + "epoch": 0.7533614873064415, + "grad_norm": 0.4413292706012726, + "learning_rate": 7.758000000000001e-05, + "loss": 0.0951, + "step": 3880 + }, + { + "epoch": 0.7553031406242415, + "grad_norm": 0.37433019280433655, + "learning_rate": 7.778e-05, + "loss": 0.0923, + "step": 3890 + }, + { + "epoch": 0.7572447939420417, + "grad_norm": 0.49588069319725037, + "learning_rate": 7.798000000000001e-05, + "loss": 0.0907, + "step": 3900 + }, + { + "epoch": 0.7591864472598417, + "grad_norm": 0.27107104659080505, + "learning_rate": 7.818000000000001e-05, + "loss": 0.0713, + "step": 3910 + }, + { + "epoch": 0.7611281005776419, + "grad_norm": 0.3341488242149353, + "learning_rate": 7.838e-05, + "loss": 0.0745, + "step": 3920 + }, + { + "epoch": 0.763069753895442, + "grad_norm": 0.307542622089386, + "learning_rate": 7.858000000000001e-05, + "loss": 0.09, + "step": 3930 + }, + { + "epoch": 0.7650114072132421, + "grad_norm": 0.3722148537635803, + "learning_rate": 7.878e-05, + "loss": 0.0897, + "step": 3940 + }, + { + "epoch": 0.7669530605310422, + "grad_norm": 0.31524547934532166, + "learning_rate": 7.897999999999999e-05, + "loss": 0.0891, + "step": 3950 + }, + { + "epoch": 0.7688947138488423, + "grad_norm": 0.3329911530017853, + "learning_rate": 7.918e-05, + "loss": 0.099, + "step": 3960 + }, + { + "epoch": 0.7708363671666424, + "grad_norm": 0.3323957324028015, + "learning_rate": 7.938e-05, + "loss": 0.0826, + "step": 3970 + }, + { + "epoch": 0.7727780204844426, + "grad_norm": 0.3036642372608185, + "learning_rate": 7.958e-05, + "loss": 0.0907, + "step": 3980 + }, + { + "epoch": 0.7747196738022426, + "grad_norm": 0.45527851581573486, + "learning_rate": 7.978e-05, + "loss": 0.0847, + "step": 3990 + }, + { + "epoch": 0.7766613271200428, + "grad_norm": 0.2687672972679138, + "learning_rate": 7.998e-05, + "loss": 0.0755, + "step": 4000 + }, + { + "epoch": 0.7786029804378428, + "grad_norm": 0.5433294177055359, + "learning_rate": 8.018e-05, + "loss": 0.0861, + "step": 4010 + }, + { + "epoch": 0.780544633755643, + "grad_norm": 0.31862369179725647, + "learning_rate": 8.038e-05, + "loss": 0.0946, + "step": 4020 + }, + { + "epoch": 0.782486287073443, + "grad_norm": 0.22564281523227692, + "learning_rate": 8.058e-05, + "loss": 0.0821, + "step": 4030 + }, + { + "epoch": 0.7844279403912432, + "grad_norm": 0.33081990480422974, + "learning_rate": 8.078e-05, + "loss": 0.0861, + "step": 4040 + }, + { + "epoch": 0.7863695937090432, + "grad_norm": 0.2720225751399994, + "learning_rate": 8.098e-05, + "loss": 0.0846, + "step": 4050 + }, + { + "epoch": 0.7883112470268434, + "grad_norm": 0.30325376987457275, + "learning_rate": 8.118e-05, + "loss": 0.089, + "step": 4060 + }, + { + "epoch": 0.7902529003446435, + "grad_norm": 0.3266007602214813, + "learning_rate": 8.138e-05, + "loss": 0.0986, + "step": 4070 + }, + { + "epoch": 0.7921945536624436, + "grad_norm": 0.3556230366230011, + "learning_rate": 8.158e-05, + "loss": 0.085, + "step": 4080 + }, + { + "epoch": 0.7941362069802437, + "grad_norm": 0.2910439372062683, + "learning_rate": 8.178e-05, + "loss": 0.1005, + "step": 4090 + }, + { + "epoch": 0.7960778602980438, + "grad_norm": 0.29079538583755493, + "learning_rate": 8.198e-05, + "loss": 0.0762, + "step": 4100 + }, + { + "epoch": 0.7980195136158439, + "grad_norm": 0.29645758867263794, + "learning_rate": 8.218e-05, + "loss": 0.0715, + "step": 4110 + }, + { + "epoch": 0.799961166933644, + "grad_norm": 0.32848072052001953, + "learning_rate": 8.238000000000001e-05, + "loss": 0.0793, + "step": 4120 + }, + { + "epoch": 0.8019028202514441, + "grad_norm": 0.4054565131664276, + "learning_rate": 8.258e-05, + "loss": 0.0707, + "step": 4130 + }, + { + "epoch": 0.8038444735692443, + "grad_norm": 0.3966512978076935, + "learning_rate": 8.278e-05, + "loss": 0.0849, + "step": 4140 + }, + { + "epoch": 0.8057861268870443, + "grad_norm": 0.45230668783187866, + "learning_rate": 8.298000000000001e-05, + "loss": 0.0874, + "step": 4150 + }, + { + "epoch": 0.8077277802048444, + "grad_norm": 0.3674554228782654, + "learning_rate": 8.318e-05, + "loss": 0.0663, + "step": 4160 + }, + { + "epoch": 0.8096694335226445, + "grad_norm": 0.29878488183021545, + "learning_rate": 8.338e-05, + "loss": 0.0679, + "step": 4170 + }, + { + "epoch": 0.8116110868404446, + "grad_norm": 0.32820606231689453, + "learning_rate": 8.358e-05, + "loss": 0.0718, + "step": 4180 + }, + { + "epoch": 0.8135527401582447, + "grad_norm": 0.30448877811431885, + "learning_rate": 8.378e-05, + "loss": 0.068, + "step": 4190 + }, + { + "epoch": 0.8154943934760448, + "grad_norm": 0.2485552430152893, + "learning_rate": 8.398e-05, + "loss": 0.0719, + "step": 4200 + }, + { + "epoch": 0.817436046793845, + "grad_norm": 0.38694778084754944, + "learning_rate": 8.418e-05, + "loss": 0.0642, + "step": 4210 + }, + { + "epoch": 0.819377700111645, + "grad_norm": 0.20621079206466675, + "learning_rate": 8.438e-05, + "loss": 0.0763, + "step": 4220 + }, + { + "epoch": 0.8213193534294452, + "grad_norm": 0.2444322556257248, + "learning_rate": 8.458e-05, + "loss": 0.0695, + "step": 4230 + }, + { + "epoch": 0.8232610067472452, + "grad_norm": 0.30332979559898376, + "learning_rate": 8.478e-05, + "loss": 0.0643, + "step": 4240 + }, + { + "epoch": 0.8252026600650454, + "grad_norm": 0.37092408537864685, + "learning_rate": 8.498e-05, + "loss": 0.0838, + "step": 4250 + }, + { + "epoch": 0.8271443133828454, + "grad_norm": 0.47846096754074097, + "learning_rate": 8.518000000000001e-05, + "loss": 0.0649, + "step": 4260 + }, + { + "epoch": 0.8290859667006456, + "grad_norm": 0.3954634964466095, + "learning_rate": 8.538e-05, + "loss": 0.0858, + "step": 4270 + }, + { + "epoch": 0.8310276200184457, + "grad_norm": 0.4284418225288391, + "learning_rate": 8.558e-05, + "loss": 0.0743, + "step": 4280 + }, + { + "epoch": 0.8329692733362458, + "grad_norm": 0.5782247185707092, + "learning_rate": 8.578000000000001e-05, + "loss": 0.0931, + "step": 4290 + }, + { + "epoch": 0.8349109266540459, + "grad_norm": 0.3219737708568573, + "learning_rate": 8.598e-05, + "loss": 0.0715, + "step": 4300 + }, + { + "epoch": 0.836852579971846, + "grad_norm": 0.3683372437953949, + "learning_rate": 8.618e-05, + "loss": 0.0695, + "step": 4310 + }, + { + "epoch": 0.8387942332896461, + "grad_norm": 0.3684869110584259, + "learning_rate": 8.638000000000001e-05, + "loss": 0.0698, + "step": 4320 + }, + { + "epoch": 0.8407358866074462, + "grad_norm": 0.3562333881855011, + "learning_rate": 8.658e-05, + "loss": 0.0684, + "step": 4330 + }, + { + "epoch": 0.8426775399252463, + "grad_norm": 0.30072200298309326, + "learning_rate": 8.678e-05, + "loss": 0.0606, + "step": 4340 + }, + { + "epoch": 0.8446191932430465, + "grad_norm": 0.27639511227607727, + "learning_rate": 8.698000000000001e-05, + "loss": 0.0613, + "step": 4350 + }, + { + "epoch": 0.8465608465608465, + "grad_norm": 0.3153053820133209, + "learning_rate": 8.718e-05, + "loss": 0.081, + "step": 4360 + }, + { + "epoch": 0.8485024998786467, + "grad_norm": 0.2901988625526428, + "learning_rate": 8.738000000000001e-05, + "loss": 0.0638, + "step": 4370 + }, + { + "epoch": 0.8504441531964467, + "grad_norm": 0.48538738489151, + "learning_rate": 8.758000000000001e-05, + "loss": 0.073, + "step": 4380 + }, + { + "epoch": 0.8523858065142469, + "grad_norm": 0.4542801082134247, + "learning_rate": 8.778e-05, + "loss": 0.0776, + "step": 4390 + }, + { + "epoch": 0.8543274598320469, + "grad_norm": 0.5198235511779785, + "learning_rate": 8.798000000000001e-05, + "loss": 0.061, + "step": 4400 + }, + { + "epoch": 0.8562691131498471, + "grad_norm": 0.4022195637226105, + "learning_rate": 8.818000000000001e-05, + "loss": 0.0729, + "step": 4410 + }, + { + "epoch": 0.8582107664676472, + "grad_norm": 0.34400907158851624, + "learning_rate": 8.838e-05, + "loss": 0.0646, + "step": 4420 + }, + { + "epoch": 0.8601524197854473, + "grad_norm": 0.32496851682662964, + "learning_rate": 8.858000000000001e-05, + "loss": 0.0556, + "step": 4430 + }, + { + "epoch": 0.8620940731032474, + "grad_norm": 0.32707279920578003, + "learning_rate": 8.878000000000001e-05, + "loss": 0.0686, + "step": 4440 + }, + { + "epoch": 0.8640357264210475, + "grad_norm": 0.36298054456710815, + "learning_rate": 8.898e-05, + "loss": 0.0607, + "step": 4450 + }, + { + "epoch": 0.8659773797388476, + "grad_norm": 0.47542956471443176, + "learning_rate": 8.918000000000001e-05, + "loss": 0.0771, + "step": 4460 + }, + { + "epoch": 0.8679190330566477, + "grad_norm": 0.4965131878852844, + "learning_rate": 8.938e-05, + "loss": 0.076, + "step": 4470 + }, + { + "epoch": 0.8698606863744478, + "grad_norm": 0.40011972188949585, + "learning_rate": 8.958e-05, + "loss": 0.0615, + "step": 4480 + }, + { + "epoch": 0.871802339692248, + "grad_norm": 0.45475712418556213, + "learning_rate": 8.978000000000001e-05, + "loss": 0.0772, + "step": 4490 + }, + { + "epoch": 0.873743993010048, + "grad_norm": 0.3300098478794098, + "learning_rate": 8.998e-05, + "loss": 0.0796, + "step": 4500 + }, + { + "epoch": 0.8756856463278482, + "grad_norm": 0.36641377210617065, + "learning_rate": 9.018000000000001e-05, + "loss": 0.0547, + "step": 4510 + }, + { + "epoch": 0.8776272996456482, + "grad_norm": 0.5128288269042969, + "learning_rate": 9.038000000000001e-05, + "loss": 0.0791, + "step": 4520 + }, + { + "epoch": 0.8795689529634484, + "grad_norm": 0.4616294801235199, + "learning_rate": 9.058e-05, + "loss": 0.0707, + "step": 4530 + }, + { + "epoch": 0.8815106062812484, + "grad_norm": 0.21777303516864777, + "learning_rate": 9.078000000000001e-05, + "loss": 0.0602, + "step": 4540 + }, + { + "epoch": 0.8834522595990486, + "grad_norm": 0.3254261016845703, + "learning_rate": 9.098000000000001e-05, + "loss": 0.0694, + "step": 4550 + }, + { + "epoch": 0.8853939129168487, + "grad_norm": 0.6571286916732788, + "learning_rate": 9.118e-05, + "loss": 0.0689, + "step": 4560 + }, + { + "epoch": 0.8873355662346488, + "grad_norm": 0.4803417921066284, + "learning_rate": 9.138e-05, + "loss": 0.0639, + "step": 4570 + }, + { + "epoch": 0.8892772195524489, + "grad_norm": 0.294588565826416, + "learning_rate": 9.158e-05, + "loss": 0.0665, + "step": 4580 + }, + { + "epoch": 0.891218872870249, + "grad_norm": 0.3508646786212921, + "learning_rate": 9.178e-05, + "loss": 0.0511, + "step": 4590 + }, + { + "epoch": 0.8931605261880491, + "grad_norm": 0.2675318717956543, + "learning_rate": 9.198e-05, + "loss": 0.0605, + "step": 4600 + }, + { + "epoch": 0.8951021795058492, + "grad_norm": 0.34682559967041016, + "learning_rate": 9.218e-05, + "loss": 0.072, + "step": 4610 + }, + { + "epoch": 0.8970438328236493, + "grad_norm": 0.34355074167251587, + "learning_rate": 9.238e-05, + "loss": 0.0739, + "step": 4620 + }, + { + "epoch": 0.8989854861414495, + "grad_norm": 0.4112480878829956, + "learning_rate": 9.258e-05, + "loss": 0.0604, + "step": 4630 + }, + { + "epoch": 0.9009271394592495, + "grad_norm": 0.3439551889896393, + "learning_rate": 9.278e-05, + "loss": 0.0646, + "step": 4640 + }, + { + "epoch": 0.9028687927770497, + "grad_norm": 0.6033586859703064, + "learning_rate": 9.298e-05, + "loss": 0.0714, + "step": 4650 + }, + { + "epoch": 0.9048104460948497, + "grad_norm": 0.4311213493347168, + "learning_rate": 9.318e-05, + "loss": 0.0741, + "step": 4660 + }, + { + "epoch": 0.9067520994126499, + "grad_norm": 0.22362740337848663, + "learning_rate": 9.338e-05, + "loss": 0.0604, + "step": 4670 + }, + { + "epoch": 0.9086937527304499, + "grad_norm": 0.3199579417705536, + "learning_rate": 9.358e-05, + "loss": 0.0593, + "step": 4680 + }, + { + "epoch": 0.9106354060482501, + "grad_norm": 0.470705509185791, + "learning_rate": 9.378e-05, + "loss": 0.0685, + "step": 4690 + }, + { + "epoch": 0.9125770593660502, + "grad_norm": 0.26145389676094055, + "learning_rate": 9.398e-05, + "loss": 0.0679, + "step": 4700 + }, + { + "epoch": 0.9145187126838503, + "grad_norm": 0.40104666352272034, + "learning_rate": 9.418e-05, + "loss": 0.062, + "step": 4710 + }, + { + "epoch": 0.9164603660016504, + "grad_norm": 0.5261479616165161, + "learning_rate": 9.438e-05, + "loss": 0.0551, + "step": 4720 + }, + { + "epoch": 0.9184020193194505, + "grad_norm": 0.4142528772354126, + "learning_rate": 9.458e-05, + "loss": 0.0652, + "step": 4730 + }, + { + "epoch": 0.9203436726372506, + "grad_norm": 0.2667190134525299, + "learning_rate": 9.478e-05, + "loss": 0.0722, + "step": 4740 + }, + { + "epoch": 0.9222853259550507, + "grad_norm": 0.21195679903030396, + "learning_rate": 9.498e-05, + "loss": 0.0563, + "step": 4750 + }, + { + "epoch": 0.9242269792728508, + "grad_norm": 0.3581194281578064, + "learning_rate": 9.518000000000001e-05, + "loss": 0.0607, + "step": 4760 + }, + { + "epoch": 0.926168632590651, + "grad_norm": 0.37581291794776917, + "learning_rate": 9.538e-05, + "loss": 0.0684, + "step": 4770 + }, + { + "epoch": 0.928110285908451, + "grad_norm": 0.31677407026290894, + "learning_rate": 9.558e-05, + "loss": 0.0649, + "step": 4780 + }, + { + "epoch": 0.9300519392262512, + "grad_norm": 0.45203515887260437, + "learning_rate": 9.578000000000001e-05, + "loss": 0.0672, + "step": 4790 + }, + { + "epoch": 0.9319935925440512, + "grad_norm": 0.38241228461265564, + "learning_rate": 9.598e-05, + "loss": 0.0651, + "step": 4800 + }, + { + "epoch": 0.9339352458618514, + "grad_norm": 0.40158316493034363, + "learning_rate": 9.618e-05, + "loss": 0.0692, + "step": 4810 + }, + { + "epoch": 0.9358768991796514, + "grad_norm": 0.4598080813884735, + "learning_rate": 9.638000000000001e-05, + "loss": 0.0563, + "step": 4820 + }, + { + "epoch": 0.9378185524974516, + "grad_norm": 0.3534424602985382, + "learning_rate": 9.658e-05, + "loss": 0.0603, + "step": 4830 + }, + { + "epoch": 0.9397602058152517, + "grad_norm": 0.21484637260437012, + "learning_rate": 9.678e-05, + "loss": 0.0607, + "step": 4840 + }, + { + "epoch": 0.9417018591330518, + "grad_norm": 0.4767545461654663, + "learning_rate": 9.698000000000001e-05, + "loss": 0.0617, + "step": 4850 + }, + { + "epoch": 0.9436435124508519, + "grad_norm": 0.2935124933719635, + "learning_rate": 9.718e-05, + "loss": 0.0591, + "step": 4860 + }, + { + "epoch": 0.945585165768652, + "grad_norm": 0.35480669140815735, + "learning_rate": 9.738e-05, + "loss": 0.0626, + "step": 4870 + }, + { + "epoch": 0.9475268190864521, + "grad_norm": 0.27477896213531494, + "learning_rate": 9.758000000000001e-05, + "loss": 0.0564, + "step": 4880 + }, + { + "epoch": 0.9494684724042523, + "grad_norm": 0.4333362281322479, + "learning_rate": 9.778e-05, + "loss": 0.0689, + "step": 4890 + }, + { + "epoch": 0.9514101257220523, + "grad_norm": 0.21560749411582947, + "learning_rate": 9.798000000000001e-05, + "loss": 0.0526, + "step": 4900 + }, + { + "epoch": 0.9533517790398525, + "grad_norm": 0.3055712878704071, + "learning_rate": 9.818000000000001e-05, + "loss": 0.0662, + "step": 4910 + }, + { + "epoch": 0.9552934323576525, + "grad_norm": 0.24438029527664185, + "learning_rate": 9.838e-05, + "loss": 0.0663, + "step": 4920 + }, + { + "epoch": 0.9572350856754527, + "grad_norm": 0.36538824439048767, + "learning_rate": 9.858000000000001e-05, + "loss": 0.0566, + "step": 4930 + }, + { + "epoch": 0.9591767389932527, + "grad_norm": 0.2990020513534546, + "learning_rate": 9.878e-05, + "loss": 0.0562, + "step": 4940 + }, + { + "epoch": 0.9611183923110529, + "grad_norm": 0.27266067266464233, + "learning_rate": 9.898e-05, + "loss": 0.0492, + "step": 4950 + }, + { + "epoch": 0.963060045628853, + "grad_norm": 0.29196420311927795, + "learning_rate": 9.918000000000001e-05, + "loss": 0.0669, + "step": 4960 + }, + { + "epoch": 0.9650016989466531, + "grad_norm": 0.42901042103767395, + "learning_rate": 9.938e-05, + "loss": 0.0612, + "step": 4970 + }, + { + "epoch": 0.9669433522644532, + "grad_norm": 0.37816131114959717, + "learning_rate": 9.958e-05, + "loss": 0.0516, + "step": 4980 + }, + { + "epoch": 0.9688850055822533, + "grad_norm": 0.3445436656475067, + "learning_rate": 9.978000000000001e-05, + "loss": 0.063, + "step": 4990 + }, + { + "epoch": 0.9708266589000534, + "grad_norm": 0.41740062832832336, + "learning_rate": 9.998e-05, + "loss": 0.0664, + "step": 5000 + }, + { + "epoch": 0.9727683122178535, + "grad_norm": 0.4032188057899475, + "learning_rate": 9.999999778549045e-05, + "loss": 0.0597, + "step": 5010 + }, + { + "epoch": 0.9747099655356536, + "grad_norm": 0.27453017234802246, + "learning_rate": 9.999999013039593e-05, + "loss": 0.0628, + "step": 5020 + }, + { + "epoch": 0.9766516188534538, + "grad_norm": 0.3467614948749542, + "learning_rate": 9.999997700737766e-05, + "loss": 0.058, + "step": 5030 + }, + { + "epoch": 0.9785932721712538, + "grad_norm": 0.3921903073787689, + "learning_rate": 9.999995841643709e-05, + "loss": 0.0574, + "step": 5040 + }, + { + "epoch": 0.980534925489054, + "grad_norm": 0.29678109288215637, + "learning_rate": 9.999993435757623e-05, + "loss": 0.0541, + "step": 5050 + }, + { + "epoch": 0.982476578806854, + "grad_norm": 0.2856873869895935, + "learning_rate": 9.999990483079773e-05, + "loss": 0.0541, + "step": 5060 + }, + { + "epoch": 0.9844182321246542, + "grad_norm": 0.2773188352584839, + "learning_rate": 9.999986983610481e-05, + "loss": 0.059, + "step": 5070 + }, + { + "epoch": 0.9863598854424542, + "grad_norm": 0.4310706853866577, + "learning_rate": 9.99998293735013e-05, + "loss": 0.0606, + "step": 5080 + }, + { + "epoch": 0.9883015387602544, + "grad_norm": 0.3140304386615753, + "learning_rate": 9.999978344299161e-05, + "loss": 0.0571, + "step": 5090 + }, + { + "epoch": 0.9902431920780544, + "grad_norm": 0.33698901534080505, + "learning_rate": 9.99997320445808e-05, + "loss": 0.0551, + "step": 5100 + }, + { + "epoch": 0.9921848453958546, + "grad_norm": 0.35239389538764954, + "learning_rate": 9.999967517827444e-05, + "loss": 0.0656, + "step": 5110 + }, + { + "epoch": 0.9941264987136547, + "grad_norm": 0.290460467338562, + "learning_rate": 9.999961284407879e-05, + "loss": 0.0551, + "step": 5120 + }, + { + "epoch": 0.9960681520314548, + "grad_norm": 0.3724539577960968, + "learning_rate": 9.999954504200067e-05, + "loss": 0.0511, + "step": 5130 + }, + { + "epoch": 0.9980098053492549, + "grad_norm": 0.2868711054325104, + "learning_rate": 9.999947177204744e-05, + "loss": 0.0541, + "step": 5140 + }, + { + "epoch": 0.999951458667055, + "grad_norm": 0.3190750181674957, + "learning_rate": 9.999939303422718e-05, + "loss": 0.0666, + "step": 5150 + }, + { + "epoch": 1.0017474879860202, + "grad_norm": 0.3182838559150696, + "learning_rate": 9.999930882854847e-05, + "loss": 0.0524, + "step": 5160 + }, + { + "epoch": 1.0036891413038203, + "grad_norm": 0.2734878659248352, + "learning_rate": 9.999921915502051e-05, + "loss": 0.0613, + "step": 5170 + }, + { + "epoch": 1.0056307946216203, + "grad_norm": 0.29829275608062744, + "learning_rate": 9.99991240136531e-05, + "loss": 0.0673, + "step": 5180 + }, + { + "epoch": 1.0075724479394204, + "grad_norm": 0.4973118305206299, + "learning_rate": 9.999902340445668e-05, + "loss": 0.0564, + "step": 5190 + }, + { + "epoch": 1.0095141012572206, + "grad_norm": 0.3299853205680847, + "learning_rate": 9.999891732744224e-05, + "loss": 0.0568, + "step": 5200 + }, + { + "epoch": 1.0114557545750207, + "grad_norm": 0.3925850987434387, + "learning_rate": 9.999880578262135e-05, + "loss": 0.0608, + "step": 5210 + }, + { + "epoch": 1.0133974078928207, + "grad_norm": 0.31341657042503357, + "learning_rate": 9.999868877000624e-05, + "loss": 0.0506, + "step": 5220 + }, + { + "epoch": 1.0153390612106208, + "grad_norm": 0.3104129731655121, + "learning_rate": 9.99985662896097e-05, + "loss": 0.0531, + "step": 5230 + }, + { + "epoch": 1.017280714528421, + "grad_norm": 0.36955738067626953, + "learning_rate": 9.999843834144513e-05, + "loss": 0.0572, + "step": 5240 + }, + { + "epoch": 1.0192223678462211, + "grad_norm": 0.6062698364257812, + "learning_rate": 9.99983049255265e-05, + "loss": 0.0537, + "step": 5250 + }, + { + "epoch": 1.0211640211640212, + "grad_norm": 0.45700544118881226, + "learning_rate": 9.999816604186843e-05, + "loss": 0.0653, + "step": 5260 + }, + { + "epoch": 1.0231056744818212, + "grad_norm": 0.22529305517673492, + "learning_rate": 9.999802169048609e-05, + "loss": 0.0525, + "step": 5270 + }, + { + "epoch": 1.0250473277996215, + "grad_norm": 0.31568610668182373, + "learning_rate": 9.999787187139527e-05, + "loss": 0.05, + "step": 5280 + }, + { + "epoch": 1.0269889811174215, + "grad_norm": 0.30549484491348267, + "learning_rate": 9.999771658461234e-05, + "loss": 0.0611, + "step": 5290 + }, + { + "epoch": 1.0289306344352216, + "grad_norm": 0.28901299834251404, + "learning_rate": 9.999755583015431e-05, + "loss": 0.0717, + "step": 5300 + }, + { + "epoch": 1.0308722877530216, + "grad_norm": 0.3047245442867279, + "learning_rate": 9.999738960803874e-05, + "loss": 0.0584, + "step": 5310 + }, + { + "epoch": 1.0328139410708217, + "grad_norm": 0.3686572313308716, + "learning_rate": 9.99972179182838e-05, + "loss": 0.0617, + "step": 5320 + }, + { + "epoch": 1.034755594388622, + "grad_norm": 0.4501430094242096, + "learning_rate": 9.99970407609083e-05, + "loss": 0.0575, + "step": 5330 + }, + { + "epoch": 1.036697247706422, + "grad_norm": 0.585284411907196, + "learning_rate": 9.999685813593159e-05, + "loss": 0.0597, + "step": 5340 + }, + { + "epoch": 1.038638901024222, + "grad_norm": 0.3215489685535431, + "learning_rate": 9.999667004337362e-05, + "loss": 0.0615, + "step": 5350 + }, + { + "epoch": 1.0405805543420223, + "grad_norm": 0.3029697835445404, + "learning_rate": 9.9996476483255e-05, + "loss": 0.0587, + "step": 5360 + }, + { + "epoch": 1.0425222076598224, + "grad_norm": 0.31280162930488586, + "learning_rate": 9.999627745559688e-05, + "loss": 0.0461, + "step": 5370 + }, + { + "epoch": 1.0444638609776224, + "grad_norm": 0.27245426177978516, + "learning_rate": 9.999607296042101e-05, + "loss": 0.058, + "step": 5380 + }, + { + "epoch": 1.0464055142954225, + "grad_norm": 0.3302372395992279, + "learning_rate": 9.99958629977498e-05, + "loss": 0.0597, + "step": 5390 + }, + { + "epoch": 1.0483471676132226, + "grad_norm": 0.3860567808151245, + "learning_rate": 9.999564756760615e-05, + "loss": 0.0646, + "step": 5400 + }, + { + "epoch": 1.0502888209310228, + "grad_norm": 0.42741817235946655, + "learning_rate": 9.999542667001366e-05, + "loss": 0.0572, + "step": 5410 + }, + { + "epoch": 1.0522304742488229, + "grad_norm": 0.27999353408813477, + "learning_rate": 9.999520030499647e-05, + "loss": 0.0555, + "step": 5420 + }, + { + "epoch": 1.054172127566623, + "grad_norm": 0.22606660425662994, + "learning_rate": 9.999496847257936e-05, + "loss": 0.0537, + "step": 5430 + }, + { + "epoch": 1.056113780884423, + "grad_norm": 0.38137513399124146, + "learning_rate": 9.999473117278764e-05, + "loss": 0.0628, + "step": 5440 + }, + { + "epoch": 1.0580554342022233, + "grad_norm": 0.16732703149318695, + "learning_rate": 9.999448840564731e-05, + "loss": 0.0593, + "step": 5450 + }, + { + "epoch": 1.0599970875200233, + "grad_norm": 0.2576126158237457, + "learning_rate": 9.999424017118488e-05, + "loss": 0.05, + "step": 5460 + }, + { + "epoch": 1.0619387408378234, + "grad_norm": 0.25189247727394104, + "learning_rate": 9.999398646942751e-05, + "loss": 0.0662, + "step": 5470 + }, + { + "epoch": 1.0638803941556234, + "grad_norm": 0.3517856299877167, + "learning_rate": 9.999372730040296e-05, + "loss": 0.0547, + "step": 5480 + }, + { + "epoch": 1.0658220474734237, + "grad_norm": 0.22108094394207, + "learning_rate": 9.999346266413953e-05, + "loss": 0.0494, + "step": 5490 + }, + { + "epoch": 1.0677637007912237, + "grad_norm": 0.32149940729141235, + "learning_rate": 9.99931925606662e-05, + "loss": 0.049, + "step": 5500 + }, + { + "epoch": 1.0697053541090238, + "grad_norm": 0.3427355885505676, + "learning_rate": 9.99929169900125e-05, + "loss": 0.0541, + "step": 5510 + }, + { + "epoch": 1.0716470074268238, + "grad_norm": 0.24608942866325378, + "learning_rate": 9.999263595220855e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.0735886607446241, + "grad_norm": 0.49799844622612, + "learning_rate": 9.99923494472851e-05, + "loss": 0.0558, + "step": 5530 + }, + { + "epoch": 1.0755303140624242, + "grad_norm": 0.3792102336883545, + "learning_rate": 9.999205747527348e-05, + "loss": 0.0552, + "step": 5540 + }, + { + "epoch": 1.0774719673802242, + "grad_norm": 0.29256609082221985, + "learning_rate": 9.999176003620561e-05, + "loss": 0.0498, + "step": 5550 + }, + { + "epoch": 1.0794136206980243, + "grad_norm": 0.30412667989730835, + "learning_rate": 9.999145713011405e-05, + "loss": 0.0584, + "step": 5560 + }, + { + "epoch": 1.0813552740158245, + "grad_norm": 0.2858879268169403, + "learning_rate": 9.999114875703186e-05, + "loss": 0.0581, + "step": 5570 + }, + { + "epoch": 1.0832969273336246, + "grad_norm": 0.32978129386901855, + "learning_rate": 9.999083491699281e-05, + "loss": 0.0529, + "step": 5580 + }, + { + "epoch": 1.0852385806514246, + "grad_norm": 0.2397681623697281, + "learning_rate": 9.999051561003123e-05, + "loss": 0.0523, + "step": 5590 + }, + { + "epoch": 1.0871802339692247, + "grad_norm": 0.3235253393650055, + "learning_rate": 9.999019083618202e-05, + "loss": 0.0602, + "step": 5600 + }, + { + "epoch": 1.089121887287025, + "grad_norm": 0.36578860878944397, + "learning_rate": 9.99898605954807e-05, + "loss": 0.053, + "step": 5610 + }, + { + "epoch": 1.091063540604825, + "grad_norm": 0.2374127060174942, + "learning_rate": 9.998952488796338e-05, + "loss": 0.0558, + "step": 5620 + }, + { + "epoch": 1.093005193922625, + "grad_norm": 0.23675839602947235, + "learning_rate": 9.998918371366676e-05, + "loss": 0.0502, + "step": 5630 + }, + { + "epoch": 1.0949468472404251, + "grad_norm": 0.44787371158599854, + "learning_rate": 9.99888370726282e-05, + "loss": 0.06, + "step": 5640 + }, + { + "epoch": 1.0968885005582254, + "grad_norm": 0.43243035674095154, + "learning_rate": 9.998848496488556e-05, + "loss": 0.0608, + "step": 5650 + }, + { + "epoch": 1.0988301538760255, + "grad_norm": 0.3519284427165985, + "learning_rate": 9.998812739047736e-05, + "loss": 0.0531, + "step": 5660 + }, + { + "epoch": 1.1007718071938255, + "grad_norm": 0.305257648229599, + "learning_rate": 9.99877643494427e-05, + "loss": 0.0483, + "step": 5670 + }, + { + "epoch": 1.1027134605116256, + "grad_norm": 0.32153913378715515, + "learning_rate": 9.998739584182128e-05, + "loss": 0.0607, + "step": 5680 + }, + { + "epoch": 1.1046551138294258, + "grad_norm": 0.26991546154022217, + "learning_rate": 9.998702186765342e-05, + "loss": 0.0492, + "step": 5690 + }, + { + "epoch": 1.1065967671472259, + "grad_norm": 0.25957292318344116, + "learning_rate": 9.998664242698e-05, + "loss": 0.0621, + "step": 5700 + }, + { + "epoch": 1.108538420465026, + "grad_norm": 0.38618698716163635, + "learning_rate": 9.998625751984251e-05, + "loss": 0.06, + "step": 5710 + }, + { + "epoch": 1.110480073782826, + "grad_norm": 0.1858365684747696, + "learning_rate": 9.998586714628307e-05, + "loss": 0.0481, + "step": 5720 + }, + { + "epoch": 1.1124217271006263, + "grad_norm": 0.3420273959636688, + "learning_rate": 9.998547130634432e-05, + "loss": 0.0597, + "step": 5730 + }, + { + "epoch": 1.1143633804184263, + "grad_norm": 0.30784207582473755, + "learning_rate": 9.99850700000696e-05, + "loss": 0.0521, + "step": 5740 + }, + { + "epoch": 1.1163050337362264, + "grad_norm": 0.4191115200519562, + "learning_rate": 9.998466322750278e-05, + "loss": 0.0555, + "step": 5750 + }, + { + "epoch": 1.1182466870540264, + "grad_norm": 0.29292359948158264, + "learning_rate": 9.998425098868834e-05, + "loss": 0.0515, + "step": 5760 + }, + { + "epoch": 1.1201883403718267, + "grad_norm": 0.2771179974079132, + "learning_rate": 9.998383328367136e-05, + "loss": 0.0521, + "step": 5770 + }, + { + "epoch": 1.1221299936896267, + "grad_norm": 0.2865660786628723, + "learning_rate": 9.99834101124975e-05, + "loss": 0.0559, + "step": 5780 + }, + { + "epoch": 1.1240716470074268, + "grad_norm": 0.3096623718738556, + "learning_rate": 9.998298147521309e-05, + "loss": 0.0529, + "step": 5790 + }, + { + "epoch": 1.1260133003252268, + "grad_norm": 0.13531656563282013, + "learning_rate": 9.998254737186496e-05, + "loss": 0.05, + "step": 5800 + }, + { + "epoch": 1.1279549536430271, + "grad_norm": 0.24478082358837128, + "learning_rate": 9.99821078025006e-05, + "loss": 0.0588, + "step": 5810 + }, + { + "epoch": 1.1298966069608272, + "grad_norm": 0.23964929580688477, + "learning_rate": 9.998166276716807e-05, + "loss": 0.0614, + "step": 5820 + }, + { + "epoch": 1.1318382602786272, + "grad_norm": 0.2804964482784271, + "learning_rate": 9.998121226591606e-05, + "loss": 0.0479, + "step": 5830 + }, + { + "epoch": 1.1337799135964273, + "grad_norm": 0.4008128345012665, + "learning_rate": 9.998075629879382e-05, + "loss": 0.0495, + "step": 5840 + }, + { + "epoch": 1.1357215669142275, + "grad_norm": 0.19999518990516663, + "learning_rate": 9.99802948658512e-05, + "loss": 0.0557, + "step": 5850 + }, + { + "epoch": 1.1376632202320276, + "grad_norm": 0.23561744391918182, + "learning_rate": 9.99798279671387e-05, + "loss": 0.051, + "step": 5860 + }, + { + "epoch": 1.1396048735498276, + "grad_norm": 0.42447853088378906, + "learning_rate": 9.997935560270734e-05, + "loss": 0.0486, + "step": 5870 + }, + { + "epoch": 1.1415465268676277, + "grad_norm": 0.27052468061447144, + "learning_rate": 9.997887777260879e-05, + "loss": 0.0591, + "step": 5880 + }, + { + "epoch": 1.143488180185428, + "grad_norm": 0.27513593435287476, + "learning_rate": 9.997839447689532e-05, + "loss": 0.0544, + "step": 5890 + }, + { + "epoch": 1.145429833503228, + "grad_norm": 0.37868040800094604, + "learning_rate": 9.997790571561978e-05, + "loss": 0.0533, + "step": 5900 + }, + { + "epoch": 1.147371486821028, + "grad_norm": 0.35527774691581726, + "learning_rate": 9.99774114888356e-05, + "loss": 0.062, + "step": 5910 + }, + { + "epoch": 1.1493131401388281, + "grad_norm": 0.24411168694496155, + "learning_rate": 9.997691179659684e-05, + "loss": 0.0572, + "step": 5920 + }, + { + "epoch": 1.1512547934566284, + "grad_norm": 0.29488056898117065, + "learning_rate": 9.997640663895815e-05, + "loss": 0.0603, + "step": 5930 + }, + { + "epoch": 1.1531964467744285, + "grad_norm": 0.2158249318599701, + "learning_rate": 9.997589601597477e-05, + "loss": 0.0525, + "step": 5940 + }, + { + "epoch": 1.1551381000922285, + "grad_norm": 0.4165487587451935, + "learning_rate": 9.997537992770252e-05, + "loss": 0.0569, + "step": 5950 + }, + { + "epoch": 1.1570797534100286, + "grad_norm": 0.3657020032405853, + "learning_rate": 9.997485837419788e-05, + "loss": 0.053, + "step": 5960 + }, + { + "epoch": 1.1590214067278288, + "grad_norm": 0.33390143513679504, + "learning_rate": 9.997433135551786e-05, + "loss": 0.0515, + "step": 5970 + }, + { + "epoch": 1.1609630600456289, + "grad_norm": 0.3236733078956604, + "learning_rate": 9.997379887172009e-05, + "loss": 0.0416, + "step": 5980 + }, + { + "epoch": 1.162904713363429, + "grad_norm": 0.7071574926376343, + "learning_rate": 9.997326092286281e-05, + "loss": 0.0538, + "step": 5990 + }, + { + "epoch": 1.164846366681229, + "grad_norm": 0.2767888009548187, + "learning_rate": 9.997271750900486e-05, + "loss": 0.0589, + "step": 6000 + }, + { + "epoch": 1.1667880199990293, + "grad_norm": 0.29244938492774963, + "learning_rate": 9.997216863020565e-05, + "loss": 0.0554, + "step": 6010 + }, + { + "epoch": 1.1687296733168293, + "grad_norm": 0.3130743205547333, + "learning_rate": 9.99716142865252e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.1706713266346294, + "grad_norm": 0.22452761232852936, + "learning_rate": 9.997105447802415e-05, + "loss": 0.0619, + "step": 6030 + }, + { + "epoch": 1.1726129799524294, + "grad_norm": 0.368327796459198, + "learning_rate": 9.997048920476373e-05, + "loss": 0.0653, + "step": 6040 + }, + { + "epoch": 1.1745546332702297, + "grad_norm": 0.30901193618774414, + "learning_rate": 9.996991846680572e-05, + "loss": 0.05, + "step": 6050 + }, + { + "epoch": 1.1764962865880297, + "grad_norm": 0.3477327525615692, + "learning_rate": 9.996934226421257e-05, + "loss": 0.0479, + "step": 6060 + }, + { + "epoch": 1.1784379399058298, + "grad_norm": 0.2303217202425003, + "learning_rate": 9.996876059704726e-05, + "loss": 0.0558, + "step": 6070 + }, + { + "epoch": 1.1803795932236298, + "grad_norm": 0.2169295996427536, + "learning_rate": 9.996817346537343e-05, + "loss": 0.0473, + "step": 6080 + }, + { + "epoch": 1.1823212465414301, + "grad_norm": 0.27642303705215454, + "learning_rate": 9.996758086925526e-05, + "loss": 0.0459, + "step": 6090 + }, + { + "epoch": 1.1842628998592302, + "grad_norm": 0.32279255986213684, + "learning_rate": 9.996698280875759e-05, + "loss": 0.0587, + "step": 6100 + }, + { + "epoch": 1.1862045531770302, + "grad_norm": 0.3808571696281433, + "learning_rate": 9.99663792839458e-05, + "loss": 0.0577, + "step": 6110 + }, + { + "epoch": 1.1881462064948303, + "grad_norm": 0.33392783999443054, + "learning_rate": 9.99657702948859e-05, + "loss": 0.0484, + "step": 6120 + }, + { + "epoch": 1.1900878598126305, + "grad_norm": 0.2921878397464752, + "learning_rate": 9.996515584164448e-05, + "loss": 0.0546, + "step": 6130 + }, + { + "epoch": 1.1920295131304306, + "grad_norm": 0.17339059710502625, + "learning_rate": 9.996453592428873e-05, + "loss": 0.0525, + "step": 6140 + }, + { + "epoch": 1.1939711664482306, + "grad_norm": 0.2966257333755493, + "learning_rate": 9.996391054288646e-05, + "loss": 0.0552, + "step": 6150 + }, + { + "epoch": 1.1959128197660307, + "grad_norm": 0.19397322833538055, + "learning_rate": 9.996327969750605e-05, + "loss": 0.0471, + "step": 6160 + }, + { + "epoch": 1.197854473083831, + "grad_norm": 0.2913099527359009, + "learning_rate": 9.996264338821649e-05, + "loss": 0.0491, + "step": 6170 + }, + { + "epoch": 1.199796126401631, + "grad_norm": 0.3895440697669983, + "learning_rate": 9.996200161508735e-05, + "loss": 0.0523, + "step": 6180 + }, + { + "epoch": 1.201737779719431, + "grad_norm": 0.3942548334598541, + "learning_rate": 9.996135437818885e-05, + "loss": 0.0492, + "step": 6190 + }, + { + "epoch": 1.2036794330372311, + "grad_norm": 0.2796578109264374, + "learning_rate": 9.996070167759175e-05, + "loss": 0.0568, + "step": 6200 + }, + { + "epoch": 1.2056210863550314, + "grad_norm": 0.31294628977775574, + "learning_rate": 9.996004351336743e-05, + "loss": 0.057, + "step": 6210 + }, + { + "epoch": 1.2075627396728315, + "grad_norm": 0.25204581022262573, + "learning_rate": 9.995937988558785e-05, + "loss": 0.0657, + "step": 6220 + }, + { + "epoch": 1.2095043929906315, + "grad_norm": 0.34203261137008667, + "learning_rate": 9.995871079432561e-05, + "loss": 0.0477, + "step": 6230 + }, + { + "epoch": 1.2114460463084316, + "grad_norm": 0.34053894877433777, + "learning_rate": 9.995803623965389e-05, + "loss": 0.0525, + "step": 6240 + }, + { + "epoch": 1.2133876996262318, + "grad_norm": 0.3352050483226776, + "learning_rate": 9.995735622164641e-05, + "loss": 0.0495, + "step": 6250 + }, + { + "epoch": 1.2153293529440319, + "grad_norm": 0.2242475152015686, + "learning_rate": 9.995667074037758e-05, + "loss": 0.0549, + "step": 6260 + }, + { + "epoch": 1.217271006261832, + "grad_norm": 0.30098655819892883, + "learning_rate": 9.995597979592232e-05, + "loss": 0.0437, + "step": 6270 + }, + { + "epoch": 1.219212659579632, + "grad_norm": 0.3822211027145386, + "learning_rate": 9.995528338835625e-05, + "loss": 0.0513, + "step": 6280 + }, + { + "epoch": 1.2211543128974323, + "grad_norm": 0.4117942750453949, + "learning_rate": 9.995458151775547e-05, + "loss": 0.0524, + "step": 6290 + }, + { + "epoch": 1.2230959662152323, + "grad_norm": 0.4916922152042389, + "learning_rate": 9.995387418419677e-05, + "loss": 0.0532, + "step": 6300 + }, + { + "epoch": 1.2250376195330324, + "grad_norm": 0.3239062428474426, + "learning_rate": 9.99531613877575e-05, + "loss": 0.0428, + "step": 6310 + }, + { + "epoch": 1.2269792728508324, + "grad_norm": 0.39545661211013794, + "learning_rate": 9.995244312851559e-05, + "loss": 0.0494, + "step": 6320 + }, + { + "epoch": 1.2289209261686325, + "grad_norm": 0.34091946482658386, + "learning_rate": 9.995171940654961e-05, + "loss": 0.0539, + "step": 6330 + }, + { + "epoch": 1.2308625794864327, + "grad_norm": 0.2678007185459137, + "learning_rate": 9.995099022193871e-05, + "loss": 0.0592, + "step": 6340 + }, + { + "epoch": 1.2328042328042328, + "grad_norm": 0.2844739556312561, + "learning_rate": 9.995025557476261e-05, + "loss": 0.0438, + "step": 6350 + }, + { + "epoch": 1.2347458861220328, + "grad_norm": 0.2497197389602661, + "learning_rate": 9.994951546510165e-05, + "loss": 0.0507, + "step": 6360 + }, + { + "epoch": 1.2366875394398331, + "grad_norm": 0.34011590480804443, + "learning_rate": 9.994876989303679e-05, + "loss": 0.0549, + "step": 6370 + }, + { + "epoch": 1.2386291927576332, + "grad_norm": 0.28280484676361084, + "learning_rate": 9.994801885864955e-05, + "loss": 0.0446, + "step": 6380 + }, + { + "epoch": 1.2405708460754332, + "grad_norm": 0.23113466799259186, + "learning_rate": 9.994726236202205e-05, + "loss": 0.0629, + "step": 6390 + }, + { + "epoch": 1.2425124993932333, + "grad_norm": 0.3290514647960663, + "learning_rate": 9.994650040323704e-05, + "loss": 0.0628, + "step": 6400 + }, + { + "epoch": 1.2444541527110333, + "grad_norm": 0.3533608019351959, + "learning_rate": 9.994573298237784e-05, + "loss": 0.0644, + "step": 6410 + }, + { + "epoch": 1.2463958060288336, + "grad_norm": 0.294447124004364, + "learning_rate": 9.994496009952837e-05, + "loss": 0.0453, + "step": 6420 + }, + { + "epoch": 1.2483374593466336, + "grad_norm": 0.48502835631370544, + "learning_rate": 9.994418175477316e-05, + "loss": 0.058, + "step": 6430 + }, + { + "epoch": 1.2502791126644337, + "grad_norm": 0.5640022158622742, + "learning_rate": 9.994339794819733e-05, + "loss": 0.0514, + "step": 6440 + }, + { + "epoch": 1.252220765982234, + "grad_norm": 0.3374970257282257, + "learning_rate": 9.994260867988658e-05, + "loss": 0.0552, + "step": 6450 + }, + { + "epoch": 1.254162419300034, + "grad_norm": 0.2787913978099823, + "learning_rate": 9.994181394992723e-05, + "loss": 0.0405, + "step": 6460 + }, + { + "epoch": 1.256104072617834, + "grad_norm": 0.42703109979629517, + "learning_rate": 9.994101375840618e-05, + "loss": 0.0446, + "step": 6470 + }, + { + "epoch": 1.2580457259356341, + "grad_norm": 0.2737075090408325, + "learning_rate": 9.994020810541098e-05, + "loss": 0.0388, + "step": 6480 + }, + { + "epoch": 1.2599873792534342, + "grad_norm": 0.2438925951719284, + "learning_rate": 9.99393969910297e-05, + "loss": 0.0477, + "step": 6490 + }, + { + "epoch": 1.2619290325712345, + "grad_norm": 0.2909867763519287, + "learning_rate": 9.993858041535104e-05, + "loss": 0.0488, + "step": 6500 + }, + { + "epoch": 1.2638706858890345, + "grad_norm": 0.34121087193489075, + "learning_rate": 9.99377583784643e-05, + "loss": 0.0568, + "step": 6510 + }, + { + "epoch": 1.2658123392068346, + "grad_norm": 0.3144819140434265, + "learning_rate": 9.993693088045939e-05, + "loss": 0.054, + "step": 6520 + }, + { + "epoch": 1.2677539925246348, + "grad_norm": 0.4358678460121155, + "learning_rate": 9.99360979214268e-05, + "loss": 0.0438, + "step": 6530 + }, + { + "epoch": 1.2696956458424349, + "grad_norm": 0.25227090716362, + "learning_rate": 9.99352595014576e-05, + "loss": 0.0521, + "step": 6540 + }, + { + "epoch": 1.271637299160235, + "grad_norm": 0.21019767224788666, + "learning_rate": 9.993441562064354e-05, + "loss": 0.054, + "step": 6550 + }, + { + "epoch": 1.273578952478035, + "grad_norm": 0.33969736099243164, + "learning_rate": 9.993356627907685e-05, + "loss": 0.042, + "step": 6560 + }, + { + "epoch": 1.275520605795835, + "grad_norm": 0.30986034870147705, + "learning_rate": 9.99327114768504e-05, + "loss": 0.058, + "step": 6570 + }, + { + "epoch": 1.2774622591136353, + "grad_norm": 0.1282600313425064, + "learning_rate": 9.99318512140577e-05, + "loss": 0.0453, + "step": 6580 + }, + { + "epoch": 1.2794039124314354, + "grad_norm": 0.21246151626110077, + "learning_rate": 9.993098549079284e-05, + "loss": 0.0542, + "step": 6590 + }, + { + "epoch": 1.2813455657492354, + "grad_norm": 0.40450531244277954, + "learning_rate": 9.993011430715047e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.2832872190670357, + "grad_norm": 0.4046102464199066, + "learning_rate": 9.992923766322586e-05, + "loss": 0.0394, + "step": 6610 + }, + { + "epoch": 1.2852288723848357, + "grad_norm": 0.3313716650009155, + "learning_rate": 9.99283555591149e-05, + "loss": 0.0461, + "step": 6620 + }, + { + "epoch": 1.2871705257026358, + "grad_norm": 0.223741352558136, + "learning_rate": 9.992746799491404e-05, + "loss": 0.0485, + "step": 6630 + }, + { + "epoch": 1.2891121790204358, + "grad_norm": 0.31067389249801636, + "learning_rate": 9.992657497072033e-05, + "loss": 0.0439, + "step": 6640 + }, + { + "epoch": 1.291053832338236, + "grad_norm": 0.3105778694152832, + "learning_rate": 9.992567648663147e-05, + "loss": 0.0402, + "step": 6650 + }, + { + "epoch": 1.2929954856560362, + "grad_norm": 0.2764989733695984, + "learning_rate": 9.992477254274568e-05, + "loss": 0.0499, + "step": 6660 + }, + { + "epoch": 1.2949371389738362, + "grad_norm": 0.2515884041786194, + "learning_rate": 9.992386313916183e-05, + "loss": 0.0485, + "step": 6670 + }, + { + "epoch": 1.2968787922916363, + "grad_norm": 0.5302994847297668, + "learning_rate": 9.992294827597934e-05, + "loss": 0.0479, + "step": 6680 + }, + { + "epoch": 1.2988204456094365, + "grad_norm": 0.425997257232666, + "learning_rate": 9.992202795329831e-05, + "loss": 0.0494, + "step": 6690 + }, + { + "epoch": 1.3007620989272366, + "grad_norm": 0.40923991799354553, + "learning_rate": 9.992110217121936e-05, + "loss": 0.0596, + "step": 6700 + }, + { + "epoch": 1.3027037522450367, + "grad_norm": 0.24352115392684937, + "learning_rate": 9.992017092984372e-05, + "loss": 0.0524, + "step": 6710 + }, + { + "epoch": 1.3046454055628367, + "grad_norm": 0.2860632538795471, + "learning_rate": 9.991923422927326e-05, + "loss": 0.043, + "step": 6720 + }, + { + "epoch": 1.3065870588806368, + "grad_norm": 0.3363390266895294, + "learning_rate": 9.991829206961037e-05, + "loss": 0.0655, + "step": 6730 + }, + { + "epoch": 1.308528712198437, + "grad_norm": 0.2556909918785095, + "learning_rate": 9.991734445095813e-05, + "loss": 0.0474, + "step": 6740 + }, + { + "epoch": 1.310470365516237, + "grad_norm": 0.40135130286216736, + "learning_rate": 9.991639137342015e-05, + "loss": 0.0464, + "step": 6750 + }, + { + "epoch": 1.3124120188340371, + "grad_norm": 0.38858264684677124, + "learning_rate": 9.991543283710064e-05, + "loss": 0.0514, + "step": 6760 + }, + { + "epoch": 1.3143536721518374, + "grad_norm": 0.3232880234718323, + "learning_rate": 9.991446884210445e-05, + "loss": 0.0556, + "step": 6770 + }, + { + "epoch": 1.3162953254696375, + "grad_norm": 0.3980470895767212, + "learning_rate": 9.9913499388537e-05, + "loss": 0.0471, + "step": 6780 + }, + { + "epoch": 1.3182369787874375, + "grad_norm": 0.3708045184612274, + "learning_rate": 9.99125244765043e-05, + "loss": 0.0436, + "step": 6790 + }, + { + "epoch": 1.3201786321052376, + "grad_norm": 0.2489119917154312, + "learning_rate": 9.991154410611296e-05, + "loss": 0.0634, + "step": 6800 + }, + { + "epoch": 1.3221202854230376, + "grad_norm": 0.3118704855442047, + "learning_rate": 9.99105582774702e-05, + "loss": 0.0556, + "step": 6810 + }, + { + "epoch": 1.3240619387408379, + "grad_norm": 0.2139330953359604, + "learning_rate": 9.990956699068384e-05, + "loss": 0.0469, + "step": 6820 + }, + { + "epoch": 1.326003592058638, + "grad_norm": 0.22830934822559357, + "learning_rate": 9.990857024586224e-05, + "loss": 0.0558, + "step": 6830 + }, + { + "epoch": 1.327945245376438, + "grad_norm": 0.3085069954395294, + "learning_rate": 9.990756804311446e-05, + "loss": 0.0467, + "step": 6840 + }, + { + "epoch": 1.3298868986942383, + "grad_norm": 0.29849398136138916, + "learning_rate": 9.990656038255006e-05, + "loss": 0.0477, + "step": 6850 + }, + { + "epoch": 1.3318285520120383, + "grad_norm": 0.418740838766098, + "learning_rate": 9.990554726427926e-05, + "loss": 0.0527, + "step": 6860 + }, + { + "epoch": 1.3337702053298384, + "grad_norm": 0.24286437034606934, + "learning_rate": 9.990452868841284e-05, + "loss": 0.041, + "step": 6870 + }, + { + "epoch": 1.3357118586476384, + "grad_norm": 0.41125357151031494, + "learning_rate": 9.99035046550622e-05, + "loss": 0.0544, + "step": 6880 + }, + { + "epoch": 1.3376535119654385, + "grad_norm": 0.3237217962741852, + "learning_rate": 9.99024751643393e-05, + "loss": 0.0504, + "step": 6890 + }, + { + "epoch": 1.3395951652832387, + "grad_norm": 0.3295341730117798, + "learning_rate": 9.990144021635677e-05, + "loss": 0.0473, + "step": 6900 + }, + { + "epoch": 1.3415368186010388, + "grad_norm": 0.2388688027858734, + "learning_rate": 9.990039981122775e-05, + "loss": 0.0579, + "step": 6910 + }, + { + "epoch": 1.3434784719188388, + "grad_norm": 0.31518450379371643, + "learning_rate": 9.989935394906602e-05, + "loss": 0.0561, + "step": 6920 + }, + { + "epoch": 1.3454201252366391, + "grad_norm": 0.2164788693189621, + "learning_rate": 9.989830262998598e-05, + "loss": 0.0433, + "step": 6930 + }, + { + "epoch": 1.3473617785544392, + "grad_norm": 0.27137333154678345, + "learning_rate": 9.989724585410259e-05, + "loss": 0.0501, + "step": 6940 + }, + { + "epoch": 1.3493034318722392, + "grad_norm": 0.301517516374588, + "learning_rate": 9.989618362153139e-05, + "loss": 0.0496, + "step": 6950 + }, + { + "epoch": 1.3512450851900393, + "grad_norm": 0.31867551803588867, + "learning_rate": 9.989511593238859e-05, + "loss": 0.045, + "step": 6960 + }, + { + "epoch": 1.3531867385078393, + "grad_norm": 0.30827945470809937, + "learning_rate": 9.98940427867909e-05, + "loss": 0.0516, + "step": 6970 + }, + { + "epoch": 1.3551283918256396, + "grad_norm": 0.26140379905700684, + "learning_rate": 9.989296418485573e-05, + "loss": 0.0494, + "step": 6980 + }, + { + "epoch": 1.3570700451434397, + "grad_norm": 0.35594692826271057, + "learning_rate": 9.989188012670101e-05, + "loss": 0.0586, + "step": 6990 + }, + { + "epoch": 1.3590116984612397, + "grad_norm": 0.2855168581008911, + "learning_rate": 9.989079061244528e-05, + "loss": 0.0494, + "step": 7000 + }, + { + "epoch": 1.36095335177904, + "grad_norm": 0.2681066691875458, + "learning_rate": 9.988969564220769e-05, + "loss": 0.0543, + "step": 7010 + }, + { + "epoch": 1.36289500509684, + "grad_norm": 0.29007282853126526, + "learning_rate": 9.988859521610801e-05, + "loss": 0.0458, + "step": 7020 + }, + { + "epoch": 1.36483665841464, + "grad_norm": 0.2895333170890808, + "learning_rate": 9.988748933426656e-05, + "loss": 0.0405, + "step": 7030 + }, + { + "epoch": 1.3667783117324401, + "grad_norm": 0.230824813246727, + "learning_rate": 9.988637799680428e-05, + "loss": 0.0491, + "step": 7040 + }, + { + "epoch": 1.3687199650502402, + "grad_norm": 0.40402528643608093, + "learning_rate": 9.98852612038427e-05, + "loss": 0.0459, + "step": 7050 + }, + { + "epoch": 1.3706616183680405, + "grad_norm": 0.20465774834156036, + "learning_rate": 9.988413895550397e-05, + "loss": 0.0419, + "step": 7060 + }, + { + "epoch": 1.3726032716858405, + "grad_norm": 0.3367931544780731, + "learning_rate": 9.98830112519108e-05, + "loss": 0.0522, + "step": 7070 + }, + { + "epoch": 1.3745449250036406, + "grad_norm": 0.2537165880203247, + "learning_rate": 9.98818780931865e-05, + "loss": 0.0466, + "step": 7080 + }, + { + "epoch": 1.3764865783214408, + "grad_norm": 0.26882800459861755, + "learning_rate": 9.988073947945502e-05, + "loss": 0.0482, + "step": 7090 + }, + { + "epoch": 1.3784282316392409, + "grad_norm": 0.24720247089862823, + "learning_rate": 9.987959541084087e-05, + "loss": 0.0443, + "step": 7100 + }, + { + "epoch": 1.380369884957041, + "grad_norm": 0.30066820979118347, + "learning_rate": 9.987844588746915e-05, + "loss": 0.058, + "step": 7110 + }, + { + "epoch": 1.382311538274841, + "grad_norm": 0.39921247959136963, + "learning_rate": 9.987729090946558e-05, + "loss": 0.0464, + "step": 7120 + }, + { + "epoch": 1.384253191592641, + "grad_norm": 0.4219141900539398, + "learning_rate": 9.987613047695647e-05, + "loss": 0.0535, + "step": 7130 + }, + { + "epoch": 1.3861948449104413, + "grad_norm": 0.41199013590812683, + "learning_rate": 9.987496459006871e-05, + "loss": 0.0544, + "step": 7140 + }, + { + "epoch": 1.3881364982282414, + "grad_norm": 0.33056679368019104, + "learning_rate": 9.987379324892982e-05, + "loss": 0.047, + "step": 7150 + }, + { + "epoch": 1.3900781515460414, + "grad_norm": 0.29083383083343506, + "learning_rate": 9.987261645366788e-05, + "loss": 0.0565, + "step": 7160 + }, + { + "epoch": 1.3920198048638417, + "grad_norm": 0.32375991344451904, + "learning_rate": 9.987143420441158e-05, + "loss": 0.0465, + "step": 7170 + }, + { + "epoch": 1.3939614581816417, + "grad_norm": 0.2662351727485657, + "learning_rate": 9.987024650129022e-05, + "loss": 0.0508, + "step": 7180 + }, + { + "epoch": 1.3959031114994418, + "grad_norm": 0.24292324483394623, + "learning_rate": 9.986905334443368e-05, + "loss": 0.0442, + "step": 7190 + }, + { + "epoch": 1.3978447648172418, + "grad_norm": 0.5163930058479309, + "learning_rate": 9.986785473397245e-05, + "loss": 0.0493, + "step": 7200 + }, + { + "epoch": 1.399786418135042, + "grad_norm": 0.4221322536468506, + "learning_rate": 9.98666506700376e-05, + "loss": 0.0463, + "step": 7210 + }, + { + "epoch": 1.4017280714528422, + "grad_norm": 0.34347593784332275, + "learning_rate": 9.986544115276081e-05, + "loss": 0.0424, + "step": 7220 + }, + { + "epoch": 1.4036697247706422, + "grad_norm": 0.16198252141475677, + "learning_rate": 9.986422618227433e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 1.4056113780884423, + "grad_norm": 0.3649376928806305, + "learning_rate": 9.986300575871106e-05, + "loss": 0.0501, + "step": 7240 + }, + { + "epoch": 1.4075530314062425, + "grad_norm": 0.2516997456550598, + "learning_rate": 9.986177988220444e-05, + "loss": 0.0521, + "step": 7250 + }, + { + "epoch": 1.4094946847240426, + "grad_norm": 0.27607229351997375, + "learning_rate": 9.986054855288856e-05, + "loss": 0.06, + "step": 7260 + }, + { + "epoch": 1.4114363380418427, + "grad_norm": 0.311637818813324, + "learning_rate": 9.985931177089802e-05, + "loss": 0.0505, + "step": 7270 + }, + { + "epoch": 1.4133779913596427, + "grad_norm": 0.22994482517242432, + "learning_rate": 9.985806953636814e-05, + "loss": 0.0496, + "step": 7280 + }, + { + "epoch": 1.4153196446774428, + "grad_norm": 0.14897021651268005, + "learning_rate": 9.985682184943471e-05, + "loss": 0.0469, + "step": 7290 + }, + { + "epoch": 1.417261297995243, + "grad_norm": 0.21591924130916595, + "learning_rate": 9.98555687102342e-05, + "loss": 0.0473, + "step": 7300 + }, + { + "epoch": 1.419202951313043, + "grad_norm": 0.3226580023765564, + "learning_rate": 9.985431011890367e-05, + "loss": 0.0431, + "step": 7310 + }, + { + "epoch": 1.4211446046308431, + "grad_norm": 0.31766271591186523, + "learning_rate": 9.985304607558075e-05, + "loss": 0.04, + "step": 7320 + }, + { + "epoch": 1.4230862579486434, + "grad_norm": 0.5049781203269958, + "learning_rate": 9.985177658040364e-05, + "loss": 0.0504, + "step": 7330 + }, + { + "epoch": 1.4250279112664435, + "grad_norm": 0.21641704440116882, + "learning_rate": 9.985050163351119e-05, + "loss": 0.0418, + "step": 7340 + }, + { + "epoch": 1.4269695645842435, + "grad_norm": 0.31539732217788696, + "learning_rate": 9.984922123504286e-05, + "loss": 0.0482, + "step": 7350 + }, + { + "epoch": 1.4289112179020436, + "grad_norm": 0.3584599792957306, + "learning_rate": 9.984793538513862e-05, + "loss": 0.0541, + "step": 7360 + }, + { + "epoch": 1.4308528712198436, + "grad_norm": 0.3415973484516144, + "learning_rate": 9.984664408393912e-05, + "loss": 0.0556, + "step": 7370 + }, + { + "epoch": 1.4327945245376439, + "grad_norm": 0.2875714600086212, + "learning_rate": 9.984534733158556e-05, + "loss": 0.0498, + "step": 7380 + }, + { + "epoch": 1.434736177855444, + "grad_norm": 0.28123733401298523, + "learning_rate": 9.984404512821977e-05, + "loss": 0.0409, + "step": 7390 + }, + { + "epoch": 1.436677831173244, + "grad_norm": 0.25532445311546326, + "learning_rate": 9.984273747398411e-05, + "loss": 0.0396, + "step": 7400 + }, + { + "epoch": 1.438619484491044, + "grad_norm": 0.33937084674835205, + "learning_rate": 9.984142436902165e-05, + "loss": 0.0487, + "step": 7410 + }, + { + "epoch": 1.440561137808844, + "grad_norm": 0.35214921832084656, + "learning_rate": 9.984010581347596e-05, + "loss": 0.0557, + "step": 7420 + }, + { + "epoch": 1.4425027911266444, + "grad_norm": 0.26142412424087524, + "learning_rate": 9.983878180749121e-05, + "loss": 0.0558, + "step": 7430 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 0.34775134921073914, + "learning_rate": 9.983745235121222e-05, + "loss": 0.0499, + "step": 7440 + }, + { + "epoch": 1.4463860977622445, + "grad_norm": 0.23416826128959656, + "learning_rate": 9.983611744478438e-05, + "loss": 0.0468, + "step": 7450 + }, + { + "epoch": 1.4483277510800447, + "grad_norm": 0.2217777520418167, + "learning_rate": 9.983477708835365e-05, + "loss": 0.0429, + "step": 7460 + }, + { + "epoch": 1.4502694043978448, + "grad_norm": 0.26306039094924927, + "learning_rate": 9.983343128206664e-05, + "loss": 0.0502, + "step": 7470 + }, + { + "epoch": 1.4522110577156448, + "grad_norm": 0.41326019167900085, + "learning_rate": 9.983208002607049e-05, + "loss": 0.047, + "step": 7480 + }, + { + "epoch": 1.454152711033445, + "grad_norm": 0.20224930346012115, + "learning_rate": 9.9830723320513e-05, + "loss": 0.0471, + "step": 7490 + }, + { + "epoch": 1.456094364351245, + "grad_norm": 0.38472986221313477, + "learning_rate": 9.982936116554254e-05, + "loss": 0.0477, + "step": 7500 + }, + { + "epoch": 1.4580360176690452, + "grad_norm": 0.2969725430011749, + "learning_rate": 9.982799356130803e-05, + "loss": 0.0384, + "step": 7510 + }, + { + "epoch": 1.4599776709868453, + "grad_norm": 0.13692162930965424, + "learning_rate": 9.982662050795908e-05, + "loss": 0.0537, + "step": 7520 + }, + { + "epoch": 1.4619193243046453, + "grad_norm": 0.30071935057640076, + "learning_rate": 9.982524200564583e-05, + "loss": 0.0629, + "step": 7530 + }, + { + "epoch": 1.4638609776224456, + "grad_norm": 0.274595707654953, + "learning_rate": 9.982385805451901e-05, + "loss": 0.0518, + "step": 7540 + }, + { + "epoch": 1.4658026309402457, + "grad_norm": 0.27716532349586487, + "learning_rate": 9.982246865472998e-05, + "loss": 0.0513, + "step": 7550 + }, + { + "epoch": 1.4677442842580457, + "grad_norm": 0.24940265715122223, + "learning_rate": 9.982107380643069e-05, + "loss": 0.0426, + "step": 7560 + }, + { + "epoch": 1.4696859375758458, + "grad_norm": 0.2195909172296524, + "learning_rate": 9.981967350977368e-05, + "loss": 0.0428, + "step": 7570 + }, + { + "epoch": 1.4716275908936458, + "grad_norm": 0.2643185555934906, + "learning_rate": 9.981826776491208e-05, + "loss": 0.0489, + "step": 7580 + }, + { + "epoch": 1.473569244211446, + "grad_norm": 0.29168710112571716, + "learning_rate": 9.98168565719996e-05, + "loss": 0.0515, + "step": 7590 + }, + { + "epoch": 1.4755108975292461, + "grad_norm": 0.38002148270606995, + "learning_rate": 9.98154399311906e-05, + "loss": 0.0458, + "step": 7600 + }, + { + "epoch": 1.4774525508470462, + "grad_norm": 0.3003154397010803, + "learning_rate": 9.981401784263997e-05, + "loss": 0.0493, + "step": 7610 + }, + { + "epoch": 1.4793942041648465, + "grad_norm": 0.3133838474750519, + "learning_rate": 9.981259030650326e-05, + "loss": 0.0384, + "step": 7620 + }, + { + "epoch": 1.4813358574826465, + "grad_norm": 0.355437695980072, + "learning_rate": 9.981115732293655e-05, + "loss": 0.0562, + "step": 7630 + }, + { + "epoch": 1.4832775108004466, + "grad_norm": 0.36235520243644714, + "learning_rate": 9.980971889209659e-05, + "loss": 0.0498, + "step": 7640 + }, + { + "epoch": 1.4852191641182466, + "grad_norm": 0.3868720829486847, + "learning_rate": 9.980827501414064e-05, + "loss": 0.05, + "step": 7650 + }, + { + "epoch": 1.4871608174360467, + "grad_norm": 0.41807782649993896, + "learning_rate": 9.980682568922663e-05, + "loss": 0.0503, + "step": 7660 + }, + { + "epoch": 1.489102470753847, + "grad_norm": 0.3706916570663452, + "learning_rate": 9.980537091751304e-05, + "loss": 0.045, + "step": 7670 + }, + { + "epoch": 1.491044124071647, + "grad_norm": 0.2616509795188904, + "learning_rate": 9.980391069915897e-05, + "loss": 0.0431, + "step": 7680 + }, + { + "epoch": 1.492985777389447, + "grad_norm": 0.36757692694664, + "learning_rate": 9.98024450343241e-05, + "loss": 0.0523, + "step": 7690 + }, + { + "epoch": 1.4949274307072473, + "grad_norm": 0.3434084951877594, + "learning_rate": 9.980097392316872e-05, + "loss": 0.0391, + "step": 7700 + }, + { + "epoch": 1.4968690840250474, + "grad_norm": 0.3782077431678772, + "learning_rate": 9.97994973658537e-05, + "loss": 0.0429, + "step": 7710 + }, + { + "epoch": 1.4988107373428474, + "grad_norm": 0.3207756578922272, + "learning_rate": 9.979801536254054e-05, + "loss": 0.0579, + "step": 7720 + }, + { + "epoch": 1.5007523906606477, + "grad_norm": 0.3647724986076355, + "learning_rate": 9.979652791339127e-05, + "loss": 0.0454, + "step": 7730 + }, + { + "epoch": 1.5026940439784475, + "grad_norm": 0.4254285991191864, + "learning_rate": 9.97950350185686e-05, + "loss": 0.0475, + "step": 7740 + }, + { + "epoch": 1.5046356972962478, + "grad_norm": 0.31515267491340637, + "learning_rate": 9.979353667823574e-05, + "loss": 0.048, + "step": 7750 + }, + { + "epoch": 1.5065773506140478, + "grad_norm": 0.3024234473705292, + "learning_rate": 9.979203289255658e-05, + "loss": 0.0492, + "step": 7760 + }, + { + "epoch": 1.508519003931848, + "grad_norm": 0.289552241563797, + "learning_rate": 9.979052366169557e-05, + "loss": 0.0431, + "step": 7770 + }, + { + "epoch": 1.5104606572496482, + "grad_norm": 0.27821168303489685, + "learning_rate": 9.978900898581775e-05, + "loss": 0.0506, + "step": 7780 + }, + { + "epoch": 1.5124023105674482, + "grad_norm": 0.2701769769191742, + "learning_rate": 9.978748886508875e-05, + "loss": 0.0582, + "step": 7790 + }, + { + "epoch": 1.5143439638852483, + "grad_norm": 0.2785208523273468, + "learning_rate": 9.978596329967484e-05, + "loss": 0.0401, + "step": 7800 + }, + { + "epoch": 1.5162856172030486, + "grad_norm": 0.31619131565093994, + "learning_rate": 9.978443228974284e-05, + "loss": 0.0525, + "step": 7810 + }, + { + "epoch": 1.5182272705208484, + "grad_norm": 0.2795034348964691, + "learning_rate": 9.978289583546015e-05, + "loss": 0.0512, + "step": 7820 + }, + { + "epoch": 1.5201689238386487, + "grad_norm": 0.25353166460990906, + "learning_rate": 9.978135393699484e-05, + "loss": 0.0508, + "step": 7830 + }, + { + "epoch": 1.5221105771564487, + "grad_norm": 0.2138972282409668, + "learning_rate": 9.977980659451548e-05, + "loss": 0.0446, + "step": 7840 + }, + { + "epoch": 1.5240522304742488, + "grad_norm": 0.2248075008392334, + "learning_rate": 9.977825380819135e-05, + "loss": 0.0573, + "step": 7850 + }, + { + "epoch": 1.525993883792049, + "grad_norm": 0.25621968507766724, + "learning_rate": 9.97766955781922e-05, + "loss": 0.0444, + "step": 7860 + }, + { + "epoch": 1.527935537109849, + "grad_norm": 0.4114599823951721, + "learning_rate": 9.977513190468848e-05, + "loss": 0.0523, + "step": 7870 + }, + { + "epoch": 1.5298771904276491, + "grad_norm": 0.2718915045261383, + "learning_rate": 9.977356278785116e-05, + "loss": 0.047, + "step": 7880 + }, + { + "epoch": 1.5318188437454494, + "grad_norm": 0.23552611470222473, + "learning_rate": 9.977198822785184e-05, + "loss": 0.0512, + "step": 7890 + }, + { + "epoch": 1.5337604970632492, + "grad_norm": 0.31293052434921265, + "learning_rate": 9.977040822486273e-05, + "loss": 0.0527, + "step": 7900 + }, + { + "epoch": 1.5357021503810495, + "grad_norm": 0.29740461707115173, + "learning_rate": 9.97688227790566e-05, + "loss": 0.0478, + "step": 7910 + }, + { + "epoch": 1.5376438036988496, + "grad_norm": 0.24667952954769135, + "learning_rate": 9.976723189060684e-05, + "loss": 0.0479, + "step": 7920 + }, + { + "epoch": 1.5395854570166496, + "grad_norm": 0.20694930851459503, + "learning_rate": 9.976563555968742e-05, + "loss": 0.0522, + "step": 7930 + }, + { + "epoch": 1.54152711033445, + "grad_norm": 0.45407915115356445, + "learning_rate": 9.976403378647292e-05, + "loss": 0.053, + "step": 7940 + }, + { + "epoch": 1.54346876365225, + "grad_norm": 0.2972606420516968, + "learning_rate": 9.97624265711385e-05, + "loss": 0.0434, + "step": 7950 + }, + { + "epoch": 1.54541041697005, + "grad_norm": 0.29464834928512573, + "learning_rate": 9.976081391385993e-05, + "loss": 0.0405, + "step": 7960 + }, + { + "epoch": 1.5473520702878503, + "grad_norm": 0.3235883116722107, + "learning_rate": 9.975919581481356e-05, + "loss": 0.0474, + "step": 7970 + }, + { + "epoch": 1.54929372360565, + "grad_norm": 0.2207026183605194, + "learning_rate": 9.975757227417634e-05, + "loss": 0.044, + "step": 7980 + }, + { + "epoch": 1.5512353769234504, + "grad_norm": 0.17374907433986664, + "learning_rate": 9.975594329212586e-05, + "loss": 0.0477, + "step": 7990 + }, + { + "epoch": 1.5531770302412504, + "grad_norm": 0.3005909025669098, + "learning_rate": 9.97543088688402e-05, + "loss": 0.0444, + "step": 8000 + }, + { + "epoch": 1.5551186835590505, + "grad_norm": 0.3198379874229431, + "learning_rate": 9.975266900449814e-05, + "loss": 0.0472, + "step": 8010 + }, + { + "epoch": 1.5570603368768507, + "grad_norm": 0.22219808399677277, + "learning_rate": 9.975102369927898e-05, + "loss": 0.0446, + "step": 8020 + }, + { + "epoch": 1.5590019901946508, + "grad_norm": 0.24377594888210297, + "learning_rate": 9.974937295336269e-05, + "loss": 0.0533, + "step": 8030 + }, + { + "epoch": 1.5609436435124509, + "grad_norm": 0.27423810958862305, + "learning_rate": 9.974771676692975e-05, + "loss": 0.0467, + "step": 8040 + }, + { + "epoch": 1.5628852968302511, + "grad_norm": 0.2851283848285675, + "learning_rate": 9.974605514016131e-05, + "loss": 0.0478, + "step": 8050 + }, + { + "epoch": 1.564826950148051, + "grad_norm": 0.2361554503440857, + "learning_rate": 9.974438807323907e-05, + "loss": 0.0577, + "step": 8060 + }, + { + "epoch": 1.5667686034658512, + "grad_norm": 0.2627582550048828, + "learning_rate": 9.974271556634535e-05, + "loss": 0.0437, + "step": 8070 + }, + { + "epoch": 1.5687102567836513, + "grad_norm": 0.12920956313610077, + "learning_rate": 9.974103761966302e-05, + "loss": 0.0592, + "step": 8080 + }, + { + "epoch": 1.5706519101014513, + "grad_norm": 0.27044618129730225, + "learning_rate": 9.973935423337563e-05, + "loss": 0.0465, + "step": 8090 + }, + { + "epoch": 1.5725935634192516, + "grad_norm": 0.2814515233039856, + "learning_rate": 9.973766540766722e-05, + "loss": 0.0463, + "step": 8100 + }, + { + "epoch": 1.5745352167370514, + "grad_norm": 0.2698323726654053, + "learning_rate": 9.97359711427225e-05, + "loss": 0.051, + "step": 8110 + }, + { + "epoch": 1.5764768700548517, + "grad_norm": 0.25057846307754517, + "learning_rate": 9.973427143872677e-05, + "loss": 0.0546, + "step": 8120 + }, + { + "epoch": 1.5784185233726518, + "grad_norm": 0.21889521181583405, + "learning_rate": 9.973256629586589e-05, + "loss": 0.0499, + "step": 8130 + }, + { + "epoch": 1.5803601766904518, + "grad_norm": 0.2553783655166626, + "learning_rate": 9.973085571432632e-05, + "loss": 0.0484, + "step": 8140 + }, + { + "epoch": 1.582301830008252, + "grad_norm": 0.22009684145450592, + "learning_rate": 9.972913969429513e-05, + "loss": 0.0376, + "step": 8150 + }, + { + "epoch": 1.5842434833260521, + "grad_norm": 0.2461891621351242, + "learning_rate": 9.972741823596e-05, + "loss": 0.0407, + "step": 8160 + }, + { + "epoch": 1.5861851366438522, + "grad_norm": 0.37978681921958923, + "learning_rate": 9.972569133950917e-05, + "loss": 0.048, + "step": 8170 + }, + { + "epoch": 1.5881267899616525, + "grad_norm": 0.3486580550670624, + "learning_rate": 9.972395900513151e-05, + "loss": 0.0408, + "step": 8180 + }, + { + "epoch": 1.5900684432794523, + "grad_norm": 0.35139524936676025, + "learning_rate": 9.972222123301645e-05, + "loss": 0.0553, + "step": 8190 + }, + { + "epoch": 1.5920100965972526, + "grad_norm": 0.44965916872024536, + "learning_rate": 9.972047802335403e-05, + "loss": 0.0424, + "step": 8200 + }, + { + "epoch": 1.5939517499150526, + "grad_norm": 0.21362516283988953, + "learning_rate": 9.971872937633488e-05, + "loss": 0.0534, + "step": 8210 + }, + { + "epoch": 1.5958934032328527, + "grad_norm": 0.29159554839134216, + "learning_rate": 9.971697529215024e-05, + "loss": 0.0437, + "step": 8220 + }, + { + "epoch": 1.597835056550653, + "grad_norm": 0.2653329074382782, + "learning_rate": 9.971521577099192e-05, + "loss": 0.0428, + "step": 8230 + }, + { + "epoch": 1.599776709868453, + "grad_norm": 0.25692036747932434, + "learning_rate": 9.971345081305236e-05, + "loss": 0.0448, + "step": 8240 + }, + { + "epoch": 1.601718363186253, + "grad_norm": 0.2656397223472595, + "learning_rate": 9.971168041852456e-05, + "loss": 0.0459, + "step": 8250 + }, + { + "epoch": 1.6036600165040533, + "grad_norm": 0.330809086561203, + "learning_rate": 9.970990458760215e-05, + "loss": 0.0467, + "step": 8260 + }, + { + "epoch": 1.6056016698218532, + "grad_norm": 0.36970779299736023, + "learning_rate": 9.970812332047929e-05, + "loss": 0.0493, + "step": 8270 + }, + { + "epoch": 1.6075433231396534, + "grad_norm": 0.24826620519161224, + "learning_rate": 9.97063366173508e-05, + "loss": 0.0501, + "step": 8280 + }, + { + "epoch": 1.6094849764574535, + "grad_norm": 0.24621473252773285, + "learning_rate": 9.970454447841207e-05, + "loss": 0.0509, + "step": 8290 + }, + { + "epoch": 1.6114266297752535, + "grad_norm": 0.27184394001960754, + "learning_rate": 9.970274690385909e-05, + "loss": 0.0378, + "step": 8300 + }, + { + "epoch": 1.6133682830930538, + "grad_norm": 0.29942211508750916, + "learning_rate": 9.970094389388844e-05, + "loss": 0.042, + "step": 8310 + }, + { + "epoch": 1.6153099364108539, + "grad_norm": 0.4382030963897705, + "learning_rate": 9.969913544869728e-05, + "loss": 0.044, + "step": 8320 + }, + { + "epoch": 1.617251589728654, + "grad_norm": 0.3388069272041321, + "learning_rate": 9.96973215684834e-05, + "loss": 0.0564, + "step": 8330 + }, + { + "epoch": 1.6191932430464542, + "grad_norm": 0.19092261791229248, + "learning_rate": 9.969550225344513e-05, + "loss": 0.0422, + "step": 8340 + }, + { + "epoch": 1.621134896364254, + "grad_norm": 0.279037207365036, + "learning_rate": 9.969367750378147e-05, + "loss": 0.0533, + "step": 8350 + }, + { + "epoch": 1.6230765496820543, + "grad_norm": 0.23022933304309845, + "learning_rate": 9.969184731969194e-05, + "loss": 0.0566, + "step": 8360 + }, + { + "epoch": 1.6250182029998543, + "grad_norm": 0.3912797272205353, + "learning_rate": 9.96900117013767e-05, + "loss": 0.0498, + "step": 8370 + }, + { + "epoch": 1.6269598563176544, + "grad_norm": 0.31406596302986145, + "learning_rate": 9.96881706490365e-05, + "loss": 0.0483, + "step": 8380 + }, + { + "epoch": 1.6289015096354547, + "grad_norm": 0.3052789270877838, + "learning_rate": 9.968632416287265e-05, + "loss": 0.0501, + "step": 8390 + }, + { + "epoch": 1.6308431629532547, + "grad_norm": 0.4190371036529541, + "learning_rate": 9.96844722430871e-05, + "loss": 0.0429, + "step": 8400 + }, + { + "epoch": 1.6327848162710548, + "grad_norm": 0.35976722836494446, + "learning_rate": 9.968261488988235e-05, + "loss": 0.0381, + "step": 8410 + }, + { + "epoch": 1.634726469588855, + "grad_norm": 0.42740604281425476, + "learning_rate": 9.968075210346155e-05, + "loss": 0.0461, + "step": 8420 + }, + { + "epoch": 1.6366681229066549, + "grad_norm": 0.21536989510059357, + "learning_rate": 9.967888388402839e-05, + "loss": 0.0475, + "step": 8430 + }, + { + "epoch": 1.6386097762244551, + "grad_norm": 0.27007046341896057, + "learning_rate": 9.967701023178717e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 1.6405514295422552, + "grad_norm": 0.276721328496933, + "learning_rate": 9.967513114694282e-05, + "loss": 0.0437, + "step": 8450 + }, + { + "epoch": 1.6424930828600552, + "grad_norm": 0.2852634787559509, + "learning_rate": 9.967324662970079e-05, + "loss": 0.0396, + "step": 8460 + }, + { + "epoch": 1.6444347361778555, + "grad_norm": 0.2872546315193176, + "learning_rate": 9.96713566802672e-05, + "loss": 0.0466, + "step": 8470 + }, + { + "epoch": 1.6463763894956556, + "grad_norm": 0.22479747235774994, + "learning_rate": 9.966946129884873e-05, + "loss": 0.0527, + "step": 8480 + }, + { + "epoch": 1.6483180428134556, + "grad_norm": 0.33399245142936707, + "learning_rate": 9.966756048565265e-05, + "loss": 0.0496, + "step": 8490 + }, + { + "epoch": 1.650259696131256, + "grad_norm": 0.2832449674606323, + "learning_rate": 9.966565424088681e-05, + "loss": 0.0502, + "step": 8500 + }, + { + "epoch": 1.6522013494490557, + "grad_norm": 0.28387871384620667, + "learning_rate": 9.96637425647597e-05, + "loss": 0.0519, + "step": 8510 + }, + { + "epoch": 1.654143002766856, + "grad_norm": 0.17717161774635315, + "learning_rate": 9.966182545748038e-05, + "loss": 0.0381, + "step": 8520 + }, + { + "epoch": 1.656084656084656, + "grad_norm": 0.2033626139163971, + "learning_rate": 9.96599029192585e-05, + "loss": 0.0373, + "step": 8530 + }, + { + "epoch": 1.658026309402456, + "grad_norm": 0.2670786380767822, + "learning_rate": 9.965797495030428e-05, + "loss": 0.0422, + "step": 8540 + }, + { + "epoch": 1.6599679627202564, + "grad_norm": 0.31507688760757446, + "learning_rate": 9.96560415508286e-05, + "loss": 0.0455, + "step": 8550 + }, + { + "epoch": 1.6619096160380564, + "grad_norm": 0.24830733239650726, + "learning_rate": 9.965410272104286e-05, + "loss": 0.0455, + "step": 8560 + }, + { + "epoch": 1.6638512693558565, + "grad_norm": 0.2674705684185028, + "learning_rate": 9.96521584611591e-05, + "loss": 0.0438, + "step": 8570 + }, + { + "epoch": 1.6657929226736568, + "grad_norm": 0.2414357215166092, + "learning_rate": 9.965020877138994e-05, + "loss": 0.0451, + "step": 8580 + }, + { + "epoch": 1.6677345759914566, + "grad_norm": 0.18167322874069214, + "learning_rate": 9.964825365194861e-05, + "loss": 0.0491, + "step": 8590 + }, + { + "epoch": 1.6696762293092569, + "grad_norm": 0.47259241342544556, + "learning_rate": 9.96462931030489e-05, + "loss": 0.0492, + "step": 8600 + }, + { + "epoch": 1.671617882627057, + "grad_norm": 0.31719908118247986, + "learning_rate": 9.96443271249052e-05, + "loss": 0.0481, + "step": 8610 + }, + { + "epoch": 1.673559535944857, + "grad_norm": 0.22568240761756897, + "learning_rate": 9.964235571773255e-05, + "loss": 0.0401, + "step": 8620 + }, + { + "epoch": 1.6755011892626572, + "grad_norm": 0.29394763708114624, + "learning_rate": 9.96403788817465e-05, + "loss": 0.0418, + "step": 8630 + }, + { + "epoch": 1.6774428425804573, + "grad_norm": 0.2842504382133484, + "learning_rate": 9.963839661716325e-05, + "loss": 0.0342, + "step": 8640 + }, + { + "epoch": 1.6793844958982573, + "grad_norm": 0.3311135768890381, + "learning_rate": 9.963640892419958e-05, + "loss": 0.0488, + "step": 8650 + }, + { + "epoch": 1.6813261492160576, + "grad_norm": 0.2752250134944916, + "learning_rate": 9.963441580307286e-05, + "loss": 0.0407, + "step": 8660 + }, + { + "epoch": 1.6832678025338574, + "grad_norm": 0.38041964173316956, + "learning_rate": 9.963241725400104e-05, + "loss": 0.0516, + "step": 8670 + }, + { + "epoch": 1.6852094558516577, + "grad_norm": 0.22376437485218048, + "learning_rate": 9.963041327720271e-05, + "loss": 0.0461, + "step": 8680 + }, + { + "epoch": 1.6871511091694578, + "grad_norm": 0.23950783908367157, + "learning_rate": 9.962840387289697e-05, + "loss": 0.0474, + "step": 8690 + }, + { + "epoch": 1.6890927624872578, + "grad_norm": 0.2913685441017151, + "learning_rate": 9.962638904130363e-05, + "loss": 0.0414, + "step": 8700 + }, + { + "epoch": 1.691034415805058, + "grad_norm": 0.3036958575248718, + "learning_rate": 9.962436878264298e-05, + "loss": 0.0377, + "step": 8710 + }, + { + "epoch": 1.6929760691228581, + "grad_norm": 0.39152684807777405, + "learning_rate": 9.962234309713598e-05, + "loss": 0.0453, + "step": 8720 + }, + { + "epoch": 1.6949177224406582, + "grad_norm": 0.27375468611717224, + "learning_rate": 9.962031198500414e-05, + "loss": 0.0473, + "step": 8730 + }, + { + "epoch": 1.6968593757584585, + "grad_norm": 0.24594862759113312, + "learning_rate": 9.961827544646958e-05, + "loss": 0.043, + "step": 8740 + }, + { + "epoch": 1.6988010290762583, + "grad_norm": 0.2418200820684433, + "learning_rate": 9.961623348175501e-05, + "loss": 0.0424, + "step": 8750 + }, + { + "epoch": 1.7007426823940586, + "grad_norm": 0.3093269169330597, + "learning_rate": 9.961418609108377e-05, + "loss": 0.0462, + "step": 8760 + }, + { + "epoch": 1.7026843357118586, + "grad_norm": 0.19358143210411072, + "learning_rate": 9.961213327467971e-05, + "loss": 0.0504, + "step": 8770 + }, + { + "epoch": 1.7046259890296587, + "grad_norm": 0.18664048612117767, + "learning_rate": 9.961007503276736e-05, + "loss": 0.0456, + "step": 8780 + }, + { + "epoch": 1.706567642347459, + "grad_norm": 0.31241166591644287, + "learning_rate": 9.960801136557179e-05, + "loss": 0.0513, + "step": 8790 + }, + { + "epoch": 1.708509295665259, + "grad_norm": 0.33681991696357727, + "learning_rate": 9.960594227331866e-05, + "loss": 0.0583, + "step": 8800 + }, + { + "epoch": 1.710450948983059, + "grad_norm": 0.22571717202663422, + "learning_rate": 9.960386775623429e-05, + "loss": 0.0443, + "step": 8810 + }, + { + "epoch": 1.7123926023008593, + "grad_norm": 0.2903292775154114, + "learning_rate": 9.96017878145455e-05, + "loss": 0.0449, + "step": 8820 + }, + { + "epoch": 1.7143342556186592, + "grad_norm": 0.2517069876194, + "learning_rate": 9.959970244847977e-05, + "loss": 0.0383, + "step": 8830 + }, + { + "epoch": 1.7162759089364594, + "grad_norm": 0.44983378052711487, + "learning_rate": 9.959761165826518e-05, + "loss": 0.0536, + "step": 8840 + }, + { + "epoch": 1.7182175622542595, + "grad_norm": 0.3028988838195801, + "learning_rate": 9.959551544413033e-05, + "loss": 0.0481, + "step": 8850 + }, + { + "epoch": 1.7201592155720595, + "grad_norm": 0.33303210139274597, + "learning_rate": 9.959341380630448e-05, + "loss": 0.0446, + "step": 8860 + }, + { + "epoch": 1.7221008688898598, + "grad_norm": 0.3182477355003357, + "learning_rate": 9.959130674501746e-05, + "loss": 0.0466, + "step": 8870 + }, + { + "epoch": 1.7240425222076599, + "grad_norm": 0.30982697010040283, + "learning_rate": 9.958919426049968e-05, + "loss": 0.0577, + "step": 8880 + }, + { + "epoch": 1.72598417552546, + "grad_norm": 0.30847373604774475, + "learning_rate": 9.958707635298219e-05, + "loss": 0.0482, + "step": 8890 + }, + { + "epoch": 1.7279258288432602, + "grad_norm": 0.26117292046546936, + "learning_rate": 9.958495302269657e-05, + "loss": 0.0383, + "step": 8900 + }, + { + "epoch": 1.72986748216106, + "grad_norm": 0.22697702050209045, + "learning_rate": 9.958282426987503e-05, + "loss": 0.0423, + "step": 8910 + }, + { + "epoch": 1.7318091354788603, + "grad_norm": 0.26917964220046997, + "learning_rate": 9.95806900947504e-05, + "loss": 0.0496, + "step": 8920 + }, + { + "epoch": 1.7337507887966603, + "grad_norm": 0.2193775326013565, + "learning_rate": 9.957855049755604e-05, + "loss": 0.0424, + "step": 8930 + }, + { + "epoch": 1.7356924421144604, + "grad_norm": 0.20793071389198303, + "learning_rate": 9.957640547852593e-05, + "loss": 0.0455, + "step": 8940 + }, + { + "epoch": 1.7376340954322607, + "grad_norm": 0.2614475190639496, + "learning_rate": 9.957425503789466e-05, + "loss": 0.0482, + "step": 8950 + }, + { + "epoch": 1.7395757487500607, + "grad_norm": 0.2248305380344391, + "learning_rate": 9.957209917589738e-05, + "loss": 0.0439, + "step": 8960 + }, + { + "epoch": 1.7415174020678608, + "grad_norm": 0.2948211133480072, + "learning_rate": 9.956993789276987e-05, + "loss": 0.0453, + "step": 8970 + }, + { + "epoch": 1.743459055385661, + "grad_norm": 0.13918347656726837, + "learning_rate": 9.956777118874847e-05, + "loss": 0.0451, + "step": 8980 + }, + { + "epoch": 1.7454007087034609, + "grad_norm": 0.3427683413028717, + "learning_rate": 9.956559906407016e-05, + "loss": 0.0517, + "step": 8990 + }, + { + "epoch": 1.7473423620212611, + "grad_norm": 0.1856931447982788, + "learning_rate": 9.956342151897245e-05, + "loss": 0.0421, + "step": 9000 + }, + { + "epoch": 1.7492840153390612, + "grad_norm": 0.27691173553466797, + "learning_rate": 9.956123855369346e-05, + "loss": 0.0488, + "step": 9010 + }, + { + "epoch": 1.7512256686568612, + "grad_norm": 0.24733340740203857, + "learning_rate": 9.955905016847196e-05, + "loss": 0.0485, + "step": 9020 + }, + { + "epoch": 1.7531673219746615, + "grad_norm": 0.2732384502887726, + "learning_rate": 9.955685636354723e-05, + "loss": 0.0415, + "step": 9030 + }, + { + "epoch": 1.7551089752924616, + "grad_norm": 0.2108873873949051, + "learning_rate": 9.95546571391592e-05, + "loss": 0.0426, + "step": 9040 + }, + { + "epoch": 1.7570506286102616, + "grad_norm": 0.39726898074150085, + "learning_rate": 9.955245249554837e-05, + "loss": 0.0542, + "step": 9050 + }, + { + "epoch": 1.758992281928062, + "grad_norm": 0.23997622728347778, + "learning_rate": 9.955024243295582e-05, + "loss": 0.044, + "step": 9060 + }, + { + "epoch": 1.7609339352458617, + "grad_norm": 0.22440314292907715, + "learning_rate": 9.954802695162328e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 1.762875588563662, + "grad_norm": 0.25953614711761475, + "learning_rate": 9.954580605179302e-05, + "loss": 0.0443, + "step": 9080 + }, + { + "epoch": 1.764817241881462, + "grad_norm": 0.25798672437667847, + "learning_rate": 9.954357973370788e-05, + "loss": 0.0451, + "step": 9090 + }, + { + "epoch": 1.766758895199262, + "grad_norm": 0.27624213695526123, + "learning_rate": 9.954134799761135e-05, + "loss": 0.0352, + "step": 9100 + }, + { + "epoch": 1.7687005485170624, + "grad_norm": 0.28703567385673523, + "learning_rate": 9.953911084374748e-05, + "loss": 0.0396, + "step": 9110 + }, + { + "epoch": 1.7706422018348624, + "grad_norm": 0.22558733820915222, + "learning_rate": 9.953686827236093e-05, + "loss": 0.0449, + "step": 9120 + }, + { + "epoch": 1.7725838551526625, + "grad_norm": 0.2655683159828186, + "learning_rate": 9.953462028369695e-05, + "loss": 0.0528, + "step": 9130 + }, + { + "epoch": 1.7745255084704628, + "grad_norm": 0.27709054946899414, + "learning_rate": 9.953236687800136e-05, + "loss": 0.0459, + "step": 9140 + }, + { + "epoch": 1.7764671617882626, + "grad_norm": 0.25492197275161743, + "learning_rate": 9.95301080555206e-05, + "loss": 0.0412, + "step": 9150 + }, + { + "epoch": 1.7784088151060629, + "grad_norm": 0.24706290662288666, + "learning_rate": 9.952784381650171e-05, + "loss": 0.0397, + "step": 9160 + }, + { + "epoch": 1.780350468423863, + "grad_norm": 0.25753048062324524, + "learning_rate": 9.952557416119226e-05, + "loss": 0.0527, + "step": 9170 + }, + { + "epoch": 1.782292121741663, + "grad_norm": 0.27744144201278687, + "learning_rate": 9.95232990898405e-05, + "loss": 0.0482, + "step": 9180 + }, + { + "epoch": 1.7842337750594632, + "grad_norm": 0.3188321590423584, + "learning_rate": 9.95210186026952e-05, + "loss": 0.0398, + "step": 9190 + }, + { + "epoch": 1.7861754283772633, + "grad_norm": 0.20578499138355255, + "learning_rate": 9.951873270000576e-05, + "loss": 0.0434, + "step": 9200 + }, + { + "epoch": 1.7881170816950633, + "grad_norm": 0.49389204382896423, + "learning_rate": 9.951644138202216e-05, + "loss": 0.0503, + "step": 9210 + }, + { + "epoch": 1.7900587350128636, + "grad_norm": 0.27407166361808777, + "learning_rate": 9.951414464899498e-05, + "loss": 0.0411, + "step": 9220 + }, + { + "epoch": 1.7920003883306634, + "grad_norm": 0.299873024225235, + "learning_rate": 9.951184250117538e-05, + "loss": 0.041, + "step": 9230 + }, + { + "epoch": 1.7939420416484637, + "grad_norm": 0.41479483246803284, + "learning_rate": 9.950953493881513e-05, + "loss": 0.0605, + "step": 9240 + }, + { + "epoch": 1.7958836949662638, + "grad_norm": 0.25655660033226013, + "learning_rate": 9.950722196216658e-05, + "loss": 0.0489, + "step": 9250 + }, + { + "epoch": 1.7978253482840638, + "grad_norm": 0.35164138674736023, + "learning_rate": 9.950490357148265e-05, + "loss": 0.0481, + "step": 9260 + }, + { + "epoch": 1.799767001601864, + "grad_norm": 0.34397420287132263, + "learning_rate": 9.950257976701692e-05, + "loss": 0.0433, + "step": 9270 + }, + { + "epoch": 1.8017086549196641, + "grad_norm": 0.3747882544994354, + "learning_rate": 9.950025054902348e-05, + "loss": 0.0495, + "step": 9280 + }, + { + "epoch": 1.8036503082374642, + "grad_norm": 0.3046642243862152, + "learning_rate": 9.949791591775706e-05, + "loss": 0.0421, + "step": 9290 + }, + { + "epoch": 1.8055919615552645, + "grad_norm": 0.26132768392562866, + "learning_rate": 9.949557587347298e-05, + "loss": 0.0537, + "step": 9300 + }, + { + "epoch": 1.8075336148730643, + "grad_norm": 0.2656953036785126, + "learning_rate": 9.949323041642713e-05, + "loss": 0.046, + "step": 9310 + }, + { + "epoch": 1.8094752681908646, + "grad_norm": 0.2849544882774353, + "learning_rate": 9.949087954687602e-05, + "loss": 0.0508, + "step": 9320 + }, + { + "epoch": 1.8114169215086646, + "grad_norm": 0.37808534502983093, + "learning_rate": 9.948852326507672e-05, + "loss": 0.0476, + "step": 9330 + }, + { + "epoch": 1.8133585748264647, + "grad_norm": 0.47259023785591125, + "learning_rate": 9.948616157128694e-05, + "loss": 0.0485, + "step": 9340 + }, + { + "epoch": 1.815300228144265, + "grad_norm": 0.23166196048259735, + "learning_rate": 9.948379446576493e-05, + "loss": 0.0437, + "step": 9350 + }, + { + "epoch": 1.817241881462065, + "grad_norm": 0.2656733989715576, + "learning_rate": 9.948142194876952e-05, + "loss": 0.0434, + "step": 9360 + }, + { + "epoch": 1.819183534779865, + "grad_norm": 0.3313795030117035, + "learning_rate": 9.947904402056024e-05, + "loss": 0.0523, + "step": 9370 + }, + { + "epoch": 1.8211251880976653, + "grad_norm": 0.27454429864883423, + "learning_rate": 9.947666068139708e-05, + "loss": 0.0409, + "step": 9380 + }, + { + "epoch": 1.8230668414154652, + "grad_norm": 0.2654266357421875, + "learning_rate": 9.947427193154071e-05, + "loss": 0.0567, + "step": 9390 + }, + { + "epoch": 1.8250084947332654, + "grad_norm": 0.2985609173774719, + "learning_rate": 9.947187777125233e-05, + "loss": 0.0491, + "step": 9400 + }, + { + "epoch": 1.8269501480510655, + "grad_norm": 0.23097823560237885, + "learning_rate": 9.946947820079377e-05, + "loss": 0.0484, + "step": 9410 + }, + { + "epoch": 1.8288918013688655, + "grad_norm": 0.2626244127750397, + "learning_rate": 9.946707322042747e-05, + "loss": 0.0457, + "step": 9420 + }, + { + "epoch": 1.8308334546866658, + "grad_norm": 0.2743435800075531, + "learning_rate": 9.94646628304164e-05, + "loss": 0.046, + "step": 9430 + }, + { + "epoch": 1.8327751080044656, + "grad_norm": 0.2522735893726349, + "learning_rate": 9.946224703102418e-05, + "loss": 0.0456, + "step": 9440 + }, + { + "epoch": 1.834716761322266, + "grad_norm": 0.31839364767074585, + "learning_rate": 9.945982582251498e-05, + "loss": 0.0484, + "step": 9450 + }, + { + "epoch": 1.836658414640066, + "grad_norm": 0.2895512580871582, + "learning_rate": 9.94573992051536e-05, + "loss": 0.0402, + "step": 9460 + }, + { + "epoch": 1.838600067957866, + "grad_norm": 0.23363155126571655, + "learning_rate": 9.94549671792054e-05, + "loss": 0.0451, + "step": 9470 + }, + { + "epoch": 1.8405417212756663, + "grad_norm": 0.26422756910324097, + "learning_rate": 9.945252974493635e-05, + "loss": 0.0388, + "step": 9480 + }, + { + "epoch": 1.8424833745934663, + "grad_norm": 0.2326052188873291, + "learning_rate": 9.9450086902613e-05, + "loss": 0.0407, + "step": 9490 + }, + { + "epoch": 1.8444250279112664, + "grad_norm": 0.34842514991760254, + "learning_rate": 9.944763865250248e-05, + "loss": 0.0417, + "step": 9500 + }, + { + "epoch": 1.8463666812290667, + "grad_norm": 0.2553196847438812, + "learning_rate": 9.944518499487254e-05, + "loss": 0.0453, + "step": 9510 + }, + { + "epoch": 1.8483083345468665, + "grad_norm": 0.4429849088191986, + "learning_rate": 9.944272592999151e-05, + "loss": 0.0395, + "step": 9520 + }, + { + "epoch": 1.8502499878646668, + "grad_norm": 0.3085503578186035, + "learning_rate": 9.94402614581283e-05, + "loss": 0.0431, + "step": 9530 + }, + { + "epoch": 1.8521916411824668, + "grad_norm": 0.22932827472686768, + "learning_rate": 9.943779157955244e-05, + "loss": 0.0441, + "step": 9540 + }, + { + "epoch": 1.8541332945002669, + "grad_norm": 0.351565420627594, + "learning_rate": 9.943531629453403e-05, + "loss": 0.0552, + "step": 9550 + }, + { + "epoch": 1.8560749478180671, + "grad_norm": 0.6458670496940613, + "learning_rate": 9.943283560334375e-05, + "loss": 0.0493, + "step": 9560 + }, + { + "epoch": 1.8580166011358672, + "grad_norm": 0.33002936840057373, + "learning_rate": 9.943034950625288e-05, + "loss": 0.0445, + "step": 9570 + }, + { + "epoch": 1.8599582544536672, + "grad_norm": 0.23084765672683716, + "learning_rate": 9.942785800353332e-05, + "loss": 0.0366, + "step": 9580 + }, + { + "epoch": 1.8618999077714675, + "grad_norm": 0.32284659147262573, + "learning_rate": 9.942536109545751e-05, + "loss": 0.0399, + "step": 9590 + }, + { + "epoch": 1.8638415610892674, + "grad_norm": 0.42354267835617065, + "learning_rate": 9.942285878229853e-05, + "loss": 0.053, + "step": 9600 + }, + { + "epoch": 1.8657832144070676, + "grad_norm": 0.35208967328071594, + "learning_rate": 9.942035106433001e-05, + "loss": 0.0441, + "step": 9610 + }, + { + "epoch": 1.8677248677248677, + "grad_norm": 0.2465812712907791, + "learning_rate": 9.94178379418262e-05, + "loss": 0.0476, + "step": 9620 + }, + { + "epoch": 1.8696665210426677, + "grad_norm": 0.24252566695213318, + "learning_rate": 9.941531941506194e-05, + "loss": 0.0424, + "step": 9630 + }, + { + "epoch": 1.871608174360468, + "grad_norm": 0.2588130533695221, + "learning_rate": 9.941279548431263e-05, + "loss": 0.0362, + "step": 9640 + }, + { + "epoch": 1.873549827678268, + "grad_norm": 0.2700667977333069, + "learning_rate": 9.941026614985431e-05, + "loss": 0.0478, + "step": 9650 + }, + { + "epoch": 1.875491480996068, + "grad_norm": 0.3727455735206604, + "learning_rate": 9.940773141196357e-05, + "loss": 0.0454, + "step": 9660 + }, + { + "epoch": 1.8774331343138684, + "grad_norm": 0.2641751766204834, + "learning_rate": 9.94051912709176e-05, + "loss": 0.0553, + "step": 9670 + }, + { + "epoch": 1.8793747876316682, + "grad_norm": 0.21501190960407257, + "learning_rate": 9.940264572699421e-05, + "loss": 0.0432, + "step": 9680 + }, + { + "epoch": 1.8813164409494685, + "grad_norm": 0.4259840250015259, + "learning_rate": 9.940009478047174e-05, + "loss": 0.0409, + "step": 9690 + }, + { + "epoch": 1.8832580942672685, + "grad_norm": 0.1845087856054306, + "learning_rate": 9.939753843162918e-05, + "loss": 0.0452, + "step": 9700 + }, + { + "epoch": 1.8851997475850686, + "grad_norm": 0.1453196257352829, + "learning_rate": 9.939497668074609e-05, + "loss": 0.038, + "step": 9710 + }, + { + "epoch": 1.8871414009028689, + "grad_norm": 0.18985560536384583, + "learning_rate": 9.93924095281026e-05, + "loss": 0.0478, + "step": 9720 + }, + { + "epoch": 1.889083054220669, + "grad_norm": 0.21943743526935577, + "learning_rate": 9.938983697397948e-05, + "loss": 0.0379, + "step": 9730 + }, + { + "epoch": 1.891024707538469, + "grad_norm": 0.42362987995147705, + "learning_rate": 9.938725901865805e-05, + "loss": 0.0391, + "step": 9740 + }, + { + "epoch": 1.8929663608562692, + "grad_norm": 0.19339051842689514, + "learning_rate": 9.93846756624202e-05, + "loss": 0.046, + "step": 9750 + }, + { + "epoch": 1.894908014174069, + "grad_norm": 0.25969767570495605, + "learning_rate": 9.938208690554849e-05, + "loss": 0.0436, + "step": 9760 + }, + { + "epoch": 1.8968496674918693, + "grad_norm": 0.17927803099155426, + "learning_rate": 9.9379492748326e-05, + "loss": 0.0446, + "step": 9770 + }, + { + "epoch": 1.8987913208096694, + "grad_norm": 0.3193296790122986, + "learning_rate": 9.937689319103641e-05, + "loss": 0.041, + "step": 9780 + }, + { + "epoch": 1.9007329741274694, + "grad_norm": 0.17727085947990417, + "learning_rate": 9.937428823396404e-05, + "loss": 0.0502, + "step": 9790 + }, + { + "epoch": 1.9026746274452697, + "grad_norm": 0.1837790608406067, + "learning_rate": 9.937167787739372e-05, + "loss": 0.0399, + "step": 9800 + }, + { + "epoch": 1.9046162807630698, + "grad_norm": 0.20336100459098816, + "learning_rate": 9.936906212161095e-05, + "loss": 0.0394, + "step": 9810 + }, + { + "epoch": 1.9065579340808698, + "grad_norm": 0.24965570867061615, + "learning_rate": 9.936644096690176e-05, + "loss": 0.0469, + "step": 9820 + }, + { + "epoch": 1.90849958739867, + "grad_norm": 0.21099507808685303, + "learning_rate": 9.936381441355282e-05, + "loss": 0.0391, + "step": 9830 + }, + { + "epoch": 1.91044124071647, + "grad_norm": 0.26020294427871704, + "learning_rate": 9.936118246185136e-05, + "loss": 0.0385, + "step": 9840 + }, + { + "epoch": 1.9123828940342702, + "grad_norm": 0.3155662715435028, + "learning_rate": 9.935854511208518e-05, + "loss": 0.0433, + "step": 9850 + }, + { + "epoch": 1.9143245473520702, + "grad_norm": 0.383134126663208, + "learning_rate": 9.935590236454272e-05, + "loss": 0.0465, + "step": 9860 + }, + { + "epoch": 1.9162662006698703, + "grad_norm": 0.22494524717330933, + "learning_rate": 9.935325421951298e-05, + "loss": 0.0434, + "step": 9870 + }, + { + "epoch": 1.9182078539876706, + "grad_norm": 0.3928385376930237, + "learning_rate": 9.935060067728557e-05, + "loss": 0.0408, + "step": 9880 + }, + { + "epoch": 1.9201495073054706, + "grad_norm": 0.23127898573875427, + "learning_rate": 9.934794173815067e-05, + "loss": 0.042, + "step": 9890 + }, + { + "epoch": 1.9220911606232707, + "grad_norm": 0.3435990810394287, + "learning_rate": 9.934527740239906e-05, + "loss": 0.0531, + "step": 9900 + }, + { + "epoch": 1.924032813941071, + "grad_norm": 0.30788543820381165, + "learning_rate": 9.934260767032209e-05, + "loss": 0.0482, + "step": 9910 + }, + { + "epoch": 1.9259744672588708, + "grad_norm": 0.20864318311214447, + "learning_rate": 9.933993254221172e-05, + "loss": 0.0399, + "step": 9920 + }, + { + "epoch": 1.927916120576671, + "grad_norm": 0.3732742369174957, + "learning_rate": 9.933725201836053e-05, + "loss": 0.047, + "step": 9930 + }, + { + "epoch": 1.929857773894471, + "grad_norm": 0.3932124078273773, + "learning_rate": 9.933456609906162e-05, + "loss": 0.054, + "step": 9940 + }, + { + "epoch": 1.9317994272122712, + "grad_norm": 0.18438509106636047, + "learning_rate": 9.933187478460875e-05, + "loss": 0.0504, + "step": 9950 + }, + { + "epoch": 1.9337410805300714, + "grad_norm": 0.33035174012184143, + "learning_rate": 9.93291780752962e-05, + "loss": 0.0448, + "step": 9960 + }, + { + "epoch": 1.9356827338478715, + "grad_norm": 0.33518892526626587, + "learning_rate": 9.932647597141893e-05, + "loss": 0.0417, + "step": 9970 + }, + { + "epoch": 1.9376243871656715, + "grad_norm": 0.24700117111206055, + "learning_rate": 9.932376847327239e-05, + "loss": 0.0485, + "step": 9980 + }, + { + "epoch": 1.9395660404834718, + "grad_norm": 0.20159848034381866, + "learning_rate": 9.932105558115268e-05, + "loss": 0.0422, + "step": 9990 + }, + { + "epoch": 1.9415076938012716, + "grad_norm": 0.3000229299068451, + "learning_rate": 9.931833729535651e-05, + "loss": 0.0431, + "step": 10000 + }, + { + "epoch": 1.943449347119072, + "grad_norm": 0.2926139831542969, + "learning_rate": 9.931561361618111e-05, + "loss": 0.0429, + "step": 10010 + }, + { + "epoch": 1.945391000436872, + "grad_norm": 0.2552880346775055, + "learning_rate": 9.931288454392435e-05, + "loss": 0.0565, + "step": 10020 + }, + { + "epoch": 1.947332653754672, + "grad_norm": 0.23960722982883453, + "learning_rate": 9.931015007888467e-05, + "loss": 0.051, + "step": 10030 + }, + { + "epoch": 1.9492743070724723, + "grad_norm": 0.2274731695652008, + "learning_rate": 9.930741022136112e-05, + "loss": 0.0418, + "step": 10040 + }, + { + "epoch": 1.9512159603902723, + "grad_norm": 0.33701178431510925, + "learning_rate": 9.930466497165333e-05, + "loss": 0.0499, + "step": 10050 + }, + { + "epoch": 1.9531576137080724, + "grad_norm": 0.21386152505874634, + "learning_rate": 9.93019143300615e-05, + "loss": 0.0413, + "step": 10060 + }, + { + "epoch": 1.9550992670258727, + "grad_norm": 0.3176353871822357, + "learning_rate": 9.929915829688644e-05, + "loss": 0.0417, + "step": 10070 + }, + { + "epoch": 1.9570409203436725, + "grad_norm": 0.2869899272918701, + "learning_rate": 9.929639687242955e-05, + "loss": 0.0434, + "step": 10080 + }, + { + "epoch": 1.9589825736614728, + "grad_norm": 0.24282768368721008, + "learning_rate": 9.929363005699281e-05, + "loss": 0.0408, + "step": 10090 + }, + { + "epoch": 1.9609242269792728, + "grad_norm": 0.23872773349285126, + "learning_rate": 9.92908578508788e-05, + "loss": 0.0402, + "step": 10100 + }, + { + "epoch": 1.9628658802970729, + "grad_norm": 0.22120344638824463, + "learning_rate": 9.928808025439069e-05, + "loss": 0.0464, + "step": 10110 + }, + { + "epoch": 1.9648075336148731, + "grad_norm": 0.21929281949996948, + "learning_rate": 9.928529726783223e-05, + "loss": 0.0448, + "step": 10120 + }, + { + "epoch": 1.9667491869326732, + "grad_norm": 0.18634839355945587, + "learning_rate": 9.928250889150774e-05, + "loss": 0.0341, + "step": 10130 + }, + { + "epoch": 1.9686908402504733, + "grad_norm": 0.3976479470729828, + "learning_rate": 9.92797151257222e-05, + "loss": 0.0374, + "step": 10140 + }, + { + "epoch": 1.9706324935682735, + "grad_norm": 0.28064242005348206, + "learning_rate": 9.927691597078108e-05, + "loss": 0.0485, + "step": 10150 + }, + { + "epoch": 1.9725741468860734, + "grad_norm": 0.2537800073623657, + "learning_rate": 9.927411142699053e-05, + "loss": 0.0437, + "step": 10160 + }, + { + "epoch": 1.9745158002038736, + "grad_norm": 0.17910411953926086, + "learning_rate": 9.927130149465725e-05, + "loss": 0.0453, + "step": 10170 + }, + { + "epoch": 1.9764574535216737, + "grad_norm": 0.1595236361026764, + "learning_rate": 9.92684861740885e-05, + "loss": 0.0462, + "step": 10180 + }, + { + "epoch": 1.9783991068394737, + "grad_norm": 0.4039172828197479, + "learning_rate": 9.926566546559217e-05, + "loss": 0.0496, + "step": 10190 + }, + { + "epoch": 1.980340760157274, + "grad_norm": 0.28875482082366943, + "learning_rate": 9.926283936947673e-05, + "loss": 0.0504, + "step": 10200 + }, + { + "epoch": 1.982282413475074, + "grad_norm": 0.190827414393425, + "learning_rate": 9.926000788605126e-05, + "loss": 0.0387, + "step": 10210 + }, + { + "epoch": 1.984224066792874, + "grad_norm": 0.24030345678329468, + "learning_rate": 9.92571710156254e-05, + "loss": 0.0434, + "step": 10220 + }, + { + "epoch": 1.9861657201106744, + "grad_norm": 0.30545976758003235, + "learning_rate": 9.925432875850936e-05, + "loss": 0.056, + "step": 10230 + }, + { + "epoch": 1.9881073734284742, + "grad_norm": 0.24663668870925903, + "learning_rate": 9.925148111501396e-05, + "loss": 0.0458, + "step": 10240 + }, + { + "epoch": 1.9900490267462745, + "grad_norm": 0.21366600692272186, + "learning_rate": 9.924862808545066e-05, + "loss": 0.0318, + "step": 10250 + }, + { + "epoch": 1.9919906800640745, + "grad_norm": 0.2915486693382263, + "learning_rate": 9.924576967013141e-05, + "loss": 0.0389, + "step": 10260 + }, + { + "epoch": 1.9939323333818746, + "grad_norm": 0.2567528188228607, + "learning_rate": 9.924290586936887e-05, + "loss": 0.0401, + "step": 10270 + }, + { + "epoch": 1.9958739866996749, + "grad_norm": 0.21560360491275787, + "learning_rate": 9.924003668347614e-05, + "loss": 0.0514, + "step": 10280 + }, + { + "epoch": 1.997815640017475, + "grad_norm": 0.16050368547439575, + "learning_rate": 9.923716211276704e-05, + "loss": 0.0421, + "step": 10290 + }, + { + "epoch": 1.999757293335275, + "grad_norm": 0.2968396544456482, + "learning_rate": 9.923428215755594e-05, + "loss": 0.0448, + "step": 10300 + }, + { + "epoch": 2.00155332265424, + "grad_norm": 0.1968073844909668, + "learning_rate": 9.923139681815775e-05, + "loss": 0.0363, + "step": 10310 + }, + { + "epoch": 2.0034949759720404, + "grad_norm": 0.3705418109893799, + "learning_rate": 9.922850609488801e-05, + "loss": 0.0462, + "step": 10320 + }, + { + "epoch": 2.0054366292898402, + "grad_norm": 0.40342727303504944, + "learning_rate": 9.922560998806287e-05, + "loss": 0.0348, + "step": 10330 + }, + { + "epoch": 2.0073782826076405, + "grad_norm": 0.2272331714630127, + "learning_rate": 9.922270849799905e-05, + "loss": 0.0363, + "step": 10340 + }, + { + "epoch": 2.0093199359254403, + "grad_norm": 0.25110161304473877, + "learning_rate": 9.92198016250138e-05, + "loss": 0.0366, + "step": 10350 + }, + { + "epoch": 2.0112615892432406, + "grad_norm": 0.24650166928768158, + "learning_rate": 9.921688936942506e-05, + "loss": 0.0394, + "step": 10360 + }, + { + "epoch": 2.013203242561041, + "grad_norm": 0.26365771889686584, + "learning_rate": 9.921397173155129e-05, + "loss": 0.0465, + "step": 10370 + }, + { + "epoch": 2.0151448958788407, + "grad_norm": 0.3948422372341156, + "learning_rate": 9.921104871171157e-05, + "loss": 0.0444, + "step": 10380 + }, + { + "epoch": 2.017086549196641, + "grad_norm": 0.23146075010299683, + "learning_rate": 9.920812031022554e-05, + "loss": 0.0519, + "step": 10390 + }, + { + "epoch": 2.0190282025144413, + "grad_norm": 0.2763524055480957, + "learning_rate": 9.920518652741348e-05, + "loss": 0.0401, + "step": 10400 + }, + { + "epoch": 2.020969855832241, + "grad_norm": 0.15341953933238983, + "learning_rate": 9.920224736359618e-05, + "loss": 0.0406, + "step": 10410 + }, + { + "epoch": 2.0229115091500414, + "grad_norm": 0.2498539388179779, + "learning_rate": 9.91993028190951e-05, + "loss": 0.0434, + "step": 10420 + }, + { + "epoch": 2.024853162467841, + "grad_norm": 0.14109106361865997, + "learning_rate": 9.919635289423222e-05, + "loss": 0.0396, + "step": 10430 + }, + { + "epoch": 2.0267948157856415, + "grad_norm": 0.31083834171295166, + "learning_rate": 9.919339758933015e-05, + "loss": 0.0413, + "step": 10440 + }, + { + "epoch": 2.0287364691034417, + "grad_norm": 0.1933247447013855, + "learning_rate": 9.919043690471209e-05, + "loss": 0.039, + "step": 10450 + }, + { + "epoch": 2.0306781224212416, + "grad_norm": 0.28862932324409485, + "learning_rate": 9.91874708407018e-05, + "loss": 0.034, + "step": 10460 + }, + { + "epoch": 2.032619775739042, + "grad_norm": 0.3081539273262024, + "learning_rate": 9.918449939762367e-05, + "loss": 0.035, + "step": 10470 + }, + { + "epoch": 2.034561429056842, + "grad_norm": 0.468769907951355, + "learning_rate": 9.91815225758026e-05, + "loss": 0.0443, + "step": 10480 + }, + { + "epoch": 2.036503082374642, + "grad_norm": 0.3022797405719757, + "learning_rate": 9.917854037556419e-05, + "loss": 0.0344, + "step": 10490 + }, + { + "epoch": 2.0384447356924422, + "grad_norm": 0.30061620473861694, + "learning_rate": 9.917555279723454e-05, + "loss": 0.0364, + "step": 10500 + }, + { + "epoch": 2.040386389010242, + "grad_norm": 0.1919950693845749, + "learning_rate": 9.917255984114036e-05, + "loss": 0.0513, + "step": 10510 + }, + { + "epoch": 2.0423280423280423, + "grad_norm": 0.3250443935394287, + "learning_rate": 9.916956150760896e-05, + "loss": 0.041, + "step": 10520 + }, + { + "epoch": 2.0442696956458426, + "grad_norm": 0.21310432255268097, + "learning_rate": 9.916655779696826e-05, + "loss": 0.0429, + "step": 10530 + }, + { + "epoch": 2.0462113489636424, + "grad_norm": 0.2674300968647003, + "learning_rate": 9.916354870954671e-05, + "loss": 0.0422, + "step": 10540 + }, + { + "epoch": 2.0481530022814427, + "grad_norm": 0.20518654584884644, + "learning_rate": 9.91605342456734e-05, + "loss": 0.0445, + "step": 10550 + }, + { + "epoch": 2.050094655599243, + "grad_norm": 0.3450964391231537, + "learning_rate": 9.915751440567795e-05, + "loss": 0.0441, + "step": 10560 + }, + { + "epoch": 2.052036308917043, + "grad_norm": 0.3273125886917114, + "learning_rate": 9.915448918989066e-05, + "loss": 0.0468, + "step": 10570 + }, + { + "epoch": 2.053977962234843, + "grad_norm": 0.2822215259075165, + "learning_rate": 9.915145859864232e-05, + "loss": 0.044, + "step": 10580 + }, + { + "epoch": 2.055919615552643, + "grad_norm": 0.32174837589263916, + "learning_rate": 9.914842263226437e-05, + "loss": 0.0437, + "step": 10590 + }, + { + "epoch": 2.057861268870443, + "grad_norm": 0.1809709370136261, + "learning_rate": 9.914538129108882e-05, + "loss": 0.0419, + "step": 10600 + }, + { + "epoch": 2.0598029221882435, + "grad_norm": 0.17803829908370972, + "learning_rate": 9.914233457544825e-05, + "loss": 0.0401, + "step": 10610 + }, + { + "epoch": 2.0617445755060433, + "grad_norm": 0.30559059977531433, + "learning_rate": 9.913928248567586e-05, + "loss": 0.0454, + "step": 10620 + }, + { + "epoch": 2.0636862288238436, + "grad_norm": 0.2686355412006378, + "learning_rate": 9.913622502210542e-05, + "loss": 0.0449, + "step": 10630 + }, + { + "epoch": 2.0656278821416434, + "grad_norm": 0.2852998673915863, + "learning_rate": 9.913316218507128e-05, + "loss": 0.0378, + "step": 10640 + }, + { + "epoch": 2.0675695354594437, + "grad_norm": 0.19581831991672516, + "learning_rate": 9.91300939749084e-05, + "loss": 0.0408, + "step": 10650 + }, + { + "epoch": 2.069511188777244, + "grad_norm": 0.18454673886299133, + "learning_rate": 9.91270203919523e-05, + "loss": 0.0385, + "step": 10660 + }, + { + "epoch": 2.0714528420950438, + "grad_norm": 0.29719117283821106, + "learning_rate": 9.912394143653912e-05, + "loss": 0.0468, + "step": 10670 + }, + { + "epoch": 2.073394495412844, + "grad_norm": 0.14797507226467133, + "learning_rate": 9.912085710900555e-05, + "loss": 0.0373, + "step": 10680 + }, + { + "epoch": 2.0753361487306443, + "grad_norm": 0.299151748418808, + "learning_rate": 9.911776740968892e-05, + "loss": 0.0441, + "step": 10690 + }, + { + "epoch": 2.077277802048444, + "grad_norm": 0.22025328874588013, + "learning_rate": 9.911467233892709e-05, + "loss": 0.0411, + "step": 10700 + }, + { + "epoch": 2.0792194553662444, + "grad_norm": 0.28312692046165466, + "learning_rate": 9.911157189705853e-05, + "loss": 0.0482, + "step": 10710 + }, + { + "epoch": 2.0811611086840447, + "grad_norm": 0.2168707400560379, + "learning_rate": 9.910846608442229e-05, + "loss": 0.0452, + "step": 10720 + }, + { + "epoch": 2.0831027620018445, + "grad_norm": 0.29155755043029785, + "learning_rate": 9.910535490135805e-05, + "loss": 0.0507, + "step": 10730 + }, + { + "epoch": 2.085044415319645, + "grad_norm": 0.2412203699350357, + "learning_rate": 9.910223834820603e-05, + "loss": 0.0414, + "step": 10740 + }, + { + "epoch": 2.0869860686374446, + "grad_norm": 0.26797300577163696, + "learning_rate": 9.909911642530703e-05, + "loss": 0.0392, + "step": 10750 + }, + { + "epoch": 2.088927721955245, + "grad_norm": 0.22448930144309998, + "learning_rate": 9.909598913300249e-05, + "loss": 0.0381, + "step": 10760 + }, + { + "epoch": 2.090869375273045, + "grad_norm": 0.2130109667778015, + "learning_rate": 9.909285647163438e-05, + "loss": 0.0421, + "step": 10770 + }, + { + "epoch": 2.092811028590845, + "grad_norm": 0.30457803606987, + "learning_rate": 9.908971844154531e-05, + "loss": 0.045, + "step": 10780 + }, + { + "epoch": 2.0947526819086453, + "grad_norm": 0.3146781325340271, + "learning_rate": 9.908657504307843e-05, + "loss": 0.0398, + "step": 10790 + }, + { + "epoch": 2.096694335226445, + "grad_norm": 0.25516200065612793, + "learning_rate": 9.908342627657751e-05, + "loss": 0.0422, + "step": 10800 + }, + { + "epoch": 2.0986359885442454, + "grad_norm": 0.15928860008716583, + "learning_rate": 9.908027214238689e-05, + "loss": 0.041, + "step": 10810 + }, + { + "epoch": 2.1005776418620457, + "grad_norm": 0.29286083579063416, + "learning_rate": 9.90771126408515e-05, + "loss": 0.0437, + "step": 10820 + }, + { + "epoch": 2.1025192951798455, + "grad_norm": 0.23694442212581635, + "learning_rate": 9.907394777231685e-05, + "loss": 0.0404, + "step": 10830 + }, + { + "epoch": 2.1044609484976458, + "grad_norm": 0.30608493089675903, + "learning_rate": 9.907077753712905e-05, + "loss": 0.0386, + "step": 10840 + }, + { + "epoch": 2.106402601815446, + "grad_norm": 0.2294103056192398, + "learning_rate": 9.906760193563482e-05, + "loss": 0.0419, + "step": 10850 + }, + { + "epoch": 2.108344255133246, + "grad_norm": 0.2824893593788147, + "learning_rate": 9.906442096818139e-05, + "loss": 0.0378, + "step": 10860 + }, + { + "epoch": 2.110285908451046, + "grad_norm": 0.29979661107063293, + "learning_rate": 9.906123463511665e-05, + "loss": 0.0381, + "step": 10870 + }, + { + "epoch": 2.112227561768846, + "grad_norm": 0.3301090896129608, + "learning_rate": 9.905804293678907e-05, + "loss": 0.0463, + "step": 10880 + }, + { + "epoch": 2.1141692150866462, + "grad_norm": 0.17310984432697296, + "learning_rate": 9.905484587354766e-05, + "loss": 0.0442, + "step": 10890 + }, + { + "epoch": 2.1161108684044465, + "grad_norm": 0.24093849956989288, + "learning_rate": 9.905164344574205e-05, + "loss": 0.0487, + "step": 10900 + }, + { + "epoch": 2.1180525217222463, + "grad_norm": 0.24934840202331543, + "learning_rate": 9.904843565372248e-05, + "loss": 0.0411, + "step": 10910 + }, + { + "epoch": 2.1199941750400466, + "grad_norm": 0.1896219402551651, + "learning_rate": 9.904522249783972e-05, + "loss": 0.0402, + "step": 10920 + }, + { + "epoch": 2.121935828357847, + "grad_norm": 0.2389296293258667, + "learning_rate": 9.904200397844517e-05, + "loss": 0.0392, + "step": 10930 + }, + { + "epoch": 2.1238774816756467, + "grad_norm": 0.25280627608299255, + "learning_rate": 9.903878009589078e-05, + "loss": 0.0434, + "step": 10940 + }, + { + "epoch": 2.125819134993447, + "grad_norm": 0.22629867494106293, + "learning_rate": 9.903555085052915e-05, + "loss": 0.0378, + "step": 10950 + }, + { + "epoch": 2.127760788311247, + "grad_norm": 0.21695594489574432, + "learning_rate": 9.903231624271338e-05, + "loss": 0.0406, + "step": 10960 + }, + { + "epoch": 2.129702441629047, + "grad_norm": 0.24524284899234772, + "learning_rate": 9.902907627279724e-05, + "loss": 0.0375, + "step": 10970 + }, + { + "epoch": 2.1316440949468474, + "grad_norm": 0.23372936248779297, + "learning_rate": 9.902583094113504e-05, + "loss": 0.0373, + "step": 10980 + }, + { + "epoch": 2.133585748264647, + "grad_norm": 0.3012290596961975, + "learning_rate": 9.902258024808168e-05, + "loss": 0.0375, + "step": 10990 + }, + { + "epoch": 2.1355274015824475, + "grad_norm": 0.3611435890197754, + "learning_rate": 9.901932419399264e-05, + "loss": 0.0422, + "step": 11000 + }, + { + "epoch": 2.1374690549002477, + "grad_norm": 0.2431531846523285, + "learning_rate": 9.9016062779224e-05, + "loss": 0.0357, + "step": 11010 + }, + { + "epoch": 2.1394107082180476, + "grad_norm": 0.2769901752471924, + "learning_rate": 9.901279600413242e-05, + "loss": 0.0401, + "step": 11020 + }, + { + "epoch": 2.141352361535848, + "grad_norm": 0.33812224864959717, + "learning_rate": 9.900952386907518e-05, + "loss": 0.0399, + "step": 11030 + }, + { + "epoch": 2.1432940148536477, + "grad_norm": 0.22572959959506989, + "learning_rate": 9.90062463744101e-05, + "loss": 0.0385, + "step": 11040 + }, + { + "epoch": 2.145235668171448, + "grad_norm": 0.1554115265607834, + "learning_rate": 9.900296352049558e-05, + "loss": 0.0448, + "step": 11050 + }, + { + "epoch": 2.1471773214892482, + "grad_norm": 0.2920364737510681, + "learning_rate": 9.899967530769065e-05, + "loss": 0.0554, + "step": 11060 + }, + { + "epoch": 2.149118974807048, + "grad_norm": 0.4229273498058319, + "learning_rate": 9.899638173635489e-05, + "loss": 0.0432, + "step": 11070 + }, + { + "epoch": 2.1510606281248483, + "grad_norm": 0.24733422696590424, + "learning_rate": 9.899308280684849e-05, + "loss": 0.0359, + "step": 11080 + }, + { + "epoch": 2.1530022814426486, + "grad_norm": 0.2684401273727417, + "learning_rate": 9.898977851953222e-05, + "loss": 0.0355, + "step": 11090 + }, + { + "epoch": 2.1549439347604484, + "grad_norm": 0.3648720979690552, + "learning_rate": 9.898646887476741e-05, + "loss": 0.044, + "step": 11100 + }, + { + "epoch": 2.1568855880782487, + "grad_norm": 0.2667017877101898, + "learning_rate": 9.898315387291603e-05, + "loss": 0.0434, + "step": 11110 + }, + { + "epoch": 2.1588272413960485, + "grad_norm": 0.2509898543357849, + "learning_rate": 9.89798335143406e-05, + "loss": 0.037, + "step": 11120 + }, + { + "epoch": 2.160768894713849, + "grad_norm": 0.27130696177482605, + "learning_rate": 9.897650779940419e-05, + "loss": 0.0364, + "step": 11130 + }, + { + "epoch": 2.162710548031649, + "grad_norm": 0.19049124419689178, + "learning_rate": 9.897317672847054e-05, + "loss": 0.0396, + "step": 11140 + }, + { + "epoch": 2.164652201349449, + "grad_norm": 0.16747266054153442, + "learning_rate": 9.89698403019039e-05, + "loss": 0.0441, + "step": 11150 + }, + { + "epoch": 2.166593854667249, + "grad_norm": 0.2660399079322815, + "learning_rate": 9.896649852006917e-05, + "loss": 0.0414, + "step": 11160 + }, + { + "epoch": 2.1685355079850495, + "grad_norm": 0.4051297903060913, + "learning_rate": 9.896315138333177e-05, + "loss": 0.0457, + "step": 11170 + }, + { + "epoch": 2.1704771613028493, + "grad_norm": 0.22890649735927582, + "learning_rate": 9.895979889205774e-05, + "loss": 0.0415, + "step": 11180 + }, + { + "epoch": 2.1724188146206496, + "grad_norm": 0.24290722608566284, + "learning_rate": 9.895644104661372e-05, + "loss": 0.0522, + "step": 11190 + }, + { + "epoch": 2.1743604679384494, + "grad_norm": 0.2674873173236847, + "learning_rate": 9.895307784736691e-05, + "loss": 0.036, + "step": 11200 + }, + { + "epoch": 2.1763021212562497, + "grad_norm": 0.25993552803993225, + "learning_rate": 9.894970929468512e-05, + "loss": 0.0555, + "step": 11210 + }, + { + "epoch": 2.17824377457405, + "grad_norm": 0.2552343010902405, + "learning_rate": 9.89463353889367e-05, + "loss": 0.0389, + "step": 11220 + }, + { + "epoch": 2.1801854278918498, + "grad_norm": 0.1922900378704071, + "learning_rate": 9.894295613049065e-05, + "loss": 0.0437, + "step": 11230 + }, + { + "epoch": 2.18212708120965, + "grad_norm": 0.182949960231781, + "learning_rate": 9.893957151971649e-05, + "loss": 0.0416, + "step": 11240 + }, + { + "epoch": 2.1840687345274503, + "grad_norm": 0.2995719909667969, + "learning_rate": 9.893618155698436e-05, + "loss": 0.036, + "step": 11250 + }, + { + "epoch": 2.18601038784525, + "grad_norm": 0.2862603962421417, + "learning_rate": 9.8932786242665e-05, + "loss": 0.039, + "step": 11260 + }, + { + "epoch": 2.1879520411630504, + "grad_norm": 0.14827537536621094, + "learning_rate": 9.89293855771297e-05, + "loss": 0.0347, + "step": 11270 + }, + { + "epoch": 2.1898936944808503, + "grad_norm": 0.291908323764801, + "learning_rate": 9.892597956075036e-05, + "loss": 0.0355, + "step": 11280 + }, + { + "epoch": 2.1918353477986505, + "grad_norm": 0.23681434988975525, + "learning_rate": 9.892256819389947e-05, + "loss": 0.0422, + "step": 11290 + }, + { + "epoch": 2.193777001116451, + "grad_norm": 0.21662963926792145, + "learning_rate": 9.891915147695006e-05, + "loss": 0.0405, + "step": 11300 + }, + { + "epoch": 2.1957186544342506, + "grad_norm": 0.23175981640815735, + "learning_rate": 9.891572941027577e-05, + "loss": 0.0392, + "step": 11310 + }, + { + "epoch": 2.197660307752051, + "grad_norm": 0.35996147990226746, + "learning_rate": 9.89123019942509e-05, + "loss": 0.0474, + "step": 11320 + }, + { + "epoch": 2.199601961069851, + "grad_norm": 0.3375934958457947, + "learning_rate": 9.89088692292502e-05, + "loss": 0.0384, + "step": 11330 + }, + { + "epoch": 2.201543614387651, + "grad_norm": 0.1997940093278885, + "learning_rate": 9.89054311156491e-05, + "loss": 0.049, + "step": 11340 + }, + { + "epoch": 2.2034852677054513, + "grad_norm": 0.23274856805801392, + "learning_rate": 9.890198765382357e-05, + "loss": 0.0373, + "step": 11350 + }, + { + "epoch": 2.205426921023251, + "grad_norm": 0.23315036296844482, + "learning_rate": 9.889853884415021e-05, + "loss": 0.0421, + "step": 11360 + }, + { + "epoch": 2.2073685743410514, + "grad_norm": 0.22231096029281616, + "learning_rate": 9.889508468700614e-05, + "loss": 0.0449, + "step": 11370 + }, + { + "epoch": 2.2093102276588517, + "grad_norm": 0.24994972348213196, + "learning_rate": 9.889162518276915e-05, + "loss": 0.0539, + "step": 11380 + }, + { + "epoch": 2.2112518809766515, + "grad_norm": 0.29225173592567444, + "learning_rate": 9.888816033181752e-05, + "loss": 0.0427, + "step": 11390 + }, + { + "epoch": 2.2131935342944518, + "grad_norm": 0.2766824960708618, + "learning_rate": 9.888469013453018e-05, + "loss": 0.0412, + "step": 11400 + }, + { + "epoch": 2.215135187612252, + "grad_norm": 0.26028740406036377, + "learning_rate": 9.888121459128663e-05, + "loss": 0.0528, + "step": 11410 + }, + { + "epoch": 2.217076840930052, + "grad_norm": 0.3203356862068176, + "learning_rate": 9.887773370246693e-05, + "loss": 0.047, + "step": 11420 + }, + { + "epoch": 2.219018494247852, + "grad_norm": 0.276254802942276, + "learning_rate": 9.887424746845177e-05, + "loss": 0.0504, + "step": 11430 + }, + { + "epoch": 2.220960147565652, + "grad_norm": 0.26180729269981384, + "learning_rate": 9.887075588962239e-05, + "loss": 0.0446, + "step": 11440 + }, + { + "epoch": 2.2229018008834522, + "grad_norm": 0.2508277893066406, + "learning_rate": 9.88672589663606e-05, + "loss": 0.0467, + "step": 11450 + }, + { + "epoch": 2.2248434542012525, + "grad_norm": 0.3178410232067108, + "learning_rate": 9.886375669904886e-05, + "loss": 0.0536, + "step": 11460 + }, + { + "epoch": 2.2267851075190523, + "grad_norm": 0.251884788274765, + "learning_rate": 9.886024908807014e-05, + "loss": 0.0339, + "step": 11470 + }, + { + "epoch": 2.2287267608368526, + "grad_norm": 0.250946968793869, + "learning_rate": 9.885673613380806e-05, + "loss": 0.0404, + "step": 11480 + }, + { + "epoch": 2.230668414154653, + "grad_norm": 0.2872620224952698, + "learning_rate": 9.885321783664676e-05, + "loss": 0.035, + "step": 11490 + }, + { + "epoch": 2.2326100674724527, + "grad_norm": 0.344888836145401, + "learning_rate": 9.884969419697101e-05, + "loss": 0.0458, + "step": 11500 + }, + { + "epoch": 2.234551720790253, + "grad_norm": 0.22256311774253845, + "learning_rate": 9.884616521516614e-05, + "loss": 0.0361, + "step": 11510 + }, + { + "epoch": 2.236493374108053, + "grad_norm": 0.18137076497077942, + "learning_rate": 9.88426308916181e-05, + "loss": 0.0485, + "step": 11520 + }, + { + "epoch": 2.238435027425853, + "grad_norm": 0.20176567137241364, + "learning_rate": 9.883909122671335e-05, + "loss": 0.046, + "step": 11530 + }, + { + "epoch": 2.2403766807436534, + "grad_norm": 0.21256938576698303, + "learning_rate": 9.883554622083904e-05, + "loss": 0.0371, + "step": 11540 + }, + { + "epoch": 2.242318334061453, + "grad_norm": 0.2991948425769806, + "learning_rate": 9.88319958743828e-05, + "loss": 0.0425, + "step": 11550 + }, + { + "epoch": 2.2442599873792535, + "grad_norm": 0.30896255373954773, + "learning_rate": 9.882844018773291e-05, + "loss": 0.0529, + "step": 11560 + }, + { + "epoch": 2.2462016406970537, + "grad_norm": 0.32034483551979065, + "learning_rate": 9.882487916127823e-05, + "loss": 0.0337, + "step": 11570 + }, + { + "epoch": 2.2481432940148536, + "grad_norm": 0.20584042370319366, + "learning_rate": 9.882131279540815e-05, + "loss": 0.0445, + "step": 11580 + }, + { + "epoch": 2.250084947332654, + "grad_norm": 0.31939497590065, + "learning_rate": 9.881774109051271e-05, + "loss": 0.0452, + "step": 11590 + }, + { + "epoch": 2.2520266006504537, + "grad_norm": 0.22086302936077118, + "learning_rate": 9.881416404698252e-05, + "loss": 0.0338, + "step": 11600 + }, + { + "epoch": 2.253968253968254, + "grad_norm": 0.2617216110229492, + "learning_rate": 9.881058166520873e-05, + "loss": 0.0379, + "step": 11610 + }, + { + "epoch": 2.2559099072860542, + "grad_norm": 0.32944685220718384, + "learning_rate": 9.880699394558311e-05, + "loss": 0.0425, + "step": 11620 + }, + { + "epoch": 2.257851560603854, + "grad_norm": 0.18866553902626038, + "learning_rate": 9.880340088849801e-05, + "loss": 0.038, + "step": 11630 + }, + { + "epoch": 2.2597932139216543, + "grad_norm": 0.2825371026992798, + "learning_rate": 9.879980249434637e-05, + "loss": 0.0382, + "step": 11640 + }, + { + "epoch": 2.261734867239454, + "grad_norm": 0.22992601990699768, + "learning_rate": 9.879619876352168e-05, + "loss": 0.0416, + "step": 11650 + }, + { + "epoch": 2.2636765205572544, + "grad_norm": 0.18645122647285461, + "learning_rate": 9.879258969641809e-05, + "loss": 0.0481, + "step": 11660 + }, + { + "epoch": 2.2656181738750547, + "grad_norm": 0.34367090463638306, + "learning_rate": 9.878897529343023e-05, + "loss": 0.0484, + "step": 11670 + }, + { + "epoch": 2.2675598271928545, + "grad_norm": 0.28868338465690613, + "learning_rate": 9.878535555495338e-05, + "loss": 0.0393, + "step": 11680 + }, + { + "epoch": 2.269501480510655, + "grad_norm": 0.24136511981487274, + "learning_rate": 9.87817304813834e-05, + "loss": 0.0522, + "step": 11690 + }, + { + "epoch": 2.271443133828455, + "grad_norm": 0.24066685140132904, + "learning_rate": 9.877810007311671e-05, + "loss": 0.0444, + "step": 11700 + }, + { + "epoch": 2.273384787146255, + "grad_norm": 0.3030815124511719, + "learning_rate": 9.877446433055035e-05, + "loss": 0.0485, + "step": 11710 + }, + { + "epoch": 2.275326440464055, + "grad_norm": 0.2122952789068222, + "learning_rate": 9.877082325408191e-05, + "loss": 0.0408, + "step": 11720 + }, + { + "epoch": 2.2772680937818555, + "grad_norm": 0.2681575119495392, + "learning_rate": 9.876717684410954e-05, + "loss": 0.0415, + "step": 11730 + }, + { + "epoch": 2.2792097470996553, + "grad_norm": 0.21529118716716766, + "learning_rate": 9.876352510103204e-05, + "loss": 0.0357, + "step": 11740 + }, + { + "epoch": 2.2811514004174556, + "grad_norm": 0.35827216506004333, + "learning_rate": 9.875986802524875e-05, + "loss": 0.0434, + "step": 11750 + }, + { + "epoch": 2.2830930537352554, + "grad_norm": 0.24774675071239471, + "learning_rate": 9.87562056171596e-05, + "loss": 0.0463, + "step": 11760 + }, + { + "epoch": 2.2850347070530557, + "grad_norm": 0.22255069017410278, + "learning_rate": 9.875253787716511e-05, + "loss": 0.0394, + "step": 11770 + }, + { + "epoch": 2.286976360370856, + "grad_norm": 0.18556830286979675, + "learning_rate": 9.874886480566637e-05, + "loss": 0.034, + "step": 11780 + }, + { + "epoch": 2.2889180136886558, + "grad_norm": 0.17128552496433258, + "learning_rate": 9.874518640306507e-05, + "loss": 0.0398, + "step": 11790 + }, + { + "epoch": 2.290859667006456, + "grad_norm": 0.2582174837589264, + "learning_rate": 9.874150266976347e-05, + "loss": 0.0532, + "step": 11800 + }, + { + "epoch": 2.292801320324256, + "grad_norm": 0.24217425286769867, + "learning_rate": 9.873781360616443e-05, + "loss": 0.0391, + "step": 11810 + }, + { + "epoch": 2.294742973642056, + "grad_norm": 0.3460756540298462, + "learning_rate": 9.873411921267137e-05, + "loss": 0.0381, + "step": 11820 + }, + { + "epoch": 2.2966846269598564, + "grad_norm": 0.300162672996521, + "learning_rate": 9.873041948968829e-05, + "loss": 0.0469, + "step": 11830 + }, + { + "epoch": 2.2986262802776563, + "grad_norm": 0.2901240289211273, + "learning_rate": 9.872671443761981e-05, + "loss": 0.0434, + "step": 11840 + }, + { + "epoch": 2.3005679335954565, + "grad_norm": 0.13106854259967804, + "learning_rate": 9.872300405687109e-05, + "loss": 0.0412, + "step": 11850 + }, + { + "epoch": 2.302509586913257, + "grad_norm": 0.27847757935523987, + "learning_rate": 9.871928834784792e-05, + "loss": 0.0364, + "step": 11860 + }, + { + "epoch": 2.3044512402310566, + "grad_norm": 0.3369109630584717, + "learning_rate": 9.871556731095661e-05, + "loss": 0.0355, + "step": 11870 + }, + { + "epoch": 2.306392893548857, + "grad_norm": 0.2085149735212326, + "learning_rate": 9.871184094660411e-05, + "loss": 0.0372, + "step": 11880 + }, + { + "epoch": 2.308334546866657, + "grad_norm": 0.3727099597454071, + "learning_rate": 9.870810925519791e-05, + "loss": 0.0355, + "step": 11890 + }, + { + "epoch": 2.310276200184457, + "grad_norm": 0.21240292489528656, + "learning_rate": 9.870437223714612e-05, + "loss": 0.0424, + "step": 11900 + }, + { + "epoch": 2.3122178535022573, + "grad_norm": 0.32667481899261475, + "learning_rate": 9.87006298928574e-05, + "loss": 0.0391, + "step": 11910 + }, + { + "epoch": 2.314159506820057, + "grad_norm": 0.2820795774459839, + "learning_rate": 9.869688222274103e-05, + "loss": 0.0369, + "step": 11920 + }, + { + "epoch": 2.3161011601378574, + "grad_norm": 0.3004360496997833, + "learning_rate": 9.869312922720681e-05, + "loss": 0.0385, + "step": 11930 + }, + { + "epoch": 2.3180428134556577, + "grad_norm": 0.21331919729709625, + "learning_rate": 9.868937090666521e-05, + "loss": 0.0401, + "step": 11940 + }, + { + "epoch": 2.3199844667734575, + "grad_norm": 0.29107069969177246, + "learning_rate": 9.86856072615272e-05, + "loss": 0.0385, + "step": 11950 + }, + { + "epoch": 2.3219261200912578, + "grad_norm": 0.22933192551136017, + "learning_rate": 9.868183829220438e-05, + "loss": 0.0409, + "step": 11960 + }, + { + "epoch": 2.3238677734090576, + "grad_norm": 0.256235808134079, + "learning_rate": 9.867806399910893e-05, + "loss": 0.0393, + "step": 11970 + }, + { + "epoch": 2.325809426726858, + "grad_norm": 0.2520841956138611, + "learning_rate": 9.867428438265356e-05, + "loss": 0.0376, + "step": 11980 + }, + { + "epoch": 2.327751080044658, + "grad_norm": 0.3014223873615265, + "learning_rate": 9.867049944325165e-05, + "loss": 0.0457, + "step": 11990 + }, + { + "epoch": 2.329692733362458, + "grad_norm": 0.20606453716754913, + "learning_rate": 9.86667091813171e-05, + "loss": 0.0381, + "step": 12000 + }, + { + "epoch": 2.3316343866802582, + "grad_norm": 0.20465807616710663, + "learning_rate": 9.866291359726438e-05, + "loss": 0.0355, + "step": 12010 + }, + { + "epoch": 2.3335760399980585, + "grad_norm": 0.3016906678676605, + "learning_rate": 9.865911269150861e-05, + "loss": 0.0389, + "step": 12020 + }, + { + "epoch": 2.3355176933158583, + "grad_norm": 0.2750893831253052, + "learning_rate": 9.865530646446544e-05, + "loss": 0.0399, + "step": 12030 + }, + { + "epoch": 2.3374593466336586, + "grad_norm": 0.2593207061290741, + "learning_rate": 9.86514949165511e-05, + "loss": 0.0416, + "step": 12040 + }, + { + "epoch": 2.339400999951459, + "grad_norm": 0.19491451978683472, + "learning_rate": 9.864767804818243e-05, + "loss": 0.0416, + "step": 12050 + }, + { + "epoch": 2.3413426532692587, + "grad_norm": 0.1798880398273468, + "learning_rate": 9.86438558597768e-05, + "loss": 0.0399, + "step": 12060 + }, + { + "epoch": 2.343284306587059, + "grad_norm": 0.3641591966152191, + "learning_rate": 9.864002835175225e-05, + "loss": 0.0465, + "step": 12070 + }, + { + "epoch": 2.345225959904859, + "grad_norm": 0.18194542825222015, + "learning_rate": 9.863619552452734e-05, + "loss": 0.0404, + "step": 12080 + }, + { + "epoch": 2.347167613222659, + "grad_norm": 0.17058973014354706, + "learning_rate": 9.863235737852119e-05, + "loss": 0.0344, + "step": 12090 + }, + { + "epoch": 2.3491092665404594, + "grad_norm": 0.17012302577495575, + "learning_rate": 9.862851391415356e-05, + "loss": 0.0392, + "step": 12100 + }, + { + "epoch": 2.351050919858259, + "grad_norm": 0.26636114716529846, + "learning_rate": 9.862466513184477e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 2.3529925731760595, + "grad_norm": 0.2800975739955902, + "learning_rate": 9.86208110320157e-05, + "loss": 0.0507, + "step": 12120 + }, + { + "epoch": 2.3549342264938593, + "grad_norm": 0.16764843463897705, + "learning_rate": 9.861695161508784e-05, + "loss": 0.0422, + "step": 12130 + }, + { + "epoch": 2.3568758798116596, + "grad_norm": 0.26224759221076965, + "learning_rate": 9.861308688148324e-05, + "loss": 0.0366, + "step": 12140 + }, + { + "epoch": 2.35881753312946, + "grad_norm": 0.3461199104785919, + "learning_rate": 9.860921683162455e-05, + "loss": 0.0558, + "step": 12150 + }, + { + "epoch": 2.3607591864472597, + "grad_norm": 0.32444557547569275, + "learning_rate": 9.860534146593499e-05, + "loss": 0.0458, + "step": 12160 + }, + { + "epoch": 2.36270083976506, + "grad_norm": 0.2528102695941925, + "learning_rate": 9.860146078483836e-05, + "loss": 0.0367, + "step": 12170 + }, + { + "epoch": 2.3646424930828602, + "grad_norm": 0.16529738903045654, + "learning_rate": 9.859757478875905e-05, + "loss": 0.0421, + "step": 12180 + }, + { + "epoch": 2.36658414640066, + "grad_norm": 0.28634533286094666, + "learning_rate": 9.859368347812204e-05, + "loss": 0.0332, + "step": 12190 + }, + { + "epoch": 2.3685257997184603, + "grad_norm": 0.19485671818256378, + "learning_rate": 9.858978685335285e-05, + "loss": 0.0316, + "step": 12200 + }, + { + "epoch": 2.3704674530362606, + "grad_norm": 0.29707661271095276, + "learning_rate": 9.858588491487763e-05, + "loss": 0.0456, + "step": 12210 + }, + { + "epoch": 2.3724091063540604, + "grad_norm": 0.24157246947288513, + "learning_rate": 9.858197766312308e-05, + "loss": 0.04, + "step": 12220 + }, + { + "epoch": 2.3743507596718607, + "grad_norm": 0.30625608563423157, + "learning_rate": 9.857806509851649e-05, + "loss": 0.04, + "step": 12230 + }, + { + "epoch": 2.3762924129896605, + "grad_norm": 0.3660636246204376, + "learning_rate": 9.857414722148574e-05, + "loss": 0.0453, + "step": 12240 + }, + { + "epoch": 2.378234066307461, + "grad_norm": 0.2620915472507477, + "learning_rate": 9.857022403245928e-05, + "loss": 0.0411, + "step": 12250 + }, + { + "epoch": 2.380175719625261, + "grad_norm": 0.26251742243766785, + "learning_rate": 9.856629553186615e-05, + "loss": 0.0432, + "step": 12260 + }, + { + "epoch": 2.382117372943061, + "grad_norm": 0.21912500262260437, + "learning_rate": 9.856236172013595e-05, + "loss": 0.0378, + "step": 12270 + }, + { + "epoch": 2.384059026260861, + "grad_norm": 0.17894482612609863, + "learning_rate": 9.85584225976989e-05, + "loss": 0.0474, + "step": 12280 + }, + { + "epoch": 2.386000679578661, + "grad_norm": 0.18768933415412903, + "learning_rate": 9.855447816498575e-05, + "loss": 0.0352, + "step": 12290 + }, + { + "epoch": 2.3879423328964613, + "grad_norm": 0.31961652636528015, + "learning_rate": 9.855052842242787e-05, + "loss": 0.038, + "step": 12300 + }, + { + "epoch": 2.3898839862142616, + "grad_norm": 0.333266943693161, + "learning_rate": 9.85465733704572e-05, + "loss": 0.0387, + "step": 12310 + }, + { + "epoch": 2.3918256395320614, + "grad_norm": 0.21190591156482697, + "learning_rate": 9.854261300950624e-05, + "loss": 0.0429, + "step": 12320 + }, + { + "epoch": 2.3937672928498617, + "grad_norm": 0.2088167518377304, + "learning_rate": 9.853864734000813e-05, + "loss": 0.037, + "step": 12330 + }, + { + "epoch": 2.395708946167662, + "grad_norm": 0.30809736251831055, + "learning_rate": 9.85346763623965e-05, + "loss": 0.0412, + "step": 12340 + }, + { + "epoch": 2.3976505994854618, + "grad_norm": 0.2624853849411011, + "learning_rate": 9.853070007710564e-05, + "loss": 0.0376, + "step": 12350 + }, + { + "epoch": 2.399592252803262, + "grad_norm": 0.18395733833312988, + "learning_rate": 9.85267184845704e-05, + "loss": 0.0401, + "step": 12360 + }, + { + "epoch": 2.4015339061210623, + "grad_norm": 0.4474465250968933, + "learning_rate": 9.852273158522616e-05, + "loss": 0.0499, + "step": 12370 + }, + { + "epoch": 2.403475559438862, + "grad_norm": 0.4525596797466278, + "learning_rate": 9.851873937950896e-05, + "loss": 0.0388, + "step": 12380 + }, + { + "epoch": 2.4054172127566624, + "grad_norm": 0.3306519389152527, + "learning_rate": 9.851474186785537e-05, + "loss": 0.0461, + "step": 12390 + }, + { + "epoch": 2.4073588660744623, + "grad_norm": 0.26897549629211426, + "learning_rate": 9.851073905070254e-05, + "loss": 0.0355, + "step": 12400 + }, + { + "epoch": 2.4093005193922625, + "grad_norm": 0.1764659732580185, + "learning_rate": 9.850673092848824e-05, + "loss": 0.0404, + "step": 12410 + }, + { + "epoch": 2.411242172710063, + "grad_norm": 0.1953483521938324, + "learning_rate": 9.850271750165077e-05, + "loss": 0.0343, + "step": 12420 + }, + { + "epoch": 2.4131838260278626, + "grad_norm": 0.2273072749376297, + "learning_rate": 9.849869877062902e-05, + "loss": 0.0375, + "step": 12430 + }, + { + "epoch": 2.415125479345663, + "grad_norm": 0.1832713484764099, + "learning_rate": 9.849467473586252e-05, + "loss": 0.0374, + "step": 12440 + }, + { + "epoch": 2.4170671326634627, + "grad_norm": 0.2107522189617157, + "learning_rate": 9.849064539779127e-05, + "loss": 0.043, + "step": 12450 + }, + { + "epoch": 2.419008785981263, + "grad_norm": 0.25547781586647034, + "learning_rate": 9.848661075685594e-05, + "loss": 0.0344, + "step": 12460 + }, + { + "epoch": 2.4209504392990633, + "grad_norm": 0.15128768980503082, + "learning_rate": 9.848257081349778e-05, + "loss": 0.0357, + "step": 12470 + }, + { + "epoch": 2.422892092616863, + "grad_norm": 0.2305450290441513, + "learning_rate": 9.847852556815856e-05, + "loss": 0.041, + "step": 12480 + }, + { + "epoch": 2.4248337459346634, + "grad_norm": 0.22432519495487213, + "learning_rate": 9.847447502128067e-05, + "loss": 0.0362, + "step": 12490 + }, + { + "epoch": 2.4267753992524637, + "grad_norm": 0.140140563249588, + "learning_rate": 9.847041917330708e-05, + "loss": 0.0414, + "step": 12500 + }, + { + "epoch": 2.4287170525702635, + "grad_norm": 0.45049312710762024, + "learning_rate": 9.846635802468132e-05, + "loss": 0.0368, + "step": 12510 + }, + { + "epoch": 2.4306587058880638, + "grad_norm": 0.1715106964111328, + "learning_rate": 9.84622915758475e-05, + "loss": 0.0437, + "step": 12520 + }, + { + "epoch": 2.432600359205864, + "grad_norm": 0.2971069812774658, + "learning_rate": 9.845821982725034e-05, + "loss": 0.0425, + "step": 12530 + }, + { + "epoch": 2.434542012523664, + "grad_norm": 0.27265700697898865, + "learning_rate": 9.845414277933514e-05, + "loss": 0.0389, + "step": 12540 + }, + { + "epoch": 2.436483665841464, + "grad_norm": 0.26858657598495483, + "learning_rate": 9.845006043254771e-05, + "loss": 0.0444, + "step": 12550 + }, + { + "epoch": 2.438425319159264, + "grad_norm": 0.34696662425994873, + "learning_rate": 9.844597278733451e-05, + "loss": 0.0394, + "step": 12560 + }, + { + "epoch": 2.4403669724770642, + "grad_norm": 0.204989954829216, + "learning_rate": 9.844187984414259e-05, + "loss": 0.0429, + "step": 12570 + }, + { + "epoch": 2.4423086257948645, + "grad_norm": 0.32350537180900574, + "learning_rate": 9.84377816034195e-05, + "loss": 0.0457, + "step": 12580 + }, + { + "epoch": 2.4442502791126643, + "grad_norm": 0.20798306167125702, + "learning_rate": 9.843367806561345e-05, + "loss": 0.0368, + "step": 12590 + }, + { + "epoch": 2.4461919324304646, + "grad_norm": 0.1827053278684616, + "learning_rate": 9.842956923117317e-05, + "loss": 0.0391, + "step": 12600 + }, + { + "epoch": 2.4481335857482645, + "grad_norm": 0.25487950444221497, + "learning_rate": 9.842545510054802e-05, + "loss": 0.0386, + "step": 12610 + }, + { + "epoch": 2.4500752390660647, + "grad_norm": 0.2876618802547455, + "learning_rate": 9.842133567418792e-05, + "loss": 0.0383, + "step": 12620 + }, + { + "epoch": 2.452016892383865, + "grad_norm": 0.20811814069747925, + "learning_rate": 9.841721095254333e-05, + "loss": 0.0351, + "step": 12630 + }, + { + "epoch": 2.453958545701665, + "grad_norm": 0.16110315918922424, + "learning_rate": 9.841308093606537e-05, + "loss": 0.0371, + "step": 12640 + }, + { + "epoch": 2.455900199019465, + "grad_norm": 0.1978152096271515, + "learning_rate": 9.840894562520565e-05, + "loss": 0.0345, + "step": 12650 + }, + { + "epoch": 2.457841852337265, + "grad_norm": 0.25268232822418213, + "learning_rate": 9.840480502041642e-05, + "loss": 0.0393, + "step": 12660 + }, + { + "epoch": 2.459783505655065, + "grad_norm": 0.14525453746318817, + "learning_rate": 9.840065912215049e-05, + "loss": 0.0375, + "step": 12670 + }, + { + "epoch": 2.4617251589728655, + "grad_norm": 0.20427528023719788, + "learning_rate": 9.839650793086124e-05, + "loss": 0.0389, + "step": 12680 + }, + { + "epoch": 2.4636668122906653, + "grad_norm": 0.15827445685863495, + "learning_rate": 9.839235144700265e-05, + "loss": 0.0422, + "step": 12690 + }, + { + "epoch": 2.4656084656084656, + "grad_norm": 0.33463430404663086, + "learning_rate": 9.838818967102926e-05, + "loss": 0.0352, + "step": 12700 + }, + { + "epoch": 2.467550118926266, + "grad_norm": 0.30123674869537354, + "learning_rate": 9.83840226033962e-05, + "loss": 0.0328, + "step": 12710 + }, + { + "epoch": 2.4694917722440657, + "grad_norm": 0.22922681272029877, + "learning_rate": 9.837985024455918e-05, + "loss": 0.0457, + "step": 12720 + }, + { + "epoch": 2.471433425561866, + "grad_norm": 0.1748860478401184, + "learning_rate": 9.837567259497447e-05, + "loss": 0.0403, + "step": 12730 + }, + { + "epoch": 2.4733750788796662, + "grad_norm": 0.15902581810951233, + "learning_rate": 9.837148965509894e-05, + "loss": 0.0424, + "step": 12740 + }, + { + "epoch": 2.475316732197466, + "grad_norm": 0.28162330389022827, + "learning_rate": 9.836730142539001e-05, + "loss": 0.0364, + "step": 12750 + }, + { + "epoch": 2.4772583855152663, + "grad_norm": 0.24068555235862732, + "learning_rate": 9.836310790630574e-05, + "loss": 0.0397, + "step": 12760 + }, + { + "epoch": 2.479200038833066, + "grad_norm": 0.2462080866098404, + "learning_rate": 9.83589090983047e-05, + "loss": 0.041, + "step": 12770 + }, + { + "epoch": 2.4811416921508664, + "grad_norm": 0.2251151204109192, + "learning_rate": 9.835470500184605e-05, + "loss": 0.0376, + "step": 12780 + }, + { + "epoch": 2.4830833454686667, + "grad_norm": 0.22156861424446106, + "learning_rate": 9.835049561738957e-05, + "loss": 0.0395, + "step": 12790 + }, + { + "epoch": 2.4850249987864665, + "grad_norm": 0.2826627492904663, + "learning_rate": 9.834628094539558e-05, + "loss": 0.0458, + "step": 12800 + }, + { + "epoch": 2.486966652104267, + "grad_norm": 0.2418656200170517, + "learning_rate": 9.834206098632499e-05, + "loss": 0.0384, + "step": 12810 + }, + { + "epoch": 2.4889083054220666, + "grad_norm": 0.28288963437080383, + "learning_rate": 9.833783574063931e-05, + "loss": 0.0416, + "step": 12820 + }, + { + "epoch": 2.490849958739867, + "grad_norm": 0.36279118061065674, + "learning_rate": 9.833360520880058e-05, + "loss": 0.0417, + "step": 12830 + }, + { + "epoch": 2.492791612057667, + "grad_norm": 0.27333202958106995, + "learning_rate": 9.832936939127144e-05, + "loss": 0.0468, + "step": 12840 + }, + { + "epoch": 2.494733265375467, + "grad_norm": 0.2738889753818512, + "learning_rate": 9.832512828851515e-05, + "loss": 0.0416, + "step": 12850 + }, + { + "epoch": 2.4966749186932673, + "grad_norm": 0.19328583776950836, + "learning_rate": 9.832088190099546e-05, + "loss": 0.0417, + "step": 12860 + }, + { + "epoch": 2.4986165720110676, + "grad_norm": 0.30066728591918945, + "learning_rate": 9.831663022917679e-05, + "loss": 0.0479, + "step": 12870 + }, + { + "epoch": 2.5005582253288674, + "grad_norm": 0.17595359683036804, + "learning_rate": 9.831237327352407e-05, + "loss": 0.041, + "step": 12880 + }, + { + "epoch": 2.5024998786466677, + "grad_norm": 0.24456612765789032, + "learning_rate": 9.830811103450286e-05, + "loss": 0.0326, + "step": 12890 + }, + { + "epoch": 2.504441531964468, + "grad_norm": 0.2817888855934143, + "learning_rate": 9.830384351257924e-05, + "loss": 0.0402, + "step": 12900 + }, + { + "epoch": 2.506383185282268, + "grad_norm": 0.24459518492221832, + "learning_rate": 9.829957070821993e-05, + "loss": 0.034, + "step": 12910 + }, + { + "epoch": 2.508324838600068, + "grad_norm": 0.29423055052757263, + "learning_rate": 9.829529262189218e-05, + "loss": 0.0399, + "step": 12920 + }, + { + "epoch": 2.510266491917868, + "grad_norm": 0.24283277988433838, + "learning_rate": 9.829100925406385e-05, + "loss": 0.0358, + "step": 12930 + }, + { + "epoch": 2.512208145235668, + "grad_norm": 0.18618683516979218, + "learning_rate": 9.828672060520333e-05, + "loss": 0.0454, + "step": 12940 + }, + { + "epoch": 2.5141497985534684, + "grad_norm": 0.14100512862205505, + "learning_rate": 9.828242667577966e-05, + "loss": 0.0285, + "step": 12950 + }, + { + "epoch": 2.5160914518712683, + "grad_norm": 0.29148945212364197, + "learning_rate": 9.82781274662624e-05, + "loss": 0.0357, + "step": 12960 + }, + { + "epoch": 2.5180331051890685, + "grad_norm": 0.21421843767166138, + "learning_rate": 9.82738229771217e-05, + "loss": 0.0386, + "step": 12970 + }, + { + "epoch": 2.5199747585068684, + "grad_norm": 0.23319341242313385, + "learning_rate": 9.826951320882829e-05, + "loss": 0.0418, + "step": 12980 + }, + { + "epoch": 2.5219164118246686, + "grad_norm": 0.2467195838689804, + "learning_rate": 9.826519816185351e-05, + "loss": 0.0438, + "step": 12990 + }, + { + "epoch": 2.523858065142469, + "grad_norm": 0.17568446695804596, + "learning_rate": 9.826087783666921e-05, + "loss": 0.0292, + "step": 13000 + }, + { + "epoch": 2.525799718460269, + "grad_norm": 0.21209290623664856, + "learning_rate": 9.825655223374787e-05, + "loss": 0.0498, + "step": 13010 + }, + { + "epoch": 2.527741371778069, + "grad_norm": 0.1760053038597107, + "learning_rate": 9.825222135356253e-05, + "loss": 0.0347, + "step": 13020 + }, + { + "epoch": 2.5296830250958693, + "grad_norm": 0.18908044695854187, + "learning_rate": 9.82478851965868e-05, + "loss": 0.0436, + "step": 13030 + }, + { + "epoch": 2.531624678413669, + "grad_norm": 0.2042984664440155, + "learning_rate": 9.82435437632949e-05, + "loss": 0.0408, + "step": 13040 + }, + { + "epoch": 2.5335663317314694, + "grad_norm": 0.34546831250190735, + "learning_rate": 9.823919705416158e-05, + "loss": 0.0488, + "step": 13050 + }, + { + "epoch": 2.5355079850492697, + "grad_norm": 0.24475133419036865, + "learning_rate": 9.82348450696622e-05, + "loss": 0.038, + "step": 13060 + }, + { + "epoch": 2.5374496383670695, + "grad_norm": 0.29207471013069153, + "learning_rate": 9.823048781027268e-05, + "loss": 0.0385, + "step": 13070 + }, + { + "epoch": 2.5393912916848698, + "grad_norm": 0.20733992755413055, + "learning_rate": 9.822612527646953e-05, + "loss": 0.0393, + "step": 13080 + }, + { + "epoch": 2.5413329450026696, + "grad_norm": 0.2014782577753067, + "learning_rate": 9.822175746872984e-05, + "loss": 0.0357, + "step": 13090 + }, + { + "epoch": 2.54327459832047, + "grad_norm": 0.19127604365348816, + "learning_rate": 9.821738438753123e-05, + "loss": 0.0396, + "step": 13100 + }, + { + "epoch": 2.54521625163827, + "grad_norm": 0.22181417047977448, + "learning_rate": 9.821300603335196e-05, + "loss": 0.0399, + "step": 13110 + }, + { + "epoch": 2.54715790495607, + "grad_norm": 0.23517897725105286, + "learning_rate": 9.820862240667085e-05, + "loss": 0.0491, + "step": 13120 + }, + { + "epoch": 2.5490995582738702, + "grad_norm": 0.20045605301856995, + "learning_rate": 9.820423350796726e-05, + "loss": 0.034, + "step": 13130 + }, + { + "epoch": 2.55104121159167, + "grad_norm": 0.27515801787376404, + "learning_rate": 9.819983933772118e-05, + "loss": 0.0468, + "step": 13140 + }, + { + "epoch": 2.5529828649094704, + "grad_norm": 0.3666194975376129, + "learning_rate": 9.819543989641314e-05, + "loss": 0.045, + "step": 13150 + }, + { + "epoch": 2.5549245182272706, + "grad_norm": 0.16741345822811127, + "learning_rate": 9.819103518452423e-05, + "loss": 0.0428, + "step": 13160 + }, + { + "epoch": 2.556866171545071, + "grad_norm": 0.21022769808769226, + "learning_rate": 9.818662520253618e-05, + "loss": 0.0409, + "step": 13170 + }, + { + "epoch": 2.5588078248628707, + "grad_norm": 0.15099020302295685, + "learning_rate": 9.818220995093126e-05, + "loss": 0.0416, + "step": 13180 + }, + { + "epoch": 2.560749478180671, + "grad_norm": 0.1715025007724762, + "learning_rate": 9.817778943019228e-05, + "loss": 0.038, + "step": 13190 + }, + { + "epoch": 2.562691131498471, + "grad_norm": 0.2542901933193207, + "learning_rate": 9.81733636408027e-05, + "loss": 0.0407, + "step": 13200 + }, + { + "epoch": 2.564632784816271, + "grad_norm": 0.2655174136161804, + "learning_rate": 9.816893258324649e-05, + "loss": 0.0372, + "step": 13210 + }, + { + "epoch": 2.5665744381340714, + "grad_norm": 0.2933461666107178, + "learning_rate": 9.816449625800823e-05, + "loss": 0.041, + "step": 13220 + }, + { + "epoch": 2.568516091451871, + "grad_norm": 0.3258202373981476, + "learning_rate": 9.816005466557308e-05, + "loss": 0.0378, + "step": 13230 + }, + { + "epoch": 2.5704577447696715, + "grad_norm": 0.29169875383377075, + "learning_rate": 9.815560780642674e-05, + "loss": 0.039, + "step": 13240 + }, + { + "epoch": 2.5723993980874713, + "grad_norm": 0.22979910671710968, + "learning_rate": 9.815115568105555e-05, + "loss": 0.0365, + "step": 13250 + }, + { + "epoch": 2.5743410514052716, + "grad_norm": 0.229916051030159, + "learning_rate": 9.814669828994638e-05, + "loss": 0.0469, + "step": 13260 + }, + { + "epoch": 2.576282704723072, + "grad_norm": 0.1697099357843399, + "learning_rate": 9.814223563358665e-05, + "loss": 0.0423, + "step": 13270 + }, + { + "epoch": 2.5782243580408717, + "grad_norm": 0.31295767426490784, + "learning_rate": 9.813776771246443e-05, + "loss": 0.0388, + "step": 13280 + }, + { + "epoch": 2.580166011358672, + "grad_norm": 0.1971392035484314, + "learning_rate": 9.813329452706829e-05, + "loss": 0.0396, + "step": 13290 + }, + { + "epoch": 2.582107664676472, + "grad_norm": 0.24022720754146576, + "learning_rate": 9.812881607788744e-05, + "loss": 0.0353, + "step": 13300 + }, + { + "epoch": 2.584049317994272, + "grad_norm": 0.2996468245983124, + "learning_rate": 9.812433236541163e-05, + "loss": 0.041, + "step": 13310 + }, + { + "epoch": 2.5859909713120723, + "grad_norm": 0.24354583024978638, + "learning_rate": 9.811984339013116e-05, + "loss": 0.0369, + "step": 13320 + }, + { + "epoch": 2.5879326246298726, + "grad_norm": 0.28509223461151123, + "learning_rate": 9.811534915253698e-05, + "loss": 0.0292, + "step": 13330 + }, + { + "epoch": 2.5898742779476724, + "grad_norm": 0.36854663491249084, + "learning_rate": 9.811084965312056e-05, + "loss": 0.0412, + "step": 13340 + }, + { + "epoch": 2.5918159312654723, + "grad_norm": 0.17614823579788208, + "learning_rate": 9.810634489237396e-05, + "loss": 0.0458, + "step": 13350 + }, + { + "epoch": 2.5937575845832725, + "grad_norm": 0.17309947311878204, + "learning_rate": 9.81018348707898e-05, + "loss": 0.0421, + "step": 13360 + }, + { + "epoch": 2.595699237901073, + "grad_norm": 0.21174271404743195, + "learning_rate": 9.809731958886131e-05, + "loss": 0.0355, + "step": 13370 + }, + { + "epoch": 2.597640891218873, + "grad_norm": 0.19656533002853394, + "learning_rate": 9.809279904708224e-05, + "loss": 0.0424, + "step": 13380 + }, + { + "epoch": 2.599582544536673, + "grad_norm": 0.23465722799301147, + "learning_rate": 9.808827324594699e-05, + "loss": 0.0443, + "step": 13390 + }, + { + "epoch": 2.601524197854473, + "grad_norm": 0.4715029001235962, + "learning_rate": 9.808374218595046e-05, + "loss": 0.0411, + "step": 13400 + }, + { + "epoch": 2.603465851172273, + "grad_norm": 0.0736365094780922, + "learning_rate": 9.80792058675882e-05, + "loss": 0.0376, + "step": 13410 + }, + { + "epoch": 2.6054075044900733, + "grad_norm": 0.32468894124031067, + "learning_rate": 9.807466429135627e-05, + "loss": 0.0444, + "step": 13420 + }, + { + "epoch": 2.6073491578078736, + "grad_norm": 0.26459065079689026, + "learning_rate": 9.807011745775132e-05, + "loss": 0.0452, + "step": 13430 + }, + { + "epoch": 2.6092908111256734, + "grad_norm": 0.22333195805549622, + "learning_rate": 9.806556536727061e-05, + "loss": 0.0348, + "step": 13440 + }, + { + "epoch": 2.6112324644434737, + "grad_norm": 0.25485867261886597, + "learning_rate": 9.806100802041193e-05, + "loss": 0.0342, + "step": 13450 + }, + { + "epoch": 2.6131741177612735, + "grad_norm": 0.28772813081741333, + "learning_rate": 9.805644541767368e-05, + "loss": 0.036, + "step": 13460 + }, + { + "epoch": 2.615115771079074, + "grad_norm": 0.29571038484573364, + "learning_rate": 9.805187755955478e-05, + "loss": 0.0443, + "step": 13470 + }, + { + "epoch": 2.617057424396874, + "grad_norm": 0.2792856693267822, + "learning_rate": 9.804730444655483e-05, + "loss": 0.0424, + "step": 13480 + }, + { + "epoch": 2.618999077714674, + "grad_norm": 0.23584653437137604, + "learning_rate": 9.804272607917388e-05, + "loss": 0.0343, + "step": 13490 + }, + { + "epoch": 2.620940731032474, + "grad_norm": 0.2753569185733795, + "learning_rate": 9.803814245791265e-05, + "loss": 0.0325, + "step": 13500 + }, + { + "epoch": 2.622882384350274, + "grad_norm": 0.2524276673793793, + "learning_rate": 9.803355358327239e-05, + "loss": 0.0371, + "step": 13510 + }, + { + "epoch": 2.6248240376680743, + "grad_norm": 0.301822692155838, + "learning_rate": 9.802895945575492e-05, + "loss": 0.0375, + "step": 13520 + }, + { + "epoch": 2.6267656909858745, + "grad_norm": 0.1943647414445877, + "learning_rate": 9.802436007586266e-05, + "loss": 0.0365, + "step": 13530 + }, + { + "epoch": 2.628707344303675, + "grad_norm": 0.17101408541202545, + "learning_rate": 9.801975544409858e-05, + "loss": 0.0329, + "step": 13540 + }, + { + "epoch": 2.6306489976214746, + "grad_norm": 0.2851826548576355, + "learning_rate": 9.801514556096625e-05, + "loss": 0.0341, + "step": 13550 + }, + { + "epoch": 2.632590650939275, + "grad_norm": 0.2652980089187622, + "learning_rate": 9.801053042696977e-05, + "loss": 0.04, + "step": 13560 + }, + { + "epoch": 2.6345323042570747, + "grad_norm": 0.30544033646583557, + "learning_rate": 9.800591004261388e-05, + "loss": 0.0377, + "step": 13570 + }, + { + "epoch": 2.636473957574875, + "grad_norm": 0.18783368170261383, + "learning_rate": 9.800128440840385e-05, + "loss": 0.0368, + "step": 13580 + }, + { + "epoch": 2.6384156108926753, + "grad_norm": 0.1724255383014679, + "learning_rate": 9.799665352484552e-05, + "loss": 0.0413, + "step": 13590 + }, + { + "epoch": 2.640357264210475, + "grad_norm": 0.32256123423576355, + "learning_rate": 9.799201739244532e-05, + "loss": 0.0439, + "step": 13600 + }, + { + "epoch": 2.6422989175282754, + "grad_norm": 0.2107088416814804, + "learning_rate": 9.798737601171025e-05, + "loss": 0.0343, + "step": 13610 + }, + { + "epoch": 2.6442405708460752, + "grad_norm": 0.23873111605644226, + "learning_rate": 9.79827293831479e-05, + "loss": 0.0351, + "step": 13620 + }, + { + "epoch": 2.6461822241638755, + "grad_norm": 0.177182137966156, + "learning_rate": 9.797807750726638e-05, + "loss": 0.0385, + "step": 13630 + }, + { + "epoch": 2.6481238774816758, + "grad_norm": 0.170808345079422, + "learning_rate": 9.797342038457446e-05, + "loss": 0.0318, + "step": 13640 + }, + { + "epoch": 2.6500655307994756, + "grad_norm": 0.21389266848564148, + "learning_rate": 9.796875801558141e-05, + "loss": 0.0361, + "step": 13650 + }, + { + "epoch": 2.652007184117276, + "grad_norm": 0.2673434019088745, + "learning_rate": 9.79640904007971e-05, + "loss": 0.0391, + "step": 13660 + }, + { + "epoch": 2.6539488374350757, + "grad_norm": 0.251534104347229, + "learning_rate": 9.795941754073199e-05, + "loss": 0.0322, + "step": 13670 + }, + { + "epoch": 2.655890490752876, + "grad_norm": 0.1839836835861206, + "learning_rate": 9.795473943589705e-05, + "loss": 0.039, + "step": 13680 + }, + { + "epoch": 2.6578321440706763, + "grad_norm": 0.24872924387454987, + "learning_rate": 9.795005608680394e-05, + "loss": 0.0362, + "step": 13690 + }, + { + "epoch": 2.6597737973884765, + "grad_norm": 0.2662915885448456, + "learning_rate": 9.794536749396477e-05, + "loss": 0.0373, + "step": 13700 + }, + { + "epoch": 2.6617154507062764, + "grad_norm": 0.30277276039123535, + "learning_rate": 9.79406736578923e-05, + "loss": 0.0368, + "step": 13710 + }, + { + "epoch": 2.6636571040240766, + "grad_norm": 0.3792738914489746, + "learning_rate": 9.793597457909984e-05, + "loss": 0.043, + "step": 13720 + }, + { + "epoch": 2.6655987573418765, + "grad_norm": 0.2198854386806488, + "learning_rate": 9.793127025810127e-05, + "loss": 0.0412, + "step": 13730 + }, + { + "epoch": 2.6675404106596767, + "grad_norm": 0.19362103939056396, + "learning_rate": 9.792656069541104e-05, + "loss": 0.04, + "step": 13740 + }, + { + "epoch": 2.669482063977477, + "grad_norm": 0.17423054575920105, + "learning_rate": 9.79218458915442e-05, + "loss": 0.038, + "step": 13750 + }, + { + "epoch": 2.671423717295277, + "grad_norm": 0.22041168808937073, + "learning_rate": 9.791712584701634e-05, + "loss": 0.0482, + "step": 13760 + }, + { + "epoch": 2.673365370613077, + "grad_norm": 0.12302055954933167, + "learning_rate": 9.791240056234364e-05, + "loss": 0.051, + "step": 13770 + }, + { + "epoch": 2.675307023930877, + "grad_norm": 0.2786458134651184, + "learning_rate": 9.790767003804283e-05, + "loss": 0.0472, + "step": 13780 + }, + { + "epoch": 2.677248677248677, + "grad_norm": 0.20186880230903625, + "learning_rate": 9.790293427463126e-05, + "loss": 0.0374, + "step": 13790 + }, + { + "epoch": 2.6791903305664775, + "grad_norm": 0.24017326533794403, + "learning_rate": 9.789819327262684e-05, + "loss": 0.0419, + "step": 13800 + }, + { + "epoch": 2.6811319838842773, + "grad_norm": 0.209945410490036, + "learning_rate": 9.7893447032548e-05, + "loss": 0.0476, + "step": 13810 + }, + { + "epoch": 2.6830736372020776, + "grad_norm": 0.17020809650421143, + "learning_rate": 9.78886955549138e-05, + "loss": 0.0339, + "step": 13820 + }, + { + "epoch": 2.6850152905198774, + "grad_norm": 0.1694844514131546, + "learning_rate": 9.788393884024387e-05, + "loss": 0.0413, + "step": 13830 + }, + { + "epoch": 2.6869569438376777, + "grad_norm": 0.2913554310798645, + "learning_rate": 9.787917688905836e-05, + "loss": 0.0472, + "step": 13840 + }, + { + "epoch": 2.688898597155478, + "grad_norm": 0.2349453717470169, + "learning_rate": 9.787440970187807e-05, + "loss": 0.0337, + "step": 13850 + }, + { + "epoch": 2.6908402504732782, + "grad_norm": 0.26497116684913635, + "learning_rate": 9.786963727922429e-05, + "loss": 0.037, + "step": 13860 + }, + { + "epoch": 2.692781903791078, + "grad_norm": 0.28286850452423096, + "learning_rate": 9.786485962161897e-05, + "loss": 0.0359, + "step": 13870 + }, + { + "epoch": 2.6947235571088783, + "grad_norm": 0.21095386147499084, + "learning_rate": 9.786007672958455e-05, + "loss": 0.0342, + "step": 13880 + }, + { + "epoch": 2.696665210426678, + "grad_norm": 0.21679918467998505, + "learning_rate": 9.78552886036441e-05, + "loss": 0.043, + "step": 13890 + }, + { + "epoch": 2.6986068637444784, + "grad_norm": 0.2669420838356018, + "learning_rate": 9.785049524432124e-05, + "loss": 0.0438, + "step": 13900 + }, + { + "epoch": 2.7005485170622787, + "grad_norm": 0.217739075422287, + "learning_rate": 9.784569665214016e-05, + "loss": 0.0404, + "step": 13910 + }, + { + "epoch": 2.7024901703800785, + "grad_norm": 0.2075047492980957, + "learning_rate": 9.784089282762563e-05, + "loss": 0.0339, + "step": 13920 + }, + { + "epoch": 2.704431823697879, + "grad_norm": 0.20159761607646942, + "learning_rate": 9.7836083771303e-05, + "loss": 0.04, + "step": 13930 + }, + { + "epoch": 2.7063734770156787, + "grad_norm": 0.14418503642082214, + "learning_rate": 9.783126948369817e-05, + "loss": 0.0344, + "step": 13940 + }, + { + "epoch": 2.708315130333479, + "grad_norm": 0.2713961601257324, + "learning_rate": 9.78264499653376e-05, + "loss": 0.0521, + "step": 13950 + }, + { + "epoch": 2.710256783651279, + "grad_norm": 0.14768514037132263, + "learning_rate": 9.782162521674838e-05, + "loss": 0.0319, + "step": 13960 + }, + { + "epoch": 2.712198436969079, + "grad_norm": 0.13384225964546204, + "learning_rate": 9.781679523845812e-05, + "loss": 0.042, + "step": 13970 + }, + { + "epoch": 2.7141400902868793, + "grad_norm": 0.276685893535614, + "learning_rate": 9.781196003099502e-05, + "loss": 0.0414, + "step": 13980 + }, + { + "epoch": 2.716081743604679, + "grad_norm": 0.18192721903324127, + "learning_rate": 9.780711959488786e-05, + "loss": 0.0325, + "step": 13990 + }, + { + "epoch": 2.7180233969224794, + "grad_norm": 0.2192801982164383, + "learning_rate": 9.780227393066599e-05, + "loss": 0.0348, + "step": 14000 + }, + { + "epoch": 2.7199650502402797, + "grad_norm": 0.2933427095413208, + "learning_rate": 9.77974230388593e-05, + "loss": 0.0397, + "step": 14010 + }, + { + "epoch": 2.72190670355808, + "grad_norm": 0.27941426634788513, + "learning_rate": 9.779256691999829e-05, + "loss": 0.0446, + "step": 14020 + }, + { + "epoch": 2.72384835687588, + "grad_norm": 0.3346688449382782, + "learning_rate": 9.778770557461403e-05, + "loss": 0.039, + "step": 14030 + }, + { + "epoch": 2.72579001019368, + "grad_norm": 0.1484583467245102, + "learning_rate": 9.778283900323812e-05, + "loss": 0.0344, + "step": 14040 + }, + { + "epoch": 2.72773166351148, + "grad_norm": 0.12695372104644775, + "learning_rate": 9.777796720640277e-05, + "loss": 0.0378, + "step": 14050 + }, + { + "epoch": 2.72967331682928, + "grad_norm": 0.21022047102451324, + "learning_rate": 9.777309018464078e-05, + "loss": 0.0361, + "step": 14060 + }, + { + "epoch": 2.7316149701470804, + "grad_norm": 0.17530573904514313, + "learning_rate": 9.776820793848547e-05, + "loss": 0.0358, + "step": 14070 + }, + { + "epoch": 2.7335566234648803, + "grad_norm": 0.16432997584342957, + "learning_rate": 9.776332046847075e-05, + "loss": 0.0463, + "step": 14080 + }, + { + "epoch": 2.7354982767826805, + "grad_norm": 0.21445105969905853, + "learning_rate": 9.775842777513111e-05, + "loss": 0.0368, + "step": 14090 + }, + { + "epoch": 2.7374399301004804, + "grad_norm": 0.17950403690338135, + "learning_rate": 9.775352985900163e-05, + "loss": 0.0434, + "step": 14100 + }, + { + "epoch": 2.7393815834182806, + "grad_norm": 0.21906740963459015, + "learning_rate": 9.774862672061791e-05, + "loss": 0.0477, + "step": 14110 + }, + { + "epoch": 2.741323236736081, + "grad_norm": 0.2322807013988495, + "learning_rate": 9.774371836051616e-05, + "loss": 0.0384, + "step": 14120 + }, + { + "epoch": 2.7432648900538807, + "grad_norm": 0.16903911530971527, + "learning_rate": 9.773880477923315e-05, + "loss": 0.0421, + "step": 14130 + }, + { + "epoch": 2.745206543371681, + "grad_norm": 0.23396658897399902, + "learning_rate": 9.773388597730623e-05, + "loss": 0.0384, + "step": 14140 + }, + { + "epoch": 2.747148196689481, + "grad_norm": 0.24056169390678406, + "learning_rate": 9.77289619552733e-05, + "loss": 0.0315, + "step": 14150 + }, + { + "epoch": 2.749089850007281, + "grad_norm": 0.21483618021011353, + "learning_rate": 9.772403271367285e-05, + "loss": 0.0407, + "step": 14160 + }, + { + "epoch": 2.7510315033250814, + "grad_norm": 0.2588040828704834, + "learning_rate": 9.771909825304396e-05, + "loss": 0.0355, + "step": 14170 + }, + { + "epoch": 2.7529731566428817, + "grad_norm": 0.3633214235305786, + "learning_rate": 9.771415857392619e-05, + "loss": 0.032, + "step": 14180 + }, + { + "epoch": 2.7549148099606815, + "grad_norm": 0.2250208854675293, + "learning_rate": 9.770921367685978e-05, + "loss": 0.0344, + "step": 14190 + }, + { + "epoch": 2.7568564632784818, + "grad_norm": 0.17344382405281067, + "learning_rate": 9.770426356238551e-05, + "loss": 0.0373, + "step": 14200 + }, + { + "epoch": 2.7587981165962816, + "grad_norm": 0.16614574193954468, + "learning_rate": 9.769930823104469e-05, + "loss": 0.0384, + "step": 14210 + }, + { + "epoch": 2.760739769914082, + "grad_norm": 0.3029290735721588, + "learning_rate": 9.769434768337926e-05, + "loss": 0.0408, + "step": 14220 + }, + { + "epoch": 2.762681423231882, + "grad_norm": 0.25267118215560913, + "learning_rate": 9.768938191993164e-05, + "loss": 0.0399, + "step": 14230 + }, + { + "epoch": 2.764623076549682, + "grad_norm": 0.299072802066803, + "learning_rate": 9.768441094124494e-05, + "loss": 0.0436, + "step": 14240 + }, + { + "epoch": 2.7665647298674823, + "grad_norm": 0.3123197555541992, + "learning_rate": 9.767943474786275e-05, + "loss": 0.0411, + "step": 14250 + }, + { + "epoch": 2.768506383185282, + "grad_norm": 0.1775164008140564, + "learning_rate": 9.767445334032923e-05, + "loss": 0.034, + "step": 14260 + }, + { + "epoch": 2.7704480365030824, + "grad_norm": 0.18920181691646576, + "learning_rate": 9.766946671918919e-05, + "loss": 0.0384, + "step": 14270 + }, + { + "epoch": 2.7723896898208826, + "grad_norm": 0.3027338981628418, + "learning_rate": 9.766447488498796e-05, + "loss": 0.0411, + "step": 14280 + }, + { + "epoch": 2.7743313431386825, + "grad_norm": 0.2351565808057785, + "learning_rate": 9.765947783827139e-05, + "loss": 0.0466, + "step": 14290 + }, + { + "epoch": 2.7762729964564827, + "grad_norm": 0.24160386621952057, + "learning_rate": 9.765447557958599e-05, + "loss": 0.039, + "step": 14300 + }, + { + "epoch": 2.7782146497742826, + "grad_norm": 0.4987974762916565, + "learning_rate": 9.764946810947879e-05, + "loss": 0.0438, + "step": 14310 + }, + { + "epoch": 2.780156303092083, + "grad_norm": 0.275923490524292, + "learning_rate": 9.764445542849738e-05, + "loss": 0.0377, + "step": 14320 + }, + { + "epoch": 2.782097956409883, + "grad_norm": 0.31996622681617737, + "learning_rate": 9.763943753718998e-05, + "loss": 0.0445, + "step": 14330 + }, + { + "epoch": 2.7840396097276834, + "grad_norm": 0.20588751137256622, + "learning_rate": 9.76344144361053e-05, + "loss": 0.0431, + "step": 14340 + }, + { + "epoch": 2.785981263045483, + "grad_norm": 0.12768115103244781, + "learning_rate": 9.762938612579269e-05, + "loss": 0.0398, + "step": 14350 + }, + { + "epoch": 2.7879229163632835, + "grad_norm": 0.21817903220653534, + "learning_rate": 9.762435260680202e-05, + "loss": 0.0424, + "step": 14360 + }, + { + "epoch": 2.7898645696810833, + "grad_norm": 0.2040734887123108, + "learning_rate": 9.761931387968373e-05, + "loss": 0.0398, + "step": 14370 + }, + { + "epoch": 2.7918062229988836, + "grad_norm": 0.21220892667770386, + "learning_rate": 9.76142699449889e-05, + "loss": 0.0396, + "step": 14380 + }, + { + "epoch": 2.793747876316684, + "grad_norm": 0.1937689483165741, + "learning_rate": 9.760922080326908e-05, + "loss": 0.0341, + "step": 14390 + }, + { + "epoch": 2.7956895296344837, + "grad_norm": 0.19774776697158813, + "learning_rate": 9.760416645507644e-05, + "loss": 0.041, + "step": 14400 + }, + { + "epoch": 2.797631182952284, + "grad_norm": 0.20217838883399963, + "learning_rate": 9.759910690096375e-05, + "loss": 0.0346, + "step": 14410 + }, + { + "epoch": 2.799572836270084, + "grad_norm": 0.3362094461917877, + "learning_rate": 9.759404214148429e-05, + "loss": 0.0339, + "step": 14420 + }, + { + "epoch": 2.801514489587884, + "grad_norm": 0.1842496544122696, + "learning_rate": 9.758897217719191e-05, + "loss": 0.0396, + "step": 14430 + }, + { + "epoch": 2.8034561429056843, + "grad_norm": 0.3189023733139038, + "learning_rate": 9.758389700864113e-05, + "loss": 0.0417, + "step": 14440 + }, + { + "epoch": 2.805397796223484, + "grad_norm": 0.2413509637117386, + "learning_rate": 9.757881663638688e-05, + "loss": 0.0369, + "step": 14450 + }, + { + "epoch": 2.8073394495412844, + "grad_norm": 0.21029433608055115, + "learning_rate": 9.757373106098478e-05, + "loss": 0.0332, + "step": 14460 + }, + { + "epoch": 2.8092811028590843, + "grad_norm": 0.21833889186382294, + "learning_rate": 9.756864028299097e-05, + "loss": 0.0355, + "step": 14470 + }, + { + "epoch": 2.8112227561768846, + "grad_norm": 0.20574896037578583, + "learning_rate": 9.75635443029622e-05, + "loss": 0.0425, + "step": 14480 + }, + { + "epoch": 2.813164409494685, + "grad_norm": 0.13620036840438843, + "learning_rate": 9.755844312145572e-05, + "loss": 0.0301, + "step": 14490 + }, + { + "epoch": 2.815106062812485, + "grad_norm": 0.17800141870975494, + "learning_rate": 9.755333673902941e-05, + "loss": 0.0354, + "step": 14500 + }, + { + "epoch": 2.817047716130285, + "grad_norm": 0.317480206489563, + "learning_rate": 9.75482251562417e-05, + "loss": 0.0371, + "step": 14510 + }, + { + "epoch": 2.818989369448085, + "grad_norm": 0.3347386419773102, + "learning_rate": 9.754310837365155e-05, + "loss": 0.0362, + "step": 14520 + }, + { + "epoch": 2.820931022765885, + "grad_norm": 0.2558448612689972, + "learning_rate": 9.753798639181856e-05, + "loss": 0.0407, + "step": 14530 + }, + { + "epoch": 2.8228726760836853, + "grad_norm": 0.23550580441951752, + "learning_rate": 9.753285921130286e-05, + "loss": 0.0399, + "step": 14540 + }, + { + "epoch": 2.8248143294014856, + "grad_norm": 0.2056812047958374, + "learning_rate": 9.752772683266512e-05, + "loss": 0.0405, + "step": 14550 + }, + { + "epoch": 2.8267559827192854, + "grad_norm": 0.22122043371200562, + "learning_rate": 9.752258925646665e-05, + "loss": 0.0429, + "step": 14560 + }, + { + "epoch": 2.8286976360370857, + "grad_norm": 0.21765506267547607, + "learning_rate": 9.751744648326926e-05, + "loss": 0.0351, + "step": 14570 + }, + { + "epoch": 2.8306392893548855, + "grad_norm": 0.23865152895450592, + "learning_rate": 9.751229851363536e-05, + "loss": 0.0419, + "step": 14580 + }, + { + "epoch": 2.832580942672686, + "grad_norm": 0.22499457001686096, + "learning_rate": 9.750714534812793e-05, + "loss": 0.0459, + "step": 14590 + }, + { + "epoch": 2.834522595990486, + "grad_norm": 0.18027758598327637, + "learning_rate": 9.750198698731053e-05, + "loss": 0.0362, + "step": 14600 + }, + { + "epoch": 2.836464249308286, + "grad_norm": 0.21722830832004547, + "learning_rate": 9.749682343174722e-05, + "loss": 0.0354, + "step": 14610 + }, + { + "epoch": 2.838405902626086, + "grad_norm": 0.20161864161491394, + "learning_rate": 9.749165468200272e-05, + "loss": 0.0433, + "step": 14620 + }, + { + "epoch": 2.840347555943886, + "grad_norm": 0.2161751687526703, + "learning_rate": 9.748648073864229e-05, + "loss": 0.0302, + "step": 14630 + }, + { + "epoch": 2.8422892092616863, + "grad_norm": 0.23245398700237274, + "learning_rate": 9.748130160223168e-05, + "loss": 0.0434, + "step": 14640 + }, + { + "epoch": 2.8442308625794865, + "grad_norm": 0.31028807163238525, + "learning_rate": 9.747611727333734e-05, + "loss": 0.0387, + "step": 14650 + }, + { + "epoch": 2.846172515897287, + "grad_norm": 0.12059907615184784, + "learning_rate": 9.74709277525262e-05, + "loss": 0.0343, + "step": 14660 + }, + { + "epoch": 2.8481141692150866, + "grad_norm": 0.2911700904369354, + "learning_rate": 9.746573304036576e-05, + "loss": 0.0388, + "step": 14670 + }, + { + "epoch": 2.850055822532887, + "grad_norm": 0.1630418747663498, + "learning_rate": 9.746053313742412e-05, + "loss": 0.042, + "step": 14680 + }, + { + "epoch": 2.8519974758506867, + "grad_norm": 0.3124994933605194, + "learning_rate": 9.745532804426994e-05, + "loss": 0.0353, + "step": 14690 + }, + { + "epoch": 2.853939129168487, + "grad_norm": 0.21004684269428253, + "learning_rate": 9.745011776147242e-05, + "loss": 0.0412, + "step": 14700 + }, + { + "epoch": 2.8558807824862873, + "grad_norm": 0.16432420909404755, + "learning_rate": 9.744490228960138e-05, + "loss": 0.0358, + "step": 14710 + }, + { + "epoch": 2.857822435804087, + "grad_norm": 0.21510834991931915, + "learning_rate": 9.743968162922713e-05, + "loss": 0.0384, + "step": 14720 + }, + { + "epoch": 2.8597640891218874, + "grad_norm": 0.14891697466373444, + "learning_rate": 9.743445578092064e-05, + "loss": 0.0385, + "step": 14730 + }, + { + "epoch": 2.8617057424396872, + "grad_norm": 0.13457320630550385, + "learning_rate": 9.742922474525338e-05, + "loss": 0.0353, + "step": 14740 + }, + { + "epoch": 2.8636473957574875, + "grad_norm": 0.28935977816581726, + "learning_rate": 9.742398852279741e-05, + "loss": 0.0389, + "step": 14750 + }, + { + "epoch": 2.8655890490752878, + "grad_norm": 0.29885396361351013, + "learning_rate": 9.741874711412535e-05, + "loss": 0.0378, + "step": 14760 + }, + { + "epoch": 2.8675307023930876, + "grad_norm": 0.20319059491157532, + "learning_rate": 9.741350051981042e-05, + "loss": 0.0368, + "step": 14770 + }, + { + "epoch": 2.869472355710888, + "grad_norm": 0.23412702977657318, + "learning_rate": 9.740824874042633e-05, + "loss": 0.046, + "step": 14780 + }, + { + "epoch": 2.8714140090286877, + "grad_norm": 0.17523185908794403, + "learning_rate": 9.740299177654746e-05, + "loss": 0.0376, + "step": 14790 + }, + { + "epoch": 2.873355662346488, + "grad_norm": 0.20318104326725006, + "learning_rate": 9.739772962874867e-05, + "loss": 0.0316, + "step": 14800 + }, + { + "epoch": 2.8752973156642883, + "grad_norm": 0.15884248912334442, + "learning_rate": 9.739246229760541e-05, + "loss": 0.0285, + "step": 14810 + }, + { + "epoch": 2.877238968982088, + "grad_norm": 0.21871502697467804, + "learning_rate": 9.738718978369376e-05, + "loss": 0.0415, + "step": 14820 + }, + { + "epoch": 2.8791806222998884, + "grad_norm": 0.2605384290218353, + "learning_rate": 9.738191208759025e-05, + "loss": 0.0361, + "step": 14830 + }, + { + "epoch": 2.881122275617688, + "grad_norm": 0.21037335693836212, + "learning_rate": 9.73766292098721e-05, + "loss": 0.0413, + "step": 14840 + }, + { + "epoch": 2.8830639289354885, + "grad_norm": 0.2957265079021454, + "learning_rate": 9.737134115111699e-05, + "loss": 0.0394, + "step": 14850 + }, + { + "epoch": 2.8850055822532887, + "grad_norm": 0.19165079295635223, + "learning_rate": 9.736604791190323e-05, + "loss": 0.0378, + "step": 14860 + }, + { + "epoch": 2.886947235571089, + "grad_norm": 0.11553101986646652, + "learning_rate": 9.73607494928097e-05, + "loss": 0.0429, + "step": 14870 + }, + { + "epoch": 2.888888888888889, + "grad_norm": 0.2138199359178543, + "learning_rate": 9.735544589441581e-05, + "loss": 0.0425, + "step": 14880 + }, + { + "epoch": 2.890830542206689, + "grad_norm": 0.22463607788085938, + "learning_rate": 9.735013711730154e-05, + "loss": 0.0342, + "step": 14890 + }, + { + "epoch": 2.892772195524489, + "grad_norm": 0.550243616104126, + "learning_rate": 9.734482316204747e-05, + "loss": 0.0323, + "step": 14900 + }, + { + "epoch": 2.894713848842289, + "grad_norm": 0.274413526058197, + "learning_rate": 9.733950402923473e-05, + "loss": 0.0343, + "step": 14910 + }, + { + "epoch": 2.8966555021600895, + "grad_norm": 0.2974182367324829, + "learning_rate": 9.7334179719445e-05, + "loss": 0.052, + "step": 14920 + }, + { + "epoch": 2.8985971554778893, + "grad_norm": 0.16498877108097076, + "learning_rate": 9.732885023326053e-05, + "loss": 0.0331, + "step": 14930 + }, + { + "epoch": 2.9005388087956896, + "grad_norm": 0.17680683732032776, + "learning_rate": 9.732351557126418e-05, + "loss": 0.034, + "step": 14940 + }, + { + "epoch": 2.9024804621134894, + "grad_norm": 0.21415308117866516, + "learning_rate": 9.731817573403929e-05, + "loss": 0.0476, + "step": 14950 + }, + { + "epoch": 2.9044221154312897, + "grad_norm": 0.15444232523441315, + "learning_rate": 9.731283072216985e-05, + "loss": 0.0372, + "step": 14960 + }, + { + "epoch": 2.90636376874909, + "grad_norm": 0.22325384616851807, + "learning_rate": 9.730748053624039e-05, + "loss": 0.041, + "step": 14970 + }, + { + "epoch": 2.90830542206689, + "grad_norm": 0.24794696271419525, + "learning_rate": 9.730212517683598e-05, + "loss": 0.0442, + "step": 14980 + }, + { + "epoch": 2.91024707538469, + "grad_norm": 0.21238790452480316, + "learning_rate": 9.729676464454228e-05, + "loss": 0.0354, + "step": 14990 + }, + { + "epoch": 2.91218872870249, + "grad_norm": 0.14431262016296387, + "learning_rate": 9.72913989399455e-05, + "loss": 0.0425, + "step": 15000 + }, + { + "epoch": 2.91413038202029, + "grad_norm": 0.1566421240568161, + "learning_rate": 9.728602806363242e-05, + "loss": 0.0317, + "step": 15010 + }, + { + "epoch": 2.9160720353380905, + "grad_norm": 0.14628054201602936, + "learning_rate": 9.728065201619043e-05, + "loss": 0.0284, + "step": 15020 + }, + { + "epoch": 2.9180136886558907, + "grad_norm": 0.24085742235183716, + "learning_rate": 9.727527079820742e-05, + "loss": 0.048, + "step": 15030 + }, + { + "epoch": 2.9199553419736906, + "grad_norm": 0.31201305985450745, + "learning_rate": 9.726988441027186e-05, + "loss": 0.036, + "step": 15040 + }, + { + "epoch": 2.921896995291491, + "grad_norm": 0.2905929684638977, + "learning_rate": 9.726449285297281e-05, + "loss": 0.0327, + "step": 15050 + }, + { + "epoch": 2.9238386486092907, + "grad_norm": 0.27224764227867126, + "learning_rate": 9.72590961268999e-05, + "loss": 0.0424, + "step": 15060 + }, + { + "epoch": 2.925780301927091, + "grad_norm": 0.24597857892513275, + "learning_rate": 9.725369423264328e-05, + "loss": 0.0314, + "step": 15070 + }, + { + "epoch": 2.927721955244891, + "grad_norm": 0.21249431371688843, + "learning_rate": 9.72482871707937e-05, + "loss": 0.0344, + "step": 15080 + }, + { + "epoch": 2.929663608562691, + "grad_norm": 0.24412007629871368, + "learning_rate": 9.724287494194247e-05, + "loss": 0.0409, + "step": 15090 + }, + { + "epoch": 2.9316052618804913, + "grad_norm": 0.1795089691877365, + "learning_rate": 9.723745754668147e-05, + "loss": 0.034, + "step": 15100 + }, + { + "epoch": 2.933546915198291, + "grad_norm": 0.22791272401809692, + "learning_rate": 9.723203498560313e-05, + "loss": 0.0387, + "step": 15110 + }, + { + "epoch": 2.9354885685160914, + "grad_norm": 0.2821199595928192, + "learning_rate": 9.722660725930046e-05, + "loss": 0.0386, + "step": 15120 + }, + { + "epoch": 2.9374302218338917, + "grad_norm": 0.18719469010829926, + "learning_rate": 9.722117436836702e-05, + "loss": 0.0337, + "step": 15130 + }, + { + "epoch": 2.9393718751516915, + "grad_norm": 0.3100469708442688, + "learning_rate": 9.721573631339696e-05, + "loss": 0.0409, + "step": 15140 + }, + { + "epoch": 2.941313528469492, + "grad_norm": 0.2494862973690033, + "learning_rate": 9.721029309498494e-05, + "loss": 0.0421, + "step": 15150 + }, + { + "epoch": 2.9432551817872916, + "grad_norm": 0.4296823740005493, + "learning_rate": 9.720484471372627e-05, + "loss": 0.0503, + "step": 15160 + }, + { + "epoch": 2.945196835105092, + "grad_norm": 0.23276758193969727, + "learning_rate": 9.719939117021673e-05, + "loss": 0.037, + "step": 15170 + }, + { + "epoch": 2.947138488422892, + "grad_norm": 0.2427251935005188, + "learning_rate": 9.719393246505275e-05, + "loss": 0.0356, + "step": 15180 + }, + { + "epoch": 2.9490801417406924, + "grad_norm": 0.2579660713672638, + "learning_rate": 9.718846859883128e-05, + "loss": 0.0375, + "step": 15190 + }, + { + "epoch": 2.9510217950584923, + "grad_norm": 0.22033533453941345, + "learning_rate": 9.718299957214982e-05, + "loss": 0.0361, + "step": 15200 + }, + { + "epoch": 2.9529634483762925, + "grad_norm": 0.18453843891620636, + "learning_rate": 9.717752538560646e-05, + "loss": 0.0288, + "step": 15210 + }, + { + "epoch": 2.9549051016940924, + "grad_norm": 0.24588841199874878, + "learning_rate": 9.717204603979986e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 2.9568467550118926, + "grad_norm": 0.21578195691108704, + "learning_rate": 9.716656153532922e-05, + "loss": 0.0381, + "step": 15230 + }, + { + "epoch": 2.958788408329693, + "grad_norm": 0.14484238624572754, + "learning_rate": 9.716107187279434e-05, + "loss": 0.0336, + "step": 15240 + }, + { + "epoch": 2.9607300616474928, + "grad_norm": 0.18966223299503326, + "learning_rate": 9.715557705279555e-05, + "loss": 0.0377, + "step": 15250 + }, + { + "epoch": 2.962671714965293, + "grad_norm": 0.1923038810491562, + "learning_rate": 9.715007707593372e-05, + "loss": 0.0347, + "step": 15260 + }, + { + "epoch": 2.964613368283093, + "grad_norm": 0.2828550338745117, + "learning_rate": 9.714457194281036e-05, + "loss": 0.0386, + "step": 15270 + }, + { + "epoch": 2.966555021600893, + "grad_norm": 0.17861433327198029, + "learning_rate": 9.713906165402751e-05, + "loss": 0.0358, + "step": 15280 + }, + { + "epoch": 2.9684966749186934, + "grad_norm": 0.1724475920200348, + "learning_rate": 9.713354621018774e-05, + "loss": 0.0458, + "step": 15290 + }, + { + "epoch": 2.9704383282364932, + "grad_norm": 0.3372080326080322, + "learning_rate": 9.712802561189422e-05, + "loss": 0.0398, + "step": 15300 + }, + { + "epoch": 2.9723799815542935, + "grad_norm": 0.22269776463508606, + "learning_rate": 9.712249985975069e-05, + "loss": 0.0318, + "step": 15310 + }, + { + "epoch": 2.9743216348720933, + "grad_norm": 0.2644989490509033, + "learning_rate": 9.71169689543614e-05, + "loss": 0.0357, + "step": 15320 + }, + { + "epoch": 2.9762632881898936, + "grad_norm": 0.25211963057518005, + "learning_rate": 9.711143289633123e-05, + "loss": 0.0371, + "step": 15330 + }, + { + "epoch": 2.978204941507694, + "grad_norm": 0.18106235563755035, + "learning_rate": 9.710589168626561e-05, + "loss": 0.0366, + "step": 15340 + }, + { + "epoch": 2.980146594825494, + "grad_norm": 0.3257196843624115, + "learning_rate": 9.710034532477048e-05, + "loss": 0.0378, + "step": 15350 + }, + { + "epoch": 2.982088248143294, + "grad_norm": 0.17488667368888855, + "learning_rate": 9.709479381245239e-05, + "loss": 0.0414, + "step": 15360 + }, + { + "epoch": 2.9840299014610943, + "grad_norm": 0.1934822052717209, + "learning_rate": 9.708923714991847e-05, + "loss": 0.0389, + "step": 15370 + }, + { + "epoch": 2.985971554778894, + "grad_norm": 0.23546265065670013, + "learning_rate": 9.708367533777638e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 2.9879132080966944, + "grad_norm": 0.2480507642030716, + "learning_rate": 9.707810837663431e-05, + "loss": 0.0372, + "step": 15390 + }, + { + "epoch": 2.9898548614144946, + "grad_norm": 0.2402900606393814, + "learning_rate": 9.707253626710113e-05, + "loss": 0.0352, + "step": 15400 + }, + { + "epoch": 2.9917965147322945, + "grad_norm": 0.340287446975708, + "learning_rate": 9.706695900978613e-05, + "loss": 0.0389, + "step": 15410 + }, + { + "epoch": 2.9937381680500947, + "grad_norm": 0.23778073489665985, + "learning_rate": 9.706137660529926e-05, + "loss": 0.0396, + "step": 15420 + }, + { + "epoch": 2.9956798213678946, + "grad_norm": 0.25633418560028076, + "learning_rate": 9.705578905425101e-05, + "loss": 0.0372, + "step": 15430 + }, + { + "epoch": 2.997621474685695, + "grad_norm": 0.21004487574100494, + "learning_rate": 9.705019635725241e-05, + "loss": 0.0362, + "step": 15440 + }, + { + "epoch": 2.999563128003495, + "grad_norm": 0.1788327544927597, + "learning_rate": 9.704459851491508e-05, + "loss": 0.0421, + "step": 15450 + }, + { + "epoch": 3.0013591573224603, + "grad_norm": 0.3305857479572296, + "learning_rate": 9.703899552785118e-05, + "loss": 0.0456, + "step": 15460 + }, + { + "epoch": 3.00330081064026, + "grad_norm": 0.16319029033184052, + "learning_rate": 9.703338739667346e-05, + "loss": 0.0365, + "step": 15470 + }, + { + "epoch": 3.0052424639580604, + "grad_norm": 0.2231966108083725, + "learning_rate": 9.70277741219952e-05, + "loss": 0.0409, + "step": 15480 + }, + { + "epoch": 3.00718411727586, + "grad_norm": 0.2660413980484009, + "learning_rate": 9.702215570443027e-05, + "loss": 0.0383, + "step": 15490 + }, + { + "epoch": 3.0091257705936605, + "grad_norm": 0.23922209441661835, + "learning_rate": 9.701653214459309e-05, + "loss": 0.0361, + "step": 15500 + }, + { + "epoch": 3.0110674239114608, + "grad_norm": 0.21643374860286713, + "learning_rate": 9.701090344309865e-05, + "loss": 0.0367, + "step": 15510 + }, + { + "epoch": 3.0130090772292606, + "grad_norm": 0.2891300320625305, + "learning_rate": 9.700526960056247e-05, + "loss": 0.0339, + "step": 15520 + }, + { + "epoch": 3.014950730547061, + "grad_norm": 0.21930204331874847, + "learning_rate": 9.699963061760068e-05, + "loss": 0.0338, + "step": 15530 + }, + { + "epoch": 3.016892383864861, + "grad_norm": 0.211860790848732, + "learning_rate": 9.699398649482997e-05, + "loss": 0.0328, + "step": 15540 + }, + { + "epoch": 3.018834037182661, + "grad_norm": 0.18025244772434235, + "learning_rate": 9.698833723286753e-05, + "loss": 0.0353, + "step": 15550 + }, + { + "epoch": 3.0207756905004612, + "grad_norm": 0.21995964646339417, + "learning_rate": 9.698268283233118e-05, + "loss": 0.038, + "step": 15560 + }, + { + "epoch": 3.022717343818261, + "grad_norm": 0.23934157192707062, + "learning_rate": 9.697702329383929e-05, + "loss": 0.0375, + "step": 15570 + }, + { + "epoch": 3.0246589971360613, + "grad_norm": 0.20732346177101135, + "learning_rate": 9.697135861801074e-05, + "loss": 0.036, + "step": 15580 + }, + { + "epoch": 3.0266006504538616, + "grad_norm": 0.1816892921924591, + "learning_rate": 9.696568880546505e-05, + "loss": 0.046, + "step": 15590 + }, + { + "epoch": 3.0285423037716614, + "grad_norm": 0.2137608826160431, + "learning_rate": 9.696001385682223e-05, + "loss": 0.0308, + "step": 15600 + }, + { + "epoch": 3.0304839570894617, + "grad_norm": 0.2756574749946594, + "learning_rate": 9.695433377270291e-05, + "loss": 0.0334, + "step": 15610 + }, + { + "epoch": 3.032425610407262, + "grad_norm": 0.2534341514110565, + "learning_rate": 9.694864855372824e-05, + "loss": 0.0389, + "step": 15620 + }, + { + "epoch": 3.034367263725062, + "grad_norm": 0.3636474907398224, + "learning_rate": 9.694295820051995e-05, + "loss": 0.0402, + "step": 15630 + }, + { + "epoch": 3.036308917042862, + "grad_norm": 0.16920483112335205, + "learning_rate": 9.693726271370032e-05, + "loss": 0.0378, + "step": 15640 + }, + { + "epoch": 3.038250570360662, + "grad_norm": 0.2614116370677948, + "learning_rate": 9.693156209389221e-05, + "loss": 0.0396, + "step": 15650 + }, + { + "epoch": 3.040192223678462, + "grad_norm": 0.29311883449554443, + "learning_rate": 9.692585634171905e-05, + "loss": 0.0504, + "step": 15660 + }, + { + "epoch": 3.0421338769962625, + "grad_norm": 0.22987915575504303, + "learning_rate": 9.692014545780476e-05, + "loss": 0.0375, + "step": 15670 + }, + { + "epoch": 3.0440755303140623, + "grad_norm": 0.15786977112293243, + "learning_rate": 9.691442944277393e-05, + "loss": 0.0388, + "step": 15680 + }, + { + "epoch": 3.0460171836318626, + "grad_norm": 0.2703127861022949, + "learning_rate": 9.690870829725162e-05, + "loss": 0.0421, + "step": 15690 + }, + { + "epoch": 3.047958836949663, + "grad_norm": 0.1660545915365219, + "learning_rate": 9.69029820218635e-05, + "loss": 0.0328, + "step": 15700 + }, + { + "epoch": 3.0499004902674627, + "grad_norm": 0.16714492440223694, + "learning_rate": 9.689725061723579e-05, + "loss": 0.0378, + "step": 15710 + }, + { + "epoch": 3.051842143585263, + "grad_norm": 0.1934463083744049, + "learning_rate": 9.689151408399527e-05, + "loss": 0.0379, + "step": 15720 + }, + { + "epoch": 3.053783796903063, + "grad_norm": 0.24797217547893524, + "learning_rate": 9.688577242276924e-05, + "loss": 0.0327, + "step": 15730 + }, + { + "epoch": 3.055725450220863, + "grad_norm": 0.1653146594762802, + "learning_rate": 9.688002563418566e-05, + "loss": 0.0316, + "step": 15740 + }, + { + "epoch": 3.0576671035386633, + "grad_norm": 0.1524249017238617, + "learning_rate": 9.687427371887293e-05, + "loss": 0.0372, + "step": 15750 + }, + { + "epoch": 3.059608756856463, + "grad_norm": 0.15125662088394165, + "learning_rate": 9.686851667746012e-05, + "loss": 0.0354, + "step": 15760 + }, + { + "epoch": 3.0615504101742634, + "grad_norm": 0.20020584762096405, + "learning_rate": 9.686275451057677e-05, + "loss": 0.0402, + "step": 15770 + }, + { + "epoch": 3.0634920634920633, + "grad_norm": 0.3480217754840851, + "learning_rate": 9.685698721885308e-05, + "loss": 0.0325, + "step": 15780 + }, + { + "epoch": 3.0654337168098635, + "grad_norm": 0.2625987231731415, + "learning_rate": 9.68512148029197e-05, + "loss": 0.0381, + "step": 15790 + }, + { + "epoch": 3.067375370127664, + "grad_norm": 0.21376760303974152, + "learning_rate": 9.684543726340791e-05, + "loss": 0.035, + "step": 15800 + }, + { + "epoch": 3.0693170234454636, + "grad_norm": 0.20543529093265533, + "learning_rate": 9.683965460094952e-05, + "loss": 0.038, + "step": 15810 + }, + { + "epoch": 3.071258676763264, + "grad_norm": 0.2275961935520172, + "learning_rate": 9.683386681617694e-05, + "loss": 0.0388, + "step": 15820 + }, + { + "epoch": 3.073200330081064, + "grad_norm": 0.24596379697322845, + "learning_rate": 9.68280739097231e-05, + "loss": 0.0395, + "step": 15830 + }, + { + "epoch": 3.075141983398864, + "grad_norm": 0.2993042469024658, + "learning_rate": 9.682227588222148e-05, + "loss": 0.0327, + "step": 15840 + }, + { + "epoch": 3.0770836367166643, + "grad_norm": 0.2394229918718338, + "learning_rate": 9.681647273430618e-05, + "loss": 0.0337, + "step": 15850 + }, + { + "epoch": 3.0790252900344646, + "grad_norm": 0.2060711830854416, + "learning_rate": 9.681066446661182e-05, + "loss": 0.0372, + "step": 15860 + }, + { + "epoch": 3.0809669433522644, + "grad_norm": 0.2567727267742157, + "learning_rate": 9.680485107977357e-05, + "loss": 0.0367, + "step": 15870 + }, + { + "epoch": 3.0829085966700647, + "grad_norm": 0.19484879076480865, + "learning_rate": 9.679903257442716e-05, + "loss": 0.0411, + "step": 15880 + }, + { + "epoch": 3.0848502499878645, + "grad_norm": 0.2504352331161499, + "learning_rate": 9.679320895120891e-05, + "loss": 0.0388, + "step": 15890 + }, + { + "epoch": 3.0867919033056648, + "grad_norm": 0.2702738642692566, + "learning_rate": 9.67873802107557e-05, + "loss": 0.0511, + "step": 15900 + }, + { + "epoch": 3.088733556623465, + "grad_norm": 0.30620628595352173, + "learning_rate": 9.67815463537049e-05, + "loss": 0.0346, + "step": 15910 + }, + { + "epoch": 3.090675209941265, + "grad_norm": 0.20674914121627808, + "learning_rate": 9.677570738069457e-05, + "loss": 0.0382, + "step": 15920 + }, + { + "epoch": 3.092616863259065, + "grad_norm": 0.16557686030864716, + "learning_rate": 9.676986329236318e-05, + "loss": 0.0392, + "step": 15930 + }, + { + "epoch": 3.094558516576865, + "grad_norm": 0.1537196785211563, + "learning_rate": 9.676401408934987e-05, + "loss": 0.0337, + "step": 15940 + }, + { + "epoch": 3.0965001698946653, + "grad_norm": 0.1811198592185974, + "learning_rate": 9.675815977229428e-05, + "loss": 0.0404, + "step": 15950 + }, + { + "epoch": 3.0984418232124655, + "grad_norm": 0.19991175830364227, + "learning_rate": 9.675230034183664e-05, + "loss": 0.0383, + "step": 15960 + }, + { + "epoch": 3.1003834765302654, + "grad_norm": 0.23265352845191956, + "learning_rate": 9.674643579861773e-05, + "loss": 0.0443, + "step": 15970 + }, + { + "epoch": 3.1023251298480656, + "grad_norm": 0.3376545011997223, + "learning_rate": 9.674056614327886e-05, + "loss": 0.0403, + "step": 15980 + }, + { + "epoch": 3.104266783165866, + "grad_norm": 0.2614001929759979, + "learning_rate": 9.673469137646198e-05, + "loss": 0.0368, + "step": 15990 + }, + { + "epoch": 3.1062084364836657, + "grad_norm": 0.2188001424074173, + "learning_rate": 9.67288114988095e-05, + "loss": 0.0395, + "step": 16000 + }, + { + "epoch": 3.108150089801466, + "grad_norm": 0.3234338164329529, + "learning_rate": 9.672292651096447e-05, + "loss": 0.0366, + "step": 16010 + }, + { + "epoch": 3.1100917431192663, + "grad_norm": 0.2993229329586029, + "learning_rate": 9.671703641357042e-05, + "loss": 0.0408, + "step": 16020 + }, + { + "epoch": 3.112033396437066, + "grad_norm": 0.1498977392911911, + "learning_rate": 9.67111412072715e-05, + "loss": 0.0382, + "step": 16030 + }, + { + "epoch": 3.1139750497548664, + "grad_norm": 0.1403276026248932, + "learning_rate": 9.670524089271242e-05, + "loss": 0.0421, + "step": 16040 + }, + { + "epoch": 3.115916703072666, + "grad_norm": 0.1790028214454651, + "learning_rate": 9.669933547053842e-05, + "loss": 0.0386, + "step": 16050 + }, + { + "epoch": 3.1178583563904665, + "grad_norm": 0.17343486845493317, + "learning_rate": 9.669342494139531e-05, + "loss": 0.0343, + "step": 16060 + }, + { + "epoch": 3.1198000097082668, + "grad_norm": 0.2216900736093521, + "learning_rate": 9.668750930592943e-05, + "loss": 0.0394, + "step": 16070 + }, + { + "epoch": 3.1217416630260666, + "grad_norm": 0.2265598326921463, + "learning_rate": 9.668158856478775e-05, + "loss": 0.0291, + "step": 16080 + }, + { + "epoch": 3.123683316343867, + "grad_norm": 0.259065717458725, + "learning_rate": 9.66756627186177e-05, + "loss": 0.0371, + "step": 16090 + }, + { + "epoch": 3.1256249696616667, + "grad_norm": 0.1633385270833969, + "learning_rate": 9.666973176806737e-05, + "loss": 0.0332, + "step": 16100 + }, + { + "epoch": 3.127566622979467, + "grad_norm": 0.1801491230726242, + "learning_rate": 9.666379571378534e-05, + "loss": 0.0408, + "step": 16110 + }, + { + "epoch": 3.1295082762972672, + "grad_norm": 0.2537270486354828, + "learning_rate": 9.665785455642076e-05, + "loss": 0.0412, + "step": 16120 + }, + { + "epoch": 3.131449929615067, + "grad_norm": 0.14508214592933655, + "learning_rate": 9.665190829662337e-05, + "loss": 0.0353, + "step": 16130 + }, + { + "epoch": 3.1333915829328673, + "grad_norm": 0.12717527151107788, + "learning_rate": 9.664595693504342e-05, + "loss": 0.033, + "step": 16140 + }, + { + "epoch": 3.1353332362506676, + "grad_norm": 0.2072364091873169, + "learning_rate": 9.664000047233175e-05, + "loss": 0.033, + "step": 16150 + }, + { + "epoch": 3.1372748895684675, + "grad_norm": 0.1843695044517517, + "learning_rate": 9.663403890913976e-05, + "loss": 0.0337, + "step": 16160 + }, + { + "epoch": 3.1392165428862677, + "grad_norm": 0.3413967192173004, + "learning_rate": 9.662807224611938e-05, + "loss": 0.0379, + "step": 16170 + }, + { + "epoch": 3.141158196204068, + "grad_norm": 0.1854998916387558, + "learning_rate": 9.662210048392311e-05, + "loss": 0.0349, + "step": 16180 + }, + { + "epoch": 3.143099849521868, + "grad_norm": 0.25129371881484985, + "learning_rate": 9.661612362320405e-05, + "loss": 0.0377, + "step": 16190 + }, + { + "epoch": 3.145041502839668, + "grad_norm": 0.16032840311527252, + "learning_rate": 9.661014166461579e-05, + "loss": 0.0306, + "step": 16200 + }, + { + "epoch": 3.146983156157468, + "grad_norm": 0.2726815640926361, + "learning_rate": 9.66041546088125e-05, + "loss": 0.0356, + "step": 16210 + }, + { + "epoch": 3.148924809475268, + "grad_norm": 0.2467789351940155, + "learning_rate": 9.659816245644895e-05, + "loss": 0.0362, + "step": 16220 + }, + { + "epoch": 3.1508664627930685, + "grad_norm": 0.1869688332080841, + "learning_rate": 9.65921652081804e-05, + "loss": 0.029, + "step": 16230 + }, + { + "epoch": 3.1528081161108683, + "grad_norm": 0.1676536351442337, + "learning_rate": 9.658616286466271e-05, + "loss": 0.0409, + "step": 16240 + }, + { + "epoch": 3.1547497694286686, + "grad_norm": 0.3809219002723694, + "learning_rate": 9.65801554265523e-05, + "loss": 0.0313, + "step": 16250 + }, + { + "epoch": 3.1566914227464684, + "grad_norm": 0.20359691977500916, + "learning_rate": 9.657414289450612e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 3.1586330760642687, + "grad_norm": 0.13335952162742615, + "learning_rate": 9.656812526918171e-05, + "loss": 0.0303, + "step": 16270 + }, + { + "epoch": 3.160574729382069, + "grad_norm": 0.21017147600650787, + "learning_rate": 9.656210255123712e-05, + "loss": 0.0463, + "step": 16280 + }, + { + "epoch": 3.162516382699869, + "grad_norm": 0.15363547205924988, + "learning_rate": 9.6556074741331e-05, + "loss": 0.0369, + "step": 16290 + }, + { + "epoch": 3.164458036017669, + "grad_norm": 0.2340639978647232, + "learning_rate": 9.655004184012256e-05, + "loss": 0.0324, + "step": 16300 + }, + { + "epoch": 3.1663996893354693, + "grad_norm": 0.27709290385246277, + "learning_rate": 9.654400384827152e-05, + "loss": 0.0419, + "step": 16310 + }, + { + "epoch": 3.168341342653269, + "grad_norm": 0.27256470918655396, + "learning_rate": 9.653796076643818e-05, + "loss": 0.0366, + "step": 16320 + }, + { + "epoch": 3.1702829959710694, + "grad_norm": 0.18032650649547577, + "learning_rate": 9.653191259528344e-05, + "loss": 0.0341, + "step": 16330 + }, + { + "epoch": 3.1722246492888693, + "grad_norm": 0.2656252384185791, + "learning_rate": 9.65258593354687e-05, + "loss": 0.0378, + "step": 16340 + }, + { + "epoch": 3.1741663026066695, + "grad_norm": 0.19389311969280243, + "learning_rate": 9.651980098765591e-05, + "loss": 0.0376, + "step": 16350 + }, + { + "epoch": 3.17610795592447, + "grad_norm": 0.18279239535331726, + "learning_rate": 9.651373755250765e-05, + "loss": 0.0417, + "step": 16360 + }, + { + "epoch": 3.1780496092422696, + "grad_norm": 0.3128945231437683, + "learning_rate": 9.650766903068697e-05, + "loss": 0.0378, + "step": 16370 + }, + { + "epoch": 3.17999126256007, + "grad_norm": 0.32739636301994324, + "learning_rate": 9.650159542285753e-05, + "loss": 0.0349, + "step": 16380 + }, + { + "epoch": 3.18193291587787, + "grad_norm": 0.2186892330646515, + "learning_rate": 9.649551672968353e-05, + "loss": 0.0338, + "step": 16390 + }, + { + "epoch": 3.18387456919567, + "grad_norm": 0.22319160401821136, + "learning_rate": 9.648943295182973e-05, + "loss": 0.0344, + "step": 16400 + }, + { + "epoch": 3.1858162225134703, + "grad_norm": 0.19712907075881958, + "learning_rate": 9.648334408996144e-05, + "loss": 0.0342, + "step": 16410 + }, + { + "epoch": 3.18775787583127, + "grad_norm": 0.23248505592346191, + "learning_rate": 9.647725014474452e-05, + "loss": 0.0327, + "step": 16420 + }, + { + "epoch": 3.1896995291490704, + "grad_norm": 0.20281434059143066, + "learning_rate": 9.64711511168454e-05, + "loss": 0.0387, + "step": 16430 + }, + { + "epoch": 3.1916411824668707, + "grad_norm": 0.18165545165538788, + "learning_rate": 9.646504700693108e-05, + "loss": 0.0321, + "step": 16440 + }, + { + "epoch": 3.1935828357846705, + "grad_norm": 0.32430359721183777, + "learning_rate": 9.645893781566907e-05, + "loss": 0.036, + "step": 16450 + }, + { + "epoch": 3.1955244891024708, + "grad_norm": 0.36816319823265076, + "learning_rate": 9.645282354372744e-05, + "loss": 0.0362, + "step": 16460 + }, + { + "epoch": 3.197466142420271, + "grad_norm": 0.22636255621910095, + "learning_rate": 9.644670419177491e-05, + "loss": 0.0347, + "step": 16470 + }, + { + "epoch": 3.199407795738071, + "grad_norm": 0.1565445363521576, + "learning_rate": 9.644057976048062e-05, + "loss": 0.04, + "step": 16480 + }, + { + "epoch": 3.201349449055871, + "grad_norm": 0.16422751545906067, + "learning_rate": 9.643445025051435e-05, + "loss": 0.0336, + "step": 16490 + }, + { + "epoch": 3.203291102373671, + "grad_norm": 0.18277403712272644, + "learning_rate": 9.642831566254641e-05, + "loss": 0.0345, + "step": 16500 + }, + { + "epoch": 3.2052327556914713, + "grad_norm": 0.25554680824279785, + "learning_rate": 9.642217599724769e-05, + "loss": 0.0393, + "step": 16510 + }, + { + "epoch": 3.2071744090092715, + "grad_norm": 0.24713082611560822, + "learning_rate": 9.64160312552896e-05, + "loss": 0.0345, + "step": 16520 + }, + { + "epoch": 3.2091160623270714, + "grad_norm": 0.2028665989637375, + "learning_rate": 9.64098814373441e-05, + "loss": 0.0456, + "step": 16530 + }, + { + "epoch": 3.2110577156448716, + "grad_norm": 0.23469515144824982, + "learning_rate": 9.640372654408374e-05, + "loss": 0.0371, + "step": 16540 + }, + { + "epoch": 3.212999368962672, + "grad_norm": 0.2714433968067169, + "learning_rate": 9.639756657618162e-05, + "loss": 0.0469, + "step": 16550 + }, + { + "epoch": 3.2149410222804717, + "grad_norm": 0.18659400939941406, + "learning_rate": 9.639140153431138e-05, + "loss": 0.0305, + "step": 16560 + }, + { + "epoch": 3.216882675598272, + "grad_norm": 0.21991287171840668, + "learning_rate": 9.638523141914721e-05, + "loss": 0.0395, + "step": 16570 + }, + { + "epoch": 3.218824328916072, + "grad_norm": 0.24652153253555298, + "learning_rate": 9.637905623136388e-05, + "loss": 0.0327, + "step": 16580 + }, + { + "epoch": 3.220765982233872, + "grad_norm": 0.19893454015254974, + "learning_rate": 9.637287597163669e-05, + "loss": 0.0369, + "step": 16590 + }, + { + "epoch": 3.2227076355516724, + "grad_norm": 0.15707436203956604, + "learning_rate": 9.63666906406415e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 3.224649288869472, + "grad_norm": 0.1911407709121704, + "learning_rate": 9.636050023905473e-05, + "loss": 0.0351, + "step": 16610 + }, + { + "epoch": 3.2265909421872725, + "grad_norm": 0.18510936200618744, + "learning_rate": 9.635430476755336e-05, + "loss": 0.0455, + "step": 16620 + }, + { + "epoch": 3.2285325955050728, + "grad_norm": 0.1944964975118637, + "learning_rate": 9.63481042268149e-05, + "loss": 0.0395, + "step": 16630 + }, + { + "epoch": 3.2304742488228726, + "grad_norm": 0.2128075659275055, + "learning_rate": 9.634189861751745e-05, + "loss": 0.0388, + "step": 16640 + }, + { + "epoch": 3.232415902140673, + "grad_norm": 0.1711507886648178, + "learning_rate": 9.633568794033967e-05, + "loss": 0.0386, + "step": 16650 + }, + { + "epoch": 3.2343575554584727, + "grad_norm": 0.15657928586006165, + "learning_rate": 9.63294721959607e-05, + "loss": 0.038, + "step": 16660 + }, + { + "epoch": 3.236299208776273, + "grad_norm": 0.17088259756565094, + "learning_rate": 9.63232513850603e-05, + "loss": 0.0378, + "step": 16670 + }, + { + "epoch": 3.2382408620940732, + "grad_norm": 0.20192265510559082, + "learning_rate": 9.631702550831878e-05, + "loss": 0.0366, + "step": 16680 + }, + { + "epoch": 3.240182515411873, + "grad_norm": 0.25673580169677734, + "learning_rate": 9.631079456641698e-05, + "loss": 0.0454, + "step": 16690 + }, + { + "epoch": 3.2421241687296734, + "grad_norm": 0.252910852432251, + "learning_rate": 9.630455856003632e-05, + "loss": 0.0388, + "step": 16700 + }, + { + "epoch": 3.2440658220474736, + "grad_norm": 0.1728261560201645, + "learning_rate": 9.629831748985876e-05, + "loss": 0.0351, + "step": 16710 + }, + { + "epoch": 3.2460074753652735, + "grad_norm": 0.2379225343465805, + "learning_rate": 9.629207135656679e-05, + "loss": 0.0373, + "step": 16720 + }, + { + "epoch": 3.2479491286830737, + "grad_norm": 0.1921250820159912, + "learning_rate": 9.628582016084353e-05, + "loss": 0.0351, + "step": 16730 + }, + { + "epoch": 3.2498907820008736, + "grad_norm": 0.16296255588531494, + "learning_rate": 9.627956390337254e-05, + "loss": 0.0327, + "step": 16740 + }, + { + "epoch": 3.251832435318674, + "grad_norm": 0.20507711172103882, + "learning_rate": 9.627330258483802e-05, + "loss": 0.0356, + "step": 16750 + }, + { + "epoch": 3.253774088636474, + "grad_norm": 0.3137287497520447, + "learning_rate": 9.62670362059247e-05, + "loss": 0.0344, + "step": 16760 + }, + { + "epoch": 3.255715741954274, + "grad_norm": 0.2800579369068146, + "learning_rate": 9.626076476731786e-05, + "loss": 0.0359, + "step": 16770 + }, + { + "epoch": 3.257657395272074, + "grad_norm": 0.18329423666000366, + "learning_rate": 9.625448826970336e-05, + "loss": 0.0364, + "step": 16780 + }, + { + "epoch": 3.259599048589874, + "grad_norm": 0.11375905573368073, + "learning_rate": 9.624820671376755e-05, + "loss": 0.0338, + "step": 16790 + }, + { + "epoch": 3.2615407019076743, + "grad_norm": 0.1271248608827591, + "learning_rate": 9.62419201001974e-05, + "loss": 0.0293, + "step": 16800 + }, + { + "epoch": 3.2634823552254746, + "grad_norm": 0.1420639008283615, + "learning_rate": 9.623562842968037e-05, + "loss": 0.03, + "step": 16810 + }, + { + "epoch": 3.2654240085432744, + "grad_norm": 0.231108620762825, + "learning_rate": 9.622933170290454e-05, + "loss": 0.0347, + "step": 16820 + }, + { + "epoch": 3.2673656618610747, + "grad_norm": 0.23094332218170166, + "learning_rate": 9.622302992055849e-05, + "loss": 0.0414, + "step": 16830 + }, + { + "epoch": 3.269307315178875, + "grad_norm": 0.13782794773578644, + "learning_rate": 9.62167230833314e-05, + "loss": 0.0386, + "step": 16840 + }, + { + "epoch": 3.271248968496675, + "grad_norm": 0.18135830760002136, + "learning_rate": 9.621041119191295e-05, + "loss": 0.031, + "step": 16850 + }, + { + "epoch": 3.273190621814475, + "grad_norm": 0.16881699860095978, + "learning_rate": 9.620409424699342e-05, + "loss": 0.0327, + "step": 16860 + }, + { + "epoch": 3.2751322751322753, + "grad_norm": 0.14436785876750946, + "learning_rate": 9.619777224926359e-05, + "loss": 0.0327, + "step": 16870 + }, + { + "epoch": 3.277073928450075, + "grad_norm": 0.24791455268859863, + "learning_rate": 9.619144519941485e-05, + "loss": 0.0302, + "step": 16880 + }, + { + "epoch": 3.2790155817678754, + "grad_norm": 0.20010371506214142, + "learning_rate": 9.618511309813912e-05, + "loss": 0.0337, + "step": 16890 + }, + { + "epoch": 3.2809572350856753, + "grad_norm": 0.2213602513074875, + "learning_rate": 9.617877594612886e-05, + "loss": 0.0301, + "step": 16900 + }, + { + "epoch": 3.2828988884034755, + "grad_norm": 0.17058290541172028, + "learning_rate": 9.617243374407707e-05, + "loss": 0.0348, + "step": 16910 + }, + { + "epoch": 3.284840541721276, + "grad_norm": 0.202434241771698, + "learning_rate": 9.616608649267736e-05, + "loss": 0.0284, + "step": 16920 + }, + { + "epoch": 3.2867821950390756, + "grad_norm": 0.23823216557502747, + "learning_rate": 9.615973419262385e-05, + "loss": 0.0367, + "step": 16930 + }, + { + "epoch": 3.288723848356876, + "grad_norm": 0.18467195332050323, + "learning_rate": 9.615337684461119e-05, + "loss": 0.0325, + "step": 16940 + }, + { + "epoch": 3.2906655016746758, + "grad_norm": 0.23833660781383514, + "learning_rate": 9.614701444933465e-05, + "loss": 0.037, + "step": 16950 + }, + { + "epoch": 3.292607154992476, + "grad_norm": 0.23326672613620758, + "learning_rate": 9.614064700748997e-05, + "loss": 0.0395, + "step": 16960 + }, + { + "epoch": 3.2945488083102763, + "grad_norm": 0.10755740851163864, + "learning_rate": 9.613427451977352e-05, + "loss": 0.0375, + "step": 16970 + }, + { + "epoch": 3.296490461628076, + "grad_norm": 0.2546355128288269, + "learning_rate": 9.612789698688216e-05, + "loss": 0.0442, + "step": 16980 + }, + { + "epoch": 3.2984321149458764, + "grad_norm": 0.214040145277977, + "learning_rate": 9.612151440951334e-05, + "loss": 0.0303, + "step": 16990 + }, + { + "epoch": 3.3003737682636767, + "grad_norm": 0.3237748146057129, + "learning_rate": 9.611512678836506e-05, + "loss": 0.0354, + "step": 17000 + }, + { + "epoch": 3.3023154215814765, + "grad_norm": 0.1488879919052124, + "learning_rate": 9.610873412413584e-05, + "loss": 0.0393, + "step": 17010 + }, + { + "epoch": 3.304257074899277, + "grad_norm": 0.26254576444625854, + "learning_rate": 9.610233641752476e-05, + "loss": 0.0377, + "step": 17020 + }, + { + "epoch": 3.306198728217077, + "grad_norm": 0.11265778541564941, + "learning_rate": 9.609593366923151e-05, + "loss": 0.0321, + "step": 17030 + }, + { + "epoch": 3.308140381534877, + "grad_norm": 0.1787158101797104, + "learning_rate": 9.608952587995625e-05, + "loss": 0.034, + "step": 17040 + }, + { + "epoch": 3.310082034852677, + "grad_norm": 0.20045919716358185, + "learning_rate": 9.608311305039972e-05, + "loss": 0.0339, + "step": 17050 + }, + { + "epoch": 3.312023688170477, + "grad_norm": 0.2248355746269226, + "learning_rate": 9.607669518126326e-05, + "loss": 0.0335, + "step": 17060 + }, + { + "epoch": 3.3139653414882773, + "grad_norm": 0.1529293805360794, + "learning_rate": 9.607027227324866e-05, + "loss": 0.0314, + "step": 17070 + }, + { + "epoch": 3.3159069948060775, + "grad_norm": 0.266495943069458, + "learning_rate": 9.606384432705837e-05, + "loss": 0.0336, + "step": 17080 + }, + { + "epoch": 3.3178486481238774, + "grad_norm": 0.22968000173568726, + "learning_rate": 9.60574113433953e-05, + "loss": 0.0372, + "step": 17090 + }, + { + "epoch": 3.3197903014416776, + "grad_norm": 0.17173868417739868, + "learning_rate": 9.6050973322963e-05, + "loss": 0.0321, + "step": 17100 + }, + { + "epoch": 3.3217319547594775, + "grad_norm": 0.14872616529464722, + "learning_rate": 9.604453026646547e-05, + "loss": 0.0425, + "step": 17110 + }, + { + "epoch": 3.3236736080772777, + "grad_norm": 0.12189745903015137, + "learning_rate": 9.603808217460735e-05, + "loss": 0.0292, + "step": 17120 + }, + { + "epoch": 3.325615261395078, + "grad_norm": 0.20577707886695862, + "learning_rate": 9.603162904809377e-05, + "loss": 0.038, + "step": 17130 + }, + { + "epoch": 3.327556914712878, + "grad_norm": 0.21882911026477814, + "learning_rate": 9.602517088763045e-05, + "loss": 0.0399, + "step": 17140 + }, + { + "epoch": 3.329498568030678, + "grad_norm": 0.2233012318611145, + "learning_rate": 9.601870769392365e-05, + "loss": 0.0347, + "step": 17150 + }, + { + "epoch": 3.3314402213484784, + "grad_norm": 0.12794533371925354, + "learning_rate": 9.601223946768017e-05, + "loss": 0.04, + "step": 17160 + }, + { + "epoch": 3.333381874666278, + "grad_norm": 0.11574213951826096, + "learning_rate": 9.600576620960734e-05, + "loss": 0.029, + "step": 17170 + }, + { + "epoch": 3.3353235279840785, + "grad_norm": 0.26341819763183594, + "learning_rate": 9.599928792041308e-05, + "loss": 0.0353, + "step": 17180 + }, + { + "epoch": 3.3372651813018788, + "grad_norm": 0.22161249816417694, + "learning_rate": 9.599280460080587e-05, + "loss": 0.0359, + "step": 17190 + }, + { + "epoch": 3.3392068346196786, + "grad_norm": 0.17274247109889984, + "learning_rate": 9.59863162514947e-05, + "loss": 0.0394, + "step": 17200 + }, + { + "epoch": 3.341148487937479, + "grad_norm": 0.2672402858734131, + "learning_rate": 9.597982287318911e-05, + "loss": 0.0328, + "step": 17210 + }, + { + "epoch": 3.3430901412552787, + "grad_norm": 0.17889682948589325, + "learning_rate": 9.597332446659923e-05, + "loss": 0.0337, + "step": 17220 + }, + { + "epoch": 3.345031794573079, + "grad_norm": 0.20979464054107666, + "learning_rate": 9.59668210324357e-05, + "loss": 0.0409, + "step": 17230 + }, + { + "epoch": 3.3469734478908792, + "grad_norm": 0.1603352129459381, + "learning_rate": 9.596031257140974e-05, + "loss": 0.0394, + "step": 17240 + }, + { + "epoch": 3.348915101208679, + "grad_norm": 0.21059289574623108, + "learning_rate": 9.59537990842331e-05, + "loss": 0.0359, + "step": 17250 + }, + { + "epoch": 3.3508567545264794, + "grad_norm": 0.24495922029018402, + "learning_rate": 9.594728057161806e-05, + "loss": 0.0347, + "step": 17260 + }, + { + "epoch": 3.352798407844279, + "grad_norm": 0.1426721215248108, + "learning_rate": 9.594075703427752e-05, + "loss": 0.0363, + "step": 17270 + }, + { + "epoch": 3.3547400611620795, + "grad_norm": 0.2013421356678009, + "learning_rate": 9.593422847292486e-05, + "loss": 0.0291, + "step": 17280 + }, + { + "epoch": 3.3566817144798797, + "grad_norm": 0.19961602985858917, + "learning_rate": 9.592769488827402e-05, + "loss": 0.0295, + "step": 17290 + }, + { + "epoch": 3.3586233677976796, + "grad_norm": 0.19987986981868744, + "learning_rate": 9.592115628103952e-05, + "loss": 0.0446, + "step": 17300 + }, + { + "epoch": 3.36056502111548, + "grad_norm": 0.26948636770248413, + "learning_rate": 9.591461265193643e-05, + "loss": 0.0342, + "step": 17310 + }, + { + "epoch": 3.36250667443328, + "grad_norm": 0.14207026362419128, + "learning_rate": 9.590806400168032e-05, + "loss": 0.0286, + "step": 17320 + }, + { + "epoch": 3.36444832775108, + "grad_norm": 0.2297673523426056, + "learning_rate": 9.590151033098735e-05, + "loss": 0.0377, + "step": 17330 + }, + { + "epoch": 3.36638998106888, + "grad_norm": 0.24731232225894928, + "learning_rate": 9.589495164057423e-05, + "loss": 0.0375, + "step": 17340 + }, + { + "epoch": 3.3683316343866805, + "grad_norm": 0.21122011542320251, + "learning_rate": 9.58883879311582e-05, + "loss": 0.0402, + "step": 17350 + }, + { + "epoch": 3.3702732877044803, + "grad_norm": 0.27568185329437256, + "learning_rate": 9.588181920345705e-05, + "loss": 0.0474, + "step": 17360 + }, + { + "epoch": 3.3722149410222806, + "grad_norm": 0.25252291560173035, + "learning_rate": 9.587524545818913e-05, + "loss": 0.034, + "step": 17370 + }, + { + "epoch": 3.3741565943400804, + "grad_norm": 0.31138014793395996, + "learning_rate": 9.586866669607335e-05, + "loss": 0.041, + "step": 17380 + }, + { + "epoch": 3.3760982476578807, + "grad_norm": 0.23987452685832977, + "learning_rate": 9.586208291782915e-05, + "loss": 0.0374, + "step": 17390 + }, + { + "epoch": 3.378039900975681, + "grad_norm": 0.4572565257549286, + "learning_rate": 9.58554941241765e-05, + "loss": 0.0399, + "step": 17400 + }, + { + "epoch": 3.379981554293481, + "grad_norm": 0.27715691924095154, + "learning_rate": 9.584890031583596e-05, + "loss": 0.0359, + "step": 17410 + }, + { + "epoch": 3.381923207611281, + "grad_norm": 0.11869527399539948, + "learning_rate": 9.584230149352861e-05, + "loss": 0.0275, + "step": 17420 + }, + { + "epoch": 3.383864860929081, + "grad_norm": 0.32075461745262146, + "learning_rate": 9.58356976579761e-05, + "loss": 0.0396, + "step": 17430 + }, + { + "epoch": 3.385806514246881, + "grad_norm": 0.33831021189689636, + "learning_rate": 9.58290888099006e-05, + "loss": 0.038, + "step": 17440 + }, + { + "epoch": 3.3877481675646814, + "grad_norm": 0.1579270362854004, + "learning_rate": 9.582247495002486e-05, + "loss": 0.0374, + "step": 17450 + }, + { + "epoch": 3.3896898208824813, + "grad_norm": 0.22050116956233978, + "learning_rate": 9.581585607907214e-05, + "loss": 0.0346, + "step": 17460 + }, + { + "epoch": 3.3916314742002815, + "grad_norm": 0.3684067130088806, + "learning_rate": 9.580923219776628e-05, + "loss": 0.0389, + "step": 17470 + }, + { + "epoch": 3.393573127518082, + "grad_norm": 0.23634742200374603, + "learning_rate": 9.580260330683167e-05, + "loss": 0.0362, + "step": 17480 + }, + { + "epoch": 3.3955147808358817, + "grad_norm": 0.22803880274295807, + "learning_rate": 9.579596940699322e-05, + "loss": 0.0463, + "step": 17490 + }, + { + "epoch": 3.397456434153682, + "grad_norm": 0.28461772203445435, + "learning_rate": 9.578933049897643e-05, + "loss": 0.035, + "step": 17500 + }, + { + "epoch": 3.399398087471482, + "grad_norm": 0.23437421023845673, + "learning_rate": 9.578268658350728e-05, + "loss": 0.0401, + "step": 17510 + }, + { + "epoch": 3.401339740789282, + "grad_norm": 0.23824210464954376, + "learning_rate": 9.577603766131235e-05, + "loss": 0.0377, + "step": 17520 + }, + { + "epoch": 3.4032813941070823, + "grad_norm": 0.17255181074142456, + "learning_rate": 9.576938373311878e-05, + "loss": 0.0354, + "step": 17530 + }, + { + "epoch": 3.405223047424882, + "grad_norm": 0.22650973498821259, + "learning_rate": 9.576272479965421e-05, + "loss": 0.038, + "step": 17540 + }, + { + "epoch": 3.4071647007426824, + "grad_norm": 0.16210906207561493, + "learning_rate": 9.575606086164687e-05, + "loss": 0.0369, + "step": 17550 + }, + { + "epoch": 3.4091063540604827, + "grad_norm": 0.28144755959510803, + "learning_rate": 9.57493919198255e-05, + "loss": 0.0402, + "step": 17560 + }, + { + "epoch": 3.4110480073782825, + "grad_norm": 0.2686938941478729, + "learning_rate": 9.57427179749194e-05, + "loss": 0.038, + "step": 17570 + }, + { + "epoch": 3.412989660696083, + "grad_norm": 0.24786612391471863, + "learning_rate": 9.573603902765846e-05, + "loss": 0.0375, + "step": 17580 + }, + { + "epoch": 3.4149313140138826, + "grad_norm": 0.23872534930706024, + "learning_rate": 9.572935507877304e-05, + "loss": 0.0289, + "step": 17590 + }, + { + "epoch": 3.416872967331683, + "grad_norm": 0.14480087161064148, + "learning_rate": 9.57226661289941e-05, + "loss": 0.0371, + "step": 17600 + }, + { + "epoch": 3.418814620649483, + "grad_norm": 0.12750738859176636, + "learning_rate": 9.571597217905315e-05, + "loss": 0.0462, + "step": 17610 + }, + { + "epoch": 3.420756273967283, + "grad_norm": 0.24535061419010162, + "learning_rate": 9.57092732296822e-05, + "loss": 0.0317, + "step": 17620 + }, + { + "epoch": 3.4226979272850833, + "grad_norm": 0.16405202448368073, + "learning_rate": 9.570256928161385e-05, + "loss": 0.0282, + "step": 17630 + }, + { + "epoch": 3.4246395806028835, + "grad_norm": 0.1835489720106125, + "learning_rate": 9.569586033558126e-05, + "loss": 0.0382, + "step": 17640 + }, + { + "epoch": 3.4265812339206834, + "grad_norm": 0.21347764134407043, + "learning_rate": 9.568914639231807e-05, + "loss": 0.0413, + "step": 17650 + }, + { + "epoch": 3.4285228872384836, + "grad_norm": 0.1614355593919754, + "learning_rate": 9.568242745255852e-05, + "loss": 0.037, + "step": 17660 + }, + { + "epoch": 3.430464540556284, + "grad_norm": 0.1906730681657791, + "learning_rate": 9.567570351703739e-05, + "loss": 0.0349, + "step": 17670 + }, + { + "epoch": 3.4324061938740837, + "grad_norm": 0.2381775826215744, + "learning_rate": 9.566897458649001e-05, + "loss": 0.0321, + "step": 17680 + }, + { + "epoch": 3.434347847191884, + "grad_norm": 0.24951890110969543, + "learning_rate": 9.566224066165221e-05, + "loss": 0.0314, + "step": 17690 + }, + { + "epoch": 3.436289500509684, + "grad_norm": 0.23775073885917664, + "learning_rate": 9.565550174326043e-05, + "loss": 0.0366, + "step": 17700 + }, + { + "epoch": 3.438231153827484, + "grad_norm": 0.3031565248966217, + "learning_rate": 9.564875783205162e-05, + "loss": 0.0461, + "step": 17710 + }, + { + "epoch": 3.4401728071452844, + "grad_norm": 0.10908597707748413, + "learning_rate": 9.564200892876328e-05, + "loss": 0.0358, + "step": 17720 + }, + { + "epoch": 3.4421144604630842, + "grad_norm": 0.11390956491231918, + "learning_rate": 9.563525503413348e-05, + "loss": 0.0288, + "step": 17730 + }, + { + "epoch": 3.4440561137808845, + "grad_norm": 0.12093869596719742, + "learning_rate": 9.562849614890079e-05, + "loss": 0.0274, + "step": 17740 + }, + { + "epoch": 3.4459977670986843, + "grad_norm": 0.1606406420469284, + "learning_rate": 9.562173227380436e-05, + "loss": 0.0341, + "step": 17750 + }, + { + "epoch": 3.4479394204164846, + "grad_norm": 0.18007676303386688, + "learning_rate": 9.561496340958389e-05, + "loss": 0.0447, + "step": 17760 + }, + { + "epoch": 3.449881073734285, + "grad_norm": 0.36562490463256836, + "learning_rate": 9.560818955697959e-05, + "loss": 0.0343, + "step": 17770 + }, + { + "epoch": 3.4518227270520847, + "grad_norm": 0.12255498021841049, + "learning_rate": 9.560141071673228e-05, + "loss": 0.0311, + "step": 17780 + }, + { + "epoch": 3.453764380369885, + "grad_norm": 0.1590954065322876, + "learning_rate": 9.559462688958323e-05, + "loss": 0.0303, + "step": 17790 + }, + { + "epoch": 3.4557060336876853, + "grad_norm": 0.24051232635974884, + "learning_rate": 9.558783807627434e-05, + "loss": 0.0433, + "step": 17800 + }, + { + "epoch": 3.457647687005485, + "grad_norm": 0.23992201685905457, + "learning_rate": 9.558104427754801e-05, + "loss": 0.0402, + "step": 17810 + }, + { + "epoch": 3.4595893403232854, + "grad_norm": 0.18021860718727112, + "learning_rate": 9.557424549414722e-05, + "loss": 0.0323, + "step": 17820 + }, + { + "epoch": 3.4615309936410856, + "grad_norm": 0.20850710570812225, + "learning_rate": 9.556744172681546e-05, + "loss": 0.0337, + "step": 17830 + }, + { + "epoch": 3.4634726469588855, + "grad_norm": 0.2685239017009735, + "learning_rate": 9.556063297629677e-05, + "loss": 0.0346, + "step": 17840 + }, + { + "epoch": 3.4654143002766857, + "grad_norm": 0.25578707456588745, + "learning_rate": 9.555381924333578e-05, + "loss": 0.0352, + "step": 17850 + }, + { + "epoch": 3.4673559535944856, + "grad_norm": 0.30072152614593506, + "learning_rate": 9.554700052867758e-05, + "loss": 0.0346, + "step": 17860 + }, + { + "epoch": 3.469297606912286, + "grad_norm": 0.17810410261154175, + "learning_rate": 9.554017683306789e-05, + "loss": 0.0353, + "step": 17870 + }, + { + "epoch": 3.471239260230086, + "grad_norm": 0.1628720611333847, + "learning_rate": 9.553334815725294e-05, + "loss": 0.0346, + "step": 17880 + }, + { + "epoch": 3.473180913547886, + "grad_norm": 0.18937532603740692, + "learning_rate": 9.552651450197949e-05, + "loss": 0.0411, + "step": 17890 + }, + { + "epoch": 3.475122566865686, + "grad_norm": 0.2073807418346405, + "learning_rate": 9.551967586799486e-05, + "loss": 0.038, + "step": 17900 + }, + { + "epoch": 3.477064220183486, + "grad_norm": 0.22810034453868866, + "learning_rate": 9.551283225604692e-05, + "loss": 0.0357, + "step": 17910 + }, + { + "epoch": 3.4790058735012863, + "grad_norm": 0.3039987087249756, + "learning_rate": 9.550598366688406e-05, + "loss": 0.043, + "step": 17920 + }, + { + "epoch": 3.4809475268190866, + "grad_norm": 0.2455424815416336, + "learning_rate": 9.549913010125526e-05, + "loss": 0.0397, + "step": 17930 + }, + { + "epoch": 3.4828891801368864, + "grad_norm": 0.15295390784740448, + "learning_rate": 9.549227155990999e-05, + "loss": 0.0334, + "step": 17940 + }, + { + "epoch": 3.4848308334546867, + "grad_norm": 0.11646267771720886, + "learning_rate": 9.548540804359828e-05, + "loss": 0.0329, + "step": 17950 + }, + { + "epoch": 3.4867724867724865, + "grad_norm": 0.19278031587600708, + "learning_rate": 9.547853955307077e-05, + "loss": 0.04, + "step": 17960 + }, + { + "epoch": 3.488714140090287, + "grad_norm": 0.12380882352590561, + "learning_rate": 9.547166608907853e-05, + "loss": 0.0314, + "step": 17970 + }, + { + "epoch": 3.490655793408087, + "grad_norm": 0.08898573368787766, + "learning_rate": 9.546478765237326e-05, + "loss": 0.0384, + "step": 17980 + }, + { + "epoch": 3.492597446725887, + "grad_norm": 0.1747208535671234, + "learning_rate": 9.545790424370715e-05, + "loss": 0.0347, + "step": 17990 + }, + { + "epoch": 3.494539100043687, + "grad_norm": 0.2195773422718048, + "learning_rate": 9.5451015863833e-05, + "loss": 0.0417, + "step": 18000 + }, + { + "epoch": 3.4964807533614874, + "grad_norm": 0.19510114192962646, + "learning_rate": 9.544412251350408e-05, + "loss": 0.0402, + "step": 18010 + }, + { + "epoch": 3.4984224066792873, + "grad_norm": 0.2243666797876358, + "learning_rate": 9.543722419347422e-05, + "loss": 0.0341, + "step": 18020 + }, + { + "epoch": 3.5003640599970876, + "grad_norm": 0.2516593337059021, + "learning_rate": 9.543032090449788e-05, + "loss": 0.0283, + "step": 18030 + }, + { + "epoch": 3.502305713314888, + "grad_norm": 0.16286243498325348, + "learning_rate": 9.542341264732992e-05, + "loss": 0.0294, + "step": 18040 + }, + { + "epoch": 3.5042473666326877, + "grad_norm": 0.21512699127197266, + "learning_rate": 9.541649942272585e-05, + "loss": 0.0392, + "step": 18050 + }, + { + "epoch": 3.506189019950488, + "grad_norm": 0.22315306961536407, + "learning_rate": 9.54095812314417e-05, + "loss": 0.0369, + "step": 18060 + }, + { + "epoch": 3.5081306732682878, + "grad_norm": 0.21351197361946106, + "learning_rate": 9.540265807423401e-05, + "loss": 0.0293, + "step": 18070 + }, + { + "epoch": 3.510072326586088, + "grad_norm": 0.20571811497211456, + "learning_rate": 9.53957299518599e-05, + "loss": 0.0361, + "step": 18080 + }, + { + "epoch": 3.5120139799038883, + "grad_norm": 0.13632188737392426, + "learning_rate": 9.5388796865077e-05, + "loss": 0.0369, + "step": 18090 + }, + { + "epoch": 3.513955633221688, + "grad_norm": 0.18737205862998962, + "learning_rate": 9.538185881464353e-05, + "loss": 0.038, + "step": 18100 + }, + { + "epoch": 3.5158972865394884, + "grad_norm": 0.24686887860298157, + "learning_rate": 9.537491580131821e-05, + "loss": 0.0368, + "step": 18110 + }, + { + "epoch": 3.5178389398572882, + "grad_norm": 0.15176492929458618, + "learning_rate": 9.53679678258603e-05, + "loss": 0.0313, + "step": 18120 + }, + { + "epoch": 3.5197805931750885, + "grad_norm": 0.2408599704504013, + "learning_rate": 9.536101488902966e-05, + "loss": 0.0339, + "step": 18130 + }, + { + "epoch": 3.521722246492889, + "grad_norm": 0.1838449239730835, + "learning_rate": 9.535405699158663e-05, + "loss": 0.0283, + "step": 18140 + }, + { + "epoch": 3.523663899810689, + "grad_norm": 0.11995450407266617, + "learning_rate": 9.53470941342921e-05, + "loss": 0.0407, + "step": 18150 + }, + { + "epoch": 3.525605553128489, + "grad_norm": 0.12861967086791992, + "learning_rate": 9.534012631790756e-05, + "loss": 0.0298, + "step": 18160 + }, + { + "epoch": 3.527547206446289, + "grad_norm": 0.16170652210712433, + "learning_rate": 9.533315354319494e-05, + "loss": 0.0312, + "step": 18170 + }, + { + "epoch": 3.529488859764089, + "grad_norm": 0.24574892222881317, + "learning_rate": 9.532617581091682e-05, + "loss": 0.0381, + "step": 18180 + }, + { + "epoch": 3.5314305130818893, + "grad_norm": 0.18513701856136322, + "learning_rate": 9.531919312183629e-05, + "loss": 0.0291, + "step": 18190 + }, + { + "epoch": 3.5333721663996895, + "grad_norm": 0.15302419662475586, + "learning_rate": 9.531220547671688e-05, + "loss": 0.0396, + "step": 18200 + }, + { + "epoch": 3.5353138197174894, + "grad_norm": 0.2682681679725647, + "learning_rate": 9.530521287632285e-05, + "loss": 0.0386, + "step": 18210 + }, + { + "epoch": 3.5372554730352896, + "grad_norm": 0.29102036356925964, + "learning_rate": 9.529821532141884e-05, + "loss": 0.0361, + "step": 18220 + }, + { + "epoch": 3.5391971263530895, + "grad_norm": 0.3111441433429718, + "learning_rate": 9.52912128127701e-05, + "loss": 0.0396, + "step": 18230 + }, + { + "epoch": 3.5411387796708897, + "grad_norm": 0.20083025097846985, + "learning_rate": 9.528420535114244e-05, + "loss": 0.0314, + "step": 18240 + }, + { + "epoch": 3.54308043298869, + "grad_norm": 0.1532052755355835, + "learning_rate": 9.527719293730215e-05, + "loss": 0.0288, + "step": 18250 + }, + { + "epoch": 3.54502208630649, + "grad_norm": 0.14050358533859253, + "learning_rate": 9.527017557201611e-05, + "loss": 0.028, + "step": 18260 + }, + { + "epoch": 3.54696373962429, + "grad_norm": 0.25597473978996277, + "learning_rate": 9.526315325605176e-05, + "loss": 0.0412, + "step": 18270 + }, + { + "epoch": 3.54890539294209, + "grad_norm": 0.12587672472000122, + "learning_rate": 9.525612599017699e-05, + "loss": 0.0355, + "step": 18280 + }, + { + "epoch": 3.5508470462598902, + "grad_norm": 0.13735522329807281, + "learning_rate": 9.524909377516033e-05, + "loss": 0.0359, + "step": 18290 + }, + { + "epoch": 3.5527886995776905, + "grad_norm": 0.2144606113433838, + "learning_rate": 9.524205661177081e-05, + "loss": 0.0487, + "step": 18300 + }, + { + "epoch": 3.5547303528954908, + "grad_norm": 0.2060600221157074, + "learning_rate": 9.523501450077801e-05, + "loss": 0.038, + "step": 18310 + }, + { + "epoch": 3.5566720062132906, + "grad_norm": 0.19012272357940674, + "learning_rate": 9.522796744295202e-05, + "loss": 0.0376, + "step": 18320 + }, + { + "epoch": 3.558613659531091, + "grad_norm": 0.264726847410202, + "learning_rate": 9.522091543906352e-05, + "loss": 0.0365, + "step": 18330 + }, + { + "epoch": 3.5605553128488907, + "grad_norm": 0.27844738960266113, + "learning_rate": 9.521385848988369e-05, + "loss": 0.0391, + "step": 18340 + }, + { + "epoch": 3.562496966166691, + "grad_norm": 0.38445624709129333, + "learning_rate": 9.520679659618428e-05, + "loss": 0.0445, + "step": 18350 + }, + { + "epoch": 3.5644386194844913, + "grad_norm": 0.2623445689678192, + "learning_rate": 9.519972975873754e-05, + "loss": 0.0371, + "step": 18360 + }, + { + "epoch": 3.566380272802291, + "grad_norm": 0.254609078168869, + "learning_rate": 9.519265797831633e-05, + "loss": 0.0379, + "step": 18370 + }, + { + "epoch": 3.5683219261200914, + "grad_norm": 0.2371920794248581, + "learning_rate": 9.518558125569399e-05, + "loss": 0.0337, + "step": 18380 + }, + { + "epoch": 3.570263579437891, + "grad_norm": 0.22442744672298431, + "learning_rate": 9.517849959164442e-05, + "loss": 0.0341, + "step": 18390 + }, + { + "epoch": 3.5722052327556915, + "grad_norm": 0.21859484910964966, + "learning_rate": 9.517141298694205e-05, + "loss": 0.0319, + "step": 18400 + }, + { + "epoch": 3.5741468860734917, + "grad_norm": 0.2089177966117859, + "learning_rate": 9.516432144236188e-05, + "loss": 0.0387, + "step": 18410 + }, + { + "epoch": 3.5760885393912916, + "grad_norm": 0.19881071150302887, + "learning_rate": 9.515722495867941e-05, + "loss": 0.0426, + "step": 18420 + }, + { + "epoch": 3.578030192709092, + "grad_norm": 0.23747046291828156, + "learning_rate": 9.515012353667072e-05, + "loss": 0.0413, + "step": 18430 + }, + { + "epoch": 3.5799718460268917, + "grad_norm": 0.20011566579341888, + "learning_rate": 9.51430171771124e-05, + "loss": 0.0337, + "step": 18440 + }, + { + "epoch": 3.581913499344692, + "grad_norm": 0.2219311147928238, + "learning_rate": 9.513590588078159e-05, + "loss": 0.0372, + "step": 18450 + }, + { + "epoch": 3.583855152662492, + "grad_norm": 0.30443188548088074, + "learning_rate": 9.512878964845597e-05, + "loss": 0.0391, + "step": 18460 + }, + { + "epoch": 3.5857968059802925, + "grad_norm": 0.2722233235836029, + "learning_rate": 9.512166848091377e-05, + "loss": 0.0359, + "step": 18470 + }, + { + "epoch": 3.5877384592980923, + "grad_norm": 0.14042209088802338, + "learning_rate": 9.511454237893376e-05, + "loss": 0.0361, + "step": 18480 + }, + { + "epoch": 3.5896801126158926, + "grad_norm": 0.18731525540351868, + "learning_rate": 9.51074113432952e-05, + "loss": 0.0359, + "step": 18490 + }, + { + "epoch": 3.5916217659336924, + "grad_norm": 0.3497833013534546, + "learning_rate": 9.510027537477797e-05, + "loss": 0.0375, + "step": 18500 + }, + { + "epoch": 3.5935634192514927, + "grad_norm": 0.23908908665180206, + "learning_rate": 9.509313447416242e-05, + "loss": 0.0342, + "step": 18510 + }, + { + "epoch": 3.595505072569293, + "grad_norm": 0.18831010162830353, + "learning_rate": 9.508598864222949e-05, + "loss": 0.0345, + "step": 18520 + }, + { + "epoch": 3.597446725887093, + "grad_norm": 0.22316937148571014, + "learning_rate": 9.507883787976062e-05, + "loss": 0.032, + "step": 18530 + }, + { + "epoch": 3.599388379204893, + "grad_norm": 0.15901435911655426, + "learning_rate": 9.507168218753781e-05, + "loss": 0.0353, + "step": 18540 + }, + { + "epoch": 3.601330032522693, + "grad_norm": 0.17235234379768372, + "learning_rate": 9.506452156634362e-05, + "loss": 0.0327, + "step": 18550 + }, + { + "epoch": 3.603271685840493, + "grad_norm": 0.2351781576871872, + "learning_rate": 9.505735601696109e-05, + "loss": 0.0338, + "step": 18560 + }, + { + "epoch": 3.6052133391582935, + "grad_norm": 0.2614324986934662, + "learning_rate": 9.505018554017385e-05, + "loss": 0.0349, + "step": 18570 + }, + { + "epoch": 3.6071549924760933, + "grad_norm": 0.16505472362041473, + "learning_rate": 9.504301013676604e-05, + "loss": 0.0282, + "step": 18580 + }, + { + "epoch": 3.6090966457938936, + "grad_norm": 0.24002093076705933, + "learning_rate": 9.503582980752238e-05, + "loss": 0.0309, + "step": 18590 + }, + { + "epoch": 3.6110382991116934, + "grad_norm": 0.16196739673614502, + "learning_rate": 9.502864455322809e-05, + "loss": 0.0438, + "step": 18600 + }, + { + "epoch": 3.6129799524294937, + "grad_norm": 0.21072633564472198, + "learning_rate": 9.502145437466891e-05, + "loss": 0.0338, + "step": 18610 + }, + { + "epoch": 3.614921605747294, + "grad_norm": 0.22694489359855652, + "learning_rate": 9.501425927263116e-05, + "loss": 0.0354, + "step": 18620 + }, + { + "epoch": 3.616863259065094, + "grad_norm": 0.15875528752803802, + "learning_rate": 9.500705924790172e-05, + "loss": 0.0353, + "step": 18630 + }, + { + "epoch": 3.618804912382894, + "grad_norm": 0.20296014845371246, + "learning_rate": 9.499985430126794e-05, + "loss": 0.0315, + "step": 18640 + }, + { + "epoch": 3.620746565700694, + "grad_norm": 0.19304326176643372, + "learning_rate": 9.499264443351775e-05, + "loss": 0.0375, + "step": 18650 + }, + { + "epoch": 3.622688219018494, + "grad_norm": 0.12978604435920715, + "learning_rate": 9.498542964543961e-05, + "loss": 0.0359, + "step": 18660 + }, + { + "epoch": 3.6246298723362944, + "grad_norm": 0.15656553208827972, + "learning_rate": 9.497820993782252e-05, + "loss": 0.0377, + "step": 18670 + }, + { + "epoch": 3.6265715256540947, + "grad_norm": 0.19687804579734802, + "learning_rate": 9.497098531145601e-05, + "loss": 0.0381, + "step": 18680 + }, + { + "epoch": 3.6285131789718945, + "grad_norm": 0.19424472749233246, + "learning_rate": 9.496375576713017e-05, + "loss": 0.0291, + "step": 18690 + }, + { + "epoch": 3.630454832289695, + "grad_norm": 0.16568274796009064, + "learning_rate": 9.49565213056356e-05, + "loss": 0.0301, + "step": 18700 + }, + { + "epoch": 3.6323964856074946, + "grad_norm": 0.18221555650234222, + "learning_rate": 9.494928192776342e-05, + "loss": 0.0439, + "step": 18710 + }, + { + "epoch": 3.634338138925295, + "grad_norm": 0.2061670571565628, + "learning_rate": 9.494203763430538e-05, + "loss": 0.035, + "step": 18720 + }, + { + "epoch": 3.636279792243095, + "grad_norm": 0.1267136186361313, + "learning_rate": 9.493478842605366e-05, + "loss": 0.0401, + "step": 18730 + }, + { + "epoch": 3.638221445560895, + "grad_norm": 0.2811046540737152, + "learning_rate": 9.492753430380105e-05, + "loss": 0.0433, + "step": 18740 + }, + { + "epoch": 3.6401630988786953, + "grad_norm": 0.23417215049266815, + "learning_rate": 9.492027526834083e-05, + "loss": 0.0352, + "step": 18750 + }, + { + "epoch": 3.642104752196495, + "grad_norm": 0.2968893051147461, + "learning_rate": 9.491301132046684e-05, + "loss": 0.0422, + "step": 18760 + }, + { + "epoch": 3.6440464055142954, + "grad_norm": 0.245815709233284, + "learning_rate": 9.490574246097345e-05, + "loss": 0.0352, + "step": 18770 + }, + { + "epoch": 3.6459880588320956, + "grad_norm": 0.2158382683992386, + "learning_rate": 9.48984686906556e-05, + "loss": 0.0339, + "step": 18780 + }, + { + "epoch": 3.6479297121498955, + "grad_norm": 0.21587160229682922, + "learning_rate": 9.489119001030871e-05, + "loss": 0.0332, + "step": 18790 + }, + { + "epoch": 3.6498713654676957, + "grad_norm": 0.15983210504055023, + "learning_rate": 9.488390642072878e-05, + "loss": 0.032, + "step": 18800 + }, + { + "epoch": 3.6518130187854956, + "grad_norm": 0.1851532757282257, + "learning_rate": 9.48766179227123e-05, + "loss": 0.0393, + "step": 18810 + }, + { + "epoch": 3.653754672103296, + "grad_norm": 0.2549534738063812, + "learning_rate": 9.486932451705636e-05, + "loss": 0.03, + "step": 18820 + }, + { + "epoch": 3.655696325421096, + "grad_norm": 0.2334885597229004, + "learning_rate": 9.486202620455857e-05, + "loss": 0.0434, + "step": 18830 + }, + { + "epoch": 3.6576379787388964, + "grad_norm": 0.1312558948993683, + "learning_rate": 9.485472298601704e-05, + "loss": 0.0337, + "step": 18840 + }, + { + "epoch": 3.6595796320566962, + "grad_norm": 0.14893607795238495, + "learning_rate": 9.484741486223043e-05, + "loss": 0.031, + "step": 18850 + }, + { + "epoch": 3.6615212853744965, + "grad_norm": 0.2690189778804779, + "learning_rate": 9.484010183399797e-05, + "loss": 0.0326, + "step": 18860 + }, + { + "epoch": 3.6634629386922963, + "grad_norm": 0.18041330575942993, + "learning_rate": 9.483278390211938e-05, + "loss": 0.0369, + "step": 18870 + }, + { + "epoch": 3.6654045920100966, + "grad_norm": 0.19307096302509308, + "learning_rate": 9.482546106739496e-05, + "loss": 0.0413, + "step": 18880 + }, + { + "epoch": 3.667346245327897, + "grad_norm": 0.2542825937271118, + "learning_rate": 9.48181333306255e-05, + "loss": 0.0384, + "step": 18890 + }, + { + "epoch": 3.6692878986456967, + "grad_norm": 0.23608516156673431, + "learning_rate": 9.481080069261237e-05, + "loss": 0.0367, + "step": 18900 + }, + { + "epoch": 3.671229551963497, + "grad_norm": 0.12217075377702713, + "learning_rate": 9.480346315415745e-05, + "loss": 0.032, + "step": 18910 + }, + { + "epoch": 3.673171205281297, + "grad_norm": 0.13826338946819305, + "learning_rate": 9.479612071606314e-05, + "loss": 0.0362, + "step": 18920 + }, + { + "epoch": 3.675112858599097, + "grad_norm": 0.19856753945350647, + "learning_rate": 9.478877337913244e-05, + "loss": 0.0398, + "step": 18930 + }, + { + "epoch": 3.6770545119168974, + "grad_norm": 0.25439178943634033, + "learning_rate": 9.478142114416881e-05, + "loss": 0.0389, + "step": 18940 + }, + { + "epoch": 3.678996165234697, + "grad_norm": 0.2979566156864166, + "learning_rate": 9.47740640119763e-05, + "loss": 0.0345, + "step": 18950 + }, + { + "epoch": 3.6809378185524975, + "grad_norm": 0.17496898770332336, + "learning_rate": 9.476670198335947e-05, + "loss": 0.0409, + "step": 18960 + }, + { + "epoch": 3.6828794718702973, + "grad_norm": 0.2640736699104309, + "learning_rate": 9.47593350591234e-05, + "loss": 0.0392, + "step": 18970 + }, + { + "epoch": 3.6848211251880976, + "grad_norm": 0.19308443367481232, + "learning_rate": 9.475196324007376e-05, + "loss": 0.0412, + "step": 18980 + }, + { + "epoch": 3.686762778505898, + "grad_norm": 0.19144456088542938, + "learning_rate": 9.474458652701669e-05, + "loss": 0.0349, + "step": 18990 + }, + { + "epoch": 3.688704431823698, + "grad_norm": 0.19774295389652252, + "learning_rate": 9.473720492075892e-05, + "loss": 0.0367, + "step": 19000 + }, + { + "epoch": 3.690646085141498, + "grad_norm": 0.12331102788448334, + "learning_rate": 9.472981842210768e-05, + "loss": 0.0415, + "step": 19010 + }, + { + "epoch": 3.692587738459298, + "grad_norm": 0.24338391423225403, + "learning_rate": 9.472242703187074e-05, + "loss": 0.0311, + "step": 19020 + }, + { + "epoch": 3.694529391777098, + "grad_norm": 0.2533773183822632, + "learning_rate": 9.471503075085643e-05, + "loss": 0.0389, + "step": 19030 + }, + { + "epoch": 3.6964710450948983, + "grad_norm": 0.1280445158481598, + "learning_rate": 9.470762957987359e-05, + "loss": 0.0338, + "step": 19040 + }, + { + "epoch": 3.6984126984126986, + "grad_norm": 0.22813062369823456, + "learning_rate": 9.470022351973158e-05, + "loss": 0.0362, + "step": 19050 + }, + { + "epoch": 3.7003543517304984, + "grad_norm": 0.1815641075372696, + "learning_rate": 9.469281257124034e-05, + "loss": 0.0373, + "step": 19060 + }, + { + "epoch": 3.7022960050482987, + "grad_norm": 0.3289199769496918, + "learning_rate": 9.46853967352103e-05, + "loss": 0.041, + "step": 19070 + }, + { + "epoch": 3.7042376583660985, + "grad_norm": 0.24086226522922516, + "learning_rate": 9.467797601245246e-05, + "loss": 0.0401, + "step": 19080 + }, + { + "epoch": 3.706179311683899, + "grad_norm": 0.2631997764110565, + "learning_rate": 9.467055040377834e-05, + "loss": 0.0426, + "step": 19090 + }, + { + "epoch": 3.708120965001699, + "grad_norm": 0.18705418705940247, + "learning_rate": 9.466311990999999e-05, + "loss": 0.0357, + "step": 19100 + }, + { + "epoch": 3.710062618319499, + "grad_norm": 0.14181040227413177, + "learning_rate": 9.465568453193e-05, + "loss": 0.0353, + "step": 19110 + }, + { + "epoch": 3.712004271637299, + "grad_norm": 0.23069114983081818, + "learning_rate": 9.464824427038148e-05, + "loss": 0.0446, + "step": 19120 + }, + { + "epoch": 3.713945924955099, + "grad_norm": 0.18365561962127686, + "learning_rate": 9.46407991261681e-05, + "loss": 0.0345, + "step": 19130 + }, + { + "epoch": 3.7158875782728993, + "grad_norm": 0.24760156869888306, + "learning_rate": 9.463334910010404e-05, + "loss": 0.0411, + "step": 19140 + }, + { + "epoch": 3.7178292315906996, + "grad_norm": 0.23609702289104462, + "learning_rate": 9.462589419300403e-05, + "loss": 0.0385, + "step": 19150 + }, + { + "epoch": 3.7197708849085, + "grad_norm": 0.21741034090518951, + "learning_rate": 9.461843440568333e-05, + "loss": 0.0387, + "step": 19160 + }, + { + "epoch": 3.7217125382262997, + "grad_norm": 0.2347104698419571, + "learning_rate": 9.461096973895773e-05, + "loss": 0.0326, + "step": 19170 + }, + { + "epoch": 3.7236541915441, + "grad_norm": 0.23595565557479858, + "learning_rate": 9.460350019364355e-05, + "loss": 0.0255, + "step": 19180 + }, + { + "epoch": 3.7255958448618998, + "grad_norm": 0.20910660922527313, + "learning_rate": 9.459602577055764e-05, + "loss": 0.0365, + "step": 19190 + }, + { + "epoch": 3.7275374981797, + "grad_norm": 0.14727449417114258, + "learning_rate": 9.45885464705174e-05, + "loss": 0.0338, + "step": 19200 + }, + { + "epoch": 3.7294791514975003, + "grad_norm": 0.18723522126674652, + "learning_rate": 9.458106229434076e-05, + "loss": 0.0369, + "step": 19210 + }, + { + "epoch": 3.7314208048153, + "grad_norm": 0.2150357961654663, + "learning_rate": 9.457357324284617e-05, + "loss": 0.0376, + "step": 19220 + }, + { + "epoch": 3.7333624581331004, + "grad_norm": 0.14783555269241333, + "learning_rate": 9.456607931685262e-05, + "loss": 0.0452, + "step": 19230 + }, + { + "epoch": 3.7353041114509002, + "grad_norm": 0.2590571939945221, + "learning_rate": 9.455858051717965e-05, + "loss": 0.0356, + "step": 19240 + }, + { + "epoch": 3.7372457647687005, + "grad_norm": 0.2196631282567978, + "learning_rate": 9.45510768446473e-05, + "loss": 0.0383, + "step": 19250 + }, + { + "epoch": 3.739187418086501, + "grad_norm": 0.32129791378974915, + "learning_rate": 9.454356830007618e-05, + "loss": 0.0337, + "step": 19260 + }, + { + "epoch": 3.7411290714043006, + "grad_norm": 0.26976892352104187, + "learning_rate": 9.45360548842874e-05, + "loss": 0.0338, + "step": 19270 + }, + { + "epoch": 3.743070724722101, + "grad_norm": 0.21213921904563904, + "learning_rate": 9.452853659810261e-05, + "loss": 0.0324, + "step": 19280 + }, + { + "epoch": 3.7450123780399007, + "grad_norm": 0.31609299778938293, + "learning_rate": 9.452101344234401e-05, + "loss": 0.0419, + "step": 19290 + }, + { + "epoch": 3.746954031357701, + "grad_norm": 0.2193802297115326, + "learning_rate": 9.451348541783431e-05, + "loss": 0.0349, + "step": 19300 + }, + { + "epoch": 3.7488956846755013, + "grad_norm": 0.22285959124565125, + "learning_rate": 9.450595252539678e-05, + "loss": 0.0358, + "step": 19310 + }, + { + "epoch": 3.7508373379933015, + "grad_norm": 0.13007138669490814, + "learning_rate": 9.449841476585518e-05, + "loss": 0.0415, + "step": 19320 + }, + { + "epoch": 3.7527789913111014, + "grad_norm": 0.25383734703063965, + "learning_rate": 9.449087214003384e-05, + "loss": 0.0399, + "step": 19330 + }, + { + "epoch": 3.7547206446289016, + "grad_norm": 0.2625589668750763, + "learning_rate": 9.448332464875765e-05, + "loss": 0.0333, + "step": 19340 + }, + { + "epoch": 3.7566622979467015, + "grad_norm": 0.2059636414051056, + "learning_rate": 9.447577229285192e-05, + "loss": 0.0316, + "step": 19350 + }, + { + "epoch": 3.7586039512645018, + "grad_norm": 0.22598406672477722, + "learning_rate": 9.446821507314261e-05, + "loss": 0.0396, + "step": 19360 + }, + { + "epoch": 3.760545604582302, + "grad_norm": 0.28237906098365784, + "learning_rate": 9.446065299045617e-05, + "loss": 0.0385, + "step": 19370 + }, + { + "epoch": 3.762487257900102, + "grad_norm": 0.2558126151561737, + "learning_rate": 9.445308604561955e-05, + "loss": 0.0357, + "step": 19380 + }, + { + "epoch": 3.764428911217902, + "grad_norm": 0.25377005338668823, + "learning_rate": 9.444551423946028e-05, + "loss": 0.035, + "step": 19390 + }, + { + "epoch": 3.766370564535702, + "grad_norm": 0.2422984540462494, + "learning_rate": 9.443793757280638e-05, + "loss": 0.0306, + "step": 19400 + }, + { + "epoch": 3.7683122178535022, + "grad_norm": 0.2570381760597229, + "learning_rate": 9.443035604648646e-05, + "loss": 0.0389, + "step": 19410 + }, + { + "epoch": 3.7702538711713025, + "grad_norm": 0.20761898159980774, + "learning_rate": 9.44227696613296e-05, + "loss": 0.0332, + "step": 19420 + }, + { + "epoch": 3.7721955244891023, + "grad_norm": 0.22831431031227112, + "learning_rate": 9.441517841816542e-05, + "loss": 0.0414, + "step": 19430 + }, + { + "epoch": 3.7741371778069026, + "grad_norm": 0.2374638468027115, + "learning_rate": 9.440758231782413e-05, + "loss": 0.0357, + "step": 19440 + }, + { + "epoch": 3.7760788311247024, + "grad_norm": 0.25701844692230225, + "learning_rate": 9.439998136113639e-05, + "loss": 0.0371, + "step": 19450 + }, + { + "epoch": 3.7780204844425027, + "grad_norm": 0.24818600714206696, + "learning_rate": 9.439237554893344e-05, + "loss": 0.0352, + "step": 19460 + }, + { + "epoch": 3.779962137760303, + "grad_norm": 0.18000073730945587, + "learning_rate": 9.438476488204705e-05, + "loss": 0.0406, + "step": 19470 + }, + { + "epoch": 3.7819037910781033, + "grad_norm": 0.19255603849887848, + "learning_rate": 9.43771493613095e-05, + "loss": 0.0297, + "step": 19480 + }, + { + "epoch": 3.783845444395903, + "grad_norm": 0.30409809947013855, + "learning_rate": 9.436952898755362e-05, + "loss": 0.0295, + "step": 19490 + }, + { + "epoch": 3.7857870977137034, + "grad_norm": 0.16080519556999207, + "learning_rate": 9.436190376161276e-05, + "loss": 0.0338, + "step": 19500 + }, + { + "epoch": 3.787728751031503, + "grad_norm": 0.22005344927310944, + "learning_rate": 9.43542736843208e-05, + "loss": 0.0352, + "step": 19510 + }, + { + "epoch": 3.7896704043493035, + "grad_norm": 0.17865276336669922, + "learning_rate": 9.434663875651216e-05, + "loss": 0.0377, + "step": 19520 + }, + { + "epoch": 3.7916120576671037, + "grad_norm": 0.22358743846416473, + "learning_rate": 9.433899897902177e-05, + "loss": 0.0327, + "step": 19530 + }, + { + "epoch": 3.7935537109849036, + "grad_norm": 0.15306922793388367, + "learning_rate": 9.433135435268511e-05, + "loss": 0.0304, + "step": 19540 + }, + { + "epoch": 3.795495364302704, + "grad_norm": 0.3382417559623718, + "learning_rate": 9.432370487833819e-05, + "loss": 0.0383, + "step": 19550 + }, + { + "epoch": 3.7974370176205037, + "grad_norm": 0.15547992289066315, + "learning_rate": 9.431605055681756e-05, + "loss": 0.0299, + "step": 19560 + }, + { + "epoch": 3.799378670938304, + "grad_norm": 0.19318950176239014, + "learning_rate": 9.430839138896026e-05, + "loss": 0.028, + "step": 19570 + }, + { + "epoch": 3.801320324256104, + "grad_norm": 0.16930192708969116, + "learning_rate": 9.43007273756039e-05, + "loss": 0.0326, + "step": 19580 + }, + { + "epoch": 3.803261977573904, + "grad_norm": 0.1831599622964859, + "learning_rate": 9.429305851758658e-05, + "loss": 0.0403, + "step": 19590 + }, + { + "epoch": 3.8052036308917043, + "grad_norm": 0.28975698351860046, + "learning_rate": 9.428538481574699e-05, + "loss": 0.034, + "step": 19600 + }, + { + "epoch": 3.807145284209504, + "grad_norm": 0.2308121919631958, + "learning_rate": 9.42777062709243e-05, + "loss": 0.0421, + "step": 19610 + }, + { + "epoch": 3.8090869375273044, + "grad_norm": 0.17828458547592163, + "learning_rate": 9.427002288395821e-05, + "loss": 0.035, + "step": 19620 + }, + { + "epoch": 3.8110285908451047, + "grad_norm": 0.22094663977622986, + "learning_rate": 9.426233465568898e-05, + "loss": 0.0307, + "step": 19630 + }, + { + "epoch": 3.812970244162905, + "grad_norm": 0.2124018669128418, + "learning_rate": 9.42546415869574e-05, + "loss": 0.0306, + "step": 19640 + }, + { + "epoch": 3.814911897480705, + "grad_norm": 0.21384884417057037, + "learning_rate": 9.424694367860473e-05, + "loss": 0.0528, + "step": 19650 + }, + { + "epoch": 3.816853550798505, + "grad_norm": 0.23187772929668427, + "learning_rate": 9.423924093147284e-05, + "loss": 0.0289, + "step": 19660 + }, + { + "epoch": 3.818795204116305, + "grad_norm": 0.13602204620838165, + "learning_rate": 9.423153334640407e-05, + "loss": 0.0352, + "step": 19670 + }, + { + "epoch": 3.820736857434105, + "grad_norm": 0.17646118998527527, + "learning_rate": 9.42238209242413e-05, + "loss": 0.0327, + "step": 19680 + }, + { + "epoch": 3.8226785107519055, + "grad_norm": 0.1923348754644394, + "learning_rate": 9.421610366582798e-05, + "loss": 0.0406, + "step": 19690 + }, + { + "epoch": 3.8246201640697053, + "grad_norm": 0.2965943515300751, + "learning_rate": 9.420838157200803e-05, + "loss": 0.0387, + "step": 19700 + }, + { + "epoch": 3.8265618173875056, + "grad_norm": 0.14774228632450104, + "learning_rate": 9.420065464362594e-05, + "loss": 0.0377, + "step": 19710 + }, + { + "epoch": 3.8285034707053054, + "grad_norm": 0.2651638984680176, + "learning_rate": 9.419292288152673e-05, + "loss": 0.037, + "step": 19720 + }, + { + "epoch": 3.8304451240231057, + "grad_norm": 0.12743881344795227, + "learning_rate": 9.418518628655588e-05, + "loss": 0.0339, + "step": 19730 + }, + { + "epoch": 3.832386777340906, + "grad_norm": 0.1326272040605545, + "learning_rate": 9.417744485955951e-05, + "loss": 0.0362, + "step": 19740 + }, + { + "epoch": 3.8343284306587058, + "grad_norm": 0.1379929631948471, + "learning_rate": 9.41696986013842e-05, + "loss": 0.04, + "step": 19750 + }, + { + "epoch": 3.836270083976506, + "grad_norm": 0.20534521341323853, + "learning_rate": 9.416194751287705e-05, + "loss": 0.0336, + "step": 19760 + }, + { + "epoch": 3.838211737294306, + "grad_norm": 0.2777542769908905, + "learning_rate": 9.415419159488572e-05, + "loss": 0.0361, + "step": 19770 + }, + { + "epoch": 3.840153390612106, + "grad_norm": 0.16731244325637817, + "learning_rate": 9.414643084825837e-05, + "loss": 0.0354, + "step": 19780 + }, + { + "epoch": 3.8420950439299064, + "grad_norm": 0.27633699774742126, + "learning_rate": 9.413866527384372e-05, + "loss": 0.0407, + "step": 19790 + }, + { + "epoch": 3.8440366972477067, + "grad_norm": 0.15595701336860657, + "learning_rate": 9.4130894872491e-05, + "loss": 0.0344, + "step": 19800 + }, + { + "epoch": 3.8459783505655065, + "grad_norm": 0.2740718424320221, + "learning_rate": 9.412311964504998e-05, + "loss": 0.0294, + "step": 19810 + }, + { + "epoch": 3.847920003883307, + "grad_norm": 0.22409896552562714, + "learning_rate": 9.411533959237091e-05, + "loss": 0.0353, + "step": 19820 + }, + { + "epoch": 3.8498616572011066, + "grad_norm": 0.1335069239139557, + "learning_rate": 9.410755471530464e-05, + "loss": 0.0532, + "step": 19830 + }, + { + "epoch": 3.851803310518907, + "grad_norm": 0.19443123042583466, + "learning_rate": 9.40997650147025e-05, + "loss": 0.0316, + "step": 19840 + }, + { + "epoch": 3.853744963836707, + "grad_norm": 0.14762192964553833, + "learning_rate": 9.409197049141637e-05, + "loss": 0.0305, + "step": 19850 + }, + { + "epoch": 3.855686617154507, + "grad_norm": 0.14231529831886292, + "learning_rate": 9.408417114629863e-05, + "loss": 0.0315, + "step": 19860 + }, + { + "epoch": 3.8576282704723073, + "grad_norm": 0.3699951469898224, + "learning_rate": 9.40763669802022e-05, + "loss": 0.0471, + "step": 19870 + }, + { + "epoch": 3.859569923790107, + "grad_norm": 0.29742953181266785, + "learning_rate": 9.406855799398056e-05, + "loss": 0.0418, + "step": 19880 + }, + { + "epoch": 3.8615115771079074, + "grad_norm": 0.24783195555210114, + "learning_rate": 9.406074418848767e-05, + "loss": 0.0369, + "step": 19890 + }, + { + "epoch": 3.8634532304257077, + "grad_norm": 0.2376369684934616, + "learning_rate": 9.405292556457805e-05, + "loss": 0.0319, + "step": 19900 + }, + { + "epoch": 3.8653948837435075, + "grad_norm": 0.23878943920135498, + "learning_rate": 9.404510212310671e-05, + "loss": 0.0346, + "step": 19910 + }, + { + "epoch": 3.8673365370613078, + "grad_norm": 0.25190502405166626, + "learning_rate": 9.403727386492924e-05, + "loss": 0.0329, + "step": 19920 + }, + { + "epoch": 3.8692781903791076, + "grad_norm": 0.4174947738647461, + "learning_rate": 9.40294407909017e-05, + "loss": 0.044, + "step": 19930 + }, + { + "epoch": 3.871219843696908, + "grad_norm": 0.18008436262607574, + "learning_rate": 9.40216029018807e-05, + "loss": 0.0371, + "step": 19940 + }, + { + "epoch": 3.873161497014708, + "grad_norm": 0.16764681041240692, + "learning_rate": 9.401376019872338e-05, + "loss": 0.0346, + "step": 19950 + }, + { + "epoch": 3.8751031503325084, + "grad_norm": 0.2759682834148407, + "learning_rate": 9.400591268228746e-05, + "loss": 0.0368, + "step": 19960 + }, + { + "epoch": 3.8770448036503082, + "grad_norm": 0.15382716059684753, + "learning_rate": 9.399806035343106e-05, + "loss": 0.0348, + "step": 19970 + }, + { + "epoch": 3.8789864569681085, + "grad_norm": 0.23020458221435547, + "learning_rate": 9.399020321301294e-05, + "loss": 0.0329, + "step": 19980 + }, + { + "epoch": 3.8809281102859083, + "grad_norm": 0.19434483349323273, + "learning_rate": 9.398234126189234e-05, + "loss": 0.0296, + "step": 19990 + }, + { + "epoch": 3.8828697636037086, + "grad_norm": 0.19342926144599915, + "learning_rate": 9.397447450092902e-05, + "loss": 0.0374, + "step": 20000 + }, + { + "epoch": 3.884811416921509, + "grad_norm": 0.19782738387584686, + "learning_rate": 9.39666029309833e-05, + "loss": 0.0364, + "step": 20010 + }, + { + "epoch": 3.8867530702393087, + "grad_norm": 0.20490039885044098, + "learning_rate": 9.395872655291596e-05, + "loss": 0.0376, + "step": 20020 + }, + { + "epoch": 3.888694723557109, + "grad_norm": 0.1357162594795227, + "learning_rate": 9.395084536758838e-05, + "loss": 0.0324, + "step": 20030 + }, + { + "epoch": 3.890636376874909, + "grad_norm": 0.15485891699790955, + "learning_rate": 9.394295937586243e-05, + "loss": 0.0446, + "step": 20040 + }, + { + "epoch": 3.892578030192709, + "grad_norm": 0.27295219898223877, + "learning_rate": 9.393506857860052e-05, + "loss": 0.0331, + "step": 20050 + }, + { + "epoch": 3.8945196835105094, + "grad_norm": 0.23857852816581726, + "learning_rate": 9.392717297666555e-05, + "loss": 0.0268, + "step": 20060 + }, + { + "epoch": 3.896461336828309, + "grad_norm": 0.24874627590179443, + "learning_rate": 9.391927257092101e-05, + "loss": 0.0374, + "step": 20070 + }, + { + "epoch": 3.8984029901461095, + "grad_norm": 0.271077424287796, + "learning_rate": 9.391136736223085e-05, + "loss": 0.0397, + "step": 20080 + }, + { + "epoch": 3.9003446434639093, + "grad_norm": 0.21257264912128448, + "learning_rate": 9.390345735145956e-05, + "loss": 0.0307, + "step": 20090 + }, + { + "epoch": 3.9022862967817096, + "grad_norm": 0.17404496669769287, + "learning_rate": 9.389554253947219e-05, + "loss": 0.0294, + "step": 20100 + }, + { + "epoch": 3.90422795009951, + "grad_norm": 0.21520502865314484, + "learning_rate": 9.388762292713428e-05, + "loss": 0.0348, + "step": 20110 + }, + { + "epoch": 3.9061696034173097, + "grad_norm": 0.17227308452129364, + "learning_rate": 9.38796985153119e-05, + "loss": 0.0341, + "step": 20120 + }, + { + "epoch": 3.90811125673511, + "grad_norm": 0.22399736940860748, + "learning_rate": 9.387176930487169e-05, + "loss": 0.0315, + "step": 20130 + }, + { + "epoch": 3.91005291005291, + "grad_norm": 0.24731740355491638, + "learning_rate": 9.386383529668072e-05, + "loss": 0.0413, + "step": 20140 + }, + { + "epoch": 3.91199456337071, + "grad_norm": 0.2952998876571655, + "learning_rate": 9.385589649160669e-05, + "loss": 0.0398, + "step": 20150 + }, + { + "epoch": 3.9139362166885103, + "grad_norm": 0.12535908818244934, + "learning_rate": 9.384795289051775e-05, + "loss": 0.0383, + "step": 20160 + }, + { + "epoch": 3.9158778700063106, + "grad_norm": 0.19065077602863312, + "learning_rate": 9.384000449428261e-05, + "loss": 0.0327, + "step": 20170 + }, + { + "epoch": 3.9178195233241104, + "grad_norm": 0.33152851462364197, + "learning_rate": 9.383205130377048e-05, + "loss": 0.0357, + "step": 20180 + }, + { + "epoch": 3.9197611766419107, + "grad_norm": 0.19719283282756805, + "learning_rate": 9.382409331985114e-05, + "loss": 0.0403, + "step": 20190 + }, + { + "epoch": 3.9217028299597105, + "grad_norm": 0.1460149884223938, + "learning_rate": 9.381613054339482e-05, + "loss": 0.0366, + "step": 20200 + }, + { + "epoch": 3.923644483277511, + "grad_norm": 0.14023126661777496, + "learning_rate": 9.380816297527235e-05, + "loss": 0.0313, + "step": 20210 + }, + { + "epoch": 3.925586136595311, + "grad_norm": 0.22469428181648254, + "learning_rate": 9.380019061635506e-05, + "loss": 0.0431, + "step": 20220 + }, + { + "epoch": 3.927527789913111, + "grad_norm": 0.2516086995601654, + "learning_rate": 9.379221346751474e-05, + "loss": 0.0272, + "step": 20230 + }, + { + "epoch": 3.929469443230911, + "grad_norm": 0.30200326442718506, + "learning_rate": 9.378423152962382e-05, + "loss": 0.0401, + "step": 20240 + }, + { + "epoch": 3.931411096548711, + "grad_norm": 0.16075672209262848, + "learning_rate": 9.377624480355517e-05, + "loss": 0.0325, + "step": 20250 + }, + { + "epoch": 3.9333527498665113, + "grad_norm": 0.13555273413658142, + "learning_rate": 9.376825329018219e-05, + "loss": 0.0332, + "step": 20260 + }, + { + "epoch": 3.9352944031843116, + "grad_norm": 0.1668381690979004, + "learning_rate": 9.376025699037884e-05, + "loss": 0.0305, + "step": 20270 + }, + { + "epoch": 3.9372360565021114, + "grad_norm": 0.14484405517578125, + "learning_rate": 9.37522559050196e-05, + "loss": 0.0349, + "step": 20280 + }, + { + "epoch": 3.9391777098199117, + "grad_norm": 0.25772860646247864, + "learning_rate": 9.37442500349794e-05, + "loss": 0.0355, + "step": 20290 + }, + { + "epoch": 3.9411193631377115, + "grad_norm": 0.1498018205165863, + "learning_rate": 9.373623938113381e-05, + "loss": 0.0346, + "step": 20300 + }, + { + "epoch": 3.9430610164555118, + "grad_norm": 0.19465216994285583, + "learning_rate": 9.372822394435883e-05, + "loss": 0.0359, + "step": 20310 + }, + { + "epoch": 3.945002669773312, + "grad_norm": 0.14734511077404022, + "learning_rate": 9.372020372553102e-05, + "loss": 0.0328, + "step": 20320 + }, + { + "epoch": 3.9469443230911123, + "grad_norm": 0.22113220393657684, + "learning_rate": 9.371217872552746e-05, + "loss": 0.0307, + "step": 20330 + }, + { + "epoch": 3.948885976408912, + "grad_norm": 0.12414927780628204, + "learning_rate": 9.370414894522576e-05, + "loss": 0.0353, + "step": 20340 + }, + { + "epoch": 3.9508276297267124, + "grad_norm": 0.20562897622585297, + "learning_rate": 9.369611438550406e-05, + "loss": 0.0364, + "step": 20350 + }, + { + "epoch": 3.9527692830445122, + "grad_norm": 0.232610285282135, + "learning_rate": 9.368807504724095e-05, + "loss": 0.0388, + "step": 20360 + }, + { + "epoch": 3.9547109363623125, + "grad_norm": 0.251353919506073, + "learning_rate": 9.368003093131565e-05, + "loss": 0.0339, + "step": 20370 + }, + { + "epoch": 3.956652589680113, + "grad_norm": 0.19874629378318787, + "learning_rate": 9.367198203860785e-05, + "loss": 0.0326, + "step": 20380 + }, + { + "epoch": 3.9585942429979126, + "grad_norm": 0.25015807151794434, + "learning_rate": 9.366392836999774e-05, + "loss": 0.0357, + "step": 20390 + }, + { + "epoch": 3.960535896315713, + "grad_norm": 0.17219214141368866, + "learning_rate": 9.365586992636607e-05, + "loss": 0.0369, + "step": 20400 + }, + { + "epoch": 3.9624775496335127, + "grad_norm": 0.20875412225723267, + "learning_rate": 9.364780670859412e-05, + "loss": 0.0374, + "step": 20410 + }, + { + "epoch": 3.964419202951313, + "grad_norm": 0.19962722063064575, + "learning_rate": 9.363973871756364e-05, + "loss": 0.0313, + "step": 20420 + }, + { + "epoch": 3.9663608562691133, + "grad_norm": 0.1896151453256607, + "learning_rate": 9.363166595415696e-05, + "loss": 0.0363, + "step": 20430 + }, + { + "epoch": 3.968302509586913, + "grad_norm": 0.22481119632720947, + "learning_rate": 9.362358841925686e-05, + "loss": 0.0353, + "step": 20440 + }, + { + "epoch": 3.9702441629047134, + "grad_norm": 0.1942598521709442, + "learning_rate": 9.361550611374674e-05, + "loss": 0.0372, + "step": 20450 + }, + { + "epoch": 3.972185816222513, + "grad_norm": 0.1768556535243988, + "learning_rate": 9.360741903851043e-05, + "loss": 0.0323, + "step": 20460 + }, + { + "epoch": 3.9741274695403135, + "grad_norm": 0.19001415371894836, + "learning_rate": 9.359932719443236e-05, + "loss": 0.0327, + "step": 20470 + }, + { + "epoch": 3.9760691228581138, + "grad_norm": 0.1707480549812317, + "learning_rate": 9.35912305823974e-05, + "loss": 0.0341, + "step": 20480 + }, + { + "epoch": 3.978010776175914, + "grad_norm": 0.15119564533233643, + "learning_rate": 9.358312920329101e-05, + "loss": 0.0334, + "step": 20490 + }, + { + "epoch": 3.979952429493714, + "grad_norm": 0.19147314131259918, + "learning_rate": 9.357502305799914e-05, + "loss": 0.03, + "step": 20500 + }, + { + "epoch": 3.981894082811514, + "grad_norm": 0.251453161239624, + "learning_rate": 9.356691214740824e-05, + "loss": 0.035, + "step": 20510 + }, + { + "epoch": 3.983835736129314, + "grad_norm": 0.16877219080924988, + "learning_rate": 9.355879647240535e-05, + "loss": 0.0394, + "step": 20520 + }, + { + "epoch": 3.9857773894471142, + "grad_norm": 0.18306972086429596, + "learning_rate": 9.355067603387798e-05, + "loss": 0.0342, + "step": 20530 + }, + { + "epoch": 3.9877190427649145, + "grad_norm": 0.22380542755126953, + "learning_rate": 9.354255083271412e-05, + "loss": 0.0341, + "step": 20540 + }, + { + "epoch": 3.9896606960827143, + "grad_norm": 0.15691910684108734, + "learning_rate": 9.353442086980239e-05, + "loss": 0.0397, + "step": 20550 + }, + { + "epoch": 3.9916023494005146, + "grad_norm": 0.17560677230358124, + "learning_rate": 9.352628614603185e-05, + "loss": 0.0317, + "step": 20560 + }, + { + "epoch": 3.9935440027183144, + "grad_norm": 0.22879457473754883, + "learning_rate": 9.351814666229209e-05, + "loss": 0.0429, + "step": 20570 + }, + { + "epoch": 3.9954856560361147, + "grad_norm": 0.1841765195131302, + "learning_rate": 9.351000241947324e-05, + "loss": 0.03, + "step": 20580 + }, + { + "epoch": 3.997427309353915, + "grad_norm": 0.27976250648498535, + "learning_rate": 9.350185341846594e-05, + "loss": 0.0386, + "step": 20590 + }, + { + "epoch": 3.999368962671715, + "grad_norm": 0.16760759055614471, + "learning_rate": 9.349369966016134e-05, + "loss": 0.0327, + "step": 20600 + }, + { + "epoch": 4.00116499199068, + "grad_norm": 0.3333459496498108, + "learning_rate": 9.348554114545117e-05, + "loss": 0.0334, + "step": 20610 + }, + { + "epoch": 4.00310664530848, + "grad_norm": 0.1859004944562912, + "learning_rate": 9.347737787522758e-05, + "loss": 0.0416, + "step": 20620 + }, + { + "epoch": 4.0050482986262805, + "grad_norm": 0.31889608502388, + "learning_rate": 9.346920985038332e-05, + "loss": 0.0407, + "step": 20630 + }, + { + "epoch": 4.006989951944081, + "grad_norm": 0.22701554000377655, + "learning_rate": 9.346103707181162e-05, + "loss": 0.0375, + "step": 20640 + }, + { + "epoch": 4.00893160526188, + "grad_norm": 0.1906045526266098, + "learning_rate": 9.345285954040626e-05, + "loss": 0.0326, + "step": 20650 + }, + { + "epoch": 4.0108732585796805, + "grad_norm": 0.2650797665119171, + "learning_rate": 9.34446772570615e-05, + "loss": 0.0334, + "step": 20660 + }, + { + "epoch": 4.012814911897481, + "grad_norm": 0.23639295995235443, + "learning_rate": 9.343649022267214e-05, + "loss": 0.0345, + "step": 20670 + }, + { + "epoch": 4.014756565215281, + "grad_norm": 0.1685684472322464, + "learning_rate": 9.342829843813353e-05, + "loss": 0.0307, + "step": 20680 + }, + { + "epoch": 4.016698218533081, + "grad_norm": 0.2601465880870819, + "learning_rate": 9.342010190434149e-05, + "loss": 0.0445, + "step": 20690 + }, + { + "epoch": 4.018639871850881, + "grad_norm": 0.20671816170215607, + "learning_rate": 9.34119006221924e-05, + "loss": 0.0288, + "step": 20700 + }, + { + "epoch": 4.020581525168681, + "grad_norm": 0.18012258410453796, + "learning_rate": 9.340369459258313e-05, + "loss": 0.0325, + "step": 20710 + }, + { + "epoch": 4.022523178486481, + "grad_norm": 0.2786230146884918, + "learning_rate": 9.339548381641106e-05, + "loss": 0.0384, + "step": 20720 + }, + { + "epoch": 4.0244648318042815, + "grad_norm": 0.17471931874752045, + "learning_rate": 9.338726829457413e-05, + "loss": 0.0293, + "step": 20730 + }, + { + "epoch": 4.026406485122082, + "grad_norm": 0.18964523077011108, + "learning_rate": 9.337904802797078e-05, + "loss": 0.032, + "step": 20740 + }, + { + "epoch": 4.028348138439881, + "grad_norm": 0.21571344137191772, + "learning_rate": 9.337082301749993e-05, + "loss": 0.0377, + "step": 20750 + }, + { + "epoch": 4.030289791757681, + "grad_norm": 0.2493242770433426, + "learning_rate": 9.336259326406109e-05, + "loss": 0.0365, + "step": 20760 + }, + { + "epoch": 4.032231445075482, + "grad_norm": 0.10913172364234924, + "learning_rate": 9.335435876855427e-05, + "loss": 0.0316, + "step": 20770 + }, + { + "epoch": 4.034173098393282, + "grad_norm": 0.23476694524288177, + "learning_rate": 9.334611953187994e-05, + "loss": 0.028, + "step": 20780 + }, + { + "epoch": 4.036114751711082, + "grad_norm": 0.17237527668476105, + "learning_rate": 9.333787555493914e-05, + "loss": 0.0401, + "step": 20790 + }, + { + "epoch": 4.0380564050288825, + "grad_norm": 0.1962307244539261, + "learning_rate": 9.332962683863345e-05, + "loss": 0.0386, + "step": 20800 + }, + { + "epoch": 4.039998058346682, + "grad_norm": 0.17237058281898499, + "learning_rate": 9.332137338386489e-05, + "loss": 0.0307, + "step": 20810 + }, + { + "epoch": 4.041939711664482, + "grad_norm": 0.22380022704601288, + "learning_rate": 9.33131151915361e-05, + "loss": 0.0402, + "step": 20820 + }, + { + "epoch": 4.0438813649822825, + "grad_norm": 0.16876189410686493, + "learning_rate": 9.330485226255012e-05, + "loss": 0.0291, + "step": 20830 + }, + { + "epoch": 4.045823018300083, + "grad_norm": 0.17966768145561218, + "learning_rate": 9.329658459781061e-05, + "loss": 0.0383, + "step": 20840 + }, + { + "epoch": 4.047764671617883, + "grad_norm": 0.17755667865276337, + "learning_rate": 9.328831219822172e-05, + "loss": 0.0297, + "step": 20850 + }, + { + "epoch": 4.049706324935682, + "grad_norm": 0.16743957996368408, + "learning_rate": 9.328003506468808e-05, + "loss": 0.0285, + "step": 20860 + }, + { + "epoch": 4.051647978253483, + "grad_norm": 0.25864654779434204, + "learning_rate": 9.327175319811488e-05, + "loss": 0.0385, + "step": 20870 + }, + { + "epoch": 4.053589631571283, + "grad_norm": 0.18599653244018555, + "learning_rate": 9.326346659940781e-05, + "loss": 0.0325, + "step": 20880 + }, + { + "epoch": 4.055531284889083, + "grad_norm": 0.16717016696929932, + "learning_rate": 9.325517526947308e-05, + "loss": 0.0301, + "step": 20890 + }, + { + "epoch": 4.0574729382068835, + "grad_norm": 0.13855499029159546, + "learning_rate": 9.32468792092174e-05, + "loss": 0.0324, + "step": 20900 + }, + { + "epoch": 4.059414591524683, + "grad_norm": 0.15961290895938873, + "learning_rate": 9.323857841954803e-05, + "loss": 0.0357, + "step": 20910 + }, + { + "epoch": 4.061356244842483, + "grad_norm": 0.15313702821731567, + "learning_rate": 9.323027290137276e-05, + "loss": 0.0266, + "step": 20920 + }, + { + "epoch": 4.063297898160283, + "grad_norm": 0.17955592274665833, + "learning_rate": 9.322196265559981e-05, + "loss": 0.0402, + "step": 20930 + }, + { + "epoch": 4.065239551478084, + "grad_norm": 0.22810234129428864, + "learning_rate": 9.321364768313803e-05, + "loss": 0.0379, + "step": 20940 + }, + { + "epoch": 4.067181204795884, + "grad_norm": 0.18233297765254974, + "learning_rate": 9.32053279848967e-05, + "loss": 0.0286, + "step": 20950 + }, + { + "epoch": 4.069122858113684, + "grad_norm": 0.18556304275989532, + "learning_rate": 9.319700356178567e-05, + "loss": 0.0371, + "step": 20960 + }, + { + "epoch": 4.071064511431484, + "grad_norm": 0.17565131187438965, + "learning_rate": 9.318867441471527e-05, + "loss": 0.0361, + "step": 20970 + }, + { + "epoch": 4.073006164749284, + "grad_norm": 0.1724223494529724, + "learning_rate": 9.318034054459637e-05, + "loss": 0.0278, + "step": 20980 + }, + { + "epoch": 4.074947818067084, + "grad_norm": 0.20785078406333923, + "learning_rate": 9.317200195234034e-05, + "loss": 0.0314, + "step": 20990 + }, + { + "epoch": 4.0768894713848844, + "grad_norm": 0.24459482729434967, + "learning_rate": 9.316365863885909e-05, + "loss": 0.0333, + "step": 21000 + }, + { + "epoch": 4.078831124702685, + "grad_norm": 0.20183688402175903, + "learning_rate": 9.315531060506502e-05, + "loss": 0.0283, + "step": 21010 + }, + { + "epoch": 4.080772778020484, + "grad_norm": 0.17964035272598267, + "learning_rate": 9.314695785187108e-05, + "loss": 0.0354, + "step": 21020 + }, + { + "epoch": 4.082714431338284, + "grad_norm": 0.2768024802207947, + "learning_rate": 9.313860038019069e-05, + "loss": 0.0417, + "step": 21030 + }, + { + "epoch": 4.084656084656085, + "grad_norm": 0.20072638988494873, + "learning_rate": 9.313023819093782e-05, + "loss": 0.0326, + "step": 21040 + }, + { + "epoch": 4.086597737973885, + "grad_norm": 0.14750972390174866, + "learning_rate": 9.312187128502695e-05, + "loss": 0.0346, + "step": 21050 + }, + { + "epoch": 4.088539391291685, + "grad_norm": 0.2168748825788498, + "learning_rate": 9.311349966337307e-05, + "loss": 0.0315, + "step": 21060 + }, + { + "epoch": 4.090481044609485, + "grad_norm": 0.17136763036251068, + "learning_rate": 9.310512332689169e-05, + "loss": 0.0316, + "step": 21070 + }, + { + "epoch": 4.092422697927285, + "grad_norm": 0.24152891337871552, + "learning_rate": 9.309674227649883e-05, + "loss": 0.0316, + "step": 21080 + }, + { + "epoch": 4.094364351245085, + "grad_norm": 0.11427599936723709, + "learning_rate": 9.308835651311103e-05, + "loss": 0.0311, + "step": 21090 + }, + { + "epoch": 4.096306004562885, + "grad_norm": 0.16600266098976135, + "learning_rate": 9.307996603764533e-05, + "loss": 0.0302, + "step": 21100 + }, + { + "epoch": 4.098247657880686, + "grad_norm": 0.1777990609407425, + "learning_rate": 9.307157085101932e-05, + "loss": 0.0293, + "step": 21110 + }, + { + "epoch": 4.100189311198486, + "grad_norm": 0.23210132122039795, + "learning_rate": 9.306317095415109e-05, + "loss": 0.0348, + "step": 21120 + }, + { + "epoch": 4.102130964516285, + "grad_norm": 0.18164995312690735, + "learning_rate": 9.305476634795922e-05, + "loss": 0.0362, + "step": 21130 + }, + { + "epoch": 4.104072617834086, + "grad_norm": 0.11163429915904999, + "learning_rate": 9.304635703336284e-05, + "loss": 0.0341, + "step": 21140 + }, + { + "epoch": 4.106014271151886, + "grad_norm": 0.1590149998664856, + "learning_rate": 9.303794301128157e-05, + "loss": 0.0321, + "step": 21150 + }, + { + "epoch": 4.107955924469686, + "grad_norm": 0.12935303151607513, + "learning_rate": 9.302952428263555e-05, + "loss": 0.0261, + "step": 21160 + }, + { + "epoch": 4.109897577787486, + "grad_norm": 0.21589943766593933, + "learning_rate": 9.302110084834545e-05, + "loss": 0.0391, + "step": 21170 + }, + { + "epoch": 4.111839231105286, + "grad_norm": 0.3096737861633301, + "learning_rate": 9.301267270933245e-05, + "loss": 0.0427, + "step": 21180 + }, + { + "epoch": 4.113780884423086, + "grad_norm": 0.14556783437728882, + "learning_rate": 9.300423986651823e-05, + "loss": 0.0359, + "step": 21190 + }, + { + "epoch": 4.115722537740886, + "grad_norm": 0.20317260921001434, + "learning_rate": 9.299580232082501e-05, + "loss": 0.0376, + "step": 21200 + }, + { + "epoch": 4.117664191058687, + "grad_norm": 0.1723712831735611, + "learning_rate": 9.298736007317547e-05, + "loss": 0.0325, + "step": 21210 + }, + { + "epoch": 4.119605844376487, + "grad_norm": 0.188360795378685, + "learning_rate": 9.297891312449288e-05, + "loss": 0.0317, + "step": 21220 + }, + { + "epoch": 4.121547497694286, + "grad_norm": 0.21352708339691162, + "learning_rate": 9.297046147570094e-05, + "loss": 0.0372, + "step": 21230 + }, + { + "epoch": 4.123489151012087, + "grad_norm": 0.16206999123096466, + "learning_rate": 9.296200512772396e-05, + "loss": 0.0335, + "step": 21240 + }, + { + "epoch": 4.125430804329887, + "grad_norm": 0.15794219076633453, + "learning_rate": 9.295354408148668e-05, + "loss": 0.0323, + "step": 21250 + }, + { + "epoch": 4.127372457647687, + "grad_norm": 0.6373804807662964, + "learning_rate": 9.294507833791441e-05, + "loss": 0.0352, + "step": 21260 + }, + { + "epoch": 4.129314110965487, + "grad_norm": 0.4457343518733978, + "learning_rate": 9.293660789793295e-05, + "loss": 0.0385, + "step": 21270 + }, + { + "epoch": 4.131255764283287, + "grad_norm": 0.10798870772123337, + "learning_rate": 9.292813276246858e-05, + "loss": 0.0307, + "step": 21280 + }, + { + "epoch": 4.133197417601087, + "grad_norm": 0.22745828330516815, + "learning_rate": 9.291965293244816e-05, + "loss": 0.0304, + "step": 21290 + }, + { + "epoch": 4.135139070918887, + "grad_norm": 0.2442256063222885, + "learning_rate": 9.291116840879904e-05, + "loss": 0.033, + "step": 21300 + }, + { + "epoch": 4.137080724236688, + "grad_norm": 0.32506221532821655, + "learning_rate": 9.290267919244904e-05, + "loss": 0.035, + "step": 21310 + }, + { + "epoch": 4.139022377554488, + "grad_norm": 0.19819606840610504, + "learning_rate": 9.289418528432655e-05, + "loss": 0.0387, + "step": 21320 + }, + { + "epoch": 4.140964030872288, + "grad_norm": 0.21910184621810913, + "learning_rate": 9.288568668536045e-05, + "loss": 0.0356, + "step": 21330 + }, + { + "epoch": 4.1429056841900875, + "grad_norm": 0.17214073240756989, + "learning_rate": 9.287718339648013e-05, + "loss": 0.0281, + "step": 21340 + }, + { + "epoch": 4.144847337507888, + "grad_norm": 0.1624045968055725, + "learning_rate": 9.28686754186155e-05, + "loss": 0.0336, + "step": 21350 + }, + { + "epoch": 4.146788990825688, + "grad_norm": 0.249050572514534, + "learning_rate": 9.286016275269698e-05, + "loss": 0.0312, + "step": 21360 + }, + { + "epoch": 4.148730644143488, + "grad_norm": 0.1984904259443283, + "learning_rate": 9.285164539965551e-05, + "loss": 0.0274, + "step": 21370 + }, + { + "epoch": 4.150672297461289, + "grad_norm": 0.18217112123966217, + "learning_rate": 9.284312336042251e-05, + "loss": 0.0343, + "step": 21380 + }, + { + "epoch": 4.152613950779088, + "grad_norm": 0.25430062413215637, + "learning_rate": 9.283459663592996e-05, + "loss": 0.035, + "step": 21390 + }, + { + "epoch": 4.154555604096888, + "grad_norm": 0.15371721982955933, + "learning_rate": 9.282606522711033e-05, + "loss": 0.0344, + "step": 21400 + }, + { + "epoch": 4.156497257414689, + "grad_norm": 0.32763341069221497, + "learning_rate": 9.281752913489657e-05, + "loss": 0.0348, + "step": 21410 + }, + { + "epoch": 4.158438910732489, + "grad_norm": 0.2027217149734497, + "learning_rate": 9.280898836022222e-05, + "loss": 0.0293, + "step": 21420 + }, + { + "epoch": 4.160380564050289, + "grad_norm": 0.34386685490608215, + "learning_rate": 9.280044290402126e-05, + "loss": 0.0376, + "step": 21430 + }, + { + "epoch": 4.162322217368089, + "grad_norm": 0.15221650898456573, + "learning_rate": 9.279189276722821e-05, + "loss": 0.0309, + "step": 21440 + }, + { + "epoch": 4.164263870685889, + "grad_norm": 0.3076215088367462, + "learning_rate": 9.278333795077812e-05, + "loss": 0.0294, + "step": 21450 + }, + { + "epoch": 4.166205524003689, + "grad_norm": 0.23586197197437286, + "learning_rate": 9.27747784556065e-05, + "loss": 0.0261, + "step": 21460 + }, + { + "epoch": 4.168147177321489, + "grad_norm": 0.1361730992794037, + "learning_rate": 9.276621428264942e-05, + "loss": 0.0345, + "step": 21470 + }, + { + "epoch": 4.17008883063929, + "grad_norm": 0.1430988311767578, + "learning_rate": 9.275764543284345e-05, + "loss": 0.0303, + "step": 21480 + }, + { + "epoch": 4.17203048395709, + "grad_norm": 0.1956777274608612, + "learning_rate": 9.274907190712566e-05, + "loss": 0.0354, + "step": 21490 + }, + { + "epoch": 4.173972137274889, + "grad_norm": 0.15191742777824402, + "learning_rate": 9.274049370643363e-05, + "loss": 0.0274, + "step": 21500 + }, + { + "epoch": 4.1759137905926895, + "grad_norm": 0.21861737966537476, + "learning_rate": 9.273191083170547e-05, + "loss": 0.032, + "step": 21510 + }, + { + "epoch": 4.17785544391049, + "grad_norm": 0.20480577647686005, + "learning_rate": 9.27233232838798e-05, + "loss": 0.0397, + "step": 21520 + }, + { + "epoch": 4.17979709722829, + "grad_norm": 0.23564130067825317, + "learning_rate": 9.27147310638957e-05, + "loss": 0.031, + "step": 21530 + }, + { + "epoch": 4.18173875054609, + "grad_norm": 0.2924972474575043, + "learning_rate": 9.270613417269286e-05, + "loss": 0.0372, + "step": 21540 + }, + { + "epoch": 4.18368040386389, + "grad_norm": 0.1851537525653839, + "learning_rate": 9.269753261121138e-05, + "loss": 0.0373, + "step": 21550 + }, + { + "epoch": 4.18562205718169, + "grad_norm": 0.2101287990808487, + "learning_rate": 9.268892638039194e-05, + "loss": 0.0385, + "step": 21560 + }, + { + "epoch": 4.18756371049949, + "grad_norm": 0.17215651273727417, + "learning_rate": 9.268031548117569e-05, + "loss": 0.0265, + "step": 21570 + }, + { + "epoch": 4.1895053638172906, + "grad_norm": 0.3067096769809723, + "learning_rate": 9.26716999145043e-05, + "loss": 0.0426, + "step": 21580 + }, + { + "epoch": 4.191447017135091, + "grad_norm": 0.20475134253501892, + "learning_rate": 9.266307968131998e-05, + "loss": 0.0375, + "step": 21590 + }, + { + "epoch": 4.19338867045289, + "grad_norm": 0.18695242702960968, + "learning_rate": 9.26544547825654e-05, + "loss": 0.0275, + "step": 21600 + }, + { + "epoch": 4.1953303237706905, + "grad_norm": 0.1589411199092865, + "learning_rate": 9.264582521918376e-05, + "loss": 0.0329, + "step": 21610 + }, + { + "epoch": 4.197271977088491, + "grad_norm": 0.14176344871520996, + "learning_rate": 9.263719099211881e-05, + "loss": 0.0361, + "step": 21620 + }, + { + "epoch": 4.199213630406291, + "grad_norm": 0.20472420752048492, + "learning_rate": 9.262855210231476e-05, + "loss": 0.0354, + "step": 21630 + }, + { + "epoch": 4.201155283724091, + "grad_norm": 0.17693950235843658, + "learning_rate": 9.261990855071633e-05, + "loss": 0.0392, + "step": 21640 + }, + { + "epoch": 4.203096937041892, + "grad_norm": 0.18654713034629822, + "learning_rate": 9.261126033826878e-05, + "loss": 0.0319, + "step": 21650 + }, + { + "epoch": 4.205038590359691, + "grad_norm": 0.2720799744129181, + "learning_rate": 9.260260746591786e-05, + "loss": 0.0356, + "step": 21660 + }, + { + "epoch": 4.206980243677491, + "grad_norm": 0.3044658303260803, + "learning_rate": 9.259394993460985e-05, + "loss": 0.0362, + "step": 21670 + }, + { + "epoch": 4.2089218969952915, + "grad_norm": 0.24976299703121185, + "learning_rate": 9.258528774529151e-05, + "loss": 0.0377, + "step": 21680 + }, + { + "epoch": 4.210863550313092, + "grad_norm": 0.23173952102661133, + "learning_rate": 9.257662089891013e-05, + "loss": 0.0308, + "step": 21690 + }, + { + "epoch": 4.212805203630892, + "grad_norm": 0.16410350799560547, + "learning_rate": 9.25679493964135e-05, + "loss": 0.0331, + "step": 21700 + }, + { + "epoch": 4.2147468569486914, + "grad_norm": 0.15623325109481812, + "learning_rate": 9.255927323874994e-05, + "loss": 0.0311, + "step": 21710 + }, + { + "epoch": 4.216688510266492, + "grad_norm": 0.24036052823066711, + "learning_rate": 9.255059242686822e-05, + "loss": 0.0285, + "step": 21720 + }, + { + "epoch": 4.218630163584292, + "grad_norm": 0.24893449246883392, + "learning_rate": 9.254190696171769e-05, + "loss": 0.0364, + "step": 21730 + }, + { + "epoch": 4.220571816902092, + "grad_norm": 0.15583430230617523, + "learning_rate": 9.25332168442482e-05, + "loss": 0.0326, + "step": 21740 + }, + { + "epoch": 4.2225134702198925, + "grad_norm": 0.14663222432136536, + "learning_rate": 9.252452207541004e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 4.224455123537692, + "grad_norm": 0.1986563503742218, + "learning_rate": 9.251582265615409e-05, + "loss": 0.034, + "step": 21760 + }, + { + "epoch": 4.226396776855492, + "grad_norm": 0.2134586125612259, + "learning_rate": 9.250711858743169e-05, + "loss": 0.0354, + "step": 21770 + }, + { + "epoch": 4.2283384301732925, + "grad_norm": 0.3400198519229889, + "learning_rate": 9.24984098701947e-05, + "loss": 0.0327, + "step": 21780 + }, + { + "epoch": 4.230280083491093, + "grad_norm": 0.24740393459796906, + "learning_rate": 9.248969650539552e-05, + "loss": 0.0369, + "step": 21790 + }, + { + "epoch": 4.232221736808893, + "grad_norm": 0.15620581805706024, + "learning_rate": 9.2480978493987e-05, + "loss": 0.0268, + "step": 21800 + }, + { + "epoch": 4.234163390126693, + "grad_norm": 0.312980055809021, + "learning_rate": 9.247225583692256e-05, + "loss": 0.0367, + "step": 21810 + }, + { + "epoch": 4.236105043444493, + "grad_norm": 0.272513747215271, + "learning_rate": 9.246352853515607e-05, + "loss": 0.0329, + "step": 21820 + }, + { + "epoch": 4.238046696762293, + "grad_norm": 0.15584488213062286, + "learning_rate": 9.245479658964194e-05, + "loss": 0.0303, + "step": 21830 + }, + { + "epoch": 4.239988350080093, + "grad_norm": 0.17077769339084625, + "learning_rate": 9.244606000133507e-05, + "loss": 0.028, + "step": 21840 + }, + { + "epoch": 4.2419300033978935, + "grad_norm": 0.21817132830619812, + "learning_rate": 9.24373187711909e-05, + "loss": 0.0337, + "step": 21850 + }, + { + "epoch": 4.243871656715694, + "grad_norm": 0.1826409548521042, + "learning_rate": 9.242857290016537e-05, + "loss": 0.0377, + "step": 21860 + }, + { + "epoch": 4.245813310033493, + "grad_norm": 0.25927016139030457, + "learning_rate": 9.241982238921488e-05, + "loss": 0.0314, + "step": 21870 + }, + { + "epoch": 4.247754963351293, + "grad_norm": 0.15132056176662445, + "learning_rate": 9.24110672392964e-05, + "loss": 0.0312, + "step": 21880 + }, + { + "epoch": 4.249696616669094, + "grad_norm": 0.2028547078371048, + "learning_rate": 9.240230745136737e-05, + "loss": 0.04, + "step": 21890 + }, + { + "epoch": 4.251638269986894, + "grad_norm": 0.260329931974411, + "learning_rate": 9.239354302638575e-05, + "loss": 0.0311, + "step": 21900 + }, + { + "epoch": 4.253579923304694, + "grad_norm": 0.1099710762500763, + "learning_rate": 9.238477396531e-05, + "loss": 0.0356, + "step": 21910 + }, + { + "epoch": 4.255521576622494, + "grad_norm": 0.2364816665649414, + "learning_rate": 9.23760002690991e-05, + "loss": 0.0322, + "step": 21920 + }, + { + "epoch": 4.257463229940294, + "grad_norm": 0.111485056579113, + "learning_rate": 9.236722193871252e-05, + "loss": 0.0353, + "step": 21930 + }, + { + "epoch": 4.259404883258094, + "grad_norm": 0.16246241331100464, + "learning_rate": 9.235843897511023e-05, + "loss": 0.0298, + "step": 21940 + }, + { + "epoch": 4.2613465365758945, + "grad_norm": 0.1776793897151947, + "learning_rate": 9.234965137925276e-05, + "loss": 0.0299, + "step": 21950 + }, + { + "epoch": 4.263288189893695, + "grad_norm": 0.0992666706442833, + "learning_rate": 9.234085915210108e-05, + "loss": 0.036, + "step": 21960 + }, + { + "epoch": 4.265229843211495, + "grad_norm": 0.19457785785198212, + "learning_rate": 9.23320622946167e-05, + "loss": 0.0346, + "step": 21970 + }, + { + "epoch": 4.267171496529294, + "grad_norm": 0.1687958985567093, + "learning_rate": 9.232326080776163e-05, + "loss": 0.0309, + "step": 21980 + }, + { + "epoch": 4.269113149847095, + "grad_norm": 0.2754058539867401, + "learning_rate": 9.23144546924984e-05, + "loss": 0.038, + "step": 21990 + }, + { + "epoch": 4.271054803164895, + "grad_norm": 0.18923448026180267, + "learning_rate": 9.230564394979e-05, + "loss": 0.0438, + "step": 22000 + }, + { + "epoch": 4.272996456482695, + "grad_norm": 0.24606242775917053, + "learning_rate": 9.22968285806e-05, + "loss": 0.0343, + "step": 22010 + }, + { + "epoch": 4.2749381098004955, + "grad_norm": 0.17630209028720856, + "learning_rate": 9.228800858589242e-05, + "loss": 0.0358, + "step": 22020 + }, + { + "epoch": 4.276879763118295, + "grad_norm": 0.16561724245548248, + "learning_rate": 9.227918396663179e-05, + "loss": 0.0274, + "step": 22030 + }, + { + "epoch": 4.278821416436095, + "grad_norm": 0.22854569554328918, + "learning_rate": 9.227035472378319e-05, + "loss": 0.0316, + "step": 22040 + }, + { + "epoch": 4.280763069753895, + "grad_norm": 0.22712038457393646, + "learning_rate": 9.226152085831213e-05, + "loss": 0.0391, + "step": 22050 + }, + { + "epoch": 4.282704723071696, + "grad_norm": 0.12096705287694931, + "learning_rate": 9.22526823711847e-05, + "loss": 0.038, + "step": 22060 + }, + { + "epoch": 4.284646376389496, + "grad_norm": 0.10439678281545639, + "learning_rate": 9.224383926336745e-05, + "loss": 0.0301, + "step": 22070 + }, + { + "epoch": 4.286588029707295, + "grad_norm": 0.17507021129131317, + "learning_rate": 9.223499153582744e-05, + "loss": 0.0316, + "step": 22080 + }, + { + "epoch": 4.288529683025096, + "grad_norm": 0.16345931589603424, + "learning_rate": 9.222613918953226e-05, + "loss": 0.0268, + "step": 22090 + }, + { + "epoch": 4.290471336342896, + "grad_norm": 0.11616241931915283, + "learning_rate": 9.221728222544999e-05, + "loss": 0.0337, + "step": 22100 + }, + { + "epoch": 4.292412989660696, + "grad_norm": 0.15001656115055084, + "learning_rate": 9.22084206445492e-05, + "loss": 0.0358, + "step": 22110 + }, + { + "epoch": 4.2943546429784965, + "grad_norm": 0.18197529017925262, + "learning_rate": 9.2199554447799e-05, + "loss": 0.027, + "step": 22120 + }, + { + "epoch": 4.296296296296296, + "grad_norm": 0.19221335649490356, + "learning_rate": 9.219068363616897e-05, + "loss": 0.0335, + "step": 22130 + }, + { + "epoch": 4.298237949614096, + "grad_norm": 0.20148824155330658, + "learning_rate": 9.218180821062919e-05, + "loss": 0.0332, + "step": 22140 + }, + { + "epoch": 4.300179602931896, + "grad_norm": 0.24894872307777405, + "learning_rate": 9.21729281721503e-05, + "loss": 0.0349, + "step": 22150 + }, + { + "epoch": 4.302121256249697, + "grad_norm": 0.22707194089889526, + "learning_rate": 9.216404352170339e-05, + "loss": 0.0347, + "step": 22160 + }, + { + "epoch": 4.304062909567497, + "grad_norm": 0.2007305920124054, + "learning_rate": 9.215515426026007e-05, + "loss": 0.0355, + "step": 22170 + }, + { + "epoch": 4.306004562885297, + "grad_norm": 0.18455997109413147, + "learning_rate": 9.214626038879246e-05, + "loss": 0.0357, + "step": 22180 + }, + { + "epoch": 4.307946216203097, + "grad_norm": 0.27010953426361084, + "learning_rate": 9.21373619082732e-05, + "loss": 0.0345, + "step": 22190 + }, + { + "epoch": 4.309887869520897, + "grad_norm": 0.21631869673728943, + "learning_rate": 9.212845881967535e-05, + "loss": 0.0376, + "step": 22200 + }, + { + "epoch": 4.311829522838697, + "grad_norm": 0.21535992622375488, + "learning_rate": 9.211955112397262e-05, + "loss": 0.0462, + "step": 22210 + }, + { + "epoch": 4.313771176156497, + "grad_norm": 0.20166106522083282, + "learning_rate": 9.211063882213909e-05, + "loss": 0.0356, + "step": 22220 + }, + { + "epoch": 4.315712829474298, + "grad_norm": 0.2696080803871155, + "learning_rate": 9.210172191514942e-05, + "loss": 0.0391, + "step": 22230 + }, + { + "epoch": 4.317654482792097, + "grad_norm": 0.16294534504413605, + "learning_rate": 9.209280040397874e-05, + "loss": 0.0329, + "step": 22240 + }, + { + "epoch": 4.319596136109897, + "grad_norm": 0.15837174654006958, + "learning_rate": 9.208387428960268e-05, + "loss": 0.0289, + "step": 22250 + }, + { + "epoch": 4.321537789427698, + "grad_norm": 0.14383989572525024, + "learning_rate": 9.20749435729974e-05, + "loss": 0.0363, + "step": 22260 + }, + { + "epoch": 4.323479442745498, + "grad_norm": 0.17511965334415436, + "learning_rate": 9.206600825513957e-05, + "loss": 0.0324, + "step": 22270 + }, + { + "epoch": 4.325421096063298, + "grad_norm": 0.2866928279399872, + "learning_rate": 9.20570683370063e-05, + "loss": 0.0386, + "step": 22280 + }, + { + "epoch": 4.327362749381098, + "grad_norm": 0.34527498483657837, + "learning_rate": 9.204812381957528e-05, + "loss": 0.0486, + "step": 22290 + }, + { + "epoch": 4.329304402698898, + "grad_norm": 0.15475137531757355, + "learning_rate": 9.203917470382465e-05, + "loss": 0.0325, + "step": 22300 + }, + { + "epoch": 4.331246056016698, + "grad_norm": 0.22418610751628876, + "learning_rate": 9.203022099073309e-05, + "loss": 0.0341, + "step": 22310 + }, + { + "epoch": 4.333187709334498, + "grad_norm": 0.25646644830703735, + "learning_rate": 9.202126268127976e-05, + "loss": 0.0352, + "step": 22320 + }, + { + "epoch": 4.335129362652299, + "grad_norm": 0.22314627468585968, + "learning_rate": 9.20122997764443e-05, + "loss": 0.0397, + "step": 22330 + }, + { + "epoch": 4.337071015970099, + "grad_norm": 0.23273858428001404, + "learning_rate": 9.200333227720692e-05, + "loss": 0.0361, + "step": 22340 + }, + { + "epoch": 4.339012669287898, + "grad_norm": 0.20805805921554565, + "learning_rate": 9.199436018454826e-05, + "loss": 0.0398, + "step": 22350 + }, + { + "epoch": 4.340954322605699, + "grad_norm": 0.14469514787197113, + "learning_rate": 9.198538349944952e-05, + "loss": 0.0373, + "step": 22360 + }, + { + "epoch": 4.342895975923499, + "grad_norm": 0.1500426083803177, + "learning_rate": 9.197640222289234e-05, + "loss": 0.0291, + "step": 22370 + }, + { + "epoch": 4.344837629241299, + "grad_norm": 0.16147689521312714, + "learning_rate": 9.196741635585895e-05, + "loss": 0.0379, + "step": 22380 + }, + { + "epoch": 4.346779282559099, + "grad_norm": 0.14968623220920563, + "learning_rate": 9.195842589933199e-05, + "loss": 0.035, + "step": 22390 + }, + { + "epoch": 4.348720935876899, + "grad_norm": 0.23327411711215973, + "learning_rate": 9.194943085429466e-05, + "loss": 0.0446, + "step": 22400 + }, + { + "epoch": 4.350662589194699, + "grad_norm": 0.2292041778564453, + "learning_rate": 9.194043122173065e-05, + "loss": 0.0371, + "step": 22410 + }, + { + "epoch": 4.352604242512499, + "grad_norm": 0.17322003841400146, + "learning_rate": 9.193142700262413e-05, + "loss": 0.0389, + "step": 22420 + }, + { + "epoch": 4.3545458958303, + "grad_norm": 0.17389263212680817, + "learning_rate": 9.192241819795979e-05, + "loss": 0.0311, + "step": 22430 + }, + { + "epoch": 4.3564875491481, + "grad_norm": 0.22571724653244019, + "learning_rate": 9.191340480872284e-05, + "loss": 0.0364, + "step": 22440 + }, + { + "epoch": 4.358429202465899, + "grad_norm": 0.16306279599666595, + "learning_rate": 9.190438683589895e-05, + "loss": 0.0288, + "step": 22450 + }, + { + "epoch": 4.3603708557836995, + "grad_norm": 0.2499823421239853, + "learning_rate": 9.189536428047432e-05, + "loss": 0.0353, + "step": 22460 + }, + { + "epoch": 4.3623125091015, + "grad_norm": 0.2018115222454071, + "learning_rate": 9.188633714343564e-05, + "loss": 0.0349, + "step": 22470 + }, + { + "epoch": 4.3642541624193, + "grad_norm": 0.28920260071754456, + "learning_rate": 9.18773054257701e-05, + "loss": 0.0305, + "step": 22480 + }, + { + "epoch": 4.3661958157371, + "grad_norm": 0.2487276941537857, + "learning_rate": 9.18682691284654e-05, + "loss": 0.0301, + "step": 22490 + }, + { + "epoch": 4.368137469054901, + "grad_norm": 0.1620475947856903, + "learning_rate": 9.185922825250974e-05, + "loss": 0.0366, + "step": 22500 + }, + { + "epoch": 4.3700791223727, + "grad_norm": 0.23283399641513824, + "learning_rate": 9.185018279889181e-05, + "loss": 0.0326, + "step": 22510 + }, + { + "epoch": 4.3720207756905, + "grad_norm": 0.3312838077545166, + "learning_rate": 9.184113276860082e-05, + "loss": 0.0319, + "step": 22520 + }, + { + "epoch": 4.373962429008301, + "grad_norm": 0.17919941246509552, + "learning_rate": 9.183207816262645e-05, + "loss": 0.0316, + "step": 22530 + }, + { + "epoch": 4.375904082326101, + "grad_norm": 0.1853739321231842, + "learning_rate": 9.182301898195891e-05, + "loss": 0.0319, + "step": 22540 + }, + { + "epoch": 4.377845735643901, + "grad_norm": 0.23701339960098267, + "learning_rate": 9.181395522758889e-05, + "loss": 0.0418, + "step": 22550 + }, + { + "epoch": 4.3797873889617005, + "grad_norm": 0.20836982131004333, + "learning_rate": 9.180488690050759e-05, + "loss": 0.0339, + "step": 22560 + }, + { + "epoch": 4.381729042279501, + "grad_norm": 0.1870199292898178, + "learning_rate": 9.179581400170671e-05, + "loss": 0.0345, + "step": 22570 + }, + { + "epoch": 4.383670695597301, + "grad_norm": 0.24109356105327606, + "learning_rate": 9.178673653217845e-05, + "loss": 0.0322, + "step": 22580 + }, + { + "epoch": 4.385612348915101, + "grad_norm": 0.2256878912448883, + "learning_rate": 9.177765449291551e-05, + "loss": 0.0388, + "step": 22590 + }, + { + "epoch": 4.387554002232902, + "grad_norm": 0.18659397959709167, + "learning_rate": 9.176856788491109e-05, + "loss": 0.0334, + "step": 22600 + }, + { + "epoch": 4.389495655550702, + "grad_norm": 0.18623070418834686, + "learning_rate": 9.175947670915887e-05, + "loss": 0.0305, + "step": 22610 + }, + { + "epoch": 4.391437308868501, + "grad_norm": 0.14781156182289124, + "learning_rate": 9.175038096665309e-05, + "loss": 0.0296, + "step": 22620 + }, + { + "epoch": 4.3933789621863015, + "grad_norm": 0.10584624111652374, + "learning_rate": 9.17412806583884e-05, + "loss": 0.029, + "step": 22630 + }, + { + "epoch": 4.395320615504102, + "grad_norm": 0.1843084841966629, + "learning_rate": 9.173217578536002e-05, + "loss": 0.0301, + "step": 22640 + }, + { + "epoch": 4.397262268821902, + "grad_norm": 0.17726631462574005, + "learning_rate": 9.172306634856362e-05, + "loss": 0.0351, + "step": 22650 + }, + { + "epoch": 4.399203922139702, + "grad_norm": 0.16950233280658722, + "learning_rate": 9.171395234899545e-05, + "loss": 0.0345, + "step": 22660 + }, + { + "epoch": 4.401145575457502, + "grad_norm": 0.2599477767944336, + "learning_rate": 9.170483378765214e-05, + "loss": 0.0327, + "step": 22670 + }, + { + "epoch": 4.403087228775302, + "grad_norm": 0.1384553164243698, + "learning_rate": 9.169571066553091e-05, + "loss": 0.034, + "step": 22680 + }, + { + "epoch": 4.405028882093102, + "grad_norm": 0.19023410975933075, + "learning_rate": 9.168658298362946e-05, + "loss": 0.0345, + "step": 22690 + }, + { + "epoch": 4.406970535410903, + "grad_norm": 0.25846412777900696, + "learning_rate": 9.167745074294598e-05, + "loss": 0.0334, + "step": 22700 + }, + { + "epoch": 4.408912188728703, + "grad_norm": 0.21981628239154816, + "learning_rate": 9.166831394447913e-05, + "loss": 0.0293, + "step": 22710 + }, + { + "epoch": 4.410853842046502, + "grad_norm": 0.17030806839466095, + "learning_rate": 9.165917258922812e-05, + "loss": 0.0272, + "step": 22720 + }, + { + "epoch": 4.4127954953643025, + "grad_norm": 0.1468973010778427, + "learning_rate": 9.165002667819262e-05, + "loss": 0.0333, + "step": 22730 + }, + { + "epoch": 4.414737148682103, + "grad_norm": 0.3478202521800995, + "learning_rate": 9.164087621237282e-05, + "loss": 0.0387, + "step": 22740 + }, + { + "epoch": 4.416678801999903, + "grad_norm": 0.19393876194953918, + "learning_rate": 9.163172119276942e-05, + "loss": 0.0339, + "step": 22750 + }, + { + "epoch": 4.418620455317703, + "grad_norm": 0.25415730476379395, + "learning_rate": 9.162256162038358e-05, + "loss": 0.0365, + "step": 22760 + }, + { + "epoch": 4.420562108635503, + "grad_norm": 0.2343069165945053, + "learning_rate": 9.161339749621698e-05, + "loss": 0.0269, + "step": 22770 + }, + { + "epoch": 4.422503761953303, + "grad_norm": 0.27085205912590027, + "learning_rate": 9.160422882127177e-05, + "loss": 0.0346, + "step": 22780 + }, + { + "epoch": 4.424445415271103, + "grad_norm": 0.1305655986070633, + "learning_rate": 9.159505559655069e-05, + "loss": 0.0315, + "step": 22790 + }, + { + "epoch": 4.4263870685889035, + "grad_norm": 0.16029031574726105, + "learning_rate": 9.158587782305684e-05, + "loss": 0.0276, + "step": 22800 + }, + { + "epoch": 4.428328721906704, + "grad_norm": 0.2822364270687103, + "learning_rate": 9.157669550179391e-05, + "loss": 0.0388, + "step": 22810 + }, + { + "epoch": 4.430270375224504, + "grad_norm": 0.2623833119869232, + "learning_rate": 9.156750863376609e-05, + "loss": 0.0396, + "step": 22820 + }, + { + "epoch": 4.4322120285423035, + "grad_norm": 0.11365760862827301, + "learning_rate": 9.155831721997801e-05, + "loss": 0.036, + "step": 22830 + }, + { + "epoch": 4.434153681860104, + "grad_norm": 0.1709112823009491, + "learning_rate": 9.154912126143484e-05, + "loss": 0.0348, + "step": 22840 + }, + { + "epoch": 4.436095335177904, + "grad_norm": 0.4681706726551056, + "learning_rate": 9.153992075914224e-05, + "loss": 0.0364, + "step": 22850 + }, + { + "epoch": 4.438036988495704, + "grad_norm": 0.1544363796710968, + "learning_rate": 9.153071571410635e-05, + "loss": 0.0304, + "step": 22860 + }, + { + "epoch": 4.4399786418135045, + "grad_norm": 0.12931688129901886, + "learning_rate": 9.152150612733384e-05, + "loss": 0.0339, + "step": 22870 + }, + { + "epoch": 4.441920295131304, + "grad_norm": 0.2020302712917328, + "learning_rate": 9.151229199983184e-05, + "loss": 0.0343, + "step": 22880 + }, + { + "epoch": 4.443861948449104, + "grad_norm": 0.20919308066368103, + "learning_rate": 9.150307333260802e-05, + "loss": 0.0298, + "step": 22890 + }, + { + "epoch": 4.4458036017669045, + "grad_norm": 0.17066824436187744, + "learning_rate": 9.149385012667048e-05, + "loss": 0.0423, + "step": 22900 + }, + { + "epoch": 4.447745255084705, + "grad_norm": 0.17014369368553162, + "learning_rate": 9.148462238302788e-05, + "loss": 0.0367, + "step": 22910 + }, + { + "epoch": 4.449686908402505, + "grad_norm": 0.27988195419311523, + "learning_rate": 9.147539010268936e-05, + "loss": 0.0405, + "step": 22920 + }, + { + "epoch": 4.451628561720305, + "grad_norm": 0.164398655295372, + "learning_rate": 9.14661532866645e-05, + "loss": 0.0418, + "step": 22930 + }, + { + "epoch": 4.453570215038105, + "grad_norm": 0.27179789543151855, + "learning_rate": 9.145691193596348e-05, + "loss": 0.0429, + "step": 22940 + }, + { + "epoch": 4.455511868355905, + "grad_norm": 0.20141848921775818, + "learning_rate": 9.144766605159691e-05, + "loss": 0.0335, + "step": 22950 + }, + { + "epoch": 4.457453521673705, + "grad_norm": 0.2532428503036499, + "learning_rate": 9.14384156345759e-05, + "loss": 0.0286, + "step": 22960 + }, + { + "epoch": 4.4593951749915055, + "grad_norm": 0.20437951385974884, + "learning_rate": 9.142916068591204e-05, + "loss": 0.0335, + "step": 22970 + }, + { + "epoch": 4.461336828309306, + "grad_norm": 0.24142539501190186, + "learning_rate": 9.141990120661746e-05, + "loss": 0.0256, + "step": 22980 + }, + { + "epoch": 4.463278481627105, + "grad_norm": 0.35170984268188477, + "learning_rate": 9.141063719770475e-05, + "loss": 0.0278, + "step": 22990 + }, + { + "epoch": 4.465220134944905, + "grad_norm": 0.13200423121452332, + "learning_rate": 9.140136866018704e-05, + "loss": 0.0415, + "step": 23000 + }, + { + "epoch": 4.467161788262706, + "grad_norm": 0.1796845942735672, + "learning_rate": 9.139209559507788e-05, + "loss": 0.0312, + "step": 23010 + }, + { + "epoch": 4.469103441580506, + "grad_norm": 0.1647598296403885, + "learning_rate": 9.13828180033914e-05, + "loss": 0.0346, + "step": 23020 + }, + { + "epoch": 4.471045094898306, + "grad_norm": 0.23070712387561798, + "learning_rate": 9.137353588614212e-05, + "loss": 0.0311, + "step": 23030 + }, + { + "epoch": 4.472986748216106, + "grad_norm": 0.24564793705940247, + "learning_rate": 9.136424924434519e-05, + "loss": 0.0395, + "step": 23040 + }, + { + "epoch": 4.474928401533906, + "grad_norm": 0.14742682874202728, + "learning_rate": 9.135495807901615e-05, + "loss": 0.0319, + "step": 23050 + }, + { + "epoch": 4.476870054851706, + "grad_norm": 0.14269067347049713, + "learning_rate": 9.134566239117108e-05, + "loss": 0.0359, + "step": 23060 + }, + { + "epoch": 4.4788117081695065, + "grad_norm": 0.22296833992004395, + "learning_rate": 9.13363621818265e-05, + "loss": 0.037, + "step": 23070 + }, + { + "epoch": 4.480753361487307, + "grad_norm": 0.18754690885543823, + "learning_rate": 9.132705745199953e-05, + "loss": 0.0322, + "step": 23080 + }, + { + "epoch": 4.482695014805106, + "grad_norm": 0.14966358244419098, + "learning_rate": 9.131774820270768e-05, + "loss": 0.0338, + "step": 23090 + }, + { + "epoch": 4.484636668122906, + "grad_norm": 0.2346290946006775, + "learning_rate": 9.130843443496901e-05, + "loss": 0.039, + "step": 23100 + }, + { + "epoch": 4.486578321440707, + "grad_norm": 0.30405497550964355, + "learning_rate": 9.129911614980206e-05, + "loss": 0.0411, + "step": 23110 + }, + { + "epoch": 4.488519974758507, + "grad_norm": 0.19003801047801971, + "learning_rate": 9.128979334822584e-05, + "loss": 0.0353, + "step": 23120 + }, + { + "epoch": 4.490461628076307, + "grad_norm": 0.15477821230888367, + "learning_rate": 9.128046603125992e-05, + "loss": 0.0338, + "step": 23130 + }, + { + "epoch": 4.4924032813941075, + "grad_norm": 0.15577073395252228, + "learning_rate": 9.12711341999243e-05, + "loss": 0.0322, + "step": 23140 + }, + { + "epoch": 4.494344934711907, + "grad_norm": 0.2638483941555023, + "learning_rate": 9.12617978552395e-05, + "loss": 0.0333, + "step": 23150 + }, + { + "epoch": 4.496286588029707, + "grad_norm": 0.15097342431545258, + "learning_rate": 9.12524569982265e-05, + "loss": 0.0273, + "step": 23160 + }, + { + "epoch": 4.498228241347507, + "grad_norm": 0.2429112195968628, + "learning_rate": 9.124311162990684e-05, + "loss": 0.0296, + "step": 23170 + }, + { + "epoch": 4.500169894665308, + "grad_norm": 0.18413987755775452, + "learning_rate": 9.12337617513025e-05, + "loss": 0.0338, + "step": 23180 + }, + { + "epoch": 4.502111547983108, + "grad_norm": 0.18675939738750458, + "learning_rate": 9.122440736343596e-05, + "loss": 0.0342, + "step": 23190 + }, + { + "epoch": 4.504053201300907, + "grad_norm": 0.17959512770175934, + "learning_rate": 9.12150484673302e-05, + "loss": 0.0364, + "step": 23200 + }, + { + "epoch": 4.505994854618708, + "grad_norm": 0.16647221148014069, + "learning_rate": 9.120568506400873e-05, + "loss": 0.0315, + "step": 23210 + }, + { + "epoch": 4.507936507936508, + "grad_norm": 0.21314328908920288, + "learning_rate": 9.119631715449548e-05, + "loss": 0.0346, + "step": 23220 + }, + { + "epoch": 4.509878161254308, + "grad_norm": 0.08998262137174606, + "learning_rate": 9.118694473981493e-05, + "loss": 0.027, + "step": 23230 + }, + { + "epoch": 4.5118198145721085, + "grad_norm": 0.1656205654144287, + "learning_rate": 9.117756782099203e-05, + "loss": 0.0355, + "step": 23240 + }, + { + "epoch": 4.513761467889909, + "grad_norm": 0.3535095453262329, + "learning_rate": 9.11681863990522e-05, + "loss": 0.0357, + "step": 23250 + }, + { + "epoch": 4.515703121207708, + "grad_norm": 0.16563041508197784, + "learning_rate": 9.115880047502142e-05, + "loss": 0.0364, + "step": 23260 + }, + { + "epoch": 4.517644774525508, + "grad_norm": 0.16016636788845062, + "learning_rate": 9.114941004992609e-05, + "loss": 0.0436, + "step": 23270 + }, + { + "epoch": 4.519586427843309, + "grad_norm": 0.19431839883327484, + "learning_rate": 9.114001512479317e-05, + "loss": 0.0313, + "step": 23280 + }, + { + "epoch": 4.521528081161109, + "grad_norm": 0.3825242519378662, + "learning_rate": 9.113061570065003e-05, + "loss": 0.0355, + "step": 23290 + }, + { + "epoch": 4.523469734478908, + "grad_norm": 0.5166106820106506, + "learning_rate": 9.112121177852459e-05, + "loss": 0.0343, + "step": 23300 + }, + { + "epoch": 4.525411387796709, + "grad_norm": 0.20302598178386688, + "learning_rate": 9.111180335944527e-05, + "loss": 0.0365, + "step": 23310 + }, + { + "epoch": 4.527353041114509, + "grad_norm": 0.1875050812959671, + "learning_rate": 9.110239044444093e-05, + "loss": 0.0349, + "step": 23320 + }, + { + "epoch": 4.529294694432309, + "grad_norm": 0.17307297885417938, + "learning_rate": 9.109297303454099e-05, + "loss": 0.0333, + "step": 23330 + }, + { + "epoch": 4.531236347750109, + "grad_norm": 0.184527188539505, + "learning_rate": 9.108355113077526e-05, + "loss": 0.0367, + "step": 23340 + }, + { + "epoch": 4.53317800106791, + "grad_norm": 0.19287656247615814, + "learning_rate": 9.107412473417419e-05, + "loss": 0.0286, + "step": 23350 + }, + { + "epoch": 4.535119654385709, + "grad_norm": 0.1406412124633789, + "learning_rate": 9.106469384576858e-05, + "loss": 0.0311, + "step": 23360 + }, + { + "epoch": 4.537061307703509, + "grad_norm": 0.1658353954553604, + "learning_rate": 9.105525846658978e-05, + "loss": 0.0305, + "step": 23370 + }, + { + "epoch": 4.53900296102131, + "grad_norm": 0.15543825924396515, + "learning_rate": 9.104581859766965e-05, + "loss": 0.0263, + "step": 23380 + }, + { + "epoch": 4.54094461433911, + "grad_norm": 0.20504607260227203, + "learning_rate": 9.10363742400405e-05, + "loss": 0.0377, + "step": 23390 + }, + { + "epoch": 4.54288626765691, + "grad_norm": 0.11488933116197586, + "learning_rate": 9.102692539473518e-05, + "loss": 0.0294, + "step": 23400 + }, + { + "epoch": 4.54482792097471, + "grad_norm": 0.19870683550834656, + "learning_rate": 9.101747206278697e-05, + "loss": 0.0476, + "step": 23410 + }, + { + "epoch": 4.54676957429251, + "grad_norm": 0.12435264885425568, + "learning_rate": 9.100801424522968e-05, + "loss": 0.0328, + "step": 23420 + }, + { + "epoch": 4.54871122761031, + "grad_norm": 0.2634901702404022, + "learning_rate": 9.099855194309762e-05, + "loss": 0.0275, + "step": 23430 + }, + { + "epoch": 4.55065288092811, + "grad_norm": 0.30174967646598816, + "learning_rate": 9.098908515742554e-05, + "loss": 0.0359, + "step": 23440 + }, + { + "epoch": 4.552594534245911, + "grad_norm": 0.1599094271659851, + "learning_rate": 9.097961388924873e-05, + "loss": 0.0273, + "step": 23450 + }, + { + "epoch": 4.554536187563711, + "grad_norm": 0.1713283210992813, + "learning_rate": 9.097013813960298e-05, + "loss": 0.0263, + "step": 23460 + }, + { + "epoch": 4.55647784088151, + "grad_norm": 0.27582013607025146, + "learning_rate": 9.09606579095245e-05, + "loss": 0.0287, + "step": 23470 + }, + { + "epoch": 4.558419494199311, + "grad_norm": 0.08826220780611038, + "learning_rate": 9.095117320005008e-05, + "loss": 0.0323, + "step": 23480 + }, + { + "epoch": 4.560361147517111, + "grad_norm": 0.22721585631370544, + "learning_rate": 9.094168401221691e-05, + "loss": 0.0273, + "step": 23490 + }, + { + "epoch": 4.562302800834911, + "grad_norm": 0.20252330601215363, + "learning_rate": 9.093219034706273e-05, + "loss": 0.0287, + "step": 23500 + }, + { + "epoch": 4.564244454152711, + "grad_norm": 0.18608105182647705, + "learning_rate": 9.092269220562577e-05, + "loss": 0.0346, + "step": 23510 + }, + { + "epoch": 4.566186107470511, + "grad_norm": 0.20816288888454437, + "learning_rate": 9.09131895889447e-05, + "loss": 0.0326, + "step": 23520 + }, + { + "epoch": 4.568127760788311, + "grad_norm": 0.15095245838165283, + "learning_rate": 9.090368249805873e-05, + "loss": 0.0292, + "step": 23530 + }, + { + "epoch": 4.570069414106111, + "grad_norm": 0.14020629227161407, + "learning_rate": 9.089417093400754e-05, + "loss": 0.0272, + "step": 23540 + }, + { + "epoch": 4.572011067423912, + "grad_norm": 0.12320244312286377, + "learning_rate": 9.088465489783131e-05, + "loss": 0.0309, + "step": 23550 + }, + { + "epoch": 4.573952720741712, + "grad_norm": 0.1618451178073883, + "learning_rate": 9.087513439057068e-05, + "loss": 0.0314, + "step": 23560 + }, + { + "epoch": 4.575894374059512, + "grad_norm": 0.21096394956111908, + "learning_rate": 9.08656094132668e-05, + "loss": 0.0286, + "step": 23570 + }, + { + "epoch": 4.5778360273773115, + "grad_norm": 0.19844450056552887, + "learning_rate": 9.085607996696134e-05, + "loss": 0.0367, + "step": 23580 + }, + { + "epoch": 4.579777680695112, + "grad_norm": 0.14877325296401978, + "learning_rate": 9.084654605269639e-05, + "loss": 0.0356, + "step": 23590 + }, + { + "epoch": 4.581719334012912, + "grad_norm": 0.1708286702632904, + "learning_rate": 9.083700767151457e-05, + "loss": 0.0289, + "step": 23600 + }, + { + "epoch": 4.583660987330712, + "grad_norm": 0.20467470586299896, + "learning_rate": 9.082746482445898e-05, + "loss": 0.0338, + "step": 23610 + }, + { + "epoch": 4.585602640648512, + "grad_norm": 0.15381403267383575, + "learning_rate": 9.081791751257325e-05, + "loss": 0.0295, + "step": 23620 + }, + { + "epoch": 4.587544293966312, + "grad_norm": 0.2169843167066574, + "learning_rate": 9.080836573690142e-05, + "loss": 0.0366, + "step": 23630 + }, + { + "epoch": 4.589485947284112, + "grad_norm": 0.14095233380794525, + "learning_rate": 9.079880949848805e-05, + "loss": 0.0401, + "step": 23640 + }, + { + "epoch": 4.591427600601913, + "grad_norm": 0.24937449395656586, + "learning_rate": 9.078924879837822e-05, + "loss": 0.0337, + "step": 23650 + }, + { + "epoch": 4.593369253919713, + "grad_norm": 0.12704022228717804, + "learning_rate": 9.077968363761747e-05, + "loss": 0.0309, + "step": 23660 + }, + { + "epoch": 4.595310907237513, + "grad_norm": 0.19923165440559387, + "learning_rate": 9.077011401725182e-05, + "loss": 0.0252, + "step": 23670 + }, + { + "epoch": 4.5972525605553125, + "grad_norm": 0.1922103464603424, + "learning_rate": 9.07605399383278e-05, + "loss": 0.0361, + "step": 23680 + }, + { + "epoch": 4.599194213873113, + "grad_norm": 0.2100202441215515, + "learning_rate": 9.075096140189243e-05, + "loss": 0.0389, + "step": 23690 + }, + { + "epoch": 4.601135867190913, + "grad_norm": 0.2081795185804367, + "learning_rate": 9.074137840899318e-05, + "loss": 0.0303, + "step": 23700 + }, + { + "epoch": 4.603077520508713, + "grad_norm": 0.22428655624389648, + "learning_rate": 9.073179096067804e-05, + "loss": 0.0347, + "step": 23710 + }, + { + "epoch": 4.605019173826514, + "grad_norm": 0.1681276112794876, + "learning_rate": 9.072219905799549e-05, + "loss": 0.0359, + "step": 23720 + }, + { + "epoch": 4.606960827144313, + "grad_norm": 0.18905627727508545, + "learning_rate": 9.071260270199447e-05, + "loss": 0.0342, + "step": 23730 + }, + { + "epoch": 4.608902480462113, + "grad_norm": 0.12882326543331146, + "learning_rate": 9.070300189372441e-05, + "loss": 0.0424, + "step": 23740 + }, + { + "epoch": 4.6108441337799135, + "grad_norm": 0.1539352834224701, + "learning_rate": 9.069339663423528e-05, + "loss": 0.0353, + "step": 23750 + }, + { + "epoch": 4.612785787097714, + "grad_norm": 0.18313217163085938, + "learning_rate": 9.068378692457747e-05, + "loss": 0.0333, + "step": 23760 + }, + { + "epoch": 4.614727440415514, + "grad_norm": 0.20014868676662445, + "learning_rate": 9.067417276580189e-05, + "loss": 0.0324, + "step": 23770 + }, + { + "epoch": 4.616669093733314, + "grad_norm": 0.370993971824646, + "learning_rate": 9.066455415895993e-05, + "loss": 0.0387, + "step": 23780 + }, + { + "epoch": 4.618610747051114, + "grad_norm": 0.2803526818752289, + "learning_rate": 9.065493110510346e-05, + "loss": 0.0459, + "step": 23790 + }, + { + "epoch": 4.620552400368914, + "grad_norm": 0.21059054136276245, + "learning_rate": 9.064530360528484e-05, + "loss": 0.0313, + "step": 23800 + }, + { + "epoch": 4.622494053686714, + "grad_norm": 0.1795136034488678, + "learning_rate": 9.063567166055695e-05, + "loss": 0.0407, + "step": 23810 + }, + { + "epoch": 4.624435707004515, + "grad_norm": 0.17856833338737488, + "learning_rate": 9.062603527197308e-05, + "loss": 0.036, + "step": 23820 + }, + { + "epoch": 4.626377360322315, + "grad_norm": 0.1794009804725647, + "learning_rate": 9.06163944405871e-05, + "loss": 0.0288, + "step": 23830 + }, + { + "epoch": 4.628319013640114, + "grad_norm": 0.21413175761699677, + "learning_rate": 9.060674916745327e-05, + "loss": 0.0298, + "step": 23840 + }, + { + "epoch": 4.6302606669579145, + "grad_norm": 0.2872111201286316, + "learning_rate": 9.05970994536264e-05, + "loss": 0.0327, + "step": 23850 + }, + { + "epoch": 4.632202320275715, + "grad_norm": 0.15861846506595612, + "learning_rate": 9.05874453001618e-05, + "loss": 0.0288, + "step": 23860 + }, + { + "epoch": 4.634143973593515, + "grad_norm": 0.2400345653295517, + "learning_rate": 9.057778670811517e-05, + "loss": 0.0395, + "step": 23870 + }, + { + "epoch": 4.636085626911315, + "grad_norm": 0.12879236042499542, + "learning_rate": 9.056812367854281e-05, + "loss": 0.0327, + "step": 23880 + }, + { + "epoch": 4.638027280229116, + "grad_norm": 0.19480644166469574, + "learning_rate": 9.055845621250143e-05, + "loss": 0.0282, + "step": 23890 + }, + { + "epoch": 4.639968933546915, + "grad_norm": 0.1973835974931717, + "learning_rate": 9.054878431104825e-05, + "loss": 0.034, + "step": 23900 + }, + { + "epoch": 4.641910586864715, + "grad_norm": 0.2276168316602707, + "learning_rate": 9.0539107975241e-05, + "loss": 0.029, + "step": 23910 + }, + { + "epoch": 4.6438522401825155, + "grad_norm": 0.14473827183246613, + "learning_rate": 9.052942720613784e-05, + "loss": 0.0336, + "step": 23920 + }, + { + "epoch": 4.645793893500316, + "grad_norm": 0.2692374289035797, + "learning_rate": 9.051974200479745e-05, + "loss": 0.0384, + "step": 23930 + }, + { + "epoch": 4.647735546818115, + "grad_norm": 0.22406597435474396, + "learning_rate": 9.051005237227901e-05, + "loss": 0.0306, + "step": 23940 + }, + { + "epoch": 4.6496772001359155, + "grad_norm": 0.23283113539218903, + "learning_rate": 9.050035830964215e-05, + "loss": 0.0348, + "step": 23950 + }, + { + "epoch": 4.651618853453716, + "grad_norm": 0.1911323368549347, + "learning_rate": 9.049065981794698e-05, + "loss": 0.0306, + "step": 23960 + }, + { + "epoch": 4.653560506771516, + "grad_norm": 0.17454896867275238, + "learning_rate": 9.048095689825414e-05, + "loss": 0.0304, + "step": 23970 + }, + { + "epoch": 4.655502160089316, + "grad_norm": 0.1777789145708084, + "learning_rate": 9.047124955162472e-05, + "loss": 0.0277, + "step": 23980 + }, + { + "epoch": 4.6574438134071166, + "grad_norm": 0.16620871424674988, + "learning_rate": 9.046153777912028e-05, + "loss": 0.0309, + "step": 23990 + }, + { + "epoch": 4.659385466724916, + "grad_norm": 0.19541971385478973, + "learning_rate": 9.045182158180292e-05, + "loss": 0.0292, + "step": 24000 + }, + { + "epoch": 4.661327120042716, + "grad_norm": 0.1567687690258026, + "learning_rate": 9.044210096073516e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 4.6632687733605165, + "grad_norm": 0.14008843898773193, + "learning_rate": 9.043237591698004e-05, + "loss": 0.0317, + "step": 24020 + }, + { + "epoch": 4.665210426678317, + "grad_norm": 0.19253216683864594, + "learning_rate": 9.04226464516011e-05, + "loss": 0.0326, + "step": 24030 + }, + { + "epoch": 4.667152079996117, + "grad_norm": 0.23184041678905487, + "learning_rate": 9.041291256566229e-05, + "loss": 0.0406, + "step": 24040 + }, + { + "epoch": 4.669093733313916, + "grad_norm": 0.20580285787582397, + "learning_rate": 9.040317426022814e-05, + "loss": 0.0291, + "step": 24050 + }, + { + "epoch": 4.671035386631717, + "grad_norm": 0.25593024492263794, + "learning_rate": 9.03934315363636e-05, + "loss": 0.0336, + "step": 24060 + }, + { + "epoch": 4.672977039949517, + "grad_norm": 0.22761030495166779, + "learning_rate": 9.038368439513409e-05, + "loss": 0.029, + "step": 24070 + }, + { + "epoch": 4.674918693267317, + "grad_norm": 0.23578622937202454, + "learning_rate": 9.03739328376056e-05, + "loss": 0.031, + "step": 24080 + }, + { + "epoch": 4.6768603465851175, + "grad_norm": 0.19530242681503296, + "learning_rate": 9.036417686484451e-05, + "loss": 0.0311, + "step": 24090 + }, + { + "epoch": 4.678801999902918, + "grad_norm": 0.276134729385376, + "learning_rate": 9.035441647791773e-05, + "loss": 0.0358, + "step": 24100 + }, + { + "epoch": 4.680743653220717, + "grad_norm": 0.12988172471523285, + "learning_rate": 9.034465167789263e-05, + "loss": 0.0318, + "step": 24110 + }, + { + "epoch": 4.6826853065385174, + "grad_norm": 0.17949961125850677, + "learning_rate": 9.033488246583706e-05, + "loss": 0.0356, + "step": 24120 + }, + { + "epoch": 4.684626959856318, + "grad_norm": 0.15797516703605652, + "learning_rate": 9.032510884281941e-05, + "loss": 0.0446, + "step": 24130 + }, + { + "epoch": 4.686568613174118, + "grad_norm": 0.12246522307395935, + "learning_rate": 9.031533080990848e-05, + "loss": 0.028, + "step": 24140 + }, + { + "epoch": 4.688510266491917, + "grad_norm": 0.1932101994752884, + "learning_rate": 9.030554836817358e-05, + "loss": 0.0348, + "step": 24150 + }, + { + "epoch": 4.690451919809718, + "grad_norm": 0.16160845756530762, + "learning_rate": 9.029576151868451e-05, + "loss": 0.0319, + "step": 24160 + }, + { + "epoch": 4.692393573127518, + "grad_norm": 0.2695138156414032, + "learning_rate": 9.028597026251155e-05, + "loss": 0.0307, + "step": 24170 + }, + { + "epoch": 4.694335226445318, + "grad_norm": 0.11630969494581223, + "learning_rate": 9.027617460072547e-05, + "loss": 0.0299, + "step": 24180 + }, + { + "epoch": 4.6962768797631185, + "grad_norm": 0.2514195442199707, + "learning_rate": 9.026637453439745e-05, + "loss": 0.0315, + "step": 24190 + }, + { + "epoch": 4.698218533080919, + "grad_norm": 0.22950688004493713, + "learning_rate": 9.025657006459927e-05, + "loss": 0.0318, + "step": 24200 + }, + { + "epoch": 4.700160186398718, + "grad_norm": 0.23051127791404724, + "learning_rate": 9.024676119240311e-05, + "loss": 0.0265, + "step": 24210 + }, + { + "epoch": 4.702101839716518, + "grad_norm": 0.17701232433319092, + "learning_rate": 9.023694791888166e-05, + "loss": 0.0364, + "step": 24220 + }, + { + "epoch": 4.704043493034319, + "grad_norm": 0.3445071578025818, + "learning_rate": 9.022713024510808e-05, + "loss": 0.0332, + "step": 24230 + }, + { + "epoch": 4.705985146352119, + "grad_norm": 0.19380077719688416, + "learning_rate": 9.021730817215601e-05, + "loss": 0.0338, + "step": 24240 + }, + { + "epoch": 4.707926799669919, + "grad_norm": 0.2552260458469391, + "learning_rate": 9.02074817010996e-05, + "loss": 0.0323, + "step": 24250 + }, + { + "epoch": 4.709868452987719, + "grad_norm": 0.1674281358718872, + "learning_rate": 9.019765083301342e-05, + "loss": 0.0334, + "step": 24260 + }, + { + "epoch": 4.711810106305519, + "grad_norm": 0.22341574728488922, + "learning_rate": 9.01878155689726e-05, + "loss": 0.0295, + "step": 24270 + }, + { + "epoch": 4.713751759623319, + "grad_norm": 0.2724861800670624, + "learning_rate": 9.017797591005268e-05, + "loss": 0.0274, + "step": 24280 + }, + { + "epoch": 4.715693412941119, + "grad_norm": 0.1493113785982132, + "learning_rate": 9.016813185732972e-05, + "loss": 0.0319, + "step": 24290 + }, + { + "epoch": 4.71763506625892, + "grad_norm": 0.1611071676015854, + "learning_rate": 9.015828341188027e-05, + "loss": 0.0362, + "step": 24300 + }, + { + "epoch": 4.71957671957672, + "grad_norm": 0.283888041973114, + "learning_rate": 9.01484305747813e-05, + "loss": 0.0392, + "step": 24310 + }, + { + "epoch": 4.721518372894519, + "grad_norm": 0.2958276569843292, + "learning_rate": 9.013857334711033e-05, + "loss": 0.0329, + "step": 24320 + }, + { + "epoch": 4.72346002621232, + "grad_norm": 0.19193026423454285, + "learning_rate": 9.012871172994534e-05, + "loss": 0.0319, + "step": 24330 + }, + { + "epoch": 4.72540167953012, + "grad_norm": 0.18292628228664398, + "learning_rate": 9.011884572436476e-05, + "loss": 0.0325, + "step": 24340 + }, + { + "epoch": 4.72734333284792, + "grad_norm": 0.20194725692272186, + "learning_rate": 9.010897533144754e-05, + "loss": 0.0384, + "step": 24350 + }, + { + "epoch": 4.7292849861657205, + "grad_norm": 0.1923699826002121, + "learning_rate": 9.009910055227306e-05, + "loss": 0.0321, + "step": 24360 + }, + { + "epoch": 4.73122663948352, + "grad_norm": 0.19716522097587585, + "learning_rate": 9.008922138792124e-05, + "loss": 0.0355, + "step": 24370 + }, + { + "epoch": 4.73316829280132, + "grad_norm": 0.16021427512168884, + "learning_rate": 9.007933783947244e-05, + "loss": 0.0357, + "step": 24380 + }, + { + "epoch": 4.73510994611912, + "grad_norm": 0.13476812839508057, + "learning_rate": 9.006944990800752e-05, + "loss": 0.0289, + "step": 24390 + }, + { + "epoch": 4.737051599436921, + "grad_norm": 0.20078805088996887, + "learning_rate": 9.005955759460779e-05, + "loss": 0.0295, + "step": 24400 + }, + { + "epoch": 4.738993252754721, + "grad_norm": 0.06484012305736542, + "learning_rate": 9.004966090035508e-05, + "loss": 0.0292, + "step": 24410 + }, + { + "epoch": 4.740934906072521, + "grad_norm": 0.18522650003433228, + "learning_rate": 9.003975982633166e-05, + "loss": 0.0303, + "step": 24420 + }, + { + "epoch": 4.742876559390321, + "grad_norm": 0.2725616693496704, + "learning_rate": 9.00298543736203e-05, + "loss": 0.0355, + "step": 24430 + }, + { + "epoch": 4.744818212708121, + "grad_norm": 0.2048788219690323, + "learning_rate": 9.001994454330427e-05, + "loss": 0.0336, + "step": 24440 + }, + { + "epoch": 4.746759866025921, + "grad_norm": 0.28007084131240845, + "learning_rate": 9.001003033646727e-05, + "loss": 0.0363, + "step": 24450 + }, + { + "epoch": 4.748701519343721, + "grad_norm": 0.13720360398292542, + "learning_rate": 9.00001117541935e-05, + "loss": 0.0368, + "step": 24460 + }, + { + "epoch": 4.750643172661521, + "grad_norm": 0.21086367964744568, + "learning_rate": 8.999018879756764e-05, + "loss": 0.0327, + "step": 24470 + }, + { + "epoch": 4.752584825979321, + "grad_norm": 0.29310664534568787, + "learning_rate": 8.998026146767487e-05, + "loss": 0.0328, + "step": 24480 + }, + { + "epoch": 4.754526479297121, + "grad_norm": 0.13775870203971863, + "learning_rate": 8.99703297656008e-05, + "loss": 0.0332, + "step": 24490 + }, + { + "epoch": 4.756468132614922, + "grad_norm": 0.2056814283132553, + "learning_rate": 8.996039369243156e-05, + "loss": 0.03, + "step": 24500 + }, + { + "epoch": 4.758409785932722, + "grad_norm": 0.206271693110466, + "learning_rate": 8.995045324925378e-05, + "loss": 0.0361, + "step": 24510 + }, + { + "epoch": 4.760351439250522, + "grad_norm": 0.12944962084293365, + "learning_rate": 8.994050843715448e-05, + "loss": 0.0284, + "step": 24520 + }, + { + "epoch": 4.762293092568322, + "grad_norm": 0.12701748311519623, + "learning_rate": 8.993055925722121e-05, + "loss": 0.0313, + "step": 24530 + }, + { + "epoch": 4.764234745886122, + "grad_norm": 0.18884629011154175, + "learning_rate": 8.992060571054202e-05, + "loss": 0.0314, + "step": 24540 + }, + { + "epoch": 4.766176399203922, + "grad_norm": 0.152491495013237, + "learning_rate": 8.991064779820542e-05, + "loss": 0.0302, + "step": 24550 + }, + { + "epoch": 4.768118052521722, + "grad_norm": 0.24310506880283356, + "learning_rate": 8.990068552130036e-05, + "loss": 0.035, + "step": 24560 + }, + { + "epoch": 4.770059705839523, + "grad_norm": 0.20404213666915894, + "learning_rate": 8.989071888091634e-05, + "loss": 0.0334, + "step": 24570 + }, + { + "epoch": 4.772001359157322, + "grad_norm": 0.22653734683990479, + "learning_rate": 8.988074787814329e-05, + "loss": 0.0284, + "step": 24580 + }, + { + "epoch": 4.773943012475122, + "grad_norm": 0.1773272007703781, + "learning_rate": 8.987077251407158e-05, + "loss": 0.0335, + "step": 24590 + }, + { + "epoch": 4.775884665792923, + "grad_norm": 0.2035023719072342, + "learning_rate": 8.986079278979216e-05, + "loss": 0.0345, + "step": 24600 + }, + { + "epoch": 4.777826319110723, + "grad_norm": 0.15773198008537292, + "learning_rate": 8.985080870639635e-05, + "loss": 0.0338, + "step": 24610 + }, + { + "epoch": 4.779767972428523, + "grad_norm": 0.1763548105955124, + "learning_rate": 8.984082026497603e-05, + "loss": 0.0287, + "step": 24620 + }, + { + "epoch": 4.781709625746323, + "grad_norm": 0.17863373458385468, + "learning_rate": 8.98308274666235e-05, + "loss": 0.0321, + "step": 24630 + }, + { + "epoch": 4.783651279064123, + "grad_norm": 0.23783458769321442, + "learning_rate": 8.982083031243155e-05, + "loss": 0.035, + "step": 24640 + }, + { + "epoch": 4.785592932381923, + "grad_norm": 0.15935203433036804, + "learning_rate": 8.98108288034935e-05, + "loss": 0.0288, + "step": 24650 + }, + { + "epoch": 4.787534585699723, + "grad_norm": 0.14331486821174622, + "learning_rate": 8.980082294090305e-05, + "loss": 0.0292, + "step": 24660 + }, + { + "epoch": 4.789476239017524, + "grad_norm": 0.25847098231315613, + "learning_rate": 8.979081272575443e-05, + "loss": 0.0284, + "step": 24670 + }, + { + "epoch": 4.791417892335324, + "grad_norm": 0.2018406093120575, + "learning_rate": 8.978079815914236e-05, + "loss": 0.0315, + "step": 24680 + }, + { + "epoch": 4.793359545653123, + "grad_norm": 0.3235059678554535, + "learning_rate": 8.977077924216202e-05, + "loss": 0.0284, + "step": 24690 + }, + { + "epoch": 4.7953011989709236, + "grad_norm": 0.20306214690208435, + "learning_rate": 8.976075597590905e-05, + "loss": 0.0414, + "step": 24700 + }, + { + "epoch": 4.797242852288724, + "grad_norm": 0.22947587072849274, + "learning_rate": 8.975072836147958e-05, + "loss": 0.0324, + "step": 24710 + }, + { + "epoch": 4.799184505606524, + "grad_norm": 0.1477012038230896, + "learning_rate": 8.974069639997025e-05, + "loss": 0.0297, + "step": 24720 + }, + { + "epoch": 4.801126158924324, + "grad_norm": 0.2763909101486206, + "learning_rate": 8.973066009247808e-05, + "loss": 0.0309, + "step": 24730 + }, + { + "epoch": 4.803067812242125, + "grad_norm": 0.22565725445747375, + "learning_rate": 8.972061944010066e-05, + "loss": 0.0259, + "step": 24740 + }, + { + "epoch": 4.805009465559924, + "grad_norm": 0.09968414157629013, + "learning_rate": 8.971057444393603e-05, + "loss": 0.0288, + "step": 24750 + }, + { + "epoch": 4.806951118877724, + "grad_norm": 0.1768580973148346, + "learning_rate": 8.970052510508268e-05, + "loss": 0.0441, + "step": 24760 + }, + { + "epoch": 4.808892772195525, + "grad_norm": 0.14431679248809814, + "learning_rate": 8.969047142463959e-05, + "loss": 0.0337, + "step": 24770 + }, + { + "epoch": 4.810834425513325, + "grad_norm": 0.2428230494260788, + "learning_rate": 8.968041340370621e-05, + "loss": 0.0376, + "step": 24780 + }, + { + "epoch": 4.812776078831124, + "grad_norm": 0.24011369049549103, + "learning_rate": 8.96703510433825e-05, + "loss": 0.033, + "step": 24790 + }, + { + "epoch": 4.8147177321489245, + "grad_norm": 0.2100374698638916, + "learning_rate": 8.966028434476883e-05, + "loss": 0.0439, + "step": 24800 + }, + { + "epoch": 4.816659385466725, + "grad_norm": 0.3392895460128784, + "learning_rate": 8.96502133089661e-05, + "loss": 0.0394, + "step": 24810 + }, + { + "epoch": 4.818601038784525, + "grad_norm": 0.2570626139640808, + "learning_rate": 8.964013793707564e-05, + "loss": 0.0302, + "step": 24820 + }, + { + "epoch": 4.820542692102325, + "grad_norm": 0.14173723757266998, + "learning_rate": 8.963005823019932e-05, + "loss": 0.0322, + "step": 24830 + }, + { + "epoch": 4.822484345420126, + "grad_norm": 0.2207336723804474, + "learning_rate": 8.961997418943939e-05, + "loss": 0.0367, + "step": 24840 + }, + { + "epoch": 4.824425998737925, + "grad_norm": 0.18441566824913025, + "learning_rate": 8.960988581589865e-05, + "loss": 0.0322, + "step": 24850 + }, + { + "epoch": 4.826367652055725, + "grad_norm": 0.126241534948349, + "learning_rate": 8.959979311068037e-05, + "loss": 0.0256, + "step": 24860 + }, + { + "epoch": 4.8283093053735255, + "grad_norm": 0.34223806858062744, + "learning_rate": 8.958969607488823e-05, + "loss": 0.0413, + "step": 24870 + }, + { + "epoch": 4.830250958691326, + "grad_norm": 0.1696474552154541, + "learning_rate": 8.957959470962647e-05, + "loss": 0.0347, + "step": 24880 + }, + { + "epoch": 4.832192612009126, + "grad_norm": 0.1744728535413742, + "learning_rate": 8.956948901599971e-05, + "loss": 0.0302, + "step": 24890 + }, + { + "epoch": 4.8341342653269255, + "grad_norm": 0.2784549593925476, + "learning_rate": 8.955937899511315e-05, + "loss": 0.029, + "step": 24900 + }, + { + "epoch": 4.836075918644726, + "grad_norm": 0.31290796399116516, + "learning_rate": 8.954926464807238e-05, + "loss": 0.0282, + "step": 24910 + }, + { + "epoch": 4.838017571962526, + "grad_norm": 0.1970120668411255, + "learning_rate": 8.953914597598347e-05, + "loss": 0.0316, + "step": 24920 + }, + { + "epoch": 4.839959225280326, + "grad_norm": 0.25873127579689026, + "learning_rate": 8.952902297995303e-05, + "loss": 0.0388, + "step": 24930 + }, + { + "epoch": 4.841900878598127, + "grad_norm": 0.12852220237255096, + "learning_rate": 8.951889566108804e-05, + "loss": 0.0249, + "step": 24940 + }, + { + "epoch": 4.843842531915927, + "grad_norm": 0.16389261186122894, + "learning_rate": 8.950876402049606e-05, + "loss": 0.0346, + "step": 24950 + }, + { + "epoch": 4.845784185233726, + "grad_norm": 0.2018161118030548, + "learning_rate": 8.949862805928504e-05, + "loss": 0.0304, + "step": 24960 + }, + { + "epoch": 4.8477258385515265, + "grad_norm": 0.15598461031913757, + "learning_rate": 8.948848777856343e-05, + "loss": 0.0314, + "step": 24970 + }, + { + "epoch": 4.849667491869327, + "grad_norm": 0.18946103751659393, + "learning_rate": 8.947834317944017e-05, + "loss": 0.03, + "step": 24980 + }, + { + "epoch": 4.851609145187127, + "grad_norm": 0.14106373488903046, + "learning_rate": 8.946819426302466e-05, + "loss": 0.0256, + "step": 24990 + }, + { + "epoch": 4.853550798504927, + "grad_norm": 0.24162381887435913, + "learning_rate": 8.945804103042676e-05, + "loss": 0.0337, + "step": 25000 + }, + { + "epoch": 4.855492451822727, + "grad_norm": 0.15710586309432983, + "learning_rate": 8.944788348275681e-05, + "loss": 0.0275, + "step": 25010 + }, + { + "epoch": 4.857434105140527, + "grad_norm": 0.2408047467470169, + "learning_rate": 8.943772162112565e-05, + "loss": 0.0261, + "step": 25020 + }, + { + "epoch": 4.859375758458327, + "grad_norm": 0.2051103115081787, + "learning_rate": 8.942755544664454e-05, + "loss": 0.0317, + "step": 25030 + }, + { + "epoch": 4.8613174117761275, + "grad_norm": 0.15815100073814392, + "learning_rate": 8.941738496042525e-05, + "loss": 0.0271, + "step": 25040 + }, + { + "epoch": 4.863259065093928, + "grad_norm": 0.2303641140460968, + "learning_rate": 8.940721016357999e-05, + "loss": 0.0295, + "step": 25050 + }, + { + "epoch": 4.865200718411728, + "grad_norm": 0.16023795306682587, + "learning_rate": 8.939703105722148e-05, + "loss": 0.0314, + "step": 25060 + }, + { + "epoch": 4.8671423717295275, + "grad_norm": 0.2856203317642212, + "learning_rate": 8.93868476424629e-05, + "loss": 0.031, + "step": 25070 + }, + { + "epoch": 4.869084025047328, + "grad_norm": 0.24200944602489471, + "learning_rate": 8.937665992041786e-05, + "loss": 0.0284, + "step": 25080 + }, + { + "epoch": 4.871025678365128, + "grad_norm": 0.220608651638031, + "learning_rate": 8.93664678922005e-05, + "loss": 0.0279, + "step": 25090 + }, + { + "epoch": 4.872967331682928, + "grad_norm": 0.21810601651668549, + "learning_rate": 8.93562715589254e-05, + "loss": 0.0349, + "step": 25100 + }, + { + "epoch": 4.874908985000728, + "grad_norm": 0.10102331638336182, + "learning_rate": 8.934607092170762e-05, + "loss": 0.0328, + "step": 25110 + }, + { + "epoch": 4.876850638318528, + "grad_norm": 0.12485083192586899, + "learning_rate": 8.933586598166266e-05, + "loss": 0.0298, + "step": 25120 + }, + { + "epoch": 4.878792291636328, + "grad_norm": 0.1934269517660141, + "learning_rate": 8.932565673990655e-05, + "loss": 0.0302, + "step": 25130 + }, + { + "epoch": 4.8807339449541285, + "grad_norm": 0.2588981091976166, + "learning_rate": 8.931544319755574e-05, + "loss": 0.0393, + "step": 25140 + }, + { + "epoch": 4.882675598271929, + "grad_norm": 0.15199720859527588, + "learning_rate": 8.930522535572718e-05, + "loss": 0.0351, + "step": 25150 + }, + { + "epoch": 4.884617251589729, + "grad_norm": 0.16384786367416382, + "learning_rate": 8.929500321553826e-05, + "loss": 0.0295, + "step": 25160 + }, + { + "epoch": 4.886558904907528, + "grad_norm": 0.1650274097919464, + "learning_rate": 8.928477677810686e-05, + "loss": 0.0307, + "step": 25170 + }, + { + "epoch": 4.888500558225329, + "grad_norm": 0.2022310048341751, + "learning_rate": 8.927454604455137e-05, + "loss": 0.0291, + "step": 25180 + }, + { + "epoch": 4.890442211543129, + "grad_norm": 0.2208854705095291, + "learning_rate": 8.926431101599053e-05, + "loss": 0.0325, + "step": 25190 + }, + { + "epoch": 4.892383864860929, + "grad_norm": 0.1351090669631958, + "learning_rate": 8.925407169354369e-05, + "loss": 0.0376, + "step": 25200 + }, + { + "epoch": 4.8943255181787295, + "grad_norm": 0.16943886876106262, + "learning_rate": 8.92438280783306e-05, + "loss": 0.0353, + "step": 25210 + }, + { + "epoch": 4.896267171496529, + "grad_norm": 0.23862215876579285, + "learning_rate": 8.923358017147146e-05, + "loss": 0.0276, + "step": 25220 + }, + { + "epoch": 4.898208824814329, + "grad_norm": 0.30109891295433044, + "learning_rate": 8.922332797408697e-05, + "loss": 0.0298, + "step": 25230 + }, + { + "epoch": 4.9001504781321295, + "grad_norm": 0.2399410903453827, + "learning_rate": 8.921307148729831e-05, + "loss": 0.0375, + "step": 25240 + }, + { + "epoch": 4.90209213144993, + "grad_norm": 0.16893379390239716, + "learning_rate": 8.920281071222712e-05, + "loss": 0.0335, + "step": 25250 + }, + { + "epoch": 4.90403378476773, + "grad_norm": 0.18379147350788116, + "learning_rate": 8.919254564999548e-05, + "loss": 0.0327, + "step": 25260 + }, + { + "epoch": 4.90597543808553, + "grad_norm": 0.24652959406375885, + "learning_rate": 8.918227630172598e-05, + "loss": 0.0307, + "step": 25270 + }, + { + "epoch": 4.90791709140333, + "grad_norm": 0.17542216181755066, + "learning_rate": 8.917200266854165e-05, + "loss": 0.03, + "step": 25280 + }, + { + "epoch": 4.90985874472113, + "grad_norm": 0.1602097600698471, + "learning_rate": 8.9161724751566e-05, + "loss": 0.0346, + "step": 25290 + }, + { + "epoch": 4.91180039803893, + "grad_norm": 0.21226409077644348, + "learning_rate": 8.915144255192302e-05, + "loss": 0.031, + "step": 25300 + }, + { + "epoch": 4.9137420513567305, + "grad_norm": 0.28084394335746765, + "learning_rate": 8.914115607073714e-05, + "loss": 0.0266, + "step": 25310 + }, + { + "epoch": 4.91568370467453, + "grad_norm": 0.1879352480173111, + "learning_rate": 8.913086530913327e-05, + "loss": 0.0371, + "step": 25320 + }, + { + "epoch": 4.91762535799233, + "grad_norm": 0.2036208063364029, + "learning_rate": 8.912057026823681e-05, + "loss": 0.031, + "step": 25330 + }, + { + "epoch": 4.91956701131013, + "grad_norm": 0.11136353760957718, + "learning_rate": 8.91102709491736e-05, + "loss": 0.0334, + "step": 25340 + }, + { + "epoch": 4.921508664627931, + "grad_norm": 0.15005041658878326, + "learning_rate": 8.909996735306996e-05, + "loss": 0.0316, + "step": 25350 + }, + { + "epoch": 4.923450317945731, + "grad_norm": 0.12320563942193985, + "learning_rate": 8.908965948105268e-05, + "loss": 0.0283, + "step": 25360 + }, + { + "epoch": 4.925391971263531, + "grad_norm": 0.24833732843399048, + "learning_rate": 8.907934733424901e-05, + "loss": 0.0366, + "step": 25370 + }, + { + "epoch": 4.927333624581331, + "grad_norm": 0.1569041907787323, + "learning_rate": 8.906903091378666e-05, + "loss": 0.0311, + "step": 25380 + }, + { + "epoch": 4.929275277899131, + "grad_norm": 0.2965238690376282, + "learning_rate": 8.905871022079384e-05, + "loss": 0.0428, + "step": 25390 + }, + { + "epoch": 4.931216931216931, + "grad_norm": 0.22858069837093353, + "learning_rate": 8.90483852563992e-05, + "loss": 0.0324, + "step": 25400 + }, + { + "epoch": 4.933158584534731, + "grad_norm": 0.1279359757900238, + "learning_rate": 8.903805602173185e-05, + "loss": 0.033, + "step": 25410 + }, + { + "epoch": 4.935100237852532, + "grad_norm": 0.15190374851226807, + "learning_rate": 8.902772251792137e-05, + "loss": 0.0329, + "step": 25420 + }, + { + "epoch": 4.937041891170331, + "grad_norm": 0.2267152965068817, + "learning_rate": 8.901738474609786e-05, + "loss": 0.0331, + "step": 25430 + }, + { + "epoch": 4.938983544488131, + "grad_norm": 0.13507407903671265, + "learning_rate": 8.900704270739179e-05, + "loss": 0.0278, + "step": 25440 + }, + { + "epoch": 4.940925197805932, + "grad_norm": 0.217319518327713, + "learning_rate": 8.89966964029342e-05, + "loss": 0.0318, + "step": 25450 + }, + { + "epoch": 4.942866851123732, + "grad_norm": 0.20784661173820496, + "learning_rate": 8.898634583385652e-05, + "loss": 0.0299, + "step": 25460 + }, + { + "epoch": 4.944808504441532, + "grad_norm": 0.1906619518995285, + "learning_rate": 8.897599100129065e-05, + "loss": 0.0355, + "step": 25470 + }, + { + "epoch": 4.9467501577593325, + "grad_norm": 0.24104462563991547, + "learning_rate": 8.896563190636903e-05, + "loss": 0.0317, + "step": 25480 + }, + { + "epoch": 4.948691811077132, + "grad_norm": 0.22603900730609894, + "learning_rate": 8.895526855022448e-05, + "loss": 0.0316, + "step": 25490 + }, + { + "epoch": 4.950633464394932, + "grad_norm": 0.2522924244403839, + "learning_rate": 8.894490093399033e-05, + "loss": 0.0333, + "step": 25500 + }, + { + "epoch": 4.952575117712732, + "grad_norm": 0.21110226213932037, + "learning_rate": 8.893452905880035e-05, + "loss": 0.0321, + "step": 25510 + }, + { + "epoch": 4.954516771030533, + "grad_norm": 0.25045064091682434, + "learning_rate": 8.892415292578883e-05, + "loss": 0.0384, + "step": 25520 + }, + { + "epoch": 4.956458424348333, + "grad_norm": 0.1394871026277542, + "learning_rate": 8.891377253609046e-05, + "loss": 0.0293, + "step": 25530 + }, + { + "epoch": 4.958400077666132, + "grad_norm": 0.24090808629989624, + "learning_rate": 8.890338789084043e-05, + "loss": 0.0266, + "step": 25540 + }, + { + "epoch": 4.960341730983933, + "grad_norm": 0.1777510792016983, + "learning_rate": 8.88929989911744e-05, + "loss": 0.0325, + "step": 25550 + }, + { + "epoch": 4.962283384301733, + "grad_norm": 0.269822359085083, + "learning_rate": 8.888260583822847e-05, + "loss": 0.041, + "step": 25560 + }, + { + "epoch": 4.964225037619533, + "grad_norm": 0.2159549593925476, + "learning_rate": 8.887220843313921e-05, + "loss": 0.0305, + "step": 25570 + }, + { + "epoch": 4.966166690937333, + "grad_norm": 0.25916293263435364, + "learning_rate": 8.88618067770437e-05, + "loss": 0.0357, + "step": 25580 + }, + { + "epoch": 4.968108344255134, + "grad_norm": 0.259062796831131, + "learning_rate": 8.885140087107942e-05, + "loss": 0.0378, + "step": 25590 + }, + { + "epoch": 4.970049997572933, + "grad_norm": 0.1364634931087494, + "learning_rate": 8.884099071638436e-05, + "loss": 0.0252, + "step": 25600 + }, + { + "epoch": 4.971991650890733, + "grad_norm": 0.19973276555538177, + "learning_rate": 8.883057631409695e-05, + "loss": 0.0467, + "step": 25610 + }, + { + "epoch": 4.973933304208534, + "grad_norm": 0.1821584552526474, + "learning_rate": 8.882015766535608e-05, + "loss": 0.029, + "step": 25620 + }, + { + "epoch": 4.975874957526334, + "grad_norm": 0.1345699280500412, + "learning_rate": 8.880973477130115e-05, + "loss": 0.0357, + "step": 25630 + }, + { + "epoch": 4.977816610844133, + "grad_norm": 0.2662254869937897, + "learning_rate": 8.879930763307197e-05, + "loss": 0.038, + "step": 25640 + }, + { + "epoch": 4.979758264161934, + "grad_norm": 0.25853514671325684, + "learning_rate": 8.878887625180884e-05, + "loss": 0.0359, + "step": 25650 + }, + { + "epoch": 4.981699917479734, + "grad_norm": 0.14029809832572937, + "learning_rate": 8.877844062865253e-05, + "loss": 0.0288, + "step": 25660 + }, + { + "epoch": 4.983641570797534, + "grad_norm": 0.22229209542274475, + "learning_rate": 8.876800076474424e-05, + "loss": 0.0407, + "step": 25670 + }, + { + "epoch": 4.985583224115334, + "grad_norm": 0.12612031400203705, + "learning_rate": 8.875755666122568e-05, + "loss": 0.0394, + "step": 25680 + }, + { + "epoch": 4.987524877433135, + "grad_norm": 0.08774629980325699, + "learning_rate": 8.8747108319239e-05, + "loss": 0.0329, + "step": 25690 + }, + { + "epoch": 4.989466530750934, + "grad_norm": 0.16741561889648438, + "learning_rate": 8.87366557399268e-05, + "loss": 0.0315, + "step": 25700 + }, + { + "epoch": 4.991408184068734, + "grad_norm": 0.22175171971321106, + "learning_rate": 8.872619892443217e-05, + "loss": 0.0318, + "step": 25710 + }, + { + "epoch": 4.993349837386535, + "grad_norm": 0.2167983204126358, + "learning_rate": 8.871573787389865e-05, + "loss": 0.0343, + "step": 25720 + }, + { + "epoch": 4.995291490704335, + "grad_norm": 0.1614050716161728, + "learning_rate": 8.870527258947024e-05, + "loss": 0.0338, + "step": 25730 + }, + { + "epoch": 4.997233144022135, + "grad_norm": 0.1891735941171646, + "learning_rate": 8.869480307229143e-05, + "loss": 0.0366, + "step": 25740 + }, + { + "epoch": 4.9991747973399345, + "grad_norm": 0.2173306941986084, + "learning_rate": 8.868432932350712e-05, + "loss": 0.0355, + "step": 25750 + }, + { + "epoch": 5.0009708266589, + "grad_norm": 0.15086375176906586, + "learning_rate": 8.867385134426272e-05, + "loss": 0.028, + "step": 25760 + }, + { + "epoch": 5.0029124799767, + "grad_norm": 0.1587013304233551, + "learning_rate": 8.866336913570407e-05, + "loss": 0.0329, + "step": 25770 + }, + { + "epoch": 5.004854133294501, + "grad_norm": 0.16269530355930328, + "learning_rate": 8.865288269897751e-05, + "loss": 0.0351, + "step": 25780 + }, + { + "epoch": 5.0067957866123, + "grad_norm": 0.27686402201652527, + "learning_rate": 8.864239203522981e-05, + "loss": 0.0353, + "step": 25790 + }, + { + "epoch": 5.0087374399301, + "grad_norm": 0.10287122428417206, + "learning_rate": 8.863189714560822e-05, + "loss": 0.026, + "step": 25800 + }, + { + "epoch": 5.010679093247901, + "grad_norm": 0.1535065919160843, + "learning_rate": 8.862139803126043e-05, + "loss": 0.0293, + "step": 25810 + }, + { + "epoch": 5.012620746565701, + "grad_norm": 0.16634243726730347, + "learning_rate": 8.861089469333463e-05, + "loss": 0.0296, + "step": 25820 + }, + { + "epoch": 5.014562399883501, + "grad_norm": 0.19557330012321472, + "learning_rate": 8.860038713297944e-05, + "loss": 0.0345, + "step": 25830 + }, + { + "epoch": 5.0165040532013006, + "grad_norm": 0.11878351122140884, + "learning_rate": 8.858987535134394e-05, + "loss": 0.0303, + "step": 25840 + }, + { + "epoch": 5.018445706519101, + "grad_norm": 0.16624996066093445, + "learning_rate": 8.857935934957769e-05, + "loss": 0.0348, + "step": 25850 + }, + { + "epoch": 5.020387359836901, + "grad_norm": 0.24149759113788605, + "learning_rate": 8.856883912883071e-05, + "loss": 0.0339, + "step": 25860 + }, + { + "epoch": 5.022329013154701, + "grad_norm": 0.16471457481384277, + "learning_rate": 8.855831469025346e-05, + "loss": 0.0288, + "step": 25870 + }, + { + "epoch": 5.024270666472502, + "grad_norm": 0.16555975377559662, + "learning_rate": 8.854778603499689e-05, + "loss": 0.0279, + "step": 25880 + }, + { + "epoch": 5.026212319790301, + "grad_norm": 0.23665253818035126, + "learning_rate": 8.85372531642124e-05, + "loss": 0.0289, + "step": 25890 + }, + { + "epoch": 5.028153973108101, + "grad_norm": 0.13602110743522644, + "learning_rate": 8.852671607905185e-05, + "loss": 0.0288, + "step": 25900 + }, + { + "epoch": 5.030095626425902, + "grad_norm": 0.12527233362197876, + "learning_rate": 8.851617478066754e-05, + "loss": 0.0287, + "step": 25910 + }, + { + "epoch": 5.032037279743702, + "grad_norm": 0.20299586653709412, + "learning_rate": 8.850562927021227e-05, + "loss": 0.0294, + "step": 25920 + }, + { + "epoch": 5.033978933061502, + "grad_norm": 0.22301241755485535, + "learning_rate": 8.849507954883928e-05, + "loss": 0.0334, + "step": 25930 + }, + { + "epoch": 5.035920586379302, + "grad_norm": 0.22117498517036438, + "learning_rate": 8.848452561770226e-05, + "loss": 0.0343, + "step": 25940 + }, + { + "epoch": 5.037862239697102, + "grad_norm": 0.1707780957221985, + "learning_rate": 8.847396747795538e-05, + "loss": 0.0354, + "step": 25950 + }, + { + "epoch": 5.039803893014902, + "grad_norm": 0.1840467005968094, + "learning_rate": 8.846340513075327e-05, + "loss": 0.0313, + "step": 25960 + }, + { + "epoch": 5.041745546332702, + "grad_norm": 0.20218536257743835, + "learning_rate": 8.845283857725099e-05, + "loss": 0.0305, + "step": 25970 + }, + { + "epoch": 5.043687199650503, + "grad_norm": 0.1987713724374771, + "learning_rate": 8.844226781860409e-05, + "loss": 0.0359, + "step": 25980 + }, + { + "epoch": 5.045628852968303, + "grad_norm": 0.22331704199314117, + "learning_rate": 8.84316928559686e-05, + "loss": 0.0321, + "step": 25990 + }, + { + "epoch": 5.047570506286102, + "grad_norm": 0.16789592802524567, + "learning_rate": 8.842111369050094e-05, + "loss": 0.0314, + "step": 26000 + }, + { + "epoch": 5.0495121596039025, + "grad_norm": 0.20940819382667542, + "learning_rate": 8.841053032335808e-05, + "loss": 0.0395, + "step": 26010 + }, + { + "epoch": 5.051453812921703, + "grad_norm": 0.1460288017988205, + "learning_rate": 8.839994275569735e-05, + "loss": 0.0313, + "step": 26020 + }, + { + "epoch": 5.053395466239503, + "grad_norm": 0.19262340664863586, + "learning_rate": 8.838935098867662e-05, + "loss": 0.0353, + "step": 26030 + }, + { + "epoch": 5.055337119557303, + "grad_norm": 0.22559617459774017, + "learning_rate": 8.837875502345418e-05, + "loss": 0.0314, + "step": 26040 + }, + { + "epoch": 5.057278772875103, + "grad_norm": 0.18959450721740723, + "learning_rate": 8.83681548611888e-05, + "loss": 0.0269, + "step": 26050 + }, + { + "epoch": 5.059220426192903, + "grad_norm": 0.15562519431114197, + "learning_rate": 8.835755050303969e-05, + "loss": 0.0407, + "step": 26060 + }, + { + "epoch": 5.061162079510703, + "grad_norm": 0.13815540075302124, + "learning_rate": 8.834694195016653e-05, + "loss": 0.0343, + "step": 26070 + }, + { + "epoch": 5.063103732828504, + "grad_norm": 0.23897887766361237, + "learning_rate": 8.833632920372942e-05, + "loss": 0.0273, + "step": 26080 + }, + { + "epoch": 5.065045386146304, + "grad_norm": 0.2146979123353958, + "learning_rate": 8.832571226488903e-05, + "loss": 0.0336, + "step": 26090 + }, + { + "epoch": 5.066987039464104, + "grad_norm": 0.15159134566783905, + "learning_rate": 8.831509113480634e-05, + "loss": 0.035, + "step": 26100 + }, + { + "epoch": 5.0689286927819035, + "grad_norm": 0.19149480760097504, + "learning_rate": 8.83044658146429e-05, + "loss": 0.0364, + "step": 26110 + }, + { + "epoch": 5.070870346099704, + "grad_norm": 0.2882765829563141, + "learning_rate": 8.829383630556067e-05, + "loss": 0.034, + "step": 26120 + }, + { + "epoch": 5.072811999417504, + "grad_norm": 0.19150808453559875, + "learning_rate": 8.828320260872207e-05, + "loss": 0.0257, + "step": 26130 + }, + { + "epoch": 5.074753652735304, + "grad_norm": 0.15966564416885376, + "learning_rate": 8.827256472529e-05, + "loss": 0.0292, + "step": 26140 + }, + { + "epoch": 5.076695306053105, + "grad_norm": 0.24928030371665955, + "learning_rate": 8.826192265642778e-05, + "loss": 0.0308, + "step": 26150 + }, + { + "epoch": 5.078636959370904, + "grad_norm": 0.23656527698040009, + "learning_rate": 8.825127640329923e-05, + "loss": 0.0336, + "step": 26160 + }, + { + "epoch": 5.080578612688704, + "grad_norm": 0.14307573437690735, + "learning_rate": 8.824062596706861e-05, + "loss": 0.0338, + "step": 26170 + }, + { + "epoch": 5.0825202660065045, + "grad_norm": 0.16826961934566498, + "learning_rate": 8.822997134890062e-05, + "loss": 0.0351, + "step": 26180 + }, + { + "epoch": 5.084461919324305, + "grad_norm": 0.2478516846895218, + "learning_rate": 8.821931254996044e-05, + "loss": 0.0284, + "step": 26190 + }, + { + "epoch": 5.086403572642105, + "grad_norm": 0.1656726896762848, + "learning_rate": 8.82086495714137e-05, + "loss": 0.0372, + "step": 26200 + }, + { + "epoch": 5.0883452259599045, + "grad_norm": 0.1915220469236374, + "learning_rate": 8.81979824144265e-05, + "loss": 0.0341, + "step": 26210 + }, + { + "epoch": 5.090286879277705, + "grad_norm": 0.23817752301692963, + "learning_rate": 8.818731108016536e-05, + "loss": 0.0331, + "step": 26220 + }, + { + "epoch": 5.092228532595505, + "grad_norm": 0.29346105456352234, + "learning_rate": 8.81766355697973e-05, + "loss": 0.0255, + "step": 26230 + }, + { + "epoch": 5.094170185913305, + "grad_norm": 0.20582696795463562, + "learning_rate": 8.816595588448977e-05, + "loss": 0.0365, + "step": 26240 + }, + { + "epoch": 5.0961118392311056, + "grad_norm": 0.20824037492275238, + "learning_rate": 8.81552720254107e-05, + "loss": 0.0375, + "step": 26250 + }, + { + "epoch": 5.098053492548906, + "grad_norm": 0.1912957727909088, + "learning_rate": 8.814458399372842e-05, + "loss": 0.0359, + "step": 26260 + }, + { + "epoch": 5.099995145866705, + "grad_norm": 0.19093304872512817, + "learning_rate": 8.813389179061181e-05, + "loss": 0.032, + "step": 26270 + }, + { + "epoch": 5.1019367991845055, + "grad_norm": 0.19404445588588715, + "learning_rate": 8.812319541723012e-05, + "loss": 0.0348, + "step": 26280 + }, + { + "epoch": 5.103878452502306, + "grad_norm": 0.22219963371753693, + "learning_rate": 8.811249487475309e-05, + "loss": 0.0284, + "step": 26290 + }, + { + "epoch": 5.105820105820106, + "grad_norm": 0.20051462948322296, + "learning_rate": 8.810179016435092e-05, + "loss": 0.0326, + "step": 26300 + }, + { + "epoch": 5.107761759137906, + "grad_norm": 0.1501709371805191, + "learning_rate": 8.809108128719428e-05, + "loss": 0.0298, + "step": 26310 + }, + { + "epoch": 5.109703412455706, + "grad_norm": 0.23035898804664612, + "learning_rate": 8.808036824445424e-05, + "loss": 0.028, + "step": 26320 + }, + { + "epoch": 5.111645065773506, + "grad_norm": 0.2943940758705139, + "learning_rate": 8.806965103730238e-05, + "loss": 0.0337, + "step": 26330 + }, + { + "epoch": 5.113586719091306, + "grad_norm": 0.17435812950134277, + "learning_rate": 8.805892966691074e-05, + "loss": 0.039, + "step": 26340 + }, + { + "epoch": 5.1155283724091065, + "grad_norm": 0.14115220308303833, + "learning_rate": 8.804820413445175e-05, + "loss": 0.0271, + "step": 26350 + }, + { + "epoch": 5.117470025726907, + "grad_norm": 0.18857493996620178, + "learning_rate": 8.803747444109837e-05, + "loss": 0.0382, + "step": 26360 + }, + { + "epoch": 5.119411679044706, + "grad_norm": 0.1855785846710205, + "learning_rate": 8.802674058802399e-05, + "loss": 0.0269, + "step": 26370 + }, + { + "epoch": 5.1213533323625064, + "grad_norm": 0.20173397660255432, + "learning_rate": 8.801600257640241e-05, + "loss": 0.0263, + "step": 26380 + }, + { + "epoch": 5.123294985680307, + "grad_norm": 0.20088011026382446, + "learning_rate": 8.800526040740795e-05, + "loss": 0.0315, + "step": 26390 + }, + { + "epoch": 5.125236638998107, + "grad_norm": 0.23597368597984314, + "learning_rate": 8.799451408221535e-05, + "loss": 0.0401, + "step": 26400 + }, + { + "epoch": 5.127178292315907, + "grad_norm": 0.09488996863365173, + "learning_rate": 8.798376360199982e-05, + "loss": 0.033, + "step": 26410 + }, + { + "epoch": 5.1291199456337075, + "grad_norm": 0.21200670301914215, + "learning_rate": 8.797300896793701e-05, + "loss": 0.0305, + "step": 26420 + }, + { + "epoch": 5.131061598951507, + "grad_norm": 0.15544304251670837, + "learning_rate": 8.796225018120302e-05, + "loss": 0.031, + "step": 26430 + }, + { + "epoch": 5.133003252269307, + "grad_norm": 0.19761589169502258, + "learning_rate": 8.795148724297444e-05, + "loss": 0.0358, + "step": 26440 + }, + { + "epoch": 5.1349449055871075, + "grad_norm": 0.13803423941135406, + "learning_rate": 8.794072015442825e-05, + "loss": 0.0307, + "step": 26450 + }, + { + "epoch": 5.136886558904908, + "grad_norm": 0.2502194344997406, + "learning_rate": 8.792994891674198e-05, + "loss": 0.037, + "step": 26460 + }, + { + "epoch": 5.138828212222708, + "grad_norm": 0.16274775564670563, + "learning_rate": 8.79191735310935e-05, + "loss": 0.0316, + "step": 26470 + }, + { + "epoch": 5.140769865540507, + "grad_norm": 0.1540401577949524, + "learning_rate": 8.790839399866122e-05, + "loss": 0.0403, + "step": 26480 + }, + { + "epoch": 5.142711518858308, + "grad_norm": 0.17668142914772034, + "learning_rate": 8.789761032062397e-05, + "loss": 0.0339, + "step": 26490 + }, + { + "epoch": 5.144653172176108, + "grad_norm": 0.11875941604375839, + "learning_rate": 8.788682249816103e-05, + "loss": 0.0309, + "step": 26500 + }, + { + "epoch": 5.146594825493908, + "grad_norm": 0.13879211246967316, + "learning_rate": 8.787603053245215e-05, + "loss": 0.0313, + "step": 26510 + }, + { + "epoch": 5.1485364788117085, + "grad_norm": 0.20562851428985596, + "learning_rate": 8.78652344246775e-05, + "loss": 0.0379, + "step": 26520 + }, + { + "epoch": 5.150478132129508, + "grad_norm": 0.14988577365875244, + "learning_rate": 8.785443417601776e-05, + "loss": 0.0297, + "step": 26530 + }, + { + "epoch": 5.152419785447308, + "grad_norm": 0.16265250742435455, + "learning_rate": 8.784362978765401e-05, + "loss": 0.0345, + "step": 26540 + }, + { + "epoch": 5.154361438765108, + "grad_norm": 0.32054561376571655, + "learning_rate": 8.783282126076779e-05, + "loss": 0.0444, + "step": 26550 + }, + { + "epoch": 5.156303092082909, + "grad_norm": 0.10751073807477951, + "learning_rate": 8.782200859654112e-05, + "loss": 0.0355, + "step": 26560 + }, + { + "epoch": 5.158244745400709, + "grad_norm": 0.17656782269477844, + "learning_rate": 8.781119179615646e-05, + "loss": 0.0322, + "step": 26570 + }, + { + "epoch": 5.160186398718508, + "grad_norm": 0.22041180729866028, + "learning_rate": 8.780037086079674e-05, + "loss": 0.0353, + "step": 26580 + }, + { + "epoch": 5.162128052036309, + "grad_norm": 0.18413160741329193, + "learning_rate": 8.778954579164527e-05, + "loss": 0.0349, + "step": 26590 + }, + { + "epoch": 5.164069705354109, + "grad_norm": 0.22808973491191864, + "learning_rate": 8.777871658988588e-05, + "loss": 0.0385, + "step": 26600 + }, + { + "epoch": 5.166011358671909, + "grad_norm": 0.19456616044044495, + "learning_rate": 8.776788325670285e-05, + "loss": 0.0303, + "step": 26610 + }, + { + "epoch": 5.1679530119897095, + "grad_norm": 0.22805656492710114, + "learning_rate": 8.775704579328089e-05, + "loss": 0.0303, + "step": 26620 + }, + { + "epoch": 5.16989466530751, + "grad_norm": 0.18781551718711853, + "learning_rate": 8.774620420080517e-05, + "loss": 0.0276, + "step": 26630 + }, + { + "epoch": 5.171836318625309, + "grad_norm": 0.21478019654750824, + "learning_rate": 8.773535848046131e-05, + "loss": 0.0339, + "step": 26640 + }, + { + "epoch": 5.173777971943109, + "grad_norm": 0.15308958292007446, + "learning_rate": 8.772450863343538e-05, + "loss": 0.0347, + "step": 26650 + }, + { + "epoch": 5.17571962526091, + "grad_norm": 0.23003296554088593, + "learning_rate": 8.77136546609139e-05, + "loss": 0.0416, + "step": 26660 + }, + { + "epoch": 5.17766127857871, + "grad_norm": 0.2536243498325348, + "learning_rate": 8.770279656408385e-05, + "loss": 0.0332, + "step": 26670 + }, + { + "epoch": 5.17960293189651, + "grad_norm": 0.15056757628917694, + "learning_rate": 8.769193434413265e-05, + "loss": 0.0307, + "step": 26680 + }, + { + "epoch": 5.18154458521431, + "grad_norm": 0.15747815370559692, + "learning_rate": 8.76810680022482e-05, + "loss": 0.0316, + "step": 26690 + }, + { + "epoch": 5.18348623853211, + "grad_norm": 0.2232719212770462, + "learning_rate": 8.767019753961878e-05, + "loss": 0.029, + "step": 26700 + }, + { + "epoch": 5.18542789184991, + "grad_norm": 0.26556262373924255, + "learning_rate": 8.765932295743321e-05, + "loss": 0.0351, + "step": 26710 + }, + { + "epoch": 5.18736954516771, + "grad_norm": 0.13855156302452087, + "learning_rate": 8.764844425688068e-05, + "loss": 0.0287, + "step": 26720 + }, + { + "epoch": 5.189311198485511, + "grad_norm": 0.17613685131072998, + "learning_rate": 8.763756143915092e-05, + "loss": 0.0294, + "step": 26730 + }, + { + "epoch": 5.191252851803311, + "grad_norm": 0.12874320149421692, + "learning_rate": 8.7626674505434e-05, + "loss": 0.0264, + "step": 26740 + }, + { + "epoch": 5.19319450512111, + "grad_norm": 0.12397172302007675, + "learning_rate": 8.761578345692053e-05, + "loss": 0.0289, + "step": 26750 + }, + { + "epoch": 5.195136158438911, + "grad_norm": 0.15132319927215576, + "learning_rate": 8.760488829480156e-05, + "loss": 0.0291, + "step": 26760 + }, + { + "epoch": 5.197077811756711, + "grad_norm": 0.274628609418869, + "learning_rate": 8.759398902026854e-05, + "loss": 0.038, + "step": 26770 + }, + { + "epoch": 5.199019465074511, + "grad_norm": 0.15731622278690338, + "learning_rate": 8.758308563451339e-05, + "loss": 0.0299, + "step": 26780 + }, + { + "epoch": 5.2009611183923115, + "grad_norm": 0.17374546825885773, + "learning_rate": 8.75721781387285e-05, + "loss": 0.029, + "step": 26790 + }, + { + "epoch": 5.202902771710111, + "grad_norm": 0.15741132199764252, + "learning_rate": 8.75612665341067e-05, + "loss": 0.0245, + "step": 26800 + }, + { + "epoch": 5.204844425027911, + "grad_norm": 0.22281382977962494, + "learning_rate": 8.755035082184126e-05, + "loss": 0.0285, + "step": 26810 + }, + { + "epoch": 5.206786078345711, + "grad_norm": 0.1706848442554474, + "learning_rate": 8.753943100312592e-05, + "loss": 0.031, + "step": 26820 + }, + { + "epoch": 5.208727731663512, + "grad_norm": 0.18685488402843475, + "learning_rate": 8.752850707915484e-05, + "loss": 0.0315, + "step": 26830 + }, + { + "epoch": 5.210669384981312, + "grad_norm": 0.1533070057630539, + "learning_rate": 8.751757905112264e-05, + "loss": 0.0324, + "step": 26840 + }, + { + "epoch": 5.212611038299111, + "grad_norm": 0.1449110358953476, + "learning_rate": 8.75066469202244e-05, + "loss": 0.035, + "step": 26850 + }, + { + "epoch": 5.214552691616912, + "grad_norm": 0.26306387782096863, + "learning_rate": 8.749571068765567e-05, + "loss": 0.0333, + "step": 26860 + }, + { + "epoch": 5.216494344934712, + "grad_norm": 0.1946239173412323, + "learning_rate": 8.748477035461238e-05, + "loss": 0.0352, + "step": 26870 + }, + { + "epoch": 5.218435998252512, + "grad_norm": 0.20773491263389587, + "learning_rate": 8.747382592229095e-05, + "loss": 0.0304, + "step": 26880 + }, + { + "epoch": 5.220377651570312, + "grad_norm": 0.16133831441402435, + "learning_rate": 8.746287739188828e-05, + "loss": 0.0337, + "step": 26890 + }, + { + "epoch": 5.222319304888112, + "grad_norm": 0.25565361976623535, + "learning_rate": 8.745192476460165e-05, + "loss": 0.0362, + "step": 26900 + }, + { + "epoch": 5.224260958205912, + "grad_norm": 0.12219846993684769, + "learning_rate": 8.744096804162882e-05, + "loss": 0.0297, + "step": 26910 + }, + { + "epoch": 5.226202611523712, + "grad_norm": 0.18813450634479523, + "learning_rate": 8.743000722416804e-05, + "loss": 0.0356, + "step": 26920 + }, + { + "epoch": 5.228144264841513, + "grad_norm": 0.20849540829658508, + "learning_rate": 8.741904231341793e-05, + "loss": 0.0387, + "step": 26930 + }, + { + "epoch": 5.230085918159313, + "grad_norm": 0.24499765038490295, + "learning_rate": 8.740807331057762e-05, + "loss": 0.0321, + "step": 26940 + }, + { + "epoch": 5.232027571477113, + "grad_norm": 0.10482950508594513, + "learning_rate": 8.739710021684667e-05, + "loss": 0.0239, + "step": 26950 + }, + { + "epoch": 5.2339692247949126, + "grad_norm": 0.20482735335826874, + "learning_rate": 8.738612303342503e-05, + "loss": 0.0315, + "step": 26960 + }, + { + "epoch": 5.235910878112713, + "grad_norm": 0.190626859664917, + "learning_rate": 8.73751417615132e-05, + "loss": 0.0277, + "step": 26970 + }, + { + "epoch": 5.237852531430513, + "grad_norm": 0.2292417734861374, + "learning_rate": 8.736415640231208e-05, + "loss": 0.0376, + "step": 26980 + }, + { + "epoch": 5.239794184748313, + "grad_norm": 0.14189551770687103, + "learning_rate": 8.735316695702297e-05, + "loss": 0.0334, + "step": 26990 + }, + { + "epoch": 5.241735838066114, + "grad_norm": 0.23760667443275452, + "learning_rate": 8.734217342684769e-05, + "loss": 0.0341, + "step": 27000 + }, + { + "epoch": 5.243677491383913, + "grad_norm": 0.1247677132487297, + "learning_rate": 8.733117581298847e-05, + "loss": 0.0328, + "step": 27010 + }, + { + "epoch": 5.245619144701713, + "grad_norm": 0.2249530404806137, + "learning_rate": 8.732017411664796e-05, + "loss": 0.0348, + "step": 27020 + }, + { + "epoch": 5.247560798019514, + "grad_norm": 0.15590739250183105, + "learning_rate": 8.730916833902936e-05, + "loss": 0.0366, + "step": 27030 + }, + { + "epoch": 5.249502451337314, + "grad_norm": 0.12521050870418549, + "learning_rate": 8.729815848133618e-05, + "loss": 0.0282, + "step": 27040 + }, + { + "epoch": 5.251444104655114, + "grad_norm": 0.21889451146125793, + "learning_rate": 8.728714454477247e-05, + "loss": 0.0295, + "step": 27050 + }, + { + "epoch": 5.253385757972914, + "grad_norm": 0.14734508097171783, + "learning_rate": 8.727612653054269e-05, + "loss": 0.0321, + "step": 27060 + }, + { + "epoch": 5.255327411290714, + "grad_norm": 0.23519843816757202, + "learning_rate": 8.726510443985176e-05, + "loss": 0.0268, + "step": 27070 + }, + { + "epoch": 5.257269064608514, + "grad_norm": 0.14699915051460266, + "learning_rate": 8.725407827390503e-05, + "loss": 0.0333, + "step": 27080 + }, + { + "epoch": 5.259210717926314, + "grad_norm": 0.17909742891788483, + "learning_rate": 8.724304803390833e-05, + "loss": 0.0361, + "step": 27090 + }, + { + "epoch": 5.261152371244115, + "grad_norm": 0.12517967820167542, + "learning_rate": 8.723201372106788e-05, + "loss": 0.0324, + "step": 27100 + }, + { + "epoch": 5.263094024561915, + "grad_norm": 0.1602180451154709, + "learning_rate": 8.722097533659038e-05, + "loss": 0.0368, + "step": 27110 + }, + { + "epoch": 5.265035677879714, + "grad_norm": 0.17736804485321045, + "learning_rate": 8.720993288168299e-05, + "loss": 0.0379, + "step": 27120 + }, + { + "epoch": 5.2669773311975145, + "grad_norm": 0.1625392884016037, + "learning_rate": 8.719888635755327e-05, + "loss": 0.03, + "step": 27130 + }, + { + "epoch": 5.268918984515315, + "grad_norm": 0.14253011345863342, + "learning_rate": 8.718783576540928e-05, + "loss": 0.0259, + "step": 27140 + }, + { + "epoch": 5.270860637833115, + "grad_norm": 0.22424258291721344, + "learning_rate": 8.717678110645948e-05, + "loss": 0.0385, + "step": 27150 + }, + { + "epoch": 5.272802291150915, + "grad_norm": 0.2669633626937866, + "learning_rate": 8.716572238191279e-05, + "loss": 0.0334, + "step": 27160 + }, + { + "epoch": 5.274743944468715, + "grad_norm": 0.1927465796470642, + "learning_rate": 8.715465959297857e-05, + "loss": 0.0355, + "step": 27170 + }, + { + "epoch": 5.276685597786515, + "grad_norm": 0.1629289984703064, + "learning_rate": 8.714359274086665e-05, + "loss": 0.0259, + "step": 27180 + }, + { + "epoch": 5.278627251104315, + "grad_norm": 0.36754441261291504, + "learning_rate": 8.713252182678726e-05, + "loss": 0.0325, + "step": 27190 + }, + { + "epoch": 5.280568904422116, + "grad_norm": 0.21474367380142212, + "learning_rate": 8.712144685195112e-05, + "loss": 0.0318, + "step": 27200 + }, + { + "epoch": 5.282510557739916, + "grad_norm": 0.208513543009758, + "learning_rate": 8.711036781756936e-05, + "loss": 0.0323, + "step": 27210 + }, + { + "epoch": 5.284452211057715, + "grad_norm": 0.18921546638011932, + "learning_rate": 8.709928472485357e-05, + "loss": 0.0313, + "step": 27220 + }, + { + "epoch": 5.2863938643755155, + "grad_norm": 0.19126754999160767, + "learning_rate": 8.708819757501579e-05, + "loss": 0.037, + "step": 27230 + }, + { + "epoch": 5.288335517693316, + "grad_norm": 0.21516510844230652, + "learning_rate": 8.707710636926846e-05, + "loss": 0.0366, + "step": 27240 + }, + { + "epoch": 5.290277171011116, + "grad_norm": 0.16076324880123138, + "learning_rate": 8.706601110882455e-05, + "loss": 0.0327, + "step": 27250 + }, + { + "epoch": 5.292218824328916, + "grad_norm": 0.25619783997535706, + "learning_rate": 8.705491179489738e-05, + "loss": 0.031, + "step": 27260 + }, + { + "epoch": 5.294160477646717, + "grad_norm": 0.13441021740436554, + "learning_rate": 8.704380842870077e-05, + "loss": 0.0352, + "step": 27270 + }, + { + "epoch": 5.296102130964516, + "grad_norm": 0.18776725232601166, + "learning_rate": 8.703270101144895e-05, + "loss": 0.0366, + "step": 27280 + }, + { + "epoch": 5.298043784282316, + "grad_norm": 0.22265902161598206, + "learning_rate": 8.702158954435664e-05, + "loss": 0.0302, + "step": 27290 + }, + { + "epoch": 5.2999854376001165, + "grad_norm": 0.17718175053596497, + "learning_rate": 8.701047402863896e-05, + "loss": 0.0289, + "step": 27300 + }, + { + "epoch": 5.301927090917917, + "grad_norm": 0.14619335532188416, + "learning_rate": 8.699935446551148e-05, + "loss": 0.0387, + "step": 27310 + }, + { + "epoch": 5.303868744235717, + "grad_norm": 0.18319369852542877, + "learning_rate": 8.698823085619022e-05, + "loss": 0.0333, + "step": 27320 + }, + { + "epoch": 5.3058103975535165, + "grad_norm": 0.18103450536727905, + "learning_rate": 8.697710320189166e-05, + "loss": 0.03, + "step": 27330 + }, + { + "epoch": 5.307752050871317, + "grad_norm": 0.15954247117042542, + "learning_rate": 8.696597150383268e-05, + "loss": 0.0296, + "step": 27340 + }, + { + "epoch": 5.309693704189117, + "grad_norm": 0.2648673951625824, + "learning_rate": 8.695483576323063e-05, + "loss": 0.0368, + "step": 27350 + }, + { + "epoch": 5.311635357506917, + "grad_norm": 0.2143164724111557, + "learning_rate": 8.69436959813033e-05, + "loss": 0.0298, + "step": 27360 + }, + { + "epoch": 5.313577010824718, + "grad_norm": 0.1827014833688736, + "learning_rate": 8.693255215926892e-05, + "loss": 0.0312, + "step": 27370 + }, + { + "epoch": 5.315518664142517, + "grad_norm": 0.19347047805786133, + "learning_rate": 8.692140429834617e-05, + "loss": 0.032, + "step": 27380 + }, + { + "epoch": 5.317460317460317, + "grad_norm": 0.2754537761211395, + "learning_rate": 8.691025239975415e-05, + "loss": 0.0383, + "step": 27390 + }, + { + "epoch": 5.3194019707781175, + "grad_norm": 0.22791607677936554, + "learning_rate": 8.689909646471243e-05, + "loss": 0.035, + "step": 27400 + }, + { + "epoch": 5.321343624095918, + "grad_norm": 0.2829844355583191, + "learning_rate": 8.688793649444099e-05, + "loss": 0.0328, + "step": 27410 + }, + { + "epoch": 5.323285277413718, + "grad_norm": 0.25241512060165405, + "learning_rate": 8.687677249016029e-05, + "loss": 0.0362, + "step": 27420 + }, + { + "epoch": 5.325226930731517, + "grad_norm": 0.16105133295059204, + "learning_rate": 8.686560445309118e-05, + "loss": 0.0282, + "step": 27430 + }, + { + "epoch": 5.327168584049318, + "grad_norm": 0.2352343499660492, + "learning_rate": 8.685443238445499e-05, + "loss": 0.0284, + "step": 27440 + }, + { + "epoch": 5.329110237367118, + "grad_norm": 0.20374703407287598, + "learning_rate": 8.68432562854735e-05, + "loss": 0.0342, + "step": 27450 + }, + { + "epoch": 5.331051890684918, + "grad_norm": 0.19076310098171234, + "learning_rate": 8.683207615736887e-05, + "loss": 0.0304, + "step": 27460 + }, + { + "epoch": 5.3329935440027185, + "grad_norm": 0.2010902315378189, + "learning_rate": 8.682089200136379e-05, + "loss": 0.0397, + "step": 27470 + }, + { + "epoch": 5.334935197320519, + "grad_norm": 0.24038049578666687, + "learning_rate": 8.680970381868132e-05, + "loss": 0.0317, + "step": 27480 + }, + { + "epoch": 5.336876850638318, + "grad_norm": 0.19339002668857574, + "learning_rate": 8.679851161054498e-05, + "loss": 0.0368, + "step": 27490 + }, + { + "epoch": 5.3388185039561185, + "grad_norm": 0.27252519130706787, + "learning_rate": 8.678731537817873e-05, + "loss": 0.0357, + "step": 27500 + }, + { + "epoch": 5.340760157273919, + "grad_norm": 0.1930498331785202, + "learning_rate": 8.677611512280697e-05, + "loss": 0.0319, + "step": 27510 + }, + { + "epoch": 5.342701810591719, + "grad_norm": 0.2063945084810257, + "learning_rate": 8.676491084565457e-05, + "loss": 0.0307, + "step": 27520 + }, + { + "epoch": 5.344643463909519, + "grad_norm": 0.19960106909275055, + "learning_rate": 8.675370254794678e-05, + "loss": 0.034, + "step": 27530 + }, + { + "epoch": 5.346585117227319, + "grad_norm": 0.09223666042089462, + "learning_rate": 8.674249023090935e-05, + "loss": 0.0264, + "step": 27540 + }, + { + "epoch": 5.348526770545119, + "grad_norm": 0.24930423498153687, + "learning_rate": 8.673127389576843e-05, + "loss": 0.0308, + "step": 27550 + }, + { + "epoch": 5.350468423862919, + "grad_norm": 0.1449226289987564, + "learning_rate": 8.67200535437506e-05, + "loss": 0.0298, + "step": 27560 + }, + { + "epoch": 5.3524100771807195, + "grad_norm": 0.2261963039636612, + "learning_rate": 8.670882917608296e-05, + "loss": 0.0303, + "step": 27570 + }, + { + "epoch": 5.35435173049852, + "grad_norm": 0.20250296592712402, + "learning_rate": 8.669760079399292e-05, + "loss": 0.0264, + "step": 27580 + }, + { + "epoch": 5.35629338381632, + "grad_norm": 0.20327286422252655, + "learning_rate": 8.668636839870845e-05, + "loss": 0.0283, + "step": 27590 + }, + { + "epoch": 5.358235037134119, + "grad_norm": 0.19296513497829437, + "learning_rate": 8.667513199145789e-05, + "loss": 0.0349, + "step": 27600 + }, + { + "epoch": 5.36017669045192, + "grad_norm": 0.16078458726406097, + "learning_rate": 8.666389157347002e-05, + "loss": 0.0373, + "step": 27610 + }, + { + "epoch": 5.36211834376972, + "grad_norm": 0.20788590610027313, + "learning_rate": 8.66526471459741e-05, + "loss": 0.0403, + "step": 27620 + }, + { + "epoch": 5.36405999708752, + "grad_norm": 0.16152796149253845, + "learning_rate": 8.66413987101998e-05, + "loss": 0.0292, + "step": 27630 + }, + { + "epoch": 5.3660016504053205, + "grad_norm": 0.236563578248024, + "learning_rate": 8.663014626737723e-05, + "loss": 0.0331, + "step": 27640 + }, + { + "epoch": 5.36794330372312, + "grad_norm": 0.1813763827085495, + "learning_rate": 8.661888981873691e-05, + "loss": 0.037, + "step": 27650 + }, + { + "epoch": 5.36988495704092, + "grad_norm": 0.11343374848365784, + "learning_rate": 8.660762936550988e-05, + "loss": 0.0253, + "step": 27660 + }, + { + "epoch": 5.37182661035872, + "grad_norm": 0.23968251049518585, + "learning_rate": 8.659636490892753e-05, + "loss": 0.0384, + "step": 27670 + }, + { + "epoch": 5.373768263676521, + "grad_norm": 0.2624916136264801, + "learning_rate": 8.658509645022174e-05, + "loss": 0.0318, + "step": 27680 + }, + { + "epoch": 5.375709916994321, + "grad_norm": 0.16146083176136017, + "learning_rate": 8.657382399062481e-05, + "loss": 0.0317, + "step": 27690 + }, + { + "epoch": 5.37765157031212, + "grad_norm": 0.23806743323802948, + "learning_rate": 8.656254753136946e-05, + "loss": 0.0317, + "step": 27700 + }, + { + "epoch": 5.379593223629921, + "grad_norm": 0.18474538624286652, + "learning_rate": 8.655126707368891e-05, + "loss": 0.0362, + "step": 27710 + }, + { + "epoch": 5.381534876947721, + "grad_norm": 0.0940910056233406, + "learning_rate": 8.653998261881672e-05, + "loss": 0.0284, + "step": 27720 + }, + { + "epoch": 5.383476530265521, + "grad_norm": 0.19858328998088837, + "learning_rate": 8.652869416798699e-05, + "loss": 0.0304, + "step": 27730 + }, + { + "epoch": 5.3854181835833215, + "grad_norm": 0.12784363329410553, + "learning_rate": 8.651740172243417e-05, + "loss": 0.033, + "step": 27740 + }, + { + "epoch": 5.387359836901121, + "grad_norm": 0.19105590879917145, + "learning_rate": 8.65061052833932e-05, + "loss": 0.0279, + "step": 27750 + }, + { + "epoch": 5.389301490218921, + "grad_norm": 0.18741662800312042, + "learning_rate": 8.649480485209945e-05, + "loss": 0.03, + "step": 27760 + }, + { + "epoch": 5.391243143536721, + "grad_norm": 0.20106641948223114, + "learning_rate": 8.64835004297887e-05, + "loss": 0.0302, + "step": 27770 + }, + { + "epoch": 5.393184796854522, + "grad_norm": 0.2771984338760376, + "learning_rate": 8.64721920176972e-05, + "loss": 0.0405, + "step": 27780 + }, + { + "epoch": 5.395126450172322, + "grad_norm": 0.18765078485012054, + "learning_rate": 8.646087961706164e-05, + "loss": 0.0323, + "step": 27790 + }, + { + "epoch": 5.397068103490122, + "grad_norm": 0.27682793140411377, + "learning_rate": 8.644956322911908e-05, + "loss": 0.0318, + "step": 27800 + }, + { + "epoch": 5.399009756807922, + "grad_norm": 0.15356384217739105, + "learning_rate": 8.643824285510709e-05, + "loss": 0.0352, + "step": 27810 + }, + { + "epoch": 5.400951410125722, + "grad_norm": 0.16803774237632751, + "learning_rate": 8.642691849626364e-05, + "loss": 0.0331, + "step": 27820 + }, + { + "epoch": 5.402893063443522, + "grad_norm": 0.2959275245666504, + "learning_rate": 8.641559015382717e-05, + "loss": 0.0326, + "step": 27830 + }, + { + "epoch": 5.404834716761322, + "grad_norm": 0.20469462871551514, + "learning_rate": 8.640425782903649e-05, + "loss": 0.0334, + "step": 27840 + }, + { + "epoch": 5.406776370079123, + "grad_norm": 0.2044079452753067, + "learning_rate": 8.639292152313091e-05, + "loss": 0.035, + "step": 27850 + }, + { + "epoch": 5.408718023396922, + "grad_norm": 0.1784180849790573, + "learning_rate": 8.638158123735015e-05, + "loss": 0.0345, + "step": 27860 + }, + { + "epoch": 5.410659676714722, + "grad_norm": 0.15511643886566162, + "learning_rate": 8.637023697293436e-05, + "loss": 0.0278, + "step": 27870 + }, + { + "epoch": 5.412601330032523, + "grad_norm": 0.14983460307121277, + "learning_rate": 8.635888873112414e-05, + "loss": 0.0315, + "step": 27880 + }, + { + "epoch": 5.414542983350323, + "grad_norm": 0.23199313879013062, + "learning_rate": 8.634753651316052e-05, + "loss": 0.0365, + "step": 27890 + }, + { + "epoch": 5.416484636668123, + "grad_norm": 0.11804462969303131, + "learning_rate": 8.633618032028496e-05, + "loss": 0.0334, + "step": 27900 + }, + { + "epoch": 5.4184262899859235, + "grad_norm": 0.170469731092453, + "learning_rate": 8.632482015373934e-05, + "loss": 0.0315, + "step": 27910 + }, + { + "epoch": 5.420367943303723, + "grad_norm": 0.10598254203796387, + "learning_rate": 8.6313456014766e-05, + "loss": 0.0309, + "step": 27920 + }, + { + "epoch": 5.422309596621523, + "grad_norm": 0.11230511963367462, + "learning_rate": 8.630208790460771e-05, + "loss": 0.0294, + "step": 27930 + }, + { + "epoch": 5.424251249939323, + "grad_norm": 0.1741533875465393, + "learning_rate": 8.629071582450768e-05, + "loss": 0.0312, + "step": 27940 + }, + { + "epoch": 5.426192903257124, + "grad_norm": 0.11147715151309967, + "learning_rate": 8.62793397757095e-05, + "loss": 0.0281, + "step": 27950 + }, + { + "epoch": 5.428134556574924, + "grad_norm": 0.29273420572280884, + "learning_rate": 8.626795975945729e-05, + "loss": 0.0284, + "step": 27960 + }, + { + "epoch": 5.430076209892723, + "grad_norm": 0.15062125027179718, + "learning_rate": 8.625657577699551e-05, + "loss": 0.0253, + "step": 27970 + }, + { + "epoch": 5.432017863210524, + "grad_norm": 0.20146627724170685, + "learning_rate": 8.624518782956914e-05, + "loss": 0.033, + "step": 27980 + }, + { + "epoch": 5.433959516528324, + "grad_norm": 0.2045658528804779, + "learning_rate": 8.62337959184235e-05, + "loss": 0.0251, + "step": 27990 + }, + { + "epoch": 5.435901169846124, + "grad_norm": 0.07547573000192642, + "learning_rate": 8.622240004480441e-05, + "loss": 0.0307, + "step": 28000 + }, + { + "epoch": 5.437842823163924, + "grad_norm": 0.19977082312107086, + "learning_rate": 8.621100020995814e-05, + "loss": 0.0235, + "step": 28010 + }, + { + "epoch": 5.439784476481724, + "grad_norm": 0.1946919560432434, + "learning_rate": 8.619959641513132e-05, + "loss": 0.0381, + "step": 28020 + }, + { + "epoch": 5.441726129799524, + "grad_norm": 0.21256054937839508, + "learning_rate": 8.618818866157105e-05, + "loss": 0.0286, + "step": 28030 + }, + { + "epoch": 5.443667783117324, + "grad_norm": 0.2355918437242508, + "learning_rate": 8.617677695052487e-05, + "loss": 0.0316, + "step": 28040 + }, + { + "epoch": 5.445609436435125, + "grad_norm": 0.16625498235225677, + "learning_rate": 8.616536128324078e-05, + "loss": 0.0304, + "step": 28050 + }, + { + "epoch": 5.447551089752925, + "grad_norm": 0.21209020912647247, + "learning_rate": 8.615394166096712e-05, + "loss": 0.0315, + "step": 28060 + }, + { + "epoch": 5.449492743070724, + "grad_norm": 0.17689985036849976, + "learning_rate": 8.614251808495279e-05, + "loss": 0.0319, + "step": 28070 + }, + { + "epoch": 5.451434396388525, + "grad_norm": 0.17036476731300354, + "learning_rate": 8.6131090556447e-05, + "loss": 0.0353, + "step": 28080 + }, + { + "epoch": 5.453376049706325, + "grad_norm": 0.12247289717197418, + "learning_rate": 8.611965907669947e-05, + "loss": 0.0254, + "step": 28090 + }, + { + "epoch": 5.455317703024125, + "grad_norm": 0.12962377071380615, + "learning_rate": 8.610822364696034e-05, + "loss": 0.0297, + "step": 28100 + }, + { + "epoch": 5.457259356341925, + "grad_norm": 0.2917938530445099, + "learning_rate": 8.609678426848015e-05, + "loss": 0.0312, + "step": 28110 + }, + { + "epoch": 5.459201009659726, + "grad_norm": 0.12697888910770416, + "learning_rate": 8.60853409425099e-05, + "loss": 0.029, + "step": 28120 + }, + { + "epoch": 5.461142662977525, + "grad_norm": 0.23704494535923004, + "learning_rate": 8.607389367030104e-05, + "loss": 0.0293, + "step": 28130 + }, + { + "epoch": 5.463084316295325, + "grad_norm": 0.22823593020439148, + "learning_rate": 8.606244245310538e-05, + "loss": 0.0375, + "step": 28140 + }, + { + "epoch": 5.465025969613126, + "grad_norm": 0.19808755815029144, + "learning_rate": 8.605098729217525e-05, + "loss": 0.037, + "step": 28150 + }, + { + "epoch": 5.466967622930926, + "grad_norm": 0.2160843461751938, + "learning_rate": 8.603952818876335e-05, + "loss": 0.0271, + "step": 28160 + }, + { + "epoch": 5.468909276248726, + "grad_norm": 0.20430287718772888, + "learning_rate": 8.602806514412281e-05, + "loss": 0.0281, + "step": 28170 + }, + { + "epoch": 5.4708509295665255, + "grad_norm": 0.1725231409072876, + "learning_rate": 8.601659815950726e-05, + "loss": 0.0318, + "step": 28180 + }, + { + "epoch": 5.472792582884326, + "grad_norm": 0.12482280284166336, + "learning_rate": 8.600512723617067e-05, + "loss": 0.0228, + "step": 28190 + }, + { + "epoch": 5.474734236202126, + "grad_norm": 0.1978358030319214, + "learning_rate": 8.59936523753675e-05, + "loss": 0.0339, + "step": 28200 + }, + { + "epoch": 5.476675889519926, + "grad_norm": 0.12338658422231674, + "learning_rate": 8.598217357835264e-05, + "loss": 0.0301, + "step": 28210 + }, + { + "epoch": 5.478617542837727, + "grad_norm": 0.19960808753967285, + "learning_rate": 8.597069084638135e-05, + "loss": 0.0256, + "step": 28220 + }, + { + "epoch": 5.480559196155527, + "grad_norm": 0.12193078547716141, + "learning_rate": 8.595920418070939e-05, + "loss": 0.0283, + "step": 28230 + }, + { + "epoch": 5.482500849473326, + "grad_norm": 0.12721705436706543, + "learning_rate": 8.594771358259295e-05, + "loss": 0.0303, + "step": 28240 + }, + { + "epoch": 5.4844425027911266, + "grad_norm": 0.15781863033771515, + "learning_rate": 8.593621905328858e-05, + "loss": 0.0285, + "step": 28250 + }, + { + "epoch": 5.486384156108927, + "grad_norm": 0.1021123081445694, + "learning_rate": 8.592472059405333e-05, + "loss": 0.0263, + "step": 28260 + }, + { + "epoch": 5.488325809426727, + "grad_norm": 0.19490599632263184, + "learning_rate": 8.591321820614464e-05, + "loss": 0.0346, + "step": 28270 + }, + { + "epoch": 5.490267462744527, + "grad_norm": 0.19842372834682465, + "learning_rate": 8.590171189082041e-05, + "loss": 0.0338, + "step": 28280 + }, + { + "epoch": 5.492209116062327, + "grad_norm": 0.22860151529312134, + "learning_rate": 8.589020164933894e-05, + "loss": 0.0331, + "step": 28290 + }, + { + "epoch": 5.494150769380127, + "grad_norm": 0.17525137960910797, + "learning_rate": 8.587868748295898e-05, + "loss": 0.0264, + "step": 28300 + }, + { + "epoch": 5.496092422697927, + "grad_norm": 0.19944365322589874, + "learning_rate": 8.586716939293971e-05, + "loss": 0.0367, + "step": 28310 + }, + { + "epoch": 5.498034076015728, + "grad_norm": 0.1589127629995346, + "learning_rate": 8.58556473805407e-05, + "loss": 0.0265, + "step": 28320 + }, + { + "epoch": 5.499975729333528, + "grad_norm": 0.1499607414007187, + "learning_rate": 8.584412144702202e-05, + "loss": 0.0349, + "step": 28330 + }, + { + "epoch": 5.501917382651327, + "grad_norm": 0.28253671526908875, + "learning_rate": 8.58325915936441e-05, + "loss": 0.0329, + "step": 28340 + }, + { + "epoch": 5.5038590359691275, + "grad_norm": 0.2627990245819092, + "learning_rate": 8.582105782166783e-05, + "loss": 0.0287, + "step": 28350 + }, + { + "epoch": 5.505800689286928, + "grad_norm": 0.11893458664417267, + "learning_rate": 8.580952013235455e-05, + "loss": 0.0275, + "step": 28360 + }, + { + "epoch": 5.507742342604728, + "grad_norm": 0.2385786920785904, + "learning_rate": 8.579797852696596e-05, + "loss": 0.0321, + "step": 28370 + }, + { + "epoch": 5.509683995922528, + "grad_norm": 0.18251584470272064, + "learning_rate": 8.578643300676428e-05, + "loss": 0.0343, + "step": 28380 + }, + { + "epoch": 5.511625649240328, + "grad_norm": 0.24119067192077637, + "learning_rate": 8.577488357301209e-05, + "loss": 0.0326, + "step": 28390 + }, + { + "epoch": 5.513567302558128, + "grad_norm": 0.08077303320169449, + "learning_rate": 8.576333022697242e-05, + "loss": 0.0294, + "step": 28400 + }, + { + "epoch": 5.515508955875928, + "grad_norm": 0.1966727077960968, + "learning_rate": 8.575177296990873e-05, + "loss": 0.03, + "step": 28410 + }, + { + "epoch": 5.5174506091937285, + "grad_norm": 0.1454380601644516, + "learning_rate": 8.574021180308489e-05, + "loss": 0.0237, + "step": 28420 + }, + { + "epoch": 5.519392262511529, + "grad_norm": 0.2650313675403595, + "learning_rate": 8.572864672776523e-05, + "loss": 0.0339, + "step": 28430 + }, + { + "epoch": 5.521333915829329, + "grad_norm": 0.09125682711601257, + "learning_rate": 8.571707774521447e-05, + "loss": 0.0281, + "step": 28440 + }, + { + "epoch": 5.5232755691471285, + "grad_norm": 0.23085463047027588, + "learning_rate": 8.57055048566978e-05, + "loss": 0.037, + "step": 28450 + }, + { + "epoch": 5.525217222464929, + "grad_norm": 0.20115646719932556, + "learning_rate": 8.569392806348078e-05, + "loss": 0.033, + "step": 28460 + }, + { + "epoch": 5.527158875782729, + "grad_norm": 0.14797598123550415, + "learning_rate": 8.568234736682947e-05, + "loss": 0.0311, + "step": 28470 + }, + { + "epoch": 5.529100529100529, + "grad_norm": 0.1539364606142044, + "learning_rate": 8.567076276801029e-05, + "loss": 0.0312, + "step": 28480 + }, + { + "epoch": 5.53104218241833, + "grad_norm": 0.2487635314464569, + "learning_rate": 8.565917426829013e-05, + "loss": 0.036, + "step": 28490 + }, + { + "epoch": 5.532983835736129, + "grad_norm": 0.20602044463157654, + "learning_rate": 8.564758186893628e-05, + "loss": 0.0408, + "step": 28500 + }, + { + "epoch": 5.534925489053929, + "grad_norm": 0.1688597947359085, + "learning_rate": 8.563598557121649e-05, + "loss": 0.0332, + "step": 28510 + }, + { + "epoch": 5.5368671423717295, + "grad_norm": 0.18624797463417053, + "learning_rate": 8.562438537639888e-05, + "loss": 0.0339, + "step": 28520 + }, + { + "epoch": 5.53880879568953, + "grad_norm": 0.1215522289276123, + "learning_rate": 8.561278128575206e-05, + "loss": 0.0334, + "step": 28530 + }, + { + "epoch": 5.54075044900733, + "grad_norm": 0.19978117942810059, + "learning_rate": 8.5601173300545e-05, + "loss": 0.027, + "step": 28540 + }, + { + "epoch": 5.54269210232513, + "grad_norm": 0.16019408404827118, + "learning_rate": 8.558956142204717e-05, + "loss": 0.0284, + "step": 28550 + }, + { + "epoch": 5.54463375564293, + "grad_norm": 0.15061019361019135, + "learning_rate": 8.55779456515284e-05, + "loss": 0.0328, + "step": 28560 + }, + { + "epoch": 5.54657540896073, + "grad_norm": 0.26002466678619385, + "learning_rate": 8.556632599025898e-05, + "loss": 0.0213, + "step": 28570 + }, + { + "epoch": 5.54851706227853, + "grad_norm": 0.2363291233778, + "learning_rate": 8.555470243950964e-05, + "loss": 0.0356, + "step": 28580 + }, + { + "epoch": 5.5504587155963305, + "grad_norm": 0.1304263472557068, + "learning_rate": 8.554307500055148e-05, + "loss": 0.0251, + "step": 28590 + }, + { + "epoch": 5.55240036891413, + "grad_norm": 0.22468271851539612, + "learning_rate": 8.553144367465609e-05, + "loss": 0.0284, + "step": 28600 + }, + { + "epoch": 5.55434202223193, + "grad_norm": 0.15017808973789215, + "learning_rate": 8.551980846309544e-05, + "loss": 0.027, + "step": 28610 + }, + { + "epoch": 5.5562836755497305, + "grad_norm": 0.15781474113464355, + "learning_rate": 8.550816936714193e-05, + "loss": 0.0244, + "step": 28620 + }, + { + "epoch": 5.558225328867531, + "grad_norm": 0.1839340627193451, + "learning_rate": 8.549652638806841e-05, + "loss": 0.0317, + "step": 28630 + }, + { + "epoch": 5.560166982185331, + "grad_norm": 0.08404222875833511, + "learning_rate": 8.548487952714812e-05, + "loss": 0.0422, + "step": 28640 + }, + { + "epoch": 5.562108635503131, + "grad_norm": 0.28214043378829956, + "learning_rate": 8.547322878565478e-05, + "loss": 0.0333, + "step": 28650 + }, + { + "epoch": 5.564050288820931, + "grad_norm": 0.3273789882659912, + "learning_rate": 8.546157416486245e-05, + "loss": 0.0334, + "step": 28660 + }, + { + "epoch": 5.565991942138731, + "grad_norm": 0.2314649373292923, + "learning_rate": 8.54499156660457e-05, + "loss": 0.0298, + "step": 28670 + }, + { + "epoch": 5.567933595456531, + "grad_norm": 0.14258459210395813, + "learning_rate": 8.543825329047947e-05, + "loss": 0.0276, + "step": 28680 + }, + { + "epoch": 5.5698752487743315, + "grad_norm": 0.24401596188545227, + "learning_rate": 8.542658703943913e-05, + "loss": 0.0426, + "step": 28690 + }, + { + "epoch": 5.571816902092132, + "grad_norm": 0.23706363141536713, + "learning_rate": 8.541491691420051e-05, + "loss": 0.0279, + "step": 28700 + }, + { + "epoch": 5.573758555409931, + "grad_norm": 0.15503594279289246, + "learning_rate": 8.54032429160398e-05, + "loss": 0.0295, + "step": 28710 + }, + { + "epoch": 5.575700208727731, + "grad_norm": 0.3059500455856323, + "learning_rate": 8.539156504623369e-05, + "loss": 0.033, + "step": 28720 + }, + { + "epoch": 5.577641862045532, + "grad_norm": 0.20680804550647736, + "learning_rate": 8.537988330605923e-05, + "loss": 0.0386, + "step": 28730 + }, + { + "epoch": 5.579583515363332, + "grad_norm": 0.206232488155365, + "learning_rate": 8.536819769679393e-05, + "loss": 0.0292, + "step": 28740 + }, + { + "epoch": 5.581525168681132, + "grad_norm": 0.29472801089286804, + "learning_rate": 8.53565082197157e-05, + "loss": 0.0273, + "step": 28750 + }, + { + "epoch": 5.5834668219989325, + "grad_norm": 0.16522538661956787, + "learning_rate": 8.534481487610289e-05, + "loss": 0.0288, + "step": 28760 + }, + { + "epoch": 5.585408475316732, + "grad_norm": 0.2532913386821747, + "learning_rate": 8.533311766723428e-05, + "loss": 0.035, + "step": 28770 + }, + { + "epoch": 5.587350128634532, + "grad_norm": 0.12113380432128906, + "learning_rate": 8.532141659438901e-05, + "loss": 0.0319, + "step": 28780 + }, + { + "epoch": 5.5892917819523324, + "grad_norm": 0.16701175272464752, + "learning_rate": 8.530971165884675e-05, + "loss": 0.0254, + "step": 28790 + }, + { + "epoch": 5.591233435270133, + "grad_norm": 0.22009491920471191, + "learning_rate": 8.529800286188752e-05, + "loss": 0.033, + "step": 28800 + }, + { + "epoch": 5.593175088587933, + "grad_norm": 0.11936478316783905, + "learning_rate": 8.528629020479175e-05, + "loss": 0.0299, + "step": 28810 + }, + { + "epoch": 5.595116741905732, + "grad_norm": 0.23042358458042145, + "learning_rate": 8.527457368884033e-05, + "loss": 0.0378, + "step": 28820 + }, + { + "epoch": 5.597058395223533, + "grad_norm": 0.22708603739738464, + "learning_rate": 8.526285331531458e-05, + "loss": 0.0266, + "step": 28830 + }, + { + "epoch": 5.599000048541333, + "grad_norm": 0.09449120610952377, + "learning_rate": 8.525112908549621e-05, + "loss": 0.0295, + "step": 28840 + }, + { + "epoch": 5.600941701859133, + "grad_norm": 0.1327795833349228, + "learning_rate": 8.523940100066735e-05, + "loss": 0.0275, + "step": 28850 + }, + { + "epoch": 5.6028833551769335, + "grad_norm": 0.18500319123268127, + "learning_rate": 8.52276690621106e-05, + "loss": 0.0251, + "step": 28860 + }, + { + "epoch": 5.604825008494734, + "grad_norm": 0.22036516666412354, + "learning_rate": 8.521593327110889e-05, + "loss": 0.0252, + "step": 28870 + }, + { + "epoch": 5.606766661812533, + "grad_norm": 0.21968533098697662, + "learning_rate": 8.520419362894569e-05, + "loss": 0.0312, + "step": 28880 + }, + { + "epoch": 5.608708315130333, + "grad_norm": 0.19512662291526794, + "learning_rate": 8.51924501369048e-05, + "loss": 0.0321, + "step": 28890 + }, + { + "epoch": 5.610649968448134, + "grad_norm": 0.1977577805519104, + "learning_rate": 8.518070279627047e-05, + "loss": 0.0281, + "step": 28900 + }, + { + "epoch": 5.612591621765934, + "grad_norm": 0.16095830500125885, + "learning_rate": 8.516895160832737e-05, + "loss": 0.0403, + "step": 28910 + }, + { + "epoch": 5.614533275083733, + "grad_norm": 0.12269501388072968, + "learning_rate": 8.515719657436061e-05, + "loss": 0.0348, + "step": 28920 + }, + { + "epoch": 5.616474928401534, + "grad_norm": 0.1604570895433426, + "learning_rate": 8.514543769565568e-05, + "loss": 0.0335, + "step": 28930 + }, + { + "epoch": 5.618416581719334, + "grad_norm": 0.15120114386081696, + "learning_rate": 8.513367497349853e-05, + "loss": 0.0322, + "step": 28940 + }, + { + "epoch": 5.620358235037134, + "grad_norm": 0.18620705604553223, + "learning_rate": 8.51219084091755e-05, + "loss": 0.0332, + "step": 28950 + }, + { + "epoch": 5.622299888354934, + "grad_norm": 0.13111956417560577, + "learning_rate": 8.511013800397338e-05, + "loss": 0.0268, + "step": 28960 + }, + { + "epoch": 5.624241541672735, + "grad_norm": 0.19632716476917267, + "learning_rate": 8.509836375917937e-05, + "loss": 0.0458, + "step": 28970 + }, + { + "epoch": 5.626183194990534, + "grad_norm": 0.16489143669605255, + "learning_rate": 8.508658567608104e-05, + "loss": 0.0297, + "step": 28980 + }, + { + "epoch": 5.628124848308334, + "grad_norm": 0.2858222424983978, + "learning_rate": 8.507480375596647e-05, + "loss": 0.0316, + "step": 28990 + }, + { + "epoch": 5.630066501626135, + "grad_norm": 0.2320927381515503, + "learning_rate": 8.506301800012408e-05, + "loss": 0.0286, + "step": 29000 + }, + { + "epoch": 5.632008154943935, + "grad_norm": 0.18890982866287231, + "learning_rate": 8.505122840984278e-05, + "loss": 0.0314, + "step": 29010 + }, + { + "epoch": 5.633949808261735, + "grad_norm": 0.1395580768585205, + "learning_rate": 8.503943498641182e-05, + "loss": 0.0335, + "step": 29020 + }, + { + "epoch": 5.635891461579535, + "grad_norm": 0.17069382965564728, + "learning_rate": 8.502763773112095e-05, + "loss": 0.0291, + "step": 29030 + }, + { + "epoch": 5.637833114897335, + "grad_norm": 0.13181807100772858, + "learning_rate": 8.501583664526026e-05, + "loss": 0.0244, + "step": 29040 + }, + { + "epoch": 5.639774768215135, + "grad_norm": 0.24742917716503143, + "learning_rate": 8.500403173012032e-05, + "loss": 0.0295, + "step": 29050 + }, + { + "epoch": 5.641716421532935, + "grad_norm": 0.23225454986095428, + "learning_rate": 8.499222298699211e-05, + "loss": 0.0344, + "step": 29060 + }, + { + "epoch": 5.643658074850736, + "grad_norm": 0.24199619889259338, + "learning_rate": 8.498041041716701e-05, + "loss": 0.0411, + "step": 29070 + }, + { + "epoch": 5.645599728168536, + "grad_norm": 0.16393554210662842, + "learning_rate": 8.496859402193681e-05, + "loss": 0.0292, + "step": 29080 + }, + { + "epoch": 5.647541381486335, + "grad_norm": 0.20084623992443085, + "learning_rate": 8.495677380259374e-05, + "loss": 0.031, + "step": 29090 + }, + { + "epoch": 5.649483034804136, + "grad_norm": 0.12491276860237122, + "learning_rate": 8.494494976043045e-05, + "loss": 0.0281, + "step": 29100 + }, + { + "epoch": 5.651424688121936, + "grad_norm": 0.25177615880966187, + "learning_rate": 8.493312189673998e-05, + "loss": 0.0419, + "step": 29110 + }, + { + "epoch": 5.653366341439736, + "grad_norm": 0.15612486004829407, + "learning_rate": 8.492129021281584e-05, + "loss": 0.0297, + "step": 29120 + }, + { + "epoch": 5.655307994757536, + "grad_norm": 0.15625858306884766, + "learning_rate": 8.490945470995188e-05, + "loss": 0.0345, + "step": 29130 + }, + { + "epoch": 5.657249648075336, + "grad_norm": 0.20436525344848633, + "learning_rate": 8.489761538944247e-05, + "loss": 0.0326, + "step": 29140 + }, + { + "epoch": 5.659191301393136, + "grad_norm": 0.11758942157030106, + "learning_rate": 8.48857722525823e-05, + "loss": 0.0302, + "step": 29150 + }, + { + "epoch": 5.661132954710936, + "grad_norm": 0.192983940243721, + "learning_rate": 8.487392530066652e-05, + "loss": 0.0331, + "step": 29160 + }, + { + "epoch": 5.663074608028737, + "grad_norm": 0.18576504290103912, + "learning_rate": 8.486207453499069e-05, + "loss": 0.0284, + "step": 29170 + }, + { + "epoch": 5.665016261346537, + "grad_norm": 0.19327451288700104, + "learning_rate": 8.485021995685082e-05, + "loss": 0.0269, + "step": 29180 + }, + { + "epoch": 5.666957914664337, + "grad_norm": 0.18412970006465912, + "learning_rate": 8.483836156754328e-05, + "loss": 0.0279, + "step": 29190 + }, + { + "epoch": 5.668899567982137, + "grad_norm": 0.0918130874633789, + "learning_rate": 8.482649936836491e-05, + "loss": 0.0294, + "step": 29200 + }, + { + "epoch": 5.670841221299937, + "grad_norm": 0.1519562005996704, + "learning_rate": 8.481463336061293e-05, + "loss": 0.0231, + "step": 29210 + }, + { + "epoch": 5.672782874617737, + "grad_norm": 0.20789887011051178, + "learning_rate": 8.480276354558496e-05, + "loss": 0.0389, + "step": 29220 + }, + { + "epoch": 5.674724527935537, + "grad_norm": 0.12184341251850128, + "learning_rate": 8.479088992457913e-05, + "loss": 0.0302, + "step": 29230 + }, + { + "epoch": 5.676666181253337, + "grad_norm": 0.13352392613887787, + "learning_rate": 8.477901249889387e-05, + "loss": 0.026, + "step": 29240 + }, + { + "epoch": 5.678607834571137, + "grad_norm": 0.15029264986515045, + "learning_rate": 8.47671312698281e-05, + "loss": 0.0411, + "step": 29250 + }, + { + "epoch": 5.680549487888937, + "grad_norm": 0.24794729053974152, + "learning_rate": 8.475524623868112e-05, + "loss": 0.0353, + "step": 29260 + }, + { + "epoch": 5.682491141206738, + "grad_norm": 0.27991631627082825, + "learning_rate": 8.474335740675266e-05, + "loss": 0.0402, + "step": 29270 + }, + { + "epoch": 5.684432794524538, + "grad_norm": 0.1853950470685959, + "learning_rate": 8.473146477534289e-05, + "loss": 0.0308, + "step": 29280 + }, + { + "epoch": 5.686374447842338, + "grad_norm": 0.41041457653045654, + "learning_rate": 8.471956834575232e-05, + "loss": 0.0323, + "step": 29290 + }, + { + "epoch": 5.6883161011601375, + "grad_norm": 0.09830740094184875, + "learning_rate": 8.470766811928197e-05, + "loss": 0.0345, + "step": 29300 + }, + { + "epoch": 5.690257754477938, + "grad_norm": 0.14750806987285614, + "learning_rate": 8.469576409723323e-05, + "loss": 0.0261, + "step": 29310 + }, + { + "epoch": 5.692199407795738, + "grad_norm": 0.22560210525989532, + "learning_rate": 8.468385628090788e-05, + "loss": 0.0324, + "step": 29320 + }, + { + "epoch": 5.694141061113538, + "grad_norm": 0.21497830748558044, + "learning_rate": 8.467194467160815e-05, + "loss": 0.032, + "step": 29330 + }, + { + "epoch": 5.696082714431339, + "grad_norm": 0.15039516985416412, + "learning_rate": 8.466002927063667e-05, + "loss": 0.0257, + "step": 29340 + }, + { + "epoch": 5.698024367749138, + "grad_norm": 0.21822181344032288, + "learning_rate": 8.464811007929651e-05, + "loss": 0.0309, + "step": 29350 + }, + { + "epoch": 5.699966021066938, + "grad_norm": 0.16659851372241974, + "learning_rate": 8.463618709889114e-05, + "loss": 0.0293, + "step": 29360 + }, + { + "epoch": 5.7019076743847386, + "grad_norm": 0.22404257953166962, + "learning_rate": 8.462426033072442e-05, + "loss": 0.0267, + "step": 29370 + }, + { + "epoch": 5.703849327702539, + "grad_norm": 0.25775304436683655, + "learning_rate": 8.461232977610061e-05, + "loss": 0.0331, + "step": 29380 + }, + { + "epoch": 5.705790981020339, + "grad_norm": 0.10437467694282532, + "learning_rate": 8.46003954363245e-05, + "loss": 0.0314, + "step": 29390 + }, + { + "epoch": 5.707732634338139, + "grad_norm": 0.1503445953130722, + "learning_rate": 8.458845731270115e-05, + "loss": 0.0335, + "step": 29400 + }, + { + "epoch": 5.709674287655939, + "grad_norm": 0.20275269448757172, + "learning_rate": 8.45765154065361e-05, + "loss": 0.0436, + "step": 29410 + }, + { + "epoch": 5.711615940973739, + "grad_norm": 0.13658571243286133, + "learning_rate": 8.456456971913532e-05, + "loss": 0.0292, + "step": 29420 + }, + { + "epoch": 5.713557594291539, + "grad_norm": 0.15180252492427826, + "learning_rate": 8.455262025180517e-05, + "loss": 0.0291, + "step": 29430 + }, + { + "epoch": 5.71549924760934, + "grad_norm": 0.21921858191490173, + "learning_rate": 8.454066700585242e-05, + "loss": 0.042, + "step": 29440 + }, + { + "epoch": 5.717440900927139, + "grad_norm": 0.22832922637462616, + "learning_rate": 8.452870998258423e-05, + "loss": 0.0288, + "step": 29450 + }, + { + "epoch": 5.719382554244939, + "grad_norm": 0.16080209612846375, + "learning_rate": 8.451674918330825e-05, + "loss": 0.0314, + "step": 29460 + }, + { + "epoch": 5.7213242075627395, + "grad_norm": 0.24477632343769073, + "learning_rate": 8.450478460933246e-05, + "loss": 0.0412, + "step": 29470 + }, + { + "epoch": 5.72326586088054, + "grad_norm": 0.1685037612915039, + "learning_rate": 8.449281626196532e-05, + "loss": 0.0376, + "step": 29480 + }, + { + "epoch": 5.72520751419834, + "grad_norm": 0.19988365471363068, + "learning_rate": 8.448084414251564e-05, + "loss": 0.035, + "step": 29490 + }, + { + "epoch": 5.72714916751614, + "grad_norm": 0.17968717217445374, + "learning_rate": 8.446886825229271e-05, + "loss": 0.0337, + "step": 29500 + }, + { + "epoch": 5.72909082083394, + "grad_norm": 0.1342824548482895, + "learning_rate": 8.445688859260615e-05, + "loss": 0.0316, + "step": 29510 + }, + { + "epoch": 5.73103247415174, + "grad_norm": 0.168105810880661, + "learning_rate": 8.444490516476606e-05, + "loss": 0.0346, + "step": 29520 + }, + { + "epoch": 5.73297412746954, + "grad_norm": 0.3409203588962555, + "learning_rate": 8.443291797008293e-05, + "loss": 0.0389, + "step": 29530 + }, + { + "epoch": 5.7349157807873405, + "grad_norm": 0.25137367844581604, + "learning_rate": 8.442092700986765e-05, + "loss": 0.0397, + "step": 29540 + }, + { + "epoch": 5.736857434105141, + "grad_norm": 0.15089507400989532, + "learning_rate": 8.440893228543156e-05, + "loss": 0.0336, + "step": 29550 + }, + { + "epoch": 5.73879908742294, + "grad_norm": 0.23265643417835236, + "learning_rate": 8.439693379808638e-05, + "loss": 0.0286, + "step": 29560 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 0.2991223931312561, + "learning_rate": 8.43849315491442e-05, + "loss": 0.032, + "step": 29570 + }, + { + "epoch": 5.742682394058541, + "grad_norm": 0.1542273759841919, + "learning_rate": 8.437292553991763e-05, + "loss": 0.0316, + "step": 29580 + }, + { + "epoch": 5.744624047376341, + "grad_norm": 0.23037809133529663, + "learning_rate": 8.436091577171959e-05, + "loss": 0.0277, + "step": 29590 + }, + { + "epoch": 5.746565700694141, + "grad_norm": 0.16908694803714752, + "learning_rate": 8.434890224586347e-05, + "loss": 0.0284, + "step": 29600 + }, + { + "epoch": 5.748507354011942, + "grad_norm": 0.2623303532600403, + "learning_rate": 8.433688496366303e-05, + "loss": 0.0301, + "step": 29610 + }, + { + "epoch": 5.750449007329741, + "grad_norm": 0.1772368997335434, + "learning_rate": 8.432486392643248e-05, + "loss": 0.0303, + "step": 29620 + }, + { + "epoch": 5.752390660647541, + "grad_norm": 0.14497137069702148, + "learning_rate": 8.431283913548643e-05, + "loss": 0.0325, + "step": 29630 + }, + { + "epoch": 5.7543323139653415, + "grad_norm": 0.16695836186408997, + "learning_rate": 8.430081059213985e-05, + "loss": 0.0311, + "step": 29640 + }, + { + "epoch": 5.756273967283142, + "grad_norm": 0.18265685439109802, + "learning_rate": 8.428877829770823e-05, + "loss": 0.0312, + "step": 29650 + }, + { + "epoch": 5.758215620600942, + "grad_norm": 0.2991071343421936, + "learning_rate": 8.427674225350735e-05, + "loss": 0.0362, + "step": 29660 + }, + { + "epoch": 5.760157273918741, + "grad_norm": 0.20770803093910217, + "learning_rate": 8.426470246085347e-05, + "loss": 0.0319, + "step": 29670 + }, + { + "epoch": 5.762098927236542, + "grad_norm": 0.24027590453624725, + "learning_rate": 8.425265892106324e-05, + "loss": 0.03, + "step": 29680 + }, + { + "epoch": 5.764040580554342, + "grad_norm": 0.19181346893310547, + "learning_rate": 8.424061163545374e-05, + "loss": 0.0351, + "step": 29690 + }, + { + "epoch": 5.765982233872142, + "grad_norm": 0.2953486740589142, + "learning_rate": 8.422856060534243e-05, + "loss": 0.0315, + "step": 29700 + }, + { + "epoch": 5.7679238871899425, + "grad_norm": 0.17827092111110687, + "learning_rate": 8.421650583204718e-05, + "loss": 0.031, + "step": 29710 + }, + { + "epoch": 5.769865540507743, + "grad_norm": 0.17209841310977936, + "learning_rate": 8.420444731688633e-05, + "loss": 0.0338, + "step": 29720 + }, + { + "epoch": 5.771807193825542, + "grad_norm": 0.20551186800003052, + "learning_rate": 8.419238506117852e-05, + "loss": 0.0295, + "step": 29730 + }, + { + "epoch": 5.7737488471433425, + "grad_norm": 0.22397680580615997, + "learning_rate": 8.418031906624289e-05, + "loss": 0.0295, + "step": 29740 + }, + { + "epoch": 5.775690500461143, + "grad_norm": 0.1763957440853119, + "learning_rate": 8.416824933339898e-05, + "loss": 0.0298, + "step": 29750 + }, + { + "epoch": 5.777632153778943, + "grad_norm": 0.20973224937915802, + "learning_rate": 8.415617586396667e-05, + "loss": 0.0261, + "step": 29760 + }, + { + "epoch": 5.779573807096742, + "grad_norm": 0.278302401304245, + "learning_rate": 8.414409865926632e-05, + "loss": 0.03, + "step": 29770 + }, + { + "epoch": 5.781515460414543, + "grad_norm": 0.26034146547317505, + "learning_rate": 8.413201772061867e-05, + "loss": 0.0337, + "step": 29780 + }, + { + "epoch": 5.783457113732343, + "grad_norm": 0.1608598232269287, + "learning_rate": 8.411993304934488e-05, + "loss": 0.0314, + "step": 29790 + }, + { + "epoch": 5.785398767050143, + "grad_norm": 0.2621002495288849, + "learning_rate": 8.410784464676654e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 5.7873404203679435, + "grad_norm": 0.1862478107213974, + "learning_rate": 8.409575251420556e-05, + "loss": 0.0278, + "step": 29810 + }, + { + "epoch": 5.789282073685744, + "grad_norm": 0.12934927642345428, + "learning_rate": 8.408365665298435e-05, + "loss": 0.0312, + "step": 29820 + }, + { + "epoch": 5.791223727003543, + "grad_norm": 0.3335045278072357, + "learning_rate": 8.40715570644257e-05, + "loss": 0.0349, + "step": 29830 + }, + { + "epoch": 5.793165380321343, + "grad_norm": 0.21333549916744232, + "learning_rate": 8.40594537498528e-05, + "loss": 0.035, + "step": 29840 + }, + { + "epoch": 5.795107033639144, + "grad_norm": 0.1879519671201706, + "learning_rate": 8.404734671058924e-05, + "loss": 0.0296, + "step": 29850 + }, + { + "epoch": 5.797048686956944, + "grad_norm": 0.2100723832845688, + "learning_rate": 8.403523594795902e-05, + "loss": 0.035, + "step": 29860 + }, + { + "epoch": 5.798990340274744, + "grad_norm": 0.1517438143491745, + "learning_rate": 8.402312146328659e-05, + "loss": 0.0251, + "step": 29870 + }, + { + "epoch": 5.800931993592544, + "grad_norm": 0.169762521982193, + "learning_rate": 8.401100325789675e-05, + "loss": 0.0309, + "step": 29880 + }, + { + "epoch": 5.802873646910344, + "grad_norm": 0.14354580640792847, + "learning_rate": 8.399888133311472e-05, + "loss": 0.0352, + "step": 29890 + }, + { + "epoch": 5.804815300228144, + "grad_norm": 0.16634631156921387, + "learning_rate": 8.398675569026613e-05, + "loss": 0.0313, + "step": 29900 + }, + { + "epoch": 5.8067569535459445, + "grad_norm": 0.1506464034318924, + "learning_rate": 8.397462633067705e-05, + "loss": 0.0285, + "step": 29910 + }, + { + "epoch": 5.808698606863745, + "grad_norm": 0.14835508167743683, + "learning_rate": 8.396249325567392e-05, + "loss": 0.0301, + "step": 29920 + }, + { + "epoch": 5.810640260181545, + "grad_norm": 0.23141281306743622, + "learning_rate": 8.395035646658357e-05, + "loss": 0.0251, + "step": 29930 + }, + { + "epoch": 5.812581913499344, + "grad_norm": 0.17414656281471252, + "learning_rate": 8.39382159647333e-05, + "loss": 0.0274, + "step": 29940 + }, + { + "epoch": 5.814523566817145, + "grad_norm": 0.19520623981952667, + "learning_rate": 8.392607175145075e-05, + "loss": 0.0287, + "step": 29950 + }, + { + "epoch": 5.816465220134945, + "grad_norm": 0.21409347653388977, + "learning_rate": 8.3913923828064e-05, + "loss": 0.0284, + "step": 29960 + }, + { + "epoch": 5.818406873452745, + "grad_norm": 0.1668071299791336, + "learning_rate": 8.390177219590152e-05, + "loss": 0.0317, + "step": 29970 + }, + { + "epoch": 5.8203485267705455, + "grad_norm": 0.2462553232908249, + "learning_rate": 8.388961685629222e-05, + "loss": 0.0354, + "step": 29980 + }, + { + "epoch": 5.822290180088345, + "grad_norm": 0.2478836476802826, + "learning_rate": 8.387745781056536e-05, + "loss": 0.0294, + "step": 29990 + }, + { + "epoch": 5.824231833406145, + "grad_norm": 0.19641409814357758, + "learning_rate": 8.386529506005065e-05, + "loss": 0.0365, + "step": 30000 + }, + { + "epoch": 5.826173486723945, + "grad_norm": 0.1877579689025879, + "learning_rate": 8.38531286060782e-05, + "loss": 0.0286, + "step": 30010 + }, + { + "epoch": 5.828115140041746, + "grad_norm": 0.17898806929588318, + "learning_rate": 8.384095844997849e-05, + "loss": 0.0255, + "step": 30020 + }, + { + "epoch": 5.830056793359546, + "grad_norm": 0.23618628084659576, + "learning_rate": 8.382878459308245e-05, + "loss": 0.0334, + "step": 30030 + }, + { + "epoch": 5.831998446677346, + "grad_norm": 0.22219057381153107, + "learning_rate": 8.381660703672138e-05, + "loss": 0.0236, + "step": 30040 + }, + { + "epoch": 5.833940099995146, + "grad_norm": 0.16750474274158478, + "learning_rate": 8.380442578222702e-05, + "loss": 0.0272, + "step": 30050 + }, + { + "epoch": 5.835881753312946, + "grad_norm": 0.1287144422531128, + "learning_rate": 8.379224083093146e-05, + "loss": 0.0415, + "step": 30060 + }, + { + "epoch": 5.837823406630746, + "grad_norm": 0.1525782346725464, + "learning_rate": 8.378005218416727e-05, + "loss": 0.0309, + "step": 30070 + }, + { + "epoch": 5.839765059948546, + "grad_norm": 0.3484955132007599, + "learning_rate": 8.376785984326735e-05, + "loss": 0.0338, + "step": 30080 + }, + { + "epoch": 5.841706713266346, + "grad_norm": 0.11412868648767471, + "learning_rate": 8.375566380956506e-05, + "loss": 0.0313, + "step": 30090 + }, + { + "epoch": 5.843648366584146, + "grad_norm": 0.10681269317865372, + "learning_rate": 8.374346408439411e-05, + "loss": 0.0306, + "step": 30100 + }, + { + "epoch": 5.845590019901946, + "grad_norm": 0.13267026841640472, + "learning_rate": 8.373126066908868e-05, + "loss": 0.0268, + "step": 30110 + }, + { + "epoch": 5.847531673219747, + "grad_norm": 0.1523614376783371, + "learning_rate": 8.371905356498326e-05, + "loss": 0.0303, + "step": 30120 + }, + { + "epoch": 5.849473326537547, + "grad_norm": 0.2465251088142395, + "learning_rate": 8.370684277341288e-05, + "loss": 0.0312, + "step": 30130 + }, + { + "epoch": 5.851414979855347, + "grad_norm": 0.16347268223762512, + "learning_rate": 8.369462829571282e-05, + "loss": 0.0312, + "step": 30140 + }, + { + "epoch": 5.853356633173147, + "grad_norm": 0.18480968475341797, + "learning_rate": 8.36824101332189e-05, + "loss": 0.036, + "step": 30150 + }, + { + "epoch": 5.855298286490947, + "grad_norm": 0.24608498811721802, + "learning_rate": 8.367018828726721e-05, + "loss": 0.0324, + "step": 30160 + }, + { + "epoch": 5.857239939808747, + "grad_norm": 0.15034523606300354, + "learning_rate": 8.365796275919438e-05, + "loss": 0.0293, + "step": 30170 + }, + { + "epoch": 5.859181593126547, + "grad_norm": 0.11669853329658508, + "learning_rate": 8.364573355033734e-05, + "loss": 0.0334, + "step": 30180 + }, + { + "epoch": 5.861123246444348, + "grad_norm": 0.15219919383525848, + "learning_rate": 8.363350066203346e-05, + "loss": 0.0242, + "step": 30190 + }, + { + "epoch": 5.863064899762147, + "grad_norm": 0.20297811925411224, + "learning_rate": 8.362126409562053e-05, + "loss": 0.0308, + "step": 30200 + }, + { + "epoch": 5.865006553079947, + "grad_norm": 0.152302548289299, + "learning_rate": 8.360902385243667e-05, + "loss": 0.0286, + "step": 30210 + }, + { + "epoch": 5.866948206397748, + "grad_norm": 0.1320721060037613, + "learning_rate": 8.359677993382052e-05, + "loss": 0.0404, + "step": 30220 + }, + { + "epoch": 5.868889859715548, + "grad_norm": 0.21738646924495697, + "learning_rate": 8.358453234111103e-05, + "loss": 0.035, + "step": 30230 + }, + { + "epoch": 5.870831513033348, + "grad_norm": 0.2027326375246048, + "learning_rate": 8.357228107564756e-05, + "loss": 0.032, + "step": 30240 + }, + { + "epoch": 5.872773166351148, + "grad_norm": 0.27783188223838806, + "learning_rate": 8.356002613876993e-05, + "loss": 0.0341, + "step": 30250 + }, + { + "epoch": 5.874714819668948, + "grad_norm": 0.21355463564395905, + "learning_rate": 8.35477675318183e-05, + "loss": 0.0256, + "step": 30260 + }, + { + "epoch": 5.876656472986748, + "grad_norm": 0.21550583839416504, + "learning_rate": 8.353550525613323e-05, + "loss": 0.0337, + "step": 30270 + }, + { + "epoch": 5.878598126304548, + "grad_norm": 0.2185061275959015, + "learning_rate": 8.352323931305572e-05, + "loss": 0.0335, + "step": 30280 + }, + { + "epoch": 5.880539779622349, + "grad_norm": 0.19217655062675476, + "learning_rate": 8.351096970392717e-05, + "loss": 0.027, + "step": 30290 + }, + { + "epoch": 5.882481432940149, + "grad_norm": 0.18502764403820038, + "learning_rate": 8.349869643008937e-05, + "loss": 0.0348, + "step": 30300 + }, + { + "epoch": 5.884423086257948, + "grad_norm": 0.15581014752388, + "learning_rate": 8.348641949288449e-05, + "loss": 0.0288, + "step": 30310 + }, + { + "epoch": 5.886364739575749, + "grad_norm": 0.13887357711791992, + "learning_rate": 8.34741388936551e-05, + "loss": 0.0225, + "step": 30320 + }, + { + "epoch": 5.888306392893549, + "grad_norm": 0.19200469553470612, + "learning_rate": 8.346185463374423e-05, + "loss": 0.0269, + "step": 30330 + }, + { + "epoch": 5.890248046211349, + "grad_norm": 0.19645774364471436, + "learning_rate": 8.344956671449524e-05, + "loss": 0.0269, + "step": 30340 + }, + { + "epoch": 5.892189699529149, + "grad_norm": 0.1530791074037552, + "learning_rate": 8.343727513725192e-05, + "loss": 0.0296, + "step": 30350 + }, + { + "epoch": 5.89413135284695, + "grad_norm": 0.20044225454330444, + "learning_rate": 8.342497990335847e-05, + "loss": 0.0271, + "step": 30360 + }, + { + "epoch": 5.896073006164749, + "grad_norm": 0.17832259833812714, + "learning_rate": 8.341268101415946e-05, + "loss": 0.03, + "step": 30370 + }, + { + "epoch": 5.898014659482549, + "grad_norm": 0.21114994585514069, + "learning_rate": 8.34003784709999e-05, + "loss": 0.0312, + "step": 30380 + }, + { + "epoch": 5.89995631280035, + "grad_norm": 0.3038488030433655, + "learning_rate": 8.338807227522516e-05, + "loss": 0.034, + "step": 30390 + }, + { + "epoch": 5.90189796611815, + "grad_norm": 0.13354896008968353, + "learning_rate": 8.337576242818103e-05, + "loss": 0.0335, + "step": 30400 + }, + { + "epoch": 5.903839619435949, + "grad_norm": 0.152410089969635, + "learning_rate": 8.336344893121372e-05, + "loss": 0.0268, + "step": 30410 + }, + { + "epoch": 5.9057812727537495, + "grad_norm": 0.13639646768569946, + "learning_rate": 8.335113178566977e-05, + "loss": 0.0254, + "step": 30420 + }, + { + "epoch": 5.90772292607155, + "grad_norm": 0.150845468044281, + "learning_rate": 8.33388109928962e-05, + "loss": 0.031, + "step": 30430 + }, + { + "epoch": 5.90966457938935, + "grad_norm": 0.15227511525154114, + "learning_rate": 8.33264865542404e-05, + "loss": 0.0232, + "step": 30440 + }, + { + "epoch": 5.91160623270715, + "grad_norm": 0.1811702400445938, + "learning_rate": 8.331415847105013e-05, + "loss": 0.0289, + "step": 30450 + }, + { + "epoch": 5.913547886024951, + "grad_norm": 0.17806842923164368, + "learning_rate": 8.330182674467357e-05, + "loss": 0.0335, + "step": 30460 + }, + { + "epoch": 5.91548953934275, + "grad_norm": 0.1911558210849762, + "learning_rate": 8.32894913764593e-05, + "loss": 0.0292, + "step": 30470 + }, + { + "epoch": 5.91743119266055, + "grad_norm": 0.15945881605148315, + "learning_rate": 8.327715236775633e-05, + "loss": 0.0306, + "step": 30480 + }, + { + "epoch": 5.919372845978351, + "grad_norm": 0.2380877435207367, + "learning_rate": 8.326480971991398e-05, + "loss": 0.0354, + "step": 30490 + }, + { + "epoch": 5.921314499296151, + "grad_norm": 0.31091439723968506, + "learning_rate": 8.325246343428206e-05, + "loss": 0.0362, + "step": 30500 + }, + { + "epoch": 5.923256152613951, + "grad_norm": 0.20974032580852509, + "learning_rate": 8.324011351221072e-05, + "loss": 0.0335, + "step": 30510 + }, + { + "epoch": 5.9251978059317505, + "grad_norm": 0.15750457346439362, + "learning_rate": 8.322775995505057e-05, + "loss": 0.0279, + "step": 30520 + }, + { + "epoch": 5.927139459249551, + "grad_norm": 0.12405212968587875, + "learning_rate": 8.321540276415254e-05, + "loss": 0.0355, + "step": 30530 + }, + { + "epoch": 5.929081112567351, + "grad_norm": 0.2711668908596039, + "learning_rate": 8.320304194086798e-05, + "loss": 0.0324, + "step": 30540 + }, + { + "epoch": 5.931022765885151, + "grad_norm": 0.21462242305278778, + "learning_rate": 8.31906774865487e-05, + "loss": 0.0304, + "step": 30550 + }, + { + "epoch": 5.932964419202952, + "grad_norm": 0.2256188541650772, + "learning_rate": 8.317830940254682e-05, + "loss": 0.0337, + "step": 30560 + }, + { + "epoch": 5.934906072520752, + "grad_norm": 0.13999390602111816, + "learning_rate": 8.316593769021491e-05, + "loss": 0.0336, + "step": 30570 + }, + { + "epoch": 5.936847725838551, + "grad_norm": 0.20760352909564972, + "learning_rate": 8.315356235090592e-05, + "loss": 0.0392, + "step": 30580 + }, + { + "epoch": 5.9387893791563515, + "grad_norm": 0.16125038266181946, + "learning_rate": 8.314118338597319e-05, + "loss": 0.0355, + "step": 30590 + }, + { + "epoch": 5.940731032474152, + "grad_norm": 0.14495807886123657, + "learning_rate": 8.312880079677048e-05, + "loss": 0.0303, + "step": 30600 + }, + { + "epoch": 5.942672685791952, + "grad_norm": 0.3125211000442505, + "learning_rate": 8.311641458465191e-05, + "loss": 0.03, + "step": 30610 + }, + { + "epoch": 5.9446143391097515, + "grad_norm": 0.19698476791381836, + "learning_rate": 8.310402475097205e-05, + "loss": 0.0357, + "step": 30620 + }, + { + "epoch": 5.946555992427552, + "grad_norm": 0.18336845934391022, + "learning_rate": 8.309163129708581e-05, + "loss": 0.0384, + "step": 30630 + }, + { + "epoch": 5.948497645745352, + "grad_norm": 0.22776572406291962, + "learning_rate": 8.307923422434852e-05, + "loss": 0.0392, + "step": 30640 + }, + { + "epoch": 5.950439299063152, + "grad_norm": 0.11138027906417847, + "learning_rate": 8.30668335341159e-05, + "loss": 0.0417, + "step": 30650 + }, + { + "epoch": 5.9523809523809526, + "grad_norm": 0.2930234372615814, + "learning_rate": 8.305442922774408e-05, + "loss": 0.0263, + "step": 30660 + }, + { + "epoch": 5.954322605698753, + "grad_norm": 0.21029067039489746, + "learning_rate": 8.304202130658959e-05, + "loss": 0.0312, + "step": 30670 + }, + { + "epoch": 5.956264259016552, + "grad_norm": 0.20713594555854797, + "learning_rate": 8.302960977200931e-05, + "loss": 0.0264, + "step": 30680 + }, + { + "epoch": 5.9582059123343525, + "grad_norm": 0.15724235773086548, + "learning_rate": 8.301719462536058e-05, + "loss": 0.0277, + "step": 30690 + }, + { + "epoch": 5.960147565652153, + "grad_norm": 0.21397794783115387, + "learning_rate": 8.300477586800108e-05, + "loss": 0.0337, + "step": 30700 + }, + { + "epoch": 5.962089218969953, + "grad_norm": 0.19234201312065125, + "learning_rate": 8.299235350128892e-05, + "loss": 0.0269, + "step": 30710 + }, + { + "epoch": 5.964030872287753, + "grad_norm": 0.1294204741716385, + "learning_rate": 8.297992752658258e-05, + "loss": 0.0268, + "step": 30720 + }, + { + "epoch": 5.965972525605553, + "grad_norm": 0.239422008395195, + "learning_rate": 8.296749794524094e-05, + "loss": 0.0336, + "step": 30730 + }, + { + "epoch": 5.967914178923353, + "grad_norm": 0.1870928406715393, + "learning_rate": 8.295506475862332e-05, + "loss": 0.03, + "step": 30740 + }, + { + "epoch": 5.969855832241153, + "grad_norm": 0.1314457803964615, + "learning_rate": 8.294262796808933e-05, + "loss": 0.027, + "step": 30750 + }, + { + "epoch": 5.9717974855589535, + "grad_norm": 0.1788250058889389, + "learning_rate": 8.293018757499909e-05, + "loss": 0.0283, + "step": 30760 + }, + { + "epoch": 5.973739138876754, + "grad_norm": 0.15645956993103027, + "learning_rate": 8.291774358071305e-05, + "loss": 0.0267, + "step": 30770 + }, + { + "epoch": 5.975680792194554, + "grad_norm": 0.20546039938926697, + "learning_rate": 8.290529598659205e-05, + "loss": 0.038, + "step": 30780 + }, + { + "epoch": 5.977622445512353, + "grad_norm": 0.16651326417922974, + "learning_rate": 8.289284479399738e-05, + "loss": 0.0365, + "step": 30790 + }, + { + "epoch": 5.979564098830154, + "grad_norm": 0.26560455560684204, + "learning_rate": 8.288039000429064e-05, + "loss": 0.0319, + "step": 30800 + }, + { + "epoch": 5.981505752147954, + "grad_norm": 0.18104857206344604, + "learning_rate": 8.286793161883388e-05, + "loss": 0.0343, + "step": 30810 + }, + { + "epoch": 5.983447405465754, + "grad_norm": 0.21850381791591644, + "learning_rate": 8.285546963898954e-05, + "loss": 0.0262, + "step": 30820 + }, + { + "epoch": 5.9853890587835545, + "grad_norm": 0.14695845544338226, + "learning_rate": 8.284300406612044e-05, + "loss": 0.0323, + "step": 30830 + }, + { + "epoch": 5.987330712101354, + "grad_norm": 0.20953962206840515, + "learning_rate": 8.283053490158978e-05, + "loss": 0.0316, + "step": 30840 + }, + { + "epoch": 5.989272365419154, + "grad_norm": 0.14373694360256195, + "learning_rate": 8.28180621467612e-05, + "loss": 0.0247, + "step": 30850 + }, + { + "epoch": 5.9912140187369545, + "grad_norm": 0.10612989962100983, + "learning_rate": 8.280558580299867e-05, + "loss": 0.0269, + "step": 30860 + }, + { + "epoch": 5.993155672054755, + "grad_norm": 0.170515239238739, + "learning_rate": 8.279310587166661e-05, + "loss": 0.027, + "step": 30870 + }, + { + "epoch": 5.995097325372555, + "grad_norm": 0.20614169538021088, + "learning_rate": 8.278062235412978e-05, + "loss": 0.0284, + "step": 30880 + }, + { + "epoch": 5.997038978690355, + "grad_norm": 0.19539621472358704, + "learning_rate": 8.276813525175339e-05, + "loss": 0.0316, + "step": 30890 + }, + { + "epoch": 5.998980632008155, + "grad_norm": 0.13394440710544586, + "learning_rate": 8.2755644565903e-05, + "loss": 0.0323, + "step": 30900 + }, + { + "epoch": 6.00077666132712, + "grad_norm": 0.27754369378089905, + "learning_rate": 8.274315029794454e-05, + "loss": 0.0253, + "step": 30910 + }, + { + "epoch": 6.002718314644921, + "grad_norm": 0.26431769132614136, + "learning_rate": 8.273065244924443e-05, + "loss": 0.0272, + "step": 30920 + }, + { + "epoch": 6.00465996796272, + "grad_norm": 0.16649189591407776, + "learning_rate": 8.271815102116936e-05, + "loss": 0.0254, + "step": 30930 + }, + { + "epoch": 6.00660162128052, + "grad_norm": 0.1983540952205658, + "learning_rate": 8.270564601508648e-05, + "loss": 0.0333, + "step": 30940 + }, + { + "epoch": 6.0085432745983205, + "grad_norm": 0.14748385548591614, + "learning_rate": 8.269313743236333e-05, + "loss": 0.0306, + "step": 30950 + }, + { + "epoch": 6.010484927916121, + "grad_norm": 0.2337389886379242, + "learning_rate": 8.268062527436783e-05, + "loss": 0.027, + "step": 30960 + }, + { + "epoch": 6.012426581233921, + "grad_norm": 0.19993793964385986, + "learning_rate": 8.266810954246828e-05, + "loss": 0.0347, + "step": 30970 + }, + { + "epoch": 6.01436823455172, + "grad_norm": 0.21594299376010895, + "learning_rate": 8.265559023803338e-05, + "loss": 0.0381, + "step": 30980 + }, + { + "epoch": 6.016309887869521, + "grad_norm": 0.16470736265182495, + "learning_rate": 8.264306736243223e-05, + "loss": 0.0281, + "step": 30990 + }, + { + "epoch": 6.018251541187321, + "grad_norm": 0.23877309262752533, + "learning_rate": 8.263054091703432e-05, + "loss": 0.0253, + "step": 31000 + }, + { + "epoch": 6.020193194505121, + "grad_norm": 0.2223202884197235, + "learning_rate": 8.26180109032095e-05, + "loss": 0.0271, + "step": 31010 + }, + { + "epoch": 6.0221348478229215, + "grad_norm": 0.14531370997428894, + "learning_rate": 8.260547732232807e-05, + "loss": 0.0273, + "step": 31020 + }, + { + "epoch": 6.024076501140721, + "grad_norm": 0.17427124083042145, + "learning_rate": 8.259294017576063e-05, + "loss": 0.0369, + "step": 31030 + }, + { + "epoch": 6.026018154458521, + "grad_norm": 0.24197496473789215, + "learning_rate": 8.258039946487828e-05, + "loss": 0.0288, + "step": 31040 + }, + { + "epoch": 6.0279598077763215, + "grad_norm": 0.325630784034729, + "learning_rate": 8.256785519105241e-05, + "loss": 0.0391, + "step": 31050 + }, + { + "epoch": 6.029901461094122, + "grad_norm": 0.20629684627056122, + "learning_rate": 8.255530735565488e-05, + "loss": 0.0394, + "step": 31060 + }, + { + "epoch": 6.031843114411922, + "grad_norm": 0.11327040195465088, + "learning_rate": 8.254275596005787e-05, + "loss": 0.0302, + "step": 31070 + }, + { + "epoch": 6.033784767729722, + "grad_norm": 0.18798109889030457, + "learning_rate": 8.2530201005634e-05, + "loss": 0.0324, + "step": 31080 + }, + { + "epoch": 6.035726421047522, + "grad_norm": 0.1338767260313034, + "learning_rate": 8.251764249375626e-05, + "loss": 0.0323, + "step": 31090 + }, + { + "epoch": 6.037668074365322, + "grad_norm": 0.21186771988868713, + "learning_rate": 8.250508042579803e-05, + "loss": 0.0271, + "step": 31100 + }, + { + "epoch": 6.039609727683122, + "grad_norm": 0.18070441484451294, + "learning_rate": 8.249251480313307e-05, + "loss": 0.0315, + "step": 31110 + }, + { + "epoch": 6.0415513810009225, + "grad_norm": 0.20550377666950226, + "learning_rate": 8.247994562713555e-05, + "loss": 0.0243, + "step": 31120 + }, + { + "epoch": 6.043493034318723, + "grad_norm": 0.2169986218214035, + "learning_rate": 8.246737289917998e-05, + "loss": 0.0299, + "step": 31130 + }, + { + "epoch": 6.045434687636522, + "grad_norm": 0.17071154713630676, + "learning_rate": 8.245479662064135e-05, + "loss": 0.0285, + "step": 31140 + }, + { + "epoch": 6.047376340954322, + "grad_norm": 0.19631625711917877, + "learning_rate": 8.244221679289496e-05, + "loss": 0.029, + "step": 31150 + }, + { + "epoch": 6.049317994272123, + "grad_norm": 0.16780290007591248, + "learning_rate": 8.242963341731652e-05, + "loss": 0.0349, + "step": 31160 + }, + { + "epoch": 6.051259647589923, + "grad_norm": 0.20480884611606598, + "learning_rate": 8.24170464952821e-05, + "loss": 0.0332, + "step": 31170 + }, + { + "epoch": 6.053201300907723, + "grad_norm": 0.184340700507164, + "learning_rate": 8.240445602816824e-05, + "loss": 0.0293, + "step": 31180 + }, + { + "epoch": 6.055142954225523, + "grad_norm": 0.1577940136194229, + "learning_rate": 8.239186201735179e-05, + "loss": 0.0276, + "step": 31190 + }, + { + "epoch": 6.057084607543323, + "grad_norm": 0.21619661152362823, + "learning_rate": 8.237926446420998e-05, + "loss": 0.0279, + "step": 31200 + }, + { + "epoch": 6.059026260861123, + "grad_norm": 0.12062963098287582, + "learning_rate": 8.236666337012052e-05, + "loss": 0.0221, + "step": 31210 + }, + { + "epoch": 6.060967914178923, + "grad_norm": 0.30584481358528137, + "learning_rate": 8.23540587364614e-05, + "loss": 0.0343, + "step": 31220 + }, + { + "epoch": 6.062909567496724, + "grad_norm": 0.1622442901134491, + "learning_rate": 8.234145056461107e-05, + "loss": 0.0302, + "step": 31230 + }, + { + "epoch": 6.064851220814524, + "grad_norm": 0.17570936679840088, + "learning_rate": 8.232883885594832e-05, + "loss": 0.0303, + "step": 31240 + }, + { + "epoch": 6.066792874132323, + "grad_norm": 0.1595296561717987, + "learning_rate": 8.231622361185236e-05, + "loss": 0.0256, + "step": 31250 + }, + { + "epoch": 6.068734527450124, + "grad_norm": 0.14751383662223816, + "learning_rate": 8.230360483370278e-05, + "loss": 0.0256, + "step": 31260 + }, + { + "epoch": 6.070676180767924, + "grad_norm": 0.10396761447191238, + "learning_rate": 8.229098252287953e-05, + "loss": 0.0255, + "step": 31270 + }, + { + "epoch": 6.072617834085724, + "grad_norm": 0.12255352735519409, + "learning_rate": 8.2278356680763e-05, + "loss": 0.0305, + "step": 31280 + }, + { + "epoch": 6.0745594874035245, + "grad_norm": 0.1320217251777649, + "learning_rate": 8.22657273087339e-05, + "loss": 0.0272, + "step": 31290 + }, + { + "epoch": 6.076501140721324, + "grad_norm": 0.15606336295604706, + "learning_rate": 8.225309440817336e-05, + "loss": 0.0377, + "step": 31300 + }, + { + "epoch": 6.078442794039124, + "grad_norm": 0.17228876054286957, + "learning_rate": 8.224045798046293e-05, + "loss": 0.0272, + "step": 31310 + }, + { + "epoch": 6.080384447356924, + "grad_norm": 0.13043789565563202, + "learning_rate": 8.22278180269845e-05, + "loss": 0.0291, + "step": 31320 + }, + { + "epoch": 6.082326100674725, + "grad_norm": 0.11990831792354584, + "learning_rate": 8.221517454912031e-05, + "loss": 0.0351, + "step": 31330 + }, + { + "epoch": 6.084267753992525, + "grad_norm": 0.12492592632770538, + "learning_rate": 8.220252754825308e-05, + "loss": 0.0334, + "step": 31340 + }, + { + "epoch": 6.086209407310324, + "grad_norm": 0.1325184553861618, + "learning_rate": 8.218987702576586e-05, + "loss": 0.0317, + "step": 31350 + }, + { + "epoch": 6.088151060628125, + "grad_norm": 0.08860103040933609, + "learning_rate": 8.217722298304207e-05, + "loss": 0.0271, + "step": 31360 + }, + { + "epoch": 6.090092713945925, + "grad_norm": 0.09150291234254837, + "learning_rate": 8.216456542146557e-05, + "loss": 0.0262, + "step": 31370 + }, + { + "epoch": 6.092034367263725, + "grad_norm": 0.11652892827987671, + "learning_rate": 8.215190434242055e-05, + "loss": 0.0295, + "step": 31380 + }, + { + "epoch": 6.093976020581525, + "grad_norm": 0.15858979523181915, + "learning_rate": 8.213923974729161e-05, + "loss": 0.0316, + "step": 31390 + }, + { + "epoch": 6.095917673899326, + "grad_norm": 0.15216407179832458, + "learning_rate": 8.212657163746373e-05, + "loss": 0.0288, + "step": 31400 + }, + { + "epoch": 6.097859327217125, + "grad_norm": 0.2531457245349884, + "learning_rate": 8.211390001432227e-05, + "loss": 0.0361, + "step": 31410 + }, + { + "epoch": 6.099800980534925, + "grad_norm": 0.11525558680295944, + "learning_rate": 8.210122487925297e-05, + "loss": 0.0281, + "step": 31420 + }, + { + "epoch": 6.101742633852726, + "grad_norm": 0.1261812001466751, + "learning_rate": 8.208854623364202e-05, + "loss": 0.0262, + "step": 31430 + }, + { + "epoch": 6.103684287170526, + "grad_norm": 0.13592740893363953, + "learning_rate": 8.207586407887589e-05, + "loss": 0.0336, + "step": 31440 + }, + { + "epoch": 6.105625940488326, + "grad_norm": 0.19082209467887878, + "learning_rate": 8.206317841634148e-05, + "loss": 0.0316, + "step": 31450 + }, + { + "epoch": 6.107567593806126, + "grad_norm": 0.11300941556692123, + "learning_rate": 8.205048924742609e-05, + "loss": 0.0277, + "step": 31460 + }, + { + "epoch": 6.109509247123926, + "grad_norm": 0.15280728042125702, + "learning_rate": 8.203779657351738e-05, + "loss": 0.0351, + "step": 31470 + }, + { + "epoch": 6.111450900441726, + "grad_norm": 0.13281890749931335, + "learning_rate": 8.20251003960034e-05, + "loss": 0.0276, + "step": 31480 + }, + { + "epoch": 6.113392553759526, + "grad_norm": 0.13500376045703888, + "learning_rate": 8.201240071627258e-05, + "loss": 0.0288, + "step": 31490 + }, + { + "epoch": 6.115334207077327, + "grad_norm": 0.1973990947008133, + "learning_rate": 8.199969753571377e-05, + "loss": 0.03, + "step": 31500 + }, + { + "epoch": 6.117275860395126, + "grad_norm": 0.17553424835205078, + "learning_rate": 8.198699085571615e-05, + "loss": 0.0283, + "step": 31510 + }, + { + "epoch": 6.119217513712926, + "grad_norm": 0.13929085433483124, + "learning_rate": 8.197428067766928e-05, + "loss": 0.0247, + "step": 31520 + }, + { + "epoch": 6.121159167030727, + "grad_norm": 0.20957253873348236, + "learning_rate": 8.196156700296316e-05, + "loss": 0.0341, + "step": 31530 + }, + { + "epoch": 6.123100820348527, + "grad_norm": 0.1994895339012146, + "learning_rate": 8.194884983298814e-05, + "loss": 0.0323, + "step": 31540 + }, + { + "epoch": 6.125042473666327, + "grad_norm": 0.3761920630931854, + "learning_rate": 8.193612916913491e-05, + "loss": 0.0311, + "step": 31550 + }, + { + "epoch": 6.1269841269841265, + "grad_norm": 0.29855233430862427, + "learning_rate": 8.192340501279463e-05, + "loss": 0.0343, + "step": 31560 + }, + { + "epoch": 6.128925780301927, + "grad_norm": 0.20374301075935364, + "learning_rate": 8.191067736535876e-05, + "loss": 0.0345, + "step": 31570 + }, + { + "epoch": 6.130867433619727, + "grad_norm": 0.1643836945295334, + "learning_rate": 8.18979462282192e-05, + "loss": 0.0309, + "step": 31580 + }, + { + "epoch": 6.132809086937527, + "grad_norm": 0.11623607575893402, + "learning_rate": 8.188521160276819e-05, + "loss": 0.0332, + "step": 31590 + }, + { + "epoch": 6.134750740255328, + "grad_norm": 0.1643979847431183, + "learning_rate": 8.187247349039837e-05, + "loss": 0.0305, + "step": 31600 + }, + { + "epoch": 6.136692393573128, + "grad_norm": 0.19269362092018127, + "learning_rate": 8.185973189250278e-05, + "loss": 0.0348, + "step": 31610 + }, + { + "epoch": 6.138634046890927, + "grad_norm": 0.13622435927391052, + "learning_rate": 8.184698681047482e-05, + "loss": 0.0337, + "step": 31620 + }, + { + "epoch": 6.140575700208728, + "grad_norm": 0.18557657301425934, + "learning_rate": 8.183423824570827e-05, + "loss": 0.0334, + "step": 31630 + }, + { + "epoch": 6.142517353526528, + "grad_norm": 0.13867342472076416, + "learning_rate": 8.182148619959725e-05, + "loss": 0.0375, + "step": 31640 + }, + { + "epoch": 6.144459006844328, + "grad_norm": 0.10930222272872925, + "learning_rate": 8.180873067353636e-05, + "loss": 0.0286, + "step": 31650 + }, + { + "epoch": 6.146400660162128, + "grad_norm": 0.2164786159992218, + "learning_rate": 8.179597166892052e-05, + "loss": 0.0352, + "step": 31660 + }, + { + "epoch": 6.148342313479928, + "grad_norm": 0.15528631210327148, + "learning_rate": 8.178320918714501e-05, + "loss": 0.0289, + "step": 31670 + }, + { + "epoch": 6.150283966797728, + "grad_norm": 0.22149862349033356, + "learning_rate": 8.177044322960554e-05, + "loss": 0.0288, + "step": 31680 + }, + { + "epoch": 6.152225620115528, + "grad_norm": 0.1895286589860916, + "learning_rate": 8.175767379769816e-05, + "loss": 0.0348, + "step": 31690 + }, + { + "epoch": 6.154167273433329, + "grad_norm": 0.1691332310438156, + "learning_rate": 8.174490089281932e-05, + "loss": 0.0262, + "step": 31700 + }, + { + "epoch": 6.156108926751129, + "grad_norm": 0.2664792537689209, + "learning_rate": 8.173212451636584e-05, + "loss": 0.0273, + "step": 31710 + }, + { + "epoch": 6.158050580068929, + "grad_norm": 0.2534017562866211, + "learning_rate": 8.171934466973493e-05, + "loss": 0.0343, + "step": 31720 + }, + { + "epoch": 6.1599922333867285, + "grad_norm": 0.1419450342655182, + "learning_rate": 8.170656135432418e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 6.161933886704529, + "grad_norm": 0.18167045712471008, + "learning_rate": 8.169377457153155e-05, + "loss": 0.0326, + "step": 31740 + }, + { + "epoch": 6.163875540022329, + "grad_norm": 0.12594656646251678, + "learning_rate": 8.168098432275539e-05, + "loss": 0.0316, + "step": 31750 + }, + { + "epoch": 6.165817193340129, + "grad_norm": 0.23707951605319977, + "learning_rate": 8.166819060939442e-05, + "loss": 0.0359, + "step": 31760 + }, + { + "epoch": 6.16775884665793, + "grad_norm": 0.14845019578933716, + "learning_rate": 8.165539343284772e-05, + "loss": 0.0348, + "step": 31770 + }, + { + "epoch": 6.169700499975729, + "grad_norm": 0.17741484940052032, + "learning_rate": 8.16425927945148e-05, + "loss": 0.0228, + "step": 31780 + }, + { + "epoch": 6.171642153293529, + "grad_norm": 0.16811901330947876, + "learning_rate": 8.162978869579551e-05, + "loss": 0.0362, + "step": 31790 + }, + { + "epoch": 6.1735838066113295, + "grad_norm": 0.23125775158405304, + "learning_rate": 8.161698113809007e-05, + "loss": 0.0307, + "step": 31800 + }, + { + "epoch": 6.17552545992913, + "grad_norm": 0.18805928528308868, + "learning_rate": 8.160417012279911e-05, + "loss": 0.0305, + "step": 31810 + }, + { + "epoch": 6.17746711324693, + "grad_norm": 0.1630890816450119, + "learning_rate": 8.159135565132363e-05, + "loss": 0.0349, + "step": 31820 + }, + { + "epoch": 6.1794087665647295, + "grad_norm": 0.14951983094215393, + "learning_rate": 8.157853772506498e-05, + "loss": 0.0266, + "step": 31830 + }, + { + "epoch": 6.18135041988253, + "grad_norm": 0.1873738318681717, + "learning_rate": 8.156571634542494e-05, + "loss": 0.0295, + "step": 31840 + }, + { + "epoch": 6.18329207320033, + "grad_norm": 0.2027851939201355, + "learning_rate": 8.15528915138056e-05, + "loss": 0.0377, + "step": 31850 + }, + { + "epoch": 6.18523372651813, + "grad_norm": 0.1204376369714737, + "learning_rate": 8.154006323160949e-05, + "loss": 0.0447, + "step": 31860 + }, + { + "epoch": 6.187175379835931, + "grad_norm": 0.13397054374217987, + "learning_rate": 8.152723150023949e-05, + "loss": 0.0271, + "step": 31870 + }, + { + "epoch": 6.18911703315373, + "grad_norm": 0.27990373969078064, + "learning_rate": 8.151439632109886e-05, + "loss": 0.0259, + "step": 31880 + }, + { + "epoch": 6.19105868647153, + "grad_norm": 0.1554562747478485, + "learning_rate": 8.150155769559122e-05, + "loss": 0.0274, + "step": 31890 + }, + { + "epoch": 6.1930003397893305, + "grad_norm": 0.24541467428207397, + "learning_rate": 8.148871562512058e-05, + "loss": 0.0371, + "step": 31900 + }, + { + "epoch": 6.194941993107131, + "grad_norm": 0.2720147371292114, + "learning_rate": 8.147587011109136e-05, + "loss": 0.0371, + "step": 31910 + }, + { + "epoch": 6.196883646424931, + "grad_norm": 0.1532709300518036, + "learning_rate": 8.14630211549083e-05, + "loss": 0.033, + "step": 31920 + }, + { + "epoch": 6.198825299742731, + "grad_norm": 0.166277214884758, + "learning_rate": 8.145016875797655e-05, + "loss": 0.0284, + "step": 31930 + }, + { + "epoch": 6.200766953060531, + "grad_norm": 0.2078407257795334, + "learning_rate": 8.143731292170164e-05, + "loss": 0.0325, + "step": 31940 + }, + { + "epoch": 6.202708606378331, + "grad_norm": 0.16804270446300507, + "learning_rate": 8.142445364748944e-05, + "loss": 0.0291, + "step": 31950 + }, + { + "epoch": 6.204650259696131, + "grad_norm": 0.21785643696784973, + "learning_rate": 8.141159093674624e-05, + "loss": 0.0273, + "step": 31960 + }, + { + "epoch": 6.2065919130139315, + "grad_norm": 0.37529805302619934, + "learning_rate": 8.139872479087869e-05, + "loss": 0.034, + "step": 31970 + }, + { + "epoch": 6.208533566331732, + "grad_norm": 0.20243875682353973, + "learning_rate": 8.13858552112938e-05, + "loss": 0.0269, + "step": 31980 + }, + { + "epoch": 6.210475219649531, + "grad_norm": 0.13753725588321686, + "learning_rate": 8.137298219939895e-05, + "loss": 0.0274, + "step": 31990 + }, + { + "epoch": 6.2124168729673315, + "grad_norm": 0.15889795124530792, + "learning_rate": 8.136010575660196e-05, + "loss": 0.0338, + "step": 32000 + }, + { + "epoch": 6.214358526285132, + "grad_norm": 0.12413620203733444, + "learning_rate": 8.134722588431095e-05, + "loss": 0.0274, + "step": 32010 + }, + { + "epoch": 6.216300179602932, + "grad_norm": 0.21593518555164337, + "learning_rate": 8.133434258393444e-05, + "loss": 0.0281, + "step": 32020 + }, + { + "epoch": 6.218241832920732, + "grad_norm": 0.14273206889629364, + "learning_rate": 8.132145585688134e-05, + "loss": 0.0249, + "step": 32030 + }, + { + "epoch": 6.220183486238533, + "grad_norm": 0.16121837496757507, + "learning_rate": 8.130856570456093e-05, + "loss": 0.0223, + "step": 32040 + }, + { + "epoch": 6.222125139556332, + "grad_norm": 0.1817539483308792, + "learning_rate": 8.129567212838283e-05, + "loss": 0.0295, + "step": 32050 + }, + { + "epoch": 6.224066792874132, + "grad_norm": 0.14318035542964935, + "learning_rate": 8.128277512975708e-05, + "loss": 0.0277, + "step": 32060 + }, + { + "epoch": 6.2260084461919325, + "grad_norm": 0.21533139050006866, + "learning_rate": 8.126987471009408e-05, + "loss": 0.0319, + "step": 32070 + }, + { + "epoch": 6.227950099509733, + "grad_norm": 0.07689645141363144, + "learning_rate": 8.125697087080459e-05, + "loss": 0.0243, + "step": 32080 + }, + { + "epoch": 6.229891752827533, + "grad_norm": 0.10231161117553711, + "learning_rate": 8.124406361329976e-05, + "loss": 0.0278, + "step": 32090 + }, + { + "epoch": 6.231833406145332, + "grad_norm": 0.23206380009651184, + "learning_rate": 8.123115293899111e-05, + "loss": 0.0334, + "step": 32100 + }, + { + "epoch": 6.233775059463133, + "grad_norm": 0.14483003318309784, + "learning_rate": 8.121823884929055e-05, + "loss": 0.0296, + "step": 32110 + }, + { + "epoch": 6.235716712780933, + "grad_norm": 0.15290692448616028, + "learning_rate": 8.12053213456103e-05, + "loss": 0.0249, + "step": 32120 + }, + { + "epoch": 6.237658366098733, + "grad_norm": 0.17629346251487732, + "learning_rate": 8.119240042936303e-05, + "loss": 0.0252, + "step": 32130 + }, + { + "epoch": 6.2396000194165335, + "grad_norm": 0.14510907232761383, + "learning_rate": 8.117947610196175e-05, + "loss": 0.0304, + "step": 32140 + }, + { + "epoch": 6.241541672734333, + "grad_norm": 0.22342173755168915, + "learning_rate": 8.116654836481982e-05, + "loss": 0.0342, + "step": 32150 + }, + { + "epoch": 6.243483326052133, + "grad_norm": 0.12980760633945465, + "learning_rate": 8.115361721935106e-05, + "loss": 0.0347, + "step": 32160 + }, + { + "epoch": 6.2454249793699335, + "grad_norm": 0.29117944836616516, + "learning_rate": 8.114068266696953e-05, + "loss": 0.036, + "step": 32170 + }, + { + "epoch": 6.247366632687734, + "grad_norm": 0.14421656727790833, + "learning_rate": 8.112774470908978e-05, + "loss": 0.0285, + "step": 32180 + }, + { + "epoch": 6.249308286005534, + "grad_norm": 0.1940900683403015, + "learning_rate": 8.111480334712665e-05, + "loss": 0.03, + "step": 32190 + }, + { + "epoch": 6.251249939323333, + "grad_norm": 0.1250244677066803, + "learning_rate": 8.110185858249542e-05, + "loss": 0.0258, + "step": 32200 + }, + { + "epoch": 6.253191592641134, + "grad_norm": 0.23541562259197235, + "learning_rate": 8.108891041661168e-05, + "loss": 0.0288, + "step": 32210 + }, + { + "epoch": 6.255133245958934, + "grad_norm": 0.17143301665782928, + "learning_rate": 8.107595885089146e-05, + "loss": 0.0378, + "step": 32220 + }, + { + "epoch": 6.257074899276734, + "grad_norm": 0.14015652239322662, + "learning_rate": 8.106300388675108e-05, + "loss": 0.0275, + "step": 32230 + }, + { + "epoch": 6.2590165525945345, + "grad_norm": 0.22341930866241455, + "learning_rate": 8.105004552560731e-05, + "loss": 0.0327, + "step": 32240 + }, + { + "epoch": 6.260958205912335, + "grad_norm": 0.19440093636512756, + "learning_rate": 8.103708376887724e-05, + "loss": 0.0305, + "step": 32250 + }, + { + "epoch": 6.262899859230134, + "grad_norm": 0.13034746050834656, + "learning_rate": 8.102411861797836e-05, + "loss": 0.0303, + "step": 32260 + }, + { + "epoch": 6.264841512547934, + "grad_norm": 0.16110947728157043, + "learning_rate": 8.101115007432851e-05, + "loss": 0.0279, + "step": 32270 + }, + { + "epoch": 6.266783165865735, + "grad_norm": 0.1975105255842209, + "learning_rate": 8.09981781393459e-05, + "loss": 0.0357, + "step": 32280 + }, + { + "epoch": 6.268724819183535, + "grad_norm": 0.2287583351135254, + "learning_rate": 8.098520281444915e-05, + "loss": 0.035, + "step": 32290 + }, + { + "epoch": 6.270666472501335, + "grad_norm": 0.11917458474636078, + "learning_rate": 8.09722241010572e-05, + "loss": 0.0306, + "step": 32300 + }, + { + "epoch": 6.272608125819135, + "grad_norm": 0.13476994633674622, + "learning_rate": 8.095924200058939e-05, + "loss": 0.0305, + "step": 32310 + }, + { + "epoch": 6.274549779136935, + "grad_norm": 0.1651223599910736, + "learning_rate": 8.094625651446541e-05, + "loss": 0.0328, + "step": 32320 + }, + { + "epoch": 6.276491432454735, + "grad_norm": 0.15504945814609528, + "learning_rate": 8.093326764410536e-05, + "loss": 0.0259, + "step": 32330 + }, + { + "epoch": 6.2784330857725354, + "grad_norm": 0.11294306069612503, + "learning_rate": 8.092027539092966e-05, + "loss": 0.0261, + "step": 32340 + }, + { + "epoch": 6.280374739090336, + "grad_norm": 0.2394629716873169, + "learning_rate": 8.090727975635913e-05, + "loss": 0.0302, + "step": 32350 + }, + { + "epoch": 6.282316392408136, + "grad_norm": 0.15555822849273682, + "learning_rate": 8.089428074181497e-05, + "loss": 0.0383, + "step": 32360 + }, + { + "epoch": 6.284258045725935, + "grad_norm": 0.18920578062534332, + "learning_rate": 8.088127834871871e-05, + "loss": 0.0268, + "step": 32370 + }, + { + "epoch": 6.286199699043736, + "grad_norm": 0.18678215146064758, + "learning_rate": 8.086827257849226e-05, + "loss": 0.0321, + "step": 32380 + }, + { + "epoch": 6.288141352361536, + "grad_norm": 0.16873617470264435, + "learning_rate": 8.085526343255795e-05, + "loss": 0.034, + "step": 32390 + }, + { + "epoch": 6.290083005679336, + "grad_norm": 0.2476726919412613, + "learning_rate": 8.084225091233842e-05, + "loss": 0.0373, + "step": 32400 + }, + { + "epoch": 6.2920246589971365, + "grad_norm": 0.17276233434677124, + "learning_rate": 8.082923501925668e-05, + "loss": 0.0244, + "step": 32410 + }, + { + "epoch": 6.293966312314936, + "grad_norm": 0.13620609045028687, + "learning_rate": 8.081621575473617e-05, + "loss": 0.0268, + "step": 32420 + }, + { + "epoch": 6.295907965632736, + "grad_norm": 0.17670249938964844, + "learning_rate": 8.080319312020064e-05, + "loss": 0.0276, + "step": 32430 + }, + { + "epoch": 6.297849618950536, + "grad_norm": 0.15706805884838104, + "learning_rate": 8.079016711707421e-05, + "loss": 0.0311, + "step": 32440 + }, + { + "epoch": 6.299791272268337, + "grad_norm": 0.17171500623226166, + "learning_rate": 8.077713774678139e-05, + "loss": 0.0258, + "step": 32450 + }, + { + "epoch": 6.301732925586137, + "grad_norm": 0.16095660626888275, + "learning_rate": 8.076410501074707e-05, + "loss": 0.0276, + "step": 32460 + }, + { + "epoch": 6.303674578903936, + "grad_norm": 0.1516689658164978, + "learning_rate": 8.075106891039647e-05, + "loss": 0.0338, + "step": 32470 + }, + { + "epoch": 6.305616232221737, + "grad_norm": 0.20984821021556854, + "learning_rate": 8.073802944715523e-05, + "loss": 0.0256, + "step": 32480 + }, + { + "epoch": 6.307557885539537, + "grad_norm": 0.18925316631793976, + "learning_rate": 8.072498662244929e-05, + "loss": 0.0351, + "step": 32490 + }, + { + "epoch": 6.309499538857337, + "grad_norm": 0.16658446192741394, + "learning_rate": 8.0711940437705e-05, + "loss": 0.0264, + "step": 32500 + }, + { + "epoch": 6.311441192175137, + "grad_norm": 0.23851032555103302, + "learning_rate": 8.06988908943491e-05, + "loss": 0.0338, + "step": 32510 + }, + { + "epoch": 6.313382845492937, + "grad_norm": 0.16937533020973206, + "learning_rate": 8.068583799380863e-05, + "loss": 0.0339, + "step": 32520 + }, + { + "epoch": 6.315324498810737, + "grad_norm": 0.18878456950187683, + "learning_rate": 8.067278173751104e-05, + "loss": 0.0354, + "step": 32530 + }, + { + "epoch": 6.317266152128537, + "grad_norm": 0.1789567470550537, + "learning_rate": 8.065972212688417e-05, + "loss": 0.0362, + "step": 32540 + }, + { + "epoch": 6.319207805446338, + "grad_norm": 0.16601639986038208, + "learning_rate": 8.064665916335618e-05, + "loss": 0.0231, + "step": 32550 + }, + { + "epoch": 6.321149458764138, + "grad_norm": 0.17182500660419464, + "learning_rate": 8.063359284835564e-05, + "loss": 0.0245, + "step": 32560 + }, + { + "epoch": 6.323091112081938, + "grad_norm": 0.10585848987102509, + "learning_rate": 8.062052318331142e-05, + "loss": 0.036, + "step": 32570 + }, + { + "epoch": 6.325032765399738, + "grad_norm": 0.22327575087547302, + "learning_rate": 8.060745016965283e-05, + "loss": 0.0302, + "step": 32580 + }, + { + "epoch": 6.326974418717538, + "grad_norm": 0.17499157786369324, + "learning_rate": 8.059437380880952e-05, + "loss": 0.0251, + "step": 32590 + }, + { + "epoch": 6.328916072035338, + "grad_norm": 0.1785002201795578, + "learning_rate": 8.058129410221146e-05, + "loss": 0.028, + "step": 32600 + }, + { + "epoch": 6.330857725353138, + "grad_norm": 0.15380114316940308, + "learning_rate": 8.056821105128908e-05, + "loss": 0.0297, + "step": 32610 + }, + { + "epoch": 6.332799378670939, + "grad_norm": 0.14792807400226593, + "learning_rate": 8.05551246574731e-05, + "loss": 0.0306, + "step": 32620 + }, + { + "epoch": 6.334741031988738, + "grad_norm": 0.1413191705942154, + "learning_rate": 8.05420349221946e-05, + "loss": 0.0288, + "step": 32630 + }, + { + "epoch": 6.336682685306538, + "grad_norm": 0.14221321046352386, + "learning_rate": 8.05289418468851e-05, + "loss": 0.0291, + "step": 32640 + }, + { + "epoch": 6.338624338624339, + "grad_norm": 0.17333932220935822, + "learning_rate": 8.051584543297642e-05, + "loss": 0.0229, + "step": 32650 + }, + { + "epoch": 6.340565991942139, + "grad_norm": 0.16483387351036072, + "learning_rate": 8.050274568190074e-05, + "loss": 0.0362, + "step": 32660 + }, + { + "epoch": 6.342507645259939, + "grad_norm": 0.19792625308036804, + "learning_rate": 8.048964259509067e-05, + "loss": 0.0325, + "step": 32670 + }, + { + "epoch": 6.3444492985777385, + "grad_norm": 0.15080693364143372, + "learning_rate": 8.047653617397914e-05, + "loss": 0.0317, + "step": 32680 + }, + { + "epoch": 6.346390951895539, + "grad_norm": 0.17799994349479675, + "learning_rate": 8.046342641999941e-05, + "loss": 0.0238, + "step": 32690 + }, + { + "epoch": 6.348332605213339, + "grad_norm": 0.15744256973266602, + "learning_rate": 8.045031333458517e-05, + "loss": 0.0275, + "step": 32700 + }, + { + "epoch": 6.350274258531139, + "grad_norm": 0.23370012640953064, + "learning_rate": 8.043719691917047e-05, + "loss": 0.0301, + "step": 32710 + }, + { + "epoch": 6.35221591184894, + "grad_norm": 0.1410970240831375, + "learning_rate": 8.042407717518966e-05, + "loss": 0.0269, + "step": 32720 + }, + { + "epoch": 6.354157565166739, + "grad_norm": 0.1591562032699585, + "learning_rate": 8.041095410407751e-05, + "loss": 0.031, + "step": 32730 + }, + { + "epoch": 6.356099218484539, + "grad_norm": 0.15446920692920685, + "learning_rate": 8.039782770726913e-05, + "loss": 0.0302, + "step": 32740 + }, + { + "epoch": 6.35804087180234, + "grad_norm": 0.1541350930929184, + "learning_rate": 8.038469798620004e-05, + "loss": 0.0274, + "step": 32750 + }, + { + "epoch": 6.35998252512014, + "grad_norm": 0.21505752205848694, + "learning_rate": 8.037156494230604e-05, + "loss": 0.0328, + "step": 32760 + }, + { + "epoch": 6.36192417843794, + "grad_norm": 0.2255742996931076, + "learning_rate": 8.035842857702338e-05, + "loss": 0.0341, + "step": 32770 + }, + { + "epoch": 6.36386583175574, + "grad_norm": 0.1617579609155655, + "learning_rate": 8.03452888917886e-05, + "loss": 0.0246, + "step": 32780 + }, + { + "epoch": 6.36580748507354, + "grad_norm": 0.13210931420326233, + "learning_rate": 8.033214588803866e-05, + "loss": 0.0305, + "step": 32790 + }, + { + "epoch": 6.36774913839134, + "grad_norm": 0.24119700491428375, + "learning_rate": 8.031899956721083e-05, + "loss": 0.0304, + "step": 32800 + }, + { + "epoch": 6.36969079170914, + "grad_norm": 0.22313609719276428, + "learning_rate": 8.030584993074282e-05, + "loss": 0.0307, + "step": 32810 + }, + { + "epoch": 6.371632445026941, + "grad_norm": 0.28163382411003113, + "learning_rate": 8.02926969800726e-05, + "loss": 0.0391, + "step": 32820 + }, + { + "epoch": 6.373574098344741, + "grad_norm": 0.13485251367092133, + "learning_rate": 8.027954071663859e-05, + "loss": 0.0242, + "step": 32830 + }, + { + "epoch": 6.37551575166254, + "grad_norm": 0.17582671344280243, + "learning_rate": 8.026638114187954e-05, + "loss": 0.0258, + "step": 32840 + }, + { + "epoch": 6.3774574049803405, + "grad_norm": 0.17687197029590607, + "learning_rate": 8.025321825723456e-05, + "loss": 0.0261, + "step": 32850 + }, + { + "epoch": 6.379399058298141, + "grad_norm": 0.15874193608760834, + "learning_rate": 8.02400520641431e-05, + "loss": 0.0264, + "step": 32860 + }, + { + "epoch": 6.381340711615941, + "grad_norm": 0.219004824757576, + "learning_rate": 8.022688256404501e-05, + "loss": 0.028, + "step": 32870 + }, + { + "epoch": 6.383282364933741, + "grad_norm": 0.15464861690998077, + "learning_rate": 8.02137097583805e-05, + "loss": 0.0341, + "step": 32880 + }, + { + "epoch": 6.385224018251542, + "grad_norm": 0.1583421528339386, + "learning_rate": 8.02005336485901e-05, + "loss": 0.029, + "step": 32890 + }, + { + "epoch": 6.387165671569341, + "grad_norm": 0.18545331060886383, + "learning_rate": 8.018735423611476e-05, + "loss": 0.0296, + "step": 32900 + }, + { + "epoch": 6.389107324887141, + "grad_norm": 0.08875702321529388, + "learning_rate": 8.017417152239574e-05, + "loss": 0.0225, + "step": 32910 + }, + { + "epoch": 6.3910489782049416, + "grad_norm": 0.17889361083507538, + "learning_rate": 8.01609855088747e-05, + "loss": 0.0293, + "step": 32920 + }, + { + "epoch": 6.392990631522742, + "grad_norm": 0.2081173211336136, + "learning_rate": 8.014779619699362e-05, + "loss": 0.0321, + "step": 32930 + }, + { + "epoch": 6.394932284840542, + "grad_norm": 0.25298935174942017, + "learning_rate": 8.013460358819489e-05, + "loss": 0.0266, + "step": 32940 + }, + { + "epoch": 6.3968739381583415, + "grad_norm": 0.22723980247974396, + "learning_rate": 8.01214076839212e-05, + "loss": 0.0358, + "step": 32950 + }, + { + "epoch": 6.398815591476142, + "grad_norm": 0.24276690185070038, + "learning_rate": 8.010820848561565e-05, + "loss": 0.0379, + "step": 32960 + }, + { + "epoch": 6.400757244793942, + "grad_norm": 0.14551974833011627, + "learning_rate": 8.009500599472171e-05, + "loss": 0.0284, + "step": 32970 + }, + { + "epoch": 6.402698898111742, + "grad_norm": 0.13303543627262115, + "learning_rate": 8.008180021268314e-05, + "loss": 0.0337, + "step": 32980 + }, + { + "epoch": 6.404640551429543, + "grad_norm": 0.1537947803735733, + "learning_rate": 8.006859114094414e-05, + "loss": 0.0327, + "step": 32990 + }, + { + "epoch": 6.406582204747342, + "grad_norm": 0.15053853392601013, + "learning_rate": 8.005537878094921e-05, + "loss": 0.0307, + "step": 33000 + }, + { + "epoch": 6.408523858065142, + "grad_norm": 0.1682431548833847, + "learning_rate": 8.004216313414323e-05, + "loss": 0.0316, + "step": 33010 + }, + { + "epoch": 6.4104655113829425, + "grad_norm": 0.20535153150558472, + "learning_rate": 8.002894420197149e-05, + "loss": 0.025, + "step": 33020 + }, + { + "epoch": 6.412407164700743, + "grad_norm": 0.2208167314529419, + "learning_rate": 8.001572198587954e-05, + "loss": 0.0359, + "step": 33030 + }, + { + "epoch": 6.414348818018543, + "grad_norm": 0.14209890365600586, + "learning_rate": 8.000249648731338e-05, + "loss": 0.0218, + "step": 33040 + }, + { + "epoch": 6.4162904713363424, + "grad_norm": 0.3042738735675812, + "learning_rate": 7.998926770771928e-05, + "loss": 0.0314, + "step": 33050 + }, + { + "epoch": 6.418232124654143, + "grad_norm": 0.15514405071735382, + "learning_rate": 7.997603564854397e-05, + "loss": 0.0317, + "step": 33060 + }, + { + "epoch": 6.420173777971943, + "grad_norm": 0.0869312658905983, + "learning_rate": 7.996280031123448e-05, + "loss": 0.0223, + "step": 33070 + }, + { + "epoch": 6.422115431289743, + "grad_norm": 0.16167402267456055, + "learning_rate": 7.994956169723818e-05, + "loss": 0.0299, + "step": 33080 + }, + { + "epoch": 6.4240570846075435, + "grad_norm": 0.1971796154975891, + "learning_rate": 7.993631980800285e-05, + "loss": 0.0277, + "step": 33090 + }, + { + "epoch": 6.425998737925344, + "grad_norm": 0.13995404541492462, + "learning_rate": 7.992307464497659e-05, + "loss": 0.034, + "step": 33100 + }, + { + "epoch": 6.427940391243143, + "grad_norm": 0.231758713722229, + "learning_rate": 7.990982620960787e-05, + "loss": 0.0286, + "step": 33110 + }, + { + "epoch": 6.4298820445609435, + "grad_norm": 0.27954933047294617, + "learning_rate": 7.989657450334554e-05, + "loss": 0.0293, + "step": 33120 + }, + { + "epoch": 6.431823697878744, + "grad_norm": 0.27827635407447815, + "learning_rate": 7.988331952763877e-05, + "loss": 0.0244, + "step": 33130 + }, + { + "epoch": 6.433765351196544, + "grad_norm": 0.250764936208725, + "learning_rate": 7.987006128393709e-05, + "loss": 0.0335, + "step": 33140 + }, + { + "epoch": 6.435707004514344, + "grad_norm": 0.14790579676628113, + "learning_rate": 7.985679977369043e-05, + "loss": 0.032, + "step": 33150 + }, + { + "epoch": 6.437648657832144, + "grad_norm": 0.1678600013256073, + "learning_rate": 7.984353499834902e-05, + "loss": 0.0313, + "step": 33160 + }, + { + "epoch": 6.439590311149944, + "grad_norm": 0.21519596874713898, + "learning_rate": 7.983026695936351e-05, + "loss": 0.0349, + "step": 33170 + }, + { + "epoch": 6.441531964467744, + "grad_norm": 0.22578422725200653, + "learning_rate": 7.981699565818486e-05, + "loss": 0.0217, + "step": 33180 + }, + { + "epoch": 6.4434736177855445, + "grad_norm": 0.21260832250118256, + "learning_rate": 7.980372109626437e-05, + "loss": 0.0248, + "step": 33190 + }, + { + "epoch": 6.445415271103345, + "grad_norm": 0.19561181962490082, + "learning_rate": 7.979044327505375e-05, + "loss": 0.0336, + "step": 33200 + }, + { + "epoch": 6.447356924421145, + "grad_norm": 0.17373016476631165, + "learning_rate": 7.977716219600506e-05, + "loss": 0.0302, + "step": 33210 + }, + { + "epoch": 6.449298577738944, + "grad_norm": 0.27783629298210144, + "learning_rate": 7.97638778605707e-05, + "loss": 0.0268, + "step": 33220 + }, + { + "epoch": 6.451240231056745, + "grad_norm": 0.19145609438419342, + "learning_rate": 7.975059027020338e-05, + "loss": 0.026, + "step": 33230 + }, + { + "epoch": 6.453181884374545, + "grad_norm": 0.13304060697555542, + "learning_rate": 7.973729942635623e-05, + "loss": 0.0336, + "step": 33240 + }, + { + "epoch": 6.455123537692345, + "grad_norm": 0.1742095798254013, + "learning_rate": 7.972400533048273e-05, + "loss": 0.032, + "step": 33250 + }, + { + "epoch": 6.4570651910101455, + "grad_norm": 0.10753912478685379, + "learning_rate": 7.97107079840367e-05, + "loss": 0.0294, + "step": 33260 + }, + { + "epoch": 6.459006844327945, + "grad_norm": 0.1981819123029709, + "learning_rate": 7.969740738847231e-05, + "loss": 0.0304, + "step": 33270 + }, + { + "epoch": 6.460948497645745, + "grad_norm": 0.13317210972309113, + "learning_rate": 7.968410354524411e-05, + "loss": 0.0281, + "step": 33280 + }, + { + "epoch": 6.4628901509635455, + "grad_norm": 0.12028073519468307, + "learning_rate": 7.967079645580697e-05, + "loss": 0.0264, + "step": 33290 + }, + { + "epoch": 6.464831804281346, + "grad_norm": 0.19251416623592377, + "learning_rate": 7.965748612161612e-05, + "loss": 0.0329, + "step": 33300 + }, + { + "epoch": 6.466773457599146, + "grad_norm": 0.1956917941570282, + "learning_rate": 7.96441725441272e-05, + "loss": 0.032, + "step": 33310 + }, + { + "epoch": 6.468715110916945, + "grad_norm": 0.23458032310009003, + "learning_rate": 7.963085572479614e-05, + "loss": 0.0285, + "step": 33320 + }, + { + "epoch": 6.470656764234746, + "grad_norm": 0.16376356780529022, + "learning_rate": 7.961753566507924e-05, + "loss": 0.0253, + "step": 33330 + }, + { + "epoch": 6.472598417552546, + "grad_norm": 0.26147782802581787, + "learning_rate": 7.960421236643316e-05, + "loss": 0.033, + "step": 33340 + }, + { + "epoch": 6.474540070870346, + "grad_norm": 0.21698908507823944, + "learning_rate": 7.959088583031496e-05, + "loss": 0.0278, + "step": 33350 + }, + { + "epoch": 6.4764817241881465, + "grad_norm": 0.16479480266571045, + "learning_rate": 7.957755605818194e-05, + "loss": 0.0263, + "step": 33360 + }, + { + "epoch": 6.478423377505946, + "grad_norm": 0.10540761798620224, + "learning_rate": 7.956422305149185e-05, + "loss": 0.0216, + "step": 33370 + }, + { + "epoch": 6.480365030823746, + "grad_norm": 0.15146704018115997, + "learning_rate": 7.95508868117028e-05, + "loss": 0.031, + "step": 33380 + }, + { + "epoch": 6.482306684141546, + "grad_norm": 0.19217967987060547, + "learning_rate": 7.953754734027318e-05, + "loss": 0.0341, + "step": 33390 + }, + { + "epoch": 6.484248337459347, + "grad_norm": 0.23228098452091217, + "learning_rate": 7.952420463866182e-05, + "loss": 0.0268, + "step": 33400 + }, + { + "epoch": 6.486189990777147, + "grad_norm": 0.1331324577331543, + "learning_rate": 7.951085870832782e-05, + "loss": 0.0369, + "step": 33410 + }, + { + "epoch": 6.488131644094947, + "grad_norm": 0.19076383113861084, + "learning_rate": 7.949750955073067e-05, + "loss": 0.0267, + "step": 33420 + }, + { + "epoch": 6.490073297412747, + "grad_norm": 0.1010788083076477, + "learning_rate": 7.948415716733022e-05, + "loss": 0.022, + "step": 33430 + }, + { + "epoch": 6.492014950730547, + "grad_norm": 0.25067612528800964, + "learning_rate": 7.947080155958669e-05, + "loss": 0.0326, + "step": 33440 + }, + { + "epoch": 6.493956604048347, + "grad_norm": 0.10984073579311371, + "learning_rate": 7.94574427289606e-05, + "loss": 0.0309, + "step": 33450 + }, + { + "epoch": 6.4958982573661475, + "grad_norm": 0.12081495672464371, + "learning_rate": 7.944408067691284e-05, + "loss": 0.0252, + "step": 33460 + }, + { + "epoch": 6.497839910683948, + "grad_norm": 0.17617087066173553, + "learning_rate": 7.943071540490473e-05, + "loss": 0.0247, + "step": 33470 + }, + { + "epoch": 6.499781564001747, + "grad_norm": 0.14340154826641083, + "learning_rate": 7.94173469143978e-05, + "loss": 0.0262, + "step": 33480 + }, + { + "epoch": 6.501723217319547, + "grad_norm": 0.16683240234851837, + "learning_rate": 7.940397520685406e-05, + "loss": 0.0249, + "step": 33490 + }, + { + "epoch": 6.503664870637348, + "grad_norm": 0.14302271604537964, + "learning_rate": 7.939060028373577e-05, + "loss": 0.027, + "step": 33500 + }, + { + "epoch": 6.505606523955148, + "grad_norm": 0.1052699089050293, + "learning_rate": 7.937722214650565e-05, + "loss": 0.0296, + "step": 33510 + }, + { + "epoch": 6.507548177272948, + "grad_norm": 0.19758296012878418, + "learning_rate": 7.936384079662666e-05, + "loss": 0.0297, + "step": 33520 + }, + { + "epoch": 6.5094898305907485, + "grad_norm": 0.25572648644447327, + "learning_rate": 7.93504562355622e-05, + "loss": 0.0357, + "step": 33530 + }, + { + "epoch": 6.511431483908548, + "grad_norm": 0.2044427990913391, + "learning_rate": 7.933706846477599e-05, + "loss": 0.0347, + "step": 33540 + }, + { + "epoch": 6.513373137226348, + "grad_norm": 0.12934136390686035, + "learning_rate": 7.932367748573206e-05, + "loss": 0.0212, + "step": 33550 + }, + { + "epoch": 6.515314790544148, + "grad_norm": 0.2595832347869873, + "learning_rate": 7.931028329989485e-05, + "loss": 0.0287, + "step": 33560 + }, + { + "epoch": 6.517256443861949, + "grad_norm": 0.1559179425239563, + "learning_rate": 7.929688590872913e-05, + "loss": 0.0317, + "step": 33570 + }, + { + "epoch": 6.519198097179748, + "grad_norm": 0.17617376148700714, + "learning_rate": 7.928348531370003e-05, + "loss": 0.0326, + "step": 33580 + }, + { + "epoch": 6.521139750497548, + "grad_norm": 0.2164783924818039, + "learning_rate": 7.927008151627297e-05, + "loss": 0.0297, + "step": 33590 + }, + { + "epoch": 6.523081403815349, + "grad_norm": 0.17747417092323303, + "learning_rate": 7.925667451791383e-05, + "loss": 0.0339, + "step": 33600 + }, + { + "epoch": 6.525023057133149, + "grad_norm": 0.1477184295654297, + "learning_rate": 7.924326432008874e-05, + "loss": 0.0243, + "step": 33610 + }, + { + "epoch": 6.526964710450949, + "grad_norm": 0.17985063791275024, + "learning_rate": 7.922985092426422e-05, + "loss": 0.0367, + "step": 33620 + }, + { + "epoch": 6.528906363768749, + "grad_norm": 0.17444156110286713, + "learning_rate": 7.921643433190717e-05, + "loss": 0.0317, + "step": 33630 + }, + { + "epoch": 6.530848017086549, + "grad_norm": 0.221188485622406, + "learning_rate": 7.920301454448478e-05, + "loss": 0.0267, + "step": 33640 + }, + { + "epoch": 6.532789670404349, + "grad_norm": 0.21064119040966034, + "learning_rate": 7.918959156346461e-05, + "loss": 0.0307, + "step": 33650 + }, + { + "epoch": 6.534731323722149, + "grad_norm": 0.18124128878116608, + "learning_rate": 7.91761653903146e-05, + "loss": 0.0288, + "step": 33660 + }, + { + "epoch": 6.53667297703995, + "grad_norm": 0.20955733954906464, + "learning_rate": 7.916273602650302e-05, + "loss": 0.0309, + "step": 33670 + }, + { + "epoch": 6.53861463035775, + "grad_norm": 0.08313922584056854, + "learning_rate": 7.914930347349847e-05, + "loss": 0.0323, + "step": 33680 + }, + { + "epoch": 6.540556283675549, + "grad_norm": 0.12527458369731903, + "learning_rate": 7.913586773276992e-05, + "loss": 0.0281, + "step": 33690 + }, + { + "epoch": 6.54249793699335, + "grad_norm": 0.13222834467887878, + "learning_rate": 7.912242880578667e-05, + "loss": 0.0323, + "step": 33700 + }, + { + "epoch": 6.54443959031115, + "grad_norm": 0.1395120471715927, + "learning_rate": 7.910898669401839e-05, + "loss": 0.0282, + "step": 33710 + }, + { + "epoch": 6.54638124362895, + "grad_norm": 0.21630850434303284, + "learning_rate": 7.909554139893511e-05, + "loss": 0.0263, + "step": 33720 + }, + { + "epoch": 6.54832289694675, + "grad_norm": 0.19028551876544952, + "learning_rate": 7.908209292200715e-05, + "loss": 0.0257, + "step": 33730 + }, + { + "epoch": 6.550264550264551, + "grad_norm": 0.24975977838039398, + "learning_rate": 7.906864126470523e-05, + "loss": 0.025, + "step": 33740 + }, + { + "epoch": 6.55220620358235, + "grad_norm": 0.21052823960781097, + "learning_rate": 7.905518642850041e-05, + "loss": 0.0331, + "step": 33750 + }, + { + "epoch": 6.55414785690015, + "grad_norm": 0.20668916404247284, + "learning_rate": 7.904172841486409e-05, + "loss": 0.0319, + "step": 33760 + }, + { + "epoch": 6.556089510217951, + "grad_norm": 0.3003540635108948, + "learning_rate": 7.902826722526801e-05, + "loss": 0.0355, + "step": 33770 + }, + { + "epoch": 6.558031163535751, + "grad_norm": 0.2535806894302368, + "learning_rate": 7.901480286118427e-05, + "loss": 0.0308, + "step": 33780 + }, + { + "epoch": 6.559972816853551, + "grad_norm": 0.12641341984272003, + "learning_rate": 7.900133532408531e-05, + "loss": 0.0337, + "step": 33790 + }, + { + "epoch": 6.5619144701713505, + "grad_norm": 0.10422774404287338, + "learning_rate": 7.898786461544395e-05, + "loss": 0.0367, + "step": 33800 + }, + { + "epoch": 6.563856123489151, + "grad_norm": 0.16965417563915253, + "learning_rate": 7.897439073673325e-05, + "loss": 0.0316, + "step": 33810 + }, + { + "epoch": 6.565797776806951, + "grad_norm": 0.21965016424655914, + "learning_rate": 7.896091368942677e-05, + "loss": 0.0283, + "step": 33820 + }, + { + "epoch": 6.567739430124751, + "grad_norm": 0.24169965088367462, + "learning_rate": 7.894743347499832e-05, + "loss": 0.0282, + "step": 33830 + }, + { + "epoch": 6.569681083442552, + "grad_norm": 0.17500513792037964, + "learning_rate": 7.893395009492203e-05, + "loss": 0.0295, + "step": 33840 + }, + { + "epoch": 6.571622736760352, + "grad_norm": 0.13759396970272064, + "learning_rate": 7.892046355067248e-05, + "loss": 0.0283, + "step": 33850 + }, + { + "epoch": 6.573564390078151, + "grad_norm": 0.16969315707683563, + "learning_rate": 7.890697384372451e-05, + "loss": 0.0282, + "step": 33860 + }, + { + "epoch": 6.575506043395952, + "grad_norm": 0.12122692167758942, + "learning_rate": 7.889348097555336e-05, + "loss": 0.0316, + "step": 33870 + }, + { + "epoch": 6.577447696713752, + "grad_norm": 0.2690393626689911, + "learning_rate": 7.887998494763455e-05, + "loss": 0.0308, + "step": 33880 + }, + { + "epoch": 6.579389350031552, + "grad_norm": 0.1191205382347107, + "learning_rate": 7.886648576144404e-05, + "loss": 0.0282, + "step": 33890 + }, + { + "epoch": 6.5813310033493515, + "grad_norm": 0.23019421100616455, + "learning_rate": 7.885298341845802e-05, + "loss": 0.0266, + "step": 33900 + }, + { + "epoch": 6.583272656667152, + "grad_norm": 0.1599297970533371, + "learning_rate": 7.883947792015311e-05, + "loss": 0.0302, + "step": 33910 + }, + { + "epoch": 6.585214309984952, + "grad_norm": 0.18450488150119781, + "learning_rate": 7.882596926800628e-05, + "loss": 0.0318, + "step": 33920 + }, + { + "epoch": 6.587155963302752, + "grad_norm": 0.240425705909729, + "learning_rate": 7.881245746349477e-05, + "loss": 0.0333, + "step": 33930 + }, + { + "epoch": 6.589097616620553, + "grad_norm": 0.17370860278606415, + "learning_rate": 7.879894250809623e-05, + "loss": 0.0314, + "step": 33940 + }, + { + "epoch": 6.591039269938353, + "grad_norm": 0.13033828139305115, + "learning_rate": 7.878542440328865e-05, + "loss": 0.0264, + "step": 33950 + }, + { + "epoch": 6.592980923256152, + "grad_norm": 0.25845226645469666, + "learning_rate": 7.877190315055031e-05, + "loss": 0.031, + "step": 33960 + }, + { + "epoch": 6.5949225765739525, + "grad_norm": 0.2438809871673584, + "learning_rate": 7.875837875135991e-05, + "loss": 0.0354, + "step": 33970 + }, + { + "epoch": 6.596864229891753, + "grad_norm": 0.4251873791217804, + "learning_rate": 7.874485120719646e-05, + "loss": 0.0351, + "step": 33980 + }, + { + "epoch": 6.598805883209553, + "grad_norm": 0.23566870391368866, + "learning_rate": 7.873132051953928e-05, + "loss": 0.0306, + "step": 33990 + }, + { + "epoch": 6.600747536527353, + "grad_norm": 0.16232942044734955, + "learning_rate": 7.87177866898681e-05, + "loss": 0.031, + "step": 34000 + }, + { + "epoch": 6.602689189845153, + "grad_norm": 0.1643732190132141, + "learning_rate": 7.870424971966294e-05, + "loss": 0.0309, + "step": 34010 + }, + { + "epoch": 6.604630843162953, + "grad_norm": 0.15166950225830078, + "learning_rate": 7.869070961040419e-05, + "loss": 0.0317, + "step": 34020 + }, + { + "epoch": 6.606572496480753, + "grad_norm": 0.2329782098531723, + "learning_rate": 7.867716636357257e-05, + "loss": 0.0312, + "step": 34030 + }, + { + "epoch": 6.608514149798554, + "grad_norm": 0.2049921154975891, + "learning_rate": 7.866361998064915e-05, + "loss": 0.0309, + "step": 34040 + }, + { + "epoch": 6.610455803116354, + "grad_norm": 0.22948701679706573, + "learning_rate": 7.865007046311534e-05, + "loss": 0.0285, + "step": 34050 + }, + { + "epoch": 6.612397456434154, + "grad_norm": 0.08032187819480896, + "learning_rate": 7.86365178124529e-05, + "loss": 0.0265, + "step": 34060 + }, + { + "epoch": 6.6143391097519535, + "grad_norm": 0.1655253916978836, + "learning_rate": 7.862296203014394e-05, + "loss": 0.0312, + "step": 34070 + }, + { + "epoch": 6.616280763069754, + "grad_norm": 0.10823177546262741, + "learning_rate": 7.860940311767088e-05, + "loss": 0.0251, + "step": 34080 + }, + { + "epoch": 6.618222416387554, + "grad_norm": 0.12644144892692566, + "learning_rate": 7.85958410765165e-05, + "loss": 0.0293, + "step": 34090 + }, + { + "epoch": 6.620164069705354, + "grad_norm": 0.19112473726272583, + "learning_rate": 7.858227590816394e-05, + "loss": 0.0272, + "step": 34100 + }, + { + "epoch": 6.622105723023155, + "grad_norm": 0.17945687472820282, + "learning_rate": 7.856870761409664e-05, + "loss": 0.0326, + "step": 34110 + }, + { + "epoch": 6.624047376340954, + "grad_norm": 0.20942726731300354, + "learning_rate": 7.855513619579846e-05, + "loss": 0.0286, + "step": 34120 + }, + { + "epoch": 6.625989029658754, + "grad_norm": 0.12302982807159424, + "learning_rate": 7.85415616547535e-05, + "loss": 0.024, + "step": 34130 + }, + { + "epoch": 6.6279306829765545, + "grad_norm": 0.10281126946210861, + "learning_rate": 7.852798399244627e-05, + "loss": 0.0257, + "step": 34140 + }, + { + "epoch": 6.629872336294355, + "grad_norm": 0.20746482908725739, + "learning_rate": 7.851440321036161e-05, + "loss": 0.0371, + "step": 34150 + }, + { + "epoch": 6.631813989612155, + "grad_norm": 0.12957827746868134, + "learning_rate": 7.850081930998468e-05, + "loss": 0.0291, + "step": 34160 + }, + { + "epoch": 6.633755642929955, + "grad_norm": 0.21866290271282196, + "learning_rate": 7.8487232292801e-05, + "loss": 0.0279, + "step": 34170 + }, + { + "epoch": 6.635697296247755, + "grad_norm": 0.15668703615665436, + "learning_rate": 7.847364216029642e-05, + "loss": 0.0202, + "step": 34180 + }, + { + "epoch": 6.637638949565555, + "grad_norm": 0.24481455981731415, + "learning_rate": 7.846004891395716e-05, + "loss": 0.0404, + "step": 34190 + }, + { + "epoch": 6.639580602883355, + "grad_norm": 0.0991225615143776, + "learning_rate": 7.844645255526972e-05, + "loss": 0.0323, + "step": 34200 + }, + { + "epoch": 6.6415222562011555, + "grad_norm": 0.14725808799266815, + "learning_rate": 7.843285308572101e-05, + "loss": 0.025, + "step": 34210 + }, + { + "epoch": 6.643463909518955, + "grad_norm": 0.12703613936901093, + "learning_rate": 7.841925050679823e-05, + "loss": 0.0302, + "step": 34220 + }, + { + "epoch": 6.645405562836755, + "grad_norm": 0.13153298199176788, + "learning_rate": 7.840564481998895e-05, + "loss": 0.0339, + "step": 34230 + }, + { + "epoch": 6.6473472161545555, + "grad_norm": 0.14177122712135315, + "learning_rate": 7.839203602678105e-05, + "loss": 0.0326, + "step": 34240 + }, + { + "epoch": 6.649288869472356, + "grad_norm": 0.11848311126232147, + "learning_rate": 7.837842412866279e-05, + "loss": 0.0278, + "step": 34250 + }, + { + "epoch": 6.651230522790156, + "grad_norm": 0.17316456139087677, + "learning_rate": 7.836480912712272e-05, + "loss": 0.0361, + "step": 34260 + }, + { + "epoch": 6.653172176107956, + "grad_norm": 0.1922834813594818, + "learning_rate": 7.835119102364976e-05, + "loss": 0.0354, + "step": 34270 + }, + { + "epoch": 6.655113829425756, + "grad_norm": 0.15328367054462433, + "learning_rate": 7.83375698197332e-05, + "loss": 0.0352, + "step": 34280 + }, + { + "epoch": 6.657055482743556, + "grad_norm": 0.1851423680782318, + "learning_rate": 7.83239455168626e-05, + "loss": 0.0218, + "step": 34290 + }, + { + "epoch": 6.658997136061356, + "grad_norm": 0.1994243860244751, + "learning_rate": 7.83103181165279e-05, + "loss": 0.0318, + "step": 34300 + }, + { + "epoch": 6.6609387893791565, + "grad_norm": 0.16085051000118256, + "learning_rate": 7.829668762021937e-05, + "loss": 0.0346, + "step": 34310 + }, + { + "epoch": 6.662880442696957, + "grad_norm": 0.20418322086334229, + "learning_rate": 7.828305402942764e-05, + "loss": 0.0292, + "step": 34320 + }, + { + "epoch": 6.664822096014756, + "grad_norm": 0.18738245964050293, + "learning_rate": 7.826941734564363e-05, + "loss": 0.0347, + "step": 34330 + }, + { + "epoch": 6.666763749332556, + "grad_norm": 0.19346685707569122, + "learning_rate": 7.825577757035865e-05, + "loss": 0.0234, + "step": 34340 + }, + { + "epoch": 6.668705402650357, + "grad_norm": 0.15124069154262543, + "learning_rate": 7.824213470506431e-05, + "loss": 0.0281, + "step": 34350 + }, + { + "epoch": 6.670647055968157, + "grad_norm": 0.1938045620918274, + "learning_rate": 7.822848875125257e-05, + "loss": 0.0309, + "step": 34360 + }, + { + "epoch": 6.672588709285957, + "grad_norm": 0.1508161723613739, + "learning_rate": 7.821483971041576e-05, + "loss": 0.0385, + "step": 34370 + }, + { + "epoch": 6.6745303626037575, + "grad_norm": 0.18730910122394562, + "learning_rate": 7.820118758404649e-05, + "loss": 0.0283, + "step": 34380 + }, + { + "epoch": 6.676472015921557, + "grad_norm": 0.15057002007961273, + "learning_rate": 7.818753237363776e-05, + "loss": 0.022, + "step": 34390 + }, + { + "epoch": 6.678413669239357, + "grad_norm": 0.1364169865846634, + "learning_rate": 7.817387408068286e-05, + "loss": 0.0335, + "step": 34400 + }, + { + "epoch": 6.6803553225571575, + "grad_norm": 0.16068409383296967, + "learning_rate": 7.816021270667544e-05, + "loss": 0.0315, + "step": 34410 + }, + { + "epoch": 6.682296975874958, + "grad_norm": 0.15488745272159576, + "learning_rate": 7.81465482531095e-05, + "loss": 0.0344, + "step": 34420 + }, + { + "epoch": 6.684238629192758, + "grad_norm": 0.15332278609275818, + "learning_rate": 7.813288072147938e-05, + "loss": 0.0271, + "step": 34430 + }, + { + "epoch": 6.686180282510557, + "grad_norm": 0.13410818576812744, + "learning_rate": 7.811921011327972e-05, + "loss": 0.0321, + "step": 34440 + }, + { + "epoch": 6.688121935828358, + "grad_norm": 0.13541767001152039, + "learning_rate": 7.810553643000549e-05, + "loss": 0.0322, + "step": 34450 + }, + { + "epoch": 6.690063589146158, + "grad_norm": 0.1282317191362381, + "learning_rate": 7.809185967315206e-05, + "loss": 0.0279, + "step": 34460 + }, + { + "epoch": 6.692005242463958, + "grad_norm": 0.24424125254154205, + "learning_rate": 7.80781798442151e-05, + "loss": 0.0298, + "step": 34470 + }, + { + "epoch": 6.6939468957817585, + "grad_norm": 0.17410476505756378, + "learning_rate": 7.806449694469061e-05, + "loss": 0.022, + "step": 34480 + }, + { + "epoch": 6.695888549099559, + "grad_norm": 0.22729218006134033, + "learning_rate": 7.805081097607492e-05, + "loss": 0.0375, + "step": 34490 + }, + { + "epoch": 6.697830202417358, + "grad_norm": 0.20530864596366882, + "learning_rate": 7.803712193986474e-05, + "loss": 0.0361, + "step": 34500 + }, + { + "epoch": 6.699771855735158, + "grad_norm": 0.16624604165554047, + "learning_rate": 7.802342983755702e-05, + "loss": 0.0272, + "step": 34510 + }, + { + "epoch": 6.701713509052959, + "grad_norm": 0.100346140563488, + "learning_rate": 7.800973467064918e-05, + "loss": 0.0275, + "step": 34520 + }, + { + "epoch": 6.703655162370759, + "grad_norm": 0.11354120075702667, + "learning_rate": 7.799603644063884e-05, + "loss": 0.0337, + "step": 34530 + }, + { + "epoch": 6.705596815688558, + "grad_norm": 0.2466568499803543, + "learning_rate": 7.798233514902405e-05, + "loss": 0.0323, + "step": 34540 + }, + { + "epoch": 6.707538469006359, + "grad_norm": 0.14900851249694824, + "learning_rate": 7.796863079730318e-05, + "loss": 0.0313, + "step": 34550 + }, + { + "epoch": 6.709480122324159, + "grad_norm": 0.18032750487327576, + "learning_rate": 7.795492338697488e-05, + "loss": 0.0236, + "step": 34560 + }, + { + "epoch": 6.711421775641959, + "grad_norm": 0.12512585520744324, + "learning_rate": 7.794121291953819e-05, + "loss": 0.0353, + "step": 34570 + }, + { + "epoch": 6.7133634289597595, + "grad_norm": 0.2030434012413025, + "learning_rate": 7.792749939649246e-05, + "loss": 0.0286, + "step": 34580 + }, + { + "epoch": 6.71530508227756, + "grad_norm": 0.20925673842430115, + "learning_rate": 7.79137828193374e-05, + "loss": 0.0326, + "step": 34590 + }, + { + "epoch": 6.717246735595359, + "grad_norm": 0.22306346893310547, + "learning_rate": 7.790006318957301e-05, + "loss": 0.0259, + "step": 34600 + }, + { + "epoch": 6.719188388913159, + "grad_norm": 0.17353713512420654, + "learning_rate": 7.788634050869965e-05, + "loss": 0.0322, + "step": 34610 + }, + { + "epoch": 6.72113004223096, + "grad_norm": 0.10393532365560532, + "learning_rate": 7.787261477821803e-05, + "loss": 0.0421, + "step": 34620 + }, + { + "epoch": 6.72307169554876, + "grad_norm": 0.20031747221946716, + "learning_rate": 7.785888599962916e-05, + "loss": 0.035, + "step": 34630 + }, + { + "epoch": 6.72501334886656, + "grad_norm": 0.2774578034877777, + "learning_rate": 7.784515417443439e-05, + "loss": 0.0261, + "step": 34640 + }, + { + "epoch": 6.72695500218436, + "grad_norm": 0.16058284044265747, + "learning_rate": 7.783141930413545e-05, + "loss": 0.0279, + "step": 34650 + }, + { + "epoch": 6.72889665550216, + "grad_norm": 0.21110571920871735, + "learning_rate": 7.78176813902343e-05, + "loss": 0.0368, + "step": 34660 + }, + { + "epoch": 6.73083830881996, + "grad_norm": 0.18277513980865479, + "learning_rate": 7.780394043423336e-05, + "loss": 0.0399, + "step": 34670 + }, + { + "epoch": 6.73277996213776, + "grad_norm": 0.21650417149066925, + "learning_rate": 7.77901964376353e-05, + "loss": 0.0329, + "step": 34680 + }, + { + "epoch": 6.734721615455561, + "grad_norm": 0.2566540241241455, + "learning_rate": 7.777644940194316e-05, + "loss": 0.0335, + "step": 34690 + }, + { + "epoch": 6.736663268773361, + "grad_norm": 0.1986154019832611, + "learning_rate": 7.776269932866023e-05, + "loss": 0.0301, + "step": 34700 + }, + { + "epoch": 6.73860492209116, + "grad_norm": 0.16489708423614502, + "learning_rate": 7.774894621929026e-05, + "loss": 0.0266, + "step": 34710 + }, + { + "epoch": 6.740546575408961, + "grad_norm": 0.1433258056640625, + "learning_rate": 7.773519007533725e-05, + "loss": 0.0363, + "step": 34720 + }, + { + "epoch": 6.742488228726761, + "grad_norm": 0.14168810844421387, + "learning_rate": 7.772143089830556e-05, + "loss": 0.0274, + "step": 34730 + }, + { + "epoch": 6.744429882044561, + "grad_norm": 0.19599424302577972, + "learning_rate": 7.770766868969985e-05, + "loss": 0.0323, + "step": 34740 + }, + { + "epoch": 6.746371535362361, + "grad_norm": 0.14160221815109253, + "learning_rate": 7.769390345102518e-05, + "loss": 0.0259, + "step": 34750 + }, + { + "epoch": 6.748313188680161, + "grad_norm": 0.11284924298524857, + "learning_rate": 7.768013518378683e-05, + "loss": 0.0262, + "step": 34760 + }, + { + "epoch": 6.750254841997961, + "grad_norm": 0.13973015546798706, + "learning_rate": 7.766636388949053e-05, + "loss": 0.028, + "step": 34770 + }, + { + "epoch": 6.752196495315761, + "grad_norm": 0.15168578922748566, + "learning_rate": 7.765258956964229e-05, + "loss": 0.0234, + "step": 34780 + }, + { + "epoch": 6.754138148633562, + "grad_norm": 0.08807583153247833, + "learning_rate": 7.76388122257484e-05, + "loss": 0.0224, + "step": 34790 + }, + { + "epoch": 6.756079801951362, + "grad_norm": 0.17177607119083405, + "learning_rate": 7.762503185931558e-05, + "loss": 0.0301, + "step": 34800 + }, + { + "epoch": 6.758021455269161, + "grad_norm": 0.12241638451814651, + "learning_rate": 7.76112484718508e-05, + "loss": 0.0278, + "step": 34810 + }, + { + "epoch": 6.759963108586962, + "grad_norm": 0.15146145224571228, + "learning_rate": 7.75974620648614e-05, + "loss": 0.0312, + "step": 34820 + }, + { + "epoch": 6.761904761904762, + "grad_norm": 0.1986989974975586, + "learning_rate": 7.758367263985503e-05, + "loss": 0.0325, + "step": 34830 + }, + { + "epoch": 6.763846415222562, + "grad_norm": 0.15177832543849945, + "learning_rate": 7.75698801983397e-05, + "loss": 0.0316, + "step": 34840 + }, + { + "epoch": 6.765788068540362, + "grad_norm": 0.24527764320373535, + "learning_rate": 7.755608474182372e-05, + "loss": 0.0316, + "step": 34850 + }, + { + "epoch": 6.767729721858162, + "grad_norm": 0.20786109566688538, + "learning_rate": 7.754228627181574e-05, + "loss": 0.0337, + "step": 34860 + }, + { + "epoch": 6.769671375175962, + "grad_norm": 0.13976895809173584, + "learning_rate": 7.752848478982476e-05, + "loss": 0.0304, + "step": 34870 + }, + { + "epoch": 6.771613028493762, + "grad_norm": 0.15582279860973358, + "learning_rate": 7.751468029736006e-05, + "loss": 0.0249, + "step": 34880 + }, + { + "epoch": 6.773554681811563, + "grad_norm": 0.20110663771629333, + "learning_rate": 7.750087279593129e-05, + "loss": 0.0267, + "step": 34890 + }, + { + "epoch": 6.775496335129363, + "grad_norm": 0.15879575908184052, + "learning_rate": 7.748706228704843e-05, + "loss": 0.0295, + "step": 34900 + }, + { + "epoch": 6.777437988447163, + "grad_norm": 0.21615062654018402, + "learning_rate": 7.747324877222176e-05, + "loss": 0.027, + "step": 34910 + }, + { + "epoch": 6.7793796417649625, + "grad_norm": 0.17005464434623718, + "learning_rate": 7.745943225296188e-05, + "loss": 0.0292, + "step": 34920 + }, + { + "epoch": 6.781321295082763, + "grad_norm": 0.15940099954605103, + "learning_rate": 7.744561273077981e-05, + "loss": 0.0311, + "step": 34930 + }, + { + "epoch": 6.783262948400563, + "grad_norm": 0.15813486278057098, + "learning_rate": 7.743179020718678e-05, + "loss": 0.0259, + "step": 34940 + }, + { + "epoch": 6.785204601718363, + "grad_norm": 0.15824539959430695, + "learning_rate": 7.741796468369443e-05, + "loss": 0.0304, + "step": 34950 + }, + { + "epoch": 6.787146255036164, + "grad_norm": 0.22812585532665253, + "learning_rate": 7.740413616181466e-05, + "loss": 0.0307, + "step": 34960 + }, + { + "epoch": 6.789087908353963, + "grad_norm": 0.1425376981496811, + "learning_rate": 7.739030464305978e-05, + "loss": 0.0321, + "step": 34970 + }, + { + "epoch": 6.791029561671763, + "grad_norm": 0.22515089809894562, + "learning_rate": 7.737647012894235e-05, + "loss": 0.026, + "step": 34980 + }, + { + "epoch": 6.792971214989564, + "grad_norm": 0.16100694239139557, + "learning_rate": 7.736263262097532e-05, + "loss": 0.0271, + "step": 34990 + }, + { + "epoch": 6.794912868307364, + "grad_norm": 0.1442883163690567, + "learning_rate": 7.734879212067192e-05, + "loss": 0.0225, + "step": 35000 + }, + { + "epoch": 6.796854521625164, + "grad_norm": 0.09149777889251709, + "learning_rate": 7.733494862954573e-05, + "loss": 0.0231, + "step": 35010 + }, + { + "epoch": 6.798796174942964, + "grad_norm": 0.1646118462085724, + "learning_rate": 7.732110214911066e-05, + "loss": 0.027, + "step": 35020 + }, + { + "epoch": 6.800737828260764, + "grad_norm": 0.16483914852142334, + "learning_rate": 7.730725268088094e-05, + "loss": 0.027, + "step": 35030 + }, + { + "epoch": 6.802679481578564, + "grad_norm": 0.12430781871080399, + "learning_rate": 7.729340022637112e-05, + "loss": 0.0258, + "step": 35040 + }, + { + "epoch": 6.804621134896364, + "grad_norm": 0.19447392225265503, + "learning_rate": 7.727954478709607e-05, + "loss": 0.0294, + "step": 35050 + }, + { + "epoch": 6.806562788214165, + "grad_norm": 0.15028473734855652, + "learning_rate": 7.726568636457103e-05, + "loss": 0.0233, + "step": 35060 + }, + { + "epoch": 6.808504441531964, + "grad_norm": 0.22221724689006805, + "learning_rate": 7.725182496031153e-05, + "loss": 0.0359, + "step": 35070 + }, + { + "epoch": 6.810446094849764, + "grad_norm": 0.29479148983955383, + "learning_rate": 7.723796057583341e-05, + "loss": 0.0384, + "step": 35080 + }, + { + "epoch": 6.8123877481675645, + "grad_norm": 0.20874983072280884, + "learning_rate": 7.722409321265287e-05, + "loss": 0.0378, + "step": 35090 + }, + { + "epoch": 6.814329401485365, + "grad_norm": 0.10976245999336243, + "learning_rate": 7.721022287228645e-05, + "loss": 0.0296, + "step": 35100 + }, + { + "epoch": 6.816271054803165, + "grad_norm": 0.16887325048446655, + "learning_rate": 7.719634955625094e-05, + "loss": 0.0311, + "step": 35110 + }, + { + "epoch": 6.818212708120965, + "grad_norm": 0.13958488404750824, + "learning_rate": 7.718247326606355e-05, + "loss": 0.028, + "step": 35120 + }, + { + "epoch": 6.820154361438765, + "grad_norm": 0.2273133546113968, + "learning_rate": 7.716859400324175e-05, + "loss": 0.0292, + "step": 35130 + }, + { + "epoch": 6.822096014756565, + "grad_norm": 0.1300773024559021, + "learning_rate": 7.715471176930335e-05, + "loss": 0.0288, + "step": 35140 + }, + { + "epoch": 6.824037668074365, + "grad_norm": 0.2846824526786804, + "learning_rate": 7.714082656576651e-05, + "loss": 0.0309, + "step": 35150 + }, + { + "epoch": 6.825979321392166, + "grad_norm": 0.22312501072883606, + "learning_rate": 7.712693839414968e-05, + "loss": 0.0307, + "step": 35160 + }, + { + "epoch": 6.827920974709966, + "grad_norm": 0.16117005050182343, + "learning_rate": 7.711304725597164e-05, + "loss": 0.031, + "step": 35170 + }, + { + "epoch": 6.829862628027765, + "grad_norm": 0.16516771912574768, + "learning_rate": 7.709915315275151e-05, + "loss": 0.0287, + "step": 35180 + }, + { + "epoch": 6.8318042813455655, + "grad_norm": 0.25636547803878784, + "learning_rate": 7.708525608600876e-05, + "loss": 0.0258, + "step": 35190 + }, + { + "epoch": 6.833745934663366, + "grad_norm": 0.1950451284646988, + "learning_rate": 7.707135605726311e-05, + "loss": 0.025, + "step": 35200 + }, + { + "epoch": 6.835687587981166, + "grad_norm": 0.18931539356708527, + "learning_rate": 7.705745306803466e-05, + "loss": 0.0299, + "step": 35210 + }, + { + "epoch": 6.837629241298966, + "grad_norm": 0.11663661897182465, + "learning_rate": 7.704354711984383e-05, + "loss": 0.0239, + "step": 35220 + }, + { + "epoch": 6.839570894616767, + "grad_norm": 0.18663011491298676, + "learning_rate": 7.702963821421133e-05, + "loss": 0.0281, + "step": 35230 + }, + { + "epoch": 6.841512547934566, + "grad_norm": 0.14212432503700256, + "learning_rate": 7.701572635265826e-05, + "loss": 0.0253, + "step": 35240 + }, + { + "epoch": 6.843454201252366, + "grad_norm": 0.16161231696605682, + "learning_rate": 7.700181153670596e-05, + "loss": 0.0274, + "step": 35250 + }, + { + "epoch": 6.8453958545701665, + "grad_norm": 0.13850852847099304, + "learning_rate": 7.698789376787616e-05, + "loss": 0.0284, + "step": 35260 + }, + { + "epoch": 6.847337507887967, + "grad_norm": 0.1670788675546646, + "learning_rate": 7.697397304769085e-05, + "loss": 0.0292, + "step": 35270 + }, + { + "epoch": 6.849279161205767, + "grad_norm": 0.18529759347438812, + "learning_rate": 7.696004937767241e-05, + "loss": 0.033, + "step": 35280 + }, + { + "epoch": 6.8512208145235665, + "grad_norm": 0.2394055277109146, + "learning_rate": 7.694612275934352e-05, + "loss": 0.0289, + "step": 35290 + }, + { + "epoch": 6.853162467841367, + "grad_norm": 0.1329985111951828, + "learning_rate": 7.693219319422714e-05, + "loss": 0.024, + "step": 35300 + }, + { + "epoch": 6.855104121159167, + "grad_norm": 0.18663693964481354, + "learning_rate": 7.69182606838466e-05, + "loss": 0.0301, + "step": 35310 + }, + { + "epoch": 6.857045774476967, + "grad_norm": 0.19544197618961334, + "learning_rate": 7.690432522972558e-05, + "loss": 0.0312, + "step": 35320 + }, + { + "epoch": 6.8589874277947676, + "grad_norm": 0.1305239051580429, + "learning_rate": 7.689038683338796e-05, + "loss": 0.0309, + "step": 35330 + }, + { + "epoch": 6.860929081112568, + "grad_norm": 0.104022316634655, + "learning_rate": 7.687644549635808e-05, + "loss": 0.038, + "step": 35340 + }, + { + "epoch": 6.862870734430367, + "grad_norm": 0.170240581035614, + "learning_rate": 7.686250122016053e-05, + "loss": 0.023, + "step": 35350 + }, + { + "epoch": 6.8648123877481675, + "grad_norm": 0.21316513419151306, + "learning_rate": 7.684855400632022e-05, + "loss": 0.0301, + "step": 35360 + }, + { + "epoch": 6.866754041065968, + "grad_norm": 0.12500451505184174, + "learning_rate": 7.683460385636243e-05, + "loss": 0.0285, + "step": 35370 + }, + { + "epoch": 6.868695694383768, + "grad_norm": 0.1506001204252243, + "learning_rate": 7.68206507718127e-05, + "loss": 0.0311, + "step": 35380 + }, + { + "epoch": 6.870637347701567, + "grad_norm": 0.21025146543979645, + "learning_rate": 7.680669475419693e-05, + "loss": 0.03, + "step": 35390 + }, + { + "epoch": 6.872579001019368, + "grad_norm": 0.12616652250289917, + "learning_rate": 7.679273580504132e-05, + "loss": 0.0369, + "step": 35400 + }, + { + "epoch": 6.874520654337168, + "grad_norm": 0.18443459272384644, + "learning_rate": 7.67787739258724e-05, + "loss": 0.0306, + "step": 35410 + }, + { + "epoch": 6.876462307654968, + "grad_norm": 0.14918698370456696, + "learning_rate": 7.676480911821705e-05, + "loss": 0.032, + "step": 35420 + }, + { + "epoch": 6.8784039609727685, + "grad_norm": 0.10508673638105392, + "learning_rate": 7.675084138360238e-05, + "loss": 0.027, + "step": 35430 + }, + { + "epoch": 6.880345614290569, + "grad_norm": 0.17001491785049438, + "learning_rate": 7.673687072355592e-05, + "loss": 0.0327, + "step": 35440 + }, + { + "epoch": 6.882287267608368, + "grad_norm": 0.1854696124792099, + "learning_rate": 7.67228971396055e-05, + "loss": 0.0327, + "step": 35450 + }, + { + "epoch": 6.8842289209261684, + "grad_norm": 0.1449102908372879, + "learning_rate": 7.670892063327922e-05, + "loss": 0.0226, + "step": 35460 + }, + { + "epoch": 6.886170574243969, + "grad_norm": 0.16758617758750916, + "learning_rate": 7.669494120610552e-05, + "loss": 0.0333, + "step": 35470 + }, + { + "epoch": 6.888112227561769, + "grad_norm": 0.11909189075231552, + "learning_rate": 7.66809588596132e-05, + "loss": 0.0283, + "step": 35480 + }, + { + "epoch": 6.890053880879569, + "grad_norm": 0.12135133147239685, + "learning_rate": 7.666697359533132e-05, + "loss": 0.031, + "step": 35490 + }, + { + "epoch": 6.891995534197369, + "grad_norm": 0.17113108932971954, + "learning_rate": 7.665298541478932e-05, + "loss": 0.0273, + "step": 35500 + }, + { + "epoch": 6.893937187515169, + "grad_norm": 0.1437770426273346, + "learning_rate": 7.66389943195169e-05, + "loss": 0.0262, + "step": 35510 + }, + { + "epoch": 6.895878840832969, + "grad_norm": 0.15873423218727112, + "learning_rate": 7.66250003110441e-05, + "loss": 0.0312, + "step": 35520 + }, + { + "epoch": 6.8978204941507695, + "grad_norm": 0.18604472279548645, + "learning_rate": 7.66110033909013e-05, + "loss": 0.0341, + "step": 35530 + }, + { + "epoch": 6.89976214746857, + "grad_norm": 0.19904546439647675, + "learning_rate": 7.659700356061918e-05, + "loss": 0.0366, + "step": 35540 + }, + { + "epoch": 6.90170380078637, + "grad_norm": 0.14556752145290375, + "learning_rate": 7.658300082172875e-05, + "loss": 0.02, + "step": 35550 + }, + { + "epoch": 6.903645454104169, + "grad_norm": 0.18559664487838745, + "learning_rate": 7.65689951757613e-05, + "loss": 0.0332, + "step": 35560 + }, + { + "epoch": 6.90558710742197, + "grad_norm": 0.16698388755321503, + "learning_rate": 7.655498662424849e-05, + "loss": 0.0326, + "step": 35570 + }, + { + "epoch": 6.90752876073977, + "grad_norm": 0.19300644099712372, + "learning_rate": 7.654097516872227e-05, + "loss": 0.0279, + "step": 35580 + }, + { + "epoch": 6.90947041405757, + "grad_norm": 0.20577386021614075, + "learning_rate": 7.65269608107149e-05, + "loss": 0.031, + "step": 35590 + }, + { + "epoch": 6.9114120673753705, + "grad_norm": 0.22112782299518585, + "learning_rate": 7.6512943551759e-05, + "loss": 0.0358, + "step": 35600 + }, + { + "epoch": 6.91335372069317, + "grad_norm": 0.19321729242801666, + "learning_rate": 7.649892339338743e-05, + "loss": 0.0343, + "step": 35610 + }, + { + "epoch": 6.91529537401097, + "grad_norm": 0.1436993032693863, + "learning_rate": 7.648490033713344e-05, + "loss": 0.0257, + "step": 35620 + }, + { + "epoch": 6.91723702732877, + "grad_norm": 0.18832798302173615, + "learning_rate": 7.647087438453058e-05, + "loss": 0.0262, + "step": 35630 + }, + { + "epoch": 6.919178680646571, + "grad_norm": 0.2533597946166992, + "learning_rate": 7.64568455371127e-05, + "loss": 0.0275, + "step": 35640 + }, + { + "epoch": 6.921120333964371, + "grad_norm": 0.15794780850410461, + "learning_rate": 7.644281379641396e-05, + "loss": 0.028, + "step": 35650 + }, + { + "epoch": 6.923061987282171, + "grad_norm": 0.1710764765739441, + "learning_rate": 7.642877916396887e-05, + "loss": 0.0299, + "step": 35660 + }, + { + "epoch": 6.925003640599971, + "grad_norm": 0.31367331743240356, + "learning_rate": 7.641474164131221e-05, + "loss": 0.0331, + "step": 35670 + }, + { + "epoch": 6.926945293917771, + "grad_norm": 0.192983940243721, + "learning_rate": 7.640070122997913e-05, + "loss": 0.0282, + "step": 35680 + }, + { + "epoch": 6.928886947235571, + "grad_norm": 0.12825386226177216, + "learning_rate": 7.638665793150505e-05, + "loss": 0.0337, + "step": 35690 + }, + { + "epoch": 6.9308286005533715, + "grad_norm": 0.2493455708026886, + "learning_rate": 7.637261174742574e-05, + "loss": 0.0447, + "step": 35700 + }, + { + "epoch": 6.932770253871171, + "grad_norm": 0.223856121301651, + "learning_rate": 7.635856267927725e-05, + "loss": 0.0326, + "step": 35710 + }, + { + "epoch": 6.934711907188971, + "grad_norm": 0.10137344151735306, + "learning_rate": 7.634451072859597e-05, + "loss": 0.0212, + "step": 35720 + }, + { + "epoch": 6.936653560506771, + "grad_norm": 0.15541066229343414, + "learning_rate": 7.633045589691863e-05, + "loss": 0.0245, + "step": 35730 + }, + { + "epoch": 6.938595213824572, + "grad_norm": 0.21802416443824768, + "learning_rate": 7.63163981857822e-05, + "loss": 0.0253, + "step": 35740 + }, + { + "epoch": 6.940536867142372, + "grad_norm": 0.2763529419898987, + "learning_rate": 7.630233759672403e-05, + "loss": 0.0329, + "step": 35750 + }, + { + "epoch": 6.942478520460172, + "grad_norm": 0.21498261392116547, + "learning_rate": 7.628827413128177e-05, + "loss": 0.0285, + "step": 35760 + }, + { + "epoch": 6.944420173777972, + "grad_norm": 0.1692885309457779, + "learning_rate": 7.627420779099338e-05, + "loss": 0.0326, + "step": 35770 + }, + { + "epoch": 6.946361827095772, + "grad_norm": 0.10897502303123474, + "learning_rate": 7.626013857739711e-05, + "loss": 0.0303, + "step": 35780 + }, + { + "epoch": 6.948303480413572, + "grad_norm": 0.13797423243522644, + "learning_rate": 7.624606649203158e-05, + "loss": 0.0291, + "step": 35790 + }, + { + "epoch": 6.950245133731372, + "grad_norm": 0.234788715839386, + "learning_rate": 7.623199153643569e-05, + "loss": 0.0364, + "step": 35800 + }, + { + "epoch": 6.952186787049173, + "grad_norm": 0.1862485110759735, + "learning_rate": 7.621791371214863e-05, + "loss": 0.0261, + "step": 35810 + }, + { + "epoch": 6.954128440366972, + "grad_norm": 0.2023027241230011, + "learning_rate": 7.620383302070995e-05, + "loss": 0.0326, + "step": 35820 + }, + { + "epoch": 6.956070093684772, + "grad_norm": 0.174139603972435, + "learning_rate": 7.61897494636595e-05, + "loss": 0.032, + "step": 35830 + }, + { + "epoch": 6.958011747002573, + "grad_norm": 0.3136448860168457, + "learning_rate": 7.617566304253739e-05, + "loss": 0.0388, + "step": 35840 + }, + { + "epoch": 6.959953400320373, + "grad_norm": 0.21147924661636353, + "learning_rate": 7.616157375888416e-05, + "loss": 0.0372, + "step": 35850 + }, + { + "epoch": 6.961895053638173, + "grad_norm": 0.2234170138835907, + "learning_rate": 7.614748161424053e-05, + "loss": 0.0278, + "step": 35860 + }, + { + "epoch": 6.9638367069559735, + "grad_norm": 0.14377689361572266, + "learning_rate": 7.613338661014763e-05, + "loss": 0.0345, + "step": 35870 + }, + { + "epoch": 6.965778360273773, + "grad_norm": 0.12798085808753967, + "learning_rate": 7.611928874814686e-05, + "loss": 0.0346, + "step": 35880 + }, + { + "epoch": 6.967720013591573, + "grad_norm": 0.1796179711818695, + "learning_rate": 7.610518802977996e-05, + "loss": 0.0315, + "step": 35890 + }, + { + "epoch": 6.969661666909373, + "grad_norm": 0.24897722899913788, + "learning_rate": 7.609108445658893e-05, + "loss": 0.0285, + "step": 35900 + }, + { + "epoch": 6.971603320227174, + "grad_norm": 0.12852253019809723, + "learning_rate": 7.607697803011612e-05, + "loss": 0.0316, + "step": 35910 + }, + { + "epoch": 6.973544973544973, + "grad_norm": 0.15952853858470917, + "learning_rate": 7.606286875190421e-05, + "loss": 0.0297, + "step": 35920 + }, + { + "epoch": 6.975486626862773, + "grad_norm": 0.22232139110565186, + "learning_rate": 7.604875662349617e-05, + "loss": 0.0343, + "step": 35930 + }, + { + "epoch": 6.977428280180574, + "grad_norm": 0.14162059128284454, + "learning_rate": 7.603464164643525e-05, + "loss": 0.0361, + "step": 35940 + }, + { + "epoch": 6.979369933498374, + "grad_norm": 0.182095468044281, + "learning_rate": 7.602052382226507e-05, + "loss": 0.0306, + "step": 35950 + }, + { + "epoch": 6.981311586816174, + "grad_norm": 0.1596217006444931, + "learning_rate": 7.600640315252954e-05, + "loss": 0.0294, + "step": 35960 + }, + { + "epoch": 6.983253240133974, + "grad_norm": 0.17167140543460846, + "learning_rate": 7.599227963877284e-05, + "loss": 0.0282, + "step": 35970 + }, + { + "epoch": 6.985194893451774, + "grad_norm": 0.20545829832553864, + "learning_rate": 7.597815328253954e-05, + "loss": 0.0271, + "step": 35980 + }, + { + "epoch": 6.987136546769574, + "grad_norm": 0.16700923442840576, + "learning_rate": 7.596402408537443e-05, + "loss": 0.0324, + "step": 35990 + }, + { + "epoch": 6.989078200087374, + "grad_norm": 0.11344216018915176, + "learning_rate": 7.59498920488227e-05, + "loss": 0.0229, + "step": 36000 + }, + { + "epoch": 6.991019853405175, + "grad_norm": 0.1392388790845871, + "learning_rate": 7.593575717442979e-05, + "loss": 0.0289, + "step": 36010 + }, + { + "epoch": 6.992961506722975, + "grad_norm": 0.10625333338975906, + "learning_rate": 7.592161946374147e-05, + "loss": 0.029, + "step": 36020 + }, + { + "epoch": 6.994903160040774, + "grad_norm": 0.2035549432039261, + "learning_rate": 7.590747891830381e-05, + "loss": 0.0328, + "step": 36030 + }, + { + "epoch": 6.9968448133585746, + "grad_norm": 0.2085263729095459, + "learning_rate": 7.58933355396632e-05, + "loss": 0.0276, + "step": 36040 + }, + { + "epoch": 6.998786466676375, + "grad_norm": 0.1423065960407257, + "learning_rate": 7.587918932936636e-05, + "loss": 0.0313, + "step": 36050 + }, + { + "epoch": 7.00058249599534, + "grad_norm": 0.2808956503868103, + "learning_rate": 7.586504028896028e-05, + "loss": 0.0311, + "step": 36060 + }, + { + "epoch": 7.00252414931314, + "grad_norm": 0.2745082974433899, + "learning_rate": 7.585088841999228e-05, + "loss": 0.0308, + "step": 36070 + }, + { + "epoch": 7.00446580263094, + "grad_norm": 0.16468863189220428, + "learning_rate": 7.583673372400999e-05, + "loss": 0.0258, + "step": 36080 + }, + { + "epoch": 7.00640745594874, + "grad_norm": 0.14952224493026733, + "learning_rate": 7.582257620256134e-05, + "loss": 0.0318, + "step": 36090 + }, + { + "epoch": 7.008349109266541, + "grad_norm": 0.1987837553024292, + "learning_rate": 7.580841585719458e-05, + "loss": 0.0364, + "step": 36100 + }, + { + "epoch": 7.010290762584341, + "grad_norm": 0.1416655033826828, + "learning_rate": 7.579425268945825e-05, + "loss": 0.0309, + "step": 36110 + }, + { + "epoch": 7.01223241590214, + "grad_norm": 0.20667551457881927, + "learning_rate": 7.578008670090127e-05, + "loss": 0.0296, + "step": 36120 + }, + { + "epoch": 7.014174069219941, + "grad_norm": 0.20384246110916138, + "learning_rate": 7.576591789307272e-05, + "loss": 0.029, + "step": 36130 + }, + { + "epoch": 7.016115722537741, + "grad_norm": 0.11724309623241425, + "learning_rate": 7.575174626752216e-05, + "loss": 0.0244, + "step": 36140 + }, + { + "epoch": 7.018057375855541, + "grad_norm": 0.1394442617893219, + "learning_rate": 7.573757182579934e-05, + "loss": 0.028, + "step": 36150 + }, + { + "epoch": 7.019999029173341, + "grad_norm": 0.18321514129638672, + "learning_rate": 7.572339456945435e-05, + "loss": 0.0275, + "step": 36160 + }, + { + "epoch": 7.021940682491141, + "grad_norm": 0.11061292141675949, + "learning_rate": 7.570921450003762e-05, + "loss": 0.0343, + "step": 36170 + }, + { + "epoch": 7.023882335808941, + "grad_norm": 0.18301844596862793, + "learning_rate": 7.569503161909983e-05, + "loss": 0.026, + "step": 36180 + }, + { + "epoch": 7.025823989126741, + "grad_norm": 0.09318552166223526, + "learning_rate": 7.568084592819202e-05, + "loss": 0.0273, + "step": 36190 + }, + { + "epoch": 7.027765642444542, + "grad_norm": 0.18281152844429016, + "learning_rate": 7.566665742886551e-05, + "loss": 0.0242, + "step": 36200 + }, + { + "epoch": 7.029707295762342, + "grad_norm": 0.16466225683689117, + "learning_rate": 7.565246612267194e-05, + "loss": 0.035, + "step": 36210 + }, + { + "epoch": 7.031648949080142, + "grad_norm": 0.293043851852417, + "learning_rate": 7.563827201116325e-05, + "loss": 0.0263, + "step": 36220 + }, + { + "epoch": 7.0335906023979415, + "grad_norm": 0.23788876831531525, + "learning_rate": 7.562407509589164e-05, + "loss": 0.0339, + "step": 36230 + }, + { + "epoch": 7.035532255715742, + "grad_norm": 0.2032715231180191, + "learning_rate": 7.560987537840973e-05, + "loss": 0.0284, + "step": 36240 + }, + { + "epoch": 7.037473909033542, + "grad_norm": 0.11461687088012695, + "learning_rate": 7.559567286027036e-05, + "loss": 0.0297, + "step": 36250 + }, + { + "epoch": 7.039415562351342, + "grad_norm": 0.22252680361270905, + "learning_rate": 7.558146754302668e-05, + "loss": 0.0283, + "step": 36260 + }, + { + "epoch": 7.041357215669143, + "grad_norm": 0.2031634896993637, + "learning_rate": 7.556725942823217e-05, + "loss": 0.0267, + "step": 36270 + }, + { + "epoch": 7.043298868986942, + "grad_norm": 0.1522190421819687, + "learning_rate": 7.555304851744061e-05, + "loss": 0.0366, + "step": 36280 + }, + { + "epoch": 7.045240522304742, + "grad_norm": 0.2856155335903168, + "learning_rate": 7.553883481220608e-05, + "loss": 0.0368, + "step": 36290 + }, + { + "epoch": 7.047182175622543, + "grad_norm": 0.15315531194210052, + "learning_rate": 7.552461831408298e-05, + "loss": 0.0232, + "step": 36300 + }, + { + "epoch": 7.049123828940343, + "grad_norm": 0.2738569676876068, + "learning_rate": 7.551039902462599e-05, + "loss": 0.0306, + "step": 36310 + }, + { + "epoch": 7.051065482258143, + "grad_norm": 0.22186999022960663, + "learning_rate": 7.549617694539011e-05, + "loss": 0.0313, + "step": 36320 + }, + { + "epoch": 7.0530071355759425, + "grad_norm": 0.1893969625234604, + "learning_rate": 7.548195207793066e-05, + "loss": 0.0299, + "step": 36330 + }, + { + "epoch": 7.054948788893743, + "grad_norm": 0.227090522646904, + "learning_rate": 7.546772442380323e-05, + "loss": 0.0269, + "step": 36340 + }, + { + "epoch": 7.056890442211543, + "grad_norm": 0.2436380386352539, + "learning_rate": 7.545349398456376e-05, + "loss": 0.0277, + "step": 36350 + }, + { + "epoch": 7.058832095529343, + "grad_norm": 0.2574445903301239, + "learning_rate": 7.543926076176845e-05, + "loss": 0.0301, + "step": 36360 + }, + { + "epoch": 7.060773748847144, + "grad_norm": 0.2597096264362335, + "learning_rate": 7.542502475697385e-05, + "loss": 0.033, + "step": 36370 + }, + { + "epoch": 7.062715402164944, + "grad_norm": 0.18964172899723053, + "learning_rate": 7.541078597173675e-05, + "loss": 0.0359, + "step": 36380 + }, + { + "epoch": 7.064657055482743, + "grad_norm": 0.14598259329795837, + "learning_rate": 7.539654440761431e-05, + "loss": 0.024, + "step": 36390 + }, + { + "epoch": 7.0665987088005435, + "grad_norm": 0.20010431110858917, + "learning_rate": 7.538230006616395e-05, + "loss": 0.0275, + "step": 36400 + }, + { + "epoch": 7.068540362118344, + "grad_norm": 0.16244623064994812, + "learning_rate": 7.536805294894342e-05, + "loss": 0.0236, + "step": 36410 + }, + { + "epoch": 7.070482015436144, + "grad_norm": 0.14044030010700226, + "learning_rate": 7.535380305751076e-05, + "loss": 0.0244, + "step": 36420 + }, + { + "epoch": 7.072423668753944, + "grad_norm": 0.19342383742332458, + "learning_rate": 7.533955039342431e-05, + "loss": 0.0244, + "step": 36430 + }, + { + "epoch": 7.074365322071744, + "grad_norm": 0.15717940032482147, + "learning_rate": 7.532529495824274e-05, + "loss": 0.0262, + "step": 36440 + }, + { + "epoch": 7.076306975389544, + "grad_norm": 0.1696855127811432, + "learning_rate": 7.531103675352497e-05, + "loss": 0.0314, + "step": 36450 + }, + { + "epoch": 7.078248628707344, + "grad_norm": 0.1726742535829544, + "learning_rate": 7.529677578083029e-05, + "loss": 0.031, + "step": 36460 + }, + { + "epoch": 7.0801902820251446, + "grad_norm": 0.3349723517894745, + "learning_rate": 7.528251204171823e-05, + "loss": 0.0353, + "step": 36470 + }, + { + "epoch": 7.082131935342945, + "grad_norm": 0.17261473834514618, + "learning_rate": 7.526824553774867e-05, + "loss": 0.0304, + "step": 36480 + }, + { + "epoch": 7.084073588660744, + "grad_norm": 0.1758783906698227, + "learning_rate": 7.525397627048178e-05, + "loss": 0.0271, + "step": 36490 + }, + { + "epoch": 7.0860152419785445, + "grad_norm": 0.150572270154953, + "learning_rate": 7.523970424147802e-05, + "loss": 0.0276, + "step": 36500 + }, + { + "epoch": 7.087956895296345, + "grad_norm": 0.15643933415412903, + "learning_rate": 7.522542945229813e-05, + "loss": 0.0308, + "step": 36510 + }, + { + "epoch": 7.089898548614145, + "grad_norm": 0.29570114612579346, + "learning_rate": 7.521115190450321e-05, + "loss": 0.0274, + "step": 36520 + }, + { + "epoch": 7.091840201931945, + "grad_norm": 0.2876463234424591, + "learning_rate": 7.519687159965462e-05, + "loss": 0.0332, + "step": 36530 + }, + { + "epoch": 7.093781855249746, + "grad_norm": 0.2273808717727661, + "learning_rate": 7.518258853931403e-05, + "loss": 0.0269, + "step": 36540 + }, + { + "epoch": 7.095723508567545, + "grad_norm": 0.17902050912380219, + "learning_rate": 7.516830272504342e-05, + "loss": 0.0262, + "step": 36550 + }, + { + "epoch": 7.097665161885345, + "grad_norm": 0.15267984569072723, + "learning_rate": 7.515401415840509e-05, + "loss": 0.0261, + "step": 36560 + }, + { + "epoch": 7.0996068152031455, + "grad_norm": 0.2055000215768814, + "learning_rate": 7.513972284096155e-05, + "loss": 0.0199, + "step": 36570 + }, + { + "epoch": 7.101548468520946, + "grad_norm": 0.18984580039978027, + "learning_rate": 7.512542877427574e-05, + "loss": 0.0349, + "step": 36580 + }, + { + "epoch": 7.103490121838746, + "grad_norm": 0.1773250550031662, + "learning_rate": 7.51111319599108e-05, + "loss": 0.0321, + "step": 36590 + }, + { + "epoch": 7.1054317751565454, + "grad_norm": 0.2266409546136856, + "learning_rate": 7.509683239943024e-05, + "loss": 0.0372, + "step": 36600 + }, + { + "epoch": 7.107373428474346, + "grad_norm": 0.26992711424827576, + "learning_rate": 7.50825300943978e-05, + "loss": 0.0347, + "step": 36610 + }, + { + "epoch": 7.109315081792146, + "grad_norm": 0.17020843923091888, + "learning_rate": 7.50682250463776e-05, + "loss": 0.0271, + "step": 36620 + }, + { + "epoch": 7.111256735109946, + "grad_norm": 0.1609017699956894, + "learning_rate": 7.5053917256934e-05, + "loss": 0.032, + "step": 36630 + }, + { + "epoch": 7.1131983884277465, + "grad_norm": 0.2556333541870117, + "learning_rate": 7.503960672763166e-05, + "loss": 0.0252, + "step": 36640 + }, + { + "epoch": 7.115140041745546, + "grad_norm": 0.2460157871246338, + "learning_rate": 7.502529346003559e-05, + "loss": 0.0261, + "step": 36650 + }, + { + "epoch": 7.117081695063346, + "grad_norm": 0.17333637177944183, + "learning_rate": 7.501097745571106e-05, + "loss": 0.0317, + "step": 36660 + }, + { + "epoch": 7.1190233483811465, + "grad_norm": 0.21071574091911316, + "learning_rate": 7.499665871622361e-05, + "loss": 0.0301, + "step": 36670 + }, + { + "epoch": 7.120965001698947, + "grad_norm": 0.2835758626461029, + "learning_rate": 7.498233724313917e-05, + "loss": 0.0284, + "step": 36680 + }, + { + "epoch": 7.122906655016747, + "grad_norm": 0.2349063754081726, + "learning_rate": 7.496801303802387e-05, + "loss": 0.0232, + "step": 36690 + }, + { + "epoch": 7.124848308334547, + "grad_norm": 0.26323026418685913, + "learning_rate": 7.495368610244423e-05, + "loss": 0.0278, + "step": 36700 + }, + { + "epoch": 7.126789961652347, + "grad_norm": 0.15506567060947418, + "learning_rate": 7.493935643796697e-05, + "loss": 0.0241, + "step": 36710 + }, + { + "epoch": 7.128731614970147, + "grad_norm": 0.1394454389810562, + "learning_rate": 7.49250240461592e-05, + "loss": 0.028, + "step": 36720 + }, + { + "epoch": 7.130673268287947, + "grad_norm": 0.09993193298578262, + "learning_rate": 7.491068892858826e-05, + "loss": 0.0263, + "step": 36730 + }, + { + "epoch": 7.1326149216057475, + "grad_norm": 0.1903925985097885, + "learning_rate": 7.489635108682184e-05, + "loss": 0.0334, + "step": 36740 + }, + { + "epoch": 7.134556574923548, + "grad_norm": 0.15556831657886505, + "learning_rate": 7.48820105224279e-05, + "loss": 0.0324, + "step": 36750 + }, + { + "epoch": 7.136498228241347, + "grad_norm": 0.22493498027324677, + "learning_rate": 7.48676672369747e-05, + "loss": 0.0305, + "step": 36760 + }, + { + "epoch": 7.138439881559147, + "grad_norm": 0.13288959860801697, + "learning_rate": 7.485332123203079e-05, + "loss": 0.0274, + "step": 36770 + }, + { + "epoch": 7.140381534876948, + "grad_norm": 0.19796417653560638, + "learning_rate": 7.483897250916504e-05, + "loss": 0.0317, + "step": 36780 + }, + { + "epoch": 7.142323188194748, + "grad_norm": 0.17493195831775665, + "learning_rate": 7.48246210699466e-05, + "loss": 0.0324, + "step": 36790 + }, + { + "epoch": 7.144264841512548, + "grad_norm": 0.21407116949558258, + "learning_rate": 7.481026691594492e-05, + "loss": 0.0328, + "step": 36800 + }, + { + "epoch": 7.146206494830348, + "grad_norm": 0.13860024511814117, + "learning_rate": 7.479591004872975e-05, + "loss": 0.031, + "step": 36810 + }, + { + "epoch": 7.148148148148148, + "grad_norm": 0.19148607552051544, + "learning_rate": 7.478155046987114e-05, + "loss": 0.0293, + "step": 36820 + }, + { + "epoch": 7.150089801465948, + "grad_norm": 0.12127980589866638, + "learning_rate": 7.476718818093944e-05, + "loss": 0.0285, + "step": 36830 + }, + { + "epoch": 7.1520314547837485, + "grad_norm": 0.1772058755159378, + "learning_rate": 7.475282318350527e-05, + "loss": 0.0328, + "step": 36840 + }, + { + "epoch": 7.153973108101549, + "grad_norm": 0.13158951699733734, + "learning_rate": 7.473845547913957e-05, + "loss": 0.0211, + "step": 36850 + }, + { + "epoch": 7.155914761419348, + "grad_norm": 0.2615951895713806, + "learning_rate": 7.472408506941357e-05, + "loss": 0.0338, + "step": 36860 + }, + { + "epoch": 7.157856414737148, + "grad_norm": 0.1878732144832611, + "learning_rate": 7.470971195589881e-05, + "loss": 0.025, + "step": 36870 + }, + { + "epoch": 7.159798068054949, + "grad_norm": 0.16335493326187134, + "learning_rate": 7.46953361401671e-05, + "loss": 0.025, + "step": 36880 + }, + { + "epoch": 7.161739721372749, + "grad_norm": 0.2749117612838745, + "learning_rate": 7.468095762379055e-05, + "loss": 0.0344, + "step": 36890 + }, + { + "epoch": 7.163681374690549, + "grad_norm": 0.23469872772693634, + "learning_rate": 7.466657640834158e-05, + "loss": 0.0242, + "step": 36900 + }, + { + "epoch": 7.1656230280083495, + "grad_norm": 0.23131944239139557, + "learning_rate": 7.46521924953929e-05, + "loss": 0.0295, + "step": 36910 + }, + { + "epoch": 7.167564681326149, + "grad_norm": 0.23185792565345764, + "learning_rate": 7.463780588651752e-05, + "loss": 0.0356, + "step": 36920 + }, + { + "epoch": 7.169506334643949, + "grad_norm": 0.1926761120557785, + "learning_rate": 7.462341658328873e-05, + "loss": 0.0262, + "step": 36930 + }, + { + "epoch": 7.171447987961749, + "grad_norm": 0.09373583644628525, + "learning_rate": 7.460902458728012e-05, + "loss": 0.0226, + "step": 36940 + }, + { + "epoch": 7.17338964127955, + "grad_norm": 0.15113665163516998, + "learning_rate": 7.459462990006558e-05, + "loss": 0.0252, + "step": 36950 + }, + { + "epoch": 7.17533129459735, + "grad_norm": 0.22303883731365204, + "learning_rate": 7.458023252321929e-05, + "loss": 0.0237, + "step": 36960 + }, + { + "epoch": 7.177272947915149, + "grad_norm": 0.20018042623996735, + "learning_rate": 7.456583245831573e-05, + "loss": 0.0283, + "step": 36970 + }, + { + "epoch": 7.17921460123295, + "grad_norm": 0.10329657793045044, + "learning_rate": 7.455142970692968e-05, + "loss": 0.0294, + "step": 36980 + }, + { + "epoch": 7.18115625455075, + "grad_norm": 0.11039308458566666, + "learning_rate": 7.453702427063615e-05, + "loss": 0.0328, + "step": 36990 + }, + { + "epoch": 7.18309790786855, + "grad_norm": 0.19120389223098755, + "learning_rate": 7.452261615101057e-05, + "loss": 0.0315, + "step": 37000 + }, + { + "epoch": 7.1850395611863505, + "grad_norm": 0.11957593262195587, + "learning_rate": 7.450820534962856e-05, + "loss": 0.0235, + "step": 37010 + }, + { + "epoch": 7.186981214504151, + "grad_norm": 0.12496119737625122, + "learning_rate": 7.449379186806604e-05, + "loss": 0.0249, + "step": 37020 + }, + { + "epoch": 7.18892286782195, + "grad_norm": 0.17755578458309174, + "learning_rate": 7.44793757078993e-05, + "loss": 0.0333, + "step": 37030 + }, + { + "epoch": 7.19086452113975, + "grad_norm": 0.15496380627155304, + "learning_rate": 7.44649568707048e-05, + "loss": 0.0288, + "step": 37040 + }, + { + "epoch": 7.192806174457551, + "grad_norm": 0.21483877301216125, + "learning_rate": 7.445053535805942e-05, + "loss": 0.0291, + "step": 37050 + }, + { + "epoch": 7.194747827775351, + "grad_norm": 0.16811582446098328, + "learning_rate": 7.443611117154024e-05, + "loss": 0.0218, + "step": 37060 + }, + { + "epoch": 7.196689481093151, + "grad_norm": 0.15685203671455383, + "learning_rate": 7.442168431272469e-05, + "loss": 0.0326, + "step": 37070 + }, + { + "epoch": 7.198631134410951, + "grad_norm": 0.194184347987175, + "learning_rate": 7.440725478319044e-05, + "loss": 0.0352, + "step": 37080 + }, + { + "epoch": 7.200572787728751, + "grad_norm": 0.16173462569713593, + "learning_rate": 7.439282258451553e-05, + "loss": 0.0323, + "step": 37090 + }, + { + "epoch": 7.202514441046551, + "grad_norm": 0.31155455112457275, + "learning_rate": 7.43783877182782e-05, + "loss": 0.026, + "step": 37100 + }, + { + "epoch": 7.204456094364351, + "grad_norm": 0.24218447506427765, + "learning_rate": 7.436395018605707e-05, + "loss": 0.0214, + "step": 37110 + }, + { + "epoch": 7.206397747682152, + "grad_norm": 0.10226308554410934, + "learning_rate": 7.434950998943094e-05, + "loss": 0.0351, + "step": 37120 + }, + { + "epoch": 7.208339400999951, + "grad_norm": 0.20702971518039703, + "learning_rate": 7.433506712997902e-05, + "loss": 0.0271, + "step": 37130 + }, + { + "epoch": 7.210281054317751, + "grad_norm": 0.17925406992435455, + "learning_rate": 7.432062160928077e-05, + "loss": 0.0271, + "step": 37140 + }, + { + "epoch": 7.212222707635552, + "grad_norm": 0.1732388287782669, + "learning_rate": 7.430617342891588e-05, + "loss": 0.0214, + "step": 37150 + }, + { + "epoch": 7.214164360953352, + "grad_norm": 0.15463852882385254, + "learning_rate": 7.429172259046442e-05, + "loss": 0.0246, + "step": 37160 + }, + { + "epoch": 7.216106014271152, + "grad_norm": 0.13102708756923676, + "learning_rate": 7.42772690955067e-05, + "loss": 0.0342, + "step": 37170 + }, + { + "epoch": 7.2180476675889516, + "grad_norm": 0.14300768077373505, + "learning_rate": 7.426281294562332e-05, + "loss": 0.031, + "step": 37180 + }, + { + "epoch": 7.219989320906752, + "grad_norm": 0.24371610581874847, + "learning_rate": 7.424835414239519e-05, + "loss": 0.0344, + "step": 37190 + }, + { + "epoch": 7.221930974224552, + "grad_norm": 0.24510474503040314, + "learning_rate": 7.423389268740352e-05, + "loss": 0.0277, + "step": 37200 + }, + { + "epoch": 7.223872627542352, + "grad_norm": 0.1359211653470993, + "learning_rate": 7.42194285822298e-05, + "loss": 0.0308, + "step": 37210 + }, + { + "epoch": 7.225814280860153, + "grad_norm": 0.19208283722400665, + "learning_rate": 7.420496182845576e-05, + "loss": 0.031, + "step": 37220 + }, + { + "epoch": 7.227755934177953, + "grad_norm": 0.21595372259616852, + "learning_rate": 7.419049242766348e-05, + "loss": 0.0305, + "step": 37230 + }, + { + "epoch": 7.229697587495752, + "grad_norm": 0.10903077572584152, + "learning_rate": 7.417602038143533e-05, + "loss": 0.026, + "step": 37240 + }, + { + "epoch": 7.231639240813553, + "grad_norm": 0.1095404326915741, + "learning_rate": 7.416154569135393e-05, + "loss": 0.0252, + "step": 37250 + }, + { + "epoch": 7.233580894131353, + "grad_norm": 0.19404231011867523, + "learning_rate": 7.414706835900222e-05, + "loss": 0.0277, + "step": 37260 + }, + { + "epoch": 7.235522547449153, + "grad_norm": 0.10942542552947998, + "learning_rate": 7.413258838596345e-05, + "loss": 0.0261, + "step": 37270 + }, + { + "epoch": 7.237464200766953, + "grad_norm": 0.39146745204925537, + "learning_rate": 7.411810577382106e-05, + "loss": 0.0291, + "step": 37280 + }, + { + "epoch": 7.239405854084753, + "grad_norm": 0.19801047444343567, + "learning_rate": 7.410362052415889e-05, + "loss": 0.0274, + "step": 37290 + }, + { + "epoch": 7.241347507402553, + "grad_norm": 0.18719618022441864, + "learning_rate": 7.408913263856102e-05, + "loss": 0.0257, + "step": 37300 + }, + { + "epoch": 7.243289160720353, + "grad_norm": 0.19472533464431763, + "learning_rate": 7.407464211861183e-05, + "loss": 0.024, + "step": 37310 + }, + { + "epoch": 7.245230814038154, + "grad_norm": 0.1194811761379242, + "learning_rate": 7.406014896589597e-05, + "loss": 0.0256, + "step": 37320 + }, + { + "epoch": 7.247172467355954, + "grad_norm": 0.24261267483234406, + "learning_rate": 7.404565318199838e-05, + "loss": 0.0307, + "step": 37330 + }, + { + "epoch": 7.249114120673754, + "grad_norm": 0.17855796217918396, + "learning_rate": 7.403115476850432e-05, + "loss": 0.0322, + "step": 37340 + }, + { + "epoch": 7.2510557739915535, + "grad_norm": 0.2223322093486786, + "learning_rate": 7.401665372699932e-05, + "loss": 0.0379, + "step": 37350 + }, + { + "epoch": 7.252997427309354, + "grad_norm": 0.13753828406333923, + "learning_rate": 7.400215005906916e-05, + "loss": 0.0344, + "step": 37360 + }, + { + "epoch": 7.254939080627154, + "grad_norm": 0.21891586482524872, + "learning_rate": 7.398764376629996e-05, + "loss": 0.0267, + "step": 37370 + }, + { + "epoch": 7.256880733944954, + "grad_norm": 0.22092044353485107, + "learning_rate": 7.39731348502781e-05, + "loss": 0.0269, + "step": 37380 + }, + { + "epoch": 7.258822387262755, + "grad_norm": 0.16126857697963715, + "learning_rate": 7.395862331259028e-05, + "loss": 0.0288, + "step": 37390 + }, + { + "epoch": 7.260764040580554, + "grad_norm": 0.09445105493068695, + "learning_rate": 7.394410915482341e-05, + "loss": 0.0322, + "step": 37400 + }, + { + "epoch": 7.262705693898354, + "grad_norm": 0.20297132432460785, + "learning_rate": 7.392959237856476e-05, + "loss": 0.0306, + "step": 37410 + }, + { + "epoch": 7.264647347216155, + "grad_norm": 0.13751693069934845, + "learning_rate": 7.391507298540187e-05, + "loss": 0.0282, + "step": 37420 + }, + { + "epoch": 7.266589000533955, + "grad_norm": 0.14093804359436035, + "learning_rate": 7.390055097692257e-05, + "loss": 0.0274, + "step": 37430 + }, + { + "epoch": 7.268530653851755, + "grad_norm": 0.12985995411872864, + "learning_rate": 7.388602635471494e-05, + "loss": 0.0265, + "step": 37440 + }, + { + "epoch": 7.2704723071695545, + "grad_norm": 0.21893537044525146, + "learning_rate": 7.387149912036738e-05, + "loss": 0.0279, + "step": 37450 + }, + { + "epoch": 7.272413960487355, + "grad_norm": 0.08275198191404343, + "learning_rate": 7.385696927546854e-05, + "loss": 0.023, + "step": 37460 + }, + { + "epoch": 7.274355613805155, + "grad_norm": 0.17548921704292297, + "learning_rate": 7.384243682160743e-05, + "loss": 0.0283, + "step": 37470 + }, + { + "epoch": 7.276297267122955, + "grad_norm": 0.2265455275774002, + "learning_rate": 7.382790176037329e-05, + "loss": 0.0241, + "step": 37480 + }, + { + "epoch": 7.278238920440756, + "grad_norm": 0.09379564970731735, + "learning_rate": 7.381336409335562e-05, + "loss": 0.023, + "step": 37490 + }, + { + "epoch": 7.280180573758555, + "grad_norm": 0.225638285279274, + "learning_rate": 7.379882382214426e-05, + "loss": 0.0266, + "step": 37500 + }, + { + "epoch": 7.282122227076355, + "grad_norm": 0.157902792096138, + "learning_rate": 7.378428094832931e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 7.2840638803941555, + "grad_norm": 0.10365783423185349, + "learning_rate": 7.376973547350114e-05, + "loss": 0.0292, + "step": 37520 + }, + { + "epoch": 7.286005533711956, + "grad_norm": 0.14230775833129883, + "learning_rate": 7.375518739925046e-05, + "loss": 0.0259, + "step": 37530 + }, + { + "epoch": 7.287947187029756, + "grad_norm": 0.21131393313407898, + "learning_rate": 7.374063672716818e-05, + "loss": 0.0303, + "step": 37540 + }, + { + "epoch": 7.289888840347556, + "grad_norm": 0.28438031673431396, + "learning_rate": 7.372608345884558e-05, + "loss": 0.0233, + "step": 37550 + }, + { + "epoch": 7.291830493665356, + "grad_norm": 0.16904133558273315, + "learning_rate": 7.371152759587415e-05, + "loss": 0.0285, + "step": 37560 + }, + { + "epoch": 7.293772146983156, + "grad_norm": 0.17062941193580627, + "learning_rate": 7.369696913984572e-05, + "loss": 0.0294, + "step": 37570 + }, + { + "epoch": 7.295713800300956, + "grad_norm": 0.1945980191230774, + "learning_rate": 7.368240809235237e-05, + "loss": 0.0263, + "step": 37580 + }, + { + "epoch": 7.297655453618757, + "grad_norm": 0.2274242639541626, + "learning_rate": 7.366784445498649e-05, + "loss": 0.0376, + "step": 37590 + }, + { + "epoch": 7.299597106936557, + "grad_norm": 0.13020992279052734, + "learning_rate": 7.36532782293407e-05, + "loss": 0.0246, + "step": 37600 + }, + { + "epoch": 7.301538760254356, + "grad_norm": 0.21479085087776184, + "learning_rate": 7.363870941700797e-05, + "loss": 0.0327, + "step": 37610 + }, + { + "epoch": 7.3034804135721565, + "grad_norm": 0.23161081969738007, + "learning_rate": 7.362413801958152e-05, + "loss": 0.0313, + "step": 37620 + }, + { + "epoch": 7.305422066889957, + "grad_norm": 0.16649234294891357, + "learning_rate": 7.360956403865486e-05, + "loss": 0.0255, + "step": 37630 + }, + { + "epoch": 7.307363720207757, + "grad_norm": 0.25625526905059814, + "learning_rate": 7.359498747582177e-05, + "loss": 0.0267, + "step": 37640 + }, + { + "epoch": 7.309305373525557, + "grad_norm": 0.16099347174167633, + "learning_rate": 7.358040833267634e-05, + "loss": 0.0309, + "step": 37650 + }, + { + "epoch": 7.311247026843358, + "grad_norm": 0.161004438996315, + "learning_rate": 7.356582661081288e-05, + "loss": 0.0294, + "step": 37660 + }, + { + "epoch": 7.313188680161157, + "grad_norm": 0.13343043625354767, + "learning_rate": 7.355124231182607e-05, + "loss": 0.0286, + "step": 37670 + }, + { + "epoch": 7.315130333478957, + "grad_norm": 0.1696847379207611, + "learning_rate": 7.353665543731083e-05, + "loss": 0.0276, + "step": 37680 + }, + { + "epoch": 7.3170719867967575, + "grad_norm": 0.17915315926074982, + "learning_rate": 7.352206598886229e-05, + "loss": 0.0297, + "step": 37690 + }, + { + "epoch": 7.319013640114558, + "grad_norm": 0.16830278933048248, + "learning_rate": 7.350747396807601e-05, + "loss": 0.0295, + "step": 37700 + }, + { + "epoch": 7.320955293432358, + "grad_norm": 0.1170060858130455, + "learning_rate": 7.349287937654772e-05, + "loss": 0.0274, + "step": 37710 + }, + { + "epoch": 7.3228969467501575, + "grad_norm": 0.16797856986522675, + "learning_rate": 7.347828221587345e-05, + "loss": 0.0281, + "step": 37720 + }, + { + "epoch": 7.324838600067958, + "grad_norm": 0.1277090311050415, + "learning_rate": 7.346368248764955e-05, + "loss": 0.0261, + "step": 37730 + }, + { + "epoch": 7.326780253385758, + "grad_norm": 0.21857532858848572, + "learning_rate": 7.34490801934726e-05, + "loss": 0.0313, + "step": 37740 + }, + { + "epoch": 7.328721906703558, + "grad_norm": 0.13978753983974457, + "learning_rate": 7.343447533493947e-05, + "loss": 0.0339, + "step": 37750 + }, + { + "epoch": 7.3306635600213585, + "grad_norm": 0.19928598403930664, + "learning_rate": 7.341986791364739e-05, + "loss": 0.025, + "step": 37760 + }, + { + "epoch": 7.332605213339158, + "grad_norm": 0.17499548196792603, + "learning_rate": 7.340525793119374e-05, + "loss": 0.0302, + "step": 37770 + }, + { + "epoch": 7.334546866656958, + "grad_norm": 0.17045335471630096, + "learning_rate": 7.33906453891763e-05, + "loss": 0.0215, + "step": 37780 + }, + { + "epoch": 7.3364885199747585, + "grad_norm": 0.15448226034641266, + "learning_rate": 7.337603028919303e-05, + "loss": 0.0244, + "step": 37790 + }, + { + "epoch": 7.338430173292559, + "grad_norm": 0.1516563445329666, + "learning_rate": 7.336141263284225e-05, + "loss": 0.0299, + "step": 37800 + }, + { + "epoch": 7.340371826610359, + "grad_norm": 0.24757015705108643, + "learning_rate": 7.334679242172249e-05, + "loss": 0.0276, + "step": 37810 + }, + { + "epoch": 7.342313479928158, + "grad_norm": 0.39398857951164246, + "learning_rate": 7.33321696574326e-05, + "loss": 0.0273, + "step": 37820 + }, + { + "epoch": 7.344255133245959, + "grad_norm": 0.19090500473976135, + "learning_rate": 7.331754434157174e-05, + "loss": 0.0314, + "step": 37830 + }, + { + "epoch": 7.346196786563759, + "grad_norm": 0.18344014883041382, + "learning_rate": 7.330291647573928e-05, + "loss": 0.0321, + "step": 37840 + }, + { + "epoch": 7.348138439881559, + "grad_norm": 0.15279309451580048, + "learning_rate": 7.32882860615349e-05, + "loss": 0.0284, + "step": 37850 + }, + { + "epoch": 7.3500800931993595, + "grad_norm": 0.2010318636894226, + "learning_rate": 7.327365310055858e-05, + "loss": 0.0359, + "step": 37860 + }, + { + "epoch": 7.35202174651716, + "grad_norm": 0.16939377784729004, + "learning_rate": 7.325901759441055e-05, + "loss": 0.026, + "step": 37870 + }, + { + "epoch": 7.353963399834959, + "grad_norm": 0.1851704716682434, + "learning_rate": 7.324437954469132e-05, + "loss": 0.0296, + "step": 37880 + }, + { + "epoch": 7.355905053152759, + "grad_norm": 0.2139330506324768, + "learning_rate": 7.32297389530017e-05, + "loss": 0.0281, + "step": 37890 + }, + { + "epoch": 7.35784670647056, + "grad_norm": 0.21079206466674805, + "learning_rate": 7.321509582094274e-05, + "loss": 0.0252, + "step": 37900 + }, + { + "epoch": 7.35978835978836, + "grad_norm": 0.40090397000312805, + "learning_rate": 7.320045015011582e-05, + "loss": 0.0293, + "step": 37910 + }, + { + "epoch": 7.36173001310616, + "grad_norm": 0.17527323961257935, + "learning_rate": 7.318580194212253e-05, + "loss": 0.0318, + "step": 37920 + }, + { + "epoch": 7.36367166642396, + "grad_norm": 0.12055344134569168, + "learning_rate": 7.317115119856483e-05, + "loss": 0.0281, + "step": 37930 + }, + { + "epoch": 7.36561331974176, + "grad_norm": 0.17892564833164215, + "learning_rate": 7.315649792104485e-05, + "loss": 0.0314, + "step": 37940 + }, + { + "epoch": 7.36755497305956, + "grad_norm": 0.2492782175540924, + "learning_rate": 7.314184211116507e-05, + "loss": 0.025, + "step": 37950 + }, + { + "epoch": 7.3694966263773605, + "grad_norm": 0.10046923905611038, + "learning_rate": 7.312718377052823e-05, + "loss": 0.0279, + "step": 37960 + }, + { + "epoch": 7.371438279695161, + "grad_norm": 0.11894357949495316, + "learning_rate": 7.311252290073737e-05, + "loss": 0.0265, + "step": 37970 + }, + { + "epoch": 7.37337993301296, + "grad_norm": 0.08404558151960373, + "learning_rate": 7.309785950339572e-05, + "loss": 0.0253, + "step": 37980 + }, + { + "epoch": 7.37532158633076, + "grad_norm": 0.16629022359848022, + "learning_rate": 7.308319358010692e-05, + "loss": 0.0191, + "step": 37990 + }, + { + "epoch": 7.377263239648561, + "grad_norm": 0.13787224888801575, + "learning_rate": 7.306852513247474e-05, + "loss": 0.0272, + "step": 38000 + }, + { + "epoch": 7.379204892966361, + "grad_norm": 0.11125738173723221, + "learning_rate": 7.305385416210336e-05, + "loss": 0.0262, + "step": 38010 + }, + { + "epoch": 7.381146546284161, + "grad_norm": 0.21288533508777618, + "learning_rate": 7.303918067059711e-05, + "loss": 0.0238, + "step": 38020 + }, + { + "epoch": 7.383088199601961, + "grad_norm": 0.20718561112880707, + "learning_rate": 7.302450465956075e-05, + "loss": 0.0321, + "step": 38030 + }, + { + "epoch": 7.385029852919761, + "grad_norm": 0.1748870313167572, + "learning_rate": 7.300982613059914e-05, + "loss": 0.0325, + "step": 38040 + }, + { + "epoch": 7.386971506237561, + "grad_norm": 0.28297922015190125, + "learning_rate": 7.299514508531757e-05, + "loss": 0.0311, + "step": 38050 + }, + { + "epoch": 7.388913159555361, + "grad_norm": 0.27939310669898987, + "learning_rate": 7.298046152532148e-05, + "loss": 0.0315, + "step": 38060 + }, + { + "epoch": 7.390854812873162, + "grad_norm": 0.22359076142311096, + "learning_rate": 7.296577545221668e-05, + "loss": 0.0347, + "step": 38070 + }, + { + "epoch": 7.392796466190962, + "grad_norm": 0.14256873726844788, + "learning_rate": 7.29510868676092e-05, + "loss": 0.0278, + "step": 38080 + }, + { + "epoch": 7.394738119508761, + "grad_norm": 0.17994390428066254, + "learning_rate": 7.293639577310538e-05, + "loss": 0.0327, + "step": 38090 + }, + { + "epoch": 7.396679772826562, + "grad_norm": 0.18644799292087555, + "learning_rate": 7.292170217031179e-05, + "loss": 0.0219, + "step": 38100 + }, + { + "epoch": 7.398621426144362, + "grad_norm": 0.16376811265945435, + "learning_rate": 7.290700606083532e-05, + "loss": 0.0244, + "step": 38110 + }, + { + "epoch": 7.400563079462162, + "grad_norm": 0.1772133857011795, + "learning_rate": 7.28923074462831e-05, + "loss": 0.0229, + "step": 38120 + }, + { + "epoch": 7.4025047327799625, + "grad_norm": 0.1788024753332138, + "learning_rate": 7.287760632826255e-05, + "loss": 0.025, + "step": 38130 + }, + { + "epoch": 7.404446386097762, + "grad_norm": 0.19337061047554016, + "learning_rate": 7.286290270838139e-05, + "loss": 0.0324, + "step": 38140 + }, + { + "epoch": 7.406388039415562, + "grad_norm": 0.12140896916389465, + "learning_rate": 7.284819658824756e-05, + "loss": 0.0261, + "step": 38150 + }, + { + "epoch": 7.408329692733362, + "grad_norm": 0.21204660832881927, + "learning_rate": 7.283348796946929e-05, + "loss": 0.0274, + "step": 38160 + }, + { + "epoch": 7.410271346051163, + "grad_norm": 0.09978124499320984, + "learning_rate": 7.28187768536551e-05, + "loss": 0.0285, + "step": 38170 + }, + { + "epoch": 7.412212999368963, + "grad_norm": 0.11657197028398514, + "learning_rate": 7.280406324241379e-05, + "loss": 0.0281, + "step": 38180 + }, + { + "epoch": 7.414154652686763, + "grad_norm": 0.2134203165769577, + "learning_rate": 7.278934713735441e-05, + "loss": 0.0288, + "step": 38190 + }, + { + "epoch": 7.416096306004563, + "grad_norm": 0.16985288262367249, + "learning_rate": 7.277462854008629e-05, + "loss": 0.0253, + "step": 38200 + }, + { + "epoch": 7.418037959322363, + "grad_norm": 0.21608410775661469, + "learning_rate": 7.275990745221904e-05, + "loss": 0.0339, + "step": 38210 + }, + { + "epoch": 7.419979612640163, + "grad_norm": 0.22138656675815582, + "learning_rate": 7.274518387536252e-05, + "loss": 0.025, + "step": 38220 + }, + { + "epoch": 7.421921265957963, + "grad_norm": 0.17093485593795776, + "learning_rate": 7.273045781112689e-05, + "loss": 0.0249, + "step": 38230 + }, + { + "epoch": 7.423862919275764, + "grad_norm": 0.2048567235469818, + "learning_rate": 7.271572926112257e-05, + "loss": 0.0307, + "step": 38240 + }, + { + "epoch": 7.425804572593563, + "grad_norm": 0.16129560768604279, + "learning_rate": 7.270099822696024e-05, + "loss": 0.027, + "step": 38250 + }, + { + "epoch": 7.427746225911363, + "grad_norm": 0.19752447307109833, + "learning_rate": 7.26862647102509e-05, + "loss": 0.0293, + "step": 38260 + }, + { + "epoch": 7.429687879229164, + "grad_norm": 0.203399196267128, + "learning_rate": 7.267152871260573e-05, + "loss": 0.03, + "step": 38270 + }, + { + "epoch": 7.431629532546964, + "grad_norm": 0.10940667241811752, + "learning_rate": 7.265679023563629e-05, + "loss": 0.0277, + "step": 38280 + }, + { + "epoch": 7.433571185864764, + "grad_norm": 0.16157923638820648, + "learning_rate": 7.264204928095433e-05, + "loss": 0.0258, + "step": 38290 + }, + { + "epoch": 7.435512839182564, + "grad_norm": 0.15131616592407227, + "learning_rate": 7.262730585017188e-05, + "loss": 0.0301, + "step": 38300 + }, + { + "epoch": 7.437454492500364, + "grad_norm": 0.188639834523201, + "learning_rate": 7.261255994490131e-05, + "loss": 0.0269, + "step": 38310 + }, + { + "epoch": 7.439396145818164, + "grad_norm": 0.26292020082473755, + "learning_rate": 7.259781156675517e-05, + "loss": 0.028, + "step": 38320 + }, + { + "epoch": 7.441337799135964, + "grad_norm": 0.14634071290493011, + "learning_rate": 7.258306071734633e-05, + "loss": 0.0295, + "step": 38330 + }, + { + "epoch": 7.443279452453765, + "grad_norm": 0.1751640886068344, + "learning_rate": 7.256830739828792e-05, + "loss": 0.0344, + "step": 38340 + }, + { + "epoch": 7.445221105771564, + "grad_norm": 0.14195410907268524, + "learning_rate": 7.255355161119336e-05, + "loss": 0.0303, + "step": 38350 + }, + { + "epoch": 7.447162759089364, + "grad_norm": 0.21670468151569366, + "learning_rate": 7.253879335767628e-05, + "loss": 0.0319, + "step": 38360 + }, + { + "epoch": 7.449104412407165, + "grad_norm": 0.1569194197654724, + "learning_rate": 7.252403263935067e-05, + "loss": 0.0298, + "step": 38370 + }, + { + "epoch": 7.451046065724965, + "grad_norm": 0.13375067710876465, + "learning_rate": 7.25092694578307e-05, + "loss": 0.026, + "step": 38380 + }, + { + "epoch": 7.452987719042765, + "grad_norm": 0.18648791313171387, + "learning_rate": 7.249450381473087e-05, + "loss": 0.0321, + "step": 38390 + }, + { + "epoch": 7.454929372360565, + "grad_norm": 0.17022408545017242, + "learning_rate": 7.247973571166593e-05, + "loss": 0.028, + "step": 38400 + }, + { + "epoch": 7.456871025678365, + "grad_norm": 0.13525988161563873, + "learning_rate": 7.246496515025089e-05, + "loss": 0.0248, + "step": 38410 + }, + { + "epoch": 7.458812678996165, + "grad_norm": 0.26117491722106934, + "learning_rate": 7.245019213210101e-05, + "loss": 0.0238, + "step": 38420 + }, + { + "epoch": 7.460754332313965, + "grad_norm": 0.3246883451938629, + "learning_rate": 7.24354166588319e-05, + "loss": 0.0307, + "step": 38430 + }, + { + "epoch": 7.462695985631766, + "grad_norm": 0.16524097323417664, + "learning_rate": 7.242063873205935e-05, + "loss": 0.0201, + "step": 38440 + }, + { + "epoch": 7.464637638949566, + "grad_norm": 0.16053932905197144, + "learning_rate": 7.240585835339946e-05, + "loss": 0.0286, + "step": 38450 + }, + { + "epoch": 7.466579292267365, + "grad_norm": 0.19162580370903015, + "learning_rate": 7.239107552446857e-05, + "loss": 0.0278, + "step": 38460 + }, + { + "epoch": 7.4685209455851655, + "grad_norm": 0.15466706454753876, + "learning_rate": 7.237629024688334e-05, + "loss": 0.0253, + "step": 38470 + }, + { + "epoch": 7.470462598902966, + "grad_norm": 0.17739331722259521, + "learning_rate": 7.236150252226064e-05, + "loss": 0.0269, + "step": 38480 + }, + { + "epoch": 7.472404252220766, + "grad_norm": 0.22527258098125458, + "learning_rate": 7.234671235221765e-05, + "loss": 0.0305, + "step": 38490 + }, + { + "epoch": 7.474345905538566, + "grad_norm": 0.17901580035686493, + "learning_rate": 7.233191973837179e-05, + "loss": 0.0354, + "step": 38500 + }, + { + "epoch": 7.476287558856367, + "grad_norm": 0.23073461651802063, + "learning_rate": 7.231712468234078e-05, + "loss": 0.0291, + "step": 38510 + }, + { + "epoch": 7.478229212174166, + "grad_norm": 0.23415473103523254, + "learning_rate": 7.230232718574254e-05, + "loss": 0.0294, + "step": 38520 + }, + { + "epoch": 7.480170865491966, + "grad_norm": 0.22971220314502716, + "learning_rate": 7.228752725019535e-05, + "loss": 0.024, + "step": 38530 + }, + { + "epoch": 7.482112518809767, + "grad_norm": 0.12869879603385925, + "learning_rate": 7.227272487731769e-05, + "loss": 0.0307, + "step": 38540 + }, + { + "epoch": 7.484054172127567, + "grad_norm": 0.18661995232105255, + "learning_rate": 7.225792006872831e-05, + "loss": 0.0248, + "step": 38550 + }, + { + "epoch": 7.485995825445367, + "grad_norm": 0.11305641382932663, + "learning_rate": 7.224311282604628e-05, + "loss": 0.0344, + "step": 38560 + }, + { + "epoch": 7.4879374787631665, + "grad_norm": 0.19101516902446747, + "learning_rate": 7.222830315089085e-05, + "loss": 0.0271, + "step": 38570 + }, + { + "epoch": 7.489879132080967, + "grad_norm": 0.25325554609298706, + "learning_rate": 7.22134910448816e-05, + "loss": 0.024, + "step": 38580 + }, + { + "epoch": 7.491820785398767, + "grad_norm": 0.18223141133785248, + "learning_rate": 7.219867650963839e-05, + "loss": 0.0329, + "step": 38590 + }, + { + "epoch": 7.493762438716567, + "grad_norm": 0.11203067749738693, + "learning_rate": 7.21838595467813e-05, + "loss": 0.0303, + "step": 38600 + }, + { + "epoch": 7.495704092034368, + "grad_norm": 0.19016870856285095, + "learning_rate": 7.216904015793067e-05, + "loss": 0.0256, + "step": 38610 + }, + { + "epoch": 7.497645745352167, + "grad_norm": 0.19430950284004211, + "learning_rate": 7.215421834470713e-05, + "loss": 0.032, + "step": 38620 + }, + { + "epoch": 7.499587398669967, + "grad_norm": 0.35830214619636536, + "learning_rate": 7.213939410873158e-05, + "loss": 0.0291, + "step": 38630 + }, + { + "epoch": 7.5015290519877675, + "grad_norm": 0.2358475625514984, + "learning_rate": 7.212456745162519e-05, + "loss": 0.0216, + "step": 38640 + }, + { + "epoch": 7.503470705305568, + "grad_norm": 0.19295962154865265, + "learning_rate": 7.210973837500937e-05, + "loss": 0.0261, + "step": 38650 + }, + { + "epoch": 7.505412358623368, + "grad_norm": 0.23490378260612488, + "learning_rate": 7.209490688050578e-05, + "loss": 0.0269, + "step": 38660 + }, + { + "epoch": 7.5073540119411675, + "grad_norm": 0.1510562300682068, + "learning_rate": 7.208007296973641e-05, + "loss": 0.0251, + "step": 38670 + }, + { + "epoch": 7.509295665258968, + "grad_norm": 0.28068211674690247, + "learning_rate": 7.206523664432345e-05, + "loss": 0.0301, + "step": 38680 + }, + { + "epoch": 7.511237318576768, + "grad_norm": 0.16075563430786133, + "learning_rate": 7.205039790588939e-05, + "loss": 0.0268, + "step": 38690 + }, + { + "epoch": 7.513178971894568, + "grad_norm": 0.1785326898097992, + "learning_rate": 7.203555675605697e-05, + "loss": 0.0284, + "step": 38700 + }, + { + "epoch": 7.515120625212369, + "grad_norm": 0.141700878739357, + "learning_rate": 7.202071319644917e-05, + "loss": 0.0316, + "step": 38710 + }, + { + "epoch": 7.517062278530169, + "grad_norm": 0.25841203331947327, + "learning_rate": 7.200586722868932e-05, + "loss": 0.0234, + "step": 38720 + }, + { + "epoch": 7.519003931847968, + "grad_norm": 0.21759283542633057, + "learning_rate": 7.199101885440088e-05, + "loss": 0.0266, + "step": 38730 + }, + { + "epoch": 7.5209455851657685, + "grad_norm": 0.15198922157287598, + "learning_rate": 7.197616807520771e-05, + "loss": 0.0245, + "step": 38740 + }, + { + "epoch": 7.522887238483569, + "grad_norm": 0.18797026574611664, + "learning_rate": 7.196131489273381e-05, + "loss": 0.0271, + "step": 38750 + }, + { + "epoch": 7.524828891801369, + "grad_norm": 0.15660132467746735, + "learning_rate": 7.194645930860355e-05, + "loss": 0.028, + "step": 38760 + }, + { + "epoch": 7.526770545119169, + "grad_norm": 0.14730234444141388, + "learning_rate": 7.19316013244415e-05, + "loss": 0.0281, + "step": 38770 + }, + { + "epoch": 7.528712198436969, + "grad_norm": 0.1771114021539688, + "learning_rate": 7.191674094187248e-05, + "loss": 0.0281, + "step": 38780 + }, + { + "epoch": 7.530653851754769, + "grad_norm": 0.1823849081993103, + "learning_rate": 7.190187816252165e-05, + "loss": 0.0298, + "step": 38790 + }, + { + "epoch": 7.532595505072569, + "grad_norm": 0.2283593714237213, + "learning_rate": 7.188701298801435e-05, + "loss": 0.0277, + "step": 38800 + }, + { + "epoch": 7.5345371583903695, + "grad_norm": 0.2595245838165283, + "learning_rate": 7.18721454199762e-05, + "loss": 0.0358, + "step": 38810 + }, + { + "epoch": 7.53647881170817, + "grad_norm": 0.2515411078929901, + "learning_rate": 7.185727546003312e-05, + "loss": 0.0244, + "step": 38820 + }, + { + "epoch": 7.53842046502597, + "grad_norm": 0.16943877935409546, + "learning_rate": 7.184240310981126e-05, + "loss": 0.034, + "step": 38830 + }, + { + "epoch": 7.5403621183437695, + "grad_norm": 0.1495865434408188, + "learning_rate": 7.182752837093704e-05, + "loss": 0.0272, + "step": 38840 + }, + { + "epoch": 7.54230377166157, + "grad_norm": 0.2410697191953659, + "learning_rate": 7.181265124503711e-05, + "loss": 0.0268, + "step": 38850 + }, + { + "epoch": 7.54424542497937, + "grad_norm": 0.1400655210018158, + "learning_rate": 7.179777173373847e-05, + "loss": 0.0234, + "step": 38860 + }, + { + "epoch": 7.54618707829717, + "grad_norm": 0.21965433657169342, + "learning_rate": 7.178288983866826e-05, + "loss": 0.0247, + "step": 38870 + }, + { + "epoch": 7.54812873161497, + "grad_norm": 0.1971740871667862, + "learning_rate": 7.176800556145397e-05, + "loss": 0.0271, + "step": 38880 + }, + { + "epoch": 7.55007038493277, + "grad_norm": 0.12404996901750565, + "learning_rate": 7.175311890372334e-05, + "loss": 0.0353, + "step": 38890 + }, + { + "epoch": 7.55201203825057, + "grad_norm": 0.12771864235401154, + "learning_rate": 7.17382298671043e-05, + "loss": 0.0271, + "step": 38900 + }, + { + "epoch": 7.5539536915683705, + "grad_norm": 0.24850310385227203, + "learning_rate": 7.172333845322515e-05, + "loss": 0.027, + "step": 38910 + }, + { + "epoch": 7.555895344886171, + "grad_norm": 0.18650943040847778, + "learning_rate": 7.170844466371436e-05, + "loss": 0.0206, + "step": 38920 + }, + { + "epoch": 7.557836998203971, + "grad_norm": 0.13239550590515137, + "learning_rate": 7.16935485002007e-05, + "loss": 0.0272, + "step": 38930 + }, + { + "epoch": 7.55977865152177, + "grad_norm": 0.3019721508026123, + "learning_rate": 7.167864996431319e-05, + "loss": 0.0369, + "step": 38940 + }, + { + "epoch": 7.561720304839571, + "grad_norm": 0.23454223573207855, + "learning_rate": 7.166374905768111e-05, + "loss": 0.0298, + "step": 38950 + }, + { + "epoch": 7.563661958157371, + "grad_norm": 0.15501241385936737, + "learning_rate": 7.164884578193401e-05, + "loss": 0.0358, + "step": 38960 + }, + { + "epoch": 7.565603611475171, + "grad_norm": 0.20347492396831512, + "learning_rate": 7.163394013870168e-05, + "loss": 0.0281, + "step": 38970 + }, + { + "epoch": 7.5675452647929715, + "grad_norm": 0.2898329198360443, + "learning_rate": 7.161903212961419e-05, + "loss": 0.037, + "step": 38980 + }, + { + "epoch": 7.569486918110771, + "grad_norm": 0.1909138262271881, + "learning_rate": 7.160412175630183e-05, + "loss": 0.026, + "step": 38990 + }, + { + "epoch": 7.571428571428571, + "grad_norm": 0.14981216192245483, + "learning_rate": 7.158920902039521e-05, + "loss": 0.0235, + "step": 39000 + }, + { + "epoch": 7.5733702247463714, + "grad_norm": 0.18579886853694916, + "learning_rate": 7.157429392352514e-05, + "loss": 0.0295, + "step": 39010 + }, + { + "epoch": 7.575311878064172, + "grad_norm": 0.0883009135723114, + "learning_rate": 7.155937646732275e-05, + "loss": 0.0218, + "step": 39020 + }, + { + "epoch": 7.577253531381972, + "grad_norm": 0.29365500807762146, + "learning_rate": 7.154445665341933e-05, + "loss": 0.0299, + "step": 39030 + }, + { + "epoch": 7.579195184699772, + "grad_norm": 0.12071356922388077, + "learning_rate": 7.152953448344654e-05, + "loss": 0.026, + "step": 39040 + }, + { + "epoch": 7.581136838017572, + "grad_norm": 0.1850924789905548, + "learning_rate": 7.151460995903624e-05, + "loss": 0.0222, + "step": 39050 + }, + { + "epoch": 7.583078491335372, + "grad_norm": 0.13706514239311218, + "learning_rate": 7.149968308182052e-05, + "loss": 0.0248, + "step": 39060 + }, + { + "epoch": 7.585020144653172, + "grad_norm": 0.2512175738811493, + "learning_rate": 7.14847538534318e-05, + "loss": 0.0308, + "step": 39070 + }, + { + "epoch": 7.5869617979709725, + "grad_norm": 0.12706363201141357, + "learning_rate": 7.14698222755027e-05, + "loss": 0.027, + "step": 39080 + }, + { + "epoch": 7.588903451288773, + "grad_norm": 0.20881341397762299, + "learning_rate": 7.14548883496661e-05, + "loss": 0.0253, + "step": 39090 + }, + { + "epoch": 7.590845104606572, + "grad_norm": 0.22737856209278107, + "learning_rate": 7.143995207755517e-05, + "loss": 0.0382, + "step": 39100 + }, + { + "epoch": 7.592786757924372, + "grad_norm": 0.21641579270362854, + "learning_rate": 7.142501346080333e-05, + "loss": 0.0279, + "step": 39110 + }, + { + "epoch": 7.594728411242173, + "grad_norm": 0.19298164546489716, + "learning_rate": 7.141007250104421e-05, + "loss": 0.0369, + "step": 39120 + }, + { + "epoch": 7.596670064559973, + "grad_norm": 0.18574371933937073, + "learning_rate": 7.139512919991176e-05, + "loss": 0.0301, + "step": 39130 + }, + { + "epoch": 7.598611717877773, + "grad_norm": 0.23536311089992523, + "learning_rate": 7.138018355904014e-05, + "loss": 0.0387, + "step": 39140 + }, + { + "epoch": 7.6005533711955735, + "grad_norm": 0.18839667737483978, + "learning_rate": 7.13652355800638e-05, + "loss": 0.0261, + "step": 39150 + }, + { + "epoch": 7.602495024513373, + "grad_norm": 0.2138085961341858, + "learning_rate": 7.135028526461743e-05, + "loss": 0.0297, + "step": 39160 + }, + { + "epoch": 7.604436677831173, + "grad_norm": 0.18470017611980438, + "learning_rate": 7.133533261433594e-05, + "loss": 0.0312, + "step": 39170 + }, + { + "epoch": 7.606378331148973, + "grad_norm": 0.18608500063419342, + "learning_rate": 7.132037763085457e-05, + "loss": 0.0247, + "step": 39180 + }, + { + "epoch": 7.608319984466774, + "grad_norm": 0.1544347107410431, + "learning_rate": 7.130542031580875e-05, + "loss": 0.0248, + "step": 39190 + }, + { + "epoch": 7.610261637784573, + "grad_norm": 0.1691783219575882, + "learning_rate": 7.12904606708342e-05, + "loss": 0.0271, + "step": 39200 + }, + { + "epoch": 7.612203291102373, + "grad_norm": 0.16014984250068665, + "learning_rate": 7.127549869756687e-05, + "loss": 0.0277, + "step": 39210 + }, + { + "epoch": 7.614144944420174, + "grad_norm": 0.15313081443309784, + "learning_rate": 7.126053439764299e-05, + "loss": 0.0255, + "step": 39220 + }, + { + "epoch": 7.616086597737974, + "grad_norm": 0.3460787534713745, + "learning_rate": 7.124556777269904e-05, + "loss": 0.0255, + "step": 39230 + }, + { + "epoch": 7.618028251055774, + "grad_norm": 0.25470736622810364, + "learning_rate": 7.123059882437174e-05, + "loss": 0.0283, + "step": 39240 + }, + { + "epoch": 7.6199699043735745, + "grad_norm": 0.11193574965000153, + "learning_rate": 7.121562755429807e-05, + "loss": 0.0328, + "step": 39250 + }, + { + "epoch": 7.621911557691374, + "grad_norm": 0.18901604413986206, + "learning_rate": 7.120065396411528e-05, + "loss": 0.0308, + "step": 39260 + }, + { + "epoch": 7.623853211009174, + "grad_norm": 0.13127091526985168, + "learning_rate": 7.118567805546084e-05, + "loss": 0.0258, + "step": 39270 + }, + { + "epoch": 7.625794864326974, + "grad_norm": 0.15081167221069336, + "learning_rate": 7.117069982997248e-05, + "loss": 0.025, + "step": 39280 + }, + { + "epoch": 7.627736517644775, + "grad_norm": 0.15990552306175232, + "learning_rate": 7.115571928928825e-05, + "loss": 0.0219, + "step": 39290 + }, + { + "epoch": 7.629678170962575, + "grad_norm": 0.10793888568878174, + "learning_rate": 7.114073643504635e-05, + "loss": 0.0246, + "step": 39300 + }, + { + "epoch": 7.631619824280374, + "grad_norm": 0.14396411180496216, + "learning_rate": 7.11257512688853e-05, + "loss": 0.03, + "step": 39310 + }, + { + "epoch": 7.633561477598175, + "grad_norm": 0.4442632496356964, + "learning_rate": 7.111076379244384e-05, + "loss": 0.029, + "step": 39320 + }, + { + "epoch": 7.635503130915975, + "grad_norm": 0.21338950097560883, + "learning_rate": 7.109577400736101e-05, + "loss": 0.0255, + "step": 39330 + }, + { + "epoch": 7.637444784233775, + "grad_norm": 0.19445885717868805, + "learning_rate": 7.108078191527605e-05, + "loss": 0.035, + "step": 39340 + }, + { + "epoch": 7.639386437551575, + "grad_norm": 0.1342901885509491, + "learning_rate": 7.106578751782847e-05, + "loss": 0.0264, + "step": 39350 + }, + { + "epoch": 7.641328090869376, + "grad_norm": 0.12022947520017624, + "learning_rate": 7.105079081665803e-05, + "loss": 0.0231, + "step": 39360 + }, + { + "epoch": 7.643269744187175, + "grad_norm": 0.2070641815662384, + "learning_rate": 7.103579181340476e-05, + "loss": 0.028, + "step": 39370 + }, + { + "epoch": 7.645211397504975, + "grad_norm": 0.15403538942337036, + "learning_rate": 7.102079050970893e-05, + "loss": 0.0239, + "step": 39380 + }, + { + "epoch": 7.647153050822776, + "grad_norm": 0.2563916742801666, + "learning_rate": 7.100578690721105e-05, + "loss": 0.0384, + "step": 39390 + }, + { + "epoch": 7.649094704140576, + "grad_norm": 0.11303606629371643, + "learning_rate": 7.09907810075519e-05, + "loss": 0.0223, + "step": 39400 + }, + { + "epoch": 7.651036357458376, + "grad_norm": 0.2034497857093811, + "learning_rate": 7.097577281237249e-05, + "loss": 0.0379, + "step": 39410 + }, + { + "epoch": 7.652978010776176, + "grad_norm": 0.2688402831554413, + "learning_rate": 7.09607623233141e-05, + "loss": 0.0274, + "step": 39420 + }, + { + "epoch": 7.654919664093976, + "grad_norm": 0.18338321149349213, + "learning_rate": 7.094574954201827e-05, + "loss": 0.0267, + "step": 39430 + }, + { + "epoch": 7.656861317411776, + "grad_norm": 0.2035408467054367, + "learning_rate": 7.093073447012675e-05, + "loss": 0.0341, + "step": 39440 + }, + { + "epoch": 7.658802970729576, + "grad_norm": 0.1897236406803131, + "learning_rate": 7.09157171092816e-05, + "loss": 0.0248, + "step": 39450 + }, + { + "epoch": 7.660744624047377, + "grad_norm": 0.11124586313962936, + "learning_rate": 7.090069746112504e-05, + "loss": 0.0357, + "step": 39460 + }, + { + "epoch": 7.662686277365177, + "grad_norm": 0.1475754976272583, + "learning_rate": 7.088567552729965e-05, + "loss": 0.024, + "step": 39470 + }, + { + "epoch": 7.664627930682976, + "grad_norm": 0.14264893531799316, + "learning_rate": 7.087065130944818e-05, + "loss": 0.0303, + "step": 39480 + }, + { + "epoch": 7.666569584000777, + "grad_norm": 0.20841741561889648, + "learning_rate": 7.085562480921366e-05, + "loss": 0.0324, + "step": 39490 + }, + { + "epoch": 7.668511237318577, + "grad_norm": 0.15285493433475494, + "learning_rate": 7.084059602823937e-05, + "loss": 0.033, + "step": 39500 + }, + { + "epoch": 7.670452890636377, + "grad_norm": 0.1506563425064087, + "learning_rate": 7.082556496816882e-05, + "loss": 0.0279, + "step": 39510 + }, + { + "epoch": 7.6723945439541765, + "grad_norm": 0.1595456451177597, + "learning_rate": 7.081053163064582e-05, + "loss": 0.0301, + "step": 39520 + }, + { + "epoch": 7.674336197271977, + "grad_norm": 0.1828005313873291, + "learning_rate": 7.079549601731434e-05, + "loss": 0.0271, + "step": 39530 + }, + { + "epoch": 7.676277850589777, + "grad_norm": 0.09962645918130875, + "learning_rate": 7.07804581298187e-05, + "loss": 0.0298, + "step": 39540 + }, + { + "epoch": 7.678219503907577, + "grad_norm": 0.09849396347999573, + "learning_rate": 7.07654179698034e-05, + "loss": 0.0291, + "step": 39550 + }, + { + "epoch": 7.680161157225378, + "grad_norm": 0.18018187582492828, + "learning_rate": 7.075037553891321e-05, + "loss": 0.0375, + "step": 39560 + }, + { + "epoch": 7.682102810543178, + "grad_norm": 0.2283305674791336, + "learning_rate": 7.073533083879315e-05, + "loss": 0.0274, + "step": 39570 + }, + { + "epoch": 7.684044463860977, + "grad_norm": 0.1865517944097519, + "learning_rate": 7.072028387108849e-05, + "loss": 0.0305, + "step": 39580 + }, + { + "epoch": 7.6859861171787776, + "grad_norm": 0.13483788073062897, + "learning_rate": 7.070523463744472e-05, + "loss": 0.0277, + "step": 39590 + }, + { + "epoch": 7.687927770496578, + "grad_norm": 0.24890661239624023, + "learning_rate": 7.069018313950763e-05, + "loss": 0.0312, + "step": 39600 + }, + { + "epoch": 7.689869423814378, + "grad_norm": 0.21865162253379822, + "learning_rate": 7.067512937892322e-05, + "loss": 0.0332, + "step": 39610 + }, + { + "epoch": 7.691811077132178, + "grad_norm": 0.19650210440158844, + "learning_rate": 7.066007335733775e-05, + "loss": 0.0243, + "step": 39620 + }, + { + "epoch": 7.693752730449978, + "grad_norm": 0.13224798440933228, + "learning_rate": 7.064501507639772e-05, + "loss": 0.0239, + "step": 39630 + }, + { + "epoch": 7.695694383767778, + "grad_norm": 0.1564997434616089, + "learning_rate": 7.062995453774987e-05, + "loss": 0.0221, + "step": 39640 + }, + { + "epoch": 7.697636037085578, + "grad_norm": 0.1448875516653061, + "learning_rate": 7.061489174304121e-05, + "loss": 0.0313, + "step": 39650 + }, + { + "epoch": 7.699577690403379, + "grad_norm": 0.29732272028923035, + "learning_rate": 7.0599826693919e-05, + "loss": 0.0307, + "step": 39660 + }, + { + "epoch": 7.701519343721179, + "grad_norm": 0.18256577849388123, + "learning_rate": 7.05847593920307e-05, + "loss": 0.0226, + "step": 39670 + }, + { + "epoch": 7.703460997038979, + "grad_norm": 0.179616317152977, + "learning_rate": 7.056968983902406e-05, + "loss": 0.0305, + "step": 39680 + }, + { + "epoch": 7.7054026503567785, + "grad_norm": 0.17942120134830475, + "learning_rate": 7.055461803654706e-05, + "loss": 0.0316, + "step": 39690 + }, + { + "epoch": 7.707344303674579, + "grad_norm": 0.23532575368881226, + "learning_rate": 7.053954398624794e-05, + "loss": 0.0329, + "step": 39700 + }, + { + "epoch": 7.709285956992379, + "grad_norm": 0.15090368688106537, + "learning_rate": 7.052446768977518e-05, + "loss": 0.0318, + "step": 39710 + }, + { + "epoch": 7.711227610310179, + "grad_norm": 0.13166168332099915, + "learning_rate": 7.050938914877748e-05, + "loss": 0.0314, + "step": 39720 + }, + { + "epoch": 7.71316926362798, + "grad_norm": 0.2281075417995453, + "learning_rate": 7.049430836490382e-05, + "loss": 0.032, + "step": 39730 + }, + { + "epoch": 7.715110916945779, + "grad_norm": 0.15370815992355347, + "learning_rate": 7.047922533980342e-05, + "loss": 0.0296, + "step": 39740 + }, + { + "epoch": 7.717052570263579, + "grad_norm": 0.23921237885951996, + "learning_rate": 7.046414007512571e-05, + "loss": 0.037, + "step": 39750 + }, + { + "epoch": 7.7189942235813795, + "grad_norm": 0.15648765861988068, + "learning_rate": 7.044905257252042e-05, + "loss": 0.0244, + "step": 39760 + }, + { + "epoch": 7.72093587689918, + "grad_norm": 0.20609354972839355, + "learning_rate": 7.04339628336375e-05, + "loss": 0.0325, + "step": 39770 + }, + { + "epoch": 7.72287753021698, + "grad_norm": 0.24257196485996246, + "learning_rate": 7.041887086012711e-05, + "loss": 0.0236, + "step": 39780 + }, + { + "epoch": 7.72481918353478, + "grad_norm": 0.14406773447990417, + "learning_rate": 7.040377665363969e-05, + "loss": 0.0268, + "step": 39790 + }, + { + "epoch": 7.72676083685258, + "grad_norm": 0.1937863975763321, + "learning_rate": 7.038868021582594e-05, + "loss": 0.0247, + "step": 39800 + }, + { + "epoch": 7.72870249017038, + "grad_norm": 0.13090632855892181, + "learning_rate": 7.037358154833679e-05, + "loss": 0.0286, + "step": 39810 + }, + { + "epoch": 7.73064414348818, + "grad_norm": 0.17788957059383392, + "learning_rate": 7.035848065282339e-05, + "loss": 0.024, + "step": 39820 + }, + { + "epoch": 7.732585796805981, + "grad_norm": 0.12293180823326111, + "learning_rate": 7.034337753093714e-05, + "loss": 0.0225, + "step": 39830 + }, + { + "epoch": 7.73452745012378, + "grad_norm": 0.20564424991607666, + "learning_rate": 7.032827218432972e-05, + "loss": 0.0314, + "step": 39840 + }, + { + "epoch": 7.73646910344158, + "grad_norm": 0.1698569655418396, + "learning_rate": 7.031316461465302e-05, + "loss": 0.0279, + "step": 39850 + }, + { + "epoch": 7.7384107567593805, + "grad_norm": 0.16668403148651123, + "learning_rate": 7.029805482355915e-05, + "loss": 0.0279, + "step": 39860 + }, + { + "epoch": 7.740352410077181, + "grad_norm": 0.1608809232711792, + "learning_rate": 7.028294281270055e-05, + "loss": 0.0267, + "step": 39870 + }, + { + "epoch": 7.742294063394981, + "grad_norm": 0.1835881471633911, + "learning_rate": 7.026782858372982e-05, + "loss": 0.0228, + "step": 39880 + }, + { + "epoch": 7.744235716712781, + "grad_norm": 0.18817521631717682, + "learning_rate": 7.025271213829982e-05, + "loss": 0.0295, + "step": 39890 + }, + { + "epoch": 7.746177370030581, + "grad_norm": 0.16358114778995514, + "learning_rate": 7.023759347806366e-05, + "loss": 0.0255, + "step": 39900 + }, + { + "epoch": 7.748119023348381, + "grad_norm": 0.19025200605392456, + "learning_rate": 7.022247260467472e-05, + "loss": 0.0269, + "step": 39910 + }, + { + "epoch": 7.750060676666181, + "grad_norm": 0.12760859727859497, + "learning_rate": 7.020734951978658e-05, + "loss": 0.0322, + "step": 39920 + }, + { + "epoch": 7.7520023299839815, + "grad_norm": 0.28136396408081055, + "learning_rate": 7.019222422505307e-05, + "loss": 0.03, + "step": 39930 + }, + { + "epoch": 7.753943983301782, + "grad_norm": 0.20438699424266815, + "learning_rate": 7.017709672212827e-05, + "loss": 0.0289, + "step": 39940 + }, + { + "epoch": 7.755885636619581, + "grad_norm": 0.11917057633399963, + "learning_rate": 7.016196701266652e-05, + "loss": 0.0249, + "step": 39950 + }, + { + "epoch": 7.7578272899373815, + "grad_norm": 0.1886441707611084, + "learning_rate": 7.014683509832238e-05, + "loss": 0.0263, + "step": 39960 + }, + { + "epoch": 7.759768943255182, + "grad_norm": 0.17372125387191772, + "learning_rate": 7.013170098075063e-05, + "loss": 0.0293, + "step": 39970 + }, + { + "epoch": 7.761710596572982, + "grad_norm": 0.279472291469574, + "learning_rate": 7.011656466160632e-05, + "loss": 0.0273, + "step": 39980 + }, + { + "epoch": 7.763652249890782, + "grad_norm": 0.10107189416885376, + "learning_rate": 7.010142614254475e-05, + "loss": 0.0267, + "step": 39990 + }, + { + "epoch": 7.765593903208583, + "grad_norm": 0.25937938690185547, + "learning_rate": 7.008628542522147e-05, + "loss": 0.0296, + "step": 40000 + }, + { + "epoch": 7.767535556526382, + "grad_norm": 0.17390339076519012, + "learning_rate": 7.007114251129218e-05, + "loss": 0.0272, + "step": 40010 + }, + { + "epoch": 7.769477209844182, + "grad_norm": 0.1696099042892456, + "learning_rate": 7.005599740241293e-05, + "loss": 0.0262, + "step": 40020 + }, + { + "epoch": 7.7714188631619825, + "grad_norm": 0.23928840458393097, + "learning_rate": 7.004085010023996e-05, + "loss": 0.0325, + "step": 40030 + }, + { + "epoch": 7.773360516479783, + "grad_norm": 0.07947216182947159, + "learning_rate": 7.002570060642976e-05, + "loss": 0.03, + "step": 40040 + }, + { + "epoch": 7.775302169797582, + "grad_norm": 0.2546564042568207, + "learning_rate": 7.001054892263903e-05, + "loss": 0.0319, + "step": 40050 + }, + { + "epoch": 7.777243823115382, + "grad_norm": 0.16561509668827057, + "learning_rate": 6.999539505052477e-05, + "loss": 0.0285, + "step": 40060 + }, + { + "epoch": 7.779185476433183, + "grad_norm": 0.1700277328491211, + "learning_rate": 6.998023899174418e-05, + "loss": 0.0299, + "step": 40070 + }, + { + "epoch": 7.781127129750983, + "grad_norm": 0.20625078678131104, + "learning_rate": 6.996508074795467e-05, + "loss": 0.0287, + "step": 40080 + }, + { + "epoch": 7.783068783068783, + "grad_norm": 0.13029532134532928, + "learning_rate": 6.994992032081396e-05, + "loss": 0.0241, + "step": 40090 + }, + { + "epoch": 7.7850104363865835, + "grad_norm": 0.17760910093784332, + "learning_rate": 6.993475771197995e-05, + "loss": 0.0244, + "step": 40100 + }, + { + "epoch": 7.786952089704383, + "grad_norm": 0.1824548840522766, + "learning_rate": 6.991959292311082e-05, + "loss": 0.0288, + "step": 40110 + }, + { + "epoch": 7.788893743022183, + "grad_norm": 0.16331690549850464, + "learning_rate": 6.990442595586495e-05, + "loss": 0.0296, + "step": 40120 + }, + { + "epoch": 7.7908353963399835, + "grad_norm": 0.29392629861831665, + "learning_rate": 6.988925681190098e-05, + "loss": 0.0288, + "step": 40130 + }, + { + "epoch": 7.792777049657784, + "grad_norm": 0.17756542563438416, + "learning_rate": 6.987408549287778e-05, + "loss": 0.0284, + "step": 40140 + }, + { + "epoch": 7.794718702975584, + "grad_norm": 0.10160905122756958, + "learning_rate": 6.985891200045449e-05, + "loss": 0.0278, + "step": 40150 + }, + { + "epoch": 7.796660356293383, + "grad_norm": 0.10374601930379868, + "learning_rate": 6.984373633629045e-05, + "loss": 0.0253, + "step": 40160 + }, + { + "epoch": 7.798602009611184, + "grad_norm": 0.18399016559123993, + "learning_rate": 6.98285585020452e-05, + "loss": 0.0229, + "step": 40170 + }, + { + "epoch": 7.800543662928984, + "grad_norm": 0.18814809620380402, + "learning_rate": 6.981337849937864e-05, + "loss": 0.0285, + "step": 40180 + }, + { + "epoch": 7.802485316246784, + "grad_norm": 0.2504492700099945, + "learning_rate": 6.979819632995078e-05, + "loss": 0.025, + "step": 40190 + }, + { + "epoch": 7.8044269695645845, + "grad_norm": 0.23162488639354706, + "learning_rate": 6.978301199542193e-05, + "loss": 0.0334, + "step": 40200 + }, + { + "epoch": 7.806368622882385, + "grad_norm": 0.19217786192893982, + "learning_rate": 6.976782549745263e-05, + "loss": 0.0323, + "step": 40210 + }, + { + "epoch": 7.808310276200184, + "grad_norm": 0.194997176527977, + "learning_rate": 6.975263683770367e-05, + "loss": 0.0257, + "step": 40220 + }, + { + "epoch": 7.810251929517984, + "grad_norm": 0.13735121488571167, + "learning_rate": 6.973744601783604e-05, + "loss": 0.021, + "step": 40230 + }, + { + "epoch": 7.812193582835785, + "grad_norm": 0.23318924009799957, + "learning_rate": 6.972225303951097e-05, + "loss": 0.0263, + "step": 40240 + }, + { + "epoch": 7.814135236153585, + "grad_norm": 0.12311150878667831, + "learning_rate": 6.970705790438998e-05, + "loss": 0.0217, + "step": 40250 + }, + { + "epoch": 7.816076889471385, + "grad_norm": 0.23368346691131592, + "learning_rate": 6.969186061413477e-05, + "loss": 0.0257, + "step": 40260 + }, + { + "epoch": 7.818018542789185, + "grad_norm": 0.1267135590314865, + "learning_rate": 6.967666117040727e-05, + "loss": 0.028, + "step": 40270 + }, + { + "epoch": 7.819960196106985, + "grad_norm": 0.1688273698091507, + "learning_rate": 6.966145957486972e-05, + "loss": 0.0334, + "step": 40280 + }, + { + "epoch": 7.821901849424785, + "grad_norm": 0.08396698534488678, + "learning_rate": 6.964625582918449e-05, + "loss": 0.03, + "step": 40290 + }, + { + "epoch": 7.823843502742585, + "grad_norm": 0.1786169558763504, + "learning_rate": 6.963104993501425e-05, + "loss": 0.0316, + "step": 40300 + }, + { + "epoch": 7.825785156060386, + "grad_norm": 0.12270987033843994, + "learning_rate": 6.961584189402192e-05, + "loss": 0.0272, + "step": 40310 + }, + { + "epoch": 7.827726809378186, + "grad_norm": 0.11438368260860443, + "learning_rate": 6.96006317078706e-05, + "loss": 0.0349, + "step": 40320 + }, + { + "epoch": 7.829668462695985, + "grad_norm": 0.13765040040016174, + "learning_rate": 6.958541937822367e-05, + "loss": 0.0299, + "step": 40330 + }, + { + "epoch": 7.831610116013786, + "grad_norm": 0.07660172879695892, + "learning_rate": 6.957020490674473e-05, + "loss": 0.0274, + "step": 40340 + }, + { + "epoch": 7.833551769331586, + "grad_norm": 0.10950667411088943, + "learning_rate": 6.95549882950976e-05, + "loss": 0.0256, + "step": 40350 + }, + { + "epoch": 7.835493422649386, + "grad_norm": 0.15706850588321686, + "learning_rate": 6.953976954494635e-05, + "loss": 0.0397, + "step": 40360 + }, + { + "epoch": 7.837435075967186, + "grad_norm": 0.1503974348306656, + "learning_rate": 6.952454865795528e-05, + "loss": 0.0242, + "step": 40370 + }, + { + "epoch": 7.839376729284986, + "grad_norm": 0.18901702761650085, + "learning_rate": 6.950932563578892e-05, + "loss": 0.0299, + "step": 40380 + }, + { + "epoch": 7.841318382602786, + "grad_norm": 0.3391530513763428, + "learning_rate": 6.949410048011206e-05, + "loss": 0.0245, + "step": 40390 + }, + { + "epoch": 7.843260035920586, + "grad_norm": 0.1126910150051117, + "learning_rate": 6.947887319258966e-05, + "loss": 0.0338, + "step": 40400 + }, + { + "epoch": 7.845201689238387, + "grad_norm": 0.13377083837985992, + "learning_rate": 6.946364377488696e-05, + "loss": 0.0367, + "step": 40410 + }, + { + "epoch": 7.847143342556187, + "grad_norm": 0.13542377948760986, + "learning_rate": 6.944841222866947e-05, + "loss": 0.0283, + "step": 40420 + }, + { + "epoch": 7.849084995873986, + "grad_norm": 0.32364708185195923, + "learning_rate": 6.943317855560284e-05, + "loss": 0.0297, + "step": 40430 + }, + { + "epoch": 7.851026649191787, + "grad_norm": 0.17811837792396545, + "learning_rate": 6.941794275735302e-05, + "loss": 0.0276, + "step": 40440 + }, + { + "epoch": 7.852968302509587, + "grad_norm": 0.25189393758773804, + "learning_rate": 6.94027048355862e-05, + "loss": 0.0321, + "step": 40450 + }, + { + "epoch": 7.854909955827387, + "grad_norm": 0.14521215856075287, + "learning_rate": 6.938746479196871e-05, + "loss": 0.0318, + "step": 40460 + }, + { + "epoch": 7.856851609145187, + "grad_norm": 0.1427156627178192, + "learning_rate": 6.937222262816724e-05, + "loss": 0.0256, + "step": 40470 + }, + { + "epoch": 7.858793262462987, + "grad_norm": 0.17923372983932495, + "learning_rate": 6.935697834584865e-05, + "loss": 0.028, + "step": 40480 + }, + { + "epoch": 7.860734915780787, + "grad_norm": 0.19077517092227936, + "learning_rate": 6.934173194667998e-05, + "loss": 0.0279, + "step": 40490 + }, + { + "epoch": 7.862676569098587, + "grad_norm": 0.1529773473739624, + "learning_rate": 6.93264834323286e-05, + "loss": 0.0208, + "step": 40500 + }, + { + "epoch": 7.864618222416388, + "grad_norm": 0.16183316707611084, + "learning_rate": 6.931123280446206e-05, + "loss": 0.0234, + "step": 40510 + }, + { + "epoch": 7.866559875734188, + "grad_norm": 0.16177977621555328, + "learning_rate": 6.929598006474811e-05, + "loss": 0.027, + "step": 40520 + }, + { + "epoch": 7.868501529051988, + "grad_norm": 0.2506939172744751, + "learning_rate": 6.92807252148548e-05, + "loss": 0.0226, + "step": 40530 + }, + { + "epoch": 7.870443182369788, + "grad_norm": 0.18496371805667877, + "learning_rate": 6.926546825645037e-05, + "loss": 0.0327, + "step": 40540 + }, + { + "epoch": 7.872384835687588, + "grad_norm": 0.23066891729831696, + "learning_rate": 6.92502091912033e-05, + "loss": 0.028, + "step": 40550 + }, + { + "epoch": 7.874326489005388, + "grad_norm": 0.15035594999790192, + "learning_rate": 6.923494802078229e-05, + "loss": 0.0228, + "step": 40560 + }, + { + "epoch": 7.876268142323188, + "grad_norm": 0.1414550244808197, + "learning_rate": 6.921968474685628e-05, + "loss": 0.03, + "step": 40570 + }, + { + "epoch": 7.878209795640989, + "grad_norm": 0.14065924286842346, + "learning_rate": 6.920441937109445e-05, + "loss": 0.0191, + "step": 40580 + }, + { + "epoch": 7.880151448958788, + "grad_norm": 0.15829476714134216, + "learning_rate": 6.91891518951662e-05, + "loss": 0.0306, + "step": 40590 + }, + { + "epoch": 7.882093102276588, + "grad_norm": 0.20231342315673828, + "learning_rate": 6.917388232074114e-05, + "loss": 0.0315, + "step": 40600 + }, + { + "epoch": 7.884034755594389, + "grad_norm": 0.1541995108127594, + "learning_rate": 6.915861064948914e-05, + "loss": 0.0284, + "step": 40610 + }, + { + "epoch": 7.885976408912189, + "grad_norm": 0.175252303481102, + "learning_rate": 6.914333688308025e-05, + "loss": 0.0247, + "step": 40620 + }, + { + "epoch": 7.887918062229989, + "grad_norm": 0.1222141683101654, + "learning_rate": 6.912806102318487e-05, + "loss": 0.0279, + "step": 40630 + }, + { + "epoch": 7.889859715547789, + "grad_norm": 0.09158585220575333, + "learning_rate": 6.911278307147347e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 7.891801368865589, + "grad_norm": 0.16865983605384827, + "learning_rate": 6.909750302961684e-05, + "loss": 0.0253, + "step": 40650 + }, + { + "epoch": 7.893743022183389, + "grad_norm": 0.1013462170958519, + "learning_rate": 6.9082220899286e-05, + "loss": 0.0317, + "step": 40660 + }, + { + "epoch": 7.895684675501189, + "grad_norm": 0.27724701166152954, + "learning_rate": 6.906693668215219e-05, + "loss": 0.0289, + "step": 40670 + }, + { + "epoch": 7.89762632881899, + "grad_norm": 0.20405705273151398, + "learning_rate": 6.905165037988683e-05, + "loss": 0.0346, + "step": 40680 + }, + { + "epoch": 7.899567982136789, + "grad_norm": 0.22351527214050293, + "learning_rate": 6.903636199416164e-05, + "loss": 0.0285, + "step": 40690 + }, + { + "epoch": 7.901509635454589, + "grad_norm": 0.19941136240959167, + "learning_rate": 6.902107152664851e-05, + "loss": 0.0277, + "step": 40700 + }, + { + "epoch": 7.90345128877239, + "grad_norm": 0.18945486843585968, + "learning_rate": 6.90057789790196e-05, + "loss": 0.0322, + "step": 40710 + }, + { + "epoch": 7.90539294209019, + "grad_norm": 0.27546635270118713, + "learning_rate": 6.899048435294728e-05, + "loss": 0.0304, + "step": 40720 + }, + { + "epoch": 7.90733459540799, + "grad_norm": 0.12582656741142273, + "learning_rate": 6.897518765010415e-05, + "loss": 0.0277, + "step": 40730 + }, + { + "epoch": 7.90927624872579, + "grad_norm": 0.29052817821502686, + "learning_rate": 6.895988887216302e-05, + "loss": 0.0254, + "step": 40740 + }, + { + "epoch": 7.91121790204359, + "grad_norm": 0.13756923377513885, + "learning_rate": 6.894458802079694e-05, + "loss": 0.0233, + "step": 40750 + }, + { + "epoch": 7.91315955536139, + "grad_norm": 0.11785595864057541, + "learning_rate": 6.892928509767922e-05, + "loss": 0.0254, + "step": 40760 + }, + { + "epoch": 7.91510120867919, + "grad_norm": 0.24851036071777344, + "learning_rate": 6.891398010448333e-05, + "loss": 0.028, + "step": 40770 + }, + { + "epoch": 7.917042861996991, + "grad_norm": 0.12843777239322662, + "learning_rate": 6.8898673042883e-05, + "loss": 0.0293, + "step": 40780 + }, + { + "epoch": 7.918984515314791, + "grad_norm": 0.15208712220191956, + "learning_rate": 6.888336391455222e-05, + "loss": 0.0196, + "step": 40790 + }, + { + "epoch": 7.92092616863259, + "grad_norm": 0.24607160687446594, + "learning_rate": 6.886805272116513e-05, + "loss": 0.0223, + "step": 40800 + }, + { + "epoch": 7.9228678219503905, + "grad_norm": 0.16969697177410126, + "learning_rate": 6.885273946439617e-05, + "loss": 0.0229, + "step": 40810 + }, + { + "epoch": 7.924809475268191, + "grad_norm": 0.2701255977153778, + "learning_rate": 6.883742414591998e-05, + "loss": 0.0265, + "step": 40820 + }, + { + "epoch": 7.926751128585991, + "grad_norm": 0.160742849111557, + "learning_rate": 6.88221067674114e-05, + "loss": 0.032, + "step": 40830 + }, + { + "epoch": 7.928692781903791, + "grad_norm": 0.132036030292511, + "learning_rate": 6.88067873305455e-05, + "loss": 0.0326, + "step": 40840 + }, + { + "epoch": 7.930634435221592, + "grad_norm": 0.21490038931369781, + "learning_rate": 6.879146583699765e-05, + "loss": 0.0314, + "step": 40850 + }, + { + "epoch": 7.932576088539391, + "grad_norm": 0.249811053276062, + "learning_rate": 6.877614228844334e-05, + "loss": 0.0272, + "step": 40860 + }, + { + "epoch": 7.934517741857191, + "grad_norm": 0.14185799658298492, + "learning_rate": 6.876081668655832e-05, + "loss": 0.0277, + "step": 40870 + }, + { + "epoch": 7.9364593951749915, + "grad_norm": 0.13842064142227173, + "learning_rate": 6.874548903301861e-05, + "loss": 0.0223, + "step": 40880 + }, + { + "epoch": 7.938401048492792, + "grad_norm": 0.18821464478969574, + "learning_rate": 6.873015932950039e-05, + "loss": 0.0289, + "step": 40890 + }, + { + "epoch": 7.940342701810592, + "grad_norm": 0.1511618196964264, + "learning_rate": 6.871482757768012e-05, + "loss": 0.02, + "step": 40900 + }, + { + "epoch": 7.9422843551283915, + "grad_norm": 0.2558678686618805, + "learning_rate": 6.869949377923442e-05, + "loss": 0.0301, + "step": 40910 + }, + { + "epoch": 7.944226008446192, + "grad_norm": 0.22738471627235413, + "learning_rate": 6.868415793584022e-05, + "loss": 0.0261, + "step": 40920 + }, + { + "epoch": 7.946167661763992, + "grad_norm": 0.28241968154907227, + "learning_rate": 6.86688200491746e-05, + "loss": 0.0236, + "step": 40930 + }, + { + "epoch": 7.948109315081792, + "grad_norm": 0.1767837405204773, + "learning_rate": 6.865348012091488e-05, + "loss": 0.0254, + "step": 40940 + }, + { + "epoch": 7.950050968399593, + "grad_norm": 0.16126801073551178, + "learning_rate": 6.86381381527386e-05, + "loss": 0.0269, + "step": 40950 + }, + { + "epoch": 7.951992621717393, + "grad_norm": 0.2527468800544739, + "learning_rate": 6.862279414632359e-05, + "loss": 0.0252, + "step": 40960 + }, + { + "epoch": 7.953934275035192, + "grad_norm": 0.1593000441789627, + "learning_rate": 6.860744810334778e-05, + "loss": 0.0205, + "step": 40970 + }, + { + "epoch": 7.9558759283529925, + "grad_norm": 0.17562335729599, + "learning_rate": 6.859210002548943e-05, + "loss": 0.0254, + "step": 40980 + }, + { + "epoch": 7.957817581670793, + "grad_norm": 0.21386466920375824, + "learning_rate": 6.8576749914427e-05, + "loss": 0.0269, + "step": 40990 + }, + { + "epoch": 7.959759234988593, + "grad_norm": 0.23308835923671722, + "learning_rate": 6.85613977718391e-05, + "loss": 0.0261, + "step": 41000 + }, + { + "epoch": 7.961700888306392, + "grad_norm": 0.1290193349123001, + "learning_rate": 6.854604359940464e-05, + "loss": 0.0272, + "step": 41010 + }, + { + "epoch": 7.963642541624193, + "grad_norm": 0.14558623731136322, + "learning_rate": 6.853068739880276e-05, + "loss": 0.0357, + "step": 41020 + }, + { + "epoch": 7.965584194941993, + "grad_norm": 0.2003004103899002, + "learning_rate": 6.851532917171276e-05, + "loss": 0.0252, + "step": 41030 + }, + { + "epoch": 7.967525848259793, + "grad_norm": 0.2368324249982834, + "learning_rate": 6.849996891981418e-05, + "loss": 0.0262, + "step": 41040 + }, + { + "epoch": 7.9694675015775935, + "grad_norm": 0.20499645173549652, + "learning_rate": 6.848460664478684e-05, + "loss": 0.0256, + "step": 41050 + }, + { + "epoch": 7.971409154895394, + "grad_norm": 0.11282561719417572, + "learning_rate": 6.846924234831069e-05, + "loss": 0.0294, + "step": 41060 + }, + { + "epoch": 7.973350808213193, + "grad_norm": 0.20620855689048767, + "learning_rate": 6.845387603206597e-05, + "loss": 0.0235, + "step": 41070 + }, + { + "epoch": 7.9752924615309935, + "grad_norm": 0.1284402757883072, + "learning_rate": 6.843850769773312e-05, + "loss": 0.0257, + "step": 41080 + }, + { + "epoch": 7.977234114848794, + "grad_norm": 0.1924726516008377, + "learning_rate": 6.842313734699278e-05, + "loss": 0.0303, + "step": 41090 + }, + { + "epoch": 7.979175768166594, + "grad_norm": 0.11398079991340637, + "learning_rate": 6.840776498152584e-05, + "loss": 0.0278, + "step": 41100 + }, + { + "epoch": 7.981117421484394, + "grad_norm": 0.17424847185611725, + "learning_rate": 6.83923906030134e-05, + "loss": 0.0269, + "step": 41110 + }, + { + "epoch": 7.983059074802194, + "grad_norm": 0.11794821172952652, + "learning_rate": 6.837701421313677e-05, + "loss": 0.0291, + "step": 41120 + }, + { + "epoch": 7.985000728119994, + "grad_norm": 0.21200214326381683, + "learning_rate": 6.83616358135775e-05, + "loss": 0.0332, + "step": 41130 + }, + { + "epoch": 7.986942381437794, + "grad_norm": 0.10399593412876129, + "learning_rate": 6.834625540601734e-05, + "loss": 0.0195, + "step": 41140 + }, + { + "epoch": 7.9888840347555945, + "grad_norm": 0.2098480761051178, + "learning_rate": 6.833087299213829e-05, + "loss": 0.0449, + "step": 41150 + }, + { + "epoch": 7.990825688073395, + "grad_norm": 0.18051017820835114, + "learning_rate": 6.83154885736225e-05, + "loss": 0.0253, + "step": 41160 + }, + { + "epoch": 7.992767341391195, + "grad_norm": 0.14039532840251923, + "learning_rate": 6.830010215215244e-05, + "loss": 0.0284, + "step": 41170 + }, + { + "epoch": 7.994708994708994, + "grad_norm": 0.12691377103328705, + "learning_rate": 6.828471372941072e-05, + "loss": 0.0252, + "step": 41180 + }, + { + "epoch": 7.996650648026795, + "grad_norm": 0.22790534794330597, + "learning_rate": 6.826932330708018e-05, + "loss": 0.0279, + "step": 41190 + }, + { + "epoch": 7.998592301344595, + "grad_norm": 0.16820213198661804, + "learning_rate": 6.825393088684393e-05, + "loss": 0.0306, + "step": 41200 + }, + { + "epoch": 8.00038833066356, + "grad_norm": 0.16879057884216309, + "learning_rate": 6.823853647038524e-05, + "loss": 0.0323, + "step": 41210 + }, + { + "epoch": 8.00232998398136, + "grad_norm": 0.12176863104104996, + "learning_rate": 6.822314005938762e-05, + "loss": 0.0288, + "step": 41220 + }, + { + "epoch": 8.00427163729916, + "grad_norm": 0.17231915891170502, + "learning_rate": 6.820774165553479e-05, + "loss": 0.0296, + "step": 41230 + }, + { + "epoch": 8.00621329061696, + "grad_norm": 0.1504489630460739, + "learning_rate": 6.819234126051072e-05, + "loss": 0.0289, + "step": 41240 + }, + { + "epoch": 8.00815494393476, + "grad_norm": 0.1520255208015442, + "learning_rate": 6.817693887599956e-05, + "loss": 0.0233, + "step": 41250 + }, + { + "epoch": 8.010096597252561, + "grad_norm": 0.12207664549350739, + "learning_rate": 6.816153450368568e-05, + "loss": 0.028, + "step": 41260 + }, + { + "epoch": 8.012038250570361, + "grad_norm": 0.16559487581253052, + "learning_rate": 6.814612814525372e-05, + "loss": 0.0298, + "step": 41270 + }, + { + "epoch": 8.013979903888162, + "grad_norm": 0.1341675966978073, + "learning_rate": 6.813071980238846e-05, + "loss": 0.0278, + "step": 41280 + }, + { + "epoch": 8.01592155720596, + "grad_norm": 0.16493134200572968, + "learning_rate": 6.811530947677492e-05, + "loss": 0.0257, + "step": 41290 + }, + { + "epoch": 8.01786321052376, + "grad_norm": 0.1891603320837021, + "learning_rate": 6.809989717009839e-05, + "loss": 0.0249, + "step": 41300 + }, + { + "epoch": 8.01980486384156, + "grad_norm": 0.2685517370700836, + "learning_rate": 6.808448288404431e-05, + "loss": 0.0376, + "step": 41310 + }, + { + "epoch": 8.021746517159361, + "grad_norm": 0.2612510323524475, + "learning_rate": 6.806906662029838e-05, + "loss": 0.0258, + "step": 41320 + }, + { + "epoch": 8.023688170477161, + "grad_norm": 0.23491741716861725, + "learning_rate": 6.805364838054647e-05, + "loss": 0.0274, + "step": 41330 + }, + { + "epoch": 8.025629823794961, + "grad_norm": 0.1311972588300705, + "learning_rate": 6.803822816647474e-05, + "loss": 0.0245, + "step": 41340 + }, + { + "epoch": 8.027571477112762, + "grad_norm": 0.1641254872083664, + "learning_rate": 6.802280597976949e-05, + "loss": 0.0268, + "step": 41350 + }, + { + "epoch": 8.029513130430562, + "grad_norm": 0.1432117074728012, + "learning_rate": 6.800738182211727e-05, + "loss": 0.0258, + "step": 41360 + }, + { + "epoch": 8.031454783748362, + "grad_norm": 0.16097988188266754, + "learning_rate": 6.799195569520485e-05, + "loss": 0.0296, + "step": 41370 + }, + { + "epoch": 8.033396437066163, + "grad_norm": 0.26602354645729065, + "learning_rate": 6.79765276007192e-05, + "loss": 0.03, + "step": 41380 + }, + { + "epoch": 8.035338090383961, + "grad_norm": 0.12759247422218323, + "learning_rate": 6.796109754034753e-05, + "loss": 0.0242, + "step": 41390 + }, + { + "epoch": 8.037279743701761, + "grad_norm": 0.18765263259410858, + "learning_rate": 6.794566551577724e-05, + "loss": 0.0253, + "step": 41400 + }, + { + "epoch": 8.039221397019562, + "grad_norm": 0.17229335010051727, + "learning_rate": 6.793023152869594e-05, + "loss": 0.0285, + "step": 41410 + }, + { + "epoch": 8.041163050337362, + "grad_norm": 0.1767314225435257, + "learning_rate": 6.791479558079147e-05, + "loss": 0.0268, + "step": 41420 + }, + { + "epoch": 8.043104703655162, + "grad_norm": 0.18798616528511047, + "learning_rate": 6.789935767375191e-05, + "loss": 0.0296, + "step": 41430 + }, + { + "epoch": 8.045046356972962, + "grad_norm": 0.14472553133964539, + "learning_rate": 6.78839178092655e-05, + "loss": 0.0282, + "step": 41440 + }, + { + "epoch": 8.046988010290763, + "grad_norm": 0.17633689939975739, + "learning_rate": 6.786847598902072e-05, + "loss": 0.027, + "step": 41450 + }, + { + "epoch": 8.048929663608563, + "grad_norm": 0.17645584046840668, + "learning_rate": 6.785303221470627e-05, + "loss": 0.029, + "step": 41460 + }, + { + "epoch": 8.050871316926363, + "grad_norm": 0.13308265805244446, + "learning_rate": 6.783758648801108e-05, + "loss": 0.0283, + "step": 41470 + }, + { + "epoch": 8.052812970244164, + "grad_norm": 0.21850524842739105, + "learning_rate": 6.782213881062422e-05, + "loss": 0.0279, + "step": 41480 + }, + { + "epoch": 8.054754623561964, + "grad_norm": 0.2493409961462021, + "learning_rate": 6.780668918423508e-05, + "loss": 0.0257, + "step": 41490 + }, + { + "epoch": 8.056696276879762, + "grad_norm": 0.1802072674036026, + "learning_rate": 6.779123761053317e-05, + "loss": 0.0257, + "step": 41500 + }, + { + "epoch": 8.058637930197563, + "grad_norm": 0.16417503356933594, + "learning_rate": 6.777578409120826e-05, + "loss": 0.0209, + "step": 41510 + }, + { + "epoch": 8.060579583515363, + "grad_norm": 0.15600402653217316, + "learning_rate": 6.776032862795033e-05, + "loss": 0.0211, + "step": 41520 + }, + { + "epoch": 8.062521236833163, + "grad_norm": 0.3204577565193176, + "learning_rate": 6.774487122244956e-05, + "loss": 0.0309, + "step": 41530 + }, + { + "epoch": 8.064462890150963, + "grad_norm": 0.16113097965717316, + "learning_rate": 6.772941187639637e-05, + "loss": 0.0289, + "step": 41540 + }, + { + "epoch": 8.066404543468764, + "grad_norm": 0.1445164680480957, + "learning_rate": 6.771395059148134e-05, + "loss": 0.0271, + "step": 41550 + }, + { + "epoch": 8.068346196786564, + "grad_norm": 0.219603031873703, + "learning_rate": 6.76984873693953e-05, + "loss": 0.0343, + "step": 41560 + }, + { + "epoch": 8.070287850104364, + "grad_norm": 0.13899725675582886, + "learning_rate": 6.768302221182931e-05, + "loss": 0.0263, + "step": 41570 + }, + { + "epoch": 8.072229503422164, + "grad_norm": 0.19179868698120117, + "learning_rate": 6.766755512047457e-05, + "loss": 0.0331, + "step": 41580 + }, + { + "epoch": 8.074171156739965, + "grad_norm": 0.16209277510643005, + "learning_rate": 6.765208609702259e-05, + "loss": 0.0331, + "step": 41590 + }, + { + "epoch": 8.076112810057765, + "grad_norm": 0.1337200105190277, + "learning_rate": 6.763661514316499e-05, + "loss": 0.0258, + "step": 41600 + }, + { + "epoch": 8.078054463375564, + "grad_norm": 0.12894698977470398, + "learning_rate": 6.76211422605937e-05, + "loss": 0.0247, + "step": 41610 + }, + { + "epoch": 8.079996116693364, + "grad_norm": 0.29303404688835144, + "learning_rate": 6.760566745100076e-05, + "loss": 0.0257, + "step": 41620 + }, + { + "epoch": 8.081937770011164, + "grad_norm": 0.12958769500255585, + "learning_rate": 6.75901907160785e-05, + "loss": 0.0281, + "step": 41630 + }, + { + "epoch": 8.083879423328964, + "grad_norm": 0.2705373466014862, + "learning_rate": 6.757471205751943e-05, + "loss": 0.0282, + "step": 41640 + }, + { + "epoch": 8.085821076646765, + "grad_norm": 0.14219428598880768, + "learning_rate": 6.75592314770163e-05, + "loss": 0.0319, + "step": 41650 + }, + { + "epoch": 8.087762729964565, + "grad_norm": 0.28569087386131287, + "learning_rate": 6.7543748976262e-05, + "loss": 0.0258, + "step": 41660 + }, + { + "epoch": 8.089704383282365, + "grad_norm": 0.17415280640125275, + "learning_rate": 6.752826455694968e-05, + "loss": 0.0255, + "step": 41670 + }, + { + "epoch": 8.091646036600165, + "grad_norm": 0.19130605459213257, + "learning_rate": 6.751277822077271e-05, + "loss": 0.0278, + "step": 41680 + }, + { + "epoch": 8.093587689917966, + "grad_norm": 0.24783070385456085, + "learning_rate": 6.749728996942465e-05, + "loss": 0.03, + "step": 41690 + }, + { + "epoch": 8.095529343235766, + "grad_norm": 0.17884932458400726, + "learning_rate": 6.748179980459924e-05, + "loss": 0.0301, + "step": 41700 + }, + { + "epoch": 8.097470996553565, + "grad_norm": 0.21430835127830505, + "learning_rate": 6.746630772799052e-05, + "loss": 0.0236, + "step": 41710 + }, + { + "epoch": 8.099412649871365, + "grad_norm": 0.19735318422317505, + "learning_rate": 6.745081374129262e-05, + "loss": 0.0256, + "step": 41720 + }, + { + "epoch": 8.101354303189165, + "grad_norm": 0.16158203780651093, + "learning_rate": 6.743531784619998e-05, + "loss": 0.0203, + "step": 41730 + }, + { + "epoch": 8.103295956506965, + "grad_norm": 0.22778812050819397, + "learning_rate": 6.741982004440719e-05, + "loss": 0.022, + "step": 41740 + }, + { + "epoch": 8.105237609824766, + "grad_norm": 0.1448313444852829, + "learning_rate": 6.740432033760907e-05, + "loss": 0.0241, + "step": 41750 + }, + { + "epoch": 8.107179263142566, + "grad_norm": 0.17851364612579346, + "learning_rate": 6.738881872750066e-05, + "loss": 0.0196, + "step": 41760 + }, + { + "epoch": 8.109120916460366, + "grad_norm": 0.13721778988838196, + "learning_rate": 6.737331521577715e-05, + "loss": 0.0277, + "step": 41770 + }, + { + "epoch": 8.111062569778166, + "grad_norm": 0.17972803115844727, + "learning_rate": 6.735780980413403e-05, + "loss": 0.0225, + "step": 41780 + }, + { + "epoch": 8.113004223095967, + "grad_norm": 0.15351194143295288, + "learning_rate": 6.734230249426692e-05, + "loss": 0.0274, + "step": 41790 + }, + { + "epoch": 8.114945876413767, + "grad_norm": 0.26010802388191223, + "learning_rate": 6.732679328787168e-05, + "loss": 0.0271, + "step": 41800 + }, + { + "epoch": 8.116887529731567, + "grad_norm": 0.19195348024368286, + "learning_rate": 6.731128218664438e-05, + "loss": 0.022, + "step": 41810 + }, + { + "epoch": 8.118829183049366, + "grad_norm": 0.18432004749774933, + "learning_rate": 6.729576919228129e-05, + "loss": 0.0332, + "step": 41820 + }, + { + "epoch": 8.120770836367166, + "grad_norm": 0.14060258865356445, + "learning_rate": 6.728025430647888e-05, + "loss": 0.0252, + "step": 41830 + }, + { + "epoch": 8.122712489684966, + "grad_norm": 0.21165090799331665, + "learning_rate": 6.726473753093383e-05, + "loss": 0.0257, + "step": 41840 + }, + { + "epoch": 8.124654143002767, + "grad_norm": 0.19022256135940552, + "learning_rate": 6.724921886734305e-05, + "loss": 0.0358, + "step": 41850 + }, + { + "epoch": 8.126595796320567, + "grad_norm": 0.12281109392642975, + "learning_rate": 6.723369831740363e-05, + "loss": 0.0239, + "step": 41860 + }, + { + "epoch": 8.128537449638367, + "grad_norm": 0.24182626605033875, + "learning_rate": 6.721817588281288e-05, + "loss": 0.0231, + "step": 41870 + }, + { + "epoch": 8.130479102956167, + "grad_norm": 0.22696049511432648, + "learning_rate": 6.720265156526828e-05, + "loss": 0.031, + "step": 41880 + }, + { + "epoch": 8.132420756273968, + "grad_norm": 0.1857869178056717, + "learning_rate": 6.718712536646758e-05, + "loss": 0.0334, + "step": 41890 + }, + { + "epoch": 8.134362409591768, + "grad_norm": 0.11807309091091156, + "learning_rate": 6.71715972881087e-05, + "loss": 0.0259, + "step": 41900 + }, + { + "epoch": 8.136304062909568, + "grad_norm": 0.10735898464918137, + "learning_rate": 6.715606733188974e-05, + "loss": 0.026, + "step": 41910 + }, + { + "epoch": 8.138245716227368, + "grad_norm": 0.22502966225147247, + "learning_rate": 6.714053549950909e-05, + "loss": 0.03, + "step": 41920 + }, + { + "epoch": 8.140187369545167, + "grad_norm": 0.21329368650913239, + "learning_rate": 6.71250017926652e-05, + "loss": 0.0361, + "step": 41930 + }, + { + "epoch": 8.142129022862967, + "grad_norm": 0.08683369308710098, + "learning_rate": 6.71094662130569e-05, + "loss": 0.0307, + "step": 41940 + }, + { + "epoch": 8.144070676180768, + "grad_norm": 0.20837309956550598, + "learning_rate": 6.709392876238307e-05, + "loss": 0.0223, + "step": 41950 + }, + { + "epoch": 8.146012329498568, + "grad_norm": 0.12666690349578857, + "learning_rate": 6.707838944234289e-05, + "loss": 0.0324, + "step": 41960 + }, + { + "epoch": 8.147953982816368, + "grad_norm": 0.25787293910980225, + "learning_rate": 6.706284825463574e-05, + "loss": 0.029, + "step": 41970 + }, + { + "epoch": 8.149895636134168, + "grad_norm": 0.15558986365795135, + "learning_rate": 6.704730520096114e-05, + "loss": 0.0268, + "step": 41980 + }, + { + "epoch": 8.151837289451969, + "grad_norm": 0.15428008139133453, + "learning_rate": 6.703176028301888e-05, + "loss": 0.0267, + "step": 41990 + }, + { + "epoch": 8.153778942769769, + "grad_norm": 0.2080444097518921, + "learning_rate": 6.701621350250892e-05, + "loss": 0.0458, + "step": 42000 + }, + { + "epoch": 8.15572059608757, + "grad_norm": 0.1515149027109146, + "learning_rate": 6.700066486113144e-05, + "loss": 0.027, + "step": 42010 + }, + { + "epoch": 8.15766224940537, + "grad_norm": 0.19097782671451569, + "learning_rate": 6.69851143605868e-05, + "loss": 0.0208, + "step": 42020 + }, + { + "epoch": 8.159603902723168, + "grad_norm": 0.19297344982624054, + "learning_rate": 6.696956200257561e-05, + "loss": 0.0276, + "step": 42030 + }, + { + "epoch": 8.161545556040968, + "grad_norm": 0.18564139306545258, + "learning_rate": 6.695400778879863e-05, + "loss": 0.0215, + "step": 42040 + }, + { + "epoch": 8.163487209358768, + "grad_norm": 0.2947467267513275, + "learning_rate": 6.693845172095683e-05, + "loss": 0.0315, + "step": 42050 + }, + { + "epoch": 8.165428862676569, + "grad_norm": 0.23317809402942657, + "learning_rate": 6.692289380075142e-05, + "loss": 0.026, + "step": 42060 + }, + { + "epoch": 8.167370515994369, + "grad_norm": 0.15701152384281158, + "learning_rate": 6.690733402988379e-05, + "loss": 0.0259, + "step": 42070 + }, + { + "epoch": 8.16931216931217, + "grad_norm": 0.16329114139080048, + "learning_rate": 6.689177241005553e-05, + "loss": 0.0228, + "step": 42080 + }, + { + "epoch": 8.17125382262997, + "grad_norm": 0.2056419849395752, + "learning_rate": 6.687620894296844e-05, + "loss": 0.0233, + "step": 42090 + }, + { + "epoch": 8.17319547594777, + "grad_norm": 0.23299947381019592, + "learning_rate": 6.686064363032451e-05, + "loss": 0.0305, + "step": 42100 + }, + { + "epoch": 8.17513712926557, + "grad_norm": 0.1811285763978958, + "learning_rate": 6.684507647382596e-05, + "loss": 0.0258, + "step": 42110 + }, + { + "epoch": 8.17707878258337, + "grad_norm": 0.2334316372871399, + "learning_rate": 6.682950747517513e-05, + "loss": 0.031, + "step": 42120 + }, + { + "epoch": 8.17902043590117, + "grad_norm": 0.15796388685703278, + "learning_rate": 6.68139366360747e-05, + "loss": 0.0266, + "step": 42130 + }, + { + "epoch": 8.18096208921897, + "grad_norm": 0.194207563996315, + "learning_rate": 6.679836395822744e-05, + "loss": 0.0265, + "step": 42140 + }, + { + "epoch": 8.18290374253677, + "grad_norm": 0.12793223559856415, + "learning_rate": 6.678278944333633e-05, + "loss": 0.0304, + "step": 42150 + }, + { + "epoch": 8.18484539585457, + "grad_norm": 0.1375778764486313, + "learning_rate": 6.676721309310462e-05, + "loss": 0.0395, + "step": 42160 + }, + { + "epoch": 8.18678704917237, + "grad_norm": 0.19843900203704834, + "learning_rate": 6.67516349092357e-05, + "loss": 0.0329, + "step": 42170 + }, + { + "epoch": 8.18872870249017, + "grad_norm": 0.2316676825284958, + "learning_rate": 6.673605489343315e-05, + "loss": 0.0234, + "step": 42180 + }, + { + "epoch": 8.19067035580797, + "grad_norm": 0.23170550167560577, + "learning_rate": 6.672047304740081e-05, + "loss": 0.0266, + "step": 42190 + }, + { + "epoch": 8.19261200912577, + "grad_norm": 0.14847593009471893, + "learning_rate": 6.670488937284268e-05, + "loss": 0.0282, + "step": 42200 + }, + { + "epoch": 8.194553662443571, + "grad_norm": 0.1435343325138092, + "learning_rate": 6.668930387146294e-05, + "loss": 0.0275, + "step": 42210 + }, + { + "epoch": 8.196495315761371, + "grad_norm": 0.24359974265098572, + "learning_rate": 6.667371654496605e-05, + "loss": 0.0261, + "step": 42220 + }, + { + "epoch": 8.198436969079172, + "grad_norm": 0.24028262495994568, + "learning_rate": 6.665812739505659e-05, + "loss": 0.0265, + "step": 42230 + }, + { + "epoch": 8.200378622396972, + "grad_norm": 0.1740829050540924, + "learning_rate": 6.664253642343935e-05, + "loss": 0.0267, + "step": 42240 + }, + { + "epoch": 8.20232027571477, + "grad_norm": 0.1860286146402359, + "learning_rate": 6.662694363181935e-05, + "loss": 0.0287, + "step": 42250 + }, + { + "epoch": 8.20426192903257, + "grad_norm": 0.19148191809654236, + "learning_rate": 6.66113490219018e-05, + "loss": 0.0259, + "step": 42260 + }, + { + "epoch": 8.206203582350371, + "grad_norm": 0.20429733395576477, + "learning_rate": 6.659575259539207e-05, + "loss": 0.0309, + "step": 42270 + }, + { + "epoch": 8.208145235668171, + "grad_norm": 0.18474651873111725, + "learning_rate": 6.65801543539958e-05, + "loss": 0.027, + "step": 42280 + }, + { + "epoch": 8.210086888985971, + "grad_norm": 0.2311110943555832, + "learning_rate": 6.65645542994188e-05, + "loss": 0.0271, + "step": 42290 + }, + { + "epoch": 8.212028542303772, + "grad_norm": 0.2132435441017151, + "learning_rate": 6.654895243336702e-05, + "loss": 0.0273, + "step": 42300 + }, + { + "epoch": 8.213970195621572, + "grad_norm": 0.19587478041648865, + "learning_rate": 6.65333487575467e-05, + "loss": 0.0293, + "step": 42310 + }, + { + "epoch": 8.215911848939372, + "grad_norm": 0.12851116061210632, + "learning_rate": 6.651774327366422e-05, + "loss": 0.0286, + "step": 42320 + }, + { + "epoch": 8.217853502257173, + "grad_norm": 0.21282045543193817, + "learning_rate": 6.650213598342613e-05, + "loss": 0.0218, + "step": 42330 + }, + { + "epoch": 8.219795155574973, + "grad_norm": 0.15376605093479156, + "learning_rate": 6.648652688853931e-05, + "loss": 0.0302, + "step": 42340 + }, + { + "epoch": 8.221736808892771, + "grad_norm": 0.23987513780593872, + "learning_rate": 6.647091599071066e-05, + "loss": 0.0315, + "step": 42350 + }, + { + "epoch": 8.223678462210572, + "grad_norm": 0.1200513169169426, + "learning_rate": 6.645530329164743e-05, + "loss": 0.0215, + "step": 42360 + }, + { + "epoch": 8.225620115528372, + "grad_norm": 0.09028749912977219, + "learning_rate": 6.643968879305694e-05, + "loss": 0.0222, + "step": 42370 + }, + { + "epoch": 8.227561768846172, + "grad_norm": 0.19007697701454163, + "learning_rate": 6.642407249664683e-05, + "loss": 0.0242, + "step": 42380 + }, + { + "epoch": 8.229503422163972, + "grad_norm": 0.20335637032985687, + "learning_rate": 6.640845440412483e-05, + "loss": 0.0286, + "step": 42390 + }, + { + "epoch": 8.231445075481773, + "grad_norm": 0.16792510449886322, + "learning_rate": 6.639283451719893e-05, + "loss": 0.0271, + "step": 42400 + }, + { + "epoch": 8.233386728799573, + "grad_norm": 0.207283616065979, + "learning_rate": 6.637721283757729e-05, + "loss": 0.0302, + "step": 42410 + }, + { + "epoch": 8.235328382117373, + "grad_norm": 0.09226012229919434, + "learning_rate": 6.636158936696828e-05, + "loss": 0.0211, + "step": 42420 + }, + { + "epoch": 8.237270035435174, + "grad_norm": 0.19640742242336273, + "learning_rate": 6.634596410708047e-05, + "loss": 0.0267, + "step": 42430 + }, + { + "epoch": 8.239211688752974, + "grad_norm": 0.17917051911354065, + "learning_rate": 6.633033705962257e-05, + "loss": 0.0292, + "step": 42440 + }, + { + "epoch": 8.241153342070774, + "grad_norm": 0.14698734879493713, + "learning_rate": 6.631470822630359e-05, + "loss": 0.023, + "step": 42450 + }, + { + "epoch": 8.243094995388573, + "grad_norm": 0.22111229598522186, + "learning_rate": 6.629907760883263e-05, + "loss": 0.0295, + "step": 42460 + }, + { + "epoch": 8.245036648706373, + "grad_norm": 0.1793995499610901, + "learning_rate": 6.628344520891905e-05, + "loss": 0.0296, + "step": 42470 + }, + { + "epoch": 8.246978302024173, + "grad_norm": 0.15680941939353943, + "learning_rate": 6.62678110282724e-05, + "loss": 0.0295, + "step": 42480 + }, + { + "epoch": 8.248919955341973, + "grad_norm": 0.19050870835781097, + "learning_rate": 6.625217506860237e-05, + "loss": 0.0366, + "step": 42490 + }, + { + "epoch": 8.250861608659774, + "grad_norm": 0.23852089047431946, + "learning_rate": 6.623653733161892e-05, + "loss": 0.0252, + "step": 42500 + }, + { + "epoch": 8.252803261977574, + "grad_norm": 0.14280903339385986, + "learning_rate": 6.622089781903217e-05, + "loss": 0.025, + "step": 42510 + }, + { + "epoch": 8.254744915295374, + "grad_norm": 0.14903317391872406, + "learning_rate": 6.620525653255241e-05, + "loss": 0.0309, + "step": 42520 + }, + { + "epoch": 8.256686568613175, + "grad_norm": 0.2349361926317215, + "learning_rate": 6.618961347389015e-05, + "loss": 0.0318, + "step": 42530 + }, + { + "epoch": 8.258628221930975, + "grad_norm": 0.25681188702583313, + "learning_rate": 6.617396864475613e-05, + "loss": 0.0297, + "step": 42540 + }, + { + "epoch": 8.260569875248775, + "grad_norm": 0.17399878799915314, + "learning_rate": 6.61583220468612e-05, + "loss": 0.0273, + "step": 42550 + }, + { + "epoch": 8.262511528566574, + "grad_norm": 0.1514386534690857, + "learning_rate": 6.614267368191645e-05, + "loss": 0.028, + "step": 42560 + }, + { + "epoch": 8.264453181884374, + "grad_norm": 0.19892823696136475, + "learning_rate": 6.61270235516332e-05, + "loss": 0.0218, + "step": 42570 + }, + { + "epoch": 8.266394835202174, + "grad_norm": 0.15188442170619965, + "learning_rate": 6.611137165772287e-05, + "loss": 0.0257, + "step": 42580 + }, + { + "epoch": 8.268336488519974, + "grad_norm": 0.24047651886940002, + "learning_rate": 6.60957180018972e-05, + "loss": 0.0356, + "step": 42590 + }, + { + "epoch": 8.270278141837775, + "grad_norm": 0.19980305433273315, + "learning_rate": 6.608006258586797e-05, + "loss": 0.0267, + "step": 42600 + }, + { + "epoch": 8.272219795155575, + "grad_norm": 0.24362994730472565, + "learning_rate": 6.606440541134731e-05, + "loss": 0.0278, + "step": 42610 + }, + { + "epoch": 8.274161448473375, + "grad_norm": 0.1287357658147812, + "learning_rate": 6.60487464800474e-05, + "loss": 0.0209, + "step": 42620 + }, + { + "epoch": 8.276103101791175, + "grad_norm": 0.23176275193691254, + "learning_rate": 6.603308579368071e-05, + "loss": 0.0241, + "step": 42630 + }, + { + "epoch": 8.278044755108976, + "grad_norm": 0.19333821535110474, + "learning_rate": 6.601742335395987e-05, + "loss": 0.0252, + "step": 42640 + }, + { + "epoch": 8.279986408426776, + "grad_norm": 0.2145097851753235, + "learning_rate": 6.600175916259769e-05, + "loss": 0.0266, + "step": 42650 + }, + { + "epoch": 8.281928061744576, + "grad_norm": 0.12161748856306076, + "learning_rate": 6.598609322130718e-05, + "loss": 0.0238, + "step": 42660 + }, + { + "epoch": 8.283869715062375, + "grad_norm": 0.1785440295934677, + "learning_rate": 6.597042553180155e-05, + "loss": 0.0256, + "step": 42670 + }, + { + "epoch": 8.285811368380175, + "grad_norm": 0.14551955461502075, + "learning_rate": 6.59547560957942e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 8.287753021697975, + "grad_norm": 0.14039050042629242, + "learning_rate": 6.59390849149987e-05, + "loss": 0.0229, + "step": 42690 + }, + { + "epoch": 8.289694675015776, + "grad_norm": 0.0908256322145462, + "learning_rate": 6.592341199112886e-05, + "loss": 0.0266, + "step": 42700 + }, + { + "epoch": 8.291636328333576, + "grad_norm": 0.14729976654052734, + "learning_rate": 6.590773732589861e-05, + "loss": 0.0216, + "step": 42710 + }, + { + "epoch": 8.293577981651376, + "grad_norm": 0.2700798213481903, + "learning_rate": 6.589206092102213e-05, + "loss": 0.039, + "step": 42720 + }, + { + "epoch": 8.295519634969176, + "grad_norm": 0.15492738783359528, + "learning_rate": 6.587638277821375e-05, + "loss": 0.0292, + "step": 42730 + }, + { + "epoch": 8.297461288286977, + "grad_norm": 0.08289746195077896, + "learning_rate": 6.586070289918803e-05, + "loss": 0.028, + "step": 42740 + }, + { + "epoch": 8.299402941604777, + "grad_norm": 0.20480456948280334, + "learning_rate": 6.584502128565968e-05, + "loss": 0.0286, + "step": 42750 + }, + { + "epoch": 8.301344594922577, + "grad_norm": 0.10657639056444168, + "learning_rate": 6.582933793934363e-05, + "loss": 0.0253, + "step": 42760 + }, + { + "epoch": 8.303286248240378, + "grad_norm": 0.15614083409309387, + "learning_rate": 6.581365286195498e-05, + "loss": 0.0265, + "step": 42770 + }, + { + "epoch": 8.305227901558176, + "grad_norm": 0.10411017388105392, + "learning_rate": 6.579796605520903e-05, + "loss": 0.0203, + "step": 42780 + }, + { + "epoch": 8.307169554875976, + "grad_norm": 0.16126573085784912, + "learning_rate": 6.578227752082127e-05, + "loss": 0.0285, + "step": 42790 + }, + { + "epoch": 8.309111208193777, + "grad_norm": 0.22927962243556976, + "learning_rate": 6.576658726050735e-05, + "loss": 0.0301, + "step": 42800 + }, + { + "epoch": 8.311052861511577, + "grad_norm": 0.14666755497455597, + "learning_rate": 6.575089527598316e-05, + "loss": 0.0256, + "step": 42810 + }, + { + "epoch": 8.312994514829377, + "grad_norm": 0.16200187802314758, + "learning_rate": 6.573520156896476e-05, + "loss": 0.0274, + "step": 42820 + }, + { + "epoch": 8.314936168147177, + "grad_norm": 0.17111891508102417, + "learning_rate": 6.571950614116835e-05, + "loss": 0.0295, + "step": 42830 + }, + { + "epoch": 8.316877821464978, + "grad_norm": 0.10907305032014847, + "learning_rate": 6.570380899431039e-05, + "loss": 0.0288, + "step": 42840 + }, + { + "epoch": 8.318819474782778, + "grad_norm": 0.16422732174396515, + "learning_rate": 6.568811013010749e-05, + "loss": 0.0257, + "step": 42850 + }, + { + "epoch": 8.320761128100578, + "grad_norm": 0.19934937357902527, + "learning_rate": 6.567240955027645e-05, + "loss": 0.0296, + "step": 42860 + }, + { + "epoch": 8.322702781418378, + "grad_norm": 0.19803163409233093, + "learning_rate": 6.565670725653427e-05, + "loss": 0.0256, + "step": 42870 + }, + { + "epoch": 8.324644434736179, + "grad_norm": 0.1468823254108429, + "learning_rate": 6.564100325059812e-05, + "loss": 0.0254, + "step": 42880 + }, + { + "epoch": 8.326586088053977, + "grad_norm": 0.13660180568695068, + "learning_rate": 6.562529753418538e-05, + "loss": 0.0323, + "step": 42890 + }, + { + "epoch": 8.328527741371778, + "grad_norm": 0.2641386091709137, + "learning_rate": 6.56095901090136e-05, + "loss": 0.0315, + "step": 42900 + }, + { + "epoch": 8.330469394689578, + "grad_norm": 0.20985394716262817, + "learning_rate": 6.55938809768005e-05, + "loss": 0.0286, + "step": 42910 + }, + { + "epoch": 8.332411048007378, + "grad_norm": 0.12419769167900085, + "learning_rate": 6.557817013926402e-05, + "loss": 0.0259, + "step": 42920 + }, + { + "epoch": 8.334352701325178, + "grad_norm": 0.08551289141178131, + "learning_rate": 6.556245759812227e-05, + "loss": 0.0229, + "step": 42930 + }, + { + "epoch": 8.336294354642979, + "grad_norm": 0.1873006820678711, + "learning_rate": 6.554674335509357e-05, + "loss": 0.0279, + "step": 42940 + }, + { + "epoch": 8.338236007960779, + "grad_norm": 0.1215183436870575, + "learning_rate": 6.553102741189638e-05, + "loss": 0.028, + "step": 42950 + }, + { + "epoch": 8.34017766127858, + "grad_norm": 0.20463192462921143, + "learning_rate": 6.55153097702494e-05, + "loss": 0.0258, + "step": 42960 + }, + { + "epoch": 8.34211931459638, + "grad_norm": 0.14152246713638306, + "learning_rate": 6.549959043187146e-05, + "loss": 0.0243, + "step": 42970 + }, + { + "epoch": 8.34406096791418, + "grad_norm": 0.2753293514251709, + "learning_rate": 6.54838693984816e-05, + "loss": 0.0251, + "step": 42980 + }, + { + "epoch": 8.346002621231978, + "grad_norm": 0.10512479394674301, + "learning_rate": 6.546814667179909e-05, + "loss": 0.0267, + "step": 42990 + }, + { + "epoch": 8.347944274549778, + "grad_norm": 0.16737912595272064, + "learning_rate": 6.545242225354328e-05, + "loss": 0.0256, + "step": 43000 + }, + { + "epoch": 8.349885927867579, + "grad_norm": 0.13906964659690857, + "learning_rate": 6.543669614543383e-05, + "loss": 0.0279, + "step": 43010 + }, + { + "epoch": 8.351827581185379, + "grad_norm": 0.22635680437088013, + "learning_rate": 6.542096834919049e-05, + "loss": 0.0319, + "step": 43020 + }, + { + "epoch": 8.35376923450318, + "grad_norm": 0.12701521813869476, + "learning_rate": 6.540523886653324e-05, + "loss": 0.0243, + "step": 43030 + }, + { + "epoch": 8.35571088782098, + "grad_norm": 0.15352892875671387, + "learning_rate": 6.538950769918222e-05, + "loss": 0.0299, + "step": 43040 + }, + { + "epoch": 8.35765254113878, + "grad_norm": 0.1946544647216797, + "learning_rate": 6.537377484885779e-05, + "loss": 0.0264, + "step": 43050 + }, + { + "epoch": 8.35959419445658, + "grad_norm": 0.11953651160001755, + "learning_rate": 6.535804031728045e-05, + "loss": 0.0234, + "step": 43060 + }, + { + "epoch": 8.36153584777438, + "grad_norm": 0.12283200025558472, + "learning_rate": 6.534230410617092e-05, + "loss": 0.0248, + "step": 43070 + }, + { + "epoch": 8.36347750109218, + "grad_norm": 0.1690419465303421, + "learning_rate": 6.532656621725007e-05, + "loss": 0.0236, + "step": 43080 + }, + { + "epoch": 8.365419154409981, + "grad_norm": 0.1130574643611908, + "learning_rate": 6.531082665223898e-05, + "loss": 0.0301, + "step": 43090 + }, + { + "epoch": 8.36736080772778, + "grad_norm": 0.17220142483711243, + "learning_rate": 6.529508541285889e-05, + "loss": 0.0224, + "step": 43100 + }, + { + "epoch": 8.36930246104558, + "grad_norm": 0.20985785126686096, + "learning_rate": 6.527934250083128e-05, + "loss": 0.031, + "step": 43110 + }, + { + "epoch": 8.37124411436338, + "grad_norm": 0.20462404191493988, + "learning_rate": 6.526359791787775e-05, + "loss": 0.0279, + "step": 43120 + }, + { + "epoch": 8.37318576768118, + "grad_norm": 0.20215846598148346, + "learning_rate": 6.524785166572009e-05, + "loss": 0.0257, + "step": 43130 + }, + { + "epoch": 8.37512742099898, + "grad_norm": 0.23012416064739227, + "learning_rate": 6.523210374608029e-05, + "loss": 0.0241, + "step": 43140 + }, + { + "epoch": 8.37706907431678, + "grad_norm": 0.19319188594818115, + "learning_rate": 6.521635416068054e-05, + "loss": 0.0256, + "step": 43150 + }, + { + "epoch": 8.379010727634581, + "grad_norm": 0.15571564435958862, + "learning_rate": 6.520060291124317e-05, + "loss": 0.0291, + "step": 43160 + }, + { + "epoch": 8.380952380952381, + "grad_norm": 0.29223600029945374, + "learning_rate": 6.518484999949072e-05, + "loss": 0.0228, + "step": 43170 + }, + { + "epoch": 8.382894034270182, + "grad_norm": 0.10183652490377426, + "learning_rate": 6.516909542714591e-05, + "loss": 0.036, + "step": 43180 + }, + { + "epoch": 8.384835687587982, + "grad_norm": 0.21278776228427887, + "learning_rate": 6.515333919593161e-05, + "loss": 0.0318, + "step": 43190 + }, + { + "epoch": 8.38677734090578, + "grad_norm": 0.17547547817230225, + "learning_rate": 6.513758130757094e-05, + "loss": 0.0246, + "step": 43200 + }, + { + "epoch": 8.38871899422358, + "grad_norm": 0.21567393839359283, + "learning_rate": 6.512182176378713e-05, + "loss": 0.0261, + "step": 43210 + }, + { + "epoch": 8.390660647541381, + "grad_norm": 0.22367992997169495, + "learning_rate": 6.510606056630362e-05, + "loss": 0.0341, + "step": 43220 + }, + { + "epoch": 8.392602300859181, + "grad_norm": 0.16549620032310486, + "learning_rate": 6.509029771684405e-05, + "loss": 0.0249, + "step": 43230 + }, + { + "epoch": 8.394543954176982, + "grad_norm": 0.194001242518425, + "learning_rate": 6.50745332171322e-05, + "loss": 0.0293, + "step": 43240 + }, + { + "epoch": 8.396485607494782, + "grad_norm": 0.15785016119480133, + "learning_rate": 6.505876706889207e-05, + "loss": 0.0309, + "step": 43250 + }, + { + "epoch": 8.398427260812582, + "grad_norm": 0.14469340443611145, + "learning_rate": 6.504299927384781e-05, + "loss": 0.0235, + "step": 43260 + }, + { + "epoch": 8.400368914130382, + "grad_norm": 0.17866173386573792, + "learning_rate": 6.502722983372378e-05, + "loss": 0.0222, + "step": 43270 + }, + { + "epoch": 8.402310567448183, + "grad_norm": 0.19136443734169006, + "learning_rate": 6.501145875024446e-05, + "loss": 0.0273, + "step": 43280 + }, + { + "epoch": 8.404252220765983, + "grad_norm": 0.11577443778514862, + "learning_rate": 6.499568602513459e-05, + "loss": 0.0315, + "step": 43290 + }, + { + "epoch": 8.406193874083783, + "grad_norm": 0.16605320572853088, + "learning_rate": 6.497991166011903e-05, + "loss": 0.0264, + "step": 43300 + }, + { + "epoch": 8.408135527401582, + "grad_norm": 0.12810075283050537, + "learning_rate": 6.49641356569229e-05, + "loss": 0.0231, + "step": 43310 + }, + { + "epoch": 8.410077180719382, + "grad_norm": 0.18264473974704742, + "learning_rate": 6.494835801727135e-05, + "loss": 0.0262, + "step": 43320 + }, + { + "epoch": 8.412018834037182, + "grad_norm": 0.07102327048778534, + "learning_rate": 6.493257874288986e-05, + "loss": 0.0246, + "step": 43330 + }, + { + "epoch": 8.413960487354982, + "grad_norm": 0.2751195430755615, + "learning_rate": 6.491679783550399e-05, + "loss": 0.0292, + "step": 43340 + }, + { + "epoch": 8.415902140672783, + "grad_norm": 0.15032462775707245, + "learning_rate": 6.490101529683957e-05, + "loss": 0.0352, + "step": 43350 + }, + { + "epoch": 8.417843793990583, + "grad_norm": 0.181746244430542, + "learning_rate": 6.488523112862248e-05, + "loss": 0.0298, + "step": 43360 + }, + { + "epoch": 8.419785447308383, + "grad_norm": 0.23322521150112152, + "learning_rate": 6.486944533257891e-05, + "loss": 0.0296, + "step": 43370 + }, + { + "epoch": 8.421727100626184, + "grad_norm": 0.1302754133939743, + "learning_rate": 6.485365791043517e-05, + "loss": 0.0312, + "step": 43380 + }, + { + "epoch": 8.423668753943984, + "grad_norm": 0.2193375527858734, + "learning_rate": 6.483786886391772e-05, + "loss": 0.0248, + "step": 43390 + }, + { + "epoch": 8.425610407261784, + "grad_norm": 0.17572437226772308, + "learning_rate": 6.482207819475323e-05, + "loss": 0.0321, + "step": 43400 + }, + { + "epoch": 8.427552060579583, + "grad_norm": 0.20295946300029755, + "learning_rate": 6.480628590466857e-05, + "loss": 0.0319, + "step": 43410 + }, + { + "epoch": 8.429493713897383, + "grad_norm": 0.18248088657855988, + "learning_rate": 6.479049199539074e-05, + "loss": 0.0282, + "step": 43420 + }, + { + "epoch": 8.431435367215183, + "grad_norm": 0.15618182718753815, + "learning_rate": 6.477469646864693e-05, + "loss": 0.0238, + "step": 43430 + }, + { + "epoch": 8.433377020532983, + "grad_norm": 0.15643300116062164, + "learning_rate": 6.475889932616454e-05, + "loss": 0.0307, + "step": 43440 + }, + { + "epoch": 8.435318673850784, + "grad_norm": 0.1290995478630066, + "learning_rate": 6.474310056967111e-05, + "loss": 0.025, + "step": 43450 + }, + { + "epoch": 8.437260327168584, + "grad_norm": 0.177011176943779, + "learning_rate": 6.472730020089437e-05, + "loss": 0.0265, + "step": 43460 + }, + { + "epoch": 8.439201980486384, + "grad_norm": 0.1729496270418167, + "learning_rate": 6.471149822156222e-05, + "loss": 0.0287, + "step": 43470 + }, + { + "epoch": 8.441143633804185, + "grad_norm": 0.27438977360725403, + "learning_rate": 6.469569463340275e-05, + "loss": 0.0287, + "step": 43480 + }, + { + "epoch": 8.443085287121985, + "grad_norm": 0.23286840319633484, + "learning_rate": 6.467988943814418e-05, + "loss": 0.0277, + "step": 43490 + }, + { + "epoch": 8.445026940439785, + "grad_norm": 0.22669756412506104, + "learning_rate": 6.4664082637515e-05, + "loss": 0.0279, + "step": 43500 + }, + { + "epoch": 8.446968593757585, + "grad_norm": 0.16202673316001892, + "learning_rate": 6.464827423324377e-05, + "loss": 0.0265, + "step": 43510 + }, + { + "epoch": 8.448910247075384, + "grad_norm": 0.22033149003982544, + "learning_rate": 6.463246422705931e-05, + "loss": 0.0278, + "step": 43520 + }, + { + "epoch": 8.450851900393184, + "grad_norm": 0.07379832863807678, + "learning_rate": 6.461665262069055e-05, + "loss": 0.024, + "step": 43530 + }, + { + "epoch": 8.452793553710984, + "grad_norm": 0.11943241208791733, + "learning_rate": 6.460083941586665e-05, + "loss": 0.0314, + "step": 43540 + }, + { + "epoch": 8.454735207028785, + "grad_norm": 0.18466946482658386, + "learning_rate": 6.45850246143169e-05, + "loss": 0.0217, + "step": 43550 + }, + { + "epoch": 8.456676860346585, + "grad_norm": 0.16442731022834778, + "learning_rate": 6.456920821777077e-05, + "loss": 0.0232, + "step": 43560 + }, + { + "epoch": 8.458618513664385, + "grad_norm": 0.277180552482605, + "learning_rate": 6.455339022795796e-05, + "loss": 0.0329, + "step": 43570 + }, + { + "epoch": 8.460560166982185, + "grad_norm": 0.14132724702358246, + "learning_rate": 6.453757064660825e-05, + "loss": 0.0284, + "step": 43580 + }, + { + "epoch": 8.462501820299986, + "grad_norm": 0.2879060208797455, + "learning_rate": 6.452174947545169e-05, + "loss": 0.0266, + "step": 43590 + }, + { + "epoch": 8.464443473617786, + "grad_norm": 0.1743650883436203, + "learning_rate": 6.450592671621842e-05, + "loss": 0.0271, + "step": 43600 + }, + { + "epoch": 8.466385126935586, + "grad_norm": 0.22296389937400818, + "learning_rate": 6.449010237063882e-05, + "loss": 0.0237, + "step": 43610 + }, + { + "epoch": 8.468326780253387, + "grad_norm": 0.1075686514377594, + "learning_rate": 6.447427644044343e-05, + "loss": 0.0249, + "step": 43620 + }, + { + "epoch": 8.470268433571185, + "grad_norm": 0.11293010413646698, + "learning_rate": 6.44584489273629e-05, + "loss": 0.0323, + "step": 43630 + }, + { + "epoch": 8.472210086888985, + "grad_norm": 0.1594226211309433, + "learning_rate": 6.444261983312817e-05, + "loss": 0.0231, + "step": 43640 + }, + { + "epoch": 8.474151740206786, + "grad_norm": 0.19333282113075256, + "learning_rate": 6.442678915947023e-05, + "loss": 0.0244, + "step": 43650 + }, + { + "epoch": 8.476093393524586, + "grad_norm": 0.18889367580413818, + "learning_rate": 6.441095690812032e-05, + "loss": 0.028, + "step": 43660 + }, + { + "epoch": 8.478035046842386, + "grad_norm": 0.17352573573589325, + "learning_rate": 6.439512308080982e-05, + "loss": 0.0274, + "step": 43670 + }, + { + "epoch": 8.479976700160186, + "grad_norm": 0.0826498344540596, + "learning_rate": 6.43792876792703e-05, + "loss": 0.0243, + "step": 43680 + }, + { + "epoch": 8.481918353477987, + "grad_norm": 0.133263498544693, + "learning_rate": 6.43634507052335e-05, + "loss": 0.0274, + "step": 43690 + }, + { + "epoch": 8.483860006795787, + "grad_norm": 0.1395699679851532, + "learning_rate": 6.434761216043135e-05, + "loss": 0.0271, + "step": 43700 + }, + { + "epoch": 8.485801660113587, + "grad_norm": 0.1683696061372757, + "learning_rate": 6.433177204659589e-05, + "loss": 0.0347, + "step": 43710 + }, + { + "epoch": 8.487743313431388, + "grad_norm": 0.09828544408082962, + "learning_rate": 6.431593036545938e-05, + "loss": 0.0241, + "step": 43720 + }, + { + "epoch": 8.489684966749188, + "grad_norm": 0.17342813313007355, + "learning_rate": 6.430008711875426e-05, + "loss": 0.0323, + "step": 43730 + }, + { + "epoch": 8.491626620066986, + "grad_norm": 0.14902658760547638, + "learning_rate": 6.42842423082131e-05, + "loss": 0.0264, + "step": 43740 + }, + { + "epoch": 8.493568273384787, + "grad_norm": 0.1720750331878662, + "learning_rate": 6.426839593556868e-05, + "loss": 0.0316, + "step": 43750 + }, + { + "epoch": 8.495509926702587, + "grad_norm": 0.2728644907474518, + "learning_rate": 6.425254800255392e-05, + "loss": 0.0323, + "step": 43760 + }, + { + "epoch": 8.497451580020387, + "grad_norm": 0.14953741431236267, + "learning_rate": 6.423669851090196e-05, + "loss": 0.0252, + "step": 43770 + }, + { + "epoch": 8.499393233338187, + "grad_norm": 0.24346503615379333, + "learning_rate": 6.422084746234604e-05, + "loss": 0.0275, + "step": 43780 + }, + { + "epoch": 8.501334886655988, + "grad_norm": 0.13585765659809113, + "learning_rate": 6.420499485861965e-05, + "loss": 0.0243, + "step": 43790 + }, + { + "epoch": 8.503276539973788, + "grad_norm": 0.1681412011384964, + "learning_rate": 6.418914070145636e-05, + "loss": 0.0267, + "step": 43800 + }, + { + "epoch": 8.505218193291588, + "grad_norm": 0.2540476322174072, + "learning_rate": 6.417328499258996e-05, + "loss": 0.0232, + "step": 43810 + }, + { + "epoch": 8.507159846609389, + "grad_norm": 0.17615796625614166, + "learning_rate": 6.415742773375446e-05, + "loss": 0.0315, + "step": 43820 + }, + { + "epoch": 8.509101499927189, + "grad_norm": 0.2114793062210083, + "learning_rate": 6.414156892668393e-05, + "loss": 0.0273, + "step": 43830 + }, + { + "epoch": 8.511043153244987, + "grad_norm": 0.17025801539421082, + "learning_rate": 6.412570857311267e-05, + "loss": 0.0195, + "step": 43840 + }, + { + "epoch": 8.512984806562788, + "grad_norm": 0.27584975957870483, + "learning_rate": 6.410984667477518e-05, + "loss": 0.0277, + "step": 43850 + }, + { + "epoch": 8.514926459880588, + "grad_norm": 0.24439023435115814, + "learning_rate": 6.409398323340607e-05, + "loss": 0.0328, + "step": 43860 + }, + { + "epoch": 8.516868113198388, + "grad_norm": 0.1547064632177353, + "learning_rate": 6.407811825074012e-05, + "loss": 0.0298, + "step": 43870 + }, + { + "epoch": 8.518809766516188, + "grad_norm": 0.20913031697273254, + "learning_rate": 6.406225172851234e-05, + "loss": 0.0262, + "step": 43880 + }, + { + "epoch": 8.520751419833989, + "grad_norm": 0.2070881575345993, + "learning_rate": 6.404638366845786e-05, + "loss": 0.0284, + "step": 43890 + }, + { + "epoch": 8.522693073151789, + "grad_norm": 0.2380957007408142, + "learning_rate": 6.403051407231196e-05, + "loss": 0.0268, + "step": 43900 + }, + { + "epoch": 8.52463472646959, + "grad_norm": 0.22623611986637115, + "learning_rate": 6.401464294181016e-05, + "loss": 0.0212, + "step": 43910 + }, + { + "epoch": 8.52657637978739, + "grad_norm": 0.0978052169084549, + "learning_rate": 6.399877027868808e-05, + "loss": 0.0239, + "step": 43920 + }, + { + "epoch": 8.52851803310519, + "grad_norm": 0.12714867293834686, + "learning_rate": 6.39828960846815e-05, + "loss": 0.02, + "step": 43930 + }, + { + "epoch": 8.53045968642299, + "grad_norm": 0.13819850981235504, + "learning_rate": 6.396702036152644e-05, + "loss": 0.0231, + "step": 43940 + }, + { + "epoch": 8.532401339740789, + "grad_norm": 0.10191885381937027, + "learning_rate": 6.395114311095905e-05, + "loss": 0.025, + "step": 43950 + }, + { + "epoch": 8.534342993058589, + "grad_norm": 0.16838766634464264, + "learning_rate": 6.393526433471559e-05, + "loss": 0.031, + "step": 43960 + }, + { + "epoch": 8.536284646376389, + "grad_norm": 0.21337653696537018, + "learning_rate": 6.391938403453262e-05, + "loss": 0.0418, + "step": 43970 + }, + { + "epoch": 8.53822629969419, + "grad_norm": 0.1472337245941162, + "learning_rate": 6.390350221214671e-05, + "loss": 0.031, + "step": 43980 + }, + { + "epoch": 8.54016795301199, + "grad_norm": 0.1267649531364441, + "learning_rate": 6.38876188692947e-05, + "loss": 0.0207, + "step": 43990 + }, + { + "epoch": 8.54210960632979, + "grad_norm": 0.2140641212463379, + "learning_rate": 6.387173400771358e-05, + "loss": 0.0247, + "step": 44000 + }, + { + "epoch": 8.54405125964759, + "grad_norm": 0.20046818256378174, + "learning_rate": 6.385584762914047e-05, + "loss": 0.0235, + "step": 44010 + }, + { + "epoch": 8.54599291296539, + "grad_norm": 0.16049738228321075, + "learning_rate": 6.38399597353127e-05, + "loss": 0.0294, + "step": 44020 + }, + { + "epoch": 8.54793456628319, + "grad_norm": 0.12608377635478973, + "learning_rate": 6.382407032796775e-05, + "loss": 0.0275, + "step": 44030 + }, + { + "epoch": 8.549876219600991, + "grad_norm": 0.22340373694896698, + "learning_rate": 6.380817940884325e-05, + "loss": 0.0316, + "step": 44040 + }, + { + "epoch": 8.55181787291879, + "grad_norm": 0.17683127522468567, + "learning_rate": 6.379228697967702e-05, + "loss": 0.0248, + "step": 44050 + }, + { + "epoch": 8.55375952623659, + "grad_norm": 0.15896090865135193, + "learning_rate": 6.3776393042207e-05, + "loss": 0.0264, + "step": 44060 + }, + { + "epoch": 8.55570117955439, + "grad_norm": 0.17995567619800568, + "learning_rate": 6.376049759817137e-05, + "loss": 0.0256, + "step": 44070 + }, + { + "epoch": 8.55764283287219, + "grad_norm": 0.15370307862758636, + "learning_rate": 6.37446006493084e-05, + "loss": 0.0268, + "step": 44080 + }, + { + "epoch": 8.55958448618999, + "grad_norm": 0.12752008438110352, + "learning_rate": 6.372870219735656e-05, + "loss": 0.0263, + "step": 44090 + }, + { + "epoch": 8.56152613950779, + "grad_norm": 0.16228023171424866, + "learning_rate": 6.371280224405452e-05, + "loss": 0.0307, + "step": 44100 + }, + { + "epoch": 8.563467792825591, + "grad_norm": 0.2795092761516571, + "learning_rate": 6.369690079114101e-05, + "loss": 0.0348, + "step": 44110 + }, + { + "epoch": 8.565409446143391, + "grad_norm": 0.15773102641105652, + "learning_rate": 6.368099784035504e-05, + "loss": 0.0218, + "step": 44120 + }, + { + "epoch": 8.567351099461192, + "grad_norm": 0.1740136295557022, + "learning_rate": 6.366509339343572e-05, + "loss": 0.0202, + "step": 44130 + }, + { + "epoch": 8.569292752778992, + "grad_norm": 0.21124009788036346, + "learning_rate": 6.364918745212232e-05, + "loss": 0.0243, + "step": 44140 + }, + { + "epoch": 8.571234406096792, + "grad_norm": 0.09621188789606094, + "learning_rate": 6.363328001815431e-05, + "loss": 0.0247, + "step": 44150 + }, + { + "epoch": 8.57317605941459, + "grad_norm": 0.12043247371912003, + "learning_rate": 6.361737109327128e-05, + "loss": 0.0283, + "step": 44160 + }, + { + "epoch": 8.575117712732391, + "grad_norm": 0.14766855537891388, + "learning_rate": 6.360146067921301e-05, + "loss": 0.0245, + "step": 44170 + }, + { + "epoch": 8.577059366050191, + "grad_norm": 0.12462431192398071, + "learning_rate": 6.35855487777195e-05, + "loss": 0.025, + "step": 44180 + }, + { + "epoch": 8.579001019367992, + "grad_norm": 0.12997011840343475, + "learning_rate": 6.356963539053076e-05, + "loss": 0.022, + "step": 44190 + }, + { + "epoch": 8.580942672685792, + "grad_norm": 0.1671791970729828, + "learning_rate": 6.355372051938711e-05, + "loss": 0.0284, + "step": 44200 + }, + { + "epoch": 8.582884326003592, + "grad_norm": 0.13076047599315643, + "learning_rate": 6.353780416602894e-05, + "loss": 0.0246, + "step": 44210 + }, + { + "epoch": 8.584825979321392, + "grad_norm": 0.2725337743759155, + "learning_rate": 6.352188633219689e-05, + "loss": 0.033, + "step": 44220 + }, + { + "epoch": 8.586767632639193, + "grad_norm": 0.28869402408599854, + "learning_rate": 6.350596701963166e-05, + "loss": 0.0361, + "step": 44230 + }, + { + "epoch": 8.588709285956993, + "grad_norm": 0.19374416768550873, + "learning_rate": 6.349004623007419e-05, + "loss": 0.0251, + "step": 44240 + }, + { + "epoch": 8.590650939274793, + "grad_norm": 0.2408381551504135, + "learning_rate": 6.347412396526555e-05, + "loss": 0.027, + "step": 44250 + }, + { + "epoch": 8.592592592592592, + "grad_norm": 0.17304059863090515, + "learning_rate": 6.345820022694696e-05, + "loss": 0.0271, + "step": 44260 + }, + { + "epoch": 8.594534245910392, + "grad_norm": 0.13858437538146973, + "learning_rate": 6.344227501685984e-05, + "loss": 0.0223, + "step": 44270 + }, + { + "epoch": 8.596475899228192, + "grad_norm": 0.20733502507209778, + "learning_rate": 6.342634833674572e-05, + "loss": 0.0275, + "step": 44280 + }, + { + "epoch": 8.598417552545992, + "grad_norm": 0.17392104864120483, + "learning_rate": 6.341042018834635e-05, + "loss": 0.0267, + "step": 44290 + }, + { + "epoch": 8.600359205863793, + "grad_norm": 0.2639080286026001, + "learning_rate": 6.339449057340359e-05, + "loss": 0.0248, + "step": 44300 + }, + { + "epoch": 8.602300859181593, + "grad_norm": 0.20609495043754578, + "learning_rate": 6.337855949365945e-05, + "loss": 0.0327, + "step": 44310 + }, + { + "epoch": 8.604242512499393, + "grad_norm": 0.24089179933071136, + "learning_rate": 6.336262695085619e-05, + "loss": 0.0299, + "step": 44320 + }, + { + "epoch": 8.606184165817194, + "grad_norm": 0.20204490423202515, + "learning_rate": 6.334669294673612e-05, + "loss": 0.0311, + "step": 44330 + }, + { + "epoch": 8.608125819134994, + "grad_norm": 0.24624280631542206, + "learning_rate": 6.333075748304179e-05, + "loss": 0.0243, + "step": 44340 + }, + { + "epoch": 8.610067472452794, + "grad_norm": 0.09950239211320877, + "learning_rate": 6.331482056151585e-05, + "loss": 0.0211, + "step": 44350 + }, + { + "epoch": 8.612009125770594, + "grad_norm": 0.199923574924469, + "learning_rate": 6.329888218390117e-05, + "loss": 0.0265, + "step": 44360 + }, + { + "epoch": 8.613950779088395, + "grad_norm": 0.1561402827501297, + "learning_rate": 6.328294235194072e-05, + "loss": 0.0271, + "step": 44370 + }, + { + "epoch": 8.615892432406193, + "grad_norm": 0.14098180830478668, + "learning_rate": 6.326700106737765e-05, + "loss": 0.0296, + "step": 44380 + }, + { + "epoch": 8.617834085723993, + "grad_norm": 0.1319088339805603, + "learning_rate": 6.325105833195531e-05, + "loss": 0.021, + "step": 44390 + }, + { + "epoch": 8.619775739041794, + "grad_norm": 0.23861296474933624, + "learning_rate": 6.323511414741715e-05, + "loss": 0.0272, + "step": 44400 + }, + { + "epoch": 8.621717392359594, + "grad_norm": 0.11526218801736832, + "learning_rate": 6.321916851550678e-05, + "loss": 0.0227, + "step": 44410 + }, + { + "epoch": 8.623659045677394, + "grad_norm": 0.11725404858589172, + "learning_rate": 6.320322143796806e-05, + "loss": 0.0215, + "step": 44420 + }, + { + "epoch": 8.625600698995195, + "grad_norm": 0.22073222696781158, + "learning_rate": 6.318727291654488e-05, + "loss": 0.0308, + "step": 44430 + }, + { + "epoch": 8.627542352312995, + "grad_norm": 0.14971092343330383, + "learning_rate": 6.317132295298134e-05, + "loss": 0.0327, + "step": 44440 + }, + { + "epoch": 8.629484005630795, + "grad_norm": 0.110050730407238, + "learning_rate": 6.315537154902173e-05, + "loss": 0.0296, + "step": 44450 + }, + { + "epoch": 8.631425658948595, + "grad_norm": 0.15145277976989746, + "learning_rate": 6.313941870641048e-05, + "loss": 0.0275, + "step": 44460 + }, + { + "epoch": 8.633367312266396, + "grad_norm": 0.13257865607738495, + "learning_rate": 6.312346442689214e-05, + "loss": 0.0222, + "step": 44470 + }, + { + "epoch": 8.635308965584194, + "grad_norm": 0.17036359012126923, + "learning_rate": 6.310750871221147e-05, + "loss": 0.0255, + "step": 44480 + }, + { + "epoch": 8.637250618901994, + "grad_norm": 0.1104377880692482, + "learning_rate": 6.309155156411335e-05, + "loss": 0.0313, + "step": 44490 + }, + { + "epoch": 8.639192272219795, + "grad_norm": 0.14031165838241577, + "learning_rate": 6.307559298434284e-05, + "loss": 0.0259, + "step": 44500 + }, + { + "epoch": 8.641133925537595, + "grad_norm": 0.1424546092748642, + "learning_rate": 6.305963297464512e-05, + "loss": 0.0329, + "step": 44510 + }, + { + "epoch": 8.643075578855395, + "grad_norm": 0.18832707405090332, + "learning_rate": 6.304367153676561e-05, + "loss": 0.0329, + "step": 44520 + }, + { + "epoch": 8.645017232173196, + "grad_norm": 0.14775702357292175, + "learning_rate": 6.302770867244978e-05, + "loss": 0.0245, + "step": 44530 + }, + { + "epoch": 8.646958885490996, + "grad_norm": 0.0782972201704979, + "learning_rate": 6.301174438344328e-05, + "loss": 0.0258, + "step": 44540 + }, + { + "epoch": 8.648900538808796, + "grad_norm": 0.2577759921550751, + "learning_rate": 6.299577867149202e-05, + "loss": 0.0276, + "step": 44550 + }, + { + "epoch": 8.650842192126596, + "grad_norm": 0.29113033413887024, + "learning_rate": 6.297981153834192e-05, + "loss": 0.0262, + "step": 44560 + }, + { + "epoch": 8.652783845444397, + "grad_norm": 0.12486786395311356, + "learning_rate": 6.296384298573916e-05, + "loss": 0.0197, + "step": 44570 + }, + { + "epoch": 8.654725498762197, + "grad_norm": 0.1442822515964508, + "learning_rate": 6.294787301543001e-05, + "loss": 0.0253, + "step": 44580 + }, + { + "epoch": 8.656667152079995, + "grad_norm": 0.09871888160705566, + "learning_rate": 6.293190162916095e-05, + "loss": 0.0231, + "step": 44590 + }, + { + "epoch": 8.658608805397796, + "grad_norm": 0.19800913333892822, + "learning_rate": 6.291592882867855e-05, + "loss": 0.0258, + "step": 44600 + }, + { + "epoch": 8.660550458715596, + "grad_norm": 0.16245347261428833, + "learning_rate": 6.28999546157296e-05, + "loss": 0.0245, + "step": 44610 + }, + { + "epoch": 8.662492112033396, + "grad_norm": 0.21559762954711914, + "learning_rate": 6.288397899206102e-05, + "loss": 0.0263, + "step": 44620 + }, + { + "epoch": 8.664433765351196, + "grad_norm": 0.22539915144443512, + "learning_rate": 6.286800195941984e-05, + "loss": 0.0292, + "step": 44630 + }, + { + "epoch": 8.666375418668997, + "grad_norm": 0.19904190301895142, + "learning_rate": 6.285202351955334e-05, + "loss": 0.0307, + "step": 44640 + }, + { + "epoch": 8.668317071986797, + "grad_norm": 0.16693677008152008, + "learning_rate": 6.283604367420887e-05, + "loss": 0.0283, + "step": 44650 + }, + { + "epoch": 8.670258725304597, + "grad_norm": 0.11192738264799118, + "learning_rate": 6.282006242513394e-05, + "loss": 0.0237, + "step": 44660 + }, + { + "epoch": 8.672200378622398, + "grad_norm": 0.16561850905418396, + "learning_rate": 6.280407977407628e-05, + "loss": 0.0324, + "step": 44670 + }, + { + "epoch": 8.674142031940198, + "grad_norm": 0.16220787167549133, + "learning_rate": 6.27880957227837e-05, + "loss": 0.0247, + "step": 44680 + }, + { + "epoch": 8.676083685257996, + "grad_norm": 0.2784339487552643, + "learning_rate": 6.27721102730042e-05, + "loss": 0.0257, + "step": 44690 + }, + { + "epoch": 8.678025338575797, + "grad_norm": 0.15080876648426056, + "learning_rate": 6.275612342648592e-05, + "loss": 0.0299, + "step": 44700 + }, + { + "epoch": 8.679966991893597, + "grad_norm": 0.20663700997829437, + "learning_rate": 6.274013518497716e-05, + "loss": 0.0215, + "step": 44710 + }, + { + "epoch": 8.681908645211397, + "grad_norm": 0.1857842653989792, + "learning_rate": 6.272414555022636e-05, + "loss": 0.0233, + "step": 44720 + }, + { + "epoch": 8.683850298529197, + "grad_norm": 0.2011844664812088, + "learning_rate": 6.270815452398215e-05, + "loss": 0.0254, + "step": 44730 + }, + { + "epoch": 8.685791951846998, + "grad_norm": 0.2315995693206787, + "learning_rate": 6.269216210799326e-05, + "loss": 0.0294, + "step": 44740 + }, + { + "epoch": 8.687733605164798, + "grad_norm": 0.22546711564064026, + "learning_rate": 6.26761683040086e-05, + "loss": 0.0283, + "step": 44750 + }, + { + "epoch": 8.689675258482598, + "grad_norm": 0.18663613498210907, + "learning_rate": 6.266017311377723e-05, + "loss": 0.0387, + "step": 44760 + }, + { + "epoch": 8.691616911800399, + "grad_norm": 0.16122600436210632, + "learning_rate": 6.264417653904839e-05, + "loss": 0.0245, + "step": 44770 + }, + { + "epoch": 8.693558565118199, + "grad_norm": 0.19092725217342377, + "learning_rate": 6.262817858157139e-05, + "loss": 0.0245, + "step": 44780 + }, + { + "epoch": 8.695500218435999, + "grad_norm": 0.19971127808094025, + "learning_rate": 6.261217924309576e-05, + "loss": 0.0291, + "step": 44790 + }, + { + "epoch": 8.697441871753798, + "grad_norm": 0.17548777163028717, + "learning_rate": 6.259617852537118e-05, + "loss": 0.0299, + "step": 44800 + }, + { + "epoch": 8.699383525071598, + "grad_norm": 0.17922212183475494, + "learning_rate": 6.258017643014747e-05, + "loss": 0.0302, + "step": 44810 + }, + { + "epoch": 8.701325178389398, + "grad_norm": 0.24372705817222595, + "learning_rate": 6.256417295917456e-05, + "loss": 0.0294, + "step": 44820 + }, + { + "epoch": 8.703266831707198, + "grad_norm": 0.1324862688779831, + "learning_rate": 6.254816811420258e-05, + "loss": 0.0327, + "step": 44830 + }, + { + "epoch": 8.705208485024999, + "grad_norm": 0.14080585539340973, + "learning_rate": 6.253216189698183e-05, + "loss": 0.0251, + "step": 44840 + }, + { + "epoch": 8.707150138342799, + "grad_norm": 0.15310613811016083, + "learning_rate": 6.251615430926267e-05, + "loss": 0.0219, + "step": 44850 + }, + { + "epoch": 8.7090917916606, + "grad_norm": 0.16536684334278107, + "learning_rate": 6.25001453527957e-05, + "loss": 0.0234, + "step": 44860 + }, + { + "epoch": 8.7110334449784, + "grad_norm": 0.240324467420578, + "learning_rate": 6.248413502933164e-05, + "loss": 0.0313, + "step": 44870 + }, + { + "epoch": 8.7129750982962, + "grad_norm": 0.20638887584209442, + "learning_rate": 6.246812334062133e-05, + "loss": 0.0262, + "step": 44880 + }, + { + "epoch": 8.714916751614, + "grad_norm": 0.16072200238704681, + "learning_rate": 6.245211028841579e-05, + "loss": 0.0225, + "step": 44890 + }, + { + "epoch": 8.716858404931799, + "grad_norm": 0.15852290391921997, + "learning_rate": 6.24360958744662e-05, + "loss": 0.0338, + "step": 44900 + }, + { + "epoch": 8.718800058249599, + "grad_norm": 0.20794713497161865, + "learning_rate": 6.242008010052387e-05, + "loss": 0.0331, + "step": 44910 + }, + { + "epoch": 8.720741711567399, + "grad_norm": 0.16942644119262695, + "learning_rate": 6.240406296834024e-05, + "loss": 0.0289, + "step": 44920 + }, + { + "epoch": 8.7226833648852, + "grad_norm": 0.24854038655757904, + "learning_rate": 6.238804447966694e-05, + "loss": 0.0305, + "step": 44930 + }, + { + "epoch": 8.724625018203, + "grad_norm": 0.20418858528137207, + "learning_rate": 6.237202463625573e-05, + "loss": 0.0203, + "step": 44940 + }, + { + "epoch": 8.7265666715208, + "grad_norm": 0.1449132263660431, + "learning_rate": 6.235600343985848e-05, + "loss": 0.0239, + "step": 44950 + }, + { + "epoch": 8.7285083248386, + "grad_norm": 0.24199533462524414, + "learning_rate": 6.233998089222729e-05, + "loss": 0.0295, + "step": 44960 + }, + { + "epoch": 8.7304499781564, + "grad_norm": 0.18472658097743988, + "learning_rate": 6.232395699511433e-05, + "loss": 0.0234, + "step": 44970 + }, + { + "epoch": 8.7323916314742, + "grad_norm": 0.1855306774377823, + "learning_rate": 6.230793175027199e-05, + "loss": 0.0225, + "step": 44980 + }, + { + "epoch": 8.734333284792001, + "grad_norm": 0.16345909237861633, + "learning_rate": 6.22919051594527e-05, + "loss": 0.0337, + "step": 44990 + }, + { + "epoch": 8.736274938109801, + "grad_norm": 0.12559708952903748, + "learning_rate": 6.227587722440917e-05, + "loss": 0.0238, + "step": 45000 + }, + { + "epoch": 8.738216591427602, + "grad_norm": 0.18455572426319122, + "learning_rate": 6.225984794689414e-05, + "loss": 0.0264, + "step": 45010 + }, + { + "epoch": 8.7401582447454, + "grad_norm": 0.1206340342760086, + "learning_rate": 6.22438173286606e-05, + "loss": 0.0241, + "step": 45020 + }, + { + "epoch": 8.7420998980632, + "grad_norm": 0.1405380368232727, + "learning_rate": 6.222778537146159e-05, + "loss": 0.0229, + "step": 45030 + }, + { + "epoch": 8.744041551381, + "grad_norm": 0.1282503604888916, + "learning_rate": 6.221175207705037e-05, + "loss": 0.0276, + "step": 45040 + }, + { + "epoch": 8.7459832046988, + "grad_norm": 0.1361822783946991, + "learning_rate": 6.219571744718027e-05, + "loss": 0.0223, + "step": 45050 + }, + { + "epoch": 8.747924858016601, + "grad_norm": 0.18942001461982727, + "learning_rate": 6.217968148360487e-05, + "loss": 0.028, + "step": 45060 + }, + { + "epoch": 8.749866511334401, + "grad_norm": 0.23998065292835236, + "learning_rate": 6.216364418807782e-05, + "loss": 0.026, + "step": 45070 + }, + { + "epoch": 8.751808164652202, + "grad_norm": 0.2894456386566162, + "learning_rate": 6.214760556235292e-05, + "loss": 0.0244, + "step": 45080 + }, + { + "epoch": 8.753749817970002, + "grad_norm": 0.240095853805542, + "learning_rate": 6.213156560818414e-05, + "loss": 0.0285, + "step": 45090 + }, + { + "epoch": 8.755691471287802, + "grad_norm": 0.2848142385482788, + "learning_rate": 6.211552432732559e-05, + "loss": 0.026, + "step": 45100 + }, + { + "epoch": 8.7576331246056, + "grad_norm": 0.1575227826833725, + "learning_rate": 6.20994817215315e-05, + "loss": 0.0235, + "step": 45110 + }, + { + "epoch": 8.759574777923401, + "grad_norm": 0.22637899219989777, + "learning_rate": 6.208343779255628e-05, + "loss": 0.0207, + "step": 45120 + }, + { + "epoch": 8.761516431241201, + "grad_norm": 0.17876599729061127, + "learning_rate": 6.206739254215449e-05, + "loss": 0.0219, + "step": 45130 + }, + { + "epoch": 8.763458084559002, + "grad_norm": 0.1281043440103531, + "learning_rate": 6.205134597208077e-05, + "loss": 0.0225, + "step": 45140 + }, + { + "epoch": 8.765399737876802, + "grad_norm": 0.22113190591335297, + "learning_rate": 6.203529808409e-05, + "loss": 0.0325, + "step": 45150 + }, + { + "epoch": 8.767341391194602, + "grad_norm": 0.2167409062385559, + "learning_rate": 6.201924887993712e-05, + "loss": 0.0248, + "step": 45160 + }, + { + "epoch": 8.769283044512402, + "grad_norm": 0.262026846408844, + "learning_rate": 6.200319836137725e-05, + "loss": 0.0223, + "step": 45170 + }, + { + "epoch": 8.771224697830203, + "grad_norm": 0.23821702599525452, + "learning_rate": 6.198714653016565e-05, + "loss": 0.0262, + "step": 45180 + }, + { + "epoch": 8.773166351148003, + "grad_norm": 0.15758590400218964, + "learning_rate": 6.197109338805774e-05, + "loss": 0.022, + "step": 45190 + }, + { + "epoch": 8.775108004465803, + "grad_norm": 0.2231239527463913, + "learning_rate": 6.195503893680903e-05, + "loss": 0.0259, + "step": 45200 + }, + { + "epoch": 8.777049657783603, + "grad_norm": 0.20518708229064941, + "learning_rate": 6.193898317817524e-05, + "loss": 0.0249, + "step": 45210 + }, + { + "epoch": 8.778991311101404, + "grad_norm": 0.1105138510465622, + "learning_rate": 6.192292611391221e-05, + "loss": 0.0242, + "step": 45220 + }, + { + "epoch": 8.780932964419202, + "grad_norm": 0.2462611198425293, + "learning_rate": 6.190686774577591e-05, + "loss": 0.0292, + "step": 45230 + }, + { + "epoch": 8.782874617737003, + "grad_norm": 0.18370525538921356, + "learning_rate": 6.189080807552245e-05, + "loss": 0.0235, + "step": 45240 + }, + { + "epoch": 8.784816271054803, + "grad_norm": 0.17018021643161774, + "learning_rate": 6.187474710490809e-05, + "loss": 0.0231, + "step": 45250 + }, + { + "epoch": 8.786757924372603, + "grad_norm": 0.18350349366664886, + "learning_rate": 6.185868483568926e-05, + "loss": 0.027, + "step": 45260 + }, + { + "epoch": 8.788699577690403, + "grad_norm": 0.20151518285274506, + "learning_rate": 6.184262126962245e-05, + "loss": 0.0226, + "step": 45270 + }, + { + "epoch": 8.790641231008204, + "grad_norm": 0.14037388563156128, + "learning_rate": 6.182655640846442e-05, + "loss": 0.0306, + "step": 45280 + }, + { + "epoch": 8.792582884326004, + "grad_norm": 0.22930960357189178, + "learning_rate": 6.181049025397196e-05, + "loss": 0.0245, + "step": 45290 + }, + { + "epoch": 8.794524537643804, + "grad_norm": 0.18060548603534698, + "learning_rate": 6.179442280790202e-05, + "loss": 0.023, + "step": 45300 + }, + { + "epoch": 8.796466190961604, + "grad_norm": 0.30307313799858093, + "learning_rate": 6.177835407201174e-05, + "loss": 0.0235, + "step": 45310 + }, + { + "epoch": 8.798407844279405, + "grad_norm": 0.13736361265182495, + "learning_rate": 6.176228404805839e-05, + "loss": 0.0211, + "step": 45320 + }, + { + "epoch": 8.800349497597203, + "grad_norm": 0.16248495876789093, + "learning_rate": 6.174621273779932e-05, + "loss": 0.0342, + "step": 45330 + }, + { + "epoch": 8.802291150915003, + "grad_norm": 0.18090765178203583, + "learning_rate": 6.17301401429921e-05, + "loss": 0.029, + "step": 45340 + }, + { + "epoch": 8.804232804232804, + "grad_norm": 0.14767387509346008, + "learning_rate": 6.171406626539436e-05, + "loss": 0.0225, + "step": 45350 + }, + { + "epoch": 8.806174457550604, + "grad_norm": 0.21346992254257202, + "learning_rate": 6.169799110676398e-05, + "loss": 0.0251, + "step": 45360 + }, + { + "epoch": 8.808116110868404, + "grad_norm": 0.47828012704849243, + "learning_rate": 6.168191466885885e-05, + "loss": 0.0316, + "step": 45370 + }, + { + "epoch": 8.810057764186205, + "grad_norm": 0.08820274472236633, + "learning_rate": 6.166583695343711e-05, + "loss": 0.0278, + "step": 45380 + }, + { + "epoch": 8.811999417504005, + "grad_norm": 0.18368060886859894, + "learning_rate": 6.164975796225698e-05, + "loss": 0.0309, + "step": 45390 + }, + { + "epoch": 8.813941070821805, + "grad_norm": 0.24209272861480713, + "learning_rate": 6.163367769707683e-05, + "loss": 0.0306, + "step": 45400 + }, + { + "epoch": 8.815882724139605, + "grad_norm": 0.13379891216754913, + "learning_rate": 6.161759615965519e-05, + "loss": 0.0275, + "step": 45410 + }, + { + "epoch": 8.817824377457406, + "grad_norm": 0.19732451438903809, + "learning_rate": 6.16015133517507e-05, + "loss": 0.0271, + "step": 45420 + }, + { + "epoch": 8.819766030775206, + "grad_norm": 0.14049877226352692, + "learning_rate": 6.158542927512214e-05, + "loss": 0.0277, + "step": 45430 + }, + { + "epoch": 8.821707684093004, + "grad_norm": 0.28834497928619385, + "learning_rate": 6.156934393152846e-05, + "loss": 0.0316, + "step": 45440 + }, + { + "epoch": 8.823649337410805, + "grad_norm": 0.10604466497898102, + "learning_rate": 6.15532573227287e-05, + "loss": 0.0296, + "step": 45450 + }, + { + "epoch": 8.825590990728605, + "grad_norm": 0.22275042533874512, + "learning_rate": 6.153716945048212e-05, + "loss": 0.0315, + "step": 45460 + }, + { + "epoch": 8.827532644046405, + "grad_norm": 0.1592061072587967, + "learning_rate": 6.152108031654802e-05, + "loss": 0.0258, + "step": 45470 + }, + { + "epoch": 8.829474297364206, + "grad_norm": 0.13987129926681519, + "learning_rate": 6.15049899226859e-05, + "loss": 0.0288, + "step": 45480 + }, + { + "epoch": 8.831415950682006, + "grad_norm": 0.20056912302970886, + "learning_rate": 6.148889827065537e-05, + "loss": 0.0277, + "step": 45490 + }, + { + "epoch": 8.833357603999806, + "grad_norm": 0.15303626656532288, + "learning_rate": 6.147280536221622e-05, + "loss": 0.0277, + "step": 45500 + }, + { + "epoch": 8.835299257317606, + "grad_norm": 0.1800311952829361, + "learning_rate": 6.145671119912832e-05, + "loss": 0.0305, + "step": 45510 + }, + { + "epoch": 8.837240910635407, + "grad_norm": 0.13011904060840607, + "learning_rate": 6.144061578315169e-05, + "loss": 0.025, + "step": 45520 + }, + { + "epoch": 8.839182563953207, + "grad_norm": 0.23101796209812164, + "learning_rate": 6.142451911604654e-05, + "loss": 0.0306, + "step": 45530 + }, + { + "epoch": 8.841124217271005, + "grad_norm": 0.15900790691375732, + "learning_rate": 6.140842119957315e-05, + "loss": 0.0231, + "step": 45540 + }, + { + "epoch": 8.843065870588806, + "grad_norm": 0.18207624554634094, + "learning_rate": 6.139232203549197e-05, + "loss": 0.0259, + "step": 45550 + }, + { + "epoch": 8.845007523906606, + "grad_norm": 0.21022260189056396, + "learning_rate": 6.137622162556357e-05, + "loss": 0.0231, + "step": 45560 + }, + { + "epoch": 8.846949177224406, + "grad_norm": 0.20820525288581848, + "learning_rate": 6.13601199715487e-05, + "loss": 0.0302, + "step": 45570 + }, + { + "epoch": 8.848890830542206, + "grad_norm": 0.07218022644519806, + "learning_rate": 6.134401707520816e-05, + "loss": 0.0323, + "step": 45580 + }, + { + "epoch": 8.850832483860007, + "grad_norm": 0.13682031631469727, + "learning_rate": 6.132791293830298e-05, + "loss": 0.0221, + "step": 45590 + }, + { + "epoch": 8.852774137177807, + "grad_norm": 0.2364613115787506, + "learning_rate": 6.131180756259428e-05, + "loss": 0.0306, + "step": 45600 + }, + { + "epoch": 8.854715790495607, + "grad_norm": 0.1807117611169815, + "learning_rate": 6.129570094984331e-05, + "loss": 0.0282, + "step": 45610 + }, + { + "epoch": 8.856657443813408, + "grad_norm": 0.17547592520713806, + "learning_rate": 6.127959310181145e-05, + "loss": 0.0236, + "step": 45620 + }, + { + "epoch": 8.858599097131208, + "grad_norm": 0.18282340466976166, + "learning_rate": 6.126348402026026e-05, + "loss": 0.0286, + "step": 45630 + }, + { + "epoch": 8.860540750449008, + "grad_norm": 0.22920741140842438, + "learning_rate": 6.12473737069514e-05, + "loss": 0.0268, + "step": 45640 + }, + { + "epoch": 8.862482403766807, + "grad_norm": 0.17765145003795624, + "learning_rate": 6.123126216364665e-05, + "loss": 0.0241, + "step": 45650 + }, + { + "epoch": 8.864424057084607, + "grad_norm": 0.1853027641773224, + "learning_rate": 6.121514939210797e-05, + "loss": 0.022, + "step": 45660 + }, + { + "epoch": 8.866365710402407, + "grad_norm": 0.18548625707626343, + "learning_rate": 6.119903539409741e-05, + "loss": 0.026, + "step": 45670 + }, + { + "epoch": 8.868307363720207, + "grad_norm": 0.28616008162498474, + "learning_rate": 6.118292017137716e-05, + "loss": 0.0263, + "step": 45680 + }, + { + "epoch": 8.870249017038008, + "grad_norm": 0.2522358000278473, + "learning_rate": 6.116680372570959e-05, + "loss": 0.0231, + "step": 45690 + }, + { + "epoch": 8.872190670355808, + "grad_norm": 0.29727306962013245, + "learning_rate": 6.115068605885713e-05, + "loss": 0.0291, + "step": 45700 + }, + { + "epoch": 8.874132323673608, + "grad_norm": 0.2351083904504776, + "learning_rate": 6.113456717258243e-05, + "loss": 0.0238, + "step": 45710 + }, + { + "epoch": 8.876073976991409, + "grad_norm": 0.2521604299545288, + "learning_rate": 6.11184470686482e-05, + "loss": 0.0318, + "step": 45720 + }, + { + "epoch": 8.878015630309209, + "grad_norm": 0.17012418806552887, + "learning_rate": 6.110232574881731e-05, + "loss": 0.0275, + "step": 45730 + }, + { + "epoch": 8.879957283627009, + "grad_norm": 0.1430356800556183, + "learning_rate": 6.108620321485277e-05, + "loss": 0.0264, + "step": 45740 + }, + { + "epoch": 8.881898936944808, + "grad_norm": 0.12883256375789642, + "learning_rate": 6.107007946851773e-05, + "loss": 0.0256, + "step": 45750 + }, + { + "epoch": 8.883840590262608, + "grad_norm": 0.23455704748630524, + "learning_rate": 6.105395451157542e-05, + "loss": 0.0251, + "step": 45760 + }, + { + "epoch": 8.885782243580408, + "grad_norm": 0.10737639665603638, + "learning_rate": 6.103782834578928e-05, + "loss": 0.0236, + "step": 45770 + }, + { + "epoch": 8.887723896898208, + "grad_norm": 0.18056099116802216, + "learning_rate": 6.102170097292281e-05, + "loss": 0.0267, + "step": 45780 + }, + { + "epoch": 8.889665550216009, + "grad_norm": 0.11901843547821045, + "learning_rate": 6.1005572394739716e-05, + "loss": 0.0279, + "step": 45790 + }, + { + "epoch": 8.891607203533809, + "grad_norm": 0.14483773708343506, + "learning_rate": 6.0989442613003765e-05, + "loss": 0.0232, + "step": 45800 + }, + { + "epoch": 8.89354885685161, + "grad_norm": 0.2284281700849533, + "learning_rate": 6.097331162947888e-05, + "loss": 0.0294, + "step": 45810 + }, + { + "epoch": 8.89549051016941, + "grad_norm": 0.16272257268428802, + "learning_rate": 6.095717944592914e-05, + "loss": 0.0233, + "step": 45820 + }, + { + "epoch": 8.89743216348721, + "grad_norm": 0.13676097989082336, + "learning_rate": 6.094104606411873e-05, + "loss": 0.0322, + "step": 45830 + }, + { + "epoch": 8.89937381680501, + "grad_norm": 0.08985556662082672, + "learning_rate": 6.0924911485811966e-05, + "loss": 0.0288, + "step": 45840 + }, + { + "epoch": 8.90131547012281, + "grad_norm": 0.1592593789100647, + "learning_rate": 6.090877571277331e-05, + "loss": 0.0279, + "step": 45850 + }, + { + "epoch": 8.90325712344061, + "grad_norm": 0.13729064166545868, + "learning_rate": 6.089263874676736e-05, + "loss": 0.0256, + "step": 45860 + }, + { + "epoch": 8.905198776758409, + "grad_norm": 0.1287968009710312, + "learning_rate": 6.0876500589558796e-05, + "loss": 0.0233, + "step": 45870 + }, + { + "epoch": 8.90714043007621, + "grad_norm": 0.10805711895227432, + "learning_rate": 6.086036124291248e-05, + "loss": 0.0249, + "step": 45880 + }, + { + "epoch": 8.90908208339401, + "grad_norm": 0.12583498656749725, + "learning_rate": 6.084422070859339e-05, + "loss": 0.0229, + "step": 45890 + }, + { + "epoch": 8.91102373671181, + "grad_norm": 0.19096782803535461, + "learning_rate": 6.082807898836663e-05, + "loss": 0.0279, + "step": 45900 + }, + { + "epoch": 8.91296539002961, + "grad_norm": 0.17573517560958862, + "learning_rate": 6.081193608399742e-05, + "loss": 0.0227, + "step": 45910 + }, + { + "epoch": 8.91490704334741, + "grad_norm": 0.16635973751544952, + "learning_rate": 6.0795791997251164e-05, + "loss": 0.0277, + "step": 45920 + }, + { + "epoch": 8.91684869666521, + "grad_norm": 0.1372900903224945, + "learning_rate": 6.0779646729893294e-05, + "loss": 0.0227, + "step": 45930 + }, + { + "epoch": 8.918790349983011, + "grad_norm": 0.22105567157268524, + "learning_rate": 6.0763500283689476e-05, + "loss": 0.0239, + "step": 45940 + }, + { + "epoch": 8.920732003300811, + "grad_norm": 0.20382070541381836, + "learning_rate": 6.0747352660405455e-05, + "loss": 0.0277, + "step": 45950 + }, + { + "epoch": 8.922673656618612, + "grad_norm": 0.23524972796440125, + "learning_rate": 6.073120386180709e-05, + "loss": 0.0265, + "step": 45960 + }, + { + "epoch": 8.92461530993641, + "grad_norm": 0.1786951720714569, + "learning_rate": 6.0715053889660425e-05, + "loss": 0.0289, + "step": 45970 + }, + { + "epoch": 8.92655696325421, + "grad_norm": 0.1146051287651062, + "learning_rate": 6.069890274573157e-05, + "loss": 0.0247, + "step": 45980 + }, + { + "epoch": 8.92849861657201, + "grad_norm": 0.08767542243003845, + "learning_rate": 6.068275043178679e-05, + "loss": 0.0228, + "step": 45990 + }, + { + "epoch": 8.93044026988981, + "grad_norm": 0.21066786348819733, + "learning_rate": 6.066659694959248e-05, + "loss": 0.028, + "step": 46000 + }, + { + "epoch": 8.932381923207611, + "grad_norm": 0.10529998689889908, + "learning_rate": 6.0650442300915176e-05, + "loss": 0.0252, + "step": 46010 + }, + { + "epoch": 8.934323576525411, + "grad_norm": 0.2934766709804535, + "learning_rate": 6.063428648752152e-05, + "loss": 0.0333, + "step": 46020 + }, + { + "epoch": 8.936265229843212, + "grad_norm": 0.33198073506355286, + "learning_rate": 6.0618129511178266e-05, + "loss": 0.0251, + "step": 46030 + }, + { + "epoch": 8.938206883161012, + "grad_norm": 0.22203931212425232, + "learning_rate": 6.060197137365234e-05, + "loss": 0.0256, + "step": 46040 + }, + { + "epoch": 8.940148536478812, + "grad_norm": 0.13841181993484497, + "learning_rate": 6.058581207671077e-05, + "loss": 0.0242, + "step": 46050 + }, + { + "epoch": 8.942090189796613, + "grad_norm": 0.17912234365940094, + "learning_rate": 6.056965162212072e-05, + "loss": 0.0292, + "step": 46060 + }, + { + "epoch": 8.944031843114413, + "grad_norm": 0.15959963202476501, + "learning_rate": 6.055349001164943e-05, + "loss": 0.0227, + "step": 46070 + }, + { + "epoch": 8.945973496432211, + "grad_norm": 0.22635868191719055, + "learning_rate": 6.0537327247064347e-05, + "loss": 0.0291, + "step": 46080 + }, + { + "epoch": 8.947915149750012, + "grad_norm": 0.14387406408786774, + "learning_rate": 6.0521163330133014e-05, + "loss": 0.0285, + "step": 46090 + }, + { + "epoch": 8.949856803067812, + "grad_norm": 0.23154620826244354, + "learning_rate": 6.050499826262306e-05, + "loss": 0.032, + "step": 46100 + }, + { + "epoch": 8.951798456385612, + "grad_norm": 0.12293318659067154, + "learning_rate": 6.0488832046302294e-05, + "loss": 0.0389, + "step": 46110 + }, + { + "epoch": 8.953740109703412, + "grad_norm": 0.16578762233257294, + "learning_rate": 6.0472664682938626e-05, + "loss": 0.0259, + "step": 46120 + }, + { + "epoch": 8.955681763021213, + "grad_norm": 0.18746918439865112, + "learning_rate": 6.045649617430009e-05, + "loss": 0.035, + "step": 46130 + }, + { + "epoch": 8.957623416339013, + "grad_norm": 0.1329531967639923, + "learning_rate": 6.0440326522154866e-05, + "loss": 0.0248, + "step": 46140 + }, + { + "epoch": 8.959565069656813, + "grad_norm": 0.21299688518047333, + "learning_rate": 6.0424155728271224e-05, + "loss": 0.0328, + "step": 46150 + }, + { + "epoch": 8.961506722974613, + "grad_norm": 0.18330907821655273, + "learning_rate": 6.040798379441758e-05, + "loss": 0.0242, + "step": 46160 + }, + { + "epoch": 8.963448376292414, + "grad_norm": 0.17915241420269012, + "learning_rate": 6.0391810722362485e-05, + "loss": 0.0266, + "step": 46170 + }, + { + "epoch": 8.965390029610212, + "grad_norm": 0.233330637216568, + "learning_rate": 6.037563651387458e-05, + "loss": 0.0261, + "step": 46180 + }, + { + "epoch": 8.967331682928013, + "grad_norm": 0.16662882268428802, + "learning_rate": 6.0359461170722666e-05, + "loss": 0.0299, + "step": 46190 + }, + { + "epoch": 8.969273336245813, + "grad_norm": 0.17164799571037292, + "learning_rate": 6.034328469467566e-05, + "loss": 0.0246, + "step": 46200 + }, + { + "epoch": 8.971214989563613, + "grad_norm": 0.19551071524620056, + "learning_rate": 6.0327107087502596e-05, + "loss": 0.0344, + "step": 46210 + }, + { + "epoch": 8.973156642881413, + "grad_norm": 0.2173694521188736, + "learning_rate": 6.031092835097262e-05, + "loss": 0.0251, + "step": 46220 + }, + { + "epoch": 8.975098296199214, + "grad_norm": 0.1984950751066208, + "learning_rate": 6.0294748486855024e-05, + "loss": 0.0299, + "step": 46230 + }, + { + "epoch": 8.977039949517014, + "grad_norm": 0.13630902767181396, + "learning_rate": 6.0278567496919216e-05, + "loss": 0.0266, + "step": 46240 + }, + { + "epoch": 8.978981602834814, + "grad_norm": 0.16445158421993256, + "learning_rate": 6.026238538293472e-05, + "loss": 0.027, + "step": 46250 + }, + { + "epoch": 8.980923256152614, + "grad_norm": 0.28139209747314453, + "learning_rate": 6.024620214667118e-05, + "loss": 0.0237, + "step": 46260 + }, + { + "epoch": 8.982864909470415, + "grad_norm": 0.12562023103237152, + "learning_rate": 6.0230017789898384e-05, + "loss": 0.0287, + "step": 46270 + }, + { + "epoch": 8.984806562788215, + "grad_norm": 0.24017930030822754, + "learning_rate": 6.021383231438622e-05, + "loss": 0.0252, + "step": 46280 + }, + { + "epoch": 8.986748216106013, + "grad_norm": 0.1120583787560463, + "learning_rate": 6.0197645721904704e-05, + "loss": 0.0309, + "step": 46290 + }, + { + "epoch": 8.988689869423814, + "grad_norm": 0.1851133406162262, + "learning_rate": 6.0181458014224e-05, + "loss": 0.0322, + "step": 46300 + }, + { + "epoch": 8.990631522741614, + "grad_norm": 0.15784485638141632, + "learning_rate": 6.016526919311435e-05, + "loss": 0.0304, + "step": 46310 + }, + { + "epoch": 8.992573176059414, + "grad_norm": 0.19302517175674438, + "learning_rate": 6.014907926034613e-05, + "loss": 0.0323, + "step": 46320 + }, + { + "epoch": 8.994514829377215, + "grad_norm": 0.12110082060098648, + "learning_rate": 6.0132888217689875e-05, + "loss": 0.0264, + "step": 46330 + }, + { + "epoch": 8.996456482695015, + "grad_norm": 0.18065938353538513, + "learning_rate": 6.01166960669162e-05, + "loss": 0.0238, + "step": 46340 + }, + { + "epoch": 8.998398136012815, + "grad_norm": 0.17107336223125458, + "learning_rate": 6.0100502809795845e-05, + "loss": 0.0227, + "step": 46350 + }, + { + "epoch": 9.00019416533178, + "grad_norm": 0.1270778626203537, + "learning_rate": 6.00843084480997e-05, + "loss": 0.0217, + "step": 46360 + }, + { + "epoch": 9.00213581864958, + "grad_norm": 0.19826799631118774, + "learning_rate": 6.0068112983598736e-05, + "loss": 0.0295, + "step": 46370 + }, + { + "epoch": 9.00407747196738, + "grad_norm": 0.15275351703166962, + "learning_rate": 6.005191641806407e-05, + "loss": 0.0252, + "step": 46380 + }, + { + "epoch": 9.00601912528518, + "grad_norm": 0.16445675492286682, + "learning_rate": 6.003571875326694e-05, + "loss": 0.0175, + "step": 46390 + }, + { + "epoch": 9.00796077860298, + "grad_norm": 0.20565173029899597, + "learning_rate": 6.00195199909787e-05, + "loss": 0.0316, + "step": 46400 + }, + { + "epoch": 9.00990243192078, + "grad_norm": 0.12811632454395294, + "learning_rate": 6.000332013297082e-05, + "loss": 0.0226, + "step": 46410 + }, + { + "epoch": 9.01184408523858, + "grad_norm": 0.13405245542526245, + "learning_rate": 5.998711918101487e-05, + "loss": 0.0365, + "step": 46420 + }, + { + "epoch": 9.01378573855638, + "grad_norm": 0.15119051933288574, + "learning_rate": 5.997091713688261e-05, + "loss": 0.0309, + "step": 46430 + }, + { + "epoch": 9.015727391874181, + "grad_norm": 0.1392679065465927, + "learning_rate": 5.9954714002345836e-05, + "loss": 0.0276, + "step": 46440 + }, + { + "epoch": 9.017669045191981, + "grad_norm": 0.17640738189220428, + "learning_rate": 5.993850977917649e-05, + "loss": 0.0289, + "step": 46450 + }, + { + "epoch": 9.019610698509782, + "grad_norm": 0.1286066472530365, + "learning_rate": 5.992230446914667e-05, + "loss": 0.0273, + "step": 46460 + }, + { + "epoch": 9.021552351827582, + "grad_norm": 0.2244495004415512, + "learning_rate": 5.990609807402855e-05, + "loss": 0.0351, + "step": 46470 + }, + { + "epoch": 9.02349400514538, + "grad_norm": 0.2391398698091507, + "learning_rate": 5.988989059559443e-05, + "loss": 0.0212, + "step": 46480 + }, + { + "epoch": 9.02543565846318, + "grad_norm": 0.19524843990802765, + "learning_rate": 5.987368203561675e-05, + "loss": 0.023, + "step": 46490 + }, + { + "epoch": 9.027377311780981, + "grad_norm": 0.22225531935691833, + "learning_rate": 5.9857472395868055e-05, + "loss": 0.0339, + "step": 46500 + }, + { + "epoch": 9.029318965098781, + "grad_norm": 0.18180209398269653, + "learning_rate": 5.9841261678120983e-05, + "loss": 0.0343, + "step": 46510 + }, + { + "epoch": 9.031260618416582, + "grad_norm": 0.2514437437057495, + "learning_rate": 5.982504988414834e-05, + "loss": 0.0252, + "step": 46520 + }, + { + "epoch": 9.033202271734382, + "grad_norm": 0.34068629145622253, + "learning_rate": 5.9808837015723015e-05, + "loss": 0.0409, + "step": 46530 + }, + { + "epoch": 9.035143925052182, + "grad_norm": 0.16028140485286713, + "learning_rate": 5.9792623074618016e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 9.037085578369982, + "grad_norm": 0.15825478732585907, + "learning_rate": 5.977640806260648e-05, + "loss": 0.0263, + "step": 46550 + }, + { + "epoch": 9.039027231687783, + "grad_norm": 0.09346158802509308, + "learning_rate": 5.976019198146166e-05, + "loss": 0.0289, + "step": 46560 + }, + { + "epoch": 9.040968885005583, + "grad_norm": 0.2136273831129074, + "learning_rate": 5.974397483295692e-05, + "loss": 0.0313, + "step": 46570 + }, + { + "epoch": 9.042910538323383, + "grad_norm": 0.12170963734388351, + "learning_rate": 5.972775661886575e-05, + "loss": 0.0206, + "step": 46580 + }, + { + "epoch": 9.044852191641182, + "grad_norm": 0.19601742923259735, + "learning_rate": 5.971153734096173e-05, + "loss": 0.0246, + "step": 46590 + }, + { + "epoch": 9.046793844958982, + "grad_norm": 0.2613515853881836, + "learning_rate": 5.969531700101859e-05, + "loss": 0.032, + "step": 46600 + }, + { + "epoch": 9.048735498276782, + "grad_norm": 0.1019047200679779, + "learning_rate": 5.9679095600810155e-05, + "loss": 0.0222, + "step": 46610 + }, + { + "epoch": 9.050677151594583, + "grad_norm": 0.10923628509044647, + "learning_rate": 5.9662873142110384e-05, + "loss": 0.0281, + "step": 46620 + }, + { + "epoch": 9.052618804912383, + "grad_norm": 0.30609554052352905, + "learning_rate": 5.964664962669333e-05, + "loss": 0.0252, + "step": 46630 + }, + { + "epoch": 9.054560458230183, + "grad_norm": 0.1482456624507904, + "learning_rate": 5.9630425056333186e-05, + "loss": 0.0296, + "step": 46640 + }, + { + "epoch": 9.056502111547983, + "grad_norm": 0.21073728799819946, + "learning_rate": 5.961419943280422e-05, + "loss": 0.0257, + "step": 46650 + }, + { + "epoch": 9.058443764865784, + "grad_norm": 0.18408626317977905, + "learning_rate": 5.959797275788087e-05, + "loss": 0.0276, + "step": 46660 + }, + { + "epoch": 9.060385418183584, + "grad_norm": 0.12127804011106491, + "learning_rate": 5.958174503333765e-05, + "loss": 0.0221, + "step": 46670 + }, + { + "epoch": 9.062327071501384, + "grad_norm": 0.22617605328559875, + "learning_rate": 5.9565516260949195e-05, + "loss": 0.0252, + "step": 46680 + }, + { + "epoch": 9.064268724819183, + "grad_norm": 0.1309843212366104, + "learning_rate": 5.9549286442490273e-05, + "loss": 0.0221, + "step": 46690 + }, + { + "epoch": 9.066210378136983, + "grad_norm": 0.13982398808002472, + "learning_rate": 5.953305557973572e-05, + "loss": 0.0211, + "step": 46700 + }, + { + "epoch": 9.068152031454783, + "grad_norm": 0.19910961389541626, + "learning_rate": 5.9516823674460564e-05, + "loss": 0.0288, + "step": 46710 + }, + { + "epoch": 9.070093684772583, + "grad_norm": 0.16726046800613403, + "learning_rate": 5.9500590728439875e-05, + "loss": 0.0222, + "step": 46720 + }, + { + "epoch": 9.072035338090384, + "grad_norm": 0.12974244356155396, + "learning_rate": 5.948435674344886e-05, + "loss": 0.0216, + "step": 46730 + }, + { + "epoch": 9.073976991408184, + "grad_norm": 0.1506296545267105, + "learning_rate": 5.946812172126285e-05, + "loss": 0.0271, + "step": 46740 + }, + { + "epoch": 9.075918644725984, + "grad_norm": 0.13079601526260376, + "learning_rate": 5.945188566365729e-05, + "loss": 0.0227, + "step": 46750 + }, + { + "epoch": 9.077860298043785, + "grad_norm": 0.14677481353282928, + "learning_rate": 5.943564857240773e-05, + "loss": 0.0287, + "step": 46760 + }, + { + "epoch": 9.079801951361585, + "grad_norm": 0.18370698392391205, + "learning_rate": 5.941941044928981e-05, + "loss": 0.0296, + "step": 46770 + }, + { + "epoch": 9.081743604679385, + "grad_norm": 0.09985506534576416, + "learning_rate": 5.940317129607935e-05, + "loss": 0.0237, + "step": 46780 + }, + { + "epoch": 9.083685257997185, + "grad_norm": 0.30443814396858215, + "learning_rate": 5.9386931114552204e-05, + "loss": 0.0373, + "step": 46790 + }, + { + "epoch": 9.085626911314984, + "grad_norm": 0.1692449301481247, + "learning_rate": 5.937068990648438e-05, + "loss": 0.0246, + "step": 46800 + }, + { + "epoch": 9.087568564632784, + "grad_norm": 0.29017871618270874, + "learning_rate": 5.935444767365199e-05, + "loss": 0.0259, + "step": 46810 + }, + { + "epoch": 9.089510217950584, + "grad_norm": 0.20591461658477783, + "learning_rate": 5.933820441783129e-05, + "loss": 0.0347, + "step": 46820 + }, + { + "epoch": 9.091451871268385, + "grad_norm": 0.1368168443441391, + "learning_rate": 5.932196014079857e-05, + "loss": 0.0221, + "step": 46830 + }, + { + "epoch": 9.093393524586185, + "grad_norm": 0.1324397176504135, + "learning_rate": 5.930571484433032e-05, + "loss": 0.0272, + "step": 46840 + }, + { + "epoch": 9.095335177903985, + "grad_norm": 0.2436484694480896, + "learning_rate": 5.928946853020309e-05, + "loss": 0.0318, + "step": 46850 + }, + { + "epoch": 9.097276831221786, + "grad_norm": 0.14009258151054382, + "learning_rate": 5.927322120019352e-05, + "loss": 0.0265, + "step": 46860 + }, + { + "epoch": 9.099218484539586, + "grad_norm": 0.1651400327682495, + "learning_rate": 5.925697285607844e-05, + "loss": 0.0232, + "step": 46870 + }, + { + "epoch": 9.101160137857386, + "grad_norm": 0.19855310022830963, + "learning_rate": 5.924072349963472e-05, + "loss": 0.0287, + "step": 46880 + }, + { + "epoch": 9.103101791175186, + "grad_norm": 0.22046591341495514, + "learning_rate": 5.922447313263938e-05, + "loss": 0.025, + "step": 46890 + }, + { + "epoch": 9.105043444492987, + "grad_norm": 0.18067654967308044, + "learning_rate": 5.920822175686952e-05, + "loss": 0.0234, + "step": 46900 + }, + { + "epoch": 9.106985097810785, + "grad_norm": 0.13870133459568024, + "learning_rate": 5.919196937410239e-05, + "loss": 0.0233, + "step": 46910 + }, + { + "epoch": 9.108926751128585, + "grad_norm": 0.161329984664917, + "learning_rate": 5.917571598611529e-05, + "loss": 0.0297, + "step": 46920 + }, + { + "epoch": 9.110868404446386, + "grad_norm": 0.13587330281734467, + "learning_rate": 5.9159461594685686e-05, + "loss": 0.027, + "step": 46930 + }, + { + "epoch": 9.112810057764186, + "grad_norm": 0.1383013129234314, + "learning_rate": 5.914320620159114e-05, + "loss": 0.0244, + "step": 46940 + }, + { + "epoch": 9.114751711081986, + "grad_norm": 0.14059409499168396, + "learning_rate": 5.912694980860931e-05, + "loss": 0.0268, + "step": 46950 + }, + { + "epoch": 9.116693364399787, + "grad_norm": 0.1496448665857315, + "learning_rate": 5.9110692417517964e-05, + "loss": 0.0239, + "step": 46960 + }, + { + "epoch": 9.118635017717587, + "grad_norm": 0.18320026993751526, + "learning_rate": 5.9094434030095e-05, + "loss": 0.0317, + "step": 46970 + }, + { + "epoch": 9.120576671035387, + "grad_norm": 0.19455783069133759, + "learning_rate": 5.907817464811841e-05, + "loss": 0.0235, + "step": 46980 + }, + { + "epoch": 9.122518324353187, + "grad_norm": 0.2458461970090866, + "learning_rate": 5.906191427336627e-05, + "loss": 0.0306, + "step": 46990 + }, + { + "epoch": 9.124459977670988, + "grad_norm": 0.13085198402404785, + "learning_rate": 5.9045652907616835e-05, + "loss": 0.027, + "step": 47000 + }, + { + "epoch": 9.126401630988786, + "grad_norm": 0.10007555037736893, + "learning_rate": 5.902939055264838e-05, + "loss": 0.0274, + "step": 47010 + }, + { + "epoch": 9.128343284306586, + "grad_norm": 0.12384022772312164, + "learning_rate": 5.901312721023935e-05, + "loss": 0.0268, + "step": 47020 + }, + { + "epoch": 9.130284937624387, + "grad_norm": 0.0929393619298935, + "learning_rate": 5.8996862882168294e-05, + "loss": 0.0243, + "step": 47030 + }, + { + "epoch": 9.132226590942187, + "grad_norm": 0.14373551309108734, + "learning_rate": 5.8980597570213826e-05, + "loss": 0.0268, + "step": 47040 + }, + { + "epoch": 9.134168244259987, + "grad_norm": 0.10234443098306656, + "learning_rate": 5.896433127615471e-05, + "loss": 0.0275, + "step": 47050 + }, + { + "epoch": 9.136109897577787, + "grad_norm": 0.18533697724342346, + "learning_rate": 5.894806400176981e-05, + "loss": 0.0272, + "step": 47060 + }, + { + "epoch": 9.138051550895588, + "grad_norm": 0.2741023600101471, + "learning_rate": 5.893179574883808e-05, + "loss": 0.0255, + "step": 47070 + }, + { + "epoch": 9.139993204213388, + "grad_norm": 0.18932707607746124, + "learning_rate": 5.8915526519138585e-05, + "loss": 0.0286, + "step": 47080 + }, + { + "epoch": 9.141934857531188, + "grad_norm": 0.35022979974746704, + "learning_rate": 5.889925631445053e-05, + "loss": 0.0265, + "step": 47090 + }, + { + "epoch": 9.143876510848989, + "grad_norm": 0.15964283049106598, + "learning_rate": 5.888298513655318e-05, + "loss": 0.0214, + "step": 47100 + }, + { + "epoch": 9.145818164166789, + "grad_norm": 0.19344750046730042, + "learning_rate": 5.886671298722592e-05, + "loss": 0.0225, + "step": 47110 + }, + { + "epoch": 9.147759817484587, + "grad_norm": 0.19802838563919067, + "learning_rate": 5.885043986824828e-05, + "loss": 0.0249, + "step": 47120 + }, + { + "epoch": 9.149701470802388, + "grad_norm": 0.8707701563835144, + "learning_rate": 5.883416578139982e-05, + "loss": 0.0229, + "step": 47130 + }, + { + "epoch": 9.151643124120188, + "grad_norm": 0.18017920851707458, + "learning_rate": 5.881789072846029e-05, + "loss": 0.0294, + "step": 47140 + }, + { + "epoch": 9.153584777437988, + "grad_norm": 0.20825079083442688, + "learning_rate": 5.8801614711209474e-05, + "loss": 0.0239, + "step": 47150 + }, + { + "epoch": 9.155526430755788, + "grad_norm": 0.1659684181213379, + "learning_rate": 5.878533773142732e-05, + "loss": 0.0284, + "step": 47160 + }, + { + "epoch": 9.157468084073589, + "grad_norm": 0.18840765953063965, + "learning_rate": 5.8769059790893834e-05, + "loss": 0.0276, + "step": 47170 + }, + { + "epoch": 9.159409737391389, + "grad_norm": 0.14845065772533417, + "learning_rate": 5.8752780891389145e-05, + "loss": 0.0206, + "step": 47180 + }, + { + "epoch": 9.16135139070919, + "grad_norm": 0.33954837918281555, + "learning_rate": 5.873650103469351e-05, + "loss": 0.038, + "step": 47190 + }, + { + "epoch": 9.16329304402699, + "grad_norm": 0.1287592202425003, + "learning_rate": 5.872022022258725e-05, + "loss": 0.0312, + "step": 47200 + }, + { + "epoch": 9.16523469734479, + "grad_norm": 0.2039954662322998, + "learning_rate": 5.8703938456850814e-05, + "loss": 0.0303, + "step": 47210 + }, + { + "epoch": 9.16717635066259, + "grad_norm": 0.17379368841648102, + "learning_rate": 5.868765573926477e-05, + "loss": 0.0241, + "step": 47220 + }, + { + "epoch": 9.169118003980389, + "grad_norm": 0.2405364215373993, + "learning_rate": 5.867137207160974e-05, + "loss": 0.0324, + "step": 47230 + }, + { + "epoch": 9.171059657298189, + "grad_norm": 0.1762675791978836, + "learning_rate": 5.86550874556665e-05, + "loss": 0.027, + "step": 47240 + }, + { + "epoch": 9.17300131061599, + "grad_norm": 0.14324083924293518, + "learning_rate": 5.863880189321592e-05, + "loss": 0.0252, + "step": 47250 + }, + { + "epoch": 9.17494296393379, + "grad_norm": 0.11847824603319168, + "learning_rate": 5.862251538603894e-05, + "loss": 0.0271, + "step": 47260 + }, + { + "epoch": 9.17688461725159, + "grad_norm": 0.06011451035737991, + "learning_rate": 5.8606227935916656e-05, + "loss": 0.0238, + "step": 47270 + }, + { + "epoch": 9.17882627056939, + "grad_norm": 0.16802357137203217, + "learning_rate": 5.858993954463021e-05, + "loss": 0.0219, + "step": 47280 + }, + { + "epoch": 9.18076792388719, + "grad_norm": 0.16382233798503876, + "learning_rate": 5.8573650213960896e-05, + "loss": 0.0377, + "step": 47290 + }, + { + "epoch": 9.18270957720499, + "grad_norm": 0.1624467372894287, + "learning_rate": 5.8557359945690084e-05, + "loss": 0.0247, + "step": 47300 + }, + { + "epoch": 9.18465123052279, + "grad_norm": 0.13915576040744781, + "learning_rate": 5.8541068741599246e-05, + "loss": 0.0271, + "step": 47310 + }, + { + "epoch": 9.186592883840591, + "grad_norm": 0.34427085518836975, + "learning_rate": 5.8524776603469985e-05, + "loss": 0.0304, + "step": 47320 + }, + { + "epoch": 9.18853453715839, + "grad_norm": 0.10195261985063553, + "learning_rate": 5.850848353308397e-05, + "loss": 0.0255, + "step": 47330 + }, + { + "epoch": 9.19047619047619, + "grad_norm": 0.22677424550056458, + "learning_rate": 5.849218953222297e-05, + "loss": 0.0262, + "step": 47340 + }, + { + "epoch": 9.19241784379399, + "grad_norm": 0.11036282777786255, + "learning_rate": 5.847589460266891e-05, + "loss": 0.0281, + "step": 47350 + }, + { + "epoch": 9.19435949711179, + "grad_norm": 0.22685852646827698, + "learning_rate": 5.845959874620375e-05, + "loss": 0.0291, + "step": 47360 + }, + { + "epoch": 9.19630115042959, + "grad_norm": 0.15024615824222565, + "learning_rate": 5.84433019646096e-05, + "loss": 0.0342, + "step": 47370 + }, + { + "epoch": 9.198242803747391, + "grad_norm": 0.20524591207504272, + "learning_rate": 5.842700425966863e-05, + "loss": 0.0292, + "step": 47380 + }, + { + "epoch": 9.200184457065191, + "grad_norm": 0.16774386167526245, + "learning_rate": 5.841070563316315e-05, + "loss": 0.0281, + "step": 47390 + }, + { + "epoch": 9.202126110382991, + "grad_norm": 0.16951650381088257, + "learning_rate": 5.839440608687554e-05, + "loss": 0.0259, + "step": 47400 + }, + { + "epoch": 9.204067763700792, + "grad_norm": 0.15360955893993378, + "learning_rate": 5.837810562258831e-05, + "loss": 0.0224, + "step": 47410 + }, + { + "epoch": 9.206009417018592, + "grad_norm": 0.22555190324783325, + "learning_rate": 5.836180424208405e-05, + "loss": 0.0269, + "step": 47420 + }, + { + "epoch": 9.207951070336392, + "grad_norm": 0.17055320739746094, + "learning_rate": 5.834550194714543e-05, + "loss": 0.0241, + "step": 47430 + }, + { + "epoch": 9.20989272365419, + "grad_norm": 0.167196124792099, + "learning_rate": 5.832919873955528e-05, + "loss": 0.0299, + "step": 47440 + }, + { + "epoch": 9.211834376971991, + "grad_norm": 0.25428804755210876, + "learning_rate": 5.831289462109648e-05, + "loss": 0.0301, + "step": 47450 + }, + { + "epoch": 9.213776030289791, + "grad_norm": 0.17853565514087677, + "learning_rate": 5.829658959355202e-05, + "loss": 0.0326, + "step": 47460 + }, + { + "epoch": 9.215717683607592, + "grad_norm": 0.24842627346515656, + "learning_rate": 5.8280283658704995e-05, + "loss": 0.0208, + "step": 47470 + }, + { + "epoch": 9.217659336925392, + "grad_norm": 0.2480815052986145, + "learning_rate": 5.826397681833859e-05, + "loss": 0.0217, + "step": 47480 + }, + { + "epoch": 9.219600990243192, + "grad_norm": 0.12931512296199799, + "learning_rate": 5.824766907423612e-05, + "loss": 0.0286, + "step": 47490 + }, + { + "epoch": 9.221542643560992, + "grad_norm": 0.26412421464920044, + "learning_rate": 5.8231360428180945e-05, + "loss": 0.0256, + "step": 47500 + }, + { + "epoch": 9.223484296878793, + "grad_norm": 0.12526746094226837, + "learning_rate": 5.821505088195658e-05, + "loss": 0.0276, + "step": 47510 + }, + { + "epoch": 9.225425950196593, + "grad_norm": 0.1725071519613266, + "learning_rate": 5.819874043734661e-05, + "loss": 0.0257, + "step": 47520 + }, + { + "epoch": 9.227367603514393, + "grad_norm": 0.1775413304567337, + "learning_rate": 5.8182429096134695e-05, + "loss": 0.0252, + "step": 47530 + }, + { + "epoch": 9.229309256832192, + "grad_norm": 0.19043006002902985, + "learning_rate": 5.816611686010465e-05, + "loss": 0.0338, + "step": 47540 + }, + { + "epoch": 9.231250910149992, + "grad_norm": 0.1309259831905365, + "learning_rate": 5.814980373104033e-05, + "loss": 0.0234, + "step": 47550 + }, + { + "epoch": 9.233192563467792, + "grad_norm": 0.16583319008350372, + "learning_rate": 5.813348971072572e-05, + "loss": 0.0281, + "step": 47560 + }, + { + "epoch": 9.235134216785593, + "grad_norm": 0.2103344053030014, + "learning_rate": 5.811717480094492e-05, + "loss": 0.0278, + "step": 47570 + }, + { + "epoch": 9.237075870103393, + "grad_norm": 0.17746110260486603, + "learning_rate": 5.810085900348209e-05, + "loss": 0.0279, + "step": 47580 + }, + { + "epoch": 9.239017523421193, + "grad_norm": 0.26422199606895447, + "learning_rate": 5.8084542320121483e-05, + "loss": 0.0298, + "step": 47590 + }, + { + "epoch": 9.240959176738993, + "grad_norm": 0.1960185170173645, + "learning_rate": 5.8068224752647497e-05, + "loss": 0.0247, + "step": 47600 + }, + { + "epoch": 9.242900830056794, + "grad_norm": 0.24432408809661865, + "learning_rate": 5.805190630284457e-05, + "loss": 0.027, + "step": 47610 + }, + { + "epoch": 9.244842483374594, + "grad_norm": 0.0905044674873352, + "learning_rate": 5.803558697249729e-05, + "loss": 0.0223, + "step": 47620 + }, + { + "epoch": 9.246784136692394, + "grad_norm": 0.13557223975658417, + "learning_rate": 5.8019266763390276e-05, + "loss": 0.0359, + "step": 47630 + }, + { + "epoch": 9.248725790010194, + "grad_norm": 0.17130477726459503, + "learning_rate": 5.800294567730833e-05, + "loss": 0.0282, + "step": 47640 + }, + { + "epoch": 9.250667443327993, + "grad_norm": 0.18100948631763458, + "learning_rate": 5.798662371603628e-05, + "loss": 0.0265, + "step": 47650 + }, + { + "epoch": 9.252609096645793, + "grad_norm": 0.18536530435085297, + "learning_rate": 5.797030088135904e-05, + "loss": 0.0315, + "step": 47660 + }, + { + "epoch": 9.254550749963594, + "grad_norm": 0.22876110672950745, + "learning_rate": 5.7953977175061705e-05, + "loss": 0.035, + "step": 47670 + }, + { + "epoch": 9.256492403281394, + "grad_norm": 0.17066560685634613, + "learning_rate": 5.793765259892938e-05, + "loss": 0.0269, + "step": 47680 + }, + { + "epoch": 9.258434056599194, + "grad_norm": 0.18435296416282654, + "learning_rate": 5.792132715474729e-05, + "loss": 0.0273, + "step": 47690 + }, + { + "epoch": 9.260375709916994, + "grad_norm": 0.16039708256721497, + "learning_rate": 5.790500084430078e-05, + "loss": 0.0366, + "step": 47700 + }, + { + "epoch": 9.262317363234795, + "grad_norm": 0.14166902005672455, + "learning_rate": 5.7888673669375274e-05, + "loss": 0.0309, + "step": 47710 + }, + { + "epoch": 9.264259016552595, + "grad_norm": 0.2584669291973114, + "learning_rate": 5.787234563175625e-05, + "loss": 0.0254, + "step": 47720 + }, + { + "epoch": 9.266200669870395, + "grad_norm": 0.28869420289993286, + "learning_rate": 5.7856016733229355e-05, + "loss": 0.0304, + "step": 47730 + }, + { + "epoch": 9.268142323188195, + "grad_norm": 0.16954176127910614, + "learning_rate": 5.7839686975580297e-05, + "loss": 0.023, + "step": 47740 + }, + { + "epoch": 9.270083976505996, + "grad_norm": 0.18046411871910095, + "learning_rate": 5.782335636059484e-05, + "loss": 0.0293, + "step": 47750 + }, + { + "epoch": 9.272025629823794, + "grad_norm": 0.17617736756801605, + "learning_rate": 5.780702489005889e-05, + "loss": 0.0258, + "step": 47760 + }, + { + "epoch": 9.273967283141594, + "grad_norm": 0.1572423130273819, + "learning_rate": 5.779069256575845e-05, + "loss": 0.0351, + "step": 47770 + }, + { + "epoch": 9.275908936459395, + "grad_norm": 0.17757758498191833, + "learning_rate": 5.7774359389479574e-05, + "loss": 0.0346, + "step": 47780 + }, + { + "epoch": 9.277850589777195, + "grad_norm": 0.4187242090702057, + "learning_rate": 5.775802536300845e-05, + "loss": 0.0288, + "step": 47790 + }, + { + "epoch": 9.279792243094995, + "grad_norm": 0.1631302684545517, + "learning_rate": 5.774169048813134e-05, + "loss": 0.0318, + "step": 47800 + }, + { + "epoch": 9.281733896412796, + "grad_norm": 0.15326561033725739, + "learning_rate": 5.77253547666346e-05, + "loss": 0.0264, + "step": 47810 + }, + { + "epoch": 9.283675549730596, + "grad_norm": 0.10491926223039627, + "learning_rate": 5.770901820030465e-05, + "loss": 0.0225, + "step": 47820 + }, + { + "epoch": 9.285617203048396, + "grad_norm": 0.17897234857082367, + "learning_rate": 5.769268079092809e-05, + "loss": 0.0303, + "step": 47830 + }, + { + "epoch": 9.287558856366196, + "grad_norm": 0.19280272722244263, + "learning_rate": 5.767634254029151e-05, + "loss": 0.0244, + "step": 47840 + }, + { + "epoch": 9.289500509683997, + "grad_norm": 0.13376657664775848, + "learning_rate": 5.7660003450181655e-05, + "loss": 0.0337, + "step": 47850 + }, + { + "epoch": 9.291442163001797, + "grad_norm": 0.1361461877822876, + "learning_rate": 5.764366352238534e-05, + "loss": 0.0264, + "step": 47860 + }, + { + "epoch": 9.293383816319595, + "grad_norm": 0.23623794317245483, + "learning_rate": 5.7627322758689474e-05, + "loss": 0.0307, + "step": 47870 + }, + { + "epoch": 9.295325469637396, + "grad_norm": 0.24123598635196686, + "learning_rate": 5.761098116088105e-05, + "loss": 0.0253, + "step": 47880 + }, + { + "epoch": 9.297267122955196, + "grad_norm": 0.12835241854190826, + "learning_rate": 5.759463873074717e-05, + "loss": 0.0218, + "step": 47890 + }, + { + "epoch": 9.299208776272996, + "grad_norm": 0.16654609143733978, + "learning_rate": 5.757829547007504e-05, + "loss": 0.0299, + "step": 47900 + }, + { + "epoch": 9.301150429590797, + "grad_norm": 0.1541460007429123, + "learning_rate": 5.756195138065189e-05, + "loss": 0.0307, + "step": 47910 + }, + { + "epoch": 9.303092082908597, + "grad_norm": 0.08415880054235458, + "learning_rate": 5.754560646426511e-05, + "loss": 0.0192, + "step": 47920 + }, + { + "epoch": 9.305033736226397, + "grad_norm": 0.14310045540332794, + "learning_rate": 5.752926072270216e-05, + "loss": 0.0272, + "step": 47930 + }, + { + "epoch": 9.306975389544197, + "grad_norm": 0.18558746576309204, + "learning_rate": 5.7512914157750563e-05, + "loss": 0.0248, + "step": 47940 + }, + { + "epoch": 9.308917042861998, + "grad_norm": 0.10252557694911957, + "learning_rate": 5.749656677119798e-05, + "loss": 0.0249, + "step": 47950 + }, + { + "epoch": 9.310858696179798, + "grad_norm": 0.22810940444469452, + "learning_rate": 5.7480218564832125e-05, + "loss": 0.0257, + "step": 47960 + }, + { + "epoch": 9.312800349497596, + "grad_norm": 0.11468461900949478, + "learning_rate": 5.746386954044082e-05, + "loss": 0.0215, + "step": 47970 + }, + { + "epoch": 9.314742002815397, + "grad_norm": 0.11900336295366287, + "learning_rate": 5.744751969981195e-05, + "loss": 0.0306, + "step": 47980 + }, + { + "epoch": 9.316683656133197, + "grad_norm": 0.12410027533769608, + "learning_rate": 5.7431169044733526e-05, + "loss": 0.0274, + "step": 47990 + }, + { + "epoch": 9.318625309450997, + "grad_norm": 0.2266850620508194, + "learning_rate": 5.741481757699364e-05, + "loss": 0.032, + "step": 48000 + }, + { + "epoch": 9.320566962768797, + "grad_norm": 0.23546524345874786, + "learning_rate": 5.7398465298380434e-05, + "loss": 0.0297, + "step": 48010 + }, + { + "epoch": 9.322508616086598, + "grad_norm": 0.10507258772850037, + "learning_rate": 5.7382112210682193e-05, + "loss": 0.0259, + "step": 48020 + }, + { + "epoch": 9.324450269404398, + "grad_norm": 0.178772434592247, + "learning_rate": 5.7365758315687266e-05, + "loss": 0.0261, + "step": 48030 + }, + { + "epoch": 9.326391922722198, + "grad_norm": 0.18133904039859772, + "learning_rate": 5.734940361518407e-05, + "loss": 0.0279, + "step": 48040 + }, + { + "epoch": 9.328333576039999, + "grad_norm": 0.13631819188594818, + "learning_rate": 5.733304811096116e-05, + "loss": 0.0355, + "step": 48050 + }, + { + "epoch": 9.330275229357799, + "grad_norm": 0.08606553077697754, + "learning_rate": 5.731669180480713e-05, + "loss": 0.0292, + "step": 48060 + }, + { + "epoch": 9.3322168826756, + "grad_norm": 0.1575414538383484, + "learning_rate": 5.730033469851067e-05, + "loss": 0.0173, + "step": 48070 + }, + { + "epoch": 9.334158535993398, + "grad_norm": 0.1638353019952774, + "learning_rate": 5.7283976793860607e-05, + "loss": 0.0277, + "step": 48080 + }, + { + "epoch": 9.336100189311198, + "grad_norm": 0.14018312096595764, + "learning_rate": 5.726761809264577e-05, + "loss": 0.0275, + "step": 48090 + }, + { + "epoch": 9.338041842628998, + "grad_norm": 0.22754952311515808, + "learning_rate": 5.7251258596655155e-05, + "loss": 0.0263, + "step": 48100 + }, + { + "epoch": 9.339983495946798, + "grad_norm": 0.21420755982398987, + "learning_rate": 5.72348983076778e-05, + "loss": 0.0276, + "step": 48110 + }, + { + "epoch": 9.341925149264599, + "grad_norm": 0.1455446481704712, + "learning_rate": 5.7218537227502854e-05, + "loss": 0.0264, + "step": 48120 + }, + { + "epoch": 9.343866802582399, + "grad_norm": 0.1906912475824356, + "learning_rate": 5.720217535791951e-05, + "loss": 0.0251, + "step": 48130 + }, + { + "epoch": 9.3458084559002, + "grad_norm": 0.15124262869358063, + "learning_rate": 5.718581270071711e-05, + "loss": 0.0224, + "step": 48140 + }, + { + "epoch": 9.347750109218, + "grad_norm": 0.17838452756404877, + "learning_rate": 5.716944925768505e-05, + "loss": 0.0307, + "step": 48150 + }, + { + "epoch": 9.3496917625358, + "grad_norm": 0.08418498933315277, + "learning_rate": 5.7153085030612786e-05, + "loss": 0.0259, + "step": 48160 + }, + { + "epoch": 9.3516334158536, + "grad_norm": 0.164537712931633, + "learning_rate": 5.713672002128989e-05, + "loss": 0.0261, + "step": 48170 + }, + { + "epoch": 9.353575069171399, + "grad_norm": 0.20341964066028595, + "learning_rate": 5.712035423150602e-05, + "loss": 0.0399, + "step": 48180 + }, + { + "epoch": 9.355516722489199, + "grad_norm": 0.2644931375980377, + "learning_rate": 5.710398766305094e-05, + "loss": 0.0279, + "step": 48190 + }, + { + "epoch": 9.357458375807, + "grad_norm": 0.11192839592695236, + "learning_rate": 5.708762031771442e-05, + "loss": 0.0206, + "step": 48200 + }, + { + "epoch": 9.3594000291248, + "grad_norm": 0.24934245645999908, + "learning_rate": 5.707125219728642e-05, + "loss": 0.0301, + "step": 48210 + }, + { + "epoch": 9.3613416824426, + "grad_norm": 0.1995455026626587, + "learning_rate": 5.7054883303556905e-05, + "loss": 0.0292, + "step": 48220 + }, + { + "epoch": 9.3632833357604, + "grad_norm": 0.2190834879875183, + "learning_rate": 5.703851363831595e-05, + "loss": 0.0246, + "step": 48230 + }, + { + "epoch": 9.3652249890782, + "grad_norm": 0.12730196118354797, + "learning_rate": 5.702214320335374e-05, + "loss": 0.0292, + "step": 48240 + }, + { + "epoch": 9.367166642396, + "grad_norm": 0.18086916208267212, + "learning_rate": 5.7005772000460514e-05, + "loss": 0.0275, + "step": 48250 + }, + { + "epoch": 9.3691082957138, + "grad_norm": 0.16395743191242218, + "learning_rate": 5.698940003142656e-05, + "loss": 0.0272, + "step": 48260 + }, + { + "epoch": 9.371049949031601, + "grad_norm": 0.25708243250846863, + "learning_rate": 5.697302729804236e-05, + "loss": 0.025, + "step": 48270 + }, + { + "epoch": 9.372991602349401, + "grad_norm": 0.20320922136306763, + "learning_rate": 5.695665380209837e-05, + "loss": 0.0267, + "step": 48280 + }, + { + "epoch": 9.3749332556672, + "grad_norm": 0.2766076326370239, + "learning_rate": 5.6940279545385165e-05, + "loss": 0.0232, + "step": 48290 + }, + { + "epoch": 9.376874908985, + "grad_norm": 0.13362526893615723, + "learning_rate": 5.692390452969344e-05, + "loss": 0.0195, + "step": 48300 + }, + { + "epoch": 9.3788165623028, + "grad_norm": 0.11813269555568695, + "learning_rate": 5.690752875681392e-05, + "loss": 0.0274, + "step": 48310 + }, + { + "epoch": 9.3807582156206, + "grad_norm": 0.20150640606880188, + "learning_rate": 5.6891152228537435e-05, + "loss": 0.0231, + "step": 48320 + }, + { + "epoch": 9.382699868938401, + "grad_norm": 0.18579794466495514, + "learning_rate": 5.687477494665492e-05, + "loss": 0.0259, + "step": 48330 + }, + { + "epoch": 9.384641522256201, + "grad_norm": 0.11510023474693298, + "learning_rate": 5.685839691295733e-05, + "loss": 0.0241, + "step": 48340 + }, + { + "epoch": 9.386583175574001, + "grad_norm": 0.13130778074264526, + "learning_rate": 5.6842018129235786e-05, + "loss": 0.0327, + "step": 48350 + }, + { + "epoch": 9.388524828891802, + "grad_norm": 0.2001517117023468, + "learning_rate": 5.6825638597281404e-05, + "loss": 0.0253, + "step": 48360 + }, + { + "epoch": 9.390466482209602, + "grad_norm": 0.11839043349027634, + "learning_rate": 5.680925831888546e-05, + "loss": 0.0215, + "step": 48370 + }, + { + "epoch": 9.392408135527402, + "grad_norm": 0.20033252239227295, + "learning_rate": 5.6792877295839274e-05, + "loss": 0.0326, + "step": 48380 + }, + { + "epoch": 9.3943497888452, + "grad_norm": 0.1994612067937851, + "learning_rate": 5.6776495529934224e-05, + "loss": 0.0333, + "step": 48390 + }, + { + "epoch": 9.396291442163001, + "grad_norm": 0.2068672627210617, + "learning_rate": 5.6760113022961824e-05, + "loss": 0.0234, + "step": 48400 + }, + { + "epoch": 9.398233095480801, + "grad_norm": 0.1302889883518219, + "learning_rate": 5.6743729776713617e-05, + "loss": 0.0237, + "step": 48410 + }, + { + "epoch": 9.400174748798602, + "grad_norm": 0.20671837031841278, + "learning_rate": 5.672734579298126e-05, + "loss": 0.0259, + "step": 48420 + }, + { + "epoch": 9.402116402116402, + "grad_norm": 0.16231656074523926, + "learning_rate": 5.671096107355649e-05, + "loss": 0.0245, + "step": 48430 + }, + { + "epoch": 9.404058055434202, + "grad_norm": 0.13516272604465485, + "learning_rate": 5.66945756202311e-05, + "loss": 0.026, + "step": 48440 + }, + { + "epoch": 9.405999708752002, + "grad_norm": 0.19839422404766083, + "learning_rate": 5.667818943479699e-05, + "loss": 0.024, + "step": 48450 + }, + { + "epoch": 9.407941362069803, + "grad_norm": 0.14894849061965942, + "learning_rate": 5.666180251904612e-05, + "loss": 0.0229, + "step": 48460 + }, + { + "epoch": 9.409883015387603, + "grad_norm": 0.40724533796310425, + "learning_rate": 5.6645414874770555e-05, + "loss": 0.0305, + "step": 48470 + }, + { + "epoch": 9.411824668705403, + "grad_norm": 0.17334513366222382, + "learning_rate": 5.66290265037624e-05, + "loss": 0.0227, + "step": 48480 + }, + { + "epoch": 9.413766322023204, + "grad_norm": 0.21354492008686066, + "learning_rate": 5.661263740781386e-05, + "loss": 0.029, + "step": 48490 + }, + { + "epoch": 9.415707975341002, + "grad_norm": 0.1394340991973877, + "learning_rate": 5.6596247588717254e-05, + "loss": 0.0247, + "step": 48500 + }, + { + "epoch": 9.417649628658802, + "grad_norm": 0.14565297961235046, + "learning_rate": 5.6579857048264926e-05, + "loss": 0.0226, + "step": 48510 + }, + { + "epoch": 9.419591281976603, + "grad_norm": 0.1938851922750473, + "learning_rate": 5.6563465788249314e-05, + "loss": 0.0277, + "step": 48520 + }, + { + "epoch": 9.421532935294403, + "grad_norm": 0.11615218222141266, + "learning_rate": 5.6547073810462956e-05, + "loss": 0.0288, + "step": 48530 + }, + { + "epoch": 9.423474588612203, + "grad_norm": 0.10623469948768616, + "learning_rate": 5.653068111669846e-05, + "loss": 0.0203, + "step": 48540 + }, + { + "epoch": 9.425416241930003, + "grad_norm": 0.1438930779695511, + "learning_rate": 5.651428770874848e-05, + "loss": 0.0252, + "step": 48550 + }, + { + "epoch": 9.427357895247804, + "grad_norm": 0.2952445447444916, + "learning_rate": 5.64978935884058e-05, + "loss": 0.0273, + "step": 48560 + }, + { + "epoch": 9.429299548565604, + "grad_norm": 0.20873712003231049, + "learning_rate": 5.6481498757463244e-05, + "loss": 0.0228, + "step": 48570 + }, + { + "epoch": 9.431241201883404, + "grad_norm": 0.19164392352104187, + "learning_rate": 5.646510321771373e-05, + "loss": 0.0229, + "step": 48580 + }, + { + "epoch": 9.433182855201204, + "grad_norm": 0.16485103964805603, + "learning_rate": 5.644870697095024e-05, + "loss": 0.0218, + "step": 48590 + }, + { + "epoch": 9.435124508519005, + "grad_norm": 0.1575695127248764, + "learning_rate": 5.643231001896586e-05, + "loss": 0.027, + "step": 48600 + }, + { + "epoch": 9.437066161836803, + "grad_norm": 0.29965609312057495, + "learning_rate": 5.6415912363553726e-05, + "loss": 0.0268, + "step": 48610 + }, + { + "epoch": 9.439007815154604, + "grad_norm": 0.2796837091445923, + "learning_rate": 5.639951400650706e-05, + "loss": 0.0263, + "step": 48620 + }, + { + "epoch": 9.440949468472404, + "grad_norm": 0.15966123342514038, + "learning_rate": 5.6383114949619165e-05, + "loss": 0.0258, + "step": 48630 + }, + { + "epoch": 9.442891121790204, + "grad_norm": 0.2226765751838684, + "learning_rate": 5.636671519468342e-05, + "loss": 0.0297, + "step": 48640 + }, + { + "epoch": 9.444832775108004, + "grad_norm": 0.14472994208335876, + "learning_rate": 5.635031474349327e-05, + "loss": 0.0295, + "step": 48650 + }, + { + "epoch": 9.446774428425805, + "grad_norm": 0.09684714674949646, + "learning_rate": 5.6333913597842246e-05, + "loss": 0.0245, + "step": 48660 + }, + { + "epoch": 9.448716081743605, + "grad_norm": 0.25464966893196106, + "learning_rate": 5.6317511759523955e-05, + "loss": 0.0261, + "step": 48670 + }, + { + "epoch": 9.450657735061405, + "grad_norm": 0.20682673156261444, + "learning_rate": 5.630110923033207e-05, + "loss": 0.0242, + "step": 48680 + }, + { + "epoch": 9.452599388379205, + "grad_norm": 0.09131097048521042, + "learning_rate": 5.628470601206036e-05, + "loss": 0.0242, + "step": 48690 + }, + { + "epoch": 9.454541041697006, + "grad_norm": 0.220163032412529, + "learning_rate": 5.626830210650266e-05, + "loss": 0.0285, + "step": 48700 + }, + { + "epoch": 9.456482695014806, + "grad_norm": 0.20928244292736053, + "learning_rate": 5.625189751545285e-05, + "loss": 0.0255, + "step": 48710 + }, + { + "epoch": 9.458424348332604, + "grad_norm": 0.128305584192276, + "learning_rate": 5.6235492240704936e-05, + "loss": 0.0264, + "step": 48720 + }, + { + "epoch": 9.460366001650405, + "grad_norm": 0.23637257516384125, + "learning_rate": 5.621908628405296e-05, + "loss": 0.0268, + "step": 48730 + }, + { + "epoch": 9.462307654968205, + "grad_norm": 0.20031829178333282, + "learning_rate": 5.620267964729106e-05, + "loss": 0.0224, + "step": 48740 + }, + { + "epoch": 9.464249308286005, + "grad_norm": 0.15461739897727966, + "learning_rate": 5.618627233221344e-05, + "loss": 0.0261, + "step": 48750 + }, + { + "epoch": 9.466190961603806, + "grad_norm": 0.18264324963092804, + "learning_rate": 5.616986434061438e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 9.468132614921606, + "grad_norm": 0.16167056560516357, + "learning_rate": 5.615345567428822e-05, + "loss": 0.0311, + "step": 48770 + }, + { + "epoch": 9.470074268239406, + "grad_norm": 0.24242378771305084, + "learning_rate": 5.613704633502941e-05, + "loss": 0.0332, + "step": 48780 + }, + { + "epoch": 9.472015921557206, + "grad_norm": 0.12831245362758636, + "learning_rate": 5.612063632463245e-05, + "loss": 0.0305, + "step": 48790 + }, + { + "epoch": 9.473957574875007, + "grad_norm": 0.14246995747089386, + "learning_rate": 5.610422564489188e-05, + "loss": 0.0281, + "step": 48800 + }, + { + "epoch": 9.475899228192807, + "grad_norm": 0.1459207683801651, + "learning_rate": 5.608781429760239e-05, + "loss": 0.0279, + "step": 48810 + }, + { + "epoch": 9.477840881510605, + "grad_norm": 0.17315298318862915, + "learning_rate": 5.607140228455866e-05, + "loss": 0.0248, + "step": 48820 + }, + { + "epoch": 9.479782534828406, + "grad_norm": 0.16365528106689453, + "learning_rate": 5.605498960755553e-05, + "loss": 0.0247, + "step": 48830 + }, + { + "epoch": 9.481724188146206, + "grad_norm": 0.20460748672485352, + "learning_rate": 5.603857626838782e-05, + "loss": 0.0267, + "step": 48840 + }, + { + "epoch": 9.483665841464006, + "grad_norm": 0.08868443220853806, + "learning_rate": 5.60221622688505e-05, + "loss": 0.0346, + "step": 48850 + }, + { + "epoch": 9.485607494781807, + "grad_norm": 0.2707081139087677, + "learning_rate": 5.6005747610738565e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 9.487549148099607, + "grad_norm": 0.23829296231269836, + "learning_rate": 5.59893322958471e-05, + "loss": 0.0206, + "step": 48870 + }, + { + "epoch": 9.489490801417407, + "grad_norm": 0.18477842211723328, + "learning_rate": 5.5972916325971256e-05, + "loss": 0.0238, + "step": 48880 + }, + { + "epoch": 9.491432454735207, + "grad_norm": 0.23181578516960144, + "learning_rate": 5.595649970290628e-05, + "loss": 0.0248, + "step": 48890 + }, + { + "epoch": 9.493374108053008, + "grad_norm": 0.3081984221935272, + "learning_rate": 5.5940082428447426e-05, + "loss": 0.0314, + "step": 48900 + }, + { + "epoch": 9.495315761370808, + "grad_norm": 0.286953330039978, + "learning_rate": 5.592366450439012e-05, + "loss": 0.0296, + "step": 48910 + }, + { + "epoch": 9.497257414688608, + "grad_norm": 0.10926362127065659, + "learning_rate": 5.590724593252975e-05, + "loss": 0.0237, + "step": 48920 + }, + { + "epoch": 9.499199068006407, + "grad_norm": 0.23592399060726166, + "learning_rate": 5.589082671466184e-05, + "loss": 0.0228, + "step": 48930 + }, + { + "epoch": 9.501140721324207, + "grad_norm": 0.15675921738147736, + "learning_rate": 5.587440685258199e-05, + "loss": 0.0348, + "step": 48940 + }, + { + "epoch": 9.503082374642007, + "grad_norm": 0.22418710589408875, + "learning_rate": 5.585798634808583e-05, + "loss": 0.0265, + "step": 48950 + }, + { + "epoch": 9.505024027959807, + "grad_norm": 0.138553187251091, + "learning_rate": 5.584156520296909e-05, + "loss": 0.0285, + "step": 48960 + }, + { + "epoch": 9.506965681277608, + "grad_norm": 0.1780085563659668, + "learning_rate": 5.582514341902757e-05, + "loss": 0.0236, + "step": 48970 + }, + { + "epoch": 9.508907334595408, + "grad_norm": 0.11711210012435913, + "learning_rate": 5.580872099805713e-05, + "loss": 0.0226, + "step": 48980 + }, + { + "epoch": 9.510848987913208, + "grad_norm": 0.20871926844120026, + "learning_rate": 5.5792297941853674e-05, + "loss": 0.0254, + "step": 48990 + }, + { + "epoch": 9.512790641231009, + "grad_norm": 0.24462206661701202, + "learning_rate": 5.5775874252213247e-05, + "loss": 0.0276, + "step": 49000 + }, + { + "epoch": 9.514732294548809, + "grad_norm": 0.12239263951778412, + "learning_rate": 5.575944993093189e-05, + "loss": 0.0212, + "step": 49010 + }, + { + "epoch": 9.51667394786661, + "grad_norm": 0.10828784853219986, + "learning_rate": 5.574302497980574e-05, + "loss": 0.0253, + "step": 49020 + }, + { + "epoch": 9.518615601184408, + "grad_norm": 0.22741790115833282, + "learning_rate": 5.5726599400631005e-05, + "loss": 0.0241, + "step": 49030 + }, + { + "epoch": 9.520557254502208, + "grad_norm": 0.15646347403526306, + "learning_rate": 5.571017319520399e-05, + "loss": 0.0236, + "step": 49040 + }, + { + "epoch": 9.522498907820008, + "grad_norm": 0.31730321049690247, + "learning_rate": 5.5693746365320984e-05, + "loss": 0.0277, + "step": 49050 + }, + { + "epoch": 9.524440561137808, + "grad_norm": 0.1476486623287201, + "learning_rate": 5.567731891277848e-05, + "loss": 0.0282, + "step": 49060 + }, + { + "epoch": 9.526382214455609, + "grad_norm": 0.19641737639904022, + "learning_rate": 5.566089083937288e-05, + "loss": 0.0328, + "step": 49070 + }, + { + "epoch": 9.528323867773409, + "grad_norm": 0.16229739785194397, + "learning_rate": 5.564446214690079e-05, + "loss": 0.0237, + "step": 49080 + }, + { + "epoch": 9.53026552109121, + "grad_norm": 0.15208236873149872, + "learning_rate": 5.56280328371588e-05, + "loss": 0.0293, + "step": 49090 + }, + { + "epoch": 9.53220717440901, + "grad_norm": 0.15131624042987823, + "learning_rate": 5.5611602911943596e-05, + "loss": 0.0274, + "step": 49100 + }, + { + "epoch": 9.53414882772681, + "grad_norm": 0.18076975643634796, + "learning_rate": 5.559517237305194e-05, + "loss": 0.0258, + "step": 49110 + }, + { + "epoch": 9.53609048104461, + "grad_norm": 0.2123313546180725, + "learning_rate": 5.557874122228064e-05, + "loss": 0.0226, + "step": 49120 + }, + { + "epoch": 9.53803213436241, + "grad_norm": 0.13575726747512817, + "learning_rate": 5.55623094614266e-05, + "loss": 0.023, + "step": 49130 + }, + { + "epoch": 9.539973787680209, + "grad_norm": 0.21665583550930023, + "learning_rate": 5.5545877092286743e-05, + "loss": 0.0243, + "step": 49140 + }, + { + "epoch": 9.54191544099801, + "grad_norm": 0.090555839240551, + "learning_rate": 5.552944411665811e-05, + "loss": 0.0221, + "step": 49150 + }, + { + "epoch": 9.54385709431581, + "grad_norm": 0.15469160676002502, + "learning_rate": 5.5513010536337795e-05, + "loss": 0.0221, + "step": 49160 + }, + { + "epoch": 9.54579874763361, + "grad_norm": 0.22660000622272491, + "learning_rate": 5.5496576353122934e-05, + "loss": 0.0211, + "step": 49170 + }, + { + "epoch": 9.54774040095141, + "grad_norm": 0.22964054346084595, + "learning_rate": 5.548014156881074e-05, + "loss": 0.0274, + "step": 49180 + }, + { + "epoch": 9.54968205426921, + "grad_norm": 0.1728970855474472, + "learning_rate": 5.5463706185198494e-05, + "loss": 0.0327, + "step": 49190 + }, + { + "epoch": 9.55162370758701, + "grad_norm": 0.17834480106830597, + "learning_rate": 5.544727020408358e-05, + "loss": 0.0312, + "step": 49200 + }, + { + "epoch": 9.55356536090481, + "grad_norm": 0.22164104878902435, + "learning_rate": 5.5430833627263366e-05, + "loss": 0.0293, + "step": 49210 + }, + { + "epoch": 9.555507014222611, + "grad_norm": 0.14601346850395203, + "learning_rate": 5.5414396456535354e-05, + "loss": 0.024, + "step": 49220 + }, + { + "epoch": 9.557448667540411, + "grad_norm": 0.15384796261787415, + "learning_rate": 5.539795869369709e-05, + "loss": 0.0265, + "step": 49230 + }, + { + "epoch": 9.55939032085821, + "grad_norm": 0.17531390488147736, + "learning_rate": 5.538152034054618e-05, + "loss": 0.0314, + "step": 49240 + }, + { + "epoch": 9.56133197417601, + "grad_norm": 0.13173808157444, + "learning_rate": 5.536508139888028e-05, + "loss": 0.0285, + "step": 49250 + }, + { + "epoch": 9.56327362749381, + "grad_norm": 0.2058638036251068, + "learning_rate": 5.534864187049716e-05, + "loss": 0.0252, + "step": 49260 + }, + { + "epoch": 9.56521528081161, + "grad_norm": 0.2461201697587967, + "learning_rate": 5.5332201757194615e-05, + "loss": 0.0283, + "step": 49270 + }, + { + "epoch": 9.567156934129411, + "grad_norm": 0.14754973351955414, + "learning_rate": 5.531576106077049e-05, + "loss": 0.0235, + "step": 49280 + }, + { + "epoch": 9.569098587447211, + "grad_norm": 0.19510585069656372, + "learning_rate": 5.529931978302272e-05, + "loss": 0.019, + "step": 49290 + }, + { + "epoch": 9.571040240765011, + "grad_norm": 0.17448952794075012, + "learning_rate": 5.528287792574932e-05, + "loss": 0.0227, + "step": 49300 + }, + { + "epoch": 9.572981894082812, + "grad_norm": 0.18582157790660858, + "learning_rate": 5.5266435490748324e-05, + "loss": 0.0232, + "step": 49310 + }, + { + "epoch": 9.574923547400612, + "grad_norm": 0.17176222801208496, + "learning_rate": 5.524999247981787e-05, + "loss": 0.0253, + "step": 49320 + }, + { + "epoch": 9.576865200718412, + "grad_norm": 0.15030664205551147, + "learning_rate": 5.523354889475613e-05, + "loss": 0.0314, + "step": 49330 + }, + { + "epoch": 9.578806854036213, + "grad_norm": 0.2837184965610504, + "learning_rate": 5.521710473736134e-05, + "loss": 0.0259, + "step": 49340 + }, + { + "epoch": 9.580748507354013, + "grad_norm": 0.25380563735961914, + "learning_rate": 5.5200660009431826e-05, + "loss": 0.0299, + "step": 49350 + }, + { + "epoch": 9.582690160671811, + "grad_norm": 0.11424683779478073, + "learning_rate": 5.518421471276596e-05, + "loss": 0.0238, + "step": 49360 + }, + { + "epoch": 9.584631813989612, + "grad_norm": 0.1478542685508728, + "learning_rate": 5.5167768849162174e-05, + "loss": 0.0267, + "step": 49370 + }, + { + "epoch": 9.586573467307412, + "grad_norm": 0.24314315617084503, + "learning_rate": 5.515132242041893e-05, + "loss": 0.0292, + "step": 49380 + }, + { + "epoch": 9.588515120625212, + "grad_norm": 0.19025704264640808, + "learning_rate": 5.513487542833483e-05, + "loss": 0.0323, + "step": 49390 + }, + { + "epoch": 9.590456773943012, + "grad_norm": 0.19913144409656525, + "learning_rate": 5.5118427874708466e-05, + "loss": 0.0229, + "step": 49400 + }, + { + "epoch": 9.592398427260813, + "grad_norm": 0.13077910244464874, + "learning_rate": 5.510197976133853e-05, + "loss": 0.0247, + "step": 49410 + }, + { + "epoch": 9.594340080578613, + "grad_norm": 0.15004123747348785, + "learning_rate": 5.508553109002376e-05, + "loss": 0.0201, + "step": 49420 + }, + { + "epoch": 9.596281733896413, + "grad_norm": 0.19914564490318298, + "learning_rate": 5.5069081862562957e-05, + "loss": 0.0239, + "step": 49430 + }, + { + "epoch": 9.598223387214214, + "grad_norm": 0.1427185982465744, + "learning_rate": 5.5052632080754965e-05, + "loss": 0.0258, + "step": 49440 + }, + { + "epoch": 9.600165040532014, + "grad_norm": 0.18793341517448425, + "learning_rate": 5.503618174639874e-05, + "loss": 0.0273, + "step": 49450 + }, + { + "epoch": 9.602106693849812, + "grad_norm": 0.1297074556350708, + "learning_rate": 5.501973086129325e-05, + "loss": 0.0335, + "step": 49460 + }, + { + "epoch": 9.604048347167613, + "grad_norm": 0.1582275778055191, + "learning_rate": 5.500327942723753e-05, + "loss": 0.0225, + "step": 49470 + }, + { + "epoch": 9.605990000485413, + "grad_norm": 0.13691940903663635, + "learning_rate": 5.498682744603071e-05, + "loss": 0.0214, + "step": 49480 + }, + { + "epoch": 9.607931653803213, + "grad_norm": 0.15465179085731506, + "learning_rate": 5.497037491947194e-05, + "loss": 0.0227, + "step": 49490 + }, + { + "epoch": 9.609873307121013, + "grad_norm": 0.2472810298204422, + "learning_rate": 5.4953921849360424e-05, + "loss": 0.0278, + "step": 49500 + }, + { + "epoch": 9.611814960438814, + "grad_norm": 0.13816985487937927, + "learning_rate": 5.493746823749547e-05, + "loss": 0.0206, + "step": 49510 + }, + { + "epoch": 9.613756613756614, + "grad_norm": 0.09861411154270172, + "learning_rate": 5.49210140856764e-05, + "loss": 0.0194, + "step": 49520 + }, + { + "epoch": 9.615698267074414, + "grad_norm": 0.15617330372333527, + "learning_rate": 5.4904559395702647e-05, + "loss": 0.0196, + "step": 49530 + }, + { + "epoch": 9.617639920392214, + "grad_norm": 0.12533660233020782, + "learning_rate": 5.488810416937364e-05, + "loss": 0.0188, + "step": 49540 + }, + { + "epoch": 9.619581573710015, + "grad_norm": 0.12351743876934052, + "learning_rate": 5.4871648408488926e-05, + "loss": 0.0243, + "step": 49550 + }, + { + "epoch": 9.621523227027815, + "grad_norm": 0.14662107825279236, + "learning_rate": 5.485519211484807e-05, + "loss": 0.0198, + "step": 49560 + }, + { + "epoch": 9.623464880345614, + "grad_norm": 0.29044851660728455, + "learning_rate": 5.4838735290250676e-05, + "loss": 0.0296, + "step": 49570 + }, + { + "epoch": 9.625406533663414, + "grad_norm": 0.14574314653873444, + "learning_rate": 5.482227793649649e-05, + "loss": 0.0253, + "step": 49580 + }, + { + "epoch": 9.627348186981214, + "grad_norm": 0.19071824848651886, + "learning_rate": 5.480582005538524e-05, + "loss": 0.0299, + "step": 49590 + }, + { + "epoch": 9.629289840299014, + "grad_norm": 0.210913747549057, + "learning_rate": 5.478936164871671e-05, + "loss": 0.0279, + "step": 49600 + }, + { + "epoch": 9.631231493616815, + "grad_norm": 0.13160865008831024, + "learning_rate": 5.477290271829081e-05, + "loss": 0.0244, + "step": 49610 + }, + { + "epoch": 9.633173146934615, + "grad_norm": 0.18156510591506958, + "learning_rate": 5.475644326590744e-05, + "loss": 0.0256, + "step": 49620 + }, + { + "epoch": 9.635114800252415, + "grad_norm": 0.1467639058828354, + "learning_rate": 5.473998329336658e-05, + "loss": 0.0238, + "step": 49630 + }, + { + "epoch": 9.637056453570215, + "grad_norm": 0.15905971825122833, + "learning_rate": 5.4723522802468286e-05, + "loss": 0.0242, + "step": 49640 + }, + { + "epoch": 9.638998106888016, + "grad_norm": 0.1975431889295578, + "learning_rate": 5.470706179501264e-05, + "loss": 0.0235, + "step": 49650 + }, + { + "epoch": 9.640939760205816, + "grad_norm": 0.19322609901428223, + "learning_rate": 5.4690600272799776e-05, + "loss": 0.0204, + "step": 49660 + }, + { + "epoch": 9.642881413523614, + "grad_norm": 0.13460074365139008, + "learning_rate": 5.467413823762993e-05, + "loss": 0.0279, + "step": 49670 + }, + { + "epoch": 9.644823066841415, + "grad_norm": 0.11204443126916885, + "learning_rate": 5.465767569130335e-05, + "loss": 0.0249, + "step": 49680 + }, + { + "epoch": 9.646764720159215, + "grad_norm": 0.17872294783592224, + "learning_rate": 5.464121263562036e-05, + "loss": 0.0305, + "step": 49690 + }, + { + "epoch": 9.648706373477015, + "grad_norm": 0.16053897142410278, + "learning_rate": 5.4624749072381343e-05, + "loss": 0.0299, + "step": 49700 + }, + { + "epoch": 9.650648026794816, + "grad_norm": 0.1985166221857071, + "learning_rate": 5.460828500338672e-05, + "loss": 0.031, + "step": 49710 + }, + { + "epoch": 9.652589680112616, + "grad_norm": 0.20457755029201508, + "learning_rate": 5.459182043043698e-05, + "loss": 0.0318, + "step": 49720 + }, + { + "epoch": 9.654531333430416, + "grad_norm": 0.17956148087978363, + "learning_rate": 5.457535535533265e-05, + "loss": 0.0228, + "step": 49730 + }, + { + "epoch": 9.656472986748216, + "grad_norm": 0.3141503632068634, + "learning_rate": 5.4558889779874334e-05, + "loss": 0.0207, + "step": 49740 + }, + { + "epoch": 9.658414640066017, + "grad_norm": 0.0977109745144844, + "learning_rate": 5.454242370586269e-05, + "loss": 0.0185, + "step": 49750 + }, + { + "epoch": 9.660356293383817, + "grad_norm": 0.11544504761695862, + "learning_rate": 5.452595713509843e-05, + "loss": 0.0287, + "step": 49760 + }, + { + "epoch": 9.662297946701617, + "grad_norm": 0.18805210292339325, + "learning_rate": 5.4509490069382274e-05, + "loss": 0.0256, + "step": 49770 + }, + { + "epoch": 9.664239600019416, + "grad_norm": 0.17836792767047882, + "learning_rate": 5.449302251051509e-05, + "loss": 0.0233, + "step": 49780 + }, + { + "epoch": 9.666181253337216, + "grad_norm": 0.1927485466003418, + "learning_rate": 5.44765544602977e-05, + "loss": 0.0275, + "step": 49790 + }, + { + "epoch": 9.668122906655016, + "grad_norm": 0.18234632909297943, + "learning_rate": 5.446008592053107e-05, + "loss": 0.0257, + "step": 49800 + }, + { + "epoch": 9.670064559972817, + "grad_norm": 0.20840369164943695, + "learning_rate": 5.4443616893016135e-05, + "loss": 0.0262, + "step": 49810 + }, + { + "epoch": 9.672006213290617, + "grad_norm": 0.09229418635368347, + "learning_rate": 5.4427147379553934e-05, + "loss": 0.0207, + "step": 49820 + }, + { + "epoch": 9.673947866608417, + "grad_norm": 0.1785234957933426, + "learning_rate": 5.441067738194556e-05, + "loss": 0.0267, + "step": 49830 + }, + { + "epoch": 9.675889519926217, + "grad_norm": 0.15375196933746338, + "learning_rate": 5.439420690199214e-05, + "loss": 0.0201, + "step": 49840 + }, + { + "epoch": 9.677831173244018, + "grad_norm": 0.09212907403707504, + "learning_rate": 5.437773594149484e-05, + "loss": 0.0208, + "step": 49850 + }, + { + "epoch": 9.679772826561818, + "grad_norm": 0.17985479533672333, + "learning_rate": 5.436126450225495e-05, + "loss": 0.0268, + "step": 49860 + }, + { + "epoch": 9.681714479879618, + "grad_norm": 0.16673067212104797, + "learning_rate": 5.434479258607371e-05, + "loss": 0.0316, + "step": 49870 + }, + { + "epoch": 9.683656133197417, + "grad_norm": 0.1591537743806839, + "learning_rate": 5.43283201947525e-05, + "loss": 0.0273, + "step": 49880 + }, + { + "epoch": 9.685597786515217, + "grad_norm": 0.0791800394654274, + "learning_rate": 5.431184733009268e-05, + "loss": 0.029, + "step": 49890 + }, + { + "epoch": 9.687539439833017, + "grad_norm": 0.19337379932403564, + "learning_rate": 5.4295373993895736e-05, + "loss": 0.0271, + "step": 49900 + }, + { + "epoch": 9.689481093150818, + "grad_norm": 0.16503044962882996, + "learning_rate": 5.4278900187963157e-05, + "loss": 0.0193, + "step": 49910 + }, + { + "epoch": 9.691422746468618, + "grad_norm": 0.15664730966091156, + "learning_rate": 5.426242591409646e-05, + "loss": 0.0288, + "step": 49920 + }, + { + "epoch": 9.693364399786418, + "grad_norm": 0.15061424672603607, + "learning_rate": 5.42459511740973e-05, + "loss": 0.0225, + "step": 49930 + }, + { + "epoch": 9.695306053104218, + "grad_norm": 0.2339954525232315, + "learning_rate": 5.422947596976729e-05, + "loss": 0.0241, + "step": 49940 + }, + { + "epoch": 9.697247706422019, + "grad_norm": 0.1815415471792221, + "learning_rate": 5.4213000302908134e-05, + "loss": 0.0386, + "step": 49950 + }, + { + "epoch": 9.699189359739819, + "grad_norm": 0.3195284307003021, + "learning_rate": 5.419652417532162e-05, + "loss": 0.0295, + "step": 49960 + }, + { + "epoch": 9.70113101305762, + "grad_norm": 0.22564265131950378, + "learning_rate": 5.4180047588809534e-05, + "loss": 0.0264, + "step": 49970 + }, + { + "epoch": 9.70307266637542, + "grad_norm": 0.16694079339504242, + "learning_rate": 5.4163570545173704e-05, + "loss": 0.0238, + "step": 49980 + }, + { + "epoch": 9.70501431969322, + "grad_norm": 0.1844356209039688, + "learning_rate": 5.414709304621608e-05, + "loss": 0.0277, + "step": 49990 + }, + { + "epoch": 9.706955973011018, + "grad_norm": 0.17349237203598022, + "learning_rate": 5.413061509373858e-05, + "loss": 0.0261, + "step": 50000 + }, + { + "epoch": 9.708897626328818, + "grad_norm": 0.18731606006622314, + "learning_rate": 5.411413668954324e-05, + "loss": 0.0359, + "step": 50010 + }, + { + "epoch": 9.710839279646619, + "grad_norm": 0.1668590009212494, + "learning_rate": 5.4097657835432083e-05, + "loss": 0.0298, + "step": 50020 + }, + { + "epoch": 9.712780932964419, + "grad_norm": 0.10795105993747711, + "learning_rate": 5.408117853320723e-05, + "loss": 0.0203, + "step": 50030 + }, + { + "epoch": 9.71472258628222, + "grad_norm": 0.12315544486045837, + "learning_rate": 5.406469878467084e-05, + "loss": 0.0257, + "step": 50040 + }, + { + "epoch": 9.71666423960002, + "grad_norm": 0.17089851200580597, + "learning_rate": 5.404821859162509e-05, + "loss": 0.0276, + "step": 50050 + }, + { + "epoch": 9.71860589291782, + "grad_norm": 0.24636580049991608, + "learning_rate": 5.403173795587225e-05, + "loss": 0.0223, + "step": 50060 + }, + { + "epoch": 9.72054754623562, + "grad_norm": 0.14709728956222534, + "learning_rate": 5.4015256879214606e-05, + "loss": 0.028, + "step": 50070 + }, + { + "epoch": 9.72248919955342, + "grad_norm": 0.27801474928855896, + "learning_rate": 5.39987753634545e-05, + "loss": 0.0234, + "step": 50080 + }, + { + "epoch": 9.72443085287122, + "grad_norm": 0.13438469171524048, + "learning_rate": 5.3982293410394335e-05, + "loss": 0.0222, + "step": 50090 + }, + { + "epoch": 9.72637250618902, + "grad_norm": 0.16749398410320282, + "learning_rate": 5.396581102183655e-05, + "loss": 0.0262, + "step": 50100 + }, + { + "epoch": 9.72831415950682, + "grad_norm": 0.21490134298801422, + "learning_rate": 5.3949328199583634e-05, + "loss": 0.0257, + "step": 50110 + }, + { + "epoch": 9.73025581282462, + "grad_norm": 0.12141857296228409, + "learning_rate": 5.393284494543813e-05, + "loss": 0.0281, + "step": 50120 + }, + { + "epoch": 9.73219746614242, + "grad_norm": 0.15381349623203278, + "learning_rate": 5.391636126120262e-05, + "loss": 0.0274, + "step": 50130 + }, + { + "epoch": 9.73413911946022, + "grad_norm": 0.23192401230335236, + "learning_rate": 5.389987714867971e-05, + "loss": 0.0236, + "step": 50140 + }, + { + "epoch": 9.73608077277802, + "grad_norm": 0.1461729109287262, + "learning_rate": 5.3883392609672114e-05, + "loss": 0.0198, + "step": 50150 + }, + { + "epoch": 9.73802242609582, + "grad_norm": 0.16483108699321747, + "learning_rate": 5.3866907645982545e-05, + "loss": 0.0237, + "step": 50160 + }, + { + "epoch": 9.739964079413621, + "grad_norm": 0.20845037698745728, + "learning_rate": 5.385042225941375e-05, + "loss": 0.025, + "step": 50170 + }, + { + "epoch": 9.741905732731421, + "grad_norm": 0.18595798313617706, + "learning_rate": 5.3833936451768585e-05, + "loss": 0.0194, + "step": 50180 + }, + { + "epoch": 9.743847386049222, + "grad_norm": 0.18177703022956848, + "learning_rate": 5.3817450224849884e-05, + "loss": 0.0223, + "step": 50190 + }, + { + "epoch": 9.745789039367022, + "grad_norm": 0.1344742774963379, + "learning_rate": 5.380096358046056e-05, + "loss": 0.0271, + "step": 50200 + }, + { + "epoch": 9.74773069268482, + "grad_norm": 0.16825246810913086, + "learning_rate": 5.378447652040359e-05, + "loss": 0.0297, + "step": 50210 + }, + { + "epoch": 9.74967234600262, + "grad_norm": 0.2680754065513611, + "learning_rate": 5.376798904648195e-05, + "loss": 0.033, + "step": 50220 + }, + { + "epoch": 9.751613999320421, + "grad_norm": 0.32169049978256226, + "learning_rate": 5.375150116049869e-05, + "loss": 0.0377, + "step": 50230 + }, + { + "epoch": 9.753555652638221, + "grad_norm": 0.1570441871881485, + "learning_rate": 5.373501286425691e-05, + "loss": 0.0244, + "step": 50240 + }, + { + "epoch": 9.755497305956021, + "grad_norm": 0.13560453057289124, + "learning_rate": 5.3718524159559726e-05, + "loss": 0.023, + "step": 50250 + }, + { + "epoch": 9.757438959273822, + "grad_norm": 0.18500612676143646, + "learning_rate": 5.370203504821034e-05, + "loss": 0.0228, + "step": 50260 + }, + { + "epoch": 9.759380612591622, + "grad_norm": 0.16394881904125214, + "learning_rate": 5.368554553201196e-05, + "loss": 0.0228, + "step": 50270 + }, + { + "epoch": 9.761322265909422, + "grad_norm": 0.19625037908554077, + "learning_rate": 5.3669055612767874e-05, + "loss": 0.0229, + "step": 50280 + }, + { + "epoch": 9.763263919227223, + "grad_norm": 0.09883078187704086, + "learning_rate": 5.3652565292281376e-05, + "loss": 0.0311, + "step": 50290 + }, + { + "epoch": 9.765205572545023, + "grad_norm": 0.17065350711345673, + "learning_rate": 5.3636074572355824e-05, + "loss": 0.0317, + "step": 50300 + }, + { + "epoch": 9.767147225862821, + "grad_norm": 0.21706214547157288, + "learning_rate": 5.361958345479463e-05, + "loss": 0.0297, + "step": 50310 + }, + { + "epoch": 9.769088879180622, + "grad_norm": 0.281301349401474, + "learning_rate": 5.360309194140123e-05, + "loss": 0.0295, + "step": 50320 + }, + { + "epoch": 9.771030532498422, + "grad_norm": 0.15856817364692688, + "learning_rate": 5.3586600033979106e-05, + "loss": 0.0239, + "step": 50330 + }, + { + "epoch": 9.772972185816222, + "grad_norm": 0.14954982697963715, + "learning_rate": 5.357010773433181e-05, + "loss": 0.022, + "step": 50340 + }, + { + "epoch": 9.774913839134022, + "grad_norm": 0.1912699192762375, + "learning_rate": 5.3553615044262895e-05, + "loss": 0.022, + "step": 50350 + }, + { + "epoch": 9.776855492451823, + "grad_norm": 0.18571852147579193, + "learning_rate": 5.3537121965575964e-05, + "loss": 0.0225, + "step": 50360 + }, + { + "epoch": 9.778797145769623, + "grad_norm": 0.14995796978473663, + "learning_rate": 5.352062850007471e-05, + "loss": 0.0209, + "step": 50370 + }, + { + "epoch": 9.780738799087423, + "grad_norm": 0.19229891896247864, + "learning_rate": 5.350413464956282e-05, + "loss": 0.0231, + "step": 50380 + }, + { + "epoch": 9.782680452405224, + "grad_norm": 0.1747620701789856, + "learning_rate": 5.348764041584403e-05, + "loss": 0.0226, + "step": 50390 + }, + { + "epoch": 9.784622105723024, + "grad_norm": 0.23852691054344177, + "learning_rate": 5.347114580072212e-05, + "loss": 0.0266, + "step": 50400 + }, + { + "epoch": 9.786563759040824, + "grad_norm": 0.27821680903434753, + "learning_rate": 5.345465080600093e-05, + "loss": 0.0324, + "step": 50410 + }, + { + "epoch": 9.788505412358623, + "grad_norm": 0.2653883397579193, + "learning_rate": 5.3438155433484326e-05, + "loss": 0.0185, + "step": 50420 + }, + { + "epoch": 9.790447065676423, + "grad_norm": 0.15362127125263214, + "learning_rate": 5.3421659684976197e-05, + "loss": 0.0221, + "step": 50430 + }, + { + "epoch": 9.792388718994223, + "grad_norm": 0.22901636362075806, + "learning_rate": 5.340516356228052e-05, + "loss": 0.0227, + "step": 50440 + }, + { + "epoch": 9.794330372312023, + "grad_norm": 0.14960356056690216, + "learning_rate": 5.338866706720128e-05, + "loss": 0.0259, + "step": 50450 + }, + { + "epoch": 9.796272025629824, + "grad_norm": 0.2821138799190521, + "learning_rate": 5.337217020154249e-05, + "loss": 0.0255, + "step": 50460 + }, + { + "epoch": 9.798213678947624, + "grad_norm": 0.2545718252658844, + "learning_rate": 5.335567296710825e-05, + "loss": 0.025, + "step": 50470 + }, + { + "epoch": 9.800155332265424, + "grad_norm": 0.20143593847751617, + "learning_rate": 5.333917536570265e-05, + "loss": 0.0253, + "step": 50480 + }, + { + "epoch": 9.802096985583225, + "grad_norm": 0.11536460369825363, + "learning_rate": 5.332267739912986e-05, + "loss": 0.0202, + "step": 50490 + }, + { + "epoch": 9.804038638901025, + "grad_norm": 0.18332527577877045, + "learning_rate": 5.330617906919405e-05, + "loss": 0.0331, + "step": 50500 + }, + { + "epoch": 9.805980292218825, + "grad_norm": 0.22747348248958588, + "learning_rate": 5.328968037769949e-05, + "loss": 0.0239, + "step": 50510 + }, + { + "epoch": 9.807921945536624, + "grad_norm": 0.17527472972869873, + "learning_rate": 5.327318132645042e-05, + "loss": 0.0262, + "step": 50520 + }, + { + "epoch": 9.809863598854424, + "grad_norm": 0.2948116362094879, + "learning_rate": 5.325668191725116e-05, + "loss": 0.0293, + "step": 50530 + }, + { + "epoch": 9.811805252172224, + "grad_norm": 0.14818443357944489, + "learning_rate": 5.324018215190606e-05, + "loss": 0.0264, + "step": 50540 + }, + { + "epoch": 9.813746905490024, + "grad_norm": 0.2140810340642929, + "learning_rate": 5.3223682032219515e-05, + "loss": 0.0297, + "step": 50550 + }, + { + "epoch": 9.815688558807825, + "grad_norm": 0.13876526057720184, + "learning_rate": 5.320718155999595e-05, + "loss": 0.0251, + "step": 50560 + }, + { + "epoch": 9.817630212125625, + "grad_norm": 0.142886683344841, + "learning_rate": 5.3190680737039835e-05, + "loss": 0.0259, + "step": 50570 + }, + { + "epoch": 9.819571865443425, + "grad_norm": 0.14664216339588165, + "learning_rate": 5.317417956515567e-05, + "loss": 0.0201, + "step": 50580 + }, + { + "epoch": 9.821513518761225, + "grad_norm": 0.17417533695697784, + "learning_rate": 5.315767804614798e-05, + "loss": 0.0273, + "step": 50590 + }, + { + "epoch": 9.823455172079026, + "grad_norm": 0.18556030094623566, + "learning_rate": 5.3141176181821395e-05, + "loss": 0.024, + "step": 50600 + }, + { + "epoch": 9.825396825396826, + "grad_norm": 0.18712200224399567, + "learning_rate": 5.31246739739805e-05, + "loss": 0.0283, + "step": 50610 + }, + { + "epoch": 9.827338478714626, + "grad_norm": 0.09641329199075699, + "learning_rate": 5.310817142442994e-05, + "loss": 0.0267, + "step": 50620 + }, + { + "epoch": 9.829280132032425, + "grad_norm": 0.1836899071931839, + "learning_rate": 5.309166853497445e-05, + "loss": 0.0299, + "step": 50630 + }, + { + "epoch": 9.831221785350225, + "grad_norm": 0.30010464787483215, + "learning_rate": 5.307516530741873e-05, + "loss": 0.0294, + "step": 50640 + }, + { + "epoch": 9.833163438668025, + "grad_norm": 0.15775971114635468, + "learning_rate": 5.305866174356754e-05, + "loss": 0.0219, + "step": 50650 + }, + { + "epoch": 9.835105091985826, + "grad_norm": 0.16308552026748657, + "learning_rate": 5.304215784522571e-05, + "loss": 0.0271, + "step": 50660 + }, + { + "epoch": 9.837046745303626, + "grad_norm": 0.15432341396808624, + "learning_rate": 5.302565361419808e-05, + "loss": 0.0241, + "step": 50670 + }, + { + "epoch": 9.838988398621426, + "grad_norm": 0.17623864114284515, + "learning_rate": 5.3009149052289507e-05, + "loss": 0.0221, + "step": 50680 + }, + { + "epoch": 9.840930051939226, + "grad_norm": 0.20969361066818237, + "learning_rate": 5.299264416130493e-05, + "loss": 0.0244, + "step": 50690 + }, + { + "epoch": 9.842871705257027, + "grad_norm": 0.18121355772018433, + "learning_rate": 5.297613894304928e-05, + "loss": 0.026, + "step": 50700 + }, + { + "epoch": 9.844813358574827, + "grad_norm": 0.17253640294075012, + "learning_rate": 5.2959633399327534e-05, + "loss": 0.0241, + "step": 50710 + }, + { + "epoch": 9.846755011892627, + "grad_norm": 0.11667514592409134, + "learning_rate": 5.294312753194476e-05, + "loss": 0.0223, + "step": 50720 + }, + { + "epoch": 9.848696665210426, + "grad_norm": 0.22999204695224762, + "learning_rate": 5.292662134270596e-05, + "loss": 0.0247, + "step": 50730 + }, + { + "epoch": 9.850638318528226, + "grad_norm": 0.295337438583374, + "learning_rate": 5.291011483341626e-05, + "loss": 0.0314, + "step": 50740 + }, + { + "epoch": 9.852579971846026, + "grad_norm": 0.09504961967468262, + "learning_rate": 5.2893608005880767e-05, + "loss": 0.0276, + "step": 50750 + }, + { + "epoch": 9.854521625163827, + "grad_norm": 0.1256539225578308, + "learning_rate": 5.287710086190467e-05, + "loss": 0.0258, + "step": 50760 + }, + { + "epoch": 9.856463278481627, + "grad_norm": 0.28815028071403503, + "learning_rate": 5.286059340329314e-05, + "loss": 0.0325, + "step": 50770 + }, + { + "epoch": 9.858404931799427, + "grad_norm": 0.17904792726039886, + "learning_rate": 5.284408563185141e-05, + "loss": 0.0241, + "step": 50780 + }, + { + "epoch": 9.860346585117227, + "grad_norm": 0.17398548126220703, + "learning_rate": 5.282757754938476e-05, + "loss": 0.0219, + "step": 50790 + }, + { + "epoch": 9.862288238435028, + "grad_norm": 0.183556467294693, + "learning_rate": 5.281106915769849e-05, + "loss": 0.0221, + "step": 50800 + }, + { + "epoch": 9.864229891752828, + "grad_norm": 0.20484906435012817, + "learning_rate": 5.2794560458597897e-05, + "loss": 0.0311, + "step": 50810 + }, + { + "epoch": 9.866171545070628, + "grad_norm": 0.10251286625862122, + "learning_rate": 5.27780514538884e-05, + "loss": 0.0209, + "step": 50820 + }, + { + "epoch": 9.868113198388428, + "grad_norm": 0.11062789708375931, + "learning_rate": 5.2761542145375365e-05, + "loss": 0.0227, + "step": 50830 + }, + { + "epoch": 9.870054851706229, + "grad_norm": 0.2121044248342514, + "learning_rate": 5.274503253486421e-05, + "loss": 0.0259, + "step": 50840 + }, + { + "epoch": 9.871996505024027, + "grad_norm": 0.09744908660650253, + "learning_rate": 5.272852262416046e-05, + "loss": 0.0331, + "step": 50850 + }, + { + "epoch": 9.873938158341828, + "grad_norm": 0.15168219804763794, + "learning_rate": 5.2712012415069555e-05, + "loss": 0.0271, + "step": 50860 + }, + { + "epoch": 9.875879811659628, + "grad_norm": 0.21021509170532227, + "learning_rate": 5.269550190939705e-05, + "loss": 0.0301, + "step": 50870 + }, + { + "epoch": 9.877821464977428, + "grad_norm": 0.17838984727859497, + "learning_rate": 5.267899110894852e-05, + "loss": 0.036, + "step": 50880 + }, + { + "epoch": 9.879763118295228, + "grad_norm": 0.19834275543689728, + "learning_rate": 5.266248001552955e-05, + "loss": 0.0256, + "step": 50890 + }, + { + "epoch": 9.881704771613029, + "grad_norm": 0.23399919271469116, + "learning_rate": 5.264596863094575e-05, + "loss": 0.0299, + "step": 50900 + }, + { + "epoch": 9.883646424930829, + "grad_norm": 0.11973799020051956, + "learning_rate": 5.262945695700282e-05, + "loss": 0.0232, + "step": 50910 + }, + { + "epoch": 9.88558807824863, + "grad_norm": 0.22377359867095947, + "learning_rate": 5.261294499550643e-05, + "loss": 0.0351, + "step": 50920 + }, + { + "epoch": 9.88752973156643, + "grad_norm": 0.149162158370018, + "learning_rate": 5.25964327482623e-05, + "loss": 0.0242, + "step": 50930 + }, + { + "epoch": 9.88947138488423, + "grad_norm": 0.11091526597738266, + "learning_rate": 5.257992021707617e-05, + "loss": 0.0243, + "step": 50940 + }, + { + "epoch": 9.891413038202028, + "grad_norm": 0.18883149325847626, + "learning_rate": 5.256340740375387e-05, + "loss": 0.0238, + "step": 50950 + }, + { + "epoch": 9.893354691519828, + "grad_norm": 0.3536356985569, + "learning_rate": 5.254689431010117e-05, + "loss": 0.0236, + "step": 50960 + }, + { + "epoch": 9.895296344837629, + "grad_norm": 0.2049991488456726, + "learning_rate": 5.253038093792395e-05, + "loss": 0.0226, + "step": 50970 + }, + { + "epoch": 9.897237998155429, + "grad_norm": 0.17206327617168427, + "learning_rate": 5.251386728902806e-05, + "loss": 0.0225, + "step": 50980 + }, + { + "epoch": 9.89917965147323, + "grad_norm": 0.15273410081863403, + "learning_rate": 5.2497353365219446e-05, + "loss": 0.0246, + "step": 50990 + }, + { + "epoch": 9.90112130479103, + "grad_norm": 0.1301601231098175, + "learning_rate": 5.2480839168304e-05, + "loss": 0.0272, + "step": 51000 + }, + { + "epoch": 9.90306295810883, + "grad_norm": 0.2162867784500122, + "learning_rate": 5.246432470008773e-05, + "loss": 0.0293, + "step": 51010 + }, + { + "epoch": 9.90500461142663, + "grad_norm": 0.10492607206106186, + "learning_rate": 5.2447809962376616e-05, + "loss": 0.0289, + "step": 51020 + }, + { + "epoch": 9.90694626474443, + "grad_norm": 0.14687176048755646, + "learning_rate": 5.243129495697666e-05, + "loss": 0.0318, + "step": 51030 + }, + { + "epoch": 9.90888791806223, + "grad_norm": 0.11780894547700882, + "learning_rate": 5.241477968569397e-05, + "loss": 0.0312, + "step": 51040 + }, + { + "epoch": 9.910829571380031, + "grad_norm": 0.17400714755058289, + "learning_rate": 5.23982641503346e-05, + "loss": 0.0182, + "step": 51050 + }, + { + "epoch": 9.91277122469783, + "grad_norm": 0.11702916026115417, + "learning_rate": 5.238174835270465e-05, + "loss": 0.0327, + "step": 51060 + }, + { + "epoch": 9.91471287801563, + "grad_norm": 0.13976052403450012, + "learning_rate": 5.23652322946103e-05, + "loss": 0.0321, + "step": 51070 + }, + { + "epoch": 9.91665453133343, + "grad_norm": 0.14566048979759216, + "learning_rate": 5.234871597785771e-05, + "loss": 0.0288, + "step": 51080 + }, + { + "epoch": 9.91859618465123, + "grad_norm": 0.09198077023029327, + "learning_rate": 5.233219940425306e-05, + "loss": 0.0265, + "step": 51090 + }, + { + "epoch": 9.92053783796903, + "grad_norm": 0.15705065429210663, + "learning_rate": 5.2315682575602576e-05, + "loss": 0.0253, + "step": 51100 + }, + { + "epoch": 9.92247949128683, + "grad_norm": 0.15053679049015045, + "learning_rate": 5.229916549371254e-05, + "loss": 0.0238, + "step": 51110 + }, + { + "epoch": 9.924421144604631, + "grad_norm": 0.1456732600927353, + "learning_rate": 5.228264816038924e-05, + "loss": 0.0293, + "step": 51120 + }, + { + "epoch": 9.926362797922431, + "grad_norm": 0.13294629752635956, + "learning_rate": 5.2266130577438935e-05, + "loss": 0.0261, + "step": 51130 + }, + { + "epoch": 9.928304451240232, + "grad_norm": 0.22715474665164948, + "learning_rate": 5.224961274666802e-05, + "loss": 0.0271, + "step": 51140 + }, + { + "epoch": 9.930246104558032, + "grad_norm": 0.19859521090984344, + "learning_rate": 5.223309466988283e-05, + "loss": 0.0268, + "step": 51150 + }, + { + "epoch": 9.93218775787583, + "grad_norm": 0.1736355870962143, + "learning_rate": 5.221657634888976e-05, + "loss": 0.0306, + "step": 51160 + }, + { + "epoch": 9.93412941119363, + "grad_norm": 0.20586873590946198, + "learning_rate": 5.2200057785495235e-05, + "loss": 0.0235, + "step": 51170 + }, + { + "epoch": 9.936071064511431, + "grad_norm": 0.09036766737699509, + "learning_rate": 5.2183538981505706e-05, + "loss": 0.0231, + "step": 51180 + }, + { + "epoch": 9.938012717829231, + "grad_norm": 0.23842936754226685, + "learning_rate": 5.216701993872762e-05, + "loss": 0.0284, + "step": 51190 + }, + { + "epoch": 9.939954371147032, + "grad_norm": 0.10383347421884537, + "learning_rate": 5.215050065896752e-05, + "loss": 0.0227, + "step": 51200 + }, + { + "epoch": 9.941896024464832, + "grad_norm": 0.1392977088689804, + "learning_rate": 5.2133981144031875e-05, + "loss": 0.0255, + "step": 51210 + }, + { + "epoch": 9.943837677782632, + "grad_norm": 0.21853093802928925, + "learning_rate": 5.211746139572727e-05, + "loss": 0.0248, + "step": 51220 + }, + { + "epoch": 9.945779331100432, + "grad_norm": 0.10772169381380081, + "learning_rate": 5.210094141586026e-05, + "loss": 0.0243, + "step": 51230 + }, + { + "epoch": 9.947720984418233, + "grad_norm": 0.1341540664434433, + "learning_rate": 5.2084421206237476e-05, + "loss": 0.0251, + "step": 51240 + }, + { + "epoch": 9.949662637736033, + "grad_norm": 0.24796636402606964, + "learning_rate": 5.20679007686655e-05, + "loss": 0.0226, + "step": 51250 + }, + { + "epoch": 9.951604291053833, + "grad_norm": 0.31407874822616577, + "learning_rate": 5.205138010495102e-05, + "loss": 0.0311, + "step": 51260 + }, + { + "epoch": 9.953545944371632, + "grad_norm": 0.2171928584575653, + "learning_rate": 5.20348592169007e-05, + "loss": 0.023, + "step": 51270 + }, + { + "epoch": 9.955487597689432, + "grad_norm": 0.12402195483446121, + "learning_rate": 5.201833810632123e-05, + "loss": 0.0265, + "step": 51280 + }, + { + "epoch": 9.957429251007232, + "grad_norm": 0.15069639682769775, + "learning_rate": 5.200181677501933e-05, + "loss": 0.0182, + "step": 51290 + }, + { + "epoch": 9.959370904325032, + "grad_norm": 0.17736808955669403, + "learning_rate": 5.1985295224801777e-05, + "loss": 0.0288, + "step": 51300 + }, + { + "epoch": 9.961312557642833, + "grad_norm": 0.17545872926712036, + "learning_rate": 5.196877345747531e-05, + "loss": 0.0228, + "step": 51310 + }, + { + "epoch": 9.963254210960633, + "grad_norm": 0.12622074782848358, + "learning_rate": 5.195225147484674e-05, + "loss": 0.0257, + "step": 51320 + }, + { + "epoch": 9.965195864278433, + "grad_norm": 0.08760702610015869, + "learning_rate": 5.193572927872288e-05, + "loss": 0.0211, + "step": 51330 + }, + { + "epoch": 9.967137517596234, + "grad_norm": 0.20418278872966766, + "learning_rate": 5.191920687091059e-05, + "loss": 0.0271, + "step": 51340 + }, + { + "epoch": 9.969079170914034, + "grad_norm": 0.2045222520828247, + "learning_rate": 5.1902684253216695e-05, + "loss": 0.0215, + "step": 51350 + }, + { + "epoch": 9.971020824231834, + "grad_norm": 0.195500448346138, + "learning_rate": 5.1886161427448134e-05, + "loss": 0.0245, + "step": 51360 + }, + { + "epoch": 9.972962477549633, + "grad_norm": 0.13092565536499023, + "learning_rate": 5.186963839541179e-05, + "loss": 0.0287, + "step": 51370 + }, + { + "epoch": 9.974904130867433, + "grad_norm": 0.14680984616279602, + "learning_rate": 5.185311515891459e-05, + "loss": 0.0263, + "step": 51380 + }, + { + "epoch": 9.976845784185233, + "grad_norm": 0.15116803348064423, + "learning_rate": 5.1836591719763516e-05, + "loss": 0.0288, + "step": 51390 + }, + { + "epoch": 9.978787437503033, + "grad_norm": 0.2562921643257141, + "learning_rate": 5.182006807976554e-05, + "loss": 0.0276, + "step": 51400 + }, + { + "epoch": 9.980729090820834, + "grad_norm": 0.17401398718357086, + "learning_rate": 5.180354424072763e-05, + "loss": 0.0259, + "step": 51410 + }, + { + "epoch": 9.982670744138634, + "grad_norm": 0.2006758451461792, + "learning_rate": 5.178702020445686e-05, + "loss": 0.0213, + "step": 51420 + }, + { + "epoch": 9.984612397456434, + "grad_norm": 0.1822323501110077, + "learning_rate": 5.177049597276024e-05, + "loss": 0.0225, + "step": 51430 + }, + { + "epoch": 9.986554050774235, + "grad_norm": 0.21852196753025055, + "learning_rate": 5.175397154744482e-05, + "loss": 0.0243, + "step": 51440 + }, + { + "epoch": 9.988495704092035, + "grad_norm": 0.1453770250082016, + "learning_rate": 5.1737446930317736e-05, + "loss": 0.0296, + "step": 51450 + }, + { + "epoch": 9.990437357409835, + "grad_norm": 0.16980963945388794, + "learning_rate": 5.172092212318604e-05, + "loss": 0.025, + "step": 51460 + }, + { + "epoch": 9.992379010727635, + "grad_norm": 0.1355520337820053, + "learning_rate": 5.170439712785692e-05, + "loss": 0.0331, + "step": 51470 + }, + { + "epoch": 9.994320664045436, + "grad_norm": 0.2429930567741394, + "learning_rate": 5.1687871946137467e-05, + "loss": 0.0278, + "step": 51480 + }, + { + "epoch": 9.996262317363234, + "grad_norm": 0.18491294980049133, + "learning_rate": 5.16713465798349e-05, + "loss": 0.0253, + "step": 51490 + }, + { + "epoch": 9.998203970681034, + "grad_norm": 0.15400578081607819, + "learning_rate": 5.165482103075637e-05, + "loss": 0.0225, + "step": 51500 + }, + { + "epoch": 10.0, + "grad_norm": 0.05598721653223038, + "learning_rate": 5.163829530070909e-05, + "loss": 0.0227, + "step": 51510 + }, + { + "epoch": 10.0019416533178, + "grad_norm": 0.19615451991558075, + "learning_rate": 5.162176939150032e-05, + "loss": 0.0216, + "step": 51520 + }, + { + "epoch": 10.0038833066356, + "grad_norm": 0.18824104964733124, + "learning_rate": 5.160524330493728e-05, + "loss": 0.0251, + "step": 51530 + }, + { + "epoch": 10.0058249599534, + "grad_norm": 0.15071408450603485, + "learning_rate": 5.158871704282724e-05, + "loss": 0.0224, + "step": 51540 + }, + { + "epoch": 10.007766613271201, + "grad_norm": 0.29105034470558167, + "learning_rate": 5.15721906069775e-05, + "loss": 0.0274, + "step": 51550 + }, + { + "epoch": 10.009708266589001, + "grad_norm": 0.23438507318496704, + "learning_rate": 5.155566399919536e-05, + "loss": 0.0301, + "step": 51560 + }, + { + "epoch": 10.0116499199068, + "grad_norm": 0.13639160990715027, + "learning_rate": 5.153913722128813e-05, + "loss": 0.0297, + "step": 51570 + }, + { + "epoch": 10.0135915732246, + "grad_norm": 0.17015783488750458, + "learning_rate": 5.1522610275063176e-05, + "loss": 0.0258, + "step": 51580 + }, + { + "epoch": 10.0155332265424, + "grad_norm": 0.11101607233285904, + "learning_rate": 5.150608316232786e-05, + "loss": 0.0229, + "step": 51590 + }, + { + "epoch": 10.0174748798602, + "grad_norm": 0.0680772066116333, + "learning_rate": 5.148955588488955e-05, + "loss": 0.0247, + "step": 51600 + }, + { + "epoch": 10.019416533178001, + "grad_norm": 0.1816963404417038, + "learning_rate": 5.147302844455564e-05, + "loss": 0.0231, + "step": 51610 + }, + { + "epoch": 10.021358186495801, + "grad_norm": 0.19409726560115814, + "learning_rate": 5.145650084313356e-05, + "loss": 0.0216, + "step": 51620 + }, + { + "epoch": 10.023299839813602, + "grad_norm": 0.13092616200447083, + "learning_rate": 5.1439973082430734e-05, + "loss": 0.0215, + "step": 51630 + }, + { + "epoch": 10.025241493131402, + "grad_norm": 0.20098817348480225, + "learning_rate": 5.142344516425461e-05, + "loss": 0.0231, + "step": 51640 + }, + { + "epoch": 10.027183146449202, + "grad_norm": 0.11725042760372162, + "learning_rate": 5.1406917090412673e-05, + "loss": 0.0237, + "step": 51650 + }, + { + "epoch": 10.029124799767002, + "grad_norm": 0.18855975568294525, + "learning_rate": 5.1390388862712404e-05, + "loss": 0.0235, + "step": 51660 + }, + { + "epoch": 10.0310664530848, + "grad_norm": 0.16372200846672058, + "learning_rate": 5.137386048296129e-05, + "loss": 0.0276, + "step": 51670 + }, + { + "epoch": 10.033008106402601, + "grad_norm": 0.20253099501132965, + "learning_rate": 5.135733195296686e-05, + "loss": 0.0339, + "step": 51680 + }, + { + "epoch": 10.034949759720401, + "grad_norm": 0.16372698545455933, + "learning_rate": 5.134080327453664e-05, + "loss": 0.0271, + "step": 51690 + }, + { + "epoch": 10.036891413038202, + "grad_norm": 0.1852947324514389, + "learning_rate": 5.1324274449478204e-05, + "loss": 0.0271, + "step": 51700 + }, + { + "epoch": 10.038833066356002, + "grad_norm": 0.20602038502693176, + "learning_rate": 5.130774547959909e-05, + "loss": 0.0266, + "step": 51710 + }, + { + "epoch": 10.040774719673802, + "grad_norm": 0.09783291816711426, + "learning_rate": 5.129121636670692e-05, + "loss": 0.0297, + "step": 51720 + }, + { + "epoch": 10.042716372991602, + "grad_norm": 0.21204859018325806, + "learning_rate": 5.1274687112609246e-05, + "loss": 0.0257, + "step": 51730 + }, + { + "epoch": 10.044658026309403, + "grad_norm": 0.15392707288265228, + "learning_rate": 5.125815771911372e-05, + "loss": 0.0224, + "step": 51740 + }, + { + "epoch": 10.046599679627203, + "grad_norm": 0.12463170289993286, + "learning_rate": 5.1241628188027956e-05, + "loss": 0.0262, + "step": 51750 + }, + { + "epoch": 10.048541332945003, + "grad_norm": 0.23447605967521667, + "learning_rate": 5.1225098521159585e-05, + "loss": 0.0261, + "step": 51760 + }, + { + "epoch": 10.050482986262804, + "grad_norm": 0.14699827134609222, + "learning_rate": 5.12085687203163e-05, + "loss": 0.027, + "step": 51770 + }, + { + "epoch": 10.052424639580602, + "grad_norm": 0.2044990062713623, + "learning_rate": 5.119203878730575e-05, + "loss": 0.0281, + "step": 51780 + }, + { + "epoch": 10.054366292898402, + "grad_norm": 0.16936273872852325, + "learning_rate": 5.117550872393564e-05, + "loss": 0.0244, + "step": 51790 + }, + { + "epoch": 10.056307946216203, + "grad_norm": 0.18033628165721893, + "learning_rate": 5.1158978532013645e-05, + "loss": 0.0223, + "step": 51800 + }, + { + "epoch": 10.058249599534003, + "grad_norm": 0.12156856060028076, + "learning_rate": 5.114244821334751e-05, + "loss": 0.0263, + "step": 51810 + }, + { + "epoch": 10.060191252851803, + "grad_norm": 0.22705847024917603, + "learning_rate": 5.112591776974497e-05, + "loss": 0.0296, + "step": 51820 + }, + { + "epoch": 10.062132906169603, + "grad_norm": 0.14037632942199707, + "learning_rate": 5.110938720301374e-05, + "loss": 0.0311, + "step": 51830 + }, + { + "epoch": 10.064074559487404, + "grad_norm": 0.1791774332523346, + "learning_rate": 5.1092856514961595e-05, + "loss": 0.0281, + "step": 51840 + }, + { + "epoch": 10.066016212805204, + "grad_norm": 0.21413253247737885, + "learning_rate": 5.107632570739631e-05, + "loss": 0.0194, + "step": 51850 + }, + { + "epoch": 10.067957866123004, + "grad_norm": 0.142309308052063, + "learning_rate": 5.1059794782125645e-05, + "loss": 0.0198, + "step": 51860 + }, + { + "epoch": 10.069899519440805, + "grad_norm": 0.20498089492321014, + "learning_rate": 5.104326374095744e-05, + "loss": 0.0231, + "step": 51870 + }, + { + "epoch": 10.071841172758605, + "grad_norm": 0.1671755611896515, + "learning_rate": 5.102673258569947e-05, + "loss": 0.019, + "step": 51880 + }, + { + "epoch": 10.073782826076403, + "grad_norm": 0.19177168607711792, + "learning_rate": 5.101020131815956e-05, + "loss": 0.0251, + "step": 51890 + }, + { + "epoch": 10.075724479394204, + "grad_norm": 0.23368923366069794, + "learning_rate": 5.099366994014556e-05, + "loss": 0.0346, + "step": 51900 + }, + { + "epoch": 10.077666132712004, + "grad_norm": 0.22178956866264343, + "learning_rate": 5.097713845346532e-05, + "loss": 0.0243, + "step": 51910 + }, + { + "epoch": 10.079607786029804, + "grad_norm": 0.10468930006027222, + "learning_rate": 5.096060685992668e-05, + "loss": 0.0212, + "step": 51920 + }, + { + "epoch": 10.081549439347604, + "grad_norm": 0.2129671722650528, + "learning_rate": 5.0944075161337514e-05, + "loss": 0.0209, + "step": 51930 + }, + { + "epoch": 10.083491092665405, + "grad_norm": 0.16852857172489166, + "learning_rate": 5.09275433595057e-05, + "loss": 0.0229, + "step": 51940 + }, + { + "epoch": 10.085432745983205, + "grad_norm": 0.16637912392616272, + "learning_rate": 5.0911011456239157e-05, + "loss": 0.0342, + "step": 51950 + }, + { + "epoch": 10.087374399301005, + "grad_norm": 0.16449786722660065, + "learning_rate": 5.089447945334574e-05, + "loss": 0.0201, + "step": 51960 + }, + { + "epoch": 10.089316052618805, + "grad_norm": 0.1627381145954132, + "learning_rate": 5.087794735263343e-05, + "loss": 0.0212, + "step": 51970 + }, + { + "epoch": 10.091257705936606, + "grad_norm": 0.1611672043800354, + "learning_rate": 5.086141515591011e-05, + "loss": 0.0258, + "step": 51980 + }, + { + "epoch": 10.093199359254404, + "grad_norm": 0.2062760889530182, + "learning_rate": 5.084488286498371e-05, + "loss": 0.026, + "step": 51990 + }, + { + "epoch": 10.095141012572205, + "grad_norm": 0.25026118755340576, + "learning_rate": 5.08283504816622e-05, + "loss": 0.0268, + "step": 52000 + }, + { + "epoch": 10.097082665890005, + "grad_norm": 0.14410673081874847, + "learning_rate": 5.081181800775353e-05, + "loss": 0.0258, + "step": 52010 + }, + { + "epoch": 10.099024319207805, + "grad_norm": 0.2756834626197815, + "learning_rate": 5.079528544506565e-05, + "loss": 0.0245, + "step": 52020 + }, + { + "epoch": 10.100965972525605, + "grad_norm": 0.17021480202674866, + "learning_rate": 5.077875279540657e-05, + "loss": 0.021, + "step": 52030 + }, + { + "epoch": 10.102907625843406, + "grad_norm": 0.15931329131126404, + "learning_rate": 5.076222006058424e-05, + "loss": 0.0252, + "step": 52040 + }, + { + "epoch": 10.104849279161206, + "grad_norm": 0.19132813811302185, + "learning_rate": 5.074568724240667e-05, + "loss": 0.0279, + "step": 52050 + }, + { + "epoch": 10.106790932479006, + "grad_norm": 0.19666671752929688, + "learning_rate": 5.072915434268187e-05, + "loss": 0.0311, + "step": 52060 + }, + { + "epoch": 10.108732585796806, + "grad_norm": 0.22790853679180145, + "learning_rate": 5.071262136321785e-05, + "loss": 0.027, + "step": 52070 + }, + { + "epoch": 10.110674239114607, + "grad_norm": 0.18591049313545227, + "learning_rate": 5.069608830582262e-05, + "loss": 0.0228, + "step": 52080 + }, + { + "epoch": 10.112615892432407, + "grad_norm": 0.13218937814235687, + "learning_rate": 5.067955517230423e-05, + "loss": 0.0242, + "step": 52090 + }, + { + "epoch": 10.114557545750205, + "grad_norm": 0.22328083217144012, + "learning_rate": 5.06630219644707e-05, + "loss": 0.0249, + "step": 52100 + }, + { + "epoch": 10.116499199068006, + "grad_norm": 0.2150278091430664, + "learning_rate": 5.064648868413008e-05, + "loss": 0.0189, + "step": 52110 + }, + { + "epoch": 10.118440852385806, + "grad_norm": 0.13739070296287537, + "learning_rate": 5.062995533309044e-05, + "loss": 0.0216, + "step": 52120 + }, + { + "epoch": 10.120382505703606, + "grad_norm": 0.16245317459106445, + "learning_rate": 5.061342191315983e-05, + "loss": 0.0289, + "step": 52130 + }, + { + "epoch": 10.122324159021407, + "grad_norm": 0.1889868974685669, + "learning_rate": 5.0596888426146325e-05, + "loss": 0.0225, + "step": 52140 + }, + { + "epoch": 10.124265812339207, + "grad_norm": 0.1180649995803833, + "learning_rate": 5.058035487385798e-05, + "loss": 0.0186, + "step": 52150 + }, + { + "epoch": 10.126207465657007, + "grad_norm": 0.1232222393155098, + "learning_rate": 5.0563821258102915e-05, + "loss": 0.028, + "step": 52160 + }, + { + "epoch": 10.128149118974807, + "grad_norm": 0.10247425734996796, + "learning_rate": 5.05472875806892e-05, + "loss": 0.0323, + "step": 52170 + }, + { + "epoch": 10.130090772292608, + "grad_norm": 0.11457817256450653, + "learning_rate": 5.053075384342494e-05, + "loss": 0.0271, + "step": 52180 + }, + { + "epoch": 10.132032425610408, + "grad_norm": 0.22532393038272858, + "learning_rate": 5.0514220048118234e-05, + "loss": 0.0229, + "step": 52190 + }, + { + "epoch": 10.133974078928208, + "grad_norm": 0.1290034055709839, + "learning_rate": 5.049768619657721e-05, + "loss": 0.0259, + "step": 52200 + }, + { + "epoch": 10.135915732246007, + "grad_norm": 0.17566150426864624, + "learning_rate": 5.048115229060996e-05, + "loss": 0.0331, + "step": 52210 + }, + { + "epoch": 10.137857385563807, + "grad_norm": 0.2062370926141739, + "learning_rate": 5.046461833202462e-05, + "loss": 0.0237, + "step": 52220 + }, + { + "epoch": 10.139799038881607, + "grad_norm": 0.14965733885765076, + "learning_rate": 5.044808432262932e-05, + "loss": 0.0298, + "step": 52230 + }, + { + "epoch": 10.141740692199408, + "grad_norm": 0.08650621026754379, + "learning_rate": 5.043155026423219e-05, + "loss": 0.0253, + "step": 52240 + }, + { + "epoch": 10.143682345517208, + "grad_norm": 0.1320127695798874, + "learning_rate": 5.041501615864137e-05, + "loss": 0.0277, + "step": 52250 + }, + { + "epoch": 10.145623998835008, + "grad_norm": 0.11721818149089813, + "learning_rate": 5.039848200766502e-05, + "loss": 0.0203, + "step": 52260 + }, + { + "epoch": 10.147565652152808, + "grad_norm": 0.11422178894281387, + "learning_rate": 5.038194781311125e-05, + "loss": 0.0284, + "step": 52270 + }, + { + "epoch": 10.149507305470609, + "grad_norm": 0.16787005960941315, + "learning_rate": 5.0365413576788264e-05, + "loss": 0.0172, + "step": 52280 + }, + { + "epoch": 10.151448958788409, + "grad_norm": 0.15367555618286133, + "learning_rate": 5.0348879300504194e-05, + "loss": 0.0231, + "step": 52290 + }, + { + "epoch": 10.15339061210621, + "grad_norm": 0.13236302137374878, + "learning_rate": 5.033234498606721e-05, + "loss": 0.0297, + "step": 52300 + }, + { + "epoch": 10.155332265424008, + "grad_norm": 0.21927183866500854, + "learning_rate": 5.0315810635285476e-05, + "loss": 0.0254, + "step": 52310 + }, + { + "epoch": 10.157273918741808, + "grad_norm": 0.1314874291419983, + "learning_rate": 5.0299276249967164e-05, + "loss": 0.03, + "step": 52320 + }, + { + "epoch": 10.159215572059608, + "grad_norm": 0.2583661675453186, + "learning_rate": 5.0282741831920454e-05, + "loss": 0.0286, + "step": 52330 + }, + { + "epoch": 10.161157225377409, + "grad_norm": 0.2648869752883911, + "learning_rate": 5.026620738295351e-05, + "loss": 0.0228, + "step": 52340 + }, + { + "epoch": 10.163098878695209, + "grad_norm": 0.20521946251392365, + "learning_rate": 5.024967290487455e-05, + "loss": 0.0274, + "step": 52350 + }, + { + "epoch": 10.165040532013009, + "grad_norm": 0.17189525067806244, + "learning_rate": 5.023313839949172e-05, + "loss": 0.0276, + "step": 52360 + }, + { + "epoch": 10.16698218533081, + "grad_norm": 0.14121361076831818, + "learning_rate": 5.0216603868613224e-05, + "loss": 0.0266, + "step": 52370 + }, + { + "epoch": 10.16892383864861, + "grad_norm": 0.1806129515171051, + "learning_rate": 5.020006931404726e-05, + "loss": 0.0228, + "step": 52380 + }, + { + "epoch": 10.17086549196641, + "grad_norm": 0.1223258450627327, + "learning_rate": 5.0183534737602014e-05, + "loss": 0.0212, + "step": 52390 + }, + { + "epoch": 10.17280714528421, + "grad_norm": 0.20392456650733948, + "learning_rate": 5.0167000141085686e-05, + "loss": 0.0241, + "step": 52400 + }, + { + "epoch": 10.17474879860201, + "grad_norm": 0.13131295144557953, + "learning_rate": 5.015046552630647e-05, + "loss": 0.03, + "step": 52410 + }, + { + "epoch": 10.176690451919809, + "grad_norm": 0.08514326810836792, + "learning_rate": 5.013393089507258e-05, + "loss": 0.0222, + "step": 52420 + }, + { + "epoch": 10.17863210523761, + "grad_norm": 0.08725172281265259, + "learning_rate": 5.011739624919219e-05, + "loss": 0.0252, + "step": 52430 + }, + { + "epoch": 10.18057375855541, + "grad_norm": 0.2875996530056, + "learning_rate": 5.0100861590473545e-05, + "loss": 0.0272, + "step": 52440 + }, + { + "epoch": 10.18251541187321, + "grad_norm": 0.15158510208129883, + "learning_rate": 5.008432692072482e-05, + "loss": 0.0301, + "step": 52450 + }, + { + "epoch": 10.18445706519101, + "grad_norm": 0.24238227307796478, + "learning_rate": 5.006779224175423e-05, + "loss": 0.029, + "step": 52460 + }, + { + "epoch": 10.18639871850881, + "grad_norm": 0.13724325597286224, + "learning_rate": 5.005125755536998e-05, + "loss": 0.0261, + "step": 52470 + }, + { + "epoch": 10.18834037182661, + "grad_norm": 0.1650007665157318, + "learning_rate": 5.003472286338029e-05, + "loss": 0.0259, + "step": 52480 + }, + { + "epoch": 10.19028202514441, + "grad_norm": 0.09679479151964188, + "learning_rate": 5.001818816759335e-05, + "loss": 0.0226, + "step": 52490 + }, + { + "epoch": 10.192223678462211, + "grad_norm": 0.24841715395450592, + "learning_rate": 5.000165346981738e-05, + "loss": 0.0233, + "step": 52500 + }, + { + "epoch": 10.194165331780011, + "grad_norm": 0.14463436603546143, + "learning_rate": 4.9985118771860595e-05, + "loss": 0.0273, + "step": 52510 + }, + { + "epoch": 10.196106985097812, + "grad_norm": 0.2504689395427704, + "learning_rate": 4.996858407553119e-05, + "loss": 0.0271, + "step": 52520 + }, + { + "epoch": 10.19804863841561, + "grad_norm": 0.1371837705373764, + "learning_rate": 4.99520493826374e-05, + "loss": 0.0251, + "step": 52530 + }, + { + "epoch": 10.19999029173341, + "grad_norm": 0.14717473089694977, + "learning_rate": 4.9935514694987384e-05, + "loss": 0.0305, + "step": 52540 + }, + { + "epoch": 10.20193194505121, + "grad_norm": 0.2571682631969452, + "learning_rate": 4.991898001438941e-05, + "loss": 0.0226, + "step": 52550 + }, + { + "epoch": 10.203873598369011, + "grad_norm": 0.19061151146888733, + "learning_rate": 4.990244534265166e-05, + "loss": 0.0225, + "step": 52560 + }, + { + "epoch": 10.205815251686811, + "grad_norm": 0.06291787326335907, + "learning_rate": 4.988591068158234e-05, + "loss": 0.02, + "step": 52570 + }, + { + "epoch": 10.207756905004612, + "grad_norm": 0.2230106145143509, + "learning_rate": 4.986937603298965e-05, + "loss": 0.0292, + "step": 52580 + }, + { + "epoch": 10.209698558322412, + "grad_norm": 0.16128182411193848, + "learning_rate": 4.985284139868181e-05, + "loss": 0.0275, + "step": 52590 + }, + { + "epoch": 10.211640211640212, + "grad_norm": 0.12054084986448288, + "learning_rate": 4.9836306780467e-05, + "loss": 0.0248, + "step": 52600 + }, + { + "epoch": 10.213581864958012, + "grad_norm": 0.1452690064907074, + "learning_rate": 4.981977218015343e-05, + "loss": 0.023, + "step": 52610 + }, + { + "epoch": 10.215523518275813, + "grad_norm": 0.1392006278038025, + "learning_rate": 4.980323759954932e-05, + "loss": 0.024, + "step": 52620 + }, + { + "epoch": 10.217465171593611, + "grad_norm": 0.16257713735103607, + "learning_rate": 4.978670304046284e-05, + "loss": 0.0277, + "step": 52630 + }, + { + "epoch": 10.219406824911411, + "grad_norm": 0.2422265261411667, + "learning_rate": 4.9770168504702195e-05, + "loss": 0.0283, + "step": 52640 + }, + { + "epoch": 10.221348478229212, + "grad_norm": 0.1902187019586563, + "learning_rate": 4.975363399407559e-05, + "loss": 0.0218, + "step": 52650 + }, + { + "epoch": 10.223290131547012, + "grad_norm": 0.1540481150150299, + "learning_rate": 4.9737099510391194e-05, + "loss": 0.0332, + "step": 52660 + }, + { + "epoch": 10.225231784864812, + "grad_norm": 0.16646425426006317, + "learning_rate": 4.972056505545719e-05, + "loss": 0.0298, + "step": 52670 + }, + { + "epoch": 10.227173438182612, + "grad_norm": 0.1464407742023468, + "learning_rate": 4.9704030631081775e-05, + "loss": 0.0261, + "step": 52680 + }, + { + "epoch": 10.229115091500413, + "grad_norm": 0.17283977568149567, + "learning_rate": 4.968749623907313e-05, + "loss": 0.0215, + "step": 52690 + }, + { + "epoch": 10.231056744818213, + "grad_norm": 0.27787482738494873, + "learning_rate": 4.967096188123943e-05, + "loss": 0.0301, + "step": 52700 + }, + { + "epoch": 10.232998398136013, + "grad_norm": 0.21372035145759583, + "learning_rate": 4.965442755938884e-05, + "loss": 0.0273, + "step": 52710 + }, + { + "epoch": 10.234940051453814, + "grad_norm": 0.15455424785614014, + "learning_rate": 4.963789327532954e-05, + "loss": 0.0214, + "step": 52720 + }, + { + "epoch": 10.236881704771614, + "grad_norm": 0.17509761452674866, + "learning_rate": 4.962135903086968e-05, + "loss": 0.0223, + "step": 52730 + }, + { + "epoch": 10.238823358089412, + "grad_norm": 0.24899321794509888, + "learning_rate": 4.960482482781741e-05, + "loss": 0.0284, + "step": 52740 + }, + { + "epoch": 10.240765011407213, + "grad_norm": 0.22559766471385956, + "learning_rate": 4.958829066798092e-05, + "loss": 0.0258, + "step": 52750 + }, + { + "epoch": 10.242706664725013, + "grad_norm": 0.3101564943790436, + "learning_rate": 4.957175655316835e-05, + "loss": 0.0308, + "step": 52760 + }, + { + "epoch": 10.244648318042813, + "grad_norm": 0.3210146725177765, + "learning_rate": 4.9555222485187834e-05, + "loss": 0.0286, + "step": 52770 + }, + { + "epoch": 10.246589971360613, + "grad_norm": 0.1545766443014145, + "learning_rate": 4.953868846584752e-05, + "loss": 0.0201, + "step": 52780 + }, + { + "epoch": 10.248531624678414, + "grad_norm": 0.19898022711277008, + "learning_rate": 4.9522154496955544e-05, + "loss": 0.0239, + "step": 52790 + }, + { + "epoch": 10.250473277996214, + "grad_norm": 0.19055254757404327, + "learning_rate": 4.950562058032001e-05, + "loss": 0.0291, + "step": 52800 + }, + { + "epoch": 10.252414931314014, + "grad_norm": 0.36369067430496216, + "learning_rate": 4.948908671774909e-05, + "loss": 0.0224, + "step": 52810 + }, + { + "epoch": 10.254356584631815, + "grad_norm": 0.21718047559261322, + "learning_rate": 4.947255291105087e-05, + "loss": 0.0264, + "step": 52820 + }, + { + "epoch": 10.256298237949615, + "grad_norm": 0.2280074954032898, + "learning_rate": 4.945601916203348e-05, + "loss": 0.031, + "step": 52830 + }, + { + "epoch": 10.258239891267415, + "grad_norm": 0.22289608418941498, + "learning_rate": 4.9439485472505016e-05, + "loss": 0.0235, + "step": 52840 + }, + { + "epoch": 10.260181544585214, + "grad_norm": 0.19208456575870514, + "learning_rate": 4.942295184427356e-05, + "loss": 0.0241, + "step": 52850 + }, + { + "epoch": 10.262123197903014, + "grad_norm": 0.18663758039474487, + "learning_rate": 4.9406418279147234e-05, + "loss": 0.0287, + "step": 52860 + }, + { + "epoch": 10.264064851220814, + "grad_norm": 0.1286361813545227, + "learning_rate": 4.9389884778934116e-05, + "loss": 0.0252, + "step": 52870 + }, + { + "epoch": 10.266006504538614, + "grad_norm": 0.21689799427986145, + "learning_rate": 4.937335134544228e-05, + "loss": 0.0231, + "step": 52880 + }, + { + "epoch": 10.267948157856415, + "grad_norm": 0.24066345393657684, + "learning_rate": 4.935681798047979e-05, + "loss": 0.03, + "step": 52890 + }, + { + "epoch": 10.269889811174215, + "grad_norm": 0.10063197463750839, + "learning_rate": 4.934028468585473e-05, + "loss": 0.0317, + "step": 52900 + }, + { + "epoch": 10.271831464492015, + "grad_norm": 0.18333210051059723, + "learning_rate": 4.932375146337514e-05, + "loss": 0.0235, + "step": 52910 + }, + { + "epoch": 10.273773117809816, + "grad_norm": 0.2989475727081299, + "learning_rate": 4.930721831484908e-05, + "loss": 0.0284, + "step": 52920 + }, + { + "epoch": 10.275714771127616, + "grad_norm": 0.15884706377983093, + "learning_rate": 4.929068524208456e-05, + "loss": 0.025, + "step": 52930 + }, + { + "epoch": 10.277656424445416, + "grad_norm": 0.10638707131147385, + "learning_rate": 4.927415224688965e-05, + "loss": 0.0201, + "step": 52940 + }, + { + "epoch": 10.279598077763215, + "grad_norm": 0.16796307265758514, + "learning_rate": 4.9257619331072375e-05, + "loss": 0.0252, + "step": 52950 + }, + { + "epoch": 10.281539731081015, + "grad_norm": 0.16968050599098206, + "learning_rate": 4.924108649644072e-05, + "loss": 0.0297, + "step": 52960 + }, + { + "epoch": 10.283481384398815, + "grad_norm": 0.1450064778327942, + "learning_rate": 4.922455374480271e-05, + "loss": 0.0256, + "step": 52970 + }, + { + "epoch": 10.285423037716615, + "grad_norm": 0.19098570942878723, + "learning_rate": 4.920802107796634e-05, + "loss": 0.0225, + "step": 52980 + }, + { + "epoch": 10.287364691034416, + "grad_norm": 0.16768433153629303, + "learning_rate": 4.9191488497739564e-05, + "loss": 0.0207, + "step": 52990 + }, + { + "epoch": 10.289306344352216, + "grad_norm": 0.11768428981304169, + "learning_rate": 4.917495600593042e-05, + "loss": 0.0336, + "step": 53000 + }, + { + "epoch": 10.291247997670016, + "grad_norm": 0.1650218516588211, + "learning_rate": 4.915842360434684e-05, + "loss": 0.0244, + "step": 53010 + }, + { + "epoch": 10.293189650987816, + "grad_norm": 0.16634629666805267, + "learning_rate": 4.914189129479679e-05, + "loss": 0.0214, + "step": 53020 + }, + { + "epoch": 10.295131304305617, + "grad_norm": 0.16816335916519165, + "learning_rate": 4.912535907908822e-05, + "loss": 0.0257, + "step": 53030 + }, + { + "epoch": 10.297072957623417, + "grad_norm": 0.17029023170471191, + "learning_rate": 4.910882695902907e-05, + "loss": 0.0309, + "step": 53040 + }, + { + "epoch": 10.299014610941217, + "grad_norm": 0.15186059474945068, + "learning_rate": 4.9092294936427234e-05, + "loss": 0.0313, + "step": 53050 + }, + { + "epoch": 10.300956264259016, + "grad_norm": 0.21353241801261902, + "learning_rate": 4.9075763013090685e-05, + "loss": 0.0209, + "step": 53060 + }, + { + "epoch": 10.302897917576816, + "grad_norm": 0.18535888195037842, + "learning_rate": 4.9059231190827305e-05, + "loss": 0.0216, + "step": 53070 + }, + { + "epoch": 10.304839570894616, + "grad_norm": 0.3392988443374634, + "learning_rate": 4.904269947144499e-05, + "loss": 0.0335, + "step": 53080 + }, + { + "epoch": 10.306781224212417, + "grad_norm": 0.12428943812847137, + "learning_rate": 4.902616785675161e-05, + "loss": 0.0225, + "step": 53090 + }, + { + "epoch": 10.308722877530217, + "grad_norm": 0.15509654581546783, + "learning_rate": 4.900963634855505e-05, + "loss": 0.0206, + "step": 53100 + }, + { + "epoch": 10.310664530848017, + "grad_norm": 0.12676823139190674, + "learning_rate": 4.8993104948663173e-05, + "loss": 0.0273, + "step": 53110 + }, + { + "epoch": 10.312606184165817, + "grad_norm": 0.1257760226726532, + "learning_rate": 4.8976573658883823e-05, + "loss": 0.0237, + "step": 53120 + }, + { + "epoch": 10.314547837483618, + "grad_norm": 0.20132264494895935, + "learning_rate": 4.896004248102484e-05, + "loss": 0.0215, + "step": 53130 + }, + { + "epoch": 10.316489490801418, + "grad_norm": 0.16813528537750244, + "learning_rate": 4.8943511416894045e-05, + "loss": 0.0252, + "step": 53140 + }, + { + "epoch": 10.318431144119218, + "grad_norm": 0.1781943440437317, + "learning_rate": 4.892698046829926e-05, + "loss": 0.0255, + "step": 53150 + }, + { + "epoch": 10.320372797437017, + "grad_norm": 0.15116707980632782, + "learning_rate": 4.8910449637048275e-05, + "loss": 0.03, + "step": 53160 + }, + { + "epoch": 10.322314450754817, + "grad_norm": 0.22739216685295105, + "learning_rate": 4.8893918924948884e-05, + "loss": 0.0234, + "step": 53170 + }, + { + "epoch": 10.324256104072617, + "grad_norm": 0.11056627333164215, + "learning_rate": 4.8877388333808835e-05, + "loss": 0.0225, + "step": 53180 + }, + { + "epoch": 10.326197757390418, + "grad_norm": 0.22237996757030487, + "learning_rate": 4.8860857865435925e-05, + "loss": 0.0247, + "step": 53190 + }, + { + "epoch": 10.328139410708218, + "grad_norm": 0.10539597272872925, + "learning_rate": 4.884432752163789e-05, + "loss": 0.0204, + "step": 53200 + }, + { + "epoch": 10.330081064026018, + "grad_norm": 0.2305692434310913, + "learning_rate": 4.8827797304222466e-05, + "loss": 0.0313, + "step": 53210 + }, + { + "epoch": 10.332022717343818, + "grad_norm": 0.11388994008302689, + "learning_rate": 4.8811267214997366e-05, + "loss": 0.0299, + "step": 53220 + }, + { + "epoch": 10.333964370661619, + "grad_norm": 0.24223759770393372, + "learning_rate": 4.879473725577029e-05, + "loss": 0.0228, + "step": 53230 + }, + { + "epoch": 10.335906023979419, + "grad_norm": 0.19617730379104614, + "learning_rate": 4.877820742834891e-05, + "loss": 0.0208, + "step": 53240 + }, + { + "epoch": 10.33784767729722, + "grad_norm": 0.25551837682724, + "learning_rate": 4.8761677734540956e-05, + "loss": 0.0232, + "step": 53250 + }, + { + "epoch": 10.33978933061502, + "grad_norm": 0.1697172224521637, + "learning_rate": 4.874514817615407e-05, + "loss": 0.0289, + "step": 53260 + }, + { + "epoch": 10.341730983932818, + "grad_norm": 0.20353034138679504, + "learning_rate": 4.872861875499588e-05, + "loss": 0.0341, + "step": 53270 + }, + { + "epoch": 10.343672637250618, + "grad_norm": 0.14989343285560608, + "learning_rate": 4.871208947287404e-05, + "loss": 0.0234, + "step": 53280 + }, + { + "epoch": 10.345614290568419, + "grad_norm": 0.14519424736499786, + "learning_rate": 4.869556033159615e-05, + "loss": 0.0366, + "step": 53290 + }, + { + "epoch": 10.347555943886219, + "grad_norm": 0.24737218022346497, + "learning_rate": 4.8679031332969835e-05, + "loss": 0.0308, + "step": 53300 + }, + { + "epoch": 10.349497597204019, + "grad_norm": 0.11275754868984222, + "learning_rate": 4.8662502478802635e-05, + "loss": 0.0231, + "step": 53310 + }, + { + "epoch": 10.35143925052182, + "grad_norm": 0.15263120830059052, + "learning_rate": 4.8645973770902176e-05, + "loss": 0.0232, + "step": 53320 + }, + { + "epoch": 10.35338090383962, + "grad_norm": 0.19460079073905945, + "learning_rate": 4.862944521107599e-05, + "loss": 0.0254, + "step": 53330 + }, + { + "epoch": 10.35532255715742, + "grad_norm": 0.12305038422346115, + "learning_rate": 4.861291680113162e-05, + "loss": 0.0236, + "step": 53340 + }, + { + "epoch": 10.35726421047522, + "grad_norm": 0.12485070526599884, + "learning_rate": 4.8596388542876563e-05, + "loss": 0.028, + "step": 53350 + }, + { + "epoch": 10.35920586379302, + "grad_norm": 0.11864934116601944, + "learning_rate": 4.857986043811836e-05, + "loss": 0.0211, + "step": 53360 + }, + { + "epoch": 10.36114751711082, + "grad_norm": 0.20170170068740845, + "learning_rate": 4.856333248866448e-05, + "loss": 0.0285, + "step": 53370 + }, + { + "epoch": 10.36308917042862, + "grad_norm": 0.15246210992336273, + "learning_rate": 4.85468046963224e-05, + "loss": 0.0283, + "step": 53380 + }, + { + "epoch": 10.36503082374642, + "grad_norm": 0.18268638849258423, + "learning_rate": 4.853027706289957e-05, + "loss": 0.0225, + "step": 53390 + }, + { + "epoch": 10.36697247706422, + "grad_norm": 0.2674362361431122, + "learning_rate": 4.851374959020344e-05, + "loss": 0.0256, + "step": 53400 + }, + { + "epoch": 10.36891413038202, + "grad_norm": 0.150828018784523, + "learning_rate": 4.849722228004141e-05, + "loss": 0.0253, + "step": 53410 + }, + { + "epoch": 10.37085578369982, + "grad_norm": 0.17614026367664337, + "learning_rate": 4.848069513422091e-05, + "loss": 0.0274, + "step": 53420 + }, + { + "epoch": 10.37279743701762, + "grad_norm": 0.2658246159553528, + "learning_rate": 4.8464168154549294e-05, + "loss": 0.0296, + "step": 53430 + }, + { + "epoch": 10.37473909033542, + "grad_norm": 0.22288112342357635, + "learning_rate": 4.844764134283392e-05, + "loss": 0.0264, + "step": 53440 + }, + { + "epoch": 10.376680743653221, + "grad_norm": 0.13527293503284454, + "learning_rate": 4.843111470088218e-05, + "loss": 0.0273, + "step": 53450 + }, + { + "epoch": 10.378622396971021, + "grad_norm": 0.10640635341405869, + "learning_rate": 4.841458823050137e-05, + "loss": 0.0262, + "step": 53460 + }, + { + "epoch": 10.380564050288822, + "grad_norm": 0.17235074937343597, + "learning_rate": 4.8398061933498816e-05, + "loss": 0.0234, + "step": 53470 + }, + { + "epoch": 10.382505703606622, + "grad_norm": 0.12565572559833527, + "learning_rate": 4.83815358116818e-05, + "loss": 0.0182, + "step": 53480 + }, + { + "epoch": 10.38444735692442, + "grad_norm": 0.23687542974948883, + "learning_rate": 4.8365009866857584e-05, + "loss": 0.0259, + "step": 53490 + }, + { + "epoch": 10.38638901024222, + "grad_norm": 0.14704450964927673, + "learning_rate": 4.834848410083342e-05, + "loss": 0.0176, + "step": 53500 + }, + { + "epoch": 10.388330663560021, + "grad_norm": 0.12249636650085449, + "learning_rate": 4.833195851541657e-05, + "loss": 0.0255, + "step": 53510 + }, + { + "epoch": 10.390272316877821, + "grad_norm": 0.12745331227779388, + "learning_rate": 4.831543311241423e-05, + "loss": 0.0262, + "step": 53520 + }, + { + "epoch": 10.392213970195622, + "grad_norm": 0.17895473539829254, + "learning_rate": 4.829890789363359e-05, + "loss": 0.0214, + "step": 53530 + }, + { + "epoch": 10.394155623513422, + "grad_norm": 0.32287415862083435, + "learning_rate": 4.828238286088182e-05, + "loss": 0.0263, + "step": 53540 + }, + { + "epoch": 10.396097276831222, + "grad_norm": 0.11309900879859924, + "learning_rate": 4.826585801596609e-05, + "loss": 0.0264, + "step": 53550 + }, + { + "epoch": 10.398038930149022, + "grad_norm": 0.11324533075094223, + "learning_rate": 4.8249333360693486e-05, + "loss": 0.0204, + "step": 53560 + }, + { + "epoch": 10.399980583466823, + "grad_norm": 0.1634945273399353, + "learning_rate": 4.8232808896871185e-05, + "loss": 0.0294, + "step": 53570 + }, + { + "epoch": 10.401922236784623, + "grad_norm": 0.4339398145675659, + "learning_rate": 4.821628462630624e-05, + "loss": 0.0295, + "step": 53580 + }, + { + "epoch": 10.403863890102421, + "grad_norm": 0.15767662227153778, + "learning_rate": 4.8199760550805735e-05, + "loss": 0.0266, + "step": 53590 + }, + { + "epoch": 10.405805543420222, + "grad_norm": 0.14544463157653809, + "learning_rate": 4.818323667217669e-05, + "loss": 0.0237, + "step": 53600 + }, + { + "epoch": 10.407747196738022, + "grad_norm": 0.1712036281824112, + "learning_rate": 4.8166712992226165e-05, + "loss": 0.027, + "step": 53610 + }, + { + "epoch": 10.409688850055822, + "grad_norm": 0.1303112506866455, + "learning_rate": 4.8150189512761154e-05, + "loss": 0.0293, + "step": 53620 + }, + { + "epoch": 10.411630503373623, + "grad_norm": 0.11896079778671265, + "learning_rate": 4.813366623558862e-05, + "loss": 0.0247, + "step": 53630 + }, + { + "epoch": 10.413572156691423, + "grad_norm": 0.10759659856557846, + "learning_rate": 4.811714316251554e-05, + "loss": 0.022, + "step": 53640 + }, + { + "epoch": 10.415513810009223, + "grad_norm": 0.19120965898036957, + "learning_rate": 4.810062029534886e-05, + "loss": 0.0262, + "step": 53650 + }, + { + "epoch": 10.417455463327023, + "grad_norm": 0.1837225705385208, + "learning_rate": 4.8084097635895505e-05, + "loss": 0.0285, + "step": 53660 + }, + { + "epoch": 10.419397116644824, + "grad_norm": 0.2693427801132202, + "learning_rate": 4.8067575185962335e-05, + "loss": 0.0264, + "step": 53670 + }, + { + "epoch": 10.421338769962624, + "grad_norm": 0.20692752301692963, + "learning_rate": 4.805105294735625e-05, + "loss": 0.0363, + "step": 53680 + }, + { + "epoch": 10.423280423280424, + "grad_norm": 0.1338297426700592, + "learning_rate": 4.803453092188406e-05, + "loss": 0.0209, + "step": 53690 + }, + { + "epoch": 10.425222076598223, + "grad_norm": 0.14276084303855896, + "learning_rate": 4.801800911135263e-05, + "loss": 0.0265, + "step": 53700 + }, + { + "epoch": 10.427163729916023, + "grad_norm": 0.24517615139484406, + "learning_rate": 4.8001487517568744e-05, + "loss": 0.0305, + "step": 53710 + }, + { + "epoch": 10.429105383233823, + "grad_norm": 0.18891467154026031, + "learning_rate": 4.798496614233918e-05, + "loss": 0.0237, + "step": 53720 + }, + { + "epoch": 10.431047036551623, + "grad_norm": 0.1545296609401703, + "learning_rate": 4.796844498747069e-05, + "loss": 0.033, + "step": 53730 + }, + { + "epoch": 10.432988689869424, + "grad_norm": 0.157765731215477, + "learning_rate": 4.795192405477e-05, + "loss": 0.0225, + "step": 53740 + }, + { + "epoch": 10.434930343187224, + "grad_norm": 0.12317518144845963, + "learning_rate": 4.79354033460438e-05, + "loss": 0.0242, + "step": 53750 + }, + { + "epoch": 10.436871996505024, + "grad_norm": 0.3953395485877991, + "learning_rate": 4.79188828630988e-05, + "loss": 0.0357, + "step": 53760 + }, + { + "epoch": 10.438813649822825, + "grad_norm": 0.09654375910758972, + "learning_rate": 4.790236260774165e-05, + "loss": 0.0303, + "step": 53770 + }, + { + "epoch": 10.440755303140625, + "grad_norm": 0.20073288679122925, + "learning_rate": 4.788584258177896e-05, + "loss": 0.024, + "step": 53780 + }, + { + "epoch": 10.442696956458425, + "grad_norm": 0.24667459726333618, + "learning_rate": 4.7869322787017356e-05, + "loss": 0.0236, + "step": 53790 + }, + { + "epoch": 10.444638609776224, + "grad_norm": 0.2781980335712433, + "learning_rate": 4.78528032252634e-05, + "loss": 0.0239, + "step": 53800 + }, + { + "epoch": 10.446580263094024, + "grad_norm": 0.1292855143547058, + "learning_rate": 4.783628389832366e-05, + "loss": 0.0269, + "step": 53810 + }, + { + "epoch": 10.448521916411824, + "grad_norm": 0.208623006939888, + "learning_rate": 4.7819764808004644e-05, + "loss": 0.0227, + "step": 53820 + }, + { + "epoch": 10.450463569729624, + "grad_norm": 0.1211472898721695, + "learning_rate": 4.780324595611289e-05, + "loss": 0.0236, + "step": 53830 + }, + { + "epoch": 10.452405223047425, + "grad_norm": 0.15062077343463898, + "learning_rate": 4.7786727344454854e-05, + "loss": 0.0206, + "step": 53840 + }, + { + "epoch": 10.454346876365225, + "grad_norm": 0.17309147119522095, + "learning_rate": 4.777020897483697e-05, + "loss": 0.0244, + "step": 53850 + }, + { + "epoch": 10.456288529683025, + "grad_norm": 0.2433672994375229, + "learning_rate": 4.775369084906569e-05, + "loss": 0.0236, + "step": 53860 + }, + { + "epoch": 10.458230183000826, + "grad_norm": 0.27346745133399963, + "learning_rate": 4.773717296894742e-05, + "loss": 0.0251, + "step": 53870 + }, + { + "epoch": 10.460171836318626, + "grad_norm": 0.0987197756767273, + "learning_rate": 4.772065533628848e-05, + "loss": 0.0251, + "step": 53880 + }, + { + "epoch": 10.462113489636426, + "grad_norm": 0.12252558022737503, + "learning_rate": 4.7704137952895265e-05, + "loss": 0.0332, + "step": 53890 + }, + { + "epoch": 10.464055142954226, + "grad_norm": 0.16600042581558228, + "learning_rate": 4.768762082057407e-05, + "loss": 0.0269, + "step": 53900 + }, + { + "epoch": 10.465996796272025, + "grad_norm": 0.1391725093126297, + "learning_rate": 4.7671103941131186e-05, + "loss": 0.0213, + "step": 53910 + }, + { + "epoch": 10.467938449589825, + "grad_norm": 0.15352369844913483, + "learning_rate": 4.765458731637287e-05, + "loss": 0.0223, + "step": 53920 + }, + { + "epoch": 10.469880102907625, + "grad_norm": 0.13718226552009583, + "learning_rate": 4.763807094810537e-05, + "loss": 0.023, + "step": 53930 + }, + { + "epoch": 10.471821756225426, + "grad_norm": 0.16560319066047668, + "learning_rate": 4.762155483813487e-05, + "loss": 0.0266, + "step": 53940 + }, + { + "epoch": 10.473763409543226, + "grad_norm": 0.18316420912742615, + "learning_rate": 4.760503898826754e-05, + "loss": 0.0204, + "step": 53950 + }, + { + "epoch": 10.475705062861026, + "grad_norm": 0.1140185222029686, + "learning_rate": 4.758852340030957e-05, + "loss": 0.0224, + "step": 53960 + }, + { + "epoch": 10.477646716178826, + "grad_norm": 0.1295527070760727, + "learning_rate": 4.757200807606705e-05, + "loss": 0.0227, + "step": 53970 + }, + { + "epoch": 10.479588369496627, + "grad_norm": 0.10476531088352203, + "learning_rate": 4.755549301734607e-05, + "loss": 0.0295, + "step": 53980 + }, + { + "epoch": 10.481530022814427, + "grad_norm": 0.15116436779499054, + "learning_rate": 4.753897822595271e-05, + "loss": 0.0252, + "step": 53990 + }, + { + "epoch": 10.483471676132227, + "grad_norm": 0.12051970511674881, + "learning_rate": 4.7522463703692994e-05, + "loss": 0.027, + "step": 54000 + }, + { + "epoch": 10.485413329450026, + "grad_norm": 0.17040082812309265, + "learning_rate": 4.75059494523729e-05, + "loss": 0.0274, + "step": 54010 + }, + { + "epoch": 10.487354982767826, + "grad_norm": 0.1689983755350113, + "learning_rate": 4.748943547379843e-05, + "loss": 0.0255, + "step": 54020 + }, + { + "epoch": 10.489296636085626, + "grad_norm": 0.14835593104362488, + "learning_rate": 4.7472921769775535e-05, + "loss": 0.0276, + "step": 54030 + }, + { + "epoch": 10.491238289403427, + "grad_norm": 0.22554239630699158, + "learning_rate": 4.745640834211011e-05, + "loss": 0.0255, + "step": 54040 + }, + { + "epoch": 10.493179942721227, + "grad_norm": 0.17515012621879578, + "learning_rate": 4.7439895192608054e-05, + "loss": 0.021, + "step": 54050 + }, + { + "epoch": 10.495121596039027, + "grad_norm": 0.12526187300682068, + "learning_rate": 4.7423382323075186e-05, + "loss": 0.0232, + "step": 54060 + }, + { + "epoch": 10.497063249356827, + "grad_norm": 0.2376967817544937, + "learning_rate": 4.740686973531737e-05, + "loss": 0.0256, + "step": 54070 + }, + { + "epoch": 10.499004902674628, + "grad_norm": 0.19564072787761688, + "learning_rate": 4.7390357431140375e-05, + "loss": 0.0219, + "step": 54080 + }, + { + "epoch": 10.500946555992428, + "grad_norm": 0.21394401788711548, + "learning_rate": 4.737384541234998e-05, + "loss": 0.0251, + "step": 54090 + }, + { + "epoch": 10.502888209310228, + "grad_norm": 0.13280242681503296, + "learning_rate": 4.735733368075189e-05, + "loss": 0.0272, + "step": 54100 + }, + { + "epoch": 10.504829862628029, + "grad_norm": 0.2315320372581482, + "learning_rate": 4.734082223815182e-05, + "loss": 0.0269, + "step": 54110 + }, + { + "epoch": 10.506771515945829, + "grad_norm": 0.11393146216869354, + "learning_rate": 4.7324311086355436e-05, + "loss": 0.028, + "step": 54120 + }, + { + "epoch": 10.508713169263627, + "grad_norm": 0.20576421916484833, + "learning_rate": 4.730780022716836e-05, + "loss": 0.0261, + "step": 54130 + }, + { + "epoch": 10.510654822581428, + "grad_norm": 0.146884024143219, + "learning_rate": 4.729128966239619e-05, + "loss": 0.0272, + "step": 54140 + }, + { + "epoch": 10.512596475899228, + "grad_norm": 0.2327466756105423, + "learning_rate": 4.7274779393844536e-05, + "loss": 0.0227, + "step": 54150 + }, + { + "epoch": 10.514538129217028, + "grad_norm": 0.21910040080547333, + "learning_rate": 4.725826942331891e-05, + "loss": 0.0273, + "step": 54160 + }, + { + "epoch": 10.516479782534828, + "grad_norm": 0.10524764657020569, + "learning_rate": 4.724175975262481e-05, + "loss": 0.025, + "step": 54170 + }, + { + "epoch": 10.518421435852629, + "grad_norm": 0.23956480622291565, + "learning_rate": 4.722525038356772e-05, + "loss": 0.0261, + "step": 54180 + }, + { + "epoch": 10.520363089170429, + "grad_norm": 0.13689306378364563, + "learning_rate": 4.7208741317953075e-05, + "loss": 0.0351, + "step": 54190 + }, + { + "epoch": 10.52230474248823, + "grad_norm": 0.24379026889801025, + "learning_rate": 4.7192232557586264e-05, + "loss": 0.0281, + "step": 54200 + }, + { + "epoch": 10.52424639580603, + "grad_norm": 0.1406620293855667, + "learning_rate": 4.717572410427269e-05, + "loss": 0.0371, + "step": 54210 + }, + { + "epoch": 10.52618804912383, + "grad_norm": 0.20003525912761688, + "learning_rate": 4.71592159598177e-05, + "loss": 0.0261, + "step": 54220 + }, + { + "epoch": 10.528129702441628, + "grad_norm": 0.17562557756900787, + "learning_rate": 4.714270812602657e-05, + "loss": 0.0271, + "step": 54230 + }, + { + "epoch": 10.530071355759429, + "grad_norm": 0.20881250500679016, + "learning_rate": 4.712620060470458e-05, + "loss": 0.0254, + "step": 54240 + }, + { + "epoch": 10.532013009077229, + "grad_norm": 0.1546546071767807, + "learning_rate": 4.7109693397656986e-05, + "loss": 0.0277, + "step": 54250 + }, + { + "epoch": 10.533954662395029, + "grad_norm": 0.19727329909801483, + "learning_rate": 4.709318650668894e-05, + "loss": 0.0222, + "step": 54260 + }, + { + "epoch": 10.53589631571283, + "grad_norm": 0.1515333205461502, + "learning_rate": 4.7076679933605676e-05, + "loss": 0.0301, + "step": 54270 + }, + { + "epoch": 10.53783796903063, + "grad_norm": 0.16509903967380524, + "learning_rate": 4.70601736802123e-05, + "loss": 0.0234, + "step": 54280 + }, + { + "epoch": 10.53977962234843, + "grad_norm": 0.13570623099803925, + "learning_rate": 4.704366774831391e-05, + "loss": 0.0268, + "step": 54290 + }, + { + "epoch": 10.54172127566623, + "grad_norm": 0.17374826967716217, + "learning_rate": 4.702716213971557e-05, + "loss": 0.0291, + "step": 54300 + }, + { + "epoch": 10.54366292898403, + "grad_norm": 0.12339568883180618, + "learning_rate": 4.70106568562223e-05, + "loss": 0.0243, + "step": 54310 + }, + { + "epoch": 10.54560458230183, + "grad_norm": 0.3507354259490967, + "learning_rate": 4.6994151899639105e-05, + "loss": 0.0286, + "step": 54320 + }, + { + "epoch": 10.547546235619631, + "grad_norm": 0.13597369194030762, + "learning_rate": 4.697764727177093e-05, + "loss": 0.0188, + "step": 54330 + }, + { + "epoch": 10.54948788893743, + "grad_norm": 0.19008538126945496, + "learning_rate": 4.696114297442272e-05, + "loss": 0.022, + "step": 54340 + }, + { + "epoch": 10.55142954225523, + "grad_norm": 0.2425709068775177, + "learning_rate": 4.694463900939933e-05, + "loss": 0.0289, + "step": 54350 + }, + { + "epoch": 10.55337119557303, + "grad_norm": 0.20455481112003326, + "learning_rate": 4.692813537850564e-05, + "loss": 0.0348, + "step": 54360 + }, + { + "epoch": 10.55531284889083, + "grad_norm": 0.1641455739736557, + "learning_rate": 4.6911632083546445e-05, + "loss": 0.0297, + "step": 54370 + }, + { + "epoch": 10.55725450220863, + "grad_norm": 0.17881669104099274, + "learning_rate": 4.6895129126326525e-05, + "loss": 0.025, + "step": 54380 + }, + { + "epoch": 10.55919615552643, + "grad_norm": 0.15290653705596924, + "learning_rate": 4.687862650865059e-05, + "loss": 0.0205, + "step": 54390 + }, + { + "epoch": 10.561137808844231, + "grad_norm": 0.23842540383338928, + "learning_rate": 4.686212423232339e-05, + "loss": 0.0281, + "step": 54400 + }, + { + "epoch": 10.563079462162031, + "grad_norm": 0.09225936979055405, + "learning_rate": 4.6845622299149565e-05, + "loss": 0.0215, + "step": 54410 + }, + { + "epoch": 10.565021115479832, + "grad_norm": 0.19561569392681122, + "learning_rate": 4.682912071093374e-05, + "loss": 0.022, + "step": 54420 + }, + { + "epoch": 10.566962768797632, + "grad_norm": 0.2014070749282837, + "learning_rate": 4.681261946948052e-05, + "loss": 0.0254, + "step": 54430 + }, + { + "epoch": 10.56890442211543, + "grad_norm": 0.10010844469070435, + "learning_rate": 4.679611857659443e-05, + "loss": 0.025, + "step": 54440 + }, + { + "epoch": 10.57084607543323, + "grad_norm": 0.1734624058008194, + "learning_rate": 4.6779618034079975e-05, + "loss": 0.0251, + "step": 54450 + }, + { + "epoch": 10.572787728751031, + "grad_norm": 0.19600673019886017, + "learning_rate": 4.6763117843741675e-05, + "loss": 0.024, + "step": 54460 + }, + { + "epoch": 10.574729382068831, + "grad_norm": 0.23232728242874146, + "learning_rate": 4.674661800738394e-05, + "loss": 0.0296, + "step": 54470 + }, + { + "epoch": 10.576671035386632, + "grad_norm": 0.15634751319885254, + "learning_rate": 4.673011852681117e-05, + "loss": 0.0196, + "step": 54480 + }, + { + "epoch": 10.578612688704432, + "grad_norm": 0.2971816658973694, + "learning_rate": 4.671361940382772e-05, + "loss": 0.0283, + "step": 54490 + }, + { + "epoch": 10.580554342022232, + "grad_norm": 0.09349570423364639, + "learning_rate": 4.669712064023791e-05, + "loss": 0.0251, + "step": 54500 + }, + { + "epoch": 10.582495995340032, + "grad_norm": 0.14488527178764343, + "learning_rate": 4.668062223784602e-05, + "loss": 0.0221, + "step": 54510 + }, + { + "epoch": 10.584437648657833, + "grad_norm": 0.1874062418937683, + "learning_rate": 4.6664124198456274e-05, + "loss": 0.0314, + "step": 54520 + }, + { + "epoch": 10.586379301975633, + "grad_norm": 0.17188669741153717, + "learning_rate": 4.6647626523872905e-05, + "loss": 0.0267, + "step": 54530 + }, + { + "epoch": 10.588320955293433, + "grad_norm": 0.13693498075008392, + "learning_rate": 4.6631129215900066e-05, + "loss": 0.0178, + "step": 54540 + }, + { + "epoch": 10.590262608611232, + "grad_norm": 0.17703069746494293, + "learning_rate": 4.661463227634187e-05, + "loss": 0.0278, + "step": 54550 + }, + { + "epoch": 10.592204261929032, + "grad_norm": 0.08368969708681107, + "learning_rate": 4.659813570700238e-05, + "loss": 0.0256, + "step": 54560 + }, + { + "epoch": 10.594145915246832, + "grad_norm": 0.11021411418914795, + "learning_rate": 4.658163950968568e-05, + "loss": 0.0225, + "step": 54570 + }, + { + "epoch": 10.596087568564633, + "grad_norm": 0.24277904629707336, + "learning_rate": 4.6565143686195726e-05, + "loss": 0.0309, + "step": 54580 + }, + { + "epoch": 10.598029221882433, + "grad_norm": 0.1857050508260727, + "learning_rate": 4.654864823833649e-05, + "loss": 0.029, + "step": 54590 + }, + { + "epoch": 10.599970875200233, + "grad_norm": 0.1305861622095108, + "learning_rate": 4.65321531679119e-05, + "loss": 0.0262, + "step": 54600 + }, + { + "epoch": 10.601912528518033, + "grad_norm": 0.2059132307767868, + "learning_rate": 4.6515658476725834e-05, + "loss": 0.0294, + "step": 54610 + }, + { + "epoch": 10.603854181835834, + "grad_norm": 0.1291099339723587, + "learning_rate": 4.649916416658212e-05, + "loss": 0.021, + "step": 54620 + }, + { + "epoch": 10.605795835153634, + "grad_norm": 0.07089049369096756, + "learning_rate": 4.648267023928455e-05, + "loss": 0.024, + "step": 54630 + }, + { + "epoch": 10.607737488471434, + "grad_norm": 0.1537701040506363, + "learning_rate": 4.646617669663687e-05, + "loss": 0.0246, + "step": 54640 + }, + { + "epoch": 10.609679141789233, + "grad_norm": 0.23314036428928375, + "learning_rate": 4.644968354044277e-05, + "loss": 0.0225, + "step": 54650 + }, + { + "epoch": 10.611620795107033, + "grad_norm": 0.27359700202941895, + "learning_rate": 4.643319077250596e-05, + "loss": 0.0259, + "step": 54660 + }, + { + "epoch": 10.613562448424833, + "grad_norm": 0.0783952996134758, + "learning_rate": 4.6416698394630055e-05, + "loss": 0.0186, + "step": 54670 + }, + { + "epoch": 10.615504101742633, + "grad_norm": 0.22040186822414398, + "learning_rate": 4.640020640861862e-05, + "loss": 0.0246, + "step": 54680 + }, + { + "epoch": 10.617445755060434, + "grad_norm": 0.26173290610313416, + "learning_rate": 4.6383714816275195e-05, + "loss": 0.0295, + "step": 54690 + }, + { + "epoch": 10.619387408378234, + "grad_norm": 0.2105872929096222, + "learning_rate": 4.636722361940329e-05, + "loss": 0.0273, + "step": 54700 + }, + { + "epoch": 10.621329061696034, + "grad_norm": 0.23058751225471497, + "learning_rate": 4.635073281980632e-05, + "loss": 0.0289, + "step": 54710 + }, + { + "epoch": 10.623270715013835, + "grad_norm": 0.1356731653213501, + "learning_rate": 4.633424241928775e-05, + "loss": 0.0346, + "step": 54720 + }, + { + "epoch": 10.625212368331635, + "grad_norm": 0.18289420008659363, + "learning_rate": 4.6317752419650914e-05, + "loss": 0.0288, + "step": 54730 + }, + { + "epoch": 10.627154021649435, + "grad_norm": 0.20748086273670197, + "learning_rate": 4.630126282269914e-05, + "loss": 0.025, + "step": 54740 + }, + { + "epoch": 10.629095674967235, + "grad_norm": 0.15469981729984283, + "learning_rate": 4.628477363023569e-05, + "loss": 0.0272, + "step": 54750 + }, + { + "epoch": 10.631037328285034, + "grad_norm": 0.13519974052906036, + "learning_rate": 4.626828484406383e-05, + "loss": 0.0241, + "step": 54760 + }, + { + "epoch": 10.632978981602834, + "grad_norm": 0.17861437797546387, + "learning_rate": 4.625179646598669e-05, + "loss": 0.024, + "step": 54770 + }, + { + "epoch": 10.634920634920634, + "grad_norm": 0.1383136361837387, + "learning_rate": 4.6235308497807487e-05, + "loss": 0.0252, + "step": 54780 + }, + { + "epoch": 10.636862288238435, + "grad_norm": 0.19027920067310333, + "learning_rate": 4.6218820941329276e-05, + "loss": 0.025, + "step": 54790 + }, + { + "epoch": 10.638803941556235, + "grad_norm": 0.17633526027202606, + "learning_rate": 4.620233379835513e-05, + "loss": 0.027, + "step": 54800 + }, + { + "epoch": 10.640745594874035, + "grad_norm": 0.16263602674007416, + "learning_rate": 4.618584707068804e-05, + "loss": 0.03, + "step": 54810 + }, + { + "epoch": 10.642687248191836, + "grad_norm": 0.20759296417236328, + "learning_rate": 4.616936076013099e-05, + "loss": 0.0263, + "step": 54820 + }, + { + "epoch": 10.644628901509636, + "grad_norm": 0.1497931182384491, + "learning_rate": 4.615287486848688e-05, + "loss": 0.0257, + "step": 54830 + }, + { + "epoch": 10.646570554827436, + "grad_norm": 0.2869642376899719, + "learning_rate": 4.6136389397558584e-05, + "loss": 0.0273, + "step": 54840 + }, + { + "epoch": 10.648512208145236, + "grad_norm": 0.16277532279491425, + "learning_rate": 4.6119904349148935e-05, + "loss": 0.0293, + "step": 54850 + }, + { + "epoch": 10.650453861463035, + "grad_norm": 0.2241763323545456, + "learning_rate": 4.6103419725060726e-05, + "loss": 0.0242, + "step": 54860 + }, + { + "epoch": 10.652395514780835, + "grad_norm": 0.18066906929016113, + "learning_rate": 4.608693552709667e-05, + "loss": 0.0255, + "step": 54870 + }, + { + "epoch": 10.654337168098635, + "grad_norm": 0.12352555990219116, + "learning_rate": 4.607045175705947e-05, + "loss": 0.0257, + "step": 54880 + }, + { + "epoch": 10.656278821416436, + "grad_norm": 0.15359100699424744, + "learning_rate": 4.605396841675176e-05, + "loss": 0.0189, + "step": 54890 + }, + { + "epoch": 10.658220474734236, + "grad_norm": 0.21786655485630035, + "learning_rate": 4.6037485507976094e-05, + "loss": 0.0253, + "step": 54900 + }, + { + "epoch": 10.660162128052036, + "grad_norm": 0.1356218457221985, + "learning_rate": 4.602100303253509e-05, + "loss": 0.0219, + "step": 54910 + }, + { + "epoch": 10.662103781369836, + "grad_norm": 0.3246844410896301, + "learning_rate": 4.600452099223122e-05, + "loss": 0.0266, + "step": 54920 + }, + { + "epoch": 10.664045434687637, + "grad_norm": 0.2031717449426651, + "learning_rate": 4.598803938886691e-05, + "loss": 0.0324, + "step": 54930 + }, + { + "epoch": 10.665987088005437, + "grad_norm": 0.17522595822811127, + "learning_rate": 4.597155822424459e-05, + "loss": 0.0233, + "step": 54940 + }, + { + "epoch": 10.667928741323237, + "grad_norm": 0.07127680629491806, + "learning_rate": 4.59550775001666e-05, + "loss": 0.0219, + "step": 54950 + }, + { + "epoch": 10.669870394641038, + "grad_norm": 0.10145461559295654, + "learning_rate": 4.5938597218435234e-05, + "loss": 0.0205, + "step": 54960 + }, + { + "epoch": 10.671812047958838, + "grad_norm": 0.1936151087284088, + "learning_rate": 4.592211738085279e-05, + "loss": 0.0267, + "step": 54970 + }, + { + "epoch": 10.673753701276636, + "grad_norm": 0.1040283590555191, + "learning_rate": 4.590563798922146e-05, + "loss": 0.0289, + "step": 54980 + }, + { + "epoch": 10.675695354594437, + "grad_norm": 0.15916213393211365, + "learning_rate": 4.5889159045343404e-05, + "loss": 0.0257, + "step": 54990 + }, + { + "epoch": 10.677637007912237, + "grad_norm": 0.1351243257522583, + "learning_rate": 4.5872680551020734e-05, + "loss": 0.0271, + "step": 55000 + }, + { + "epoch": 10.679578661230037, + "grad_norm": 0.16744214296340942, + "learning_rate": 4.585620250805552e-05, + "loss": 0.0278, + "step": 55010 + }, + { + "epoch": 10.681520314547837, + "grad_norm": 0.20143435895442963, + "learning_rate": 4.583972491824975e-05, + "loss": 0.026, + "step": 55020 + }, + { + "epoch": 10.683461967865638, + "grad_norm": 0.12119381129741669, + "learning_rate": 4.582324778340541e-05, + "loss": 0.0222, + "step": 55030 + }, + { + "epoch": 10.685403621183438, + "grad_norm": 0.1459367275238037, + "learning_rate": 4.5806771105324424e-05, + "loss": 0.0276, + "step": 55040 + }, + { + "epoch": 10.687345274501238, + "grad_norm": 0.15731924772262573, + "learning_rate": 4.579029488580865e-05, + "loss": 0.0293, + "step": 55050 + }, + { + "epoch": 10.689286927819039, + "grad_norm": 0.07035990059375763, + "learning_rate": 4.5773819126659876e-05, + "loss": 0.0196, + "step": 55060 + }, + { + "epoch": 10.691228581136839, + "grad_norm": 0.18590980768203735, + "learning_rate": 4.575734382967991e-05, + "loss": 0.0232, + "step": 55070 + }, + { + "epoch": 10.693170234454637, + "grad_norm": 0.12931478023529053, + "learning_rate": 4.574086899667044e-05, + "loss": 0.0223, + "step": 55080 + }, + { + "epoch": 10.695111887772438, + "grad_norm": 0.13667982816696167, + "learning_rate": 4.5724394629433106e-05, + "loss": 0.0233, + "step": 55090 + }, + { + "epoch": 10.697053541090238, + "grad_norm": 0.19656100869178772, + "learning_rate": 4.570792072976958e-05, + "loss": 0.0184, + "step": 55100 + }, + { + "epoch": 10.698995194408038, + "grad_norm": 0.13129600882530212, + "learning_rate": 4.569144729948138e-05, + "loss": 0.032, + "step": 55110 + }, + { + "epoch": 10.700936847725838, + "grad_norm": 0.23064365983009338, + "learning_rate": 4.5674974340370033e-05, + "loss": 0.0277, + "step": 55120 + }, + { + "epoch": 10.702878501043639, + "grad_norm": 0.13834132254123688, + "learning_rate": 4.565850185423699e-05, + "loss": 0.0202, + "step": 55130 + }, + { + "epoch": 10.704820154361439, + "grad_norm": 0.16155105829238892, + "learning_rate": 4.564202984288365e-05, + "loss": 0.0269, + "step": 55140 + }, + { + "epoch": 10.70676180767924, + "grad_norm": 0.283810555934906, + "learning_rate": 4.5625558308111356e-05, + "loss": 0.0259, + "step": 55150 + }, + { + "epoch": 10.70870346099704, + "grad_norm": 0.18152980506420135, + "learning_rate": 4.560908725172145e-05, + "loss": 0.0202, + "step": 55160 + }, + { + "epoch": 10.71064511431484, + "grad_norm": 0.21758483350276947, + "learning_rate": 4.559261667551516e-05, + "loss": 0.0223, + "step": 55170 + }, + { + "epoch": 10.71258676763264, + "grad_norm": 0.2983839213848114, + "learning_rate": 4.5576146581293685e-05, + "loss": 0.0239, + "step": 55180 + }, + { + "epoch": 10.714528420950439, + "grad_norm": 0.16246864199638367, + "learning_rate": 4.5559676970858164e-05, + "loss": 0.0195, + "step": 55190 + }, + { + "epoch": 10.716470074268239, + "grad_norm": 0.20448628067970276, + "learning_rate": 4.5543207846009705e-05, + "loss": 0.029, + "step": 55200 + }, + { + "epoch": 10.718411727586039, + "grad_norm": 0.18333368003368378, + "learning_rate": 4.552673920854933e-05, + "loss": 0.0266, + "step": 55210 + }, + { + "epoch": 10.72035338090384, + "grad_norm": 0.15691660344600677, + "learning_rate": 4.5510271060278006e-05, + "loss": 0.0261, + "step": 55220 + }, + { + "epoch": 10.72229503422164, + "grad_norm": 0.1940966546535492, + "learning_rate": 4.5493803402996704e-05, + "loss": 0.0262, + "step": 55230 + }, + { + "epoch": 10.72423668753944, + "grad_norm": 0.11924688518047333, + "learning_rate": 4.5477336238506295e-05, + "loss": 0.0213, + "step": 55240 + }, + { + "epoch": 10.72617834085724, + "grad_norm": 0.18528388440608978, + "learning_rate": 4.546086956860759e-05, + "loss": 0.0216, + "step": 55250 + }, + { + "epoch": 10.72811999417504, + "grad_norm": 0.2682265341281891, + "learning_rate": 4.5444403395101374e-05, + "loss": 0.0247, + "step": 55260 + }, + { + "epoch": 10.73006164749284, + "grad_norm": 0.2525326907634735, + "learning_rate": 4.542793771978834e-05, + "loss": 0.0257, + "step": 55270 + }, + { + "epoch": 10.732003300810641, + "grad_norm": 0.08485419303178787, + "learning_rate": 4.541147254446916e-05, + "loss": 0.0237, + "step": 55280 + }, + { + "epoch": 10.73394495412844, + "grad_norm": 0.2495793253183365, + "learning_rate": 4.539500787094445e-05, + "loss": 0.0237, + "step": 55290 + }, + { + "epoch": 10.73588660744624, + "grad_norm": 0.2490960657596588, + "learning_rate": 4.537854370101475e-05, + "loss": 0.029, + "step": 55300 + }, + { + "epoch": 10.73782826076404, + "grad_norm": 0.13250724971294403, + "learning_rate": 4.5362080036480566e-05, + "loss": 0.0299, + "step": 55310 + }, + { + "epoch": 10.73976991408184, + "grad_norm": 0.12523555755615234, + "learning_rate": 4.534561687914235e-05, + "loss": 0.0225, + "step": 55320 + }, + { + "epoch": 10.74171156739964, + "grad_norm": 0.24009136855602264, + "learning_rate": 4.5329154230800466e-05, + "loss": 0.0224, + "step": 55330 + }, + { + "epoch": 10.74365322071744, + "grad_norm": 0.22744466364383698, + "learning_rate": 4.531269209325525e-05, + "loss": 0.0247, + "step": 55340 + }, + { + "epoch": 10.745594874035241, + "grad_norm": 0.14146146178245544, + "learning_rate": 4.529623046830696e-05, + "loss": 0.0289, + "step": 55350 + }, + { + "epoch": 10.747536527353041, + "grad_norm": 0.14018936455249786, + "learning_rate": 4.5279769357755866e-05, + "loss": 0.0301, + "step": 55360 + }, + { + "epoch": 10.749478180670842, + "grad_norm": 0.18970249593257904, + "learning_rate": 4.5263308763402084e-05, + "loss": 0.0248, + "step": 55370 + }, + { + "epoch": 10.751419833988642, + "grad_norm": 0.17059263586997986, + "learning_rate": 4.524684868704574e-05, + "loss": 0.0206, + "step": 55380 + }, + { + "epoch": 10.753361487306442, + "grad_norm": 0.1961117386817932, + "learning_rate": 4.523038913048688e-05, + "loss": 0.0245, + "step": 55390 + }, + { + "epoch": 10.75530314062424, + "grad_norm": 0.11994529515504837, + "learning_rate": 4.521393009552549e-05, + "loss": 0.0261, + "step": 55400 + }, + { + "epoch": 10.757244793942041, + "grad_norm": 0.11567891389131546, + "learning_rate": 4.5197471583961484e-05, + "loss": 0.0211, + "step": 55410 + }, + { + "epoch": 10.759186447259841, + "grad_norm": 0.21428021788597107, + "learning_rate": 4.51810135975948e-05, + "loss": 0.0208, + "step": 55420 + }, + { + "epoch": 10.761128100577642, + "grad_norm": 0.13422709703445435, + "learning_rate": 4.516455613822521e-05, + "loss": 0.0245, + "step": 55430 + }, + { + "epoch": 10.763069753895442, + "grad_norm": 0.2253941297531128, + "learning_rate": 4.514809920765249e-05, + "loss": 0.0258, + "step": 55440 + }, + { + "epoch": 10.765011407213242, + "grad_norm": 0.10325181484222412, + "learning_rate": 4.5131642807676356e-05, + "loss": 0.0252, + "step": 55450 + }, + { + "epoch": 10.766953060531042, + "grad_norm": 0.11227639764547348, + "learning_rate": 4.511518694009644e-05, + "loss": 0.0235, + "step": 55460 + }, + { + "epoch": 10.768894713848843, + "grad_norm": 0.12500697374343872, + "learning_rate": 4.509873160671231e-05, + "loss": 0.032, + "step": 55470 + }, + { + "epoch": 10.770836367166643, + "grad_norm": 0.17388689517974854, + "learning_rate": 4.508227680932356e-05, + "loss": 0.0233, + "step": 55480 + }, + { + "epoch": 10.772778020484443, + "grad_norm": 0.1694231480360031, + "learning_rate": 4.5065822549729605e-05, + "loss": 0.0244, + "step": 55490 + }, + { + "epoch": 10.774719673802242, + "grad_norm": 0.27666226029396057, + "learning_rate": 4.504936882972989e-05, + "loss": 0.0242, + "step": 55500 + }, + { + "epoch": 10.776661327120042, + "grad_norm": 0.2578411400318146, + "learning_rate": 4.503291565112374e-05, + "loss": 0.026, + "step": 55510 + }, + { + "epoch": 10.778602980437842, + "grad_norm": 0.22344322502613068, + "learning_rate": 4.501646301571048e-05, + "loss": 0.0261, + "step": 55520 + }, + { + "epoch": 10.780544633755643, + "grad_norm": 0.17404869198799133, + "learning_rate": 4.500001092528933e-05, + "loss": 0.0304, + "step": 55530 + }, + { + "epoch": 10.782486287073443, + "grad_norm": 0.14670245349407196, + "learning_rate": 4.4983559381659455e-05, + "loss": 0.0243, + "step": 55540 + }, + { + "epoch": 10.784427940391243, + "grad_norm": 0.1636301726102829, + "learning_rate": 4.496710838661999e-05, + "loss": 0.0276, + "step": 55550 + }, + { + "epoch": 10.786369593709043, + "grad_norm": 0.29113680124282837, + "learning_rate": 4.495065794196999e-05, + "loss": 0.0274, + "step": 55560 + }, + { + "epoch": 10.788311247026844, + "grad_norm": 0.12360081821680069, + "learning_rate": 4.4934208049508455e-05, + "loss": 0.0222, + "step": 55570 + }, + { + "epoch": 10.790252900344644, + "grad_norm": 0.11365436762571335, + "learning_rate": 4.4917758711034303e-05, + "loss": 0.0232, + "step": 55580 + }, + { + "epoch": 10.792194553662444, + "grad_norm": 0.1953737884759903, + "learning_rate": 4.490130992834642e-05, + "loss": 0.0263, + "step": 55590 + }, + { + "epoch": 10.794136206980244, + "grad_norm": 0.12030297517776489, + "learning_rate": 4.488486170324359e-05, + "loss": 0.0277, + "step": 55600 + }, + { + "epoch": 10.796077860298045, + "grad_norm": 0.1413915902376175, + "learning_rate": 4.486841403752463e-05, + "loss": 0.0239, + "step": 55610 + }, + { + "epoch": 10.798019513615843, + "grad_norm": 0.21282178163528442, + "learning_rate": 4.4851966932988174e-05, + "loss": 0.0345, + "step": 55620 + }, + { + "epoch": 10.799961166933643, + "grad_norm": 0.17479072511196136, + "learning_rate": 4.4835520391432894e-05, + "loss": 0.0242, + "step": 55630 + }, + { + "epoch": 10.801902820251444, + "grad_norm": 0.13246972858905792, + "learning_rate": 4.481907441465733e-05, + "loss": 0.023, + "step": 55640 + }, + { + "epoch": 10.803844473569244, + "grad_norm": 0.23911236226558685, + "learning_rate": 4.480262900446e-05, + "loss": 0.0229, + "step": 55650 + }, + { + "epoch": 10.805786126887044, + "grad_norm": 0.1553921103477478, + "learning_rate": 4.478618416263933e-05, + "loss": 0.0255, + "step": 55660 + }, + { + "epoch": 10.807727780204845, + "grad_norm": 0.12034262716770172, + "learning_rate": 4.476973989099373e-05, + "loss": 0.0201, + "step": 55670 + }, + { + "epoch": 10.809669433522645, + "grad_norm": 0.13328352570533752, + "learning_rate": 4.475329619132153e-05, + "loss": 0.0252, + "step": 55680 + }, + { + "epoch": 10.811611086840445, + "grad_norm": 0.14244890213012695, + "learning_rate": 4.4736853065420956e-05, + "loss": 0.0286, + "step": 55690 + }, + { + "epoch": 10.813552740158245, + "grad_norm": 0.16051435470581055, + "learning_rate": 4.472041051509023e-05, + "loss": 0.0274, + "step": 55700 + }, + { + "epoch": 10.815494393476044, + "grad_norm": 0.1545262187719345, + "learning_rate": 4.470396854212746e-05, + "loss": 0.024, + "step": 55710 + }, + { + "epoch": 10.817436046793844, + "grad_norm": 0.15279768407344818, + "learning_rate": 4.468752714833072e-05, + "loss": 0.0235, + "step": 55720 + }, + { + "epoch": 10.819377700111644, + "grad_norm": 0.2574901580810547, + "learning_rate": 4.467108633549803e-05, + "loss": 0.0248, + "step": 55730 + }, + { + "epoch": 10.821319353429445, + "grad_norm": 0.2181915044784546, + "learning_rate": 4.4654646105427335e-05, + "loss": 0.025, + "step": 55740 + }, + { + "epoch": 10.823261006747245, + "grad_norm": 0.11168371140956879, + "learning_rate": 4.463820645991651e-05, + "loss": 0.0243, + "step": 55750 + }, + { + "epoch": 10.825202660065045, + "grad_norm": 0.1771726906299591, + "learning_rate": 4.462176740076334e-05, + "loss": 0.0215, + "step": 55760 + }, + { + "epoch": 10.827144313382846, + "grad_norm": 0.1385628581047058, + "learning_rate": 4.4605328929765624e-05, + "loss": 0.025, + "step": 55770 + }, + { + "epoch": 10.829085966700646, + "grad_norm": 0.17029419541358948, + "learning_rate": 4.458889104872102e-05, + "loss": 0.0267, + "step": 55780 + }, + { + "epoch": 10.831027620018446, + "grad_norm": 0.15406474471092224, + "learning_rate": 4.457245375942715e-05, + "loss": 0.0246, + "step": 55790 + }, + { + "epoch": 10.832969273336246, + "grad_norm": 0.1739126443862915, + "learning_rate": 4.455601706368157e-05, + "loss": 0.0293, + "step": 55800 + }, + { + "epoch": 10.834910926654047, + "grad_norm": 0.2981250584125519, + "learning_rate": 4.453958096328179e-05, + "loss": 0.0227, + "step": 55810 + }, + { + "epoch": 10.836852579971847, + "grad_norm": 0.1807013303041458, + "learning_rate": 4.452314546002523e-05, + "loss": 0.0269, + "step": 55820 + }, + { + "epoch": 10.838794233289645, + "grad_norm": 0.22651414573192596, + "learning_rate": 4.4506710555709246e-05, + "loss": 0.0314, + "step": 55830 + }, + { + "epoch": 10.840735886607446, + "grad_norm": 0.1773669421672821, + "learning_rate": 4.449027625213114e-05, + "loss": 0.0233, + "step": 55840 + }, + { + "epoch": 10.842677539925246, + "grad_norm": 0.1906171292066574, + "learning_rate": 4.4473842551088135e-05, + "loss": 0.025, + "step": 55850 + }, + { + "epoch": 10.844619193243046, + "grad_norm": 0.2590939402580261, + "learning_rate": 4.445740945437738e-05, + "loss": 0.0329, + "step": 55860 + }, + { + "epoch": 10.846560846560847, + "grad_norm": 0.06512361019849777, + "learning_rate": 4.444097696379602e-05, + "loss": 0.0242, + "step": 55870 + }, + { + "epoch": 10.848502499878647, + "grad_norm": 0.19095203280448914, + "learning_rate": 4.442454508114106e-05, + "loss": 0.0247, + "step": 55880 + }, + { + "epoch": 10.850444153196447, + "grad_norm": 0.13295812904834747, + "learning_rate": 4.4408113808209465e-05, + "loss": 0.0285, + "step": 55890 + }, + { + "epoch": 10.852385806514247, + "grad_norm": 0.17635750770568848, + "learning_rate": 4.439168314679813e-05, + "loss": 0.03, + "step": 55900 + }, + { + "epoch": 10.854327459832048, + "grad_norm": 0.12393493205308914, + "learning_rate": 4.43752530987039e-05, + "loss": 0.0226, + "step": 55910 + }, + { + "epoch": 10.856269113149848, + "grad_norm": 0.24200817942619324, + "learning_rate": 4.4358823665723515e-05, + "loss": 0.0237, + "step": 55920 + }, + { + "epoch": 10.858210766467646, + "grad_norm": 0.1710403561592102, + "learning_rate": 4.434239484965371e-05, + "loss": 0.0265, + "step": 55930 + }, + { + "epoch": 10.860152419785447, + "grad_norm": 0.19844278693199158, + "learning_rate": 4.4325966652291103e-05, + "loss": 0.0278, + "step": 55940 + }, + { + "epoch": 10.862094073103247, + "grad_norm": 0.12483780086040497, + "learning_rate": 4.430953907543225e-05, + "loss": 0.0262, + "step": 55950 + }, + { + "epoch": 10.864035726421047, + "grad_norm": 0.15908053517341614, + "learning_rate": 4.4293112120873645e-05, + "loss": 0.02, + "step": 55960 + }, + { + "epoch": 10.865977379738847, + "grad_norm": 0.13543425500392914, + "learning_rate": 4.427668579041172e-05, + "loss": 0.0275, + "step": 55970 + }, + { + "epoch": 10.867919033056648, + "grad_norm": 0.20784154534339905, + "learning_rate": 4.4260260085842816e-05, + "loss": 0.0272, + "step": 55980 + }, + { + "epoch": 10.869860686374448, + "grad_norm": 0.1300036907196045, + "learning_rate": 4.4243835008963256e-05, + "loss": 0.0165, + "step": 55990 + }, + { + "epoch": 10.871802339692248, + "grad_norm": 0.10483308881521225, + "learning_rate": 4.422741056156925e-05, + "loss": 0.0209, + "step": 56000 + }, + { + "epoch": 10.873743993010049, + "grad_norm": 0.13053305447101593, + "learning_rate": 4.421098674545693e-05, + "loss": 0.0215, + "step": 56010 + }, + { + "epoch": 10.875685646327849, + "grad_norm": 0.14310969412326813, + "learning_rate": 4.419456356242241e-05, + "loss": 0.0244, + "step": 56020 + }, + { + "epoch": 10.87762729964565, + "grad_norm": 0.17825600504875183, + "learning_rate": 4.4178141014261685e-05, + "loss": 0.0219, + "step": 56030 + }, + { + "epoch": 10.879568952963448, + "grad_norm": 0.10252035409212112, + "learning_rate": 4.416171910277071e-05, + "loss": 0.03, + "step": 56040 + }, + { + "epoch": 10.881510606281248, + "grad_norm": 0.12201020866632462, + "learning_rate": 4.414529782974533e-05, + "loss": 0.0298, + "step": 56050 + }, + { + "epoch": 10.883452259599048, + "grad_norm": 0.2589929699897766, + "learning_rate": 4.41288771969814e-05, + "loss": 0.027, + "step": 56060 + }, + { + "epoch": 10.885393912916848, + "grad_norm": 0.14768481254577637, + "learning_rate": 4.411245720627464e-05, + "loss": 0.023, + "step": 56070 + }, + { + "epoch": 10.887335566234649, + "grad_norm": 0.17639607191085815, + "learning_rate": 4.40960378594207e-05, + "loss": 0.0244, + "step": 56080 + }, + { + "epoch": 10.889277219552449, + "grad_norm": 0.1801632046699524, + "learning_rate": 4.407961915821519e-05, + "loss": 0.0275, + "step": 56090 + }, + { + "epoch": 10.89121887287025, + "grad_norm": 0.11338438838720322, + "learning_rate": 4.406320110445362e-05, + "loss": 0.0216, + "step": 56100 + }, + { + "epoch": 10.89316052618805, + "grad_norm": 0.16744625568389893, + "learning_rate": 4.404678369993144e-05, + "loss": 0.0236, + "step": 56110 + }, + { + "epoch": 10.89510217950585, + "grad_norm": 0.14147566258907318, + "learning_rate": 4.403036694644406e-05, + "loss": 0.0214, + "step": 56120 + }, + { + "epoch": 10.89704383282365, + "grad_norm": 0.15253737568855286, + "learning_rate": 4.4013950845786764e-05, + "loss": 0.0244, + "step": 56130 + }, + { + "epoch": 10.898985486141449, + "grad_norm": 0.08865177631378174, + "learning_rate": 4.399753539975482e-05, + "loss": 0.019, + "step": 56140 + }, + { + "epoch": 10.900927139459249, + "grad_norm": 0.0990108922123909, + "learning_rate": 4.398112061014337e-05, + "loss": 0.0211, + "step": 56150 + }, + { + "epoch": 10.90286879277705, + "grad_norm": 0.164214625954628, + "learning_rate": 4.396470647874753e-05, + "loss": 0.0291, + "step": 56160 + }, + { + "epoch": 10.90481044609485, + "grad_norm": 0.14431054890155792, + "learning_rate": 4.394829300736229e-05, + "loss": 0.0209, + "step": 56170 + }, + { + "epoch": 10.90675209941265, + "grad_norm": 0.13809499144554138, + "learning_rate": 4.393188019778265e-05, + "loss": 0.0233, + "step": 56180 + }, + { + "epoch": 10.90869375273045, + "grad_norm": 0.18529318273067474, + "learning_rate": 4.391546805180347e-05, + "loss": 0.0254, + "step": 56190 + }, + { + "epoch": 10.91063540604825, + "grad_norm": 0.1330181509256363, + "learning_rate": 4.389905657121955e-05, + "loss": 0.0298, + "step": 56200 + }, + { + "epoch": 10.91257705936605, + "grad_norm": 0.13290834426879883, + "learning_rate": 4.3882645757825635e-05, + "loss": 0.0274, + "step": 56210 + }, + { + "epoch": 10.91451871268385, + "grad_norm": 0.203896164894104, + "learning_rate": 4.386623561341637e-05, + "loss": 0.021, + "step": 56220 + }, + { + "epoch": 10.916460366001651, + "grad_norm": 0.10916180908679962, + "learning_rate": 4.384982613978637e-05, + "loss": 0.0232, + "step": 56230 + }, + { + "epoch": 10.918402019319451, + "grad_norm": 0.19621163606643677, + "learning_rate": 4.383341733873012e-05, + "loss": 0.0254, + "step": 56240 + }, + { + "epoch": 10.92034367263725, + "grad_norm": 0.12158776074647903, + "learning_rate": 4.381700921204209e-05, + "loss": 0.0258, + "step": 56250 + }, + { + "epoch": 10.92228532595505, + "grad_norm": 0.18289950489997864, + "learning_rate": 4.380060176151663e-05, + "loss": 0.022, + "step": 56260 + }, + { + "epoch": 10.92422697927285, + "grad_norm": 0.09427635371685028, + "learning_rate": 4.378419498894805e-05, + "loss": 0.0226, + "step": 56270 + }, + { + "epoch": 10.92616863259065, + "grad_norm": 0.08931384980678558, + "learning_rate": 4.376778889613056e-05, + "loss": 0.0192, + "step": 56280 + }, + { + "epoch": 10.928110285908451, + "grad_norm": 0.12370869517326355, + "learning_rate": 4.375138348485831e-05, + "loss": 0.0205, + "step": 56290 + }, + { + "epoch": 10.930051939226251, + "grad_norm": 0.1341947615146637, + "learning_rate": 4.3734978756925346e-05, + "loss": 0.026, + "step": 56300 + }, + { + "epoch": 10.931993592544051, + "grad_norm": 0.16344833374023438, + "learning_rate": 4.3718574714125706e-05, + "loss": 0.0206, + "step": 56310 + }, + { + "epoch": 10.933935245861852, + "grad_norm": 0.26004040241241455, + "learning_rate": 4.370217135825329e-05, + "loss": 0.0271, + "step": 56320 + }, + { + "epoch": 10.935876899179652, + "grad_norm": 0.23639072477817535, + "learning_rate": 4.368576869110194e-05, + "loss": 0.0248, + "step": 56330 + }, + { + "epoch": 10.937818552497452, + "grad_norm": 0.22128701210021973, + "learning_rate": 4.366936671446544e-05, + "loss": 0.0285, + "step": 56340 + }, + { + "epoch": 10.93976020581525, + "grad_norm": 0.1296374350786209, + "learning_rate": 4.3652965430137474e-05, + "loss": 0.0204, + "step": 56350 + }, + { + "epoch": 10.941701859133051, + "grad_norm": 0.18867704272270203, + "learning_rate": 4.3636564839911646e-05, + "loss": 0.019, + "step": 56360 + }, + { + "epoch": 10.943643512450851, + "grad_norm": 0.1404188722372055, + "learning_rate": 4.3620164945581545e-05, + "loss": 0.0248, + "step": 56370 + }, + { + "epoch": 10.945585165768652, + "grad_norm": 0.12424081563949585, + "learning_rate": 4.360376574894061e-05, + "loss": 0.019, + "step": 56380 + }, + { + "epoch": 10.947526819086452, + "grad_norm": 0.20112213492393494, + "learning_rate": 4.358736725178224e-05, + "loss": 0.0179, + "step": 56390 + }, + { + "epoch": 10.949468472404252, + "grad_norm": 0.18454039096832275, + "learning_rate": 4.357096945589974e-05, + "loss": 0.0272, + "step": 56400 + }, + { + "epoch": 10.951410125722052, + "grad_norm": 0.17318622767925262, + "learning_rate": 4.355457236308636e-05, + "loss": 0.0322, + "step": 56410 + }, + { + "epoch": 10.953351779039853, + "grad_norm": 0.22169248759746552, + "learning_rate": 4.353817597513526e-05, + "loss": 0.0248, + "step": 56420 + }, + { + "epoch": 10.955293432357653, + "grad_norm": 0.19795450568199158, + "learning_rate": 4.352178029383948e-05, + "loss": 0.0278, + "step": 56430 + }, + { + "epoch": 10.957235085675453, + "grad_norm": 0.266185462474823, + "learning_rate": 4.3505385320992105e-05, + "loss": 0.02, + "step": 56440 + }, + { + "epoch": 10.959176738993254, + "grad_norm": 0.14077693223953247, + "learning_rate": 4.348899105838602e-05, + "loss": 0.0332, + "step": 56450 + }, + { + "epoch": 10.961118392311054, + "grad_norm": 0.2982840836048126, + "learning_rate": 4.3472597507814087e-05, + "loss": 0.03, + "step": 56460 + }, + { + "epoch": 10.963060045628852, + "grad_norm": 0.11789120733737946, + "learning_rate": 4.3456204671069066e-05, + "loss": 0.0248, + "step": 56470 + }, + { + "epoch": 10.965001698946653, + "grad_norm": 0.18211308121681213, + "learning_rate": 4.343981254994367e-05, + "loss": 0.0228, + "step": 56480 + }, + { + "epoch": 10.966943352264453, + "grad_norm": 0.10491756349802017, + "learning_rate": 4.34234211462305e-05, + "loss": 0.021, + "step": 56490 + }, + { + "epoch": 10.968885005582253, + "grad_norm": 0.16184741258621216, + "learning_rate": 4.340703046172213e-05, + "loss": 0.0251, + "step": 56500 + }, + { + "epoch": 10.970826658900053, + "grad_norm": 0.19279702007770538, + "learning_rate": 4.339064049821097e-05, + "loss": 0.0278, + "step": 56510 + }, + { + "epoch": 10.972768312217854, + "grad_norm": 0.27567100524902344, + "learning_rate": 4.3374251257489446e-05, + "loss": 0.0289, + "step": 56520 + }, + { + "epoch": 10.974709965535654, + "grad_norm": 0.19364173710346222, + "learning_rate": 4.335786274134984e-05, + "loss": 0.0262, + "step": 56530 + }, + { + "epoch": 10.976651618853454, + "grad_norm": 0.22852970659732819, + "learning_rate": 4.334147495158438e-05, + "loss": 0.0265, + "step": 56540 + }, + { + "epoch": 10.978593272171254, + "grad_norm": 0.1381288468837738, + "learning_rate": 4.33250878899852e-05, + "loss": 0.0227, + "step": 56550 + }, + { + "epoch": 10.980534925489055, + "grad_norm": 0.1490766704082489, + "learning_rate": 4.330870155834437e-05, + "loss": 0.0215, + "step": 56560 + }, + { + "epoch": 10.982476578806853, + "grad_norm": 0.13038696348667145, + "learning_rate": 4.3292315958453884e-05, + "loss": 0.0204, + "step": 56570 + }, + { + "epoch": 10.984418232124654, + "grad_norm": 0.14761461317539215, + "learning_rate": 4.3275931092105646e-05, + "loss": 0.0215, + "step": 56580 + }, + { + "epoch": 10.986359885442454, + "grad_norm": 0.15992440283298492, + "learning_rate": 4.325954696109147e-05, + "loss": 0.0242, + "step": 56590 + }, + { + "epoch": 10.988301538760254, + "grad_norm": 0.13682767748832703, + "learning_rate": 4.32431635672031e-05, + "loss": 0.0202, + "step": 56600 + }, + { + "epoch": 10.990243192078054, + "grad_norm": 0.15865974128246307, + "learning_rate": 4.3226780912232214e-05, + "loss": 0.0265, + "step": 56610 + }, + { + "epoch": 10.992184845395855, + "grad_norm": 0.19116376340389252, + "learning_rate": 4.321039899797036e-05, + "loss": 0.0212, + "step": 56620 + }, + { + "epoch": 10.994126498713655, + "grad_norm": 0.18420280516147614, + "learning_rate": 4.319401782620908e-05, + "loss": 0.0247, + "step": 56630 + }, + { + "epoch": 10.996068152031455, + "grad_norm": 0.17618680000305176, + "learning_rate": 4.317763739873978e-05, + "loss": 0.0259, + "step": 56640 + }, + { + "epoch": 10.998009805349255, + "grad_norm": 0.15839320421218872, + "learning_rate": 4.31612577173538e-05, + "loss": 0.0217, + "step": 56650 + }, + { + "epoch": 10.999951458667056, + "grad_norm": 0.23249976336956024, + "learning_rate": 4.314487878384239e-05, + "loss": 0.0331, + "step": 56660 + }, + { + "epoch": 11.00174748798602, + "grad_norm": 0.2216552048921585, + "learning_rate": 4.3128500599996736e-05, + "loss": 0.0176, + "step": 56670 + }, + { + "epoch": 11.00368914130382, + "grad_norm": 0.23188433051109314, + "learning_rate": 4.31121231676079e-05, + "loss": 0.0254, + "step": 56680 + }, + { + "epoch": 11.00563079462162, + "grad_norm": 0.1543400138616562, + "learning_rate": 4.309574648846694e-05, + "loss": 0.0236, + "step": 56690 + }, + { + "epoch": 11.00757244793942, + "grad_norm": 0.142267107963562, + "learning_rate": 4.3079370564364755e-05, + "loss": 0.0244, + "step": 56700 + }, + { + "epoch": 11.00951410125722, + "grad_norm": 0.21440744400024414, + "learning_rate": 4.306299539709221e-05, + "loss": 0.0256, + "step": 56710 + }, + { + "epoch": 11.011455754575021, + "grad_norm": 0.13742071390151978, + "learning_rate": 4.3046620988440046e-05, + "loss": 0.0235, + "step": 56720 + }, + { + "epoch": 11.013397407892821, + "grad_norm": 0.29659461975097656, + "learning_rate": 4.303024734019897e-05, + "loss": 0.0225, + "step": 56730 + }, + { + "epoch": 11.015339061210621, + "grad_norm": 0.1309509426355362, + "learning_rate": 4.3013874454159565e-05, + "loss": 0.0273, + "step": 56740 + }, + { + "epoch": 11.017280714528422, + "grad_norm": 0.16269148886203766, + "learning_rate": 4.299750233211233e-05, + "loss": 0.0247, + "step": 56750 + }, + { + "epoch": 11.01922236784622, + "grad_norm": 0.19254818558692932, + "learning_rate": 4.2981130975847715e-05, + "loss": 0.0227, + "step": 56760 + }, + { + "epoch": 11.02116402116402, + "grad_norm": 0.18022890388965607, + "learning_rate": 4.2964760387156076e-05, + "loss": 0.0255, + "step": 56770 + }, + { + "epoch": 11.02310567448182, + "grad_norm": 0.0950978472828865, + "learning_rate": 4.294839056782767e-05, + "loss": 0.0253, + "step": 56780 + }, + { + "epoch": 11.025047327799621, + "grad_norm": 0.17387500405311584, + "learning_rate": 4.293202151965266e-05, + "loss": 0.0255, + "step": 56790 + }, + { + "epoch": 11.026988981117421, + "grad_norm": 0.16373705863952637, + "learning_rate": 4.291565324442115e-05, + "loss": 0.0257, + "step": 56800 + }, + { + "epoch": 11.028930634435222, + "grad_norm": 0.18411797285079956, + "learning_rate": 4.2899285743923135e-05, + "loss": 0.0189, + "step": 56810 + }, + { + "epoch": 11.030872287753022, + "grad_norm": 0.23705141246318817, + "learning_rate": 4.288291901994857e-05, + "loss": 0.0297, + "step": 56820 + }, + { + "epoch": 11.032813941070822, + "grad_norm": 0.17799542844295502, + "learning_rate": 4.286655307428729e-05, + "loss": 0.0258, + "step": 56830 + }, + { + "epoch": 11.034755594388622, + "grad_norm": 0.22338061034679413, + "learning_rate": 4.285018790872903e-05, + "loss": 0.0285, + "step": 56840 + }, + { + "epoch": 11.036697247706423, + "grad_norm": 0.20597554743289948, + "learning_rate": 4.283382352506347e-05, + "loss": 0.0235, + "step": 56850 + }, + { + "epoch": 11.038638901024223, + "grad_norm": 0.10739754140377045, + "learning_rate": 4.2817459925080205e-05, + "loss": 0.022, + "step": 56860 + }, + { + "epoch": 11.040580554342021, + "grad_norm": 0.1966058760881424, + "learning_rate": 4.28010971105687e-05, + "loss": 0.0261, + "step": 56870 + }, + { + "epoch": 11.042522207659822, + "grad_norm": 0.1840880662202835, + "learning_rate": 4.278473508331841e-05, + "loss": 0.0278, + "step": 56880 + }, + { + "epoch": 11.044463860977622, + "grad_norm": 0.1528264433145523, + "learning_rate": 4.276837384511864e-05, + "loss": 0.0312, + "step": 56890 + }, + { + "epoch": 11.046405514295422, + "grad_norm": 0.2581161558628082, + "learning_rate": 4.275201339775864e-05, + "loss": 0.0245, + "step": 56900 + }, + { + "epoch": 11.048347167613223, + "grad_norm": 0.17296920716762543, + "learning_rate": 4.273565374302756e-05, + "loss": 0.0191, + "step": 56910 + }, + { + "epoch": 11.050288820931023, + "grad_norm": 0.21464723348617554, + "learning_rate": 4.2719294882714454e-05, + "loss": 0.0276, + "step": 56920 + }, + { + "epoch": 11.052230474248823, + "grad_norm": 0.1849418580532074, + "learning_rate": 4.270293681860831e-05, + "loss": 0.0287, + "step": 56930 + }, + { + "epoch": 11.054172127566623, + "grad_norm": 0.12664812803268433, + "learning_rate": 4.268657955249802e-05, + "loss": 0.0205, + "step": 56940 + }, + { + "epoch": 11.056113780884424, + "grad_norm": 0.16246499121189117, + "learning_rate": 4.267022308617241e-05, + "loss": 0.0197, + "step": 56950 + }, + { + "epoch": 11.058055434202224, + "grad_norm": 0.15575774013996124, + "learning_rate": 4.265386742142018e-05, + "loss": 0.0293, + "step": 56960 + }, + { + "epoch": 11.059997087520022, + "grad_norm": 0.15584060549736023, + "learning_rate": 4.263751256002996e-05, + "loss": 0.0258, + "step": 56970 + }, + { + "epoch": 11.061938740837823, + "grad_norm": 0.08256980776786804, + "learning_rate": 4.26211585037903e-05, + "loss": 0.0266, + "step": 56980 + }, + { + "epoch": 11.063880394155623, + "grad_norm": 0.23149719834327698, + "learning_rate": 4.2604805254489656e-05, + "loss": 0.0315, + "step": 56990 + }, + { + "epoch": 11.065822047473423, + "grad_norm": 0.15892896056175232, + "learning_rate": 4.258845281391638e-05, + "loss": 0.0235, + "step": 57000 + }, + { + "epoch": 11.067763700791224, + "grad_norm": 0.318860799074173, + "learning_rate": 4.257210118385876e-05, + "loss": 0.0247, + "step": 57010 + }, + { + "epoch": 11.069705354109024, + "grad_norm": 0.1456792950630188, + "learning_rate": 4.2555750366104994e-05, + "loss": 0.0198, + "step": 57020 + }, + { + "epoch": 11.071647007426824, + "grad_norm": 0.44949087500572205, + "learning_rate": 4.2539400362443184e-05, + "loss": 0.0254, + "step": 57030 + }, + { + "epoch": 11.073588660744624, + "grad_norm": 0.16443884372711182, + "learning_rate": 4.252305117466133e-05, + "loss": 0.0234, + "step": 57040 + }, + { + "epoch": 11.075530314062425, + "grad_norm": 0.14189191162586212, + "learning_rate": 4.2506702804547356e-05, + "loss": 0.0256, + "step": 57050 + }, + { + "epoch": 11.077471967380225, + "grad_norm": 0.2744057774543762, + "learning_rate": 4.249035525388907e-05, + "loss": 0.0291, + "step": 57060 + }, + { + "epoch": 11.079413620698025, + "grad_norm": 0.11909061670303345, + "learning_rate": 4.247400852447427e-05, + "loss": 0.0218, + "step": 57070 + }, + { + "epoch": 11.081355274015824, + "grad_norm": 0.2371351420879364, + "learning_rate": 4.245766261809059e-05, + "loss": 0.0342, + "step": 57080 + }, + { + "epoch": 11.083296927333624, + "grad_norm": 0.26812633872032166, + "learning_rate": 4.244131753652558e-05, + "loss": 0.0298, + "step": 57090 + }, + { + "epoch": 11.085238580651424, + "grad_norm": 0.19832521677017212, + "learning_rate": 4.242497328156672e-05, + "loss": 0.023, + "step": 57100 + }, + { + "epoch": 11.087180233969224, + "grad_norm": 0.14666493237018585, + "learning_rate": 4.240862985500139e-05, + "loss": 0.0196, + "step": 57110 + }, + { + "epoch": 11.089121887287025, + "grad_norm": 0.122244693338871, + "learning_rate": 4.2392287258616894e-05, + "loss": 0.0236, + "step": 57120 + }, + { + "epoch": 11.091063540604825, + "grad_norm": 0.14569760859012604, + "learning_rate": 4.2375945494200396e-05, + "loss": 0.0207, + "step": 57130 + }, + { + "epoch": 11.093005193922625, + "grad_norm": 0.2489551454782486, + "learning_rate": 4.2359604563539054e-05, + "loss": 0.0276, + "step": 57140 + }, + { + "epoch": 11.094946847240426, + "grad_norm": 0.10197612643241882, + "learning_rate": 4.2343264468419885e-05, + "loss": 0.0202, + "step": 57150 + }, + { + "epoch": 11.096888500558226, + "grad_norm": 0.17277653515338898, + "learning_rate": 4.232692521062979e-05, + "loss": 0.0306, + "step": 57160 + }, + { + "epoch": 11.098830153876026, + "grad_norm": 0.11940374225378036, + "learning_rate": 4.231058679195562e-05, + "loss": 0.0221, + "step": 57170 + }, + { + "epoch": 11.100771807193826, + "grad_norm": 0.16777993738651276, + "learning_rate": 4.229424921418411e-05, + "loss": 0.026, + "step": 57180 + }, + { + "epoch": 11.102713460511625, + "grad_norm": 0.2864263951778412, + "learning_rate": 4.2277912479101916e-05, + "loss": 0.0236, + "step": 57190 + }, + { + "epoch": 11.104655113829425, + "grad_norm": 0.19606108963489532, + "learning_rate": 4.226157658849562e-05, + "loss": 0.0278, + "step": 57200 + }, + { + "epoch": 11.106596767147225, + "grad_norm": 0.20294462144374847, + "learning_rate": 4.2245241544151664e-05, + "loss": 0.0234, + "step": 57210 + }, + { + "epoch": 11.108538420465026, + "grad_norm": 0.18713633716106415, + "learning_rate": 4.2228907347856425e-05, + "loss": 0.0254, + "step": 57220 + }, + { + "epoch": 11.110480073782826, + "grad_norm": 0.1719260811805725, + "learning_rate": 4.2212574001396214e-05, + "loss": 0.021, + "step": 57230 + }, + { + "epoch": 11.112421727100626, + "grad_norm": 0.19203662872314453, + "learning_rate": 4.21962415065572e-05, + "loss": 0.0309, + "step": 57240 + }, + { + "epoch": 11.114363380418427, + "grad_norm": 0.2586146593093872, + "learning_rate": 4.217990986512548e-05, + "loss": 0.0305, + "step": 57250 + }, + { + "epoch": 11.116305033736227, + "grad_norm": 0.10655723512172699, + "learning_rate": 4.216357907888704e-05, + "loss": 0.0226, + "step": 57260 + }, + { + "epoch": 11.118246687054027, + "grad_norm": 0.20272178947925568, + "learning_rate": 4.2147249149627824e-05, + "loss": 0.0277, + "step": 57270 + }, + { + "epoch": 11.120188340371827, + "grad_norm": 0.11329931765794754, + "learning_rate": 4.213092007913364e-05, + "loss": 0.0199, + "step": 57280 + }, + { + "epoch": 11.122129993689626, + "grad_norm": 0.1503034234046936, + "learning_rate": 4.2114591869190205e-05, + "loss": 0.0252, + "step": 57290 + }, + { + "epoch": 11.124071647007426, + "grad_norm": 0.1797308772802353, + "learning_rate": 4.209826452158315e-05, + "loss": 0.0223, + "step": 57300 + }, + { + "epoch": 11.126013300325226, + "grad_norm": 0.19195757806301117, + "learning_rate": 4.2081938038098e-05, + "loss": 0.0234, + "step": 57310 + }, + { + "epoch": 11.127954953643027, + "grad_norm": 0.15464307367801666, + "learning_rate": 4.2065612420520185e-05, + "loss": 0.0278, + "step": 57320 + }, + { + "epoch": 11.129896606960827, + "grad_norm": 0.19721141457557678, + "learning_rate": 4.204928767063509e-05, + "loss": 0.0258, + "step": 57330 + }, + { + "epoch": 11.131838260278627, + "grad_norm": 0.2911233603954315, + "learning_rate": 4.2032963790227944e-05, + "loss": 0.0254, + "step": 57340 + }, + { + "epoch": 11.133779913596427, + "grad_norm": 0.149281308054924, + "learning_rate": 4.2016640781083887e-05, + "loss": 0.028, + "step": 57350 + }, + { + "epoch": 11.135721566914228, + "grad_norm": 0.2325287014245987, + "learning_rate": 4.2000318644988004e-05, + "loss": 0.0319, + "step": 57360 + }, + { + "epoch": 11.137663220232028, + "grad_norm": 0.16327835619449615, + "learning_rate": 4.198399738372524e-05, + "loss": 0.0255, + "step": 57370 + }, + { + "epoch": 11.139604873549828, + "grad_norm": 0.17526185512542725, + "learning_rate": 4.196767699908045e-05, + "loss": 0.0307, + "step": 57380 + }, + { + "epoch": 11.141546526867629, + "grad_norm": 0.14995355904102325, + "learning_rate": 4.195135749283845e-05, + "loss": 0.0225, + "step": 57390 + }, + { + "epoch": 11.143488180185427, + "grad_norm": 0.15403051674365997, + "learning_rate": 4.1935038866783894e-05, + "loss": 0.0217, + "step": 57400 + }, + { + "epoch": 11.145429833503227, + "grad_norm": 0.22494889795780182, + "learning_rate": 4.1918721122701355e-05, + "loss": 0.022, + "step": 57410 + }, + { + "epoch": 11.147371486821028, + "grad_norm": 0.1374790370464325, + "learning_rate": 4.190240426237533e-05, + "loss": 0.0241, + "step": 57420 + }, + { + "epoch": 11.149313140138828, + "grad_norm": 0.18619632720947266, + "learning_rate": 4.188608828759019e-05, + "loss": 0.0282, + "step": 57430 + }, + { + "epoch": 11.151254793456628, + "grad_norm": 0.17070753872394562, + "learning_rate": 4.1869773200130246e-05, + "loss": 0.0252, + "step": 57440 + }, + { + "epoch": 11.153196446774428, + "grad_norm": 0.07204219698905945, + "learning_rate": 4.185345900177967e-05, + "loss": 0.0234, + "step": 57450 + }, + { + "epoch": 11.155138100092229, + "grad_norm": 0.20455537736415863, + "learning_rate": 4.183714569432258e-05, + "loss": 0.0269, + "step": 57460 + }, + { + "epoch": 11.157079753410029, + "grad_norm": 0.1541885882616043, + "learning_rate": 4.182083327954295e-05, + "loss": 0.0235, + "step": 57470 + }, + { + "epoch": 11.15902140672783, + "grad_norm": 0.19924448430538177, + "learning_rate": 4.1804521759224714e-05, + "loss": 0.0216, + "step": 57480 + }, + { + "epoch": 11.16096306004563, + "grad_norm": 0.14329935610294342, + "learning_rate": 4.178821113515165e-05, + "loss": 0.0245, + "step": 57490 + }, + { + "epoch": 11.16290471336343, + "grad_norm": 0.19720128178596497, + "learning_rate": 4.1771901409107476e-05, + "loss": 0.0206, + "step": 57500 + }, + { + "epoch": 11.164846366681228, + "grad_norm": 0.157124862074852, + "learning_rate": 4.175559258287577e-05, + "loss": 0.0222, + "step": 57510 + }, + { + "epoch": 11.166788019999029, + "grad_norm": 0.13846611976623535, + "learning_rate": 4.173928465824008e-05, + "loss": 0.0269, + "step": 57520 + }, + { + "epoch": 11.168729673316829, + "grad_norm": 0.1999945342540741, + "learning_rate": 4.17229776369838e-05, + "loss": 0.0281, + "step": 57530 + }, + { + "epoch": 11.17067132663463, + "grad_norm": 0.12509697675704956, + "learning_rate": 4.170667152089025e-05, + "loss": 0.0241, + "step": 57540 + }, + { + "epoch": 11.17261297995243, + "grad_norm": 0.36553117632865906, + "learning_rate": 4.1690366311742626e-05, + "loss": 0.0298, + "step": 57550 + }, + { + "epoch": 11.17455463327023, + "grad_norm": 0.11722781509160995, + "learning_rate": 4.1674062011324055e-05, + "loss": 0.0277, + "step": 57560 + }, + { + "epoch": 11.17649628658803, + "grad_norm": 0.20951442420482635, + "learning_rate": 4.165775862141752e-05, + "loss": 0.024, + "step": 57570 + }, + { + "epoch": 11.17843793990583, + "grad_norm": 0.20283809304237366, + "learning_rate": 4.164145614380598e-05, + "loss": 0.0215, + "step": 57580 + }, + { + "epoch": 11.18037959322363, + "grad_norm": 0.1738319993019104, + "learning_rate": 4.162515458027223e-05, + "loss": 0.0311, + "step": 57590 + }, + { + "epoch": 11.18232124654143, + "grad_norm": 0.1751876026391983, + "learning_rate": 4.160885393259899e-05, + "loss": 0.0236, + "step": 57600 + }, + { + "epoch": 11.18426289985923, + "grad_norm": 0.1910989135503769, + "learning_rate": 4.159255420256886e-05, + "loss": 0.028, + "step": 57610 + }, + { + "epoch": 11.18620455317703, + "grad_norm": 0.2002355456352234, + "learning_rate": 4.157625539196436e-05, + "loss": 0.024, + "step": 57620 + }, + { + "epoch": 11.18814620649483, + "grad_norm": 0.24997973442077637, + "learning_rate": 4.155995750256791e-05, + "loss": 0.0212, + "step": 57630 + }, + { + "epoch": 11.19008785981263, + "grad_norm": 0.10767284035682678, + "learning_rate": 4.15436605361618e-05, + "loss": 0.023, + "step": 57640 + }, + { + "epoch": 11.19202951313043, + "grad_norm": 0.20093192160129547, + "learning_rate": 4.152736449452827e-05, + "loss": 0.0224, + "step": 57650 + }, + { + "epoch": 11.19397116644823, + "grad_norm": 0.145492285490036, + "learning_rate": 4.151106937944943e-05, + "loss": 0.0173, + "step": 57660 + }, + { + "epoch": 11.195912819766031, + "grad_norm": 0.15802937746047974, + "learning_rate": 4.149477519270727e-05, + "loss": 0.022, + "step": 57670 + }, + { + "epoch": 11.197854473083831, + "grad_norm": 0.09694857895374298, + "learning_rate": 4.147848193608371e-05, + "loss": 0.021, + "step": 57680 + }, + { + "epoch": 11.199796126401631, + "grad_norm": 0.16820022463798523, + "learning_rate": 4.146218961136056e-05, + "loss": 0.0207, + "step": 57690 + }, + { + "epoch": 11.201737779719432, + "grad_norm": 0.22163017094135284, + "learning_rate": 4.14458982203195e-05, + "loss": 0.0244, + "step": 57700 + }, + { + "epoch": 11.203679433037232, + "grad_norm": 0.2758379578590393, + "learning_rate": 4.1429607764742165e-05, + "loss": 0.0259, + "step": 57710 + }, + { + "epoch": 11.20562108635503, + "grad_norm": 0.14237381517887115, + "learning_rate": 4.141331824641003e-05, + "loss": 0.0281, + "step": 57720 + }, + { + "epoch": 11.20756273967283, + "grad_norm": 0.14157697558403015, + "learning_rate": 4.139702966710452e-05, + "loss": 0.0215, + "step": 57730 + }, + { + "epoch": 11.209504392990631, + "grad_norm": 0.18142327666282654, + "learning_rate": 4.13807420286069e-05, + "loss": 0.0263, + "step": 57740 + }, + { + "epoch": 11.211446046308431, + "grad_norm": 0.13671353459358215, + "learning_rate": 4.136445533269837e-05, + "loss": 0.0319, + "step": 57750 + }, + { + "epoch": 11.213387699626232, + "grad_norm": 0.0987921804189682, + "learning_rate": 4.134816958116003e-05, + "loss": 0.0244, + "step": 57760 + }, + { + "epoch": 11.215329352944032, + "grad_norm": 0.1391935646533966, + "learning_rate": 4.133188477577283e-05, + "loss": 0.0211, + "step": 57770 + }, + { + "epoch": 11.217271006261832, + "grad_norm": 0.1500079184770584, + "learning_rate": 4.131560091831771e-05, + "loss": 0.0243, + "step": 57780 + }, + { + "epoch": 11.219212659579632, + "grad_norm": 0.14570681750774384, + "learning_rate": 4.1299318010575405e-05, + "loss": 0.0257, + "step": 57790 + }, + { + "epoch": 11.221154312897433, + "grad_norm": 0.16036154329776764, + "learning_rate": 4.128303605432661e-05, + "loss": 0.0275, + "step": 57800 + }, + { + "epoch": 11.223095966215233, + "grad_norm": 0.17841407656669617, + "learning_rate": 4.126675505135188e-05, + "loss": 0.0238, + "step": 57810 + }, + { + "epoch": 11.225037619533033, + "grad_norm": 0.19139231741428375, + "learning_rate": 4.125047500343169e-05, + "loss": 0.0209, + "step": 57820 + }, + { + "epoch": 11.226979272850832, + "grad_norm": 0.1340862214565277, + "learning_rate": 4.1234195912346376e-05, + "loss": 0.0263, + "step": 57830 + }, + { + "epoch": 11.228920926168632, + "grad_norm": 0.12932001054286957, + "learning_rate": 4.1217917779876235e-05, + "loss": 0.0205, + "step": 57840 + }, + { + "epoch": 11.230862579486432, + "grad_norm": 0.12589719891548157, + "learning_rate": 4.120164060780141e-05, + "loss": 0.026, + "step": 57850 + }, + { + "epoch": 11.232804232804233, + "grad_norm": 0.21346518397331238, + "learning_rate": 4.1185364397901935e-05, + "loss": 0.0313, + "step": 57860 + }, + { + "epoch": 11.234745886122033, + "grad_norm": 0.16716903448104858, + "learning_rate": 4.116908915195776e-05, + "loss": 0.0285, + "step": 57870 + }, + { + "epoch": 11.236687539439833, + "grad_norm": 0.07832382619380951, + "learning_rate": 4.115281487174872e-05, + "loss": 0.02, + "step": 57880 + }, + { + "epoch": 11.238629192757633, + "grad_norm": 0.1372905671596527, + "learning_rate": 4.113654155905452e-05, + "loss": 0.0194, + "step": 57890 + }, + { + "epoch": 11.240570846075434, + "grad_norm": 0.25666543841362, + "learning_rate": 4.112026921565483e-05, + "loss": 0.0204, + "step": 57900 + }, + { + "epoch": 11.242512499393234, + "grad_norm": 0.1845369189977646, + "learning_rate": 4.1103997843329147e-05, + "loss": 0.0232, + "step": 57910 + }, + { + "epoch": 11.244454152711034, + "grad_norm": 0.19518250226974487, + "learning_rate": 4.108772744385689e-05, + "loss": 0.021, + "step": 57920 + }, + { + "epoch": 11.246395806028833, + "grad_norm": 0.3107609152793884, + "learning_rate": 4.1071458019017346e-05, + "loss": 0.0271, + "step": 57930 + }, + { + "epoch": 11.248337459346633, + "grad_norm": 0.13134035468101501, + "learning_rate": 4.1055189570589734e-05, + "loss": 0.0217, + "step": 57940 + }, + { + "epoch": 11.250279112664433, + "grad_norm": 0.1291692554950714, + "learning_rate": 4.103892210035315e-05, + "loss": 0.0211, + "step": 57950 + }, + { + "epoch": 11.252220765982234, + "grad_norm": 0.16067823767662048, + "learning_rate": 4.102265561008656e-05, + "loss": 0.0245, + "step": 57960 + }, + { + "epoch": 11.254162419300034, + "grad_norm": 0.3422442674636841, + "learning_rate": 4.100639010156885e-05, + "loss": 0.0205, + "step": 57970 + }, + { + "epoch": 11.256104072617834, + "grad_norm": 0.10832452774047852, + "learning_rate": 4.099012557657882e-05, + "loss": 0.0247, + "step": 57980 + }, + { + "epoch": 11.258045725935634, + "grad_norm": 0.21449819207191467, + "learning_rate": 4.09738620368951e-05, + "loss": 0.027, + "step": 57990 + }, + { + "epoch": 11.259987379253435, + "grad_norm": 0.13470150530338287, + "learning_rate": 4.095759948429626e-05, + "loss": 0.0251, + "step": 58000 + }, + { + "epoch": 11.261929032571235, + "grad_norm": 0.10948371887207031, + "learning_rate": 4.094133792056075e-05, + "loss": 0.0259, + "step": 58010 + }, + { + "epoch": 11.263870685889035, + "grad_norm": 0.22918583452701569, + "learning_rate": 4.092507734746688e-05, + "loss": 0.0293, + "step": 58020 + }, + { + "epoch": 11.265812339206835, + "grad_norm": 0.13797235488891602, + "learning_rate": 4.090881776679293e-05, + "loss": 0.0198, + "step": 58030 + }, + { + "epoch": 11.267753992524634, + "grad_norm": 0.1294940561056137, + "learning_rate": 4.0892559180316994e-05, + "loss": 0.0238, + "step": 58040 + }, + { + "epoch": 11.269695645842434, + "grad_norm": 0.11842739582061768, + "learning_rate": 4.08763015898171e-05, + "loss": 0.0209, + "step": 58050 + }, + { + "epoch": 11.271637299160234, + "grad_norm": 0.14923523366451263, + "learning_rate": 4.0860044997071145e-05, + "loss": 0.0212, + "step": 58060 + }, + { + "epoch": 11.273578952478035, + "grad_norm": 0.18743938207626343, + "learning_rate": 4.084378940385693e-05, + "loss": 0.0224, + "step": 58070 + }, + { + "epoch": 11.275520605795835, + "grad_norm": 0.11087440699338913, + "learning_rate": 4.082753481195211e-05, + "loss": 0.0233, + "step": 58080 + }, + { + "epoch": 11.277462259113635, + "grad_norm": 0.18760648369789124, + "learning_rate": 4.0811281223134324e-05, + "loss": 0.0233, + "step": 58090 + }, + { + "epoch": 11.279403912431436, + "grad_norm": 0.20332187414169312, + "learning_rate": 4.0795028639181e-05, + "loss": 0.0221, + "step": 58100 + }, + { + "epoch": 11.281345565749236, + "grad_norm": 0.19690150022506714, + "learning_rate": 4.077877706186951e-05, + "loss": 0.0224, + "step": 58110 + }, + { + "epoch": 11.283287219067036, + "grad_norm": 0.14781305193901062, + "learning_rate": 4.076252649297709e-05, + "loss": 0.023, + "step": 58120 + }, + { + "epoch": 11.285228872384836, + "grad_norm": 0.29745858907699585, + "learning_rate": 4.07462769342809e-05, + "loss": 0.0222, + "step": 58130 + }, + { + "epoch": 11.287170525702635, + "grad_norm": 0.17214331030845642, + "learning_rate": 4.073002838755793e-05, + "loss": 0.0277, + "step": 58140 + }, + { + "epoch": 11.289112179020435, + "grad_norm": 0.13539743423461914, + "learning_rate": 4.0713780854585115e-05, + "loss": 0.0264, + "step": 58150 + }, + { + "epoch": 11.291053832338235, + "grad_norm": 0.16070358455181122, + "learning_rate": 4.069753433713927e-05, + "loss": 0.017, + "step": 58160 + }, + { + "epoch": 11.292995485656036, + "grad_norm": 0.16186366975307465, + "learning_rate": 4.068128883699709e-05, + "loss": 0.0287, + "step": 58170 + }, + { + "epoch": 11.294937138973836, + "grad_norm": 0.16240879893302917, + "learning_rate": 4.0665044355935155e-05, + "loss": 0.0298, + "step": 58180 + }, + { + "epoch": 11.296878792291636, + "grad_norm": 0.1474623680114746, + "learning_rate": 4.064880089572993e-05, + "loss": 0.0251, + "step": 58190 + }, + { + "epoch": 11.298820445609437, + "grad_norm": 0.17617784440517426, + "learning_rate": 4.063255845815779e-05, + "loss": 0.0235, + "step": 58200 + }, + { + "epoch": 11.300762098927237, + "grad_norm": 0.1437406986951828, + "learning_rate": 4.0616317044994945e-05, + "loss": 0.0237, + "step": 58210 + }, + { + "epoch": 11.302703752245037, + "grad_norm": 0.20564480125904083, + "learning_rate": 4.0600076658017585e-05, + "loss": 0.0299, + "step": 58220 + }, + { + "epoch": 11.304645405562837, + "grad_norm": 0.18799571692943573, + "learning_rate": 4.0583837299001706e-05, + "loss": 0.0228, + "step": 58230 + }, + { + "epoch": 11.306587058880638, + "grad_norm": 0.2039646953344345, + "learning_rate": 4.056759896972322e-05, + "loss": 0.0231, + "step": 58240 + }, + { + "epoch": 11.308528712198436, + "grad_norm": 0.23211026191711426, + "learning_rate": 4.055136167195793e-05, + "loss": 0.0243, + "step": 58250 + }, + { + "epoch": 11.310470365516236, + "grad_norm": 0.166098952293396, + "learning_rate": 4.053512540748152e-05, + "loss": 0.027, + "step": 58260 + }, + { + "epoch": 11.312412018834037, + "grad_norm": 0.20777136087417603, + "learning_rate": 4.051889017806955e-05, + "loss": 0.0204, + "step": 58270 + }, + { + "epoch": 11.314353672151837, + "grad_norm": 0.14493197202682495, + "learning_rate": 4.0502655985497516e-05, + "loss": 0.0235, + "step": 58280 + }, + { + "epoch": 11.316295325469637, + "grad_norm": 0.18026070296764374, + "learning_rate": 4.048642283154074e-05, + "loss": 0.0189, + "step": 58290 + }, + { + "epoch": 11.318236978787438, + "grad_norm": 0.20059382915496826, + "learning_rate": 4.0470190717974464e-05, + "loss": 0.0253, + "step": 58300 + }, + { + "epoch": 11.320178632105238, + "grad_norm": 0.1555122286081314, + "learning_rate": 4.045395964657381e-05, + "loss": 0.0218, + "step": 58310 + }, + { + "epoch": 11.322120285423038, + "grad_norm": 0.18820223212242126, + "learning_rate": 4.043772961911376e-05, + "loss": 0.0216, + "step": 58320 + }, + { + "epoch": 11.324061938740838, + "grad_norm": 0.13743123412132263, + "learning_rate": 4.042150063736924e-05, + "loss": 0.0235, + "step": 58330 + }, + { + "epoch": 11.326003592058639, + "grad_norm": 0.13443435728549957, + "learning_rate": 4.0405272703114976e-05, + "loss": 0.0211, + "step": 58340 + }, + { + "epoch": 11.327945245376439, + "grad_norm": 0.15964123606681824, + "learning_rate": 4.038904581812569e-05, + "loss": 0.0234, + "step": 58350 + }, + { + "epoch": 11.329886898694237, + "grad_norm": 0.1464417278766632, + "learning_rate": 4.03728199841759e-05, + "loss": 0.0282, + "step": 58360 + }, + { + "epoch": 11.331828552012038, + "grad_norm": 0.30307623744010925, + "learning_rate": 4.0356595203040044e-05, + "loss": 0.027, + "step": 58370 + }, + { + "epoch": 11.333770205329838, + "grad_norm": 0.11453693360090256, + "learning_rate": 4.0340371476492426e-05, + "loss": 0.019, + "step": 58380 + }, + { + "epoch": 11.335711858647638, + "grad_norm": 0.24554742872714996, + "learning_rate": 4.032414880630727e-05, + "loss": 0.0246, + "step": 58390 + }, + { + "epoch": 11.337653511965438, + "grad_norm": 0.2016633003950119, + "learning_rate": 4.0307927194258646e-05, + "loss": 0.0303, + "step": 58400 + }, + { + "epoch": 11.339595165283239, + "grad_norm": 0.1862475723028183, + "learning_rate": 4.029170664212054e-05, + "loss": 0.0209, + "step": 58410 + }, + { + "epoch": 11.341536818601039, + "grad_norm": 0.18118475377559662, + "learning_rate": 4.0275487151666794e-05, + "loss": 0.0246, + "step": 58420 + }, + { + "epoch": 11.34347847191884, + "grad_norm": 0.1465563327074051, + "learning_rate": 4.025926872467116e-05, + "loss": 0.0225, + "step": 58430 + }, + { + "epoch": 11.34542012523664, + "grad_norm": 0.16130602359771729, + "learning_rate": 4.024305136290725e-05, + "loss": 0.0254, + "step": 58440 + }, + { + "epoch": 11.34736177855444, + "grad_norm": 0.1830226331949234, + "learning_rate": 4.022683506814858e-05, + "loss": 0.0238, + "step": 58450 + }, + { + "epoch": 11.34930343187224, + "grad_norm": 0.15461114048957825, + "learning_rate": 4.021061984216853e-05, + "loss": 0.0261, + "step": 58460 + }, + { + "epoch": 11.351245085190039, + "grad_norm": 0.19656214118003845, + "learning_rate": 4.019440568674035e-05, + "loss": 0.022, + "step": 58470 + }, + { + "epoch": 11.353186738507839, + "grad_norm": 0.1532265692949295, + "learning_rate": 4.017819260363724e-05, + "loss": 0.0214, + "step": 58480 + }, + { + "epoch": 11.35512839182564, + "grad_norm": 0.11229655891656876, + "learning_rate": 4.016198059463223e-05, + "loss": 0.0249, + "step": 58490 + }, + { + "epoch": 11.35707004514344, + "grad_norm": 0.24685531854629517, + "learning_rate": 4.0145769661498233e-05, + "loss": 0.021, + "step": 58500 + }, + { + "epoch": 11.35901169846124, + "grad_norm": 0.1319931149482727, + "learning_rate": 4.012955980600804e-05, + "loss": 0.0244, + "step": 58510 + }, + { + "epoch": 11.36095335177904, + "grad_norm": 0.3190854787826538, + "learning_rate": 4.0113351029934355e-05, + "loss": 0.029, + "step": 58520 + }, + { + "epoch": 11.36289500509684, + "grad_norm": 0.17934764921665192, + "learning_rate": 4.0097143335049714e-05, + "loss": 0.0207, + "step": 58530 + }, + { + "epoch": 11.36483665841464, + "grad_norm": 0.09616737067699432, + "learning_rate": 4.0080936723126606e-05, + "loss": 0.0296, + "step": 58540 + }, + { + "epoch": 11.36677831173244, + "grad_norm": 0.2204774171113968, + "learning_rate": 4.006473119593735e-05, + "loss": 0.021, + "step": 58550 + }, + { + "epoch": 11.368719965050241, + "grad_norm": 0.18741174042224884, + "learning_rate": 4.004852675525415e-05, + "loss": 0.0256, + "step": 58560 + }, + { + "epoch": 11.37066161836804, + "grad_norm": 0.13466422259807587, + "learning_rate": 4.00323234028491e-05, + "loss": 0.0215, + "step": 58570 + }, + { + "epoch": 11.37260327168584, + "grad_norm": 0.20159055292606354, + "learning_rate": 4.0016121140494186e-05, + "loss": 0.0285, + "step": 58580 + }, + { + "epoch": 11.37454492500364, + "grad_norm": 0.13303105533123016, + "learning_rate": 3.999991996996122e-05, + "loss": 0.0219, + "step": 58590 + }, + { + "epoch": 11.37648657832144, + "grad_norm": 0.18648330867290497, + "learning_rate": 3.9983719893022e-05, + "loss": 0.0196, + "step": 58600 + }, + { + "epoch": 11.37842823163924, + "grad_norm": 0.12281802296638489, + "learning_rate": 3.996752091144812e-05, + "loss": 0.0204, + "step": 58610 + }, + { + "epoch": 11.380369884957041, + "grad_norm": 0.19280189275741577, + "learning_rate": 3.9951323027011055e-05, + "loss": 0.0267, + "step": 58620 + }, + { + "epoch": 11.382311538274841, + "grad_norm": 0.20571117103099823, + "learning_rate": 3.993512624148219e-05, + "loss": 0.0213, + "step": 58630 + }, + { + "epoch": 11.384253191592641, + "grad_norm": 0.16276773810386658, + "learning_rate": 3.99189305566328e-05, + "loss": 0.0255, + "step": 58640 + }, + { + "epoch": 11.386194844910442, + "grad_norm": 0.17863599956035614, + "learning_rate": 3.990273597423401e-05, + "loss": 0.0224, + "step": 58650 + }, + { + "epoch": 11.388136498228242, + "grad_norm": 0.19194801151752472, + "learning_rate": 3.988654249605681e-05, + "loss": 0.0349, + "step": 58660 + }, + { + "epoch": 11.390078151546042, + "grad_norm": 0.1979411542415619, + "learning_rate": 3.987035012387213e-05, + "loss": 0.0237, + "step": 58670 + }, + { + "epoch": 11.39201980486384, + "grad_norm": 0.15408076345920563, + "learning_rate": 3.9854158859450735e-05, + "loss": 0.0214, + "step": 58680 + }, + { + "epoch": 11.393961458181641, + "grad_norm": 0.16891032457351685, + "learning_rate": 3.983796870456328e-05, + "loss": 0.0225, + "step": 58690 + }, + { + "epoch": 11.395903111499441, + "grad_norm": 0.07719181478023529, + "learning_rate": 3.982177966098028e-05, + "loss": 0.0233, + "step": 58700 + }, + { + "epoch": 11.397844764817242, + "grad_norm": 0.14915068447589874, + "learning_rate": 3.980559173047216e-05, + "loss": 0.0277, + "step": 58710 + }, + { + "epoch": 11.399786418135042, + "grad_norm": 0.13980470597743988, + "learning_rate": 3.9789404914809176e-05, + "loss": 0.0217, + "step": 58720 + }, + { + "epoch": 11.401728071452842, + "grad_norm": 0.22908756136894226, + "learning_rate": 3.9773219215761544e-05, + "loss": 0.0222, + "step": 58730 + }, + { + "epoch": 11.403669724770642, + "grad_norm": 0.3157366216182709, + "learning_rate": 3.975703463509928e-05, + "loss": 0.0206, + "step": 58740 + }, + { + "epoch": 11.405611378088443, + "grad_norm": 0.17687274515628815, + "learning_rate": 3.974085117459231e-05, + "loss": 0.0198, + "step": 58750 + }, + { + "epoch": 11.407553031406243, + "grad_norm": 0.21687307953834534, + "learning_rate": 3.972466883601043e-05, + "loss": 0.0232, + "step": 58760 + }, + { + "epoch": 11.409494684724043, + "grad_norm": 0.17175713181495667, + "learning_rate": 3.9708487621123316e-05, + "loss": 0.028, + "step": 58770 + }, + { + "epoch": 11.411436338041842, + "grad_norm": 0.1771903932094574, + "learning_rate": 3.969230753170051e-05, + "loss": 0.0258, + "step": 58780 + }, + { + "epoch": 11.413377991359642, + "grad_norm": 0.1826559603214264, + "learning_rate": 3.967612856951146e-05, + "loss": 0.0217, + "step": 58790 + }, + { + "epoch": 11.415319644677442, + "grad_norm": 0.13108889758586884, + "learning_rate": 3.9659950736325476e-05, + "loss": 0.0276, + "step": 58800 + }, + { + "epoch": 11.417261297995243, + "grad_norm": 0.2999287247657776, + "learning_rate": 3.964377403391174e-05, + "loss": 0.0268, + "step": 58810 + }, + { + "epoch": 11.419202951313043, + "grad_norm": 0.12780602276325226, + "learning_rate": 3.9627598464039303e-05, + "loss": 0.0219, + "step": 58820 + }, + { + "epoch": 11.421144604630843, + "grad_norm": 0.1414797604084015, + "learning_rate": 3.9611424028477096e-05, + "loss": 0.0223, + "step": 58830 + }, + { + "epoch": 11.423086257948643, + "grad_norm": 0.16992658376693726, + "learning_rate": 3.959525072899394e-05, + "loss": 0.0208, + "step": 58840 + }, + { + "epoch": 11.425027911266444, + "grad_norm": 0.27475836873054504, + "learning_rate": 3.9579078567358506e-05, + "loss": 0.0252, + "step": 58850 + }, + { + "epoch": 11.426969564584244, + "grad_norm": 0.23451238870620728, + "learning_rate": 3.956290754533939e-05, + "loss": 0.0221, + "step": 58860 + }, + { + "epoch": 11.428911217902044, + "grad_norm": 0.21014533936977386, + "learning_rate": 3.954673766470501e-05, + "loss": 0.0266, + "step": 58870 + }, + { + "epoch": 11.430852871219845, + "grad_norm": 0.15346622467041016, + "learning_rate": 3.9530568927223674e-05, + "loss": 0.0233, + "step": 58880 + }, + { + "epoch": 11.432794524537643, + "grad_norm": 0.09613638371229172, + "learning_rate": 3.95144013346636e-05, + "loss": 0.0209, + "step": 58890 + }, + { + "epoch": 11.434736177855443, + "grad_norm": 0.21950022876262665, + "learning_rate": 3.949823488879282e-05, + "loss": 0.0257, + "step": 58900 + }, + { + "epoch": 11.436677831173244, + "grad_norm": 0.25986966490745544, + "learning_rate": 3.9482069591379276e-05, + "loss": 0.0307, + "step": 58910 + }, + { + "epoch": 11.438619484491044, + "grad_norm": 0.12610459327697754, + "learning_rate": 3.9465905444190784e-05, + "loss": 0.0241, + "step": 58920 + }, + { + "epoch": 11.440561137808844, + "grad_norm": 0.23521167039871216, + "learning_rate": 3.944974244899505e-05, + "loss": 0.0225, + "step": 58930 + }, + { + "epoch": 11.442502791126644, + "grad_norm": 0.1589908003807068, + "learning_rate": 3.9433580607559616e-05, + "loss": 0.0238, + "step": 58940 + }, + { + "epoch": 11.444444444444445, + "grad_norm": 0.2737111449241638, + "learning_rate": 3.941741992165193e-05, + "loss": 0.0251, + "step": 58950 + }, + { + "epoch": 11.446386097762245, + "grad_norm": 0.215650737285614, + "learning_rate": 3.940126039303928e-05, + "loss": 0.0252, + "step": 58960 + }, + { + "epoch": 11.448327751080045, + "grad_norm": 0.13568376004695892, + "learning_rate": 3.938510202348886e-05, + "loss": 0.0196, + "step": 58970 + }, + { + "epoch": 11.450269404397845, + "grad_norm": 0.14447475969791412, + "learning_rate": 3.93689448147677e-05, + "loss": 0.0223, + "step": 58980 + }, + { + "epoch": 11.452211057715644, + "grad_norm": 0.11326269060373306, + "learning_rate": 3.935278876864278e-05, + "loss": 0.0237, + "step": 58990 + }, + { + "epoch": 11.454152711033444, + "grad_norm": 0.18780159950256348, + "learning_rate": 3.933663388688087e-05, + "loss": 0.029, + "step": 59000 + }, + { + "epoch": 11.456094364351245, + "grad_norm": 0.14206993579864502, + "learning_rate": 3.9320480171248645e-05, + "loss": 0.0196, + "step": 59010 + }, + { + "epoch": 11.458036017669045, + "grad_norm": 0.2138024866580963, + "learning_rate": 3.930432762351265e-05, + "loss": 0.0268, + "step": 59020 + }, + { + "epoch": 11.459977670986845, + "grad_norm": 0.19245240092277527, + "learning_rate": 3.928817624543931e-05, + "loss": 0.0278, + "step": 59030 + }, + { + "epoch": 11.461919324304645, + "grad_norm": 0.19745051860809326, + "learning_rate": 3.9272026038794885e-05, + "loss": 0.0234, + "step": 59040 + }, + { + "epoch": 11.463860977622446, + "grad_norm": 0.1770852953195572, + "learning_rate": 3.9255877005345586e-05, + "loss": 0.0258, + "step": 59050 + }, + { + "epoch": 11.465802630940246, + "grad_norm": 0.2213110327720642, + "learning_rate": 3.923972914685743e-05, + "loss": 0.0264, + "step": 59060 + }, + { + "epoch": 11.467744284258046, + "grad_norm": 0.11106176674365997, + "learning_rate": 3.922358246509631e-05, + "loss": 0.0224, + "step": 59070 + }, + { + "epoch": 11.469685937575846, + "grad_norm": 0.2423941045999527, + "learning_rate": 3.920743696182801e-05, + "loss": 0.0214, + "step": 59080 + }, + { + "epoch": 11.471627590893647, + "grad_norm": 0.1962684839963913, + "learning_rate": 3.9191292638818164e-05, + "loss": 0.0227, + "step": 59090 + }, + { + "epoch": 11.473569244211445, + "grad_norm": 0.090970978140831, + "learning_rate": 3.9175149497832316e-05, + "loss": 0.0185, + "step": 59100 + }, + { + "epoch": 11.475510897529245, + "grad_norm": 0.14938025176525116, + "learning_rate": 3.915900754063584e-05, + "loss": 0.0192, + "step": 59110 + }, + { + "epoch": 11.477452550847046, + "grad_norm": 0.16736681759357452, + "learning_rate": 3.9142866768994014e-05, + "loss": 0.0294, + "step": 59120 + }, + { + "epoch": 11.479394204164846, + "grad_norm": 0.13384251296520233, + "learning_rate": 3.912672718467193e-05, + "loss": 0.0241, + "step": 59130 + }, + { + "epoch": 11.481335857482646, + "grad_norm": 0.12561936676502228, + "learning_rate": 3.911058878943463e-05, + "loss": 0.0234, + "step": 59140 + }, + { + "epoch": 11.483277510800447, + "grad_norm": 0.14539386332035065, + "learning_rate": 3.9094451585046974e-05, + "loss": 0.0213, + "step": 59150 + }, + { + "epoch": 11.485219164118247, + "grad_norm": 0.14726904034614563, + "learning_rate": 3.90783155732737e-05, + "loss": 0.0227, + "step": 59160 + }, + { + "epoch": 11.487160817436047, + "grad_norm": 0.10003631561994553, + "learning_rate": 3.9062180755879373e-05, + "loss": 0.0282, + "step": 59170 + }, + { + "epoch": 11.489102470753847, + "grad_norm": 0.14282728731632233, + "learning_rate": 3.904604713462855e-05, + "loss": 0.0273, + "step": 59180 + }, + { + "epoch": 11.491044124071648, + "grad_norm": 0.15545660257339478, + "learning_rate": 3.902991471128554e-05, + "loss": 0.0237, + "step": 59190 + }, + { + "epoch": 11.492985777389448, + "grad_norm": 0.08776077628135681, + "learning_rate": 3.901378348761456e-05, + "loss": 0.0266, + "step": 59200 + }, + { + "epoch": 11.494927430707246, + "grad_norm": 0.1584693342447281, + "learning_rate": 3.899765346537971e-05, + "loss": 0.0251, + "step": 59210 + }, + { + "epoch": 11.496869084025047, + "grad_norm": 0.2401718944311142, + "learning_rate": 3.898152464634493e-05, + "loss": 0.0231, + "step": 59220 + }, + { + "epoch": 11.498810737342847, + "grad_norm": 0.19256079196929932, + "learning_rate": 3.896539703227402e-05, + "loss": 0.0235, + "step": 59230 + }, + { + "epoch": 11.500752390660647, + "grad_norm": 0.2266107201576233, + "learning_rate": 3.894927062493073e-05, + "loss": 0.0249, + "step": 59240 + }, + { + "epoch": 11.502694043978448, + "grad_norm": 0.15869100391864777, + "learning_rate": 3.893314542607859e-05, + "loss": 0.0224, + "step": 59250 + }, + { + "epoch": 11.504635697296248, + "grad_norm": 0.205480694770813, + "learning_rate": 3.891702143748103e-05, + "loss": 0.0256, + "step": 59260 + }, + { + "epoch": 11.506577350614048, + "grad_norm": 0.15823780000209808, + "learning_rate": 3.890089866090133e-05, + "loss": 0.0225, + "step": 59270 + }, + { + "epoch": 11.508519003931848, + "grad_norm": 0.16071529686450958, + "learning_rate": 3.888477709810267e-05, + "loss": 0.0219, + "step": 59280 + }, + { + "epoch": 11.510460657249649, + "grad_norm": 0.12414044886827469, + "learning_rate": 3.886865675084806e-05, + "loss": 0.0224, + "step": 59290 + }, + { + "epoch": 11.512402310567449, + "grad_norm": 0.3465592861175537, + "learning_rate": 3.885253762090043e-05, + "loss": 0.0278, + "step": 59300 + }, + { + "epoch": 11.51434396388525, + "grad_norm": 0.13468237221240997, + "learning_rate": 3.8836419710022534e-05, + "loss": 0.0236, + "step": 59310 + }, + { + "epoch": 11.516285617203048, + "grad_norm": 0.1877397745847702, + "learning_rate": 3.882030301997698e-05, + "loss": 0.0301, + "step": 59320 + }, + { + "epoch": 11.518227270520848, + "grad_norm": 0.2775793671607971, + "learning_rate": 3.880418755252629e-05, + "loss": 0.0244, + "step": 59330 + }, + { + "epoch": 11.520168923838648, + "grad_norm": 0.1743113100528717, + "learning_rate": 3.87880733094328e-05, + "loss": 0.0233, + "step": 59340 + }, + { + "epoch": 11.522110577156448, + "grad_norm": 0.1221703365445137, + "learning_rate": 3.877196029245877e-05, + "loss": 0.0233, + "step": 59350 + }, + { + "epoch": 11.524052230474249, + "grad_norm": 0.192588210105896, + "learning_rate": 3.875584850336627e-05, + "loss": 0.0252, + "step": 59360 + }, + { + "epoch": 11.525993883792049, + "grad_norm": 0.1791846752166748, + "learning_rate": 3.873973794391728e-05, + "loss": 0.0259, + "step": 59370 + }, + { + "epoch": 11.52793553710985, + "grad_norm": 0.24287210404872894, + "learning_rate": 3.872362861587361e-05, + "loss": 0.0266, + "step": 59380 + }, + { + "epoch": 11.52987719042765, + "grad_norm": 0.22864042222499847, + "learning_rate": 3.870752052099698e-05, + "loss": 0.0209, + "step": 59390 + }, + { + "epoch": 11.53181884374545, + "grad_norm": 0.1740761548280716, + "learning_rate": 3.8691413661048916e-05, + "loss": 0.0191, + "step": 59400 + }, + { + "epoch": 11.53376049706325, + "grad_norm": 0.1385725885629654, + "learning_rate": 3.867530803779085e-05, + "loss": 0.0228, + "step": 59410 + }, + { + "epoch": 11.535702150381049, + "grad_norm": 0.15762339532375336, + "learning_rate": 3.8659203652984054e-05, + "loss": 0.0271, + "step": 59420 + }, + { + "epoch": 11.537643803698849, + "grad_norm": 0.20748458802700043, + "learning_rate": 3.864310050838972e-05, + "loss": 0.0235, + "step": 59430 + }, + { + "epoch": 11.53958545701665, + "grad_norm": 0.1728503406047821, + "learning_rate": 3.8626998605768835e-05, + "loss": 0.0253, + "step": 59440 + }, + { + "epoch": 11.54152711033445, + "grad_norm": 0.16173046827316284, + "learning_rate": 3.8610897946882294e-05, + "loss": 0.0301, + "step": 59450 + }, + { + "epoch": 11.54346876365225, + "grad_norm": 0.24445897340774536, + "learning_rate": 3.859479853349082e-05, + "loss": 0.026, + "step": 59460 + }, + { + "epoch": 11.54541041697005, + "grad_norm": 0.20445555448532104, + "learning_rate": 3.857870036735503e-05, + "loss": 0.0254, + "step": 59470 + }, + { + "epoch": 11.54735207028785, + "grad_norm": 0.13151642680168152, + "learning_rate": 3.856260345023538e-05, + "loss": 0.0227, + "step": 59480 + }, + { + "epoch": 11.54929372360565, + "grad_norm": 0.1596391648054123, + "learning_rate": 3.854650778389224e-05, + "loss": 0.0285, + "step": 59490 + }, + { + "epoch": 11.55123537692345, + "grad_norm": 0.2063235342502594, + "learning_rate": 3.8530413370085786e-05, + "loss": 0.0221, + "step": 59500 + }, + { + "epoch": 11.553177030241251, + "grad_norm": 0.12705211341381073, + "learning_rate": 3.851432021057608e-05, + "loss": 0.0214, + "step": 59510 + }, + { + "epoch": 11.555118683559051, + "grad_norm": 0.15698207914829254, + "learning_rate": 3.849822830712306e-05, + "loss": 0.0325, + "step": 59520 + }, + { + "epoch": 11.55706033687685, + "grad_norm": 0.21972569823265076, + "learning_rate": 3.848213766148649e-05, + "loss": 0.0233, + "step": 59530 + }, + { + "epoch": 11.55900199019465, + "grad_norm": 0.17520108819007874, + "learning_rate": 3.846604827542602e-05, + "loss": 0.0196, + "step": 59540 + }, + { + "epoch": 11.56094364351245, + "grad_norm": 0.11386563628911972, + "learning_rate": 3.844996015070117e-05, + "loss": 0.022, + "step": 59550 + }, + { + "epoch": 11.56288529683025, + "grad_norm": 0.142514169216156, + "learning_rate": 3.843387328907132e-05, + "loss": 0.0238, + "step": 59560 + }, + { + "epoch": 11.564826950148051, + "grad_norm": 0.19727136194705963, + "learning_rate": 3.8417787692295696e-05, + "loss": 0.0233, + "step": 59570 + }, + { + "epoch": 11.566768603465851, + "grad_norm": 0.12918728590011597, + "learning_rate": 3.84017033621334e-05, + "loss": 0.0264, + "step": 59580 + }, + { + "epoch": 11.568710256783652, + "grad_norm": 0.2424289882183075, + "learning_rate": 3.838562030034338e-05, + "loss": 0.0266, + "step": 59590 + }, + { + "epoch": 11.570651910101452, + "grad_norm": 0.07507125288248062, + "learning_rate": 3.836953850868447e-05, + "loss": 0.0272, + "step": 59600 + }, + { + "epoch": 11.572593563419252, + "grad_norm": 0.17176704108715057, + "learning_rate": 3.835345798891533e-05, + "loss": 0.0269, + "step": 59610 + }, + { + "epoch": 11.574535216737052, + "grad_norm": 0.11792410910129547, + "learning_rate": 3.833737874279452e-05, + "loss": 0.0284, + "step": 59620 + }, + { + "epoch": 11.57647687005485, + "grad_norm": 0.1435583084821701, + "learning_rate": 3.8321300772080433e-05, + "loss": 0.0261, + "step": 59630 + }, + { + "epoch": 11.578418523372651, + "grad_norm": 0.23023273050785065, + "learning_rate": 3.8305224078531345e-05, + "loss": 0.0258, + "step": 59640 + }, + { + "epoch": 11.580360176690451, + "grad_norm": 0.39080455899238586, + "learning_rate": 3.828914866390537e-05, + "loss": 0.0221, + "step": 59650 + }, + { + "epoch": 11.582301830008252, + "grad_norm": 0.1126798763871193, + "learning_rate": 3.827307452996048e-05, + "loss": 0.0196, + "step": 59660 + }, + { + "epoch": 11.584243483326052, + "grad_norm": 0.11725811660289764, + "learning_rate": 3.825700167845454e-05, + "loss": 0.0259, + "step": 59670 + }, + { + "epoch": 11.586185136643852, + "grad_norm": 0.1609870195388794, + "learning_rate": 3.8240930111145214e-05, + "loss": 0.0264, + "step": 59680 + }, + { + "epoch": 11.588126789961652, + "grad_norm": 0.1652299165725708, + "learning_rate": 3.8224859829790116e-05, + "loss": 0.0266, + "step": 59690 + }, + { + "epoch": 11.590068443279453, + "grad_norm": 0.21013975143432617, + "learning_rate": 3.820879083614664e-05, + "loss": 0.0275, + "step": 59700 + }, + { + "epoch": 11.592010096597253, + "grad_norm": 0.19678936898708344, + "learning_rate": 3.819272313197208e-05, + "loss": 0.0263, + "step": 59710 + }, + { + "epoch": 11.593951749915053, + "grad_norm": 0.15230996906757355, + "learning_rate": 3.817665671902355e-05, + "loss": 0.0224, + "step": 59720 + }, + { + "epoch": 11.595893403232854, + "grad_norm": 0.12560665607452393, + "learning_rate": 3.816059159905808e-05, + "loss": 0.0322, + "step": 59730 + }, + { + "epoch": 11.597835056550652, + "grad_norm": 0.24192549288272858, + "learning_rate": 3.8144527773832476e-05, + "loss": 0.0341, + "step": 59740 + }, + { + "epoch": 11.599776709868452, + "grad_norm": 0.1420890837907791, + "learning_rate": 3.812846524510352e-05, + "loss": 0.0295, + "step": 59750 + }, + { + "epoch": 11.601718363186253, + "grad_norm": 0.20944088697433472, + "learning_rate": 3.811240401462775e-05, + "loss": 0.024, + "step": 59760 + }, + { + "epoch": 11.603660016504053, + "grad_norm": 0.10599706321954727, + "learning_rate": 3.809634408416162e-05, + "loss": 0.02, + "step": 59770 + }, + { + "epoch": 11.605601669821853, + "grad_norm": 0.18510887026786804, + "learning_rate": 3.8080285455461394e-05, + "loss": 0.0222, + "step": 59780 + }, + { + "epoch": 11.607543323139653, + "grad_norm": 0.1098499745130539, + "learning_rate": 3.806422813028323e-05, + "loss": 0.0201, + "step": 59790 + }, + { + "epoch": 11.609484976457454, + "grad_norm": 0.18300099670886993, + "learning_rate": 3.8048172110383114e-05, + "loss": 0.0239, + "step": 59800 + }, + { + "epoch": 11.611426629775254, + "grad_norm": 0.102833591401577, + "learning_rate": 3.8032117397516944e-05, + "loss": 0.0267, + "step": 59810 + }, + { + "epoch": 11.613368283093054, + "grad_norm": 0.12799598276615143, + "learning_rate": 3.801606399344042e-05, + "loss": 0.0191, + "step": 59820 + }, + { + "epoch": 11.615309936410855, + "grad_norm": 0.12808942794799805, + "learning_rate": 3.800001189990913e-05, + "loss": 0.0296, + "step": 59830 + }, + { + "epoch": 11.617251589728653, + "grad_norm": 0.17568518221378326, + "learning_rate": 3.798396111867847e-05, + "loss": 0.0178, + "step": 59840 + }, + { + "epoch": 11.619193243046453, + "grad_norm": 0.12463025748729706, + "learning_rate": 3.796791165150378e-05, + "loss": 0.0199, + "step": 59850 + }, + { + "epoch": 11.621134896364254, + "grad_norm": 0.07171319425106049, + "learning_rate": 3.795186350014018e-05, + "loss": 0.0215, + "step": 59860 + }, + { + "epoch": 11.623076549682054, + "grad_norm": 0.24386896193027496, + "learning_rate": 3.793581666634266e-05, + "loss": 0.0273, + "step": 59870 + }, + { + "epoch": 11.625018202999854, + "grad_norm": 0.14181576669216156, + "learning_rate": 3.791977115186609e-05, + "loss": 0.0263, + "step": 59880 + }, + { + "epoch": 11.626959856317654, + "grad_norm": 0.13916990160942078, + "learning_rate": 3.79037269584652e-05, + "loss": 0.0228, + "step": 59890 + }, + { + "epoch": 11.628901509635455, + "grad_norm": 0.07671840488910675, + "learning_rate": 3.7887684087894545e-05, + "loss": 0.0219, + "step": 59900 + }, + { + "epoch": 11.630843162953255, + "grad_norm": 0.17676372826099396, + "learning_rate": 3.787164254190854e-05, + "loss": 0.0214, + "step": 59910 + }, + { + "epoch": 11.632784816271055, + "grad_norm": 0.15373839437961578, + "learning_rate": 3.7855602322261473e-05, + "loss": 0.0298, + "step": 59920 + }, + { + "epoch": 11.634726469588855, + "grad_norm": 0.16887597739696503, + "learning_rate": 3.783956343070746e-05, + "loss": 0.0241, + "step": 59930 + }, + { + "epoch": 11.636668122906656, + "grad_norm": 0.1506093144416809, + "learning_rate": 3.7823525869000526e-05, + "loss": 0.0276, + "step": 59940 + }, + { + "epoch": 11.638609776224456, + "grad_norm": 0.30291953682899475, + "learning_rate": 3.78074896388945e-05, + "loss": 0.0229, + "step": 59950 + }, + { + "epoch": 11.640551429542255, + "grad_norm": 0.2262040078639984, + "learning_rate": 3.7791454742143075e-05, + "loss": 0.0229, + "step": 59960 + }, + { + "epoch": 11.642493082860055, + "grad_norm": 0.15956030786037445, + "learning_rate": 3.7775421180499795e-05, + "loss": 0.0271, + "step": 59970 + }, + { + "epoch": 11.644434736177855, + "grad_norm": 0.18739432096481323, + "learning_rate": 3.7759388955718086e-05, + "loss": 0.0223, + "step": 59980 + }, + { + "epoch": 11.646376389495655, + "grad_norm": 0.18376298248767853, + "learning_rate": 3.774335806955117e-05, + "loss": 0.0231, + "step": 59990 + }, + { + "epoch": 11.648318042813456, + "grad_norm": 0.10308967530727386, + "learning_rate": 3.7727328523752215e-05, + "loss": 0.0231, + "step": 60000 + }, + { + "epoch": 11.650259696131256, + "grad_norm": 0.23213064670562744, + "learning_rate": 3.771130032007415e-05, + "loss": 0.0228, + "step": 60010 + }, + { + "epoch": 11.652201349449056, + "grad_norm": 0.32635676860809326, + "learning_rate": 3.769527346026982e-05, + "loss": 0.035, + "step": 60020 + }, + { + "epoch": 11.654143002766856, + "grad_norm": 0.11537773907184601, + "learning_rate": 3.767924794609188e-05, + "loss": 0.0188, + "step": 60030 + }, + { + "epoch": 11.656084656084657, + "grad_norm": 0.2919673025608063, + "learning_rate": 3.7663223779292854e-05, + "loss": 0.0276, + "step": 60040 + }, + { + "epoch": 11.658026309402457, + "grad_norm": 0.19685517251491547, + "learning_rate": 3.764720096162512e-05, + "loss": 0.0229, + "step": 60050 + }, + { + "epoch": 11.659967962720255, + "grad_norm": 0.19979450106620789, + "learning_rate": 3.763117949484092e-05, + "loss": 0.0319, + "step": 60060 + }, + { + "epoch": 11.661909616038056, + "grad_norm": 0.12711967527866364, + "learning_rate": 3.7615159380692344e-05, + "loss": 0.0254, + "step": 60070 + }, + { + "epoch": 11.663851269355856, + "grad_norm": 0.13973964750766754, + "learning_rate": 3.759914062093132e-05, + "loss": 0.0219, + "step": 60080 + }, + { + "epoch": 11.665792922673656, + "grad_norm": 0.1844879686832428, + "learning_rate": 3.7583123217309615e-05, + "loss": 0.0264, + "step": 60090 + }, + { + "epoch": 11.667734575991457, + "grad_norm": 0.23667386174201965, + "learning_rate": 3.7567107171578904e-05, + "loss": 0.0223, + "step": 60100 + }, + { + "epoch": 11.669676229309257, + "grad_norm": 0.1653008908033371, + "learning_rate": 3.755109248549066e-05, + "loss": 0.0218, + "step": 60110 + }, + { + "epoch": 11.671617882627057, + "grad_norm": 0.22096985578536987, + "learning_rate": 3.75350791607962e-05, + "loss": 0.0218, + "step": 60120 + }, + { + "epoch": 11.673559535944857, + "grad_norm": 0.1345764398574829, + "learning_rate": 3.751906719924676e-05, + "loss": 0.0215, + "step": 60130 + }, + { + "epoch": 11.675501189262658, + "grad_norm": 0.21937212347984314, + "learning_rate": 3.750305660259337e-05, + "loss": 0.0267, + "step": 60140 + }, + { + "epoch": 11.677442842580458, + "grad_norm": 0.09866105020046234, + "learning_rate": 3.748704737258693e-05, + "loss": 0.0228, + "step": 60150 + }, + { + "epoch": 11.679384495898258, + "grad_norm": 0.14970190823078156, + "learning_rate": 3.747103951097816e-05, + "loss": 0.0196, + "step": 60160 + }, + { + "epoch": 11.681326149216057, + "grad_norm": 0.18115834891796112, + "learning_rate": 3.745503301951768e-05, + "loss": 0.0225, + "step": 60170 + }, + { + "epoch": 11.683267802533857, + "grad_norm": 0.23655681312084198, + "learning_rate": 3.743902789995592e-05, + "loss": 0.0261, + "step": 60180 + }, + { + "epoch": 11.685209455851657, + "grad_norm": 0.09875829517841339, + "learning_rate": 3.7423024154043183e-05, + "loss": 0.0249, + "step": 60190 + }, + { + "epoch": 11.687151109169458, + "grad_norm": 0.1987859308719635, + "learning_rate": 3.740702178352963e-05, + "loss": 0.0248, + "step": 60200 + }, + { + "epoch": 11.689092762487258, + "grad_norm": 0.15332061052322388, + "learning_rate": 3.739102079016523e-05, + "loss": 0.0299, + "step": 60210 + }, + { + "epoch": 11.691034415805058, + "grad_norm": 0.14651046693325043, + "learning_rate": 3.7375021175699846e-05, + "loss": 0.0292, + "step": 60220 + }, + { + "epoch": 11.692976069122858, + "grad_norm": 0.12567223608493805, + "learning_rate": 3.735902294188317e-05, + "loss": 0.0242, + "step": 60230 + }, + { + "epoch": 11.694917722440659, + "grad_norm": 0.32948336005210876, + "learning_rate": 3.734302609046473e-05, + "loss": 0.0271, + "step": 60240 + }, + { + "epoch": 11.696859375758459, + "grad_norm": 0.10301379859447479, + "learning_rate": 3.732703062319391e-05, + "loss": 0.0184, + "step": 60250 + }, + { + "epoch": 11.69880102907626, + "grad_norm": 0.23302243649959564, + "learning_rate": 3.731103654181999e-05, + "loss": 0.0298, + "step": 60260 + }, + { + "epoch": 11.700742682394058, + "grad_norm": 0.3282927870750427, + "learning_rate": 3.729504384809203e-05, + "loss": 0.031, + "step": 60270 + }, + { + "epoch": 11.702684335711858, + "grad_norm": 0.2371399998664856, + "learning_rate": 3.727905254375898e-05, + "loss": 0.0254, + "step": 60280 + }, + { + "epoch": 11.704625989029658, + "grad_norm": 0.21692143380641937, + "learning_rate": 3.7263062630569614e-05, + "loss": 0.031, + "step": 60290 + }, + { + "epoch": 11.706567642347459, + "grad_norm": 0.18172065913677216, + "learning_rate": 3.724707411027256e-05, + "loss": 0.0257, + "step": 60300 + }, + { + "epoch": 11.708509295665259, + "grad_norm": 0.28121325373649597, + "learning_rate": 3.723108698461631e-05, + "loss": 0.034, + "step": 60310 + }, + { + "epoch": 11.710450948983059, + "grad_norm": 0.1770496666431427, + "learning_rate": 3.72151012553492e-05, + "loss": 0.0257, + "step": 60320 + }, + { + "epoch": 11.71239260230086, + "grad_norm": 0.2681506872177124, + "learning_rate": 3.719911692421939e-05, + "loss": 0.0318, + "step": 60330 + }, + { + "epoch": 11.71433425561866, + "grad_norm": 0.24116945266723633, + "learning_rate": 3.71831339929749e-05, + "loss": 0.0281, + "step": 60340 + }, + { + "epoch": 11.71627590893646, + "grad_norm": 0.12838606536388397, + "learning_rate": 3.716715246336362e-05, + "loss": 0.0258, + "step": 60350 + }, + { + "epoch": 11.71821756225426, + "grad_norm": 0.1838877946138382, + "learning_rate": 3.715117233713324e-05, + "loss": 0.0254, + "step": 60360 + }, + { + "epoch": 11.72015921557206, + "grad_norm": 0.19316403567790985, + "learning_rate": 3.713519361603135e-05, + "loss": 0.0303, + "step": 60370 + }, + { + "epoch": 11.722100868889859, + "grad_norm": 0.16213124990463257, + "learning_rate": 3.711921630180532e-05, + "loss": 0.0277, + "step": 60380 + }, + { + "epoch": 11.72404252220766, + "grad_norm": 0.0975290909409523, + "learning_rate": 3.710324039620245e-05, + "loss": 0.0268, + "step": 60390 + }, + { + "epoch": 11.72598417552546, + "grad_norm": 0.15117771923542023, + "learning_rate": 3.708726590096982e-05, + "loss": 0.028, + "step": 60400 + }, + { + "epoch": 11.72792582884326, + "grad_norm": 0.23458024859428406, + "learning_rate": 3.707129281785437e-05, + "loss": 0.0262, + "step": 60410 + }, + { + "epoch": 11.72986748216106, + "grad_norm": 0.09967557340860367, + "learning_rate": 3.705532114860291e-05, + "loss": 0.0239, + "step": 60420 + }, + { + "epoch": 11.73180913547886, + "grad_norm": 0.24275583028793335, + "learning_rate": 3.703935089496207e-05, + "loss": 0.0259, + "step": 60430 + }, + { + "epoch": 11.73375078879666, + "grad_norm": 0.12954461574554443, + "learning_rate": 3.702338205867829e-05, + "loss": 0.0203, + "step": 60440 + }, + { + "epoch": 11.73569244211446, + "grad_norm": 0.16139666736125946, + "learning_rate": 3.7007414641497976e-05, + "loss": 0.0323, + "step": 60450 + }, + { + "epoch": 11.737634095432261, + "grad_norm": 0.10349852591753006, + "learning_rate": 3.699144864516726e-05, + "loss": 0.0304, + "step": 60460 + }, + { + "epoch": 11.739575748750061, + "grad_norm": 0.11177331954240799, + "learning_rate": 3.6975484071432153e-05, + "loss": 0.0236, + "step": 60470 + }, + { + "epoch": 11.74151740206786, + "grad_norm": 0.19943739473819733, + "learning_rate": 3.695952092203854e-05, + "loss": 0.0188, + "step": 60480 + }, + { + "epoch": 11.74345905538566, + "grad_norm": 0.16868644952774048, + "learning_rate": 3.69435591987321e-05, + "loss": 0.0219, + "step": 60490 + }, + { + "epoch": 11.74540070870346, + "grad_norm": 0.1816667914390564, + "learning_rate": 3.6927598903258374e-05, + "loss": 0.0171, + "step": 60500 + }, + { + "epoch": 11.74734236202126, + "grad_norm": 0.16763493418693542, + "learning_rate": 3.691164003736281e-05, + "loss": 0.0241, + "step": 60510 + }, + { + "epoch": 11.749284015339061, + "grad_norm": 0.2971729636192322, + "learning_rate": 3.689568260279061e-05, + "loss": 0.0269, + "step": 60520 + }, + { + "epoch": 11.751225668656861, + "grad_norm": 0.2225055992603302, + "learning_rate": 3.687972660128686e-05, + "loss": 0.0232, + "step": 60530 + }, + { + "epoch": 11.753167321974662, + "grad_norm": 0.1368275135755539, + "learning_rate": 3.686377203459648e-05, + "loss": 0.0317, + "step": 60540 + }, + { + "epoch": 11.755108975292462, + "grad_norm": 0.1893022656440735, + "learning_rate": 3.6847818904464226e-05, + "loss": 0.0202, + "step": 60550 + }, + { + "epoch": 11.757050628610262, + "grad_norm": 0.13082551956176758, + "learning_rate": 3.683186721263474e-05, + "loss": 0.0261, + "step": 60560 + }, + { + "epoch": 11.758992281928062, + "grad_norm": 0.13518905639648438, + "learning_rate": 3.6815916960852435e-05, + "loss": 0.0278, + "step": 60570 + }, + { + "epoch": 11.760933935245863, + "grad_norm": 0.13628710806369781, + "learning_rate": 3.679996815086165e-05, + "loss": 0.0198, + "step": 60580 + }, + { + "epoch": 11.762875588563663, + "grad_norm": 0.29890376329421997, + "learning_rate": 3.678402078440649e-05, + "loss": 0.0221, + "step": 60590 + }, + { + "epoch": 11.764817241881461, + "grad_norm": 0.18297143280506134, + "learning_rate": 3.676807486323096e-05, + "loss": 0.0264, + "step": 60600 + }, + { + "epoch": 11.766758895199262, + "grad_norm": 0.1984526664018631, + "learning_rate": 3.6752130389078864e-05, + "loss": 0.026, + "step": 60610 + }, + { + "epoch": 11.768700548517062, + "grad_norm": 0.0890333279967308, + "learning_rate": 3.673618736369388e-05, + "loss": 0.0297, + "step": 60620 + }, + { + "epoch": 11.770642201834862, + "grad_norm": 0.19620470702648163, + "learning_rate": 3.6720245788819474e-05, + "loss": 0.0211, + "step": 60630 + }, + { + "epoch": 11.772583855152662, + "grad_norm": 0.1421767771244049, + "learning_rate": 3.6704305666199044e-05, + "loss": 0.0249, + "step": 60640 + }, + { + "epoch": 11.774525508470463, + "grad_norm": 0.13060010969638824, + "learning_rate": 3.668836699757576e-05, + "loss": 0.0207, + "step": 60650 + }, + { + "epoch": 11.776467161788263, + "grad_norm": 0.20984232425689697, + "learning_rate": 3.667242978469265e-05, + "loss": 0.0263, + "step": 60660 + }, + { + "epoch": 11.778408815106063, + "grad_norm": 0.20433732867240906, + "learning_rate": 3.6656494029292575e-05, + "loss": 0.0309, + "step": 60670 + }, + { + "epoch": 11.780350468423864, + "grad_norm": 0.18099920451641083, + "learning_rate": 3.664055973311825e-05, + "loss": 0.0255, + "step": 60680 + }, + { + "epoch": 11.782292121741664, + "grad_norm": 0.19483111798763275, + "learning_rate": 3.662462689791221e-05, + "loss": 0.0216, + "step": 60690 + }, + { + "epoch": 11.784233775059462, + "grad_norm": 0.1312481313943863, + "learning_rate": 3.660869552541689e-05, + "loss": 0.0243, + "step": 60700 + }, + { + "epoch": 11.786175428377263, + "grad_norm": 0.1632033735513687, + "learning_rate": 3.659276561737448e-05, + "loss": 0.0239, + "step": 60710 + }, + { + "epoch": 11.788117081695063, + "grad_norm": 0.23926040530204773, + "learning_rate": 3.657683717552708e-05, + "loss": 0.0224, + "step": 60720 + }, + { + "epoch": 11.790058735012863, + "grad_norm": 0.2492612898349762, + "learning_rate": 3.6560910201616574e-05, + "loss": 0.0295, + "step": 60730 + }, + { + "epoch": 11.792000388330663, + "grad_norm": 0.12514328956604004, + "learning_rate": 3.654498469738472e-05, + "loss": 0.0259, + "step": 60740 + }, + { + "epoch": 11.793942041648464, + "grad_norm": 0.10681623220443726, + "learning_rate": 3.6529060664573126e-05, + "loss": 0.0255, + "step": 60750 + }, + { + "epoch": 11.795883694966264, + "grad_norm": 0.10627970844507217, + "learning_rate": 3.6513138104923176e-05, + "loss": 0.0241, + "step": 60760 + }, + { + "epoch": 11.797825348284064, + "grad_norm": 0.2362518310546875, + "learning_rate": 3.6497217020176176e-05, + "loss": 0.0287, + "step": 60770 + }, + { + "epoch": 11.799767001601865, + "grad_norm": 0.19271697103977203, + "learning_rate": 3.648129741207323e-05, + "loss": 0.0234, + "step": 60780 + }, + { + "epoch": 11.801708654919665, + "grad_norm": 0.2530425488948822, + "learning_rate": 3.646537928235527e-05, + "loss": 0.03, + "step": 60790 + }, + { + "epoch": 11.803650308237465, + "grad_norm": 0.18411372601985931, + "learning_rate": 3.6449462632763075e-05, + "loss": 0.0256, + "step": 60800 + }, + { + "epoch": 11.805591961555264, + "grad_norm": 0.11562281847000122, + "learning_rate": 3.643354746503729e-05, + "loss": 0.0255, + "step": 60810 + }, + { + "epoch": 11.807533614873064, + "grad_norm": 0.27546370029449463, + "learning_rate": 3.6417633780918345e-05, + "loss": 0.0249, + "step": 60820 + }, + { + "epoch": 11.809475268190864, + "grad_norm": 0.12289613485336304, + "learning_rate": 3.640172158214655e-05, + "loss": 0.0269, + "step": 60830 + }, + { + "epoch": 11.811416921508664, + "grad_norm": 0.14373566210269928, + "learning_rate": 3.6385810870462035e-05, + "loss": 0.0255, + "step": 60840 + }, + { + "epoch": 11.813358574826465, + "grad_norm": 0.13249340653419495, + "learning_rate": 3.636990164760479e-05, + "loss": 0.024, + "step": 60850 + }, + { + "epoch": 11.815300228144265, + "grad_norm": 0.22727622091770172, + "learning_rate": 3.6353993915314596e-05, + "loss": 0.0276, + "step": 60860 + }, + { + "epoch": 11.817241881462065, + "grad_norm": 0.17393074929714203, + "learning_rate": 3.6338087675331124e-05, + "loss": 0.0236, + "step": 60870 + }, + { + "epoch": 11.819183534779865, + "grad_norm": 0.13861414790153503, + "learning_rate": 3.632218292939383e-05, + "loss": 0.0273, + "step": 60880 + }, + { + "epoch": 11.821125188097666, + "grad_norm": 0.17442680895328522, + "learning_rate": 3.6306279679242033e-05, + "loss": 0.0221, + "step": 60890 + }, + { + "epoch": 11.823066841415466, + "grad_norm": 0.137144073843956, + "learning_rate": 3.629037792661492e-05, + "loss": 0.0193, + "step": 60900 + }, + { + "epoch": 11.825008494733265, + "grad_norm": 0.11211548745632172, + "learning_rate": 3.627447767325147e-05, + "loss": 0.0206, + "step": 60910 + }, + { + "epoch": 11.826950148051065, + "grad_norm": 0.12433627247810364, + "learning_rate": 3.62585789208905e-05, + "loss": 0.0219, + "step": 60920 + }, + { + "epoch": 11.828891801368865, + "grad_norm": 0.12319124490022659, + "learning_rate": 3.624268167127068e-05, + "loss": 0.0304, + "step": 60930 + }, + { + "epoch": 11.830833454686665, + "grad_norm": 0.15467944741249084, + "learning_rate": 3.622678592613051e-05, + "loss": 0.0267, + "step": 60940 + }, + { + "epoch": 11.832775108004466, + "grad_norm": 0.2335837334394455, + "learning_rate": 3.6210891687208304e-05, + "loss": 0.0249, + "step": 60950 + }, + { + "epoch": 11.834716761322266, + "grad_norm": 0.15577368438243866, + "learning_rate": 3.619499895624228e-05, + "loss": 0.0193, + "step": 60960 + }, + { + "epoch": 11.836658414640066, + "grad_norm": 0.3429565727710724, + "learning_rate": 3.61791077349704e-05, + "loss": 0.0303, + "step": 60970 + }, + { + "epoch": 11.838600067957866, + "grad_norm": 0.1538398563861847, + "learning_rate": 3.616321802513053e-05, + "loss": 0.0212, + "step": 60980 + }, + { + "epoch": 11.840541721275667, + "grad_norm": 0.11023618280887604, + "learning_rate": 3.6147329828460325e-05, + "loss": 0.0249, + "step": 60990 + }, + { + "epoch": 11.842483374593467, + "grad_norm": 0.14354372024536133, + "learning_rate": 3.6131443146697306e-05, + "loss": 0.0216, + "step": 61000 + }, + { + "epoch": 11.844425027911267, + "grad_norm": 0.21372847259044647, + "learning_rate": 3.611555798157879e-05, + "loss": 0.0225, + "step": 61010 + }, + { + "epoch": 11.846366681229066, + "grad_norm": 0.2186114490032196, + "learning_rate": 3.6099674334842e-05, + "loss": 0.0232, + "step": 61020 + }, + { + "epoch": 11.848308334546866, + "grad_norm": 0.13967275619506836, + "learning_rate": 3.6083792208223934e-05, + "loss": 0.0231, + "step": 61030 + }, + { + "epoch": 11.850249987864666, + "grad_norm": 0.14456415176391602, + "learning_rate": 3.606791160346142e-05, + "loss": 0.0274, + "step": 61040 + }, + { + "epoch": 11.852191641182467, + "grad_norm": 0.12518343329429626, + "learning_rate": 3.6052032522291134e-05, + "loss": 0.0204, + "step": 61050 + }, + { + "epoch": 11.854133294500267, + "grad_norm": 0.22171808779239655, + "learning_rate": 3.603615496644962e-05, + "loss": 0.0281, + "step": 61060 + }, + { + "epoch": 11.856074947818067, + "grad_norm": 0.08795158565044403, + "learning_rate": 3.60202789376732e-05, + "loss": 0.0231, + "step": 61070 + }, + { + "epoch": 11.858016601135867, + "grad_norm": 0.14259269833564758, + "learning_rate": 3.6004404437698036e-05, + "loss": 0.0207, + "step": 61080 + }, + { + "epoch": 11.859958254453668, + "grad_norm": 0.1748451441526413, + "learning_rate": 3.5988531468260184e-05, + "loss": 0.0257, + "step": 61090 + }, + { + "epoch": 11.861899907771468, + "grad_norm": 0.20885410904884338, + "learning_rate": 3.597266003109546e-05, + "loss": 0.0222, + "step": 61100 + }, + { + "epoch": 11.863841561089268, + "grad_norm": 0.2559202313423157, + "learning_rate": 3.5956790127939557e-05, + "loss": 0.0331, + "step": 61110 + }, + { + "epoch": 11.865783214407067, + "grad_norm": 0.10782154649496078, + "learning_rate": 3.594092176052796e-05, + "loss": 0.024, + "step": 61120 + }, + { + "epoch": 11.867724867724867, + "grad_norm": 0.1239183247089386, + "learning_rate": 3.592505493059603e-05, + "loss": 0.0279, + "step": 61130 + }, + { + "epoch": 11.869666521042667, + "grad_norm": 0.09265150129795074, + "learning_rate": 3.5909189639878905e-05, + "loss": 0.0178, + "step": 61140 + }, + { + "epoch": 11.871608174360468, + "grad_norm": 0.1882154494524002, + "learning_rate": 3.5893325890111626e-05, + "loss": 0.0316, + "step": 61150 + }, + { + "epoch": 11.873549827678268, + "grad_norm": 0.14185841381549835, + "learning_rate": 3.5877463683029034e-05, + "loss": 0.0265, + "step": 61160 + }, + { + "epoch": 11.875491480996068, + "grad_norm": 0.23836325109004974, + "learning_rate": 3.586160302036578e-05, + "loss": 0.0192, + "step": 61170 + }, + { + "epoch": 11.877433134313868, + "grad_norm": 0.1806061565876007, + "learning_rate": 3.584574390385636e-05, + "loss": 0.031, + "step": 61180 + }, + { + "epoch": 11.879374787631669, + "grad_norm": 0.1143728643655777, + "learning_rate": 3.582988633523511e-05, + "loss": 0.0171, + "step": 61190 + }, + { + "epoch": 11.881316440949469, + "grad_norm": 0.18952687084674835, + "learning_rate": 3.581403031623616e-05, + "loss": 0.0239, + "step": 61200 + }, + { + "epoch": 11.88325809426727, + "grad_norm": 0.10226095467805862, + "learning_rate": 3.579817584859354e-05, + "loss": 0.0383, + "step": 61210 + }, + { + "epoch": 11.88519974758507, + "grad_norm": 0.224937304854393, + "learning_rate": 3.578232293404107e-05, + "loss": 0.0248, + "step": 61220 + }, + { + "epoch": 11.887141400902868, + "grad_norm": 0.17282401025295258, + "learning_rate": 3.5766471574312385e-05, + "loss": 0.0216, + "step": 61230 + }, + { + "epoch": 11.889083054220668, + "grad_norm": 0.08252257108688354, + "learning_rate": 3.5750621771140964e-05, + "loss": 0.0217, + "step": 61240 + }, + { + "epoch": 11.891024707538469, + "grad_norm": 0.17573241889476776, + "learning_rate": 3.5734773526260115e-05, + "loss": 0.0208, + "step": 61250 + }, + { + "epoch": 11.892966360856269, + "grad_norm": 0.09636412560939789, + "learning_rate": 3.5718926841402996e-05, + "loss": 0.0184, + "step": 61260 + }, + { + "epoch": 11.894908014174069, + "grad_norm": 0.14154687523841858, + "learning_rate": 3.5703081718302555e-05, + "loss": 0.0254, + "step": 61270 + }, + { + "epoch": 11.89684966749187, + "grad_norm": 0.2163955271244049, + "learning_rate": 3.568723815869162e-05, + "loss": 0.0179, + "step": 61280 + }, + { + "epoch": 11.89879132080967, + "grad_norm": 0.12726548314094543, + "learning_rate": 3.5671396164302777e-05, + "loss": 0.021, + "step": 61290 + }, + { + "epoch": 11.90073297412747, + "grad_norm": 0.15093374252319336, + "learning_rate": 3.565555573686853e-05, + "loss": 0.0194, + "step": 61300 + }, + { + "epoch": 11.90267462744527, + "grad_norm": 0.3517008125782013, + "learning_rate": 3.563971687812113e-05, + "loss": 0.0221, + "step": 61310 + }, + { + "epoch": 11.90461628076307, + "grad_norm": 0.22838962078094482, + "learning_rate": 3.5623879589792706e-05, + "loss": 0.0258, + "step": 61320 + }, + { + "epoch": 11.906557934080869, + "grad_norm": 0.09258966892957687, + "learning_rate": 3.5608043873615185e-05, + "loss": 0.0264, + "step": 61330 + }, + { + "epoch": 11.90849958739867, + "grad_norm": 0.1686827689409256, + "learning_rate": 3.559220973132035e-05, + "loss": 0.0271, + "step": 61340 + }, + { + "epoch": 11.91044124071647, + "grad_norm": 0.26184993982315063, + "learning_rate": 3.557637716463981e-05, + "loss": 0.0258, + "step": 61350 + }, + { + "epoch": 11.91238289403427, + "grad_norm": 0.21371005475521088, + "learning_rate": 3.556054617530497e-05, + "loss": 0.0246, + "step": 61360 + }, + { + "epoch": 11.91432454735207, + "grad_norm": 0.1376568228006363, + "learning_rate": 3.5544716765047106e-05, + "loss": 0.0257, + "step": 61370 + }, + { + "epoch": 11.91626620066987, + "grad_norm": 0.1848272681236267, + "learning_rate": 3.5528888935597274e-05, + "loss": 0.0209, + "step": 61380 + }, + { + "epoch": 11.91820785398767, + "grad_norm": 0.17138336598873138, + "learning_rate": 3.551306268868636e-05, + "loss": 0.027, + "step": 61390 + }, + { + "epoch": 11.92014950730547, + "grad_norm": 0.15934331715106964, + "learning_rate": 3.5497238026045166e-05, + "loss": 0.0255, + "step": 61400 + }, + { + "epoch": 11.922091160623271, + "grad_norm": 0.12437772005796432, + "learning_rate": 3.548141494940421e-05, + "loss": 0.0237, + "step": 61410 + }, + { + "epoch": 11.924032813941071, + "grad_norm": 0.17449668049812317, + "learning_rate": 3.546559346049388e-05, + "loss": 0.0253, + "step": 61420 + }, + { + "epoch": 11.925974467258872, + "grad_norm": 0.13601136207580566, + "learning_rate": 3.544977356104441e-05, + "loss": 0.0251, + "step": 61430 + }, + { + "epoch": 11.927916120576672, + "grad_norm": 0.1635892689228058, + "learning_rate": 3.543395525278582e-05, + "loss": 0.0261, + "step": 61440 + }, + { + "epoch": 11.92985777389447, + "grad_norm": 0.12074629962444305, + "learning_rate": 3.5418138537447985e-05, + "loss": 0.0249, + "step": 61450 + }, + { + "epoch": 11.93179942721227, + "grad_norm": 0.2017955332994461, + "learning_rate": 3.5402323416760574e-05, + "loss": 0.0312, + "step": 61460 + }, + { + "epoch": 11.933741080530071, + "grad_norm": 0.1709159016609192, + "learning_rate": 3.538650989245316e-05, + "loss": 0.0261, + "step": 61470 + }, + { + "epoch": 11.935682733847871, + "grad_norm": 0.14459329843521118, + "learning_rate": 3.537069796625504e-05, + "loss": 0.0252, + "step": 61480 + }, + { + "epoch": 11.937624387165672, + "grad_norm": 0.18511876463890076, + "learning_rate": 3.53548876398954e-05, + "loss": 0.0244, + "step": 61490 + }, + { + "epoch": 11.939566040483472, + "grad_norm": 0.2779194712638855, + "learning_rate": 3.5339078915103214e-05, + "loss": 0.023, + "step": 61500 + }, + { + "epoch": 11.941507693801272, + "grad_norm": 0.2847335934638977, + "learning_rate": 3.532327179360733e-05, + "loss": 0.0244, + "step": 61510 + }, + { + "epoch": 11.943449347119072, + "grad_norm": 0.19938218593597412, + "learning_rate": 3.530746627713636e-05, + "loss": 0.0235, + "step": 61520 + }, + { + "epoch": 11.945391000436873, + "grad_norm": 0.2726835310459137, + "learning_rate": 3.5291662367418805e-05, + "loss": 0.0266, + "step": 61530 + }, + { + "epoch": 11.947332653754673, + "grad_norm": 0.13599471747875214, + "learning_rate": 3.527586006618293e-05, + "loss": 0.0206, + "step": 61540 + }, + { + "epoch": 11.949274307072471, + "grad_norm": 0.17915719747543335, + "learning_rate": 3.526005937515687e-05, + "loss": 0.0342, + "step": 61550 + }, + { + "epoch": 11.951215960390272, + "grad_norm": 0.16739386320114136, + "learning_rate": 3.524426029606856e-05, + "loss": 0.0249, + "step": 61560 + }, + { + "epoch": 11.953157613708072, + "grad_norm": 0.22015570104122162, + "learning_rate": 3.5228462830645746e-05, + "loss": 0.0261, + "step": 61570 + }, + { + "epoch": 11.955099267025872, + "grad_norm": 0.2439931035041809, + "learning_rate": 3.5212666980616036e-05, + "loss": 0.0206, + "step": 61580 + }, + { + "epoch": 11.957040920343672, + "grad_norm": 0.26347705721855164, + "learning_rate": 3.519687274770681e-05, + "loss": 0.0346, + "step": 61590 + }, + { + "epoch": 11.958982573661473, + "grad_norm": 0.1776318997144699, + "learning_rate": 3.518108013364535e-05, + "loss": 0.0201, + "step": 61600 + }, + { + "epoch": 11.960924226979273, + "grad_norm": 0.1060875877737999, + "learning_rate": 3.516528914015868e-05, + "loss": 0.0224, + "step": 61610 + }, + { + "epoch": 11.962865880297073, + "grad_norm": 0.16272825002670288, + "learning_rate": 3.5149499768973694e-05, + "loss": 0.0242, + "step": 61620 + }, + { + "epoch": 11.964807533614874, + "grad_norm": 0.09995388984680176, + "learning_rate": 3.513371202181708e-05, + "loss": 0.019, + "step": 61630 + }, + { + "epoch": 11.966749186932674, + "grad_norm": 0.1479283571243286, + "learning_rate": 3.511792590041537e-05, + "loss": 0.0203, + "step": 61640 + }, + { + "epoch": 11.968690840250474, + "grad_norm": 0.14355693757534027, + "learning_rate": 3.510214140649488e-05, + "loss": 0.0197, + "step": 61650 + }, + { + "epoch": 11.970632493568273, + "grad_norm": 0.18941575288772583, + "learning_rate": 3.508635854178184e-05, + "loss": 0.0276, + "step": 61660 + }, + { + "epoch": 11.972574146886073, + "grad_norm": 0.2518247663974762, + "learning_rate": 3.50705773080022e-05, + "loss": 0.0269, + "step": 61670 + }, + { + "epoch": 11.974515800203873, + "grad_norm": 0.10700744390487671, + "learning_rate": 3.505479770688178e-05, + "loss": 0.0297, + "step": 61680 + }, + { + "epoch": 11.976457453521673, + "grad_norm": 0.21830740571022034, + "learning_rate": 3.503901974014622e-05, + "loss": 0.0239, + "step": 61690 + }, + { + "epoch": 11.978399106839474, + "grad_norm": 0.14902223646640778, + "learning_rate": 3.502324340952095e-05, + "loss": 0.0222, + "step": 61700 + }, + { + "epoch": 11.980340760157274, + "grad_norm": 0.21944040060043335, + "learning_rate": 3.500746871673126e-05, + "loss": 0.0223, + "step": 61710 + }, + { + "epoch": 11.982282413475074, + "grad_norm": 0.20703837275505066, + "learning_rate": 3.499169566350227e-05, + "loss": 0.0247, + "step": 61720 + }, + { + "epoch": 11.984224066792875, + "grad_norm": 0.1807044893503189, + "learning_rate": 3.4975924251558876e-05, + "loss": 0.0219, + "step": 61730 + }, + { + "epoch": 11.986165720110675, + "grad_norm": 0.252412348985672, + "learning_rate": 3.496015448262582e-05, + "loss": 0.0223, + "step": 61740 + }, + { + "epoch": 11.988107373428475, + "grad_norm": 0.13905887305736542, + "learning_rate": 3.494438635842764e-05, + "loss": 0.0279, + "step": 61750 + }, + { + "epoch": 11.990049026746274, + "grad_norm": 0.17785663902759552, + "learning_rate": 3.492861988068874e-05, + "loss": 0.0225, + "step": 61760 + }, + { + "epoch": 11.991990680064074, + "grad_norm": 0.1930645853281021, + "learning_rate": 3.491285505113332e-05, + "loss": 0.0276, + "step": 61770 + }, + { + "epoch": 11.993932333381874, + "grad_norm": 0.1468254029750824, + "learning_rate": 3.4897091871485376e-05, + "loss": 0.0217, + "step": 61780 + }, + { + "epoch": 11.995873986699674, + "grad_norm": 0.19059179723262787, + "learning_rate": 3.488133034346875e-05, + "loss": 0.0206, + "step": 61790 + }, + { + "epoch": 11.997815640017475, + "grad_norm": 0.10740312933921814, + "learning_rate": 3.486557046880713e-05, + "loss": 0.0231, + "step": 61800 + }, + { + "epoch": 11.999757293335275, + "grad_norm": 0.1384621411561966, + "learning_rate": 3.484981224922395e-05, + "loss": 0.0233, + "step": 61810 + }, + { + "epoch": 12.00155332265424, + "grad_norm": 0.13552357256412506, + "learning_rate": 3.4834055686442544e-05, + "loss": 0.027, + "step": 61820 + }, + { + "epoch": 12.00349497597204, + "grad_norm": 0.1731899529695511, + "learning_rate": 3.4818300782186e-05, + "loss": 0.0254, + "step": 61830 + }, + { + "epoch": 12.005436629289841, + "grad_norm": 0.1561589390039444, + "learning_rate": 3.4802547538177236e-05, + "loss": 0.0207, + "step": 61840 + }, + { + "epoch": 12.00737828260764, + "grad_norm": 0.19992421567440033, + "learning_rate": 3.478679595613905e-05, + "loss": 0.0211, + "step": 61850 + }, + { + "epoch": 12.00931993592544, + "grad_norm": 0.1667265146970749, + "learning_rate": 3.4771046037793975e-05, + "loss": 0.021, + "step": 61860 + }, + { + "epoch": 12.01126158924324, + "grad_norm": 0.23795431852340698, + "learning_rate": 3.4755297784864416e-05, + "loss": 0.0248, + "step": 61870 + }, + { + "epoch": 12.01320324256104, + "grad_norm": 0.1340435892343521, + "learning_rate": 3.473955119907257e-05, + "loss": 0.0188, + "step": 61880 + }, + { + "epoch": 12.01514489587884, + "grad_norm": 0.09260783344507217, + "learning_rate": 3.472380628214046e-05, + "loss": 0.026, + "step": 61890 + }, + { + "epoch": 12.017086549196641, + "grad_norm": 0.2734070122241974, + "learning_rate": 3.4708063035789906e-05, + "loss": 0.027, + "step": 61900 + }, + { + "epoch": 12.019028202514441, + "grad_norm": 0.19667945802211761, + "learning_rate": 3.4692321461742614e-05, + "loss": 0.0244, + "step": 61910 + }, + { + "epoch": 12.020969855832242, + "grad_norm": 0.17815245687961578, + "learning_rate": 3.4676581561720023e-05, + "loss": 0.0305, + "step": 61920 + }, + { + "epoch": 12.022911509150042, + "grad_norm": 0.18895690143108368, + "learning_rate": 3.466084333744344e-05, + "loss": 0.022, + "step": 61930 + }, + { + "epoch": 12.024853162467842, + "grad_norm": 0.1599978506565094, + "learning_rate": 3.4645106790633964e-05, + "loss": 0.0228, + "step": 61940 + }, + { + "epoch": 12.026794815785642, + "grad_norm": 0.14658990502357483, + "learning_rate": 3.4629371923012526e-05, + "loss": 0.02, + "step": 61950 + }, + { + "epoch": 12.02873646910344, + "grad_norm": 0.17698656022548676, + "learning_rate": 3.4613638736299856e-05, + "loss": 0.0305, + "step": 61960 + }, + { + "epoch": 12.030678122421241, + "grad_norm": 0.10875401645898819, + "learning_rate": 3.459790723221652e-05, + "loss": 0.0203, + "step": 61970 + }, + { + "epoch": 12.032619775739041, + "grad_norm": 0.22282220423221588, + "learning_rate": 3.45821774124829e-05, + "loss": 0.0266, + "step": 61980 + }, + { + "epoch": 12.034561429056842, + "grad_norm": 0.2063220739364624, + "learning_rate": 3.456644927881918e-05, + "loss": 0.0248, + "step": 61990 + }, + { + "epoch": 12.036503082374642, + "grad_norm": 0.2003994584083557, + "learning_rate": 3.4550722832945354e-05, + "loss": 0.0238, + "step": 62000 + }, + { + "epoch": 12.038444735692442, + "grad_norm": 0.07806701213121414, + "learning_rate": 3.453499807658126e-05, + "loss": 0.0242, + "step": 62010 + }, + { + "epoch": 12.040386389010242, + "grad_norm": 0.11291953921318054, + "learning_rate": 3.451927501144653e-05, + "loss": 0.0265, + "step": 62020 + }, + { + "epoch": 12.042328042328043, + "grad_norm": 0.16432075202465057, + "learning_rate": 3.450355363926059e-05, + "loss": 0.0208, + "step": 62030 + }, + { + "epoch": 12.044269695645843, + "grad_norm": 0.17409852147102356, + "learning_rate": 3.448783396174273e-05, + "loss": 0.0231, + "step": 62040 + }, + { + "epoch": 12.046211348963643, + "grad_norm": 0.23743659257888794, + "learning_rate": 3.4472115980612034e-05, + "loss": 0.0267, + "step": 62050 + }, + { + "epoch": 12.048153002281442, + "grad_norm": 0.18213611841201782, + "learning_rate": 3.4456399697587396e-05, + "loss": 0.0216, + "step": 62060 + }, + { + "epoch": 12.050094655599242, + "grad_norm": 0.1544632762670517, + "learning_rate": 3.4440685114387506e-05, + "loss": 0.0294, + "step": 62070 + }, + { + "epoch": 12.052036308917042, + "grad_norm": 0.14022448658943176, + "learning_rate": 3.442497223273091e-05, + "loss": 0.0239, + "step": 62080 + }, + { + "epoch": 12.053977962234843, + "grad_norm": 0.1749206930398941, + "learning_rate": 3.4409261054335926e-05, + "loss": 0.0273, + "step": 62090 + }, + { + "epoch": 12.055919615552643, + "grad_norm": 0.15574081242084503, + "learning_rate": 3.439355158092069e-05, + "loss": 0.0173, + "step": 62100 + }, + { + "epoch": 12.057861268870443, + "grad_norm": 0.14570853114128113, + "learning_rate": 3.437784381420322e-05, + "loss": 0.0212, + "step": 62110 + }, + { + "epoch": 12.059802922188243, + "grad_norm": 0.1870241016149521, + "learning_rate": 3.436213775590126e-05, + "loss": 0.0181, + "step": 62120 + }, + { + "epoch": 12.061744575506044, + "grad_norm": 0.16875147819519043, + "learning_rate": 3.4346433407732405e-05, + "loss": 0.0219, + "step": 62130 + }, + { + "epoch": 12.063686228823844, + "grad_norm": 0.22031359374523163, + "learning_rate": 3.433073077141405e-05, + "loss": 0.0291, + "step": 62140 + }, + { + "epoch": 12.065627882141644, + "grad_norm": 0.2754877209663391, + "learning_rate": 3.431502984866343e-05, + "loss": 0.0278, + "step": 62150 + }, + { + "epoch": 12.067569535459445, + "grad_norm": 0.18736545741558075, + "learning_rate": 3.429933064119753e-05, + "loss": 0.0292, + "step": 62160 + }, + { + "epoch": 12.069511188777243, + "grad_norm": 0.16384804248809814, + "learning_rate": 3.4283633150733255e-05, + "loss": 0.0225, + "step": 62170 + }, + { + "epoch": 12.071452842095043, + "grad_norm": 0.23115883767604828, + "learning_rate": 3.4267937378987225e-05, + "loss": 0.0285, + "step": 62180 + }, + { + "epoch": 12.073394495412844, + "grad_norm": 0.2388353943824768, + "learning_rate": 3.425224332767592e-05, + "loss": 0.0287, + "step": 62190 + }, + { + "epoch": 12.075336148730644, + "grad_norm": 0.07471220940351486, + "learning_rate": 3.4236550998515595e-05, + "loss": 0.0226, + "step": 62200 + }, + { + "epoch": 12.077277802048444, + "grad_norm": 0.2104075700044632, + "learning_rate": 3.422086039322235e-05, + "loss": 0.0228, + "step": 62210 + }, + { + "epoch": 12.079219455366244, + "grad_norm": 0.1945919394493103, + "learning_rate": 3.420517151351209e-05, + "loss": 0.0295, + "step": 62220 + }, + { + "epoch": 12.081161108684045, + "grad_norm": 0.20082776248455048, + "learning_rate": 3.418948436110053e-05, + "loss": 0.0267, + "step": 62230 + }, + { + "epoch": 12.083102762001845, + "grad_norm": 0.15512681007385254, + "learning_rate": 3.417379893770319e-05, + "loss": 0.0265, + "step": 62240 + }, + { + "epoch": 12.085044415319645, + "grad_norm": 0.1422979235649109, + "learning_rate": 3.41581152450354e-05, + "loss": 0.0221, + "step": 62250 + }, + { + "epoch": 12.086986068637446, + "grad_norm": 0.1972709447145462, + "learning_rate": 3.414243328481231e-05, + "loss": 0.0252, + "step": 62260 + }, + { + "epoch": 12.088927721955244, + "grad_norm": 0.1434035450220108, + "learning_rate": 3.412675305874888e-05, + "loss": 0.0222, + "step": 62270 + }, + { + "epoch": 12.090869375273044, + "grad_norm": 0.1933736652135849, + "learning_rate": 3.4111074568559876e-05, + "loss": 0.0238, + "step": 62280 + }, + { + "epoch": 12.092811028590845, + "grad_norm": 0.15087853372097015, + "learning_rate": 3.4095397815959846e-05, + "loss": 0.0241, + "step": 62290 + }, + { + "epoch": 12.094752681908645, + "grad_norm": 0.17164841294288635, + "learning_rate": 3.407972280266322e-05, + "loss": 0.0299, + "step": 62300 + }, + { + "epoch": 12.096694335226445, + "grad_norm": 0.13588501513004303, + "learning_rate": 3.4064049530384166e-05, + "loss": 0.0292, + "step": 62310 + }, + { + "epoch": 12.098635988544245, + "grad_norm": 0.11072229593992233, + "learning_rate": 3.40483780008367e-05, + "loss": 0.0276, + "step": 62320 + }, + { + "epoch": 12.100577641862046, + "grad_norm": 0.1456560492515564, + "learning_rate": 3.4032708215734635e-05, + "loss": 0.0241, + "step": 62330 + }, + { + "epoch": 12.102519295179846, + "grad_norm": 0.11351900547742844, + "learning_rate": 3.4017040176791584e-05, + "loss": 0.0174, + "step": 62340 + }, + { + "epoch": 12.104460948497646, + "grad_norm": 0.1974814385175705, + "learning_rate": 3.4001373885720976e-05, + "loss": 0.024, + "step": 62350 + }, + { + "epoch": 12.106402601815446, + "grad_norm": 0.22596219182014465, + "learning_rate": 3.398570934423608e-05, + "loss": 0.0212, + "step": 62360 + }, + { + "epoch": 12.108344255133247, + "grad_norm": 0.16684353351593018, + "learning_rate": 3.397004655404994e-05, + "loss": 0.0203, + "step": 62370 + }, + { + "epoch": 12.110285908451045, + "grad_norm": 0.128892183303833, + "learning_rate": 3.395438551687541e-05, + "loss": 0.0198, + "step": 62380 + }, + { + "epoch": 12.112227561768846, + "grad_norm": 0.20940542221069336, + "learning_rate": 3.393872623442515e-05, + "loss": 0.0267, + "step": 62390 + }, + { + "epoch": 12.114169215086646, + "grad_norm": 0.15240828692913055, + "learning_rate": 3.392306870841164e-05, + "loss": 0.0258, + "step": 62400 + }, + { + "epoch": 12.116110868404446, + "grad_norm": 0.2818796932697296, + "learning_rate": 3.390741294054715e-05, + "loss": 0.023, + "step": 62410 + }, + { + "epoch": 12.118052521722246, + "grad_norm": 0.11622022837400436, + "learning_rate": 3.38917589325438e-05, + "loss": 0.0238, + "step": 62420 + }, + { + "epoch": 12.119994175040047, + "grad_norm": 0.24621643126010895, + "learning_rate": 3.387610668611349e-05, + "loss": 0.018, + "step": 62430 + }, + { + "epoch": 12.121935828357847, + "grad_norm": 0.27070775628089905, + "learning_rate": 3.38604562029679e-05, + "loss": 0.0265, + "step": 62440 + }, + { + "epoch": 12.123877481675647, + "grad_norm": 0.1501847803592682, + "learning_rate": 3.384480748481855e-05, + "loss": 0.0292, + "step": 62450 + }, + { + "epoch": 12.125819134993447, + "grad_norm": 0.1583959311246872, + "learning_rate": 3.382916053337676e-05, + "loss": 0.0197, + "step": 62460 + }, + { + "epoch": 12.127760788311248, + "grad_norm": 0.14255428314208984, + "learning_rate": 3.381351535035366e-05, + "loss": 0.0204, + "step": 62470 + }, + { + "epoch": 12.129702441629048, + "grad_norm": 0.10540088266134262, + "learning_rate": 3.379787193746019e-05, + "loss": 0.0293, + "step": 62480 + }, + { + "epoch": 12.131644094946846, + "grad_norm": 0.1370476931333542, + "learning_rate": 3.378223029640709e-05, + "loss": 0.024, + "step": 62490 + }, + { + "epoch": 12.133585748264647, + "grad_norm": 0.17088904976844788, + "learning_rate": 3.376659042890489e-05, + "loss": 0.022, + "step": 62500 + }, + { + "epoch": 12.135527401582447, + "grad_norm": 0.0955924466252327, + "learning_rate": 3.375095233666397e-05, + "loss": 0.0215, + "step": 62510 + }, + { + "epoch": 12.137469054900247, + "grad_norm": 0.17029094696044922, + "learning_rate": 3.373531602139447e-05, + "loss": 0.0218, + "step": 62520 + }, + { + "epoch": 12.139410708218048, + "grad_norm": 0.10436850041151047, + "learning_rate": 3.371968148480635e-05, + "loss": 0.025, + "step": 62530 + }, + { + "epoch": 12.141352361535848, + "grad_norm": 0.17620569467544556, + "learning_rate": 3.370404872860937e-05, + "loss": 0.0224, + "step": 62540 + }, + { + "epoch": 12.143294014853648, + "grad_norm": 0.17841869592666626, + "learning_rate": 3.3688417754513145e-05, + "loss": 0.0232, + "step": 62550 + }, + { + "epoch": 12.145235668171448, + "grad_norm": 0.1800888180732727, + "learning_rate": 3.367278856422703e-05, + "loss": 0.0189, + "step": 62560 + }, + { + "epoch": 12.147177321489249, + "grad_norm": 0.09616659581661224, + "learning_rate": 3.365716115946021e-05, + "loss": 0.0199, + "step": 62570 + }, + { + "epoch": 12.149118974807049, + "grad_norm": 0.2257307469844818, + "learning_rate": 3.3641535541921666e-05, + "loss": 0.0229, + "step": 62580 + }, + { + "epoch": 12.151060628124847, + "grad_norm": 0.14122368395328522, + "learning_rate": 3.36259117133202e-05, + "loss": 0.0241, + "step": 62590 + }, + { + "epoch": 12.153002281442648, + "grad_norm": 0.17090953886508942, + "learning_rate": 3.36102896753644e-05, + "loss": 0.029, + "step": 62600 + }, + { + "epoch": 12.154943934760448, + "grad_norm": 0.197663813829422, + "learning_rate": 3.35946694297627e-05, + "loss": 0.0187, + "step": 62610 + }, + { + "epoch": 12.156885588078248, + "grad_norm": 0.14470773935317993, + "learning_rate": 3.357905097822328e-05, + "loss": 0.0218, + "step": 62620 + }, + { + "epoch": 12.158827241396049, + "grad_norm": 0.0996636375784874, + "learning_rate": 3.356343432245416e-05, + "loss": 0.0213, + "step": 62630 + }, + { + "epoch": 12.160768894713849, + "grad_norm": 0.1294059455394745, + "learning_rate": 3.354781946416315e-05, + "loss": 0.0231, + "step": 62640 + }, + { + "epoch": 12.162710548031649, + "grad_norm": 0.1670428365468979, + "learning_rate": 3.3532206405057864e-05, + "loss": 0.0302, + "step": 62650 + }, + { + "epoch": 12.16465220134945, + "grad_norm": 0.1889348030090332, + "learning_rate": 3.351659514684573e-05, + "loss": 0.0247, + "step": 62660 + }, + { + "epoch": 12.16659385466725, + "grad_norm": 0.183235302567482, + "learning_rate": 3.350098569123395e-05, + "loss": 0.0262, + "step": 62670 + }, + { + "epoch": 12.16853550798505, + "grad_norm": 0.13260070979595184, + "learning_rate": 3.348537803992958e-05, + "loss": 0.0212, + "step": 62680 + }, + { + "epoch": 12.17047716130285, + "grad_norm": 0.1507459580898285, + "learning_rate": 3.346977219463945e-05, + "loss": 0.0192, + "step": 62690 + }, + { + "epoch": 12.172418814620649, + "grad_norm": 0.16756895184516907, + "learning_rate": 3.345416815707017e-05, + "loss": 0.0201, + "step": 62700 + }, + { + "epoch": 12.174360467938449, + "grad_norm": 0.1909259855747223, + "learning_rate": 3.3438565928928185e-05, + "loss": 0.0225, + "step": 62710 + }, + { + "epoch": 12.17630212125625, + "grad_norm": 0.22870776057243347, + "learning_rate": 3.3422965511919736e-05, + "loss": 0.0221, + "step": 62720 + }, + { + "epoch": 12.17824377457405, + "grad_norm": 0.15137892961502075, + "learning_rate": 3.3407366907750846e-05, + "loss": 0.0239, + "step": 62730 + }, + { + "epoch": 12.18018542789185, + "grad_norm": 0.155155211687088, + "learning_rate": 3.339177011812737e-05, + "loss": 0.0196, + "step": 62740 + }, + { + "epoch": 12.18212708120965, + "grad_norm": 0.1858144998550415, + "learning_rate": 3.3376175144754935e-05, + "loss": 0.0237, + "step": 62750 + }, + { + "epoch": 12.18406873452745, + "grad_norm": 0.11708983033895493, + "learning_rate": 3.3360581989339004e-05, + "loss": 0.0233, + "step": 62760 + }, + { + "epoch": 12.18601038784525, + "grad_norm": 0.13067243993282318, + "learning_rate": 3.3344990653584804e-05, + "loss": 0.0207, + "step": 62770 + }, + { + "epoch": 12.18795204116305, + "grad_norm": 0.1691751480102539, + "learning_rate": 3.332940113919739e-05, + "loss": 0.0228, + "step": 62780 + }, + { + "epoch": 12.189893694480851, + "grad_norm": 0.18888166546821594, + "learning_rate": 3.331381344788161e-05, + "loss": 0.0237, + "step": 62790 + }, + { + "epoch": 12.191835347798651, + "grad_norm": 0.2964826226234436, + "learning_rate": 3.3298227581342076e-05, + "loss": 0.0221, + "step": 62800 + }, + { + "epoch": 12.19377700111645, + "grad_norm": 0.12580856680870056, + "learning_rate": 3.328264354128329e-05, + "loss": 0.0224, + "step": 62810 + }, + { + "epoch": 12.19571865443425, + "grad_norm": 0.12758716940879822, + "learning_rate": 3.3267061329409465e-05, + "loss": 0.029, + "step": 62820 + }, + { + "epoch": 12.19766030775205, + "grad_norm": 0.23947465419769287, + "learning_rate": 3.325148094742466e-05, + "loss": 0.0232, + "step": 62830 + }, + { + "epoch": 12.19960196106985, + "grad_norm": 0.16408458352088928, + "learning_rate": 3.323590239703271e-05, + "loss": 0.0203, + "step": 62840 + }, + { + "epoch": 12.201543614387651, + "grad_norm": 0.08823619037866592, + "learning_rate": 3.322032567993728e-05, + "loss": 0.0212, + "step": 62850 + }, + { + "epoch": 12.203485267705451, + "grad_norm": 0.17075881361961365, + "learning_rate": 3.3204750797841774e-05, + "loss": 0.0274, + "step": 62860 + }, + { + "epoch": 12.205426921023252, + "grad_norm": 0.2634963095188141, + "learning_rate": 3.31891777524495e-05, + "loss": 0.0229, + "step": 62870 + }, + { + "epoch": 12.207368574341052, + "grad_norm": 0.2209424078464508, + "learning_rate": 3.317360654546347e-05, + "loss": 0.0249, + "step": 62880 + }, + { + "epoch": 12.209310227658852, + "grad_norm": 0.16600774228572845, + "learning_rate": 3.3158037178586534e-05, + "loss": 0.0295, + "step": 62890 + }, + { + "epoch": 12.211251880976652, + "grad_norm": 0.10964582860469818, + "learning_rate": 3.3142469653521316e-05, + "loss": 0.0212, + "step": 62900 + }, + { + "epoch": 12.21319353429445, + "grad_norm": 0.14215900003910065, + "learning_rate": 3.3126903971970284e-05, + "loss": 0.0215, + "step": 62910 + }, + { + "epoch": 12.215135187612251, + "grad_norm": 0.11241460591554642, + "learning_rate": 3.311134013563564e-05, + "loss": 0.02, + "step": 62920 + }, + { + "epoch": 12.217076840930051, + "grad_norm": 0.1639251857995987, + "learning_rate": 3.309577814621947e-05, + "loss": 0.0221, + "step": 62930 + }, + { + "epoch": 12.219018494247852, + "grad_norm": 0.30141425132751465, + "learning_rate": 3.308021800542359e-05, + "loss": 0.0215, + "step": 62940 + }, + { + "epoch": 12.220960147565652, + "grad_norm": 0.11261779069900513, + "learning_rate": 3.306465971494963e-05, + "loss": 0.0314, + "step": 62950 + }, + { + "epoch": 12.222901800883452, + "grad_norm": 0.20871390402317047, + "learning_rate": 3.304910327649901e-05, + "loss": 0.0197, + "step": 62960 + }, + { + "epoch": 12.224843454201253, + "grad_norm": 0.11770880222320557, + "learning_rate": 3.303354869177297e-05, + "loss": 0.0194, + "step": 62970 + }, + { + "epoch": 12.226785107519053, + "grad_norm": 0.26325318217277527, + "learning_rate": 3.301799596247255e-05, + "loss": 0.026, + "step": 62980 + }, + { + "epoch": 12.228726760836853, + "grad_norm": 0.07625202089548111, + "learning_rate": 3.300244509029854e-05, + "loss": 0.0282, + "step": 62990 + }, + { + "epoch": 12.230668414154653, + "grad_norm": 0.2308763712644577, + "learning_rate": 3.2986896076951585e-05, + "loss": 0.0252, + "step": 63000 + }, + { + "epoch": 12.232610067472454, + "grad_norm": 0.14444345235824585, + "learning_rate": 3.297134892413211e-05, + "loss": 0.0204, + "step": 63010 + }, + { + "epoch": 12.234551720790252, + "grad_norm": 0.2384175956249237, + "learning_rate": 3.295580363354031e-05, + "loss": 0.0287, + "step": 63020 + }, + { + "epoch": 12.236493374108052, + "grad_norm": 0.18024608492851257, + "learning_rate": 3.2940260206876194e-05, + "loss": 0.0267, + "step": 63030 + }, + { + "epoch": 12.238435027425853, + "grad_norm": 0.13891851902008057, + "learning_rate": 3.292471864583958e-05, + "loss": 0.0197, + "step": 63040 + }, + { + "epoch": 12.240376680743653, + "grad_norm": 0.10807763785123825, + "learning_rate": 3.290917895213004e-05, + "loss": 0.0201, + "step": 63050 + }, + { + "epoch": 12.242318334061453, + "grad_norm": 0.13629662990570068, + "learning_rate": 3.289364112744701e-05, + "loss": 0.0242, + "step": 63060 + }, + { + "epoch": 12.244259987379253, + "grad_norm": 0.17187029123306274, + "learning_rate": 3.287810517348967e-05, + "loss": 0.0313, + "step": 63070 + }, + { + "epoch": 12.246201640697054, + "grad_norm": 0.1938142627477646, + "learning_rate": 3.2862571091957014e-05, + "loss": 0.0251, + "step": 63080 + }, + { + "epoch": 12.248143294014854, + "grad_norm": 0.13925860822200775, + "learning_rate": 3.284703888454781e-05, + "loss": 0.026, + "step": 63090 + }, + { + "epoch": 12.250084947332654, + "grad_norm": 0.16108669340610504, + "learning_rate": 3.2831508552960646e-05, + "loss": 0.0172, + "step": 63100 + }, + { + "epoch": 12.252026600650455, + "grad_norm": 0.232346311211586, + "learning_rate": 3.281598009889387e-05, + "loss": 0.0201, + "step": 63110 + }, + { + "epoch": 12.253968253968253, + "grad_norm": 0.26810842752456665, + "learning_rate": 3.28004535240457e-05, + "loss": 0.0321, + "step": 63120 + }, + { + "epoch": 12.255909907286053, + "grad_norm": 0.13015665113925934, + "learning_rate": 3.278492883011407e-05, + "loss": 0.0219, + "step": 63130 + }, + { + "epoch": 12.257851560603854, + "grad_norm": 0.16061699390411377, + "learning_rate": 3.276940601879675e-05, + "loss": 0.0269, + "step": 63140 + }, + { + "epoch": 12.259793213921654, + "grad_norm": 0.18252220749855042, + "learning_rate": 3.275388509179127e-05, + "loss": 0.032, + "step": 63150 + }, + { + "epoch": 12.261734867239454, + "grad_norm": 0.08736129850149155, + "learning_rate": 3.2738366050794985e-05, + "loss": 0.0288, + "step": 63160 + }, + { + "epoch": 12.263676520557254, + "grad_norm": 0.17521487176418304, + "learning_rate": 3.2722848897505034e-05, + "loss": 0.0199, + "step": 63170 + }, + { + "epoch": 12.265618173875055, + "grad_norm": 0.15318338572978973, + "learning_rate": 3.2707333633618356e-05, + "loss": 0.0227, + "step": 63180 + }, + { + "epoch": 12.267559827192855, + "grad_norm": 0.137571319937706, + "learning_rate": 3.269182026083167e-05, + "loss": 0.0234, + "step": 63190 + }, + { + "epoch": 12.269501480510655, + "grad_norm": 0.15947887301445007, + "learning_rate": 3.2676308780841516e-05, + "loss": 0.0283, + "step": 63200 + }, + { + "epoch": 12.271443133828456, + "grad_norm": 0.36167648434638977, + "learning_rate": 3.266079919534416e-05, + "loss": 0.0305, + "step": 63210 + }, + { + "epoch": 12.273384787146256, + "grad_norm": 0.2332013100385666, + "learning_rate": 3.264529150603575e-05, + "loss": 0.0209, + "step": 63220 + }, + { + "epoch": 12.275326440464054, + "grad_norm": 0.19039539992809296, + "learning_rate": 3.262978571461217e-05, + "loss": 0.0278, + "step": 63230 + }, + { + "epoch": 12.277268093781855, + "grad_norm": 0.12248498946428299, + "learning_rate": 3.26142818227691e-05, + "loss": 0.0179, + "step": 63240 + }, + { + "epoch": 12.279209747099655, + "grad_norm": 0.2894190549850464, + "learning_rate": 3.259877983220203e-05, + "loss": 0.0245, + "step": 63250 + }, + { + "epoch": 12.281151400417455, + "grad_norm": 0.16787797212600708, + "learning_rate": 3.2583279744606246e-05, + "loss": 0.0205, + "step": 63260 + }, + { + "epoch": 12.283093053735255, + "grad_norm": 0.23387356102466583, + "learning_rate": 3.25677815616768e-05, + "loss": 0.0252, + "step": 63270 + }, + { + "epoch": 12.285034707053056, + "grad_norm": 0.12100182473659515, + "learning_rate": 3.255228528510856e-05, + "loss": 0.0212, + "step": 63280 + }, + { + "epoch": 12.286976360370856, + "grad_norm": 0.1407461017370224, + "learning_rate": 3.253679091659617e-05, + "loss": 0.0187, + "step": 63290 + }, + { + "epoch": 12.288918013688656, + "grad_norm": 0.17204558849334717, + "learning_rate": 3.252129845783404e-05, + "loss": 0.0227, + "step": 63300 + }, + { + "epoch": 12.290859667006456, + "grad_norm": 0.19899982213974, + "learning_rate": 3.2505807910516456e-05, + "loss": 0.0218, + "step": 63310 + }, + { + "epoch": 12.292801320324257, + "grad_norm": 0.2671632468700409, + "learning_rate": 3.249031927633742e-05, + "loss": 0.0205, + "step": 63320 + }, + { + "epoch": 12.294742973642057, + "grad_norm": 0.25814205408096313, + "learning_rate": 3.2474832556990737e-05, + "loss": 0.024, + "step": 63330 + }, + { + "epoch": 12.296684626959856, + "grad_norm": 0.1700914353132248, + "learning_rate": 3.245934775417002e-05, + "loss": 0.0267, + "step": 63340 + }, + { + "epoch": 12.298626280277656, + "grad_norm": 0.17677004635334015, + "learning_rate": 3.244386486956866e-05, + "loss": 0.0251, + "step": 63350 + }, + { + "epoch": 12.300567933595456, + "grad_norm": 0.1758670061826706, + "learning_rate": 3.2428383904879846e-05, + "loss": 0.0174, + "step": 63360 + }, + { + "epoch": 12.302509586913256, + "grad_norm": 0.15205448865890503, + "learning_rate": 3.241290486179652e-05, + "loss": 0.0199, + "step": 63370 + }, + { + "epoch": 12.304451240231057, + "grad_norm": 0.17980873584747314, + "learning_rate": 3.239742774201151e-05, + "loss": 0.0316, + "step": 63380 + }, + { + "epoch": 12.306392893548857, + "grad_norm": 0.1436787247657776, + "learning_rate": 3.238195254721733e-05, + "loss": 0.0285, + "step": 63390 + }, + { + "epoch": 12.308334546866657, + "grad_norm": 0.15966910123825073, + "learning_rate": 3.236647927910633e-05, + "loss": 0.0222, + "step": 63400 + }, + { + "epoch": 12.310276200184457, + "grad_norm": 0.18776899576187134, + "learning_rate": 3.2351007939370646e-05, + "loss": 0.022, + "step": 63410 + }, + { + "epoch": 12.312217853502258, + "grad_norm": 0.20296500623226166, + "learning_rate": 3.2335538529702184e-05, + "loss": 0.0227, + "step": 63420 + }, + { + "epoch": 12.314159506820058, + "grad_norm": 0.19262894988059998, + "learning_rate": 3.232007105179268e-05, + "loss": 0.0285, + "step": 63430 + }, + { + "epoch": 12.316101160137858, + "grad_norm": 0.2028058022260666, + "learning_rate": 3.230460550733363e-05, + "loss": 0.0268, + "step": 63440 + }, + { + "epoch": 12.318042813455657, + "grad_norm": 0.19966888427734375, + "learning_rate": 3.2289141898016315e-05, + "loss": 0.0243, + "step": 63450 + }, + { + "epoch": 12.319984466773457, + "grad_norm": 0.2298109084367752, + "learning_rate": 3.227368022553181e-05, + "loss": 0.0218, + "step": 63460 + }, + { + "epoch": 12.321926120091257, + "grad_norm": 0.2644821107387543, + "learning_rate": 3.225822049157099e-05, + "loss": 0.0261, + "step": 63470 + }, + { + "epoch": 12.323867773409058, + "grad_norm": 0.15262404084205627, + "learning_rate": 3.2242762697824494e-05, + "loss": 0.0209, + "step": 63480 + }, + { + "epoch": 12.325809426726858, + "grad_norm": 0.14786851406097412, + "learning_rate": 3.222730684598279e-05, + "loss": 0.0239, + "step": 63490 + }, + { + "epoch": 12.327751080044658, + "grad_norm": 0.11001919955015182, + "learning_rate": 3.221185293773605e-05, + "loss": 0.0244, + "step": 63500 + }, + { + "epoch": 12.329692733362458, + "grad_norm": 0.09407351166009903, + "learning_rate": 3.2196400974774354e-05, + "loss": 0.0197, + "step": 63510 + }, + { + "epoch": 12.331634386680259, + "grad_norm": 0.11238384246826172, + "learning_rate": 3.2180950958787485e-05, + "loss": 0.0248, + "step": 63520 + }, + { + "epoch": 12.333576039998059, + "grad_norm": 0.1708657294511795, + "learning_rate": 3.2165502891465024e-05, + "loss": 0.0266, + "step": 63530 + }, + { + "epoch": 12.33551769331586, + "grad_norm": 0.2045331597328186, + "learning_rate": 3.215005677449636e-05, + "loss": 0.0218, + "step": 63540 + }, + { + "epoch": 12.337459346633658, + "grad_norm": 0.17522171139717102, + "learning_rate": 3.2134612609570644e-05, + "loss": 0.0256, + "step": 63550 + }, + { + "epoch": 12.339400999951458, + "grad_norm": 0.1795661300420761, + "learning_rate": 3.211917039837682e-05, + "loss": 0.0238, + "step": 63560 + }, + { + "epoch": 12.341342653269258, + "grad_norm": 0.1901790052652359, + "learning_rate": 3.210373014260365e-05, + "loss": 0.031, + "step": 63570 + }, + { + "epoch": 12.343284306587059, + "grad_norm": 0.17348505556583405, + "learning_rate": 3.208829184393965e-05, + "loss": 0.0235, + "step": 63580 + }, + { + "epoch": 12.345225959904859, + "grad_norm": 0.0955720990896225, + "learning_rate": 3.207285550407312e-05, + "loss": 0.0214, + "step": 63590 + }, + { + "epoch": 12.347167613222659, + "grad_norm": 0.1694360375404358, + "learning_rate": 3.205742112469217e-05, + "loss": 0.0253, + "step": 63600 + }, + { + "epoch": 12.34910926654046, + "grad_norm": 0.31819722056388855, + "learning_rate": 3.204198870748465e-05, + "loss": 0.0219, + "step": 63610 + }, + { + "epoch": 12.35105091985826, + "grad_norm": 0.13192494213581085, + "learning_rate": 3.202655825413824e-05, + "loss": 0.0188, + "step": 63620 + }, + { + "epoch": 12.35299257317606, + "grad_norm": 0.1981651335954666, + "learning_rate": 3.2011129766340416e-05, + "loss": 0.0224, + "step": 63630 + }, + { + "epoch": 12.35493422649386, + "grad_norm": 0.17251473665237427, + "learning_rate": 3.1995703245778384e-05, + "loss": 0.0224, + "step": 63640 + }, + { + "epoch": 12.35687587981166, + "grad_norm": 0.1736515462398529, + "learning_rate": 3.1980278694139186e-05, + "loss": 0.0296, + "step": 63650 + }, + { + "epoch": 12.358817533129459, + "grad_norm": 0.1457744538784027, + "learning_rate": 3.196485611310962e-05, + "loss": 0.0249, + "step": 63660 + }, + { + "epoch": 12.36075918644726, + "grad_norm": 0.2833820581436157, + "learning_rate": 3.1949435504376257e-05, + "loss": 0.0285, + "step": 63670 + }, + { + "epoch": 12.36270083976506, + "grad_norm": 0.12295127660036087, + "learning_rate": 3.1934016869625505e-05, + "loss": 0.0228, + "step": 63680 + }, + { + "epoch": 12.36464249308286, + "grad_norm": 0.13429129123687744, + "learning_rate": 3.191860021054349e-05, + "loss": 0.0271, + "step": 63690 + }, + { + "epoch": 12.36658414640066, + "grad_norm": 0.2489231377840042, + "learning_rate": 3.190318552881618e-05, + "loss": 0.0272, + "step": 63700 + }, + { + "epoch": 12.36852579971846, + "grad_norm": 0.20513156056404114, + "learning_rate": 3.1887772826129285e-05, + "loss": 0.0207, + "step": 63710 + }, + { + "epoch": 12.37046745303626, + "grad_norm": 0.42527297139167786, + "learning_rate": 3.1872362104168335e-05, + "loss": 0.0199, + "step": 63720 + }, + { + "epoch": 12.37240910635406, + "grad_norm": 0.2165602445602417, + "learning_rate": 3.185695336461861e-05, + "loss": 0.0253, + "step": 63730 + }, + { + "epoch": 12.374350759671861, + "grad_norm": 0.05156926065683365, + "learning_rate": 3.184154660916519e-05, + "loss": 0.027, + "step": 63740 + }, + { + "epoch": 12.376292412989661, + "grad_norm": 0.11221352964639664, + "learning_rate": 3.1826141839492906e-05, + "loss": 0.022, + "step": 63750 + }, + { + "epoch": 12.37823406630746, + "grad_norm": 0.13735635578632355, + "learning_rate": 3.181073905728644e-05, + "loss": 0.0196, + "step": 63760 + }, + { + "epoch": 12.38017571962526, + "grad_norm": 0.21365812420845032, + "learning_rate": 3.1795338264230214e-05, + "loss": 0.025, + "step": 63770 + }, + { + "epoch": 12.38211737294306, + "grad_norm": 0.19627802073955536, + "learning_rate": 3.1779939462008426e-05, + "loss": 0.022, + "step": 63780 + }, + { + "epoch": 12.38405902626086, + "grad_norm": 0.15826846659183502, + "learning_rate": 3.1764542652305066e-05, + "loss": 0.0226, + "step": 63790 + }, + { + "epoch": 12.386000679578661, + "grad_norm": 0.1651902049779892, + "learning_rate": 3.17491478368039e-05, + "loss": 0.0285, + "step": 63800 + }, + { + "epoch": 12.387942332896461, + "grad_norm": 0.15156929194927216, + "learning_rate": 3.173375501718847e-05, + "loss": 0.0296, + "step": 63810 + }, + { + "epoch": 12.389883986214262, + "grad_norm": 0.12723854184150696, + "learning_rate": 3.171836419514215e-05, + "loss": 0.0179, + "step": 63820 + }, + { + "epoch": 12.391825639532062, + "grad_norm": 0.29071876406669617, + "learning_rate": 3.170297537234803e-05, + "loss": 0.0238, + "step": 63830 + }, + { + "epoch": 12.393767292849862, + "grad_norm": 0.18716776371002197, + "learning_rate": 3.168758855048901e-05, + "loss": 0.0214, + "step": 63840 + }, + { + "epoch": 12.395708946167662, + "grad_norm": 0.15759818255901337, + "learning_rate": 3.1672203731247785e-05, + "loss": 0.0207, + "step": 63850 + }, + { + "epoch": 12.397650599485463, + "grad_norm": 0.15700924396514893, + "learning_rate": 3.16568209163068e-05, + "loss": 0.0219, + "step": 63860 + }, + { + "epoch": 12.399592252803261, + "grad_norm": 0.25813356041908264, + "learning_rate": 3.1641440107348295e-05, + "loss": 0.0223, + "step": 63870 + }, + { + "epoch": 12.401533906121061, + "grad_norm": 0.24195152521133423, + "learning_rate": 3.162606130605429e-05, + "loss": 0.0273, + "step": 63880 + }, + { + "epoch": 12.403475559438862, + "grad_norm": 0.19894026219844818, + "learning_rate": 3.16106845141066e-05, + "loss": 0.0279, + "step": 63890 + }, + { + "epoch": 12.405417212756662, + "grad_norm": 0.1924581080675125, + "learning_rate": 3.159530973318682e-05, + "loss": 0.022, + "step": 63900 + }, + { + "epoch": 12.407358866074462, + "grad_norm": 0.20022577047348022, + "learning_rate": 3.1579936964976286e-05, + "loss": 0.0225, + "step": 63910 + }, + { + "epoch": 12.409300519392263, + "grad_norm": 0.12302602827548981, + "learning_rate": 3.156456621115615e-05, + "loss": 0.0294, + "step": 63920 + }, + { + "epoch": 12.411242172710063, + "grad_norm": 0.22358323633670807, + "learning_rate": 3.154919747340734e-05, + "loss": 0.0317, + "step": 63930 + }, + { + "epoch": 12.413183826027863, + "grad_norm": 0.16725285351276398, + "learning_rate": 3.153383075341054e-05, + "loss": 0.0216, + "step": 63940 + }, + { + "epoch": 12.415125479345663, + "grad_norm": 0.1322431117296219, + "learning_rate": 3.1518466052846264e-05, + "loss": 0.0229, + "step": 63950 + }, + { + "epoch": 12.417067132663464, + "grad_norm": 0.0744406208395958, + "learning_rate": 3.150310337339474e-05, + "loss": 0.0267, + "step": 63960 + }, + { + "epoch": 12.419008785981264, + "grad_norm": 0.22873534262180328, + "learning_rate": 3.148774271673603e-05, + "loss": 0.0254, + "step": 63970 + }, + { + "epoch": 12.420950439299062, + "grad_norm": 0.24126963317394257, + "learning_rate": 3.147238408454995e-05, + "loss": 0.0217, + "step": 63980 + }, + { + "epoch": 12.422892092616863, + "grad_norm": 0.1850498616695404, + "learning_rate": 3.1457027478516086e-05, + "loss": 0.0213, + "step": 63990 + }, + { + "epoch": 12.424833745934663, + "grad_norm": 0.35500097274780273, + "learning_rate": 3.144167290031382e-05, + "loss": 0.0193, + "step": 64000 + }, + { + "epoch": 12.426775399252463, + "grad_norm": 0.16481991112232208, + "learning_rate": 3.142632035162227e-05, + "loss": 0.0289, + "step": 64010 + }, + { + "epoch": 12.428717052570263, + "grad_norm": 0.1332075297832489, + "learning_rate": 3.141096983412044e-05, + "loss": 0.0195, + "step": 64020 + }, + { + "epoch": 12.430658705888064, + "grad_norm": 0.18494530022144318, + "learning_rate": 3.139562134948699e-05, + "loss": 0.0201, + "step": 64030 + }, + { + "epoch": 12.432600359205864, + "grad_norm": 0.25354769825935364, + "learning_rate": 3.1380274899400424e-05, + "loss": 0.0249, + "step": 64040 + }, + { + "epoch": 12.434542012523664, + "grad_norm": 0.2533554434776306, + "learning_rate": 3.1364930485539e-05, + "loss": 0.0242, + "step": 64050 + }, + { + "epoch": 12.436483665841465, + "grad_norm": 0.12214088439941406, + "learning_rate": 3.134958810958076e-05, + "loss": 0.0246, + "step": 64060 + }, + { + "epoch": 12.438425319159265, + "grad_norm": 0.27670228481292725, + "learning_rate": 3.13342477732035e-05, + "loss": 0.0262, + "step": 64070 + }, + { + "epoch": 12.440366972477065, + "grad_norm": 0.1525597870349884, + "learning_rate": 3.131890947808487e-05, + "loss": 0.0228, + "step": 64080 + }, + { + "epoch": 12.442308625794864, + "grad_norm": 0.20525489747524261, + "learning_rate": 3.13035732259022e-05, + "loss": 0.0256, + "step": 64090 + }, + { + "epoch": 12.444250279112664, + "grad_norm": 0.20419470965862274, + "learning_rate": 3.1288239018332656e-05, + "loss": 0.0216, + "step": 64100 + }, + { + "epoch": 12.446191932430464, + "grad_norm": 0.16645607352256775, + "learning_rate": 3.1272906857053164e-05, + "loss": 0.0213, + "step": 64110 + }, + { + "epoch": 12.448133585748264, + "grad_norm": 0.15949635207653046, + "learning_rate": 3.125757674374042e-05, + "loss": 0.0268, + "step": 64120 + }, + { + "epoch": 12.450075239066065, + "grad_norm": 0.17885667085647583, + "learning_rate": 3.124224868007088e-05, + "loss": 0.0166, + "step": 64130 + }, + { + "epoch": 12.452016892383865, + "grad_norm": 0.1042613685131073, + "learning_rate": 3.1226922667720854e-05, + "loss": 0.022, + "step": 64140 + }, + { + "epoch": 12.453958545701665, + "grad_norm": 0.24690508842468262, + "learning_rate": 3.121159870836633e-05, + "loss": 0.0186, + "step": 64150 + }, + { + "epoch": 12.455900199019466, + "grad_norm": 0.16928108036518097, + "learning_rate": 3.119627680368312e-05, + "loss": 0.0255, + "step": 64160 + }, + { + "epoch": 12.457841852337266, + "grad_norm": 0.2454218715429306, + "learning_rate": 3.1180956955346816e-05, + "loss": 0.0259, + "step": 64170 + }, + { + "epoch": 12.459783505655066, + "grad_norm": 0.1722285896539688, + "learning_rate": 3.116563916503276e-05, + "loss": 0.0218, + "step": 64180 + }, + { + "epoch": 12.461725158972865, + "grad_norm": 0.20146699249744415, + "learning_rate": 3.1150323434416095e-05, + "loss": 0.0298, + "step": 64190 + }, + { + "epoch": 12.463666812290665, + "grad_norm": 0.15848827362060547, + "learning_rate": 3.1135009765171696e-05, + "loss": 0.0224, + "step": 64200 + }, + { + "epoch": 12.465608465608465, + "grad_norm": 0.16571933031082153, + "learning_rate": 3.111969815897429e-05, + "loss": 0.0289, + "step": 64210 + }, + { + "epoch": 12.467550118926265, + "grad_norm": 0.08944110572338104, + "learning_rate": 3.1104388617498295e-05, + "loss": 0.0192, + "step": 64220 + }, + { + "epoch": 12.469491772244066, + "grad_norm": 0.18187052011489868, + "learning_rate": 3.1089081142417966e-05, + "loss": 0.0202, + "step": 64230 + }, + { + "epoch": 12.471433425561866, + "grad_norm": 0.573434591293335, + "learning_rate": 3.1073775735407285e-05, + "loss": 0.0235, + "step": 64240 + }, + { + "epoch": 12.473375078879666, + "grad_norm": 0.20023714005947113, + "learning_rate": 3.1058472398140035e-05, + "loss": 0.0304, + "step": 64250 + }, + { + "epoch": 12.475316732197467, + "grad_norm": 0.21721619367599487, + "learning_rate": 3.1043171132289736e-05, + "loss": 0.0196, + "step": 64260 + }, + { + "epoch": 12.477258385515267, + "grad_norm": 0.12348467856645584, + "learning_rate": 3.1027871939529774e-05, + "loss": 0.0273, + "step": 64270 + }, + { + "epoch": 12.479200038833067, + "grad_norm": 0.291385680437088, + "learning_rate": 3.1012574821533214e-05, + "loss": 0.0233, + "step": 64280 + }, + { + "epoch": 12.481141692150867, + "grad_norm": 0.17560602724552155, + "learning_rate": 3.099727977997291e-05, + "loss": 0.0245, + "step": 64290 + }, + { + "epoch": 12.483083345468666, + "grad_norm": 0.15891025960445404, + "learning_rate": 3.098198681652154e-05, + "loss": 0.0279, + "step": 64300 + }, + { + "epoch": 12.485024998786466, + "grad_norm": 0.24916046857833862, + "learning_rate": 3.0966695932851475e-05, + "loss": 0.0262, + "step": 64310 + }, + { + "epoch": 12.486966652104266, + "grad_norm": 0.18265336751937866, + "learning_rate": 3.0951407130634926e-05, + "loss": 0.0303, + "step": 64320 + }, + { + "epoch": 12.488908305422067, + "grad_norm": 0.20533119142055511, + "learning_rate": 3.0936120411543864e-05, + "loss": 0.0248, + "step": 64330 + }, + { + "epoch": 12.490849958739867, + "grad_norm": 0.08745497465133667, + "learning_rate": 3.0920835777250015e-05, + "loss": 0.0285, + "step": 64340 + }, + { + "epoch": 12.492791612057667, + "grad_norm": 0.20919322967529297, + "learning_rate": 3.090555322942488e-05, + "loss": 0.0272, + "step": 64350 + }, + { + "epoch": 12.494733265375467, + "grad_norm": 0.2584279179573059, + "learning_rate": 3.089027276973973e-05, + "loss": 0.0302, + "step": 64360 + }, + { + "epoch": 12.496674918693268, + "grad_norm": 0.18335776031017303, + "learning_rate": 3.0874994399865615e-05, + "loss": 0.0236, + "step": 64370 + }, + { + "epoch": 12.498616572011068, + "grad_norm": 0.16925473511219025, + "learning_rate": 3.085971812147337e-05, + "loss": 0.019, + "step": 64380 + }, + { + "epoch": 12.500558225328868, + "grad_norm": 0.19957949221134186, + "learning_rate": 3.0844443936233557e-05, + "loss": 0.0194, + "step": 64390 + }, + { + "epoch": 12.502499878646667, + "grad_norm": 0.12335006147623062, + "learning_rate": 3.082917184581657e-05, + "loss": 0.0203, + "step": 64400 + }, + { + "epoch": 12.504441531964467, + "grad_norm": 0.1671629399061203, + "learning_rate": 3.0813901851892513e-05, + "loss": 0.0194, + "step": 64410 + }, + { + "epoch": 12.506383185282267, + "grad_norm": 0.21057723462581635, + "learning_rate": 3.0798633956131316e-05, + "loss": 0.0241, + "step": 64420 + }, + { + "epoch": 12.508324838600068, + "grad_norm": 0.21244513988494873, + "learning_rate": 3.078336816020264e-05, + "loss": 0.0189, + "step": 64430 + }, + { + "epoch": 12.510266491917868, + "grad_norm": 0.044500503689050674, + "learning_rate": 3.0768104465775935e-05, + "loss": 0.0225, + "step": 64440 + }, + { + "epoch": 12.512208145235668, + "grad_norm": 0.2362666130065918, + "learning_rate": 3.0752842874520396e-05, + "loss": 0.0191, + "step": 64450 + }, + { + "epoch": 12.514149798553468, + "grad_norm": 0.26228249073028564, + "learning_rate": 3.073758338810503e-05, + "loss": 0.0247, + "step": 64460 + }, + { + "epoch": 12.516091451871269, + "grad_norm": 0.11571791023015976, + "learning_rate": 3.072232600819859e-05, + "loss": 0.0212, + "step": 64470 + }, + { + "epoch": 12.518033105189069, + "grad_norm": 0.2036418616771698, + "learning_rate": 3.070707073646959e-05, + "loss": 0.0215, + "step": 64480 + }, + { + "epoch": 12.51997475850687, + "grad_norm": 0.17154847085475922, + "learning_rate": 3.069181757458633e-05, + "loss": 0.0255, + "step": 64490 + }, + { + "epoch": 12.52191641182467, + "grad_norm": 0.1270340383052826, + "learning_rate": 3.067656652421687e-05, + "loss": 0.0313, + "step": 64500 + }, + { + "epoch": 12.523858065142468, + "grad_norm": 0.13946160674095154, + "learning_rate": 3.0661317587029025e-05, + "loss": 0.02, + "step": 64510 + }, + { + "epoch": 12.525799718460268, + "grad_norm": 0.16687685251235962, + "learning_rate": 3.064607076469043e-05, + "loss": 0.0147, + "step": 64520 + }, + { + "epoch": 12.527741371778069, + "grad_norm": 0.15253320336341858, + "learning_rate": 3.063082605886845e-05, + "loss": 0.0215, + "step": 64530 + }, + { + "epoch": 12.529683025095869, + "grad_norm": 0.15350128710269928, + "learning_rate": 3.061558347123019e-05, + "loss": 0.023, + "step": 64540 + }, + { + "epoch": 12.53162467841367, + "grad_norm": 0.15283627808094025, + "learning_rate": 3.060034300344259e-05, + "loss": 0.022, + "step": 64550 + }, + { + "epoch": 12.53356633173147, + "grad_norm": 0.24672088027000427, + "learning_rate": 3.058510465717231e-05, + "loss": 0.0213, + "step": 64560 + }, + { + "epoch": 12.53550798504927, + "grad_norm": 0.22893311083316803, + "learning_rate": 3.0569868434085804e-05, + "loss": 0.0219, + "step": 64570 + }, + { + "epoch": 12.53744963836707, + "grad_norm": 0.10111208260059357, + "learning_rate": 3.055463433584924e-05, + "loss": 0.023, + "step": 64580 + }, + { + "epoch": 12.53939129168487, + "grad_norm": 0.2074316143989563, + "learning_rate": 3.0539402364128656e-05, + "loss": 0.033, + "step": 64590 + }, + { + "epoch": 12.54133294500267, + "grad_norm": 0.1225058063864708, + "learning_rate": 3.052417252058977e-05, + "loss": 0.016, + "step": 64600 + }, + { + "epoch": 12.543274598320469, + "grad_norm": 0.14242441952228546, + "learning_rate": 3.0508944806898092e-05, + "loss": 0.0268, + "step": 64610 + }, + { + "epoch": 12.54521625163827, + "grad_norm": 0.3040463626384735, + "learning_rate": 3.04937192247189e-05, + "loss": 0.025, + "step": 64620 + }, + { + "epoch": 12.54715790495607, + "grad_norm": 0.14220651984214783, + "learning_rate": 3.047849577571725e-05, + "loss": 0.0251, + "step": 64630 + }, + { + "epoch": 12.54909955827387, + "grad_norm": 0.19118721783161163, + "learning_rate": 3.046327446155795e-05, + "loss": 0.0236, + "step": 64640 + }, + { + "epoch": 12.55104121159167, + "grad_norm": 0.19366158545017242, + "learning_rate": 3.0448055283905587e-05, + "loss": 0.0196, + "step": 64650 + }, + { + "epoch": 12.55298286490947, + "grad_norm": 0.1836445927619934, + "learning_rate": 3.0432838244424494e-05, + "loss": 0.0222, + "step": 64660 + }, + { + "epoch": 12.55492451822727, + "grad_norm": 0.14083774387836456, + "learning_rate": 3.04176233447788e-05, + "loss": 0.0249, + "step": 64670 + }, + { + "epoch": 12.556866171545071, + "grad_norm": 0.2747848331928253, + "learning_rate": 3.040241058663238e-05, + "loss": 0.0257, + "step": 64680 + }, + { + "epoch": 12.558807824862871, + "grad_norm": 0.16614896059036255, + "learning_rate": 3.038719997164887e-05, + "loss": 0.0281, + "step": 64690 + }, + { + "epoch": 12.560749478180671, + "grad_norm": 0.24040408432483673, + "learning_rate": 3.0371991501491683e-05, + "loss": 0.0274, + "step": 64700 + }, + { + "epoch": 12.562691131498472, + "grad_norm": 0.208902046084404, + "learning_rate": 3.0356785177823975e-05, + "loss": 0.0193, + "step": 64710 + }, + { + "epoch": 12.564632784816272, + "grad_norm": 0.21419695019721985, + "learning_rate": 3.0341581002308726e-05, + "loss": 0.0204, + "step": 64720 + }, + { + "epoch": 12.56657443813407, + "grad_norm": 0.24735109508037567, + "learning_rate": 3.0326378976608627e-05, + "loss": 0.0302, + "step": 64730 + }, + { + "epoch": 12.56851609145187, + "grad_norm": 0.14005428552627563, + "learning_rate": 3.0311179102386146e-05, + "loss": 0.025, + "step": 64740 + }, + { + "epoch": 12.570457744769671, + "grad_norm": 0.17430002987384796, + "learning_rate": 3.0295981381303513e-05, + "loss": 0.0223, + "step": 64750 + }, + { + "epoch": 12.572399398087471, + "grad_norm": 0.1537391096353531, + "learning_rate": 3.0280785815022727e-05, + "loss": 0.0227, + "step": 64760 + }, + { + "epoch": 12.574341051405272, + "grad_norm": 0.2464696764945984, + "learning_rate": 3.0265592405205544e-05, + "loss": 0.0272, + "step": 64770 + }, + { + "epoch": 12.576282704723072, + "grad_norm": 0.18142420053482056, + "learning_rate": 3.0250401153513517e-05, + "loss": 0.0298, + "step": 64780 + }, + { + "epoch": 12.578224358040872, + "grad_norm": 0.13508115708827972, + "learning_rate": 3.0235212061607936e-05, + "loss": 0.0273, + "step": 64790 + }, + { + "epoch": 12.580166011358672, + "grad_norm": 0.2603246569633484, + "learning_rate": 3.0220025131149842e-05, + "loss": 0.027, + "step": 64800 + }, + { + "epoch": 12.582107664676473, + "grad_norm": 0.17557071149349213, + "learning_rate": 3.020484036380007e-05, + "loss": 0.0256, + "step": 64810 + }, + { + "epoch": 12.584049317994273, + "grad_norm": 0.17627008259296417, + "learning_rate": 3.0189657761219188e-05, + "loss": 0.0204, + "step": 64820 + }, + { + "epoch": 12.585990971312071, + "grad_norm": 0.17118147015571594, + "learning_rate": 3.017447732506753e-05, + "loss": 0.023, + "step": 64830 + }, + { + "epoch": 12.587932624629872, + "grad_norm": 0.20913037657737732, + "learning_rate": 3.015929905700524e-05, + "loss": 0.0222, + "step": 64840 + }, + { + "epoch": 12.589874277947672, + "grad_norm": 0.2723805010318756, + "learning_rate": 3.014412295869218e-05, + "loss": 0.0264, + "step": 64850 + }, + { + "epoch": 12.591815931265472, + "grad_norm": 0.2946603000164032, + "learning_rate": 3.0128949031787977e-05, + "loss": 0.029, + "step": 64860 + }, + { + "epoch": 12.593757584583273, + "grad_norm": 0.1448388546705246, + "learning_rate": 3.0113777277952022e-05, + "loss": 0.0218, + "step": 64870 + }, + { + "epoch": 12.595699237901073, + "grad_norm": 0.11546837538480759, + "learning_rate": 3.0098607698843496e-05, + "loss": 0.0264, + "step": 64880 + }, + { + "epoch": 12.597640891218873, + "grad_norm": 0.17167794704437256, + "learning_rate": 3.0083440296121312e-05, + "loss": 0.0263, + "step": 64890 + }, + { + "epoch": 12.599582544536673, + "grad_norm": 0.16321179270744324, + "learning_rate": 3.0068275071444142e-05, + "loss": 0.0222, + "step": 64900 + }, + { + "epoch": 12.601524197854474, + "grad_norm": 0.19260472059249878, + "learning_rate": 3.0053112026470442e-05, + "loss": 0.0238, + "step": 64910 + }, + { + "epoch": 12.603465851172274, + "grad_norm": 0.14619743824005127, + "learning_rate": 3.0037951162858434e-05, + "loss": 0.0224, + "step": 64920 + }, + { + "epoch": 12.605407504490074, + "grad_norm": 0.20594599843025208, + "learning_rate": 3.002279248226607e-05, + "loss": 0.0176, + "step": 64930 + }, + { + "epoch": 12.607349157807873, + "grad_norm": 0.22833313047885895, + "learning_rate": 3.000763598635109e-05, + "loss": 0.0226, + "step": 64940 + }, + { + "epoch": 12.609290811125673, + "grad_norm": 0.1114019900560379, + "learning_rate": 2.999248167677098e-05, + "loss": 0.0244, + "step": 64950 + }, + { + "epoch": 12.611232464443473, + "grad_norm": 0.24610954523086548, + "learning_rate": 2.9977329555182976e-05, + "loss": 0.0156, + "step": 64960 + }, + { + "epoch": 12.613174117761274, + "grad_norm": 0.14162348210811615, + "learning_rate": 2.9962179623244124e-05, + "loss": 0.0194, + "step": 64970 + }, + { + "epoch": 12.615115771079074, + "grad_norm": 0.1324779987335205, + "learning_rate": 2.9947031882611175e-05, + "loss": 0.0225, + "step": 64980 + }, + { + "epoch": 12.617057424396874, + "grad_norm": 0.2459503561258316, + "learning_rate": 2.9931886334940685e-05, + "loss": 0.0274, + "step": 64990 + }, + { + "epoch": 12.618999077714674, + "grad_norm": 0.21744997799396515, + "learning_rate": 2.9916742981888923e-05, + "loss": 0.027, + "step": 65000 + }, + { + "epoch": 12.620940731032475, + "grad_norm": 0.18467679619789124, + "learning_rate": 2.990160182511196e-05, + "loss": 0.025, + "step": 65010 + }, + { + "epoch": 12.622882384350275, + "grad_norm": 0.15900830924510956, + "learning_rate": 2.9886462866265574e-05, + "loss": 0.0226, + "step": 65020 + }, + { + "epoch": 12.624824037668075, + "grad_norm": 0.22186125814914703, + "learning_rate": 2.9871326107005392e-05, + "loss": 0.0254, + "step": 65030 + }, + { + "epoch": 12.626765690985874, + "grad_norm": 0.12660226225852966, + "learning_rate": 2.9856191548986735e-05, + "loss": 0.0306, + "step": 65040 + }, + { + "epoch": 12.628707344303674, + "grad_norm": 0.16126908361911774, + "learning_rate": 2.9841059193864672e-05, + "loss": 0.0187, + "step": 65050 + }, + { + "epoch": 12.630648997621474, + "grad_norm": 0.15567240118980408, + "learning_rate": 2.982592904329407e-05, + "loss": 0.0209, + "step": 65060 + }, + { + "epoch": 12.632590650939274, + "grad_norm": 0.1683681756258011, + "learning_rate": 2.981080109892954e-05, + "loss": 0.0257, + "step": 65070 + }, + { + "epoch": 12.634532304257075, + "grad_norm": 0.15687492489814758, + "learning_rate": 2.9795675362425434e-05, + "loss": 0.0256, + "step": 65080 + }, + { + "epoch": 12.636473957574875, + "grad_norm": 0.28322434425354004, + "learning_rate": 2.97805518354359e-05, + "loss": 0.0291, + "step": 65090 + }, + { + "epoch": 12.638415610892675, + "grad_norm": 0.20778414607048035, + "learning_rate": 2.9765430519614825e-05, + "loss": 0.0232, + "step": 65100 + }, + { + "epoch": 12.640357264210476, + "grad_norm": 0.1669037789106369, + "learning_rate": 2.975031141661584e-05, + "loss": 0.0263, + "step": 65110 + }, + { + "epoch": 12.642298917528276, + "grad_norm": 0.20340897142887115, + "learning_rate": 2.9735194528092347e-05, + "loss": 0.0241, + "step": 65120 + }, + { + "epoch": 12.644240570846076, + "grad_norm": 0.1171436607837677, + "learning_rate": 2.9720079855697525e-05, + "loss": 0.0229, + "step": 65130 + }, + { + "epoch": 12.646182224163876, + "grad_norm": 0.1373230367898941, + "learning_rate": 2.970496740108427e-05, + "loss": 0.0226, + "step": 65140 + }, + { + "epoch": 12.648123877481675, + "grad_norm": 0.10877729952335358, + "learning_rate": 2.968985716590527e-05, + "loss": 0.0272, + "step": 65150 + }, + { + "epoch": 12.650065530799475, + "grad_norm": 0.1789378523826599, + "learning_rate": 2.9674749151812953e-05, + "loss": 0.0364, + "step": 65160 + }, + { + "epoch": 12.652007184117275, + "grad_norm": 0.24315063655376434, + "learning_rate": 2.965964336045951e-05, + "loss": 0.029, + "step": 65170 + }, + { + "epoch": 12.653948837435076, + "grad_norm": 0.2097417563199997, + "learning_rate": 2.9644539793496894e-05, + "loss": 0.0251, + "step": 65180 + }, + { + "epoch": 12.655890490752876, + "grad_norm": 0.2543206214904785, + "learning_rate": 2.9629438452576807e-05, + "loss": 0.0221, + "step": 65190 + }, + { + "epoch": 12.657832144070676, + "grad_norm": 0.1545957624912262, + "learning_rate": 2.9614339339350698e-05, + "loss": 0.0318, + "step": 65200 + }, + { + "epoch": 12.659773797388477, + "grad_norm": 0.12419403344392776, + "learning_rate": 2.959924245546979e-05, + "loss": 0.0227, + "step": 65210 + }, + { + "epoch": 12.661715450706277, + "grad_norm": 0.2379770278930664, + "learning_rate": 2.9584147802585038e-05, + "loss": 0.0245, + "step": 65220 + }, + { + "epoch": 12.663657104024077, + "grad_norm": 0.17345625162124634, + "learning_rate": 2.956905538234721e-05, + "loss": 0.02, + "step": 65230 + }, + { + "epoch": 12.665598757341877, + "grad_norm": 0.15139353275299072, + "learning_rate": 2.9553965196406764e-05, + "loss": 0.0236, + "step": 65240 + }, + { + "epoch": 12.667540410659676, + "grad_norm": 0.32661330699920654, + "learning_rate": 2.9538877246413943e-05, + "loss": 0.0206, + "step": 65250 + }, + { + "epoch": 12.669482063977476, + "grad_norm": 0.23590444028377533, + "learning_rate": 2.952379153401874e-05, + "loss": 0.0217, + "step": 65260 + }, + { + "epoch": 12.671423717295276, + "grad_norm": 0.1366998851299286, + "learning_rate": 2.9508708060870914e-05, + "loss": 0.0201, + "step": 65270 + }, + { + "epoch": 12.673365370613077, + "grad_norm": 0.13201157748699188, + "learning_rate": 2.9493626828619948e-05, + "loss": 0.0243, + "step": 65280 + }, + { + "epoch": 12.675307023930877, + "grad_norm": 0.29744264483451843, + "learning_rate": 2.9478547838915137e-05, + "loss": 0.0278, + "step": 65290 + }, + { + "epoch": 12.677248677248677, + "grad_norm": 0.22842596471309662, + "learning_rate": 2.946347109340548e-05, + "loss": 0.0262, + "step": 65300 + }, + { + "epoch": 12.679190330566477, + "grad_norm": 0.14290490746498108, + "learning_rate": 2.944839659373976e-05, + "loss": 0.0269, + "step": 65310 + }, + { + "epoch": 12.681131983884278, + "grad_norm": 0.21571789681911469, + "learning_rate": 2.943332434156647e-05, + "loss": 0.0218, + "step": 65320 + }, + { + "epoch": 12.683073637202078, + "grad_norm": 0.134928360581398, + "learning_rate": 2.9418254338533912e-05, + "loss": 0.0185, + "step": 65330 + }, + { + "epoch": 12.685015290519878, + "grad_norm": 0.14709743857383728, + "learning_rate": 2.940318658629011e-05, + "loss": 0.0211, + "step": 65340 + }, + { + "epoch": 12.686956943837679, + "grad_norm": 0.23078404366970062, + "learning_rate": 2.938812108648287e-05, + "loss": 0.0271, + "step": 65350 + }, + { + "epoch": 12.688898597155477, + "grad_norm": 0.1814977377653122, + "learning_rate": 2.937305784075972e-05, + "loss": 0.0273, + "step": 65360 + }, + { + "epoch": 12.690840250473277, + "grad_norm": 0.23410604894161224, + "learning_rate": 2.9357996850767937e-05, + "loss": 0.0233, + "step": 65370 + }, + { + "epoch": 12.692781903791078, + "grad_norm": 0.16857440769672394, + "learning_rate": 2.9342938118154595e-05, + "loss": 0.0161, + "step": 65380 + }, + { + "epoch": 12.694723557108878, + "grad_norm": 0.148453950881958, + "learning_rate": 2.9327881644566485e-05, + "loss": 0.0228, + "step": 65390 + }, + { + "epoch": 12.696665210426678, + "grad_norm": 0.21906879544258118, + "learning_rate": 2.9312827431650157e-05, + "loss": 0.0246, + "step": 65400 + }, + { + "epoch": 12.698606863744478, + "grad_norm": 0.21339106559753418, + "learning_rate": 2.92977754810519e-05, + "loss": 0.0251, + "step": 65410 + }, + { + "epoch": 12.700548517062279, + "grad_norm": 0.14383219182491302, + "learning_rate": 2.928272579441781e-05, + "loss": 0.0268, + "step": 65420 + }, + { + "epoch": 12.702490170380079, + "grad_norm": 0.19620785117149353, + "learning_rate": 2.9267678373393676e-05, + "loss": 0.0299, + "step": 65430 + }, + { + "epoch": 12.70443182369788, + "grad_norm": 0.19200600683689117, + "learning_rate": 2.925263321962507e-05, + "loss": 0.0218, + "step": 65440 + }, + { + "epoch": 12.70637347701568, + "grad_norm": 0.17580071091651917, + "learning_rate": 2.9237590334757292e-05, + "loss": 0.0231, + "step": 65450 + }, + { + "epoch": 12.708315130333478, + "grad_norm": 0.19008181989192963, + "learning_rate": 2.922254972043543e-05, + "loss": 0.0201, + "step": 65460 + }, + { + "epoch": 12.710256783651278, + "grad_norm": 0.2029113620519638, + "learning_rate": 2.9207511378304265e-05, + "loss": 0.0222, + "step": 65470 + }, + { + "epoch": 12.712198436969079, + "grad_norm": 0.1007395088672638, + "learning_rate": 2.9192475310008428e-05, + "loss": 0.0212, + "step": 65480 + }, + { + "epoch": 12.714140090286879, + "grad_norm": 0.18204674124717712, + "learning_rate": 2.917744151719219e-05, + "loss": 0.0231, + "step": 65490 + }, + { + "epoch": 12.71608174360468, + "grad_norm": 0.13547584414482117, + "learning_rate": 2.916241000149964e-05, + "loss": 0.0246, + "step": 65500 + }, + { + "epoch": 12.71802339692248, + "grad_norm": 0.24544857442378998, + "learning_rate": 2.9147380764574622e-05, + "loss": 0.0272, + "step": 65510 + }, + { + "epoch": 12.71996505024028, + "grad_norm": 0.2264019399881363, + "learning_rate": 2.9132353808060674e-05, + "loss": 0.0218, + "step": 65520 + }, + { + "epoch": 12.72190670355808, + "grad_norm": 0.13903126120567322, + "learning_rate": 2.9117329133601134e-05, + "loss": 0.0214, + "step": 65530 + }, + { + "epoch": 12.72384835687588, + "grad_norm": 0.21490821242332458, + "learning_rate": 2.9102306742839093e-05, + "loss": 0.0229, + "step": 65540 + }, + { + "epoch": 12.72579001019368, + "grad_norm": 0.18766376376152039, + "learning_rate": 2.9087286637417375e-05, + "loss": 0.0303, + "step": 65550 + }, + { + "epoch": 12.72773166351148, + "grad_norm": 0.1973390132188797, + "learning_rate": 2.9072268818978537e-05, + "loss": 0.0247, + "step": 65560 + }, + { + "epoch": 12.729673316829281, + "grad_norm": 0.2516203820705414, + "learning_rate": 2.905725328916493e-05, + "loss": 0.0236, + "step": 65570 + }, + { + "epoch": 12.73161497014708, + "grad_norm": 0.15078796446323395, + "learning_rate": 2.9042240049618608e-05, + "loss": 0.0237, + "step": 65580 + }, + { + "epoch": 12.73355662346488, + "grad_norm": 0.14218801259994507, + "learning_rate": 2.9027229101981413e-05, + "loss": 0.0204, + "step": 65590 + }, + { + "epoch": 12.73549827678268, + "grad_norm": 0.12228502333164215, + "learning_rate": 2.901222044789488e-05, + "loss": 0.0225, + "step": 65600 + }, + { + "epoch": 12.73743993010048, + "grad_norm": 0.20458900928497314, + "learning_rate": 2.89972140890004e-05, + "loss": 0.0263, + "step": 65610 + }, + { + "epoch": 12.73938158341828, + "grad_norm": 0.20587950944900513, + "learning_rate": 2.8982210026938988e-05, + "loss": 0.0267, + "step": 65620 + }, + { + "epoch": 12.741323236736081, + "grad_norm": 0.21739180386066437, + "learning_rate": 2.896720826335151e-05, + "loss": 0.0236, + "step": 65630 + }, + { + "epoch": 12.743264890053881, + "grad_norm": 0.2497396320104599, + "learning_rate": 2.8952208799878493e-05, + "loss": 0.0191, + "step": 65640 + }, + { + "epoch": 12.745206543371681, + "grad_norm": 0.1943778544664383, + "learning_rate": 2.89372116381603e-05, + "loss": 0.0212, + "step": 65650 + }, + { + "epoch": 12.747148196689482, + "grad_norm": 0.26742690801620483, + "learning_rate": 2.8922216779836932e-05, + "loss": 0.0321, + "step": 65660 + }, + { + "epoch": 12.749089850007282, + "grad_norm": 0.2045307606458664, + "learning_rate": 2.8907224226548285e-05, + "loss": 0.0227, + "step": 65670 + }, + { + "epoch": 12.75103150332508, + "grad_norm": 0.2423778474330902, + "learning_rate": 2.8892233979933868e-05, + "loss": 0.0314, + "step": 65680 + }, + { + "epoch": 12.75297315664288, + "grad_norm": 0.10987994819879532, + "learning_rate": 2.8877246041633026e-05, + "loss": 0.0209, + "step": 65690 + }, + { + "epoch": 12.754914809960681, + "grad_norm": 0.1886376589536667, + "learning_rate": 2.886226041328478e-05, + "loss": 0.0273, + "step": 65700 + }, + { + "epoch": 12.756856463278481, + "grad_norm": 0.13109107315540314, + "learning_rate": 2.8847277096527957e-05, + "loss": 0.0246, + "step": 65710 + }, + { + "epoch": 12.758798116596282, + "grad_norm": 0.2852749228477478, + "learning_rate": 2.8832296093001105e-05, + "loss": 0.0209, + "step": 65720 + }, + { + "epoch": 12.760739769914082, + "grad_norm": 0.10708723217248917, + "learning_rate": 2.8817317404342547e-05, + "loss": 0.0202, + "step": 65730 + }, + { + "epoch": 12.762681423231882, + "grad_norm": 0.10154986381530762, + "learning_rate": 2.880234103219029e-05, + "loss": 0.0217, + "step": 65740 + }, + { + "epoch": 12.764623076549682, + "grad_norm": 0.1605578511953354, + "learning_rate": 2.878736697818215e-05, + "loss": 0.0252, + "step": 65750 + }, + { + "epoch": 12.766564729867483, + "grad_norm": 0.2146139144897461, + "learning_rate": 2.877239524395568e-05, + "loss": 0.0183, + "step": 65760 + }, + { + "epoch": 12.768506383185283, + "grad_norm": 0.25854483246803284, + "learning_rate": 2.8757425831148134e-05, + "loss": 0.0262, + "step": 65770 + }, + { + "epoch": 12.770448036503083, + "grad_norm": 0.20008394122123718, + "learning_rate": 2.874245874139658e-05, + "loss": 0.0197, + "step": 65780 + }, + { + "epoch": 12.772389689820882, + "grad_norm": 0.07424318790435791, + "learning_rate": 2.8727493976337737e-05, + "loss": 0.0245, + "step": 65790 + }, + { + "epoch": 12.774331343138682, + "grad_norm": 0.17650257050991058, + "learning_rate": 2.8712531537608218e-05, + "loss": 0.028, + "step": 65800 + }, + { + "epoch": 12.776272996456482, + "grad_norm": 0.1333223581314087, + "learning_rate": 2.8697571426844217e-05, + "loss": 0.0276, + "step": 65810 + }, + { + "epoch": 12.778214649774283, + "grad_norm": 0.12265574932098389, + "learning_rate": 2.86826136456818e-05, + "loss": 0.0206, + "step": 65820 + }, + { + "epoch": 12.780156303092083, + "grad_norm": 0.11293309181928635, + "learning_rate": 2.8667658195756685e-05, + "loss": 0.0236, + "step": 65830 + }, + { + "epoch": 12.782097956409883, + "grad_norm": 0.1117052435874939, + "learning_rate": 2.8652705078704412e-05, + "loss": 0.0176, + "step": 65840 + }, + { + "epoch": 12.784039609727683, + "grad_norm": 0.11600609868764877, + "learning_rate": 2.863775429616018e-05, + "loss": 0.0155, + "step": 65850 + }, + { + "epoch": 12.785981263045484, + "grad_norm": 0.2085554450750351, + "learning_rate": 2.862280584975906e-05, + "loss": 0.0226, + "step": 65860 + }, + { + "epoch": 12.787922916363284, + "grad_norm": 0.1767098307609558, + "learning_rate": 2.8607859741135735e-05, + "loss": 0.0209, + "step": 65870 + }, + { + "epoch": 12.789864569681084, + "grad_norm": 0.24172456562519073, + "learning_rate": 2.8592915971924723e-05, + "loss": 0.0205, + "step": 65880 + }, + { + "epoch": 12.791806222998883, + "grad_norm": 0.17588521540164948, + "learning_rate": 2.8577974543760212e-05, + "loss": 0.0259, + "step": 65890 + }, + { + "epoch": 12.793747876316683, + "grad_norm": 0.17745989561080933, + "learning_rate": 2.8563035458276222e-05, + "loss": 0.0214, + "step": 65900 + }, + { + "epoch": 12.795689529634483, + "grad_norm": 0.08338428288698196, + "learning_rate": 2.8548098717106424e-05, + "loss": 0.0293, + "step": 65910 + }, + { + "epoch": 12.797631182952284, + "grad_norm": 0.12580713629722595, + "learning_rate": 2.85331643218843e-05, + "loss": 0.0279, + "step": 65920 + }, + { + "epoch": 12.799572836270084, + "grad_norm": 0.17300474643707275, + "learning_rate": 2.8518232274243046e-05, + "loss": 0.0254, + "step": 65930 + }, + { + "epoch": 12.801514489587884, + "grad_norm": 0.22865605354309082, + "learning_rate": 2.8503302575815638e-05, + "loss": 0.029, + "step": 65940 + }, + { + "epoch": 12.803456142905684, + "grad_norm": 0.18337994813919067, + "learning_rate": 2.8488375228234716e-05, + "loss": 0.0225, + "step": 65950 + }, + { + "epoch": 12.805397796223485, + "grad_norm": 0.12754878401756287, + "learning_rate": 2.8473450233132738e-05, + "loss": 0.0218, + "step": 65960 + }, + { + "epoch": 12.807339449541285, + "grad_norm": 0.157621368765831, + "learning_rate": 2.8458527592141898e-05, + "loss": 0.0273, + "step": 65970 + }, + { + "epoch": 12.809281102859085, + "grad_norm": 0.22013680636882782, + "learning_rate": 2.8443607306894073e-05, + "loss": 0.0229, + "step": 65980 + }, + { + "epoch": 12.811222756176885, + "grad_norm": 0.1585218757390976, + "learning_rate": 2.8428689379020946e-05, + "loss": 0.0199, + "step": 65990 + }, + { + "epoch": 12.813164409494684, + "grad_norm": 0.18127641081809998, + "learning_rate": 2.841377381015391e-05, + "loss": 0.0169, + "step": 66000 + }, + { + "epoch": 12.815106062812484, + "grad_norm": 0.1667250096797943, + "learning_rate": 2.8398860601924144e-05, + "loss": 0.0278, + "step": 66010 + }, + { + "epoch": 12.817047716130284, + "grad_norm": 0.16819603741168976, + "learning_rate": 2.838394975596248e-05, + "loss": 0.0248, + "step": 66020 + }, + { + "epoch": 12.818989369448085, + "grad_norm": 0.18985901772975922, + "learning_rate": 2.83690412738996e-05, + "loss": 0.0212, + "step": 66030 + }, + { + "epoch": 12.820931022765885, + "grad_norm": 0.1231437623500824, + "learning_rate": 2.835413515736579e-05, + "loss": 0.0226, + "step": 66040 + }, + { + "epoch": 12.822872676083685, + "grad_norm": 0.19615884125232697, + "learning_rate": 2.8339231407991267e-05, + "loss": 0.0241, + "step": 66050 + }, + { + "epoch": 12.824814329401486, + "grad_norm": 0.17278103530406952, + "learning_rate": 2.832433002740581e-05, + "loss": 0.0251, + "step": 66060 + }, + { + "epoch": 12.826755982719286, + "grad_norm": 0.2511650621891022, + "learning_rate": 2.8309431017239047e-05, + "loss": 0.0256, + "step": 66070 + }, + { + "epoch": 12.828697636037086, + "grad_norm": 0.12402454018592834, + "learning_rate": 2.829453437912028e-05, + "loss": 0.024, + "step": 66080 + }, + { + "epoch": 12.830639289354886, + "grad_norm": 0.142817422747612, + "learning_rate": 2.8279640114678617e-05, + "loss": 0.022, + "step": 66090 + }, + { + "epoch": 12.832580942672685, + "grad_norm": 0.20124852657318115, + "learning_rate": 2.8264748225542836e-05, + "loss": 0.0217, + "step": 66100 + }, + { + "epoch": 12.834522595990485, + "grad_norm": 0.11849503964185715, + "learning_rate": 2.8249858713341516e-05, + "loss": 0.0263, + "step": 66110 + }, + { + "epoch": 12.836464249308285, + "grad_norm": 0.269143283367157, + "learning_rate": 2.823497157970294e-05, + "loss": 0.0185, + "step": 66120 + }, + { + "epoch": 12.838405902626086, + "grad_norm": 0.18576835095882416, + "learning_rate": 2.8220086826255164e-05, + "loss": 0.0244, + "step": 66130 + }, + { + "epoch": 12.840347555943886, + "grad_norm": 0.13158467411994934, + "learning_rate": 2.8205204454625934e-05, + "loss": 0.0205, + "step": 66140 + }, + { + "epoch": 12.842289209261686, + "grad_norm": 0.14827977120876312, + "learning_rate": 2.8190324466442798e-05, + "loss": 0.0198, + "step": 66150 + }, + { + "epoch": 12.844230862579487, + "grad_norm": 0.13660430908203125, + "learning_rate": 2.8175446863332965e-05, + "loss": 0.0223, + "step": 66160 + }, + { + "epoch": 12.846172515897287, + "grad_norm": 0.16149596869945526, + "learning_rate": 2.816057164692345e-05, + "loss": 0.0187, + "step": 66170 + }, + { + "epoch": 12.848114169215087, + "grad_norm": 0.12370273470878601, + "learning_rate": 2.8145698818840992e-05, + "loss": 0.0209, + "step": 66180 + }, + { + "epoch": 12.850055822532887, + "grad_norm": 0.15151822566986084, + "learning_rate": 2.8130828380712065e-05, + "loss": 0.023, + "step": 66190 + }, + { + "epoch": 12.851997475850688, + "grad_norm": 0.13902932405471802, + "learning_rate": 2.811596033416285e-05, + "loss": 0.0218, + "step": 66200 + }, + { + "epoch": 12.853939129168488, + "grad_norm": 0.10963690280914307, + "learning_rate": 2.8101094680819307e-05, + "loss": 0.0206, + "step": 66210 + }, + { + "epoch": 12.855880782486286, + "grad_norm": 0.17805235087871552, + "learning_rate": 2.808623142230715e-05, + "loss": 0.0214, + "step": 66220 + }, + { + "epoch": 12.857822435804087, + "grad_norm": 0.13471783697605133, + "learning_rate": 2.8071370560251757e-05, + "loss": 0.0285, + "step": 66230 + }, + { + "epoch": 12.859764089121887, + "grad_norm": 0.12997712194919586, + "learning_rate": 2.805651209627831e-05, + "loss": 0.0209, + "step": 66240 + }, + { + "epoch": 12.861705742439687, + "grad_norm": 0.13061532378196716, + "learning_rate": 2.804165603201171e-05, + "loss": 0.0233, + "step": 66250 + }, + { + "epoch": 12.863647395757488, + "grad_norm": 0.17599163949489594, + "learning_rate": 2.802680236907661e-05, + "loss": 0.0205, + "step": 66260 + }, + { + "epoch": 12.865589049075288, + "grad_norm": 0.16720522940158844, + "learning_rate": 2.8011951109097344e-05, + "loss": 0.0225, + "step": 66270 + }, + { + "epoch": 12.867530702393088, + "grad_norm": 0.1386972814798355, + "learning_rate": 2.7997102253698072e-05, + "loss": 0.0222, + "step": 66280 + }, + { + "epoch": 12.869472355710888, + "grad_norm": 0.23928774893283844, + "learning_rate": 2.7982255804502588e-05, + "loss": 0.0232, + "step": 66290 + }, + { + "epoch": 12.871414009028689, + "grad_norm": 0.19939640164375305, + "learning_rate": 2.7967411763134516e-05, + "loss": 0.031, + "step": 66300 + }, + { + "epoch": 12.873355662346487, + "grad_norm": 0.14439773559570312, + "learning_rate": 2.7952570131217166e-05, + "loss": 0.0237, + "step": 66310 + }, + { + "epoch": 12.875297315664287, + "grad_norm": 0.12631753087043762, + "learning_rate": 2.7937730910373618e-05, + "loss": 0.0226, + "step": 66320 + }, + { + "epoch": 12.877238968982088, + "grad_norm": 0.107735775411129, + "learning_rate": 2.7922894102226628e-05, + "loss": 0.0156, + "step": 66330 + }, + { + "epoch": 12.879180622299888, + "grad_norm": 0.14486263692378998, + "learning_rate": 2.790805970839876e-05, + "loss": 0.0246, + "step": 66340 + }, + { + "epoch": 12.881122275617688, + "grad_norm": 0.16524580121040344, + "learning_rate": 2.789322773051225e-05, + "loss": 0.0255, + "step": 66350 + }, + { + "epoch": 12.883063928935488, + "grad_norm": 0.25838425755500793, + "learning_rate": 2.7878398170189124e-05, + "loss": 0.0258, + "step": 66360 + }, + { + "epoch": 12.885005582253289, + "grad_norm": 0.15391862392425537, + "learning_rate": 2.7863571029051112e-05, + "loss": 0.031, + "step": 66370 + }, + { + "epoch": 12.886947235571089, + "grad_norm": 0.16610005497932434, + "learning_rate": 2.784874630871971e-05, + "loss": 0.0202, + "step": 66380 + }, + { + "epoch": 12.88888888888889, + "grad_norm": 0.22411666810512543, + "learning_rate": 2.7833924010816086e-05, + "loss": 0.0283, + "step": 66390 + }, + { + "epoch": 12.89083054220669, + "grad_norm": 0.170858234167099, + "learning_rate": 2.7819104136961226e-05, + "loss": 0.0338, + "step": 66400 + }, + { + "epoch": 12.89277219552449, + "grad_norm": 0.2782066762447357, + "learning_rate": 2.780428668877577e-05, + "loss": 0.0268, + "step": 66410 + }, + { + "epoch": 12.89471384884229, + "grad_norm": 0.16483399271965027, + "learning_rate": 2.7789471667880152e-05, + "loss": 0.0224, + "step": 66420 + }, + { + "epoch": 12.896655502160089, + "grad_norm": 0.09849468618631363, + "learning_rate": 2.7774659075894523e-05, + "loss": 0.0267, + "step": 66430 + }, + { + "epoch": 12.898597155477889, + "grad_norm": 0.0705692395567894, + "learning_rate": 2.7759848914438778e-05, + "loss": 0.0224, + "step": 66440 + }, + { + "epoch": 12.90053880879569, + "grad_norm": 0.10286679118871689, + "learning_rate": 2.7745041185132496e-05, + "loss": 0.0203, + "step": 66450 + }, + { + "epoch": 12.90248046211349, + "grad_norm": 0.2132958620786667, + "learning_rate": 2.773023588959506e-05, + "loss": 0.0285, + "step": 66460 + }, + { + "epoch": 12.90442211543129, + "grad_norm": 0.12410389631986618, + "learning_rate": 2.7715433029445555e-05, + "loss": 0.0232, + "step": 66470 + }, + { + "epoch": 12.90636376874909, + "grad_norm": 0.24933481216430664, + "learning_rate": 2.7700632606302766e-05, + "loss": 0.0244, + "step": 66480 + }, + { + "epoch": 12.90830542206689, + "grad_norm": 0.12304230779409409, + "learning_rate": 2.768583462178528e-05, + "loss": 0.0233, + "step": 66490 + }, + { + "epoch": 12.91024707538469, + "grad_norm": 0.25193244218826294, + "learning_rate": 2.767103907751136e-05, + "loss": 0.0271, + "step": 66500 + }, + { + "epoch": 12.91218872870249, + "grad_norm": 0.09389819949865341, + "learning_rate": 2.7656245975099065e-05, + "loss": 0.0193, + "step": 66510 + }, + { + "epoch": 12.914130382020291, + "grad_norm": 0.18161644041538239, + "learning_rate": 2.7641455316166093e-05, + "loss": 0.0276, + "step": 66520 + }, + { + "epoch": 12.91607203533809, + "grad_norm": 0.1309601217508316, + "learning_rate": 2.7626667102329973e-05, + "loss": 0.0217, + "step": 66530 + }, + { + "epoch": 12.91801368865589, + "grad_norm": 0.14387363195419312, + "learning_rate": 2.761188133520788e-05, + "loss": 0.0197, + "step": 66540 + }, + { + "epoch": 12.91995534197369, + "grad_norm": 0.19160138070583344, + "learning_rate": 2.7597098016416784e-05, + "loss": 0.0344, + "step": 66550 + }, + { + "epoch": 12.92189699529149, + "grad_norm": 0.16607710719108582, + "learning_rate": 2.7582317147573368e-05, + "loss": 0.027, + "step": 66560 + }, + { + "epoch": 12.92383864860929, + "grad_norm": 0.14147016406059265, + "learning_rate": 2.7567538730294064e-05, + "loss": 0.0216, + "step": 66570 + }, + { + "epoch": 12.925780301927091, + "grad_norm": 0.1510186791419983, + "learning_rate": 2.7552762766194973e-05, + "loss": 0.0215, + "step": 66580 + }, + { + "epoch": 12.927721955244891, + "grad_norm": 0.14012885093688965, + "learning_rate": 2.753798925689202e-05, + "loss": 0.0184, + "step": 66590 + }, + { + "epoch": 12.929663608562691, + "grad_norm": 0.2750057876110077, + "learning_rate": 2.7523218204000774e-05, + "loss": 0.0331, + "step": 66600 + }, + { + "epoch": 12.931605261880492, + "grad_norm": 0.34764623641967773, + "learning_rate": 2.7508449609136612e-05, + "loss": 0.021, + "step": 66610 + }, + { + "epoch": 12.933546915198292, + "grad_norm": 0.06156230717897415, + "learning_rate": 2.749368347391454e-05, + "loss": 0.0224, + "step": 66620 + }, + { + "epoch": 12.935488568516092, + "grad_norm": 0.09940534830093384, + "learning_rate": 2.7478919799949443e-05, + "loss": 0.0198, + "step": 66630 + }, + { + "epoch": 12.93743022183389, + "grad_norm": 0.1148456558585167, + "learning_rate": 2.7464158588855803e-05, + "loss": 0.0303, + "step": 66640 + }, + { + "epoch": 12.939371875151691, + "grad_norm": 0.14608345925807953, + "learning_rate": 2.7449399842247918e-05, + "loss": 0.0269, + "step": 66650 + }, + { + "epoch": 12.941313528469491, + "grad_norm": 0.17039352655410767, + "learning_rate": 2.7434643561739736e-05, + "loss": 0.024, + "step": 66660 + }, + { + "epoch": 12.943255181787292, + "grad_norm": 0.1690853238105774, + "learning_rate": 2.7419889748945016e-05, + "loss": 0.0268, + "step": 66670 + }, + { + "epoch": 12.945196835105092, + "grad_norm": 0.19329297542572021, + "learning_rate": 2.7405138405477203e-05, + "loss": 0.0273, + "step": 66680 + }, + { + "epoch": 12.947138488422892, + "grad_norm": 0.2542104423046112, + "learning_rate": 2.73903895329495e-05, + "loss": 0.0218, + "step": 66690 + }, + { + "epoch": 12.949080141740692, + "grad_norm": 0.16483314335346222, + "learning_rate": 2.737564313297478e-05, + "loss": 0.0233, + "step": 66700 + }, + { + "epoch": 12.951021795058493, + "grad_norm": 0.17870335280895233, + "learning_rate": 2.736089920716572e-05, + "loss": 0.0229, + "step": 66710 + }, + { + "epoch": 12.952963448376293, + "grad_norm": 0.1723444163799286, + "learning_rate": 2.7346157757134706e-05, + "loss": 0.0229, + "step": 66720 + }, + { + "epoch": 12.954905101694093, + "grad_norm": 0.2027643322944641, + "learning_rate": 2.733141878449379e-05, + "loss": 0.0297, + "step": 66730 + }, + { + "epoch": 12.956846755011892, + "grad_norm": 0.1918172538280487, + "learning_rate": 2.7316682290854844e-05, + "loss": 0.0253, + "step": 66740 + }, + { + "epoch": 12.958788408329692, + "grad_norm": 0.2616194784641266, + "learning_rate": 2.7301948277829414e-05, + "loss": 0.0343, + "step": 66750 + }, + { + "epoch": 12.960730061647492, + "grad_norm": 0.1431165635585785, + "learning_rate": 2.7287216747028814e-05, + "loss": 0.0287, + "step": 66760 + }, + { + "epoch": 12.962671714965293, + "grad_norm": 0.13121873140335083, + "learning_rate": 2.7272487700064025e-05, + "loss": 0.0235, + "step": 66770 + }, + { + "epoch": 12.964613368283093, + "grad_norm": 0.15852972865104675, + "learning_rate": 2.725776113854583e-05, + "loss": 0.0254, + "step": 66780 + }, + { + "epoch": 12.966555021600893, + "grad_norm": 0.1755647361278534, + "learning_rate": 2.7243037064084664e-05, + "loss": 0.0214, + "step": 66790 + }, + { + "epoch": 12.968496674918693, + "grad_norm": 0.2545931935310364, + "learning_rate": 2.7228315478290766e-05, + "loss": 0.0241, + "step": 66800 + }, + { + "epoch": 12.970438328236494, + "grad_norm": 0.1618879735469818, + "learning_rate": 2.721359638277402e-05, + "loss": 0.0317, + "step": 66810 + }, + { + "epoch": 12.972379981554294, + "grad_norm": 0.07544434815645218, + "learning_rate": 2.7198879779144148e-05, + "loss": 0.0191, + "step": 66820 + }, + { + "epoch": 12.974321634872094, + "grad_norm": 0.14500685036182404, + "learning_rate": 2.7184165669010485e-05, + "loss": 0.0235, + "step": 66830 + }, + { + "epoch": 12.976263288189894, + "grad_norm": 0.27478912472724915, + "learning_rate": 2.716945405398218e-05, + "loss": 0.0197, + "step": 66840 + }, + { + "epoch": 12.978204941507693, + "grad_norm": 0.2247217893600464, + "learning_rate": 2.715474493566803e-05, + "loss": 0.0233, + "step": 66850 + }, + { + "epoch": 12.980146594825493, + "grad_norm": 0.21573805809020996, + "learning_rate": 2.7140038315676652e-05, + "loss": 0.0232, + "step": 66860 + }, + { + "epoch": 12.982088248143294, + "grad_norm": 0.10698512196540833, + "learning_rate": 2.7125334195616275e-05, + "loss": 0.0202, + "step": 66870 + }, + { + "epoch": 12.984029901461094, + "grad_norm": 0.14306320250034332, + "learning_rate": 2.7110632577094997e-05, + "loss": 0.0227, + "step": 66880 + }, + { + "epoch": 12.985971554778894, + "grad_norm": 0.19433648884296417, + "learning_rate": 2.709593346172051e-05, + "loss": 0.0245, + "step": 66890 + }, + { + "epoch": 12.987913208096694, + "grad_norm": 0.14128433167934418, + "learning_rate": 2.708123685110032e-05, + "loss": 0.019, + "step": 66900 + }, + { + "epoch": 12.989854861414495, + "grad_norm": 0.1447797566652298, + "learning_rate": 2.7066542746841588e-05, + "loss": 0.0238, + "step": 66910 + }, + { + "epoch": 12.991796514732295, + "grad_norm": 0.1792125403881073, + "learning_rate": 2.7051851150551267e-05, + "loss": 0.0226, + "step": 66920 + }, + { + "epoch": 12.993738168050095, + "grad_norm": 0.15211601555347443, + "learning_rate": 2.7037162063835996e-05, + "loss": 0.0251, + "step": 66930 + }, + { + "epoch": 12.995679821367895, + "grad_norm": 0.22120346128940582, + "learning_rate": 2.702247548830218e-05, + "loss": 0.025, + "step": 66940 + }, + { + "epoch": 12.997621474685694, + "grad_norm": 0.19635355472564697, + "learning_rate": 2.7007791425555883e-05, + "loss": 0.0207, + "step": 66950 + }, + { + "epoch": 12.999563128003494, + "grad_norm": 0.16582892835140228, + "learning_rate": 2.6993109877202945e-05, + "loss": 0.0215, + "step": 66960 + }, + { + "epoch": 13.00135915732246, + "grad_norm": 0.21285022795200348, + "learning_rate": 2.6978430844848935e-05, + "loss": 0.0203, + "step": 66970 + }, + { + "epoch": 13.00330081064026, + "grad_norm": 0.16722224652767181, + "learning_rate": 2.69637543300991e-05, + "loss": 0.0282, + "step": 66980 + }, + { + "epoch": 13.00524246395806, + "grad_norm": 0.12664802372455597, + "learning_rate": 2.694908033455848e-05, + "loss": 0.0257, + "step": 66990 + }, + { + "epoch": 13.00718411727586, + "grad_norm": 0.21778342127799988, + "learning_rate": 2.6934408859831733e-05, + "loss": 0.0266, + "step": 67000 + }, + { + "epoch": 13.009125770593661, + "grad_norm": 0.15460295975208282, + "learning_rate": 2.69197399075234e-05, + "loss": 0.022, + "step": 67010 + }, + { + "epoch": 13.011067423911461, + "grad_norm": 0.2149609476327896, + "learning_rate": 2.6905073479237584e-05, + "loss": 0.0298, + "step": 67020 + }, + { + "epoch": 13.013009077229261, + "grad_norm": 0.1960848867893219, + "learning_rate": 2.6890409576578235e-05, + "loss": 0.0286, + "step": 67030 + }, + { + "epoch": 13.01495073054706, + "grad_norm": 0.11736689507961273, + "learning_rate": 2.687574820114892e-05, + "loss": 0.0227, + "step": 67040 + }, + { + "epoch": 13.01689238386486, + "grad_norm": 0.20746921002864838, + "learning_rate": 2.686108935455305e-05, + "loss": 0.0189, + "step": 67050 + }, + { + "epoch": 13.01883403718266, + "grad_norm": 0.16668567061424255, + "learning_rate": 2.6846433038393603e-05, + "loss": 0.0257, + "step": 67060 + }, + { + "epoch": 13.02077569050046, + "grad_norm": 0.1783992201089859, + "learning_rate": 2.6831779254273482e-05, + "loss": 0.0208, + "step": 67070 + }, + { + "epoch": 13.022717343818261, + "grad_norm": 0.23620182275772095, + "learning_rate": 2.6817128003795122e-05, + "loss": 0.0221, + "step": 67080 + }, + { + "epoch": 13.024658997136061, + "grad_norm": 0.20253592729568481, + "learning_rate": 2.680247928856081e-05, + "loss": 0.0277, + "step": 67090 + }, + { + "epoch": 13.026600650453862, + "grad_norm": 0.12900254130363464, + "learning_rate": 2.6787833110172466e-05, + "loss": 0.0215, + "step": 67100 + }, + { + "epoch": 13.028542303771662, + "grad_norm": 0.10704639554023743, + "learning_rate": 2.6773189470231807e-05, + "loss": 0.0225, + "step": 67110 + }, + { + "epoch": 13.030483957089462, + "grad_norm": 0.19864866137504578, + "learning_rate": 2.675854837034022e-05, + "loss": 0.0219, + "step": 67120 + }, + { + "epoch": 13.032425610407262, + "grad_norm": 0.16225799918174744, + "learning_rate": 2.6743909812098833e-05, + "loss": 0.0217, + "step": 67130 + }, + { + "epoch": 13.034367263725063, + "grad_norm": 0.19743938744068146, + "learning_rate": 2.67292737971085e-05, + "loss": 0.018, + "step": 67140 + }, + { + "epoch": 13.036308917042861, + "grad_norm": 0.1146761029958725, + "learning_rate": 2.671464032696982e-05, + "loss": 0.0206, + "step": 67150 + }, + { + "epoch": 13.038250570360661, + "grad_norm": 0.22728274762630463, + "learning_rate": 2.6700009403283055e-05, + "loss": 0.0211, + "step": 67160 + }, + { + "epoch": 13.040192223678462, + "grad_norm": 0.13123086094856262, + "learning_rate": 2.6685381027648214e-05, + "loss": 0.0234, + "step": 67170 + }, + { + "epoch": 13.042133876996262, + "grad_norm": 0.22735968232154846, + "learning_rate": 2.6670755201665086e-05, + "loss": 0.0184, + "step": 67180 + }, + { + "epoch": 13.044075530314062, + "grad_norm": 0.17917414009571075, + "learning_rate": 2.6656131926933066e-05, + "loss": 0.0242, + "step": 67190 + }, + { + "epoch": 13.046017183631863, + "grad_norm": 0.23059962689876556, + "learning_rate": 2.664151120505136e-05, + "loss": 0.0291, + "step": 67200 + }, + { + "epoch": 13.047958836949663, + "grad_norm": 0.3132423758506775, + "learning_rate": 2.6626893037618873e-05, + "loss": 0.0246, + "step": 67210 + }, + { + "epoch": 13.049900490267463, + "grad_norm": 0.11611654609441757, + "learning_rate": 2.6612277426234243e-05, + "loss": 0.0202, + "step": 67220 + }, + { + "epoch": 13.051842143585263, + "grad_norm": 0.233942449092865, + "learning_rate": 2.659766437249577e-05, + "loss": 0.021, + "step": 67230 + }, + { + "epoch": 13.053783796903064, + "grad_norm": 0.21905332803726196, + "learning_rate": 2.6583053878001557e-05, + "loss": 0.0213, + "step": 67240 + }, + { + "epoch": 13.055725450220864, + "grad_norm": 0.19117194414138794, + "learning_rate": 2.6568445944349323e-05, + "loss": 0.0285, + "step": 67250 + }, + { + "epoch": 13.057667103538662, + "grad_norm": 0.14561410248279572, + "learning_rate": 2.6553840573136657e-05, + "loss": 0.0269, + "step": 67260 + }, + { + "epoch": 13.059608756856463, + "grad_norm": 0.07806218415498734, + "learning_rate": 2.6539237765960706e-05, + "loss": 0.0214, + "step": 67270 + }, + { + "epoch": 13.061550410174263, + "grad_norm": 0.1024571880698204, + "learning_rate": 2.6524637524418465e-05, + "loss": 0.0223, + "step": 67280 + }, + { + "epoch": 13.063492063492063, + "grad_norm": 0.06893007457256317, + "learning_rate": 2.651003985010655e-05, + "loss": 0.0178, + "step": 67290 + }, + { + "epoch": 13.065433716809864, + "grad_norm": 0.18015296757221222, + "learning_rate": 2.649544474462138e-05, + "loss": 0.0279, + "step": 67300 + }, + { + "epoch": 13.067375370127664, + "grad_norm": 0.1635616421699524, + "learning_rate": 2.6480852209559015e-05, + "loss": 0.03, + "step": 67310 + }, + { + "epoch": 13.069317023445464, + "grad_norm": 0.12841005623340607, + "learning_rate": 2.6466262246515296e-05, + "loss": 0.0256, + "step": 67320 + }, + { + "epoch": 13.071258676763264, + "grad_norm": 0.1500343531370163, + "learning_rate": 2.6451674857085746e-05, + "loss": 0.0217, + "step": 67330 + }, + { + "epoch": 13.073200330081065, + "grad_norm": 0.09018564224243164, + "learning_rate": 2.6437090042865655e-05, + "loss": 0.027, + "step": 67340 + }, + { + "epoch": 13.075141983398865, + "grad_norm": 0.18594719469547272, + "learning_rate": 2.6422507805449947e-05, + "loss": 0.0217, + "step": 67350 + }, + { + "epoch": 13.077083636716663, + "grad_norm": 0.20128437876701355, + "learning_rate": 2.6407928146433357e-05, + "loss": 0.0213, + "step": 67360 + }, + { + "epoch": 13.079025290034464, + "grad_norm": 0.12192277610301971, + "learning_rate": 2.639335106741026e-05, + "loss": 0.0188, + "step": 67370 + }, + { + "epoch": 13.080966943352264, + "grad_norm": 0.13669805228710175, + "learning_rate": 2.6378776569974796e-05, + "loss": 0.0227, + "step": 67380 + }, + { + "epoch": 13.082908596670064, + "grad_norm": 0.17003601789474487, + "learning_rate": 2.6364204655720814e-05, + "loss": 0.0296, + "step": 67390 + }, + { + "epoch": 13.084850249987865, + "grad_norm": 0.3626402020454407, + "learning_rate": 2.6349635326241894e-05, + "loss": 0.0235, + "step": 67400 + }, + { + "epoch": 13.086791903305665, + "grad_norm": 0.13006770610809326, + "learning_rate": 2.633506858313127e-05, + "loss": 0.0233, + "step": 67410 + }, + { + "epoch": 13.088733556623465, + "grad_norm": 0.13173674046993256, + "learning_rate": 2.6320504427981975e-05, + "loss": 0.0243, + "step": 67420 + }, + { + "epoch": 13.090675209941265, + "grad_norm": 0.1815057098865509, + "learning_rate": 2.6305942862386734e-05, + "loss": 0.0228, + "step": 67430 + }, + { + "epoch": 13.092616863259066, + "grad_norm": 0.17010332643985748, + "learning_rate": 2.629138388793793e-05, + "loss": 0.0191, + "step": 67440 + }, + { + "epoch": 13.094558516576866, + "grad_norm": 0.19402343034744263, + "learning_rate": 2.6276827506227743e-05, + "loss": 0.0243, + "step": 67450 + }, + { + "epoch": 13.096500169894666, + "grad_norm": 0.08050544559955597, + "learning_rate": 2.6262273718848037e-05, + "loss": 0.0248, + "step": 67460 + }, + { + "epoch": 13.098441823212465, + "grad_norm": 0.16884289681911469, + "learning_rate": 2.6247722527390407e-05, + "loss": 0.017, + "step": 67470 + }, + { + "epoch": 13.100383476530265, + "grad_norm": 0.20670819282531738, + "learning_rate": 2.6233173933446104e-05, + "loss": 0.0265, + "step": 67480 + }, + { + "epoch": 13.102325129848065, + "grad_norm": 0.15972577035427094, + "learning_rate": 2.6218627938606188e-05, + "loss": 0.0286, + "step": 67490 + }, + { + "epoch": 13.104266783165865, + "grad_norm": 0.24423210322856903, + "learning_rate": 2.620408454446135e-05, + "loss": 0.0177, + "step": 67500 + }, + { + "epoch": 13.106208436483666, + "grad_norm": 0.15794947743415833, + "learning_rate": 2.6189543752602046e-05, + "loss": 0.0213, + "step": 67510 + }, + { + "epoch": 13.108150089801466, + "grad_norm": 0.31429165601730347, + "learning_rate": 2.6175005564618444e-05, + "loss": 0.0219, + "step": 67520 + }, + { + "epoch": 13.110091743119266, + "grad_norm": 0.20317308604717255, + "learning_rate": 2.6160469982100428e-05, + "loss": 0.0262, + "step": 67530 + }, + { + "epoch": 13.112033396437067, + "grad_norm": 0.09418261796236038, + "learning_rate": 2.6145937006637555e-05, + "loss": 0.0219, + "step": 67540 + }, + { + "epoch": 13.113975049754867, + "grad_norm": 0.09461037814617157, + "learning_rate": 2.6131406639819166e-05, + "loss": 0.0242, + "step": 67550 + }, + { + "epoch": 13.115916703072667, + "grad_norm": 0.13622340559959412, + "learning_rate": 2.611687888323424e-05, + "loss": 0.0266, + "step": 67560 + }, + { + "epoch": 13.117858356390466, + "grad_norm": 0.14504532516002655, + "learning_rate": 2.610235373847153e-05, + "loss": 0.0223, + "step": 67570 + }, + { + "epoch": 13.119800009708266, + "grad_norm": 0.1529482901096344, + "learning_rate": 2.6087831207119494e-05, + "loss": 0.0224, + "step": 67580 + }, + { + "epoch": 13.121741663026066, + "grad_norm": 0.12369491159915924, + "learning_rate": 2.60733112907663e-05, + "loss": 0.0185, + "step": 67590 + }, + { + "epoch": 13.123683316343866, + "grad_norm": 0.10691945999860764, + "learning_rate": 2.6058793990999786e-05, + "loss": 0.0248, + "step": 67600 + }, + { + "epoch": 13.125624969661667, + "grad_norm": 0.20264185965061188, + "learning_rate": 2.6044279309407593e-05, + "loss": 0.0268, + "step": 67610 + }, + { + "epoch": 13.127566622979467, + "grad_norm": 0.2296820431947708, + "learning_rate": 2.6029767247576976e-05, + "loss": 0.0206, + "step": 67620 + }, + { + "epoch": 13.129508276297267, + "grad_norm": 0.1861986219882965, + "learning_rate": 2.601525780709497e-05, + "loss": 0.0212, + "step": 67630 + }, + { + "epoch": 13.131449929615068, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6000750989548318e-05, + "loss": 0.0275, + "step": 67640 + }, + { + "epoch": 13.133391582932868, + "grad_norm": 0.11544580012559891, + "learning_rate": 2.5986246796523462e-05, + "loss": 0.018, + "step": 67650 + }, + { + "epoch": 13.135333236250668, + "grad_norm": 0.14951485395431519, + "learning_rate": 2.5971745229606546e-05, + "loss": 0.0213, + "step": 67660 + }, + { + "epoch": 13.137274889568468, + "grad_norm": 0.13272379338741302, + "learning_rate": 2.5957246290383442e-05, + "loss": 0.0164, + "step": 67670 + }, + { + "epoch": 13.139216542886267, + "grad_norm": 0.24534913897514343, + "learning_rate": 2.594274998043975e-05, + "loss": 0.0297, + "step": 67680 + }, + { + "epoch": 13.141158196204067, + "grad_norm": 0.14085040986537933, + "learning_rate": 2.5928256301360733e-05, + "loss": 0.0189, + "step": 67690 + }, + { + "epoch": 13.143099849521867, + "grad_norm": 0.2525012791156769, + "learning_rate": 2.591376525473142e-05, + "loss": 0.0207, + "step": 67700 + }, + { + "epoch": 13.145041502839668, + "grad_norm": 0.10015721619129181, + "learning_rate": 2.5899276842136523e-05, + "loss": 0.0214, + "step": 67710 + }, + { + "epoch": 13.146983156157468, + "grad_norm": 0.08435633778572083, + "learning_rate": 2.5884791065160495e-05, + "loss": 0.0194, + "step": 67720 + }, + { + "epoch": 13.148924809475268, + "grad_norm": 0.21842563152313232, + "learning_rate": 2.5870307925387436e-05, + "loss": 0.0216, + "step": 67730 + }, + { + "epoch": 13.150866462793068, + "grad_norm": 0.2380625605583191, + "learning_rate": 2.585582742440125e-05, + "loss": 0.0237, + "step": 67740 + }, + { + "epoch": 13.152808116110869, + "grad_norm": 0.14660829305648804, + "learning_rate": 2.5841349563785455e-05, + "loss": 0.0169, + "step": 67750 + }, + { + "epoch": 13.154749769428669, + "grad_norm": 0.14055182039737701, + "learning_rate": 2.5826874345123355e-05, + "loss": 0.0158, + "step": 67760 + }, + { + "epoch": 13.15669142274647, + "grad_norm": 0.17404529452323914, + "learning_rate": 2.581240176999792e-05, + "loss": 0.0253, + "step": 67770 + }, + { + "epoch": 13.15863307606427, + "grad_norm": 0.2224411815404892, + "learning_rate": 2.579793183999189e-05, + "loss": 0.0217, + "step": 67780 + }, + { + "epoch": 13.160574729382068, + "grad_norm": 0.19892069697380066, + "learning_rate": 2.5783464556687627e-05, + "loss": 0.0209, + "step": 67790 + }, + { + "epoch": 13.162516382699868, + "grad_norm": 0.20761631429195404, + "learning_rate": 2.576899992166729e-05, + "loss": 0.0227, + "step": 67800 + }, + { + "epoch": 13.164458036017669, + "grad_norm": 0.17011494934558868, + "learning_rate": 2.5754537936512668e-05, + "loss": 0.0316, + "step": 67810 + }, + { + "epoch": 13.166399689335469, + "grad_norm": 0.13002322614192963, + "learning_rate": 2.574007860280535e-05, + "loss": 0.0193, + "step": 67820 + }, + { + "epoch": 13.16834134265327, + "grad_norm": 0.16933566331863403, + "learning_rate": 2.5725621922126518e-05, + "loss": 0.0249, + "step": 67830 + }, + { + "epoch": 13.17028299597107, + "grad_norm": 0.24322062730789185, + "learning_rate": 2.571116789605722e-05, + "loss": 0.0224, + "step": 67840 + }, + { + "epoch": 13.17222464928887, + "grad_norm": 0.08197551220655441, + "learning_rate": 2.5696716526178054e-05, + "loss": 0.0261, + "step": 67850 + }, + { + "epoch": 13.17416630260667, + "grad_norm": 0.1357717514038086, + "learning_rate": 2.568226781406946e-05, + "loss": 0.023, + "step": 67860 + }, + { + "epoch": 13.17610795592447, + "grad_norm": 0.2287503182888031, + "learning_rate": 2.5667821761311474e-05, + "loss": 0.0196, + "step": 67870 + }, + { + "epoch": 13.17804960924227, + "grad_norm": 0.23530130088329315, + "learning_rate": 2.5653378369483917e-05, + "loss": 0.0205, + "step": 67880 + }, + { + "epoch": 13.179991262560069, + "grad_norm": 0.2312161773443222, + "learning_rate": 2.5638937640166292e-05, + "loss": 0.0238, + "step": 67890 + }, + { + "epoch": 13.18193291587787, + "grad_norm": 0.14585408568382263, + "learning_rate": 2.5624499574937842e-05, + "loss": 0.0265, + "step": 67900 + }, + { + "epoch": 13.18387456919567, + "grad_norm": 0.12375572323799133, + "learning_rate": 2.5610064175377456e-05, + "loss": 0.0245, + "step": 67910 + }, + { + "epoch": 13.18581622251347, + "grad_norm": 0.2285776287317276, + "learning_rate": 2.5595631443063782e-05, + "loss": 0.0211, + "step": 67920 + }, + { + "epoch": 13.18775787583127, + "grad_norm": 0.24861231446266174, + "learning_rate": 2.5581201379575177e-05, + "loss": 0.0226, + "step": 67930 + }, + { + "epoch": 13.18969952914907, + "grad_norm": 0.22021667659282684, + "learning_rate": 2.5566773986489655e-05, + "loss": 0.032, + "step": 67940 + }, + { + "epoch": 13.19164118246687, + "grad_norm": 0.15854954719543457, + "learning_rate": 2.5552349265385e-05, + "loss": 0.0226, + "step": 67950 + }, + { + "epoch": 13.193582835784671, + "grad_norm": 0.152474045753479, + "learning_rate": 2.5537927217838675e-05, + "loss": 0.0187, + "step": 67960 + }, + { + "epoch": 13.195524489102471, + "grad_norm": 0.08981902152299881, + "learning_rate": 2.5523507845427864e-05, + "loss": 0.0233, + "step": 67970 + }, + { + "epoch": 13.197466142420271, + "grad_norm": 0.14259862899780273, + "learning_rate": 2.5509091149729413e-05, + "loss": 0.019, + "step": 67980 + }, + { + "epoch": 13.199407795738072, + "grad_norm": 0.11840429902076721, + "learning_rate": 2.5494677132319962e-05, + "loss": 0.0268, + "step": 67990 + }, + { + "epoch": 13.20134944905587, + "grad_norm": 0.1369611769914627, + "learning_rate": 2.548026579477575e-05, + "loss": 0.0242, + "step": 68000 + }, + { + "epoch": 13.20329110237367, + "grad_norm": 0.19921022653579712, + "learning_rate": 2.5465857138672834e-05, + "loss": 0.0294, + "step": 68010 + }, + { + "epoch": 13.20523275569147, + "grad_norm": 0.30207690596580505, + "learning_rate": 2.5451451165586848e-05, + "loss": 0.0293, + "step": 68020 + }, + { + "epoch": 13.207174409009271, + "grad_norm": 0.1503565013408661, + "learning_rate": 2.543704787709329e-05, + "loss": 0.0194, + "step": 68030 + }, + { + "epoch": 13.209116062327071, + "grad_norm": 0.15439631044864655, + "learning_rate": 2.542264727476723e-05, + "loss": 0.021, + "step": 68040 + }, + { + "epoch": 13.211057715644872, + "grad_norm": 0.15105865895748138, + "learning_rate": 2.5408249360183535e-05, + "loss": 0.0305, + "step": 68050 + }, + { + "epoch": 13.212999368962672, + "grad_norm": 0.178850457072258, + "learning_rate": 2.5393854134916694e-05, + "loss": 0.0323, + "step": 68060 + }, + { + "epoch": 13.214941022280472, + "grad_norm": 0.12227541953325272, + "learning_rate": 2.5379461600540993e-05, + "loss": 0.0229, + "step": 68070 + }, + { + "epoch": 13.216882675598272, + "grad_norm": 0.0982542559504509, + "learning_rate": 2.5365071758630317e-05, + "loss": 0.015, + "step": 68080 + }, + { + "epoch": 13.218824328916073, + "grad_norm": 0.1432112753391266, + "learning_rate": 2.535068461075839e-05, + "loss": 0.019, + "step": 68090 + }, + { + "epoch": 13.220765982233873, + "grad_norm": 0.1543458253145218, + "learning_rate": 2.5336300158498516e-05, + "loss": 0.0298, + "step": 68100 + }, + { + "epoch": 13.222707635551672, + "grad_norm": 0.1377125084400177, + "learning_rate": 2.5321918403423793e-05, + "loss": 0.021, + "step": 68110 + }, + { + "epoch": 13.224649288869472, + "grad_norm": 0.13884417712688446, + "learning_rate": 2.5307539347106957e-05, + "loss": 0.0194, + "step": 68120 + }, + { + "epoch": 13.226590942187272, + "grad_norm": 0.11207544803619385, + "learning_rate": 2.529316299112049e-05, + "loss": 0.023, + "step": 68130 + }, + { + "epoch": 13.228532595505072, + "grad_norm": 0.13441963493824005, + "learning_rate": 2.5278789337036567e-05, + "loss": 0.0265, + "step": 68140 + }, + { + "epoch": 13.230474248822873, + "grad_norm": 0.22627750039100647, + "learning_rate": 2.52644183864271e-05, + "loss": 0.0164, + "step": 68150 + }, + { + "epoch": 13.232415902140673, + "grad_norm": 0.11958245187997818, + "learning_rate": 2.525005014086363e-05, + "loss": 0.0241, + "step": 68160 + }, + { + "epoch": 13.234357555458473, + "grad_norm": 0.1491071879863739, + "learning_rate": 2.5235684601917465e-05, + "loss": 0.0191, + "step": 68170 + }, + { + "epoch": 13.236299208776273, + "grad_norm": 0.2334280163049698, + "learning_rate": 2.5221321771159618e-05, + "loss": 0.0222, + "step": 68180 + }, + { + "epoch": 13.238240862094074, + "grad_norm": 0.2130013108253479, + "learning_rate": 2.5206961650160743e-05, + "loss": 0.0254, + "step": 68190 + }, + { + "epoch": 13.240182515411874, + "grad_norm": 0.23212824761867523, + "learning_rate": 2.519260424049129e-05, + "loss": 0.0176, + "step": 68200 + }, + { + "epoch": 13.242124168729672, + "grad_norm": 0.2567809820175171, + "learning_rate": 2.5178249543721304e-05, + "loss": 0.0243, + "step": 68210 + }, + { + "epoch": 13.244065822047473, + "grad_norm": 0.1785610020160675, + "learning_rate": 2.5163897561420668e-05, + "loss": 0.0274, + "step": 68220 + }, + { + "epoch": 13.246007475365273, + "grad_norm": 0.27397602796554565, + "learning_rate": 2.514954829515883e-05, + "loss": 0.0207, + "step": 68230 + }, + { + "epoch": 13.247949128683073, + "grad_norm": 0.09127113968133926, + "learning_rate": 2.513520174650505e-05, + "loss": 0.03, + "step": 68240 + }, + { + "epoch": 13.249890782000874, + "grad_norm": 0.22772115468978882, + "learning_rate": 2.5120857917028207e-05, + "loss": 0.0249, + "step": 68250 + }, + { + "epoch": 13.251832435318674, + "grad_norm": 0.14188063144683838, + "learning_rate": 2.5106516808296953e-05, + "loss": 0.0177, + "step": 68260 + }, + { + "epoch": 13.253774088636474, + "grad_norm": 0.14189697802066803, + "learning_rate": 2.5092178421879552e-05, + "loss": 0.0183, + "step": 68270 + }, + { + "epoch": 13.255715741954274, + "grad_norm": 0.19932159781455994, + "learning_rate": 2.5077842759344116e-05, + "loss": 0.0177, + "step": 68280 + }, + { + "epoch": 13.257657395272075, + "grad_norm": 0.37837889790534973, + "learning_rate": 2.5063509822258304e-05, + "loss": 0.027, + "step": 68290 + }, + { + "epoch": 13.259599048589875, + "grad_norm": 0.17045438289642334, + "learning_rate": 2.5049179612189576e-05, + "loss": 0.0256, + "step": 68300 + }, + { + "epoch": 13.261540701907675, + "grad_norm": 0.18833020329475403, + "learning_rate": 2.5034852130705043e-05, + "loss": 0.0301, + "step": 68310 + }, + { + "epoch": 13.263482355225474, + "grad_norm": 0.2744807302951813, + "learning_rate": 2.5020527379371572e-05, + "loss": 0.0197, + "step": 68320 + }, + { + "epoch": 13.265424008543274, + "grad_norm": 0.2065567523241043, + "learning_rate": 2.500620535975564e-05, + "loss": 0.0212, + "step": 68330 + }, + { + "epoch": 13.267365661861074, + "grad_norm": 0.1696881800889969, + "learning_rate": 2.4991886073423515e-05, + "loss": 0.0248, + "step": 68340 + }, + { + "epoch": 13.269307315178875, + "grad_norm": 0.1660327911376953, + "learning_rate": 2.4977569521941134e-05, + "loss": 0.0191, + "step": 68350 + }, + { + "epoch": 13.271248968496675, + "grad_norm": 0.1376262605190277, + "learning_rate": 2.4963255706874143e-05, + "loss": 0.0229, + "step": 68360 + }, + { + "epoch": 13.273190621814475, + "grad_norm": 0.2307380586862564, + "learning_rate": 2.4948944629787853e-05, + "loss": 0.0251, + "step": 68370 + }, + { + "epoch": 13.275132275132275, + "grad_norm": 0.16063450276851654, + "learning_rate": 2.493463629224731e-05, + "loss": 0.019, + "step": 68380 + }, + { + "epoch": 13.277073928450076, + "grad_norm": 0.18985486030578613, + "learning_rate": 2.492033069581728e-05, + "loss": 0.0235, + "step": 68390 + }, + { + "epoch": 13.279015581767876, + "grad_norm": 0.1429871767759323, + "learning_rate": 2.4906027842062137e-05, + "loss": 0.0213, + "step": 68400 + }, + { + "epoch": 13.280957235085676, + "grad_norm": 0.1464131623506546, + "learning_rate": 2.48917277325461e-05, + "loss": 0.0235, + "step": 68410 + }, + { + "epoch": 13.282898888403476, + "grad_norm": 0.22128961980342865, + "learning_rate": 2.487743036883295e-05, + "loss": 0.0302, + "step": 68420 + }, + { + "epoch": 13.284840541721275, + "grad_norm": 0.2325209230184555, + "learning_rate": 2.486313575248626e-05, + "loss": 0.0265, + "step": 68430 + }, + { + "epoch": 13.286782195039075, + "grad_norm": 0.3590739965438843, + "learning_rate": 2.4848843885069235e-05, + "loss": 0.0348, + "step": 68440 + }, + { + "epoch": 13.288723848356875, + "grad_norm": 0.09394557029008865, + "learning_rate": 2.483455476814484e-05, + "loss": 0.0192, + "step": 68450 + }, + { + "epoch": 13.290665501674676, + "grad_norm": 0.3361545205116272, + "learning_rate": 2.4820268403275672e-05, + "loss": 0.0221, + "step": 68460 + }, + { + "epoch": 13.292607154992476, + "grad_norm": 0.2665104866027832, + "learning_rate": 2.480598479202413e-05, + "loss": 0.0188, + "step": 68470 + }, + { + "epoch": 13.294548808310276, + "grad_norm": 0.0908730998635292, + "learning_rate": 2.4791703935952193e-05, + "loss": 0.0232, + "step": 68480 + }, + { + "epoch": 13.296490461628077, + "grad_norm": 0.16664788126945496, + "learning_rate": 2.477742583662163e-05, + "loss": 0.0209, + "step": 68490 + }, + { + "epoch": 13.298432114945877, + "grad_norm": 0.18516601622104645, + "learning_rate": 2.476315049559384e-05, + "loss": 0.0278, + "step": 68500 + }, + { + "epoch": 13.300373768263677, + "grad_norm": 0.13973885774612427, + "learning_rate": 2.4748877914429987e-05, + "loss": 0.024, + "step": 68510 + }, + { + "epoch": 13.302315421581477, + "grad_norm": 0.08838848769664764, + "learning_rate": 2.4734608094690857e-05, + "loss": 0.0204, + "step": 68520 + }, + { + "epoch": 13.304257074899276, + "grad_norm": 0.15474705398082733, + "learning_rate": 2.4720341037937006e-05, + "loss": 0.0218, + "step": 68530 + }, + { + "epoch": 13.306198728217076, + "grad_norm": 0.2057352364063263, + "learning_rate": 2.4706076745728645e-05, + "loss": 0.0233, + "step": 68540 + }, + { + "epoch": 13.308140381534876, + "grad_norm": 0.22895309329032898, + "learning_rate": 2.469181521962573e-05, + "loss": 0.0251, + "step": 68550 + }, + { + "epoch": 13.310082034852677, + "grad_norm": 0.22836105525493622, + "learning_rate": 2.467755646118783e-05, + "loss": 0.0254, + "step": 68560 + }, + { + "epoch": 13.312023688170477, + "grad_norm": 0.18750326335430145, + "learning_rate": 2.46633004719743e-05, + "loss": 0.0228, + "step": 68570 + }, + { + "epoch": 13.313965341488277, + "grad_norm": 0.30729472637176514, + "learning_rate": 2.4649047253544123e-05, + "loss": 0.0238, + "step": 68580 + }, + { + "epoch": 13.315906994806078, + "grad_norm": 0.1749935895204544, + "learning_rate": 2.4634796807456016e-05, + "loss": 0.026, + "step": 68590 + }, + { + "epoch": 13.317848648123878, + "grad_norm": 0.16254740953445435, + "learning_rate": 2.46205491352684e-05, + "loss": 0.0194, + "step": 68600 + }, + { + "epoch": 13.319790301441678, + "grad_norm": 0.3116527199745178, + "learning_rate": 2.4606304238539375e-05, + "loss": 0.02, + "step": 68610 + }, + { + "epoch": 13.321731954759478, + "grad_norm": 0.1750473529100418, + "learning_rate": 2.4592062118826753e-05, + "loss": 0.0206, + "step": 68620 + }, + { + "epoch": 13.323673608077279, + "grad_norm": 0.13405704498291016, + "learning_rate": 2.4577822777688004e-05, + "loss": 0.0278, + "step": 68630 + }, + { + "epoch": 13.325615261395077, + "grad_norm": 0.17063425481319427, + "learning_rate": 2.4563586216680347e-05, + "loss": 0.0213, + "step": 68640 + }, + { + "epoch": 13.327556914712877, + "grad_norm": 0.2725241780281067, + "learning_rate": 2.4549352437360622e-05, + "loss": 0.0239, + "step": 68650 + }, + { + "epoch": 13.329498568030678, + "grad_norm": 0.14839395880699158, + "learning_rate": 2.4535121441285493e-05, + "loss": 0.0196, + "step": 68660 + }, + { + "epoch": 13.331440221348478, + "grad_norm": 0.19642989337444305, + "learning_rate": 2.4520893230011172e-05, + "loss": 0.0236, + "step": 68670 + }, + { + "epoch": 13.333381874666278, + "grad_norm": 0.18892084062099457, + "learning_rate": 2.450666780509368e-05, + "loss": 0.0216, + "step": 68680 + }, + { + "epoch": 13.335323527984078, + "grad_norm": 0.15907220542430878, + "learning_rate": 2.4492445168088656e-05, + "loss": 0.0215, + "step": 68690 + }, + { + "epoch": 13.337265181301879, + "grad_norm": 0.23762337863445282, + "learning_rate": 2.447822532055149e-05, + "loss": 0.0257, + "step": 68700 + }, + { + "epoch": 13.339206834619679, + "grad_norm": 0.21609215438365936, + "learning_rate": 2.4464008264037212e-05, + "loss": 0.0211, + "step": 68710 + }, + { + "epoch": 13.34114848793748, + "grad_norm": 0.2774966061115265, + "learning_rate": 2.4449794000100605e-05, + "loss": 0.034, + "step": 68720 + }, + { + "epoch": 13.34309014125528, + "grad_norm": 0.1500200629234314, + "learning_rate": 2.4435582530296108e-05, + "loss": 0.0263, + "step": 68730 + }, + { + "epoch": 13.345031794573078, + "grad_norm": 0.15308339893817902, + "learning_rate": 2.4421373856177887e-05, + "loss": 0.0271, + "step": 68740 + }, + { + "epoch": 13.346973447890878, + "grad_norm": 0.18978825211524963, + "learning_rate": 2.440716797929975e-05, + "loss": 0.0227, + "step": 68750 + }, + { + "epoch": 13.348915101208679, + "grad_norm": 0.1706586629152298, + "learning_rate": 2.439296490121526e-05, + "loss": 0.0218, + "step": 68760 + }, + { + "epoch": 13.350856754526479, + "grad_norm": 0.18625648319721222, + "learning_rate": 2.4378764623477608e-05, + "loss": 0.024, + "step": 68770 + }, + { + "epoch": 13.35279840784428, + "grad_norm": 0.17890511453151703, + "learning_rate": 2.436456714763974e-05, + "loss": 0.0281, + "step": 68780 + }, + { + "epoch": 13.35474006116208, + "grad_norm": 0.2734157145023346, + "learning_rate": 2.4350372475254262e-05, + "loss": 0.0205, + "step": 68790 + }, + { + "epoch": 13.35668171447988, + "grad_norm": 0.15055881440639496, + "learning_rate": 2.433618060787351e-05, + "loss": 0.0229, + "step": 68800 + }, + { + "epoch": 13.35862336779768, + "grad_norm": 0.1459416300058365, + "learning_rate": 2.432199154704944e-05, + "loss": 0.0246, + "step": 68810 + }, + { + "epoch": 13.36056502111548, + "grad_norm": 0.28429368138313293, + "learning_rate": 2.430780529433377e-05, + "loss": 0.0235, + "step": 68820 + }, + { + "epoch": 13.36250667443328, + "grad_norm": 0.14674335718154907, + "learning_rate": 2.42936218512779e-05, + "loss": 0.0179, + "step": 68830 + }, + { + "epoch": 13.36444832775108, + "grad_norm": 0.1935519427061081, + "learning_rate": 2.427944121943288e-05, + "loss": 0.0287, + "step": 68840 + }, + { + "epoch": 13.36638998106888, + "grad_norm": 0.1967054307460785, + "learning_rate": 2.4265263400349497e-05, + "loss": 0.0191, + "step": 68850 + }, + { + "epoch": 13.36833163438668, + "grad_norm": 0.18851053714752197, + "learning_rate": 2.4251088395578213e-05, + "loss": 0.0271, + "step": 68860 + }, + { + "epoch": 13.37027328770448, + "grad_norm": 0.18397191166877747, + "learning_rate": 2.4236916206669214e-05, + "loss": 0.0182, + "step": 68870 + }, + { + "epoch": 13.37221494102228, + "grad_norm": 0.13107642531394958, + "learning_rate": 2.4222746835172304e-05, + "loss": 0.0283, + "step": 68880 + }, + { + "epoch": 13.37415659434008, + "grad_norm": 0.23770861327648163, + "learning_rate": 2.4208580282637065e-05, + "loss": 0.0277, + "step": 68890 + }, + { + "epoch": 13.37609824765788, + "grad_norm": 0.17922073602676392, + "learning_rate": 2.4194416550612687e-05, + "loss": 0.0221, + "step": 68900 + }, + { + "epoch": 13.378039900975681, + "grad_norm": 0.14918451011180878, + "learning_rate": 2.418025564064812e-05, + "loss": 0.0296, + "step": 68910 + }, + { + "epoch": 13.379981554293481, + "grad_norm": 0.20995520055294037, + "learning_rate": 2.4166097554291978e-05, + "loss": 0.0307, + "step": 68920 + }, + { + "epoch": 13.381923207611282, + "grad_norm": 0.13787293434143066, + "learning_rate": 2.4151942293092583e-05, + "loss": 0.0179, + "step": 68930 + }, + { + "epoch": 13.383864860929082, + "grad_norm": 0.15265734493732452, + "learning_rate": 2.41377898585979e-05, + "loss": 0.0284, + "step": 68940 + }, + { + "epoch": 13.385806514246882, + "grad_norm": 0.14457270503044128, + "learning_rate": 2.4123640252355652e-05, + "loss": 0.0194, + "step": 68950 + }, + { + "epoch": 13.38774816756468, + "grad_norm": 0.23642154037952423, + "learning_rate": 2.410949347591318e-05, + "loss": 0.0246, + "step": 68960 + }, + { + "epoch": 13.38968982088248, + "grad_norm": 0.10034576803445816, + "learning_rate": 2.4095349530817578e-05, + "loss": 0.0229, + "step": 68970 + }, + { + "epoch": 13.391631474200281, + "grad_norm": 0.09798238426446915, + "learning_rate": 2.4081208418615603e-05, + "loss": 0.0176, + "step": 68980 + }, + { + "epoch": 13.393573127518081, + "grad_norm": 0.16288138926029205, + "learning_rate": 2.4067070140853725e-05, + "loss": 0.0283, + "step": 68990 + }, + { + "epoch": 13.395514780835882, + "grad_norm": 0.2292538732290268, + "learning_rate": 2.405293469907805e-05, + "loss": 0.0218, + "step": 69000 + }, + { + "epoch": 13.397456434153682, + "grad_norm": 0.17066998779773712, + "learning_rate": 2.4038802094834446e-05, + "loss": 0.0235, + "step": 69010 + }, + { + "epoch": 13.399398087471482, + "grad_norm": 0.23797234892845154, + "learning_rate": 2.4024672329668395e-05, + "loss": 0.0241, + "step": 69020 + }, + { + "epoch": 13.401339740789282, + "grad_norm": 0.16829703748226166, + "learning_rate": 2.4010545405125125e-05, + "loss": 0.0196, + "step": 69030 + }, + { + "epoch": 13.403281394107083, + "grad_norm": 0.21426016092300415, + "learning_rate": 2.3996421322749528e-05, + "loss": 0.0213, + "step": 69040 + }, + { + "epoch": 13.405223047424883, + "grad_norm": 0.1166730746626854, + "learning_rate": 2.3982300084086224e-05, + "loss": 0.0259, + "step": 69050 + }, + { + "epoch": 13.407164700742683, + "grad_norm": 0.14238646626472473, + "learning_rate": 2.396818169067944e-05, + "loss": 0.0233, + "step": 69060 + }, + { + "epoch": 13.409106354060482, + "grad_norm": 0.1392161250114441, + "learning_rate": 2.3954066144073178e-05, + "loss": 0.0236, + "step": 69070 + }, + { + "epoch": 13.411048007378282, + "grad_norm": 0.1301308274269104, + "learning_rate": 2.39399534458111e-05, + "loss": 0.0222, + "step": 69080 + }, + { + "epoch": 13.412989660696082, + "grad_norm": 0.23285983502864838, + "learning_rate": 2.392584359743651e-05, + "loss": 0.0227, + "step": 69090 + }, + { + "epoch": 13.414931314013883, + "grad_norm": 0.10524749755859375, + "learning_rate": 2.3911736600492463e-05, + "loss": 0.0195, + "step": 69100 + }, + { + "epoch": 13.416872967331683, + "grad_norm": 0.10645534098148346, + "learning_rate": 2.3897632456521672e-05, + "loss": 0.0262, + "step": 69110 + }, + { + "epoch": 13.418814620649483, + "grad_norm": 0.20202071964740753, + "learning_rate": 2.388353116706658e-05, + "loss": 0.0254, + "step": 69120 + }, + { + "epoch": 13.420756273967283, + "grad_norm": 0.18133661150932312, + "learning_rate": 2.3869432733669232e-05, + "loss": 0.0241, + "step": 69130 + }, + { + "epoch": 13.422697927285084, + "grad_norm": 0.09555380046367645, + "learning_rate": 2.385533715787145e-05, + "loss": 0.0218, + "step": 69140 + }, + { + "epoch": 13.424639580602884, + "grad_norm": 0.2512674033641815, + "learning_rate": 2.3841244441214673e-05, + "loss": 0.0222, + "step": 69150 + }, + { + "epoch": 13.426581233920684, + "grad_norm": 0.09048847109079361, + "learning_rate": 2.3827154585240073e-05, + "loss": 0.0186, + "step": 69160 + }, + { + "epoch": 13.428522887238483, + "grad_norm": 0.20654912292957306, + "learning_rate": 2.3813067591488498e-05, + "loss": 0.0268, + "step": 69170 + }, + { + "epoch": 13.430464540556283, + "grad_norm": 0.2642638385295868, + "learning_rate": 2.3798983461500494e-05, + "loss": 0.0235, + "step": 69180 + }, + { + "epoch": 13.432406193874083, + "grad_norm": 0.1920076161623001, + "learning_rate": 2.3784902196816256e-05, + "loss": 0.0213, + "step": 69190 + }, + { + "epoch": 13.434347847191884, + "grad_norm": 0.07393290847539902, + "learning_rate": 2.3770823798975716e-05, + "loss": 0.0332, + "step": 69200 + }, + { + "epoch": 13.436289500509684, + "grad_norm": 0.15021094679832458, + "learning_rate": 2.3756748269518424e-05, + "loss": 0.0199, + "step": 69210 + }, + { + "epoch": 13.438231153827484, + "grad_norm": 0.14685770869255066, + "learning_rate": 2.374267560998371e-05, + "loss": 0.0226, + "step": 69220 + }, + { + "epoch": 13.440172807145284, + "grad_norm": 0.13027243316173553, + "learning_rate": 2.3728605821910477e-05, + "loss": 0.0212, + "step": 69230 + }, + { + "epoch": 13.442114460463085, + "grad_norm": 0.1927196979522705, + "learning_rate": 2.3714538906837453e-05, + "loss": 0.0218, + "step": 69240 + }, + { + "epoch": 13.444056113780885, + "grad_norm": 0.13191565871238708, + "learning_rate": 2.3700474866302913e-05, + "loss": 0.0258, + "step": 69250 + }, + { + "epoch": 13.445997767098685, + "grad_norm": 0.1816159039735794, + "learning_rate": 2.3686413701844917e-05, + "loss": 0.0292, + "step": 69260 + }, + { + "epoch": 13.447939420416485, + "grad_norm": 0.14019939303398132, + "learning_rate": 2.367235541500114e-05, + "loss": 0.0198, + "step": 69270 + }, + { + "epoch": 13.449881073734284, + "grad_norm": 0.1779610961675644, + "learning_rate": 2.3658300007308993e-05, + "loss": 0.0228, + "step": 69280 + }, + { + "epoch": 13.451822727052084, + "grad_norm": 0.13029718399047852, + "learning_rate": 2.364424748030555e-05, + "loss": 0.0229, + "step": 69290 + }, + { + "epoch": 13.453764380369885, + "grad_norm": 0.16662073135375977, + "learning_rate": 2.3630197835527595e-05, + "loss": 0.0193, + "step": 69300 + }, + { + "epoch": 13.455706033687685, + "grad_norm": 0.21701470017433167, + "learning_rate": 2.3616151074511546e-05, + "loss": 0.0238, + "step": 69310 + }, + { + "epoch": 13.457647687005485, + "grad_norm": 0.19402927160263062, + "learning_rate": 2.3602107198793545e-05, + "loss": 0.0208, + "step": 69320 + }, + { + "epoch": 13.459589340323285, + "grad_norm": 0.27070289850234985, + "learning_rate": 2.3588066209909427e-05, + "loss": 0.0243, + "step": 69330 + }, + { + "epoch": 13.461530993641086, + "grad_norm": 0.19405917823314667, + "learning_rate": 2.3574028109394662e-05, + "loss": 0.0228, + "step": 69340 + }, + { + "epoch": 13.463472646958886, + "grad_norm": 0.2263776808977127, + "learning_rate": 2.3559992898784454e-05, + "loss": 0.0198, + "step": 69350 + }, + { + "epoch": 13.465414300276686, + "grad_norm": 0.17227418720722198, + "learning_rate": 2.354596057961367e-05, + "loss": 0.02, + "step": 69360 + }, + { + "epoch": 13.467355953594486, + "grad_norm": 0.18196138739585876, + "learning_rate": 2.353193115341688e-05, + "loss": 0.0206, + "step": 69370 + }, + { + "epoch": 13.469297606912285, + "grad_norm": 0.1212795078754425, + "learning_rate": 2.3517904621728294e-05, + "loss": 0.0276, + "step": 69380 + }, + { + "epoch": 13.471239260230085, + "grad_norm": 0.18073557317256927, + "learning_rate": 2.3503880986081855e-05, + "loss": 0.0252, + "step": 69390 + }, + { + "epoch": 13.473180913547885, + "grad_norm": 0.1265541911125183, + "learning_rate": 2.3489860248011136e-05, + "loss": 0.0226, + "step": 69400 + }, + { + "epoch": 13.475122566865686, + "grad_norm": 0.2000221461057663, + "learning_rate": 2.3475842409049464e-05, + "loss": 0.0234, + "step": 69410 + }, + { + "epoch": 13.477064220183486, + "grad_norm": 0.22059772908687592, + "learning_rate": 2.346182747072975e-05, + "loss": 0.0281, + "step": 69420 + }, + { + "epoch": 13.479005873501286, + "grad_norm": 0.17589884996414185, + "learning_rate": 2.3447815434584718e-05, + "loss": 0.022, + "step": 69430 + }, + { + "epoch": 13.480947526819087, + "grad_norm": 0.15858472883701324, + "learning_rate": 2.3433806302146656e-05, + "loss": 0.0199, + "step": 69440 + }, + { + "epoch": 13.482889180136887, + "grad_norm": 0.28899046778678894, + "learning_rate": 2.3419800074947612e-05, + "loss": 0.0258, + "step": 69450 + }, + { + "epoch": 13.484830833454687, + "grad_norm": 0.15530264377593994, + "learning_rate": 2.3405796754519248e-05, + "loss": 0.0225, + "step": 69460 + }, + { + "epoch": 13.486772486772487, + "grad_norm": 0.19494308531284332, + "learning_rate": 2.3391796342392992e-05, + "loss": 0.0267, + "step": 69470 + }, + { + "epoch": 13.488714140090288, + "grad_norm": 0.18831424415111542, + "learning_rate": 2.337779884009984e-05, + "loss": 0.0257, + "step": 69480 + }, + { + "epoch": 13.490655793408086, + "grad_norm": 0.18684223294258118, + "learning_rate": 2.3363804249170624e-05, + "loss": 0.0229, + "step": 69490 + }, + { + "epoch": 13.492597446725886, + "grad_norm": 0.273855060338974, + "learning_rate": 2.3349812571135703e-05, + "loss": 0.0255, + "step": 69500 + }, + { + "epoch": 13.494539100043687, + "grad_norm": 0.17906484007835388, + "learning_rate": 2.333582380752523e-05, + "loss": 0.0187, + "step": 69510 + }, + { + "epoch": 13.496480753361487, + "grad_norm": 0.16371826827526093, + "learning_rate": 2.3321837959868963e-05, + "loss": 0.0184, + "step": 69520 + }, + { + "epoch": 13.498422406679287, + "grad_norm": 0.24277840554714203, + "learning_rate": 2.3307855029696383e-05, + "loss": 0.0216, + "step": 69530 + }, + { + "epoch": 13.500364059997088, + "grad_norm": 0.1966646909713745, + "learning_rate": 2.3293875018536648e-05, + "loss": 0.0227, + "step": 69540 + }, + { + "epoch": 13.502305713314888, + "grad_norm": 0.14212334156036377, + "learning_rate": 2.327989792791861e-05, + "loss": 0.0253, + "step": 69550 + }, + { + "epoch": 13.504247366632688, + "grad_norm": 0.2413787990808487, + "learning_rate": 2.326592375937074e-05, + "loss": 0.0243, + "step": 69560 + }, + { + "epoch": 13.506189019950488, + "grad_norm": 0.17208407819271088, + "learning_rate": 2.3251952514421248e-05, + "loss": 0.0226, + "step": 69570 + }, + { + "epoch": 13.508130673268289, + "grad_norm": 0.26632896065711975, + "learning_rate": 2.323798419459804e-05, + "loss": 0.0294, + "step": 69580 + }, + { + "epoch": 13.510072326586087, + "grad_norm": 0.17470957338809967, + "learning_rate": 2.3224018801428625e-05, + "loss": 0.0245, + "step": 69590 + }, + { + "epoch": 13.512013979903887, + "grad_norm": 0.22522172331809998, + "learning_rate": 2.3210056336440273e-05, + "loss": 0.0288, + "step": 69600 + }, + { + "epoch": 13.513955633221688, + "grad_norm": 0.15403634309768677, + "learning_rate": 2.3196096801159844e-05, + "loss": 0.0276, + "step": 69610 + }, + { + "epoch": 13.515897286539488, + "grad_norm": 0.16726452112197876, + "learning_rate": 2.3182140197114012e-05, + "loss": 0.0228, + "step": 69620 + }, + { + "epoch": 13.517838939857288, + "grad_norm": 0.24128201603889465, + "learning_rate": 2.3168186525828985e-05, + "loss": 0.0225, + "step": 69630 + }, + { + "epoch": 13.519780593175089, + "grad_norm": 0.21044830977916718, + "learning_rate": 2.3154235788830757e-05, + "loss": 0.0273, + "step": 69640 + }, + { + "epoch": 13.521722246492889, + "grad_norm": 0.17690639197826385, + "learning_rate": 2.3140287987644927e-05, + "loss": 0.0217, + "step": 69650 + }, + { + "epoch": 13.523663899810689, + "grad_norm": 0.09444458037614822, + "learning_rate": 2.3126343123796833e-05, + "loss": 0.0177, + "step": 69660 + }, + { + "epoch": 13.52560555312849, + "grad_norm": 0.15242336690425873, + "learning_rate": 2.3112401198811418e-05, + "loss": 0.0166, + "step": 69670 + }, + { + "epoch": 13.52754720644629, + "grad_norm": 0.1298152357339859, + "learning_rate": 2.3098462214213418e-05, + "loss": 0.0288, + "step": 69680 + }, + { + "epoch": 13.52948885976409, + "grad_norm": 0.14646472036838531, + "learning_rate": 2.308452617152712e-05, + "loss": 0.023, + "step": 69690 + }, + { + "epoch": 13.53143051308189, + "grad_norm": 0.16243360936641693, + "learning_rate": 2.3070593072276603e-05, + "loss": 0.0201, + "step": 69700 + }, + { + "epoch": 13.533372166399689, + "grad_norm": 0.1432775855064392, + "learning_rate": 2.3056662917985518e-05, + "loss": 0.0257, + "step": 69710 + }, + { + "epoch": 13.535313819717489, + "grad_norm": 0.16989396512508392, + "learning_rate": 2.3042735710177283e-05, + "loss": 0.0224, + "step": 69720 + }, + { + "epoch": 13.53725547303529, + "grad_norm": 0.1428575962781906, + "learning_rate": 2.302881145037492e-05, + "loss": 0.0178, + "step": 69730 + }, + { + "epoch": 13.53919712635309, + "grad_norm": 0.18975843489170074, + "learning_rate": 2.3014890140101192e-05, + "loss": 0.0177, + "step": 69740 + }, + { + "epoch": 13.54113877967089, + "grad_norm": 0.21855542063713074, + "learning_rate": 2.30009717808785e-05, + "loss": 0.0213, + "step": 69750 + }, + { + "epoch": 13.54308043298869, + "grad_norm": 0.10782883316278458, + "learning_rate": 2.2987056374228967e-05, + "loss": 0.0215, + "step": 69760 + }, + { + "epoch": 13.54502208630649, + "grad_norm": 0.20091421902179718, + "learning_rate": 2.297314392167431e-05, + "loss": 0.0209, + "step": 69770 + }, + { + "epoch": 13.54696373962429, + "grad_norm": 0.1946532279253006, + "learning_rate": 2.295923442473601e-05, + "loss": 0.0203, + "step": 69780 + }, + { + "epoch": 13.54890539294209, + "grad_norm": 0.2004549354314804, + "learning_rate": 2.294532788493519e-05, + "loss": 0.0179, + "step": 69790 + }, + { + "epoch": 13.550847046259891, + "grad_norm": 0.15496915578842163, + "learning_rate": 2.293142430379262e-05, + "loss": 0.0229, + "step": 69800 + }, + { + "epoch": 13.55278869957769, + "grad_norm": 0.1538456380367279, + "learning_rate": 2.291752368282879e-05, + "loss": 0.0201, + "step": 69810 + }, + { + "epoch": 13.55473035289549, + "grad_norm": 0.2967134416103363, + "learning_rate": 2.290362602356385e-05, + "loss": 0.021, + "step": 69820 + }, + { + "epoch": 13.55667200621329, + "grad_norm": 0.21002620458602905, + "learning_rate": 2.288973132751765e-05, + "loss": 0.0203, + "step": 69830 + }, + { + "epoch": 13.55861365953109, + "grad_norm": 0.19398953020572662, + "learning_rate": 2.287583959620965e-05, + "loss": 0.0262, + "step": 69840 + }, + { + "epoch": 13.56055531284889, + "grad_norm": 0.14324435591697693, + "learning_rate": 2.2861950831159074e-05, + "loss": 0.0214, + "step": 69850 + }, + { + "epoch": 13.562496966166691, + "grad_norm": 0.13842783868312836, + "learning_rate": 2.284806503388472e-05, + "loss": 0.0287, + "step": 69860 + }, + { + "epoch": 13.564438619484491, + "grad_norm": 0.1935422420501709, + "learning_rate": 2.283418220590518e-05, + "loss": 0.0279, + "step": 69870 + }, + { + "epoch": 13.566380272802292, + "grad_norm": 0.0929589718580246, + "learning_rate": 2.282030234873862e-05, + "loss": 0.0208, + "step": 69880 + }, + { + "epoch": 13.568321926120092, + "grad_norm": 0.20460176467895508, + "learning_rate": 2.2806425463902948e-05, + "loss": 0.0236, + "step": 69890 + }, + { + "epoch": 13.570263579437892, + "grad_norm": 0.33424368500709534, + "learning_rate": 2.279255155291568e-05, + "loss": 0.0286, + "step": 69900 + }, + { + "epoch": 13.572205232755692, + "grad_norm": 0.2228831648826599, + "learning_rate": 2.277868061729409e-05, + "loss": 0.0224, + "step": 69910 + }, + { + "epoch": 13.57414688607349, + "grad_norm": 0.14538517594337463, + "learning_rate": 2.276481265855504e-05, + "loss": 0.0241, + "step": 69920 + }, + { + "epoch": 13.576088539391291, + "grad_norm": 0.11858521401882172, + "learning_rate": 2.2750947678215128e-05, + "loss": 0.0218, + "step": 69930 + }, + { + "epoch": 13.578030192709091, + "grad_norm": 0.2536510229110718, + "learning_rate": 2.2737085677790616e-05, + "loss": 0.028, + "step": 69940 + }, + { + "epoch": 13.579971846026892, + "grad_norm": 0.14402341842651367, + "learning_rate": 2.2723226658797443e-05, + "loss": 0.0198, + "step": 69950 + }, + { + "epoch": 13.581913499344692, + "grad_norm": 0.15396077930927277, + "learning_rate": 2.270937062275117e-05, + "loss": 0.021, + "step": 69960 + }, + { + "epoch": 13.583855152662492, + "grad_norm": 0.11144743859767914, + "learning_rate": 2.2695517571167113e-05, + "loss": 0.0264, + "step": 69970 + }, + { + "epoch": 13.585796805980292, + "grad_norm": 0.11537624150514603, + "learning_rate": 2.268166750556019e-05, + "loss": 0.0215, + "step": 69980 + }, + { + "epoch": 13.587738459298093, + "grad_norm": 0.15271389484405518, + "learning_rate": 2.266782042744503e-05, + "loss": 0.0196, + "step": 69990 + }, + { + "epoch": 13.589680112615893, + "grad_norm": 0.19057799875736237, + "learning_rate": 2.2653976338335935e-05, + "loss": 0.0241, + "step": 70000 + }, + { + "epoch": 13.591621765933693, + "grad_norm": 0.1288520246744156, + "learning_rate": 2.2640135239746895e-05, + "loss": 0.0216, + "step": 70010 + }, + { + "epoch": 13.593563419251492, + "grad_norm": 0.16369809210300446, + "learning_rate": 2.2626297133191515e-05, + "loss": 0.0212, + "step": 70020 + }, + { + "epoch": 13.595505072569292, + "grad_norm": 0.22443737089633942, + "learning_rate": 2.261246202018312e-05, + "loss": 0.0243, + "step": 70030 + }, + { + "epoch": 13.597446725887092, + "grad_norm": 0.32064729928970337, + "learning_rate": 2.2598629902234718e-05, + "loss": 0.0231, + "step": 70040 + }, + { + "epoch": 13.599388379204893, + "grad_norm": 0.2881312966346741, + "learning_rate": 2.258480078085894e-05, + "loss": 0.024, + "step": 70050 + }, + { + "epoch": 13.601330032522693, + "grad_norm": 0.15370024740695953, + "learning_rate": 2.2570974657568126e-05, + "loss": 0.0251, + "step": 70060 + }, + { + "epoch": 13.603271685840493, + "grad_norm": 0.14874430000782013, + "learning_rate": 2.2557151533874286e-05, + "loss": 0.0278, + "step": 70070 + }, + { + "epoch": 13.605213339158293, + "grad_norm": 0.1748114675283432, + "learning_rate": 2.2543331411289116e-05, + "loss": 0.0272, + "step": 70080 + }, + { + "epoch": 13.607154992476094, + "grad_norm": 0.15684247016906738, + "learning_rate": 2.252951429132392e-05, + "loss": 0.0241, + "step": 70090 + }, + { + "epoch": 13.609096645793894, + "grad_norm": 0.18482227623462677, + "learning_rate": 2.2515700175489762e-05, + "loss": 0.0239, + "step": 70100 + }, + { + "epoch": 13.611038299111694, + "grad_norm": 0.16799093782901764, + "learning_rate": 2.2501889065297288e-05, + "loss": 0.0246, + "step": 70110 + }, + { + "epoch": 13.612979952429495, + "grad_norm": 0.12355061620473862, + "learning_rate": 2.2488080962256887e-05, + "loss": 0.0218, + "step": 70120 + }, + { + "epoch": 13.614921605747293, + "grad_norm": 0.1478356122970581, + "learning_rate": 2.2474275867878585e-05, + "loss": 0.0297, + "step": 70130 + }, + { + "epoch": 13.616863259065093, + "grad_norm": 0.19367782771587372, + "learning_rate": 2.246047378367211e-05, + "loss": 0.0225, + "step": 70140 + }, + { + "epoch": 13.618804912382894, + "grad_norm": 0.18557989597320557, + "learning_rate": 2.2446674711146798e-05, + "loss": 0.0263, + "step": 70150 + }, + { + "epoch": 13.620746565700694, + "grad_norm": 0.12031278759241104, + "learning_rate": 2.2432878651811734e-05, + "loss": 0.0236, + "step": 70160 + }, + { + "epoch": 13.622688219018494, + "grad_norm": 0.1556462198495865, + "learning_rate": 2.2419085607175594e-05, + "loss": 0.0281, + "step": 70170 + }, + { + "epoch": 13.624629872336294, + "grad_norm": 0.15462574362754822, + "learning_rate": 2.240529557874678e-05, + "loss": 0.0224, + "step": 70180 + }, + { + "epoch": 13.626571525654095, + "grad_norm": 0.27968502044677734, + "learning_rate": 2.2391508568033358e-05, + "loss": 0.0251, + "step": 70190 + }, + { + "epoch": 13.628513178971895, + "grad_norm": 0.15110786259174347, + "learning_rate": 2.237772457654307e-05, + "loss": 0.0195, + "step": 70200 + }, + { + "epoch": 13.630454832289695, + "grad_norm": 0.19144366681575775, + "learning_rate": 2.236394360578327e-05, + "loss": 0.025, + "step": 70210 + }, + { + "epoch": 13.632396485607496, + "grad_norm": 0.18963705003261566, + "learning_rate": 2.2350165657261073e-05, + "loss": 0.023, + "step": 70220 + }, + { + "epoch": 13.634338138925294, + "grad_norm": 0.1343594640493393, + "learning_rate": 2.2336390732483164e-05, + "loss": 0.0268, + "step": 70230 + }, + { + "epoch": 13.636279792243094, + "grad_norm": 0.1154949963092804, + "learning_rate": 2.2322618832955976e-05, + "loss": 0.0233, + "step": 70240 + }, + { + "epoch": 13.638221445560895, + "grad_norm": 0.12292136996984482, + "learning_rate": 2.2308849960185578e-05, + "loss": 0.0279, + "step": 70250 + }, + { + "epoch": 13.640163098878695, + "grad_norm": 0.1995978206396103, + "learning_rate": 2.229508411567774e-05, + "loss": 0.0294, + "step": 70260 + }, + { + "epoch": 13.642104752196495, + "grad_norm": 0.1562691628932953, + "learning_rate": 2.2281321300937823e-05, + "loss": 0.0262, + "step": 70270 + }, + { + "epoch": 13.644046405514295, + "grad_norm": 0.16874919831752777, + "learning_rate": 2.2267561517470937e-05, + "loss": 0.0211, + "step": 70280 + }, + { + "epoch": 13.645988058832096, + "grad_norm": 0.061015065759420395, + "learning_rate": 2.2253804766781845e-05, + "loss": 0.0263, + "step": 70290 + }, + { + "epoch": 13.647929712149896, + "grad_norm": 0.17484548687934875, + "learning_rate": 2.2240051050374933e-05, + "loss": 0.0231, + "step": 70300 + }, + { + "epoch": 13.649871365467696, + "grad_norm": 0.16453954577445984, + "learning_rate": 2.22263003697543e-05, + "loss": 0.0273, + "step": 70310 + }, + { + "epoch": 13.651813018785496, + "grad_norm": 0.18921509385108948, + "learning_rate": 2.2212552726423702e-05, + "loss": 0.0184, + "step": 70320 + }, + { + "epoch": 13.653754672103297, + "grad_norm": 0.19155284762382507, + "learning_rate": 2.2198808121886578e-05, + "loss": 0.0264, + "step": 70330 + }, + { + "epoch": 13.655696325421095, + "grad_norm": 0.2882726192474365, + "learning_rate": 2.218506655764598e-05, + "loss": 0.0312, + "step": 70340 + }, + { + "epoch": 13.657637978738896, + "grad_norm": 0.13021230697631836, + "learning_rate": 2.2171328035204704e-05, + "loss": 0.0232, + "step": 70350 + }, + { + "epoch": 13.659579632056696, + "grad_norm": 0.11786351352930069, + "learning_rate": 2.2157592556065133e-05, + "loss": 0.0178, + "step": 70360 + }, + { + "epoch": 13.661521285374496, + "grad_norm": 0.24968385696411133, + "learning_rate": 2.2143860121729382e-05, + "loss": 0.0175, + "step": 70370 + }, + { + "epoch": 13.663462938692296, + "grad_norm": 0.17163537442684174, + "learning_rate": 2.2130130733699206e-05, + "loss": 0.0296, + "step": 70380 + }, + { + "epoch": 13.665404592010097, + "grad_norm": 0.12597139179706573, + "learning_rate": 2.2116404393476052e-05, + "loss": 0.0257, + "step": 70390 + }, + { + "epoch": 13.667346245327897, + "grad_norm": 0.20219534635543823, + "learning_rate": 2.2102681102560967e-05, + "loss": 0.0195, + "step": 70400 + }, + { + "epoch": 13.669287898645697, + "grad_norm": 0.11714940518140793, + "learning_rate": 2.208896086245476e-05, + "loss": 0.0245, + "step": 70410 + }, + { + "epoch": 13.671229551963497, + "grad_norm": 0.17330530285835266, + "learning_rate": 2.2075243674657808e-05, + "loss": 0.0283, + "step": 70420 + }, + { + "epoch": 13.673171205281298, + "grad_norm": 0.12623627483844757, + "learning_rate": 2.206152954067024e-05, + "loss": 0.0239, + "step": 70430 + }, + { + "epoch": 13.675112858599096, + "grad_norm": 0.15443316102027893, + "learning_rate": 2.2047818461991758e-05, + "loss": 0.0238, + "step": 70440 + }, + { + "epoch": 13.677054511916896, + "grad_norm": 0.1470535844564438, + "learning_rate": 2.2034110440121864e-05, + "loss": 0.0211, + "step": 70450 + }, + { + "epoch": 13.678996165234697, + "grad_norm": 0.08970816433429718, + "learning_rate": 2.2020405476559585e-05, + "loss": 0.0203, + "step": 70460 + }, + { + "epoch": 13.680937818552497, + "grad_norm": 0.15464423596858978, + "learning_rate": 2.2006703572803722e-05, + "loss": 0.0186, + "step": 70470 + }, + { + "epoch": 13.682879471870297, + "grad_norm": 0.1788945198059082, + "learning_rate": 2.1993004730352646e-05, + "loss": 0.0214, + "step": 70480 + }, + { + "epoch": 13.684821125188098, + "grad_norm": 0.16433373093605042, + "learning_rate": 2.1979308950704464e-05, + "loss": 0.0252, + "step": 70490 + }, + { + "epoch": 13.686762778505898, + "grad_norm": 0.21598705649375916, + "learning_rate": 2.1965616235356923e-05, + "loss": 0.0259, + "step": 70500 + }, + { + "epoch": 13.688704431823698, + "grad_norm": 0.17330843210220337, + "learning_rate": 2.1951926585807465e-05, + "loss": 0.0225, + "step": 70510 + }, + { + "epoch": 13.690646085141498, + "grad_norm": 0.1412438154220581, + "learning_rate": 2.193824000355313e-05, + "loss": 0.0229, + "step": 70520 + }, + { + "epoch": 13.692587738459299, + "grad_norm": 0.13859087228775024, + "learning_rate": 2.192455649009067e-05, + "loss": 0.0246, + "step": 70530 + }, + { + "epoch": 13.694529391777099, + "grad_norm": 0.15590544044971466, + "learning_rate": 2.1910876046916523e-05, + "loss": 0.0244, + "step": 70540 + }, + { + "epoch": 13.6964710450949, + "grad_norm": 0.12738287448883057, + "learning_rate": 2.1897198675526725e-05, + "loss": 0.0249, + "step": 70550 + }, + { + "epoch": 13.698412698412698, + "grad_norm": 0.20194248855113983, + "learning_rate": 2.1883524377417024e-05, + "loss": 0.0192, + "step": 70560 + }, + { + "epoch": 13.700354351730498, + "grad_norm": 0.18115200102329254, + "learning_rate": 2.1869853154082825e-05, + "loss": 0.0222, + "step": 70570 + }, + { + "epoch": 13.702296005048298, + "grad_norm": 0.1395024210214615, + "learning_rate": 2.1856185007019215e-05, + "loss": 0.0325, + "step": 70580 + }, + { + "epoch": 13.704237658366099, + "grad_norm": 0.1894669532775879, + "learning_rate": 2.1842519937720874e-05, + "loss": 0.0242, + "step": 70590 + }, + { + "epoch": 13.706179311683899, + "grad_norm": 0.32827228307724, + "learning_rate": 2.182885794768224e-05, + "loss": 0.0221, + "step": 70600 + }, + { + "epoch": 13.708120965001699, + "grad_norm": 0.17571164667606354, + "learning_rate": 2.1815199038397326e-05, + "loss": 0.0241, + "step": 70610 + }, + { + "epoch": 13.7100626183195, + "grad_norm": 0.26090681552886963, + "learning_rate": 2.1801543211359893e-05, + "loss": 0.0287, + "step": 70620 + }, + { + "epoch": 13.7120042716373, + "grad_norm": 0.18737013638019562, + "learning_rate": 2.178789046806326e-05, + "loss": 0.0244, + "step": 70630 + }, + { + "epoch": 13.7139459249551, + "grad_norm": 0.10792474448680878, + "learning_rate": 2.1774240810000547e-05, + "loss": 0.0251, + "step": 70640 + }, + { + "epoch": 13.7158875782729, + "grad_norm": 0.09979475289583206, + "learning_rate": 2.17605942386644e-05, + "loss": 0.0193, + "step": 70650 + }, + { + "epoch": 13.717829231590699, + "grad_norm": 0.16170577704906464, + "learning_rate": 2.174695075554723e-05, + "loss": 0.0276, + "step": 70660 + }, + { + "epoch": 13.719770884908499, + "grad_norm": 0.10586127638816833, + "learning_rate": 2.173331036214103e-05, + "loss": 0.0183, + "step": 70670 + }, + { + "epoch": 13.7217125382263, + "grad_norm": 0.14897461235523224, + "learning_rate": 2.171967305993753e-05, + "loss": 0.0222, + "step": 70680 + }, + { + "epoch": 13.7236541915441, + "grad_norm": 0.18546752631664276, + "learning_rate": 2.1706038850428024e-05, + "loss": 0.0212, + "step": 70690 + }, + { + "epoch": 13.7255958448619, + "grad_norm": 0.14501513540744781, + "learning_rate": 2.169240773510361e-05, + "loss": 0.0235, + "step": 70700 + }, + { + "epoch": 13.7275374981797, + "grad_norm": 0.3710647523403168, + "learning_rate": 2.1678779715454906e-05, + "loss": 0.0238, + "step": 70710 + }, + { + "epoch": 13.7294791514975, + "grad_norm": 0.16551896929740906, + "learning_rate": 2.166515479297229e-05, + "loss": 0.0236, + "step": 70720 + }, + { + "epoch": 13.7314208048153, + "grad_norm": 0.14635683596134186, + "learning_rate": 2.1651532969145732e-05, + "loss": 0.0211, + "step": 70730 + }, + { + "epoch": 13.7333624581331, + "grad_norm": 0.1427820324897766, + "learning_rate": 2.1637914245464903e-05, + "loss": 0.0248, + "step": 70740 + }, + { + "epoch": 13.735304111450901, + "grad_norm": 0.29837822914123535, + "learning_rate": 2.1624298623419133e-05, + "loss": 0.0325, + "step": 70750 + }, + { + "epoch": 13.737245764768701, + "grad_norm": 0.2513333261013031, + "learning_rate": 2.1610686104497413e-05, + "loss": 0.0199, + "step": 70760 + }, + { + "epoch": 13.7391874180865, + "grad_norm": 0.14872610569000244, + "learning_rate": 2.1597076690188366e-05, + "loss": 0.0257, + "step": 70770 + }, + { + "epoch": 13.7411290714043, + "grad_norm": 0.2290005385875702, + "learning_rate": 2.1583470381980315e-05, + "loss": 0.0172, + "step": 70780 + }, + { + "epoch": 13.7430707247221, + "grad_norm": 0.2549475431442261, + "learning_rate": 2.156986718136123e-05, + "loss": 0.0225, + "step": 70790 + }, + { + "epoch": 13.7450123780399, + "grad_norm": 0.17099200189113617, + "learning_rate": 2.155626708981871e-05, + "loss": 0.0226, + "step": 70800 + }, + { + "epoch": 13.746954031357701, + "grad_norm": 0.16638639569282532, + "learning_rate": 2.1542670108840075e-05, + "loss": 0.0169, + "step": 70810 + }, + { + "epoch": 13.748895684675501, + "grad_norm": 0.1946866661310196, + "learning_rate": 2.152907623991221e-05, + "loss": 0.0226, + "step": 70820 + }, + { + "epoch": 13.750837337993302, + "grad_norm": 0.23005956411361694, + "learning_rate": 2.1515485484521802e-05, + "loss": 0.0175, + "step": 70830 + }, + { + "epoch": 13.752778991311102, + "grad_norm": 0.16156278550624847, + "learning_rate": 2.150189784415506e-05, + "loss": 0.022, + "step": 70840 + }, + { + "epoch": 13.754720644628902, + "grad_norm": 0.13529421389102936, + "learning_rate": 2.148831332029794e-05, + "loss": 0.0214, + "step": 70850 + }, + { + "epoch": 13.756662297946702, + "grad_norm": 0.10178156197071075, + "learning_rate": 2.147473191443599e-05, + "loss": 0.0172, + "step": 70860 + }, + { + "epoch": 13.7586039512645, + "grad_norm": 0.21320796012878418, + "learning_rate": 2.1461153628054492e-05, + "loss": 0.0178, + "step": 70870 + }, + { + "epoch": 13.760545604582301, + "grad_norm": 0.0911235362291336, + "learning_rate": 2.1447578462638284e-05, + "loss": 0.0212, + "step": 70880 + }, + { + "epoch": 13.762487257900101, + "grad_norm": 0.26986074447631836, + "learning_rate": 2.143400641967201e-05, + "loss": 0.0204, + "step": 70890 + }, + { + "epoch": 13.764428911217902, + "grad_norm": 0.12101934850215912, + "learning_rate": 2.1420437500639823e-05, + "loss": 0.0223, + "step": 70900 + }, + { + "epoch": 13.766370564535702, + "grad_norm": 0.1287418156862259, + "learning_rate": 2.1406871707025646e-05, + "loss": 0.025, + "step": 70910 + }, + { + "epoch": 13.768312217853502, + "grad_norm": 0.21093380451202393, + "learning_rate": 2.1393309040312963e-05, + "loss": 0.0233, + "step": 70920 + }, + { + "epoch": 13.770253871171303, + "grad_norm": 0.13856744766235352, + "learning_rate": 2.137974950198502e-05, + "loss": 0.0202, + "step": 70930 + }, + { + "epoch": 13.772195524489103, + "grad_norm": 0.1505834460258484, + "learning_rate": 2.1366193093524617e-05, + "loss": 0.0193, + "step": 70940 + }, + { + "epoch": 13.774137177806903, + "grad_norm": 0.15093320608139038, + "learning_rate": 2.1352639816414288e-05, + "loss": 0.0252, + "step": 70950 + }, + { + "epoch": 13.776078831124703, + "grad_norm": 0.1715390533208847, + "learning_rate": 2.1339089672136202e-05, + "loss": 0.021, + "step": 70960 + }, + { + "epoch": 13.778020484442504, + "grad_norm": 0.16880668699741364, + "learning_rate": 2.1325542662172193e-05, + "loss": 0.0177, + "step": 70970 + }, + { + "epoch": 13.779962137760302, + "grad_norm": 0.13600316643714905, + "learning_rate": 2.1311998788003712e-05, + "loss": 0.0203, + "step": 70980 + }, + { + "epoch": 13.781903791078102, + "grad_norm": 0.14662286639213562, + "learning_rate": 2.1298458051111907e-05, + "loss": 0.024, + "step": 70990 + }, + { + "epoch": 13.783845444395903, + "grad_norm": 0.224696084856987, + "learning_rate": 2.1284920452977602e-05, + "loss": 0.0268, + "step": 71000 + }, + { + "epoch": 13.785787097713703, + "grad_norm": 0.10508545488119125, + "learning_rate": 2.12713859950812e-05, + "loss": 0.0205, + "step": 71010 + }, + { + "epoch": 13.787728751031503, + "grad_norm": 0.1947932243347168, + "learning_rate": 2.1257854678902832e-05, + "loss": 0.021, + "step": 71020 + }, + { + "epoch": 13.789670404349303, + "grad_norm": 0.1453395038843155, + "learning_rate": 2.124432650592227e-05, + "loss": 0.0294, + "step": 71030 + }, + { + "epoch": 13.791612057667104, + "grad_norm": 0.143920436501503, + "learning_rate": 2.1230801477618937e-05, + "loss": 0.0155, + "step": 71040 + }, + { + "epoch": 13.793553710984904, + "grad_norm": 0.19512994587421417, + "learning_rate": 2.121727959547189e-05, + "loss": 0.0216, + "step": 71050 + }, + { + "epoch": 13.795495364302704, + "grad_norm": 0.1657339334487915, + "learning_rate": 2.1203760860959886e-05, + "loss": 0.0187, + "step": 71060 + }, + { + "epoch": 13.797437017620505, + "grad_norm": 0.233073890209198, + "learning_rate": 2.119024527556127e-05, + "loss": 0.0198, + "step": 71070 + }, + { + "epoch": 13.799378670938303, + "grad_norm": 0.16176851093769073, + "learning_rate": 2.117673284075415e-05, + "loss": 0.021, + "step": 71080 + }, + { + "epoch": 13.801320324256103, + "grad_norm": 0.15276704728603363, + "learning_rate": 2.1163223558016165e-05, + "loss": 0.0253, + "step": 71090 + }, + { + "epoch": 13.803261977573904, + "grad_norm": 0.18483982980251312, + "learning_rate": 2.1149717428824722e-05, + "loss": 0.0169, + "step": 71100 + }, + { + "epoch": 13.805203630891704, + "grad_norm": 0.17131537199020386, + "learning_rate": 2.1136214454656784e-05, + "loss": 0.0235, + "step": 71110 + }, + { + "epoch": 13.807145284209504, + "grad_norm": 0.11901730298995972, + "learning_rate": 2.1122714636989056e-05, + "loss": 0.0231, + "step": 71120 + }, + { + "epoch": 13.809086937527304, + "grad_norm": 0.22831203043460846, + "learning_rate": 2.1109217977297823e-05, + "loss": 0.0267, + "step": 71130 + }, + { + "epoch": 13.811028590845105, + "grad_norm": 0.17784282565116882, + "learning_rate": 2.1095724477059077e-05, + "loss": 0.0211, + "step": 71140 + }, + { + "epoch": 13.812970244162905, + "grad_norm": 0.19595621526241302, + "learning_rate": 2.1082234137748434e-05, + "loss": 0.0244, + "step": 71150 + }, + { + "epoch": 13.814911897480705, + "grad_norm": 0.09897895157337189, + "learning_rate": 2.106874696084122e-05, + "loss": 0.0243, + "step": 71160 + }, + { + "epoch": 13.816853550798506, + "grad_norm": 0.20668955147266388, + "learning_rate": 2.1055262947812315e-05, + "loss": 0.0256, + "step": 71170 + }, + { + "epoch": 13.818795204116306, + "grad_norm": 0.19746892154216766, + "learning_rate": 2.104178210013636e-05, + "loss": 0.0225, + "step": 71180 + }, + { + "epoch": 13.820736857434106, + "grad_norm": 0.1657024621963501, + "learning_rate": 2.1028304419287554e-05, + "loss": 0.022, + "step": 71190 + }, + { + "epoch": 13.822678510751905, + "grad_norm": 0.28388798236846924, + "learning_rate": 2.101482990673982e-05, + "loss": 0.0246, + "step": 71200 + }, + { + "epoch": 13.824620164069705, + "grad_norm": 0.172817200422287, + "learning_rate": 2.1001358563966707e-05, + "loss": 0.0198, + "step": 71210 + }, + { + "epoch": 13.826561817387505, + "grad_norm": 0.3641606569290161, + "learning_rate": 2.098789039244144e-05, + "loss": 0.0188, + "step": 71220 + }, + { + "epoch": 13.828503470705305, + "grad_norm": 0.2074432373046875, + "learning_rate": 2.097442539363685e-05, + "loss": 0.0283, + "step": 71230 + }, + { + "epoch": 13.830445124023106, + "grad_norm": 0.17891769111156464, + "learning_rate": 2.0960963569025448e-05, + "loss": 0.0258, + "step": 71240 + }, + { + "epoch": 13.832386777340906, + "grad_norm": 0.18410924077033997, + "learning_rate": 2.0947504920079435e-05, + "loss": 0.0202, + "step": 71250 + }, + { + "epoch": 13.834328430658706, + "grad_norm": 0.20747129619121552, + "learning_rate": 2.093404944827059e-05, + "loss": 0.0218, + "step": 71260 + }, + { + "epoch": 13.836270083976506, + "grad_norm": 0.22612152993679047, + "learning_rate": 2.0920597155070394e-05, + "loss": 0.0209, + "step": 71270 + }, + { + "epoch": 13.838211737294307, + "grad_norm": 0.19575221836566925, + "learning_rate": 2.090714804194997e-05, + "loss": 0.0225, + "step": 71280 + }, + { + "epoch": 13.840153390612107, + "grad_norm": 0.17620721459388733, + "learning_rate": 2.0893702110380114e-05, + "loss": 0.0323, + "step": 71290 + }, + { + "epoch": 13.842095043929906, + "grad_norm": 0.14829488098621368, + "learning_rate": 2.0880259361831216e-05, + "loss": 0.0208, + "step": 71300 + }, + { + "epoch": 13.844036697247706, + "grad_norm": 0.2665937840938568, + "learning_rate": 2.0866819797773397e-05, + "loss": 0.0225, + "step": 71310 + }, + { + "epoch": 13.845978350565506, + "grad_norm": 0.26306867599487305, + "learning_rate": 2.085338341967634e-05, + "loss": 0.0273, + "step": 71320 + }, + { + "epoch": 13.847920003883306, + "grad_norm": 0.1309150755405426, + "learning_rate": 2.083995022900946e-05, + "loss": 0.0179, + "step": 71330 + }, + { + "epoch": 13.849861657201107, + "grad_norm": 0.14186601340770721, + "learning_rate": 2.0826520227241768e-05, + "loss": 0.0215, + "step": 71340 + }, + { + "epoch": 13.851803310518907, + "grad_norm": 0.17224878072738647, + "learning_rate": 2.081309341584199e-05, + "loss": 0.0244, + "step": 71350 + }, + { + "epoch": 13.853744963836707, + "grad_norm": 0.18447718024253845, + "learning_rate": 2.0799669796278413e-05, + "loss": 0.0213, + "step": 71360 + }, + { + "epoch": 13.855686617154507, + "grad_norm": 0.2085709124803543, + "learning_rate": 2.078624937001906e-05, + "loss": 0.0231, + "step": 71370 + }, + { + "epoch": 13.857628270472308, + "grad_norm": 0.2758846879005432, + "learning_rate": 2.0772832138531544e-05, + "loss": 0.0256, + "step": 71380 + }, + { + "epoch": 13.859569923790108, + "grad_norm": 0.08890267461538315, + "learning_rate": 2.0759418103283158e-05, + "loss": 0.0213, + "step": 71390 + }, + { + "epoch": 13.861511577107908, + "grad_norm": 0.18356719613075256, + "learning_rate": 2.0746007265740846e-05, + "loss": 0.0257, + "step": 71400 + }, + { + "epoch": 13.863453230425707, + "grad_norm": 0.21782174706459045, + "learning_rate": 2.0732599627371215e-05, + "loss": 0.0236, + "step": 71410 + }, + { + "epoch": 13.865394883743507, + "grad_norm": 0.20004960894584656, + "learning_rate": 2.0719195189640468e-05, + "loss": 0.0253, + "step": 71420 + }, + { + "epoch": 13.867336537061307, + "grad_norm": 0.11261799186468124, + "learning_rate": 2.0705793954014528e-05, + "loss": 0.0238, + "step": 71430 + }, + { + "epoch": 13.869278190379108, + "grad_norm": 0.21699441969394684, + "learning_rate": 2.06923959219589e-05, + "loss": 0.0203, + "step": 71440 + }, + { + "epoch": 13.871219843696908, + "grad_norm": 0.1432344615459442, + "learning_rate": 2.0679001094938793e-05, + "loss": 0.0267, + "step": 71450 + }, + { + "epoch": 13.873161497014708, + "grad_norm": 0.09898660331964493, + "learning_rate": 2.066560947441904e-05, + "loss": 0.019, + "step": 71460 + }, + { + "epoch": 13.875103150332508, + "grad_norm": 0.16489244997501373, + "learning_rate": 2.0652221061864152e-05, + "loss": 0.0243, + "step": 71470 + }, + { + "epoch": 13.877044803650309, + "grad_norm": 0.24068081378936768, + "learning_rate": 2.0638835858738226e-05, + "loss": 0.0278, + "step": 71480 + }, + { + "epoch": 13.878986456968109, + "grad_norm": 0.23365169763565063, + "learning_rate": 2.0625453866505068e-05, + "loss": 0.0178, + "step": 71490 + }, + { + "epoch": 13.88092811028591, + "grad_norm": 0.15587638318538666, + "learning_rate": 2.0612075086628123e-05, + "loss": 0.0181, + "step": 71500 + }, + { + "epoch": 13.882869763603708, + "grad_norm": 0.1563904881477356, + "learning_rate": 2.059869952057045e-05, + "loss": 0.0236, + "step": 71510 + }, + { + "epoch": 13.884811416921508, + "grad_norm": 0.33308884501457214, + "learning_rate": 2.0585327169794793e-05, + "loss": 0.0243, + "step": 71520 + }, + { + "epoch": 13.886753070239308, + "grad_norm": 0.4473436176776886, + "learning_rate": 2.0571958035763534e-05, + "loss": 0.0278, + "step": 71530 + }, + { + "epoch": 13.888694723557109, + "grad_norm": 0.24789167940616608, + "learning_rate": 2.0558592119938715e-05, + "loss": 0.0225, + "step": 71540 + }, + { + "epoch": 13.890636376874909, + "grad_norm": 0.1831316202878952, + "learning_rate": 2.054522942378198e-05, + "loss": 0.0249, + "step": 71550 + }, + { + "epoch": 13.892578030192709, + "grad_norm": 0.163544163107872, + "learning_rate": 2.053186994875469e-05, + "loss": 0.023, + "step": 71560 + }, + { + "epoch": 13.89451968351051, + "grad_norm": 0.15818729996681213, + "learning_rate": 2.0518513696317775e-05, + "loss": 0.0212, + "step": 71570 + }, + { + "epoch": 13.89646133682831, + "grad_norm": 0.09673548489809036, + "learning_rate": 2.050516066793188e-05, + "loss": 0.0212, + "step": 71580 + }, + { + "epoch": 13.89840299014611, + "grad_norm": 0.23692788183689117, + "learning_rate": 2.0491810865057276e-05, + "loss": 0.0211, + "step": 71590 + }, + { + "epoch": 13.90034464346391, + "grad_norm": 0.4280378818511963, + "learning_rate": 2.0478464289153882e-05, + "loss": 0.0187, + "step": 71600 + }, + { + "epoch": 13.90228629678171, + "grad_norm": 0.11166799813508987, + "learning_rate": 2.0465120941681232e-05, + "loss": 0.0228, + "step": 71610 + }, + { + "epoch": 13.904227950099509, + "grad_norm": 0.11658123135566711, + "learning_rate": 2.0451780824098566e-05, + "loss": 0.0235, + "step": 71620 + }, + { + "epoch": 13.90616960341731, + "grad_norm": 0.295386403799057, + "learning_rate": 2.043844393786471e-05, + "loss": 0.0229, + "step": 71630 + }, + { + "epoch": 13.90811125673511, + "grad_norm": 0.18125483393669128, + "learning_rate": 2.0425110284438193e-05, + "loss": 0.0215, + "step": 71640 + }, + { + "epoch": 13.91005291005291, + "grad_norm": 0.18001365661621094, + "learning_rate": 2.041177986527712e-05, + "loss": 0.0199, + "step": 71650 + }, + { + "epoch": 13.91199456337071, + "grad_norm": 0.18682222068309784, + "learning_rate": 2.0398452681839347e-05, + "loss": 0.0215, + "step": 71660 + }, + { + "epoch": 13.91393621668851, + "grad_norm": 0.2949202060699463, + "learning_rate": 2.0385128735582272e-05, + "loss": 0.0269, + "step": 71670 + }, + { + "epoch": 13.91587787000631, + "grad_norm": 0.16480684280395508, + "learning_rate": 2.0371808027963003e-05, + "loss": 0.0235, + "step": 71680 + }, + { + "epoch": 13.91781952332411, + "grad_norm": 0.14105407893657684, + "learning_rate": 2.035849056043825e-05, + "loss": 0.02, + "step": 71690 + }, + { + "epoch": 13.919761176641911, + "grad_norm": 0.14900898933410645, + "learning_rate": 2.03451763344644e-05, + "loss": 0.0266, + "step": 71700 + }, + { + "epoch": 13.921702829959711, + "grad_norm": 0.14076337218284607, + "learning_rate": 2.0331865351497482e-05, + "loss": 0.0267, + "step": 71710 + }, + { + "epoch": 13.92364448327751, + "grad_norm": 0.49140554666519165, + "learning_rate": 2.0318557612993185e-05, + "loss": 0.0165, + "step": 71720 + }, + { + "epoch": 13.92558613659531, + "grad_norm": 0.23059651255607605, + "learning_rate": 2.0305253120406782e-05, + "loss": 0.0218, + "step": 71730 + }, + { + "epoch": 13.92752778991311, + "grad_norm": 0.14050838351249695, + "learning_rate": 2.029195187519325e-05, + "loss": 0.0185, + "step": 71740 + }, + { + "epoch": 13.92946944323091, + "grad_norm": 0.3365269601345062, + "learning_rate": 2.0278653878807217e-05, + "loss": 0.0213, + "step": 71750 + }, + { + "epoch": 13.931411096548711, + "grad_norm": 0.20610834658145905, + "learning_rate": 2.0265359132702892e-05, + "loss": 0.0252, + "step": 71760 + }, + { + "epoch": 13.933352749866511, + "grad_norm": 0.17740380764007568, + "learning_rate": 2.0252067638334182e-05, + "loss": 0.0226, + "step": 71770 + }, + { + "epoch": 13.935294403184312, + "grad_norm": 0.14976397156715393, + "learning_rate": 2.023877939715464e-05, + "loss": 0.0244, + "step": 71780 + }, + { + "epoch": 13.937236056502112, + "grad_norm": 0.20128227770328522, + "learning_rate": 2.0225494410617448e-05, + "loss": 0.0181, + "step": 71790 + }, + { + "epoch": 13.939177709819912, + "grad_norm": 0.2173001766204834, + "learning_rate": 2.02122126801754e-05, + "loss": 0.021, + "step": 71800 + }, + { + "epoch": 13.941119363137712, + "grad_norm": 0.18680784106254578, + "learning_rate": 2.0198934207281018e-05, + "loss": 0.0291, + "step": 71810 + }, + { + "epoch": 13.943061016455513, + "grad_norm": 0.16495539247989655, + "learning_rate": 2.0185658993386357e-05, + "loss": 0.0211, + "step": 71820 + }, + { + "epoch": 13.945002669773311, + "grad_norm": 0.1582752764225006, + "learning_rate": 2.0172387039943226e-05, + "loss": 0.0223, + "step": 71830 + }, + { + "epoch": 13.946944323091111, + "grad_norm": 0.2906574606895447, + "learning_rate": 2.0159118348402968e-05, + "loss": 0.0242, + "step": 71840 + }, + { + "epoch": 13.948885976408912, + "grad_norm": 0.12757839262485504, + "learning_rate": 2.0145852920216697e-05, + "loss": 0.0195, + "step": 71850 + }, + { + "epoch": 13.950827629726712, + "grad_norm": 0.16517750918865204, + "learning_rate": 2.0132590756835045e-05, + "loss": 0.0237, + "step": 71860 + }, + { + "epoch": 13.952769283044512, + "grad_norm": 0.10146144032478333, + "learning_rate": 2.0119331859708386e-05, + "loss": 0.0202, + "step": 71870 + }, + { + "epoch": 13.954710936362313, + "grad_norm": 0.12258946150541306, + "learning_rate": 2.0106076230286654e-05, + "loss": 0.0282, + "step": 71880 + }, + { + "epoch": 13.956652589680113, + "grad_norm": 0.12853123247623444, + "learning_rate": 2.0092823870019488e-05, + "loss": 0.0225, + "step": 71890 + }, + { + "epoch": 13.958594242997913, + "grad_norm": 0.11548801511526108, + "learning_rate": 2.0079574780356115e-05, + "loss": 0.0199, + "step": 71900 + }, + { + "epoch": 13.960535896315713, + "grad_norm": 0.08169446885585785, + "learning_rate": 2.0066328962745486e-05, + "loss": 0.0221, + "step": 71910 + }, + { + "epoch": 13.962477549633514, + "grad_norm": 0.12638747692108154, + "learning_rate": 2.00530864186361e-05, + "loss": 0.0173, + "step": 71920 + }, + { + "epoch": 13.964419202951312, + "grad_norm": 0.20420262217521667, + "learning_rate": 2.003984714947618e-05, + "loss": 0.0226, + "step": 71930 + }, + { + "epoch": 13.966360856269112, + "grad_norm": 0.1335720717906952, + "learning_rate": 2.0026611156713515e-05, + "loss": 0.0202, + "step": 71940 + }, + { + "epoch": 13.968302509586913, + "grad_norm": 0.12610527873039246, + "learning_rate": 2.0013378441795584e-05, + "loss": 0.021, + "step": 71950 + }, + { + "epoch": 13.970244162904713, + "grad_norm": 0.2689569890499115, + "learning_rate": 2.0000149006169506e-05, + "loss": 0.0237, + "step": 71960 + }, + { + "epoch": 13.972185816222513, + "grad_norm": 0.2109922170639038, + "learning_rate": 1.9986922851282048e-05, + "loss": 0.0241, + "step": 71970 + }, + { + "epoch": 13.974127469540313, + "grad_norm": 0.19833366572856903, + "learning_rate": 1.997369997857956e-05, + "loss": 0.0206, + "step": 71980 + }, + { + "epoch": 13.976069122858114, + "grad_norm": 0.1693352609872818, + "learning_rate": 1.9960480389508103e-05, + "loss": 0.0238, + "step": 71990 + }, + { + "epoch": 13.978010776175914, + "grad_norm": 0.15186135470867157, + "learning_rate": 1.9947264085513363e-05, + "loss": 0.0186, + "step": 72000 + }, + { + "epoch": 13.979952429493714, + "grad_norm": 0.22318141162395477, + "learning_rate": 1.993405106804062e-05, + "loss": 0.0252, + "step": 72010 + }, + { + "epoch": 13.981894082811515, + "grad_norm": 0.254976361989975, + "learning_rate": 1.9920841338534867e-05, + "loss": 0.0225, + "step": 72020 + }, + { + "epoch": 13.983835736129315, + "grad_norm": 0.13239240646362305, + "learning_rate": 1.9907634898440646e-05, + "loss": 0.0266, + "step": 72030 + }, + { + "epoch": 13.985777389447115, + "grad_norm": 0.11307204514741898, + "learning_rate": 1.989443174920227e-05, + "loss": 0.0179, + "step": 72040 + }, + { + "epoch": 13.987719042764914, + "grad_norm": 0.22240102291107178, + "learning_rate": 1.988123189226355e-05, + "loss": 0.0257, + "step": 72050 + }, + { + "epoch": 13.989660696082714, + "grad_norm": 0.2628770172595978, + "learning_rate": 1.986803532906805e-05, + "loss": 0.0267, + "step": 72060 + }, + { + "epoch": 13.991602349400514, + "grad_norm": 0.15042099356651306, + "learning_rate": 1.985484206105889e-05, + "loss": 0.0238, + "step": 72070 + }, + { + "epoch": 13.993544002718314, + "grad_norm": 0.15311910212039948, + "learning_rate": 1.9841652089678887e-05, + "loss": 0.0212, + "step": 72080 + }, + { + "epoch": 13.995485656036115, + "grad_norm": 0.16161218285560608, + "learning_rate": 1.9828465416370435e-05, + "loss": 0.0205, + "step": 72090 + }, + { + "epoch": 13.997427309353915, + "grad_norm": 0.19322848320007324, + "learning_rate": 1.9815282042575685e-05, + "loss": 0.0233, + "step": 72100 + }, + { + "epoch": 13.999368962671715, + "grad_norm": 0.26987695693969727, + "learning_rate": 1.9802101969736285e-05, + "loss": 0.0258, + "step": 72110 + }, + { + "epoch": 14.00116499199068, + "grad_norm": 0.11626899242401123, + "learning_rate": 1.9788925199293633e-05, + "loss": 0.0139, + "step": 72120 + }, + { + "epoch": 14.00310664530848, + "grad_norm": 0.22027084231376648, + "learning_rate": 1.977575173268868e-05, + "loss": 0.0255, + "step": 72130 + }, + { + "epoch": 14.00504829862628, + "grad_norm": 0.16210290789604187, + "learning_rate": 1.9762581571362094e-05, + "loss": 0.0247, + "step": 72140 + }, + { + "epoch": 14.00698995194408, + "grad_norm": 0.16161392629146576, + "learning_rate": 1.9749414716754112e-05, + "loss": 0.0246, + "step": 72150 + }, + { + "epoch": 14.00893160526188, + "grad_norm": 0.14959542453289032, + "learning_rate": 1.973625117030465e-05, + "loss": 0.0188, + "step": 72160 + }, + { + "epoch": 14.01087325857968, + "grad_norm": 0.12666679918766022, + "learning_rate": 1.972309093345326e-05, + "loss": 0.0209, + "step": 72170 + }, + { + "epoch": 14.01281491189748, + "grad_norm": 0.2031024545431137, + "learning_rate": 1.970993400763915e-05, + "loss": 0.0296, + "step": 72180 + }, + { + "epoch": 14.014756565215281, + "grad_norm": 0.1868605762720108, + "learning_rate": 1.969678039430109e-05, + "loss": 0.027, + "step": 72190 + }, + { + "epoch": 14.016698218533081, + "grad_norm": 0.18345539271831512, + "learning_rate": 1.9683630094877564e-05, + "loss": 0.0243, + "step": 72200 + }, + { + "epoch": 14.018639871850882, + "grad_norm": 0.12832824885845184, + "learning_rate": 1.9670483110806696e-05, + "loss": 0.0201, + "step": 72210 + }, + { + "epoch": 14.020581525168682, + "grad_norm": 0.16468381881713867, + "learning_rate": 1.9657339443526172e-05, + "loss": 0.0219, + "step": 72220 + }, + { + "epoch": 14.022523178486482, + "grad_norm": 0.1525866836309433, + "learning_rate": 1.964419909447339e-05, + "loss": 0.0216, + "step": 72230 + }, + { + "epoch": 14.02446483180428, + "grad_norm": 0.15418152511119843, + "learning_rate": 1.963106206508536e-05, + "loss": 0.0199, + "step": 72240 + }, + { + "epoch": 14.02640648512208, + "grad_norm": 0.15182895958423615, + "learning_rate": 1.9617928356798738e-05, + "loss": 0.0205, + "step": 72250 + }, + { + "epoch": 14.028348138439881, + "grad_norm": 0.15054598450660706, + "learning_rate": 1.960479797104977e-05, + "loss": 0.0211, + "step": 72260 + }, + { + "epoch": 14.030289791757681, + "grad_norm": 0.15659303963184357, + "learning_rate": 1.9591670909274418e-05, + "loss": 0.0246, + "step": 72270 + }, + { + "epoch": 14.032231445075482, + "grad_norm": 0.20461595058441162, + "learning_rate": 1.9578547172908184e-05, + "loss": 0.0227, + "step": 72280 + }, + { + "epoch": 14.034173098393282, + "grad_norm": 0.11171772330999374, + "learning_rate": 1.9565426763386325e-05, + "loss": 0.0215, + "step": 72290 + }, + { + "epoch": 14.036114751711082, + "grad_norm": 0.288463830947876, + "learning_rate": 1.9552309682143622e-05, + "loss": 0.0253, + "step": 72300 + }, + { + "epoch": 14.038056405028883, + "grad_norm": 0.14763104915618896, + "learning_rate": 1.9539195930614574e-05, + "loss": 0.0224, + "step": 72310 + }, + { + "epoch": 14.039998058346683, + "grad_norm": 0.1708507239818573, + "learning_rate": 1.952608551023324e-05, + "loss": 0.0214, + "step": 72320 + }, + { + "epoch": 14.041939711664483, + "grad_norm": 0.23145566880702972, + "learning_rate": 1.9512978422433397e-05, + "loss": 0.0232, + "step": 72330 + }, + { + "epoch": 14.043881364982282, + "grad_norm": 0.15786674618721008, + "learning_rate": 1.9499874668648376e-05, + "loss": 0.0254, + "step": 72340 + }, + { + "epoch": 14.045823018300082, + "grad_norm": 0.31304436922073364, + "learning_rate": 1.9486774250311213e-05, + "loss": 0.0257, + "step": 72350 + }, + { + "epoch": 14.047764671617882, + "grad_norm": 0.24153809249401093, + "learning_rate": 1.9473677168854537e-05, + "loss": 0.0236, + "step": 72360 + }, + { + "epoch": 14.049706324935682, + "grad_norm": 0.16806690394878387, + "learning_rate": 1.946058342571065e-05, + "loss": 0.0267, + "step": 72370 + }, + { + "epoch": 14.051647978253483, + "grad_norm": 0.20074574649333954, + "learning_rate": 1.9447493022311424e-05, + "loss": 0.0185, + "step": 72380 + }, + { + "epoch": 14.053589631571283, + "grad_norm": 0.19564469158649445, + "learning_rate": 1.9434405960088442e-05, + "loss": 0.0243, + "step": 72390 + }, + { + "epoch": 14.055531284889083, + "grad_norm": 0.16997146606445312, + "learning_rate": 1.9421322240472857e-05, + "loss": 0.0215, + "step": 72400 + }, + { + "epoch": 14.057472938206883, + "grad_norm": 0.12547704577445984, + "learning_rate": 1.9408241864895494e-05, + "loss": 0.0187, + "step": 72410 + }, + { + "epoch": 14.059414591524684, + "grad_norm": 0.10827334225177765, + "learning_rate": 1.939516483478681e-05, + "loss": 0.0148, + "step": 72420 + }, + { + "epoch": 14.061356244842484, + "grad_norm": 0.29219281673431396, + "learning_rate": 1.9382091151576902e-05, + "loss": 0.0291, + "step": 72430 + }, + { + "epoch": 14.063297898160284, + "grad_norm": 0.17565888166427612, + "learning_rate": 1.9369020816695454e-05, + "loss": 0.0201, + "step": 72440 + }, + { + "epoch": 14.065239551478083, + "grad_norm": 0.201976478099823, + "learning_rate": 1.935595383157184e-05, + "loss": 0.0235, + "step": 72450 + }, + { + "epoch": 14.067181204795883, + "grad_norm": 0.17244327068328857, + "learning_rate": 1.9342890197635057e-05, + "loss": 0.024, + "step": 72460 + }, + { + "epoch": 14.069122858113683, + "grad_norm": 0.20186738669872284, + "learning_rate": 1.9329829916313685e-05, + "loss": 0.0215, + "step": 72470 + }, + { + "epoch": 14.071064511431484, + "grad_norm": 0.17541401088237762, + "learning_rate": 1.9316772989036013e-05, + "loss": 0.0236, + "step": 72480 + }, + { + "epoch": 14.073006164749284, + "grad_norm": 0.2216850370168686, + "learning_rate": 1.930371941722991e-05, + "loss": 0.0262, + "step": 72490 + }, + { + "epoch": 14.074947818067084, + "grad_norm": 0.14884787797927856, + "learning_rate": 1.929066920232292e-05, + "loss": 0.0183, + "step": 72500 + }, + { + "epoch": 14.076889471384884, + "grad_norm": 0.09531791508197784, + "learning_rate": 1.9277622345742152e-05, + "loss": 0.0181, + "step": 72510 + }, + { + "epoch": 14.078831124702685, + "grad_norm": 0.20396259427070618, + "learning_rate": 1.926457884891444e-05, + "loss": 0.0205, + "step": 72520 + }, + { + "epoch": 14.080772778020485, + "grad_norm": 0.18681570887565613, + "learning_rate": 1.9251538713266147e-05, + "loss": 0.0244, + "step": 72530 + }, + { + "epoch": 14.082714431338285, + "grad_norm": 0.26675570011138916, + "learning_rate": 1.9238501940223358e-05, + "loss": 0.0276, + "step": 72540 + }, + { + "epoch": 14.084656084656086, + "grad_norm": 0.17083466053009033, + "learning_rate": 1.922546853121175e-05, + "loss": 0.0219, + "step": 72550 + }, + { + "epoch": 14.086597737973884, + "grad_norm": 0.2630194127559662, + "learning_rate": 1.9212438487656647e-05, + "loss": 0.0296, + "step": 72560 + }, + { + "epoch": 14.088539391291684, + "grad_norm": 0.14784091711044312, + "learning_rate": 1.9199411810982965e-05, + "loss": 0.0322, + "step": 72570 + }, + { + "epoch": 14.090481044609485, + "grad_norm": 0.23500242829322815, + "learning_rate": 1.918638850261532e-05, + "loss": 0.0223, + "step": 72580 + }, + { + "epoch": 14.092422697927285, + "grad_norm": 0.13199089467525482, + "learning_rate": 1.9173368563977885e-05, + "loss": 0.0223, + "step": 72590 + }, + { + "epoch": 14.094364351245085, + "grad_norm": 0.11504950374364853, + "learning_rate": 1.9160351996494513e-05, + "loss": 0.0201, + "step": 72600 + }, + { + "epoch": 14.096306004562885, + "grad_norm": 0.1804160624742508, + "learning_rate": 1.9147338801588677e-05, + "loss": 0.0221, + "step": 72610 + }, + { + "epoch": 14.098247657880686, + "grad_norm": 0.18362213671207428, + "learning_rate": 1.913432898068351e-05, + "loss": 0.023, + "step": 72620 + }, + { + "epoch": 14.100189311198486, + "grad_norm": 0.17890781164169312, + "learning_rate": 1.912132253520169e-05, + "loss": 0.0202, + "step": 72630 + }, + { + "epoch": 14.102130964516286, + "grad_norm": 0.16482937335968018, + "learning_rate": 1.910831946656564e-05, + "loss": 0.0236, + "step": 72640 + }, + { + "epoch": 14.104072617834087, + "grad_norm": 0.11576279252767563, + "learning_rate": 1.909531977619731e-05, + "loss": 0.0233, + "step": 72650 + }, + { + "epoch": 14.106014271151885, + "grad_norm": 0.22395826876163483, + "learning_rate": 1.908232346551834e-05, + "loss": 0.0262, + "step": 72660 + }, + { + "epoch": 14.107955924469685, + "grad_norm": 0.1477569341659546, + "learning_rate": 1.9069330535949993e-05, + "loss": 0.0258, + "step": 72670 + }, + { + "epoch": 14.109897577787486, + "grad_norm": 0.263710081577301, + "learning_rate": 1.9056340988913167e-05, + "loss": 0.0303, + "step": 72680 + }, + { + "epoch": 14.111839231105286, + "grad_norm": 0.20139513909816742, + "learning_rate": 1.9043354825828357e-05, + "loss": 0.0238, + "step": 72690 + }, + { + "epoch": 14.113780884423086, + "grad_norm": 0.11503886431455612, + "learning_rate": 1.9030372048115724e-05, + "loss": 0.0247, + "step": 72700 + }, + { + "epoch": 14.115722537740886, + "grad_norm": 0.27213531732559204, + "learning_rate": 1.9017392657195056e-05, + "loss": 0.0232, + "step": 72710 + }, + { + "epoch": 14.117664191058687, + "grad_norm": 0.13697516918182373, + "learning_rate": 1.9004416654485724e-05, + "loss": 0.0243, + "step": 72720 + }, + { + "epoch": 14.119605844376487, + "grad_norm": 0.1700529158115387, + "learning_rate": 1.8991444041406785e-05, + "loss": 0.021, + "step": 72730 + }, + { + "epoch": 14.121547497694287, + "grad_norm": 0.15851137042045593, + "learning_rate": 1.8978474819376912e-05, + "loss": 0.0257, + "step": 72740 + }, + { + "epoch": 14.123489151012087, + "grad_norm": 0.19767336547374725, + "learning_rate": 1.8965508989814406e-05, + "loss": 0.0268, + "step": 72750 + }, + { + "epoch": 14.125430804329888, + "grad_norm": 0.12335030734539032, + "learning_rate": 1.8952546554137158e-05, + "loss": 0.0193, + "step": 72760 + }, + { + "epoch": 14.127372457647686, + "grad_norm": 0.2497037798166275, + "learning_rate": 1.8939587513762757e-05, + "loss": 0.0234, + "step": 72770 + }, + { + "epoch": 14.129314110965487, + "grad_norm": 0.12259945273399353, + "learning_rate": 1.892663187010835e-05, + "loss": 0.0191, + "step": 72780 + }, + { + "epoch": 14.131255764283287, + "grad_norm": 0.23721946775913239, + "learning_rate": 1.891367962459077e-05, + "loss": 0.022, + "step": 72790 + }, + { + "epoch": 14.133197417601087, + "grad_norm": 0.19985243678092957, + "learning_rate": 1.890073077862644e-05, + "loss": 0.0201, + "step": 72800 + }, + { + "epoch": 14.135139070918887, + "grad_norm": 0.31691238284111023, + "learning_rate": 1.888778533363146e-05, + "loss": 0.0216, + "step": 72810 + }, + { + "epoch": 14.137080724236688, + "grad_norm": 0.327632874250412, + "learning_rate": 1.8874843291021478e-05, + "loss": 0.0216, + "step": 72820 + }, + { + "epoch": 14.139022377554488, + "grad_norm": 0.08533140271902084, + "learning_rate": 1.8861904652211853e-05, + "loss": 0.0217, + "step": 72830 + }, + { + "epoch": 14.140964030872288, + "grad_norm": 0.2819227874279022, + "learning_rate": 1.8848969418617507e-05, + "loss": 0.0281, + "step": 72840 + }, + { + "epoch": 14.142905684190088, + "grad_norm": 0.18646736443042755, + "learning_rate": 1.8836037591653044e-05, + "loss": 0.0216, + "step": 72850 + }, + { + "epoch": 14.144847337507889, + "grad_norm": 0.2908288538455963, + "learning_rate": 1.882310917273262e-05, + "loss": 0.0273, + "step": 72860 + }, + { + "epoch": 14.146788990825687, + "grad_norm": 0.24558530747890472, + "learning_rate": 1.881018416327014e-05, + "loss": 0.025, + "step": 72870 + }, + { + "epoch": 14.148730644143487, + "grad_norm": 0.16636472940444946, + "learning_rate": 1.8797262564679014e-05, + "loss": 0.023, + "step": 72880 + }, + { + "epoch": 14.150672297461288, + "grad_norm": 0.20203718543052673, + "learning_rate": 1.8784344378372355e-05, + "loss": 0.0242, + "step": 72890 + }, + { + "epoch": 14.152613950779088, + "grad_norm": 0.10372552275657654, + "learning_rate": 1.8771429605762843e-05, + "loss": 0.0238, + "step": 72900 + }, + { + "epoch": 14.154555604096888, + "grad_norm": 0.1916566640138626, + "learning_rate": 1.8758518248262835e-05, + "loss": 0.0195, + "step": 72910 + }, + { + "epoch": 14.156497257414689, + "grad_norm": 0.19946818053722382, + "learning_rate": 1.8745610307284305e-05, + "loss": 0.0228, + "step": 72920 + }, + { + "epoch": 14.158438910732489, + "grad_norm": 0.22151029109954834, + "learning_rate": 1.8732705784238862e-05, + "loss": 0.0264, + "step": 72930 + }, + { + "epoch": 14.160380564050289, + "grad_norm": 0.23542894423007965, + "learning_rate": 1.8719804680537682e-05, + "loss": 0.0226, + "step": 72940 + }, + { + "epoch": 14.16232221736809, + "grad_norm": 0.13524971902370453, + "learning_rate": 1.870690699759164e-05, + "loss": 0.0229, + "step": 72950 + }, + { + "epoch": 14.16426387068589, + "grad_norm": 0.1478710025548935, + "learning_rate": 1.8694012736811213e-05, + "loss": 0.0188, + "step": 72960 + }, + { + "epoch": 14.16620552400369, + "grad_norm": 0.12867730855941772, + "learning_rate": 1.8681121899606463e-05, + "loss": 0.027, + "step": 72970 + }, + { + "epoch": 14.168147177321488, + "grad_norm": 0.2083105593919754, + "learning_rate": 1.8668234487387144e-05, + "loss": 0.0203, + "step": 72980 + }, + { + "epoch": 14.170088830639289, + "grad_norm": 0.18606127798557281, + "learning_rate": 1.8655350501562592e-05, + "loss": 0.0221, + "step": 72990 + }, + { + "epoch": 14.172030483957089, + "grad_norm": 0.09017699956893921, + "learning_rate": 1.8642469943541796e-05, + "loss": 0.0257, + "step": 73000 + }, + { + "epoch": 14.17397213727489, + "grad_norm": 0.2781207263469696, + "learning_rate": 1.8629592814733328e-05, + "loss": 0.0261, + "step": 73010 + }, + { + "epoch": 14.17591379059269, + "grad_norm": 0.18428438901901245, + "learning_rate": 1.8616719116545438e-05, + "loss": 0.023, + "step": 73020 + }, + { + "epoch": 14.17785544391049, + "grad_norm": 0.16165851056575775, + "learning_rate": 1.8603848850385952e-05, + "loss": 0.028, + "step": 73030 + }, + { + "epoch": 14.17979709722829, + "grad_norm": 0.1752750128507614, + "learning_rate": 1.859098201766236e-05, + "loss": 0.0197, + "step": 73040 + }, + { + "epoch": 14.18173875054609, + "grad_norm": 0.17974603176116943, + "learning_rate": 1.8578118619781726e-05, + "loss": 0.0218, + "step": 73050 + }, + { + "epoch": 14.18368040386389, + "grad_norm": 0.15809667110443115, + "learning_rate": 1.856525865815083e-05, + "loss": 0.0189, + "step": 73060 + }, + { + "epoch": 14.185622057181691, + "grad_norm": 0.18209601938724518, + "learning_rate": 1.8552402134175966e-05, + "loss": 0.0228, + "step": 73070 + }, + { + "epoch": 14.187563710499491, + "grad_norm": 0.1516064703464508, + "learning_rate": 1.8539549049263143e-05, + "loss": 0.0282, + "step": 73080 + }, + { + "epoch": 14.18950536381729, + "grad_norm": 0.15256839990615845, + "learning_rate": 1.8526699404817917e-05, + "loss": 0.0196, + "step": 73090 + }, + { + "epoch": 14.19144701713509, + "grad_norm": 0.13844755291938782, + "learning_rate": 1.8513853202245546e-05, + "loss": 0.0211, + "step": 73100 + }, + { + "epoch": 14.19338867045289, + "grad_norm": 0.19285240769386292, + "learning_rate": 1.850101044295081e-05, + "loss": 0.0187, + "step": 73110 + }, + { + "epoch": 14.19533032377069, + "grad_norm": 0.33836475014686584, + "learning_rate": 1.848817112833825e-05, + "loss": 0.0313, + "step": 73120 + }, + { + "epoch": 14.19727197708849, + "grad_norm": 0.159011572599411, + "learning_rate": 1.8475335259811905e-05, + "loss": 0.0198, + "step": 73130 + }, + { + "epoch": 14.199213630406291, + "grad_norm": 0.17512203752994537, + "learning_rate": 1.8462502838775508e-05, + "loss": 0.0242, + "step": 73140 + }, + { + "epoch": 14.201155283724091, + "grad_norm": 0.16246412694454193, + "learning_rate": 1.8449673866632378e-05, + "loss": 0.0213, + "step": 73150 + }, + { + "epoch": 14.203096937041892, + "grad_norm": 0.22236797213554382, + "learning_rate": 1.8436848344785473e-05, + "loss": 0.0183, + "step": 73160 + }, + { + "epoch": 14.205038590359692, + "grad_norm": 0.2045494019985199, + "learning_rate": 1.8424026274637378e-05, + "loss": 0.0226, + "step": 73170 + }, + { + "epoch": 14.206980243677492, + "grad_norm": 0.32630616426467896, + "learning_rate": 1.8411207657590312e-05, + "loss": 0.0274, + "step": 73180 + }, + { + "epoch": 14.20892189699529, + "grad_norm": 0.20488877594470978, + "learning_rate": 1.839839249504607e-05, + "loss": 0.0241, + "step": 73190 + }, + { + "epoch": 14.210863550313091, + "grad_norm": 0.2072322815656662, + "learning_rate": 1.8385580788406108e-05, + "loss": 0.021, + "step": 73200 + }, + { + "epoch": 14.212805203630891, + "grad_norm": 0.09828626364469528, + "learning_rate": 1.837277253907152e-05, + "loss": 0.0172, + "step": 73210 + }, + { + "epoch": 14.214746856948691, + "grad_norm": 0.16511893272399902, + "learning_rate": 1.8359967748442957e-05, + "loss": 0.0178, + "step": 73220 + }, + { + "epoch": 14.216688510266492, + "grad_norm": 0.19925443828105927, + "learning_rate": 1.8347166417920774e-05, + "loss": 0.0206, + "step": 73230 + }, + { + "epoch": 14.218630163584292, + "grad_norm": 0.24442742764949799, + "learning_rate": 1.833436854890484e-05, + "loss": 0.0237, + "step": 73240 + }, + { + "epoch": 14.220571816902092, + "grad_norm": 0.11961230635643005, + "learning_rate": 1.8321574142794796e-05, + "loss": 0.0259, + "step": 73250 + }, + { + "epoch": 14.222513470219893, + "grad_norm": 0.21058370172977448, + "learning_rate": 1.8308783200989748e-05, + "loss": 0.0229, + "step": 73260 + }, + { + "epoch": 14.224455123537693, + "grad_norm": 0.14678236842155457, + "learning_rate": 1.829599572488855e-05, + "loss": 0.0213, + "step": 73270 + }, + { + "epoch": 14.226396776855493, + "grad_norm": 0.21616677939891815, + "learning_rate": 1.828321171588957e-05, + "loss": 0.0191, + "step": 73280 + }, + { + "epoch": 14.228338430173293, + "grad_norm": 0.2592501640319824, + "learning_rate": 1.827043117539089e-05, + "loss": 0.0301, + "step": 73290 + }, + { + "epoch": 14.230280083491092, + "grad_norm": 0.17039169371128082, + "learning_rate": 1.8257654104790116e-05, + "loss": 0.0244, + "step": 73300 + }, + { + "epoch": 14.232221736808892, + "grad_norm": 0.19585587084293365, + "learning_rate": 1.82448805054846e-05, + "loss": 0.0259, + "step": 73310 + }, + { + "epoch": 14.234163390126692, + "grad_norm": 0.11977425962686539, + "learning_rate": 1.823211037887119e-05, + "loss": 0.0249, + "step": 73320 + }, + { + "epoch": 14.236105043444493, + "grad_norm": 0.2772769331932068, + "learning_rate": 1.821934372634645e-05, + "loss": 0.0256, + "step": 73330 + }, + { + "epoch": 14.238046696762293, + "grad_norm": 0.2922762632369995, + "learning_rate": 1.8206580549306472e-05, + "loss": 0.0265, + "step": 73340 + }, + { + "epoch": 14.239988350080093, + "grad_norm": 0.2288954257965088, + "learning_rate": 1.819382084914707e-05, + "loss": 0.0199, + "step": 73350 + }, + { + "epoch": 14.241930003397894, + "grad_norm": 0.1684059351682663, + "learning_rate": 1.8181064627263577e-05, + "loss": 0.0273, + "step": 73360 + }, + { + "epoch": 14.243871656715694, + "grad_norm": 0.14267824590206146, + "learning_rate": 1.8168311885051022e-05, + "loss": 0.0239, + "step": 73370 + }, + { + "epoch": 14.245813310033494, + "grad_norm": 0.18256591260433197, + "learning_rate": 1.815556262390402e-05, + "loss": 0.0222, + "step": 73380 + }, + { + "epoch": 14.247754963351294, + "grad_norm": 0.10363206267356873, + "learning_rate": 1.8142816845216826e-05, + "loss": 0.0209, + "step": 73390 + }, + { + "epoch": 14.249696616669095, + "grad_norm": 0.15973789989948273, + "learning_rate": 1.8130074550383276e-05, + "loss": 0.022, + "step": 73400 + }, + { + "epoch": 14.251638269986893, + "grad_norm": 0.21577739715576172, + "learning_rate": 1.8117335740796848e-05, + "loss": 0.0225, + "step": 73410 + }, + { + "epoch": 14.253579923304693, + "grad_norm": 0.1901625692844391, + "learning_rate": 1.810460041785067e-05, + "loss": 0.0254, + "step": 73420 + }, + { + "epoch": 14.255521576622494, + "grad_norm": 0.15024201571941376, + "learning_rate": 1.809186858293742e-05, + "loss": 0.0237, + "step": 73430 + }, + { + "epoch": 14.257463229940294, + "grad_norm": 0.1440981924533844, + "learning_rate": 1.8079140237449448e-05, + "loss": 0.0182, + "step": 73440 + }, + { + "epoch": 14.259404883258094, + "grad_norm": 0.2343718409538269, + "learning_rate": 1.8066415382778707e-05, + "loss": 0.0219, + "step": 73450 + }, + { + "epoch": 14.261346536575894, + "grad_norm": 0.19662034511566162, + "learning_rate": 1.8053694020316786e-05, + "loss": 0.0213, + "step": 73460 + }, + { + "epoch": 14.263288189893695, + "grad_norm": 0.18280433118343353, + "learning_rate": 1.8040976151454837e-05, + "loss": 0.0253, + "step": 73470 + }, + { + "epoch": 14.265229843211495, + "grad_norm": 0.14228950440883636, + "learning_rate": 1.8028261777583704e-05, + "loss": 0.0202, + "step": 73480 + }, + { + "epoch": 14.267171496529295, + "grad_norm": 0.2632450759410858, + "learning_rate": 1.8015550900093758e-05, + "loss": 0.0253, + "step": 73490 + }, + { + "epoch": 14.269113149847096, + "grad_norm": 0.2078668773174286, + "learning_rate": 1.800284352037512e-05, + "loss": 0.0233, + "step": 73500 + }, + { + "epoch": 14.271054803164894, + "grad_norm": 0.24692486226558685, + "learning_rate": 1.799013963981738e-05, + "loss": 0.0295, + "step": 73510 + }, + { + "epoch": 14.272996456482694, + "grad_norm": 0.24690166115760803, + "learning_rate": 1.797743925980987e-05, + "loss": 0.0264, + "step": 73520 + }, + { + "epoch": 14.274938109800495, + "grad_norm": 0.24658674001693726, + "learning_rate": 1.7964742381741435e-05, + "loss": 0.0185, + "step": 73530 + }, + { + "epoch": 14.276879763118295, + "grad_norm": 0.2136848121881485, + "learning_rate": 1.7952049007000627e-05, + "loss": 0.0239, + "step": 73540 + }, + { + "epoch": 14.278821416436095, + "grad_norm": 0.154246985912323, + "learning_rate": 1.7939359136975535e-05, + "loss": 0.0162, + "step": 73550 + }, + { + "epoch": 14.280763069753895, + "grad_norm": 0.11796466261148453, + "learning_rate": 1.7926672773053932e-05, + "loss": 0.023, + "step": 73560 + }, + { + "epoch": 14.282704723071696, + "grad_norm": 0.1480465978384018, + "learning_rate": 1.791398991662317e-05, + "loss": 0.0226, + "step": 73570 + }, + { + "epoch": 14.284646376389496, + "grad_norm": 0.14997969567775726, + "learning_rate": 1.7901310569070244e-05, + "loss": 0.0196, + "step": 73580 + }, + { + "epoch": 14.286588029707296, + "grad_norm": 0.10067664086818695, + "learning_rate": 1.7888634731781724e-05, + "loss": 0.0177, + "step": 73590 + }, + { + "epoch": 14.288529683025097, + "grad_norm": 0.24583160877227783, + "learning_rate": 1.7875962406143844e-05, + "loss": 0.0258, + "step": 73600 + }, + { + "epoch": 14.290471336342897, + "grad_norm": 0.2608129680156708, + "learning_rate": 1.7863293593542403e-05, + "loss": 0.0238, + "step": 73610 + }, + { + "epoch": 14.292412989660695, + "grad_norm": 0.1714700311422348, + "learning_rate": 1.7850628295362853e-05, + "loss": 0.0256, + "step": 73620 + }, + { + "epoch": 14.294354642978496, + "grad_norm": 0.23224619030952454, + "learning_rate": 1.7837966512990263e-05, + "loss": 0.0295, + "step": 73630 + }, + { + "epoch": 14.296296296296296, + "grad_norm": 0.2922183573246002, + "learning_rate": 1.782530824780931e-05, + "loss": 0.0281, + "step": 73640 + }, + { + "epoch": 14.298237949614096, + "grad_norm": 0.18548963963985443, + "learning_rate": 1.7812653501204263e-05, + "loss": 0.0218, + "step": 73650 + }, + { + "epoch": 14.300179602931896, + "grad_norm": 0.26413393020629883, + "learning_rate": 1.780000227455903e-05, + "loss": 0.0227, + "step": 73660 + }, + { + "epoch": 14.302121256249697, + "grad_norm": 0.26166674494743347, + "learning_rate": 1.778735456925716e-05, + "loss": 0.0206, + "step": 73670 + }, + { + "epoch": 14.304062909567497, + "grad_norm": 0.13577285408973694, + "learning_rate": 1.7774710386681737e-05, + "loss": 0.0262, + "step": 73680 + }, + { + "epoch": 14.306004562885297, + "grad_norm": 0.11447370797395706, + "learning_rate": 1.7762069728215545e-05, + "loss": 0.0228, + "step": 73690 + }, + { + "epoch": 14.307946216203097, + "grad_norm": 0.16558881103992462, + "learning_rate": 1.7749432595240932e-05, + "loss": 0.0212, + "step": 73700 + }, + { + "epoch": 14.309887869520898, + "grad_norm": 0.13937066495418549, + "learning_rate": 1.773679898913991e-05, + "loss": 0.028, + "step": 73710 + }, + { + "epoch": 14.311829522838696, + "grad_norm": 0.35891175270080566, + "learning_rate": 1.7724168911294015e-05, + "loss": 0.0253, + "step": 73720 + }, + { + "epoch": 14.313771176156497, + "grad_norm": 0.1651078760623932, + "learning_rate": 1.7711542363084504e-05, + "loss": 0.0164, + "step": 73730 + }, + { + "epoch": 14.315712829474297, + "grad_norm": 0.2429250180721283, + "learning_rate": 1.7698919345892158e-05, + "loss": 0.0211, + "step": 73740 + }, + { + "epoch": 14.317654482792097, + "grad_norm": 0.19416333734989166, + "learning_rate": 1.768629986109743e-05, + "loss": 0.018, + "step": 73750 + }, + { + "epoch": 14.319596136109897, + "grad_norm": 0.10445486009120941, + "learning_rate": 1.7673683910080357e-05, + "loss": 0.0209, + "step": 73760 + }, + { + "epoch": 14.321537789427698, + "grad_norm": 0.11536348611116409, + "learning_rate": 1.7661071494220638e-05, + "loss": 0.0232, + "step": 73770 + }, + { + "epoch": 14.323479442745498, + "grad_norm": 0.19522151350975037, + "learning_rate": 1.7648462614897498e-05, + "loss": 0.0184, + "step": 73780 + }, + { + "epoch": 14.325421096063298, + "grad_norm": 0.2606154680252075, + "learning_rate": 1.7635857273489863e-05, + "loss": 0.0296, + "step": 73790 + }, + { + "epoch": 14.327362749381098, + "grad_norm": 0.1289704293012619, + "learning_rate": 1.7623255471376198e-05, + "loss": 0.0185, + "step": 73800 + }, + { + "epoch": 14.329304402698899, + "grad_norm": 0.1109551265835762, + "learning_rate": 1.7610657209934646e-05, + "loss": 0.0179, + "step": 73810 + }, + { + "epoch": 14.331246056016699, + "grad_norm": 0.06517098844051361, + "learning_rate": 1.7598062490542915e-05, + "loss": 0.0275, + "step": 73820 + }, + { + "epoch": 14.333187709334497, + "grad_norm": 0.2684691548347473, + "learning_rate": 1.7585471314578377e-05, + "loss": 0.0222, + "step": 73830 + }, + { + "epoch": 14.335129362652298, + "grad_norm": 0.2399524748325348, + "learning_rate": 1.7572883683417945e-05, + "loss": 0.0264, + "step": 73840 + }, + { + "epoch": 14.337071015970098, + "grad_norm": 0.18604621291160583, + "learning_rate": 1.7560299598438212e-05, + "loss": 0.0333, + "step": 73850 + }, + { + "epoch": 14.339012669287898, + "grad_norm": 0.2512335479259491, + "learning_rate": 1.7547719061015334e-05, + "loss": 0.0247, + "step": 73860 + }, + { + "epoch": 14.340954322605699, + "grad_norm": 0.22062574326992035, + "learning_rate": 1.75351420725251e-05, + "loss": 0.0235, + "step": 73870 + }, + { + "epoch": 14.342895975923499, + "grad_norm": 0.2653123438358307, + "learning_rate": 1.7522568634342923e-05, + "loss": 0.0211, + "step": 73880 + }, + { + "epoch": 14.3448376292413, + "grad_norm": 0.20441216230392456, + "learning_rate": 1.7509998747843825e-05, + "loss": 0.019, + "step": 73890 + }, + { + "epoch": 14.3467792825591, + "grad_norm": 0.1534610539674759, + "learning_rate": 1.7497432414402403e-05, + "loss": 0.0194, + "step": 73900 + }, + { + "epoch": 14.3487209358769, + "grad_norm": 0.190581277012825, + "learning_rate": 1.7484869635392903e-05, + "loss": 0.0205, + "step": 73910 + }, + { + "epoch": 14.3506625891947, + "grad_norm": 0.16159245371818542, + "learning_rate": 1.747231041218919e-05, + "loss": 0.0186, + "step": 73920 + }, + { + "epoch": 14.3526042425125, + "grad_norm": 0.203901007771492, + "learning_rate": 1.7459754746164692e-05, + "loss": 0.0212, + "step": 73930 + }, + { + "epoch": 14.354545895830299, + "grad_norm": 0.13881602883338928, + "learning_rate": 1.744720263869249e-05, + "loss": 0.019, + "step": 73940 + }, + { + "epoch": 14.356487549148099, + "grad_norm": 0.1369694471359253, + "learning_rate": 1.7434654091145263e-05, + "loss": 0.0208, + "step": 73950 + }, + { + "epoch": 14.3584292024659, + "grad_norm": 0.09445571154356003, + "learning_rate": 1.742210910489532e-05, + "loss": 0.0267, + "step": 73960 + }, + { + "epoch": 14.3603708557837, + "grad_norm": 0.27388590574264526, + "learning_rate": 1.7409567681314532e-05, + "loss": 0.022, + "step": 73970 + }, + { + "epoch": 14.3623125091015, + "grad_norm": 0.1849314421415329, + "learning_rate": 1.7397029821774434e-05, + "loss": 0.0235, + "step": 73980 + }, + { + "epoch": 14.3642541624193, + "grad_norm": 0.20321719348430634, + "learning_rate": 1.7384495527646126e-05, + "loss": 0.0222, + "step": 73990 + }, + { + "epoch": 14.3661958157371, + "grad_norm": 0.17628207802772522, + "learning_rate": 1.737196480030035e-05, + "loss": 0.0282, + "step": 74000 + }, + { + "epoch": 14.3681374690549, + "grad_norm": 0.15283170342445374, + "learning_rate": 1.7359437641107447e-05, + "loss": 0.0182, + "step": 74010 + }, + { + "epoch": 14.370079122372701, + "grad_norm": 0.15388847887516022, + "learning_rate": 1.7346914051437387e-05, + "loss": 0.0189, + "step": 74020 + }, + { + "epoch": 14.372020775690501, + "grad_norm": 0.19734443724155426, + "learning_rate": 1.733439403265969e-05, + "loss": 0.0232, + "step": 74030 + }, + { + "epoch": 14.373962429008301, + "grad_norm": 0.2486744225025177, + "learning_rate": 1.732187758614357e-05, + "loss": 0.0271, + "step": 74040 + }, + { + "epoch": 14.3759040823261, + "grad_norm": 0.1378381848335266, + "learning_rate": 1.730936471325777e-05, + "loss": 0.0276, + "step": 74050 + }, + { + "epoch": 14.3778457356439, + "grad_norm": 0.21796460449695587, + "learning_rate": 1.7296855415370718e-05, + "loss": 0.0266, + "step": 74060 + }, + { + "epoch": 14.3797873889617, + "grad_norm": 0.09021934866905212, + "learning_rate": 1.728434969385035e-05, + "loss": 0.0263, + "step": 74070 + }, + { + "epoch": 14.3817290422795, + "grad_norm": 0.24908611178398132, + "learning_rate": 1.7271847550064346e-05, + "loss": 0.0199, + "step": 74080 + }, + { + "epoch": 14.383670695597301, + "grad_norm": 0.16743715107440948, + "learning_rate": 1.725934898537987e-05, + "loss": 0.0313, + "step": 74090 + }, + { + "epoch": 14.385612348915101, + "grad_norm": 0.2657108008861542, + "learning_rate": 1.7246854001163783e-05, + "loss": 0.0178, + "step": 74100 + }, + { + "epoch": 14.387554002232902, + "grad_norm": 0.15267634391784668, + "learning_rate": 1.7234362598782476e-05, + "loss": 0.0209, + "step": 74110 + }, + { + "epoch": 14.389495655550702, + "grad_norm": 0.12620244920253754, + "learning_rate": 1.722187477960202e-05, + "loss": 0.0221, + "step": 74120 + }, + { + "epoch": 14.391437308868502, + "grad_norm": 0.14317573606967926, + "learning_rate": 1.7209390544988057e-05, + "loss": 0.019, + "step": 74130 + }, + { + "epoch": 14.393378962186302, + "grad_norm": 0.1163363829255104, + "learning_rate": 1.7196909896305856e-05, + "loss": 0.024, + "step": 74140 + }, + { + "epoch": 14.395320615504101, + "grad_norm": 0.12854649126529694, + "learning_rate": 1.718443283492026e-05, + "loss": 0.0224, + "step": 74150 + }, + { + "epoch": 14.397262268821901, + "grad_norm": 0.15308547019958496, + "learning_rate": 1.7171959362195746e-05, + "loss": 0.0256, + "step": 74160 + }, + { + "epoch": 14.399203922139701, + "grad_norm": 0.27067235112190247, + "learning_rate": 1.715948947949642e-05, + "loss": 0.0263, + "step": 74170 + }, + { + "epoch": 14.401145575457502, + "grad_norm": 0.18173478543758392, + "learning_rate": 1.714702318818593e-05, + "loss": 0.026, + "step": 74180 + }, + { + "epoch": 14.403087228775302, + "grad_norm": 0.2598602771759033, + "learning_rate": 1.7134560489627592e-05, + "loss": 0.0251, + "step": 74190 + }, + { + "epoch": 14.405028882093102, + "grad_norm": 0.20965178310871124, + "learning_rate": 1.712210138518431e-05, + "loss": 0.0206, + "step": 74200 + }, + { + "epoch": 14.406970535410903, + "grad_norm": 0.1883600652217865, + "learning_rate": 1.71096458762186e-05, + "loss": 0.0197, + "step": 74210 + }, + { + "epoch": 14.408912188728703, + "grad_norm": 0.16893094778060913, + "learning_rate": 1.7097193964092552e-05, + "loss": 0.0262, + "step": 74220 + }, + { + "epoch": 14.410853842046503, + "grad_norm": 0.15546798706054688, + "learning_rate": 1.708474565016792e-05, + "loss": 0.0249, + "step": 74230 + }, + { + "epoch": 14.412795495364303, + "grad_norm": 0.20030932128429413, + "learning_rate": 1.7072300935806e-05, + "loss": 0.0231, + "step": 74240 + }, + { + "epoch": 14.414737148682104, + "grad_norm": 0.1386367231607437, + "learning_rate": 1.705985982236776e-05, + "loss": 0.0181, + "step": 74250 + }, + { + "epoch": 14.416678801999902, + "grad_norm": 0.17341899871826172, + "learning_rate": 1.7047422311213697e-05, + "loss": 0.0251, + "step": 74260 + }, + { + "epoch": 14.418620455317702, + "grad_norm": 0.15095864236354828, + "learning_rate": 1.703498840370401e-05, + "loss": 0.023, + "step": 74270 + }, + { + "epoch": 14.420562108635503, + "grad_norm": 0.13914579153060913, + "learning_rate": 1.702255810119842e-05, + "loss": 0.0229, + "step": 74280 + }, + { + "epoch": 14.422503761953303, + "grad_norm": 0.22384123504161835, + "learning_rate": 1.7010131405056305e-05, + "loss": 0.0304, + "step": 74290 + }, + { + "epoch": 14.424445415271103, + "grad_norm": 0.22809690237045288, + "learning_rate": 1.6997708316636613e-05, + "loss": 0.0279, + "step": 74300 + }, + { + "epoch": 14.426387068588904, + "grad_norm": 0.21748541295528412, + "learning_rate": 1.6985288837297932e-05, + "loss": 0.02, + "step": 74310 + }, + { + "epoch": 14.428328721906704, + "grad_norm": 0.14172521233558655, + "learning_rate": 1.6972872968398394e-05, + "loss": 0.0201, + "step": 74320 + }, + { + "epoch": 14.430270375224504, + "grad_norm": 0.1315937042236328, + "learning_rate": 1.6960460711295845e-05, + "loss": 0.0197, + "step": 74330 + }, + { + "epoch": 14.432212028542304, + "grad_norm": 0.19564485549926758, + "learning_rate": 1.6948052067347623e-05, + "loss": 0.0192, + "step": 74340 + }, + { + "epoch": 14.434153681860105, + "grad_norm": 0.12635438144207, + "learning_rate": 1.693564703791075e-05, + "loss": 0.0247, + "step": 74350 + }, + { + "epoch": 14.436095335177903, + "grad_norm": 0.2935437858104706, + "learning_rate": 1.6923245624341793e-05, + "loss": 0.0222, + "step": 74360 + }, + { + "epoch": 14.438036988495703, + "grad_norm": 0.1028546616435051, + "learning_rate": 1.6910847827996962e-05, + "loss": 0.0243, + "step": 74370 + }, + { + "epoch": 14.439978641813504, + "grad_norm": 0.17026151716709137, + "learning_rate": 1.689845365023206e-05, + "loss": 0.0215, + "step": 74380 + }, + { + "epoch": 14.441920295131304, + "grad_norm": 0.1475948989391327, + "learning_rate": 1.6886063092402505e-05, + "loss": 0.0221, + "step": 74390 + }, + { + "epoch": 14.443861948449104, + "grad_norm": 0.14138337969779968, + "learning_rate": 1.6873676155863317e-05, + "loss": 0.0211, + "step": 74400 + }, + { + "epoch": 14.445803601766904, + "grad_norm": 0.11192113161087036, + "learning_rate": 1.6861292841969085e-05, + "loss": 0.0222, + "step": 74410 + }, + { + "epoch": 14.447745255084705, + "grad_norm": 0.231907919049263, + "learning_rate": 1.684891315207406e-05, + "loss": 0.0304, + "step": 74420 + }, + { + "epoch": 14.449686908402505, + "grad_norm": 0.14172737300395966, + "learning_rate": 1.6836537087532034e-05, + "loss": 0.022, + "step": 74430 + }, + { + "epoch": 14.451628561720305, + "grad_norm": 0.10006345808506012, + "learning_rate": 1.6824164649696465e-05, + "loss": 0.0172, + "step": 74440 + }, + { + "epoch": 14.453570215038106, + "grad_norm": 0.2334979623556137, + "learning_rate": 1.681179583992034e-05, + "loss": 0.0278, + "step": 74450 + }, + { + "epoch": 14.455511868355906, + "grad_norm": 0.18101349472999573, + "learning_rate": 1.679943065955636e-05, + "loss": 0.0181, + "step": 74460 + }, + { + "epoch": 14.457453521673704, + "grad_norm": 0.09781799465417862, + "learning_rate": 1.678706910995671e-05, + "loss": 0.016, + "step": 74470 + }, + { + "epoch": 14.459395174991505, + "grad_norm": 0.2076522260904312, + "learning_rate": 1.6774711192473262e-05, + "loss": 0.0223, + "step": 74480 + }, + { + "epoch": 14.461336828309305, + "grad_norm": 0.22882239520549774, + "learning_rate": 1.676235690845743e-05, + "loss": 0.021, + "step": 74490 + }, + { + "epoch": 14.463278481627105, + "grad_norm": 0.2317926436662674, + "learning_rate": 1.6750006259260285e-05, + "loss": 0.0255, + "step": 74500 + }, + { + "epoch": 14.465220134944905, + "grad_norm": 0.1707993894815445, + "learning_rate": 1.673765924623244e-05, + "loss": 0.0228, + "step": 74510 + }, + { + "epoch": 14.467161788262706, + "grad_norm": 0.21219304203987122, + "learning_rate": 1.67253158707242e-05, + "loss": 0.0205, + "step": 74520 + }, + { + "epoch": 14.469103441580506, + "grad_norm": 0.13901810348033905, + "learning_rate": 1.671297613408537e-05, + "loss": 0.0241, + "step": 74530 + }, + { + "epoch": 14.471045094898306, + "grad_norm": 0.12823975086212158, + "learning_rate": 1.6700640037665444e-05, + "loss": 0.0204, + "step": 74540 + }, + { + "epoch": 14.472986748216107, + "grad_norm": 0.14756077527999878, + "learning_rate": 1.668830758281344e-05, + "loss": 0.0228, + "step": 74550 + }, + { + "epoch": 14.474928401533907, + "grad_norm": 0.23696212470531464, + "learning_rate": 1.6675978770878052e-05, + "loss": 0.0329, + "step": 74560 + }, + { + "epoch": 14.476870054851707, + "grad_norm": 0.14806507527828217, + "learning_rate": 1.6663653603207506e-05, + "loss": 0.0205, + "step": 74570 + }, + { + "epoch": 14.478811708169506, + "grad_norm": 0.2110086977481842, + "learning_rate": 1.6651332081149685e-05, + "loss": 0.0175, + "step": 74580 + }, + { + "epoch": 14.480753361487306, + "grad_norm": 0.21455426514148712, + "learning_rate": 1.6639014206052046e-05, + "loss": 0.0174, + "step": 74590 + }, + { + "epoch": 14.482695014805106, + "grad_norm": 0.15386609733104706, + "learning_rate": 1.6626699979261657e-05, + "loss": 0.0197, + "step": 74600 + }, + { + "epoch": 14.484636668122906, + "grad_norm": 0.24477843940258026, + "learning_rate": 1.6614389402125197e-05, + "loss": 0.0223, + "step": 74610 + }, + { + "epoch": 14.486578321440707, + "grad_norm": 0.14472328126430511, + "learning_rate": 1.66020824759889e-05, + "loss": 0.0216, + "step": 74620 + }, + { + "epoch": 14.488519974758507, + "grad_norm": 0.16390281915664673, + "learning_rate": 1.6589779202198675e-05, + "loss": 0.0312, + "step": 74630 + }, + { + "epoch": 14.490461628076307, + "grad_norm": 0.19684097170829773, + "learning_rate": 1.6577479582099924e-05, + "loss": 0.0197, + "step": 74640 + }, + { + "epoch": 14.492403281394107, + "grad_norm": 0.17183850705623627, + "learning_rate": 1.6565183617037788e-05, + "loss": 0.018, + "step": 74650 + }, + { + "epoch": 14.494344934711908, + "grad_norm": 0.2563095688819885, + "learning_rate": 1.6552891308356894e-05, + "loss": 0.0215, + "step": 74660 + }, + { + "epoch": 14.496286588029708, + "grad_norm": 0.148410826921463, + "learning_rate": 1.6540602657401526e-05, + "loss": 0.0233, + "step": 74670 + }, + { + "epoch": 14.498228241347508, + "grad_norm": 0.10088548809289932, + "learning_rate": 1.6528317665515537e-05, + "loss": 0.0214, + "step": 74680 + }, + { + "epoch": 14.500169894665307, + "grad_norm": 0.13868610560894012, + "learning_rate": 1.6516036334042416e-05, + "loss": 0.0145, + "step": 74690 + }, + { + "epoch": 14.502111547983107, + "grad_norm": 0.12362781912088394, + "learning_rate": 1.650375866432519e-05, + "loss": 0.0262, + "step": 74700 + }, + { + "epoch": 14.504053201300907, + "grad_norm": 0.14304283261299133, + "learning_rate": 1.649148465770659e-05, + "loss": 0.0194, + "step": 74710 + }, + { + "epoch": 14.505994854618708, + "grad_norm": 0.19868794083595276, + "learning_rate": 1.647921431552883e-05, + "loss": 0.0256, + "step": 74720 + }, + { + "epoch": 14.507936507936508, + "grad_norm": 0.17103807628154755, + "learning_rate": 1.646694763913381e-05, + "loss": 0.026, + "step": 74730 + }, + { + "epoch": 14.509878161254308, + "grad_norm": 0.14710839092731476, + "learning_rate": 1.6454684629862966e-05, + "loss": 0.024, + "step": 74740 + }, + { + "epoch": 14.511819814572108, + "grad_norm": 0.1720084846019745, + "learning_rate": 1.6442425289057388e-05, + "loss": 0.0235, + "step": 74750 + }, + { + "epoch": 14.513761467889909, + "grad_norm": 0.08533796668052673, + "learning_rate": 1.6430169618057718e-05, + "loss": 0.017, + "step": 74760 + }, + { + "epoch": 14.515703121207709, + "grad_norm": 0.10171707719564438, + "learning_rate": 1.641791761820422e-05, + "loss": 0.0298, + "step": 74770 + }, + { + "epoch": 14.51764477452551, + "grad_norm": 0.12841631472110748, + "learning_rate": 1.640566929083676e-05, + "loss": 0.0221, + "step": 74780 + }, + { + "epoch": 14.519586427843308, + "grad_norm": 0.16161881387233734, + "learning_rate": 1.6393424637294823e-05, + "loss": 0.0226, + "step": 74790 + }, + { + "epoch": 14.521528081161108, + "grad_norm": 0.15735280513763428, + "learning_rate": 1.638118365891742e-05, + "loss": 0.0239, + "step": 74800 + }, + { + "epoch": 14.523469734478908, + "grad_norm": 0.18759390711784363, + "learning_rate": 1.636894635704323e-05, + "loss": 0.0227, + "step": 74810 + }, + { + "epoch": 14.525411387796709, + "grad_norm": 0.13472795486450195, + "learning_rate": 1.6356712733010517e-05, + "loss": 0.0162, + "step": 74820 + }, + { + "epoch": 14.527353041114509, + "grad_norm": 0.2633356750011444, + "learning_rate": 1.6344482788157096e-05, + "loss": 0.0186, + "step": 74830 + }, + { + "epoch": 14.52929469443231, + "grad_norm": 0.19316087663173676, + "learning_rate": 1.6332256523820444e-05, + "loss": 0.0241, + "step": 74840 + }, + { + "epoch": 14.53123634775011, + "grad_norm": 0.1870635747909546, + "learning_rate": 1.6320033941337597e-05, + "loss": 0.0231, + "step": 74850 + }, + { + "epoch": 14.53317800106791, + "grad_norm": 0.20942698419094086, + "learning_rate": 1.6307815042045216e-05, + "loss": 0.0243, + "step": 74860 + }, + { + "epoch": 14.53511965438571, + "grad_norm": 0.21978610754013062, + "learning_rate": 1.6295599827279513e-05, + "loss": 0.0237, + "step": 74870 + }, + { + "epoch": 14.53706130770351, + "grad_norm": 0.21586157381534576, + "learning_rate": 1.6283388298376356e-05, + "loss": 0.0211, + "step": 74880 + }, + { + "epoch": 14.53900296102131, + "grad_norm": 0.26662495732307434, + "learning_rate": 1.6271180456671125e-05, + "loss": 0.0269, + "step": 74890 + }, + { + "epoch": 14.540944614339109, + "grad_norm": 0.1726795732975006, + "learning_rate": 1.625897630349893e-05, + "loss": 0.0218, + "step": 74900 + }, + { + "epoch": 14.54288626765691, + "grad_norm": 0.15299522876739502, + "learning_rate": 1.6246775840194335e-05, + "loss": 0.0223, + "step": 74910 + }, + { + "epoch": 14.54482792097471, + "grad_norm": 0.16735568642616272, + "learning_rate": 1.6234579068091604e-05, + "loss": 0.0204, + "step": 74920 + }, + { + "epoch": 14.54676957429251, + "grad_norm": 0.176646888256073, + "learning_rate": 1.6222385988524524e-05, + "loss": 0.018, + "step": 74930 + }, + { + "epoch": 14.54871122761031, + "grad_norm": 0.2536616325378418, + "learning_rate": 1.6210196602826544e-05, + "loss": 0.0273, + "step": 74940 + }, + { + "epoch": 14.55065288092811, + "grad_norm": 0.12481925636529922, + "learning_rate": 1.6198010912330646e-05, + "loss": 0.0237, + "step": 74950 + }, + { + "epoch": 14.55259453424591, + "grad_norm": 0.3112456798553467, + "learning_rate": 1.6185828918369456e-05, + "loss": 0.0301, + "step": 74960 + }, + { + "epoch": 14.554536187563711, + "grad_norm": 0.178793802857399, + "learning_rate": 1.617365062227517e-05, + "loss": 0.0228, + "step": 74970 + }, + { + "epoch": 14.556477840881511, + "grad_norm": 0.11614150553941727, + "learning_rate": 1.616147602537962e-05, + "loss": 0.0277, + "step": 74980 + }, + { + "epoch": 14.558419494199311, + "grad_norm": 0.13620750606060028, + "learning_rate": 1.614930512901415e-05, + "loss": 0.0207, + "step": 74990 + }, + { + "epoch": 14.56036114751711, + "grad_norm": 0.1504443734884262, + "learning_rate": 1.6137137934509804e-05, + "loss": 0.025, + "step": 75000 + }, + { + "epoch": 14.56230280083491, + "grad_norm": 0.20949704945087433, + "learning_rate": 1.6124974443197123e-05, + "loss": 0.0242, + "step": 75010 + }, + { + "epoch": 14.56424445415271, + "grad_norm": 0.2421049028635025, + "learning_rate": 1.6112814656406306e-05, + "loss": 0.0294, + "step": 75020 + }, + { + "epoch": 14.56618610747051, + "grad_norm": 0.20083141326904297, + "learning_rate": 1.6100658575467125e-05, + "loss": 0.0224, + "step": 75030 + }, + { + "epoch": 14.568127760788311, + "grad_norm": 0.10412740707397461, + "learning_rate": 1.6088506201708974e-05, + "loss": 0.02, + "step": 75040 + }, + { + "epoch": 14.570069414106111, + "grad_norm": 0.18133127689361572, + "learning_rate": 1.6076357536460785e-05, + "loss": 0.0168, + "step": 75050 + }, + { + "epoch": 14.572011067423912, + "grad_norm": 0.1529085487127304, + "learning_rate": 1.606421258105113e-05, + "loss": 0.0194, + "step": 75060 + }, + { + "epoch": 14.573952720741712, + "grad_norm": 0.21529962122440338, + "learning_rate": 1.6052071336808183e-05, + "loss": 0.0249, + "step": 75070 + }, + { + "epoch": 14.575894374059512, + "grad_norm": 0.11184683442115784, + "learning_rate": 1.6039933805059658e-05, + "loss": 0.0255, + "step": 75080 + }, + { + "epoch": 14.577836027377312, + "grad_norm": 0.12197065353393555, + "learning_rate": 1.6027799987132915e-05, + "loss": 0.0175, + "step": 75090 + }, + { + "epoch": 14.579777680695113, + "grad_norm": 0.078574538230896, + "learning_rate": 1.6015669884354888e-05, + "loss": 0.02, + "step": 75100 + }, + { + "epoch": 14.581719334012911, + "grad_norm": 0.15273578464984894, + "learning_rate": 1.6003543498052125e-05, + "loss": 0.0266, + "step": 75110 + }, + { + "epoch": 14.583660987330711, + "grad_norm": 0.1495855450630188, + "learning_rate": 1.5991420829550715e-05, + "loss": 0.0225, + "step": 75120 + }, + { + "epoch": 14.585602640648512, + "grad_norm": 0.1193271353840828, + "learning_rate": 1.5979301880176405e-05, + "loss": 0.0181, + "step": 75130 + }, + { + "epoch": 14.587544293966312, + "grad_norm": 0.1623242050409317, + "learning_rate": 1.5967186651254485e-05, + "loss": 0.019, + "step": 75140 + }, + { + "epoch": 14.589485947284112, + "grad_norm": 0.31590133905410767, + "learning_rate": 1.5955075144109853e-05, + "loss": 0.0284, + "step": 75150 + }, + { + "epoch": 14.591427600601913, + "grad_norm": 0.13327565789222717, + "learning_rate": 1.594296736006703e-05, + "loss": 0.0209, + "step": 75160 + }, + { + "epoch": 14.593369253919713, + "grad_norm": 0.21965958178043365, + "learning_rate": 1.59308633004501e-05, + "loss": 0.02, + "step": 75170 + }, + { + "epoch": 14.595310907237513, + "grad_norm": 0.19209755957126617, + "learning_rate": 1.5918762966582724e-05, + "loss": 0.0216, + "step": 75180 + }, + { + "epoch": 14.597252560555313, + "grad_norm": 0.19525893032550812, + "learning_rate": 1.5906666359788203e-05, + "loss": 0.0254, + "step": 75190 + }, + { + "epoch": 14.599194213873114, + "grad_norm": 0.23021367192268372, + "learning_rate": 1.589457348138938e-05, + "loss": 0.0201, + "step": 75200 + }, + { + "epoch": 14.601135867190912, + "grad_norm": 0.16945421695709229, + "learning_rate": 1.588248433270872e-05, + "loss": 0.0255, + "step": 75210 + }, + { + "epoch": 14.603077520508712, + "grad_norm": 0.14038902521133423, + "learning_rate": 1.587039891506828e-05, + "loss": 0.0227, + "step": 75220 + }, + { + "epoch": 14.605019173826513, + "grad_norm": 0.1401933878660202, + "learning_rate": 1.5858317229789715e-05, + "loss": 0.0218, + "step": 75230 + }, + { + "epoch": 14.606960827144313, + "grad_norm": 0.169049471616745, + "learning_rate": 1.584623927819423e-05, + "loss": 0.0196, + "step": 75240 + }, + { + "epoch": 14.608902480462113, + "grad_norm": 0.1393425464630127, + "learning_rate": 1.583416506160269e-05, + "loss": 0.0229, + "step": 75250 + }, + { + "epoch": 14.610844133779914, + "grad_norm": 0.08699537068605423, + "learning_rate": 1.5822094581335472e-05, + "loss": 0.0216, + "step": 75260 + }, + { + "epoch": 14.612785787097714, + "grad_norm": 0.31033098697662354, + "learning_rate": 1.5810027838712604e-05, + "loss": 0.0228, + "step": 75270 + }, + { + "epoch": 14.614727440415514, + "grad_norm": 0.1898641288280487, + "learning_rate": 1.5797964835053687e-05, + "loss": 0.0183, + "step": 75280 + }, + { + "epoch": 14.616669093733314, + "grad_norm": 0.23647430539131165, + "learning_rate": 1.5785905571677933e-05, + "loss": 0.0267, + "step": 75290 + }, + { + "epoch": 14.618610747051115, + "grad_norm": 0.16324134171009064, + "learning_rate": 1.5773850049904086e-05, + "loss": 0.0231, + "step": 75300 + }, + { + "epoch": 14.620552400368915, + "grad_norm": 0.1146799772977829, + "learning_rate": 1.5761798271050543e-05, + "loss": 0.0222, + "step": 75310 + }, + { + "epoch": 14.622494053686715, + "grad_norm": 0.1775435209274292, + "learning_rate": 1.5749750236435277e-05, + "loss": 0.0207, + "step": 75320 + }, + { + "epoch": 14.624435707004514, + "grad_norm": 0.2590726613998413, + "learning_rate": 1.5737705947375815e-05, + "loss": 0.0226, + "step": 75330 + }, + { + "epoch": 14.626377360322314, + "grad_norm": 0.11920076608657837, + "learning_rate": 1.5725665405189322e-05, + "loss": 0.0259, + "step": 75340 + }, + { + "epoch": 14.628319013640114, + "grad_norm": 0.15911433100700378, + "learning_rate": 1.571362861119253e-05, + "loss": 0.0232, + "step": 75350 + }, + { + "epoch": 14.630260666957914, + "grad_norm": 0.2202887386083603, + "learning_rate": 1.570159556670177e-05, + "loss": 0.0243, + "step": 75360 + }, + { + "epoch": 14.632202320275715, + "grad_norm": 0.13091765344142914, + "learning_rate": 1.5689566273032946e-05, + "loss": 0.0229, + "step": 75370 + }, + { + "epoch": 14.634143973593515, + "grad_norm": 0.17768394947052002, + "learning_rate": 1.5677540731501583e-05, + "loss": 0.0239, + "step": 75380 + }, + { + "epoch": 14.636085626911315, + "grad_norm": 0.13903379440307617, + "learning_rate": 1.5665518943422746e-05, + "loss": 0.0185, + "step": 75390 + }, + { + "epoch": 14.638027280229116, + "grad_norm": 0.20408478379249573, + "learning_rate": 1.5653500910111134e-05, + "loss": 0.0256, + "step": 75400 + }, + { + "epoch": 14.639968933546916, + "grad_norm": 0.20162390172481537, + "learning_rate": 1.5641486632881023e-05, + "loss": 0.0212, + "step": 75410 + }, + { + "epoch": 14.641910586864716, + "grad_norm": 0.12318796664476395, + "learning_rate": 1.56294761130463e-05, + "loss": 0.0194, + "step": 75420 + }, + { + "epoch": 14.643852240182515, + "grad_norm": 0.10145532339811325, + "learning_rate": 1.5617469351920373e-05, + "loss": 0.0226, + "step": 75430 + }, + { + "epoch": 14.645793893500315, + "grad_norm": 0.28760144114494324, + "learning_rate": 1.5605466350816315e-05, + "loss": 0.0289, + "step": 75440 + }, + { + "epoch": 14.647735546818115, + "grad_norm": 0.11680039018392563, + "learning_rate": 1.559346711104674e-05, + "loss": 0.0193, + "step": 75450 + }, + { + "epoch": 14.649677200135915, + "grad_norm": 0.19131655991077423, + "learning_rate": 1.5581471633923883e-05, + "loss": 0.0189, + "step": 75460 + }, + { + "epoch": 14.651618853453716, + "grad_norm": 0.2319885641336441, + "learning_rate": 1.5569479920759505e-05, + "loss": 0.0209, + "step": 75470 + }, + { + "epoch": 14.653560506771516, + "grad_norm": 0.18991760909557343, + "learning_rate": 1.5557491972865072e-05, + "loss": 0.0223, + "step": 75480 + }, + { + "epoch": 14.655502160089316, + "grad_norm": 0.24424895644187927, + "learning_rate": 1.554550779155152e-05, + "loss": 0.0209, + "step": 75490 + }, + { + "epoch": 14.657443813407117, + "grad_norm": 0.19141826033592224, + "learning_rate": 1.5533527378129453e-05, + "loss": 0.0204, + "step": 75500 + }, + { + "epoch": 14.659385466724917, + "grad_norm": 0.1851653754711151, + "learning_rate": 1.552155073390899e-05, + "loss": 0.0236, + "step": 75510 + }, + { + "epoch": 14.661327120042717, + "grad_norm": 0.21159522235393524, + "learning_rate": 1.5509577860199903e-05, + "loss": 0.0213, + "step": 75520 + }, + { + "epoch": 14.663268773360517, + "grad_norm": 0.22140024602413177, + "learning_rate": 1.5497608758311525e-05, + "loss": 0.0239, + "step": 75530 + }, + { + "epoch": 14.665210426678316, + "grad_norm": 0.18066471815109253, + "learning_rate": 1.5485643429552797e-05, + "loss": 0.0268, + "step": 75540 + }, + { + "epoch": 14.667152079996116, + "grad_norm": 0.17800474166870117, + "learning_rate": 1.5473681875232197e-05, + "loss": 0.0225, + "step": 75550 + }, + { + "epoch": 14.669093733313916, + "grad_norm": 0.2774960994720459, + "learning_rate": 1.5461724096657837e-05, + "loss": 0.0165, + "step": 75560 + }, + { + "epoch": 14.671035386631717, + "grad_norm": 0.28252097964286804, + "learning_rate": 1.5449770095137416e-05, + "loss": 0.02, + "step": 75570 + }, + { + "epoch": 14.672977039949517, + "grad_norm": 0.14569731056690216, + "learning_rate": 1.5437819871978178e-05, + "loss": 0.0199, + "step": 75580 + }, + { + "epoch": 14.674918693267317, + "grad_norm": 0.15791179239749908, + "learning_rate": 1.5425873428486992e-05, + "loss": 0.0243, + "step": 75590 + }, + { + "epoch": 14.676860346585118, + "grad_norm": 0.20194584131240845, + "learning_rate": 1.5413930765970312e-05, + "loss": 0.0182, + "step": 75600 + }, + { + "epoch": 14.678801999902918, + "grad_norm": 0.18464188277721405, + "learning_rate": 1.5401991885734175e-05, + "loss": 0.0223, + "step": 75610 + }, + { + "epoch": 14.680743653220718, + "grad_norm": 0.19670860469341278, + "learning_rate": 1.5390056789084174e-05, + "loss": 0.0165, + "step": 75620 + }, + { + "epoch": 14.682685306538518, + "grad_norm": 0.200324147939682, + "learning_rate": 1.5378125477325545e-05, + "loss": 0.0215, + "step": 75630 + }, + { + "epoch": 14.684626959856317, + "grad_norm": 0.12302065640687943, + "learning_rate": 1.5366197951763046e-05, + "loss": 0.0178, + "step": 75640 + }, + { + "epoch": 14.686568613174117, + "grad_norm": 0.15679726004600525, + "learning_rate": 1.5354274213701075e-05, + "loss": 0.0278, + "step": 75650 + }, + { + "epoch": 14.688510266491917, + "grad_norm": 0.19107964634895325, + "learning_rate": 1.5342354264443566e-05, + "loss": 0.0214, + "step": 75660 + }, + { + "epoch": 14.690451919809718, + "grad_norm": 0.32030314207077026, + "learning_rate": 1.533043810529411e-05, + "loss": 0.0247, + "step": 75670 + }, + { + "epoch": 14.692393573127518, + "grad_norm": 0.35444164276123047, + "learning_rate": 1.5318525737555807e-05, + "loss": 0.0221, + "step": 75680 + }, + { + "epoch": 14.694335226445318, + "grad_norm": 0.14587321877479553, + "learning_rate": 1.53066171625314e-05, + "loss": 0.0197, + "step": 75690 + }, + { + "epoch": 14.696276879763118, + "grad_norm": 0.1653880923986435, + "learning_rate": 1.529471238152317e-05, + "loss": 0.0225, + "step": 75700 + }, + { + "epoch": 14.698218533080919, + "grad_norm": 0.1894708275794983, + "learning_rate": 1.5282811395833023e-05, + "loss": 0.0223, + "step": 75710 + }, + { + "epoch": 14.700160186398719, + "grad_norm": 0.19224734604358673, + "learning_rate": 1.5270914206762397e-05, + "loss": 0.0288, + "step": 75720 + }, + { + "epoch": 14.70210183971652, + "grad_norm": 0.2815077304840088, + "learning_rate": 1.525902081561241e-05, + "loss": 0.0201, + "step": 75730 + }, + { + "epoch": 14.70404349303432, + "grad_norm": 0.20817792415618896, + "learning_rate": 1.5247131223683664e-05, + "loss": 0.0224, + "step": 75740 + }, + { + "epoch": 14.705985146352118, + "grad_norm": 0.16891661286354065, + "learning_rate": 1.5235245432276408e-05, + "loss": 0.0245, + "step": 75750 + }, + { + "epoch": 14.707926799669918, + "grad_norm": 0.2688045799732208, + "learning_rate": 1.5223363442690424e-05, + "loss": 0.0189, + "step": 75760 + }, + { + "epoch": 14.709868452987719, + "grad_norm": 0.21970441937446594, + "learning_rate": 1.5211485256225133e-05, + "loss": 0.017, + "step": 75770 + }, + { + "epoch": 14.711810106305519, + "grad_norm": 0.11987601220607758, + "learning_rate": 1.519961087417951e-05, + "loss": 0.0194, + "step": 75780 + }, + { + "epoch": 14.71375175962332, + "grad_norm": 0.13988490402698517, + "learning_rate": 1.5187740297852137e-05, + "loss": 0.0183, + "step": 75790 + }, + { + "epoch": 14.71569341294112, + "grad_norm": 0.1474798172712326, + "learning_rate": 1.517587352854113e-05, + "loss": 0.0196, + "step": 75800 + }, + { + "epoch": 14.71763506625892, + "grad_norm": 0.34122124314308167, + "learning_rate": 1.516401056754423e-05, + "loss": 0.0294, + "step": 75810 + }, + { + "epoch": 14.71957671957672, + "grad_norm": 0.18152698874473572, + "learning_rate": 1.5152151416158783e-05, + "loss": 0.0223, + "step": 75820 + }, + { + "epoch": 14.72151837289452, + "grad_norm": 0.11679630726575851, + "learning_rate": 1.5140296075681642e-05, + "loss": 0.0178, + "step": 75830 + }, + { + "epoch": 14.72346002621232, + "grad_norm": 0.17902666330337524, + "learning_rate": 1.5128444547409331e-05, + "loss": 0.0262, + "step": 75840 + }, + { + "epoch": 14.725401679530119, + "grad_norm": 0.14615024626255035, + "learning_rate": 1.5116596832637864e-05, + "loss": 0.0176, + "step": 75850 + }, + { + "epoch": 14.72734333284792, + "grad_norm": 0.26876774430274963, + "learning_rate": 1.5104752932662952e-05, + "loss": 0.0317, + "step": 75860 + }, + { + "epoch": 14.72928498616572, + "grad_norm": 0.23336610198020935, + "learning_rate": 1.5092912848779778e-05, + "loss": 0.0207, + "step": 75870 + }, + { + "epoch": 14.73122663948352, + "grad_norm": 0.1561601758003235, + "learning_rate": 1.5081076582283194e-05, + "loss": 0.0209, + "step": 75880 + }, + { + "epoch": 14.73316829280132, + "grad_norm": 0.1598028689622879, + "learning_rate": 1.5069244134467553e-05, + "loss": 0.023, + "step": 75890 + }, + { + "epoch": 14.73510994611912, + "grad_norm": 0.2273743897676468, + "learning_rate": 1.5057415506626876e-05, + "loss": 0.0245, + "step": 75900 + }, + { + "epoch": 14.73705159943692, + "grad_norm": 0.16832756996154785, + "learning_rate": 1.5045590700054668e-05, + "loss": 0.0178, + "step": 75910 + }, + { + "epoch": 14.738993252754721, + "grad_norm": 0.11652843654155731, + "learning_rate": 1.5033769716044144e-05, + "loss": 0.0258, + "step": 75920 + }, + { + "epoch": 14.740934906072521, + "grad_norm": 0.20910154283046722, + "learning_rate": 1.5021952555887975e-05, + "loss": 0.0212, + "step": 75930 + }, + { + "epoch": 14.742876559390321, + "grad_norm": 0.23128719627857208, + "learning_rate": 1.50101392208785e-05, + "loss": 0.0187, + "step": 75940 + }, + { + "epoch": 14.744818212708122, + "grad_norm": 0.2121761590242386, + "learning_rate": 1.4998329712307579e-05, + "loss": 0.0354, + "step": 75950 + }, + { + "epoch": 14.74675986602592, + "grad_norm": 0.16249392926692963, + "learning_rate": 1.4986524031466708e-05, + "loss": 0.0168, + "step": 75960 + }, + { + "epoch": 14.74870151934372, + "grad_norm": 0.3084944486618042, + "learning_rate": 1.4974722179646916e-05, + "loss": 0.0234, + "step": 75970 + }, + { + "epoch": 14.75064317266152, + "grad_norm": 0.10204707831144333, + "learning_rate": 1.4962924158138847e-05, + "loss": 0.0223, + "step": 75980 + }, + { + "epoch": 14.752584825979321, + "grad_norm": 0.16151732206344604, + "learning_rate": 1.4951129968232714e-05, + "loss": 0.0177, + "step": 75990 + }, + { + "epoch": 14.754526479297121, + "grad_norm": 0.11579529196023941, + "learning_rate": 1.4939339611218329e-05, + "loss": 0.024, + "step": 76000 + }, + { + "epoch": 14.756468132614922, + "grad_norm": 0.1589973419904709, + "learning_rate": 1.4927553088385032e-05, + "loss": 0.0225, + "step": 76010 + }, + { + "epoch": 14.758409785932722, + "grad_norm": 0.12997601926326752, + "learning_rate": 1.4915770401021801e-05, + "loss": 0.0199, + "step": 76020 + }, + { + "epoch": 14.760351439250522, + "grad_norm": 0.13845987617969513, + "learning_rate": 1.4903991550417185e-05, + "loss": 0.0251, + "step": 76030 + }, + { + "epoch": 14.762293092568322, + "grad_norm": 0.2161923050880432, + "learning_rate": 1.4892216537859277e-05, + "loss": 0.0214, + "step": 76040 + }, + { + "epoch": 14.764234745886123, + "grad_norm": 0.1065414771437645, + "learning_rate": 1.4880445364635775e-05, + "loss": 0.02, + "step": 76050 + }, + { + "epoch": 14.766176399203921, + "grad_norm": 0.14053645730018616, + "learning_rate": 1.4868678032033973e-05, + "loss": 0.016, + "step": 76060 + }, + { + "epoch": 14.768118052521721, + "grad_norm": 0.24066655337810516, + "learning_rate": 1.4856914541340739e-05, + "loss": 0.0247, + "step": 76070 + }, + { + "epoch": 14.770059705839522, + "grad_norm": 0.12474646419286728, + "learning_rate": 1.4845154893842472e-05, + "loss": 0.019, + "step": 76080 + }, + { + "epoch": 14.772001359157322, + "grad_norm": 0.13805563747882843, + "learning_rate": 1.4833399090825234e-05, + "loss": 0.027, + "step": 76090 + }, + { + "epoch": 14.773943012475122, + "grad_norm": 0.14095991849899292, + "learning_rate": 1.4821647133574562e-05, + "loss": 0.021, + "step": 76100 + }, + { + "epoch": 14.775884665792923, + "grad_norm": 0.2367318719625473, + "learning_rate": 1.4809899023375706e-05, + "loss": 0.022, + "step": 76110 + }, + { + "epoch": 14.777826319110723, + "grad_norm": 0.1487371176481247, + "learning_rate": 1.4798154761513366e-05, + "loss": 0.0248, + "step": 76120 + }, + { + "epoch": 14.779767972428523, + "grad_norm": 0.239067941904068, + "learning_rate": 1.4786414349271916e-05, + "loss": 0.0266, + "step": 76130 + }, + { + "epoch": 14.781709625746323, + "grad_norm": 0.24736067652702332, + "learning_rate": 1.4774677787935226e-05, + "loss": 0.0256, + "step": 76140 + }, + { + "epoch": 14.783651279064124, + "grad_norm": 0.14188796281814575, + "learning_rate": 1.4762945078786838e-05, + "loss": 0.0264, + "step": 76150 + }, + { + "epoch": 14.785592932381924, + "grad_norm": 0.2515716850757599, + "learning_rate": 1.4751216223109777e-05, + "loss": 0.0257, + "step": 76160 + }, + { + "epoch": 14.787534585699724, + "grad_norm": 0.1735125035047531, + "learning_rate": 1.4739491222186714e-05, + "loss": 0.0256, + "step": 76170 + }, + { + "epoch": 14.789476239017523, + "grad_norm": 0.25045880675315857, + "learning_rate": 1.4727770077299879e-05, + "loss": 0.021, + "step": 76180 + }, + { + "epoch": 14.791417892335323, + "grad_norm": 0.15059128403663635, + "learning_rate": 1.4716052789731094e-05, + "loss": 0.0219, + "step": 76190 + }, + { + "epoch": 14.793359545653123, + "grad_norm": 0.20650707185268402, + "learning_rate": 1.470433936076171e-05, + "loss": 0.0263, + "step": 76200 + }, + { + "epoch": 14.795301198970924, + "grad_norm": 0.15574763715267181, + "learning_rate": 1.4692629791672719e-05, + "loss": 0.0182, + "step": 76210 + }, + { + "epoch": 14.797242852288724, + "grad_norm": 0.22379553318023682, + "learning_rate": 1.468092408374464e-05, + "loss": 0.0232, + "step": 76220 + }, + { + "epoch": 14.799184505606524, + "grad_norm": 0.24149106442928314, + "learning_rate": 1.4669222238257602e-05, + "loss": 0.0315, + "step": 76230 + }, + { + "epoch": 14.801126158924324, + "grad_norm": 0.2532881498336792, + "learning_rate": 1.4657524256491301e-05, + "loss": 0.0234, + "step": 76240 + }, + { + "epoch": 14.803067812242125, + "grad_norm": 0.10953740775585175, + "learning_rate": 1.464583013972502e-05, + "loss": 0.0176, + "step": 76250 + }, + { + "epoch": 14.805009465559925, + "grad_norm": 0.20497764647006989, + "learning_rate": 1.4634139889237591e-05, + "loss": 0.0299, + "step": 76260 + }, + { + "epoch": 14.806951118877725, + "grad_norm": 0.13668911159038544, + "learning_rate": 1.4622453506307448e-05, + "loss": 0.0298, + "step": 76270 + }, + { + "epoch": 14.808892772195524, + "grad_norm": 0.16580216586589813, + "learning_rate": 1.461077099221262e-05, + "loss": 0.0211, + "step": 76280 + }, + { + "epoch": 14.810834425513324, + "grad_norm": 0.20710110664367676, + "learning_rate": 1.4599092348230652e-05, + "loss": 0.0218, + "step": 76290 + }, + { + "epoch": 14.812776078831124, + "grad_norm": 0.1707145869731903, + "learning_rate": 1.4587417575638724e-05, + "loss": 0.0257, + "step": 76300 + }, + { + "epoch": 14.814717732148925, + "grad_norm": 0.2208695113658905, + "learning_rate": 1.4575746675713563e-05, + "loss": 0.024, + "step": 76310 + }, + { + "epoch": 14.816659385466725, + "grad_norm": 0.05804787948727608, + "learning_rate": 1.4564079649731504e-05, + "loss": 0.0268, + "step": 76320 + }, + { + "epoch": 14.818601038784525, + "grad_norm": 0.27227288484573364, + "learning_rate": 1.4552416498968402e-05, + "loss": 0.0252, + "step": 76330 + }, + { + "epoch": 14.820542692102325, + "grad_norm": 0.16379688680171967, + "learning_rate": 1.4540757224699747e-05, + "loss": 0.024, + "step": 76340 + }, + { + "epoch": 14.822484345420126, + "grad_norm": 0.1551695019006729, + "learning_rate": 1.452910182820056e-05, + "loss": 0.0217, + "step": 76350 + }, + { + "epoch": 14.824425998737926, + "grad_norm": 0.15913723409175873, + "learning_rate": 1.4517450310745468e-05, + "loss": 0.0166, + "step": 76360 + }, + { + "epoch": 14.826367652055726, + "grad_norm": 0.17260770499706268, + "learning_rate": 1.450580267360866e-05, + "loss": 0.0212, + "step": 76370 + }, + { + "epoch": 14.828309305373526, + "grad_norm": 0.28299617767333984, + "learning_rate": 1.4494158918063922e-05, + "loss": 0.0324, + "step": 76380 + }, + { + "epoch": 14.830250958691325, + "grad_norm": 0.15780238807201385, + "learning_rate": 1.4482519045384568e-05, + "loss": 0.0171, + "step": 76390 + }, + { + "epoch": 14.832192612009125, + "grad_norm": 0.19566066563129425, + "learning_rate": 1.4470883056843548e-05, + "loss": 0.0341, + "step": 76400 + }, + { + "epoch": 14.834134265326925, + "grad_norm": 0.1618126481771469, + "learning_rate": 1.4459250953713322e-05, + "loss": 0.0192, + "step": 76410 + }, + { + "epoch": 14.836075918644726, + "grad_norm": 0.13623829185962677, + "learning_rate": 1.444762273726598e-05, + "loss": 0.0289, + "step": 76420 + }, + { + "epoch": 14.838017571962526, + "grad_norm": 0.19103282690048218, + "learning_rate": 1.443599840877316e-05, + "loss": 0.0267, + "step": 76430 + }, + { + "epoch": 14.839959225280326, + "grad_norm": 0.3487781286239624, + "learning_rate": 1.4424377969506103e-05, + "loss": 0.0358, + "step": 76440 + }, + { + "epoch": 14.841900878598127, + "grad_norm": 0.1267346441745758, + "learning_rate": 1.4412761420735565e-05, + "loss": 0.017, + "step": 76450 + }, + { + "epoch": 14.843842531915927, + "grad_norm": 0.2092766910791397, + "learning_rate": 1.4401148763731953e-05, + "loss": 0.0228, + "step": 76460 + }, + { + "epoch": 14.845784185233727, + "grad_norm": 0.2538582980632782, + "learning_rate": 1.4389539999765173e-05, + "loss": 0.028, + "step": 76470 + }, + { + "epoch": 14.847725838551527, + "grad_norm": 0.12601977586746216, + "learning_rate": 1.4377935130104758e-05, + "loss": 0.0198, + "step": 76480 + }, + { + "epoch": 14.849667491869326, + "grad_norm": 0.11885770410299301, + "learning_rate": 1.4366334156019801e-05, + "loss": 0.0203, + "step": 76490 + }, + { + "epoch": 14.851609145187126, + "grad_norm": 0.14125953614711761, + "learning_rate": 1.435473707877898e-05, + "loss": 0.0219, + "step": 76500 + }, + { + "epoch": 14.853550798504926, + "grad_norm": 0.6260254979133606, + "learning_rate": 1.4343143899650508e-05, + "loss": 0.0232, + "step": 76510 + }, + { + "epoch": 14.855492451822727, + "grad_norm": 0.1719396710395813, + "learning_rate": 1.4331554619902204e-05, + "loss": 0.0263, + "step": 76520 + }, + { + "epoch": 14.857434105140527, + "grad_norm": 0.303230881690979, + "learning_rate": 1.4319969240801484e-05, + "loss": 0.0266, + "step": 76530 + }, + { + "epoch": 14.859375758458327, + "grad_norm": 0.08070583641529083, + "learning_rate": 1.430838776361526e-05, + "loss": 0.0208, + "step": 76540 + }, + { + "epoch": 14.861317411776128, + "grad_norm": 0.14982417225837708, + "learning_rate": 1.4296810189610093e-05, + "loss": 0.0217, + "step": 76550 + }, + { + "epoch": 14.863259065093928, + "grad_norm": 0.13401249051094055, + "learning_rate": 1.4285236520052087e-05, + "loss": 0.0198, + "step": 76560 + }, + { + "epoch": 14.865200718411728, + "grad_norm": 0.27612099051475525, + "learning_rate": 1.4273666756206933e-05, + "loss": 0.0286, + "step": 76570 + }, + { + "epoch": 14.867142371729528, + "grad_norm": 0.11914682388305664, + "learning_rate": 1.4262100899339848e-05, + "loss": 0.0267, + "step": 76580 + }, + { + "epoch": 14.869084025047329, + "grad_norm": 0.10639733076095581, + "learning_rate": 1.4250538950715696e-05, + "loss": 0.0236, + "step": 76590 + }, + { + "epoch": 14.871025678365127, + "grad_norm": 0.25476694107055664, + "learning_rate": 1.4238980911598842e-05, + "loss": 0.0298, + "step": 76600 + }, + { + "epoch": 14.872967331682927, + "grad_norm": 0.14257550239562988, + "learning_rate": 1.4227426783253267e-05, + "loss": 0.0229, + "step": 76610 + }, + { + "epoch": 14.874908985000728, + "grad_norm": 0.2339712530374527, + "learning_rate": 1.4215876566942521e-05, + "loss": 0.0224, + "step": 76620 + }, + { + "epoch": 14.876850638318528, + "grad_norm": 0.19527986645698547, + "learning_rate": 1.4204330263929721e-05, + "loss": 0.0269, + "step": 76630 + }, + { + "epoch": 14.878792291636328, + "grad_norm": 0.1987067312002182, + "learning_rate": 1.4192787875477537e-05, + "loss": 0.0176, + "step": 76640 + }, + { + "epoch": 14.880733944954128, + "grad_norm": 0.2635473608970642, + "learning_rate": 1.4181249402848246e-05, + "loss": 0.0201, + "step": 76650 + }, + { + "epoch": 14.882675598271929, + "grad_norm": 0.05875702202320099, + "learning_rate": 1.4169714847303656e-05, + "loss": 0.017, + "step": 76660 + }, + { + "epoch": 14.884617251589729, + "grad_norm": 0.19308336079120636, + "learning_rate": 1.4158184210105197e-05, + "loss": 0.0237, + "step": 76670 + }, + { + "epoch": 14.88655890490753, + "grad_norm": 0.21630136668682098, + "learning_rate": 1.4146657492513787e-05, + "loss": 0.0232, + "step": 76680 + }, + { + "epoch": 14.88850055822533, + "grad_norm": 0.13994891941547394, + "learning_rate": 1.4135134695790047e-05, + "loss": 0.0159, + "step": 76690 + }, + { + "epoch": 14.890442211543128, + "grad_norm": 0.20169247686862946, + "learning_rate": 1.4123615821194037e-05, + "loss": 0.0197, + "step": 76700 + }, + { + "epoch": 14.892383864860928, + "grad_norm": 0.21295280754566193, + "learning_rate": 1.4112100869985473e-05, + "loss": 0.0214, + "step": 76710 + }, + { + "epoch": 14.894325518178729, + "grad_norm": 0.113213449716568, + "learning_rate": 1.4100589843423589e-05, + "loss": 0.0182, + "step": 76720 + }, + { + "epoch": 14.896267171496529, + "grad_norm": 0.25324705243110657, + "learning_rate": 1.408908274276723e-05, + "loss": 0.0214, + "step": 76730 + }, + { + "epoch": 14.89820882481433, + "grad_norm": 0.12991711497306824, + "learning_rate": 1.4077579569274786e-05, + "loss": 0.0206, + "step": 76740 + }, + { + "epoch": 14.90015047813213, + "grad_norm": 0.23480753600597382, + "learning_rate": 1.4066080324204245e-05, + "loss": 0.0261, + "step": 76750 + }, + { + "epoch": 14.90209213144993, + "grad_norm": 0.2529859244823456, + "learning_rate": 1.405458500881312e-05, + "loss": 0.0254, + "step": 76760 + }, + { + "epoch": 14.90403378476773, + "grad_norm": 0.11742053180932999, + "learning_rate": 1.4043093624358538e-05, + "loss": 0.0257, + "step": 76770 + }, + { + "epoch": 14.90597543808553, + "grad_norm": 0.1091168001294136, + "learning_rate": 1.4031606172097189e-05, + "loss": 0.0286, + "step": 76780 + }, + { + "epoch": 14.90791709140333, + "grad_norm": 0.14397600293159485, + "learning_rate": 1.40201226532853e-05, + "loss": 0.0243, + "step": 76790 + }, + { + "epoch": 14.90985874472113, + "grad_norm": 0.19860833883285522, + "learning_rate": 1.4008643069178701e-05, + "loss": 0.0312, + "step": 76800 + }, + { + "epoch": 14.911800398038931, + "grad_norm": 0.16933298110961914, + "learning_rate": 1.399716742103278e-05, + "loss": 0.0248, + "step": 76810 + }, + { + "epoch": 14.91374205135673, + "grad_norm": 0.15127278864383698, + "learning_rate": 1.398569571010252e-05, + "loss": 0.0185, + "step": 76820 + }, + { + "epoch": 14.91568370467453, + "grad_norm": 0.12204398959875107, + "learning_rate": 1.3974227937642414e-05, + "loss": 0.0171, + "step": 76830 + }, + { + "epoch": 14.91762535799233, + "grad_norm": 0.18826249241828918, + "learning_rate": 1.3962764104906595e-05, + "loss": 0.0216, + "step": 76840 + }, + { + "epoch": 14.91956701131013, + "grad_norm": 0.21285754442214966, + "learning_rate": 1.3951304213148696e-05, + "loss": 0.0196, + "step": 76850 + }, + { + "epoch": 14.92150866462793, + "grad_norm": 0.16296935081481934, + "learning_rate": 1.3939848263621979e-05, + "loss": 0.0231, + "step": 76860 + }, + { + "epoch": 14.923450317945731, + "grad_norm": 0.2001381516456604, + "learning_rate": 1.3928396257579219e-05, + "loss": 0.0246, + "step": 76870 + }, + { + "epoch": 14.925391971263531, + "grad_norm": 0.14483006298542023, + "learning_rate": 1.3916948196272833e-05, + "loss": 0.0173, + "step": 76880 + }, + { + "epoch": 14.927333624581332, + "grad_norm": 0.11926526576280594, + "learning_rate": 1.3905504080954734e-05, + "loss": 0.024, + "step": 76890 + }, + { + "epoch": 14.929275277899132, + "grad_norm": 0.24996066093444824, + "learning_rate": 1.3894063912876454e-05, + "loss": 0.0292, + "step": 76900 + }, + { + "epoch": 14.93121693121693, + "grad_norm": 0.21541590988636017, + "learning_rate": 1.3882627693289046e-05, + "loss": 0.0209, + "step": 76910 + }, + { + "epoch": 14.93315858453473, + "grad_norm": 0.22297362983226776, + "learning_rate": 1.3871195423443183e-05, + "loss": 0.021, + "step": 76920 + }, + { + "epoch": 14.93510023785253, + "grad_norm": 0.22614572942256927, + "learning_rate": 1.3859767104589044e-05, + "loss": 0.0201, + "step": 76930 + }, + { + "epoch": 14.937041891170331, + "grad_norm": 0.1298176795244217, + "learning_rate": 1.3848342737976467e-05, + "loss": 0.0186, + "step": 76940 + }, + { + "epoch": 14.938983544488131, + "grad_norm": 0.17235955595970154, + "learning_rate": 1.3836922324854757e-05, + "loss": 0.0209, + "step": 76950 + }, + { + "epoch": 14.940925197805932, + "grad_norm": 0.2094111144542694, + "learning_rate": 1.382550586647287e-05, + "loss": 0.0228, + "step": 76960 + }, + { + "epoch": 14.942866851123732, + "grad_norm": 0.16367107629776, + "learning_rate": 1.3814093364079257e-05, + "loss": 0.0249, + "step": 76970 + }, + { + "epoch": 14.944808504441532, + "grad_norm": 0.1774241328239441, + "learning_rate": 1.3802684818921991e-05, + "loss": 0.0229, + "step": 76980 + }, + { + "epoch": 14.946750157759332, + "grad_norm": 0.18684791028499603, + "learning_rate": 1.3791280232248687e-05, + "loss": 0.0212, + "step": 76990 + }, + { + "epoch": 14.948691811077133, + "grad_norm": 0.17341038584709167, + "learning_rate": 1.3779879605306555e-05, + "loss": 0.0255, + "step": 77000 + }, + { + "epoch": 14.950633464394933, + "grad_norm": 0.2785375416278839, + "learning_rate": 1.376848293934232e-05, + "loss": 0.0301, + "step": 77010 + }, + { + "epoch": 14.952575117712733, + "grad_norm": 0.2042786180973053, + "learning_rate": 1.3757090235602316e-05, + "loss": 0.0253, + "step": 77020 + }, + { + "epoch": 14.954516771030532, + "grad_norm": 0.2285882979631424, + "learning_rate": 1.3745701495332447e-05, + "loss": 0.0218, + "step": 77030 + }, + { + "epoch": 14.956458424348332, + "grad_norm": 0.2803051769733429, + "learning_rate": 1.3734316719778145e-05, + "loss": 0.0271, + "step": 77040 + }, + { + "epoch": 14.958400077666132, + "grad_norm": 0.17399248480796814, + "learning_rate": 1.3722935910184454e-05, + "loss": 0.0172, + "step": 77050 + }, + { + "epoch": 14.960341730983933, + "grad_norm": 0.16757610440254211, + "learning_rate": 1.3711559067795921e-05, + "loss": 0.0191, + "step": 77060 + }, + { + "epoch": 14.962283384301733, + "grad_norm": 0.08635014295578003, + "learning_rate": 1.3700186193856756e-05, + "loss": 0.0217, + "step": 77070 + }, + { + "epoch": 14.964225037619533, + "grad_norm": 0.09781196713447571, + "learning_rate": 1.368881728961064e-05, + "loss": 0.0195, + "step": 77080 + }, + { + "epoch": 14.966166690937333, + "grad_norm": 0.21719205379486084, + "learning_rate": 1.3677452356300885e-05, + "loss": 0.0151, + "step": 77090 + }, + { + "epoch": 14.968108344255134, + "grad_norm": 0.2723887264728546, + "learning_rate": 1.3666091395170321e-05, + "loss": 0.029, + "step": 77100 + }, + { + "epoch": 14.970049997572934, + "grad_norm": 0.152408167719841, + "learning_rate": 1.3654734407461384e-05, + "loss": 0.0214, + "step": 77110 + }, + { + "epoch": 14.971991650890734, + "grad_norm": 0.27230924367904663, + "learning_rate": 1.3643381394416022e-05, + "loss": 0.0277, + "step": 77120 + }, + { + "epoch": 14.973933304208533, + "grad_norm": 0.2360486388206482, + "learning_rate": 1.3632032357275838e-05, + "loss": 0.0244, + "step": 77130 + }, + { + "epoch": 14.975874957526333, + "grad_norm": 0.29786914587020874, + "learning_rate": 1.3620687297281898e-05, + "loss": 0.0263, + "step": 77140 + }, + { + "epoch": 14.977816610844133, + "grad_norm": 0.48841413855552673, + "learning_rate": 1.3609346215674917e-05, + "loss": 0.0226, + "step": 77150 + }, + { + "epoch": 14.979758264161934, + "grad_norm": 0.1957114338874817, + "learning_rate": 1.3598009113695109e-05, + "loss": 0.023, + "step": 77160 + }, + { + "epoch": 14.981699917479734, + "grad_norm": 0.1877567023038864, + "learning_rate": 1.358667599258231e-05, + "loss": 0.0241, + "step": 77170 + }, + { + "epoch": 14.983641570797534, + "grad_norm": 0.2151319980621338, + "learning_rate": 1.357534685357586e-05, + "loss": 0.0229, + "step": 77180 + }, + { + "epoch": 14.985583224115334, + "grad_norm": 0.14345189929008484, + "learning_rate": 1.3564021697914715e-05, + "loss": 0.0248, + "step": 77190 + }, + { + "epoch": 14.987524877433135, + "grad_norm": 0.10840703547000885, + "learning_rate": 1.3552700526837381e-05, + "loss": 0.0246, + "step": 77200 + }, + { + "epoch": 14.989466530750935, + "grad_norm": 0.18736128509044647, + "learning_rate": 1.3541383341581925e-05, + "loss": 0.0211, + "step": 77210 + }, + { + "epoch": 14.991408184068735, + "grad_norm": 0.22447951138019562, + "learning_rate": 1.3530070143385965e-05, + "loss": 0.0181, + "step": 77220 + }, + { + "epoch": 14.993349837386535, + "grad_norm": 0.2381797730922699, + "learning_rate": 1.3518760933486702e-05, + "loss": 0.0256, + "step": 77230 + }, + { + "epoch": 14.995291490704334, + "grad_norm": 0.19094781577587128, + "learning_rate": 1.3507455713120908e-05, + "loss": 0.0198, + "step": 77240 + }, + { + "epoch": 14.997233144022134, + "grad_norm": 0.14764811098575592, + "learning_rate": 1.3496154483524876e-05, + "loss": 0.0216, + "step": 77250 + }, + { + "epoch": 14.999174797339935, + "grad_norm": 0.21537183225154877, + "learning_rate": 1.3484857245934506e-05, + "loss": 0.0237, + "step": 77260 + }, + { + "epoch": 15.0009708266589, + "grad_norm": 0.12887531518936157, + "learning_rate": 1.3473564001585248e-05, + "loss": 0.023, + "step": 77270 + }, + { + "epoch": 15.0029124799767, + "grad_norm": 0.17731204628944397, + "learning_rate": 1.3462274751712129e-05, + "loss": 0.0267, + "step": 77280 + }, + { + "epoch": 15.0048541332945, + "grad_norm": 0.12871114909648895, + "learning_rate": 1.3450989497549698e-05, + "loss": 0.023, + "step": 77290 + }, + { + "epoch": 15.006795786612301, + "grad_norm": 0.16700123250484467, + "learning_rate": 1.3439708240332115e-05, + "loss": 0.0224, + "step": 77300 + }, + { + "epoch": 15.008737439930101, + "grad_norm": 0.2857915759086609, + "learning_rate": 1.3428430981293043e-05, + "loss": 0.0258, + "step": 77310 + }, + { + "epoch": 15.0106790932479, + "grad_norm": 0.11961174011230469, + "learning_rate": 1.3417157721665808e-05, + "loss": 0.0201, + "step": 77320 + }, + { + "epoch": 15.0126207465657, + "grad_norm": 0.21666842699050903, + "learning_rate": 1.3405888462683186e-05, + "loss": 0.0227, + "step": 77330 + }, + { + "epoch": 15.0145623998835, + "grad_norm": 0.1507921665906906, + "learning_rate": 1.3394623205577595e-05, + "loss": 0.0179, + "step": 77340 + }, + { + "epoch": 15.0165040532013, + "grad_norm": 0.2610705494880676, + "learning_rate": 1.3383361951580963e-05, + "loss": 0.0251, + "step": 77350 + }, + { + "epoch": 15.0184457065191, + "grad_norm": 0.13830021023750305, + "learning_rate": 1.337210470192483e-05, + "loss": 0.019, + "step": 77360 + }, + { + "epoch": 15.020387359836901, + "grad_norm": 0.16670866310596466, + "learning_rate": 1.3360851457840245e-05, + "loss": 0.0232, + "step": 77370 + }, + { + "epoch": 15.022329013154701, + "grad_norm": 0.2893085181713104, + "learning_rate": 1.3349602220557856e-05, + "loss": 0.022, + "step": 77380 + }, + { + "epoch": 15.024270666472502, + "grad_norm": 0.26028352975845337, + "learning_rate": 1.3338356991307865e-05, + "loss": 0.0195, + "step": 77390 + }, + { + "epoch": 15.026212319790302, + "grad_norm": 0.25368383526802063, + "learning_rate": 1.3327115771320048e-05, + "loss": 0.026, + "step": 77400 + }, + { + "epoch": 15.028153973108102, + "grad_norm": 0.14527176320552826, + "learning_rate": 1.3315878561823697e-05, + "loss": 0.0209, + "step": 77410 + }, + { + "epoch": 15.030095626425902, + "grad_norm": 0.17126449942588806, + "learning_rate": 1.330464536404773e-05, + "loss": 0.0197, + "step": 77420 + }, + { + "epoch": 15.032037279743701, + "grad_norm": 0.18281064927577972, + "learning_rate": 1.3293416179220563e-05, + "loss": 0.0198, + "step": 77430 + }, + { + "epoch": 15.033978933061501, + "grad_norm": 0.18316252529621124, + "learning_rate": 1.328219100857021e-05, + "loss": 0.0227, + "step": 77440 + }, + { + "epoch": 15.035920586379302, + "grad_norm": 0.151127427816391, + "learning_rate": 1.327096985332425e-05, + "loss": 0.0207, + "step": 77450 + }, + { + "epoch": 15.037862239697102, + "grad_norm": 0.27287301421165466, + "learning_rate": 1.3259752714709816e-05, + "loss": 0.0221, + "step": 77460 + }, + { + "epoch": 15.039803893014902, + "grad_norm": 0.30489271879196167, + "learning_rate": 1.3248539593953569e-05, + "loss": 0.0196, + "step": 77470 + }, + { + "epoch": 15.041745546332702, + "grad_norm": 0.2828127443790436, + "learning_rate": 1.323733049228178e-05, + "loss": 0.0362, + "step": 77480 + }, + { + "epoch": 15.043687199650503, + "grad_norm": 0.13159535825252533, + "learning_rate": 1.3226125410920276e-05, + "loss": 0.0252, + "step": 77490 + }, + { + "epoch": 15.045628852968303, + "grad_norm": 0.17011862993240356, + "learning_rate": 1.3214924351094388e-05, + "loss": 0.0236, + "step": 77500 + }, + { + "epoch": 15.047570506286103, + "grad_norm": 0.224320188164711, + "learning_rate": 1.3203727314029069e-05, + "loss": 0.0201, + "step": 77510 + }, + { + "epoch": 15.049512159603903, + "grad_norm": 0.07201744616031647, + "learning_rate": 1.3192534300948811e-05, + "loss": 0.0228, + "step": 77520 + }, + { + "epoch": 15.051453812921704, + "grad_norm": 0.12813086807727814, + "learning_rate": 1.3181345313077675e-05, + "loss": 0.0228, + "step": 77530 + }, + { + "epoch": 15.053395466239502, + "grad_norm": 0.10248711705207825, + "learning_rate": 1.3170160351639243e-05, + "loss": 0.0229, + "step": 77540 + }, + { + "epoch": 15.055337119557302, + "grad_norm": 0.28607577085494995, + "learning_rate": 1.3158979417856721e-05, + "loss": 0.0251, + "step": 77550 + }, + { + "epoch": 15.057278772875103, + "grad_norm": 0.16848532855510712, + "learning_rate": 1.31478025129528e-05, + "loss": 0.0278, + "step": 77560 + }, + { + "epoch": 15.059220426192903, + "grad_norm": 0.22780561447143555, + "learning_rate": 1.3136629638149794e-05, + "loss": 0.0186, + "step": 77570 + }, + { + "epoch": 15.061162079510703, + "grad_norm": 0.15498878061771393, + "learning_rate": 1.3125460794669547e-05, + "loss": 0.0314, + "step": 77580 + }, + { + "epoch": 15.063103732828504, + "grad_norm": 0.3611994683742523, + "learning_rate": 1.3114295983733483e-05, + "loss": 0.0228, + "step": 77590 + }, + { + "epoch": 15.065045386146304, + "grad_norm": 0.1641540825366974, + "learning_rate": 1.3103135206562534e-05, + "loss": 0.0253, + "step": 77600 + }, + { + "epoch": 15.066987039464104, + "grad_norm": 0.21672958135604858, + "learning_rate": 1.3091978464377264e-05, + "loss": 0.0199, + "step": 77610 + }, + { + "epoch": 15.068928692781904, + "grad_norm": 0.1581309288740158, + "learning_rate": 1.3080825758397724e-05, + "loss": 0.02, + "step": 77620 + }, + { + "epoch": 15.070870346099705, + "grad_norm": 0.2832014560699463, + "learning_rate": 1.3069677089843568e-05, + "loss": 0.0258, + "step": 77630 + }, + { + "epoch": 15.072811999417503, + "grad_norm": 0.2319786250591278, + "learning_rate": 1.305853245993401e-05, + "loss": 0.0246, + "step": 77640 + }, + { + "epoch": 15.074753652735303, + "grad_norm": 0.16743923723697662, + "learning_rate": 1.3047391869887809e-05, + "loss": 0.0251, + "step": 77650 + }, + { + "epoch": 15.076695306053104, + "grad_norm": 0.21079203486442566, + "learning_rate": 1.3036255320923263e-05, + "loss": 0.0232, + "step": 77660 + }, + { + "epoch": 15.078636959370904, + "grad_norm": 0.24657665193080902, + "learning_rate": 1.3025122814258273e-05, + "loss": 0.0191, + "step": 77670 + }, + { + "epoch": 15.080578612688704, + "grad_norm": 0.2284107804298401, + "learning_rate": 1.3013994351110249e-05, + "loss": 0.0239, + "step": 77680 + }, + { + "epoch": 15.082520266006505, + "grad_norm": 0.1676860898733139, + "learning_rate": 1.3002869932696194e-05, + "loss": 0.0241, + "step": 77690 + }, + { + "epoch": 15.084461919324305, + "grad_norm": 0.229390949010849, + "learning_rate": 1.2991749560232664e-05, + "loss": 0.0167, + "step": 77700 + }, + { + "epoch": 15.086403572642105, + "grad_norm": 0.09328868240118027, + "learning_rate": 1.2980633234935768e-05, + "loss": 0.0158, + "step": 77710 + }, + { + "epoch": 15.088345225959905, + "grad_norm": 0.11601803451776505, + "learning_rate": 1.2969520958021153e-05, + "loss": 0.0258, + "step": 77720 + }, + { + "epoch": 15.090286879277706, + "grad_norm": 0.2348545789718628, + "learning_rate": 1.295841273070405e-05, + "loss": 0.0192, + "step": 77730 + }, + { + "epoch": 15.092228532595506, + "grad_norm": 0.44421032071113586, + "learning_rate": 1.2947308554199255e-05, + "loss": 0.0274, + "step": 77740 + }, + { + "epoch": 15.094170185913304, + "grad_norm": 0.1973612755537033, + "learning_rate": 1.2936208429721075e-05, + "loss": 0.0299, + "step": 77750 + }, + { + "epoch": 15.096111839231105, + "grad_norm": 0.19022992253303528, + "learning_rate": 1.2925112358483416e-05, + "loss": 0.0247, + "step": 77760 + }, + { + "epoch": 15.098053492548905, + "grad_norm": 0.13283589482307434, + "learning_rate": 1.2914020341699728e-05, + "loss": 0.0206, + "step": 77770 + }, + { + "epoch": 15.099995145866705, + "grad_norm": 0.22621259093284607, + "learning_rate": 1.2902932380583032e-05, + "loss": 0.0251, + "step": 77780 + }, + { + "epoch": 15.101936799184505, + "grad_norm": 0.12780585885047913, + "learning_rate": 1.2891848476345863e-05, + "loss": 0.0207, + "step": 77790 + }, + { + "epoch": 15.103878452502306, + "grad_norm": 0.23102115094661713, + "learning_rate": 1.2880768630200368e-05, + "loss": 0.0204, + "step": 77800 + }, + { + "epoch": 15.105820105820106, + "grad_norm": 0.11325860023498535, + "learning_rate": 1.286969284335819e-05, + "loss": 0.0199, + "step": 77810 + }, + { + "epoch": 15.107761759137906, + "grad_norm": 0.1530570238828659, + "learning_rate": 1.2858621117030578e-05, + "loss": 0.0189, + "step": 77820 + }, + { + "epoch": 15.109703412455707, + "grad_norm": 0.2629915177822113, + "learning_rate": 1.2847553452428324e-05, + "loss": 0.0226, + "step": 77830 + }, + { + "epoch": 15.111645065773507, + "grad_norm": 0.37749263644218445, + "learning_rate": 1.283648985076178e-05, + "loss": 0.0233, + "step": 77840 + }, + { + "epoch": 15.113586719091305, + "grad_norm": 0.15647700428962708, + "learning_rate": 1.2825430313240817e-05, + "loss": 0.0219, + "step": 77850 + }, + { + "epoch": 15.115528372409106, + "grad_norm": 0.21560393273830414, + "learning_rate": 1.281437484107491e-05, + "loss": 0.0227, + "step": 77860 + }, + { + "epoch": 15.117470025726906, + "grad_norm": 0.2632741332054138, + "learning_rate": 1.2803323435473058e-05, + "loss": 0.025, + "step": 77870 + }, + { + "epoch": 15.119411679044706, + "grad_norm": 0.15077677369117737, + "learning_rate": 1.2792276097643841e-05, + "loss": 0.0235, + "step": 77880 + }, + { + "epoch": 15.121353332362506, + "grad_norm": 0.19968076050281525, + "learning_rate": 1.2781232828795342e-05, + "loss": 0.0204, + "step": 77890 + }, + { + "epoch": 15.123294985680307, + "grad_norm": 0.08824581652879715, + "learning_rate": 1.2770193630135292e-05, + "loss": 0.0216, + "step": 77900 + }, + { + "epoch": 15.125236638998107, + "grad_norm": 0.17312371730804443, + "learning_rate": 1.2759158502870877e-05, + "loss": 0.0203, + "step": 77910 + }, + { + "epoch": 15.127178292315907, + "grad_norm": 0.168802872300148, + "learning_rate": 1.2748127448208913e-05, + "loss": 0.0218, + "step": 77920 + }, + { + "epoch": 15.129119945633708, + "grad_norm": 0.2507579028606415, + "learning_rate": 1.2737100467355706e-05, + "loss": 0.024, + "step": 77930 + }, + { + "epoch": 15.131061598951508, + "grad_norm": 0.2572559416294098, + "learning_rate": 1.2726077561517169e-05, + "loss": 0.0193, + "step": 77940 + }, + { + "epoch": 15.133003252269308, + "grad_norm": 0.16580712795257568, + "learning_rate": 1.2715058731898755e-05, + "loss": 0.0215, + "step": 77950 + }, + { + "epoch": 15.134944905587107, + "grad_norm": 0.18251627683639526, + "learning_rate": 1.2704043979705471e-05, + "loss": 0.0228, + "step": 77960 + }, + { + "epoch": 15.136886558904907, + "grad_norm": 0.14203502237796783, + "learning_rate": 1.269303330614185e-05, + "loss": 0.0243, + "step": 77970 + }, + { + "epoch": 15.138828212222707, + "grad_norm": 0.20812532305717468, + "learning_rate": 1.2682026712412015e-05, + "loss": 0.0198, + "step": 77980 + }, + { + "epoch": 15.140769865540507, + "grad_norm": 0.21539635956287384, + "learning_rate": 1.2671024199719644e-05, + "loss": 0.0209, + "step": 77990 + }, + { + "epoch": 15.142711518858308, + "grad_norm": 0.1530768722295761, + "learning_rate": 1.2660025769267936e-05, + "loss": 0.0213, + "step": 78000 + }, + { + "epoch": 15.144653172176108, + "grad_norm": 0.24147693812847137, + "learning_rate": 1.2649031422259667e-05, + "loss": 0.0198, + "step": 78010 + }, + { + "epoch": 15.146594825493908, + "grad_norm": 0.6158980131149292, + "learning_rate": 1.2638041159897163e-05, + "loss": 0.023, + "step": 78020 + }, + { + "epoch": 15.148536478811709, + "grad_norm": 0.2568095028400421, + "learning_rate": 1.262705498338232e-05, + "loss": 0.0224, + "step": 78030 + }, + { + "epoch": 15.150478132129509, + "grad_norm": 0.1479850560426712, + "learning_rate": 1.2616072893916542e-05, + "loss": 0.0338, + "step": 78040 + }, + { + "epoch": 15.152419785447309, + "grad_norm": 0.2537541091442108, + "learning_rate": 1.2605094892700836e-05, + "loss": 0.0268, + "step": 78050 + }, + { + "epoch": 15.15436143876511, + "grad_norm": 0.18707573413848877, + "learning_rate": 1.259412098093572e-05, + "loss": 0.0188, + "step": 78060 + }, + { + "epoch": 15.156303092082908, + "grad_norm": 0.17571015655994415, + "learning_rate": 1.258315115982131e-05, + "loss": 0.0216, + "step": 78070 + }, + { + "epoch": 15.158244745400708, + "grad_norm": 0.21603715419769287, + "learning_rate": 1.2572185430557203e-05, + "loss": 0.0207, + "step": 78080 + }, + { + "epoch": 15.160186398718508, + "grad_norm": 0.17889223992824554, + "learning_rate": 1.2561223794342659e-05, + "loss": 0.02, + "step": 78090 + }, + { + "epoch": 15.162128052036309, + "grad_norm": 0.27247104048728943, + "learning_rate": 1.255026625237637e-05, + "loss": 0.0288, + "step": 78100 + }, + { + "epoch": 15.164069705354109, + "grad_norm": 0.08436835557222366, + "learning_rate": 1.2539312805856678e-05, + "loss": 0.022, + "step": 78110 + }, + { + "epoch": 15.16601135867191, + "grad_norm": 0.12900365889072418, + "learning_rate": 1.25283634559814e-05, + "loss": 0.0178, + "step": 78120 + }, + { + "epoch": 15.16795301198971, + "grad_norm": 0.22510264813899994, + "learning_rate": 1.2517418203947972e-05, + "loss": 0.0207, + "step": 78130 + }, + { + "epoch": 15.16989466530751, + "grad_norm": 0.27568525075912476, + "learning_rate": 1.2506477050953308e-05, + "loss": 0.0291, + "step": 78140 + }, + { + "epoch": 15.17183631862531, + "grad_norm": 0.22866278886795044, + "learning_rate": 1.2495539998193972e-05, + "loss": 0.0239, + "step": 78150 + }, + { + "epoch": 15.17377797194311, + "grad_norm": 0.16897134482860565, + "learning_rate": 1.2484607046865975e-05, + "loss": 0.0221, + "step": 78160 + }, + { + "epoch": 15.17571962526091, + "grad_norm": 0.1901993751525879, + "learning_rate": 1.2473678198164968e-05, + "loss": 0.0198, + "step": 78170 + }, + { + "epoch": 15.177661278578709, + "grad_norm": 0.20042477548122406, + "learning_rate": 1.2462753453286075e-05, + "loss": 0.0226, + "step": 78180 + }, + { + "epoch": 15.17960293189651, + "grad_norm": 0.2419590801000595, + "learning_rate": 1.2451832813424031e-05, + "loss": 0.0242, + "step": 78190 + }, + { + "epoch": 15.18154458521431, + "grad_norm": 0.1757732629776001, + "learning_rate": 1.2440916279773095e-05, + "loss": 0.0238, + "step": 78200 + }, + { + "epoch": 15.18348623853211, + "grad_norm": 0.12815086543560028, + "learning_rate": 1.2430003853527101e-05, + "loss": 0.0233, + "step": 78210 + }, + { + "epoch": 15.18542789184991, + "grad_norm": 0.22626207768917084, + "learning_rate": 1.2419095535879383e-05, + "loss": 0.0305, + "step": 78220 + }, + { + "epoch": 15.18736954516771, + "grad_norm": 0.15490520000457764, + "learning_rate": 1.2408191328022878e-05, + "loss": 0.0193, + "step": 78230 + }, + { + "epoch": 15.18931119848551, + "grad_norm": 0.17197874188423157, + "learning_rate": 1.2397291231150066e-05, + "loss": 0.026, + "step": 78240 + }, + { + "epoch": 15.191252851803311, + "grad_norm": 0.28220149874687195, + "learning_rate": 1.2386395246452937e-05, + "loss": 0.0194, + "step": 78250 + }, + { + "epoch": 15.193194505121111, + "grad_norm": 0.18602018058300018, + "learning_rate": 1.2375503375123083e-05, + "loss": 0.0241, + "step": 78260 + }, + { + "epoch": 15.195136158438912, + "grad_norm": 0.1285603791475296, + "learning_rate": 1.2364615618351583e-05, + "loss": 0.0159, + "step": 78270 + }, + { + "epoch": 15.19707781175671, + "grad_norm": 0.18452416360378265, + "learning_rate": 1.2353731977329169e-05, + "loss": 0.0219, + "step": 78280 + }, + { + "epoch": 15.19901946507451, + "grad_norm": 0.17182299494743347, + "learning_rate": 1.2342852453246007e-05, + "loss": 0.0322, + "step": 78290 + }, + { + "epoch": 15.20096111839231, + "grad_norm": 0.15768152475357056, + "learning_rate": 1.2331977047291898e-05, + "loss": 0.0212, + "step": 78300 + }, + { + "epoch": 15.20290277171011, + "grad_norm": 0.17670868337154388, + "learning_rate": 1.2321105760656137e-05, + "loss": 0.0226, + "step": 78310 + }, + { + "epoch": 15.204844425027911, + "grad_norm": 0.09531630575656891, + "learning_rate": 1.2310238594527613e-05, + "loss": 0.0214, + "step": 78320 + }, + { + "epoch": 15.206786078345711, + "grad_norm": 0.11374442279338837, + "learning_rate": 1.2299375550094693e-05, + "loss": 0.0233, + "step": 78330 + }, + { + "epoch": 15.208727731663512, + "grad_norm": 0.1743302047252655, + "learning_rate": 1.2288516628545421e-05, + "loss": 0.0264, + "step": 78340 + }, + { + "epoch": 15.210669384981312, + "grad_norm": 0.2489149570465088, + "learning_rate": 1.2277661831067255e-05, + "loss": 0.0206, + "step": 78350 + }, + { + "epoch": 15.212611038299112, + "grad_norm": 0.12601758539676666, + "learning_rate": 1.2266811158847286e-05, + "loss": 0.0162, + "step": 78360 + }, + { + "epoch": 15.214552691616912, + "grad_norm": 0.2822719216346741, + "learning_rate": 1.2255964613072107e-05, + "loss": 0.0196, + "step": 78370 + }, + { + "epoch": 15.216494344934713, + "grad_norm": 0.24450689554214478, + "learning_rate": 1.2245122194927905e-05, + "loss": 0.0252, + "step": 78380 + }, + { + "epoch": 15.218435998252511, + "grad_norm": 0.14436474442481995, + "learning_rate": 1.2234283905600364e-05, + "loss": 0.0233, + "step": 78390 + }, + { + "epoch": 15.220377651570312, + "grad_norm": 0.21703706681728363, + "learning_rate": 1.222344974627475e-05, + "loss": 0.0201, + "step": 78400 + }, + { + "epoch": 15.222319304888112, + "grad_norm": 0.2311210036277771, + "learning_rate": 1.2212619718135875e-05, + "loss": 0.0209, + "step": 78410 + }, + { + "epoch": 15.224260958205912, + "grad_norm": 0.12827658653259277, + "learning_rate": 1.2201793822368118e-05, + "loss": 0.0261, + "step": 78420 + }, + { + "epoch": 15.226202611523712, + "grad_norm": 0.16806954145431519, + "learning_rate": 1.2190972060155337e-05, + "loss": 0.0215, + "step": 78430 + }, + { + "epoch": 15.228144264841513, + "grad_norm": 0.1698845624923706, + "learning_rate": 1.2180154432681013e-05, + "loss": 0.0212, + "step": 78440 + }, + { + "epoch": 15.230085918159313, + "grad_norm": 0.23690670728683472, + "learning_rate": 1.2169340941128148e-05, + "loss": 0.03, + "step": 78450 + }, + { + "epoch": 15.232027571477113, + "grad_norm": 0.22398559749126434, + "learning_rate": 1.2158531586679267e-05, + "loss": 0.0174, + "step": 78460 + }, + { + "epoch": 15.233969224794913, + "grad_norm": 0.13434240221977234, + "learning_rate": 1.2147726370516477e-05, + "loss": 0.0181, + "step": 78470 + }, + { + "epoch": 15.235910878112714, + "grad_norm": 0.2002863734960556, + "learning_rate": 1.2136925293821422e-05, + "loss": 0.0236, + "step": 78480 + }, + { + "epoch": 15.237852531430512, + "grad_norm": 0.13015638291835785, + "learning_rate": 1.2126128357775306e-05, + "loss": 0.0224, + "step": 78490 + }, + { + "epoch": 15.239794184748312, + "grad_norm": 0.17105330526828766, + "learning_rate": 1.2115335563558838e-05, + "loss": 0.0258, + "step": 78500 + }, + { + "epoch": 15.241735838066113, + "grad_norm": 0.22366969287395477, + "learning_rate": 1.2104546912352327e-05, + "loss": 0.0232, + "step": 78510 + }, + { + "epoch": 15.243677491383913, + "grad_norm": 0.10250787436962128, + "learning_rate": 1.2093762405335557e-05, + "loss": 0.0202, + "step": 78520 + }, + { + "epoch": 15.245619144701713, + "grad_norm": 0.15276993811130524, + "learning_rate": 1.2082982043687974e-05, + "loss": 0.0237, + "step": 78530 + }, + { + "epoch": 15.247560798019514, + "grad_norm": 0.1593102365732193, + "learning_rate": 1.207220582858845e-05, + "loss": 0.0167, + "step": 78540 + }, + { + "epoch": 15.249502451337314, + "grad_norm": 0.22556661069393158, + "learning_rate": 1.206143376121549e-05, + "loss": 0.0226, + "step": 78550 + }, + { + "epoch": 15.251444104655114, + "grad_norm": 0.24943745136260986, + "learning_rate": 1.2050665842747078e-05, + "loss": 0.0238, + "step": 78560 + }, + { + "epoch": 15.253385757972914, + "grad_norm": 0.14904889464378357, + "learning_rate": 1.2039902074360804e-05, + "loss": 0.0185, + "step": 78570 + }, + { + "epoch": 15.255327411290715, + "grad_norm": 0.1389000564813614, + "learning_rate": 1.2029142457233755e-05, + "loss": 0.0202, + "step": 78580 + }, + { + "epoch": 15.257269064608515, + "grad_norm": 0.16324874758720398, + "learning_rate": 1.2018386992542601e-05, + "loss": 0.0227, + "step": 78590 + }, + { + "epoch": 15.259210717926313, + "grad_norm": 0.22057537734508514, + "learning_rate": 1.2007635681463542e-05, + "loss": 0.0275, + "step": 78600 + }, + { + "epoch": 15.261152371244114, + "grad_norm": 0.20916011929512024, + "learning_rate": 1.1996888525172334e-05, + "loss": 0.0227, + "step": 78610 + }, + { + "epoch": 15.263094024561914, + "grad_norm": 0.13061530888080597, + "learning_rate": 1.198614552484425e-05, + "loss": 0.0215, + "step": 78620 + }, + { + "epoch": 15.265035677879714, + "grad_norm": 0.19035488367080688, + "learning_rate": 1.197540668165415e-05, + "loss": 0.0232, + "step": 78630 + }, + { + "epoch": 15.266977331197515, + "grad_norm": 0.18029984831809998, + "learning_rate": 1.1964671996776395e-05, + "loss": 0.0196, + "step": 78640 + }, + { + "epoch": 15.268918984515315, + "grad_norm": 0.3519718050956726, + "learning_rate": 1.1953941471384922e-05, + "loss": 0.0327, + "step": 78650 + }, + { + "epoch": 15.270860637833115, + "grad_norm": 0.16479119658470154, + "learning_rate": 1.1943215106653205e-05, + "loss": 0.0245, + "step": 78660 + }, + { + "epoch": 15.272802291150915, + "grad_norm": 0.15331268310546875, + "learning_rate": 1.193249290375429e-05, + "loss": 0.02, + "step": 78670 + }, + { + "epoch": 15.274743944468716, + "grad_norm": 0.18724985420703888, + "learning_rate": 1.1921774863860696e-05, + "loss": 0.0242, + "step": 78680 + }, + { + "epoch": 15.276685597786516, + "grad_norm": 0.30903780460357666, + "learning_rate": 1.1911060988144556e-05, + "loss": 0.0305, + "step": 78690 + }, + { + "epoch": 15.278627251104314, + "grad_norm": 0.2026868611574173, + "learning_rate": 1.1900351277777533e-05, + "loss": 0.0235, + "step": 78700 + }, + { + "epoch": 15.280568904422115, + "grad_norm": 0.1917039155960083, + "learning_rate": 1.1889645733930798e-05, + "loss": 0.0197, + "step": 78710 + }, + { + "epoch": 15.282510557739915, + "grad_norm": 0.21632236242294312, + "learning_rate": 1.1878944357775107e-05, + "loss": 0.0217, + "step": 78720 + }, + { + "epoch": 15.284452211057715, + "grad_norm": 0.29926469922065735, + "learning_rate": 1.1868247150480749e-05, + "loss": 0.0182, + "step": 78730 + }, + { + "epoch": 15.286393864375516, + "grad_norm": 0.3954163193702698, + "learning_rate": 1.1857554113217567e-05, + "loss": 0.0194, + "step": 78740 + }, + { + "epoch": 15.288335517693316, + "grad_norm": 0.33582475781440735, + "learning_rate": 1.1846865247154903e-05, + "loss": 0.0193, + "step": 78750 + }, + { + "epoch": 15.290277171011116, + "grad_norm": 0.23251838982105255, + "learning_rate": 1.1836180553461711e-05, + "loss": 0.0175, + "step": 78760 + }, + { + "epoch": 15.292218824328916, + "grad_norm": 0.2847832143306732, + "learning_rate": 1.1825500033306419e-05, + "loss": 0.0252, + "step": 78770 + }, + { + "epoch": 15.294160477646717, + "grad_norm": 0.20446448028087616, + "learning_rate": 1.1814823687857052e-05, + "loss": 0.0232, + "step": 78780 + }, + { + "epoch": 15.296102130964517, + "grad_norm": 0.1587536484003067, + "learning_rate": 1.1804151518281153e-05, + "loss": 0.02, + "step": 78790 + }, + { + "epoch": 15.298043784282317, + "grad_norm": 0.06163037568330765, + "learning_rate": 1.179348352574583e-05, + "loss": 0.0182, + "step": 78800 + }, + { + "epoch": 15.299985437600116, + "grad_norm": 0.2572789788246155, + "learning_rate": 1.1782819711417697e-05, + "loss": 0.0203, + "step": 78810 + }, + { + "epoch": 15.301927090917916, + "grad_norm": 0.12890255451202393, + "learning_rate": 1.1772160076462951e-05, + "loss": 0.0191, + "step": 78820 + }, + { + "epoch": 15.303868744235716, + "grad_norm": 0.28545504808425903, + "learning_rate": 1.1761504622047293e-05, + "loss": 0.0265, + "step": 78830 + }, + { + "epoch": 15.305810397553516, + "grad_norm": 0.215108260512352, + "learning_rate": 1.1750853349335999e-05, + "loss": 0.0171, + "step": 78840 + }, + { + "epoch": 15.307752050871317, + "grad_norm": 0.3404950499534607, + "learning_rate": 1.1740206259493874e-05, + "loss": 0.0172, + "step": 78850 + }, + { + "epoch": 15.309693704189117, + "grad_norm": 0.1302623152732849, + "learning_rate": 1.1729563353685286e-05, + "loss": 0.0215, + "step": 78860 + }, + { + "epoch": 15.311635357506917, + "grad_norm": 0.18044693768024445, + "learning_rate": 1.1718924633074091e-05, + "loss": 0.0199, + "step": 78870 + }, + { + "epoch": 15.313577010824718, + "grad_norm": 0.0893653929233551, + "learning_rate": 1.1708290098823765e-05, + "loss": 0.0183, + "step": 78880 + }, + { + "epoch": 15.315518664142518, + "grad_norm": 0.2400577962398529, + "learning_rate": 1.1697659752097245e-05, + "loss": 0.0256, + "step": 78890 + }, + { + "epoch": 15.317460317460318, + "grad_norm": 0.16649484634399414, + "learning_rate": 1.1687033594057068e-05, + "loss": 0.0193, + "step": 78900 + }, + { + "epoch": 15.319401970778118, + "grad_norm": 0.18193529546260834, + "learning_rate": 1.1676411625865297e-05, + "loss": 0.0229, + "step": 78910 + }, + { + "epoch": 15.321343624095917, + "grad_norm": 0.18609605729579926, + "learning_rate": 1.1665793848683542e-05, + "loss": 0.0172, + "step": 78920 + }, + { + "epoch": 15.323285277413717, + "grad_norm": 0.2781553268432617, + "learning_rate": 1.1655180263672927e-05, + "loss": 0.0217, + "step": 78930 + }, + { + "epoch": 15.325226930731517, + "grad_norm": 0.22263483703136444, + "learning_rate": 1.1644570871994142e-05, + "loss": 0.016, + "step": 78940 + }, + { + "epoch": 15.327168584049318, + "grad_norm": 0.19160422682762146, + "learning_rate": 1.163396567480744e-05, + "loss": 0.0249, + "step": 78950 + }, + { + "epoch": 15.329110237367118, + "grad_norm": 0.13120301067829132, + "learning_rate": 1.1623364673272552e-05, + "loss": 0.0237, + "step": 78960 + }, + { + "epoch": 15.331051890684918, + "grad_norm": 0.12279228866100311, + "learning_rate": 1.1612767868548807e-05, + "loss": 0.0215, + "step": 78970 + }, + { + "epoch": 15.332993544002719, + "grad_norm": 0.204635888338089, + "learning_rate": 1.1602175261795057e-05, + "loss": 0.0217, + "step": 78980 + }, + { + "epoch": 15.334935197320519, + "grad_norm": 0.1224849745631218, + "learning_rate": 1.1591586854169696e-05, + "loss": 0.0194, + "step": 78990 + }, + { + "epoch": 15.336876850638319, + "grad_norm": 0.38813021779060364, + "learning_rate": 1.158100264683064e-05, + "loss": 0.0301, + "step": 79000 + }, + { + "epoch": 15.33881850395612, + "grad_norm": 0.18216897547245026, + "learning_rate": 1.1570422640935386e-05, + "loss": 0.0184, + "step": 79010 + }, + { + "epoch": 15.34076015727392, + "grad_norm": 0.2259826958179474, + "learning_rate": 1.155984683764092e-05, + "loss": 0.0236, + "step": 79020 + }, + { + "epoch": 15.342701810591718, + "grad_norm": 0.18563932180404663, + "learning_rate": 1.1549275238103812e-05, + "loss": 0.0206, + "step": 79030 + }, + { + "epoch": 15.344643463909518, + "grad_norm": 0.1795501410961151, + "learning_rate": 1.1538707843480152e-05, + "loss": 0.0191, + "step": 79040 + }, + { + "epoch": 15.346585117227319, + "grad_norm": 0.20315276086330414, + "learning_rate": 1.1528144654925593e-05, + "loss": 0.0276, + "step": 79050 + }, + { + "epoch": 15.348526770545119, + "grad_norm": 0.23087328672409058, + "learning_rate": 1.1517585673595277e-05, + "loss": 0.0229, + "step": 79060 + }, + { + "epoch": 15.35046842386292, + "grad_norm": 0.11431434750556946, + "learning_rate": 1.150703090064395e-05, + "loss": 0.0199, + "step": 79070 + }, + { + "epoch": 15.35241007718072, + "grad_norm": 0.12729449570178986, + "learning_rate": 1.1496480337225835e-05, + "loss": 0.0161, + "step": 79080 + }, + { + "epoch": 15.35435173049852, + "grad_norm": 0.1526130586862564, + "learning_rate": 1.1485933984494751e-05, + "loss": 0.0243, + "step": 79090 + }, + { + "epoch": 15.35629338381632, + "grad_norm": 0.17466090619564056, + "learning_rate": 1.1475391843604e-05, + "loss": 0.0226, + "step": 79100 + }, + { + "epoch": 15.35823503713412, + "grad_norm": 0.20463767647743225, + "learning_rate": 1.1464853915706503e-05, + "loss": 0.0232, + "step": 79110 + }, + { + "epoch": 15.36017669045192, + "grad_norm": 0.10285758972167969, + "learning_rate": 1.1454320201954626e-05, + "loss": 0.0221, + "step": 79120 + }, + { + "epoch": 15.362118343769719, + "grad_norm": 0.24148720502853394, + "learning_rate": 1.1443790703500356e-05, + "loss": 0.0177, + "step": 79130 + }, + { + "epoch": 15.36405999708752, + "grad_norm": 0.18488281965255737, + "learning_rate": 1.1433265421495154e-05, + "loss": 0.023, + "step": 79140 + }, + { + "epoch": 15.36600165040532, + "grad_norm": 0.23934412002563477, + "learning_rate": 1.1422744357090065e-05, + "loss": 0.023, + "step": 79150 + }, + { + "epoch": 15.36794330372312, + "grad_norm": 0.1201092004776001, + "learning_rate": 1.1412227511435647e-05, + "loss": 0.0231, + "step": 79160 + }, + { + "epoch": 15.36988495704092, + "grad_norm": 0.2364889234304428, + "learning_rate": 1.1401714885682025e-05, + "loss": 0.0234, + "step": 79170 + }, + { + "epoch": 15.37182661035872, + "grad_norm": 0.46381720900535583, + "learning_rate": 1.1391206480978823e-05, + "loss": 0.0245, + "step": 79180 + }, + { + "epoch": 15.37376826367652, + "grad_norm": 0.12768013775348663, + "learning_rate": 1.1380702298475227e-05, + "loss": 0.0191, + "step": 79190 + }, + { + "epoch": 15.375709916994321, + "grad_norm": 0.18825659155845642, + "learning_rate": 1.1370202339319985e-05, + "loss": 0.0211, + "step": 79200 + }, + { + "epoch": 15.377651570312121, + "grad_norm": 0.2251601666212082, + "learning_rate": 1.1359706604661319e-05, + "loss": 0.0255, + "step": 79210 + }, + { + "epoch": 15.379593223629922, + "grad_norm": 0.19644112884998322, + "learning_rate": 1.1349215095647042e-05, + "loss": 0.0204, + "step": 79220 + }, + { + "epoch": 15.381534876947722, + "grad_norm": 0.2128390520811081, + "learning_rate": 1.1338727813424493e-05, + "loss": 0.0271, + "step": 79230 + }, + { + "epoch": 15.38347653026552, + "grad_norm": 0.15784533321857452, + "learning_rate": 1.132824475914056e-05, + "loss": 0.0226, + "step": 79240 + }, + { + "epoch": 15.38541818358332, + "grad_norm": 0.27361586689949036, + "learning_rate": 1.131776593394162e-05, + "loss": 0.024, + "step": 79250 + }, + { + "epoch": 15.38735983690112, + "grad_norm": 0.20661956071853638, + "learning_rate": 1.1307291338973657e-05, + "loss": 0.0216, + "step": 79260 + }, + { + "epoch": 15.389301490218921, + "grad_norm": 0.2264523208141327, + "learning_rate": 1.1296820975382121e-05, + "loss": 0.0244, + "step": 79270 + }, + { + "epoch": 15.391243143536721, + "grad_norm": 0.2568355202674866, + "learning_rate": 1.1286354844312074e-05, + "loss": 0.0229, + "step": 79280 + }, + { + "epoch": 15.393184796854522, + "grad_norm": 0.175021231174469, + "learning_rate": 1.1275892946908023e-05, + "loss": 0.0224, + "step": 79290 + }, + { + "epoch": 15.395126450172322, + "grad_norm": 0.16837817430496216, + "learning_rate": 1.126543528431413e-05, + "loss": 0.0236, + "step": 79300 + }, + { + "epoch": 15.397068103490122, + "grad_norm": 0.08889447152614594, + "learning_rate": 1.125498185767398e-05, + "loss": 0.0177, + "step": 79310 + }, + { + "epoch": 15.399009756807922, + "grad_norm": 0.12000635266304016, + "learning_rate": 1.1244532668130781e-05, + "loss": 0.0215, + "step": 79320 + }, + { + "epoch": 15.400951410125723, + "grad_norm": 0.17899420857429504, + "learning_rate": 1.1234087716827203e-05, + "loss": 0.0229, + "step": 79330 + }, + { + "epoch": 15.402893063443521, + "grad_norm": 0.20883740484714508, + "learning_rate": 1.1223647004905524e-05, + "loss": 0.0178, + "step": 79340 + }, + { + "epoch": 15.404834716761322, + "grad_norm": 0.1720387190580368, + "learning_rate": 1.1213210533507485e-05, + "loss": 0.023, + "step": 79350 + }, + { + "epoch": 15.406776370079122, + "grad_norm": 0.15996694564819336, + "learning_rate": 1.1202778303774447e-05, + "loss": 0.022, + "step": 79360 + }, + { + "epoch": 15.408718023396922, + "grad_norm": 0.16699685156345367, + "learning_rate": 1.1192350316847234e-05, + "loss": 0.0212, + "step": 79370 + }, + { + "epoch": 15.410659676714722, + "grad_norm": 0.17236189544200897, + "learning_rate": 1.1181926573866258e-05, + "loss": 0.0242, + "step": 79380 + }, + { + "epoch": 15.412601330032523, + "grad_norm": 0.211061030626297, + "learning_rate": 1.1171507075971416e-05, + "loss": 0.0211, + "step": 79390 + }, + { + "epoch": 15.414542983350323, + "grad_norm": 0.2324594408273697, + "learning_rate": 1.1161091824302177e-05, + "loss": 0.0236, + "step": 79400 + }, + { + "epoch": 15.416484636668123, + "grad_norm": 0.1455221176147461, + "learning_rate": 1.115068081999755e-05, + "loss": 0.02, + "step": 79410 + }, + { + "epoch": 15.418426289985923, + "grad_norm": 0.16639572381973267, + "learning_rate": 1.1140274064196071e-05, + "loss": 0.023, + "step": 79420 + }, + { + "epoch": 15.420367943303724, + "grad_norm": 0.13764923810958862, + "learning_rate": 1.112987155803578e-05, + "loss": 0.0255, + "step": 79430 + }, + { + "epoch": 15.422309596621524, + "grad_norm": 0.19674080610275269, + "learning_rate": 1.1119473302654298e-05, + "loss": 0.0234, + "step": 79440 + }, + { + "epoch": 15.424251249939323, + "grad_norm": 0.20278030633926392, + "learning_rate": 1.1109079299188769e-05, + "loss": 0.0225, + "step": 79450 + }, + { + "epoch": 15.426192903257123, + "grad_norm": 0.1532328873872757, + "learning_rate": 1.1098689548775847e-05, + "loss": 0.0245, + "step": 79460 + }, + { + "epoch": 15.428134556574923, + "grad_norm": 0.15466096997261047, + "learning_rate": 1.1088304052551762e-05, + "loss": 0.0245, + "step": 79470 + }, + { + "epoch": 15.430076209892723, + "grad_norm": 0.22263599932193756, + "learning_rate": 1.1077922811652215e-05, + "loss": 0.0254, + "step": 79480 + }, + { + "epoch": 15.432017863210524, + "grad_norm": 0.14287257194519043, + "learning_rate": 1.1067545827212534e-05, + "loss": 0.0196, + "step": 79490 + }, + { + "epoch": 15.433959516528324, + "grad_norm": 0.15848177671432495, + "learning_rate": 1.1057173100367496e-05, + "loss": 0.02, + "step": 79500 + }, + { + "epoch": 15.435901169846124, + "grad_norm": 0.2544611096382141, + "learning_rate": 1.1046804632251474e-05, + "loss": 0.0217, + "step": 79510 + }, + { + "epoch": 15.437842823163924, + "grad_norm": 0.10041555017232895, + "learning_rate": 1.1036440423998318e-05, + "loss": 0.0198, + "step": 79520 + }, + { + "epoch": 15.439784476481725, + "grad_norm": 0.12479431182146072, + "learning_rate": 1.1026080476741469e-05, + "loss": 0.022, + "step": 79530 + }, + { + "epoch": 15.441726129799525, + "grad_norm": 0.15311545133590698, + "learning_rate": 1.1015724791613836e-05, + "loss": 0.0225, + "step": 79540 + }, + { + "epoch": 15.443667783117325, + "grad_norm": 0.13533198833465576, + "learning_rate": 1.1005373369747951e-05, + "loss": 0.0217, + "step": 79550 + }, + { + "epoch": 15.445609436435124, + "grad_norm": 0.14206476509571075, + "learning_rate": 1.09950262122758e-05, + "loss": 0.0201, + "step": 79560 + }, + { + "epoch": 15.447551089752924, + "grad_norm": 0.21699835360050201, + "learning_rate": 1.0984683320328948e-05, + "loss": 0.0257, + "step": 79570 + }, + { + "epoch": 15.449492743070724, + "grad_norm": 0.1568334698677063, + "learning_rate": 1.0974344695038458e-05, + "loss": 0.0312, + "step": 79580 + }, + { + "epoch": 15.451434396388525, + "grad_norm": 0.12019465118646622, + "learning_rate": 1.0964010337534968e-05, + "loss": 0.0192, + "step": 79590 + }, + { + "epoch": 15.453376049706325, + "grad_norm": 0.10807188600301743, + "learning_rate": 1.0953680248948611e-05, + "loss": 0.0176, + "step": 79600 + }, + { + "epoch": 15.455317703024125, + "grad_norm": 0.21880221366882324, + "learning_rate": 1.0943354430409069e-05, + "loss": 0.0211, + "step": 79610 + }, + { + "epoch": 15.457259356341925, + "grad_norm": 0.21225497126579285, + "learning_rate": 1.0933032883045574e-05, + "loss": 0.0208, + "step": 79620 + }, + { + "epoch": 15.459201009659726, + "grad_norm": 0.18900218605995178, + "learning_rate": 1.0922715607986872e-05, + "loss": 0.0326, + "step": 79630 + }, + { + "epoch": 15.461142662977526, + "grad_norm": 0.09742976725101471, + "learning_rate": 1.0912402606361234e-05, + "loss": 0.0225, + "step": 79640 + }, + { + "epoch": 15.463084316295326, + "grad_norm": 0.1562129408121109, + "learning_rate": 1.090209387929647e-05, + "loss": 0.0252, + "step": 79650 + }, + { + "epoch": 15.465025969613126, + "grad_norm": 0.14043214917182922, + "learning_rate": 1.0891789427919957e-05, + "loss": 0.0204, + "step": 79660 + }, + { + "epoch": 15.466967622930925, + "grad_norm": 0.18205557763576508, + "learning_rate": 1.088148925335854e-05, + "loss": 0.0172, + "step": 79670 + }, + { + "epoch": 15.468909276248725, + "grad_norm": 0.1559341698884964, + "learning_rate": 1.087119335673864e-05, + "loss": 0.0208, + "step": 79680 + }, + { + "epoch": 15.470850929566526, + "grad_norm": 0.19294866919517517, + "learning_rate": 1.0860901739186208e-05, + "loss": 0.021, + "step": 79690 + }, + { + "epoch": 15.472792582884326, + "grad_norm": 0.20131006836891174, + "learning_rate": 1.085061440182673e-05, + "loss": 0.0209, + "step": 79700 + }, + { + "epoch": 15.474734236202126, + "grad_norm": 0.19990423321723938, + "learning_rate": 1.0840331345785193e-05, + "loss": 0.0246, + "step": 79710 + }, + { + "epoch": 15.476675889519926, + "grad_norm": 0.09354723244905472, + "learning_rate": 1.083005257218615e-05, + "loss": 0.0277, + "step": 79720 + }, + { + "epoch": 15.478617542837727, + "grad_norm": 0.0961817279458046, + "learning_rate": 1.0819778082153642e-05, + "loss": 0.0241, + "step": 79730 + }, + { + "epoch": 15.480559196155527, + "grad_norm": 0.1655864119529724, + "learning_rate": 1.0809507876811325e-05, + "loss": 0.0227, + "step": 79740 + }, + { + "epoch": 15.482500849473327, + "grad_norm": 0.24122077226638794, + "learning_rate": 1.0799241957282292e-05, + "loss": 0.0257, + "step": 79750 + }, + { + "epoch": 15.484442502791127, + "grad_norm": 0.0820898711681366, + "learning_rate": 1.078898032468923e-05, + "loss": 0.0225, + "step": 79760 + }, + { + "epoch": 15.486384156108926, + "grad_norm": 0.14904288947582245, + "learning_rate": 1.0778722980154315e-05, + "loss": 0.0256, + "step": 79770 + }, + { + "epoch": 15.488325809426726, + "grad_norm": 0.16661347448825836, + "learning_rate": 1.0768469924799302e-05, + "loss": 0.0212, + "step": 79780 + }, + { + "epoch": 15.490267462744526, + "grad_norm": 0.1996821016073227, + "learning_rate": 1.0758221159745419e-05, + "loss": 0.0178, + "step": 79790 + }, + { + "epoch": 15.492209116062327, + "grad_norm": 0.3992796540260315, + "learning_rate": 1.074797668611347e-05, + "loss": 0.0199, + "step": 79800 + }, + { + "epoch": 15.494150769380127, + "grad_norm": 0.28043410181999207, + "learning_rate": 1.0737736505023777e-05, + "loss": 0.0261, + "step": 79810 + }, + { + "epoch": 15.496092422697927, + "grad_norm": 0.1640961766242981, + "learning_rate": 1.0727500617596203e-05, + "loss": 0.0249, + "step": 79820 + }, + { + "epoch": 15.498034076015728, + "grad_norm": 0.13006259500980377, + "learning_rate": 1.0717269024950104e-05, + "loss": 0.0246, + "step": 79830 + }, + { + "epoch": 15.499975729333528, + "grad_norm": 0.17387883365154266, + "learning_rate": 1.070704172820442e-05, + "loss": 0.0206, + "step": 79840 + }, + { + "epoch": 15.501917382651328, + "grad_norm": 0.18474045395851135, + "learning_rate": 1.0696818728477559e-05, + "loss": 0.0259, + "step": 79850 + }, + { + "epoch": 15.503859035969128, + "grad_norm": 0.26657381653785706, + "learning_rate": 1.0686600026887511e-05, + "loss": 0.0297, + "step": 79860 + }, + { + "epoch": 15.505800689286929, + "grad_norm": 0.2484980821609497, + "learning_rate": 1.0676385624551777e-05, + "loss": 0.018, + "step": 79870 + }, + { + "epoch": 15.507742342604727, + "grad_norm": 0.2373620867729187, + "learning_rate": 1.0666175522587402e-05, + "loss": 0.0235, + "step": 79880 + }, + { + "epoch": 15.509683995922527, + "grad_norm": 0.08786848187446594, + "learning_rate": 1.0655969722110925e-05, + "loss": 0.0209, + "step": 79890 + }, + { + "epoch": 15.511625649240328, + "grad_norm": 0.1725875288248062, + "learning_rate": 1.0645768224238439e-05, + "loss": 0.0299, + "step": 79900 + }, + { + "epoch": 15.513567302558128, + "grad_norm": 0.11897052824497223, + "learning_rate": 1.063557103008559e-05, + "loss": 0.0191, + "step": 79910 + }, + { + "epoch": 15.515508955875928, + "grad_norm": 0.27943968772888184, + "learning_rate": 1.0625378140767494e-05, + "loss": 0.0189, + "step": 79920 + }, + { + "epoch": 15.517450609193729, + "grad_norm": 0.15767920017242432, + "learning_rate": 1.0615189557398846e-05, + "loss": 0.0203, + "step": 79930 + }, + { + "epoch": 15.519392262511529, + "grad_norm": 0.17832493782043457, + "learning_rate": 1.060500528109386e-05, + "loss": 0.0179, + "step": 79940 + }, + { + "epoch": 15.521333915829329, + "grad_norm": 0.21693117916584015, + "learning_rate": 1.0594825312966272e-05, + "loss": 0.019, + "step": 79950 + }, + { + "epoch": 15.52327556914713, + "grad_norm": 0.05958405137062073, + "learning_rate": 1.0584649654129336e-05, + "loss": 0.0244, + "step": 79960 + }, + { + "epoch": 15.52521722246493, + "grad_norm": 0.11648624390363693, + "learning_rate": 1.0574478305695863e-05, + "loss": 0.0196, + "step": 79970 + }, + { + "epoch": 15.527158875782728, + "grad_norm": 0.14179491996765137, + "learning_rate": 1.0564311268778154e-05, + "loss": 0.0235, + "step": 79980 + }, + { + "epoch": 15.529100529100528, + "grad_norm": 0.1709437072277069, + "learning_rate": 1.0554148544488073e-05, + "loss": 0.0284, + "step": 79990 + }, + { + "epoch": 15.531042182418329, + "grad_norm": 0.1653372198343277, + "learning_rate": 1.0543990133937004e-05, + "loss": 0.0149, + "step": 80000 + }, + { + "epoch": 15.532983835736129, + "grad_norm": 0.23093944787979126, + "learning_rate": 1.0533836038235868e-05, + "loss": 0.0271, + "step": 80010 + }, + { + "epoch": 15.53492548905393, + "grad_norm": 0.3644031286239624, + "learning_rate": 1.052368625849507e-05, + "loss": 0.0203, + "step": 80020 + }, + { + "epoch": 15.53686714237173, + "grad_norm": 0.1292496621608734, + "learning_rate": 1.05135407958246e-05, + "loss": 0.0176, + "step": 80030 + }, + { + "epoch": 15.53880879568953, + "grad_norm": 0.17976364493370056, + "learning_rate": 1.0503399651333934e-05, + "loss": 0.0248, + "step": 80040 + }, + { + "epoch": 15.54075044900733, + "grad_norm": 0.16690556704998016, + "learning_rate": 1.04932628261321e-05, + "loss": 0.0193, + "step": 80050 + }, + { + "epoch": 15.54269210232513, + "grad_norm": 0.2546330392360687, + "learning_rate": 1.0483130321327644e-05, + "loss": 0.0228, + "step": 80060 + }, + { + "epoch": 15.54463375564293, + "grad_norm": 0.16494831442832947, + "learning_rate": 1.0473002138028654e-05, + "loss": 0.026, + "step": 80070 + }, + { + "epoch": 15.54657540896073, + "grad_norm": 0.1787133812904358, + "learning_rate": 1.0462878277342713e-05, + "loss": 0.0202, + "step": 80080 + }, + { + "epoch": 15.54851706227853, + "grad_norm": 0.18621303141117096, + "learning_rate": 1.0452758740376972e-05, + "loss": 0.0191, + "step": 80090 + }, + { + "epoch": 15.55045871559633, + "grad_norm": 0.10563928633928299, + "learning_rate": 1.0442643528238067e-05, + "loss": 0.0189, + "step": 80100 + }, + { + "epoch": 15.55240036891413, + "grad_norm": 0.16931450366973877, + "learning_rate": 1.0432532642032188e-05, + "loss": 0.0226, + "step": 80110 + }, + { + "epoch": 15.55434202223193, + "grad_norm": 0.1988125592470169, + "learning_rate": 1.042242608286505e-05, + "loss": 0.0248, + "step": 80120 + }, + { + "epoch": 15.55628367554973, + "grad_norm": 0.19464978575706482, + "learning_rate": 1.041232385184191e-05, + "loss": 0.0223, + "step": 80130 + }, + { + "epoch": 15.55822532886753, + "grad_norm": 0.22699487209320068, + "learning_rate": 1.0402225950067495e-05, + "loss": 0.0252, + "step": 80140 + }, + { + "epoch": 15.560166982185331, + "grad_norm": 0.18421146273612976, + "learning_rate": 1.0392132378646119e-05, + "loss": 0.0222, + "step": 80150 + }, + { + "epoch": 15.562108635503131, + "grad_norm": 0.22264480590820312, + "learning_rate": 1.0382043138681607e-05, + "loss": 0.0237, + "step": 80160 + }, + { + "epoch": 15.564050288820932, + "grad_norm": 0.2811559736728668, + "learning_rate": 1.0371958231277284e-05, + "loss": 0.0211, + "step": 80170 + }, + { + "epoch": 15.565991942138732, + "grad_norm": 0.1739753633737564, + "learning_rate": 1.0361877657536024e-05, + "loss": 0.0196, + "step": 80180 + }, + { + "epoch": 15.56793359545653, + "grad_norm": 0.078675277531147, + "learning_rate": 1.035180141856023e-05, + "loss": 0.0183, + "step": 80190 + }, + { + "epoch": 15.56987524877433, + "grad_norm": 0.1577284336090088, + "learning_rate": 1.0341729515451836e-05, + "loss": 0.0207, + "step": 80200 + }, + { + "epoch": 15.57181690209213, + "grad_norm": 0.14051620662212372, + "learning_rate": 1.0331661949312265e-05, + "loss": 0.0216, + "step": 80210 + }, + { + "epoch": 15.573758555409931, + "grad_norm": 0.08819518983364105, + "learning_rate": 1.0321598721242514e-05, + "loss": 0.019, + "step": 80220 + }, + { + "epoch": 15.575700208727731, + "grad_norm": 0.1555490344762802, + "learning_rate": 1.0311539832343064e-05, + "loss": 0.0197, + "step": 80230 + }, + { + "epoch": 15.577641862045532, + "grad_norm": 0.3124142289161682, + "learning_rate": 1.0301485283713942e-05, + "loss": 0.0233, + "step": 80240 + }, + { + "epoch": 15.579583515363332, + "grad_norm": 0.1252514272928238, + "learning_rate": 1.0291435076454703e-05, + "loss": 0.0232, + "step": 80250 + }, + { + "epoch": 15.581525168681132, + "grad_norm": 0.21215878427028656, + "learning_rate": 1.028138921166444e-05, + "loss": 0.0198, + "step": 80260 + }, + { + "epoch": 15.583466821998933, + "grad_norm": 0.14206436276435852, + "learning_rate": 1.0271347690441719e-05, + "loss": 0.023, + "step": 80270 + }, + { + "epoch": 15.585408475316733, + "grad_norm": 0.24189524352550507, + "learning_rate": 1.0261310513884698e-05, + "loss": 0.0188, + "step": 80280 + }, + { + "epoch": 15.587350128634533, + "grad_norm": 0.2014293670654297, + "learning_rate": 1.0251277683090998e-05, + "loss": 0.0217, + "step": 80290 + }, + { + "epoch": 15.589291781952333, + "grad_norm": 0.13361258804798126, + "learning_rate": 1.0241249199157826e-05, + "loss": 0.0206, + "step": 80300 + }, + { + "epoch": 15.591233435270132, + "grad_norm": 0.15391108393669128, + "learning_rate": 1.0231225063181837e-05, + "loss": 0.0188, + "step": 80310 + }, + { + "epoch": 15.593175088587932, + "grad_norm": 0.34131482243537903, + "learning_rate": 1.0221205276259316e-05, + "loss": 0.024, + "step": 80320 + }, + { + "epoch": 15.595116741905732, + "grad_norm": 0.15350696444511414, + "learning_rate": 1.0211189839485958e-05, + "loss": 0.028, + "step": 80330 + }, + { + "epoch": 15.597058395223533, + "grad_norm": 0.25452324748039246, + "learning_rate": 1.0201178753957075e-05, + "loss": 0.0237, + "step": 80340 + }, + { + "epoch": 15.599000048541333, + "grad_norm": 0.1184617131948471, + "learning_rate": 1.0191172020767432e-05, + "loss": 0.0182, + "step": 80350 + }, + { + "epoch": 15.600941701859133, + "grad_norm": 0.22460594773292542, + "learning_rate": 1.0181169641011362e-05, + "loss": 0.0198, + "step": 80360 + }, + { + "epoch": 15.602883355176933, + "grad_norm": 0.1829737424850464, + "learning_rate": 1.0171171615782721e-05, + "loss": 0.019, + "step": 80370 + }, + { + "epoch": 15.604825008494734, + "grad_norm": 0.13037057220935822, + "learning_rate": 1.0161177946174866e-05, + "loss": 0.0193, + "step": 80380 + }, + { + "epoch": 15.606766661812534, + "grad_norm": 0.20795783400535583, + "learning_rate": 1.0151188633280706e-05, + "loss": 0.0232, + "step": 80390 + }, + { + "epoch": 15.608708315130334, + "grad_norm": 0.0788927674293518, + "learning_rate": 1.0141203678192635e-05, + "loss": 0.0217, + "step": 80400 + }, + { + "epoch": 15.610649968448133, + "grad_norm": 0.12781007587909698, + "learning_rate": 1.0131223082002612e-05, + "loss": 0.0284, + "step": 80410 + }, + { + "epoch": 15.612591621765933, + "grad_norm": 0.12450864911079407, + "learning_rate": 1.012124684580208e-05, + "loss": 0.0161, + "step": 80420 + }, + { + "epoch": 15.614533275083733, + "grad_norm": 0.23816542327404022, + "learning_rate": 1.0111274970682039e-05, + "loss": 0.0268, + "step": 80430 + }, + { + "epoch": 15.616474928401534, + "grad_norm": 0.05958109349012375, + "learning_rate": 1.010130745773299e-05, + "loss": 0.0222, + "step": 80440 + }, + { + "epoch": 15.618416581719334, + "grad_norm": 0.15771397948265076, + "learning_rate": 1.0091344308044986e-05, + "loss": 0.0211, + "step": 80450 + }, + { + "epoch": 15.620358235037134, + "grad_norm": 0.23574717342853546, + "learning_rate": 1.0081385522707554e-05, + "loss": 0.0236, + "step": 80460 + }, + { + "epoch": 15.622299888354934, + "grad_norm": 0.15993806719779968, + "learning_rate": 1.0071431102809803e-05, + "loss": 0.0185, + "step": 80470 + }, + { + "epoch": 15.624241541672735, + "grad_norm": 0.18990051746368408, + "learning_rate": 1.0061481049440297e-05, + "loss": 0.0208, + "step": 80480 + }, + { + "epoch": 15.626183194990535, + "grad_norm": 0.13782978057861328, + "learning_rate": 1.0051535363687187e-05, + "loss": 0.0186, + "step": 80490 + }, + { + "epoch": 15.628124848308335, + "grad_norm": 0.07959933578968048, + "learning_rate": 1.0041594046638087e-05, + "loss": 0.0182, + "step": 80500 + }, + { + "epoch": 15.630066501626136, + "grad_norm": 0.15231257677078247, + "learning_rate": 1.0031657099380204e-05, + "loss": 0.0193, + "step": 80510 + }, + { + "epoch": 15.632008154943934, + "grad_norm": 0.22699569165706635, + "learning_rate": 1.0021724523000204e-05, + "loss": 0.023, + "step": 80520 + }, + { + "epoch": 15.633949808261734, + "grad_norm": 0.12041130661964417, + "learning_rate": 1.001179631858431e-05, + "loss": 0.0201, + "step": 80530 + }, + { + "epoch": 15.635891461579535, + "grad_norm": 0.1772070974111557, + "learning_rate": 1.0001872487218245e-05, + "loss": 0.0177, + "step": 80540 + }, + { + "epoch": 15.637833114897335, + "grad_norm": 0.16035759449005127, + "learning_rate": 9.991953029987272e-06, + "loss": 0.0243, + "step": 80550 + }, + { + "epoch": 15.639774768215135, + "grad_norm": 0.14414621889591217, + "learning_rate": 9.982037947976147e-06, + "loss": 0.0208, + "step": 80560 + }, + { + "epoch": 15.641716421532935, + "grad_norm": 0.198325514793396, + "learning_rate": 9.972127242269208e-06, + "loss": 0.0166, + "step": 80570 + }, + { + "epoch": 15.643658074850736, + "grad_norm": 0.1846969723701477, + "learning_rate": 9.962220913950238e-06, + "loss": 0.0229, + "step": 80580 + }, + { + "epoch": 15.645599728168536, + "grad_norm": 0.16912831366062164, + "learning_rate": 9.952318964102591e-06, + "loss": 0.0181, + "step": 80590 + }, + { + "epoch": 15.647541381486336, + "grad_norm": 0.23977424204349518, + "learning_rate": 9.942421393809148e-06, + "loss": 0.0267, + "step": 80600 + }, + { + "epoch": 15.649483034804136, + "grad_norm": 0.1852746605873108, + "learning_rate": 9.932528204152264e-06, + "loss": 0.0273, + "step": 80610 + }, + { + "epoch": 15.651424688121935, + "grad_norm": 0.13047900795936584, + "learning_rate": 9.922639396213856e-06, + "loss": 0.0208, + "step": 80620 + }, + { + "epoch": 15.653366341439735, + "grad_norm": 0.13068516552448273, + "learning_rate": 9.912754971075344e-06, + "loss": 0.0242, + "step": 80630 + }, + { + "epoch": 15.655307994757536, + "grad_norm": 0.21818409860134125, + "learning_rate": 9.902874929817696e-06, + "loss": 0.0236, + "step": 80640 + }, + { + "epoch": 15.657249648075336, + "grad_norm": 0.060085728764534, + "learning_rate": 9.892999273521342e-06, + "loss": 0.021, + "step": 80650 + }, + { + "epoch": 15.659191301393136, + "grad_norm": 0.1372332125902176, + "learning_rate": 9.883128003266302e-06, + "loss": 0.0273, + "step": 80660 + }, + { + "epoch": 15.661132954710936, + "grad_norm": 0.12304391711950302, + "learning_rate": 9.87326112013206e-06, + "loss": 0.0227, + "step": 80670 + }, + { + "epoch": 15.663074608028737, + "grad_norm": 0.1934748739004135, + "learning_rate": 9.863398625197662e-06, + "loss": 0.0274, + "step": 80680 + }, + { + "epoch": 15.665016261346537, + "grad_norm": 0.21374334394931793, + "learning_rate": 9.853540519541615e-06, + "loss": 0.0277, + "step": 80690 + }, + { + "epoch": 15.666957914664337, + "grad_norm": 0.0811886116862297, + "learning_rate": 9.843686804242053e-06, + "loss": 0.0195, + "step": 80700 + }, + { + "epoch": 15.668899567982137, + "grad_norm": 0.1157001256942749, + "learning_rate": 9.833837480376506e-06, + "loss": 0.0257, + "step": 80710 + }, + { + "epoch": 15.670841221299938, + "grad_norm": 0.1877552568912506, + "learning_rate": 9.82399254902212e-06, + "loss": 0.0217, + "step": 80720 + }, + { + "epoch": 15.672782874617736, + "grad_norm": 0.29636427760124207, + "learning_rate": 9.814152011255494e-06, + "loss": 0.023, + "step": 80730 + }, + { + "epoch": 15.674724527935536, + "grad_norm": 0.29033133387565613, + "learning_rate": 9.804315868152797e-06, + "loss": 0.0208, + "step": 80740 + }, + { + "epoch": 15.676666181253337, + "grad_norm": 0.27257850766181946, + "learning_rate": 9.79448412078966e-06, + "loss": 0.0265, + "step": 80750 + }, + { + "epoch": 15.678607834571137, + "grad_norm": 0.13891002535820007, + "learning_rate": 9.784656770241312e-06, + "loss": 0.0188, + "step": 80760 + }, + { + "epoch": 15.680549487888937, + "grad_norm": 0.11768442392349243, + "learning_rate": 9.77483381758243e-06, + "loss": 0.0207, + "step": 80770 + }, + { + "epoch": 15.682491141206738, + "grad_norm": 0.12070979923009872, + "learning_rate": 9.76501526388725e-06, + "loss": 0.0205, + "step": 80780 + }, + { + "epoch": 15.684432794524538, + "grad_norm": 0.18805228173732758, + "learning_rate": 9.7552011102295e-06, + "loss": 0.0214, + "step": 80790 + }, + { + "epoch": 15.686374447842338, + "grad_norm": 0.16119831800460815, + "learning_rate": 9.745391357682449e-06, + "loss": 0.0215, + "step": 80800 + }, + { + "epoch": 15.688316101160138, + "grad_norm": 0.10648077726364136, + "learning_rate": 9.735586007318875e-06, + "loss": 0.0226, + "step": 80810 + }, + { + "epoch": 15.690257754477939, + "grad_norm": 0.18825416266918182, + "learning_rate": 9.725785060211096e-06, + "loss": 0.0242, + "step": 80820 + }, + { + "epoch": 15.692199407795737, + "grad_norm": 0.16692744195461273, + "learning_rate": 9.715988517430896e-06, + "loss": 0.0255, + "step": 80830 + }, + { + "epoch": 15.694141061113537, + "grad_norm": 0.16864198446273804, + "learning_rate": 9.706196380049625e-06, + "loss": 0.018, + "step": 80840 + }, + { + "epoch": 15.696082714431338, + "grad_norm": 0.2824241816997528, + "learning_rate": 9.69640864913815e-06, + "loss": 0.0314, + "step": 80850 + }, + { + "epoch": 15.698024367749138, + "grad_norm": 0.09728111326694489, + "learning_rate": 9.686625325766818e-06, + "loss": 0.0223, + "step": 80860 + }, + { + "epoch": 15.699966021066938, + "grad_norm": 0.16792158782482147, + "learning_rate": 9.67684641100554e-06, + "loss": 0.0314, + "step": 80870 + }, + { + "epoch": 15.701907674384739, + "grad_norm": 0.1390063613653183, + "learning_rate": 9.667071905923692e-06, + "loss": 0.02, + "step": 80880 + }, + { + "epoch": 15.703849327702539, + "grad_norm": 0.19918760657310486, + "learning_rate": 9.657301811590242e-06, + "loss": 0.0276, + "step": 80890 + }, + { + "epoch": 15.705790981020339, + "grad_norm": 0.22633877396583557, + "learning_rate": 9.647536129073603e-06, + "loss": 0.0244, + "step": 80900 + }, + { + "epoch": 15.70773263433814, + "grad_norm": 0.26459479331970215, + "learning_rate": 9.637774859441751e-06, + "loss": 0.0271, + "step": 80910 + }, + { + "epoch": 15.70967428765594, + "grad_norm": 0.06291802227497101, + "learning_rate": 9.628018003762152e-06, + "loss": 0.0241, + "step": 80920 + }, + { + "epoch": 15.71161594097374, + "grad_norm": 0.1856285035610199, + "learning_rate": 9.618265563101813e-06, + "loss": 0.0255, + "step": 80930 + }, + { + "epoch": 15.713557594291538, + "grad_norm": 0.16417627036571503, + "learning_rate": 9.608517538527211e-06, + "loss": 0.0245, + "step": 80940 + }, + { + "epoch": 15.715499247609339, + "grad_norm": 0.14585186541080475, + "learning_rate": 9.598773931104431e-06, + "loss": 0.0179, + "step": 80950 + }, + { + "epoch": 15.717440900927139, + "grad_norm": 0.11678126454353333, + "learning_rate": 9.589034741898983e-06, + "loss": 0.0223, + "step": 80960 + }, + { + "epoch": 15.71938255424494, + "grad_norm": 0.108711838722229, + "learning_rate": 9.579299971975947e-06, + "loss": 0.0261, + "step": 80970 + }, + { + "epoch": 15.72132420756274, + "grad_norm": 0.19420933723449707, + "learning_rate": 9.569569622399882e-06, + "loss": 0.0246, + "step": 80980 + }, + { + "epoch": 15.72326586088054, + "grad_norm": 0.17335985600948334, + "learning_rate": 9.559843694234916e-06, + "loss": 0.0193, + "step": 80990 + }, + { + "epoch": 15.72520751419834, + "grad_norm": 0.14536751806735992, + "learning_rate": 9.550122188544625e-06, + "loss": 0.0205, + "step": 81000 + }, + { + "epoch": 15.72714916751614, + "grad_norm": 0.2064066231250763, + "learning_rate": 9.540405106392152e-06, + "loss": 0.0205, + "step": 81010 + }, + { + "epoch": 15.72909082083394, + "grad_norm": 0.17032694816589355, + "learning_rate": 9.53069244884015e-06, + "loss": 0.0226, + "step": 81020 + }, + { + "epoch": 15.73103247415174, + "grad_norm": 0.15879715979099274, + "learning_rate": 9.52098421695079e-06, + "loss": 0.029, + "step": 81030 + }, + { + "epoch": 15.73297412746954, + "grad_norm": 0.17257696390151978, + "learning_rate": 9.511280411785728e-06, + "loss": 0.022, + "step": 81040 + }, + { + "epoch": 15.73491578078734, + "grad_norm": 0.13080616295337677, + "learning_rate": 9.50158103440616e-06, + "loss": 0.0232, + "step": 81050 + }, + { + "epoch": 15.73685743410514, + "grad_norm": 0.17017312347888947, + "learning_rate": 9.491886085872815e-06, + "loss": 0.0233, + "step": 81060 + }, + { + "epoch": 15.73879908742294, + "grad_norm": 0.1555790901184082, + "learning_rate": 9.482195567245895e-06, + "loss": 0.0207, + "step": 81070 + }, + { + "epoch": 15.74074074074074, + "grad_norm": 0.19869397580623627, + "learning_rate": 9.472509479585151e-06, + "loss": 0.0242, + "step": 81080 + }, + { + "epoch": 15.74268239405854, + "grad_norm": 0.195403054356575, + "learning_rate": 9.462827823949833e-06, + "loss": 0.0226, + "step": 81090 + }, + { + "epoch": 15.744624047376341, + "grad_norm": 0.08230079710483551, + "learning_rate": 9.453150601398735e-06, + "loss": 0.0204, + "step": 81100 + }, + { + "epoch": 15.746565700694141, + "grad_norm": 0.14190825819969177, + "learning_rate": 9.44347781299011e-06, + "loss": 0.0214, + "step": 81110 + }, + { + "epoch": 15.748507354011942, + "grad_norm": 0.09462994337081909, + "learning_rate": 9.43380945978179e-06, + "loss": 0.0197, + "step": 81120 + }, + { + "epoch": 15.750449007329742, + "grad_norm": 0.1533605009317398, + "learning_rate": 9.424145542831047e-06, + "loss": 0.0213, + "step": 81130 + }, + { + "epoch": 15.752390660647542, + "grad_norm": 0.20770171284675598, + "learning_rate": 9.414486063194778e-06, + "loss": 0.0217, + "step": 81140 + }, + { + "epoch": 15.754332313965342, + "grad_norm": 0.1833520382642746, + "learning_rate": 9.404831021929273e-06, + "loss": 0.0196, + "step": 81150 + }, + { + "epoch": 15.756273967283141, + "grad_norm": 0.23390871286392212, + "learning_rate": 9.395180420090427e-06, + "loss": 0.0256, + "step": 81160 + }, + { + "epoch": 15.758215620600941, + "grad_norm": 0.1622077375650406, + "learning_rate": 9.385534258733591e-06, + "loss": 0.0251, + "step": 81170 + }, + { + "epoch": 15.760157273918741, + "grad_norm": 0.21857570111751556, + "learning_rate": 9.375892538913677e-06, + "loss": 0.0243, + "step": 81180 + }, + { + "epoch": 15.762098927236542, + "grad_norm": 0.23531576991081238, + "learning_rate": 9.36625526168507e-06, + "loss": 0.0199, + "step": 81190 + }, + { + "epoch": 15.764040580554342, + "grad_norm": 0.22035691142082214, + "learning_rate": 9.356622428101686e-06, + "loss": 0.0287, + "step": 81200 + }, + { + "epoch": 15.765982233872142, + "grad_norm": 0.15611529350280762, + "learning_rate": 9.346994039216972e-06, + "loss": 0.0192, + "step": 81210 + }, + { + "epoch": 15.767923887189943, + "grad_norm": 0.13128222525119781, + "learning_rate": 9.337370096083886e-06, + "loss": 0.0148, + "step": 81220 + }, + { + "epoch": 15.769865540507743, + "grad_norm": 0.09671221673488617, + "learning_rate": 9.327750599754854e-06, + "loss": 0.0184, + "step": 81230 + }, + { + "epoch": 15.771807193825543, + "grad_norm": 0.20216478407382965, + "learning_rate": 9.318135551281875e-06, + "loss": 0.0255, + "step": 81240 + }, + { + "epoch": 15.773748847143343, + "grad_norm": 0.23095670342445374, + "learning_rate": 9.308524951716424e-06, + "loss": 0.0225, + "step": 81250 + }, + { + "epoch": 15.775690500461142, + "grad_norm": 0.17709574103355408, + "learning_rate": 9.298918802109497e-06, + "loss": 0.0224, + "step": 81260 + }, + { + "epoch": 15.777632153778942, + "grad_norm": 0.11536017060279846, + "learning_rate": 9.289317103511625e-06, + "loss": 0.0206, + "step": 81270 + }, + { + "epoch": 15.779573807096742, + "grad_norm": 0.18869121372699738, + "learning_rate": 9.279719856972835e-06, + "loss": 0.019, + "step": 81280 + }, + { + "epoch": 15.781515460414543, + "grad_norm": 0.13944987952709198, + "learning_rate": 9.270127063542645e-06, + "loss": 0.0216, + "step": 81290 + }, + { + "epoch": 15.783457113732343, + "grad_norm": 0.15615470707416534, + "learning_rate": 9.260538724270124e-06, + "loss": 0.0223, + "step": 81300 + }, + { + "epoch": 15.785398767050143, + "grad_norm": 0.13671334087848663, + "learning_rate": 9.250954840203852e-06, + "loss": 0.026, + "step": 81310 + }, + { + "epoch": 15.787340420367943, + "grad_norm": 0.3211415410041809, + "learning_rate": 9.241375412391878e-06, + "loss": 0.0207, + "step": 81320 + }, + { + "epoch": 15.789282073685744, + "grad_norm": 0.26527905464172363, + "learning_rate": 9.23180044188181e-06, + "loss": 0.0244, + "step": 81330 + }, + { + "epoch": 15.791223727003544, + "grad_norm": 0.3402283489704132, + "learning_rate": 9.222229929720755e-06, + "loss": 0.0216, + "step": 81340 + }, + { + "epoch": 15.793165380321344, + "grad_norm": 0.23919577896595, + "learning_rate": 9.212663876955336e-06, + "loss": 0.0234, + "step": 81350 + }, + { + "epoch": 15.795107033639145, + "grad_norm": 0.1609787791967392, + "learning_rate": 9.20310228463166e-06, + "loss": 0.0203, + "step": 81360 + }, + { + "epoch": 15.797048686956943, + "grad_norm": 0.1435048133134842, + "learning_rate": 9.19354515379539e-06, + "loss": 0.0199, + "step": 81370 + }, + { + "epoch": 15.798990340274743, + "grad_norm": 0.17631670832633972, + "learning_rate": 9.183992485491666e-06, + "loss": 0.0209, + "step": 81380 + }, + { + "epoch": 15.800931993592544, + "grad_norm": 0.17891371250152588, + "learning_rate": 9.174444280765154e-06, + "loss": 0.0176, + "step": 81390 + }, + { + "epoch": 15.802873646910344, + "grad_norm": 0.14433525502681732, + "learning_rate": 9.164900540660032e-06, + "loss": 0.0247, + "step": 81400 + }, + { + "epoch": 15.804815300228144, + "grad_norm": 0.11159155517816544, + "learning_rate": 9.155361266220008e-06, + "loss": 0.0176, + "step": 81410 + }, + { + "epoch": 15.806756953545944, + "grad_norm": 0.1392231583595276, + "learning_rate": 9.145826458488254e-06, + "loss": 0.0229, + "step": 81420 + }, + { + "epoch": 15.808698606863745, + "grad_norm": 0.22030232846736908, + "learning_rate": 9.136296118507509e-06, + "loss": 0.0211, + "step": 81430 + }, + { + "epoch": 15.810640260181545, + "grad_norm": 0.2926202714443207, + "learning_rate": 9.126770247319966e-06, + "loss": 0.0299, + "step": 81440 + }, + { + "epoch": 15.812581913499345, + "grad_norm": 0.17668212950229645, + "learning_rate": 9.11724884596738e-06, + "loss": 0.0238, + "step": 81450 + }, + { + "epoch": 15.814523566817146, + "grad_norm": 0.15644004940986633, + "learning_rate": 9.107731915490996e-06, + "loss": 0.0214, + "step": 81460 + }, + { + "epoch": 15.816465220134944, + "grad_norm": 0.11603613197803497, + "learning_rate": 9.098219456931579e-06, + "loss": 0.0167, + "step": 81470 + }, + { + "epoch": 15.818406873452744, + "grad_norm": 0.18646948039531708, + "learning_rate": 9.088711471329376e-06, + "loss": 0.0261, + "step": 81480 + }, + { + "epoch": 15.820348526770545, + "grad_norm": 0.3040325343608856, + "learning_rate": 9.079207959724184e-06, + "loss": 0.0241, + "step": 81490 + }, + { + "epoch": 15.822290180088345, + "grad_norm": 0.3031758964061737, + "learning_rate": 9.069708923155279e-06, + "loss": 0.0292, + "step": 81500 + }, + { + "epoch": 15.824231833406145, + "grad_norm": 0.23486320674419403, + "learning_rate": 9.060214362661468e-06, + "loss": 0.0244, + "step": 81510 + }, + { + "epoch": 15.826173486723945, + "grad_norm": 0.24042651057243347, + "learning_rate": 9.050724279281064e-06, + "loss": 0.025, + "step": 81520 + }, + { + "epoch": 15.828115140041746, + "grad_norm": 0.14744769036769867, + "learning_rate": 9.041238674051894e-06, + "loss": 0.0258, + "step": 81530 + }, + { + "epoch": 15.830056793359546, + "grad_norm": 0.13963468372821808, + "learning_rate": 9.031757548011277e-06, + "loss": 0.0213, + "step": 81540 + }, + { + "epoch": 15.831998446677346, + "grad_norm": 0.1266099363565445, + "learning_rate": 9.022280902196051e-06, + "loss": 0.0167, + "step": 81550 + }, + { + "epoch": 15.833940099995147, + "grad_norm": 0.13131672143936157, + "learning_rate": 9.012808737642592e-06, + "loss": 0.0176, + "step": 81560 + }, + { + "epoch": 15.835881753312947, + "grad_norm": 0.21413558721542358, + "learning_rate": 9.003341055386739e-06, + "loss": 0.024, + "step": 81570 + }, + { + "epoch": 15.837823406630745, + "grad_norm": 0.13605307042598724, + "learning_rate": 8.993877856463862e-06, + "loss": 0.0325, + "step": 81580 + }, + { + "epoch": 15.839765059948546, + "grad_norm": 0.15838737785816193, + "learning_rate": 8.984419141908856e-06, + "loss": 0.0241, + "step": 81590 + }, + { + "epoch": 15.841706713266346, + "grad_norm": 0.1618547886610031, + "learning_rate": 8.974964912756117e-06, + "loss": 0.0208, + "step": 81600 + }, + { + "epoch": 15.843648366584146, + "grad_norm": 0.18116459250450134, + "learning_rate": 8.965515170039523e-06, + "loss": 0.0148, + "step": 81610 + }, + { + "epoch": 15.845590019901946, + "grad_norm": 0.13189946115016937, + "learning_rate": 8.956069914792514e-06, + "loss": 0.0259, + "step": 81620 + }, + { + "epoch": 15.847531673219747, + "grad_norm": 0.20284615457057953, + "learning_rate": 8.946629148047969e-06, + "loss": 0.0236, + "step": 81630 + }, + { + "epoch": 15.849473326537547, + "grad_norm": 0.17010807991027832, + "learning_rate": 8.93719287083834e-06, + "loss": 0.0221, + "step": 81640 + }, + { + "epoch": 15.851414979855347, + "grad_norm": 0.15520039200782776, + "learning_rate": 8.927761084195569e-06, + "loss": 0.0161, + "step": 81650 + }, + { + "epoch": 15.853356633173147, + "grad_norm": 0.202140212059021, + "learning_rate": 8.918333789151096e-06, + "loss": 0.0218, + "step": 81660 + }, + { + "epoch": 15.855298286490948, + "grad_norm": 0.13372886180877686, + "learning_rate": 8.90891098673587e-06, + "loss": 0.0216, + "step": 81670 + }, + { + "epoch": 15.857239939808746, + "grad_norm": 0.1981658786535263, + "learning_rate": 8.89949267798037e-06, + "loss": 0.0242, + "step": 81680 + }, + { + "epoch": 15.859181593126547, + "grad_norm": 0.11549074202775955, + "learning_rate": 8.890078863914542e-06, + "loss": 0.0182, + "step": 81690 + }, + { + "epoch": 15.861123246444347, + "grad_norm": 0.1852417141199112, + "learning_rate": 8.880669545567894e-06, + "loss": 0.0248, + "step": 81700 + }, + { + "epoch": 15.863064899762147, + "grad_norm": 0.22484181821346283, + "learning_rate": 8.871264723969381e-06, + "loss": 0.0233, + "step": 81710 + }, + { + "epoch": 15.865006553079947, + "grad_norm": 0.10548753291368484, + "learning_rate": 8.86186440014754e-06, + "loss": 0.0201, + "step": 81720 + }, + { + "epoch": 15.866948206397748, + "grad_norm": 0.161876380443573, + "learning_rate": 8.852468575130351e-06, + "loss": 0.0224, + "step": 81730 + }, + { + "epoch": 15.868889859715548, + "grad_norm": 0.24597282707691193, + "learning_rate": 8.843077249945347e-06, + "loss": 0.0245, + "step": 81740 + }, + { + "epoch": 15.870831513033348, + "grad_norm": 0.08605965226888657, + "learning_rate": 8.833690425619523e-06, + "loss": 0.0237, + "step": 81750 + }, + { + "epoch": 15.872773166351148, + "grad_norm": 0.14818233251571655, + "learning_rate": 8.82430810317942e-06, + "loss": 0.0239, + "step": 81760 + }, + { + "epoch": 15.874714819668949, + "grad_norm": 0.2730647921562195, + "learning_rate": 8.814930283651074e-06, + "loss": 0.0218, + "step": 81770 + }, + { + "epoch": 15.876656472986749, + "grad_norm": 0.18664176762104034, + "learning_rate": 8.805556968060048e-06, + "loss": 0.0224, + "step": 81780 + }, + { + "epoch": 15.87859812630455, + "grad_norm": 0.16076548397541046, + "learning_rate": 8.796188157431362e-06, + "loss": 0.0168, + "step": 81790 + }, + { + "epoch": 15.880539779622348, + "grad_norm": 0.13409313559532166, + "learning_rate": 8.786823852789595e-06, + "loss": 0.026, + "step": 81800 + }, + { + "epoch": 15.882481432940148, + "grad_norm": 0.15870997309684753, + "learning_rate": 8.777464055158813e-06, + "loss": 0.0266, + "step": 81810 + }, + { + "epoch": 15.884423086257948, + "grad_norm": 0.12057510018348694, + "learning_rate": 8.768108765562578e-06, + "loss": 0.0198, + "step": 81820 + }, + { + "epoch": 15.886364739575749, + "grad_norm": 0.14155937731266022, + "learning_rate": 8.758757985023975e-06, + "loss": 0.0231, + "step": 81830 + }, + { + "epoch": 15.888306392893549, + "grad_norm": 0.1679464727640152, + "learning_rate": 8.749411714565598e-06, + "loss": 0.021, + "step": 81840 + }, + { + "epoch": 15.89024804621135, + "grad_norm": 0.2163173407316208, + "learning_rate": 8.74006995520954e-06, + "loss": 0.0211, + "step": 81850 + }, + { + "epoch": 15.89218969952915, + "grad_norm": 0.2192554473876953, + "learning_rate": 8.730732707977397e-06, + "loss": 0.02, + "step": 81860 + }, + { + "epoch": 15.89413135284695, + "grad_norm": 0.22812385857105255, + "learning_rate": 8.721399973890281e-06, + "loss": 0.0214, + "step": 81870 + }, + { + "epoch": 15.89607300616475, + "grad_norm": 0.18997976183891296, + "learning_rate": 8.712071753968792e-06, + "loss": 0.0229, + "step": 81880 + }, + { + "epoch": 15.89801465948255, + "grad_norm": 0.17082719504833221, + "learning_rate": 8.702748049233072e-06, + "loss": 0.0203, + "step": 81890 + }, + { + "epoch": 15.899956312800349, + "grad_norm": 0.19963543117046356, + "learning_rate": 8.693428860702718e-06, + "loss": 0.023, + "step": 81900 + }, + { + "epoch": 15.901897966118149, + "grad_norm": 0.17702610790729523, + "learning_rate": 8.684114189396898e-06, + "loss": 0.0269, + "step": 81910 + }, + { + "epoch": 15.90383961943595, + "grad_norm": 0.09577999264001846, + "learning_rate": 8.674804036334222e-06, + "loss": 0.0203, + "step": 81920 + }, + { + "epoch": 15.90578127275375, + "grad_norm": 0.24966776371002197, + "learning_rate": 8.665498402532857e-06, + "loss": 0.0307, + "step": 81930 + }, + { + "epoch": 15.90772292607155, + "grad_norm": 0.23579014837741852, + "learning_rate": 8.656197289010432e-06, + "loss": 0.0244, + "step": 81940 + }, + { + "epoch": 15.90966457938935, + "grad_norm": 0.127461239695549, + "learning_rate": 8.646900696784122e-06, + "loss": 0.0218, + "step": 81950 + }, + { + "epoch": 15.91160623270715, + "grad_norm": 0.1644071787595749, + "learning_rate": 8.637608626870553e-06, + "loss": 0.0189, + "step": 81960 + }, + { + "epoch": 15.91354788602495, + "grad_norm": 0.22245396673679352, + "learning_rate": 8.628321080285944e-06, + "loss": 0.0236, + "step": 81970 + }, + { + "epoch": 15.915489539342751, + "grad_norm": 0.2532748878002167, + "learning_rate": 8.619038058045925e-06, + "loss": 0.0256, + "step": 81980 + }, + { + "epoch": 15.917431192660551, + "grad_norm": 0.17410404980182648, + "learning_rate": 8.609759561165704e-06, + "loss": 0.017, + "step": 81990 + }, + { + "epoch": 15.919372845978351, + "grad_norm": 0.13033483922481537, + "learning_rate": 8.600485590659934e-06, + "loss": 0.0227, + "step": 82000 + }, + { + "epoch": 15.92131449929615, + "grad_norm": 0.15385161340236664, + "learning_rate": 8.591216147542813e-06, + "loss": 0.0181, + "step": 82010 + }, + { + "epoch": 15.92325615261395, + "grad_norm": 0.09689129889011383, + "learning_rate": 8.581951232828045e-06, + "loss": 0.0189, + "step": 82020 + }, + { + "epoch": 15.92519780593175, + "grad_norm": 0.13956938683986664, + "learning_rate": 8.572690847528825e-06, + "loss": 0.0236, + "step": 82030 + }, + { + "epoch": 15.92713945924955, + "grad_norm": 0.17925983667373657, + "learning_rate": 8.563434992657838e-06, + "loss": 0.0221, + "step": 82040 + }, + { + "epoch": 15.929081112567351, + "grad_norm": 0.10955511033535004, + "learning_rate": 8.554183669227307e-06, + "loss": 0.0188, + "step": 82050 + }, + { + "epoch": 15.931022765885151, + "grad_norm": 0.2112000584602356, + "learning_rate": 8.544936878248937e-06, + "loss": 0.0219, + "step": 82060 + }, + { + "epoch": 15.932964419202952, + "grad_norm": 0.1840696632862091, + "learning_rate": 8.535694620733942e-06, + "loss": 0.0225, + "step": 82070 + }, + { + "epoch": 15.934906072520752, + "grad_norm": 0.15014509856700897, + "learning_rate": 8.526456897693053e-06, + "loss": 0.0232, + "step": 82080 + }, + { + "epoch": 15.936847725838552, + "grad_norm": 0.10237637907266617, + "learning_rate": 8.517223710136456e-06, + "loss": 0.0207, + "step": 82090 + }, + { + "epoch": 15.938789379156352, + "grad_norm": 0.2948358654975891, + "learning_rate": 8.507995059073931e-06, + "loss": 0.0247, + "step": 82100 + }, + { + "epoch": 15.940731032474151, + "grad_norm": 0.24229177832603455, + "learning_rate": 8.498770945514673e-06, + "loss": 0.0216, + "step": 82110 + }, + { + "epoch": 15.942672685791951, + "grad_norm": 0.29216042160987854, + "learning_rate": 8.489551370467436e-06, + "loss": 0.0287, + "step": 82120 + }, + { + "epoch": 15.944614339109751, + "grad_norm": 0.16977819800376892, + "learning_rate": 8.480336334940442e-06, + "loss": 0.0198, + "step": 82130 + }, + { + "epoch": 15.946555992427552, + "grad_norm": 0.1834949105978012, + "learning_rate": 8.471125839941458e-06, + "loss": 0.0268, + "step": 82140 + }, + { + "epoch": 15.948497645745352, + "grad_norm": 0.16385163366794586, + "learning_rate": 8.461919886477682e-06, + "loss": 0.0222, + "step": 82150 + }, + { + "epoch": 15.950439299063152, + "grad_norm": 0.19129827618598938, + "learning_rate": 8.452718475555927e-06, + "loss": 0.0194, + "step": 82160 + }, + { + "epoch": 15.952380952380953, + "grad_norm": 0.17357909679412842, + "learning_rate": 8.443521608182398e-06, + "loss": 0.0227, + "step": 82170 + }, + { + "epoch": 15.954322605698753, + "grad_norm": 0.18689264357089996, + "learning_rate": 8.43432928536288e-06, + "loss": 0.0251, + "step": 82180 + }, + { + "epoch": 15.956264259016553, + "grad_norm": 0.27432769536972046, + "learning_rate": 8.425141508102607e-06, + "loss": 0.0199, + "step": 82190 + }, + { + "epoch": 15.958205912334353, + "grad_norm": 0.1158883273601532, + "learning_rate": 8.415958277406365e-06, + "loss": 0.0206, + "step": 82200 + }, + { + "epoch": 15.960147565652154, + "grad_norm": 0.1896868795156479, + "learning_rate": 8.406779594278392e-06, + "loss": 0.0255, + "step": 82210 + }, + { + "epoch": 15.962089218969952, + "grad_norm": 0.16565845906734467, + "learning_rate": 8.397605459722468e-06, + "loss": 0.0247, + "step": 82220 + }, + { + "epoch": 15.964030872287752, + "grad_norm": 0.11358430981636047, + "learning_rate": 8.388435874741863e-06, + "loss": 0.0171, + "step": 82230 + }, + { + "epoch": 15.965972525605553, + "grad_norm": 0.2805183529853821, + "learning_rate": 8.37927084033936e-06, + "loss": 0.0243, + "step": 82240 + }, + { + "epoch": 15.967914178923353, + "grad_norm": 0.18304891884326935, + "learning_rate": 8.370110357517208e-06, + "loss": 0.0247, + "step": 82250 + }, + { + "epoch": 15.969855832241153, + "grad_norm": 0.20418818295001984, + "learning_rate": 8.360954427277201e-06, + "loss": 0.0206, + "step": 82260 + }, + { + "epoch": 15.971797485558954, + "grad_norm": 0.2638707756996155, + "learning_rate": 8.351803050620627e-06, + "loss": 0.0241, + "step": 82270 + }, + { + "epoch": 15.973739138876754, + "grad_norm": 0.12614226341247559, + "learning_rate": 8.34265622854824e-06, + "loss": 0.0191, + "step": 82280 + }, + { + "epoch": 15.975680792194554, + "grad_norm": 0.15293635427951813, + "learning_rate": 8.333513962060335e-06, + "loss": 0.0172, + "step": 82290 + }, + { + "epoch": 15.977622445512354, + "grad_norm": 0.27111876010894775, + "learning_rate": 8.324376252156701e-06, + "loss": 0.0315, + "step": 82300 + }, + { + "epoch": 15.979564098830155, + "grad_norm": 0.13770148158073425, + "learning_rate": 8.315243099836633e-06, + "loss": 0.0231, + "step": 82310 + }, + { + "epoch": 15.981505752147953, + "grad_norm": 0.19458362460136414, + "learning_rate": 8.3061145060989e-06, + "loss": 0.021, + "step": 82320 + }, + { + "epoch": 15.983447405465753, + "grad_norm": 0.14244906604290009, + "learning_rate": 8.296990471941806e-06, + "loss": 0.0243, + "step": 82330 + }, + { + "epoch": 15.985389058783554, + "grad_norm": 0.22996066510677338, + "learning_rate": 8.287870998363112e-06, + "loss": 0.0238, + "step": 82340 + }, + { + "epoch": 15.987330712101354, + "grad_norm": 0.15604005753993988, + "learning_rate": 8.278756086360156e-06, + "loss": 0.0211, + "step": 82350 + }, + { + "epoch": 15.989272365419154, + "grad_norm": 0.15709319710731506, + "learning_rate": 8.269645736929693e-06, + "loss": 0.0207, + "step": 82360 + }, + { + "epoch": 15.991214018736954, + "grad_norm": 0.09989932924509048, + "learning_rate": 8.260539951068048e-06, + "loss": 0.0272, + "step": 82370 + }, + { + "epoch": 15.993155672054755, + "grad_norm": 0.14779362082481384, + "learning_rate": 8.251438729770983e-06, + "loss": 0.02, + "step": 82380 + }, + { + "epoch": 15.995097325372555, + "grad_norm": 0.18020911514759064, + "learning_rate": 8.242342074033826e-06, + "loss": 0.0211, + "step": 82390 + }, + { + "epoch": 15.997038978690355, + "grad_norm": 0.2562488317489624, + "learning_rate": 8.233249984851344e-06, + "loss": 0.0266, + "step": 82400 + }, + { + "epoch": 15.998980632008156, + "grad_norm": 0.17446032166481018, + "learning_rate": 8.224162463217849e-06, + "loss": 0.0207, + "step": 82410 + }, + { + "epoch": 16.00077666132712, + "grad_norm": 0.17149515450000763, + "learning_rate": 8.215079510127139e-06, + "loss": 0.0218, + "step": 82420 + }, + { + "epoch": 16.00271831464492, + "grad_norm": 0.1726093292236328, + "learning_rate": 8.206001126572521e-06, + "loss": 0.0238, + "step": 82430 + }, + { + "epoch": 16.00465996796272, + "grad_norm": 0.14945049583911896, + "learning_rate": 8.196927313546776e-06, + "loss": 0.0223, + "step": 82440 + }, + { + "epoch": 16.00660162128052, + "grad_norm": 0.22552011907100677, + "learning_rate": 8.18785807204222e-06, + "loss": 0.0223, + "step": 82450 + }, + { + "epoch": 16.00854327459832, + "grad_norm": 0.15474526584148407, + "learning_rate": 8.178793403050627e-06, + "loss": 0.0166, + "step": 82460 + }, + { + "epoch": 16.01048492791612, + "grad_norm": 0.227552592754364, + "learning_rate": 8.16973330756331e-06, + "loss": 0.0207, + "step": 82470 + }, + { + "epoch": 16.01242658123392, + "grad_norm": 0.2229868322610855, + "learning_rate": 8.16067778657107e-06, + "loss": 0.0263, + "step": 82480 + }, + { + "epoch": 16.01436823455172, + "grad_norm": 0.19302469491958618, + "learning_rate": 8.151626841064214e-06, + "loss": 0.0185, + "step": 82490 + }, + { + "epoch": 16.01630988786952, + "grad_norm": 0.2685433328151703, + "learning_rate": 8.142580472032518e-06, + "loss": 0.0286, + "step": 82500 + }, + { + "epoch": 16.018251541187322, + "grad_norm": 0.1706138700246811, + "learning_rate": 8.133538680465285e-06, + "loss": 0.0188, + "step": 82510 + }, + { + "epoch": 16.020193194505122, + "grad_norm": 0.17008371651172638, + "learning_rate": 8.124501467351336e-06, + "loss": 0.0188, + "step": 82520 + }, + { + "epoch": 16.022134847822922, + "grad_norm": 0.2669728696346283, + "learning_rate": 8.11546883367893e-06, + "loss": 0.0248, + "step": 82530 + }, + { + "epoch": 16.024076501140723, + "grad_norm": 0.565573513507843, + "learning_rate": 8.106440780435882e-06, + "loss": 0.0244, + "step": 82540 + }, + { + "epoch": 16.026018154458523, + "grad_norm": 0.08596926927566528, + "learning_rate": 8.097417308609485e-06, + "loss": 0.0185, + "step": 82550 + }, + { + "epoch": 16.027959807776323, + "grad_norm": 0.14626139402389526, + "learning_rate": 8.088398419186544e-06, + "loss": 0.0249, + "step": 82560 + }, + { + "epoch": 16.02990146109412, + "grad_norm": 0.1917155683040619, + "learning_rate": 8.079384113153321e-06, + "loss": 0.0226, + "step": 82570 + }, + { + "epoch": 16.03184311441192, + "grad_norm": 0.2006208598613739, + "learning_rate": 8.070374391495644e-06, + "loss": 0.0183, + "step": 82580 + }, + { + "epoch": 16.03378476772972, + "grad_norm": 0.11382768303155899, + "learning_rate": 8.061369255198764e-06, + "loss": 0.0198, + "step": 82590 + }, + { + "epoch": 16.03572642104752, + "grad_norm": 0.2262917309999466, + "learning_rate": 8.052368705247493e-06, + "loss": 0.0196, + "step": 82600 + }, + { + "epoch": 16.03766807436532, + "grad_norm": 0.13515304028987885, + "learning_rate": 8.043372742626114e-06, + "loss": 0.0233, + "step": 82610 + }, + { + "epoch": 16.03960972768312, + "grad_norm": 0.38890278339385986, + "learning_rate": 8.034381368318422e-06, + "loss": 0.0183, + "step": 82620 + }, + { + "epoch": 16.04155138100092, + "grad_norm": 0.128348246216774, + "learning_rate": 8.025394583307677e-06, + "loss": 0.0184, + "step": 82630 + }, + { + "epoch": 16.043493034318722, + "grad_norm": 0.2092525064945221, + "learning_rate": 8.016412388576683e-06, + "loss": 0.0284, + "step": 82640 + }, + { + "epoch": 16.045434687636522, + "grad_norm": 0.1867913156747818, + "learning_rate": 8.007434785107699e-06, + "loss": 0.0203, + "step": 82650 + }, + { + "epoch": 16.047376340954322, + "grad_norm": 0.24685876071453094, + "learning_rate": 7.998461773882509e-06, + "loss": 0.0256, + "step": 82660 + }, + { + "epoch": 16.049317994272123, + "grad_norm": 0.062332287430763245, + "learning_rate": 7.989493355882393e-06, + "loss": 0.0168, + "step": 82670 + }, + { + "epoch": 16.051259647589923, + "grad_norm": 0.12502078711986542, + "learning_rate": 7.980529532088137e-06, + "loss": 0.0202, + "step": 82680 + }, + { + "epoch": 16.053201300907723, + "grad_norm": 0.12994782626628876, + "learning_rate": 7.97157030347998e-06, + "loss": 0.0209, + "step": 82690 + }, + { + "epoch": 16.055142954225524, + "grad_norm": 0.13097330927848816, + "learning_rate": 7.96261567103772e-06, + "loss": 0.0173, + "step": 82700 + }, + { + "epoch": 16.057084607543324, + "grad_norm": 0.3394632935523987, + "learning_rate": 7.953665635740593e-06, + "loss": 0.0254, + "step": 82710 + }, + { + "epoch": 16.059026260861124, + "grad_norm": 0.2462237924337387, + "learning_rate": 7.94472019856738e-06, + "loss": 0.0239, + "step": 82720 + }, + { + "epoch": 16.060967914178924, + "grad_norm": 0.12212710082530975, + "learning_rate": 7.935779360496337e-06, + "loss": 0.0243, + "step": 82730 + }, + { + "epoch": 16.062909567496725, + "grad_norm": 0.22391322255134583, + "learning_rate": 7.92684312250523e-06, + "loss": 0.0207, + "step": 82740 + }, + { + "epoch": 16.064851220814525, + "grad_norm": 0.10132184624671936, + "learning_rate": 7.917911485571289e-06, + "loss": 0.0187, + "step": 82750 + }, + { + "epoch": 16.066792874132325, + "grad_norm": 0.19127674400806427, + "learning_rate": 7.90898445067128e-06, + "loss": 0.0281, + "step": 82760 + }, + { + "epoch": 16.068734527450125, + "grad_norm": 0.11697342246770859, + "learning_rate": 7.900062018781462e-06, + "loss": 0.0181, + "step": 82770 + }, + { + "epoch": 16.070676180767922, + "grad_norm": 0.19045281410217285, + "learning_rate": 7.891144190877553e-06, + "loss": 0.0228, + "step": 82780 + }, + { + "epoch": 16.072617834085722, + "grad_norm": 0.21889479458332062, + "learning_rate": 7.882230967934806e-06, + "loss": 0.0246, + "step": 82790 + }, + { + "epoch": 16.074559487403523, + "grad_norm": 0.18560650944709778, + "learning_rate": 7.873322350927959e-06, + "loss": 0.019, + "step": 82800 + }, + { + "epoch": 16.076501140721323, + "grad_norm": 0.23954324424266815, + "learning_rate": 7.864418340831253e-06, + "loss": 0.0237, + "step": 82810 + }, + { + "epoch": 16.078442794039123, + "grad_norm": 0.15906167030334473, + "learning_rate": 7.855518938618395e-06, + "loss": 0.0189, + "step": 82820 + }, + { + "epoch": 16.080384447356924, + "grad_norm": 0.1935584992170334, + "learning_rate": 7.846624145262633e-06, + "loss": 0.0209, + "step": 82830 + }, + { + "epoch": 16.082326100674724, + "grad_norm": 0.20337910950183868, + "learning_rate": 7.837733961736672e-06, + "loss": 0.0201, + "step": 82840 + }, + { + "epoch": 16.084267753992524, + "grad_norm": 0.23747749626636505, + "learning_rate": 7.828848389012727e-06, + "loss": 0.019, + "step": 82850 + }, + { + "epoch": 16.086209407310324, + "grad_norm": 0.17627079784870148, + "learning_rate": 7.819967428062524e-06, + "loss": 0.0247, + "step": 82860 + }, + { + "epoch": 16.088151060628125, + "grad_norm": 0.20006738603115082, + "learning_rate": 7.811091079857274e-06, + "loss": 0.028, + "step": 82870 + }, + { + "epoch": 16.090092713945925, + "grad_norm": 0.25880858302116394, + "learning_rate": 7.802219345367667e-06, + "loss": 0.0237, + "step": 82880 + }, + { + "epoch": 16.092034367263725, + "grad_norm": 0.27975594997406006, + "learning_rate": 7.793352225563916e-06, + "loss": 0.0225, + "step": 82890 + }, + { + "epoch": 16.093976020581525, + "grad_norm": 0.3220199942588806, + "learning_rate": 7.7844897214157e-06, + "loss": 0.0236, + "step": 82900 + }, + { + "epoch": 16.095917673899326, + "grad_norm": 0.21197204291820526, + "learning_rate": 7.775631833892228e-06, + "loss": 0.026, + "step": 82910 + }, + { + "epoch": 16.097859327217126, + "grad_norm": 0.14854758977890015, + "learning_rate": 7.76677856396215e-06, + "loss": 0.0213, + "step": 82920 + }, + { + "epoch": 16.099800980534926, + "grad_norm": 0.162013977766037, + "learning_rate": 7.757929912593697e-06, + "loss": 0.0199, + "step": 82930 + }, + { + "epoch": 16.101742633852727, + "grad_norm": 0.11215558648109436, + "learning_rate": 7.749085880754509e-06, + "loss": 0.0214, + "step": 82940 + }, + { + "epoch": 16.103684287170527, + "grad_norm": 0.13546621799468994, + "learning_rate": 7.740246469411777e-06, + "loss": 0.0201, + "step": 82950 + }, + { + "epoch": 16.105625940488327, + "grad_norm": 0.18547935783863068, + "learning_rate": 7.731411679532142e-06, + "loss": 0.0236, + "step": 82960 + }, + { + "epoch": 16.107567593806127, + "grad_norm": 0.17559298872947693, + "learning_rate": 7.722581512081779e-06, + "loss": 0.0176, + "step": 82970 + }, + { + "epoch": 16.109509247123928, + "grad_norm": 0.2080354243516922, + "learning_rate": 7.713755968026331e-06, + "loss": 0.0212, + "step": 82980 + }, + { + "epoch": 16.111450900441728, + "grad_norm": 0.2327686846256256, + "learning_rate": 7.704935048330969e-06, + "loss": 0.0191, + "step": 82990 + }, + { + "epoch": 16.113392553759525, + "grad_norm": 0.17695845663547516, + "learning_rate": 7.696118753960313e-06, + "loss": 0.021, + "step": 83000 + }, + { + "epoch": 16.115334207077325, + "grad_norm": 0.2353699803352356, + "learning_rate": 7.687307085878503e-06, + "loss": 0.0198, + "step": 83010 + }, + { + "epoch": 16.117275860395125, + "grad_norm": 0.140261709690094, + "learning_rate": 7.678500045049186e-06, + "loss": 0.0213, + "step": 83020 + }, + { + "epoch": 16.119217513712925, + "grad_norm": 0.1148105263710022, + "learning_rate": 7.669697632435463e-06, + "loss": 0.0162, + "step": 83030 + }, + { + "epoch": 16.121159167030726, + "grad_norm": 0.2970634400844574, + "learning_rate": 7.660899848999963e-06, + "loss": 0.0228, + "step": 83040 + }, + { + "epoch": 16.123100820348526, + "grad_norm": 0.20039896667003632, + "learning_rate": 7.652106695704803e-06, + "loss": 0.0144, + "step": 83050 + }, + { + "epoch": 16.125042473666326, + "grad_norm": 0.21443365514278412, + "learning_rate": 7.643318173511598e-06, + "loss": 0.0244, + "step": 83060 + }, + { + "epoch": 16.126984126984127, + "grad_norm": 0.18997490406036377, + "learning_rate": 7.634534283381423e-06, + "loss": 0.0285, + "step": 83070 + }, + { + "epoch": 16.128925780301927, + "grad_norm": 0.13533784449100494, + "learning_rate": 7.625755026274889e-06, + "loss": 0.0165, + "step": 83080 + }, + { + "epoch": 16.130867433619727, + "grad_norm": 0.23809999227523804, + "learning_rate": 7.616980403152069e-06, + "loss": 0.0239, + "step": 83090 + }, + { + "epoch": 16.132809086937527, + "grad_norm": 0.2011236697435379, + "learning_rate": 7.608210414972561e-06, + "loss": 0.0224, + "step": 83100 + }, + { + "epoch": 16.134750740255328, + "grad_norm": 0.1873779147863388, + "learning_rate": 7.599445062695404e-06, + "loss": 0.0209, + "step": 83110 + }, + { + "epoch": 16.136692393573128, + "grad_norm": 0.275983065366745, + "learning_rate": 7.590684347279214e-06, + "loss": 0.0257, + "step": 83120 + }, + { + "epoch": 16.138634046890928, + "grad_norm": 0.21736130118370056, + "learning_rate": 7.581928269682004e-06, + "loss": 0.0213, + "step": 83130 + }, + { + "epoch": 16.14057570020873, + "grad_norm": 0.16795258224010468, + "learning_rate": 7.573176830861361e-06, + "loss": 0.0225, + "step": 83140 + }, + { + "epoch": 16.14251735352653, + "grad_norm": 0.16184261441230774, + "learning_rate": 7.5644300317743e-06, + "loss": 0.0238, + "step": 83150 + }, + { + "epoch": 16.14445900684433, + "grad_norm": 0.10532806813716888, + "learning_rate": 7.555687873377376e-06, + "loss": 0.0182, + "step": 83160 + }, + { + "epoch": 16.14640066016213, + "grad_norm": 0.12946800887584686, + "learning_rate": 7.546950356626592e-06, + "loss": 0.0204, + "step": 83170 + }, + { + "epoch": 16.14834231347993, + "grad_norm": 0.18722061812877655, + "learning_rate": 7.538217482477516e-06, + "loss": 0.0191, + "step": 83180 + }, + { + "epoch": 16.15028396679773, + "grad_norm": 0.18131467700004578, + "learning_rate": 7.529489251885119e-06, + "loss": 0.0186, + "step": 83190 + }, + { + "epoch": 16.15222562011553, + "grad_norm": 0.24097736179828644, + "learning_rate": 7.520765665803941e-06, + "loss": 0.0219, + "step": 83200 + }, + { + "epoch": 16.154167273433327, + "grad_norm": 0.1803445965051651, + "learning_rate": 7.512046725187949e-06, + "loss": 0.0242, + "step": 83210 + }, + { + "epoch": 16.156108926751127, + "grad_norm": 0.2011333853006363, + "learning_rate": 7.503332430990645e-06, + "loss": 0.0184, + "step": 83220 + }, + { + "epoch": 16.158050580068927, + "grad_norm": 0.15974737703800201, + "learning_rate": 7.494622784165017e-06, + "loss": 0.0189, + "step": 83230 + }, + { + "epoch": 16.159992233386728, + "grad_norm": 0.1580742597579956, + "learning_rate": 7.485917785663543e-06, + "loss": 0.0256, + "step": 83240 + }, + { + "epoch": 16.161933886704528, + "grad_norm": 0.21305087208747864, + "learning_rate": 7.477217436438172e-06, + "loss": 0.0261, + "step": 83250 + }, + { + "epoch": 16.163875540022328, + "grad_norm": 0.16249346733093262, + "learning_rate": 7.46852173744037e-06, + "loss": 0.0231, + "step": 83260 + }, + { + "epoch": 16.16581719334013, + "grad_norm": 0.16418041288852692, + "learning_rate": 7.4598306896211004e-06, + "loss": 0.0196, + "step": 83270 + }, + { + "epoch": 16.16775884665793, + "grad_norm": 0.21266748011112213, + "learning_rate": 7.451144293930773e-06, + "loss": 0.0206, + "step": 83280 + }, + { + "epoch": 16.16970049997573, + "grad_norm": 0.22284109890460968, + "learning_rate": 7.4424625513193455e-06, + "loss": 0.0277, + "step": 83290 + }, + { + "epoch": 16.17164215329353, + "grad_norm": 0.23941411077976227, + "learning_rate": 7.433785462736209e-06, + "loss": 0.0211, + "step": 83300 + }, + { + "epoch": 16.17358380661133, + "grad_norm": 0.1845758557319641, + "learning_rate": 7.425113029130315e-06, + "loss": 0.0234, + "step": 83310 + }, + { + "epoch": 16.17552545992913, + "grad_norm": 0.11509165167808533, + "learning_rate": 7.416445251450044e-06, + "loss": 0.0216, + "step": 83320 + }, + { + "epoch": 16.17746711324693, + "grad_norm": 0.1460636854171753, + "learning_rate": 7.4077821306433035e-06, + "loss": 0.0243, + "step": 83330 + }, + { + "epoch": 16.17940876656473, + "grad_norm": 0.2189902514219284, + "learning_rate": 7.399123667657465e-06, + "loss": 0.0218, + "step": 83340 + }, + { + "epoch": 16.18135041988253, + "grad_norm": 0.2416185587644577, + "learning_rate": 7.390469863439426e-06, + "loss": 0.0182, + "step": 83350 + }, + { + "epoch": 16.18329207320033, + "grad_norm": 0.15278543531894684, + "learning_rate": 7.381820718935512e-06, + "loss": 0.0182, + "step": 83360 + }, + { + "epoch": 16.18523372651813, + "grad_norm": 0.1601364016532898, + "learning_rate": 7.373176235091633e-06, + "loss": 0.0216, + "step": 83370 + }, + { + "epoch": 16.18717537983593, + "grad_norm": 0.3970271050930023, + "learning_rate": 7.364536412853101e-06, + "loss": 0.0219, + "step": 83380 + }, + { + "epoch": 16.18911703315373, + "grad_norm": 0.2827443778514862, + "learning_rate": 7.355901253164776e-06, + "loss": 0.0223, + "step": 83390 + }, + { + "epoch": 16.191058686471532, + "grad_norm": 0.14167092740535736, + "learning_rate": 7.34727075697097e-06, + "loss": 0.0229, + "step": 83400 + }, + { + "epoch": 16.193000339789332, + "grad_norm": 0.14907582104206085, + "learning_rate": 7.3386449252155095e-06, + "loss": 0.0211, + "step": 83410 + }, + { + "epoch": 16.19494199310713, + "grad_norm": 0.12699247896671295, + "learning_rate": 7.330023758841692e-06, + "loss": 0.0198, + "step": 83420 + }, + { + "epoch": 16.19688364642493, + "grad_norm": 0.24015778303146362, + "learning_rate": 7.321407258792323e-06, + "loss": 0.0265, + "step": 83430 + }, + { + "epoch": 16.19882529974273, + "grad_norm": 0.18107064068317413, + "learning_rate": 7.312795426009694e-06, + "loss": 0.0211, + "step": 83440 + }, + { + "epoch": 16.20076695306053, + "grad_norm": 0.17983685433864594, + "learning_rate": 7.304188261435585e-06, + "loss": 0.0232, + "step": 83450 + }, + { + "epoch": 16.20270860637833, + "grad_norm": 0.23138119280338287, + "learning_rate": 7.295585766011242e-06, + "loss": 0.0292, + "step": 83460 + }, + { + "epoch": 16.20465025969613, + "grad_norm": 0.12998400628566742, + "learning_rate": 7.286987940677436e-06, + "loss": 0.0216, + "step": 83470 + }, + { + "epoch": 16.20659191301393, + "grad_norm": 0.17189881205558777, + "learning_rate": 7.278394786374426e-06, + "loss": 0.0251, + "step": 83480 + }, + { + "epoch": 16.20853356633173, + "grad_norm": 0.17136453092098236, + "learning_rate": 7.269806304041915e-06, + "loss": 0.0187, + "step": 83490 + }, + { + "epoch": 16.21047521964953, + "grad_norm": 0.08966103196144104, + "learning_rate": 7.261222494619147e-06, + "loss": 0.0185, + "step": 83500 + }, + { + "epoch": 16.21241687296733, + "grad_norm": 0.23871009051799774, + "learning_rate": 7.252643359044825e-06, + "loss": 0.0215, + "step": 83510 + }, + { + "epoch": 16.21435852628513, + "grad_norm": 0.17522914707660675, + "learning_rate": 7.244068898257172e-06, + "loss": 0.0217, + "step": 83520 + }, + { + "epoch": 16.216300179602932, + "grad_norm": 0.17390520870685577, + "learning_rate": 7.2354991131938495e-06, + "loss": 0.0215, + "step": 83530 + }, + { + "epoch": 16.218241832920732, + "grad_norm": 0.15179398655891418, + "learning_rate": 7.2269340047920574e-06, + "loss": 0.0222, + "step": 83540 + }, + { + "epoch": 16.220183486238533, + "grad_norm": 0.13549646735191345, + "learning_rate": 7.218373573988435e-06, + "loss": 0.0239, + "step": 83550 + }, + { + "epoch": 16.222125139556333, + "grad_norm": 0.15921209752559662, + "learning_rate": 7.209817821719178e-06, + "loss": 0.022, + "step": 83560 + }, + { + "epoch": 16.224066792874133, + "grad_norm": 0.18122375011444092, + "learning_rate": 7.201266748919894e-06, + "loss": 0.0231, + "step": 83570 + }, + { + "epoch": 16.226008446191933, + "grad_norm": 0.32668226957321167, + "learning_rate": 7.19272035652575e-06, + "loss": 0.0273, + "step": 83580 + }, + { + "epoch": 16.227950099509734, + "grad_norm": 0.1546870917081833, + "learning_rate": 7.184178645471334e-06, + "loss": 0.0228, + "step": 83590 + }, + { + "epoch": 16.229891752827534, + "grad_norm": 0.15875285863876343, + "learning_rate": 7.175641616690775e-06, + "loss": 0.0235, + "step": 83600 + }, + { + "epoch": 16.231833406145334, + "grad_norm": 0.2208355814218521, + "learning_rate": 7.167109271117661e-06, + "loss": 0.0259, + "step": 83610 + }, + { + "epoch": 16.233775059463134, + "grad_norm": 0.11254078149795532, + "learning_rate": 7.158581609685072e-06, + "loss": 0.022, + "step": 83620 + }, + { + "epoch": 16.23571671278093, + "grad_norm": 0.1953449845314026, + "learning_rate": 7.150058633325591e-06, + "loss": 0.026, + "step": 83630 + }, + { + "epoch": 16.23765836609873, + "grad_norm": 0.09887132793664932, + "learning_rate": 7.141540342971287e-06, + "loss": 0.0241, + "step": 83640 + }, + { + "epoch": 16.23960001941653, + "grad_norm": 0.18685020506381989, + "learning_rate": 7.133026739553677e-06, + "loss": 0.0213, + "step": 83650 + }, + { + "epoch": 16.241541672734332, + "grad_norm": 0.20687764883041382, + "learning_rate": 7.124517824003829e-06, + "loss": 0.0215, + "step": 83660 + }, + { + "epoch": 16.243483326052132, + "grad_norm": 0.3156892657279968, + "learning_rate": 7.116013597252236e-06, + "loss": 0.0281, + "step": 83670 + }, + { + "epoch": 16.245424979369933, + "grad_norm": 0.183713898062706, + "learning_rate": 7.107514060228921e-06, + "loss": 0.0244, + "step": 83680 + }, + { + "epoch": 16.247366632687733, + "grad_norm": 0.14010313153266907, + "learning_rate": 7.099019213863378e-06, + "loss": 0.0175, + "step": 83690 + }, + { + "epoch": 16.249308286005533, + "grad_norm": 0.14723680913448334, + "learning_rate": 7.090529059084611e-06, + "loss": 0.0193, + "step": 83700 + }, + { + "epoch": 16.251249939323333, + "grad_norm": 0.18826667964458466, + "learning_rate": 7.082043596821058e-06, + "loss": 0.0191, + "step": 83710 + }, + { + "epoch": 16.253191592641134, + "grad_norm": 0.18686354160308838, + "learning_rate": 7.07356282800069e-06, + "loss": 0.024, + "step": 83720 + }, + { + "epoch": 16.255133245958934, + "grad_norm": 0.1432783007621765, + "learning_rate": 7.0650867535509645e-06, + "loss": 0.0215, + "step": 83730 + }, + { + "epoch": 16.257074899276734, + "grad_norm": 0.20601704716682434, + "learning_rate": 7.0566153743987895e-06, + "loss": 0.0228, + "step": 83740 + }, + { + "epoch": 16.259016552594534, + "grad_norm": 0.22462129592895508, + "learning_rate": 7.04814869147059e-06, + "loss": 0.0281, + "step": 83750 + }, + { + "epoch": 16.260958205912335, + "grad_norm": 0.18684929609298706, + "learning_rate": 7.0396867056922725e-06, + "loss": 0.0228, + "step": 83760 + }, + { + "epoch": 16.262899859230135, + "grad_norm": 0.15217436850070953, + "learning_rate": 7.031229417989244e-06, + "loss": 0.0157, + "step": 83770 + }, + { + "epoch": 16.264841512547935, + "grad_norm": 0.19984552264213562, + "learning_rate": 7.022776829286348e-06, + "loss": 0.0223, + "step": 83780 + }, + { + "epoch": 16.266783165865736, + "grad_norm": 0.16204167902469635, + "learning_rate": 7.014328940507975e-06, + "loss": 0.0246, + "step": 83790 + }, + { + "epoch": 16.268724819183536, + "grad_norm": 0.20909641683101654, + "learning_rate": 7.005885752577945e-06, + "loss": 0.0233, + "step": 83800 + }, + { + "epoch": 16.270666472501336, + "grad_norm": 0.21053563058376312, + "learning_rate": 6.997447266419615e-06, + "loss": 0.0175, + "step": 83810 + }, + { + "epoch": 16.272608125819136, + "grad_norm": 0.19904457032680511, + "learning_rate": 6.989013482955797e-06, + "loss": 0.0162, + "step": 83820 + }, + { + "epoch": 16.274549779136937, + "grad_norm": 0.21165744960308075, + "learning_rate": 6.980584403108808e-06, + "loss": 0.0195, + "step": 83830 + }, + { + "epoch": 16.276491432454737, + "grad_norm": 0.24322101473808289, + "learning_rate": 6.972160027800417e-06, + "loss": 0.02, + "step": 83840 + }, + { + "epoch": 16.278433085772534, + "grad_norm": 0.1918199062347412, + "learning_rate": 6.963740357951931e-06, + "loss": 0.0272, + "step": 83850 + }, + { + "epoch": 16.280374739090334, + "grad_norm": 0.26913559436798096, + "learning_rate": 6.95532539448408e-06, + "loss": 0.028, + "step": 83860 + }, + { + "epoch": 16.282316392408134, + "grad_norm": 0.07935002446174622, + "learning_rate": 6.946915138317128e-06, + "loss": 0.0216, + "step": 83870 + }, + { + "epoch": 16.284258045725934, + "grad_norm": 0.13417723774909973, + "learning_rate": 6.9385095903708065e-06, + "loss": 0.0198, + "step": 83880 + }, + { + "epoch": 16.286199699043735, + "grad_norm": 0.1607789695262909, + "learning_rate": 6.930108751564346e-06, + "loss": 0.02, + "step": 83890 + }, + { + "epoch": 16.288141352361535, + "grad_norm": 0.24954591691493988, + "learning_rate": 6.921712622816428e-06, + "loss": 0.0267, + "step": 83900 + }, + { + "epoch": 16.290083005679335, + "grad_norm": 0.26082223653793335, + "learning_rate": 6.913321205045259e-06, + "loss": 0.0245, + "step": 83910 + }, + { + "epoch": 16.292024658997136, + "grad_norm": 0.16918078064918518, + "learning_rate": 6.904934499168491e-06, + "loss": 0.0272, + "step": 83920 + }, + { + "epoch": 16.293966312314936, + "grad_norm": 0.20324987173080444, + "learning_rate": 6.896552506103299e-06, + "loss": 0.0228, + "step": 83930 + }, + { + "epoch": 16.295907965632736, + "grad_norm": 0.16167007386684418, + "learning_rate": 6.888175226766313e-06, + "loss": 0.02, + "step": 83940 + }, + { + "epoch": 16.297849618950536, + "grad_norm": 0.4267142713069916, + "learning_rate": 6.879802662073686e-06, + "loss": 0.0204, + "step": 83950 + }, + { + "epoch": 16.299791272268337, + "grad_norm": 0.26951754093170166, + "learning_rate": 6.871434812940991e-06, + "loss": 0.0224, + "step": 83960 + }, + { + "epoch": 16.301732925586137, + "grad_norm": 0.28904253244400024, + "learning_rate": 6.863071680283345e-06, + "loss": 0.0158, + "step": 83970 + }, + { + "epoch": 16.303674578903937, + "grad_norm": 0.2281988114118576, + "learning_rate": 6.854713265015339e-06, + "loss": 0.0275, + "step": 83980 + }, + { + "epoch": 16.305616232221738, + "grad_norm": 0.13558825850486755, + "learning_rate": 6.8463595680510025e-06, + "loss": 0.0193, + "step": 83990 + }, + { + "epoch": 16.307557885539538, + "grad_norm": 0.20475411415100098, + "learning_rate": 6.838010590303911e-06, + "loss": 0.0191, + "step": 84000 + }, + { + "epoch": 16.309499538857338, + "grad_norm": 0.13014499843120575, + "learning_rate": 6.829666332687085e-06, + "loss": 0.0162, + "step": 84010 + }, + { + "epoch": 16.31144119217514, + "grad_norm": 0.20545974373817444, + "learning_rate": 6.821326796113059e-06, + "loss": 0.0224, + "step": 84020 + }, + { + "epoch": 16.31338284549294, + "grad_norm": 0.2170015275478363, + "learning_rate": 6.812991981493799e-06, + "loss": 0.0217, + "step": 84030 + }, + { + "epoch": 16.31532449881074, + "grad_norm": 0.16838273406028748, + "learning_rate": 6.804661889740821e-06, + "loss": 0.0288, + "step": 84040 + }, + { + "epoch": 16.31726615212854, + "grad_norm": 0.1921374350786209, + "learning_rate": 6.7963365217650625e-06, + "loss": 0.0233, + "step": 84050 + }, + { + "epoch": 16.319207805446336, + "grad_norm": 0.09585386514663696, + "learning_rate": 6.788015878476983e-06, + "loss": 0.0247, + "step": 84060 + }, + { + "epoch": 16.321149458764136, + "grad_norm": 0.16013231873512268, + "learning_rate": 6.779699960786523e-06, + "loss": 0.0264, + "step": 84070 + }, + { + "epoch": 16.323091112081936, + "grad_norm": 0.11263155192136765, + "learning_rate": 6.771388769603104e-06, + "loss": 0.0181, + "step": 84080 + }, + { + "epoch": 16.325032765399737, + "grad_norm": 0.16332034766674042, + "learning_rate": 6.763082305835605e-06, + "loss": 0.0249, + "step": 84090 + }, + { + "epoch": 16.326974418717537, + "grad_norm": 0.117873914539814, + "learning_rate": 6.754780570392433e-06, + "loss": 0.0209, + "step": 84100 + }, + { + "epoch": 16.328916072035337, + "grad_norm": 0.19154302775859833, + "learning_rate": 6.7464835641814215e-06, + "loss": 0.0182, + "step": 84110 + }, + { + "epoch": 16.330857725353138, + "grad_norm": 0.16891179978847504, + "learning_rate": 6.7381912881099515e-06, + "loss": 0.0224, + "step": 84120 + }, + { + "epoch": 16.332799378670938, + "grad_norm": 0.19295088946819305, + "learning_rate": 6.72990374308482e-06, + "loss": 0.0229, + "step": 84130 + }, + { + "epoch": 16.334741031988738, + "grad_norm": 0.2842915654182434, + "learning_rate": 6.721620930012379e-06, + "loss": 0.0245, + "step": 84140 + }, + { + "epoch": 16.33668268530654, + "grad_norm": 0.20997989177703857, + "learning_rate": 6.713342849798393e-06, + "loss": 0.0291, + "step": 84150 + }, + { + "epoch": 16.33862433862434, + "grad_norm": 0.14424946904182434, + "learning_rate": 6.705069503348166e-06, + "loss": 0.0212, + "step": 84160 + }, + { + "epoch": 16.34056599194214, + "grad_norm": 0.19698259234428406, + "learning_rate": 6.696800891566435e-06, + "loss": 0.0206, + "step": 84170 + }, + { + "epoch": 16.34250764525994, + "grad_norm": 0.13762179017066956, + "learning_rate": 6.688537015357449e-06, + "loss": 0.0249, + "step": 84180 + }, + { + "epoch": 16.34444929857774, + "grad_norm": 0.17616428434848785, + "learning_rate": 6.680277875624941e-06, + "loss": 0.0178, + "step": 84190 + }, + { + "epoch": 16.34639095189554, + "grad_norm": 0.12785208225250244, + "learning_rate": 6.672023473272121e-06, + "loss": 0.021, + "step": 84200 + }, + { + "epoch": 16.34833260521334, + "grad_norm": 0.23222123086452484, + "learning_rate": 6.663773809201667e-06, + "loss": 0.0225, + "step": 84210 + }, + { + "epoch": 16.35027425853114, + "grad_norm": 0.1943386346101761, + "learning_rate": 6.655528884315748e-06, + "loss": 0.0247, + "step": 84220 + }, + { + "epoch": 16.35221591184894, + "grad_norm": 0.14509351551532745, + "learning_rate": 6.647288699516036e-06, + "loss": 0.0196, + "step": 84230 + }, + { + "epoch": 16.35415756516674, + "grad_norm": 0.17976689338684082, + "learning_rate": 6.639053255703642e-06, + "loss": 0.0172, + "step": 84240 + }, + { + "epoch": 16.35609921848454, + "grad_norm": 0.1264963001012802, + "learning_rate": 6.6308225537791925e-06, + "loss": 0.0229, + "step": 84250 + }, + { + "epoch": 16.35804087180234, + "grad_norm": 0.11867012083530426, + "learning_rate": 6.622596594642777e-06, + "loss": 0.0229, + "step": 84260 + }, + { + "epoch": 16.359982525120138, + "grad_norm": 0.15031032264232635, + "learning_rate": 6.614375379193999e-06, + "loss": 0.0183, + "step": 84270 + }, + { + "epoch": 16.36192417843794, + "grad_norm": 0.14035825431346893, + "learning_rate": 6.606158908331889e-06, + "loss": 0.0151, + "step": 84280 + }, + { + "epoch": 16.36386583175574, + "grad_norm": 0.1494300216436386, + "learning_rate": 6.597947182955005e-06, + "loss": 0.0232, + "step": 84290 + }, + { + "epoch": 16.36580748507354, + "grad_norm": 0.26378965377807617, + "learning_rate": 6.589740203961359e-06, + "loss": 0.0194, + "step": 84300 + }, + { + "epoch": 16.36774913839134, + "grad_norm": 0.17452193796634674, + "learning_rate": 6.581537972248464e-06, + "loss": 0.027, + "step": 84310 + }, + { + "epoch": 16.36969079170914, + "grad_norm": 0.2527880370616913, + "learning_rate": 6.573340488713275e-06, + "loss": 0.0238, + "step": 84320 + }, + { + "epoch": 16.37163244502694, + "grad_norm": 0.17634007334709167, + "learning_rate": 6.565147754252304e-06, + "loss": 0.0188, + "step": 84330 + }, + { + "epoch": 16.37357409834474, + "grad_norm": 0.15700142085552216, + "learning_rate": 6.556959769761462e-06, + "loss": 0.0203, + "step": 84340 + }, + { + "epoch": 16.37551575166254, + "grad_norm": 0.18441826105117798, + "learning_rate": 6.548776536136192e-06, + "loss": 0.0197, + "step": 84350 + }, + { + "epoch": 16.37745740498034, + "grad_norm": 0.15369707345962524, + "learning_rate": 6.54059805427138e-06, + "loss": 0.0235, + "step": 84360 + }, + { + "epoch": 16.37939905829814, + "grad_norm": 0.2092394083738327, + "learning_rate": 6.532424325061437e-06, + "loss": 0.0227, + "step": 84370 + }, + { + "epoch": 16.38134071161594, + "grad_norm": 0.13647912442684174, + "learning_rate": 6.5242553494001935e-06, + "loss": 0.0208, + "step": 84380 + }, + { + "epoch": 16.38328236493374, + "grad_norm": 0.15787287056446075, + "learning_rate": 6.516091128181046e-06, + "loss": 0.0209, + "step": 84390 + }, + { + "epoch": 16.38522401825154, + "grad_norm": 0.14525337517261505, + "learning_rate": 6.507931662296785e-06, + "loss": 0.0208, + "step": 84400 + }, + { + "epoch": 16.387165671569342, + "grad_norm": 0.20751513540744781, + "learning_rate": 6.4997769526397354e-06, + "loss": 0.0216, + "step": 84410 + }, + { + "epoch": 16.389107324887142, + "grad_norm": 0.12998022139072418, + "learning_rate": 6.491627000101663e-06, + "loss": 0.0208, + "step": 84420 + }, + { + "epoch": 16.391048978204942, + "grad_norm": 0.18299435079097748, + "learning_rate": 6.4834818055738545e-06, + "loss": 0.022, + "step": 84430 + }, + { + "epoch": 16.392990631522743, + "grad_norm": 0.21897034347057343, + "learning_rate": 6.4753413699470465e-06, + "loss": 0.0244, + "step": 84440 + }, + { + "epoch": 16.394932284840543, + "grad_norm": 0.1410062611103058, + "learning_rate": 6.467205694111478e-06, + "loss": 0.0166, + "step": 84450 + }, + { + "epoch": 16.396873938158343, + "grad_norm": 0.21672764420509338, + "learning_rate": 6.459074778956836e-06, + "loss": 0.0216, + "step": 84460 + }, + { + "epoch": 16.398815591476144, + "grad_norm": 0.10290736705064774, + "learning_rate": 6.450948625372311e-06, + "loss": 0.0175, + "step": 84470 + }, + { + "epoch": 16.400757244793944, + "grad_norm": 0.17442144453525543, + "learning_rate": 6.442827234246585e-06, + "loss": 0.0219, + "step": 84480 + }, + { + "epoch": 16.40269889811174, + "grad_norm": 0.1331513375043869, + "learning_rate": 6.43471060646777e-06, + "loss": 0.0243, + "step": 84490 + }, + { + "epoch": 16.40464055142954, + "grad_norm": 0.17160987854003906, + "learning_rate": 6.426598742923512e-06, + "loss": 0.0216, + "step": 84500 + }, + { + "epoch": 16.40658220474734, + "grad_norm": 0.2083979845046997, + "learning_rate": 6.418491644500885e-06, + "loss": 0.0328, + "step": 84510 + }, + { + "epoch": 16.40852385806514, + "grad_norm": 0.09995948523283005, + "learning_rate": 6.410389312086512e-06, + "loss": 0.0217, + "step": 84520 + }, + { + "epoch": 16.41046551138294, + "grad_norm": 0.142198383808136, + "learning_rate": 6.4022917465664135e-06, + "loss": 0.0184, + "step": 84530 + }, + { + "epoch": 16.412407164700742, + "grad_norm": 0.21718527376651764, + "learning_rate": 6.3941989488261575e-06, + "loss": 0.0232, + "step": 84540 + }, + { + "epoch": 16.414348818018542, + "grad_norm": 0.2457498162984848, + "learning_rate": 6.386110919750726e-06, + "loss": 0.0245, + "step": 84550 + }, + { + "epoch": 16.416290471336342, + "grad_norm": 0.11810588091611862, + "learning_rate": 6.378027660224651e-06, + "loss": 0.0227, + "step": 84560 + }, + { + "epoch": 16.418232124654143, + "grad_norm": 0.16387403011322021, + "learning_rate": 6.369949171131856e-06, + "loss": 0.0261, + "step": 84570 + }, + { + "epoch": 16.420173777971943, + "grad_norm": 0.11084593832492828, + "learning_rate": 6.361875453355848e-06, + "loss": 0.0218, + "step": 84580 + }, + { + "epoch": 16.422115431289743, + "grad_norm": 0.24214789271354675, + "learning_rate": 6.353806507779525e-06, + "loss": 0.0233, + "step": 84590 + }, + { + "epoch": 16.424057084607544, + "grad_norm": 0.14684411883354187, + "learning_rate": 6.3457423352853044e-06, + "loss": 0.0211, + "step": 84600 + }, + { + "epoch": 16.425998737925344, + "grad_norm": 0.1241581067442894, + "learning_rate": 6.337682936755062e-06, + "loss": 0.0221, + "step": 84610 + }, + { + "epoch": 16.427940391243144, + "grad_norm": 0.1730828434228897, + "learning_rate": 6.329628313070179e-06, + "loss": 0.0157, + "step": 84620 + }, + { + "epoch": 16.429882044560944, + "grad_norm": 0.11468833684921265, + "learning_rate": 6.321578465111477e-06, + "loss": 0.0222, + "step": 84630 + }, + { + "epoch": 16.431823697878745, + "grad_norm": 0.13632774353027344, + "learning_rate": 6.31353339375928e-06, + "loss": 0.02, + "step": 84640 + }, + { + "epoch": 16.433765351196545, + "grad_norm": 0.23659217357635498, + "learning_rate": 6.305493099893395e-06, + "loss": 0.0205, + "step": 84650 + }, + { + "epoch": 16.435707004514345, + "grad_norm": 0.182918518781662, + "learning_rate": 6.297457584393096e-06, + "loss": 0.0241, + "step": 84660 + }, + { + "epoch": 16.437648657832145, + "grad_norm": 0.19691848754882812, + "learning_rate": 6.289426848137126e-06, + "loss": 0.025, + "step": 84670 + }, + { + "epoch": 16.439590311149946, + "grad_norm": 0.19935448467731476, + "learning_rate": 6.281400892003713e-06, + "loss": 0.0196, + "step": 84680 + }, + { + "epoch": 16.441531964467746, + "grad_norm": 0.23902250826358795, + "learning_rate": 6.273379716870581e-06, + "loss": 0.0227, + "step": 84690 + }, + { + "epoch": 16.443473617785543, + "grad_norm": 0.19797436892986298, + "learning_rate": 6.265363323614887e-06, + "loss": 0.0205, + "step": 84700 + }, + { + "epoch": 16.445415271103343, + "grad_norm": 0.2423199713230133, + "learning_rate": 6.257351713113308e-06, + "loss": 0.0167, + "step": 84710 + }, + { + "epoch": 16.447356924421143, + "grad_norm": 0.1501845270395279, + "learning_rate": 6.249344886241981e-06, + "loss": 0.0173, + "step": 84720 + }, + { + "epoch": 16.449298577738944, + "grad_norm": 0.21908128261566162, + "learning_rate": 6.2413428438765295e-06, + "loss": 0.0271, + "step": 84730 + }, + { + "epoch": 16.451240231056744, + "grad_norm": 0.21329180896282196, + "learning_rate": 6.233345586892026e-06, + "loss": 0.0205, + "step": 84740 + }, + { + "epoch": 16.453181884374544, + "grad_norm": 0.14963874220848083, + "learning_rate": 6.225353116163052e-06, + "loss": 0.017, + "step": 84750 + }, + { + "epoch": 16.455123537692344, + "grad_norm": 0.11909466981887817, + "learning_rate": 6.217365432563627e-06, + "loss": 0.0207, + "step": 84760 + }, + { + "epoch": 16.457065191010145, + "grad_norm": 0.16530638933181763, + "learning_rate": 6.209382536967318e-06, + "loss": 0.03, + "step": 84770 + }, + { + "epoch": 16.459006844327945, + "grad_norm": 0.18778301775455475, + "learning_rate": 6.201404430247082e-06, + "loss": 0.0211, + "step": 84780 + }, + { + "epoch": 16.460948497645745, + "grad_norm": 0.26357972621917725, + "learning_rate": 6.1934311132754165e-06, + "loss": 0.0226, + "step": 84790 + }, + { + "epoch": 16.462890150963545, + "grad_norm": 0.16043707728385925, + "learning_rate": 6.185462586924257e-06, + "loss": 0.0181, + "step": 84800 + }, + { + "epoch": 16.464831804281346, + "grad_norm": 0.15959011018276215, + "learning_rate": 6.177498852065039e-06, + "loss": 0.0193, + "step": 84810 + }, + { + "epoch": 16.466773457599146, + "grad_norm": 0.11165212839841843, + "learning_rate": 6.169539909568655e-06, + "loss": 0.0189, + "step": 84820 + }, + { + "epoch": 16.468715110916946, + "grad_norm": 0.05580144748091698, + "learning_rate": 6.161585760305488e-06, + "loss": 0.0171, + "step": 84830 + }, + { + "epoch": 16.470656764234747, + "grad_norm": 0.15510688722133636, + "learning_rate": 6.153636405145391e-06, + "loss": 0.0202, + "step": 84840 + }, + { + "epoch": 16.472598417552547, + "grad_norm": 0.30100134015083313, + "learning_rate": 6.145691844957702e-06, + "loss": 0.0268, + "step": 84850 + }, + { + "epoch": 16.474540070870347, + "grad_norm": 0.15433652698993683, + "learning_rate": 6.1377520806112145e-06, + "loss": 0.0137, + "step": 84860 + }, + { + "epoch": 16.476481724188147, + "grad_norm": 0.1852855086326599, + "learning_rate": 6.129817112974218e-06, + "loss": 0.0183, + "step": 84870 + }, + { + "epoch": 16.478423377505948, + "grad_norm": 0.24044348299503326, + "learning_rate": 6.121886942914457e-06, + "loss": 0.021, + "step": 84880 + }, + { + "epoch": 16.480365030823748, + "grad_norm": 0.13244536519050598, + "learning_rate": 6.11396157129917e-06, + "loss": 0.0183, + "step": 84890 + }, + { + "epoch": 16.482306684141548, + "grad_norm": 0.13564945757389069, + "learning_rate": 6.106040998995066e-06, + "loss": 0.0266, + "step": 84900 + }, + { + "epoch": 16.484248337459345, + "grad_norm": 0.130117267370224, + "learning_rate": 6.098125226868329e-06, + "loss": 0.0208, + "step": 84910 + }, + { + "epoch": 16.486189990777145, + "grad_norm": 0.20193956792354584, + "learning_rate": 6.090214255784604e-06, + "loss": 0.023, + "step": 84920 + }, + { + "epoch": 16.488131644094945, + "grad_norm": 0.17626157402992249, + "learning_rate": 6.0823080866090346e-06, + "loss": 0.0213, + "step": 84930 + }, + { + "epoch": 16.490073297412746, + "grad_norm": 0.5057781934738159, + "learning_rate": 6.074406720206233e-06, + "loss": 0.0227, + "step": 84940 + }, + { + "epoch": 16.492014950730546, + "grad_norm": 0.41445252299308777, + "learning_rate": 6.066510157440259e-06, + "loss": 0.0191, + "step": 84950 + }, + { + "epoch": 16.493956604048346, + "grad_norm": 0.19429700076580048, + "learning_rate": 6.058618399174676e-06, + "loss": 0.0239, + "step": 84960 + }, + { + "epoch": 16.495898257366147, + "grad_norm": 0.22594931721687317, + "learning_rate": 6.050731446272528e-06, + "loss": 0.0272, + "step": 84970 + }, + { + "epoch": 16.497839910683947, + "grad_norm": 0.15016885101795197, + "learning_rate": 6.042849299596321e-06, + "loss": 0.0273, + "step": 84980 + }, + { + "epoch": 16.499781564001747, + "grad_norm": 0.1606650948524475, + "learning_rate": 6.034971960008013e-06, + "loss": 0.0166, + "step": 84990 + }, + { + "epoch": 16.501723217319547, + "grad_norm": 0.126137837767601, + "learning_rate": 6.027099428369082e-06, + "loss": 0.021, + "step": 85000 + }, + { + "epoch": 16.503664870637348, + "grad_norm": 0.18850447237491608, + "learning_rate": 6.019231705540435e-06, + "loss": 0.0183, + "step": 85010 + }, + { + "epoch": 16.505606523955148, + "grad_norm": 0.16713012754917145, + "learning_rate": 6.011368792382488e-06, + "loss": 0.0208, + "step": 85020 + }, + { + "epoch": 16.507548177272948, + "grad_norm": 0.14434346556663513, + "learning_rate": 6.00351068975511e-06, + "loss": 0.0165, + "step": 85030 + }, + { + "epoch": 16.50948983059075, + "grad_norm": 0.16570642590522766, + "learning_rate": 5.9956573985176674e-06, + "loss": 0.021, + "step": 85040 + }, + { + "epoch": 16.51143148390855, + "grad_norm": 0.16196013987064362, + "learning_rate": 5.987808919528959e-06, + "loss": 0.0164, + "step": 85050 + }, + { + "epoch": 16.51337313722635, + "grad_norm": 0.2328518033027649, + "learning_rate": 5.979965253647307e-06, + "loss": 0.0234, + "step": 85060 + }, + { + "epoch": 16.51531479054415, + "grad_norm": 0.21235430240631104, + "learning_rate": 5.9721264017304655e-06, + "loss": 0.0261, + "step": 85070 + }, + { + "epoch": 16.51725644386195, + "grad_norm": 0.19345974922180176, + "learning_rate": 5.964292364635682e-06, + "loss": 0.0163, + "step": 85080 + }, + { + "epoch": 16.51919809717975, + "grad_norm": 0.1610259711742401, + "learning_rate": 5.956463143219682e-06, + "loss": 0.0171, + "step": 85090 + }, + { + "epoch": 16.52113975049755, + "grad_norm": 0.21579697728157043, + "learning_rate": 5.948638738338663e-06, + "loss": 0.0218, + "step": 85100 + }, + { + "epoch": 16.52308140381535, + "grad_norm": 0.20223282277584076, + "learning_rate": 5.940819150848271e-06, + "loss": 0.0197, + "step": 85110 + }, + { + "epoch": 16.525023057133147, + "grad_norm": 0.15147405862808228, + "learning_rate": 5.933004381603663e-06, + "loss": 0.021, + "step": 85120 + }, + { + "epoch": 16.526964710450947, + "grad_norm": 0.17390435934066772, + "learning_rate": 5.92519443145943e-06, + "loss": 0.0216, + "step": 85130 + }, + { + "epoch": 16.528906363768748, + "grad_norm": 0.09365788102149963, + "learning_rate": 5.917389301269671e-06, + "loss": 0.0216, + "step": 85140 + }, + { + "epoch": 16.530848017086548, + "grad_norm": 0.18031424283981323, + "learning_rate": 5.9095889918879435e-06, + "loss": 0.0226, + "step": 85150 + }, + { + "epoch": 16.532789670404348, + "grad_norm": 0.22249506413936615, + "learning_rate": 5.901793504167286e-06, + "loss": 0.022, + "step": 85160 + }, + { + "epoch": 16.53473132372215, + "grad_norm": 0.1373085379600525, + "learning_rate": 5.894002838960172e-06, + "loss": 0.0238, + "step": 85170 + }, + { + "epoch": 16.53667297703995, + "grad_norm": 0.21339215338230133, + "learning_rate": 5.886216997118604e-06, + "loss": 0.0206, + "step": 85180 + }, + { + "epoch": 16.53861463035775, + "grad_norm": 0.24069932103157043, + "learning_rate": 5.878435979494023e-06, + "loss": 0.0252, + "step": 85190 + }, + { + "epoch": 16.54055628367555, + "grad_norm": 0.30998027324676514, + "learning_rate": 5.870659786937344e-06, + "loss": 0.0227, + "step": 85200 + }, + { + "epoch": 16.54249793699335, + "grad_norm": 0.1881847381591797, + "learning_rate": 5.862888420298962e-06, + "loss": 0.0176, + "step": 85210 + }, + { + "epoch": 16.54443959031115, + "grad_norm": 0.21716736257076263, + "learning_rate": 5.855121880428743e-06, + "loss": 0.0253, + "step": 85220 + }, + { + "epoch": 16.54638124362895, + "grad_norm": 0.20811189711093903, + "learning_rate": 5.84736016817603e-06, + "loss": 0.0321, + "step": 85230 + }, + { + "epoch": 16.54832289694675, + "grad_norm": 0.12093234062194824, + "learning_rate": 5.839603284389622e-06, + "loss": 0.0191, + "step": 85240 + }, + { + "epoch": 16.55026455026455, + "grad_norm": 0.12843190133571625, + "learning_rate": 5.831851229917812e-06, + "loss": 0.0233, + "step": 85250 + }, + { + "epoch": 16.55220620358235, + "grad_norm": 0.2395116686820984, + "learning_rate": 5.824104005608338e-06, + "loss": 0.0173, + "step": 85260 + }, + { + "epoch": 16.55414785690015, + "grad_norm": 0.2791869044303894, + "learning_rate": 5.81636161230843e-06, + "loss": 0.0203, + "step": 85270 + }, + { + "epoch": 16.55608951021795, + "grad_norm": 0.20596624910831451, + "learning_rate": 5.808624050864791e-06, + "loss": 0.0217, + "step": 85280 + }, + { + "epoch": 16.55803116353575, + "grad_norm": 0.23187972605228424, + "learning_rate": 5.800891322123592e-06, + "loss": 0.0215, + "step": 85290 + }, + { + "epoch": 16.559972816853552, + "grad_norm": 0.19466136395931244, + "learning_rate": 5.793163426930454e-06, + "loss": 0.0182, + "step": 85300 + }, + { + "epoch": 16.561914470171352, + "grad_norm": 0.21261091530323029, + "learning_rate": 5.785440366130512e-06, + "loss": 0.0275, + "step": 85310 + }, + { + "epoch": 16.563856123489153, + "grad_norm": 0.15235473215579987, + "learning_rate": 5.777722140568326e-06, + "loss": 0.0202, + "step": 85320 + }, + { + "epoch": 16.565797776806953, + "grad_norm": 0.1846180260181427, + "learning_rate": 5.7700087510879685e-06, + "loss": 0.0234, + "step": 85330 + }, + { + "epoch": 16.56773943012475, + "grad_norm": 0.12189226597547531, + "learning_rate": 5.762300198532933e-06, + "loss": 0.0201, + "step": 85340 + }, + { + "epoch": 16.56968108344255, + "grad_norm": 0.18044304847717285, + "learning_rate": 5.7545964837462606e-06, + "loss": 0.024, + "step": 85350 + }, + { + "epoch": 16.57162273676035, + "grad_norm": 0.2344408929347992, + "learning_rate": 5.746897607570384e-06, + "loss": 0.0235, + "step": 85360 + }, + { + "epoch": 16.57356439007815, + "grad_norm": 0.16475790739059448, + "learning_rate": 5.7392035708472626e-06, + "loss": 0.0243, + "step": 85370 + }, + { + "epoch": 16.57550604339595, + "grad_norm": 0.2484104335308075, + "learning_rate": 5.731514374418278e-06, + "loss": 0.0215, + "step": 85380 + }, + { + "epoch": 16.57744769671375, + "grad_norm": 0.21288715302944183, + "learning_rate": 5.7238300191243325e-06, + "loss": 0.0214, + "step": 85390 + }, + { + "epoch": 16.57938935003155, + "grad_norm": 0.10013901442289352, + "learning_rate": 5.716150505805762e-06, + "loss": 0.0209, + "step": 85400 + }, + { + "epoch": 16.58133100334935, + "grad_norm": 0.12342092394828796, + "learning_rate": 5.708475835302407e-06, + "loss": 0.024, + "step": 85410 + }, + { + "epoch": 16.583272656667152, + "grad_norm": 0.11526236683130264, + "learning_rate": 5.70080600845353e-06, + "loss": 0.0222, + "step": 85420 + }, + { + "epoch": 16.585214309984952, + "grad_norm": 0.18566590547561646, + "learning_rate": 5.693141026097909e-06, + "loss": 0.0216, + "step": 85430 + }, + { + "epoch": 16.587155963302752, + "grad_norm": 0.1686856895685196, + "learning_rate": 5.6854808890737745e-06, + "loss": 0.0197, + "step": 85440 + }, + { + "epoch": 16.589097616620553, + "grad_norm": 0.26060551404953003, + "learning_rate": 5.6778255982188225e-06, + "loss": 0.0232, + "step": 85450 + }, + { + "epoch": 16.591039269938353, + "grad_norm": 0.1269253045320511, + "learning_rate": 5.670175154370217e-06, + "loss": 0.0256, + "step": 85460 + }, + { + "epoch": 16.592980923256153, + "grad_norm": 0.08082862943410873, + "learning_rate": 5.662529558364616e-06, + "loss": 0.0179, + "step": 85470 + }, + { + "epoch": 16.594922576573953, + "grad_norm": 0.12392407655715942, + "learning_rate": 5.654888811038128e-06, + "loss": 0.0247, + "step": 85480 + }, + { + "epoch": 16.596864229891754, + "grad_norm": 0.16165584325790405, + "learning_rate": 5.647252913226314e-06, + "loss": 0.0187, + "step": 85490 + }, + { + "epoch": 16.598805883209554, + "grad_norm": 0.2726539373397827, + "learning_rate": 5.639621865764255e-06, + "loss": 0.0237, + "step": 85500 + }, + { + "epoch": 16.600747536527354, + "grad_norm": 0.21943311393260956, + "learning_rate": 5.63199566948644e-06, + "loss": 0.019, + "step": 85510 + }, + { + "epoch": 16.602689189845155, + "grad_norm": 0.15807561576366425, + "learning_rate": 5.6243743252268835e-06, + "loss": 0.0233, + "step": 85520 + }, + { + "epoch": 16.604630843162955, + "grad_norm": 0.1912565976381302, + "learning_rate": 5.616757833819009e-06, + "loss": 0.0212, + "step": 85530 + }, + { + "epoch": 16.606572496480755, + "grad_norm": 0.14811214804649353, + "learning_rate": 5.609146196095794e-06, + "loss": 0.0167, + "step": 85540 + }, + { + "epoch": 16.608514149798552, + "grad_norm": 0.13066737353801727, + "learning_rate": 5.6015394128896005e-06, + "loss": 0.02, + "step": 85550 + }, + { + "epoch": 16.610455803116352, + "grad_norm": 0.11976262181997299, + "learning_rate": 5.593937485032308e-06, + "loss": 0.0162, + "step": 85560 + }, + { + "epoch": 16.612397456434152, + "grad_norm": 0.21560367941856384, + "learning_rate": 5.586340413355246e-06, + "loss": 0.0221, + "step": 85570 + }, + { + "epoch": 16.614339109751953, + "grad_norm": 0.25863635540008545, + "learning_rate": 5.578748198689227e-06, + "loss": 0.0213, + "step": 85580 + }, + { + "epoch": 16.616280763069753, + "grad_norm": 0.1328052282333374, + "learning_rate": 5.571160841864504e-06, + "loss": 0.0186, + "step": 85590 + }, + { + "epoch": 16.618222416387553, + "grad_norm": 0.2567328214645386, + "learning_rate": 5.563578343710846e-06, + "loss": 0.027, + "step": 85600 + }, + { + "epoch": 16.620164069705353, + "grad_norm": 0.1615471988916397, + "learning_rate": 5.556000705057446e-06, + "loss": 0.0208, + "step": 85610 + }, + { + "epoch": 16.622105723023154, + "grad_norm": 0.24126465618610382, + "learning_rate": 5.548427926732996e-06, + "loss": 0.0249, + "step": 85620 + }, + { + "epoch": 16.624047376340954, + "grad_norm": 0.1313343197107315, + "learning_rate": 5.540860009565629e-06, + "loss": 0.0192, + "step": 85630 + }, + { + "epoch": 16.625989029658754, + "grad_norm": 0.16127553582191467, + "learning_rate": 5.533296954382966e-06, + "loss": 0.0246, + "step": 85640 + }, + { + "epoch": 16.627930682976555, + "grad_norm": 0.1749519258737564, + "learning_rate": 5.525738762012095e-06, + "loss": 0.028, + "step": 85650 + }, + { + "epoch": 16.629872336294355, + "grad_norm": 0.20860639214515686, + "learning_rate": 5.518185433279577e-06, + "loss": 0.0207, + "step": 85660 + }, + { + "epoch": 16.631813989612155, + "grad_norm": 0.22103312611579895, + "learning_rate": 5.510636969011407e-06, + "loss": 0.0221, + "step": 85670 + }, + { + "epoch": 16.633755642929955, + "grad_norm": 0.27785155177116394, + "learning_rate": 5.503093370033096e-06, + "loss": 0.0232, + "step": 85680 + }, + { + "epoch": 16.635697296247756, + "grad_norm": 0.1497649848461151, + "learning_rate": 5.495554637169603e-06, + "loss": 0.0196, + "step": 85690 + }, + { + "epoch": 16.637638949565556, + "grad_norm": 0.20886990427970886, + "learning_rate": 5.488020771245328e-06, + "loss": 0.0184, + "step": 85700 + }, + { + "epoch": 16.639580602883356, + "grad_norm": 0.19626681506633759, + "learning_rate": 5.480491773084195e-06, + "loss": 0.0214, + "step": 85710 + }, + { + "epoch": 16.641522256201156, + "grad_norm": 0.2115306407213211, + "learning_rate": 5.472967643509525e-06, + "loss": 0.0261, + "step": 85720 + }, + { + "epoch": 16.643463909518957, + "grad_norm": 0.17481473088264465, + "learning_rate": 5.46544838334419e-06, + "loss": 0.0216, + "step": 85730 + }, + { + "epoch": 16.645405562836757, + "grad_norm": 0.259300172328949, + "learning_rate": 5.4579339934104525e-06, + "loss": 0.0238, + "step": 85740 + }, + { + "epoch": 16.647347216154557, + "grad_norm": 0.13839764893054962, + "learning_rate": 5.450424474530091e-06, + "loss": 0.0199, + "step": 85750 + }, + { + "epoch": 16.649288869472358, + "grad_norm": 0.12203522771596909, + "learning_rate": 5.442919827524323e-06, + "loss": 0.0258, + "step": 85760 + }, + { + "epoch": 16.651230522790154, + "grad_norm": 0.1395544409751892, + "learning_rate": 5.435420053213863e-06, + "loss": 0.0213, + "step": 85770 + }, + { + "epoch": 16.653172176107955, + "grad_norm": 0.23431475460529327, + "learning_rate": 5.427925152418845e-06, + "loss": 0.0244, + "step": 85780 + }, + { + "epoch": 16.655113829425755, + "grad_norm": 0.18731626868247986, + "learning_rate": 5.420435125958934e-06, + "loss": 0.0202, + "step": 85790 + }, + { + "epoch": 16.657055482743555, + "grad_norm": 0.31637337803840637, + "learning_rate": 5.412949974653209e-06, + "loss": 0.0194, + "step": 85800 + }, + { + "epoch": 16.658997136061355, + "grad_norm": 0.22411610186100006, + "learning_rate": 5.405469699320248e-06, + "loss": 0.0238, + "step": 85810 + }, + { + "epoch": 16.660938789379156, + "grad_norm": 0.1389450579881668, + "learning_rate": 5.397994300778064e-06, + "loss": 0.0216, + "step": 85820 + }, + { + "epoch": 16.662880442696956, + "grad_norm": 0.1724151223897934, + "learning_rate": 5.39052377984417e-06, + "loss": 0.0185, + "step": 85830 + }, + { + "epoch": 16.664822096014756, + "grad_norm": 0.21518802642822266, + "learning_rate": 5.383058137335523e-06, + "loss": 0.0229, + "step": 85840 + }, + { + "epoch": 16.666763749332556, + "grad_norm": 0.14021773636341095, + "learning_rate": 5.375597374068553e-06, + "loss": 0.0251, + "step": 85850 + }, + { + "epoch": 16.668705402650357, + "grad_norm": 0.2092789262533188, + "learning_rate": 5.368141490859163e-06, + "loss": 0.0238, + "step": 85860 + }, + { + "epoch": 16.670647055968157, + "grad_norm": 0.11984389275312424, + "learning_rate": 5.360690488522729e-06, + "loss": 0.0165, + "step": 85870 + }, + { + "epoch": 16.672588709285957, + "grad_norm": 0.26550260186195374, + "learning_rate": 5.353244367874055e-06, + "loss": 0.0255, + "step": 85880 + }, + { + "epoch": 16.674530362603758, + "grad_norm": 0.1815270334482193, + "learning_rate": 5.345803129727456e-06, + "loss": 0.0269, + "step": 85890 + }, + { + "epoch": 16.676472015921558, + "grad_norm": 0.12723873555660248, + "learning_rate": 5.338366774896697e-06, + "loss": 0.0181, + "step": 85900 + }, + { + "epoch": 16.678413669239358, + "grad_norm": 0.1341090053319931, + "learning_rate": 5.330935304194995e-06, + "loss": 0.0249, + "step": 85910 + }, + { + "epoch": 16.68035532255716, + "grad_norm": 0.13192223012447357, + "learning_rate": 5.323508718435044e-06, + "loss": 0.0197, + "step": 85920 + }, + { + "epoch": 16.68229697587496, + "grad_norm": 0.21896179020404816, + "learning_rate": 5.316087018429011e-06, + "loss": 0.0199, + "step": 85930 + }, + { + "epoch": 16.68423862919276, + "grad_norm": 0.19515152275562286, + "learning_rate": 5.30867020498853e-06, + "loss": 0.0254, + "step": 85940 + }, + { + "epoch": 16.68618028251056, + "grad_norm": 0.5046396255493164, + "learning_rate": 5.30125827892467e-06, + "loss": 0.021, + "step": 85950 + }, + { + "epoch": 16.68812193582836, + "grad_norm": 0.11863772571086884, + "learning_rate": 5.293851241048015e-06, + "loss": 0.0177, + "step": 85960 + }, + { + "epoch": 16.690063589146156, + "grad_norm": 0.21062448620796204, + "learning_rate": 5.286449092168555e-06, + "loss": 0.0204, + "step": 85970 + }, + { + "epoch": 16.692005242463956, + "grad_norm": 0.23211218416690826, + "learning_rate": 5.2790518330958105e-06, + "loss": 0.0189, + "step": 85980 + }, + { + "epoch": 16.693946895781757, + "grad_norm": 0.09132777899503708, + "learning_rate": 5.271659464638717e-06, + "loss": 0.0184, + "step": 85990 + }, + { + "epoch": 16.695888549099557, + "grad_norm": 0.21881882846355438, + "learning_rate": 5.264271987605701e-06, + "loss": 0.0186, + "step": 86000 + }, + { + "epoch": 16.697830202417357, + "grad_norm": 0.2063170075416565, + "learning_rate": 5.256889402804632e-06, + "loss": 0.0216, + "step": 86010 + }, + { + "epoch": 16.699771855735158, + "grad_norm": 0.1224319338798523, + "learning_rate": 5.249511711042881e-06, + "loss": 0.0161, + "step": 86020 + }, + { + "epoch": 16.701713509052958, + "grad_norm": 0.20958930253982544, + "learning_rate": 5.242138913127237e-06, + "loss": 0.0209, + "step": 86030 + }, + { + "epoch": 16.703655162370758, + "grad_norm": 0.22578118741512299, + "learning_rate": 5.234771009863987e-06, + "loss": 0.0261, + "step": 86040 + }, + { + "epoch": 16.70559681568856, + "grad_norm": 0.19687488675117493, + "learning_rate": 5.227408002058881e-06, + "loss": 0.0201, + "step": 86050 + }, + { + "epoch": 16.70753846900636, + "grad_norm": 0.19825148582458496, + "learning_rate": 5.220049890517126e-06, + "loss": 0.0272, + "step": 86060 + }, + { + "epoch": 16.70948012232416, + "grad_norm": 0.1884675920009613, + "learning_rate": 5.212696676043377e-06, + "loss": 0.0194, + "step": 86070 + }, + { + "epoch": 16.71142177564196, + "grad_norm": 0.26335909962654114, + "learning_rate": 5.205348359441798e-06, + "loss": 0.0235, + "step": 86080 + }, + { + "epoch": 16.71336342895976, + "grad_norm": 0.22394010424613953, + "learning_rate": 5.198004941515966e-06, + "loss": 0.0199, + "step": 86090 + }, + { + "epoch": 16.71530508227756, + "grad_norm": 0.10779982060194016, + "learning_rate": 5.190666423068952e-06, + "loss": 0.0228, + "step": 86100 + }, + { + "epoch": 16.71724673559536, + "grad_norm": 0.11939582973718643, + "learning_rate": 5.183332804903284e-06, + "loss": 0.0199, + "step": 86110 + }, + { + "epoch": 16.71918838891316, + "grad_norm": 0.17952856421470642, + "learning_rate": 5.176004087820979e-06, + "loss": 0.0204, + "step": 86120 + }, + { + "epoch": 16.72113004223096, + "grad_norm": 0.15271100401878357, + "learning_rate": 5.168680272623455e-06, + "loss": 0.0232, + "step": 86130 + }, + { + "epoch": 16.72307169554876, + "grad_norm": 0.1669381558895111, + "learning_rate": 5.161361360111661e-06, + "loss": 0.0302, + "step": 86140 + }, + { + "epoch": 16.72501334886656, + "grad_norm": 0.13935033977031708, + "learning_rate": 5.154047351085984e-06, + "loss": 0.0212, + "step": 86150 + }, + { + "epoch": 16.72695500218436, + "grad_norm": 0.15098851919174194, + "learning_rate": 5.1467382463462385e-06, + "loss": 0.021, + "step": 86160 + }, + { + "epoch": 16.72889665550216, + "grad_norm": 0.1431257426738739, + "learning_rate": 5.139434046691788e-06, + "loss": 0.0177, + "step": 86170 + }, + { + "epoch": 16.730838308819962, + "grad_norm": 0.18544535338878632, + "learning_rate": 5.13213475292137e-06, + "loss": 0.0233, + "step": 86180 + }, + { + "epoch": 16.73277996213776, + "grad_norm": 0.12620528042316437, + "learning_rate": 5.12484036583325e-06, + "loss": 0.0211, + "step": 86190 + }, + { + "epoch": 16.73472161545556, + "grad_norm": 0.23640501499176025, + "learning_rate": 5.117550886225109e-06, + "loss": 0.0204, + "step": 86200 + }, + { + "epoch": 16.73666326877336, + "grad_norm": 0.26855704188346863, + "learning_rate": 5.11026631489413e-06, + "loss": 0.0208, + "step": 86210 + }, + { + "epoch": 16.73860492209116, + "grad_norm": 0.22698713839054108, + "learning_rate": 5.102986652636926e-06, + "loss": 0.0229, + "step": 86220 + }, + { + "epoch": 16.74054657540896, + "grad_norm": 0.17349351942539215, + "learning_rate": 5.095711900249606e-06, + "loss": 0.0194, + "step": 86230 + }, + { + "epoch": 16.74248822872676, + "grad_norm": 0.27077168226242065, + "learning_rate": 5.088442058527715e-06, + "loss": 0.0213, + "step": 86240 + }, + { + "epoch": 16.74442988204456, + "grad_norm": 0.184342160820961, + "learning_rate": 5.08117712826629e-06, + "loss": 0.0205, + "step": 86250 + }, + { + "epoch": 16.74637153536236, + "grad_norm": 0.15133966505527496, + "learning_rate": 5.073917110259785e-06, + "loss": 0.0163, + "step": 86260 + }, + { + "epoch": 16.74831318868016, + "grad_norm": 0.13834455609321594, + "learning_rate": 5.066662005302175e-06, + "loss": 0.0292, + "step": 86270 + }, + { + "epoch": 16.75025484199796, + "grad_norm": 0.10970335453748703, + "learning_rate": 5.059411814186838e-06, + "loss": 0.0206, + "step": 86280 + }, + { + "epoch": 16.75219649531576, + "grad_norm": 0.1376248151063919, + "learning_rate": 5.05216653770666e-06, + "loss": 0.0201, + "step": 86290 + }, + { + "epoch": 16.75413814863356, + "grad_norm": 0.18753564357757568, + "learning_rate": 5.044926176653969e-06, + "loss": 0.0196, + "step": 86300 + }, + { + "epoch": 16.756079801951362, + "grad_norm": 0.14027349650859833, + "learning_rate": 5.037690731820571e-06, + "loss": 0.026, + "step": 86310 + }, + { + "epoch": 16.758021455269162, + "grad_norm": 0.21055291593074799, + "learning_rate": 5.030460203997711e-06, + "loss": 0.0217, + "step": 86320 + }, + { + "epoch": 16.759963108586962, + "grad_norm": 0.27554214000701904, + "learning_rate": 5.0232345939761174e-06, + "loss": 0.0261, + "step": 86330 + }, + { + "epoch": 16.761904761904763, + "grad_norm": 0.1564033180475235, + "learning_rate": 5.016013902545957e-06, + "loss": 0.0241, + "step": 86340 + }, + { + "epoch": 16.763846415222563, + "grad_norm": 0.15463317930698395, + "learning_rate": 5.008798130496889e-06, + "loss": 0.0232, + "step": 86350 + }, + { + "epoch": 16.765788068540363, + "grad_norm": 0.22347994148731232, + "learning_rate": 5.001587278618008e-06, + "loss": 0.0202, + "step": 86360 + }, + { + "epoch": 16.767729721858164, + "grad_norm": 0.13551905751228333, + "learning_rate": 4.994381347697891e-06, + "loss": 0.0203, + "step": 86370 + }, + { + "epoch": 16.769671375175964, + "grad_norm": 0.13109157979488373, + "learning_rate": 4.987180338524572e-06, + "loss": 0.0176, + "step": 86380 + }, + { + "epoch": 16.771613028493764, + "grad_norm": 0.17574085295200348, + "learning_rate": 4.9799842518855275e-06, + "loss": 0.0192, + "step": 86390 + }, + { + "epoch": 16.77355468181156, + "grad_norm": 0.23757609724998474, + "learning_rate": 4.9727930885677285e-06, + "loss": 0.0198, + "step": 86400 + }, + { + "epoch": 16.77549633512936, + "grad_norm": 0.19861885905265808, + "learning_rate": 4.965606849357573e-06, + "loss": 0.0274, + "step": 86410 + }, + { + "epoch": 16.77743798844716, + "grad_norm": 0.3280782103538513, + "learning_rate": 4.9584255350409445e-06, + "loss": 0.0185, + "step": 86420 + }, + { + "epoch": 16.77937964176496, + "grad_norm": 0.14314399659633636, + "learning_rate": 4.951249146403175e-06, + "loss": 0.0189, + "step": 86430 + }, + { + "epoch": 16.781321295082762, + "grad_norm": 0.20117633044719696, + "learning_rate": 4.9440776842290815e-06, + "loss": 0.0252, + "step": 86440 + }, + { + "epoch": 16.783262948400562, + "grad_norm": 0.17958533763885498, + "learning_rate": 4.936911149302903e-06, + "loss": 0.0255, + "step": 86450 + }, + { + "epoch": 16.785204601718362, + "grad_norm": 0.1839921921491623, + "learning_rate": 4.929749542408379e-06, + "loss": 0.0243, + "step": 86460 + }, + { + "epoch": 16.787146255036163, + "grad_norm": 0.3568752110004425, + "learning_rate": 4.922592864328673e-06, + "loss": 0.0281, + "step": 86470 + }, + { + "epoch": 16.789087908353963, + "grad_norm": 0.06734620779752731, + "learning_rate": 4.915441115846436e-06, + "loss": 0.0157, + "step": 86480 + }, + { + "epoch": 16.791029561671763, + "grad_norm": 0.20284371078014374, + "learning_rate": 4.908294297743782e-06, + "loss": 0.0245, + "step": 86490 + }, + { + "epoch": 16.792971214989564, + "grad_norm": 0.18649186193943024, + "learning_rate": 4.901152410802273e-06, + "loss": 0.0178, + "step": 86500 + }, + { + "epoch": 16.794912868307364, + "grad_norm": 0.13134413957595825, + "learning_rate": 4.894015455802919e-06, + "loss": 0.0215, + "step": 86510 + }, + { + "epoch": 16.796854521625164, + "grad_norm": 0.09975214302539825, + "learning_rate": 4.8868834335262346e-06, + "loss": 0.0245, + "step": 86520 + }, + { + "epoch": 16.798796174942964, + "grad_norm": 0.24750006198883057, + "learning_rate": 4.87975634475214e-06, + "loss": 0.0216, + "step": 86530 + }, + { + "epoch": 16.800737828260765, + "grad_norm": 0.11289766430854797, + "learning_rate": 4.8726341902600614e-06, + "loss": 0.0181, + "step": 86540 + }, + { + "epoch": 16.802679481578565, + "grad_norm": 0.17430956661701202, + "learning_rate": 4.865516970828843e-06, + "loss": 0.0256, + "step": 86550 + }, + { + "epoch": 16.804621134896365, + "grad_norm": 0.10757862031459808, + "learning_rate": 4.85840468723685e-06, + "loss": 0.0157, + "step": 86560 + }, + { + "epoch": 16.806562788214165, + "grad_norm": 0.1203526109457016, + "learning_rate": 4.85129734026184e-06, + "loss": 0.0222, + "step": 86570 + }, + { + "epoch": 16.808504441531966, + "grad_norm": 0.1411086618900299, + "learning_rate": 4.844194930681073e-06, + "loss": 0.0201, + "step": 86580 + }, + { + "epoch": 16.810446094849766, + "grad_norm": 0.2158084511756897, + "learning_rate": 4.837097459271267e-06, + "loss": 0.0223, + "step": 86590 + }, + { + "epoch": 16.812387748167566, + "grad_norm": 0.22794882953166962, + "learning_rate": 4.830004926808573e-06, + "loss": 0.0192, + "step": 86600 + }, + { + "epoch": 16.814329401485367, + "grad_norm": 0.21852412819862366, + "learning_rate": 4.822917334068627e-06, + "loss": 0.0207, + "step": 86610 + }, + { + "epoch": 16.816271054803163, + "grad_norm": 0.4781781733036041, + "learning_rate": 4.81583468182652e-06, + "loss": 0.0204, + "step": 86620 + }, + { + "epoch": 16.818212708120964, + "grad_norm": 0.241501122713089, + "learning_rate": 4.808756970856809e-06, + "loss": 0.0206, + "step": 86630 + }, + { + "epoch": 16.820154361438764, + "grad_norm": 0.2365991473197937, + "learning_rate": 4.801684201933476e-06, + "loss": 0.0234, + "step": 86640 + }, + { + "epoch": 16.822096014756564, + "grad_norm": 0.11460074782371521, + "learning_rate": 4.794616375830019e-06, + "loss": 0.0153, + "step": 86650 + }, + { + "epoch": 16.824037668074364, + "grad_norm": 0.20253732800483704, + "learning_rate": 4.787553493319341e-06, + "loss": 0.0183, + "step": 86660 + }, + { + "epoch": 16.825979321392165, + "grad_norm": 0.20822161436080933, + "learning_rate": 4.780495555173836e-06, + "loss": 0.0156, + "step": 86670 + }, + { + "epoch": 16.827920974709965, + "grad_norm": 0.27253392338752747, + "learning_rate": 4.7734425621653475e-06, + "loss": 0.0226, + "step": 86680 + }, + { + "epoch": 16.829862628027765, + "grad_norm": 0.14259739220142365, + "learning_rate": 4.766394515065198e-06, + "loss": 0.0209, + "step": 86690 + }, + { + "epoch": 16.831804281345565, + "grad_norm": 0.1809067577123642, + "learning_rate": 4.75935141464412e-06, + "loss": 0.0204, + "step": 86700 + }, + { + "epoch": 16.833745934663366, + "grad_norm": 0.16990973055362701, + "learning_rate": 4.75231326167237e-06, + "loss": 0.0249, + "step": 86710 + }, + { + "epoch": 16.835687587981166, + "grad_norm": 0.2598649263381958, + "learning_rate": 4.745280056919599e-06, + "loss": 0.0247, + "step": 86720 + }, + { + "epoch": 16.837629241298966, + "grad_norm": 0.11582615226507187, + "learning_rate": 4.738251801154975e-06, + "loss": 0.0226, + "step": 86730 + }, + { + "epoch": 16.839570894616767, + "grad_norm": 0.21130020916461945, + "learning_rate": 4.731228495147061e-06, + "loss": 0.0239, + "step": 86740 + }, + { + "epoch": 16.841512547934567, + "grad_norm": 0.2633974850177765, + "learning_rate": 4.7242101396639595e-06, + "loss": 0.0201, + "step": 86750 + }, + { + "epoch": 16.843454201252367, + "grad_norm": 0.23545826971530914, + "learning_rate": 4.717196735473151e-06, + "loss": 0.0193, + "step": 86760 + }, + { + "epoch": 16.845395854570167, + "grad_norm": 0.15583810210227966, + "learning_rate": 4.710188283341643e-06, + "loss": 0.0161, + "step": 86770 + }, + { + "epoch": 16.847337507887968, + "grad_norm": 0.21191050112247467, + "learning_rate": 4.703184784035835e-06, + "loss": 0.0216, + "step": 86780 + }, + { + "epoch": 16.849279161205768, + "grad_norm": 0.15318293869495392, + "learning_rate": 4.696186238321642e-06, + "loss": 0.0249, + "step": 86790 + }, + { + "epoch": 16.85122081452357, + "grad_norm": 0.18150822818279266, + "learning_rate": 4.689192646964408e-06, + "loss": 0.0239, + "step": 86800 + }, + { + "epoch": 16.85316246784137, + "grad_norm": 0.09377574920654297, + "learning_rate": 4.682204010728952e-06, + "loss": 0.0226, + "step": 86810 + }, + { + "epoch": 16.855104121159165, + "grad_norm": 0.11083613336086273, + "learning_rate": 4.67522033037952e-06, + "loss": 0.0221, + "step": 86820 + }, + { + "epoch": 16.857045774476966, + "grad_norm": 0.16862766444683075, + "learning_rate": 4.668241606679846e-06, + "loss": 0.0195, + "step": 86830 + }, + { + "epoch": 16.858987427794766, + "grad_norm": 0.14313355088233948, + "learning_rate": 4.66126784039313e-06, + "loss": 0.0158, + "step": 86840 + }, + { + "epoch": 16.860929081112566, + "grad_norm": 0.17655588686466217, + "learning_rate": 4.6542990322819836e-06, + "loss": 0.0247, + "step": 86850 + }, + { + "epoch": 16.862870734430366, + "grad_norm": 0.14149005711078644, + "learning_rate": 4.64733518310852e-06, + "loss": 0.0188, + "step": 86860 + }, + { + "epoch": 16.864812387748167, + "grad_norm": 0.16480623185634613, + "learning_rate": 4.640376293634291e-06, + "loss": 0.0227, + "step": 86870 + }, + { + "epoch": 16.866754041065967, + "grad_norm": 0.19907909631729126, + "learning_rate": 4.633422364620327e-06, + "loss": 0.0285, + "step": 86880 + }, + { + "epoch": 16.868695694383767, + "grad_norm": 0.22708973288536072, + "learning_rate": 4.6264733968270745e-06, + "loss": 0.0218, + "step": 86890 + }, + { + "epoch": 16.870637347701567, + "grad_norm": 0.15210512280464172, + "learning_rate": 4.619529391014482e-06, + "loss": 0.0219, + "step": 86900 + }, + { + "epoch": 16.872579001019368, + "grad_norm": 0.06583259254693985, + "learning_rate": 4.61259034794192e-06, + "loss": 0.0229, + "step": 86910 + }, + { + "epoch": 16.874520654337168, + "grad_norm": 0.1271398961544037, + "learning_rate": 4.605656268368247e-06, + "loss": 0.0205, + "step": 86920 + }, + { + "epoch": 16.87646230765497, + "grad_norm": 0.29435479640960693, + "learning_rate": 4.598727153051735e-06, + "loss": 0.0196, + "step": 86930 + }, + { + "epoch": 16.87840396097277, + "grad_norm": 0.29002514481544495, + "learning_rate": 4.591803002750178e-06, + "loss": 0.0258, + "step": 86940 + }, + { + "epoch": 16.88034561429057, + "grad_norm": 0.20502141118049622, + "learning_rate": 4.58488381822077e-06, + "loss": 0.0219, + "step": 86950 + }, + { + "epoch": 16.88228726760837, + "grad_norm": 0.17176279425621033, + "learning_rate": 4.5779696002201946e-06, + "loss": 0.0233, + "step": 86960 + }, + { + "epoch": 16.88422892092617, + "grad_norm": 0.18701927363872528, + "learning_rate": 4.571060349504563e-06, + "loss": 0.0215, + "step": 86970 + }, + { + "epoch": 16.88617057424397, + "grad_norm": 0.274078369140625, + "learning_rate": 4.564156066829478e-06, + "loss": 0.0211, + "step": 86980 + }, + { + "epoch": 16.88811222756177, + "grad_norm": 0.22014346718788147, + "learning_rate": 4.557256752949957e-06, + "loss": 0.0269, + "step": 86990 + }, + { + "epoch": 16.89005388087957, + "grad_norm": 0.14935672283172607, + "learning_rate": 4.550362408620534e-06, + "loss": 0.0226, + "step": 87000 + }, + { + "epoch": 16.89199553419737, + "grad_norm": 0.16455966234207153, + "learning_rate": 4.543473034595136e-06, + "loss": 0.018, + "step": 87010 + }, + { + "epoch": 16.89393718751517, + "grad_norm": 0.27929699420928955, + "learning_rate": 4.536588631627192e-06, + "loss": 0.0253, + "step": 87020 + }, + { + "epoch": 16.89587884083297, + "grad_norm": 0.11995965987443924, + "learning_rate": 4.5297092004695565e-06, + "loss": 0.021, + "step": 87030 + }, + { + "epoch": 16.897820494150768, + "grad_norm": 0.18159672617912292, + "learning_rate": 4.522834741874555e-06, + "loss": 0.0211, + "step": 87040 + }, + { + "epoch": 16.899762147468568, + "grad_norm": 0.24479736387729645, + "learning_rate": 4.515965256593974e-06, + "loss": 0.021, + "step": 87050 + }, + { + "epoch": 16.90170380078637, + "grad_norm": 0.11566213518381119, + "learning_rate": 4.509100745379058e-06, + "loss": 0.0198, + "step": 87060 + }, + { + "epoch": 16.90364545410417, + "grad_norm": 0.1697847843170166, + "learning_rate": 4.502241208980479e-06, + "loss": 0.0197, + "step": 87070 + }, + { + "epoch": 16.90558710742197, + "grad_norm": 0.1488920897245407, + "learning_rate": 4.495386648148398e-06, + "loss": 0.02, + "step": 87080 + }, + { + "epoch": 16.90752876073977, + "grad_norm": 0.1657141149044037, + "learning_rate": 4.488537063632431e-06, + "loss": 0.0223, + "step": 87090 + }, + { + "epoch": 16.90947041405757, + "grad_norm": 0.22819358110427856, + "learning_rate": 4.481692456181608e-06, + "loss": 0.0222, + "step": 87100 + }, + { + "epoch": 16.91141206737537, + "grad_norm": 0.1832670420408249, + "learning_rate": 4.474852826544479e-06, + "loss": 0.0259, + "step": 87110 + }, + { + "epoch": 16.91335372069317, + "grad_norm": 0.22815647721290588, + "learning_rate": 4.468018175468974e-06, + "loss": 0.0237, + "step": 87120 + }, + { + "epoch": 16.91529537401097, + "grad_norm": 0.18389011919498444, + "learning_rate": 4.461188503702568e-06, + "loss": 0.0198, + "step": 87130 + }, + { + "epoch": 16.91723702732877, + "grad_norm": 0.18530556559562683, + "learning_rate": 4.454363811992107e-06, + "loss": 0.0207, + "step": 87140 + }, + { + "epoch": 16.91917868064657, + "grad_norm": 0.1185055747628212, + "learning_rate": 4.447544101083956e-06, + "loss": 0.0165, + "step": 87150 + }, + { + "epoch": 16.92112033396437, + "grad_norm": 0.12273137271404266, + "learning_rate": 4.440729371723884e-06, + "loss": 0.0178, + "step": 87160 + }, + { + "epoch": 16.92306198728217, + "grad_norm": 0.13803821802139282, + "learning_rate": 4.433919624657162e-06, + "loss": 0.0192, + "step": 87170 + }, + { + "epoch": 16.92500364059997, + "grad_norm": 0.24327757954597473, + "learning_rate": 4.427114860628462e-06, + "loss": 0.0266, + "step": 87180 + }, + { + "epoch": 16.926945293917772, + "grad_norm": 0.2523009777069092, + "learning_rate": 4.420315080381982e-06, + "loss": 0.0186, + "step": 87190 + }, + { + "epoch": 16.928886947235572, + "grad_norm": 0.17174597084522247, + "learning_rate": 4.4135202846613045e-06, + "loss": 0.0218, + "step": 87200 + }, + { + "epoch": 16.930828600553372, + "grad_norm": 0.22540636360645294, + "learning_rate": 4.406730474209525e-06, + "loss": 0.0262, + "step": 87210 + }, + { + "epoch": 16.932770253871173, + "grad_norm": 0.17030371725559235, + "learning_rate": 4.399945649769138e-06, + "loss": 0.016, + "step": 87220 + }, + { + "epoch": 16.934711907188973, + "grad_norm": 0.16862204670906067, + "learning_rate": 4.393165812082151e-06, + "loss": 0.0352, + "step": 87230 + }, + { + "epoch": 16.936653560506773, + "grad_norm": 0.11926623433828354, + "learning_rate": 4.386390961889969e-06, + "loss": 0.0197, + "step": 87240 + }, + { + "epoch": 16.93859521382457, + "grad_norm": 0.15553295612335205, + "learning_rate": 4.3796210999334895e-06, + "loss": 0.0197, + "step": 87250 + }, + { + "epoch": 16.94053686714237, + "grad_norm": 0.12491293251514435, + "learning_rate": 4.372856226953065e-06, + "loss": 0.0204, + "step": 87260 + }, + { + "epoch": 16.94247852046017, + "grad_norm": 0.09598934650421143, + "learning_rate": 4.366096343688486e-06, + "loss": 0.0231, + "step": 87270 + }, + { + "epoch": 16.94442017377797, + "grad_norm": 0.4094280004501343, + "learning_rate": 4.359341450878995e-06, + "loss": 0.0249, + "step": 87280 + }, + { + "epoch": 16.94636182709577, + "grad_norm": 0.2043982595205307, + "learning_rate": 4.352591549263302e-06, + "loss": 0.0211, + "step": 87290 + }, + { + "epoch": 16.94830348041357, + "grad_norm": 0.15599901974201202, + "learning_rate": 4.345846639579582e-06, + "loss": 0.0206, + "step": 87300 + }, + { + "epoch": 16.95024513373137, + "grad_norm": 0.21944497525691986, + "learning_rate": 4.339106722565417e-06, + "loss": 0.0225, + "step": 87310 + }, + { + "epoch": 16.952186787049172, + "grad_norm": 0.12806323170661926, + "learning_rate": 4.332371798957896e-06, + "loss": 0.0228, + "step": 87320 + }, + { + "epoch": 16.954128440366972, + "grad_norm": 0.1821986883878708, + "learning_rate": 4.3256418694935295e-06, + "loss": 0.0211, + "step": 87330 + }, + { + "epoch": 16.956070093684772, + "grad_norm": 0.11917965114116669, + "learning_rate": 4.3189169349083125e-06, + "loss": 0.0182, + "step": 87340 + }, + { + "epoch": 16.958011747002573, + "grad_norm": 0.12581346929073334, + "learning_rate": 4.312196995937645e-06, + "loss": 0.0207, + "step": 87350 + }, + { + "epoch": 16.959953400320373, + "grad_norm": 0.16023360192775726, + "learning_rate": 4.305482053316434e-06, + "loss": 0.0261, + "step": 87360 + }, + { + "epoch": 16.961895053638173, + "grad_norm": 0.14504115283489227, + "learning_rate": 4.2987721077789925e-06, + "loss": 0.0191, + "step": 87370 + }, + { + "epoch": 16.963836706955973, + "grad_norm": 0.1868792474269867, + "learning_rate": 4.2920671600591325e-06, + "loss": 0.0253, + "step": 87380 + }, + { + "epoch": 16.965778360273774, + "grad_norm": 0.0894448533654213, + "learning_rate": 4.285367210890084e-06, + "loss": 0.0284, + "step": 87390 + }, + { + "epoch": 16.967720013591574, + "grad_norm": 0.10047277808189392, + "learning_rate": 4.278672261004551e-06, + "loss": 0.0156, + "step": 87400 + }, + { + "epoch": 16.969661666909374, + "grad_norm": 0.24072545766830444, + "learning_rate": 4.2719823111346704e-06, + "loss": 0.0236, + "step": 87410 + }, + { + "epoch": 16.971603320227175, + "grad_norm": 0.24563103914260864, + "learning_rate": 4.265297362012061e-06, + "loss": 0.0267, + "step": 87420 + }, + { + "epoch": 16.973544973544975, + "grad_norm": 0.1773175597190857, + "learning_rate": 4.258617414367766e-06, + "loss": 0.0249, + "step": 87430 + }, + { + "epoch": 16.975486626862775, + "grad_norm": 0.22163622081279755, + "learning_rate": 4.251942468932296e-06, + "loss": 0.0203, + "step": 87440 + }, + { + "epoch": 16.977428280180575, + "grad_norm": 0.23460809886455536, + "learning_rate": 4.2452725264356175e-06, + "loss": 0.0224, + "step": 87450 + }, + { + "epoch": 16.979369933498376, + "grad_norm": 0.11561181396245956, + "learning_rate": 4.2386075876071575e-06, + "loss": 0.023, + "step": 87460 + }, + { + "epoch": 16.981311586816172, + "grad_norm": 0.1514660269021988, + "learning_rate": 4.231947653175755e-06, + "loss": 0.0236, + "step": 87470 + }, + { + "epoch": 16.983253240133973, + "grad_norm": 0.32729899883270264, + "learning_rate": 4.225292723869761e-06, + "loss": 0.0239, + "step": 87480 + }, + { + "epoch": 16.985194893451773, + "grad_norm": 0.14266866445541382, + "learning_rate": 4.218642800416916e-06, + "loss": 0.0208, + "step": 87490 + }, + { + "epoch": 16.987136546769573, + "grad_norm": 0.18460644781589508, + "learning_rate": 4.211997883544472e-06, + "loss": 0.0218, + "step": 87500 + }, + { + "epoch": 16.989078200087373, + "grad_norm": 0.2296205312013626, + "learning_rate": 4.205357973979096e-06, + "loss": 0.0211, + "step": 87510 + }, + { + "epoch": 16.991019853405174, + "grad_norm": 0.1432405263185501, + "learning_rate": 4.198723072446925e-06, + "loss": 0.0231, + "step": 87520 + }, + { + "epoch": 16.992961506722974, + "grad_norm": 0.23946021497249603, + "learning_rate": 4.192093179673534e-06, + "loss": 0.0179, + "step": 87530 + }, + { + "epoch": 16.994903160040774, + "grad_norm": 0.09013189375400543, + "learning_rate": 4.185468296383959e-06, + "loss": 0.0215, + "step": 87540 + }, + { + "epoch": 16.996844813358575, + "grad_norm": 0.22529731690883636, + "learning_rate": 4.1788484233027005e-06, + "loss": 0.0269, + "step": 87550 + }, + { + "epoch": 16.998786466676375, + "grad_norm": 0.18090149760246277, + "learning_rate": 4.172233561153677e-06, + "loss": 0.0194, + "step": 87560 + }, + { + "epoch": 17.00058249599534, + "grad_norm": 0.20154114067554474, + "learning_rate": 4.165623710660288e-06, + "loss": 0.0208, + "step": 87570 + }, + { + "epoch": 17.00252414931314, + "grad_norm": 0.13897141814231873, + "learning_rate": 4.1590188725453835e-06, + "loss": 0.0266, + "step": 87580 + }, + { + "epoch": 17.00446580263094, + "grad_norm": 0.17362654209136963, + "learning_rate": 4.152419047531259e-06, + "loss": 0.0201, + "step": 87590 + }, + { + "epoch": 17.00640745594874, + "grad_norm": 0.12489502876996994, + "learning_rate": 4.145824236339651e-06, + "loss": 0.0212, + "step": 87600 + }, + { + "epoch": 17.00834910926654, + "grad_norm": 0.15460409224033356, + "learning_rate": 4.1392344396917695e-06, + "loss": 0.0227, + "step": 87610 + }, + { + "epoch": 17.01029076258434, + "grad_norm": 0.14537116885185242, + "learning_rate": 4.132649658308252e-06, + "loss": 0.0166, + "step": 87620 + }, + { + "epoch": 17.01223241590214, + "grad_norm": 0.2231191247701645, + "learning_rate": 4.126069892909213e-06, + "loss": 0.0214, + "step": 87630 + }, + { + "epoch": 17.01417406921994, + "grad_norm": 0.17932651937007904, + "learning_rate": 4.119495144214191e-06, + "loss": 0.0212, + "step": 87640 + }, + { + "epoch": 17.01611572253774, + "grad_norm": 0.2404509335756302, + "learning_rate": 4.112925412942214e-06, + "loss": 0.0234, + "step": 87650 + }, + { + "epoch": 17.01805737585554, + "grad_norm": 0.10963451862335205, + "learning_rate": 4.106360699811712e-06, + "loss": 0.0285, + "step": 87660 + }, + { + "epoch": 17.01999902917334, + "grad_norm": 0.24322007596492767, + "learning_rate": 4.099801005540616e-06, + "loss": 0.0179, + "step": 87670 + }, + { + "epoch": 17.02194068249114, + "grad_norm": 0.3983820974826813, + "learning_rate": 4.093246330846256e-06, + "loss": 0.0199, + "step": 87680 + }, + { + "epoch": 17.023882335808942, + "grad_norm": 0.19615578651428223, + "learning_rate": 4.086696676445467e-06, + "loss": 0.0231, + "step": 87690 + }, + { + "epoch": 17.025823989126742, + "grad_norm": 0.25152069330215454, + "learning_rate": 4.0801520430544914e-06, + "loss": 0.0241, + "step": 87700 + }, + { + "epoch": 17.027765642444542, + "grad_norm": 0.15339893102645874, + "learning_rate": 4.073612431389062e-06, + "loss": 0.0232, + "step": 87710 + }, + { + "epoch": 17.029707295762343, + "grad_norm": 0.1850423812866211, + "learning_rate": 4.067077842164319e-06, + "loss": 0.0242, + "step": 87720 + }, + { + "epoch": 17.031648949080143, + "grad_norm": 0.3455301821231842, + "learning_rate": 4.0605482760948945e-06, + "loss": 0.0274, + "step": 87730 + }, + { + "epoch": 17.033590602397943, + "grad_norm": 0.1766459196805954, + "learning_rate": 4.054023733894829e-06, + "loss": 0.0273, + "step": 87740 + }, + { + "epoch": 17.035532255715744, + "grad_norm": 0.18582777678966522, + "learning_rate": 4.047504216277653e-06, + "loss": 0.0254, + "step": 87750 + }, + { + "epoch": 17.03747390903354, + "grad_norm": 0.12994521856307983, + "learning_rate": 4.040989723956323e-06, + "loss": 0.0228, + "step": 87760 + }, + { + "epoch": 17.03941556235134, + "grad_norm": 0.22828176617622375, + "learning_rate": 4.03448025764327e-06, + "loss": 0.019, + "step": 87770 + }, + { + "epoch": 17.04135721566914, + "grad_norm": 0.15519237518310547, + "learning_rate": 4.027975818050339e-06, + "loss": 0.0235, + "step": 87780 + }, + { + "epoch": 17.04329886898694, + "grad_norm": 0.21625587344169617, + "learning_rate": 4.021476405888858e-06, + "loss": 0.0225, + "step": 87790 + }, + { + "epoch": 17.04524052230474, + "grad_norm": 0.21731437742710114, + "learning_rate": 4.01498202186959e-06, + "loss": 0.0199, + "step": 87800 + }, + { + "epoch": 17.04718217562254, + "grad_norm": 0.2201395332813263, + "learning_rate": 4.008492666702751e-06, + "loss": 0.0234, + "step": 87810 + }, + { + "epoch": 17.049123828940342, + "grad_norm": 0.16721171140670776, + "learning_rate": 4.002008341098001e-06, + "loss": 0.0213, + "step": 87820 + }, + { + "epoch": 17.051065482258142, + "grad_norm": 0.15406601130962372, + "learning_rate": 3.995529045764462e-06, + "loss": 0.0176, + "step": 87830 + }, + { + "epoch": 17.053007135575942, + "grad_norm": 0.11497609317302704, + "learning_rate": 3.989054781410712e-06, + "loss": 0.0191, + "step": 87840 + }, + { + "epoch": 17.054948788893743, + "grad_norm": 0.13929927349090576, + "learning_rate": 3.982585548744744e-06, + "loss": 0.0235, + "step": 87850 + }, + { + "epoch": 17.056890442211543, + "grad_norm": 0.09974364191293716, + "learning_rate": 3.9761213484740435e-06, + "loss": 0.0215, + "step": 87860 + }, + { + "epoch": 17.058832095529343, + "grad_norm": 0.13791070878505707, + "learning_rate": 3.9696621813055055e-06, + "loss": 0.0248, + "step": 87870 + }, + { + "epoch": 17.060773748847144, + "grad_norm": 0.1685357540845871, + "learning_rate": 3.9632080479455044e-06, + "loss": 0.0226, + "step": 87880 + }, + { + "epoch": 17.062715402164944, + "grad_norm": 0.20286312699317932, + "learning_rate": 3.956758949099859e-06, + "loss": 0.0273, + "step": 87890 + }, + { + "epoch": 17.064657055482744, + "grad_norm": 0.17129085958003998, + "learning_rate": 3.9503148854738385e-06, + "loss": 0.0207, + "step": 87900 + }, + { + "epoch": 17.066598708800544, + "grad_norm": 0.13675643503665924, + "learning_rate": 3.943875857772139e-06, + "loss": 0.0215, + "step": 87910 + }, + { + "epoch": 17.068540362118345, + "grad_norm": 0.13717719912528992, + "learning_rate": 3.9374418666989375e-06, + "loss": 0.0205, + "step": 87920 + }, + { + "epoch": 17.070482015436145, + "grad_norm": 0.30173787474632263, + "learning_rate": 3.931012912957838e-06, + "loss": 0.0202, + "step": 87930 + }, + { + "epoch": 17.072423668753945, + "grad_norm": 0.13707822561264038, + "learning_rate": 3.9245889972519045e-06, + "loss": 0.0204, + "step": 87940 + }, + { + "epoch": 17.074365322071746, + "grad_norm": 0.16425926983356476, + "learning_rate": 3.9181701202836265e-06, + "loss": 0.0205, + "step": 87950 + }, + { + "epoch": 17.076306975389546, + "grad_norm": 0.2312711477279663, + "learning_rate": 3.911756282755003e-06, + "loss": 0.0241, + "step": 87960 + }, + { + "epoch": 17.078248628707346, + "grad_norm": 0.15815213322639465, + "learning_rate": 3.905347485367411e-06, + "loss": 0.0214, + "step": 87970 + }, + { + "epoch": 17.080190282025143, + "grad_norm": 0.13223567605018616, + "learning_rate": 3.898943728821725e-06, + "loss": 0.0227, + "step": 87980 + }, + { + "epoch": 17.082131935342943, + "grad_norm": 0.13662198185920715, + "learning_rate": 3.892545013818227e-06, + "loss": 0.024, + "step": 87990 + }, + { + "epoch": 17.084073588660743, + "grad_norm": 0.09766262769699097, + "learning_rate": 3.886151341056687e-06, + "loss": 0.0205, + "step": 88000 + }, + { + "epoch": 17.086015241978544, + "grad_norm": 0.22297537326812744, + "learning_rate": 3.8797627112363066e-06, + "loss": 0.0181, + "step": 88010 + }, + { + "epoch": 17.087956895296344, + "grad_norm": 0.19295518100261688, + "learning_rate": 3.873379125055743e-06, + "loss": 0.0249, + "step": 88020 + }, + { + "epoch": 17.089898548614144, + "grad_norm": 0.1858977973461151, + "learning_rate": 3.867000583213082e-06, + "loss": 0.0152, + "step": 88030 + }, + { + "epoch": 17.091840201931944, + "grad_norm": 0.2799648344516754, + "learning_rate": 3.860627086405871e-06, + "loss": 0.0247, + "step": 88040 + }, + { + "epoch": 17.093781855249745, + "grad_norm": 0.2965911626815796, + "learning_rate": 3.854258635331126e-06, + "loss": 0.0271, + "step": 88050 + }, + { + "epoch": 17.095723508567545, + "grad_norm": 0.1797155886888504, + "learning_rate": 3.8478952306852665e-06, + "loss": 0.0257, + "step": 88060 + }, + { + "epoch": 17.097665161885345, + "grad_norm": 0.16268683969974518, + "learning_rate": 3.841536873164198e-06, + "loss": 0.026, + "step": 88070 + }, + { + "epoch": 17.099606815203146, + "grad_norm": 0.23519285023212433, + "learning_rate": 3.8351835634632626e-06, + "loss": 0.0179, + "step": 88080 + }, + { + "epoch": 17.101548468520946, + "grad_norm": 0.232171431183815, + "learning_rate": 3.828835302277245e-06, + "loss": 0.0224, + "step": 88090 + }, + { + "epoch": 17.103490121838746, + "grad_norm": 0.18557198345661163, + "learning_rate": 3.822492090300378e-06, + "loss": 0.0246, + "step": 88100 + }, + { + "epoch": 17.105431775156546, + "grad_norm": 0.15115702152252197, + "learning_rate": 3.8161539282263566e-06, + "loss": 0.02, + "step": 88110 + }, + { + "epoch": 17.107373428474347, + "grad_norm": 0.24219375848770142, + "learning_rate": 3.8098208167482995e-06, + "loss": 0.027, + "step": 88120 + }, + { + "epoch": 17.109315081792147, + "grad_norm": 0.23162716627120972, + "learning_rate": 3.803492756558796e-06, + "loss": 0.018, + "step": 88130 + }, + { + "epoch": 17.111256735109947, + "grad_norm": 0.12730556726455688, + "learning_rate": 3.797169748349855e-06, + "loss": 0.019, + "step": 88140 + }, + { + "epoch": 17.113198388427747, + "grad_norm": 0.1574045866727829, + "learning_rate": 3.790851792812977e-06, + "loss": 0.0222, + "step": 88150 + }, + { + "epoch": 17.115140041745548, + "grad_norm": 0.20453546941280365, + "learning_rate": 3.784538890639072e-06, + "loss": 0.0225, + "step": 88160 + }, + { + "epoch": 17.117081695063348, + "grad_norm": 0.1688413769006729, + "learning_rate": 3.7782310425185153e-06, + "loss": 0.0173, + "step": 88170 + }, + { + "epoch": 17.11902334838115, + "grad_norm": 0.12119920551776886, + "learning_rate": 3.77192824914111e-06, + "loss": 0.029, + "step": 88180 + }, + { + "epoch": 17.120965001698945, + "grad_norm": 0.15664704144001007, + "learning_rate": 3.7656305111961333e-06, + "loss": 0.0187, + "step": 88190 + }, + { + "epoch": 17.122906655016745, + "grad_norm": 0.18633463978767395, + "learning_rate": 3.7593378293722782e-06, + "loss": 0.0247, + "step": 88200 + }, + { + "epoch": 17.124848308334546, + "grad_norm": 0.24085932970046997, + "learning_rate": 3.7530502043577333e-06, + "loss": 0.0175, + "step": 88210 + }, + { + "epoch": 17.126789961652346, + "grad_norm": 0.1331506371498108, + "learning_rate": 3.746767636840076e-06, + "loss": 0.0169, + "step": 88220 + }, + { + "epoch": 17.128731614970146, + "grad_norm": 0.1415100246667862, + "learning_rate": 3.74049012750638e-06, + "loss": 0.0162, + "step": 88230 + }, + { + "epoch": 17.130673268287946, + "grad_norm": 0.2775866389274597, + "learning_rate": 3.734217677043128e-06, + "loss": 0.0188, + "step": 88240 + }, + { + "epoch": 17.132614921605747, + "grad_norm": 0.10455576330423355, + "learning_rate": 3.727950286136267e-06, + "loss": 0.0224, + "step": 88250 + }, + { + "epoch": 17.134556574923547, + "grad_norm": 0.25140827894210815, + "learning_rate": 3.7216879554711925e-06, + "loss": 0.021, + "step": 88260 + }, + { + "epoch": 17.136498228241347, + "grad_norm": 0.17281340062618256, + "learning_rate": 3.715430685732757e-06, + "loss": 0.0207, + "step": 88270 + }, + { + "epoch": 17.138439881559147, + "grad_norm": 0.1342649906873703, + "learning_rate": 3.7091784776052243e-06, + "loss": 0.0199, + "step": 88280 + }, + { + "epoch": 17.140381534876948, + "grad_norm": 0.17370839416980743, + "learning_rate": 3.7029313317723356e-06, + "loss": 0.022, + "step": 88290 + }, + { + "epoch": 17.142323188194748, + "grad_norm": 0.17309565842151642, + "learning_rate": 3.6966892489172787e-06, + "loss": 0.0217, + "step": 88300 + }, + { + "epoch": 17.14426484151255, + "grad_norm": 0.1528434455394745, + "learning_rate": 3.6904522297226575e-06, + "loss": 0.0207, + "step": 88310 + }, + { + "epoch": 17.14620649483035, + "grad_norm": 0.22530138492584229, + "learning_rate": 3.6842202748705647e-06, + "loss": 0.0148, + "step": 88320 + }, + { + "epoch": 17.14814814814815, + "grad_norm": 0.16943228244781494, + "learning_rate": 3.677993385042494e-06, + "loss": 0.0217, + "step": 88330 + }, + { + "epoch": 17.15008980146595, + "grad_norm": 0.22307728230953217, + "learning_rate": 3.67177156091944e-06, + "loss": 0.0187, + "step": 88340 + }, + { + "epoch": 17.15203145478375, + "grad_norm": 0.26785844564437866, + "learning_rate": 3.6655548031817866e-06, + "loss": 0.0213, + "step": 88350 + }, + { + "epoch": 17.15397310810155, + "grad_norm": 0.23421545326709747, + "learning_rate": 3.6593431125094e-06, + "loss": 0.0196, + "step": 88360 + }, + { + "epoch": 17.15591476141935, + "grad_norm": 0.2636798024177551, + "learning_rate": 3.6531364895815766e-06, + "loss": 0.0208, + "step": 88370 + }, + { + "epoch": 17.15785641473715, + "grad_norm": 0.24334366619586945, + "learning_rate": 3.6469349350770677e-06, + "loss": 0.025, + "step": 88380 + }, + { + "epoch": 17.15979806805495, + "grad_norm": 0.1590246707201004, + "learning_rate": 3.6407384496740525e-06, + "loss": 0.0214, + "step": 88390 + }, + { + "epoch": 17.161739721372747, + "grad_norm": 0.12613771855831146, + "learning_rate": 3.6345470340501954e-06, + "loss": 0.0177, + "step": 88400 + }, + { + "epoch": 17.163681374690547, + "grad_norm": 0.21861806511878967, + "learning_rate": 3.6283606888825537e-06, + "loss": 0.0186, + "step": 88410 + }, + { + "epoch": 17.165623028008348, + "grad_norm": 0.16571423411369324, + "learning_rate": 3.622179414847676e-06, + "loss": 0.0248, + "step": 88420 + }, + { + "epoch": 17.167564681326148, + "grad_norm": 0.20314054191112518, + "learning_rate": 3.616003212621527e-06, + "loss": 0.0224, + "step": 88430 + }, + { + "epoch": 17.16950633464395, + "grad_norm": 0.21954242885112762, + "learning_rate": 3.609832082879533e-06, + "loss": 0.0284, + "step": 88440 + }, + { + "epoch": 17.17144798796175, + "grad_norm": 0.17326636612415314, + "learning_rate": 3.6036660262965485e-06, + "loss": 0.0193, + "step": 88450 + }, + { + "epoch": 17.17338964127955, + "grad_norm": 0.1685439646244049, + "learning_rate": 3.597505043546895e-06, + "loss": 0.0196, + "step": 88460 + }, + { + "epoch": 17.17533129459735, + "grad_norm": 0.11986982077360153, + "learning_rate": 3.5913491353043227e-06, + "loss": 0.0206, + "step": 88470 + }, + { + "epoch": 17.17727294791515, + "grad_norm": 0.1863524168729782, + "learning_rate": 3.5851983022420375e-06, + "loss": 0.0201, + "step": 88480 + }, + { + "epoch": 17.17921460123295, + "grad_norm": 0.16067418456077576, + "learning_rate": 3.5790525450326797e-06, + "loss": 0.0203, + "step": 88490 + }, + { + "epoch": 17.18115625455075, + "grad_norm": 0.16624802350997925, + "learning_rate": 3.572911864348344e-06, + "loss": 0.0231, + "step": 88500 + }, + { + "epoch": 17.18309790786855, + "grad_norm": 0.17291662096977234, + "learning_rate": 3.5667762608605717e-06, + "loss": 0.0174, + "step": 88510 + }, + { + "epoch": 17.18503956118635, + "grad_norm": 0.21307668089866638, + "learning_rate": 3.5606457352403255e-06, + "loss": 0.0284, + "step": 88520 + }, + { + "epoch": 17.18698121450415, + "grad_norm": 0.26371023058891296, + "learning_rate": 3.554520288158042e-06, + "loss": 0.0231, + "step": 88530 + }, + { + "epoch": 17.18892286782195, + "grad_norm": 0.160572350025177, + "learning_rate": 3.5483999202835903e-06, + "loss": 0.0212, + "step": 88540 + }, + { + "epoch": 17.19086452113975, + "grad_norm": 0.22572799026966095, + "learning_rate": 3.5422846322862913e-06, + "loss": 0.0202, + "step": 88550 + }, + { + "epoch": 17.19280617445755, + "grad_norm": 0.2046472728252411, + "learning_rate": 3.5361744248348925e-06, + "loss": 0.0201, + "step": 88560 + }, + { + "epoch": 17.194747827775352, + "grad_norm": 0.2489827573299408, + "learning_rate": 3.5300692985976047e-06, + "loss": 0.0211, + "step": 88570 + }, + { + "epoch": 17.196689481093152, + "grad_norm": 0.1479320228099823, + "learning_rate": 3.52396925424206e-06, + "loss": 0.0254, + "step": 88580 + }, + { + "epoch": 17.198631134410952, + "grad_norm": 0.14298292994499207, + "learning_rate": 3.5178742924353747e-06, + "loss": 0.0224, + "step": 88590 + }, + { + "epoch": 17.200572787728753, + "grad_norm": 0.13571257889270782, + "learning_rate": 3.511784413844066e-06, + "loss": 0.0183, + "step": 88600 + }, + { + "epoch": 17.202514441046553, + "grad_norm": 0.2057771235704422, + "learning_rate": 3.5056996191341287e-06, + "loss": 0.0255, + "step": 88610 + }, + { + "epoch": 17.20445609436435, + "grad_norm": 0.17908020317554474, + "learning_rate": 3.4996199089709692e-06, + "loss": 0.0228, + "step": 88620 + }, + { + "epoch": 17.20639774768215, + "grad_norm": 0.11141464859247208, + "learning_rate": 3.4935452840194725e-06, + "loss": 0.0237, + "step": 88630 + }, + { + "epoch": 17.20833940099995, + "grad_norm": 0.2568124532699585, + "learning_rate": 3.48747574494393e-06, + "loss": 0.0292, + "step": 88640 + }, + { + "epoch": 17.21028105431775, + "grad_norm": 0.2080863118171692, + "learning_rate": 3.4814112924081154e-06, + "loss": 0.0169, + "step": 88650 + }, + { + "epoch": 17.21222270763555, + "grad_norm": 0.12000516802072525, + "learning_rate": 3.475351927075221e-06, + "loss": 0.0181, + "step": 88660 + }, + { + "epoch": 17.21416436095335, + "grad_norm": 0.20223480463027954, + "learning_rate": 3.469297649607894e-06, + "loss": 0.0228, + "step": 88670 + }, + { + "epoch": 17.21610601427115, + "grad_norm": 0.0643446296453476, + "learning_rate": 3.463248460668217e-06, + "loss": 0.0147, + "step": 88680 + }, + { + "epoch": 17.21804766758895, + "grad_norm": 0.21617664396762848, + "learning_rate": 3.457204360917726e-06, + "loss": 0.0226, + "step": 88690 + }, + { + "epoch": 17.219989320906752, + "grad_norm": 0.20864816009998322, + "learning_rate": 3.4511653510173827e-06, + "loss": 0.0238, + "step": 88700 + }, + { + "epoch": 17.221930974224552, + "grad_norm": 0.19394105672836304, + "learning_rate": 3.4451314316276074e-06, + "loss": 0.018, + "step": 88710 + }, + { + "epoch": 17.223872627542352, + "grad_norm": 0.22730661928653717, + "learning_rate": 3.4391026034082674e-06, + "loss": 0.0247, + "step": 88720 + }, + { + "epoch": 17.225814280860153, + "grad_norm": 0.09463474154472351, + "learning_rate": 3.433078867018674e-06, + "loss": 0.0177, + "step": 88730 + }, + { + "epoch": 17.227755934177953, + "grad_norm": 0.13926540315151215, + "learning_rate": 3.427060223117551e-06, + "loss": 0.022, + "step": 88740 + }, + { + "epoch": 17.229697587495753, + "grad_norm": 0.15105688571929932, + "learning_rate": 3.4210466723631042e-06, + "loss": 0.0188, + "step": 88750 + }, + { + "epoch": 17.231639240813553, + "grad_norm": 0.18684779107570648, + "learning_rate": 3.415038215412969e-06, + "loss": 0.0179, + "step": 88760 + }, + { + "epoch": 17.233580894131354, + "grad_norm": 0.1646410971879959, + "learning_rate": 3.4090348529242088e-06, + "loss": 0.0232, + "step": 88770 + }, + { + "epoch": 17.235522547449154, + "grad_norm": 0.1489696502685547, + "learning_rate": 3.4030365855533487e-06, + "loss": 0.0163, + "step": 88780 + }, + { + "epoch": 17.237464200766954, + "grad_norm": 0.2342412918806076, + "learning_rate": 3.397043413956347e-06, + "loss": 0.0227, + "step": 88790 + }, + { + "epoch": 17.239405854084755, + "grad_norm": 0.19201812148094177, + "learning_rate": 3.3910553387886237e-06, + "loss": 0.0229, + "step": 88800 + }, + { + "epoch": 17.241347507402555, + "grad_norm": 0.20156867802143097, + "learning_rate": 3.3850723607049996e-06, + "loss": 0.0201, + "step": 88810 + }, + { + "epoch": 17.243289160720355, + "grad_norm": 0.22651703655719757, + "learning_rate": 3.3790944803597902e-06, + "loss": 0.0216, + "step": 88820 + }, + { + "epoch": 17.245230814038152, + "grad_norm": 0.17311285436153412, + "learning_rate": 3.373121698406706e-06, + "loss": 0.0186, + "step": 88830 + }, + { + "epoch": 17.247172467355952, + "grad_norm": 0.18832919001579285, + "learning_rate": 3.3671540154989302e-06, + "loss": 0.0275, + "step": 88840 + }, + { + "epoch": 17.249114120673752, + "grad_norm": 0.11477043479681015, + "learning_rate": 3.361191432289079e-06, + "loss": 0.0248, + "step": 88850 + }, + { + "epoch": 17.251055773991553, + "grad_norm": 0.22031959891319275, + "learning_rate": 3.3552339494292207e-06, + "loss": 0.0302, + "step": 88860 + }, + { + "epoch": 17.252997427309353, + "grad_norm": 0.18224912881851196, + "learning_rate": 3.349281567570839e-06, + "loss": 0.0176, + "step": 88870 + }, + { + "epoch": 17.254939080627153, + "grad_norm": 0.1551275998353958, + "learning_rate": 3.3433342873648965e-06, + "loss": 0.0206, + "step": 88880 + }, + { + "epoch": 17.256880733944953, + "grad_norm": 0.09277801215648651, + "learning_rate": 3.337392109461762e-06, + "loss": 0.015, + "step": 88890 + }, + { + "epoch": 17.258822387262754, + "grad_norm": 0.10240160673856735, + "learning_rate": 3.331455034511266e-06, + "loss": 0.0247, + "step": 88900 + }, + { + "epoch": 17.260764040580554, + "grad_norm": 0.17152950167655945, + "learning_rate": 3.325523063162689e-06, + "loss": 0.0185, + "step": 88910 + }, + { + "epoch": 17.262705693898354, + "grad_norm": 0.11868061870336533, + "learning_rate": 3.3195961960647393e-06, + "loss": 0.0176, + "step": 88920 + }, + { + "epoch": 17.264647347216155, + "grad_norm": 0.18836231529712677, + "learning_rate": 3.31367443386556e-06, + "loss": 0.0208, + "step": 88930 + }, + { + "epoch": 17.266589000533955, + "grad_norm": 0.24647098779678345, + "learning_rate": 3.3077577772127556e-06, + "loss": 0.0204, + "step": 88940 + }, + { + "epoch": 17.268530653851755, + "grad_norm": 0.2834285497665405, + "learning_rate": 3.301846226753358e-06, + "loss": 0.0217, + "step": 88950 + }, + { + "epoch": 17.270472307169555, + "grad_norm": 0.1889467090368271, + "learning_rate": 3.2959397831338444e-06, + "loss": 0.0227, + "step": 88960 + }, + { + "epoch": 17.272413960487356, + "grad_norm": 0.2534853518009186, + "learning_rate": 3.2900384470001432e-06, + "loss": 0.0159, + "step": 88970 + }, + { + "epoch": 17.274355613805156, + "grad_norm": 0.16453416645526886, + "learning_rate": 3.28414221899761e-06, + "loss": 0.0184, + "step": 88980 + }, + { + "epoch": 17.276297267122956, + "grad_norm": 0.1050553172826767, + "learning_rate": 3.278251099771046e-06, + "loss": 0.0184, + "step": 88990 + }, + { + "epoch": 17.278238920440756, + "grad_norm": 0.13087481260299683, + "learning_rate": 3.2723650899646906e-06, + "loss": 0.0214, + "step": 89000 + }, + { + "epoch": 17.280180573758557, + "grad_norm": 0.2903199791908264, + "learning_rate": 3.2664841902222466e-06, + "loss": 0.0237, + "step": 89010 + }, + { + "epoch": 17.282122227076357, + "grad_norm": 0.21715345978736877, + "learning_rate": 3.2606084011868208e-06, + "loss": 0.0222, + "step": 89020 + }, + { + "epoch": 17.284063880394157, + "grad_norm": 0.10825560241937637, + "learning_rate": 3.254737723500989e-06, + "loss": 0.0164, + "step": 89030 + }, + { + "epoch": 17.286005533711954, + "grad_norm": 0.23508936166763306, + "learning_rate": 3.2488721578067595e-06, + "loss": 0.021, + "step": 89040 + }, + { + "epoch": 17.287947187029754, + "grad_norm": 0.1174711361527443, + "learning_rate": 3.243011704745591e-06, + "loss": 0.0194, + "step": 89050 + }, + { + "epoch": 17.289888840347555, + "grad_norm": 0.17369787395000458, + "learning_rate": 3.237156364958355e-06, + "loss": 0.0197, + "step": 89060 + }, + { + "epoch": 17.291830493665355, + "grad_norm": 0.15847791731357574, + "learning_rate": 3.231306139085394e-06, + "loss": 0.0231, + "step": 89070 + }, + { + "epoch": 17.293772146983155, + "grad_norm": 0.1219094842672348, + "learning_rate": 3.2254610277664797e-06, + "loss": 0.0214, + "step": 89080 + }, + { + "epoch": 17.295713800300955, + "grad_norm": 0.1393786519765854, + "learning_rate": 3.2196210316408183e-06, + "loss": 0.0237, + "step": 89090 + }, + { + "epoch": 17.297655453618756, + "grad_norm": 0.24466781318187714, + "learning_rate": 3.2137861513470644e-06, + "loss": 0.0205, + "step": 89100 + }, + { + "epoch": 17.299597106936556, + "grad_norm": 0.16608937084674835, + "learning_rate": 3.2079563875233252e-06, + "loss": 0.0196, + "step": 89110 + }, + { + "epoch": 17.301538760254356, + "grad_norm": 0.19366590678691864, + "learning_rate": 3.2021317408071184e-06, + "loss": 0.0293, + "step": 89120 + }, + { + "epoch": 17.303480413572156, + "grad_norm": 0.19662177562713623, + "learning_rate": 3.196312211835434e-06, + "loss": 0.0244, + "step": 89130 + }, + { + "epoch": 17.305422066889957, + "grad_norm": 0.18682068586349487, + "learning_rate": 3.1904978012446686e-06, + "loss": 0.0215, + "step": 89140 + }, + { + "epoch": 17.307363720207757, + "grad_norm": 0.40863481163978577, + "learning_rate": 3.184688509670691e-06, + "loss": 0.0237, + "step": 89150 + }, + { + "epoch": 17.309305373525557, + "grad_norm": 0.32392391562461853, + "learning_rate": 3.178884337748783e-06, + "loss": 0.0208, + "step": 89160 + }, + { + "epoch": 17.311247026843358, + "grad_norm": 0.17343062162399292, + "learning_rate": 3.173085286113697e-06, + "loss": 0.0225, + "step": 89170 + }, + { + "epoch": 17.313188680161158, + "grad_norm": 0.16810342669487, + "learning_rate": 3.1672913553995986e-06, + "loss": 0.0271, + "step": 89180 + }, + { + "epoch": 17.315130333478958, + "grad_norm": 0.14537906646728516, + "learning_rate": 3.1615025462401138e-06, + "loss": 0.0321, + "step": 89190 + }, + { + "epoch": 17.31707198679676, + "grad_norm": 0.22155922651290894, + "learning_rate": 3.1557188592682808e-06, + "loss": 0.0227, + "step": 89200 + }, + { + "epoch": 17.31901364011456, + "grad_norm": 0.11899229139089584, + "learning_rate": 3.1499402951166045e-06, + "loss": 0.0161, + "step": 89210 + }, + { + "epoch": 17.32095529343236, + "grad_norm": 0.1583900898694992, + "learning_rate": 3.144166854417019e-06, + "loss": 0.0196, + "step": 89220 + }, + { + "epoch": 17.32289694675016, + "grad_norm": 0.13872544467449188, + "learning_rate": 3.1383985378009073e-06, + "loss": 0.0185, + "step": 89230 + }, + { + "epoch": 17.32483860006796, + "grad_norm": 0.1570669263601303, + "learning_rate": 3.1326353458990654e-06, + "loss": 0.0243, + "step": 89240 + }, + { + "epoch": 17.326780253385756, + "grad_norm": 0.1281091421842575, + "learning_rate": 3.126877279341761e-06, + "loss": 0.0192, + "step": 89250 + }, + { + "epoch": 17.328721906703556, + "grad_norm": 0.1581307053565979, + "learning_rate": 3.1211243387586964e-06, + "loss": 0.0218, + "step": 89260 + }, + { + "epoch": 17.330663560021357, + "grad_norm": 0.26165422797203064, + "learning_rate": 3.115376524778979e-06, + "loss": 0.0184, + "step": 89270 + }, + { + "epoch": 17.332605213339157, + "grad_norm": 0.24591945111751556, + "learning_rate": 3.109633838031195e-06, + "loss": 0.0221, + "step": 89280 + }, + { + "epoch": 17.334546866656957, + "grad_norm": 0.1560918539762497, + "learning_rate": 3.103896279143359e-06, + "loss": 0.0257, + "step": 89290 + }, + { + "epoch": 17.336488519974758, + "grad_norm": 0.16738225519657135, + "learning_rate": 3.098163848742924e-06, + "loss": 0.0194, + "step": 89300 + }, + { + "epoch": 17.338430173292558, + "grad_norm": 0.12465985864400864, + "learning_rate": 3.0924365474567675e-06, + "loss": 0.0253, + "step": 89310 + }, + { + "epoch": 17.340371826610358, + "grad_norm": 0.19915443658828735, + "learning_rate": 3.086714375911237e-06, + "loss": 0.0248, + "step": 89320 + }, + { + "epoch": 17.34231347992816, + "grad_norm": 0.16631579399108887, + "learning_rate": 3.0809973347320775e-06, + "loss": 0.0219, + "step": 89330 + }, + { + "epoch": 17.34425513324596, + "grad_norm": 0.21495220065116882, + "learning_rate": 3.075285424544516e-06, + "loss": 0.0176, + "step": 89340 + }, + { + "epoch": 17.34619678656376, + "grad_norm": 0.1414126455783844, + "learning_rate": 3.069578645973176e-06, + "loss": 0.0213, + "step": 89350 + }, + { + "epoch": 17.34813843988156, + "grad_norm": 0.1628330796957016, + "learning_rate": 3.0638769996421735e-06, + "loss": 0.0234, + "step": 89360 + }, + { + "epoch": 17.35008009319936, + "grad_norm": 0.1614503115415573, + "learning_rate": 3.058180486175005e-06, + "loss": 0.0209, + "step": 89370 + }, + { + "epoch": 17.35202174651716, + "grad_norm": 0.18696144223213196, + "learning_rate": 3.052489106194645e-06, + "loss": 0.021, + "step": 89380 + }, + { + "epoch": 17.35396339983496, + "grad_norm": 0.20497830212116241, + "learning_rate": 3.0468028603234946e-06, + "loss": 0.0178, + "step": 89390 + }, + { + "epoch": 17.35590505315276, + "grad_norm": 0.2157273143529892, + "learning_rate": 3.04112174918339e-06, + "loss": 0.0258, + "step": 89400 + }, + { + "epoch": 17.35784670647056, + "grad_norm": 0.18268844485282898, + "learning_rate": 3.0354457733955954e-06, + "loss": 0.0235, + "step": 89410 + }, + { + "epoch": 17.35978835978836, + "grad_norm": 0.13435880839824677, + "learning_rate": 3.029774933580859e-06, + "loss": 0.0299, + "step": 89420 + }, + { + "epoch": 17.36173001310616, + "grad_norm": 0.15086586773395538, + "learning_rate": 3.0241092303593065e-06, + "loss": 0.019, + "step": 89430 + }, + { + "epoch": 17.36367166642396, + "grad_norm": 0.2077203094959259, + "learning_rate": 3.018448664350554e-06, + "loss": 0.0205, + "step": 89440 + }, + { + "epoch": 17.36561331974176, + "grad_norm": 0.1220715343952179, + "learning_rate": 3.0127932361736055e-06, + "loss": 0.0151, + "step": 89450 + }, + { + "epoch": 17.367554973059562, + "grad_norm": 0.17495392262935638, + "learning_rate": 3.00714294644695e-06, + "loss": 0.0258, + "step": 89460 + }, + { + "epoch": 17.36949662637736, + "grad_norm": 0.22731786966323853, + "learning_rate": 3.0014977957884828e-06, + "loss": 0.0221, + "step": 89470 + }, + { + "epoch": 17.37143827969516, + "grad_norm": 0.18953384459018707, + "learning_rate": 2.99585778481557e-06, + "loss": 0.0184, + "step": 89480 + }, + { + "epoch": 17.37337993301296, + "grad_norm": 0.18830884993076324, + "learning_rate": 2.990222914144969e-06, + "loss": 0.0256, + "step": 89490 + }, + { + "epoch": 17.37532158633076, + "grad_norm": 0.21524731814861298, + "learning_rate": 2.984593184392914e-06, + "loss": 0.0215, + "step": 89500 + }, + { + "epoch": 17.37726323964856, + "grad_norm": 0.23029382526874542, + "learning_rate": 2.978968596175069e-06, + "loss": 0.0226, + "step": 89510 + }, + { + "epoch": 17.37920489296636, + "grad_norm": 0.3563918471336365, + "learning_rate": 2.9733491501065137e-06, + "loss": 0.0215, + "step": 89520 + }, + { + "epoch": 17.38114654628416, + "grad_norm": 0.22770822048187256, + "learning_rate": 2.9677348468018005e-06, + "loss": 0.0225, + "step": 89530 + }, + { + "epoch": 17.38308819960196, + "grad_norm": 0.13501028716564178, + "learning_rate": 2.9621256868748726e-06, + "loss": 0.0174, + "step": 89540 + }, + { + "epoch": 17.38502985291976, + "grad_norm": 0.22958478331565857, + "learning_rate": 2.956521670939183e-06, + "loss": 0.0209, + "step": 89550 + }, + { + "epoch": 17.38697150623756, + "grad_norm": 0.15202859044075012, + "learning_rate": 2.950922799607542e-06, + "loss": 0.0238, + "step": 89560 + }, + { + "epoch": 17.38891315955536, + "grad_norm": 0.28588438034057617, + "learning_rate": 2.9453290734922535e-06, + "loss": 0.0277, + "step": 89570 + }, + { + "epoch": 17.39085481287316, + "grad_norm": 0.19579268991947174, + "learning_rate": 2.939740493205029e-06, + "loss": 0.0207, + "step": 89580 + }, + { + "epoch": 17.392796466190962, + "grad_norm": 0.12926916778087616, + "learning_rate": 2.9341570593570345e-06, + "loss": 0.0205, + "step": 89590 + }, + { + "epoch": 17.394738119508762, + "grad_norm": 0.19080311059951782, + "learning_rate": 2.928578772558843e-06, + "loss": 0.0224, + "step": 89600 + }, + { + "epoch": 17.396679772826563, + "grad_norm": 0.10211766511201859, + "learning_rate": 2.923005633420528e-06, + "loss": 0.0233, + "step": 89610 + }, + { + "epoch": 17.398621426144363, + "grad_norm": 0.20937012135982513, + "learning_rate": 2.9174376425515237e-06, + "loss": 0.0185, + "step": 89620 + }, + { + "epoch": 17.400563079462163, + "grad_norm": 0.13984087109565735, + "learning_rate": 2.91187480056076e-06, + "loss": 0.0243, + "step": 89630 + }, + { + "epoch": 17.402504732779963, + "grad_norm": 0.21996194124221802, + "learning_rate": 2.9063171080565664e-06, + "loss": 0.0213, + "step": 89640 + }, + { + "epoch": 17.404446386097764, + "grad_norm": 0.14721667766571045, + "learning_rate": 2.900764565646741e-06, + "loss": 0.0269, + "step": 89650 + }, + { + "epoch": 17.406388039415564, + "grad_norm": 0.22418656945228577, + "learning_rate": 2.895217173938475e-06, + "loss": 0.0255, + "step": 89660 + }, + { + "epoch": 17.408329692733364, + "grad_norm": 0.27448710799217224, + "learning_rate": 2.8896749335384445e-06, + "loss": 0.0267, + "step": 89670 + }, + { + "epoch": 17.41027134605116, + "grad_norm": 0.22022242844104767, + "learning_rate": 2.884137845052731e-06, + "loss": 0.0227, + "step": 89680 + }, + { + "epoch": 17.41221299936896, + "grad_norm": 0.1503981351852417, + "learning_rate": 2.8786059090868734e-06, + "loss": 0.0204, + "step": 89690 + }, + { + "epoch": 17.41415465268676, + "grad_norm": 0.14115627110004425, + "learning_rate": 2.87307912624582e-06, + "loss": 0.0222, + "step": 89700 + }, + { + "epoch": 17.41609630600456, + "grad_norm": 0.172317773103714, + "learning_rate": 2.867557497133977e-06, + "loss": 0.0191, + "step": 89710 + }, + { + "epoch": 17.418037959322362, + "grad_norm": 0.16617290675640106, + "learning_rate": 2.862041022355194e-06, + "loss": 0.0202, + "step": 89720 + }, + { + "epoch": 17.419979612640162, + "grad_norm": 0.2762628495693207, + "learning_rate": 2.8565297025127226e-06, + "loss": 0.0239, + "step": 89730 + }, + { + "epoch": 17.421921265957963, + "grad_norm": 0.12730850279331207, + "learning_rate": 2.8510235382092853e-06, + "loss": 0.0192, + "step": 89740 + }, + { + "epoch": 17.423862919275763, + "grad_norm": 0.19373513758182526, + "learning_rate": 2.845522530047029e-06, + "loss": 0.0285, + "step": 89750 + }, + { + "epoch": 17.425804572593563, + "grad_norm": 0.23068653047084808, + "learning_rate": 2.8400266786275387e-06, + "loss": 0.0255, + "step": 89760 + }, + { + "epoch": 17.427746225911363, + "grad_norm": 0.1955900639295578, + "learning_rate": 2.834535984551817e-06, + "loss": 0.0215, + "step": 89770 + }, + { + "epoch": 17.429687879229164, + "grad_norm": 0.24794228374958038, + "learning_rate": 2.8290504484203397e-06, + "loss": 0.0214, + "step": 89780 + }, + { + "epoch": 17.431629532546964, + "grad_norm": 0.17854751646518707, + "learning_rate": 2.8235700708329705e-06, + "loss": 0.0227, + "step": 89790 + }, + { + "epoch": 17.433571185864764, + "grad_norm": 0.19650037586688995, + "learning_rate": 2.818094852389058e-06, + "loss": 0.0269, + "step": 89800 + }, + { + "epoch": 17.435512839182564, + "grad_norm": 0.11955279856920242, + "learning_rate": 2.8126247936873516e-06, + "loss": 0.0214, + "step": 89810 + }, + { + "epoch": 17.437454492500365, + "grad_norm": 0.19419953227043152, + "learning_rate": 2.8071598953260614e-06, + "loss": 0.0163, + "step": 89820 + }, + { + "epoch": 17.439396145818165, + "grad_norm": 0.25840437412261963, + "learning_rate": 2.8017001579028035e-06, + "loss": 0.0203, + "step": 89830 + }, + { + "epoch": 17.441337799135965, + "grad_norm": 0.124715656042099, + "learning_rate": 2.7962455820146617e-06, + "loss": 0.0233, + "step": 89840 + }, + { + "epoch": 17.443279452453766, + "grad_norm": 0.2636062800884247, + "learning_rate": 2.7907961682581253e-06, + "loss": 0.0266, + "step": 89850 + }, + { + "epoch": 17.445221105771566, + "grad_norm": 0.1675298660993576, + "learning_rate": 2.7853519172291453e-06, + "loss": 0.0225, + "step": 89860 + }, + { + "epoch": 17.447162759089366, + "grad_norm": 0.12382973730564117, + "learning_rate": 2.7799128295230904e-06, + "loss": 0.0223, + "step": 89870 + }, + { + "epoch": 17.449104412407166, + "grad_norm": 0.13025593757629395, + "learning_rate": 2.774478905734784e-06, + "loss": 0.0148, + "step": 89880 + }, + { + "epoch": 17.451046065724963, + "grad_norm": 0.1334162950515747, + "learning_rate": 2.7690501464584516e-06, + "loss": 0.0212, + "step": 89890 + }, + { + "epoch": 17.452987719042763, + "grad_norm": 0.1551835685968399, + "learning_rate": 2.763626552287796e-06, + "loss": 0.0207, + "step": 89900 + }, + { + "epoch": 17.454929372360564, + "grad_norm": 0.1471937745809555, + "learning_rate": 2.7582081238159083e-06, + "loss": 0.0223, + "step": 89910 + }, + { + "epoch": 17.456871025678364, + "grad_norm": 0.18183492124080658, + "learning_rate": 2.7527948616353548e-06, + "loss": 0.0168, + "step": 89920 + }, + { + "epoch": 17.458812678996164, + "grad_norm": 0.17179256677627563, + "learning_rate": 2.7473867663381226e-06, + "loss": 0.0186, + "step": 89930 + }, + { + "epoch": 17.460754332313964, + "grad_norm": 0.16183142364025116, + "learning_rate": 2.7419838385156383e-06, + "loss": 0.0225, + "step": 89940 + }, + { + "epoch": 17.462695985631765, + "grad_norm": 0.235368013381958, + "learning_rate": 2.7365860787587407e-06, + "loss": 0.0207, + "step": 89950 + }, + { + "epoch": 17.464637638949565, + "grad_norm": 0.14977708458900452, + "learning_rate": 2.7311934876577295e-06, + "loss": 0.0159, + "step": 89960 + }, + { + "epoch": 17.466579292267365, + "grad_norm": 0.20746780931949615, + "learning_rate": 2.725806065802339e-06, + "loss": 0.0189, + "step": 89970 + }, + { + "epoch": 17.468520945585166, + "grad_norm": 0.3730829060077667, + "learning_rate": 2.720423813781714e-06, + "loss": 0.0274, + "step": 89980 + }, + { + "epoch": 17.470462598902966, + "grad_norm": 0.13715222477912903, + "learning_rate": 2.715046732184462e-06, + "loss": 0.0237, + "step": 89990 + }, + { + "epoch": 17.472404252220766, + "grad_norm": 0.2400735318660736, + "learning_rate": 2.7096748215986013e-06, + "loss": 0.026, + "step": 90000 + }, + { + "epoch": 17.474345905538566, + "grad_norm": 0.16423968970775604, + "learning_rate": 2.704308082611612e-06, + "loss": 0.0215, + "step": 90010 + }, + { + "epoch": 17.476287558856367, + "grad_norm": 0.15028052031993866, + "learning_rate": 2.698946515810374e-06, + "loss": 0.0184, + "step": 90020 + }, + { + "epoch": 17.478229212174167, + "grad_norm": 0.19777309894561768, + "learning_rate": 2.6935901217812367e-06, + "loss": 0.0199, + "step": 90030 + }, + { + "epoch": 17.480170865491967, + "grad_norm": 0.17744463682174683, + "learning_rate": 2.688238901109952e-06, + "loss": 0.0203, + "step": 90040 + }, + { + "epoch": 17.482112518809767, + "grad_norm": 0.09402589499950409, + "learning_rate": 2.68289285438173e-06, + "loss": 0.023, + "step": 90050 + }, + { + "epoch": 17.484054172127568, + "grad_norm": 0.16294457018375397, + "learning_rate": 2.6775519821812033e-06, + "loss": 0.0303, + "step": 90060 + }, + { + "epoch": 17.485995825445368, + "grad_norm": 0.22040621936321259, + "learning_rate": 2.672216285092449e-06, + "loss": 0.0222, + "step": 90070 + }, + { + "epoch": 17.48793747876317, + "grad_norm": 0.20485860109329224, + "learning_rate": 2.666885763698962e-06, + "loss": 0.0221, + "step": 90080 + }, + { + "epoch": 17.48987913208097, + "grad_norm": 0.12951567769050598, + "learning_rate": 2.661560418583692e-06, + "loss": 0.0211, + "step": 90090 + }, + { + "epoch": 17.491820785398765, + "grad_norm": 0.2438715398311615, + "learning_rate": 2.656240250328995e-06, + "loss": 0.0234, + "step": 90100 + }, + { + "epoch": 17.493762438716566, + "grad_norm": 0.13157455623149872, + "learning_rate": 2.650925259516679e-06, + "loss": 0.0228, + "step": 90110 + }, + { + "epoch": 17.495704092034366, + "grad_norm": 0.17559541761875153, + "learning_rate": 2.645615446727995e-06, + "loss": 0.02, + "step": 90120 + }, + { + "epoch": 17.497645745352166, + "grad_norm": 0.1375848799943924, + "learning_rate": 2.6403108125436115e-06, + "loss": 0.0167, + "step": 90130 + }, + { + "epoch": 17.499587398669966, + "grad_norm": 0.2168065309524536, + "learning_rate": 2.6350113575436264e-06, + "loss": 0.0202, + "step": 90140 + }, + { + "epoch": 17.501529051987767, + "grad_norm": 0.25131523609161377, + "learning_rate": 2.629717082307598e-06, + "loss": 0.0245, + "step": 90150 + }, + { + "epoch": 17.503470705305567, + "grad_norm": 0.11563923954963684, + "learning_rate": 2.6244279874144793e-06, + "loss": 0.023, + "step": 90160 + }, + { + "epoch": 17.505412358623367, + "grad_norm": 0.11185992509126663, + "learning_rate": 2.6191440734426862e-06, + "loss": 0.0204, + "step": 90170 + }, + { + "epoch": 17.507354011941167, + "grad_norm": 0.16033786535263062, + "learning_rate": 2.613865340970062e-06, + "loss": 0.0229, + "step": 90180 + }, + { + "epoch": 17.509295665258968, + "grad_norm": 0.27225202322006226, + "learning_rate": 2.608591790573889e-06, + "loss": 0.0211, + "step": 90190 + }, + { + "epoch": 17.511237318576768, + "grad_norm": 0.24597379565238953, + "learning_rate": 2.603323422830861e-06, + "loss": 0.0254, + "step": 90200 + }, + { + "epoch": 17.51317897189457, + "grad_norm": 0.12317702919244766, + "learning_rate": 2.5980602383171183e-06, + "loss": 0.0211, + "step": 90210 + }, + { + "epoch": 17.51512062521237, + "grad_norm": 0.11784309893846512, + "learning_rate": 2.5928022376082496e-06, + "loss": 0.0219, + "step": 90220 + }, + { + "epoch": 17.51706227853017, + "grad_norm": 0.1359170526266098, + "learning_rate": 2.5875494212792505e-06, + "loss": 0.024, + "step": 90230 + }, + { + "epoch": 17.51900393184797, + "grad_norm": 0.22241300344467163, + "learning_rate": 2.5823017899045565e-06, + "loss": 0.0176, + "step": 90240 + }, + { + "epoch": 17.52094558516577, + "grad_norm": 0.1632193922996521, + "learning_rate": 2.577059344058047e-06, + "loss": 0.0193, + "step": 90250 + }, + { + "epoch": 17.52288723848357, + "grad_norm": 0.2547796964645386, + "learning_rate": 2.5718220843130415e-06, + "loss": 0.0286, + "step": 90260 + }, + { + "epoch": 17.52482889180137, + "grad_norm": 0.1678215116262436, + "learning_rate": 2.5665900112422535e-06, + "loss": 0.021, + "step": 90270 + }, + { + "epoch": 17.52677054511917, + "grad_norm": 0.14574767649173737, + "learning_rate": 2.561363125417876e-06, + "loss": 0.0202, + "step": 90280 + }, + { + "epoch": 17.52871219843697, + "grad_norm": 0.062117237597703934, + "learning_rate": 2.556141427411496e-06, + "loss": 0.0183, + "step": 90290 + }, + { + "epoch": 17.53065385175477, + "grad_norm": 0.19110557436943054, + "learning_rate": 2.550924917794156e-06, + "loss": 0.0231, + "step": 90300 + }, + { + "epoch": 17.53259550507257, + "grad_norm": 0.1215285137295723, + "learning_rate": 2.545713597136329e-06, + "loss": 0.0253, + "step": 90310 + }, + { + "epoch": 17.534537158390368, + "grad_norm": 0.08096106350421906, + "learning_rate": 2.5405074660079298e-06, + "loss": 0.0192, + "step": 90320 + }, + { + "epoch": 17.536478811708168, + "grad_norm": 0.22225520014762878, + "learning_rate": 2.5353065249782647e-06, + "loss": 0.0213, + "step": 90330 + }, + { + "epoch": 17.53842046502597, + "grad_norm": 0.19115452468395233, + "learning_rate": 2.530110774616129e-06, + "loss": 0.0267, + "step": 90340 + }, + { + "epoch": 17.54036211834377, + "grad_norm": 0.15919215977191925, + "learning_rate": 2.524920215489701e-06, + "loss": 0.0207, + "step": 90350 + }, + { + "epoch": 17.54230377166157, + "grad_norm": 0.18474432826042175, + "learning_rate": 2.519734848166627e-06, + "loss": 0.0231, + "step": 90360 + }, + { + "epoch": 17.54424542497937, + "grad_norm": 0.2728552520275116, + "learning_rate": 2.514554673213948e-06, + "loss": 0.0223, + "step": 90370 + }, + { + "epoch": 17.54618707829717, + "grad_norm": 0.1570427417755127, + "learning_rate": 2.5093796911981936e-06, + "loss": 0.0209, + "step": 90380 + }, + { + "epoch": 17.54812873161497, + "grad_norm": 0.13441452383995056, + "learning_rate": 2.5042099026852672e-06, + "loss": 0.0261, + "step": 90390 + }, + { + "epoch": 17.55007038493277, + "grad_norm": 0.16111701726913452, + "learning_rate": 2.499045308240544e-06, + "loss": 0.0193, + "step": 90400 + }, + { + "epoch": 17.55201203825057, + "grad_norm": 0.12157242745161057, + "learning_rate": 2.4938859084288003e-06, + "loss": 0.0186, + "step": 90410 + }, + { + "epoch": 17.55395369156837, + "grad_norm": 0.33300885558128357, + "learning_rate": 2.4887317038142734e-06, + "loss": 0.0227, + "step": 90420 + }, + { + "epoch": 17.55589534488617, + "grad_norm": 0.2116742879152298, + "learning_rate": 2.4835826949606124e-06, + "loss": 0.0371, + "step": 90430 + }, + { + "epoch": 17.55783699820397, + "grad_norm": 0.22576555609703064, + "learning_rate": 2.4784388824309167e-06, + "loss": 0.0165, + "step": 90440 + }, + { + "epoch": 17.55977865152177, + "grad_norm": 0.2515665292739868, + "learning_rate": 2.473300266787687e-06, + "loss": 0.0243, + "step": 90450 + }, + { + "epoch": 17.56172030483957, + "grad_norm": 0.21496297419071198, + "learning_rate": 2.46816684859289e-06, + "loss": 0.0231, + "step": 90460 + }, + { + "epoch": 17.563661958157372, + "grad_norm": 0.18137870728969574, + "learning_rate": 2.4630386284079043e-06, + "loss": 0.018, + "step": 90470 + }, + { + "epoch": 17.565603611475172, + "grad_norm": 0.5158802270889282, + "learning_rate": 2.4579156067935427e-06, + "loss": 0.022, + "step": 90480 + }, + { + "epoch": 17.567545264792972, + "grad_norm": 0.18762581050395966, + "learning_rate": 2.452797784310046e-06, + "loss": 0.0255, + "step": 90490 + }, + { + "epoch": 17.569486918110773, + "grad_norm": 0.2294623851776123, + "learning_rate": 2.447685161517105e-06, + "loss": 0.0198, + "step": 90500 + }, + { + "epoch": 17.571428571428573, + "grad_norm": 0.16501864790916443, + "learning_rate": 2.4425777389738224e-06, + "loss": 0.0207, + "step": 90510 + }, + { + "epoch": 17.573370224746373, + "grad_norm": 0.2492356300354004, + "learning_rate": 2.43747551723873e-06, + "loss": 0.0317, + "step": 90520 + }, + { + "epoch": 17.57531187806417, + "grad_norm": 0.14804232120513916, + "learning_rate": 2.4323784968698136e-06, + "loss": 0.0233, + "step": 90530 + }, + { + "epoch": 17.57725353138197, + "grad_norm": 0.1559220403432846, + "learning_rate": 2.4272866784244565e-06, + "loss": 0.0178, + "step": 90540 + }, + { + "epoch": 17.57919518469977, + "grad_norm": 0.15385207533836365, + "learning_rate": 2.422200062459518e-06, + "loss": 0.0218, + "step": 90550 + }, + { + "epoch": 17.58113683801757, + "grad_norm": 0.1974128633737564, + "learning_rate": 2.41711864953123e-06, + "loss": 0.0214, + "step": 90560 + }, + { + "epoch": 17.58307849133537, + "grad_norm": 0.21428939700126648, + "learning_rate": 2.412042440195322e-06, + "loss": 0.0165, + "step": 90570 + }, + { + "epoch": 17.58502014465317, + "grad_norm": 0.12523867189884186, + "learning_rate": 2.4069714350068993e-06, + "loss": 0.0175, + "step": 90580 + }, + { + "epoch": 17.58696179797097, + "grad_norm": 0.1743469387292862, + "learning_rate": 2.4019056345205294e-06, + "loss": 0.0234, + "step": 90590 + }, + { + "epoch": 17.588903451288772, + "grad_norm": 0.20055803656578064, + "learning_rate": 2.396845039290191e-06, + "loss": 0.0229, + "step": 90600 + }, + { + "epoch": 17.590845104606572, + "grad_norm": 0.1467529535293579, + "learning_rate": 2.3917896498693194e-06, + "loss": 0.0232, + "step": 90610 + }, + { + "epoch": 17.592786757924372, + "grad_norm": 0.13532640039920807, + "learning_rate": 2.3867394668107335e-06, + "loss": 0.0253, + "step": 90620 + }, + { + "epoch": 17.594728411242173, + "grad_norm": 0.2537076473236084, + "learning_rate": 2.3816944906667527e-06, + "loss": 0.0224, + "step": 90630 + }, + { + "epoch": 17.596670064559973, + "grad_norm": 0.17883804440498352, + "learning_rate": 2.3766547219890634e-06, + "loss": 0.0187, + "step": 90640 + }, + { + "epoch": 17.598611717877773, + "grad_norm": 0.32696565985679626, + "learning_rate": 2.371620161328819e-06, + "loss": 0.0225, + "step": 90650 + }, + { + "epoch": 17.600553371195574, + "grad_norm": 0.24963858723640442, + "learning_rate": 2.366590809236574e-06, + "loss": 0.0148, + "step": 90660 + }, + { + "epoch": 17.602495024513374, + "grad_norm": 0.16742393374443054, + "learning_rate": 2.3615666662623494e-06, + "loss": 0.0312, + "step": 90670 + }, + { + "epoch": 17.604436677831174, + "grad_norm": 0.1761654168367386, + "learning_rate": 2.3565477329555664e-06, + "loss": 0.0296, + "step": 90680 + }, + { + "epoch": 17.606378331148974, + "grad_norm": 0.2950090169906616, + "learning_rate": 2.3515340098650975e-06, + "loss": 0.0219, + "step": 90690 + }, + { + "epoch": 17.608319984466775, + "grad_norm": 0.08977602422237396, + "learning_rate": 2.3465254975392258e-06, + "loss": 0.018, + "step": 90700 + }, + { + "epoch": 17.610261637784575, + "grad_norm": 0.13904643058776855, + "learning_rate": 2.3415221965256805e-06, + "loss": 0.0187, + "step": 90710 + }, + { + "epoch": 17.612203291102375, + "grad_norm": 0.21050946414470673, + "learning_rate": 2.3365241073716237e-06, + "loss": 0.0254, + "step": 90720 + }, + { + "epoch": 17.614144944420175, + "grad_norm": 0.08982077240943909, + "learning_rate": 2.3315312306236183e-06, + "loss": 0.0225, + "step": 90730 + }, + { + "epoch": 17.616086597737976, + "grad_norm": 0.16801878809928894, + "learning_rate": 2.326543566827699e-06, + "loss": 0.0217, + "step": 90740 + }, + { + "epoch": 17.618028251055772, + "grad_norm": 0.16845650970935822, + "learning_rate": 2.3215611165292862e-06, + "loss": 0.0195, + "step": 90750 + }, + { + "epoch": 17.619969904373573, + "grad_norm": 0.16016051173210144, + "learning_rate": 2.3165838802732765e-06, + "loss": 0.0218, + "step": 90760 + }, + { + "epoch": 17.621911557691373, + "grad_norm": 0.18387913703918457, + "learning_rate": 2.311611858603957e-06, + "loss": 0.029, + "step": 90770 + }, + { + "epoch": 17.623853211009173, + "grad_norm": 0.2241392582654953, + "learning_rate": 2.3066450520650763e-06, + "loss": 0.0199, + "step": 90780 + }, + { + "epoch": 17.625794864326974, + "grad_norm": 0.22029165923595428, + "learning_rate": 2.3016834611997773e-06, + "loss": 0.0228, + "step": 90790 + }, + { + "epoch": 17.627736517644774, + "grad_norm": 0.1634422242641449, + "learning_rate": 2.296727086550665e-06, + "loss": 0.0202, + "step": 90800 + }, + { + "epoch": 17.629678170962574, + "grad_norm": 0.1970043033361435, + "learning_rate": 2.2917759286597507e-06, + "loss": 0.0289, + "step": 90810 + }, + { + "epoch": 17.631619824280374, + "grad_norm": 0.0847478061914444, + "learning_rate": 2.2868299880685006e-06, + "loss": 0.0257, + "step": 90820 + }, + { + "epoch": 17.633561477598175, + "grad_norm": 0.2601569890975952, + "learning_rate": 2.2818892653177827e-06, + "loss": 0.0217, + "step": 90830 + }, + { + "epoch": 17.635503130915975, + "grad_norm": 0.1653265506029129, + "learning_rate": 2.2769537609479196e-06, + "loss": 0.0258, + "step": 90840 + }, + { + "epoch": 17.637444784233775, + "grad_norm": 0.1826557070016861, + "learning_rate": 2.2720234754986357e-06, + "loss": 0.0227, + "step": 90850 + }, + { + "epoch": 17.639386437551575, + "grad_norm": 0.1969342976808548, + "learning_rate": 2.2670984095091108e-06, + "loss": 0.0248, + "step": 90860 + }, + { + "epoch": 17.641328090869376, + "grad_norm": 0.18289217352867126, + "learning_rate": 2.262178563517936e-06, + "loss": 0.0246, + "step": 90870 + }, + { + "epoch": 17.643269744187176, + "grad_norm": 0.18098001182079315, + "learning_rate": 2.257263938063137e-06, + "loss": 0.0193, + "step": 90880 + }, + { + "epoch": 17.645211397504976, + "grad_norm": 0.20683981478214264, + "learning_rate": 2.2523545336821715e-06, + "loss": 0.027, + "step": 90890 + }, + { + "epoch": 17.647153050822777, + "grad_norm": 0.1866675168275833, + "learning_rate": 2.247450350911939e-06, + "loss": 0.0217, + "step": 90900 + }, + { + "epoch": 17.649094704140577, + "grad_norm": 0.1907641887664795, + "learning_rate": 2.2425513902887328e-06, + "loss": 0.0235, + "step": 90910 + }, + { + "epoch": 17.651036357458377, + "grad_norm": 0.1182556226849556, + "learning_rate": 2.2376576523483015e-06, + "loss": 0.0194, + "step": 90920 + }, + { + "epoch": 17.652978010776177, + "grad_norm": 0.19157317280769348, + "learning_rate": 2.232769137625829e-06, + "loss": 0.0221, + "step": 90930 + }, + { + "epoch": 17.654919664093978, + "grad_norm": 0.14467574656009674, + "learning_rate": 2.227885846655903e-06, + "loss": 0.0204, + "step": 90940 + }, + { + "epoch": 17.656861317411774, + "grad_norm": 0.14991630613803864, + "learning_rate": 2.2230077799725523e-06, + "loss": 0.0162, + "step": 90950 + }, + { + "epoch": 17.658802970729575, + "grad_norm": 0.17423173785209656, + "learning_rate": 2.2181349381092396e-06, + "loss": 0.0245, + "step": 90960 + }, + { + "epoch": 17.660744624047375, + "grad_norm": 0.1842467337846756, + "learning_rate": 2.2132673215988552e-06, + "loss": 0.0246, + "step": 90970 + }, + { + "epoch": 17.662686277365175, + "grad_norm": 0.2476613074541092, + "learning_rate": 2.2084049309737065e-06, + "loss": 0.0208, + "step": 90980 + }, + { + "epoch": 17.664627930682975, + "grad_norm": 0.18695390224456787, + "learning_rate": 2.2035477667655513e-06, + "loss": 0.0197, + "step": 90990 + }, + { + "epoch": 17.666569584000776, + "grad_norm": 0.18067441880702972, + "learning_rate": 2.1986958295055315e-06, + "loss": 0.0235, + "step": 91000 + }, + { + "epoch": 17.668511237318576, + "grad_norm": 0.17629045248031616, + "learning_rate": 2.193849119724284e-06, + "loss": 0.023, + "step": 91010 + }, + { + "epoch": 17.670452890636376, + "grad_norm": 0.18816429376602173, + "learning_rate": 2.1890076379518175e-06, + "loss": 0.0275, + "step": 91020 + }, + { + "epoch": 17.672394543954177, + "grad_norm": 0.13262715935707092, + "learning_rate": 2.1841713847175925e-06, + "loss": 0.0222, + "step": 91030 + }, + { + "epoch": 17.674336197271977, + "grad_norm": 0.1368056684732437, + "learning_rate": 2.1793403605504968e-06, + "loss": 0.0236, + "step": 91040 + }, + { + "epoch": 17.676277850589777, + "grad_norm": 0.15212634205818176, + "learning_rate": 2.1745145659788413e-06, + "loss": 0.019, + "step": 91050 + }, + { + "epoch": 17.678219503907577, + "grad_norm": 0.19312675297260284, + "learning_rate": 2.1696940015303645e-06, + "loss": 0.022, + "step": 91060 + }, + { + "epoch": 17.680161157225378, + "grad_norm": 0.236393004655838, + "learning_rate": 2.1648786677322397e-06, + "loss": 0.0181, + "step": 91070 + }, + { + "epoch": 17.682102810543178, + "grad_norm": 0.18134890496730804, + "learning_rate": 2.1600685651110673e-06, + "loss": 0.0278, + "step": 91080 + }, + { + "epoch": 17.684044463860978, + "grad_norm": 0.17089204490184784, + "learning_rate": 2.1552636941928717e-06, + "loss": 0.0186, + "step": 91090 + }, + { + "epoch": 17.68598611717878, + "grad_norm": 0.22210733592510223, + "learning_rate": 2.1504640555031043e-06, + "loss": 0.0314, + "step": 91100 + }, + { + "epoch": 17.68792777049658, + "grad_norm": 0.12755729258060455, + "learning_rate": 2.1456696495666506e-06, + "loss": 0.0216, + "step": 91110 + }, + { + "epoch": 17.68986942381438, + "grad_norm": 0.17968140542507172, + "learning_rate": 2.1408804769078084e-06, + "loss": 0.0199, + "step": 91120 + }, + { + "epoch": 17.69181107713218, + "grad_norm": 0.1855476051568985, + "learning_rate": 2.136096538050325e-06, + "loss": 0.0192, + "step": 91130 + }, + { + "epoch": 17.69375273044998, + "grad_norm": 0.1213473379611969, + "learning_rate": 2.13131783351736e-06, + "loss": 0.0219, + "step": 91140 + }, + { + "epoch": 17.69569438376778, + "grad_norm": 0.18365077674388885, + "learning_rate": 2.126544363831512e-06, + "loss": 0.0178, + "step": 91150 + }, + { + "epoch": 17.69763603708558, + "grad_norm": 0.23753754794597626, + "learning_rate": 2.121776129514791e-06, + "loss": 0.0245, + "step": 91160 + }, + { + "epoch": 17.699577690403377, + "grad_norm": 0.14891132712364197, + "learning_rate": 2.1170131310886523e-06, + "loss": 0.0252, + "step": 91170 + }, + { + "epoch": 17.701519343721177, + "grad_norm": 0.12846103310585022, + "learning_rate": 2.112255369073968e-06, + "loss": 0.0196, + "step": 91180 + }, + { + "epoch": 17.703460997038977, + "grad_norm": 0.11054282635450363, + "learning_rate": 2.1075028439910328e-06, + "loss": 0.0206, + "step": 91190 + }, + { + "epoch": 17.705402650356778, + "grad_norm": 0.15208758413791656, + "learning_rate": 2.1027555563595813e-06, + "loss": 0.0182, + "step": 91200 + }, + { + "epoch": 17.707344303674578, + "grad_norm": 0.1407657116651535, + "learning_rate": 2.0980135066987706e-06, + "loss": 0.0134, + "step": 91210 + }, + { + "epoch": 17.709285956992378, + "grad_norm": 0.35381433367729187, + "learning_rate": 2.093276695527191e-06, + "loss": 0.0215, + "step": 91220 + }, + { + "epoch": 17.71122761031018, + "grad_norm": 0.29123538732528687, + "learning_rate": 2.0885451233628396e-06, + "loss": 0.0189, + "step": 91230 + }, + { + "epoch": 17.71316926362798, + "grad_norm": 0.12824440002441406, + "learning_rate": 2.0838187907231633e-06, + "loss": 0.0218, + "step": 91240 + }, + { + "epoch": 17.71511091694578, + "grad_norm": 0.19400262832641602, + "learning_rate": 2.0790976981250153e-06, + "loss": 0.0235, + "step": 91250 + }, + { + "epoch": 17.71705257026358, + "grad_norm": 0.2003626972436905, + "learning_rate": 2.0743818460847e-06, + "loss": 0.0274, + "step": 91260 + }, + { + "epoch": 17.71899422358138, + "grad_norm": 0.10098603367805481, + "learning_rate": 2.069671235117926e-06, + "loss": 0.0165, + "step": 91270 + }, + { + "epoch": 17.72093587689918, + "grad_norm": 0.09990088641643524, + "learning_rate": 2.0649658657398484e-06, + "loss": 0.0175, + "step": 91280 + }, + { + "epoch": 17.72287753021698, + "grad_norm": 0.17074887454509735, + "learning_rate": 2.060265738465034e-06, + "loss": 0.0207, + "step": 91290 + }, + { + "epoch": 17.72481918353478, + "grad_norm": 0.11236733198165894, + "learning_rate": 2.055570853807487e-06, + "loss": 0.014, + "step": 91300 + }, + { + "epoch": 17.72676083685258, + "grad_norm": 0.1709698736667633, + "learning_rate": 2.0508812122806196e-06, + "loss": 0.0165, + "step": 91310 + }, + { + "epoch": 17.72870249017038, + "grad_norm": 0.25990211963653564, + "learning_rate": 2.046196814397294e-06, + "loss": 0.022, + "step": 91320 + }, + { + "epoch": 17.73064414348818, + "grad_norm": 0.2697608172893524, + "learning_rate": 2.0415176606697896e-06, + "loss": 0.0234, + "step": 91330 + }, + { + "epoch": 17.73258579680598, + "grad_norm": 0.09183743596076965, + "learning_rate": 2.0368437516098126e-06, + "loss": 0.0148, + "step": 91340 + }, + { + "epoch": 17.73452745012378, + "grad_norm": 0.23138831555843353, + "learning_rate": 2.032175087728483e-06, + "loss": 0.0215, + "step": 91350 + }, + { + "epoch": 17.736469103441582, + "grad_norm": 0.11116928607225418, + "learning_rate": 2.027511669536375e-06, + "loss": 0.0198, + "step": 91360 + }, + { + "epoch": 17.738410756759382, + "grad_norm": 0.13762101531028748, + "learning_rate": 2.0228534975434643e-06, + "loss": 0.0211, + "step": 91370 + }, + { + "epoch": 17.74035241007718, + "grad_norm": 0.1624533236026764, + "learning_rate": 2.0182005722591548e-06, + "loss": 0.0174, + "step": 91380 + }, + { + "epoch": 17.74229406339498, + "grad_norm": 0.1906057447195053, + "learning_rate": 2.0135528941922944e-06, + "loss": 0.0221, + "step": 91390 + }, + { + "epoch": 17.74423571671278, + "grad_norm": 0.14238838851451874, + "learning_rate": 2.008910463851149e-06, + "loss": 0.0252, + "step": 91400 + }, + { + "epoch": 17.74617737003058, + "grad_norm": 0.1039675623178482, + "learning_rate": 2.0042732817433963e-06, + "loss": 0.0165, + "step": 91410 + }, + { + "epoch": 17.74811902334838, + "grad_norm": 0.1816033124923706, + "learning_rate": 1.999641348376158e-06, + "loss": 0.0245, + "step": 91420 + }, + { + "epoch": 17.75006067666618, + "grad_norm": 0.14349646866321564, + "learning_rate": 1.9950146642559787e-06, + "loss": 0.0213, + "step": 91430 + }, + { + "epoch": 17.75200232998398, + "grad_norm": 0.1021638885140419, + "learning_rate": 1.990393229888815e-06, + "loss": 0.0242, + "step": 91440 + }, + { + "epoch": 17.75394398330178, + "grad_norm": 0.19128775596618652, + "learning_rate": 1.9857770457800684e-06, + "loss": 0.0209, + "step": 91450 + }, + { + "epoch": 17.75588563661958, + "grad_norm": 0.1676088571548462, + "learning_rate": 1.9811661124345517e-06, + "loss": 0.0196, + "step": 91460 + }, + { + "epoch": 17.75782728993738, + "grad_norm": 0.09821417927742004, + "learning_rate": 1.976560430356522e-06, + "loss": 0.0228, + "step": 91470 + }, + { + "epoch": 17.75976894325518, + "grad_norm": 0.2087513953447342, + "learning_rate": 1.9719600000496383e-06, + "loss": 0.0237, + "step": 91480 + }, + { + "epoch": 17.761710596572982, + "grad_norm": 0.1906522810459137, + "learning_rate": 1.967364822017004e-06, + "loss": 0.0239, + "step": 91490 + }, + { + "epoch": 17.763652249890782, + "grad_norm": 0.3296050429344177, + "learning_rate": 1.962774896761127e-06, + "loss": 0.0202, + "step": 91500 + }, + { + "epoch": 17.765593903208583, + "grad_norm": 0.2198701649904251, + "learning_rate": 1.958190224783968e-06, + "loss": 0.024, + "step": 91510 + }, + { + "epoch": 17.767535556526383, + "grad_norm": 0.22489960491657257, + "learning_rate": 1.9536108065868975e-06, + "loss": 0.0255, + "step": 91520 + }, + { + "epoch": 17.769477209844183, + "grad_norm": 0.1085398867726326, + "learning_rate": 1.949036642670715e-06, + "loss": 0.0227, + "step": 91530 + }, + { + "epoch": 17.771418863161983, + "grad_norm": 0.22628557682037354, + "learning_rate": 1.9444677335356377e-06, + "loss": 0.0174, + "step": 91540 + }, + { + "epoch": 17.773360516479784, + "grad_norm": 0.22945702075958252, + "learning_rate": 1.939904079681326e-06, + "loss": 0.0215, + "step": 91550 + }, + { + "epoch": 17.775302169797584, + "grad_norm": 0.16471923887729645, + "learning_rate": 1.935345681606837e-06, + "loss": 0.0267, + "step": 91560 + }, + { + "epoch": 17.777243823115384, + "grad_norm": 0.09338859468698502, + "learning_rate": 1.930792539810694e-06, + "loss": 0.0181, + "step": 91570 + }, + { + "epoch": 17.779185476433184, + "grad_norm": 0.24578005075454712, + "learning_rate": 1.9262446547907877e-06, + "loss": 0.0232, + "step": 91580 + }, + { + "epoch": 17.781127129750985, + "grad_norm": 0.27349576354026794, + "learning_rate": 1.921702027044503e-06, + "loss": 0.0229, + "step": 91590 + }, + { + "epoch": 17.78306878306878, + "grad_norm": 0.2047071009874344, + "learning_rate": 1.917164657068593e-06, + "loss": 0.0256, + "step": 91600 + }, + { + "epoch": 17.78501043638658, + "grad_norm": 0.1222568228840828, + "learning_rate": 1.912632545359272e-06, + "loss": 0.0248, + "step": 91610 + }, + { + "epoch": 17.786952089704382, + "grad_norm": 0.08336471021175385, + "learning_rate": 1.9081056924121488e-06, + "loss": 0.0155, + "step": 91620 + }, + { + "epoch": 17.788893743022182, + "grad_norm": 0.10260267555713654, + "learning_rate": 1.9035840987222887e-06, + "loss": 0.0253, + "step": 91630 + }, + { + "epoch": 17.790835396339983, + "grad_norm": 0.148038849234581, + "learning_rate": 1.8990677647841515e-06, + "loss": 0.0212, + "step": 91640 + }, + { + "epoch": 17.792777049657783, + "grad_norm": 0.08810494840145111, + "learning_rate": 1.8945566910916534e-06, + "loss": 0.0238, + "step": 91650 + }, + { + "epoch": 17.794718702975583, + "grad_norm": 0.12622995674610138, + "learning_rate": 1.8900508781381054e-06, + "loss": 0.0181, + "step": 91660 + }, + { + "epoch": 17.796660356293383, + "grad_norm": 0.09952015429735184, + "learning_rate": 1.8855503264162633e-06, + "loss": 0.0243, + "step": 91670 + }, + { + "epoch": 17.798602009611184, + "grad_norm": 0.15381307899951935, + "learning_rate": 1.8810550364182998e-06, + "loss": 0.0221, + "step": 91680 + }, + { + "epoch": 17.800543662928984, + "grad_norm": 0.30591732263565063, + "learning_rate": 1.8765650086358112e-06, + "loss": 0.0253, + "step": 91690 + }, + { + "epoch": 17.802485316246784, + "grad_norm": 0.18196111917495728, + "learning_rate": 1.8720802435598151e-06, + "loss": 0.0205, + "step": 91700 + }, + { + "epoch": 17.804426969564584, + "grad_norm": 0.2684709131717682, + "learning_rate": 1.8676007416807694e-06, + "loss": 0.0198, + "step": 91710 + }, + { + "epoch": 17.806368622882385, + "grad_norm": 0.18127207458019257, + "learning_rate": 1.8631265034885437e-06, + "loss": 0.0256, + "step": 91720 + }, + { + "epoch": 17.808310276200185, + "grad_norm": 0.15422861278057098, + "learning_rate": 1.858657529472424e-06, + "loss": 0.0324, + "step": 91730 + }, + { + "epoch": 17.810251929517985, + "grad_norm": 0.320080041885376, + "learning_rate": 1.8541938201211473e-06, + "loss": 0.0181, + "step": 91740 + }, + { + "epoch": 17.812193582835786, + "grad_norm": 0.38605543971061707, + "learning_rate": 1.8497353759228398e-06, + "loss": 0.0275, + "step": 91750 + }, + { + "epoch": 17.814135236153586, + "grad_norm": 0.07649330049753189, + "learning_rate": 1.8452821973650835e-06, + "loss": 0.0202, + "step": 91760 + }, + { + "epoch": 17.816076889471386, + "grad_norm": 0.16691771149635315, + "learning_rate": 1.840834284934856e-06, + "loss": 0.0172, + "step": 91770 + }, + { + "epoch": 17.818018542789186, + "grad_norm": 0.20289163291454315, + "learning_rate": 1.8363916391185954e-06, + "loss": 0.0195, + "step": 91780 + }, + { + "epoch": 17.819960196106987, + "grad_norm": 0.10157956928014755, + "learning_rate": 1.8319542604021245e-06, + "loss": 0.0216, + "step": 91790 + }, + { + "epoch": 17.821901849424783, + "grad_norm": 0.11254288256168365, + "learning_rate": 1.8275221492707273e-06, + "loss": 0.0189, + "step": 91800 + }, + { + "epoch": 17.823843502742584, + "grad_norm": 0.16789260506629944, + "learning_rate": 1.8230953062090717e-06, + "loss": 0.0199, + "step": 91810 + }, + { + "epoch": 17.825785156060384, + "grad_norm": 0.24561674892902374, + "learning_rate": 1.8186737317012926e-06, + "loss": 0.0269, + "step": 91820 + }, + { + "epoch": 17.827726809378184, + "grad_norm": 0.15262651443481445, + "learning_rate": 1.8142574262308977e-06, + "loss": 0.0161, + "step": 91830 + }, + { + "epoch": 17.829668462695984, + "grad_norm": 0.1802464723587036, + "learning_rate": 1.8098463902808727e-06, + "loss": 0.0172, + "step": 91840 + }, + { + "epoch": 17.831610116013785, + "grad_norm": 0.09430526942014694, + "learning_rate": 1.805440624333593e-06, + "loss": 0.0218, + "step": 91850 + }, + { + "epoch": 17.833551769331585, + "grad_norm": 0.14556066691875458, + "learning_rate": 1.8010401288708734e-06, + "loss": 0.0205, + "step": 91860 + }, + { + "epoch": 17.835493422649385, + "grad_norm": 0.11980893462896347, + "learning_rate": 1.7966449043739343e-06, + "loss": 0.0169, + "step": 91870 + }, + { + "epoch": 17.837435075967186, + "grad_norm": 0.1248927041888237, + "learning_rate": 1.79225495132343e-06, + "loss": 0.0221, + "step": 91880 + }, + { + "epoch": 17.839376729284986, + "grad_norm": 0.22156992554664612, + "learning_rate": 1.7878702701994487e-06, + "loss": 0.0175, + "step": 91890 + }, + { + "epoch": 17.841318382602786, + "grad_norm": 0.1625455766916275, + "learning_rate": 1.783490861481496e-06, + "loss": 0.0233, + "step": 91900 + }, + { + "epoch": 17.843260035920586, + "grad_norm": 0.1417047679424286, + "learning_rate": 1.7791167256484832e-06, + "loss": 0.0207, + "step": 91910 + }, + { + "epoch": 17.845201689238387, + "grad_norm": 0.18476618826389313, + "learning_rate": 1.774747863178766e-06, + "loss": 0.0252, + "step": 91920 + }, + { + "epoch": 17.847143342556187, + "grad_norm": 0.07006014138460159, + "learning_rate": 1.7703842745501186e-06, + "loss": 0.0219, + "step": 91930 + }, + { + "epoch": 17.849084995873987, + "grad_norm": 0.17586162686347961, + "learning_rate": 1.7660259602397366e-06, + "loss": 0.0206, + "step": 91940 + }, + { + "epoch": 17.851026649191787, + "grad_norm": 0.2166103571653366, + "learning_rate": 1.7616729207242388e-06, + "loss": 0.0229, + "step": 91950 + }, + { + "epoch": 17.852968302509588, + "grad_norm": 0.0888141617178917, + "learning_rate": 1.7573251564796556e-06, + "loss": 0.0232, + "step": 91960 + }, + { + "epoch": 17.854909955827388, + "grad_norm": 0.19565589725971222, + "learning_rate": 1.7529826679814676e-06, + "loss": 0.0155, + "step": 91970 + }, + { + "epoch": 17.85685160914519, + "grad_norm": 0.2027096301317215, + "learning_rate": 1.7486454557045562e-06, + "loss": 0.0269, + "step": 91980 + }, + { + "epoch": 17.85879326246299, + "grad_norm": 0.12802453339099884, + "learning_rate": 1.7443135201232364e-06, + "loss": 0.0225, + "step": 91990 + }, + { + "epoch": 17.86073491578079, + "grad_norm": 0.1792043298482895, + "learning_rate": 1.7399868617112347e-06, + "loss": 0.0188, + "step": 92000 + }, + { + "epoch": 17.86267656909859, + "grad_norm": 0.17554526031017303, + "learning_rate": 1.7356654809417171e-06, + "loss": 0.0207, + "step": 92010 + }, + { + "epoch": 17.864618222416386, + "grad_norm": 0.16608570516109467, + "learning_rate": 1.7313493782872448e-06, + "loss": 0.0211, + "step": 92020 + }, + { + "epoch": 17.866559875734186, + "grad_norm": 0.20949678122997284, + "learning_rate": 1.7270385542198453e-06, + "loss": 0.0243, + "step": 92030 + }, + { + "epoch": 17.868501529051986, + "grad_norm": 0.15626895427703857, + "learning_rate": 1.7227330092109307e-06, + "loss": 0.0186, + "step": 92040 + }, + { + "epoch": 17.870443182369787, + "grad_norm": 0.16423286497592926, + "learning_rate": 1.718432743731352e-06, + "loss": 0.0216, + "step": 92050 + }, + { + "epoch": 17.872384835687587, + "grad_norm": 0.14603543281555176, + "learning_rate": 1.7141377582513718e-06, + "loss": 0.0182, + "step": 92060 + }, + { + "epoch": 17.874326489005387, + "grad_norm": 0.2512543499469757, + "learning_rate": 1.709848053240698e-06, + "loss": 0.0266, + "step": 92070 + }, + { + "epoch": 17.876268142323188, + "grad_norm": 0.11424887925386429, + "learning_rate": 1.7055636291684328e-06, + "loss": 0.0223, + "step": 92080 + }, + { + "epoch": 17.878209795640988, + "grad_norm": 0.19999806582927704, + "learning_rate": 1.7012844865031185e-06, + "loss": 0.0173, + "step": 92090 + }, + { + "epoch": 17.880151448958788, + "grad_norm": 0.21276246011257172, + "learning_rate": 1.6970106257127138e-06, + "loss": 0.0231, + "step": 92100 + }, + { + "epoch": 17.88209310227659, + "grad_norm": 0.17461538314819336, + "learning_rate": 1.6927420472646117e-06, + "loss": 0.0255, + "step": 92110 + }, + { + "epoch": 17.88403475559439, + "grad_norm": 0.11133269220590591, + "learning_rate": 1.6884787516256106e-06, + "loss": 0.0238, + "step": 92120 + }, + { + "epoch": 17.88597640891219, + "grad_norm": 0.1706024706363678, + "learning_rate": 1.6842207392619325e-06, + "loss": 0.0287, + "step": 92130 + }, + { + "epoch": 17.88791806222999, + "grad_norm": 0.1394994705915451, + "learning_rate": 1.679968010639238e-06, + "loss": 0.0198, + "step": 92140 + }, + { + "epoch": 17.88985971554779, + "grad_norm": 0.21746434271335602, + "learning_rate": 1.6757205662225829e-06, + "loss": 0.023, + "step": 92150 + }, + { + "epoch": 17.89180136886559, + "grad_norm": 0.2108192890882492, + "learning_rate": 1.6714784064764844e-06, + "loss": 0.0185, + "step": 92160 + }, + { + "epoch": 17.89374302218339, + "grad_norm": 0.2863318622112274, + "learning_rate": 1.6672415318648383e-06, + "loss": 0.0222, + "step": 92170 + }, + { + "epoch": 17.89568467550119, + "grad_norm": 0.16454371809959412, + "learning_rate": 1.6630099428510016e-06, + "loss": 0.024, + "step": 92180 + }, + { + "epoch": 17.89762632881899, + "grad_norm": 0.20121292769908905, + "learning_rate": 1.6587836398977153e-06, + "loss": 0.0265, + "step": 92190 + }, + { + "epoch": 17.89956798213679, + "grad_norm": 0.22644449770450592, + "learning_rate": 1.6545626234671762e-06, + "loss": 0.0233, + "step": 92200 + }, + { + "epoch": 17.90150963545459, + "grad_norm": 0.21343958377838135, + "learning_rate": 1.650346894020971e-06, + "loss": 0.0204, + "step": 92210 + }, + { + "epoch": 17.90345128877239, + "grad_norm": 0.2305063009262085, + "learning_rate": 1.6461364520201472e-06, + "loss": 0.0203, + "step": 92220 + }, + { + "epoch": 17.905392942090188, + "grad_norm": 0.09986432641744614, + "learning_rate": 1.6419312979251366e-06, + "loss": 0.0193, + "step": 92230 + }, + { + "epoch": 17.90733459540799, + "grad_norm": 0.18113034963607788, + "learning_rate": 1.6377314321958214e-06, + "loss": 0.0207, + "step": 92240 + }, + { + "epoch": 17.90927624872579, + "grad_norm": 0.1713102012872696, + "learning_rate": 1.6335368552914787e-06, + "loss": 0.0189, + "step": 92250 + }, + { + "epoch": 17.91121790204359, + "grad_norm": 0.11289876699447632, + "learning_rate": 1.6293475676708358e-06, + "loss": 0.0221, + "step": 92260 + }, + { + "epoch": 17.91315955536139, + "grad_norm": 0.13383667171001434, + "learning_rate": 1.6251635697920097e-06, + "loss": 0.0204, + "step": 92270 + }, + { + "epoch": 17.91510120867919, + "grad_norm": 0.06286497414112091, + "learning_rate": 1.6209848621125624e-06, + "loss": 0.0201, + "step": 92280 + }, + { + "epoch": 17.91704286199699, + "grad_norm": 0.2043876200914383, + "learning_rate": 1.6168114450894778e-06, + "loss": 0.0182, + "step": 92290 + }, + { + "epoch": 17.91898451531479, + "grad_norm": 0.14808747172355652, + "learning_rate": 1.612643319179158e-06, + "loss": 0.0186, + "step": 92300 + }, + { + "epoch": 17.92092616863259, + "grad_norm": 0.05560484156012535, + "learning_rate": 1.6084804848374046e-06, + "loss": 0.0156, + "step": 92310 + }, + { + "epoch": 17.92286782195039, + "grad_norm": 0.2638534903526306, + "learning_rate": 1.6043229425194817e-06, + "loss": 0.023, + "step": 92320 + }, + { + "epoch": 17.92480947526819, + "grad_norm": 0.1761423796415329, + "learning_rate": 1.6001706926800308e-06, + "loss": 0.0239, + "step": 92330 + }, + { + "epoch": 17.92675112858599, + "grad_norm": 0.1499510109424591, + "learning_rate": 1.5960237357731444e-06, + "loss": 0.0252, + "step": 92340 + }, + { + "epoch": 17.92869278190379, + "grad_norm": 0.2943415343761444, + "learning_rate": 1.5918820722523264e-06, + "loss": 0.0237, + "step": 92350 + }, + { + "epoch": 17.93063443522159, + "grad_norm": 0.25589093565940857, + "learning_rate": 1.587745702570509e-06, + "loss": 0.0156, + "step": 92360 + }, + { + "epoch": 17.932576088539392, + "grad_norm": 0.25440940260887146, + "learning_rate": 1.5836146271800356e-06, + "loss": 0.0222, + "step": 92370 + }, + { + "epoch": 17.934517741857192, + "grad_norm": 0.2637893259525299, + "learning_rate": 1.5794888465326728e-06, + "loss": 0.0214, + "step": 92380 + }, + { + "epoch": 17.936459395174992, + "grad_norm": 0.051139216870069504, + "learning_rate": 1.5753683610796155e-06, + "loss": 0.0187, + "step": 92390 + }, + { + "epoch": 17.938401048492793, + "grad_norm": 0.2909650206565857, + "learning_rate": 1.5712531712714528e-06, + "loss": 0.0243, + "step": 92400 + }, + { + "epoch": 17.940342701810593, + "grad_norm": 0.17345957458019257, + "learning_rate": 1.5671432775582474e-06, + "loss": 0.0221, + "step": 92410 + }, + { + "epoch": 17.942284355128393, + "grad_norm": 0.20546965301036835, + "learning_rate": 1.563038680389428e-06, + "loss": 0.0278, + "step": 92420 + }, + { + "epoch": 17.944226008446194, + "grad_norm": 0.2280532270669937, + "learning_rate": 1.5589393802138808e-06, + "loss": 0.0261, + "step": 92430 + }, + { + "epoch": 17.946167661763994, + "grad_norm": 0.19827650487422943, + "learning_rate": 1.554845377479891e-06, + "loss": 0.0254, + "step": 92440 + }, + { + "epoch": 17.94810931508179, + "grad_norm": 0.08595892786979675, + "learning_rate": 1.550756672635173e-06, + "loss": 0.0206, + "step": 92450 + }, + { + "epoch": 17.95005096839959, + "grad_norm": 0.16981296241283417, + "learning_rate": 1.5466732661268635e-06, + "loss": 0.0201, + "step": 92460 + }, + { + "epoch": 17.95199262171739, + "grad_norm": 0.2576039731502533, + "learning_rate": 1.5425951584015163e-06, + "loss": 0.0257, + "step": 92470 + }, + { + "epoch": 17.95393427503519, + "grad_norm": 0.1417664885520935, + "learning_rate": 1.5385223499051026e-06, + "loss": 0.0298, + "step": 92480 + }, + { + "epoch": 17.95587592835299, + "grad_norm": 0.22588299214839935, + "learning_rate": 1.5344548410830328e-06, + "loss": 0.0188, + "step": 92490 + }, + { + "epoch": 17.957817581670792, + "grad_norm": 0.1033327728509903, + "learning_rate": 1.5303926323801121e-06, + "loss": 0.024, + "step": 92500 + }, + { + "epoch": 17.959759234988592, + "grad_norm": 0.17124596238136292, + "learning_rate": 1.5263357242405795e-06, + "loss": 0.0225, + "step": 92510 + }, + { + "epoch": 17.961700888306392, + "grad_norm": 0.20322830975055695, + "learning_rate": 1.5222841171080915e-06, + "loss": 0.0285, + "step": 92520 + }, + { + "epoch": 17.963642541624193, + "grad_norm": 0.17587362229824066, + "learning_rate": 1.5182378114257268e-06, + "loss": 0.0188, + "step": 92530 + }, + { + "epoch": 17.965584194941993, + "grad_norm": 0.1576639711856842, + "learning_rate": 1.5141968076359869e-06, + "loss": 0.0202, + "step": 92540 + }, + { + "epoch": 17.967525848259793, + "grad_norm": 0.23397211730480194, + "learning_rate": 1.5101611061807908e-06, + "loss": 0.0168, + "step": 92550 + }, + { + "epoch": 17.969467501577594, + "grad_norm": 0.21108609437942505, + "learning_rate": 1.506130707501463e-06, + "loss": 0.0215, + "step": 92560 + }, + { + "epoch": 17.971409154895394, + "grad_norm": 0.13427239656448364, + "learning_rate": 1.5021056120387734e-06, + "loss": 0.0181, + "step": 92570 + }, + { + "epoch": 17.973350808213194, + "grad_norm": 0.21666856110095978, + "learning_rate": 1.4980858202329084e-06, + "loss": 0.0209, + "step": 92580 + }, + { + "epoch": 17.975292461530994, + "grad_norm": 0.20672845840454102, + "learning_rate": 1.4940713325234445e-06, + "loss": 0.0269, + "step": 92590 + }, + { + "epoch": 17.977234114848795, + "grad_norm": 0.20577529072761536, + "learning_rate": 1.4900621493494137e-06, + "loss": 0.0208, + "step": 92600 + }, + { + "epoch": 17.979175768166595, + "grad_norm": 0.16833078861236572, + "learning_rate": 1.4860582711492544e-06, + "loss": 0.0213, + "step": 92610 + }, + { + "epoch": 17.981117421484395, + "grad_norm": 0.28450673818588257, + "learning_rate": 1.4820596983608325e-06, + "loss": 0.0263, + "step": 92620 + }, + { + "epoch": 17.983059074802195, + "grad_norm": 0.1560310423374176, + "learning_rate": 1.4780664314214044e-06, + "loss": 0.0261, + "step": 92630 + }, + { + "epoch": 17.985000728119996, + "grad_norm": 0.142457976937294, + "learning_rate": 1.474078470767687e-06, + "loss": 0.0215, + "step": 92640 + }, + { + "epoch": 17.986942381437796, + "grad_norm": 0.1514746993780136, + "learning_rate": 1.470095816835787e-06, + "loss": 0.0194, + "step": 92650 + }, + { + "epoch": 17.988884034755593, + "grad_norm": 0.15792421996593475, + "learning_rate": 1.4661184700612395e-06, + "loss": 0.0258, + "step": 92660 + }, + { + "epoch": 17.990825688073393, + "grad_norm": 0.19855685532093048, + "learning_rate": 1.4621464308790133e-06, + "loss": 0.0179, + "step": 92670 + }, + { + "epoch": 17.992767341391193, + "grad_norm": 0.2169361710548401, + "learning_rate": 1.4581796997234775e-06, + "loss": 0.0193, + "step": 92680 + }, + { + "epoch": 17.994708994708994, + "grad_norm": 0.16297252476215363, + "learning_rate": 1.4542182770284297e-06, + "loss": 0.0244, + "step": 92690 + }, + { + "epoch": 17.996650648026794, + "grad_norm": 0.1884971708059311, + "learning_rate": 1.450262163227084e-06, + "loss": 0.0227, + "step": 92700 + }, + { + "epoch": 17.998592301344594, + "grad_norm": 0.2617066502571106, + "learning_rate": 1.4463113587520671e-06, + "loss": 0.0215, + "step": 92710 + }, + { + "epoch": 18.00038833066356, + "grad_norm": 0.22338196635246277, + "learning_rate": 1.4423658640354443e-06, + "loss": 0.0226, + "step": 92720 + }, + { + "epoch": 18.00232998398136, + "grad_norm": 0.14809902012348175, + "learning_rate": 1.438425679508687e-06, + "loss": 0.0193, + "step": 92730 + }, + { + "epoch": 18.00427163729916, + "grad_norm": 0.13416798412799835, + "learning_rate": 1.4344908056026895e-06, + "loss": 0.0192, + "step": 92740 + }, + { + "epoch": 18.00621329061696, + "grad_norm": 0.18149201571941376, + "learning_rate": 1.4305612427477521e-06, + "loss": 0.0221, + "step": 92750 + }, + { + "epoch": 18.00815494393476, + "grad_norm": 0.18591725826263428, + "learning_rate": 1.42663699137362e-06, + "loss": 0.0194, + "step": 92760 + }, + { + "epoch": 18.01009659725256, + "grad_norm": 0.29387524724006653, + "learning_rate": 1.4227180519094386e-06, + "loss": 0.0265, + "step": 92770 + }, + { + "epoch": 18.01203825057036, + "grad_norm": 0.2164984941482544, + "learning_rate": 1.4188044247837706e-06, + "loss": 0.0228, + "step": 92780 + }, + { + "epoch": 18.01397990388816, + "grad_norm": 0.2762722969055176, + "learning_rate": 1.4148961104246128e-06, + "loss": 0.0231, + "step": 92790 + }, + { + "epoch": 18.01592155720596, + "grad_norm": 0.17575222253799438, + "learning_rate": 1.4109931092593731e-06, + "loss": 0.0159, + "step": 92800 + }, + { + "epoch": 18.01786321052376, + "grad_norm": 0.29366007447242737, + "learning_rate": 1.4070954217148658e-06, + "loss": 0.0195, + "step": 92810 + }, + { + "epoch": 18.01980486384156, + "grad_norm": 0.14358896017074585, + "learning_rate": 1.4032030482173497e-06, + "loss": 0.0171, + "step": 92820 + }, + { + "epoch": 18.02174651715936, + "grad_norm": 0.2089298814535141, + "learning_rate": 1.3993159891924901e-06, + "loss": 0.0191, + "step": 92830 + }, + { + "epoch": 18.02368817047716, + "grad_norm": 0.15424567461013794, + "learning_rate": 1.3954342450653524e-06, + "loss": 0.0163, + "step": 92840 + }, + { + "epoch": 18.02562982379496, + "grad_norm": 0.2639167308807373, + "learning_rate": 1.3915578162604582e-06, + "loss": 0.0161, + "step": 92850 + }, + { + "epoch": 18.02757147711276, + "grad_norm": 0.12218558043241501, + "learning_rate": 1.3876867032017126e-06, + "loss": 0.0195, + "step": 92860 + }, + { + "epoch": 18.029513130430562, + "grad_norm": 0.17183975875377655, + "learning_rate": 1.383820906312472e-06, + "loss": 0.0198, + "step": 92870 + }, + { + "epoch": 18.031454783748362, + "grad_norm": 0.22836236655712128, + "learning_rate": 1.3799604260154752e-06, + "loss": 0.0228, + "step": 92880 + }, + { + "epoch": 18.033396437066163, + "grad_norm": 0.12755876779556274, + "learning_rate": 1.376105262732913e-06, + "loss": 0.024, + "step": 92890 + }, + { + "epoch": 18.035338090383963, + "grad_norm": 0.1146666407585144, + "learning_rate": 1.37225541688637e-06, + "loss": 0.0188, + "step": 92900 + }, + { + "epoch": 18.037279743701763, + "grad_norm": 0.21462562680244446, + "learning_rate": 1.3684108888968594e-06, + "loss": 0.021, + "step": 92910 + }, + { + "epoch": 18.039221397019563, + "grad_norm": 0.18850679695606232, + "learning_rate": 1.364571679184823e-06, + "loss": 0.0212, + "step": 92920 + }, + { + "epoch": 18.041163050337364, + "grad_norm": 0.16511449217796326, + "learning_rate": 1.3607377881701077e-06, + "loss": 0.0195, + "step": 92930 + }, + { + "epoch": 18.043104703655164, + "grad_norm": 0.2310611754655838, + "learning_rate": 1.3569092162719787e-06, + "loss": 0.0247, + "step": 92940 + }, + { + "epoch": 18.045046356972964, + "grad_norm": 0.18910914659500122, + "learning_rate": 1.353085963909123e-06, + "loss": 0.0184, + "step": 92950 + }, + { + "epoch": 18.04698801029076, + "grad_norm": 0.16141851246356964, + "learning_rate": 1.3492680314996454e-06, + "loss": 0.0217, + "step": 92960 + }, + { + "epoch": 18.04892966360856, + "grad_norm": 0.13012133538722992, + "learning_rate": 1.345455419461078e-06, + "loss": 0.0258, + "step": 92970 + }, + { + "epoch": 18.05087131692636, + "grad_norm": 0.18993915617465973, + "learning_rate": 1.3416481282103431e-06, + "loss": 0.0247, + "step": 92980 + }, + { + "epoch": 18.05281297024416, + "grad_norm": 0.1895151436328888, + "learning_rate": 1.337846158163819e-06, + "loss": 0.0191, + "step": 92990 + }, + { + "epoch": 18.054754623561962, + "grad_norm": 0.2046438306570053, + "learning_rate": 1.334049509737273e-06, + "loss": 0.0205, + "step": 93000 + }, + { + "epoch": 18.056696276879762, + "grad_norm": 0.15899428725242615, + "learning_rate": 1.3302581833459116e-06, + "loss": 0.0235, + "step": 93010 + }, + { + "epoch": 18.058637930197563, + "grad_norm": 0.2023995965719223, + "learning_rate": 1.3264721794043366e-06, + "loss": 0.0223, + "step": 93020 + }, + { + "epoch": 18.060579583515363, + "grad_norm": 0.2675536870956421, + "learning_rate": 1.3226914983265836e-06, + "loss": 0.0206, + "step": 93030 + }, + { + "epoch": 18.062521236833163, + "grad_norm": 0.2268034815788269, + "learning_rate": 1.3189161405260997e-06, + "loss": 0.0254, + "step": 93040 + }, + { + "epoch": 18.064462890150963, + "grad_norm": 0.13913851976394653, + "learning_rate": 1.3151461064157656e-06, + "loss": 0.0233, + "step": 93050 + }, + { + "epoch": 18.066404543468764, + "grad_norm": 0.17953607439994812, + "learning_rate": 1.3113813964078459e-06, + "loss": 0.0187, + "step": 93060 + }, + { + "epoch": 18.068346196786564, + "grad_norm": 0.20702780783176422, + "learning_rate": 1.3076220109140614e-06, + "loss": 0.0215, + "step": 93070 + }, + { + "epoch": 18.070287850104364, + "grad_norm": 0.176734060049057, + "learning_rate": 1.3038679503455275e-06, + "loss": 0.0222, + "step": 93080 + }, + { + "epoch": 18.072229503422164, + "grad_norm": 0.1903819739818573, + "learning_rate": 1.3001192151127773e-06, + "loss": 0.0201, + "step": 93090 + }, + { + "epoch": 18.074171156739965, + "grad_norm": 0.22737827897071838, + "learning_rate": 1.2963758056257657e-06, + "loss": 0.0207, + "step": 93100 + }, + { + "epoch": 18.076112810057765, + "grad_norm": 0.08887090533971786, + "learning_rate": 1.2926377222938768e-06, + "loss": 0.0173, + "step": 93110 + }, + { + "epoch": 18.078054463375565, + "grad_norm": 0.1930423229932785, + "learning_rate": 1.2889049655259e-06, + "loss": 0.0243, + "step": 93120 + }, + { + "epoch": 18.079996116693366, + "grad_norm": 0.19958849251270294, + "learning_rate": 1.2851775357300312e-06, + "loss": 0.0218, + "step": 93130 + }, + { + "epoch": 18.081937770011166, + "grad_norm": 0.2722330093383789, + "learning_rate": 1.281455433313916e-06, + "loss": 0.0219, + "step": 93140 + }, + { + "epoch": 18.083879423328966, + "grad_norm": 0.24651221930980682, + "learning_rate": 1.2777386586845796e-06, + "loss": 0.0249, + "step": 93150 + }, + { + "epoch": 18.085821076646766, + "grad_norm": 0.2315211296081543, + "learning_rate": 1.2740272122484964e-06, + "loss": 0.0221, + "step": 93160 + }, + { + "epoch": 18.087762729964563, + "grad_norm": 0.1747341901063919, + "learning_rate": 1.2703210944115308e-06, + "loss": 0.0247, + "step": 93170 + }, + { + "epoch": 18.089704383282363, + "grad_norm": 0.16007402539253235, + "learning_rate": 1.2666203055789916e-06, + "loss": 0.0224, + "step": 93180 + }, + { + "epoch": 18.091646036600164, + "grad_norm": 0.21186693012714386, + "learning_rate": 1.2629248461555832e-06, + "loss": 0.0289, + "step": 93190 + }, + { + "epoch": 18.093587689917964, + "grad_norm": 0.2395220696926117, + "learning_rate": 1.2592347165454431e-06, + "loss": 0.0147, + "step": 93200 + }, + { + "epoch": 18.095529343235764, + "grad_norm": 0.16118542850017548, + "learning_rate": 1.2555499171521102e-06, + "loss": 0.0215, + "step": 93210 + }, + { + "epoch": 18.097470996553565, + "grad_norm": 0.1415191888809204, + "learning_rate": 1.251870448378556e-06, + "loss": 0.019, + "step": 93220 + }, + { + "epoch": 18.099412649871365, + "grad_norm": 0.10206245630979538, + "learning_rate": 1.2481963106271478e-06, + "loss": 0.0203, + "step": 93230 + }, + { + "epoch": 18.101354303189165, + "grad_norm": 0.20993711054325104, + "learning_rate": 1.2445275042997084e-06, + "loss": 0.0267, + "step": 93240 + }, + { + "epoch": 18.103295956506965, + "grad_norm": 0.10267539322376251, + "learning_rate": 1.2408640297974283e-06, + "loss": 0.0202, + "step": 93250 + }, + { + "epoch": 18.105237609824766, + "grad_norm": 0.12282426655292511, + "learning_rate": 1.2372058875209535e-06, + "loss": 0.0194, + "step": 93260 + }, + { + "epoch": 18.107179263142566, + "grad_norm": 0.2327377051115036, + "learning_rate": 1.2335530778703252e-06, + "loss": 0.0203, + "step": 93270 + }, + { + "epoch": 18.109120916460366, + "grad_norm": 0.2393730729818344, + "learning_rate": 1.2299056012450182e-06, + "loss": 0.0181, + "step": 93280 + }, + { + "epoch": 18.111062569778166, + "grad_norm": 0.18662461638450623, + "learning_rate": 1.2262634580439026e-06, + "loss": 0.0263, + "step": 93290 + }, + { + "epoch": 18.113004223095967, + "grad_norm": 0.15074630081653595, + "learning_rate": 1.2226266486652926e-06, + "loss": 0.0229, + "step": 93300 + }, + { + "epoch": 18.114945876413767, + "grad_norm": 0.17175108194351196, + "learning_rate": 1.2189951735068927e-06, + "loss": 0.0224, + "step": 93310 + }, + { + "epoch": 18.116887529731567, + "grad_norm": 0.12860998511314392, + "learning_rate": 1.2153690329658407e-06, + "loss": 0.0177, + "step": 93320 + }, + { + "epoch": 18.118829183049368, + "grad_norm": 0.1767640858888626, + "learning_rate": 1.2117482274386859e-06, + "loss": 0.0192, + "step": 93330 + }, + { + "epoch": 18.120770836367168, + "grad_norm": 0.14285582304000854, + "learning_rate": 1.2081327573213953e-06, + "loss": 0.0231, + "step": 93340 + }, + { + "epoch": 18.122712489684968, + "grad_norm": 0.17382808029651642, + "learning_rate": 1.2045226230093465e-06, + "loss": 0.0232, + "step": 93350 + }, + { + "epoch": 18.12465414300277, + "grad_norm": 0.23606041073799133, + "learning_rate": 1.2009178248973296e-06, + "loss": 0.0304, + "step": 93360 + }, + { + "epoch": 18.12659579632057, + "grad_norm": 0.2489369809627533, + "learning_rate": 1.1973183633795849e-06, + "loss": 0.0219, + "step": 93370 + }, + { + "epoch": 18.128537449638365, + "grad_norm": 0.16790705919265747, + "learning_rate": 1.193724238849725e-06, + "loss": 0.0199, + "step": 93380 + }, + { + "epoch": 18.130479102956166, + "grad_norm": 0.11536549031734467, + "learning_rate": 1.190135451700808e-06, + "loss": 0.0246, + "step": 93390 + }, + { + "epoch": 18.132420756273966, + "grad_norm": 0.1433160901069641, + "learning_rate": 1.186552002325292e-06, + "loss": 0.0215, + "step": 93400 + }, + { + "epoch": 18.134362409591766, + "grad_norm": 0.1430143266916275, + "learning_rate": 1.1829738911150578e-06, + "loss": 0.0233, + "step": 93410 + }, + { + "epoch": 18.136304062909566, + "grad_norm": 0.09065959602594376, + "learning_rate": 1.1794011184613985e-06, + "loss": 0.0182, + "step": 93420 + }, + { + "epoch": 18.138245716227367, + "grad_norm": 0.16273067891597748, + "learning_rate": 1.17583368475504e-06, + "loss": 0.0301, + "step": 93430 + }, + { + "epoch": 18.140187369545167, + "grad_norm": 0.2227368950843811, + "learning_rate": 1.1722715903860982e-06, + "loss": 0.0195, + "step": 93440 + }, + { + "epoch": 18.142129022862967, + "grad_norm": 0.1367158591747284, + "learning_rate": 1.168714835744128e-06, + "loss": 0.0223, + "step": 93450 + }, + { + "epoch": 18.144070676180768, + "grad_norm": 0.21941371262073517, + "learning_rate": 1.165163421218085e-06, + "loss": 0.0222, + "step": 93460 + }, + { + "epoch": 18.146012329498568, + "grad_norm": 0.17530737817287445, + "learning_rate": 1.1616173471963477e-06, + "loss": 0.0201, + "step": 93470 + }, + { + "epoch": 18.147953982816368, + "grad_norm": 0.20689734816551208, + "learning_rate": 1.1580766140667109e-06, + "loss": 0.0305, + "step": 93480 + }, + { + "epoch": 18.14989563613417, + "grad_norm": 0.18223540484905243, + "learning_rate": 1.1545412222163764e-06, + "loss": 0.0208, + "step": 93490 + }, + { + "epoch": 18.15183728945197, + "grad_norm": 0.15531598031520844, + "learning_rate": 1.1510111720319795e-06, + "loss": 0.0165, + "step": 93500 + }, + { + "epoch": 18.15377894276977, + "grad_norm": 0.12664209306240082, + "learning_rate": 1.1474864638995663e-06, + "loss": 0.0228, + "step": 93510 + }, + { + "epoch": 18.15572059608757, + "grad_norm": 0.24115332961082458, + "learning_rate": 1.1439670982045735e-06, + "loss": 0.0222, + "step": 93520 + }, + { + "epoch": 18.15766224940537, + "grad_norm": 0.19581115245819092, + "learning_rate": 1.1404530753318877e-06, + "loss": 0.0224, + "step": 93530 + }, + { + "epoch": 18.15960390272317, + "grad_norm": 0.133926659822464, + "learning_rate": 1.136944395665801e-06, + "loss": 0.02, + "step": 93540 + }, + { + "epoch": 18.16154555604097, + "grad_norm": 0.25924208760261536, + "learning_rate": 1.1334410595900014e-06, + "loss": 0.0204, + "step": 93550 + }, + { + "epoch": 18.16348720935877, + "grad_norm": 0.22854512929916382, + "learning_rate": 1.129943067487621e-06, + "loss": 0.0279, + "step": 93560 + }, + { + "epoch": 18.16542886267657, + "grad_norm": 0.14127539098262787, + "learning_rate": 1.126450419741193e-06, + "loss": 0.0181, + "step": 93570 + }, + { + "epoch": 18.16737051599437, + "grad_norm": 0.1732436716556549, + "learning_rate": 1.1229631167326671e-06, + "loss": 0.0216, + "step": 93580 + }, + { + "epoch": 18.16931216931217, + "grad_norm": 0.2739960849285126, + "learning_rate": 1.1194811588434052e-06, + "loss": 0.0212, + "step": 93590 + }, + { + "epoch": 18.171253822629968, + "grad_norm": 0.16470865905284882, + "learning_rate": 1.1160045464542023e-06, + "loss": 0.019, + "step": 93600 + }, + { + "epoch": 18.173195475947768, + "grad_norm": 0.16575787961483002, + "learning_rate": 1.1125332799452325e-06, + "loss": 0.0226, + "step": 93610 + }, + { + "epoch": 18.17513712926557, + "grad_norm": 0.3705979585647583, + "learning_rate": 1.109067359696131e-06, + "loss": 0.0212, + "step": 93620 + }, + { + "epoch": 18.17707878258337, + "grad_norm": 0.206977978348732, + "learning_rate": 1.1056067860859166e-06, + "loss": 0.0244, + "step": 93630 + }, + { + "epoch": 18.17902043590117, + "grad_norm": 0.2553133964538574, + "learning_rate": 1.102151559493031e-06, + "loss": 0.0242, + "step": 93640 + }, + { + "epoch": 18.18096208921897, + "grad_norm": 0.2437540590763092, + "learning_rate": 1.0987016802953332e-06, + "loss": 0.0226, + "step": 93650 + }, + { + "epoch": 18.18290374253677, + "grad_norm": 0.17607282102108002, + "learning_rate": 1.0952571488701048e-06, + "loss": 0.0198, + "step": 93660 + }, + { + "epoch": 18.18484539585457, + "grad_norm": 0.20521344244480133, + "learning_rate": 1.0918179655940163e-06, + "loss": 0.0274, + "step": 93670 + }, + { + "epoch": 18.18678704917237, + "grad_norm": 0.10637766122817993, + "learning_rate": 1.088384130843184e-06, + "loss": 0.0219, + "step": 93680 + }, + { + "epoch": 18.18872870249017, + "grad_norm": 0.2837070822715759, + "learning_rate": 1.0849556449931296e-06, + "loss": 0.0225, + "step": 93690 + }, + { + "epoch": 18.19067035580797, + "grad_norm": 0.1155649721622467, + "learning_rate": 1.0815325084187866e-06, + "loss": 0.0201, + "step": 93700 + }, + { + "epoch": 18.19261200912577, + "grad_norm": 0.2643093168735504, + "learning_rate": 1.0781147214944943e-06, + "loss": 0.0238, + "step": 93710 + }, + { + "epoch": 18.19455366244357, + "grad_norm": 0.17010734975337982, + "learning_rate": 1.0747022845940313e-06, + "loss": 0.0215, + "step": 93720 + }, + { + "epoch": 18.19649531576137, + "grad_norm": 0.1424802988767624, + "learning_rate": 1.0712951980905605e-06, + "loss": 0.022, + "step": 93730 + }, + { + "epoch": 18.19843696907917, + "grad_norm": 0.14301826059818268, + "learning_rate": 1.0678934623566839e-06, + "loss": 0.0175, + "step": 93740 + }, + { + "epoch": 18.200378622396972, + "grad_norm": 0.1988416612148285, + "learning_rate": 1.0644970777644091e-06, + "loss": 0.022, + "step": 93750 + }, + { + "epoch": 18.202320275714772, + "grad_norm": 0.12879225611686707, + "learning_rate": 1.0611060446851673e-06, + "loss": 0.0159, + "step": 93760 + }, + { + "epoch": 18.204261929032572, + "grad_norm": 0.16400456428527832, + "learning_rate": 1.0577203634897836e-06, + "loss": 0.0219, + "step": 93770 + }, + { + "epoch": 18.206203582350373, + "grad_norm": 0.16232092678546906, + "learning_rate": 1.0543400345485178e-06, + "loss": 0.0175, + "step": 93780 + }, + { + "epoch": 18.208145235668173, + "grad_norm": 0.11482986807823181, + "learning_rate": 1.0509650582310404e-06, + "loss": 0.0202, + "step": 93790 + }, + { + "epoch": 18.210086888985973, + "grad_norm": 0.15169939398765564, + "learning_rate": 1.0475954349064287e-06, + "loss": 0.0167, + "step": 93800 + }, + { + "epoch": 18.21202854230377, + "grad_norm": 0.18363063037395477, + "learning_rate": 1.0442311649431825e-06, + "loss": 0.0185, + "step": 93810 + }, + { + "epoch": 18.21397019562157, + "grad_norm": 0.13871675729751587, + "learning_rate": 1.0408722487092071e-06, + "loss": 0.0242, + "step": 93820 + }, + { + "epoch": 18.21591184893937, + "grad_norm": 0.25311344861984253, + "learning_rate": 1.0375186865718424e-06, + "loss": 0.02, + "step": 93830 + }, + { + "epoch": 18.21785350225717, + "grad_norm": 0.09950946271419525, + "learning_rate": 1.0341704788978113e-06, + "loss": 0.0207, + "step": 93840 + }, + { + "epoch": 18.21979515557497, + "grad_norm": 0.1599515676498413, + "learning_rate": 1.0308276260532879e-06, + "loss": 0.0291, + "step": 93850 + }, + { + "epoch": 18.22173680889277, + "grad_norm": 0.16695131361484528, + "learning_rate": 1.0274901284038241e-06, + "loss": 0.0208, + "step": 93860 + }, + { + "epoch": 18.22367846221057, + "grad_norm": 0.47462522983551025, + "learning_rate": 1.0241579863144114e-06, + "loss": 0.0302, + "step": 93870 + }, + { + "epoch": 18.225620115528372, + "grad_norm": 0.18690137565135956, + "learning_rate": 1.0208312001494413e-06, + "loss": 0.0196, + "step": 93880 + }, + { + "epoch": 18.227561768846172, + "grad_norm": 0.16962875425815582, + "learning_rate": 1.0175097702727453e-06, + "loss": 0.0243, + "step": 93890 + }, + { + "epoch": 18.229503422163972, + "grad_norm": 0.16479846835136414, + "learning_rate": 1.0141936970475275e-06, + "loss": 0.0208, + "step": 93900 + }, + { + "epoch": 18.231445075481773, + "grad_norm": 0.1801426261663437, + "learning_rate": 1.0108829808364417e-06, + "loss": 0.0259, + "step": 93910 + }, + { + "epoch": 18.233386728799573, + "grad_norm": 0.20559726655483246, + "learning_rate": 1.0075776220015376e-06, + "loss": 0.0186, + "step": 93920 + }, + { + "epoch": 18.235328382117373, + "grad_norm": 0.27837681770324707, + "learning_rate": 1.0042776209042816e-06, + "loss": 0.0259, + "step": 93930 + }, + { + "epoch": 18.237270035435174, + "grad_norm": 0.17799706757068634, + "learning_rate": 1.000982977905568e-06, + "loss": 0.0223, + "step": 93940 + }, + { + "epoch": 18.239211688752974, + "grad_norm": 0.13785803318023682, + "learning_rate": 9.976936933656867e-07, + "loss": 0.0191, + "step": 93950 + }, + { + "epoch": 18.241153342070774, + "grad_norm": 0.11539746075868607, + "learning_rate": 9.944097676443442e-07, + "loss": 0.0189, + "step": 93960 + }, + { + "epoch": 18.243094995388574, + "grad_norm": 0.1574513018131256, + "learning_rate": 9.911312011006757e-07, + "loss": 0.0198, + "step": 93970 + }, + { + "epoch": 18.245036648706375, + "grad_norm": 0.10384595394134521, + "learning_rate": 9.878579940932109e-07, + "loss": 0.019, + "step": 93980 + }, + { + "epoch": 18.246978302024175, + "grad_norm": 0.17978620529174805, + "learning_rate": 9.84590146979908e-07, + "loss": 0.025, + "step": 93990 + }, + { + "epoch": 18.248919955341975, + "grad_norm": 0.25732800364494324, + "learning_rate": 9.813276601181365e-07, + "loss": 0.0214, + "step": 94000 + }, + { + "epoch": 18.250861608659775, + "grad_norm": 0.18818598985671997, + "learning_rate": 9.780705338646724e-07, + "loss": 0.0264, + "step": 94010 + }, + { + "epoch": 18.252803261977572, + "grad_norm": 0.21943160891532898, + "learning_rate": 9.748187685757083e-07, + "loss": 0.022, + "step": 94020 + }, + { + "epoch": 18.254744915295372, + "grad_norm": 0.18234160542488098, + "learning_rate": 9.715723646068543e-07, + "loss": 0.0209, + "step": 94030 + }, + { + "epoch": 18.256686568613173, + "grad_norm": 0.1727767437696457, + "learning_rate": 9.68331322313143e-07, + "loss": 0.029, + "step": 94040 + }, + { + "epoch": 18.258628221930973, + "grad_norm": 0.0991235077381134, + "learning_rate": 9.650956420489909e-07, + "loss": 0.0238, + "step": 94050 + }, + { + "epoch": 18.260569875248773, + "grad_norm": 0.28840282559394836, + "learning_rate": 9.618653241682595e-07, + "loss": 0.0183, + "step": 94060 + }, + { + "epoch": 18.262511528566574, + "grad_norm": 0.14417076110839844, + "learning_rate": 9.586403690242053e-07, + "loss": 0.0221, + "step": 94070 + }, + { + "epoch": 18.264453181884374, + "grad_norm": 0.16582095623016357, + "learning_rate": 9.554207769695122e-07, + "loss": 0.019, + "step": 94080 + }, + { + "epoch": 18.266394835202174, + "grad_norm": 0.0846128985285759, + "learning_rate": 9.522065483562659e-07, + "loss": 0.0236, + "step": 94090 + }, + { + "epoch": 18.268336488519974, + "grad_norm": 0.14075498282909393, + "learning_rate": 9.489976835359681e-07, + "loss": 0.019, + "step": 94100 + }, + { + "epoch": 18.270278141837775, + "grad_norm": 0.1614294797182083, + "learning_rate": 9.457941828595329e-07, + "loss": 0.0182, + "step": 94110 + }, + { + "epoch": 18.272219795155575, + "grad_norm": 0.1055249348282814, + "learning_rate": 9.425960466772965e-07, + "loss": 0.0322, + "step": 94120 + }, + { + "epoch": 18.274161448473375, + "grad_norm": 0.1627136766910553, + "learning_rate": 9.394032753390014e-07, + "loss": 0.0195, + "step": 94130 + }, + { + "epoch": 18.276103101791175, + "grad_norm": 0.19099591672420502, + "learning_rate": 9.362158691938017e-07, + "loss": 0.0202, + "step": 94140 + }, + { + "epoch": 18.278044755108976, + "grad_norm": 0.13999369740486145, + "learning_rate": 9.330338285902685e-07, + "loss": 0.0245, + "step": 94150 + }, + { + "epoch": 18.279986408426776, + "grad_norm": 0.19208386540412903, + "learning_rate": 9.298571538763845e-07, + "loss": 0.0245, + "step": 94160 + }, + { + "epoch": 18.281928061744576, + "grad_norm": 0.2149776965379715, + "learning_rate": 9.266858453995441e-07, + "loss": 0.0251, + "step": 94170 + }, + { + "epoch": 18.283869715062377, + "grad_norm": 0.15809808671474457, + "learning_rate": 9.235199035065645e-07, + "loss": 0.0241, + "step": 94180 + }, + { + "epoch": 18.285811368380177, + "grad_norm": 0.11202547699213028, + "learning_rate": 9.203593285436574e-07, + "loss": 0.0251, + "step": 94190 + }, + { + "epoch": 18.287753021697977, + "grad_norm": 0.21650999784469604, + "learning_rate": 9.172041208564686e-07, + "loss": 0.0276, + "step": 94200 + }, + { + "epoch": 18.289694675015777, + "grad_norm": 0.1558409035205841, + "learning_rate": 9.140542807900332e-07, + "loss": 0.0214, + "step": 94210 + }, + { + "epoch": 18.291636328333578, + "grad_norm": 0.27445921301841736, + "learning_rate": 9.109098086888313e-07, + "loss": 0.0184, + "step": 94220 + }, + { + "epoch": 18.293577981651374, + "grad_norm": 0.2428232580423355, + "learning_rate": 9.077707048967265e-07, + "loss": 0.0231, + "step": 94230 + }, + { + "epoch": 18.295519634969175, + "grad_norm": 0.12369578331708908, + "learning_rate": 9.046369697570057e-07, + "loss": 0.0268, + "step": 94240 + }, + { + "epoch": 18.297461288286975, + "grad_norm": 0.1647190898656845, + "learning_rate": 9.015086036123721e-07, + "loss": 0.02, + "step": 94250 + }, + { + "epoch": 18.299402941604775, + "grad_norm": 0.1469801664352417, + "learning_rate": 8.983856068049413e-07, + "loss": 0.0187, + "step": 94260 + }, + { + "epoch": 18.301344594922575, + "grad_norm": 0.10631681978702545, + "learning_rate": 8.952679796762398e-07, + "loss": 0.0313, + "step": 94270 + }, + { + "epoch": 18.303286248240376, + "grad_norm": 0.29808199405670166, + "learning_rate": 8.921557225671951e-07, + "loss": 0.025, + "step": 94280 + }, + { + "epoch": 18.305227901558176, + "grad_norm": 0.14370372891426086, + "learning_rate": 8.890488358181792e-07, + "loss": 0.0204, + "step": 94290 + }, + { + "epoch": 18.307169554875976, + "grad_norm": 0.18853065371513367, + "learning_rate": 8.859473197689372e-07, + "loss": 0.0276, + "step": 94300 + }, + { + "epoch": 18.309111208193777, + "grad_norm": 0.18704955279827118, + "learning_rate": 8.828511747586588e-07, + "loss": 0.0212, + "step": 94310 + }, + { + "epoch": 18.311052861511577, + "grad_norm": 0.13333988189697266, + "learning_rate": 8.797604011259287e-07, + "loss": 0.0183, + "step": 94320 + }, + { + "epoch": 18.312994514829377, + "grad_norm": 0.2661505341529846, + "learning_rate": 8.766749992087487e-07, + "loss": 0.0243, + "step": 94330 + }, + { + "epoch": 18.314936168147177, + "grad_norm": 0.13447771966457367, + "learning_rate": 8.735949693445378e-07, + "loss": 0.0237, + "step": 94340 + }, + { + "epoch": 18.316877821464978, + "grad_norm": 0.24393603205680847, + "learning_rate": 8.705203118701211e-07, + "loss": 0.0195, + "step": 94350 + }, + { + "epoch": 18.318819474782778, + "grad_norm": 0.2942955195903778, + "learning_rate": 8.674510271217295e-07, + "loss": 0.0185, + "step": 94360 + }, + { + "epoch": 18.320761128100578, + "grad_norm": 0.1685977727174759, + "learning_rate": 8.643871154350336e-07, + "loss": 0.0204, + "step": 94370 + }, + { + "epoch": 18.32270278141838, + "grad_norm": 0.1929691582918167, + "learning_rate": 8.613285771450818e-07, + "loss": 0.0241, + "step": 94380 + }, + { + "epoch": 18.32464443473618, + "grad_norm": 0.40061599016189575, + "learning_rate": 8.582754125863623e-07, + "loss": 0.0299, + "step": 94390 + }, + { + "epoch": 18.32658608805398, + "grad_norm": 0.21419157087802887, + "learning_rate": 8.55227622092758e-07, + "loss": 0.0192, + "step": 94400 + }, + { + "epoch": 18.32852774137178, + "grad_norm": 0.2717013657093048, + "learning_rate": 8.521852059975743e-07, + "loss": 0.024, + "step": 94410 + }, + { + "epoch": 18.33046939468958, + "grad_norm": 0.19358807802200317, + "learning_rate": 8.491481646335231e-07, + "loss": 0.022, + "step": 94420 + }, + { + "epoch": 18.33241104800738, + "grad_norm": 0.2686283588409424, + "learning_rate": 8.46116498332733e-07, + "loss": 0.0216, + "step": 94430 + }, + { + "epoch": 18.33435270132518, + "grad_norm": 0.1089208722114563, + "learning_rate": 8.430902074267333e-07, + "loss": 0.0221, + "step": 94440 + }, + { + "epoch": 18.336294354642977, + "grad_norm": 0.13515932857990265, + "learning_rate": 8.400692922464925e-07, + "loss": 0.015, + "step": 94450 + }, + { + "epoch": 18.338236007960777, + "grad_norm": 0.17907004058361053, + "learning_rate": 8.370537531223577e-07, + "loss": 0.0241, + "step": 94460 + }, + { + "epoch": 18.340177661278577, + "grad_norm": 0.24670292437076569, + "learning_rate": 8.340435903841092e-07, + "loss": 0.0269, + "step": 94470 + }, + { + "epoch": 18.342119314596378, + "grad_norm": 0.17477712035179138, + "learning_rate": 8.310388043609341e-07, + "loss": 0.0237, + "step": 94480 + }, + { + "epoch": 18.344060967914178, + "grad_norm": 0.1476282924413681, + "learning_rate": 8.280393953814303e-07, + "loss": 0.0203, + "step": 94490 + }, + { + "epoch": 18.34600262123198, + "grad_norm": 0.1424543410539627, + "learning_rate": 8.250453637736077e-07, + "loss": 0.0176, + "step": 94500 + }, + { + "epoch": 18.34794427454978, + "grad_norm": 0.16413021087646484, + "learning_rate": 8.22056709864899e-07, + "loss": 0.023, + "step": 94510 + }, + { + "epoch": 18.34988592786758, + "grad_norm": 0.12253589183092117, + "learning_rate": 8.190734339821204e-07, + "loss": 0.0265, + "step": 94520 + }, + { + "epoch": 18.35182758118538, + "grad_norm": 0.14833113551139832, + "learning_rate": 8.160955364515333e-07, + "loss": 0.0168, + "step": 94530 + }, + { + "epoch": 18.35376923450318, + "grad_norm": 0.25982508063316345, + "learning_rate": 8.131230175987936e-07, + "loss": 0.0229, + "step": 94540 + }, + { + "epoch": 18.35571088782098, + "grad_norm": 0.14619195461273193, + "learning_rate": 8.101558777489693e-07, + "loss": 0.0182, + "step": 94550 + }, + { + "epoch": 18.35765254113878, + "grad_norm": 0.2613223195075989, + "learning_rate": 8.071941172265396e-07, + "loss": 0.0216, + "step": 94560 + }, + { + "epoch": 18.35959419445658, + "grad_norm": 0.15258294343948364, + "learning_rate": 8.042377363553955e-07, + "loss": 0.0264, + "step": 94570 + }, + { + "epoch": 18.36153584777438, + "grad_norm": 0.18644660711288452, + "learning_rate": 8.01286735458856e-07, + "loss": 0.0189, + "step": 94580 + }, + { + "epoch": 18.36347750109218, + "grad_norm": 0.14589804410934448, + "learning_rate": 7.983411148596298e-07, + "loss": 0.0172, + "step": 94590 + }, + { + "epoch": 18.36541915440998, + "grad_norm": 0.13240008056163788, + "learning_rate": 7.954008748798425e-07, + "loss": 0.017, + "step": 94600 + }, + { + "epoch": 18.36736080772778, + "grad_norm": 0.18262727558612823, + "learning_rate": 7.924660158410369e-07, + "loss": 0.0248, + "step": 94610 + }, + { + "epoch": 18.36930246104558, + "grad_norm": 0.15059159696102142, + "learning_rate": 7.895365380641728e-07, + "loss": 0.0264, + "step": 94620 + }, + { + "epoch": 18.371244114363382, + "grad_norm": 0.2547950744628906, + "learning_rate": 7.866124418695941e-07, + "loss": 0.0264, + "step": 94630 + }, + { + "epoch": 18.373185767681182, + "grad_norm": 0.26111286878585815, + "learning_rate": 7.83693727577095e-07, + "loss": 0.0254, + "step": 94640 + }, + { + "epoch": 18.375127420998982, + "grad_norm": 0.21600887179374695, + "learning_rate": 7.807803955058535e-07, + "loss": 0.0291, + "step": 94650 + }, + { + "epoch": 18.37706907431678, + "grad_norm": 0.2842095196247101, + "learning_rate": 7.778724459744702e-07, + "loss": 0.0314, + "step": 94660 + }, + { + "epoch": 18.37901072763458, + "grad_norm": 0.21774110198020935, + "learning_rate": 7.749698793009464e-07, + "loss": 0.0201, + "step": 94670 + }, + { + "epoch": 18.38095238095238, + "grad_norm": 0.21152855455875397, + "learning_rate": 7.720726958027169e-07, + "loss": 0.0182, + "step": 94680 + }, + { + "epoch": 18.38289403427018, + "grad_norm": 0.14801275730133057, + "learning_rate": 7.691808957965951e-07, + "loss": 0.0185, + "step": 94690 + }, + { + "epoch": 18.38483568758798, + "grad_norm": 0.18916195631027222, + "learning_rate": 7.662944795988336e-07, + "loss": 0.0172, + "step": 94700 + }, + { + "epoch": 18.38677734090578, + "grad_norm": 0.10626525431871414, + "learning_rate": 7.634134475250854e-07, + "loss": 0.0204, + "step": 94710 + }, + { + "epoch": 18.38871899422358, + "grad_norm": 0.1494661271572113, + "learning_rate": 7.605377998904262e-07, + "loss": 0.0174, + "step": 94720 + }, + { + "epoch": 18.39066064754138, + "grad_norm": 0.14017467200756073, + "learning_rate": 7.5766753700931e-07, + "loss": 0.0201, + "step": 94730 + }, + { + "epoch": 18.39260230085918, + "grad_norm": 0.16017888486385345, + "learning_rate": 7.54802659195647e-07, + "loss": 0.0217, + "step": 94740 + }, + { + "epoch": 18.39454395417698, + "grad_norm": 0.1485542356967926, + "learning_rate": 7.519431667627253e-07, + "loss": 0.0171, + "step": 94750 + }, + { + "epoch": 18.396485607494782, + "grad_norm": 0.15700097382068634, + "learning_rate": 7.490890600232503e-07, + "loss": 0.0176, + "step": 94760 + }, + { + "epoch": 18.398427260812582, + "grad_norm": 0.1805541217327118, + "learning_rate": 7.462403392893502e-07, + "loss": 0.0185, + "step": 94770 + }, + { + "epoch": 18.400368914130382, + "grad_norm": 0.12699924409389496, + "learning_rate": 7.433970048725536e-07, + "loss": 0.0189, + "step": 94780 + }, + { + "epoch": 18.402310567448183, + "grad_norm": 0.21888618171215057, + "learning_rate": 7.405590570838062e-07, + "loss": 0.0197, + "step": 94790 + }, + { + "epoch": 18.404252220765983, + "grad_norm": 0.2024553120136261, + "learning_rate": 7.377264962334596e-07, + "loss": 0.0192, + "step": 94800 + }, + { + "epoch": 18.406193874083783, + "grad_norm": 0.24533262848854065, + "learning_rate": 7.348993226312773e-07, + "loss": 0.0265, + "step": 94810 + }, + { + "epoch": 18.408135527401583, + "grad_norm": 0.19808316230773926, + "learning_rate": 7.32077536586434e-07, + "loss": 0.0201, + "step": 94820 + }, + { + "epoch": 18.410077180719384, + "grad_norm": 0.22619493305683136, + "learning_rate": 7.292611384075221e-07, + "loss": 0.0208, + "step": 94830 + }, + { + "epoch": 18.412018834037184, + "grad_norm": 0.19517047703266144, + "learning_rate": 7.264501284025338e-07, + "loss": 0.0188, + "step": 94840 + }, + { + "epoch": 18.413960487354984, + "grad_norm": 0.21303975582122803, + "learning_rate": 7.236445068788789e-07, + "loss": 0.0264, + "step": 94850 + }, + { + "epoch": 18.415902140672785, + "grad_norm": 0.22337615489959717, + "learning_rate": 7.20844274143373e-07, + "loss": 0.0206, + "step": 94860 + }, + { + "epoch": 18.41784379399058, + "grad_norm": 0.17862272262573242, + "learning_rate": 7.180494305022489e-07, + "loss": 0.0298, + "step": 94870 + }, + { + "epoch": 18.41978544730838, + "grad_norm": 0.11950746923685074, + "learning_rate": 7.152599762611456e-07, + "loss": 0.0218, + "step": 94880 + }, + { + "epoch": 18.421727100626182, + "grad_norm": 0.1817234456539154, + "learning_rate": 7.124759117251079e-07, + "loss": 0.016, + "step": 94890 + }, + { + "epoch": 18.423668753943982, + "grad_norm": 0.10411788523197174, + "learning_rate": 7.096972371986033e-07, + "loss": 0.0198, + "step": 94900 + }, + { + "epoch": 18.425610407261782, + "grad_norm": 0.14081275463104248, + "learning_rate": 7.069239529855054e-07, + "loss": 0.0178, + "step": 94910 + }, + { + "epoch": 18.427552060579583, + "grad_norm": 0.25046083331108093, + "learning_rate": 7.041560593890939e-07, + "loss": 0.024, + "step": 94920 + }, + { + "epoch": 18.429493713897383, + "grad_norm": 0.16874995827674866, + "learning_rate": 7.013935567120599e-07, + "loss": 0.0207, + "step": 94930 + }, + { + "epoch": 18.431435367215183, + "grad_norm": 0.3847932517528534, + "learning_rate": 6.986364452565009e-07, + "loss": 0.0238, + "step": 94940 + }, + { + "epoch": 18.433377020532983, + "grad_norm": 0.12991608679294586, + "learning_rate": 6.95884725323942e-07, + "loss": 0.0231, + "step": 94950 + }, + { + "epoch": 18.435318673850784, + "grad_norm": 0.20318371057510376, + "learning_rate": 6.931383972152982e-07, + "loss": 0.0172, + "step": 94960 + }, + { + "epoch": 18.437260327168584, + "grad_norm": 0.2252999246120453, + "learning_rate": 6.903974612309127e-07, + "loss": 0.0215, + "step": 94970 + }, + { + "epoch": 18.439201980486384, + "grad_norm": 0.19572824239730835, + "learning_rate": 6.876619176705179e-07, + "loss": 0.0186, + "step": 94980 + }, + { + "epoch": 18.441143633804185, + "grad_norm": 0.1996491253376007, + "learning_rate": 6.8493176683328e-07, + "loss": 0.0284, + "step": 94990 + }, + { + "epoch": 18.443085287121985, + "grad_norm": 0.2362767606973648, + "learning_rate": 6.822070090177657e-07, + "loss": 0.0198, + "step": 95000 + }, + { + "epoch": 18.445026940439785, + "grad_norm": 0.2742530107498169, + "learning_rate": 6.794876445219367e-07, + "loss": 0.0191, + "step": 95010 + }, + { + "epoch": 18.446968593757585, + "grad_norm": 0.21378380060195923, + "learning_rate": 6.767736736431885e-07, + "loss": 0.0235, + "step": 95020 + }, + { + "epoch": 18.448910247075386, + "grad_norm": 0.19839851558208466, + "learning_rate": 6.740650966783113e-07, + "loss": 0.0255, + "step": 95030 + }, + { + "epoch": 18.450851900393186, + "grad_norm": 0.16597634553909302, + "learning_rate": 6.713619139235239e-07, + "loss": 0.0175, + "step": 95040 + }, + { + "epoch": 18.452793553710986, + "grad_norm": 0.11784008890390396, + "learning_rate": 6.686641256744286e-07, + "loss": 0.0257, + "step": 95050 + }, + { + "epoch": 18.454735207028786, + "grad_norm": 0.3366492688655853, + "learning_rate": 6.659717322260562e-07, + "loss": 0.0273, + "step": 95060 + }, + { + "epoch": 18.456676860346587, + "grad_norm": 0.1838623732328415, + "learning_rate": 6.632847338728377e-07, + "loss": 0.0185, + "step": 95070 + }, + { + "epoch": 18.458618513664383, + "grad_norm": 0.2273687720298767, + "learning_rate": 6.60603130908627e-07, + "loss": 0.0214, + "step": 95080 + }, + { + "epoch": 18.460560166982184, + "grad_norm": 0.2639268934726715, + "learning_rate": 6.579269236266783e-07, + "loss": 0.0193, + "step": 95090 + }, + { + "epoch": 18.462501820299984, + "grad_norm": 0.18573366105556488, + "learning_rate": 6.552561123196632e-07, + "loss": 0.0253, + "step": 95100 + }, + { + "epoch": 18.464443473617784, + "grad_norm": 0.14185218513011932, + "learning_rate": 6.525906972796425e-07, + "loss": 0.0208, + "step": 95110 + }, + { + "epoch": 18.466385126935585, + "grad_norm": 0.1515333652496338, + "learning_rate": 6.499306787981108e-07, + "loss": 0.0184, + "step": 95120 + }, + { + "epoch": 18.468326780253385, + "grad_norm": 0.3586294949054718, + "learning_rate": 6.472760571659631e-07, + "loss": 0.0194, + "step": 95130 + }, + { + "epoch": 18.470268433571185, + "grad_norm": 0.30399954319000244, + "learning_rate": 6.446268326735061e-07, + "loss": 0.0218, + "step": 95140 + }, + { + "epoch": 18.472210086888985, + "grad_norm": 0.17316626012325287, + "learning_rate": 6.41983005610458e-07, + "loss": 0.0222, + "step": 95150 + }, + { + "epoch": 18.474151740206786, + "grad_norm": 0.1627032607793808, + "learning_rate": 6.393445762659378e-07, + "loss": 0.0259, + "step": 95160 + }, + { + "epoch": 18.476093393524586, + "grad_norm": 0.23716270923614502, + "learning_rate": 6.36711544928481e-07, + "loss": 0.0173, + "step": 95170 + }, + { + "epoch": 18.478035046842386, + "grad_norm": 0.1337077021598816, + "learning_rate": 6.340839118860298e-07, + "loss": 0.018, + "step": 95180 + }, + { + "epoch": 18.479976700160186, + "grad_norm": 0.1934991329908371, + "learning_rate": 6.314616774259485e-07, + "loss": 0.0241, + "step": 95190 + }, + { + "epoch": 18.481918353477987, + "grad_norm": 0.14528022706508636, + "learning_rate": 6.288448418349858e-07, + "loss": 0.0212, + "step": 95200 + }, + { + "epoch": 18.483860006795787, + "grad_norm": 0.14403003454208374, + "learning_rate": 6.262334053993235e-07, + "loss": 0.0261, + "step": 95210 + }, + { + "epoch": 18.485801660113587, + "grad_norm": 0.28342172503471375, + "learning_rate": 6.236273684045501e-07, + "loss": 0.0264, + "step": 95220 + }, + { + "epoch": 18.487743313431388, + "grad_norm": 0.14235332608222961, + "learning_rate": 6.210267311356488e-07, + "loss": 0.0182, + "step": 95230 + }, + { + "epoch": 18.489684966749188, + "grad_norm": 0.2100108116865158, + "learning_rate": 6.184314938770197e-07, + "loss": 0.0202, + "step": 95240 + }, + { + "epoch": 18.491626620066988, + "grad_norm": 0.16958041489124298, + "learning_rate": 6.158416569124858e-07, + "loss": 0.0245, + "step": 95250 + }, + { + "epoch": 18.49356827338479, + "grad_norm": 0.15334230661392212, + "learning_rate": 6.132572205252541e-07, + "loss": 0.0219, + "step": 95260 + }, + { + "epoch": 18.49550992670259, + "grad_norm": 0.26288580894470215, + "learning_rate": 6.106781849979648e-07, + "loss": 0.0233, + "step": 95270 + }, + { + "epoch": 18.49745158002039, + "grad_norm": 0.19759193062782288, + "learning_rate": 6.081045506126482e-07, + "loss": 0.0217, + "step": 95280 + }, + { + "epoch": 18.49939323333819, + "grad_norm": 0.24859173595905304, + "learning_rate": 6.055363176507678e-07, + "loss": 0.0232, + "step": 95290 + }, + { + "epoch": 18.501334886655986, + "grad_norm": 0.24318112432956696, + "learning_rate": 6.029734863931602e-07, + "loss": 0.0256, + "step": 95300 + }, + { + "epoch": 18.503276539973786, + "grad_norm": 0.29252761602401733, + "learning_rate": 6.004160571201179e-07, + "loss": 0.0191, + "step": 95310 + }, + { + "epoch": 18.505218193291586, + "grad_norm": 0.13936561346054077, + "learning_rate": 5.978640301112947e-07, + "loss": 0.0196, + "step": 95320 + }, + { + "epoch": 18.507159846609387, + "grad_norm": 0.17378588020801544, + "learning_rate": 5.9531740564579e-07, + "loss": 0.0225, + "step": 95330 + }, + { + "epoch": 18.509101499927187, + "grad_norm": 0.19428882002830505, + "learning_rate": 5.927761840020918e-07, + "loss": 0.0195, + "step": 95340 + }, + { + "epoch": 18.511043153244987, + "grad_norm": 0.18431632220745087, + "learning_rate": 5.902403654581113e-07, + "loss": 0.018, + "step": 95350 + }, + { + "epoch": 18.512984806562788, + "grad_norm": 0.091387078166008, + "learning_rate": 5.877099502911542e-07, + "loss": 0.0203, + "step": 95360 + }, + { + "epoch": 18.514926459880588, + "grad_norm": 0.134566992521286, + "learning_rate": 5.851849387779496e-07, + "loss": 0.0238, + "step": 95370 + }, + { + "epoch": 18.516868113198388, + "grad_norm": 0.17857202887535095, + "learning_rate": 5.826653311946262e-07, + "loss": 0.0218, + "step": 95380 + }, + { + "epoch": 18.51880976651619, + "grad_norm": 0.37984538078308105, + "learning_rate": 5.801511278167249e-07, + "loss": 0.0241, + "step": 95390 + }, + { + "epoch": 18.52075141983399, + "grad_norm": 0.13261981308460236, + "learning_rate": 5.776423289191868e-07, + "loss": 0.0231, + "step": 95400 + }, + { + "epoch": 18.52269307315179, + "grad_norm": 0.1996171772480011, + "learning_rate": 5.751389347763869e-07, + "loss": 0.0182, + "step": 95410 + }, + { + "epoch": 18.52463472646959, + "grad_norm": 0.21594646573066711, + "learning_rate": 5.726409456620841e-07, + "loss": 0.0223, + "step": 95420 + }, + { + "epoch": 18.52657637978739, + "grad_norm": 0.21968136727809906, + "learning_rate": 5.701483618494541e-07, + "loss": 0.0218, + "step": 95430 + }, + { + "epoch": 18.52851803310519, + "grad_norm": 0.332126259803772, + "learning_rate": 5.67661183611079e-07, + "loss": 0.0215, + "step": 95440 + }, + { + "epoch": 18.53045968642299, + "grad_norm": 0.06985297799110413, + "learning_rate": 5.651794112189579e-07, + "loss": 0.0294, + "step": 95450 + }, + { + "epoch": 18.53240133974079, + "grad_norm": 0.08145968616008759, + "learning_rate": 5.627030449444903e-07, + "loss": 0.0169, + "step": 95460 + }, + { + "epoch": 18.53434299305859, + "grad_norm": 0.18935950100421906, + "learning_rate": 5.60232085058493e-07, + "loss": 0.0197, + "step": 95470 + }, + { + "epoch": 18.53628464637639, + "grad_norm": 0.14483670890331268, + "learning_rate": 5.577665318311776e-07, + "loss": 0.0192, + "step": 95480 + }, + { + "epoch": 18.53822629969419, + "grad_norm": 0.14849604666233063, + "learning_rate": 5.553063855321838e-07, + "loss": 0.0226, + "step": 95490 + }, + { + "epoch": 18.54016795301199, + "grad_norm": 0.15358349680900574, + "learning_rate": 5.528516464305467e-07, + "loss": 0.0155, + "step": 95500 + }, + { + "epoch": 18.542109606329788, + "grad_norm": 0.15012840926647186, + "learning_rate": 5.504023147947068e-07, + "loss": 0.0173, + "step": 95510 + }, + { + "epoch": 18.54405125964759, + "grad_norm": 0.11074108630418777, + "learning_rate": 5.479583908925279e-07, + "loss": 0.0224, + "step": 95520 + }, + { + "epoch": 18.54599291296539, + "grad_norm": 0.18893076479434967, + "learning_rate": 5.455198749912626e-07, + "loss": 0.0212, + "step": 95530 + }, + { + "epoch": 18.54793456628319, + "grad_norm": 0.30609574913978577, + "learning_rate": 5.430867673575979e-07, + "loss": 0.0207, + "step": 95540 + }, + { + "epoch": 18.54987621960099, + "grad_norm": 0.19155412912368774, + "learning_rate": 5.406590682575985e-07, + "loss": 0.0204, + "step": 95550 + }, + { + "epoch": 18.55181787291879, + "grad_norm": 0.18329833447933197, + "learning_rate": 5.382367779567687e-07, + "loss": 0.0218, + "step": 95560 + }, + { + "epoch": 18.55375952623659, + "grad_norm": 0.16096438467502594, + "learning_rate": 5.358198967200023e-07, + "loss": 0.0198, + "step": 95570 + }, + { + "epoch": 18.55570117955439, + "grad_norm": 0.10874314606189728, + "learning_rate": 5.334084248115989e-07, + "loss": 0.0183, + "step": 95580 + }, + { + "epoch": 18.55764283287219, + "grad_norm": 0.25527331233024597, + "learning_rate": 5.310023624952754e-07, + "loss": 0.0224, + "step": 95590 + }, + { + "epoch": 18.55958448618999, + "grad_norm": 0.2437710016965866, + "learning_rate": 5.286017100341656e-07, + "loss": 0.0251, + "step": 95600 + }, + { + "epoch": 18.56152613950779, + "grad_norm": 0.15085187554359436, + "learning_rate": 5.262064676907874e-07, + "loss": 0.0188, + "step": 95610 + }, + { + "epoch": 18.56346779282559, + "grad_norm": 0.2642273008823395, + "learning_rate": 5.238166357270924e-07, + "loss": 0.0251, + "step": 95620 + }, + { + "epoch": 18.56540944614339, + "grad_norm": 0.1578921526670456, + "learning_rate": 5.214322144044215e-07, + "loss": 0.0183, + "step": 95630 + }, + { + "epoch": 18.56735109946119, + "grad_norm": 0.1922447383403778, + "learning_rate": 5.190532039835327e-07, + "loss": 0.0214, + "step": 95640 + }, + { + "epoch": 18.569292752778992, + "grad_norm": 0.1502421647310257, + "learning_rate": 5.166796047245903e-07, + "loss": 0.02, + "step": 95650 + }, + { + "epoch": 18.571234406096792, + "grad_norm": 0.23630894720554352, + "learning_rate": 5.143114168871699e-07, + "loss": 0.0204, + "step": 95660 + }, + { + "epoch": 18.573176059414592, + "grad_norm": 0.1663961410522461, + "learning_rate": 5.119486407302532e-07, + "loss": 0.0169, + "step": 95670 + }, + { + "epoch": 18.575117712732393, + "grad_norm": 0.1354585886001587, + "learning_rate": 5.095912765122335e-07, + "loss": 0.0158, + "step": 95680 + }, + { + "epoch": 18.577059366050193, + "grad_norm": 0.15921631455421448, + "learning_rate": 5.07239324490899e-07, + "loss": 0.0191, + "step": 95690 + }, + { + "epoch": 18.579001019367993, + "grad_norm": 0.175084188580513, + "learning_rate": 5.048927849234608e-07, + "loss": 0.0228, + "step": 95700 + }, + { + "epoch": 18.580942672685794, + "grad_norm": 0.19928085803985596, + "learning_rate": 5.025516580665358e-07, + "loss": 0.0227, + "step": 95710 + }, + { + "epoch": 18.582884326003594, + "grad_norm": 0.121269591152668, + "learning_rate": 5.002159441761412e-07, + "loss": 0.0193, + "step": 95720 + }, + { + "epoch": 18.58482597932139, + "grad_norm": 0.19075584411621094, + "learning_rate": 4.978856435077062e-07, + "loss": 0.0241, + "step": 95730 + }, + { + "epoch": 18.58676763263919, + "grad_norm": 0.18667639791965485, + "learning_rate": 4.955607563160769e-07, + "loss": 0.0256, + "step": 95740 + }, + { + "epoch": 18.58870928595699, + "grad_norm": 0.19305036962032318, + "learning_rate": 4.932412828554944e-07, + "loss": 0.0206, + "step": 95750 + }, + { + "epoch": 18.59065093927479, + "grad_norm": 0.2430868148803711, + "learning_rate": 4.909272233796114e-07, + "loss": 0.0219, + "step": 95760 + }, + { + "epoch": 18.59259259259259, + "grad_norm": 0.1477147936820984, + "learning_rate": 4.88618578141492e-07, + "loss": 0.023, + "step": 95770 + }, + { + "epoch": 18.594534245910392, + "grad_norm": 0.2308138608932495, + "learning_rate": 4.863153473936011e-07, + "loss": 0.0199, + "step": 95780 + }, + { + "epoch": 18.596475899228192, + "grad_norm": 0.2191038578748703, + "learning_rate": 4.840175313878259e-07, + "loss": 0.0191, + "step": 95790 + }, + { + "epoch": 18.598417552545992, + "grad_norm": 0.11381988227367401, + "learning_rate": 4.817251303754433e-07, + "loss": 0.0226, + "step": 95800 + }, + { + "epoch": 18.600359205863793, + "grad_norm": 0.2130480855703354, + "learning_rate": 4.794381446071583e-07, + "loss": 0.023, + "step": 95810 + }, + { + "epoch": 18.602300859181593, + "grad_norm": 0.19618529081344604, + "learning_rate": 4.771565743330597e-07, + "loss": 0.0236, + "step": 95820 + }, + { + "epoch": 18.604242512499393, + "grad_norm": 0.16432158648967743, + "learning_rate": 4.7488041980266463e-07, + "loss": 0.0232, + "step": 95830 + }, + { + "epoch": 18.606184165817194, + "grad_norm": 0.23426632583141327, + "learning_rate": 4.726096812648795e-07, + "loss": 0.0231, + "step": 95840 + }, + { + "epoch": 18.608125819134994, + "grad_norm": 0.10844091325998306, + "learning_rate": 4.703443589680445e-07, + "loss": 0.0204, + "step": 95850 + }, + { + "epoch": 18.610067472452794, + "grad_norm": 0.16054849326610565, + "learning_rate": 4.6808445315987826e-07, + "loss": 0.0237, + "step": 95860 + }, + { + "epoch": 18.612009125770594, + "grad_norm": 0.290009468793869, + "learning_rate": 4.658299640875274e-07, + "loss": 0.0316, + "step": 95870 + }, + { + "epoch": 18.613950779088395, + "grad_norm": 0.08116474002599716, + "learning_rate": 4.635808919975393e-07, + "loss": 0.0251, + "step": 95880 + }, + { + "epoch": 18.615892432406195, + "grad_norm": 0.2252039909362793, + "learning_rate": 4.6133723713586706e-07, + "loss": 0.0226, + "step": 95890 + }, + { + "epoch": 18.617834085723995, + "grad_norm": 0.07795736938714981, + "learning_rate": 4.590989997478756e-07, + "loss": 0.0175, + "step": 95900 + }, + { + "epoch": 18.619775739041796, + "grad_norm": 0.22040610015392303, + "learning_rate": 4.5686618007833027e-07, + "loss": 0.0223, + "step": 95910 + }, + { + "epoch": 18.621717392359596, + "grad_norm": 0.1543586403131485, + "learning_rate": 4.54638778371419e-07, + "loss": 0.0251, + "step": 95920 + }, + { + "epoch": 18.623659045677392, + "grad_norm": 0.16191112995147705, + "learning_rate": 4.524167948707192e-07, + "loss": 0.0213, + "step": 95930 + }, + { + "epoch": 18.625600698995193, + "grad_norm": 0.21048589050769806, + "learning_rate": 4.5020022981921986e-07, + "loss": 0.0181, + "step": 95940 + }, + { + "epoch": 18.627542352312993, + "grad_norm": 0.12681828439235687, + "learning_rate": 4.47989083459327e-07, + "loss": 0.0211, + "step": 95950 + }, + { + "epoch": 18.629484005630793, + "grad_norm": 0.21754348278045654, + "learning_rate": 4.457833560328528e-07, + "loss": 0.0181, + "step": 95960 + }, + { + "epoch": 18.631425658948594, + "grad_norm": 0.09822071343660355, + "learning_rate": 4.4358304778100434e-07, + "loss": 0.02, + "step": 95970 + }, + { + "epoch": 18.633367312266394, + "grad_norm": 0.1867024153470993, + "learning_rate": 4.413881589444058e-07, + "loss": 0.0198, + "step": 95980 + }, + { + "epoch": 18.635308965584194, + "grad_norm": 0.2548544108867645, + "learning_rate": 4.391986897630873e-07, + "loss": 0.0213, + "step": 95990 + }, + { + "epoch": 18.637250618901994, + "grad_norm": 0.25729793310165405, + "learning_rate": 4.370146404764908e-07, + "loss": 0.0175, + "step": 96000 + }, + { + "epoch": 18.639192272219795, + "grad_norm": 0.10835833102464676, + "learning_rate": 4.3483601132345285e-07, + "loss": 0.0192, + "step": 96010 + }, + { + "epoch": 18.641133925537595, + "grad_norm": 0.2462715208530426, + "learning_rate": 4.32662802542233e-07, + "loss": 0.0251, + "step": 96020 + }, + { + "epoch": 18.643075578855395, + "grad_norm": 0.14670926332473755, + "learning_rate": 4.3049501437047444e-07, + "loss": 0.0217, + "step": 96030 + }, + { + "epoch": 18.645017232173196, + "grad_norm": 0.2955728769302368, + "learning_rate": 4.283326470452653e-07, + "loss": 0.0229, + "step": 96040 + }, + { + "epoch": 18.646958885490996, + "grad_norm": 0.14924228191375732, + "learning_rate": 4.2617570080306647e-07, + "loss": 0.023, + "step": 96050 + }, + { + "epoch": 18.648900538808796, + "grad_norm": 0.2080947458744049, + "learning_rate": 4.2402417587975585e-07, + "loss": 0.0206, + "step": 96060 + }, + { + "epoch": 18.650842192126596, + "grad_norm": 0.1682354360818863, + "learning_rate": 4.218780725106286e-07, + "loss": 0.0202, + "step": 96070 + }, + { + "epoch": 18.652783845444397, + "grad_norm": 0.14555686712265015, + "learning_rate": 4.197373909303748e-07, + "loss": 0.0225, + "step": 96080 + }, + { + "epoch": 18.654725498762197, + "grad_norm": 0.1873752623796463, + "learning_rate": 4.17602131373096e-07, + "loss": 0.0163, + "step": 96090 + }, + { + "epoch": 18.656667152079997, + "grad_norm": 0.12146338075399399, + "learning_rate": 4.154722940723055e-07, + "loss": 0.023, + "step": 96100 + }, + { + "epoch": 18.658608805397797, + "grad_norm": 0.14366592466831207, + "learning_rate": 4.1334787926091136e-07, + "loss": 0.0213, + "step": 96110 + }, + { + "epoch": 18.660550458715598, + "grad_norm": 0.11858862638473511, + "learning_rate": 4.1122888717124444e-07, + "loss": 0.0172, + "step": 96120 + }, + { + "epoch": 18.662492112033398, + "grad_norm": 0.1947420984506607, + "learning_rate": 4.091153180350249e-07, + "loss": 0.0197, + "step": 96130 + }, + { + "epoch": 18.6644337653512, + "grad_norm": 0.18328922986984253, + "learning_rate": 4.0700717208340125e-07, + "loss": 0.0234, + "step": 96140 + }, + { + "epoch": 18.666375418668995, + "grad_norm": 0.16596569120883942, + "learning_rate": 4.049044495469112e-07, + "loss": 0.0218, + "step": 96150 + }, + { + "epoch": 18.668317071986795, + "grad_norm": 0.18409685790538788, + "learning_rate": 4.0280715065549866e-07, + "loss": 0.0158, + "step": 96160 + }, + { + "epoch": 18.670258725304596, + "grad_norm": 0.20996616780757904, + "learning_rate": 4.0071527563853575e-07, + "loss": 0.0223, + "step": 96170 + }, + { + "epoch": 18.672200378622396, + "grad_norm": 0.12975898385047913, + "learning_rate": 3.9862882472477823e-07, + "loss": 0.0235, + "step": 96180 + }, + { + "epoch": 18.674142031940196, + "grad_norm": 0.17648182809352875, + "learning_rate": 3.9654779814239375e-07, + "loss": 0.0242, + "step": 96190 + }, + { + "epoch": 18.676083685257996, + "grad_norm": 0.12512607872486115, + "learning_rate": 3.944721961189668e-07, + "loss": 0.0235, + "step": 96200 + }, + { + "epoch": 18.678025338575797, + "grad_norm": 0.2443738877773285, + "learning_rate": 3.9240201888147697e-07, + "loss": 0.0195, + "step": 96210 + }, + { + "epoch": 18.679966991893597, + "grad_norm": 0.1881607472896576, + "learning_rate": 3.9033726665632096e-07, + "loss": 0.0171, + "step": 96220 + }, + { + "epoch": 18.681908645211397, + "grad_norm": 0.1882072240114212, + "learning_rate": 3.882779396692959e-07, + "loss": 0.0187, + "step": 96230 + }, + { + "epoch": 18.683850298529197, + "grad_norm": 0.21914859116077423, + "learning_rate": 3.8622403814559947e-07, + "loss": 0.0265, + "step": 96240 + }, + { + "epoch": 18.685791951846998, + "grad_norm": 0.1533639132976532, + "learning_rate": 3.841755623098575e-07, + "loss": 0.0201, + "step": 96250 + }, + { + "epoch": 18.687733605164798, + "grad_norm": 0.16054095327854156, + "learning_rate": 3.821325123860742e-07, + "loss": 0.023, + "step": 96260 + }, + { + "epoch": 18.6896752584826, + "grad_norm": 0.1279592663049698, + "learning_rate": 3.800948885976818e-07, + "loss": 0.0254, + "step": 96270 + }, + { + "epoch": 18.6916169118004, + "grad_norm": 0.171452596783638, + "learning_rate": 3.780626911675078e-07, + "loss": 0.0213, + "step": 96280 + }, + { + "epoch": 18.6935585651182, + "grad_norm": 0.17692792415618896, + "learning_rate": 3.76035920317791e-07, + "loss": 0.0206, + "step": 96290 + }, + { + "epoch": 18.695500218436, + "grad_norm": 0.26402053236961365, + "learning_rate": 3.740145762701819e-07, + "loss": 0.0193, + "step": 96300 + }, + { + "epoch": 18.6974418717538, + "grad_norm": 0.19240885972976685, + "learning_rate": 3.7199865924572585e-07, + "loss": 0.0194, + "step": 96310 + }, + { + "epoch": 18.6993835250716, + "grad_norm": 0.17913688719272614, + "learning_rate": 3.699881694648799e-07, + "loss": 0.0244, + "step": 96320 + }, + { + "epoch": 18.7013251783894, + "grad_norm": 0.18579964339733124, + "learning_rate": 3.679831071475126e-07, + "loss": 0.0192, + "step": 96330 + }, + { + "epoch": 18.7032668317072, + "grad_norm": 0.15934747457504272, + "learning_rate": 3.65983472512893e-07, + "loss": 0.0212, + "step": 96340 + }, + { + "epoch": 18.705208485025, + "grad_norm": 0.09420603513717651, + "learning_rate": 3.6398926577969616e-07, + "loss": 0.0168, + "step": 96350 + }, + { + "epoch": 18.707150138342797, + "grad_norm": 0.24465946853160858, + "learning_rate": 3.620004871660032e-07, + "loss": 0.0236, + "step": 96360 + }, + { + "epoch": 18.709091791660597, + "grad_norm": 0.21097487211227417, + "learning_rate": 3.600171368893124e-07, + "loss": 0.0139, + "step": 96370 + }, + { + "epoch": 18.711033444978398, + "grad_norm": 0.11928018182516098, + "learning_rate": 3.580392151665113e-07, + "loss": 0.0237, + "step": 96380 + }, + { + "epoch": 18.712975098296198, + "grad_norm": 0.1149793341755867, + "learning_rate": 3.5606672221391024e-07, + "loss": 0.0255, + "step": 96390 + }, + { + "epoch": 18.714916751614, + "grad_norm": 0.1498190313577652, + "learning_rate": 3.540996582472089e-07, + "loss": 0.0214, + "step": 96400 + }, + { + "epoch": 18.7168584049318, + "grad_norm": 0.17387385666370392, + "learning_rate": 3.521380234815297e-07, + "loss": 0.0165, + "step": 96410 + }, + { + "epoch": 18.7188000582496, + "grad_norm": 0.22507278621196747, + "learning_rate": 3.5018181813138985e-07, + "loss": 0.0253, + "step": 96420 + }, + { + "epoch": 18.7207417115674, + "grad_norm": 0.09666958451271057, + "learning_rate": 3.482310424107238e-07, + "loss": 0.0194, + "step": 96430 + }, + { + "epoch": 18.7226833648852, + "grad_norm": 0.32585522532463074, + "learning_rate": 3.4628569653285537e-07, + "loss": 0.0222, + "step": 96440 + }, + { + "epoch": 18.724625018203, + "grad_norm": 0.10769928246736526, + "learning_rate": 3.44345780710531e-07, + "loss": 0.0304, + "step": 96450 + }, + { + "epoch": 18.7265666715208, + "grad_norm": 0.17696183919906616, + "learning_rate": 3.424112951558978e-07, + "loss": 0.0147, + "step": 96460 + }, + { + "epoch": 18.7285083248386, + "grad_norm": 0.18949542939662933, + "learning_rate": 3.40482240080503e-07, + "loss": 0.0158, + "step": 96470 + }, + { + "epoch": 18.7304499781564, + "grad_norm": 0.16613979637622833, + "learning_rate": 3.3855861569531133e-07, + "loss": 0.0218, + "step": 96480 + }, + { + "epoch": 18.7323916314742, + "grad_norm": 0.17699293792247772, + "learning_rate": 3.3664042221068226e-07, + "loss": 0.0222, + "step": 96490 + }, + { + "epoch": 18.734333284792, + "grad_norm": 0.11247818171977997, + "learning_rate": 3.347276598363924e-07, + "loss": 0.022, + "step": 96500 + }, + { + "epoch": 18.7362749381098, + "grad_norm": 0.061888162046670914, + "learning_rate": 3.328203287816134e-07, + "loss": 0.0183, + "step": 96510 + }, + { + "epoch": 18.7382165914276, + "grad_norm": 0.19336889684200287, + "learning_rate": 3.3091842925493386e-07, + "loss": 0.0182, + "step": 96520 + }, + { + "epoch": 18.740158244745402, + "grad_norm": 0.24201321601867676, + "learning_rate": 3.2902196146433196e-07, + "loss": 0.025, + "step": 96530 + }, + { + "epoch": 18.742099898063202, + "grad_norm": 0.3061988055706024, + "learning_rate": 3.2713092561720836e-07, + "loss": 0.0206, + "step": 96540 + }, + { + "epoch": 18.744041551381002, + "grad_norm": 0.17315149307250977, + "learning_rate": 3.2524532192036996e-07, + "loss": 0.0217, + "step": 96550 + }, + { + "epoch": 18.745983204698803, + "grad_norm": 0.2432393729686737, + "learning_rate": 3.2336515058001836e-07, + "loss": 0.0249, + "step": 96560 + }, + { + "epoch": 18.747924858016603, + "grad_norm": 0.12313227355480194, + "learning_rate": 3.2149041180176143e-07, + "loss": 0.0193, + "step": 96570 + }, + { + "epoch": 18.7498665113344, + "grad_norm": 0.1450788527727127, + "learning_rate": 3.1962110579062954e-07, + "loss": 0.0178, + "step": 96580 + }, + { + "epoch": 18.7518081646522, + "grad_norm": 0.11810564994812012, + "learning_rate": 3.177572327510314e-07, + "loss": 0.0184, + "step": 96590 + }, + { + "epoch": 18.75374981797, + "grad_norm": 0.16801434755325317, + "learning_rate": 3.158987928868151e-07, + "loss": 0.0216, + "step": 96600 + }, + { + "epoch": 18.7556914712878, + "grad_norm": 0.30873990058898926, + "learning_rate": 3.1404578640120144e-07, + "loss": 0.0219, + "step": 96610 + }, + { + "epoch": 18.7576331246056, + "grad_norm": 0.12081781774759293, + "learning_rate": 3.1219821349683933e-07, + "loss": 0.0155, + "step": 96620 + }, + { + "epoch": 18.7595747779234, + "grad_norm": 0.1798660010099411, + "learning_rate": 3.103560743757783e-07, + "loss": 0.0265, + "step": 96630 + }, + { + "epoch": 18.7615164312412, + "grad_norm": 0.14395196735858917, + "learning_rate": 3.0851936923946834e-07, + "loss": 0.0258, + "step": 96640 + }, + { + "epoch": 18.763458084559, + "grad_norm": 0.132935032248497, + "learning_rate": 3.06688098288771e-07, + "loss": 0.0253, + "step": 96650 + }, + { + "epoch": 18.765399737876802, + "grad_norm": 0.4117497503757477, + "learning_rate": 3.0486226172394825e-07, + "loss": 0.0154, + "step": 96660 + }, + { + "epoch": 18.767341391194602, + "grad_norm": 0.21939612925052643, + "learning_rate": 3.030418597446738e-07, + "loss": 0.0243, + "step": 96670 + }, + { + "epoch": 18.769283044512402, + "grad_norm": 0.2788102626800537, + "learning_rate": 3.012268925500272e-07, + "loss": 0.0254, + "step": 96680 + }, + { + "epoch": 18.771224697830203, + "grad_norm": 0.18877211213111877, + "learning_rate": 2.994173603384831e-07, + "loss": 0.0238, + "step": 96690 + }, + { + "epoch": 18.773166351148003, + "grad_norm": 0.1551094800233841, + "learning_rate": 2.9761326330793317e-07, + "loss": 0.0265, + "step": 96700 + }, + { + "epoch": 18.775108004465803, + "grad_norm": 0.12107528001070023, + "learning_rate": 2.958146016556751e-07, + "loss": 0.0262, + "step": 96710 + }, + { + "epoch": 18.777049657783603, + "grad_norm": 0.27568283677101135, + "learning_rate": 2.94021375578396e-07, + "loss": 0.0237, + "step": 96720 + }, + { + "epoch": 18.778991311101404, + "grad_norm": 0.16919048130512238, + "learning_rate": 2.9223358527221135e-07, + "loss": 0.02, + "step": 96730 + }, + { + "epoch": 18.780932964419204, + "grad_norm": 0.2926575243473053, + "learning_rate": 2.904512309326257e-07, + "loss": 0.0254, + "step": 96740 + }, + { + "epoch": 18.782874617737004, + "grad_norm": 0.15460123121738434, + "learning_rate": 2.8867431275455545e-07, + "loss": 0.0162, + "step": 96750 + }, + { + "epoch": 18.784816271054805, + "grad_norm": 0.16845907270908356, + "learning_rate": 2.86902830932323e-07, + "loss": 0.0189, + "step": 96760 + }, + { + "epoch": 18.786757924372605, + "grad_norm": 0.18073847889900208, + "learning_rate": 2.851367856596565e-07, + "loss": 0.0187, + "step": 96770 + }, + { + "epoch": 18.7886995776904, + "grad_norm": 0.20161215960979462, + "learning_rate": 2.83376177129685e-07, + "loss": 0.0206, + "step": 96780 + }, + { + "epoch": 18.790641231008202, + "grad_norm": 0.25097712874412537, + "learning_rate": 2.8162100553494884e-07, + "loss": 0.0261, + "step": 96790 + }, + { + "epoch": 18.792582884326002, + "grad_norm": 0.1357235163450241, + "learning_rate": 2.798712710673834e-07, + "loss": 0.0204, + "step": 96800 + }, + { + "epoch": 18.794524537643802, + "grad_norm": 0.06777311861515045, + "learning_rate": 2.7812697391834676e-07, + "loss": 0.0215, + "step": 96810 + }, + { + "epoch": 18.796466190961603, + "grad_norm": 0.20424316823482513, + "learning_rate": 2.7638811427858627e-07, + "loss": 0.0201, + "step": 96820 + }, + { + "epoch": 18.798407844279403, + "grad_norm": 0.24733754992485046, + "learning_rate": 2.746546923382665e-07, + "loss": 0.0247, + "step": 96830 + }, + { + "epoch": 18.800349497597203, + "grad_norm": 0.1860211044549942, + "learning_rate": 2.729267082869469e-07, + "loss": 0.0187, + "step": 96840 + }, + { + "epoch": 18.802291150915003, + "grad_norm": 0.31234580278396606, + "learning_rate": 2.712041623135986e-07, + "loss": 0.0261, + "step": 96850 + }, + { + "epoch": 18.804232804232804, + "grad_norm": 0.21740610897541046, + "learning_rate": 2.694870546065931e-07, + "loss": 0.0227, + "step": 96860 + }, + { + "epoch": 18.806174457550604, + "grad_norm": 0.2505815625190735, + "learning_rate": 2.677753853537246e-07, + "loss": 0.0216, + "step": 96870 + }, + { + "epoch": 18.808116110868404, + "grad_norm": 0.2722338140010834, + "learning_rate": 2.660691547421601e-07, + "loss": 0.0185, + "step": 96880 + }, + { + "epoch": 18.810057764186205, + "grad_norm": 0.18868863582611084, + "learning_rate": 2.6436836295850033e-07, + "loss": 0.0292, + "step": 96890 + }, + { + "epoch": 18.811999417504005, + "grad_norm": 0.12099304795265198, + "learning_rate": 2.6267301018874093e-07, + "loss": 0.0183, + "step": 96900 + }, + { + "epoch": 18.813941070821805, + "grad_norm": 0.19523218274116516, + "learning_rate": 2.6098309661828355e-07, + "loss": 0.0264, + "step": 96910 + }, + { + "epoch": 18.815882724139605, + "grad_norm": 0.2153708040714264, + "learning_rate": 2.592986224319305e-07, + "loss": 0.0217, + "step": 96920 + }, + { + "epoch": 18.817824377457406, + "grad_norm": 0.07482931017875671, + "learning_rate": 2.576195878139009e-07, + "loss": 0.0164, + "step": 96930 + }, + { + "epoch": 18.819766030775206, + "grad_norm": 0.06592654436826706, + "learning_rate": 2.5594599294780364e-07, + "loss": 0.0174, + "step": 96940 + }, + { + "epoch": 18.821707684093006, + "grad_norm": 0.2363724559545517, + "learning_rate": 2.542778380166644e-07, + "loss": 0.0233, + "step": 96950 + }, + { + "epoch": 18.823649337410806, + "grad_norm": 0.14211533963680267, + "learning_rate": 2.526151232029095e-07, + "loss": 0.0185, + "step": 96960 + }, + { + "epoch": 18.825590990728607, + "grad_norm": 0.1384066492319107, + "learning_rate": 2.5095784868836567e-07, + "loss": 0.0154, + "step": 96970 + }, + { + "epoch": 18.827532644046407, + "grad_norm": 0.06550666689872742, + "learning_rate": 2.493060146542825e-07, + "loss": 0.019, + "step": 96980 + }, + { + "epoch": 18.829474297364207, + "grad_norm": 0.22618380188941956, + "learning_rate": 2.476596212812876e-07, + "loss": 0.0206, + "step": 96990 + }, + { + "epoch": 18.831415950682004, + "grad_norm": 0.22728331387043, + "learning_rate": 2.4601866874943703e-07, + "loss": 0.0191, + "step": 97000 + }, + { + "epoch": 18.833357603999804, + "grad_norm": 0.23845113813877106, + "learning_rate": 2.4438315723818163e-07, + "loss": 0.0276, + "step": 97010 + }, + { + "epoch": 18.835299257317605, + "grad_norm": 0.25218045711517334, + "learning_rate": 2.4275308692637833e-07, + "loss": 0.0193, + "step": 97020 + }, + { + "epoch": 18.837240910635405, + "grad_norm": 0.1982644498348236, + "learning_rate": 2.411284579922901e-07, + "loss": 0.0212, + "step": 97030 + }, + { + "epoch": 18.839182563953205, + "grad_norm": 0.11019981652498245, + "learning_rate": 2.395092706135804e-07, + "loss": 0.0215, + "step": 97040 + }, + { + "epoch": 18.841124217271005, + "grad_norm": 0.17301751673221588, + "learning_rate": 2.3789552496731872e-07, + "loss": 0.0203, + "step": 97050 + }, + { + "epoch": 18.843065870588806, + "grad_norm": 0.1743253767490387, + "learning_rate": 2.3628722122999158e-07, + "loss": 0.0215, + "step": 97060 + }, + { + "epoch": 18.845007523906606, + "grad_norm": 0.1417521834373474, + "learning_rate": 2.3468435957747503e-07, + "loss": 0.0191, + "step": 97070 + }, + { + "epoch": 18.846949177224406, + "grad_norm": 0.1904975324869156, + "learning_rate": 2.330869401850566e-07, + "loss": 0.0208, + "step": 97080 + }, + { + "epoch": 18.848890830542206, + "grad_norm": 0.18272152543067932, + "learning_rate": 2.3149496322742437e-07, + "loss": 0.0237, + "step": 97090 + }, + { + "epoch": 18.850832483860007, + "grad_norm": 0.26530882716178894, + "learning_rate": 2.2990842887868346e-07, + "loss": 0.0211, + "step": 97100 + }, + { + "epoch": 18.852774137177807, + "grad_norm": 0.2132236808538437, + "learning_rate": 2.2832733731232292e-07, + "loss": 0.0205, + "step": 97110 + }, + { + "epoch": 18.854715790495607, + "grad_norm": 0.105011485517025, + "learning_rate": 2.2675168870124886e-07, + "loss": 0.0206, + "step": 97120 + }, + { + "epoch": 18.856657443813408, + "grad_norm": 0.22367647290229797, + "learning_rate": 2.2518148321778455e-07, + "loss": 0.0207, + "step": 97130 + }, + { + "epoch": 18.858599097131208, + "grad_norm": 0.18221347033977509, + "learning_rate": 2.2361672103363706e-07, + "loss": 0.0207, + "step": 97140 + }, + { + "epoch": 18.860540750449008, + "grad_norm": 0.1379469484090805, + "learning_rate": 2.220574023199251e-07, + "loss": 0.0193, + "step": 97150 + }, + { + "epoch": 18.86248240376681, + "grad_norm": 0.18288032710552216, + "learning_rate": 2.2050352724717894e-07, + "loss": 0.0219, + "step": 97160 + }, + { + "epoch": 18.86442405708461, + "grad_norm": 0.28152382373809814, + "learning_rate": 2.1895509598532372e-07, + "loss": 0.0236, + "step": 97170 + }, + { + "epoch": 18.86636571040241, + "grad_norm": 0.09255489706993103, + "learning_rate": 2.1741210870369066e-07, + "loss": 0.0224, + "step": 97180 + }, + { + "epoch": 18.86830736372021, + "grad_norm": 0.11387412995100021, + "learning_rate": 2.1587456557102814e-07, + "loss": 0.025, + "step": 97190 + }, + { + "epoch": 18.87024901703801, + "grad_norm": 0.17528383433818817, + "learning_rate": 2.143424667554683e-07, + "loss": 0.0235, + "step": 97200 + }, + { + "epoch": 18.872190670355806, + "grad_norm": 0.1871112734079361, + "learning_rate": 2.128158124245716e-07, + "loss": 0.0226, + "step": 97210 + }, + { + "epoch": 18.874132323673606, + "grad_norm": 0.0710844099521637, + "learning_rate": 2.1129460274527668e-07, + "loss": 0.0212, + "step": 97220 + }, + { + "epoch": 18.876073976991407, + "grad_norm": 0.19182635843753815, + "learning_rate": 2.0977883788395046e-07, + "loss": 0.0356, + "step": 97230 + }, + { + "epoch": 18.878015630309207, + "grad_norm": 0.23203858733177185, + "learning_rate": 2.0826851800634928e-07, + "loss": 0.0186, + "step": 97240 + }, + { + "epoch": 18.879957283627007, + "grad_norm": 0.20655512809753418, + "learning_rate": 2.0676364327764653e-07, + "loss": 0.0224, + "step": 97250 + }, + { + "epoch": 18.881898936944808, + "grad_norm": 0.1524290293455124, + "learning_rate": 2.0526421386240502e-07, + "loss": 0.0225, + "step": 97260 + }, + { + "epoch": 18.883840590262608, + "grad_norm": 0.25949907302856445, + "learning_rate": 2.0377022992461025e-07, + "loss": 0.0263, + "step": 97270 + }, + { + "epoch": 18.885782243580408, + "grad_norm": 0.13318359851837158, + "learning_rate": 2.0228169162763156e-07, + "loss": 0.0241, + "step": 97280 + }, + { + "epoch": 18.88772389689821, + "grad_norm": 0.24925179779529572, + "learning_rate": 2.0079859913425536e-07, + "loss": 0.0223, + "step": 97290 + }, + { + "epoch": 18.88966555021601, + "grad_norm": 0.1867723912000656, + "learning_rate": 1.9932095260667417e-07, + "loss": 0.0171, + "step": 97300 + }, + { + "epoch": 18.89160720353381, + "grad_norm": 0.19721883535385132, + "learning_rate": 1.9784875220648092e-07, + "loss": 0.0171, + "step": 97310 + }, + { + "epoch": 18.89354885685161, + "grad_norm": 0.23996610939502716, + "learning_rate": 1.9638199809466907e-07, + "loss": 0.0217, + "step": 97320 + }, + { + "epoch": 18.89549051016941, + "grad_norm": 0.16650468111038208, + "learning_rate": 1.9492069043164364e-07, + "loss": 0.0216, + "step": 97330 + }, + { + "epoch": 18.89743216348721, + "grad_norm": 0.22034506499767303, + "learning_rate": 1.9346482937721567e-07, + "loss": 0.0213, + "step": 97340 + }, + { + "epoch": 18.89937381680501, + "grad_norm": 0.18482106924057007, + "learning_rate": 1.920144150905856e-07, + "loss": 0.0216, + "step": 97350 + }, + { + "epoch": 18.90131547012281, + "grad_norm": 0.09010034799575806, + "learning_rate": 1.9056944773037655e-07, + "loss": 0.0165, + "step": 97360 + }, + { + "epoch": 18.90325712344061, + "grad_norm": 0.11684071272611618, + "learning_rate": 1.8912992745460102e-07, + "loss": 0.0187, + "step": 97370 + }, + { + "epoch": 18.90519877675841, + "grad_norm": 0.18437941372394562, + "learning_rate": 1.876958544206886e-07, + "loss": 0.0233, + "step": 97380 + }, + { + "epoch": 18.90714043007621, + "grad_norm": 0.16843700408935547, + "learning_rate": 1.8626722878546942e-07, + "loss": 0.0272, + "step": 97390 + }, + { + "epoch": 18.90908208339401, + "grad_norm": 0.18167062103748322, + "learning_rate": 1.8484405070516851e-07, + "loss": 0.0229, + "step": 97400 + }, + { + "epoch": 18.91102373671181, + "grad_norm": 0.1928253471851349, + "learning_rate": 1.8342632033542805e-07, + "loss": 0.0159, + "step": 97410 + }, + { + "epoch": 18.912965390029612, + "grad_norm": 0.199929878115654, + "learning_rate": 1.8201403783129066e-07, + "loss": 0.0239, + "step": 97420 + }, + { + "epoch": 18.91490704334741, + "grad_norm": 0.14887934923171997, + "learning_rate": 1.806072033471884e-07, + "loss": 0.0254, + "step": 97430 + }, + { + "epoch": 18.91684869666521, + "grad_norm": 0.35160863399505615, + "learning_rate": 1.79205817036987e-07, + "loss": 0.0215, + "step": 97440 + }, + { + "epoch": 18.91879034998301, + "grad_norm": 0.10641259700059891, + "learning_rate": 1.7780987905393065e-07, + "loss": 0.0244, + "step": 97450 + }, + { + "epoch": 18.92073200330081, + "grad_norm": 0.17117460072040558, + "learning_rate": 1.764193895506805e-07, + "loss": 0.0241, + "step": 97460 + }, + { + "epoch": 18.92267365661861, + "grad_norm": 0.2324765920639038, + "learning_rate": 1.7503434867929824e-07, + "loss": 0.0255, + "step": 97470 + }, + { + "epoch": 18.92461530993641, + "grad_norm": 0.10777327418327332, + "learning_rate": 1.7365475659125165e-07, + "loss": 0.0214, + "step": 97480 + }, + { + "epoch": 18.92655696325421, + "grad_norm": 0.13908100128173828, + "learning_rate": 1.7228061343740332e-07, + "loss": 0.0217, + "step": 97490 + }, + { + "epoch": 18.92849861657201, + "grad_norm": 0.1761026233434677, + "learning_rate": 1.7091191936803308e-07, + "loss": 0.0223, + "step": 97500 + }, + { + "epoch": 18.93044026988981, + "grad_norm": 0.22842535376548767, + "learning_rate": 1.6954867453281563e-07, + "loss": 0.0196, + "step": 97510 + }, + { + "epoch": 18.93238192320761, + "grad_norm": 0.15655022859573364, + "learning_rate": 1.6819087908084285e-07, + "loss": 0.0163, + "step": 97520 + }, + { + "epoch": 18.93432357652541, + "grad_norm": 0.13168860971927643, + "learning_rate": 1.6683853316058485e-07, + "loss": 0.0218, + "step": 97530 + }, + { + "epoch": 18.93626522984321, + "grad_norm": 0.1845484972000122, + "learning_rate": 1.6549163691995106e-07, + "loss": 0.0213, + "step": 97540 + }, + { + "epoch": 18.938206883161012, + "grad_norm": 0.17648491263389587, + "learning_rate": 1.641501905062237e-07, + "loss": 0.0177, + "step": 97550 + }, + { + "epoch": 18.940148536478812, + "grad_norm": 0.14069876074790955, + "learning_rate": 1.628141940661021e-07, + "loss": 0.0197, + "step": 97560 + }, + { + "epoch": 18.942090189796613, + "grad_norm": 0.13690391182899475, + "learning_rate": 1.6148364774569158e-07, + "loss": 0.0197, + "step": 97570 + }, + { + "epoch": 18.944031843114413, + "grad_norm": 0.12914229929447174, + "learning_rate": 1.6015855169050355e-07, + "loss": 0.0242, + "step": 97580 + }, + { + "epoch": 18.945973496432213, + "grad_norm": 0.12433923780918121, + "learning_rate": 1.5883890604543873e-07, + "loss": 0.0244, + "step": 97590 + }, + { + "epoch": 18.947915149750013, + "grad_norm": 0.20303404331207275, + "learning_rate": 1.5752471095482057e-07, + "loss": 0.0162, + "step": 97600 + }, + { + "epoch": 18.949856803067814, + "grad_norm": 0.23725832998752594, + "learning_rate": 1.5621596656235638e-07, + "loss": 0.0258, + "step": 97610 + }, + { + "epoch": 18.951798456385614, + "grad_norm": 0.2207658290863037, + "learning_rate": 1.5491267301117608e-07, + "loss": 0.0292, + "step": 97620 + }, + { + "epoch": 18.953740109703414, + "grad_norm": 0.16419945657253265, + "learning_rate": 1.536148304438101e-07, + "loss": 0.0244, + "step": 97630 + }, + { + "epoch": 18.95568176302121, + "grad_norm": 0.15548139810562134, + "learning_rate": 1.5232243900217823e-07, + "loss": 0.0225, + "step": 97640 + }, + { + "epoch": 18.95762341633901, + "grad_norm": 0.2851684093475342, + "learning_rate": 1.51035498827623e-07, + "loss": 0.0229, + "step": 97650 + }, + { + "epoch": 18.95956506965681, + "grad_norm": 0.20061998069286346, + "learning_rate": 1.497540100608763e-07, + "loss": 0.0215, + "step": 97660 + }, + { + "epoch": 18.96150672297461, + "grad_norm": 0.18809224665164948, + "learning_rate": 1.4847797284208153e-07, + "loss": 0.0214, + "step": 97670 + }, + { + "epoch": 18.963448376292412, + "grad_norm": 0.19118981063365936, + "learning_rate": 1.472073873107882e-07, + "loss": 0.019, + "step": 97680 + }, + { + "epoch": 18.965390029610212, + "grad_norm": 0.17738230526447296, + "learning_rate": 1.459422536059407e-07, + "loss": 0.0178, + "step": 97690 + }, + { + "epoch": 18.967331682928013, + "grad_norm": 0.13263994455337524, + "learning_rate": 1.4468257186589508e-07, + "loss": 0.0211, + "step": 97700 + }, + { + "epoch": 18.969273336245813, + "grad_norm": 0.23119091987609863, + "learning_rate": 1.4342834222840217e-07, + "loss": 0.0247, + "step": 97710 + }, + { + "epoch": 18.971214989563613, + "grad_norm": 0.18647868931293488, + "learning_rate": 1.4217956483063566e-07, + "loss": 0.0203, + "step": 97720 + }, + { + "epoch": 18.973156642881413, + "grad_norm": 0.1313183307647705, + "learning_rate": 1.409362398091474e-07, + "loss": 0.0253, + "step": 97730 + }, + { + "epoch": 18.975098296199214, + "grad_norm": 0.16385811567306519, + "learning_rate": 1.3969836729990638e-07, + "loss": 0.0225, + "step": 97740 + }, + { + "epoch": 18.977039949517014, + "grad_norm": 0.23773989081382751, + "learning_rate": 1.384659474382932e-07, + "loss": 0.0248, + "step": 97750 + }, + { + "epoch": 18.978981602834814, + "grad_norm": 0.14500980079174042, + "learning_rate": 1.372389803590779e-07, + "loss": 0.0256, + "step": 97760 + }, + { + "epoch": 18.980923256152614, + "grad_norm": 0.15029552578926086, + "learning_rate": 1.360174661964364e-07, + "loss": 0.0205, + "step": 97770 + }, + { + "epoch": 18.982864909470415, + "grad_norm": 0.18284447491168976, + "learning_rate": 1.3480140508396188e-07, + "loss": 0.0338, + "step": 97780 + }, + { + "epoch": 18.984806562788215, + "grad_norm": 0.18488238751888275, + "learning_rate": 1.335907971546313e-07, + "loss": 0.019, + "step": 97790 + }, + { + "epoch": 18.986748216106015, + "grad_norm": 0.20442290604114532, + "learning_rate": 1.3238564254083875e-07, + "loss": 0.0316, + "step": 97800 + }, + { + "epoch": 18.988689869423816, + "grad_norm": 0.20198661088943481, + "learning_rate": 1.3118594137437323e-07, + "loss": 0.0197, + "step": 97810 + }, + { + "epoch": 18.990631522741616, + "grad_norm": 0.16385608911514282, + "learning_rate": 1.2999169378644094e-07, + "loss": 0.0201, + "step": 97820 + }, + { + "epoch": 18.992573176059416, + "grad_norm": 0.10333547741174698, + "learning_rate": 1.2880289990763738e-07, + "loss": 0.0182, + "step": 97830 + }, + { + "epoch": 18.994514829377216, + "grad_norm": 0.18732035160064697, + "learning_rate": 1.2761955986796968e-07, + "loss": 0.0202, + "step": 97840 + }, + { + "epoch": 18.996456482695017, + "grad_norm": 0.2354864478111267, + "learning_rate": 1.264416737968399e-07, + "loss": 0.0182, + "step": 97850 + }, + { + "epoch": 18.998398136012813, + "grad_norm": 0.12769131362438202, + "learning_rate": 1.2526924182307275e-07, + "loss": 0.0187, + "step": 97860 + }, + { + "epoch": 19.00019416533178, + "grad_norm": 0.1404564380645752, + "learning_rate": 1.2410226407487124e-07, + "loss": 0.0165, + "step": 97870 + }, + { + "epoch": 19.00213581864958, + "grad_norm": 0.0705072283744812, + "learning_rate": 1.229407406798555e-07, + "loss": 0.0193, + "step": 97880 + }, + { + "epoch": 19.00407747196738, + "grad_norm": 0.10349749028682709, + "learning_rate": 1.2178467176505726e-07, + "loss": 0.0229, + "step": 97890 + }, + { + "epoch": 19.00601912528518, + "grad_norm": 0.2366885095834732, + "learning_rate": 1.2063405745689205e-07, + "loss": 0.0201, + "step": 97900 + }, + { + "epoch": 19.007960778602982, + "grad_norm": 0.2387586086988449, + "learning_rate": 1.1948889788119812e-07, + "loss": 0.0285, + "step": 97910 + }, + { + "epoch": 19.009902431920782, + "grad_norm": 0.10837316513061523, + "learning_rate": 1.1834919316320303e-07, + "loss": 0.0193, + "step": 97920 + }, + { + "epoch": 19.011844085238582, + "grad_norm": 0.12854595482349396, + "learning_rate": 1.1721494342754048e-07, + "loss": 0.0191, + "step": 97930 + }, + { + "epoch": 19.01378573855638, + "grad_norm": 0.2682092785835266, + "learning_rate": 1.1608614879825563e-07, + "loss": 0.0224, + "step": 97940 + }, + { + "epoch": 19.01572739187418, + "grad_norm": 0.16976037621498108, + "learning_rate": 1.1496280939879422e-07, + "loss": 0.0179, + "step": 97950 + }, + { + "epoch": 19.01766904519198, + "grad_norm": 0.1499042510986328, + "learning_rate": 1.1384492535199687e-07, + "loss": 0.0175, + "step": 97960 + }, + { + "epoch": 19.01961069850978, + "grad_norm": 0.15648917853832245, + "learning_rate": 1.1273249678011578e-07, + "loss": 0.0202, + "step": 97970 + }, + { + "epoch": 19.02155235182758, + "grad_norm": 0.14722630381584167, + "learning_rate": 1.1162552380480362e-07, + "loss": 0.0213, + "step": 97980 + }, + { + "epoch": 19.02349400514538, + "grad_norm": 0.18474549055099487, + "learning_rate": 1.1052400654711359e-07, + "loss": 0.0197, + "step": 97990 + }, + { + "epoch": 19.02543565846318, + "grad_norm": 0.22385169565677643, + "learning_rate": 1.0942794512751597e-07, + "loss": 0.0292, + "step": 98000 + }, + { + "epoch": 19.02737731178098, + "grad_norm": 0.2133379727602005, + "learning_rate": 1.0833733966587045e-07, + "loss": 0.0204, + "step": 98010 + }, + { + "epoch": 19.02931896509878, + "grad_norm": 0.10539011657238007, + "learning_rate": 1.0725219028143718e-07, + "loss": 0.0169, + "step": 98020 + }, + { + "epoch": 19.03126061841658, + "grad_norm": 0.20763175189495087, + "learning_rate": 1.0617249709289345e-07, + "loss": 0.0205, + "step": 98030 + }, + { + "epoch": 19.033202271734382, + "grad_norm": 0.175946906208992, + "learning_rate": 1.0509826021831703e-07, + "loss": 0.0228, + "step": 98040 + }, + { + "epoch": 19.035143925052182, + "grad_norm": 0.1278945803642273, + "learning_rate": 1.0402947977517508e-07, + "loss": 0.0251, + "step": 98050 + }, + { + "epoch": 19.037085578369982, + "grad_norm": 0.17536918818950653, + "learning_rate": 1.0296615588035185e-07, + "loss": 0.0234, + "step": 98060 + }, + { + "epoch": 19.039027231687783, + "grad_norm": 0.1792476922273636, + "learning_rate": 1.0190828865012659e-07, + "loss": 0.0203, + "step": 98070 + }, + { + "epoch": 19.040968885005583, + "grad_norm": 0.1782170683145523, + "learning_rate": 1.008558782001956e-07, + "loss": 0.0231, + "step": 98080 + }, + { + "epoch": 19.042910538323383, + "grad_norm": 0.19308583438396454, + "learning_rate": 9.980892464563906e-08, + "loss": 0.0187, + "step": 98090 + }, + { + "epoch": 19.044852191641183, + "grad_norm": 0.12511973083019257, + "learning_rate": 9.876742810095985e-08, + "loss": 0.0215, + "step": 98100 + }, + { + "epoch": 19.046793844958984, + "grad_norm": 0.2385575920343399, + "learning_rate": 9.77313886800446e-08, + "loss": 0.0211, + "step": 98110 + }, + { + "epoch": 19.048735498276784, + "grad_norm": 0.1478600949048996, + "learning_rate": 9.670080649619717e-08, + "loss": 0.0249, + "step": 98120 + }, + { + "epoch": 19.050677151594584, + "grad_norm": 0.17946752905845642, + "learning_rate": 9.567568166212737e-08, + "loss": 0.0184, + "step": 98130 + }, + { + "epoch": 19.052618804912385, + "grad_norm": 0.15606656670570374, + "learning_rate": 9.465601428992888e-08, + "loss": 0.0191, + "step": 98140 + }, + { + "epoch": 19.05456045823018, + "grad_norm": 0.10879402607679367, + "learning_rate": 9.364180449111803e-08, + "loss": 0.0155, + "step": 98150 + }, + { + "epoch": 19.05650211154798, + "grad_norm": 0.2766000032424927, + "learning_rate": 9.263305237661169e-08, + "loss": 0.0227, + "step": 98160 + }, + { + "epoch": 19.058443764865782, + "grad_norm": 0.15710121393203735, + "learning_rate": 9.162975805671603e-08, + "loss": 0.0171, + "step": 98170 + }, + { + "epoch": 19.060385418183582, + "grad_norm": 0.18005871772766113, + "learning_rate": 9.063192164115442e-08, + "loss": 0.0228, + "step": 98180 + }, + { + "epoch": 19.062327071501382, + "grad_norm": 0.20151183009147644, + "learning_rate": 8.963954323904511e-08, + "loss": 0.0213, + "step": 98190 + }, + { + "epoch": 19.064268724819183, + "grad_norm": 0.1221851035952568, + "learning_rate": 8.865262295891796e-08, + "loss": 0.0229, + "step": 98200 + }, + { + "epoch": 19.066210378136983, + "grad_norm": 0.18030628561973572, + "learning_rate": 8.767116090870331e-08, + "loss": 0.0217, + "step": 98210 + }, + { + "epoch": 19.068152031454783, + "grad_norm": 0.13914203643798828, + "learning_rate": 8.669515719572086e-08, + "loss": 0.0168, + "step": 98220 + }, + { + "epoch": 19.070093684772583, + "grad_norm": 0.10938137769699097, + "learning_rate": 8.572461192671855e-08, + "loss": 0.0193, + "step": 98230 + }, + { + "epoch": 19.072035338090384, + "grad_norm": 0.13846145570278168, + "learning_rate": 8.475952520782815e-08, + "loss": 0.0165, + "step": 98240 + }, + { + "epoch": 19.073976991408184, + "grad_norm": 0.12425961345434189, + "learning_rate": 8.379989714458747e-08, + "loss": 0.023, + "step": 98250 + }, + { + "epoch": 19.075918644725984, + "grad_norm": 0.24095943570137024, + "learning_rate": 8.284572784194034e-08, + "loss": 0.0212, + "step": 98260 + }, + { + "epoch": 19.077860298043785, + "grad_norm": 0.21135137975215912, + "learning_rate": 8.189701740424216e-08, + "loss": 0.0209, + "step": 98270 + }, + { + "epoch": 19.079801951361585, + "grad_norm": 0.1801832914352417, + "learning_rate": 8.095376593522663e-08, + "loss": 0.0223, + "step": 98280 + }, + { + "epoch": 19.081743604679385, + "grad_norm": 0.14185331761837006, + "learning_rate": 8.001597353806123e-08, + "loss": 0.0283, + "step": 98290 + }, + { + "epoch": 19.083685257997185, + "grad_norm": 0.21077118813991547, + "learning_rate": 7.90836403152917e-08, + "loss": 0.0194, + "step": 98300 + }, + { + "epoch": 19.085626911314986, + "grad_norm": 0.1846708357334137, + "learning_rate": 7.815676636888092e-08, + "loss": 0.0207, + "step": 98310 + }, + { + "epoch": 19.087568564632786, + "grad_norm": 0.10038524121046066, + "learning_rate": 7.723535180019226e-08, + "loss": 0.0258, + "step": 98320 + }, + { + "epoch": 19.089510217950586, + "grad_norm": 0.1876441091299057, + "learning_rate": 7.6319396709984e-08, + "loss": 0.0231, + "step": 98330 + }, + { + "epoch": 19.091451871268386, + "grad_norm": 0.2181277722120285, + "learning_rate": 7.540890119843158e-08, + "loss": 0.0227, + "step": 98340 + }, + { + "epoch": 19.093393524586187, + "grad_norm": 0.1669026017189026, + "learning_rate": 7.450386536509979e-08, + "loss": 0.0253, + "step": 98350 + }, + { + "epoch": 19.095335177903983, + "grad_norm": 0.18398486077785492, + "learning_rate": 7.360428930895947e-08, + "loss": 0.0169, + "step": 98360 + }, + { + "epoch": 19.097276831221784, + "grad_norm": 0.27744656801223755, + "learning_rate": 7.271017312839302e-08, + "loss": 0.0215, + "step": 98370 + }, + { + "epoch": 19.099218484539584, + "grad_norm": 0.2534930109977722, + "learning_rate": 7.182151692118333e-08, + "loss": 0.0205, + "step": 98380 + }, + { + "epoch": 19.101160137857384, + "grad_norm": 0.2250322848558426, + "learning_rate": 7.093832078449713e-08, + "loss": 0.024, + "step": 98390 + }, + { + "epoch": 19.103101791175185, + "grad_norm": 0.2569163143634796, + "learning_rate": 7.006058481493493e-08, + "loss": 0.0253, + "step": 98400 + }, + { + "epoch": 19.105043444492985, + "grad_norm": 0.2847388982772827, + "learning_rate": 6.918830910847552e-08, + "loss": 0.0244, + "step": 98410 + }, + { + "epoch": 19.106985097810785, + "grad_norm": 0.1691528856754303, + "learning_rate": 6.832149376052033e-08, + "loss": 0.0172, + "step": 98420 + }, + { + "epoch": 19.108926751128585, + "grad_norm": 0.1045399084687233, + "learning_rate": 6.746013886584912e-08, + "loss": 0.0179, + "step": 98430 + }, + { + "epoch": 19.110868404446386, + "grad_norm": 0.11626863479614258, + "learning_rate": 6.660424451866431e-08, + "loss": 0.0229, + "step": 98440 + }, + { + "epoch": 19.112810057764186, + "grad_norm": 0.18633203208446503, + "learning_rate": 6.575381081256327e-08, + "loss": 0.0252, + "step": 98450 + }, + { + "epoch": 19.114751711081986, + "grad_norm": 0.1457909792661667, + "learning_rate": 6.49088378405549e-08, + "loss": 0.0193, + "step": 98460 + }, + { + "epoch": 19.116693364399787, + "grad_norm": 0.07884363830089569, + "learning_rate": 6.406932569503755e-08, + "loss": 0.0158, + "step": 98470 + }, + { + "epoch": 19.118635017717587, + "grad_norm": 0.2400139719247818, + "learning_rate": 6.323527446782662e-08, + "loss": 0.0236, + "step": 98480 + }, + { + "epoch": 19.120576671035387, + "grad_norm": 0.21677462756633759, + "learning_rate": 6.240668425012142e-08, + "loss": 0.0208, + "step": 98490 + }, + { + "epoch": 19.122518324353187, + "grad_norm": 0.2357797473669052, + "learning_rate": 6.158355513254388e-08, + "loss": 0.0235, + "step": 98500 + }, + { + "epoch": 19.124459977670988, + "grad_norm": 0.23508580029010773, + "learning_rate": 6.076588720510534e-08, + "loss": 0.0236, + "step": 98510 + }, + { + "epoch": 19.126401630988788, + "grad_norm": 0.1331864297389984, + "learning_rate": 5.995368055722316e-08, + "loss": 0.0184, + "step": 98520 + }, + { + "epoch": 19.128343284306588, + "grad_norm": 0.24791723489761353, + "learning_rate": 5.914693527773185e-08, + "loss": 0.021, + "step": 98530 + }, + { + "epoch": 19.13028493762439, + "grad_norm": 0.3368222117424011, + "learning_rate": 5.834565145483861e-08, + "loss": 0.0183, + "step": 98540 + }, + { + "epoch": 19.13222659094219, + "grad_norm": 0.15639786422252655, + "learning_rate": 5.754982917618446e-08, + "loss": 0.0184, + "step": 98550 + }, + { + "epoch": 19.13416824425999, + "grad_norm": 0.1893913596868515, + "learning_rate": 5.675946852878866e-08, + "loss": 0.0205, + "step": 98560 + }, + { + "epoch": 19.13610989757779, + "grad_norm": 0.22847315669059753, + "learning_rate": 5.59745695990932e-08, + "loss": 0.0225, + "step": 98570 + }, + { + "epoch": 19.138051550895586, + "grad_norm": 0.1466110348701477, + "learning_rate": 5.51951324729294e-08, + "loss": 0.0184, + "step": 98580 + }, + { + "epoch": 19.139993204213386, + "grad_norm": 0.11034779995679855, + "learning_rate": 5.4421157235529093e-08, + "loss": 0.0182, + "step": 98590 + }, + { + "epoch": 19.141934857531187, + "grad_norm": 0.10632094740867615, + "learning_rate": 5.365264397154679e-08, + "loss": 0.0162, + "step": 98600 + }, + { + "epoch": 19.143876510848987, + "grad_norm": 0.18102224171161652, + "learning_rate": 5.288959276501526e-08, + "loss": 0.0217, + "step": 98610 + }, + { + "epoch": 19.145818164166787, + "grad_norm": 0.24344459176063538, + "learning_rate": 5.2132003699378876e-08, + "loss": 0.0174, + "step": 98620 + }, + { + "epoch": 19.147759817484587, + "grad_norm": 0.11962645500898361, + "learning_rate": 5.137987685749912e-08, + "loss": 0.0217, + "step": 98630 + }, + { + "epoch": 19.149701470802388, + "grad_norm": 0.07996749132871628, + "learning_rate": 5.0633212321610227e-08, + "loss": 0.0203, + "step": 98640 + }, + { + "epoch": 19.151643124120188, + "grad_norm": 0.23670636117458344, + "learning_rate": 4.989201017338019e-08, + "loss": 0.0267, + "step": 98650 + }, + { + "epoch": 19.153584777437988, + "grad_norm": 0.21060343086719513, + "learning_rate": 4.9156270493860846e-08, + "loss": 0.021, + "step": 98660 + }, + { + "epoch": 19.15552643075579, + "grad_norm": 0.2734650671482086, + "learning_rate": 4.842599336351561e-08, + "loss": 0.0235, + "step": 98670 + }, + { + "epoch": 19.15746808407359, + "grad_norm": 0.16114795207977295, + "learning_rate": 4.7701178862197274e-08, + "loss": 0.0186, + "step": 98680 + }, + { + "epoch": 19.15940973739139, + "grad_norm": 0.15867449343204498, + "learning_rate": 4.698182706918131e-08, + "loss": 0.017, + "step": 98690 + }, + { + "epoch": 19.16135139070919, + "grad_norm": 0.17162153124809265, + "learning_rate": 4.6267938063121466e-08, + "loss": 0.0208, + "step": 98700 + }, + { + "epoch": 19.16329304402699, + "grad_norm": 0.13378623127937317, + "learning_rate": 4.555951192209973e-08, + "loss": 0.0237, + "step": 98710 + }, + { + "epoch": 19.16523469734479, + "grad_norm": 0.13628850877285004, + "learning_rate": 4.485654872358747e-08, + "loss": 0.0183, + "step": 98720 + }, + { + "epoch": 19.16717635066259, + "grad_norm": 0.18932810425758362, + "learning_rate": 4.415904854445097e-08, + "loss": 0.0184, + "step": 98730 + }, + { + "epoch": 19.16911800398039, + "grad_norm": 0.19857127964496613, + "learning_rate": 4.34670114609792e-08, + "loss": 0.0195, + "step": 98740 + }, + { + "epoch": 19.17105965729819, + "grad_norm": 0.2975034713745117, + "learning_rate": 4.278043754884498e-08, + "loss": 0.0168, + "step": 98750 + }, + { + "epoch": 19.17300131061599, + "grad_norm": 0.179962620139122, + "learning_rate": 4.209932688312712e-08, + "loss": 0.0214, + "step": 98760 + }, + { + "epoch": 19.17494296393379, + "grad_norm": 0.14529503881931305, + "learning_rate": 4.14236795383216e-08, + "loss": 0.0213, + "step": 98770 + }, + { + "epoch": 19.17688461725159, + "grad_norm": 0.28414446115493774, + "learning_rate": 4.075349558830821e-08, + "loss": 0.0238, + "step": 98780 + }, + { + "epoch": 19.178826270569388, + "grad_norm": 0.20883435010910034, + "learning_rate": 4.008877510638387e-08, + "loss": 0.0226, + "step": 98790 + }, + { + "epoch": 19.18076792388719, + "grad_norm": 0.1984362006187439, + "learning_rate": 3.942951816523488e-08, + "loss": 0.0235, + "step": 98800 + }, + { + "epoch": 19.18270957720499, + "grad_norm": 0.1329496055841446, + "learning_rate": 3.8775724836959125e-08, + "loss": 0.0173, + "step": 98810 + }, + { + "epoch": 19.18465123052279, + "grad_norm": 0.31275731325149536, + "learning_rate": 3.812739519305497e-08, + "loss": 0.0252, + "step": 98820 + }, + { + "epoch": 19.18659288384059, + "grad_norm": 0.21587859094142914, + "learning_rate": 3.748452930442126e-08, + "loss": 0.021, + "step": 98830 + }, + { + "epoch": 19.18853453715839, + "grad_norm": 0.19711586833000183, + "learning_rate": 3.6847127241362855e-08, + "loss": 0.0283, + "step": 98840 + }, + { + "epoch": 19.19047619047619, + "grad_norm": 0.12216270714998245, + "learning_rate": 3.621518907358512e-08, + "loss": 0.0179, + "step": 98850 + }, + { + "epoch": 19.19241784379399, + "grad_norm": 0.28551021218299866, + "learning_rate": 3.558871487019388e-08, + "loss": 0.0222, + "step": 98860 + }, + { + "epoch": 19.19435949711179, + "grad_norm": 0.12353220582008362, + "learning_rate": 3.4967704699701007e-08, + "loss": 0.0191, + "step": 98870 + }, + { + "epoch": 19.19630115042959, + "grad_norm": 0.12757660448551178, + "learning_rate": 3.435215863001884e-08, + "loss": 0.0195, + "step": 98880 + }, + { + "epoch": 19.19824280374739, + "grad_norm": 0.1219472885131836, + "learning_rate": 3.374207672846019e-08, + "loss": 0.0171, + "step": 98890 + }, + { + "epoch": 19.20018445706519, + "grad_norm": 0.13900135457515717, + "learning_rate": 3.313745906174392e-08, + "loss": 0.0178, + "step": 98900 + }, + { + "epoch": 19.20212611038299, + "grad_norm": 0.15024317800998688, + "learning_rate": 3.253830569599492e-08, + "loss": 0.0154, + "step": 98910 + }, + { + "epoch": 19.20406776370079, + "grad_norm": 0.21957440674304962, + "learning_rate": 3.1944616696727436e-08, + "loss": 0.0186, + "step": 98920 + }, + { + "epoch": 19.206009417018592, + "grad_norm": 0.20016489923000336, + "learning_rate": 3.135639212887287e-08, + "loss": 0.0215, + "step": 98930 + }, + { + "epoch": 19.207951070336392, + "grad_norm": 0.09911968559026718, + "learning_rate": 3.077363205675754e-08, + "loss": 0.0223, + "step": 98940 + }, + { + "epoch": 19.209892723654193, + "grad_norm": 0.10372790694236755, + "learning_rate": 3.0196336544113804e-08, + "loss": 0.0189, + "step": 98950 + }, + { + "epoch": 19.211834376971993, + "grad_norm": 0.24164018034934998, + "learning_rate": 2.9624505654063383e-08, + "loss": 0.0229, + "step": 98960 + }, + { + "epoch": 19.213776030289793, + "grad_norm": 0.14433792233467102, + "learning_rate": 2.90581394491507e-08, + "loss": 0.0254, + "step": 98970 + }, + { + "epoch": 19.215717683607593, + "grad_norm": 0.23455633223056793, + "learning_rate": 2.8497237991309545e-08, + "loss": 0.0186, + "step": 98980 + }, + { + "epoch": 19.217659336925394, + "grad_norm": 0.17451612651348114, + "learning_rate": 2.7941801341879735e-08, + "loss": 0.0201, + "step": 98990 + }, + { + "epoch": 19.21960099024319, + "grad_norm": 0.34776753187179565, + "learning_rate": 2.7391829561601578e-08, + "loss": 0.0194, + "step": 99000 + }, + { + "epoch": 19.22154264356099, + "grad_norm": 0.2327645719051361, + "learning_rate": 2.6847322710621403e-08, + "loss": 0.0229, + "step": 99010 + }, + { + "epoch": 19.22348429687879, + "grad_norm": 0.23770104348659515, + "learning_rate": 2.6308280848486022e-08, + "loss": 0.0241, + "step": 99020 + }, + { + "epoch": 19.22542595019659, + "grad_norm": 0.0985194519162178, + "learning_rate": 2.5774704034137177e-08, + "loss": 0.0172, + "step": 99030 + }, + { + "epoch": 19.22736760351439, + "grad_norm": 0.16533461213111877, + "learning_rate": 2.524659232593929e-08, + "loss": 0.0158, + "step": 99040 + }, + { + "epoch": 19.22930925683219, + "grad_norm": 0.19515420496463776, + "learning_rate": 2.472394578163506e-08, + "loss": 0.0254, + "step": 99050 + }, + { + "epoch": 19.231250910149992, + "grad_norm": 0.14853009581565857, + "learning_rate": 2.4206764458378772e-08, + "loss": 0.0216, + "step": 99060 + }, + { + "epoch": 19.233192563467792, + "grad_norm": 0.1993803232908249, + "learning_rate": 2.3695048412736285e-08, + "loss": 0.0262, + "step": 99070 + }, + { + "epoch": 19.235134216785593, + "grad_norm": 0.13895724713802338, + "learning_rate": 2.3188797700668395e-08, + "loss": 0.0204, + "step": 99080 + }, + { + "epoch": 19.237075870103393, + "grad_norm": 0.25893205404281616, + "learning_rate": 2.268801237753082e-08, + "loss": 0.0293, + "step": 99090 + }, + { + "epoch": 19.239017523421193, + "grad_norm": 0.22818098962306976, + "learning_rate": 2.2192692498090862e-08, + "loss": 0.0256, + "step": 99100 + }, + { + "epoch": 19.240959176738993, + "grad_norm": 0.17197555303573608, + "learning_rate": 2.170283811652185e-08, + "loss": 0.0203, + "step": 99110 + }, + { + "epoch": 19.242900830056794, + "grad_norm": 0.14409275352954865, + "learning_rate": 2.12184492863865e-08, + "loss": 0.0192, + "step": 99120 + }, + { + "epoch": 19.244842483374594, + "grad_norm": 0.12029476463794708, + "learning_rate": 2.073952606066465e-08, + "loss": 0.0242, + "step": 99130 + }, + { + "epoch": 19.246784136692394, + "grad_norm": 0.19965624809265137, + "learning_rate": 2.026606849171997e-08, + "loss": 0.0162, + "step": 99140 + }, + { + "epoch": 19.248725790010194, + "grad_norm": 0.1842818409204483, + "learning_rate": 1.9798076631333262e-08, + "loss": 0.0214, + "step": 99150 + }, + { + "epoch": 19.250667443327995, + "grad_norm": 0.08067276328802109, + "learning_rate": 1.933555053069136e-08, + "loss": 0.0214, + "step": 99160 + }, + { + "epoch": 19.252609096645795, + "grad_norm": 0.24595627188682556, + "learning_rate": 1.887849024036492e-08, + "loss": 0.0253, + "step": 99170 + }, + { + "epoch": 19.254550749963595, + "grad_norm": 0.17149284482002258, + "learning_rate": 1.8426895810341736e-08, + "loss": 0.0192, + "step": 99180 + }, + { + "epoch": 19.256492403281396, + "grad_norm": 0.1946643590927124, + "learning_rate": 1.798076729000453e-08, + "loss": 0.0253, + "step": 99190 + }, + { + "epoch": 19.258434056599196, + "grad_norm": 0.13637828826904297, + "learning_rate": 1.75401047281476e-08, + "loss": 0.0234, + "step": 99200 + }, + { + "epoch": 19.260375709916993, + "grad_norm": 0.12837538123130798, + "learning_rate": 1.7104908172954626e-08, + "loss": 0.0226, + "step": 99210 + }, + { + "epoch": 19.262317363234793, + "grad_norm": 0.19574539363384247, + "learning_rate": 1.6675177672020868e-08, + "loss": 0.0269, + "step": 99220 + }, + { + "epoch": 19.264259016552593, + "grad_norm": 0.11744990944862366, + "learning_rate": 1.6250913272342073e-08, + "loss": 0.0198, + "step": 99230 + }, + { + "epoch": 19.266200669870393, + "grad_norm": 0.3034004271030426, + "learning_rate": 1.5832115020314453e-08, + "loss": 0.0217, + "step": 99240 + }, + { + "epoch": 19.268142323188194, + "grad_norm": 0.30978158116340637, + "learning_rate": 1.5418782961734712e-08, + "loss": 0.0257, + "step": 99250 + }, + { + "epoch": 19.270083976505994, + "grad_norm": 0.2948550879955292, + "learning_rate": 1.501091714181113e-08, + "loss": 0.0257, + "step": 99260 + }, + { + "epoch": 19.272025629823794, + "grad_norm": 0.09797440469264984, + "learning_rate": 1.460851760513582e-08, + "loss": 0.019, + "step": 99270 + }, + { + "epoch": 19.273967283141594, + "grad_norm": 0.13263900578022003, + "learning_rate": 1.4211584395723566e-08, + "loss": 0.019, + "step": 99280 + }, + { + "epoch": 19.275908936459395, + "grad_norm": 0.15888042747974396, + "learning_rate": 1.382011755697854e-08, + "loss": 0.0223, + "step": 99290 + }, + { + "epoch": 19.277850589777195, + "grad_norm": 0.22469109296798706, + "learning_rate": 1.3434117131716494e-08, + "loss": 0.0252, + "step": 99300 + }, + { + "epoch": 19.279792243094995, + "grad_norm": 0.24479912221431732, + "learning_rate": 1.3053583162142557e-08, + "loss": 0.0221, + "step": 99310 + }, + { + "epoch": 19.281733896412796, + "grad_norm": 0.1544281244277954, + "learning_rate": 1.2678515689873438e-08, + "loss": 0.028, + "step": 99320 + }, + { + "epoch": 19.283675549730596, + "grad_norm": 0.19081014394760132, + "learning_rate": 1.2308914755931877e-08, + "loss": 0.0234, + "step": 99330 + }, + { + "epoch": 19.285617203048396, + "grad_norm": 0.21505136787891388, + "learning_rate": 1.1944780400729993e-08, + "loss": 0.0217, + "step": 99340 + }, + { + "epoch": 19.287558856366196, + "grad_norm": 0.2043776512145996, + "learning_rate": 1.1586112664085935e-08, + "loss": 0.018, + "step": 99350 + }, + { + "epoch": 19.289500509683997, + "grad_norm": 0.1443033367395401, + "learning_rate": 1.1232911585234984e-08, + "loss": 0.0189, + "step": 99360 + }, + { + "epoch": 19.291442163001797, + "grad_norm": 0.1314253807067871, + "learning_rate": 1.0885177202790697e-08, + "loss": 0.0205, + "step": 99370 + }, + { + "epoch": 19.293383816319597, + "grad_norm": 0.3084002137184143, + "learning_rate": 1.0542909554789315e-08, + "loss": 0.0243, + "step": 99380 + }, + { + "epoch": 19.295325469637397, + "grad_norm": 0.089642733335495, + "learning_rate": 1.0206108678656456e-08, + "loss": 0.0205, + "step": 99390 + }, + { + "epoch": 19.297267122955198, + "grad_norm": 0.12139718979597092, + "learning_rate": 9.874774611223769e-09, + "loss": 0.0191, + "step": 99400 + }, + { + "epoch": 19.299208776272998, + "grad_norm": 0.2419566810131073, + "learning_rate": 9.548907388728933e-09, + "loss": 0.022, + "step": 99410 + }, + { + "epoch": 19.3011504295908, + "grad_norm": 0.24190694093704224, + "learning_rate": 9.228507046804557e-09, + "loss": 0.0233, + "step": 99420 + }, + { + "epoch": 19.303092082908595, + "grad_norm": 0.20501258969306946, + "learning_rate": 8.91357362048928e-09, + "loss": 0.0248, + "step": 99430 + }, + { + "epoch": 19.305033736226395, + "grad_norm": 0.2689861059188843, + "learning_rate": 8.604107144227769e-09, + "loss": 0.0288, + "step": 99440 + }, + { + "epoch": 19.306975389544196, + "grad_norm": 0.18495047092437744, + "learning_rate": 8.300107651859623e-09, + "loss": 0.0249, + "step": 99450 + }, + { + "epoch": 19.308917042861996, + "grad_norm": 0.19953888654708862, + "learning_rate": 8.001575176630472e-09, + "loss": 0.0225, + "step": 99460 + }, + { + "epoch": 19.310858696179796, + "grad_norm": 0.18429876863956451, + "learning_rate": 7.708509751186422e-09, + "loss": 0.0176, + "step": 99470 + }, + { + "epoch": 19.312800349497596, + "grad_norm": 0.22542478144168854, + "learning_rate": 7.420911407579611e-09, + "loss": 0.0187, + "step": 99480 + }, + { + "epoch": 19.314742002815397, + "grad_norm": 0.17730459570884705, + "learning_rate": 7.13878017725711e-09, + "loss": 0.0204, + "step": 99490 + }, + { + "epoch": 19.316683656133197, + "grad_norm": 0.13185864686965942, + "learning_rate": 6.8621160910720125e-09, + "loss": 0.0242, + "step": 99500 + }, + { + "epoch": 19.318625309450997, + "grad_norm": 0.2007168084383011, + "learning_rate": 6.59091917928345e-09, + "loss": 0.0238, + "step": 99510 + }, + { + "epoch": 19.320566962768797, + "grad_norm": 0.21303687989711761, + "learning_rate": 6.32518947155103e-09, + "loss": 0.0218, + "step": 99520 + }, + { + "epoch": 19.322508616086598, + "grad_norm": 0.1202530488371849, + "learning_rate": 6.064926996929288e-09, + "loss": 0.0248, + "step": 99530 + }, + { + "epoch": 19.324450269404398, + "grad_norm": 0.1432410627603531, + "learning_rate": 5.810131783884343e-09, + "loss": 0.0234, + "step": 99540 + }, + { + "epoch": 19.3263919227222, + "grad_norm": 0.34450337290763855, + "learning_rate": 5.5608038602772415e-09, + "loss": 0.0281, + "step": 99550 + }, + { + "epoch": 19.32833357604, + "grad_norm": 0.1902855485677719, + "learning_rate": 5.316943253375062e-09, + "loss": 0.0213, + "step": 99560 + }, + { + "epoch": 19.3302752293578, + "grad_norm": 0.20433653891086578, + "learning_rate": 5.07854998984536e-09, + "loss": 0.0195, + "step": 99570 + }, + { + "epoch": 19.3322168826756, + "grad_norm": 0.1373092532157898, + "learning_rate": 4.845624095756174e-09, + "loss": 0.0209, + "step": 99580 + }, + { + "epoch": 19.3341585359934, + "grad_norm": 0.16015450656414032, + "learning_rate": 4.618165596587121e-09, + "loss": 0.0233, + "step": 99590 + }, + { + "epoch": 19.3361001893112, + "grad_norm": 0.10134819895029068, + "learning_rate": 4.396174517207197e-09, + "loss": 0.0163, + "step": 99600 + }, + { + "epoch": 19.338041842629, + "grad_norm": 0.14066849648952484, + "learning_rate": 4.179650881896979e-09, + "loss": 0.0216, + "step": 99610 + }, + { + "epoch": 19.3399834959468, + "grad_norm": 0.21496634185314178, + "learning_rate": 3.968594714331975e-09, + "loss": 0.0215, + "step": 99620 + }, + { + "epoch": 19.3419251492646, + "grad_norm": 0.1971287578344345, + "learning_rate": 3.763006037593719e-09, + "loss": 0.0228, + "step": 99630 + }, + { + "epoch": 19.343866802582397, + "grad_norm": 0.27647072076797485, + "learning_rate": 3.5628848741586786e-09, + "loss": 0.0266, + "step": 99640 + }, + { + "epoch": 19.345808455900197, + "grad_norm": 0.16461949050426483, + "learning_rate": 3.368231245926001e-09, + "loss": 0.0244, + "step": 99650 + }, + { + "epoch": 19.347750109217998, + "grad_norm": 0.21461988985538483, + "learning_rate": 3.179045174167561e-09, + "loss": 0.0215, + "step": 99660 + }, + { + "epoch": 19.349691762535798, + "grad_norm": 0.18635860085487366, + "learning_rate": 2.9953266795834657e-09, + "loss": 0.0245, + "step": 99670 + }, + { + "epoch": 19.3516334158536, + "grad_norm": 0.14093004167079926, + "learning_rate": 2.817075782263201e-09, + "loss": 0.0213, + "step": 99680 + }, + { + "epoch": 19.3535750691714, + "grad_norm": 0.2728019058704376, + "learning_rate": 2.6442925016967325e-09, + "loss": 0.0231, + "step": 99690 + }, + { + "epoch": 19.3555167224892, + "grad_norm": 0.19401812553405762, + "learning_rate": 2.4769768567800555e-09, + "loss": 0.0216, + "step": 99700 + }, + { + "epoch": 19.357458375807, + "grad_norm": 0.13364183902740479, + "learning_rate": 2.315128865809646e-09, + "loss": 0.0202, + "step": 99710 + }, + { + "epoch": 19.3594000291248, + "grad_norm": 0.549473762512207, + "learning_rate": 2.1587485464880097e-09, + "loss": 0.0252, + "step": 99720 + }, + { + "epoch": 19.3613416824426, + "grad_norm": 0.21070343255996704, + "learning_rate": 2.0078359159125813e-09, + "loss": 0.021, + "step": 99730 + }, + { + "epoch": 19.3632833357604, + "grad_norm": 0.1319783478975296, + "learning_rate": 1.8623909905923776e-09, + "loss": 0.0201, + "step": 99740 + }, + { + "epoch": 19.3652249890782, + "grad_norm": 0.15501965582370758, + "learning_rate": 1.7224137864257916e-09, + "loss": 0.0214, + "step": 99750 + }, + { + "epoch": 19.367166642396, + "grad_norm": 0.1984877735376358, + "learning_rate": 1.5879043187283505e-09, + "loss": 0.0222, + "step": 99760 + }, + { + "epoch": 19.3691082957138, + "grad_norm": 0.3083249032497406, + "learning_rate": 1.458862602204958e-09, + "loss": 0.0215, + "step": 99770 + }, + { + "epoch": 19.3710499490316, + "grad_norm": 0.18874698877334595, + "learning_rate": 1.3352886509720996e-09, + "loss": 0.0263, + "step": 99780 + }, + { + "epoch": 19.3729916023494, + "grad_norm": 0.2423279583454132, + "learning_rate": 1.2171824785356389e-09, + "loss": 0.0243, + "step": 99790 + }, + { + "epoch": 19.3749332556672, + "grad_norm": 0.261412113904953, + "learning_rate": 1.1045440978185718e-09, + "loss": 0.0164, + "step": 99800 + }, + { + "epoch": 19.376874908985002, + "grad_norm": 0.20082053542137146, + "learning_rate": 9.973735211332714e-10, + "loss": 0.0172, + "step": 99810 + }, + { + "epoch": 19.378816562302802, + "grad_norm": 0.16320447623729706, + "learning_rate": 8.956707602036929e-10, + "loss": 0.0215, + "step": 99820 + }, + { + "epoch": 19.380758215620602, + "grad_norm": 0.08018267899751663, + "learning_rate": 7.994358261542712e-10, + "loss": 0.022, + "step": 99830 + }, + { + "epoch": 19.382699868938403, + "grad_norm": 0.17163027822971344, + "learning_rate": 7.086687295043692e-10, + "loss": 0.0246, + "step": 99840 + }, + { + "epoch": 19.3846415222562, + "grad_norm": 0.22621220350265503, + "learning_rate": 6.23369480179381e-10, + "loss": 0.0288, + "step": 99850 + }, + { + "epoch": 19.386583175574, + "grad_norm": 0.1994219720363617, + "learning_rate": 5.43538087510731e-10, + "loss": 0.0218, + "step": 99860 + }, + { + "epoch": 19.3885248288918, + "grad_norm": 0.200997456908226, + "learning_rate": 4.691745602303233e-10, + "loss": 0.0214, + "step": 99870 + }, + { + "epoch": 19.3904664822096, + "grad_norm": 0.189958393573761, + "learning_rate": 4.0027890646499034e-10, + "loss": 0.0214, + "step": 99880 + }, + { + "epoch": 19.3924081355274, + "grad_norm": 0.1372307687997818, + "learning_rate": 3.368511337531466e-10, + "loss": 0.021, + "step": 99890 + }, + { + "epoch": 19.3943497888452, + "grad_norm": 0.16894178092479706, + "learning_rate": 2.788912490281348e-10, + "loss": 0.0214, + "step": 99900 + }, + { + "epoch": 19.396291442163, + "grad_norm": 0.1389956921339035, + "learning_rate": 2.2639925863487954e-10, + "loss": 0.0221, + "step": 99910 + }, + { + "epoch": 19.3982330954808, + "grad_norm": 0.2253703773021698, + "learning_rate": 1.7937516830213163e-10, + "loss": 0.0245, + "step": 99920 + }, + { + "epoch": 19.4001747487986, + "grad_norm": 0.1511986255645752, + "learning_rate": 1.3781898318687702e-10, + "loss": 0.0261, + "step": 99930 + }, + { + "epoch": 19.402116402116402, + "grad_norm": 0.15379486978054047, + "learning_rate": 1.0173070781882566e-10, + "loss": 0.0219, + "step": 99940 + }, + { + "epoch": 19.404058055434202, + "grad_norm": 0.2590492069721222, + "learning_rate": 7.111034615592261e-11, + "loss": 0.0208, + "step": 99950 + }, + { + "epoch": 19.405999708752002, + "grad_norm": 0.23277787864208221, + "learning_rate": 4.595790153993918e-11, + "loss": 0.0228, + "step": 99960 + }, + { + "epoch": 19.407941362069803, + "grad_norm": 0.08511075377464294, + "learning_rate": 2.6273376729779585e-11, + "loss": 0.0228, + "step": 99970 + }, + { + "epoch": 19.409883015387603, + "grad_norm": 0.1303296685218811, + "learning_rate": 1.2056773868174276e-11, + "loss": 0.0227, + "step": 99980 + }, + { + "epoch": 19.411824668705403, + "grad_norm": 0.14683400094509125, + "learning_rate": 3.3080945149865926e-12, + "loss": 0.0184, + "step": 99990 + }, + { + "epoch": 19.413766322023204, + "grad_norm": 0.1899387389421463, + "learning_rate": 2.7339625008337977e-14, + "loss": 0.0266, + "step": 100000 + }, + { + "epoch": 19.413766322023204, + "step": 100000, + "total_flos": 0.0, + "train_loss": 0.03547922075420618, + "train_runtime": 81469.4862, + "train_samples_per_second": 19.639, + "train_steps_per_second": 1.227 + } + ], + "logging_steps": 10, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "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": 4, + "trial_name": null, + "trial_params": null +}