diff --git "a/checkpoint-20000/trainer_state.json" "b/checkpoint-20000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-20000/trainer_state.json" @@ -0,0 +1,14034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.8828697636037086, + "eval_steps": 500, + "global_step": 20000, + "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 + } + ], + "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": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}