{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.4, "eval_steps": 500, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.497777223587036, "learning_rate": 3.6e-06, "loss": 1.1476, "step": 10 }, { "grad_norm": 0.5899417996406555, "learning_rate": 7.6e-06, "loss": 1.1021, "step": 20 }, { "grad_norm": 0.6255532503128052, "learning_rate": 1.16e-05, "loss": 1.0779, "step": 30 }, { "grad_norm": 0.5589399933815002, "learning_rate": 1.56e-05, "loss": 1.0599, "step": 40 }, { "grad_norm": 0.6063516736030579, "learning_rate": 1.9600000000000002e-05, "loss": 1.0856, "step": 50 }, { "grad_norm": 0.5875065326690674, "learning_rate": 2.36e-05, "loss": 1.0636, "step": 60 }, { "grad_norm": 0.6206650733947754, "learning_rate": 2.7600000000000003e-05, "loss": 1.0586, "step": 70 }, { "grad_norm": 0.6976884007453918, "learning_rate": 3.16e-05, "loss": 1.0581, "step": 80 }, { "grad_norm": 0.4975197911262512, "learning_rate": 3.56e-05, "loss": 1.0338, "step": 90 }, { "grad_norm": 0.5175010561943054, "learning_rate": 3.960000000000001e-05, "loss": 1.0363, "step": 100 }, { "grad_norm": 3.203188180923462, "learning_rate": 4.36e-05, "loss": 0.9943, "step": 110 }, { "grad_norm": 1.0046463012695312, "learning_rate": 4.76e-05, "loss": 0.9498, "step": 120 }, { "grad_norm": 0.824455976486206, "learning_rate": 5.16e-05, "loss": 0.9141, "step": 130 }, { "grad_norm": 0.933220386505127, "learning_rate": 5.560000000000001e-05, "loss": 0.8966, "step": 140 }, { "grad_norm": 1.6803524494171143, "learning_rate": 5.96e-05, "loss": 0.8456, "step": 150 }, { "grad_norm": 1.3335356712341309, "learning_rate": 6.36e-05, "loss": 0.7857, "step": 160 }, { "grad_norm": 1.7373239994049072, "learning_rate": 6.76e-05, "loss": 0.6714, "step": 170 }, { "grad_norm": 1.2455403804779053, "learning_rate": 7.16e-05, "loss": 0.5578, "step": 180 }, { "grad_norm": 2.673060178756714, "learning_rate": 7.560000000000001e-05, "loss": 0.4206, "step": 190 }, { "grad_norm": 1.7698984146118164, "learning_rate": 7.960000000000001e-05, "loss": 0.3168, "step": 200 }, { "grad_norm": 1.393641471862793, "learning_rate": 8.36e-05, "loss": 0.2192, "step": 210 }, { "grad_norm": 1.3259373903274536, "learning_rate": 8.76e-05, "loss": 0.1876, "step": 220 }, { "grad_norm": 0.9707556366920471, "learning_rate": 9.16e-05, "loss": 0.1548, "step": 230 }, { "grad_norm": 0.8586066961288452, "learning_rate": 9.56e-05, "loss": 0.1401, "step": 240 }, { "grad_norm": 0.8358578085899353, "learning_rate": 9.960000000000001e-05, "loss": 0.1216, "step": 250 }, { "grad_norm": 0.6485460996627808, "learning_rate": 9.999911419878559e-05, "loss": 0.1228, "step": 260 }, { "grad_norm": 1.0369858741760254, "learning_rate": 9.999605221019081e-05, "loss": 0.1396, "step": 270 }, { "grad_norm": 0.8534221053123474, "learning_rate": 9.999080323230761e-05, "loss": 0.1208, "step": 280 }, { "grad_norm": 0.8548014760017395, "learning_rate": 9.998336749474329e-05, "loss": 0.1341, "step": 290 }, { "grad_norm": 0.9402936100959778, "learning_rate": 9.997374532276107e-05, "loss": 0.102, "step": 300 }, { "grad_norm": 0.7149423956871033, "learning_rate": 9.996193713726596e-05, "loss": 0.1092, "step": 310 }, { "grad_norm": 0.7244396805763245, "learning_rate": 9.994794345478624e-05, "loss": 0.0924, "step": 320 }, { "grad_norm": 0.5502927899360657, "learning_rate": 9.99317648874509e-05, "loss": 0.0838, "step": 330 }, { "grad_norm": 0.6544800400733948, "learning_rate": 9.991340214296292e-05, "loss": 0.0873, "step": 340 }, { "grad_norm": 0.7180399894714355, "learning_rate": 9.989285602456819e-05, "loss": 0.0734, "step": 350 }, { "grad_norm": 0.6445586681365967, "learning_rate": 9.98701274310205e-05, "loss": 0.0794, "step": 360 }, { "grad_norm": 0.5352494716644287, "learning_rate": 9.984521735654218e-05, "loss": 0.0907, "step": 370 }, { "grad_norm": 0.6672635674476624, "learning_rate": 9.981812689078057e-05, "loss": 0.0848, "step": 380 }, { "grad_norm": 0.5291720628738403, "learning_rate": 9.978885721876041e-05, "loss": 0.086, "step": 390 }, { "grad_norm": 0.39044567942619324, "learning_rate": 9.975740962083198e-05, "loss": 0.0787, "step": 400 }, { "grad_norm": 0.7575682997703552, "learning_rate": 9.972378547261504e-05, "loss": 0.0836, "step": 410 }, { "grad_norm": 0.6050339341163635, "learning_rate": 9.968798624493885e-05, "loss": 0.0738, "step": 420 }, { "grad_norm": 0.594001054763794, "learning_rate": 9.965001350377753e-05, "loss": 0.0817, "step": 430 }, { "grad_norm": 0.5727436542510986, "learning_rate": 9.960986891018183e-05, "loss": 0.0728, "step": 440 }, { "grad_norm": 0.4483703076839447, "learning_rate": 9.95675542202063e-05, "loss": 0.0747, "step": 450 }, { "grad_norm": 0.3992336094379425, "learning_rate": 9.952307128483256e-05, "loss": 0.0796, "step": 460 }, { "grad_norm": 0.5178577303886414, "learning_rate": 9.947642204988835e-05, "loss": 0.0833, "step": 470 }, { "grad_norm": 0.4497125446796417, "learning_rate": 9.942760855596226e-05, "loss": 0.0787, "step": 480 }, { "grad_norm": 0.5343725085258484, "learning_rate": 9.937663293831471e-05, "loss": 0.0808, "step": 490 }, { "grad_norm": 0.47977861762046814, "learning_rate": 9.932349742678433e-05, "loss": 0.0701, "step": 500 }, { "grad_norm": 0.5831079483032227, "learning_rate": 9.926820434569051e-05, "loss": 0.0742, "step": 510 }, { "grad_norm": 0.47160646319389343, "learning_rate": 9.921075611373179e-05, "loss": 0.067, "step": 520 }, { "grad_norm": 0.35774683952331543, "learning_rate": 9.915115524387988e-05, "loss": 0.0709, "step": 530 }, { "grad_norm": 0.5747243762016296, "learning_rate": 9.908940434326997e-05, "loss": 0.0668, "step": 540 }, { "grad_norm": 0.42538684606552124, "learning_rate": 9.902550611308645e-05, "loss": 0.0676, "step": 550 }, { "grad_norm": 0.5689398646354675, "learning_rate": 9.895946334844494e-05, "loss": 0.0631, "step": 560 }, { "grad_norm": 0.648316502571106, "learning_rate": 9.889127893826989e-05, "loss": 0.0643, "step": 570 }, { "grad_norm": 0.5649459362030029, "learning_rate": 9.882095586516831e-05, "loss": 0.0679, "step": 580 }, { "grad_norm": 0.42304474115371704, "learning_rate": 9.874849720529921e-05, "loss": 0.0744, "step": 590 }, { "grad_norm": 0.5694190263748169, "learning_rate": 9.867390612823914e-05, "loss": 0.0683, "step": 600 }, { "grad_norm": 0.5124210715293884, "learning_rate": 9.859718589684344e-05, "loss": 0.0797, "step": 610 }, { "grad_norm": 0.6514238119125366, "learning_rate": 9.851833986710353e-05, "loss": 0.0644, "step": 620 }, { "grad_norm": 0.49233269691467285, "learning_rate": 9.843737148800023e-05, "loss": 0.0568, "step": 630 }, { "grad_norm": 0.40201443433761597, "learning_rate": 9.835428430135271e-05, "loss": 0.063, "step": 640 }, { "grad_norm": 0.4645858407020569, "learning_rate": 9.82690819416637e-05, "loss": 0.0641, "step": 650 }, { "grad_norm": 0.5921373963356018, "learning_rate": 9.818176813596041e-05, "loss": 0.0622, "step": 660 }, { "grad_norm": 0.5843824148178101, "learning_rate": 9.809234670363159e-05, "loss": 0.0724, "step": 670 }, { "grad_norm": 0.6411054730415344, "learning_rate": 9.800082155626034e-05, "loss": 0.0697, "step": 680 }, { "grad_norm": 0.4261399507522583, "learning_rate": 9.790719669745312e-05, "loss": 0.0641, "step": 690 }, { "grad_norm": 0.3344235122203827, "learning_rate": 9.781147622266455e-05, "loss": 0.0642, "step": 700 }, { "grad_norm": 0.4839865565299988, "learning_rate": 9.771366431901831e-05, "loss": 0.0734, "step": 710 }, { "grad_norm": 0.3950216472148895, "learning_rate": 9.761376526512394e-05, "loss": 0.063, "step": 720 }, { "grad_norm": 0.45863887667655945, "learning_rate": 9.751178343088963e-05, "loss": 0.0598, "step": 730 }, { "grad_norm": 0.414181113243103, "learning_rate": 9.740772327733123e-05, "loss": 0.0592, "step": 740 }, { "grad_norm": 0.48416581749916077, "learning_rate": 9.730158935637697e-05, "loss": 0.0576, "step": 750 }, { "grad_norm": 0.40759462118148804, "learning_rate": 9.719338631066834e-05, "loss": 0.0522, "step": 760 }, { "grad_norm": 0.4843514859676361, "learning_rate": 9.708311887335713e-05, "loss": 0.0551, "step": 770 }, { "grad_norm": 0.37705329060554504, "learning_rate": 9.697079186789823e-05, "loss": 0.0601, "step": 780 }, { "grad_norm": 0.41809579730033875, "learning_rate": 9.685641020783876e-05, "loss": 0.0567, "step": 790 }, { "grad_norm": 0.4055332839488983, "learning_rate": 9.67399788966031e-05, "loss": 0.0569, "step": 800 }, { "grad_norm": 0.384504109621048, "learning_rate": 9.662150302727395e-05, "loss": 0.0586, "step": 810 }, { "grad_norm": 0.38468998670578003, "learning_rate": 9.650098778236968e-05, "loss": 0.0531, "step": 820 }, { "grad_norm": 0.5237128734588623, "learning_rate": 9.637843843361749e-05, "loss": 0.0552, "step": 830 }, { "grad_norm": 0.4317338466644287, "learning_rate": 9.62538603417229e-05, "loss": 0.0509, "step": 840 }, { "grad_norm": 0.3871803879737854, "learning_rate": 9.612725895613526e-05, "loss": 0.0625, "step": 850 }, { "grad_norm": 0.4323013126850128, "learning_rate": 9.599863981480926e-05, "loss": 0.0604, "step": 860 }, { "grad_norm": 0.36297470331192017, "learning_rate": 9.586800854396283e-05, "loss": 0.0545, "step": 870 }, { "grad_norm": 0.37466663122177124, "learning_rate": 9.573537085783095e-05, "loss": 0.0529, "step": 880 }, { "grad_norm": 0.3929672837257385, "learning_rate": 9.560073255841571e-05, "loss": 0.05, "step": 890 }, { "grad_norm": 0.4040852189064026, "learning_rate": 9.546409953523247e-05, "loss": 0.0603, "step": 900 }, { "grad_norm": 0.6248286366462708, "learning_rate": 9.532547776505229e-05, "loss": 0.0492, "step": 910 }, { "grad_norm": 0.5975190997123718, "learning_rate": 9.518487331164048e-05, "loss": 0.0538, "step": 920 }, { "grad_norm": 0.42542824149131775, "learning_rate": 9.504229232549134e-05, "loss": 0.0523, "step": 930 }, { "grad_norm": 0.328101247549057, "learning_rate": 9.489774104355909e-05, "loss": 0.0519, "step": 940 }, { "grad_norm": 0.5055563449859619, "learning_rate": 9.475122578898507e-05, "loss": 0.0603, "step": 950 }, { "grad_norm": 0.5114820003509521, "learning_rate": 9.460275297082119e-05, "loss": 0.0486, "step": 960 }, { "grad_norm": 0.30745837092399597, "learning_rate": 9.445232908374948e-05, "loss": 0.0515, "step": 970 }, { "grad_norm": 0.5132862329483032, "learning_rate": 9.429996070779808e-05, "loss": 0.0534, "step": 980 }, { "grad_norm": 0.3776204586029053, "learning_rate": 9.414565450805333e-05, "loss": 0.0543, "step": 990 }, { "grad_norm": 0.39591264724731445, "learning_rate": 9.398941723436831e-05, "loss": 0.051, "step": 1000 }, { "grad_norm": 0.3092094957828522, "learning_rate": 9.383125572106752e-05, "loss": 0.0485, "step": 1010 }, { "grad_norm": 0.4166562855243683, "learning_rate": 9.367117688664791e-05, "loss": 0.054, "step": 1020 }, { "grad_norm": 0.38996878266334534, "learning_rate": 9.35091877334763e-05, "loss": 0.0506, "step": 1030 }, { "grad_norm": 0.4977191686630249, "learning_rate": 9.334529534748297e-05, "loss": 0.0543, "step": 1040 }, { "grad_norm": 0.47602540254592896, "learning_rate": 9.317950689785188e-05, "loss": 0.048, "step": 1050 }, { "grad_norm": 0.3852064907550812, "learning_rate": 9.301182963670688e-05, "loss": 0.0563, "step": 1060 }, { "grad_norm": 0.2994341254234314, "learning_rate": 9.284227089879456e-05, "loss": 0.0528, "step": 1070 }, { "grad_norm": 0.4053724706172943, "learning_rate": 9.26708381011634e-05, "loss": 0.0546, "step": 1080 }, { "grad_norm": 0.28891879320144653, "learning_rate": 9.249753874283937e-05, "loss": 0.0435, "step": 1090 }, { "grad_norm": 0.2600522041320801, "learning_rate": 9.232238040449779e-05, "loss": 0.0459, "step": 1100 }, { "grad_norm": 0.5400200486183167, "learning_rate": 9.214537074813181e-05, "loss": 0.0553, "step": 1110 }, { "grad_norm": 0.4278520941734314, "learning_rate": 9.196651751671724e-05, "loss": 0.0543, "step": 1120 }, { "grad_norm": 0.43360674381256104, "learning_rate": 9.178582853387384e-05, "loss": 0.058, "step": 1130 }, { "grad_norm": 0.5356864929199219, "learning_rate": 9.160331170352304e-05, "loss": 0.0607, "step": 1140 }, { "grad_norm": 0.37267163395881653, "learning_rate": 9.141897500954229e-05, "loss": 0.0527, "step": 1150 }, { "grad_norm": 0.4231492877006531, "learning_rate": 9.123282651541576e-05, "loss": 0.0515, "step": 1160 }, { "grad_norm": 0.41324272751808167, "learning_rate": 9.104487436388161e-05, "loss": 0.0477, "step": 1170 }, { "grad_norm": 0.2980903089046478, "learning_rate": 9.085512677657582e-05, "loss": 0.0545, "step": 1180 }, { "grad_norm": 0.3594469428062439, "learning_rate": 9.066359205367258e-05, "loss": 0.0519, "step": 1190 }, { "grad_norm": 0.3295166492462158, "learning_rate": 9.047027857352112e-05, "loss": 0.0489, "step": 1200 }, { "grad_norm": 0.35804498195648193, "learning_rate": 9.027519479227935e-05, "loss": 0.0542, "step": 1210 }, { "grad_norm": 0.5596129298210144, "learning_rate": 9.007834924354383e-05, "loss": 0.0492, "step": 1220 }, { "grad_norm": 0.32689106464385986, "learning_rate": 8.987975053797655e-05, "loss": 0.0478, "step": 1230 }, { "grad_norm": 0.3586772680282593, "learning_rate": 8.967940736292825e-05, "loss": 0.0518, "step": 1240 }, { "grad_norm": 0.4046182632446289, "learning_rate": 8.947732848205846e-05, "loss": 0.0561, "step": 1250 }, { "grad_norm": 0.2832528352737427, "learning_rate": 8.927352273495204e-05, "loss": 0.0481, "step": 1260 }, { "grad_norm": 0.38299012184143066, "learning_rate": 8.906799903673265e-05, "loss": 0.0543, "step": 1270 }, { "grad_norm": 0.39338234066963196, "learning_rate": 8.88607663776726e-05, "loss": 0.053, "step": 1280 }, { "grad_norm": 0.326089471578598, "learning_rate": 8.865183382279978e-05, "loss": 0.0584, "step": 1290 }, { "grad_norm": 0.5139309167861938, "learning_rate": 8.844121051150096e-05, "loss": 0.0508, "step": 1300 }, { "grad_norm": 0.2685113251209259, "learning_rate": 8.822890565712211e-05, "loss": 0.0473, "step": 1310 }, { "grad_norm": 0.2396756261587143, "learning_rate": 8.801492854656536e-05, "loss": 0.047, "step": 1320 }, { "grad_norm": 0.2990108132362366, "learning_rate": 8.779928853988268e-05, "loss": 0.0422, "step": 1330 }, { "grad_norm": 0.25585711002349854, "learning_rate": 8.758199506986655e-05, "loss": 0.0443, "step": 1340 }, { "grad_norm": 0.2606576681137085, "learning_rate": 8.73630576416373e-05, "loss": 0.0399, "step": 1350 }, { "grad_norm": 0.43027952313423157, "learning_rate": 8.714248583222726e-05, "loss": 0.0449, "step": 1360 }, { "grad_norm": 0.3107929229736328, "learning_rate": 8.692028929016196e-05, "loss": 0.039, "step": 1370 }, { "grad_norm": 0.26511290669441223, "learning_rate": 8.669647773503797e-05, "loss": 0.0466, "step": 1380 }, { "grad_norm": 0.40483543276786804, "learning_rate": 8.647106095709773e-05, "loss": 0.04, "step": 1390 }, { "grad_norm": 0.3609035909175873, "learning_rate": 8.624404881680139e-05, "loss": 0.0468, "step": 1400 }, { "grad_norm": 0.3585107624530792, "learning_rate": 8.601545124439535e-05, "loss": 0.0396, "step": 1410 }, { "grad_norm": 0.39873549342155457, "learning_rate": 8.5785278239478e-05, "loss": 0.0506, "step": 1420 }, { "grad_norm": 0.26610010862350464, "learning_rate": 8.555353987056224e-05, "loss": 0.0446, "step": 1430 }, { "grad_norm": 0.29464486241340637, "learning_rate": 8.532024627463505e-05, "loss": 0.0449, "step": 1440 }, { "grad_norm": 0.3142518699169159, "learning_rate": 8.508540765671407e-05, "loss": 0.0449, "step": 1450 }, { "grad_norm": 0.3424524962902069, "learning_rate": 8.484903428940121e-05, "loss": 0.0462, "step": 1460 }, { "grad_norm": 0.4664880037307739, "learning_rate": 8.461113651243334e-05, "loss": 0.0499, "step": 1470 }, { "grad_norm": 0.28201568126678467, "learning_rate": 8.437172473222987e-05, "loss": 0.0467, "step": 1480 }, { "grad_norm": 0.4245680868625641, "learning_rate": 8.413080942143767e-05, "loss": 0.0439, "step": 1490 }, { "grad_norm": 0.42034053802490234, "learning_rate": 8.388840111847288e-05, "loss": 0.0532, "step": 1500 }, { "grad_norm": 0.3836017847061157, "learning_rate": 8.364451042705998e-05, "loss": 0.049, "step": 1510 }, { "grad_norm": 0.26328060030937195, "learning_rate": 8.33991480157679e-05, "loss": 0.0427, "step": 1520 }, { "grad_norm": 0.38004162907600403, "learning_rate": 8.315232461754338e-05, "loss": 0.0473, "step": 1530 }, { "grad_norm": 0.45609551668167114, "learning_rate": 8.290405102924144e-05, "loss": 0.0533, "step": 1540 }, { "grad_norm": 0.20536701381206512, "learning_rate": 8.265433811115316e-05, "loss": 0.0436, "step": 1550 }, { "grad_norm": 0.3798106908798218, "learning_rate": 8.240319678653049e-05, "loss": 0.0439, "step": 1560 }, { "grad_norm": 0.28516557812690735, "learning_rate": 8.215063804110857e-05, "loss": 0.0425, "step": 1570 }, { "grad_norm": 0.27644988894462585, "learning_rate": 8.189667292262512e-05, "loss": 0.0403, "step": 1580 }, { "grad_norm": 0.40204596519470215, "learning_rate": 8.164131254033716e-05, "loss": 0.0474, "step": 1590 }, { "grad_norm": 0.43881192803382874, "learning_rate": 8.138456806453503e-05, "loss": 0.0403, "step": 1600 }, { "grad_norm": 0.2845113277435303, "learning_rate": 8.112645072605386e-05, "loss": 0.0392, "step": 1610 }, { "grad_norm": 0.29198166728019714, "learning_rate": 8.086697181578222e-05, "loss": 0.0348, "step": 1620 }, { "grad_norm": 0.3617711663246155, "learning_rate": 8.060614268416823e-05, "loss": 0.0376, "step": 1630 }, { "grad_norm": 0.404875785112381, "learning_rate": 8.034397474072309e-05, "loss": 0.0454, "step": 1640 }, { "grad_norm": 0.22489239275455475, "learning_rate": 8.008047945352193e-05, "loss": 0.0425, "step": 1650 }, { "grad_norm": 0.27072638273239136, "learning_rate": 7.981566834870225e-05, "loss": 0.0383, "step": 1660 }, { "grad_norm": 0.24278584122657776, "learning_rate": 7.954955300995961e-05, "loss": 0.0421, "step": 1670 }, { "grad_norm": 0.3021833002567291, "learning_rate": 7.928214507804104e-05, "loss": 0.0429, "step": 1680 }, { "grad_norm": 0.38383153080940247, "learning_rate": 7.901345625023576e-05, "loss": 0.0436, "step": 1690 }, { "grad_norm": 0.510800838470459, "learning_rate": 7.874349827986354e-05, "loss": 0.0414, "step": 1700 }, { "grad_norm": 0.3297543525695801, "learning_rate": 7.847228297576053e-05, "loss": 0.0436, "step": 1710 }, { "grad_norm": 0.244502454996109, "learning_rate": 7.819982220176276e-05, "loss": 0.0453, "step": 1720 }, { "grad_norm": 0.4956775903701782, "learning_rate": 7.792612787618714e-05, "loss": 0.0466, "step": 1730 }, { "grad_norm": 0.25343015789985657, "learning_rate": 7.765121197131009e-05, "loss": 0.0424, "step": 1740 }, { "grad_norm": 0.26113173365592957, "learning_rate": 7.737508651284391e-05, "loss": 0.0416, "step": 1750 }, { "grad_norm": 0.2182377576828003, "learning_rate": 7.709776357941069e-05, "loss": 0.0441, "step": 1760 }, { "grad_norm": 0.3971315026283264, "learning_rate": 7.681925530201392e-05, "loss": 0.0438, "step": 1770 }, { "grad_norm": 0.28176161646842957, "learning_rate": 7.65395738635079e-05, "loss": 0.0408, "step": 1780 }, { "grad_norm": 0.34179630875587463, "learning_rate": 7.62587314980648e-05, "loss": 0.0479, "step": 1790 }, { "grad_norm": 0.32931986451148987, "learning_rate": 7.597674049063947e-05, "loss": 0.0429, "step": 1800 }, { "grad_norm": 0.39432066679000854, "learning_rate": 7.569361317643211e-05, "loss": 0.0431, "step": 1810 }, { "grad_norm": 0.3576849699020386, "learning_rate": 7.540936194034865e-05, "loss": 0.0384, "step": 1820 }, { "grad_norm": 0.23606491088867188, "learning_rate": 7.512399921645901e-05, "loss": 0.0357, "step": 1830 }, { "grad_norm": 0.2227422446012497, "learning_rate": 7.483753748745317e-05, "loss": 0.0367, "step": 1840 }, { "grad_norm": 0.2424551248550415, "learning_rate": 7.454998928409516e-05, "loss": 0.0385, "step": 1850 }, { "grad_norm": 0.33607348799705505, "learning_rate": 7.426136718467493e-05, "loss": 0.0412, "step": 1860 }, { "grad_norm": 0.2656790018081665, "learning_rate": 7.397168381445812e-05, "loss": 0.0402, "step": 1870 }, { "grad_norm": 0.27483707666397095, "learning_rate": 7.368095184513377e-05, "loss": 0.0332, "step": 1880 }, { "grad_norm": 0.32864516973495483, "learning_rate": 7.338918399426005e-05, "loss": 0.0352, "step": 1890 }, { "grad_norm": 0.2956995964050293, "learning_rate": 7.309639302470801e-05, "loss": 0.0464, "step": 1900 }, { "grad_norm": 0.3073318302631378, "learning_rate": 7.280259174410312e-05, "loss": 0.0378, "step": 1910 }, { "grad_norm": 0.3206455409526825, "learning_rate": 7.250779300426517e-05, "loss": 0.0428, "step": 1920 }, { "grad_norm": 0.40320953726768494, "learning_rate": 7.22120097006461e-05, "loss": 0.0434, "step": 1930 }, { "grad_norm": 0.2006569355726242, "learning_rate": 7.191525477176577e-05, "loss": 0.0367, "step": 1940 }, { "grad_norm": 0.26202642917633057, "learning_rate": 7.161754119864616e-05, "loss": 0.0387, "step": 1950 }, { "grad_norm": 0.36526966094970703, "learning_rate": 7.131888200424339e-05, "loss": 0.0344, "step": 1960 }, { "grad_norm": 0.3327760696411133, "learning_rate": 7.101929025287816e-05, "loss": 0.0401, "step": 1970 }, { "grad_norm": 0.36563345789909363, "learning_rate": 7.071877904966423e-05, "loss": 0.0474, "step": 1980 }, { "grad_norm": 0.27546799182891846, "learning_rate": 7.04173615399351e-05, "loss": 0.0376, "step": 1990 }, { "grad_norm": 0.3782268464565277, "learning_rate": 7.011505090866913e-05, "loss": 0.0324, "step": 2000 } ], "logging_steps": 10, "max_steps": 5000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2000, "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": 32, "trial_name": null, "trial_params": null }