{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 4.663957595825195, "learning_rate": 9e-06, "loss": 1.1524, "step": 10 }, { "grad_norm": 4.667612075805664, "learning_rate": 1.9e-05, "loss": 1.0796, "step": 20 }, { "grad_norm": 5.802227973937988, "learning_rate": 2.9e-05, "loss": 1.1797, "step": 30 }, { "grad_norm": 3.701993703842163, "learning_rate": 3.9000000000000006e-05, "loss": 1.1, "step": 40 }, { "grad_norm": 3.461245059967041, "learning_rate": 4.9e-05, "loss": 1.1146, "step": 50 }, { "grad_norm": 1.6692336797714233, "learning_rate": 5.9e-05, "loss": 1.1527, "step": 60 }, { "grad_norm": 3.9192681312561035, "learning_rate": 6.9e-05, "loss": 1.093, "step": 70 }, { "grad_norm": 1.4972280263900757, "learning_rate": 7.900000000000001e-05, "loss": 1.236, "step": 80 }, { "grad_norm": 9.080512046813965, "learning_rate": 8.900000000000001e-05, "loss": 1.1077, "step": 90 }, { "grad_norm": 3.8250746726989746, "learning_rate": 9.900000000000001e-05, "loss": 1.1689, "step": 100 }, { "grad_norm": 3.9549026489257812, "learning_rate": 9.999446382823013e-05, "loss": 1.1124, "step": 110 }, { "grad_norm": 2.109513998031616, "learning_rate": 9.997532801828658e-05, "loss": 1.1972, "step": 120 }, { "grad_norm": 1.5042897462844849, "learning_rate": 9.99425294526634e-05, "loss": 1.1155, "step": 130 }, { "grad_norm": 3.423783779144287, "learning_rate": 9.989607709816091e-05, "loss": 1.0653, "step": 140 }, { "grad_norm": 1.6116546392440796, "learning_rate": 9.983598365438902e-05, "loss": 1.0256, "step": 150 }, { "grad_norm": 2.693556547164917, "learning_rate": 9.976226555029522e-05, "loss": 1.1256, "step": 160 }, { "grad_norm": 1.924296259880066, "learning_rate": 9.967494293967312e-05, "loss": 1.1095, "step": 170 }, { "grad_norm": 2.4951071739196777, "learning_rate": 9.95740396956525e-05, "loss": 1.1847, "step": 180 }, { "grad_norm": 2.5630929470062256, "learning_rate": 9.945958340417283e-05, "loss": 1.048, "step": 190 }, { "grad_norm": 3.097710371017456, "learning_rate": 9.93316053564413e-05, "loss": 1.1462, "step": 200 }, { "grad_norm": 2.8427069187164307, "learning_rate": 9.919014054037836e-05, "loss": 1.0916, "step": 210 }, { "grad_norm": 2.6329288482666016, "learning_rate": 9.903522763105218e-05, "loss": 1.0583, "step": 220 }, { "grad_norm": 2.420417308807373, "learning_rate": 9.886690898010535e-05, "loss": 1.039, "step": 230 }, { "grad_norm": 3.4768154621124268, "learning_rate": 9.868523060417646e-05, "loss": 1.0799, "step": 240 }, { "grad_norm": 3.935210704803467, "learning_rate": 9.849024217231935e-05, "loss": 1.0871, "step": 250 }, { "grad_norm": 1.2814865112304688, "learning_rate": 9.82819969924244e-05, "loss": 1.0821, "step": 260 }, { "grad_norm": 2.4026999473571777, "learning_rate": 9.806055199664446e-05, "loss": 1.0193, "step": 270 }, { "grad_norm": 1.2491191625595093, "learning_rate": 9.782596772583026e-05, "loss": 1.0296, "step": 280 }, { "grad_norm": 1.4779388904571533, "learning_rate": 9.757830831297914e-05, "loss": 0.9628, "step": 290 }, { "grad_norm": 1.6885578632354736, "learning_rate": 9.731764146570173e-05, "loss": 0.9186, "step": 300 }, { "grad_norm": 2.1885502338409424, "learning_rate": 9.704403844771128e-05, "loss": 0.9338, "step": 310 }, { "grad_norm": 1.7894405126571655, "learning_rate": 9.675757405934103e-05, "loss": 0.901, "step": 320 }, { "grad_norm": 1.5354323387145996, "learning_rate": 9.645832661709444e-05, "loss": 0.8692, "step": 330 }, { "grad_norm": 2.3132033348083496, "learning_rate": 9.614637793223425e-05, "loss": 0.7952, "step": 340 }, { "grad_norm": 1.8628592491149902, "learning_rate": 9.582181328841611e-05, "loss": 0.7426, "step": 350 }, { "grad_norm": 2.144160747528076, "learning_rate": 9.548472141837286e-05, "loss": 0.7966, "step": 360 }, { "grad_norm": 3.1614763736724854, "learning_rate": 9.513519447965595e-05, "loss": 0.7412, "step": 370 }, { "grad_norm": 1.981932282447815, "learning_rate": 9.477332802944044e-05, "loss": 0.7891, "step": 380 }, { "grad_norm": 2.354620933532715, "learning_rate": 9.439922099840054e-05, "loss": 0.7409, "step": 390 }, { "grad_norm": 2.1166627407073975, "learning_rate": 9.401297566366318e-05, "loss": 0.7042, "step": 400 }, { "grad_norm": 1.5316749811172485, "learning_rate": 9.36146976208462e-05, "loss": 0.718, "step": 410 }, { "grad_norm": 2.0734477043151855, "learning_rate": 9.320449575518972e-05, "loss": 0.6956, "step": 420 }, { "grad_norm": 1.30366051197052, "learning_rate": 9.278248221178798e-05, "loss": 0.7278, "step": 430 }, { "grad_norm": 1.0749855041503906, "learning_rate": 9.234877236492997e-05, "loss": 0.7503, "step": 440 }, { "grad_norm": 2.587333917617798, "learning_rate": 9.190348478655724e-05, "loss": 0.7834, "step": 450 }, { "grad_norm": 1.8653922080993652, "learning_rate": 9.144674121384757e-05, "loss": 0.7397, "step": 460 }, { "grad_norm": 2.5511720180511475, "learning_rate": 9.097866651593317e-05, "loss": 0.7529, "step": 470 }, { "grad_norm": 2.498544931411743, "learning_rate": 9.049938865976275e-05, "loss": 0.7312, "step": 480 }, { "grad_norm": 3.6733498573303223, "learning_rate": 9.000903867511666e-05, "loss": 0.6933, "step": 490 }, { "grad_norm": 1.115907073020935, "learning_rate": 8.950775061878453e-05, "loss": 0.6058, "step": 500 }, { "grad_norm": 1.8989201784133911, "learning_rate": 8.899566153791566e-05, "loss": 0.6244, "step": 510 }, { "grad_norm": 2.2771401405334473, "learning_rate": 8.84729114325516e-05, "loss": 0.5869, "step": 520 }, { "grad_norm": 1.6825555562973022, "learning_rate": 8.79396432173515e-05, "loss": 0.596, "step": 530 }, { "grad_norm": 1.4658030271530151, "learning_rate": 8.739600268252078e-05, "loss": 0.5573, "step": 540 }, { "grad_norm": 1.304313063621521, "learning_rate": 8.684213845395339e-05, "loss": 0.5757, "step": 550 }, { "grad_norm": 1.7369334697723389, "learning_rate": 8.627820195259918e-05, "loss": 0.5608, "step": 560 }, { "grad_norm": 1.1636648178100586, "learning_rate": 8.570434735306671e-05, "loss": 0.5879, "step": 570 }, { "grad_norm": 1.6585968732833862, "learning_rate": 8.512073154147362e-05, "loss": 0.5603, "step": 580 }, { "grad_norm": 1.9376773834228516, "learning_rate": 8.452751407255541e-05, "loss": 0.5006, "step": 590 }, { "grad_norm": 10.76678466796875, "learning_rate": 8.392485712604483e-05, "loss": 0.5607, "step": 600 }, { "grad_norm": 2.6736297607421875, "learning_rate": 8.331292546233362e-05, "loss": 0.4497, "step": 610 }, { "grad_norm": 2.1621875762939453, "learning_rate": 8.269188637742846e-05, "loss": 0.4839, "step": 620 }, { "grad_norm": 1.8973023891448975, "learning_rate": 8.206190965721419e-05, "loss": 0.4194, "step": 630 }, { "grad_norm": 1.6862151622772217, "learning_rate": 8.14231675310358e-05, "loss": 0.3488, "step": 640 }, { "grad_norm": 2.992509365081787, "learning_rate": 8.077583462461283e-05, "loss": 0.3396, "step": 650 }, { "grad_norm": 4.031017303466797, "learning_rate": 8.012008791229826e-05, "loss": 0.4537, "step": 660 }, { "grad_norm": 1.3632490634918213, "learning_rate": 7.945610666869568e-05, "loss": 0.3599, "step": 670 }, { "grad_norm": 1.4331293106079102, "learning_rate": 7.878407241964729e-05, "loss": 0.3029, "step": 680 }, { "grad_norm": 2.666959047317505, "learning_rate": 7.810416889260653e-05, "loss": 0.3079, "step": 690 }, { "grad_norm": 2.2112526893615723, "learning_rate": 7.741658196640892e-05, "loss": 0.3407, "step": 700 }, { "grad_norm": 2.248488426208496, "learning_rate": 7.672149962045457e-05, "loss": 0.2885, "step": 710 }, { "grad_norm": 1.4031531810760498, "learning_rate": 7.60191118833165e-05, "loss": 0.2401, "step": 720 }, { "grad_norm": 3.00880765914917, "learning_rate": 7.530961078078873e-05, "loss": 0.2748, "step": 730 }, { "grad_norm": 1.924612283706665, "learning_rate": 7.45931902833884e-05, "loss": 0.3194, "step": 740 }, { "grad_norm": 2.3214099407196045, "learning_rate": 7.387004625332608e-05, "loss": 0.2438, "step": 750 }, { "grad_norm": 1.46572744846344, "learning_rate": 7.3140376390959e-05, "loss": 0.2993, "step": 760 }, { "grad_norm": 1.2204995155334473, "learning_rate": 7.240438018074189e-05, "loss": 0.23, "step": 770 }, { "grad_norm": 1.0724091529846191, "learning_rate": 7.166225883668969e-05, "loss": 0.2436, "step": 780 }, { "grad_norm": 1.7630325555801392, "learning_rate": 7.091421524736784e-05, "loss": 0.2266, "step": 790 }, { "grad_norm": 1.5953762531280518, "learning_rate": 7.016045392042452e-05, "loss": 0.2984, "step": 800 }, { "grad_norm": 1.8009793758392334, "learning_rate": 6.940118092668022e-05, "loss": 0.2518, "step": 810 }, { "grad_norm": 1.5019856691360474, "learning_rate": 6.863660384379017e-05, "loss": 0.2146, "step": 820 }, { "grad_norm": 2.158843755722046, "learning_rate": 6.786693169949455e-05, "loss": 0.2234, "step": 830 }, { "grad_norm": 0.8007408976554871, "learning_rate": 6.709237491447249e-05, "loss": 0.2121, "step": 840 }, { "grad_norm": 0.7677724957466125, "learning_rate": 6.631314524481513e-05, "loss": 0.2001, "step": 850 }, { "grad_norm": 1.7501623630523682, "learning_rate": 6.552945572413358e-05, "loss": 0.1867, "step": 860 }, { "grad_norm": 1.8922343254089355, "learning_rate": 6.474152060531768e-05, "loss": 0.2217, "step": 870 }, { "grad_norm": 3.635430335998535, "learning_rate": 6.394955530196147e-05, "loss": 0.1905, "step": 880 }, { "grad_norm": 1.8172924518585205, "learning_rate": 6.315377632947115e-05, "loss": 0.1388, "step": 890 }, { "grad_norm": 3.030200481414795, "learning_rate": 6.235440124587198e-05, "loss": 0.2962, "step": 900 }, { "grad_norm": 1.272994041442871, "learning_rate": 6.155164859233012e-05, "loss": 0.1492, "step": 910 }, { "grad_norm": 2.5055532455444336, "learning_rate": 6.074573783340562e-05, "loss": 0.1007, "step": 920 }, { "grad_norm": 1.443792700767517, "learning_rate": 5.9936889297052986e-05, "loss": 0.0678, "step": 930 }, { "grad_norm": 1.208217740058899, "learning_rate": 5.912532411438576e-05, "loss": 0.0678, "step": 940 }, { "grad_norm": 1.4550530910491943, "learning_rate": 5.831126415922148e-05, "loss": 0.1533, "step": 950 }, { "grad_norm": 1.3415378332138062, "learning_rate": 5.74949319874235e-05, "loss": 0.1581, "step": 960 }, { "grad_norm": 1.722151279449463, "learning_rate": 5.667655077605659e-05, "loss": 0.076, "step": 970 }, { "grad_norm": 1.203705906867981, "learning_rate": 5.585634426237246e-05, "loss": 0.086, "step": 980 }, { "grad_norm": 1.511094570159912, "learning_rate": 5.5034536682642224e-05, "loss": 0.0675, "step": 990 }, { "grad_norm": 1.0288337469100952, "learning_rate": 5.4211352710852495e-05, "loss": 0.0525, "step": 1000 }, { "grad_norm": 1.886352777481079, "learning_rate": 5.3387017397281704e-05, "loss": 0.0676, "step": 1010 }, { "grad_norm": 0.8772637844085693, "learning_rate": 5.2561756106973656e-05, "loss": 0.0406, "step": 1020 }, { "grad_norm": 1.5332303047180176, "learning_rate": 5.1735794458124956e-05, "loss": 0.0518, "step": 1030 }, { "grad_norm": 1.2781940698623657, "learning_rate": 5.0909358260403186e-05, "loss": 0.0469, "step": 1040 }, { "grad_norm": 1.253562092781067, "learning_rate": 5.0082673453212914e-05, "loss": 0.0499, "step": 1050 }, { "grad_norm": 0.7091103196144104, "learning_rate": 4.925596604392603e-05, "loss": 0.0603, "step": 1060 }, { "grad_norm": 0.8219345211982727, "learning_rate": 4.8429462046093585e-05, "loss": 0.052, "step": 1070 }, { "grad_norm": 1.0253986120224, "learning_rate": 4.7603387417656026e-05, "loss": 0.0446, "step": 1080 }, { "grad_norm": 1.1790286302566528, "learning_rate": 4.677796799916845e-05, "loss": 0.041, "step": 1090 }, { "grad_norm": 1.1473667621612549, "learning_rate": 4.5953429452058135e-05, "loss": 0.0438, "step": 1100 }, { "grad_norm": 1.0651497840881348, "learning_rate": 4.5129997196930845e-05, "loss": 0.0387, "step": 1110 }, { "grad_norm": 1.6050951480865479, "learning_rate": 4.430789635194324e-05, "loss": 0.0633, "step": 1120 }, { "grad_norm": 0.9059212803840637, "learning_rate": 4.348735167125771e-05, "loss": 0.0548, "step": 1130 }, { "grad_norm": 1.545888900756836, "learning_rate": 4.2668587483596864e-05, "loss": 0.0376, "step": 1140 }, { "grad_norm": 1.7836066484451294, "learning_rate": 4.1851827630914305e-05, "loss": 0.0878, "step": 1150 }, { "grad_norm": 1.217666506767273, "learning_rate": 4.103729540719847e-05, "loss": 0.066, "step": 1160 }, { "grad_norm": 1.0235333442687988, "learning_rate": 4.0225213497426276e-05, "loss": 0.068, "step": 1170 }, { "grad_norm": 1.7279560565948486, "learning_rate": 3.9415803916683224e-05, "loss": 0.1933, "step": 1180 }, { "grad_norm": 1.3281042575836182, "learning_rate": 3.860928794946682e-05, "loss": 0.232, "step": 1190 }, { "grad_norm": 2.720484733581543, "learning_rate": 3.780588608918947e-05, "loss": 0.2442, "step": 1200 }, { "grad_norm": 1.7740724086761475, "learning_rate": 3.700581797789786e-05, "loss": 0.2427, "step": 1210 }, { "grad_norm": 1.5594145059585571, "learning_rate": 3.6209302346225006e-05, "loss": 0.3945, "step": 1220 }, { "grad_norm": 3.00752329826355, "learning_rate": 3.541655695359142e-05, "loss": 0.2323, "step": 1230 }, { "grad_norm": 1.3963649272918701, "learning_rate": 3.462779852867197e-05, "loss": 0.2093, "step": 1240 }, { "grad_norm": 1.139393925666809, "learning_rate": 3.384324271014429e-05, "loss": 0.1642, "step": 1250 }, { "grad_norm": 1.4123072624206543, "learning_rate": 3.3063103987735433e-05, "loss": 0.2023, "step": 1260 }, { "grad_norm": 1.3841872215270996, "learning_rate": 3.228759564358248e-05, "loss": 0.2041, "step": 1270 }, { "grad_norm": 0.7465299367904663, "learning_rate": 3.1516929693923315e-05, "loss": 0.1444, "step": 1280 }, { "grad_norm": 1.7622781991958618, "learning_rate": 3.075131683113352e-05, "loss": 0.2116, "step": 1290 }, { "grad_norm": 3.3836557865142822, "learning_rate": 2.999096636612518e-05, "loss": 0.2211, "step": 1300 }, { "grad_norm": 1.1476629972457886, "learning_rate": 2.9236086171123404e-05, "loss": 0.1872, "step": 1310 }, { "grad_norm": 2.681485891342163, "learning_rate": 2.8486882622836026e-05, "loss": 0.1984, "step": 1320 }, { "grad_norm": 1.6360139846801758, "learning_rate": 2.774356054603243e-05, "loss": 0.2322, "step": 1330 }, { "grad_norm": 1.6883631944656372, "learning_rate": 2.7006323157546386e-05, "loss": 0.1764, "step": 1340 }, { "grad_norm": 3.836796522140503, "learning_rate": 2.6275372010718635e-05, "loss": 0.2608, "step": 1350 }, { "grad_norm": 3.1437857151031494, "learning_rate": 2.555090694029421e-05, "loss": 0.1633, "step": 1360 }, { "grad_norm": 3.7042806148529053, "learning_rate": 2.4833126007789653e-05, "loss": 0.1778, "step": 1370 }, { "grad_norm": 3.1288084983825684, "learning_rate": 2.4122225447344875e-05, "loss": 0.2279, "step": 1380 }, { "grad_norm": 1.2500102519989014, "learning_rate": 2.341839961207482e-05, "loss": 0.1595, "step": 1390 }, { "grad_norm": 2.126539468765259, "learning_rate": 2.2721840920935196e-05, "loss": 0.1551, "step": 1400 }, { "grad_norm": 1.9851874113082886, "learning_rate": 2.2032739806117058e-05, "loss": 0.1668, "step": 1410 }, { "grad_norm": 3.7958602905273438, "learning_rate": 2.1351284660984572e-05, "loss": 0.2309, "step": 1420 }, { "grad_norm": 2.6419732570648193, "learning_rate": 2.067766178857013e-05, "loss": 0.1617, "step": 1430 }, { "grad_norm": 2.527336835861206, "learning_rate": 2.0012055350640986e-05, "loss": 0.1241, "step": 1440 }, { "grad_norm": 2.7349932193756104, "learning_rate": 1.9354647317351188e-05, "loss": 0.2024, "step": 1450 }, { "grad_norm": 1.515536904335022, "learning_rate": 1.8705617417492883e-05, "loss": 0.1136, "step": 1460 }, { "grad_norm": 5.851349353790283, "learning_rate": 1.8065143089360172e-05, "loss": 0.2145, "step": 1470 }, { "grad_norm": 1.833863377571106, "learning_rate": 1.743339943223926e-05, "loss": 0.1337, "step": 1480 }, { "grad_norm": 2.6687817573547363, "learning_rate": 1.6810559158538092e-05, "loss": 0.2251, "step": 1490 }, { "grad_norm": 1.4367228746414185, "learning_rate": 1.6196792546568472e-05, "loss": 0.2292, "step": 1500 }, { "grad_norm": 1.5391637086868286, "learning_rate": 1.5592267393993716e-05, "loss": 0.1254, "step": 1510 }, { "grad_norm": 1.3211592435836792, "learning_rate": 1.4997148971954344e-05, "loss": 0.1557, "step": 1520 }, { "grad_norm": 1.6100497245788574, "learning_rate": 1.4411599979884744e-05, "loss": 0.198, "step": 1530 }, { "grad_norm": 2.344667673110962, "learning_rate": 1.383578050103268e-05, "loss": 0.2409, "step": 1540 }, { "grad_norm": 1.045395851135254, "learning_rate": 1.3269847958694148e-05, "loss": 0.2647, "step": 1550 }, { "grad_norm": 1.9208920001983643, "learning_rate": 1.2713957073175425e-05, "loss": 0.1756, "step": 1560 }, { "grad_norm": 1.3460582494735718, "learning_rate": 1.2168259819494066e-05, "loss": 0.2055, "step": 1570 }, { "grad_norm": 1.985026478767395, "learning_rate": 1.1632905385830484e-05, "loss": 0.1652, "step": 1580 }, { "grad_norm": 2.376067638397217, "learning_rate": 1.1108040132741354e-05, "loss": 0.1727, "step": 1590 }, { "grad_norm": 3.2674777507781982, "learning_rate": 1.059380755314613e-05, "loss": 0.2962, "step": 1600 }, { "grad_norm": 1.835555076599121, "learning_rate": 1.009034823309749e-05, "loss": 0.2349, "step": 1610 }, { "grad_norm": 2.164524555206299, "learning_rate": 9.597799813346525e-06, "loss": 0.2314, "step": 1620 }, { "grad_norm": 2.483271598815918, "learning_rate": 9.116296951713133e-06, "loss": 0.2598, "step": 1630 }, { "grad_norm": 4.282663822174072, "learning_rate": 8.645971286271904e-06, "loss": 0.2772, "step": 1640 }, { "grad_norm": 1.270452618598938, "learning_rate": 8.186951399363613e-06, "loss": 0.1543, "step": 1650 }, { "grad_norm": 1.2190138101577759, "learning_rate": 7.739362782442021e-06, "loss": 0.1863, "step": 1660 }, { "grad_norm": 2.430107355117798, "learning_rate": 7.30332780176588e-06, "loss": 0.1834, "step": 1670 }, { "grad_norm": 1.1952879428863525, "learning_rate": 6.878965664945108e-06, "loss": 0.143, "step": 1680 }, { "grad_norm": 1.5379353761672974, "learning_rate": 6.466392388350695e-06, "loss": 0.2987, "step": 1690 }, { "grad_norm": 1.7996628284454346, "learning_rate": 6.0657207653969315e-06, "loss": 0.1935, "step": 1700 }, { "grad_norm": 1.3230831623077393, "learning_rate": 5.67706033570487e-06, "loss": 0.2057, "step": 1710 }, { "grad_norm": 2.430042028427124, "learning_rate": 5.300517355155215e-06, "loss": 0.1465, "step": 1720 }, { "grad_norm": 2.063180923461914, "learning_rate": 4.936194766839103e-06, "loss": 0.2015, "step": 1730 }, { "grad_norm": 0.8021765947341919, "learning_rate": 4.5841921729144424e-06, "loss": 0.2467, "step": 1740 }, { "grad_norm": 1.857146978378296, "learning_rate": 4.244605807375679e-06, "loss": 0.1839, "step": 1750 }, { "grad_norm": 1.9084161520004272, "learning_rate": 3.917528509744412e-06, "loss": 0.1634, "step": 1760 }, { "grad_norm": 3.154071807861328, "learning_rate": 3.60304969968796e-06, "loss": 0.1476, "step": 1770 }, { "grad_norm": 3.2903988361358643, "learning_rate": 3.301255352572946e-06, "loss": 0.294, "step": 1780 }, { "grad_norm": 1.6087337732315063, "learning_rate": 3.0122279759604745e-06, "loss": 0.2048, "step": 1790 }, { "grad_norm": 1.41805899143219, "learning_rate": 2.73604658704939e-06, "loss": 0.1701, "step": 1800 }, { "grad_norm": 2.2126808166503906, "learning_rate": 2.4727866910737583e-06, "loss": 0.1825, "step": 1810 }, { "grad_norm": 1.390885353088379, "learning_rate": 2.222520260660521e-06, "loss": 0.2563, "step": 1820 }, { "grad_norm": 2.4333114624023438, "learning_rate": 1.985315716152847e-06, "loss": 0.2607, "step": 1830 }, { "grad_norm": 2.5737111568450928, "learning_rate": 1.7612379069047335e-06, "loss": 0.2266, "step": 1840 }, { "grad_norm": 4.825987339019775, "learning_rate": 1.550348093551829e-06, "loss": 0.2847, "step": 1850 }, { "grad_norm": 2.4544243812561035, "learning_rate": 1.3527039312633827e-06, "loss": 0.17, "step": 1860 }, { "grad_norm": 1.0220296382904053, "learning_rate": 1.1683594539798893e-06, "loss": 0.1591, "step": 1870 }, { "grad_norm": 1.7394435405731201, "learning_rate": 9.97365059640787e-07, "loss": 0.2198, "step": 1880 }, { "grad_norm": 1.1420143842697144, "learning_rate": 8.397674964061075e-07, "loss": 0.1964, "step": 1890 }, { "grad_norm": 2.1842455863952637, "learning_rate": 6.956098498760389e-07, "loss": 0.1718, "step": 1900 }, { "grad_norm": 2.5906801223754883, "learning_rate": 5.64931531311741e-07, "loss": 0.2554, "step": 1910 }, { "grad_norm": 1.515612006187439, "learning_rate": 4.4776826686069305e-07, "loss": 0.1547, "step": 1920 }, { "grad_norm": 2.444715738296509, "learning_rate": 3.441520877895288e-07, "loss": 0.2279, "step": 1930 }, { "grad_norm": 1.8030273914337158, "learning_rate": 2.5411132172700194e-07, "loss": 0.2032, "step": 1940 }, { "grad_norm": 1.8206000328063965, "learning_rate": 1.776705849195037e-07, "loss": 0.2283, "step": 1950 }, { "grad_norm": 1.7573992013931274, "learning_rate": 1.1485077550122402e-07, "loss": 0.2943, "step": 1960 }, { "grad_norm": 2.789184808731079, "learning_rate": 6.566906778079917e-08, "loss": 0.2005, "step": 1970 }, { "grad_norm": 3.3766531944274902, "learning_rate": 3.01389075460512e-08, "loss": 0.2293, "step": 1980 }, { "grad_norm": 2.624661445617676, "learning_rate": 8.270008388022721e-09, "loss": 0.1843, "step": 1990 }, { "grad_norm": 1.9280586242675781, "learning_rate": 6.834904537900144e-11, "loss": 0.2011, "step": 2000 }, { "grad_norm": 1.9142124652862549, "learning_rate": 1.8e-06, "loss": 0.098, "step": 2010 }, { "grad_norm": 1.8881168365478516, "learning_rate": 3.8e-06, "loss": 0.0888, "step": 2020 }, { "grad_norm": 0.9192376732826233, "learning_rate": 5.8e-06, "loss": 0.0517, "step": 2030 }, { "grad_norm": 1.1629652976989746, "learning_rate": 7.8e-06, "loss": 0.0554, "step": 2040 }, { "grad_norm": 0.7949849963188171, "learning_rate": 9.800000000000001e-06, "loss": 0.0373, "step": 2050 }, { "grad_norm": 1.211435079574585, "learning_rate": 1.18e-05, "loss": 0.0366, "step": 2060 }, { "grad_norm": 0.67411869764328, "learning_rate": 1.3800000000000002e-05, "loss": 0.0329, "step": 2070 }, { "grad_norm": 0.9655842781066895, "learning_rate": 1.58e-05, "loss": 0.0299, "step": 2080 }, { "grad_norm": 0.5895830988883972, "learning_rate": 1.78e-05, "loss": 0.0339, "step": 2090 }, { "grad_norm": 1.4831079244613647, "learning_rate": 1.9800000000000004e-05, "loss": 0.0542, "step": 2100 }, { "grad_norm": 0.7277588844299316, "learning_rate": 2.18e-05, "loss": 0.0422, "step": 2110 }, { "grad_norm": 1.3350826501846313, "learning_rate": 2.38e-05, "loss": 0.0363, "step": 2120 }, { "grad_norm": 0.9681083559989929, "learning_rate": 2.58e-05, "loss": 0.0477, "step": 2130 }, { "grad_norm": 0.4943430423736572, "learning_rate": 2.7800000000000005e-05, "loss": 0.0257, "step": 2140 }, { "grad_norm": 0.7951409220695496, "learning_rate": 2.98e-05, "loss": 0.0278, "step": 2150 }, { "grad_norm": 1.134987473487854, "learning_rate": 3.18e-05, "loss": 0.0583, "step": 2160 }, { "grad_norm": 0.5520042181015015, "learning_rate": 3.38e-05, "loss": 0.0309, "step": 2170 }, { "grad_norm": 0.7561506032943726, "learning_rate": 3.58e-05, "loss": 0.026, "step": 2180 }, { "grad_norm": 1.1808117628097534, "learning_rate": 3.7800000000000004e-05, "loss": 0.0469, "step": 2190 }, { "grad_norm": 1.125457763671875, "learning_rate": 3.9800000000000005e-05, "loss": 0.0399, "step": 2200 }, { "grad_norm": 0.9655048251152039, "learning_rate": 4.18e-05, "loss": 0.0442, "step": 2210 }, { "grad_norm": 1.6301419734954834, "learning_rate": 4.38e-05, "loss": 0.0315, "step": 2220 }, { "grad_norm": 1.4245551824569702, "learning_rate": 4.58e-05, "loss": 0.0668, "step": 2230 }, { "grad_norm": 1.2428812980651855, "learning_rate": 4.78e-05, "loss": 0.0507, "step": 2240 }, { "grad_norm": 2.431546688079834, "learning_rate": 4.9800000000000004e-05, "loss": 0.0514, "step": 2250 }, { "grad_norm": 0.9127787351608276, "learning_rate": 5.1800000000000005e-05, "loss": 0.052, "step": 2260 }, { "grad_norm": 1.9670411348342896, "learning_rate": 5.380000000000001e-05, "loss": 0.0472, "step": 2270 }, { "grad_norm": 1.1853229999542236, "learning_rate": 5.580000000000001e-05, "loss": 0.0725, "step": 2280 }, { "grad_norm": 1.1005306243896484, "learning_rate": 5.7799999999999995e-05, "loss": 0.0458, "step": 2290 }, { "grad_norm": 0.49564310908317566, "learning_rate": 5.9800000000000003e-05, "loss": 0.0458, "step": 2300 }, { "grad_norm": 0.7596853375434875, "learning_rate": 6.18e-05, "loss": 0.0446, "step": 2310 }, { "grad_norm": 0.5753937363624573, "learning_rate": 6.38e-05, "loss": 0.0314, "step": 2320 }, { "grad_norm": 1.1597989797592163, "learning_rate": 6.58e-05, "loss": 0.0641, "step": 2330 }, { "grad_norm": 1.2174358367919922, "learning_rate": 6.780000000000001e-05, "loss": 0.0503, "step": 2340 }, { "grad_norm": 1.2351247072219849, "learning_rate": 6.98e-05, "loss": 0.0355, "step": 2350 }, { "grad_norm": 1.1845176219940186, "learning_rate": 7.18e-05, "loss": 0.0603, "step": 2360 }, { "grad_norm": 1.5108373165130615, "learning_rate": 7.38e-05, "loss": 0.0686, "step": 2370 }, { "grad_norm": 1.0825536251068115, "learning_rate": 7.58e-05, "loss": 0.0881, "step": 2380 }, { "grad_norm": 0.7487830519676208, "learning_rate": 7.780000000000001e-05, "loss": 0.057, "step": 2390 }, { "grad_norm": 1.3652535676956177, "learning_rate": 7.98e-05, "loss": 0.0559, "step": 2400 }, { "grad_norm": 1.285368800163269, "learning_rate": 8.18e-05, "loss": 0.0822, "step": 2410 }, { "grad_norm": 1.393035888671875, "learning_rate": 8.38e-05, "loss": 0.064, "step": 2420 }, { "grad_norm": 1.752549171447754, "learning_rate": 8.58e-05, "loss": 0.0817, "step": 2430 }, { "grad_norm": 0.7709497213363647, "learning_rate": 8.78e-05, "loss": 0.0524, "step": 2440 }, { "grad_norm": 1.29141104221344, "learning_rate": 8.98e-05, "loss": 0.0587, "step": 2450 }, { "grad_norm": 0.684868574142456, "learning_rate": 9.180000000000001e-05, "loss": 0.0461, "step": 2460 }, { "grad_norm": 1.4486169815063477, "learning_rate": 9.38e-05, "loss": 0.1188, "step": 2470 }, { "grad_norm": 0.8844814300537109, "learning_rate": 9.58e-05, "loss": 0.1112, "step": 2480 }, { "grad_norm": 1.1087321043014526, "learning_rate": 9.78e-05, "loss": 0.0606, "step": 2490 }, { "grad_norm": 1.0692814588546753, "learning_rate": 9.98e-05, "loss": 0.12, "step": 2500 }, { "grad_norm": 1.3575022220611572, "learning_rate": 9.9999778549206e-05, "loss": 0.1, "step": 2510 }, { "grad_norm": 1.0319279432296753, "learning_rate": 9.999901304280685e-05, "loss": 0.0488, "step": 2520 }, { "grad_norm": 1.097380518913269, "learning_rate": 9.999770075521164e-05, "loss": 0.0668, "step": 2530 }, { "grad_norm": 1.5954612493515015, "learning_rate": 9.99958417007713e-05, "loss": 0.0836, "step": 2540 }, { "grad_norm": 0.9431521892547607, "learning_rate": 9.999343589981615e-05, "loss": 0.0724, "step": 2550 }, { "grad_norm": 2.0689821243286133, "learning_rate": 9.999048337865568e-05, "loss": 0.0829, "step": 2560 }, { "grad_norm": 0.9491808414459229, "learning_rate": 9.998698416957815e-05, "loss": 0.0998, "step": 2570 }, { "grad_norm": 2.049532651901245, "learning_rate": 9.998293831085037e-05, "loss": 0.0633, "step": 2580 }, { "grad_norm": 3.1709156036376953, "learning_rate": 9.997834584671719e-05, "loss": 0.0563, "step": 2590 }, { "grad_norm": 1.3370786905288696, "learning_rate": 9.997320682740107e-05, "loss": 0.1264, "step": 2600 }, { "grad_norm": 1.389707326889038, "learning_rate": 9.996752130910149e-05, "loss": 0.0554, "step": 2610 }, { "grad_norm": 1.0643879175186157, "learning_rate": 9.99612893539944e-05, "loss": 0.0783, "step": 2620 }, { "grad_norm": 1.0231214761734009, "learning_rate": 9.995451103023144e-05, "loss": 0.0556, "step": 2630 }, { "grad_norm": 0.980778694152832, "learning_rate": 9.994718641193928e-05, "loss": 0.06, "step": 2640 }, { "grad_norm": 1.1619300842285156, "learning_rate": 9.993931557921874e-05, "loss": 0.055, "step": 2650 }, { "grad_norm": 1.440382719039917, "learning_rate": 9.993089861814402e-05, "loss": 0.0663, "step": 2660 }, { "grad_norm": 1.3679347038269043, "learning_rate": 9.992193562076166e-05, "loss": 0.0554, "step": 2670 }, { "grad_norm": 1.2099665403366089, "learning_rate": 9.991242668508954e-05, "loss": 0.0577, "step": 2680 }, { "grad_norm": 1.2164071798324585, "learning_rate": 9.990237191511587e-05, "loss": 0.089, "step": 2690 }, { "grad_norm": 4.013327121734619, "learning_rate": 9.989177142079802e-05, "loss": 0.0962, "step": 2700 }, { "grad_norm": 0.8404901623725891, "learning_rate": 9.988062531806126e-05, "loss": 0.0522, "step": 2710 }, { "grad_norm": 1.439043641090393, "learning_rate": 9.986893372879762e-05, "loss": 0.0376, "step": 2720 }, { "grad_norm": 1.623354434967041, "learning_rate": 9.985669678086443e-05, "loss": 0.0617, "step": 2730 }, { "grad_norm": 1.7078526020050049, "learning_rate": 9.984391460808298e-05, "loss": 0.067, "step": 2740 }, { "grad_norm": 1.0720077753067017, "learning_rate": 9.983058735023709e-05, "loss": 0.0689, "step": 2750 }, { "grad_norm": 2.2524731159210205, "learning_rate": 9.98167151530715e-05, "loss": 0.114, "step": 2760 }, { "grad_norm": 1.3048789501190186, "learning_rate": 9.980229816829034e-05, "loss": 0.0707, "step": 2770 }, { "grad_norm": 0.7458677291870117, "learning_rate": 9.978733655355544e-05, "loss": 0.0596, "step": 2780 }, { "grad_norm": 1.6696217060089111, "learning_rate": 9.977183047248464e-05, "loss": 0.0608, "step": 2790 }, { "grad_norm": 1.2139520645141602, "learning_rate": 9.975578009464992e-05, "loss": 0.1024, "step": 2800 }, { "grad_norm": 0.6063175797462463, "learning_rate": 9.97391855955757e-05, "loss": 0.0712, "step": 2810 }, { "grad_norm": 0.8264336585998535, "learning_rate": 9.972204715673669e-05, "loss": 0.0371, "step": 2820 }, { "grad_norm": 0.7754390239715576, "learning_rate": 9.970436496555617e-05, "loss": 0.0524, "step": 2830 }, { "grad_norm": 0.7724748849868774, "learning_rate": 9.968613921540373e-05, "loss": 0.0471, "step": 2840 }, { "grad_norm": 0.7410463690757751, "learning_rate": 9.966737010559326e-05, "loss": 0.0539, "step": 2850 }, { "grad_norm": 1.09246826171875, "learning_rate": 9.964805784138072e-05, "loss": 0.0477, "step": 2860 }, { "grad_norm": 0.6059792041778564, "learning_rate": 9.962820263396195e-05, "loss": 0.0648, "step": 2870 }, { "grad_norm": 1.7347480058670044, "learning_rate": 9.960780470047033e-05, "loss": 0.0979, "step": 2880 }, { "grad_norm": 0.4212770164012909, "learning_rate": 9.958686426397437e-05, "loss": 0.0527, "step": 2890 }, { "grad_norm": 3.461780309677124, "learning_rate": 9.956538155347534e-05, "loss": 0.6297, "step": 2900 }, { "grad_norm": 3.5302493572235107, "learning_rate": 9.95433568039047e-05, "loss": 0.3911, "step": 2910 }, { "grad_norm": 2.2311391830444336, "learning_rate": 9.952079025612162e-05, "loss": 0.3361, "step": 2920 }, { "grad_norm": 1.6683107614517212, "learning_rate": 9.949768215691022e-05, "loss": 0.2647, "step": 2930 }, { "grad_norm": 1.979753851890564, "learning_rate": 9.9474032758977e-05, "loss": 0.3418, "step": 2940 }, { "grad_norm": 2.5483880043029785, "learning_rate": 9.944984232094794e-05, "loss": 0.3601, "step": 2950 }, { "grad_norm": 1.6190166473388672, "learning_rate": 9.942511110736584e-05, "loss": 0.29, "step": 2960 }, { "grad_norm": 1.819980263710022, "learning_rate": 9.939983938868726e-05, "loss": 0.316, "step": 2970 }, { "grad_norm": 1.3827078342437744, "learning_rate": 9.93740274412797e-05, "loss": 0.2735, "step": 2980 }, { "grad_norm": 2.797220230102539, "learning_rate": 9.934767554741846e-05, "loss": 0.2859, "step": 2990 }, { "grad_norm": 2.9383599758148193, "learning_rate": 9.932078399528361e-05, "loss": 0.1712, "step": 3000 }, { "grad_norm": 1.9055110216140747, "learning_rate": 9.929335307895689e-05, "loss": 0.2631, "step": 3010 }, { "grad_norm": 1.5297774076461792, "learning_rate": 9.926538309841839e-05, "loss": 0.1784, "step": 3020 }, { "grad_norm": 2.2921905517578125, "learning_rate": 9.923687435954334e-05, "loss": 0.1328, "step": 3030 }, { "grad_norm": 1.52601158618927, "learning_rate": 9.920782717409873e-05, "loss": 0.1807, "step": 3040 }, { "grad_norm": 2.3993465900421143, "learning_rate": 9.917824185973994e-05, "loss": 0.2015, "step": 3050 }, { "grad_norm": 1.6294753551483154, "learning_rate": 9.914811874000723e-05, "loss": 0.2706, "step": 3060 }, { "grad_norm": 2.141359329223633, "learning_rate": 9.911745814432218e-05, "loss": 0.2403, "step": 3070 }, { "grad_norm": 3.0772581100463867, "learning_rate": 9.90862604079842e-05, "loss": 0.241, "step": 3080 }, { "grad_norm": 4.836127758026123, "learning_rate": 9.90545258721667e-05, "loss": 0.2831, "step": 3090 }, { "grad_norm": 1.815845012664795, "learning_rate": 9.90222548839135e-05, "loss": 0.2196, "step": 3100 }, { "grad_norm": 2.060487747192383, "learning_rate": 9.898944779613495e-05, "loss": 0.1687, "step": 3110 }, { "grad_norm": 3.1497676372528076, "learning_rate": 9.89561049676041e-05, "loss": 0.198, "step": 3120 }, { "grad_norm": 2.537574291229248, "learning_rate": 9.89222267629528e-05, "loss": 0.2894, "step": 3130 }, { "grad_norm": 3.6005682945251465, "learning_rate": 9.888781355266763e-05, "loss": 0.3254, "step": 3140 }, { "grad_norm": 1.9073723554611206, "learning_rate": 9.885286571308598e-05, "loss": 0.2083, "step": 3150 }, { "grad_norm": 2.5607001781463623, "learning_rate": 9.881738362639182e-05, "loss": 0.3574, "step": 3160 }, { "grad_norm": 1.1356687545776367, "learning_rate": 9.878136768061154e-05, "loss": 0.2686, "step": 3170 }, { "grad_norm": 2.5611162185668945, "learning_rate": 9.874481826960979e-05, "loss": 0.3165, "step": 3180 }, { "grad_norm": 2.709408760070801, "learning_rate": 9.870773579308503e-05, "loss": 0.2255, "step": 3190 }, { "grad_norm": 1.8751304149627686, "learning_rate": 9.867012065656533e-05, "loss": 0.2991, "step": 3200 }, { "grad_norm": 1.6324174404144287, "learning_rate": 9.863197327140376e-05, "loss": 0.2666, "step": 3210 }, { "grad_norm": 1.1242610216140747, "learning_rate": 9.859329405477403e-05, "loss": 0.2422, "step": 3220 }, { "grad_norm": 2.5316553115844727, "learning_rate": 9.855408342966585e-05, "loss": 0.1953, "step": 3230 }, { "grad_norm": 1.1990642547607422, "learning_rate": 9.851434182488033e-05, "loss": 0.1496, "step": 3240 }, { "grad_norm": 1.8103933334350586, "learning_rate": 9.84740696750253e-05, "loss": 0.2639, "step": 3250 }, { "grad_norm": 2.2604472637176514, "learning_rate": 9.843326742051055e-05, "loss": 0.1598, "step": 3260 }, { "grad_norm": 0.8064984679222107, "learning_rate": 9.839193550754297e-05, "loss": 0.1656, "step": 3270 }, { "grad_norm": 1.099325180053711, "learning_rate": 9.835007438812177e-05, "loss": 0.1081, "step": 3280 }, { "grad_norm": 1.2965447902679443, "learning_rate": 9.830768452003341e-05, "loss": 0.2265, "step": 3290 }, { "grad_norm": 2.2290050983428955, "learning_rate": 9.826476636684671e-05, "loss": 0.1419, "step": 3300 }, { "grad_norm": 1.5989066362380981, "learning_rate": 9.822132039790773e-05, "loss": 0.1946, "step": 3310 }, { "grad_norm": 2.237204074859619, "learning_rate": 9.817734708833461e-05, "loss": 0.3086, "step": 3320 }, { "grad_norm": 3.4248452186584473, "learning_rate": 9.813284691901243e-05, "loss": 0.2033, "step": 3330 }, { "grad_norm": 1.3145263195037842, "learning_rate": 9.808782037658792e-05, "loss": 0.1313, "step": 3340 }, { "grad_norm": 1.4613280296325684, "learning_rate": 9.804226795346411e-05, "loss": 0.2036, "step": 3350 }, { "grad_norm": 1.9568893909454346, "learning_rate": 9.799619014779503e-05, "loss": 0.2002, "step": 3360 }, { "grad_norm": 1.7012279033660889, "learning_rate": 9.794958746348013e-05, "loss": 0.1189, "step": 3370 }, { "grad_norm": 1.9728801250457764, "learning_rate": 9.790246041015896e-05, "loss": 0.1705, "step": 3380 }, { "grad_norm": 4.486622333526611, "learning_rate": 9.785480950320538e-05, "loss": 0.2737, "step": 3390 }, { "grad_norm": 4.9826340675354, "learning_rate": 9.78066352637221e-05, "loss": 0.2324, "step": 3400 }, { "grad_norm": 1.5212738513946533, "learning_rate": 9.775793821853488e-05, "loss": 0.1564, "step": 3410 }, { "grad_norm": 2.05259108543396, "learning_rate": 9.77087189001868e-05, "loss": 0.1495, "step": 3420 }, { "grad_norm": 1.6725037097930908, "learning_rate": 9.765897784693243e-05, "loss": 0.1146, "step": 3430 }, { "grad_norm": 1.167182207107544, "learning_rate": 9.760871560273197e-05, "loss": 0.1696, "step": 3440 }, { "grad_norm": 1.7801024913787842, "learning_rate": 9.755793271724526e-05, "loss": 0.1921, "step": 3450 }, { "grad_norm": 3.9849090576171875, "learning_rate": 9.750662974582584e-05, "loss": 0.2706, "step": 3460 }, { "grad_norm": 3.4585747718811035, "learning_rate": 9.745480724951473e-05, "loss": 0.2691, "step": 3470 }, { "grad_norm": 3.3321328163146973, "learning_rate": 9.740246579503447e-05, "loss": 0.2822, "step": 3480 }, { "grad_norm": 3.190204620361328, "learning_rate": 9.734960595478284e-05, "loss": 0.1804, "step": 3490 }, { "grad_norm": 2.083669424057007, "learning_rate": 9.729622830682657e-05, "loss": 0.2847, "step": 3500 }, { "grad_norm": 0.860984206199646, "learning_rate": 9.724233343489504e-05, "loss": 0.15, "step": 3510 }, { "grad_norm": 1.489528775215149, "learning_rate": 9.718792192837396e-05, "loss": 0.1789, "step": 3520 }, { "grad_norm": 1.998658299446106, "learning_rate": 9.713299438229886e-05, "loss": 0.2508, "step": 3530 }, { "grad_norm": 1.8065434694290161, "learning_rate": 9.707755139734855e-05, "loss": 0.2407, "step": 3540 }, { "grad_norm": 1.2814288139343262, "learning_rate": 9.702159357983866e-05, "loss": 0.2188, "step": 3550 }, { "grad_norm": 1.5360549688339233, "learning_rate": 9.696512154171492e-05, "loss": 0.1372, "step": 3560 }, { "grad_norm": 2.475228786468506, "learning_rate": 9.690813590054645e-05, "loss": 0.1378, "step": 3570 }, { "grad_norm": 3.5576469898223877, "learning_rate": 9.685063727951914e-05, "loss": 0.1509, "step": 3580 }, { "grad_norm": 1.4074150323867798, "learning_rate": 9.679262630742865e-05, "loss": 0.2013, "step": 3590 }, { "grad_norm": 1.9083681106567383, "learning_rate": 9.673410361867373e-05, "loss": 0.2901, "step": 3600 }, { "grad_norm": 0.9546309113502502, "learning_rate": 9.667506985324909e-05, "loss": 0.1309, "step": 3610 }, { "grad_norm": 1.7428703308105469, "learning_rate": 9.661552565673855e-05, "loss": 0.2489, "step": 3620 }, { "grad_norm": 3.2678346633911133, "learning_rate": 9.655547168030789e-05, "loss": 0.1849, "step": 3630 }, { "grad_norm": 1.6866271495819092, "learning_rate": 9.649490858069777e-05, "loss": 0.1729, "step": 3640 }, { "grad_norm": 3.062800884246826, "learning_rate": 9.643383702021658e-05, "loss": 0.1385, "step": 3650 }, { "grad_norm": 2.189366340637207, "learning_rate": 9.637225766673307e-05, "loss": 0.2057, "step": 3660 }, { "grad_norm": 1.5299835205078125, "learning_rate": 9.631017119366922e-05, "loss": 0.1563, "step": 3670 }, { "grad_norm": 2.367997169494629, "learning_rate": 9.624757827999273e-05, "loss": 0.1771, "step": 3680 }, { "grad_norm": 1.1537022590637207, "learning_rate": 9.618447961020971e-05, "loss": 0.2741, "step": 3690 }, { "grad_norm": 0.8300281167030334, "learning_rate": 9.612087587435707e-05, "loss": 0.1641, "step": 3700 }, { "grad_norm": 1.2957969903945923, "learning_rate": 9.605676776799508e-05, "loss": 0.1577, "step": 3710 }, { "grad_norm": 2.287971019744873, "learning_rate": 9.599215599219973e-05, "loss": 0.1715, "step": 3720 }, { "grad_norm": 1.1727265119552612, "learning_rate": 9.592704125355505e-05, "loss": 0.2075, "step": 3730 }, { "grad_norm": 1.411626935005188, "learning_rate": 9.586142426414538e-05, "loss": 0.1942, "step": 3740 }, { "grad_norm": 1.3716391324996948, "learning_rate": 9.57953057415476e-05, "loss": 0.1586, "step": 3750 }, { "grad_norm": 2.9568865299224854, "learning_rate": 9.572868640882328e-05, "loss": 0.22, "step": 3760 }, { "grad_norm": 2.1684067249298096, "learning_rate": 9.56615669945108e-05, "loss": 0.1314, "step": 3770 }, { "grad_norm": 1.519336462020874, "learning_rate": 9.55939482326173e-05, "loss": 0.1095, "step": 3780 }, { "grad_norm": 2.5453670024871826, "learning_rate": 9.552583086261069e-05, "loss": 0.153, "step": 3790 }, { "grad_norm": 1.6337180137634277, "learning_rate": 9.545721562941168e-05, "loss": 0.1433, "step": 3800 }, { "grad_norm": 1.3899997472763062, "learning_rate": 9.538810328338543e-05, "loss": 0.1037, "step": 3810 }, { "grad_norm": 1.3656458854675293, "learning_rate": 9.531849458033349e-05, "loss": 0.1073, "step": 3820 }, { "grad_norm": 5.081826686859131, "learning_rate": 9.524839028148547e-05, "loss": 0.3267, "step": 3830 }, { "grad_norm": 2.45864200592041, "learning_rate": 9.517779115349077e-05, "loss": 0.2516, "step": 3840 }, { "grad_norm": 1.7332441806793213, "learning_rate": 9.510669796841014e-05, "loss": 0.2108, "step": 3850 }, { "grad_norm": 4.1433916091918945, "learning_rate": 9.503511150370727e-05, "loss": 0.1655, "step": 3860 }, { "grad_norm": 2.413032293319702, "learning_rate": 9.496303254224024e-05, "loss": 0.1944, "step": 3870 }, { "grad_norm": 2.159942865371704, "learning_rate": 9.489046187225306e-05, "loss": 0.156, "step": 3880 }, { "grad_norm": 2.1075193881988525, "learning_rate": 9.481740028736692e-05, "loss": 0.2276, "step": 3890 }, { "grad_norm": 2.1793174743652344, "learning_rate": 9.474384858657164e-05, "loss": 0.1488, "step": 3900 }, { "grad_norm": 2.2949423789978027, "learning_rate": 9.466980757421679e-05, "loss": 0.1916, "step": 3910 }, { "grad_norm": 2.0820324420928955, "learning_rate": 9.459527806000305e-05, "loss": 0.1801, "step": 3920 }, { "grad_norm": 1.0559215545654297, "learning_rate": 9.452026085897325e-05, "loss": 0.2034, "step": 3930 }, { "grad_norm": 1.5029317140579224, "learning_rate": 9.444475679150348e-05, "loss": 0.2154, "step": 3940 }, { "grad_norm": 1.597171664237976, "learning_rate": 9.436876668329411e-05, "loss": 0.1528, "step": 3950 }, { "grad_norm": 1.3960667848587036, "learning_rate": 9.429229136536079e-05, "loss": 0.1398, "step": 3960 }, { "grad_norm": 0.9422345161437988, "learning_rate": 9.421533167402534e-05, "loss": 0.1827, "step": 3970 }, { "grad_norm": 1.683300256729126, "learning_rate": 9.413788845090666e-05, "loss": 0.1985, "step": 3980 }, { "grad_norm": 1.238467812538147, "learning_rate": 9.405996254291136e-05, "loss": 0.1353, "step": 3990 }, { "grad_norm": 3.113614320755005, "learning_rate": 9.398155480222474e-05, "loss": 0.1824, "step": 4000 }, { "grad_norm": 1.3014827966690063, "learning_rate": 9.390266608630128e-05, "loss": 0.1753, "step": 4010 }, { "grad_norm": 2.520200729370117, "learning_rate": 9.38232972578553e-05, "loss": 0.1951, "step": 4020 }, { "grad_norm": 3.058201789855957, "learning_rate": 9.374344918485164e-05, "loss": 0.1919, "step": 4030 }, { "grad_norm": 1.2199641466140747, "learning_rate": 9.366312274049602e-05, "loss": 0.1577, "step": 4040 }, { "grad_norm": 2.899874448776245, "learning_rate": 9.358231880322554e-05, "loss": 0.23, "step": 4050 }, { "grad_norm": 1.381322979927063, "learning_rate": 9.350103825669916e-05, "loss": 0.189, "step": 4060 }, { "grad_norm": 1.7286477088928223, "learning_rate": 9.341928198978787e-05, "loss": 0.1529, "step": 4070 }, { "grad_norm": 1.7477315664291382, "learning_rate": 9.333705089656512e-05, "loss": 0.26, "step": 4080 }, { "grad_norm": 1.1755719184875488, "learning_rate": 9.325434587629698e-05, "loss": 0.1443, "step": 4090 }, { "grad_norm": 1.1644933223724365, "learning_rate": 9.31711678334323e-05, "loss": 0.2175, "step": 4100 }, { "grad_norm": 1.3457353115081787, "learning_rate": 9.308751767759282e-05, "loss": 0.1504, "step": 4110 }, { "grad_norm": 2.00907564163208, "learning_rate": 9.300339632356325e-05, "loss": 0.1374, "step": 4120 }, { "grad_norm": 1.8459323644638062, "learning_rate": 9.291880469128124e-05, "loss": 0.2357, "step": 4130 }, { "grad_norm": 1.5956283807754517, "learning_rate": 9.283374370582732e-05, "loss": 0.1911, "step": 4140 }, { "grad_norm": 1.1371138095855713, "learning_rate": 9.274821429741482e-05, "loss": 0.0719, "step": 4150 }, { "grad_norm": 1.5914876461029053, "learning_rate": 9.266221740137961e-05, "loss": 0.0894, "step": 4160 }, { "grad_norm": 1.9739874601364136, "learning_rate": 9.257575395817001e-05, "loss": 0.1346, "step": 4170 }, { "grad_norm": 0.9959692358970642, "learning_rate": 9.248882491333637e-05, "loss": 0.2107, "step": 4180 }, { "grad_norm": 1.5838148593902588, "learning_rate": 9.240143121752076e-05, "loss": 0.1955, "step": 4190 }, { "grad_norm": 1.8215276002883911, "learning_rate": 9.23135738264467e-05, "loss": 0.1374, "step": 4200 }, { "grad_norm": 1.951718807220459, "learning_rate": 9.222525370090849e-05, "loss": 0.1839, "step": 4210 }, { "grad_norm": 1.5761936902999878, "learning_rate": 9.213647180676088e-05, "loss": 0.2238, "step": 4220 }, { "grad_norm": 1.5812604427337646, "learning_rate": 9.204722911490846e-05, "loss": 0.1799, "step": 4230 }, { "grad_norm": 1.2741230726242065, "learning_rate": 9.1957526601295e-05, "loss": 0.1426, "step": 4240 }, { "grad_norm": 1.5742825269699097, "learning_rate": 9.186736524689281e-05, "loss": 0.1383, "step": 4250 }, { "grad_norm": 1.7368192672729492, "learning_rate": 9.177674603769204e-05, "loss": 0.1945, "step": 4260 }, { "grad_norm": 2.617191791534424, "learning_rate": 9.168566996468983e-05, "loss": 0.1037, "step": 4270 }, { "grad_norm": 2.032783031463623, "learning_rate": 9.159413802387951e-05, "loss": 0.1538, "step": 4280 }, { "grad_norm": 1.2998415231704712, "learning_rate": 9.150215121623974e-05, "loss": 0.1614, "step": 4290 }, { "grad_norm": 1.4626550674438477, "learning_rate": 9.140971054772349e-05, "loss": 0.0895, "step": 4300 }, { "grad_norm": 1.3064746856689453, "learning_rate": 9.131681702924713e-05, "loss": 0.1052, "step": 4310 }, { "grad_norm": 0.9010644555091858, "learning_rate": 9.122347167667926e-05, "loss": 0.131, "step": 4320 }, { "grad_norm": 2.408940076828003, "learning_rate": 9.112967551082973e-05, "loss": 0.1674, "step": 4330 }, { "grad_norm": 2.5002799034118652, "learning_rate": 9.103542955743835e-05, "loss": 0.1805, "step": 4340 }, { "grad_norm": 2.0549020767211914, "learning_rate": 9.094073484716381e-05, "loss": 0.2185, "step": 4350 }, { "grad_norm": 0.9140834212303162, "learning_rate": 9.084559241557226e-05, "loss": 0.1315, "step": 4360 }, { "grad_norm": 1.1889265775680542, "learning_rate": 9.075000330312608e-05, "loss": 0.1124, "step": 4370 }, { "grad_norm": 1.2359732389450073, "learning_rate": 9.065396855517253e-05, "loss": 0.2214, "step": 4380 }, { "grad_norm": 1.1520806550979614, "learning_rate": 9.055748922193219e-05, "loss": 0.2275, "step": 4390 }, { "grad_norm": 1.4618325233459473, "learning_rate": 9.046056635848761e-05, "loss": 0.1493, "step": 4400 }, { "grad_norm": 1.588210940361023, "learning_rate": 9.036320102477169e-05, "loss": 0.171, "step": 4410 }, { "grad_norm": 2.903379201889038, "learning_rate": 9.02653942855561e-05, "loss": 0.212, "step": 4420 }, { "grad_norm": 1.6613436937332153, "learning_rate": 9.016714721043971e-05, "loss": 0.1523, "step": 4430 }, { "grad_norm": 1.2434238195419312, "learning_rate": 9.006846087383675e-05, "loss": 0.1597, "step": 4440 }, { "grad_norm": 1.5199185609817505, "learning_rate": 8.996933635496523e-05, "loss": 0.1914, "step": 4450 }, { "grad_norm": 3.4354777336120605, "learning_rate": 8.986977473783498e-05, "loss": 0.2112, "step": 4460 }, { "grad_norm": 2.2233710289001465, "learning_rate": 8.97697771112359e-05, "loss": 0.1706, "step": 4470 }, { "grad_norm": 1.7762634754180908, "learning_rate": 8.966934456872602e-05, "loss": 0.1799, "step": 4480 }, { "grad_norm": 1.1585371494293213, "learning_rate": 8.95684782086195e-05, "loss": 0.1514, "step": 4490 }, { "grad_norm": 2.3345112800598145, "learning_rate": 8.946717913397476e-05, "loss": 0.1147, "step": 4500 }, { "grad_norm": 1.2702770233154297, "learning_rate": 8.93654484525822e-05, "loss": 0.171, "step": 4510 }, { "grad_norm": 1.9842958450317383, "learning_rate": 8.926328727695226e-05, "loss": 0.1648, "step": 4520 }, { "grad_norm": 1.8156622648239136, "learning_rate": 8.916069672430319e-05, "loss": 0.1661, "step": 4530 }, { "grad_norm": 1.2517907619476318, "learning_rate": 8.905767791654884e-05, "loss": 0.1163, "step": 4540 }, { "grad_norm": 1.5316753387451172, "learning_rate": 8.895423198028638e-05, "loss": 0.114, "step": 4550 }, { "grad_norm": 1.981549620628357, "learning_rate": 8.885036004678402e-05, "loss": 0.1789, "step": 4560 }, { "grad_norm": 1.0280131101608276, "learning_rate": 8.874606325196857e-05, "loss": 0.0933, "step": 4570 }, { "grad_norm": 3.3848249912261963, "learning_rate": 8.864134273641304e-05, "loss": 0.1449, "step": 4580 }, { "grad_norm": 2.6775574684143066, "learning_rate": 8.853619964532427e-05, "loss": 0.1536, "step": 4590 }, { "grad_norm": 2.8554322719573975, "learning_rate": 8.843063512853019e-05, "loss": 0.1913, "step": 4600 }, { "grad_norm": 3.0597846508026123, "learning_rate": 8.832465034046749e-05, "loss": 0.1793, "step": 4610 }, { "grad_norm": 0.8963209390640259, "learning_rate": 8.821824644016882e-05, "loss": 0.1086, "step": 4620 }, { "grad_norm": 1.491929531097412, "learning_rate": 8.811142459125019e-05, "loss": 0.2357, "step": 4630 }, { "grad_norm": 0.696634829044342, "learning_rate": 8.800418596189822e-05, "loss": 0.1742, "step": 4640 }, { "grad_norm": 1.9884861707687378, "learning_rate": 8.789653172485737e-05, "loss": 0.1371, "step": 4650 }, { "grad_norm": 3.5188920497894287, "learning_rate": 8.778846305741715e-05, "loss": 0.1306, "step": 4660 }, { "grad_norm": 1.8176524639129639, "learning_rate": 8.767998114139918e-05, "loss": 0.1469, "step": 4670 }, { "grad_norm": 1.0082200765609741, "learning_rate": 8.757108716314429e-05, "loss": 0.098, "step": 4680 }, { "grad_norm": 5.223330974578857, "learning_rate": 8.746178231349962e-05, "loss": 0.1951, "step": 4690 }, { "grad_norm": 1.3476448059082031, "learning_rate": 8.735206778780549e-05, "loss": 0.1712, "step": 4700 }, { "grad_norm": 1.7969763278961182, "learning_rate": 8.724194478588234e-05, "loss": 0.237, "step": 4710 }, { "grad_norm": 3.4075326919555664, "learning_rate": 8.713141451201772e-05, "loss": 0.2422, "step": 4720 }, { "grad_norm": 2.7447047233581543, "learning_rate": 8.702047817495295e-05, "loss": 0.1864, "step": 4730 }, { "grad_norm": 1.1376560926437378, "learning_rate": 8.69091369878701e-05, "loss": 0.2151, "step": 4740 }, { "grad_norm": 1.358223557472229, "learning_rate": 8.679739216837849e-05, "loss": 0.3183, "step": 4750 }, { "grad_norm": 2.1688196659088135, "learning_rate": 8.66852449385016e-05, "loss": 0.1471, "step": 4760 }, { "grad_norm": 1.8279099464416504, "learning_rate": 8.657269652466356e-05, "loss": 0.2074, "step": 4770 }, { "grad_norm": 2.12509822845459, "learning_rate": 8.645974815767577e-05, "loss": 0.2105, "step": 4780 }, { "grad_norm": 2.7584822177886963, "learning_rate": 8.634640107272351e-05, "loss": 0.3359, "step": 4790 }, { "grad_norm": 2.5486831665039062, "learning_rate": 8.623265650935234e-05, "loss": 0.207, "step": 4800 }, { "grad_norm": 1.5694423913955688, "learning_rate": 8.611851571145456e-05, "loss": 0.1658, "step": 4810 }, { "grad_norm": 1.616253137588501, "learning_rate": 8.600397992725566e-05, "loss": 0.1183, "step": 4820 }, { "grad_norm": 1.9806562662124634, "learning_rate": 8.588905040930061e-05, "loss": 0.1543, "step": 4830 }, { "grad_norm": 1.443027377128601, "learning_rate": 8.577372841444022e-05, "loss": 0.171, "step": 4840 }, { "grad_norm": 1.458254098892212, "learning_rate": 8.565801520381736e-05, "loss": 0.2183, "step": 4850 }, { "grad_norm": 2.4779763221740723, "learning_rate": 8.554191204285313e-05, "loss": 0.1742, "step": 4860 }, { "grad_norm": 1.5002259016036987, "learning_rate": 8.542542020123315e-05, "loss": 0.1941, "step": 4870 }, { "grad_norm": 3.8247122764587402, "learning_rate": 8.530854095289347e-05, "loss": 0.2054, "step": 4880 }, { "grad_norm": 0.6758179068565369, "learning_rate": 8.519127557600688e-05, "loss": 0.1675, "step": 4890 }, { "grad_norm": 1.6149580478668213, "learning_rate": 8.507362535296871e-05, "loss": 0.1607, "step": 4900 }, { "grad_norm": 2.967733383178711, "learning_rate": 8.495559157038299e-05, "loss": 0.2527, "step": 4910 }, { "grad_norm": 2.3624613285064697, "learning_rate": 8.483717551904823e-05, "loss": 0.1621, "step": 4920 }, { "grad_norm": 1.3246668577194214, "learning_rate": 8.47183784939434e-05, "loss": 0.1107, "step": 4930 }, { "grad_norm": 2.3354532718658447, "learning_rate": 8.459920179421374e-05, "loss": 0.1309, "step": 4940 }, { "grad_norm": 1.4452805519104004, "learning_rate": 8.447964672315656e-05, "loss": 0.1564, "step": 4950 }, { "grad_norm": 1.6815961599349976, "learning_rate": 8.435971458820692e-05, "loss": 0.1355, "step": 4960 }, { "grad_norm": 1.1882200241088867, "learning_rate": 8.423940670092345e-05, "loss": 0.1555, "step": 4970 }, { "grad_norm": 0.9890279769897461, "learning_rate": 8.411872437697394e-05, "loss": 0.0544, "step": 4980 }, { "grad_norm": 1.198477864265442, "learning_rate": 8.399766893612096e-05, "loss": 0.0995, "step": 4990 }, { "grad_norm": 1.0018227100372314, "learning_rate": 8.38762417022074e-05, "loss": 0.0416, "step": 5000 }, { "grad_norm": 1.7873529195785522, "learning_rate": 8.375444400314204e-05, "loss": 0.0578, "step": 5010 }, { "grad_norm": 0.9942506551742554, "learning_rate": 8.3632277170885e-05, "loss": 0.0479, "step": 5020 }, { "grad_norm": 1.0294082164764404, "learning_rate": 8.350974254143318e-05, "loss": 0.0842, "step": 5030 }, { "grad_norm": 0.9019378423690796, "learning_rate": 8.338684145480566e-05, "loss": 0.0303, "step": 5040 }, { "grad_norm": 1.185564637184143, "learning_rate": 8.326357525502904e-05, "loss": 0.0239, "step": 5050 }, { "grad_norm": 0.7096260190010071, "learning_rate": 8.313994529012273e-05, "loss": 0.0352, "step": 5060 }, { "grad_norm": 0.6269180774688721, "learning_rate": 8.301595291208422e-05, "loss": 0.0719, "step": 5070 }, { "grad_norm": 1.025739312171936, "learning_rate": 8.289159947687427e-05, "loss": 0.0353, "step": 5080 }, { "grad_norm": 0.7228735685348511, "learning_rate": 8.276688634440216e-05, "loss": 0.064, "step": 5090 }, { "grad_norm": 0.9021449685096741, "learning_rate": 8.26418148785107e-05, "loss": 0.0914, "step": 5100 }, { "grad_norm": 0.6129550337791443, "learning_rate": 8.251638644696141e-05, "loss": 0.0408, "step": 5110 }, { "grad_norm": 0.7158212661743164, "learning_rate": 8.23906024214195e-05, "loss": 0.0963, "step": 5120 }, { "grad_norm": 0.6214717030525208, "learning_rate": 8.226446417743897e-05, "loss": 0.0235, "step": 5130 }, { "grad_norm": 0.676845133304596, "learning_rate": 8.213797309444742e-05, "loss": 0.0603, "step": 5140 }, { "grad_norm": 0.6965628862380981, "learning_rate": 8.201113055573105e-05, "loss": 0.0609, "step": 5150 }, { "grad_norm": 0.6506880521774292, "learning_rate": 8.188393794841958e-05, "loss": 0.0386, "step": 5160 }, { "grad_norm": 0.6985394954681396, "learning_rate": 8.175639666347094e-05, "loss": 0.033, "step": 5170 }, { "grad_norm": 1.1383644342422485, "learning_rate": 8.162850809565623e-05, "loss": 0.0349, "step": 5180 }, { "grad_norm": 0.7668223977088928, "learning_rate": 8.150027364354431e-05, "loss": 0.0267, "step": 5190 }, { "grad_norm": 0.792066216468811, "learning_rate": 8.137169470948662e-05, "loss": 0.0346, "step": 5200 }, { "grad_norm": 0.5683591365814209, "learning_rate": 8.124277269960179e-05, "loss": 0.0518, "step": 5210 }, { "grad_norm": 0.8433091044425964, "learning_rate": 8.111350902376023e-05, "loss": 0.0268, "step": 5220 }, { "grad_norm": 0.9513429403305054, "learning_rate": 8.098390509556883e-05, "loss": 0.0365, "step": 5230 }, { "grad_norm": 1.4104185104370117, "learning_rate": 8.085396233235536e-05, "loss": 0.0334, "step": 5240 }, { "grad_norm": 0.8286669850349426, "learning_rate": 8.072368215515306e-05, "loss": 0.0381, "step": 5250 }, { "grad_norm": 0.5714389085769653, "learning_rate": 8.059306598868506e-05, "loss": 0.0215, "step": 5260 }, { "grad_norm": 0.800783634185791, "learning_rate": 8.046211526134888e-05, "loss": 0.0349, "step": 5270 }, { "grad_norm": 0.8123432993888855, "learning_rate": 8.033083140520065e-05, "loss": 0.0588, "step": 5280 }, { "grad_norm": 1.6906486749649048, "learning_rate": 8.019921585593962e-05, "loss": 0.0364, "step": 5290 }, { "grad_norm": 1.1177639961242676, "learning_rate": 8.006727005289232e-05, "loss": 0.1142, "step": 5300 }, { "grad_norm": 1.6728641986846924, "learning_rate": 7.993499543899692e-05, "loss": 0.0417, "step": 5310 }, { "grad_norm": 0.7386102676391602, "learning_rate": 7.980239346078742e-05, "loss": 0.0281, "step": 5320 }, { "grad_norm": 0.44531866908073425, "learning_rate": 7.966946556837778e-05, "loss": 0.0324, "step": 5330 }, { "grad_norm": 0.5838891863822937, "learning_rate": 7.953621321544616e-05, "loss": 0.0273, "step": 5340 }, { "grad_norm": 0.7053329348564148, "learning_rate": 7.940263785921896e-05, "loss": 0.0529, "step": 5350 }, { "grad_norm": 0.7029603719711304, "learning_rate": 7.926874096045482e-05, "loss": 0.055, "step": 5360 }, { "grad_norm": 0.5235321521759033, "learning_rate": 7.913452398342881e-05, "loss": 0.046, "step": 5370 }, { "grad_norm": 0.5353266596794128, "learning_rate": 7.89999883959163e-05, "loss": 0.0249, "step": 5380 }, { "grad_norm": 0.6488685607910156, "learning_rate": 7.886513566917687e-05, "loss": 0.0263, "step": 5390 }, { "grad_norm": 0.6330428123474121, "learning_rate": 7.872996727793838e-05, "loss": 0.0667, "step": 5400 }, { "grad_norm": 0.7988374829292297, "learning_rate": 7.859448470038069e-05, "loss": 0.0234, "step": 5410 }, { "grad_norm": 0.6608654260635376, "learning_rate": 7.845868941811956e-05, "loss": 0.0184, "step": 5420 }, { "grad_norm": 0.8554064035415649, "learning_rate": 7.832258291619043e-05, "loss": 0.0391, "step": 5430 }, { "grad_norm": 0.7886276841163635, "learning_rate": 7.81861666830322e-05, "loss": 0.0277, "step": 5440 }, { "grad_norm": 0.6548476815223694, "learning_rate": 7.804944221047097e-05, "loss": 0.0768, "step": 5450 }, { "grad_norm": 0.9442992806434631, "learning_rate": 7.791241099370364e-05, "loss": 0.0291, "step": 5460 }, { "grad_norm": 0.6193794012069702, "learning_rate": 7.777507453128163e-05, "loss": 0.0479, "step": 5470 }, { "grad_norm": 1.7362347841262817, "learning_rate": 7.763743432509451e-05, "loss": 0.0403, "step": 5480 }, { "grad_norm": 0.9698314070701599, "learning_rate": 7.749949188035353e-05, "loss": 0.0261, "step": 5490 }, { "grad_norm": 0.6848409175872803, "learning_rate": 7.736124870557516e-05, "loss": 0.0255, "step": 5500 }, { "grad_norm": 1.1129796504974365, "learning_rate": 7.722270631256459e-05, "loss": 0.0275, "step": 5510 }, { "grad_norm": 0.57719886302948, "learning_rate": 7.708386621639925e-05, "loss": 0.0313, "step": 5520 }, { "grad_norm": 0.7146404981613159, "learning_rate": 7.694472993541219e-05, "loss": 0.039, "step": 5530 }, { "grad_norm": 1.8802779912948608, "learning_rate": 7.680529899117547e-05, "loss": 0.0436, "step": 5540 }, { "grad_norm": 0.6772153377532959, "learning_rate": 7.666557490848358e-05, "loss": 0.0334, "step": 5550 }, { "grad_norm": 0.9064921736717224, "learning_rate": 7.65255592153367e-05, "loss": 0.0229, "step": 5560 }, { "grad_norm": 0.8220618963241577, "learning_rate": 7.638525344292402e-05, "loss": 0.0263, "step": 5570 }, { "grad_norm": 0.7669905424118042, "learning_rate": 7.624465912560697e-05, "loss": 0.0244, "step": 5580 }, { "grad_norm": 0.6566009521484375, "learning_rate": 7.610377780090249e-05, "loss": 0.031, "step": 5590 }, { "grad_norm": 0.9565125703811646, "learning_rate": 7.596261100946618e-05, "loss": 0.0437, "step": 5600 }, { "grad_norm": 0.5573037266731262, "learning_rate": 7.582116029507542e-05, "loss": 0.0203, "step": 5610 }, { "grad_norm": 0.463133841753006, "learning_rate": 7.56794272046126e-05, "loss": 0.033, "step": 5620 }, { "grad_norm": 1.1883325576782227, "learning_rate": 7.55374132880481e-05, "loss": 0.0266, "step": 5630 }, { "grad_norm": 0.6277577877044678, "learning_rate": 7.539512009842333e-05, "loss": 0.0399, "step": 5640 }, { "grad_norm": 0.6208600401878357, "learning_rate": 7.525254919183382e-05, "loss": 0.0201, "step": 5650 }, { "grad_norm": 0.8525328636169434, "learning_rate": 7.510970212741215e-05, "loss": 0.0215, "step": 5660 }, { "grad_norm": 0.5788384079933167, "learning_rate": 7.496658046731096e-05, "loss": 0.0207, "step": 5670 }, { "grad_norm": 0.8163131475448608, "learning_rate": 7.482318577668578e-05, "loss": 0.0185, "step": 5680 }, { "grad_norm": 0.6973435282707214, "learning_rate": 7.467951962367796e-05, "loss": 0.0222, "step": 5690 }, { "grad_norm": 0.7446568608283997, "learning_rate": 7.453558357939755e-05, "loss": 0.022, "step": 5700 }, { "grad_norm": 0.42061471939086914, "learning_rate": 7.439137921790606e-05, "loss": 0.0222, "step": 5710 }, { "grad_norm": 0.5957080721855164, "learning_rate": 7.42469081161993e-05, "loss": 0.0578, "step": 5720 }, { "grad_norm": 0.7819215059280396, "learning_rate": 7.410217185419006e-05, "loss": 0.0288, "step": 5730 }, { "grad_norm": 0.8777925968170166, "learning_rate": 7.395717201469095e-05, "loss": 0.072, "step": 5740 }, { "grad_norm": 0.465981662273407, "learning_rate": 7.381191018339696e-05, "loss": 0.016, "step": 5750 }, { "grad_norm": 0.4687363803386688, "learning_rate": 7.36663879488682e-05, "loss": 0.0383, "step": 5760 }, { "grad_norm": 0.707307755947113, "learning_rate": 7.352060690251254e-05, "loss": 0.0148, "step": 5770 }, { "grad_norm": 1.1169835329055786, "learning_rate": 7.337456863856811e-05, "loss": 0.0417, "step": 5780 }, { "grad_norm": 0.823988676071167, "learning_rate": 7.3228274754086e-05, "loss": 0.018, "step": 5790 }, { "grad_norm": 0.5474166870117188, "learning_rate": 7.308172684891267e-05, "loss": 0.0383, "step": 5800 }, { "grad_norm": 1.082371473312378, "learning_rate": 7.293492652567255e-05, "loss": 0.05, "step": 5810 }, { "grad_norm": 0.5915845632553101, "learning_rate": 7.278787538975043e-05, "loss": 0.0213, "step": 5820 }, { "grad_norm": 0.3640531897544861, "learning_rate": 7.2640575049274e-05, "loss": 0.0159, "step": 5830 }, { "grad_norm": 0.7117072343826294, "learning_rate": 7.249302711509616e-05, "loss": 0.0166, "step": 5840 }, { "grad_norm": 1.9897350072860718, "learning_rate": 7.23452332007775e-05, "loss": 0.0372, "step": 5850 }, { "grad_norm": 2.9066784381866455, "learning_rate": 7.219719492256858e-05, "loss": 0.3826, "step": 5860 }, { "grad_norm": 1.62128746509552, "learning_rate": 7.20489138993923e-05, "loss": 0.2557, "step": 5870 }, { "grad_norm": 2.0654478073120117, "learning_rate": 7.190039175282614e-05, "loss": 0.1682, "step": 5880 }, { "grad_norm": 1.4620529413223267, "learning_rate": 7.175163010708455e-05, "loss": 0.164, "step": 5890 }, { "grad_norm": 1.4518001079559326, "learning_rate": 7.1602630589001e-05, "loss": 0.1137, "step": 5900 }, { "grad_norm": 1.3200379610061646, "learning_rate": 7.14533948280104e-05, "loss": 0.1151, "step": 5910 }, { "grad_norm": 1.626326560974121, "learning_rate": 7.130392445613109e-05, "loss": 0.1817, "step": 5920 }, { "grad_norm": 1.6537237167358398, "learning_rate": 7.115422110794711e-05, "loss": 0.1734, "step": 5930 }, { "grad_norm": 1.3261401653289795, "learning_rate": 7.100428642059033e-05, "loss": 0.1117, "step": 5940 }, { "grad_norm": 1.1704758405685425, "learning_rate": 7.08541220337224e-05, "loss": 0.0875, "step": 5950 }, { "grad_norm": 1.2572885751724243, "learning_rate": 7.070372958951706e-05, "loss": 0.1233, "step": 5960 }, { "grad_norm": 1.823311448097229, "learning_rate": 7.055311073264194e-05, "loss": 0.155, "step": 5970 }, { "grad_norm": 2.106067419052124, "learning_rate": 7.040226711024077e-05, "loss": 0.1339, "step": 5980 }, { "grad_norm": 2.1002001762390137, "learning_rate": 7.02512003719152e-05, "loss": 0.1027, "step": 5990 }, { "grad_norm": 1.3731050491333008, "learning_rate": 7.00999121697069e-05, "loss": 0.1848, "step": 6000 }, { "grad_norm": 2.090085744857788, "learning_rate": 6.99484041580794e-05, "loss": 0.1288, "step": 6010 }, { "grad_norm": 0.8364267349243164, "learning_rate": 6.979667799390004e-05, "loss": 0.1281, "step": 6020 }, { "grad_norm": 2.129692316055298, "learning_rate": 6.964473533642185e-05, "loss": 0.1749, "step": 6030 }, { "grad_norm": 1.1232068538665771, "learning_rate": 6.949257784726539e-05, "loss": 0.135, "step": 6040 }, { "grad_norm": 2.066246271133423, "learning_rate": 6.934020719040056e-05, "loss": 0.1318, "step": 6050 }, { "grad_norm": 0.9737508893013, "learning_rate": 6.918762503212848e-05, "loss": 0.2264, "step": 6060 }, { "grad_norm": 1.9637460708618164, "learning_rate": 6.903483304106319e-05, "loss": 0.117, "step": 6070 }, { "grad_norm": 1.2745380401611328, "learning_rate": 6.888183288811341e-05, "loss": 0.3141, "step": 6080 }, { "grad_norm": 3.584176540374756, "learning_rate": 6.87286262464643e-05, "loss": 0.203, "step": 6090 }, { "grad_norm": 4.847754001617432, "learning_rate": 6.857521479155915e-05, "loss": 0.1631, "step": 6100 }, { "grad_norm": 1.581041693687439, "learning_rate": 6.842160020108104e-05, "loss": 0.1281, "step": 6110 }, { "grad_norm": 0.6225283145904541, "learning_rate": 6.826778415493455e-05, "loss": 0.1563, "step": 6120 }, { "grad_norm": 1.0778053998947144, "learning_rate": 6.811376833522729e-05, "loss": 0.2074, "step": 6130 }, { "grad_norm": 1.758590817451477, "learning_rate": 6.795955442625159e-05, "loss": 0.1849, "step": 6140 }, { "grad_norm": 2.41597843170166, "learning_rate": 6.780514411446608e-05, "loss": 0.2211, "step": 6150 }, { "grad_norm": 2.225973129272461, "learning_rate": 6.765053908847716e-05, "loss": 0.1277, "step": 6160 }, { "grad_norm": 1.411848545074463, "learning_rate": 6.749574103902064e-05, "loss": 0.153, "step": 6170 }, { "grad_norm": 2.648810863494873, "learning_rate": 6.734075165894317e-05, "loss": 0.1681, "step": 6180 }, { "grad_norm": 1.293265461921692, "learning_rate": 6.71855726431838e-05, "loss": 0.1422, "step": 6190 }, { "grad_norm": 1.2664004564285278, "learning_rate": 6.703020568875538e-05, "loss": 0.1772, "step": 6200 }, { "grad_norm": 1.3341739177703857, "learning_rate": 6.687465249472603e-05, "loss": 0.1893, "step": 6210 }, { "grad_norm": 3.7255783081054688, "learning_rate": 6.671891476220055e-05, "loss": 0.2842, "step": 6220 }, { "grad_norm": 1.3223974704742432, "learning_rate": 6.656299419430183e-05, "loss": 0.2324, "step": 6230 }, { "grad_norm": 1.2601040601730347, "learning_rate": 6.640689249615223e-05, "loss": 0.1283, "step": 6240 }, { "grad_norm": 2.052736759185791, "learning_rate": 6.625061137485491e-05, "loss": 0.2001, "step": 6250 }, { "grad_norm": 1.5382658243179321, "learning_rate": 6.609415253947517e-05, "loss": 0.14, "step": 6260 }, { "grad_norm": 1.3842768669128418, "learning_rate": 6.593751770102178e-05, "loss": 0.1686, "step": 6270 }, { "grad_norm": 1.4178731441497803, "learning_rate": 6.578070857242823e-05, "loss": 0.1831, "step": 6280 }, { "grad_norm": 1.955704927444458, "learning_rate": 6.562372686853402e-05, "loss": 0.2035, "step": 6290 }, { "grad_norm": 1.0092883110046387, "learning_rate": 6.546657430606593e-05, "loss": 0.1499, "step": 6300 }, { "grad_norm": 1.2690070867538452, "learning_rate": 6.530925260361918e-05, "loss": 0.1493, "step": 6310 }, { "grad_norm": 2.2860379219055176, "learning_rate": 6.515176348163871e-05, "loss": 0.1957, "step": 6320 }, { "grad_norm": 1.7515034675598145, "learning_rate": 6.499410866240032e-05, "loss": 0.1536, "step": 6330 }, { "grad_norm": 1.0313317775726318, "learning_rate": 6.48362898699919e-05, "loss": 0.1665, "step": 6340 }, { "grad_norm": 1.2928482294082642, "learning_rate": 6.467830883029443e-05, "loss": 0.1854, "step": 6350 }, { "grad_norm": 2.61213755607605, "learning_rate": 6.452016727096326e-05, "loss": 0.1427, "step": 6360 }, { "grad_norm": 1.513239860534668, "learning_rate": 6.436186692140916e-05, "loss": 0.133, "step": 6370 }, { "grad_norm": 1.7199549674987793, "learning_rate": 6.420340951277938e-05, "loss": 0.1573, "step": 6380 }, { "grad_norm": 1.2692121267318726, "learning_rate": 6.404479677793874e-05, "loss": 0.1488, "step": 6390 }, { "grad_norm": 1.117326021194458, "learning_rate": 6.388603045145075e-05, "loss": 0.118, "step": 6400 }, { "grad_norm": 1.2169133424758911, "learning_rate": 6.372711226955843e-05, "loss": 0.148, "step": 6410 }, { "grad_norm": 2.0463032722473145, "learning_rate": 6.356804397016564e-05, "loss": 0.1582, "step": 6420 }, { "grad_norm": 1.8079224824905396, "learning_rate": 6.340882729281779e-05, "loss": 0.1947, "step": 6430 }, { "grad_norm": 1.4607234001159668, "learning_rate": 6.324946397868294e-05, "loss": 0.1653, "step": 6440 }, { "grad_norm": 0.779877781867981, "learning_rate": 6.308995577053276e-05, "loss": 0.1332, "step": 6450 }, { "grad_norm": 2.6102097034454346, "learning_rate": 6.293030441272347e-05, "loss": 0.1346, "step": 6460 }, { "grad_norm": 0.9398349523544312, "learning_rate": 6.277051165117677e-05, "loss": 0.0853, "step": 6470 }, { "grad_norm": 1.862646460533142, "learning_rate": 6.261057923336064e-05, "loss": 0.1215, "step": 6480 }, { "grad_norm": 0.9092372059822083, "learning_rate": 6.245050890827042e-05, "loss": 0.1236, "step": 6490 }, { "grad_norm": 3.1480414867401123, "learning_rate": 6.229030242640952e-05, "loss": 0.1459, "step": 6500 }, { "grad_norm": 1.5064548254013062, "learning_rate": 6.212996153977037e-05, "loss": 0.1446, "step": 6510 }, { "grad_norm": 1.4317164421081543, "learning_rate": 6.196948800181523e-05, "loss": 0.1403, "step": 6520 }, { "grad_norm": 0.925515353679657, "learning_rate": 6.180888356745695e-05, "loss": 0.1671, "step": 6530 }, { "grad_norm": 1.6173279285430908, "learning_rate": 6.164814999303995e-05, "loss": 0.1212, "step": 6540 }, { "grad_norm": 3.0452563762664795, "learning_rate": 6.148728903632081e-05, "loss": 0.121, "step": 6550 }, { "grad_norm": 3.201441526412964, "learning_rate": 6.132630245644921e-05, "loss": 0.2469, "step": 6560 }, { "grad_norm": 1.3046343326568604, "learning_rate": 6.116519201394857e-05, "loss": 0.1082, "step": 6570 }, { "grad_norm": 2.350104808807373, "learning_rate": 6.10039594706969e-05, "loss": 0.1415, "step": 6580 }, { "grad_norm": 0.7488396763801575, "learning_rate": 6.084260658990744e-05, "loss": 0.1168, "step": 6590 }, { "grad_norm": 2.6322073936462402, "learning_rate": 6.068113513610943e-05, "loss": 0.1458, "step": 6600 }, { "grad_norm": 0.922398567199707, "learning_rate": 6.0519546875128876e-05, "loss": 0.1167, "step": 6610 }, { "grad_norm": 1.0130202770233154, "learning_rate": 6.035784357406906e-05, "loss": 0.0964, "step": 6620 }, { "grad_norm": 1.9407148361206055, "learning_rate": 6.01960270012914e-05, "loss": 0.121, "step": 6630 }, { "grad_norm": 1.159264087677002, "learning_rate": 6.003409892639599e-05, "loss": 0.1642, "step": 6640 }, { "grad_norm": 2.1008455753326416, "learning_rate": 5.9872061120202336e-05, "loss": 0.1504, "step": 6650 }, { "grad_norm": 1.7095788717269897, "learning_rate": 5.9709915354729914e-05, "loss": 0.1337, "step": 6660 }, { "grad_norm": 1.3086369037628174, "learning_rate": 5.9547663403178824e-05, "loss": 0.1579, "step": 6670 }, { "grad_norm": 1.2651876211166382, "learning_rate": 5.9385307039910445e-05, "loss": 0.1192, "step": 6680 }, { "grad_norm": 2.553370237350464, "learning_rate": 5.922284804042792e-05, "loss": 0.1734, "step": 6690 }, { "grad_norm": 1.1105281114578247, "learning_rate": 5.906028818135687e-05, "loss": 0.0997, "step": 6700 }, { "grad_norm": 1.397526502609253, "learning_rate": 5.889762924042585e-05, "loss": 0.1236, "step": 6710 }, { "grad_norm": 1.351365327835083, "learning_rate": 5.873487299644699e-05, "loss": 0.1252, "step": 6720 }, { "grad_norm": 1.6125704050064087, "learning_rate": 5.857202122929649e-05, "loss": 0.1086, "step": 6730 }, { "grad_norm": 1.1821553707122803, "learning_rate": 5.840907571989518e-05, "loss": 0.1376, "step": 6740 }, { "grad_norm": 2.07326602935791, "learning_rate": 5.824603825018904e-05, "loss": 0.1146, "step": 6750 }, { "grad_norm": 1.0939154624938965, "learning_rate": 5.808291060312975e-05, "loss": 0.128, "step": 6760 }, { "grad_norm": 1.0586215257644653, "learning_rate": 5.7919694562655083e-05, "loss": 0.1825, "step": 6770 }, { "grad_norm": 1.2288000583648682, "learning_rate": 5.775639191366954e-05, "loss": 0.0928, "step": 6780 }, { "grad_norm": 1.187947392463684, "learning_rate": 5.75930044420247e-05, "loss": 0.098, "step": 6790 }, { "grad_norm": 1.8951289653778076, "learning_rate": 5.74295339344998e-05, "loss": 0.2122, "step": 6800 }, { "grad_norm": 2.7530951499938965, "learning_rate": 5.726598217878211e-05, "loss": 0.1919, "step": 6810 }, { "grad_norm": 0.7781674861907959, "learning_rate": 5.71023509634474e-05, "loss": 0.1964, "step": 6820 }, { "grad_norm": 1.0305484533309937, "learning_rate": 5.693864207794049e-05, "loss": 0.1634, "step": 6830 }, { "grad_norm": 3.2862069606781006, "learning_rate": 5.677485731255545e-05, "loss": 0.1575, "step": 6840 }, { "grad_norm": 2.1151626110076904, "learning_rate": 5.6610998458416296e-05, "loss": 0.11, "step": 6850 }, { "grad_norm": 0.9000088572502136, "learning_rate": 5.644706730745716e-05, "loss": 0.127, "step": 6860 }, { "grad_norm": 1.0295668840408325, "learning_rate": 5.628306565240287e-05, "loss": 0.1523, "step": 6870 }, { "grad_norm": 1.610122799873352, "learning_rate": 5.611899528674923e-05, "loss": 0.1368, "step": 6880 }, { "grad_norm": 1.7211111783981323, "learning_rate": 5.595485800474349e-05, "loss": 0.184, "step": 6890 }, { "grad_norm": 2.181696653366089, "learning_rate": 5.579065560136467e-05, "loss": 0.1263, "step": 6900 }, { "grad_norm": 2.5626747608184814, "learning_rate": 5.562638987230392e-05, "loss": 0.1017, "step": 6910 }, { "grad_norm": 1.0421947240829468, "learning_rate": 5.546206261394498e-05, "loss": 0.0857, "step": 6920 }, { "grad_norm": 1.972312331199646, "learning_rate": 5.529767562334437e-05, "loss": 0.1223, "step": 6930 }, { "grad_norm": 1.4084653854370117, "learning_rate": 5.5133230698211926e-05, "loss": 0.1391, "step": 6940 }, { "grad_norm": 2.1309430599212646, "learning_rate": 5.496872963689096e-05, "loss": 0.0728, "step": 6950 }, { "grad_norm": 1.3563735485076904, "learning_rate": 5.4804174238338756e-05, "loss": 0.1197, "step": 6960 }, { "grad_norm": 1.1981059312820435, "learning_rate": 5.463956630210678e-05, "loss": 0.1183, "step": 6970 }, { "grad_norm": 0.9127631783485413, "learning_rate": 5.4474907628321046e-05, "loss": 0.1354, "step": 6980 }, { "grad_norm": 2.7803075313568115, "learning_rate": 5.431020001766244e-05, "loss": 0.0981, "step": 6990 }, { "grad_norm": 2.54464054107666, "learning_rate": 5.4145445271346986e-05, "loss": 0.122, "step": 7000 }, { "grad_norm": 1.8784445524215698, "learning_rate": 5.398064519110622e-05, "loss": 0.2242, "step": 7010 }, { "grad_norm": 1.0987147092819214, "learning_rate": 5.3815801579167394e-05, "loss": 0.1475, "step": 7020 }, { "grad_norm": 1.1494786739349365, "learning_rate": 5.365091623823382e-05, "loss": 0.1135, "step": 7030 }, { "grad_norm": 1.1154922246932983, "learning_rate": 5.348599097146521e-05, "loss": 0.1686, "step": 7040 }, { "grad_norm": 2.908900260925293, "learning_rate": 5.3321027582457836e-05, "loss": 0.1444, "step": 7050 }, { "grad_norm": 1.4258407354354858, "learning_rate": 5.315602787522491e-05, "loss": 0.098, "step": 7060 }, { "grad_norm": 1.4264552593231201, "learning_rate": 5.299099365417678e-05, "loss": 0.0925, "step": 7070 }, { "grad_norm": 1.257734775543213, "learning_rate": 5.2825926724101236e-05, "loss": 0.0951, "step": 7080 }, { "grad_norm": 0.9101056456565857, "learning_rate": 5.26608288901438e-05, "loss": 0.1414, "step": 7090 }, { "grad_norm": 0.8639841079711914, "learning_rate": 5.24957019577879e-05, "loss": 0.1285, "step": 7100 }, { "grad_norm": 1.2864470481872559, "learning_rate": 5.2330547732835266e-05, "loss": 0.143, "step": 7110 }, { "grad_norm": 1.4703928232192993, "learning_rate": 5.2165368021385996e-05, "loss": 0.0745, "step": 7120 }, { "grad_norm": 0.89513099193573, "learning_rate": 5.200016462981897e-05, "loss": 0.1638, "step": 7130 }, { "grad_norm": 1.563665747642517, "learning_rate": 5.1834939364772015e-05, "loss": 0.1556, "step": 7140 }, { "grad_norm": 1.857216238975525, "learning_rate": 5.166969403312214e-05, "loss": 0.1279, "step": 7150 }, { "grad_norm": 1.3357319831848145, "learning_rate": 5.1504430441965844e-05, "loss": 0.0798, "step": 7160 }, { "grad_norm": 1.0861470699310303, "learning_rate": 5.133915039859923e-05, "loss": 0.1396, "step": 7170 }, { "grad_norm": 1.4017977714538574, "learning_rate": 5.1173855710498444e-05, "loss": 0.1357, "step": 7180 }, { "grad_norm": 1.3713289499282837, "learning_rate": 5.100854818529967e-05, "loss": 0.1147, "step": 7190 }, { "grad_norm": 1.111243486404419, "learning_rate": 5.084322963077951e-05, "loss": 0.1129, "step": 7200 }, { "grad_norm": 0.852666974067688, "learning_rate": 5.067790185483522e-05, "loss": 0.065, "step": 7210 }, { "grad_norm": 1.7826944589614868, "learning_rate": 5.0512566665464844e-05, "loss": 0.1737, "step": 7220 }, { "grad_norm": 0.7209739685058594, "learning_rate": 5.034722587074755e-05, "loss": 0.041, "step": 7230 }, { "grad_norm": 0.709304928779602, "learning_rate": 5.018188127882375e-05, "loss": 0.0345, "step": 7240 }, { "grad_norm": 1.207432746887207, "learning_rate": 5.0016534697875417e-05, "loss": 0.0383, "step": 7250 }, { "grad_norm": 1.0377296209335327, "learning_rate": 4.9851187936106294e-05, "loss": 0.0267, "step": 7260 }, { "grad_norm": 0.9182533621788025, "learning_rate": 4.968584280172206e-05, "loss": 0.0522, "step": 7270 }, { "grad_norm": 0.6933771371841431, "learning_rate": 4.95205011029106e-05, "loss": 0.042, "step": 7280 }, { "grad_norm": 0.3947477638721466, "learning_rate": 4.935516464782227e-05, "loss": 0.0244, "step": 7290 }, { "grad_norm": 0.8052363395690918, "learning_rate": 4.918983524455003e-05, "loss": 0.0176, "step": 7300 }, { "grad_norm": 1.3028993606567383, "learning_rate": 4.9024514701109766e-05, "loss": 0.0288, "step": 7310 }, { "grad_norm": 0.6126571297645569, "learning_rate": 4.885920482542043e-05, "loss": 0.0186, "step": 7320 }, { "grad_norm": 0.8411293625831604, "learning_rate": 4.869390742528438e-05, "loss": 0.0155, "step": 7330 }, { "grad_norm": 0.41807064414024353, "learning_rate": 4.852862430836744e-05, "loss": 0.0129, "step": 7340 }, { "grad_norm": 1.5495203733444214, "learning_rate": 4.836335728217933e-05, "loss": 0.0297, "step": 7350 }, { "grad_norm": 0.249651238322258, "learning_rate": 4.819810815405379e-05, "loss": 0.0207, "step": 7360 }, { "grad_norm": 0.6073088049888611, "learning_rate": 4.803287873112877e-05, "loss": 0.0483, "step": 7370 }, { "grad_norm": 0.858202338218689, "learning_rate": 4.786767082032681e-05, "loss": 0.0202, "step": 7380 }, { "grad_norm": 0.9322536587715149, "learning_rate": 4.77024862283351e-05, "loss": 0.0328, "step": 7390 }, { "grad_norm": 0.6784276366233826, "learning_rate": 4.753732676158593e-05, "loss": 0.0585, "step": 7400 }, { "grad_norm": 0.5207681655883789, "learning_rate": 4.737219422623672e-05, "loss": 0.0239, "step": 7410 }, { "grad_norm": 1.2801867723464966, "learning_rate": 4.720709042815044e-05, "loss": 0.0255, "step": 7420 }, { "grad_norm": 1.0369192361831665, "learning_rate": 4.704201717287578e-05, "loss": 0.0214, "step": 7430 }, { "grad_norm": 0.4536983370780945, "learning_rate": 4.6876976265627404e-05, "loss": 0.0333, "step": 7440 }, { "grad_norm": 0.5442711710929871, "learning_rate": 4.671196951126626e-05, "loss": 0.0203, "step": 7450 }, { "grad_norm": 0.28534409403800964, "learning_rate": 4.654699871427971e-05, "loss": 0.018, "step": 7460 }, { "grad_norm": 0.6043292880058289, "learning_rate": 4.6382065678762034e-05, "loss": 0.0211, "step": 7470 }, { "grad_norm": 0.7019131779670715, "learning_rate": 4.6217172208394424e-05, "loss": 0.0199, "step": 7480 }, { "grad_norm": 0.44604459404945374, "learning_rate": 4.605232010642549e-05, "loss": 0.0144, "step": 7490 }, { "grad_norm": 0.41108834743499756, "learning_rate": 4.588751117565142e-05, "loss": 0.015, "step": 7500 }, { "grad_norm": 0.34100672602653503, "learning_rate": 4.5722747218396214e-05, "loss": 0.0144, "step": 7510 }, { "grad_norm": 0.5571150779724121, "learning_rate": 4.5558030036492194e-05, "loss": 0.017, "step": 7520 }, { "grad_norm": 0.874160647392273, "learning_rate": 4.539336143125999e-05, "loss": 0.0168, "step": 7530 }, { "grad_norm": 0.5988022089004517, "learning_rate": 4.522874320348916e-05, "loss": 0.0194, "step": 7540 }, { "grad_norm": 0.39095091819763184, "learning_rate": 4.506417715341821e-05, "loss": 0.017, "step": 7550 }, { "grad_norm": 0.6540190577507019, "learning_rate": 4.489966508071511e-05, "loss": 0.0193, "step": 7560 }, { "grad_norm": 0.472965806722641, "learning_rate": 4.4735208784457575e-05, "loss": 0.0176, "step": 7570 }, { "grad_norm": 0.28557607531547546, "learning_rate": 4.457081006311325e-05, "loss": 0.0114, "step": 7580 }, { "grad_norm": 0.4235917031764984, "learning_rate": 4.440647071452027e-05, "loss": 0.0157, "step": 7590 }, { "grad_norm": 0.4166930913925171, "learning_rate": 4.424219253586737e-05, "loss": 0.0244, "step": 7600 }, { "grad_norm": 0.4047311544418335, "learning_rate": 4.407797732367443e-05, "loss": 0.0155, "step": 7610 }, { "grad_norm": 0.7077624201774597, "learning_rate": 4.391382687377268e-05, "loss": 0.0434, "step": 7620 }, { "grad_norm": 0.5773056149482727, "learning_rate": 4.374974298128512e-05, "loss": 0.0274, "step": 7630 }, { "grad_norm": 0.5004397630691528, "learning_rate": 4.358572744060699e-05, "loss": 0.0147, "step": 7640 }, { "grad_norm": 0.48977017402648926, "learning_rate": 4.342178204538588e-05, "loss": 0.0177, "step": 7650 }, { "grad_norm": 0.5832704305648804, "learning_rate": 4.325790858850241e-05, "loss": 0.0146, "step": 7660 }, { "grad_norm": 0.43779242038726807, "learning_rate": 4.309410886205043e-05, "loss": 0.042, "step": 7670 }, { "grad_norm": 0.8233016729354858, "learning_rate": 4.293038465731752e-05, "loss": 0.0161, "step": 7680 }, { "grad_norm": 0.7267255187034607, "learning_rate": 4.276673776476533e-05, "loss": 0.0199, "step": 7690 }, { "grad_norm": 0.5264241695404053, "learning_rate": 4.260316997401007e-05, "loss": 0.0137, "step": 7700 }, { "grad_norm": 0.5008209347724915, "learning_rate": 4.243968307380293e-05, "loss": 0.0118, "step": 7710 }, { "grad_norm": 0.4032913148403168, "learning_rate": 4.22762788520104e-05, "loss": 0.0119, "step": 7720 }, { "grad_norm": 0.4097349941730499, "learning_rate": 4.211295909559491e-05, "loss": 0.0132, "step": 7730 }, { "grad_norm": 1.23343026638031, "learning_rate": 4.194972559059511e-05, "loss": 0.0172, "step": 7740 }, { "grad_norm": 0.3894916772842407, "learning_rate": 4.178658012210651e-05, "loss": 0.0337, "step": 7750 }, { "grad_norm": 0.6537580490112305, "learning_rate": 4.162352447426177e-05, "loss": 0.0152, "step": 7760 }, { "grad_norm": 0.8270625472068787, "learning_rate": 4.146056043021135e-05, "loss": 0.0271, "step": 7770 }, { "grad_norm": 0.5779984593391418, "learning_rate": 4.1297689772103944e-05, "loss": 0.0345, "step": 7780 }, { "grad_norm": 0.552596390247345, "learning_rate": 4.113491428106694e-05, "loss": 0.0194, "step": 7790 }, { "grad_norm": 0.4605834186077118, "learning_rate": 4.0972235737187055e-05, "loss": 0.0199, "step": 7800 }, { "grad_norm": 1.4474873542785645, "learning_rate": 4.080965591949076e-05, "loss": 0.0319, "step": 7810 }, { "grad_norm": 0.6967477798461914, "learning_rate": 4.0647176605924924e-05, "loss": 0.03, "step": 7820 }, { "grad_norm": 0.33694809675216675, "learning_rate": 4.0484799573337255e-05, "loss": 0.0105, "step": 7830 }, { "grad_norm": 0.3757384717464447, "learning_rate": 4.032252659745699e-05, "loss": 0.0106, "step": 7840 }, { "grad_norm": 0.6247829794883728, "learning_rate": 4.016035945287539e-05, "loss": 0.0325, "step": 7850 }, { "grad_norm": 0.38609710335731506, "learning_rate": 3.999829991302635e-05, "loss": 0.0181, "step": 7860 }, { "grad_norm": 0.8181343078613281, "learning_rate": 3.983634975016707e-05, "loss": 0.0182, "step": 7870 }, { "grad_norm": 0.477130651473999, "learning_rate": 3.967451073535854e-05, "loss": 0.0104, "step": 7880 }, { "grad_norm": 0.39428213238716125, "learning_rate": 3.951278463844633e-05, "loss": 0.0179, "step": 7890 }, { "grad_norm": 0.3988570272922516, "learning_rate": 3.935117322804111e-05, "loss": 0.01, "step": 7900 }, { "grad_norm": 0.6588640213012695, "learning_rate": 3.918967827149938e-05, "loss": 0.0084, "step": 7910 }, { "grad_norm": 0.6041159629821777, "learning_rate": 3.9028301534904094e-05, "loss": 0.0153, "step": 7920 }, { "grad_norm": 0.26583540439605713, "learning_rate": 3.88670447830454e-05, "loss": 0.0081, "step": 7930 }, { "grad_norm": 0.3195621371269226, "learning_rate": 3.870590977940132e-05, "loss": 0.0122, "step": 7940 }, { "grad_norm": 0.6134281158447266, "learning_rate": 3.8544898286118404e-05, "loss": 0.0148, "step": 7950 }, { "grad_norm": 0.44025176763534546, "learning_rate": 3.838401206399257e-05, "loss": 0.0241, "step": 7960 }, { "grad_norm": 0.37634575366973877, "learning_rate": 3.822325287244975e-05, "loss": 0.0105, "step": 7970 }, { "grad_norm": 0.6684257984161377, "learning_rate": 3.8062622469526725e-05, "loss": 0.0432, "step": 7980 }, { "grad_norm": 0.4964136779308319, "learning_rate": 3.790212261185183e-05, "loss": 0.0154, "step": 7990 }, { "grad_norm": 0.40544721484184265, "learning_rate": 3.7741755054625794e-05, "loss": 0.0236, "step": 8000 }, { "grad_norm": 0.8887676000595093, "learning_rate": 3.758152155160255e-05, "loss": 0.01, "step": 8010 }, { "grad_norm": 0.30664631724357605, "learning_rate": 3.742142385506999e-05, "loss": 0.015, "step": 8020 }, { "grad_norm": 0.5146483778953552, "learning_rate": 3.72614637158309e-05, "loss": 0.0152, "step": 8030 }, { "grad_norm": 0.3090154230594635, "learning_rate": 3.710164288318371e-05, "loss": 0.012, "step": 8040 }, { "grad_norm": 0.27367129921913147, "learning_rate": 3.694196310490345e-05, "loss": 0.0088, "step": 8050 }, { "grad_norm": 0.4127488136291504, "learning_rate": 3.678242612722259e-05, "loss": 0.0156, "step": 8060 }, { "grad_norm": 0.4479810893535614, "learning_rate": 3.6623033694811953e-05, "loss": 0.0162, "step": 8070 }, { "grad_norm": 0.7118269205093384, "learning_rate": 3.6463787550761665e-05, "loss": 0.0187, "step": 8080 }, { "grad_norm": 0.23758980631828308, "learning_rate": 3.630468943656202e-05, "loss": 0.0123, "step": 8090 }, { "grad_norm": 0.5002861022949219, "learning_rate": 3.6145741092084523e-05, "loss": 0.008, "step": 8100 }, { "grad_norm": 0.8797154426574707, "learning_rate": 3.598694425556278e-05, "loss": 0.0554, "step": 8110 }, { "grad_norm": 1.276999592781067, "learning_rate": 3.58283006635736e-05, "loss": 0.3098, "step": 8120 }, { "grad_norm": 1.457051396369934, "learning_rate": 3.566981205101781e-05, "loss": 0.2763, "step": 8130 }, { "grad_norm": 1.0294983386993408, "learning_rate": 3.5511480151101556e-05, "loss": 0.2284, "step": 8140 }, { "grad_norm": 3.480923891067505, "learning_rate": 3.5353306695317104e-05, "loss": 0.2076, "step": 8150 }, { "grad_norm": 1.5514417886734009, "learning_rate": 3.519529341342402e-05, "loss": 0.1619, "step": 8160 }, { "grad_norm": 1.8099689483642578, "learning_rate": 3.503744203343026e-05, "loss": 0.0974, "step": 8170 }, { "grad_norm": 1.934470295906067, "learning_rate": 3.487975428157318e-05, "loss": 0.1223, "step": 8180 }, { "grad_norm": 1.268875002861023, "learning_rate": 3.472223188230083e-05, "loss": 0.1172, "step": 8190 }, { "grad_norm": 1.9821053743362427, "learning_rate": 3.4564876558252866e-05, "loss": 0.1459, "step": 8200 }, { "grad_norm": 1.0652520656585693, "learning_rate": 3.440769003024195e-05, "loss": 0.2134, "step": 8210 }, { "grad_norm": 1.4654064178466797, "learning_rate": 3.425067401723477e-05, "loss": 0.1443, "step": 8220 }, { "grad_norm": 1.7006343603134155, "learning_rate": 3.409383023633325e-05, "loss": 0.1132, "step": 8230 }, { "grad_norm": 1.03605055809021, "learning_rate": 3.3937160402755894e-05, "loss": 0.1048, "step": 8240 }, { "grad_norm": 1.414684772491455, "learning_rate": 3.378066622981885e-05, "loss": 0.0864, "step": 8250 }, { "grad_norm": 2.64220929145813, "learning_rate": 3.362434942891738e-05, "loss": 0.1304, "step": 8260 }, { "grad_norm": 2.0282182693481445, "learning_rate": 3.346821170950693e-05, "loss": 0.0998, "step": 8270 }, { "grad_norm": 1.8180134296417236, "learning_rate": 3.3312254779084585e-05, "loss": 0.12, "step": 8280 }, { "grad_norm": 1.0029250383377075, "learning_rate": 3.315648034317039e-05, "loss": 0.0972, "step": 8290 }, { "grad_norm": 2.0710246562957764, "learning_rate": 3.3000890105288564e-05, "loss": 0.1092, "step": 8300 }, { "grad_norm": 2.587975263595581, "learning_rate": 3.284548576694908e-05, "loss": 0.1497, "step": 8310 }, { "grad_norm": 1.1066375970840454, "learning_rate": 3.2690269027628815e-05, "loss": 0.1897, "step": 8320 }, { "grad_norm": 0.9329871535301208, "learning_rate": 3.253524158475324e-05, "loss": 0.1453, "step": 8330 }, { "grad_norm": 1.666224718093872, "learning_rate": 3.238040513367757e-05, "loss": 0.1451, "step": 8340 }, { "grad_norm": 3.2275898456573486, "learning_rate": 3.222576136766843e-05, "loss": 0.1572, "step": 8350 }, { "grad_norm": 2.669989585876465, "learning_rate": 3.2071311977885324e-05, "loss": 0.1537, "step": 8360 }, { "grad_norm": 2.3525824546813965, "learning_rate": 3.191705865336197e-05, "loss": 0.1619, "step": 8370 }, { "grad_norm": 1.1076533794403076, "learning_rate": 3.1763003080988075e-05, "loss": 0.1269, "step": 8380 }, { "grad_norm": 2.3213393688201904, "learning_rate": 3.160914694549063e-05, "loss": 0.1644, "step": 8390 }, { "grad_norm": 0.788842499256134, "learning_rate": 3.145549192941573e-05, "loss": 0.0282, "step": 8400 }, { "grad_norm": 0.6313167810440063, "learning_rate": 3.130203971310999e-05, "loss": 0.017, "step": 8410 }, { "grad_norm": 0.7856450080871582, "learning_rate": 3.114879197470225e-05, "loss": 0.0184, "step": 8420 }, { "grad_norm": 0.47100433707237244, "learning_rate": 3.0995750390085285e-05, "loss": 0.0145, "step": 8430 }, { "grad_norm": 0.8391005992889404, "learning_rate": 3.084291663289728e-05, "loss": 0.0164, "step": 8440 }, { "grad_norm": 0.3089490830898285, "learning_rate": 3.069029237450375e-05, "loss": 0.0179, "step": 8450 }, { "grad_norm": 1.398669958114624, "learning_rate": 3.053787928397911e-05, "loss": 0.0183, "step": 8460 }, { "grad_norm": 0.6086050271987915, "learning_rate": 3.0385679028088526e-05, "loss": 0.0231, "step": 8470 }, { "grad_norm": 0.5684769153594971, "learning_rate": 3.023369327126959e-05, "loss": 0.0146, "step": 8480 }, { "grad_norm": 0.6411479711532593, "learning_rate": 3.0081923675614198e-05, "loss": 0.0404, "step": 8490 }, { "grad_norm": 0.3192841410636902, "learning_rate": 2.993037190085034e-05, "loss": 0.0182, "step": 8500 }, { "grad_norm": 0.3988809883594513, "learning_rate": 2.977903960432392e-05, "loss": 0.0526, "step": 8510 }, { "grad_norm": 0.6413911581039429, "learning_rate": 2.9627928440980722e-05, "loss": 0.0541, "step": 8520 }, { "grad_norm": 0.5700685977935791, "learning_rate": 2.9477040063348183e-05, "loss": 0.0138, "step": 8530 }, { "grad_norm": 0.4260883033275604, "learning_rate": 2.9326376121517456e-05, "loss": 0.0235, "step": 8540 }, { "grad_norm": 0.6406350135803223, "learning_rate": 2.9175938263125236e-05, "loss": 0.0161, "step": 8550 }, { "grad_norm": 0.32973071932792664, "learning_rate": 2.9025728133335873e-05, "loss": 0.042, "step": 8560 }, { "grad_norm": 0.21808475255966187, "learning_rate": 2.8875747374823288e-05, "loss": 0.014, "step": 8570 }, { "grad_norm": 0.3978922367095947, "learning_rate": 2.872599762775298e-05, "loss": 0.0309, "step": 8580 }, { "grad_norm": 0.43990087509155273, "learning_rate": 2.857648052976425e-05, "loss": 0.0126, "step": 8590 }, { "grad_norm": 0.5409092903137207, "learning_rate": 2.8427197715952047e-05, "loss": 0.0168, "step": 8600 }, { "grad_norm": 0.8768467307090759, "learning_rate": 2.8278150818849393e-05, "loss": 0.0127, "step": 8610 }, { "grad_norm": 0.38725268840789795, "learning_rate": 2.812934146840922e-05, "loss": 0.014, "step": 8620 }, { "grad_norm": 0.4657934606075287, "learning_rate": 2.7980771291986764e-05, "loss": 0.0103, "step": 8630 }, { "grad_norm": 0.5855324864387512, "learning_rate": 2.783244191432167e-05, "loss": 0.0275, "step": 8640 }, { "grad_norm": 0.4409773647785187, "learning_rate": 2.768435495752022e-05, "loss": 0.0099, "step": 8650 }, { "grad_norm": 0.8642302751541138, "learning_rate": 2.753651204103771e-05, "loss": 0.0123, "step": 8660 }, { "grad_norm": 0.30015772581100464, "learning_rate": 2.7388914781660523e-05, "loss": 0.0101, "step": 8670 }, { "grad_norm": 0.8514881730079651, "learning_rate": 2.7241564793488693e-05, "loss": 0.0094, "step": 8680 }, { "grad_norm": 0.9567261934280396, "learning_rate": 2.7094463687918037e-05, "loss": 0.0127, "step": 8690 }, { "grad_norm": 0.31610217690467834, "learning_rate": 2.694761307362268e-05, "loss": 0.0079, "step": 8700 }, { "grad_norm": 0.2255050092935562, "learning_rate": 2.6801014556537467e-05, "loss": 0.0124, "step": 8710 }, { "grad_norm": 0.43634510040283203, "learning_rate": 2.6654669739840243e-05, "loss": 0.0117, "step": 8720 }, { "grad_norm": 0.4568555951118469, "learning_rate": 2.650858022393451e-05, "loss": 0.02, "step": 8730 }, { "grad_norm": 0.8556116223335266, "learning_rate": 2.6362747606431747e-05, "loss": 0.0171, "step": 8740 }, { "grad_norm": 0.48292893171310425, "learning_rate": 2.6217173482134172e-05, "loss": 0.0145, "step": 8750 }, { "grad_norm": 0.4280220568180084, "learning_rate": 2.6071859443017044e-05, "loss": 0.0073, "step": 8760 }, { "grad_norm": 0.3266158103942871, "learning_rate": 2.5926807078211414e-05, "loss": 0.0161, "step": 8770 }, { "grad_norm": 0.6052151322364807, "learning_rate": 2.5782017973986728e-05, "loss": 0.0081, "step": 8780 }, { "grad_norm": 0.49538302421569824, "learning_rate": 2.5637493713733374e-05, "loss": 0.0127, "step": 8790 }, { "grad_norm": 0.2500576674938202, "learning_rate": 2.549323587794559e-05, "loss": 0.0125, "step": 8800 }, { "grad_norm": 0.2570004463195801, "learning_rate": 2.5349246044203895e-05, "loss": 0.0108, "step": 8810 }, { "grad_norm": 0.44875362515449524, "learning_rate": 2.520552578715808e-05, "loss": 0.0242, "step": 8820 }, { "grad_norm": 0.5465111136436462, "learning_rate": 2.506207667850981e-05, "loss": 0.0193, "step": 8830 }, { "grad_norm": 0.4824475646018982, "learning_rate": 2.4918900286995555e-05, "loss": 0.0379, "step": 8840 }, { "grad_norm": 0.5685285329818726, "learning_rate": 2.4775998178369458e-05, "loss": 0.02, "step": 8850 }, { "grad_norm": 0.25589653849601746, "learning_rate": 2.4633371915386017e-05, "loss": 0.0081, "step": 8860 }, { "grad_norm": 0.3986184597015381, "learning_rate": 2.4491023057783235e-05, "loss": 0.0178, "step": 8870 }, { "grad_norm": 0.5388959050178528, "learning_rate": 2.4348953162265375e-05, "loss": 0.0169, "step": 8880 }, { "grad_norm": 0.8321703672409058, "learning_rate": 2.420716378248607e-05, "loss": 0.0329, "step": 8890 }, { "grad_norm": 0.30057698488235474, "learning_rate": 2.4065656469031266e-05, "loss": 0.0128, "step": 8900 }, { "grad_norm": 0.4876841604709625, "learning_rate": 2.3924432769402268e-05, "loss": 0.0105, "step": 8910 }, { "grad_norm": 0.2530577778816223, "learning_rate": 2.3783494227998844e-05, "loss": 0.0097, "step": 8920 }, { "grad_norm": 0.48391827940940857, "learning_rate": 2.3642842386102264e-05, "loss": 0.0117, "step": 8930 }, { "grad_norm": 0.40014317631721497, "learning_rate": 2.3502478781858567e-05, "loss": 0.0197, "step": 8940 }, { "grad_norm": 0.3997556269168854, "learning_rate": 2.3362404950261628e-05, "loss": 0.0073, "step": 8950 }, { "grad_norm": 0.37387514114379883, "learning_rate": 2.3222622423136458e-05, "loss": 0.0086, "step": 8960 }, { "grad_norm": 0.2698398232460022, "learning_rate": 2.3083132729122332e-05, "loss": 0.0174, "step": 8970 }, { "grad_norm": 1.4482982158660889, "learning_rate": 2.294393739365621e-05, "loss": 0.0156, "step": 8980 }, { "grad_norm": 0.2810406982898712, "learning_rate": 2.2805037938956e-05, "loss": 0.0067, "step": 8990 }, { "grad_norm": 0.3189747929573059, "learning_rate": 2.266643588400386e-05, "loss": 0.0199, "step": 9000 }, { "grad_norm": 0.3710552453994751, "learning_rate": 2.252813274452969e-05, "loss": 0.008, "step": 9010 }, { "grad_norm": 0.3728056252002716, "learning_rate": 2.2390130032994427e-05, "loss": 0.021, "step": 9020 }, { "grad_norm": 0.3940730392932892, "learning_rate": 2.2252429258573633e-05, "loss": 0.006, "step": 9030 }, { "grad_norm": 1.6259667873382568, "learning_rate": 2.2115031927140904e-05, "loss": 0.0287, "step": 9040 }, { "grad_norm": 0.43943995237350464, "learning_rate": 2.1977939541251463e-05, "loss": 0.0127, "step": 9050 }, { "grad_norm": 0.32675206661224365, "learning_rate": 2.1841153600125684e-05, "loss": 0.0071, "step": 9060 }, { "grad_norm": 0.28180480003356934, "learning_rate": 2.170467559963267e-05, "loss": 0.0081, "step": 9070 }, { "grad_norm": 0.5519963502883911, "learning_rate": 2.1568507032273982e-05, "loss": 0.0123, "step": 9080 }, { "grad_norm": 0.6266724467277527, "learning_rate": 2.1432649387167264e-05, "loss": 0.0207, "step": 9090 }, { "grad_norm": 1.3605291843414307, "learning_rate": 2.1297104150029973e-05, "loss": 0.0173, "step": 9100 }, { "grad_norm": 1.2567826509475708, "learning_rate": 2.116187280316307e-05, "loss": 0.0217, "step": 9110 }, { "grad_norm": 0.3632425367832184, "learning_rate": 2.1026956825434908e-05, "loss": 0.008, "step": 9120 }, { "grad_norm": 0.435910701751709, "learning_rate": 2.0892357692265017e-05, "loss": 0.0077, "step": 9130 }, { "grad_norm": 0.666904628276825, "learning_rate": 2.0758076875607947e-05, "loss": 0.0093, "step": 9140 }, { "grad_norm": 0.29546257853507996, "learning_rate": 2.0624115843937207e-05, "loss": 0.0069, "step": 9150 }, { "grad_norm": 0.20967619121074677, "learning_rate": 2.0490476062229157e-05, "loss": 0.0177, "step": 9160 }, { "grad_norm": 0.4116738736629486, "learning_rate": 2.035715899194704e-05, "loss": 0.0176, "step": 9170 }, { "grad_norm": 0.9032182693481445, "learning_rate": 2.022416609102499e-05, "loss": 0.0108, "step": 9180 }, { "grad_norm": 0.4808456003665924, "learning_rate": 2.009149881385205e-05, "loss": 0.0206, "step": 9190 }, { "grad_norm": 0.1935708373785019, "learning_rate": 1.995915861125634e-05, "loss": 0.0072, "step": 9200 }, { "grad_norm": 0.292745977640152, "learning_rate": 1.9827146930489065e-05, "loss": 0.013, "step": 9210 }, { "grad_norm": 0.6400882601737976, "learning_rate": 1.9695465215208848e-05, "loss": 0.0085, "step": 9220 }, { "grad_norm": 0.7054076194763184, "learning_rate": 1.9564114905465813e-05, "loss": 0.0088, "step": 9230 }, { "grad_norm": 0.44898533821105957, "learning_rate": 1.9433097437685936e-05, "loss": 0.0074, "step": 9240 }, { "grad_norm": 0.3004296123981476, "learning_rate": 1.930241424465521e-05, "loss": 0.0076, "step": 9250 }, { "grad_norm": 0.20809587836265564, "learning_rate": 1.9172066755504115e-05, "loss": 0.0121, "step": 9260 }, { "grad_norm": 0.365384042263031, "learning_rate": 1.9042056395691914e-05, "loss": 0.0196, "step": 9270 }, { "grad_norm": 4.157004356384277, "learning_rate": 1.8912384586991066e-05, "loss": 0.1241, "step": 9280 }, { "grad_norm": 1.9715690612792969, "learning_rate": 1.8783052747471717e-05, "loss": 0.1136, "step": 9290 }, { "grad_norm": 0.7683215141296387, "learning_rate": 1.865406229148611e-05, "loss": 0.1756, "step": 9300 }, { "grad_norm": 3.93827748298645, "learning_rate": 1.8525414629653233e-05, "loss": 0.1399, "step": 9310 }, { "grad_norm": 5.585920333862305, "learning_rate": 1.8397111168843255e-05, "loss": 0.2399, "step": 9320 }, { "grad_norm": 1.119153618812561, "learning_rate": 1.8269153312162323e-05, "loss": 0.1763, "step": 9330 }, { "grad_norm": 1.4112812280654907, "learning_rate": 1.8141542458937054e-05, "loss": 0.1467, "step": 9340 }, { "grad_norm": 1.664231300354004, "learning_rate": 1.8014280004699268e-05, "loss": 0.1338, "step": 9350 }, { "grad_norm": 2.0256919860839844, "learning_rate": 1.788736734117078e-05, "loss": 0.1049, "step": 9360 }, { "grad_norm": 2.1683149337768555, "learning_rate": 1.7760805856248152e-05, "loss": 0.1888, "step": 9370 }, { "grad_norm": 1.6567128896713257, "learning_rate": 1.7634596933987518e-05, "loss": 0.2068, "step": 9380 }, { "grad_norm": 0.9550359845161438, "learning_rate": 1.7508741954589404e-05, "loss": 0.0806, "step": 9390 }, { "grad_norm": 0.8125738501548767, "learning_rate": 1.7383242294383717e-05, "loss": 0.1498, "step": 9400 }, { "grad_norm": 1.1275123357772827, "learning_rate": 1.7258099325814632e-05, "loss": 0.1548, "step": 9410 }, { "grad_norm": 1.6616339683532715, "learning_rate": 1.7133314417425594e-05, "loss": 0.0952, "step": 9420 }, { "grad_norm": 1.3431308269500732, "learning_rate": 1.7008888933844408e-05, "loss": 0.2014, "step": 9430 }, { "grad_norm": 0.8460913300514221, "learning_rate": 1.6884824235768172e-05, "loss": 0.1494, "step": 9440 }, { "grad_norm": 0.9945228695869446, "learning_rate": 1.6761121679948592e-05, "loss": 0.0597, "step": 9450 }, { "grad_norm": 1.1904104948043823, "learning_rate": 1.663778261917695e-05, "loss": 0.0976, "step": 9460 }, { "grad_norm": 0.9898011088371277, "learning_rate": 1.651480840226952e-05, "loss": 0.0791, "step": 9470 }, { "grad_norm": 0.9907180070877075, "learning_rate": 1.639220037405258e-05, "loss": 0.1084, "step": 9480 }, { "grad_norm": 0.6528921723365784, "learning_rate": 1.6269959875347906e-05, "loss": 0.0666, "step": 9490 }, { "grad_norm": 0.9692262411117554, "learning_rate": 1.614808824295802e-05, "loss": 0.0797, "step": 9500 }, { "grad_norm": 0.778036892414093, "learning_rate": 1.602658680965152e-05, "loss": 0.0901, "step": 9510 }, { "grad_norm": 1.4358996152877808, "learning_rate": 1.5905456904148686e-05, "loss": 0.1493, "step": 9520 }, { "grad_norm": 2.228118419647217, "learning_rate": 1.57846998511067e-05, "loss": 0.1109, "step": 9530 }, { "grad_norm": 0.9565455317497253, "learning_rate": 1.566431697110538e-05, "loss": 0.1203, "step": 9540 }, { "grad_norm": 1.8373321294784546, "learning_rate": 1.554430958063259e-05, "loss": 0.0966, "step": 9550 }, { "grad_norm": 1.5351370573043823, "learning_rate": 1.5424678992069912e-05, "loss": 0.0742, "step": 9560 }, { "grad_norm": 2.528749704360962, "learning_rate": 1.5305426513678362e-05, "loss": 0.1606, "step": 9570 }, { "grad_norm": 2.2176082134246826, "learning_rate": 1.518655344958388e-05, "loss": 0.1563, "step": 9580 }, { "grad_norm": 1.5316991806030273, "learning_rate": 1.5068061099763275e-05, "loss": 0.1107, "step": 9590 }, { "grad_norm": 0.5689011216163635, "learning_rate": 1.494995076002988e-05, "loss": 0.0661, "step": 9600 }, { "grad_norm": 1.0998042821884155, "learning_rate": 1.4832223722019456e-05, "loss": 0.1445, "step": 9610 }, { "grad_norm": 1.0791428089141846, "learning_rate": 1.4714881273176035e-05, "loss": 0.0858, "step": 9620 }, { "grad_norm": 0.6619061827659607, "learning_rate": 1.4597924696737835e-05, "loss": 0.1132, "step": 9630 }, { "grad_norm": 1.8049027919769287, "learning_rate": 1.4481355271723252e-05, "loss": 0.1342, "step": 9640 }, { "grad_norm": 0.7240359783172607, "learning_rate": 1.4365174272916809e-05, "loss": 0.1323, "step": 9650 }, { "grad_norm": 2.7417116165161133, "learning_rate": 1.4249382970855319e-05, "loss": 0.1685, "step": 9660 }, { "grad_norm": 1.247850775718689, "learning_rate": 1.4133982631813903e-05, "loss": 0.1294, "step": 9670 }, { "grad_norm": 2.6065242290496826, "learning_rate": 1.4018974517792194e-05, "loss": 0.1386, "step": 9680 }, { "grad_norm": 1.7329907417297363, "learning_rate": 1.390435988650048e-05, "loss": 0.0974, "step": 9690 }, { "grad_norm": 0.5939985513687134, "learning_rate": 1.3790139991346006e-05, "loss": 0.0616, "step": 9700 }, { "grad_norm": 0.8865715265274048, "learning_rate": 1.367631608141926e-05, "loss": 0.0768, "step": 9710 }, { "grad_norm": 0.6667280197143555, "learning_rate": 1.3562889401480278e-05, "loss": 0.0871, "step": 9720 }, { "grad_norm": 0.6840537190437317, "learning_rate": 1.3449861191945074e-05, "loss": 0.1161, "step": 9730 }, { "grad_norm": 1.0476784706115723, "learning_rate": 1.3337232688872009e-05, "loss": 0.1071, "step": 9740 }, { "grad_norm": 1.2856920957565308, "learning_rate": 1.3225005123948364e-05, "loss": 0.07, "step": 9750 }, { "grad_norm": 1.6053963899612427, "learning_rate": 1.311317972447681e-05, "loss": 0.1656, "step": 9760 }, { "grad_norm": 0.6024764180183411, "learning_rate": 1.3001757713361996e-05, "loss": 0.0717, "step": 9770 }, { "grad_norm": 1.3848552703857422, "learning_rate": 1.2890740309097204e-05, "loss": 0.1072, "step": 9780 }, { "grad_norm": 0.9103549718856812, "learning_rate": 1.2780128725750944e-05, "loss": 0.0892, "step": 9790 }, { "grad_norm": 1.1624550819396973, "learning_rate": 1.266992417295379e-05, "loss": 0.0602, "step": 9800 }, { "grad_norm": 1.593658685684204, "learning_rate": 1.2560127855885073e-05, "loss": 0.1382, "step": 9810 }, { "grad_norm": 2.0817809104919434, "learning_rate": 1.2450740975259745e-05, "loss": 0.0659, "step": 9820 }, { "grad_norm": 0.5789938569068909, "learning_rate": 1.234176472731517e-05, "loss": 0.0784, "step": 9830 }, { "grad_norm": 0.35579508543014526, "learning_rate": 1.2233200303798158e-05, "loss": 0.1057, "step": 9840 }, { "grad_norm": 0.641800045967102, "learning_rate": 1.2125048891951846e-05, "loss": 0.0481, "step": 9850 }, { "grad_norm": 1.0702484846115112, "learning_rate": 1.2017311674502745e-05, "loss": 0.1244, "step": 9860 }, { "grad_norm": 0.7272056937217712, "learning_rate": 1.1909989829647822e-05, "loss": 0.0852, "step": 9870 }, { "grad_norm": 1.0328195095062256, "learning_rate": 1.1803084531041553e-05, "loss": 0.1152, "step": 9880 }, { "grad_norm": 1.2145726680755615, "learning_rate": 1.1696596947783162e-05, "loss": 0.1735, "step": 9890 }, { "grad_norm": 1.2381833791732788, "learning_rate": 1.1590528244403803e-05, "loss": 0.055, "step": 9900 }, { "grad_norm": 1.1508761644363403, "learning_rate": 1.148487958085382e-05, "loss": 0.0905, "step": 9910 }, { "grad_norm": 1.4407891035079956, "learning_rate": 1.1379652112490086e-05, "loss": 0.1218, "step": 9920 }, { "grad_norm": 1.088624119758606, "learning_rate": 1.1274846990063315e-05, "loss": 0.0664, "step": 9930 }, { "grad_norm": 0.6003662943840027, "learning_rate": 1.117046535970554e-05, "loss": 0.1576, "step": 9940 }, { "grad_norm": 1.5742018222808838, "learning_rate": 1.106650836291755e-05, "loss": 0.0729, "step": 9950 }, { "grad_norm": 1.638815999031067, "learning_rate": 1.0962977136556418e-05, "loss": 0.0818, "step": 9960 }, { "grad_norm": 1.2643394470214844, "learning_rate": 1.0859872812823024e-05, "loss": 0.0988, "step": 9970 }, { "grad_norm": 1.6043508052825928, "learning_rate": 1.0757196519249747e-05, "loss": 0.1529, "step": 9980 }, { "grad_norm": 1.6593811511993408, "learning_rate": 1.0654949378688077e-05, "loss": 0.0935, "step": 9990 }, { "grad_norm": 0.8385270833969116, "learning_rate": 1.0553132509296376e-05, "loss": 0.0747, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }