{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 13.175230566534914, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.013175230566534914, "grad_norm": 12.585679054260254, "learning_rate": 1.8e-06, "loss": 1.3094, "step": 10 }, { "epoch": 0.026350461133069828, "grad_norm": 4.774308681488037, "learning_rate": 3.8e-06, "loss": 1.0382, "step": 20 }, { "epoch": 0.039525691699604744, "grad_norm": 3.6964216232299805, "learning_rate": 5.8e-06, "loss": 0.5183, "step": 30 }, { "epoch": 0.052700922266139656, "grad_norm": 1.9297399520874023, "learning_rate": 7.8e-06, "loss": 0.3297, "step": 40 }, { "epoch": 0.06587615283267458, "grad_norm": 3.0238001346588135, "learning_rate": 9.800000000000001e-06, "loss": 0.2296, "step": 50 }, { "epoch": 0.07905138339920949, "grad_norm": 1.9905507564544678, "learning_rate": 1.18e-05, "loss": 0.1924, "step": 60 }, { "epoch": 0.0922266139657444, "grad_norm": 1.7341668605804443, "learning_rate": 1.3800000000000002e-05, "loss": 0.1222, "step": 70 }, { "epoch": 0.10540184453227931, "grad_norm": 1.1846983432769775, "learning_rate": 1.58e-05, "loss": 0.1143, "step": 80 }, { "epoch": 0.11857707509881422, "grad_norm": 0.9171633720397949, "learning_rate": 1.78e-05, "loss": 0.0895, "step": 90 }, { "epoch": 0.13175230566534915, "grad_norm": 0.9449801445007324, "learning_rate": 1.9800000000000004e-05, "loss": 0.0864, "step": 100 }, { "epoch": 0.14492753623188406, "grad_norm": 0.49553966522216797, "learning_rate": 2.18e-05, "loss": 0.073, "step": 110 }, { "epoch": 0.15810276679841898, "grad_norm": 0.9112503528594971, "learning_rate": 2.38e-05, "loss": 0.0631, "step": 120 }, { "epoch": 0.1712779973649539, "grad_norm": 1.0028133392333984, "learning_rate": 2.58e-05, "loss": 0.0668, "step": 130 }, { "epoch": 0.1844532279314888, "grad_norm": 1.2871983051300049, "learning_rate": 2.7800000000000005e-05, "loss": 0.0646, "step": 140 }, { "epoch": 0.1976284584980237, "grad_norm": 0.9798805117607117, "learning_rate": 2.98e-05, "loss": 0.0565, "step": 150 }, { "epoch": 0.21080368906455862, "grad_norm": 0.8933775424957275, "learning_rate": 3.18e-05, "loss": 0.0581, "step": 160 }, { "epoch": 0.22397891963109354, "grad_norm": 1.1488267183303833, "learning_rate": 3.38e-05, "loss": 0.0521, "step": 170 }, { "epoch": 0.23715415019762845, "grad_norm": 1.329391360282898, "learning_rate": 3.58e-05, "loss": 0.0518, "step": 180 }, { "epoch": 0.2503293807641634, "grad_norm": 0.9331446290016174, "learning_rate": 3.7800000000000004e-05, "loss": 0.05, "step": 190 }, { "epoch": 0.2635046113306983, "grad_norm": 0.7018640041351318, "learning_rate": 3.9800000000000005e-05, "loss": 0.0475, "step": 200 }, { "epoch": 0.2766798418972332, "grad_norm": 0.7023022174835205, "learning_rate": 4.18e-05, "loss": 0.0506, "step": 210 }, { "epoch": 0.2898550724637681, "grad_norm": 0.6905364394187927, "learning_rate": 4.38e-05, "loss": 0.0627, "step": 220 }, { "epoch": 0.30303030303030304, "grad_norm": 1.085239052772522, "learning_rate": 4.58e-05, "loss": 0.0481, "step": 230 }, { "epoch": 0.31620553359683795, "grad_norm": 0.8007176518440247, "learning_rate": 4.78e-05, "loss": 0.0519, "step": 240 }, { "epoch": 0.32938076416337286, "grad_norm": 1.0830175876617432, "learning_rate": 4.9800000000000004e-05, "loss": 0.0501, "step": 250 }, { "epoch": 0.3425559947299078, "grad_norm": 1.1084550619125366, "learning_rate": 5.1800000000000005e-05, "loss": 0.0553, "step": 260 }, { "epoch": 0.3557312252964427, "grad_norm": 1.091086983680725, "learning_rate": 5.380000000000001e-05, "loss": 0.0605, "step": 270 }, { "epoch": 0.3689064558629776, "grad_norm": 0.6870923638343811, "learning_rate": 5.580000000000001e-05, "loss": 0.047, "step": 280 }, { "epoch": 0.3820816864295125, "grad_norm": 0.44346657395362854, "learning_rate": 5.7799999999999995e-05, "loss": 0.0561, "step": 290 }, { "epoch": 0.3952569169960474, "grad_norm": 1.086161732673645, "learning_rate": 5.9800000000000003e-05, "loss": 0.05, "step": 300 }, { "epoch": 0.40843214756258234, "grad_norm": 0.6184456944465637, "learning_rate": 6.18e-05, "loss": 0.0505, "step": 310 }, { "epoch": 0.42160737812911725, "grad_norm": 0.7681217789649963, "learning_rate": 6.38e-05, "loss": 0.0458, "step": 320 }, { "epoch": 0.43478260869565216, "grad_norm": 0.7591136693954468, "learning_rate": 6.58e-05, "loss": 0.0504, "step": 330 }, { "epoch": 0.4479578392621871, "grad_norm": 0.7818257808685303, "learning_rate": 6.780000000000001e-05, "loss": 0.049, "step": 340 }, { "epoch": 0.461133069828722, "grad_norm": 0.929442286491394, "learning_rate": 6.98e-05, "loss": 0.0429, "step": 350 }, { "epoch": 0.4743083003952569, "grad_norm": 0.8063979148864746, "learning_rate": 7.18e-05, "loss": 0.0394, "step": 360 }, { "epoch": 0.4874835309617918, "grad_norm": 0.8431144952774048, "learning_rate": 7.38e-05, "loss": 0.048, "step": 370 }, { "epoch": 0.5006587615283268, "grad_norm": 0.6320706009864807, "learning_rate": 7.58e-05, "loss": 0.0445, "step": 380 }, { "epoch": 0.5138339920948617, "grad_norm": 0.6919974684715271, "learning_rate": 7.780000000000001e-05, "loss": 0.0515, "step": 390 }, { "epoch": 0.5270092226613966, "grad_norm": 1.066429615020752, "learning_rate": 7.98e-05, "loss": 0.048, "step": 400 }, { "epoch": 0.5401844532279315, "grad_norm": 0.7330025434494019, "learning_rate": 8.18e-05, "loss": 0.0499, "step": 410 }, { "epoch": 0.5533596837944664, "grad_norm": 0.642924427986145, "learning_rate": 8.38e-05, "loss": 0.0437, "step": 420 }, { "epoch": 0.5665349143610013, "grad_norm": 0.6752575635910034, "learning_rate": 8.58e-05, "loss": 0.0441, "step": 430 }, { "epoch": 0.5797101449275363, "grad_norm": 0.9005573391914368, "learning_rate": 8.78e-05, "loss": 0.0418, "step": 440 }, { "epoch": 0.5928853754940712, "grad_norm": 0.870056688785553, "learning_rate": 8.98e-05, "loss": 0.0451, "step": 450 }, { "epoch": 0.6060606060606061, "grad_norm": 1.2314468622207642, "learning_rate": 9.180000000000001e-05, "loss": 0.0521, "step": 460 }, { "epoch": 0.619235836627141, "grad_norm": 0.8210376501083374, "learning_rate": 9.38e-05, "loss": 0.0541, "step": 470 }, { "epoch": 0.6324110671936759, "grad_norm": 0.8480843901634216, "learning_rate": 9.58e-05, "loss": 0.0456, "step": 480 }, { "epoch": 0.6455862977602108, "grad_norm": 0.9127245545387268, "learning_rate": 9.78e-05, "loss": 0.0424, "step": 490 }, { "epoch": 0.6587615283267457, "grad_norm": 0.7549911141395569, "learning_rate": 9.98e-05, "loss": 0.0459, "step": 500 }, { "epoch": 0.6719367588932806, "grad_norm": 0.5965666174888611, "learning_rate": 9.9999778549206e-05, "loss": 0.0406, "step": 510 }, { "epoch": 0.6851119894598156, "grad_norm": 0.666847825050354, "learning_rate": 9.999901304280685e-05, "loss": 0.0417, "step": 520 }, { "epoch": 0.6982872200263505, "grad_norm": 0.4748551547527313, "learning_rate": 9.999770075521164e-05, "loss": 0.0475, "step": 530 }, { "epoch": 0.7114624505928854, "grad_norm": 1.098746418952942, "learning_rate": 9.99958417007713e-05, "loss": 0.0434, "step": 540 }, { "epoch": 0.7246376811594203, "grad_norm": 0.5824465751647949, "learning_rate": 9.999343589981615e-05, "loss": 0.0407, "step": 550 }, { "epoch": 0.7378129117259552, "grad_norm": 0.4799025058746338, "learning_rate": 9.999048337865568e-05, "loss": 0.0369, "step": 560 }, { "epoch": 0.7509881422924901, "grad_norm": 0.6298460960388184, "learning_rate": 9.998698416957815e-05, "loss": 0.0375, "step": 570 }, { "epoch": 0.764163372859025, "grad_norm": 0.6886503100395203, "learning_rate": 9.998293831085037e-05, "loss": 0.0389, "step": 580 }, { "epoch": 0.7773386034255599, "grad_norm": 0.5728000402450562, "learning_rate": 9.997834584671719e-05, "loss": 0.0417, "step": 590 }, { "epoch": 0.7905138339920948, "grad_norm": 0.7823150157928467, "learning_rate": 9.997320682740107e-05, "loss": 0.0409, "step": 600 }, { "epoch": 0.8036890645586298, "grad_norm": 0.8241381645202637, "learning_rate": 9.996752130910149e-05, "loss": 0.0459, "step": 610 }, { "epoch": 0.8168642951251647, "grad_norm": 0.5445210933685303, "learning_rate": 9.99612893539944e-05, "loss": 0.0352, "step": 620 }, { "epoch": 0.8300395256916996, "grad_norm": 0.5866085886955261, "learning_rate": 9.995451103023144e-05, "loss": 0.0295, "step": 630 }, { "epoch": 0.8432147562582345, "grad_norm": 0.7469897270202637, "learning_rate": 9.994718641193928e-05, "loss": 0.0321, "step": 640 }, { "epoch": 0.8563899868247694, "grad_norm": 0.869316041469574, "learning_rate": 9.993931557921874e-05, "loss": 0.0353, "step": 650 }, { "epoch": 0.8695652173913043, "grad_norm": 0.6988140344619751, "learning_rate": 9.993089861814402e-05, "loss": 0.0343, "step": 660 }, { "epoch": 0.8827404479578392, "grad_norm": 0.5596421360969543, "learning_rate": 9.992193562076166e-05, "loss": 0.0424, "step": 670 }, { "epoch": 0.8959156785243741, "grad_norm": 0.5170013308525085, "learning_rate": 9.991242668508954e-05, "loss": 0.0373, "step": 680 }, { "epoch": 0.9090909090909091, "grad_norm": 0.5016767382621765, "learning_rate": 9.990237191511587e-05, "loss": 0.0379, "step": 690 }, { "epoch": 0.922266139657444, "grad_norm": 0.445021390914917, "learning_rate": 9.989177142079802e-05, "loss": 0.0352, "step": 700 }, { "epoch": 0.9354413702239789, "grad_norm": 0.41842174530029297, "learning_rate": 9.988062531806126e-05, "loss": 0.0383, "step": 710 }, { "epoch": 0.9486166007905138, "grad_norm": 0.4440097212791443, "learning_rate": 9.986893372879762e-05, "loss": 0.0334, "step": 720 }, { "epoch": 0.9617918313570487, "grad_norm": 0.3641037940979004, "learning_rate": 9.985669678086443e-05, "loss": 0.034, "step": 730 }, { "epoch": 0.9749670619235836, "grad_norm": 0.5874915719032288, "learning_rate": 9.984391460808298e-05, "loss": 0.0334, "step": 740 }, { "epoch": 0.9881422924901185, "grad_norm": 0.736756443977356, "learning_rate": 9.983058735023709e-05, "loss": 0.038, "step": 750 }, { "epoch": 1.0013175230566536, "grad_norm": 0.6323824524879456, "learning_rate": 9.98167151530715e-05, "loss": 0.0344, "step": 760 }, { "epoch": 1.0144927536231885, "grad_norm": 0.4926735758781433, "learning_rate": 9.980229816829034e-05, "loss": 0.0293, "step": 770 }, { "epoch": 1.0276679841897234, "grad_norm": 0.6455330848693848, "learning_rate": 9.978733655355544e-05, "loss": 0.0282, "step": 780 }, { "epoch": 1.0408432147562583, "grad_norm": 0.7627984881401062, "learning_rate": 9.977183047248464e-05, "loss": 0.0296, "step": 790 }, { "epoch": 1.0540184453227932, "grad_norm": 0.5647096633911133, "learning_rate": 9.975578009464992e-05, "loss": 0.0349, "step": 800 }, { "epoch": 1.0671936758893281, "grad_norm": 0.3709547817707062, "learning_rate": 9.97391855955757e-05, "loss": 0.0322, "step": 810 }, { "epoch": 1.080368906455863, "grad_norm": 0.7317115068435669, "learning_rate": 9.972204715673669e-05, "loss": 0.0332, "step": 820 }, { "epoch": 1.093544137022398, "grad_norm": 0.6597622632980347, "learning_rate": 9.970436496555617e-05, "loss": 0.0338, "step": 830 }, { "epoch": 1.1067193675889329, "grad_norm": 0.4061703383922577, "learning_rate": 9.968613921540373e-05, "loss": 0.0304, "step": 840 }, { "epoch": 1.1198945981554678, "grad_norm": 0.4285595118999481, "learning_rate": 9.966737010559326e-05, "loss": 0.0374, "step": 850 }, { "epoch": 1.1330698287220027, "grad_norm": 0.5008792877197266, "learning_rate": 9.964805784138072e-05, "loss": 0.0269, "step": 860 }, { "epoch": 1.1462450592885376, "grad_norm": 0.3218404948711395, "learning_rate": 9.962820263396195e-05, "loss": 0.0333, "step": 870 }, { "epoch": 1.1594202898550725, "grad_norm": 0.550163984298706, "learning_rate": 9.960780470047033e-05, "loss": 0.0302, "step": 880 }, { "epoch": 1.1725955204216074, "grad_norm": 0.4781897962093353, "learning_rate": 9.958686426397437e-05, "loss": 0.0282, "step": 890 }, { "epoch": 1.1857707509881423, "grad_norm": 0.41555675864219666, "learning_rate": 9.956538155347534e-05, "loss": 0.0338, "step": 900 }, { "epoch": 1.1989459815546772, "grad_norm": 0.6632710099220276, "learning_rate": 9.95433568039047e-05, "loss": 0.0349, "step": 910 }, { "epoch": 1.2121212121212122, "grad_norm": 0.6058996915817261, "learning_rate": 9.952079025612162e-05, "loss": 0.0334, "step": 920 }, { "epoch": 1.225296442687747, "grad_norm": 0.46293482184410095, "learning_rate": 9.949768215691022e-05, "loss": 0.0296, "step": 930 }, { "epoch": 1.238471673254282, "grad_norm": 0.5244153738021851, "learning_rate": 9.9474032758977e-05, "loss": 0.0287, "step": 940 }, { "epoch": 1.251646903820817, "grad_norm": 0.4336259961128235, "learning_rate": 9.944984232094794e-05, "loss": 0.0327, "step": 950 }, { "epoch": 1.2648221343873518, "grad_norm": 0.5450809001922607, "learning_rate": 9.942511110736584e-05, "loss": 0.0287, "step": 960 }, { "epoch": 1.2779973649538867, "grad_norm": 0.37985944747924805, "learning_rate": 9.939983938868726e-05, "loss": 0.0294, "step": 970 }, { "epoch": 1.2911725955204216, "grad_norm": 0.6127066612243652, "learning_rate": 9.93740274412797e-05, "loss": 0.0265, "step": 980 }, { "epoch": 1.3043478260869565, "grad_norm": 0.6105954051017761, "learning_rate": 9.934767554741846e-05, "loss": 0.0341, "step": 990 }, { "epoch": 1.3175230566534915, "grad_norm": 0.350715696811676, "learning_rate": 9.932078399528361e-05, "loss": 0.026, "step": 1000 }, { "epoch": 1.3306982872200264, "grad_norm": 0.34700220823287964, "learning_rate": 9.929335307895689e-05, "loss": 0.0249, "step": 1010 }, { "epoch": 1.3438735177865613, "grad_norm": 0.2994444668292999, "learning_rate": 9.926538309841839e-05, "loss": 0.0254, "step": 1020 }, { "epoch": 1.3570487483530962, "grad_norm": 0.2695624530315399, "learning_rate": 9.923687435954334e-05, "loss": 0.0259, "step": 1030 }, { "epoch": 1.370223978919631, "grad_norm": 0.35280922055244446, "learning_rate": 9.920782717409873e-05, "loss": 0.0296, "step": 1040 }, { "epoch": 1.383399209486166, "grad_norm": 0.4313255548477173, "learning_rate": 9.917824185973994e-05, "loss": 0.0219, "step": 1050 }, { "epoch": 1.396574440052701, "grad_norm": 0.6447331309318542, "learning_rate": 9.914811874000723e-05, "loss": 0.033, "step": 1060 }, { "epoch": 1.4097496706192358, "grad_norm": 0.6714804172515869, "learning_rate": 9.911745814432218e-05, "loss": 0.0316, "step": 1070 }, { "epoch": 1.4229249011857708, "grad_norm": 0.35933995246887207, "learning_rate": 9.90862604079842e-05, "loss": 0.0288, "step": 1080 }, { "epoch": 1.4361001317523057, "grad_norm": 0.6764751672744751, "learning_rate": 9.90545258721667e-05, "loss": 0.0221, "step": 1090 }, { "epoch": 1.4492753623188406, "grad_norm": 0.4556847810745239, "learning_rate": 9.90222548839135e-05, "loss": 0.0261, "step": 1100 }, { "epoch": 1.4624505928853755, "grad_norm": 0.49368539452552795, "learning_rate": 9.898944779613495e-05, "loss": 0.0266, "step": 1110 }, { "epoch": 1.4756258234519104, "grad_norm": 0.3783435821533203, "learning_rate": 9.89561049676041e-05, "loss": 0.0259, "step": 1120 }, { "epoch": 1.4888010540184453, "grad_norm": 0.26338449120521545, "learning_rate": 9.89222267629528e-05, "loss": 0.0283, "step": 1130 }, { "epoch": 1.5019762845849802, "grad_norm": 0.4226028621196747, "learning_rate": 9.888781355266763e-05, "loss": 0.0242, "step": 1140 }, { "epoch": 1.5151515151515151, "grad_norm": 0.38758257031440735, "learning_rate": 9.885286571308598e-05, "loss": 0.0301, "step": 1150 }, { "epoch": 1.52832674571805, "grad_norm": 0.3099963366985321, "learning_rate": 9.881738362639182e-05, "loss": 0.026, "step": 1160 }, { "epoch": 1.541501976284585, "grad_norm": 0.519696056842804, "learning_rate": 9.878136768061154e-05, "loss": 0.0258, "step": 1170 }, { "epoch": 1.5546772068511199, "grad_norm": 0.4573246240615845, "learning_rate": 9.874481826960979e-05, "loss": 0.0247, "step": 1180 }, { "epoch": 1.5678524374176548, "grad_norm": 0.6357062458992004, "learning_rate": 9.870773579308503e-05, "loss": 0.0249, "step": 1190 }, { "epoch": 1.5810276679841897, "grad_norm": 0.5196475386619568, "learning_rate": 9.867012065656533e-05, "loss": 0.032, "step": 1200 }, { "epoch": 1.5942028985507246, "grad_norm": 0.7611187696456909, "learning_rate": 9.863197327140376e-05, "loss": 0.0302, "step": 1210 }, { "epoch": 1.6073781291172595, "grad_norm": 0.3952830731868744, "learning_rate": 9.859329405477403e-05, "loss": 0.0241, "step": 1220 }, { "epoch": 1.6205533596837944, "grad_norm": 0.5244176983833313, "learning_rate": 9.855408342966585e-05, "loss": 0.0285, "step": 1230 }, { "epoch": 1.6337285902503293, "grad_norm": 0.45377880334854126, "learning_rate": 9.851434182488033e-05, "loss": 0.0268, "step": 1240 }, { "epoch": 1.6469038208168643, "grad_norm": 0.35962337255477905, "learning_rate": 9.84740696750253e-05, "loss": 0.0304, "step": 1250 }, { "epoch": 1.6600790513833992, "grad_norm": 0.49337679147720337, "learning_rate": 9.843326742051055e-05, "loss": 0.0266, "step": 1260 }, { "epoch": 1.673254281949934, "grad_norm": 0.4094015061855316, "learning_rate": 9.839193550754297e-05, "loss": 0.0328, "step": 1270 }, { "epoch": 1.686429512516469, "grad_norm": 0.28054094314575195, "learning_rate": 9.835007438812177e-05, "loss": 0.0298, "step": 1280 }, { "epoch": 1.699604743083004, "grad_norm": 0.38139668107032776, "learning_rate": 9.830768452003341e-05, "loss": 0.0318, "step": 1290 }, { "epoch": 1.7127799736495388, "grad_norm": 0.3427455723285675, "learning_rate": 9.826476636684671e-05, "loss": 0.0262, "step": 1300 }, { "epoch": 1.7259552042160737, "grad_norm": 0.4167240858078003, "learning_rate": 9.822132039790773e-05, "loss": 0.026, "step": 1310 }, { "epoch": 1.7391304347826086, "grad_norm": 0.3868737816810608, "learning_rate": 9.817734708833461e-05, "loss": 0.0318, "step": 1320 }, { "epoch": 1.7523056653491436, "grad_norm": 0.2871069014072418, "learning_rate": 9.813284691901243e-05, "loss": 0.0278, "step": 1330 }, { "epoch": 1.7654808959156785, "grad_norm": 0.5268304347991943, "learning_rate": 9.808782037658792e-05, "loss": 0.027, "step": 1340 }, { "epoch": 1.7786561264822134, "grad_norm": 0.34222403168678284, "learning_rate": 9.804226795346411e-05, "loss": 0.0247, "step": 1350 }, { "epoch": 1.7918313570487485, "grad_norm": 0.21124961972236633, "learning_rate": 9.799619014779503e-05, "loss": 0.0267, "step": 1360 }, { "epoch": 1.8050065876152832, "grad_norm": 0.6499063968658447, "learning_rate": 9.794958746348013e-05, "loss": 0.0209, "step": 1370 }, { "epoch": 1.8181818181818183, "grad_norm": 0.2798737585544586, "learning_rate": 9.790246041015896e-05, "loss": 0.0237, "step": 1380 }, { "epoch": 1.831357048748353, "grad_norm": 0.5095801949501038, "learning_rate": 9.785480950320538e-05, "loss": 0.0204, "step": 1390 }, { "epoch": 1.8445322793148882, "grad_norm": 0.50572270154953, "learning_rate": 9.78066352637221e-05, "loss": 0.027, "step": 1400 }, { "epoch": 1.8577075098814229, "grad_norm": 0.3475058674812317, "learning_rate": 9.775793821853488e-05, "loss": 0.0239, "step": 1410 }, { "epoch": 1.870882740447958, "grad_norm": 0.44244056940078735, "learning_rate": 9.77087189001868e-05, "loss": 0.0289, "step": 1420 }, { "epoch": 1.8840579710144927, "grad_norm": 0.3293551206588745, "learning_rate": 9.765897784693243e-05, "loss": 0.023, "step": 1430 }, { "epoch": 1.8972332015810278, "grad_norm": 0.31059810519218445, "learning_rate": 9.760871560273197e-05, "loss": 0.0247, "step": 1440 }, { "epoch": 1.9104084321475625, "grad_norm": 0.396855890750885, "learning_rate": 9.755793271724526e-05, "loss": 0.0227, "step": 1450 }, { "epoch": 1.9235836627140976, "grad_norm": 0.3800758719444275, "learning_rate": 9.750662974582584e-05, "loss": 0.0265, "step": 1460 }, { "epoch": 1.9367588932806323, "grad_norm": 0.3450396656990051, "learning_rate": 9.745480724951473e-05, "loss": 0.0278, "step": 1470 }, { "epoch": 1.9499341238471675, "grad_norm": 0.5375139117240906, "learning_rate": 9.740246579503447e-05, "loss": 0.0291, "step": 1480 }, { "epoch": 1.9631093544137022, "grad_norm": 0.48629283905029297, "learning_rate": 9.734960595478284e-05, "loss": 0.0266, "step": 1490 }, { "epoch": 1.9762845849802373, "grad_norm": 0.5940545797348022, "learning_rate": 9.729622830682657e-05, "loss": 0.0309, "step": 1500 }, { "epoch": 1.989459815546772, "grad_norm": 0.4284568428993225, "learning_rate": 9.724233343489504e-05, "loss": 0.0303, "step": 1510 }, { "epoch": 2.002635046113307, "grad_norm": 0.649758517742157, "learning_rate": 9.718792192837396e-05, "loss": 0.0281, "step": 1520 }, { "epoch": 2.015810276679842, "grad_norm": 0.3681211471557617, "learning_rate": 9.713299438229886e-05, "loss": 0.0271, "step": 1530 }, { "epoch": 2.028985507246377, "grad_norm": 0.5985534191131592, "learning_rate": 9.707755139734855e-05, "loss": 0.0302, "step": 1540 }, { "epoch": 2.0421607378129116, "grad_norm": 0.6348494291305542, "learning_rate": 9.702159357983866e-05, "loss": 0.0358, "step": 1550 }, { "epoch": 2.0553359683794468, "grad_norm": 0.7188793420791626, "learning_rate": 9.696512154171492e-05, "loss": 0.0244, "step": 1560 }, { "epoch": 2.0685111989459815, "grad_norm": 0.7274630069732666, "learning_rate": 9.690813590054645e-05, "loss": 0.0271, "step": 1570 }, { "epoch": 2.0816864295125166, "grad_norm": 0.4827321171760559, "learning_rate": 9.685063727951914e-05, "loss": 0.0256, "step": 1580 }, { "epoch": 2.0948616600790513, "grad_norm": 0.539287805557251, "learning_rate": 9.679262630742865e-05, "loss": 0.0272, "step": 1590 }, { "epoch": 2.1080368906455864, "grad_norm": 0.34089940786361694, "learning_rate": 9.673410361867373e-05, "loss": 0.0238, "step": 1600 }, { "epoch": 2.121212121212121, "grad_norm": 0.32052725553512573, "learning_rate": 9.667506985324909e-05, "loss": 0.0208, "step": 1610 }, { "epoch": 2.1343873517786562, "grad_norm": 0.369957834482193, "learning_rate": 9.661552565673855e-05, "loss": 0.0214, "step": 1620 }, { "epoch": 2.147562582345191, "grad_norm": 0.26442891359329224, "learning_rate": 9.655547168030789e-05, "loss": 0.0195, "step": 1630 }, { "epoch": 2.160737812911726, "grad_norm": 0.47019681334495544, "learning_rate": 9.649490858069777e-05, "loss": 0.0221, "step": 1640 }, { "epoch": 2.1739130434782608, "grad_norm": 0.22871090471744537, "learning_rate": 9.643383702021658e-05, "loss": 0.0259, "step": 1650 }, { "epoch": 2.187088274044796, "grad_norm": 0.2812241315841675, "learning_rate": 9.637225766673307e-05, "loss": 0.0213, "step": 1660 }, { "epoch": 2.2002635046113306, "grad_norm": 0.472553014755249, "learning_rate": 9.631017119366922e-05, "loss": 0.0216, "step": 1670 }, { "epoch": 2.2134387351778657, "grad_norm": 0.4297308623790741, "learning_rate": 9.624757827999273e-05, "loss": 0.0239, "step": 1680 }, { "epoch": 2.2266139657444004, "grad_norm": 0.2847653329372406, "learning_rate": 9.618447961020971e-05, "loss": 0.0245, "step": 1690 }, { "epoch": 2.2397891963109355, "grad_norm": 0.3599100708961487, "learning_rate": 9.612087587435707e-05, "loss": 0.019, "step": 1700 }, { "epoch": 2.2529644268774702, "grad_norm": 0.306449830532074, "learning_rate": 9.605676776799508e-05, "loss": 0.0216, "step": 1710 }, { "epoch": 2.2661396574440054, "grad_norm": 0.40967902541160583, "learning_rate": 9.599215599219973e-05, "loss": 0.0235, "step": 1720 }, { "epoch": 2.27931488801054, "grad_norm": 0.35484758019447327, "learning_rate": 9.592704125355505e-05, "loss": 0.0218, "step": 1730 }, { "epoch": 2.292490118577075, "grad_norm": 0.3785801827907562, "learning_rate": 9.586142426414538e-05, "loss": 0.0239, "step": 1740 }, { "epoch": 2.30566534914361, "grad_norm": 0.24761465191841125, "learning_rate": 9.57953057415476e-05, "loss": 0.0244, "step": 1750 }, { "epoch": 2.318840579710145, "grad_norm": 0.22312359511852264, "learning_rate": 9.572868640882328e-05, "loss": 0.0266, "step": 1760 }, { "epoch": 2.3320158102766797, "grad_norm": 0.3827199339866638, "learning_rate": 9.56615669945108e-05, "loss": 0.0215, "step": 1770 }, { "epoch": 2.345191040843215, "grad_norm": 0.5303699374198914, "learning_rate": 9.55939482326173e-05, "loss": 0.0241, "step": 1780 }, { "epoch": 2.3583662714097495, "grad_norm": 0.2927263081073761, "learning_rate": 9.552583086261069e-05, "loss": 0.025, "step": 1790 }, { "epoch": 2.3715415019762847, "grad_norm": 0.36540526151657104, "learning_rate": 9.545721562941168e-05, "loss": 0.026, "step": 1800 }, { "epoch": 2.3847167325428194, "grad_norm": 0.3504578471183777, "learning_rate": 9.538810328338543e-05, "loss": 0.0204, "step": 1810 }, { "epoch": 2.3978919631093545, "grad_norm": 0.268880158662796, "learning_rate": 9.531849458033349e-05, "loss": 0.02, "step": 1820 }, { "epoch": 2.411067193675889, "grad_norm": 0.27113309502601624, "learning_rate": 9.524839028148547e-05, "loss": 0.0265, "step": 1830 }, { "epoch": 2.4242424242424243, "grad_norm": 0.2458828240633011, "learning_rate": 9.517779115349077e-05, "loss": 0.0217, "step": 1840 }, { "epoch": 2.437417654808959, "grad_norm": 0.41761359572410583, "learning_rate": 9.510669796841014e-05, "loss": 0.0258, "step": 1850 }, { "epoch": 2.450592885375494, "grad_norm": 0.6113036274909973, "learning_rate": 9.503511150370727e-05, "loss": 0.0241, "step": 1860 }, { "epoch": 2.463768115942029, "grad_norm": 0.4670635759830475, "learning_rate": 9.496303254224024e-05, "loss": 0.0293, "step": 1870 }, { "epoch": 2.476943346508564, "grad_norm": 0.39811426401138306, "learning_rate": 9.489046187225306e-05, "loss": 0.0241, "step": 1880 }, { "epoch": 2.4901185770750986, "grad_norm": 0.4884253442287445, "learning_rate": 9.481740028736692e-05, "loss": 0.0233, "step": 1890 }, { "epoch": 2.503293807641634, "grad_norm": 0.41597744822502136, "learning_rate": 9.474384858657164e-05, "loss": 0.0257, "step": 1900 }, { "epoch": 2.516469038208169, "grad_norm": 0.3447679877281189, "learning_rate": 9.466980757421679e-05, "loss": 0.0262, "step": 1910 }, { "epoch": 2.5296442687747036, "grad_norm": 0.3724343776702881, "learning_rate": 9.459527806000305e-05, "loss": 0.025, "step": 1920 }, { "epoch": 2.5428194993412383, "grad_norm": 0.5130154490470886, "learning_rate": 9.452026085897325e-05, "loss": 0.0268, "step": 1930 }, { "epoch": 2.5559947299077734, "grad_norm": 0.323124498128891, "learning_rate": 9.444475679150348e-05, "loss": 0.0225, "step": 1940 }, { "epoch": 2.5691699604743086, "grad_norm": 0.3052145838737488, "learning_rate": 9.436876668329411e-05, "loss": 0.0261, "step": 1950 }, { "epoch": 2.5823451910408433, "grad_norm": 0.424175500869751, "learning_rate": 9.429229136536079e-05, "loss": 0.0217, "step": 1960 }, { "epoch": 2.595520421607378, "grad_norm": 0.3193293809890747, "learning_rate": 9.421533167402534e-05, "loss": 0.0236, "step": 1970 }, { "epoch": 2.608695652173913, "grad_norm": 0.3281519412994385, "learning_rate": 9.413788845090666e-05, "loss": 0.024, "step": 1980 }, { "epoch": 2.621870882740448, "grad_norm": 0.5388944745063782, "learning_rate": 9.405996254291136e-05, "loss": 0.0254, "step": 1990 }, { "epoch": 2.635046113306983, "grad_norm": 0.43273690342903137, "learning_rate": 9.398155480222474e-05, "loss": 0.0216, "step": 2000 }, { "epoch": 2.6482213438735176, "grad_norm": 0.33787184953689575, "learning_rate": 9.390266608630128e-05, "loss": 0.0218, "step": 2010 }, { "epoch": 2.6613965744400527, "grad_norm": 0.4092657268047333, "learning_rate": 9.38232972578553e-05, "loss": 0.0205, "step": 2020 }, { "epoch": 2.674571805006588, "grad_norm": 0.3141351342201233, "learning_rate": 9.374344918485164e-05, "loss": 0.0303, "step": 2030 }, { "epoch": 2.6877470355731226, "grad_norm": 0.43514305353164673, "learning_rate": 9.366312274049602e-05, "loss": 0.0276, "step": 2040 }, { "epoch": 2.7009222661396572, "grad_norm": 0.5144362449645996, "learning_rate": 9.358231880322554e-05, "loss": 0.0245, "step": 2050 }, { "epoch": 2.7140974967061924, "grad_norm": 0.3669821619987488, "learning_rate": 9.350103825669916e-05, "loss": 0.0219, "step": 2060 }, { "epoch": 2.7272727272727275, "grad_norm": 0.5248842835426331, "learning_rate": 9.341928198978787e-05, "loss": 0.0194, "step": 2070 }, { "epoch": 2.740447957839262, "grad_norm": 0.24729865789413452, "learning_rate": 9.333705089656512e-05, "loss": 0.0218, "step": 2080 }, { "epoch": 2.753623188405797, "grad_norm": 0.2766522467136383, "learning_rate": 9.325434587629698e-05, "loss": 0.0165, "step": 2090 }, { "epoch": 2.766798418972332, "grad_norm": 0.3554515838623047, "learning_rate": 9.31711678334323e-05, "loss": 0.018, "step": 2100 }, { "epoch": 2.779973649538867, "grad_norm": 0.37791627645492554, "learning_rate": 9.308751767759282e-05, "loss": 0.0206, "step": 2110 }, { "epoch": 2.793148880105402, "grad_norm": 0.48156484961509705, "learning_rate": 9.300339632356325e-05, "loss": 0.0205, "step": 2120 }, { "epoch": 2.8063241106719365, "grad_norm": 0.21726833283901215, "learning_rate": 9.291880469128124e-05, "loss": 0.0235, "step": 2130 }, { "epoch": 2.8194993412384717, "grad_norm": 0.3837207555770874, "learning_rate": 9.283374370582732e-05, "loss": 0.0232, "step": 2140 }, { "epoch": 2.832674571805007, "grad_norm": 0.21415646374225616, "learning_rate": 9.274821429741482e-05, "loss": 0.0246, "step": 2150 }, { "epoch": 2.8458498023715415, "grad_norm": 0.2206660807132721, "learning_rate": 9.266221740137961e-05, "loss": 0.0225, "step": 2160 }, { "epoch": 2.859025032938076, "grad_norm": 0.30975693464279175, "learning_rate": 9.257575395817001e-05, "loss": 0.0192, "step": 2170 }, { "epoch": 2.8722002635046113, "grad_norm": 0.34590449929237366, "learning_rate": 9.248882491333637e-05, "loss": 0.023, "step": 2180 }, { "epoch": 2.8853754940711465, "grad_norm": 0.2704302966594696, "learning_rate": 9.240143121752076e-05, "loss": 0.0211, "step": 2190 }, { "epoch": 2.898550724637681, "grad_norm": 0.5347023606300354, "learning_rate": 9.23135738264467e-05, "loss": 0.0187, "step": 2200 }, { "epoch": 2.911725955204216, "grad_norm": 0.23387780785560608, "learning_rate": 9.222525370090849e-05, "loss": 0.02, "step": 2210 }, { "epoch": 2.924901185770751, "grad_norm": 0.285714328289032, "learning_rate": 9.213647180676088e-05, "loss": 0.0169, "step": 2220 }, { "epoch": 2.938076416337286, "grad_norm": 0.34807008504867554, "learning_rate": 9.204722911490846e-05, "loss": 0.0157, "step": 2230 }, { "epoch": 2.951251646903821, "grad_norm": 0.20837929844856262, "learning_rate": 9.1957526601295e-05, "loss": 0.02, "step": 2240 }, { "epoch": 2.9644268774703555, "grad_norm": 0.43575602769851685, "learning_rate": 9.186736524689281e-05, "loss": 0.0249, "step": 2250 }, { "epoch": 2.9776021080368906, "grad_norm": 0.3351643979549408, "learning_rate": 9.177674603769204e-05, "loss": 0.0228, "step": 2260 }, { "epoch": 2.9907773386034258, "grad_norm": 0.521143913269043, "learning_rate": 9.168566996468983e-05, "loss": 0.0232, "step": 2270 }, { "epoch": 3.0039525691699605, "grad_norm": 0.23659829795360565, "learning_rate": 9.159413802387951e-05, "loss": 0.0245, "step": 2280 }, { "epoch": 3.0171277997364956, "grad_norm": 0.3649868071079254, "learning_rate": 9.150215121623974e-05, "loss": 0.0238, "step": 2290 }, { "epoch": 3.0303030303030303, "grad_norm": 0.29252931475639343, "learning_rate": 9.140971054772349e-05, "loss": 0.0156, "step": 2300 }, { "epoch": 3.0434782608695654, "grad_norm": 0.27642548084259033, "learning_rate": 9.131681702924713e-05, "loss": 0.0211, "step": 2310 }, { "epoch": 3.0566534914361, "grad_norm": 0.1891625076532364, "learning_rate": 9.122347167667926e-05, "loss": 0.0217, "step": 2320 }, { "epoch": 3.0698287220026352, "grad_norm": 0.2757662236690521, "learning_rate": 9.112967551082973e-05, "loss": 0.0249, "step": 2330 }, { "epoch": 3.08300395256917, "grad_norm": 0.23346559703350067, "learning_rate": 9.103542955743835e-05, "loss": 0.0219, "step": 2340 }, { "epoch": 3.096179183135705, "grad_norm": 0.2930508852005005, "learning_rate": 9.094073484716381e-05, "loss": 0.0238, "step": 2350 }, { "epoch": 3.1093544137022397, "grad_norm": 0.25949984788894653, "learning_rate": 9.084559241557226e-05, "loss": 0.023, "step": 2360 }, { "epoch": 3.122529644268775, "grad_norm": 0.2743162512779236, "learning_rate": 9.075000330312608e-05, "loss": 0.0216, "step": 2370 }, { "epoch": 3.1357048748353096, "grad_norm": 0.3307958245277405, "learning_rate": 9.065396855517253e-05, "loss": 0.0167, "step": 2380 }, { "epoch": 3.1488801054018447, "grad_norm": 0.4071004390716553, "learning_rate": 9.055748922193219e-05, "loss": 0.0255, "step": 2390 }, { "epoch": 3.1620553359683794, "grad_norm": 0.2871227562427521, "learning_rate": 9.046056635848761e-05, "loss": 0.0247, "step": 2400 }, { "epoch": 3.1752305665349145, "grad_norm": 0.4296686351299286, "learning_rate": 9.036320102477169e-05, "loss": 0.0197, "step": 2410 }, { "epoch": 3.1884057971014492, "grad_norm": 0.2433115541934967, "learning_rate": 9.02653942855561e-05, "loss": 0.0207, "step": 2420 }, { "epoch": 3.2015810276679844, "grad_norm": 0.28873997926712036, "learning_rate": 9.016714721043971e-05, "loss": 0.0175, "step": 2430 }, { "epoch": 3.214756258234519, "grad_norm": 0.3427509069442749, "learning_rate": 9.006846087383675e-05, "loss": 0.0216, "step": 2440 }, { "epoch": 3.227931488801054, "grad_norm": 0.5131787061691284, "learning_rate": 8.996933635496523e-05, "loss": 0.0203, "step": 2450 }, { "epoch": 3.241106719367589, "grad_norm": 0.347695529460907, "learning_rate": 8.986977473783498e-05, "loss": 0.0217, "step": 2460 }, { "epoch": 3.254281949934124, "grad_norm": 0.3435944616794586, "learning_rate": 8.97697771112359e-05, "loss": 0.0194, "step": 2470 }, { "epoch": 3.2674571805006587, "grad_norm": 0.24768733978271484, "learning_rate": 8.966934456872602e-05, "loss": 0.0206, "step": 2480 }, { "epoch": 3.280632411067194, "grad_norm": 0.3580437898635864, "learning_rate": 8.95684782086195e-05, "loss": 0.0171, "step": 2490 }, { "epoch": 3.2938076416337285, "grad_norm": 0.38864248991012573, "learning_rate": 8.946717913397476e-05, "loss": 0.0223, "step": 2500 }, { "epoch": 3.3069828722002637, "grad_norm": 0.35056614875793457, "learning_rate": 8.93654484525822e-05, "loss": 0.0195, "step": 2510 }, { "epoch": 3.3201581027667983, "grad_norm": 0.1966569721698761, "learning_rate": 8.926328727695226e-05, "loss": 0.0192, "step": 2520 }, { "epoch": 3.3333333333333335, "grad_norm": 0.2948076128959656, "learning_rate": 8.916069672430319e-05, "loss": 0.0192, "step": 2530 }, { "epoch": 3.346508563899868, "grad_norm": 0.32685697078704834, "learning_rate": 8.905767791654884e-05, "loss": 0.0192, "step": 2540 }, { "epoch": 3.3596837944664033, "grad_norm": 0.26307734847068787, "learning_rate": 8.895423198028638e-05, "loss": 0.0237, "step": 2550 }, { "epoch": 3.372859025032938, "grad_norm": 0.27680397033691406, "learning_rate": 8.885036004678402e-05, "loss": 0.0212, "step": 2560 }, { "epoch": 3.386034255599473, "grad_norm": 0.2818705141544342, "learning_rate": 8.874606325196857e-05, "loss": 0.0168, "step": 2570 }, { "epoch": 3.399209486166008, "grad_norm": 0.2630791664123535, "learning_rate": 8.864134273641304e-05, "loss": 0.0223, "step": 2580 }, { "epoch": 3.412384716732543, "grad_norm": 0.28061074018478394, "learning_rate": 8.853619964532427e-05, "loss": 0.0219, "step": 2590 }, { "epoch": 3.4255599472990776, "grad_norm": 0.1784231811761856, "learning_rate": 8.843063512853019e-05, "loss": 0.0176, "step": 2600 }, { "epoch": 3.438735177865613, "grad_norm": 0.23167821764945984, "learning_rate": 8.832465034046749e-05, "loss": 0.0204, "step": 2610 }, { "epoch": 3.4519104084321475, "grad_norm": 0.25540560483932495, "learning_rate": 8.821824644016882e-05, "loss": 0.0185, "step": 2620 }, { "epoch": 3.4650856389986826, "grad_norm": 0.5046829581260681, "learning_rate": 8.811142459125019e-05, "loss": 0.0191, "step": 2630 }, { "epoch": 3.4782608695652173, "grad_norm": 0.45605048537254333, "learning_rate": 8.800418596189822e-05, "loss": 0.0186, "step": 2640 }, { "epoch": 3.4914361001317524, "grad_norm": 0.34779468178749084, "learning_rate": 8.789653172485737e-05, "loss": 0.0209, "step": 2650 }, { "epoch": 3.504611330698287, "grad_norm": 0.26932665705680847, "learning_rate": 8.778846305741715e-05, "loss": 0.0295, "step": 2660 }, { "epoch": 3.5177865612648223, "grad_norm": 0.2671760320663452, "learning_rate": 8.767998114139918e-05, "loss": 0.0227, "step": 2670 }, { "epoch": 3.530961791831357, "grad_norm": 0.5423169732093811, "learning_rate": 8.757108716314429e-05, "loss": 0.0211, "step": 2680 }, { "epoch": 3.544137022397892, "grad_norm": 0.603018045425415, "learning_rate": 8.746178231349962e-05, "loss": 0.0168, "step": 2690 }, { "epoch": 3.5573122529644268, "grad_norm": 0.33711379766464233, "learning_rate": 8.735206778780549e-05, "loss": 0.0231, "step": 2700 }, { "epoch": 3.570487483530962, "grad_norm": 0.4962400197982788, "learning_rate": 8.724194478588234e-05, "loss": 0.0233, "step": 2710 }, { "epoch": 3.5836627140974966, "grad_norm": 0.4521920680999756, "learning_rate": 8.713141451201772e-05, "loss": 0.0194, "step": 2720 }, { "epoch": 3.5968379446640317, "grad_norm": 0.5082101821899414, "learning_rate": 8.702047817495295e-05, "loss": 0.0207, "step": 2730 }, { "epoch": 3.6100131752305664, "grad_norm": 0.3494955003261566, "learning_rate": 8.69091369878701e-05, "loss": 0.0244, "step": 2740 }, { "epoch": 3.6231884057971016, "grad_norm": 0.359284371137619, "learning_rate": 8.679739216837849e-05, "loss": 0.0161, "step": 2750 }, { "epoch": 3.6363636363636362, "grad_norm": 0.4799230992794037, "learning_rate": 8.66852449385016e-05, "loss": 0.0178, "step": 2760 }, { "epoch": 3.6495388669301714, "grad_norm": 0.37354904413223267, "learning_rate": 8.657269652466356e-05, "loss": 0.0193, "step": 2770 }, { "epoch": 3.662714097496706, "grad_norm": 0.2906670570373535, "learning_rate": 8.645974815767577e-05, "loss": 0.0203, "step": 2780 }, { "epoch": 3.675889328063241, "grad_norm": 0.2389180213212967, "learning_rate": 8.634640107272351e-05, "loss": 0.0166, "step": 2790 }, { "epoch": 3.689064558629776, "grad_norm": 0.2554898262023926, "learning_rate": 8.623265650935234e-05, "loss": 0.0183, "step": 2800 }, { "epoch": 3.702239789196311, "grad_norm": 0.18219242990016937, "learning_rate": 8.611851571145456e-05, "loss": 0.0217, "step": 2810 }, { "epoch": 3.7154150197628457, "grad_norm": 0.292879194021225, "learning_rate": 8.600397992725566e-05, "loss": 0.0154, "step": 2820 }, { "epoch": 3.728590250329381, "grad_norm": 0.2756798565387726, "learning_rate": 8.588905040930061e-05, "loss": 0.019, "step": 2830 }, { "epoch": 3.7417654808959155, "grad_norm": 0.2696395218372345, "learning_rate": 8.577372841444022e-05, "loss": 0.0182, "step": 2840 }, { "epoch": 3.7549407114624507, "grad_norm": 0.27421027421951294, "learning_rate": 8.565801520381736e-05, "loss": 0.0177, "step": 2850 }, { "epoch": 3.7681159420289854, "grad_norm": 0.29462969303131104, "learning_rate": 8.554191204285313e-05, "loss": 0.0199, "step": 2860 }, { "epoch": 3.7812911725955205, "grad_norm": 0.39727547764778137, "learning_rate": 8.542542020123315e-05, "loss": 0.0193, "step": 2870 }, { "epoch": 3.794466403162055, "grad_norm": 0.3851800560951233, "learning_rate": 8.530854095289347e-05, "loss": 0.0196, "step": 2880 }, { "epoch": 3.8076416337285903, "grad_norm": 0.49293825030326843, "learning_rate": 8.519127557600688e-05, "loss": 0.0184, "step": 2890 }, { "epoch": 3.820816864295125, "grad_norm": 0.40799012780189514, "learning_rate": 8.507362535296871e-05, "loss": 0.0204, "step": 2900 }, { "epoch": 3.83399209486166, "grad_norm": 0.43597811460494995, "learning_rate": 8.495559157038299e-05, "loss": 0.0275, "step": 2910 }, { "epoch": 3.847167325428195, "grad_norm": 0.35586801171302795, "learning_rate": 8.483717551904823e-05, "loss": 0.0205, "step": 2920 }, { "epoch": 3.86034255599473, "grad_norm": 0.4549504816532135, "learning_rate": 8.47183784939434e-05, "loss": 0.0209, "step": 2930 }, { "epoch": 3.8735177865612647, "grad_norm": 0.4210937023162842, "learning_rate": 8.459920179421374e-05, "loss": 0.0212, "step": 2940 }, { "epoch": 3.8866930171278, "grad_norm": 0.30184030532836914, "learning_rate": 8.447964672315656e-05, "loss": 0.0186, "step": 2950 }, { "epoch": 3.8998682476943345, "grad_norm": 0.4085947275161743, "learning_rate": 8.435971458820692e-05, "loss": 0.021, "step": 2960 }, { "epoch": 3.9130434782608696, "grad_norm": 0.44483330845832825, "learning_rate": 8.423940670092345e-05, "loss": 0.0196, "step": 2970 }, { "epoch": 3.9262187088274043, "grad_norm": 0.3594276010990143, "learning_rate": 8.411872437697394e-05, "loss": 0.019, "step": 2980 }, { "epoch": 3.9393939393939394, "grad_norm": 0.33870217204093933, "learning_rate": 8.399766893612096e-05, "loss": 0.0172, "step": 2990 }, { "epoch": 3.9525691699604746, "grad_norm": 0.28566181659698486, "learning_rate": 8.38762417022074e-05, "loss": 0.0171, "step": 3000 }, { "epoch": 3.9657444005270093, "grad_norm": 0.18129827082157135, "learning_rate": 8.375444400314204e-05, "loss": 0.0182, "step": 3010 }, { "epoch": 3.978919631093544, "grad_norm": 0.471519410610199, "learning_rate": 8.3632277170885e-05, "loss": 0.0238, "step": 3020 }, { "epoch": 3.992094861660079, "grad_norm": 0.2733546793460846, "learning_rate": 8.350974254143318e-05, "loss": 0.0182, "step": 3030 }, { "epoch": 4.005270092226614, "grad_norm": 0.2888932526111603, "learning_rate": 8.338684145480566e-05, "loss": 0.0157, "step": 3040 }, { "epoch": 4.0184453227931485, "grad_norm": 0.31720566749572754, "learning_rate": 8.326357525502904e-05, "loss": 0.0175, "step": 3050 }, { "epoch": 4.031620553359684, "grad_norm": 0.2824496030807495, "learning_rate": 8.313994529012273e-05, "loss": 0.0202, "step": 3060 }, { "epoch": 4.044795783926219, "grad_norm": 0.32166388630867004, "learning_rate": 8.301595291208422e-05, "loss": 0.0225, "step": 3070 }, { "epoch": 4.057971014492754, "grad_norm": 0.5426804423332214, "learning_rate": 8.289159947687427e-05, "loss": 0.016, "step": 3080 }, { "epoch": 4.071146245059288, "grad_norm": 0.3431253433227539, "learning_rate": 8.276688634440216e-05, "loss": 0.0178, "step": 3090 }, { "epoch": 4.084321475625823, "grad_norm": 0.32741838693618774, "learning_rate": 8.26418148785107e-05, "loss": 0.0164, "step": 3100 }, { "epoch": 4.097496706192358, "grad_norm": 0.2607785165309906, "learning_rate": 8.251638644696141e-05, "loss": 0.018, "step": 3110 }, { "epoch": 4.1106719367588935, "grad_norm": 0.31585800647735596, "learning_rate": 8.23906024214195e-05, "loss": 0.0246, "step": 3120 }, { "epoch": 4.123847167325428, "grad_norm": 0.6056305170059204, "learning_rate": 8.226446417743897e-05, "loss": 0.0205, "step": 3130 }, { "epoch": 4.137022397891963, "grad_norm": 0.39446625113487244, "learning_rate": 8.213797309444742e-05, "loss": 0.0243, "step": 3140 }, { "epoch": 4.150197628458498, "grad_norm": 0.23034586012363434, "learning_rate": 8.201113055573105e-05, "loss": 0.019, "step": 3150 }, { "epoch": 4.163372859025033, "grad_norm": 0.28100505471229553, "learning_rate": 8.188393794841958e-05, "loss": 0.0142, "step": 3160 }, { "epoch": 4.176548089591567, "grad_norm": 0.19925159215927124, "learning_rate": 8.175639666347094e-05, "loss": 0.0208, "step": 3170 }, { "epoch": 4.189723320158103, "grad_norm": 0.4441159963607788, "learning_rate": 8.162850809565623e-05, "loss": 0.0176, "step": 3180 }, { "epoch": 4.202898550724638, "grad_norm": 0.2651693820953369, "learning_rate": 8.150027364354431e-05, "loss": 0.0187, "step": 3190 }, { "epoch": 4.216073781291173, "grad_norm": 0.2650001645088196, "learning_rate": 8.137169470948662e-05, "loss": 0.0219, "step": 3200 }, { "epoch": 4.229249011857707, "grad_norm": 0.17157067358493805, "learning_rate": 8.124277269960179e-05, "loss": 0.0173, "step": 3210 }, { "epoch": 4.242424242424242, "grad_norm": 0.3844333291053772, "learning_rate": 8.111350902376023e-05, "loss": 0.0172, "step": 3220 }, { "epoch": 4.255599472990777, "grad_norm": 0.29438167810440063, "learning_rate": 8.098390509556883e-05, "loss": 0.0196, "step": 3230 }, { "epoch": 4.2687747035573125, "grad_norm": 0.2370031476020813, "learning_rate": 8.085396233235536e-05, "loss": 0.0157, "step": 3240 }, { "epoch": 4.281949934123848, "grad_norm": 0.233305886387825, "learning_rate": 8.072368215515306e-05, "loss": 0.0156, "step": 3250 }, { "epoch": 4.295125164690382, "grad_norm": 0.27094778418540955, "learning_rate": 8.059306598868506e-05, "loss": 0.0196, "step": 3260 }, { "epoch": 4.308300395256917, "grad_norm": 0.25244513154029846, "learning_rate": 8.046211526134888e-05, "loss": 0.0147, "step": 3270 }, { "epoch": 4.321475625823452, "grad_norm": 0.26508960127830505, "learning_rate": 8.033083140520065e-05, "loss": 0.0155, "step": 3280 }, { "epoch": 4.334650856389986, "grad_norm": 0.24337764084339142, "learning_rate": 8.019921585593962e-05, "loss": 0.0177, "step": 3290 }, { "epoch": 4.3478260869565215, "grad_norm": 0.23213884234428406, "learning_rate": 8.006727005289232e-05, "loss": 0.0153, "step": 3300 }, { "epoch": 4.361001317523057, "grad_norm": 0.2334267944097519, "learning_rate": 7.993499543899692e-05, "loss": 0.0142, "step": 3310 }, { "epoch": 4.374176548089592, "grad_norm": 0.21493524312973022, "learning_rate": 7.980239346078742e-05, "loss": 0.017, "step": 3320 }, { "epoch": 4.387351778656127, "grad_norm": 0.23911434412002563, "learning_rate": 7.966946556837778e-05, "loss": 0.0143, "step": 3330 }, { "epoch": 4.400527009222661, "grad_norm": 0.285081684589386, "learning_rate": 7.953621321544616e-05, "loss": 0.0193, "step": 3340 }, { "epoch": 4.413702239789196, "grad_norm": 0.34651780128479004, "learning_rate": 7.940263785921896e-05, "loss": 0.0202, "step": 3350 }, { "epoch": 4.426877470355731, "grad_norm": 0.3821335732936859, "learning_rate": 7.926874096045482e-05, "loss": 0.0193, "step": 3360 }, { "epoch": 4.440052700922266, "grad_norm": 0.28446274995803833, "learning_rate": 7.913452398342881e-05, "loss": 0.0227, "step": 3370 }, { "epoch": 4.453227931488801, "grad_norm": 0.3928273916244507, "learning_rate": 7.89999883959163e-05, "loss": 0.0208, "step": 3380 }, { "epoch": 4.466403162055336, "grad_norm": 0.5543475151062012, "learning_rate": 7.886513566917687e-05, "loss": 0.0231, "step": 3390 }, { "epoch": 4.479578392621871, "grad_norm": 0.2991754710674286, "learning_rate": 7.872996727793838e-05, "loss": 0.0222, "step": 3400 }, { "epoch": 4.492753623188406, "grad_norm": 0.5026523470878601, "learning_rate": 7.859448470038069e-05, "loss": 0.0234, "step": 3410 }, { "epoch": 4.5059288537549405, "grad_norm": 0.296000599861145, "learning_rate": 7.845868941811956e-05, "loss": 0.0193, "step": 3420 }, { "epoch": 4.519104084321476, "grad_norm": 0.26779842376708984, "learning_rate": 7.832258291619043e-05, "loss": 0.0169, "step": 3430 }, { "epoch": 4.532279314888011, "grad_norm": 0.22933119535446167, "learning_rate": 7.81861666830322e-05, "loss": 0.0181, "step": 3440 }, { "epoch": 4.545454545454545, "grad_norm": 0.30201396346092224, "learning_rate": 7.804944221047097e-05, "loss": 0.0204, "step": 3450 }, { "epoch": 4.55862977602108, "grad_norm": 0.23445376753807068, "learning_rate": 7.791241099370364e-05, "loss": 0.0196, "step": 3460 }, { "epoch": 4.571805006587615, "grad_norm": 0.24794942140579224, "learning_rate": 7.777507453128163e-05, "loss": 0.0184, "step": 3470 }, { "epoch": 4.58498023715415, "grad_norm": 0.3686930537223816, "learning_rate": 7.763743432509451e-05, "loss": 0.0152, "step": 3480 }, { "epoch": 4.5981554677206855, "grad_norm": 0.6352735161781311, "learning_rate": 7.749949188035353e-05, "loss": 0.0173, "step": 3490 }, { "epoch": 4.61133069828722, "grad_norm": 0.3366442322731018, "learning_rate": 7.736124870557516e-05, "loss": 0.0245, "step": 3500 }, { "epoch": 4.624505928853755, "grad_norm": 0.3499387800693512, "learning_rate": 7.722270631256459e-05, "loss": 0.0236, "step": 3510 }, { "epoch": 4.63768115942029, "grad_norm": 0.23462465405464172, "learning_rate": 7.708386621639925e-05, "loss": 0.017, "step": 3520 }, { "epoch": 4.650856389986825, "grad_norm": 0.37723302841186523, "learning_rate": 7.694472993541219e-05, "loss": 0.0185, "step": 3530 }, { "epoch": 4.664031620553359, "grad_norm": 0.27838173508644104, "learning_rate": 7.680529899117547e-05, "loss": 0.0195, "step": 3540 }, { "epoch": 4.6772068511198945, "grad_norm": 0.17701946198940277, "learning_rate": 7.666557490848358e-05, "loss": 0.0197, "step": 3550 }, { "epoch": 4.69038208168643, "grad_norm": 0.3603031635284424, "learning_rate": 7.65255592153367e-05, "loss": 0.0207, "step": 3560 }, { "epoch": 4.703557312252965, "grad_norm": 0.37775689363479614, "learning_rate": 7.638525344292402e-05, "loss": 0.019, "step": 3570 }, { "epoch": 4.716732542819499, "grad_norm": 0.3618902862071991, "learning_rate": 7.624465912560697e-05, "loss": 0.0172, "step": 3580 }, { "epoch": 4.729907773386034, "grad_norm": 0.2325575351715088, "learning_rate": 7.610377780090249e-05, "loss": 0.0218, "step": 3590 }, { "epoch": 4.743083003952569, "grad_norm": 0.39902588725090027, "learning_rate": 7.596261100946618e-05, "loss": 0.0185, "step": 3600 }, { "epoch": 4.7562582345191045, "grad_norm": 0.41432973742485046, "learning_rate": 7.582116029507542e-05, "loss": 0.0182, "step": 3610 }, { "epoch": 4.769433465085639, "grad_norm": 0.3391708731651306, "learning_rate": 7.56794272046126e-05, "loss": 0.018, "step": 3620 }, { "epoch": 4.782608695652174, "grad_norm": 0.2337019294500351, "learning_rate": 7.55374132880481e-05, "loss": 0.017, "step": 3630 }, { "epoch": 4.795783926218709, "grad_norm": 0.4126136004924774, "learning_rate": 7.539512009842333e-05, "loss": 0.0186, "step": 3640 }, { "epoch": 4.808959156785244, "grad_norm": 0.3636666536331177, "learning_rate": 7.525254919183382e-05, "loss": 0.0204, "step": 3650 }, { "epoch": 4.822134387351778, "grad_norm": 0.34775424003601074, "learning_rate": 7.510970212741215e-05, "loss": 0.0185, "step": 3660 }, { "epoch": 4.8353096179183135, "grad_norm": 0.37673404812812805, "learning_rate": 7.496658046731096e-05, "loss": 0.0202, "step": 3670 }, { "epoch": 4.848484848484849, "grad_norm": 0.42134353518486023, "learning_rate": 7.482318577668578e-05, "loss": 0.0191, "step": 3680 }, { "epoch": 4.861660079051384, "grad_norm": 0.28770652413368225, "learning_rate": 7.467951962367796e-05, "loss": 0.0224, "step": 3690 }, { "epoch": 4.874835309617918, "grad_norm": 0.27064254879951477, "learning_rate": 7.453558357939755e-05, "loss": 0.0186, "step": 3700 }, { "epoch": 4.888010540184453, "grad_norm": 0.2361055612564087, "learning_rate": 7.439137921790606e-05, "loss": 0.0189, "step": 3710 }, { "epoch": 4.901185770750988, "grad_norm": 0.3993823826313019, "learning_rate": 7.42469081161993e-05, "loss": 0.0242, "step": 3720 }, { "epoch": 4.914361001317523, "grad_norm": 0.483935683965683, "learning_rate": 7.410217185419006e-05, "loss": 0.0173, "step": 3730 }, { "epoch": 4.927536231884058, "grad_norm": 0.24301816523075104, "learning_rate": 7.395717201469095e-05, "loss": 0.0172, "step": 3740 }, { "epoch": 4.940711462450593, "grad_norm": 0.323569118976593, "learning_rate": 7.381191018339696e-05, "loss": 0.0193, "step": 3750 }, { "epoch": 4.953886693017128, "grad_norm": 0.21056847274303436, "learning_rate": 7.36663879488682e-05, "loss": 0.0164, "step": 3760 }, { "epoch": 4.967061923583663, "grad_norm": 0.29462572932243347, "learning_rate": 7.352060690251254e-05, "loss": 0.021, "step": 3770 }, { "epoch": 4.980237154150197, "grad_norm": 0.23086123168468475, "learning_rate": 7.337456863856811e-05, "loss": 0.0174, "step": 3780 }, { "epoch": 4.993412384716732, "grad_norm": 0.254380464553833, "learning_rate": 7.3228274754086e-05, "loss": 0.0177, "step": 3790 }, { "epoch": 5.006587615283268, "grad_norm": 0.4429381489753723, "learning_rate": 7.308172684891267e-05, "loss": 0.0169, "step": 3800 }, { "epoch": 5.019762845849803, "grad_norm": 0.24621398746967316, "learning_rate": 7.293492652567255e-05, "loss": 0.0161, "step": 3810 }, { "epoch": 5.032938076416337, "grad_norm": 0.250224232673645, "learning_rate": 7.278787538975043e-05, "loss": 0.0214, "step": 3820 }, { "epoch": 5.046113306982872, "grad_norm": 0.3510632812976837, "learning_rate": 7.2640575049274e-05, "loss": 0.0209, "step": 3830 }, { "epoch": 5.059288537549407, "grad_norm": 0.2965291440486908, "learning_rate": 7.249302711509616e-05, "loss": 0.0156, "step": 3840 }, { "epoch": 5.072463768115942, "grad_norm": 0.25247883796691895, "learning_rate": 7.23452332007775e-05, "loss": 0.0204, "step": 3850 }, { "epoch": 5.085638998682477, "grad_norm": 0.36644870042800903, "learning_rate": 7.219719492256858e-05, "loss": 0.0198, "step": 3860 }, { "epoch": 5.098814229249012, "grad_norm": 0.20640870928764343, "learning_rate": 7.20489138993923e-05, "loss": 0.0208, "step": 3870 }, { "epoch": 5.111989459815547, "grad_norm": 0.26875925064086914, "learning_rate": 7.190039175282614e-05, "loss": 0.0185, "step": 3880 }, { "epoch": 5.125164690382082, "grad_norm": 0.2605835497379303, "learning_rate": 7.175163010708455e-05, "loss": 0.0152, "step": 3890 }, { "epoch": 5.138339920948616, "grad_norm": 0.24679867923259735, "learning_rate": 7.1602630589001e-05, "loss": 0.019, "step": 3900 }, { "epoch": 5.151515151515151, "grad_norm": 0.30721744894981384, "learning_rate": 7.14533948280104e-05, "loss": 0.0163, "step": 3910 }, { "epoch": 5.1646903820816865, "grad_norm": 0.20572873950004578, "learning_rate": 7.130392445613109e-05, "loss": 0.0182, "step": 3920 }, { "epoch": 5.177865612648222, "grad_norm": 0.28250861167907715, "learning_rate": 7.115422110794711e-05, "loss": 0.0178, "step": 3930 }, { "epoch": 5.191040843214756, "grad_norm": 0.2272404581308365, "learning_rate": 7.100428642059033e-05, "loss": 0.0178, "step": 3940 }, { "epoch": 5.204216073781291, "grad_norm": 0.3031019866466522, "learning_rate": 7.08541220337224e-05, "loss": 0.0215, "step": 3950 }, { "epoch": 5.217391304347826, "grad_norm": 0.30057474970817566, "learning_rate": 7.070372958951706e-05, "loss": 0.0207, "step": 3960 }, { "epoch": 5.230566534914361, "grad_norm": 0.5263017416000366, "learning_rate": 7.055311073264194e-05, "loss": 0.0192, "step": 3970 }, { "epoch": 5.2437417654808955, "grad_norm": 0.3072521686553955, "learning_rate": 7.040226711024077e-05, "loss": 0.0179, "step": 3980 }, { "epoch": 5.256916996047431, "grad_norm": 0.3801591992378235, "learning_rate": 7.02512003719152e-05, "loss": 0.0173, "step": 3990 }, { "epoch": 5.270092226613966, "grad_norm": 0.27008554339408875, "learning_rate": 7.00999121697069e-05, "loss": 0.0138, "step": 4000 }, { "epoch": 5.283267457180501, "grad_norm": 0.23966392874717712, "learning_rate": 6.99484041580794e-05, "loss": 0.0159, "step": 4010 }, { "epoch": 5.296442687747035, "grad_norm": 0.2968190610408783, "learning_rate": 6.979667799390004e-05, "loss": 0.0163, "step": 4020 }, { "epoch": 5.30961791831357, "grad_norm": 0.24101383984088898, "learning_rate": 6.964473533642185e-05, "loss": 0.0154, "step": 4030 }, { "epoch": 5.3227931488801055, "grad_norm": 0.418264240026474, "learning_rate": 6.949257784726539e-05, "loss": 0.0181, "step": 4040 }, { "epoch": 5.335968379446641, "grad_norm": 0.3220149576663971, "learning_rate": 6.934020719040056e-05, "loss": 0.014, "step": 4050 }, { "epoch": 5.349143610013175, "grad_norm": 0.2802201211452484, "learning_rate": 6.918762503212848e-05, "loss": 0.0169, "step": 4060 }, { "epoch": 5.36231884057971, "grad_norm": 0.2993554174900055, "learning_rate": 6.903483304106319e-05, "loss": 0.0214, "step": 4070 }, { "epoch": 5.375494071146245, "grad_norm": 0.30128803849220276, "learning_rate": 6.888183288811341e-05, "loss": 0.0174, "step": 4080 }, { "epoch": 5.38866930171278, "grad_norm": 0.33825552463531494, "learning_rate": 6.87286262464643e-05, "loss": 0.0171, "step": 4090 }, { "epoch": 5.4018445322793145, "grad_norm": 0.31894010305404663, "learning_rate": 6.857521479155915e-05, "loss": 0.0165, "step": 4100 }, { "epoch": 5.41501976284585, "grad_norm": 0.289244681596756, "learning_rate": 6.842160020108104e-05, "loss": 0.0175, "step": 4110 }, { "epoch": 5.428194993412385, "grad_norm": 0.2629208266735077, "learning_rate": 6.826778415493455e-05, "loss": 0.017, "step": 4120 }, { "epoch": 5.44137022397892, "grad_norm": 0.24454976618289948, "learning_rate": 6.811376833522729e-05, "loss": 0.0174, "step": 4130 }, { "epoch": 5.454545454545454, "grad_norm": 0.27494749426841736, "learning_rate": 6.795955442625159e-05, "loss": 0.0183, "step": 4140 }, { "epoch": 5.467720685111989, "grad_norm": 0.2748429775238037, "learning_rate": 6.780514411446608e-05, "loss": 0.0186, "step": 4150 }, { "epoch": 5.480895915678524, "grad_norm": 0.26624834537506104, "learning_rate": 6.765053908847716e-05, "loss": 0.013, "step": 4160 }, { "epoch": 5.4940711462450595, "grad_norm": 0.41027799248695374, "learning_rate": 6.749574103902064e-05, "loss": 0.0176, "step": 4170 }, { "epoch": 5.507246376811594, "grad_norm": 0.27950572967529297, "learning_rate": 6.734075165894317e-05, "loss": 0.0217, "step": 4180 }, { "epoch": 5.520421607378129, "grad_norm": 0.32828864455223083, "learning_rate": 6.71855726431838e-05, "loss": 0.0212, "step": 4190 }, { "epoch": 5.533596837944664, "grad_norm": 0.32206991314888, "learning_rate": 6.703020568875538e-05, "loss": 0.0208, "step": 4200 }, { "epoch": 5.546772068511199, "grad_norm": 0.2994462549686432, "learning_rate": 6.687465249472603e-05, "loss": 0.0169, "step": 4210 }, { "epoch": 5.559947299077734, "grad_norm": 0.1446998417377472, "learning_rate": 6.671891476220055e-05, "loss": 0.0163, "step": 4220 }, { "epoch": 5.573122529644269, "grad_norm": 0.2796706259250641, "learning_rate": 6.656299419430183e-05, "loss": 0.014, "step": 4230 }, { "epoch": 5.586297760210804, "grad_norm": 0.2798866033554077, "learning_rate": 6.640689249615223e-05, "loss": 0.0159, "step": 4240 }, { "epoch": 5.599472990777339, "grad_norm": 0.3122697174549103, "learning_rate": 6.625061137485491e-05, "loss": 0.0216, "step": 4250 }, { "epoch": 5.612648221343873, "grad_norm": 0.23494744300842285, "learning_rate": 6.609415253947517e-05, "loss": 0.0215, "step": 4260 }, { "epoch": 5.625823451910408, "grad_norm": 0.218988299369812, "learning_rate": 6.593751770102178e-05, "loss": 0.0158, "step": 4270 }, { "epoch": 5.638998682476943, "grad_norm": 0.2929661273956299, "learning_rate": 6.578070857242823e-05, "loss": 0.018, "step": 4280 }, { "epoch": 5.6521739130434785, "grad_norm": 0.26279154419898987, "learning_rate": 6.562372686853402e-05, "loss": 0.0158, "step": 4290 }, { "epoch": 5.665349143610014, "grad_norm": 0.35149410367012024, "learning_rate": 6.546657430606593e-05, "loss": 0.0186, "step": 4300 }, { "epoch": 5.678524374176548, "grad_norm": 0.26635152101516724, "learning_rate": 6.530925260361918e-05, "loss": 0.016, "step": 4310 }, { "epoch": 5.691699604743083, "grad_norm": 0.25784632563591003, "learning_rate": 6.515176348163871e-05, "loss": 0.0183, "step": 4320 }, { "epoch": 5.704874835309618, "grad_norm": 0.34484511613845825, "learning_rate": 6.499410866240032e-05, "loss": 0.0151, "step": 4330 }, { "epoch": 5.718050065876152, "grad_norm": 0.29317110776901245, "learning_rate": 6.48362898699919e-05, "loss": 0.0139, "step": 4340 }, { "epoch": 5.7312252964426875, "grad_norm": 0.35007333755493164, "learning_rate": 6.467830883029443e-05, "loss": 0.0218, "step": 4350 }, { "epoch": 5.744400527009223, "grad_norm": 0.17053362727165222, "learning_rate": 6.452016727096326e-05, "loss": 0.0161, "step": 4360 }, { "epoch": 5.757575757575758, "grad_norm": 0.24661651253700256, "learning_rate": 6.436186692140916e-05, "loss": 0.0166, "step": 4370 }, { "epoch": 5.770750988142293, "grad_norm": 0.3710819184780121, "learning_rate": 6.420340951277938e-05, "loss": 0.0161, "step": 4380 }, { "epoch": 5.783926218708827, "grad_norm": 0.2797921299934387, "learning_rate": 6.404479677793874e-05, "loss": 0.016, "step": 4390 }, { "epoch": 5.797101449275362, "grad_norm": 0.5894481539726257, "learning_rate": 6.388603045145075e-05, "loss": 0.0219, "step": 4400 }, { "epoch": 5.810276679841897, "grad_norm": 0.24950364232063293, "learning_rate": 6.372711226955843e-05, "loss": 0.0193, "step": 4410 }, { "epoch": 5.823451910408432, "grad_norm": 0.38150009512901306, "learning_rate": 6.356804397016564e-05, "loss": 0.0185, "step": 4420 }, { "epoch": 5.836627140974967, "grad_norm": 0.22015729546546936, "learning_rate": 6.340882729281779e-05, "loss": 0.0168, "step": 4430 }, { "epoch": 5.849802371541502, "grad_norm": 0.2161886841058731, "learning_rate": 6.324946397868294e-05, "loss": 0.019, "step": 4440 }, { "epoch": 5.862977602108037, "grad_norm": 0.31231123208999634, "learning_rate": 6.308995577053276e-05, "loss": 0.0197, "step": 4450 }, { "epoch": 5.876152832674572, "grad_norm": 0.19847756624221802, "learning_rate": 6.293030441272347e-05, "loss": 0.0159, "step": 4460 }, { "epoch": 5.8893280632411065, "grad_norm": 0.18192614614963531, "learning_rate": 6.277051165117677e-05, "loss": 0.0131, "step": 4470 }, { "epoch": 5.902503293807642, "grad_norm": 0.25896963477134705, "learning_rate": 6.261057923336064e-05, "loss": 0.0147, "step": 4480 }, { "epoch": 5.915678524374177, "grad_norm": 0.2821304202079773, "learning_rate": 6.245050890827042e-05, "loss": 0.0178, "step": 4490 }, { "epoch": 5.928853754940711, "grad_norm": 0.24604766070842743, "learning_rate": 6.229030242640952e-05, "loss": 0.0174, "step": 4500 }, { "epoch": 5.942028985507246, "grad_norm": 0.23723308742046356, "learning_rate": 6.212996153977037e-05, "loss": 0.0197, "step": 4510 }, { "epoch": 5.955204216073781, "grad_norm": 0.3339266777038574, "learning_rate": 6.196948800181523e-05, "loss": 0.0164, "step": 4520 }, { "epoch": 5.968379446640316, "grad_norm": 0.3604557514190674, "learning_rate": 6.180888356745695e-05, "loss": 0.0164, "step": 4530 }, { "epoch": 5.9815546772068515, "grad_norm": 0.398492693901062, "learning_rate": 6.164814999303995e-05, "loss": 0.0152, "step": 4540 }, { "epoch": 5.994729907773386, "grad_norm": 0.23473089933395386, "learning_rate": 6.148728903632081e-05, "loss": 0.0136, "step": 4550 }, { "epoch": 6.007905138339921, "grad_norm": 0.4985441565513611, "learning_rate": 6.132630245644921e-05, "loss": 0.0174, "step": 4560 }, { "epoch": 6.021080368906456, "grad_norm": 0.2359652817249298, "learning_rate": 6.116519201394857e-05, "loss": 0.0203, "step": 4570 }, { "epoch": 6.034255599472991, "grad_norm": 0.24522265791893005, "learning_rate": 6.10039594706969e-05, "loss": 0.0184, "step": 4580 }, { "epoch": 6.047430830039525, "grad_norm": 0.3857898414134979, "learning_rate": 6.084260658990744e-05, "loss": 0.0173, "step": 4590 }, { "epoch": 6.0606060606060606, "grad_norm": 0.26704806089401245, "learning_rate": 6.068113513610943e-05, "loss": 0.0151, "step": 4600 }, { "epoch": 6.073781291172596, "grad_norm": 0.7608732581138611, "learning_rate": 6.0519546875128876e-05, "loss": 0.0176, "step": 4610 }, { "epoch": 6.086956521739131, "grad_norm": 0.4076343774795532, "learning_rate": 6.035784357406906e-05, "loss": 0.0221, "step": 4620 }, { "epoch": 6.100131752305665, "grad_norm": 0.18973155319690704, "learning_rate": 6.01960270012914e-05, "loss": 0.0186, "step": 4630 }, { "epoch": 6.1133069828722, "grad_norm": 0.2442351132631302, "learning_rate": 6.003409892639599e-05, "loss": 0.0213, "step": 4640 }, { "epoch": 6.126482213438735, "grad_norm": 0.22191573679447174, "learning_rate": 5.9872061120202336e-05, "loss": 0.0191, "step": 4650 }, { "epoch": 6.1396574440052705, "grad_norm": 0.23175200819969177, "learning_rate": 5.9709915354729914e-05, "loss": 0.0248, "step": 4660 }, { "epoch": 6.152832674571805, "grad_norm": 0.42165979743003845, "learning_rate": 5.9547663403178824e-05, "loss": 0.0205, "step": 4670 }, { "epoch": 6.16600790513834, "grad_norm": 0.34937527775764465, "learning_rate": 5.9385307039910445e-05, "loss": 0.0158, "step": 4680 }, { "epoch": 6.179183135704875, "grad_norm": 0.21536408364772797, "learning_rate": 5.922284804042792e-05, "loss": 0.0186, "step": 4690 }, { "epoch": 6.19235836627141, "grad_norm": 0.20470696687698364, "learning_rate": 5.906028818135687e-05, "loss": 0.0204, "step": 4700 }, { "epoch": 6.205533596837944, "grad_norm": 0.35827165842056274, "learning_rate": 5.889762924042585e-05, "loss": 0.0155, "step": 4710 }, { "epoch": 6.2187088274044795, "grad_norm": 0.42314767837524414, "learning_rate": 5.873487299644699e-05, "loss": 0.019, "step": 4720 }, { "epoch": 6.231884057971015, "grad_norm": 0.3023538589477539, "learning_rate": 5.857202122929649e-05, "loss": 0.0199, "step": 4730 }, { "epoch": 6.24505928853755, "grad_norm": 0.3448214828968048, "learning_rate": 5.840907571989518e-05, "loss": 0.0133, "step": 4740 }, { "epoch": 6.258234519104084, "grad_norm": 0.1906011402606964, "learning_rate": 5.824603825018904e-05, "loss": 0.0157, "step": 4750 }, { "epoch": 6.271409749670619, "grad_norm": 0.21655221283435822, "learning_rate": 5.808291060312975e-05, "loss": 0.0151, "step": 4760 }, { "epoch": 6.284584980237154, "grad_norm": 0.26237595081329346, "learning_rate": 5.7919694562655083e-05, "loss": 0.0151, "step": 4770 }, { "epoch": 6.297760210803689, "grad_norm": 0.29299458861351013, "learning_rate": 5.775639191366954e-05, "loss": 0.0164, "step": 4780 }, { "epoch": 6.310935441370224, "grad_norm": 0.24668575823307037, "learning_rate": 5.75930044420247e-05, "loss": 0.0156, "step": 4790 }, { "epoch": 6.324110671936759, "grad_norm": 0.39310356974601746, "learning_rate": 5.74295339344998e-05, "loss": 0.0132, "step": 4800 }, { "epoch": 6.337285902503294, "grad_norm": 0.2791636884212494, "learning_rate": 5.726598217878211e-05, "loss": 0.0204, "step": 4810 }, { "epoch": 6.350461133069829, "grad_norm": 0.2549298405647278, "learning_rate": 5.71023509634474e-05, "loss": 0.0134, "step": 4820 }, { "epoch": 6.363636363636363, "grad_norm": 0.21396104991436005, "learning_rate": 5.693864207794049e-05, "loss": 0.0156, "step": 4830 }, { "epoch": 6.3768115942028984, "grad_norm": 0.2730461359024048, "learning_rate": 5.677485731255545e-05, "loss": 0.0258, "step": 4840 }, { "epoch": 6.389986824769434, "grad_norm": 0.1767221987247467, "learning_rate": 5.6610998458416296e-05, "loss": 0.0177, "step": 4850 }, { "epoch": 6.403162055335969, "grad_norm": 0.32627832889556885, "learning_rate": 5.644706730745716e-05, "loss": 0.0136, "step": 4860 }, { "epoch": 6.416337285902503, "grad_norm": 0.1690535843372345, "learning_rate": 5.628306565240287e-05, "loss": 0.0165, "step": 4870 }, { "epoch": 6.429512516469038, "grad_norm": 0.2542962431907654, "learning_rate": 5.611899528674923e-05, "loss": 0.0154, "step": 4880 }, { "epoch": 6.442687747035573, "grad_norm": 0.29230985045433044, "learning_rate": 5.595485800474349e-05, "loss": 0.0153, "step": 4890 }, { "epoch": 6.455862977602108, "grad_norm": 0.33685338497161865, "learning_rate": 5.579065560136467e-05, "loss": 0.017, "step": 4900 }, { "epoch": 6.469038208168643, "grad_norm": 0.3226190507411957, "learning_rate": 5.562638987230392e-05, "loss": 0.0172, "step": 4910 }, { "epoch": 6.482213438735178, "grad_norm": 0.2552209198474884, "learning_rate": 5.546206261394498e-05, "loss": 0.0123, "step": 4920 }, { "epoch": 6.495388669301713, "grad_norm": 0.26738399267196655, "learning_rate": 5.529767562334437e-05, "loss": 0.0159, "step": 4930 }, { "epoch": 6.508563899868248, "grad_norm": 0.4078490138053894, "learning_rate": 5.5133230698211926e-05, "loss": 0.0146, "step": 4940 }, { "epoch": 6.521739130434782, "grad_norm": 0.31933119893074036, "learning_rate": 5.496872963689096e-05, "loss": 0.0146, "step": 4950 }, { "epoch": 6.534914361001317, "grad_norm": 0.17034372687339783, "learning_rate": 5.4804174238338756e-05, "loss": 0.0172, "step": 4960 }, { "epoch": 6.5480895915678525, "grad_norm": 0.32328030467033386, "learning_rate": 5.463956630210678e-05, "loss": 0.0159, "step": 4970 }, { "epoch": 6.561264822134388, "grad_norm": 0.23999857902526855, "learning_rate": 5.4474907628321046e-05, "loss": 0.0198, "step": 4980 }, { "epoch": 6.574440052700922, "grad_norm": 0.20127585530281067, "learning_rate": 5.431020001766244e-05, "loss": 0.0129, "step": 4990 }, { "epoch": 6.587615283267457, "grad_norm": 0.3747844696044922, "learning_rate": 5.4145445271346986e-05, "loss": 0.018, "step": 5000 }, { "epoch": 6.600790513833992, "grad_norm": 0.25499704480171204, "learning_rate": 5.398064519110622e-05, "loss": 0.0145, "step": 5010 }, { "epoch": 6.613965744400527, "grad_norm": 0.20843933522701263, "learning_rate": 5.3815801579167394e-05, "loss": 0.0171, "step": 5020 }, { "epoch": 6.627140974967062, "grad_norm": 0.23514339327812195, "learning_rate": 5.365091623823382e-05, "loss": 0.0147, "step": 5030 }, { "epoch": 6.640316205533597, "grad_norm": 0.2892953157424927, "learning_rate": 5.348599097146521e-05, "loss": 0.0154, "step": 5040 }, { "epoch": 6.653491436100132, "grad_norm": 0.2587839961051941, "learning_rate": 5.3321027582457836e-05, "loss": 0.015, "step": 5050 }, { "epoch": 6.666666666666667, "grad_norm": 0.15298497676849365, "learning_rate": 5.315602787522491e-05, "loss": 0.011, "step": 5060 }, { "epoch": 6.679841897233201, "grad_norm": 0.24668270349502563, "learning_rate": 5.299099365417678e-05, "loss": 0.0148, "step": 5070 }, { "epoch": 6.693017127799736, "grad_norm": 0.19608354568481445, "learning_rate": 5.2825926724101236e-05, "loss": 0.0133, "step": 5080 }, { "epoch": 6.7061923583662715, "grad_norm": 0.28214743733406067, "learning_rate": 5.26608288901438e-05, "loss": 0.0149, "step": 5090 }, { "epoch": 6.719367588932807, "grad_norm": 0.301462858915329, "learning_rate": 5.24957019577879e-05, "loss": 0.0187, "step": 5100 }, { "epoch": 6.732542819499341, "grad_norm": 0.3126068115234375, "learning_rate": 5.2330547732835266e-05, "loss": 0.0135, "step": 5110 }, { "epoch": 6.745718050065876, "grad_norm": 0.27192267775535583, "learning_rate": 5.2165368021385996e-05, "loss": 0.0171, "step": 5120 }, { "epoch": 6.758893280632411, "grad_norm": 0.2133040726184845, "learning_rate": 5.200016462981897e-05, "loss": 0.0146, "step": 5130 }, { "epoch": 6.772068511198946, "grad_norm": 0.22072051465511322, "learning_rate": 5.1834939364772015e-05, "loss": 0.0128, "step": 5140 }, { "epoch": 6.7852437417654805, "grad_norm": 0.16883297264575958, "learning_rate": 5.166969403312214e-05, "loss": 0.0148, "step": 5150 }, { "epoch": 6.798418972332016, "grad_norm": 0.24253515899181366, "learning_rate": 5.1504430441965844e-05, "loss": 0.0158, "step": 5160 }, { "epoch": 6.811594202898551, "grad_norm": 0.23843777179718018, "learning_rate": 5.133915039859923e-05, "loss": 0.011, "step": 5170 }, { "epoch": 6.824769433465086, "grad_norm": 0.20740272104740143, "learning_rate": 5.1173855710498444e-05, "loss": 0.0123, "step": 5180 }, { "epoch": 6.837944664031621, "grad_norm": 0.1737157106399536, "learning_rate": 5.100854818529967e-05, "loss": 0.0148, "step": 5190 }, { "epoch": 6.851119894598155, "grad_norm": 0.2929360270500183, "learning_rate": 5.084322963077951e-05, "loss": 0.0201, "step": 5200 }, { "epoch": 6.86429512516469, "grad_norm": 0.33327507972717285, "learning_rate": 5.067790185483522e-05, "loss": 0.016, "step": 5210 }, { "epoch": 6.877470355731226, "grad_norm": 0.25603005290031433, "learning_rate": 5.0512566665464844e-05, "loss": 0.0164, "step": 5220 }, { "epoch": 6.89064558629776, "grad_norm": 0.2927559018135071, "learning_rate": 5.034722587074755e-05, "loss": 0.0155, "step": 5230 }, { "epoch": 6.903820816864295, "grad_norm": 0.31930333375930786, "learning_rate": 5.018188127882375e-05, "loss": 0.0128, "step": 5240 }, { "epoch": 6.91699604743083, "grad_norm": 0.1962880790233612, "learning_rate": 5.0016534697875417e-05, "loss": 0.014, "step": 5250 }, { "epoch": 6.930171277997365, "grad_norm": 0.1927347481250763, "learning_rate": 4.9851187936106294e-05, "loss": 0.015, "step": 5260 }, { "epoch": 6.9433465085639, "grad_norm": 0.3085514008998871, "learning_rate": 4.968584280172206e-05, "loss": 0.0176, "step": 5270 }, { "epoch": 6.956521739130435, "grad_norm": 0.30694425106048584, "learning_rate": 4.95205011029106e-05, "loss": 0.013, "step": 5280 }, { "epoch": 6.96969696969697, "grad_norm": 0.3337339758872986, "learning_rate": 4.935516464782227e-05, "loss": 0.0155, "step": 5290 }, { "epoch": 6.982872200263505, "grad_norm": 0.273749977350235, "learning_rate": 4.918983524455003e-05, "loss": 0.0141, "step": 5300 }, { "epoch": 6.996047430830039, "grad_norm": 0.2660110890865326, "learning_rate": 4.9024514701109766e-05, "loss": 0.0154, "step": 5310 }, { "epoch": 7.009222661396574, "grad_norm": 0.16499164700508118, "learning_rate": 4.885920482542043e-05, "loss": 0.0135, "step": 5320 }, { "epoch": 7.022397891963109, "grad_norm": 0.2095925658941269, "learning_rate": 4.869390742528438e-05, "loss": 0.0167, "step": 5330 }, { "epoch": 7.0355731225296445, "grad_norm": 0.2133823186159134, "learning_rate": 4.852862430836744e-05, "loss": 0.0155, "step": 5340 }, { "epoch": 7.048748353096179, "grad_norm": 0.42575761675834656, "learning_rate": 4.836335728217933e-05, "loss": 0.014, "step": 5350 }, { "epoch": 7.061923583662714, "grad_norm": 0.2763396203517914, "learning_rate": 4.819810815405379e-05, "loss": 0.0159, "step": 5360 }, { "epoch": 7.075098814229249, "grad_norm": 0.17784270644187927, "learning_rate": 4.803287873112877e-05, "loss": 0.0138, "step": 5370 }, { "epoch": 7.088274044795784, "grad_norm": 0.36119556427001953, "learning_rate": 4.786767082032681e-05, "loss": 0.0167, "step": 5380 }, { "epoch": 7.101449275362318, "grad_norm": 0.2838824689388275, "learning_rate": 4.77024862283351e-05, "loss": 0.0131, "step": 5390 }, { "epoch": 7.1146245059288535, "grad_norm": 0.3211531341075897, "learning_rate": 4.753732676158593e-05, "loss": 0.0122, "step": 5400 }, { "epoch": 7.127799736495389, "grad_norm": 0.2507878541946411, "learning_rate": 4.737219422623672e-05, "loss": 0.015, "step": 5410 }, { "epoch": 7.140974967061924, "grad_norm": 0.29864948987960815, "learning_rate": 4.720709042815044e-05, "loss": 0.0121, "step": 5420 }, { "epoch": 7.154150197628459, "grad_norm": 0.20546981692314148, "learning_rate": 4.704201717287578e-05, "loss": 0.0119, "step": 5430 }, { "epoch": 7.167325428194993, "grad_norm": 0.29513996839523315, "learning_rate": 4.6876976265627404e-05, "loss": 0.0135, "step": 5440 }, { "epoch": 7.180500658761528, "grad_norm": 0.252029687166214, "learning_rate": 4.671196951126626e-05, "loss": 0.0158, "step": 5450 }, { "epoch": 7.1936758893280635, "grad_norm": 0.31195035576820374, "learning_rate": 4.654699871427971e-05, "loss": 0.0144, "step": 5460 }, { "epoch": 7.206851119894598, "grad_norm": 0.2600157558917999, "learning_rate": 4.6382065678762034e-05, "loss": 0.0141, "step": 5470 }, { "epoch": 7.220026350461133, "grad_norm": 0.21147438883781433, "learning_rate": 4.6217172208394424e-05, "loss": 0.013, "step": 5480 }, { "epoch": 7.233201581027668, "grad_norm": 0.2041497379541397, "learning_rate": 4.605232010642549e-05, "loss": 0.0128, "step": 5490 }, { "epoch": 7.246376811594203, "grad_norm": 0.25358089804649353, "learning_rate": 4.588751117565142e-05, "loss": 0.0115, "step": 5500 }, { "epoch": 7.259552042160738, "grad_norm": 0.23748533427715302, "learning_rate": 4.5722747218396214e-05, "loss": 0.012, "step": 5510 }, { "epoch": 7.2727272727272725, "grad_norm": 0.2055012583732605, "learning_rate": 4.5558030036492194e-05, "loss": 0.0136, "step": 5520 }, { "epoch": 7.285902503293808, "grad_norm": 0.15180236101150513, "learning_rate": 4.539336143125999e-05, "loss": 0.0136, "step": 5530 }, { "epoch": 7.299077733860343, "grad_norm": 0.26564714312553406, "learning_rate": 4.522874320348916e-05, "loss": 0.0172, "step": 5540 }, { "epoch": 7.312252964426877, "grad_norm": 0.19193261861801147, "learning_rate": 4.506417715341821e-05, "loss": 0.0096, "step": 5550 }, { "epoch": 7.325428194993412, "grad_norm": 0.12721064686775208, "learning_rate": 4.489966508071511e-05, "loss": 0.0129, "step": 5560 }, { "epoch": 7.338603425559947, "grad_norm": 0.2680702209472656, "learning_rate": 4.4735208784457575e-05, "loss": 0.012, "step": 5570 }, { "epoch": 7.351778656126482, "grad_norm": 0.22795313596725464, "learning_rate": 4.457081006311325e-05, "loss": 0.0114, "step": 5580 }, { "epoch": 7.3649538866930175, "grad_norm": 0.24292701482772827, "learning_rate": 4.440647071452027e-05, "loss": 0.014, "step": 5590 }, { "epoch": 7.378129117259552, "grad_norm": 0.22425279021263123, "learning_rate": 4.424219253586737e-05, "loss": 0.0147, "step": 5600 }, { "epoch": 7.391304347826087, "grad_norm": 0.2584420442581177, "learning_rate": 4.407797732367443e-05, "loss": 0.0142, "step": 5610 }, { "epoch": 7.404479578392622, "grad_norm": 0.27375370264053345, "learning_rate": 4.391382687377268e-05, "loss": 0.0139, "step": 5620 }, { "epoch": 7.417654808959157, "grad_norm": 0.269481360912323, "learning_rate": 4.374974298128512e-05, "loss": 0.0131, "step": 5630 }, { "epoch": 7.430830039525691, "grad_norm": 0.200521782040596, "learning_rate": 4.358572744060699e-05, "loss": 0.0113, "step": 5640 }, { "epoch": 7.444005270092227, "grad_norm": 0.1899503469467163, "learning_rate": 4.342178204538588e-05, "loss": 0.0157, "step": 5650 }, { "epoch": 7.457180500658762, "grad_norm": 0.17898981273174286, "learning_rate": 4.325790858850241e-05, "loss": 0.0096, "step": 5660 }, { "epoch": 7.470355731225297, "grad_norm": 0.1877722591161728, "learning_rate": 4.309410886205043e-05, "loss": 0.013, "step": 5670 }, { "epoch": 7.483530961791831, "grad_norm": 0.19093987345695496, "learning_rate": 4.293038465731752e-05, "loss": 0.0135, "step": 5680 }, { "epoch": 7.496706192358366, "grad_norm": 0.32967081665992737, "learning_rate": 4.276673776476533e-05, "loss": 0.0163, "step": 5690 }, { "epoch": 7.509881422924901, "grad_norm": 0.24727611243724823, "learning_rate": 4.260316997401007e-05, "loss": 0.0151, "step": 5700 }, { "epoch": 7.5230566534914365, "grad_norm": 0.23724959790706635, "learning_rate": 4.243968307380293e-05, "loss": 0.0124, "step": 5710 }, { "epoch": 7.536231884057971, "grad_norm": 0.24676303565502167, "learning_rate": 4.22762788520104e-05, "loss": 0.0162, "step": 5720 }, { "epoch": 7.549407114624506, "grad_norm": 0.23069219291210175, "learning_rate": 4.211295909559491e-05, "loss": 0.0151, "step": 5730 }, { "epoch": 7.562582345191041, "grad_norm": 0.22480972111225128, "learning_rate": 4.194972559059511e-05, "loss": 0.0146, "step": 5740 }, { "epoch": 7.575757575757576, "grad_norm": 0.22481106221675873, "learning_rate": 4.178658012210651e-05, "loss": 0.0133, "step": 5750 }, { "epoch": 7.58893280632411, "grad_norm": 0.2546963393688202, "learning_rate": 4.162352447426177e-05, "loss": 0.0132, "step": 5760 }, { "epoch": 7.6021080368906455, "grad_norm": 0.24717223644256592, "learning_rate": 4.146056043021135e-05, "loss": 0.0145, "step": 5770 }, { "epoch": 7.615283267457181, "grad_norm": 0.20582571625709534, "learning_rate": 4.1297689772103944e-05, "loss": 0.0129, "step": 5780 }, { "epoch": 7.628458498023716, "grad_norm": 0.18852469325065613, "learning_rate": 4.113491428106694e-05, "loss": 0.013, "step": 5790 }, { "epoch": 7.64163372859025, "grad_norm": 0.2953014671802521, "learning_rate": 4.0972235737187055e-05, "loss": 0.0164, "step": 5800 }, { "epoch": 7.654808959156785, "grad_norm": 0.1740752011537552, "learning_rate": 4.080965591949076e-05, "loss": 0.0126, "step": 5810 }, { "epoch": 7.66798418972332, "grad_norm": 0.24041345715522766, "learning_rate": 4.0647176605924924e-05, "loss": 0.0129, "step": 5820 }, { "epoch": 7.681159420289855, "grad_norm": 0.25071561336517334, "learning_rate": 4.0484799573337255e-05, "loss": 0.0172, "step": 5830 }, { "epoch": 7.69433465085639, "grad_norm": 0.20551979541778564, "learning_rate": 4.032252659745699e-05, "loss": 0.0146, "step": 5840 }, { "epoch": 7.707509881422925, "grad_norm": 0.20781241357326508, "learning_rate": 4.016035945287539e-05, "loss": 0.0132, "step": 5850 }, { "epoch": 7.72068511198946, "grad_norm": 0.2976298928260803, "learning_rate": 3.999829991302635e-05, "loss": 0.0186, "step": 5860 }, { "epoch": 7.733860342555995, "grad_norm": 0.30154094099998474, "learning_rate": 3.983634975016707e-05, "loss": 0.0136, "step": 5870 }, { "epoch": 7.747035573122529, "grad_norm": 0.259523868560791, "learning_rate": 3.967451073535854e-05, "loss": 0.0144, "step": 5880 }, { "epoch": 7.7602108036890645, "grad_norm": 0.24883371591567993, "learning_rate": 3.951278463844633e-05, "loss": 0.015, "step": 5890 }, { "epoch": 7.7733860342556, "grad_norm": 0.2566511631011963, "learning_rate": 3.935117322804111e-05, "loss": 0.0137, "step": 5900 }, { "epoch": 7.786561264822135, "grad_norm": 0.2503558099269867, "learning_rate": 3.918967827149938e-05, "loss": 0.0182, "step": 5910 }, { "epoch": 7.799736495388669, "grad_norm": 0.2588229179382324, "learning_rate": 3.9028301534904094e-05, "loss": 0.0109, "step": 5920 }, { "epoch": 7.812911725955204, "grad_norm": 0.22495663166046143, "learning_rate": 3.88670447830454e-05, "loss": 0.0125, "step": 5930 }, { "epoch": 7.826086956521739, "grad_norm": 0.2702830731868744, "learning_rate": 3.870590977940132e-05, "loss": 0.0131, "step": 5940 }, { "epoch": 7.839262187088274, "grad_norm": 0.16298788785934448, "learning_rate": 3.8544898286118404e-05, "loss": 0.011, "step": 5950 }, { "epoch": 7.852437417654809, "grad_norm": 0.18812081217765808, "learning_rate": 3.838401206399257e-05, "loss": 0.0122, "step": 5960 }, { "epoch": 7.865612648221344, "grad_norm": 0.2774938642978668, "learning_rate": 3.822325287244975e-05, "loss": 0.012, "step": 5970 }, { "epoch": 7.878787878787879, "grad_norm": 0.15192358195781708, "learning_rate": 3.8062622469526725e-05, "loss": 0.0121, "step": 5980 }, { "epoch": 7.891963109354414, "grad_norm": 0.24365267157554626, "learning_rate": 3.790212261185183e-05, "loss": 0.0108, "step": 5990 }, { "epoch": 7.905138339920948, "grad_norm": 0.24914588034152985, "learning_rate": 3.7741755054625794e-05, "loss": 0.0112, "step": 6000 }, { "epoch": 7.918313570487483, "grad_norm": 0.23948980867862701, "learning_rate": 3.758152155160255e-05, "loss": 0.0115, "step": 6010 }, { "epoch": 7.9314888010540185, "grad_norm": 0.24214811623096466, "learning_rate": 3.742142385506999e-05, "loss": 0.0123, "step": 6020 }, { "epoch": 7.944664031620554, "grad_norm": 0.19386622309684753, "learning_rate": 3.72614637158309e-05, "loss": 0.0143, "step": 6030 }, { "epoch": 7.957839262187088, "grad_norm": 0.18021267652511597, "learning_rate": 3.710164288318371e-05, "loss": 0.0132, "step": 6040 }, { "epoch": 7.971014492753623, "grad_norm": 0.21891270577907562, "learning_rate": 3.694196310490345e-05, "loss": 0.0108, "step": 6050 }, { "epoch": 7.984189723320158, "grad_norm": 0.23576456308364868, "learning_rate": 3.678242612722259e-05, "loss": 0.0096, "step": 6060 }, { "epoch": 7.997364953886693, "grad_norm": 0.12668360769748688, "learning_rate": 3.6623033694811953e-05, "loss": 0.012, "step": 6070 }, { "epoch": 8.010540184453228, "grad_norm": 0.16363386809825897, "learning_rate": 3.6463787550761665e-05, "loss": 0.012, "step": 6080 }, { "epoch": 8.023715415019764, "grad_norm": 0.1997222751379013, "learning_rate": 3.630468943656202e-05, "loss": 0.0114, "step": 6090 }, { "epoch": 8.036890645586297, "grad_norm": 0.1548883616924286, "learning_rate": 3.6145741092084523e-05, "loss": 0.0124, "step": 6100 }, { "epoch": 8.050065876152832, "grad_norm": 0.2827013432979584, "learning_rate": 3.598694425556278e-05, "loss": 0.0114, "step": 6110 }, { "epoch": 8.063241106719367, "grad_norm": 0.24781610071659088, "learning_rate": 3.58283006635736e-05, "loss": 0.0129, "step": 6120 }, { "epoch": 8.076416337285902, "grad_norm": 0.22967146337032318, "learning_rate": 3.566981205101781e-05, "loss": 0.0135, "step": 6130 }, { "epoch": 8.089591567852437, "grad_norm": 0.2027362883090973, "learning_rate": 3.5511480151101556e-05, "loss": 0.0142, "step": 6140 }, { "epoch": 8.102766798418973, "grad_norm": 0.1740102767944336, "learning_rate": 3.5353306695317104e-05, "loss": 0.0116, "step": 6150 }, { "epoch": 8.115942028985508, "grad_norm": 0.1729733943939209, "learning_rate": 3.519529341342402e-05, "loss": 0.0114, "step": 6160 }, { "epoch": 8.129117259552043, "grad_norm": 0.27243903279304504, "learning_rate": 3.503744203343026e-05, "loss": 0.0116, "step": 6170 }, { "epoch": 8.142292490118576, "grad_norm": 0.36281585693359375, "learning_rate": 3.487975428157318e-05, "loss": 0.0127, "step": 6180 }, { "epoch": 8.155467720685111, "grad_norm": 0.21071209013462067, "learning_rate": 3.472223188230083e-05, "loss": 0.0143, "step": 6190 }, { "epoch": 8.168642951251647, "grad_norm": 0.16985180974006653, "learning_rate": 3.4564876558252866e-05, "loss": 0.0161, "step": 6200 }, { "epoch": 8.181818181818182, "grad_norm": 0.28265002369880676, "learning_rate": 3.440769003024195e-05, "loss": 0.0114, "step": 6210 }, { "epoch": 8.194993412384717, "grad_norm": 0.23601707816123962, "learning_rate": 3.425067401723477e-05, "loss": 0.0123, "step": 6220 }, { "epoch": 8.208168642951252, "grad_norm": 0.23068974912166595, "learning_rate": 3.409383023633325e-05, "loss": 0.0102, "step": 6230 }, { "epoch": 8.221343873517787, "grad_norm": 0.16201822459697723, "learning_rate": 3.3937160402755894e-05, "loss": 0.009, "step": 6240 }, { "epoch": 8.234519104084322, "grad_norm": 0.18543598055839539, "learning_rate": 3.378066622981885e-05, "loss": 0.0119, "step": 6250 }, { "epoch": 8.247694334650856, "grad_norm": 0.14588962495326996, "learning_rate": 3.362434942891738e-05, "loss": 0.0135, "step": 6260 }, { "epoch": 8.26086956521739, "grad_norm": 0.23218132555484772, "learning_rate": 3.346821170950693e-05, "loss": 0.0138, "step": 6270 }, { "epoch": 8.274044795783926, "grad_norm": 0.1436203271150589, "learning_rate": 3.3312254779084585e-05, "loss": 0.0119, "step": 6280 }, { "epoch": 8.287220026350461, "grad_norm": 0.20360875129699707, "learning_rate": 3.315648034317039e-05, "loss": 0.013, "step": 6290 }, { "epoch": 8.300395256916996, "grad_norm": 0.1437600553035736, "learning_rate": 3.3000890105288564e-05, "loss": 0.0113, "step": 6300 }, { "epoch": 8.313570487483531, "grad_norm": 0.3516365885734558, "learning_rate": 3.284548576694908e-05, "loss": 0.0121, "step": 6310 }, { "epoch": 8.326745718050066, "grad_norm": 0.1786649525165558, "learning_rate": 3.2690269027628815e-05, "loss": 0.0118, "step": 6320 }, { "epoch": 8.339920948616601, "grad_norm": 0.2372141033411026, "learning_rate": 3.253524158475324e-05, "loss": 0.0132, "step": 6330 }, { "epoch": 8.353096179183135, "grad_norm": 0.254221111536026, "learning_rate": 3.238040513367757e-05, "loss": 0.0117, "step": 6340 }, { "epoch": 8.36627140974967, "grad_norm": 0.2704264521598816, "learning_rate": 3.222576136766843e-05, "loss": 0.0106, "step": 6350 }, { "epoch": 8.379446640316205, "grad_norm": 0.19264473021030426, "learning_rate": 3.2071311977885324e-05, "loss": 0.0091, "step": 6360 }, { "epoch": 8.39262187088274, "grad_norm": 0.394570529460907, "learning_rate": 3.191705865336197e-05, "loss": 0.012, "step": 6370 }, { "epoch": 8.405797101449275, "grad_norm": 0.282055139541626, "learning_rate": 3.1763003080988075e-05, "loss": 0.0111, "step": 6380 }, { "epoch": 8.41897233201581, "grad_norm": 0.14342975616455078, "learning_rate": 3.160914694549063e-05, "loss": 0.0111, "step": 6390 }, { "epoch": 8.432147562582346, "grad_norm": 0.20480139553546906, "learning_rate": 3.145549192941573e-05, "loss": 0.0131, "step": 6400 }, { "epoch": 8.44532279314888, "grad_norm": 0.3057665228843689, "learning_rate": 3.130203971310999e-05, "loss": 0.0149, "step": 6410 }, { "epoch": 8.458498023715414, "grad_norm": 0.20350416004657745, "learning_rate": 3.114879197470225e-05, "loss": 0.0108, "step": 6420 }, { "epoch": 8.47167325428195, "grad_norm": 0.19562511146068573, "learning_rate": 3.0995750390085285e-05, "loss": 0.0138, "step": 6430 }, { "epoch": 8.484848484848484, "grad_norm": 0.23752659559249878, "learning_rate": 3.084291663289728e-05, "loss": 0.011, "step": 6440 }, { "epoch": 8.49802371541502, "grad_norm": 0.20719431340694427, "learning_rate": 3.069029237450375e-05, "loss": 0.0118, "step": 6450 }, { "epoch": 8.511198945981555, "grad_norm": 0.14212384819984436, "learning_rate": 3.053787928397911e-05, "loss": 0.0104, "step": 6460 }, { "epoch": 8.52437417654809, "grad_norm": 0.1687770038843155, "learning_rate": 3.0385679028088526e-05, "loss": 0.0104, "step": 6470 }, { "epoch": 8.537549407114625, "grad_norm": 0.17196719348430634, "learning_rate": 3.023369327126959e-05, "loss": 0.0104, "step": 6480 }, { "epoch": 8.55072463768116, "grad_norm": 0.0772448256611824, "learning_rate": 3.0081923675614198e-05, "loss": 0.0111, "step": 6490 }, { "epoch": 8.563899868247695, "grad_norm": 0.200760155916214, "learning_rate": 2.993037190085034e-05, "loss": 0.0108, "step": 6500 }, { "epoch": 8.577075098814229, "grad_norm": 0.14407402276992798, "learning_rate": 2.977903960432392e-05, "loss": 0.0121, "step": 6510 }, { "epoch": 8.590250329380764, "grad_norm": 0.22289133071899414, "learning_rate": 2.9627928440980722e-05, "loss": 0.0109, "step": 6520 }, { "epoch": 8.603425559947299, "grad_norm": 0.15810120105743408, "learning_rate": 2.9477040063348183e-05, "loss": 0.0134, "step": 6530 }, { "epoch": 8.616600790513834, "grad_norm": 0.2241305559873581, "learning_rate": 2.9326376121517456e-05, "loss": 0.0112, "step": 6540 }, { "epoch": 8.62977602108037, "grad_norm": 0.2314157485961914, "learning_rate": 2.9175938263125236e-05, "loss": 0.0111, "step": 6550 }, { "epoch": 8.642951251646904, "grad_norm": 0.21943719685077667, "learning_rate": 2.9025728133335873e-05, "loss": 0.012, "step": 6560 }, { "epoch": 8.65612648221344, "grad_norm": 0.14395660161972046, "learning_rate": 2.8875747374823288e-05, "loss": 0.0109, "step": 6570 }, { "epoch": 8.669301712779973, "grad_norm": 0.21410293877124786, "learning_rate": 2.872599762775298e-05, "loss": 0.0164, "step": 6580 }, { "epoch": 8.682476943346508, "grad_norm": 0.19529272615909576, "learning_rate": 2.857648052976425e-05, "loss": 0.0141, "step": 6590 }, { "epoch": 8.695652173913043, "grad_norm": 0.21148556470870972, "learning_rate": 2.8427197715952047e-05, "loss": 0.0097, "step": 6600 }, { "epoch": 8.708827404479578, "grad_norm": 0.25088655948638916, "learning_rate": 2.8278150818849393e-05, "loss": 0.0126, "step": 6610 }, { "epoch": 8.722002635046113, "grad_norm": 0.20759238302707672, "learning_rate": 2.812934146840922e-05, "loss": 0.0119, "step": 6620 }, { "epoch": 8.735177865612648, "grad_norm": 0.23086342215538025, "learning_rate": 2.7980771291986764e-05, "loss": 0.0122, "step": 6630 }, { "epoch": 8.748353096179184, "grad_norm": 0.21984222531318665, "learning_rate": 2.783244191432167e-05, "loss": 0.0137, "step": 6640 }, { "epoch": 8.761528326745719, "grad_norm": 0.16901598870754242, "learning_rate": 2.768435495752022e-05, "loss": 0.0102, "step": 6650 }, { "epoch": 8.774703557312254, "grad_norm": 0.16642488539218903, "learning_rate": 2.753651204103771e-05, "loss": 0.011, "step": 6660 }, { "epoch": 8.787878787878787, "grad_norm": 0.30090758204460144, "learning_rate": 2.7388914781660523e-05, "loss": 0.0132, "step": 6670 }, { "epoch": 8.801054018445322, "grad_norm": 0.2027425616979599, "learning_rate": 2.7241564793488693e-05, "loss": 0.01, "step": 6680 }, { "epoch": 8.814229249011857, "grad_norm": 0.17420056462287903, "learning_rate": 2.7094463687918037e-05, "loss": 0.0113, "step": 6690 }, { "epoch": 8.827404479578393, "grad_norm": 0.1992754340171814, "learning_rate": 2.694761307362268e-05, "loss": 0.0112, "step": 6700 }, { "epoch": 8.840579710144928, "grad_norm": 0.14116120338439941, "learning_rate": 2.6801014556537467e-05, "loss": 0.0088, "step": 6710 }, { "epoch": 8.853754940711463, "grad_norm": 0.15394073724746704, "learning_rate": 2.6654669739840243e-05, "loss": 0.0097, "step": 6720 }, { "epoch": 8.866930171277998, "grad_norm": 0.2150605171918869, "learning_rate": 2.650858022393451e-05, "loss": 0.0104, "step": 6730 }, { "epoch": 8.880105401844531, "grad_norm": 0.22290481626987457, "learning_rate": 2.6362747606431747e-05, "loss": 0.012, "step": 6740 }, { "epoch": 8.893280632411066, "grad_norm": 0.18290726840496063, "learning_rate": 2.6217173482134172e-05, "loss": 0.0117, "step": 6750 }, { "epoch": 8.906455862977602, "grad_norm": 0.18754500150680542, "learning_rate": 2.6071859443017044e-05, "loss": 0.009, "step": 6760 }, { "epoch": 8.919631093544137, "grad_norm": 0.18662101030349731, "learning_rate": 2.5926807078211414e-05, "loss": 0.0113, "step": 6770 }, { "epoch": 8.932806324110672, "grad_norm": 0.2117272913455963, "learning_rate": 2.5782017973986728e-05, "loss": 0.0097, "step": 6780 }, { "epoch": 8.945981554677207, "grad_norm": 0.2579967975616455, "learning_rate": 2.5637493713733374e-05, "loss": 0.0115, "step": 6790 }, { "epoch": 8.959156785243742, "grad_norm": 0.3161567747592926, "learning_rate": 2.549323587794559e-05, "loss": 0.0102, "step": 6800 }, { "epoch": 8.972332015810277, "grad_norm": 0.1549680531024933, "learning_rate": 2.5349246044203895e-05, "loss": 0.0149, "step": 6810 }, { "epoch": 8.985507246376812, "grad_norm": 0.20083005726337433, "learning_rate": 2.520552578715808e-05, "loss": 0.0132, "step": 6820 }, { "epoch": 8.998682476943346, "grad_norm": 0.1578175276517868, "learning_rate": 2.506207667850981e-05, "loss": 0.0073, "step": 6830 }, { "epoch": 9.011857707509881, "grad_norm": 0.12106601148843765, "learning_rate": 2.4918900286995555e-05, "loss": 0.0087, "step": 6840 }, { "epoch": 9.025032938076416, "grad_norm": 0.14856408536434174, "learning_rate": 2.4775998178369458e-05, "loss": 0.0093, "step": 6850 }, { "epoch": 9.038208168642951, "grad_norm": 0.21818016469478607, "learning_rate": 2.4633371915386017e-05, "loss": 0.0099, "step": 6860 }, { "epoch": 9.051383399209486, "grad_norm": 0.23136316239833832, "learning_rate": 2.4491023057783235e-05, "loss": 0.012, "step": 6870 }, { "epoch": 9.064558629776021, "grad_norm": 0.23694013059139252, "learning_rate": 2.4348953162265375e-05, "loss": 0.0135, "step": 6880 }, { "epoch": 9.077733860342557, "grad_norm": 0.17179802060127258, "learning_rate": 2.420716378248607e-05, "loss": 0.0129, "step": 6890 }, { "epoch": 9.090909090909092, "grad_norm": 0.1833650767803192, "learning_rate": 2.4065656469031266e-05, "loss": 0.0099, "step": 6900 }, { "epoch": 9.104084321475625, "grad_norm": 0.17761869728565216, "learning_rate": 2.3924432769402268e-05, "loss": 0.0114, "step": 6910 }, { "epoch": 9.11725955204216, "grad_norm": 0.21558180451393127, "learning_rate": 2.3783494227998844e-05, "loss": 0.0123, "step": 6920 }, { "epoch": 9.130434782608695, "grad_norm": 0.1749400645494461, "learning_rate": 2.3642842386102264e-05, "loss": 0.014, "step": 6930 }, { "epoch": 9.14361001317523, "grad_norm": 0.17026203870773315, "learning_rate": 2.3502478781858567e-05, "loss": 0.0089, "step": 6940 }, { "epoch": 9.156785243741766, "grad_norm": 0.2876078486442566, "learning_rate": 2.3362404950261628e-05, "loss": 0.0097, "step": 6950 }, { "epoch": 9.1699604743083, "grad_norm": 0.3452471196651459, "learning_rate": 2.3222622423136458e-05, "loss": 0.0129, "step": 6960 }, { "epoch": 9.183135704874836, "grad_norm": 0.2366582602262497, "learning_rate": 2.3083132729122332e-05, "loss": 0.0103, "step": 6970 }, { "epoch": 9.196310935441371, "grad_norm": 0.2024129182100296, "learning_rate": 2.294393739365621e-05, "loss": 0.014, "step": 6980 }, { "epoch": 9.209486166007904, "grad_norm": 0.19319729506969452, "learning_rate": 2.2805037938956e-05, "loss": 0.012, "step": 6990 }, { "epoch": 9.22266139657444, "grad_norm": 0.17540772259235382, "learning_rate": 2.266643588400386e-05, "loss": 0.0114, "step": 7000 }, { "epoch": 9.235836627140975, "grad_norm": 0.23558589816093445, "learning_rate": 2.252813274452969e-05, "loss": 0.0098, "step": 7010 }, { "epoch": 9.24901185770751, "grad_norm": 0.19168560206890106, "learning_rate": 2.2390130032994427e-05, "loss": 0.0086, "step": 7020 }, { "epoch": 9.262187088274045, "grad_norm": 0.17272160947322845, "learning_rate": 2.2252429258573633e-05, "loss": 0.0087, "step": 7030 }, { "epoch": 9.27536231884058, "grad_norm": 0.13440345227718353, "learning_rate": 2.2115031927140904e-05, "loss": 0.0096, "step": 7040 }, { "epoch": 9.288537549407115, "grad_norm": 0.25527286529541016, "learning_rate": 2.1977939541251463e-05, "loss": 0.0117, "step": 7050 }, { "epoch": 9.30171277997365, "grad_norm": 0.1481703817844391, "learning_rate": 2.1841153600125684e-05, "loss": 0.0163, "step": 7060 }, { "epoch": 9.314888010540184, "grad_norm": 0.20705603063106537, "learning_rate": 2.170467559963267e-05, "loss": 0.0099, "step": 7070 }, { "epoch": 9.328063241106719, "grad_norm": 0.10488489270210266, "learning_rate": 2.1568507032273982e-05, "loss": 0.01, "step": 7080 }, { "epoch": 9.341238471673254, "grad_norm": 0.29786285758018494, "learning_rate": 2.1432649387167264e-05, "loss": 0.01, "step": 7090 }, { "epoch": 9.354413702239789, "grad_norm": 0.2556435763835907, "learning_rate": 2.1297104150029973e-05, "loss": 0.0088, "step": 7100 }, { "epoch": 9.367588932806324, "grad_norm": 0.2757495641708374, "learning_rate": 2.116187280316307e-05, "loss": 0.0114, "step": 7110 }, { "epoch": 9.38076416337286, "grad_norm": 0.25607019662857056, "learning_rate": 2.1026956825434908e-05, "loss": 0.0112, "step": 7120 }, { "epoch": 9.393939393939394, "grad_norm": 0.30468475818634033, "learning_rate": 2.0892357692265017e-05, "loss": 0.0099, "step": 7130 }, { "epoch": 9.40711462450593, "grad_norm": 0.28744977712631226, "learning_rate": 2.0758076875607947e-05, "loss": 0.0113, "step": 7140 }, { "epoch": 9.420289855072463, "grad_norm": 0.23797324299812317, "learning_rate": 2.0624115843937207e-05, "loss": 0.0112, "step": 7150 }, { "epoch": 9.433465085638998, "grad_norm": 0.2671097218990326, "learning_rate": 2.0490476062229157e-05, "loss": 0.0101, "step": 7160 }, { "epoch": 9.446640316205533, "grad_norm": 0.24652323126792908, "learning_rate": 2.035715899194704e-05, "loss": 0.0094, "step": 7170 }, { "epoch": 9.459815546772068, "grad_norm": 0.13198544085025787, "learning_rate": 2.022416609102499e-05, "loss": 0.0101, "step": 7180 }, { "epoch": 9.472990777338604, "grad_norm": 0.15571995079517365, "learning_rate": 2.009149881385205e-05, "loss": 0.0116, "step": 7190 }, { "epoch": 9.486166007905139, "grad_norm": 0.3202069103717804, "learning_rate": 1.995915861125634e-05, "loss": 0.0122, "step": 7200 }, { "epoch": 9.499341238471674, "grad_norm": 0.28108784556388855, "learning_rate": 1.9827146930489065e-05, "loss": 0.0091, "step": 7210 }, { "epoch": 9.512516469038209, "grad_norm": 0.22256651520729065, "learning_rate": 1.9695465215208848e-05, "loss": 0.0108, "step": 7220 }, { "epoch": 9.525691699604742, "grad_norm": 0.18551883101463318, "learning_rate": 1.9564114905465813e-05, "loss": 0.0114, "step": 7230 }, { "epoch": 9.538866930171277, "grad_norm": 0.15290212631225586, "learning_rate": 1.9433097437685936e-05, "loss": 0.0094, "step": 7240 }, { "epoch": 9.552042160737813, "grad_norm": 0.3061795234680176, "learning_rate": 1.930241424465521e-05, "loss": 0.0116, "step": 7250 }, { "epoch": 9.565217391304348, "grad_norm": 0.2448814958333969, "learning_rate": 1.9172066755504115e-05, "loss": 0.0086, "step": 7260 }, { "epoch": 9.578392621870883, "grad_norm": 0.22381773591041565, "learning_rate": 1.9042056395691914e-05, "loss": 0.0111, "step": 7270 }, { "epoch": 9.591567852437418, "grad_norm": 0.14286114275455475, "learning_rate": 1.8912384586991066e-05, "loss": 0.0086, "step": 7280 }, { "epoch": 9.604743083003953, "grad_norm": 0.2133261263370514, "learning_rate": 1.8783052747471717e-05, "loss": 0.0097, "step": 7290 }, { "epoch": 9.617918313570488, "grad_norm": 0.267894446849823, "learning_rate": 1.865406229148611e-05, "loss": 0.0122, "step": 7300 }, { "epoch": 9.631093544137022, "grad_norm": 0.2295752465724945, "learning_rate": 1.8525414629653233e-05, "loss": 0.01, "step": 7310 }, { "epoch": 9.644268774703557, "grad_norm": 0.2165345996618271, "learning_rate": 1.8397111168843255e-05, "loss": 0.0105, "step": 7320 }, { "epoch": 9.657444005270092, "grad_norm": 0.21037983894348145, "learning_rate": 1.8269153312162323e-05, "loss": 0.0093, "step": 7330 }, { "epoch": 9.670619235836627, "grad_norm": 0.21100281178951263, "learning_rate": 1.8141542458937054e-05, "loss": 0.0109, "step": 7340 }, { "epoch": 9.683794466403162, "grad_norm": 0.14599454402923584, "learning_rate": 1.8014280004699268e-05, "loss": 0.01, "step": 7350 }, { "epoch": 9.696969696969697, "grad_norm": 0.15527044236660004, "learning_rate": 1.788736734117078e-05, "loss": 0.0094, "step": 7360 }, { "epoch": 9.710144927536232, "grad_norm": 0.15139532089233398, "learning_rate": 1.7760805856248152e-05, "loss": 0.0092, "step": 7370 }, { "epoch": 9.723320158102768, "grad_norm": 0.1905406415462494, "learning_rate": 1.7634596933987518e-05, "loss": 0.0099, "step": 7380 }, { "epoch": 9.7364953886693, "grad_norm": 0.24150630831718445, "learning_rate": 1.7508741954589404e-05, "loss": 0.0098, "step": 7390 }, { "epoch": 9.749670619235836, "grad_norm": 0.13083931803703308, "learning_rate": 1.7383242294383717e-05, "loss": 0.0115, "step": 7400 }, { "epoch": 9.762845849802371, "grad_norm": 0.20750683546066284, "learning_rate": 1.7258099325814632e-05, "loss": 0.0091, "step": 7410 }, { "epoch": 9.776021080368906, "grad_norm": 0.29100707173347473, "learning_rate": 1.7133314417425594e-05, "loss": 0.01, "step": 7420 }, { "epoch": 9.789196310935441, "grad_norm": 0.14742396771907806, "learning_rate": 1.7008888933844408e-05, "loss": 0.0087, "step": 7430 }, { "epoch": 9.802371541501977, "grad_norm": 0.21344038844108582, "learning_rate": 1.6884824235768172e-05, "loss": 0.0107, "step": 7440 }, { "epoch": 9.815546772068512, "grad_norm": 0.1723639816045761, "learning_rate": 1.6761121679948592e-05, "loss": 0.0091, "step": 7450 }, { "epoch": 9.828722002635047, "grad_norm": 0.11757984012365341, "learning_rate": 1.663778261917695e-05, "loss": 0.0073, "step": 7460 }, { "epoch": 9.841897233201582, "grad_norm": 0.17279432713985443, "learning_rate": 1.651480840226952e-05, "loss": 0.0109, "step": 7470 }, { "epoch": 9.855072463768115, "grad_norm": 0.17756018042564392, "learning_rate": 1.639220037405258e-05, "loss": 0.0096, "step": 7480 }, { "epoch": 9.86824769433465, "grad_norm": 0.21475255489349365, "learning_rate": 1.6269959875347906e-05, "loss": 0.0114, "step": 7490 }, { "epoch": 9.881422924901186, "grad_norm": 0.1909179538488388, "learning_rate": 1.614808824295802e-05, "loss": 0.0092, "step": 7500 }, { "epoch": 9.89459815546772, "grad_norm": 0.13779670000076294, "learning_rate": 1.602658680965152e-05, "loss": 0.0092, "step": 7510 }, { "epoch": 9.907773386034256, "grad_norm": 0.18837608397006989, "learning_rate": 1.5905456904148686e-05, "loss": 0.0112, "step": 7520 }, { "epoch": 9.920948616600791, "grad_norm": 0.1538374423980713, "learning_rate": 1.57846998511067e-05, "loss": 0.0085, "step": 7530 }, { "epoch": 9.934123847167326, "grad_norm": 0.22132106125354767, "learning_rate": 1.566431697110538e-05, "loss": 0.0116, "step": 7540 }, { "epoch": 9.94729907773386, "grad_norm": 0.19287706911563873, "learning_rate": 1.554430958063259e-05, "loss": 0.0105, "step": 7550 }, { "epoch": 9.960474308300395, "grad_norm": 0.2113484889268875, "learning_rate": 1.5424678992069912e-05, "loss": 0.0083, "step": 7560 }, { "epoch": 9.97364953886693, "grad_norm": 0.2562454640865326, "learning_rate": 1.5305426513678362e-05, "loss": 0.0086, "step": 7570 }, { "epoch": 9.986824769433465, "grad_norm": 0.13577942550182343, "learning_rate": 1.518655344958388e-05, "loss": 0.0108, "step": 7580 }, { "epoch": 10.0, "grad_norm": 0.2062060385942459, "learning_rate": 1.5068061099763275e-05, "loss": 0.0108, "step": 7590 }, { "epoch": 10.013175230566535, "grad_norm": 0.14395931363105774, "learning_rate": 1.494995076002988e-05, "loss": 0.0082, "step": 7600 }, { "epoch": 10.02635046113307, "grad_norm": 0.13919766247272491, "learning_rate": 1.4832223722019456e-05, "loss": 0.0106, "step": 7610 }, { "epoch": 10.039525691699605, "grad_norm": 0.19914333522319794, "learning_rate": 1.4714881273176035e-05, "loss": 0.009, "step": 7620 }, { "epoch": 10.05270092226614, "grad_norm": 0.15142998099327087, "learning_rate": 1.4597924696737835e-05, "loss": 0.0089, "step": 7630 }, { "epoch": 10.065876152832674, "grad_norm": 0.2267027497291565, "learning_rate": 1.4481355271723252e-05, "loss": 0.0088, "step": 7640 }, { "epoch": 10.079051383399209, "grad_norm": 0.20620562136173248, "learning_rate": 1.4365174272916809e-05, "loss": 0.0093, "step": 7650 }, { "epoch": 10.092226613965744, "grad_norm": 0.17721879482269287, "learning_rate": 1.4249382970855319e-05, "loss": 0.0106, "step": 7660 }, { "epoch": 10.10540184453228, "grad_norm": 0.21728409826755524, "learning_rate": 1.4133982631813903e-05, "loss": 0.0087, "step": 7670 }, { "epoch": 10.118577075098814, "grad_norm": 0.13773860037326813, "learning_rate": 1.4018974517792194e-05, "loss": 0.0104, "step": 7680 }, { "epoch": 10.13175230566535, "grad_norm": 0.20984046161174774, "learning_rate": 1.390435988650048e-05, "loss": 0.0103, "step": 7690 }, { "epoch": 10.144927536231885, "grad_norm": 0.14403507113456726, "learning_rate": 1.3790139991346006e-05, "loss": 0.0082, "step": 7700 }, { "epoch": 10.15810276679842, "grad_norm": 0.16830037534236908, "learning_rate": 1.367631608141926e-05, "loss": 0.0097, "step": 7710 }, { "epoch": 10.171277997364953, "grad_norm": 0.1339355707168579, "learning_rate": 1.3562889401480278e-05, "loss": 0.008, "step": 7720 }, { "epoch": 10.184453227931488, "grad_norm": 0.19045881927013397, "learning_rate": 1.3449861191945074e-05, "loss": 0.01, "step": 7730 }, { "epoch": 10.197628458498023, "grad_norm": 0.11783690750598907, "learning_rate": 1.3337232688872009e-05, "loss": 0.0108, "step": 7740 }, { "epoch": 10.210803689064559, "grad_norm": 0.15417170524597168, "learning_rate": 1.3225005123948364e-05, "loss": 0.0095, "step": 7750 }, { "epoch": 10.223978919631094, "grad_norm": 0.15006369352340698, "learning_rate": 1.311317972447681e-05, "loss": 0.0095, "step": 7760 }, { "epoch": 10.237154150197629, "grad_norm": 0.11807779222726822, "learning_rate": 1.3001757713361996e-05, "loss": 0.0082, "step": 7770 }, { "epoch": 10.250329380764164, "grad_norm": 0.18806616961956024, "learning_rate": 1.2890740309097204e-05, "loss": 0.0091, "step": 7780 }, { "epoch": 10.2635046113307, "grad_norm": 0.18147198855876923, "learning_rate": 1.2780128725750944e-05, "loss": 0.0074, "step": 7790 }, { "epoch": 10.276679841897232, "grad_norm": 0.20511873066425323, "learning_rate": 1.266992417295379e-05, "loss": 0.0092, "step": 7800 }, { "epoch": 10.289855072463768, "grad_norm": 0.16353419423103333, "learning_rate": 1.2560127855885073e-05, "loss": 0.0103, "step": 7810 }, { "epoch": 10.303030303030303, "grad_norm": 0.16834335029125214, "learning_rate": 1.2450740975259745e-05, "loss": 0.0087, "step": 7820 }, { "epoch": 10.316205533596838, "grad_norm": 0.1898210048675537, "learning_rate": 1.234176472731517e-05, "loss": 0.0106, "step": 7830 }, { "epoch": 10.329380764163373, "grad_norm": 0.23814056813716888, "learning_rate": 1.2233200303798158e-05, "loss": 0.013, "step": 7840 }, { "epoch": 10.342555994729908, "grad_norm": 0.20332714915275574, "learning_rate": 1.2125048891951846e-05, "loss": 0.0103, "step": 7850 }, { "epoch": 10.355731225296443, "grad_norm": 0.12532776594161987, "learning_rate": 1.2017311674502745e-05, "loss": 0.0076, "step": 7860 }, { "epoch": 10.368906455862978, "grad_norm": 0.12927134335041046, "learning_rate": 1.1909989829647822e-05, "loss": 0.0075, "step": 7870 }, { "epoch": 10.382081686429512, "grad_norm": 0.22005397081375122, "learning_rate": 1.1803084531041553e-05, "loss": 0.0095, "step": 7880 }, { "epoch": 10.395256916996047, "grad_norm": 0.20675811171531677, "learning_rate": 1.1696596947783162e-05, "loss": 0.0101, "step": 7890 }, { "epoch": 10.408432147562582, "grad_norm": 0.27455243468284607, "learning_rate": 1.1590528244403803e-05, "loss": 0.0093, "step": 7900 }, { "epoch": 10.421607378129117, "grad_norm": 0.23391202092170715, "learning_rate": 1.148487958085382e-05, "loss": 0.008, "step": 7910 }, { "epoch": 10.434782608695652, "grad_norm": 0.17565032839775085, "learning_rate": 1.1379652112490086e-05, "loss": 0.0099, "step": 7920 }, { "epoch": 10.447957839262187, "grad_norm": 0.14462821185588837, "learning_rate": 1.1274846990063315e-05, "loss": 0.0072, "step": 7930 }, { "epoch": 10.461133069828723, "grad_norm": 0.1348995417356491, "learning_rate": 1.117046535970554e-05, "loss": 0.0098, "step": 7940 }, { "epoch": 10.474308300395258, "grad_norm": 0.17276756465435028, "learning_rate": 1.106650836291755e-05, "loss": 0.0109, "step": 7950 }, { "epoch": 10.487483530961791, "grad_norm": 0.17429892718791962, "learning_rate": 1.0962977136556418e-05, "loss": 0.0157, "step": 7960 }, { "epoch": 10.500658761528326, "grad_norm": 0.22026081383228302, "learning_rate": 1.0859872812823024e-05, "loss": 0.0108, "step": 7970 }, { "epoch": 10.513833992094861, "grad_norm": 0.14811038970947266, "learning_rate": 1.0757196519249747e-05, "loss": 0.0119, "step": 7980 }, { "epoch": 10.527009222661396, "grad_norm": 0.19848553836345673, "learning_rate": 1.0654949378688077e-05, "loss": 0.0112, "step": 7990 }, { "epoch": 10.540184453227932, "grad_norm": 0.22085192799568176, "learning_rate": 1.0553132509296376e-05, "loss": 0.009, "step": 8000 }, { "epoch": 10.553359683794467, "grad_norm": 0.15261310338974, "learning_rate": 1.0451747024527613e-05, "loss": 0.0103, "step": 8010 }, { "epoch": 10.566534914361002, "grad_norm": 0.13214528560638428, "learning_rate": 1.0350794033117189e-05, "loss": 0.0107, "step": 8020 }, { "epoch": 10.579710144927537, "grad_norm": 0.22246448695659637, "learning_rate": 1.0250274639070856e-05, "loss": 0.0096, "step": 8030 }, { "epoch": 10.59288537549407, "grad_norm": 0.18928980827331543, "learning_rate": 1.0150189941652599e-05, "loss": 0.0093, "step": 8040 }, { "epoch": 10.606060606060606, "grad_norm": 0.15583442151546478, "learning_rate": 1.0050541035372635e-05, "loss": 0.0084, "step": 8050 }, { "epoch": 10.61923583662714, "grad_norm": 0.1976160854101181, "learning_rate": 9.951329009975458e-06, "loss": 0.0079, "step": 8060 }, { "epoch": 10.632411067193676, "grad_norm": 0.2328920215368271, "learning_rate": 9.852554950427845e-06, "loss": 0.0084, "step": 8070 }, { "epoch": 10.645586297760211, "grad_norm": 0.19469021260738373, "learning_rate": 9.754219936907105e-06, "loss": 0.0079, "step": 8080 }, { "epoch": 10.658761528326746, "grad_norm": 0.16853180527687073, "learning_rate": 9.656325044789194e-06, "loss": 0.0082, "step": 8090 }, { "epoch": 10.671936758893281, "grad_norm": 0.10395480692386627, "learning_rate": 9.55887134463697e-06, "loss": 0.0082, "step": 8100 }, { "epoch": 10.685111989459816, "grad_norm": 0.18008802831172943, "learning_rate": 9.461859902188475e-06, "loss": 0.0078, "step": 8110 }, { "epoch": 10.69828722002635, "grad_norm": 0.22644630074501038, "learning_rate": 9.365291778345303e-06, "loss": 0.009, "step": 8120 }, { "epoch": 10.711462450592885, "grad_norm": 0.10834449529647827, "learning_rate": 9.269168029160991e-06, "loss": 0.0069, "step": 8130 }, { "epoch": 10.72463768115942, "grad_norm": 0.12147624790668488, "learning_rate": 9.173489705829447e-06, "loss": 0.008, "step": 8140 }, { "epoch": 10.737812911725955, "grad_norm": 0.1636052131652832, "learning_rate": 9.078257854673516e-06, "loss": 0.0093, "step": 8150 }, { "epoch": 10.75098814229249, "grad_norm": 0.21793915331363678, "learning_rate": 8.983473517133429e-06, "loss": 0.0076, "step": 8160 }, { "epoch": 10.764163372859025, "grad_norm": 0.1545238196849823, "learning_rate": 8.889137729755537e-06, "loss": 0.0073, "step": 8170 }, { "epoch": 10.77733860342556, "grad_norm": 0.1943051815032959, "learning_rate": 8.79525152418087e-06, "loss": 0.0086, "step": 8180 }, { "epoch": 10.790513833992096, "grad_norm": 0.241092249751091, "learning_rate": 8.701815927133961e-06, "loss": 0.0097, "step": 8190 }, { "epoch": 10.803689064558629, "grad_norm": 0.16625875234603882, "learning_rate": 8.608831960411534e-06, "loss": 0.0083, "step": 8200 }, { "epoch": 10.816864295125164, "grad_norm": 0.1712971031665802, "learning_rate": 8.516300640871321e-06, "loss": 0.0086, "step": 8210 }, { "epoch": 10.8300395256917, "grad_norm": 0.19201204180717468, "learning_rate": 8.424222980421038e-06, "loss": 0.0099, "step": 8220 }, { "epoch": 10.843214756258234, "grad_norm": 0.13061735033988953, "learning_rate": 8.332599986007184e-06, "loss": 0.0074, "step": 8230 }, { "epoch": 10.85638998682477, "grad_norm": 0.18794460594654083, "learning_rate": 8.241432659604203e-06, "loss": 0.0095, "step": 8240 }, { "epoch": 10.869565217391305, "grad_norm": 0.2249092310667038, "learning_rate": 8.150721998203331e-06, "loss": 0.008, "step": 8250 }, { "epoch": 10.88274044795784, "grad_norm": 0.2575426697731018, "learning_rate": 8.06046899380184e-06, "loss": 0.0068, "step": 8260 }, { "epoch": 10.895915678524375, "grad_norm": 0.2082395702600479, "learning_rate": 7.970674633392133e-06, "loss": 0.0083, "step": 8270 }, { "epoch": 10.909090909090908, "grad_norm": 0.17624586820602417, "learning_rate": 7.881339898950924e-06, "loss": 0.0111, "step": 8280 }, { "epoch": 10.922266139657443, "grad_norm": 0.13988056778907776, "learning_rate": 7.792465767428597e-06, "loss": 0.0079, "step": 8290 }, { "epoch": 10.935441370223979, "grad_norm": 0.2388487309217453, "learning_rate": 7.704053210738376e-06, "loss": 0.0086, "step": 8300 }, { "epoch": 10.948616600790514, "grad_norm": 0.15359775722026825, "learning_rate": 7.6161031957458494e-06, "loss": 0.0051, "step": 8310 }, { "epoch": 10.961791831357049, "grad_norm": 0.28482502698898315, "learning_rate": 7.5286166842582605e-06, "loss": 0.0122, "step": 8320 }, { "epoch": 10.974967061923584, "grad_norm": 0.18701320886611938, "learning_rate": 7.4415946330140814e-06, "loss": 0.0091, "step": 8330 }, { "epoch": 10.988142292490119, "grad_norm": 0.12363678216934204, "learning_rate": 7.3550379936725644e-06, "loss": 0.0091, "step": 8340 }, { "epoch": 11.001317523056654, "grad_norm": 0.21663540601730347, "learning_rate": 7.2689477128032035e-06, "loss": 0.0098, "step": 8350 }, { "epoch": 11.014492753623188, "grad_norm": 0.18112725019454956, "learning_rate": 7.183324731875551e-06, "loss": 0.0082, "step": 8360 }, { "epoch": 11.027667984189723, "grad_norm": 0.16493700444698334, "learning_rate": 7.098169987248782e-06, "loss": 0.0078, "step": 8370 }, { "epoch": 11.040843214756258, "grad_norm": 0.13823840022087097, "learning_rate": 7.013484410161553e-06, "loss": 0.0056, "step": 8380 }, { "epoch": 11.054018445322793, "grad_norm": 0.2114834487438202, "learning_rate": 6.92926892672176e-06, "loss": 0.0098, "step": 8390 }, { "epoch": 11.067193675889328, "grad_norm": 0.2341718226671219, "learning_rate": 6.845524457896446e-06, "loss": 0.0111, "step": 8400 }, { "epoch": 11.080368906455863, "grad_norm": 0.16323405504226685, "learning_rate": 6.7622519195017165e-06, "loss": 0.0059, "step": 8410 }, { "epoch": 11.093544137022398, "grad_norm": 0.25901684165000916, "learning_rate": 6.679452222192684e-06, "loss": 0.0092, "step": 8420 }, { "epoch": 11.106719367588934, "grad_norm": 0.11953702569007874, "learning_rate": 6.597126271453579e-06, "loss": 0.0066, "step": 8430 }, { "epoch": 11.119894598155467, "grad_norm": 0.13243408501148224, "learning_rate": 6.51527496758782e-06, "loss": 0.0069, "step": 8440 }, { "epoch": 11.133069828722002, "grad_norm": 0.119338758289814, "learning_rate": 6.433899205708155e-06, "loss": 0.0079, "step": 8450 }, { "epoch": 11.146245059288537, "grad_norm": 0.21323683857917786, "learning_rate": 6.352999875726856e-06, "loss": 0.0078, "step": 8460 }, { "epoch": 11.159420289855072, "grad_norm": 0.21504490077495575, "learning_rate": 6.272577862346052e-06, "loss": 0.0062, "step": 8470 }, { "epoch": 11.172595520421607, "grad_norm": 0.17403605580329895, "learning_rate": 6.192634045047996e-06, "loss": 0.0086, "step": 8480 }, { "epoch": 11.185770750988143, "grad_norm": 0.11534730345010757, "learning_rate": 6.113169298085458e-06, "loss": 0.0138, "step": 8490 }, { "epoch": 11.198945981554678, "grad_norm": 0.1640927791595459, "learning_rate": 6.034184490472195e-06, "loss": 0.0094, "step": 8500 }, { "epoch": 11.212121212121213, "grad_norm": 0.11937487870454788, "learning_rate": 5.955680485973386e-06, "loss": 0.0082, "step": 8510 }, { "epoch": 11.225296442687746, "grad_norm": 0.1374310553073883, "learning_rate": 5.877658143096265e-06, "loss": 0.0079, "step": 8520 }, { "epoch": 11.238471673254281, "grad_norm": 0.2548101246356964, "learning_rate": 5.800118315080661e-06, "loss": 0.0084, "step": 8530 }, { "epoch": 11.251646903820816, "grad_norm": 0.1334235519170761, "learning_rate": 5.723061849889716e-06, "loss": 0.007, "step": 8540 }, { "epoch": 11.264822134387352, "grad_norm": 0.1807614266872406, "learning_rate": 5.646489590200604e-06, "loss": 0.0073, "step": 8550 }, { "epoch": 11.277997364953887, "grad_norm": 0.22683222591876984, "learning_rate": 5.570402373395256e-06, "loss": 0.0082, "step": 8560 }, { "epoch": 11.291172595520422, "grad_norm": 0.16206318140029907, "learning_rate": 5.494801031551305e-06, "loss": 0.0076, "step": 8570 }, { "epoch": 11.304347826086957, "grad_norm": 0.11189868301153183, "learning_rate": 5.41968639143291e-06, "loss": 0.0088, "step": 8580 }, { "epoch": 11.317523056653492, "grad_norm": 0.11796924471855164, "learning_rate": 5.345059274481751e-06, "loss": 0.0074, "step": 8590 }, { "epoch": 11.330698287220027, "grad_norm": 0.23435086011886597, "learning_rate": 5.270920496808002e-06, "loss": 0.0078, "step": 8600 }, { "epoch": 11.34387351778656, "grad_norm": 0.13652624189853668, "learning_rate": 5.1972708691814695e-06, "loss": 0.0078, "step": 8610 }, { "epoch": 11.357048748353096, "grad_norm": 0.17509783804416656, "learning_rate": 5.124111197022674e-06, "loss": 0.0082, "step": 8620 }, { "epoch": 11.37022397891963, "grad_norm": 0.1447538137435913, "learning_rate": 5.051442280394081e-06, "loss": 0.0087, "step": 8630 }, { "epoch": 11.383399209486166, "grad_norm": 0.11106734722852707, "learning_rate": 4.979264913991322e-06, "loss": 0.0067, "step": 8640 }, { "epoch": 11.396574440052701, "grad_norm": 0.16821962594985962, "learning_rate": 4.907579887134489e-06, "loss": 0.0076, "step": 8650 }, { "epoch": 11.409749670619236, "grad_norm": 0.11497115343809128, "learning_rate": 4.836387983759572e-06, "loss": 0.0089, "step": 8660 }, { "epoch": 11.422924901185771, "grad_norm": 0.15908224880695343, "learning_rate": 4.765689982409816e-06, "loss": 0.0083, "step": 8670 }, { "epoch": 11.436100131752305, "grad_norm": 0.12201621383428574, "learning_rate": 4.695486656227233e-06, "loss": 0.0073, "step": 8680 }, { "epoch": 11.44927536231884, "grad_norm": 0.11300718039274216, "learning_rate": 4.625778772944156e-06, "loss": 0.01, "step": 8690 }, { "epoch": 11.462450592885375, "grad_norm": 0.11321785300970078, "learning_rate": 4.556567094874825e-06, "loss": 0.0066, "step": 8700 }, { "epoch": 11.47562582345191, "grad_norm": 0.11770084500312805, "learning_rate": 4.487852378907059e-06, "loss": 0.0075, "step": 8710 }, { "epoch": 11.488801054018445, "grad_norm": 0.18582583963871002, "learning_rate": 4.419635376493986e-06, "loss": 0.0061, "step": 8720 }, { "epoch": 11.50197628458498, "grad_norm": 0.1639215350151062, "learning_rate": 4.351916833645825e-06, "loss": 0.0081, "step": 8730 }, { "epoch": 11.515151515151516, "grad_norm": 0.0973479226231575, "learning_rate": 4.284697490921691e-06, "loss": 0.0056, "step": 8740 }, { "epoch": 11.52832674571805, "grad_norm": 0.08141589164733887, "learning_rate": 4.2179780834215585e-06, "loss": 0.0076, "step": 8750 }, { "epoch": 11.541501976284586, "grad_norm": 0.29469650983810425, "learning_rate": 4.151759340778178e-06, "loss": 0.0067, "step": 8760 }, { "epoch": 11.55467720685112, "grad_norm": 0.0893554612994194, "learning_rate": 4.086041987149109e-06, "loss": 0.0076, "step": 8770 }, { "epoch": 11.567852437417654, "grad_norm": 0.16742943227291107, "learning_rate": 4.020826741208811e-06, "loss": 0.0083, "step": 8780 }, { "epoch": 11.58102766798419, "grad_norm": 0.14368514716625214, "learning_rate": 3.956114316140746e-06, "loss": 0.0099, "step": 8790 }, { "epoch": 11.594202898550725, "grad_norm": 0.17125031352043152, "learning_rate": 3.891905419629643e-06, "loss": 0.0108, "step": 8800 }, { "epoch": 11.60737812911726, "grad_norm": 0.19009582698345184, "learning_rate": 3.8282007538536946e-06, "loss": 0.0105, "step": 8810 }, { "epoch": 11.620553359683795, "grad_norm": 0.15681231021881104, "learning_rate": 3.7650010154769265e-06, "loss": 0.006, "step": 8820 }, { "epoch": 11.63372859025033, "grad_norm": 0.09388696402311325, "learning_rate": 3.7023068956415608e-06, "loss": 0.0084, "step": 8830 }, { "epoch": 11.646903820816863, "grad_norm": 0.14405575394630432, "learning_rate": 3.6401190799604303e-06, "loss": 0.0078, "step": 8840 }, { "epoch": 11.660079051383399, "grad_norm": 0.07782807946205139, "learning_rate": 3.578438248509536e-06, "loss": 0.0071, "step": 8850 }, { "epoch": 11.673254281949934, "grad_norm": 0.18701545894145966, "learning_rate": 3.5172650758205583e-06, "loss": 0.0085, "step": 8860 }, { "epoch": 11.686429512516469, "grad_norm": 0.2714499533176422, "learning_rate": 3.45660023087353e-06, "loss": 0.0072, "step": 8870 }, { "epoch": 11.699604743083004, "grad_norm": 0.14982731640338898, "learning_rate": 3.3964443770894528e-06, "loss": 0.0069, "step": 8880 }, { "epoch": 11.712779973649539, "grad_norm": 0.14660252630710602, "learning_rate": 3.3367981723231245e-06, "loss": 0.0068, "step": 8890 }, { "epoch": 11.725955204216074, "grad_norm": 0.14174430072307587, "learning_rate": 3.2776622688558746e-06, "loss": 0.0067, "step": 8900 }, { "epoch": 11.73913043478261, "grad_norm": 0.15731292963027954, "learning_rate": 3.2190373133884677e-06, "loss": 0.0093, "step": 8910 }, { "epoch": 11.752305665349144, "grad_norm": 0.14486129581928253, "learning_rate": 3.1609239470340446e-06, "loss": 0.0065, "step": 8920 }, { "epoch": 11.765480895915678, "grad_norm": 0.11417282372713089, "learning_rate": 3.1033228053110373e-06, "loss": 0.0064, "step": 8930 }, { "epoch": 11.778656126482213, "grad_norm": 0.089751698076725, "learning_rate": 3.0462345181363314e-06, "loss": 0.0067, "step": 8940 }, { "epoch": 11.791831357048748, "grad_norm": 0.17435140907764435, "learning_rate": 2.9896597098182654e-06, "loss": 0.0092, "step": 8950 }, { "epoch": 11.805006587615283, "grad_norm": 0.07187920808792114, "learning_rate": 2.933598999049891e-06, "loss": 0.0079, "step": 8960 }, { "epoch": 11.818181818181818, "grad_norm": 0.1424235999584198, "learning_rate": 2.8780529989021697e-06, "loss": 0.0083, "step": 8970 }, { "epoch": 11.831357048748353, "grad_norm": 0.12420742958784103, "learning_rate": 2.823022316817242e-06, "loss": 0.007, "step": 8980 }, { "epoch": 11.844532279314889, "grad_norm": 0.1716492623090744, "learning_rate": 2.7685075546018456e-06, "loss": 0.0105, "step": 8990 }, { "epoch": 11.857707509881424, "grad_norm": 0.2577267587184906, "learning_rate": 2.7145093084206598e-06, "loss": 0.0082, "step": 9000 }, { "epoch": 11.870882740447957, "grad_norm": 0.08861849457025528, "learning_rate": 2.661028168789892e-06, "loss": 0.0087, "step": 9010 }, { "epoch": 11.884057971014492, "grad_norm": 0.19629858434200287, "learning_rate": 2.6080647205706855e-06, "loss": 0.0088, "step": 9020 }, { "epoch": 11.897233201581027, "grad_norm": 0.1813318282365799, "learning_rate": 2.555619542962834e-06, "loss": 0.0065, "step": 9030 }, { "epoch": 11.910408432147563, "grad_norm": 0.10713133215904236, "learning_rate": 2.503693209498409e-06, "loss": 0.0052, "step": 9040 }, { "epoch": 11.923583662714098, "grad_norm": 0.17797449231147766, "learning_rate": 2.452286288035449e-06, "loss": 0.0072, "step": 9050 }, { "epoch": 11.936758893280633, "grad_norm": 0.11036095023155212, "learning_rate": 2.4013993407518363e-06, "loss": 0.0068, "step": 9060 }, { "epoch": 11.949934123847168, "grad_norm": 0.1084558367729187, "learning_rate": 2.351032924139063e-06, "loss": 0.0075, "step": 9070 }, { "epoch": 11.963109354413703, "grad_norm": 0.20317208766937256, "learning_rate": 2.30118758899619e-06, "loss": 0.0103, "step": 9080 }, { "epoch": 11.976284584980236, "grad_norm": 0.12911272048950195, "learning_rate": 2.2518638804238157e-06, "loss": 0.005, "step": 9090 }, { "epoch": 11.989459815546772, "grad_norm": 0.1062001958489418, "learning_rate": 2.203062337818118e-06, "loss": 0.0071, "step": 9100 }, { "epoch": 12.002635046113307, "grad_norm": 0.16308660805225372, "learning_rate": 2.1547834948649483e-06, "loss": 0.0112, "step": 9110 }, { "epoch": 12.015810276679842, "grad_norm": 0.18296882510185242, "learning_rate": 2.1070278795340017e-06, "loss": 0.0072, "step": 9120 }, { "epoch": 12.028985507246377, "grad_norm": 0.18580010533332825, "learning_rate": 2.059796014073029e-06, "loss": 0.0068, "step": 9130 }, { "epoch": 12.042160737812912, "grad_norm": 0.1033332422375679, "learning_rate": 2.01308841500214e-06, "loss": 0.0051, "step": 9140 }, { "epoch": 12.055335968379447, "grad_norm": 0.1866554468870163, "learning_rate": 1.9669055931081704e-06, "loss": 0.0073, "step": 9150 }, { "epoch": 12.068511198945982, "grad_norm": 0.26920342445373535, "learning_rate": 1.9212480534390507e-06, "loss": 0.0089, "step": 9160 }, { "epoch": 12.081686429512516, "grad_norm": 0.18751156330108643, "learning_rate": 1.8761162952983246e-06, "loss": 0.0085, "step": 9170 }, { "epoch": 12.09486166007905, "grad_norm": 0.3206658959388733, "learning_rate": 1.8315108122396618e-06, "loss": 0.0111, "step": 9180 }, { "epoch": 12.108036890645586, "grad_norm": 0.1985057294368744, "learning_rate": 1.787432092061475e-06, "loss": 0.009, "step": 9190 }, { "epoch": 12.121212121212121, "grad_norm": 0.17330951988697052, "learning_rate": 1.743880616801602e-06, "loss": 0.0072, "step": 9200 }, { "epoch": 12.134387351778656, "grad_norm": 0.08973265439271927, "learning_rate": 1.7008568627319865e-06, "loss": 0.0076, "step": 9210 }, { "epoch": 12.147562582345191, "grad_norm": 0.23255111277103424, "learning_rate": 1.6583613003535226e-06, "loss": 0.0078, "step": 9220 }, { "epoch": 12.160737812911727, "grad_norm": 0.09232556819915771, "learning_rate": 1.6163943943908522e-06, "loss": 0.0104, "step": 9230 }, { "epoch": 12.173913043478262, "grad_norm": 0.1634305715560913, "learning_rate": 1.5749566037873476e-06, "loss": 0.0063, "step": 9240 }, { "epoch": 12.187088274044795, "grad_norm": 0.11620501428842545, "learning_rate": 1.5340483817000428e-06, "loss": 0.0067, "step": 9250 }, { "epoch": 12.20026350461133, "grad_norm": 0.15891708433628082, "learning_rate": 1.4936701754947101e-06, "loss": 0.0085, "step": 9260 }, { "epoch": 12.213438735177865, "grad_norm": 0.21395912766456604, "learning_rate": 1.4538224267409361e-06, "loss": 0.0068, "step": 9270 }, { "epoch": 12.2266139657444, "grad_norm": 0.11466138064861298, "learning_rate": 1.414505571207314e-06, "loss": 0.0069, "step": 9280 }, { "epoch": 12.239789196310936, "grad_norm": 0.2666526734828949, "learning_rate": 1.3757200388566816e-06, "loss": 0.0064, "step": 9290 }, { "epoch": 12.25296442687747, "grad_norm": 0.10765407234430313, "learning_rate": 1.3374662538414074e-06, "loss": 0.0065, "step": 9300 }, { "epoch": 12.266139657444006, "grad_norm": 0.13771207630634308, "learning_rate": 1.2997446344987617e-06, "loss": 0.0095, "step": 9310 }, { "epoch": 12.279314888010541, "grad_norm": 0.15090085566043854, "learning_rate": 1.262555593346315e-06, "loss": 0.0077, "step": 9320 }, { "epoch": 12.292490118577074, "grad_norm": 0.15205277502536774, "learning_rate": 1.2258995370774685e-06, "loss": 0.008, "step": 9330 }, { "epoch": 12.30566534914361, "grad_norm": 0.12167327851057053, "learning_rate": 1.1897768665569798e-06, "loss": 0.0076, "step": 9340 }, { "epoch": 12.318840579710145, "grad_norm": 0.10922878235578537, "learning_rate": 1.1541879768165954e-06, "loss": 0.0079, "step": 9350 }, { "epoch": 12.33201581027668, "grad_norm": 0.0954914391040802, "learning_rate": 1.1191332570507085e-06, "loss": 0.0077, "step": 9360 }, { "epoch": 12.345191040843215, "grad_norm": 0.18894347548484802, "learning_rate": 1.0846130906121132e-06, "loss": 0.0092, "step": 9370 }, { "epoch": 12.35836627140975, "grad_norm": 0.25298139452934265, "learning_rate": 1.0506278550078131e-06, "loss": 0.0084, "step": 9380 }, { "epoch": 12.371541501976285, "grad_norm": 0.10267756134271622, "learning_rate": 1.0171779218949185e-06, "loss": 0.0088, "step": 9390 }, { "epoch": 12.38471673254282, "grad_norm": 0.19973057508468628, "learning_rate": 9.842636570765174e-07, "loss": 0.0077, "step": 9400 }, { "epoch": 12.397891963109354, "grad_norm": 0.12004857510328293, "learning_rate": 9.518854204977612e-07, "loss": 0.0064, "step": 9410 }, { "epoch": 12.411067193675889, "grad_norm": 0.1885608583688736, "learning_rate": 9.200435662418349e-07, "loss": 0.0061, "step": 9420 }, { "epoch": 12.424242424242424, "grad_norm": 0.19836807250976562, "learning_rate": 8.887384425261658e-07, "loss": 0.0076, "step": 9430 }, { "epoch": 12.437417654808959, "grad_norm": 0.0960315391421318, "learning_rate": 8.579703916985648e-07, "loss": 0.007, "step": 9440 }, { "epoch": 12.450592885375494, "grad_norm": 0.16369284689426422, "learning_rate": 8.277397502335194e-07, "loss": 0.0091, "step": 9450 }, { "epoch": 12.46376811594203, "grad_norm": 0.10038845986127853, "learning_rate": 7.980468487284675e-07, "loss": 0.0064, "step": 9460 }, { "epoch": 12.476943346508564, "grad_norm": 0.22409868240356445, "learning_rate": 7.688920119002297e-07, "loss": 0.0084, "step": 9470 }, { "epoch": 12.4901185770751, "grad_norm": 0.11471178382635117, "learning_rate": 7.402755585814269e-07, "loss": 0.0068, "step": 9480 }, { "epoch": 12.503293807641633, "grad_norm": 0.14783388376235962, "learning_rate": 7.121978017170073e-07, "loss": 0.0068, "step": 9490 }, { "epoch": 12.516469038208168, "grad_norm": 0.12338835746049881, "learning_rate": 6.846590483608306e-07, "loss": 0.009, "step": 9500 }, { "epoch": 12.529644268774703, "grad_norm": 0.17950797080993652, "learning_rate": 6.576595996722834e-07, "loss": 0.0094, "step": 9510 }, { "epoch": 12.542819499341238, "grad_norm": 0.20151561498641968, "learning_rate": 6.311997509130141e-07, "loss": 0.0083, "step": 9520 }, { "epoch": 12.555994729907773, "grad_norm": 0.2703305184841156, "learning_rate": 6.052797914436803e-07, "loss": 0.0065, "step": 9530 }, { "epoch": 12.569169960474309, "grad_norm": 0.11729317158460617, "learning_rate": 5.799000047208181e-07, "loss": 0.0077, "step": 9540 }, { "epoch": 12.582345191040844, "grad_norm": 0.13801559805870056, "learning_rate": 5.550606682937054e-07, "loss": 0.0089, "step": 9550 }, { "epoch": 12.595520421607379, "grad_norm": 0.10463900119066238, "learning_rate": 5.307620538013481e-07, "loss": 0.0084, "step": 9560 }, { "epoch": 12.608695652173914, "grad_norm": 0.13275998830795288, "learning_rate": 5.070044269694874e-07, "loss": 0.0106, "step": 9570 }, { "epoch": 12.621870882740447, "grad_norm": 0.11285616457462311, "learning_rate": 4.837880476077417e-07, "loss": 0.0068, "step": 9580 }, { "epoch": 12.635046113306982, "grad_norm": 0.10838264226913452, "learning_rate": 4.6111316960670835e-07, "loss": 0.0093, "step": 9590 }, { "epoch": 12.648221343873518, "grad_norm": 0.1535099595785141, "learning_rate": 4.389800409352218e-07, "loss": 0.0066, "step": 9600 }, { "epoch": 12.661396574440053, "grad_norm": 0.1464422047138214, "learning_rate": 4.173889036376277e-07, "loss": 0.0058, "step": 9610 }, { "epoch": 12.674571805006588, "grad_norm": 0.17395055294036865, "learning_rate": 3.963399938311463e-07, "loss": 0.0088, "step": 9620 }, { "epoch": 12.687747035573123, "grad_norm": 0.19919253885746002, "learning_rate": 3.7583354170328545e-07, "loss": 0.0073, "step": 9630 }, { "epoch": 12.700922266139658, "grad_norm": 0.10250774025917053, "learning_rate": 3.558697715093207e-07, "loss": 0.0088, "step": 9640 }, { "epoch": 12.714097496706191, "grad_norm": 0.15123838186264038, "learning_rate": 3.3644890156983576e-07, "loss": 0.0062, "step": 9650 }, { "epoch": 12.727272727272727, "grad_norm": 0.13388404250144958, "learning_rate": 3.175711442683638e-07, "loss": 0.0087, "step": 9660 }, { "epoch": 12.740447957839262, "grad_norm": 0.12210414558649063, "learning_rate": 2.9923670604902197e-07, "loss": 0.0069, "step": 9670 }, { "epoch": 12.753623188405797, "grad_norm": 0.16462716460227966, "learning_rate": 2.814457874143028e-07, "loss": 0.0068, "step": 9680 }, { "epoch": 12.766798418972332, "grad_norm": 0.18494081497192383, "learning_rate": 2.641985829228366e-07, "loss": 0.0098, "step": 9690 }, { "epoch": 12.779973649538867, "grad_norm": 0.09822367131710052, "learning_rate": 2.474952811872877e-07, "loss": 0.0066, "step": 9700 }, { "epoch": 12.793148880105402, "grad_norm": 0.19610926508903503, "learning_rate": 2.3133606487228397e-07, "loss": 0.0069, "step": 9710 }, { "epoch": 12.806324110671937, "grad_norm": 0.21302397549152374, "learning_rate": 2.157211106924295e-07, "loss": 0.0094, "step": 9720 }, { "epoch": 12.819499341238473, "grad_norm": 0.30115124583244324, "learning_rate": 2.006505894103672e-07, "loss": 0.0072, "step": 9730 }, { "epoch": 12.832674571805006, "grad_norm": 0.15520079433918, "learning_rate": 1.8612466583489696e-07, "loss": 0.0077, "step": 9740 }, { "epoch": 12.845849802371541, "grad_norm": 0.08769290894269943, "learning_rate": 1.7214349881918834e-07, "loss": 0.0056, "step": 9750 }, { "epoch": 12.859025032938076, "grad_norm": 0.09021688997745514, "learning_rate": 1.5870724125904845e-07, "loss": 0.0071, "step": 9760 }, { "epoch": 12.872200263504611, "grad_norm": 0.13328352570533752, "learning_rate": 1.4581604009124006e-07, "loss": 0.0076, "step": 9770 }, { "epoch": 12.885375494071146, "grad_norm": 0.13384944200515747, "learning_rate": 1.334700362918717e-07, "loss": 0.0061, "step": 9780 }, { "epoch": 12.898550724637682, "grad_norm": 0.13545677065849304, "learning_rate": 1.2166936487486015e-07, "loss": 0.0064, "step": 9790 }, { "epoch": 12.911725955204217, "grad_norm": 0.18425719439983368, "learning_rate": 1.1041415489045914e-07, "loss": 0.0076, "step": 9800 }, { "epoch": 12.92490118577075, "grad_norm": 0.19322925806045532, "learning_rate": 9.970452942384412e-08, "loss": 0.0076, "step": 9810 }, { "epoch": 12.938076416337285, "grad_norm": 0.1715928018093109, "learning_rate": 8.954060559375754e-08, "loss": 0.0085, "step": 9820 }, { "epoch": 12.95125164690382, "grad_norm": 0.1342555284500122, "learning_rate": 7.99224945512489e-08, "loss": 0.0057, "step": 9830 }, { "epoch": 12.964426877470355, "grad_norm": 0.2071145623922348, "learning_rate": 7.085030147843675e-08, "loss": 0.0107, "step": 9840 }, { "epoch": 12.97760210803689, "grad_norm": 0.14439895749092102, "learning_rate": 6.232412558736523e-08, "loss": 0.008, "step": 9850 }, { "epoch": 12.990777338603426, "grad_norm": 0.18717682361602783, "learning_rate": 5.434406011893822e-08, "loss": 0.0096, "step": 9860 }, { "epoch": 13.003952569169961, "grad_norm": 0.14880332350730896, "learning_rate": 4.6910192341864664e-08, "loss": 0.01, "step": 9870 }, { "epoch": 13.017127799736496, "grad_norm": 0.17164605855941772, "learning_rate": 4.0022603551737035e-08, "loss": 0.0057, "step": 9880 }, { "epoch": 13.030303030303031, "grad_norm": 0.22017565369606018, "learning_rate": 3.3681369070120985e-08, "loss": 0.0083, "step": 9890 }, { "epoch": 13.043478260869565, "grad_norm": 0.16133266687393188, "learning_rate": 2.7886558243744866e-08, "loss": 0.0095, "step": 9900 }, { "epoch": 13.0566534914361, "grad_norm": 0.14002366364002228, "learning_rate": 2.2638234443722596e-08, "loss": 0.0075, "step": 9910 }, { "epoch": 13.069828722002635, "grad_norm": 0.20343704521656036, "learning_rate": 1.7936455064887504e-08, "loss": 0.0078, "step": 9920 }, { "epoch": 13.08300395256917, "grad_norm": 0.09156999737024307, "learning_rate": 1.378127152514841e-08, "loss": 0.0083, "step": 9930 }, { "epoch": 13.096179183135705, "grad_norm": 0.16819879412651062, "learning_rate": 1.0172729264917857e-08, "loss": 0.0061, "step": 9940 }, { "epoch": 13.10935441370224, "grad_norm": 0.11606054753065109, "learning_rate": 7.1108677466458215e-09, "loss": 0.009, "step": 9950 }, { "epoch": 13.122529644268775, "grad_norm": 0.21958565711975098, "learning_rate": 4.595720454353414e-09, "loss": 0.0087, "step": 9960 }, { "epoch": 13.13570487483531, "grad_norm": 0.09595669060945511, "learning_rate": 2.627314893294264e-09, "loss": 0.0066, "step": 9970 }, { "epoch": 13.148880105401844, "grad_norm": 0.144814595580101, "learning_rate": 1.2056725896270048e-09, "loss": 0.0086, "step": 9980 }, { "epoch": 13.162055335968379, "grad_norm": 0.15428341925144196, "learning_rate": 3.308090902098826e-10, "loss": 0.007, "step": 9990 }, { "epoch": 13.175230566534914, "grad_norm": 0.1226944774389267, "learning_rate": 2.7339624120159555e-12, "loss": 0.0079, "step": 10000 }, { "epoch": 13.175230566534914, "step": 10000, "total_flos": 0.0, "train_loss": 0.021228669699653983, "train_runtime": 6638.0204, "train_samples_per_second": 48.207, "train_steps_per_second": 1.506 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 14, "save_steps": 1000, "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": 32, "trial_name": null, "trial_params": null }