{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.770433756270286, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0029507229271171437, "grad_norm": 11.860811233520508, "learning_rate": 3e-06, "loss": 1.0462, "step": 10 }, { "epoch": 0.005901445854234287, "grad_norm": 3.9827382564544678, "learning_rate": 6.333333333333334e-06, "loss": 0.7746, "step": 20 }, { "epoch": 0.00885216878135143, "grad_norm": 4.043567657470703, "learning_rate": 9.666666666666667e-06, "loss": 0.4418, "step": 30 }, { "epoch": 0.011802891708468575, "grad_norm": 1.4716283082962036, "learning_rate": 1.3000000000000001e-05, "loss": 0.2765, "step": 40 }, { "epoch": 0.014753614635585718, "grad_norm": 1.7400227785110474, "learning_rate": 1.6333333333333335e-05, "loss": 0.236, "step": 50 }, { "epoch": 0.01770433756270286, "grad_norm": 0.9207665920257568, "learning_rate": 1.9666666666666666e-05, "loss": 0.2212, "step": 60 }, { "epoch": 0.020655060489820007, "grad_norm": 1.5557454824447632, "learning_rate": 2.3000000000000003e-05, "loss": 0.1598, "step": 70 }, { "epoch": 0.02360578341693715, "grad_norm": 1.2374076843261719, "learning_rate": 2.633333333333333e-05, "loss": 0.1643, "step": 80 }, { "epoch": 0.026556506344054293, "grad_norm": 1.046493649482727, "learning_rate": 2.9666666666666672e-05, "loss": 0.1334, "step": 90 }, { "epoch": 0.029507229271171435, "grad_norm": 1.2764812707901, "learning_rate": 3.3e-05, "loss": 0.1329, "step": 100 }, { "epoch": 0.03245795219828858, "grad_norm": 1.71575927734375, "learning_rate": 3.633333333333333e-05, "loss": 0.1218, "step": 110 }, { "epoch": 0.03540867512540572, "grad_norm": 1.118678331375122, "learning_rate": 3.966666666666667e-05, "loss": 0.1076, "step": 120 }, { "epoch": 0.03835939805252287, "grad_norm": 0.7499859929084778, "learning_rate": 4.3e-05, "loss": 0.1108, "step": 130 }, { "epoch": 0.041310120979640014, "grad_norm": 0.6885630488395691, "learning_rate": 4.633333333333333e-05, "loss": 0.1043, "step": 140 }, { "epoch": 0.04426084390675716, "grad_norm": 1.3828084468841553, "learning_rate": 4.966666666666667e-05, "loss": 0.1081, "step": 150 }, { "epoch": 0.0472115668338743, "grad_norm": 1.1867084503173828, "learning_rate": 5.300000000000001e-05, "loss": 0.1087, "step": 160 }, { "epoch": 0.05016228976099144, "grad_norm": 0.7206751704216003, "learning_rate": 5.633333333333334e-05, "loss": 0.0956, "step": 170 }, { "epoch": 0.053113012688108585, "grad_norm": 1.3189795017242432, "learning_rate": 5.966666666666667e-05, "loss": 0.0997, "step": 180 }, { "epoch": 0.05606373561522573, "grad_norm": 1.300041913986206, "learning_rate": 6.3e-05, "loss": 0.0946, "step": 190 }, { "epoch": 0.05901445854234287, "grad_norm": 1.2459789514541626, "learning_rate": 6.633333333333334e-05, "loss": 0.0912, "step": 200 }, { "epoch": 0.06196518146946002, "grad_norm": 0.6687663793563843, "learning_rate": 6.966666666666668e-05, "loss": 0.0966, "step": 210 }, { "epoch": 0.06491590439657716, "grad_norm": 0.9948630332946777, "learning_rate": 7.3e-05, "loss": 0.0865, "step": 220 }, { "epoch": 0.0678666273236943, "grad_norm": 0.8532779812812805, "learning_rate": 7.633333333333334e-05, "loss": 0.0811, "step": 230 }, { "epoch": 0.07081735025081144, "grad_norm": 0.7049078941345215, "learning_rate": 7.966666666666666e-05, "loss": 0.0853, "step": 240 }, { "epoch": 0.07376807317792859, "grad_norm": 0.8651864528656006, "learning_rate": 8.3e-05, "loss": 0.0791, "step": 250 }, { "epoch": 0.07671879610504574, "grad_norm": 0.7383908629417419, "learning_rate": 8.633333333333334e-05, "loss": 0.0708, "step": 260 }, { "epoch": 0.07966951903216288, "grad_norm": 0.9129989743232727, "learning_rate": 8.966666666666666e-05, "loss": 0.0842, "step": 270 }, { "epoch": 0.08262024195928003, "grad_norm": 0.614589273929596, "learning_rate": 9.300000000000001e-05, "loss": 0.0733, "step": 280 }, { "epoch": 0.08557096488639716, "grad_norm": 1.0666412115097046, "learning_rate": 9.633333333333335e-05, "loss": 0.0882, "step": 290 }, { "epoch": 0.08852168781351431, "grad_norm": 0.8570053577423096, "learning_rate": 9.966666666666667e-05, "loss": 0.0734, "step": 300 }, { "epoch": 0.09147241074063145, "grad_norm": 0.39629802107810974, "learning_rate": 9.999938485971279e-05, "loss": 0.0771, "step": 310 }, { "epoch": 0.0944231336677486, "grad_norm": 0.5321000218391418, "learning_rate": 9.999725846827562e-05, "loss": 0.0718, "step": 320 }, { "epoch": 0.09737385659486575, "grad_norm": 0.7146950960159302, "learning_rate": 9.999361329594254e-05, "loss": 0.0686, "step": 330 }, { "epoch": 0.10032457952198288, "grad_norm": 0.8377274870872498, "learning_rate": 9.998844945344405e-05, "loss": 0.0634, "step": 340 }, { "epoch": 0.10327530244910003, "grad_norm": 0.49813681840896606, "learning_rate": 9.99817670976436e-05, "loss": 0.0603, "step": 350 }, { "epoch": 0.10622602537621717, "grad_norm": 0.799103319644928, "learning_rate": 9.997356643153303e-05, "loss": 0.0637, "step": 360 }, { "epoch": 0.10917674830333432, "grad_norm": 0.5847508311271667, "learning_rate": 9.996384770422629e-05, "loss": 0.0681, "step": 370 }, { "epoch": 0.11212747123045146, "grad_norm": 0.6221186518669128, "learning_rate": 9.995261121095194e-05, "loss": 0.0641, "step": 380 }, { "epoch": 0.1150781941575686, "grad_norm": 0.792294442653656, "learning_rate": 9.993985729304408e-05, "loss": 0.0644, "step": 390 }, { "epoch": 0.11802891708468574, "grad_norm": 0.6830067038536072, "learning_rate": 9.992558633793212e-05, "loss": 0.0594, "step": 400 }, { "epoch": 0.12097964001180289, "grad_norm": 0.738215446472168, "learning_rate": 9.990979877912891e-05, "loss": 0.0591, "step": 410 }, { "epoch": 0.12393036293892004, "grad_norm": 0.7751050591468811, "learning_rate": 9.989249509621759e-05, "loss": 0.0621, "step": 420 }, { "epoch": 0.1268810858660372, "grad_norm": 0.5941200256347656, "learning_rate": 9.987367581483705e-05, "loss": 0.0611, "step": 430 }, { "epoch": 0.1298318087931543, "grad_norm": 0.8491165041923523, "learning_rate": 9.985334150666592e-05, "loss": 0.0615, "step": 440 }, { "epoch": 0.13278253172027146, "grad_norm": 1.0582492351531982, "learning_rate": 9.983149278940526e-05, "loss": 0.061, "step": 450 }, { "epoch": 0.1357332546473886, "grad_norm": 0.5902143716812134, "learning_rate": 9.980813032675974e-05, "loss": 0.0683, "step": 460 }, { "epoch": 0.13868397757450576, "grad_norm": 0.8963476419448853, "learning_rate": 9.978325482841753e-05, "loss": 0.0684, "step": 470 }, { "epoch": 0.14163470050162288, "grad_norm": 0.5603164434432983, "learning_rate": 9.975686705002867e-05, "loss": 0.0716, "step": 480 }, { "epoch": 0.14458542342874003, "grad_norm": 0.8444815874099731, "learning_rate": 9.972896779318219e-05, "loss": 0.063, "step": 490 }, { "epoch": 0.14753614635585718, "grad_norm": 0.6846355199813843, "learning_rate": 9.969955790538175e-05, "loss": 0.0538, "step": 500 }, { "epoch": 0.15048686928297433, "grad_norm": 0.6709917187690735, "learning_rate": 9.966863828001982e-05, "loss": 0.0578, "step": 510 }, { "epoch": 0.15343759221009148, "grad_norm": 0.7113671898841858, "learning_rate": 9.963620985635065e-05, "loss": 0.0562, "step": 520 }, { "epoch": 0.1563883151372086, "grad_norm": 0.6490744948387146, "learning_rate": 9.960227361946164e-05, "loss": 0.0613, "step": 530 }, { "epoch": 0.15933903806432576, "grad_norm": 0.8408623337745667, "learning_rate": 9.95668306002435e-05, "loss": 0.0543, "step": 540 }, { "epoch": 0.1622897609914429, "grad_norm": 0.6774630546569824, "learning_rate": 9.952988187535886e-05, "loss": 0.0508, "step": 550 }, { "epoch": 0.16524048391856005, "grad_norm": 0.5782418251037598, "learning_rate": 9.949142856720961e-05, "loss": 0.0466, "step": 560 }, { "epoch": 0.1681912068456772, "grad_norm": 0.613166332244873, "learning_rate": 9.945147184390278e-05, "loss": 0.0431, "step": 570 }, { "epoch": 0.17114192977279433, "grad_norm": 0.39589640498161316, "learning_rate": 9.941001291921512e-05, "loss": 0.0527, "step": 580 }, { "epoch": 0.17409265269991148, "grad_norm": 0.5749406218528748, "learning_rate": 9.936705305255612e-05, "loss": 0.053, "step": 590 }, { "epoch": 0.17704337562702863, "grad_norm": 0.48604968190193176, "learning_rate": 9.932259354892984e-05, "loss": 0.0518, "step": 600 }, { "epoch": 0.17999409855414578, "grad_norm": 0.6258148550987244, "learning_rate": 9.927663575889521e-05, "loss": 0.0569, "step": 610 }, { "epoch": 0.1829448214812629, "grad_norm": 0.4342539310455322, "learning_rate": 9.922918107852504e-05, "loss": 0.0486, "step": 620 }, { "epoch": 0.18589554440838005, "grad_norm": 0.37068474292755127, "learning_rate": 9.918023094936363e-05, "loss": 0.0435, "step": 630 }, { "epoch": 0.1888462673354972, "grad_norm": 0.43584898114204407, "learning_rate": 9.912978685838294e-05, "loss": 0.0462, "step": 640 }, { "epoch": 0.19179699026261435, "grad_norm": 0.8911308646202087, "learning_rate": 9.90778503379374e-05, "loss": 0.0553, "step": 650 }, { "epoch": 0.1947477131897315, "grad_norm": 0.5372707843780518, "learning_rate": 9.902442296571743e-05, "loss": 0.047, "step": 660 }, { "epoch": 0.19769843611684862, "grad_norm": 0.5211854577064514, "learning_rate": 9.896950636470147e-05, "loss": 0.056, "step": 670 }, { "epoch": 0.20064915904396577, "grad_norm": 0.5618753433227539, "learning_rate": 9.891310220310666e-05, "loss": 0.0482, "step": 680 }, { "epoch": 0.20359988197108292, "grad_norm": 0.4594133198261261, "learning_rate": 9.885521219433823e-05, "loss": 0.0525, "step": 690 }, { "epoch": 0.20655060489820007, "grad_norm": 0.6655735969543457, "learning_rate": 9.879583809693738e-05, "loss": 0.0525, "step": 700 }, { "epoch": 0.2095013278253172, "grad_norm": 0.5805474519729614, "learning_rate": 9.873498171452789e-05, "loss": 0.06, "step": 710 }, { "epoch": 0.21245205075243434, "grad_norm": 0.7790851593017578, "learning_rate": 9.867264489576135e-05, "loss": 0.0529, "step": 720 }, { "epoch": 0.2154027736795515, "grad_norm": 0.5360324382781982, "learning_rate": 9.860882953426099e-05, "loss": 0.0524, "step": 730 }, { "epoch": 0.21835349660666864, "grad_norm": 0.5274640321731567, "learning_rate": 9.854353756856412e-05, "loss": 0.0408, "step": 740 }, { "epoch": 0.2213042195337858, "grad_norm": 0.4872770607471466, "learning_rate": 9.847677098206332e-05, "loss": 0.047, "step": 750 }, { "epoch": 0.2242549424609029, "grad_norm": 0.4077429175376892, "learning_rate": 9.840853180294608e-05, "loss": 0.042, "step": 760 }, { "epoch": 0.22720566538802006, "grad_norm": 0.6694160103797913, "learning_rate": 9.833882210413332e-05, "loss": 0.0469, "step": 770 }, { "epoch": 0.2301563883151372, "grad_norm": 0.5221421122550964, "learning_rate": 9.826764400321633e-05, "loss": 0.0458, "step": 780 }, { "epoch": 0.23310711124225436, "grad_norm": 0.2927101254463196, "learning_rate": 9.819499966239243e-05, "loss": 0.045, "step": 790 }, { "epoch": 0.23605783416937148, "grad_norm": 0.6948075294494629, "learning_rate": 9.812089128839938e-05, "loss": 0.0467, "step": 800 }, { "epoch": 0.23900855709648863, "grad_norm": 0.3236323893070221, "learning_rate": 9.804532113244828e-05, "loss": 0.047, "step": 810 }, { "epoch": 0.24195928002360578, "grad_norm": 0.5774182081222534, "learning_rate": 9.796829149015517e-05, "loss": 0.0483, "step": 820 }, { "epoch": 0.24491000295072293, "grad_norm": 0.8551205992698669, "learning_rate": 9.788980470147132e-05, "loss": 0.0492, "step": 830 }, { "epoch": 0.24786072587784008, "grad_norm": 0.6032458543777466, "learning_rate": 9.780986315061218e-05, "loss": 0.0446, "step": 840 }, { "epoch": 0.2508114488049572, "grad_norm": 0.71601802110672, "learning_rate": 9.772846926598491e-05, "loss": 0.0557, "step": 850 }, { "epoch": 0.2537621717320744, "grad_norm": 0.4930027723312378, "learning_rate": 9.76456255201146e-05, "loss": 0.0482, "step": 860 }, { "epoch": 0.2567128946591915, "grad_norm": 0.4278770983219147, "learning_rate": 9.756133442956923e-05, "loss": 0.0457, "step": 870 }, { "epoch": 0.2596636175863086, "grad_norm": 0.5457911491394043, "learning_rate": 9.747559855488313e-05, "loss": 0.0427, "step": 880 }, { "epoch": 0.2626143405134258, "grad_norm": 0.4986213147640228, "learning_rate": 9.73884205004793e-05, "loss": 0.0396, "step": 890 }, { "epoch": 0.2655650634405429, "grad_norm": 0.5596452951431274, "learning_rate": 9.729980291459019e-05, "loss": 0.0512, "step": 900 }, { "epoch": 0.2685157863676601, "grad_norm": 0.4846031367778778, "learning_rate": 9.720974848917735e-05, "loss": 0.0375, "step": 910 }, { "epoch": 0.2714665092947772, "grad_norm": 0.6028445363044739, "learning_rate": 9.711825995984957e-05, "loss": 0.0486, "step": 920 }, { "epoch": 0.27441723222189435, "grad_norm": 0.4547349810600281, "learning_rate": 9.702534010577991e-05, "loss": 0.0444, "step": 930 }, { "epoch": 0.2773679551490115, "grad_norm": 0.38289719820022583, "learning_rate": 9.693099174962103e-05, "loss": 0.0408, "step": 940 }, { "epoch": 0.28031867807612865, "grad_norm": 0.3622853457927704, "learning_rate": 9.683521775741977e-05, "loss": 0.0505, "step": 950 }, { "epoch": 0.28326940100324577, "grad_norm": 0.6771332025527954, "learning_rate": 9.673802103852979e-05, "loss": 0.0404, "step": 960 }, { "epoch": 0.28622012393036295, "grad_norm": 0.4589938521385193, "learning_rate": 9.663940454552342e-05, "loss": 0.0403, "step": 970 }, { "epoch": 0.28917084685748007, "grad_norm": 0.8690301775932312, "learning_rate": 9.65393712741018e-05, "loss": 0.0446, "step": 980 }, { "epoch": 0.29212156978459725, "grad_norm": 0.5445677042007446, "learning_rate": 9.6437924263004e-05, "loss": 0.0474, "step": 990 }, { "epoch": 0.29507229271171437, "grad_norm": 0.6575219035148621, "learning_rate": 9.63350665939146e-05, "loss": 0.0416, "step": 1000 }, { "epoch": 0.2980230156388315, "grad_norm": 0.42845404148101807, "learning_rate": 9.623080139137023e-05, "loss": 0.0382, "step": 1010 }, { "epoch": 0.30097373856594867, "grad_norm": 0.39788901805877686, "learning_rate": 9.612513182266447e-05, "loss": 0.0465, "step": 1020 }, { "epoch": 0.3039244614930658, "grad_norm": 0.3768392503261566, "learning_rate": 9.601806109775179e-05, "loss": 0.0412, "step": 1030 }, { "epoch": 0.30687518442018297, "grad_norm": 0.4149167537689209, "learning_rate": 9.590959246914995e-05, "loss": 0.0474, "step": 1040 }, { "epoch": 0.3098259073473001, "grad_norm": 0.44760340452194214, "learning_rate": 9.579972923184122e-05, "loss": 0.0356, "step": 1050 }, { "epoch": 0.3127766302744172, "grad_norm": 0.9699430465698242, "learning_rate": 9.568847472317232e-05, "loss": 0.0486, "step": 1060 }, { "epoch": 0.3157273532015344, "grad_norm": 0.7206645011901855, "learning_rate": 9.557583232275303e-05, "loss": 0.0415, "step": 1070 }, { "epoch": 0.3186780761286515, "grad_norm": 0.4414677321910858, "learning_rate": 9.546180545235344e-05, "loss": 0.0465, "step": 1080 }, { "epoch": 0.3216287990557687, "grad_norm": 0.8250203728675842, "learning_rate": 9.534639757580013e-05, "loss": 0.039, "step": 1090 }, { "epoch": 0.3245795219828858, "grad_norm": 0.37138691544532776, "learning_rate": 9.522961219887092e-05, "loss": 0.042, "step": 1100 }, { "epoch": 0.32753024491000293, "grad_norm": 0.49275052547454834, "learning_rate": 9.511145286918828e-05, "loss": 0.0431, "step": 1110 }, { "epoch": 0.3304809678371201, "grad_norm": 0.48843705654144287, "learning_rate": 9.499192317611167e-05, "loss": 0.0408, "step": 1120 }, { "epoch": 0.33343169076423723, "grad_norm": 0.46524518728256226, "learning_rate": 9.487102675062851e-05, "loss": 0.0486, "step": 1130 }, { "epoch": 0.3363824136913544, "grad_norm": 0.5163531303405762, "learning_rate": 9.474876726524374e-05, "loss": 0.0448, "step": 1140 }, { "epoch": 0.33933313661847153, "grad_norm": 0.32577982544898987, "learning_rate": 9.462514843386845e-05, "loss": 0.0389, "step": 1150 }, { "epoch": 0.34228385954558865, "grad_norm": 0.4565533995628357, "learning_rate": 9.450017401170689e-05, "loss": 0.0362, "step": 1160 }, { "epoch": 0.34523458247270583, "grad_norm": 0.42608538269996643, "learning_rate": 9.437384779514256e-05, "loss": 0.0397, "step": 1170 }, { "epoch": 0.34818530539982295, "grad_norm": 0.4438352882862091, "learning_rate": 9.424617362162271e-05, "loss": 0.0419, "step": 1180 }, { "epoch": 0.3511360283269401, "grad_norm": 0.2844030559062958, "learning_rate": 9.411715536954196e-05, "loss": 0.0387, "step": 1190 }, { "epoch": 0.35408675125405725, "grad_norm": 0.5422294735908508, "learning_rate": 9.39867969581243e-05, "loss": 0.0389, "step": 1200 }, { "epoch": 0.3570374741811744, "grad_norm": 0.35898223519325256, "learning_rate": 9.385510234730415e-05, "loss": 0.0421, "step": 1210 }, { "epoch": 0.35998819710829155, "grad_norm": 0.48864200711250305, "learning_rate": 9.372207553760603e-05, "loss": 0.035, "step": 1220 }, { "epoch": 0.3629389200354087, "grad_norm": 0.43427953124046326, "learning_rate": 9.358772057002312e-05, "loss": 0.0409, "step": 1230 }, { "epoch": 0.3658896429625258, "grad_norm": 0.3565417528152466, "learning_rate": 9.345204152589428e-05, "loss": 0.0375, "step": 1240 }, { "epoch": 0.368840365889643, "grad_norm": 0.21392573416233063, "learning_rate": 9.331504252678037e-05, "loss": 0.0378, "step": 1250 }, { "epoch": 0.3717910888167601, "grad_norm": 0.46816399693489075, "learning_rate": 9.317672773433876e-05, "loss": 0.0361, "step": 1260 }, { "epoch": 0.3747418117438773, "grad_norm": 0.5512992143630981, "learning_rate": 9.30371013501972e-05, "loss": 0.0416, "step": 1270 }, { "epoch": 0.3776925346709944, "grad_norm": 0.5283510684967041, "learning_rate": 9.289616761582587e-05, "loss": 0.0371, "step": 1280 }, { "epoch": 0.3806432575981115, "grad_norm": 0.5033398270606995, "learning_rate": 9.275393081240882e-05, "loss": 0.0442, "step": 1290 }, { "epoch": 0.3835939805252287, "grad_norm": 0.6394841074943542, "learning_rate": 9.261039526071374e-05, "loss": 0.0411, "step": 1300 }, { "epoch": 0.3865447034523458, "grad_norm": 0.4064463973045349, "learning_rate": 9.246556532096078e-05, "loss": 0.0427, "step": 1310 }, { "epoch": 0.389495426379463, "grad_norm": 0.46286794543266296, "learning_rate": 9.231944539269009e-05, "loss": 0.0427, "step": 1320 }, { "epoch": 0.3924461493065801, "grad_norm": 0.5626000165939331, "learning_rate": 9.217203991462815e-05, "loss": 0.0352, "step": 1330 }, { "epoch": 0.39539687223369724, "grad_norm": 0.4471147060394287, "learning_rate": 9.202335336455296e-05, "loss": 0.0392, "step": 1340 }, { "epoch": 0.3983475951608144, "grad_norm": 0.3415900766849518, "learning_rate": 9.187339025915802e-05, "loss": 0.0331, "step": 1350 }, { "epoch": 0.40129831808793154, "grad_norm": 0.39568573236465454, "learning_rate": 9.17221551539151e-05, "loss": 0.0338, "step": 1360 }, { "epoch": 0.40424904101504866, "grad_norm": 0.43257808685302734, "learning_rate": 9.156965264293586e-05, "loss": 0.0325, "step": 1370 }, { "epoch": 0.40719976394216584, "grad_norm": 0.5751752853393555, "learning_rate": 9.141588735883232e-05, "loss": 0.0356, "step": 1380 }, { "epoch": 0.41015048686928296, "grad_norm": 0.41519418358802795, "learning_rate": 9.126086397257612e-05, "loss": 0.0361, "step": 1390 }, { "epoch": 0.41310120979640014, "grad_norm": 0.5463760495185852, "learning_rate": 9.110458719335659e-05, "loss": 0.0414, "step": 1400 }, { "epoch": 0.41605193272351726, "grad_norm": 0.3883129358291626, "learning_rate": 9.094706176843777e-05, "loss": 0.031, "step": 1410 }, { "epoch": 0.4190026556506344, "grad_norm": 0.4870079457759857, "learning_rate": 9.078829248301417e-05, "loss": 0.0425, "step": 1420 }, { "epoch": 0.42195337857775156, "grad_norm": 0.4505230784416199, "learning_rate": 9.062828416006539e-05, "loss": 0.0395, "step": 1430 }, { "epoch": 0.4249041015048687, "grad_norm": 0.3301721215248108, "learning_rate": 9.046704166020961e-05, "loss": 0.0343, "step": 1440 }, { "epoch": 0.42785482443198586, "grad_norm": 0.2069552093744278, "learning_rate": 9.030456988155596e-05, "loss": 0.0375, "step": 1450 }, { "epoch": 0.430805547359103, "grad_norm": 0.28802281618118286, "learning_rate": 9.014087375955573e-05, "loss": 0.0339, "step": 1460 }, { "epoch": 0.4337562702862201, "grad_norm": 0.7400907874107361, "learning_rate": 8.997595826685243e-05, "loss": 0.0402, "step": 1470 }, { "epoch": 0.4367069932133373, "grad_norm": 0.577970027923584, "learning_rate": 8.980982841313074e-05, "loss": 0.0421, "step": 1480 }, { "epoch": 0.4396577161404544, "grad_norm": 0.3911955952644348, "learning_rate": 8.964248924496435e-05, "loss": 0.0366, "step": 1490 }, { "epoch": 0.4426084390675716, "grad_norm": 0.4710315763950348, "learning_rate": 8.947394584566258e-05, "loss": 0.0445, "step": 1500 }, { "epoch": 0.4455591619946887, "grad_norm": 0.29758352041244507, "learning_rate": 8.930420333511606e-05, "loss": 0.0408, "step": 1510 }, { "epoch": 0.4485098849218058, "grad_norm": 0.3915882706642151, "learning_rate": 8.913326686964117e-05, "loss": 0.0366, "step": 1520 }, { "epoch": 0.451460607848923, "grad_norm": 0.39474600553512573, "learning_rate": 8.89611416418234e-05, "loss": 0.0343, "step": 1530 }, { "epoch": 0.4544113307760401, "grad_norm": 0.3231770992279053, "learning_rate": 8.878783288035957e-05, "loss": 0.0403, "step": 1540 }, { "epoch": 0.4573620537031573, "grad_norm": 0.3711271286010742, "learning_rate": 8.86133458498991e-05, "loss": 0.0335, "step": 1550 }, { "epoch": 0.4603127766302744, "grad_norm": 0.39086106419563293, "learning_rate": 8.843768585088393e-05, "loss": 0.0351, "step": 1560 }, { "epoch": 0.46326349955739154, "grad_norm": 0.3677861988544464, "learning_rate": 8.82608582193877e-05, "loss": 0.035, "step": 1570 }, { "epoch": 0.4662142224845087, "grad_norm": 0.4083658754825592, "learning_rate": 8.80828683269535e-05, "loss": 0.0326, "step": 1580 }, { "epoch": 0.46916494541162584, "grad_norm": 0.5073351860046387, "learning_rate": 8.790372158043074e-05, "loss": 0.0375, "step": 1590 }, { "epoch": 0.47211566833874297, "grad_norm": 0.31964629888534546, "learning_rate": 8.772342342181095e-05, "loss": 0.0362, "step": 1600 }, { "epoch": 0.47506639126586014, "grad_norm": 0.39856019616127014, "learning_rate": 8.75419793280624e-05, "loss": 0.0318, "step": 1610 }, { "epoch": 0.47801711419297727, "grad_norm": 0.5297977328300476, "learning_rate": 8.735939481096378e-05, "loss": 0.037, "step": 1620 }, { "epoch": 0.48096783712009444, "grad_norm": 0.3231882154941559, "learning_rate": 8.717567541693673e-05, "loss": 0.0353, "step": 1630 }, { "epoch": 0.48391856004721157, "grad_norm": 0.430480420589447, "learning_rate": 8.699082672687734e-05, "loss": 0.0409, "step": 1640 }, { "epoch": 0.4868692829743287, "grad_norm": 0.27540120482444763, "learning_rate": 8.680485435598673e-05, "loss": 0.0364, "step": 1650 }, { "epoch": 0.48982000590144587, "grad_norm": 0.4706716239452362, "learning_rate": 8.661776395360029e-05, "loss": 0.0308, "step": 1660 }, { "epoch": 0.492770728828563, "grad_norm": 0.3352447748184204, "learning_rate": 8.642956120301626e-05, "loss": 0.0351, "step": 1670 }, { "epoch": 0.49572145175568016, "grad_norm": 0.23647312819957733, "learning_rate": 8.624025182132292e-05, "loss": 0.0371, "step": 1680 }, { "epoch": 0.4986721746827973, "grad_norm": 0.3626992404460907, "learning_rate": 8.604984155922506e-05, "loss": 0.0359, "step": 1690 }, { "epoch": 0.5016228976099144, "grad_norm": 0.39877787232398987, "learning_rate": 8.585833620086918e-05, "loss": 0.0309, "step": 1700 }, { "epoch": 0.5045736205370316, "grad_norm": 0.4257752001285553, "learning_rate": 8.566574156366784e-05, "loss": 0.0335, "step": 1710 }, { "epoch": 0.5075243434641488, "grad_norm": 0.6871393918991089, "learning_rate": 8.547206349812298e-05, "loss": 0.0381, "step": 1720 }, { "epoch": 0.5104750663912658, "grad_norm": 0.335425466299057, "learning_rate": 8.527730788764805e-05, "loss": 0.0367, "step": 1730 }, { "epoch": 0.513425789318383, "grad_norm": 0.34913158416748047, "learning_rate": 8.508148064838948e-05, "loss": 0.0358, "step": 1740 }, { "epoch": 0.5163765122455002, "grad_norm": 0.6184975504875183, "learning_rate": 8.488458772904684e-05, "loss": 0.04, "step": 1750 }, { "epoch": 0.5193272351726173, "grad_norm": 0.3886233866214752, "learning_rate": 8.468663511069217e-05, "loss": 0.0404, "step": 1760 }, { "epoch": 0.5222779580997344, "grad_norm": 0.6518878936767578, "learning_rate": 8.448762880658825e-05, "loss": 0.039, "step": 1770 }, { "epoch": 0.5252286810268516, "grad_norm": 0.28984302282333374, "learning_rate": 8.428757486200603e-05, "loss": 0.034, "step": 1780 }, { "epoch": 0.5281794039539687, "grad_norm": 0.36835214495658875, "learning_rate": 8.40864793540409e-05, "loss": 0.0384, "step": 1790 }, { "epoch": 0.5311301268810859, "grad_norm": 0.3769693970680237, "learning_rate": 8.388434839142813e-05, "loss": 0.0374, "step": 1800 }, { "epoch": 0.534080849808203, "grad_norm": 0.5667304992675781, "learning_rate": 8.368118811435726e-05, "loss": 0.038, "step": 1810 }, { "epoch": 0.5370315727353202, "grad_norm": 0.2845826745033264, "learning_rate": 8.347700469428564e-05, "loss": 0.0333, "step": 1820 }, { "epoch": 0.5399822956624373, "grad_norm": 0.5754022598266602, "learning_rate": 8.327180433375091e-05, "loss": 0.0348, "step": 1830 }, { "epoch": 0.5429330185895545, "grad_norm": 0.4201681911945343, "learning_rate": 8.306559326618259e-05, "loss": 0.0344, "step": 1840 }, { "epoch": 0.5458837415166716, "grad_norm": 0.43894684314727783, "learning_rate": 8.285837775571276e-05, "loss": 0.0335, "step": 1850 }, { "epoch": 0.5488344644437887, "grad_norm": 0.47914958000183105, "learning_rate": 8.265016409698573e-05, "loss": 0.0381, "step": 1860 }, { "epoch": 0.5517851873709059, "grad_norm": 0.4709184169769287, "learning_rate": 8.244095861496686e-05, "loss": 0.0393, "step": 1870 }, { "epoch": 0.554735910298023, "grad_norm": 0.4779587686061859, "learning_rate": 8.223076766475035e-05, "loss": 0.0326, "step": 1880 }, { "epoch": 0.5576866332251401, "grad_norm": 0.35479798913002014, "learning_rate": 8.201959763136633e-05, "loss": 0.0321, "step": 1890 }, { "epoch": 0.5606373561522573, "grad_norm": 0.5290229320526123, "learning_rate": 8.180745492958674e-05, "loss": 0.0408, "step": 1900 }, { "epoch": 0.5635880790793745, "grad_norm": 0.5257147550582886, "learning_rate": 8.159434600373061e-05, "loss": 0.0363, "step": 1910 }, { "epoch": 0.5665388020064915, "grad_norm": 0.44541099667549133, "learning_rate": 8.138027732746818e-05, "loss": 0.0376, "step": 1920 }, { "epoch": 0.5694895249336087, "grad_norm": 0.3628045618534088, "learning_rate": 8.116525540362434e-05, "loss": 0.033, "step": 1930 }, { "epoch": 0.5724402478607259, "grad_norm": 0.3127179443836212, "learning_rate": 8.094928676398101e-05, "loss": 0.0368, "step": 1940 }, { "epoch": 0.5753909707878431, "grad_norm": 0.29914721846580505, "learning_rate": 8.073237796907882e-05, "loss": 0.0367, "step": 1950 }, { "epoch": 0.5783416937149601, "grad_norm": 0.5147581100463867, "learning_rate": 8.051453560801772e-05, "loss": 0.0308, "step": 1960 }, { "epoch": 0.5812924166420773, "grad_norm": 0.4027189612388611, "learning_rate": 8.029576629825687e-05, "loss": 0.0381, "step": 1970 }, { "epoch": 0.5842431395691945, "grad_norm": 0.35352823138237, "learning_rate": 8.007607668541362e-05, "loss": 0.0405, "step": 1980 }, { "epoch": 0.5871938624963116, "grad_norm": 0.33770787715911865, "learning_rate": 7.985547344306161e-05, "loss": 0.0363, "step": 1990 }, { "epoch": 0.5901445854234287, "grad_norm": 0.44872477650642395, "learning_rate": 7.963396327252812e-05, "loss": 0.0299, "step": 2000 }, { "epoch": 0.5930953083505459, "grad_norm": 0.3611343502998352, "learning_rate": 7.941155290269038e-05, "loss": 0.0321, "step": 2010 }, { "epoch": 0.596046031277663, "grad_norm": 0.39241594076156616, "learning_rate": 7.918824908977123e-05, "loss": 0.0422, "step": 2020 }, { "epoch": 0.5989967542047802, "grad_norm": 0.29873213171958923, "learning_rate": 7.896405861713394e-05, "loss": 0.0369, "step": 2030 }, { "epoch": 0.6019474771318973, "grad_norm": 0.2608635127544403, "learning_rate": 7.873898829507606e-05, "loss": 0.0345, "step": 2040 }, { "epoch": 0.6048982000590145, "grad_norm": 0.37617939710617065, "learning_rate": 7.851304496062254e-05, "loss": 0.0308, "step": 2050 }, { "epoch": 0.6078489229861316, "grad_norm": 0.33235302567481995, "learning_rate": 7.828623547731818e-05, "loss": 0.0341, "step": 2060 }, { "epoch": 0.6107996459132488, "grad_norm": 0.5229690670967102, "learning_rate": 7.80585667350189e-05, "loss": 0.0331, "step": 2070 }, { "epoch": 0.6137503688403659, "grad_norm": 0.47838494181632996, "learning_rate": 7.783004564968263e-05, "loss": 0.0281, "step": 2080 }, { "epoch": 0.616701091767483, "grad_norm": 0.25658509135246277, "learning_rate": 7.760067916315921e-05, "loss": 0.0307, "step": 2090 }, { "epoch": 0.6196518146946002, "grad_norm": 0.3120635151863098, "learning_rate": 7.737047424297941e-05, "loss": 0.0312, "step": 2100 }, { "epoch": 0.6226025376217174, "grad_norm": 0.2760956585407257, "learning_rate": 7.713943788214337e-05, "loss": 0.0321, "step": 2110 }, { "epoch": 0.6255532605488344, "grad_norm": 0.4365791976451874, "learning_rate": 7.690757709890812e-05, "loss": 0.0313, "step": 2120 }, { "epoch": 0.6285039834759516, "grad_norm": 0.4906895160675049, "learning_rate": 7.66748989365744e-05, "loss": 0.03, "step": 2130 }, { "epoch": 0.6314547064030688, "grad_norm": 0.34254807233810425, "learning_rate": 7.644141046327271e-05, "loss": 0.0317, "step": 2140 }, { "epoch": 0.6344054293301858, "grad_norm": 0.23449936509132385, "learning_rate": 7.620711877174866e-05, "loss": 0.0351, "step": 2150 }, { "epoch": 0.637356152257303, "grad_norm": 0.3942568302154541, "learning_rate": 7.597203097914732e-05, "loss": 0.0317, "step": 2160 }, { "epoch": 0.6403068751844202, "grad_norm": 0.4756617546081543, "learning_rate": 7.573615422679726e-05, "loss": 0.0284, "step": 2170 }, { "epoch": 0.6432575981115374, "grad_norm": 0.3713917136192322, "learning_rate": 7.549949567999345e-05, "loss": 0.032, "step": 2180 }, { "epoch": 0.6462083210386544, "grad_norm": 0.2236933410167694, "learning_rate": 7.526206252777968e-05, "loss": 0.0286, "step": 2190 }, { "epoch": 0.6491590439657716, "grad_norm": 0.24449186027050018, "learning_rate": 7.50238619827301e-05, "loss": 0.0333, "step": 2200 }, { "epoch": 0.6521097668928888, "grad_norm": 0.26938706636428833, "learning_rate": 7.478490128073022e-05, "loss": 0.0277, "step": 2210 }, { "epoch": 0.6550604898200059, "grad_norm": 0.2308829426765442, "learning_rate": 7.454518768075704e-05, "loss": 0.0292, "step": 2220 }, { "epoch": 0.658011212747123, "grad_norm": 0.39452090859413147, "learning_rate": 7.430472846465856e-05, "loss": 0.0321, "step": 2230 }, { "epoch": 0.6609619356742402, "grad_norm": 0.3481849730014801, "learning_rate": 7.406353093693253e-05, "loss": 0.0319, "step": 2240 }, { "epoch": 0.6639126586013573, "grad_norm": 0.6345138549804688, "learning_rate": 7.382160242450469e-05, "loss": 0.0365, "step": 2250 }, { "epoch": 0.6668633815284745, "grad_norm": 0.3432180881500244, "learning_rate": 7.357895027650598e-05, "loss": 0.0333, "step": 2260 }, { "epoch": 0.6698141044555916, "grad_norm": 0.3641951084136963, "learning_rate": 7.333558186404958e-05, "loss": 0.0285, "step": 2270 }, { "epoch": 0.6727648273827088, "grad_norm": 0.4424731731414795, "learning_rate": 7.309150458000668e-05, "loss": 0.0313, "step": 2280 }, { "epoch": 0.6757155503098259, "grad_norm": 0.26001280546188354, "learning_rate": 7.284672583878219e-05, "loss": 0.0285, "step": 2290 }, { "epoch": 0.6786662732369431, "grad_norm": 0.3934877812862396, "learning_rate": 7.260125307608929e-05, "loss": 0.0277, "step": 2300 }, { "epoch": 0.6816169961640602, "grad_norm": 0.2620293200016022, "learning_rate": 7.235509374872373e-05, "loss": 0.034, "step": 2310 }, { "epoch": 0.6845677190911773, "grad_norm": 0.5262678861618042, "learning_rate": 7.210825533433719e-05, "loss": 0.0277, "step": 2320 }, { "epoch": 0.6875184420182945, "grad_norm": 0.2813281714916229, "learning_rate": 7.186074533121013e-05, "loss": 0.0331, "step": 2330 }, { "epoch": 0.6904691649454117, "grad_norm": 0.3299836814403534, "learning_rate": 7.161257125802413e-05, "loss": 0.0298, "step": 2340 }, { "epoch": 0.6934198878725287, "grad_norm": 0.2901538014411926, "learning_rate": 7.136374065363334e-05, "loss": 0.0348, "step": 2350 }, { "epoch": 0.6963706107996459, "grad_norm": 0.33680474758148193, "learning_rate": 7.11142610768356e-05, "loss": 0.0312, "step": 2360 }, { "epoch": 0.6993213337267631, "grad_norm": 0.2652488946914673, "learning_rate": 7.086414010614276e-05, "loss": 0.0303, "step": 2370 }, { "epoch": 0.7022720566538802, "grad_norm": 0.37170788645744324, "learning_rate": 7.061338533955043e-05, "loss": 0.0268, "step": 2380 }, { "epoch": 0.7052227795809973, "grad_norm": 0.5011565089225769, "learning_rate": 7.036200439430725e-05, "loss": 0.0324, "step": 2390 }, { "epoch": 0.7081735025081145, "grad_norm": 0.2641090452671051, "learning_rate": 7.01100049066835e-05, "loss": 0.0332, "step": 2400 }, { "epoch": 0.7111242254352317, "grad_norm": 0.3112793564796448, "learning_rate": 6.985739453173903e-05, "loss": 0.0251, "step": 2410 }, { "epoch": 0.7140749483623487, "grad_norm": 0.2889801561832428, "learning_rate": 6.960418094309085e-05, "loss": 0.0271, "step": 2420 }, { "epoch": 0.7170256712894659, "grad_norm": 0.41144901514053345, "learning_rate": 6.93503718326799e-05, "loss": 0.0298, "step": 2430 }, { "epoch": 0.7199763942165831, "grad_norm": 0.31838467717170715, "learning_rate": 6.909597491053751e-05, "loss": 0.0296, "step": 2440 }, { "epoch": 0.7229271171437002, "grad_norm": 0.44234567880630493, "learning_rate": 6.884099790455113e-05, "loss": 0.0292, "step": 2450 }, { "epoch": 0.7258778400708173, "grad_norm": 0.41713687777519226, "learning_rate": 6.858544856022952e-05, "loss": 0.0339, "step": 2460 }, { "epoch": 0.7288285629979345, "grad_norm": 0.2954857349395752, "learning_rate": 6.83293346404676e-05, "loss": 0.0305, "step": 2470 }, { "epoch": 0.7317792859250516, "grad_norm": 0.39784836769104004, "learning_rate": 6.80726639253105e-05, "loss": 0.0288, "step": 2480 }, { "epoch": 0.7347300088521688, "grad_norm": 0.3012574017047882, "learning_rate": 6.781544421171732e-05, "loss": 0.0276, "step": 2490 }, { "epoch": 0.737680731779286, "grad_norm": 0.3239777088165283, "learning_rate": 6.755768331332424e-05, "loss": 0.034, "step": 2500 }, { "epoch": 0.740631454706403, "grad_norm": 0.3193805515766144, "learning_rate": 6.729938906020713e-05, "loss": 0.0317, "step": 2510 }, { "epoch": 0.7435821776335202, "grad_norm": 0.43207693099975586, "learning_rate": 6.704056929864376e-05, "loss": 0.0254, "step": 2520 }, { "epoch": 0.7465329005606374, "grad_norm": 0.5076596736907959, "learning_rate": 6.67812318908754e-05, "loss": 0.028, "step": 2530 }, { "epoch": 0.7494836234877545, "grad_norm": 0.48384392261505127, "learning_rate": 6.6521384714868e-05, "loss": 0.0333, "step": 2540 }, { "epoch": 0.7524343464148716, "grad_norm": 0.3058372735977173, "learning_rate": 6.626103566407295e-05, "loss": 0.0349, "step": 2550 }, { "epoch": 0.7553850693419888, "grad_norm": 0.253572016954422, "learning_rate": 6.600019264718713e-05, "loss": 0.028, "step": 2560 }, { "epoch": 0.758335792269106, "grad_norm": 0.23040670156478882, "learning_rate": 6.573886358791285e-05, "loss": 0.0299, "step": 2570 }, { "epoch": 0.761286515196223, "grad_norm": 0.2834416925907135, "learning_rate": 6.547705642471703e-05, "loss": 0.0309, "step": 2580 }, { "epoch": 0.7642372381233402, "grad_norm": 0.2733057737350464, "learning_rate": 6.521477911059008e-05, "loss": 0.028, "step": 2590 }, { "epoch": 0.7671879610504574, "grad_norm": 0.32252931594848633, "learning_rate": 6.495203961280434e-05, "loss": 0.0277, "step": 2600 }, { "epoch": 0.7701386839775745, "grad_norm": 0.2437136024236679, "learning_rate": 6.468884591267204e-05, "loss": 0.0274, "step": 2610 }, { "epoch": 0.7730894069046916, "grad_norm": 0.32221049070358276, "learning_rate": 6.44252060053028e-05, "loss": 0.0271, "step": 2620 }, { "epoch": 0.7760401298318088, "grad_norm": 0.3222113847732544, "learning_rate": 6.416112789936086e-05, "loss": 0.0252, "step": 2630 }, { "epoch": 0.778990852758926, "grad_norm": 0.33199211955070496, "learning_rate": 6.389661961682173e-05, "loss": 0.0242, "step": 2640 }, { "epoch": 0.7819415756860431, "grad_norm": 0.29851701855659485, "learning_rate": 6.363168919272846e-05, "loss": 0.0282, "step": 2650 }, { "epoch": 0.7848922986131602, "grad_norm": 0.3832946717739105, "learning_rate": 6.336634467494768e-05, "loss": 0.0363, "step": 2660 }, { "epoch": 0.7878430215402774, "grad_norm": 0.2601946294307709, "learning_rate": 6.310059412392505e-05, "loss": 0.0314, "step": 2670 }, { "epoch": 0.7907937444673945, "grad_norm": 0.22252275049686432, "learning_rate": 6.283444561244042e-05, "loss": 0.0305, "step": 2680 }, { "epoch": 0.7937444673945117, "grad_norm": 0.44154414534568787, "learning_rate": 6.256790722536251e-05, "loss": 0.0253, "step": 2690 }, { "epoch": 0.7966951903216288, "grad_norm": 0.40197300910949707, "learning_rate": 6.230098705940354e-05, "loss": 0.0304, "step": 2700 }, { "epoch": 0.7996459132487459, "grad_norm": 0.4446711838245392, "learning_rate": 6.203369322287306e-05, "loss": 0.0294, "step": 2710 }, { "epoch": 0.8025966361758631, "grad_norm": 0.37209516763687134, "learning_rate": 6.17660338354317e-05, "loss": 0.0269, "step": 2720 }, { "epoch": 0.8055473591029803, "grad_norm": 0.6373627781867981, "learning_rate": 6.149801702784456e-05, "loss": 0.0295, "step": 2730 }, { "epoch": 0.8084980820300973, "grad_norm": 0.2860022783279419, "learning_rate": 6.122965094173424e-05, "loss": 0.0297, "step": 2740 }, { "epoch": 0.8114488049572145, "grad_norm": 0.3376583158969879, "learning_rate": 6.0960943729333374e-05, "loss": 0.0268, "step": 2750 }, { "epoch": 0.8143995278843317, "grad_norm": 0.2861047089099884, "learning_rate": 6.069190355323717e-05, "loss": 0.0262, "step": 2760 }, { "epoch": 0.8173502508114489, "grad_norm": 0.34502729773521423, "learning_rate": 6.042253858615532e-05, "loss": 0.026, "step": 2770 }, { "epoch": 0.8203009737385659, "grad_norm": 0.3573007583618164, "learning_rate": 6.015285701066382e-05, "loss": 0.0285, "step": 2780 }, { "epoch": 0.8232516966656831, "grad_norm": 0.32636600732803345, "learning_rate": 5.988286701895631e-05, "loss": 0.0248, "step": 2790 }, { "epoch": 0.8262024195928003, "grad_norm": 0.3002491593360901, "learning_rate": 5.961257681259535e-05, "loss": 0.0277, "step": 2800 }, { "epoch": 0.8291531425199173, "grad_norm": 0.20530137419700623, "learning_rate": 5.934199460226317e-05, "loss": 0.027, "step": 2810 }, { "epoch": 0.8321038654470345, "grad_norm": 0.40788331627845764, "learning_rate": 5.9071128607512285e-05, "loss": 0.0242, "step": 2820 }, { "epoch": 0.8350545883741517, "grad_norm": 0.2836095690727234, "learning_rate": 5.8799987056515804e-05, "loss": 0.0271, "step": 2830 }, { "epoch": 0.8380053113012688, "grad_norm": 0.2927771806716919, "learning_rate": 5.8528578185817514e-05, "loss": 0.0291, "step": 2840 }, { "epoch": 0.8409560342283859, "grad_norm": 0.43577444553375244, "learning_rate": 5.825691024008162e-05, "loss": 0.0273, "step": 2850 }, { "epoch": 0.8439067571555031, "grad_norm": 0.24480371177196503, "learning_rate": 5.798499147184233e-05, "loss": 0.0277, "step": 2860 }, { "epoch": 0.8468574800826203, "grad_norm": 0.32603973150253296, "learning_rate": 5.771283014125317e-05, "loss": 0.0245, "step": 2870 }, { "epoch": 0.8498082030097374, "grad_norm": 0.4056181311607361, "learning_rate": 5.7440434515836064e-05, "loss": 0.0255, "step": 2880 }, { "epoch": 0.8527589259368545, "grad_norm": 0.4071770906448364, "learning_rate": 5.7167812870230094e-05, "loss": 0.0253, "step": 2890 }, { "epoch": 0.8557096488639717, "grad_norm": 0.22184453904628754, "learning_rate": 5.689497348594035e-05, "loss": 0.0277, "step": 2900 }, { "epoch": 0.8586603717910888, "grad_norm": 0.4091993272304535, "learning_rate": 5.662192465108613e-05, "loss": 0.0321, "step": 2910 }, { "epoch": 0.861611094718206, "grad_norm": 0.42448750138282776, "learning_rate": 5.634867466014932e-05, "loss": 0.0266, "step": 2920 }, { "epoch": 0.8645618176453231, "grad_norm": 0.20825569331645966, "learning_rate": 5.607523181372234e-05, "loss": 0.0267, "step": 2930 }, { "epoch": 0.8675125405724402, "grad_norm": 0.2742401957511902, "learning_rate": 5.5801604418256117e-05, "loss": 0.0258, "step": 2940 }, { "epoch": 0.8704632634995574, "grad_norm": 0.3009909987449646, "learning_rate": 5.552780078580756e-05, "loss": 0.0244, "step": 2950 }, { "epoch": 0.8734139864266746, "grad_norm": 0.2381351888179779, "learning_rate": 5.525382923378728e-05, "loss": 0.0298, "step": 2960 }, { "epoch": 0.8763647093537916, "grad_norm": 0.4504906237125397, "learning_rate": 5.49796980847068e-05, "loss": 0.0306, "step": 2970 }, { "epoch": 0.8793154322809088, "grad_norm": 0.349475622177124, "learning_rate": 5.470541566592573e-05, "loss": 0.0236, "step": 2980 }, { "epoch": 0.882266155208026, "grad_norm": 0.261816143989563, "learning_rate": 5.443099030939887e-05, "loss": 0.022, "step": 2990 }, { "epoch": 0.8852168781351432, "grad_norm": 0.27858564257621765, "learning_rate": 5.415643035142309e-05, "loss": 0.0297, "step": 3000 }, { "epoch": 0.8881676010622602, "grad_norm": 0.3191457986831665, "learning_rate": 5.3881744132384104e-05, "loss": 0.0221, "step": 3010 }, { "epoch": 0.8911183239893774, "grad_norm": 0.29666879773139954, "learning_rate": 5.360693999650303e-05, "loss": 0.0304, "step": 3020 }, { "epoch": 0.8940690469164946, "grad_norm": 0.2829201817512512, "learning_rate": 5.3332026291583016e-05, "loss": 0.0226, "step": 3030 }, { "epoch": 0.8970197698436116, "grad_norm": 0.2625159025192261, "learning_rate": 5.305701136875566e-05, "loss": 0.0219, "step": 3040 }, { "epoch": 0.8999704927707288, "grad_norm": 0.19914016127586365, "learning_rate": 5.278190358222721e-05, "loss": 0.0286, "step": 3050 }, { "epoch": 0.902921215697846, "grad_norm": 0.3129836618900299, "learning_rate": 5.25067112890249e-05, "loss": 0.0276, "step": 3060 }, { "epoch": 0.9058719386249631, "grad_norm": 0.33484506607055664, "learning_rate": 5.2231442848743064e-05, "loss": 0.0307, "step": 3070 }, { "epoch": 0.9088226615520802, "grad_norm": 0.3785524070262909, "learning_rate": 5.1956106623289145e-05, "loss": 0.0251, "step": 3080 }, { "epoch": 0.9117733844791974, "grad_norm": 0.31567320227622986, "learning_rate": 5.168071097662972e-05, "loss": 0.029, "step": 3090 }, { "epoch": 0.9147241074063146, "grad_norm": 0.2716624438762665, "learning_rate": 5.1405264274536445e-05, "loss": 0.0291, "step": 3100 }, { "epoch": 0.9176748303334317, "grad_norm": 0.3030311167240143, "learning_rate": 5.112977488433188e-05, "loss": 0.031, "step": 3110 }, { "epoch": 0.9206255532605488, "grad_norm": 0.4680795669555664, "learning_rate": 5.085425117463533e-05, "loss": 0.0258, "step": 3120 }, { "epoch": 0.923576276187666, "grad_norm": 0.2229551374912262, "learning_rate": 5.057870151510864e-05, "loss": 0.0289, "step": 3130 }, { "epoch": 0.9265269991147831, "grad_norm": 0.5324947834014893, "learning_rate": 5.030313427620197e-05, "loss": 0.0256, "step": 3140 }, { "epoch": 0.9294777220419003, "grad_norm": 0.4284992516040802, "learning_rate": 5.0027557828899426e-05, "loss": 0.0239, "step": 3150 }, { "epoch": 0.9324284449690174, "grad_norm": 0.38140439987182617, "learning_rate": 4.975198054446492e-05, "loss": 0.0245, "step": 3160 }, { "epoch": 0.9353791678961345, "grad_norm": 0.2657966613769531, "learning_rate": 4.947641079418773e-05, "loss": 0.029, "step": 3170 }, { "epoch": 0.9383298908232517, "grad_norm": 0.23872986435890198, "learning_rate": 4.920085694912828e-05, "loss": 0.0277, "step": 3180 }, { "epoch": 0.9412806137503689, "grad_norm": 0.42396003007888794, "learning_rate": 4.892532737986387e-05, "loss": 0.0269, "step": 3190 }, { "epoch": 0.9442313366774859, "grad_norm": 0.4265054166316986, "learning_rate": 4.864983045623434e-05, "loss": 0.0293, "step": 3200 }, { "epoch": 0.9471820596046031, "grad_norm": 0.33946704864501953, "learning_rate": 4.837437454708784e-05, "loss": 0.0283, "step": 3210 }, { "epoch": 0.9501327825317203, "grad_norm": 0.2541551887989044, "learning_rate": 4.809896802002662e-05, "loss": 0.0232, "step": 3220 }, { "epoch": 0.9530835054588375, "grad_norm": 0.22080323100090027, "learning_rate": 4.7823619241152854e-05, "loss": 0.0261, "step": 3230 }, { "epoch": 0.9560342283859545, "grad_norm": 0.49365678429603577, "learning_rate": 4.754833657481445e-05, "loss": 0.0251, "step": 3240 }, { "epoch": 0.9589849513130717, "grad_norm": 0.43560582399368286, "learning_rate": 4.7273128383351015e-05, "loss": 0.0254, "step": 3250 }, { "epoch": 0.9619356742401889, "grad_norm": 0.20434625446796417, "learning_rate": 4.699800302683981e-05, "loss": 0.0289, "step": 3260 }, { "epoch": 0.964886397167306, "grad_norm": 0.37063729763031006, "learning_rate": 4.6722968862841806e-05, "loss": 0.0245, "step": 3270 }, { "epoch": 0.9678371200944231, "grad_norm": 0.37795308232307434, "learning_rate": 4.6448034246147754e-05, "loss": 0.0225, "step": 3280 }, { "epoch": 0.9707878430215403, "grad_norm": 0.3016812205314636, "learning_rate": 4.6173207528524476e-05, "loss": 0.0289, "step": 3290 }, { "epoch": 0.9737385659486574, "grad_norm": 0.4373134970664978, "learning_rate": 4.58984970584611e-05, "loss": 0.0224, "step": 3300 }, { "epoch": 0.9766892888757746, "grad_norm": 0.2669011354446411, "learning_rate": 4.562391118091544e-05, "loss": 0.0253, "step": 3310 }, { "epoch": 0.9796400118028917, "grad_norm": 0.4083000421524048, "learning_rate": 4.534945823706056e-05, "loss": 0.0234, "step": 3320 }, { "epoch": 0.9825907347300088, "grad_norm": 0.3307209312915802, "learning_rate": 4.507514656403137e-05, "loss": 0.0233, "step": 3330 }, { "epoch": 0.985541457657126, "grad_norm": 0.3889909088611603, "learning_rate": 4.480098449467132e-05, "loss": 0.0235, "step": 3340 }, { "epoch": 0.9884921805842432, "grad_norm": 0.36005377769470215, "learning_rate": 4.452698035727929e-05, "loss": 0.0259, "step": 3350 }, { "epoch": 0.9914429035113603, "grad_norm": 0.3942202627658844, "learning_rate": 4.425314247535668e-05, "loss": 0.0246, "step": 3360 }, { "epoch": 0.9943936264384774, "grad_norm": 0.314887672662735, "learning_rate": 4.3979479167354477e-05, "loss": 0.0299, "step": 3370 }, { "epoch": 0.9973443493655946, "grad_norm": 0.2141730636358261, "learning_rate": 4.370599874642055e-05, "loss": 0.0249, "step": 3380 }, { "epoch": 1.0002950722927118, "grad_norm": 0.29817038774490356, "learning_rate": 4.3432709520147205e-05, "loss": 0.027, "step": 3390 }, { "epoch": 1.0032457952198288, "grad_norm": 0.3092167377471924, "learning_rate": 4.315961979031875e-05, "loss": 0.0286, "step": 3400 }, { "epoch": 1.006196518146946, "grad_norm": 0.25228509306907654, "learning_rate": 4.2886737852659325e-05, "loss": 0.0267, "step": 3410 }, { "epoch": 1.0091472410740632, "grad_norm": 0.22625741362571716, "learning_rate": 4.261407199658093e-05, "loss": 0.0249, "step": 3420 }, { "epoch": 1.0120979640011802, "grad_norm": 0.3815089166164398, "learning_rate": 4.234163050493158e-05, "loss": 0.0235, "step": 3430 }, { "epoch": 1.0150486869282975, "grad_norm": 0.3839840292930603, "learning_rate": 4.2069421653743706e-05, "loss": 0.0271, "step": 3440 }, { "epoch": 1.0179994098554146, "grad_norm": 0.2331916093826294, "learning_rate": 4.179745371198276e-05, "loss": 0.0214, "step": 3450 }, { "epoch": 1.0209501327825317, "grad_norm": 0.3271558880805969, "learning_rate": 4.1525734941296026e-05, "loss": 0.0284, "step": 3460 }, { "epoch": 1.023900855709649, "grad_norm": 0.2440626174211502, "learning_rate": 4.125427359576162e-05, "loss": 0.0189, "step": 3470 }, { "epoch": 1.026851578636766, "grad_norm": 0.3395729064941406, "learning_rate": 4.0983077921637815e-05, "loss": 0.0243, "step": 3480 }, { "epoch": 1.029802301563883, "grad_norm": 0.22356589138507843, "learning_rate": 4.07121561571125e-05, "loss": 0.0231, "step": 3490 }, { "epoch": 1.0327530244910004, "grad_norm": 0.24800588190555573, "learning_rate": 4.044151653205292e-05, "loss": 0.0278, "step": 3500 }, { "epoch": 1.0357037474181174, "grad_norm": 0.3692575693130493, "learning_rate": 4.0171167267755696e-05, "loss": 0.0269, "step": 3510 }, { "epoch": 1.0386544703452345, "grad_norm": 0.2397177666425705, "learning_rate": 3.9901116576697083e-05, "loss": 0.0204, "step": 3520 }, { "epoch": 1.0416051932723518, "grad_norm": 0.318625271320343, "learning_rate": 3.963137266228349e-05, "loss": 0.0228, "step": 3530 }, { "epoch": 1.0445559161994689, "grad_norm": 0.29952913522720337, "learning_rate": 3.93619437186023e-05, "loss": 0.0269, "step": 3540 }, { "epoch": 1.047506639126586, "grad_norm": 0.2282438427209854, "learning_rate": 3.9092837930172884e-05, "loss": 0.0267, "step": 3550 }, { "epoch": 1.0504573620537032, "grad_norm": 0.3359334170818329, "learning_rate": 3.8824063471698105e-05, "loss": 0.025, "step": 3560 }, { "epoch": 1.0534080849808203, "grad_norm": 0.3965560495853424, "learning_rate": 3.855562850781589e-05, "loss": 0.0269, "step": 3570 }, { "epoch": 1.0563588079079373, "grad_norm": 0.24504753947257996, "learning_rate": 3.828754119285123e-05, "loss": 0.024, "step": 3580 }, { "epoch": 1.0593095308350546, "grad_norm": 0.19007989764213562, "learning_rate": 3.801980967056851e-05, "loss": 0.0241, "step": 3590 }, { "epoch": 1.0622602537621717, "grad_norm": 0.22689615190029144, "learning_rate": 3.77524420739241e-05, "loss": 0.0244, "step": 3600 }, { "epoch": 1.065210976689289, "grad_norm": 0.3528243899345398, "learning_rate": 3.748544652481927e-05, "loss": 0.0209, "step": 3610 }, { "epoch": 1.068161699616406, "grad_norm": 0.3866553008556366, "learning_rate": 3.721883113385353e-05, "loss": 0.0225, "step": 3620 }, { "epoch": 1.0711124225435231, "grad_norm": 0.24558000266551971, "learning_rate": 3.695260400007819e-05, "loss": 0.0223, "step": 3630 }, { "epoch": 1.0740631454706402, "grad_norm": 0.24806998670101166, "learning_rate": 3.6686773210750385e-05, "loss": 0.0247, "step": 3640 }, { "epoch": 1.0770138683977575, "grad_norm": 0.3183746039867401, "learning_rate": 3.642134684108737e-05, "loss": 0.0244, "step": 3650 }, { "epoch": 1.0799645913248745, "grad_norm": 0.2888622283935547, "learning_rate": 3.615633295402123e-05, "loss": 0.0194, "step": 3660 }, { "epoch": 1.0829153142519918, "grad_norm": 0.2269098162651062, "learning_rate": 3.5891739599953945e-05, "loss": 0.0206, "step": 3670 }, { "epoch": 1.085866037179109, "grad_norm": 0.29047688841819763, "learning_rate": 3.5627574816512846e-05, "loss": 0.0275, "step": 3680 }, { "epoch": 1.088816760106226, "grad_norm": 0.24343855679035187, "learning_rate": 3.536384662830648e-05, "loss": 0.0248, "step": 3690 }, { "epoch": 1.0917674830333433, "grad_norm": 0.21406111121177673, "learning_rate": 3.5100563046680764e-05, "loss": 0.0207, "step": 3700 }, { "epoch": 1.0947182059604603, "grad_norm": 0.22464735805988312, "learning_rate": 3.483773206947572e-05, "loss": 0.0271, "step": 3710 }, { "epoch": 1.0976689288875774, "grad_norm": 0.1557023823261261, "learning_rate": 3.457536168078247e-05, "loss": 0.0246, "step": 3720 }, { "epoch": 1.1006196518146947, "grad_norm": 0.2940058708190918, "learning_rate": 3.431345985070067e-05, "loss": 0.0183, "step": 3730 }, { "epoch": 1.1035703747418117, "grad_norm": 0.2519182860851288, "learning_rate": 3.40520345350965e-05, "loss": 0.0271, "step": 3740 }, { "epoch": 1.1065210976689288, "grad_norm": 0.2663120925426483, "learning_rate": 3.379109367536089e-05, "loss": 0.0208, "step": 3750 }, { "epoch": 1.109471820596046, "grad_norm": 0.35298481583595276, "learning_rate": 3.3530645198168295e-05, "loss": 0.0224, "step": 3760 }, { "epoch": 1.1124225435231632, "grad_norm": 0.37713515758514404, "learning_rate": 3.327069701523595e-05, "loss": 0.0281, "step": 3770 }, { "epoch": 1.1153732664502802, "grad_norm": 0.17818234860897064, "learning_rate": 3.301125702308353e-05, "loss": 0.0209, "step": 3780 }, { "epoch": 1.1183239893773975, "grad_norm": 0.2551743686199188, "learning_rate": 3.275233310279321e-05, "loss": 0.0223, "step": 3790 }, { "epoch": 1.1212747123045146, "grad_norm": 0.37746578454971313, "learning_rate": 3.249393311977037e-05, "loss": 0.0234, "step": 3800 }, { "epoch": 1.1242254352316317, "grad_norm": 0.25601494312286377, "learning_rate": 3.223606492350451e-05, "loss": 0.0185, "step": 3810 }, { "epoch": 1.127176158158749, "grad_norm": 0.24753884971141815, "learning_rate": 3.197873634733096e-05, "loss": 0.0233, "step": 3820 }, { "epoch": 1.130126881085866, "grad_norm": 0.19221985340118408, "learning_rate": 3.172195520819285e-05, "loss": 0.0239, "step": 3830 }, { "epoch": 1.1330776040129833, "grad_norm": 0.3169882297515869, "learning_rate": 3.146572930640362e-05, "loss": 0.024, "step": 3840 }, { "epoch": 1.1360283269401004, "grad_norm": 0.4285431206226349, "learning_rate": 3.121006642541014e-05, "loss": 0.0208, "step": 3850 }, { "epoch": 1.1389790498672174, "grad_norm": 0.2852340340614319, "learning_rate": 3.095497433155626e-05, "loss": 0.0229, "step": 3860 }, { "epoch": 1.1419297727943345, "grad_norm": 0.43975260853767395, "learning_rate": 3.070046077384682e-05, "loss": 0.0245, "step": 3870 }, { "epoch": 1.1448804957214518, "grad_norm": 0.3585070073604584, "learning_rate": 3.0446533483712304e-05, "loss": 0.0242, "step": 3880 }, { "epoch": 1.1478312186485689, "grad_norm": 0.31226640939712524, "learning_rate": 3.0193200174774038e-05, "loss": 0.0203, "step": 3890 }, { "epoch": 1.1507819415756861, "grad_norm": 0.3078920841217041, "learning_rate": 2.994046854260974e-05, "loss": 0.0235, "step": 3900 }, { "epoch": 1.1537326645028032, "grad_norm": 0.29201143980026245, "learning_rate": 2.9688346264519866e-05, "loss": 0.0212, "step": 3910 }, { "epoch": 1.1566833874299203, "grad_norm": 0.22376073896884918, "learning_rate": 2.943684099929436e-05, "loss": 0.0204, "step": 3920 }, { "epoch": 1.1596341103570376, "grad_norm": 0.21112863719463348, "learning_rate": 2.918596038697995e-05, "loss": 0.0223, "step": 3930 }, { "epoch": 1.1625848332841546, "grad_norm": 0.3708503246307373, "learning_rate": 2.8935712048648112e-05, "loss": 0.0259, "step": 3940 }, { "epoch": 1.1655355562112717, "grad_norm": 0.3241925537586212, "learning_rate": 2.8686103586163626e-05, "loss": 0.0238, "step": 3950 }, { "epoch": 1.168486279138389, "grad_norm": 0.24407649040222168, "learning_rate": 2.843714258195346e-05, "loss": 0.0218, "step": 3960 }, { "epoch": 1.171437002065506, "grad_norm": 0.24018298089504242, "learning_rate": 2.8188836598776662e-05, "loss": 0.0254, "step": 3970 }, { "epoch": 1.1743877249926231, "grad_norm": 0.2403210997581482, "learning_rate": 2.7941193179494484e-05, "loss": 0.0194, "step": 3980 }, { "epoch": 1.1773384479197404, "grad_norm": 0.18335258960723877, "learning_rate": 2.7694219846841262e-05, "loss": 0.0185, "step": 3990 }, { "epoch": 1.1802891708468575, "grad_norm": 0.23028570413589478, "learning_rate": 2.7447924103195976e-05, "loss": 0.0208, "step": 4000 }, { "epoch": 1.1832398937739745, "grad_norm": 0.23867817223072052, "learning_rate": 2.7202313430354253e-05, "loss": 0.0223, "step": 4010 }, { "epoch": 1.1861906167010918, "grad_norm": 0.2862341105937958, "learning_rate": 2.695739528930111e-05, "loss": 0.021, "step": 4020 }, { "epoch": 1.189141339628209, "grad_norm": 0.2866331934928894, "learning_rate": 2.67131771199844e-05, "loss": 0.0215, "step": 4030 }, { "epoch": 1.192092062555326, "grad_norm": 0.18745006620883942, "learning_rate": 2.6469666341088677e-05, "loss": 0.0251, "step": 4040 }, { "epoch": 1.1950427854824432, "grad_norm": 0.19981816411018372, "learning_rate": 2.6226870349809885e-05, "loss": 0.0196, "step": 4050 }, { "epoch": 1.1979935084095603, "grad_norm": 0.2717253267765045, "learning_rate": 2.5984796521630737e-05, "loss": 0.0239, "step": 4060 }, { "epoch": 1.2009442313366776, "grad_norm": 0.27966541051864624, "learning_rate": 2.574345221009653e-05, "loss": 0.0207, "step": 4070 }, { "epoch": 1.2038949542637947, "grad_norm": 0.24904650449752808, "learning_rate": 2.5502844746591804e-05, "loss": 0.019, "step": 4080 }, { "epoch": 1.2068456771909117, "grad_norm": 0.18721486628055573, "learning_rate": 2.526298144011775e-05, "loss": 0.0225, "step": 4090 }, { "epoch": 1.2097964001180288, "grad_norm": 0.2366674393415451, "learning_rate": 2.5023869577070013e-05, "loss": 0.0232, "step": 4100 }, { "epoch": 1.212747123045146, "grad_norm": 0.18318581581115723, "learning_rate": 2.478551642101743e-05, "loss": 0.022, "step": 4110 }, { "epoch": 1.2156978459722632, "grad_norm": 0.1958722472190857, "learning_rate": 2.4547929212481435e-05, "loss": 0.0212, "step": 4120 }, { "epoch": 1.2186485688993804, "grad_norm": 0.28014546632766724, "learning_rate": 2.4311115168716013e-05, "loss": 0.0182, "step": 4130 }, { "epoch": 1.2215992918264975, "grad_norm": 0.22706721723079681, "learning_rate": 2.4075081483488494e-05, "loss": 0.0223, "step": 4140 }, { "epoch": 1.2245500147536146, "grad_norm": 0.18561671674251556, "learning_rate": 2.3839835326861104e-05, "loss": 0.0193, "step": 4150 }, { "epoch": 1.2275007376807319, "grad_norm": 0.15958668291568756, "learning_rate": 2.3605383844972966e-05, "loss": 0.0171, "step": 4160 }, { "epoch": 1.230451460607849, "grad_norm": 0.25699806213378906, "learning_rate": 2.3371734159823284e-05, "loss": 0.0227, "step": 4170 }, { "epoch": 1.233402183534966, "grad_norm": 0.23318181931972504, "learning_rate": 2.3138893369054766e-05, "loss": 0.0226, "step": 4180 }, { "epoch": 1.2363529064620833, "grad_norm": 0.36779430508613586, "learning_rate": 2.2906868545738102e-05, "loss": 0.0217, "step": 4190 }, { "epoch": 1.2393036293892004, "grad_norm": 0.1963030993938446, "learning_rate": 2.2675666738157186e-05, "loss": 0.0182, "step": 4200 }, { "epoch": 1.2422543523163174, "grad_norm": 0.2196776121854782, "learning_rate": 2.2445294969594844e-05, "loss": 0.0171, "step": 4210 }, { "epoch": 1.2452050752434347, "grad_norm": 0.22019082307815552, "learning_rate": 2.22157602381196e-05, "loss": 0.0179, "step": 4220 }, { "epoch": 1.2481557981705518, "grad_norm": 0.14729483425617218, "learning_rate": 2.1987069516373098e-05, "loss": 0.0176, "step": 4230 }, { "epoch": 1.251106521097669, "grad_norm": 0.20216135680675507, "learning_rate": 2.1759229751358217e-05, "loss": 0.0224, "step": 4240 }, { "epoch": 1.2540572440247861, "grad_norm": 0.22608602046966553, "learning_rate": 2.1532247864228084e-05, "loss": 0.0202, "step": 4250 }, { "epoch": 1.2570079669519032, "grad_norm": 0.14543867111206055, "learning_rate": 2.1306130750075865e-05, "loss": 0.023, "step": 4260 }, { "epoch": 1.2599586898790203, "grad_norm": 0.2126445770263672, "learning_rate": 2.1080885277725236e-05, "loss": 0.0199, "step": 4270 }, { "epoch": 1.2629094128061376, "grad_norm": 0.2856239974498749, "learning_rate": 2.085651828952175e-05, "loss": 0.0207, "step": 4280 }, { "epoch": 1.2658601357332546, "grad_norm": 0.3021923899650574, "learning_rate": 2.063303660112506e-05, "loss": 0.0185, "step": 4290 }, { "epoch": 1.268810858660372, "grad_norm": 0.28485947847366333, "learning_rate": 2.0410447001301753e-05, "loss": 0.0207, "step": 4300 }, { "epoch": 1.271761581587489, "grad_norm": 0.3055853843688965, "learning_rate": 2.0188756251719203e-05, "loss": 0.0181, "step": 4310 }, { "epoch": 1.274712304514606, "grad_norm": 0.26596036553382874, "learning_rate": 1.9967971086740195e-05, "loss": 0.0207, "step": 4320 }, { "epoch": 1.277663027441723, "grad_norm": 0.20866192877292633, "learning_rate": 1.974809821321827e-05, "loss": 0.0176, "step": 4330 }, { "epoch": 1.2806137503688404, "grad_norm": 0.19133244454860687, "learning_rate": 1.9529144310294023e-05, "loss": 0.0186, "step": 4340 }, { "epoch": 1.2835644732959575, "grad_norm": 0.2650998830795288, "learning_rate": 1.9311116029192278e-05, "loss": 0.0189, "step": 4350 }, { "epoch": 1.2865151962230748, "grad_norm": 0.36844655871391296, "learning_rate": 1.909401999301993e-05, "loss": 0.0189, "step": 4360 }, { "epoch": 1.2894659191501918, "grad_norm": 0.26102837920188904, "learning_rate": 1.887786279656482e-05, "loss": 0.0197, "step": 4370 }, { "epoch": 1.2924166420773089, "grad_norm": 0.19810092449188232, "learning_rate": 1.8662651006095387e-05, "loss": 0.0195, "step": 4380 }, { "epoch": 1.295367365004426, "grad_norm": 0.21970173716545105, "learning_rate": 1.8448391159161204e-05, "loss": 0.0203, "step": 4390 }, { "epoch": 1.2983180879315432, "grad_norm": 0.32644402980804443, "learning_rate": 1.8235089764394408e-05, "loss": 0.0245, "step": 4400 }, { "epoch": 1.3012688108586603, "grad_norm": 0.19420267641544342, "learning_rate": 1.8022753301311935e-05, "loss": 0.0194, "step": 4410 }, { "epoch": 1.3042195337857776, "grad_norm": 0.20820944011211395, "learning_rate": 1.7811388220118707e-05, "loss": 0.02, "step": 4420 }, { "epoch": 1.3071702567128947, "grad_norm": 0.28606945276260376, "learning_rate": 1.7601000941511757e-05, "loss": 0.022, "step": 4430 }, { "epoch": 1.3101209796400117, "grad_norm": 0.3081311583518982, "learning_rate": 1.7391597856485083e-05, "loss": 0.024, "step": 4440 }, { "epoch": 1.313071702567129, "grad_norm": 0.15141037106513977, "learning_rate": 1.7183185326135543e-05, "loss": 0.0165, "step": 4450 }, { "epoch": 1.316022425494246, "grad_norm": 0.23260480165481567, "learning_rate": 1.6975769681469705e-05, "loss": 0.0198, "step": 4460 }, { "epoch": 1.3189731484213634, "grad_norm": 0.21865080296993256, "learning_rate": 1.676935722321139e-05, "loss": 0.0198, "step": 4470 }, { "epoch": 1.3219238713484804, "grad_norm": 0.19756555557250977, "learning_rate": 1.6563954221610355e-05, "loss": 0.0186, "step": 4480 }, { "epoch": 1.3248745942755975, "grad_norm": 0.2762605845928192, "learning_rate": 1.6359566916251845e-05, "loss": 0.0216, "step": 4490 }, { "epoch": 1.3278253172027146, "grad_norm": 0.22259187698364258, "learning_rate": 1.615620151586697e-05, "loss": 0.019, "step": 4500 }, { "epoch": 1.3307760401298319, "grad_norm": 0.2643989324569702, "learning_rate": 1.5953864198144135e-05, "loss": 0.0212, "step": 4510 }, { "epoch": 1.333726763056949, "grad_norm": 0.2703520357608795, "learning_rate": 1.5752561109541447e-05, "loss": 0.0178, "step": 4520 }, { "epoch": 1.3366774859840662, "grad_norm": 0.2908737063407898, "learning_rate": 1.5552298365099882e-05, "loss": 0.0184, "step": 4530 }, { "epoch": 1.3396282089111833, "grad_norm": 0.29694125056266785, "learning_rate": 1.5353082048257596e-05, "loss": 0.0186, "step": 4540 }, { "epoch": 1.3425789318383003, "grad_norm": 0.1732797622680664, "learning_rate": 1.5154918210665148e-05, "loss": 0.0167, "step": 4550 }, { "epoch": 1.3455296547654174, "grad_norm": 0.2552129924297333, "learning_rate": 1.4957812872001614e-05, "loss": 0.0161, "step": 4560 }, { "epoch": 1.3484803776925347, "grad_norm": 0.17536737024784088, "learning_rate": 1.4761772019791748e-05, "loss": 0.0216, "step": 4570 }, { "epoch": 1.3514311006196518, "grad_norm": 0.289426326751709, "learning_rate": 1.4566801609224096e-05, "loss": 0.0201, "step": 4580 }, { "epoch": 1.354381823546769, "grad_norm": 0.21466957032680511, "learning_rate": 1.4372907562970079e-05, "loss": 0.0215, "step": 4590 }, { "epoch": 1.3573325464738861, "grad_norm": 0.35995760560035706, "learning_rate": 1.4180095771004154e-05, "loss": 0.0203, "step": 4600 }, { "epoch": 1.3602832694010032, "grad_norm": 0.317004919052124, "learning_rate": 1.3988372090424773e-05, "loss": 0.0196, "step": 4610 }, { "epoch": 1.3632339923281203, "grad_norm": 0.17685896158218384, "learning_rate": 1.3797742345276521e-05, "loss": 0.0208, "step": 4620 }, { "epoch": 1.3661847152552375, "grad_norm": 0.17452839016914368, "learning_rate": 1.3608212326373249e-05, "loss": 0.0196, "step": 4630 }, { "epoch": 1.3691354381823546, "grad_norm": 0.2645012140274048, "learning_rate": 1.3419787791122062e-05, "loss": 0.0186, "step": 4640 }, { "epoch": 1.372086161109472, "grad_norm": 0.2119213342666626, "learning_rate": 1.323247446334847e-05, "loss": 0.02, "step": 4650 }, { "epoch": 1.375036884036589, "grad_norm": 0.3236892819404602, "learning_rate": 1.3046278033122577e-05, "loss": 0.0271, "step": 4660 }, { "epoch": 1.377987606963706, "grad_norm": 0.3879804015159607, "learning_rate": 1.286120415658611e-05, "loss": 0.0181, "step": 4670 }, { "epoch": 1.3809383298908233, "grad_norm": 0.22897237539291382, "learning_rate": 1.2677258455780683e-05, "loss": 0.0141, "step": 4680 }, { "epoch": 1.3838890528179404, "grad_norm": 0.19873371720314026, "learning_rate": 1.2494446518477022e-05, "loss": 0.0187, "step": 4690 }, { "epoch": 1.3868397757450575, "grad_norm": 0.287604957818985, "learning_rate": 1.2312773898005175e-05, "loss": 0.0211, "step": 4700 }, { "epoch": 1.3897904986721747, "grad_norm": 0.26559850573539734, "learning_rate": 1.2132246113085822e-05, "loss": 0.0179, "step": 4710 }, { "epoch": 1.3927412215992918, "grad_norm": 0.3130019009113312, "learning_rate": 1.1952868647662696e-05, "loss": 0.0214, "step": 4720 }, { "epoch": 1.3956919445264089, "grad_norm": 0.2792266309261322, "learning_rate": 1.1774646950735913e-05, "loss": 0.0195, "step": 4730 }, { "epoch": 1.3986426674535262, "grad_norm": 0.2873581349849701, "learning_rate": 1.1597586436196473e-05, "loss": 0.0174, "step": 4740 }, { "epoch": 1.4015933903806432, "grad_norm": 0.296114057302475, "learning_rate": 1.1421692482661856e-05, "loss": 0.0217, "step": 4750 }, { "epoch": 1.4045441133077605, "grad_norm": 0.19929592311382294, "learning_rate": 1.124697043331256e-05, "loss": 0.0183, "step": 4760 }, { "epoch": 1.4074948362348776, "grad_norm": 0.2661643326282501, "learning_rate": 1.107342559572977e-05, "loss": 0.0185, "step": 4770 }, { "epoch": 1.4104455591619947, "grad_norm": 0.3681509494781494, "learning_rate": 1.090106324173426e-05, "loss": 0.0193, "step": 4780 }, { "epoch": 1.4133962820891117, "grad_norm": 0.35690200328826904, "learning_rate": 1.0729888607226113e-05, "loss": 0.0178, "step": 4790 }, { "epoch": 1.416347005016229, "grad_norm": 0.2802860736846924, "learning_rate": 1.0559906892025745e-05, "loss": 0.02, "step": 4800 }, { "epoch": 1.419297727943346, "grad_norm": 0.27629420161247253, "learning_rate": 1.0391123259715906e-05, "loss": 0.0179, "step": 4810 }, { "epoch": 1.4222484508704634, "grad_norm": 0.21866770088672638, "learning_rate": 1.0223542837484839e-05, "loss": 0.016, "step": 4820 }, { "epoch": 1.4251991737975804, "grad_norm": 0.2228262722492218, "learning_rate": 1.0057170715970559e-05, "loss": 0.0195, "step": 4830 }, { "epoch": 1.4281498967246975, "grad_norm": 0.24920450150966644, "learning_rate": 9.892011949106172e-06, "loss": 0.0245, "step": 4840 }, { "epoch": 1.4311006196518146, "grad_norm": 0.18863655626773834, "learning_rate": 9.728071553966339e-06, "loss": 0.0185, "step": 4850 }, { "epoch": 1.4340513425789319, "grad_norm": 0.27390772104263306, "learning_rate": 9.56535451061496e-06, "loss": 0.0177, "step": 4860 }, { "epoch": 1.437002065506049, "grad_norm": 0.20433449745178223, "learning_rate": 9.403865761953779e-06, "loss": 0.017, "step": 4870 }, { "epoch": 1.4399527884331662, "grad_norm": 0.23858758807182312, "learning_rate": 9.243610213572285e-06, "loss": 0.0198, "step": 4880 }, { "epoch": 1.4429035113602833, "grad_norm": 0.3207745850086212, "learning_rate": 9.084592733598735e-06, "loss": 0.0174, "step": 4890 }, { "epoch": 1.4458542342874003, "grad_norm": 0.19312088191509247, "learning_rate": 8.92681815255219e-06, "loss": 0.0176, "step": 4900 }, { "epoch": 1.4488049572145176, "grad_norm": 0.1280824989080429, "learning_rate": 8.770291263195819e-06, "loss": 0.0177, "step": 4910 }, { "epoch": 1.4517556801416347, "grad_norm": 0.238336443901062, "learning_rate": 8.615016820391342e-06, "loss": 0.0153, "step": 4920 }, { "epoch": 1.4547064030687518, "grad_norm": 0.19052039086818695, "learning_rate": 8.460999540954517e-06, "loss": 0.0179, "step": 4930 }, { "epoch": 1.457657125995869, "grad_norm": 0.29646188020706177, "learning_rate": 8.308244103511909e-06, "loss": 0.018, "step": 4940 }, { "epoch": 1.4606078489229861, "grad_norm": 0.3154849708080292, "learning_rate": 8.156755148358764e-06, "loss": 0.0179, "step": 4950 }, { "epoch": 1.4635585718501032, "grad_norm": 0.28589123487472534, "learning_rate": 8.00653727731801e-06, "loss": 0.0144, "step": 4960 }, { "epoch": 1.4665092947772205, "grad_norm": 0.15766900777816772, "learning_rate": 7.857595053600513e-06, "loss": 0.0224, "step": 4970 }, { "epoch": 1.4694600177043375, "grad_norm": 0.22361356019973755, "learning_rate": 7.709933001666431e-06, "loss": 0.0229, "step": 4980 }, { "epoch": 1.4724107406314548, "grad_norm": 0.18924807012081146, "learning_rate": 7.56355560708778e-06, "loss": 0.0178, "step": 4990 }, { "epoch": 1.475361463558572, "grad_norm": 0.2892850935459137, "learning_rate": 7.418467316412158e-06, "loss": 0.0178, "step": 5000 }, { "epoch": 1.478312186485689, "grad_norm": 0.2627609074115753, "learning_rate": 7.2746725370277435e-06, "loss": 0.021, "step": 5010 }, { "epoch": 1.481262909412806, "grad_norm": 0.23089174926280975, "learning_rate": 7.132175637029293e-06, "loss": 0.0203, "step": 5020 }, { "epoch": 1.4842136323399233, "grad_norm": 0.22778406739234924, "learning_rate": 6.9909809450855345e-06, "loss": 0.018, "step": 5030 }, { "epoch": 1.4871643552670404, "grad_norm": 0.2813952565193176, "learning_rate": 6.851092750307686e-06, "loss": 0.0176, "step": 5040 }, { "epoch": 1.4901150781941577, "grad_norm": 0.24083396792411804, "learning_rate": 6.712515302119077e-06, "loss": 0.0182, "step": 5050 }, { "epoch": 1.4930658011212747, "grad_norm": 0.24483950436115265, "learning_rate": 6.575252810126143e-06, "loss": 0.0184, "step": 5060 }, { "epoch": 1.4960165240483918, "grad_norm": 0.24604901671409607, "learning_rate": 6.439309443990532e-06, "loss": 0.0139, "step": 5070 }, { "epoch": 1.4989672469755089, "grad_norm": 0.2011985033750534, "learning_rate": 6.304689333302416e-06, "loss": 0.0179, "step": 5080 }, { "epoch": 1.5019179699026262, "grad_norm": 0.1611073762178421, "learning_rate": 6.171396567455051e-06, "loss": 0.0161, "step": 5090 }, { "epoch": 1.5048686928297434, "grad_norm": 0.17760014533996582, "learning_rate": 6.039435195520604e-06, "loss": 0.0202, "step": 5100 }, { "epoch": 1.5078194157568605, "grad_norm": 0.16185684502124786, "learning_rate": 5.908809226127054e-06, "loss": 0.0158, "step": 5110 }, { "epoch": 1.5107701386839776, "grad_norm": 0.2905328571796417, "learning_rate": 5.779522627336537e-06, "loss": 0.022, "step": 5120 }, { "epoch": 1.5137208616110946, "grad_norm": 0.1555669605731964, "learning_rate": 5.651579326524709e-06, "loss": 0.0156, "step": 5130 }, { "epoch": 1.5166715845382117, "grad_norm": 0.26654595136642456, "learning_rate": 5.524983210261481e-06, "loss": 0.0186, "step": 5140 }, { "epoch": 1.519622307465329, "grad_norm": 0.19950677454471588, "learning_rate": 5.399738124192988e-06, "loss": 0.0146, "step": 5150 }, { "epoch": 1.5225730303924463, "grad_norm": 0.2644423246383667, "learning_rate": 5.2758478729247164e-06, "loss": 0.0175, "step": 5160 }, { "epoch": 1.5255237533195634, "grad_norm": 0.21380996704101562, "learning_rate": 5.153316219905946e-06, "loss": 0.015, "step": 5170 }, { "epoch": 1.5284744762466804, "grad_norm": 0.2527577579021454, "learning_rate": 5.032146887315448e-06, "loss": 0.0154, "step": 5180 }, { "epoch": 1.5314251991737975, "grad_norm": 0.24711045622825623, "learning_rate": 4.91234355594839e-06, "loss": 0.0182, "step": 5190 }, { "epoch": 1.5343759221009146, "grad_norm": 0.19659630954265594, "learning_rate": 4.7939098651045235e-06, "loss": 0.0208, "step": 5200 }, { "epoch": 1.5373266450280318, "grad_norm": 0.19166162610054016, "learning_rate": 4.67684941247768e-06, "loss": 0.0175, "step": 5210 }, { "epoch": 1.5402773679551491, "grad_norm": 0.24067266285419464, "learning_rate": 4.5611657540464036e-06, "loss": 0.0184, "step": 5220 }, { "epoch": 1.5432280908822662, "grad_norm": 0.17422425746917725, "learning_rate": 4.446862403965984e-06, "loss": 0.017, "step": 5230 }, { "epoch": 1.5461788138093833, "grad_norm": 0.1915813684463501, "learning_rate": 4.333942834461702e-06, "loss": 0.0143, "step": 5240 }, { "epoch": 1.5491295367365003, "grad_norm": 0.1772589534521103, "learning_rate": 4.222410475723326e-06, "loss": 0.014, "step": 5250 }, { "epoch": 1.5520802596636176, "grad_norm": 0.13021902740001678, "learning_rate": 4.112268715800943e-06, "loss": 0.0169, "step": 5260 }, { "epoch": 1.5550309825907347, "grad_norm": 0.20913878083229065, "learning_rate": 4.003520900502028e-06, "loss": 0.0169, "step": 5270 }, { "epoch": 1.557981705517852, "grad_norm": 0.14371579885482788, "learning_rate": 3.8961703332898e-06, "loss": 0.0171, "step": 5280 }, { "epoch": 1.560932428444969, "grad_norm": 0.21924927830696106, "learning_rate": 3.790220275182854e-06, "loss": 0.0192, "step": 5290 }, { "epoch": 1.5638831513720861, "grad_norm": 0.2913877069950104, "learning_rate": 3.685673944656176e-06, "loss": 0.0169, "step": 5300 }, { "epoch": 1.5668338742992032, "grad_norm": 0.18612533807754517, "learning_rate": 3.582534517543268e-06, "loss": 0.0157, "step": 5310 }, { "epoch": 1.5697845972263205, "grad_norm": 0.28482428193092346, "learning_rate": 3.4808051269397512e-06, "loss": 0.0176, "step": 5320 }, { "epoch": 1.5727353201534378, "grad_norm": 0.2025042325258255, "learning_rate": 3.380488863108183e-06, "loss": 0.0159, "step": 5330 }, { "epoch": 1.5756860430805548, "grad_norm": 0.1597754955291748, "learning_rate": 3.2815887733841365e-06, "loss": 0.0162, "step": 5340 }, { "epoch": 1.5786367660076719, "grad_norm": 0.175078347325325, "learning_rate": 3.1841078620836683e-06, "loss": 0.0165, "step": 5350 }, { "epoch": 1.581587488934789, "grad_norm": 0.1488451063632965, "learning_rate": 3.0880490904120874e-06, "loss": 0.0162, "step": 5360 }, { "epoch": 1.584538211861906, "grad_norm": 0.13998070359230042, "learning_rate": 2.9934153763739205e-06, "loss": 0.0195, "step": 5370 }, { "epoch": 1.5874889347890233, "grad_norm": 0.21413278579711914, "learning_rate": 2.9002095946843277e-06, "loss": 0.0166, "step": 5380 }, { "epoch": 1.5904396577161406, "grad_norm": 0.28923460841178894, "learning_rate": 2.8084345766817676e-06, "loss": 0.0174, "step": 5390 }, { "epoch": 1.5933903806432577, "grad_norm": 0.24130022525787354, "learning_rate": 2.718093110241976e-06, "loss": 0.0166, "step": 5400 }, { "epoch": 1.5963411035703747, "grad_norm": 0.2335025519132614, "learning_rate": 2.6291879396933004e-06, "loss": 0.0193, "step": 5410 }, { "epoch": 1.5992918264974918, "grad_norm": 0.23496149480342865, "learning_rate": 2.541721765733318e-06, "loss": 0.0165, "step": 5420 }, { "epoch": 1.6022425494246089, "grad_norm": 0.30696988105773926, "learning_rate": 2.455697245346783e-06, "loss": 0.0152, "step": 5430 }, { "epoch": 1.6051932723517262, "grad_norm": 0.16828006505966187, "learning_rate": 2.371116991724953e-06, "loss": 0.0178, "step": 5440 }, { "epoch": 1.6081439952788434, "grad_norm": 0.19028416275978088, "learning_rate": 2.2879835741861586e-06, "loss": 0.0195, "step": 5450 }, { "epoch": 1.6110947182059605, "grad_norm": 0.2158205509185791, "learning_rate": 2.206299518097804e-06, "loss": 0.0148, "step": 5460 }, { "epoch": 1.6140454411330776, "grad_norm": 0.1675664186477661, "learning_rate": 2.1260673047996227e-06, "loss": 0.0179, "step": 5470 }, { "epoch": 1.6169961640601946, "grad_norm": 0.28710392117500305, "learning_rate": 2.047289371528299e-06, "loss": 0.0164, "step": 5480 }, { "epoch": 1.619946886987312, "grad_norm": 0.2290625125169754, "learning_rate": 1.96996811134344e-06, "loss": 0.0132, "step": 5490 }, { "epoch": 1.622897609914429, "grad_norm": 0.13754595816135406, "learning_rate": 1.8941058730549132e-06, "loss": 0.0177, "step": 5500 }, { "epoch": 1.6258483328415463, "grad_norm": 0.2654421329498291, "learning_rate": 1.8197049611514194e-06, "loss": 0.0184, "step": 5510 }, { "epoch": 1.6287990557686634, "grad_norm": 0.11221639811992645, "learning_rate": 1.7467676357305561e-06, "loss": 0.0142, "step": 5520 }, { "epoch": 1.6317497786957804, "grad_norm": 0.2562015652656555, "learning_rate": 1.6752961124301415e-06, "loss": 0.0252, "step": 5530 }, { "epoch": 1.6347005016228975, "grad_norm": 0.11668754369020462, "learning_rate": 1.6052925623609049e-06, "loss": 0.016, "step": 5540 }, { "epoch": 1.6376512245500148, "grad_norm": 0.21560530364513397, "learning_rate": 1.5367591120405256e-06, "loss": 0.0165, "step": 5550 }, { "epoch": 1.640601947477132, "grad_norm": 0.2209858000278473, "learning_rate": 1.4696978433290653e-06, "loss": 0.0155, "step": 5560 }, { "epoch": 1.6435526704042491, "grad_norm": 0.15884585678577423, "learning_rate": 1.4041107933656928e-06, "loss": 0.0127, "step": 5570 }, { "epoch": 1.6465033933313662, "grad_norm": 0.20755048096179962, "learning_rate": 1.339999954506821e-06, "loss": 0.0168, "step": 5580 }, { "epoch": 1.6494541162584833, "grad_norm": 0.21209847927093506, "learning_rate": 1.2773672742655784e-06, "loss": 0.0161, "step": 5590 }, { "epoch": 1.6524048391856003, "grad_norm": 0.22342316806316376, "learning_rate": 1.2162146552526399e-06, "loss": 0.0163, "step": 5600 }, { "epoch": 1.6553555621127176, "grad_norm": 0.2544778287410736, "learning_rate": 1.1565439551184664e-06, "loss": 0.0195, "step": 5610 }, { "epoch": 1.658306285039835, "grad_norm": 0.2241518646478653, "learning_rate": 1.0983569864968346e-06, "loss": 0.018, "step": 5620 }, { "epoch": 1.661257007966952, "grad_norm": 0.142816424369812, "learning_rate": 1.0416555169497688e-06, "loss": 0.0147, "step": 5630 }, { "epoch": 1.664207730894069, "grad_norm": 0.2596038579940796, "learning_rate": 9.864412689139123e-07, "loss": 0.0218, "step": 5640 }, { "epoch": 1.667158453821186, "grad_norm": 0.28301334381103516, "learning_rate": 9.327159196481138e-07, "loss": 0.0132, "step": 5650 }, { "epoch": 1.6701091767483032, "grad_norm": 0.18248535692691803, "learning_rate": 8.804811011825398e-07, "loss": 0.0174, "step": 5660 }, { "epoch": 1.6730598996754205, "grad_norm": 0.2579980790615082, "learning_rate": 8.297384002690866e-07, "loss": 0.0172, "step": 5670 }, { "epoch": 1.6760106226025377, "grad_norm": 0.19259390234947205, "learning_rate": 7.804893583331696e-07, "loss": 0.0184, "step": 5680 }, { "epoch": 1.6789613455296548, "grad_norm": 0.18520526587963104, "learning_rate": 7.32735471426893e-07, "loss": 0.0181, "step": 5690 }, { "epoch": 1.6819120684567719, "grad_norm": 0.130854994058609, "learning_rate": 6.864781901836259e-07, "loss": 0.016, "step": 5700 }, { "epoch": 1.684862791383889, "grad_norm": 0.2609096169471741, "learning_rate": 6.417189197739093e-07, "loss": 0.0188, "step": 5710 }, { "epoch": 1.6878135143110062, "grad_norm": 0.1988040804862976, "learning_rate": 5.984590198627849e-07, "loss": 0.0223, "step": 5720 }, { "epoch": 1.6907642372381233, "grad_norm": 0.17318883538246155, "learning_rate": 5.566998045685112e-07, "loss": 0.0193, "step": 5730 }, { "epoch": 1.6937149601652406, "grad_norm": 0.20778967440128326, "learning_rate": 5.164425424226016e-07, "loss": 0.0193, "step": 5740 }, { "epoch": 1.6966656830923577, "grad_norm": 0.1542847752571106, "learning_rate": 4.776884563313266e-07, "loss": 0.0141, "step": 5750 }, { "epoch": 1.6996164060194747, "grad_norm": 0.1506889909505844, "learning_rate": 4.404387235385443e-07, "loss": 0.0164, "step": 5760 }, { "epoch": 1.7025671289465918, "grad_norm": 0.1802382916212082, "learning_rate": 4.0469447558995065e-07, "loss": 0.0189, "step": 5770 }, { "epoch": 1.705517851873709, "grad_norm": 0.20460198819637299, "learning_rate": 3.7045679829870175e-07, "loss": 0.0189, "step": 5780 }, { "epoch": 1.7084685748008264, "grad_norm": 0.24809370934963226, "learning_rate": 3.377267317124233e-07, "loss": 0.0184, "step": 5790 }, { "epoch": 1.7114192977279434, "grad_norm": 0.16671574115753174, "learning_rate": 3.0650527008162513e-07, "loss": 0.0125, "step": 5800 }, { "epoch": 1.7143700206550605, "grad_norm": 0.15032552182674408, "learning_rate": 2.767933618295082e-07, "loss": 0.0128, "step": 5810 }, { "epoch": 1.7173207435821776, "grad_norm": 0.1787770688533783, "learning_rate": 2.485919095231326e-07, "loss": 0.0159, "step": 5820 }, { "epoch": 1.7202714665092946, "grad_norm": 0.24622216820716858, "learning_rate": 2.219017698460002e-07, "loss": 0.0191, "step": 5830 }, { "epoch": 1.723222189436412, "grad_norm": 0.2202993631362915, "learning_rate": 1.9672375357206452e-07, "loss": 0.0171, "step": 5840 }, { "epoch": 1.7261729123635292, "grad_norm": 0.27309519052505493, "learning_rate": 1.73058625541056e-07, "loss": 0.0198, "step": 5850 }, { "epoch": 1.7291236352906463, "grad_norm": 0.18019090592861176, "learning_rate": 1.5090710463527836e-07, "loss": 0.0164, "step": 5860 }, { "epoch": 1.7320743582177633, "grad_norm": 0.15442153811454773, "learning_rate": 1.3026986375776485e-07, "loss": 0.0189, "step": 5870 }, { "epoch": 1.7350250811448804, "grad_norm": 0.27966898679733276, "learning_rate": 1.1114752981183917e-07, "loss": 0.0227, "step": 5880 }, { "epoch": 1.7379758040719975, "grad_norm": 0.21470524370670319, "learning_rate": 9.354068368204739e-08, "loss": 0.016, "step": 5890 }, { "epoch": 1.7409265269991148, "grad_norm": 0.12310141324996948, "learning_rate": 7.744986021656076e-08, "loss": 0.0168, "step": 5900 }, { "epoch": 1.743877249926232, "grad_norm": 0.17956896126270294, "learning_rate": 6.287554821087783e-08, "loss": 0.0163, "step": 5910 }, { "epoch": 1.7468279728533491, "grad_norm": 0.14682123064994812, "learning_rate": 4.981819039300284e-08, "loss": 0.0155, "step": 5920 }, { "epoch": 1.7497786957804662, "grad_norm": 0.23914210498332977, "learning_rate": 3.827818341000655e-08, "loss": 0.0185, "step": 5930 }, { "epoch": 1.7527294187075833, "grad_norm": 0.2295534461736679, "learning_rate": 2.8255877815946963e-08, "loss": 0.0154, "step": 5940 }, { "epoch": 1.7556801416347005, "grad_norm": 0.14401505887508392, "learning_rate": 1.9751578061244504e-08, "loss": 0.0158, "step": 5950 }, { "epoch": 1.7586308645618176, "grad_norm": 0.23302072286605835, "learning_rate": 1.2765542483417214e-08, "loss": 0.016, "step": 5960 }, { "epoch": 1.761581587488935, "grad_norm": 0.11384108662605286, "learning_rate": 7.2979832992592365e-09, "loss": 0.0189, "step": 5970 }, { "epoch": 1.764532310416052, "grad_norm": 0.22752645611763, "learning_rate": 3.349066598362649e-09, "loss": 0.0205, "step": 5980 }, { "epoch": 1.767483033343169, "grad_norm": 0.19875404238700867, "learning_rate": 9.189123380826114e-10, "loss": 0.0157, "step": 5990 }, { "epoch": 1.770433756270286, "grad_norm": 0.22663816809654236, "learning_rate": 7.594339912486703e-12, "loss": 0.0172, "step": 6000 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 2000, "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 }