Buckets:
| { | |
| "best_global_step": 130000, | |
| "best_metric": 0.7778754830360413, | |
| "best_model_checkpoint": "cache/medium/checkpoints/checkpoint-130000", | |
| "epoch": 2.6650266502665025, | |
| "eval_steps": 5000, | |
| "global_step": 130000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001025010250102501, | |
| "grad_norm": 1.1566449403762817, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.37623008728027346, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.002050020500205002, | |
| "grad_norm": 1.1058133840560913, | |
| "learning_rate": 1.9800000000000004e-05, | |
| "loss": 0.3689537048339844, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.003075030750307503, | |
| "grad_norm": 1.193008303642273, | |
| "learning_rate": 2.98e-05, | |
| "loss": 0.3648074722290039, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.004100041000410004, | |
| "grad_norm": 1.2246938943862915, | |
| "learning_rate": 3.9800000000000005e-05, | |
| "loss": 0.3658078384399414, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.005125051250512505, | |
| "grad_norm": 1.091650366783142, | |
| "learning_rate": 4.9800000000000004e-05, | |
| "loss": 0.37023128509521486, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.006150061500615006, | |
| "grad_norm": 1.1708260774612427, | |
| "learning_rate": 5.9800000000000003e-05, | |
| "loss": 0.36595996856689456, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.007175071750717507, | |
| "grad_norm": 1.133751392364502, | |
| "learning_rate": 6.98e-05, | |
| "loss": 0.36910999298095704, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.008200082000820008, | |
| "grad_norm": 1.0874284505844116, | |
| "learning_rate": 7.98e-05, | |
| "loss": 0.36752891540527344, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.00922509225092251, | |
| "grad_norm": 1.067591667175293, | |
| "learning_rate": 8.98e-05, | |
| "loss": 0.3795060348510742, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.01025010250102501, | |
| "grad_norm": 1.1766570806503296, | |
| "learning_rate": 9.98e-05, | |
| "loss": 0.38381763458251955, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.011275112751127511, | |
| "grad_norm": 1.12791907787323, | |
| "learning_rate": 9.996640153592979e-05, | |
| "loss": 0.37818374633789065, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.012300123001230012, | |
| "grad_norm": 1.1951355934143066, | |
| "learning_rate": 9.993211738891936e-05, | |
| "loss": 0.38590068817138673, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.013325133251332513, | |
| "grad_norm": 1.1941357851028442, | |
| "learning_rate": 9.989783324190894e-05, | |
| "loss": 0.3888316345214844, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.014350143501435014, | |
| "grad_norm": 1.1433979272842407, | |
| "learning_rate": 9.986354909489853e-05, | |
| "loss": 0.39156036376953124, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.015375153751537515, | |
| "grad_norm": 1.1588531732559204, | |
| "learning_rate": 9.98292649478881e-05, | |
| "loss": 0.3940975952148438, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.016400164001640016, | |
| "grad_norm": 1.160005807876587, | |
| "learning_rate": 9.979498080087768e-05, | |
| "loss": 0.3919621276855469, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.017425174251742517, | |
| "grad_norm": 1.2051060199737549, | |
| "learning_rate": 9.976069665386726e-05, | |
| "loss": 0.38980655670166015, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.01845018450184502, | |
| "grad_norm": 1.1003724336624146, | |
| "learning_rate": 9.972641250685684e-05, | |
| "loss": 0.3979477310180664, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.01947519475194752, | |
| "grad_norm": 1.1645162105560303, | |
| "learning_rate": 9.969212835984641e-05, | |
| "loss": 0.392269401550293, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.02050020500205002, | |
| "grad_norm": 1.1680755615234375, | |
| "learning_rate": 9.965784421283599e-05, | |
| "loss": 0.3947479248046875, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.02152521525215252, | |
| "grad_norm": 1.1269372701644897, | |
| "learning_rate": 9.962356006582558e-05, | |
| "loss": 0.39560958862304685, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.022550225502255022, | |
| "grad_norm": 1.1528249979019165, | |
| "learning_rate": 9.958927591881514e-05, | |
| "loss": 0.39358020782470704, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.023575235752357523, | |
| "grad_norm": 1.1549291610717773, | |
| "learning_rate": 9.955499177180471e-05, | |
| "loss": 0.39642642974853515, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.024600246002460024, | |
| "grad_norm": 1.1308238506317139, | |
| "learning_rate": 9.95207076247943e-05, | |
| "loss": 0.3923631286621094, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.025625256252562525, | |
| "grad_norm": 1.1449800729751587, | |
| "learning_rate": 9.948642347778388e-05, | |
| "loss": 0.3856623458862305, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.026650266502665027, | |
| "grad_norm": 1.1634982824325562, | |
| "learning_rate": 9.945213933077345e-05, | |
| "loss": 0.3931452941894531, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.027675276752767528, | |
| "grad_norm": 1.2422840595245361, | |
| "learning_rate": 9.941785518376302e-05, | |
| "loss": 0.3958161544799805, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.02870028700287003, | |
| "grad_norm": 1.1141917705535889, | |
| "learning_rate": 9.938357103675261e-05, | |
| "loss": 0.3923382949829102, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.02972529725297253, | |
| "grad_norm": 1.0814729928970337, | |
| "learning_rate": 9.934928688974219e-05, | |
| "loss": 0.3869150924682617, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.03075030750307503, | |
| "grad_norm": 1.0617198944091797, | |
| "learning_rate": 9.931500274273176e-05, | |
| "loss": 0.3899386215209961, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.03177531775317753, | |
| "grad_norm": 1.0915844440460205, | |
| "learning_rate": 9.928071859572135e-05, | |
| "loss": 0.39418777465820315, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.03280032800328003, | |
| "grad_norm": 1.098642110824585, | |
| "learning_rate": 9.924643444871093e-05, | |
| "loss": 0.39837581634521485, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.033825338253382534, | |
| "grad_norm": 1.0125762224197388, | |
| "learning_rate": 9.92121503017005e-05, | |
| "loss": 0.3964727020263672, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.034850348503485035, | |
| "grad_norm": 1.0763367414474487, | |
| "learning_rate": 9.917786615469007e-05, | |
| "loss": 0.39296024322509765, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.035875358753587536, | |
| "grad_norm": 1.1015264987945557, | |
| "learning_rate": 9.914358200767966e-05, | |
| "loss": 0.3892650604248047, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.03690036900369004, | |
| "grad_norm": 1.040937900543213, | |
| "learning_rate": 9.910929786066924e-05, | |
| "loss": 0.3961595916748047, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.03792537925379254, | |
| "grad_norm": 1.1278302669525146, | |
| "learning_rate": 9.90750137136588e-05, | |
| "loss": 0.3906804275512695, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.03895038950389504, | |
| "grad_norm": 1.1163142919540405, | |
| "learning_rate": 9.904072956664839e-05, | |
| "loss": 0.40465126037597654, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.03997539975399754, | |
| "grad_norm": 1.0773122310638428, | |
| "learning_rate": 9.900644541963796e-05, | |
| "loss": 0.39360107421875, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.04100041000410004, | |
| "grad_norm": 1.0944104194641113, | |
| "learning_rate": 9.897216127262754e-05, | |
| "loss": 0.39632961273193357, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.04202542025420254, | |
| "grad_norm": 1.125977873802185, | |
| "learning_rate": 9.893787712561711e-05, | |
| "loss": 0.39472366333007813, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.04305043050430504, | |
| "grad_norm": 1.0370248556137085, | |
| "learning_rate": 9.89035929786067e-05, | |
| "loss": 0.3972608184814453, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.044075440754407544, | |
| "grad_norm": 1.053231120109558, | |
| "learning_rate": 9.886930883159627e-05, | |
| "loss": 0.39653865814208983, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.045100451004510045, | |
| "grad_norm": 1.0601778030395508, | |
| "learning_rate": 9.883502468458585e-05, | |
| "loss": 0.3921388626098633, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.046125461254612546, | |
| "grad_norm": 1.0939007997512817, | |
| "learning_rate": 9.880074053757544e-05, | |
| "loss": 0.39368816375732424, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.04715047150471505, | |
| "grad_norm": 0.9953641891479492, | |
| "learning_rate": 9.876645639056501e-05, | |
| "loss": 0.38961708068847656, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.04817548175481755, | |
| "grad_norm": 1.124768853187561, | |
| "learning_rate": 9.873217224355459e-05, | |
| "loss": 0.394456672668457, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.04920049200492005, | |
| "grad_norm": 1.1697200536727905, | |
| "learning_rate": 9.869788809654416e-05, | |
| "loss": 0.3938796234130859, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.05022550225502255, | |
| "grad_norm": 1.1384869813919067, | |
| "learning_rate": 9.866360394953375e-05, | |
| "loss": 0.39263980865478515, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.05125051250512505, | |
| "grad_norm": 1.117107629776001, | |
| "learning_rate": 9.862931980252332e-05, | |
| "loss": 0.39507198333740234, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.05227552275522755, | |
| "grad_norm": 0.988071858882904, | |
| "learning_rate": 9.859503565551289e-05, | |
| "loss": 0.392652587890625, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.05330053300533005, | |
| "grad_norm": 1.1315289735794067, | |
| "learning_rate": 9.856075150850247e-05, | |
| "loss": 0.39317310333251954, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.054325543255432554, | |
| "grad_norm": 1.1007463932037354, | |
| "learning_rate": 9.852646736149205e-05, | |
| "loss": 0.396639404296875, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.055350553505535055, | |
| "grad_norm": 1.1778672933578491, | |
| "learning_rate": 9.849218321448162e-05, | |
| "loss": 0.39288509368896485, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.056375563755637556, | |
| "grad_norm": 1.0827738046646118, | |
| "learning_rate": 9.845789906747121e-05, | |
| "loss": 0.39339523315429686, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.05740057400574006, | |
| "grad_norm": 1.0607283115386963, | |
| "learning_rate": 9.842361492046079e-05, | |
| "loss": 0.3987705612182617, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.05842558425584256, | |
| "grad_norm": 0.9470078349113464, | |
| "learning_rate": 9.838933077345036e-05, | |
| "loss": 0.3905525207519531, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.05945059450594506, | |
| "grad_norm": 1.0350408554077148, | |
| "learning_rate": 9.835504662643994e-05, | |
| "loss": 0.3879140853881836, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.06047560475604756, | |
| "grad_norm": 1.056072473526001, | |
| "learning_rate": 9.832076247942952e-05, | |
| "loss": 0.38652332305908205, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.06150061500615006, | |
| "grad_norm": 1.0818918943405151, | |
| "learning_rate": 9.82864783324191e-05, | |
| "loss": 0.3907636642456055, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.06252562525625256, | |
| "grad_norm": 1.1062462329864502, | |
| "learning_rate": 9.825219418540867e-05, | |
| "loss": 0.39525394439697265, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.06355063550635506, | |
| "grad_norm": 1.0979666709899902, | |
| "learning_rate": 9.821791003839825e-05, | |
| "loss": 0.3915418243408203, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.06457564575645756, | |
| "grad_norm": 1.0048003196716309, | |
| "learning_rate": 9.818362589138784e-05, | |
| "loss": 0.3946091842651367, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.06560065600656007, | |
| "grad_norm": 1.0149345397949219, | |
| "learning_rate": 9.814934174437741e-05, | |
| "loss": 0.3916042709350586, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.06662566625666257, | |
| "grad_norm": 1.103816270828247, | |
| "learning_rate": 9.811505759736697e-05, | |
| "loss": 0.39914703369140625, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.06765067650676507, | |
| "grad_norm": 0.9210329651832581, | |
| "learning_rate": 9.808077345035656e-05, | |
| "loss": 0.39466846466064454, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.06867568675686757, | |
| "grad_norm": 0.9868373274803162, | |
| "learning_rate": 9.804648930334613e-05, | |
| "loss": 0.3939936065673828, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.06970069700697007, | |
| "grad_norm": 1.116458773612976, | |
| "learning_rate": 9.801220515633571e-05, | |
| "loss": 0.39745388031005857, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.07072570725707257, | |
| "grad_norm": 1.0137195587158203, | |
| "learning_rate": 9.79779210093253e-05, | |
| "loss": 0.3907796478271484, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.07175071750717507, | |
| "grad_norm": 1.033501386642456, | |
| "learning_rate": 9.794363686231487e-05, | |
| "loss": 0.38831123352050784, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.07277572775727757, | |
| "grad_norm": 1.176866054534912, | |
| "learning_rate": 9.790935271530445e-05, | |
| "loss": 0.39310401916503906, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.07380073800738007, | |
| "grad_norm": 0.9767327904701233, | |
| "learning_rate": 9.787506856829402e-05, | |
| "loss": 0.3894063949584961, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.07482574825748257, | |
| "grad_norm": 0.9704754948616028, | |
| "learning_rate": 9.784078442128361e-05, | |
| "loss": 0.38792617797851564, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.07585075850758508, | |
| "grad_norm": 1.0809571743011475, | |
| "learning_rate": 9.780650027427318e-05, | |
| "loss": 0.3975833511352539, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.07687576875768758, | |
| "grad_norm": 1.03998863697052, | |
| "learning_rate": 9.777221612726276e-05, | |
| "loss": 0.38746932983398436, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.07790077900779008, | |
| "grad_norm": 1.028603196144104, | |
| "learning_rate": 9.773793198025233e-05, | |
| "loss": 0.3898578643798828, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.07892578925789258, | |
| "grad_norm": 1.0210936069488525, | |
| "learning_rate": 9.770364783324192e-05, | |
| "loss": 0.39279804229736326, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.07995079950799508, | |
| "grad_norm": 1.0008447170257568, | |
| "learning_rate": 9.76693636862315e-05, | |
| "loss": 0.3922360229492188, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.08097580975809758, | |
| "grad_norm": 1.0721122026443481, | |
| "learning_rate": 9.763507953922106e-05, | |
| "loss": 0.3890448760986328, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.08200082000820008, | |
| "grad_norm": 0.9801535606384277, | |
| "learning_rate": 9.760079539221065e-05, | |
| "loss": 0.3894508743286133, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.08302583025830258, | |
| "grad_norm": 0.9887153506278992, | |
| "learning_rate": 9.756651124520022e-05, | |
| "loss": 0.39537227630615235, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.08405084050840508, | |
| "grad_norm": 0.9575806856155396, | |
| "learning_rate": 9.75322270981898e-05, | |
| "loss": 0.39471725463867186, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.08507585075850758, | |
| "grad_norm": 1.0587745904922485, | |
| "learning_rate": 9.749794295117938e-05, | |
| "loss": 0.3896254348754883, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.08610086100861009, | |
| "grad_norm": 1.176320195198059, | |
| "learning_rate": 9.746365880416896e-05, | |
| "loss": 0.39947822570800784, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.08712587125871259, | |
| "grad_norm": 0.9555501341819763, | |
| "learning_rate": 9.742937465715853e-05, | |
| "loss": 0.3889955520629883, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.08815088150881509, | |
| "grad_norm": 1.0682436227798462, | |
| "learning_rate": 9.739509051014811e-05, | |
| "loss": 0.39127803802490235, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.08917589175891759, | |
| "grad_norm": 0.8655234575271606, | |
| "learning_rate": 9.73608063631377e-05, | |
| "loss": 0.38822830200195313, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.09020090200902009, | |
| "grad_norm": 1.0104262828826904, | |
| "learning_rate": 9.732652221612727e-05, | |
| "loss": 0.3923531723022461, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.09122591225912259, | |
| "grad_norm": 1.022980809211731, | |
| "learning_rate": 9.729223806911685e-05, | |
| "loss": 0.3968360137939453, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.09225092250922509, | |
| "grad_norm": 1.023424506187439, | |
| "learning_rate": 9.725795392210642e-05, | |
| "loss": 0.3920841598510742, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.09327593275932759, | |
| "grad_norm": 1.0785950422286987, | |
| "learning_rate": 9.722366977509601e-05, | |
| "loss": 0.3953189468383789, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.0943009430094301, | |
| "grad_norm": 0.9991852045059204, | |
| "learning_rate": 9.718938562808558e-05, | |
| "loss": 0.39019107818603516, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.0953259532595326, | |
| "grad_norm": 1.0142594575881958, | |
| "learning_rate": 9.715510148107514e-05, | |
| "loss": 0.3869612503051758, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.0963509635096351, | |
| "grad_norm": 1.0702730417251587, | |
| "learning_rate": 9.712081733406473e-05, | |
| "loss": 0.3899692535400391, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.0973759737597376, | |
| "grad_norm": 1.0036945343017578, | |
| "learning_rate": 9.708653318705431e-05, | |
| "loss": 0.3927028656005859, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.0984009840098401, | |
| "grad_norm": 0.9355688095092773, | |
| "learning_rate": 9.705224904004388e-05, | |
| "loss": 0.39271892547607423, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.0994259942599426, | |
| "grad_norm": 0.9752117395401001, | |
| "learning_rate": 9.701796489303347e-05, | |
| "loss": 0.3905238723754883, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.1004510045100451, | |
| "grad_norm": 1.0007034540176392, | |
| "learning_rate": 9.698368074602304e-05, | |
| "loss": 0.3868400192260742, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.1014760147601476, | |
| "grad_norm": 1.154180645942688, | |
| "learning_rate": 9.694939659901262e-05, | |
| "loss": 0.39537399291992187, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.1025010250102501, | |
| "grad_norm": 1.0214035511016846, | |
| "learning_rate": 9.69151124520022e-05, | |
| "loss": 0.39350807189941406, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.1025010250102501, | |
| "eval_loss": 0.4790654182434082, | |
| "eval_runtime": 4.5547, | |
| "eval_samples_per_second": 5488.829, | |
| "eval_steps_per_second": 5.489, | |
| "eval_teacher_cosine_loss": 0.293754518032074, | |
| "eval_teacher_cosine_mean": 0.706245481967926, | |
| "eval_teacher_cosine_p10": 0.6258020401000977, | |
| "eval_teacher_cosine_p50": 0.7147992849349976, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.1035260352603526, | |
| "grad_norm": 1.0060973167419434, | |
| "learning_rate": 9.688082830499178e-05, | |
| "loss": 0.3890859603881836, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.1045510455104551, | |
| "grad_norm": 0.9707800149917603, | |
| "learning_rate": 9.684654415798136e-05, | |
| "loss": 0.39221298217773437, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.1055760557605576, | |
| "grad_norm": 0.9550859928131104, | |
| "learning_rate": 9.681226001097093e-05, | |
| "loss": 0.3898042678833008, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.1066010660106601, | |
| "grad_norm": 1.072510838508606, | |
| "learning_rate": 9.67779758639605e-05, | |
| "loss": 0.3880596542358398, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.10762607626076261, | |
| "grad_norm": 1.0140739679336548, | |
| "learning_rate": 9.67436917169501e-05, | |
| "loss": 0.38892894744873047, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.10865108651086511, | |
| "grad_norm": 1.1300958395004272, | |
| "learning_rate": 9.670940756993967e-05, | |
| "loss": 0.3930231475830078, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.10967609676096761, | |
| "grad_norm": 0.9301708340644836, | |
| "learning_rate": 9.667512342292923e-05, | |
| "loss": 0.3831196975708008, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.11070110701107011, | |
| "grad_norm": 1.0083624124526978, | |
| "learning_rate": 9.664083927591882e-05, | |
| "loss": 0.3794632720947266, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.11172611726117261, | |
| "grad_norm": 0.9093284606933594, | |
| "learning_rate": 9.66065551289084e-05, | |
| "loss": 0.3853170013427734, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.11275112751127511, | |
| "grad_norm": 1.0244650840759277, | |
| "learning_rate": 9.657227098189797e-05, | |
| "loss": 0.39216903686523436, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.11377613776137761, | |
| "grad_norm": 1.0149824619293213, | |
| "learning_rate": 9.653798683488756e-05, | |
| "loss": 0.3916678237915039, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.11480114801148011, | |
| "grad_norm": 0.9391900897026062, | |
| "learning_rate": 9.650370268787713e-05, | |
| "loss": 0.39551502227783203, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.11582615826158262, | |
| "grad_norm": 0.9509477615356445, | |
| "learning_rate": 9.64694185408667e-05, | |
| "loss": 0.3881892013549805, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.11685116851168512, | |
| "grad_norm": 0.9446685910224915, | |
| "learning_rate": 9.643513439385628e-05, | |
| "loss": 0.3790993118286133, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.11787617876178762, | |
| "grad_norm": 0.9401887655258179, | |
| "learning_rate": 9.640085024684587e-05, | |
| "loss": 0.39570098876953125, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.11890118901189012, | |
| "grad_norm": 0.947109580039978, | |
| "learning_rate": 9.636656609983544e-05, | |
| "loss": 0.39298431396484373, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.11992619926199262, | |
| "grad_norm": 0.9778124690055847, | |
| "learning_rate": 9.633228195282502e-05, | |
| "loss": 0.38612052917480466, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.12095120951209512, | |
| "grad_norm": 0.9742032289505005, | |
| "learning_rate": 9.629799780581459e-05, | |
| "loss": 0.38578262329101565, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.12197621976219762, | |
| "grad_norm": 0.920020341873169, | |
| "learning_rate": 9.626371365880418e-05, | |
| "loss": 0.37897506713867185, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.12300123001230012, | |
| "grad_norm": 1.0522706508636475, | |
| "learning_rate": 9.622942951179376e-05, | |
| "loss": 0.38796062469482423, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.12402624026240262, | |
| "grad_norm": 0.9806820154190063, | |
| "learning_rate": 9.619514536478332e-05, | |
| "loss": 0.3841505432128906, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.12505125051250512, | |
| "grad_norm": 1.051397681236267, | |
| "learning_rate": 9.61608612177729e-05, | |
| "loss": 0.3859896469116211, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.12607626076260764, | |
| "grad_norm": 1.0001674890518188, | |
| "learning_rate": 9.612657707076248e-05, | |
| "loss": 0.38761005401611326, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.12710127101271013, | |
| "grad_norm": 0.9256671667098999, | |
| "learning_rate": 9.609229292375205e-05, | |
| "loss": 0.39095249176025393, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.12812628126281264, | |
| "grad_norm": 0.9498622417449951, | |
| "learning_rate": 9.605800877674164e-05, | |
| "loss": 0.3878898620605469, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.12915129151291513, | |
| "grad_norm": 0.964492678642273, | |
| "learning_rate": 9.602372462973122e-05, | |
| "loss": 0.3858418655395508, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.13017630176301764, | |
| "grad_norm": 0.9889938831329346, | |
| "learning_rate": 9.598944048272079e-05, | |
| "loss": 0.38567405700683594, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.13120131201312013, | |
| "grad_norm": 1.0020108222961426, | |
| "learning_rate": 9.595515633571037e-05, | |
| "loss": 0.3937076187133789, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.13222632226322265, | |
| "grad_norm": 0.9936574697494507, | |
| "learning_rate": 9.592087218869996e-05, | |
| "loss": 0.3861484146118164, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.13325133251332513, | |
| "grad_norm": 0.998080313205719, | |
| "learning_rate": 9.588658804168953e-05, | |
| "loss": 0.3910287094116211, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.13427634276342765, | |
| "grad_norm": 1.040822148323059, | |
| "learning_rate": 9.58523038946791e-05, | |
| "loss": 0.38510765075683595, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.13530135301353013, | |
| "grad_norm": 1.0341635942459106, | |
| "learning_rate": 9.581801974766868e-05, | |
| "loss": 0.3902041244506836, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.13632636326363265, | |
| "grad_norm": 0.9677271842956543, | |
| "learning_rate": 9.578373560065827e-05, | |
| "loss": 0.3900999450683594, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.13735137351373514, | |
| "grad_norm": 0.9766192436218262, | |
| "learning_rate": 9.574945145364784e-05, | |
| "loss": 0.3803682708740234, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.13837638376383765, | |
| "grad_norm": 0.90891432762146, | |
| "learning_rate": 9.57151673066374e-05, | |
| "loss": 0.38466361999511717, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.13940139401394014, | |
| "grad_norm": 1.044262170791626, | |
| "learning_rate": 9.568088315962699e-05, | |
| "loss": 0.3881848907470703, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.14042640426404265, | |
| "grad_norm": 0.8678585290908813, | |
| "learning_rate": 9.564659901261657e-05, | |
| "loss": 0.38272193908691404, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.14145141451414514, | |
| "grad_norm": 0.9656084775924683, | |
| "learning_rate": 9.561231486560614e-05, | |
| "loss": 0.38340423583984373, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.14247642476424766, | |
| "grad_norm": 1.095802903175354, | |
| "learning_rate": 9.557803071859573e-05, | |
| "loss": 0.39173069000244143, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.14350143501435014, | |
| "grad_norm": 0.9511718153953552, | |
| "learning_rate": 9.55437465715853e-05, | |
| "loss": 0.38641986846923826, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.14452644526445266, | |
| "grad_norm": 1.0598710775375366, | |
| "learning_rate": 9.550946242457488e-05, | |
| "loss": 0.38396541595458983, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.14555145551455514, | |
| "grad_norm": 1.023289680480957, | |
| "learning_rate": 9.547517827756445e-05, | |
| "loss": 0.38503433227539063, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.14657646576465766, | |
| "grad_norm": 0.8889958262443542, | |
| "learning_rate": 9.544089413055404e-05, | |
| "loss": 0.38462825775146486, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.14760147601476015, | |
| "grad_norm": 0.907673716545105, | |
| "learning_rate": 9.540660998354362e-05, | |
| "loss": 0.3838591003417969, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.14862648626486266, | |
| "grad_norm": 1.0963199138641357, | |
| "learning_rate": 9.537232583653319e-05, | |
| "loss": 0.38962028503417967, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.14965149651496515, | |
| "grad_norm": 1.0007156133651733, | |
| "learning_rate": 9.533804168952277e-05, | |
| "loss": 0.3871998977661133, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.15067650676506766, | |
| "grad_norm": 0.9626657962799072, | |
| "learning_rate": 9.530375754251235e-05, | |
| "loss": 0.3868283462524414, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.15170151701517015, | |
| "grad_norm": 1.041344165802002, | |
| "learning_rate": 9.526947339550193e-05, | |
| "loss": 0.38434059143066407, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.15272652726527267, | |
| "grad_norm": 0.9025008082389832, | |
| "learning_rate": 9.52351892484915e-05, | |
| "loss": 0.38416252136230467, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.15375153751537515, | |
| "grad_norm": 1.0520020723342896, | |
| "learning_rate": 9.520090510148108e-05, | |
| "loss": 0.3882419586181641, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.15477654776547767, | |
| "grad_norm": 1.0362555980682373, | |
| "learning_rate": 9.516662095447065e-05, | |
| "loss": 0.38122001647949216, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.15580155801558015, | |
| "grad_norm": 0.9644442200660706, | |
| "learning_rate": 9.513233680746023e-05, | |
| "loss": 0.38314281463623046, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.15682656826568267, | |
| "grad_norm": 0.8997851610183716, | |
| "learning_rate": 9.509805266044982e-05, | |
| "loss": 0.38658527374267576, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.15785157851578516, | |
| "grad_norm": 1.0372231006622314, | |
| "learning_rate": 9.506376851343939e-05, | |
| "loss": 0.38490631103515627, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.15887658876588767, | |
| "grad_norm": 0.9239746332168579, | |
| "learning_rate": 9.502948436642897e-05, | |
| "loss": 0.3856174850463867, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.15990159901599016, | |
| "grad_norm": 0.9766268134117126, | |
| "learning_rate": 9.499520021941854e-05, | |
| "loss": 0.37917984008789063, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.16092660926609267, | |
| "grad_norm": 0.9248613119125366, | |
| "learning_rate": 9.496091607240813e-05, | |
| "loss": 0.385213623046875, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.16195161951619516, | |
| "grad_norm": 1.0540366172790527, | |
| "learning_rate": 9.49266319253977e-05, | |
| "loss": 0.39073848724365234, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.16297662976629768, | |
| "grad_norm": 1.0271457433700562, | |
| "learning_rate": 9.489234777838728e-05, | |
| "loss": 0.38568016052246096, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.16400164001640016, | |
| "grad_norm": 0.8689510226249695, | |
| "learning_rate": 9.485806363137685e-05, | |
| "loss": 0.38491676330566404, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.16502665026650268, | |
| "grad_norm": 0.9900168776512146, | |
| "learning_rate": 9.482377948436644e-05, | |
| "loss": 0.3833469009399414, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.16605166051660517, | |
| "grad_norm": 0.9545993804931641, | |
| "learning_rate": 9.478949533735602e-05, | |
| "loss": 0.37658924102783203, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.16707667076670768, | |
| "grad_norm": 0.9692700505256653, | |
| "learning_rate": 9.475521119034559e-05, | |
| "loss": 0.38341827392578126, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.16810168101681017, | |
| "grad_norm": 0.9968641996383667, | |
| "learning_rate": 9.472092704333516e-05, | |
| "loss": 0.3886453247070312, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.16912669126691268, | |
| "grad_norm": 0.9806464314460754, | |
| "learning_rate": 9.468664289632474e-05, | |
| "loss": 0.3858697509765625, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.17015170151701517, | |
| "grad_norm": 1.010970115661621, | |
| "learning_rate": 9.465235874931431e-05, | |
| "loss": 0.3842259216308594, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.17117671176711768, | |
| "grad_norm": 0.8758835196495056, | |
| "learning_rate": 9.46180746023039e-05, | |
| "loss": 0.3848011016845703, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.17220172201722017, | |
| "grad_norm": 0.971757709980011, | |
| "learning_rate": 9.458379045529348e-05, | |
| "loss": 0.38819068908691406, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.1732267322673227, | |
| "grad_norm": 1.021390676498413, | |
| "learning_rate": 9.454950630828305e-05, | |
| "loss": 0.383485107421875, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.17425174251742517, | |
| "grad_norm": 0.9065835475921631, | |
| "learning_rate": 9.451522216127263e-05, | |
| "loss": 0.38290172576904297, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.1752767527675277, | |
| "grad_norm": 0.9806657433509827, | |
| "learning_rate": 9.448093801426221e-05, | |
| "loss": 0.38694095611572266, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.17630176301763018, | |
| "grad_norm": 0.9851366281509399, | |
| "learning_rate": 9.444665386725179e-05, | |
| "loss": 0.3823180389404297, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.1773267732677327, | |
| "grad_norm": 0.9909108877182007, | |
| "learning_rate": 9.441236972024136e-05, | |
| "loss": 0.38766963958740236, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.17835178351783518, | |
| "grad_norm": 0.9420474767684937, | |
| "learning_rate": 9.437808557323094e-05, | |
| "loss": 0.3867580795288086, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.1793767937679377, | |
| "grad_norm": 0.9109618663787842, | |
| "learning_rate": 9.434380142622053e-05, | |
| "loss": 0.3817281723022461, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.18040180401804018, | |
| "grad_norm": 0.8160777688026428, | |
| "learning_rate": 9.43095172792101e-05, | |
| "loss": 0.38428955078125, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.1814268142681427, | |
| "grad_norm": 1.074156641960144, | |
| "learning_rate": 9.427523313219968e-05, | |
| "loss": 0.3834354782104492, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.18245182451824518, | |
| "grad_norm": 1.0000030994415283, | |
| "learning_rate": 9.424094898518925e-05, | |
| "loss": 0.3791364669799805, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.1834768347683477, | |
| "grad_norm": 0.8356959819793701, | |
| "learning_rate": 9.420666483817883e-05, | |
| "loss": 0.37940494537353514, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.18450184501845018, | |
| "grad_norm": 0.9983715415000916, | |
| "learning_rate": 9.41723806911684e-05, | |
| "loss": 0.3863615417480469, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.1855268552685527, | |
| "grad_norm": 0.9836549758911133, | |
| "learning_rate": 9.413809654415799e-05, | |
| "loss": 0.382811393737793, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.18655186551865519, | |
| "grad_norm": 0.878430187702179, | |
| "learning_rate": 9.410381239714756e-05, | |
| "loss": 0.38379634857177736, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.1875768757687577, | |
| "grad_norm": 0.9808958172798157, | |
| "learning_rate": 9.406952825013714e-05, | |
| "loss": 0.3824180221557617, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.1886018860188602, | |
| "grad_norm": 0.8256483674049377, | |
| "learning_rate": 9.403524410312671e-05, | |
| "loss": 0.3791842269897461, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.1896268962689627, | |
| "grad_norm": 0.9319470524787903, | |
| "learning_rate": 9.40009599561163e-05, | |
| "loss": 0.3782553482055664, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.1906519065190652, | |
| "grad_norm": 0.9519809484481812, | |
| "learning_rate": 9.396667580910588e-05, | |
| "loss": 0.38426536560058594, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.1916769167691677, | |
| "grad_norm": 0.9517083168029785, | |
| "learning_rate": 9.393239166209545e-05, | |
| "loss": 0.3869959259033203, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.1927019270192702, | |
| "grad_norm": 0.8581302165985107, | |
| "learning_rate": 9.389810751508504e-05, | |
| "loss": 0.3790604782104492, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.1937269372693727, | |
| "grad_norm": 0.9409393072128296, | |
| "learning_rate": 9.386382336807461e-05, | |
| "loss": 0.3730373001098633, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.1947519475194752, | |
| "grad_norm": 0.9260959625244141, | |
| "learning_rate": 9.382953922106419e-05, | |
| "loss": 0.38074058532714844, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.1957769577695777, | |
| "grad_norm": 0.8830330967903137, | |
| "learning_rate": 9.379525507405376e-05, | |
| "loss": 0.37637657165527344, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.1968019680196802, | |
| "grad_norm": 0.9540618658065796, | |
| "learning_rate": 9.376097092704334e-05, | |
| "loss": 0.3863861083984375, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.1978269782697827, | |
| "grad_norm": 0.873328447341919, | |
| "learning_rate": 9.372668678003291e-05, | |
| "loss": 0.38958278656005857, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.1988519885198852, | |
| "grad_norm": 0.8584098219871521, | |
| "learning_rate": 9.369240263302249e-05, | |
| "loss": 0.379111213684082, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.1998769987699877, | |
| "grad_norm": 0.872678279876709, | |
| "learning_rate": 9.365811848601208e-05, | |
| "loss": 0.37820552825927733, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.2009020090200902, | |
| "grad_norm": 0.8778783082962036, | |
| "learning_rate": 9.362383433900165e-05, | |
| "loss": 0.3825270462036133, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.20192701927019271, | |
| "grad_norm": 0.8765313029289246, | |
| "learning_rate": 9.358955019199122e-05, | |
| "loss": 0.38493602752685546, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.2029520295202952, | |
| "grad_norm": 1.020390510559082, | |
| "learning_rate": 9.35552660449808e-05, | |
| "loss": 0.3772980499267578, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.20397703977039772, | |
| "grad_norm": 0.975033164024353, | |
| "learning_rate": 9.352098189797039e-05, | |
| "loss": 0.3829964065551758, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.2050020500205002, | |
| "grad_norm": 0.9375325441360474, | |
| "learning_rate": 9.348669775095996e-05, | |
| "loss": 0.378480110168457, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.2050020500205002, | |
| "eval_loss": 0.46537917852401733, | |
| "eval_runtime": 4.4505, | |
| "eval_samples_per_second": 5617.307, | |
| "eval_steps_per_second": 5.617, | |
| "eval_teacher_cosine_loss": 0.288166344165802, | |
| "eval_teacher_cosine_mean": 0.711833655834198, | |
| "eval_teacher_cosine_p10": 0.6314711570739746, | |
| "eval_teacher_cosine_p50": 0.720482349395752, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.20602706027060272, | |
| "grad_norm": 0.8948764801025391, | |
| "learning_rate": 9.345241360394954e-05, | |
| "loss": 0.37587287902832034, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.2070520705207052, | |
| "grad_norm": 0.9858198761940002, | |
| "learning_rate": 9.341812945693913e-05, | |
| "loss": 0.3785189437866211, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.20807708077080772, | |
| "grad_norm": 0.9548747539520264, | |
| "learning_rate": 9.33838453099287e-05, | |
| "loss": 0.3788298034667969, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.2091020910209102, | |
| "grad_norm": 0.9604178071022034, | |
| "learning_rate": 9.334956116291827e-05, | |
| "loss": 0.3860102844238281, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.21012710127101272, | |
| "grad_norm": 0.9090835452079773, | |
| "learning_rate": 9.331527701590785e-05, | |
| "loss": 0.3831613540649414, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.2111521115211152, | |
| "grad_norm": 1.008748173713684, | |
| "learning_rate": 9.328099286889742e-05, | |
| "loss": 0.37774810791015623, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.21217712177121772, | |
| "grad_norm": 0.9406893253326416, | |
| "learning_rate": 9.3246708721887e-05, | |
| "loss": 0.38153438568115233, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.2132021320213202, | |
| "grad_norm": 0.8962439298629761, | |
| "learning_rate": 9.321242457487657e-05, | |
| "loss": 0.3762447738647461, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.21422714227142273, | |
| "grad_norm": 0.8759261965751648, | |
| "learning_rate": 9.317814042786616e-05, | |
| "loss": 0.3887854766845703, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.21525215252152521, | |
| "grad_norm": 0.8931547999382019, | |
| "learning_rate": 9.314385628085574e-05, | |
| "loss": 0.3788555908203125, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.21627716277162773, | |
| "grad_norm": 0.9626860022544861, | |
| "learning_rate": 9.310957213384531e-05, | |
| "loss": 0.3834539794921875, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.21730217302173022, | |
| "grad_norm": 0.9040349721908569, | |
| "learning_rate": 9.307528798683489e-05, | |
| "loss": 0.3825030136108398, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.21832718327183273, | |
| "grad_norm": 0.9993950128555298, | |
| "learning_rate": 9.304100383982447e-05, | |
| "loss": 0.37788055419921873, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.21935219352193522, | |
| "grad_norm": 0.9242931008338928, | |
| "learning_rate": 9.300671969281405e-05, | |
| "loss": 0.37859432220458983, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.22037720377203773, | |
| "grad_norm": 0.959726870059967, | |
| "learning_rate": 9.297243554580362e-05, | |
| "loss": 0.3868673706054688, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.22140221402214022, | |
| "grad_norm": 0.9452869892120361, | |
| "learning_rate": 9.293815139879321e-05, | |
| "loss": 0.37610702514648436, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.22242722427224274, | |
| "grad_norm": 0.8363068103790283, | |
| "learning_rate": 9.290386725178279e-05, | |
| "loss": 0.3854954528808594, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.22345223452234522, | |
| "grad_norm": 0.9416174292564392, | |
| "learning_rate": 9.286958310477236e-05, | |
| "loss": 0.3781816864013672, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.22447724477244774, | |
| "grad_norm": 0.9125226736068726, | |
| "learning_rate": 9.283529895776194e-05, | |
| "loss": 0.38271350860595704, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.22550225502255022, | |
| "grad_norm": 0.9020611047744751, | |
| "learning_rate": 9.280101481075151e-05, | |
| "loss": 0.3829726028442383, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.22652726527265274, | |
| "grad_norm": 0.8089579343795776, | |
| "learning_rate": 9.276673066374109e-05, | |
| "loss": 0.3817304992675781, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.22755227552275523, | |
| "grad_norm": 1.0000059604644775, | |
| "learning_rate": 9.273244651673066e-05, | |
| "loss": 0.3752499771118164, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.22857728577285774, | |
| "grad_norm": 0.8663347363471985, | |
| "learning_rate": 9.269816236972025e-05, | |
| "loss": 0.3772423553466797, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.22960229602296023, | |
| "grad_norm": 1.007620096206665, | |
| "learning_rate": 9.266387822270982e-05, | |
| "loss": 0.378299674987793, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.23062730627306274, | |
| "grad_norm": 0.9828184843063354, | |
| "learning_rate": 9.26295940756994e-05, | |
| "loss": 0.3822395706176758, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.23165231652316523, | |
| "grad_norm": 0.8999512195587158, | |
| "learning_rate": 9.259530992868897e-05, | |
| "loss": 0.37898773193359375, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.23267732677326775, | |
| "grad_norm": 0.8549731373786926, | |
| "learning_rate": 9.256102578167856e-05, | |
| "loss": 0.38248577117919924, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 0.23370233702337023, | |
| "grad_norm": 0.8490684628486633, | |
| "learning_rate": 9.252674163466814e-05, | |
| "loss": 0.3778285217285156, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.23472734727347275, | |
| "grad_norm": 1.0765396356582642, | |
| "learning_rate": 9.249245748765771e-05, | |
| "loss": 0.3751084899902344, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 0.23575235752357523, | |
| "grad_norm": 0.8854128122329712, | |
| "learning_rate": 9.24581733406473e-05, | |
| "loss": 0.3815226364135742, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.23677736777367775, | |
| "grad_norm": 0.9462901949882507, | |
| "learning_rate": 9.242388919363687e-05, | |
| "loss": 0.376729850769043, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 0.23780237802378024, | |
| "grad_norm": 0.9277807474136353, | |
| "learning_rate": 9.238960504662645e-05, | |
| "loss": 0.3751403427124023, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.23882738827388275, | |
| "grad_norm": 0.8764863610267639, | |
| "learning_rate": 9.235532089961602e-05, | |
| "loss": 0.3718427658081055, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 0.23985239852398524, | |
| "grad_norm": 0.951400101184845, | |
| "learning_rate": 9.23210367526056e-05, | |
| "loss": 0.37830867767333987, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.24087740877408775, | |
| "grad_norm": 1.0767115354537964, | |
| "learning_rate": 9.228675260559517e-05, | |
| "loss": 0.3772780990600586, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 0.24190241902419024, | |
| "grad_norm": 0.9252443313598633, | |
| "learning_rate": 9.225246845858475e-05, | |
| "loss": 0.37594383239746093, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.24292742927429276, | |
| "grad_norm": 0.9442302584648132, | |
| "learning_rate": 9.221818431157433e-05, | |
| "loss": 0.37919326782226564, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 0.24395243952439524, | |
| "grad_norm": 0.9696224927902222, | |
| "learning_rate": 9.218390016456391e-05, | |
| "loss": 0.3784609603881836, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.24497744977449776, | |
| "grad_norm": 1.030659794807434, | |
| "learning_rate": 9.214961601755348e-05, | |
| "loss": 0.38156280517578123, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 0.24600246002460024, | |
| "grad_norm": 0.974122941493988, | |
| "learning_rate": 9.211533187054306e-05, | |
| "loss": 0.3759873199462891, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.24702747027470276, | |
| "grad_norm": 0.8502260446548462, | |
| "learning_rate": 9.208104772353265e-05, | |
| "loss": 0.3790750503540039, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 0.24805248052480525, | |
| "grad_norm": 0.9666852951049805, | |
| "learning_rate": 9.204676357652222e-05, | |
| "loss": 0.3703215789794922, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.24907749077490776, | |
| "grad_norm": 0.8803427815437317, | |
| "learning_rate": 9.20124794295118e-05, | |
| "loss": 0.37341415405273437, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 0.25010250102501025, | |
| "grad_norm": 1.0092825889587402, | |
| "learning_rate": 9.197819528250138e-05, | |
| "loss": 0.37458995819091795, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.25112751127511274, | |
| "grad_norm": 0.9279468655586243, | |
| "learning_rate": 9.194391113549096e-05, | |
| "loss": 0.3745867156982422, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 0.2521525215252153, | |
| "grad_norm": 0.856102705001831, | |
| "learning_rate": 9.190962698848053e-05, | |
| "loss": 0.3768229675292969, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.25317753177531777, | |
| "grad_norm": 0.9179368019104004, | |
| "learning_rate": 9.187534284147011e-05, | |
| "loss": 0.3743790817260742, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 0.25420254202542025, | |
| "grad_norm": 1.0300084352493286, | |
| "learning_rate": 9.184105869445968e-05, | |
| "loss": 0.37301589965820314, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.25522755227552274, | |
| "grad_norm": 0.9511299729347229, | |
| "learning_rate": 9.180677454744926e-05, | |
| "loss": 0.3752445602416992, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 0.2562525625256253, | |
| "grad_norm": 0.9306531548500061, | |
| "learning_rate": 9.177249040043883e-05, | |
| "loss": 0.37558921813964846, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.25727757277572777, | |
| "grad_norm": 0.9262721538543701, | |
| "learning_rate": 9.173820625342842e-05, | |
| "loss": 0.37717815399169924, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 0.25830258302583026, | |
| "grad_norm": 0.905663788318634, | |
| "learning_rate": 9.1703922106418e-05, | |
| "loss": 0.37981117248535157, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 0.25932759327593274, | |
| "grad_norm": 0.9691590070724487, | |
| "learning_rate": 9.166963795940757e-05, | |
| "loss": 0.37886699676513674, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 0.2603526035260353, | |
| "grad_norm": 0.8920459747314453, | |
| "learning_rate": 9.163535381239714e-05, | |
| "loss": 0.3829997253417969, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 0.2613776137761378, | |
| "grad_norm": 0.938231885433197, | |
| "learning_rate": 9.160106966538673e-05, | |
| "loss": 0.3744136047363281, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 0.26240262402624026, | |
| "grad_norm": 0.8784525990486145, | |
| "learning_rate": 9.156678551837631e-05, | |
| "loss": 0.3803242111206055, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 0.26342763427634275, | |
| "grad_norm": 0.9013773202896118, | |
| "learning_rate": 9.153250137136588e-05, | |
| "loss": 0.3807582473754883, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 0.2644526445264453, | |
| "grad_norm": 0.9393748641014099, | |
| "learning_rate": 9.149821722435547e-05, | |
| "loss": 0.36834522247314455, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 0.2654776547765478, | |
| "grad_norm": 0.9058448672294617, | |
| "learning_rate": 9.146393307734505e-05, | |
| "loss": 0.376998176574707, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 0.26650266502665027, | |
| "grad_norm": 0.7872834801673889, | |
| "learning_rate": 9.142964893033462e-05, | |
| "loss": 0.3794910430908203, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.26752767527675275, | |
| "grad_norm": 0.9153306484222412, | |
| "learning_rate": 9.13953647833242e-05, | |
| "loss": 0.3744646453857422, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 0.2685526855268553, | |
| "grad_norm": 0.896645188331604, | |
| "learning_rate": 9.136108063631378e-05, | |
| "loss": 0.3749618911743164, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 0.2695776957769578, | |
| "grad_norm": 0.9123072028160095, | |
| "learning_rate": 9.132679648930334e-05, | |
| "loss": 0.377745246887207, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 0.27060270602706027, | |
| "grad_norm": 0.9632300138473511, | |
| "learning_rate": 9.129251234229292e-05, | |
| "loss": 0.3756327819824219, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.27162771627716276, | |
| "grad_norm": 0.8565627932548523, | |
| "learning_rate": 9.125822819528251e-05, | |
| "loss": 0.38645397186279296, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 0.2726527265272653, | |
| "grad_norm": 0.911180853843689, | |
| "learning_rate": 9.122394404827208e-05, | |
| "loss": 0.37197341918945315, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 0.2736777367773678, | |
| "grad_norm": 0.9202716946601868, | |
| "learning_rate": 9.118965990126166e-05, | |
| "loss": 0.37791946411132815, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 0.2747027470274703, | |
| "grad_norm": 0.8607870936393738, | |
| "learning_rate": 9.115537575425123e-05, | |
| "loss": 0.3766735076904297, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 0.27572775727757276, | |
| "grad_norm": 0.8086125254631042, | |
| "learning_rate": 9.112109160724082e-05, | |
| "loss": 0.37189990997314454, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 0.2767527675276753, | |
| "grad_norm": 0.8633895516395569, | |
| "learning_rate": 9.10868074602304e-05, | |
| "loss": 0.36816112518310545, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 0.9373879432678223, | |
| "learning_rate": 9.105252331321997e-05, | |
| "loss": 0.3766764068603516, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 0.2788027880278803, | |
| "grad_norm": 0.9496821165084839, | |
| "learning_rate": 9.101823916620956e-05, | |
| "loss": 0.3716035079956055, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 0.27982779827798276, | |
| "grad_norm": 0.938170313835144, | |
| "learning_rate": 9.098395501919913e-05, | |
| "loss": 0.37355762481689453, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 0.2808528085280853, | |
| "grad_norm": 0.9260206818580627, | |
| "learning_rate": 9.09496708721887e-05, | |
| "loss": 0.3742988586425781, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 0.2818778187781878, | |
| "grad_norm": 0.9371470808982849, | |
| "learning_rate": 9.091538672517828e-05, | |
| "loss": 0.3769746017456055, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 0.2829028290282903, | |
| "grad_norm": 0.8824672102928162, | |
| "learning_rate": 9.088110257816787e-05, | |
| "loss": 0.37410057067871094, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 0.28392783927839277, | |
| "grad_norm": 0.9268654584884644, | |
| "learning_rate": 9.084681843115743e-05, | |
| "loss": 0.3803014373779297, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 0.2849528495284953, | |
| "grad_norm": 0.833306610584259, | |
| "learning_rate": 9.0812534284147e-05, | |
| "loss": 0.3746840667724609, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 0.2859778597785978, | |
| "grad_norm": 0.9149842262268066, | |
| "learning_rate": 9.07782501371366e-05, | |
| "loss": 0.3733023071289063, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 0.2870028700287003, | |
| "grad_norm": 0.8075073957443237, | |
| "learning_rate": 9.074396599012617e-05, | |
| "loss": 0.37573966979980467, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.2880278802788028, | |
| "grad_norm": 0.9132227897644043, | |
| "learning_rate": 9.070968184311574e-05, | |
| "loss": 0.3778294372558594, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 0.2890528905289053, | |
| "grad_norm": 0.8698980212211609, | |
| "learning_rate": 9.067539769610532e-05, | |
| "loss": 0.37292243957519533, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 0.2900779007790078, | |
| "grad_norm": 0.9611607789993286, | |
| "learning_rate": 9.06411135490949e-05, | |
| "loss": 0.3745908737182617, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 0.2911029110291103, | |
| "grad_norm": 0.9108705520629883, | |
| "learning_rate": 9.060682940208448e-05, | |
| "loss": 0.37185867309570314, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 0.2921279212792128, | |
| "grad_norm": 0.9152095317840576, | |
| "learning_rate": 9.057254525507406e-05, | |
| "loss": 0.3720784378051758, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 0.2931529315293153, | |
| "grad_norm": 0.9379122257232666, | |
| "learning_rate": 9.053826110806364e-05, | |
| "loss": 0.3681333541870117, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 0.2941779417794178, | |
| "grad_norm": 0.9308175444602966, | |
| "learning_rate": 9.050397696105322e-05, | |
| "loss": 0.37542362213134767, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 0.2952029520295203, | |
| "grad_norm": 0.9414443373680115, | |
| "learning_rate": 9.046969281404279e-05, | |
| "loss": 0.3708903121948242, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 0.2962279622796228, | |
| "grad_norm": 0.9574779272079468, | |
| "learning_rate": 9.043540866703237e-05, | |
| "loss": 0.3728959655761719, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 0.2972529725297253, | |
| "grad_norm": 0.8834381699562073, | |
| "learning_rate": 9.040112452002196e-05, | |
| "loss": 0.3803950119018555, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.2982779827798278, | |
| "grad_norm": 0.8980011940002441, | |
| "learning_rate": 9.036684037301152e-05, | |
| "loss": 0.38026893615722657, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 0.2993029930299303, | |
| "grad_norm": 0.8272210955619812, | |
| "learning_rate": 9.033255622600109e-05, | |
| "loss": 0.3767722702026367, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 0.3003280032800328, | |
| "grad_norm": 0.9115729928016663, | |
| "learning_rate": 9.029827207899068e-05, | |
| "loss": 0.37336029052734376, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 0.3013530135301353, | |
| "grad_norm": 0.9003026485443115, | |
| "learning_rate": 9.026398793198025e-05, | |
| "loss": 0.37091796875, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 0.3023780237802378, | |
| "grad_norm": 0.9348158240318298, | |
| "learning_rate": 9.022970378496983e-05, | |
| "loss": 0.3767798233032227, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 0.3034030340303403, | |
| "grad_norm": 0.9327846765518188, | |
| "learning_rate": 9.01954196379594e-05, | |
| "loss": 0.3753174209594727, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 0.3044280442804428, | |
| "grad_norm": 0.7624741196632385, | |
| "learning_rate": 9.016113549094899e-05, | |
| "loss": 0.3686408233642578, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 0.30545305453054533, | |
| "grad_norm": 0.9010864496231079, | |
| "learning_rate": 9.012685134393857e-05, | |
| "loss": 0.36874263763427734, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 0.3064780647806478, | |
| "grad_norm": 0.9010471701622009, | |
| "learning_rate": 9.009256719692814e-05, | |
| "loss": 0.3715047836303711, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 0.3075030750307503, | |
| "grad_norm": 0.8821438550949097, | |
| "learning_rate": 9.005828304991773e-05, | |
| "loss": 0.3756497192382813, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.3075030750307503, | |
| "eval_loss": 0.45619136095046997, | |
| "eval_runtime": 4.3976, | |
| "eval_samples_per_second": 5684.904, | |
| "eval_steps_per_second": 5.685, | |
| "eval_teacher_cosine_loss": 0.2815589904785156, | |
| "eval_teacher_cosine_mean": 0.7184410095214844, | |
| "eval_teacher_cosine_p10": 0.6389575004577637, | |
| "eval_teacher_cosine_p50": 0.7274641990661621, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.3085280852808528, | |
| "grad_norm": 0.9257699251174927, | |
| "learning_rate": 9.00239989029073e-05, | |
| "loss": 0.3711713790893555, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 0.30955309553095534, | |
| "grad_norm": 0.8627203106880188, | |
| "learning_rate": 8.998971475589688e-05, | |
| "loss": 0.36991249084472655, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 0.3105781057810578, | |
| "grad_norm": 0.8617995381355286, | |
| "learning_rate": 8.995543060888645e-05, | |
| "loss": 0.3682516479492188, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 0.3116031160311603, | |
| "grad_norm": 0.8817408680915833, | |
| "learning_rate": 8.992114646187604e-05, | |
| "loss": 0.37412437438964846, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 0.3126281262812628, | |
| "grad_norm": 0.8762063384056091, | |
| "learning_rate": 8.98868623148656e-05, | |
| "loss": 0.37474369049072265, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 0.31365313653136534, | |
| "grad_norm": 0.9961568713188171, | |
| "learning_rate": 8.985257816785518e-05, | |
| "loss": 0.3744321060180664, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 0.3146781467814678, | |
| "grad_norm": 0.9071103930473328, | |
| "learning_rate": 8.981829402084477e-05, | |
| "loss": 0.3736297607421875, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 0.3157031570315703, | |
| "grad_norm": 0.8751444816589355, | |
| "learning_rate": 8.978400987383434e-05, | |
| "loss": 0.3731893157958984, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 0.3167281672816728, | |
| "grad_norm": 0.8277934789657593, | |
| "learning_rate": 8.974972572682392e-05, | |
| "loss": 0.377271728515625, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 0.31775317753177534, | |
| "grad_norm": 0.887535572052002, | |
| "learning_rate": 8.971544157981349e-05, | |
| "loss": 0.3714146041870117, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.31877818778187783, | |
| "grad_norm": 0.87473464012146, | |
| "learning_rate": 8.968115743280308e-05, | |
| "loss": 0.3753161239624023, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 0.3198031980319803, | |
| "grad_norm": 0.9824237823486328, | |
| "learning_rate": 8.964687328579265e-05, | |
| "loss": 0.3717261505126953, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 0.3208282082820828, | |
| "grad_norm": 0.8578912019729614, | |
| "learning_rate": 8.961258913878223e-05, | |
| "loss": 0.37559310913085936, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 0.32185321853218535, | |
| "grad_norm": 0.9741605520248413, | |
| "learning_rate": 8.957830499177182e-05, | |
| "loss": 0.3719668960571289, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 0.32287822878228783, | |
| "grad_norm": 0.7923543453216553, | |
| "learning_rate": 8.954402084476139e-05, | |
| "loss": 0.3701831817626953, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 0.3239032390323903, | |
| "grad_norm": 0.8646412491798401, | |
| "learning_rate": 8.950973669775097e-05, | |
| "loss": 0.3755689239501953, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 0.3249282492824928, | |
| "grad_norm": 0.9215549826622009, | |
| "learning_rate": 8.947545255074054e-05, | |
| "loss": 0.37338077545166015, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 0.32595325953259535, | |
| "grad_norm": 0.9665416479110718, | |
| "learning_rate": 8.944116840373013e-05, | |
| "loss": 0.3735879135131836, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 0.32697826978269784, | |
| "grad_norm": 0.8452497720718384, | |
| "learning_rate": 8.940688425671969e-05, | |
| "loss": 0.37112060546875, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 0.3280032800328003, | |
| "grad_norm": 0.8731474876403809, | |
| "learning_rate": 8.937260010970926e-05, | |
| "loss": 0.3701690673828125, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.3290282902829028, | |
| "grad_norm": 0.8352526426315308, | |
| "learning_rate": 8.933831596269885e-05, | |
| "loss": 0.3731770706176758, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 0.33005330053300536, | |
| "grad_norm": 0.8384972810745239, | |
| "learning_rate": 8.930403181568843e-05, | |
| "loss": 0.3679676055908203, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 0.33107831078310784, | |
| "grad_norm": 0.8718621134757996, | |
| "learning_rate": 8.9269747668678e-05, | |
| "loss": 0.3753276062011719, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 0.33210332103321033, | |
| "grad_norm": 0.8473393321037292, | |
| "learning_rate": 8.923546352166759e-05, | |
| "loss": 0.372968864440918, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 0.3331283312833128, | |
| "grad_norm": 0.9392423629760742, | |
| "learning_rate": 8.920117937465717e-05, | |
| "loss": 0.37619026184082033, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 0.33415334153341536, | |
| "grad_norm": 0.8857057094573975, | |
| "learning_rate": 8.916689522764674e-05, | |
| "loss": 0.3712038040161133, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 0.33517835178351785, | |
| "grad_norm": 0.7681214809417725, | |
| "learning_rate": 8.913261108063631e-05, | |
| "loss": 0.36811851501464843, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 0.33620336203362033, | |
| "grad_norm": 0.9219163656234741, | |
| "learning_rate": 8.90983269336259e-05, | |
| "loss": 0.36732242584228514, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 0.3372283722837228, | |
| "grad_norm": 0.8490789532661438, | |
| "learning_rate": 8.906404278661548e-05, | |
| "loss": 0.3747453308105469, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 0.33825338253382536, | |
| "grad_norm": 0.8938300609588623, | |
| "learning_rate": 8.902975863960505e-05, | |
| "loss": 0.3666362762451172, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.33927839278392785, | |
| "grad_norm": 0.9085167050361633, | |
| "learning_rate": 8.899547449259463e-05, | |
| "loss": 0.373651008605957, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 0.34030340303403034, | |
| "grad_norm": 0.8874534368515015, | |
| "learning_rate": 8.896119034558422e-05, | |
| "loss": 0.37278167724609373, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 0.3413284132841328, | |
| "grad_norm": 0.8592978119850159, | |
| "learning_rate": 8.892690619857378e-05, | |
| "loss": 0.37098949432373046, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 0.34235342353423537, | |
| "grad_norm": 0.83613520860672, | |
| "learning_rate": 8.889262205156335e-05, | |
| "loss": 0.3769094085693359, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 0.34337843378433786, | |
| "grad_norm": 0.8706909418106079, | |
| "learning_rate": 8.885833790455294e-05, | |
| "loss": 0.3733229064941406, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 0.34440344403444034, | |
| "grad_norm": 0.9130774736404419, | |
| "learning_rate": 8.882405375754251e-05, | |
| "loss": 0.3727190399169922, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 0.34542845428454283, | |
| "grad_norm": 0.9043731093406677, | |
| "learning_rate": 8.878976961053209e-05, | |
| "loss": 0.3768247985839844, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 0.3464534645346454, | |
| "grad_norm": 0.9237262606620789, | |
| "learning_rate": 8.875548546352168e-05, | |
| "loss": 0.37340038299560546, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 0.34747847478474786, | |
| "grad_norm": 0.933336615562439, | |
| "learning_rate": 8.872120131651125e-05, | |
| "loss": 0.37377540588378905, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 0.34850348503485035, | |
| "grad_norm": 0.8860220909118652, | |
| "learning_rate": 8.868691716950083e-05, | |
| "loss": 0.37401512145996096, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.34952849528495283, | |
| "grad_norm": 0.8620807528495789, | |
| "learning_rate": 8.86526330224904e-05, | |
| "loss": 0.3739821624755859, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 0.3505535055350554, | |
| "grad_norm": 0.8988568186759949, | |
| "learning_rate": 8.861834887547999e-05, | |
| "loss": 0.36935020446777345, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 0.35157851578515786, | |
| "grad_norm": 0.8636974692344666, | |
| "learning_rate": 8.858406472846956e-05, | |
| "loss": 0.36726520538330076, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 0.35260352603526035, | |
| "grad_norm": 0.877202570438385, | |
| "learning_rate": 8.854978058145914e-05, | |
| "loss": 0.37382724761962893, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 0.35362853628536284, | |
| "grad_norm": 0.8962008357048035, | |
| "learning_rate": 8.851549643444871e-05, | |
| "loss": 0.3772966766357422, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 0.3546535465354654, | |
| "grad_norm": 0.9151343107223511, | |
| "learning_rate": 8.84812122874383e-05, | |
| "loss": 0.36620166778564456, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 0.35567855678556787, | |
| "grad_norm": 0.8452996611595154, | |
| "learning_rate": 8.844692814042786e-05, | |
| "loss": 0.36557750701904296, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 0.35670356703567035, | |
| "grad_norm": 0.9622381925582886, | |
| "learning_rate": 8.841264399341744e-05, | |
| "loss": 0.37210052490234374, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 0.35772857728577284, | |
| "grad_norm": 0.9372038841247559, | |
| "learning_rate": 8.837835984640703e-05, | |
| "loss": 0.3714407348632813, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 0.3587535875358754, | |
| "grad_norm": 0.8182379007339478, | |
| "learning_rate": 8.83440756993966e-05, | |
| "loss": 0.36388587951660156, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.35977859778597787, | |
| "grad_norm": 0.884780764579773, | |
| "learning_rate": 8.830979155238618e-05, | |
| "loss": 0.3660946273803711, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 0.36080360803608036, | |
| "grad_norm": 0.8666535019874573, | |
| "learning_rate": 8.827550740537576e-05, | |
| "loss": 0.37296184539794924, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 0.36182861828618285, | |
| "grad_norm": 0.8815746307373047, | |
| "learning_rate": 8.824122325836534e-05, | |
| "loss": 0.3715848922729492, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 0.3628536285362854, | |
| "grad_norm": 0.8872948884963989, | |
| "learning_rate": 8.820693911135491e-05, | |
| "loss": 0.3682816696166992, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 0.3638786387863879, | |
| "grad_norm": 0.8765414357185364, | |
| "learning_rate": 8.817265496434449e-05, | |
| "loss": 0.3699205017089844, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 0.36490364903649036, | |
| "grad_norm": 0.9134261608123779, | |
| "learning_rate": 8.813837081733408e-05, | |
| "loss": 0.37452091217041017, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 0.36592865928659285, | |
| "grad_norm": 0.7621716856956482, | |
| "learning_rate": 8.810408667032365e-05, | |
| "loss": 0.3741279602050781, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 0.3669536695366954, | |
| "grad_norm": 0.8742589950561523, | |
| "learning_rate": 8.806980252331323e-05, | |
| "loss": 0.36470718383789064, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 0.3679786797867979, | |
| "grad_norm": 0.9553512930870056, | |
| "learning_rate": 8.80355183763028e-05, | |
| "loss": 0.36966686248779296, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 0.36900369003690037, | |
| "grad_norm": 0.893338680267334, | |
| "learning_rate": 8.800123422929239e-05, | |
| "loss": 0.3742272186279297, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.37002870028700285, | |
| "grad_norm": 0.9760714769363403, | |
| "learning_rate": 8.796695008228196e-05, | |
| "loss": 0.37147285461425783, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 0.3710537105371054, | |
| "grad_norm": 0.8242753744125366, | |
| "learning_rate": 8.793266593527152e-05, | |
| "loss": 0.3680310821533203, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 0.3720787207872079, | |
| "grad_norm": 0.8709667921066284, | |
| "learning_rate": 8.789838178826111e-05, | |
| "loss": 0.3707925033569336, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 0.37310373103731037, | |
| "grad_norm": 0.916915237903595, | |
| "learning_rate": 8.786409764125069e-05, | |
| "loss": 0.3664524078369141, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 0.37412874128741286, | |
| "grad_norm": 0.9148959517478943, | |
| "learning_rate": 8.782981349424026e-05, | |
| "loss": 0.3651688766479492, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 0.3751537515375154, | |
| "grad_norm": 0.9615899324417114, | |
| "learning_rate": 8.779552934722985e-05, | |
| "loss": 0.3720450973510742, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 0.3761787617876179, | |
| "grad_norm": 0.8949769139289856, | |
| "learning_rate": 8.776124520021942e-05, | |
| "loss": 0.37032054901123046, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 0.3772037720377204, | |
| "grad_norm": 0.8993897438049316, | |
| "learning_rate": 8.7726961053209e-05, | |
| "loss": 0.36923397064208985, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 0.37822878228782286, | |
| "grad_norm": 0.9236711263656616, | |
| "learning_rate": 8.769267690619857e-05, | |
| "loss": 0.3678728485107422, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 0.3792537925379254, | |
| "grad_norm": 0.7910723090171814, | |
| "learning_rate": 8.765839275918816e-05, | |
| "loss": 0.36641387939453124, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.3802788027880279, | |
| "grad_norm": 0.93461674451828, | |
| "learning_rate": 8.762410861217774e-05, | |
| "loss": 0.36756076812744143, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 0.3813038130381304, | |
| "grad_norm": 0.9021868705749512, | |
| "learning_rate": 8.758982446516731e-05, | |
| "loss": 0.3747684097290039, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 0.38232882328823287, | |
| "grad_norm": 0.9199862480163574, | |
| "learning_rate": 8.755554031815689e-05, | |
| "loss": 0.3728667449951172, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 0.3833538335383354, | |
| "grad_norm": 0.8526470065116882, | |
| "learning_rate": 8.752125617114647e-05, | |
| "loss": 0.36846576690673827, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 0.3843788437884379, | |
| "grad_norm": 0.8709496855735779, | |
| "learning_rate": 8.748697202413605e-05, | |
| "loss": 0.3699648666381836, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 0.3854038540385404, | |
| "grad_norm": 0.8371281027793884, | |
| "learning_rate": 8.745268787712561e-05, | |
| "loss": 0.36740345001220703, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 0.38642886428864287, | |
| "grad_norm": 0.8197815418243408, | |
| "learning_rate": 8.74184037301152e-05, | |
| "loss": 0.3676946258544922, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 0.3874538745387454, | |
| "grad_norm": 0.9089030027389526, | |
| "learning_rate": 8.738411958310477e-05, | |
| "loss": 0.36753597259521487, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 0.3884788847888479, | |
| "grad_norm": 0.7963665127754211, | |
| "learning_rate": 8.734983543609435e-05, | |
| "loss": 0.36824161529541016, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 0.3895038950389504, | |
| "grad_norm": 0.8730947971343994, | |
| "learning_rate": 8.731555128908394e-05, | |
| "loss": 0.3718975067138672, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.3905289052890529, | |
| "grad_norm": 0.8167917132377625, | |
| "learning_rate": 8.728126714207351e-05, | |
| "loss": 0.36955986022949217, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 0.3915539155391554, | |
| "grad_norm": 1.0277336835861206, | |
| "learning_rate": 8.724698299506309e-05, | |
| "loss": 0.3669309234619141, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 0.3925789257892579, | |
| "grad_norm": 0.8616814017295837, | |
| "learning_rate": 8.721269884805266e-05, | |
| "loss": 0.3725516891479492, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 0.3936039360393604, | |
| "grad_norm": 0.895061731338501, | |
| "learning_rate": 8.717841470104225e-05, | |
| "loss": 0.36698299407958984, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 0.3946289462894629, | |
| "grad_norm": 0.8269166350364685, | |
| "learning_rate": 8.714413055403182e-05, | |
| "loss": 0.36657798767089844, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 0.3956539565395654, | |
| "grad_norm": 0.8734097480773926, | |
| "learning_rate": 8.71098464070214e-05, | |
| "loss": 0.37505271911621096, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 0.3966789667896679, | |
| "grad_norm": 0.793544590473175, | |
| "learning_rate": 8.707556226001097e-05, | |
| "loss": 0.37141807556152345, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 0.3977039770397704, | |
| "grad_norm": 0.811577558517456, | |
| "learning_rate": 8.704127811300056e-05, | |
| "loss": 0.36417709350585936, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 0.3987289872898729, | |
| "grad_norm": 0.81349778175354, | |
| "learning_rate": 8.700699396599014e-05, | |
| "loss": 0.36803470611572264, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 0.3997539975399754, | |
| "grad_norm": 0.8975096940994263, | |
| "learning_rate": 8.69727098189797e-05, | |
| "loss": 0.37214141845703125, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.4007790077900779, | |
| "grad_norm": 0.8204998970031738, | |
| "learning_rate": 8.693842567196928e-05, | |
| "loss": 0.3682279205322266, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 0.4018040180401804, | |
| "grad_norm": 0.9531434178352356, | |
| "learning_rate": 8.690414152495886e-05, | |
| "loss": 0.3753878402709961, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 0.4028290282902829, | |
| "grad_norm": 0.8048810958862305, | |
| "learning_rate": 8.686985737794843e-05, | |
| "loss": 0.3661195373535156, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 0.40385403854038543, | |
| "grad_norm": 0.8216005563735962, | |
| "learning_rate": 8.683557323093802e-05, | |
| "loss": 0.3661447525024414, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 0.4048790487904879, | |
| "grad_norm": 0.9691771268844604, | |
| "learning_rate": 8.68012890839276e-05, | |
| "loss": 0.3639635467529297, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 0.4059040590405904, | |
| "grad_norm": 0.8534151911735535, | |
| "learning_rate": 8.676700493691717e-05, | |
| "loss": 0.36912094116210936, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 0.4069290692906929, | |
| "grad_norm": 0.8257306814193726, | |
| "learning_rate": 8.673272078990675e-05, | |
| "loss": 0.36109989166259765, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 0.40795407954079543, | |
| "grad_norm": 0.8341092467308044, | |
| "learning_rate": 8.669843664289633e-05, | |
| "loss": 0.3666196060180664, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 0.4089790897908979, | |
| "grad_norm": 0.9771219491958618, | |
| "learning_rate": 8.666415249588591e-05, | |
| "loss": 0.3715673828125, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 0.4100041000410004, | |
| "grad_norm": 0.8525674343109131, | |
| "learning_rate": 8.662986834887548e-05, | |
| "loss": 0.3710095977783203, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.4100041000410004, | |
| "eval_loss": 0.44470059871673584, | |
| "eval_runtime": 4.3982, | |
| "eval_samples_per_second": 5684.167, | |
| "eval_steps_per_second": 5.684, | |
| "eval_teacher_cosine_loss": 0.2772529721260071, | |
| "eval_teacher_cosine_mean": 0.7227470278739929, | |
| "eval_teacher_cosine_p10": 0.6429551839828491, | |
| "eval_teacher_cosine_p50": 0.7318383455276489, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.4110291102911029, | |
| "grad_norm": 0.8950290679931641, | |
| "learning_rate": 8.659558420186506e-05, | |
| "loss": 0.3704460906982422, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 0.41205412054120544, | |
| "grad_norm": 0.899596095085144, | |
| "learning_rate": 8.656130005485465e-05, | |
| "loss": 0.3710757064819336, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 0.4130791307913079, | |
| "grad_norm": 0.9266843199729919, | |
| "learning_rate": 8.652701590784422e-05, | |
| "loss": 0.36684986114501955, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 0.4141041410414104, | |
| "grad_norm": 0.91863614320755, | |
| "learning_rate": 8.649273176083378e-05, | |
| "loss": 0.36687538146972654, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 0.4151291512915129, | |
| "grad_norm": 0.8501536250114441, | |
| "learning_rate": 8.645844761382337e-05, | |
| "loss": 0.362510986328125, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 0.41615416154161544, | |
| "grad_norm": 0.7653887271881104, | |
| "learning_rate": 8.642416346681295e-05, | |
| "loss": 0.37177783966064454, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 0.41717917179171793, | |
| "grad_norm": 0.9987759590148926, | |
| "learning_rate": 8.638987931980252e-05, | |
| "loss": 0.36352001190185546, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 0.4182041820418204, | |
| "grad_norm": 0.8876786231994629, | |
| "learning_rate": 8.635559517279211e-05, | |
| "loss": 0.36670082092285156, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 0.4192291922919229, | |
| "grad_norm": 0.9351732730865479, | |
| "learning_rate": 8.632131102578168e-05, | |
| "loss": 0.3672980499267578, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 0.42025420254202545, | |
| "grad_norm": 0.8635382652282715, | |
| "learning_rate": 8.628702687877126e-05, | |
| "loss": 0.3605247116088867, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.42127921279212793, | |
| "grad_norm": 0.9113935232162476, | |
| "learning_rate": 8.625274273176083e-05, | |
| "loss": 0.3625196838378906, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 0.4223042230422304, | |
| "grad_norm": 0.8909702897071838, | |
| "learning_rate": 8.621845858475042e-05, | |
| "loss": 0.36710758209228517, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 0.4233292332923329, | |
| "grad_norm": 0.9444068670272827, | |
| "learning_rate": 8.618417443774e-05, | |
| "loss": 0.36613998413085935, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 0.42435424354243545, | |
| "grad_norm": 0.8394571542739868, | |
| "learning_rate": 8.614989029072957e-05, | |
| "loss": 0.36490360260009763, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 0.42537925379253794, | |
| "grad_norm": 0.9182170033454895, | |
| "learning_rate": 8.611560614371915e-05, | |
| "loss": 0.3664513397216797, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 0.4264042640426404, | |
| "grad_norm": 0.7615907788276672, | |
| "learning_rate": 8.608132199670873e-05, | |
| "loss": 0.3545575332641602, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 0.4274292742927429, | |
| "grad_norm": 0.7987914085388184, | |
| "learning_rate": 8.604703784969831e-05, | |
| "loss": 0.3588758087158203, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 0.42845428454284545, | |
| "grad_norm": 0.8668729662895203, | |
| "learning_rate": 8.601275370268787e-05, | |
| "loss": 0.36827835083007815, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 0.42947929479294794, | |
| "grad_norm": 0.8589150905609131, | |
| "learning_rate": 8.597846955567746e-05, | |
| "loss": 0.36506805419921873, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 0.43050430504305043, | |
| "grad_norm": 0.841946005821228, | |
| "learning_rate": 8.594418540866703e-05, | |
| "loss": 0.37263389587402346, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.4315293152931529, | |
| "grad_norm": 0.8532674908638, | |
| "learning_rate": 8.590990126165661e-05, | |
| "loss": 0.37519485473632813, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 0.43255432554325546, | |
| "grad_norm": 0.8013874292373657, | |
| "learning_rate": 8.58756171146462e-05, | |
| "loss": 0.36339168548583983, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 0.43357933579335795, | |
| "grad_norm": 0.9031952619552612, | |
| "learning_rate": 8.584133296763577e-05, | |
| "loss": 0.36240840911865235, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 0.43460434604346043, | |
| "grad_norm": 0.8340909481048584, | |
| "learning_rate": 8.580704882062534e-05, | |
| "loss": 0.3635038757324219, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 0.4356293562935629, | |
| "grad_norm": 0.8390151858329773, | |
| "learning_rate": 8.577276467361492e-05, | |
| "loss": 0.3616721725463867, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 0.43665436654366546, | |
| "grad_norm": 0.8581411242485046, | |
| "learning_rate": 8.573848052660451e-05, | |
| "loss": 0.3666685485839844, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 0.43767937679376795, | |
| "grad_norm": 0.8377926349639893, | |
| "learning_rate": 8.570419637959408e-05, | |
| "loss": 0.3683164978027344, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 0.43870438704387044, | |
| "grad_norm": 0.8628886342048645, | |
| "learning_rate": 8.566991223258366e-05, | |
| "loss": 0.36182125091552736, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 0.4397293972939729, | |
| "grad_norm": 0.7596575617790222, | |
| "learning_rate": 8.563562808557323e-05, | |
| "loss": 0.3642469024658203, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 0.44075440754407547, | |
| "grad_norm": 0.8620837926864624, | |
| "learning_rate": 8.560134393856282e-05, | |
| "loss": 0.36394199371337893, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.44177941779417795, | |
| "grad_norm": 0.8858438730239868, | |
| "learning_rate": 8.55670597915524e-05, | |
| "loss": 0.3652682113647461, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 0.44280442804428044, | |
| "grad_norm": 0.9359121322631836, | |
| "learning_rate": 8.553277564454196e-05, | |
| "loss": 0.36370033264160156, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 0.4438294382943829, | |
| "grad_norm": 0.8444784283638, | |
| "learning_rate": 8.549849149753154e-05, | |
| "loss": 0.35973655700683593, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 0.44485444854448547, | |
| "grad_norm": 0.9217962622642517, | |
| "learning_rate": 8.546420735052112e-05, | |
| "loss": 0.36759464263916014, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 0.44587945879458796, | |
| "grad_norm": 0.8227363228797913, | |
| "learning_rate": 8.54299232035107e-05, | |
| "loss": 0.36530227661132814, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 0.44690446904469044, | |
| "grad_norm": 0.9252040982246399, | |
| "learning_rate": 8.539563905650028e-05, | |
| "loss": 0.3683010101318359, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 0.44792947929479293, | |
| "grad_norm": 0.8653552532196045, | |
| "learning_rate": 8.536135490948986e-05, | |
| "loss": 0.37175540924072265, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 0.4489544895448955, | |
| "grad_norm": 0.8379127979278564, | |
| "learning_rate": 8.532707076247943e-05, | |
| "loss": 0.36971298217773435, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 0.44997949979499796, | |
| "grad_norm": 0.8649946451187134, | |
| "learning_rate": 8.5292786615469e-05, | |
| "loss": 0.36593544006347656, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 0.45100451004510045, | |
| "grad_norm": 0.8102148771286011, | |
| "learning_rate": 8.52585024684586e-05, | |
| "loss": 0.36611740112304686, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.45202952029520294, | |
| "grad_norm": 0.9124872088432312, | |
| "learning_rate": 8.522421832144817e-05, | |
| "loss": 0.3684893798828125, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 0.4530545305453055, | |
| "grad_norm": 0.7713510394096375, | |
| "learning_rate": 8.518993417443774e-05, | |
| "loss": 0.35966949462890624, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 0.45407954079540797, | |
| "grad_norm": 0.9526383876800537, | |
| "learning_rate": 8.515565002742732e-05, | |
| "loss": 0.36799461364746094, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 0.45510455104551045, | |
| "grad_norm": 0.8538844585418701, | |
| "learning_rate": 8.51213658804169e-05, | |
| "loss": 0.3650806427001953, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 0.45612956129561294, | |
| "grad_norm": 0.8621478080749512, | |
| "learning_rate": 8.508708173340648e-05, | |
| "loss": 0.3745027160644531, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 0.4571545715457155, | |
| "grad_norm": 0.8080431222915649, | |
| "learning_rate": 8.505279758639604e-05, | |
| "loss": 0.36575820922851565, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 0.45817958179581797, | |
| "grad_norm": 0.8665605187416077, | |
| "learning_rate": 8.501851343938563e-05, | |
| "loss": 0.3654668426513672, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 0.45920459204592046, | |
| "grad_norm": 0.8259551525115967, | |
| "learning_rate": 8.49842292923752e-05, | |
| "loss": 0.3593029022216797, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 0.46022960229602294, | |
| "grad_norm": 0.8555912375450134, | |
| "learning_rate": 8.494994514536478e-05, | |
| "loss": 0.3595264434814453, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 0.4612546125461255, | |
| "grad_norm": 0.8761295080184937, | |
| "learning_rate": 8.491566099835437e-05, | |
| "loss": 0.35779743194580077, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.462279622796228, | |
| "grad_norm": 0.8698954582214355, | |
| "learning_rate": 8.488137685134394e-05, | |
| "loss": 0.3585723114013672, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 0.46330463304633046, | |
| "grad_norm": 0.9927098751068115, | |
| "learning_rate": 8.484709270433352e-05, | |
| "loss": 0.3644823455810547, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 0.46432964329643295, | |
| "grad_norm": 0.8049850463867188, | |
| "learning_rate": 8.481280855732309e-05, | |
| "loss": 0.3603937530517578, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 0.4653546535465355, | |
| "grad_norm": 0.8526210188865662, | |
| "learning_rate": 8.477852441031268e-05, | |
| "loss": 0.36474998474121095, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 0.466379663796638, | |
| "grad_norm": 0.8220229148864746, | |
| "learning_rate": 8.474424026330226e-05, | |
| "loss": 0.3630109405517578, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 0.46740467404674046, | |
| "grad_norm": 0.8481012582778931, | |
| "learning_rate": 8.470995611629183e-05, | |
| "loss": 0.36333370208740234, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 0.46842968429684295, | |
| "grad_norm": 0.8143265247344971, | |
| "learning_rate": 8.467567196928142e-05, | |
| "loss": 0.36565284729003905, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 0.4694546945469455, | |
| "grad_norm": 0.8416976928710938, | |
| "learning_rate": 8.464138782227099e-05, | |
| "loss": 0.3654035186767578, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 0.470479704797048, | |
| "grad_norm": 0.9457855224609375, | |
| "learning_rate": 8.460710367526057e-05, | |
| "loss": 0.36474658966064455, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 0.47150471504715047, | |
| "grad_norm": 0.8547295331954956, | |
| "learning_rate": 8.457281952825014e-05, | |
| "loss": 0.35817806243896483, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.47252972529725296, | |
| "grad_norm": 0.9635514616966248, | |
| "learning_rate": 8.453853538123972e-05, | |
| "loss": 0.3619247055053711, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 0.4735547355473555, | |
| "grad_norm": 0.8368202447891235, | |
| "learning_rate": 8.450425123422929e-05, | |
| "loss": 0.3631948089599609, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 0.474579745797458, | |
| "grad_norm": 0.8579609990119934, | |
| "learning_rate": 8.446996708721887e-05, | |
| "loss": 0.358449821472168, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 0.4756047560475605, | |
| "grad_norm": 0.8721423745155334, | |
| "learning_rate": 8.443568294020845e-05, | |
| "loss": 0.3620936965942383, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 0.47662976629766296, | |
| "grad_norm": 0.8403780460357666, | |
| "learning_rate": 8.440139879319803e-05, | |
| "loss": 0.3651991271972656, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 0.4776547765477655, | |
| "grad_norm": 0.8912323713302612, | |
| "learning_rate": 8.43671146461876e-05, | |
| "loss": 0.35714229583740237, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 0.478679786797868, | |
| "grad_norm": 0.8357952833175659, | |
| "learning_rate": 8.433283049917718e-05, | |
| "loss": 0.36668243408203127, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 0.4797047970479705, | |
| "grad_norm": 0.8607558608055115, | |
| "learning_rate": 8.429854635216677e-05, | |
| "loss": 0.36373176574707033, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 0.48072980729807296, | |
| "grad_norm": 0.8239311575889587, | |
| "learning_rate": 8.426426220515634e-05, | |
| "loss": 0.36499122619628904, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 0.4817548175481755, | |
| "grad_norm": 0.7704447507858276, | |
| "learning_rate": 8.422997805814592e-05, | |
| "loss": 0.36263248443603513, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.482779827798278, | |
| "grad_norm": 0.8514408469200134, | |
| "learning_rate": 8.41956939111355e-05, | |
| "loss": 0.36284027099609373, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 0.4838048380483805, | |
| "grad_norm": 0.8505940437316895, | |
| "learning_rate": 8.416140976412508e-05, | |
| "loss": 0.36450679779052736, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 0.48482984829848297, | |
| "grad_norm": 0.7906241416931152, | |
| "learning_rate": 8.412712561711465e-05, | |
| "loss": 0.36706554412841796, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 0.4858548585485855, | |
| "grad_norm": 0.8834350109100342, | |
| "learning_rate": 8.409284147010423e-05, | |
| "loss": 0.35738918304443357, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 0.486879868798688, | |
| "grad_norm": 0.8502029776573181, | |
| "learning_rate": 8.40585573230938e-05, | |
| "loss": 0.3631878280639648, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 0.4879048790487905, | |
| "grad_norm": 0.8504689335823059, | |
| "learning_rate": 8.402427317608338e-05, | |
| "loss": 0.360937614440918, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 0.488929889298893, | |
| "grad_norm": 0.8153953552246094, | |
| "learning_rate": 8.398998902907295e-05, | |
| "loss": 0.3639552688598633, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 0.4899548995489955, | |
| "grad_norm": 0.8356646299362183, | |
| "learning_rate": 8.395570488206254e-05, | |
| "loss": 0.3627224349975586, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 0.490979909799098, | |
| "grad_norm": 0.87696373462677, | |
| "learning_rate": 8.392142073505212e-05, | |
| "loss": 0.35759166717529295, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 0.4920049200492005, | |
| "grad_norm": 0.8025882244110107, | |
| "learning_rate": 8.388713658804169e-05, | |
| "loss": 0.3614153289794922, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.493029930299303, | |
| "grad_norm": 0.7722679972648621, | |
| "learning_rate": 8.385285244103127e-05, | |
| "loss": 0.360185546875, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 0.4940549405494055, | |
| "grad_norm": 0.7664601802825928, | |
| "learning_rate": 8.381856829402085e-05, | |
| "loss": 0.3594731140136719, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 0.495079950799508, | |
| "grad_norm": 0.8804201483726501, | |
| "learning_rate": 8.378428414701043e-05, | |
| "loss": 0.36097015380859376, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 0.4961049610496105, | |
| "grad_norm": 0.8388698697090149, | |
| "learning_rate": 8.375e-05, | |
| "loss": 0.3733727264404297, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 0.497129971299713, | |
| "grad_norm": 0.9241569638252258, | |
| "learning_rate": 8.371571585298959e-05, | |
| "loss": 0.3627790832519531, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 0.4981549815498155, | |
| "grad_norm": 0.9334095120429993, | |
| "learning_rate": 8.368143170597917e-05, | |
| "loss": 0.3637585830688477, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 0.499179991799918, | |
| "grad_norm": 0.7697033882141113, | |
| "learning_rate": 8.364714755896874e-05, | |
| "loss": 0.3658308029174805, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 0.5002050020500205, | |
| "grad_norm": 0.7736930847167969, | |
| "learning_rate": 8.361286341195832e-05, | |
| "loss": 0.35859806060791016, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 0.501230012300123, | |
| "grad_norm": 0.8636450171470642, | |
| "learning_rate": 8.357857926494789e-05, | |
| "loss": 0.35948280334472654, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 0.5022550225502255, | |
| "grad_norm": 0.8940028548240662, | |
| "learning_rate": 8.354429511793746e-05, | |
| "loss": 0.36054698944091795, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.503280032800328, | |
| "grad_norm": 0.9075922966003418, | |
| "learning_rate": 8.351001097092704e-05, | |
| "loss": 0.35775009155273435, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 0.5043050430504306, | |
| "grad_norm": 0.8745831847190857, | |
| "learning_rate": 8.347572682391663e-05, | |
| "loss": 0.360748291015625, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 0.505330053300533, | |
| "grad_norm": 0.9447755813598633, | |
| "learning_rate": 8.34414426769062e-05, | |
| "loss": 0.3613899230957031, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 0.5063550635506355, | |
| "grad_norm": 0.8780161738395691, | |
| "learning_rate": 8.340715852989578e-05, | |
| "loss": 0.3551458740234375, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 0.507380073800738, | |
| "grad_norm": 0.8721882700920105, | |
| "learning_rate": 8.337287438288535e-05, | |
| "loss": 0.36248302459716797, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 0.5084050840508405, | |
| "grad_norm": 0.7997353076934814, | |
| "learning_rate": 8.333859023587494e-05, | |
| "loss": 0.36661109924316404, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 0.509430094300943, | |
| "grad_norm": 0.9125805497169495, | |
| "learning_rate": 8.330430608886451e-05, | |
| "loss": 0.3618216705322266, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 0.5104551045510455, | |
| "grad_norm": 0.8255252838134766, | |
| "learning_rate": 8.327002194185409e-05, | |
| "loss": 0.3615468597412109, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 0.511480114801148, | |
| "grad_norm": 0.8176280856132507, | |
| "learning_rate": 8.323573779484368e-05, | |
| "loss": 0.36482738494873046, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 0.5125051250512506, | |
| "grad_norm": 0.8737911581993103, | |
| "learning_rate": 8.320145364783325e-05, | |
| "loss": 0.36439056396484376, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.5125051250512506, | |
| "eval_loss": 0.4389941394329071, | |
| "eval_runtime": 4.4356, | |
| "eval_samples_per_second": 5636.256, | |
| "eval_steps_per_second": 5.636, | |
| "eval_teacher_cosine_loss": 0.2751823663711548, | |
| "eval_teacher_cosine_mean": 0.7248176336288452, | |
| "eval_teacher_cosine_p10": 0.6459316611289978, | |
| "eval_teacher_cosine_p50": 0.733785092830658, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.513530135301353, | |
| "grad_norm": 0.7743839025497437, | |
| "learning_rate": 8.316716950082283e-05, | |
| "loss": 0.3658085632324219, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 0.5145551455514555, | |
| "grad_norm": 0.8161033987998962, | |
| "learning_rate": 8.31328853538124e-05, | |
| "loss": 0.35796260833740234, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 0.515580155801558, | |
| "grad_norm": 0.9121061563491821, | |
| "learning_rate": 8.309860120680198e-05, | |
| "loss": 0.35813514709472655, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 0.5166051660516605, | |
| "grad_norm": 0.7575347423553467, | |
| "learning_rate": 8.306431705979155e-05, | |
| "loss": 0.3563520812988281, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 0.517630176301763, | |
| "grad_norm": 0.824091374874115, | |
| "learning_rate": 8.303003291278113e-05, | |
| "loss": 0.35860725402832033, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 0.5186551865518655, | |
| "grad_norm": 0.8979877829551697, | |
| "learning_rate": 8.299574876577071e-05, | |
| "loss": 0.36064044952392577, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 0.519680196801968, | |
| "grad_norm": 0.8735562562942505, | |
| "learning_rate": 8.296146461876029e-05, | |
| "loss": 0.36381290435791014, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 0.5207052070520706, | |
| "grad_norm": 0.8899359107017517, | |
| "learning_rate": 8.292718047174986e-05, | |
| "loss": 0.36132225036621096, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 0.5217302173021731, | |
| "grad_norm": 0.8510153293609619, | |
| "learning_rate": 8.289289632473944e-05, | |
| "loss": 0.3581420516967773, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 0.5227552275522755, | |
| "grad_norm": 0.9191192388534546, | |
| "learning_rate": 8.285861217772903e-05, | |
| "loss": 0.35945365905761717, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.523780237802378, | |
| "grad_norm": 0.8325066566467285, | |
| "learning_rate": 8.28243280307186e-05, | |
| "loss": 0.3569718170166016, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 0.5248052480524805, | |
| "grad_norm": 0.7860215902328491, | |
| "learning_rate": 8.279004388370818e-05, | |
| "loss": 0.35934043884277345, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 0.525830258302583, | |
| "grad_norm": 0.9151034951210022, | |
| "learning_rate": 8.275575973669776e-05, | |
| "loss": 0.3618041610717773, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 0.5268552685526855, | |
| "grad_norm": 0.8099998235702515, | |
| "learning_rate": 8.272147558968734e-05, | |
| "loss": 0.354765739440918, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 0.527880278802788, | |
| "grad_norm": 0.8213974237442017, | |
| "learning_rate": 8.268719144267691e-05, | |
| "loss": 0.3602980804443359, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 0.5289052890528906, | |
| "grad_norm": 0.8709638714790344, | |
| "learning_rate": 8.265290729566649e-05, | |
| "loss": 0.3622092056274414, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 0.5299302993029931, | |
| "grad_norm": 0.7902204394340515, | |
| "learning_rate": 8.261862314865606e-05, | |
| "loss": 0.3606731414794922, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 0.5309553095530956, | |
| "grad_norm": 0.8207998871803284, | |
| "learning_rate": 8.258433900164564e-05, | |
| "loss": 0.35970375061035154, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 0.531980319803198, | |
| "grad_norm": 0.8831226229667664, | |
| "learning_rate": 8.255005485463521e-05, | |
| "loss": 0.36464622497558596, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 0.5330053300533005, | |
| "grad_norm": 0.7910507917404175, | |
| "learning_rate": 8.25157707076248e-05, | |
| "loss": 0.3535409164428711, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.534030340303403, | |
| "grad_norm": 0.9025859236717224, | |
| "learning_rate": 8.248148656061438e-05, | |
| "loss": 0.3638542938232422, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 0.5350553505535055, | |
| "grad_norm": 0.8940946459770203, | |
| "learning_rate": 8.244720241360395e-05, | |
| "loss": 0.3621247482299805, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 0.536080360803608, | |
| "grad_norm": 0.8906770944595337, | |
| "learning_rate": 8.241291826659352e-05, | |
| "loss": 0.36541515350341797, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 0.5371053710537106, | |
| "grad_norm": 0.8171073198318481, | |
| "learning_rate": 8.237863411958311e-05, | |
| "loss": 0.360729866027832, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 0.5381303813038131, | |
| "grad_norm": 0.9031678438186646, | |
| "learning_rate": 8.234434997257269e-05, | |
| "loss": 0.35964839935302734, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 0.5391553915539156, | |
| "grad_norm": 0.8438989520072937, | |
| "learning_rate": 8.231006582556226e-05, | |
| "loss": 0.35663795471191406, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 0.540180401804018, | |
| "grad_norm": 0.9370900988578796, | |
| "learning_rate": 8.227578167855185e-05, | |
| "loss": 0.3602850723266602, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 0.5412054120541205, | |
| "grad_norm": 1.0119235515594482, | |
| "learning_rate": 8.224149753154143e-05, | |
| "loss": 0.3577724456787109, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 0.542230422304223, | |
| "grad_norm": 0.8136317133903503, | |
| "learning_rate": 8.2207213384531e-05, | |
| "loss": 0.3647231292724609, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 0.5432554325543255, | |
| "grad_norm": 0.8323021531105042, | |
| "learning_rate": 8.217292923752057e-05, | |
| "loss": 0.3632455825805664, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.544280442804428, | |
| "grad_norm": 0.901305615901947, | |
| "learning_rate": 8.213864509051015e-05, | |
| "loss": 0.362430305480957, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 0.5453054530545306, | |
| "grad_norm": 0.8658463954925537, | |
| "learning_rate": 8.210436094349972e-05, | |
| "loss": 0.3622901153564453, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 0.5463304633046331, | |
| "grad_norm": 0.852361798286438, | |
| "learning_rate": 8.20700767964893e-05, | |
| "loss": 0.3547652816772461, | |
| "step": 26650 | |
| }, | |
| { | |
| "epoch": 0.5473554735547356, | |
| "grad_norm": 0.8690440654754639, | |
| "learning_rate": 8.203579264947889e-05, | |
| "loss": 0.3563945770263672, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 0.5483804838048381, | |
| "grad_norm": 0.8117708563804626, | |
| "learning_rate": 8.200150850246846e-05, | |
| "loss": 0.36114036560058593, | |
| "step": 26750 | |
| }, | |
| { | |
| "epoch": 0.5494054940549405, | |
| "grad_norm": 0.871225118637085, | |
| "learning_rate": 8.196722435545804e-05, | |
| "loss": 0.3577569580078125, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 0.550430504305043, | |
| "grad_norm": 0.8727680444717407, | |
| "learning_rate": 8.193294020844761e-05, | |
| "loss": 0.3637746047973633, | |
| "step": 26850 | |
| }, | |
| { | |
| "epoch": 0.5514555145551455, | |
| "grad_norm": 0.9404217600822449, | |
| "learning_rate": 8.18986560614372e-05, | |
| "loss": 0.3608892822265625, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 0.552480524805248, | |
| "grad_norm": 0.793751060962677, | |
| "learning_rate": 8.186437191442677e-05, | |
| "loss": 0.3594549560546875, | |
| "step": 26950 | |
| }, | |
| { | |
| "epoch": 0.5535055350553506, | |
| "grad_norm": 0.8895848989486694, | |
| "learning_rate": 8.183008776741635e-05, | |
| "loss": 0.36035865783691406, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.5545305453054531, | |
| "grad_norm": 0.8901317715644836, | |
| "learning_rate": 8.179580362040594e-05, | |
| "loss": 0.3637228012084961, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.773557186126709, | |
| "learning_rate": 8.176151947339551e-05, | |
| "loss": 0.36049617767333986, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 0.5565805658056581, | |
| "grad_norm": 0.8264882564544678, | |
| "learning_rate": 8.172723532638509e-05, | |
| "loss": 0.3556612777709961, | |
| "step": 27150 | |
| }, | |
| { | |
| "epoch": 0.5576055760557606, | |
| "grad_norm": 0.7710375785827637, | |
| "learning_rate": 8.169295117937466e-05, | |
| "loss": 0.3596794128417969, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 0.558630586305863, | |
| "grad_norm": 0.8807925581932068, | |
| "learning_rate": 8.165866703236424e-05, | |
| "loss": 0.35717517852783204, | |
| "step": 27250 | |
| }, | |
| { | |
| "epoch": 0.5596555965559655, | |
| "grad_norm": 0.7816828489303589, | |
| "learning_rate": 8.162438288535381e-05, | |
| "loss": 0.3548538589477539, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 0.560680606806068, | |
| "grad_norm": 0.8368867039680481, | |
| "learning_rate": 8.159009873834338e-05, | |
| "loss": 0.35506591796875, | |
| "step": 27350 | |
| }, | |
| { | |
| "epoch": 0.5617056170561706, | |
| "grad_norm": 0.8687714338302612, | |
| "learning_rate": 8.155581459133297e-05, | |
| "loss": 0.3566728973388672, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 0.5627306273062731, | |
| "grad_norm": 0.9154672622680664, | |
| "learning_rate": 8.152153044432255e-05, | |
| "loss": 0.3574481201171875, | |
| "step": 27450 | |
| }, | |
| { | |
| "epoch": 0.5637556375563756, | |
| "grad_norm": 0.7781676650047302, | |
| "learning_rate": 8.148724629731212e-05, | |
| "loss": 0.35708251953125, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.5647806478064781, | |
| "grad_norm": 0.7617968916893005, | |
| "learning_rate": 8.14529621503017e-05, | |
| "loss": 0.3614686584472656, | |
| "step": 27550 | |
| }, | |
| { | |
| "epoch": 0.5658056580565806, | |
| "grad_norm": 0.7849624752998352, | |
| "learning_rate": 8.141867800329129e-05, | |
| "loss": 0.3591892242431641, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 0.566830668306683, | |
| "grad_norm": 0.8132139444351196, | |
| "learning_rate": 8.138439385628086e-05, | |
| "loss": 0.35954967498779294, | |
| "step": 27650 | |
| }, | |
| { | |
| "epoch": 0.5678556785567855, | |
| "grad_norm": 0.8808278441429138, | |
| "learning_rate": 8.135010970927043e-05, | |
| "loss": 0.350660400390625, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 0.568880688806888, | |
| "grad_norm": 0.8600214719772339, | |
| "learning_rate": 8.131582556226002e-05, | |
| "loss": 0.3579346466064453, | |
| "step": 27750 | |
| }, | |
| { | |
| "epoch": 0.5699056990569906, | |
| "grad_norm": 0.84795081615448, | |
| "learning_rate": 8.12815414152496e-05, | |
| "loss": 0.35815021514892575, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 0.5709307093070931, | |
| "grad_norm": 0.7525407075881958, | |
| "learning_rate": 8.124725726823917e-05, | |
| "loss": 0.3572280502319336, | |
| "step": 27850 | |
| }, | |
| { | |
| "epoch": 0.5719557195571956, | |
| "grad_norm": 0.8173540234565735, | |
| "learning_rate": 8.121297312122875e-05, | |
| "loss": 0.3543804168701172, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 0.5729807298072981, | |
| "grad_norm": 0.8048562407493591, | |
| "learning_rate": 8.117868897421832e-05, | |
| "loss": 0.3592356491088867, | |
| "step": 27950 | |
| }, | |
| { | |
| "epoch": 0.5740057400574006, | |
| "grad_norm": 0.8439728617668152, | |
| "learning_rate": 8.11444048272079e-05, | |
| "loss": 0.35078819274902345, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.5750307503075031, | |
| "grad_norm": 0.7652221918106079, | |
| "learning_rate": 8.111012068019747e-05, | |
| "loss": 0.3577986145019531, | |
| "step": 28050 | |
| }, | |
| { | |
| "epoch": 0.5760557605576055, | |
| "grad_norm": 0.9582123756408691, | |
| "learning_rate": 8.107583653318706e-05, | |
| "loss": 0.3653752899169922, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 0.577080770807708, | |
| "grad_norm": 0.7919740080833435, | |
| "learning_rate": 8.104155238617663e-05, | |
| "loss": 0.3548568344116211, | |
| "step": 28150 | |
| }, | |
| { | |
| "epoch": 0.5781057810578106, | |
| "grad_norm": 0.8621631860733032, | |
| "learning_rate": 8.100726823916621e-05, | |
| "loss": 0.3536335754394531, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 0.5791307913079131, | |
| "grad_norm": 0.8584179282188416, | |
| "learning_rate": 8.097298409215578e-05, | |
| "loss": 0.354282112121582, | |
| "step": 28250 | |
| }, | |
| { | |
| "epoch": 0.5801558015580156, | |
| "grad_norm": 0.8936502933502197, | |
| "learning_rate": 8.093869994514537e-05, | |
| "loss": 0.36104148864746094, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 0.5811808118081181, | |
| "grad_norm": 0.8981918692588806, | |
| "learning_rate": 8.090441579813495e-05, | |
| "loss": 0.3560706329345703, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 0.5822058220582206, | |
| "grad_norm": 0.8531879186630249, | |
| "learning_rate": 8.087013165112452e-05, | |
| "loss": 0.3585588836669922, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 0.5832308323083231, | |
| "grad_norm": 0.8343385457992554, | |
| "learning_rate": 8.083584750411411e-05, | |
| "loss": 0.3554458999633789, | |
| "step": 28450 | |
| }, | |
| { | |
| "epoch": 0.5842558425584256, | |
| "grad_norm": 0.8878923654556274, | |
| "learning_rate": 8.080156335710368e-05, | |
| "loss": 0.3606267547607422, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.585280852808528, | |
| "grad_norm": 0.8992790579795837, | |
| "learning_rate": 8.076727921009326e-05, | |
| "loss": 0.3589632797241211, | |
| "step": 28550 | |
| }, | |
| { | |
| "epoch": 0.5863058630586306, | |
| "grad_norm": 0.8206989169120789, | |
| "learning_rate": 8.073299506308283e-05, | |
| "loss": 0.35453445434570313, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 0.5873308733087331, | |
| "grad_norm": 0.795693576335907, | |
| "learning_rate": 8.069871091607242e-05, | |
| "loss": 0.3590379333496094, | |
| "step": 28650 | |
| }, | |
| { | |
| "epoch": 0.5883558835588356, | |
| "grad_norm": 0.823332667350769, | |
| "learning_rate": 8.066442676906198e-05, | |
| "loss": 0.3627606964111328, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 0.5893808938089381, | |
| "grad_norm": 0.8944718837738037, | |
| "learning_rate": 8.063014262205156e-05, | |
| "loss": 0.3562588882446289, | |
| "step": 28750 | |
| }, | |
| { | |
| "epoch": 0.5904059040590406, | |
| "grad_norm": 0.8963759541511536, | |
| "learning_rate": 8.059585847504115e-05, | |
| "loss": 0.3564021301269531, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 0.5914309143091431, | |
| "grad_norm": 0.8016635179519653, | |
| "learning_rate": 8.056157432803072e-05, | |
| "loss": 0.35980003356933593, | |
| "step": 28850 | |
| }, | |
| { | |
| "epoch": 0.5924559245592456, | |
| "grad_norm": 0.8858565092086792, | |
| "learning_rate": 8.05272901810203e-05, | |
| "loss": 0.3569320297241211, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 0.593480934809348, | |
| "grad_norm": 0.9644514918327332, | |
| "learning_rate": 8.049300603400987e-05, | |
| "loss": 0.3493239212036133, | |
| "step": 28950 | |
| }, | |
| { | |
| "epoch": 0.5945059450594506, | |
| "grad_norm": 0.9269139766693115, | |
| "learning_rate": 8.045872188699946e-05, | |
| "loss": 0.3607057189941406, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.5955309553095531, | |
| "grad_norm": 0.782023549079895, | |
| "learning_rate": 8.042443773998903e-05, | |
| "loss": 0.3580877685546875, | |
| "step": 29050 | |
| }, | |
| { | |
| "epoch": 0.5965559655596556, | |
| "grad_norm": 0.8470706343650818, | |
| "learning_rate": 8.039015359297861e-05, | |
| "loss": 0.3600139617919922, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 0.5975809758097581, | |
| "grad_norm": 0.8379719257354736, | |
| "learning_rate": 8.03558694459682e-05, | |
| "loss": 0.35603939056396483, | |
| "step": 29150 | |
| }, | |
| { | |
| "epoch": 0.5986059860598606, | |
| "grad_norm": 0.7803846597671509, | |
| "learning_rate": 8.032158529895777e-05, | |
| "loss": 0.35749279022216796, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 0.5996309963099631, | |
| "grad_norm": 0.8948676586151123, | |
| "learning_rate": 8.028730115194735e-05, | |
| "loss": 0.3520519256591797, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 0.6006560065600656, | |
| "grad_norm": 0.8582128286361694, | |
| "learning_rate": 8.025301700493692e-05, | |
| "loss": 0.35623767852783206, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 0.6016810168101681, | |
| "grad_norm": 0.7041678428649902, | |
| "learning_rate": 8.021873285792651e-05, | |
| "loss": 0.356219596862793, | |
| "step": 29350 | |
| }, | |
| { | |
| "epoch": 0.6027060270602707, | |
| "grad_norm": 0.857638418674469, | |
| "learning_rate": 8.018444871091607e-05, | |
| "loss": 0.3607267761230469, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 0.6037310373103731, | |
| "grad_norm": 0.9416924118995667, | |
| "learning_rate": 8.015016456390564e-05, | |
| "loss": 0.35737953186035154, | |
| "step": 29450 | |
| }, | |
| { | |
| "epoch": 0.6047560475604756, | |
| "grad_norm": 0.8850821256637573, | |
| "learning_rate": 8.011588041689523e-05, | |
| "loss": 0.35307548522949217, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.6057810578105781, | |
| "grad_norm": 0.8739897608757019, | |
| "learning_rate": 8.008159626988481e-05, | |
| "loss": 0.3554246139526367, | |
| "step": 29550 | |
| }, | |
| { | |
| "epoch": 0.6068060680606806, | |
| "grad_norm": 0.9212623238563538, | |
| "learning_rate": 8.004731212287438e-05, | |
| "loss": 0.35762710571289064, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 0.6078310783107831, | |
| "grad_norm": 0.8691321611404419, | |
| "learning_rate": 8.001302797586397e-05, | |
| "loss": 0.35037574768066404, | |
| "step": 29650 | |
| }, | |
| { | |
| "epoch": 0.6088560885608856, | |
| "grad_norm": 0.8812757730484009, | |
| "learning_rate": 7.997874382885354e-05, | |
| "loss": 0.3612977600097656, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 0.6098810988109881, | |
| "grad_norm": 0.8670889735221863, | |
| "learning_rate": 7.994445968184312e-05, | |
| "loss": 0.35367000579833985, | |
| "step": 29750 | |
| }, | |
| { | |
| "epoch": 0.6109061090610907, | |
| "grad_norm": 0.8355680108070374, | |
| "learning_rate": 7.99101755348327e-05, | |
| "loss": 0.35907081604003904, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 0.6119311193111932, | |
| "grad_norm": 0.7381357550621033, | |
| "learning_rate": 7.987589138782228e-05, | |
| "loss": 0.3606062698364258, | |
| "step": 29850 | |
| }, | |
| { | |
| "epoch": 0.6129561295612956, | |
| "grad_norm": 0.8983296751976013, | |
| "learning_rate": 7.984160724081186e-05, | |
| "loss": 0.3548773193359375, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 0.6139811398113981, | |
| "grad_norm": 0.8917942643165588, | |
| "learning_rate": 7.980732309380143e-05, | |
| "loss": 0.3514129257202148, | |
| "step": 29950 | |
| }, | |
| { | |
| "epoch": 0.6150061500615006, | |
| "grad_norm": 0.840478777885437, | |
| "learning_rate": 7.9773038946791e-05, | |
| "loss": 0.3561463928222656, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.6150061500615006, | |
| "eval_loss": 0.4298919141292572, | |
| "eval_runtime": 4.3941, | |
| "eval_samples_per_second": 5689.401, | |
| "eval_steps_per_second": 5.689, | |
| "eval_teacher_cosine_loss": 0.26899826526641846, | |
| "eval_teacher_cosine_mean": 0.7310017347335815, | |
| "eval_teacher_cosine_p10": 0.6527736783027649, | |
| "eval_teacher_cosine_p50": 0.7399643659591675, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.6160311603116031, | |
| "grad_norm": 0.9246795773506165, | |
| "learning_rate": 7.97387547997806e-05, | |
| "loss": 0.3593678283691406, | |
| "step": 30050 | |
| }, | |
| { | |
| "epoch": 0.6170561705617056, | |
| "grad_norm": 0.8073226809501648, | |
| "learning_rate": 7.970447065277016e-05, | |
| "loss": 0.35351463317871096, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 0.6180811808118081, | |
| "grad_norm": 0.7920362949371338, | |
| "learning_rate": 7.967018650575973e-05, | |
| "loss": 0.35560050964355466, | |
| "step": 30150 | |
| }, | |
| { | |
| "epoch": 0.6191061910619107, | |
| "grad_norm": 0.842374861240387, | |
| "learning_rate": 7.963590235874932e-05, | |
| "loss": 0.35549949645996093, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 0.6201312013120132, | |
| "grad_norm": 0.8177646994590759, | |
| "learning_rate": 7.96016182117389e-05, | |
| "loss": 0.3581734848022461, | |
| "step": 30250 | |
| }, | |
| { | |
| "epoch": 0.6211562115621156, | |
| "grad_norm": 0.9136655330657959, | |
| "learning_rate": 7.956733406472847e-05, | |
| "loss": 0.35842212677001956, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 0.6221812218122181, | |
| "grad_norm": 0.9235081076622009, | |
| "learning_rate": 7.953304991771806e-05, | |
| "loss": 0.355417366027832, | |
| "step": 30350 | |
| }, | |
| { | |
| "epoch": 0.6232062320623206, | |
| "grad_norm": 0.8277457356452942, | |
| "learning_rate": 7.949876577070763e-05, | |
| "loss": 0.35267631530761717, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 0.6242312423124231, | |
| "grad_norm": 0.7737902402877808, | |
| "learning_rate": 7.94644816236972e-05, | |
| "loss": 0.35218448638916017, | |
| "step": 30450 | |
| }, | |
| { | |
| "epoch": 0.6252562525625256, | |
| "grad_norm": 0.8352388143539429, | |
| "learning_rate": 7.943019747668678e-05, | |
| "loss": 0.35176475524902345, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.6262812628126281, | |
| "grad_norm": 0.8856598138809204, | |
| "learning_rate": 7.939591332967637e-05, | |
| "loss": 0.3551509857177734, | |
| "step": 30550 | |
| }, | |
| { | |
| "epoch": 0.6273062730627307, | |
| "grad_norm": 0.8500962853431702, | |
| "learning_rate": 7.936162918266594e-05, | |
| "loss": 0.3560435104370117, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 0.6283312833128332, | |
| "grad_norm": 0.8450675010681152, | |
| "learning_rate": 7.932734503565552e-05, | |
| "loss": 0.3535994338989258, | |
| "step": 30650 | |
| }, | |
| { | |
| "epoch": 0.6293562935629357, | |
| "grad_norm": 0.8164230585098267, | |
| "learning_rate": 7.929306088864509e-05, | |
| "loss": 0.3492850112915039, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 0.6303813038130381, | |
| "grad_norm": 0.9341080784797668, | |
| "learning_rate": 7.925877674163468e-05, | |
| "loss": 0.35301616668701175, | |
| "step": 30750 | |
| }, | |
| { | |
| "epoch": 0.6314063140631406, | |
| "grad_norm": 0.7725896239280701, | |
| "learning_rate": 7.922449259462424e-05, | |
| "loss": 0.35379093170166015, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 0.6324313243132431, | |
| "grad_norm": 0.8028572797775269, | |
| "learning_rate": 7.919020844761382e-05, | |
| "loss": 0.358525390625, | |
| "step": 30850 | |
| }, | |
| { | |
| "epoch": 0.6334563345633456, | |
| "grad_norm": 0.7972704172134399, | |
| "learning_rate": 7.91559243006034e-05, | |
| "loss": 0.3579832458496094, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 0.6344813448134481, | |
| "grad_norm": 0.7610266208648682, | |
| "learning_rate": 7.912164015359298e-05, | |
| "loss": 0.3583976364135742, | |
| "step": 30950 | |
| }, | |
| { | |
| "epoch": 0.6355063550635507, | |
| "grad_norm": 0.8298759460449219, | |
| "learning_rate": 7.908735600658255e-05, | |
| "loss": 0.35183197021484375, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.6365313653136532, | |
| "grad_norm": 0.8527406454086304, | |
| "learning_rate": 7.905307185957214e-05, | |
| "loss": 0.35047210693359376, | |
| "step": 31050 | |
| }, | |
| { | |
| "epoch": 0.6375563755637557, | |
| "grad_norm": 0.9645882844924927, | |
| "learning_rate": 7.901878771256172e-05, | |
| "loss": 0.3544742584228516, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 0.6385813858138581, | |
| "grad_norm": 0.7952987551689148, | |
| "learning_rate": 7.898450356555129e-05, | |
| "loss": 0.35582889556884767, | |
| "step": 31150 | |
| }, | |
| { | |
| "epoch": 0.6396063960639606, | |
| "grad_norm": 0.8509933352470398, | |
| "learning_rate": 7.895021941854087e-05, | |
| "loss": 0.36014842987060547, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 0.6406314063140631, | |
| "grad_norm": 0.7833871841430664, | |
| "learning_rate": 7.891593527153046e-05, | |
| "loss": 0.3513349151611328, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 0.6416564165641656, | |
| "grad_norm": 0.7752136588096619, | |
| "learning_rate": 7.888165112452003e-05, | |
| "loss": 0.3604141616821289, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 0.6426814268142681, | |
| "grad_norm": 0.853550910949707, | |
| "learning_rate": 7.88473669775096e-05, | |
| "loss": 0.3557135009765625, | |
| "step": 31350 | |
| }, | |
| { | |
| "epoch": 0.6437064370643707, | |
| "grad_norm": 0.815820574760437, | |
| "learning_rate": 7.881308283049918e-05, | |
| "loss": 0.3530382919311523, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 0.6447314473144732, | |
| "grad_norm": 0.8135126233100891, | |
| "learning_rate": 7.877879868348877e-05, | |
| "loss": 0.3535493469238281, | |
| "step": 31450 | |
| }, | |
| { | |
| "epoch": 0.6457564575645757, | |
| "grad_norm": 0.7749558687210083, | |
| "learning_rate": 7.874451453647833e-05, | |
| "loss": 0.3513792419433594, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.6467814678146782, | |
| "grad_norm": 0.8742313385009766, | |
| "learning_rate": 7.87102303894679e-05, | |
| "loss": 0.35767345428466796, | |
| "step": 31550 | |
| }, | |
| { | |
| "epoch": 0.6478064780647806, | |
| "grad_norm": 0.7352192997932434, | |
| "learning_rate": 7.867594624245749e-05, | |
| "loss": 0.34823787689208985, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 0.6488314883148831, | |
| "grad_norm": 0.8540325164794922, | |
| "learning_rate": 7.864166209544707e-05, | |
| "loss": 0.35097164154052735, | |
| "step": 31650 | |
| }, | |
| { | |
| "epoch": 0.6498564985649856, | |
| "grad_norm": 0.8557196259498596, | |
| "learning_rate": 7.860737794843664e-05, | |
| "loss": 0.35128978729248045, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 0.6508815088150881, | |
| "grad_norm": 0.8197920918464661, | |
| "learning_rate": 7.857309380142623e-05, | |
| "loss": 0.3552590560913086, | |
| "step": 31750 | |
| }, | |
| { | |
| "epoch": 0.6519065190651907, | |
| "grad_norm": 0.9171255826950073, | |
| "learning_rate": 7.85388096544158e-05, | |
| "loss": 0.35281253814697267, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 0.6529315293152932, | |
| "grad_norm": 0.842154860496521, | |
| "learning_rate": 7.850452550740538e-05, | |
| "loss": 0.3565026092529297, | |
| "step": 31850 | |
| }, | |
| { | |
| "epoch": 0.6539565395653957, | |
| "grad_norm": 0.8630461096763611, | |
| "learning_rate": 7.847024136039495e-05, | |
| "loss": 0.35101253509521485, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 0.6549815498154982, | |
| "grad_norm": 0.9228710532188416, | |
| "learning_rate": 7.843595721338454e-05, | |
| "loss": 0.3504985809326172, | |
| "step": 31950 | |
| }, | |
| { | |
| "epoch": 0.6560065600656007, | |
| "grad_norm": 0.7992339730262756, | |
| "learning_rate": 7.840167306637412e-05, | |
| "loss": 0.35441749572753906, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.6570315703157031, | |
| "grad_norm": 0.7641682624816895, | |
| "learning_rate": 7.836738891936369e-05, | |
| "loss": 0.3570366287231445, | |
| "step": 32050 | |
| }, | |
| { | |
| "epoch": 0.6580565805658056, | |
| "grad_norm": 0.8735712766647339, | |
| "learning_rate": 7.833310477235327e-05, | |
| "loss": 0.3487904739379883, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 0.6590815908159081, | |
| "grad_norm": 0.8495463132858276, | |
| "learning_rate": 7.829882062534285e-05, | |
| "loss": 0.34899894714355467, | |
| "step": 32150 | |
| }, | |
| { | |
| "epoch": 0.6601066010660107, | |
| "grad_norm": 0.8347789645195007, | |
| "learning_rate": 7.826453647833242e-05, | |
| "loss": 0.3556767654418945, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 0.6611316113161132, | |
| "grad_norm": 0.875018835067749, | |
| "learning_rate": 7.823025233132199e-05, | |
| "loss": 0.3562625885009766, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 0.6621566215662157, | |
| "grad_norm": 0.8710410594940186, | |
| "learning_rate": 7.819596818431158e-05, | |
| "loss": 0.35376251220703125, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 0.6631816318163182, | |
| "grad_norm": 0.8194191455841064, | |
| "learning_rate": 7.816168403730115e-05, | |
| "loss": 0.35417415618896486, | |
| "step": 32350 | |
| }, | |
| { | |
| "epoch": 0.6642066420664207, | |
| "grad_norm": 0.8622740507125854, | |
| "learning_rate": 7.812739989029073e-05, | |
| "loss": 0.35683162689208986, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 0.6652316523165231, | |
| "grad_norm": 0.9038733243942261, | |
| "learning_rate": 7.809311574328032e-05, | |
| "loss": 0.3521404647827148, | |
| "step": 32450 | |
| }, | |
| { | |
| "epoch": 0.6662566625666256, | |
| "grad_norm": 0.8288673758506775, | |
| "learning_rate": 7.805883159626989e-05, | |
| "loss": 0.35355560302734373, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.6672816728167281, | |
| "grad_norm": 0.8682975769042969, | |
| "learning_rate": 7.802454744925947e-05, | |
| "loss": 0.3517645263671875, | |
| "step": 32550 | |
| }, | |
| { | |
| "epoch": 0.6683066830668307, | |
| "grad_norm": 0.8398820161819458, | |
| "learning_rate": 7.799026330224904e-05, | |
| "loss": 0.35899986267089845, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 0.6693316933169332, | |
| "grad_norm": 0.8845388293266296, | |
| "learning_rate": 7.795597915523863e-05, | |
| "loss": 0.352743034362793, | |
| "step": 32650 | |
| }, | |
| { | |
| "epoch": 0.6703567035670357, | |
| "grad_norm": 0.8872183561325073, | |
| "learning_rate": 7.79216950082282e-05, | |
| "loss": 0.35604019165039064, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 0.6713817138171382, | |
| "grad_norm": 0.9090275168418884, | |
| "learning_rate": 7.788741086121778e-05, | |
| "loss": 0.3581103515625, | |
| "step": 32750 | |
| }, | |
| { | |
| "epoch": 0.6724067240672407, | |
| "grad_norm": 0.8390264511108398, | |
| "learning_rate": 7.785312671420735e-05, | |
| "loss": 0.35154487609863283, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 0.6734317343173432, | |
| "grad_norm": 0.8008032441139221, | |
| "learning_rate": 7.781884256719694e-05, | |
| "loss": 0.3580214691162109, | |
| "step": 32850 | |
| }, | |
| { | |
| "epoch": 0.6744567445674456, | |
| "grad_norm": 0.8097127079963684, | |
| "learning_rate": 7.77845584201865e-05, | |
| "loss": 0.3546445846557617, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 0.6754817548175481, | |
| "grad_norm": 0.859113335609436, | |
| "learning_rate": 7.775027427317608e-05, | |
| "loss": 0.3563142395019531, | |
| "step": 32950 | |
| }, | |
| { | |
| "epoch": 0.6765067650676507, | |
| "grad_norm": 0.8614320755004883, | |
| "learning_rate": 7.771599012616566e-05, | |
| "loss": 0.35070980072021485, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.6775317753177532, | |
| "grad_norm": 0.8451588153839111, | |
| "learning_rate": 7.768170597915524e-05, | |
| "loss": 0.35453773498535157, | |
| "step": 33050 | |
| }, | |
| { | |
| "epoch": 0.6785567855678557, | |
| "grad_norm": 0.827163815498352, | |
| "learning_rate": 7.764742183214481e-05, | |
| "loss": 0.3556865310668945, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 0.6795817958179582, | |
| "grad_norm": 0.7939395308494568, | |
| "learning_rate": 7.76131376851344e-05, | |
| "loss": 0.35079715728759764, | |
| "step": 33150 | |
| }, | |
| { | |
| "epoch": 0.6806068060680607, | |
| "grad_norm": 0.8408690094947815, | |
| "learning_rate": 7.757885353812398e-05, | |
| "loss": 0.35393112182617187, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 0.6816318163181632, | |
| "grad_norm": 0.8484756350517273, | |
| "learning_rate": 7.754456939111355e-05, | |
| "loss": 0.3522013473510742, | |
| "step": 33250 | |
| }, | |
| { | |
| "epoch": 0.6826568265682657, | |
| "grad_norm": 0.8404049277305603, | |
| "learning_rate": 7.751028524410313e-05, | |
| "loss": 0.3599639892578125, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 0.6836818368183681, | |
| "grad_norm": 0.8588556051254272, | |
| "learning_rate": 7.747600109709271e-05, | |
| "loss": 0.3599920654296875, | |
| "step": 33350 | |
| }, | |
| { | |
| "epoch": 0.6847068470684707, | |
| "grad_norm": 0.8720951676368713, | |
| "learning_rate": 7.744171695008229e-05, | |
| "loss": 0.3526485061645508, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 0.6857318573185732, | |
| "grad_norm": 0.8624259829521179, | |
| "learning_rate": 7.740743280307186e-05, | |
| "loss": 0.35140888214111327, | |
| "step": 33450 | |
| }, | |
| { | |
| "epoch": 0.6867568675686757, | |
| "grad_norm": 0.7727066278457642, | |
| "learning_rate": 7.737314865606144e-05, | |
| "loss": 0.3447951889038086, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.6877818778187782, | |
| "grad_norm": 0.6971118450164795, | |
| "learning_rate": 7.733886450905103e-05, | |
| "loss": 0.3536207580566406, | |
| "step": 33550 | |
| }, | |
| { | |
| "epoch": 0.6888068880688807, | |
| "grad_norm": 0.841401994228363, | |
| "learning_rate": 7.730458036204059e-05, | |
| "loss": 0.3530032730102539, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 0.6898318983189832, | |
| "grad_norm": 0.7993757724761963, | |
| "learning_rate": 7.727029621503016e-05, | |
| "loss": 0.3522896194458008, | |
| "step": 33650 | |
| }, | |
| { | |
| "epoch": 0.6908569085690857, | |
| "grad_norm": 0.829692006111145, | |
| "learning_rate": 7.723601206801975e-05, | |
| "loss": 0.35281139373779297, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 0.6918819188191881, | |
| "grad_norm": 0.8604363799095154, | |
| "learning_rate": 7.720172792100933e-05, | |
| "loss": 0.35130794525146486, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 0.6929069290692907, | |
| "grad_norm": 0.8765594959259033, | |
| "learning_rate": 7.71674437739989e-05, | |
| "loss": 0.3549665069580078, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 0.6939319393193932, | |
| "grad_norm": 0.7937105298042297, | |
| "learning_rate": 7.713315962698849e-05, | |
| "loss": 0.35160835266113283, | |
| "step": 33850 | |
| }, | |
| { | |
| "epoch": 0.6949569495694957, | |
| "grad_norm": 0.8403233885765076, | |
| "learning_rate": 7.709887547997806e-05, | |
| "loss": 0.35586540222167967, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 0.6959819598195982, | |
| "grad_norm": 0.7591594457626343, | |
| "learning_rate": 7.706459133296764e-05, | |
| "loss": 0.3506972885131836, | |
| "step": 33950 | |
| }, | |
| { | |
| "epoch": 0.6970069700697007, | |
| "grad_norm": 0.8192698955535889, | |
| "learning_rate": 7.703030718595721e-05, | |
| "loss": 0.3495545196533203, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.6980319803198032, | |
| "grad_norm": 0.9285784363746643, | |
| "learning_rate": 7.69960230389468e-05, | |
| "loss": 0.35372390747070315, | |
| "step": 34050 | |
| }, | |
| { | |
| "epoch": 0.6990569905699057, | |
| "grad_norm": 0.8133446574211121, | |
| "learning_rate": 7.696173889193638e-05, | |
| "loss": 0.3541581726074219, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 0.7000820008200082, | |
| "grad_norm": 0.8204118013381958, | |
| "learning_rate": 7.692745474492595e-05, | |
| "loss": 0.3458837127685547, | |
| "step": 34150 | |
| }, | |
| { | |
| "epoch": 0.7011070110701108, | |
| "grad_norm": 0.7600720524787903, | |
| "learning_rate": 7.689317059791552e-05, | |
| "loss": 0.3474576187133789, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 0.7021320213202132, | |
| "grad_norm": 0.8301897644996643, | |
| "learning_rate": 7.685888645090511e-05, | |
| "loss": 0.35668689727783204, | |
| "step": 34250 | |
| }, | |
| { | |
| "epoch": 0.7031570315703157, | |
| "grad_norm": 0.856426477432251, | |
| "learning_rate": 7.682460230389469e-05, | |
| "loss": 0.3498827362060547, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 0.7041820418204182, | |
| "grad_norm": 0.8601041436195374, | |
| "learning_rate": 7.679031815688425e-05, | |
| "loss": 0.35324199676513673, | |
| "step": 34350 | |
| }, | |
| { | |
| "epoch": 0.7052070520705207, | |
| "grad_norm": 0.8945884704589844, | |
| "learning_rate": 7.675603400987384e-05, | |
| "loss": 0.3542506790161133, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 0.7062320623206232, | |
| "grad_norm": 0.8104718923568726, | |
| "learning_rate": 7.672174986286341e-05, | |
| "loss": 0.354705810546875, | |
| "step": 34450 | |
| }, | |
| { | |
| "epoch": 0.7072570725707257, | |
| "grad_norm": 0.808803915977478, | |
| "learning_rate": 7.668746571585299e-05, | |
| "loss": 0.35393505096435546, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.7082820828208282, | |
| "grad_norm": 0.7353049516677856, | |
| "learning_rate": 7.665318156884257e-05, | |
| "loss": 0.3468206787109375, | |
| "step": 34550 | |
| }, | |
| { | |
| "epoch": 0.7093070930709308, | |
| "grad_norm": 0.751739501953125, | |
| "learning_rate": 7.661889742183215e-05, | |
| "loss": 0.34675010681152346, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 0.7103321033210332, | |
| "grad_norm": 0.8857824206352234, | |
| "learning_rate": 7.658461327482172e-05, | |
| "loss": 0.3526826095581055, | |
| "step": 34650 | |
| }, | |
| { | |
| "epoch": 0.7113571135711357, | |
| "grad_norm": 0.8352428674697876, | |
| "learning_rate": 7.65503291278113e-05, | |
| "loss": 0.348548583984375, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 0.7123821238212382, | |
| "grad_norm": 0.7677897214889526, | |
| "learning_rate": 7.651604498080089e-05, | |
| "loss": 0.35046283721923827, | |
| "step": 34750 | |
| }, | |
| { | |
| "epoch": 0.7134071340713407, | |
| "grad_norm": 0.8878645896911621, | |
| "learning_rate": 7.648176083379046e-05, | |
| "loss": 0.34630340576171875, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 0.7144321443214432, | |
| "grad_norm": 0.7644602656364441, | |
| "learning_rate": 7.644747668678004e-05, | |
| "loss": 0.34488227844238284, | |
| "step": 34850 | |
| }, | |
| { | |
| "epoch": 0.7154571545715457, | |
| "grad_norm": 0.782006025314331, | |
| "learning_rate": 7.641319253976961e-05, | |
| "loss": 0.3503366470336914, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 0.7164821648216482, | |
| "grad_norm": 0.8230557441711426, | |
| "learning_rate": 7.63789083927592e-05, | |
| "loss": 0.34512527465820314, | |
| "step": 34950 | |
| }, | |
| { | |
| "epoch": 0.7175071750717508, | |
| "grad_norm": 0.8710888624191284, | |
| "learning_rate": 7.634462424574877e-05, | |
| "loss": 0.3471884536743164, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.7175071750717508, | |
| "eval_loss": 0.42443960905075073, | |
| "eval_runtime": 4.4231, | |
| "eval_samples_per_second": 5652.202, | |
| "eval_steps_per_second": 5.652, | |
| "eval_teacher_cosine_loss": 0.26427584886550903, | |
| "eval_teacher_cosine_mean": 0.735724151134491, | |
| "eval_teacher_cosine_p10": 0.6578198671340942, | |
| "eval_teacher_cosine_p50": 0.7444288730621338, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.7185321853218533, | |
| "grad_norm": 0.8859487771987915, | |
| "learning_rate": 7.631034009873834e-05, | |
| "loss": 0.35407520294189454, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 0.7195571955719557, | |
| "grad_norm": 0.8879075646400452, | |
| "learning_rate": 7.627605595172792e-05, | |
| "loss": 0.35130149841308594, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 0.7205822058220582, | |
| "grad_norm": 0.845038890838623, | |
| "learning_rate": 7.62417718047175e-05, | |
| "loss": 0.34823848724365236, | |
| "step": 35150 | |
| }, | |
| { | |
| "epoch": 0.7216072160721607, | |
| "grad_norm": 0.7380141615867615, | |
| "learning_rate": 7.620748765770707e-05, | |
| "loss": 0.3503864669799805, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 0.7226322263222632, | |
| "grad_norm": 0.9759238362312317, | |
| "learning_rate": 7.617320351069666e-05, | |
| "loss": 0.3549625396728516, | |
| "step": 35250 | |
| }, | |
| { | |
| "epoch": 0.7236572365723657, | |
| "grad_norm": 0.8182779550552368, | |
| "learning_rate": 7.613891936368624e-05, | |
| "loss": 0.34774044036865237, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 0.7246822468224682, | |
| "grad_norm": 0.797929048538208, | |
| "learning_rate": 7.610463521667581e-05, | |
| "loss": 0.34825763702392576, | |
| "step": 35350 | |
| }, | |
| { | |
| "epoch": 0.7257072570725708, | |
| "grad_norm": 0.8060253262519836, | |
| "learning_rate": 7.607035106966539e-05, | |
| "loss": 0.3439775848388672, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 0.7267322673226733, | |
| "grad_norm": 0.8674017190933228, | |
| "learning_rate": 7.603606692265497e-05, | |
| "loss": 0.3493878173828125, | |
| "step": 35450 | |
| }, | |
| { | |
| "epoch": 0.7277572775727758, | |
| "grad_norm": 0.8006479144096375, | |
| "learning_rate": 7.600178277564455e-05, | |
| "loss": 0.349954833984375, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.7287822878228782, | |
| "grad_norm": 0.7979263663291931, | |
| "learning_rate": 7.596749862863412e-05, | |
| "loss": 0.3423832702636719, | |
| "step": 35550 | |
| }, | |
| { | |
| "epoch": 0.7298072980729807, | |
| "grad_norm": 0.8829553127288818, | |
| "learning_rate": 7.59332144816237e-05, | |
| "loss": 0.348779296875, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 0.7308323083230832, | |
| "grad_norm": 0.850635290145874, | |
| "learning_rate": 7.589893033461329e-05, | |
| "loss": 0.3477918243408203, | |
| "step": 35650 | |
| }, | |
| { | |
| "epoch": 0.7318573185731857, | |
| "grad_norm": 0.8203648328781128, | |
| "learning_rate": 7.586464618760286e-05, | |
| "loss": 0.3432146453857422, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 0.7328823288232882, | |
| "grad_norm": 0.792127251625061, | |
| "learning_rate": 7.583036204059242e-05, | |
| "loss": 0.34449928283691406, | |
| "step": 35750 | |
| }, | |
| { | |
| "epoch": 0.7339073390733908, | |
| "grad_norm": 0.857633650302887, | |
| "learning_rate": 7.579607789358201e-05, | |
| "loss": 0.34855850219726564, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 0.7349323493234933, | |
| "grad_norm": 0.7725878953933716, | |
| "learning_rate": 7.576179374657158e-05, | |
| "loss": 0.34717330932617185, | |
| "step": 35850 | |
| }, | |
| { | |
| "epoch": 0.7359573595735958, | |
| "grad_norm": 0.8133530616760254, | |
| "learning_rate": 7.572750959956116e-05, | |
| "loss": 0.34999565124511717, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 0.7369823698236982, | |
| "grad_norm": 0.8477935194969177, | |
| "learning_rate": 7.569322545255075e-05, | |
| "loss": 0.3467626953125, | |
| "step": 35950 | |
| }, | |
| { | |
| "epoch": 0.7380073800738007, | |
| "grad_norm": 0.8187217712402344, | |
| "learning_rate": 7.565894130554032e-05, | |
| "loss": 0.35038536071777343, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.7390323903239032, | |
| "grad_norm": 0.7854519486427307, | |
| "learning_rate": 7.56246571585299e-05, | |
| "loss": 0.3512165069580078, | |
| "step": 36050 | |
| }, | |
| { | |
| "epoch": 0.7400574005740057, | |
| "grad_norm": 0.8167485594749451, | |
| "learning_rate": 7.559037301151947e-05, | |
| "loss": 0.35295101165771486, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 0.7410824108241082, | |
| "grad_norm": 0.7738946676254272, | |
| "learning_rate": 7.555608886450906e-05, | |
| "loss": 0.35010246276855467, | |
| "step": 36150 | |
| }, | |
| { | |
| "epoch": 0.7421074210742108, | |
| "grad_norm": 0.8674753308296204, | |
| "learning_rate": 7.552180471749863e-05, | |
| "loss": 0.35264347076416014, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 0.7431324313243133, | |
| "grad_norm": 0.8231740593910217, | |
| "learning_rate": 7.548752057048821e-05, | |
| "loss": 0.34949455261230467, | |
| "step": 36250 | |
| }, | |
| { | |
| "epoch": 0.7441574415744158, | |
| "grad_norm": 0.781320333480835, | |
| "learning_rate": 7.54532364234778e-05, | |
| "loss": 0.3494378662109375, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 0.7451824518245183, | |
| "grad_norm": 0.823767900466919, | |
| "learning_rate": 7.541895227646737e-05, | |
| "loss": 0.34346694946289064, | |
| "step": 36350 | |
| }, | |
| { | |
| "epoch": 0.7462074620746207, | |
| "grad_norm": 0.8600749969482422, | |
| "learning_rate": 7.538466812945695e-05, | |
| "loss": 0.3537895965576172, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 0.7472324723247232, | |
| "grad_norm": 0.8500552773475647, | |
| "learning_rate": 7.535038398244652e-05, | |
| "loss": 0.3546759033203125, | |
| "step": 36450 | |
| }, | |
| { | |
| "epoch": 0.7482574825748257, | |
| "grad_norm": 0.7751675844192505, | |
| "learning_rate": 7.53160998354361e-05, | |
| "loss": 0.35273712158203124, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.7492824928249282, | |
| "grad_norm": 0.8140791058540344, | |
| "learning_rate": 7.528181568842567e-05, | |
| "loss": 0.3535601806640625, | |
| "step": 36550 | |
| }, | |
| { | |
| "epoch": 0.7503075030750308, | |
| "grad_norm": 0.6779204607009888, | |
| "learning_rate": 7.524753154141525e-05, | |
| "loss": 0.34812538146972655, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 0.7513325133251333, | |
| "grad_norm": 0.7942777276039124, | |
| "learning_rate": 7.521324739440483e-05, | |
| "loss": 0.3493399429321289, | |
| "step": 36650 | |
| }, | |
| { | |
| "epoch": 0.7523575235752358, | |
| "grad_norm": 0.8550200462341309, | |
| "learning_rate": 7.517896324739441e-05, | |
| "loss": 0.35684398651123045, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 0.7533825338253383, | |
| "grad_norm": 0.7836914658546448, | |
| "learning_rate": 7.514467910038398e-05, | |
| "loss": 0.3528653335571289, | |
| "step": 36750 | |
| }, | |
| { | |
| "epoch": 0.7544075440754408, | |
| "grad_norm": 0.8352323174476624, | |
| "learning_rate": 7.511039495337356e-05, | |
| "loss": 0.3538046646118164, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 0.7554325543255432, | |
| "grad_norm": 0.9169872403144836, | |
| "learning_rate": 7.507611080636315e-05, | |
| "loss": 0.3514836883544922, | |
| "step": 36850 | |
| }, | |
| { | |
| "epoch": 0.7564575645756457, | |
| "grad_norm": 0.7737328410148621, | |
| "learning_rate": 7.504182665935272e-05, | |
| "loss": 0.3467452621459961, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 0.7574825748257482, | |
| "grad_norm": 0.805257260799408, | |
| "learning_rate": 7.50075425123423e-05, | |
| "loss": 0.34786018371582034, | |
| "step": 36950 | |
| }, | |
| { | |
| "epoch": 0.7585075850758508, | |
| "grad_norm": 0.8026707172393799, | |
| "learning_rate": 7.497325836533188e-05, | |
| "loss": 0.34376163482666017, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.7595325953259533, | |
| "grad_norm": 0.8708282113075256, | |
| "learning_rate": 7.493897421832146e-05, | |
| "loss": 0.3510934066772461, | |
| "step": 37050 | |
| }, | |
| { | |
| "epoch": 0.7605576055760558, | |
| "grad_norm": 0.7575963735580444, | |
| "learning_rate": 7.490469007131103e-05, | |
| "loss": 0.3523604965209961, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 0.7615826158261583, | |
| "grad_norm": 0.872189462184906, | |
| "learning_rate": 7.487040592430061e-05, | |
| "loss": 0.3487224960327148, | |
| "step": 37150 | |
| }, | |
| { | |
| "epoch": 0.7626076260762608, | |
| "grad_norm": 0.7935099601745605, | |
| "learning_rate": 7.483612177729018e-05, | |
| "loss": 0.3471009063720703, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 0.7636326363263632, | |
| "grad_norm": 0.7949592471122742, | |
| "learning_rate": 7.480183763027976e-05, | |
| "loss": 0.3517554092407227, | |
| "step": 37250 | |
| }, | |
| { | |
| "epoch": 0.7646576465764657, | |
| "grad_norm": 0.8350103497505188, | |
| "learning_rate": 7.476755348326933e-05, | |
| "loss": 0.35215656280517577, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 0.7656826568265682, | |
| "grad_norm": 0.7591427564620972, | |
| "learning_rate": 7.473326933625892e-05, | |
| "loss": 0.3487773132324219, | |
| "step": 37350 | |
| }, | |
| { | |
| "epoch": 0.7667076670766708, | |
| "grad_norm": 0.9126777052879333, | |
| "learning_rate": 7.46989851892485e-05, | |
| "loss": 0.35323802947998045, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 0.7677326773267733, | |
| "grad_norm": 0.8218392729759216, | |
| "learning_rate": 7.466470104223807e-05, | |
| "loss": 0.3453704833984375, | |
| "step": 37450 | |
| }, | |
| { | |
| "epoch": 0.7687576875768758, | |
| "grad_norm": 0.8318965435028076, | |
| "learning_rate": 7.463041689522764e-05, | |
| "loss": 0.3498828887939453, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.7697826978269783, | |
| "grad_norm": 0.7620861530303955, | |
| "learning_rate": 7.459613274821723e-05, | |
| "loss": 0.34893093109130857, | |
| "step": 37550 | |
| }, | |
| { | |
| "epoch": 0.7708077080770808, | |
| "grad_norm": 0.8421209454536438, | |
| "learning_rate": 7.456184860120681e-05, | |
| "loss": 0.3458150863647461, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 0.7718327183271833, | |
| "grad_norm": 0.8842194080352783, | |
| "learning_rate": 7.452756445419638e-05, | |
| "loss": 0.34812103271484374, | |
| "step": 37650 | |
| }, | |
| { | |
| "epoch": 0.7728577285772857, | |
| "grad_norm": 0.7631191611289978, | |
| "learning_rate": 7.449328030718597e-05, | |
| "loss": 0.3484158706665039, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 0.7738827388273882, | |
| "grad_norm": 0.9068435430526733, | |
| "learning_rate": 7.445899616017555e-05, | |
| "loss": 0.34199817657470705, | |
| "step": 37750 | |
| }, | |
| { | |
| "epoch": 0.7749077490774908, | |
| "grad_norm": 0.9269093871116638, | |
| "learning_rate": 7.442471201316512e-05, | |
| "loss": 0.3446169662475586, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 0.7759327593275933, | |
| "grad_norm": 0.8941023349761963, | |
| "learning_rate": 7.43904278661547e-05, | |
| "loss": 0.3502924728393555, | |
| "step": 37850 | |
| }, | |
| { | |
| "epoch": 0.7769577695776958, | |
| "grad_norm": 0.8585652709007263, | |
| "learning_rate": 7.435614371914427e-05, | |
| "loss": 0.3490323257446289, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 0.7779827798277983, | |
| "grad_norm": 0.7989552617073059, | |
| "learning_rate": 7.432185957213384e-05, | |
| "loss": 0.34680171966552736, | |
| "step": 37950 | |
| }, | |
| { | |
| "epoch": 0.7790077900779008, | |
| "grad_norm": 0.8973004221916199, | |
| "learning_rate": 7.428757542512342e-05, | |
| "loss": 0.3500716400146484, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.7800328003280033, | |
| "grad_norm": 0.8058493137359619, | |
| "learning_rate": 7.425329127811301e-05, | |
| "loss": 0.34310436248779297, | |
| "step": 38050 | |
| }, | |
| { | |
| "epoch": 0.7810578105781057, | |
| "grad_norm": 0.8422321677207947, | |
| "learning_rate": 7.421900713110258e-05, | |
| "loss": 0.34766632080078125, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 0.7820828208282082, | |
| "grad_norm": 0.7820338010787964, | |
| "learning_rate": 7.418472298409216e-05, | |
| "loss": 0.3461764144897461, | |
| "step": 38150 | |
| }, | |
| { | |
| "epoch": 0.7831078310783108, | |
| "grad_norm": 0.764457643032074, | |
| "learning_rate": 7.415043883708173e-05, | |
| "loss": 0.3451896286010742, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 0.7841328413284133, | |
| "grad_norm": 0.8611803650856018, | |
| "learning_rate": 7.411615469007132e-05, | |
| "loss": 0.3440092086791992, | |
| "step": 38250 | |
| }, | |
| { | |
| "epoch": 0.7851578515785158, | |
| "grad_norm": 0.7152115106582642, | |
| "learning_rate": 7.40818705430609e-05, | |
| "loss": 0.33897369384765624, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 0.7861828618286183, | |
| "grad_norm": 0.8746068477630615, | |
| "learning_rate": 7.404758639605047e-05, | |
| "loss": 0.34935367584228516, | |
| "step": 38350 | |
| }, | |
| { | |
| "epoch": 0.7872078720787208, | |
| "grad_norm": 0.7815963625907898, | |
| "learning_rate": 7.401330224904006e-05, | |
| "loss": 0.34895614624023436, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 0.7882328823288233, | |
| "grad_norm": 0.7449670433998108, | |
| "learning_rate": 7.397901810202963e-05, | |
| "loss": 0.35156856536865233, | |
| "step": 38450 | |
| }, | |
| { | |
| "epoch": 0.7892578925789258, | |
| "grad_norm": 0.7672573924064636, | |
| "learning_rate": 7.39447339550192e-05, | |
| "loss": 0.34383018493652345, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.7902829028290282, | |
| "grad_norm": 0.814353883266449, | |
| "learning_rate": 7.391044980800878e-05, | |
| "loss": 0.35046504974365233, | |
| "step": 38550 | |
| }, | |
| { | |
| "epoch": 0.7913079130791308, | |
| "grad_norm": 0.7673587203025818, | |
| "learning_rate": 7.387616566099836e-05, | |
| "loss": 0.3459416580200195, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 0.7923329233292333, | |
| "grad_norm": 0.8047495484352112, | |
| "learning_rate": 7.384188151398793e-05, | |
| "loss": 0.3412811279296875, | |
| "step": 38650 | |
| }, | |
| { | |
| "epoch": 0.7933579335793358, | |
| "grad_norm": 0.8241590261459351, | |
| "learning_rate": 7.38075973669775e-05, | |
| "loss": 0.3456698226928711, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 0.7943829438294383, | |
| "grad_norm": 0.8381426334381104, | |
| "learning_rate": 7.37733132199671e-05, | |
| "loss": 0.3491682434082031, | |
| "step": 38750 | |
| }, | |
| { | |
| "epoch": 0.7954079540795408, | |
| "grad_norm": 0.8259062767028809, | |
| "learning_rate": 7.373902907295667e-05, | |
| "loss": 0.35304359436035154, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 0.7964329643296433, | |
| "grad_norm": 0.8970834016799927, | |
| "learning_rate": 7.370474492594624e-05, | |
| "loss": 0.3441598892211914, | |
| "step": 38850 | |
| }, | |
| { | |
| "epoch": 0.7974579745797458, | |
| "grad_norm": 0.8075913190841675, | |
| "learning_rate": 7.367046077893582e-05, | |
| "loss": 0.34804004669189453, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 0.7984829848298483, | |
| "grad_norm": 0.7509746551513672, | |
| "learning_rate": 7.36361766319254e-05, | |
| "loss": 0.35109031677246094, | |
| "step": 38950 | |
| }, | |
| { | |
| "epoch": 0.7995079950799509, | |
| "grad_norm": 0.7941098809242249, | |
| "learning_rate": 7.360189248491498e-05, | |
| "loss": 0.3441288375854492, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.8005330053300533, | |
| "grad_norm": 0.7938528656959534, | |
| "learning_rate": 7.356760833790456e-05, | |
| "loss": 0.34931072235107424, | |
| "step": 39050 | |
| }, | |
| { | |
| "epoch": 0.8015580155801558, | |
| "grad_norm": 0.736039400100708, | |
| "learning_rate": 7.353332419089414e-05, | |
| "loss": 0.34772510528564454, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 0.8025830258302583, | |
| "grad_norm": 0.7714667320251465, | |
| "learning_rate": 7.349904004388372e-05, | |
| "loss": 0.34220405578613283, | |
| "step": 39150 | |
| }, | |
| { | |
| "epoch": 0.8036080360803608, | |
| "grad_norm": 0.7959954738616943, | |
| "learning_rate": 7.346475589687329e-05, | |
| "loss": 0.34389007568359375, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 0.8046330463304633, | |
| "grad_norm": 0.7825502753257751, | |
| "learning_rate": 7.343047174986287e-05, | |
| "loss": 0.3540479278564453, | |
| "step": 39250 | |
| }, | |
| { | |
| "epoch": 0.8056580565805658, | |
| "grad_norm": 0.7958180904388428, | |
| "learning_rate": 7.339618760285244e-05, | |
| "loss": 0.3500849914550781, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 0.8066830668306683, | |
| "grad_norm": 0.9446352124214172, | |
| "learning_rate": 7.336190345584202e-05, | |
| "loss": 0.3434780120849609, | |
| "step": 39350 | |
| }, | |
| { | |
| "epoch": 0.8077080770807709, | |
| "grad_norm": 0.7788706421852112, | |
| "learning_rate": 7.332761930883159e-05, | |
| "loss": 0.35138893127441406, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 0.8087330873308733, | |
| "grad_norm": 0.8599708676338196, | |
| "learning_rate": 7.329333516182118e-05, | |
| "loss": 0.3533333206176758, | |
| "step": 39450 | |
| }, | |
| { | |
| "epoch": 0.8097580975809758, | |
| "grad_norm": 0.8119699954986572, | |
| "learning_rate": 7.325905101481075e-05, | |
| "loss": 0.3414528274536133, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.8107831078310783, | |
| "grad_norm": 0.7969697713851929, | |
| "learning_rate": 7.322476686780033e-05, | |
| "loss": 0.34422290802001954, | |
| "step": 39550 | |
| }, | |
| { | |
| "epoch": 0.8118081180811808, | |
| "grad_norm": 0.8129152655601501, | |
| "learning_rate": 7.31904827207899e-05, | |
| "loss": 0.3414501953125, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 0.8128331283312833, | |
| "grad_norm": 0.7618581056594849, | |
| "learning_rate": 7.315619857377949e-05, | |
| "loss": 0.34989917755126954, | |
| "step": 39650 | |
| }, | |
| { | |
| "epoch": 0.8138581385813858, | |
| "grad_norm": 0.8868136405944824, | |
| "learning_rate": 7.312191442676907e-05, | |
| "loss": 0.3422850036621094, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 0.8148831488314883, | |
| "grad_norm": 0.7645081877708435, | |
| "learning_rate": 7.308763027975864e-05, | |
| "loss": 0.3426713180541992, | |
| "step": 39750 | |
| }, | |
| { | |
| "epoch": 0.8159081590815909, | |
| "grad_norm": 0.8448659777641296, | |
| "learning_rate": 7.305334613274823e-05, | |
| "loss": 0.34805381774902344, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 0.8169331693316934, | |
| "grad_norm": 0.8921565413475037, | |
| "learning_rate": 7.30190619857378e-05, | |
| "loss": 0.34674064636230467, | |
| "step": 39850 | |
| }, | |
| { | |
| "epoch": 0.8179581795817958, | |
| "grad_norm": 0.828484058380127, | |
| "learning_rate": 7.298477783872738e-05, | |
| "loss": 0.3486968231201172, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 0.8189831898318983, | |
| "grad_norm": 0.8645332455635071, | |
| "learning_rate": 7.295049369171695e-05, | |
| "loss": 0.34676685333251955, | |
| "step": 39950 | |
| }, | |
| { | |
| "epoch": 0.8200082000820008, | |
| "grad_norm": 0.8277470469474792, | |
| "learning_rate": 7.291620954470653e-05, | |
| "loss": 0.3453370666503906, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.8200082000820008, | |
| "eval_loss": 0.41712766885757446, | |
| "eval_runtime": 4.4046, | |
| "eval_samples_per_second": 5675.918, | |
| "eval_steps_per_second": 5.676, | |
| "eval_teacher_cosine_loss": 0.26259028911590576, | |
| "eval_teacher_cosine_mean": 0.7374097108840942, | |
| "eval_teacher_cosine_p10": 0.6600194573402405, | |
| "eval_teacher_cosine_p50": 0.7462900876998901, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.8210332103321033, | |
| "grad_norm": 0.8038929104804993, | |
| "learning_rate": 7.28819253976961e-05, | |
| "loss": 0.34076431274414065, | |
| "step": 40050 | |
| }, | |
| { | |
| "epoch": 0.8220582205822058, | |
| "grad_norm": 0.9361200332641602, | |
| "learning_rate": 7.284764125068568e-05, | |
| "loss": 0.3482931518554688, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 0.8230832308323083, | |
| "grad_norm": 0.7499080896377563, | |
| "learning_rate": 7.281335710367527e-05, | |
| "loss": 0.34473251342773437, | |
| "step": 40150 | |
| }, | |
| { | |
| "epoch": 0.8241082410824109, | |
| "grad_norm": 0.7993378043174744, | |
| "learning_rate": 7.277907295666484e-05, | |
| "loss": 0.3444669723510742, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 0.8251332513325134, | |
| "grad_norm": 0.7516816854476929, | |
| "learning_rate": 7.274478880965442e-05, | |
| "loss": 0.3476390838623047, | |
| "step": 40250 | |
| }, | |
| { | |
| "epoch": 0.8261582615826158, | |
| "grad_norm": 0.792417585849762, | |
| "learning_rate": 7.271050466264399e-05, | |
| "loss": 0.3489529037475586, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 0.8271832718327183, | |
| "grad_norm": 0.8179230093955994, | |
| "learning_rate": 7.267622051563358e-05, | |
| "loss": 0.3541193389892578, | |
| "step": 40350 | |
| }, | |
| { | |
| "epoch": 0.8282082820828208, | |
| "grad_norm": 0.7739876508712769, | |
| "learning_rate": 7.264193636862315e-05, | |
| "loss": 0.3437077713012695, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 0.8292332923329233, | |
| "grad_norm": 0.8429916501045227, | |
| "learning_rate": 7.260765222161273e-05, | |
| "loss": 0.3403946304321289, | |
| "step": 40450 | |
| }, | |
| { | |
| "epoch": 0.8302583025830258, | |
| "grad_norm": 0.8695170283317566, | |
| "learning_rate": 7.257336807460232e-05, | |
| "loss": 0.35067737579345704, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.8312833128331283, | |
| "grad_norm": 0.8550466299057007, | |
| "learning_rate": 7.253908392759189e-05, | |
| "loss": 0.341976203918457, | |
| "step": 40550 | |
| }, | |
| { | |
| "epoch": 0.8323083230832309, | |
| "grad_norm": 0.7927470803260803, | |
| "learning_rate": 7.250479978058147e-05, | |
| "loss": 0.347991943359375, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.8122792840003967, | |
| "learning_rate": 7.247051563357104e-05, | |
| "loss": 0.34443653106689454, | |
| "step": 40650 | |
| }, | |
| { | |
| "epoch": 0.8343583435834359, | |
| "grad_norm": 0.8034294247627258, | |
| "learning_rate": 7.243623148656062e-05, | |
| "loss": 0.34653743743896487, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 0.8353833538335383, | |
| "grad_norm": 0.7267901301383972, | |
| "learning_rate": 7.240194733955019e-05, | |
| "loss": 0.35085575103759764, | |
| "step": 40750 | |
| }, | |
| { | |
| "epoch": 0.8364083640836408, | |
| "grad_norm": 0.752204179763794, | |
| "learning_rate": 7.236766319253976e-05, | |
| "loss": 0.3495679473876953, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 0.8374333743337433, | |
| "grad_norm": 0.8167332410812378, | |
| "learning_rate": 7.233337904552935e-05, | |
| "loss": 0.3441920852661133, | |
| "step": 40850 | |
| }, | |
| { | |
| "epoch": 0.8384583845838458, | |
| "grad_norm": 0.8485610485076904, | |
| "learning_rate": 7.229909489851893e-05, | |
| "loss": 0.3445806121826172, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 0.8394833948339483, | |
| "grad_norm": 0.7576408386230469, | |
| "learning_rate": 7.22648107515085e-05, | |
| "loss": 0.3391580581665039, | |
| "step": 40950 | |
| }, | |
| { | |
| "epoch": 0.8405084050840509, | |
| "grad_norm": 0.7663058638572693, | |
| "learning_rate": 7.223052660449808e-05, | |
| "loss": 0.3444937133789063, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.8415334153341534, | |
| "grad_norm": 0.8309179544448853, | |
| "learning_rate": 7.219624245748767e-05, | |
| "loss": 0.3439421844482422, | |
| "step": 41050 | |
| }, | |
| { | |
| "epoch": 0.8425584255842559, | |
| "grad_norm": 0.9276794791221619, | |
| "learning_rate": 7.216195831047724e-05, | |
| "loss": 0.3422798538208008, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 0.8435834358343584, | |
| "grad_norm": 0.875667154788971, | |
| "learning_rate": 7.212767416346681e-05, | |
| "loss": 0.3367802810668945, | |
| "step": 41150 | |
| }, | |
| { | |
| "epoch": 0.8446084460844608, | |
| "grad_norm": 0.8705291152000427, | |
| "learning_rate": 7.20933900164564e-05, | |
| "loss": 0.3491183090209961, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 0.8456334563345633, | |
| "grad_norm": 0.9101588129997253, | |
| "learning_rate": 7.205910586944598e-05, | |
| "loss": 0.34627834320068357, | |
| "step": 41250 | |
| }, | |
| { | |
| "epoch": 0.8466584665846658, | |
| "grad_norm": 0.7876223921775818, | |
| "learning_rate": 7.202482172243555e-05, | |
| "loss": 0.3485475158691406, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 0.8476834768347683, | |
| "grad_norm": 0.7891747951507568, | |
| "learning_rate": 7.199053757542513e-05, | |
| "loss": 0.3373814392089844, | |
| "step": 41350 | |
| }, | |
| { | |
| "epoch": 0.8487084870848709, | |
| "grad_norm": 0.8372437357902527, | |
| "learning_rate": 7.19562534284147e-05, | |
| "loss": 0.34700016021728514, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 0.8497334973349734, | |
| "grad_norm": 0.8303970694541931, | |
| "learning_rate": 7.192196928140428e-05, | |
| "loss": 0.34442615509033203, | |
| "step": 41450 | |
| }, | |
| { | |
| "epoch": 0.8507585075850759, | |
| "grad_norm": 0.855299711227417, | |
| "learning_rate": 7.188768513439385e-05, | |
| "loss": 0.34732799530029296, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.8517835178351784, | |
| "grad_norm": 0.7483532428741455, | |
| "learning_rate": 7.185340098738344e-05, | |
| "loss": 0.3414728927612305, | |
| "step": 41550 | |
| }, | |
| { | |
| "epoch": 0.8528085280852808, | |
| "grad_norm": 0.8602964282035828, | |
| "learning_rate": 7.181911684037301e-05, | |
| "loss": 0.3432622528076172, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 0.8538335383353833, | |
| "grad_norm": 0.8208170533180237, | |
| "learning_rate": 7.178483269336259e-05, | |
| "loss": 0.3455703353881836, | |
| "step": 41650 | |
| }, | |
| { | |
| "epoch": 0.8548585485854858, | |
| "grad_norm": 0.8216872811317444, | |
| "learning_rate": 7.175054854635216e-05, | |
| "loss": 0.3389241027832031, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 0.8558835588355883, | |
| "grad_norm": 0.7815793752670288, | |
| "learning_rate": 7.171626439934175e-05, | |
| "loss": 0.3391556167602539, | |
| "step": 41750 | |
| }, | |
| { | |
| "epoch": 0.8569085690856909, | |
| "grad_norm": 0.8208360075950623, | |
| "learning_rate": 7.168198025233133e-05, | |
| "loss": 0.34248348236083986, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 0.8579335793357934, | |
| "grad_norm": 0.7469902634620667, | |
| "learning_rate": 7.16476961053209e-05, | |
| "loss": 0.3409922790527344, | |
| "step": 41850 | |
| }, | |
| { | |
| "epoch": 0.8589585895858959, | |
| "grad_norm": 0.7633910179138184, | |
| "learning_rate": 7.161341195831049e-05, | |
| "loss": 0.339487190246582, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 0.8599835998359984, | |
| "grad_norm": 0.8420964479446411, | |
| "learning_rate": 7.157912781130006e-05, | |
| "loss": 0.34453960418701174, | |
| "step": 41950 | |
| }, | |
| { | |
| "epoch": 0.8610086100861009, | |
| "grad_norm": 0.7952750325202942, | |
| "learning_rate": 7.154484366428964e-05, | |
| "loss": 0.3403425598144531, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.8620336203362033, | |
| "grad_norm": 0.8091248273849487, | |
| "learning_rate": 7.151055951727921e-05, | |
| "loss": 0.3415816116333008, | |
| "step": 42050 | |
| }, | |
| { | |
| "epoch": 0.8630586305863058, | |
| "grad_norm": 0.8487741947174072, | |
| "learning_rate": 7.147627537026879e-05, | |
| "loss": 0.3439748764038086, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 0.8640836408364083, | |
| "grad_norm": 0.8201131820678711, | |
| "learning_rate": 7.144199122325836e-05, | |
| "loss": 0.345859375, | |
| "step": 42150 | |
| }, | |
| { | |
| "epoch": 0.8651086510865109, | |
| "grad_norm": 0.8775553703308105, | |
| "learning_rate": 7.140770707624794e-05, | |
| "loss": 0.3436273956298828, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 0.8661336613366134, | |
| "grad_norm": 0.8008050322532654, | |
| "learning_rate": 7.137342292923753e-05, | |
| "loss": 0.3514908981323242, | |
| "step": 42250 | |
| }, | |
| { | |
| "epoch": 0.8671586715867159, | |
| "grad_norm": 0.8790284395217896, | |
| "learning_rate": 7.13391387822271e-05, | |
| "loss": 0.3442119216918945, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 0.8681836818368184, | |
| "grad_norm": 0.8865100145339966, | |
| "learning_rate": 7.130485463521667e-05, | |
| "loss": 0.34529834747314453, | |
| "step": 42350 | |
| }, | |
| { | |
| "epoch": 0.8692086920869209, | |
| "grad_norm": 0.9767987132072449, | |
| "learning_rate": 7.127057048820625e-05, | |
| "loss": 0.34931221008300783, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 0.8702337023370234, | |
| "grad_norm": 0.805337131023407, | |
| "learning_rate": 7.123628634119584e-05, | |
| "loss": 0.3474878311157227, | |
| "step": 42450 | |
| }, | |
| { | |
| "epoch": 0.8712587125871258, | |
| "grad_norm": 0.8012571930885315, | |
| "learning_rate": 7.120200219418541e-05, | |
| "loss": 0.34498981475830076, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.8722837228372283, | |
| "grad_norm": 0.7793713808059692, | |
| "learning_rate": 7.116771804717499e-05, | |
| "loss": 0.34025714874267576, | |
| "step": 42550 | |
| }, | |
| { | |
| "epoch": 0.8733087330873309, | |
| "grad_norm": 0.8102912306785583, | |
| "learning_rate": 7.113343390016458e-05, | |
| "loss": 0.34836277008056643, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 0.8743337433374334, | |
| "grad_norm": 0.8289583325386047, | |
| "learning_rate": 7.109914975315415e-05, | |
| "loss": 0.343720588684082, | |
| "step": 42650 | |
| }, | |
| { | |
| "epoch": 0.8753587535875359, | |
| "grad_norm": 0.7556430101394653, | |
| "learning_rate": 7.106486560614372e-05, | |
| "loss": 0.34643470764160156, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 0.8763837638376384, | |
| "grad_norm": 0.8377882242202759, | |
| "learning_rate": 7.10305814591333e-05, | |
| "loss": 0.3493313980102539, | |
| "step": 42750 | |
| }, | |
| { | |
| "epoch": 0.8774087740877409, | |
| "grad_norm": 0.8487588167190552, | |
| "learning_rate": 7.099629731212287e-05, | |
| "loss": 0.33952983856201174, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 0.8784337843378434, | |
| "grad_norm": 0.8280442357063293, | |
| "learning_rate": 7.096201316511245e-05, | |
| "loss": 0.34414817810058596, | |
| "step": 42850 | |
| }, | |
| { | |
| "epoch": 0.8794587945879458, | |
| "grad_norm": 0.8658256530761719, | |
| "learning_rate": 7.092772901810202e-05, | |
| "loss": 0.3405352783203125, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 0.8804838048380483, | |
| "grad_norm": 0.7546058297157288, | |
| "learning_rate": 7.089344487109161e-05, | |
| "loss": 0.3408969497680664, | |
| "step": 42950 | |
| }, | |
| { | |
| "epoch": 0.8815088150881509, | |
| "grad_norm": 0.8004677295684814, | |
| "learning_rate": 7.085916072408119e-05, | |
| "loss": 0.3468672561645508, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.8825338253382534, | |
| "grad_norm": 0.8871493339538574, | |
| "learning_rate": 7.082487657707076e-05, | |
| "loss": 0.3443138885498047, | |
| "step": 43050 | |
| }, | |
| { | |
| "epoch": 0.8835588355883559, | |
| "grad_norm": 0.812741219997406, | |
| "learning_rate": 7.079059243006035e-05, | |
| "loss": 0.3491249084472656, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 0.8845838458384584, | |
| "grad_norm": 0.8690420389175415, | |
| "learning_rate": 7.075630828304992e-05, | |
| "loss": 0.34076526641845706, | |
| "step": 43150 | |
| }, | |
| { | |
| "epoch": 0.8856088560885609, | |
| "grad_norm": 0.7516093254089355, | |
| "learning_rate": 7.07220241360395e-05, | |
| "loss": 0.3397038269042969, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 0.8866338663386634, | |
| "grad_norm": 0.7591049075126648, | |
| "learning_rate": 7.068773998902907e-05, | |
| "loss": 0.34129810333251953, | |
| "step": 43250 | |
| }, | |
| { | |
| "epoch": 0.8876588765887659, | |
| "grad_norm": 0.8539579510688782, | |
| "learning_rate": 7.065345584201866e-05, | |
| "loss": 0.3465434646606445, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 0.8886838868388683, | |
| "grad_norm": 0.7801506519317627, | |
| "learning_rate": 7.061917169500824e-05, | |
| "loss": 0.33541633605957033, | |
| "step": 43350 | |
| }, | |
| { | |
| "epoch": 0.8897088970889709, | |
| "grad_norm": 0.8458453416824341, | |
| "learning_rate": 7.058488754799781e-05, | |
| "loss": 0.3383928298950195, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 0.8907339073390734, | |
| "grad_norm": 0.8315999507904053, | |
| "learning_rate": 7.055060340098739e-05, | |
| "loss": 0.3435313034057617, | |
| "step": 43450 | |
| }, | |
| { | |
| "epoch": 0.8917589175891759, | |
| "grad_norm": 0.7546735405921936, | |
| "learning_rate": 7.051631925397696e-05, | |
| "loss": 0.3429303359985352, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.8927839278392784, | |
| "grad_norm": 0.8558475971221924, | |
| "learning_rate": 7.048203510696654e-05, | |
| "loss": 0.34082763671875, | |
| "step": 43550 | |
| }, | |
| { | |
| "epoch": 0.8938089380893809, | |
| "grad_norm": 0.8185769319534302, | |
| "learning_rate": 7.044775095995611e-05, | |
| "loss": 0.3431754684448242, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 0.8948339483394834, | |
| "grad_norm": 0.7818198800086975, | |
| "learning_rate": 7.04134668129457e-05, | |
| "loss": 0.34071399688720705, | |
| "step": 43650 | |
| }, | |
| { | |
| "epoch": 0.8958589585895859, | |
| "grad_norm": 0.7697409987449646, | |
| "learning_rate": 7.037918266593527e-05, | |
| "loss": 0.3403765869140625, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 0.8968839688396884, | |
| "grad_norm": 0.9100814461708069, | |
| "learning_rate": 7.034489851892485e-05, | |
| "loss": 0.34470596313476565, | |
| "step": 43750 | |
| }, | |
| { | |
| "epoch": 0.897908979089791, | |
| "grad_norm": 0.8283818960189819, | |
| "learning_rate": 7.031061437191444e-05, | |
| "loss": 0.3400896453857422, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 0.8989339893398934, | |
| "grad_norm": 0.8208755254745483, | |
| "learning_rate": 7.027633022490401e-05, | |
| "loss": 0.33513545989990234, | |
| "step": 43850 | |
| }, | |
| { | |
| "epoch": 0.8999589995899959, | |
| "grad_norm": 0.8187302947044373, | |
| "learning_rate": 7.024204607789359e-05, | |
| "loss": 0.34179447174072264, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 0.9009840098400984, | |
| "grad_norm": 0.8155043721199036, | |
| "learning_rate": 7.020776193088316e-05, | |
| "loss": 0.3377865982055664, | |
| "step": 43950 | |
| }, | |
| { | |
| "epoch": 0.9020090200902009, | |
| "grad_norm": 0.7414567470550537, | |
| "learning_rate": 7.017347778387275e-05, | |
| "loss": 0.34257026672363283, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.9030340303403034, | |
| "grad_norm": 0.8276044130325317, | |
| "learning_rate": 7.013919363686232e-05, | |
| "loss": 0.3399489212036133, | |
| "step": 44050 | |
| }, | |
| { | |
| "epoch": 0.9040590405904059, | |
| "grad_norm": 0.8294392824172974, | |
| "learning_rate": 7.01049094898519e-05, | |
| "loss": 0.3398480224609375, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 0.9050840508405084, | |
| "grad_norm": 0.8833320736885071, | |
| "learning_rate": 7.007062534284147e-05, | |
| "loss": 0.3415219497680664, | |
| "step": 44150 | |
| }, | |
| { | |
| "epoch": 0.906109061090611, | |
| "grad_norm": 0.8286173939704895, | |
| "learning_rate": 7.003634119583105e-05, | |
| "loss": 0.3419487762451172, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 0.9071340713407134, | |
| "grad_norm": 0.8375602960586548, | |
| "learning_rate": 7.000205704882062e-05, | |
| "loss": 0.34328163146972657, | |
| "step": 44250 | |
| }, | |
| { | |
| "epoch": 0.9081590815908159, | |
| "grad_norm": 0.7297822833061218, | |
| "learning_rate": 6.99677729018102e-05, | |
| "loss": 0.33917510986328125, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 0.9091840918409184, | |
| "grad_norm": 0.7560060024261475, | |
| "learning_rate": 6.993348875479978e-05, | |
| "loss": 0.33717903137207034, | |
| "step": 44350 | |
| }, | |
| { | |
| "epoch": 0.9102091020910209, | |
| "grad_norm": 0.8967326283454895, | |
| "learning_rate": 6.989920460778936e-05, | |
| "loss": 0.3421474838256836, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 0.9112341123411234, | |
| "grad_norm": 0.8593087196350098, | |
| "learning_rate": 6.986492046077893e-05, | |
| "loss": 0.34114315032958986, | |
| "step": 44450 | |
| }, | |
| { | |
| "epoch": 0.9122591225912259, | |
| "grad_norm": 0.861524760723114, | |
| "learning_rate": 6.983063631376852e-05, | |
| "loss": 0.34765380859375, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.9132841328413284, | |
| "grad_norm": 0.8305798172950745, | |
| "learning_rate": 6.97963521667581e-05, | |
| "loss": 0.34759834289550784, | |
| "step": 44550 | |
| }, | |
| { | |
| "epoch": 0.914309143091431, | |
| "grad_norm": 0.8140143752098083, | |
| "learning_rate": 6.976206801974767e-05, | |
| "loss": 0.3470936965942383, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 0.9153341533415335, | |
| "grad_norm": 0.7895699739456177, | |
| "learning_rate": 6.972778387273725e-05, | |
| "loss": 0.3395873260498047, | |
| "step": 44650 | |
| }, | |
| { | |
| "epoch": 0.9163591635916359, | |
| "grad_norm": 0.8499590754508972, | |
| "learning_rate": 6.969349972572683e-05, | |
| "loss": 0.34457645416259763, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 0.9173841738417384, | |
| "grad_norm": 0.7332067489624023, | |
| "learning_rate": 6.965921557871641e-05, | |
| "loss": 0.3384964370727539, | |
| "step": 44750 | |
| }, | |
| { | |
| "epoch": 0.9184091840918409, | |
| "grad_norm": 0.9319055080413818, | |
| "learning_rate": 6.962493143170598e-05, | |
| "loss": 0.34340629577636717, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 0.9194341943419434, | |
| "grad_norm": 0.7451433539390564, | |
| "learning_rate": 6.959064728469556e-05, | |
| "loss": 0.3443481063842773, | |
| "step": 44850 | |
| }, | |
| { | |
| "epoch": 0.9204592045920459, | |
| "grad_norm": 0.850803792476654, | |
| "learning_rate": 6.955636313768515e-05, | |
| "loss": 0.33997459411621095, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 0.9214842148421484, | |
| "grad_norm": 0.7563380599021912, | |
| "learning_rate": 6.952207899067471e-05, | |
| "loss": 0.3377606964111328, | |
| "step": 44950 | |
| }, | |
| { | |
| "epoch": 0.922509225092251, | |
| "grad_norm": 0.8730197548866272, | |
| "learning_rate": 6.948779484366428e-05, | |
| "loss": 0.3412102508544922, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.922509225092251, | |
| "eval_loss": 0.41288453340530396, | |
| "eval_runtime": 4.4229, | |
| "eval_samples_per_second": 5652.452, | |
| "eval_steps_per_second": 5.652, | |
| "eval_teacher_cosine_loss": 0.2582279443740845, | |
| "eval_teacher_cosine_mean": 0.7417720556259155, | |
| "eval_teacher_cosine_p10": 0.6647007465362549, | |
| "eval_teacher_cosine_p50": 0.7507580518722534, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.9235342353423535, | |
| "grad_norm": 0.8670917749404907, | |
| "learning_rate": 6.945351069665387e-05, | |
| "loss": 0.34338233947753904, | |
| "step": 45050 | |
| }, | |
| { | |
| "epoch": 0.924559245592456, | |
| "grad_norm": 0.8930084705352783, | |
| "learning_rate": 6.941922654964345e-05, | |
| "loss": 0.3400748825073242, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 0.9255842558425584, | |
| "grad_norm": 0.7936744689941406, | |
| "learning_rate": 6.938494240263302e-05, | |
| "loss": 0.3428042221069336, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 0.9266092660926609, | |
| "grad_norm": 0.7987287640571594, | |
| "learning_rate": 6.935065825562261e-05, | |
| "loss": 0.3390156173706055, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 0.9276342763427634, | |
| "grad_norm": 0.8535095453262329, | |
| "learning_rate": 6.931637410861218e-05, | |
| "loss": 0.34122440338134763, | |
| "step": 45250 | |
| }, | |
| { | |
| "epoch": 0.9286592865928659, | |
| "grad_norm": 0.8350134491920471, | |
| "learning_rate": 6.928208996160176e-05, | |
| "loss": 0.3407657241821289, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 0.9296842968429684, | |
| "grad_norm": 0.7424451112747192, | |
| "learning_rate": 6.924780581459133e-05, | |
| "loss": 0.3390554428100586, | |
| "step": 45350 | |
| }, | |
| { | |
| "epoch": 0.930709307093071, | |
| "grad_norm": 0.8034331798553467, | |
| "learning_rate": 6.921352166758092e-05, | |
| "loss": 0.3391274642944336, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 0.9317343173431735, | |
| "grad_norm": 0.8002919554710388, | |
| "learning_rate": 6.91792375205705e-05, | |
| "loss": 0.34332645416259766, | |
| "step": 45450 | |
| }, | |
| { | |
| "epoch": 0.932759327593276, | |
| "grad_norm": 0.8195891380310059, | |
| "learning_rate": 6.914495337356007e-05, | |
| "loss": 0.3412936782836914, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.9337843378433784, | |
| "grad_norm": 0.8300272226333618, | |
| "learning_rate": 6.911066922654965e-05, | |
| "loss": 0.34317604064941404, | |
| "step": 45550 | |
| }, | |
| { | |
| "epoch": 0.9348093480934809, | |
| "grad_norm": 0.8461131453514099, | |
| "learning_rate": 6.907638507953923e-05, | |
| "loss": 0.3414227294921875, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 0.9358343583435834, | |
| "grad_norm": 0.8538811206817627, | |
| "learning_rate": 6.90421009325288e-05, | |
| "loss": 0.34242958068847656, | |
| "step": 45650 | |
| }, | |
| { | |
| "epoch": 0.9368593685936859, | |
| "grad_norm": 0.8598376512527466, | |
| "learning_rate": 6.900781678551837e-05, | |
| "loss": 0.34188133239746094, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 0.9378843788437884, | |
| "grad_norm": 0.8520790934562683, | |
| "learning_rate": 6.897353263850796e-05, | |
| "loss": 0.33959362030029294, | |
| "step": 45750 | |
| }, | |
| { | |
| "epoch": 0.938909389093891, | |
| "grad_norm": 0.7374169230461121, | |
| "learning_rate": 6.893924849149753e-05, | |
| "loss": 0.34431224822998047, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 0.9399343993439935, | |
| "grad_norm": 0.7990703582763672, | |
| "learning_rate": 6.890496434448711e-05, | |
| "loss": 0.34029354095458986, | |
| "step": 45850 | |
| }, | |
| { | |
| "epoch": 0.940959409594096, | |
| "grad_norm": 0.7686148881912231, | |
| "learning_rate": 6.88706801974767e-05, | |
| "loss": 0.341511116027832, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 0.9419844198441985, | |
| "grad_norm": 0.88141268491745, | |
| "learning_rate": 6.883639605046627e-05, | |
| "loss": 0.3350189208984375, | |
| "step": 45950 | |
| }, | |
| { | |
| "epoch": 0.9430094300943009, | |
| "grad_norm": 0.8354436755180359, | |
| "learning_rate": 6.880211190345584e-05, | |
| "loss": 0.34227134704589846, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.9440344403444034, | |
| "grad_norm": 0.8428378701210022, | |
| "learning_rate": 6.876782775644542e-05, | |
| "loss": 0.34108432769775393, | |
| "step": 46050 | |
| }, | |
| { | |
| "epoch": 0.9450594505945059, | |
| "grad_norm": 0.7658049464225769, | |
| "learning_rate": 6.873354360943501e-05, | |
| "loss": 0.3356414031982422, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 0.9460844608446084, | |
| "grad_norm": 0.766192615032196, | |
| "learning_rate": 6.869925946242458e-05, | |
| "loss": 0.3383335113525391, | |
| "step": 46150 | |
| }, | |
| { | |
| "epoch": 0.947109471094711, | |
| "grad_norm": 0.865257978439331, | |
| "learning_rate": 6.866497531541416e-05, | |
| "loss": 0.3394303894042969, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 0.9481344813448135, | |
| "grad_norm": 0.8257219791412354, | |
| "learning_rate": 6.863069116840373e-05, | |
| "loss": 0.3415960693359375, | |
| "step": 46250 | |
| }, | |
| { | |
| "epoch": 0.949159491594916, | |
| "grad_norm": 0.7919724583625793, | |
| "learning_rate": 6.859640702139332e-05, | |
| "loss": 0.34070381164550784, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 0.9501845018450185, | |
| "grad_norm": 0.827800989151001, | |
| "learning_rate": 6.856212287438288e-05, | |
| "loss": 0.3399745178222656, | |
| "step": 46350 | |
| }, | |
| { | |
| "epoch": 0.951209512095121, | |
| "grad_norm": 0.7611798048019409, | |
| "learning_rate": 6.852783872737246e-05, | |
| "loss": 0.33984378814697264, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 0.9522345223452234, | |
| "grad_norm": 0.8684995174407959, | |
| "learning_rate": 6.849355458036204e-05, | |
| "loss": 0.33557121276855467, | |
| "step": 46450 | |
| }, | |
| { | |
| "epoch": 0.9532595325953259, | |
| "grad_norm": 0.8896166682243347, | |
| "learning_rate": 6.845927043335162e-05, | |
| "loss": 0.34928062438964846, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.9542845428454284, | |
| "grad_norm": 0.816207766532898, | |
| "learning_rate": 6.84249862863412e-05, | |
| "loss": 0.33950084686279297, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 0.955309553095531, | |
| "grad_norm": 0.8799117207527161, | |
| "learning_rate": 6.839070213933078e-05, | |
| "loss": 0.3416341400146484, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 0.9563345633456335, | |
| "grad_norm": 0.8032182455062866, | |
| "learning_rate": 6.835641799232036e-05, | |
| "loss": 0.34092498779296876, | |
| "step": 46650 | |
| }, | |
| { | |
| "epoch": 0.957359573595736, | |
| "grad_norm": 0.7889217734336853, | |
| "learning_rate": 6.832213384530993e-05, | |
| "loss": 0.3443967056274414, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 0.9583845838458385, | |
| "grad_norm": 0.7487931847572327, | |
| "learning_rate": 6.82878496982995e-05, | |
| "loss": 0.3397322463989258, | |
| "step": 46750 | |
| }, | |
| { | |
| "epoch": 0.959409594095941, | |
| "grad_norm": 0.829073965549469, | |
| "learning_rate": 6.82535655512891e-05, | |
| "loss": 0.34414398193359375, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 0.9604346043460434, | |
| "grad_norm": 0.8456899523735046, | |
| "learning_rate": 6.821928140427867e-05, | |
| "loss": 0.3397352600097656, | |
| "step": 46850 | |
| }, | |
| { | |
| "epoch": 0.9614596145961459, | |
| "grad_norm": 0.7813683748245239, | |
| "learning_rate": 6.818499725726824e-05, | |
| "loss": 0.337515754699707, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 0.9624846248462484, | |
| "grad_norm": 0.8231688737869263, | |
| "learning_rate": 6.815071311025782e-05, | |
| "loss": 0.3421532821655273, | |
| "step": 46950 | |
| }, | |
| { | |
| "epoch": 0.963509635096351, | |
| "grad_norm": 0.7784870266914368, | |
| "learning_rate": 6.81164289632474e-05, | |
| "loss": 0.3328588104248047, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.9645346453464535, | |
| "grad_norm": 0.8055475950241089, | |
| "learning_rate": 6.808214481623697e-05, | |
| "loss": 0.34021720886230467, | |
| "step": 47050 | |
| }, | |
| { | |
| "epoch": 0.965559655596556, | |
| "grad_norm": 0.8010567426681519, | |
| "learning_rate": 6.804786066922654e-05, | |
| "loss": 0.34084171295166016, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 0.9665846658466585, | |
| "grad_norm": 0.7855961918830872, | |
| "learning_rate": 6.801357652221613e-05, | |
| "loss": 0.3428207778930664, | |
| "step": 47150 | |
| }, | |
| { | |
| "epoch": 0.967609676096761, | |
| "grad_norm": 0.6983242630958557, | |
| "learning_rate": 6.79792923752057e-05, | |
| "loss": 0.3392998504638672, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 0.9686346863468634, | |
| "grad_norm": 0.8925724029541016, | |
| "learning_rate": 6.794500822819528e-05, | |
| "loss": 0.3436629104614258, | |
| "step": 47250 | |
| }, | |
| { | |
| "epoch": 0.9696596965969659, | |
| "grad_norm": 0.7431129217147827, | |
| "learning_rate": 6.791072408118487e-05, | |
| "loss": 0.3411537170410156, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 0.9706847068470684, | |
| "grad_norm": 0.8046679496765137, | |
| "learning_rate": 6.787643993417444e-05, | |
| "loss": 0.3391756057739258, | |
| "step": 47350 | |
| }, | |
| { | |
| "epoch": 0.971709717097171, | |
| "grad_norm": 0.7404874563217163, | |
| "learning_rate": 6.784215578716402e-05, | |
| "loss": 0.33607963562011717, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 0.9727347273472735, | |
| "grad_norm": 0.8749111294746399, | |
| "learning_rate": 6.780787164015359e-05, | |
| "loss": 0.3393808746337891, | |
| "step": 47450 | |
| }, | |
| { | |
| "epoch": 0.973759737597376, | |
| "grad_norm": 0.8132859468460083, | |
| "learning_rate": 6.777358749314318e-05, | |
| "loss": 0.34185691833496096, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.9747847478474785, | |
| "grad_norm": 0.855075478553772, | |
| "learning_rate": 6.773930334613276e-05, | |
| "loss": 0.34292003631591794, | |
| "step": 47550 | |
| }, | |
| { | |
| "epoch": 0.975809758097581, | |
| "grad_norm": 0.8510366678237915, | |
| "learning_rate": 6.770501919912233e-05, | |
| "loss": 0.3399077987670898, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 0.9768347683476835, | |
| "grad_norm": 0.7676845192909241, | |
| "learning_rate": 6.76707350521119e-05, | |
| "loss": 0.34026222229003905, | |
| "step": 47650 | |
| }, | |
| { | |
| "epoch": 0.977859778597786, | |
| "grad_norm": 0.8032896518707275, | |
| "learning_rate": 6.763645090510149e-05, | |
| "loss": 0.3417753982543945, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 0.9788847888478884, | |
| "grad_norm": 0.7645705938339233, | |
| "learning_rate": 6.760216675809105e-05, | |
| "loss": 0.3384263610839844, | |
| "step": 47750 | |
| }, | |
| { | |
| "epoch": 0.979909799097991, | |
| "grad_norm": 0.7977441549301147, | |
| "learning_rate": 6.756788261108063e-05, | |
| "loss": 0.3397220993041992, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 0.9809348093480935, | |
| "grad_norm": 0.8725268244743347, | |
| "learning_rate": 6.753359846407022e-05, | |
| "loss": 0.3451786804199219, | |
| "step": 47850 | |
| }, | |
| { | |
| "epoch": 0.981959819598196, | |
| "grad_norm": 0.8435919284820557, | |
| "learning_rate": 6.749931431705979e-05, | |
| "loss": 0.34383113861083986, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 0.9829848298482985, | |
| "grad_norm": 0.7496271133422852, | |
| "learning_rate": 6.746503017004937e-05, | |
| "loss": 0.34511505126953124, | |
| "step": 47950 | |
| }, | |
| { | |
| "epoch": 0.984009840098401, | |
| "grad_norm": 0.7936993837356567, | |
| "learning_rate": 6.743074602303895e-05, | |
| "loss": 0.34265598297119143, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.9850348503485035, | |
| "grad_norm": 0.8614635467529297, | |
| "learning_rate": 6.739646187602853e-05, | |
| "loss": 0.3446725082397461, | |
| "step": 48050 | |
| }, | |
| { | |
| "epoch": 0.986059860598606, | |
| "grad_norm": 0.8128381967544556, | |
| "learning_rate": 6.73621777290181e-05, | |
| "loss": 0.34219295501708985, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 0.9870848708487084, | |
| "grad_norm": 0.8764263987541199, | |
| "learning_rate": 6.732789358200768e-05, | |
| "loss": 0.3455628204345703, | |
| "step": 48150 | |
| }, | |
| { | |
| "epoch": 0.988109881098811, | |
| "grad_norm": 0.807208240032196, | |
| "learning_rate": 6.729360943499727e-05, | |
| "loss": 0.3426730346679687, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 0.9891348913489135, | |
| "grad_norm": 0.7538514733314514, | |
| "learning_rate": 6.725932528798684e-05, | |
| "loss": 0.3391271591186523, | |
| "step": 48250 | |
| }, | |
| { | |
| "epoch": 0.990159901599016, | |
| "grad_norm": 0.8558005690574646, | |
| "learning_rate": 6.722504114097642e-05, | |
| "loss": 0.33498668670654297, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 0.9911849118491185, | |
| "grad_norm": 0.8771311640739441, | |
| "learning_rate": 6.719075699396599e-05, | |
| "loss": 0.337501220703125, | |
| "step": 48350 | |
| }, | |
| { | |
| "epoch": 0.992209922099221, | |
| "grad_norm": 0.7802230715751648, | |
| "learning_rate": 6.715647284695558e-05, | |
| "loss": 0.3388097381591797, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 0.9932349323493235, | |
| "grad_norm": 0.8378593325614929, | |
| "learning_rate": 6.712218869994514e-05, | |
| "loss": 0.334377326965332, | |
| "step": 48450 | |
| }, | |
| { | |
| "epoch": 0.994259942599426, | |
| "grad_norm": 0.9049150347709656, | |
| "learning_rate": 6.708790455293472e-05, | |
| "loss": 0.33694694519042967, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.9952849528495284, | |
| "grad_norm": 0.8142573237419128, | |
| "learning_rate": 6.70536204059243e-05, | |
| "loss": 0.34378395080566404, | |
| "step": 48550 | |
| }, | |
| { | |
| "epoch": 0.996309963099631, | |
| "grad_norm": 0.8853530883789062, | |
| "learning_rate": 6.701933625891388e-05, | |
| "loss": 0.3445872116088867, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 0.9973349733497335, | |
| "grad_norm": 0.8539954423904419, | |
| "learning_rate": 6.698505211190345e-05, | |
| "loss": 0.34083343505859376, | |
| "step": 48650 | |
| }, | |
| { | |
| "epoch": 0.998359983599836, | |
| "grad_norm": 0.7120438814163208, | |
| "learning_rate": 6.695076796489304e-05, | |
| "loss": 0.33607505798339843, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 0.9993849938499385, | |
| "grad_norm": 0.8253554701805115, | |
| "learning_rate": 6.691648381788262e-05, | |
| "loss": 0.3397009658813477, | |
| "step": 48750 | |
| }, | |
| { | |
| "epoch": 1.000410004100041, | |
| "grad_norm": 0.7532312273979187, | |
| "learning_rate": 6.688219967087219e-05, | |
| "loss": 0.3362251663208008, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 1.0014350143501436, | |
| "grad_norm": 0.8266898393630981, | |
| "learning_rate": 6.684791552386176e-05, | |
| "loss": 0.33053276062011716, | |
| "step": 48850 | |
| }, | |
| { | |
| "epoch": 1.002460024600246, | |
| "grad_norm": 0.8132497072219849, | |
| "learning_rate": 6.681363137685135e-05, | |
| "loss": 0.3304967498779297, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 1.0034850348503486, | |
| "grad_norm": 0.7800465822219849, | |
| "learning_rate": 6.677934722984093e-05, | |
| "loss": 0.32845794677734375, | |
| "step": 48950 | |
| }, | |
| { | |
| "epoch": 1.004510045100451, | |
| "grad_norm": 0.7818922400474548, | |
| "learning_rate": 6.67450630828305e-05, | |
| "loss": 0.3264786148071289, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 1.0055350553505535, | |
| "grad_norm": 0.7955949902534485, | |
| "learning_rate": 6.671077893582008e-05, | |
| "loss": 0.33372188568115235, | |
| "step": 49050 | |
| }, | |
| { | |
| "epoch": 1.006560065600656, | |
| "grad_norm": 0.8309062719345093, | |
| "learning_rate": 6.667649478880967e-05, | |
| "loss": 0.3315741729736328, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 1.0075850758507585, | |
| "grad_norm": 0.8513662219047546, | |
| "learning_rate": 6.664221064179923e-05, | |
| "loss": 0.33255374908447266, | |
| "step": 49150 | |
| }, | |
| { | |
| "epoch": 1.0086100861008611, | |
| "grad_norm": 0.8101342916488647, | |
| "learning_rate": 6.66079264947888e-05, | |
| "loss": 0.3335658264160156, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 1.0096350963509635, | |
| "grad_norm": 0.8653038740158081, | |
| "learning_rate": 6.657364234777839e-05, | |
| "loss": 0.33112110137939454, | |
| "step": 49250 | |
| }, | |
| { | |
| "epoch": 1.010660106601066, | |
| "grad_norm": 0.85350501537323, | |
| "learning_rate": 6.653935820076796e-05, | |
| "loss": 0.3337070465087891, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 1.0116851168511685, | |
| "grad_norm": 0.750427782535553, | |
| "learning_rate": 6.650507405375754e-05, | |
| "loss": 0.32760292053222656, | |
| "step": 49350 | |
| }, | |
| { | |
| "epoch": 1.012710127101271, | |
| "grad_norm": 0.8079749941825867, | |
| "learning_rate": 6.647078990674713e-05, | |
| "loss": 0.33652645111083984, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 1.0137351373513734, | |
| "grad_norm": 0.8437575697898865, | |
| "learning_rate": 6.64365057597367e-05, | |
| "loss": 0.3321171951293945, | |
| "step": 49450 | |
| }, | |
| { | |
| "epoch": 1.014760147601476, | |
| "grad_norm": 0.7019612789154053, | |
| "learning_rate": 6.640222161272628e-05, | |
| "loss": 0.32951507568359373, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 1.0157851578515784, | |
| "grad_norm": 0.7848524451255798, | |
| "learning_rate": 6.636793746571585e-05, | |
| "loss": 0.33918540954589843, | |
| "step": 49550 | |
| }, | |
| { | |
| "epoch": 1.016810168101681, | |
| "grad_norm": 0.8270383477210999, | |
| "learning_rate": 6.633365331870544e-05, | |
| "loss": 0.32835037231445313, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 1.0178351783517836, | |
| "grad_norm": 0.760426938533783, | |
| "learning_rate": 6.629936917169501e-05, | |
| "loss": 0.3362861633300781, | |
| "step": 49650 | |
| }, | |
| { | |
| "epoch": 1.018860188601886, | |
| "grad_norm": 0.7765412330627441, | |
| "learning_rate": 6.626508502468459e-05, | |
| "loss": 0.3345751571655273, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 1.0198851988519886, | |
| "grad_norm": 0.8224347233772278, | |
| "learning_rate": 6.623080087767418e-05, | |
| "loss": 0.32922367095947264, | |
| "step": 49750 | |
| }, | |
| { | |
| "epoch": 1.020910209102091, | |
| "grad_norm": 0.9070143699645996, | |
| "learning_rate": 6.619651673066375e-05, | |
| "loss": 0.3320350646972656, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 1.0219352193521936, | |
| "grad_norm": 0.7810198664665222, | |
| "learning_rate": 6.616223258365331e-05, | |
| "loss": 0.3291141891479492, | |
| "step": 49850 | |
| }, | |
| { | |
| "epoch": 1.022960229602296, | |
| "grad_norm": 0.7284088730812073, | |
| "learning_rate": 6.61279484366429e-05, | |
| "loss": 0.3341518020629883, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 1.0239852398523985, | |
| "grad_norm": 0.8296566605567932, | |
| "learning_rate": 6.609366428963248e-05, | |
| "loss": 0.3288591003417969, | |
| "step": 49950 | |
| }, | |
| { | |
| "epoch": 1.0250102501025011, | |
| "grad_norm": 0.8086618185043335, | |
| "learning_rate": 6.605938014262205e-05, | |
| "loss": 0.3367390823364258, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 1.0250102501025011, | |
| "eval_loss": 0.4105064868927002, | |
| "eval_runtime": 4.4832, | |
| "eval_samples_per_second": 5576.352, | |
| "eval_steps_per_second": 5.576, | |
| "eval_teacher_cosine_loss": 0.25726622343063354, | |
| "eval_teacher_cosine_mean": 0.7427337765693665, | |
| "eval_teacher_cosine_p10": 0.6649437546730042, | |
| "eval_teacher_cosine_p50": 0.7521957159042358, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 1.0260352603526035, | |
| "grad_norm": 0.7820824384689331, | |
| "learning_rate": 6.602509599561163e-05, | |
| "loss": 0.3357306671142578, | |
| "step": 50050 | |
| }, | |
| { | |
| "epoch": 1.027060270602706, | |
| "grad_norm": 0.8739521503448486, | |
| "learning_rate": 6.599081184860121e-05, | |
| "loss": 0.33369003295898436, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 1.0280852808528085, | |
| "grad_norm": 0.7873463034629822, | |
| "learning_rate": 6.595652770159079e-05, | |
| "loss": 0.3271013259887695, | |
| "step": 50150 | |
| }, | |
| { | |
| "epoch": 1.029110291102911, | |
| "grad_norm": 0.847722053527832, | |
| "learning_rate": 6.592224355458036e-05, | |
| "loss": 0.3344841003417969, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 1.0301353013530135, | |
| "grad_norm": 0.7994325160980225, | |
| "learning_rate": 6.588795940756994e-05, | |
| "loss": 0.3308522033691406, | |
| "step": 50250 | |
| }, | |
| { | |
| "epoch": 1.031160311603116, | |
| "grad_norm": 0.8272190690040588, | |
| "learning_rate": 6.585367526055953e-05, | |
| "loss": 0.3290030288696289, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 1.0321853218532184, | |
| "grad_norm": 0.78369140625, | |
| "learning_rate": 6.58193911135491e-05, | |
| "loss": 0.3344311904907227, | |
| "step": 50350 | |
| }, | |
| { | |
| "epoch": 1.033210332103321, | |
| "grad_norm": 0.8008580803871155, | |
| "learning_rate": 6.578510696653868e-05, | |
| "loss": 0.3259135818481445, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 1.0342353423534236, | |
| "grad_norm": 0.8687888383865356, | |
| "learning_rate": 6.575082281952826e-05, | |
| "loss": 0.33363807678222657, | |
| "step": 50450 | |
| }, | |
| { | |
| "epoch": 1.035260352603526, | |
| "grad_norm": 0.9110373258590698, | |
| "learning_rate": 6.571653867251784e-05, | |
| "loss": 0.3260751724243164, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 1.0362853628536286, | |
| "grad_norm": 0.8350563049316406, | |
| "learning_rate": 6.568225452550741e-05, | |
| "loss": 0.3334164810180664, | |
| "step": 50550 | |
| }, | |
| { | |
| "epoch": 1.037310373103731, | |
| "grad_norm": 0.8459510803222656, | |
| "learning_rate": 6.564797037849699e-05, | |
| "loss": 0.3354366302490234, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 1.0383353833538336, | |
| "grad_norm": 0.8334582448005676, | |
| "learning_rate": 6.561368623148656e-05, | |
| "loss": 0.33314617156982423, | |
| "step": 50650 | |
| }, | |
| { | |
| "epoch": 1.039360393603936, | |
| "grad_norm": 0.7499675154685974, | |
| "learning_rate": 6.557940208447614e-05, | |
| "loss": 0.3337816619873047, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 1.0403854038540385, | |
| "grad_norm": 0.8296809196472168, | |
| "learning_rate": 6.554511793746571e-05, | |
| "loss": 0.32563720703125, | |
| "step": 50750 | |
| }, | |
| { | |
| "epoch": 1.0414104141041411, | |
| "grad_norm": 0.8647814989089966, | |
| "learning_rate": 6.55108337904553e-05, | |
| "loss": 0.32740543365478514, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 1.0424354243542435, | |
| "grad_norm": 0.8768616318702698, | |
| "learning_rate": 6.547654964344487e-05, | |
| "loss": 0.3330791473388672, | |
| "step": 50850 | |
| }, | |
| { | |
| "epoch": 1.0434604346043461, | |
| "grad_norm": 0.8242044448852539, | |
| "learning_rate": 6.544226549643445e-05, | |
| "loss": 0.33479488372802735, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 1.0444854448544485, | |
| "grad_norm": 0.7757652997970581, | |
| "learning_rate": 6.540798134942402e-05, | |
| "loss": 0.3292919921875, | |
| "step": 50950 | |
| }, | |
| { | |
| "epoch": 1.045510455104551, | |
| "grad_norm": 0.8209587335586548, | |
| "learning_rate": 6.537369720241361e-05, | |
| "loss": 0.33092041015625, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.0465354653546535, | |
| "grad_norm": 0.7978271245956421, | |
| "learning_rate": 6.533941305540319e-05, | |
| "loss": 0.32867340087890623, | |
| "step": 51050 | |
| }, | |
| { | |
| "epoch": 1.047560475604756, | |
| "grad_norm": 0.8840471506118774, | |
| "learning_rate": 6.530512890839276e-05, | |
| "loss": 0.33462760925292967, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 1.0485854858548584, | |
| "grad_norm": 0.7766537070274353, | |
| "learning_rate": 6.527084476138235e-05, | |
| "loss": 0.33895130157470704, | |
| "step": 51150 | |
| }, | |
| { | |
| "epoch": 1.049610496104961, | |
| "grad_norm": 0.8571930527687073, | |
| "learning_rate": 6.523656061437192e-05, | |
| "loss": 0.33560428619384763, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 1.0506355063550636, | |
| "grad_norm": 0.7688530087471008, | |
| "learning_rate": 6.52022764673615e-05, | |
| "loss": 0.3329842758178711, | |
| "step": 51250 | |
| }, | |
| { | |
| "epoch": 1.051660516605166, | |
| "grad_norm": 0.7348241209983826, | |
| "learning_rate": 6.516799232035107e-05, | |
| "loss": 0.33077308654785154, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 1.0526855268552686, | |
| "grad_norm": 0.8242869973182678, | |
| "learning_rate": 6.513370817334065e-05, | |
| "loss": 0.3299036407470703, | |
| "step": 51350 | |
| }, | |
| { | |
| "epoch": 1.053710537105371, | |
| "grad_norm": 0.8798810839653015, | |
| "learning_rate": 6.509942402633022e-05, | |
| "loss": 0.32433429718017576, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 1.0547355473554736, | |
| "grad_norm": 0.9281744956970215, | |
| "learning_rate": 6.50651398793198e-05, | |
| "loss": 0.33434730529785156, | |
| "step": 51450 | |
| }, | |
| { | |
| "epoch": 1.055760557605576, | |
| "grad_norm": 0.839369535446167, | |
| "learning_rate": 6.503085573230939e-05, | |
| "loss": 0.33418094635009765, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 1.0567855678556786, | |
| "grad_norm": 0.8913266658782959, | |
| "learning_rate": 6.499657158529896e-05, | |
| "loss": 0.3327432632446289, | |
| "step": 51550 | |
| }, | |
| { | |
| "epoch": 1.0578105781057812, | |
| "grad_norm": 0.847016453742981, | |
| "learning_rate": 6.496228743828854e-05, | |
| "loss": 0.3308758544921875, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 1.0588355883558835, | |
| "grad_norm": 0.7980254292488098, | |
| "learning_rate": 6.492800329127811e-05, | |
| "loss": 0.3327946090698242, | |
| "step": 51650 | |
| }, | |
| { | |
| "epoch": 1.0598605986059861, | |
| "grad_norm": 0.8209347724914551, | |
| "learning_rate": 6.48937191442677e-05, | |
| "loss": 0.33655754089355466, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 1.0608856088560885, | |
| "grad_norm": 0.8379464149475098, | |
| "learning_rate": 6.485943499725727e-05, | |
| "loss": 0.3324146270751953, | |
| "step": 51750 | |
| }, | |
| { | |
| "epoch": 1.0619106191061911, | |
| "grad_norm": 0.7903445363044739, | |
| "learning_rate": 6.482515085024685e-05, | |
| "loss": 0.3290152740478516, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 1.0629356293562935, | |
| "grad_norm": 0.8222147822380066, | |
| "learning_rate": 6.479086670323644e-05, | |
| "loss": 0.32505207061767577, | |
| "step": 51850 | |
| }, | |
| { | |
| "epoch": 1.063960639606396, | |
| "grad_norm": 0.8345584869384766, | |
| "learning_rate": 6.475658255622601e-05, | |
| "loss": 0.3379618072509766, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 1.0649856498564985, | |
| "grad_norm": 0.831203043460846, | |
| "learning_rate": 6.472229840921559e-05, | |
| "loss": 0.3329928970336914, | |
| "step": 51950 | |
| }, | |
| { | |
| "epoch": 1.066010660106601, | |
| "grad_norm": 0.8929985761642456, | |
| "learning_rate": 6.468801426220516e-05, | |
| "loss": 0.3283989334106445, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 1.0670356703567037, | |
| "grad_norm": 0.8127049207687378, | |
| "learning_rate": 6.465373011519474e-05, | |
| "loss": 0.33351539611816405, | |
| "step": 52050 | |
| }, | |
| { | |
| "epoch": 1.068060680606806, | |
| "grad_norm": 0.8155128955841064, | |
| "learning_rate": 6.461944596818431e-05, | |
| "loss": 0.327579345703125, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 1.0690856908569086, | |
| "grad_norm": 0.7758855223655701, | |
| "learning_rate": 6.458516182117388e-05, | |
| "loss": 0.3336928176879883, | |
| "step": 52150 | |
| }, | |
| { | |
| "epoch": 1.070110701107011, | |
| "grad_norm": 0.7681171298027039, | |
| "learning_rate": 6.455087767416347e-05, | |
| "loss": 0.3333876419067383, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 1.0711357113571136, | |
| "grad_norm": 0.8418704867362976, | |
| "learning_rate": 6.451659352715305e-05, | |
| "loss": 0.32921504974365234, | |
| "step": 52250 | |
| }, | |
| { | |
| "epoch": 1.072160721607216, | |
| "grad_norm": 0.8479698896408081, | |
| "learning_rate": 6.448230938014262e-05, | |
| "loss": 0.33223762512207033, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 1.0731857318573186, | |
| "grad_norm": 0.9115731120109558, | |
| "learning_rate": 6.44480252331322e-05, | |
| "loss": 0.32863014221191406, | |
| "step": 52350 | |
| }, | |
| { | |
| "epoch": 1.0742107421074212, | |
| "grad_norm": 0.8291541934013367, | |
| "learning_rate": 6.441374108612179e-05, | |
| "loss": 0.32942676544189453, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 1.0752357523575236, | |
| "grad_norm": 0.7475597262382507, | |
| "learning_rate": 6.437945693911136e-05, | |
| "loss": 0.3333176803588867, | |
| "step": 52450 | |
| }, | |
| { | |
| "epoch": 1.0762607626076262, | |
| "grad_norm": 0.8054630756378174, | |
| "learning_rate": 6.434517279210093e-05, | |
| "loss": 0.33060565948486326, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 1.0772857728577285, | |
| "grad_norm": 0.8592610955238342, | |
| "learning_rate": 6.431088864509052e-05, | |
| "loss": 0.3371944808959961, | |
| "step": 52550 | |
| }, | |
| { | |
| "epoch": 1.0783107831078311, | |
| "grad_norm": 0.8010218739509583, | |
| "learning_rate": 6.42766044980801e-05, | |
| "loss": 0.32953815460205077, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 1.0793357933579335, | |
| "grad_norm": 0.7844801545143127, | |
| "learning_rate": 6.424232035106967e-05, | |
| "loss": 0.33063697814941406, | |
| "step": 52650 | |
| }, | |
| { | |
| "epoch": 1.080360803608036, | |
| "grad_norm": 0.8183125853538513, | |
| "learning_rate": 6.420803620405925e-05, | |
| "loss": 0.33582832336425783, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 1.0813858138581385, | |
| "grad_norm": 0.8097880482673645, | |
| "learning_rate": 6.417375205704882e-05, | |
| "loss": 0.3283766555786133, | |
| "step": 52750 | |
| }, | |
| { | |
| "epoch": 1.082410824108241, | |
| "grad_norm": 0.850816011428833, | |
| "learning_rate": 6.41394679100384e-05, | |
| "loss": 0.3322663116455078, | |
| "step": 52800 | |
| }, | |
| { | |
| "epoch": 1.0834358343583437, | |
| "grad_norm": 0.8583124279975891, | |
| "learning_rate": 6.410518376302797e-05, | |
| "loss": 0.3321968078613281, | |
| "step": 52850 | |
| }, | |
| { | |
| "epoch": 1.084460844608446, | |
| "grad_norm": 0.7499483823776245, | |
| "learning_rate": 6.407089961601756e-05, | |
| "loss": 0.33192115783691406, | |
| "step": 52900 | |
| }, | |
| { | |
| "epoch": 1.0854858548585486, | |
| "grad_norm": 0.7568508386611938, | |
| "learning_rate": 6.403661546900713e-05, | |
| "loss": 0.3324843215942383, | |
| "step": 52950 | |
| }, | |
| { | |
| "epoch": 1.086510865108651, | |
| "grad_norm": 0.913845956325531, | |
| "learning_rate": 6.400233132199671e-05, | |
| "loss": 0.33451251983642577, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 1.0875358753587536, | |
| "grad_norm": 0.8257799744606018, | |
| "learning_rate": 6.396804717498628e-05, | |
| "loss": 0.34046268463134766, | |
| "step": 53050 | |
| }, | |
| { | |
| "epoch": 1.088560885608856, | |
| "grad_norm": 0.920488178730011, | |
| "learning_rate": 6.393376302797587e-05, | |
| "loss": 0.33624126434326174, | |
| "step": 53100 | |
| }, | |
| { | |
| "epoch": 1.0895858958589586, | |
| "grad_norm": 0.8326303362846375, | |
| "learning_rate": 6.389947888096545e-05, | |
| "loss": 0.3313284683227539, | |
| "step": 53150 | |
| }, | |
| { | |
| "epoch": 1.090610906109061, | |
| "grad_norm": 0.7235586643218994, | |
| "learning_rate": 6.386519473395502e-05, | |
| "loss": 0.3317373275756836, | |
| "step": 53200 | |
| }, | |
| { | |
| "epoch": 1.0916359163591636, | |
| "grad_norm": 0.8453660011291504, | |
| "learning_rate": 6.383091058694461e-05, | |
| "loss": 0.3292432022094727, | |
| "step": 53250 | |
| }, | |
| { | |
| "epoch": 1.0926609266092662, | |
| "grad_norm": 0.9193964600563049, | |
| "learning_rate": 6.379662643993418e-05, | |
| "loss": 0.32905914306640627, | |
| "step": 53300 | |
| }, | |
| { | |
| "epoch": 1.0936859368593685, | |
| "grad_norm": 0.8209521770477295, | |
| "learning_rate": 6.376234229292376e-05, | |
| "loss": 0.3345983123779297, | |
| "step": 53350 | |
| }, | |
| { | |
| "epoch": 1.0947109471094711, | |
| "grad_norm": 0.7486148476600647, | |
| "learning_rate": 6.372805814591333e-05, | |
| "loss": 0.3347195053100586, | |
| "step": 53400 | |
| }, | |
| { | |
| "epoch": 1.0957359573595735, | |
| "grad_norm": 0.7953851819038391, | |
| "learning_rate": 6.369377399890291e-05, | |
| "loss": 0.33252445220947263, | |
| "step": 53450 | |
| }, | |
| { | |
| "epoch": 1.0967609676096761, | |
| "grad_norm": 0.7317273616790771, | |
| "learning_rate": 6.365948985189248e-05, | |
| "loss": 0.3319861602783203, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 1.0977859778597785, | |
| "grad_norm": 0.8502221703529358, | |
| "learning_rate": 6.362520570488206e-05, | |
| "loss": 0.33589332580566406, | |
| "step": 53550 | |
| }, | |
| { | |
| "epoch": 1.098810988109881, | |
| "grad_norm": 0.8646501898765564, | |
| "learning_rate": 6.359092155787165e-05, | |
| "loss": 0.3323219680786133, | |
| "step": 53600 | |
| }, | |
| { | |
| "epoch": 1.0998359983599837, | |
| "grad_norm": 0.8983181715011597, | |
| "learning_rate": 6.355663741086122e-05, | |
| "loss": 0.3326095199584961, | |
| "step": 53650 | |
| }, | |
| { | |
| "epoch": 1.100861008610086, | |
| "grad_norm": 0.8646010160446167, | |
| "learning_rate": 6.35223532638508e-05, | |
| "loss": 0.32858699798583985, | |
| "step": 53700 | |
| }, | |
| { | |
| "epoch": 1.1018860188601887, | |
| "grad_norm": 0.7975472211837769, | |
| "learning_rate": 6.348806911684037e-05, | |
| "loss": 0.33134807586669923, | |
| "step": 53750 | |
| }, | |
| { | |
| "epoch": 1.102911029110291, | |
| "grad_norm": 0.7708473205566406, | |
| "learning_rate": 6.345378496982996e-05, | |
| "loss": 0.32811927795410156, | |
| "step": 53800 | |
| }, | |
| { | |
| "epoch": 1.1039360393603936, | |
| "grad_norm": 0.8618834018707275, | |
| "learning_rate": 6.341950082281953e-05, | |
| "loss": 0.3295989227294922, | |
| "step": 53850 | |
| }, | |
| { | |
| "epoch": 1.104961049610496, | |
| "grad_norm": 0.8573905229568481, | |
| "learning_rate": 6.338521667580911e-05, | |
| "loss": 0.33238658905029295, | |
| "step": 53900 | |
| }, | |
| { | |
| "epoch": 1.1059860598605986, | |
| "grad_norm": 0.9029597043991089, | |
| "learning_rate": 6.33509325287987e-05, | |
| "loss": 0.33339744567871094, | |
| "step": 53950 | |
| }, | |
| { | |
| "epoch": 1.1070110701107012, | |
| "grad_norm": 0.8479767441749573, | |
| "learning_rate": 6.331664838178827e-05, | |
| "loss": 0.32906696319580075, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.1080360803608036, | |
| "grad_norm": 0.7513405680656433, | |
| "learning_rate": 6.328236423477785e-05, | |
| "loss": 0.3311982727050781, | |
| "step": 54050 | |
| }, | |
| { | |
| "epoch": 1.1090610906109062, | |
| "grad_norm": 0.8105347156524658, | |
| "learning_rate": 6.324808008776742e-05, | |
| "loss": 0.33865703582763673, | |
| "step": 54100 | |
| }, | |
| { | |
| "epoch": 1.1100861008610086, | |
| "grad_norm": 0.8898072242736816, | |
| "learning_rate": 6.3213795940757e-05, | |
| "loss": 0.33245800018310545, | |
| "step": 54150 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.8270716071128845, | |
| "learning_rate": 6.317951179374657e-05, | |
| "loss": 0.3283497619628906, | |
| "step": 54200 | |
| }, | |
| { | |
| "epoch": 1.1121361213612135, | |
| "grad_norm": 0.9061145186424255, | |
| "learning_rate": 6.314522764673614e-05, | |
| "loss": 0.33055049896240235, | |
| "step": 54250 | |
| }, | |
| { | |
| "epoch": 1.1131611316113161, | |
| "grad_norm": 0.8416737914085388, | |
| "learning_rate": 6.311094349972573e-05, | |
| "loss": 0.3307025146484375, | |
| "step": 54300 | |
| }, | |
| { | |
| "epoch": 1.1141861418614185, | |
| "grad_norm": 0.8140227198600769, | |
| "learning_rate": 6.307665935271531e-05, | |
| "loss": 0.33488998413085935, | |
| "step": 54350 | |
| }, | |
| { | |
| "epoch": 1.115211152111521, | |
| "grad_norm": 0.9350836873054504, | |
| "learning_rate": 6.304237520570488e-05, | |
| "loss": 0.33722270965576173, | |
| "step": 54400 | |
| }, | |
| { | |
| "epoch": 1.1162361623616237, | |
| "grad_norm": 0.8020601272583008, | |
| "learning_rate": 6.300809105869446e-05, | |
| "loss": 0.3286500549316406, | |
| "step": 54450 | |
| }, | |
| { | |
| "epoch": 1.117261172611726, | |
| "grad_norm": 0.7725200653076172, | |
| "learning_rate": 6.297380691168404e-05, | |
| "loss": 0.3293790435791016, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 1.1182861828618287, | |
| "grad_norm": 0.8838649988174438, | |
| "learning_rate": 6.293952276467362e-05, | |
| "loss": 0.3336125183105469, | |
| "step": 54550 | |
| }, | |
| { | |
| "epoch": 1.119311193111931, | |
| "grad_norm": 0.7584066390991211, | |
| "learning_rate": 6.29052386176632e-05, | |
| "loss": 0.33170696258544924, | |
| "step": 54600 | |
| }, | |
| { | |
| "epoch": 1.1203362033620337, | |
| "grad_norm": 0.8115083575248718, | |
| "learning_rate": 6.287095447065278e-05, | |
| "loss": 0.33287025451660157, | |
| "step": 54650 | |
| }, | |
| { | |
| "epoch": 1.121361213612136, | |
| "grad_norm": 0.794979453086853, | |
| "learning_rate": 6.283667032364236e-05, | |
| "loss": 0.32832927703857423, | |
| "step": 54700 | |
| }, | |
| { | |
| "epoch": 1.1223862238622386, | |
| "grad_norm": 0.7886215448379517, | |
| "learning_rate": 6.280238617663193e-05, | |
| "loss": 0.33019577026367186, | |
| "step": 54750 | |
| }, | |
| { | |
| "epoch": 1.123411234112341, | |
| "grad_norm": 0.7897105813026428, | |
| "learning_rate": 6.27681020296215e-05, | |
| "loss": 0.3316835403442383, | |
| "step": 54800 | |
| }, | |
| { | |
| "epoch": 1.1244362443624436, | |
| "grad_norm": 0.8683794736862183, | |
| "learning_rate": 6.273381788261108e-05, | |
| "loss": 0.32726833343505857, | |
| "step": 54850 | |
| }, | |
| { | |
| "epoch": 1.1254612546125462, | |
| "grad_norm": 0.8658251166343689, | |
| "learning_rate": 6.269953373560066e-05, | |
| "loss": 0.3290073013305664, | |
| "step": 54900 | |
| }, | |
| { | |
| "epoch": 1.1264862648626486, | |
| "grad_norm": 0.8676416277885437, | |
| "learning_rate": 6.266524958859023e-05, | |
| "loss": 0.3309788131713867, | |
| "step": 54950 | |
| }, | |
| { | |
| "epoch": 1.1275112751127512, | |
| "grad_norm": 0.9222783446311951, | |
| "learning_rate": 6.263096544157982e-05, | |
| "loss": 0.33274337768554685, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 1.1275112751127512, | |
| "eval_loss": 0.40561673045158386, | |
| "eval_runtime": 4.5148, | |
| "eval_samples_per_second": 5537.326, | |
| "eval_steps_per_second": 5.537, | |
| "eval_teacher_cosine_loss": 0.2544936537742615, | |
| "eval_teacher_cosine_mean": 0.7455063462257385, | |
| "eval_teacher_cosine_p10": 0.6672899127006531, | |
| "eval_teacher_cosine_p50": 0.7551168203353882, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 1.1285362853628536, | |
| "grad_norm": 0.8280739784240723, | |
| "learning_rate": 6.25966812945694e-05, | |
| "loss": 0.3278068542480469, | |
| "step": 55050 | |
| }, | |
| { | |
| "epoch": 1.1295612956129562, | |
| "grad_norm": 0.8689860105514526, | |
| "learning_rate": 6.256239714755897e-05, | |
| "loss": 0.3320955276489258, | |
| "step": 55100 | |
| }, | |
| { | |
| "epoch": 1.1305863058630585, | |
| "grad_norm": 0.8158032894134521, | |
| "learning_rate": 6.252811300054854e-05, | |
| "loss": 0.3334500503540039, | |
| "step": 55150 | |
| }, | |
| { | |
| "epoch": 1.1316113161131611, | |
| "grad_norm": 0.7937714457511902, | |
| "learning_rate": 6.249382885353813e-05, | |
| "loss": 0.3324356842041016, | |
| "step": 55200 | |
| }, | |
| { | |
| "epoch": 1.1326363263632637, | |
| "grad_norm": 0.7214861512184143, | |
| "learning_rate": 6.24595447065277e-05, | |
| "loss": 0.3298004150390625, | |
| "step": 55250 | |
| }, | |
| { | |
| "epoch": 1.133661336613366, | |
| "grad_norm": 0.817425549030304, | |
| "learning_rate": 6.242526055951728e-05, | |
| "loss": 0.3300360107421875, | |
| "step": 55300 | |
| }, | |
| { | |
| "epoch": 1.1346863468634687, | |
| "grad_norm": 0.7910701036453247, | |
| "learning_rate": 6.239097641250687e-05, | |
| "loss": 0.32433349609375, | |
| "step": 55350 | |
| }, | |
| { | |
| "epoch": 1.135711357113571, | |
| "grad_norm": 0.7492151260375977, | |
| "learning_rate": 6.235669226549644e-05, | |
| "loss": 0.3307527542114258, | |
| "step": 55400 | |
| }, | |
| { | |
| "epoch": 1.1367363673636737, | |
| "grad_norm": 0.8596962094306946, | |
| "learning_rate": 6.232240811848602e-05, | |
| "loss": 0.3331085205078125, | |
| "step": 55450 | |
| }, | |
| { | |
| "epoch": 1.137761377613776, | |
| "grad_norm": 0.9153379201889038, | |
| "learning_rate": 6.228812397147559e-05, | |
| "loss": 0.3335791015625, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 1.1387863878638786, | |
| "grad_norm": 0.8540279865264893, | |
| "learning_rate": 6.225383982446517e-05, | |
| "loss": 0.3284048843383789, | |
| "step": 55550 | |
| }, | |
| { | |
| "epoch": 1.1398113981139812, | |
| "grad_norm": 0.8452389240264893, | |
| "learning_rate": 6.221955567745474e-05, | |
| "loss": 0.33166065216064455, | |
| "step": 55600 | |
| }, | |
| { | |
| "epoch": 1.1408364083640836, | |
| "grad_norm": 0.8229838013648987, | |
| "learning_rate": 6.218527153044432e-05, | |
| "loss": 0.3274332809448242, | |
| "step": 55650 | |
| }, | |
| { | |
| "epoch": 1.1418614186141862, | |
| "grad_norm": 0.8455730676651001, | |
| "learning_rate": 6.21509873834339e-05, | |
| "loss": 0.32831058502197263, | |
| "step": 55700 | |
| }, | |
| { | |
| "epoch": 1.1428864288642886, | |
| "grad_norm": 0.9262994527816772, | |
| "learning_rate": 6.211670323642348e-05, | |
| "loss": 0.3324673843383789, | |
| "step": 55750 | |
| }, | |
| { | |
| "epoch": 1.1439114391143912, | |
| "grad_norm": 0.8037222027778625, | |
| "learning_rate": 6.208241908941305e-05, | |
| "loss": 0.3330752944946289, | |
| "step": 55800 | |
| }, | |
| { | |
| "epoch": 1.1449364493644936, | |
| "grad_norm": 0.7826458215713501, | |
| "learning_rate": 6.204813494240263e-05, | |
| "loss": 0.32536170959472654, | |
| "step": 55850 | |
| }, | |
| { | |
| "epoch": 1.1459614596145962, | |
| "grad_norm": 0.7451727390289307, | |
| "learning_rate": 6.201385079539222e-05, | |
| "loss": 0.3311782455444336, | |
| "step": 55900 | |
| }, | |
| { | |
| "epoch": 1.1469864698646988, | |
| "grad_norm": 0.8275009393692017, | |
| "learning_rate": 6.197956664838179e-05, | |
| "loss": 0.32733860015869143, | |
| "step": 55950 | |
| }, | |
| { | |
| "epoch": 1.1480114801148011, | |
| "grad_norm": 0.7221106290817261, | |
| "learning_rate": 6.194528250137137e-05, | |
| "loss": 0.3324686813354492, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 1.1490364903649037, | |
| "grad_norm": 0.8201016187667847, | |
| "learning_rate": 6.191099835436096e-05, | |
| "loss": 0.33275001525878906, | |
| "step": 56050 | |
| }, | |
| { | |
| "epoch": 1.1500615006150061, | |
| "grad_norm": 0.911514163017273, | |
| "learning_rate": 6.187671420735053e-05, | |
| "loss": 0.33460781097412107, | |
| "step": 56100 | |
| }, | |
| { | |
| "epoch": 1.1510865108651087, | |
| "grad_norm": 0.8894955515861511, | |
| "learning_rate": 6.18424300603401e-05, | |
| "loss": 0.32276912689208986, | |
| "step": 56150 | |
| }, | |
| { | |
| "epoch": 1.152111521115211, | |
| "grad_norm": 0.7889685034751892, | |
| "learning_rate": 6.180814591332968e-05, | |
| "loss": 0.3323189163208008, | |
| "step": 56200 | |
| }, | |
| { | |
| "epoch": 1.1531365313653137, | |
| "grad_norm": 0.8331491351127625, | |
| "learning_rate": 6.177386176631925e-05, | |
| "loss": 0.3296329498291016, | |
| "step": 56250 | |
| }, | |
| { | |
| "epoch": 1.154161541615416, | |
| "grad_norm": 0.8493555784225464, | |
| "learning_rate": 6.173957761930883e-05, | |
| "loss": 0.3329767608642578, | |
| "step": 56300 | |
| }, | |
| { | |
| "epoch": 1.1551865518655187, | |
| "grad_norm": 0.8830084800720215, | |
| "learning_rate": 6.17052934722984e-05, | |
| "loss": 0.3304352569580078, | |
| "step": 56350 | |
| }, | |
| { | |
| "epoch": 1.156211562115621, | |
| "grad_norm": 0.9141111969947815, | |
| "learning_rate": 6.167100932528799e-05, | |
| "loss": 0.33026588439941407, | |
| "step": 56400 | |
| }, | |
| { | |
| "epoch": 1.1572365723657236, | |
| "grad_norm": 0.7311347723007202, | |
| "learning_rate": 6.163672517827757e-05, | |
| "loss": 0.32671092987060546, | |
| "step": 56450 | |
| }, | |
| { | |
| "epoch": 1.1582615826158262, | |
| "grad_norm": 0.8988407254219055, | |
| "learning_rate": 6.160244103126714e-05, | |
| "loss": 0.33409019470214846, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 1.1592865928659286, | |
| "grad_norm": 0.8051838278770447, | |
| "learning_rate": 6.156815688425673e-05, | |
| "loss": 0.32466949462890626, | |
| "step": 56550 | |
| }, | |
| { | |
| "epoch": 1.1603116031160312, | |
| "grad_norm": 0.766719400882721, | |
| "learning_rate": 6.15338727372463e-05, | |
| "loss": 0.3337104797363281, | |
| "step": 56600 | |
| }, | |
| { | |
| "epoch": 1.1613366133661336, | |
| "grad_norm": 0.8739268779754639, | |
| "learning_rate": 6.149958859023588e-05, | |
| "loss": 0.3314841842651367, | |
| "step": 56650 | |
| }, | |
| { | |
| "epoch": 1.1623616236162362, | |
| "grad_norm": 0.7714674472808838, | |
| "learning_rate": 6.146530444322545e-05, | |
| "loss": 0.32541336059570314, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 1.1633866338663386, | |
| "grad_norm": 0.8535489439964294, | |
| "learning_rate": 6.143102029621504e-05, | |
| "loss": 0.3324369812011719, | |
| "step": 56750 | |
| }, | |
| { | |
| "epoch": 1.1644116441164412, | |
| "grad_norm": 0.8496703505516052, | |
| "learning_rate": 6.139673614920462e-05, | |
| "loss": 0.32721839904785155, | |
| "step": 56800 | |
| }, | |
| { | |
| "epoch": 1.1654366543665438, | |
| "grad_norm": 0.7496902346611023, | |
| "learning_rate": 6.136245200219419e-05, | |
| "loss": 0.33139427185058595, | |
| "step": 56850 | |
| }, | |
| { | |
| "epoch": 1.1664616646166461, | |
| "grad_norm": 0.7812837362289429, | |
| "learning_rate": 6.132816785518377e-05, | |
| "loss": 0.3255027389526367, | |
| "step": 56900 | |
| }, | |
| { | |
| "epoch": 1.1674866748667487, | |
| "grad_norm": 0.7950329184532166, | |
| "learning_rate": 6.129388370817334e-05, | |
| "loss": 0.32295726776123046, | |
| "step": 56950 | |
| }, | |
| { | |
| "epoch": 1.168511685116851, | |
| "grad_norm": 0.8041350841522217, | |
| "learning_rate": 6.125959956116291e-05, | |
| "loss": 0.3290506744384766, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 1.1695366953669537, | |
| "grad_norm": 0.8516210913658142, | |
| "learning_rate": 6.122531541415249e-05, | |
| "loss": 0.3315174102783203, | |
| "step": 57050 | |
| }, | |
| { | |
| "epoch": 1.170561705617056, | |
| "grad_norm": 0.7783889770507812, | |
| "learning_rate": 6.119103126714208e-05, | |
| "loss": 0.33206024169921877, | |
| "step": 57100 | |
| }, | |
| { | |
| "epoch": 1.1715867158671587, | |
| "grad_norm": 0.8401105403900146, | |
| "learning_rate": 6.115674712013165e-05, | |
| "loss": 0.33236587524414063, | |
| "step": 57150 | |
| }, | |
| { | |
| "epoch": 1.1726117261172613, | |
| "grad_norm": 0.7813949584960938, | |
| "learning_rate": 6.112246297312123e-05, | |
| "loss": 0.32738136291503905, | |
| "step": 57200 | |
| }, | |
| { | |
| "epoch": 1.1736367363673637, | |
| "grad_norm": 0.7983943223953247, | |
| "learning_rate": 6.108817882611082e-05, | |
| "loss": 0.3247266006469727, | |
| "step": 57250 | |
| }, | |
| { | |
| "epoch": 1.1746617466174663, | |
| "grad_norm": 0.8523854613304138, | |
| "learning_rate": 6.105389467910039e-05, | |
| "loss": 0.33392494201660156, | |
| "step": 57300 | |
| }, | |
| { | |
| "epoch": 1.1756867568675686, | |
| "grad_norm": 0.8423400521278381, | |
| "learning_rate": 6.1019610532089965e-05, | |
| "loss": 0.32623321533203126, | |
| "step": 57350 | |
| }, | |
| { | |
| "epoch": 1.1767117671176712, | |
| "grad_norm": 0.8958165645599365, | |
| "learning_rate": 6.0985326385079546e-05, | |
| "loss": 0.3322746276855469, | |
| "step": 57400 | |
| }, | |
| { | |
| "epoch": 1.1777367773677736, | |
| "grad_norm": 0.8603180050849915, | |
| "learning_rate": 6.095104223806912e-05, | |
| "loss": 0.33123390197753905, | |
| "step": 57450 | |
| }, | |
| { | |
| "epoch": 1.1787617876178762, | |
| "grad_norm": 0.8342885971069336, | |
| "learning_rate": 6.09167580910587e-05, | |
| "loss": 0.33096492767333985, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 1.1797867978679788, | |
| "grad_norm": 0.8071539998054504, | |
| "learning_rate": 6.088247394404828e-05, | |
| "loss": 0.32896678924560546, | |
| "step": 57550 | |
| }, | |
| { | |
| "epoch": 1.1808118081180812, | |
| "grad_norm": 0.8804877996444702, | |
| "learning_rate": 6.084818979703786e-05, | |
| "loss": 0.3319522857666016, | |
| "step": 57600 | |
| }, | |
| { | |
| "epoch": 1.1818368183681838, | |
| "grad_norm": 0.8537464141845703, | |
| "learning_rate": 6.081390565002743e-05, | |
| "loss": 0.3308625030517578, | |
| "step": 57650 | |
| }, | |
| { | |
| "epoch": 1.1828618286182861, | |
| "grad_norm": 0.9388439059257507, | |
| "learning_rate": 6.0779621503017e-05, | |
| "loss": 0.3344725036621094, | |
| "step": 57700 | |
| }, | |
| { | |
| "epoch": 1.1838868388683887, | |
| "grad_norm": 0.8253179788589478, | |
| "learning_rate": 6.074533735600658e-05, | |
| "loss": 0.3262712860107422, | |
| "step": 57750 | |
| }, | |
| { | |
| "epoch": 1.1849118491184911, | |
| "grad_norm": 0.8718785643577576, | |
| "learning_rate": 6.071105320899616e-05, | |
| "loss": 0.3324061584472656, | |
| "step": 57800 | |
| }, | |
| { | |
| "epoch": 1.1859368593685937, | |
| "grad_norm": 0.7682190537452698, | |
| "learning_rate": 6.067676906198574e-05, | |
| "loss": 0.32571834564208985, | |
| "step": 57850 | |
| }, | |
| { | |
| "epoch": 1.186961869618696, | |
| "grad_norm": 0.77012699842453, | |
| "learning_rate": 6.064248491497532e-05, | |
| "loss": 0.32980003356933596, | |
| "step": 57900 | |
| }, | |
| { | |
| "epoch": 1.1879868798687987, | |
| "grad_norm": 0.7941640019416809, | |
| "learning_rate": 6.0608200767964895e-05, | |
| "loss": 0.3314422607421875, | |
| "step": 57950 | |
| }, | |
| { | |
| "epoch": 1.189011890118901, | |
| "grad_norm": 0.7940207719802856, | |
| "learning_rate": 6.057391662095448e-05, | |
| "loss": 0.33427345275878906, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 1.1900369003690037, | |
| "grad_norm": 0.9056081175804138, | |
| "learning_rate": 6.053963247394405e-05, | |
| "loss": 0.3220090103149414, | |
| "step": 58050 | |
| }, | |
| { | |
| "epoch": 1.1910619106191063, | |
| "grad_norm": 0.7956807613372803, | |
| "learning_rate": 6.050534832693363e-05, | |
| "loss": 0.33605045318603516, | |
| "step": 58100 | |
| }, | |
| { | |
| "epoch": 1.1920869208692086, | |
| "grad_norm": 0.7694193720817566, | |
| "learning_rate": 6.047106417992321e-05, | |
| "loss": 0.32954719543457034, | |
| "step": 58150 | |
| }, | |
| { | |
| "epoch": 1.1931119311193112, | |
| "grad_norm": 0.8306771516799927, | |
| "learning_rate": 6.043678003291279e-05, | |
| "loss": 0.3237570571899414, | |
| "step": 58200 | |
| }, | |
| { | |
| "epoch": 1.1941369413694136, | |
| "grad_norm": 0.851830244064331, | |
| "learning_rate": 6.0402495885902364e-05, | |
| "loss": 0.33425289154052734, | |
| "step": 58250 | |
| }, | |
| { | |
| "epoch": 1.1951619516195162, | |
| "grad_norm": 0.820610761642456, | |
| "learning_rate": 6.0368211738891945e-05, | |
| "loss": 0.32722747802734375, | |
| "step": 58300 | |
| }, | |
| { | |
| "epoch": 1.1961869618696186, | |
| "grad_norm": 0.8808788061141968, | |
| "learning_rate": 6.033392759188151e-05, | |
| "loss": 0.33020065307617186, | |
| "step": 58350 | |
| }, | |
| { | |
| "epoch": 1.1972119721197212, | |
| "grad_norm": 0.7599654793739319, | |
| "learning_rate": 6.029964344487109e-05, | |
| "loss": 0.33122352600097654, | |
| "step": 58400 | |
| }, | |
| { | |
| "epoch": 1.1982369823698238, | |
| "grad_norm": 0.8254241943359375, | |
| "learning_rate": 6.026535929786067e-05, | |
| "loss": 0.33072624206542967, | |
| "step": 58450 | |
| }, | |
| { | |
| "epoch": 1.1992619926199262, | |
| "grad_norm": 0.791110634803772, | |
| "learning_rate": 6.0231075150850244e-05, | |
| "loss": 0.33087337493896485, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 1.2002870028700288, | |
| "grad_norm": 0.7769493460655212, | |
| "learning_rate": 6.0196791003839825e-05, | |
| "loss": 0.3358992385864258, | |
| "step": 58550 | |
| }, | |
| { | |
| "epoch": 1.2013120131201311, | |
| "grad_norm": 0.7905113101005554, | |
| "learning_rate": 6.016250685682941e-05, | |
| "loss": 0.33145919799804685, | |
| "step": 58600 | |
| }, | |
| { | |
| "epoch": 1.2023370233702337, | |
| "grad_norm": 0.8688026070594788, | |
| "learning_rate": 6.012822270981898e-05, | |
| "loss": 0.330220947265625, | |
| "step": 58650 | |
| }, | |
| { | |
| "epoch": 1.2033620336203361, | |
| "grad_norm": 0.810697615146637, | |
| "learning_rate": 6.009393856280856e-05, | |
| "loss": 0.33379665374755857, | |
| "step": 58700 | |
| }, | |
| { | |
| "epoch": 1.2043870438704387, | |
| "grad_norm": 0.8053363561630249, | |
| "learning_rate": 6.005965441579814e-05, | |
| "loss": 0.333529052734375, | |
| "step": 58750 | |
| }, | |
| { | |
| "epoch": 1.2054120541205413, | |
| "grad_norm": 0.8194394111633301, | |
| "learning_rate": 6.002537026878772e-05, | |
| "loss": 0.3333978271484375, | |
| "step": 58800 | |
| }, | |
| { | |
| "epoch": 1.2064370643706437, | |
| "grad_norm": 0.7876539826393127, | |
| "learning_rate": 5.9991086121777294e-05, | |
| "loss": 0.33216468811035155, | |
| "step": 58850 | |
| }, | |
| { | |
| "epoch": 1.2074620746207463, | |
| "grad_norm": 0.841381311416626, | |
| "learning_rate": 5.9956801974766875e-05, | |
| "loss": 0.33051036834716796, | |
| "step": 58900 | |
| }, | |
| { | |
| "epoch": 1.2084870848708487, | |
| "grad_norm": 0.768565833568573, | |
| "learning_rate": 5.992251782775645e-05, | |
| "loss": 0.33341762542724607, | |
| "step": 58950 | |
| }, | |
| { | |
| "epoch": 1.2095120951209513, | |
| "grad_norm": 0.8852304816246033, | |
| "learning_rate": 5.988823368074603e-05, | |
| "loss": 0.32806304931640623, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 1.2105371053710536, | |
| "grad_norm": 0.8292588591575623, | |
| "learning_rate": 5.98539495337356e-05, | |
| "loss": 0.33504638671875, | |
| "step": 59050 | |
| }, | |
| { | |
| "epoch": 1.2115621156211562, | |
| "grad_norm": 0.8059162497520447, | |
| "learning_rate": 5.9819665386725174e-05, | |
| "loss": 0.32167530059814453, | |
| "step": 59100 | |
| }, | |
| { | |
| "epoch": 1.2125871258712588, | |
| "grad_norm": 0.7463501691818237, | |
| "learning_rate": 5.9785381239714756e-05, | |
| "loss": 0.32884601593017576, | |
| "step": 59150 | |
| }, | |
| { | |
| "epoch": 1.2136121361213612, | |
| "grad_norm": 0.752579927444458, | |
| "learning_rate": 5.975109709270433e-05, | |
| "loss": 0.3293482208251953, | |
| "step": 59200 | |
| }, | |
| { | |
| "epoch": 1.2146371463714638, | |
| "grad_norm": 0.8358730673789978, | |
| "learning_rate": 5.971681294569391e-05, | |
| "loss": 0.3359712600708008, | |
| "step": 59250 | |
| }, | |
| { | |
| "epoch": 1.2156621566215662, | |
| "grad_norm": 0.8508303761482239, | |
| "learning_rate": 5.968252879868349e-05, | |
| "loss": 0.3283417892456055, | |
| "step": 59300 | |
| }, | |
| { | |
| "epoch": 1.2166871668716688, | |
| "grad_norm": 0.8609314560890198, | |
| "learning_rate": 5.964824465167307e-05, | |
| "loss": 0.3321079254150391, | |
| "step": 59350 | |
| }, | |
| { | |
| "epoch": 1.2177121771217712, | |
| "grad_norm": 0.8971276879310608, | |
| "learning_rate": 5.961396050466265e-05, | |
| "loss": 0.3250263214111328, | |
| "step": 59400 | |
| }, | |
| { | |
| "epoch": 1.2187371873718738, | |
| "grad_norm": 0.870012640953064, | |
| "learning_rate": 5.9579676357652224e-05, | |
| "loss": 0.32972404479980466, | |
| "step": 59450 | |
| }, | |
| { | |
| "epoch": 1.2197621976219761, | |
| "grad_norm": 0.7674804329872131, | |
| "learning_rate": 5.9545392210641806e-05, | |
| "loss": 0.33057853698730466, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 1.2207872078720787, | |
| "grad_norm": 0.8699909448623657, | |
| "learning_rate": 5.951110806363138e-05, | |
| "loss": 0.33232521057128905, | |
| "step": 59550 | |
| }, | |
| { | |
| "epoch": 1.221812218122181, | |
| "grad_norm": 0.8710799813270569, | |
| "learning_rate": 5.947682391662096e-05, | |
| "loss": 0.3280634307861328, | |
| "step": 59600 | |
| }, | |
| { | |
| "epoch": 1.2228372283722837, | |
| "grad_norm": 0.7630007266998291, | |
| "learning_rate": 5.9442539769610537e-05, | |
| "loss": 0.3303402328491211, | |
| "step": 59650 | |
| }, | |
| { | |
| "epoch": 1.2238622386223863, | |
| "grad_norm": 0.8189054727554321, | |
| "learning_rate": 5.940825562260012e-05, | |
| "loss": 0.32570648193359375, | |
| "step": 59700 | |
| }, | |
| { | |
| "epoch": 1.2248872488724887, | |
| "grad_norm": 0.8107175827026367, | |
| "learning_rate": 5.9373971475589686e-05, | |
| "loss": 0.3279362869262695, | |
| "step": 59750 | |
| }, | |
| { | |
| "epoch": 1.2259122591225913, | |
| "grad_norm": 0.8830100893974304, | |
| "learning_rate": 5.933968732857926e-05, | |
| "loss": 0.33183971405029294, | |
| "step": 59800 | |
| }, | |
| { | |
| "epoch": 1.2269372693726937, | |
| "grad_norm": 0.7496248483657837, | |
| "learning_rate": 5.930540318156884e-05, | |
| "loss": 0.32977775573730467, | |
| "step": 59850 | |
| }, | |
| { | |
| "epoch": 1.2279622796227962, | |
| "grad_norm": 0.8617159724235535, | |
| "learning_rate": 5.927111903455842e-05, | |
| "loss": 0.32601634979248045, | |
| "step": 59900 | |
| }, | |
| { | |
| "epoch": 1.2289872898728986, | |
| "grad_norm": 0.8602395057678223, | |
| "learning_rate": 5.9236834887548e-05, | |
| "loss": 0.3260711669921875, | |
| "step": 59950 | |
| }, | |
| { | |
| "epoch": 1.2300123001230012, | |
| "grad_norm": 0.749657392501831, | |
| "learning_rate": 5.920255074053758e-05, | |
| "loss": 0.325057258605957, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 1.2300123001230012, | |
| "eval_loss": 0.39810439944267273, | |
| "eval_runtime": 4.5099, | |
| "eval_samples_per_second": 5543.376, | |
| "eval_steps_per_second": 5.543, | |
| "eval_teacher_cosine_loss": 0.25358372926712036, | |
| "eval_teacher_cosine_mean": 0.7464162707328796, | |
| "eval_teacher_cosine_p10": 0.668035089969635, | |
| "eval_teacher_cosine_p50": 0.7564405202865601, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 1.2310373103731038, | |
| "grad_norm": 0.8218343257904053, | |
| "learning_rate": 5.9168266593527154e-05, | |
| "loss": 0.32841732025146486, | |
| "step": 60050 | |
| }, | |
| { | |
| "epoch": 1.2320623206232062, | |
| "grad_norm": 0.8128740191459656, | |
| "learning_rate": 5.9133982446516736e-05, | |
| "loss": 0.3253915786743164, | |
| "step": 60100 | |
| }, | |
| { | |
| "epoch": 1.2330873308733088, | |
| "grad_norm": 0.8702530860900879, | |
| "learning_rate": 5.909969829950631e-05, | |
| "loss": 0.33133655548095703, | |
| "step": 60150 | |
| }, | |
| { | |
| "epoch": 1.2341123411234112, | |
| "grad_norm": 0.8210967183113098, | |
| "learning_rate": 5.906541415249589e-05, | |
| "loss": 0.32437156677246093, | |
| "step": 60200 | |
| }, | |
| { | |
| "epoch": 1.2351373513735138, | |
| "grad_norm": 0.7465206980705261, | |
| "learning_rate": 5.903113000548547e-05, | |
| "loss": 0.33278926849365237, | |
| "step": 60250 | |
| }, | |
| { | |
| "epoch": 1.2361623616236161, | |
| "grad_norm": 0.8605986833572388, | |
| "learning_rate": 5.899684585847505e-05, | |
| "loss": 0.3351786804199219, | |
| "step": 60300 | |
| }, | |
| { | |
| "epoch": 1.2371873718737187, | |
| "grad_norm": 0.7875062227249146, | |
| "learning_rate": 5.896256171146462e-05, | |
| "loss": 0.3240493392944336, | |
| "step": 60350 | |
| }, | |
| { | |
| "epoch": 1.2382123821238213, | |
| "grad_norm": 0.7319639921188354, | |
| "learning_rate": 5.8928277564454204e-05, | |
| "loss": 0.3273191452026367, | |
| "step": 60400 | |
| }, | |
| { | |
| "epoch": 1.2392373923739237, | |
| "grad_norm": 0.8413223028182983, | |
| "learning_rate": 5.889399341744377e-05, | |
| "loss": 0.3261093521118164, | |
| "step": 60450 | |
| }, | |
| { | |
| "epoch": 1.2402624026240263, | |
| "grad_norm": 0.81492680311203, | |
| "learning_rate": 5.885970927043335e-05, | |
| "loss": 0.3290855407714844, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 1.2412874128741287, | |
| "grad_norm": 0.7862895727157593, | |
| "learning_rate": 5.882542512342293e-05, | |
| "loss": 0.32848175048828127, | |
| "step": 60550 | |
| }, | |
| { | |
| "epoch": 1.2423124231242313, | |
| "grad_norm": 0.8700463175773621, | |
| "learning_rate": 5.87911409764125e-05, | |
| "loss": 0.32747100830078124, | |
| "step": 60600 | |
| }, | |
| { | |
| "epoch": 1.2433374333743337, | |
| "grad_norm": 0.8070055842399597, | |
| "learning_rate": 5.8756856829402085e-05, | |
| "loss": 0.3304486083984375, | |
| "step": 60650 | |
| }, | |
| { | |
| "epoch": 1.2443624436244363, | |
| "grad_norm": 0.8466408848762512, | |
| "learning_rate": 5.8722572682391666e-05, | |
| "loss": 0.3225171661376953, | |
| "step": 60700 | |
| }, | |
| { | |
| "epoch": 1.2453874538745389, | |
| "grad_norm": 0.7944824695587158, | |
| "learning_rate": 5.868828853538124e-05, | |
| "loss": 0.32632110595703123, | |
| "step": 60750 | |
| }, | |
| { | |
| "epoch": 1.2464124641246412, | |
| "grad_norm": 0.8876914978027344, | |
| "learning_rate": 5.865400438837082e-05, | |
| "loss": 0.326151123046875, | |
| "step": 60800 | |
| }, | |
| { | |
| "epoch": 1.2474374743747438, | |
| "grad_norm": 0.8217604160308838, | |
| "learning_rate": 5.86197202413604e-05, | |
| "loss": 0.32424713134765626, | |
| "step": 60850 | |
| }, | |
| { | |
| "epoch": 1.2484624846248462, | |
| "grad_norm": 0.7867749929428101, | |
| "learning_rate": 5.858543609434998e-05, | |
| "loss": 0.33268280029296876, | |
| "step": 60900 | |
| }, | |
| { | |
| "epoch": 1.2494874948749488, | |
| "grad_norm": 0.8116174340248108, | |
| "learning_rate": 5.855115194733955e-05, | |
| "loss": 0.3296493148803711, | |
| "step": 60950 | |
| }, | |
| { | |
| "epoch": 1.2505125051250512, | |
| "grad_norm": 0.8088791370391846, | |
| "learning_rate": 5.8516867800329135e-05, | |
| "loss": 0.32690437316894533, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 1.2515375153751538, | |
| "grad_norm": 0.8264136910438538, | |
| "learning_rate": 5.848258365331871e-05, | |
| "loss": 0.3257485580444336, | |
| "step": 61050 | |
| }, | |
| { | |
| "epoch": 1.2525625256252564, | |
| "grad_norm": 0.8766059279441833, | |
| "learning_rate": 5.844829950630829e-05, | |
| "loss": 0.32831268310546874, | |
| "step": 61100 | |
| }, | |
| { | |
| "epoch": 1.2535875358753588, | |
| "grad_norm": 0.9072951078414917, | |
| "learning_rate": 5.841401535929787e-05, | |
| "loss": 0.3331643676757812, | |
| "step": 61150 | |
| }, | |
| { | |
| "epoch": 1.2546125461254611, | |
| "grad_norm": 0.8960299491882324, | |
| "learning_rate": 5.8379731212287433e-05, | |
| "loss": 0.32320083618164064, | |
| "step": 61200 | |
| }, | |
| { | |
| "epoch": 1.2556375563755637, | |
| "grad_norm": 0.7984836101531982, | |
| "learning_rate": 5.8345447065277015e-05, | |
| "loss": 0.3268982696533203, | |
| "step": 61250 | |
| }, | |
| { | |
| "epoch": 1.2566625666256663, | |
| "grad_norm": 0.9406589269638062, | |
| "learning_rate": 5.8311162918266596e-05, | |
| "loss": 0.3303249740600586, | |
| "step": 61300 | |
| }, | |
| { | |
| "epoch": 1.2576875768757687, | |
| "grad_norm": 0.748437762260437, | |
| "learning_rate": 5.827687877125617e-05, | |
| "loss": 0.3233198165893555, | |
| "step": 61350 | |
| }, | |
| { | |
| "epoch": 1.2587125871258713, | |
| "grad_norm": 0.761820912361145, | |
| "learning_rate": 5.824259462424575e-05, | |
| "loss": 0.32345603942871093, | |
| "step": 61400 | |
| }, | |
| { | |
| "epoch": 1.2597375973759737, | |
| "grad_norm": 0.8321351408958435, | |
| "learning_rate": 5.820831047723533e-05, | |
| "loss": 0.3252406692504883, | |
| "step": 61450 | |
| }, | |
| { | |
| "epoch": 1.2607626076260763, | |
| "grad_norm": 0.8007746338844299, | |
| "learning_rate": 5.817402633022491e-05, | |
| "loss": 0.3279948043823242, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 1.2617876178761787, | |
| "grad_norm": 0.8780990242958069, | |
| "learning_rate": 5.8139742183214483e-05, | |
| "loss": 0.3284341049194336, | |
| "step": 61550 | |
| }, | |
| { | |
| "epoch": 1.2628126281262813, | |
| "grad_norm": 0.7659162878990173, | |
| "learning_rate": 5.8105458036204065e-05, | |
| "loss": 0.3275317764282227, | |
| "step": 61600 | |
| }, | |
| { | |
| "epoch": 1.2638376383763839, | |
| "grad_norm": 0.8799045085906982, | |
| "learning_rate": 5.807117388919364e-05, | |
| "loss": 0.32882537841796877, | |
| "step": 61650 | |
| }, | |
| { | |
| "epoch": 1.2648626486264862, | |
| "grad_norm": 0.8847517371177673, | |
| "learning_rate": 5.803688974218322e-05, | |
| "loss": 0.330797004699707, | |
| "step": 61700 | |
| }, | |
| { | |
| "epoch": 1.2658876588765888, | |
| "grad_norm": 0.8502843379974365, | |
| "learning_rate": 5.8002605595172796e-05, | |
| "loss": 0.32946624755859377, | |
| "step": 61750 | |
| }, | |
| { | |
| "epoch": 1.2669126691266912, | |
| "grad_norm": 0.882281482219696, | |
| "learning_rate": 5.796832144816238e-05, | |
| "loss": 0.32732337951660156, | |
| "step": 61800 | |
| }, | |
| { | |
| "epoch": 1.2679376793767938, | |
| "grad_norm": 0.7853183746337891, | |
| "learning_rate": 5.793403730115196e-05, | |
| "loss": 0.32719070434570313, | |
| "step": 61850 | |
| }, | |
| { | |
| "epoch": 1.2689626896268962, | |
| "grad_norm": 0.7982047200202942, | |
| "learning_rate": 5.789975315414152e-05, | |
| "loss": 0.3261737442016602, | |
| "step": 61900 | |
| }, | |
| { | |
| "epoch": 1.2699876998769988, | |
| "grad_norm": 0.7903969883918762, | |
| "learning_rate": 5.78654690071311e-05, | |
| "loss": 0.3306828308105469, | |
| "step": 61950 | |
| }, | |
| { | |
| "epoch": 1.2710127101271014, | |
| "grad_norm": 0.7593696713447571, | |
| "learning_rate": 5.783118486012068e-05, | |
| "loss": 0.32361331939697263, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 1.2720377203772038, | |
| "grad_norm": 0.8642659783363342, | |
| "learning_rate": 5.779690071311026e-05, | |
| "loss": 0.33044631958007814, | |
| "step": 62050 | |
| }, | |
| { | |
| "epoch": 1.2730627306273063, | |
| "grad_norm": 0.8054972887039185, | |
| "learning_rate": 5.776261656609984e-05, | |
| "loss": 0.3259759521484375, | |
| "step": 62100 | |
| }, | |
| { | |
| "epoch": 1.2740877408774087, | |
| "grad_norm": 0.7991418838500977, | |
| "learning_rate": 5.7728332419089414e-05, | |
| "loss": 0.32941570281982424, | |
| "step": 62150 | |
| }, | |
| { | |
| "epoch": 1.2751127511275113, | |
| "grad_norm": 0.8188523650169373, | |
| "learning_rate": 5.7694048272078995e-05, | |
| "loss": 0.33065425872802734, | |
| "step": 62200 | |
| }, | |
| { | |
| "epoch": 1.2761377613776137, | |
| "grad_norm": 0.8273463845252991, | |
| "learning_rate": 5.765976412506857e-05, | |
| "loss": 0.3247711181640625, | |
| "step": 62250 | |
| }, | |
| { | |
| "epoch": 1.2771627716277163, | |
| "grad_norm": 0.8663382530212402, | |
| "learning_rate": 5.762547997805815e-05, | |
| "loss": 0.3324728012084961, | |
| "step": 62300 | |
| }, | |
| { | |
| "epoch": 1.278187781877819, | |
| "grad_norm": 0.8284932971000671, | |
| "learning_rate": 5.7591195831047726e-05, | |
| "loss": 0.3265801239013672, | |
| "step": 62350 | |
| }, | |
| { | |
| "epoch": 1.2792127921279213, | |
| "grad_norm": 0.8603318929672241, | |
| "learning_rate": 5.755691168403731e-05, | |
| "loss": 0.3292204284667969, | |
| "step": 62400 | |
| }, | |
| { | |
| "epoch": 1.2802378023780236, | |
| "grad_norm": 0.8416708111763, | |
| "learning_rate": 5.752262753702688e-05, | |
| "loss": 0.3335953140258789, | |
| "step": 62450 | |
| }, | |
| { | |
| "epoch": 1.2812628126281262, | |
| "grad_norm": 0.9232722520828247, | |
| "learning_rate": 5.7488343390016464e-05, | |
| "loss": 0.32983375549316407, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 1.2822878228782288, | |
| "grad_norm": 0.8240601420402527, | |
| "learning_rate": 5.7454059243006045e-05, | |
| "loss": 0.3208028793334961, | |
| "step": 62550 | |
| }, | |
| { | |
| "epoch": 1.2833128331283312, | |
| "grad_norm": 0.8200990557670593, | |
| "learning_rate": 5.7419775095995606e-05, | |
| "loss": 0.3276394271850586, | |
| "step": 62600 | |
| }, | |
| { | |
| "epoch": 1.2843378433784338, | |
| "grad_norm": 0.7965755462646484, | |
| "learning_rate": 5.738549094898519e-05, | |
| "loss": 0.32493370056152343, | |
| "step": 62650 | |
| }, | |
| { | |
| "epoch": 1.2853628536285364, | |
| "grad_norm": 0.906743586063385, | |
| "learning_rate": 5.735120680197477e-05, | |
| "loss": 0.32568138122558593, | |
| "step": 62700 | |
| }, | |
| { | |
| "epoch": 1.2863878638786388, | |
| "grad_norm": 0.8408162593841553, | |
| "learning_rate": 5.7316922654964344e-05, | |
| "loss": 0.32078582763671876, | |
| "step": 62750 | |
| }, | |
| { | |
| "epoch": 1.2874128741287412, | |
| "grad_norm": 0.8064471483230591, | |
| "learning_rate": 5.7282638507953925e-05, | |
| "loss": 0.32127185821533205, | |
| "step": 62800 | |
| }, | |
| { | |
| "epoch": 1.2884378843788438, | |
| "grad_norm": 0.8866981863975525, | |
| "learning_rate": 5.72483543609435e-05, | |
| "loss": 0.3302082824707031, | |
| "step": 62850 | |
| }, | |
| { | |
| "epoch": 1.2894628946289464, | |
| "grad_norm": 0.8829692602157593, | |
| "learning_rate": 5.721407021393308e-05, | |
| "loss": 0.33002376556396484, | |
| "step": 62900 | |
| }, | |
| { | |
| "epoch": 1.2904879048790487, | |
| "grad_norm": 0.8297115564346313, | |
| "learning_rate": 5.7179786066922656e-05, | |
| "loss": 0.326523551940918, | |
| "step": 62950 | |
| }, | |
| { | |
| "epoch": 1.2915129151291513, | |
| "grad_norm": 0.7863450050354004, | |
| "learning_rate": 5.714550191991224e-05, | |
| "loss": 0.32475475311279295, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 1.2925379253792537, | |
| "grad_norm": 0.8154489398002625, | |
| "learning_rate": 5.711121777290181e-05, | |
| "loss": 0.3230597686767578, | |
| "step": 63050 | |
| }, | |
| { | |
| "epoch": 1.2935629356293563, | |
| "grad_norm": 0.8872915506362915, | |
| "learning_rate": 5.7076933625891394e-05, | |
| "loss": 0.3317874908447266, | |
| "step": 63100 | |
| }, | |
| { | |
| "epoch": 1.2945879458794587, | |
| "grad_norm": 0.8664490580558777, | |
| "learning_rate": 5.704264947888097e-05, | |
| "loss": 0.3267560195922852, | |
| "step": 63150 | |
| }, | |
| { | |
| "epoch": 1.2956129561295613, | |
| "grad_norm": 0.7445337772369385, | |
| "learning_rate": 5.700836533187055e-05, | |
| "loss": 0.32788379669189455, | |
| "step": 63200 | |
| }, | |
| { | |
| "epoch": 1.2966379663796639, | |
| "grad_norm": 0.885084331035614, | |
| "learning_rate": 5.697408118486013e-05, | |
| "loss": 0.32551509857177735, | |
| "step": 63250 | |
| }, | |
| { | |
| "epoch": 1.2976629766297663, | |
| "grad_norm": 0.8188558220863342, | |
| "learning_rate": 5.693979703784969e-05, | |
| "loss": 0.32466777801513674, | |
| "step": 63300 | |
| }, | |
| { | |
| "epoch": 1.2986879868798689, | |
| "grad_norm": 0.8803619146347046, | |
| "learning_rate": 5.6905512890839274e-05, | |
| "loss": 0.32592967987060545, | |
| "step": 63350 | |
| }, | |
| { | |
| "epoch": 1.2997129971299712, | |
| "grad_norm": 0.866905927658081, | |
| "learning_rate": 5.6871228743828856e-05, | |
| "loss": 0.324463996887207, | |
| "step": 63400 | |
| }, | |
| { | |
| "epoch": 1.3007380073800738, | |
| "grad_norm": 0.8225587010383606, | |
| "learning_rate": 5.683694459681843e-05, | |
| "loss": 0.3299306869506836, | |
| "step": 63450 | |
| }, | |
| { | |
| "epoch": 1.3017630176301762, | |
| "grad_norm": 0.8648690581321716, | |
| "learning_rate": 5.680266044980801e-05, | |
| "loss": 0.3305857849121094, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 1.3027880278802788, | |
| "grad_norm": 0.8619871735572815, | |
| "learning_rate": 5.6768376302797586e-05, | |
| "loss": 0.3247836685180664, | |
| "step": 63550 | |
| }, | |
| { | |
| "epoch": 1.3038130381303814, | |
| "grad_norm": 0.8033298850059509, | |
| "learning_rate": 5.673409215578717e-05, | |
| "loss": 0.32981861114501954, | |
| "step": 63600 | |
| }, | |
| { | |
| "epoch": 1.3048380483804838, | |
| "grad_norm": 0.7464938759803772, | |
| "learning_rate": 5.669980800877674e-05, | |
| "loss": 0.3258791351318359, | |
| "step": 63650 | |
| }, | |
| { | |
| "epoch": 1.3058630586305864, | |
| "grad_norm": 0.7822047472000122, | |
| "learning_rate": 5.6665523861766324e-05, | |
| "loss": 0.33139286041259763, | |
| "step": 63700 | |
| }, | |
| { | |
| "epoch": 1.3068880688806888, | |
| "grad_norm": 0.8651247024536133, | |
| "learning_rate": 5.66312397147559e-05, | |
| "loss": 0.3299574279785156, | |
| "step": 63750 | |
| }, | |
| { | |
| "epoch": 1.3079130791307914, | |
| "grad_norm": 0.753192126750946, | |
| "learning_rate": 5.659695556774548e-05, | |
| "loss": 0.3282101058959961, | |
| "step": 63800 | |
| }, | |
| { | |
| "epoch": 1.3089380893808937, | |
| "grad_norm": 0.8179863095283508, | |
| "learning_rate": 5.6562671420735055e-05, | |
| "loss": 0.3286818313598633, | |
| "step": 63850 | |
| }, | |
| { | |
| "epoch": 1.3099630996309963, | |
| "grad_norm": 0.792097270488739, | |
| "learning_rate": 5.6528387273724636e-05, | |
| "loss": 0.3291008377075195, | |
| "step": 63900 | |
| }, | |
| { | |
| "epoch": 1.310988109881099, | |
| "grad_norm": 0.7775896787643433, | |
| "learning_rate": 5.649410312671422e-05, | |
| "loss": 0.32300037384033203, | |
| "step": 63950 | |
| }, | |
| { | |
| "epoch": 1.3120131201312013, | |
| "grad_norm": 0.7928282022476196, | |
| "learning_rate": 5.645981897970378e-05, | |
| "loss": 0.33330310821533204, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 1.3130381303813037, | |
| "grad_norm": 0.8812917470932007, | |
| "learning_rate": 5.642553483269336e-05, | |
| "loss": 0.3244797897338867, | |
| "step": 64050 | |
| }, | |
| { | |
| "epoch": 1.3140631406314063, | |
| "grad_norm": 0.850044846534729, | |
| "learning_rate": 5.639125068568294e-05, | |
| "loss": 0.32575374603271484, | |
| "step": 64100 | |
| }, | |
| { | |
| "epoch": 1.3150881508815089, | |
| "grad_norm": 0.8299598097801208, | |
| "learning_rate": 5.635696653867252e-05, | |
| "loss": 0.32380908966064453, | |
| "step": 64150 | |
| }, | |
| { | |
| "epoch": 1.3161131611316113, | |
| "grad_norm": 0.7505680322647095, | |
| "learning_rate": 5.63226823916621e-05, | |
| "loss": 0.3237628173828125, | |
| "step": 64200 | |
| }, | |
| { | |
| "epoch": 1.3171381713817139, | |
| "grad_norm": 0.8305618762969971, | |
| "learning_rate": 5.628839824465167e-05, | |
| "loss": 0.33267108917236327, | |
| "step": 64250 | |
| }, | |
| { | |
| "epoch": 1.3181631816318164, | |
| "grad_norm": 0.7733376026153564, | |
| "learning_rate": 5.6254114097641254e-05, | |
| "loss": 0.3252303314208984, | |
| "step": 64300 | |
| }, | |
| { | |
| "epoch": 1.3191881918819188, | |
| "grad_norm": 0.852988064289093, | |
| "learning_rate": 5.621982995063083e-05, | |
| "loss": 0.32650276184082033, | |
| "step": 64350 | |
| }, | |
| { | |
| "epoch": 1.3202132021320212, | |
| "grad_norm": 0.895401120185852, | |
| "learning_rate": 5.618554580362041e-05, | |
| "loss": 0.3286096954345703, | |
| "step": 64400 | |
| }, | |
| { | |
| "epoch": 1.3212382123821238, | |
| "grad_norm": 0.8217565417289734, | |
| "learning_rate": 5.6151261656609985e-05, | |
| "loss": 0.3232681655883789, | |
| "step": 64450 | |
| }, | |
| { | |
| "epoch": 1.3222632226322264, | |
| "grad_norm": 0.9198343753814697, | |
| "learning_rate": 5.611697750959957e-05, | |
| "loss": 0.32986549377441404, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 1.3232882328823288, | |
| "grad_norm": 0.8182905316352844, | |
| "learning_rate": 5.608269336258915e-05, | |
| "loss": 0.32429985046386717, | |
| "step": 64550 | |
| }, | |
| { | |
| "epoch": 1.3243132431324314, | |
| "grad_norm": 0.8533055186271667, | |
| "learning_rate": 5.604840921557872e-05, | |
| "loss": 0.3250510787963867, | |
| "step": 64600 | |
| }, | |
| { | |
| "epoch": 1.3253382533825337, | |
| "grad_norm": 0.7308329343795776, | |
| "learning_rate": 5.6014125068568304e-05, | |
| "loss": 0.3276877593994141, | |
| "step": 64650 | |
| }, | |
| { | |
| "epoch": 1.3263632636326363, | |
| "grad_norm": 0.79646235704422, | |
| "learning_rate": 5.597984092155787e-05, | |
| "loss": 0.3235052490234375, | |
| "step": 64700 | |
| }, | |
| { | |
| "epoch": 1.3273882738827387, | |
| "grad_norm": 0.8857685923576355, | |
| "learning_rate": 5.594555677454745e-05, | |
| "loss": 0.32531276702880857, | |
| "step": 64750 | |
| }, | |
| { | |
| "epoch": 1.3284132841328413, | |
| "grad_norm": 0.8716639280319214, | |
| "learning_rate": 5.591127262753703e-05, | |
| "loss": 0.32700321197509763, | |
| "step": 64800 | |
| }, | |
| { | |
| "epoch": 1.329438294382944, | |
| "grad_norm": 0.9236313700675964, | |
| "learning_rate": 5.58769884805266e-05, | |
| "loss": 0.32999488830566404, | |
| "step": 64850 | |
| }, | |
| { | |
| "epoch": 1.3304633046330463, | |
| "grad_norm": 0.768204927444458, | |
| "learning_rate": 5.5842704333516185e-05, | |
| "loss": 0.32388065338134764, | |
| "step": 64900 | |
| }, | |
| { | |
| "epoch": 1.331488314883149, | |
| "grad_norm": 0.9428282976150513, | |
| "learning_rate": 5.580842018650576e-05, | |
| "loss": 0.3244235610961914, | |
| "step": 64950 | |
| }, | |
| { | |
| "epoch": 1.3325133251332513, | |
| "grad_norm": 0.8863661289215088, | |
| "learning_rate": 5.577413603949534e-05, | |
| "loss": 0.32580875396728515, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 1.3325133251332513, | |
| "eval_loss": 0.39588069915771484, | |
| "eval_runtime": 4.4902, | |
| "eval_samples_per_second": 5567.628, | |
| "eval_steps_per_second": 5.568, | |
| "eval_teacher_cosine_loss": 0.25079721212387085, | |
| "eval_teacher_cosine_mean": 0.7492027878761292, | |
| "eval_teacher_cosine_p10": 0.6726008057594299, | |
| "eval_teacher_cosine_p50": 0.7587714195251465, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 1.3335383353833539, | |
| "grad_norm": 0.833237886428833, | |
| "learning_rate": 5.5739851892484915e-05, | |
| "loss": 0.3242361068725586, | |
| "step": 65050 | |
| }, | |
| { | |
| "epoch": 1.3345633456334562, | |
| "grad_norm": 0.88962721824646, | |
| "learning_rate": 5.57055677454745e-05, | |
| "loss": 0.32839942932128907, | |
| "step": 65100 | |
| }, | |
| { | |
| "epoch": 1.3355883558835588, | |
| "grad_norm": 0.8793926239013672, | |
| "learning_rate": 5.567128359846407e-05, | |
| "loss": 0.32906513214111327, | |
| "step": 65150 | |
| }, | |
| { | |
| "epoch": 1.3366133661336614, | |
| "grad_norm": 0.819770336151123, | |
| "learning_rate": 5.563699945145365e-05, | |
| "loss": 0.32637088775634765, | |
| "step": 65200 | |
| }, | |
| { | |
| "epoch": 1.3376383763837638, | |
| "grad_norm": 0.8931409120559692, | |
| "learning_rate": 5.5602715304443235e-05, | |
| "loss": 0.3282789611816406, | |
| "step": 65250 | |
| }, | |
| { | |
| "epoch": 1.3386633866338664, | |
| "grad_norm": 0.814875066280365, | |
| "learning_rate": 5.556843115743281e-05, | |
| "loss": 0.33237514495849607, | |
| "step": 65300 | |
| }, | |
| { | |
| "epoch": 1.3396883968839688, | |
| "grad_norm": 0.7590654492378235, | |
| "learning_rate": 5.553414701042239e-05, | |
| "loss": 0.3232054138183594, | |
| "step": 65350 | |
| }, | |
| { | |
| "epoch": 1.3407134071340714, | |
| "grad_norm": 0.8904417753219604, | |
| "learning_rate": 5.549986286341196e-05, | |
| "loss": 0.32497577667236327, | |
| "step": 65400 | |
| }, | |
| { | |
| "epoch": 1.3417384173841738, | |
| "grad_norm": 0.9005507230758667, | |
| "learning_rate": 5.5465578716401533e-05, | |
| "loss": 0.3207798767089844, | |
| "step": 65450 | |
| }, | |
| { | |
| "epoch": 1.3427634276342764, | |
| "grad_norm": 0.8197437524795532, | |
| "learning_rate": 5.5431294569391115e-05, | |
| "loss": 0.3246816253662109, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 1.343788437884379, | |
| "grad_norm": 0.787700891494751, | |
| "learning_rate": 5.539701042238069e-05, | |
| "loss": 0.3220206832885742, | |
| "step": 65550 | |
| }, | |
| { | |
| "epoch": 1.3448134481344813, | |
| "grad_norm": 0.9401196837425232, | |
| "learning_rate": 5.536272627537027e-05, | |
| "loss": 0.32266864776611326, | |
| "step": 65600 | |
| }, | |
| { | |
| "epoch": 1.3458384583845837, | |
| "grad_norm": 0.8708605170249939, | |
| "learning_rate": 5.5328442128359846e-05, | |
| "loss": 0.3265773010253906, | |
| "step": 65650 | |
| }, | |
| { | |
| "epoch": 1.3468634686346863, | |
| "grad_norm": 0.8541357517242432, | |
| "learning_rate": 5.529415798134943e-05, | |
| "loss": 0.32489498138427736, | |
| "step": 65700 | |
| }, | |
| { | |
| "epoch": 1.347888478884789, | |
| "grad_norm": 0.776192843914032, | |
| "learning_rate": 5.5259873834339e-05, | |
| "loss": 0.3257966613769531, | |
| "step": 65750 | |
| }, | |
| { | |
| "epoch": 1.3489134891348913, | |
| "grad_norm": 0.8461732268333435, | |
| "learning_rate": 5.522558968732858e-05, | |
| "loss": 0.3218109130859375, | |
| "step": 65800 | |
| }, | |
| { | |
| "epoch": 1.3499384993849939, | |
| "grad_norm": 0.8257558941841125, | |
| "learning_rate": 5.519130554031816e-05, | |
| "loss": 0.33038642883300784, | |
| "step": 65850 | |
| }, | |
| { | |
| "epoch": 1.3509635096350965, | |
| "grad_norm": 0.7846958637237549, | |
| "learning_rate": 5.515702139330774e-05, | |
| "loss": 0.32293087005615234, | |
| "step": 65900 | |
| }, | |
| { | |
| "epoch": 1.3519885198851989, | |
| "grad_norm": 0.743119478225708, | |
| "learning_rate": 5.512273724629732e-05, | |
| "loss": 0.32599189758300784, | |
| "step": 65950 | |
| }, | |
| { | |
| "epoch": 1.3530135301353012, | |
| "grad_norm": 0.8826002478599548, | |
| "learning_rate": 5.5088453099286896e-05, | |
| "loss": 0.32416046142578125, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 1.3540385403854038, | |
| "grad_norm": 0.7558027505874634, | |
| "learning_rate": 5.505416895227648e-05, | |
| "loss": 0.32051322937011717, | |
| "step": 66050 | |
| }, | |
| { | |
| "epoch": 1.3550635506355064, | |
| "grad_norm": 0.766213595867157, | |
| "learning_rate": 5.501988480526605e-05, | |
| "loss": 0.31972488403320315, | |
| "step": 66100 | |
| }, | |
| { | |
| "epoch": 1.3560885608856088, | |
| "grad_norm": 0.8968978524208069, | |
| "learning_rate": 5.498560065825562e-05, | |
| "loss": 0.32700267791748044, | |
| "step": 66150 | |
| }, | |
| { | |
| "epoch": 1.3571135711357114, | |
| "grad_norm": 0.8148743510246277, | |
| "learning_rate": 5.49513165112452e-05, | |
| "loss": 0.3243909072875977, | |
| "step": 66200 | |
| }, | |
| { | |
| "epoch": 1.3581385813858138, | |
| "grad_norm": 0.8513030409812927, | |
| "learning_rate": 5.4917032364234776e-05, | |
| "loss": 0.3275979232788086, | |
| "step": 66250 | |
| }, | |
| { | |
| "epoch": 1.3591635916359164, | |
| "grad_norm": 0.7605785131454468, | |
| "learning_rate": 5.488274821722436e-05, | |
| "loss": 0.32051094055175783, | |
| "step": 66300 | |
| }, | |
| { | |
| "epoch": 1.3601886018860188, | |
| "grad_norm": 0.842176079750061, | |
| "learning_rate": 5.484846407021393e-05, | |
| "loss": 0.33005111694335937, | |
| "step": 66350 | |
| }, | |
| { | |
| "epoch": 1.3612136121361214, | |
| "grad_norm": 0.8882763385772705, | |
| "learning_rate": 5.4814179923203514e-05, | |
| "loss": 0.3262485122680664, | |
| "step": 66400 | |
| }, | |
| { | |
| "epoch": 1.362238622386224, | |
| "grad_norm": 0.8196696639060974, | |
| "learning_rate": 5.477989577619309e-05, | |
| "loss": 0.32551109313964843, | |
| "step": 66450 | |
| }, | |
| { | |
| "epoch": 1.3632636326363263, | |
| "grad_norm": 0.8026140928268433, | |
| "learning_rate": 5.474561162918267e-05, | |
| "loss": 0.32536327362060546, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 1.364288642886429, | |
| "grad_norm": 0.8378697633743286, | |
| "learning_rate": 5.4711327482172244e-05, | |
| "loss": 0.3225536727905273, | |
| "step": 66550 | |
| }, | |
| { | |
| "epoch": 1.3653136531365313, | |
| "grad_norm": 0.8614795207977295, | |
| "learning_rate": 5.4677043335161826e-05, | |
| "loss": 0.32576980590820315, | |
| "step": 66600 | |
| }, | |
| { | |
| "epoch": 1.366338663386634, | |
| "grad_norm": 0.878044843673706, | |
| "learning_rate": 5.464275918815141e-05, | |
| "loss": 0.32534324645996093, | |
| "step": 66650 | |
| }, | |
| { | |
| "epoch": 1.3673636736367363, | |
| "grad_norm": 0.9303902983665466, | |
| "learning_rate": 5.460847504114098e-05, | |
| "loss": 0.328780403137207, | |
| "step": 66700 | |
| }, | |
| { | |
| "epoch": 1.3683886838868389, | |
| "grad_norm": 0.837350070476532, | |
| "learning_rate": 5.4574190894130564e-05, | |
| "loss": 0.3230078125, | |
| "step": 66750 | |
| }, | |
| { | |
| "epoch": 1.3694136941369415, | |
| "grad_norm": 0.8433408737182617, | |
| "learning_rate": 5.453990674712014e-05, | |
| "loss": 0.32327033996582033, | |
| "step": 66800 | |
| }, | |
| { | |
| "epoch": 1.3704387043870438, | |
| "grad_norm": 0.7934775352478027, | |
| "learning_rate": 5.4505622600109706e-05, | |
| "loss": 0.32381595611572267, | |
| "step": 66850 | |
| }, | |
| { | |
| "epoch": 1.3714637146371464, | |
| "grad_norm": 0.8823449611663818, | |
| "learning_rate": 5.447133845309929e-05, | |
| "loss": 0.325301399230957, | |
| "step": 66900 | |
| }, | |
| { | |
| "epoch": 1.3724887248872488, | |
| "grad_norm": 0.7905043959617615, | |
| "learning_rate": 5.443705430608886e-05, | |
| "loss": 0.32528247833251955, | |
| "step": 66950 | |
| }, | |
| { | |
| "epoch": 1.3735137351373514, | |
| "grad_norm": 0.8246670961380005, | |
| "learning_rate": 5.4402770159078444e-05, | |
| "loss": 0.3306504440307617, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 1.3745387453874538, | |
| "grad_norm": 0.8536582589149475, | |
| "learning_rate": 5.436848601206802e-05, | |
| "loss": 0.3234693145751953, | |
| "step": 67050 | |
| }, | |
| { | |
| "epoch": 1.3755637556375564, | |
| "grad_norm": 0.7841454744338989, | |
| "learning_rate": 5.43342018650576e-05, | |
| "loss": 0.3267879867553711, | |
| "step": 67100 | |
| }, | |
| { | |
| "epoch": 1.376588765887659, | |
| "grad_norm": 0.812348484992981, | |
| "learning_rate": 5.4299917718047175e-05, | |
| "loss": 0.32729995727539063, | |
| "step": 67150 | |
| }, | |
| { | |
| "epoch": 1.3776137761377614, | |
| "grad_norm": 0.8351894021034241, | |
| "learning_rate": 5.4265633571036756e-05, | |
| "loss": 0.32767864227294924, | |
| "step": 67200 | |
| }, | |
| { | |
| "epoch": 1.3786387863878637, | |
| "grad_norm": 0.7939273715019226, | |
| "learning_rate": 5.423134942402633e-05, | |
| "loss": 0.32908283233642577, | |
| "step": 67250 | |
| }, | |
| { | |
| "epoch": 1.3796637966379663, | |
| "grad_norm": 0.7924728989601135, | |
| "learning_rate": 5.419706527701591e-05, | |
| "loss": 0.3241147994995117, | |
| "step": 67300 | |
| }, | |
| { | |
| "epoch": 1.380688806888069, | |
| "grad_norm": 0.7914777398109436, | |
| "learning_rate": 5.4162781130005494e-05, | |
| "loss": 0.32814430236816405, | |
| "step": 67350 | |
| }, | |
| { | |
| "epoch": 1.3817138171381713, | |
| "grad_norm": 0.7896748185157776, | |
| "learning_rate": 5.412849698299507e-05, | |
| "loss": 0.3303387451171875, | |
| "step": 67400 | |
| }, | |
| { | |
| "epoch": 1.382738827388274, | |
| "grad_norm": 0.7297735810279846, | |
| "learning_rate": 5.409421283598465e-05, | |
| "loss": 0.32080089569091796, | |
| "step": 67450 | |
| }, | |
| { | |
| "epoch": 1.3837638376383765, | |
| "grad_norm": 0.7153830528259277, | |
| "learning_rate": 5.4059928688974225e-05, | |
| "loss": 0.3245745849609375, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 1.384788847888479, | |
| "grad_norm": 0.8354848623275757, | |
| "learning_rate": 5.402564454196379e-05, | |
| "loss": 0.327910041809082, | |
| "step": 67550 | |
| }, | |
| { | |
| "epoch": 1.3858138581385813, | |
| "grad_norm": 0.8918254375457764, | |
| "learning_rate": 5.3991360394953374e-05, | |
| "loss": 0.3211099624633789, | |
| "step": 67600 | |
| }, | |
| { | |
| "epoch": 1.3868388683886839, | |
| "grad_norm": 0.8696297407150269, | |
| "learning_rate": 5.395707624794295e-05, | |
| "loss": 0.32040863037109374, | |
| "step": 67650 | |
| }, | |
| { | |
| "epoch": 1.3878638786387865, | |
| "grad_norm": 0.8778762817382812, | |
| "learning_rate": 5.392279210093253e-05, | |
| "loss": 0.3228469467163086, | |
| "step": 67700 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.9433466196060181, | |
| "learning_rate": 5.3888507953922105e-05, | |
| "loss": 0.3228094482421875, | |
| "step": 67750 | |
| }, | |
| { | |
| "epoch": 1.3899138991389914, | |
| "grad_norm": 0.7952296733856201, | |
| "learning_rate": 5.3854223806911686e-05, | |
| "loss": 0.3224370574951172, | |
| "step": 67800 | |
| }, | |
| { | |
| "epoch": 1.3909389093890938, | |
| "grad_norm": 0.8082793951034546, | |
| "learning_rate": 5.381993965990126e-05, | |
| "loss": 0.3262384033203125, | |
| "step": 67850 | |
| }, | |
| { | |
| "epoch": 1.3919639196391964, | |
| "grad_norm": 0.9086599946022034, | |
| "learning_rate": 5.378565551289084e-05, | |
| "loss": 0.32342315673828126, | |
| "step": 67900 | |
| }, | |
| { | |
| "epoch": 1.3929889298892988, | |
| "grad_norm": 0.8279297947883606, | |
| "learning_rate": 5.3751371365880424e-05, | |
| "loss": 0.32221973419189454, | |
| "step": 67950 | |
| }, | |
| { | |
| "epoch": 1.3940139401394014, | |
| "grad_norm": 0.85975581407547, | |
| "learning_rate": 5.371708721887e-05, | |
| "loss": 0.32570705413818357, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 1.395038950389504, | |
| "grad_norm": 0.8757076263427734, | |
| "learning_rate": 5.368280307185958e-05, | |
| "loss": 0.32869308471679687, | |
| "step": 68050 | |
| }, | |
| { | |
| "epoch": 1.3960639606396064, | |
| "grad_norm": 0.8919777274131775, | |
| "learning_rate": 5.3648518924849155e-05, | |
| "loss": 0.3239681625366211, | |
| "step": 68100 | |
| }, | |
| { | |
| "epoch": 1.397088970889709, | |
| "grad_norm": 0.8017241358757019, | |
| "learning_rate": 5.3614234777838736e-05, | |
| "loss": 0.32130081176757813, | |
| "step": 68150 | |
| }, | |
| { | |
| "epoch": 1.3981139811398113, | |
| "grad_norm": 0.8513978719711304, | |
| "learning_rate": 5.357995063082831e-05, | |
| "loss": 0.32170398712158205, | |
| "step": 68200 | |
| }, | |
| { | |
| "epoch": 1.399138991389914, | |
| "grad_norm": 0.7696248888969421, | |
| "learning_rate": 5.354566648381788e-05, | |
| "loss": 0.3252886962890625, | |
| "step": 68250 | |
| }, | |
| { | |
| "epoch": 1.4001640016400163, | |
| "grad_norm": 0.84288489818573, | |
| "learning_rate": 5.351138233680746e-05, | |
| "loss": 0.3276533126831055, | |
| "step": 68300 | |
| }, | |
| { | |
| "epoch": 1.401189011890119, | |
| "grad_norm": 0.8318451046943665, | |
| "learning_rate": 5.3477098189797035e-05, | |
| "loss": 0.32150787353515625, | |
| "step": 68350 | |
| }, | |
| { | |
| "epoch": 1.4022140221402215, | |
| "grad_norm": 0.8222660422325134, | |
| "learning_rate": 5.344281404278662e-05, | |
| "loss": 0.3197616386413574, | |
| "step": 68400 | |
| }, | |
| { | |
| "epoch": 1.4032390323903239, | |
| "grad_norm": 0.829852283000946, | |
| "learning_rate": 5.340852989577619e-05, | |
| "loss": 0.3262361526489258, | |
| "step": 68450 | |
| }, | |
| { | |
| "epoch": 1.4042640426404265, | |
| "grad_norm": 0.9323288798332214, | |
| "learning_rate": 5.337424574876577e-05, | |
| "loss": 0.3259796905517578, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 1.4052890528905289, | |
| "grad_norm": 0.786021888256073, | |
| "learning_rate": 5.333996160175535e-05, | |
| "loss": 0.32123626708984376, | |
| "step": 68550 | |
| }, | |
| { | |
| "epoch": 1.4063140631406315, | |
| "grad_norm": 0.7671592831611633, | |
| "learning_rate": 5.330567745474493e-05, | |
| "loss": 0.32137245178222656, | |
| "step": 68600 | |
| }, | |
| { | |
| "epoch": 1.4073390733907338, | |
| "grad_norm": 0.8092140555381775, | |
| "learning_rate": 5.327139330773451e-05, | |
| "loss": 0.32300045013427736, | |
| "step": 68650 | |
| }, | |
| { | |
| "epoch": 1.4083640836408364, | |
| "grad_norm": 0.7742304801940918, | |
| "learning_rate": 5.3237109160724085e-05, | |
| "loss": 0.3263935852050781, | |
| "step": 68700 | |
| }, | |
| { | |
| "epoch": 1.409389093890939, | |
| "grad_norm": 0.7967414259910583, | |
| "learning_rate": 5.320282501371367e-05, | |
| "loss": 0.32161540985107423, | |
| "step": 68750 | |
| }, | |
| { | |
| "epoch": 1.4104141041410414, | |
| "grad_norm": 0.7483751773834229, | |
| "learning_rate": 5.316854086670324e-05, | |
| "loss": 0.3208003997802734, | |
| "step": 68800 | |
| }, | |
| { | |
| "epoch": 1.4114391143911438, | |
| "grad_norm": 0.8223891854286194, | |
| "learning_rate": 5.313425671969282e-05, | |
| "loss": 0.32317386627197264, | |
| "step": 68850 | |
| }, | |
| { | |
| "epoch": 1.4124641246412464, | |
| "grad_norm": 0.8667693734169006, | |
| "learning_rate": 5.30999725726824e-05, | |
| "loss": 0.32322269439697265, | |
| "step": 68900 | |
| }, | |
| { | |
| "epoch": 1.413489134891349, | |
| "grad_norm": 0.9177883267402649, | |
| "learning_rate": 5.3065688425671965e-05, | |
| "loss": 0.32764148712158203, | |
| "step": 68950 | |
| }, | |
| { | |
| "epoch": 1.4145141451414514, | |
| "grad_norm": 0.7942080497741699, | |
| "learning_rate": 5.303140427866155e-05, | |
| "loss": 0.32547779083251954, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 1.415539155391554, | |
| "grad_norm": 0.8054535388946533, | |
| "learning_rate": 5.299712013165112e-05, | |
| "loss": 0.3248989868164063, | |
| "step": 69050 | |
| }, | |
| { | |
| "epoch": 1.4165641656416565, | |
| "grad_norm": 0.8348788619041443, | |
| "learning_rate": 5.29628359846407e-05, | |
| "loss": 0.3258544921875, | |
| "step": 69100 | |
| }, | |
| { | |
| "epoch": 1.417589175891759, | |
| "grad_norm": 0.8743811249732971, | |
| "learning_rate": 5.292855183763028e-05, | |
| "loss": 0.3214908218383789, | |
| "step": 69150 | |
| }, | |
| { | |
| "epoch": 1.4186141861418613, | |
| "grad_norm": 0.7922078967094421, | |
| "learning_rate": 5.289426769061986e-05, | |
| "loss": 0.3264510345458984, | |
| "step": 69200 | |
| }, | |
| { | |
| "epoch": 1.419639196391964, | |
| "grad_norm": 0.8300161957740784, | |
| "learning_rate": 5.2859983543609434e-05, | |
| "loss": 0.32866275787353516, | |
| "step": 69250 | |
| }, | |
| { | |
| "epoch": 1.4206642066420665, | |
| "grad_norm": 0.8617637157440186, | |
| "learning_rate": 5.2825699396599015e-05, | |
| "loss": 0.3229903030395508, | |
| "step": 69300 | |
| }, | |
| { | |
| "epoch": 1.4216892168921689, | |
| "grad_norm": 0.8252127170562744, | |
| "learning_rate": 5.27914152495886e-05, | |
| "loss": 0.3215915679931641, | |
| "step": 69350 | |
| }, | |
| { | |
| "epoch": 1.4227142271422715, | |
| "grad_norm": 0.8279685974121094, | |
| "learning_rate": 5.275713110257817e-05, | |
| "loss": 0.33011085510253907, | |
| "step": 69400 | |
| }, | |
| { | |
| "epoch": 1.4237392373923738, | |
| "grad_norm": 0.8190934658050537, | |
| "learning_rate": 5.272284695556775e-05, | |
| "loss": 0.3247592163085937, | |
| "step": 69450 | |
| }, | |
| { | |
| "epoch": 1.4247642476424764, | |
| "grad_norm": 0.8781667351722717, | |
| "learning_rate": 5.268856280855733e-05, | |
| "loss": 0.3197508430480957, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 1.4257892578925788, | |
| "grad_norm": 0.829992949962616, | |
| "learning_rate": 5.265427866154691e-05, | |
| "loss": 0.327332763671875, | |
| "step": 69550 | |
| }, | |
| { | |
| "epoch": 1.4268142681426814, | |
| "grad_norm": 0.8778061270713806, | |
| "learning_rate": 5.2619994514536484e-05, | |
| "loss": 0.323226203918457, | |
| "step": 69600 | |
| }, | |
| { | |
| "epoch": 1.427839278392784, | |
| "grad_norm": 0.8193259239196777, | |
| "learning_rate": 5.258571036752605e-05, | |
| "loss": 0.32503116607666016, | |
| "step": 69650 | |
| }, | |
| { | |
| "epoch": 1.4288642886428864, | |
| "grad_norm": 0.808738648891449, | |
| "learning_rate": 5.255142622051563e-05, | |
| "loss": 0.32555351257324217, | |
| "step": 69700 | |
| }, | |
| { | |
| "epoch": 1.429889298892989, | |
| "grad_norm": 0.8136274218559265, | |
| "learning_rate": 5.251714207350521e-05, | |
| "loss": 0.3252926254272461, | |
| "step": 69750 | |
| }, | |
| { | |
| "epoch": 1.4309143091430914, | |
| "grad_norm": 0.9110633134841919, | |
| "learning_rate": 5.248285792649479e-05, | |
| "loss": 0.3224606704711914, | |
| "step": 69800 | |
| }, | |
| { | |
| "epoch": 1.431939319393194, | |
| "grad_norm": 0.737113893032074, | |
| "learning_rate": 5.2448573779484364e-05, | |
| "loss": 0.3237055587768555, | |
| "step": 69850 | |
| }, | |
| { | |
| "epoch": 1.4329643296432963, | |
| "grad_norm": 0.8181567192077637, | |
| "learning_rate": 5.2414289632473946e-05, | |
| "loss": 0.3288494873046875, | |
| "step": 69900 | |
| }, | |
| { | |
| "epoch": 1.433989339893399, | |
| "grad_norm": 0.7982308864593506, | |
| "learning_rate": 5.238000548546352e-05, | |
| "loss": 0.32074440002441407, | |
| "step": 69950 | |
| }, | |
| { | |
| "epoch": 1.4350143501435015, | |
| "grad_norm": 0.8481606841087341, | |
| "learning_rate": 5.23457213384531e-05, | |
| "loss": 0.32541091918945314, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 1.4350143501435015, | |
| "eval_loss": 0.39375633001327515, | |
| "eval_runtime": 4.5129, | |
| "eval_samples_per_second": 5539.616, | |
| "eval_steps_per_second": 5.54, | |
| "eval_teacher_cosine_loss": 0.24818599224090576, | |
| "eval_teacher_cosine_mean": 0.7518140077590942, | |
| "eval_teacher_cosine_p10": 0.6747055053710938, | |
| "eval_teacher_cosine_p50": 0.7617442607879639, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 1.436039360393604, | |
| "grad_norm": 0.9407992362976074, | |
| "learning_rate": 5.231143719144268e-05, | |
| "loss": 0.32606609344482423, | |
| "step": 70050 | |
| }, | |
| { | |
| "epoch": 1.4370643706437065, | |
| "grad_norm": 0.8172420859336853, | |
| "learning_rate": 5.227715304443226e-05, | |
| "loss": 0.32359210968017577, | |
| "step": 70100 | |
| }, | |
| { | |
| "epoch": 1.4380893808938089, | |
| "grad_norm": 0.843425989151001, | |
| "learning_rate": 5.224286889742184e-05, | |
| "loss": 0.32859848022460936, | |
| "step": 70150 | |
| }, | |
| { | |
| "epoch": 1.4391143911439115, | |
| "grad_norm": 0.8608160018920898, | |
| "learning_rate": 5.2208584750411414e-05, | |
| "loss": 0.32490303039550783, | |
| "step": 70200 | |
| }, | |
| { | |
| "epoch": 1.4401394013940139, | |
| "grad_norm": 0.8055329322814941, | |
| "learning_rate": 5.2174300603400996e-05, | |
| "loss": 0.3219798278808594, | |
| "step": 70250 | |
| }, | |
| { | |
| "epoch": 1.4411644116441165, | |
| "grad_norm": 0.840088963508606, | |
| "learning_rate": 5.214001645639057e-05, | |
| "loss": 0.32554107666015625, | |
| "step": 70300 | |
| }, | |
| { | |
| "epoch": 1.442189421894219, | |
| "grad_norm": 0.7542402744293213, | |
| "learning_rate": 5.210573230938014e-05, | |
| "loss": 0.3228450393676758, | |
| "step": 70350 | |
| }, | |
| { | |
| "epoch": 1.4432144321443214, | |
| "grad_norm": 0.7770607471466064, | |
| "learning_rate": 5.207144816236972e-05, | |
| "loss": 0.3206169128417969, | |
| "step": 70400 | |
| }, | |
| { | |
| "epoch": 1.4442394423944238, | |
| "grad_norm": 0.8323031663894653, | |
| "learning_rate": 5.2037164015359294e-05, | |
| "loss": 0.3175970458984375, | |
| "step": 70450 | |
| }, | |
| { | |
| "epoch": 1.4452644526445264, | |
| "grad_norm": 0.8296452164649963, | |
| "learning_rate": 5.2002879868348876e-05, | |
| "loss": 0.3247274017333984, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 1.446289462894629, | |
| "grad_norm": 0.7410755157470703, | |
| "learning_rate": 5.196859572133845e-05, | |
| "loss": 0.3221352767944336, | |
| "step": 70550 | |
| }, | |
| { | |
| "epoch": 1.4473144731447314, | |
| "grad_norm": 0.7597193717956543, | |
| "learning_rate": 5.193431157432803e-05, | |
| "loss": 0.32127349853515624, | |
| "step": 70600 | |
| }, | |
| { | |
| "epoch": 1.448339483394834, | |
| "grad_norm": 0.7934571504592896, | |
| "learning_rate": 5.190002742731761e-05, | |
| "loss": 0.31878501892089844, | |
| "step": 70650 | |
| }, | |
| { | |
| "epoch": 1.4493644936449366, | |
| "grad_norm": 0.8169525265693665, | |
| "learning_rate": 5.186574328030719e-05, | |
| "loss": 0.3176326942443848, | |
| "step": 70700 | |
| }, | |
| { | |
| "epoch": 1.450389503895039, | |
| "grad_norm": 0.8655585050582886, | |
| "learning_rate": 5.183145913329677e-05, | |
| "loss": 0.3293414306640625, | |
| "step": 70750 | |
| }, | |
| { | |
| "epoch": 1.4514145141451413, | |
| "grad_norm": 0.9415244460105896, | |
| "learning_rate": 5.1797174986286344e-05, | |
| "loss": 0.329102897644043, | |
| "step": 70800 | |
| }, | |
| { | |
| "epoch": 1.452439524395244, | |
| "grad_norm": 0.8756157755851746, | |
| "learning_rate": 5.1762890839275926e-05, | |
| "loss": 0.32319149017333987, | |
| "step": 70850 | |
| }, | |
| { | |
| "epoch": 1.4534645346453465, | |
| "grad_norm": 0.8652890920639038, | |
| "learning_rate": 5.17286066922655e-05, | |
| "loss": 0.32893829345703124, | |
| "step": 70900 | |
| }, | |
| { | |
| "epoch": 1.454489544895449, | |
| "grad_norm": 0.8373616933822632, | |
| "learning_rate": 5.169432254525508e-05, | |
| "loss": 0.33355567932128904, | |
| "step": 70950 | |
| }, | |
| { | |
| "epoch": 1.4555145551455515, | |
| "grad_norm": 0.7425225377082825, | |
| "learning_rate": 5.166003839824466e-05, | |
| "loss": 0.3148170471191406, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 1.4565395653956539, | |
| "grad_norm": 0.8008820414543152, | |
| "learning_rate": 5.1625754251234225e-05, | |
| "loss": 0.3212063217163086, | |
| "step": 71050 | |
| }, | |
| { | |
| "epoch": 1.4575645756457565, | |
| "grad_norm": 0.8627517819404602, | |
| "learning_rate": 5.1591470104223806e-05, | |
| "loss": 0.3259848403930664, | |
| "step": 71100 | |
| }, | |
| { | |
| "epoch": 1.4585895858958589, | |
| "grad_norm": 0.7379549741744995, | |
| "learning_rate": 5.155718595721338e-05, | |
| "loss": 0.3224420928955078, | |
| "step": 71150 | |
| }, | |
| { | |
| "epoch": 1.4596145961459615, | |
| "grad_norm": 0.838890552520752, | |
| "learning_rate": 5.152290181020296e-05, | |
| "loss": 0.3219120788574219, | |
| "step": 71200 | |
| }, | |
| { | |
| "epoch": 1.460639606396064, | |
| "grad_norm": 0.8152722120285034, | |
| "learning_rate": 5.148861766319254e-05, | |
| "loss": 0.31721355438232424, | |
| "step": 71250 | |
| }, | |
| { | |
| "epoch": 1.4616646166461664, | |
| "grad_norm": 0.8347107172012329, | |
| "learning_rate": 5.145433351618212e-05, | |
| "loss": 0.3306005859375, | |
| "step": 71300 | |
| }, | |
| { | |
| "epoch": 1.462689626896269, | |
| "grad_norm": 0.8349406719207764, | |
| "learning_rate": 5.14200493691717e-05, | |
| "loss": 0.32376636505126954, | |
| "step": 71350 | |
| }, | |
| { | |
| "epoch": 1.4637146371463714, | |
| "grad_norm": 0.8464040756225586, | |
| "learning_rate": 5.1385765222161275e-05, | |
| "loss": 0.3224752426147461, | |
| "step": 71400 | |
| }, | |
| { | |
| "epoch": 1.464739647396474, | |
| "grad_norm": 0.811945915222168, | |
| "learning_rate": 5.1351481075150856e-05, | |
| "loss": 0.31966629028320315, | |
| "step": 71450 | |
| }, | |
| { | |
| "epoch": 1.4657646576465764, | |
| "grad_norm": 0.9541268348693848, | |
| "learning_rate": 5.131719692814043e-05, | |
| "loss": 0.3267343521118164, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 1.466789667896679, | |
| "grad_norm": 0.8406468033790588, | |
| "learning_rate": 5.128291278113001e-05, | |
| "loss": 0.32012245178222654, | |
| "step": 71550 | |
| }, | |
| { | |
| "epoch": 1.4678146781467816, | |
| "grad_norm": 0.7580136060714722, | |
| "learning_rate": 5.124862863411959e-05, | |
| "loss": 0.3237984085083008, | |
| "step": 71600 | |
| }, | |
| { | |
| "epoch": 1.468839688396884, | |
| "grad_norm": 0.8599867820739746, | |
| "learning_rate": 5.121434448710917e-05, | |
| "loss": 0.3265164566040039, | |
| "step": 71650 | |
| }, | |
| { | |
| "epoch": 1.4698646986469865, | |
| "grad_norm": 0.811047375202179, | |
| "learning_rate": 5.118006034009874e-05, | |
| "loss": 0.3289006805419922, | |
| "step": 71700 | |
| }, | |
| { | |
| "epoch": 1.470889708897089, | |
| "grad_norm": 0.8274995684623718, | |
| "learning_rate": 5.1145776193088325e-05, | |
| "loss": 0.32238533020019533, | |
| "step": 71750 | |
| }, | |
| { | |
| "epoch": 1.4719147191471915, | |
| "grad_norm": 0.863638699054718, | |
| "learning_rate": 5.111149204607789e-05, | |
| "loss": 0.32924819946289063, | |
| "step": 71800 | |
| }, | |
| { | |
| "epoch": 1.472939729397294, | |
| "grad_norm": 0.9082023501396179, | |
| "learning_rate": 5.107720789906747e-05, | |
| "loss": 0.325850715637207, | |
| "step": 71850 | |
| }, | |
| { | |
| "epoch": 1.4739647396473965, | |
| "grad_norm": 0.798706591129303, | |
| "learning_rate": 5.104292375205705e-05, | |
| "loss": 0.3200459671020508, | |
| "step": 71900 | |
| }, | |
| { | |
| "epoch": 1.474989749897499, | |
| "grad_norm": 0.7852233052253723, | |
| "learning_rate": 5.1008639605046623e-05, | |
| "loss": 0.3252104187011719, | |
| "step": 71950 | |
| }, | |
| { | |
| "epoch": 1.4760147601476015, | |
| "grad_norm": 0.8240165710449219, | |
| "learning_rate": 5.0974355458036205e-05, | |
| "loss": 0.31657413482666014, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 1.4770397703977038, | |
| "grad_norm": 0.8072550892829895, | |
| "learning_rate": 5.0940071311025786e-05, | |
| "loss": 0.3240640640258789, | |
| "step": 72050 | |
| }, | |
| { | |
| "epoch": 1.4780647806478064, | |
| "grad_norm": 0.8864871263504028, | |
| "learning_rate": 5.090578716401536e-05, | |
| "loss": 0.3234889602661133, | |
| "step": 72100 | |
| }, | |
| { | |
| "epoch": 1.479089790897909, | |
| "grad_norm": 0.9024125933647156, | |
| "learning_rate": 5.087150301700494e-05, | |
| "loss": 0.324779167175293, | |
| "step": 72150 | |
| }, | |
| { | |
| "epoch": 1.4801148011480114, | |
| "grad_norm": 0.826238751411438, | |
| "learning_rate": 5.083721886999452e-05, | |
| "loss": 0.31973369598388673, | |
| "step": 72200 | |
| }, | |
| { | |
| "epoch": 1.481139811398114, | |
| "grad_norm": 0.8518129587173462, | |
| "learning_rate": 5.08029347229841e-05, | |
| "loss": 0.32566558837890625, | |
| "step": 72250 | |
| }, | |
| { | |
| "epoch": 1.4821648216482166, | |
| "grad_norm": 0.8558894991874695, | |
| "learning_rate": 5.0768650575973673e-05, | |
| "loss": 0.3187210845947266, | |
| "step": 72300 | |
| }, | |
| { | |
| "epoch": 1.483189831898319, | |
| "grad_norm": 0.7581382989883423, | |
| "learning_rate": 5.0734366428963255e-05, | |
| "loss": 0.31517763137817384, | |
| "step": 72350 | |
| }, | |
| { | |
| "epoch": 1.4842148421484214, | |
| "grad_norm": 0.9150123000144958, | |
| "learning_rate": 5.070008228195283e-05, | |
| "loss": 0.32226043701171875, | |
| "step": 72400 | |
| }, | |
| { | |
| "epoch": 1.485239852398524, | |
| "grad_norm": 0.8550363183021545, | |
| "learning_rate": 5.066579813494241e-05, | |
| "loss": 0.3218573760986328, | |
| "step": 72450 | |
| }, | |
| { | |
| "epoch": 1.4862648626486266, | |
| "grad_norm": 0.8347415328025818, | |
| "learning_rate": 5.063151398793198e-05, | |
| "loss": 0.32767398834228517, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 1.487289872898729, | |
| "grad_norm": 0.804771900177002, | |
| "learning_rate": 5.0597229840921554e-05, | |
| "loss": 0.32101150512695314, | |
| "step": 72550 | |
| }, | |
| { | |
| "epoch": 1.4883148831488315, | |
| "grad_norm": 0.7870937585830688, | |
| "learning_rate": 5.0562945693911135e-05, | |
| "loss": 0.3219620132446289, | |
| "step": 72600 | |
| }, | |
| { | |
| "epoch": 1.489339893398934, | |
| "grad_norm": 0.8917014598846436, | |
| "learning_rate": 5.052866154690071e-05, | |
| "loss": 0.3207341384887695, | |
| "step": 72650 | |
| }, | |
| { | |
| "epoch": 1.4903649036490365, | |
| "grad_norm": 0.7857219576835632, | |
| "learning_rate": 5.049437739989029e-05, | |
| "loss": 0.320240364074707, | |
| "step": 72700 | |
| }, | |
| { | |
| "epoch": 1.4913899138991389, | |
| "grad_norm": 0.8973723649978638, | |
| "learning_rate": 5.046009325287987e-05, | |
| "loss": 0.32663536071777344, | |
| "step": 72750 | |
| }, | |
| { | |
| "epoch": 1.4924149241492415, | |
| "grad_norm": 0.7418645620346069, | |
| "learning_rate": 5.042580910586945e-05, | |
| "loss": 0.3241929626464844, | |
| "step": 72800 | |
| }, | |
| { | |
| "epoch": 1.493439934399344, | |
| "grad_norm": 0.8751468062400818, | |
| "learning_rate": 5.039152495885903e-05, | |
| "loss": 0.3228106307983398, | |
| "step": 72850 | |
| }, | |
| { | |
| "epoch": 1.4944649446494465, | |
| "grad_norm": 0.8268604874610901, | |
| "learning_rate": 5.0357240811848604e-05, | |
| "loss": 0.3227836990356445, | |
| "step": 72900 | |
| }, | |
| { | |
| "epoch": 1.495489954899549, | |
| "grad_norm": 0.767557680606842, | |
| "learning_rate": 5.0322956664838185e-05, | |
| "loss": 0.3211869430541992, | |
| "step": 72950 | |
| }, | |
| { | |
| "epoch": 1.4965149651496514, | |
| "grad_norm": 0.9542518854141235, | |
| "learning_rate": 5.028867251782776e-05, | |
| "loss": 0.32011390686035157, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 1.497539975399754, | |
| "grad_norm": 0.8736904263496399, | |
| "learning_rate": 5.025438837081734e-05, | |
| "loss": 0.32777896881103513, | |
| "step": 73050 | |
| }, | |
| { | |
| "epoch": 1.4985649856498564, | |
| "grad_norm": 0.8682650923728943, | |
| "learning_rate": 5.0220104223806916e-05, | |
| "loss": 0.3216840744018555, | |
| "step": 73100 | |
| }, | |
| { | |
| "epoch": 1.499589995899959, | |
| "grad_norm": 0.8211861252784729, | |
| "learning_rate": 5.01858200767965e-05, | |
| "loss": 0.3232941436767578, | |
| "step": 73150 | |
| }, | |
| { | |
| "epoch": 1.5006150061500616, | |
| "grad_norm": 0.772685706615448, | |
| "learning_rate": 5.0151535929786065e-05, | |
| "loss": 0.3221567916870117, | |
| "step": 73200 | |
| }, | |
| { | |
| "epoch": 1.501640016400164, | |
| "grad_norm": 0.8587966561317444, | |
| "learning_rate": 5.011725178277564e-05, | |
| "loss": 0.3243228530883789, | |
| "step": 73250 | |
| }, | |
| { | |
| "epoch": 1.5026650266502664, | |
| "grad_norm": 0.81226646900177, | |
| "learning_rate": 5.008296763576522e-05, | |
| "loss": 0.3203650665283203, | |
| "step": 73300 | |
| }, | |
| { | |
| "epoch": 1.503690036900369, | |
| "grad_norm": 0.8040902614593506, | |
| "learning_rate": 5.0048683488754796e-05, | |
| "loss": 0.31609243392944336, | |
| "step": 73350 | |
| }, | |
| { | |
| "epoch": 1.5047150471504716, | |
| "grad_norm": 0.7672544717788696, | |
| "learning_rate": 5.001439934174438e-05, | |
| "loss": 0.32147529602050784, | |
| "step": 73400 | |
| }, | |
| { | |
| "epoch": 1.505740057400574, | |
| "grad_norm": 0.9493066668510437, | |
| "learning_rate": 4.998011519473396e-05, | |
| "loss": 0.3256067657470703, | |
| "step": 73450 | |
| }, | |
| { | |
| "epoch": 1.5067650676506765, | |
| "grad_norm": 0.8621537685394287, | |
| "learning_rate": 4.9945831047723534e-05, | |
| "loss": 0.3190264129638672, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 1.5077900779007791, | |
| "grad_norm": 0.8428702354431152, | |
| "learning_rate": 4.9911546900713115e-05, | |
| "loss": 0.3234341049194336, | |
| "step": 73550 | |
| }, | |
| { | |
| "epoch": 1.5088150881508815, | |
| "grad_norm": 0.8564041256904602, | |
| "learning_rate": 4.987726275370269e-05, | |
| "loss": 0.3214791488647461, | |
| "step": 73600 | |
| }, | |
| { | |
| "epoch": 1.5098400984009839, | |
| "grad_norm": 0.9037203788757324, | |
| "learning_rate": 4.984297860669227e-05, | |
| "loss": 0.3186996650695801, | |
| "step": 73650 | |
| }, | |
| { | |
| "epoch": 1.5108651086510865, | |
| "grad_norm": 0.8297356963157654, | |
| "learning_rate": 4.980869445968184e-05, | |
| "loss": 0.32877620697021487, | |
| "step": 73700 | |
| }, | |
| { | |
| "epoch": 1.511890118901189, | |
| "grad_norm": 0.8550005555152893, | |
| "learning_rate": 4.977441031267142e-05, | |
| "loss": 0.32242507934570314, | |
| "step": 73750 | |
| }, | |
| { | |
| "epoch": 1.5129151291512914, | |
| "grad_norm": 0.7329958081245422, | |
| "learning_rate": 4.9740126165661e-05, | |
| "loss": 0.3191924285888672, | |
| "step": 73800 | |
| }, | |
| { | |
| "epoch": 1.513940139401394, | |
| "grad_norm": 0.8045824766159058, | |
| "learning_rate": 4.970584201865058e-05, | |
| "loss": 0.32192474365234375, | |
| "step": 73850 | |
| }, | |
| { | |
| "epoch": 1.5149651496514966, | |
| "grad_norm": 0.8658392429351807, | |
| "learning_rate": 4.967155787164016e-05, | |
| "loss": 0.31969024658203127, | |
| "step": 73900 | |
| }, | |
| { | |
| "epoch": 1.515990159901599, | |
| "grad_norm": 0.8708179593086243, | |
| "learning_rate": 4.963727372462973e-05, | |
| "loss": 0.32359527587890624, | |
| "step": 73950 | |
| }, | |
| { | |
| "epoch": 1.5170151701517014, | |
| "grad_norm": 0.8628482818603516, | |
| "learning_rate": 4.9602989577619315e-05, | |
| "loss": 0.3163245391845703, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 1.518040180401804, | |
| "grad_norm": 0.8732318878173828, | |
| "learning_rate": 4.956870543060888e-05, | |
| "loss": 0.32051082611083986, | |
| "step": 74050 | |
| }, | |
| { | |
| "epoch": 1.5190651906519066, | |
| "grad_norm": 0.8582825660705566, | |
| "learning_rate": 4.9534421283598464e-05, | |
| "loss": 0.3230252456665039, | |
| "step": 74100 | |
| }, | |
| { | |
| "epoch": 1.520090200902009, | |
| "grad_norm": 0.8784299492835999, | |
| "learning_rate": 4.9500137136588046e-05, | |
| "loss": 0.3149135971069336, | |
| "step": 74150 | |
| }, | |
| { | |
| "epoch": 1.5211152111521116, | |
| "grad_norm": 0.7432060241699219, | |
| "learning_rate": 4.946585298957762e-05, | |
| "loss": 0.3244295501708984, | |
| "step": 74200 | |
| }, | |
| { | |
| "epoch": 1.5221402214022142, | |
| "grad_norm": 0.8517181873321533, | |
| "learning_rate": 4.94315688425672e-05, | |
| "loss": 0.32057872772216794, | |
| "step": 74250 | |
| }, | |
| { | |
| "epoch": 1.5231652316523165, | |
| "grad_norm": 0.799624502658844, | |
| "learning_rate": 4.9397284695556777e-05, | |
| "loss": 0.31837175369262694, | |
| "step": 74300 | |
| }, | |
| { | |
| "epoch": 1.524190241902419, | |
| "grad_norm": 0.840546727180481, | |
| "learning_rate": 4.936300054854636e-05, | |
| "loss": 0.3227960205078125, | |
| "step": 74350 | |
| }, | |
| { | |
| "epoch": 1.5252152521525215, | |
| "grad_norm": 0.7657656073570251, | |
| "learning_rate": 4.932871640153593e-05, | |
| "loss": 0.32157192230224607, | |
| "step": 74400 | |
| }, | |
| { | |
| "epoch": 1.5262402624026241, | |
| "grad_norm": 0.7886798977851868, | |
| "learning_rate": 4.929443225452551e-05, | |
| "loss": 0.31677234649658204, | |
| "step": 74450 | |
| }, | |
| { | |
| "epoch": 1.5272652726527265, | |
| "grad_norm": 0.8501538634300232, | |
| "learning_rate": 4.926014810751509e-05, | |
| "loss": 0.3218958282470703, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 1.5282902829028289, | |
| "grad_norm": 0.9173591732978821, | |
| "learning_rate": 4.9225863960504664e-05, | |
| "loss": 0.3136588668823242, | |
| "step": 74550 | |
| }, | |
| { | |
| "epoch": 1.5293152931529317, | |
| "grad_norm": 0.8487765192985535, | |
| "learning_rate": 4.9191579813494245e-05, | |
| "loss": 0.3261243438720703, | |
| "step": 74600 | |
| }, | |
| { | |
| "epoch": 1.530340303403034, | |
| "grad_norm": 0.8181816935539246, | |
| "learning_rate": 4.915729566648382e-05, | |
| "loss": 0.31758182525634765, | |
| "step": 74650 | |
| }, | |
| { | |
| "epoch": 1.5313653136531364, | |
| "grad_norm": 0.899936318397522, | |
| "learning_rate": 4.91230115194734e-05, | |
| "loss": 0.320896110534668, | |
| "step": 74700 | |
| }, | |
| { | |
| "epoch": 1.532390323903239, | |
| "grad_norm": 0.8115948438644409, | |
| "learning_rate": 4.9088727372462976e-05, | |
| "loss": 0.3224074935913086, | |
| "step": 74750 | |
| }, | |
| { | |
| "epoch": 1.5334153341533416, | |
| "grad_norm": 0.8222758769989014, | |
| "learning_rate": 4.905444322545255e-05, | |
| "loss": 0.31692960739135745, | |
| "step": 74800 | |
| }, | |
| { | |
| "epoch": 1.534440344403444, | |
| "grad_norm": 0.860182523727417, | |
| "learning_rate": 4.902015907844213e-05, | |
| "loss": 0.3194332504272461, | |
| "step": 74850 | |
| }, | |
| { | |
| "epoch": 1.5354653546535464, | |
| "grad_norm": 0.7902491688728333, | |
| "learning_rate": 4.898587493143171e-05, | |
| "loss": 0.31940689086914065, | |
| "step": 74900 | |
| }, | |
| { | |
| "epoch": 1.536490364903649, | |
| "grad_norm": 0.7965166568756104, | |
| "learning_rate": 4.895159078442129e-05, | |
| "loss": 0.3183931922912598, | |
| "step": 74950 | |
| }, | |
| { | |
| "epoch": 1.5375153751537516, | |
| "grad_norm": 0.8543437123298645, | |
| "learning_rate": 4.891730663741086e-05, | |
| "loss": 0.3151415824890137, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 1.5375153751537516, | |
| "eval_loss": 0.38501670956611633, | |
| "eval_runtime": 4.4949, | |
| "eval_samples_per_second": 5561.888, | |
| "eval_steps_per_second": 5.562, | |
| "eval_teacher_cosine_loss": 0.24623775482177734, | |
| "eval_teacher_cosine_mean": 0.7537622451782227, | |
| "eval_teacher_cosine_p10": 0.6764745116233826, | |
| "eval_teacher_cosine_p50": 0.7636899948120117, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 1.538540385403854, | |
| "grad_norm": 0.780753493309021, | |
| "learning_rate": 4.8883022490400444e-05, | |
| "loss": 0.319671688079834, | |
| "step": 75050 | |
| }, | |
| { | |
| "epoch": 1.5395653956539566, | |
| "grad_norm": 0.8383253216743469, | |
| "learning_rate": 4.884873834339002e-05, | |
| "loss": 0.3204848861694336, | |
| "step": 75100 | |
| }, | |
| { | |
| "epoch": 1.5405904059040592, | |
| "grad_norm": 0.8321828246116638, | |
| "learning_rate": 4.8814454196379594e-05, | |
| "loss": 0.31752824783325195, | |
| "step": 75150 | |
| }, | |
| { | |
| "epoch": 1.5416154161541615, | |
| "grad_norm": 0.808409571647644, | |
| "learning_rate": 4.8780170049369175e-05, | |
| "loss": 0.3209188842773438, | |
| "step": 75200 | |
| }, | |
| { | |
| "epoch": 1.542640426404264, | |
| "grad_norm": 0.9598098397254944, | |
| "learning_rate": 4.874588590235875e-05, | |
| "loss": 0.31988815307617186, | |
| "step": 75250 | |
| }, | |
| { | |
| "epoch": 1.5436654366543665, | |
| "grad_norm": 0.8654833436012268, | |
| "learning_rate": 4.871160175534833e-05, | |
| "loss": 0.31215457916259765, | |
| "step": 75300 | |
| }, | |
| { | |
| "epoch": 1.544690446904469, | |
| "grad_norm": 0.8033096194267273, | |
| "learning_rate": 4.8677317608337906e-05, | |
| "loss": 0.31854236602783204, | |
| "step": 75350 | |
| }, | |
| { | |
| "epoch": 1.5457154571545715, | |
| "grad_norm": 0.7928189039230347, | |
| "learning_rate": 4.864303346132749e-05, | |
| "loss": 0.3226872253417969, | |
| "step": 75400 | |
| }, | |
| { | |
| "epoch": 1.546740467404674, | |
| "grad_norm": 0.8464558720588684, | |
| "learning_rate": 4.860874931431706e-05, | |
| "loss": 0.3185890197753906, | |
| "step": 75450 | |
| }, | |
| { | |
| "epoch": 1.5477654776547767, | |
| "grad_norm": 0.894307017326355, | |
| "learning_rate": 4.857446516730664e-05, | |
| "loss": 0.32289566040039064, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 1.548790487904879, | |
| "grad_norm": 0.8126587271690369, | |
| "learning_rate": 4.854018102029622e-05, | |
| "loss": 0.3175657653808594, | |
| "step": 75550 | |
| }, | |
| { | |
| "epoch": 1.5498154981549814, | |
| "grad_norm": 0.8079051375389099, | |
| "learning_rate": 4.850589687328579e-05, | |
| "loss": 0.3221469497680664, | |
| "step": 75600 | |
| }, | |
| { | |
| "epoch": 1.550840508405084, | |
| "grad_norm": 0.9454007744789124, | |
| "learning_rate": 4.8471612726275375e-05, | |
| "loss": 0.3207063674926758, | |
| "step": 75650 | |
| }, | |
| { | |
| "epoch": 1.5518655186551866, | |
| "grad_norm": 0.8760682940483093, | |
| "learning_rate": 4.843732857926495e-05, | |
| "loss": 0.31334426879882815, | |
| "step": 75700 | |
| }, | |
| { | |
| "epoch": 1.552890528905289, | |
| "grad_norm": 0.8810198307037354, | |
| "learning_rate": 4.840304443225453e-05, | |
| "loss": 0.32138355255126955, | |
| "step": 75750 | |
| }, | |
| { | |
| "epoch": 1.5539155391553916, | |
| "grad_norm": 0.8131755590438843, | |
| "learning_rate": 4.8368760285244106e-05, | |
| "loss": 0.32288604736328125, | |
| "step": 75800 | |
| }, | |
| { | |
| "epoch": 1.5549405494054942, | |
| "grad_norm": 0.8832893967628479, | |
| "learning_rate": 4.833447613823368e-05, | |
| "loss": 0.3257476425170898, | |
| "step": 75850 | |
| }, | |
| { | |
| "epoch": 1.5559655596555966, | |
| "grad_norm": 0.7937950491905212, | |
| "learning_rate": 4.830019199122326e-05, | |
| "loss": 0.31896984100341796, | |
| "step": 75900 | |
| }, | |
| { | |
| "epoch": 1.556990569905699, | |
| "grad_norm": 0.8258074522018433, | |
| "learning_rate": 4.8265907844212836e-05, | |
| "loss": 0.32163631439208984, | |
| "step": 75950 | |
| }, | |
| { | |
| "epoch": 1.5580155801558015, | |
| "grad_norm": 0.7663768529891968, | |
| "learning_rate": 4.823162369720242e-05, | |
| "loss": 0.32326377868652345, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 1.5590405904059041, | |
| "grad_norm": 0.7842888236045837, | |
| "learning_rate": 4.819733955019199e-05, | |
| "loss": 0.32083873748779296, | |
| "step": 76050 | |
| }, | |
| { | |
| "epoch": 1.5600656006560065, | |
| "grad_norm": 0.8103510141372681, | |
| "learning_rate": 4.8163055403181574e-05, | |
| "loss": 0.3219564819335938, | |
| "step": 76100 | |
| }, | |
| { | |
| "epoch": 1.561090610906109, | |
| "grad_norm": 0.8093062043190002, | |
| "learning_rate": 4.812877125617115e-05, | |
| "loss": 0.31376943588256834, | |
| "step": 76150 | |
| }, | |
| { | |
| "epoch": 1.5621156211562117, | |
| "grad_norm": 0.8659927248954773, | |
| "learning_rate": 4.8094487109160723e-05, | |
| "loss": 0.32738101959228516, | |
| "step": 76200 | |
| }, | |
| { | |
| "epoch": 1.563140631406314, | |
| "grad_norm": 0.885398805141449, | |
| "learning_rate": 4.8060202962150305e-05, | |
| "loss": 0.32239830017089843, | |
| "step": 76250 | |
| }, | |
| { | |
| "epoch": 1.5641656416564165, | |
| "grad_norm": 0.8417985439300537, | |
| "learning_rate": 4.802591881513988e-05, | |
| "loss": 0.3258563995361328, | |
| "step": 76300 | |
| }, | |
| { | |
| "epoch": 1.565190651906519, | |
| "grad_norm": 0.7951918244361877, | |
| "learning_rate": 4.799163466812946e-05, | |
| "loss": 0.3176189231872559, | |
| "step": 76350 | |
| }, | |
| { | |
| "epoch": 1.5662156621566217, | |
| "grad_norm": 0.8387738466262817, | |
| "learning_rate": 4.7957350521119036e-05, | |
| "loss": 0.3212513732910156, | |
| "step": 76400 | |
| }, | |
| { | |
| "epoch": 1.567240672406724, | |
| "grad_norm": 0.8465948104858398, | |
| "learning_rate": 4.792306637410862e-05, | |
| "loss": 0.3156487655639648, | |
| "step": 76450 | |
| }, | |
| { | |
| "epoch": 1.5682656826568264, | |
| "grad_norm": 0.7404701113700867, | |
| "learning_rate": 4.788878222709819e-05, | |
| "loss": 0.3192993927001953, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 1.569290692906929, | |
| "grad_norm": 0.8457310795783997, | |
| "learning_rate": 4.785449808008777e-05, | |
| "loss": 0.3215878677368164, | |
| "step": 76550 | |
| }, | |
| { | |
| "epoch": 1.5703157031570316, | |
| "grad_norm": 0.8577420711517334, | |
| "learning_rate": 4.782021393307735e-05, | |
| "loss": 0.31313199996948243, | |
| "step": 76600 | |
| }, | |
| { | |
| "epoch": 1.571340713407134, | |
| "grad_norm": 0.929417073726654, | |
| "learning_rate": 4.778592978606692e-05, | |
| "loss": 0.3176727867126465, | |
| "step": 76650 | |
| }, | |
| { | |
| "epoch": 1.5723657236572366, | |
| "grad_norm": 0.8129697442054749, | |
| "learning_rate": 4.7751645639056504e-05, | |
| "loss": 0.3164213752746582, | |
| "step": 76700 | |
| }, | |
| { | |
| "epoch": 1.5733907339073392, | |
| "grad_norm": 0.7792359590530396, | |
| "learning_rate": 4.771736149204608e-05, | |
| "loss": 0.3246908187866211, | |
| "step": 76750 | |
| }, | |
| { | |
| "epoch": 1.5744157441574416, | |
| "grad_norm": 0.7388657927513123, | |
| "learning_rate": 4.768307734503566e-05, | |
| "loss": 0.31682104110717774, | |
| "step": 76800 | |
| }, | |
| { | |
| "epoch": 1.575440754407544, | |
| "grad_norm": 0.9094868898391724, | |
| "learning_rate": 4.7648793198025235e-05, | |
| "loss": 0.32040748596191404, | |
| "step": 76850 | |
| }, | |
| { | |
| "epoch": 1.5764657646576465, | |
| "grad_norm": 0.764255166053772, | |
| "learning_rate": 4.761450905101481e-05, | |
| "loss": 0.3155165100097656, | |
| "step": 76900 | |
| }, | |
| { | |
| "epoch": 1.5774907749077491, | |
| "grad_norm": 0.8642795085906982, | |
| "learning_rate": 4.758022490400439e-05, | |
| "loss": 0.31511314392089845, | |
| "step": 76950 | |
| }, | |
| { | |
| "epoch": 1.5785157851578515, | |
| "grad_norm": 0.8035222291946411, | |
| "learning_rate": 4.7545940756993966e-05, | |
| "loss": 0.3142196083068848, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 1.5795407954079541, | |
| "grad_norm": 0.9460496306419373, | |
| "learning_rate": 4.751165660998355e-05, | |
| "loss": 0.3174629020690918, | |
| "step": 77050 | |
| }, | |
| { | |
| "epoch": 1.5805658056580567, | |
| "grad_norm": 0.9577794075012207, | |
| "learning_rate": 4.747737246297312e-05, | |
| "loss": 0.3174264907836914, | |
| "step": 77100 | |
| }, | |
| { | |
| "epoch": 1.581590815908159, | |
| "grad_norm": 0.9185048341751099, | |
| "learning_rate": 4.7443088315962704e-05, | |
| "loss": 0.31767179489135744, | |
| "step": 77150 | |
| }, | |
| { | |
| "epoch": 1.5826158261582615, | |
| "grad_norm": 0.7938192486763, | |
| "learning_rate": 4.740880416895228e-05, | |
| "loss": 0.31591192245483396, | |
| "step": 77200 | |
| }, | |
| { | |
| "epoch": 1.583640836408364, | |
| "grad_norm": 0.8933728933334351, | |
| "learning_rate": 4.737452002194185e-05, | |
| "loss": 0.3204296112060547, | |
| "step": 77250 | |
| }, | |
| { | |
| "epoch": 1.5846658466584667, | |
| "grad_norm": 0.8625443577766418, | |
| "learning_rate": 4.7340235874931435e-05, | |
| "loss": 0.3168534278869629, | |
| "step": 77300 | |
| }, | |
| { | |
| "epoch": 1.585690856908569, | |
| "grad_norm": 0.9073539972305298, | |
| "learning_rate": 4.730595172792101e-05, | |
| "loss": 0.31664215087890624, | |
| "step": 77350 | |
| }, | |
| { | |
| "epoch": 1.5867158671586716, | |
| "grad_norm": 0.8036795854568481, | |
| "learning_rate": 4.727166758091059e-05, | |
| "loss": 0.3167709732055664, | |
| "step": 77400 | |
| }, | |
| { | |
| "epoch": 1.5877408774087742, | |
| "grad_norm": 0.9465228319168091, | |
| "learning_rate": 4.7237383433900165e-05, | |
| "loss": 0.321316032409668, | |
| "step": 77450 | |
| }, | |
| { | |
| "epoch": 1.5887658876588766, | |
| "grad_norm": 0.7671904563903809, | |
| "learning_rate": 4.720309928688975e-05, | |
| "loss": 0.3174594879150391, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 1.589790897908979, | |
| "grad_norm": 0.784227192401886, | |
| "learning_rate": 4.716881513987932e-05, | |
| "loss": 0.3221343231201172, | |
| "step": 77550 | |
| }, | |
| { | |
| "epoch": 1.5908159081590816, | |
| "grad_norm": 0.8529446125030518, | |
| "learning_rate": 4.7134530992868896e-05, | |
| "loss": 0.32070777893066404, | |
| "step": 77600 | |
| }, | |
| { | |
| "epoch": 1.5918409184091842, | |
| "grad_norm": 0.9602042436599731, | |
| "learning_rate": 4.710024684585848e-05, | |
| "loss": 0.31466983795166015, | |
| "step": 77650 | |
| }, | |
| { | |
| "epoch": 1.5928659286592866, | |
| "grad_norm": 0.8479489684104919, | |
| "learning_rate": 4.706596269884805e-05, | |
| "loss": 0.32226348876953126, | |
| "step": 77700 | |
| }, | |
| { | |
| "epoch": 1.593890938909389, | |
| "grad_norm": 0.8461215496063232, | |
| "learning_rate": 4.7031678551837634e-05, | |
| "loss": 0.3161444091796875, | |
| "step": 77750 | |
| }, | |
| { | |
| "epoch": 1.5949159491594918, | |
| "grad_norm": 0.8258674740791321, | |
| "learning_rate": 4.699739440482721e-05, | |
| "loss": 0.31805326461791994, | |
| "step": 77800 | |
| }, | |
| { | |
| "epoch": 1.5959409594095941, | |
| "grad_norm": 0.8341616988182068, | |
| "learning_rate": 4.696311025781679e-05, | |
| "loss": 0.31697954177856447, | |
| "step": 77850 | |
| }, | |
| { | |
| "epoch": 1.5969659696596965, | |
| "grad_norm": 0.7861951589584351, | |
| "learning_rate": 4.6928826110806365e-05, | |
| "loss": 0.32761966705322265, | |
| "step": 77900 | |
| }, | |
| { | |
| "epoch": 1.597990979909799, | |
| "grad_norm": 0.8059016466140747, | |
| "learning_rate": 4.689454196379594e-05, | |
| "loss": 0.31410076141357424, | |
| "step": 77950 | |
| }, | |
| { | |
| "epoch": 1.5990159901599017, | |
| "grad_norm": 0.8756787776947021, | |
| "learning_rate": 4.686025781678552e-05, | |
| "loss": 0.32003135681152345, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 1.600041000410004, | |
| "grad_norm": 0.8756090402603149, | |
| "learning_rate": 4.6825973669775096e-05, | |
| "loss": 0.3205356979370117, | |
| "step": 78050 | |
| }, | |
| { | |
| "epoch": 1.6010660106601065, | |
| "grad_norm": 0.8523690700531006, | |
| "learning_rate": 4.679168952276468e-05, | |
| "loss": 0.3137340927124023, | |
| "step": 78100 | |
| }, | |
| { | |
| "epoch": 1.602091020910209, | |
| "grad_norm": 0.8872647881507874, | |
| "learning_rate": 4.675740537575425e-05, | |
| "loss": 0.31629999160766603, | |
| "step": 78150 | |
| }, | |
| { | |
| "epoch": 1.6031160311603116, | |
| "grad_norm": 0.8466306328773499, | |
| "learning_rate": 4.672312122874383e-05, | |
| "loss": 0.31956207275390625, | |
| "step": 78200 | |
| }, | |
| { | |
| "epoch": 1.604141041410414, | |
| "grad_norm": 0.879272997379303, | |
| "learning_rate": 4.668883708173341e-05, | |
| "loss": 0.31973594665527344, | |
| "step": 78250 | |
| }, | |
| { | |
| "epoch": 1.6051660516605166, | |
| "grad_norm": 0.7967390418052673, | |
| "learning_rate": 4.665455293472298e-05, | |
| "loss": 0.3170758247375488, | |
| "step": 78300 | |
| }, | |
| { | |
| "epoch": 1.6061910619106192, | |
| "grad_norm": 0.8612989783287048, | |
| "learning_rate": 4.6620268787712564e-05, | |
| "loss": 0.32166156768798826, | |
| "step": 78350 | |
| }, | |
| { | |
| "epoch": 1.6072160721607216, | |
| "grad_norm": 0.8083723783493042, | |
| "learning_rate": 4.658598464070214e-05, | |
| "loss": 0.3237931442260742, | |
| "step": 78400 | |
| }, | |
| { | |
| "epoch": 1.608241082410824, | |
| "grad_norm": 0.6994765400886536, | |
| "learning_rate": 4.655170049369172e-05, | |
| "loss": 0.31633190155029295, | |
| "step": 78450 | |
| }, | |
| { | |
| "epoch": 1.6092660926609266, | |
| "grad_norm": 0.8314966559410095, | |
| "learning_rate": 4.6517416346681295e-05, | |
| "loss": 0.3165827560424805, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 1.6102911029110292, | |
| "grad_norm": 0.7824987769126892, | |
| "learning_rate": 4.6483132199670876e-05, | |
| "loss": 0.3194588088989258, | |
| "step": 78550 | |
| }, | |
| { | |
| "epoch": 1.6113161131611315, | |
| "grad_norm": 0.8485012650489807, | |
| "learning_rate": 4.644884805266045e-05, | |
| "loss": 0.3204211807250977, | |
| "step": 78600 | |
| }, | |
| { | |
| "epoch": 1.6123411234112341, | |
| "grad_norm": 0.8031493425369263, | |
| "learning_rate": 4.6414563905650026e-05, | |
| "loss": 0.3155327796936035, | |
| "step": 78650 | |
| }, | |
| { | |
| "epoch": 1.6133661336613367, | |
| "grad_norm": 0.8817875385284424, | |
| "learning_rate": 4.638027975863961e-05, | |
| "loss": 0.31789566040039063, | |
| "step": 78700 | |
| }, | |
| { | |
| "epoch": 1.6143911439114391, | |
| "grad_norm": 0.7712017893791199, | |
| "learning_rate": 4.634599561162918e-05, | |
| "loss": 0.3121475601196289, | |
| "step": 78750 | |
| }, | |
| { | |
| "epoch": 1.6154161541615415, | |
| "grad_norm": 0.8248951435089111, | |
| "learning_rate": 4.6311711464618764e-05, | |
| "loss": 0.32087928771972657, | |
| "step": 78800 | |
| }, | |
| { | |
| "epoch": 1.616441164411644, | |
| "grad_norm": 0.8763154745101929, | |
| "learning_rate": 4.627742731760834e-05, | |
| "loss": 0.31577957153320313, | |
| "step": 78850 | |
| }, | |
| { | |
| "epoch": 1.6174661746617467, | |
| "grad_norm": 0.8384722471237183, | |
| "learning_rate": 4.624314317059792e-05, | |
| "loss": 0.3196330833435059, | |
| "step": 78900 | |
| }, | |
| { | |
| "epoch": 1.618491184911849, | |
| "grad_norm": 0.8924349546432495, | |
| "learning_rate": 4.6208859023587494e-05, | |
| "loss": 0.31931554794311523, | |
| "step": 78950 | |
| }, | |
| { | |
| "epoch": 1.6195161951619517, | |
| "grad_norm": 0.7989446520805359, | |
| "learning_rate": 4.617457487657707e-05, | |
| "loss": 0.31598247528076173, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 1.6205412054120543, | |
| "grad_norm": 0.8299421668052673, | |
| "learning_rate": 4.614029072956665e-05, | |
| "loss": 0.3145192718505859, | |
| "step": 79050 | |
| }, | |
| { | |
| "epoch": 1.6215662156621566, | |
| "grad_norm": 0.900879442691803, | |
| "learning_rate": 4.6106006582556225e-05, | |
| "loss": 0.3245106887817383, | |
| "step": 79100 | |
| }, | |
| { | |
| "epoch": 1.622591225912259, | |
| "grad_norm": 0.8314200043678284, | |
| "learning_rate": 4.607172243554581e-05, | |
| "loss": 0.318758602142334, | |
| "step": 79150 | |
| }, | |
| { | |
| "epoch": 1.6236162361623616, | |
| "grad_norm": 0.8202230334281921, | |
| "learning_rate": 4.603743828853538e-05, | |
| "loss": 0.32000885009765623, | |
| "step": 79200 | |
| }, | |
| { | |
| "epoch": 1.6246412464124642, | |
| "grad_norm": 0.9234268069267273, | |
| "learning_rate": 4.600315414152496e-05, | |
| "loss": 0.31623172760009766, | |
| "step": 79250 | |
| }, | |
| { | |
| "epoch": 1.6256662566625666, | |
| "grad_norm": 0.8224514126777649, | |
| "learning_rate": 4.596886999451454e-05, | |
| "loss": 0.3130918884277344, | |
| "step": 79300 | |
| }, | |
| { | |
| "epoch": 1.626691266912669, | |
| "grad_norm": 0.8268124461174011, | |
| "learning_rate": 4.593458584750411e-05, | |
| "loss": 0.31718255996704103, | |
| "step": 79350 | |
| }, | |
| { | |
| "epoch": 1.6277162771627718, | |
| "grad_norm": 0.8159886598587036, | |
| "learning_rate": 4.5900301700493694e-05, | |
| "loss": 0.3219700241088867, | |
| "step": 79400 | |
| }, | |
| { | |
| "epoch": 1.6287412874128742, | |
| "grad_norm": 0.7524468302726746, | |
| "learning_rate": 4.586601755348327e-05, | |
| "loss": 0.3117045783996582, | |
| "step": 79450 | |
| }, | |
| { | |
| "epoch": 1.6297662976629765, | |
| "grad_norm": 0.7985491156578064, | |
| "learning_rate": 4.583173340647285e-05, | |
| "loss": 0.3176197242736816, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 1.6307913079130791, | |
| "grad_norm": 0.7812376618385315, | |
| "learning_rate": 4.5797449259462425e-05, | |
| "loss": 0.3173367118835449, | |
| "step": 79550 | |
| }, | |
| { | |
| "epoch": 1.6318163181631817, | |
| "grad_norm": 0.8427786231040955, | |
| "learning_rate": 4.5763165112452006e-05, | |
| "loss": 0.3192551803588867, | |
| "step": 79600 | |
| }, | |
| { | |
| "epoch": 1.632841328413284, | |
| "grad_norm": 0.8390555381774902, | |
| "learning_rate": 4.572888096544158e-05, | |
| "loss": 0.3193316650390625, | |
| "step": 79650 | |
| }, | |
| { | |
| "epoch": 1.6338663386633865, | |
| "grad_norm": 0.8913291692733765, | |
| "learning_rate": 4.569459681843116e-05, | |
| "loss": 0.32151405334472655, | |
| "step": 79700 | |
| }, | |
| { | |
| "epoch": 1.634891348913489, | |
| "grad_norm": 0.8743128180503845, | |
| "learning_rate": 4.566031267142074e-05, | |
| "loss": 0.31926259994506834, | |
| "step": 79750 | |
| }, | |
| { | |
| "epoch": 1.6359163591635917, | |
| "grad_norm": 0.8355011343955994, | |
| "learning_rate": 4.562602852441031e-05, | |
| "loss": 0.31993555068969726, | |
| "step": 79800 | |
| }, | |
| { | |
| "epoch": 1.636941369413694, | |
| "grad_norm": 0.7864542007446289, | |
| "learning_rate": 4.559174437739989e-05, | |
| "loss": 0.31949954986572265, | |
| "step": 79850 | |
| }, | |
| { | |
| "epoch": 1.6379663796637967, | |
| "grad_norm": 0.9152125716209412, | |
| "learning_rate": 4.555746023038947e-05, | |
| "loss": 0.31908416748046875, | |
| "step": 79900 | |
| }, | |
| { | |
| "epoch": 1.6389913899138993, | |
| "grad_norm": 0.8982939720153809, | |
| "learning_rate": 4.552317608337905e-05, | |
| "loss": 0.3132705307006836, | |
| "step": 79950 | |
| }, | |
| { | |
| "epoch": 1.6400164001640016, | |
| "grad_norm": 0.8152450919151306, | |
| "learning_rate": 4.5488891936368624e-05, | |
| "loss": 0.31204761505126954, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 1.6400164001640016, | |
| "eval_loss": 0.38376733660697937, | |
| "eval_runtime": 4.4859, | |
| "eval_samples_per_second": 5572.965, | |
| "eval_steps_per_second": 5.573, | |
| "eval_teacher_cosine_loss": 0.24258661270141602, | |
| "eval_teacher_cosine_mean": 0.757413387298584, | |
| "eval_teacher_cosine_p10": 0.6804119944572449, | |
| "eval_teacher_cosine_p50": 0.7671465873718262, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 1.641041410414104, | |
| "grad_norm": 0.756226658821106, | |
| "learning_rate": 4.5454607789358205e-05, | |
| "loss": 0.3194527053833008, | |
| "step": 80050 | |
| }, | |
| { | |
| "epoch": 1.6420664206642066, | |
| "grad_norm": 0.8725249171257019, | |
| "learning_rate": 4.542032364234778e-05, | |
| "loss": 0.3204443359375, | |
| "step": 80100 | |
| }, | |
| { | |
| "epoch": 1.6430914309143092, | |
| "grad_norm": 0.8264670372009277, | |
| "learning_rate": 4.5386039495337355e-05, | |
| "loss": 0.3188313102722168, | |
| "step": 80150 | |
| }, | |
| { | |
| "epoch": 1.6441164411644116, | |
| "grad_norm": 0.7963871955871582, | |
| "learning_rate": 4.5351755348326936e-05, | |
| "loss": 0.31985857009887697, | |
| "step": 80200 | |
| }, | |
| { | |
| "epoch": 1.6451414514145142, | |
| "grad_norm": 0.8287241458892822, | |
| "learning_rate": 4.531747120131651e-05, | |
| "loss": 0.31706186294555666, | |
| "step": 80250 | |
| }, | |
| { | |
| "epoch": 1.6461664616646168, | |
| "grad_norm": 0.8354255557060242, | |
| "learning_rate": 4.528318705430609e-05, | |
| "loss": 0.3155230712890625, | |
| "step": 80300 | |
| }, | |
| { | |
| "epoch": 1.6471914719147192, | |
| "grad_norm": 0.8692095279693604, | |
| "learning_rate": 4.524890290729567e-05, | |
| "loss": 0.31708459854125975, | |
| "step": 80350 | |
| }, | |
| { | |
| "epoch": 1.6482164821648215, | |
| "grad_norm": 0.8380481600761414, | |
| "learning_rate": 4.521461876028525e-05, | |
| "loss": 0.3208130645751953, | |
| "step": 80400 | |
| }, | |
| { | |
| "epoch": 1.6492414924149241, | |
| "grad_norm": 0.8951699137687683, | |
| "learning_rate": 4.5180334613274823e-05, | |
| "loss": 0.32354423522949216, | |
| "step": 80450 | |
| }, | |
| { | |
| "epoch": 1.6502665026650267, | |
| "grad_norm": 0.8498013615608215, | |
| "learning_rate": 4.51460504662644e-05, | |
| "loss": 0.3154960250854492, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 1.651291512915129, | |
| "grad_norm": 0.9198926091194153, | |
| "learning_rate": 4.511176631925398e-05, | |
| "loss": 0.3150771903991699, | |
| "step": 80550 | |
| }, | |
| { | |
| "epoch": 1.6523165231652317, | |
| "grad_norm": 0.8761058449745178, | |
| "learning_rate": 4.5077482172243554e-05, | |
| "loss": 0.31855541229248047, | |
| "step": 80600 | |
| }, | |
| { | |
| "epoch": 1.6533415334153343, | |
| "grad_norm": 0.7531996369361877, | |
| "learning_rate": 4.5043198025233136e-05, | |
| "loss": 0.31778488159179685, | |
| "step": 80650 | |
| }, | |
| { | |
| "epoch": 1.6543665436654367, | |
| "grad_norm": 0.806756854057312, | |
| "learning_rate": 4.500891387822271e-05, | |
| "loss": 0.3184314155578613, | |
| "step": 80700 | |
| }, | |
| { | |
| "epoch": 1.655391553915539, | |
| "grad_norm": 0.8317981958389282, | |
| "learning_rate": 4.497462973121229e-05, | |
| "loss": 0.31850061416625974, | |
| "step": 80750 | |
| }, | |
| { | |
| "epoch": 1.6564165641656416, | |
| "grad_norm": 0.8248363733291626, | |
| "learning_rate": 4.4940345584201867e-05, | |
| "loss": 0.31640424728393557, | |
| "step": 80800 | |
| }, | |
| { | |
| "epoch": 1.6574415744157442, | |
| "grad_norm": 0.756259024143219, | |
| "learning_rate": 4.490606143719144e-05, | |
| "loss": 0.32035667419433594, | |
| "step": 80850 | |
| }, | |
| { | |
| "epoch": 1.6584665846658466, | |
| "grad_norm": 0.807367742061615, | |
| "learning_rate": 4.487177729018102e-05, | |
| "loss": 0.3208369445800781, | |
| "step": 80900 | |
| }, | |
| { | |
| "epoch": 1.659491594915949, | |
| "grad_norm": 0.8890630602836609, | |
| "learning_rate": 4.48374931431706e-05, | |
| "loss": 0.320812873840332, | |
| "step": 80950 | |
| }, | |
| { | |
| "epoch": 1.6605166051660518, | |
| "grad_norm": 0.8126643896102905, | |
| "learning_rate": 4.480320899616018e-05, | |
| "loss": 0.3125026702880859, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 1.6615416154161542, | |
| "grad_norm": 0.861640453338623, | |
| "learning_rate": 4.476892484914976e-05, | |
| "loss": 0.3149283981323242, | |
| "step": 81050 | |
| }, | |
| { | |
| "epoch": 1.6625666256662566, | |
| "grad_norm": 0.8799784183502197, | |
| "learning_rate": 4.4734640702139335e-05, | |
| "loss": 0.3186873626708984, | |
| "step": 81100 | |
| }, | |
| { | |
| "epoch": 1.6635916359163592, | |
| "grad_norm": 0.8192854523658752, | |
| "learning_rate": 4.470035655512891e-05, | |
| "loss": 0.3126109504699707, | |
| "step": 81150 | |
| }, | |
| { | |
| "epoch": 1.6646166461664618, | |
| "grad_norm": 0.8817254900932312, | |
| "learning_rate": 4.4666072408118485e-05, | |
| "loss": 0.31717355728149416, | |
| "step": 81200 | |
| }, | |
| { | |
| "epoch": 1.6656416564165641, | |
| "grad_norm": 0.8115575313568115, | |
| "learning_rate": 4.4631788261108066e-05, | |
| "loss": 0.3170262336730957, | |
| "step": 81250 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.9080811738967896, | |
| "learning_rate": 4.459750411409764e-05, | |
| "loss": 0.3217843246459961, | |
| "step": 81300 | |
| }, | |
| { | |
| "epoch": 1.6676916769167691, | |
| "grad_norm": 0.9741719365119934, | |
| "learning_rate": 4.456321996708722e-05, | |
| "loss": 0.31655244827270507, | |
| "step": 81350 | |
| }, | |
| { | |
| "epoch": 1.6687166871668717, | |
| "grad_norm": 0.7824592590332031, | |
| "learning_rate": 4.4528935820076804e-05, | |
| "loss": 0.3123858642578125, | |
| "step": 81400 | |
| }, | |
| { | |
| "epoch": 1.669741697416974, | |
| "grad_norm": 0.9863879680633545, | |
| "learning_rate": 4.449465167306638e-05, | |
| "loss": 0.3167323303222656, | |
| "step": 81450 | |
| }, | |
| { | |
| "epoch": 1.6707667076670767, | |
| "grad_norm": 0.8062689900398254, | |
| "learning_rate": 4.446036752605595e-05, | |
| "loss": 0.3201549530029297, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 1.6717917179171793, | |
| "grad_norm": 0.8985300660133362, | |
| "learning_rate": 4.442608337904553e-05, | |
| "loss": 0.3168154144287109, | |
| "step": 81550 | |
| }, | |
| { | |
| "epoch": 1.6728167281672817, | |
| "grad_norm": 0.8367087841033936, | |
| "learning_rate": 4.439179923203511e-05, | |
| "loss": 0.3198018836975098, | |
| "step": 81600 | |
| }, | |
| { | |
| "epoch": 1.673841738417384, | |
| "grad_norm": 0.7990803718566895, | |
| "learning_rate": 4.4357515085024684e-05, | |
| "loss": 0.3202411651611328, | |
| "step": 81650 | |
| }, | |
| { | |
| "epoch": 1.6748667486674866, | |
| "grad_norm": 0.8815580010414124, | |
| "learning_rate": 4.4323230938014265e-05, | |
| "loss": 0.3173732376098633, | |
| "step": 81700 | |
| }, | |
| { | |
| "epoch": 1.6758917589175892, | |
| "grad_norm": 0.8835466504096985, | |
| "learning_rate": 4.428894679100385e-05, | |
| "loss": 0.3147778701782227, | |
| "step": 81750 | |
| }, | |
| { | |
| "epoch": 1.6769167691676916, | |
| "grad_norm": 0.8459696769714355, | |
| "learning_rate": 4.425466264399342e-05, | |
| "loss": 0.31777645111083985, | |
| "step": 81800 | |
| }, | |
| { | |
| "epoch": 1.6779417794177942, | |
| "grad_norm": 0.8562559485435486, | |
| "learning_rate": 4.4220378496982996e-05, | |
| "loss": 0.3226267623901367, | |
| "step": 81850 | |
| }, | |
| { | |
| "epoch": 1.6789667896678968, | |
| "grad_norm": 0.8353957533836365, | |
| "learning_rate": 4.418609434997257e-05, | |
| "loss": 0.3125286483764648, | |
| "step": 81900 | |
| }, | |
| { | |
| "epoch": 1.6799917999179992, | |
| "grad_norm": 0.8082168102264404, | |
| "learning_rate": 4.415181020296215e-05, | |
| "loss": 0.31829252243041994, | |
| "step": 81950 | |
| }, | |
| { | |
| "epoch": 1.6810168101681016, | |
| "grad_norm": 0.80135178565979, | |
| "learning_rate": 4.411752605595173e-05, | |
| "loss": 0.3184130096435547, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 1.6820418204182042, | |
| "grad_norm": 0.9615581631660461, | |
| "learning_rate": 4.408324190894131e-05, | |
| "loss": 0.3139524459838867, | |
| "step": 82050 | |
| }, | |
| { | |
| "epoch": 1.6830668306683068, | |
| "grad_norm": 0.9389336109161377, | |
| "learning_rate": 4.404895776193089e-05, | |
| "loss": 0.31543596267700197, | |
| "step": 82100 | |
| }, | |
| { | |
| "epoch": 1.6840918409184091, | |
| "grad_norm": 0.88881915807724, | |
| "learning_rate": 4.4014673614920465e-05, | |
| "loss": 0.3115476989746094, | |
| "step": 82150 | |
| }, | |
| { | |
| "epoch": 1.6851168511685117, | |
| "grad_norm": 0.8746258616447449, | |
| "learning_rate": 4.398038946791004e-05, | |
| "loss": 0.31994855880737305, | |
| "step": 82200 | |
| }, | |
| { | |
| "epoch": 1.6861418614186143, | |
| "grad_norm": 0.9569081664085388, | |
| "learning_rate": 4.3946105320899614e-05, | |
| "loss": 0.3162792205810547, | |
| "step": 82250 | |
| }, | |
| { | |
| "epoch": 1.6871668716687167, | |
| "grad_norm": 0.8175927400588989, | |
| "learning_rate": 4.3911821173889196e-05, | |
| "loss": 0.32186553955078123, | |
| "step": 82300 | |
| }, | |
| { | |
| "epoch": 1.688191881918819, | |
| "grad_norm": 0.8115511536598206, | |
| "learning_rate": 4.387753702687877e-05, | |
| "loss": 0.3201601791381836, | |
| "step": 82350 | |
| }, | |
| { | |
| "epoch": 1.6892168921689217, | |
| "grad_norm": 0.9138662815093994, | |
| "learning_rate": 4.384325287986835e-05, | |
| "loss": 0.31565954208374025, | |
| "step": 82400 | |
| }, | |
| { | |
| "epoch": 1.6902419024190243, | |
| "grad_norm": 0.82341468334198, | |
| "learning_rate": 4.380896873285793e-05, | |
| "loss": 0.31377153396606444, | |
| "step": 82450 | |
| }, | |
| { | |
| "epoch": 1.6912669126691267, | |
| "grad_norm": 0.7550086379051208, | |
| "learning_rate": 4.377468458584751e-05, | |
| "loss": 0.31041648864746096, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 1.692291922919229, | |
| "grad_norm": 0.9127821922302246, | |
| "learning_rate": 4.374040043883708e-05, | |
| "loss": 0.31694808959960935, | |
| "step": 82550 | |
| }, | |
| { | |
| "epoch": 1.6933169331693319, | |
| "grad_norm": 0.7476402521133423, | |
| "learning_rate": 4.370611629182666e-05, | |
| "loss": 0.31502790451049806, | |
| "step": 82600 | |
| }, | |
| { | |
| "epoch": 1.6943419434194342, | |
| "grad_norm": 0.838546633720398, | |
| "learning_rate": 4.367183214481624e-05, | |
| "loss": 0.31489654541015627, | |
| "step": 82650 | |
| }, | |
| { | |
| "epoch": 1.6953669536695366, | |
| "grad_norm": 0.8744800686836243, | |
| "learning_rate": 4.3637547997805814e-05, | |
| "loss": 0.319176082611084, | |
| "step": 82700 | |
| }, | |
| { | |
| "epoch": 1.6963919639196392, | |
| "grad_norm": 0.8490485548973083, | |
| "learning_rate": 4.3603263850795395e-05, | |
| "loss": 0.31279666900634767, | |
| "step": 82750 | |
| }, | |
| { | |
| "epoch": 1.6974169741697418, | |
| "grad_norm": 0.7393656969070435, | |
| "learning_rate": 4.3568979703784976e-05, | |
| "loss": 0.31138097763061523, | |
| "step": 82800 | |
| }, | |
| { | |
| "epoch": 1.6984419844198442, | |
| "grad_norm": 0.8192431926727295, | |
| "learning_rate": 4.353469555677455e-05, | |
| "loss": 0.3154014778137207, | |
| "step": 82850 | |
| }, | |
| { | |
| "epoch": 1.6994669946699466, | |
| "grad_norm": 0.9681013226509094, | |
| "learning_rate": 4.3500411409764126e-05, | |
| "loss": 0.3145168685913086, | |
| "step": 82900 | |
| }, | |
| { | |
| "epoch": 1.7004920049200491, | |
| "grad_norm": 0.864546000957489, | |
| "learning_rate": 4.34661272627537e-05, | |
| "loss": 0.3185692024230957, | |
| "step": 82950 | |
| }, | |
| { | |
| "epoch": 1.7015170151701517, | |
| "grad_norm": 0.8715412616729736, | |
| "learning_rate": 4.343184311574328e-05, | |
| "loss": 0.317348518371582, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 1.7025420254202541, | |
| "grad_norm": 0.869002103805542, | |
| "learning_rate": 4.339755896873286e-05, | |
| "loss": 0.31150882720947265, | |
| "step": 83050 | |
| }, | |
| { | |
| "epoch": 1.7035670356703567, | |
| "grad_norm": 0.815898597240448, | |
| "learning_rate": 4.336327482172244e-05, | |
| "loss": 0.3211828231811523, | |
| "step": 83100 | |
| }, | |
| { | |
| "epoch": 1.7045920459204593, | |
| "grad_norm": 0.8185902833938599, | |
| "learning_rate": 4.332899067471202e-05, | |
| "loss": 0.3158570289611816, | |
| "step": 83150 | |
| }, | |
| { | |
| "epoch": 1.7056170561705617, | |
| "grad_norm": 0.8341085314750671, | |
| "learning_rate": 4.3294706527701594e-05, | |
| "loss": 0.3156291961669922, | |
| "step": 83200 | |
| }, | |
| { | |
| "epoch": 1.706642066420664, | |
| "grad_norm": 0.9095550775527954, | |
| "learning_rate": 4.326042238069117e-05, | |
| "loss": 0.31606042861938477, | |
| "step": 83250 | |
| }, | |
| { | |
| "epoch": 1.7076670766707667, | |
| "grad_norm": 0.852431058883667, | |
| "learning_rate": 4.3226138233680744e-05, | |
| "loss": 0.3126823043823242, | |
| "step": 83300 | |
| }, | |
| { | |
| "epoch": 1.7086920869208693, | |
| "grad_norm": 0.8617656230926514, | |
| "learning_rate": 4.3191854086670325e-05, | |
| "loss": 0.3163833236694336, | |
| "step": 83350 | |
| }, | |
| { | |
| "epoch": 1.7097170971709716, | |
| "grad_norm": 0.8765053153038025, | |
| "learning_rate": 4.31575699396599e-05, | |
| "loss": 0.3166981315612793, | |
| "step": 83400 | |
| }, | |
| { | |
| "epoch": 1.7107421074210742, | |
| "grad_norm": 0.897761881351471, | |
| "learning_rate": 4.312328579264948e-05, | |
| "loss": 0.321356315612793, | |
| "step": 83450 | |
| }, | |
| { | |
| "epoch": 1.7117671176711768, | |
| "grad_norm": 0.8739738464355469, | |
| "learning_rate": 4.308900164563906e-05, | |
| "loss": 0.3181402587890625, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 1.7127921279212792, | |
| "grad_norm": 0.943489134311676, | |
| "learning_rate": 4.305471749862864e-05, | |
| "loss": 0.31579544067382814, | |
| "step": 83550 | |
| }, | |
| { | |
| "epoch": 1.7138171381713816, | |
| "grad_norm": 0.9860527515411377, | |
| "learning_rate": 4.302043335161821e-05, | |
| "loss": 0.31612697601318357, | |
| "step": 83600 | |
| }, | |
| { | |
| "epoch": 1.7148421484214842, | |
| "grad_norm": 0.8367102146148682, | |
| "learning_rate": 4.298614920460779e-05, | |
| "loss": 0.31945114135742186, | |
| "step": 83650 | |
| }, | |
| { | |
| "epoch": 1.7158671586715868, | |
| "grad_norm": 0.8427906036376953, | |
| "learning_rate": 4.295186505759737e-05, | |
| "loss": 0.30977874755859375, | |
| "step": 83700 | |
| }, | |
| { | |
| "epoch": 1.7168921689216892, | |
| "grad_norm": 0.8448742628097534, | |
| "learning_rate": 4.291758091058694e-05, | |
| "loss": 0.3168920135498047, | |
| "step": 83750 | |
| }, | |
| { | |
| "epoch": 1.7179171791717918, | |
| "grad_norm": 0.8090877532958984, | |
| "learning_rate": 4.2883296763576525e-05, | |
| "loss": 0.3086431694030762, | |
| "step": 83800 | |
| }, | |
| { | |
| "epoch": 1.7189421894218944, | |
| "grad_norm": 0.9004344344139099, | |
| "learning_rate": 4.2849012616566106e-05, | |
| "loss": 0.31566299438476564, | |
| "step": 83850 | |
| }, | |
| { | |
| "epoch": 1.7199671996719967, | |
| "grad_norm": 0.8169884085655212, | |
| "learning_rate": 4.281472846955568e-05, | |
| "loss": 0.3168793296813965, | |
| "step": 83900 | |
| }, | |
| { | |
| "epoch": 1.7209922099220991, | |
| "grad_norm": 0.8668081760406494, | |
| "learning_rate": 4.2780444322545255e-05, | |
| "loss": 0.3174347496032715, | |
| "step": 83950 | |
| }, | |
| { | |
| "epoch": 1.7220172201722017, | |
| "grad_norm": 0.8710331916809082, | |
| "learning_rate": 4.274616017553483e-05, | |
| "loss": 0.31934385299682616, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 1.7230422304223043, | |
| "grad_norm": 0.8071451783180237, | |
| "learning_rate": 4.271187602852441e-05, | |
| "loss": 0.3175623893737793, | |
| "step": 84050 | |
| }, | |
| { | |
| "epoch": 1.7240672406724067, | |
| "grad_norm": 0.7706942558288574, | |
| "learning_rate": 4.2677591881513986e-05, | |
| "loss": 0.3150172233581543, | |
| "step": 84100 | |
| }, | |
| { | |
| "epoch": 1.725092250922509, | |
| "grad_norm": 0.8234313726425171, | |
| "learning_rate": 4.264330773450357e-05, | |
| "loss": 0.31650840759277343, | |
| "step": 84150 | |
| }, | |
| { | |
| "epoch": 1.7261172611726119, | |
| "grad_norm": 0.9378411769866943, | |
| "learning_rate": 4.260902358749315e-05, | |
| "loss": 0.3209552764892578, | |
| "step": 84200 | |
| }, | |
| { | |
| "epoch": 1.7271422714227143, | |
| "grad_norm": 0.856643557548523, | |
| "learning_rate": 4.2574739440482724e-05, | |
| "loss": 0.3155815315246582, | |
| "step": 84250 | |
| }, | |
| { | |
| "epoch": 1.7281672816728166, | |
| "grad_norm": 0.8629660606384277, | |
| "learning_rate": 4.25404552934723e-05, | |
| "loss": 0.3127460479736328, | |
| "step": 84300 | |
| }, | |
| { | |
| "epoch": 1.7291922919229192, | |
| "grad_norm": 0.7943331003189087, | |
| "learning_rate": 4.250617114646187e-05, | |
| "loss": 0.31839431762695314, | |
| "step": 84350 | |
| }, | |
| { | |
| "epoch": 1.7302173021730218, | |
| "grad_norm": 0.79832923412323, | |
| "learning_rate": 4.2471886999451455e-05, | |
| "loss": 0.31596994400024414, | |
| "step": 84400 | |
| }, | |
| { | |
| "epoch": 1.7312423124231242, | |
| "grad_norm": 0.8989018797874451, | |
| "learning_rate": 4.2437602852441036e-05, | |
| "loss": 0.316107292175293, | |
| "step": 84450 | |
| }, | |
| { | |
| "epoch": 1.7322673226732266, | |
| "grad_norm": 0.8412982225418091, | |
| "learning_rate": 4.240331870543061e-05, | |
| "loss": 0.3221923828125, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 1.7332923329233292, | |
| "grad_norm": 0.8065756559371948, | |
| "learning_rate": 4.236903455842019e-05, | |
| "loss": 0.309579963684082, | |
| "step": 84550 | |
| }, | |
| { | |
| "epoch": 1.7343173431734318, | |
| "grad_norm": 0.8151276111602783, | |
| "learning_rate": 4.233475041140977e-05, | |
| "loss": 0.3146748733520508, | |
| "step": 84600 | |
| }, | |
| { | |
| "epoch": 1.7353423534235342, | |
| "grad_norm": 0.8241016864776611, | |
| "learning_rate": 4.230046626439934e-05, | |
| "loss": 0.3137947463989258, | |
| "step": 84650 | |
| }, | |
| { | |
| "epoch": 1.7363673636736368, | |
| "grad_norm": 0.8123605847358704, | |
| "learning_rate": 4.2266182117388917e-05, | |
| "loss": 0.31762941360473634, | |
| "step": 84700 | |
| }, | |
| { | |
| "epoch": 1.7373923739237394, | |
| "grad_norm": 0.7920235991477966, | |
| "learning_rate": 4.22318979703785e-05, | |
| "loss": 0.31415367126464844, | |
| "step": 84750 | |
| }, | |
| { | |
| "epoch": 1.7384173841738417, | |
| "grad_norm": 0.8220278024673462, | |
| "learning_rate": 4.219761382336808e-05, | |
| "loss": 0.31424888610839846, | |
| "step": 84800 | |
| }, | |
| { | |
| "epoch": 1.739442394423944, | |
| "grad_norm": 0.7766710519790649, | |
| "learning_rate": 4.2163329676357654e-05, | |
| "loss": 0.3149404525756836, | |
| "step": 84850 | |
| }, | |
| { | |
| "epoch": 1.7404674046740467, | |
| "grad_norm": 0.8865143060684204, | |
| "learning_rate": 4.2129045529347236e-05, | |
| "loss": 0.31772159576416015, | |
| "step": 84900 | |
| }, | |
| { | |
| "epoch": 1.7414924149241493, | |
| "grad_norm": 0.8703212141990662, | |
| "learning_rate": 4.209476138233681e-05, | |
| "loss": 0.3192734146118164, | |
| "step": 84950 | |
| }, | |
| { | |
| "epoch": 1.7425174251742517, | |
| "grad_norm": 0.9038416743278503, | |
| "learning_rate": 4.206047723532639e-05, | |
| "loss": 0.31425621032714846, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 1.7425174251742517, | |
| "eval_loss": 0.3802510201931, | |
| "eval_runtime": 4.483, | |
| "eval_samples_per_second": 5576.596, | |
| "eval_steps_per_second": 5.577, | |
| "eval_teacher_cosine_loss": 0.2406424880027771, | |
| "eval_teacher_cosine_mean": 0.7593575119972229, | |
| "eval_teacher_cosine_p10": 0.682523787021637, | |
| "eval_teacher_cosine_p50": 0.7692955732345581, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 1.7435424354243543, | |
| "grad_norm": 0.803948163986206, | |
| "learning_rate": 4.202619308831596e-05, | |
| "loss": 0.3113981056213379, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 1.7445674456744569, | |
| "grad_norm": 0.9213075637817383, | |
| "learning_rate": 4.199190894130554e-05, | |
| "loss": 0.32190635681152346, | |
| "step": 85100 | |
| }, | |
| { | |
| "epoch": 1.7455924559245592, | |
| "grad_norm": 0.9158979058265686, | |
| "learning_rate": 4.195762479429512e-05, | |
| "loss": 0.3134999084472656, | |
| "step": 85150 | |
| }, | |
| { | |
| "epoch": 1.7466174661746616, | |
| "grad_norm": 0.8875510096549988, | |
| "learning_rate": 4.19233406472847e-05, | |
| "loss": 0.31230342864990235, | |
| "step": 85200 | |
| }, | |
| { | |
| "epoch": 1.7476424764247642, | |
| "grad_norm": 0.8591718077659607, | |
| "learning_rate": 4.188905650027428e-05, | |
| "loss": 0.31289531707763674, | |
| "step": 85250 | |
| }, | |
| { | |
| "epoch": 1.7486674866748668, | |
| "grad_norm": 0.8327205181121826, | |
| "learning_rate": 4.1854772353263854e-05, | |
| "loss": 0.3196695899963379, | |
| "step": 85300 | |
| }, | |
| { | |
| "epoch": 1.7496924969249692, | |
| "grad_norm": 0.8299922943115234, | |
| "learning_rate": 4.1820488206253435e-05, | |
| "loss": 0.31357904434204104, | |
| "step": 85350 | |
| }, | |
| { | |
| "epoch": 1.7507175071750718, | |
| "grad_norm": 0.8735456466674805, | |
| "learning_rate": 4.1786204059243e-05, | |
| "loss": 0.3163282775878906, | |
| "step": 85400 | |
| }, | |
| { | |
| "epoch": 1.7517425174251744, | |
| "grad_norm": 0.8367878794670105, | |
| "learning_rate": 4.1751919912232584e-05, | |
| "loss": 0.3163716125488281, | |
| "step": 85450 | |
| }, | |
| { | |
| "epoch": 1.7527675276752768, | |
| "grad_norm": 0.866584062576294, | |
| "learning_rate": 4.1717635765222166e-05, | |
| "loss": 0.31564319610595704, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 1.7537925379253791, | |
| "grad_norm": 0.9058444499969482, | |
| "learning_rate": 4.168335161821174e-05, | |
| "loss": 0.31540067672729494, | |
| "step": 85550 | |
| }, | |
| { | |
| "epoch": 1.7548175481754817, | |
| "grad_norm": 0.8175299763679504, | |
| "learning_rate": 4.164906747120132e-05, | |
| "loss": 0.3222164535522461, | |
| "step": 85600 | |
| }, | |
| { | |
| "epoch": 1.7558425584255843, | |
| "grad_norm": 0.8078338503837585, | |
| "learning_rate": 4.16147833241909e-05, | |
| "loss": 0.3136418151855469, | |
| "step": 85650 | |
| }, | |
| { | |
| "epoch": 1.7568675686756867, | |
| "grad_norm": 0.8302335143089294, | |
| "learning_rate": 4.158049917718048e-05, | |
| "loss": 0.3215254592895508, | |
| "step": 85700 | |
| }, | |
| { | |
| "epoch": 1.757892578925789, | |
| "grad_norm": 0.9128769636154175, | |
| "learning_rate": 4.1546215030170046e-05, | |
| "loss": 0.3118702697753906, | |
| "step": 85750 | |
| }, | |
| { | |
| "epoch": 1.758917589175892, | |
| "grad_norm": 0.8639569878578186, | |
| "learning_rate": 4.151193088315963e-05, | |
| "loss": 0.31508848190307615, | |
| "step": 85800 | |
| }, | |
| { | |
| "epoch": 1.7599425994259943, | |
| "grad_norm": 0.7977743148803711, | |
| "learning_rate": 4.147764673614921e-05, | |
| "loss": 0.30792913436889646, | |
| "step": 85850 | |
| }, | |
| { | |
| "epoch": 1.7609676096760967, | |
| "grad_norm": 0.8640775084495544, | |
| "learning_rate": 4.1443362589138784e-05, | |
| "loss": 0.31704881668090823, | |
| "step": 85900 | |
| }, | |
| { | |
| "epoch": 1.7619926199261993, | |
| "grad_norm": 0.8910190463066101, | |
| "learning_rate": 4.1409078442128365e-05, | |
| "loss": 0.31418630599975583, | |
| "step": 85950 | |
| }, | |
| { | |
| "epoch": 1.7630176301763019, | |
| "grad_norm": 0.8158953785896301, | |
| "learning_rate": 4.137479429511794e-05, | |
| "loss": 0.3146304512023926, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 1.7640426404264042, | |
| "grad_norm": 0.8988664150238037, | |
| "learning_rate": 4.134051014810752e-05, | |
| "loss": 0.3185373687744141, | |
| "step": 86050 | |
| }, | |
| { | |
| "epoch": 1.7650676506765066, | |
| "grad_norm": 0.8888332843780518, | |
| "learning_rate": 4.130622600109709e-05, | |
| "loss": 0.31441482543945315, | |
| "step": 86100 | |
| }, | |
| { | |
| "epoch": 1.7660926609266092, | |
| "grad_norm": 0.8594223856925964, | |
| "learning_rate": 4.127194185408667e-05, | |
| "loss": 0.31578617095947265, | |
| "step": 86150 | |
| }, | |
| { | |
| "epoch": 1.7671176711767118, | |
| "grad_norm": 0.837615966796875, | |
| "learning_rate": 4.123765770707625e-05, | |
| "loss": 0.3176837348937988, | |
| "step": 86200 | |
| }, | |
| { | |
| "epoch": 1.7681426814268142, | |
| "grad_norm": 0.8743645548820496, | |
| "learning_rate": 4.120337356006583e-05, | |
| "loss": 0.3102174186706543, | |
| "step": 86250 | |
| }, | |
| { | |
| "epoch": 1.7691676916769168, | |
| "grad_norm": 0.8491408824920654, | |
| "learning_rate": 4.116908941305541e-05, | |
| "loss": 0.3153289985656738, | |
| "step": 86300 | |
| }, | |
| { | |
| "epoch": 1.7701927019270194, | |
| "grad_norm": 0.8732030391693115, | |
| "learning_rate": 4.113480526604498e-05, | |
| "loss": 0.3086025619506836, | |
| "step": 86350 | |
| }, | |
| { | |
| "epoch": 1.7712177121771218, | |
| "grad_norm": 0.9224467873573303, | |
| "learning_rate": 4.1100521119034565e-05, | |
| "loss": 0.3165803146362305, | |
| "step": 86400 | |
| }, | |
| { | |
| "epoch": 1.7722427224272241, | |
| "grad_norm": 0.8902477622032166, | |
| "learning_rate": 4.106623697202413e-05, | |
| "loss": 0.3028519439697266, | |
| "step": 86450 | |
| }, | |
| { | |
| "epoch": 1.7732677326773267, | |
| "grad_norm": 0.7227466106414795, | |
| "learning_rate": 4.1031952825013714e-05, | |
| "loss": 0.31874622344970704, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 1.7742927429274293, | |
| "grad_norm": 0.8855125904083252, | |
| "learning_rate": 4.0997668678003296e-05, | |
| "loss": 0.3133975028991699, | |
| "step": 86550 | |
| }, | |
| { | |
| "epoch": 1.7753177531775317, | |
| "grad_norm": 0.7666642069816589, | |
| "learning_rate": 4.096338453099287e-05, | |
| "loss": 0.31351842880249026, | |
| "step": 86600 | |
| }, | |
| { | |
| "epoch": 1.7763427634276343, | |
| "grad_norm": 0.8613619804382324, | |
| "learning_rate": 4.092910038398245e-05, | |
| "loss": 0.3170045852661133, | |
| "step": 86650 | |
| }, | |
| { | |
| "epoch": 1.777367773677737, | |
| "grad_norm": 0.8553327918052673, | |
| "learning_rate": 4.0894816236972026e-05, | |
| "loss": 0.3146914863586426, | |
| "step": 86700 | |
| }, | |
| { | |
| "epoch": 1.7783927839278393, | |
| "grad_norm": 0.8119992613792419, | |
| "learning_rate": 4.086053208996161e-05, | |
| "loss": 0.3164428901672363, | |
| "step": 86750 | |
| }, | |
| { | |
| "epoch": 1.7794177941779417, | |
| "grad_norm": 0.8417336940765381, | |
| "learning_rate": 4.0826247942951176e-05, | |
| "loss": 0.31529386520385744, | |
| "step": 86800 | |
| }, | |
| { | |
| "epoch": 1.7804428044280443, | |
| "grad_norm": 0.9219096302986145, | |
| "learning_rate": 4.079196379594076e-05, | |
| "loss": 0.31346363067626953, | |
| "step": 86850 | |
| }, | |
| { | |
| "epoch": 1.7814678146781469, | |
| "grad_norm": 0.7500044703483582, | |
| "learning_rate": 4.075767964893034e-05, | |
| "loss": 0.31310190200805665, | |
| "step": 86900 | |
| }, | |
| { | |
| "epoch": 1.7824928249282492, | |
| "grad_norm": 0.8900172114372253, | |
| "learning_rate": 4.0723395501919913e-05, | |
| "loss": 0.31632951736450193, | |
| "step": 86950 | |
| }, | |
| { | |
| "epoch": 1.7835178351783518, | |
| "grad_norm": 0.8893556594848633, | |
| "learning_rate": 4.0689111354909495e-05, | |
| "loss": 0.3125075149536133, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 1.7845428454284544, | |
| "grad_norm": 0.9052404761314392, | |
| "learning_rate": 4.065482720789907e-05, | |
| "loss": 0.31383333206176756, | |
| "step": 87050 | |
| }, | |
| { | |
| "epoch": 1.7855678556785568, | |
| "grad_norm": 0.8347253203392029, | |
| "learning_rate": 4.062054306088865e-05, | |
| "loss": 0.3178304290771484, | |
| "step": 87100 | |
| }, | |
| { | |
| "epoch": 1.7865928659286592, | |
| "grad_norm": 0.7786207795143127, | |
| "learning_rate": 4.058625891387822e-05, | |
| "loss": 0.31245769500732423, | |
| "step": 87150 | |
| }, | |
| { | |
| "epoch": 1.7876178761787618, | |
| "grad_norm": 0.9215298295021057, | |
| "learning_rate": 4.05519747668678e-05, | |
| "loss": 0.3158077621459961, | |
| "step": 87200 | |
| }, | |
| { | |
| "epoch": 1.7886428864288644, | |
| "grad_norm": 0.8535469770431519, | |
| "learning_rate": 4.051769061985738e-05, | |
| "loss": 0.3166618537902832, | |
| "step": 87250 | |
| }, | |
| { | |
| "epoch": 1.7896678966789668, | |
| "grad_norm": 0.8999650478363037, | |
| "learning_rate": 4.048340647284696e-05, | |
| "loss": 0.3174513626098633, | |
| "step": 87300 | |
| }, | |
| { | |
| "epoch": 1.7906929069290691, | |
| "grad_norm": 0.9462832808494568, | |
| "learning_rate": 4.044912232583654e-05, | |
| "loss": 0.3125883674621582, | |
| "step": 87350 | |
| }, | |
| { | |
| "epoch": 1.791717917179172, | |
| "grad_norm": 0.8073539137840271, | |
| "learning_rate": 4.041483817882611e-05, | |
| "loss": 0.312409496307373, | |
| "step": 87400 | |
| }, | |
| { | |
| "epoch": 1.7927429274292743, | |
| "grad_norm": 0.8617629408836365, | |
| "learning_rate": 4.0380554031815694e-05, | |
| "loss": 0.31249921798706054, | |
| "step": 87450 | |
| }, | |
| { | |
| "epoch": 1.7937679376793767, | |
| "grad_norm": 0.8780073523521423, | |
| "learning_rate": 4.034626988480526e-05, | |
| "loss": 0.30783977508544924, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 1.7947929479294793, | |
| "grad_norm": 0.8536025285720825, | |
| "learning_rate": 4.0311985737794844e-05, | |
| "loss": 0.3151261901855469, | |
| "step": 87550 | |
| }, | |
| { | |
| "epoch": 1.795817958179582, | |
| "grad_norm": 0.8345057964324951, | |
| "learning_rate": 4.0277701590784425e-05, | |
| "loss": 0.3101858901977539, | |
| "step": 87600 | |
| }, | |
| { | |
| "epoch": 1.7968429684296843, | |
| "grad_norm": 0.8096317052841187, | |
| "learning_rate": 4.0243417443774e-05, | |
| "loss": 0.31848995208740233, | |
| "step": 87650 | |
| }, | |
| { | |
| "epoch": 1.7978679786797866, | |
| "grad_norm": 0.8475655913352966, | |
| "learning_rate": 4.020913329676358e-05, | |
| "loss": 0.31384050369262695, | |
| "step": 87700 | |
| }, | |
| { | |
| "epoch": 1.7988929889298892, | |
| "grad_norm": 0.7717837691307068, | |
| "learning_rate": 4.0174849149753156e-05, | |
| "loss": 0.3147487258911133, | |
| "step": 87750 | |
| }, | |
| { | |
| "epoch": 1.7999179991799918, | |
| "grad_norm": 0.8336133360862732, | |
| "learning_rate": 4.014056500274274e-05, | |
| "loss": 0.3127629470825195, | |
| "step": 87800 | |
| }, | |
| { | |
| "epoch": 1.8009430094300942, | |
| "grad_norm": 0.8419049382209778, | |
| "learning_rate": 4.010628085573231e-05, | |
| "loss": 0.31497175216674805, | |
| "step": 87850 | |
| }, | |
| { | |
| "epoch": 1.8019680196801968, | |
| "grad_norm": 0.8552167415618896, | |
| "learning_rate": 4.007199670872189e-05, | |
| "loss": 0.30863014221191404, | |
| "step": 87900 | |
| }, | |
| { | |
| "epoch": 1.8029930299302994, | |
| "grad_norm": 0.8534164428710938, | |
| "learning_rate": 4.003771256171147e-05, | |
| "loss": 0.31941667556762693, | |
| "step": 87950 | |
| }, | |
| { | |
| "epoch": 1.8040180401804018, | |
| "grad_norm": 0.7904766201972961, | |
| "learning_rate": 4.000342841470104e-05, | |
| "loss": 0.31321609497070313, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 1.8050430504305042, | |
| "grad_norm": 0.7977574467658997, | |
| "learning_rate": 3.9969144267690625e-05, | |
| "loss": 0.31240190505981447, | |
| "step": 88050 | |
| }, | |
| { | |
| "epoch": 1.8060680606806068, | |
| "grad_norm": 0.8627797961235046, | |
| "learning_rate": 3.99348601206802e-05, | |
| "loss": 0.3102535057067871, | |
| "step": 88100 | |
| }, | |
| { | |
| "epoch": 1.8070930709307094, | |
| "grad_norm": 0.8236440420150757, | |
| "learning_rate": 3.990057597366978e-05, | |
| "loss": 0.3128690338134766, | |
| "step": 88150 | |
| }, | |
| { | |
| "epoch": 1.8081180811808117, | |
| "grad_norm": 0.9193926453590393, | |
| "learning_rate": 3.9866291826659355e-05, | |
| "loss": 0.31826129913330076, | |
| "step": 88200 | |
| }, | |
| { | |
| "epoch": 1.8091430914309143, | |
| "grad_norm": 0.8584532141685486, | |
| "learning_rate": 3.983200767964893e-05, | |
| "loss": 0.3132435417175293, | |
| "step": 88250 | |
| }, | |
| { | |
| "epoch": 1.810168101681017, | |
| "grad_norm": 0.8374156355857849, | |
| "learning_rate": 3.979772353263851e-05, | |
| "loss": 0.3113271331787109, | |
| "step": 88300 | |
| }, | |
| { | |
| "epoch": 1.8111931119311193, | |
| "grad_norm": 0.8007743954658508, | |
| "learning_rate": 3.9763439385628086e-05, | |
| "loss": 0.30762434005737305, | |
| "step": 88350 | |
| }, | |
| { | |
| "epoch": 1.8122181221812217, | |
| "grad_norm": 0.8414162993431091, | |
| "learning_rate": 3.972915523861767e-05, | |
| "loss": 0.3169694709777832, | |
| "step": 88400 | |
| }, | |
| { | |
| "epoch": 1.8132431324313243, | |
| "grad_norm": 0.9336441159248352, | |
| "learning_rate": 3.969487109160724e-05, | |
| "loss": 0.31157283782958983, | |
| "step": 88450 | |
| }, | |
| { | |
| "epoch": 1.8142681426814269, | |
| "grad_norm": 0.8306562304496765, | |
| "learning_rate": 3.9660586944596824e-05, | |
| "loss": 0.30917240142822267, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 1.8152931529315293, | |
| "grad_norm": 0.7718772888183594, | |
| "learning_rate": 3.96263027975864e-05, | |
| "loss": 0.31303939819335935, | |
| "step": 88550 | |
| }, | |
| { | |
| "epoch": 1.8163181631816319, | |
| "grad_norm": 0.8513990044593811, | |
| "learning_rate": 3.959201865057597e-05, | |
| "loss": 0.3108053207397461, | |
| "step": 88600 | |
| }, | |
| { | |
| "epoch": 1.8173431734317345, | |
| "grad_norm": 0.8888369798660278, | |
| "learning_rate": 3.9557734503565555e-05, | |
| "loss": 0.31366466522216796, | |
| "step": 88650 | |
| }, | |
| { | |
| "epoch": 1.8183681836818368, | |
| "grad_norm": 0.8713319301605225, | |
| "learning_rate": 3.952345035655513e-05, | |
| "loss": 0.3139806747436523, | |
| "step": 88700 | |
| }, | |
| { | |
| "epoch": 1.8193931939319392, | |
| "grad_norm": 0.8174113631248474, | |
| "learning_rate": 3.948916620954471e-05, | |
| "loss": 0.3138068389892578, | |
| "step": 88750 | |
| }, | |
| { | |
| "epoch": 1.8204182041820418, | |
| "grad_norm": 0.8743528723716736, | |
| "learning_rate": 3.9454882062534286e-05, | |
| "loss": 0.3167488670349121, | |
| "step": 88800 | |
| }, | |
| { | |
| "epoch": 1.8214432144321444, | |
| "grad_norm": 0.8850277066230774, | |
| "learning_rate": 3.942059791552387e-05, | |
| "loss": 0.3157675361633301, | |
| "step": 88850 | |
| }, | |
| { | |
| "epoch": 1.8224682246822468, | |
| "grad_norm": 0.7077646255493164, | |
| "learning_rate": 3.938631376851344e-05, | |
| "loss": 0.31389339447021486, | |
| "step": 88900 | |
| }, | |
| { | |
| "epoch": 1.8234932349323492, | |
| "grad_norm": 0.8304405808448792, | |
| "learning_rate": 3.9352029621503017e-05, | |
| "loss": 0.31324703216552735, | |
| "step": 88950 | |
| }, | |
| { | |
| "epoch": 1.824518245182452, | |
| "grad_norm": 0.9053293466567993, | |
| "learning_rate": 3.93177454744926e-05, | |
| "loss": 0.3113551139831543, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 1.8255432554325544, | |
| "grad_norm": 0.8750311136245728, | |
| "learning_rate": 3.928346132748217e-05, | |
| "loss": 0.311038875579834, | |
| "step": 89050 | |
| }, | |
| { | |
| "epoch": 1.8265682656826567, | |
| "grad_norm": 0.8520994186401367, | |
| "learning_rate": 3.9249177180471754e-05, | |
| "loss": 0.310667839050293, | |
| "step": 89100 | |
| }, | |
| { | |
| "epoch": 1.8275932759327593, | |
| "grad_norm": 0.8947039842605591, | |
| "learning_rate": 3.921489303346133e-05, | |
| "loss": 0.30818552017211914, | |
| "step": 89150 | |
| }, | |
| { | |
| "epoch": 1.828618286182862, | |
| "grad_norm": 0.8681294918060303, | |
| "learning_rate": 3.918060888645091e-05, | |
| "loss": 0.30531021118164064, | |
| "step": 89200 | |
| }, | |
| { | |
| "epoch": 1.8296432964329643, | |
| "grad_norm": 0.8250561356544495, | |
| "learning_rate": 3.9146324739440485e-05, | |
| "loss": 0.3144289207458496, | |
| "step": 89250 | |
| }, | |
| { | |
| "epoch": 1.8306683066830667, | |
| "grad_norm": 0.784483790397644, | |
| "learning_rate": 3.911204059243006e-05, | |
| "loss": 0.31360071182250976, | |
| "step": 89300 | |
| }, | |
| { | |
| "epoch": 1.8316933169331693, | |
| "grad_norm": 0.913860023021698, | |
| "learning_rate": 3.907775644541964e-05, | |
| "loss": 0.31291168212890624, | |
| "step": 89350 | |
| }, | |
| { | |
| "epoch": 1.8327183271832719, | |
| "grad_norm": 0.87013179063797, | |
| "learning_rate": 3.9043472298409216e-05, | |
| "loss": 0.3136296844482422, | |
| "step": 89400 | |
| }, | |
| { | |
| "epoch": 1.8337433374333743, | |
| "grad_norm": 0.8029663562774658, | |
| "learning_rate": 3.90091881513988e-05, | |
| "loss": 0.3160441970825195, | |
| "step": 89450 | |
| }, | |
| { | |
| "epoch": 1.8347683476834769, | |
| "grad_norm": 0.8349223732948303, | |
| "learning_rate": 3.897490400438837e-05, | |
| "loss": 0.3168368339538574, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 1.8357933579335795, | |
| "grad_norm": 0.926318883895874, | |
| "learning_rate": 3.8940619857377954e-05, | |
| "loss": 0.31265701293945314, | |
| "step": 89550 | |
| }, | |
| { | |
| "epoch": 1.8368183681836818, | |
| "grad_norm": 0.8480372428894043, | |
| "learning_rate": 3.890633571036753e-05, | |
| "loss": 0.3110447120666504, | |
| "step": 89600 | |
| }, | |
| { | |
| "epoch": 1.8378433784337842, | |
| "grad_norm": 0.932624876499176, | |
| "learning_rate": 3.88720515633571e-05, | |
| "loss": 0.3132493591308594, | |
| "step": 89650 | |
| }, | |
| { | |
| "epoch": 1.8388683886838868, | |
| "grad_norm": 0.901386022567749, | |
| "learning_rate": 3.8837767416346684e-05, | |
| "loss": 0.311138801574707, | |
| "step": 89700 | |
| }, | |
| { | |
| "epoch": 1.8398933989339894, | |
| "grad_norm": 0.9316921234130859, | |
| "learning_rate": 3.880348326933626e-05, | |
| "loss": 0.31236949920654294, | |
| "step": 89750 | |
| }, | |
| { | |
| "epoch": 1.8409184091840918, | |
| "grad_norm": 0.9109938740730286, | |
| "learning_rate": 3.876919912232584e-05, | |
| "loss": 0.31103551864624024, | |
| "step": 89800 | |
| }, | |
| { | |
| "epoch": 1.8419434194341944, | |
| "grad_norm": 0.9348140954971313, | |
| "learning_rate": 3.8734914975315415e-05, | |
| "loss": 0.31293119430541994, | |
| "step": 89850 | |
| }, | |
| { | |
| "epoch": 1.842968429684297, | |
| "grad_norm": 0.842801034450531, | |
| "learning_rate": 3.8700630828305e-05, | |
| "loss": 0.30986536026000977, | |
| "step": 89900 | |
| }, | |
| { | |
| "epoch": 1.8439934399343993, | |
| "grad_norm": 0.8806145787239075, | |
| "learning_rate": 3.866634668129457e-05, | |
| "loss": 0.3130190658569336, | |
| "step": 89950 | |
| }, | |
| { | |
| "epoch": 1.8450184501845017, | |
| "grad_norm": 0.7844831347465515, | |
| "learning_rate": 3.8632062534284146e-05, | |
| "loss": 0.3079430770874023, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 1.8450184501845017, | |
| "eval_loss": 0.3754900097846985, | |
| "eval_runtime": 4.4595, | |
| "eval_samples_per_second": 5606.019, | |
| "eval_steps_per_second": 5.606, | |
| "eval_teacher_cosine_loss": 0.23911607265472412, | |
| "eval_teacher_cosine_mean": 0.7608839273452759, | |
| "eval_teacher_cosine_p10": 0.6841239929199219, | |
| "eval_teacher_cosine_p50": 0.771006166934967, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 1.8460434604346043, | |
| "grad_norm": 0.8056272864341736, | |
| "learning_rate": 3.859777838727373e-05, | |
| "loss": 0.30923244476318357, | |
| "step": 90050 | |
| }, | |
| { | |
| "epoch": 1.847068470684707, | |
| "grad_norm": 0.7544776201248169, | |
| "learning_rate": 3.85634942402633e-05, | |
| "loss": 0.3132191276550293, | |
| "step": 90100 | |
| }, | |
| { | |
| "epoch": 1.8480934809348093, | |
| "grad_norm": 0.8794836401939392, | |
| "learning_rate": 3.8529210093252884e-05, | |
| "loss": 0.3096630668640137, | |
| "step": 90150 | |
| }, | |
| { | |
| "epoch": 1.849118491184912, | |
| "grad_norm": 0.7869011163711548, | |
| "learning_rate": 3.849492594624246e-05, | |
| "loss": 0.305856876373291, | |
| "step": 90200 | |
| }, | |
| { | |
| "epoch": 1.8501435014350145, | |
| "grad_norm": 0.7537678480148315, | |
| "learning_rate": 3.846064179923204e-05, | |
| "loss": 0.31104705810546873, | |
| "step": 90250 | |
| }, | |
| { | |
| "epoch": 1.8511685116851169, | |
| "grad_norm": 1.005659580230713, | |
| "learning_rate": 3.8426357652221615e-05, | |
| "loss": 0.3173058891296387, | |
| "step": 90300 | |
| }, | |
| { | |
| "epoch": 1.8521935219352192, | |
| "grad_norm": 0.9292976260185242, | |
| "learning_rate": 3.839207350521119e-05, | |
| "loss": 0.3144725227355957, | |
| "step": 90350 | |
| }, | |
| { | |
| "epoch": 1.8532185321853218, | |
| "grad_norm": 0.8585899472236633, | |
| "learning_rate": 3.835778935820077e-05, | |
| "loss": 0.31907258987426756, | |
| "step": 90400 | |
| }, | |
| { | |
| "epoch": 1.8542435424354244, | |
| "grad_norm": 0.8002761602401733, | |
| "learning_rate": 3.8323505211190346e-05, | |
| "loss": 0.30756336212158203, | |
| "step": 90450 | |
| }, | |
| { | |
| "epoch": 1.8552685526855268, | |
| "grad_norm": 0.827458918094635, | |
| "learning_rate": 3.828922106417993e-05, | |
| "loss": 0.31659698486328125, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 1.8562935629356292, | |
| "grad_norm": 0.9097380638122559, | |
| "learning_rate": 3.82549369171695e-05, | |
| "loss": 0.3115439605712891, | |
| "step": 90550 | |
| }, | |
| { | |
| "epoch": 1.857318573185732, | |
| "grad_norm": 0.8812702298164368, | |
| "learning_rate": 3.822065277015908e-05, | |
| "loss": 0.30891057968139646, | |
| "step": 90600 | |
| }, | |
| { | |
| "epoch": 1.8583435834358344, | |
| "grad_norm": 0.8173109292984009, | |
| "learning_rate": 3.818636862314866e-05, | |
| "loss": 0.3152191734313965, | |
| "step": 90650 | |
| }, | |
| { | |
| "epoch": 1.8593685936859368, | |
| "grad_norm": 0.8117443323135376, | |
| "learning_rate": 3.815208447613823e-05, | |
| "loss": 0.3070538139343262, | |
| "step": 90700 | |
| }, | |
| { | |
| "epoch": 1.8603936039360394, | |
| "grad_norm": 0.8733604550361633, | |
| "learning_rate": 3.8117800329127814e-05, | |
| "loss": 0.30416648864746093, | |
| "step": 90750 | |
| }, | |
| { | |
| "epoch": 1.861418614186142, | |
| "grad_norm": 0.9064242243766785, | |
| "learning_rate": 3.808351618211739e-05, | |
| "loss": 0.3043059349060059, | |
| "step": 90800 | |
| }, | |
| { | |
| "epoch": 1.8624436244362443, | |
| "grad_norm": 0.8272138833999634, | |
| "learning_rate": 3.804923203510697e-05, | |
| "loss": 0.31302532196044924, | |
| "step": 90850 | |
| }, | |
| { | |
| "epoch": 1.8634686346863467, | |
| "grad_norm": 0.9128270745277405, | |
| "learning_rate": 3.8014947888096545e-05, | |
| "loss": 0.3142374801635742, | |
| "step": 90900 | |
| }, | |
| { | |
| "epoch": 1.8644936449364493, | |
| "grad_norm": 0.8989775776863098, | |
| "learning_rate": 3.7980663741086126e-05, | |
| "loss": 0.3138454055786133, | |
| "step": 90950 | |
| }, | |
| { | |
| "epoch": 1.865518655186552, | |
| "grad_norm": 0.8963668346405029, | |
| "learning_rate": 3.79463795940757e-05, | |
| "loss": 0.30724027633666995, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 1.8665436654366543, | |
| "grad_norm": 0.8935912847518921, | |
| "learning_rate": 3.7912095447065276e-05, | |
| "loss": 0.31382911682128906, | |
| "step": 91050 | |
| }, | |
| { | |
| "epoch": 1.8675686756867569, | |
| "grad_norm": 0.90042644739151, | |
| "learning_rate": 3.787781130005486e-05, | |
| "loss": 0.3047069549560547, | |
| "step": 91100 | |
| }, | |
| { | |
| "epoch": 1.8685936859368595, | |
| "grad_norm": 0.8628675937652588, | |
| "learning_rate": 3.784352715304443e-05, | |
| "loss": 0.31003982543945313, | |
| "step": 91150 | |
| }, | |
| { | |
| "epoch": 1.8696186961869619, | |
| "grad_norm": 0.9214758276939392, | |
| "learning_rate": 3.7809243006034013e-05, | |
| "loss": 0.30603530883789065, | |
| "step": 91200 | |
| }, | |
| { | |
| "epoch": 1.8706437064370642, | |
| "grad_norm": 0.9035608768463135, | |
| "learning_rate": 3.777495885902359e-05, | |
| "loss": 0.30740699768066404, | |
| "step": 91250 | |
| }, | |
| { | |
| "epoch": 1.8716687166871668, | |
| "grad_norm": 0.8686066269874573, | |
| "learning_rate": 3.774067471201317e-05, | |
| "loss": 0.31121316909790037, | |
| "step": 91300 | |
| }, | |
| { | |
| "epoch": 1.8726937269372694, | |
| "grad_norm": 0.8364519476890564, | |
| "learning_rate": 3.7706390565002744e-05, | |
| "loss": 0.3125648498535156, | |
| "step": 91350 | |
| }, | |
| { | |
| "epoch": 1.8737187371873718, | |
| "grad_norm": 0.9814826250076294, | |
| "learning_rate": 3.767210641799232e-05, | |
| "loss": 0.3133916473388672, | |
| "step": 91400 | |
| }, | |
| { | |
| "epoch": 1.8747437474374744, | |
| "grad_norm": 0.8157342076301575, | |
| "learning_rate": 3.76378222709819e-05, | |
| "loss": 0.3074573707580566, | |
| "step": 91450 | |
| }, | |
| { | |
| "epoch": 1.875768757687577, | |
| "grad_norm": 0.8193994164466858, | |
| "learning_rate": 3.7603538123971475e-05, | |
| "loss": 0.31188503265380857, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 1.8767937679376794, | |
| "grad_norm": 0.7453115582466125, | |
| "learning_rate": 3.756925397696106e-05, | |
| "loss": 0.31023818969726563, | |
| "step": 91550 | |
| }, | |
| { | |
| "epoch": 1.8778187781877818, | |
| "grad_norm": 0.8813400864601135, | |
| "learning_rate": 3.753496982995063e-05, | |
| "loss": 0.3045553970336914, | |
| "step": 91600 | |
| }, | |
| { | |
| "epoch": 1.8788437884378844, | |
| "grad_norm": 0.7892765998840332, | |
| "learning_rate": 3.750068568294021e-05, | |
| "loss": 0.31083791732788085, | |
| "step": 91650 | |
| }, | |
| { | |
| "epoch": 1.879868798687987, | |
| "grad_norm": 0.890006422996521, | |
| "learning_rate": 3.746640153592979e-05, | |
| "loss": 0.3087233352661133, | |
| "step": 91700 | |
| }, | |
| { | |
| "epoch": 1.8808938089380893, | |
| "grad_norm": 0.9137267470359802, | |
| "learning_rate": 3.743211738891936e-05, | |
| "loss": 0.3125697326660156, | |
| "step": 91750 | |
| }, | |
| { | |
| "epoch": 1.881918819188192, | |
| "grad_norm": 0.7856095433235168, | |
| "learning_rate": 3.7397833241908944e-05, | |
| "loss": 0.31155101776123045, | |
| "step": 91800 | |
| }, | |
| { | |
| "epoch": 1.8829438294382945, | |
| "grad_norm": 0.8004966974258423, | |
| "learning_rate": 3.736354909489852e-05, | |
| "loss": 0.3110293960571289, | |
| "step": 91850 | |
| }, | |
| { | |
| "epoch": 1.883968839688397, | |
| "grad_norm": 0.9304362535476685, | |
| "learning_rate": 3.73292649478881e-05, | |
| "loss": 0.31033958435058595, | |
| "step": 91900 | |
| }, | |
| { | |
| "epoch": 1.8849938499384993, | |
| "grad_norm": 0.8364766836166382, | |
| "learning_rate": 3.7294980800877675e-05, | |
| "loss": 0.3089278221130371, | |
| "step": 91950 | |
| }, | |
| { | |
| "epoch": 1.8860188601886019, | |
| "grad_norm": 0.8403012752532959, | |
| "learning_rate": 3.7260696653867256e-05, | |
| "loss": 0.314017333984375, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 1.8870438704387045, | |
| "grad_norm": 0.9320840835571289, | |
| "learning_rate": 3.722641250685683e-05, | |
| "loss": 0.3071384620666504, | |
| "step": 92050 | |
| }, | |
| { | |
| "epoch": 1.8880688806888068, | |
| "grad_norm": 0.9273920059204102, | |
| "learning_rate": 3.7192128359846405e-05, | |
| "loss": 0.31021919250488283, | |
| "step": 92100 | |
| }, | |
| { | |
| "epoch": 1.8890938909389092, | |
| "grad_norm": 0.8838326334953308, | |
| "learning_rate": 3.715784421283599e-05, | |
| "loss": 0.3120521926879883, | |
| "step": 92150 | |
| }, | |
| { | |
| "epoch": 1.890118901189012, | |
| "grad_norm": 0.9118202924728394, | |
| "learning_rate": 3.712356006582556e-05, | |
| "loss": 0.3099850654602051, | |
| "step": 92200 | |
| }, | |
| { | |
| "epoch": 1.8911439114391144, | |
| "grad_norm": 0.8146107196807861, | |
| "learning_rate": 3.708927591881514e-05, | |
| "loss": 0.3118232536315918, | |
| "step": 92250 | |
| }, | |
| { | |
| "epoch": 1.8921689216892168, | |
| "grad_norm": 0.8996651768684387, | |
| "learning_rate": 3.705499177180472e-05, | |
| "loss": 0.3095920181274414, | |
| "step": 92300 | |
| }, | |
| { | |
| "epoch": 1.8931939319393194, | |
| "grad_norm": 0.8883601427078247, | |
| "learning_rate": 3.70207076247943e-05, | |
| "loss": 0.31027421951293943, | |
| "step": 92350 | |
| }, | |
| { | |
| "epoch": 1.894218942189422, | |
| "grad_norm": 0.8014428615570068, | |
| "learning_rate": 3.6986423477783874e-05, | |
| "loss": 0.3093454551696777, | |
| "step": 92400 | |
| }, | |
| { | |
| "epoch": 1.8952439524395244, | |
| "grad_norm": 0.8479575514793396, | |
| "learning_rate": 3.695213933077345e-05, | |
| "loss": 0.30914844512939454, | |
| "step": 92450 | |
| }, | |
| { | |
| "epoch": 1.8962689626896267, | |
| "grad_norm": 0.8123742341995239, | |
| "learning_rate": 3.691785518376303e-05, | |
| "loss": 0.30591440200805664, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 1.8972939729397293, | |
| "grad_norm": 0.9305078983306885, | |
| "learning_rate": 3.6883571036752605e-05, | |
| "loss": 0.3118933296203613, | |
| "step": 92550 | |
| }, | |
| { | |
| "epoch": 1.898318983189832, | |
| "grad_norm": 0.8503822088241577, | |
| "learning_rate": 3.6849286889742186e-05, | |
| "loss": 0.30845970153808594, | |
| "step": 92600 | |
| }, | |
| { | |
| "epoch": 1.8993439934399343, | |
| "grad_norm": 0.8877168893814087, | |
| "learning_rate": 3.681500274273176e-05, | |
| "loss": 0.31343606948852537, | |
| "step": 92650 | |
| }, | |
| { | |
| "epoch": 1.900369003690037, | |
| "grad_norm": 0.8192229270935059, | |
| "learning_rate": 3.678071859572134e-05, | |
| "loss": 0.3070513725280762, | |
| "step": 92700 | |
| }, | |
| { | |
| "epoch": 1.9013940139401395, | |
| "grad_norm": 0.8203495740890503, | |
| "learning_rate": 3.674643444871092e-05, | |
| "loss": 0.306214599609375, | |
| "step": 92750 | |
| }, | |
| { | |
| "epoch": 1.902419024190242, | |
| "grad_norm": 0.8231675624847412, | |
| "learning_rate": 3.671215030170049e-05, | |
| "loss": 0.3100650978088379, | |
| "step": 92800 | |
| }, | |
| { | |
| "epoch": 1.9034440344403443, | |
| "grad_norm": 0.9266271591186523, | |
| "learning_rate": 3.667786615469007e-05, | |
| "loss": 0.3130646705627441, | |
| "step": 92850 | |
| }, | |
| { | |
| "epoch": 1.9044690446904469, | |
| "grad_norm": 0.7471237182617188, | |
| "learning_rate": 3.664358200767965e-05, | |
| "loss": 0.3050639915466309, | |
| "step": 92900 | |
| }, | |
| { | |
| "epoch": 1.9054940549405495, | |
| "grad_norm": 0.7709605097770691, | |
| "learning_rate": 3.660929786066923e-05, | |
| "loss": 0.31078922271728515, | |
| "step": 92950 | |
| }, | |
| { | |
| "epoch": 1.9065190651906518, | |
| "grad_norm": 0.850380539894104, | |
| "learning_rate": 3.6575013713658804e-05, | |
| "loss": 0.30950761795043946, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 1.9075440754407544, | |
| "grad_norm": 0.895480751991272, | |
| "learning_rate": 3.6540729566648386e-05, | |
| "loss": 0.30479257583618163, | |
| "step": 93050 | |
| }, | |
| { | |
| "epoch": 1.908569085690857, | |
| "grad_norm": 0.8471795320510864, | |
| "learning_rate": 3.650644541963796e-05, | |
| "loss": 0.3094027900695801, | |
| "step": 93100 | |
| }, | |
| { | |
| "epoch": 1.9095940959409594, | |
| "grad_norm": 0.9467073678970337, | |
| "learning_rate": 3.6472161272627535e-05, | |
| "loss": 0.3046421051025391, | |
| "step": 93150 | |
| }, | |
| { | |
| "epoch": 1.9106191061910618, | |
| "grad_norm": 0.8307456374168396, | |
| "learning_rate": 3.6437877125617117e-05, | |
| "loss": 0.306127986907959, | |
| "step": 93200 | |
| }, | |
| { | |
| "epoch": 1.9116441164411644, | |
| "grad_norm": 0.9045378565788269, | |
| "learning_rate": 3.640359297860669e-05, | |
| "loss": 0.31133806228637695, | |
| "step": 93250 | |
| }, | |
| { | |
| "epoch": 1.912669126691267, | |
| "grad_norm": 0.8128647208213806, | |
| "learning_rate": 3.636930883159627e-05, | |
| "loss": 0.30952239990234376, | |
| "step": 93300 | |
| }, | |
| { | |
| "epoch": 1.9136941369413694, | |
| "grad_norm": 0.9846119284629822, | |
| "learning_rate": 3.633502468458585e-05, | |
| "loss": 0.3077969551086426, | |
| "step": 93350 | |
| }, | |
| { | |
| "epoch": 1.914719147191472, | |
| "grad_norm": 0.8450021743774414, | |
| "learning_rate": 3.630074053757543e-05, | |
| "loss": 0.3144134712219238, | |
| "step": 93400 | |
| }, | |
| { | |
| "epoch": 1.9157441574415746, | |
| "grad_norm": 0.8476465344429016, | |
| "learning_rate": 3.6266456390565004e-05, | |
| "loss": 0.3113539695739746, | |
| "step": 93450 | |
| }, | |
| { | |
| "epoch": 1.916769167691677, | |
| "grad_norm": 0.8455567359924316, | |
| "learning_rate": 3.623217224355458e-05, | |
| "loss": 0.30703603744506835, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 1.9177941779417793, | |
| "grad_norm": 0.8679808974266052, | |
| "learning_rate": 3.619788809654416e-05, | |
| "loss": 0.3133483695983887, | |
| "step": 93550 | |
| }, | |
| { | |
| "epoch": 1.918819188191882, | |
| "grad_norm": 0.8686735033988953, | |
| "learning_rate": 3.6163603949533734e-05, | |
| "loss": 0.3086144638061523, | |
| "step": 93600 | |
| }, | |
| { | |
| "epoch": 1.9198441984419845, | |
| "grad_norm": 0.8812918663024902, | |
| "learning_rate": 3.6129319802523316e-05, | |
| "loss": 0.3122638893127441, | |
| "step": 93650 | |
| }, | |
| { | |
| "epoch": 1.9208692086920869, | |
| "grad_norm": 0.8492827415466309, | |
| "learning_rate": 3.609503565551289e-05, | |
| "loss": 0.3102984237670898, | |
| "step": 93700 | |
| }, | |
| { | |
| "epoch": 1.9218942189421893, | |
| "grad_norm": 0.8537363409996033, | |
| "learning_rate": 3.606075150850247e-05, | |
| "loss": 0.30946544647216795, | |
| "step": 93750 | |
| }, | |
| { | |
| "epoch": 1.922919229192292, | |
| "grad_norm": 0.8917142748832703, | |
| "learning_rate": 3.602646736149205e-05, | |
| "loss": 0.3128114318847656, | |
| "step": 93800 | |
| }, | |
| { | |
| "epoch": 1.9239442394423945, | |
| "grad_norm": 0.8372023701667786, | |
| "learning_rate": 3.599218321448162e-05, | |
| "loss": 0.3060482215881348, | |
| "step": 93850 | |
| }, | |
| { | |
| "epoch": 1.9249692496924968, | |
| "grad_norm": 0.8893210291862488, | |
| "learning_rate": 3.59578990674712e-05, | |
| "loss": 0.30943252563476564, | |
| "step": 93900 | |
| }, | |
| { | |
| "epoch": 1.9259942599425994, | |
| "grad_norm": 0.9047923684120178, | |
| "learning_rate": 3.592361492046078e-05, | |
| "loss": 0.3131132698059082, | |
| "step": 93950 | |
| }, | |
| { | |
| "epoch": 1.927019270192702, | |
| "grad_norm": 0.8870962858200073, | |
| "learning_rate": 3.588933077345036e-05, | |
| "loss": 0.3083792495727539, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 1.9280442804428044, | |
| "grad_norm": 0.9483633637428284, | |
| "learning_rate": 3.5855046626439934e-05, | |
| "loss": 0.31549835205078125, | |
| "step": 94050 | |
| }, | |
| { | |
| "epoch": 1.9290692906929068, | |
| "grad_norm": 0.8161036968231201, | |
| "learning_rate": 3.5820762479429515e-05, | |
| "loss": 0.31491214752197266, | |
| "step": 94100 | |
| }, | |
| { | |
| "epoch": 1.9300943009430094, | |
| "grad_norm": 0.8917114734649658, | |
| "learning_rate": 3.578647833241909e-05, | |
| "loss": 0.31423368453979494, | |
| "step": 94150 | |
| }, | |
| { | |
| "epoch": 1.931119311193112, | |
| "grad_norm": 0.9384127259254456, | |
| "learning_rate": 3.5752194185408665e-05, | |
| "loss": 0.3117011070251465, | |
| "step": 94200 | |
| }, | |
| { | |
| "epoch": 1.9321443214432144, | |
| "grad_norm": 0.9139137268066406, | |
| "learning_rate": 3.5717910038398246e-05, | |
| "loss": 0.31228559494018554, | |
| "step": 94250 | |
| }, | |
| { | |
| "epoch": 1.933169331693317, | |
| "grad_norm": 0.9403935074806213, | |
| "learning_rate": 3.568362589138782e-05, | |
| "loss": 0.3054885673522949, | |
| "step": 94300 | |
| }, | |
| { | |
| "epoch": 1.9341943419434195, | |
| "grad_norm": 0.9717183113098145, | |
| "learning_rate": 3.56493417443774e-05, | |
| "loss": 0.3085132026672363, | |
| "step": 94350 | |
| }, | |
| { | |
| "epoch": 1.935219352193522, | |
| "grad_norm": 0.916111171245575, | |
| "learning_rate": 3.561505759736698e-05, | |
| "loss": 0.3123166275024414, | |
| "step": 94400 | |
| }, | |
| { | |
| "epoch": 1.9362443624436243, | |
| "grad_norm": 1.0178413391113281, | |
| "learning_rate": 3.558077345035656e-05, | |
| "loss": 0.3084921455383301, | |
| "step": 94450 | |
| }, | |
| { | |
| "epoch": 1.937269372693727, | |
| "grad_norm": 0.8317949771881104, | |
| "learning_rate": 3.554648930334614e-05, | |
| "loss": 0.311522216796875, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 1.9382943829438295, | |
| "grad_norm": 0.8449988961219788, | |
| "learning_rate": 3.551220515633571e-05, | |
| "loss": 0.3088962173461914, | |
| "step": 94550 | |
| }, | |
| { | |
| "epoch": 1.9393193931939319, | |
| "grad_norm": 0.9030212759971619, | |
| "learning_rate": 3.547792100932529e-05, | |
| "loss": 0.3060719871520996, | |
| "step": 94600 | |
| }, | |
| { | |
| "epoch": 1.9403444034440345, | |
| "grad_norm": 0.7787507772445679, | |
| "learning_rate": 3.5443636862314864e-05, | |
| "loss": 0.30561012268066406, | |
| "step": 94650 | |
| }, | |
| { | |
| "epoch": 1.941369413694137, | |
| "grad_norm": 0.8310166597366333, | |
| "learning_rate": 3.5409352715304446e-05, | |
| "loss": 0.3025302505493164, | |
| "step": 94700 | |
| }, | |
| { | |
| "epoch": 1.9423944239442394, | |
| "grad_norm": 0.9542337656021118, | |
| "learning_rate": 3.537506856829402e-05, | |
| "loss": 0.30956947326660156, | |
| "step": 94750 | |
| }, | |
| { | |
| "epoch": 1.9434194341943418, | |
| "grad_norm": 0.8787341713905334, | |
| "learning_rate": 3.53407844212836e-05, | |
| "loss": 0.312775764465332, | |
| "step": 94800 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.8923402428627014, | |
| "learning_rate": 3.530650027427318e-05, | |
| "loss": 0.30923460006713865, | |
| "step": 94850 | |
| }, | |
| { | |
| "epoch": 1.945469454694547, | |
| "grad_norm": 0.9525632858276367, | |
| "learning_rate": 3.527221612726275e-05, | |
| "loss": 0.3075278663635254, | |
| "step": 94900 | |
| }, | |
| { | |
| "epoch": 1.9464944649446494, | |
| "grad_norm": 0.8809632658958435, | |
| "learning_rate": 3.523793198025233e-05, | |
| "loss": 0.31187726974487306, | |
| "step": 94950 | |
| }, | |
| { | |
| "epoch": 1.947519475194752, | |
| "grad_norm": 0.8053218722343445, | |
| "learning_rate": 3.520364783324191e-05, | |
| "loss": 0.30278476715087893, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 1.947519475194752, | |
| "eval_loss": 0.37200942635536194, | |
| "eval_runtime": 4.4874, | |
| "eval_samples_per_second": 5571.133, | |
| "eval_steps_per_second": 5.571, | |
| "eval_teacher_cosine_loss": 0.2359941601753235, | |
| "eval_teacher_cosine_mean": 0.7640058398246765, | |
| "eval_teacher_cosine_p10": 0.6876839399337769, | |
| "eval_teacher_cosine_p50": 0.7738487124443054, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 1.9485444854448546, | |
| "grad_norm": 0.8830417990684509, | |
| "learning_rate": 3.516936368623149e-05, | |
| "loss": 0.307344970703125, | |
| "step": 95050 | |
| }, | |
| { | |
| "epoch": 1.949569495694957, | |
| "grad_norm": 0.8577626347541809, | |
| "learning_rate": 3.5135079539221063e-05, | |
| "loss": 0.3111692428588867, | |
| "step": 95100 | |
| }, | |
| { | |
| "epoch": 1.9505945059450593, | |
| "grad_norm": 0.9102058410644531, | |
| "learning_rate": 3.5100795392210645e-05, | |
| "loss": 0.3128496551513672, | |
| "step": 95150 | |
| }, | |
| { | |
| "epoch": 1.951619516195162, | |
| "grad_norm": 0.8377068042755127, | |
| "learning_rate": 3.5066511245200226e-05, | |
| "loss": 0.30486541748046875, | |
| "step": 95200 | |
| }, | |
| { | |
| "epoch": 1.9526445264452645, | |
| "grad_norm": 0.752038836479187, | |
| "learning_rate": 3.5032227098189794e-05, | |
| "loss": 0.3124969482421875, | |
| "step": 95250 | |
| }, | |
| { | |
| "epoch": 1.953669536695367, | |
| "grad_norm": 0.9308105111122131, | |
| "learning_rate": 3.4997942951179376e-05, | |
| "loss": 0.31303287506103517, | |
| "step": 95300 | |
| }, | |
| { | |
| "epoch": 1.9546945469454693, | |
| "grad_norm": 0.8750619292259216, | |
| "learning_rate": 3.496365880416895e-05, | |
| "loss": 0.30979873657226564, | |
| "step": 95350 | |
| }, | |
| { | |
| "epoch": 1.9557195571955721, | |
| "grad_norm": 0.9156538844108582, | |
| "learning_rate": 3.492937465715853e-05, | |
| "loss": 0.3062991905212402, | |
| "step": 95400 | |
| }, | |
| { | |
| "epoch": 1.9567445674456745, | |
| "grad_norm": 0.9094713926315308, | |
| "learning_rate": 3.489509051014811e-05, | |
| "loss": 0.3141656494140625, | |
| "step": 95450 | |
| }, | |
| { | |
| "epoch": 1.9577695776957769, | |
| "grad_norm": 0.8753242492675781, | |
| "learning_rate": 3.486080636313769e-05, | |
| "loss": 0.3104344177246094, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 1.9587945879458795, | |
| "grad_norm": 0.9141530990600586, | |
| "learning_rate": 3.482652221612727e-05, | |
| "loss": 0.310899543762207, | |
| "step": 95550 | |
| }, | |
| { | |
| "epoch": 1.959819598195982, | |
| "grad_norm": 0.8031003475189209, | |
| "learning_rate": 3.4792238069116844e-05, | |
| "loss": 0.3075739479064941, | |
| "step": 95600 | |
| }, | |
| { | |
| "epoch": 1.9608446084460844, | |
| "grad_norm": 0.8298653960227966, | |
| "learning_rate": 3.475795392210642e-05, | |
| "loss": 0.3037712478637695, | |
| "step": 95650 | |
| }, | |
| { | |
| "epoch": 1.9618696186961868, | |
| "grad_norm": 0.9140118360519409, | |
| "learning_rate": 3.4723669775095994e-05, | |
| "loss": 0.31177892684936526, | |
| "step": 95700 | |
| }, | |
| { | |
| "epoch": 1.9628946289462894, | |
| "grad_norm": 0.8468354344367981, | |
| "learning_rate": 3.4689385628085575e-05, | |
| "loss": 0.3055906295776367, | |
| "step": 95750 | |
| }, | |
| { | |
| "epoch": 1.963919639196392, | |
| "grad_norm": 0.8834452033042908, | |
| "learning_rate": 3.465510148107515e-05, | |
| "loss": 0.31394960403442385, | |
| "step": 95800 | |
| }, | |
| { | |
| "epoch": 1.9649446494464944, | |
| "grad_norm": 0.8386774063110352, | |
| "learning_rate": 3.462081733406473e-05, | |
| "loss": 0.31188426971435546, | |
| "step": 95850 | |
| }, | |
| { | |
| "epoch": 1.965969659696597, | |
| "grad_norm": 0.9671270251274109, | |
| "learning_rate": 3.458653318705431e-05, | |
| "loss": 0.31365394592285156, | |
| "step": 95900 | |
| }, | |
| { | |
| "epoch": 1.9669946699466996, | |
| "grad_norm": 0.920083224773407, | |
| "learning_rate": 3.455224904004389e-05, | |
| "loss": 0.31267255783081055, | |
| "step": 95950 | |
| }, | |
| { | |
| "epoch": 1.968019680196802, | |
| "grad_norm": 0.8485410213470459, | |
| "learning_rate": 3.451796489303346e-05, | |
| "loss": 0.31441535949707033, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 1.9690446904469043, | |
| "grad_norm": 0.9852578639984131, | |
| "learning_rate": 3.448368074602304e-05, | |
| "loss": 0.30556182861328124, | |
| "step": 96050 | |
| }, | |
| { | |
| "epoch": 1.970069700697007, | |
| "grad_norm": 0.8162198066711426, | |
| "learning_rate": 3.444939659901262e-05, | |
| "loss": 0.3070684432983398, | |
| "step": 96100 | |
| }, | |
| { | |
| "epoch": 1.9710947109471095, | |
| "grad_norm": 0.8950273990631104, | |
| "learning_rate": 3.441511245200219e-05, | |
| "loss": 0.30584016799926755, | |
| "step": 96150 | |
| }, | |
| { | |
| "epoch": 1.972119721197212, | |
| "grad_norm": 0.8224257826805115, | |
| "learning_rate": 3.4380828304991775e-05, | |
| "loss": 0.3116300392150879, | |
| "step": 96200 | |
| }, | |
| { | |
| "epoch": 1.9731447314473145, | |
| "grad_norm": 0.8795046210289001, | |
| "learning_rate": 3.4346544157981356e-05, | |
| "loss": 0.3134246063232422, | |
| "step": 96250 | |
| }, | |
| { | |
| "epoch": 1.974169741697417, | |
| "grad_norm": 0.8823623657226562, | |
| "learning_rate": 3.431226001097093e-05, | |
| "loss": 0.3050746726989746, | |
| "step": 96300 | |
| }, | |
| { | |
| "epoch": 1.9751947519475195, | |
| "grad_norm": 0.7794287204742432, | |
| "learning_rate": 3.4277975863960505e-05, | |
| "loss": 0.3022525405883789, | |
| "step": 96350 | |
| }, | |
| { | |
| "epoch": 1.9762197621976219, | |
| "grad_norm": 0.8751199245452881, | |
| "learning_rate": 3.424369171695008e-05, | |
| "loss": 0.3108419990539551, | |
| "step": 96400 | |
| }, | |
| { | |
| "epoch": 1.9772447724477245, | |
| "grad_norm": 0.9428775310516357, | |
| "learning_rate": 3.420940756993966e-05, | |
| "loss": 0.30823818206787107, | |
| "step": 96450 | |
| }, | |
| { | |
| "epoch": 1.978269782697827, | |
| "grad_norm": 0.831886887550354, | |
| "learning_rate": 3.4175123422929236e-05, | |
| "loss": 0.3073104476928711, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 1.9792947929479294, | |
| "grad_norm": 0.855381190776825, | |
| "learning_rate": 3.414083927591882e-05, | |
| "loss": 0.3087195014953613, | |
| "step": 96550 | |
| }, | |
| { | |
| "epoch": 1.980319803198032, | |
| "grad_norm": 0.7549914717674255, | |
| "learning_rate": 3.41065551289084e-05, | |
| "loss": 0.30560577392578125, | |
| "step": 96600 | |
| }, | |
| { | |
| "epoch": 1.9813448134481346, | |
| "grad_norm": 0.8518986105918884, | |
| "learning_rate": 3.4072270981897974e-05, | |
| "loss": 0.3083632850646973, | |
| "step": 96650 | |
| }, | |
| { | |
| "epoch": 1.982369823698237, | |
| "grad_norm": 0.8370254039764404, | |
| "learning_rate": 3.403798683488755e-05, | |
| "loss": 0.3091820526123047, | |
| "step": 96700 | |
| }, | |
| { | |
| "epoch": 1.9833948339483394, | |
| "grad_norm": 0.7821494340896606, | |
| "learning_rate": 3.400370268787712e-05, | |
| "loss": 0.3103960418701172, | |
| "step": 96750 | |
| }, | |
| { | |
| "epoch": 1.984419844198442, | |
| "grad_norm": 0.8332588076591492, | |
| "learning_rate": 3.3969418540866705e-05, | |
| "loss": 0.305260066986084, | |
| "step": 96800 | |
| }, | |
| { | |
| "epoch": 1.9854448544485446, | |
| "grad_norm": 0.7623105645179749, | |
| "learning_rate": 3.393513439385628e-05, | |
| "loss": 0.30901412963867186, | |
| "step": 96850 | |
| }, | |
| { | |
| "epoch": 1.986469864698647, | |
| "grad_norm": 0.8535319566726685, | |
| "learning_rate": 3.390085024684586e-05, | |
| "loss": 0.3020808792114258, | |
| "step": 96900 | |
| }, | |
| { | |
| "epoch": 1.9874948749487493, | |
| "grad_norm": 0.8604022264480591, | |
| "learning_rate": 3.386656609983544e-05, | |
| "loss": 0.3043686866760254, | |
| "step": 96950 | |
| }, | |
| { | |
| "epoch": 1.9885198851988521, | |
| "grad_norm": 1.0008717775344849, | |
| "learning_rate": 3.383228195282502e-05, | |
| "loss": 0.31059438705444337, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 1.9895448954489545, | |
| "grad_norm": 0.8843833804130554, | |
| "learning_rate": 3.379799780581459e-05, | |
| "loss": 0.3072913932800293, | |
| "step": 97050 | |
| }, | |
| { | |
| "epoch": 1.990569905699057, | |
| "grad_norm": 0.8130398392677307, | |
| "learning_rate": 3.3763713658804166e-05, | |
| "loss": 0.3111440277099609, | |
| "step": 97100 | |
| }, | |
| { | |
| "epoch": 1.9915949159491595, | |
| "grad_norm": 0.8369132280349731, | |
| "learning_rate": 3.372942951179375e-05, | |
| "loss": 0.30324390411376956, | |
| "step": 97150 | |
| }, | |
| { | |
| "epoch": 1.992619926199262, | |
| "grad_norm": 0.878365159034729, | |
| "learning_rate": 3.369514536478332e-05, | |
| "loss": 0.3123606109619141, | |
| "step": 97200 | |
| }, | |
| { | |
| "epoch": 1.9936449364493645, | |
| "grad_norm": 1.0073654651641846, | |
| "learning_rate": 3.3660861217772904e-05, | |
| "loss": 0.3085115242004395, | |
| "step": 97250 | |
| }, | |
| { | |
| "epoch": 1.9946699466994668, | |
| "grad_norm": 0.845204770565033, | |
| "learning_rate": 3.3626577070762486e-05, | |
| "loss": 0.3065038299560547, | |
| "step": 97300 | |
| }, | |
| { | |
| "epoch": 1.9956949569495694, | |
| "grad_norm": 0.8621194362640381, | |
| "learning_rate": 3.359229292375206e-05, | |
| "loss": 0.30835060119628904, | |
| "step": 97350 | |
| }, | |
| { | |
| "epoch": 1.996719967199672, | |
| "grad_norm": 0.8676398396492004, | |
| "learning_rate": 3.3558008776741635e-05, | |
| "loss": 0.31082786560058595, | |
| "step": 97400 | |
| }, | |
| { | |
| "epoch": 1.9977449774497744, | |
| "grad_norm": 0.8452362418174744, | |
| "learning_rate": 3.352372462973121e-05, | |
| "loss": 0.31110443115234376, | |
| "step": 97450 | |
| }, | |
| { | |
| "epoch": 1.998769987699877, | |
| "grad_norm": 0.9320103526115417, | |
| "learning_rate": 3.348944048272079e-05, | |
| "loss": 0.30736167907714845, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 1.9997949979499796, | |
| "grad_norm": 0.8884273171424866, | |
| "learning_rate": 3.3455156335710366e-05, | |
| "loss": 0.3039799499511719, | |
| "step": 97550 | |
| }, | |
| { | |
| "epoch": 2.000820008200082, | |
| "grad_norm": 0.9731688499450684, | |
| "learning_rate": 3.342087218869995e-05, | |
| "loss": 0.2994588851928711, | |
| "step": 97600 | |
| }, | |
| { | |
| "epoch": 2.0018450184501844, | |
| "grad_norm": 0.7750595211982727, | |
| "learning_rate": 3.338658804168953e-05, | |
| "loss": 0.2978740882873535, | |
| "step": 97650 | |
| }, | |
| { | |
| "epoch": 2.002870028700287, | |
| "grad_norm": 0.9050639867782593, | |
| "learning_rate": 3.3352303894679104e-05, | |
| "loss": 0.2981520462036133, | |
| "step": 97700 | |
| }, | |
| { | |
| "epoch": 2.0038950389503896, | |
| "grad_norm": 0.913205087184906, | |
| "learning_rate": 3.331801974766868e-05, | |
| "loss": 0.29863437652587893, | |
| "step": 97750 | |
| }, | |
| { | |
| "epoch": 2.004920049200492, | |
| "grad_norm": 0.885442316532135, | |
| "learning_rate": 3.328373560065825e-05, | |
| "loss": 0.3001203727722168, | |
| "step": 97800 | |
| }, | |
| { | |
| "epoch": 2.0059450594505943, | |
| "grad_norm": 0.7808282375335693, | |
| "learning_rate": 3.3249451453647834e-05, | |
| "loss": 0.2957579040527344, | |
| "step": 97850 | |
| }, | |
| { | |
| "epoch": 2.006970069700697, | |
| "grad_norm": 0.7141406536102295, | |
| "learning_rate": 3.3215167306637416e-05, | |
| "loss": 0.29652692794799806, | |
| "step": 97900 | |
| }, | |
| { | |
| "epoch": 2.0079950799507995, | |
| "grad_norm": 0.941078782081604, | |
| "learning_rate": 3.318088315962699e-05, | |
| "loss": 0.29779409408569335, | |
| "step": 97950 | |
| }, | |
| { | |
| "epoch": 2.009020090200902, | |
| "grad_norm": 0.7214632034301758, | |
| "learning_rate": 3.314659901261657e-05, | |
| "loss": 0.2947999572753906, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 2.0100451004510047, | |
| "grad_norm": 0.756691575050354, | |
| "learning_rate": 3.311231486560615e-05, | |
| "loss": 0.29664838790893555, | |
| "step": 98050 | |
| }, | |
| { | |
| "epoch": 2.011070110701107, | |
| "grad_norm": 0.8934739828109741, | |
| "learning_rate": 3.307803071859572e-05, | |
| "loss": 0.29196809768676757, | |
| "step": 98100 | |
| }, | |
| { | |
| "epoch": 2.0120951209512095, | |
| "grad_norm": 0.8826506733894348, | |
| "learning_rate": 3.3043746571585296e-05, | |
| "loss": 0.2988884735107422, | |
| "step": 98150 | |
| }, | |
| { | |
| "epoch": 2.013120131201312, | |
| "grad_norm": 0.751434326171875, | |
| "learning_rate": 3.300946242457488e-05, | |
| "loss": 0.29375429153442384, | |
| "step": 98200 | |
| }, | |
| { | |
| "epoch": 2.0141451414514147, | |
| "grad_norm": 0.8168004751205444, | |
| "learning_rate": 3.297517827756446e-05, | |
| "loss": 0.297548885345459, | |
| "step": 98250 | |
| }, | |
| { | |
| "epoch": 2.015170151701517, | |
| "grad_norm": 0.8839119076728821, | |
| "learning_rate": 3.2940894130554034e-05, | |
| "loss": 0.29352888107299807, | |
| "step": 98300 | |
| }, | |
| { | |
| "epoch": 2.0161951619516194, | |
| "grad_norm": 0.85572749376297, | |
| "learning_rate": 3.2906609983543615e-05, | |
| "loss": 0.29529296875, | |
| "step": 98350 | |
| }, | |
| { | |
| "epoch": 2.0172201722017222, | |
| "grad_norm": 0.8636766672134399, | |
| "learning_rate": 3.287232583653319e-05, | |
| "loss": 0.2968739891052246, | |
| "step": 98400 | |
| }, | |
| { | |
| "epoch": 2.0182451824518246, | |
| "grad_norm": 0.8150589466094971, | |
| "learning_rate": 3.2838041689522765e-05, | |
| "loss": 0.29872766494750974, | |
| "step": 98450 | |
| }, | |
| { | |
| "epoch": 2.019270192701927, | |
| "grad_norm": 0.8949007391929626, | |
| "learning_rate": 3.280375754251234e-05, | |
| "loss": 0.29323535919189453, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 2.0202952029520294, | |
| "grad_norm": 0.860114574432373, | |
| "learning_rate": 3.276947339550192e-05, | |
| "loss": 0.2919896507263184, | |
| "step": 98550 | |
| }, | |
| { | |
| "epoch": 2.021320213202132, | |
| "grad_norm": 0.9874735474586487, | |
| "learning_rate": 3.27351892484915e-05, | |
| "loss": 0.2995409393310547, | |
| "step": 98600 | |
| }, | |
| { | |
| "epoch": 2.0223452234522346, | |
| "grad_norm": 0.930522084236145, | |
| "learning_rate": 3.270090510148108e-05, | |
| "loss": 0.296389102935791, | |
| "step": 98650 | |
| }, | |
| { | |
| "epoch": 2.023370233702337, | |
| "grad_norm": 0.8596113324165344, | |
| "learning_rate": 3.266662095447066e-05, | |
| "loss": 0.30019216537475585, | |
| "step": 98700 | |
| }, | |
| { | |
| "epoch": 2.0243952439524393, | |
| "grad_norm": 0.7825292944908142, | |
| "learning_rate": 3.263233680746023e-05, | |
| "loss": 0.29431861877441406, | |
| "step": 98750 | |
| }, | |
| { | |
| "epoch": 2.025420254202542, | |
| "grad_norm": 0.7561672329902649, | |
| "learning_rate": 3.259805266044981e-05, | |
| "loss": 0.2950853729248047, | |
| "step": 98800 | |
| }, | |
| { | |
| "epoch": 2.0264452644526445, | |
| "grad_norm": 0.8370718955993652, | |
| "learning_rate": 3.256376851343938e-05, | |
| "loss": 0.29708311080932615, | |
| "step": 98850 | |
| }, | |
| { | |
| "epoch": 2.027470274702747, | |
| "grad_norm": 0.8547458052635193, | |
| "learning_rate": 3.2529484366428964e-05, | |
| "loss": 0.2958340835571289, | |
| "step": 98900 | |
| }, | |
| { | |
| "epoch": 2.0284952849528497, | |
| "grad_norm": 0.8593144416809082, | |
| "learning_rate": 3.2495200219418545e-05, | |
| "loss": 0.29616153717041016, | |
| "step": 98950 | |
| }, | |
| { | |
| "epoch": 2.029520295202952, | |
| "grad_norm": 0.877373218536377, | |
| "learning_rate": 3.246091607240812e-05, | |
| "loss": 0.29845537185668947, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 2.0305453054530544, | |
| "grad_norm": 0.8134665489196777, | |
| "learning_rate": 3.24266319253977e-05, | |
| "loss": 0.2981045150756836, | |
| "step": 99050 | |
| }, | |
| { | |
| "epoch": 2.031570315703157, | |
| "grad_norm": 0.8229598999023438, | |
| "learning_rate": 3.2392347778387276e-05, | |
| "loss": 0.30264711380004883, | |
| "step": 99100 | |
| }, | |
| { | |
| "epoch": 2.0325953259532596, | |
| "grad_norm": 0.9266455173492432, | |
| "learning_rate": 3.235806363137685e-05, | |
| "loss": 0.29558837890625, | |
| "step": 99150 | |
| }, | |
| { | |
| "epoch": 2.033620336203362, | |
| "grad_norm": 0.8518116474151611, | |
| "learning_rate": 3.2323779484366426e-05, | |
| "loss": 0.29767307281494143, | |
| "step": 99200 | |
| }, | |
| { | |
| "epoch": 2.0346453464534644, | |
| "grad_norm": 0.9033432006835938, | |
| "learning_rate": 3.228949533735601e-05, | |
| "loss": 0.30157108306884767, | |
| "step": 99250 | |
| }, | |
| { | |
| "epoch": 2.035670356703567, | |
| "grad_norm": 0.8224518299102783, | |
| "learning_rate": 3.225521119034559e-05, | |
| "loss": 0.3006077194213867, | |
| "step": 99300 | |
| }, | |
| { | |
| "epoch": 2.0366953669536696, | |
| "grad_norm": 0.912680983543396, | |
| "learning_rate": 3.222092704333516e-05, | |
| "loss": 0.2950175666809082, | |
| "step": 99350 | |
| }, | |
| { | |
| "epoch": 2.037720377203772, | |
| "grad_norm": 0.9501999020576477, | |
| "learning_rate": 3.2186642896324745e-05, | |
| "loss": 0.2949297332763672, | |
| "step": 99400 | |
| }, | |
| { | |
| "epoch": 2.0387453874538743, | |
| "grad_norm": 0.8039262294769287, | |
| "learning_rate": 3.215235874931432e-05, | |
| "loss": 0.2945002365112305, | |
| "step": 99450 | |
| }, | |
| { | |
| "epoch": 2.039770397703977, | |
| "grad_norm": 0.9155386090278625, | |
| "learning_rate": 3.2118074602303894e-05, | |
| "loss": 0.30047800064086916, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 2.0407954079540795, | |
| "grad_norm": 0.8987196683883667, | |
| "learning_rate": 3.208379045529347e-05, | |
| "loss": 0.2962318992614746, | |
| "step": 99550 | |
| }, | |
| { | |
| "epoch": 2.041820418204182, | |
| "grad_norm": 0.9211351871490479, | |
| "learning_rate": 3.204950630828305e-05, | |
| "loss": 0.2993268775939941, | |
| "step": 99600 | |
| }, | |
| { | |
| "epoch": 2.0428454284542847, | |
| "grad_norm": 0.877439022064209, | |
| "learning_rate": 3.201522216127263e-05, | |
| "loss": 0.301232795715332, | |
| "step": 99650 | |
| }, | |
| { | |
| "epoch": 2.043870438704387, | |
| "grad_norm": 0.8397917151451111, | |
| "learning_rate": 3.1980938014262207e-05, | |
| "loss": 0.2943710708618164, | |
| "step": 99700 | |
| }, | |
| { | |
| "epoch": 2.0448954489544895, | |
| "grad_norm": 0.8376466035842896, | |
| "learning_rate": 3.194665386725179e-05, | |
| "loss": 0.29772487640380857, | |
| "step": 99750 | |
| }, | |
| { | |
| "epoch": 2.045920459204592, | |
| "grad_norm": 0.9279513359069824, | |
| "learning_rate": 3.191236972024136e-05, | |
| "loss": 0.30280778884887694, | |
| "step": 99800 | |
| }, | |
| { | |
| "epoch": 2.0469454694546947, | |
| "grad_norm": 0.9428771734237671, | |
| "learning_rate": 3.187808557323094e-05, | |
| "loss": 0.29872325897216795, | |
| "step": 99850 | |
| }, | |
| { | |
| "epoch": 2.047970479704797, | |
| "grad_norm": 0.8505126237869263, | |
| "learning_rate": 3.184380142622051e-05, | |
| "loss": 0.2999704933166504, | |
| "step": 99900 | |
| }, | |
| { | |
| "epoch": 2.0489954899548994, | |
| "grad_norm": 0.974773108959198, | |
| "learning_rate": 3.1809517279210094e-05, | |
| "loss": 0.2945591163635254, | |
| "step": 99950 | |
| }, | |
| { | |
| "epoch": 2.0500205002050023, | |
| "grad_norm": 0.8723935484886169, | |
| "learning_rate": 3.1775233132199675e-05, | |
| "loss": 0.2935123062133789, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 2.0500205002050023, | |
| "eval_loss": 0.36836329102516174, | |
| "eval_runtime": 4.5435, | |
| "eval_samples_per_second": 5502.351, | |
| "eval_steps_per_second": 5.502, | |
| "eval_teacher_cosine_loss": 0.23333603143692017, | |
| "eval_teacher_cosine_mean": 0.7666639685630798, | |
| "eval_teacher_cosine_p10": 0.6911764740943909, | |
| "eval_teacher_cosine_p50": 0.7766528129577637, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 2.0510455104551046, | |
| "grad_norm": 0.9227980971336365, | |
| "learning_rate": 3.174094898518925e-05, | |
| "loss": 0.2997748374938965, | |
| "step": 100050 | |
| }, | |
| { | |
| "epoch": 2.052070520705207, | |
| "grad_norm": 0.8654810190200806, | |
| "learning_rate": 3.170666483817883e-05, | |
| "loss": 0.2998298454284668, | |
| "step": 100100 | |
| }, | |
| { | |
| "epoch": 2.0530955309553094, | |
| "grad_norm": 0.7885472774505615, | |
| "learning_rate": 3.1672380691168406e-05, | |
| "loss": 0.30011375427246095, | |
| "step": 100150 | |
| }, | |
| { | |
| "epoch": 2.054120541205412, | |
| "grad_norm": 0.8101473450660706, | |
| "learning_rate": 3.163809654415798e-05, | |
| "loss": 0.2917229461669922, | |
| "step": 100200 | |
| }, | |
| { | |
| "epoch": 2.0551455514555146, | |
| "grad_norm": 0.8751871585845947, | |
| "learning_rate": 3.1603812397147555e-05, | |
| "loss": 0.295882511138916, | |
| "step": 100250 | |
| }, | |
| { | |
| "epoch": 2.056170561705617, | |
| "grad_norm": 0.7858721613883972, | |
| "learning_rate": 3.156952825013714e-05, | |
| "loss": 0.29738924026489255, | |
| "step": 100300 | |
| }, | |
| { | |
| "epoch": 2.0571955719557193, | |
| "grad_norm": 0.86143958568573, | |
| "learning_rate": 3.153524410312672e-05, | |
| "loss": 0.2994497108459473, | |
| "step": 100350 | |
| }, | |
| { | |
| "epoch": 2.058220582205822, | |
| "grad_norm": 0.8117907643318176, | |
| "learning_rate": 3.150095995611629e-05, | |
| "loss": 0.29323495864868165, | |
| "step": 100400 | |
| }, | |
| { | |
| "epoch": 2.0592455924559245, | |
| "grad_norm": 0.8647395968437195, | |
| "learning_rate": 3.1466675809105874e-05, | |
| "loss": 0.30399858474731445, | |
| "step": 100450 | |
| }, | |
| { | |
| "epoch": 2.060270602706027, | |
| "grad_norm": 0.8773109316825867, | |
| "learning_rate": 3.143239166209545e-05, | |
| "loss": 0.29932346343994143, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 2.0612956129561297, | |
| "grad_norm": 0.8890672326087952, | |
| "learning_rate": 3.1398107515085024e-05, | |
| "loss": 0.297420654296875, | |
| "step": 100550 | |
| }, | |
| { | |
| "epoch": 2.062320623206232, | |
| "grad_norm": 0.8729727864265442, | |
| "learning_rate": 3.13638233680746e-05, | |
| "loss": 0.2948910140991211, | |
| "step": 100600 | |
| }, | |
| { | |
| "epoch": 2.0633456334563345, | |
| "grad_norm": 0.8633873462677002, | |
| "learning_rate": 3.132953922106418e-05, | |
| "loss": 0.3047377395629883, | |
| "step": 100650 | |
| }, | |
| { | |
| "epoch": 2.064370643706437, | |
| "grad_norm": 0.930671215057373, | |
| "learning_rate": 3.129525507405376e-05, | |
| "loss": 0.29417608261108397, | |
| "step": 100700 | |
| }, | |
| { | |
| "epoch": 2.0653956539565397, | |
| "grad_norm": 0.8550683259963989, | |
| "learning_rate": 3.1260970927043336e-05, | |
| "loss": 0.3000674057006836, | |
| "step": 100750 | |
| }, | |
| { | |
| "epoch": 2.066420664206642, | |
| "grad_norm": 0.8169131875038147, | |
| "learning_rate": 3.122668678003292e-05, | |
| "loss": 0.29356937408447265, | |
| "step": 100800 | |
| }, | |
| { | |
| "epoch": 2.0674456744567444, | |
| "grad_norm": 0.8119945526123047, | |
| "learning_rate": 3.119240263302249e-05, | |
| "loss": 0.3004438781738281, | |
| "step": 100850 | |
| }, | |
| { | |
| "epoch": 2.0684706847068473, | |
| "grad_norm": 0.9278579950332642, | |
| "learning_rate": 3.1158118486012074e-05, | |
| "loss": 0.3005023956298828, | |
| "step": 100900 | |
| }, | |
| { | |
| "epoch": 2.0694956949569496, | |
| "grad_norm": 0.9286501407623291, | |
| "learning_rate": 3.112383433900164e-05, | |
| "loss": 0.3030779457092285, | |
| "step": 100950 | |
| }, | |
| { | |
| "epoch": 2.070520705207052, | |
| "grad_norm": 0.8241747617721558, | |
| "learning_rate": 3.108955019199122e-05, | |
| "loss": 0.2943849563598633, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 2.0715457154571544, | |
| "grad_norm": 0.7966886758804321, | |
| "learning_rate": 3.1055266044980805e-05, | |
| "loss": 0.2953671646118164, | |
| "step": 101050 | |
| }, | |
| { | |
| "epoch": 2.072570725707257, | |
| "grad_norm": 0.8759568929672241, | |
| "learning_rate": 3.102098189797038e-05, | |
| "loss": 0.29533891677856444, | |
| "step": 101100 | |
| }, | |
| { | |
| "epoch": 2.0735957359573596, | |
| "grad_norm": 0.8626474738121033, | |
| "learning_rate": 3.098669775095996e-05, | |
| "loss": 0.29581178665161134, | |
| "step": 101150 | |
| }, | |
| { | |
| "epoch": 2.074620746207462, | |
| "grad_norm": 0.8856799602508545, | |
| "learning_rate": 3.0952413603949536e-05, | |
| "loss": 0.29646963119506836, | |
| "step": 101200 | |
| }, | |
| { | |
| "epoch": 2.0756457564575648, | |
| "grad_norm": 0.9650260210037231, | |
| "learning_rate": 3.091812945693912e-05, | |
| "loss": 0.29811605453491213, | |
| "step": 101250 | |
| }, | |
| { | |
| "epoch": 2.076670766707667, | |
| "grad_norm": 0.8552830219268799, | |
| "learning_rate": 3.088384530992869e-05, | |
| "loss": 0.2955424118041992, | |
| "step": 101300 | |
| }, | |
| { | |
| "epoch": 2.0776957769577695, | |
| "grad_norm": 0.8032104969024658, | |
| "learning_rate": 3.0849561162918266e-05, | |
| "loss": 0.2985131072998047, | |
| "step": 101350 | |
| }, | |
| { | |
| "epoch": 2.078720787207872, | |
| "grad_norm": 0.8130046129226685, | |
| "learning_rate": 3.081527701590785e-05, | |
| "loss": 0.29903520584106447, | |
| "step": 101400 | |
| }, | |
| { | |
| "epoch": 2.0797457974579747, | |
| "grad_norm": 0.8150535821914673, | |
| "learning_rate": 3.078099286889742e-05, | |
| "loss": 0.2959913635253906, | |
| "step": 101450 | |
| }, | |
| { | |
| "epoch": 2.080770807708077, | |
| "grad_norm": 0.9675926566123962, | |
| "learning_rate": 3.0746708721887004e-05, | |
| "loss": 0.30132427215576174, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 2.0817958179581795, | |
| "grad_norm": 0.9190462827682495, | |
| "learning_rate": 3.071242457487658e-05, | |
| "loss": 0.2941686058044434, | |
| "step": 101550 | |
| }, | |
| { | |
| "epoch": 2.0828208282082823, | |
| "grad_norm": 0.8580654859542847, | |
| "learning_rate": 3.067814042786616e-05, | |
| "loss": 0.2972437858581543, | |
| "step": 101600 | |
| }, | |
| { | |
| "epoch": 2.0838458384583847, | |
| "grad_norm": 0.8958724737167358, | |
| "learning_rate": 3.0643856280855735e-05, | |
| "loss": 0.29701982498168944, | |
| "step": 101650 | |
| }, | |
| { | |
| "epoch": 2.084870848708487, | |
| "grad_norm": 0.8746614456176758, | |
| "learning_rate": 3.060957213384531e-05, | |
| "loss": 0.30209733963012697, | |
| "step": 101700 | |
| }, | |
| { | |
| "epoch": 2.0858958589585894, | |
| "grad_norm": 0.8729932308197021, | |
| "learning_rate": 3.057528798683489e-05, | |
| "loss": 0.2958634185791016, | |
| "step": 101750 | |
| }, | |
| { | |
| "epoch": 2.0869208692086922, | |
| "grad_norm": 0.8924493193626404, | |
| "learning_rate": 3.0541003839824466e-05, | |
| "loss": 0.296065673828125, | |
| "step": 101800 | |
| }, | |
| { | |
| "epoch": 2.0879458794587946, | |
| "grad_norm": 0.844535231590271, | |
| "learning_rate": 3.0506719692814044e-05, | |
| "loss": 0.3003213882446289, | |
| "step": 101850 | |
| }, | |
| { | |
| "epoch": 2.088970889708897, | |
| "grad_norm": 0.8254181742668152, | |
| "learning_rate": 3.0472435545803625e-05, | |
| "loss": 0.2965537071228027, | |
| "step": 101900 | |
| }, | |
| { | |
| "epoch": 2.089995899959, | |
| "grad_norm": 0.9020160436630249, | |
| "learning_rate": 3.0438151398793203e-05, | |
| "loss": 0.29859384536743167, | |
| "step": 101950 | |
| }, | |
| { | |
| "epoch": 2.091020910209102, | |
| "grad_norm": 0.7991710305213928, | |
| "learning_rate": 3.0403867251782775e-05, | |
| "loss": 0.29725839614868166, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 2.0920459204592046, | |
| "grad_norm": 0.9045502543449402, | |
| "learning_rate": 3.0369583104772353e-05, | |
| "loss": 0.3019831085205078, | |
| "step": 102050 | |
| }, | |
| { | |
| "epoch": 2.093070930709307, | |
| "grad_norm": 0.9006917476654053, | |
| "learning_rate": 3.033529895776193e-05, | |
| "loss": 0.2990254783630371, | |
| "step": 102100 | |
| }, | |
| { | |
| "epoch": 2.0940959409594098, | |
| "grad_norm": 0.8556710481643677, | |
| "learning_rate": 3.030101481075151e-05, | |
| "loss": 0.2969277000427246, | |
| "step": 102150 | |
| }, | |
| { | |
| "epoch": 2.095120951209512, | |
| "grad_norm": 0.8513312935829163, | |
| "learning_rate": 3.0266730663741087e-05, | |
| "loss": 0.2990144348144531, | |
| "step": 102200 | |
| }, | |
| { | |
| "epoch": 2.0961459614596145, | |
| "grad_norm": 0.808119535446167, | |
| "learning_rate": 3.023244651673067e-05, | |
| "loss": 0.2949010276794434, | |
| "step": 102250 | |
| }, | |
| { | |
| "epoch": 2.097170971709717, | |
| "grad_norm": 0.9329097270965576, | |
| "learning_rate": 3.0198162369720247e-05, | |
| "loss": 0.29863386154174804, | |
| "step": 102300 | |
| }, | |
| { | |
| "epoch": 2.0981959819598197, | |
| "grad_norm": 0.8151674866676331, | |
| "learning_rate": 3.0163878222709818e-05, | |
| "loss": 0.29982542037963866, | |
| "step": 102350 | |
| }, | |
| { | |
| "epoch": 2.099220992209922, | |
| "grad_norm": 0.9259713888168335, | |
| "learning_rate": 3.0129594075699396e-05, | |
| "loss": 0.2935900115966797, | |
| "step": 102400 | |
| }, | |
| { | |
| "epoch": 2.1002460024600245, | |
| "grad_norm": 0.8324610590934753, | |
| "learning_rate": 3.0095309928688974e-05, | |
| "loss": 0.2974943923950195, | |
| "step": 102450 | |
| }, | |
| { | |
| "epoch": 2.1012710127101273, | |
| "grad_norm": 0.9000000953674316, | |
| "learning_rate": 3.0061025781678552e-05, | |
| "loss": 0.2966069030761719, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 2.1022960229602297, | |
| "grad_norm": 0.8474639654159546, | |
| "learning_rate": 3.002674163466813e-05, | |
| "loss": 0.3018353080749512, | |
| "step": 102550 | |
| }, | |
| { | |
| "epoch": 2.103321033210332, | |
| "grad_norm": 0.9050176739692688, | |
| "learning_rate": 2.9992457487657712e-05, | |
| "loss": 0.2975497817993164, | |
| "step": 102600 | |
| }, | |
| { | |
| "epoch": 2.1043460434604344, | |
| "grad_norm": 0.8110204339027405, | |
| "learning_rate": 2.995817334064729e-05, | |
| "loss": 0.29933670043945315, | |
| "step": 102650 | |
| }, | |
| { | |
| "epoch": 2.1053710537105372, | |
| "grad_norm": 0.9333590269088745, | |
| "learning_rate": 2.992388919363686e-05, | |
| "loss": 0.2930521774291992, | |
| "step": 102700 | |
| }, | |
| { | |
| "epoch": 2.1063960639606396, | |
| "grad_norm": 0.8833275437355042, | |
| "learning_rate": 2.988960504662644e-05, | |
| "loss": 0.29713825225830076, | |
| "step": 102750 | |
| }, | |
| { | |
| "epoch": 2.107421074210742, | |
| "grad_norm": 0.8093939423561096, | |
| "learning_rate": 2.9855320899616017e-05, | |
| "loss": 0.2971360206604004, | |
| "step": 102800 | |
| }, | |
| { | |
| "epoch": 2.108446084460845, | |
| "grad_norm": 0.8521679639816284, | |
| "learning_rate": 2.9821036752605595e-05, | |
| "loss": 0.296489372253418, | |
| "step": 102850 | |
| }, | |
| { | |
| "epoch": 2.109471094710947, | |
| "grad_norm": 0.9115025997161865, | |
| "learning_rate": 2.9786752605595174e-05, | |
| "loss": 0.30106369018554685, | |
| "step": 102900 | |
| }, | |
| { | |
| "epoch": 2.1104961049610496, | |
| "grad_norm": 0.8913432955741882, | |
| "learning_rate": 2.9752468458584755e-05, | |
| "loss": 0.3071391487121582, | |
| "step": 102950 | |
| }, | |
| { | |
| "epoch": 2.111521115211152, | |
| "grad_norm": 0.8022323250770569, | |
| "learning_rate": 2.9718184311574333e-05, | |
| "loss": 0.29422821044921876, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 2.1125461254612548, | |
| "grad_norm": 0.852857232093811, | |
| "learning_rate": 2.9683900164563904e-05, | |
| "loss": 0.29639497756958005, | |
| "step": 103050 | |
| }, | |
| { | |
| "epoch": 2.113571135711357, | |
| "grad_norm": 0.9254240393638611, | |
| "learning_rate": 2.9649616017553482e-05, | |
| "loss": 0.2950699234008789, | |
| "step": 103100 | |
| }, | |
| { | |
| "epoch": 2.1145961459614595, | |
| "grad_norm": 0.8702476620674133, | |
| "learning_rate": 2.961533187054306e-05, | |
| "loss": 0.29631526947021486, | |
| "step": 103150 | |
| }, | |
| { | |
| "epoch": 2.1156211562115623, | |
| "grad_norm": 0.8514348268508911, | |
| "learning_rate": 2.958104772353264e-05, | |
| "loss": 0.29250268936157225, | |
| "step": 103200 | |
| }, | |
| { | |
| "epoch": 2.1166461664616647, | |
| "grad_norm": 0.8276752233505249, | |
| "learning_rate": 2.9546763576522217e-05, | |
| "loss": 0.29325483322143553, | |
| "step": 103250 | |
| }, | |
| { | |
| "epoch": 2.117671176711767, | |
| "grad_norm": 0.9394628405570984, | |
| "learning_rate": 2.9512479429511798e-05, | |
| "loss": 0.29581371307373044, | |
| "step": 103300 | |
| }, | |
| { | |
| "epoch": 2.1186961869618695, | |
| "grad_norm": 0.7935223579406738, | |
| "learning_rate": 2.9478195282501376e-05, | |
| "loss": 0.29940160751342776, | |
| "step": 103350 | |
| }, | |
| { | |
| "epoch": 2.1197211972119723, | |
| "grad_norm": 0.8723791837692261, | |
| "learning_rate": 2.9443911135490948e-05, | |
| "loss": 0.2967808151245117, | |
| "step": 103400 | |
| }, | |
| { | |
| "epoch": 2.1207462074620747, | |
| "grad_norm": 0.8441775441169739, | |
| "learning_rate": 2.9409626988480526e-05, | |
| "loss": 0.29500282287597657, | |
| "step": 103450 | |
| }, | |
| { | |
| "epoch": 2.121771217712177, | |
| "grad_norm": 0.9139922857284546, | |
| "learning_rate": 2.9375342841470104e-05, | |
| "loss": 0.29827117919921875, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 2.1227962279622794, | |
| "grad_norm": 0.8601665496826172, | |
| "learning_rate": 2.9341058694459682e-05, | |
| "loss": 0.29784025192260744, | |
| "step": 103550 | |
| }, | |
| { | |
| "epoch": 2.1238212382123822, | |
| "grad_norm": 0.8663135170936584, | |
| "learning_rate": 2.9306774547449263e-05, | |
| "loss": 0.29910276412963865, | |
| "step": 103600 | |
| }, | |
| { | |
| "epoch": 2.1248462484624846, | |
| "grad_norm": 0.8638542890548706, | |
| "learning_rate": 2.927249040043884e-05, | |
| "loss": 0.2969480895996094, | |
| "step": 103650 | |
| }, | |
| { | |
| "epoch": 2.125871258712587, | |
| "grad_norm": 0.7919973134994507, | |
| "learning_rate": 2.923820625342842e-05, | |
| "loss": 0.2963202285766602, | |
| "step": 103700 | |
| }, | |
| { | |
| "epoch": 2.12689626896269, | |
| "grad_norm": 0.8650715947151184, | |
| "learning_rate": 2.920392210641799e-05, | |
| "loss": 0.29715681076049805, | |
| "step": 103750 | |
| }, | |
| { | |
| "epoch": 2.127921279212792, | |
| "grad_norm": 0.9250510334968567, | |
| "learning_rate": 2.916963795940757e-05, | |
| "loss": 0.2962819290161133, | |
| "step": 103800 | |
| }, | |
| { | |
| "epoch": 2.1289462894628945, | |
| "grad_norm": 0.8462968468666077, | |
| "learning_rate": 2.9135353812397147e-05, | |
| "loss": 0.2974299049377441, | |
| "step": 103850 | |
| }, | |
| { | |
| "epoch": 2.129971299712997, | |
| "grad_norm": 0.8234403133392334, | |
| "learning_rate": 2.9101069665386725e-05, | |
| "loss": 0.29649465560913085, | |
| "step": 103900 | |
| }, | |
| { | |
| "epoch": 2.1309963099630997, | |
| "grad_norm": 0.7610822319984436, | |
| "learning_rate": 2.9066785518376307e-05, | |
| "loss": 0.29418701171875, | |
| "step": 103950 | |
| }, | |
| { | |
| "epoch": 2.132021320213202, | |
| "grad_norm": 0.9581247568130493, | |
| "learning_rate": 2.9032501371365885e-05, | |
| "loss": 0.30057228088378907, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 2.1330463304633045, | |
| "grad_norm": 0.8769311308860779, | |
| "learning_rate": 2.8998217224355463e-05, | |
| "loss": 0.29846960067749023, | |
| "step": 104050 | |
| }, | |
| { | |
| "epoch": 2.1340713407134073, | |
| "grad_norm": 0.953406572341919, | |
| "learning_rate": 2.8963933077345034e-05, | |
| "loss": 0.29461614608764647, | |
| "step": 104100 | |
| }, | |
| { | |
| "epoch": 2.1350963509635097, | |
| "grad_norm": 0.9159896373748779, | |
| "learning_rate": 2.8929648930334612e-05, | |
| "loss": 0.29942970275878905, | |
| "step": 104150 | |
| }, | |
| { | |
| "epoch": 2.136121361213612, | |
| "grad_norm": 0.8949552774429321, | |
| "learning_rate": 2.889536478332419e-05, | |
| "loss": 0.2938087463378906, | |
| "step": 104200 | |
| }, | |
| { | |
| "epoch": 2.1371463714637144, | |
| "grad_norm": 0.7950694561004639, | |
| "learning_rate": 2.8861080636313768e-05, | |
| "loss": 0.29575170516967775, | |
| "step": 104250 | |
| }, | |
| { | |
| "epoch": 2.1381713817138173, | |
| "grad_norm": 0.857952892780304, | |
| "learning_rate": 2.882679648930335e-05, | |
| "loss": 0.293023853302002, | |
| "step": 104300 | |
| }, | |
| { | |
| "epoch": 2.1391963919639196, | |
| "grad_norm": 0.9615322947502136, | |
| "learning_rate": 2.8792512342292928e-05, | |
| "loss": 0.29327873229980467, | |
| "step": 104350 | |
| }, | |
| { | |
| "epoch": 2.140221402214022, | |
| "grad_norm": 0.9281526803970337, | |
| "learning_rate": 2.8758228195282506e-05, | |
| "loss": 0.29798423767089843, | |
| "step": 104400 | |
| }, | |
| { | |
| "epoch": 2.141246412464125, | |
| "grad_norm": 0.969214677810669, | |
| "learning_rate": 2.8723944048272077e-05, | |
| "loss": 0.2944455528259277, | |
| "step": 104450 | |
| }, | |
| { | |
| "epoch": 2.142271422714227, | |
| "grad_norm": 0.7796619534492493, | |
| "learning_rate": 2.8689659901261655e-05, | |
| "loss": 0.29821718215942383, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 2.1432964329643296, | |
| "grad_norm": 0.8747615814208984, | |
| "learning_rate": 2.8655375754251233e-05, | |
| "loss": 0.29270177841186523, | |
| "step": 104550 | |
| }, | |
| { | |
| "epoch": 2.144321443214432, | |
| "grad_norm": 0.8056252002716064, | |
| "learning_rate": 2.862109160724081e-05, | |
| "loss": 0.2975976181030273, | |
| "step": 104600 | |
| }, | |
| { | |
| "epoch": 2.145346453464535, | |
| "grad_norm": 0.9199638366699219, | |
| "learning_rate": 2.8586807460230393e-05, | |
| "loss": 0.29958267211914064, | |
| "step": 104650 | |
| }, | |
| { | |
| "epoch": 2.146371463714637, | |
| "grad_norm": 0.7882837057113647, | |
| "learning_rate": 2.855252331321997e-05, | |
| "loss": 0.29708974838256835, | |
| "step": 104700 | |
| }, | |
| { | |
| "epoch": 2.1473964739647395, | |
| "grad_norm": 0.8830605149269104, | |
| "learning_rate": 2.851823916620955e-05, | |
| "loss": 0.2960023880004883, | |
| "step": 104750 | |
| }, | |
| { | |
| "epoch": 2.1484214842148424, | |
| "grad_norm": 0.8034010529518127, | |
| "learning_rate": 2.848395501919912e-05, | |
| "loss": 0.2947233200073242, | |
| "step": 104800 | |
| }, | |
| { | |
| "epoch": 2.1494464944649447, | |
| "grad_norm": 0.8090293407440186, | |
| "learning_rate": 2.84496708721887e-05, | |
| "loss": 0.29248819351196287, | |
| "step": 104850 | |
| }, | |
| { | |
| "epoch": 2.150471504715047, | |
| "grad_norm": 0.8797180652618408, | |
| "learning_rate": 2.8415386725178277e-05, | |
| "loss": 0.2915445327758789, | |
| "step": 104900 | |
| }, | |
| { | |
| "epoch": 2.1514965149651495, | |
| "grad_norm": 0.8586041331291199, | |
| "learning_rate": 2.8381102578167855e-05, | |
| "loss": 0.3007703971862793, | |
| "step": 104950 | |
| }, | |
| { | |
| "epoch": 2.1525215252152523, | |
| "grad_norm": 0.9227334856987, | |
| "learning_rate": 2.8346818431157436e-05, | |
| "loss": 0.29264835357666014, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 2.1525215252152523, | |
| "eval_loss": 0.36609575152397156, | |
| "eval_runtime": 4.5453, | |
| "eval_samples_per_second": 5500.15, | |
| "eval_steps_per_second": 5.5, | |
| "eval_teacher_cosine_loss": 0.23152732849121094, | |
| "eval_teacher_cosine_mean": 0.7684726715087891, | |
| "eval_teacher_cosine_p10": 0.6927378177642822, | |
| "eval_teacher_cosine_p50": 0.7785059213638306, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 2.1535465354653547, | |
| "grad_norm": 0.8654130101203918, | |
| "learning_rate": 2.8312534284147014e-05, | |
| "loss": 0.2975589179992676, | |
| "step": 105050 | |
| }, | |
| { | |
| "epoch": 2.154571545715457, | |
| "grad_norm": 0.8066343069076538, | |
| "learning_rate": 2.8278250137136592e-05, | |
| "loss": 0.29931987762451173, | |
| "step": 105100 | |
| }, | |
| { | |
| "epoch": 2.15559655596556, | |
| "grad_norm": 0.9070913791656494, | |
| "learning_rate": 2.8243965990126164e-05, | |
| "loss": 0.29504390716552736, | |
| "step": 105150 | |
| }, | |
| { | |
| "epoch": 2.1566215662156623, | |
| "grad_norm": 0.8948898911476135, | |
| "learning_rate": 2.8209681843115742e-05, | |
| "loss": 0.29416522979736326, | |
| "step": 105200 | |
| }, | |
| { | |
| "epoch": 2.1576465764657646, | |
| "grad_norm": 0.8565952181816101, | |
| "learning_rate": 2.817539769610532e-05, | |
| "loss": 0.293405818939209, | |
| "step": 105250 | |
| }, | |
| { | |
| "epoch": 2.158671586715867, | |
| "grad_norm": 0.8996642827987671, | |
| "learning_rate": 2.81411135490949e-05, | |
| "loss": 0.2932826232910156, | |
| "step": 105300 | |
| }, | |
| { | |
| "epoch": 2.15969659696597, | |
| "grad_norm": 0.9096159338951111, | |
| "learning_rate": 2.810682940208448e-05, | |
| "loss": 0.294307861328125, | |
| "step": 105350 | |
| }, | |
| { | |
| "epoch": 2.160721607216072, | |
| "grad_norm": 0.9305816888809204, | |
| "learning_rate": 2.8072545255074057e-05, | |
| "loss": 0.2943918228149414, | |
| "step": 105400 | |
| }, | |
| { | |
| "epoch": 2.1617466174661746, | |
| "grad_norm": 0.9332500696182251, | |
| "learning_rate": 2.8038261108063636e-05, | |
| "loss": 0.2917665672302246, | |
| "step": 105450 | |
| }, | |
| { | |
| "epoch": 2.162771627716277, | |
| "grad_norm": 0.9485348463058472, | |
| "learning_rate": 2.8003976961053207e-05, | |
| "loss": 0.2978641128540039, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 2.1637966379663798, | |
| "grad_norm": 0.7976488471031189, | |
| "learning_rate": 2.7969692814042785e-05, | |
| "loss": 0.29764236450195314, | |
| "step": 105550 | |
| }, | |
| { | |
| "epoch": 2.164821648216482, | |
| "grad_norm": 0.8732777833938599, | |
| "learning_rate": 2.7935408667032363e-05, | |
| "loss": 0.2987928199768066, | |
| "step": 105600 | |
| }, | |
| { | |
| "epoch": 2.1658466584665845, | |
| "grad_norm": 0.8364903330802917, | |
| "learning_rate": 2.7901124520021944e-05, | |
| "loss": 0.2965454292297363, | |
| "step": 105650 | |
| }, | |
| { | |
| "epoch": 2.1668716687166873, | |
| "grad_norm": 0.971548318862915, | |
| "learning_rate": 2.7866840373011523e-05, | |
| "loss": 0.29494882583618165, | |
| "step": 105700 | |
| }, | |
| { | |
| "epoch": 2.1678966789667897, | |
| "grad_norm": 0.9323691725730896, | |
| "learning_rate": 2.78325562260011e-05, | |
| "loss": 0.29035879135131837, | |
| "step": 105750 | |
| }, | |
| { | |
| "epoch": 2.168921689216892, | |
| "grad_norm": 0.9512751698493958, | |
| "learning_rate": 2.779827207899068e-05, | |
| "loss": 0.29584331512451173, | |
| "step": 105800 | |
| }, | |
| { | |
| "epoch": 2.1699466994669945, | |
| "grad_norm": 0.7855687141418457, | |
| "learning_rate": 2.776398793198025e-05, | |
| "loss": 0.2937512969970703, | |
| "step": 105850 | |
| }, | |
| { | |
| "epoch": 2.1709717097170973, | |
| "grad_norm": 0.8793435096740723, | |
| "learning_rate": 2.7729703784969828e-05, | |
| "loss": 0.2979957389831543, | |
| "step": 105900 | |
| }, | |
| { | |
| "epoch": 2.1719967199671997, | |
| "grad_norm": 0.8838885426521301, | |
| "learning_rate": 2.7695419637959406e-05, | |
| "loss": 0.2974848747253418, | |
| "step": 105950 | |
| }, | |
| { | |
| "epoch": 2.173021730217302, | |
| "grad_norm": 0.8843762874603271, | |
| "learning_rate": 2.7661135490948988e-05, | |
| "loss": 0.2933164596557617, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 2.174046740467405, | |
| "grad_norm": 0.9306745529174805, | |
| "learning_rate": 2.7626851343938566e-05, | |
| "loss": 0.29968494415283203, | |
| "step": 106050 | |
| }, | |
| { | |
| "epoch": 2.1750717507175072, | |
| "grad_norm": 0.9780585169792175, | |
| "learning_rate": 2.7592567196928144e-05, | |
| "loss": 0.29551042556762697, | |
| "step": 106100 | |
| }, | |
| { | |
| "epoch": 2.1760967609676096, | |
| "grad_norm": 0.8371595740318298, | |
| "learning_rate": 2.7558283049917722e-05, | |
| "loss": 0.2956245040893555, | |
| "step": 106150 | |
| }, | |
| { | |
| "epoch": 2.177121771217712, | |
| "grad_norm": 0.800273060798645, | |
| "learning_rate": 2.7523998902907293e-05, | |
| "loss": 0.29739866256713865, | |
| "step": 106200 | |
| }, | |
| { | |
| "epoch": 2.178146781467815, | |
| "grad_norm": 0.9268495440483093, | |
| "learning_rate": 2.748971475589687e-05, | |
| "loss": 0.2946196746826172, | |
| "step": 106250 | |
| }, | |
| { | |
| "epoch": 2.179171791717917, | |
| "grad_norm": 0.8785091042518616, | |
| "learning_rate": 2.745543060888645e-05, | |
| "loss": 0.2939936637878418, | |
| "step": 106300 | |
| }, | |
| { | |
| "epoch": 2.1801968019680196, | |
| "grad_norm": 0.8797817826271057, | |
| "learning_rate": 2.742114646187603e-05, | |
| "loss": 0.30013439178466794, | |
| "step": 106350 | |
| }, | |
| { | |
| "epoch": 2.181221812218122, | |
| "grad_norm": 0.9020611047744751, | |
| "learning_rate": 2.738686231486561e-05, | |
| "loss": 0.29708511352539063, | |
| "step": 106400 | |
| }, | |
| { | |
| "epoch": 2.1822468224682248, | |
| "grad_norm": 0.8190785050392151, | |
| "learning_rate": 2.7352578167855187e-05, | |
| "loss": 0.2952339363098144, | |
| "step": 106450 | |
| }, | |
| { | |
| "epoch": 2.183271832718327, | |
| "grad_norm": 0.8604353070259094, | |
| "learning_rate": 2.7318294020844765e-05, | |
| "loss": 0.29489984512329104, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 2.1842968429684295, | |
| "grad_norm": 0.8518834114074707, | |
| "learning_rate": 2.7284009873834343e-05, | |
| "loss": 0.2984670257568359, | |
| "step": 106550 | |
| }, | |
| { | |
| "epoch": 2.1853218532185323, | |
| "grad_norm": 0.8734896183013916, | |
| "learning_rate": 2.7249725726823915e-05, | |
| "loss": 0.2972888565063477, | |
| "step": 106600 | |
| }, | |
| { | |
| "epoch": 2.1863468634686347, | |
| "grad_norm": 0.869484007358551, | |
| "learning_rate": 2.7215441579813493e-05, | |
| "loss": 0.2954715728759766, | |
| "step": 106650 | |
| }, | |
| { | |
| "epoch": 2.187371873718737, | |
| "grad_norm": 0.8709007501602173, | |
| "learning_rate": 2.7181157432803074e-05, | |
| "loss": 0.2952753257751465, | |
| "step": 106700 | |
| }, | |
| { | |
| "epoch": 2.1883968839688395, | |
| "grad_norm": 0.9275204539299011, | |
| "learning_rate": 2.7146873285792652e-05, | |
| "loss": 0.29618587493896487, | |
| "step": 106750 | |
| }, | |
| { | |
| "epoch": 2.1894218942189423, | |
| "grad_norm": 0.8196974992752075, | |
| "learning_rate": 2.711258913878223e-05, | |
| "loss": 0.2953327560424805, | |
| "step": 106800 | |
| }, | |
| { | |
| "epoch": 2.1904469044690447, | |
| "grad_norm": 0.7997073531150818, | |
| "learning_rate": 2.707830499177181e-05, | |
| "loss": 0.296455078125, | |
| "step": 106850 | |
| }, | |
| { | |
| "epoch": 2.191471914719147, | |
| "grad_norm": 0.9379832744598389, | |
| "learning_rate": 2.7044020844761386e-05, | |
| "loss": 0.29215272903442385, | |
| "step": 106900 | |
| }, | |
| { | |
| "epoch": 2.19249692496925, | |
| "grad_norm": 0.9899487495422363, | |
| "learning_rate": 2.7009736697750958e-05, | |
| "loss": 0.29423337936401367, | |
| "step": 106950 | |
| }, | |
| { | |
| "epoch": 2.1935219352193522, | |
| "grad_norm": 0.9138458967208862, | |
| "learning_rate": 2.697545255074054e-05, | |
| "loss": 0.29839458465576174, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 2.1945469454694546, | |
| "grad_norm": 0.8896054029464722, | |
| "learning_rate": 2.6941168403730117e-05, | |
| "loss": 0.2940182113647461, | |
| "step": 107050 | |
| }, | |
| { | |
| "epoch": 2.195571955719557, | |
| "grad_norm": 0.8583347797393799, | |
| "learning_rate": 2.6906884256719695e-05, | |
| "loss": 0.29643489837646486, | |
| "step": 107100 | |
| }, | |
| { | |
| "epoch": 2.19659696596966, | |
| "grad_norm": 0.838973343372345, | |
| "learning_rate": 2.6872600109709273e-05, | |
| "loss": 0.2932909393310547, | |
| "step": 107150 | |
| }, | |
| { | |
| "epoch": 2.197621976219762, | |
| "grad_norm": 0.9327914118766785, | |
| "learning_rate": 2.683831596269885e-05, | |
| "loss": 0.2920139503479004, | |
| "step": 107200 | |
| }, | |
| { | |
| "epoch": 2.1986469864698646, | |
| "grad_norm": 0.8599159121513367, | |
| "learning_rate": 2.680403181568843e-05, | |
| "loss": 0.29998361587524414, | |
| "step": 107250 | |
| }, | |
| { | |
| "epoch": 2.1996719967199674, | |
| "grad_norm": 0.9132016897201538, | |
| "learning_rate": 2.6769747668678e-05, | |
| "loss": 0.29563203811645505, | |
| "step": 107300 | |
| }, | |
| { | |
| "epoch": 2.2006970069700698, | |
| "grad_norm": 0.8162052035331726, | |
| "learning_rate": 2.6735463521667582e-05, | |
| "loss": 0.3004001998901367, | |
| "step": 107350 | |
| }, | |
| { | |
| "epoch": 2.201722017220172, | |
| "grad_norm": 0.8499268293380737, | |
| "learning_rate": 2.670117937465716e-05, | |
| "loss": 0.29571243286132814, | |
| "step": 107400 | |
| }, | |
| { | |
| "epoch": 2.2027470274702745, | |
| "grad_norm": 0.8278290629386902, | |
| "learning_rate": 2.666689522764674e-05, | |
| "loss": 0.29378049850463867, | |
| "step": 107450 | |
| }, | |
| { | |
| "epoch": 2.2037720377203773, | |
| "grad_norm": 0.9030365347862244, | |
| "learning_rate": 2.6632611080636317e-05, | |
| "loss": 0.2982151794433594, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 2.2047970479704797, | |
| "grad_norm": 0.8522258996963501, | |
| "learning_rate": 2.6598326933625895e-05, | |
| "loss": 0.3015831756591797, | |
| "step": 107550 | |
| }, | |
| { | |
| "epoch": 2.205822058220582, | |
| "grad_norm": 0.8651204705238342, | |
| "learning_rate": 2.6564042786615473e-05, | |
| "loss": 0.2943475914001465, | |
| "step": 107600 | |
| }, | |
| { | |
| "epoch": 2.206847068470685, | |
| "grad_norm": 0.9146361351013184, | |
| "learning_rate": 2.6529758639605044e-05, | |
| "loss": 0.2942600059509277, | |
| "step": 107650 | |
| }, | |
| { | |
| "epoch": 2.2078720787207873, | |
| "grad_norm": 0.9256983995437622, | |
| "learning_rate": 2.6495474492594626e-05, | |
| "loss": 0.2942539405822754, | |
| "step": 107700 | |
| }, | |
| { | |
| "epoch": 2.2088970889708897, | |
| "grad_norm": 0.9270093441009521, | |
| "learning_rate": 2.6461190345584204e-05, | |
| "loss": 0.2998019981384277, | |
| "step": 107750 | |
| }, | |
| { | |
| "epoch": 2.209922099220992, | |
| "grad_norm": 0.81485515832901, | |
| "learning_rate": 2.6426906198573782e-05, | |
| "loss": 0.29529743194580077, | |
| "step": 107800 | |
| }, | |
| { | |
| "epoch": 2.210947109471095, | |
| "grad_norm": 0.8073440790176392, | |
| "learning_rate": 2.639262205156336e-05, | |
| "loss": 0.2978074645996094, | |
| "step": 107850 | |
| }, | |
| { | |
| "epoch": 2.2119721197211972, | |
| "grad_norm": 0.8900967836380005, | |
| "learning_rate": 2.6358337904552938e-05, | |
| "loss": 0.29368783950805666, | |
| "step": 107900 | |
| }, | |
| { | |
| "epoch": 2.2129971299712996, | |
| "grad_norm": 0.917927622795105, | |
| "learning_rate": 2.6324053757542516e-05, | |
| "loss": 0.2981483650207519, | |
| "step": 107950 | |
| }, | |
| { | |
| "epoch": 2.2140221402214024, | |
| "grad_norm": 0.9126695990562439, | |
| "learning_rate": 2.6289769610532087e-05, | |
| "loss": 0.2960077476501465, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 2.215047150471505, | |
| "grad_norm": 0.9732317924499512, | |
| "learning_rate": 2.625548546352167e-05, | |
| "loss": 0.2927455139160156, | |
| "step": 108050 | |
| }, | |
| { | |
| "epoch": 2.216072160721607, | |
| "grad_norm": 0.8393977284431458, | |
| "learning_rate": 2.6221201316511247e-05, | |
| "loss": 0.29182962417602537, | |
| "step": 108100 | |
| }, | |
| { | |
| "epoch": 2.2170971709717096, | |
| "grad_norm": 0.8862981200218201, | |
| "learning_rate": 2.6186917169500825e-05, | |
| "loss": 0.29085391998291016, | |
| "step": 108150 | |
| }, | |
| { | |
| "epoch": 2.2181221812218124, | |
| "grad_norm": 0.8738946318626404, | |
| "learning_rate": 2.6152633022490403e-05, | |
| "loss": 0.30129114151000974, | |
| "step": 108200 | |
| }, | |
| { | |
| "epoch": 2.2191471914719147, | |
| "grad_norm": 0.9143874645233154, | |
| "learning_rate": 2.611834887547998e-05, | |
| "loss": 0.29225454330444334, | |
| "step": 108250 | |
| }, | |
| { | |
| "epoch": 2.220172201722017, | |
| "grad_norm": 0.8038158416748047, | |
| "learning_rate": 2.608406472846956e-05, | |
| "loss": 0.29569541931152343, | |
| "step": 108300 | |
| }, | |
| { | |
| "epoch": 2.22119721197212, | |
| "grad_norm": 0.8750113248825073, | |
| "learning_rate": 2.604978058145913e-05, | |
| "loss": 0.29836496353149417, | |
| "step": 108350 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.8441888093948364, | |
| "learning_rate": 2.6015496434448712e-05, | |
| "loss": 0.2913064002990723, | |
| "step": 108400 | |
| }, | |
| { | |
| "epoch": 2.2232472324723247, | |
| "grad_norm": 0.9294244647026062, | |
| "learning_rate": 2.598121228743829e-05, | |
| "loss": 0.2969842529296875, | |
| "step": 108450 | |
| }, | |
| { | |
| "epoch": 2.224272242722427, | |
| "grad_norm": 0.810257077217102, | |
| "learning_rate": 2.5946928140427868e-05, | |
| "loss": 0.29390581130981447, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 2.22529725297253, | |
| "grad_norm": 0.8865947127342224, | |
| "learning_rate": 2.5912643993417446e-05, | |
| "loss": 0.2921183776855469, | |
| "step": 108550 | |
| }, | |
| { | |
| "epoch": 2.2263222632226323, | |
| "grad_norm": 0.866049587726593, | |
| "learning_rate": 2.5878359846407024e-05, | |
| "loss": 0.2926931381225586, | |
| "step": 108600 | |
| }, | |
| { | |
| "epoch": 2.2273472734727346, | |
| "grad_norm": 0.7930173873901367, | |
| "learning_rate": 2.5844075699396602e-05, | |
| "loss": 0.3009513282775879, | |
| "step": 108650 | |
| }, | |
| { | |
| "epoch": 2.228372283722837, | |
| "grad_norm": 0.881267786026001, | |
| "learning_rate": 2.5809791552386177e-05, | |
| "loss": 0.2992196273803711, | |
| "step": 108700 | |
| }, | |
| { | |
| "epoch": 2.22939729397294, | |
| "grad_norm": 0.8533804416656494, | |
| "learning_rate": 2.5775507405375755e-05, | |
| "loss": 0.29853626251220705, | |
| "step": 108750 | |
| }, | |
| { | |
| "epoch": 2.230422304223042, | |
| "grad_norm": 0.8839023113250732, | |
| "learning_rate": 2.5741223258365333e-05, | |
| "loss": 0.29716306686401367, | |
| "step": 108800 | |
| }, | |
| { | |
| "epoch": 2.2314473144731446, | |
| "grad_norm": 0.9902378916740417, | |
| "learning_rate": 2.570693911135491e-05, | |
| "loss": 0.30288604736328123, | |
| "step": 108850 | |
| }, | |
| { | |
| "epoch": 2.2324723247232474, | |
| "grad_norm": 0.8992528319358826, | |
| "learning_rate": 2.567265496434449e-05, | |
| "loss": 0.297005672454834, | |
| "step": 108900 | |
| }, | |
| { | |
| "epoch": 2.23349733497335, | |
| "grad_norm": 0.8139254450798035, | |
| "learning_rate": 2.5638370817334068e-05, | |
| "loss": 0.28948963165283204, | |
| "step": 108950 | |
| }, | |
| { | |
| "epoch": 2.234522345223452, | |
| "grad_norm": 0.8563607335090637, | |
| "learning_rate": 2.5604086670323646e-05, | |
| "loss": 0.29211824417114257, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 2.2355473554735545, | |
| "grad_norm": 0.9101682901382446, | |
| "learning_rate": 2.556980252331322e-05, | |
| "loss": 0.2971038246154785, | |
| "step": 109050 | |
| }, | |
| { | |
| "epoch": 2.2365723657236574, | |
| "grad_norm": 0.9029500484466553, | |
| "learning_rate": 2.55355183763028e-05, | |
| "loss": 0.29678089141845704, | |
| "step": 109100 | |
| }, | |
| { | |
| "epoch": 2.2375973759737597, | |
| "grad_norm": 0.9404675960540771, | |
| "learning_rate": 2.5501234229292377e-05, | |
| "loss": 0.2894249725341797, | |
| "step": 109150 | |
| }, | |
| { | |
| "epoch": 2.238622386223862, | |
| "grad_norm": 0.855964720249176, | |
| "learning_rate": 2.5466950082281955e-05, | |
| "loss": 0.2954045677185059, | |
| "step": 109200 | |
| }, | |
| { | |
| "epoch": 2.239647396473965, | |
| "grad_norm": 0.9263198971748352, | |
| "learning_rate": 2.5432665935271533e-05, | |
| "loss": 0.29440643310546877, | |
| "step": 109250 | |
| }, | |
| { | |
| "epoch": 2.2406724067240673, | |
| "grad_norm": 0.8884220123291016, | |
| "learning_rate": 2.539838178826111e-05, | |
| "loss": 0.29876121520996096, | |
| "step": 109300 | |
| }, | |
| { | |
| "epoch": 2.2416974169741697, | |
| "grad_norm": 0.9323673844337463, | |
| "learning_rate": 2.536409764125069e-05, | |
| "loss": 0.30023441314697263, | |
| "step": 109350 | |
| }, | |
| { | |
| "epoch": 2.242722427224272, | |
| "grad_norm": 0.8877379894256592, | |
| "learning_rate": 2.5329813494240264e-05, | |
| "loss": 0.2989007568359375, | |
| "step": 109400 | |
| }, | |
| { | |
| "epoch": 2.243747437474375, | |
| "grad_norm": 0.8307398557662964, | |
| "learning_rate": 2.5295529347229842e-05, | |
| "loss": 0.2949715232849121, | |
| "step": 109450 | |
| }, | |
| { | |
| "epoch": 2.2447724477244773, | |
| "grad_norm": 0.8021062016487122, | |
| "learning_rate": 2.526124520021942e-05, | |
| "loss": 0.29329662322998046, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 2.2457974579745796, | |
| "grad_norm": 0.913710355758667, | |
| "learning_rate": 2.5226961053208998e-05, | |
| "loss": 0.2986569595336914, | |
| "step": 109550 | |
| }, | |
| { | |
| "epoch": 2.246822468224682, | |
| "grad_norm": 0.8524134159088135, | |
| "learning_rate": 2.5192676906198576e-05, | |
| "loss": 0.2954893112182617, | |
| "step": 109600 | |
| }, | |
| { | |
| "epoch": 2.247847478474785, | |
| "grad_norm": 0.9402430653572083, | |
| "learning_rate": 2.5158392759188154e-05, | |
| "loss": 0.29814781188964845, | |
| "step": 109650 | |
| }, | |
| { | |
| "epoch": 2.248872488724887, | |
| "grad_norm": 0.8851281404495239, | |
| "learning_rate": 2.5124108612177732e-05, | |
| "loss": 0.2899352264404297, | |
| "step": 109700 | |
| }, | |
| { | |
| "epoch": 2.2498974989749896, | |
| "grad_norm": 0.9330788254737854, | |
| "learning_rate": 2.5089824465167307e-05, | |
| "loss": 0.30306114196777345, | |
| "step": 109750 | |
| }, | |
| { | |
| "epoch": 2.2509225092250924, | |
| "grad_norm": 0.838013231754303, | |
| "learning_rate": 2.5055540318156885e-05, | |
| "loss": 0.29630992889404295, | |
| "step": 109800 | |
| }, | |
| { | |
| "epoch": 2.251947519475195, | |
| "grad_norm": 0.8707270622253418, | |
| "learning_rate": 2.5021256171146463e-05, | |
| "loss": 0.29819602966308595, | |
| "step": 109850 | |
| }, | |
| { | |
| "epoch": 2.252972529725297, | |
| "grad_norm": 0.8493938446044922, | |
| "learning_rate": 2.498697202413604e-05, | |
| "loss": 0.29686717987060546, | |
| "step": 109900 | |
| }, | |
| { | |
| "epoch": 2.2539975399753995, | |
| "grad_norm": 0.9356685280799866, | |
| "learning_rate": 2.495268787712562e-05, | |
| "loss": 0.30005903244018556, | |
| "step": 109950 | |
| }, | |
| { | |
| "epoch": 2.2550225502255024, | |
| "grad_norm": 0.9322360157966614, | |
| "learning_rate": 2.4918403730115194e-05, | |
| "loss": 0.2951397132873535, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 2.2550225502255024, | |
| "eval_loss": 0.363637238740921, | |
| "eval_runtime": 4.7375, | |
| "eval_samples_per_second": 5277.043, | |
| "eval_steps_per_second": 5.277, | |
| "eval_teacher_cosine_loss": 0.2290428876876831, | |
| "eval_teacher_cosine_mean": 0.7709571123123169, | |
| "eval_teacher_cosine_p10": 0.6954448819160461, | |
| "eval_teacher_cosine_p50": 0.7811345458030701, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 2.2560475604756047, | |
| "grad_norm": 0.8784653544425964, | |
| "learning_rate": 2.4884119583104772e-05, | |
| "loss": 0.29851919174194336, | |
| "step": 110050 | |
| }, | |
| { | |
| "epoch": 2.257072570725707, | |
| "grad_norm": 0.8443608283996582, | |
| "learning_rate": 2.484983543609435e-05, | |
| "loss": 0.2934509468078613, | |
| "step": 110100 | |
| }, | |
| { | |
| "epoch": 2.25809758097581, | |
| "grad_norm": 0.8912509679794312, | |
| "learning_rate": 2.4815551289083928e-05, | |
| "loss": 0.2991896629333496, | |
| "step": 110150 | |
| }, | |
| { | |
| "epoch": 2.2591225912259123, | |
| "grad_norm": 0.8745754361152649, | |
| "learning_rate": 2.4781267142073506e-05, | |
| "loss": 0.2874060249328613, | |
| "step": 110200 | |
| }, | |
| { | |
| "epoch": 2.2601476014760147, | |
| "grad_norm": 0.8356772065162659, | |
| "learning_rate": 2.4746982995063084e-05, | |
| "loss": 0.2966134834289551, | |
| "step": 110250 | |
| }, | |
| { | |
| "epoch": 2.261172611726117, | |
| "grad_norm": 0.8158166408538818, | |
| "learning_rate": 2.4712698848052662e-05, | |
| "loss": 0.29480520248413083, | |
| "step": 110300 | |
| }, | |
| { | |
| "epoch": 2.26219762197622, | |
| "grad_norm": 0.8606014251708984, | |
| "learning_rate": 2.467841470104224e-05, | |
| "loss": 0.29517900466918945, | |
| "step": 110350 | |
| }, | |
| { | |
| "epoch": 2.2632226322263223, | |
| "grad_norm": 0.852946400642395, | |
| "learning_rate": 2.4644130554031815e-05, | |
| "loss": 0.30126453399658204, | |
| "step": 110400 | |
| }, | |
| { | |
| "epoch": 2.2642476424764246, | |
| "grad_norm": 0.8945220708847046, | |
| "learning_rate": 2.4609846407021393e-05, | |
| "loss": 0.2993128204345703, | |
| "step": 110450 | |
| }, | |
| { | |
| "epoch": 2.2652726527265274, | |
| "grad_norm": 0.915103018283844, | |
| "learning_rate": 2.4575562260010975e-05, | |
| "loss": 0.29677162170410154, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 2.26629766297663, | |
| "grad_norm": 0.8710775971412659, | |
| "learning_rate": 2.454127811300055e-05, | |
| "loss": 0.2949982261657715, | |
| "step": 110550 | |
| }, | |
| { | |
| "epoch": 2.267322673226732, | |
| "grad_norm": 0.9300380349159241, | |
| "learning_rate": 2.4506993965990127e-05, | |
| "loss": 0.2874764633178711, | |
| "step": 110600 | |
| }, | |
| { | |
| "epoch": 2.2683476834768346, | |
| "grad_norm": 0.8918837308883667, | |
| "learning_rate": 2.4472709818979706e-05, | |
| "loss": 0.292415771484375, | |
| "step": 110650 | |
| }, | |
| { | |
| "epoch": 2.2693726937269374, | |
| "grad_norm": 0.9860808849334717, | |
| "learning_rate": 2.4438425671969284e-05, | |
| "loss": 0.29298479080200196, | |
| "step": 110700 | |
| }, | |
| { | |
| "epoch": 2.2703977039770398, | |
| "grad_norm": 0.9560282826423645, | |
| "learning_rate": 2.440414152495886e-05, | |
| "loss": 0.2969422149658203, | |
| "step": 110750 | |
| }, | |
| { | |
| "epoch": 2.271422714227142, | |
| "grad_norm": 0.9083956480026245, | |
| "learning_rate": 2.4369857377948436e-05, | |
| "loss": 0.29253263473510743, | |
| "step": 110800 | |
| }, | |
| { | |
| "epoch": 2.272447724477245, | |
| "grad_norm": 0.9328821301460266, | |
| "learning_rate": 2.4335573230938018e-05, | |
| "loss": 0.3002961349487305, | |
| "step": 110850 | |
| }, | |
| { | |
| "epoch": 2.2734727347273473, | |
| "grad_norm": 0.8655358552932739, | |
| "learning_rate": 2.4301289083927593e-05, | |
| "loss": 0.29180707931518557, | |
| "step": 110900 | |
| }, | |
| { | |
| "epoch": 2.2744977449774497, | |
| "grad_norm": 0.8965252041816711, | |
| "learning_rate": 2.426700493691717e-05, | |
| "loss": 0.2947185516357422, | |
| "step": 110950 | |
| }, | |
| { | |
| "epoch": 2.275522755227552, | |
| "grad_norm": 0.7509618401527405, | |
| "learning_rate": 2.423272078990675e-05, | |
| "loss": 0.29259157180786133, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 2.276547765477655, | |
| "grad_norm": 0.9174707531929016, | |
| "learning_rate": 2.4198436642896327e-05, | |
| "loss": 0.29327775955200197, | |
| "step": 111050 | |
| }, | |
| { | |
| "epoch": 2.2775727757277573, | |
| "grad_norm": 0.9395931959152222, | |
| "learning_rate": 2.41641524958859e-05, | |
| "loss": 0.2999496841430664, | |
| "step": 111100 | |
| }, | |
| { | |
| "epoch": 2.2785977859778597, | |
| "grad_norm": 0.8677461743354797, | |
| "learning_rate": 2.412986834887548e-05, | |
| "loss": 0.2965157699584961, | |
| "step": 111150 | |
| }, | |
| { | |
| "epoch": 2.2796227962279625, | |
| "grad_norm": 0.8833332657814026, | |
| "learning_rate": 2.409558420186506e-05, | |
| "loss": 0.29086498260498045, | |
| "step": 111200 | |
| }, | |
| { | |
| "epoch": 2.280647806478065, | |
| "grad_norm": 0.9139562845230103, | |
| "learning_rate": 2.4061300054854636e-05, | |
| "loss": 0.29327739715576173, | |
| "step": 111250 | |
| }, | |
| { | |
| "epoch": 2.2816728167281672, | |
| "grad_norm": 0.9009184837341309, | |
| "learning_rate": 2.4027015907844214e-05, | |
| "loss": 0.2898719596862793, | |
| "step": 111300 | |
| }, | |
| { | |
| "epoch": 2.2826978269782696, | |
| "grad_norm": 0.9005762338638306, | |
| "learning_rate": 2.3992731760833792e-05, | |
| "loss": 0.29697771072387696, | |
| "step": 111350 | |
| }, | |
| { | |
| "epoch": 2.2837228372283724, | |
| "grad_norm": 0.9099572896957397, | |
| "learning_rate": 2.395844761382337e-05, | |
| "loss": 0.2919172859191895, | |
| "step": 111400 | |
| }, | |
| { | |
| "epoch": 2.284747847478475, | |
| "grad_norm": 0.9277542233467102, | |
| "learning_rate": 2.3924163466812945e-05, | |
| "loss": 0.2959120750427246, | |
| "step": 111450 | |
| }, | |
| { | |
| "epoch": 2.285772857728577, | |
| "grad_norm": 0.9506464004516602, | |
| "learning_rate": 2.3889879319802523e-05, | |
| "loss": 0.28755401611328124, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 2.28679786797868, | |
| "grad_norm": 0.7757502794265747, | |
| "learning_rate": 2.3855595172792104e-05, | |
| "loss": 0.2941597366333008, | |
| "step": 111550 | |
| }, | |
| { | |
| "epoch": 2.2878228782287824, | |
| "grad_norm": 0.826040506362915, | |
| "learning_rate": 2.382131102578168e-05, | |
| "loss": 0.2937843894958496, | |
| "step": 111600 | |
| }, | |
| { | |
| "epoch": 2.2888478884788848, | |
| "grad_norm": 1.0090352296829224, | |
| "learning_rate": 2.3787026878771257e-05, | |
| "loss": 0.2939146041870117, | |
| "step": 111650 | |
| }, | |
| { | |
| "epoch": 2.289872898728987, | |
| "grad_norm": 0.9168022871017456, | |
| "learning_rate": 2.3752742731760835e-05, | |
| "loss": 0.3014184761047363, | |
| "step": 111700 | |
| }, | |
| { | |
| "epoch": 2.29089790897909, | |
| "grad_norm": 0.8913780450820923, | |
| "learning_rate": 2.3718458584750413e-05, | |
| "loss": 0.29110225677490237, | |
| "step": 111750 | |
| }, | |
| { | |
| "epoch": 2.2919229192291923, | |
| "grad_norm": 0.8325012922286987, | |
| "learning_rate": 2.3684174437739988e-05, | |
| "loss": 0.29531763076782225, | |
| "step": 111800 | |
| }, | |
| { | |
| "epoch": 2.2929479294792947, | |
| "grad_norm": 0.9322215914726257, | |
| "learning_rate": 2.3649890290729566e-05, | |
| "loss": 0.29935842514038086, | |
| "step": 111850 | |
| }, | |
| { | |
| "epoch": 2.2939729397293975, | |
| "grad_norm": 0.9170472025871277, | |
| "learning_rate": 2.3615606143719148e-05, | |
| "loss": 0.29206357955932616, | |
| "step": 111900 | |
| }, | |
| { | |
| "epoch": 2.2949979499795, | |
| "grad_norm": 0.9267598986625671, | |
| "learning_rate": 2.3581321996708722e-05, | |
| "loss": 0.29269065856933596, | |
| "step": 111950 | |
| }, | |
| { | |
| "epoch": 2.2960229602296023, | |
| "grad_norm": 0.8630905747413635, | |
| "learning_rate": 2.35470378496983e-05, | |
| "loss": 0.2904055213928223, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 2.2970479704797047, | |
| "grad_norm": 0.8597891330718994, | |
| "learning_rate": 2.351275370268788e-05, | |
| "loss": 0.2949575614929199, | |
| "step": 112050 | |
| }, | |
| { | |
| "epoch": 2.2980729807298075, | |
| "grad_norm": 0.8262439370155334, | |
| "learning_rate": 2.3478469555677456e-05, | |
| "loss": 0.293453311920166, | |
| "step": 112100 | |
| }, | |
| { | |
| "epoch": 2.29909799097991, | |
| "grad_norm": 0.8436186909675598, | |
| "learning_rate": 2.344418540866703e-05, | |
| "loss": 0.2979419136047363, | |
| "step": 112150 | |
| }, | |
| { | |
| "epoch": 2.3001230012300122, | |
| "grad_norm": 0.8903818130493164, | |
| "learning_rate": 2.3409901261656613e-05, | |
| "loss": 0.29177375793457033, | |
| "step": 112200 | |
| }, | |
| { | |
| "epoch": 2.3011480114801146, | |
| "grad_norm": 0.874401330947876, | |
| "learning_rate": 2.337561711464619e-05, | |
| "loss": 0.2937587547302246, | |
| "step": 112250 | |
| }, | |
| { | |
| "epoch": 2.3021730217302174, | |
| "grad_norm": 0.9006122350692749, | |
| "learning_rate": 2.3341332967635765e-05, | |
| "loss": 0.2969514274597168, | |
| "step": 112300 | |
| }, | |
| { | |
| "epoch": 2.30319803198032, | |
| "grad_norm": 0.9065960049629211, | |
| "learning_rate": 2.3307048820625344e-05, | |
| "loss": 0.29403152465820315, | |
| "step": 112350 | |
| }, | |
| { | |
| "epoch": 2.304223042230422, | |
| "grad_norm": 0.8452473282814026, | |
| "learning_rate": 2.327276467361492e-05, | |
| "loss": 0.2929822540283203, | |
| "step": 112400 | |
| }, | |
| { | |
| "epoch": 2.3052480524805246, | |
| "grad_norm": 0.9072423577308655, | |
| "learning_rate": 2.32384805266045e-05, | |
| "loss": 0.2971394729614258, | |
| "step": 112450 | |
| }, | |
| { | |
| "epoch": 2.3062730627306274, | |
| "grad_norm": 0.8025612831115723, | |
| "learning_rate": 2.3204196379594074e-05, | |
| "loss": 0.29441213607788086, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 2.3072980729807298, | |
| "grad_norm": 0.9351097345352173, | |
| "learning_rate": 2.3169912232583656e-05, | |
| "loss": 0.29533016204833984, | |
| "step": 112550 | |
| }, | |
| { | |
| "epoch": 2.308323083230832, | |
| "grad_norm": 0.8632520437240601, | |
| "learning_rate": 2.3135628085573234e-05, | |
| "loss": 0.29122209548950195, | |
| "step": 112600 | |
| }, | |
| { | |
| "epoch": 2.309348093480935, | |
| "grad_norm": 0.8932254314422607, | |
| "learning_rate": 2.310134393856281e-05, | |
| "loss": 0.300871467590332, | |
| "step": 112650 | |
| }, | |
| { | |
| "epoch": 2.3103731037310373, | |
| "grad_norm": 0.9460945725440979, | |
| "learning_rate": 2.3067059791552387e-05, | |
| "loss": 0.29631719589233396, | |
| "step": 112700 | |
| }, | |
| { | |
| "epoch": 2.3113981139811397, | |
| "grad_norm": 0.8980906009674072, | |
| "learning_rate": 2.3032775644541965e-05, | |
| "loss": 0.2943337059020996, | |
| "step": 112750 | |
| }, | |
| { | |
| "epoch": 2.312423124231242, | |
| "grad_norm": 0.8904237151145935, | |
| "learning_rate": 2.2998491497531543e-05, | |
| "loss": 0.2887349891662598, | |
| "step": 112800 | |
| }, | |
| { | |
| "epoch": 2.313448134481345, | |
| "grad_norm": 0.8973768949508667, | |
| "learning_rate": 2.2964207350521118e-05, | |
| "loss": 0.2956345748901367, | |
| "step": 112850 | |
| }, | |
| { | |
| "epoch": 2.3144731447314473, | |
| "grad_norm": 0.8929799199104309, | |
| "learning_rate": 2.29299232035107e-05, | |
| "loss": 0.29816999435424807, | |
| "step": 112900 | |
| }, | |
| { | |
| "epoch": 2.3154981549815496, | |
| "grad_norm": 0.8004156351089478, | |
| "learning_rate": 2.2895639056500277e-05, | |
| "loss": 0.29101139068603515, | |
| "step": 112950 | |
| }, | |
| { | |
| "epoch": 2.3165231652316525, | |
| "grad_norm": 0.8118010759353638, | |
| "learning_rate": 2.2861354909489855e-05, | |
| "loss": 0.2907235145568848, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 2.317548175481755, | |
| "grad_norm": 0.9495766162872314, | |
| "learning_rate": 2.282707076247943e-05, | |
| "loss": 0.29859151840209963, | |
| "step": 113050 | |
| }, | |
| { | |
| "epoch": 2.3185731857318572, | |
| "grad_norm": 0.8808241486549377, | |
| "learning_rate": 2.2792786615469008e-05, | |
| "loss": 0.29405879974365234, | |
| "step": 113100 | |
| }, | |
| { | |
| "epoch": 2.3195981959819596, | |
| "grad_norm": 0.8774096369743347, | |
| "learning_rate": 2.2758502468458586e-05, | |
| "loss": 0.29400638580322264, | |
| "step": 113150 | |
| }, | |
| { | |
| "epoch": 2.3206232062320624, | |
| "grad_norm": 1.0070782899856567, | |
| "learning_rate": 2.272421832144816e-05, | |
| "loss": 0.29808822631835935, | |
| "step": 113200 | |
| }, | |
| { | |
| "epoch": 2.321648216482165, | |
| "grad_norm": 0.84535151720047, | |
| "learning_rate": 2.2689934174437742e-05, | |
| "loss": 0.2944942092895508, | |
| "step": 113250 | |
| }, | |
| { | |
| "epoch": 2.322673226732267, | |
| "grad_norm": 0.7699443101882935, | |
| "learning_rate": 2.265565002742732e-05, | |
| "loss": 0.2915530014038086, | |
| "step": 113300 | |
| }, | |
| { | |
| "epoch": 2.32369823698237, | |
| "grad_norm": 0.8253020644187927, | |
| "learning_rate": 2.26213658804169e-05, | |
| "loss": 0.2933192825317383, | |
| "step": 113350 | |
| }, | |
| { | |
| "epoch": 2.3247232472324724, | |
| "grad_norm": 0.793982982635498, | |
| "learning_rate": 2.2587081733406473e-05, | |
| "loss": 0.2940970802307129, | |
| "step": 113400 | |
| }, | |
| { | |
| "epoch": 2.3257482574825747, | |
| "grad_norm": 0.9602380394935608, | |
| "learning_rate": 2.255279758639605e-05, | |
| "loss": 0.2923820114135742, | |
| "step": 113450 | |
| }, | |
| { | |
| "epoch": 2.326773267732677, | |
| "grad_norm": 0.8266251087188721, | |
| "learning_rate": 2.251851343938563e-05, | |
| "loss": 0.29248043060302736, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 2.32779827798278, | |
| "grad_norm": 0.854598879814148, | |
| "learning_rate": 2.2484229292375204e-05, | |
| "loss": 0.28931671142578125, | |
| "step": 113550 | |
| }, | |
| { | |
| "epoch": 2.3288232882328823, | |
| "grad_norm": 0.8739018440246582, | |
| "learning_rate": 2.2449945145364785e-05, | |
| "loss": 0.2929952049255371, | |
| "step": 113600 | |
| }, | |
| { | |
| "epoch": 2.3298482984829847, | |
| "grad_norm": 0.8802686333656311, | |
| "learning_rate": 2.2415660998354364e-05, | |
| "loss": 0.294363899230957, | |
| "step": 113650 | |
| }, | |
| { | |
| "epoch": 2.3308733087330875, | |
| "grad_norm": 0.8650929927825928, | |
| "learning_rate": 2.238137685134394e-05, | |
| "loss": 0.2925260353088379, | |
| "step": 113700 | |
| }, | |
| { | |
| "epoch": 2.33189831898319, | |
| "grad_norm": 0.8947234153747559, | |
| "learning_rate": 2.2347092704333516e-05, | |
| "loss": 0.2961161994934082, | |
| "step": 113750 | |
| }, | |
| { | |
| "epoch": 2.3329233292332923, | |
| "grad_norm": 0.990831732749939, | |
| "learning_rate": 2.2312808557323094e-05, | |
| "loss": 0.29639720916748047, | |
| "step": 113800 | |
| }, | |
| { | |
| "epoch": 2.3339483394833946, | |
| "grad_norm": 0.9146309494972229, | |
| "learning_rate": 2.2278524410312673e-05, | |
| "loss": 0.29371110916137694, | |
| "step": 113850 | |
| }, | |
| { | |
| "epoch": 2.3349733497334975, | |
| "grad_norm": 0.927509605884552, | |
| "learning_rate": 2.224424026330225e-05, | |
| "loss": 0.2914425849914551, | |
| "step": 113900 | |
| }, | |
| { | |
| "epoch": 2.3359983599836, | |
| "grad_norm": 0.8845198750495911, | |
| "learning_rate": 2.220995611629183e-05, | |
| "loss": 0.28975683212280273, | |
| "step": 113950 | |
| }, | |
| { | |
| "epoch": 2.337023370233702, | |
| "grad_norm": 0.894236147403717, | |
| "learning_rate": 2.2175671969281407e-05, | |
| "loss": 0.29471799850463865, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 2.338048380483805, | |
| "grad_norm": 0.9900534152984619, | |
| "learning_rate": 2.2141387822270985e-05, | |
| "loss": 0.29380016326904296, | |
| "step": 114050 | |
| }, | |
| { | |
| "epoch": 2.3390733907339074, | |
| "grad_norm": 0.999396800994873, | |
| "learning_rate": 2.210710367526056e-05, | |
| "loss": 0.2967346572875977, | |
| "step": 114100 | |
| }, | |
| { | |
| "epoch": 2.34009840098401, | |
| "grad_norm": 0.8471775054931641, | |
| "learning_rate": 2.2072819528250138e-05, | |
| "loss": 0.29183599472045896, | |
| "step": 114150 | |
| }, | |
| { | |
| "epoch": 2.341123411234112, | |
| "grad_norm": 0.9373655319213867, | |
| "learning_rate": 2.2038535381239716e-05, | |
| "loss": 0.2953369331359863, | |
| "step": 114200 | |
| }, | |
| { | |
| "epoch": 2.342148421484215, | |
| "grad_norm": 0.8588882684707642, | |
| "learning_rate": 2.2004251234229294e-05, | |
| "loss": 0.29708721160888674, | |
| "step": 114250 | |
| }, | |
| { | |
| "epoch": 2.3431734317343174, | |
| "grad_norm": 0.8311331272125244, | |
| "learning_rate": 2.1969967087218872e-05, | |
| "loss": 0.2991793441772461, | |
| "step": 114300 | |
| }, | |
| { | |
| "epoch": 2.3441984419844197, | |
| "grad_norm": 1.0317838191986084, | |
| "learning_rate": 2.193568294020845e-05, | |
| "loss": 0.28927556991577147, | |
| "step": 114350 | |
| }, | |
| { | |
| "epoch": 2.3452234522345226, | |
| "grad_norm": 0.9265230894088745, | |
| "learning_rate": 2.1901398793198028e-05, | |
| "loss": 0.29731901168823244, | |
| "step": 114400 | |
| }, | |
| { | |
| "epoch": 2.346248462484625, | |
| "grad_norm": 0.8775085806846619, | |
| "learning_rate": 2.1867114646187603e-05, | |
| "loss": 0.29290285110473635, | |
| "step": 114450 | |
| }, | |
| { | |
| "epoch": 2.3472734727347273, | |
| "grad_norm": 0.8913845419883728, | |
| "learning_rate": 2.183283049917718e-05, | |
| "loss": 0.2941425895690918, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 2.3482984829848297, | |
| "grad_norm": 0.9549135565757751, | |
| "learning_rate": 2.179854635216676e-05, | |
| "loss": 0.29423208236694337, | |
| "step": 114550 | |
| }, | |
| { | |
| "epoch": 2.3493234932349325, | |
| "grad_norm": 0.9462767243385315, | |
| "learning_rate": 2.1764262205156337e-05, | |
| "loss": 0.29479867935180665, | |
| "step": 114600 | |
| }, | |
| { | |
| "epoch": 2.350348503485035, | |
| "grad_norm": 0.9303880333900452, | |
| "learning_rate": 2.1729978058145915e-05, | |
| "loss": 0.2936443519592285, | |
| "step": 114650 | |
| }, | |
| { | |
| "epoch": 2.3513735137351373, | |
| "grad_norm": 0.9457836151123047, | |
| "learning_rate": 2.1695693911135493e-05, | |
| "loss": 0.29363698959350587, | |
| "step": 114700 | |
| }, | |
| { | |
| "epoch": 2.35239852398524, | |
| "grad_norm": 0.9594712257385254, | |
| "learning_rate": 2.166140976412507e-05, | |
| "loss": 0.29267578125, | |
| "step": 114750 | |
| }, | |
| { | |
| "epoch": 2.3534235342353425, | |
| "grad_norm": 1.0051660537719727, | |
| "learning_rate": 2.1627125617114646e-05, | |
| "loss": 0.28777292251586917, | |
| "step": 114800 | |
| }, | |
| { | |
| "epoch": 2.354448544485445, | |
| "grad_norm": 0.8472592830657959, | |
| "learning_rate": 2.1592841470104224e-05, | |
| "loss": 0.28751956939697265, | |
| "step": 114850 | |
| }, | |
| { | |
| "epoch": 2.355473554735547, | |
| "grad_norm": 1.017433762550354, | |
| "learning_rate": 2.1558557323093802e-05, | |
| "loss": 0.29304271697998047, | |
| "step": 114900 | |
| }, | |
| { | |
| "epoch": 2.35649856498565, | |
| "grad_norm": 0.9065322875976562, | |
| "learning_rate": 2.152427317608338e-05, | |
| "loss": 0.29421253204345704, | |
| "step": 114950 | |
| }, | |
| { | |
| "epoch": 2.3575235752357524, | |
| "grad_norm": 0.8683282732963562, | |
| "learning_rate": 2.148998902907296e-05, | |
| "loss": 0.2927424621582031, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 2.3575235752357524, | |
| "eval_loss": 0.3602697551250458, | |
| "eval_runtime": 4.6017, | |
| "eval_samples_per_second": 5432.754, | |
| "eval_steps_per_second": 5.433, | |
| "eval_teacher_cosine_loss": 0.22771388292312622, | |
| "eval_teacher_cosine_mean": 0.7722861170768738, | |
| "eval_teacher_cosine_p10": 0.6963392496109009, | |
| "eval_teacher_cosine_p50": 0.7820363640785217, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 2.3585485854858548, | |
| "grad_norm": 0.9129708409309387, | |
| "learning_rate": 2.1455704882062536e-05, | |
| "loss": 0.29556554794311524, | |
| "step": 115050 | |
| }, | |
| { | |
| "epoch": 2.3595735957359576, | |
| "grad_norm": 0.8830533623695374, | |
| "learning_rate": 2.1421420735052114e-05, | |
| "loss": 0.2875001335144043, | |
| "step": 115100 | |
| }, | |
| { | |
| "epoch": 2.36059860598606, | |
| "grad_norm": 0.8924604058265686, | |
| "learning_rate": 2.138713658804169e-05, | |
| "loss": 0.29439243316650393, | |
| "step": 115150 | |
| }, | |
| { | |
| "epoch": 2.3616236162361623, | |
| "grad_norm": 0.8811038136482239, | |
| "learning_rate": 2.1352852441031267e-05, | |
| "loss": 0.29714260101318357, | |
| "step": 115200 | |
| }, | |
| { | |
| "epoch": 2.3626486264862647, | |
| "grad_norm": 0.8209221363067627, | |
| "learning_rate": 2.1318568294020845e-05, | |
| "loss": 0.2943692588806152, | |
| "step": 115250 | |
| }, | |
| { | |
| "epoch": 2.3636736367363675, | |
| "grad_norm": 0.9032281041145325, | |
| "learning_rate": 2.1284284147010423e-05, | |
| "loss": 0.2942492485046387, | |
| "step": 115300 | |
| }, | |
| { | |
| "epoch": 2.36469864698647, | |
| "grad_norm": 0.8705187439918518, | |
| "learning_rate": 2.125e-05, | |
| "loss": 0.29160184860229493, | |
| "step": 115350 | |
| }, | |
| { | |
| "epoch": 2.3657236572365723, | |
| "grad_norm": 0.9203627705574036, | |
| "learning_rate": 2.121571585298958e-05, | |
| "loss": 0.2941219139099121, | |
| "step": 115400 | |
| }, | |
| { | |
| "epoch": 2.3667486674866747, | |
| "grad_norm": 0.8155919909477234, | |
| "learning_rate": 2.1181431705979158e-05, | |
| "loss": 0.29027332305908204, | |
| "step": 115450 | |
| }, | |
| { | |
| "epoch": 2.3677736777367775, | |
| "grad_norm": 0.8501454591751099, | |
| "learning_rate": 2.1147147558968732e-05, | |
| "loss": 0.29604459762573243, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 2.36879868798688, | |
| "grad_norm": 0.8721468448638916, | |
| "learning_rate": 2.111286341195831e-05, | |
| "loss": 0.2915157127380371, | |
| "step": 115550 | |
| }, | |
| { | |
| "epoch": 2.3698236982369822, | |
| "grad_norm": 0.835985541343689, | |
| "learning_rate": 2.107857926494789e-05, | |
| "loss": 0.29483240127563476, | |
| "step": 115600 | |
| }, | |
| { | |
| "epoch": 2.3708487084870846, | |
| "grad_norm": 0.9704428911209106, | |
| "learning_rate": 2.1044295117937467e-05, | |
| "loss": 0.29968265533447264, | |
| "step": 115650 | |
| }, | |
| { | |
| "epoch": 2.3718737187371874, | |
| "grad_norm": 0.8427543044090271, | |
| "learning_rate": 2.1010010970927045e-05, | |
| "loss": 0.2954860687255859, | |
| "step": 115700 | |
| }, | |
| { | |
| "epoch": 2.37289872898729, | |
| "grad_norm": 0.8587090969085693, | |
| "learning_rate": 2.0975726823916623e-05, | |
| "loss": 0.2916620826721191, | |
| "step": 115750 | |
| }, | |
| { | |
| "epoch": 2.373923739237392, | |
| "grad_norm": 0.8952643275260925, | |
| "learning_rate": 2.09414426769062e-05, | |
| "loss": 0.2911642456054688, | |
| "step": 115800 | |
| }, | |
| { | |
| "epoch": 2.374948749487495, | |
| "grad_norm": 0.9616943001747131, | |
| "learning_rate": 2.0907158529895776e-05, | |
| "loss": 0.299618034362793, | |
| "step": 115850 | |
| }, | |
| { | |
| "epoch": 2.3759737597375974, | |
| "grad_norm": 0.8490601778030396, | |
| "learning_rate": 2.0872874382885354e-05, | |
| "loss": 0.2925430870056152, | |
| "step": 115900 | |
| }, | |
| { | |
| "epoch": 2.3769987699876998, | |
| "grad_norm": 0.8581642508506775, | |
| "learning_rate": 2.0838590235874932e-05, | |
| "loss": 0.2913741874694824, | |
| "step": 115950 | |
| }, | |
| { | |
| "epoch": 2.378023780237802, | |
| "grad_norm": 0.9862703680992126, | |
| "learning_rate": 2.080430608886451e-05, | |
| "loss": 0.2934566307067871, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 2.379048790487905, | |
| "grad_norm": 0.9100529551506042, | |
| "learning_rate": 2.0770021941854088e-05, | |
| "loss": 0.2927321243286133, | |
| "step": 116050 | |
| }, | |
| { | |
| "epoch": 2.3800738007380073, | |
| "grad_norm": 0.7915745973587036, | |
| "learning_rate": 2.0735737794843666e-05, | |
| "loss": 0.2922407341003418, | |
| "step": 116100 | |
| }, | |
| { | |
| "epoch": 2.3810988109881097, | |
| "grad_norm": 0.8829566836357117, | |
| "learning_rate": 2.0701453647833244e-05, | |
| "loss": 0.2892936325073242, | |
| "step": 116150 | |
| }, | |
| { | |
| "epoch": 2.3821238212382125, | |
| "grad_norm": 0.881794273853302, | |
| "learning_rate": 2.066716950082282e-05, | |
| "loss": 0.2972703742980957, | |
| "step": 116200 | |
| }, | |
| { | |
| "epoch": 2.383148831488315, | |
| "grad_norm": 0.9483255743980408, | |
| "learning_rate": 2.0632885353812397e-05, | |
| "loss": 0.29258113861083984, | |
| "step": 116250 | |
| }, | |
| { | |
| "epoch": 2.3841738417384173, | |
| "grad_norm": 0.9662613272666931, | |
| "learning_rate": 2.0598601206801975e-05, | |
| "loss": 0.29773143768310545, | |
| "step": 116300 | |
| }, | |
| { | |
| "epoch": 2.3851988519885197, | |
| "grad_norm": 0.9392400979995728, | |
| "learning_rate": 2.0564317059791553e-05, | |
| "loss": 0.28997442245483396, | |
| "step": 116350 | |
| }, | |
| { | |
| "epoch": 2.3862238622386225, | |
| "grad_norm": 0.8928371071815491, | |
| "learning_rate": 2.053003291278113e-05, | |
| "loss": 0.2915007400512695, | |
| "step": 116400 | |
| }, | |
| { | |
| "epoch": 2.387248872488725, | |
| "grad_norm": 0.8190632462501526, | |
| "learning_rate": 2.049574876577071e-05, | |
| "loss": 0.2923478889465332, | |
| "step": 116450 | |
| }, | |
| { | |
| "epoch": 2.3882738827388272, | |
| "grad_norm": 0.8389042615890503, | |
| "learning_rate": 2.0461464618760287e-05, | |
| "loss": 0.29061323165893554, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 2.38929889298893, | |
| "grad_norm": 0.9212855100631714, | |
| "learning_rate": 2.0427180471749862e-05, | |
| "loss": 0.29221090316772463, | |
| "step": 116550 | |
| }, | |
| { | |
| "epoch": 2.3903239032390324, | |
| "grad_norm": 0.9059509634971619, | |
| "learning_rate": 2.039289632473944e-05, | |
| "loss": 0.2972230339050293, | |
| "step": 116600 | |
| }, | |
| { | |
| "epoch": 2.391348913489135, | |
| "grad_norm": 0.8734839558601379, | |
| "learning_rate": 2.0358612177729018e-05, | |
| "loss": 0.28962446212768556, | |
| "step": 116650 | |
| }, | |
| { | |
| "epoch": 2.392373923739237, | |
| "grad_norm": 0.9795411825180054, | |
| "learning_rate": 2.0324328030718596e-05, | |
| "loss": 0.29362436294555666, | |
| "step": 116700 | |
| }, | |
| { | |
| "epoch": 2.39339893398934, | |
| "grad_norm": 0.8593661785125732, | |
| "learning_rate": 2.0290043883708174e-05, | |
| "loss": 0.29214454650878907, | |
| "step": 116750 | |
| }, | |
| { | |
| "epoch": 2.3944239442394424, | |
| "grad_norm": 0.8470709323883057, | |
| "learning_rate": 2.0255759736697752e-05, | |
| "loss": 0.2937801933288574, | |
| "step": 116800 | |
| }, | |
| { | |
| "epoch": 2.3954489544895448, | |
| "grad_norm": 0.8460237979888916, | |
| "learning_rate": 2.022147558968733e-05, | |
| "loss": 0.29781177520751956, | |
| "step": 116850 | |
| }, | |
| { | |
| "epoch": 2.3964739647396476, | |
| "grad_norm": 0.9038801193237305, | |
| "learning_rate": 2.0187191442676905e-05, | |
| "loss": 0.2895893287658691, | |
| "step": 116900 | |
| }, | |
| { | |
| "epoch": 2.39749897498975, | |
| "grad_norm": 0.9412305951118469, | |
| "learning_rate": 2.0152907295666483e-05, | |
| "loss": 0.29265048980712893, | |
| "step": 116950 | |
| }, | |
| { | |
| "epoch": 2.3985239852398523, | |
| "grad_norm": 0.8094897866249084, | |
| "learning_rate": 2.011862314865606e-05, | |
| "loss": 0.2922881889343262, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 2.3995489954899547, | |
| "grad_norm": 0.8426933884620667, | |
| "learning_rate": 2.0084339001645643e-05, | |
| "loss": 0.2950929832458496, | |
| "step": 117050 | |
| }, | |
| { | |
| "epoch": 2.4005740057400575, | |
| "grad_norm": 0.8780723810195923, | |
| "learning_rate": 2.0050054854635218e-05, | |
| "loss": 0.29447498321533205, | |
| "step": 117100 | |
| }, | |
| { | |
| "epoch": 2.40159901599016, | |
| "grad_norm": 0.8872747421264648, | |
| "learning_rate": 2.0015770707624796e-05, | |
| "loss": 0.29911561965942385, | |
| "step": 117150 | |
| }, | |
| { | |
| "epoch": 2.4026240262402623, | |
| "grad_norm": 0.8839578032493591, | |
| "learning_rate": 1.9981486560614374e-05, | |
| "loss": 0.2898472023010254, | |
| "step": 117200 | |
| }, | |
| { | |
| "epoch": 2.403649036490365, | |
| "grad_norm": 0.8593679666519165, | |
| "learning_rate": 1.994720241360395e-05, | |
| "loss": 0.2932898712158203, | |
| "step": 117250 | |
| }, | |
| { | |
| "epoch": 2.4046740467404675, | |
| "grad_norm": 0.8161453008651733, | |
| "learning_rate": 1.9912918266593527e-05, | |
| "loss": 0.2934571838378906, | |
| "step": 117300 | |
| }, | |
| { | |
| "epoch": 2.40569905699057, | |
| "grad_norm": 0.8808152079582214, | |
| "learning_rate": 1.9878634119583105e-05, | |
| "loss": 0.29043291091918944, | |
| "step": 117350 | |
| }, | |
| { | |
| "epoch": 2.4067240672406722, | |
| "grad_norm": 0.917083740234375, | |
| "learning_rate": 1.9844349972572686e-05, | |
| "loss": 0.2938059616088867, | |
| "step": 117400 | |
| }, | |
| { | |
| "epoch": 2.407749077490775, | |
| "grad_norm": 1.044682502746582, | |
| "learning_rate": 1.981006582556226e-05, | |
| "loss": 0.289477653503418, | |
| "step": 117450 | |
| }, | |
| { | |
| "epoch": 2.4087740877408774, | |
| "grad_norm": 0.8029192686080933, | |
| "learning_rate": 1.977578167855184e-05, | |
| "loss": 0.29184272766113284, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 2.40979909799098, | |
| "grad_norm": 0.9607747197151184, | |
| "learning_rate": 1.9741497531541417e-05, | |
| "loss": 0.29469545364379884, | |
| "step": 117550 | |
| }, | |
| { | |
| "epoch": 2.4108241082410826, | |
| "grad_norm": 0.8039106130599976, | |
| "learning_rate": 1.970721338453099e-05, | |
| "loss": 0.28531320571899416, | |
| "step": 117600 | |
| }, | |
| { | |
| "epoch": 2.411849118491185, | |
| "grad_norm": 0.8869017958641052, | |
| "learning_rate": 1.967292923752057e-05, | |
| "loss": 0.29690942764282224, | |
| "step": 117650 | |
| }, | |
| { | |
| "epoch": 2.4128741287412874, | |
| "grad_norm": 0.9642576575279236, | |
| "learning_rate": 1.9638645090510148e-05, | |
| "loss": 0.2916181182861328, | |
| "step": 117700 | |
| }, | |
| { | |
| "epoch": 2.4138991389913897, | |
| "grad_norm": 1.0243313312530518, | |
| "learning_rate": 1.960436094349973e-05, | |
| "loss": 0.29469757080078124, | |
| "step": 117750 | |
| }, | |
| { | |
| "epoch": 2.4149241492414926, | |
| "grad_norm": 0.8645861148834229, | |
| "learning_rate": 1.9570076796489304e-05, | |
| "loss": 0.290510311126709, | |
| "step": 117800 | |
| }, | |
| { | |
| "epoch": 2.415949159491595, | |
| "grad_norm": 0.8079422116279602, | |
| "learning_rate": 1.9535792649478882e-05, | |
| "loss": 0.29283885955810546, | |
| "step": 117850 | |
| }, | |
| { | |
| "epoch": 2.4169741697416973, | |
| "grad_norm": 0.9206503629684448, | |
| "learning_rate": 1.950150850246846e-05, | |
| "loss": 0.29508222579956056, | |
| "step": 117900 | |
| }, | |
| { | |
| "epoch": 2.4179991799918, | |
| "grad_norm": 0.8677538633346558, | |
| "learning_rate": 1.9467224355458035e-05, | |
| "loss": 0.2935431289672852, | |
| "step": 117950 | |
| }, | |
| { | |
| "epoch": 2.4190241902419025, | |
| "grad_norm": 0.941310703754425, | |
| "learning_rate": 1.9432940208447613e-05, | |
| "loss": 0.2908061218261719, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 2.420049200492005, | |
| "grad_norm": 0.7821337580680847, | |
| "learning_rate": 1.939865606143719e-05, | |
| "loss": 0.2879571533203125, | |
| "step": 118050 | |
| }, | |
| { | |
| "epoch": 2.4210742107421073, | |
| "grad_norm": 0.960529625415802, | |
| "learning_rate": 1.9364371914426772e-05, | |
| "loss": 0.2931944274902344, | |
| "step": 118100 | |
| }, | |
| { | |
| "epoch": 2.42209922099221, | |
| "grad_norm": 1.005776047706604, | |
| "learning_rate": 1.9330087767416347e-05, | |
| "loss": 0.29163496017456053, | |
| "step": 118150 | |
| }, | |
| { | |
| "epoch": 2.4231242312423125, | |
| "grad_norm": 0.8607179522514343, | |
| "learning_rate": 1.9295803620405925e-05, | |
| "loss": 0.2902206039428711, | |
| "step": 118200 | |
| }, | |
| { | |
| "epoch": 2.424149241492415, | |
| "grad_norm": 0.8992176651954651, | |
| "learning_rate": 1.9261519473395503e-05, | |
| "loss": 0.28801124572753906, | |
| "step": 118250 | |
| }, | |
| { | |
| "epoch": 2.4251742517425177, | |
| "grad_norm": 0.8680164813995361, | |
| "learning_rate": 1.922723532638508e-05, | |
| "loss": 0.28865560531616213, | |
| "step": 118300 | |
| }, | |
| { | |
| "epoch": 2.42619926199262, | |
| "grad_norm": 0.8016251921653748, | |
| "learning_rate": 1.9192951179374656e-05, | |
| "loss": 0.2953912353515625, | |
| "step": 118350 | |
| }, | |
| { | |
| "epoch": 2.4272242722427224, | |
| "grad_norm": 0.9715549945831299, | |
| "learning_rate": 1.9158667032364234e-05, | |
| "loss": 0.2914541053771973, | |
| "step": 118400 | |
| }, | |
| { | |
| "epoch": 2.428249282492825, | |
| "grad_norm": 0.9597553014755249, | |
| "learning_rate": 1.9124382885353816e-05, | |
| "loss": 0.28842910766601565, | |
| "step": 118450 | |
| }, | |
| { | |
| "epoch": 2.4292742927429276, | |
| "grad_norm": 0.8453289270401001, | |
| "learning_rate": 1.909009873834339e-05, | |
| "loss": 0.29361419677734374, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 2.43029930299303, | |
| "grad_norm": 0.9309425950050354, | |
| "learning_rate": 1.905581459133297e-05, | |
| "loss": 0.2908750915527344, | |
| "step": 118550 | |
| }, | |
| { | |
| "epoch": 2.4313243132431324, | |
| "grad_norm": 0.816315233707428, | |
| "learning_rate": 1.9021530444322547e-05, | |
| "loss": 0.291048583984375, | |
| "step": 118600 | |
| }, | |
| { | |
| "epoch": 2.4323493234932347, | |
| "grad_norm": 0.9039331078529358, | |
| "learning_rate": 1.8987246297312125e-05, | |
| "loss": 0.29104415893554686, | |
| "step": 118650 | |
| }, | |
| { | |
| "epoch": 2.4333743337433376, | |
| "grad_norm": 0.8663307428359985, | |
| "learning_rate": 1.89529621503017e-05, | |
| "loss": 0.2929690361022949, | |
| "step": 118700 | |
| }, | |
| { | |
| "epoch": 2.43439934399344, | |
| "grad_norm": 0.8221769332885742, | |
| "learning_rate": 1.891867800329128e-05, | |
| "loss": 0.29125825881958006, | |
| "step": 118750 | |
| }, | |
| { | |
| "epoch": 2.4354243542435423, | |
| "grad_norm": 0.9713382124900818, | |
| "learning_rate": 1.888439385628086e-05, | |
| "loss": 0.29267024993896484, | |
| "step": 118800 | |
| }, | |
| { | |
| "epoch": 2.4364493644936447, | |
| "grad_norm": 0.8706762790679932, | |
| "learning_rate": 1.8850109709270434e-05, | |
| "loss": 0.28921051025390626, | |
| "step": 118850 | |
| }, | |
| { | |
| "epoch": 2.4374743747437475, | |
| "grad_norm": 0.8794642686843872, | |
| "learning_rate": 1.881582556226001e-05, | |
| "loss": 0.2908515930175781, | |
| "step": 118900 | |
| }, | |
| { | |
| "epoch": 2.43849938499385, | |
| "grad_norm": 0.9034032821655273, | |
| "learning_rate": 1.878154141524959e-05, | |
| "loss": 0.2972625923156738, | |
| "step": 118950 | |
| }, | |
| { | |
| "epoch": 2.4395243952439523, | |
| "grad_norm": 0.9497014284133911, | |
| "learning_rate": 1.8747257268239168e-05, | |
| "loss": 0.29750564575195315, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 2.440549405494055, | |
| "grad_norm": 0.9945149421691895, | |
| "learning_rate": 1.8712973121228743e-05, | |
| "loss": 0.29619005203247073, | |
| "step": 119050 | |
| }, | |
| { | |
| "epoch": 2.4415744157441575, | |
| "grad_norm": 0.8491098880767822, | |
| "learning_rate": 1.8678688974218324e-05, | |
| "loss": 0.28987951278686525, | |
| "step": 119100 | |
| }, | |
| { | |
| "epoch": 2.44259942599426, | |
| "grad_norm": 0.8848347663879395, | |
| "learning_rate": 1.8644404827207902e-05, | |
| "loss": 0.28769662857055667, | |
| "step": 119150 | |
| }, | |
| { | |
| "epoch": 2.443624436244362, | |
| "grad_norm": 0.9135397672653198, | |
| "learning_rate": 1.8610120680197477e-05, | |
| "loss": 0.2879890823364258, | |
| "step": 119200 | |
| }, | |
| { | |
| "epoch": 2.444649446494465, | |
| "grad_norm": 0.9636138081550598, | |
| "learning_rate": 1.8575836533187055e-05, | |
| "loss": 0.29050533294677733, | |
| "step": 119250 | |
| }, | |
| { | |
| "epoch": 2.4456744567445674, | |
| "grad_norm": 0.8019124865531921, | |
| "learning_rate": 1.8541552386176633e-05, | |
| "loss": 0.29510711669921874, | |
| "step": 119300 | |
| }, | |
| { | |
| "epoch": 2.44669946699467, | |
| "grad_norm": 0.9680748581886292, | |
| "learning_rate": 1.850726823916621e-05, | |
| "loss": 0.2901690101623535, | |
| "step": 119350 | |
| }, | |
| { | |
| "epoch": 2.4477244772447726, | |
| "grad_norm": 0.9040891528129578, | |
| "learning_rate": 1.8472984092155786e-05, | |
| "loss": 0.2927972602844238, | |
| "step": 119400 | |
| }, | |
| { | |
| "epoch": 2.448749487494875, | |
| "grad_norm": 0.9282403588294983, | |
| "learning_rate": 1.8438699945145367e-05, | |
| "loss": 0.29395462036132813, | |
| "step": 119450 | |
| }, | |
| { | |
| "epoch": 2.4497744977449774, | |
| "grad_norm": 0.9067701697349548, | |
| "learning_rate": 1.8404415798134945e-05, | |
| "loss": 0.2911212158203125, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 2.4507995079950797, | |
| "grad_norm": 0.9468734264373779, | |
| "learning_rate": 1.837013165112452e-05, | |
| "loss": 0.28661479949951174, | |
| "step": 119550 | |
| }, | |
| { | |
| "epoch": 2.4518245182451825, | |
| "grad_norm": 0.7977584004402161, | |
| "learning_rate": 1.8335847504114098e-05, | |
| "loss": 0.2909583282470703, | |
| "step": 119600 | |
| }, | |
| { | |
| "epoch": 2.452849528495285, | |
| "grad_norm": 0.9013761281967163, | |
| "learning_rate": 1.8301563357103676e-05, | |
| "loss": 0.2918566131591797, | |
| "step": 119650 | |
| }, | |
| { | |
| "epoch": 2.4538745387453873, | |
| "grad_norm": 0.8633672595024109, | |
| "learning_rate": 1.8267279210093254e-05, | |
| "loss": 0.29373287200927733, | |
| "step": 119700 | |
| }, | |
| { | |
| "epoch": 2.45489954899549, | |
| "grad_norm": 0.8338020443916321, | |
| "learning_rate": 1.823299506308283e-05, | |
| "loss": 0.29567461013793944, | |
| "step": 119750 | |
| }, | |
| { | |
| "epoch": 2.4559245592455925, | |
| "grad_norm": 0.8901137709617615, | |
| "learning_rate": 1.819871091607241e-05, | |
| "loss": 0.2912013626098633, | |
| "step": 119800 | |
| }, | |
| { | |
| "epoch": 2.456949569495695, | |
| "grad_norm": 0.9873080849647522, | |
| "learning_rate": 1.816442676906199e-05, | |
| "loss": 0.28778026580810545, | |
| "step": 119850 | |
| }, | |
| { | |
| "epoch": 2.4579745797457973, | |
| "grad_norm": 0.8672187924385071, | |
| "learning_rate": 1.8130142622051563e-05, | |
| "loss": 0.29535192489624024, | |
| "step": 119900 | |
| }, | |
| { | |
| "epoch": 2.4589995899959, | |
| "grad_norm": 0.9047064781188965, | |
| "learning_rate": 1.809585847504114e-05, | |
| "loss": 0.2901751518249512, | |
| "step": 119950 | |
| }, | |
| { | |
| "epoch": 2.4600246002460024, | |
| "grad_norm": 0.8977329730987549, | |
| "learning_rate": 1.806157432803072e-05, | |
| "loss": 0.2880943489074707, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 2.4600246002460024, | |
| "eval_loss": 0.3573911786079407, | |
| "eval_runtime": 4.5618, | |
| "eval_samples_per_second": 5480.344, | |
| "eval_steps_per_second": 5.48, | |
| "eval_teacher_cosine_loss": 0.22581195831298828, | |
| "eval_teacher_cosine_mean": 0.7741880416870117, | |
| "eval_teacher_cosine_p10": 0.6979688405990601, | |
| "eval_teacher_cosine_p50": 0.7844141721725464, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 2.461049610496105, | |
| "grad_norm": 0.8681636452674866, | |
| "learning_rate": 1.8027290181020297e-05, | |
| "loss": 0.2924513244628906, | |
| "step": 120050 | |
| }, | |
| { | |
| "epoch": 2.4620746207462076, | |
| "grad_norm": 0.8753847479820251, | |
| "learning_rate": 1.7993006034009872e-05, | |
| "loss": 0.29531843185424805, | |
| "step": 120100 | |
| }, | |
| { | |
| "epoch": 2.46309963099631, | |
| "grad_norm": 0.9279679656028748, | |
| "learning_rate": 1.7958721886999454e-05, | |
| "loss": 0.290020694732666, | |
| "step": 120150 | |
| }, | |
| { | |
| "epoch": 2.4641246412464124, | |
| "grad_norm": 0.8626305460929871, | |
| "learning_rate": 1.7924437739989032e-05, | |
| "loss": 0.29288551330566404, | |
| "step": 120200 | |
| }, | |
| { | |
| "epoch": 2.4651496514965148, | |
| "grad_norm": 0.9551226496696472, | |
| "learning_rate": 1.7890153592978606e-05, | |
| "loss": 0.2864166259765625, | |
| "step": 120250 | |
| }, | |
| { | |
| "epoch": 2.4661746617466176, | |
| "grad_norm": 0.8886475563049316, | |
| "learning_rate": 1.7855869445968185e-05, | |
| "loss": 0.29375648498535156, | |
| "step": 120300 | |
| }, | |
| { | |
| "epoch": 2.46719967199672, | |
| "grad_norm": 0.8824247121810913, | |
| "learning_rate": 1.7821585298957763e-05, | |
| "loss": 0.29665777206420896, | |
| "step": 120350 | |
| }, | |
| { | |
| "epoch": 2.4682246822468223, | |
| "grad_norm": 0.9661949276924133, | |
| "learning_rate": 1.778730115194734e-05, | |
| "loss": 0.2966621208190918, | |
| "step": 120400 | |
| }, | |
| { | |
| "epoch": 2.469249692496925, | |
| "grad_norm": 0.9454455375671387, | |
| "learning_rate": 1.775301700493692e-05, | |
| "loss": 0.289537239074707, | |
| "step": 120450 | |
| }, | |
| { | |
| "epoch": 2.4702747027470275, | |
| "grad_norm": 0.983143150806427, | |
| "learning_rate": 1.7718732857926497e-05, | |
| "loss": 0.28971467971801756, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 2.47129971299713, | |
| "grad_norm": 0.867202877998352, | |
| "learning_rate": 1.7684448710916075e-05, | |
| "loss": 0.287410831451416, | |
| "step": 120550 | |
| }, | |
| { | |
| "epoch": 2.4723247232472323, | |
| "grad_norm": 0.952211320400238, | |
| "learning_rate": 1.765016456390565e-05, | |
| "loss": 0.29342222213745117, | |
| "step": 120600 | |
| }, | |
| { | |
| "epoch": 2.473349733497335, | |
| "grad_norm": 0.9632975459098816, | |
| "learning_rate": 1.7615880416895228e-05, | |
| "loss": 0.29439937591552734, | |
| "step": 120650 | |
| }, | |
| { | |
| "epoch": 2.4743747437474375, | |
| "grad_norm": 0.8762933611869812, | |
| "learning_rate": 1.7581596269884806e-05, | |
| "loss": 0.28978254318237306, | |
| "step": 120700 | |
| }, | |
| { | |
| "epoch": 2.47539975399754, | |
| "grad_norm": 0.8218929767608643, | |
| "learning_rate": 1.7547312122874384e-05, | |
| "loss": 0.2883963203430176, | |
| "step": 120750 | |
| }, | |
| { | |
| "epoch": 2.4764247642476427, | |
| "grad_norm": 1.0789754390716553, | |
| "learning_rate": 1.7513027975863962e-05, | |
| "loss": 0.2899725914001465, | |
| "step": 120800 | |
| }, | |
| { | |
| "epoch": 2.477449774497745, | |
| "grad_norm": 0.9881124496459961, | |
| "learning_rate": 1.747874382885354e-05, | |
| "loss": 0.29112930297851564, | |
| "step": 120850 | |
| }, | |
| { | |
| "epoch": 2.4784747847478474, | |
| "grad_norm": 1.0431567430496216, | |
| "learning_rate": 1.7444459681843118e-05, | |
| "loss": 0.29119268417358396, | |
| "step": 120900 | |
| }, | |
| { | |
| "epoch": 2.47949979499795, | |
| "grad_norm": 0.9441935420036316, | |
| "learning_rate": 1.7410175534832693e-05, | |
| "loss": 0.2844077491760254, | |
| "step": 120950 | |
| }, | |
| { | |
| "epoch": 2.4805248052480526, | |
| "grad_norm": 0.9924372434616089, | |
| "learning_rate": 1.737589138782227e-05, | |
| "loss": 0.29877248764038083, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 2.481549815498155, | |
| "grad_norm": 1.0113941431045532, | |
| "learning_rate": 1.734160724081185e-05, | |
| "loss": 0.28608146667480466, | |
| "step": 121050 | |
| }, | |
| { | |
| "epoch": 2.4825748257482574, | |
| "grad_norm": 0.8419355750083923, | |
| "learning_rate": 1.7307323093801427e-05, | |
| "loss": 0.2857310676574707, | |
| "step": 121100 | |
| }, | |
| { | |
| "epoch": 2.48359983599836, | |
| "grad_norm": 0.8440735340118408, | |
| "learning_rate": 1.7273038946791005e-05, | |
| "loss": 0.286184139251709, | |
| "step": 121150 | |
| }, | |
| { | |
| "epoch": 2.4846248462484626, | |
| "grad_norm": 0.9082350134849548, | |
| "learning_rate": 1.7238754799780583e-05, | |
| "loss": 0.29266704559326173, | |
| "step": 121200 | |
| }, | |
| { | |
| "epoch": 2.485649856498565, | |
| "grad_norm": 0.8908433318138123, | |
| "learning_rate": 1.720447065277016e-05, | |
| "loss": 0.28848312377929686, | |
| "step": 121250 | |
| }, | |
| { | |
| "epoch": 2.4866748667486673, | |
| "grad_norm": 0.9164854884147644, | |
| "learning_rate": 1.717018650575974e-05, | |
| "loss": 0.2876583099365234, | |
| "step": 121300 | |
| }, | |
| { | |
| "epoch": 2.48769987699877, | |
| "grad_norm": 0.9307411909103394, | |
| "learning_rate": 1.7135902358749314e-05, | |
| "loss": 0.2894133949279785, | |
| "step": 121350 | |
| }, | |
| { | |
| "epoch": 2.4887248872488725, | |
| "grad_norm": 0.8509969711303711, | |
| "learning_rate": 1.7101618211738892e-05, | |
| "loss": 0.28459482192993163, | |
| "step": 121400 | |
| }, | |
| { | |
| "epoch": 2.489749897498975, | |
| "grad_norm": 0.999386727809906, | |
| "learning_rate": 1.706733406472847e-05, | |
| "loss": 0.29040523529052736, | |
| "step": 121450 | |
| }, | |
| { | |
| "epoch": 2.4907749077490777, | |
| "grad_norm": 0.8954871296882629, | |
| "learning_rate": 1.703304991771805e-05, | |
| "loss": 0.3008843421936035, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 2.49179991799918, | |
| "grad_norm": 0.8739713430404663, | |
| "learning_rate": 1.6998765770707626e-05, | |
| "loss": 0.2850011444091797, | |
| "step": 121550 | |
| }, | |
| { | |
| "epoch": 2.4928249282492825, | |
| "grad_norm": 0.9396263957023621, | |
| "learning_rate": 1.6964481623697205e-05, | |
| "loss": 0.2910104751586914, | |
| "step": 121600 | |
| }, | |
| { | |
| "epoch": 2.493849938499385, | |
| "grad_norm": 0.9240145087242126, | |
| "learning_rate": 1.6930197476686783e-05, | |
| "loss": 0.29323114395141603, | |
| "step": 121650 | |
| }, | |
| { | |
| "epoch": 2.4948749487494877, | |
| "grad_norm": 0.9708922505378723, | |
| "learning_rate": 1.6895913329676357e-05, | |
| "loss": 0.2876408004760742, | |
| "step": 121700 | |
| }, | |
| { | |
| "epoch": 2.49589995899959, | |
| "grad_norm": 0.8289485573768616, | |
| "learning_rate": 1.6861629182665935e-05, | |
| "loss": 0.28655014038085935, | |
| "step": 121750 | |
| }, | |
| { | |
| "epoch": 2.4969249692496924, | |
| "grad_norm": 0.8318237066268921, | |
| "learning_rate": 1.6827345035655514e-05, | |
| "loss": 0.2943228530883789, | |
| "step": 121800 | |
| }, | |
| { | |
| "epoch": 2.497949979499795, | |
| "grad_norm": 0.9992183446884155, | |
| "learning_rate": 1.679306088864509e-05, | |
| "loss": 0.2935347557067871, | |
| "step": 121850 | |
| }, | |
| { | |
| "epoch": 2.4989749897498976, | |
| "grad_norm": 0.9203330874443054, | |
| "learning_rate": 1.675877674163467e-05, | |
| "loss": 0.29014535903930666, | |
| "step": 121900 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.8969088792800903, | |
| "learning_rate": 1.6724492594624248e-05, | |
| "loss": 0.2844480895996094, | |
| "step": 121950 | |
| }, | |
| { | |
| "epoch": 2.5010250102501024, | |
| "grad_norm": 0.8596603870391846, | |
| "learning_rate": 1.6690208447613826e-05, | |
| "loss": 0.28979154586791994, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 2.5020500205002048, | |
| "grad_norm": 0.8600313067436218, | |
| "learning_rate": 1.66559243006034e-05, | |
| "loss": 0.2954870796203613, | |
| "step": 122050 | |
| }, | |
| { | |
| "epoch": 2.5030750307503076, | |
| "grad_norm": 0.8832524418830872, | |
| "learning_rate": 1.662164015359298e-05, | |
| "loss": 0.29178409576416015, | |
| "step": 122100 | |
| }, | |
| { | |
| "epoch": 2.50410004100041, | |
| "grad_norm": 0.8712535500526428, | |
| "learning_rate": 1.6587356006582557e-05, | |
| "loss": 0.28951618194580075, | |
| "step": 122150 | |
| }, | |
| { | |
| "epoch": 2.5051250512505128, | |
| "grad_norm": 0.8481578230857849, | |
| "learning_rate": 1.6553071859572135e-05, | |
| "loss": 0.2872966766357422, | |
| "step": 122200 | |
| }, | |
| { | |
| "epoch": 2.506150061500615, | |
| "grad_norm": 0.9046580791473389, | |
| "learning_rate": 1.6518787712561713e-05, | |
| "loss": 0.2872707748413086, | |
| "step": 122250 | |
| }, | |
| { | |
| "epoch": 2.5071750717507175, | |
| "grad_norm": 0.8534570932388306, | |
| "learning_rate": 1.648450356555129e-05, | |
| "loss": 0.2899884033203125, | |
| "step": 122300 | |
| }, | |
| { | |
| "epoch": 2.50820008200082, | |
| "grad_norm": 0.8494232296943665, | |
| "learning_rate": 1.645021941854087e-05, | |
| "loss": 0.29238492965698243, | |
| "step": 122350 | |
| }, | |
| { | |
| "epoch": 2.5092250922509223, | |
| "grad_norm": 0.9296335577964783, | |
| "learning_rate": 1.6415935271530444e-05, | |
| "loss": 0.2918229866027832, | |
| "step": 122400 | |
| }, | |
| { | |
| "epoch": 2.510250102501025, | |
| "grad_norm": 0.9118334054946899, | |
| "learning_rate": 1.6381651124520022e-05, | |
| "loss": 0.29319421768188475, | |
| "step": 122450 | |
| }, | |
| { | |
| "epoch": 2.5112751127511275, | |
| "grad_norm": 1.0155823230743408, | |
| "learning_rate": 1.63473669775096e-05, | |
| "loss": 0.2911719512939453, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 2.5123001230012303, | |
| "grad_norm": 0.9896712899208069, | |
| "learning_rate": 1.6313082830499178e-05, | |
| "loss": 0.29345903396606443, | |
| "step": 122550 | |
| }, | |
| { | |
| "epoch": 2.5133251332513327, | |
| "grad_norm": 0.8693930506706238, | |
| "learning_rate": 1.6278798683488756e-05, | |
| "loss": 0.29266006469726563, | |
| "step": 122600 | |
| }, | |
| { | |
| "epoch": 2.514350143501435, | |
| "grad_norm": 0.7712063193321228, | |
| "learning_rate": 1.6244514536478334e-05, | |
| "loss": 0.2877850723266602, | |
| "step": 122650 | |
| }, | |
| { | |
| "epoch": 2.5153751537515374, | |
| "grad_norm": 0.9704263210296631, | |
| "learning_rate": 1.6210230389467912e-05, | |
| "loss": 0.28948055267333983, | |
| "step": 122700 | |
| }, | |
| { | |
| "epoch": 2.51640016400164, | |
| "grad_norm": 0.9746927618980408, | |
| "learning_rate": 1.6175946242457487e-05, | |
| "loss": 0.28762619018554686, | |
| "step": 122750 | |
| }, | |
| { | |
| "epoch": 2.5174251742517426, | |
| "grad_norm": 0.9486260414123535, | |
| "learning_rate": 1.6141662095447065e-05, | |
| "loss": 0.2916549110412598, | |
| "step": 122800 | |
| }, | |
| { | |
| "epoch": 2.518450184501845, | |
| "grad_norm": 0.9005641937255859, | |
| "learning_rate": 1.6107377948436643e-05, | |
| "loss": 0.2913349533081055, | |
| "step": 122850 | |
| }, | |
| { | |
| "epoch": 2.5194751947519474, | |
| "grad_norm": 0.8470512628555298, | |
| "learning_rate": 1.607309380142622e-05, | |
| "loss": 0.29011892318725585, | |
| "step": 122900 | |
| }, | |
| { | |
| "epoch": 2.52050020500205, | |
| "grad_norm": 0.9296560883522034, | |
| "learning_rate": 1.60388096544158e-05, | |
| "loss": 0.2879182815551758, | |
| "step": 122950 | |
| }, | |
| { | |
| "epoch": 2.5215252152521526, | |
| "grad_norm": 0.9072111248970032, | |
| "learning_rate": 1.6004525507405377e-05, | |
| "loss": 0.2932208824157715, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 2.522550225502255, | |
| "grad_norm": 0.9888075590133667, | |
| "learning_rate": 1.5970241360394955e-05, | |
| "loss": 0.2833209419250488, | |
| "step": 123050 | |
| }, | |
| { | |
| "epoch": 2.5235752357523573, | |
| "grad_norm": 0.8224952816963196, | |
| "learning_rate": 1.593595721338453e-05, | |
| "loss": 0.2885819625854492, | |
| "step": 123100 | |
| }, | |
| { | |
| "epoch": 2.52460024600246, | |
| "grad_norm": 1.033673882484436, | |
| "learning_rate": 1.5901673066374108e-05, | |
| "loss": 0.2931675910949707, | |
| "step": 123150 | |
| }, | |
| { | |
| "epoch": 2.5256252562525625, | |
| "grad_norm": 0.9990005493164062, | |
| "learning_rate": 1.5867388919363686e-05, | |
| "loss": 0.29214986801147463, | |
| "step": 123200 | |
| }, | |
| { | |
| "epoch": 2.526650266502665, | |
| "grad_norm": 0.9784498810768127, | |
| "learning_rate": 1.5833104772353264e-05, | |
| "loss": 0.28983661651611325, | |
| "step": 123250 | |
| }, | |
| { | |
| "epoch": 2.5276752767527677, | |
| "grad_norm": 0.9017473459243774, | |
| "learning_rate": 1.5798820625342843e-05, | |
| "loss": 0.291893367767334, | |
| "step": 123300 | |
| }, | |
| { | |
| "epoch": 2.52870028700287, | |
| "grad_norm": 0.924109935760498, | |
| "learning_rate": 1.576453647833242e-05, | |
| "loss": 0.28658069610595704, | |
| "step": 123350 | |
| }, | |
| { | |
| "epoch": 2.5297252972529725, | |
| "grad_norm": 0.9996832609176636, | |
| "learning_rate": 1.5730252331322e-05, | |
| "loss": 0.2898409843444824, | |
| "step": 123400 | |
| }, | |
| { | |
| "epoch": 2.530750307503075, | |
| "grad_norm": 0.9308768510818481, | |
| "learning_rate": 1.5695968184311573e-05, | |
| "loss": 0.29080766677856446, | |
| "step": 123450 | |
| }, | |
| { | |
| "epoch": 2.5317753177531777, | |
| "grad_norm": 0.9657359719276428, | |
| "learning_rate": 1.566168403730115e-05, | |
| "loss": 0.28270708084106444, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 2.53280032800328, | |
| "grad_norm": 0.91163569688797, | |
| "learning_rate": 1.562739989029073e-05, | |
| "loss": 0.28729700088500976, | |
| "step": 123550 | |
| }, | |
| { | |
| "epoch": 2.5338253382533824, | |
| "grad_norm": 0.9579923152923584, | |
| "learning_rate": 1.5593115743280308e-05, | |
| "loss": 0.29202396392822266, | |
| "step": 123600 | |
| }, | |
| { | |
| "epoch": 2.5348503485034852, | |
| "grad_norm": 0.9233018159866333, | |
| "learning_rate": 1.5558831596269886e-05, | |
| "loss": 0.2864284896850586, | |
| "step": 123650 | |
| }, | |
| { | |
| "epoch": 2.5358753587535876, | |
| "grad_norm": 0.9077503681182861, | |
| "learning_rate": 1.5524547449259464e-05, | |
| "loss": 0.28452043533325194, | |
| "step": 123700 | |
| }, | |
| { | |
| "epoch": 2.53690036900369, | |
| "grad_norm": 0.9166804552078247, | |
| "learning_rate": 1.5490263302249042e-05, | |
| "loss": 0.2857227897644043, | |
| "step": 123750 | |
| }, | |
| { | |
| "epoch": 2.5379253792537924, | |
| "grad_norm": 0.8766891956329346, | |
| "learning_rate": 1.5455979155238617e-05, | |
| "loss": 0.2913432502746582, | |
| "step": 123800 | |
| }, | |
| { | |
| "epoch": 2.538950389503895, | |
| "grad_norm": 0.8360671997070312, | |
| "learning_rate": 1.5421695008228195e-05, | |
| "loss": 0.28093355178833007, | |
| "step": 123850 | |
| }, | |
| { | |
| "epoch": 2.5399753997539976, | |
| "grad_norm": 0.9075208306312561, | |
| "learning_rate": 1.5387410861217773e-05, | |
| "loss": 0.28754926681518556, | |
| "step": 123900 | |
| }, | |
| { | |
| "epoch": 2.5410004100041, | |
| "grad_norm": 0.9108579754829407, | |
| "learning_rate": 1.5353126714207354e-05, | |
| "loss": 0.2900448989868164, | |
| "step": 123950 | |
| }, | |
| { | |
| "epoch": 2.5420254202542027, | |
| "grad_norm": 0.7434515953063965, | |
| "learning_rate": 1.531884256719693e-05, | |
| "loss": 0.2947080039978027, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 2.543050430504305, | |
| "grad_norm": 0.9114060997962952, | |
| "learning_rate": 1.5284558420186507e-05, | |
| "loss": 0.28562557220458984, | |
| "step": 124050 | |
| }, | |
| { | |
| "epoch": 2.5440754407544075, | |
| "grad_norm": 0.8588831424713135, | |
| "learning_rate": 1.5250274273176085e-05, | |
| "loss": 0.28704544067382814, | |
| "step": 124100 | |
| }, | |
| { | |
| "epoch": 2.54510045100451, | |
| "grad_norm": 0.9164804816246033, | |
| "learning_rate": 1.5215990126165661e-05, | |
| "loss": 0.2893808937072754, | |
| "step": 124150 | |
| }, | |
| { | |
| "epoch": 2.5461254612546127, | |
| "grad_norm": 0.9556722640991211, | |
| "learning_rate": 1.518170597915524e-05, | |
| "loss": 0.29119874954223635, | |
| "step": 124200 | |
| }, | |
| { | |
| "epoch": 2.547150471504715, | |
| "grad_norm": 0.8894028067588806, | |
| "learning_rate": 1.5147421832144818e-05, | |
| "loss": 0.2921468925476074, | |
| "step": 124250 | |
| }, | |
| { | |
| "epoch": 2.5481754817548175, | |
| "grad_norm": 0.968216061592102, | |
| "learning_rate": 1.5113137685134396e-05, | |
| "loss": 0.28494062423706057, | |
| "step": 124300 | |
| }, | |
| { | |
| "epoch": 2.5492004920049203, | |
| "grad_norm": 0.9217172265052795, | |
| "learning_rate": 1.5078853538123972e-05, | |
| "loss": 0.2888154983520508, | |
| "step": 124350 | |
| }, | |
| { | |
| "epoch": 2.5502255022550226, | |
| "grad_norm": 1.0175808668136597, | |
| "learning_rate": 1.504456939111355e-05, | |
| "loss": 0.2908037567138672, | |
| "step": 124400 | |
| }, | |
| { | |
| "epoch": 2.551250512505125, | |
| "grad_norm": 0.8861957788467407, | |
| "learning_rate": 1.5010285244103128e-05, | |
| "loss": 0.2881911087036133, | |
| "step": 124450 | |
| }, | |
| { | |
| "epoch": 2.5522755227552274, | |
| "grad_norm": 0.8603388667106628, | |
| "learning_rate": 1.4976001097092705e-05, | |
| "loss": 0.2926754951477051, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 2.55330053300533, | |
| "grad_norm": 0.8782898783683777, | |
| "learning_rate": 1.4941716950082283e-05, | |
| "loss": 0.29313562393188475, | |
| "step": 124550 | |
| }, | |
| { | |
| "epoch": 2.5543255432554326, | |
| "grad_norm": 0.9232538938522339, | |
| "learning_rate": 1.4907432803071861e-05, | |
| "loss": 0.290765266418457, | |
| "step": 124600 | |
| }, | |
| { | |
| "epoch": 2.555350553505535, | |
| "grad_norm": 1.0029239654541016, | |
| "learning_rate": 1.4873148656061439e-05, | |
| "loss": 0.2882759666442871, | |
| "step": 124650 | |
| }, | |
| { | |
| "epoch": 2.556375563755638, | |
| "grad_norm": 0.9692409634590149, | |
| "learning_rate": 1.4838864509051015e-05, | |
| "loss": 0.2892047119140625, | |
| "step": 124700 | |
| }, | |
| { | |
| "epoch": 2.55740057400574, | |
| "grad_norm": 0.8958144187927246, | |
| "learning_rate": 1.4804580362040593e-05, | |
| "loss": 0.2927612113952637, | |
| "step": 124750 | |
| }, | |
| { | |
| "epoch": 2.5584255842558425, | |
| "grad_norm": 0.9015181660652161, | |
| "learning_rate": 1.4770296215030172e-05, | |
| "loss": 0.28911336898803713, | |
| "step": 124800 | |
| }, | |
| { | |
| "epoch": 2.559450594505945, | |
| "grad_norm": 0.9216517806053162, | |
| "learning_rate": 1.4736012068019748e-05, | |
| "loss": 0.29164674758911135, | |
| "step": 124850 | |
| }, | |
| { | |
| "epoch": 2.5604756047560473, | |
| "grad_norm": 0.9058089256286621, | |
| "learning_rate": 1.4701727921009326e-05, | |
| "loss": 0.29210403442382815, | |
| "step": 124900 | |
| }, | |
| { | |
| "epoch": 2.56150061500615, | |
| "grad_norm": 0.8299572467803955, | |
| "learning_rate": 1.4667443773998904e-05, | |
| "loss": 0.29126682281494143, | |
| "step": 124950 | |
| }, | |
| { | |
| "epoch": 2.5625256252562525, | |
| "grad_norm": 0.8630422949790955, | |
| "learning_rate": 1.4633159626988482e-05, | |
| "loss": 0.28574405670166014, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 2.5625256252562525, | |
| "eval_loss": 0.35482868552207947, | |
| "eval_runtime": 4.5942, | |
| "eval_samples_per_second": 5441.629, | |
| "eval_steps_per_second": 5.442, | |
| "eval_teacher_cosine_loss": 0.22441154718399048, | |
| "eval_teacher_cosine_mean": 0.7755884528160095, | |
| "eval_teacher_cosine_p10": 0.6993184089660645, | |
| "eval_teacher_cosine_p50": 0.7857810258865356, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 2.5635506355063553, | |
| "grad_norm": 0.8793892860412598, | |
| "learning_rate": 1.4598875479978059e-05, | |
| "loss": 0.28699670791625975, | |
| "step": 125050 | |
| }, | |
| { | |
| "epoch": 2.5645756457564577, | |
| "grad_norm": 0.887455403804779, | |
| "learning_rate": 1.4564591332967637e-05, | |
| "loss": 0.2897138404846191, | |
| "step": 125100 | |
| }, | |
| { | |
| "epoch": 2.56560065600656, | |
| "grad_norm": 0.9993875622749329, | |
| "learning_rate": 1.4530307185957215e-05, | |
| "loss": 0.2911001014709473, | |
| "step": 125150 | |
| }, | |
| { | |
| "epoch": 2.5666256662566624, | |
| "grad_norm": 0.815069317817688, | |
| "learning_rate": 1.4496023038946791e-05, | |
| "loss": 0.2847999954223633, | |
| "step": 125200 | |
| }, | |
| { | |
| "epoch": 2.567650676506765, | |
| "grad_norm": 0.9466104507446289, | |
| "learning_rate": 1.446173889193637e-05, | |
| "loss": 0.29033891677856444, | |
| "step": 125250 | |
| }, | |
| { | |
| "epoch": 2.5686756867568676, | |
| "grad_norm": 0.911689043045044, | |
| "learning_rate": 1.4427454744925947e-05, | |
| "loss": 0.2867807960510254, | |
| "step": 125300 | |
| }, | |
| { | |
| "epoch": 2.56970069700697, | |
| "grad_norm": 0.9113633632659912, | |
| "learning_rate": 1.4393170597915525e-05, | |
| "loss": 0.2908939743041992, | |
| "step": 125350 | |
| }, | |
| { | |
| "epoch": 2.570725707257073, | |
| "grad_norm": 0.8580791354179382, | |
| "learning_rate": 1.4358886450905102e-05, | |
| "loss": 0.2868094635009766, | |
| "step": 125400 | |
| }, | |
| { | |
| "epoch": 2.571750717507175, | |
| "grad_norm": 0.894975483417511, | |
| "learning_rate": 1.432460230389468e-05, | |
| "loss": 0.2962175750732422, | |
| "step": 125450 | |
| }, | |
| { | |
| "epoch": 2.5727757277572776, | |
| "grad_norm": 0.8675337433815002, | |
| "learning_rate": 1.4290318156884258e-05, | |
| "loss": 0.2925552749633789, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 2.57380073800738, | |
| "grad_norm": 0.830657422542572, | |
| "learning_rate": 1.4256034009873834e-05, | |
| "loss": 0.28648574829101564, | |
| "step": 125550 | |
| }, | |
| { | |
| "epoch": 2.5748257482574823, | |
| "grad_norm": 0.9442114233970642, | |
| "learning_rate": 1.4221749862863412e-05, | |
| "loss": 0.28440305709838865, | |
| "step": 125600 | |
| }, | |
| { | |
| "epoch": 2.575850758507585, | |
| "grad_norm": 1.050042986869812, | |
| "learning_rate": 1.418746571585299e-05, | |
| "loss": 0.29291107177734377, | |
| "step": 125650 | |
| }, | |
| { | |
| "epoch": 2.5768757687576875, | |
| "grad_norm": 0.8491564989089966, | |
| "learning_rate": 1.4153181568842569e-05, | |
| "loss": 0.28838003158569336, | |
| "step": 125700 | |
| }, | |
| { | |
| "epoch": 2.5779007790077904, | |
| "grad_norm": 0.7950577139854431, | |
| "learning_rate": 1.4118897421832145e-05, | |
| "loss": 0.2834554100036621, | |
| "step": 125750 | |
| }, | |
| { | |
| "epoch": 2.5789257892578927, | |
| "grad_norm": 0.9213470220565796, | |
| "learning_rate": 1.4084613274821723e-05, | |
| "loss": 0.29096059799194335, | |
| "step": 125800 | |
| }, | |
| { | |
| "epoch": 2.579950799507995, | |
| "grad_norm": 0.8824624419212341, | |
| "learning_rate": 1.4050329127811301e-05, | |
| "loss": 0.29198875427246096, | |
| "step": 125850 | |
| }, | |
| { | |
| "epoch": 2.5809758097580975, | |
| "grad_norm": 0.9599575996398926, | |
| "learning_rate": 1.4016044980800878e-05, | |
| "loss": 0.28924137115478515, | |
| "step": 125900 | |
| }, | |
| { | |
| "epoch": 2.5820008200082, | |
| "grad_norm": 0.8263136744499207, | |
| "learning_rate": 1.3981760833790456e-05, | |
| "loss": 0.28431766510009765, | |
| "step": 125950 | |
| }, | |
| { | |
| "epoch": 2.5830258302583027, | |
| "grad_norm": 0.9727679491043091, | |
| "learning_rate": 1.3947476686780034e-05, | |
| "loss": 0.290102596282959, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 2.584050840508405, | |
| "grad_norm": 0.7776066064834595, | |
| "learning_rate": 1.3913192539769612e-05, | |
| "loss": 0.28874576568603516, | |
| "step": 126050 | |
| }, | |
| { | |
| "epoch": 2.5850758507585074, | |
| "grad_norm": 0.8617852330207825, | |
| "learning_rate": 1.3878908392759188e-05, | |
| "loss": 0.29063032150268553, | |
| "step": 126100 | |
| }, | |
| { | |
| "epoch": 2.5861008610086103, | |
| "grad_norm": 0.9641491770744324, | |
| "learning_rate": 1.3844624245748766e-05, | |
| "loss": 0.29044986724853517, | |
| "step": 126150 | |
| }, | |
| { | |
| "epoch": 2.5871258712587126, | |
| "grad_norm": 0.8334475159645081, | |
| "learning_rate": 1.3810340098738344e-05, | |
| "loss": 0.2834018325805664, | |
| "step": 126200 | |
| }, | |
| { | |
| "epoch": 2.588150881508815, | |
| "grad_norm": 0.9625629782676697, | |
| "learning_rate": 1.377605595172792e-05, | |
| "loss": 0.2851939392089844, | |
| "step": 126250 | |
| }, | |
| { | |
| "epoch": 2.5891758917589174, | |
| "grad_norm": 0.9652614593505859, | |
| "learning_rate": 1.3741771804717499e-05, | |
| "loss": 0.2918891525268555, | |
| "step": 126300 | |
| }, | |
| { | |
| "epoch": 2.59020090200902, | |
| "grad_norm": 0.9606708288192749, | |
| "learning_rate": 1.3707487657707077e-05, | |
| "loss": 0.2889429473876953, | |
| "step": 126350 | |
| }, | |
| { | |
| "epoch": 2.5912259122591226, | |
| "grad_norm": 0.889956533908844, | |
| "learning_rate": 1.3673203510696655e-05, | |
| "loss": 0.2902523040771484, | |
| "step": 126400 | |
| }, | |
| { | |
| "epoch": 2.592250922509225, | |
| "grad_norm": 0.8522166013717651, | |
| "learning_rate": 1.3638919363686231e-05, | |
| "loss": 0.2906357002258301, | |
| "step": 126450 | |
| }, | |
| { | |
| "epoch": 2.5932759327593278, | |
| "grad_norm": 0.8673744201660156, | |
| "learning_rate": 1.360463521667581e-05, | |
| "loss": 0.29321680068969724, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 2.59430094300943, | |
| "grad_norm": 0.9032505750656128, | |
| "learning_rate": 1.3570351069665388e-05, | |
| "loss": 0.28727428436279295, | |
| "step": 126550 | |
| }, | |
| { | |
| "epoch": 2.5953259532595325, | |
| "grad_norm": 0.9062594175338745, | |
| "learning_rate": 1.3536066922654966e-05, | |
| "loss": 0.2899449348449707, | |
| "step": 126600 | |
| }, | |
| { | |
| "epoch": 2.596350963509635, | |
| "grad_norm": 0.9251903295516968, | |
| "learning_rate": 1.3501782775644542e-05, | |
| "loss": 0.2797533988952637, | |
| "step": 126650 | |
| }, | |
| { | |
| "epoch": 2.5973759737597377, | |
| "grad_norm": 0.8733513355255127, | |
| "learning_rate": 1.346749862863412e-05, | |
| "loss": 0.28341827392578123, | |
| "step": 126700 | |
| }, | |
| { | |
| "epoch": 2.59840098400984, | |
| "grad_norm": 0.8493253588676453, | |
| "learning_rate": 1.3433214481623698e-05, | |
| "loss": 0.28889219284057616, | |
| "step": 126750 | |
| }, | |
| { | |
| "epoch": 2.5994259942599425, | |
| "grad_norm": 0.8332608938217163, | |
| "learning_rate": 1.3398930334613275e-05, | |
| "loss": 0.2901733779907227, | |
| "step": 126800 | |
| }, | |
| { | |
| "epoch": 2.6004510045100453, | |
| "grad_norm": 0.9012063145637512, | |
| "learning_rate": 1.3364646187602853e-05, | |
| "loss": 0.29180164337158204, | |
| "step": 126850 | |
| }, | |
| { | |
| "epoch": 2.6014760147601477, | |
| "grad_norm": 0.9540715217590332, | |
| "learning_rate": 1.333036204059243e-05, | |
| "loss": 0.2928956604003906, | |
| "step": 126900 | |
| }, | |
| { | |
| "epoch": 2.60250102501025, | |
| "grad_norm": 0.8621730208396912, | |
| "learning_rate": 1.3296077893582009e-05, | |
| "loss": 0.29229259490966797, | |
| "step": 126950 | |
| }, | |
| { | |
| "epoch": 2.6035260352603524, | |
| "grad_norm": 0.9612332582473755, | |
| "learning_rate": 1.3261793746571585e-05, | |
| "loss": 0.29197967529296875, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 2.6045510455104552, | |
| "grad_norm": 0.8159084916114807, | |
| "learning_rate": 1.3227509599561163e-05, | |
| "loss": 0.2872229194641113, | |
| "step": 127050 | |
| }, | |
| { | |
| "epoch": 2.6055760557605576, | |
| "grad_norm": 1.003429651260376, | |
| "learning_rate": 1.3193225452550741e-05, | |
| "loss": 0.2902134704589844, | |
| "step": 127100 | |
| }, | |
| { | |
| "epoch": 2.60660106601066, | |
| "grad_norm": 0.8841830492019653, | |
| "learning_rate": 1.3158941305540318e-05, | |
| "loss": 0.2877028846740723, | |
| "step": 127150 | |
| }, | |
| { | |
| "epoch": 2.607626076260763, | |
| "grad_norm": 0.953175961971283, | |
| "learning_rate": 1.3124657158529896e-05, | |
| "loss": 0.2872025108337402, | |
| "step": 127200 | |
| }, | |
| { | |
| "epoch": 2.608651086510865, | |
| "grad_norm": 0.8366156816482544, | |
| "learning_rate": 1.3090373011519474e-05, | |
| "loss": 0.29553956985473634, | |
| "step": 127250 | |
| }, | |
| { | |
| "epoch": 2.6096760967609676, | |
| "grad_norm": 0.8093856573104858, | |
| "learning_rate": 1.3056088864509052e-05, | |
| "loss": 0.29251909255981445, | |
| "step": 127300 | |
| }, | |
| { | |
| "epoch": 2.61070110701107, | |
| "grad_norm": 0.9371615052223206, | |
| "learning_rate": 1.3021804717498628e-05, | |
| "loss": 0.28456111907958986, | |
| "step": 127350 | |
| }, | |
| { | |
| "epoch": 2.6117261172611728, | |
| "grad_norm": 0.931952714920044, | |
| "learning_rate": 1.2987520570488207e-05, | |
| "loss": 0.29139841079711915, | |
| "step": 127400 | |
| }, | |
| { | |
| "epoch": 2.612751127511275, | |
| "grad_norm": 1.0688202381134033, | |
| "learning_rate": 1.2953236423477785e-05, | |
| "loss": 0.2879133987426758, | |
| "step": 127450 | |
| }, | |
| { | |
| "epoch": 2.6137761377613775, | |
| "grad_norm": 0.9129714965820312, | |
| "learning_rate": 1.2918952276467361e-05, | |
| "loss": 0.28966991424560545, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 2.6148011480114803, | |
| "grad_norm": 0.9581945538520813, | |
| "learning_rate": 1.2884668129456939e-05, | |
| "loss": 0.2860833549499512, | |
| "step": 127550 | |
| }, | |
| { | |
| "epoch": 2.6158261582615827, | |
| "grad_norm": 0.9344260692596436, | |
| "learning_rate": 1.2850383982446517e-05, | |
| "loss": 0.28462226867675783, | |
| "step": 127600 | |
| }, | |
| { | |
| "epoch": 2.616851168511685, | |
| "grad_norm": 0.8806621432304382, | |
| "learning_rate": 1.2816099835436097e-05, | |
| "loss": 0.28714704513549805, | |
| "step": 127650 | |
| }, | |
| { | |
| "epoch": 2.6178761787617875, | |
| "grad_norm": 0.9804829359054565, | |
| "learning_rate": 1.2781815688425672e-05, | |
| "loss": 0.28752138137817385, | |
| "step": 127700 | |
| }, | |
| { | |
| "epoch": 2.6189011890118903, | |
| "grad_norm": 0.9272515773773193, | |
| "learning_rate": 1.274753154141525e-05, | |
| "loss": 0.2881778717041016, | |
| "step": 127750 | |
| }, | |
| { | |
| "epoch": 2.6199261992619927, | |
| "grad_norm": 0.9217604994773865, | |
| "learning_rate": 1.2713247394404828e-05, | |
| "loss": 0.28796537399291994, | |
| "step": 127800 | |
| }, | |
| { | |
| "epoch": 2.620951209512095, | |
| "grad_norm": 0.8775289058685303, | |
| "learning_rate": 1.2678963247394404e-05, | |
| "loss": 0.2868126106262207, | |
| "step": 127850 | |
| }, | |
| { | |
| "epoch": 2.621976219762198, | |
| "grad_norm": 0.8614349365234375, | |
| "learning_rate": 1.2644679100383982e-05, | |
| "loss": 0.2893955612182617, | |
| "step": 127900 | |
| }, | |
| { | |
| "epoch": 2.6230012300123002, | |
| "grad_norm": 0.9851211309432983, | |
| "learning_rate": 1.261039495337356e-05, | |
| "loss": 0.290030403137207, | |
| "step": 127950 | |
| }, | |
| { | |
| "epoch": 2.6240262402624026, | |
| "grad_norm": 0.9210476279258728, | |
| "learning_rate": 1.257611080636314e-05, | |
| "loss": 0.2892621421813965, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 2.625051250512505, | |
| "grad_norm": 0.9033406972885132, | |
| "learning_rate": 1.2541826659352715e-05, | |
| "loss": 0.28687347412109376, | |
| "step": 128050 | |
| }, | |
| { | |
| "epoch": 2.6260762607626074, | |
| "grad_norm": 0.8436753153800964, | |
| "learning_rate": 1.2507542512342293e-05, | |
| "loss": 0.28788496017456056, | |
| "step": 128100 | |
| }, | |
| { | |
| "epoch": 2.62710127101271, | |
| "grad_norm": 0.9875116348266602, | |
| "learning_rate": 1.2473258365331871e-05, | |
| "loss": 0.28867164611816404, | |
| "step": 128150 | |
| }, | |
| { | |
| "epoch": 2.6281262812628126, | |
| "grad_norm": 0.9269770979881287, | |
| "learning_rate": 1.2438974218321449e-05, | |
| "loss": 0.28845993041992185, | |
| "step": 128200 | |
| }, | |
| { | |
| "epoch": 2.6291512915129154, | |
| "grad_norm": 1.033141016960144, | |
| "learning_rate": 1.2404690071311026e-05, | |
| "loss": 0.289724063873291, | |
| "step": 128250 | |
| }, | |
| { | |
| "epoch": 2.6301763017630178, | |
| "grad_norm": 0.9881303310394287, | |
| "learning_rate": 1.2370405924300604e-05, | |
| "loss": 0.28294586181640624, | |
| "step": 128300 | |
| }, | |
| { | |
| "epoch": 2.63120131201312, | |
| "grad_norm": 0.9082476496696472, | |
| "learning_rate": 1.2336121777290182e-05, | |
| "loss": 0.2839402961730957, | |
| "step": 128350 | |
| }, | |
| { | |
| "epoch": 2.6322263222632225, | |
| "grad_norm": 0.8308191895484924, | |
| "learning_rate": 1.230183763027976e-05, | |
| "loss": 0.2906006622314453, | |
| "step": 128400 | |
| }, | |
| { | |
| "epoch": 2.633251332513325, | |
| "grad_norm": 1.0668548345565796, | |
| "learning_rate": 1.2267553483269336e-05, | |
| "loss": 0.2903538131713867, | |
| "step": 128450 | |
| }, | |
| { | |
| "epoch": 2.6342763427634277, | |
| "grad_norm": 0.9971005320549011, | |
| "learning_rate": 1.2233269336258916e-05, | |
| "loss": 0.2850461006164551, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 2.63530135301353, | |
| "grad_norm": 0.8975290656089783, | |
| "learning_rate": 1.2198985189248492e-05, | |
| "loss": 0.2886587905883789, | |
| "step": 128550 | |
| }, | |
| { | |
| "epoch": 2.636326363263633, | |
| "grad_norm": 0.9387692809104919, | |
| "learning_rate": 1.2164701042238069e-05, | |
| "loss": 0.2873807907104492, | |
| "step": 128600 | |
| }, | |
| { | |
| "epoch": 2.6373513735137353, | |
| "grad_norm": 0.8485702276229858, | |
| "learning_rate": 1.2130416895227647e-05, | |
| "loss": 0.2889641571044922, | |
| "step": 128650 | |
| }, | |
| { | |
| "epoch": 2.6383763837638377, | |
| "grad_norm": 0.925188422203064, | |
| "learning_rate": 1.2096132748217225e-05, | |
| "loss": 0.28785240173339843, | |
| "step": 128700 | |
| }, | |
| { | |
| "epoch": 2.63940139401394, | |
| "grad_norm": 0.868857741355896, | |
| "learning_rate": 1.2061848601206803e-05, | |
| "loss": 0.28929450988769534, | |
| "step": 128750 | |
| }, | |
| { | |
| "epoch": 2.6404264042640424, | |
| "grad_norm": 0.8896534442901611, | |
| "learning_rate": 1.202756445419638e-05, | |
| "loss": 0.28911289215087893, | |
| "step": 128800 | |
| }, | |
| { | |
| "epoch": 2.6414514145141452, | |
| "grad_norm": 0.9776099324226379, | |
| "learning_rate": 1.1993280307185959e-05, | |
| "loss": 0.2883522224426269, | |
| "step": 128850 | |
| }, | |
| { | |
| "epoch": 2.6424764247642476, | |
| "grad_norm": 0.9896165728569031, | |
| "learning_rate": 1.1958996160175536e-05, | |
| "loss": 0.2928770446777344, | |
| "step": 128900 | |
| }, | |
| { | |
| "epoch": 2.6435014350143504, | |
| "grad_norm": 0.9856406450271606, | |
| "learning_rate": 1.1924712013165112e-05, | |
| "loss": 0.28488271713256835, | |
| "step": 128950 | |
| }, | |
| { | |
| "epoch": 2.644526445264453, | |
| "grad_norm": 0.807998538017273, | |
| "learning_rate": 1.189042786615469e-05, | |
| "loss": 0.2859556770324707, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 2.645551455514555, | |
| "grad_norm": 0.8164665699005127, | |
| "learning_rate": 1.1856143719144268e-05, | |
| "loss": 0.28765356063842773, | |
| "step": 129050 | |
| }, | |
| { | |
| "epoch": 2.6465764657646575, | |
| "grad_norm": 0.863982617855072, | |
| "learning_rate": 1.1821859572133846e-05, | |
| "loss": 0.2849854850769043, | |
| "step": 129100 | |
| }, | |
| { | |
| "epoch": 2.64760147601476, | |
| "grad_norm": 0.9726262092590332, | |
| "learning_rate": 1.1787575425123423e-05, | |
| "loss": 0.28894098281860353, | |
| "step": 129150 | |
| }, | |
| { | |
| "epoch": 2.6486264862648627, | |
| "grad_norm": 0.9398086667060852, | |
| "learning_rate": 1.1753291278113002e-05, | |
| "loss": 0.28884761810302734, | |
| "step": 129200 | |
| }, | |
| { | |
| "epoch": 2.649651496514965, | |
| "grad_norm": 0.9247099161148071, | |
| "learning_rate": 1.1719007131102579e-05, | |
| "loss": 0.29109188079833986, | |
| "step": 129250 | |
| }, | |
| { | |
| "epoch": 2.6506765067650675, | |
| "grad_norm": 0.9300678372383118, | |
| "learning_rate": 1.1684722984092155e-05, | |
| "loss": 0.290528678894043, | |
| "step": 129300 | |
| }, | |
| { | |
| "epoch": 2.6517015170151703, | |
| "grad_norm": 0.8407086730003357, | |
| "learning_rate": 1.1650438837081735e-05, | |
| "loss": 0.28729839324951173, | |
| "step": 129350 | |
| }, | |
| { | |
| "epoch": 2.6527265272652727, | |
| "grad_norm": 0.9626300930976868, | |
| "learning_rate": 1.1616154690071311e-05, | |
| "loss": 0.28685842514038085, | |
| "step": 129400 | |
| }, | |
| { | |
| "epoch": 2.653751537515375, | |
| "grad_norm": 0.998223066329956, | |
| "learning_rate": 1.158187054306089e-05, | |
| "loss": 0.28632095336914065, | |
| "step": 129450 | |
| }, | |
| { | |
| "epoch": 2.6547765477654774, | |
| "grad_norm": 0.8346276879310608, | |
| "learning_rate": 1.1547586396050466e-05, | |
| "loss": 0.2858513641357422, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 2.6558015580155803, | |
| "grad_norm": 0.7909831404685974, | |
| "learning_rate": 1.1513302249040046e-05, | |
| "loss": 0.28778514862060545, | |
| "step": 129550 | |
| }, | |
| { | |
| "epoch": 2.6568265682656826, | |
| "grad_norm": 0.9689843058586121, | |
| "learning_rate": 1.1479018102029622e-05, | |
| "loss": 0.29156511306762695, | |
| "step": 129600 | |
| }, | |
| { | |
| "epoch": 2.657851578515785, | |
| "grad_norm": 0.9500753879547119, | |
| "learning_rate": 1.1444733955019198e-05, | |
| "loss": 0.2844500350952148, | |
| "step": 129650 | |
| }, | |
| { | |
| "epoch": 2.658876588765888, | |
| "grad_norm": 0.854477047920227, | |
| "learning_rate": 1.1410449808008778e-05, | |
| "loss": 0.28869091033935546, | |
| "step": 129700 | |
| }, | |
| { | |
| "epoch": 2.65990159901599, | |
| "grad_norm": 0.8713935017585754, | |
| "learning_rate": 1.1376165660998355e-05, | |
| "loss": 0.2917887115478516, | |
| "step": 129750 | |
| }, | |
| { | |
| "epoch": 2.6609266092660926, | |
| "grad_norm": 0.969587504863739, | |
| "learning_rate": 1.1341881513987933e-05, | |
| "loss": 0.29288694381713865, | |
| "step": 129800 | |
| }, | |
| { | |
| "epoch": 2.661951619516195, | |
| "grad_norm": 0.9154466390609741, | |
| "learning_rate": 1.1307597366977509e-05, | |
| "loss": 0.28700933456420896, | |
| "step": 129850 | |
| }, | |
| { | |
| "epoch": 2.662976629766298, | |
| "grad_norm": 0.8307504653930664, | |
| "learning_rate": 1.1273313219967089e-05, | |
| "loss": 0.28137033462524413, | |
| "step": 129900 | |
| }, | |
| { | |
| "epoch": 2.6640016400164, | |
| "grad_norm": 0.8211974501609802, | |
| "learning_rate": 1.1239029072956665e-05, | |
| "loss": 0.28781124114990236, | |
| "step": 129950 | |
| }, | |
| { | |
| "epoch": 2.6650266502665025, | |
| "grad_norm": 0.9595004320144653, | |
| "learning_rate": 1.1204744925946243e-05, | |
| "loss": 0.2841991424560547, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 2.6650266502665025, | |
| "eval_loss": 0.35195282101631165, | |
| "eval_runtime": 4.6154, | |
| "eval_samples_per_second": 5416.697, | |
| "eval_steps_per_second": 5.417, | |
| "eval_teacher_cosine_loss": 0.22212451696395874, | |
| "eval_teacher_cosine_mean": 0.7778754830360413, | |
| "eval_teacher_cosine_p10": 0.7017282247543335, | |
| "eval_teacher_cosine_p50": 0.788249135017395, | |
| "step": 130000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 146340, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 512, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |
Xet Storage Details
- Size:
- 499 kB
- Xet hash:
- a6d6be16cbffa04750549feb3ec62f90c58f0dbe55a144737230d9eb6b2f2c0b
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.