diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,4697 @@ +{ + "best_global_step": 6500, + "best_metric": 0.8433617949485779, + "best_model_checkpoint": "outputs/vlm-age-rating-qwen25vl/checkpoint-6500", + "epoch": 1.8900843268391974, + "eval_steps": 500, + "global_step": 6500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002907822041291073, + "grad_norm": 4.79532527923584, + "learning_rate": 8.695652173913044e-06, + "loss": 2.4354520797729493, + "step": 10 + }, + { + "epoch": 0.005815644082582146, + "grad_norm": 4.398916721343994, + "learning_rate": 1.8357487922705315e-05, + "loss": 2.348348045349121, + "step": 20 + }, + { + "epoch": 0.008723466123873218, + "grad_norm": 3.8363561630249023, + "learning_rate": 2.8019323671497587e-05, + "loss": 2.2402442932128905, + "step": 30 + }, + { + "epoch": 0.011631288165164292, + "grad_norm": 3.9996328353881836, + "learning_rate": 3.7681159420289856e-05, + "loss": 2.0806962966918947, + "step": 40 + }, + { + "epoch": 0.014539110206455364, + "grad_norm": 3.623289108276367, + "learning_rate": 4.7342995169082125e-05, + "loss": 1.9209724426269532, + "step": 50 + }, + { + "epoch": 0.017446932247746436, + "grad_norm": 3.6283390522003174, + "learning_rate": 5.7004830917874394e-05, + "loss": 1.8231483459472657, + "step": 60 + }, + { + "epoch": 0.020354754289037512, + "grad_norm": 3.6183929443359375, + "learning_rate": 6.666666666666667e-05, + "loss": 1.7064151763916016, + "step": 70 + }, + { + "epoch": 0.023262576330328584, + "grad_norm": 4.0795063972473145, + "learning_rate": 7.632850241545893e-05, + "loss": 1.6121023178100586, + "step": 80 + }, + { + "epoch": 0.026170398371619656, + "grad_norm": 3.4447927474975586, + "learning_rate": 8.599033816425122e-05, + "loss": 1.6030376434326172, + "step": 90 + }, + { + "epoch": 0.02907822041291073, + "grad_norm": 3.6108040809631348, + "learning_rate": 9.565217391304348e-05, + "loss": 1.5841608047485352, + "step": 100 + }, + { + "epoch": 0.0319860424542018, + "grad_norm": 3.598888635635376, + "learning_rate": 0.00010531400966183576, + "loss": 1.4951010704040528, + "step": 110 + }, + { + "epoch": 0.03489386449549287, + "grad_norm": 3.70912766456604, + "learning_rate": 0.00011497584541062802, + "loss": 1.482225227355957, + "step": 120 + }, + { + "epoch": 0.03780168653678395, + "grad_norm": 3.4434823989868164, + "learning_rate": 0.0001246376811594203, + "loss": 1.5170270919799804, + "step": 130 + }, + { + "epoch": 0.040709508578075024, + "grad_norm": 3.361293315887451, + "learning_rate": 0.00013429951690821257, + "loss": 1.4422802925109863, + "step": 140 + }, + { + "epoch": 0.043617330619366096, + "grad_norm": 3.285003900527954, + "learning_rate": 0.00014396135265700482, + "loss": 1.4413420677185058, + "step": 150 + }, + { + "epoch": 0.04652515266065717, + "grad_norm": 3.6356050968170166, + "learning_rate": 0.0001536231884057971, + "loss": 1.4118947982788086, + "step": 160 + }, + { + "epoch": 0.04943297470194824, + "grad_norm": 3.8388562202453613, + "learning_rate": 0.00016328502415458937, + "loss": 1.4001873970031737, + "step": 170 + }, + { + "epoch": 0.05234079674323931, + "grad_norm": 3.3433616161346436, + "learning_rate": 0.00017294685990338165, + "loss": 1.3969547271728515, + "step": 180 + }, + { + "epoch": 0.055248618784530384, + "grad_norm": 3.4669270515441895, + "learning_rate": 0.00018260869565217392, + "loss": 1.337031650543213, + "step": 190 + }, + { + "epoch": 0.05815644082582146, + "grad_norm": 3.1025021076202393, + "learning_rate": 0.0001922705314009662, + "loss": 1.435082244873047, + "step": 200 + }, + { + "epoch": 0.061064262867112536, + "grad_norm": 3.3911843299865723, + "learning_rate": 0.00019994003897466647, + "loss": 1.392270278930664, + "step": 210 + }, + { + "epoch": 0.0639720849084036, + "grad_norm": 3.142547130584717, + "learning_rate": 0.0001996402338479988, + "loss": 1.243046760559082, + "step": 220 + }, + { + "epoch": 0.06687990694969467, + "grad_norm": 3.35446834564209, + "learning_rate": 0.00019934042872133114, + "loss": 1.367037868499756, + "step": 230 + }, + { + "epoch": 0.06978772899098575, + "grad_norm": 3.458134651184082, + "learning_rate": 0.0001990406235946635, + "loss": 1.3419992446899414, + "step": 240 + }, + { + "epoch": 0.07269555103227683, + "grad_norm": 3.352386474609375, + "learning_rate": 0.0001987408184679958, + "loss": 1.2986759185791015, + "step": 250 + }, + { + "epoch": 0.0756033730735679, + "grad_norm": 3.1780903339385986, + "learning_rate": 0.00019844101334132814, + "loss": 1.2988534927368165, + "step": 260 + }, + { + "epoch": 0.07851119511485898, + "grad_norm": 2.86576247215271, + "learning_rate": 0.00019814120821466048, + "loss": 1.3682787895202637, + "step": 270 + }, + { + "epoch": 0.08141901715615005, + "grad_norm": 3.3476641178131104, + "learning_rate": 0.0001978414030879928, + "loss": 1.3652713775634766, + "step": 280 + }, + { + "epoch": 0.08432683919744112, + "grad_norm": 3.2246811389923096, + "learning_rate": 0.00019754159796132515, + "loss": 1.2679337501525878, + "step": 290 + }, + { + "epoch": 0.08723466123873219, + "grad_norm": 3.134486675262451, + "learning_rate": 0.00019724179283465748, + "loss": 1.211684513092041, + "step": 300 + }, + { + "epoch": 0.09014248328002326, + "grad_norm": 3.142294406890869, + "learning_rate": 0.00019694198770798982, + "loss": 1.3083333015441894, + "step": 310 + }, + { + "epoch": 0.09305030532131434, + "grad_norm": 3.2110912799835205, + "learning_rate": 0.00019664218258132215, + "loss": 1.2491901397705079, + "step": 320 + }, + { + "epoch": 0.09595812736260541, + "grad_norm": 3.0253570079803467, + "learning_rate": 0.00019634237745465449, + "loss": 1.3311482429504395, + "step": 330 + }, + { + "epoch": 0.09886594940389648, + "grad_norm": 2.7756271362304688, + "learning_rate": 0.00019604257232798682, + "loss": 1.2524259567260743, + "step": 340 + }, + { + "epoch": 0.10177377144518755, + "grad_norm": 3.243698835372925, + "learning_rate": 0.00019574276720131915, + "loss": 1.293147087097168, + "step": 350 + }, + { + "epoch": 0.10468159348647862, + "grad_norm": 2.8504862785339355, + "learning_rate": 0.0001954429620746515, + "loss": 1.2383965492248534, + "step": 360 + }, + { + "epoch": 0.1075894155277697, + "grad_norm": 3.280348062515259, + "learning_rate": 0.00019514315694798382, + "loss": 1.2218168258666993, + "step": 370 + }, + { + "epoch": 0.11049723756906077, + "grad_norm": 3.1593716144561768, + "learning_rate": 0.00019484335182131616, + "loss": 1.2068678855895996, + "step": 380 + }, + { + "epoch": 0.11340505961035184, + "grad_norm": 3.1796255111694336, + "learning_rate": 0.0001945435466946485, + "loss": 1.2717905998229981, + "step": 390 + }, + { + "epoch": 0.11631288165164291, + "grad_norm": 2.8659069538116455, + "learning_rate": 0.00019424374156798083, + "loss": 1.2112377166748047, + "step": 400 + }, + { + "epoch": 0.119220703692934, + "grad_norm": 2.9140565395355225, + "learning_rate": 0.00019394393644131313, + "loss": 1.2414008140563966, + "step": 410 + }, + { + "epoch": 0.12212852573422507, + "grad_norm": 3.0246622562408447, + "learning_rate": 0.0001936441313146455, + "loss": 1.2327194213867188, + "step": 420 + }, + { + "epoch": 0.12503634777551614, + "grad_norm": 2.7067604064941406, + "learning_rate": 0.00019334432618797783, + "loss": 1.1890941619873048, + "step": 430 + }, + { + "epoch": 0.1279441698168072, + "grad_norm": 3.0323450565338135, + "learning_rate": 0.00019304452106131016, + "loss": 1.2328312873840332, + "step": 440 + }, + { + "epoch": 0.1308519918580983, + "grad_norm": 3.0487864017486572, + "learning_rate": 0.0001927447159346425, + "loss": 1.1881980895996094, + "step": 450 + }, + { + "epoch": 0.13375981389938935, + "grad_norm": 3.0193893909454346, + "learning_rate": 0.00019244491080797483, + "loss": 1.2217585563659668, + "step": 460 + }, + { + "epoch": 0.13666763594068043, + "grad_norm": 3.0121376514434814, + "learning_rate": 0.00019214510568130717, + "loss": 1.1492167472839356, + "step": 470 + }, + { + "epoch": 0.1395754579819715, + "grad_norm": 2.9854133129119873, + "learning_rate": 0.0001918453005546395, + "loss": 1.2008344650268554, + "step": 480 + }, + { + "epoch": 0.14248328002326258, + "grad_norm": 2.982191562652588, + "learning_rate": 0.0001915454954279718, + "loss": 1.1686691284179687, + "step": 490 + }, + { + "epoch": 0.14539110206455366, + "grad_norm": 4.596578598022461, + "learning_rate": 0.00019124569030130417, + "loss": 1.2213269233703614, + "step": 500 + }, + { + "epoch": 0.14539110206455366, + "eval_loss": 1.1948493719100952, + "eval_runtime": 471.9433, + "eval_samples_per_second": 7.287, + "eval_steps_per_second": 7.287, + "step": 500 + }, + { + "epoch": 0.14829892410584472, + "grad_norm": 2.8607027530670166, + "learning_rate": 0.0001909458851746365, + "loss": 1.2157353401184081, + "step": 510 + }, + { + "epoch": 0.1512067461471358, + "grad_norm": 3.3106703758239746, + "learning_rate": 0.0001906460800479688, + "loss": 1.2436732292175292, + "step": 520 + }, + { + "epoch": 0.15411456818842686, + "grad_norm": 2.8542511463165283, + "learning_rate": 0.00019034627492130118, + "loss": 1.2832223892211914, + "step": 530 + }, + { + "epoch": 0.15702239022971795, + "grad_norm": 2.961954355239868, + "learning_rate": 0.0001900464697946335, + "loss": 1.214921474456787, + "step": 540 + }, + { + "epoch": 0.159930212271009, + "grad_norm": 3.2760446071624756, + "learning_rate": 0.00018974666466796582, + "loss": 1.2085216522216797, + "step": 550 + }, + { + "epoch": 0.1628380343123001, + "grad_norm": 3.117536783218384, + "learning_rate": 0.00018944685954129818, + "loss": 1.1909247398376466, + "step": 560 + }, + { + "epoch": 0.16574585635359115, + "grad_norm": 3.1558895111083984, + "learning_rate": 0.00018914705441463049, + "loss": 1.2084429740905762, + "step": 570 + }, + { + "epoch": 0.16865367839488224, + "grad_norm": 2.6539628505706787, + "learning_rate": 0.00018884724928796282, + "loss": 1.2048931121826172, + "step": 580 + }, + { + "epoch": 0.1715615004361733, + "grad_norm": 2.4215893745422363, + "learning_rate": 0.00018854744416129518, + "loss": 1.1528879165649415, + "step": 590 + }, + { + "epoch": 0.17446932247746438, + "grad_norm": 3.247122049331665, + "learning_rate": 0.0001882476390346275, + "loss": 1.1167863845825194, + "step": 600 + }, + { + "epoch": 0.17737714451875544, + "grad_norm": 2.5841548442840576, + "learning_rate": 0.00018794783390795982, + "loss": 1.1510201454162599, + "step": 610 + }, + { + "epoch": 0.18028496656004653, + "grad_norm": 2.8135194778442383, + "learning_rate": 0.00018764802878129219, + "loss": 1.1705083847045898, + "step": 620 + }, + { + "epoch": 0.1831927886013376, + "grad_norm": 2.7703614234924316, + "learning_rate": 0.0001873482236546245, + "loss": 1.1498478889465331, + "step": 630 + }, + { + "epoch": 0.18610061064262867, + "grad_norm": 2.933802843093872, + "learning_rate": 0.00018704841852795685, + "loss": 1.1263715744018554, + "step": 640 + }, + { + "epoch": 0.18900843268391973, + "grad_norm": 2.8813698291778564, + "learning_rate": 0.00018674861340128916, + "loss": 1.0713846206665039, + "step": 650 + }, + { + "epoch": 0.19191625472521082, + "grad_norm": 4.469653606414795, + "learning_rate": 0.0001864488082746215, + "loss": 1.1526763916015625, + "step": 660 + }, + { + "epoch": 0.1948240767665019, + "grad_norm": 2.608485698699951, + "learning_rate": 0.00018614900314795386, + "loss": 1.1115928649902345, + "step": 670 + }, + { + "epoch": 0.19773189880779296, + "grad_norm": 2.5391645431518555, + "learning_rate": 0.00018584919802128617, + "loss": 1.1768034934997558, + "step": 680 + }, + { + "epoch": 0.20063972084908405, + "grad_norm": 2.8083078861236572, + "learning_rate": 0.0001855493928946185, + "loss": 1.1693111419677735, + "step": 690 + }, + { + "epoch": 0.2035475428903751, + "grad_norm": 2.7247753143310547, + "learning_rate": 0.00018524958776795086, + "loss": 1.1793179512023926, + "step": 700 + }, + { + "epoch": 0.2064553649316662, + "grad_norm": 2.5489275455474854, + "learning_rate": 0.00018494978264128317, + "loss": 1.153579044342041, + "step": 710 + }, + { + "epoch": 0.20936318697295725, + "grad_norm": 2.848568916320801, + "learning_rate": 0.0001846499775146155, + "loss": 1.1616141319274902, + "step": 720 + }, + { + "epoch": 0.21227100901424834, + "grad_norm": 2.6661322116851807, + "learning_rate": 0.00018435017238794784, + "loss": 1.185004997253418, + "step": 730 + }, + { + "epoch": 0.2151788310555394, + "grad_norm": 2.6399037837982178, + "learning_rate": 0.00018405036726128017, + "loss": 1.1467523574829102, + "step": 740 + }, + { + "epoch": 0.21808665309683048, + "grad_norm": 2.8910253047943115, + "learning_rate": 0.0001837505621346125, + "loss": 1.1423175811767579, + "step": 750 + }, + { + "epoch": 0.22099447513812154, + "grad_norm": 2.6750359535217285, + "learning_rate": 0.00018345075700794484, + "loss": 1.1023540496826172, + "step": 760 + }, + { + "epoch": 0.22390229717941262, + "grad_norm": 2.7184813022613525, + "learning_rate": 0.00018315095188127718, + "loss": 1.0655659675598144, + "step": 770 + }, + { + "epoch": 0.22681011922070368, + "grad_norm": 2.352343797683716, + "learning_rate": 0.0001828511467546095, + "loss": 1.0686527252197267, + "step": 780 + }, + { + "epoch": 0.22971794126199477, + "grad_norm": 2.9609851837158203, + "learning_rate": 0.00018255134162794185, + "loss": 1.0448270797729493, + "step": 790 + }, + { + "epoch": 0.23262576330328583, + "grad_norm": 2.629925489425659, + "learning_rate": 0.00018225153650127418, + "loss": 1.2246409416198731, + "step": 800 + }, + { + "epoch": 0.2355335853445769, + "grad_norm": 2.443338632583618, + "learning_rate": 0.00018195173137460651, + "loss": 1.0720739364624023, + "step": 810 + }, + { + "epoch": 0.238441407385868, + "grad_norm": 2.612340211868286, + "learning_rate": 0.00018165192624793885, + "loss": 1.135690975189209, + "step": 820 + }, + { + "epoch": 0.24134922942715906, + "grad_norm": 2.6910789012908936, + "learning_rate": 0.00018135212112127118, + "loss": 1.160903263092041, + "step": 830 + }, + { + "epoch": 0.24425705146845014, + "grad_norm": 2.729325532913208, + "learning_rate": 0.00018105231599460352, + "loss": 1.1131114959716797, + "step": 840 + }, + { + "epoch": 0.2471648735097412, + "grad_norm": 2.6346912384033203, + "learning_rate": 0.00018075251086793585, + "loss": 1.1584989547729492, + "step": 850 + }, + { + "epoch": 0.2500726955510323, + "grad_norm": 2.4764983654022217, + "learning_rate": 0.0001804527057412682, + "loss": 1.1217921257019043, + "step": 860 + }, + { + "epoch": 0.2529805175923234, + "grad_norm": 2.8104140758514404, + "learning_rate": 0.00018015290061460052, + "loss": 1.1900800704956054, + "step": 870 + }, + { + "epoch": 0.2558883396336144, + "grad_norm": 2.6423113346099854, + "learning_rate": 0.00017985309548793286, + "loss": 1.1654984474182128, + "step": 880 + }, + { + "epoch": 0.2587961616749055, + "grad_norm": 2.456171751022339, + "learning_rate": 0.0001795532903612652, + "loss": 1.1631184577941895, + "step": 890 + }, + { + "epoch": 0.2617039837161966, + "grad_norm": 2.355860471725464, + "learning_rate": 0.00017925348523459752, + "loss": 1.1408035278320312, + "step": 900 + }, + { + "epoch": 0.26461180575748766, + "grad_norm": 2.7640347480773926, + "learning_rate": 0.00017895368010792986, + "loss": 1.095857810974121, + "step": 910 + }, + { + "epoch": 0.2675196277987787, + "grad_norm": 2.518118143081665, + "learning_rate": 0.0001786538749812622, + "loss": 1.1492120742797851, + "step": 920 + }, + { + "epoch": 0.2704274498400698, + "grad_norm": 2.478942632675171, + "learning_rate": 0.00017835406985459453, + "loss": 1.095020580291748, + "step": 930 + }, + { + "epoch": 0.27333527188136086, + "grad_norm": 2.5483644008636475, + "learning_rate": 0.00017805426472792686, + "loss": 1.1031085014343263, + "step": 940 + }, + { + "epoch": 0.27624309392265195, + "grad_norm": 3.0256600379943848, + "learning_rate": 0.00017775445960125917, + "loss": 1.1728601455688477, + "step": 950 + }, + { + "epoch": 0.279150915963943, + "grad_norm": 2.417307138442993, + "learning_rate": 0.00017745465447459153, + "loss": 1.1077165603637695, + "step": 960 + }, + { + "epoch": 0.28205873800523407, + "grad_norm": 2.5772206783294678, + "learning_rate": 0.00017715484934792387, + "loss": 1.1050203323364258, + "step": 970 + }, + { + "epoch": 0.28496656004652515, + "grad_norm": 2.4329328536987305, + "learning_rate": 0.00017685504422125617, + "loss": 1.0694819450378419, + "step": 980 + }, + { + "epoch": 0.28787438208781624, + "grad_norm": 2.2022688388824463, + "learning_rate": 0.00017655523909458854, + "loss": 1.0653936386108398, + "step": 990 + }, + { + "epoch": 0.2907822041291073, + "grad_norm": 2.5784685611724854, + "learning_rate": 0.00017625543396792087, + "loss": 1.1188321113586426, + "step": 1000 + }, + { + "epoch": 0.2907822041291073, + "eval_loss": 1.088085651397705, + "eval_runtime": 466.4025, + "eval_samples_per_second": 7.373, + "eval_steps_per_second": 7.373, + "step": 1000 + }, + { + "epoch": 0.29369002617039836, + "grad_norm": 2.1797893047332764, + "learning_rate": 0.00017595562884125318, + "loss": 1.0641416549682616, + "step": 1010 + }, + { + "epoch": 0.29659784821168944, + "grad_norm": 2.351658344268799, + "learning_rate": 0.00017565582371458554, + "loss": 1.0726565361022948, + "step": 1020 + }, + { + "epoch": 0.2995056702529805, + "grad_norm": 2.5299222469329834, + "learning_rate": 0.00017535601858791785, + "loss": 1.110872459411621, + "step": 1030 + }, + { + "epoch": 0.3024134922942716, + "grad_norm": 2.492405652999878, + "learning_rate": 0.00017505621346125018, + "loss": 1.0856511116027832, + "step": 1040 + }, + { + "epoch": 0.30532131433556264, + "grad_norm": 2.490410089492798, + "learning_rate": 0.00017475640833458254, + "loss": 1.053286075592041, + "step": 1050 + }, + { + "epoch": 0.30822913637685373, + "grad_norm": 2.451176404953003, + "learning_rate": 0.00017445660320791485, + "loss": 1.1099212646484375, + "step": 1060 + }, + { + "epoch": 0.3111369584181448, + "grad_norm": 2.2699105739593506, + "learning_rate": 0.0001741567980812472, + "loss": 0.9613182067871093, + "step": 1070 + }, + { + "epoch": 0.3140447804594359, + "grad_norm": 2.1652886867523193, + "learning_rate": 0.00017385699295457955, + "loss": 1.0921488761901856, + "step": 1080 + }, + { + "epoch": 0.31695260250072693, + "grad_norm": 2.385770797729492, + "learning_rate": 0.00017355718782791185, + "loss": 1.1041201591491698, + "step": 1090 + }, + { + "epoch": 0.319860424542018, + "grad_norm": 2.5070347785949707, + "learning_rate": 0.00017325738270124421, + "loss": 1.0626919746398926, + "step": 1100 + }, + { + "epoch": 0.3227682465833091, + "grad_norm": 2.5580434799194336, + "learning_rate": 0.00017295757757457652, + "loss": 1.0355568885803224, + "step": 1110 + }, + { + "epoch": 0.3256760686246002, + "grad_norm": 2.285900592803955, + "learning_rate": 0.00017265777244790886, + "loss": 1.094072151184082, + "step": 1120 + }, + { + "epoch": 0.3285838906658912, + "grad_norm": 2.2862985134124756, + "learning_rate": 0.00017235796732124122, + "loss": 1.1321526527404786, + "step": 1130 + }, + { + "epoch": 0.3314917127071823, + "grad_norm": 2.3698503971099854, + "learning_rate": 0.00017205816219457353, + "loss": 1.1155936241149902, + "step": 1140 + }, + { + "epoch": 0.3343995347484734, + "grad_norm": 2.2256312370300293, + "learning_rate": 0.00017175835706790586, + "loss": 1.0627290725708007, + "step": 1150 + }, + { + "epoch": 0.3373073567897645, + "grad_norm": 2.3896291255950928, + "learning_rate": 0.00017145855194123822, + "loss": 1.0547872543334962, + "step": 1160 + }, + { + "epoch": 0.34021517883105556, + "grad_norm": 2.1930463314056396, + "learning_rate": 0.00017115874681457053, + "loss": 1.0243175506591797, + "step": 1170 + }, + { + "epoch": 0.3431230008723466, + "grad_norm": 2.0768635272979736, + "learning_rate": 0.00017085894168790286, + "loss": 1.1152023315429687, + "step": 1180 + }, + { + "epoch": 0.3460308229136377, + "grad_norm": 2.179349422454834, + "learning_rate": 0.0001705591365612352, + "loss": 1.1170848846435546, + "step": 1190 + }, + { + "epoch": 0.34893864495492877, + "grad_norm": 2.2244808673858643, + "learning_rate": 0.00017025933143456753, + "loss": 0.964411735534668, + "step": 1200 + }, + { + "epoch": 0.35184646699621985, + "grad_norm": 2.39132022857666, + "learning_rate": 0.00016995952630789987, + "loss": 1.0051309585571289, + "step": 1210 + }, + { + "epoch": 0.3547542890375109, + "grad_norm": 2.1408185958862305, + "learning_rate": 0.0001696597211812322, + "loss": 1.1022598266601562, + "step": 1220 + }, + { + "epoch": 0.35766211107880197, + "grad_norm": 2.3732504844665527, + "learning_rate": 0.00016935991605456454, + "loss": 1.0292579650878906, + "step": 1230 + }, + { + "epoch": 0.36056993312009306, + "grad_norm": 2.5366053581237793, + "learning_rate": 0.00016906011092789687, + "loss": 1.0912357330322267, + "step": 1240 + }, + { + "epoch": 0.36347775516138414, + "grad_norm": 2.2400059700012207, + "learning_rate": 0.0001687603058012292, + "loss": 1.033323383331299, + "step": 1250 + }, + { + "epoch": 0.3663855772026752, + "grad_norm": 2.2703261375427246, + "learning_rate": 0.00016846050067456154, + "loss": 1.0109487533569337, + "step": 1260 + }, + { + "epoch": 0.36929339924396626, + "grad_norm": 2.280935764312744, + "learning_rate": 0.00016816069554789387, + "loss": 1.0813608169555664, + "step": 1270 + }, + { + "epoch": 0.37220122128525734, + "grad_norm": 2.168682098388672, + "learning_rate": 0.0001678608904212262, + "loss": 1.0790273666381835, + "step": 1280 + }, + { + "epoch": 0.37510904332654843, + "grad_norm": 2.1177940368652344, + "learning_rate": 0.00016756108529455854, + "loss": 1.031337356567383, + "step": 1290 + }, + { + "epoch": 0.37801686536783946, + "grad_norm": 2.4730405807495117, + "learning_rate": 0.00016726128016789088, + "loss": 1.129223918914795, + "step": 1300 + }, + { + "epoch": 0.38092468740913055, + "grad_norm": 2.095201015472412, + "learning_rate": 0.0001669614750412232, + "loss": 1.097676658630371, + "step": 1310 + }, + { + "epoch": 0.38383250945042163, + "grad_norm": 5.1267242431640625, + "learning_rate": 0.00016666166991455555, + "loss": 1.0067487716674806, + "step": 1320 + }, + { + "epoch": 0.3867403314917127, + "grad_norm": 2.3142173290252686, + "learning_rate": 0.00016636186478788788, + "loss": 1.0111748695373535, + "step": 1330 + }, + { + "epoch": 0.3896481535330038, + "grad_norm": 2.5803937911987305, + "learning_rate": 0.00016606205966122022, + "loss": 1.0414213180541991, + "step": 1340 + }, + { + "epoch": 0.39255597557429484, + "grad_norm": 2.057889223098755, + "learning_rate": 0.00016576225453455255, + "loss": 1.0709516525268554, + "step": 1350 + }, + { + "epoch": 0.3954637976155859, + "grad_norm": 2.221109628677368, + "learning_rate": 0.00016546244940788488, + "loss": 0.9829123497009278, + "step": 1360 + }, + { + "epoch": 0.398371619656877, + "grad_norm": 2.2963709831237793, + "learning_rate": 0.00016516264428121722, + "loss": 1.0530911445617677, + "step": 1370 + }, + { + "epoch": 0.4012794416981681, + "grad_norm": 2.225609064102173, + "learning_rate": 0.00016486283915454955, + "loss": 0.9976913452148437, + "step": 1380 + }, + { + "epoch": 0.4041872637394591, + "grad_norm": 2.186084032058716, + "learning_rate": 0.0001645630340278819, + "loss": 0.9746930122375488, + "step": 1390 + }, + { + "epoch": 0.4070950857807502, + "grad_norm": 2.3998475074768066, + "learning_rate": 0.00016426322890121422, + "loss": 0.995127010345459, + "step": 1400 + }, + { + "epoch": 0.4100029078220413, + "grad_norm": 2.2103660106658936, + "learning_rate": 0.00016396342377454653, + "loss": 0.9968692779541015, + "step": 1410 + }, + { + "epoch": 0.4129107298633324, + "grad_norm": 2.228457450866699, + "learning_rate": 0.0001636636186478789, + "loss": 1.0345491409301757, + "step": 1420 + }, + { + "epoch": 0.4158185519046234, + "grad_norm": 2.2369489669799805, + "learning_rate": 0.00016336381352121123, + "loss": 1.0916296005249024, + "step": 1430 + }, + { + "epoch": 0.4187263739459145, + "grad_norm": 2.4540905952453613, + "learning_rate": 0.00016306400839454353, + "loss": 1.0946255683898927, + "step": 1440 + }, + { + "epoch": 0.4216341959872056, + "grad_norm": 2.272212505340576, + "learning_rate": 0.0001627642032678759, + "loss": 1.0748573303222657, + "step": 1450 + }, + { + "epoch": 0.42454201802849667, + "grad_norm": 2.1080758571624756, + "learning_rate": 0.00016246439814120823, + "loss": 1.0456165313720702, + "step": 1460 + }, + { + "epoch": 0.42744984006978776, + "grad_norm": 2.1176912784576416, + "learning_rate": 0.00016216459301454056, + "loss": 0.9982593536376954, + "step": 1470 + }, + { + "epoch": 0.4303576621110788, + "grad_norm": 2.0624117851257324, + "learning_rate": 0.0001618647878878729, + "loss": 1.067337417602539, + "step": 1480 + }, + { + "epoch": 0.4332654841523699, + "grad_norm": 2.2433207035064697, + "learning_rate": 0.0001615649827612052, + "loss": 1.1099421501159668, + "step": 1490 + }, + { + "epoch": 0.43617330619366096, + "grad_norm": 2.0459365844726562, + "learning_rate": 0.00016126517763453757, + "loss": 1.0837010383605956, + "step": 1500 + }, + { + "epoch": 0.43617330619366096, + "eval_loss": 1.0330508947372437, + "eval_runtime": 467.9014, + "eval_samples_per_second": 7.35, + "eval_steps_per_second": 7.35, + "step": 1500 + }, + { + "epoch": 0.43908112823495205, + "grad_norm": 1.9743572473526, + "learning_rate": 0.0001609653725078699, + "loss": 0.9442779541015625, + "step": 1510 + }, + { + "epoch": 0.4419889502762431, + "grad_norm": 2.259799003601074, + "learning_rate": 0.0001606655673812022, + "loss": 1.0716489791870116, + "step": 1520 + }, + { + "epoch": 0.44489677231753416, + "grad_norm": 2.1605937480926514, + "learning_rate": 0.00016036576225453457, + "loss": 1.0564517974853516, + "step": 1530 + }, + { + "epoch": 0.44780459435882525, + "grad_norm": 2.2314250469207764, + "learning_rate": 0.0001600659571278669, + "loss": 1.0048583030700684, + "step": 1540 + }, + { + "epoch": 0.45071241640011633, + "grad_norm": 2.099571943283081, + "learning_rate": 0.0001597661520011992, + "loss": 1.0006441116333007, + "step": 1550 + }, + { + "epoch": 0.45362023844140736, + "grad_norm": 1.9105018377304077, + "learning_rate": 0.00015946634687453157, + "loss": 0.9975058555603027, + "step": 1560 + }, + { + "epoch": 0.45652806048269845, + "grad_norm": 2.057384729385376, + "learning_rate": 0.00015916654174786388, + "loss": 1.087107276916504, + "step": 1570 + }, + { + "epoch": 0.45943588252398954, + "grad_norm": 1.9773719310760498, + "learning_rate": 0.00015886673662119622, + "loss": 1.0378877639770507, + "step": 1580 + }, + { + "epoch": 0.4623437045652806, + "grad_norm": 2.122605323791504, + "learning_rate": 0.00015856693149452858, + "loss": 1.0330044746398925, + "step": 1590 + }, + { + "epoch": 0.46525152660657165, + "grad_norm": 2.409942388534546, + "learning_rate": 0.00015826712636786089, + "loss": 1.0367450714111328, + "step": 1600 + }, + { + "epoch": 0.46815934864786274, + "grad_norm": 2.112682342529297, + "learning_rate": 0.00015796732124119322, + "loss": 1.0601228713989257, + "step": 1610 + }, + { + "epoch": 0.4710671706891538, + "grad_norm": 2.8438005447387695, + "learning_rate": 0.00015766751611452558, + "loss": 1.013437271118164, + "step": 1620 + }, + { + "epoch": 0.4739749927304449, + "grad_norm": 2.1778664588928223, + "learning_rate": 0.0001573677109878579, + "loss": 0.9827444076538085, + "step": 1630 + }, + { + "epoch": 0.476882814771736, + "grad_norm": 2.261019468307495, + "learning_rate": 0.00015706790586119022, + "loss": 1.019342803955078, + "step": 1640 + }, + { + "epoch": 0.479790636813027, + "grad_norm": 2.01497483253479, + "learning_rate": 0.00015676810073452258, + "loss": 0.9160719871520996, + "step": 1650 + }, + { + "epoch": 0.4826984588543181, + "grad_norm": 2.2591161727905273, + "learning_rate": 0.0001564682956078549, + "loss": 0.9794828414916992, + "step": 1660 + }, + { + "epoch": 0.4856062808956092, + "grad_norm": 2.303205966949463, + "learning_rate": 0.00015616849048118725, + "loss": 1.070561408996582, + "step": 1670 + }, + { + "epoch": 0.4885141029369003, + "grad_norm": 2.075136661529541, + "learning_rate": 0.00015586868535451956, + "loss": 1.0001187324523926, + "step": 1680 + }, + { + "epoch": 0.4914219249781913, + "grad_norm": 2.0339841842651367, + "learning_rate": 0.0001555688802278519, + "loss": 1.042809009552002, + "step": 1690 + }, + { + "epoch": 0.4943297470194824, + "grad_norm": 2.23016357421875, + "learning_rate": 0.00015526907510118426, + "loss": 1.0348269462585449, + "step": 1700 + }, + { + "epoch": 0.4972375690607735, + "grad_norm": 2.620288372039795, + "learning_rate": 0.00015496926997451656, + "loss": 0.9250534057617188, + "step": 1710 + }, + { + "epoch": 0.5001453911020646, + "grad_norm": 2.0876739025115967, + "learning_rate": 0.0001546694648478489, + "loss": 1.0507851600646974, + "step": 1720 + }, + { + "epoch": 0.5030532131433556, + "grad_norm": 2.0192840099334717, + "learning_rate": 0.00015436965972118126, + "loss": 0.9346181869506835, + "step": 1730 + }, + { + "epoch": 0.5059610351846467, + "grad_norm": 2.198378086090088, + "learning_rate": 0.00015406985459451357, + "loss": 1.008394145965576, + "step": 1740 + }, + { + "epoch": 0.5088688572259378, + "grad_norm": 2.028977394104004, + "learning_rate": 0.0001537700494678459, + "loss": 1.0137055397033692, + "step": 1750 + }, + { + "epoch": 0.5117766792672288, + "grad_norm": 2.091458320617676, + "learning_rate": 0.00015347024434117824, + "loss": 0.9982250213623047, + "step": 1760 + }, + { + "epoch": 0.51468450130852, + "grad_norm": 1.993230938911438, + "learning_rate": 0.00015317043921451057, + "loss": 0.9881864547729492, + "step": 1770 + }, + { + "epoch": 0.517592323349811, + "grad_norm": 1.9999781847000122, + "learning_rate": 0.0001528706340878429, + "loss": 0.9195417404174805, + "step": 1780 + }, + { + "epoch": 0.520500145391102, + "grad_norm": 2.0336172580718994, + "learning_rate": 0.00015257082896117524, + "loss": 1.040501880645752, + "step": 1790 + }, + { + "epoch": 0.5234079674323932, + "grad_norm": 2.3573317527770996, + "learning_rate": 0.00015227102383450758, + "loss": 1.0388559341430663, + "step": 1800 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 2.2667224407196045, + "learning_rate": 0.0001519712187078399, + "loss": 1.0280171394348145, + "step": 1810 + }, + { + "epoch": 0.5292236115149753, + "grad_norm": 2.0019094944000244, + "learning_rate": 0.00015167141358117224, + "loss": 1.054083251953125, + "step": 1820 + }, + { + "epoch": 0.5321314335562664, + "grad_norm": 2.1444737911224365, + "learning_rate": 0.00015137160845450458, + "loss": 1.0175910949707032, + "step": 1830 + }, + { + "epoch": 0.5350392555975574, + "grad_norm": 2.2335095405578613, + "learning_rate": 0.0001510718033278369, + "loss": 1.0406922340393066, + "step": 1840 + }, + { + "epoch": 0.5379470776388485, + "grad_norm": 1.984129786491394, + "learning_rate": 0.00015077199820116925, + "loss": 0.9368982315063477, + "step": 1850 + }, + { + "epoch": 0.5408548996801396, + "grad_norm": 1.9685107469558716, + "learning_rate": 0.00015047219307450158, + "loss": 0.9616594314575195, + "step": 1860 + }, + { + "epoch": 0.5437627217214307, + "grad_norm": 1.945898413658142, + "learning_rate": 0.00015017238794783392, + "loss": 0.8819991111755371, + "step": 1870 + }, + { + "epoch": 0.5466705437627217, + "grad_norm": 1.999085783958435, + "learning_rate": 0.00014987258282116625, + "loss": 0.9666972160339355, + "step": 1880 + }, + { + "epoch": 0.5495783658040128, + "grad_norm": 2.3765146732330322, + "learning_rate": 0.00014957277769449859, + "loss": 0.959471607208252, + "step": 1890 + }, + { + "epoch": 0.5524861878453039, + "grad_norm": 2.3136610984802246, + "learning_rate": 0.00014927297256783092, + "loss": 1.027439785003662, + "step": 1900 + }, + { + "epoch": 0.5553940098865949, + "grad_norm": 2.017866373062134, + "learning_rate": 0.00014897316744116325, + "loss": 0.9374051094055176, + "step": 1910 + }, + { + "epoch": 0.558301831927886, + "grad_norm": 2.01604962348938, + "learning_rate": 0.0001486733623144956, + "loss": 0.9679468154907227, + "step": 1920 + }, + { + "epoch": 0.5612096539691771, + "grad_norm": 2.088538646697998, + "learning_rate": 0.00014837355718782792, + "loss": 1.0370773315429687, + "step": 1930 + }, + { + "epoch": 0.5641174760104681, + "grad_norm": 2.233410120010376, + "learning_rate": 0.00014807375206116026, + "loss": 1.0147683143615722, + "step": 1940 + }, + { + "epoch": 0.5670252980517593, + "grad_norm": 2.323624610900879, + "learning_rate": 0.0001477739469344926, + "loss": 0.8868412017822266, + "step": 1950 + }, + { + "epoch": 0.5699331200930503, + "grad_norm": 2.008387327194214, + "learning_rate": 0.00014747414180782493, + "loss": 1.042736530303955, + "step": 1960 + }, + { + "epoch": 0.5728409421343413, + "grad_norm": 2.069681406021118, + "learning_rate": 0.00014717433668115726, + "loss": 1.0332537651062013, + "step": 1970 + }, + { + "epoch": 0.5757487641756325, + "grad_norm": 2.0033297538757324, + "learning_rate": 0.00014687453155448957, + "loss": 1.033948040008545, + "step": 1980 + }, + { + "epoch": 0.5786565862169235, + "grad_norm": 2.320786237716675, + "learning_rate": 0.00014657472642782193, + "loss": 0.9264739036560059, + "step": 1990 + }, + { + "epoch": 0.5815644082582146, + "grad_norm": 2.1512808799743652, + "learning_rate": 0.00014627492130115427, + "loss": 0.9679861068725586, + "step": 2000 + }, + { + "epoch": 0.5815644082582146, + "eval_loss": 0.9911443591117859, + "eval_runtime": 466.7402, + "eval_samples_per_second": 7.368, + "eval_steps_per_second": 7.368, + "step": 2000 + }, + { + "epoch": 0.5844722302995057, + "grad_norm": 2.1983020305633545, + "learning_rate": 0.00014597511617448657, + "loss": 0.9210161209106446, + "step": 2010 + }, + { + "epoch": 0.5873800523407967, + "grad_norm": 2.4881699085235596, + "learning_rate": 0.00014567531104781893, + "loss": 0.9766405105590821, + "step": 2020 + }, + { + "epoch": 0.5902878743820879, + "grad_norm": 2.123229503631592, + "learning_rate": 0.00014537550592115127, + "loss": 0.9692230224609375, + "step": 2030 + }, + { + "epoch": 0.5931956964233789, + "grad_norm": 2.089709758758545, + "learning_rate": 0.00014507570079448358, + "loss": 0.9490997314453125, + "step": 2040 + }, + { + "epoch": 0.5961035184646699, + "grad_norm": 2.2401466369628906, + "learning_rate": 0.00014477589566781594, + "loss": 1.010519027709961, + "step": 2050 + }, + { + "epoch": 0.599011340505961, + "grad_norm": 1.967027187347412, + "learning_rate": 0.00014447609054114825, + "loss": 0.9748605728149414, + "step": 2060 + }, + { + "epoch": 0.6019191625472521, + "grad_norm": 2.350550889968872, + "learning_rate": 0.0001441762854144806, + "loss": 1.0137447357177733, + "step": 2070 + }, + { + "epoch": 0.6048269845885432, + "grad_norm": 2.527585744857788, + "learning_rate": 0.00014387648028781294, + "loss": 0.9678420066833496, + "step": 2080 + }, + { + "epoch": 0.6077348066298343, + "grad_norm": 2.1940855979919434, + "learning_rate": 0.00014357667516114525, + "loss": 0.9550240516662598, + "step": 2090 + }, + { + "epoch": 0.6106426286711253, + "grad_norm": 1.9863537549972534, + "learning_rate": 0.0001432768700344776, + "loss": 0.941413688659668, + "step": 2100 + }, + { + "epoch": 0.6135504507124164, + "grad_norm": 2.2011659145355225, + "learning_rate": 0.00014297706490780994, + "loss": 0.947486686706543, + "step": 2110 + }, + { + "epoch": 0.6164582727537075, + "grad_norm": 2.1515135765075684, + "learning_rate": 0.00014267725978114225, + "loss": 0.9121341705322266, + "step": 2120 + }, + { + "epoch": 0.6193660947949986, + "grad_norm": 2.002952814102173, + "learning_rate": 0.00014237745465447461, + "loss": 0.9708009719848633, + "step": 2130 + }, + { + "epoch": 0.6222739168362896, + "grad_norm": 1.953120231628418, + "learning_rate": 0.00014207764952780692, + "loss": 0.9927172660827637, + "step": 2140 + }, + { + "epoch": 0.6251817388775807, + "grad_norm": 1.9617403745651245, + "learning_rate": 0.00014177784440113926, + "loss": 0.9382636070251464, + "step": 2150 + }, + { + "epoch": 0.6280895609188718, + "grad_norm": 2.2497262954711914, + "learning_rate": 0.00014147803927447162, + "loss": 0.9852985382080078, + "step": 2160 + }, + { + "epoch": 0.6309973829601628, + "grad_norm": 2.1998257637023926, + "learning_rate": 0.00014117823414780392, + "loss": 1.0225582122802734, + "step": 2170 + }, + { + "epoch": 0.6339052050014539, + "grad_norm": 1.931475281715393, + "learning_rate": 0.00014087842902113626, + "loss": 0.951665210723877, + "step": 2180 + }, + { + "epoch": 0.636813027042745, + "grad_norm": 2.1101760864257812, + "learning_rate": 0.00014057862389446862, + "loss": 0.9361721038818359, + "step": 2190 + }, + { + "epoch": 0.639720849084036, + "grad_norm": 2.0846989154815674, + "learning_rate": 0.00014027881876780093, + "loss": 0.9642247200012207, + "step": 2200 + }, + { + "epoch": 0.6426286711253272, + "grad_norm": 2.221968650817871, + "learning_rate": 0.00013997901364113326, + "loss": 0.9930216789245605, + "step": 2210 + }, + { + "epoch": 0.6455364931666182, + "grad_norm": 1.9288430213928223, + "learning_rate": 0.0001396792085144656, + "loss": 0.934235668182373, + "step": 2220 + }, + { + "epoch": 0.6484443152079092, + "grad_norm": 1.9532197713851929, + "learning_rate": 0.00013937940338779793, + "loss": 0.9925549507141114, + "step": 2230 + }, + { + "epoch": 0.6513521372492004, + "grad_norm": 1.9111248254776, + "learning_rate": 0.00013907959826113027, + "loss": 0.9969470024108886, + "step": 2240 + }, + { + "epoch": 0.6542599592904914, + "grad_norm": 2.285212278366089, + "learning_rate": 0.0001387797931344626, + "loss": 0.9836911201477051, + "step": 2250 + }, + { + "epoch": 0.6571677813317824, + "grad_norm": 2.1522858142852783, + "learning_rate": 0.00013847998800779494, + "loss": 0.9622378349304199, + "step": 2260 + }, + { + "epoch": 0.6600756033730736, + "grad_norm": 2.1318211555480957, + "learning_rate": 0.0001381801828811273, + "loss": 0.9809438705444335, + "step": 2270 + }, + { + "epoch": 0.6629834254143646, + "grad_norm": 2.0285794734954834, + "learning_rate": 0.0001378803777544596, + "loss": 1.0077838897705078, + "step": 2280 + }, + { + "epoch": 0.6658912474556558, + "grad_norm": 1.9738699197769165, + "learning_rate": 0.00013758057262779194, + "loss": 0.9947647094726563, + "step": 2290 + }, + { + "epoch": 0.6687990694969468, + "grad_norm": 2.0177745819091797, + "learning_rate": 0.00013728076750112427, + "loss": 1.000431442260742, + "step": 2300 + }, + { + "epoch": 0.6717068915382378, + "grad_norm": 2.348642110824585, + "learning_rate": 0.0001369809623744566, + "loss": 1.0012446403503419, + "step": 2310 + }, + { + "epoch": 0.674614713579529, + "grad_norm": 2.1052684783935547, + "learning_rate": 0.00013668115724778894, + "loss": 1.0000034332275392, + "step": 2320 + }, + { + "epoch": 0.67752253562082, + "grad_norm": 2.2087337970733643, + "learning_rate": 0.00013638135212112128, + "loss": 0.946139907836914, + "step": 2330 + }, + { + "epoch": 0.6804303576621111, + "grad_norm": 2.344235420227051, + "learning_rate": 0.0001360815469944536, + "loss": 0.9389586448669434, + "step": 2340 + }, + { + "epoch": 0.6833381797034022, + "grad_norm": 1.8961925506591797, + "learning_rate": 0.00013578174186778595, + "loss": 0.9666190147399902, + "step": 2350 + }, + { + "epoch": 0.6862460017446932, + "grad_norm": 2.05000901222229, + "learning_rate": 0.00013548193674111828, + "loss": 0.9825595855712891, + "step": 2360 + }, + { + "epoch": 0.6891538237859843, + "grad_norm": 2.0229718685150146, + "learning_rate": 0.00013518213161445061, + "loss": 1.0510747909545899, + "step": 2370 + }, + { + "epoch": 0.6920616458272754, + "grad_norm": 1.9893312454223633, + "learning_rate": 0.00013488232648778295, + "loss": 0.954042911529541, + "step": 2380 + }, + { + "epoch": 0.6949694678685664, + "grad_norm": 1.9798680543899536, + "learning_rate": 0.00013458252136111528, + "loss": 0.9377657890319824, + "step": 2390 + }, + { + "epoch": 0.6978772899098575, + "grad_norm": 2.1432321071624756, + "learning_rate": 0.00013428271623444762, + "loss": 0.9713227272033691, + "step": 2400 + }, + { + "epoch": 0.7007851119511486, + "grad_norm": 2.111288070678711, + "learning_rate": 0.00013398291110777995, + "loss": 0.9865769386291504, + "step": 2410 + }, + { + "epoch": 0.7036929339924397, + "grad_norm": 1.9824166297912598, + "learning_rate": 0.0001336831059811123, + "loss": 0.9995452880859375, + "step": 2420 + }, + { + "epoch": 0.7066007560337307, + "grad_norm": 1.9680949449539185, + "learning_rate": 0.00013338330085444462, + "loss": 0.909880256652832, + "step": 2430 + }, + { + "epoch": 0.7095085780750218, + "grad_norm": 2.0280842781066895, + "learning_rate": 0.00013308349572777693, + "loss": 0.8853329658508301, + "step": 2440 + }, + { + "epoch": 0.7124164001163129, + "grad_norm": 1.9791338443756104, + "learning_rate": 0.0001327836906011093, + "loss": 1.0020163536071778, + "step": 2450 + }, + { + "epoch": 0.7153242221576039, + "grad_norm": 2.158463954925537, + "learning_rate": 0.00013248388547444163, + "loss": 0.915585708618164, + "step": 2460 + }, + { + "epoch": 0.7182320441988951, + "grad_norm": 2.0322635173797607, + "learning_rate": 0.00013218408034777396, + "loss": 0.9437061309814453, + "step": 2470 + }, + { + "epoch": 0.7211398662401861, + "grad_norm": 2.1616227626800537, + "learning_rate": 0.0001318842752211063, + "loss": 0.9552411079406739, + "step": 2480 + }, + { + "epoch": 0.7240476882814771, + "grad_norm": 1.9678977727890015, + "learning_rate": 0.00013158447009443863, + "loss": 0.9155937194824219, + "step": 2490 + }, + { + "epoch": 0.7269555103227683, + "grad_norm": 1.7541477680206299, + "learning_rate": 0.00013128466496777096, + "loss": 0.9028853416442871, + "step": 2500 + }, + { + "epoch": 0.7269555103227683, + "eval_loss": 0.9591814279556274, + "eval_runtime": 473.0242, + "eval_samples_per_second": 7.27, + "eval_steps_per_second": 7.27, + "step": 2500 + }, + { + "epoch": 0.7298633323640593, + "grad_norm": 2.0649845600128174, + "learning_rate": 0.0001309848598411033, + "loss": 0.9972336769104004, + "step": 2510 + }, + { + "epoch": 0.7327711544053503, + "grad_norm": 1.8874973058700562, + "learning_rate": 0.0001306850547144356, + "loss": 0.9446205139160156, + "step": 2520 + }, + { + "epoch": 0.7356789764466415, + "grad_norm": 2.4050185680389404, + "learning_rate": 0.00013038524958776797, + "loss": 1.0329988479614258, + "step": 2530 + }, + { + "epoch": 0.7385867984879325, + "grad_norm": 2.0193111896514893, + "learning_rate": 0.0001300854444611003, + "loss": 0.910318660736084, + "step": 2540 + }, + { + "epoch": 0.7414946205292237, + "grad_norm": 2.376051902770996, + "learning_rate": 0.0001297856393344326, + "loss": 1.0247613906860351, + "step": 2550 + }, + { + "epoch": 0.7444024425705147, + "grad_norm": 2.1105611324310303, + "learning_rate": 0.00012948583420776497, + "loss": 1.016015625, + "step": 2560 + }, + { + "epoch": 0.7473102646118057, + "grad_norm": 2.1081364154815674, + "learning_rate": 0.0001291860290810973, + "loss": 0.909546184539795, + "step": 2570 + }, + { + "epoch": 0.7502180866530969, + "grad_norm": 2.046095132827759, + "learning_rate": 0.0001288862239544296, + "loss": 0.9204464912414551, + "step": 2580 + }, + { + "epoch": 0.7531259086943879, + "grad_norm": 1.9615705013275146, + "learning_rate": 0.00012858641882776197, + "loss": 0.9793522834777832, + "step": 2590 + }, + { + "epoch": 0.7560337307356789, + "grad_norm": 1.9187848567962646, + "learning_rate": 0.00012828661370109428, + "loss": 0.9414368629455566, + "step": 2600 + }, + { + "epoch": 0.7589415527769701, + "grad_norm": 2.0125985145568848, + "learning_rate": 0.00012798680857442662, + "loss": 1.0450970649719238, + "step": 2610 + }, + { + "epoch": 0.7618493748182611, + "grad_norm": 2.2543137073516846, + "learning_rate": 0.00012768700344775898, + "loss": 1.0270273208618164, + "step": 2620 + }, + { + "epoch": 0.7647571968595522, + "grad_norm": 2.00054669380188, + "learning_rate": 0.00012738719832109128, + "loss": 1.0244030952453613, + "step": 2630 + }, + { + "epoch": 0.7676650189008433, + "grad_norm": 1.8571268320083618, + "learning_rate": 0.00012708739319442362, + "loss": 0.9479835510253907, + "step": 2640 + }, + { + "epoch": 0.7705728409421343, + "grad_norm": 2.0844404697418213, + "learning_rate": 0.00012678758806775598, + "loss": 0.8867239952087402, + "step": 2650 + }, + { + "epoch": 0.7734806629834254, + "grad_norm": 2.023630142211914, + "learning_rate": 0.0001264877829410883, + "loss": 0.9323092460632324, + "step": 2660 + }, + { + "epoch": 0.7763884850247165, + "grad_norm": 1.935514211654663, + "learning_rate": 0.00012618797781442065, + "loss": 0.997464942932129, + "step": 2670 + }, + { + "epoch": 0.7792963070660076, + "grad_norm": 2.1317570209503174, + "learning_rate": 0.00012588817268775296, + "loss": 0.9762091636657715, + "step": 2680 + }, + { + "epoch": 0.7822041291072986, + "grad_norm": 2.094515323638916, + "learning_rate": 0.0001255883675610853, + "loss": 0.919368839263916, + "step": 2690 + }, + { + "epoch": 0.7851119511485897, + "grad_norm": 2.0645945072174072, + "learning_rate": 0.00012528856243441765, + "loss": 0.917721176147461, + "step": 2700 + }, + { + "epoch": 0.7880197731898808, + "grad_norm": 2.178105592727661, + "learning_rate": 0.00012498875730774996, + "loss": 0.9504012107849121, + "step": 2710 + }, + { + "epoch": 0.7909275952311718, + "grad_norm": 2.3576605319976807, + "learning_rate": 0.0001246889521810823, + "loss": 1.0032535552978517, + "step": 2720 + }, + { + "epoch": 0.7938354172724629, + "grad_norm": 1.9736145734786987, + "learning_rate": 0.00012438914705441466, + "loss": 0.9342514991760253, + "step": 2730 + }, + { + "epoch": 0.796743239313754, + "grad_norm": 2.074565887451172, + "learning_rate": 0.00012408934192774696, + "loss": 0.951146125793457, + "step": 2740 + }, + { + "epoch": 0.799651061355045, + "grad_norm": 2.1545727252960205, + "learning_rate": 0.0001237895368010793, + "loss": 0.9272260665893555, + "step": 2750 + }, + { + "epoch": 0.8025588833963362, + "grad_norm": 2.0214531421661377, + "learning_rate": 0.00012348973167441163, + "loss": 0.9915397644042969, + "step": 2760 + }, + { + "epoch": 0.8054667054376272, + "grad_norm": 2.0707271099090576, + "learning_rate": 0.00012318992654774397, + "loss": 0.9817545890808106, + "step": 2770 + }, + { + "epoch": 0.8083745274789182, + "grad_norm": 1.9729819297790527, + "learning_rate": 0.0001228901214210763, + "loss": 0.9666746139526368, + "step": 2780 + }, + { + "epoch": 0.8112823495202094, + "grad_norm": 2.0239417552948, + "learning_rate": 0.00012259031629440864, + "loss": 0.9174615859985351, + "step": 2790 + }, + { + "epoch": 0.8141901715615004, + "grad_norm": 2.0677332878112793, + "learning_rate": 0.00012229051116774097, + "loss": 0.9489200592041016, + "step": 2800 + }, + { + "epoch": 0.8170979936027916, + "grad_norm": 2.0116071701049805, + "learning_rate": 0.0001219907060410733, + "loss": 0.8491521835327148, + "step": 2810 + }, + { + "epoch": 0.8200058156440826, + "grad_norm": 2.005683422088623, + "learning_rate": 0.00012169090091440565, + "loss": 0.8922024726867676, + "step": 2820 + }, + { + "epoch": 0.8229136376853736, + "grad_norm": 2.201763868331909, + "learning_rate": 0.00012139109578773797, + "loss": 0.9258706092834472, + "step": 2830 + }, + { + "epoch": 0.8258214597266648, + "grad_norm": 1.9857138395309448, + "learning_rate": 0.0001210912906610703, + "loss": 0.9566280364990234, + "step": 2840 + }, + { + "epoch": 0.8287292817679558, + "grad_norm": 2.107966899871826, + "learning_rate": 0.00012079148553440264, + "loss": 0.924662971496582, + "step": 2850 + }, + { + "epoch": 0.8316371038092468, + "grad_norm": 2.1448616981506348, + "learning_rate": 0.00012049168040773498, + "loss": 0.9880090713500976, + "step": 2860 + }, + { + "epoch": 0.834544925850538, + "grad_norm": 2.1623342037200928, + "learning_rate": 0.0001201918752810673, + "loss": 0.9296030044555664, + "step": 2870 + }, + { + "epoch": 0.837452747891829, + "grad_norm": 2.149792194366455, + "learning_rate": 0.00011989207015439965, + "loss": 0.8926197052001953, + "step": 2880 + }, + { + "epoch": 0.8403605699331201, + "grad_norm": 1.7601749897003174, + "learning_rate": 0.00011959226502773198, + "loss": 0.937494945526123, + "step": 2890 + }, + { + "epoch": 0.8432683919744112, + "grad_norm": 1.8277206420898438, + "learning_rate": 0.00011929245990106433, + "loss": 0.9348894119262695, + "step": 2900 + }, + { + "epoch": 0.8461762140157022, + "grad_norm": 2.417834520339966, + "learning_rate": 0.00011899265477439665, + "loss": 0.9014430046081543, + "step": 2910 + }, + { + "epoch": 0.8490840360569933, + "grad_norm": 1.9275659322738647, + "learning_rate": 0.00011869284964772897, + "loss": 0.904062557220459, + "step": 2920 + }, + { + "epoch": 0.8519918580982844, + "grad_norm": 1.8992774486541748, + "learning_rate": 0.00011839304452106132, + "loss": 0.9319318771362305, + "step": 2930 + }, + { + "epoch": 0.8548996801395755, + "grad_norm": 1.7830348014831543, + "learning_rate": 0.00011809323939439365, + "loss": 0.9675899505615234, + "step": 2940 + }, + { + "epoch": 0.8578075021808665, + "grad_norm": 1.9002785682678223, + "learning_rate": 0.00011779343426772598, + "loss": 0.9884864807128906, + "step": 2950 + }, + { + "epoch": 0.8607153242221576, + "grad_norm": 2.256084442138672, + "learning_rate": 0.00011749362914105832, + "loss": 0.9169980049133301, + "step": 2960 + }, + { + "epoch": 0.8636231462634487, + "grad_norm": 1.977953553199768, + "learning_rate": 0.00011719382401439066, + "loss": 0.9288657188415528, + "step": 2970 + }, + { + "epoch": 0.8665309683047397, + "grad_norm": 2.223893404006958, + "learning_rate": 0.00011689401888772298, + "loss": 0.8921388626098633, + "step": 2980 + }, + { + "epoch": 0.8694387903460308, + "grad_norm": 1.8232818841934204, + "learning_rate": 0.00011659421376105533, + "loss": 0.9274946212768554, + "step": 2990 + }, + { + "epoch": 0.8723466123873219, + "grad_norm": 2.229280471801758, + "learning_rate": 0.00011629440863438765, + "loss": 0.9193262100219727, + "step": 3000 + }, + { + "epoch": 0.8723466123873219, + "eval_loss": 0.9303778409957886, + "eval_runtime": 469.5057, + "eval_samples_per_second": 7.325, + "eval_steps_per_second": 7.325, + "step": 3000 + }, + { + "epoch": 0.875254434428613, + "grad_norm": 1.7871650457382202, + "learning_rate": 0.00011599460350771998, + "loss": 1.0009597778320312, + "step": 3010 + }, + { + "epoch": 0.8781622564699041, + "grad_norm": 1.9837925434112549, + "learning_rate": 0.00011569479838105233, + "loss": 0.9461586952209473, + "step": 3020 + }, + { + "epoch": 0.8810700785111951, + "grad_norm": 2.1220059394836426, + "learning_rate": 0.00011539499325438465, + "loss": 0.8776027679443359, + "step": 3030 + }, + { + "epoch": 0.8839779005524862, + "grad_norm": 2.0386621952056885, + "learning_rate": 0.00011509518812771699, + "loss": 0.8935223579406738, + "step": 3040 + }, + { + "epoch": 0.8868857225937773, + "grad_norm": 2.0495152473449707, + "learning_rate": 0.00011479538300104933, + "loss": 0.9414477348327637, + "step": 3050 + }, + { + "epoch": 0.8897935446350683, + "grad_norm": 2.1676785945892334, + "learning_rate": 0.00011449557787438165, + "loss": 0.9668932914733886, + "step": 3060 + }, + { + "epoch": 0.8927013666763594, + "grad_norm": 1.8518208265304565, + "learning_rate": 0.00011419577274771398, + "loss": 0.83909912109375, + "step": 3070 + }, + { + "epoch": 0.8956091887176505, + "grad_norm": 2.6842968463897705, + "learning_rate": 0.00011389596762104632, + "loss": 1.001734161376953, + "step": 3080 + }, + { + "epoch": 0.8985170107589415, + "grad_norm": 2.093907594680786, + "learning_rate": 0.00011359616249437866, + "loss": 0.9194964408874512, + "step": 3090 + }, + { + "epoch": 0.9014248328002327, + "grad_norm": 1.7096366882324219, + "learning_rate": 0.000113296357367711, + "loss": 0.8888837814331054, + "step": 3100 + }, + { + "epoch": 0.9043326548415237, + "grad_norm": 2.0421509742736816, + "learning_rate": 0.00011299655224104333, + "loss": 0.9483534812927246, + "step": 3110 + }, + { + "epoch": 0.9072404768828147, + "grad_norm": 2.220430612564087, + "learning_rate": 0.00011269674711437566, + "loss": 0.855626106262207, + "step": 3120 + }, + { + "epoch": 0.9101482989241059, + "grad_norm": 2.721151351928711, + "learning_rate": 0.00011239694198770801, + "loss": 0.9696638107299804, + "step": 3130 + }, + { + "epoch": 0.9130561209653969, + "grad_norm": 2.0278444290161133, + "learning_rate": 0.00011209713686104033, + "loss": 0.9626810073852539, + "step": 3140 + }, + { + "epoch": 0.915963943006688, + "grad_norm": 1.9929981231689453, + "learning_rate": 0.00011179733173437265, + "loss": 0.9121280670166015, + "step": 3150 + }, + { + "epoch": 0.9188717650479791, + "grad_norm": 1.8903363943099976, + "learning_rate": 0.000111497526607705, + "loss": 0.8926810264587403, + "step": 3160 + }, + { + "epoch": 0.9217795870892701, + "grad_norm": 1.9016224145889282, + "learning_rate": 0.00011119772148103733, + "loss": 0.9811866760253907, + "step": 3170 + }, + { + "epoch": 0.9246874091305612, + "grad_norm": 1.7538851499557495, + "learning_rate": 0.00011089791635436966, + "loss": 0.9020210266113281, + "step": 3180 + }, + { + "epoch": 0.9275952311718523, + "grad_norm": 1.8952163457870483, + "learning_rate": 0.000110598111227702, + "loss": 0.9186564445495605, + "step": 3190 + }, + { + "epoch": 0.9305030532131433, + "grad_norm": 2.0162456035614014, + "learning_rate": 0.00011029830610103434, + "loss": 0.9249841690063476, + "step": 3200 + }, + { + "epoch": 0.9334108752544344, + "grad_norm": 1.9944981336593628, + "learning_rate": 0.00010999850097436666, + "loss": 0.8584202766418457, + "step": 3210 + }, + { + "epoch": 0.9363186972957255, + "grad_norm": 1.9651786088943481, + "learning_rate": 0.000109698695847699, + "loss": 0.8806270599365235, + "step": 3220 + }, + { + "epoch": 0.9392265193370166, + "grad_norm": 1.7801518440246582, + "learning_rate": 0.00010939889072103133, + "loss": 0.9265996932983398, + "step": 3230 + }, + { + "epoch": 0.9421343413783076, + "grad_norm": 2.065995454788208, + "learning_rate": 0.00010909908559436366, + "loss": 0.9499953269958497, + "step": 3240 + }, + { + "epoch": 0.9450421634195987, + "grad_norm": 1.797984004020691, + "learning_rate": 0.00010879928046769601, + "loss": 0.8817252159118653, + "step": 3250 + }, + { + "epoch": 0.9479499854608898, + "grad_norm": 1.7669072151184082, + "learning_rate": 0.00010849947534102833, + "loss": 0.9481887817382812, + "step": 3260 + }, + { + "epoch": 0.9508578075021809, + "grad_norm": 2.1424496173858643, + "learning_rate": 0.00010819967021436067, + "loss": 0.9072481155395508, + "step": 3270 + }, + { + "epoch": 0.953765629543472, + "grad_norm": 2.006564140319824, + "learning_rate": 0.00010789986508769301, + "loss": 0.9786685943603516, + "step": 3280 + }, + { + "epoch": 0.956673451584763, + "grad_norm": 1.8808116912841797, + "learning_rate": 0.00010760005996102533, + "loss": 0.9131714820861816, + "step": 3290 + }, + { + "epoch": 0.959581273626054, + "grad_norm": 1.8156124353408813, + "learning_rate": 0.00010730025483435768, + "loss": 0.9058806419372558, + "step": 3300 + }, + { + "epoch": 0.9624890956673452, + "grad_norm": 2.1244983673095703, + "learning_rate": 0.00010700044970769, + "loss": 0.8985400199890137, + "step": 3310 + }, + { + "epoch": 0.9653969177086362, + "grad_norm": 1.979709506034851, + "learning_rate": 0.00010670064458102234, + "loss": 0.8887692451477051, + "step": 3320 + }, + { + "epoch": 0.9683047397499273, + "grad_norm": 1.8038535118103027, + "learning_rate": 0.00010640083945435469, + "loss": 0.9474887847900391, + "step": 3330 + }, + { + "epoch": 0.9712125617912184, + "grad_norm": 1.883305549621582, + "learning_rate": 0.00010610103432768701, + "loss": 0.8624442100524903, + "step": 3340 + }, + { + "epoch": 0.9741203838325094, + "grad_norm": 1.9134957790374756, + "learning_rate": 0.00010580122920101934, + "loss": 0.8966679573059082, + "step": 3350 + }, + { + "epoch": 0.9770282058738006, + "grad_norm": 1.8334492444992065, + "learning_rate": 0.00010550142407435169, + "loss": 0.8644783020019531, + "step": 3360 + }, + { + "epoch": 0.9799360279150916, + "grad_norm": 1.8988165855407715, + "learning_rate": 0.00010520161894768401, + "loss": 0.8418782234191895, + "step": 3370 + }, + { + "epoch": 0.9828438499563826, + "grad_norm": 2.0219290256500244, + "learning_rate": 0.00010490181382101633, + "loss": 0.9593500137329102, + "step": 3380 + }, + { + "epoch": 0.9857516719976738, + "grad_norm": 1.9475761651992798, + "learning_rate": 0.00010460200869434868, + "loss": 0.9046772956848145, + "step": 3390 + }, + { + "epoch": 0.9886594940389648, + "grad_norm": 2.156895637512207, + "learning_rate": 0.00010430220356768101, + "loss": 0.9402565956115723, + "step": 3400 + }, + { + "epoch": 0.9915673160802558, + "grad_norm": 1.8304574489593506, + "learning_rate": 0.00010400239844101334, + "loss": 0.9041817665100098, + "step": 3410 + }, + { + "epoch": 0.994475138121547, + "grad_norm": 2.0752060413360596, + "learning_rate": 0.00010370259331434568, + "loss": 0.8676543235778809, + "step": 3420 + }, + { + "epoch": 0.997382960162838, + "grad_norm": 1.845677375793457, + "learning_rate": 0.00010340278818767802, + "loss": 0.8220011711120605, + "step": 3430 + }, + { + "epoch": 1.0002907822041291, + "grad_norm": 1.7879046201705933, + "learning_rate": 0.00010310298306101034, + "loss": 0.9136262893676758, + "step": 3440 + }, + { + "epoch": 1.0031986042454202, + "grad_norm": 1.9771437644958496, + "learning_rate": 0.00010280317793434269, + "loss": 0.7697383880615234, + "step": 3450 + }, + { + "epoch": 1.0061064262867112, + "grad_norm": 2.23533296585083, + "learning_rate": 0.00010250337280767501, + "loss": 0.8118080139160156, + "step": 3460 + }, + { + "epoch": 1.0090142483280022, + "grad_norm": 1.9873441457748413, + "learning_rate": 0.00010220356768100734, + "loss": 0.7812850475311279, + "step": 3470 + }, + { + "epoch": 1.0119220703692935, + "grad_norm": 1.8505840301513672, + "learning_rate": 0.00010190376255433969, + "loss": 0.8191473960876465, + "step": 3480 + }, + { + "epoch": 1.0148298924105845, + "grad_norm": 2.0484778881073, + "learning_rate": 0.00010160395742767201, + "loss": 0.8035045623779297, + "step": 3490 + }, + { + "epoch": 1.0177377144518756, + "grad_norm": 1.9655097723007202, + "learning_rate": 0.00010130415230100436, + "loss": 0.779402208328247, + "step": 3500 + }, + { + "epoch": 1.0177377144518756, + "eval_loss": 0.9158970713615417, + "eval_runtime": 474.9541, + "eval_samples_per_second": 7.241, + "eval_steps_per_second": 7.241, + "step": 3500 + }, + { + "epoch": 1.0206455364931666, + "grad_norm": 2.0883100032806396, + "learning_rate": 0.0001010043471743367, + "loss": 0.8522077560424804, + "step": 3510 + }, + { + "epoch": 1.0235533585344576, + "grad_norm": 2.098511219024658, + "learning_rate": 0.00010070454204766901, + "loss": 0.8504983901977539, + "step": 3520 + }, + { + "epoch": 1.0264611805757489, + "grad_norm": 2.101414442062378, + "learning_rate": 0.00010040473692100136, + "loss": 0.7838150978088378, + "step": 3530 + }, + { + "epoch": 1.02936900261704, + "grad_norm": 2.2732677459716797, + "learning_rate": 0.00010010493179433368, + "loss": 0.7565544605255127, + "step": 3540 + }, + { + "epoch": 1.032276824658331, + "grad_norm": 1.695768117904663, + "learning_rate": 9.980512666766603e-05, + "loss": 0.7418520450592041, + "step": 3550 + }, + { + "epoch": 1.035184646699622, + "grad_norm": 2.0946264266967773, + "learning_rate": 9.950532154099835e-05, + "loss": 0.820562744140625, + "step": 3560 + }, + { + "epoch": 1.038092468740913, + "grad_norm": 2.103559732437134, + "learning_rate": 9.920551641433069e-05, + "loss": 0.769464635848999, + "step": 3570 + }, + { + "epoch": 1.041000290782204, + "grad_norm": 1.9463759660720825, + "learning_rate": 9.890571128766302e-05, + "loss": 0.7375322818756104, + "step": 3580 + }, + { + "epoch": 1.0439081128234953, + "grad_norm": 2.0933845043182373, + "learning_rate": 9.860590616099536e-05, + "loss": 0.7640562534332276, + "step": 3590 + }, + { + "epoch": 1.0468159348647863, + "grad_norm": 1.9825164079666138, + "learning_rate": 9.830610103432769e-05, + "loss": 0.7955950736999512, + "step": 3600 + }, + { + "epoch": 1.0497237569060773, + "grad_norm": 1.9863218069076538, + "learning_rate": 9.800629590766003e-05, + "loss": 0.7707761764526367, + "step": 3610 + }, + { + "epoch": 1.0526315789473684, + "grad_norm": 1.9356553554534912, + "learning_rate": 9.770649078099236e-05, + "loss": 0.7836559772491455, + "step": 3620 + }, + { + "epoch": 1.0555394009886594, + "grad_norm": 2.0146119594573975, + "learning_rate": 9.74066856543247e-05, + "loss": 0.7523280143737793, + "step": 3630 + }, + { + "epoch": 1.0584472230299506, + "grad_norm": 2.056215763092041, + "learning_rate": 9.710688052765703e-05, + "loss": 0.8155484199523926, + "step": 3640 + }, + { + "epoch": 1.0613550450712417, + "grad_norm": 1.8928337097167969, + "learning_rate": 9.680707540098936e-05, + "loss": 0.8006336212158203, + "step": 3650 + }, + { + "epoch": 1.0642628671125327, + "grad_norm": 1.931458592414856, + "learning_rate": 9.65072702743217e-05, + "loss": 0.7413979530334472, + "step": 3660 + }, + { + "epoch": 1.0671706891538237, + "grad_norm": 2.162804126739502, + "learning_rate": 9.620746514765403e-05, + "loss": 0.7883041381835938, + "step": 3670 + }, + { + "epoch": 1.0700785111951148, + "grad_norm": 2.1737864017486572, + "learning_rate": 9.590766002098637e-05, + "loss": 0.7595384597778321, + "step": 3680 + }, + { + "epoch": 1.072986333236406, + "grad_norm": 2.1503849029541016, + "learning_rate": 9.560785489431869e-05, + "loss": 0.8216732025146485, + "step": 3690 + }, + { + "epoch": 1.075894155277697, + "grad_norm": 1.9446245431900024, + "learning_rate": 9.530804976765104e-05, + "loss": 0.8016197204589843, + "step": 3700 + }, + { + "epoch": 1.078801977318988, + "grad_norm": 1.8525793552398682, + "learning_rate": 9.500824464098337e-05, + "loss": 0.8090981483459473, + "step": 3710 + }, + { + "epoch": 1.0817097993602791, + "grad_norm": 1.8631248474121094, + "learning_rate": 9.470843951431569e-05, + "loss": 0.7200882434844971, + "step": 3720 + }, + { + "epoch": 1.0846176214015701, + "grad_norm": 2.06455135345459, + "learning_rate": 9.440863438764803e-05, + "loss": 0.7492884635925293, + "step": 3730 + }, + { + "epoch": 1.0875254434428614, + "grad_norm": 2.1832897663116455, + "learning_rate": 9.410882926098037e-05, + "loss": 0.8133129119873047, + "step": 3740 + }, + { + "epoch": 1.0904332654841524, + "grad_norm": 1.9299908876419067, + "learning_rate": 9.380902413431271e-05, + "loss": 0.7218931674957275, + "step": 3750 + }, + { + "epoch": 1.0933410875254435, + "grad_norm": 1.8345733880996704, + "learning_rate": 9.350921900764503e-05, + "loss": 0.8038671493530274, + "step": 3760 + }, + { + "epoch": 1.0962489095667345, + "grad_norm": 2.200580596923828, + "learning_rate": 9.320941388097736e-05, + "loss": 0.758363676071167, + "step": 3770 + }, + { + "epoch": 1.0991567316080255, + "grad_norm": 1.9002389907836914, + "learning_rate": 9.290960875430971e-05, + "loss": 0.7319043636322021, + "step": 3780 + }, + { + "epoch": 1.1020645536493165, + "grad_norm": 2.178980827331543, + "learning_rate": 9.260980362764203e-05, + "loss": 0.8144195556640625, + "step": 3790 + }, + { + "epoch": 1.1049723756906078, + "grad_norm": 1.777524709701538, + "learning_rate": 9.230999850097437e-05, + "loss": 0.7428917407989502, + "step": 3800 + }, + { + "epoch": 1.1078801977318988, + "grad_norm": 2.269186019897461, + "learning_rate": 9.20101933743067e-05, + "loss": 0.7581167221069336, + "step": 3810 + }, + { + "epoch": 1.1107880197731899, + "grad_norm": 2.0214948654174805, + "learning_rate": 9.171038824763904e-05, + "loss": 0.786471939086914, + "step": 3820 + }, + { + "epoch": 1.113695841814481, + "grad_norm": 1.8476779460906982, + "learning_rate": 9.141058312097137e-05, + "loss": 0.7075447559356689, + "step": 3830 + }, + { + "epoch": 1.1166036638557721, + "grad_norm": 1.7384737730026245, + "learning_rate": 9.11107779943037e-05, + "loss": 0.7290010452270508, + "step": 3840 + }, + { + "epoch": 1.1195114858970632, + "grad_norm": 2.1066110134124756, + "learning_rate": 9.081097286763604e-05, + "loss": 0.7164917469024659, + "step": 3850 + }, + { + "epoch": 1.1224193079383542, + "grad_norm": 2.2718801498413086, + "learning_rate": 9.051116774096837e-05, + "loss": 0.7832975864410401, + "step": 3860 + }, + { + "epoch": 1.1253271299796452, + "grad_norm": 2.0165557861328125, + "learning_rate": 9.021136261430071e-05, + "loss": 0.840385913848877, + "step": 3870 + }, + { + "epoch": 1.1282349520209363, + "grad_norm": 1.9852454662322998, + "learning_rate": 8.991155748763304e-05, + "loss": 0.8062166213989258, + "step": 3880 + }, + { + "epoch": 1.1311427740622273, + "grad_norm": 1.8891923427581787, + "learning_rate": 8.961175236096538e-05, + "loss": 0.860920524597168, + "step": 3890 + }, + { + "epoch": 1.1340505961035185, + "grad_norm": 2.2840018272399902, + "learning_rate": 8.931194723429771e-05, + "loss": 0.8350888252258301, + "step": 3900 + }, + { + "epoch": 1.1369584181448096, + "grad_norm": 1.9931566715240479, + "learning_rate": 8.901214210763005e-05, + "loss": 0.7368578910827637, + "step": 3910 + }, + { + "epoch": 1.1398662401861006, + "grad_norm": 2.0133721828460693, + "learning_rate": 8.871233698096237e-05, + "loss": 0.8287955284118652, + "step": 3920 + }, + { + "epoch": 1.1427740622273916, + "grad_norm": 2.1664857864379883, + "learning_rate": 8.841253185429472e-05, + "loss": 0.7541079044342041, + "step": 3930 + }, + { + "epoch": 1.1456818842686827, + "grad_norm": 2.1204898357391357, + "learning_rate": 8.811272672762705e-05, + "loss": 0.7905386447906494, + "step": 3940 + }, + { + "epoch": 1.148589706309974, + "grad_norm": 1.8248870372772217, + "learning_rate": 8.781292160095938e-05, + "loss": 0.7721247673034668, + "step": 3950 + }, + { + "epoch": 1.151497528351265, + "grad_norm": 1.8345229625701904, + "learning_rate": 8.75131164742917e-05, + "loss": 0.7757219314575196, + "step": 3960 + }, + { + "epoch": 1.154405350392556, + "grad_norm": 1.6888972520828247, + "learning_rate": 8.721331134762405e-05, + "loss": 0.7593977451324463, + "step": 3970 + }, + { + "epoch": 1.157313172433847, + "grad_norm": 1.9662666320800781, + "learning_rate": 8.691350622095639e-05, + "loss": 0.7545000076293945, + "step": 3980 + }, + { + "epoch": 1.160220994475138, + "grad_norm": 1.987597107887268, + "learning_rate": 8.661370109428871e-05, + "loss": 0.7677037239074707, + "step": 3990 + }, + { + "epoch": 1.163128816516429, + "grad_norm": 2.0462353229522705, + "learning_rate": 8.631389596762104e-05, + "loss": 0.7688228130340576, + "step": 4000 + }, + { + "epoch": 1.163128816516429, + "eval_loss": 0.902101457118988, + "eval_runtime": 471.9941, + "eval_samples_per_second": 7.286, + "eval_steps_per_second": 7.286, + "step": 4000 + }, + { + "epoch": 1.1660366385577203, + "grad_norm": 2.0174319744110107, + "learning_rate": 8.601409084095339e-05, + "loss": 0.6873229026794434, + "step": 4010 + }, + { + "epoch": 1.1689444605990114, + "grad_norm": 2.1310856342315674, + "learning_rate": 8.571428571428571e-05, + "loss": 0.7570897579193115, + "step": 4020 + }, + { + "epoch": 1.1718522826403024, + "grad_norm": 2.121331214904785, + "learning_rate": 8.541448058761805e-05, + "loss": 0.8037228584289551, + "step": 4030 + }, + { + "epoch": 1.1747601046815934, + "grad_norm": 1.866264820098877, + "learning_rate": 8.511467546095038e-05, + "loss": 0.7070183277130127, + "step": 4040 + }, + { + "epoch": 1.1776679267228847, + "grad_norm": 2.1696619987487793, + "learning_rate": 8.481487033428273e-05, + "loss": 0.7907981872558594, + "step": 4050 + }, + { + "epoch": 1.1805757487641757, + "grad_norm": 1.9860618114471436, + "learning_rate": 8.451506520761505e-05, + "loss": 0.801030445098877, + "step": 4060 + }, + { + "epoch": 1.1834835708054667, + "grad_norm": 1.92996084690094, + "learning_rate": 8.421526008094739e-05, + "loss": 0.7623518943786621, + "step": 4070 + }, + { + "epoch": 1.1863913928467578, + "grad_norm": 2.483771800994873, + "learning_rate": 8.391545495427972e-05, + "loss": 0.7155436992645263, + "step": 4080 + }, + { + "epoch": 1.1892992148880488, + "grad_norm": 2.0799410343170166, + "learning_rate": 8.361564982761205e-05, + "loss": 0.7853653430938721, + "step": 4090 + }, + { + "epoch": 1.1922070369293398, + "grad_norm": 1.9911246299743652, + "learning_rate": 8.331584470094439e-05, + "loss": 0.7899494171142578, + "step": 4100 + }, + { + "epoch": 1.195114858970631, + "grad_norm": 2.0691795349121094, + "learning_rate": 8.301603957427672e-05, + "loss": 0.7937345027923584, + "step": 4110 + }, + { + "epoch": 1.198022681011922, + "grad_norm": 1.9724209308624268, + "learning_rate": 8.271623444760906e-05, + "loss": 0.7867683887481689, + "step": 4120 + }, + { + "epoch": 1.2009305030532131, + "grad_norm": 2.0218544006347656, + "learning_rate": 8.241642932094139e-05, + "loss": 0.7292540550231934, + "step": 4130 + }, + { + "epoch": 1.2038383250945042, + "grad_norm": 1.8983079195022583, + "learning_rate": 8.211662419427373e-05, + "loss": 0.7957502841949463, + "step": 4140 + }, + { + "epoch": 1.2067461471357952, + "grad_norm": 2.0544803142547607, + "learning_rate": 8.181681906760606e-05, + "loss": 0.8023088455200196, + "step": 4150 + }, + { + "epoch": 1.2096539691770865, + "grad_norm": 2.2147371768951416, + "learning_rate": 8.15170139409384e-05, + "loss": 0.7521897315979004, + "step": 4160 + }, + { + "epoch": 1.2125617912183775, + "grad_norm": 2.114715814590454, + "learning_rate": 8.121720881427073e-05, + "loss": 0.7361032485961914, + "step": 4170 + }, + { + "epoch": 1.2154696132596685, + "grad_norm": 2.0806241035461426, + "learning_rate": 8.091740368760306e-05, + "loss": 0.810362434387207, + "step": 4180 + }, + { + "epoch": 1.2183774353009595, + "grad_norm": 2.1156203746795654, + "learning_rate": 8.061759856093539e-05, + "loss": 0.7738988876342774, + "step": 4190 + }, + { + "epoch": 1.2212852573422506, + "grad_norm": 2.063788652420044, + "learning_rate": 8.031779343426773e-05, + "loss": 0.7684555053710938, + "step": 4200 + }, + { + "epoch": 1.2241930793835416, + "grad_norm": 1.8935747146606445, + "learning_rate": 8.001798830760007e-05, + "loss": 0.7432861328125, + "step": 4210 + }, + { + "epoch": 1.2271009014248329, + "grad_norm": 1.9250372648239136, + "learning_rate": 7.971818318093239e-05, + "loss": 0.752553653717041, + "step": 4220 + }, + { + "epoch": 1.2300087234661239, + "grad_norm": 1.8770952224731445, + "learning_rate": 7.941837805426474e-05, + "loss": 0.7726155281066894, + "step": 4230 + }, + { + "epoch": 1.232916545507415, + "grad_norm": 1.8295116424560547, + "learning_rate": 7.911857292759707e-05, + "loss": 0.7889208793640137, + "step": 4240 + }, + { + "epoch": 1.235824367548706, + "grad_norm": 1.932578682899475, + "learning_rate": 7.88187678009294e-05, + "loss": 0.8311320304870605, + "step": 4250 + }, + { + "epoch": 1.2387321895899972, + "grad_norm": 1.9207606315612793, + "learning_rate": 7.851896267426173e-05, + "loss": 0.7485177516937256, + "step": 4260 + }, + { + "epoch": 1.2416400116312882, + "grad_norm": 1.9031519889831543, + "learning_rate": 7.821915754759408e-05, + "loss": 0.7838140010833741, + "step": 4270 + }, + { + "epoch": 1.2445478336725793, + "grad_norm": 1.8704465627670288, + "learning_rate": 7.791935242092641e-05, + "loss": 0.7774901390075684, + "step": 4280 + }, + { + "epoch": 1.2474556557138703, + "grad_norm": 1.819907784461975, + "learning_rate": 7.761954729425873e-05, + "loss": 0.750859260559082, + "step": 4290 + }, + { + "epoch": 1.2503634777551613, + "grad_norm": 1.9019205570220947, + "learning_rate": 7.731974216759107e-05, + "loss": 0.780170202255249, + "step": 4300 + }, + { + "epoch": 1.2532712997964524, + "grad_norm": 2.030393362045288, + "learning_rate": 7.701993704092341e-05, + "loss": 0.8301087379455566, + "step": 4310 + }, + { + "epoch": 1.2561791218377436, + "grad_norm": 1.842457890510559, + "learning_rate": 7.672013191425573e-05, + "loss": 0.7692018985748291, + "step": 4320 + }, + { + "epoch": 1.2590869438790346, + "grad_norm": 1.8982383012771606, + "learning_rate": 7.642032678758807e-05, + "loss": 0.781231689453125, + "step": 4330 + }, + { + "epoch": 1.2619947659203257, + "grad_norm": 1.8713147640228271, + "learning_rate": 7.61205216609204e-05, + "loss": 0.762873363494873, + "step": 4340 + }, + { + "epoch": 1.2649025879616167, + "grad_norm": 2.0454752445220947, + "learning_rate": 7.582071653425275e-05, + "loss": 0.7406270980834961, + "step": 4350 + }, + { + "epoch": 1.267810410002908, + "grad_norm": 2.4180965423583984, + "learning_rate": 7.552091140758507e-05, + "loss": 0.7520180702209472, + "step": 4360 + }, + { + "epoch": 1.270718232044199, + "grad_norm": 1.986022710800171, + "learning_rate": 7.52211062809174e-05, + "loss": 0.8284387588500977, + "step": 4370 + }, + { + "epoch": 1.27362605408549, + "grad_norm": 1.760955810546875, + "learning_rate": 7.492130115424974e-05, + "loss": 0.7012954711914062, + "step": 4380 + }, + { + "epoch": 1.276533876126781, + "grad_norm": 1.8714261054992676, + "learning_rate": 7.462149602758208e-05, + "loss": 0.7935136795043946, + "step": 4390 + }, + { + "epoch": 1.279441698168072, + "grad_norm": 1.9436787366867065, + "learning_rate": 7.432169090091441e-05, + "loss": 0.7934539318084717, + "step": 4400 + }, + { + "epoch": 1.282349520209363, + "grad_norm": 1.9582066535949707, + "learning_rate": 7.402188577424674e-05, + "loss": 0.7850748538970947, + "step": 4410 + }, + { + "epoch": 1.2852573422506541, + "grad_norm": 1.9535675048828125, + "learning_rate": 7.372208064757908e-05, + "loss": 0.7332793712615967, + "step": 4420 + }, + { + "epoch": 1.2881651642919454, + "grad_norm": 1.7450060844421387, + "learning_rate": 7.342227552091141e-05, + "loss": 0.7462188720703125, + "step": 4430 + }, + { + "epoch": 1.2910729863332364, + "grad_norm": 1.8309375047683716, + "learning_rate": 7.312247039424375e-05, + "loss": 0.7612934112548828, + "step": 4440 + }, + { + "epoch": 1.2939808083745274, + "grad_norm": 2.1091737747192383, + "learning_rate": 7.282266526757608e-05, + "loss": 0.7570091247558594, + "step": 4450 + }, + { + "epoch": 1.2968886304158185, + "grad_norm": 1.933371901512146, + "learning_rate": 7.252286014090842e-05, + "loss": 0.7872342586517334, + "step": 4460 + }, + { + "epoch": 1.2997964524571097, + "grad_norm": 1.857756495475769, + "learning_rate": 7.222305501424075e-05, + "loss": 0.7867274761199952, + "step": 4470 + }, + { + "epoch": 1.3027042744984008, + "grad_norm": 1.8640364408493042, + "learning_rate": 7.192324988757309e-05, + "loss": 0.7677565574645996, + "step": 4480 + }, + { + "epoch": 1.3056120965396918, + "grad_norm": 1.8762586116790771, + "learning_rate": 7.162344476090541e-05, + "loss": 0.7633658885955811, + "step": 4490 + }, + { + "epoch": 1.3085199185809828, + "grad_norm": 2.0363333225250244, + "learning_rate": 7.132363963423776e-05, + "loss": 0.7173254013061523, + "step": 4500 + }, + { + "epoch": 1.3085199185809828, + "eval_loss": 0.8845488429069519, + "eval_runtime": 473.8247, + "eval_samples_per_second": 7.258, + "eval_steps_per_second": 7.258, + "step": 4500 + }, + { + "epoch": 1.3114277406222739, + "grad_norm": 2.0255188941955566, + "learning_rate": 7.102383450757009e-05, + "loss": 0.7590040683746337, + "step": 4510 + }, + { + "epoch": 1.3143355626635649, + "grad_norm": 2.052431583404541, + "learning_rate": 7.072402938090241e-05, + "loss": 0.7589908599853515, + "step": 4520 + }, + { + "epoch": 1.3172433847048561, + "grad_norm": 1.9324688911437988, + "learning_rate": 7.042422425423475e-05, + "loss": 0.7061916828155518, + "step": 4530 + }, + { + "epoch": 1.3201512067461472, + "grad_norm": 1.881585717201233, + "learning_rate": 7.012441912756709e-05, + "loss": 0.7603434562683106, + "step": 4540 + }, + { + "epoch": 1.3230590287874382, + "grad_norm": 2.1515471935272217, + "learning_rate": 6.982461400089941e-05, + "loss": 0.7740952491760253, + "step": 4550 + }, + { + "epoch": 1.3259668508287292, + "grad_norm": 1.9391210079193115, + "learning_rate": 6.952480887423175e-05, + "loss": 0.7603271007537842, + "step": 4560 + }, + { + "epoch": 1.3288746728700205, + "grad_norm": 1.9450794458389282, + "learning_rate": 6.922500374756408e-05, + "loss": 0.8049670219421386, + "step": 4570 + }, + { + "epoch": 1.3317824949113115, + "grad_norm": 2.054507255554199, + "learning_rate": 6.892519862089643e-05, + "loss": 0.7841194152832032, + "step": 4580 + }, + { + "epoch": 1.3346903169526025, + "grad_norm": 1.9239962100982666, + "learning_rate": 6.862539349422875e-05, + "loss": 0.7888126850128174, + "step": 4590 + }, + { + "epoch": 1.3375981389938936, + "grad_norm": 2.1011009216308594, + "learning_rate": 6.832558836756109e-05, + "loss": 0.7314376354217529, + "step": 4600 + }, + { + "epoch": 1.3405059610351846, + "grad_norm": 2.109041213989258, + "learning_rate": 6.802578324089342e-05, + "loss": 0.8058440208435058, + "step": 4610 + }, + { + "epoch": 1.3434137830764756, + "grad_norm": 1.9356318712234497, + "learning_rate": 6.772597811422576e-05, + "loss": 0.7847652435302734, + "step": 4620 + }, + { + "epoch": 1.3463216051177667, + "grad_norm": 2.092618942260742, + "learning_rate": 6.742617298755809e-05, + "loss": 0.7178999423980713, + "step": 4630 + }, + { + "epoch": 1.349229427159058, + "grad_norm": 1.8008778095245361, + "learning_rate": 6.712636786089042e-05, + "loss": 0.7678854465484619, + "step": 4640 + }, + { + "epoch": 1.352137249200349, + "grad_norm": 2.20794415473938, + "learning_rate": 6.682656273422276e-05, + "loss": 0.7475598335266114, + "step": 4650 + }, + { + "epoch": 1.35504507124164, + "grad_norm": 1.785436749458313, + "learning_rate": 6.65267576075551e-05, + "loss": 0.7328590393066406, + "step": 4660 + }, + { + "epoch": 1.357952893282931, + "grad_norm": 1.9333852529525757, + "learning_rate": 6.622695248088743e-05, + "loss": 0.7973165035247802, + "step": 4670 + }, + { + "epoch": 1.3608607153242223, + "grad_norm": 2.149812698364258, + "learning_rate": 6.592714735421976e-05, + "loss": 0.7523110389709473, + "step": 4680 + }, + { + "epoch": 1.3637685373655133, + "grad_norm": 2.0981409549713135, + "learning_rate": 6.56273422275521e-05, + "loss": 0.7891485691070557, + "step": 4690 + }, + { + "epoch": 1.3666763594068043, + "grad_norm": 2.0596706867218018, + "learning_rate": 6.532753710088443e-05, + "loss": 0.7680598735809326, + "step": 4700 + }, + { + "epoch": 1.3695841814480954, + "grad_norm": 2.0713438987731934, + "learning_rate": 6.502773197421677e-05, + "loss": 0.7212025165557862, + "step": 4710 + }, + { + "epoch": 1.3724920034893864, + "grad_norm": 1.933114767074585, + "learning_rate": 6.472792684754909e-05, + "loss": 0.8231717109680176, + "step": 4720 + }, + { + "epoch": 1.3753998255306774, + "grad_norm": 1.9748258590698242, + "learning_rate": 6.442812172088144e-05, + "loss": 0.710457706451416, + "step": 4730 + }, + { + "epoch": 1.3783076475719687, + "grad_norm": 1.99483323097229, + "learning_rate": 6.412831659421377e-05, + "loss": 0.7033157348632812, + "step": 4740 + }, + { + "epoch": 1.3812154696132597, + "grad_norm": 2.1739165782928467, + "learning_rate": 6.382851146754609e-05, + "loss": 0.7652640819549561, + "step": 4750 + }, + { + "epoch": 1.3841232916545507, + "grad_norm": 1.9803380966186523, + "learning_rate": 6.352870634087842e-05, + "loss": 0.7610398769378662, + "step": 4760 + }, + { + "epoch": 1.3870311136958418, + "grad_norm": 1.8965667486190796, + "learning_rate": 6.322890121421077e-05, + "loss": 0.8127297401428223, + "step": 4770 + }, + { + "epoch": 1.389938935737133, + "grad_norm": 2.0516648292541504, + "learning_rate": 6.292909608754311e-05, + "loss": 0.7877964973449707, + "step": 4780 + }, + { + "epoch": 1.392846757778424, + "grad_norm": 1.9544559717178345, + "learning_rate": 6.262929096087543e-05, + "loss": 0.7798377513885498, + "step": 4790 + }, + { + "epoch": 1.395754579819715, + "grad_norm": 1.7984205484390259, + "learning_rate": 6.232948583420776e-05, + "loss": 0.7166555404663086, + "step": 4800 + }, + { + "epoch": 1.398662401861006, + "grad_norm": 1.914433479309082, + "learning_rate": 6.202968070754011e-05, + "loss": 0.7964422225952148, + "step": 4810 + }, + { + "epoch": 1.4015702239022971, + "grad_norm": 2.067776679992676, + "learning_rate": 6.172987558087243e-05, + "loss": 0.7551724910736084, + "step": 4820 + }, + { + "epoch": 1.4044780459435882, + "grad_norm": 1.9109352827072144, + "learning_rate": 6.143007045420477e-05, + "loss": 0.7385217666625976, + "step": 4830 + }, + { + "epoch": 1.4073858679848792, + "grad_norm": 1.9591761827468872, + "learning_rate": 6.11302653275371e-05, + "loss": 0.721226167678833, + "step": 4840 + }, + { + "epoch": 1.4102936900261704, + "grad_norm": 2.115058183670044, + "learning_rate": 6.0830460200869435e-05, + "loss": 0.6777582168579102, + "step": 4850 + }, + { + "epoch": 1.4132015120674615, + "grad_norm": 2.06552791595459, + "learning_rate": 6.053065507420177e-05, + "loss": 0.726964282989502, + "step": 4860 + }, + { + "epoch": 1.4161093341087525, + "grad_norm": 2.1735804080963135, + "learning_rate": 6.0230849947534104e-05, + "loss": 0.7624166011810303, + "step": 4870 + }, + { + "epoch": 1.4190171561500435, + "grad_norm": 2.2036550045013428, + "learning_rate": 5.9931044820866446e-05, + "loss": 0.7692136287689209, + "step": 4880 + }, + { + "epoch": 1.4219249781913348, + "grad_norm": 1.8746955394744873, + "learning_rate": 5.9631239694198773e-05, + "loss": 0.7410698890686035, + "step": 4890 + }, + { + "epoch": 1.4248328002326258, + "grad_norm": 1.9036871194839478, + "learning_rate": 5.933143456753111e-05, + "loss": 0.7848501205444336, + "step": 4900 + }, + { + "epoch": 1.4277406222739168, + "grad_norm": 1.8983385562896729, + "learning_rate": 5.903162944086344e-05, + "loss": 0.7688620090484619, + "step": 4910 + }, + { + "epoch": 1.4306484443152079, + "grad_norm": 2.1584537029266357, + "learning_rate": 5.873182431419577e-05, + "loss": 0.7612662315368652, + "step": 4920 + }, + { + "epoch": 1.433556266356499, + "grad_norm": 2.033599376678467, + "learning_rate": 5.843201918752811e-05, + "loss": 0.7598844528198242, + "step": 4930 + }, + { + "epoch": 1.43646408839779, + "grad_norm": 1.9845435619354248, + "learning_rate": 5.8132214060860446e-05, + "loss": 0.7194857597351074, + "step": 4940 + }, + { + "epoch": 1.4393719104390812, + "grad_norm": 2.1257405281066895, + "learning_rate": 5.7832408934192774e-05, + "loss": 0.7137130737304688, + "step": 4950 + }, + { + "epoch": 1.4422797324803722, + "grad_norm": 1.8075517416000366, + "learning_rate": 5.753260380752511e-05, + "loss": 0.7500784873962403, + "step": 4960 + }, + { + "epoch": 1.4451875545216633, + "grad_norm": 2.136146068572998, + "learning_rate": 5.723279868085745e-05, + "loss": 0.7785260677337646, + "step": 4970 + }, + { + "epoch": 1.4480953765629543, + "grad_norm": 1.9265216588974, + "learning_rate": 5.6932993554189784e-05, + "loss": 0.6893830299377441, + "step": 4980 + }, + { + "epoch": 1.4510031986042455, + "grad_norm": 2.246067762374878, + "learning_rate": 5.663318842752211e-05, + "loss": 0.7811415672302247, + "step": 4990 + }, + { + "epoch": 1.4539110206455366, + "grad_norm": 2.068582534790039, + "learning_rate": 5.6333383300854446e-05, + "loss": 0.7605808258056641, + "step": 5000 + }, + { + "epoch": 1.4539110206455366, + "eval_loss": 0.8719142079353333, + "eval_runtime": 472.3067, + "eval_samples_per_second": 7.281, + "eval_steps_per_second": 7.281, + "step": 5000 + }, + { + "epoch": 1.4568188426868276, + "grad_norm": 1.819368600845337, + "learning_rate": 5.603357817418679e-05, + "loss": 0.6737122535705566, + "step": 5010 + }, + { + "epoch": 1.4597266647281186, + "grad_norm": 4.519766330718994, + "learning_rate": 5.573377304751911e-05, + "loss": 0.723225736618042, + "step": 5020 + }, + { + "epoch": 1.4626344867694097, + "grad_norm": 2.129854202270508, + "learning_rate": 5.543396792085145e-05, + "loss": 0.744614839553833, + "step": 5030 + }, + { + "epoch": 1.4655423088107007, + "grad_norm": 1.8777213096618652, + "learning_rate": 5.5134162794183784e-05, + "loss": 0.7135731220245362, + "step": 5040 + }, + { + "epoch": 1.4684501308519917, + "grad_norm": 1.9724328517913818, + "learning_rate": 5.483435766751611e-05, + "loss": 0.754638957977295, + "step": 5050 + }, + { + "epoch": 1.471357952893283, + "grad_norm": 1.8722237348556519, + "learning_rate": 5.4534552540848446e-05, + "loss": 0.7721257209777832, + "step": 5060 + }, + { + "epoch": 1.474265774934574, + "grad_norm": 2.2810893058776855, + "learning_rate": 5.423474741418079e-05, + "loss": 0.7189100742340088, + "step": 5070 + }, + { + "epoch": 1.477173596975865, + "grad_norm": 2.171797513961792, + "learning_rate": 5.393494228751312e-05, + "loss": 0.7469087600708008, + "step": 5080 + }, + { + "epoch": 1.4800814190171563, + "grad_norm": 1.837860107421875, + "learning_rate": 5.363513716084545e-05, + "loss": 0.743229341506958, + "step": 5090 + }, + { + "epoch": 1.4829892410584473, + "grad_norm": 1.96957528591156, + "learning_rate": 5.3335332034177784e-05, + "loss": 0.7477756977081299, + "step": 5100 + }, + { + "epoch": 1.4858970630997383, + "grad_norm": 2.0596463680267334, + "learning_rate": 5.3035526907510126e-05, + "loss": 0.7668298721313477, + "step": 5110 + }, + { + "epoch": 1.4888048851410294, + "grad_norm": 1.9528030157089233, + "learning_rate": 5.2735721780842453e-05, + "loss": 0.7141449451446533, + "step": 5120 + }, + { + "epoch": 1.4917127071823204, + "grad_norm": 1.8499971628189087, + "learning_rate": 5.243591665417479e-05, + "loss": 0.7491943359375, + "step": 5130 + }, + { + "epoch": 1.4946205292236114, + "grad_norm": 2.016645908355713, + "learning_rate": 5.213611152750712e-05, + "loss": 0.7747302055358887, + "step": 5140 + }, + { + "epoch": 1.4975283512649025, + "grad_norm": 1.8727567195892334, + "learning_rate": 5.183630640083945e-05, + "loss": 0.7703737735748291, + "step": 5150 + }, + { + "epoch": 1.5004361733061935, + "grad_norm": 1.9545795917510986, + "learning_rate": 5.153650127417179e-05, + "loss": 0.7513375282287598, + "step": 5160 + }, + { + "epoch": 1.5033439953474848, + "grad_norm": 2.101614236831665, + "learning_rate": 5.1236696147504126e-05, + "loss": 0.7481746196746826, + "step": 5170 + }, + { + "epoch": 1.5062518173887758, + "grad_norm": 1.8322865962982178, + "learning_rate": 5.093689102083646e-05, + "loss": 0.6998108386993408, + "step": 5180 + }, + { + "epoch": 1.509159639430067, + "grad_norm": 2.0042383670806885, + "learning_rate": 5.063708589416879e-05, + "loss": 0.7009212017059326, + "step": 5190 + }, + { + "epoch": 1.512067461471358, + "grad_norm": 2.0147664546966553, + "learning_rate": 5.033728076750113e-05, + "loss": 0.6757394313812256, + "step": 5200 + }, + { + "epoch": 1.514975283512649, + "grad_norm": 2.185133218765259, + "learning_rate": 5.0037475640833464e-05, + "loss": 0.8208953857421875, + "step": 5210 + }, + { + "epoch": 1.5178831055539401, + "grad_norm": 1.9892646074295044, + "learning_rate": 4.97376705141658e-05, + "loss": 0.7156825065612793, + "step": 5220 + }, + { + "epoch": 1.5207909275952312, + "grad_norm": 2.1808860301971436, + "learning_rate": 4.9437865387498126e-05, + "loss": 0.7173275470733642, + "step": 5230 + }, + { + "epoch": 1.5236987496365222, + "grad_norm": 2.2999000549316406, + "learning_rate": 4.913806026083047e-05, + "loss": 0.7130680084228516, + "step": 5240 + }, + { + "epoch": 1.5266065716778132, + "grad_norm": 1.9804385900497437, + "learning_rate": 4.8838255134162795e-05, + "loss": 0.7353660106658936, + "step": 5250 + }, + { + "epoch": 1.5295143937191042, + "grad_norm": 2.0175087451934814, + "learning_rate": 4.853845000749513e-05, + "loss": 0.6823455333709717, + "step": 5260 + }, + { + "epoch": 1.5324222157603955, + "grad_norm": 1.9505001306533813, + "learning_rate": 4.8238644880827464e-05, + "loss": 0.7360805034637451, + "step": 5270 + }, + { + "epoch": 1.5353300378016865, + "grad_norm": 2.100215435028076, + "learning_rate": 4.79388397541598e-05, + "loss": 0.761193037033081, + "step": 5280 + }, + { + "epoch": 1.5382378598429776, + "grad_norm": 1.9561179876327515, + "learning_rate": 4.763903462749213e-05, + "loss": 0.8232732772827148, + "step": 5290 + }, + { + "epoch": 1.5411456818842688, + "grad_norm": 2.07002592086792, + "learning_rate": 4.733922950082447e-05, + "loss": 0.7684964656829834, + "step": 5300 + }, + { + "epoch": 1.5440535039255598, + "grad_norm": 2.175922393798828, + "learning_rate": 4.7039424374156795e-05, + "loss": 0.743954849243164, + "step": 5310 + }, + { + "epoch": 1.5469613259668509, + "grad_norm": 1.8928070068359375, + "learning_rate": 4.673961924748914e-05, + "loss": 0.8028405189514161, + "step": 5320 + }, + { + "epoch": 1.549869148008142, + "grad_norm": 1.8495947122573853, + "learning_rate": 4.6439814120821464e-05, + "loss": 0.7481747150421143, + "step": 5330 + }, + { + "epoch": 1.552776970049433, + "grad_norm": 2.0875566005706787, + "learning_rate": 4.6140008994153806e-05, + "loss": 0.7654068470001221, + "step": 5340 + }, + { + "epoch": 1.555684792090724, + "grad_norm": 2.132359027862549, + "learning_rate": 4.5840203867486133e-05, + "loss": 0.7879802703857421, + "step": 5350 + }, + { + "epoch": 1.558592614132015, + "grad_norm": 1.9833087921142578, + "learning_rate": 4.554039874081847e-05, + "loss": 0.7630073547363281, + "step": 5360 + }, + { + "epoch": 1.561500436173306, + "grad_norm": 2.261916399002075, + "learning_rate": 4.52405936141508e-05, + "loss": 0.7673455238342285, + "step": 5370 + }, + { + "epoch": 1.5644082582145973, + "grad_norm": 1.9294761419296265, + "learning_rate": 4.494078848748314e-05, + "loss": 0.7796091556549072, + "step": 5380 + }, + { + "epoch": 1.5673160802558883, + "grad_norm": 1.8864835500717163, + "learning_rate": 4.464098336081548e-05, + "loss": 0.7843995094299316, + "step": 5390 + }, + { + "epoch": 1.5702239022971796, + "grad_norm": 1.6363269090652466, + "learning_rate": 4.4341178234147806e-05, + "loss": 0.7239855766296387, + "step": 5400 + }, + { + "epoch": 1.5731317243384706, + "grad_norm": 1.910172462463379, + "learning_rate": 4.404137310748014e-05, + "loss": 0.684738302230835, + "step": 5410 + }, + { + "epoch": 1.5760395463797616, + "grad_norm": 1.936583161354065, + "learning_rate": 4.3741567980812475e-05, + "loss": 0.7819816112518311, + "step": 5420 + }, + { + "epoch": 1.5789473684210527, + "grad_norm": 2.261993169784546, + "learning_rate": 4.344176285414481e-05, + "loss": 0.7665531158447265, + "step": 5430 + }, + { + "epoch": 1.5818551904623437, + "grad_norm": 2.1516878604888916, + "learning_rate": 4.314195772747714e-05, + "loss": 0.7250272274017334, + "step": 5440 + }, + { + "epoch": 1.5847630125036347, + "grad_norm": 1.9359904527664185, + "learning_rate": 4.284215260080948e-05, + "loss": 0.7476921081542969, + "step": 5450 + }, + { + "epoch": 1.5876708345449257, + "grad_norm": 1.9610286951065063, + "learning_rate": 4.2542347474141806e-05, + "loss": 0.747294282913208, + "step": 5460 + }, + { + "epoch": 1.5905786565862168, + "grad_norm": 1.8940902948379517, + "learning_rate": 4.224254234747415e-05, + "loss": 0.7050137519836426, + "step": 5470 + }, + { + "epoch": 1.593486478627508, + "grad_norm": 2.1721606254577637, + "learning_rate": 4.1942737220806475e-05, + "loss": 0.7259575366973877, + "step": 5480 + }, + { + "epoch": 1.596394300668799, + "grad_norm": 2.0317630767822266, + "learning_rate": 4.164293209413881e-05, + "loss": 0.6914929389953614, + "step": 5490 + }, + { + "epoch": 1.59930212271009, + "grad_norm": 2.04887056350708, + "learning_rate": 4.1343126967471144e-05, + "loss": 0.7758480548858643, + "step": 5500 + }, + { + "epoch": 1.59930212271009, + "eval_loss": 0.8598825931549072, + "eval_runtime": 474.1318, + "eval_samples_per_second": 7.253, + "eval_steps_per_second": 7.253, + "step": 5500 + }, + { + "epoch": 1.6022099447513813, + "grad_norm": 1.7526671886444092, + "learning_rate": 4.104332184080348e-05, + "loss": 0.6790880680084228, + "step": 5510 + }, + { + "epoch": 1.6051177667926724, + "grad_norm": 2.0832862854003906, + "learning_rate": 4.074351671413581e-05, + "loss": 0.7566723346710205, + "step": 5520 + }, + { + "epoch": 1.6080255888339634, + "grad_norm": 1.8792216777801514, + "learning_rate": 4.044371158746815e-05, + "loss": 0.6554183959960938, + "step": 5530 + }, + { + "epoch": 1.6109334108752544, + "grad_norm": 2.0853688716888428, + "learning_rate": 4.014390646080048e-05, + "loss": 0.7421646118164062, + "step": 5540 + }, + { + "epoch": 1.6138412329165455, + "grad_norm": 2.5147030353546143, + "learning_rate": 3.984410133413282e-05, + "loss": 0.7599525451660156, + "step": 5550 + }, + { + "epoch": 1.6167490549578365, + "grad_norm": 2.236278772354126, + "learning_rate": 3.954429620746515e-05, + "loss": 0.8087862014770508, + "step": 5560 + }, + { + "epoch": 1.6196568769991275, + "grad_norm": 1.9631987810134888, + "learning_rate": 3.9244491080797486e-05, + "loss": 0.7550980091094971, + "step": 5570 + }, + { + "epoch": 1.6225646990404188, + "grad_norm": 1.8814702033996582, + "learning_rate": 3.894468595412982e-05, + "loss": 0.7769564628601074, + "step": 5580 + }, + { + "epoch": 1.6254725210817098, + "grad_norm": 1.9273751974105835, + "learning_rate": 3.864488082746215e-05, + "loss": 0.7157449722290039, + "step": 5590 + }, + { + "epoch": 1.6283803431230008, + "grad_norm": 1.9550398588180542, + "learning_rate": 3.834507570079449e-05, + "loss": 0.793599796295166, + "step": 5600 + }, + { + "epoch": 1.631288165164292, + "grad_norm": 1.9033679962158203, + "learning_rate": 3.804527057412682e-05, + "loss": 0.7436941146850586, + "step": 5610 + }, + { + "epoch": 1.6341959872055831, + "grad_norm": 1.9171406030654907, + "learning_rate": 3.774546544745916e-05, + "loss": 0.6863879203796387, + "step": 5620 + }, + { + "epoch": 1.6371038092468742, + "grad_norm": 1.7775182723999023, + "learning_rate": 3.7445660320791486e-05, + "loss": 0.7105512142181396, + "step": 5630 + }, + { + "epoch": 1.6400116312881652, + "grad_norm": 2.081458568572998, + "learning_rate": 3.714585519412382e-05, + "loss": 0.679707384109497, + "step": 5640 + }, + { + "epoch": 1.6429194533294562, + "grad_norm": 2.0358259677886963, + "learning_rate": 3.6846050067456155e-05, + "loss": 0.7309046745300293, + "step": 5650 + }, + { + "epoch": 1.6458272753707472, + "grad_norm": 2.1495471000671387, + "learning_rate": 3.654624494078849e-05, + "loss": 0.678877878189087, + "step": 5660 + }, + { + "epoch": 1.6487350974120383, + "grad_norm": 2.19714617729187, + "learning_rate": 3.6246439814120824e-05, + "loss": 0.724186372756958, + "step": 5670 + }, + { + "epoch": 1.6516429194533293, + "grad_norm": 1.930059552192688, + "learning_rate": 3.594663468745316e-05, + "loss": 0.7309486389160156, + "step": 5680 + }, + { + "epoch": 1.6545507414946206, + "grad_norm": 2.045048236846924, + "learning_rate": 3.5646829560785486e-05, + "loss": 0.7321044921875, + "step": 5690 + }, + { + "epoch": 1.6574585635359116, + "grad_norm": 1.8372198343276978, + "learning_rate": 3.534702443411783e-05, + "loss": 0.690158462524414, + "step": 5700 + }, + { + "epoch": 1.6603663855772028, + "grad_norm": 2.2258548736572266, + "learning_rate": 3.5047219307450155e-05, + "loss": 0.6976329803466796, + "step": 5710 + }, + { + "epoch": 1.6632742076184939, + "grad_norm": 1.861602544784546, + "learning_rate": 3.4747414180782496e-05, + "loss": 0.7170249462127686, + "step": 5720 + }, + { + "epoch": 1.666182029659785, + "grad_norm": 1.9709389209747314, + "learning_rate": 3.4447609054114824e-05, + "loss": 0.7029520511627197, + "step": 5730 + }, + { + "epoch": 1.669089851701076, + "grad_norm": 1.8129233121871948, + "learning_rate": 3.414780392744716e-05, + "loss": 0.7850728988647461, + "step": 5740 + }, + { + "epoch": 1.671997673742367, + "grad_norm": 2.106326103210449, + "learning_rate": 3.384799880077949e-05, + "loss": 0.8082990646362305, + "step": 5750 + }, + { + "epoch": 1.674905495783658, + "grad_norm": 1.839580774307251, + "learning_rate": 3.354819367411183e-05, + "loss": 0.7232970237731934, + "step": 5760 + }, + { + "epoch": 1.677813317824949, + "grad_norm": 2.027674674987793, + "learning_rate": 3.324838854744416e-05, + "loss": 0.7808068752288818, + "step": 5770 + }, + { + "epoch": 1.68072113986624, + "grad_norm": 2.2386841773986816, + "learning_rate": 3.29485834207765e-05, + "loss": 0.711140775680542, + "step": 5780 + }, + { + "epoch": 1.6836289619075313, + "grad_norm": 2.136626958847046, + "learning_rate": 3.264877829410883e-05, + "loss": 0.7360716342926026, + "step": 5790 + }, + { + "epoch": 1.6865367839488223, + "grad_norm": 2.048578977584839, + "learning_rate": 3.2348973167441166e-05, + "loss": 0.710478401184082, + "step": 5800 + }, + { + "epoch": 1.6894446059901134, + "grad_norm": 1.9906225204467773, + "learning_rate": 3.20491680407735e-05, + "loss": 0.7086396217346191, + "step": 5810 + }, + { + "epoch": 1.6923524280314046, + "grad_norm": 1.964224934577942, + "learning_rate": 3.1749362914105835e-05, + "loss": 0.7486394882202149, + "step": 5820 + }, + { + "epoch": 1.6952602500726957, + "grad_norm": 1.887143850326538, + "learning_rate": 3.144955778743817e-05, + "loss": 0.6863440990447998, + "step": 5830 + }, + { + "epoch": 1.6981680721139867, + "grad_norm": 2.00508713722229, + "learning_rate": 3.11497526607705e-05, + "loss": 0.6853577613830566, + "step": 5840 + }, + { + "epoch": 1.7010758941552777, + "grad_norm": 2.0328657627105713, + "learning_rate": 3.084994753410284e-05, + "loss": 0.7252396583557129, + "step": 5850 + }, + { + "epoch": 1.7039837161965687, + "grad_norm": 2.1302683353424072, + "learning_rate": 3.0550142407435166e-05, + "loss": 0.7472237586975098, + "step": 5860 + }, + { + "epoch": 1.7068915382378598, + "grad_norm": 2.106820821762085, + "learning_rate": 3.0250337280767504e-05, + "loss": 0.723507022857666, + "step": 5870 + }, + { + "epoch": 1.7097993602791508, + "grad_norm": 2.0230047702789307, + "learning_rate": 2.9950532154099835e-05, + "loss": 0.6797329425811768, + "step": 5880 + }, + { + "epoch": 1.7127071823204418, + "grad_norm": 2.121976613998413, + "learning_rate": 2.965072702743217e-05, + "loss": 0.7394673347473144, + "step": 5890 + }, + { + "epoch": 1.715615004361733, + "grad_norm": 1.8224197626113892, + "learning_rate": 2.9350921900764504e-05, + "loss": 0.758949613571167, + "step": 5900 + }, + { + "epoch": 1.7185228264030241, + "grad_norm": 1.832812786102295, + "learning_rate": 2.905111677409684e-05, + "loss": 0.724758243560791, + "step": 5910 + }, + { + "epoch": 1.7214306484443154, + "grad_norm": 2.112184524536133, + "learning_rate": 2.8751311647429173e-05, + "loss": 0.7687956809997558, + "step": 5920 + }, + { + "epoch": 1.7243384704856064, + "grad_norm": 2.5346004962921143, + "learning_rate": 2.8451506520761507e-05, + "loss": 0.721607780456543, + "step": 5930 + }, + { + "epoch": 1.7272462925268974, + "grad_norm": 1.9310683012008667, + "learning_rate": 2.815170139409384e-05, + "loss": 0.750163459777832, + "step": 5940 + }, + { + "epoch": 1.7301541145681885, + "grad_norm": 2.023061752319336, + "learning_rate": 2.7851896267426176e-05, + "loss": 0.7398929119110107, + "step": 5950 + }, + { + "epoch": 1.7330619366094795, + "grad_norm": 1.9423258304595947, + "learning_rate": 2.7552091140758507e-05, + "loss": 0.7022686004638672, + "step": 5960 + }, + { + "epoch": 1.7359697586507705, + "grad_norm": 1.9257018566131592, + "learning_rate": 2.7252286014090845e-05, + "loss": 0.760784387588501, + "step": 5970 + }, + { + "epoch": 1.7388775806920616, + "grad_norm": 1.9650042057037354, + "learning_rate": 2.6952480887423176e-05, + "loss": 0.7613511562347413, + "step": 5980 + }, + { + "epoch": 1.7417854027333526, + "grad_norm": 2.08854341506958, + "learning_rate": 2.6652675760755508e-05, + "loss": 0.7318148612976074, + "step": 5990 + }, + { + "epoch": 1.7446932247746438, + "grad_norm": 2.0090866088867188, + "learning_rate": 2.6352870634087845e-05, + "loss": 0.7362863540649414, + "step": 6000 + }, + { + "epoch": 1.7446932247746438, + "eval_loss": 0.8498985171318054, + "eval_runtime": 471.5916, + "eval_samples_per_second": 7.292, + "eval_steps_per_second": 7.292, + "step": 6000 + }, + { + "epoch": 1.7476010468159349, + "grad_norm": 1.9278331995010376, + "learning_rate": 2.6053065507420177e-05, + "loss": 0.6806503295898437, + "step": 6010 + }, + { + "epoch": 1.750508868857226, + "grad_norm": 2.056682825088501, + "learning_rate": 2.5753260380752514e-05, + "loss": 0.722940731048584, + "step": 6020 + }, + { + "epoch": 1.7534166908985171, + "grad_norm": 1.9063704013824463, + "learning_rate": 2.5453455254084846e-05, + "loss": 0.6798623085021973, + "step": 6030 + }, + { + "epoch": 1.7563245129398082, + "grad_norm": 1.8881698846817017, + "learning_rate": 2.5153650127417177e-05, + "loss": 0.7303658485412597, + "step": 6040 + }, + { + "epoch": 1.7592323349810992, + "grad_norm": 2.0086960792541504, + "learning_rate": 2.4853845000749515e-05, + "loss": 0.7225898265838623, + "step": 6050 + }, + { + "epoch": 1.7621401570223902, + "grad_norm": 2.1029181480407715, + "learning_rate": 2.4554039874081846e-05, + "loss": 0.7572491645812989, + "step": 6060 + }, + { + "epoch": 1.7650479790636813, + "grad_norm": 1.9289077520370483, + "learning_rate": 2.425423474741418e-05, + "loss": 0.780693531036377, + "step": 6070 + }, + { + "epoch": 1.7679558011049723, + "grad_norm": 2.0310537815093994, + "learning_rate": 2.3954429620746515e-05, + "loss": 0.7033182144165039, + "step": 6080 + }, + { + "epoch": 1.7708636231462633, + "grad_norm": 2.0626399517059326, + "learning_rate": 2.365462449407885e-05, + "loss": 0.7750870227813721, + "step": 6090 + }, + { + "epoch": 1.7737714451875544, + "grad_norm": 1.9674657583236694, + "learning_rate": 2.3354819367411184e-05, + "loss": 0.7491857051849365, + "step": 6100 + }, + { + "epoch": 1.7766792672288456, + "grad_norm": 2.028435707092285, + "learning_rate": 2.3055014240743518e-05, + "loss": 0.7668484687805176, + "step": 6110 + }, + { + "epoch": 1.7795870892701366, + "grad_norm": 1.8845252990722656, + "learning_rate": 2.2755209114075853e-05, + "loss": 0.7411411285400391, + "step": 6120 + }, + { + "epoch": 1.782494911311428, + "grad_norm": 2.013805627822876, + "learning_rate": 2.2455403987408187e-05, + "loss": 0.7249200344085693, + "step": 6130 + }, + { + "epoch": 1.785402733352719, + "grad_norm": 2.006671667098999, + "learning_rate": 2.2155598860740522e-05, + "loss": 0.7829705715179444, + "step": 6140 + }, + { + "epoch": 1.78831055539401, + "grad_norm": 2.0205607414245605, + "learning_rate": 2.1855793734072856e-05, + "loss": 0.7822526454925537, + "step": 6150 + }, + { + "epoch": 1.791218377435301, + "grad_norm": 1.9972530603408813, + "learning_rate": 2.1555988607405187e-05, + "loss": 0.7459287166595459, + "step": 6160 + }, + { + "epoch": 1.794126199476592, + "grad_norm": 2.1279802322387695, + "learning_rate": 2.1256183480737522e-05, + "loss": 0.7397714614868164, + "step": 6170 + }, + { + "epoch": 1.797034021517883, + "grad_norm": 2.1933655738830566, + "learning_rate": 2.0956378354069856e-05, + "loss": 0.6955758571624756, + "step": 6180 + }, + { + "epoch": 1.799941843559174, + "grad_norm": 1.9316877126693726, + "learning_rate": 2.065657322740219e-05, + "loss": 0.727911376953125, + "step": 6190 + }, + { + "epoch": 1.8028496656004651, + "grad_norm": 1.8654370307922363, + "learning_rate": 2.0356768100734525e-05, + "loss": 0.6596095561981201, + "step": 6200 + }, + { + "epoch": 1.8057574876417564, + "grad_norm": 2.046983003616333, + "learning_rate": 2.0056962974066856e-05, + "loss": 0.7094531059265137, + "step": 6210 + }, + { + "epoch": 1.8086653096830474, + "grad_norm": 2.1500179767608643, + "learning_rate": 1.975715784739919e-05, + "loss": 0.7008066654205323, + "step": 6220 + }, + { + "epoch": 1.8115731317243384, + "grad_norm": 2.3866703510284424, + "learning_rate": 1.9457352720731525e-05, + "loss": 0.7228167533874512, + "step": 6230 + }, + { + "epoch": 1.8144809537656297, + "grad_norm": 2.0172855854034424, + "learning_rate": 1.915754759406386e-05, + "loss": 0.6767440795898437, + "step": 6240 + }, + { + "epoch": 1.8173887758069207, + "grad_norm": 1.8794561624526978, + "learning_rate": 1.8857742467396194e-05, + "loss": 0.7008272647857666, + "step": 6250 + }, + { + "epoch": 1.8202965978482117, + "grad_norm": 2.0600626468658447, + "learning_rate": 1.8557937340728526e-05, + "loss": 0.7082679748535157, + "step": 6260 + }, + { + "epoch": 1.8232044198895028, + "grad_norm": 1.869964599609375, + "learning_rate": 1.825813221406086e-05, + "loss": 0.7198378562927246, + "step": 6270 + }, + { + "epoch": 1.8261122419307938, + "grad_norm": 1.9929425716400146, + "learning_rate": 1.7958327087393195e-05, + "loss": 0.7699549674987793, + "step": 6280 + }, + { + "epoch": 1.8290200639720848, + "grad_norm": 1.9782027006149292, + "learning_rate": 1.765852196072553e-05, + "loss": 0.6976008892059327, + "step": 6290 + }, + { + "epoch": 1.8319278860133759, + "grad_norm": 2.1247029304504395, + "learning_rate": 1.7358716834057864e-05, + "loss": 0.690690565109253, + "step": 6300 + }, + { + "epoch": 1.834835708054667, + "grad_norm": 1.8903204202651978, + "learning_rate": 1.7058911707390195e-05, + "loss": 0.7616622447967529, + "step": 6310 + }, + { + "epoch": 1.8377435300959581, + "grad_norm": 2.0401735305786133, + "learning_rate": 1.675910658072253e-05, + "loss": 0.6766151428222656, + "step": 6320 + }, + { + "epoch": 1.8406513521372492, + "grad_norm": 1.9508399963378906, + "learning_rate": 1.6459301454054864e-05, + "loss": 0.7218796730041503, + "step": 6330 + }, + { + "epoch": 1.8435591741785404, + "grad_norm": 2.1602697372436523, + "learning_rate": 1.6159496327387198e-05, + "loss": 0.753945779800415, + "step": 6340 + }, + { + "epoch": 1.8464669962198315, + "grad_norm": 2.0386970043182373, + "learning_rate": 1.5859691200719533e-05, + "loss": 0.7185985088348389, + "step": 6350 + }, + { + "epoch": 1.8493748182611225, + "grad_norm": 1.8012441396713257, + "learning_rate": 1.5559886074051867e-05, + "loss": 0.777646541595459, + "step": 6360 + }, + { + "epoch": 1.8522826403024135, + "grad_norm": 1.8146647214889526, + "learning_rate": 1.52600809473842e-05, + "loss": 0.7361874580383301, + "step": 6370 + }, + { + "epoch": 1.8551904623437045, + "grad_norm": 2.180405616760254, + "learning_rate": 1.4960275820716534e-05, + "loss": 0.6907386779785156, + "step": 6380 + }, + { + "epoch": 1.8580982843849956, + "grad_norm": 2.1181304454803467, + "learning_rate": 1.4660470694048869e-05, + "loss": 0.6983653545379639, + "step": 6390 + }, + { + "epoch": 1.8610061064262866, + "grad_norm": 1.8451800346374512, + "learning_rate": 1.4360665567381203e-05, + "loss": 0.6979443073272705, + "step": 6400 + }, + { + "epoch": 1.8639139284675776, + "grad_norm": 2.0219805240631104, + "learning_rate": 1.4060860440713536e-05, + "loss": 0.6582244396209717, + "step": 6410 + }, + { + "epoch": 1.866821750508869, + "grad_norm": 1.9723000526428223, + "learning_rate": 1.376105531404587e-05, + "loss": 0.7287851333618164, + "step": 6420 + }, + { + "epoch": 1.86972957255016, + "grad_norm": 2.274547815322876, + "learning_rate": 1.3461250187378205e-05, + "loss": 0.7135519027709961, + "step": 6430 + }, + { + "epoch": 1.872637394591451, + "grad_norm": 1.9583326578140259, + "learning_rate": 1.316144506071054e-05, + "loss": 0.6913717746734619, + "step": 6440 + }, + { + "epoch": 1.8755452166327422, + "grad_norm": 2.0431573390960693, + "learning_rate": 1.2861639934042874e-05, + "loss": 0.7512425422668457, + "step": 6450 + }, + { + "epoch": 1.8784530386740332, + "grad_norm": 2.096263885498047, + "learning_rate": 1.2561834807375205e-05, + "loss": 0.705587911605835, + "step": 6460 + }, + { + "epoch": 1.8813608607153243, + "grad_norm": 1.940988302230835, + "learning_rate": 1.226202968070754e-05, + "loss": 0.7097304344177247, + "step": 6470 + }, + { + "epoch": 1.8842686827566153, + "grad_norm": 1.9227349758148193, + "learning_rate": 1.1962224554039874e-05, + "loss": 0.7371804237365722, + "step": 6480 + }, + { + "epoch": 1.8871765047979063, + "grad_norm": 1.936657428741455, + "learning_rate": 1.1662419427372209e-05, + "loss": 0.763831090927124, + "step": 6490 + }, + { + "epoch": 1.8900843268391974, + "grad_norm": 2.0900590419769287, + "learning_rate": 1.1362614300704542e-05, + "loss": 0.7716608047485352, + "step": 6500 + }, + { + "epoch": 1.8900843268391974, + "eval_loss": 0.8433617949485779, + "eval_runtime": 473.6198, + "eval_samples_per_second": 7.261, + "eval_steps_per_second": 7.261, + "step": 6500 + } + ], + "logging_steps": 10, + "max_steps": 6878, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.785931355956032e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}