diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,70034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 6.551791915088777, + "eval_steps": 500, + "global_step": 100000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0006551791915088776, + "grad_norm": 1.804868221282959, + "learning_rate": 1.8e-07, + "loss": 1.465, + "step": 10 + }, + { + "epoch": 0.0013103583830177553, + "grad_norm": 0.5462529063224792, + "learning_rate": 3.8e-07, + "loss": 1.2127, + "step": 20 + }, + { + "epoch": 0.001965537574526633, + "grad_norm": 0.6241850256919861, + "learning_rate": 5.8e-07, + "loss": 1.0918, + "step": 30 + }, + { + "epoch": 0.0026207167660355106, + "grad_norm": 1.1952224969863892, + "learning_rate": 7.8e-07, + "loss": 1.0696, + "step": 40 + }, + { + "epoch": 0.0032758959575443884, + "grad_norm": 1.1293015480041504, + "learning_rate": 9.8e-07, + "loss": 1.2151, + "step": 50 + }, + { + "epoch": 0.003931075149053266, + "grad_norm": 1.4693841934204102, + "learning_rate": 1.18e-06, + "loss": 1.2503, + "step": 60 + }, + { + "epoch": 0.004586254340562144, + "grad_norm": 0.4732355773448944, + "learning_rate": 1.3800000000000001e-06, + "loss": 1.0424, + "step": 70 + }, + { + "epoch": 0.005241433532071021, + "grad_norm": 1.668462872505188, + "learning_rate": 1.5800000000000003e-06, + "loss": 1.1919, + "step": 80 + }, + { + "epoch": 0.005896612723579899, + "grad_norm": 0.48098990321159363, + "learning_rate": 1.7800000000000001e-06, + "loss": 1.2546, + "step": 90 + }, + { + "epoch": 0.006551791915088777, + "grad_norm": 1.2686821222305298, + "learning_rate": 1.98e-06, + "loss": 1.5056, + "step": 100 + }, + { + "epoch": 0.007206971106597655, + "grad_norm": 0.3098691701889038, + "learning_rate": 2.1800000000000003e-06, + "loss": 1.1809, + "step": 110 + }, + { + "epoch": 0.007862150298106533, + "grad_norm": 1.8648558855056763, + "learning_rate": 2.38e-06, + "loss": 1.341, + "step": 120 + }, + { + "epoch": 0.00851732948961541, + "grad_norm": 0.4359284043312073, + "learning_rate": 2.5800000000000003e-06, + "loss": 1.3649, + "step": 130 + }, + { + "epoch": 0.009172508681124288, + "grad_norm": 0.40405988693237305, + "learning_rate": 2.78e-06, + "loss": 1.1425, + "step": 140 + }, + { + "epoch": 0.009827687872633165, + "grad_norm": 0.5246503353118896, + "learning_rate": 2.9800000000000003e-06, + "loss": 1.145, + "step": 150 + }, + { + "epoch": 0.010482867064142042, + "grad_norm": 0.7996739149093628, + "learning_rate": 3.1800000000000005e-06, + "loss": 1.3076, + "step": 160 + }, + { + "epoch": 0.011138046255650921, + "grad_norm": 1.405401349067688, + "learning_rate": 3.38e-06, + "loss": 1.3294, + "step": 170 + }, + { + "epoch": 0.011793225447159798, + "grad_norm": 2.2338318824768066, + "learning_rate": 3.58e-06, + "loss": 1.2499, + "step": 180 + }, + { + "epoch": 0.012448404638668677, + "grad_norm": 1.0245256423950195, + "learning_rate": 3.7800000000000002e-06, + "loss": 1.0985, + "step": 190 + }, + { + "epoch": 0.013103583830177554, + "grad_norm": 1.721191644668579, + "learning_rate": 3.98e-06, + "loss": 1.5595, + "step": 200 + }, + { + "epoch": 0.01375876302168643, + "grad_norm": 2.5674519538879395, + "learning_rate": 4.18e-06, + "loss": 1.3005, + "step": 210 + }, + { + "epoch": 0.01441394221319531, + "grad_norm": 3.184626579284668, + "learning_rate": 4.38e-06, + "loss": 1.1717, + "step": 220 + }, + { + "epoch": 0.015069121404704186, + "grad_norm": 1.2653679847717285, + "learning_rate": 4.58e-06, + "loss": 1.1619, + "step": 230 + }, + { + "epoch": 0.015724300596213065, + "grad_norm": 1.0437815189361572, + "learning_rate": 4.780000000000001e-06, + "loss": 0.9659, + "step": 240 + }, + { + "epoch": 0.016379479787721942, + "grad_norm": 1.3685284852981567, + "learning_rate": 4.98e-06, + "loss": 1.1575, + "step": 250 + }, + { + "epoch": 0.01703465897923082, + "grad_norm": 3.6203811168670654, + "learning_rate": 5.18e-06, + "loss": 1.3075, + "step": 260 + }, + { + "epoch": 0.017689838170739696, + "grad_norm": 2.4867584705352783, + "learning_rate": 5.38e-06, + "loss": 0.9825, + "step": 270 + }, + { + "epoch": 0.018345017362248577, + "grad_norm": 2.092862129211426, + "learning_rate": 5.580000000000001e-06, + "loss": 1.1472, + "step": 280 + }, + { + "epoch": 0.019000196553757454, + "grad_norm": 2.0016863346099854, + "learning_rate": 5.78e-06, + "loss": 0.8115, + "step": 290 + }, + { + "epoch": 0.01965537574526633, + "grad_norm": 1.5526832342147827, + "learning_rate": 5.98e-06, + "loss": 0.9646, + "step": 300 + }, + { + "epoch": 0.020310554936775208, + "grad_norm": 2.1613259315490723, + "learning_rate": 6.18e-06, + "loss": 0.9332, + "step": 310 + }, + { + "epoch": 0.020965734128284085, + "grad_norm": 1.3020223379135132, + "learning_rate": 6.38e-06, + "loss": 0.7171, + "step": 320 + }, + { + "epoch": 0.021620913319792965, + "grad_norm": 1.837485432624817, + "learning_rate": 6.58e-06, + "loss": 0.9526, + "step": 330 + }, + { + "epoch": 0.022276092511301842, + "grad_norm": 1.194470763206482, + "learning_rate": 6.78e-06, + "loss": 0.6036, + "step": 340 + }, + { + "epoch": 0.02293127170281072, + "grad_norm": 1.9005872011184692, + "learning_rate": 6.98e-06, + "loss": 0.6584, + "step": 350 + }, + { + "epoch": 0.023586450894319596, + "grad_norm": 1.2289026975631714, + "learning_rate": 7.180000000000001e-06, + "loss": 0.6156, + "step": 360 + }, + { + "epoch": 0.024241630085828473, + "grad_norm": 1.1380399465560913, + "learning_rate": 7.3800000000000005e-06, + "loss": 0.6313, + "step": 370 + }, + { + "epoch": 0.024896809277337353, + "grad_norm": 1.2419391870498657, + "learning_rate": 7.580000000000001e-06, + "loss": 0.6012, + "step": 380 + }, + { + "epoch": 0.02555198846884623, + "grad_norm": 0.934105396270752, + "learning_rate": 7.78e-06, + "loss": 0.668, + "step": 390 + }, + { + "epoch": 0.026207167660355107, + "grad_norm": 1.6521079540252686, + "learning_rate": 7.98e-06, + "loss": 0.6498, + "step": 400 + }, + { + "epoch": 0.026862346851863984, + "grad_norm": 1.7710031270980835, + "learning_rate": 8.18e-06, + "loss": 0.5962, + "step": 410 + }, + { + "epoch": 0.02751752604337286, + "grad_norm": 1.3258167505264282, + "learning_rate": 8.380000000000001e-06, + "loss": 0.6294, + "step": 420 + }, + { + "epoch": 0.028172705234881742, + "grad_norm": 1.5607084035873413, + "learning_rate": 8.580000000000001e-06, + "loss": 0.5492, + "step": 430 + }, + { + "epoch": 0.02882788442639062, + "grad_norm": 0.9044286012649536, + "learning_rate": 8.78e-06, + "loss": 0.453, + "step": 440 + }, + { + "epoch": 0.029483063617899496, + "grad_norm": 1.4217227697372437, + "learning_rate": 8.98e-06, + "loss": 0.4627, + "step": 450 + }, + { + "epoch": 0.030138242809408373, + "grad_norm": 1.0413070917129517, + "learning_rate": 9.180000000000002e-06, + "loss": 0.4887, + "step": 460 + }, + { + "epoch": 0.03079342200091725, + "grad_norm": 1.306139349937439, + "learning_rate": 9.38e-06, + "loss": 0.4412, + "step": 470 + }, + { + "epoch": 0.03144860119242613, + "grad_norm": 1.7404978275299072, + "learning_rate": 9.58e-06, + "loss": 0.5238, + "step": 480 + }, + { + "epoch": 0.03210378038393501, + "grad_norm": 1.6194000244140625, + "learning_rate": 9.78e-06, + "loss": 0.417, + "step": 490 + }, + { + "epoch": 0.032758959575443884, + "grad_norm": 1.0762255191802979, + "learning_rate": 9.980000000000001e-06, + "loss": 0.4586, + "step": 500 + }, + { + "epoch": 0.03341413876695276, + "grad_norm": 1.0854724645614624, + "learning_rate": 1.018e-05, + "loss": 0.4811, + "step": 510 + }, + { + "epoch": 0.03406931795846164, + "grad_norm": 1.134367823600769, + "learning_rate": 1.038e-05, + "loss": 0.315, + "step": 520 + }, + { + "epoch": 0.034724497149970515, + "grad_norm": 1.4098827838897705, + "learning_rate": 1.058e-05, + "loss": 0.3222, + "step": 530 + }, + { + "epoch": 0.03537967634147939, + "grad_norm": 2.3033673763275146, + "learning_rate": 1.0780000000000002e-05, + "loss": 0.4047, + "step": 540 + }, + { + "epoch": 0.03603485553298827, + "grad_norm": 1.0668187141418457, + "learning_rate": 1.098e-05, + "loss": 0.2567, + "step": 550 + }, + { + "epoch": 0.03669003472449715, + "grad_norm": 1.3863497972488403, + "learning_rate": 1.118e-05, + "loss": 0.3701, + "step": 560 + }, + { + "epoch": 0.03734521391600603, + "grad_norm": 1.1854925155639648, + "learning_rate": 1.1380000000000001e-05, + "loss": 0.3121, + "step": 570 + }, + { + "epoch": 0.03800039310751491, + "grad_norm": 1.6623371839523315, + "learning_rate": 1.1580000000000001e-05, + "loss": 0.4207, + "step": 580 + }, + { + "epoch": 0.038655572299023784, + "grad_norm": 1.3982959985733032, + "learning_rate": 1.178e-05, + "loss": 0.3108, + "step": 590 + }, + { + "epoch": 0.03931075149053266, + "grad_norm": 1.537261724472046, + "learning_rate": 1.198e-05, + "loss": 0.3872, + "step": 600 + }, + { + "epoch": 0.03996593068204154, + "grad_norm": 1.4267938137054443, + "learning_rate": 1.2180000000000002e-05, + "loss": 0.3042, + "step": 610 + }, + { + "epoch": 0.040621109873550415, + "grad_norm": 0.7126345038414001, + "learning_rate": 1.238e-05, + "loss": 0.403, + "step": 620 + }, + { + "epoch": 0.04127628906505929, + "grad_norm": 1.4390127658843994, + "learning_rate": 1.258e-05, + "loss": 0.3264, + "step": 630 + }, + { + "epoch": 0.04193146825656817, + "grad_norm": 1.1396368741989136, + "learning_rate": 1.278e-05, + "loss": 0.3327, + "step": 640 + }, + { + "epoch": 0.042586647448077046, + "grad_norm": 1.009263277053833, + "learning_rate": 1.2980000000000001e-05, + "loss": 0.3559, + "step": 650 + }, + { + "epoch": 0.04324182663958593, + "grad_norm": 1.8853129148483276, + "learning_rate": 1.3180000000000001e-05, + "loss": 0.2751, + "step": 660 + }, + { + "epoch": 0.04389700583109481, + "grad_norm": 1.1308565139770508, + "learning_rate": 1.338e-05, + "loss": 0.2519, + "step": 670 + }, + { + "epoch": 0.044552185022603684, + "grad_norm": 1.1398775577545166, + "learning_rate": 1.358e-05, + "loss": 0.3027, + "step": 680 + }, + { + "epoch": 0.04520736421411256, + "grad_norm": 1.3917056322097778, + "learning_rate": 1.3780000000000002e-05, + "loss": 0.3569, + "step": 690 + }, + { + "epoch": 0.04586254340562144, + "grad_norm": 0.9985350966453552, + "learning_rate": 1.3980000000000002e-05, + "loss": 0.2821, + "step": 700 + }, + { + "epoch": 0.046517722597130315, + "grad_norm": 1.8001519441604614, + "learning_rate": 1.4180000000000001e-05, + "loss": 0.216, + "step": 710 + }, + { + "epoch": 0.04717290178863919, + "grad_norm": 1.7771563529968262, + "learning_rate": 1.4380000000000001e-05, + "loss": 0.3155, + "step": 720 + }, + { + "epoch": 0.04782808098014807, + "grad_norm": 1.3022303581237793, + "learning_rate": 1.4580000000000003e-05, + "loss": 0.3351, + "step": 730 + }, + { + "epoch": 0.048483260171656946, + "grad_norm": 0.976335346698761, + "learning_rate": 1.4779999999999999e-05, + "loss": 0.2456, + "step": 740 + }, + { + "epoch": 0.04913843936316582, + "grad_norm": 0.7914659380912781, + "learning_rate": 1.4979999999999999e-05, + "loss": 0.2806, + "step": 750 + }, + { + "epoch": 0.04979361855467471, + "grad_norm": 1.626084804534912, + "learning_rate": 1.518e-05, + "loss": 0.3928, + "step": 760 + }, + { + "epoch": 0.050448797746183584, + "grad_norm": 4.636074542999268, + "learning_rate": 1.538e-05, + "loss": 0.2797, + "step": 770 + }, + { + "epoch": 0.05110397693769246, + "grad_norm": 2.158555746078491, + "learning_rate": 1.558e-05, + "loss": 0.2975, + "step": 780 + }, + { + "epoch": 0.05175915612920134, + "grad_norm": 1.0077499151229858, + "learning_rate": 1.578e-05, + "loss": 0.3009, + "step": 790 + }, + { + "epoch": 0.052414335320710215, + "grad_norm": 1.563056230545044, + "learning_rate": 1.598e-05, + "loss": 0.2893, + "step": 800 + }, + { + "epoch": 0.05306951451221909, + "grad_norm": 2.527817487716675, + "learning_rate": 1.618e-05, + "loss": 0.3372, + "step": 810 + }, + { + "epoch": 0.05372469370372797, + "grad_norm": 1.221137523651123, + "learning_rate": 1.6380000000000002e-05, + "loss": 0.2917, + "step": 820 + }, + { + "epoch": 0.054379872895236846, + "grad_norm": 1.5230129957199097, + "learning_rate": 1.658e-05, + "loss": 0.3079, + "step": 830 + }, + { + "epoch": 0.05503505208674572, + "grad_norm": 1.3653696775436401, + "learning_rate": 1.6780000000000002e-05, + "loss": 0.262, + "step": 840 + }, + { + "epoch": 0.0556902312782546, + "grad_norm": 2.8329689502716064, + "learning_rate": 1.698e-05, + "loss": 0.3886, + "step": 850 + }, + { + "epoch": 0.056345410469763484, + "grad_norm": 1.8468819856643677, + "learning_rate": 1.718e-05, + "loss": 0.2933, + "step": 860 + }, + { + "epoch": 0.05700058966127236, + "grad_norm": 0.9159661531448364, + "learning_rate": 1.7380000000000003e-05, + "loss": 0.2627, + "step": 870 + }, + { + "epoch": 0.05765576885278124, + "grad_norm": 1.686079740524292, + "learning_rate": 1.758e-05, + "loss": 0.2715, + "step": 880 + }, + { + "epoch": 0.058310948044290115, + "grad_norm": 2.984532117843628, + "learning_rate": 1.7780000000000003e-05, + "loss": 0.2576, + "step": 890 + }, + { + "epoch": 0.05896612723579899, + "grad_norm": 1.0476961135864258, + "learning_rate": 1.798e-05, + "loss": 0.2899, + "step": 900 + }, + { + "epoch": 0.05962130642730787, + "grad_norm": 2.002732276916504, + "learning_rate": 1.818e-05, + "loss": 0.2107, + "step": 910 + }, + { + "epoch": 0.060276485618816746, + "grad_norm": 1.7389798164367676, + "learning_rate": 1.838e-05, + "loss": 0.2545, + "step": 920 + }, + { + "epoch": 0.06093166481032562, + "grad_norm": 1.0608752965927124, + "learning_rate": 1.858e-05, + "loss": 0.2337, + "step": 930 + }, + { + "epoch": 0.0615868440018345, + "grad_norm": 1.663006067276001, + "learning_rate": 1.878e-05, + "loss": 0.2552, + "step": 940 + }, + { + "epoch": 0.062242023193343377, + "grad_norm": 2.031991481781006, + "learning_rate": 1.898e-05, + "loss": 0.3511, + "step": 950 + }, + { + "epoch": 0.06289720238485226, + "grad_norm": 1.4748591184616089, + "learning_rate": 1.918e-05, + "loss": 0.2823, + "step": 960 + }, + { + "epoch": 0.06355238157636113, + "grad_norm": 0.6544787883758545, + "learning_rate": 1.938e-05, + "loss": 0.2402, + "step": 970 + }, + { + "epoch": 0.06420756076787001, + "grad_norm": 1.7777209281921387, + "learning_rate": 1.9580000000000002e-05, + "loss": 0.2372, + "step": 980 + }, + { + "epoch": 0.06486273995937888, + "grad_norm": 1.6940807104110718, + "learning_rate": 1.978e-05, + "loss": 0.2958, + "step": 990 + }, + { + "epoch": 0.06551791915088777, + "grad_norm": 1.009351372718811, + "learning_rate": 1.9980000000000002e-05, + "loss": 0.2554, + "step": 1000 + }, + { + "epoch": 0.06617309834239665, + "grad_norm": 1.37665855884552, + "learning_rate": 2.0180000000000003e-05, + "loss": 0.2251, + "step": 1010 + }, + { + "epoch": 0.06682827753390552, + "grad_norm": 2.3706014156341553, + "learning_rate": 2.038e-05, + "loss": 0.2446, + "step": 1020 + }, + { + "epoch": 0.0674834567254144, + "grad_norm": 1.8638423681259155, + "learning_rate": 2.0580000000000003e-05, + "loss": 0.2373, + "step": 1030 + }, + { + "epoch": 0.06813863591692328, + "grad_norm": 0.8583676218986511, + "learning_rate": 2.078e-05, + "loss": 0.1779, + "step": 1040 + }, + { + "epoch": 0.06879381510843216, + "grad_norm": 2.566725730895996, + "learning_rate": 2.098e-05, + "loss": 0.2578, + "step": 1050 + }, + { + "epoch": 0.06944899429994103, + "grad_norm": 2.476090669631958, + "learning_rate": 2.118e-05, + "loss": 0.2268, + "step": 1060 + }, + { + "epoch": 0.07010417349144991, + "grad_norm": 1.2037192583084106, + "learning_rate": 2.138e-05, + "loss": 0.2153, + "step": 1070 + }, + { + "epoch": 0.07075935268295878, + "grad_norm": 1.9854589700698853, + "learning_rate": 2.158e-05, + "loss": 0.2469, + "step": 1080 + }, + { + "epoch": 0.07141453187446767, + "grad_norm": 2.506619930267334, + "learning_rate": 2.178e-05, + "loss": 0.2057, + "step": 1090 + }, + { + "epoch": 0.07206971106597654, + "grad_norm": 1.5133029222488403, + "learning_rate": 2.198e-05, + "loss": 0.1879, + "step": 1100 + }, + { + "epoch": 0.07272489025748542, + "grad_norm": 4.859869956970215, + "learning_rate": 2.218e-05, + "loss": 0.2832, + "step": 1110 + }, + { + "epoch": 0.0733800694489943, + "grad_norm": 1.6624574661254883, + "learning_rate": 2.2380000000000003e-05, + "loss": 0.2834, + "step": 1120 + }, + { + "epoch": 0.07403524864050318, + "grad_norm": 2.0120415687561035, + "learning_rate": 2.258e-05, + "loss": 0.1805, + "step": 1130 + }, + { + "epoch": 0.07469042783201206, + "grad_norm": 1.2187970876693726, + "learning_rate": 2.2780000000000002e-05, + "loss": 0.2518, + "step": 1140 + }, + { + "epoch": 0.07534560702352093, + "grad_norm": 1.3481252193450928, + "learning_rate": 2.298e-05, + "loss": 0.2326, + "step": 1150 + }, + { + "epoch": 0.07600078621502981, + "grad_norm": 1.5089207887649536, + "learning_rate": 2.318e-05, + "loss": 0.2638, + "step": 1160 + }, + { + "epoch": 0.07665596540653868, + "grad_norm": 1.3499252796173096, + "learning_rate": 2.3380000000000003e-05, + "loss": 0.2168, + "step": 1170 + }, + { + "epoch": 0.07731114459804757, + "grad_norm": 2.576601028442383, + "learning_rate": 2.358e-05, + "loss": 0.3133, + "step": 1180 + }, + { + "epoch": 0.07796632378955644, + "grad_norm": 1.5811117887496948, + "learning_rate": 2.3780000000000003e-05, + "loss": 0.265, + "step": 1190 + }, + { + "epoch": 0.07862150298106532, + "grad_norm": 1.72284734249115, + "learning_rate": 2.398e-05, + "loss": 0.2104, + "step": 1200 + }, + { + "epoch": 0.07927668217257419, + "grad_norm": 1.401344656944275, + "learning_rate": 2.418e-05, + "loss": 0.2353, + "step": 1210 + }, + { + "epoch": 0.07993186136408308, + "grad_norm": 1.6355459690093994, + "learning_rate": 2.438e-05, + "loss": 0.216, + "step": 1220 + }, + { + "epoch": 0.08058704055559196, + "grad_norm": 1.0584959983825684, + "learning_rate": 2.4580000000000002e-05, + "loss": 0.2397, + "step": 1230 + }, + { + "epoch": 0.08124221974710083, + "grad_norm": 2.8211474418640137, + "learning_rate": 2.478e-05, + "loss": 0.2169, + "step": 1240 + }, + { + "epoch": 0.08189739893860971, + "grad_norm": 2.355088949203491, + "learning_rate": 2.498e-05, + "loss": 0.1639, + "step": 1250 + }, + { + "epoch": 0.08255257813011858, + "grad_norm": 2.1800267696380615, + "learning_rate": 2.5180000000000003e-05, + "loss": 0.1678, + "step": 1260 + }, + { + "epoch": 0.08320775732162747, + "grad_norm": 0.7784948348999023, + "learning_rate": 2.5380000000000004e-05, + "loss": 0.2105, + "step": 1270 + }, + { + "epoch": 0.08386293651313634, + "grad_norm": 1.651266098022461, + "learning_rate": 2.5580000000000002e-05, + "loss": 0.1948, + "step": 1280 + }, + { + "epoch": 0.08451811570464522, + "grad_norm": 4.736125469207764, + "learning_rate": 2.5779999999999997e-05, + "loss": 0.2627, + "step": 1290 + }, + { + "epoch": 0.08517329489615409, + "grad_norm": 1.3027892112731934, + "learning_rate": 2.598e-05, + "loss": 0.2227, + "step": 1300 + }, + { + "epoch": 0.08582847408766298, + "grad_norm": 1.1085600852966309, + "learning_rate": 2.618e-05, + "loss": 0.2283, + "step": 1310 + }, + { + "epoch": 0.08648365327917186, + "grad_norm": 1.1086338758468628, + "learning_rate": 2.6379999999999998e-05, + "loss": 0.2101, + "step": 1320 + }, + { + "epoch": 0.08713883247068073, + "grad_norm": 1.4554002285003662, + "learning_rate": 2.658e-05, + "loss": 0.1836, + "step": 1330 + }, + { + "epoch": 0.08779401166218961, + "grad_norm": 3.929887294769287, + "learning_rate": 2.678e-05, + "loss": 0.2027, + "step": 1340 + }, + { + "epoch": 0.08844919085369848, + "grad_norm": 1.6334865093231201, + "learning_rate": 2.698e-05, + "loss": 0.2269, + "step": 1350 + }, + { + "epoch": 0.08910437004520737, + "grad_norm": 1.001843810081482, + "learning_rate": 2.718e-05, + "loss": 0.1753, + "step": 1360 + }, + { + "epoch": 0.08975954923671624, + "grad_norm": 1.4153096675872803, + "learning_rate": 2.738e-05, + "loss": 0.2063, + "step": 1370 + }, + { + "epoch": 0.09041472842822512, + "grad_norm": 2.177656888961792, + "learning_rate": 2.758e-05, + "loss": 0.1928, + "step": 1380 + }, + { + "epoch": 0.09106990761973399, + "grad_norm": 2.2738373279571533, + "learning_rate": 2.778e-05, + "loss": 0.2109, + "step": 1390 + }, + { + "epoch": 0.09172508681124288, + "grad_norm": 1.5940674543380737, + "learning_rate": 2.798e-05, + "loss": 0.1491, + "step": 1400 + }, + { + "epoch": 0.09238026600275175, + "grad_norm": 0.8375024795532227, + "learning_rate": 2.818e-05, + "loss": 0.1912, + "step": 1410 + }, + { + "epoch": 0.09303544519426063, + "grad_norm": 2.575969934463501, + "learning_rate": 2.8380000000000003e-05, + "loss": 0.2077, + "step": 1420 + }, + { + "epoch": 0.09369062438576951, + "grad_norm": 1.515857458114624, + "learning_rate": 2.858e-05, + "loss": 0.215, + "step": 1430 + }, + { + "epoch": 0.09434580357727838, + "grad_norm": 1.4387102127075195, + "learning_rate": 2.8780000000000002e-05, + "loss": 0.2072, + "step": 1440 + }, + { + "epoch": 0.09500098276878727, + "grad_norm": 2.8099513053894043, + "learning_rate": 2.898e-05, + "loss": 0.2155, + "step": 1450 + }, + { + "epoch": 0.09565616196029614, + "grad_norm": 2.0418131351470947, + "learning_rate": 2.9180000000000002e-05, + "loss": 0.2541, + "step": 1460 + }, + { + "epoch": 0.09631134115180502, + "grad_norm": 0.9775047898292542, + "learning_rate": 2.9380000000000003e-05, + "loss": 0.171, + "step": 1470 + }, + { + "epoch": 0.09696652034331389, + "grad_norm": 1.1508287191390991, + "learning_rate": 2.958e-05, + "loss": 0.1939, + "step": 1480 + }, + { + "epoch": 0.09762169953482278, + "grad_norm": 0.8230220079421997, + "learning_rate": 2.9780000000000003e-05, + "loss": 0.1846, + "step": 1490 + }, + { + "epoch": 0.09827687872633165, + "grad_norm": 1.9762450456619263, + "learning_rate": 2.998e-05, + "loss": 0.2545, + "step": 1500 + }, + { + "epoch": 0.09893205791784053, + "grad_norm": 0.7681851983070374, + "learning_rate": 3.0180000000000002e-05, + "loss": 0.1571, + "step": 1510 + }, + { + "epoch": 0.09958723710934941, + "grad_norm": 0.9042073488235474, + "learning_rate": 3.0380000000000004e-05, + "loss": 0.1998, + "step": 1520 + }, + { + "epoch": 0.10024241630085828, + "grad_norm": 1.5681191682815552, + "learning_rate": 3.058e-05, + "loss": 0.2435, + "step": 1530 + }, + { + "epoch": 0.10089759549236717, + "grad_norm": 1.4539704322814941, + "learning_rate": 3.078e-05, + "loss": 0.2173, + "step": 1540 + }, + { + "epoch": 0.10155277468387604, + "grad_norm": 1.5105812549591064, + "learning_rate": 3.0980000000000005e-05, + "loss": 0.2512, + "step": 1550 + }, + { + "epoch": 0.10220795387538492, + "grad_norm": 1.7017266750335693, + "learning_rate": 3.118e-05, + "loss": 0.227, + "step": 1560 + }, + { + "epoch": 0.10286313306689379, + "grad_norm": 1.389731526374817, + "learning_rate": 3.138e-05, + "loss": 0.2124, + "step": 1570 + }, + { + "epoch": 0.10351831225840268, + "grad_norm": 1.3085849285125732, + "learning_rate": 3.1580000000000006e-05, + "loss": 0.1405, + "step": 1580 + }, + { + "epoch": 0.10417349144991155, + "grad_norm": 0.9735742211341858, + "learning_rate": 3.1780000000000004e-05, + "loss": 0.2395, + "step": 1590 + }, + { + "epoch": 0.10482867064142043, + "grad_norm": 1.165263056755066, + "learning_rate": 3.198e-05, + "loss": 0.1959, + "step": 1600 + }, + { + "epoch": 0.1054838498329293, + "grad_norm": 1.8542650938034058, + "learning_rate": 3.218e-05, + "loss": 0.1833, + "step": 1610 + }, + { + "epoch": 0.10613902902443818, + "grad_norm": 1.5433002710342407, + "learning_rate": 3.238e-05, + "loss": 0.2462, + "step": 1620 + }, + { + "epoch": 0.10679420821594707, + "grad_norm": 2.0836150646209717, + "learning_rate": 3.2579999999999996e-05, + "loss": 0.1829, + "step": 1630 + }, + { + "epoch": 0.10744938740745594, + "grad_norm": 1.3217159509658813, + "learning_rate": 3.278e-05, + "loss": 0.1678, + "step": 1640 + }, + { + "epoch": 0.10810456659896482, + "grad_norm": 1.567156195640564, + "learning_rate": 3.298e-05, + "loss": 0.2482, + "step": 1650 + }, + { + "epoch": 0.10875974579047369, + "grad_norm": 1.452773094177246, + "learning_rate": 3.318e-05, + "loss": 0.161, + "step": 1660 + }, + { + "epoch": 0.10941492498198258, + "grad_norm": 1.6922581195831299, + "learning_rate": 3.338e-05, + "loss": 0.2037, + "step": 1670 + }, + { + "epoch": 0.11007010417349145, + "grad_norm": 1.0042247772216797, + "learning_rate": 3.358e-05, + "loss": 0.1963, + "step": 1680 + }, + { + "epoch": 0.11072528336500033, + "grad_norm": 2.8825161457061768, + "learning_rate": 3.378e-05, + "loss": 0.2038, + "step": 1690 + }, + { + "epoch": 0.1113804625565092, + "grad_norm": 1.107530117034912, + "learning_rate": 3.398e-05, + "loss": 0.272, + "step": 1700 + }, + { + "epoch": 0.11203564174801808, + "grad_norm": 2.0809009075164795, + "learning_rate": 3.418e-05, + "loss": 0.2069, + "step": 1710 + }, + { + "epoch": 0.11269082093952697, + "grad_norm": 2.3265416622161865, + "learning_rate": 3.438e-05, + "loss": 0.1907, + "step": 1720 + }, + { + "epoch": 0.11334600013103584, + "grad_norm": 1.609960913658142, + "learning_rate": 3.4580000000000004e-05, + "loss": 0.1775, + "step": 1730 + }, + { + "epoch": 0.11400117932254472, + "grad_norm": 0.6146945357322693, + "learning_rate": 3.478e-05, + "loss": 0.1915, + "step": 1740 + }, + { + "epoch": 0.11465635851405359, + "grad_norm": 2.4051966667175293, + "learning_rate": 3.498e-05, + "loss": 0.203, + "step": 1750 + }, + { + "epoch": 0.11531153770556248, + "grad_norm": 1.3040859699249268, + "learning_rate": 3.518e-05, + "loss": 0.2213, + "step": 1760 + }, + { + "epoch": 0.11596671689707135, + "grad_norm": 1.5166075229644775, + "learning_rate": 3.5380000000000003e-05, + "loss": 0.2246, + "step": 1770 + }, + { + "epoch": 0.11662189608858023, + "grad_norm": 1.8945187330245972, + "learning_rate": 3.558e-05, + "loss": 0.2016, + "step": 1780 + }, + { + "epoch": 0.1172770752800891, + "grad_norm": 1.765882134437561, + "learning_rate": 3.578e-05, + "loss": 0.2135, + "step": 1790 + }, + { + "epoch": 0.11793225447159798, + "grad_norm": 2.236666202545166, + "learning_rate": 3.5980000000000004e-05, + "loss": 0.1654, + "step": 1800 + }, + { + "epoch": 0.11858743366310685, + "grad_norm": 1.9755510091781616, + "learning_rate": 3.618e-05, + "loss": 0.2284, + "step": 1810 + }, + { + "epoch": 0.11924261285461574, + "grad_norm": 1.2482941150665283, + "learning_rate": 3.638e-05, + "loss": 0.1453, + "step": 1820 + }, + { + "epoch": 0.11989779204612462, + "grad_norm": 1.554876446723938, + "learning_rate": 3.6580000000000006e-05, + "loss": 0.2257, + "step": 1830 + }, + { + "epoch": 0.12055297123763349, + "grad_norm": 1.95467209815979, + "learning_rate": 3.6780000000000004e-05, + "loss": 0.1998, + "step": 1840 + }, + { + "epoch": 0.12120815042914238, + "grad_norm": 0.8210131525993347, + "learning_rate": 3.698e-05, + "loss": 0.1823, + "step": 1850 + }, + { + "epoch": 0.12186332962065125, + "grad_norm": 1.8162206411361694, + "learning_rate": 3.7180000000000007e-05, + "loss": 0.14, + "step": 1860 + }, + { + "epoch": 0.12251850881216013, + "grad_norm": 1.3243019580841064, + "learning_rate": 3.7380000000000005e-05, + "loss": 0.1592, + "step": 1870 + }, + { + "epoch": 0.123173688003669, + "grad_norm": 5.006040573120117, + "learning_rate": 3.758e-05, + "loss": 0.2033, + "step": 1880 + }, + { + "epoch": 0.12382886719517788, + "grad_norm": 2.278228998184204, + "learning_rate": 3.778000000000001e-05, + "loss": 0.1961, + "step": 1890 + }, + { + "epoch": 0.12448404638668675, + "grad_norm": 1.329985499382019, + "learning_rate": 3.7980000000000006e-05, + "loss": 0.1832, + "step": 1900 + }, + { + "epoch": 0.12513922557819562, + "grad_norm": 2.41444993019104, + "learning_rate": 3.818e-05, + "loss": 0.1636, + "step": 1910 + }, + { + "epoch": 0.12579440476970452, + "grad_norm": 2.2457187175750732, + "learning_rate": 3.838e-05, + "loss": 0.1758, + "step": 1920 + }, + { + "epoch": 0.1264495839612134, + "grad_norm": 1.4449230432510376, + "learning_rate": 3.858e-05, + "loss": 0.1766, + "step": 1930 + }, + { + "epoch": 0.12710476315272226, + "grad_norm": 1.1978055238723755, + "learning_rate": 3.878e-05, + "loss": 0.2369, + "step": 1940 + }, + { + "epoch": 0.12775994234423116, + "grad_norm": 1.0653358697891235, + "learning_rate": 3.898e-05, + "loss": 0.1727, + "step": 1950 + }, + { + "epoch": 0.12841512153574003, + "grad_norm": 0.6688687801361084, + "learning_rate": 3.918e-05, + "loss": 0.1832, + "step": 1960 + }, + { + "epoch": 0.1290703007272489, + "grad_norm": 3.0809099674224854, + "learning_rate": 3.938e-05, + "loss": 0.2018, + "step": 1970 + }, + { + "epoch": 0.12972547991875777, + "grad_norm": 2.3768062591552734, + "learning_rate": 3.958e-05, + "loss": 0.1823, + "step": 1980 + }, + { + "epoch": 0.13038065911026667, + "grad_norm": 1.9399824142456055, + "learning_rate": 3.978e-05, + "loss": 0.148, + "step": 1990 + }, + { + "epoch": 0.13103583830177554, + "grad_norm": 1.4165918827056885, + "learning_rate": 3.998e-05, + "loss": 0.1555, + "step": 2000 + }, + { + "epoch": 0.1316910174932844, + "grad_norm": 1.9110829830169678, + "learning_rate": 4.018e-05, + "loss": 0.1897, + "step": 2010 + }, + { + "epoch": 0.1323461966847933, + "grad_norm": 2.0675418376922607, + "learning_rate": 4.038e-05, + "loss": 0.1877, + "step": 2020 + }, + { + "epoch": 0.13300137587630217, + "grad_norm": 0.9109375476837158, + "learning_rate": 4.058e-05, + "loss": 0.1639, + "step": 2030 + }, + { + "epoch": 0.13365655506781104, + "grad_norm": 1.2759534120559692, + "learning_rate": 4.078e-05, + "loss": 0.2153, + "step": 2040 + }, + { + "epoch": 0.13431173425931991, + "grad_norm": 1.4455032348632812, + "learning_rate": 4.0980000000000004e-05, + "loss": 0.1914, + "step": 2050 + }, + { + "epoch": 0.1349669134508288, + "grad_norm": 1.8373464345932007, + "learning_rate": 4.118e-05, + "loss": 0.1955, + "step": 2060 + }, + { + "epoch": 0.13562209264233768, + "grad_norm": 2.123349905014038, + "learning_rate": 4.138e-05, + "loss": 0.1613, + "step": 2070 + }, + { + "epoch": 0.13627727183384655, + "grad_norm": 1.5149427652359009, + "learning_rate": 4.1580000000000005e-05, + "loss": 0.2243, + "step": 2080 + }, + { + "epoch": 0.13693245102535542, + "grad_norm": 2.430596113204956, + "learning_rate": 4.178e-05, + "loss": 0.1671, + "step": 2090 + }, + { + "epoch": 0.13758763021686432, + "grad_norm": 1.2629282474517822, + "learning_rate": 4.198e-05, + "loss": 0.1944, + "step": 2100 + }, + { + "epoch": 0.1382428094083732, + "grad_norm": 1.129300832748413, + "learning_rate": 4.2180000000000006e-05, + "loss": 0.2055, + "step": 2110 + }, + { + "epoch": 0.13889798859988206, + "grad_norm": 1.1666574478149414, + "learning_rate": 4.2380000000000004e-05, + "loss": 0.1937, + "step": 2120 + }, + { + "epoch": 0.13955316779139096, + "grad_norm": 1.5008031129837036, + "learning_rate": 4.258e-05, + "loss": 0.194, + "step": 2130 + }, + { + "epoch": 0.14020834698289983, + "grad_norm": 3.863701343536377, + "learning_rate": 4.278e-05, + "loss": 0.1842, + "step": 2140 + }, + { + "epoch": 0.1408635261744087, + "grad_norm": 4.390907287597656, + "learning_rate": 4.2980000000000005e-05, + "loss": 0.1781, + "step": 2150 + }, + { + "epoch": 0.14151870536591757, + "grad_norm": 2.0250678062438965, + "learning_rate": 4.318e-05, + "loss": 0.2231, + "step": 2160 + }, + { + "epoch": 0.14217388455742647, + "grad_norm": 1.7647753953933716, + "learning_rate": 4.338e-05, + "loss": 0.2294, + "step": 2170 + }, + { + "epoch": 0.14282906374893534, + "grad_norm": 1.5749119520187378, + "learning_rate": 4.3580000000000006e-05, + "loss": 0.1548, + "step": 2180 + }, + { + "epoch": 0.1434842429404442, + "grad_norm": 2.385044813156128, + "learning_rate": 4.3780000000000004e-05, + "loss": 0.1436, + "step": 2190 + }, + { + "epoch": 0.14413942213195308, + "grad_norm": 2.1540536880493164, + "learning_rate": 4.398e-05, + "loss": 0.1767, + "step": 2200 + }, + { + "epoch": 0.14479460132346197, + "grad_norm": 1.863830804824829, + "learning_rate": 4.418000000000001e-05, + "loss": 0.1889, + "step": 2210 + }, + { + "epoch": 0.14544978051497084, + "grad_norm": 1.8047069311141968, + "learning_rate": 4.438e-05, + "loss": 0.1878, + "step": 2220 + }, + { + "epoch": 0.14610495970647971, + "grad_norm": 2.1893537044525146, + "learning_rate": 4.458e-05, + "loss": 0.1794, + "step": 2230 + }, + { + "epoch": 0.1467601388979886, + "grad_norm": 1.3070318698883057, + "learning_rate": 4.478e-05, + "loss": 0.172, + "step": 2240 + }, + { + "epoch": 0.14741531808949748, + "grad_norm": 3.534369468688965, + "learning_rate": 4.498e-05, + "loss": 0.2162, + "step": 2250 + }, + { + "epoch": 0.14807049728100635, + "grad_norm": 1.172663688659668, + "learning_rate": 4.518e-05, + "loss": 0.2135, + "step": 2260 + }, + { + "epoch": 0.14872567647251522, + "grad_norm": 3.160845994949341, + "learning_rate": 4.538e-05, + "loss": 0.247, + "step": 2270 + }, + { + "epoch": 0.14938085566402412, + "grad_norm": 2.766282796859741, + "learning_rate": 4.558e-05, + "loss": 0.24, + "step": 2280 + }, + { + "epoch": 0.150036034855533, + "grad_norm": 1.431797981262207, + "learning_rate": 4.578e-05, + "loss": 0.1348, + "step": 2290 + }, + { + "epoch": 0.15069121404704186, + "grad_norm": 1.8124114274978638, + "learning_rate": 4.5980000000000004e-05, + "loss": 0.1781, + "step": 2300 + }, + { + "epoch": 0.15134639323855073, + "grad_norm": 1.4705984592437744, + "learning_rate": 4.618e-05, + "loss": 0.2649, + "step": 2310 + }, + { + "epoch": 0.15200157243005963, + "grad_norm": 1.8528354167938232, + "learning_rate": 4.638e-05, + "loss": 0.2243, + "step": 2320 + }, + { + "epoch": 0.1526567516215685, + "grad_norm": 2.0011472702026367, + "learning_rate": 4.6580000000000005e-05, + "loss": 0.2113, + "step": 2330 + }, + { + "epoch": 0.15331193081307737, + "grad_norm": 2.2429723739624023, + "learning_rate": 4.678e-05, + "loss": 0.1454, + "step": 2340 + }, + { + "epoch": 0.15396711000458627, + "grad_norm": 0.5895823240280151, + "learning_rate": 4.698e-05, + "loss": 0.1775, + "step": 2350 + }, + { + "epoch": 0.15462228919609514, + "grad_norm": 2.3359038829803467, + "learning_rate": 4.718e-05, + "loss": 0.1902, + "step": 2360 + }, + { + "epoch": 0.155277468387604, + "grad_norm": 1.204166054725647, + "learning_rate": 4.7380000000000004e-05, + "loss": 0.1628, + "step": 2370 + }, + { + "epoch": 0.15593264757911288, + "grad_norm": 1.934403896331787, + "learning_rate": 4.758e-05, + "loss": 0.158, + "step": 2380 + }, + { + "epoch": 0.15658782677062177, + "grad_norm": 1.441035509109497, + "learning_rate": 4.778e-05, + "loss": 0.1893, + "step": 2390 + }, + { + "epoch": 0.15724300596213064, + "grad_norm": 1.2663317918777466, + "learning_rate": 4.7980000000000005e-05, + "loss": 0.1996, + "step": 2400 + }, + { + "epoch": 0.15789818515363951, + "grad_norm": 2.124370813369751, + "learning_rate": 4.818e-05, + "loss": 0.1628, + "step": 2410 + }, + { + "epoch": 0.15855336434514838, + "grad_norm": 1.6121026277542114, + "learning_rate": 4.838e-05, + "loss": 0.1777, + "step": 2420 + }, + { + "epoch": 0.15920854353665728, + "grad_norm": 1.350885272026062, + "learning_rate": 4.8580000000000006e-05, + "loss": 0.1835, + "step": 2430 + }, + { + "epoch": 0.15986372272816615, + "grad_norm": 1.15628981590271, + "learning_rate": 4.8780000000000004e-05, + "loss": 0.1169, + "step": 2440 + }, + { + "epoch": 0.16051890191967502, + "grad_norm": 1.471205234527588, + "learning_rate": 4.898e-05, + "loss": 0.1465, + "step": 2450 + }, + { + "epoch": 0.16117408111118392, + "grad_norm": 1.600447416305542, + "learning_rate": 4.918000000000001e-05, + "loss": 0.1472, + "step": 2460 + }, + { + "epoch": 0.1618292603026928, + "grad_norm": 1.0723706483840942, + "learning_rate": 4.9380000000000005e-05, + "loss": 0.1809, + "step": 2470 + }, + { + "epoch": 0.16248443949420166, + "grad_norm": 0.8974424600601196, + "learning_rate": 4.958e-05, + "loss": 0.2053, + "step": 2480 + }, + { + "epoch": 0.16313961868571053, + "grad_norm": 1.2110673189163208, + "learning_rate": 4.978e-05, + "loss": 0.1412, + "step": 2490 + }, + { + "epoch": 0.16379479787721943, + "grad_norm": 2.544933795928955, + "learning_rate": 4.9980000000000006e-05, + "loss": 0.163, + "step": 2500 + }, + { + "epoch": 0.1644499770687283, + "grad_norm": 1.0978542566299438, + "learning_rate": 5.0180000000000004e-05, + "loss": 0.1803, + "step": 2510 + }, + { + "epoch": 0.16510515626023717, + "grad_norm": 0.9357240200042725, + "learning_rate": 5.038e-05, + "loss": 0.1884, + "step": 2520 + }, + { + "epoch": 0.16576033545174607, + "grad_norm": 1.8413302898406982, + "learning_rate": 5.058000000000001e-05, + "loss": 0.2045, + "step": 2530 + }, + { + "epoch": 0.16641551464325494, + "grad_norm": 2.4470601081848145, + "learning_rate": 5.0780000000000005e-05, + "loss": 0.1684, + "step": 2540 + }, + { + "epoch": 0.1670706938347638, + "grad_norm": 1.479347586631775, + "learning_rate": 5.098e-05, + "loss": 0.1394, + "step": 2550 + }, + { + "epoch": 0.16772587302627268, + "grad_norm": 1.2006157636642456, + "learning_rate": 5.118000000000001e-05, + "loss": 0.1563, + "step": 2560 + }, + { + "epoch": 0.16838105221778157, + "grad_norm": 1.9761444330215454, + "learning_rate": 5.1380000000000006e-05, + "loss": 0.1521, + "step": 2570 + }, + { + "epoch": 0.16903623140929044, + "grad_norm": 2.1197240352630615, + "learning_rate": 5.1580000000000004e-05, + "loss": 0.1498, + "step": 2580 + }, + { + "epoch": 0.16969141060079931, + "grad_norm": 1.0805834531784058, + "learning_rate": 5.178000000000001e-05, + "loss": 0.1938, + "step": 2590 + }, + { + "epoch": 0.17034658979230818, + "grad_norm": 1.875270962715149, + "learning_rate": 5.198000000000001e-05, + "loss": 0.178, + "step": 2600 + }, + { + "epoch": 0.17100176898381708, + "grad_norm": 1.588253140449524, + "learning_rate": 5.2180000000000005e-05, + "loss": 0.1796, + "step": 2610 + }, + { + "epoch": 0.17165694817532595, + "grad_norm": 2.4022834300994873, + "learning_rate": 5.238000000000001e-05, + "loss": 0.184, + "step": 2620 + }, + { + "epoch": 0.17231212736683482, + "grad_norm": 0.9719442129135132, + "learning_rate": 5.258000000000001e-05, + "loss": 0.179, + "step": 2630 + }, + { + "epoch": 0.17296730655834372, + "grad_norm": 1.2704960107803345, + "learning_rate": 5.2780000000000006e-05, + "loss": 0.2571, + "step": 2640 + }, + { + "epoch": 0.1736224857498526, + "grad_norm": 1.2971549034118652, + "learning_rate": 5.2980000000000004e-05, + "loss": 0.1388, + "step": 2650 + }, + { + "epoch": 0.17427766494136146, + "grad_norm": 1.4920430183410645, + "learning_rate": 5.318000000000001e-05, + "loss": 0.1647, + "step": 2660 + }, + { + "epoch": 0.17493284413287033, + "grad_norm": 0.8554392457008362, + "learning_rate": 5.338000000000001e-05, + "loss": 0.1666, + "step": 2670 + }, + { + "epoch": 0.17558802332437923, + "grad_norm": 1.4415838718414307, + "learning_rate": 5.3580000000000005e-05, + "loss": 0.1368, + "step": 2680 + }, + { + "epoch": 0.1762432025158881, + "grad_norm": 1.7498157024383545, + "learning_rate": 5.378e-05, + "loss": 0.1832, + "step": 2690 + }, + { + "epoch": 0.17689838170739697, + "grad_norm": 1.887214183807373, + "learning_rate": 5.3979999999999995e-05, + "loss": 0.1508, + "step": 2700 + }, + { + "epoch": 0.17755356089890584, + "grad_norm": 1.5778199434280396, + "learning_rate": 5.418e-05, + "loss": 0.1644, + "step": 2710 + }, + { + "epoch": 0.17820874009041474, + "grad_norm": 1.8328790664672852, + "learning_rate": 5.438e-05, + "loss": 0.2063, + "step": 2720 + }, + { + "epoch": 0.1788639192819236, + "grad_norm": 1.4377741813659668, + "learning_rate": 5.4579999999999996e-05, + "loss": 0.1691, + "step": 2730 + }, + { + "epoch": 0.17951909847343248, + "grad_norm": 2.2693772315979004, + "learning_rate": 5.478e-05, + "loss": 0.1791, + "step": 2740 + }, + { + "epoch": 0.18017427766494137, + "grad_norm": 2.378222942352295, + "learning_rate": 5.498e-05, + "loss": 0.1856, + "step": 2750 + }, + { + "epoch": 0.18082945685645024, + "grad_norm": 0.8347851634025574, + "learning_rate": 5.518e-05, + "loss": 0.1535, + "step": 2760 + }, + { + "epoch": 0.1814846360479591, + "grad_norm": 3.5646393299102783, + "learning_rate": 5.538e-05, + "loss": 0.196, + "step": 2770 + }, + { + "epoch": 0.18213981523946798, + "grad_norm": 3.3354785442352295, + "learning_rate": 5.558e-05, + "loss": 0.2336, + "step": 2780 + }, + { + "epoch": 0.18279499443097688, + "grad_norm": 2.466275215148926, + "learning_rate": 5.578e-05, + "loss": 0.1813, + "step": 2790 + }, + { + "epoch": 0.18345017362248575, + "grad_norm": 2.3226630687713623, + "learning_rate": 5.5979999999999996e-05, + "loss": 0.1622, + "step": 2800 + }, + { + "epoch": 0.18410535281399462, + "grad_norm": 2.02966570854187, + "learning_rate": 5.618e-05, + "loss": 0.1208, + "step": 2810 + }, + { + "epoch": 0.1847605320055035, + "grad_norm": 2.414386510848999, + "learning_rate": 5.638e-05, + "loss": 0.144, + "step": 2820 + }, + { + "epoch": 0.1854157111970124, + "grad_norm": 2.8894433975219727, + "learning_rate": 5.658e-05, + "loss": 0.1899, + "step": 2830 + }, + { + "epoch": 0.18607089038852126, + "grad_norm": 2.5677649974823, + "learning_rate": 5.678e-05, + "loss": 0.2055, + "step": 2840 + }, + { + "epoch": 0.18672606958003013, + "grad_norm": 1.2904149293899536, + "learning_rate": 5.698e-05, + "loss": 0.1729, + "step": 2850 + }, + { + "epoch": 0.18738124877153903, + "grad_norm": 1.5107752084732056, + "learning_rate": 5.718e-05, + "loss": 0.1365, + "step": 2860 + }, + { + "epoch": 0.1880364279630479, + "grad_norm": 2.9473066329956055, + "learning_rate": 5.738e-05, + "loss": 0.1387, + "step": 2870 + }, + { + "epoch": 0.18869160715455677, + "grad_norm": 1.3935139179229736, + "learning_rate": 5.758e-05, + "loss": 0.1401, + "step": 2880 + }, + { + "epoch": 0.18934678634606564, + "grad_norm": 1.9837114810943604, + "learning_rate": 5.778e-05, + "loss": 0.2122, + "step": 2890 + }, + { + "epoch": 0.19000196553757454, + "grad_norm": 2.7451510429382324, + "learning_rate": 5.7980000000000004e-05, + "loss": 0.1968, + "step": 2900 + }, + { + "epoch": 0.1906571447290834, + "grad_norm": 0.8039019703865051, + "learning_rate": 5.818e-05, + "loss": 0.1877, + "step": 2910 + }, + { + "epoch": 0.19131232392059228, + "grad_norm": 3.8243813514709473, + "learning_rate": 5.838e-05, + "loss": 0.191, + "step": 2920 + }, + { + "epoch": 0.19196750311210117, + "grad_norm": 1.9841928482055664, + "learning_rate": 5.858e-05, + "loss": 0.1536, + "step": 2930 + }, + { + "epoch": 0.19262268230361004, + "grad_norm": 1.3431735038757324, + "learning_rate": 5.878e-05, + "loss": 0.1958, + "step": 2940 + }, + { + "epoch": 0.1932778614951189, + "grad_norm": 3.577378511428833, + "learning_rate": 5.898e-05, + "loss": 0.2004, + "step": 2950 + }, + { + "epoch": 0.19393304068662778, + "grad_norm": 1.2788830995559692, + "learning_rate": 5.918e-05, + "loss": 0.1611, + "step": 2960 + }, + { + "epoch": 0.19458821987813668, + "grad_norm": 1.449887752532959, + "learning_rate": 5.9380000000000004e-05, + "loss": 0.1835, + "step": 2970 + }, + { + "epoch": 0.19524339906964555, + "grad_norm": 0.5969293117523193, + "learning_rate": 5.958e-05, + "loss": 0.1828, + "step": 2980 + }, + { + "epoch": 0.19589857826115442, + "grad_norm": 1.7055848836898804, + "learning_rate": 5.978e-05, + "loss": 0.271, + "step": 2990 + }, + { + "epoch": 0.1965537574526633, + "grad_norm": 3.0351054668426514, + "learning_rate": 5.9980000000000005e-05, + "loss": 0.1445, + "step": 3000 + }, + { + "epoch": 0.1972089366441722, + "grad_norm": 1.1497769355773926, + "learning_rate": 6.018e-05, + "loss": 0.126, + "step": 3010 + }, + { + "epoch": 0.19786411583568106, + "grad_norm": 2.901716470718384, + "learning_rate": 6.038e-05, + "loss": 0.1459, + "step": 3020 + }, + { + "epoch": 0.19851929502718993, + "grad_norm": 1.7987468242645264, + "learning_rate": 6.0580000000000006e-05, + "loss": 0.2252, + "step": 3030 + }, + { + "epoch": 0.19917447421869883, + "grad_norm": 1.559998869895935, + "learning_rate": 6.0780000000000004e-05, + "loss": 0.1436, + "step": 3040 + }, + { + "epoch": 0.1998296534102077, + "grad_norm": 1.8523563146591187, + "learning_rate": 6.098e-05, + "loss": 0.1356, + "step": 3050 + }, + { + "epoch": 0.20048483260171657, + "grad_norm": 1.9314311742782593, + "learning_rate": 6.118000000000001e-05, + "loss": 0.1397, + "step": 3060 + }, + { + "epoch": 0.20114001179322544, + "grad_norm": 4.659631252288818, + "learning_rate": 6.138e-05, + "loss": 0.1872, + "step": 3070 + }, + { + "epoch": 0.20179519098473434, + "grad_norm": 1.751927375793457, + "learning_rate": 6.158e-05, + "loss": 0.1856, + "step": 3080 + }, + { + "epoch": 0.2024503701762432, + "grad_norm": 0.6744887828826904, + "learning_rate": 6.178000000000001e-05, + "loss": 0.1834, + "step": 3090 + }, + { + "epoch": 0.20310554936775208, + "grad_norm": 2.5551395416259766, + "learning_rate": 6.198e-05, + "loss": 0.2028, + "step": 3100 + }, + { + "epoch": 0.20376072855926095, + "grad_norm": 1.248826265335083, + "learning_rate": 6.218e-05, + "loss": 0.1483, + "step": 3110 + }, + { + "epoch": 0.20441590775076984, + "grad_norm": 1.9274080991744995, + "learning_rate": 6.238000000000001e-05, + "loss": 0.2219, + "step": 3120 + }, + { + "epoch": 0.2050710869422787, + "grad_norm": 0.8917430639266968, + "learning_rate": 6.258e-05, + "loss": 0.1457, + "step": 3130 + }, + { + "epoch": 0.20572626613378758, + "grad_norm": 2.7115836143493652, + "learning_rate": 6.278e-05, + "loss": 0.1588, + "step": 3140 + }, + { + "epoch": 0.20638144532529648, + "grad_norm": 3.8847908973693848, + "learning_rate": 6.298000000000001e-05, + "loss": 0.216, + "step": 3150 + }, + { + "epoch": 0.20703662451680535, + "grad_norm": 1.5824801921844482, + "learning_rate": 6.318e-05, + "loss": 0.1616, + "step": 3160 + }, + { + "epoch": 0.20769180370831422, + "grad_norm": 3.924818992614746, + "learning_rate": 6.338e-05, + "loss": 0.192, + "step": 3170 + }, + { + "epoch": 0.2083469828998231, + "grad_norm": 1.8622838258743286, + "learning_rate": 6.358000000000001e-05, + "loss": 0.1727, + "step": 3180 + }, + { + "epoch": 0.209002162091332, + "grad_norm": 2.8894197940826416, + "learning_rate": 6.378e-05, + "loss": 0.1362, + "step": 3190 + }, + { + "epoch": 0.20965734128284086, + "grad_norm": 1.380409836769104, + "learning_rate": 6.398000000000001e-05, + "loss": 0.1247, + "step": 3200 + }, + { + "epoch": 0.21031252047434973, + "grad_norm": 2.0957531929016113, + "learning_rate": 6.418000000000001e-05, + "loss": 0.1307, + "step": 3210 + }, + { + "epoch": 0.2109676996658586, + "grad_norm": 1.537599802017212, + "learning_rate": 6.438e-05, + "loss": 0.1552, + "step": 3220 + }, + { + "epoch": 0.2116228788573675, + "grad_norm": 2.7240681648254395, + "learning_rate": 6.458000000000001e-05, + "loss": 0.1419, + "step": 3230 + }, + { + "epoch": 0.21227805804887637, + "grad_norm": 1.4931658506393433, + "learning_rate": 6.478000000000001e-05, + "loss": 0.1618, + "step": 3240 + }, + { + "epoch": 0.21293323724038524, + "grad_norm": 2.8971524238586426, + "learning_rate": 6.498e-05, + "loss": 0.1681, + "step": 3250 + }, + { + "epoch": 0.21358841643189413, + "grad_norm": 2.02109956741333, + "learning_rate": 6.518000000000001e-05, + "loss": 0.1508, + "step": 3260 + }, + { + "epoch": 0.214243595623403, + "grad_norm": 2.05175518989563, + "learning_rate": 6.538000000000001e-05, + "loss": 0.1695, + "step": 3270 + }, + { + "epoch": 0.21489877481491187, + "grad_norm": 1.384790062904358, + "learning_rate": 6.558e-05, + "loss": 0.1305, + "step": 3280 + }, + { + "epoch": 0.21555395400642074, + "grad_norm": 2.1418557167053223, + "learning_rate": 6.578000000000001e-05, + "loss": 0.1832, + "step": 3290 + }, + { + "epoch": 0.21620913319792964, + "grad_norm": 2.769618511199951, + "learning_rate": 6.598e-05, + "loss": 0.187, + "step": 3300 + }, + { + "epoch": 0.2168643123894385, + "grad_norm": 1.2914879322052002, + "learning_rate": 6.618e-05, + "loss": 0.1134, + "step": 3310 + }, + { + "epoch": 0.21751949158094738, + "grad_norm": 1.489392876625061, + "learning_rate": 6.638e-05, + "loss": 0.1432, + "step": 3320 + }, + { + "epoch": 0.21817467077245628, + "grad_norm": 1.4254696369171143, + "learning_rate": 6.658e-05, + "loss": 0.1338, + "step": 3330 + }, + { + "epoch": 0.21882984996396515, + "grad_norm": 4.549455165863037, + "learning_rate": 6.678e-05, + "loss": 0.1799, + "step": 3340 + }, + { + "epoch": 0.21948502915547402, + "grad_norm": 1.5503714084625244, + "learning_rate": 6.698e-05, + "loss": 0.136, + "step": 3350 + }, + { + "epoch": 0.2201402083469829, + "grad_norm": 1.5452555418014526, + "learning_rate": 6.718e-05, + "loss": 0.1466, + "step": 3360 + }, + { + "epoch": 0.2207953875384918, + "grad_norm": 1.0497398376464844, + "learning_rate": 6.738e-05, + "loss": 0.1791, + "step": 3370 + }, + { + "epoch": 0.22145056673000066, + "grad_norm": 2.9195966720581055, + "learning_rate": 6.758e-05, + "loss": 0.1803, + "step": 3380 + }, + { + "epoch": 0.22210574592150953, + "grad_norm": 1.0214263200759888, + "learning_rate": 6.778e-05, + "loss": 0.1494, + "step": 3390 + }, + { + "epoch": 0.2227609251130184, + "grad_norm": 2.4124338626861572, + "learning_rate": 6.798e-05, + "loss": 0.1884, + "step": 3400 + }, + { + "epoch": 0.2234161043045273, + "grad_norm": 1.176680564880371, + "learning_rate": 6.818e-05, + "loss": 0.1727, + "step": 3410 + }, + { + "epoch": 0.22407128349603617, + "grad_norm": 2.070592164993286, + "learning_rate": 6.838e-05, + "loss": 0.1721, + "step": 3420 + }, + { + "epoch": 0.22472646268754504, + "grad_norm": 1.052329421043396, + "learning_rate": 6.858e-05, + "loss": 0.1722, + "step": 3430 + }, + { + "epoch": 0.22538164187905393, + "grad_norm": 3.458174467086792, + "learning_rate": 6.878e-05, + "loss": 0.1854, + "step": 3440 + }, + { + "epoch": 0.2260368210705628, + "grad_norm": 1.214791178703308, + "learning_rate": 6.898e-05, + "loss": 0.1811, + "step": 3450 + }, + { + "epoch": 0.22669200026207167, + "grad_norm": 2.053131341934204, + "learning_rate": 6.918e-05, + "loss": 0.1633, + "step": 3460 + }, + { + "epoch": 0.22734717945358054, + "grad_norm": 1.5885206460952759, + "learning_rate": 6.938e-05, + "loss": 0.1588, + "step": 3470 + }, + { + "epoch": 0.22800235864508944, + "grad_norm": 1.7747735977172852, + "learning_rate": 6.958e-05, + "loss": 0.1158, + "step": 3480 + }, + { + "epoch": 0.2286575378365983, + "grad_norm": 2.1645543575286865, + "learning_rate": 6.978e-05, + "loss": 0.1513, + "step": 3490 + }, + { + "epoch": 0.22931271702810718, + "grad_norm": 3.0476267337799072, + "learning_rate": 6.998e-05, + "loss": 0.1382, + "step": 3500 + }, + { + "epoch": 0.22996789621961605, + "grad_norm": 2.7087514400482178, + "learning_rate": 7.018e-05, + "loss": 0.1663, + "step": 3510 + }, + { + "epoch": 0.23062307541112495, + "grad_norm": 1.6026946306228638, + "learning_rate": 7.038e-05, + "loss": 0.1841, + "step": 3520 + }, + { + "epoch": 0.23127825460263382, + "grad_norm": 1.3814374208450317, + "learning_rate": 7.058e-05, + "loss": 0.231, + "step": 3530 + }, + { + "epoch": 0.2319334337941427, + "grad_norm": 1.4405162334442139, + "learning_rate": 7.078e-05, + "loss": 0.1451, + "step": 3540 + }, + { + "epoch": 0.2325886129856516, + "grad_norm": 1.0640065670013428, + "learning_rate": 7.098e-05, + "loss": 0.1353, + "step": 3550 + }, + { + "epoch": 0.23324379217716046, + "grad_norm": 2.812781810760498, + "learning_rate": 7.118e-05, + "loss": 0.1584, + "step": 3560 + }, + { + "epoch": 0.23389897136866933, + "grad_norm": 0.709271252155304, + "learning_rate": 7.138e-05, + "loss": 0.1837, + "step": 3570 + }, + { + "epoch": 0.2345541505601782, + "grad_norm": 2.481513023376465, + "learning_rate": 7.158e-05, + "loss": 0.1585, + "step": 3580 + }, + { + "epoch": 0.2352093297516871, + "grad_norm": 0.8657827377319336, + "learning_rate": 7.178000000000001e-05, + "loss": 0.1238, + "step": 3590 + }, + { + "epoch": 0.23586450894319597, + "grad_norm": 1.6376718282699585, + "learning_rate": 7.198e-05, + "loss": 0.1252, + "step": 3600 + }, + { + "epoch": 0.23651968813470484, + "grad_norm": 1.3994781970977783, + "learning_rate": 7.218e-05, + "loss": 0.1559, + "step": 3610 + }, + { + "epoch": 0.2371748673262137, + "grad_norm": 4.2085981369018555, + "learning_rate": 7.238000000000001e-05, + "loss": 0.1666, + "step": 3620 + }, + { + "epoch": 0.2378300465177226, + "grad_norm": 1.7117937803268433, + "learning_rate": 7.258e-05, + "loss": 0.1453, + "step": 3630 + }, + { + "epoch": 0.23848522570923147, + "grad_norm": 2.963873863220215, + "learning_rate": 7.278e-05, + "loss": 0.179, + "step": 3640 + }, + { + "epoch": 0.23914040490074034, + "grad_norm": 1.1248881816864014, + "learning_rate": 7.298000000000001e-05, + "loss": 0.1163, + "step": 3650 + }, + { + "epoch": 0.23979558409224924, + "grad_norm": 1.6471372842788696, + "learning_rate": 7.318e-05, + "loss": 0.1664, + "step": 3660 + }, + { + "epoch": 0.2404507632837581, + "grad_norm": 2.8147122859954834, + "learning_rate": 7.338e-05, + "loss": 0.1923, + "step": 3670 + }, + { + "epoch": 0.24110594247526698, + "grad_norm": 1.658155918121338, + "learning_rate": 7.358000000000001e-05, + "loss": 0.1306, + "step": 3680 + }, + { + "epoch": 0.24176112166677585, + "grad_norm": 1.5100269317626953, + "learning_rate": 7.378e-05, + "loss": 0.1612, + "step": 3690 + }, + { + "epoch": 0.24241630085828475, + "grad_norm": 1.3265793323516846, + "learning_rate": 7.398e-05, + "loss": 0.1408, + "step": 3700 + }, + { + "epoch": 0.24307148004979362, + "grad_norm": 2.2092418670654297, + "learning_rate": 7.418000000000001e-05, + "loss": 0.1494, + "step": 3710 + }, + { + "epoch": 0.2437266592413025, + "grad_norm": 1.419681191444397, + "learning_rate": 7.438e-05, + "loss": 0.2227, + "step": 3720 + }, + { + "epoch": 0.24438183843281136, + "grad_norm": 1.8030469417572021, + "learning_rate": 7.458000000000001e-05, + "loss": 0.1993, + "step": 3730 + }, + { + "epoch": 0.24503701762432026, + "grad_norm": 1.7267769575119019, + "learning_rate": 7.478e-05, + "loss": 0.1559, + "step": 3740 + }, + { + "epoch": 0.24569219681582913, + "grad_norm": 1.0911484956741333, + "learning_rate": 7.498e-05, + "loss": 0.1904, + "step": 3750 + }, + { + "epoch": 0.246347376007338, + "grad_norm": 4.056103706359863, + "learning_rate": 7.518000000000001e-05, + "loss": 0.1595, + "step": 3760 + }, + { + "epoch": 0.2470025551988469, + "grad_norm": 0.7594374418258667, + "learning_rate": 7.538e-05, + "loss": 0.1276, + "step": 3770 + }, + { + "epoch": 0.24765773439035577, + "grad_norm": 1.6607263088226318, + "learning_rate": 7.558e-05, + "loss": 0.1395, + "step": 3780 + }, + { + "epoch": 0.24831291358186464, + "grad_norm": 1.2656879425048828, + "learning_rate": 7.578000000000001e-05, + "loss": 0.1344, + "step": 3790 + }, + { + "epoch": 0.2489680927733735, + "grad_norm": 1.1405283212661743, + "learning_rate": 7.598e-05, + "loss": 0.1828, + "step": 3800 + }, + { + "epoch": 0.2496232719648824, + "grad_norm": 2.4148523807525635, + "learning_rate": 7.618e-05, + "loss": 0.1891, + "step": 3810 + }, + { + "epoch": 0.25027845115639125, + "grad_norm": 2.7064971923828125, + "learning_rate": 7.638000000000001e-05, + "loss": 0.1674, + "step": 3820 + }, + { + "epoch": 0.25093363034790017, + "grad_norm": 1.2596509456634521, + "learning_rate": 7.658e-05, + "loss": 0.1719, + "step": 3830 + }, + { + "epoch": 0.25158880953940904, + "grad_norm": 1.7497053146362305, + "learning_rate": 7.678000000000001e-05, + "loss": 0.19, + "step": 3840 + }, + { + "epoch": 0.2522439887309179, + "grad_norm": 1.5782902240753174, + "learning_rate": 7.698000000000001e-05, + "loss": 0.1752, + "step": 3850 + }, + { + "epoch": 0.2528991679224268, + "grad_norm": 1.1684191226959229, + "learning_rate": 7.718e-05, + "loss": 0.1463, + "step": 3860 + }, + { + "epoch": 0.25355434711393565, + "grad_norm": 1.8197124004364014, + "learning_rate": 7.738000000000001e-05, + "loss": 0.1781, + "step": 3870 + }, + { + "epoch": 0.2542095263054445, + "grad_norm": 1.801611065864563, + "learning_rate": 7.758000000000001e-05, + "loss": 0.1175, + "step": 3880 + }, + { + "epoch": 0.2548647054969534, + "grad_norm": 1.240411639213562, + "learning_rate": 7.778e-05, + "loss": 0.1441, + "step": 3890 + }, + { + "epoch": 0.2555198846884623, + "grad_norm": 3.4137840270996094, + "learning_rate": 7.798000000000001e-05, + "loss": 0.1707, + "step": 3900 + }, + { + "epoch": 0.2561750638799712, + "grad_norm": 1.1545884609222412, + "learning_rate": 7.818000000000001e-05, + "loss": 0.1179, + "step": 3910 + }, + { + "epoch": 0.25683024307148006, + "grad_norm": 4.102241516113281, + "learning_rate": 7.838e-05, + "loss": 0.109, + "step": 3920 + }, + { + "epoch": 0.25748542226298893, + "grad_norm": 0.8324083089828491, + "learning_rate": 7.858000000000001e-05, + "loss": 0.1726, + "step": 3930 + }, + { + "epoch": 0.2581406014544978, + "grad_norm": 1.2043119668960571, + "learning_rate": 7.878e-05, + "loss": 0.1584, + "step": 3940 + }, + { + "epoch": 0.25879578064600667, + "grad_norm": 3.7131147384643555, + "learning_rate": 7.897999999999999e-05, + "loss": 0.1642, + "step": 3950 + }, + { + "epoch": 0.25945095983751554, + "grad_norm": 1.0852019786834717, + "learning_rate": 7.918e-05, + "loss": 0.103, + "step": 3960 + }, + { + "epoch": 0.26010613902902446, + "grad_norm": 2.3075366020202637, + "learning_rate": 7.938e-05, + "loss": 0.1245, + "step": 3970 + }, + { + "epoch": 0.26076131822053333, + "grad_norm": 1.1030235290527344, + "learning_rate": 7.958e-05, + "loss": 0.1327, + "step": 3980 + }, + { + "epoch": 0.2614164974120422, + "grad_norm": 1.9360440969467163, + "learning_rate": 7.978e-05, + "loss": 0.1585, + "step": 3990 + }, + { + "epoch": 0.2620716766035511, + "grad_norm": 1.3773764371871948, + "learning_rate": 7.998e-05, + "loss": 0.1468, + "step": 4000 + }, + { + "epoch": 0.26272685579505994, + "grad_norm": 1.1586949825286865, + "learning_rate": 8.018e-05, + "loss": 0.1777, + "step": 4010 + }, + { + "epoch": 0.2633820349865688, + "grad_norm": 1.4444698095321655, + "learning_rate": 8.038e-05, + "loss": 0.1104, + "step": 4020 + }, + { + "epoch": 0.2640372141780777, + "grad_norm": 1.1999707221984863, + "learning_rate": 8.058e-05, + "loss": 0.1914, + "step": 4030 + }, + { + "epoch": 0.2646923933695866, + "grad_norm": 1.3994086980819702, + "learning_rate": 8.078e-05, + "loss": 0.1213, + "step": 4040 + }, + { + "epoch": 0.2653475725610955, + "grad_norm": 1.3709372282028198, + "learning_rate": 8.098e-05, + "loss": 0.1598, + "step": 4050 + }, + { + "epoch": 0.26600275175260435, + "grad_norm": 2.5795037746429443, + "learning_rate": 8.118e-05, + "loss": 0.1766, + "step": 4060 + }, + { + "epoch": 0.2666579309441132, + "grad_norm": 1.5037721395492554, + "learning_rate": 8.138e-05, + "loss": 0.1689, + "step": 4070 + }, + { + "epoch": 0.2673131101356221, + "grad_norm": 1.1711986064910889, + "learning_rate": 8.158e-05, + "loss": 0.1351, + "step": 4080 + }, + { + "epoch": 0.26796828932713096, + "grad_norm": 1.818021297454834, + "learning_rate": 8.178e-05, + "loss": 0.1704, + "step": 4090 + }, + { + "epoch": 0.26862346851863983, + "grad_norm": 1.4975029230117798, + "learning_rate": 8.198e-05, + "loss": 0.1446, + "step": 4100 + }, + { + "epoch": 0.2692786477101487, + "grad_norm": 3.4287662506103516, + "learning_rate": 8.218e-05, + "loss": 0.1419, + "step": 4110 + }, + { + "epoch": 0.2699338269016576, + "grad_norm": 1.9511228799819946, + "learning_rate": 8.238000000000001e-05, + "loss": 0.149, + "step": 4120 + }, + { + "epoch": 0.2705890060931665, + "grad_norm": 0.7081625461578369, + "learning_rate": 8.258e-05, + "loss": 0.0945, + "step": 4130 + }, + { + "epoch": 0.27124418528467537, + "grad_norm": 1.733394742012024, + "learning_rate": 8.278e-05, + "loss": 0.1241, + "step": 4140 + }, + { + "epoch": 0.27189936447618424, + "grad_norm": 2.441138982772827, + "learning_rate": 8.298000000000001e-05, + "loss": 0.2558, + "step": 4150 + }, + { + "epoch": 0.2725545436676931, + "grad_norm": 1.005645990371704, + "learning_rate": 8.318e-05, + "loss": 0.1679, + "step": 4160 + }, + { + "epoch": 0.273209722859202, + "grad_norm": 0.6987016201019287, + "learning_rate": 8.338e-05, + "loss": 0.1438, + "step": 4170 + }, + { + "epoch": 0.27386490205071085, + "grad_norm": 0.9902043342590332, + "learning_rate": 8.358e-05, + "loss": 0.1297, + "step": 4180 + }, + { + "epoch": 0.27452008124221977, + "grad_norm": 3.2190651893615723, + "learning_rate": 8.378e-05, + "loss": 0.1654, + "step": 4190 + }, + { + "epoch": 0.27517526043372864, + "grad_norm": 0.5338564515113831, + "learning_rate": 8.398e-05, + "loss": 0.1438, + "step": 4200 + }, + { + "epoch": 0.2758304396252375, + "grad_norm": 2.6433866024017334, + "learning_rate": 8.418e-05, + "loss": 0.1474, + "step": 4210 + }, + { + "epoch": 0.2764856188167464, + "grad_norm": 1.8548399209976196, + "learning_rate": 8.438e-05, + "loss": 0.1648, + "step": 4220 + }, + { + "epoch": 0.27714079800825525, + "grad_norm": 1.0515336990356445, + "learning_rate": 8.458e-05, + "loss": 0.137, + "step": 4230 + }, + { + "epoch": 0.2777959771997641, + "grad_norm": 1.9309583902359009, + "learning_rate": 8.478e-05, + "loss": 0.1576, + "step": 4240 + }, + { + "epoch": 0.278451156391273, + "grad_norm": 1.2727653980255127, + "learning_rate": 8.498e-05, + "loss": 0.1583, + "step": 4250 + }, + { + "epoch": 0.2791063355827819, + "grad_norm": 1.688522219657898, + "learning_rate": 8.518000000000001e-05, + "loss": 0.1417, + "step": 4260 + }, + { + "epoch": 0.2797615147742908, + "grad_norm": 1.6817774772644043, + "learning_rate": 8.538e-05, + "loss": 0.1032, + "step": 4270 + }, + { + "epoch": 0.28041669396579966, + "grad_norm": 0.9618755578994751, + "learning_rate": 8.558e-05, + "loss": 0.1992, + "step": 4280 + }, + { + "epoch": 0.2810718731573085, + "grad_norm": 1.4168503284454346, + "learning_rate": 8.578000000000001e-05, + "loss": 0.1587, + "step": 4290 + }, + { + "epoch": 0.2817270523488174, + "grad_norm": 1.2016980648040771, + "learning_rate": 8.598e-05, + "loss": 0.1327, + "step": 4300 + }, + { + "epoch": 0.28238223154032627, + "grad_norm": 1.1283951997756958, + "learning_rate": 8.618e-05, + "loss": 0.1215, + "step": 4310 + }, + { + "epoch": 0.28303741073183514, + "grad_norm": 0.5870518088340759, + "learning_rate": 8.638000000000001e-05, + "loss": 0.1374, + "step": 4320 + }, + { + "epoch": 0.283692589923344, + "grad_norm": 1.1566929817199707, + "learning_rate": 8.658e-05, + "loss": 0.14, + "step": 4330 + }, + { + "epoch": 0.28434776911485293, + "grad_norm": 1.3193163871765137, + "learning_rate": 8.678e-05, + "loss": 0.1307, + "step": 4340 + }, + { + "epoch": 0.2850029483063618, + "grad_norm": 1.1601078510284424, + "learning_rate": 8.698000000000001e-05, + "loss": 0.1637, + "step": 4350 + }, + { + "epoch": 0.2856581274978707, + "grad_norm": 2.918916702270508, + "learning_rate": 8.718e-05, + "loss": 0.2172, + "step": 4360 + }, + { + "epoch": 0.28631330668937954, + "grad_norm": 1.448207139968872, + "learning_rate": 8.738000000000001e-05, + "loss": 0.1839, + "step": 4370 + }, + { + "epoch": 0.2869684858808884, + "grad_norm": 0.7659546732902527, + "learning_rate": 8.758000000000001e-05, + "loss": 0.1351, + "step": 4380 + }, + { + "epoch": 0.2876236650723973, + "grad_norm": 2.1401641368865967, + "learning_rate": 8.778e-05, + "loss": 0.1313, + "step": 4390 + }, + { + "epoch": 0.28827884426390615, + "grad_norm": 1.097296118736267, + "learning_rate": 8.798000000000001e-05, + "loss": 0.0859, + "step": 4400 + }, + { + "epoch": 0.2889340234554151, + "grad_norm": 1.05695378780365, + "learning_rate": 8.818000000000001e-05, + "loss": 0.14, + "step": 4410 + }, + { + "epoch": 0.28958920264692395, + "grad_norm": 1.703461766242981, + "learning_rate": 8.838e-05, + "loss": 0.1697, + "step": 4420 + }, + { + "epoch": 0.2902443818384328, + "grad_norm": 1.1152857542037964, + "learning_rate": 8.858000000000001e-05, + "loss": 0.2071, + "step": 4430 + }, + { + "epoch": 0.2908995610299417, + "grad_norm": 0.46094396710395813, + "learning_rate": 8.878000000000001e-05, + "loss": 0.1364, + "step": 4440 + }, + { + "epoch": 0.29155474022145056, + "grad_norm": 0.9174224138259888, + "learning_rate": 8.898e-05, + "loss": 0.1304, + "step": 4450 + }, + { + "epoch": 0.29220991941295943, + "grad_norm": 1.2063994407653809, + "learning_rate": 8.918000000000001e-05, + "loss": 0.1082, + "step": 4460 + }, + { + "epoch": 0.2928650986044683, + "grad_norm": 1.6451334953308105, + "learning_rate": 8.938e-05, + "loss": 0.1137, + "step": 4470 + }, + { + "epoch": 0.2935202777959772, + "grad_norm": 1.3914252519607544, + "learning_rate": 8.958e-05, + "loss": 0.1563, + "step": 4480 + }, + { + "epoch": 0.2941754569874861, + "grad_norm": 2.1779699325561523, + "learning_rate": 8.978000000000001e-05, + "loss": 0.1553, + "step": 4490 + }, + { + "epoch": 0.29483063617899496, + "grad_norm": 2.6128666400909424, + "learning_rate": 8.998e-05, + "loss": 0.1397, + "step": 4500 + }, + { + "epoch": 0.29548581537050383, + "grad_norm": 2.018366575241089, + "learning_rate": 9.018000000000001e-05, + "loss": 0.1483, + "step": 4510 + }, + { + "epoch": 0.2961409945620127, + "grad_norm": 0.5061533451080322, + "learning_rate": 9.038000000000001e-05, + "loss": 0.1741, + "step": 4520 + }, + { + "epoch": 0.2967961737535216, + "grad_norm": 0.9906481504440308, + "learning_rate": 9.058e-05, + "loss": 0.1393, + "step": 4530 + }, + { + "epoch": 0.29745135294503044, + "grad_norm": 0.7185648083686829, + "learning_rate": 9.078000000000001e-05, + "loss": 0.1985, + "step": 4540 + }, + { + "epoch": 0.29810653213653937, + "grad_norm": 1.0696905851364136, + "learning_rate": 9.098000000000001e-05, + "loss": 0.1304, + "step": 4550 + }, + { + "epoch": 0.29876171132804824, + "grad_norm": 1.2915929555892944, + "learning_rate": 9.118e-05, + "loss": 0.1592, + "step": 4560 + }, + { + "epoch": 0.2994168905195571, + "grad_norm": 2.280205249786377, + "learning_rate": 9.138e-05, + "loss": 0.1067, + "step": 4570 + }, + { + "epoch": 0.300072069711066, + "grad_norm": 1.226305365562439, + "learning_rate": 9.158e-05, + "loss": 0.1688, + "step": 4580 + }, + { + "epoch": 0.30072724890257485, + "grad_norm": 2.479609489440918, + "learning_rate": 9.178e-05, + "loss": 0.1439, + "step": 4590 + }, + { + "epoch": 0.3013824280940837, + "grad_norm": 0.8052397966384888, + "learning_rate": 9.198e-05, + "loss": 0.1948, + "step": 4600 + }, + { + "epoch": 0.3020376072855926, + "grad_norm": 0.6335459351539612, + "learning_rate": 9.218e-05, + "loss": 0.1103, + "step": 4610 + }, + { + "epoch": 0.30269278647710146, + "grad_norm": 2.3129849433898926, + "learning_rate": 9.238e-05, + "loss": 0.1153, + "step": 4620 + }, + { + "epoch": 0.3033479656686104, + "grad_norm": 1.2992221117019653, + "learning_rate": 9.258e-05, + "loss": 0.1448, + "step": 4630 + }, + { + "epoch": 0.30400314486011926, + "grad_norm": 1.5396881103515625, + "learning_rate": 9.278e-05, + "loss": 0.128, + "step": 4640 + }, + { + "epoch": 0.3046583240516281, + "grad_norm": 1.9114267826080322, + "learning_rate": 9.298e-05, + "loss": 0.1251, + "step": 4650 + }, + { + "epoch": 0.305313503243137, + "grad_norm": 0.9093480706214905, + "learning_rate": 9.318e-05, + "loss": 0.1688, + "step": 4660 + }, + { + "epoch": 0.30596868243464587, + "grad_norm": 2.573664665222168, + "learning_rate": 9.338e-05, + "loss": 0.1412, + "step": 4670 + }, + { + "epoch": 0.30662386162615474, + "grad_norm": 0.9161596894264221, + "learning_rate": 9.358e-05, + "loss": 0.1835, + "step": 4680 + }, + { + "epoch": 0.3072790408176636, + "grad_norm": 0.7955771088600159, + "learning_rate": 9.378e-05, + "loss": 0.1533, + "step": 4690 + }, + { + "epoch": 0.30793422000917253, + "grad_norm": 1.0355494022369385, + "learning_rate": 9.398e-05, + "loss": 0.1906, + "step": 4700 + }, + { + "epoch": 0.3085893992006814, + "grad_norm": 7.218301773071289, + "learning_rate": 9.418e-05, + "loss": 0.1674, + "step": 4710 + }, + { + "epoch": 0.3092445783921903, + "grad_norm": 1.200500249862671, + "learning_rate": 9.438e-05, + "loss": 0.1357, + "step": 4720 + }, + { + "epoch": 0.30989975758369914, + "grad_norm": 1.23796546459198, + "learning_rate": 9.458e-05, + "loss": 0.1606, + "step": 4730 + }, + { + "epoch": 0.310554936775208, + "grad_norm": 1.8856940269470215, + "learning_rate": 9.478e-05, + "loss": 0.129, + "step": 4740 + }, + { + "epoch": 0.3112101159667169, + "grad_norm": 1.046579122543335, + "learning_rate": 9.498e-05, + "loss": 0.1209, + "step": 4750 + }, + { + "epoch": 0.31186529515822575, + "grad_norm": 1.3501943349838257, + "learning_rate": 9.518000000000001e-05, + "loss": 0.117, + "step": 4760 + }, + { + "epoch": 0.3125204743497347, + "grad_norm": 1.262633204460144, + "learning_rate": 9.538e-05, + "loss": 0.1362, + "step": 4770 + }, + { + "epoch": 0.31317565354124355, + "grad_norm": 1.4725573062896729, + "learning_rate": 9.558e-05, + "loss": 0.1736, + "step": 4780 + }, + { + "epoch": 0.3138308327327524, + "grad_norm": 1.370465636253357, + "learning_rate": 9.578000000000001e-05, + "loss": 0.1285, + "step": 4790 + }, + { + "epoch": 0.3144860119242613, + "grad_norm": 1.897956371307373, + "learning_rate": 9.598e-05, + "loss": 0.1362, + "step": 4800 + }, + { + "epoch": 0.31514119111577016, + "grad_norm": 1.4950143098831177, + "learning_rate": 9.618e-05, + "loss": 0.1575, + "step": 4810 + }, + { + "epoch": 0.31579637030727903, + "grad_norm": 1.4705278873443604, + "learning_rate": 9.638000000000001e-05, + "loss": 0.1581, + "step": 4820 + }, + { + "epoch": 0.3164515494987879, + "grad_norm": 2.2909488677978516, + "learning_rate": 9.658e-05, + "loss": 0.1292, + "step": 4830 + }, + { + "epoch": 0.31710672869029677, + "grad_norm": 2.0422136783599854, + "learning_rate": 9.678e-05, + "loss": 0.1291, + "step": 4840 + }, + { + "epoch": 0.3177619078818057, + "grad_norm": 1.6872895956039429, + "learning_rate": 9.698000000000001e-05, + "loss": 0.129, + "step": 4850 + }, + { + "epoch": 0.31841708707331456, + "grad_norm": 1.1081595420837402, + "learning_rate": 9.718e-05, + "loss": 0.118, + "step": 4860 + }, + { + "epoch": 0.31907226626482343, + "grad_norm": 2.090327262878418, + "learning_rate": 9.738e-05, + "loss": 0.1401, + "step": 4870 + }, + { + "epoch": 0.3197274454563323, + "grad_norm": 1.21531343460083, + "learning_rate": 9.758000000000001e-05, + "loss": 0.1323, + "step": 4880 + }, + { + "epoch": 0.3203826246478412, + "grad_norm": 0.8055250644683838, + "learning_rate": 9.778e-05, + "loss": 0.1673, + "step": 4890 + }, + { + "epoch": 0.32103780383935004, + "grad_norm": 1.3795466423034668, + "learning_rate": 9.798000000000001e-05, + "loss": 0.1629, + "step": 4900 + }, + { + "epoch": 0.3216929830308589, + "grad_norm": 1.1943589448928833, + "learning_rate": 9.818000000000001e-05, + "loss": 0.1409, + "step": 4910 + }, + { + "epoch": 0.32234816222236784, + "grad_norm": 1.4050604104995728, + "learning_rate": 9.838e-05, + "loss": 0.1283, + "step": 4920 + }, + { + "epoch": 0.3230033414138767, + "grad_norm": 1.0482237339019775, + "learning_rate": 9.858000000000001e-05, + "loss": 0.1755, + "step": 4930 + }, + { + "epoch": 0.3236585206053856, + "grad_norm": 0.6986154913902283, + "learning_rate": 9.878e-05, + "loss": 0.1643, + "step": 4940 + }, + { + "epoch": 0.32431369979689445, + "grad_norm": 1.0826584100723267, + "learning_rate": 9.898e-05, + "loss": 0.1052, + "step": 4950 + }, + { + "epoch": 0.3249688789884033, + "grad_norm": 0.8819634914398193, + "learning_rate": 9.918000000000001e-05, + "loss": 0.1611, + "step": 4960 + }, + { + "epoch": 0.3256240581799122, + "grad_norm": 0.8781566023826599, + "learning_rate": 9.938e-05, + "loss": 0.1612, + "step": 4970 + }, + { + "epoch": 0.32627923737142106, + "grad_norm": 0.8286423087120056, + "learning_rate": 9.958e-05, + "loss": 0.0711, + "step": 4980 + }, + { + "epoch": 0.32693441656293, + "grad_norm": 1.0799449682235718, + "learning_rate": 9.978000000000001e-05, + "loss": 0.1106, + "step": 4990 + }, + { + "epoch": 0.32758959575443886, + "grad_norm": 2.4590301513671875, + "learning_rate": 9.998e-05, + "loss": 0.1431, + "step": 5000 + }, + { + "epoch": 0.3282447749459477, + "grad_norm": 1.8527584075927734, + "learning_rate": 9.999999778549045e-05, + "loss": 0.1598, + "step": 5010 + }, + { + "epoch": 0.3288999541374566, + "grad_norm": 1.6811898946762085, + "learning_rate": 9.999999013039593e-05, + "loss": 0.1487, + "step": 5020 + }, + { + "epoch": 0.32955513332896547, + "grad_norm": 0.6961972117424011, + "learning_rate": 9.999997700737766e-05, + "loss": 0.1015, + "step": 5030 + }, + { + "epoch": 0.33021031252047434, + "grad_norm": 0.6481174230575562, + "learning_rate": 9.999995841643709e-05, + "loss": 0.1125, + "step": 5040 + }, + { + "epoch": 0.3308654917119832, + "grad_norm": 1.4466450214385986, + "learning_rate": 9.999993435757623e-05, + "loss": 0.1126, + "step": 5050 + }, + { + "epoch": 0.33152067090349213, + "grad_norm": 1.5542027950286865, + "learning_rate": 9.999990483079773e-05, + "loss": 0.1353, + "step": 5060 + }, + { + "epoch": 0.332175850095001, + "grad_norm": 2.0507278442382812, + "learning_rate": 9.999986983610481e-05, + "loss": 0.1547, + "step": 5070 + }, + { + "epoch": 0.33283102928650987, + "grad_norm": 1.3246705532073975, + "learning_rate": 9.99998293735013e-05, + "loss": 0.1288, + "step": 5080 + }, + { + "epoch": 0.33348620847801874, + "grad_norm": 1.1588687896728516, + "learning_rate": 9.999978344299161e-05, + "loss": 0.1041, + "step": 5090 + }, + { + "epoch": 0.3341413876695276, + "grad_norm": 1.5442335605621338, + "learning_rate": 9.99997320445808e-05, + "loss": 0.1582, + "step": 5100 + }, + { + "epoch": 0.3347965668610365, + "grad_norm": 1.0099518299102783, + "learning_rate": 9.999967517827444e-05, + "loss": 0.1072, + "step": 5110 + }, + { + "epoch": 0.33545174605254535, + "grad_norm": 0.7770519256591797, + "learning_rate": 9.999961284407879e-05, + "loss": 0.1648, + "step": 5120 + }, + { + "epoch": 0.3361069252440542, + "grad_norm": 1.4146250486373901, + "learning_rate": 9.999954504200067e-05, + "loss": 0.1338, + "step": 5130 + }, + { + "epoch": 0.33676210443556315, + "grad_norm": 1.6459146738052368, + "learning_rate": 9.999947177204744e-05, + "loss": 0.1687, + "step": 5140 + }, + { + "epoch": 0.337417283627072, + "grad_norm": 0.9897658228874207, + "learning_rate": 9.999939303422718e-05, + "loss": 0.1501, + "step": 5150 + }, + { + "epoch": 0.3380724628185809, + "grad_norm": 1.3201764822006226, + "learning_rate": 9.999930882854847e-05, + "loss": 0.1313, + "step": 5160 + }, + { + "epoch": 0.33872764201008976, + "grad_norm": 1.6178340911865234, + "learning_rate": 9.999921915502051e-05, + "loss": 0.1365, + "step": 5170 + }, + { + "epoch": 0.33938282120159863, + "grad_norm": 0.9334473013877869, + "learning_rate": 9.99991240136531e-05, + "loss": 0.1502, + "step": 5180 + }, + { + "epoch": 0.3400380003931075, + "grad_norm": 0.6862757205963135, + "learning_rate": 9.999902340445668e-05, + "loss": 0.142, + "step": 5190 + }, + { + "epoch": 0.34069317958461637, + "grad_norm": 1.6129131317138672, + "learning_rate": 9.999891732744224e-05, + "loss": 0.1626, + "step": 5200 + }, + { + "epoch": 0.3413483587761253, + "grad_norm": 1.0765351057052612, + "learning_rate": 9.999880578262135e-05, + "loss": 0.14, + "step": 5210 + }, + { + "epoch": 0.34200353796763416, + "grad_norm": 1.1877622604370117, + "learning_rate": 9.999868877000624e-05, + "loss": 0.1418, + "step": 5220 + }, + { + "epoch": 0.34265871715914303, + "grad_norm": 1.001304030418396, + "learning_rate": 9.99985662896097e-05, + "loss": 0.1099, + "step": 5230 + }, + { + "epoch": 0.3433138963506519, + "grad_norm": 1.3699877262115479, + "learning_rate": 9.999843834144513e-05, + "loss": 0.1136, + "step": 5240 + }, + { + "epoch": 0.3439690755421608, + "grad_norm": 1.1385679244995117, + "learning_rate": 9.99983049255265e-05, + "loss": 0.1202, + "step": 5250 + }, + { + "epoch": 0.34462425473366964, + "grad_norm": 0.7427843809127808, + "learning_rate": 9.999816604186843e-05, + "loss": 0.1195, + "step": 5260 + }, + { + "epoch": 0.3452794339251785, + "grad_norm": 1.2690964937210083, + "learning_rate": 9.999802169048609e-05, + "loss": 0.1148, + "step": 5270 + }, + { + "epoch": 0.34593461311668744, + "grad_norm": 1.2692551612854004, + "learning_rate": 9.999787187139527e-05, + "loss": 0.1588, + "step": 5280 + }, + { + "epoch": 0.3465897923081963, + "grad_norm": 2.8705244064331055, + "learning_rate": 9.999771658461234e-05, + "loss": 0.1081, + "step": 5290 + }, + { + "epoch": 0.3472449714997052, + "grad_norm": 2.551131248474121, + "learning_rate": 9.999755583015431e-05, + "loss": 0.1077, + "step": 5300 + }, + { + "epoch": 0.34790015069121405, + "grad_norm": 2.188859224319458, + "learning_rate": 9.999738960803874e-05, + "loss": 0.1219, + "step": 5310 + }, + { + "epoch": 0.3485553298827229, + "grad_norm": 2.0064408779144287, + "learning_rate": 9.99972179182838e-05, + "loss": 0.1346, + "step": 5320 + }, + { + "epoch": 0.3492105090742318, + "grad_norm": 0.678026020526886, + "learning_rate": 9.99970407609083e-05, + "loss": 0.1086, + "step": 5330 + }, + { + "epoch": 0.34986568826574066, + "grad_norm": 1.1718271970748901, + "learning_rate": 9.999685813593159e-05, + "loss": 0.0912, + "step": 5340 + }, + { + "epoch": 0.3505208674572496, + "grad_norm": 1.0009785890579224, + "learning_rate": 9.999667004337362e-05, + "loss": 0.1067, + "step": 5350 + }, + { + "epoch": 0.35117604664875846, + "grad_norm": 0.6089062094688416, + "learning_rate": 9.9996476483255e-05, + "loss": 0.0982, + "step": 5360 + }, + { + "epoch": 0.3518312258402673, + "grad_norm": 0.8668644428253174, + "learning_rate": 9.999627745559688e-05, + "loss": 0.1358, + "step": 5370 + }, + { + "epoch": 0.3524864050317762, + "grad_norm": 2.0708513259887695, + "learning_rate": 9.999607296042101e-05, + "loss": 0.0882, + "step": 5380 + }, + { + "epoch": 0.35314158422328507, + "grad_norm": 0.762812614440918, + "learning_rate": 9.99958629977498e-05, + "loss": 0.1151, + "step": 5390 + }, + { + "epoch": 0.35379676341479394, + "grad_norm": 1.286628246307373, + "learning_rate": 9.999564756760615e-05, + "loss": 0.1592, + "step": 5400 + }, + { + "epoch": 0.3544519426063028, + "grad_norm": 1.151953935623169, + "learning_rate": 9.999542667001366e-05, + "loss": 0.1209, + "step": 5410 + }, + { + "epoch": 0.3551071217978117, + "grad_norm": 1.3202859163284302, + "learning_rate": 9.999520030499647e-05, + "loss": 0.1593, + "step": 5420 + }, + { + "epoch": 0.3557623009893206, + "grad_norm": 0.5973538160324097, + "learning_rate": 9.999496847257936e-05, + "loss": 0.1176, + "step": 5430 + }, + { + "epoch": 0.35641748018082947, + "grad_norm": 1.0574216842651367, + "learning_rate": 9.999473117278764e-05, + "loss": 0.0893, + "step": 5440 + }, + { + "epoch": 0.35707265937233834, + "grad_norm": 0.758423924446106, + "learning_rate": 9.999448840564731e-05, + "loss": 0.1111, + "step": 5450 + }, + { + "epoch": 0.3577278385638472, + "grad_norm": 0.8484379053115845, + "learning_rate": 9.999424017118488e-05, + "loss": 0.1367, + "step": 5460 + }, + { + "epoch": 0.3583830177553561, + "grad_norm": 1.58161461353302, + "learning_rate": 9.999398646942751e-05, + "loss": 0.1185, + "step": 5470 + }, + { + "epoch": 0.35903819694686495, + "grad_norm": 0.879215657711029, + "learning_rate": 9.999372730040296e-05, + "loss": 0.1027, + "step": 5480 + }, + { + "epoch": 0.3596933761383738, + "grad_norm": 1.148396611213684, + "learning_rate": 9.999346266413953e-05, + "loss": 0.0988, + "step": 5490 + }, + { + "epoch": 0.36034855532988275, + "grad_norm": 0.8058022856712341, + "learning_rate": 9.99931925606662e-05, + "loss": 0.1054, + "step": 5500 + }, + { + "epoch": 0.3610037345213916, + "grad_norm": 0.8967760801315308, + "learning_rate": 9.99929169900125e-05, + "loss": 0.0978, + "step": 5510 + }, + { + "epoch": 0.3616589137129005, + "grad_norm": 0.9036872982978821, + "learning_rate": 9.999263595220855e-05, + "loss": 0.1044, + "step": 5520 + }, + { + "epoch": 0.36231409290440936, + "grad_norm": 1.2660173177719116, + "learning_rate": 9.99923494472851e-05, + "loss": 0.1334, + "step": 5530 + }, + { + "epoch": 0.3629692720959182, + "grad_norm": 0.7284460663795471, + "learning_rate": 9.999205747527348e-05, + "loss": 0.0882, + "step": 5540 + }, + { + "epoch": 0.3636244512874271, + "grad_norm": 1.5239609479904175, + "learning_rate": 9.999176003620561e-05, + "loss": 0.1263, + "step": 5550 + }, + { + "epoch": 0.36427963047893597, + "grad_norm": 0.5141845941543579, + "learning_rate": 9.999145713011405e-05, + "loss": 0.1247, + "step": 5560 + }, + { + "epoch": 0.3649348096704449, + "grad_norm": 1.515821099281311, + "learning_rate": 9.999114875703186e-05, + "loss": 0.106, + "step": 5570 + }, + { + "epoch": 0.36558998886195376, + "grad_norm": 0.6357114315032959, + "learning_rate": 9.999083491699281e-05, + "loss": 0.0986, + "step": 5580 + }, + { + "epoch": 0.36624516805346263, + "grad_norm": 0.6075281500816345, + "learning_rate": 9.999051561003123e-05, + "loss": 0.1443, + "step": 5590 + }, + { + "epoch": 0.3669003472449715, + "grad_norm": 1.9954540729522705, + "learning_rate": 9.999019083618202e-05, + "loss": 0.092, + "step": 5600 + }, + { + "epoch": 0.3675555264364804, + "grad_norm": 0.7516798377037048, + "learning_rate": 9.99898605954807e-05, + "loss": 0.1112, + "step": 5610 + }, + { + "epoch": 0.36821070562798924, + "grad_norm": 1.4810736179351807, + "learning_rate": 9.998952488796338e-05, + "loss": 0.1413, + "step": 5620 + }, + { + "epoch": 0.3688658848194981, + "grad_norm": 0.8783677816390991, + "learning_rate": 9.998918371366676e-05, + "loss": 0.1056, + "step": 5630 + }, + { + "epoch": 0.369521064011007, + "grad_norm": 0.6852719187736511, + "learning_rate": 9.99888370726282e-05, + "loss": 0.1033, + "step": 5640 + }, + { + "epoch": 0.3701762432025159, + "grad_norm": 0.4219054579734802, + "learning_rate": 9.998848496488556e-05, + "loss": 0.1019, + "step": 5650 + }, + { + "epoch": 0.3708314223940248, + "grad_norm": 1.254939079284668, + "learning_rate": 9.998812739047736e-05, + "loss": 0.0886, + "step": 5660 + }, + { + "epoch": 0.37148660158553365, + "grad_norm": 0.8786806464195251, + "learning_rate": 9.99877643494427e-05, + "loss": 0.1484, + "step": 5670 + }, + { + "epoch": 0.3721417807770425, + "grad_norm": 1.7643705606460571, + "learning_rate": 9.998739584182128e-05, + "loss": 0.1014, + "step": 5680 + }, + { + "epoch": 0.3727969599685514, + "grad_norm": 0.9754090905189514, + "learning_rate": 9.998702186765342e-05, + "loss": 0.0961, + "step": 5690 + }, + { + "epoch": 0.37345213916006026, + "grad_norm": 1.1691508293151855, + "learning_rate": 9.998664242698e-05, + "loss": 0.073, + "step": 5700 + }, + { + "epoch": 0.37410731835156913, + "grad_norm": 0.5739044547080994, + "learning_rate": 9.998625751984251e-05, + "loss": 0.0875, + "step": 5710 + }, + { + "epoch": 0.37476249754307805, + "grad_norm": 1.2009127140045166, + "learning_rate": 9.998586714628307e-05, + "loss": 0.0778, + "step": 5720 + }, + { + "epoch": 0.3754176767345869, + "grad_norm": 1.7604800462722778, + "learning_rate": 9.998547130634432e-05, + "loss": 0.1109, + "step": 5730 + }, + { + "epoch": 0.3760728559260958, + "grad_norm": 1.1640541553497314, + "learning_rate": 9.99850700000696e-05, + "loss": 0.1028, + "step": 5740 + }, + { + "epoch": 0.37672803511760466, + "grad_norm": 0.5109245181083679, + "learning_rate": 9.998466322750278e-05, + "loss": 0.1267, + "step": 5750 + }, + { + "epoch": 0.37738321430911353, + "grad_norm": 1.1438277959823608, + "learning_rate": 9.998425098868834e-05, + "loss": 0.128, + "step": 5760 + }, + { + "epoch": 0.3780383935006224, + "grad_norm": 0.5898852944374084, + "learning_rate": 9.998383328367136e-05, + "loss": 0.1337, + "step": 5770 + }, + { + "epoch": 0.3786935726921313, + "grad_norm": 1.224158525466919, + "learning_rate": 9.99834101124975e-05, + "loss": 0.1428, + "step": 5780 + }, + { + "epoch": 0.3793487518836402, + "grad_norm": 0.7096456289291382, + "learning_rate": 9.998298147521309e-05, + "loss": 0.1248, + "step": 5790 + }, + { + "epoch": 0.38000393107514907, + "grad_norm": 1.3166065216064453, + "learning_rate": 9.998254737186496e-05, + "loss": 0.1254, + "step": 5800 + }, + { + "epoch": 0.38065911026665794, + "grad_norm": 1.5239399671554565, + "learning_rate": 9.99821078025006e-05, + "loss": 0.1222, + "step": 5810 + }, + { + "epoch": 0.3813142894581668, + "grad_norm": 2.0747790336608887, + "learning_rate": 9.998166276716807e-05, + "loss": 0.1211, + "step": 5820 + }, + { + "epoch": 0.3819694686496757, + "grad_norm": 1.119019865989685, + "learning_rate": 9.998121226591606e-05, + "loss": 0.0613, + "step": 5830 + }, + { + "epoch": 0.38262464784118455, + "grad_norm": 1.054911494255066, + "learning_rate": 9.998075629879382e-05, + "loss": 0.0986, + "step": 5840 + }, + { + "epoch": 0.3832798270326934, + "grad_norm": 0.6126089692115784, + "learning_rate": 9.99802948658512e-05, + "loss": 0.1698, + "step": 5850 + }, + { + "epoch": 0.38393500622420235, + "grad_norm": 0.8110992908477783, + "learning_rate": 9.99798279671387e-05, + "loss": 0.1019, + "step": 5860 + }, + { + "epoch": 0.3845901854157112, + "grad_norm": 0.8765296936035156, + "learning_rate": 9.997935560270734e-05, + "loss": 0.134, + "step": 5870 + }, + { + "epoch": 0.3852453646072201, + "grad_norm": 0.7867010831832886, + "learning_rate": 9.997887777260879e-05, + "loss": 0.1242, + "step": 5880 + }, + { + "epoch": 0.38590054379872896, + "grad_norm": 2.181455373764038, + "learning_rate": 9.997839447689532e-05, + "loss": 0.1167, + "step": 5890 + }, + { + "epoch": 0.3865557229902378, + "grad_norm": 0.8233321905136108, + "learning_rate": 9.997790571561978e-05, + "loss": 0.0991, + "step": 5900 + }, + { + "epoch": 0.3872109021817467, + "grad_norm": 0.662626326084137, + "learning_rate": 9.99774114888356e-05, + "loss": 0.0974, + "step": 5910 + }, + { + "epoch": 0.38786608137325557, + "grad_norm": 0.8471824526786804, + "learning_rate": 9.997691179659684e-05, + "loss": 0.1326, + "step": 5920 + }, + { + "epoch": 0.38852126056476444, + "grad_norm": 0.8508082628250122, + "learning_rate": 9.997640663895815e-05, + "loss": 0.0922, + "step": 5930 + }, + { + "epoch": 0.38917643975627336, + "grad_norm": 1.918066382408142, + "learning_rate": 9.997589601597477e-05, + "loss": 0.1276, + "step": 5940 + }, + { + "epoch": 0.38983161894778223, + "grad_norm": 0.8431815505027771, + "learning_rate": 9.997537992770252e-05, + "loss": 0.1226, + "step": 5950 + }, + { + "epoch": 0.3904867981392911, + "grad_norm": 1.3359280824661255, + "learning_rate": 9.997485837419788e-05, + "loss": 0.124, + "step": 5960 + }, + { + "epoch": 0.3911419773308, + "grad_norm": 1.5926513671875, + "learning_rate": 9.997433135551786e-05, + "loss": 0.1443, + "step": 5970 + }, + { + "epoch": 0.39179715652230884, + "grad_norm": 1.2440940141677856, + "learning_rate": 9.997379887172009e-05, + "loss": 0.1211, + "step": 5980 + }, + { + "epoch": 0.3924523357138177, + "grad_norm": 1.4363884925842285, + "learning_rate": 9.997326092286281e-05, + "loss": 0.128, + "step": 5990 + }, + { + "epoch": 0.3931075149053266, + "grad_norm": 1.7993299961090088, + "learning_rate": 9.997271750900486e-05, + "loss": 0.1058, + "step": 6000 + }, + { + "epoch": 0.3937626940968355, + "grad_norm": 0.896467924118042, + "learning_rate": 9.997216863020565e-05, + "loss": 0.1075, + "step": 6010 + }, + { + "epoch": 0.3944178732883444, + "grad_norm": 0.3631545901298523, + "learning_rate": 9.99716142865252e-05, + "loss": 0.1107, + "step": 6020 + }, + { + "epoch": 0.39507305247985325, + "grad_norm": 0.6186422109603882, + "learning_rate": 9.997105447802415e-05, + "loss": 0.0938, + "step": 6030 + }, + { + "epoch": 0.3957282316713621, + "grad_norm": 0.9157413244247437, + "learning_rate": 9.997048920476373e-05, + "loss": 0.0974, + "step": 6040 + }, + { + "epoch": 0.396383410862871, + "grad_norm": 1.3931951522827148, + "learning_rate": 9.996991846680572e-05, + "loss": 0.1256, + "step": 6050 + }, + { + "epoch": 0.39703859005437986, + "grad_norm": 1.0801360607147217, + "learning_rate": 9.996934226421257e-05, + "loss": 0.0908, + "step": 6060 + }, + { + "epoch": 0.39769376924588873, + "grad_norm": 2.1857738494873047, + "learning_rate": 9.996876059704726e-05, + "loss": 0.1217, + "step": 6070 + }, + { + "epoch": 0.39834894843739765, + "grad_norm": 1.396328330039978, + "learning_rate": 9.996817346537343e-05, + "loss": 0.1152, + "step": 6080 + }, + { + "epoch": 0.3990041276289065, + "grad_norm": 0.8219853639602661, + "learning_rate": 9.996758086925526e-05, + "loss": 0.1045, + "step": 6090 + }, + { + "epoch": 0.3996593068204154, + "grad_norm": 1.6711665391921997, + "learning_rate": 9.996698280875759e-05, + "loss": 0.1239, + "step": 6100 + }, + { + "epoch": 0.40031448601192426, + "grad_norm": 0.6516497731208801, + "learning_rate": 9.99663792839458e-05, + "loss": 0.1326, + "step": 6110 + }, + { + "epoch": 0.40096966520343313, + "grad_norm": 0.6820084452629089, + "learning_rate": 9.99657702948859e-05, + "loss": 0.0824, + "step": 6120 + }, + { + "epoch": 0.401624844394942, + "grad_norm": 0.7127842307090759, + "learning_rate": 9.996515584164448e-05, + "loss": 0.0983, + "step": 6130 + }, + { + "epoch": 0.4022800235864509, + "grad_norm": 1.1312298774719238, + "learning_rate": 9.996453592428873e-05, + "loss": 0.1239, + "step": 6140 + }, + { + "epoch": 0.40293520277795974, + "grad_norm": 4.648057460784912, + "learning_rate": 9.996391054288646e-05, + "loss": 0.0906, + "step": 6150 + }, + { + "epoch": 0.40359038196946867, + "grad_norm": 2.395592451095581, + "learning_rate": 9.996327969750605e-05, + "loss": 0.1166, + "step": 6160 + }, + { + "epoch": 0.40424556116097754, + "grad_norm": 1.2276068925857544, + "learning_rate": 9.996264338821649e-05, + "loss": 0.1048, + "step": 6170 + }, + { + "epoch": 0.4049007403524864, + "grad_norm": 1.3626935482025146, + "learning_rate": 9.996200161508735e-05, + "loss": 0.1151, + "step": 6180 + }, + { + "epoch": 0.4055559195439953, + "grad_norm": 1.4251824617385864, + "learning_rate": 9.996135437818885e-05, + "loss": 0.1152, + "step": 6190 + }, + { + "epoch": 0.40621109873550415, + "grad_norm": 1.6418689489364624, + "learning_rate": 9.996070167759175e-05, + "loss": 0.1077, + "step": 6200 + }, + { + "epoch": 0.406866277927013, + "grad_norm": 1.6412138938903809, + "learning_rate": 9.996004351336743e-05, + "loss": 0.1515, + "step": 6210 + }, + { + "epoch": 0.4075214571185219, + "grad_norm": 0.37006136775016785, + "learning_rate": 9.995937988558785e-05, + "loss": 0.0877, + "step": 6220 + }, + { + "epoch": 0.4081766363100308, + "grad_norm": 0.7965475916862488, + "learning_rate": 9.995871079432561e-05, + "loss": 0.1281, + "step": 6230 + }, + { + "epoch": 0.4088318155015397, + "grad_norm": 1.7747098207473755, + "learning_rate": 9.995803623965389e-05, + "loss": 0.1009, + "step": 6240 + }, + { + "epoch": 0.40948699469304856, + "grad_norm": 0.5538225769996643, + "learning_rate": 9.995735622164641e-05, + "loss": 0.0821, + "step": 6250 + }, + { + "epoch": 0.4101421738845574, + "grad_norm": 1.2739512920379639, + "learning_rate": 9.995667074037758e-05, + "loss": 0.111, + "step": 6260 + }, + { + "epoch": 0.4107973530760663, + "grad_norm": 1.0575718879699707, + "learning_rate": 9.995597979592232e-05, + "loss": 0.1014, + "step": 6270 + }, + { + "epoch": 0.41145253226757517, + "grad_norm": 0.8387269973754883, + "learning_rate": 9.995528338835625e-05, + "loss": 0.1342, + "step": 6280 + }, + { + "epoch": 0.41210771145908404, + "grad_norm": 0.8791985511779785, + "learning_rate": 9.995458151775547e-05, + "loss": 0.0819, + "step": 6290 + }, + { + "epoch": 0.41276289065059296, + "grad_norm": 0.9614390730857849, + "learning_rate": 9.995387418419677e-05, + "loss": 0.0939, + "step": 6300 + }, + { + "epoch": 0.41341806984210183, + "grad_norm": 1.671506404876709, + "learning_rate": 9.99531613877575e-05, + "loss": 0.0669, + "step": 6310 + }, + { + "epoch": 0.4140732490336107, + "grad_norm": 2.1071810722351074, + "learning_rate": 9.995244312851559e-05, + "loss": 0.1075, + "step": 6320 + }, + { + "epoch": 0.41472842822511957, + "grad_norm": 1.0975126028060913, + "learning_rate": 9.995171940654961e-05, + "loss": 0.1136, + "step": 6330 + }, + { + "epoch": 0.41538360741662844, + "grad_norm": 0.6586717963218689, + "learning_rate": 9.995099022193871e-05, + "loss": 0.1068, + "step": 6340 + }, + { + "epoch": 0.4160387866081373, + "grad_norm": 0.9574748873710632, + "learning_rate": 9.995025557476261e-05, + "loss": 0.096, + "step": 6350 + }, + { + "epoch": 0.4166939657996462, + "grad_norm": 1.7618354558944702, + "learning_rate": 9.994951546510165e-05, + "loss": 0.0998, + "step": 6360 + }, + { + "epoch": 0.4173491449911551, + "grad_norm": 1.2486107349395752, + "learning_rate": 9.994876989303679e-05, + "loss": 0.0875, + "step": 6370 + }, + { + "epoch": 0.418004324182664, + "grad_norm": 1.4967548847198486, + "learning_rate": 9.994801885864955e-05, + "loss": 0.0873, + "step": 6380 + }, + { + "epoch": 0.41865950337417285, + "grad_norm": 1.1251213550567627, + "learning_rate": 9.994726236202205e-05, + "loss": 0.1399, + "step": 6390 + }, + { + "epoch": 0.4193146825656817, + "grad_norm": 1.2293003797531128, + "learning_rate": 9.994650040323704e-05, + "loss": 0.0889, + "step": 6400 + }, + { + "epoch": 0.4199698617571906, + "grad_norm": 0.7436334490776062, + "learning_rate": 9.994573298237784e-05, + "loss": 0.0907, + "step": 6410 + }, + { + "epoch": 0.42062504094869946, + "grad_norm": 1.1720842123031616, + "learning_rate": 9.994496009952837e-05, + "loss": 0.1508, + "step": 6420 + }, + { + "epoch": 0.42128022014020833, + "grad_norm": 0.7169551253318787, + "learning_rate": 9.994418175477316e-05, + "loss": 0.072, + "step": 6430 + }, + { + "epoch": 0.4219353993317172, + "grad_norm": 0.7229377627372742, + "learning_rate": 9.994339794819733e-05, + "loss": 0.0982, + "step": 6440 + }, + { + "epoch": 0.4225905785232261, + "grad_norm": 0.523935854434967, + "learning_rate": 9.994260867988658e-05, + "loss": 0.0675, + "step": 6450 + }, + { + "epoch": 0.423245757714735, + "grad_norm": 0.8462164998054504, + "learning_rate": 9.994181394992723e-05, + "loss": 0.094, + "step": 6460 + }, + { + "epoch": 0.42390093690624386, + "grad_norm": 1.1228363513946533, + "learning_rate": 9.994101375840618e-05, + "loss": 0.1148, + "step": 6470 + }, + { + "epoch": 0.42455611609775273, + "grad_norm": 0.9980197548866272, + "learning_rate": 9.994020810541098e-05, + "loss": 0.0936, + "step": 6480 + }, + { + "epoch": 0.4252112952892616, + "grad_norm": 0.6342594623565674, + "learning_rate": 9.99393969910297e-05, + "loss": 0.1279, + "step": 6490 + }, + { + "epoch": 0.4258664744807705, + "grad_norm": 2.033534288406372, + "learning_rate": 9.993858041535104e-05, + "loss": 0.0835, + "step": 6500 + }, + { + "epoch": 0.42652165367227934, + "grad_norm": 0.593916654586792, + "learning_rate": 9.99377583784643e-05, + "loss": 0.1414, + "step": 6510 + }, + { + "epoch": 0.42717683286378827, + "grad_norm": 0.7929761409759521, + "learning_rate": 9.993693088045939e-05, + "loss": 0.15, + "step": 6520 + }, + { + "epoch": 0.42783201205529714, + "grad_norm": 1.8248257637023926, + "learning_rate": 9.99360979214268e-05, + "loss": 0.0883, + "step": 6530 + }, + { + "epoch": 0.428487191246806, + "grad_norm": 0.8858325481414795, + "learning_rate": 9.99352595014576e-05, + "loss": 0.1261, + "step": 6540 + }, + { + "epoch": 0.4291423704383149, + "grad_norm": 1.1433091163635254, + "learning_rate": 9.993441562064354e-05, + "loss": 0.1016, + "step": 6550 + }, + { + "epoch": 0.42979754962982375, + "grad_norm": 0.8156454563140869, + "learning_rate": 9.993356627907685e-05, + "loss": 0.1222, + "step": 6560 + }, + { + "epoch": 0.4304527288213326, + "grad_norm": 0.5646882653236389, + "learning_rate": 9.99327114768504e-05, + "loss": 0.104, + "step": 6570 + }, + { + "epoch": 0.4311079080128415, + "grad_norm": 0.6943109631538391, + "learning_rate": 9.99318512140577e-05, + "loss": 0.0881, + "step": 6580 + }, + { + "epoch": 0.4317630872043504, + "grad_norm": 0.6117822527885437, + "learning_rate": 9.993098549079284e-05, + "loss": 0.0866, + "step": 6590 + }, + { + "epoch": 0.4324182663958593, + "grad_norm": 0.6713815927505493, + "learning_rate": 9.993011430715047e-05, + "loss": 0.0822, + "step": 6600 + }, + { + "epoch": 0.43307344558736816, + "grad_norm": 0.5744658708572388, + "learning_rate": 9.992923766322586e-05, + "loss": 0.084, + "step": 6610 + }, + { + "epoch": 0.433728624778877, + "grad_norm": 1.1191011667251587, + "learning_rate": 9.99283555591149e-05, + "loss": 0.0884, + "step": 6620 + }, + { + "epoch": 0.4343838039703859, + "grad_norm": 0.9858352541923523, + "learning_rate": 9.992746799491404e-05, + "loss": 0.1116, + "step": 6630 + }, + { + "epoch": 0.43503898316189477, + "grad_norm": 0.5387372970581055, + "learning_rate": 9.992657497072033e-05, + "loss": 0.0842, + "step": 6640 + }, + { + "epoch": 0.43569416235340364, + "grad_norm": 2.281222343444824, + "learning_rate": 9.992567648663147e-05, + "loss": 0.0656, + "step": 6650 + }, + { + "epoch": 0.43634934154491256, + "grad_norm": 1.0955073833465576, + "learning_rate": 9.992477254274568e-05, + "loss": 0.0747, + "step": 6660 + }, + { + "epoch": 0.43700452073642143, + "grad_norm": 0.8766685128211975, + "learning_rate": 9.992386313916183e-05, + "loss": 0.1127, + "step": 6670 + }, + { + "epoch": 0.4376596999279303, + "grad_norm": 0.6184762120246887, + "learning_rate": 9.992294827597934e-05, + "loss": 0.074, + "step": 6680 + }, + { + "epoch": 0.43831487911943917, + "grad_norm": 0.8691467046737671, + "learning_rate": 9.992202795329831e-05, + "loss": 0.0876, + "step": 6690 + }, + { + "epoch": 0.43897005831094804, + "grad_norm": 2.838791847229004, + "learning_rate": 9.992110217121936e-05, + "loss": 0.0728, + "step": 6700 + }, + { + "epoch": 0.4396252375024569, + "grad_norm": 0.8513792157173157, + "learning_rate": 9.992017092984372e-05, + "loss": 0.0858, + "step": 6710 + }, + { + "epoch": 0.4402804166939658, + "grad_norm": 1.4541633129119873, + "learning_rate": 9.991923422927326e-05, + "loss": 0.104, + "step": 6720 + }, + { + "epoch": 0.44093559588547465, + "grad_norm": 1.898011326789856, + "learning_rate": 9.991829206961037e-05, + "loss": 0.1225, + "step": 6730 + }, + { + "epoch": 0.4415907750769836, + "grad_norm": 0.8336621522903442, + "learning_rate": 9.991734445095813e-05, + "loss": 0.0754, + "step": 6740 + }, + { + "epoch": 0.44224595426849245, + "grad_norm": 1.6872975826263428, + "learning_rate": 9.991639137342015e-05, + "loss": 0.095, + "step": 6750 + }, + { + "epoch": 0.4429011334600013, + "grad_norm": 1.1406311988830566, + "learning_rate": 9.991543283710064e-05, + "loss": 0.0724, + "step": 6760 + }, + { + "epoch": 0.4435563126515102, + "grad_norm": 0.7642385363578796, + "learning_rate": 9.991446884210445e-05, + "loss": 0.0698, + "step": 6770 + }, + { + "epoch": 0.44421149184301906, + "grad_norm": 1.7113796472549438, + "learning_rate": 9.9913499388537e-05, + "loss": 0.0986, + "step": 6780 + }, + { + "epoch": 0.4448666710345279, + "grad_norm": 1.0370217561721802, + "learning_rate": 9.99125244765043e-05, + "loss": 0.0778, + "step": 6790 + }, + { + "epoch": 0.4455218502260368, + "grad_norm": 0.696824312210083, + "learning_rate": 9.991154410611296e-05, + "loss": 0.0977, + "step": 6800 + }, + { + "epoch": 0.4461770294175457, + "grad_norm": 1.038401484489441, + "learning_rate": 9.99105582774702e-05, + "loss": 0.0729, + "step": 6810 + }, + { + "epoch": 0.4468322086090546, + "grad_norm": 1.19156813621521, + "learning_rate": 9.990956699068384e-05, + "loss": 0.1018, + "step": 6820 + }, + { + "epoch": 0.44748738780056346, + "grad_norm": 1.1636008024215698, + "learning_rate": 9.990857024586224e-05, + "loss": 0.1093, + "step": 6830 + }, + { + "epoch": 0.44814256699207233, + "grad_norm": 0.6067702770233154, + "learning_rate": 9.990756804311446e-05, + "loss": 0.0848, + "step": 6840 + }, + { + "epoch": 0.4487977461835812, + "grad_norm": 0.8366654515266418, + "learning_rate": 9.990656038255006e-05, + "loss": 0.119, + "step": 6850 + }, + { + "epoch": 0.4494529253750901, + "grad_norm": 0.8921740055084229, + "learning_rate": 9.990554726427926e-05, + "loss": 0.0887, + "step": 6860 + }, + { + "epoch": 0.45010810456659894, + "grad_norm": 0.6565149426460266, + "learning_rate": 9.990452868841284e-05, + "loss": 0.0805, + "step": 6870 + }, + { + "epoch": 0.45076328375810787, + "grad_norm": 0.4171282649040222, + "learning_rate": 9.99035046550622e-05, + "loss": 0.075, + "step": 6880 + }, + { + "epoch": 0.45141846294961674, + "grad_norm": 0.6770170331001282, + "learning_rate": 9.99024751643393e-05, + "loss": 0.1233, + "step": 6890 + }, + { + "epoch": 0.4520736421411256, + "grad_norm": 0.674313485622406, + "learning_rate": 9.990144021635677e-05, + "loss": 0.0909, + "step": 6900 + }, + { + "epoch": 0.4527288213326345, + "grad_norm": 3.898275852203369, + "learning_rate": 9.990039981122775e-05, + "loss": 0.1209, + "step": 6910 + }, + { + "epoch": 0.45338400052414335, + "grad_norm": 1.4668830633163452, + "learning_rate": 9.989935394906602e-05, + "loss": 0.0942, + "step": 6920 + }, + { + "epoch": 0.4540391797156522, + "grad_norm": 0.2785137891769409, + "learning_rate": 9.989830262998598e-05, + "loss": 0.0876, + "step": 6930 + }, + { + "epoch": 0.4546943589071611, + "grad_norm": 0.8282185792922974, + "learning_rate": 9.989724585410259e-05, + "loss": 0.1162, + "step": 6940 + }, + { + "epoch": 0.45534953809866996, + "grad_norm": 0.49943292140960693, + "learning_rate": 9.989618362153139e-05, + "loss": 0.0786, + "step": 6950 + }, + { + "epoch": 0.4560047172901789, + "grad_norm": 0.8030149340629578, + "learning_rate": 9.989511593238859e-05, + "loss": 0.0807, + "step": 6960 + }, + { + "epoch": 0.45665989648168775, + "grad_norm": 1.2915822267532349, + "learning_rate": 9.98940427867909e-05, + "loss": 0.1009, + "step": 6970 + }, + { + "epoch": 0.4573150756731966, + "grad_norm": 0.8632746934890747, + "learning_rate": 9.989296418485573e-05, + "loss": 0.0775, + "step": 6980 + }, + { + "epoch": 0.4579702548647055, + "grad_norm": 0.5829554796218872, + "learning_rate": 9.989188012670101e-05, + "loss": 0.0763, + "step": 6990 + }, + { + "epoch": 0.45862543405621436, + "grad_norm": 0.9277575612068176, + "learning_rate": 9.989079061244528e-05, + "loss": 0.0873, + "step": 7000 + }, + { + "epoch": 0.45928061324772324, + "grad_norm": 0.8872811198234558, + "learning_rate": 9.988969564220769e-05, + "loss": 0.0826, + "step": 7010 + }, + { + "epoch": 0.4599357924392321, + "grad_norm": 0.4385394752025604, + "learning_rate": 9.988859521610801e-05, + "loss": 0.0887, + "step": 7020 + }, + { + "epoch": 0.46059097163074103, + "grad_norm": 1.6956044435501099, + "learning_rate": 9.988748933426656e-05, + "loss": 0.0824, + "step": 7030 + }, + { + "epoch": 0.4612461508222499, + "grad_norm": 1.2903300523757935, + "learning_rate": 9.988637799680428e-05, + "loss": 0.1269, + "step": 7040 + }, + { + "epoch": 0.46190133001375877, + "grad_norm": 0.8339647054672241, + "learning_rate": 9.98852612038427e-05, + "loss": 0.0792, + "step": 7050 + }, + { + "epoch": 0.46255650920526764, + "grad_norm": 0.5446783304214478, + "learning_rate": 9.988413895550397e-05, + "loss": 0.0586, + "step": 7060 + }, + { + "epoch": 0.4632116883967765, + "grad_norm": 0.62513667345047, + "learning_rate": 9.98830112519108e-05, + "loss": 0.0679, + "step": 7070 + }, + { + "epoch": 0.4638668675882854, + "grad_norm": 0.553907573223114, + "learning_rate": 9.98818780931865e-05, + "loss": 0.0965, + "step": 7080 + }, + { + "epoch": 0.46452204677979425, + "grad_norm": 0.9144310355186462, + "learning_rate": 9.988073947945502e-05, + "loss": 0.127, + "step": 7090 + }, + { + "epoch": 0.4651772259713032, + "grad_norm": 0.6517556309700012, + "learning_rate": 9.987959541084087e-05, + "loss": 0.1046, + "step": 7100 + }, + { + "epoch": 0.46583240516281205, + "grad_norm": 0.694530189037323, + "learning_rate": 9.987844588746915e-05, + "loss": 0.0931, + "step": 7110 + }, + { + "epoch": 0.4664875843543209, + "grad_norm": 1.2795195579528809, + "learning_rate": 9.987729090946558e-05, + "loss": 0.0915, + "step": 7120 + }, + { + "epoch": 0.4671427635458298, + "grad_norm": 0.8495104312896729, + "learning_rate": 9.987613047695647e-05, + "loss": 0.0675, + "step": 7130 + }, + { + "epoch": 0.46779794273733866, + "grad_norm": 1.1382215023040771, + "learning_rate": 9.987496459006871e-05, + "loss": 0.0694, + "step": 7140 + }, + { + "epoch": 0.4684531219288475, + "grad_norm": 0.41733318567276, + "learning_rate": 9.987379324892982e-05, + "loss": 0.0931, + "step": 7150 + }, + { + "epoch": 0.4691083011203564, + "grad_norm": 0.7303067445755005, + "learning_rate": 9.987261645366788e-05, + "loss": 0.1196, + "step": 7160 + }, + { + "epoch": 0.4697634803118653, + "grad_norm": 1.4135780334472656, + "learning_rate": 9.987143420441158e-05, + "loss": 0.1009, + "step": 7170 + }, + { + "epoch": 0.4704186595033742, + "grad_norm": 0.7097006440162659, + "learning_rate": 9.987024650129022e-05, + "loss": 0.1082, + "step": 7180 + }, + { + "epoch": 0.47107383869488306, + "grad_norm": 0.876897394657135, + "learning_rate": 9.986905334443368e-05, + "loss": 0.1073, + "step": 7190 + }, + { + "epoch": 0.47172901788639193, + "grad_norm": 0.3723180294036865, + "learning_rate": 9.986785473397245e-05, + "loss": 0.0883, + "step": 7200 + }, + { + "epoch": 0.4723841970779008, + "grad_norm": 0.8644726276397705, + "learning_rate": 9.98666506700376e-05, + "loss": 0.1182, + "step": 7210 + }, + { + "epoch": 0.4730393762694097, + "grad_norm": 1.4409370422363281, + "learning_rate": 9.986544115276081e-05, + "loss": 0.0707, + "step": 7220 + }, + { + "epoch": 0.47369455546091854, + "grad_norm": 1.0237318277359009, + "learning_rate": 9.986422618227433e-05, + "loss": 0.0789, + "step": 7230 + }, + { + "epoch": 0.4743497346524274, + "grad_norm": 0.5112522840499878, + "learning_rate": 9.986300575871106e-05, + "loss": 0.0577, + "step": 7240 + }, + { + "epoch": 0.47500491384393634, + "grad_norm": 0.6120023727416992, + "learning_rate": 9.986177988220444e-05, + "loss": 0.0639, + "step": 7250 + }, + { + "epoch": 0.4756600930354452, + "grad_norm": 0.9435589909553528, + "learning_rate": 9.986054855288856e-05, + "loss": 0.0709, + "step": 7260 + }, + { + "epoch": 0.4763152722269541, + "grad_norm": 2.4650070667266846, + "learning_rate": 9.985931177089802e-05, + "loss": 0.1465, + "step": 7270 + }, + { + "epoch": 0.47697045141846295, + "grad_norm": 1.3930546045303345, + "learning_rate": 9.985806953636814e-05, + "loss": 0.1026, + "step": 7280 + }, + { + "epoch": 0.4776256306099718, + "grad_norm": 1.162932276725769, + "learning_rate": 9.985682184943471e-05, + "loss": 0.0844, + "step": 7290 + }, + { + "epoch": 0.4782808098014807, + "grad_norm": 0.4651147723197937, + "learning_rate": 9.98555687102342e-05, + "loss": 0.0755, + "step": 7300 + }, + { + "epoch": 0.47893598899298956, + "grad_norm": 0.9652953147888184, + "learning_rate": 9.985431011890367e-05, + "loss": 0.0611, + "step": 7310 + }, + { + "epoch": 0.4795911681844985, + "grad_norm": 0.6520761251449585, + "learning_rate": 9.985304607558075e-05, + "loss": 0.1272, + "step": 7320 + }, + { + "epoch": 0.48024634737600735, + "grad_norm": 1.2580307722091675, + "learning_rate": 9.985177658040364e-05, + "loss": 0.1147, + "step": 7330 + }, + { + "epoch": 0.4809015265675162, + "grad_norm": 0.3914817273616791, + "learning_rate": 9.985050163351119e-05, + "loss": 0.0892, + "step": 7340 + }, + { + "epoch": 0.4815567057590251, + "grad_norm": 0.5663081407546997, + "learning_rate": 9.984922123504286e-05, + "loss": 0.0928, + "step": 7350 + }, + { + "epoch": 0.48221188495053396, + "grad_norm": 0.99325031042099, + "learning_rate": 9.984793538513862e-05, + "loss": 0.0922, + "step": 7360 + }, + { + "epoch": 0.48286706414204283, + "grad_norm": 0.873366117477417, + "learning_rate": 9.984664408393912e-05, + "loss": 0.1067, + "step": 7370 + }, + { + "epoch": 0.4835222433335517, + "grad_norm": 0.8472453355789185, + "learning_rate": 9.984534733158556e-05, + "loss": 0.1185, + "step": 7380 + }, + { + "epoch": 0.48417742252506063, + "grad_norm": 0.8856337666511536, + "learning_rate": 9.984404512821977e-05, + "loss": 0.0616, + "step": 7390 + }, + { + "epoch": 0.4848326017165695, + "grad_norm": 0.5571925044059753, + "learning_rate": 9.984273747398411e-05, + "loss": 0.0817, + "step": 7400 + }, + { + "epoch": 0.48548778090807837, + "grad_norm": 0.4322974681854248, + "learning_rate": 9.984142436902165e-05, + "loss": 0.0958, + "step": 7410 + }, + { + "epoch": 0.48614296009958724, + "grad_norm": 0.9718981981277466, + "learning_rate": 9.984010581347596e-05, + "loss": 0.077, + "step": 7420 + }, + { + "epoch": 0.4867981392910961, + "grad_norm": 0.23679722845554352, + "learning_rate": 9.983878180749121e-05, + "loss": 0.1038, + "step": 7430 + }, + { + "epoch": 0.487453318482605, + "grad_norm": 0.490441232919693, + "learning_rate": 9.983745235121222e-05, + "loss": 0.0809, + "step": 7440 + }, + { + "epoch": 0.48810849767411385, + "grad_norm": 1.6489852666854858, + "learning_rate": 9.983611744478438e-05, + "loss": 0.0569, + "step": 7450 + }, + { + "epoch": 0.4887636768656227, + "grad_norm": 1.1161770820617676, + "learning_rate": 9.983477708835365e-05, + "loss": 0.0895, + "step": 7460 + }, + { + "epoch": 0.48941885605713165, + "grad_norm": 0.709377110004425, + "learning_rate": 9.983343128206664e-05, + "loss": 0.0993, + "step": 7470 + }, + { + "epoch": 0.4900740352486405, + "grad_norm": 1.3923457860946655, + "learning_rate": 9.983208002607049e-05, + "loss": 0.0775, + "step": 7480 + }, + { + "epoch": 0.4907292144401494, + "grad_norm": 0.9146208763122559, + "learning_rate": 9.9830723320513e-05, + "loss": 0.0547, + "step": 7490 + }, + { + "epoch": 0.49138439363165826, + "grad_norm": 0.7479075193405151, + "learning_rate": 9.982936116554254e-05, + "loss": 0.0683, + "step": 7500 + }, + { + "epoch": 0.4920395728231671, + "grad_norm": 0.49663668870925903, + "learning_rate": 9.982799356130803e-05, + "loss": 0.0728, + "step": 7510 + }, + { + "epoch": 0.492694752014676, + "grad_norm": 1.2406212091445923, + "learning_rate": 9.982662050795908e-05, + "loss": 0.1069, + "step": 7520 + }, + { + "epoch": 0.49334993120618487, + "grad_norm": 0.6170276403427124, + "learning_rate": 9.982524200564583e-05, + "loss": 0.0782, + "step": 7530 + }, + { + "epoch": 0.4940051103976938, + "grad_norm": 0.5455721020698547, + "learning_rate": 9.982385805451901e-05, + "loss": 0.0859, + "step": 7540 + }, + { + "epoch": 0.49466028958920266, + "grad_norm": 1.5098469257354736, + "learning_rate": 9.982246865472998e-05, + "loss": 0.0761, + "step": 7550 + }, + { + "epoch": 0.49531546878071153, + "grad_norm": 0.797351598739624, + "learning_rate": 9.982107380643069e-05, + "loss": 0.0629, + "step": 7560 + }, + { + "epoch": 0.4959706479722204, + "grad_norm": 1.073926329612732, + "learning_rate": 9.981967350977368e-05, + "loss": 0.081, + "step": 7570 + }, + { + "epoch": 0.49662582716372927, + "grad_norm": 0.6648547649383545, + "learning_rate": 9.981826776491208e-05, + "loss": 0.0811, + "step": 7580 + }, + { + "epoch": 0.49728100635523814, + "grad_norm": 0.8206454515457153, + "learning_rate": 9.98168565719996e-05, + "loss": 0.0757, + "step": 7590 + }, + { + "epoch": 0.497936185546747, + "grad_norm": 1.4741157293319702, + "learning_rate": 9.98154399311906e-05, + "loss": 0.0727, + "step": 7600 + }, + { + "epoch": 0.49859136473825594, + "grad_norm": 0.3561849296092987, + "learning_rate": 9.981401784263997e-05, + "loss": 0.0881, + "step": 7610 + }, + { + "epoch": 0.4992465439297648, + "grad_norm": 0.42724719643592834, + "learning_rate": 9.981259030650326e-05, + "loss": 0.081, + "step": 7620 + }, + { + "epoch": 0.4999017231212737, + "grad_norm": 1.0347739458084106, + "learning_rate": 9.981115732293655e-05, + "loss": 0.0792, + "step": 7630 + }, + { + "epoch": 0.5005569023127825, + "grad_norm": 1.420472502708435, + "learning_rate": 9.980971889209659e-05, + "loss": 0.107, + "step": 7640 + }, + { + "epoch": 0.5012120815042914, + "grad_norm": 1.7326488494873047, + "learning_rate": 9.980827501414064e-05, + "loss": 0.0819, + "step": 7650 + }, + { + "epoch": 0.5018672606958003, + "grad_norm": 1.2043975591659546, + "learning_rate": 9.980682568922663e-05, + "loss": 0.0846, + "step": 7660 + }, + { + "epoch": 0.5025224398873092, + "grad_norm": 0.6115202903747559, + "learning_rate": 9.980537091751304e-05, + "loss": 0.0898, + "step": 7670 + }, + { + "epoch": 0.5031776190788181, + "grad_norm": 0.7387716174125671, + "learning_rate": 9.980391069915897e-05, + "loss": 0.088, + "step": 7680 + }, + { + "epoch": 0.5038327982703269, + "grad_norm": 0.588518500328064, + "learning_rate": 9.98024450343241e-05, + "loss": 0.0818, + "step": 7690 + }, + { + "epoch": 0.5044879774618358, + "grad_norm": 1.263738989830017, + "learning_rate": 9.980097392316872e-05, + "loss": 0.0895, + "step": 7700 + }, + { + "epoch": 0.5051431566533446, + "grad_norm": 0.7770864963531494, + "learning_rate": 9.97994973658537e-05, + "loss": 0.0643, + "step": 7710 + }, + { + "epoch": 0.5057983358448536, + "grad_norm": 0.7232218980789185, + "learning_rate": 9.979801536254054e-05, + "loss": 0.0658, + "step": 7720 + }, + { + "epoch": 0.5064535150363625, + "grad_norm": 0.921306312084198, + "learning_rate": 9.979652791339127e-05, + "loss": 0.0602, + "step": 7730 + }, + { + "epoch": 0.5071086942278713, + "grad_norm": 0.6976714730262756, + "learning_rate": 9.97950350185686e-05, + "loss": 0.0965, + "step": 7740 + }, + { + "epoch": 0.5077638734193802, + "grad_norm": 0.8672352433204651, + "learning_rate": 9.979353667823574e-05, + "loss": 0.0554, + "step": 7750 + }, + { + "epoch": 0.508419052610889, + "grad_norm": 0.9167852401733398, + "learning_rate": 9.979203289255658e-05, + "loss": 0.1163, + "step": 7760 + }, + { + "epoch": 0.509074231802398, + "grad_norm": 0.9916808605194092, + "learning_rate": 9.979052366169557e-05, + "loss": 0.0954, + "step": 7770 + }, + { + "epoch": 0.5097294109939068, + "grad_norm": 0.7340328693389893, + "learning_rate": 9.978900898581775e-05, + "loss": 0.0639, + "step": 7780 + }, + { + "epoch": 0.5103845901854157, + "grad_norm": 1.0508185625076294, + "learning_rate": 9.978748886508875e-05, + "loss": 0.0579, + "step": 7790 + }, + { + "epoch": 0.5110397693769246, + "grad_norm": 0.4471210539340973, + "learning_rate": 9.978596329967484e-05, + "loss": 0.0706, + "step": 7800 + }, + { + "epoch": 0.5116949485684334, + "grad_norm": 0.8721002340316772, + "learning_rate": 9.978443228974284e-05, + "loss": 0.0993, + "step": 7810 + }, + { + "epoch": 0.5123501277599424, + "grad_norm": 0.883975088596344, + "learning_rate": 9.978289583546015e-05, + "loss": 0.0567, + "step": 7820 + }, + { + "epoch": 0.5130053069514512, + "grad_norm": 1.0761932134628296, + "learning_rate": 9.978135393699484e-05, + "loss": 0.1074, + "step": 7830 + }, + { + "epoch": 0.5136604861429601, + "grad_norm": 0.32462847232818604, + "learning_rate": 9.977980659451548e-05, + "loss": 0.0825, + "step": 7840 + }, + { + "epoch": 0.5143156653344689, + "grad_norm": 1.292065978050232, + "learning_rate": 9.977825380819135e-05, + "loss": 0.0682, + "step": 7850 + }, + { + "epoch": 0.5149708445259779, + "grad_norm": 0.7397595643997192, + "learning_rate": 9.97766955781922e-05, + "loss": 0.0546, + "step": 7860 + }, + { + "epoch": 0.5156260237174868, + "grad_norm": 0.37062907218933105, + "learning_rate": 9.977513190468848e-05, + "loss": 0.0562, + "step": 7870 + }, + { + "epoch": 0.5162812029089956, + "grad_norm": 1.3367611169815063, + "learning_rate": 9.977356278785116e-05, + "loss": 0.0741, + "step": 7880 + }, + { + "epoch": 0.5169363821005045, + "grad_norm": 0.7810423970222473, + "learning_rate": 9.977198822785184e-05, + "loss": 0.1228, + "step": 7890 + }, + { + "epoch": 0.5175915612920133, + "grad_norm": 0.6126615405082703, + "learning_rate": 9.977040822486273e-05, + "loss": 0.0477, + "step": 7900 + }, + { + "epoch": 0.5182467404835223, + "grad_norm": 1.0230940580368042, + "learning_rate": 9.97688227790566e-05, + "loss": 0.0856, + "step": 7910 + }, + { + "epoch": 0.5189019196750311, + "grad_norm": 0.8766168355941772, + "learning_rate": 9.976723189060684e-05, + "loss": 0.1079, + "step": 7920 + }, + { + "epoch": 0.51955709886654, + "grad_norm": 0.813398540019989, + "learning_rate": 9.976563555968742e-05, + "loss": 0.0815, + "step": 7930 + }, + { + "epoch": 0.5202122780580489, + "grad_norm": 0.32452312111854553, + "learning_rate": 9.976403378647292e-05, + "loss": 0.109, + "step": 7940 + }, + { + "epoch": 0.5208674572495577, + "grad_norm": 1.0478284358978271, + "learning_rate": 9.97624265711385e-05, + "loss": 0.1029, + "step": 7950 + }, + { + "epoch": 0.5215226364410667, + "grad_norm": 0.7821952700614929, + "learning_rate": 9.976081391385993e-05, + "loss": 0.0456, + "step": 7960 + }, + { + "epoch": 0.5221778156325755, + "grad_norm": 0.7724446058273315, + "learning_rate": 9.975919581481356e-05, + "loss": 0.0712, + "step": 7970 + }, + { + "epoch": 0.5228329948240844, + "grad_norm": 0.7962720990180969, + "learning_rate": 9.975757227417634e-05, + "loss": 0.1413, + "step": 7980 + }, + { + "epoch": 0.5234881740155932, + "grad_norm": 1.0645952224731445, + "learning_rate": 9.975594329212586e-05, + "loss": 0.0803, + "step": 7990 + }, + { + "epoch": 0.5241433532071021, + "grad_norm": 0.4187980592250824, + "learning_rate": 9.97543088688402e-05, + "loss": 0.0625, + "step": 8000 + }, + { + "epoch": 0.5247985323986111, + "grad_norm": 0.560476541519165, + "learning_rate": 9.975266900449814e-05, + "loss": 0.084, + "step": 8010 + }, + { + "epoch": 0.5254537115901199, + "grad_norm": 0.6050432324409485, + "learning_rate": 9.975102369927898e-05, + "loss": 0.0624, + "step": 8020 + }, + { + "epoch": 0.5261088907816288, + "grad_norm": 0.6608274579048157, + "learning_rate": 9.974937295336269e-05, + "loss": 0.095, + "step": 8030 + }, + { + "epoch": 0.5267640699731376, + "grad_norm": 0.46296927332878113, + "learning_rate": 9.974771676692975e-05, + "loss": 0.0991, + "step": 8040 + }, + { + "epoch": 0.5274192491646466, + "grad_norm": 0.8809294700622559, + "learning_rate": 9.974605514016131e-05, + "loss": 0.0727, + "step": 8050 + }, + { + "epoch": 0.5280744283561554, + "grad_norm": 1.5689514875411987, + "learning_rate": 9.974438807323907e-05, + "loss": 0.062, + "step": 8060 + }, + { + "epoch": 0.5287296075476643, + "grad_norm": 0.23065902292728424, + "learning_rate": 9.974271556634535e-05, + "loss": 0.0829, + "step": 8070 + }, + { + "epoch": 0.5293847867391732, + "grad_norm": 0.8139961957931519, + "learning_rate": 9.974103761966302e-05, + "loss": 0.0611, + "step": 8080 + }, + { + "epoch": 0.530039965930682, + "grad_norm": 0.47768574953079224, + "learning_rate": 9.973935423337563e-05, + "loss": 0.0428, + "step": 8090 + }, + { + "epoch": 0.530695145122191, + "grad_norm": 0.7580574154853821, + "learning_rate": 9.973766540766722e-05, + "loss": 0.0458, + "step": 8100 + }, + { + "epoch": 0.5313503243136998, + "grad_norm": 0.7710840106010437, + "learning_rate": 9.97359711427225e-05, + "loss": 0.0696, + "step": 8110 + }, + { + "epoch": 0.5320055035052087, + "grad_norm": 0.6634746193885803, + "learning_rate": 9.973427143872677e-05, + "loss": 0.0821, + "step": 8120 + }, + { + "epoch": 0.5326606826967175, + "grad_norm": 0.6555359959602356, + "learning_rate": 9.973256629586589e-05, + "loss": 0.0594, + "step": 8130 + }, + { + "epoch": 0.5333158618882264, + "grad_norm": 0.9500135779380798, + "learning_rate": 9.973085571432632e-05, + "loss": 0.0826, + "step": 8140 + }, + { + "epoch": 0.5339710410797353, + "grad_norm": 1.186127781867981, + "learning_rate": 9.972913969429513e-05, + "loss": 0.1042, + "step": 8150 + }, + { + "epoch": 0.5346262202712442, + "grad_norm": 0.29740777611732483, + "learning_rate": 9.972741823596e-05, + "loss": 0.0847, + "step": 8160 + }, + { + "epoch": 0.5352813994627531, + "grad_norm": 0.7914127707481384, + "learning_rate": 9.972569133950917e-05, + "loss": 0.0784, + "step": 8170 + }, + { + "epoch": 0.5359365786542619, + "grad_norm": 0.6166767477989197, + "learning_rate": 9.972395900513151e-05, + "loss": 0.0786, + "step": 8180 + }, + { + "epoch": 0.5365917578457708, + "grad_norm": 0.287585973739624, + "learning_rate": 9.972222123301645e-05, + "loss": 0.0701, + "step": 8190 + }, + { + "epoch": 0.5372469370372797, + "grad_norm": 0.5762342214584351, + "learning_rate": 9.972047802335403e-05, + "loss": 0.044, + "step": 8200 + }, + { + "epoch": 0.5379021162287886, + "grad_norm": 0.6411095857620239, + "learning_rate": 9.971872937633488e-05, + "loss": 0.0711, + "step": 8210 + }, + { + "epoch": 0.5385572954202974, + "grad_norm": 1.2706223726272583, + "learning_rate": 9.971697529215024e-05, + "loss": 0.1024, + "step": 8220 + }, + { + "epoch": 0.5392124746118063, + "grad_norm": 0.48651406168937683, + "learning_rate": 9.971521577099192e-05, + "loss": 0.0687, + "step": 8230 + }, + { + "epoch": 0.5398676538033153, + "grad_norm": 0.5664197206497192, + "learning_rate": 9.971345081305236e-05, + "loss": 0.0879, + "step": 8240 + }, + { + "epoch": 0.5405228329948241, + "grad_norm": 0.4421285390853882, + "learning_rate": 9.971168041852456e-05, + "loss": 0.0481, + "step": 8250 + }, + { + "epoch": 0.541178012186333, + "grad_norm": 0.895587682723999, + "learning_rate": 9.970990458760215e-05, + "loss": 0.061, + "step": 8260 + }, + { + "epoch": 0.5418331913778418, + "grad_norm": 0.6101402044296265, + "learning_rate": 9.970812332047929e-05, + "loss": 0.0988, + "step": 8270 + }, + { + "epoch": 0.5424883705693507, + "grad_norm": 0.7041201591491699, + "learning_rate": 9.97063366173508e-05, + "loss": 0.0415, + "step": 8280 + }, + { + "epoch": 0.5431435497608595, + "grad_norm": 1.1475251913070679, + "learning_rate": 9.970454447841207e-05, + "loss": 0.0807, + "step": 8290 + }, + { + "epoch": 0.5437987289523685, + "grad_norm": 0.8536934852600098, + "learning_rate": 9.970274690385909e-05, + "loss": 0.0907, + "step": 8300 + }, + { + "epoch": 0.5444539081438774, + "grad_norm": 1.2763111591339111, + "learning_rate": 9.970094389388844e-05, + "loss": 0.1085, + "step": 8310 + }, + { + "epoch": 0.5451090873353862, + "grad_norm": 1.5681908130645752, + "learning_rate": 9.969913544869728e-05, + "loss": 0.0985, + "step": 8320 + }, + { + "epoch": 0.5457642665268951, + "grad_norm": 1.5142990350723267, + "learning_rate": 9.96973215684834e-05, + "loss": 0.0681, + "step": 8330 + }, + { + "epoch": 0.546419445718404, + "grad_norm": 0.4772714078426361, + "learning_rate": 9.969550225344513e-05, + "loss": 0.0743, + "step": 8340 + }, + { + "epoch": 0.5470746249099129, + "grad_norm": 0.6852735280990601, + "learning_rate": 9.969367750378147e-05, + "loss": 0.0748, + "step": 8350 + }, + { + "epoch": 0.5477298041014217, + "grad_norm": 1.0914340019226074, + "learning_rate": 9.969184731969194e-05, + "loss": 0.0819, + "step": 8360 + }, + { + "epoch": 0.5483849832929306, + "grad_norm": 0.9785975217819214, + "learning_rate": 9.96900117013767e-05, + "loss": 0.0586, + "step": 8370 + }, + { + "epoch": 0.5490401624844395, + "grad_norm": 0.6585258841514587, + "learning_rate": 9.96881706490365e-05, + "loss": 0.0795, + "step": 8380 + }, + { + "epoch": 0.5496953416759484, + "grad_norm": 1.1437153816223145, + "learning_rate": 9.968632416287265e-05, + "loss": 0.0808, + "step": 8390 + }, + { + "epoch": 0.5503505208674573, + "grad_norm": 1.9844082593917847, + "learning_rate": 9.96844722430871e-05, + "loss": 0.0673, + "step": 8400 + }, + { + "epoch": 0.5510057000589661, + "grad_norm": 0.6028240323066711, + "learning_rate": 9.968261488988235e-05, + "loss": 0.0659, + "step": 8410 + }, + { + "epoch": 0.551660879250475, + "grad_norm": 0.5887708067893982, + "learning_rate": 9.968075210346155e-05, + "loss": 0.0838, + "step": 8420 + }, + { + "epoch": 0.5523160584419838, + "grad_norm": 0.6402873396873474, + "learning_rate": 9.967888388402839e-05, + "loss": 0.1004, + "step": 8430 + }, + { + "epoch": 0.5529712376334928, + "grad_norm": 1.1275421380996704, + "learning_rate": 9.967701023178717e-05, + "loss": 0.0699, + "step": 8440 + }, + { + "epoch": 0.5536264168250017, + "grad_norm": 1.324973702430725, + "learning_rate": 9.967513114694282e-05, + "loss": 0.0919, + "step": 8450 + }, + { + "epoch": 0.5542815960165105, + "grad_norm": 1.079539179801941, + "learning_rate": 9.967324662970079e-05, + "loss": 0.07, + "step": 8460 + }, + { + "epoch": 0.5549367752080194, + "grad_norm": 0.880542516708374, + "learning_rate": 9.96713566802672e-05, + "loss": 0.0744, + "step": 8470 + }, + { + "epoch": 0.5555919543995282, + "grad_norm": 1.0542476177215576, + "learning_rate": 9.966946129884873e-05, + "loss": 0.0688, + "step": 8480 + }, + { + "epoch": 0.5562471335910372, + "grad_norm": 0.8128005862236023, + "learning_rate": 9.966756048565265e-05, + "loss": 0.048, + "step": 8490 + }, + { + "epoch": 0.556902312782546, + "grad_norm": 0.5629258155822754, + "learning_rate": 9.966565424088681e-05, + "loss": 0.0929, + "step": 8500 + }, + { + "epoch": 0.5575574919740549, + "grad_norm": 0.81717449426651, + "learning_rate": 9.96637425647597e-05, + "loss": 0.0556, + "step": 8510 + }, + { + "epoch": 0.5582126711655638, + "grad_norm": 0.6798516511917114, + "learning_rate": 9.966182545748038e-05, + "loss": 0.0562, + "step": 8520 + }, + { + "epoch": 0.5588678503570726, + "grad_norm": 0.7413844466209412, + "learning_rate": 9.96599029192585e-05, + "loss": 0.1064, + "step": 8530 + }, + { + "epoch": 0.5595230295485816, + "grad_norm": 0.7336576581001282, + "learning_rate": 9.965797495030428e-05, + "loss": 0.0631, + "step": 8540 + }, + { + "epoch": 0.5601782087400904, + "grad_norm": 0.7582850456237793, + "learning_rate": 9.96560415508286e-05, + "loss": 0.0449, + "step": 8550 + }, + { + "epoch": 0.5608333879315993, + "grad_norm": 0.3353594243526459, + "learning_rate": 9.965410272104286e-05, + "loss": 0.0432, + "step": 8560 + }, + { + "epoch": 0.5614885671231081, + "grad_norm": 0.5461919903755188, + "learning_rate": 9.96521584611591e-05, + "loss": 0.0615, + "step": 8570 + }, + { + "epoch": 0.562143746314617, + "grad_norm": 1.7283936738967896, + "learning_rate": 9.965020877138994e-05, + "loss": 0.0871, + "step": 8580 + }, + { + "epoch": 0.562798925506126, + "grad_norm": 0.963506817817688, + "learning_rate": 9.964825365194861e-05, + "loss": 0.0617, + "step": 8590 + }, + { + "epoch": 0.5634541046976348, + "grad_norm": 1.8822720050811768, + "learning_rate": 9.96462931030489e-05, + "loss": 0.0757, + "step": 8600 + }, + { + "epoch": 0.5641092838891437, + "grad_norm": 0.35969164967536926, + "learning_rate": 9.96443271249052e-05, + "loss": 0.0609, + "step": 8610 + }, + { + "epoch": 0.5647644630806525, + "grad_norm": 1.2140121459960938, + "learning_rate": 9.964235571773255e-05, + "loss": 0.0623, + "step": 8620 + }, + { + "epoch": 0.5654196422721615, + "grad_norm": 1.0485846996307373, + "learning_rate": 9.96403788817465e-05, + "loss": 0.0714, + "step": 8630 + }, + { + "epoch": 0.5660748214636703, + "grad_norm": 0.9979734420776367, + "learning_rate": 9.963839661716325e-05, + "loss": 0.0771, + "step": 8640 + }, + { + "epoch": 0.5667300006551792, + "grad_norm": 1.6107581853866577, + "learning_rate": 9.963640892419958e-05, + "loss": 0.0793, + "step": 8650 + }, + { + "epoch": 0.567385179846688, + "grad_norm": 0.6750025153160095, + "learning_rate": 9.963441580307286e-05, + "loss": 0.0847, + "step": 8660 + }, + { + "epoch": 0.5680403590381969, + "grad_norm": 0.6074758768081665, + "learning_rate": 9.963241725400104e-05, + "loss": 0.0474, + "step": 8670 + }, + { + "epoch": 0.5686955382297059, + "grad_norm": 1.5094566345214844, + "learning_rate": 9.963041327720271e-05, + "loss": 0.0472, + "step": 8680 + }, + { + "epoch": 0.5693507174212147, + "grad_norm": 0.7529876232147217, + "learning_rate": 9.962840387289697e-05, + "loss": 0.0801, + "step": 8690 + }, + { + "epoch": 0.5700058966127236, + "grad_norm": 1.1150997877120972, + "learning_rate": 9.962638904130363e-05, + "loss": 0.0841, + "step": 8700 + }, + { + "epoch": 0.5706610758042324, + "grad_norm": 0.5043185353279114, + "learning_rate": 9.962436878264298e-05, + "loss": 0.0924, + "step": 8710 + }, + { + "epoch": 0.5713162549957413, + "grad_norm": 1.5951225757598877, + "learning_rate": 9.962234309713598e-05, + "loss": 0.0767, + "step": 8720 + }, + { + "epoch": 0.5719714341872502, + "grad_norm": 0.5490920543670654, + "learning_rate": 9.962031198500414e-05, + "loss": 0.0611, + "step": 8730 + }, + { + "epoch": 0.5726266133787591, + "grad_norm": 2.3193471431732178, + "learning_rate": 9.961827544646958e-05, + "loss": 0.0618, + "step": 8740 + }, + { + "epoch": 0.573281792570268, + "grad_norm": 0.6018168926239014, + "learning_rate": 9.961623348175501e-05, + "loss": 0.0393, + "step": 8750 + }, + { + "epoch": 0.5739369717617768, + "grad_norm": 0.5578179359436035, + "learning_rate": 9.961418609108377e-05, + "loss": 0.0833, + "step": 8760 + }, + { + "epoch": 0.5745921509532858, + "grad_norm": 0.9199391603469849, + "learning_rate": 9.961213327467971e-05, + "loss": 0.0837, + "step": 8770 + }, + { + "epoch": 0.5752473301447946, + "grad_norm": 0.8956995606422424, + "learning_rate": 9.961007503276736e-05, + "loss": 0.0679, + "step": 8780 + }, + { + "epoch": 0.5759025093363035, + "grad_norm": 0.5426045656204224, + "learning_rate": 9.960801136557179e-05, + "loss": 0.0667, + "step": 8790 + }, + { + "epoch": 0.5765576885278123, + "grad_norm": 0.467263400554657, + "learning_rate": 9.960594227331866e-05, + "loss": 0.0492, + "step": 8800 + }, + { + "epoch": 0.5772128677193212, + "grad_norm": 0.3937716484069824, + "learning_rate": 9.960386775623429e-05, + "loss": 0.0515, + "step": 8810 + }, + { + "epoch": 0.5778680469108302, + "grad_norm": 0.8585725426673889, + "learning_rate": 9.96017878145455e-05, + "loss": 0.0664, + "step": 8820 + }, + { + "epoch": 0.578523226102339, + "grad_norm": 0.3727948069572449, + "learning_rate": 9.959970244847977e-05, + "loss": 0.0339, + "step": 8830 + }, + { + "epoch": 0.5791784052938479, + "grad_norm": 0.7736207246780396, + "learning_rate": 9.959761165826518e-05, + "loss": 0.053, + "step": 8840 + }, + { + "epoch": 0.5798335844853567, + "grad_norm": 0.6750363707542419, + "learning_rate": 9.959551544413033e-05, + "loss": 0.0687, + "step": 8850 + }, + { + "epoch": 0.5804887636768656, + "grad_norm": 0.9430448412895203, + "learning_rate": 9.959341380630448e-05, + "loss": 0.06, + "step": 8860 + }, + { + "epoch": 0.5811439428683745, + "grad_norm": 0.77512127161026, + "learning_rate": 9.959130674501746e-05, + "loss": 0.0704, + "step": 8870 + }, + { + "epoch": 0.5817991220598834, + "grad_norm": 0.6806347370147705, + "learning_rate": 9.958919426049968e-05, + "loss": 0.0778, + "step": 8880 + }, + { + "epoch": 0.5824543012513923, + "grad_norm": 0.6905287504196167, + "learning_rate": 9.958707635298219e-05, + "loss": 0.0525, + "step": 8890 + }, + { + "epoch": 0.5831094804429011, + "grad_norm": 0.4164761006832123, + "learning_rate": 9.958495302269657e-05, + "loss": 0.0818, + "step": 8900 + }, + { + "epoch": 0.58376465963441, + "grad_norm": 0.3138507902622223, + "learning_rate": 9.958282426987503e-05, + "loss": 0.0996, + "step": 8910 + }, + { + "epoch": 0.5844198388259189, + "grad_norm": 0.7195194959640503, + "learning_rate": 9.95806900947504e-05, + "loss": 0.0714, + "step": 8920 + }, + { + "epoch": 0.5850750180174278, + "grad_norm": 0.8047299385070801, + "learning_rate": 9.957855049755604e-05, + "loss": 0.0562, + "step": 8930 + }, + { + "epoch": 0.5857301972089366, + "grad_norm": 0.9355923533439636, + "learning_rate": 9.957640547852593e-05, + "loss": 0.0543, + "step": 8940 + }, + { + "epoch": 0.5863853764004455, + "grad_norm": 0.5526629090309143, + "learning_rate": 9.957425503789466e-05, + "loss": 0.0546, + "step": 8950 + }, + { + "epoch": 0.5870405555919544, + "grad_norm": 0.8224440217018127, + "learning_rate": 9.957209917589738e-05, + "loss": 0.0713, + "step": 8960 + }, + { + "epoch": 0.5876957347834633, + "grad_norm": 0.45570823550224304, + "learning_rate": 9.956993789276987e-05, + "loss": 0.0765, + "step": 8970 + }, + { + "epoch": 0.5883509139749722, + "grad_norm": 1.0493866205215454, + "learning_rate": 9.956777118874847e-05, + "loss": 0.0614, + "step": 8980 + }, + { + "epoch": 0.589006093166481, + "grad_norm": 0.5854618549346924, + "learning_rate": 9.956559906407016e-05, + "loss": 0.1218, + "step": 8990 + }, + { + "epoch": 0.5896612723579899, + "grad_norm": 0.7330891489982605, + "learning_rate": 9.956342151897245e-05, + "loss": 0.0511, + "step": 9000 + }, + { + "epoch": 0.5903164515494987, + "grad_norm": 0.9613815546035767, + "learning_rate": 9.956123855369346e-05, + "loss": 0.0708, + "step": 9010 + }, + { + "epoch": 0.5909716307410077, + "grad_norm": 0.8503867983818054, + "learning_rate": 9.955905016847196e-05, + "loss": 0.0698, + "step": 9020 + }, + { + "epoch": 0.5916268099325166, + "grad_norm": 1.5952963829040527, + "learning_rate": 9.955685636354723e-05, + "loss": 0.0782, + "step": 9030 + }, + { + "epoch": 0.5922819891240254, + "grad_norm": 0.45266225934028625, + "learning_rate": 9.95546571391592e-05, + "loss": 0.1039, + "step": 9040 + }, + { + "epoch": 0.5929371683155343, + "grad_norm": 1.8827322721481323, + "learning_rate": 9.955245249554837e-05, + "loss": 0.0745, + "step": 9050 + }, + { + "epoch": 0.5935923475070431, + "grad_norm": 1.0570570230484009, + "learning_rate": 9.955024243295582e-05, + "loss": 0.0753, + "step": 9060 + }, + { + "epoch": 0.5942475266985521, + "grad_norm": 0.670112133026123, + "learning_rate": 9.954802695162328e-05, + "loss": 0.0886, + "step": 9070 + }, + { + "epoch": 0.5949027058900609, + "grad_norm": 0.9260128736495972, + "learning_rate": 9.954580605179302e-05, + "loss": 0.0738, + "step": 9080 + }, + { + "epoch": 0.5955578850815698, + "grad_norm": 0.7489612698554993, + "learning_rate": 9.954357973370788e-05, + "loss": 0.0743, + "step": 9090 + }, + { + "epoch": 0.5962130642730787, + "grad_norm": 0.6623049378395081, + "learning_rate": 9.954134799761135e-05, + "loss": 0.0622, + "step": 9100 + }, + { + "epoch": 0.5968682434645876, + "grad_norm": 0.3814474940299988, + "learning_rate": 9.953911084374748e-05, + "loss": 0.0565, + "step": 9110 + }, + { + "epoch": 0.5975234226560965, + "grad_norm": 0.9237895011901855, + "learning_rate": 9.953686827236093e-05, + "loss": 0.0854, + "step": 9120 + }, + { + "epoch": 0.5981786018476053, + "grad_norm": 1.1899951696395874, + "learning_rate": 9.953462028369695e-05, + "loss": 0.1002, + "step": 9130 + }, + { + "epoch": 0.5988337810391142, + "grad_norm": 0.8747096657752991, + "learning_rate": 9.953236687800136e-05, + "loss": 0.0935, + "step": 9140 + }, + { + "epoch": 0.599488960230623, + "grad_norm": 0.5959799885749817, + "learning_rate": 9.95301080555206e-05, + "loss": 0.0623, + "step": 9150 + }, + { + "epoch": 0.600144139422132, + "grad_norm": 0.5984750986099243, + "learning_rate": 9.952784381650171e-05, + "loss": 0.1221, + "step": 9160 + }, + { + "epoch": 0.6007993186136408, + "grad_norm": 0.8432673215866089, + "learning_rate": 9.952557416119226e-05, + "loss": 0.056, + "step": 9170 + }, + { + "epoch": 0.6014544978051497, + "grad_norm": 0.9011417627334595, + "learning_rate": 9.95232990898405e-05, + "loss": 0.0477, + "step": 9180 + }, + { + "epoch": 0.6021096769966586, + "grad_norm": 0.9434717893600464, + "learning_rate": 9.95210186026952e-05, + "loss": 0.0701, + "step": 9190 + }, + { + "epoch": 0.6027648561881674, + "grad_norm": 0.23065930604934692, + "learning_rate": 9.951873270000576e-05, + "loss": 0.0495, + "step": 9200 + }, + { + "epoch": 0.6034200353796764, + "grad_norm": 0.9471765160560608, + "learning_rate": 9.951644138202216e-05, + "loss": 0.064, + "step": 9210 + }, + { + "epoch": 0.6040752145711852, + "grad_norm": 0.6939331889152527, + "learning_rate": 9.951414464899498e-05, + "loss": 0.0742, + "step": 9220 + }, + { + "epoch": 0.6047303937626941, + "grad_norm": 0.653128445148468, + "learning_rate": 9.951184250117538e-05, + "loss": 0.1055, + "step": 9230 + }, + { + "epoch": 0.6053855729542029, + "grad_norm": 0.5352967977523804, + "learning_rate": 9.950953493881513e-05, + "loss": 0.0758, + "step": 9240 + }, + { + "epoch": 0.6060407521457118, + "grad_norm": 0.7601375579833984, + "learning_rate": 9.950722196216658e-05, + "loss": 0.0544, + "step": 9250 + }, + { + "epoch": 0.6066959313372208, + "grad_norm": 1.274409532546997, + "learning_rate": 9.950490357148265e-05, + "loss": 0.0377, + "step": 9260 + }, + { + "epoch": 0.6073511105287296, + "grad_norm": 0.6958012580871582, + "learning_rate": 9.950257976701692e-05, + "loss": 0.0479, + "step": 9270 + }, + { + "epoch": 0.6080062897202385, + "grad_norm": 1.2639878988265991, + "learning_rate": 9.950025054902348e-05, + "loss": 0.0564, + "step": 9280 + }, + { + "epoch": 0.6086614689117473, + "grad_norm": 0.426265686750412, + "learning_rate": 9.949791591775706e-05, + "loss": 0.0444, + "step": 9290 + }, + { + "epoch": 0.6093166481032563, + "grad_norm": 1.817304015159607, + "learning_rate": 9.949557587347298e-05, + "loss": 0.115, + "step": 9300 + }, + { + "epoch": 0.6099718272947651, + "grad_norm": 0.31961899995803833, + "learning_rate": 9.949323041642713e-05, + "loss": 0.0773, + "step": 9310 + }, + { + "epoch": 0.610627006486274, + "grad_norm": 1.3357486724853516, + "learning_rate": 9.949087954687602e-05, + "loss": 0.0954, + "step": 9320 + }, + { + "epoch": 0.6112821856777829, + "grad_norm": 0.6790017485618591, + "learning_rate": 9.948852326507672e-05, + "loss": 0.0642, + "step": 9330 + }, + { + "epoch": 0.6119373648692917, + "grad_norm": 2.0793685913085938, + "learning_rate": 9.948616157128694e-05, + "loss": 0.0933, + "step": 9340 + }, + { + "epoch": 0.6125925440608007, + "grad_norm": 0.5191910862922668, + "learning_rate": 9.948379446576493e-05, + "loss": 0.0986, + "step": 9350 + }, + { + "epoch": 0.6132477232523095, + "grad_norm": 1.4682646989822388, + "learning_rate": 9.948142194876952e-05, + "loss": 0.0944, + "step": 9360 + }, + { + "epoch": 0.6139029024438184, + "grad_norm": 0.7524102926254272, + "learning_rate": 9.947904402056024e-05, + "loss": 0.0852, + "step": 9370 + }, + { + "epoch": 0.6145580816353272, + "grad_norm": 0.6552098393440247, + "learning_rate": 9.947666068139708e-05, + "loss": 0.0784, + "step": 9380 + }, + { + "epoch": 0.6152132608268361, + "grad_norm": 0.7321749925613403, + "learning_rate": 9.947427193154071e-05, + "loss": 0.1084, + "step": 9390 + }, + { + "epoch": 0.6158684400183451, + "grad_norm": 0.8988286852836609, + "learning_rate": 9.947187777125233e-05, + "loss": 0.0699, + "step": 9400 + }, + { + "epoch": 0.6165236192098539, + "grad_norm": 0.6470881700515747, + "learning_rate": 9.946947820079377e-05, + "loss": 0.0888, + "step": 9410 + }, + { + "epoch": 0.6171787984013628, + "grad_norm": 0.9129390120506287, + "learning_rate": 9.946707322042747e-05, + "loss": 0.0693, + "step": 9420 + }, + { + "epoch": 0.6178339775928716, + "grad_norm": 0.8282991647720337, + "learning_rate": 9.94646628304164e-05, + "loss": 0.0696, + "step": 9430 + }, + { + "epoch": 0.6184891567843805, + "grad_norm": 0.7054755687713623, + "learning_rate": 9.946224703102418e-05, + "loss": 0.0891, + "step": 9440 + }, + { + "epoch": 0.6191443359758894, + "grad_norm": 0.22279804944992065, + "learning_rate": 9.945982582251498e-05, + "loss": 0.0601, + "step": 9450 + }, + { + "epoch": 0.6197995151673983, + "grad_norm": 0.8886716365814209, + "learning_rate": 9.94573992051536e-05, + "loss": 0.1096, + "step": 9460 + }, + { + "epoch": 0.6204546943589072, + "grad_norm": 0.6012411713600159, + "learning_rate": 9.94549671792054e-05, + "loss": 0.0547, + "step": 9470 + }, + { + "epoch": 0.621109873550416, + "grad_norm": 0.772514283657074, + "learning_rate": 9.945252974493635e-05, + "loss": 0.0373, + "step": 9480 + }, + { + "epoch": 0.621765052741925, + "grad_norm": 0.5370466113090515, + "learning_rate": 9.9450086902613e-05, + "loss": 0.0981, + "step": 9490 + }, + { + "epoch": 0.6224202319334338, + "grad_norm": 0.6634767055511475, + "learning_rate": 9.944763865250248e-05, + "loss": 0.0651, + "step": 9500 + }, + { + "epoch": 0.6230754111249427, + "grad_norm": 0.1662437617778778, + "learning_rate": 9.944518499487254e-05, + "loss": 0.1028, + "step": 9510 + }, + { + "epoch": 0.6237305903164515, + "grad_norm": 0.4161169230937958, + "learning_rate": 9.944272592999151e-05, + "loss": 0.0526, + "step": 9520 + }, + { + "epoch": 0.6243857695079604, + "grad_norm": 0.560184895992279, + "learning_rate": 9.94402614581283e-05, + "loss": 0.0885, + "step": 9530 + }, + { + "epoch": 0.6250409486994694, + "grad_norm": 0.8875340223312378, + "learning_rate": 9.943779157955244e-05, + "loss": 0.0683, + "step": 9540 + }, + { + "epoch": 0.6256961278909782, + "grad_norm": 1.9632666110992432, + "learning_rate": 9.943531629453403e-05, + "loss": 0.0902, + "step": 9550 + }, + { + "epoch": 0.6263513070824871, + "grad_norm": 0.7780519127845764, + "learning_rate": 9.943283560334375e-05, + "loss": 0.1143, + "step": 9560 + }, + { + "epoch": 0.6270064862739959, + "grad_norm": 1.1248900890350342, + "learning_rate": 9.943034950625288e-05, + "loss": 0.0554, + "step": 9570 + }, + { + "epoch": 0.6276616654655048, + "grad_norm": 0.7456064224243164, + "learning_rate": 9.942785800353332e-05, + "loss": 0.0544, + "step": 9580 + }, + { + "epoch": 0.6283168446570137, + "grad_norm": 0.7202377319335938, + "learning_rate": 9.942536109545751e-05, + "loss": 0.0699, + "step": 9590 + }, + { + "epoch": 0.6289720238485226, + "grad_norm": 0.30385929346084595, + "learning_rate": 9.942285878229853e-05, + "loss": 0.0676, + "step": 9600 + }, + { + "epoch": 0.6296272030400315, + "grad_norm": 0.43793371319770813, + "learning_rate": 9.942035106433001e-05, + "loss": 0.0687, + "step": 9610 + }, + { + "epoch": 0.6302823822315403, + "grad_norm": 0.23927530646324158, + "learning_rate": 9.94178379418262e-05, + "loss": 0.0809, + "step": 9620 + }, + { + "epoch": 0.6309375614230492, + "grad_norm": 0.37316566705703735, + "learning_rate": 9.941531941506194e-05, + "loss": 0.0971, + "step": 9630 + }, + { + "epoch": 0.6315927406145581, + "grad_norm": 0.6713839769363403, + "learning_rate": 9.941279548431263e-05, + "loss": 0.0712, + "step": 9640 + }, + { + "epoch": 0.632247919806067, + "grad_norm": 1.1073435544967651, + "learning_rate": 9.941026614985431e-05, + "loss": 0.0913, + "step": 9650 + }, + { + "epoch": 0.6329030989975758, + "grad_norm": 0.9665677547454834, + "learning_rate": 9.940773141196357e-05, + "loss": 0.0911, + "step": 9660 + }, + { + "epoch": 0.6335582781890847, + "grad_norm": 0.8016369342803955, + "learning_rate": 9.94051912709176e-05, + "loss": 0.0735, + "step": 9670 + }, + { + "epoch": 0.6342134573805935, + "grad_norm": 0.6325954794883728, + "learning_rate": 9.940264572699421e-05, + "loss": 0.0666, + "step": 9680 + }, + { + "epoch": 0.6348686365721025, + "grad_norm": 0.8656646013259888, + "learning_rate": 9.940009478047174e-05, + "loss": 0.055, + "step": 9690 + }, + { + "epoch": 0.6355238157636114, + "grad_norm": 0.9206833243370056, + "learning_rate": 9.939753843162918e-05, + "loss": 0.0635, + "step": 9700 + }, + { + "epoch": 0.6361789949551202, + "grad_norm": 1.2021689414978027, + "learning_rate": 9.939497668074609e-05, + "loss": 0.0848, + "step": 9710 + }, + { + "epoch": 0.6368341741466291, + "grad_norm": 0.585694432258606, + "learning_rate": 9.93924095281026e-05, + "loss": 0.0769, + "step": 9720 + }, + { + "epoch": 0.6374893533381379, + "grad_norm": 0.3353469669818878, + "learning_rate": 9.938983697397948e-05, + "loss": 0.0845, + "step": 9730 + }, + { + "epoch": 0.6381445325296469, + "grad_norm": 0.7161203622817993, + "learning_rate": 9.938725901865805e-05, + "loss": 0.0913, + "step": 9740 + }, + { + "epoch": 0.6387997117211557, + "grad_norm": 1.0121372938156128, + "learning_rate": 9.93846756624202e-05, + "loss": 0.0861, + "step": 9750 + }, + { + "epoch": 0.6394548909126646, + "grad_norm": 0.6922144889831543, + "learning_rate": 9.938208690554849e-05, + "loss": 0.0944, + "step": 9760 + }, + { + "epoch": 0.6401100701041735, + "grad_norm": 0.5228187441825867, + "learning_rate": 9.9379492748326e-05, + "loss": 0.0756, + "step": 9770 + }, + { + "epoch": 0.6407652492956823, + "grad_norm": 0.8822422027587891, + "learning_rate": 9.937689319103641e-05, + "loss": 0.0956, + "step": 9780 + }, + { + "epoch": 0.6414204284871913, + "grad_norm": 0.7908045649528503, + "learning_rate": 9.937428823396404e-05, + "loss": 0.0542, + "step": 9790 + }, + { + "epoch": 0.6420756076787001, + "grad_norm": 0.5906897783279419, + "learning_rate": 9.937167787739372e-05, + "loss": 0.0747, + "step": 9800 + }, + { + "epoch": 0.642730786870209, + "grad_norm": 0.854347825050354, + "learning_rate": 9.936906212161095e-05, + "loss": 0.0643, + "step": 9810 + }, + { + "epoch": 0.6433859660617178, + "grad_norm": 0.9236022233963013, + "learning_rate": 9.936644096690176e-05, + "loss": 0.1019, + "step": 9820 + }, + { + "epoch": 0.6440411452532268, + "grad_norm": 0.7923506498336792, + "learning_rate": 9.936381441355282e-05, + "loss": 0.0718, + "step": 9830 + }, + { + "epoch": 0.6446963244447357, + "grad_norm": 1.199886441230774, + "learning_rate": 9.936118246185136e-05, + "loss": 0.0632, + "step": 9840 + }, + { + "epoch": 0.6453515036362445, + "grad_norm": 0.7486660480499268, + "learning_rate": 9.935854511208518e-05, + "loss": 0.045, + "step": 9850 + }, + { + "epoch": 0.6460066828277534, + "grad_norm": 0.7027573585510254, + "learning_rate": 9.935590236454272e-05, + "loss": 0.0705, + "step": 9860 + }, + { + "epoch": 0.6466618620192622, + "grad_norm": 0.46722567081451416, + "learning_rate": 9.935325421951298e-05, + "loss": 0.0552, + "step": 9870 + }, + { + "epoch": 0.6473170412107712, + "grad_norm": 1.1473817825317383, + "learning_rate": 9.935060067728557e-05, + "loss": 0.0891, + "step": 9880 + }, + { + "epoch": 0.64797222040228, + "grad_norm": 0.911948025226593, + "learning_rate": 9.934794173815067e-05, + "loss": 0.0877, + "step": 9890 + }, + { + "epoch": 0.6486273995937889, + "grad_norm": 0.5844174027442932, + "learning_rate": 9.934527740239906e-05, + "loss": 0.047, + "step": 9900 + }, + { + "epoch": 0.6492825787852978, + "grad_norm": 0.36306607723236084, + "learning_rate": 9.934260767032209e-05, + "loss": 0.0731, + "step": 9910 + }, + { + "epoch": 0.6499377579768066, + "grad_norm": 0.9279423356056213, + "learning_rate": 9.933993254221172e-05, + "loss": 0.095, + "step": 9920 + }, + { + "epoch": 0.6505929371683156, + "grad_norm": 0.5920722484588623, + "learning_rate": 9.933725201836053e-05, + "loss": 0.0624, + "step": 9930 + }, + { + "epoch": 0.6512481163598244, + "grad_norm": 0.8080114722251892, + "learning_rate": 9.933456609906162e-05, + "loss": 0.0612, + "step": 9940 + }, + { + "epoch": 0.6519032955513333, + "grad_norm": 0.9863518476486206, + "learning_rate": 9.933187478460875e-05, + "loss": 0.0708, + "step": 9950 + }, + { + "epoch": 0.6525584747428421, + "grad_norm": 0.7436809539794922, + "learning_rate": 9.93291780752962e-05, + "loss": 0.0928, + "step": 9960 + }, + { + "epoch": 0.653213653934351, + "grad_norm": 0.6209130883216858, + "learning_rate": 9.932647597141893e-05, + "loss": 0.0438, + "step": 9970 + }, + { + "epoch": 0.65386883312586, + "grad_norm": 0.5072580575942993, + "learning_rate": 9.932376847327239e-05, + "loss": 0.0807, + "step": 9980 + }, + { + "epoch": 0.6545240123173688, + "grad_norm": 0.443006694316864, + "learning_rate": 9.932105558115268e-05, + "loss": 0.0368, + "step": 9990 + }, + { + "epoch": 0.6551791915088777, + "grad_norm": 0.3210964798927307, + "learning_rate": 9.931833729535651e-05, + "loss": 0.0346, + "step": 10000 + }, + { + "epoch": 0.6558343707003865, + "grad_norm": 0.6528835296630859, + "learning_rate": 9.931561361618111e-05, + "loss": 0.0844, + "step": 10010 + }, + { + "epoch": 0.6564895498918955, + "grad_norm": 0.7609911561012268, + "learning_rate": 9.931288454392435e-05, + "loss": 0.0616, + "step": 10020 + }, + { + "epoch": 0.6571447290834043, + "grad_norm": 0.9895387887954712, + "learning_rate": 9.931015007888467e-05, + "loss": 0.0631, + "step": 10030 + }, + { + "epoch": 0.6577999082749132, + "grad_norm": 1.1014769077301025, + "learning_rate": 9.930741022136112e-05, + "loss": 0.0706, + "step": 10040 + }, + { + "epoch": 0.6584550874664221, + "grad_norm": 1.954559564590454, + "learning_rate": 9.930466497165333e-05, + "loss": 0.0996, + "step": 10050 + }, + { + "epoch": 0.6591102666579309, + "grad_norm": 1.2580156326293945, + "learning_rate": 9.93019143300615e-05, + "loss": 0.0754, + "step": 10060 + }, + { + "epoch": 0.6597654458494399, + "grad_norm": 0.9650811553001404, + "learning_rate": 9.929915829688644e-05, + "loss": 0.0811, + "step": 10070 + }, + { + "epoch": 0.6604206250409487, + "grad_norm": 0.2578173279762268, + "learning_rate": 9.929639687242955e-05, + "loss": 0.062, + "step": 10080 + }, + { + "epoch": 0.6610758042324576, + "grad_norm": 0.47148922085762024, + "learning_rate": 9.929363005699281e-05, + "loss": 0.0693, + "step": 10090 + }, + { + "epoch": 0.6617309834239664, + "grad_norm": 1.1240735054016113, + "learning_rate": 9.92908578508788e-05, + "loss": 0.0584, + "step": 10100 + }, + { + "epoch": 0.6623861626154753, + "grad_norm": 0.4830908179283142, + "learning_rate": 9.928808025439069e-05, + "loss": 0.0541, + "step": 10110 + }, + { + "epoch": 0.6630413418069843, + "grad_norm": 1.2094179391860962, + "learning_rate": 9.928529726783223e-05, + "loss": 0.0554, + "step": 10120 + }, + { + "epoch": 0.6636965209984931, + "grad_norm": 0.6858000755310059, + "learning_rate": 9.928250889150774e-05, + "loss": 0.0499, + "step": 10130 + }, + { + "epoch": 0.664351700190002, + "grad_norm": 0.20488876104354858, + "learning_rate": 9.92797151257222e-05, + "loss": 0.0708, + "step": 10140 + }, + { + "epoch": 0.6650068793815108, + "grad_norm": 0.6084568500518799, + "learning_rate": 9.927691597078108e-05, + "loss": 0.0443, + "step": 10150 + }, + { + "epoch": 0.6656620585730197, + "grad_norm": 0.8041579127311707, + "learning_rate": 9.927411142699053e-05, + "loss": 0.0629, + "step": 10160 + }, + { + "epoch": 0.6663172377645286, + "grad_norm": 0.945881187915802, + "learning_rate": 9.927130149465725e-05, + "loss": 0.073, + "step": 10170 + }, + { + "epoch": 0.6669724169560375, + "grad_norm": 0.43363767862319946, + "learning_rate": 9.92684861740885e-05, + "loss": 0.0496, + "step": 10180 + }, + { + "epoch": 0.6676275961475464, + "grad_norm": 0.6021831035614014, + "learning_rate": 9.926566546559217e-05, + "loss": 0.0334, + "step": 10190 + }, + { + "epoch": 0.6682827753390552, + "grad_norm": 0.3324200510978699, + "learning_rate": 9.926283936947673e-05, + "loss": 0.0674, + "step": 10200 + }, + { + "epoch": 0.6689379545305641, + "grad_norm": 0.7713364362716675, + "learning_rate": 9.926000788605126e-05, + "loss": 0.0918, + "step": 10210 + }, + { + "epoch": 0.669593133722073, + "grad_norm": 1.022048830986023, + "learning_rate": 9.92571710156254e-05, + "loss": 0.0785, + "step": 10220 + }, + { + "epoch": 0.6702483129135819, + "grad_norm": 0.9834038615226746, + "learning_rate": 9.925432875850936e-05, + "loss": 0.0521, + "step": 10230 + }, + { + "epoch": 0.6709034921050907, + "grad_norm": 0.6301253437995911, + "learning_rate": 9.925148111501396e-05, + "loss": 0.0453, + "step": 10240 + }, + { + "epoch": 0.6715586712965996, + "grad_norm": 1.0116667747497559, + "learning_rate": 9.924862808545066e-05, + "loss": 0.0841, + "step": 10250 + }, + { + "epoch": 0.6722138504881084, + "grad_norm": 0.9091498255729675, + "learning_rate": 9.924576967013141e-05, + "loss": 0.0762, + "step": 10260 + }, + { + "epoch": 0.6728690296796174, + "grad_norm": 0.8933075666427612, + "learning_rate": 9.924290586936887e-05, + "loss": 0.0646, + "step": 10270 + }, + { + "epoch": 0.6735242088711263, + "grad_norm": 0.682296633720398, + "learning_rate": 9.924003668347614e-05, + "loss": 0.084, + "step": 10280 + }, + { + "epoch": 0.6741793880626351, + "grad_norm": 0.9174949526786804, + "learning_rate": 9.923716211276704e-05, + "loss": 0.0915, + "step": 10290 + }, + { + "epoch": 0.674834567254144, + "grad_norm": 0.23098523914813995, + "learning_rate": 9.923428215755594e-05, + "loss": 0.0848, + "step": 10300 + }, + { + "epoch": 0.6754897464456528, + "grad_norm": 0.7185708284378052, + "learning_rate": 9.923139681815775e-05, + "loss": 0.0701, + "step": 10310 + }, + { + "epoch": 0.6761449256371618, + "grad_norm": 0.7687132954597473, + "learning_rate": 9.922850609488801e-05, + "loss": 0.0413, + "step": 10320 + }, + { + "epoch": 0.6768001048286706, + "grad_norm": 0.7559346556663513, + "learning_rate": 9.922560998806287e-05, + "loss": 0.0719, + "step": 10330 + }, + { + "epoch": 0.6774552840201795, + "grad_norm": 1.2165477275848389, + "learning_rate": 9.922270849799905e-05, + "loss": 0.0584, + "step": 10340 + }, + { + "epoch": 0.6781104632116884, + "grad_norm": 0.7038992047309875, + "learning_rate": 9.92198016250138e-05, + "loss": 0.0531, + "step": 10350 + }, + { + "epoch": 0.6787656424031973, + "grad_norm": 0.34646761417388916, + "learning_rate": 9.921688936942506e-05, + "loss": 0.0726, + "step": 10360 + }, + { + "epoch": 0.6794208215947062, + "grad_norm": 1.0250216722488403, + "learning_rate": 9.921397173155129e-05, + "loss": 0.0877, + "step": 10370 + }, + { + "epoch": 0.680076000786215, + "grad_norm": 0.8098081946372986, + "learning_rate": 9.921104871171157e-05, + "loss": 0.1222, + "step": 10380 + }, + { + "epoch": 0.6807311799777239, + "grad_norm": 0.9933450222015381, + "learning_rate": 9.920812031022554e-05, + "loss": 0.0561, + "step": 10390 + }, + { + "epoch": 0.6813863591692327, + "grad_norm": 0.6944500803947449, + "learning_rate": 9.920518652741348e-05, + "loss": 0.0489, + "step": 10400 + }, + { + "epoch": 0.6820415383607417, + "grad_norm": 0.8557353019714355, + "learning_rate": 9.920224736359618e-05, + "loss": 0.0637, + "step": 10410 + }, + { + "epoch": 0.6826967175522506, + "grad_norm": 0.315826416015625, + "learning_rate": 9.91993028190951e-05, + "loss": 0.0559, + "step": 10420 + }, + { + "epoch": 0.6833518967437594, + "grad_norm": 1.157729148864746, + "learning_rate": 9.919635289423222e-05, + "loss": 0.0737, + "step": 10430 + }, + { + "epoch": 0.6840070759352683, + "grad_norm": 0.9369429349899292, + "learning_rate": 9.919339758933015e-05, + "loss": 0.068, + "step": 10440 + }, + { + "epoch": 0.6846622551267771, + "grad_norm": 1.3324368000030518, + "learning_rate": 9.919043690471209e-05, + "loss": 0.0688, + "step": 10450 + }, + { + "epoch": 0.6853174343182861, + "grad_norm": 0.9081782102584839, + "learning_rate": 9.91874708407018e-05, + "loss": 0.0626, + "step": 10460 + }, + { + "epoch": 0.6859726135097949, + "grad_norm": 0.526563823223114, + "learning_rate": 9.918449939762367e-05, + "loss": 0.0715, + "step": 10470 + }, + { + "epoch": 0.6866277927013038, + "grad_norm": 0.2926543354988098, + "learning_rate": 9.91815225758026e-05, + "loss": 0.0382, + "step": 10480 + }, + { + "epoch": 0.6872829718928127, + "grad_norm": 0.3331846594810486, + "learning_rate": 9.917854037556419e-05, + "loss": 0.0722, + "step": 10490 + }, + { + "epoch": 0.6879381510843215, + "grad_norm": 0.13026654720306396, + "learning_rate": 9.917555279723454e-05, + "loss": 0.0466, + "step": 10500 + }, + { + "epoch": 0.6885933302758305, + "grad_norm": 0.7604377269744873, + "learning_rate": 9.917255984114036e-05, + "loss": 0.0655, + "step": 10510 + }, + { + "epoch": 0.6892485094673393, + "grad_norm": 0.2936903238296509, + "learning_rate": 9.916956150760896e-05, + "loss": 0.0725, + "step": 10520 + }, + { + "epoch": 0.6899036886588482, + "grad_norm": 0.8423580527305603, + "learning_rate": 9.916655779696826e-05, + "loss": 0.0607, + "step": 10530 + }, + { + "epoch": 0.690558867850357, + "grad_norm": 1.6415365934371948, + "learning_rate": 9.916354870954671e-05, + "loss": 0.1203, + "step": 10540 + }, + { + "epoch": 0.691214047041866, + "grad_norm": 0.835259735584259, + "learning_rate": 9.91605342456734e-05, + "loss": 0.0769, + "step": 10550 + }, + { + "epoch": 0.6918692262333749, + "grad_norm": 0.5425803661346436, + "learning_rate": 9.915751440567795e-05, + "loss": 0.0632, + "step": 10560 + }, + { + "epoch": 0.6925244054248837, + "grad_norm": 0.34817004203796387, + "learning_rate": 9.915448918989066e-05, + "loss": 0.0764, + "step": 10570 + }, + { + "epoch": 0.6931795846163926, + "grad_norm": 1.4936268329620361, + "learning_rate": 9.915145859864232e-05, + "loss": 0.0852, + "step": 10580 + }, + { + "epoch": 0.6938347638079014, + "grad_norm": 0.629213273525238, + "learning_rate": 9.914842263226437e-05, + "loss": 0.0971, + "step": 10590 + }, + { + "epoch": 0.6944899429994104, + "grad_norm": 0.9336279630661011, + "learning_rate": 9.914538129108882e-05, + "loss": 0.0607, + "step": 10600 + }, + { + "epoch": 0.6951451221909192, + "grad_norm": 0.5682711601257324, + "learning_rate": 9.914233457544825e-05, + "loss": 0.0857, + "step": 10610 + }, + { + "epoch": 0.6958003013824281, + "grad_norm": 1.521209955215454, + "learning_rate": 9.913928248567586e-05, + "loss": 0.0766, + "step": 10620 + }, + { + "epoch": 0.696455480573937, + "grad_norm": 0.5576208233833313, + "learning_rate": 9.913622502210542e-05, + "loss": 0.0858, + "step": 10630 + }, + { + "epoch": 0.6971106597654458, + "grad_norm": 1.3288923501968384, + "learning_rate": 9.913316218507128e-05, + "loss": 0.1006, + "step": 10640 + }, + { + "epoch": 0.6977658389569548, + "grad_norm": 0.8860986232757568, + "learning_rate": 9.91300939749084e-05, + "loss": 0.0451, + "step": 10650 + }, + { + "epoch": 0.6984210181484636, + "grad_norm": 1.1577712297439575, + "learning_rate": 9.91270203919523e-05, + "loss": 0.0959, + "step": 10660 + }, + { + "epoch": 0.6990761973399725, + "grad_norm": 0.5856321454048157, + "learning_rate": 9.912394143653912e-05, + "loss": 0.0732, + "step": 10670 + }, + { + "epoch": 0.6997313765314813, + "grad_norm": 1.2094820737838745, + "learning_rate": 9.912085710900555e-05, + "loss": 0.0729, + "step": 10680 + }, + { + "epoch": 0.7003865557229902, + "grad_norm": 0.6097927689552307, + "learning_rate": 9.911776740968892e-05, + "loss": 0.0413, + "step": 10690 + }, + { + "epoch": 0.7010417349144992, + "grad_norm": 1.5615991353988647, + "learning_rate": 9.911467233892709e-05, + "loss": 0.0879, + "step": 10700 + }, + { + "epoch": 0.701696914106008, + "grad_norm": 0.2832181453704834, + "learning_rate": 9.911157189705853e-05, + "loss": 0.0604, + "step": 10710 + }, + { + "epoch": 0.7023520932975169, + "grad_norm": 0.5674335956573486, + "learning_rate": 9.910846608442229e-05, + "loss": 0.0869, + "step": 10720 + }, + { + "epoch": 0.7030072724890257, + "grad_norm": 0.738660991191864, + "learning_rate": 9.910535490135805e-05, + "loss": 0.0899, + "step": 10730 + }, + { + "epoch": 0.7036624516805347, + "grad_norm": 0.6670799851417542, + "learning_rate": 9.910223834820603e-05, + "loss": 0.089, + "step": 10740 + }, + { + "epoch": 0.7043176308720435, + "grad_norm": 0.8842169642448425, + "learning_rate": 9.909911642530703e-05, + "loss": 0.0703, + "step": 10750 + }, + { + "epoch": 0.7049728100635524, + "grad_norm": 0.8229660391807556, + "learning_rate": 9.909598913300249e-05, + "loss": 0.0518, + "step": 10760 + }, + { + "epoch": 0.7056279892550612, + "grad_norm": 0.3472031056880951, + "learning_rate": 9.909285647163438e-05, + "loss": 0.053, + "step": 10770 + }, + { + "epoch": 0.7062831684465701, + "grad_norm": 0.6167739629745483, + "learning_rate": 9.908971844154531e-05, + "loss": 0.0688, + "step": 10780 + }, + { + "epoch": 0.7069383476380791, + "grad_norm": 0.28739380836486816, + "learning_rate": 9.908657504307843e-05, + "loss": 0.0363, + "step": 10790 + }, + { + "epoch": 0.7075935268295879, + "grad_norm": 0.533054769039154, + "learning_rate": 9.908342627657751e-05, + "loss": 0.0428, + "step": 10800 + }, + { + "epoch": 0.7082487060210968, + "grad_norm": 1.2700632810592651, + "learning_rate": 9.908027214238689e-05, + "loss": 0.0736, + "step": 10810 + }, + { + "epoch": 0.7089038852126056, + "grad_norm": 1.1613316535949707, + "learning_rate": 9.90771126408515e-05, + "loss": 0.0583, + "step": 10820 + }, + { + "epoch": 0.7095590644041145, + "grad_norm": 1.0795722007751465, + "learning_rate": 9.907394777231685e-05, + "loss": 0.1041, + "step": 10830 + }, + { + "epoch": 0.7102142435956234, + "grad_norm": 0.4355883300304413, + "learning_rate": 9.907077753712905e-05, + "loss": 0.0631, + "step": 10840 + }, + { + "epoch": 0.7108694227871323, + "grad_norm": 0.3509097993373871, + "learning_rate": 9.906760193563482e-05, + "loss": 0.0601, + "step": 10850 + }, + { + "epoch": 0.7115246019786412, + "grad_norm": 0.8449415564537048, + "learning_rate": 9.906442096818139e-05, + "loss": 0.0759, + "step": 10860 + }, + { + "epoch": 0.71217978117015, + "grad_norm": 0.3019065856933594, + "learning_rate": 9.906123463511665e-05, + "loss": 0.0922, + "step": 10870 + }, + { + "epoch": 0.7128349603616589, + "grad_norm": 0.8554864525794983, + "learning_rate": 9.905804293678907e-05, + "loss": 0.1141, + "step": 10880 + }, + { + "epoch": 0.7134901395531678, + "grad_norm": 0.5403743982315063, + "learning_rate": 9.905484587354766e-05, + "loss": 0.0601, + "step": 10890 + }, + { + "epoch": 0.7141453187446767, + "grad_norm": 0.35310614109039307, + "learning_rate": 9.905164344574205e-05, + "loss": 0.0495, + "step": 10900 + }, + { + "epoch": 0.7148004979361855, + "grad_norm": 1.1814618110656738, + "learning_rate": 9.904843565372248e-05, + "loss": 0.0822, + "step": 10910 + }, + { + "epoch": 0.7154556771276944, + "grad_norm": 0.739508867263794, + "learning_rate": 9.904522249783972e-05, + "loss": 0.0453, + "step": 10920 + }, + { + "epoch": 0.7161108563192033, + "grad_norm": 0.6586981415748596, + "learning_rate": 9.904200397844517e-05, + "loss": 0.0365, + "step": 10930 + }, + { + "epoch": 0.7167660355107122, + "grad_norm": 0.610660195350647, + "learning_rate": 9.903878009589078e-05, + "loss": 0.062, + "step": 10940 + }, + { + "epoch": 0.7174212147022211, + "grad_norm": 0.5311826467514038, + "learning_rate": 9.903555085052915e-05, + "loss": 0.0682, + "step": 10950 + }, + { + "epoch": 0.7180763938937299, + "grad_norm": 0.3531237542629242, + "learning_rate": 9.903231624271338e-05, + "loss": 0.0672, + "step": 10960 + }, + { + "epoch": 0.7187315730852388, + "grad_norm": 0.8045540452003479, + "learning_rate": 9.902907627279724e-05, + "loss": 0.0643, + "step": 10970 + }, + { + "epoch": 0.7193867522767476, + "grad_norm": 0.9198805093765259, + "learning_rate": 9.902583094113504e-05, + "loss": 0.0717, + "step": 10980 + }, + { + "epoch": 0.7200419314682566, + "grad_norm": 0.24727238714694977, + "learning_rate": 9.902258024808168e-05, + "loss": 0.0478, + "step": 10990 + }, + { + "epoch": 0.7206971106597655, + "grad_norm": 0.4290715456008911, + "learning_rate": 9.901932419399264e-05, + "loss": 0.044, + "step": 11000 + }, + { + "epoch": 0.7213522898512743, + "grad_norm": 2.638254165649414, + "learning_rate": 9.9016062779224e-05, + "loss": 0.0873, + "step": 11010 + }, + { + "epoch": 0.7220074690427832, + "grad_norm": 1.037346601486206, + "learning_rate": 9.901279600413242e-05, + "loss": 0.0482, + "step": 11020 + }, + { + "epoch": 0.722662648234292, + "grad_norm": 0.3130969703197479, + "learning_rate": 9.900952386907518e-05, + "loss": 0.0479, + "step": 11030 + }, + { + "epoch": 0.723317827425801, + "grad_norm": 0.3741985857486725, + "learning_rate": 9.90062463744101e-05, + "loss": 0.0762, + "step": 11040 + }, + { + "epoch": 0.7239730066173098, + "grad_norm": 0.4156497120857239, + "learning_rate": 9.900296352049558e-05, + "loss": 0.0941, + "step": 11050 + }, + { + "epoch": 0.7246281858088187, + "grad_norm": 0.9379881620407104, + "learning_rate": 9.899967530769065e-05, + "loss": 0.0545, + "step": 11060 + }, + { + "epoch": 0.7252833650003276, + "grad_norm": 0.7171107530593872, + "learning_rate": 9.899638173635489e-05, + "loss": 0.0909, + "step": 11070 + }, + { + "epoch": 0.7259385441918365, + "grad_norm": 0.6484657526016235, + "learning_rate": 9.899308280684849e-05, + "loss": 0.0582, + "step": 11080 + }, + { + "epoch": 0.7265937233833454, + "grad_norm": 0.6288986802101135, + "learning_rate": 9.898977851953222e-05, + "loss": 0.0572, + "step": 11090 + }, + { + "epoch": 0.7272489025748542, + "grad_norm": 0.6718704700469971, + "learning_rate": 9.898646887476741e-05, + "loss": 0.0805, + "step": 11100 + }, + { + "epoch": 0.7279040817663631, + "grad_norm": 0.39433547854423523, + "learning_rate": 9.898315387291603e-05, + "loss": 0.0726, + "step": 11110 + }, + { + "epoch": 0.7285592609578719, + "grad_norm": 0.7601620554924011, + "learning_rate": 9.89798335143406e-05, + "loss": 0.0512, + "step": 11120 + }, + { + "epoch": 0.7292144401493809, + "grad_norm": 0.8300408124923706, + "learning_rate": 9.897650779940419e-05, + "loss": 0.0497, + "step": 11130 + }, + { + "epoch": 0.7298696193408898, + "grad_norm": 1.1292845010757446, + "learning_rate": 9.897317672847054e-05, + "loss": 0.0719, + "step": 11140 + }, + { + "epoch": 0.7305247985323986, + "grad_norm": 0.8077048659324646, + "learning_rate": 9.89698403019039e-05, + "loss": 0.0629, + "step": 11150 + }, + { + "epoch": 0.7311799777239075, + "grad_norm": 0.6272961497306824, + "learning_rate": 9.896649852006917e-05, + "loss": 0.046, + "step": 11160 + }, + { + "epoch": 0.7318351569154163, + "grad_norm": 0.8412403464317322, + "learning_rate": 9.896315138333177e-05, + "loss": 0.0483, + "step": 11170 + }, + { + "epoch": 0.7324903361069253, + "grad_norm": 0.7861549854278564, + "learning_rate": 9.895979889205774e-05, + "loss": 0.0541, + "step": 11180 + }, + { + "epoch": 0.7331455152984341, + "grad_norm": 0.43130820989608765, + "learning_rate": 9.895644104661372e-05, + "loss": 0.0525, + "step": 11190 + }, + { + "epoch": 0.733800694489943, + "grad_norm": 0.17008890211582184, + "learning_rate": 9.895307784736691e-05, + "loss": 0.0662, + "step": 11200 + }, + { + "epoch": 0.7344558736814519, + "grad_norm": 0.4551737606525421, + "learning_rate": 9.894970929468512e-05, + "loss": 0.0431, + "step": 11210 + }, + { + "epoch": 0.7351110528729607, + "grad_norm": 0.8134713768959045, + "learning_rate": 9.89463353889367e-05, + "loss": 0.059, + "step": 11220 + }, + { + "epoch": 0.7357662320644697, + "grad_norm": 0.36902284622192383, + "learning_rate": 9.894295613049065e-05, + "loss": 0.1179, + "step": 11230 + }, + { + "epoch": 0.7364214112559785, + "grad_norm": 0.42335203289985657, + "learning_rate": 9.893957151971649e-05, + "loss": 0.0701, + "step": 11240 + }, + { + "epoch": 0.7370765904474874, + "grad_norm": 0.29900500178337097, + "learning_rate": 9.893618155698436e-05, + "loss": 0.0592, + "step": 11250 + }, + { + "epoch": 0.7377317696389962, + "grad_norm": 1.4258873462677002, + "learning_rate": 9.8932786242665e-05, + "loss": 0.0584, + "step": 11260 + }, + { + "epoch": 0.7383869488305052, + "grad_norm": 0.33801984786987305, + "learning_rate": 9.89293855771297e-05, + "loss": 0.0791, + "step": 11270 + }, + { + "epoch": 0.739042128022014, + "grad_norm": 0.9691896438598633, + "learning_rate": 9.892597956075036e-05, + "loss": 0.081, + "step": 11280 + }, + { + "epoch": 0.7396973072135229, + "grad_norm": 1.046712040901184, + "learning_rate": 9.892256819389947e-05, + "loss": 0.0802, + "step": 11290 + }, + { + "epoch": 0.7403524864050318, + "grad_norm": 0.8705103397369385, + "learning_rate": 9.891915147695006e-05, + "loss": 0.0827, + "step": 11300 + }, + { + "epoch": 0.7410076655965406, + "grad_norm": 1.1775574684143066, + "learning_rate": 9.891572941027577e-05, + "loss": 0.0623, + "step": 11310 + }, + { + "epoch": 0.7416628447880496, + "grad_norm": 0.6036169528961182, + "learning_rate": 9.89123019942509e-05, + "loss": 0.0599, + "step": 11320 + }, + { + "epoch": 0.7423180239795584, + "grad_norm": 0.9088860750198364, + "learning_rate": 9.89088692292502e-05, + "loss": 0.1155, + "step": 11330 + }, + { + "epoch": 0.7429732031710673, + "grad_norm": 0.4484582245349884, + "learning_rate": 9.89054311156491e-05, + "loss": 0.0559, + "step": 11340 + }, + { + "epoch": 0.7436283823625761, + "grad_norm": 0.4833897054195404, + "learning_rate": 9.890198765382357e-05, + "loss": 0.0923, + "step": 11350 + }, + { + "epoch": 0.744283561554085, + "grad_norm": 0.510776937007904, + "learning_rate": 9.889853884415021e-05, + "loss": 0.0686, + "step": 11360 + }, + { + "epoch": 0.744938740745594, + "grad_norm": 0.5318233370780945, + "learning_rate": 9.889508468700614e-05, + "loss": 0.066, + "step": 11370 + }, + { + "epoch": 0.7455939199371028, + "grad_norm": 0.9386031627655029, + "learning_rate": 9.889162518276915e-05, + "loss": 0.0582, + "step": 11380 + }, + { + "epoch": 0.7462490991286117, + "grad_norm": 0.9004435539245605, + "learning_rate": 9.888816033181752e-05, + "loss": 0.0634, + "step": 11390 + }, + { + "epoch": 0.7469042783201205, + "grad_norm": 1.3467686176300049, + "learning_rate": 9.888469013453018e-05, + "loss": 0.0736, + "step": 11400 + }, + { + "epoch": 0.7475594575116294, + "grad_norm": 1.2069320678710938, + "learning_rate": 9.888121459128663e-05, + "loss": 0.0643, + "step": 11410 + }, + { + "epoch": 0.7482146367031383, + "grad_norm": 1.0083979368209839, + "learning_rate": 9.887773370246693e-05, + "loss": 0.0624, + "step": 11420 + }, + { + "epoch": 0.7488698158946472, + "grad_norm": 0.7105551362037659, + "learning_rate": 9.887424746845177e-05, + "loss": 0.0739, + "step": 11430 + }, + { + "epoch": 0.7495249950861561, + "grad_norm": 1.0993916988372803, + "learning_rate": 9.887075588962239e-05, + "loss": 0.0778, + "step": 11440 + }, + { + "epoch": 0.7501801742776649, + "grad_norm": 0.9630560874938965, + "learning_rate": 9.88672589663606e-05, + "loss": 0.0748, + "step": 11450 + }, + { + "epoch": 0.7508353534691738, + "grad_norm": 0.9497333765029907, + "learning_rate": 9.886375669904886e-05, + "loss": 0.0695, + "step": 11460 + }, + { + "epoch": 0.7514905326606827, + "grad_norm": 1.300858974456787, + "learning_rate": 9.886024908807014e-05, + "loss": 0.0691, + "step": 11470 + }, + { + "epoch": 0.7521457118521916, + "grad_norm": 0.654536783695221, + "learning_rate": 9.885673613380806e-05, + "loss": 0.0646, + "step": 11480 + }, + { + "epoch": 0.7528008910437004, + "grad_norm": 0.6681713461875916, + "learning_rate": 9.885321783664676e-05, + "loss": 0.065, + "step": 11490 + }, + { + "epoch": 0.7534560702352093, + "grad_norm": 0.9725565314292908, + "learning_rate": 9.884969419697101e-05, + "loss": 0.0711, + "step": 11500 + }, + { + "epoch": 0.7541112494267183, + "grad_norm": 0.43574202060699463, + "learning_rate": 9.884616521516614e-05, + "loss": 0.0393, + "step": 11510 + }, + { + "epoch": 0.7547664286182271, + "grad_norm": 1.6250193119049072, + "learning_rate": 9.88426308916181e-05, + "loss": 0.0645, + "step": 11520 + }, + { + "epoch": 0.755421607809736, + "grad_norm": 1.1976591348648071, + "learning_rate": 9.883909122671335e-05, + "loss": 0.1353, + "step": 11530 + }, + { + "epoch": 0.7560767870012448, + "grad_norm": 0.9356075525283813, + "learning_rate": 9.883554622083904e-05, + "loss": 0.0463, + "step": 11540 + }, + { + "epoch": 0.7567319661927537, + "grad_norm": 1.2190067768096924, + "learning_rate": 9.88319958743828e-05, + "loss": 0.0691, + "step": 11550 + }, + { + "epoch": 0.7573871453842626, + "grad_norm": 0.24552957713603973, + "learning_rate": 9.882844018773291e-05, + "loss": 0.0502, + "step": 11560 + }, + { + "epoch": 0.7580423245757715, + "grad_norm": 0.5455625057220459, + "learning_rate": 9.882487916127823e-05, + "loss": 0.0702, + "step": 11570 + }, + { + "epoch": 0.7586975037672804, + "grad_norm": 0.6401561498641968, + "learning_rate": 9.882131279540815e-05, + "loss": 0.0437, + "step": 11580 + }, + { + "epoch": 0.7593526829587892, + "grad_norm": 0.35239332914352417, + "learning_rate": 9.881774109051271e-05, + "loss": 0.1003, + "step": 11590 + }, + { + "epoch": 0.7600078621502981, + "grad_norm": 0.4812977910041809, + "learning_rate": 9.881416404698252e-05, + "loss": 0.059, + "step": 11600 + }, + { + "epoch": 0.760663041341807, + "grad_norm": 0.4734530746936798, + "learning_rate": 9.881058166520873e-05, + "loss": 0.0548, + "step": 11610 + }, + { + "epoch": 0.7613182205333159, + "grad_norm": 0.6526734828948975, + "learning_rate": 9.880699394558311e-05, + "loss": 0.0611, + "step": 11620 + }, + { + "epoch": 0.7619733997248247, + "grad_norm": 1.3914482593536377, + "learning_rate": 9.880340088849801e-05, + "loss": 0.0526, + "step": 11630 + }, + { + "epoch": 0.7626285789163336, + "grad_norm": 1.3396388292312622, + "learning_rate": 9.879980249434637e-05, + "loss": 0.0673, + "step": 11640 + }, + { + "epoch": 0.7632837581078425, + "grad_norm": 0.8525916934013367, + "learning_rate": 9.879619876352168e-05, + "loss": 0.0896, + "step": 11650 + }, + { + "epoch": 0.7639389372993514, + "grad_norm": 0.44395121932029724, + "learning_rate": 9.879258969641809e-05, + "loss": 0.058, + "step": 11660 + }, + { + "epoch": 0.7645941164908603, + "grad_norm": 1.8288313150405884, + "learning_rate": 9.878897529343023e-05, + "loss": 0.1015, + "step": 11670 + }, + { + "epoch": 0.7652492956823691, + "grad_norm": 0.6201252937316895, + "learning_rate": 9.878535555495338e-05, + "loss": 0.0659, + "step": 11680 + }, + { + "epoch": 0.765904474873878, + "grad_norm": 0.644300103187561, + "learning_rate": 9.87817304813834e-05, + "loss": 0.0556, + "step": 11690 + }, + { + "epoch": 0.7665596540653868, + "grad_norm": 0.3150923550128937, + "learning_rate": 9.877810007311671e-05, + "loss": 0.0638, + "step": 11700 + }, + { + "epoch": 0.7672148332568958, + "grad_norm": 0.7382038831710815, + "learning_rate": 9.877446433055035e-05, + "loss": 0.0408, + "step": 11710 + }, + { + "epoch": 0.7678700124484047, + "grad_norm": 0.6523844599723816, + "learning_rate": 9.877082325408191e-05, + "loss": 0.0684, + "step": 11720 + }, + { + "epoch": 0.7685251916399135, + "grad_norm": 0.7167715430259705, + "learning_rate": 9.876717684410954e-05, + "loss": 0.0573, + "step": 11730 + }, + { + "epoch": 0.7691803708314224, + "grad_norm": 0.26429006457328796, + "learning_rate": 9.876352510103204e-05, + "loss": 0.0687, + "step": 11740 + }, + { + "epoch": 0.7698355500229312, + "grad_norm": 1.1594114303588867, + "learning_rate": 9.875986802524875e-05, + "loss": 0.0374, + "step": 11750 + }, + { + "epoch": 0.7704907292144402, + "grad_norm": 0.7486523389816284, + "learning_rate": 9.87562056171596e-05, + "loss": 0.0594, + "step": 11760 + }, + { + "epoch": 0.771145908405949, + "grad_norm": 0.5457315444946289, + "learning_rate": 9.875253787716511e-05, + "loss": 0.0405, + "step": 11770 + }, + { + "epoch": 0.7718010875974579, + "grad_norm": 1.1460548639297485, + "learning_rate": 9.874886480566637e-05, + "loss": 0.0444, + "step": 11780 + }, + { + "epoch": 0.7724562667889667, + "grad_norm": 0.40328311920166016, + "learning_rate": 9.874518640306507e-05, + "loss": 0.0609, + "step": 11790 + }, + { + "epoch": 0.7731114459804757, + "grad_norm": 0.7628024816513062, + "learning_rate": 9.874150266976347e-05, + "loss": 0.0585, + "step": 11800 + }, + { + "epoch": 0.7737666251719846, + "grad_norm": 0.7153297662734985, + "learning_rate": 9.873781360616443e-05, + "loss": 0.0898, + "step": 11810 + }, + { + "epoch": 0.7744218043634934, + "grad_norm": 1.3416972160339355, + "learning_rate": 9.873411921267137e-05, + "loss": 0.0576, + "step": 11820 + }, + { + "epoch": 0.7750769835550023, + "grad_norm": 0.18572741746902466, + "learning_rate": 9.873041948968829e-05, + "loss": 0.0358, + "step": 11830 + }, + { + "epoch": 0.7757321627465111, + "grad_norm": 0.7675787806510925, + "learning_rate": 9.872671443761981e-05, + "loss": 0.0399, + "step": 11840 + }, + { + "epoch": 0.7763873419380201, + "grad_norm": 0.4240221381187439, + "learning_rate": 9.872300405687109e-05, + "loss": 0.0446, + "step": 11850 + }, + { + "epoch": 0.7770425211295289, + "grad_norm": 0.5478525161743164, + "learning_rate": 9.871928834784792e-05, + "loss": 0.042, + "step": 11860 + }, + { + "epoch": 0.7776977003210378, + "grad_norm": 0.5901211500167847, + "learning_rate": 9.871556731095661e-05, + "loss": 0.0978, + "step": 11870 + }, + { + "epoch": 0.7783528795125467, + "grad_norm": 0.8560400605201721, + "learning_rate": 9.871184094660411e-05, + "loss": 0.0681, + "step": 11880 + }, + { + "epoch": 0.7790080587040555, + "grad_norm": 0.44048166275024414, + "learning_rate": 9.870810925519791e-05, + "loss": 0.0669, + "step": 11890 + }, + { + "epoch": 0.7796632378955645, + "grad_norm": 0.7876608371734619, + "learning_rate": 9.870437223714612e-05, + "loss": 0.0481, + "step": 11900 + }, + { + "epoch": 0.7803184170870733, + "grad_norm": 0.22582761943340302, + "learning_rate": 9.87006298928574e-05, + "loss": 0.0609, + "step": 11910 + }, + { + "epoch": 0.7809735962785822, + "grad_norm": 1.0888128280639648, + "learning_rate": 9.869688222274103e-05, + "loss": 0.0587, + "step": 11920 + }, + { + "epoch": 0.781628775470091, + "grad_norm": 0.5261009335517883, + "learning_rate": 9.869312922720681e-05, + "loss": 0.056, + "step": 11930 + }, + { + "epoch": 0.7822839546616, + "grad_norm": 1.4940760135650635, + "learning_rate": 9.868937090666521e-05, + "loss": 0.0423, + "step": 11940 + }, + { + "epoch": 0.7829391338531089, + "grad_norm": 0.8198509216308594, + "learning_rate": 9.86856072615272e-05, + "loss": 0.0729, + "step": 11950 + }, + { + "epoch": 0.7835943130446177, + "grad_norm": 0.3592664897441864, + "learning_rate": 9.868183829220438e-05, + "loss": 0.0811, + "step": 11960 + }, + { + "epoch": 0.7842494922361266, + "grad_norm": 0.2775837182998657, + "learning_rate": 9.867806399910893e-05, + "loss": 0.0646, + "step": 11970 + }, + { + "epoch": 0.7849046714276354, + "grad_norm": 0.8185948133468628, + "learning_rate": 9.867428438265356e-05, + "loss": 0.0702, + "step": 11980 + }, + { + "epoch": 0.7855598506191444, + "grad_norm": 1.0236600637435913, + "learning_rate": 9.867049944325165e-05, + "loss": 0.0511, + "step": 11990 + }, + { + "epoch": 0.7862150298106532, + "grad_norm": 0.7525148987770081, + "learning_rate": 9.86667091813171e-05, + "loss": 0.0389, + "step": 12000 + }, + { + "epoch": 0.7868702090021621, + "grad_norm": 1.0738627910614014, + "learning_rate": 9.866291359726438e-05, + "loss": 0.051, + "step": 12010 + }, + { + "epoch": 0.787525388193671, + "grad_norm": 0.8436263203620911, + "learning_rate": 9.865911269150861e-05, + "loss": 0.0382, + "step": 12020 + }, + { + "epoch": 0.7881805673851798, + "grad_norm": 0.9784340858459473, + "learning_rate": 9.865530646446544e-05, + "loss": 0.0342, + "step": 12030 + }, + { + "epoch": 0.7888357465766888, + "grad_norm": 0.4408501386642456, + "learning_rate": 9.86514949165511e-05, + "loss": 0.0427, + "step": 12040 + }, + { + "epoch": 0.7894909257681976, + "grad_norm": 0.6748088598251343, + "learning_rate": 9.864767804818243e-05, + "loss": 0.068, + "step": 12050 + }, + { + "epoch": 0.7901461049597065, + "grad_norm": 1.1133543252944946, + "learning_rate": 9.86438558597768e-05, + "loss": 0.0676, + "step": 12060 + }, + { + "epoch": 0.7908012841512153, + "grad_norm": 0.39960020780563354, + "learning_rate": 9.864002835175225e-05, + "loss": 0.0645, + "step": 12070 + }, + { + "epoch": 0.7914564633427242, + "grad_norm": 0.4675722122192383, + "learning_rate": 9.863619552452734e-05, + "loss": 0.0339, + "step": 12080 + }, + { + "epoch": 0.7921116425342332, + "grad_norm": 0.6435307264328003, + "learning_rate": 9.863235737852119e-05, + "loss": 0.0741, + "step": 12090 + }, + { + "epoch": 0.792766821725742, + "grad_norm": 0.37774863839149475, + "learning_rate": 9.862851391415356e-05, + "loss": 0.049, + "step": 12100 + }, + { + "epoch": 0.7934220009172509, + "grad_norm": 1.4046707153320312, + "learning_rate": 9.862466513184477e-05, + "loss": 0.071, + "step": 12110 + }, + { + "epoch": 0.7940771801087597, + "grad_norm": 1.5061200857162476, + "learning_rate": 9.86208110320157e-05, + "loss": 0.067, + "step": 12120 + }, + { + "epoch": 0.7947323593002686, + "grad_norm": 0.7131142616271973, + "learning_rate": 9.861695161508784e-05, + "loss": 0.0591, + "step": 12130 + }, + { + "epoch": 0.7953875384917775, + "grad_norm": 0.9600571393966675, + "learning_rate": 9.861308688148324e-05, + "loss": 0.0661, + "step": 12140 + }, + { + "epoch": 0.7960427176832864, + "grad_norm": 0.31425774097442627, + "learning_rate": 9.860921683162455e-05, + "loss": 0.0664, + "step": 12150 + }, + { + "epoch": 0.7966978968747953, + "grad_norm": 1.0485790967941284, + "learning_rate": 9.860534146593499e-05, + "loss": 0.0696, + "step": 12160 + }, + { + "epoch": 0.7973530760663041, + "grad_norm": 0.5167580246925354, + "learning_rate": 9.860146078483836e-05, + "loss": 0.0697, + "step": 12170 + }, + { + "epoch": 0.798008255257813, + "grad_norm": 0.8679238557815552, + "learning_rate": 9.859757478875905e-05, + "loss": 0.0679, + "step": 12180 + }, + { + "epoch": 0.7986634344493219, + "grad_norm": 0.5310705304145813, + "learning_rate": 9.859368347812204e-05, + "loss": 0.0717, + "step": 12190 + }, + { + "epoch": 0.7993186136408308, + "grad_norm": 0.30686548352241516, + "learning_rate": 9.858978685335285e-05, + "loss": 0.0562, + "step": 12200 + }, + { + "epoch": 0.7999737928323396, + "grad_norm": 0.9645583033561707, + "learning_rate": 9.858588491487763e-05, + "loss": 0.0805, + "step": 12210 + }, + { + "epoch": 0.8006289720238485, + "grad_norm": 0.5635125041007996, + "learning_rate": 9.858197766312308e-05, + "loss": 0.0869, + "step": 12220 + }, + { + "epoch": 0.8012841512153575, + "grad_norm": 0.8317805528640747, + "learning_rate": 9.857806509851649e-05, + "loss": 0.0444, + "step": 12230 + }, + { + "epoch": 0.8019393304068663, + "grad_norm": 0.39037221670150757, + "learning_rate": 9.857414722148574e-05, + "loss": 0.0717, + "step": 12240 + }, + { + "epoch": 0.8025945095983752, + "grad_norm": 0.6396064758300781, + "learning_rate": 9.857022403245928e-05, + "loss": 0.0475, + "step": 12250 + }, + { + "epoch": 0.803249688789884, + "grad_norm": 0.41368481516838074, + "learning_rate": 9.856629553186615e-05, + "loss": 0.0582, + "step": 12260 + }, + { + "epoch": 0.8039048679813929, + "grad_norm": 0.6115670800209045, + "learning_rate": 9.856236172013595e-05, + "loss": 0.0618, + "step": 12270 + }, + { + "epoch": 0.8045600471729017, + "grad_norm": 0.5229902863502502, + "learning_rate": 9.85584225976989e-05, + "loss": 0.0473, + "step": 12280 + }, + { + "epoch": 0.8052152263644107, + "grad_norm": 0.5029119253158569, + "learning_rate": 9.855447816498575e-05, + "loss": 0.0381, + "step": 12290 + }, + { + "epoch": 0.8058704055559195, + "grad_norm": 0.8754163384437561, + "learning_rate": 9.855052842242787e-05, + "loss": 0.0804, + "step": 12300 + }, + { + "epoch": 0.8065255847474284, + "grad_norm": 0.2603570520877838, + "learning_rate": 9.85465733704572e-05, + "loss": 0.0607, + "step": 12310 + }, + { + "epoch": 0.8071807639389373, + "grad_norm": 0.39909499883651733, + "learning_rate": 9.854261300950624e-05, + "loss": 0.0643, + "step": 12320 + }, + { + "epoch": 0.8078359431304462, + "grad_norm": 0.7048105001449585, + "learning_rate": 9.853864734000813e-05, + "loss": 0.0445, + "step": 12330 + }, + { + "epoch": 0.8084911223219551, + "grad_norm": 0.6492025852203369, + "learning_rate": 9.85346763623965e-05, + "loss": 0.087, + "step": 12340 + }, + { + "epoch": 0.8091463015134639, + "grad_norm": 0.973595142364502, + "learning_rate": 9.853070007710564e-05, + "loss": 0.0632, + "step": 12350 + }, + { + "epoch": 0.8098014807049728, + "grad_norm": 0.942960798740387, + "learning_rate": 9.85267184845704e-05, + "loss": 0.0973, + "step": 12360 + }, + { + "epoch": 0.8104566598964816, + "grad_norm": 0.24829868972301483, + "learning_rate": 9.852273158522616e-05, + "loss": 0.0387, + "step": 12370 + }, + { + "epoch": 0.8111118390879906, + "grad_norm": 0.6125770211219788, + "learning_rate": 9.851873937950896e-05, + "loss": 0.0606, + "step": 12380 + }, + { + "epoch": 0.8117670182794995, + "grad_norm": 0.4633290767669678, + "learning_rate": 9.851474186785537e-05, + "loss": 0.0456, + "step": 12390 + }, + { + "epoch": 0.8124221974710083, + "grad_norm": 1.1667193174362183, + "learning_rate": 9.851073905070254e-05, + "loss": 0.0542, + "step": 12400 + }, + { + "epoch": 0.8130773766625172, + "grad_norm": 0.4073387682437897, + "learning_rate": 9.850673092848824e-05, + "loss": 0.0517, + "step": 12410 + }, + { + "epoch": 0.813732555854026, + "grad_norm": 0.35114917159080505, + "learning_rate": 9.850271750165077e-05, + "loss": 0.0324, + "step": 12420 + }, + { + "epoch": 0.814387735045535, + "grad_norm": 1.9997045993804932, + "learning_rate": 9.849869877062902e-05, + "loss": 0.0781, + "step": 12430 + }, + { + "epoch": 0.8150429142370438, + "grad_norm": 0.3553565442562103, + "learning_rate": 9.849467473586252e-05, + "loss": 0.0372, + "step": 12440 + }, + { + "epoch": 0.8156980934285527, + "grad_norm": 0.4593048393726349, + "learning_rate": 9.849064539779127e-05, + "loss": 0.0481, + "step": 12450 + }, + { + "epoch": 0.8163532726200616, + "grad_norm": 0.5446400046348572, + "learning_rate": 9.848661075685594e-05, + "loss": 0.0628, + "step": 12460 + }, + { + "epoch": 0.8170084518115704, + "grad_norm": 0.45974358916282654, + "learning_rate": 9.848257081349778e-05, + "loss": 0.0525, + "step": 12470 + }, + { + "epoch": 0.8176636310030794, + "grad_norm": 0.6194895505905151, + "learning_rate": 9.847852556815856e-05, + "loss": 0.0319, + "step": 12480 + }, + { + "epoch": 0.8183188101945882, + "grad_norm": 0.1989911049604416, + "learning_rate": 9.847447502128067e-05, + "loss": 0.0456, + "step": 12490 + }, + { + "epoch": 0.8189739893860971, + "grad_norm": 1.128719687461853, + "learning_rate": 9.847041917330708e-05, + "loss": 0.0532, + "step": 12500 + }, + { + "epoch": 0.8196291685776059, + "grad_norm": 0.9512436985969543, + "learning_rate": 9.846635802468132e-05, + "loss": 0.0754, + "step": 12510 + }, + { + "epoch": 0.8202843477691149, + "grad_norm": 0.7564404010772705, + "learning_rate": 9.84622915758475e-05, + "loss": 0.0537, + "step": 12520 + }, + { + "epoch": 0.8209395269606238, + "grad_norm": 1.7267171144485474, + "learning_rate": 9.845821982725034e-05, + "loss": 0.0817, + "step": 12530 + }, + { + "epoch": 0.8215947061521326, + "grad_norm": 0.43982747197151184, + "learning_rate": 9.845414277933514e-05, + "loss": 0.0555, + "step": 12540 + }, + { + "epoch": 0.8222498853436415, + "grad_norm": 0.8948685526847839, + "learning_rate": 9.845006043254771e-05, + "loss": 0.0685, + "step": 12550 + }, + { + "epoch": 0.8229050645351503, + "grad_norm": 1.3273180723190308, + "learning_rate": 9.844597278733451e-05, + "loss": 0.0749, + "step": 12560 + }, + { + "epoch": 0.8235602437266593, + "grad_norm": 0.76951003074646, + "learning_rate": 9.844187984414259e-05, + "loss": 0.0908, + "step": 12570 + }, + { + "epoch": 0.8242154229181681, + "grad_norm": 0.8103411793708801, + "learning_rate": 9.84377816034195e-05, + "loss": 0.0552, + "step": 12580 + }, + { + "epoch": 0.824870602109677, + "grad_norm": 0.6340231895446777, + "learning_rate": 9.843367806561345e-05, + "loss": 0.0561, + "step": 12590 + }, + { + "epoch": 0.8255257813011859, + "grad_norm": 0.5345150232315063, + "learning_rate": 9.842956923117317e-05, + "loss": 0.0462, + "step": 12600 + }, + { + "epoch": 0.8261809604926947, + "grad_norm": 0.8412685990333557, + "learning_rate": 9.842545510054802e-05, + "loss": 0.0519, + "step": 12610 + }, + { + "epoch": 0.8268361396842037, + "grad_norm": 0.5431321859359741, + "learning_rate": 9.842133567418792e-05, + "loss": 0.0534, + "step": 12620 + }, + { + "epoch": 0.8274913188757125, + "grad_norm": 0.34815967082977295, + "learning_rate": 9.841721095254333e-05, + "loss": 0.0675, + "step": 12630 + }, + { + "epoch": 0.8281464980672214, + "grad_norm": 0.3265923261642456, + "learning_rate": 9.841308093606537e-05, + "loss": 0.0392, + "step": 12640 + }, + { + "epoch": 0.8288016772587302, + "grad_norm": 0.3944855034351349, + "learning_rate": 9.840894562520565e-05, + "loss": 0.0354, + "step": 12650 + }, + { + "epoch": 0.8294568564502391, + "grad_norm": 1.2355951070785522, + "learning_rate": 9.840480502041642e-05, + "loss": 0.0496, + "step": 12660 + }, + { + "epoch": 0.8301120356417481, + "grad_norm": 0.8850134015083313, + "learning_rate": 9.840065912215049e-05, + "loss": 0.0798, + "step": 12670 + }, + { + "epoch": 0.8307672148332569, + "grad_norm": 0.7371264696121216, + "learning_rate": 9.839650793086124e-05, + "loss": 0.047, + "step": 12680 + }, + { + "epoch": 0.8314223940247658, + "grad_norm": 0.6307826042175293, + "learning_rate": 9.839235144700265e-05, + "loss": 0.084, + "step": 12690 + }, + { + "epoch": 0.8320775732162746, + "grad_norm": 0.38936373591423035, + "learning_rate": 9.838818967102926e-05, + "loss": 0.055, + "step": 12700 + }, + { + "epoch": 0.8327327524077835, + "grad_norm": 0.33244436979293823, + "learning_rate": 9.83840226033962e-05, + "loss": 0.0861, + "step": 12710 + }, + { + "epoch": 0.8333879315992924, + "grad_norm": 0.3175358474254608, + "learning_rate": 9.837985024455918e-05, + "loss": 0.0792, + "step": 12720 + }, + { + "epoch": 0.8340431107908013, + "grad_norm": 1.1068274974822998, + "learning_rate": 9.837567259497447e-05, + "loss": 0.0781, + "step": 12730 + }, + { + "epoch": 0.8346982899823102, + "grad_norm": 0.3622835576534271, + "learning_rate": 9.837148965509894e-05, + "loss": 0.0486, + "step": 12740 + }, + { + "epoch": 0.835353469173819, + "grad_norm": 0.48460379242897034, + "learning_rate": 9.836730142539001e-05, + "loss": 0.0894, + "step": 12750 + }, + { + "epoch": 0.836008648365328, + "grad_norm": 0.773766279220581, + "learning_rate": 9.836310790630574e-05, + "loss": 0.0621, + "step": 12760 + }, + { + "epoch": 0.8366638275568368, + "grad_norm": 0.5161145329475403, + "learning_rate": 9.83589090983047e-05, + "loss": 0.0355, + "step": 12770 + }, + { + "epoch": 0.8373190067483457, + "grad_norm": 0.7409093976020813, + "learning_rate": 9.835470500184605e-05, + "loss": 0.0472, + "step": 12780 + }, + { + "epoch": 0.8379741859398545, + "grad_norm": 1.0706168413162231, + "learning_rate": 9.835049561738957e-05, + "loss": 0.0546, + "step": 12790 + }, + { + "epoch": 0.8386293651313634, + "grad_norm": 0.49312958121299744, + "learning_rate": 9.834628094539558e-05, + "loss": 0.0484, + "step": 12800 + }, + { + "epoch": 0.8392845443228723, + "grad_norm": 0.3340025544166565, + "learning_rate": 9.834206098632499e-05, + "loss": 0.0722, + "step": 12810 + }, + { + "epoch": 0.8399397235143812, + "grad_norm": 0.5467402935028076, + "learning_rate": 9.833783574063931e-05, + "loss": 0.05, + "step": 12820 + }, + { + "epoch": 0.8405949027058901, + "grad_norm": 0.46424102783203125, + "learning_rate": 9.833360520880058e-05, + "loss": 0.0765, + "step": 12830 + }, + { + "epoch": 0.8412500818973989, + "grad_norm": 0.5830073356628418, + "learning_rate": 9.832936939127144e-05, + "loss": 0.0376, + "step": 12840 + }, + { + "epoch": 0.8419052610889078, + "grad_norm": 0.9205835461616516, + "learning_rate": 9.832512828851515e-05, + "loss": 0.0867, + "step": 12850 + }, + { + "epoch": 0.8425604402804167, + "grad_norm": 0.6714364290237427, + "learning_rate": 9.832088190099546e-05, + "loss": 0.0672, + "step": 12860 + }, + { + "epoch": 0.8432156194719256, + "grad_norm": 0.9595742225646973, + "learning_rate": 9.831663022917679e-05, + "loss": 0.1018, + "step": 12870 + }, + { + "epoch": 0.8438707986634344, + "grad_norm": 0.3786083459854126, + "learning_rate": 9.831237327352407e-05, + "loss": 0.0879, + "step": 12880 + }, + { + "epoch": 0.8445259778549433, + "grad_norm": 0.6756841540336609, + "learning_rate": 9.830811103450286e-05, + "loss": 0.0564, + "step": 12890 + }, + { + "epoch": 0.8451811570464522, + "grad_norm": 0.6869654059410095, + "learning_rate": 9.830384351257924e-05, + "loss": 0.0542, + "step": 12900 + }, + { + "epoch": 0.8458363362379611, + "grad_norm": 0.2952326536178589, + "learning_rate": 9.829957070821993e-05, + "loss": 0.0683, + "step": 12910 + }, + { + "epoch": 0.84649151542947, + "grad_norm": 1.5048037767410278, + "learning_rate": 9.829529262189218e-05, + "loss": 0.0422, + "step": 12920 + }, + { + "epoch": 0.8471466946209788, + "grad_norm": 0.33393165469169617, + "learning_rate": 9.829100925406385e-05, + "loss": 0.0616, + "step": 12930 + }, + { + "epoch": 0.8478018738124877, + "grad_norm": 1.0152533054351807, + "learning_rate": 9.828672060520333e-05, + "loss": 0.0973, + "step": 12940 + }, + { + "epoch": 0.8484570530039965, + "grad_norm": 0.4776148796081543, + "learning_rate": 9.828242667577966e-05, + "loss": 0.0479, + "step": 12950 + }, + { + "epoch": 0.8491122321955055, + "grad_norm": 0.6377302408218384, + "learning_rate": 9.82781274662624e-05, + "loss": 0.0752, + "step": 12960 + }, + { + "epoch": 0.8497674113870144, + "grad_norm": 0.534373939037323, + "learning_rate": 9.82738229771217e-05, + "loss": 0.0727, + "step": 12970 + }, + { + "epoch": 0.8504225905785232, + "grad_norm": 0.9538039565086365, + "learning_rate": 9.826951320882829e-05, + "loss": 0.0504, + "step": 12980 + }, + { + "epoch": 0.8510777697700321, + "grad_norm": 0.17428424954414368, + "learning_rate": 9.826519816185351e-05, + "loss": 0.0515, + "step": 12990 + }, + { + "epoch": 0.851732948961541, + "grad_norm": 0.5955262780189514, + "learning_rate": 9.826087783666921e-05, + "loss": 0.1177, + "step": 13000 + }, + { + "epoch": 0.8523881281530499, + "grad_norm": 0.44511905312538147, + "learning_rate": 9.825655223374787e-05, + "loss": 0.0785, + "step": 13010 + }, + { + "epoch": 0.8530433073445587, + "grad_norm": 0.30567848682403564, + "learning_rate": 9.825222135356253e-05, + "loss": 0.0611, + "step": 13020 + }, + { + "epoch": 0.8536984865360676, + "grad_norm": 0.7070454359054565, + "learning_rate": 9.82478851965868e-05, + "loss": 0.0388, + "step": 13030 + }, + { + "epoch": 0.8543536657275765, + "grad_norm": 0.8737286925315857, + "learning_rate": 9.82435437632949e-05, + "loss": 0.0823, + "step": 13040 + }, + { + "epoch": 0.8550088449190854, + "grad_norm": 1.0255030393600464, + "learning_rate": 9.823919705416158e-05, + "loss": 0.0664, + "step": 13050 + }, + { + "epoch": 0.8556640241105943, + "grad_norm": 0.5407454967498779, + "learning_rate": 9.82348450696622e-05, + "loss": 0.0695, + "step": 13060 + }, + { + "epoch": 0.8563192033021031, + "grad_norm": 1.280232548713684, + "learning_rate": 9.823048781027268e-05, + "loss": 0.0646, + "step": 13070 + }, + { + "epoch": 0.856974382493612, + "grad_norm": 0.25676456093788147, + "learning_rate": 9.822612527646953e-05, + "loss": 0.0985, + "step": 13080 + }, + { + "epoch": 0.8576295616851208, + "grad_norm": 0.439117968082428, + "learning_rate": 9.822175746872984e-05, + "loss": 0.0468, + "step": 13090 + }, + { + "epoch": 0.8582847408766298, + "grad_norm": 0.45098787546157837, + "learning_rate": 9.821738438753123e-05, + "loss": 0.0492, + "step": 13100 + }, + { + "epoch": 0.8589399200681387, + "grad_norm": 0.2909688651561737, + "learning_rate": 9.821300603335196e-05, + "loss": 0.0314, + "step": 13110 + }, + { + "epoch": 0.8595950992596475, + "grad_norm": 0.9544477462768555, + "learning_rate": 9.820862240667085e-05, + "loss": 0.0797, + "step": 13120 + }, + { + "epoch": 0.8602502784511564, + "grad_norm": 0.5244776606559753, + "learning_rate": 9.820423350796726e-05, + "loss": 0.0482, + "step": 13130 + }, + { + "epoch": 0.8609054576426652, + "grad_norm": 1.0914827585220337, + "learning_rate": 9.819983933772118e-05, + "loss": 0.0581, + "step": 13140 + }, + { + "epoch": 0.8615606368341742, + "grad_norm": 0.3136231005191803, + "learning_rate": 9.819543989641314e-05, + "loss": 0.0611, + "step": 13150 + }, + { + "epoch": 0.862215816025683, + "grad_norm": 0.9224333167076111, + "learning_rate": 9.819103518452423e-05, + "loss": 0.0576, + "step": 13160 + }, + { + "epoch": 0.8628709952171919, + "grad_norm": 1.573154330253601, + "learning_rate": 9.818662520253618e-05, + "loss": 0.0806, + "step": 13170 + }, + { + "epoch": 0.8635261744087008, + "grad_norm": 0.6558083295822144, + "learning_rate": 9.818220995093126e-05, + "loss": 0.0434, + "step": 13180 + }, + { + "epoch": 0.8641813536002096, + "grad_norm": 0.6741963624954224, + "learning_rate": 9.817778943019228e-05, + "loss": 0.049, + "step": 13190 + }, + { + "epoch": 0.8648365327917186, + "grad_norm": 0.4944705665111542, + "learning_rate": 9.81733636408027e-05, + "loss": 0.0352, + "step": 13200 + }, + { + "epoch": 0.8654917119832274, + "grad_norm": 1.4044557809829712, + "learning_rate": 9.816893258324649e-05, + "loss": 0.1063, + "step": 13210 + }, + { + "epoch": 0.8661468911747363, + "grad_norm": 0.6655785441398621, + "learning_rate": 9.816449625800823e-05, + "loss": 0.0703, + "step": 13220 + }, + { + "epoch": 0.8668020703662451, + "grad_norm": 0.5290679335594177, + "learning_rate": 9.816005466557308e-05, + "loss": 0.063, + "step": 13230 + }, + { + "epoch": 0.867457249557754, + "grad_norm": 0.7582171559333801, + "learning_rate": 9.815560780642674e-05, + "loss": 0.0454, + "step": 13240 + }, + { + "epoch": 0.868112428749263, + "grad_norm": 1.0689880847930908, + "learning_rate": 9.815115568105555e-05, + "loss": 0.0848, + "step": 13250 + }, + { + "epoch": 0.8687676079407718, + "grad_norm": 0.9697751402854919, + "learning_rate": 9.814669828994638e-05, + "loss": 0.0799, + "step": 13260 + }, + { + "epoch": 0.8694227871322807, + "grad_norm": 0.525002658367157, + "learning_rate": 9.814223563358665e-05, + "loss": 0.0692, + "step": 13270 + }, + { + "epoch": 0.8700779663237895, + "grad_norm": 0.7700520157814026, + "learning_rate": 9.813776771246443e-05, + "loss": 0.0683, + "step": 13280 + }, + { + "epoch": 0.8707331455152985, + "grad_norm": 0.5289952754974365, + "learning_rate": 9.813329452706829e-05, + "loss": 0.0717, + "step": 13290 + }, + { + "epoch": 0.8713883247068073, + "grad_norm": 0.69666588306427, + "learning_rate": 9.812881607788744e-05, + "loss": 0.0421, + "step": 13300 + }, + { + "epoch": 0.8720435038983162, + "grad_norm": 0.3858712613582611, + "learning_rate": 9.812433236541163e-05, + "loss": 0.0436, + "step": 13310 + }, + { + "epoch": 0.8726986830898251, + "grad_norm": 0.9216866493225098, + "learning_rate": 9.811984339013116e-05, + "loss": 0.0695, + "step": 13320 + }, + { + "epoch": 0.8733538622813339, + "grad_norm": 0.64073246717453, + "learning_rate": 9.811534915253698e-05, + "loss": 0.047, + "step": 13330 + }, + { + "epoch": 0.8740090414728429, + "grad_norm": 0.7014086842536926, + "learning_rate": 9.811084965312056e-05, + "loss": 0.0689, + "step": 13340 + }, + { + "epoch": 0.8746642206643517, + "grad_norm": 0.7429068684577942, + "learning_rate": 9.810634489237396e-05, + "loss": 0.0531, + "step": 13350 + }, + { + "epoch": 0.8753193998558606, + "grad_norm": 0.58442622423172, + "learning_rate": 9.81018348707898e-05, + "loss": 0.0497, + "step": 13360 + }, + { + "epoch": 0.8759745790473694, + "grad_norm": 0.7069791555404663, + "learning_rate": 9.809731958886131e-05, + "loss": 0.0747, + "step": 13370 + }, + { + "epoch": 0.8766297582388783, + "grad_norm": 0.24819274246692657, + "learning_rate": 9.809279904708224e-05, + "loss": 0.0602, + "step": 13380 + }, + { + "epoch": 0.8772849374303872, + "grad_norm": 0.33216947317123413, + "learning_rate": 9.808827324594699e-05, + "loss": 0.0295, + "step": 13390 + }, + { + "epoch": 0.8779401166218961, + "grad_norm": 0.3282032907009125, + "learning_rate": 9.808374218595046e-05, + "loss": 0.0662, + "step": 13400 + }, + { + "epoch": 0.878595295813405, + "grad_norm": 1.1391125917434692, + "learning_rate": 9.80792058675882e-05, + "loss": 0.0492, + "step": 13410 + }, + { + "epoch": 0.8792504750049138, + "grad_norm": 0.6271048784255981, + "learning_rate": 9.807466429135627e-05, + "loss": 0.053, + "step": 13420 + }, + { + "epoch": 0.8799056541964227, + "grad_norm": 1.2676204442977905, + "learning_rate": 9.807011745775132e-05, + "loss": 0.0393, + "step": 13430 + }, + { + "epoch": 0.8805608333879316, + "grad_norm": 0.6731472611427307, + "learning_rate": 9.806556536727061e-05, + "loss": 0.0755, + "step": 13440 + }, + { + "epoch": 0.8812160125794405, + "grad_norm": 0.5623183250427246, + "learning_rate": 9.806100802041193e-05, + "loss": 0.052, + "step": 13450 + }, + { + "epoch": 0.8818711917709493, + "grad_norm": 0.685060977935791, + "learning_rate": 9.805644541767368e-05, + "loss": 0.0435, + "step": 13460 + }, + { + "epoch": 0.8825263709624582, + "grad_norm": 1.4547805786132812, + "learning_rate": 9.805187755955478e-05, + "loss": 0.082, + "step": 13470 + }, + { + "epoch": 0.8831815501539672, + "grad_norm": 0.4079425036907196, + "learning_rate": 9.804730444655483e-05, + "loss": 0.0502, + "step": 13480 + }, + { + "epoch": 0.883836729345476, + "grad_norm": 1.6266140937805176, + "learning_rate": 9.804272607917388e-05, + "loss": 0.0949, + "step": 13490 + }, + { + "epoch": 0.8844919085369849, + "grad_norm": 0.6460483074188232, + "learning_rate": 9.803814245791265e-05, + "loss": 0.0708, + "step": 13500 + }, + { + "epoch": 0.8851470877284937, + "grad_norm": 1.380408763885498, + "learning_rate": 9.803355358327239e-05, + "loss": 0.1119, + "step": 13510 + }, + { + "epoch": 0.8858022669200026, + "grad_norm": 0.8618347644805908, + "learning_rate": 9.802895945575492e-05, + "loss": 0.0645, + "step": 13520 + }, + { + "epoch": 0.8864574461115114, + "grad_norm": 0.7020772695541382, + "learning_rate": 9.802436007586266e-05, + "loss": 0.0377, + "step": 13530 + }, + { + "epoch": 0.8871126253030204, + "grad_norm": 0.9587872624397278, + "learning_rate": 9.801975544409858e-05, + "loss": 0.0685, + "step": 13540 + }, + { + "epoch": 0.8877678044945293, + "grad_norm": 0.21202658116817474, + "learning_rate": 9.801514556096625e-05, + "loss": 0.0494, + "step": 13550 + }, + { + "epoch": 0.8884229836860381, + "grad_norm": 0.3799341917037964, + "learning_rate": 9.801053042696977e-05, + "loss": 0.0627, + "step": 13560 + }, + { + "epoch": 0.889078162877547, + "grad_norm": 1.960816502571106, + "learning_rate": 9.800591004261388e-05, + "loss": 0.0798, + "step": 13570 + }, + { + "epoch": 0.8897333420690559, + "grad_norm": 0.3215174078941345, + "learning_rate": 9.800128440840385e-05, + "loss": 0.0497, + "step": 13580 + }, + { + "epoch": 0.8903885212605648, + "grad_norm": 1.012431263923645, + "learning_rate": 9.799665352484552e-05, + "loss": 0.0629, + "step": 13590 + }, + { + "epoch": 0.8910437004520736, + "grad_norm": 1.190861463546753, + "learning_rate": 9.799201739244532e-05, + "loss": 0.0529, + "step": 13600 + }, + { + "epoch": 0.8916988796435825, + "grad_norm": 0.5402487516403198, + "learning_rate": 9.798737601171025e-05, + "loss": 0.0528, + "step": 13610 + }, + { + "epoch": 0.8923540588350914, + "grad_norm": 0.644109845161438, + "learning_rate": 9.79827293831479e-05, + "loss": 0.0812, + "step": 13620 + }, + { + "epoch": 0.8930092380266003, + "grad_norm": 0.5392053127288818, + "learning_rate": 9.797807750726638e-05, + "loss": 0.0521, + "step": 13630 + }, + { + "epoch": 0.8936644172181092, + "grad_norm": 1.463032841682434, + "learning_rate": 9.797342038457446e-05, + "loss": 0.0736, + "step": 13640 + }, + { + "epoch": 0.894319596409618, + "grad_norm": 0.26403629779815674, + "learning_rate": 9.796875801558141e-05, + "loss": 0.0414, + "step": 13650 + }, + { + "epoch": 0.8949747756011269, + "grad_norm": 0.48549842834472656, + "learning_rate": 9.79640904007971e-05, + "loss": 0.0446, + "step": 13660 + }, + { + "epoch": 0.8956299547926357, + "grad_norm": 0.30364108085632324, + "learning_rate": 9.795941754073199e-05, + "loss": 0.0512, + "step": 13670 + }, + { + "epoch": 0.8962851339841447, + "grad_norm": 0.49193698167800903, + "learning_rate": 9.795473943589705e-05, + "loss": 0.0615, + "step": 13680 + }, + { + "epoch": 0.8969403131756536, + "grad_norm": 1.4529949426651, + "learning_rate": 9.795005608680394e-05, + "loss": 0.0796, + "step": 13690 + }, + { + "epoch": 0.8975954923671624, + "grad_norm": 0.3160261809825897, + "learning_rate": 9.794536749396477e-05, + "loss": 0.054, + "step": 13700 + }, + { + "epoch": 0.8982506715586713, + "grad_norm": 0.21727754175662994, + "learning_rate": 9.79406736578923e-05, + "loss": 0.0678, + "step": 13710 + }, + { + "epoch": 0.8989058507501801, + "grad_norm": 0.8706192970275879, + "learning_rate": 9.793597457909984e-05, + "loss": 0.057, + "step": 13720 + }, + { + "epoch": 0.8995610299416891, + "grad_norm": 0.7375606894493103, + "learning_rate": 9.793127025810127e-05, + "loss": 0.0574, + "step": 13730 + }, + { + "epoch": 0.9002162091331979, + "grad_norm": 0.49535253643989563, + "learning_rate": 9.792656069541104e-05, + "loss": 0.0669, + "step": 13740 + }, + { + "epoch": 0.9008713883247068, + "grad_norm": 1.0181152820587158, + "learning_rate": 9.79218458915442e-05, + "loss": 0.0682, + "step": 13750 + }, + { + "epoch": 0.9015265675162157, + "grad_norm": 0.5514432191848755, + "learning_rate": 9.791712584701634e-05, + "loss": 0.0782, + "step": 13760 + }, + { + "epoch": 0.9021817467077246, + "grad_norm": 0.23553955554962158, + "learning_rate": 9.791240056234364e-05, + "loss": 0.0368, + "step": 13770 + }, + { + "epoch": 0.9028369258992335, + "grad_norm": 0.7723908424377441, + "learning_rate": 9.790767003804283e-05, + "loss": 0.055, + "step": 13780 + }, + { + "epoch": 0.9034921050907423, + "grad_norm": 0.383605420589447, + "learning_rate": 9.790293427463126e-05, + "loss": 0.0737, + "step": 13790 + }, + { + "epoch": 0.9041472842822512, + "grad_norm": 1.0306973457336426, + "learning_rate": 9.789819327262684e-05, + "loss": 0.0767, + "step": 13800 + }, + { + "epoch": 0.90480246347376, + "grad_norm": 1.8213235139846802, + "learning_rate": 9.7893447032548e-05, + "loss": 0.0679, + "step": 13810 + }, + { + "epoch": 0.905457642665269, + "grad_norm": 1.8191838264465332, + "learning_rate": 9.78886955549138e-05, + "loss": 0.0435, + "step": 13820 + }, + { + "epoch": 0.9061128218567779, + "grad_norm": 0.17842352390289307, + "learning_rate": 9.788393884024387e-05, + "loss": 0.0626, + "step": 13830 + }, + { + "epoch": 0.9067680010482867, + "grad_norm": 0.47060054540634155, + "learning_rate": 9.787917688905836e-05, + "loss": 0.0456, + "step": 13840 + }, + { + "epoch": 0.9074231802397956, + "grad_norm": 0.6013217568397522, + "learning_rate": 9.787440970187807e-05, + "loss": 0.0787, + "step": 13850 + }, + { + "epoch": 0.9080783594313044, + "grad_norm": 0.26338931918144226, + "learning_rate": 9.786963727922429e-05, + "loss": 0.0413, + "step": 13860 + }, + { + "epoch": 0.9087335386228134, + "grad_norm": 0.39435043931007385, + "learning_rate": 9.786485962161897e-05, + "loss": 0.062, + "step": 13870 + }, + { + "epoch": 0.9093887178143222, + "grad_norm": 0.5804308652877808, + "learning_rate": 9.786007672958455e-05, + "loss": 0.0737, + "step": 13880 + }, + { + "epoch": 0.9100438970058311, + "grad_norm": 0.8728767037391663, + "learning_rate": 9.78552886036441e-05, + "loss": 0.0642, + "step": 13890 + }, + { + "epoch": 0.9106990761973399, + "grad_norm": 1.0028715133666992, + "learning_rate": 9.785049524432124e-05, + "loss": 0.1065, + "step": 13900 + }, + { + "epoch": 0.9113542553888488, + "grad_norm": 0.5261638164520264, + "learning_rate": 9.784569665214016e-05, + "loss": 0.0656, + "step": 13910 + }, + { + "epoch": 0.9120094345803578, + "grad_norm": 0.7655388712882996, + "learning_rate": 9.784089282762563e-05, + "loss": 0.0403, + "step": 13920 + }, + { + "epoch": 0.9126646137718666, + "grad_norm": 1.1390801668167114, + "learning_rate": 9.7836083771303e-05, + "loss": 0.0613, + "step": 13930 + }, + { + "epoch": 0.9133197929633755, + "grad_norm": 0.4684610366821289, + "learning_rate": 9.783126948369817e-05, + "loss": 0.0485, + "step": 13940 + }, + { + "epoch": 0.9139749721548843, + "grad_norm": 0.957876980304718, + "learning_rate": 9.78264499653376e-05, + "loss": 0.0943, + "step": 13950 + }, + { + "epoch": 0.9146301513463932, + "grad_norm": 0.7017455697059631, + "learning_rate": 9.782162521674838e-05, + "loss": 0.0754, + "step": 13960 + }, + { + "epoch": 0.9152853305379021, + "grad_norm": 0.6837944984436035, + "learning_rate": 9.781679523845812e-05, + "loss": 0.0419, + "step": 13970 + }, + { + "epoch": 0.915940509729411, + "grad_norm": 0.592918872833252, + "learning_rate": 9.781196003099502e-05, + "loss": 0.0805, + "step": 13980 + }, + { + "epoch": 0.9165956889209199, + "grad_norm": 1.0513006448745728, + "learning_rate": 9.780711959488786e-05, + "loss": 0.0552, + "step": 13990 + }, + { + "epoch": 0.9172508681124287, + "grad_norm": 0.9420386552810669, + "learning_rate": 9.780227393066599e-05, + "loss": 0.0476, + "step": 14000 + }, + { + "epoch": 0.9179060473039377, + "grad_norm": 0.5085736513137817, + "learning_rate": 9.77974230388593e-05, + "loss": 0.0529, + "step": 14010 + }, + { + "epoch": 0.9185612264954465, + "grad_norm": 0.4151586890220642, + "learning_rate": 9.779256691999829e-05, + "loss": 0.0389, + "step": 14020 + }, + { + "epoch": 0.9192164056869554, + "grad_norm": 0.4591405391693115, + "learning_rate": 9.778770557461403e-05, + "loss": 0.0613, + "step": 14030 + }, + { + "epoch": 0.9198715848784642, + "grad_norm": 0.7750739455223083, + "learning_rate": 9.778283900323812e-05, + "loss": 0.0565, + "step": 14040 + }, + { + "epoch": 0.9205267640699731, + "grad_norm": 0.659959077835083, + "learning_rate": 9.777796720640277e-05, + "loss": 0.0419, + "step": 14050 + }, + { + "epoch": 0.9211819432614821, + "grad_norm": 0.3736690580844879, + "learning_rate": 9.777309018464078e-05, + "loss": 0.0644, + "step": 14060 + }, + { + "epoch": 0.9218371224529909, + "grad_norm": 0.7124815583229065, + "learning_rate": 9.776820793848547e-05, + "loss": 0.0733, + "step": 14070 + }, + { + "epoch": 0.9224923016444998, + "grad_norm": 0.825305700302124, + "learning_rate": 9.776332046847075e-05, + "loss": 0.0581, + "step": 14080 + }, + { + "epoch": 0.9231474808360086, + "grad_norm": 0.39884376525878906, + "learning_rate": 9.775842777513111e-05, + "loss": 0.0725, + "step": 14090 + }, + { + "epoch": 0.9238026600275175, + "grad_norm": 0.5714738368988037, + "learning_rate": 9.775352985900163e-05, + "loss": 0.0528, + "step": 14100 + }, + { + "epoch": 0.9244578392190264, + "grad_norm": 0.6809442043304443, + "learning_rate": 9.774862672061791e-05, + "loss": 0.0697, + "step": 14110 + }, + { + "epoch": 0.9251130184105353, + "grad_norm": 0.5488811731338501, + "learning_rate": 9.774371836051616e-05, + "loss": 0.0503, + "step": 14120 + }, + { + "epoch": 0.9257681976020442, + "grad_norm": 0.5339645147323608, + "learning_rate": 9.773880477923315e-05, + "loss": 0.0432, + "step": 14130 + }, + { + "epoch": 0.926423376793553, + "grad_norm": 0.6278372406959534, + "learning_rate": 9.773388597730623e-05, + "loss": 0.0631, + "step": 14140 + }, + { + "epoch": 0.927078555985062, + "grad_norm": 0.36219972372055054, + "learning_rate": 9.77289619552733e-05, + "loss": 0.0398, + "step": 14150 + }, + { + "epoch": 0.9277337351765708, + "grad_norm": 0.10526380687952042, + "learning_rate": 9.772403271367285e-05, + "loss": 0.0464, + "step": 14160 + }, + { + "epoch": 0.9283889143680797, + "grad_norm": 0.5503729581832886, + "learning_rate": 9.771909825304396e-05, + "loss": 0.0398, + "step": 14170 + }, + { + "epoch": 0.9290440935595885, + "grad_norm": 0.9139094948768616, + "learning_rate": 9.771415857392619e-05, + "loss": 0.061, + "step": 14180 + }, + { + "epoch": 0.9296992727510974, + "grad_norm": 0.33286499977111816, + "learning_rate": 9.770921367685978e-05, + "loss": 0.0366, + "step": 14190 + }, + { + "epoch": 0.9303544519426064, + "grad_norm": 0.9812914729118347, + "learning_rate": 9.770426356238551e-05, + "loss": 0.0523, + "step": 14200 + }, + { + "epoch": 0.9310096311341152, + "grad_norm": 0.8805528879165649, + "learning_rate": 9.769930823104469e-05, + "loss": 0.0267, + "step": 14210 + }, + { + "epoch": 0.9316648103256241, + "grad_norm": 1.0021806955337524, + "learning_rate": 9.769434768337926e-05, + "loss": 0.0766, + "step": 14220 + }, + { + "epoch": 0.9323199895171329, + "grad_norm": 0.8471100926399231, + "learning_rate": 9.768938191993164e-05, + "loss": 0.0475, + "step": 14230 + }, + { + "epoch": 0.9329751687086418, + "grad_norm": 0.34899139404296875, + "learning_rate": 9.768441094124494e-05, + "loss": 0.0479, + "step": 14240 + }, + { + "epoch": 0.9336303479001506, + "grad_norm": 0.39603540301322937, + "learning_rate": 9.767943474786275e-05, + "loss": 0.0447, + "step": 14250 + }, + { + "epoch": 0.9342855270916596, + "grad_norm": 0.6794832348823547, + "learning_rate": 9.767445334032923e-05, + "loss": 0.0704, + "step": 14260 + }, + { + "epoch": 0.9349407062831685, + "grad_norm": 0.3377482295036316, + "learning_rate": 9.766946671918919e-05, + "loss": 0.0512, + "step": 14270 + }, + { + "epoch": 0.9355958854746773, + "grad_norm": 0.6252317428588867, + "learning_rate": 9.766447488498796e-05, + "loss": 0.057, + "step": 14280 + }, + { + "epoch": 0.9362510646661862, + "grad_norm": 0.6153748035430908, + "learning_rate": 9.765947783827139e-05, + "loss": 0.0961, + "step": 14290 + }, + { + "epoch": 0.936906243857695, + "grad_norm": 1.231326937675476, + "learning_rate": 9.765447557958599e-05, + "loss": 0.0673, + "step": 14300 + }, + { + "epoch": 0.937561423049204, + "grad_norm": 1.1353164911270142, + "learning_rate": 9.764946810947879e-05, + "loss": 0.0686, + "step": 14310 + }, + { + "epoch": 0.9382166022407128, + "grad_norm": 1.0182372331619263, + "learning_rate": 9.764445542849738e-05, + "loss": 0.0614, + "step": 14320 + }, + { + "epoch": 0.9388717814322217, + "grad_norm": 1.6563359498977661, + "learning_rate": 9.763943753718998e-05, + "loss": 0.049, + "step": 14330 + }, + { + "epoch": 0.9395269606237306, + "grad_norm": 0.8329321146011353, + "learning_rate": 9.76344144361053e-05, + "loss": 0.1079, + "step": 14340 + }, + { + "epoch": 0.9401821398152395, + "grad_norm": 0.48746129870414734, + "learning_rate": 9.762938612579269e-05, + "loss": 0.0444, + "step": 14350 + }, + { + "epoch": 0.9408373190067484, + "grad_norm": 0.6642868518829346, + "learning_rate": 9.762435260680202e-05, + "loss": 0.1013, + "step": 14360 + }, + { + "epoch": 0.9414924981982572, + "grad_norm": 0.4791405498981476, + "learning_rate": 9.761931387968373e-05, + "loss": 0.0686, + "step": 14370 + }, + { + "epoch": 0.9421476773897661, + "grad_norm": 2.4330735206604004, + "learning_rate": 9.76142699449889e-05, + "loss": 0.0563, + "step": 14380 + }, + { + "epoch": 0.9428028565812749, + "grad_norm": 0.35079115629196167, + "learning_rate": 9.760922080326908e-05, + "loss": 0.0568, + "step": 14390 + }, + { + "epoch": 0.9434580357727839, + "grad_norm": 1.1228197813034058, + "learning_rate": 9.760416645507644e-05, + "loss": 0.0842, + "step": 14400 + }, + { + "epoch": 0.9441132149642927, + "grad_norm": 0.8639314770698547, + "learning_rate": 9.759910690096375e-05, + "loss": 0.0462, + "step": 14410 + }, + { + "epoch": 0.9447683941558016, + "grad_norm": 0.37461721897125244, + "learning_rate": 9.759404214148429e-05, + "loss": 0.0491, + "step": 14420 + }, + { + "epoch": 0.9454235733473105, + "grad_norm": 1.12396240234375, + "learning_rate": 9.758897217719191e-05, + "loss": 0.0792, + "step": 14430 + }, + { + "epoch": 0.9460787525388193, + "grad_norm": 0.4566383957862854, + "learning_rate": 9.758389700864113e-05, + "loss": 0.0587, + "step": 14440 + }, + { + "epoch": 0.9467339317303283, + "grad_norm": 0.5906858444213867, + "learning_rate": 9.757881663638688e-05, + "loss": 0.0672, + "step": 14450 + }, + { + "epoch": 0.9473891109218371, + "grad_norm": 0.9844226241111755, + "learning_rate": 9.757373106098478e-05, + "loss": 0.0606, + "step": 14460 + }, + { + "epoch": 0.948044290113346, + "grad_norm": 0.7120530605316162, + "learning_rate": 9.756864028299097e-05, + "loss": 0.0488, + "step": 14470 + }, + { + "epoch": 0.9486994693048548, + "grad_norm": 0.6574244499206543, + "learning_rate": 9.75635443029622e-05, + "loss": 0.0651, + "step": 14480 + }, + { + "epoch": 0.9493546484963638, + "grad_norm": 0.24904200434684753, + "learning_rate": 9.755844312145572e-05, + "loss": 0.0644, + "step": 14490 + }, + { + "epoch": 0.9500098276878727, + "grad_norm": 0.5605410933494568, + "learning_rate": 9.755333673902941e-05, + "loss": 0.0393, + "step": 14500 + }, + { + "epoch": 0.9506650068793815, + "grad_norm": 1.0432780981063843, + "learning_rate": 9.75482251562417e-05, + "loss": 0.0654, + "step": 14510 + }, + { + "epoch": 0.9513201860708904, + "grad_norm": 0.49794289469718933, + "learning_rate": 9.754310837365155e-05, + "loss": 0.0296, + "step": 14520 + }, + { + "epoch": 0.9519753652623992, + "grad_norm": 0.18592588603496552, + "learning_rate": 9.753798639181856e-05, + "loss": 0.055, + "step": 14530 + }, + { + "epoch": 0.9526305444539082, + "grad_norm": 0.7303996682167053, + "learning_rate": 9.753285921130286e-05, + "loss": 0.0507, + "step": 14540 + }, + { + "epoch": 0.953285723645417, + "grad_norm": 0.5292498469352722, + "learning_rate": 9.752772683266512e-05, + "loss": 0.0517, + "step": 14550 + }, + { + "epoch": 0.9539409028369259, + "grad_norm": 0.7551082968711853, + "learning_rate": 9.752258925646665e-05, + "loss": 0.0488, + "step": 14560 + }, + { + "epoch": 0.9545960820284348, + "grad_norm": 0.6818788647651672, + "learning_rate": 9.751744648326926e-05, + "loss": 0.035, + "step": 14570 + }, + { + "epoch": 0.9552512612199436, + "grad_norm": 0.47669270634651184, + "learning_rate": 9.751229851363536e-05, + "loss": 0.0619, + "step": 14580 + }, + { + "epoch": 0.9559064404114526, + "grad_norm": 0.264909565448761, + "learning_rate": 9.750714534812793e-05, + "loss": 0.0648, + "step": 14590 + }, + { + "epoch": 0.9565616196029614, + "grad_norm": 0.7237739562988281, + "learning_rate": 9.750198698731053e-05, + "loss": 0.0659, + "step": 14600 + }, + { + "epoch": 0.9572167987944703, + "grad_norm": 0.6081687211990356, + "learning_rate": 9.749682343174722e-05, + "loss": 0.06, + "step": 14610 + }, + { + "epoch": 0.9578719779859791, + "grad_norm": 0.3512479364871979, + "learning_rate": 9.749165468200272e-05, + "loss": 0.0555, + "step": 14620 + }, + { + "epoch": 0.958527157177488, + "grad_norm": 0.7961754202842712, + "learning_rate": 9.748648073864229e-05, + "loss": 0.0608, + "step": 14630 + }, + { + "epoch": 0.959182336368997, + "grad_norm": 0.5924041867256165, + "learning_rate": 9.748130160223168e-05, + "loss": 0.0435, + "step": 14640 + }, + { + "epoch": 0.9598375155605058, + "grad_norm": 0.17603300511837006, + "learning_rate": 9.747611727333734e-05, + "loss": 0.0587, + "step": 14650 + }, + { + "epoch": 0.9604926947520147, + "grad_norm": 0.8254342675209045, + "learning_rate": 9.74709277525262e-05, + "loss": 0.0662, + "step": 14660 + }, + { + "epoch": 0.9611478739435235, + "grad_norm": 0.3475823402404785, + "learning_rate": 9.746573304036576e-05, + "loss": 0.0582, + "step": 14670 + }, + { + "epoch": 0.9618030531350324, + "grad_norm": 0.5557885766029358, + "learning_rate": 9.746053313742412e-05, + "loss": 0.0709, + "step": 14680 + }, + { + "epoch": 0.9624582323265413, + "grad_norm": 0.16174621880054474, + "learning_rate": 9.745532804426994e-05, + "loss": 0.0513, + "step": 14690 + }, + { + "epoch": 0.9631134115180502, + "grad_norm": 1.4115376472473145, + "learning_rate": 9.745011776147242e-05, + "loss": 0.0712, + "step": 14700 + }, + { + "epoch": 0.9637685907095591, + "grad_norm": 1.0193588733673096, + "learning_rate": 9.744490228960138e-05, + "loss": 0.0631, + "step": 14710 + }, + { + "epoch": 0.9644237699010679, + "grad_norm": 0.6263638734817505, + "learning_rate": 9.743968162922713e-05, + "loss": 0.0451, + "step": 14720 + }, + { + "epoch": 0.9650789490925769, + "grad_norm": 0.8783289194107056, + "learning_rate": 9.743445578092064e-05, + "loss": 0.0547, + "step": 14730 + }, + { + "epoch": 0.9657341282840857, + "grad_norm": 0.37476468086242676, + "learning_rate": 9.742922474525338e-05, + "loss": 0.0728, + "step": 14740 + }, + { + "epoch": 0.9663893074755946, + "grad_norm": 0.7073484063148499, + "learning_rate": 9.742398852279741e-05, + "loss": 0.052, + "step": 14750 + }, + { + "epoch": 0.9670444866671034, + "grad_norm": 0.42330822348594666, + "learning_rate": 9.741874711412535e-05, + "loss": 0.0408, + "step": 14760 + }, + { + "epoch": 0.9676996658586123, + "grad_norm": 0.37546685338020325, + "learning_rate": 9.741350051981042e-05, + "loss": 0.0567, + "step": 14770 + }, + { + "epoch": 0.9683548450501213, + "grad_norm": 0.9165253639221191, + "learning_rate": 9.740824874042633e-05, + "loss": 0.0637, + "step": 14780 + }, + { + "epoch": 0.9690100242416301, + "grad_norm": 0.394199013710022, + "learning_rate": 9.740299177654746e-05, + "loss": 0.0516, + "step": 14790 + }, + { + "epoch": 0.969665203433139, + "grad_norm": 0.49190762639045715, + "learning_rate": 9.739772962874867e-05, + "loss": 0.047, + "step": 14800 + }, + { + "epoch": 0.9703203826246478, + "grad_norm": 0.9931155443191528, + "learning_rate": 9.739246229760541e-05, + "loss": 0.065, + "step": 14810 + }, + { + "epoch": 0.9709755618161567, + "grad_norm": 0.4841979742050171, + "learning_rate": 9.738718978369376e-05, + "loss": 0.0705, + "step": 14820 + }, + { + "epoch": 0.9716307410076656, + "grad_norm": 0.3157121539115906, + "learning_rate": 9.738191208759025e-05, + "loss": 0.0349, + "step": 14830 + }, + { + "epoch": 0.9722859201991745, + "grad_norm": 1.076965570449829, + "learning_rate": 9.73766292098721e-05, + "loss": 0.0498, + "step": 14840 + }, + { + "epoch": 0.9729410993906834, + "grad_norm": 0.2656017541885376, + "learning_rate": 9.737134115111699e-05, + "loss": 0.072, + "step": 14850 + }, + { + "epoch": 0.9735962785821922, + "grad_norm": 0.41680657863616943, + "learning_rate": 9.736604791190323e-05, + "loss": 0.0656, + "step": 14860 + }, + { + "epoch": 0.9742514577737011, + "grad_norm": 0.6578284502029419, + "learning_rate": 9.73607494928097e-05, + "loss": 0.0638, + "step": 14870 + }, + { + "epoch": 0.97490663696521, + "grad_norm": 0.5628801584243774, + "learning_rate": 9.735544589441581e-05, + "loss": 0.0431, + "step": 14880 + }, + { + "epoch": 0.9755618161567189, + "grad_norm": 0.9896408915519714, + "learning_rate": 9.735013711730154e-05, + "loss": 0.0749, + "step": 14890 + }, + { + "epoch": 0.9762169953482277, + "grad_norm": 0.24205443263053894, + "learning_rate": 9.734482316204747e-05, + "loss": 0.0538, + "step": 14900 + }, + { + "epoch": 0.9768721745397366, + "grad_norm": 0.3022863268852234, + "learning_rate": 9.733950402923473e-05, + "loss": 0.0443, + "step": 14910 + }, + { + "epoch": 0.9775273537312454, + "grad_norm": 0.5883574485778809, + "learning_rate": 9.7334179719445e-05, + "loss": 0.0462, + "step": 14920 + }, + { + "epoch": 0.9781825329227544, + "grad_norm": 1.2118014097213745, + "learning_rate": 9.732885023326053e-05, + "loss": 0.0624, + "step": 14930 + }, + { + "epoch": 0.9788377121142633, + "grad_norm": 1.047371506690979, + "learning_rate": 9.732351557126418e-05, + "loss": 0.0395, + "step": 14940 + }, + { + "epoch": 0.9794928913057721, + "grad_norm": 1.101252794265747, + "learning_rate": 9.731817573403929e-05, + "loss": 0.0453, + "step": 14950 + }, + { + "epoch": 0.980148070497281, + "grad_norm": 0.9259678721427917, + "learning_rate": 9.731283072216985e-05, + "loss": 0.0636, + "step": 14960 + }, + { + "epoch": 0.9808032496887898, + "grad_norm": 0.9893555641174316, + "learning_rate": 9.730748053624039e-05, + "loss": 0.0609, + "step": 14970 + }, + { + "epoch": 0.9814584288802988, + "grad_norm": 0.7126653790473938, + "learning_rate": 9.730212517683598e-05, + "loss": 0.0428, + "step": 14980 + }, + { + "epoch": 0.9821136080718076, + "grad_norm": 1.106996774673462, + "learning_rate": 9.729676464454228e-05, + "loss": 0.0791, + "step": 14990 + }, + { + "epoch": 0.9827687872633165, + "grad_norm": 0.45108863711357117, + "learning_rate": 9.72913989399455e-05, + "loss": 0.0499, + "step": 15000 + }, + { + "epoch": 0.9834239664548254, + "grad_norm": 0.7780125737190247, + "learning_rate": 9.728602806363242e-05, + "loss": 0.0487, + "step": 15010 + }, + { + "epoch": 0.9840791456463343, + "grad_norm": 0.4825277030467987, + "learning_rate": 9.728065201619043e-05, + "loss": 0.048, + "step": 15020 + }, + { + "epoch": 0.9847343248378432, + "grad_norm": 0.6079219579696655, + "learning_rate": 9.727527079820742e-05, + "loss": 0.0864, + "step": 15030 + }, + { + "epoch": 0.985389504029352, + "grad_norm": 0.23465560376644135, + "learning_rate": 9.726988441027186e-05, + "loss": 0.0407, + "step": 15040 + }, + { + "epoch": 0.9860446832208609, + "grad_norm": 0.9421533942222595, + "learning_rate": 9.726449285297281e-05, + "loss": 0.0511, + "step": 15050 + }, + { + "epoch": 0.9866998624123697, + "grad_norm": 0.7679508328437805, + "learning_rate": 9.72590961268999e-05, + "loss": 0.0621, + "step": 15060 + }, + { + "epoch": 0.9873550416038787, + "grad_norm": 0.7024896740913391, + "learning_rate": 9.725369423264328e-05, + "loss": 0.0641, + "step": 15070 + }, + { + "epoch": 0.9880102207953876, + "grad_norm": 0.9180717468261719, + "learning_rate": 9.72482871707937e-05, + "loss": 0.0884, + "step": 15080 + }, + { + "epoch": 0.9886653999868964, + "grad_norm": 0.5065802931785583, + "learning_rate": 9.724287494194247e-05, + "loss": 0.073, + "step": 15090 + }, + { + "epoch": 0.9893205791784053, + "grad_norm": 0.7572296261787415, + "learning_rate": 9.723745754668147e-05, + "loss": 0.0693, + "step": 15100 + }, + { + "epoch": 0.9899757583699141, + "grad_norm": 0.21203699707984924, + "learning_rate": 9.723203498560313e-05, + "loss": 0.0793, + "step": 15110 + }, + { + "epoch": 0.9906309375614231, + "grad_norm": 0.5227267742156982, + "learning_rate": 9.722660725930046e-05, + "loss": 0.0671, + "step": 15120 + }, + { + "epoch": 0.9912861167529319, + "grad_norm": 0.27316853404045105, + "learning_rate": 9.722117436836702e-05, + "loss": 0.0687, + "step": 15130 + }, + { + "epoch": 0.9919412959444408, + "grad_norm": 0.28959622979164124, + "learning_rate": 9.721573631339696e-05, + "loss": 0.0373, + "step": 15140 + }, + { + "epoch": 0.9925964751359497, + "grad_norm": 0.4155622124671936, + "learning_rate": 9.721029309498494e-05, + "loss": 0.0525, + "step": 15150 + }, + { + "epoch": 0.9932516543274585, + "grad_norm": 0.33131536841392517, + "learning_rate": 9.720484471372627e-05, + "loss": 0.043, + "step": 15160 + }, + { + "epoch": 0.9939068335189675, + "grad_norm": 0.09261411428451538, + "learning_rate": 9.719939117021673e-05, + "loss": 0.0987, + "step": 15170 + }, + { + "epoch": 0.9945620127104763, + "grad_norm": 0.9157260656356812, + "learning_rate": 9.719393246505275e-05, + "loss": 0.046, + "step": 15180 + }, + { + "epoch": 0.9952171919019852, + "grad_norm": 0.5514065027236938, + "learning_rate": 9.718846859883128e-05, + "loss": 0.0533, + "step": 15190 + }, + { + "epoch": 0.995872371093494, + "grad_norm": 0.25650349259376526, + "learning_rate": 9.718299957214982e-05, + "loss": 0.0754, + "step": 15200 + }, + { + "epoch": 0.996527550285003, + "grad_norm": 0.46071678400039673, + "learning_rate": 9.717752538560646e-05, + "loss": 0.0702, + "step": 15210 + }, + { + "epoch": 0.9971827294765119, + "grad_norm": 0.7096025347709656, + "learning_rate": 9.717204603979986e-05, + "loss": 0.0694, + "step": 15220 + }, + { + "epoch": 0.9978379086680207, + "grad_norm": 0.8043963313102722, + "learning_rate": 9.716656153532922e-05, + "loss": 0.0384, + "step": 15230 + }, + { + "epoch": 0.9984930878595296, + "grad_norm": 0.4492166340351105, + "learning_rate": 9.716107187279434e-05, + "loss": 0.0279, + "step": 15240 + }, + { + "epoch": 0.9991482670510384, + "grad_norm": 0.6637885570526123, + "learning_rate": 9.715557705279555e-05, + "loss": 0.0606, + "step": 15250 + }, + { + "epoch": 0.9998034462425474, + "grad_norm": 0.5001099109649658, + "learning_rate": 9.715007707593372e-05, + "loss": 0.0415, + "step": 15260 + }, + { + "epoch": 1.0004586254340562, + "grad_norm": 0.1925588846206665, + "learning_rate": 9.714457194281036e-05, + "loss": 0.0566, + "step": 15270 + }, + { + "epoch": 1.001113804625565, + "grad_norm": 0.5997272729873657, + "learning_rate": 9.713906165402751e-05, + "loss": 0.0559, + "step": 15280 + }, + { + "epoch": 1.001768983817074, + "grad_norm": 0.5595455169677734, + "learning_rate": 9.713354621018774e-05, + "loss": 0.0382, + "step": 15290 + }, + { + "epoch": 1.0024241630085828, + "grad_norm": 0.3343864381313324, + "learning_rate": 9.712802561189422e-05, + "loss": 0.0556, + "step": 15300 + }, + { + "epoch": 1.0030793422000917, + "grad_norm": 0.7999988794326782, + "learning_rate": 9.712249985975069e-05, + "loss": 0.0343, + "step": 15310 + }, + { + "epoch": 1.0037345213916007, + "grad_norm": 1.5372861623764038, + "learning_rate": 9.71169689543614e-05, + "loss": 0.0653, + "step": 15320 + }, + { + "epoch": 1.0043897005831095, + "grad_norm": 0.5015921592712402, + "learning_rate": 9.711143289633123e-05, + "loss": 0.0487, + "step": 15330 + }, + { + "epoch": 1.0050448797746183, + "grad_norm": 0.4378429353237152, + "learning_rate": 9.710589168626561e-05, + "loss": 0.0459, + "step": 15340 + }, + { + "epoch": 1.0057000589661271, + "grad_norm": 0.6718723773956299, + "learning_rate": 9.710034532477048e-05, + "loss": 0.0444, + "step": 15350 + }, + { + "epoch": 1.0063552381576362, + "grad_norm": 1.0090510845184326, + "learning_rate": 9.709479381245239e-05, + "loss": 0.0525, + "step": 15360 + }, + { + "epoch": 1.007010417349145, + "grad_norm": 0.3842965364456177, + "learning_rate": 9.708923714991847e-05, + "loss": 0.0689, + "step": 15370 + }, + { + "epoch": 1.0076655965406538, + "grad_norm": 1.0429259538650513, + "learning_rate": 9.708367533777638e-05, + "loss": 0.0637, + "step": 15380 + }, + { + "epoch": 1.0083207757321628, + "grad_norm": 0.9014490246772766, + "learning_rate": 9.707810837663431e-05, + "loss": 0.059, + "step": 15390 + }, + { + "epoch": 1.0089759549236716, + "grad_norm": 0.6956924200057983, + "learning_rate": 9.707253626710113e-05, + "loss": 0.0668, + "step": 15400 + }, + { + "epoch": 1.0096311341151805, + "grad_norm": 0.6650571823120117, + "learning_rate": 9.706695900978613e-05, + "loss": 0.064, + "step": 15410 + }, + { + "epoch": 1.0102863133066893, + "grad_norm": 0.39251238107681274, + "learning_rate": 9.706137660529926e-05, + "loss": 0.0409, + "step": 15420 + }, + { + "epoch": 1.0109414924981983, + "grad_norm": 0.3231596052646637, + "learning_rate": 9.705578905425101e-05, + "loss": 0.0312, + "step": 15430 + }, + { + "epoch": 1.0115966716897071, + "grad_norm": 0.4595090448856354, + "learning_rate": 9.705019635725241e-05, + "loss": 0.0485, + "step": 15440 + }, + { + "epoch": 1.012251850881216, + "grad_norm": 0.4490564167499542, + "learning_rate": 9.704459851491508e-05, + "loss": 0.0645, + "step": 15450 + }, + { + "epoch": 1.012907030072725, + "grad_norm": 0.8977745175361633, + "learning_rate": 9.703899552785118e-05, + "loss": 0.0673, + "step": 15460 + }, + { + "epoch": 1.0135622092642338, + "grad_norm": 0.5210543870925903, + "learning_rate": 9.703338739667346e-05, + "loss": 0.0648, + "step": 15470 + }, + { + "epoch": 1.0142173884557426, + "grad_norm": 0.8921528458595276, + "learning_rate": 9.70277741219952e-05, + "loss": 0.065, + "step": 15480 + }, + { + "epoch": 1.0148725676472514, + "grad_norm": 0.15063409507274628, + "learning_rate": 9.702215570443027e-05, + "loss": 0.0643, + "step": 15490 + }, + { + "epoch": 1.0155277468387605, + "grad_norm": 1.2168816328048706, + "learning_rate": 9.701653214459309e-05, + "loss": 0.0569, + "step": 15500 + }, + { + "epoch": 1.0161829260302693, + "grad_norm": 0.6239476203918457, + "learning_rate": 9.701090344309865e-05, + "loss": 0.0264, + "step": 15510 + }, + { + "epoch": 1.016838105221778, + "grad_norm": 0.9249889850616455, + "learning_rate": 9.700526960056247e-05, + "loss": 0.0454, + "step": 15520 + }, + { + "epoch": 1.0174932844132871, + "grad_norm": 0.4423693120479584, + "learning_rate": 9.699963061760068e-05, + "loss": 0.0603, + "step": 15530 + }, + { + "epoch": 1.018148463604796, + "grad_norm": 0.9155156016349792, + "learning_rate": 9.699398649482997e-05, + "loss": 0.0684, + "step": 15540 + }, + { + "epoch": 1.0188036427963048, + "grad_norm": 0.4504965543746948, + "learning_rate": 9.698833723286753e-05, + "loss": 0.0373, + "step": 15550 + }, + { + "epoch": 1.0194588219878136, + "grad_norm": 0.5500404238700867, + "learning_rate": 9.698268283233118e-05, + "loss": 0.0702, + "step": 15560 + }, + { + "epoch": 1.0201140011793226, + "grad_norm": 0.5919736623764038, + "learning_rate": 9.697702329383929e-05, + "loss": 0.0653, + "step": 15570 + }, + { + "epoch": 1.0207691803708314, + "grad_norm": 0.49739792943000793, + "learning_rate": 9.697135861801074e-05, + "loss": 0.0465, + "step": 15580 + }, + { + "epoch": 1.0214243595623402, + "grad_norm": 0.5352333188056946, + "learning_rate": 9.696568880546505e-05, + "loss": 0.0496, + "step": 15590 + }, + { + "epoch": 1.0220795387538493, + "grad_norm": 0.23843690752983093, + "learning_rate": 9.696001385682223e-05, + "loss": 0.0377, + "step": 15600 + }, + { + "epoch": 1.022734717945358, + "grad_norm": 0.884663462638855, + "learning_rate": 9.695433377270291e-05, + "loss": 0.0511, + "step": 15610 + }, + { + "epoch": 1.023389897136867, + "grad_norm": 0.44047537446022034, + "learning_rate": 9.694864855372824e-05, + "loss": 0.038, + "step": 15620 + }, + { + "epoch": 1.0240450763283757, + "grad_norm": 0.9589326977729797, + "learning_rate": 9.694295820051995e-05, + "loss": 0.062, + "step": 15630 + }, + { + "epoch": 1.0247002555198848, + "grad_norm": 0.4542221128940582, + "learning_rate": 9.693726271370032e-05, + "loss": 0.0393, + "step": 15640 + }, + { + "epoch": 1.0253554347113936, + "grad_norm": 0.5850716233253479, + "learning_rate": 9.693156209389221e-05, + "loss": 0.0317, + "step": 15650 + }, + { + "epoch": 1.0260106139029024, + "grad_norm": 0.29659533500671387, + "learning_rate": 9.692585634171905e-05, + "loss": 0.0566, + "step": 15660 + }, + { + "epoch": 1.0266657930944114, + "grad_norm": 0.2918514907360077, + "learning_rate": 9.692014545780476e-05, + "loss": 0.0367, + "step": 15670 + }, + { + "epoch": 1.0273209722859202, + "grad_norm": 1.0493500232696533, + "learning_rate": 9.691442944277393e-05, + "loss": 0.0716, + "step": 15680 + }, + { + "epoch": 1.027976151477429, + "grad_norm": 0.39490899443626404, + "learning_rate": 9.690870829725162e-05, + "loss": 0.0478, + "step": 15690 + }, + { + "epoch": 1.0286313306689379, + "grad_norm": 0.7581951022148132, + "learning_rate": 9.69029820218635e-05, + "loss": 0.0725, + "step": 15700 + }, + { + "epoch": 1.029286509860447, + "grad_norm": 0.9907164573669434, + "learning_rate": 9.689725061723579e-05, + "loss": 0.0772, + "step": 15710 + }, + { + "epoch": 1.0299416890519557, + "grad_norm": 0.35023191571235657, + "learning_rate": 9.689151408399527e-05, + "loss": 0.0447, + "step": 15720 + }, + { + "epoch": 1.0305968682434645, + "grad_norm": 0.23471756279468536, + "learning_rate": 9.688577242276924e-05, + "loss": 0.0546, + "step": 15730 + }, + { + "epoch": 1.0312520474349736, + "grad_norm": 0.9907585382461548, + "learning_rate": 9.688002563418566e-05, + "loss": 0.0589, + "step": 15740 + }, + { + "epoch": 1.0319072266264824, + "grad_norm": 0.7399505972862244, + "learning_rate": 9.687427371887293e-05, + "loss": 0.0566, + "step": 15750 + }, + { + "epoch": 1.0325624058179912, + "grad_norm": 0.7956912517547607, + "learning_rate": 9.686851667746012e-05, + "loss": 0.0312, + "step": 15760 + }, + { + "epoch": 1.0332175850095, + "grad_norm": 0.6377907991409302, + "learning_rate": 9.686275451057677e-05, + "loss": 0.0692, + "step": 15770 + }, + { + "epoch": 1.033872764201009, + "grad_norm": 0.3312484622001648, + "learning_rate": 9.685698721885308e-05, + "loss": 0.0426, + "step": 15780 + }, + { + "epoch": 1.0345279433925179, + "grad_norm": 0.6556704044342041, + "learning_rate": 9.68512148029197e-05, + "loss": 0.0387, + "step": 15790 + }, + { + "epoch": 1.0351831225840267, + "grad_norm": 0.27309173345565796, + "learning_rate": 9.684543726340791e-05, + "loss": 0.0638, + "step": 15800 + }, + { + "epoch": 1.0358383017755357, + "grad_norm": 0.43227750062942505, + "learning_rate": 9.683965460094952e-05, + "loss": 0.1201, + "step": 15810 + }, + { + "epoch": 1.0364934809670445, + "grad_norm": 0.2523287236690521, + "learning_rate": 9.683386681617694e-05, + "loss": 0.0559, + "step": 15820 + }, + { + "epoch": 1.0371486601585533, + "grad_norm": 0.9118742346763611, + "learning_rate": 9.68280739097231e-05, + "loss": 0.0521, + "step": 15830 + }, + { + "epoch": 1.0378038393500622, + "grad_norm": 0.6253945827484131, + "learning_rate": 9.682227588222148e-05, + "loss": 0.0319, + "step": 15840 + }, + { + "epoch": 1.0384590185415712, + "grad_norm": 1.012669324874878, + "learning_rate": 9.681647273430618e-05, + "loss": 0.0798, + "step": 15850 + }, + { + "epoch": 1.03911419773308, + "grad_norm": 0.8481881618499756, + "learning_rate": 9.681066446661182e-05, + "loss": 0.0591, + "step": 15860 + }, + { + "epoch": 1.0397693769245888, + "grad_norm": 0.3410697281360626, + "learning_rate": 9.680485107977357e-05, + "loss": 0.0525, + "step": 15870 + }, + { + "epoch": 1.0404245561160979, + "grad_norm": 0.4943464994430542, + "learning_rate": 9.679903257442716e-05, + "loss": 0.0448, + "step": 15880 + }, + { + "epoch": 1.0410797353076067, + "grad_norm": 1.5602381229400635, + "learning_rate": 9.679320895120891e-05, + "loss": 0.0724, + "step": 15890 + }, + { + "epoch": 1.0417349144991155, + "grad_norm": 0.31514328718185425, + "learning_rate": 9.67873802107557e-05, + "loss": 0.062, + "step": 15900 + }, + { + "epoch": 1.0423900936906243, + "grad_norm": 0.4788703918457031, + "learning_rate": 9.67815463537049e-05, + "loss": 0.0478, + "step": 15910 + }, + { + "epoch": 1.0430452728821333, + "grad_norm": 0.3865969777107239, + "learning_rate": 9.677570738069457e-05, + "loss": 0.0559, + "step": 15920 + }, + { + "epoch": 1.0437004520736421, + "grad_norm": 0.41647428274154663, + "learning_rate": 9.676986329236318e-05, + "loss": 0.0608, + "step": 15930 + }, + { + "epoch": 1.044355631265151, + "grad_norm": 0.748646080493927, + "learning_rate": 9.676401408934987e-05, + "loss": 0.1037, + "step": 15940 + }, + { + "epoch": 1.04501081045666, + "grad_norm": 1.0008667707443237, + "learning_rate": 9.675815977229428e-05, + "loss": 0.047, + "step": 15950 + }, + { + "epoch": 1.0456659896481688, + "grad_norm": 0.9880648851394653, + "learning_rate": 9.675230034183664e-05, + "loss": 0.0589, + "step": 15960 + }, + { + "epoch": 1.0463211688396776, + "grad_norm": 0.30245906114578247, + "learning_rate": 9.674643579861773e-05, + "loss": 0.0447, + "step": 15970 + }, + { + "epoch": 1.0469763480311864, + "grad_norm": 0.9661765098571777, + "learning_rate": 9.674056614327886e-05, + "loss": 0.0531, + "step": 15980 + }, + { + "epoch": 1.0476315272226955, + "grad_norm": 0.6142928600311279, + "learning_rate": 9.673469137646198e-05, + "loss": 0.0321, + "step": 15990 + }, + { + "epoch": 1.0482867064142043, + "grad_norm": 0.14501795172691345, + "learning_rate": 9.67288114988095e-05, + "loss": 0.0656, + "step": 16000 + }, + { + "epoch": 1.048941885605713, + "grad_norm": 0.6470243334770203, + "learning_rate": 9.672292651096447e-05, + "loss": 0.0444, + "step": 16010 + }, + { + "epoch": 1.0495970647972221, + "grad_norm": 0.22925227880477905, + "learning_rate": 9.671703641357042e-05, + "loss": 0.0416, + "step": 16020 + }, + { + "epoch": 1.050252243988731, + "grad_norm": 0.8200265169143677, + "learning_rate": 9.67111412072715e-05, + "loss": 0.0612, + "step": 16030 + }, + { + "epoch": 1.0509074231802398, + "grad_norm": 0.647597074508667, + "learning_rate": 9.670524089271242e-05, + "loss": 0.0566, + "step": 16040 + }, + { + "epoch": 1.0515626023717486, + "grad_norm": 2.147754192352295, + "learning_rate": 9.669933547053842e-05, + "loss": 0.0443, + "step": 16050 + }, + { + "epoch": 1.0522177815632576, + "grad_norm": 2.5494678020477295, + "learning_rate": 9.669342494139531e-05, + "loss": 0.1089, + "step": 16060 + }, + { + "epoch": 1.0528729607547664, + "grad_norm": 1.2066696882247925, + "learning_rate": 9.668750930592943e-05, + "loss": 0.0618, + "step": 16070 + }, + { + "epoch": 1.0535281399462753, + "grad_norm": 0.35088837146759033, + "learning_rate": 9.668158856478775e-05, + "loss": 0.0619, + "step": 16080 + }, + { + "epoch": 1.0541833191377843, + "grad_norm": 0.41922107338905334, + "learning_rate": 9.66756627186177e-05, + "loss": 0.067, + "step": 16090 + }, + { + "epoch": 1.054838498329293, + "grad_norm": 1.291082501411438, + "learning_rate": 9.666973176806737e-05, + "loss": 0.0443, + "step": 16100 + }, + { + "epoch": 1.055493677520802, + "grad_norm": 0.4770950973033905, + "learning_rate": 9.666379571378534e-05, + "loss": 0.1079, + "step": 16110 + }, + { + "epoch": 1.0561488567123107, + "grad_norm": 0.4131571650505066, + "learning_rate": 9.665785455642076e-05, + "loss": 0.0514, + "step": 16120 + }, + { + "epoch": 1.0568040359038198, + "grad_norm": 0.4468565285205841, + "learning_rate": 9.665190829662337e-05, + "loss": 0.027, + "step": 16130 + }, + { + "epoch": 1.0574592150953286, + "grad_norm": 0.1926049143075943, + "learning_rate": 9.664595693504342e-05, + "loss": 0.0312, + "step": 16140 + }, + { + "epoch": 1.0581143942868374, + "grad_norm": 0.6884435415267944, + "learning_rate": 9.664000047233175e-05, + "loss": 0.0611, + "step": 16150 + }, + { + "epoch": 1.0587695734783464, + "grad_norm": 0.38569605350494385, + "learning_rate": 9.663403890913976e-05, + "loss": 0.043, + "step": 16160 + }, + { + "epoch": 1.0594247526698553, + "grad_norm": 0.46383774280548096, + "learning_rate": 9.662807224611938e-05, + "loss": 0.0674, + "step": 16170 + }, + { + "epoch": 1.060079931861364, + "grad_norm": 0.351988285779953, + "learning_rate": 9.662210048392311e-05, + "loss": 0.0498, + "step": 16180 + }, + { + "epoch": 1.0607351110528729, + "grad_norm": 0.3224245607852936, + "learning_rate": 9.661612362320405e-05, + "loss": 0.0884, + "step": 16190 + }, + { + "epoch": 1.061390290244382, + "grad_norm": 1.7253715991973877, + "learning_rate": 9.661014166461579e-05, + "loss": 0.0655, + "step": 16200 + }, + { + "epoch": 1.0620454694358907, + "grad_norm": 0.5329586267471313, + "learning_rate": 9.66041546088125e-05, + "loss": 0.0799, + "step": 16210 + }, + { + "epoch": 1.0627006486273995, + "grad_norm": 0.5441992878913879, + "learning_rate": 9.659816245644895e-05, + "loss": 0.0493, + "step": 16220 + }, + { + "epoch": 1.0633558278189086, + "grad_norm": 0.683830738067627, + "learning_rate": 9.65921652081804e-05, + "loss": 0.0667, + "step": 16230 + }, + { + "epoch": 1.0640110070104174, + "grad_norm": 0.442516565322876, + "learning_rate": 9.658616286466271e-05, + "loss": 0.0365, + "step": 16240 + }, + { + "epoch": 1.0646661862019262, + "grad_norm": 0.26811063289642334, + "learning_rate": 9.65801554265523e-05, + "loss": 0.0496, + "step": 16250 + }, + { + "epoch": 1.065321365393435, + "grad_norm": 0.5877781510353088, + "learning_rate": 9.657414289450612e-05, + "loss": 0.0545, + "step": 16260 + }, + { + "epoch": 1.065976544584944, + "grad_norm": 0.441048264503479, + "learning_rate": 9.656812526918171e-05, + "loss": 0.0481, + "step": 16270 + }, + { + "epoch": 1.0666317237764529, + "grad_norm": 1.1812959909439087, + "learning_rate": 9.656210255123712e-05, + "loss": 0.05, + "step": 16280 + }, + { + "epoch": 1.0672869029679617, + "grad_norm": 0.8458932638168335, + "learning_rate": 9.6556074741331e-05, + "loss": 0.0582, + "step": 16290 + }, + { + "epoch": 1.0679420821594707, + "grad_norm": 0.5652876496315002, + "learning_rate": 9.655004184012256e-05, + "loss": 0.0501, + "step": 16300 + }, + { + "epoch": 1.0685972613509795, + "grad_norm": 0.760123610496521, + "learning_rate": 9.654400384827152e-05, + "loss": 0.0496, + "step": 16310 + }, + { + "epoch": 1.0692524405424884, + "grad_norm": 0.7901782393455505, + "learning_rate": 9.653796076643818e-05, + "loss": 0.0472, + "step": 16320 + }, + { + "epoch": 1.0699076197339972, + "grad_norm": 0.4466434717178345, + "learning_rate": 9.653191259528344e-05, + "loss": 0.0516, + "step": 16330 + }, + { + "epoch": 1.0705627989255062, + "grad_norm": 0.3053244650363922, + "learning_rate": 9.65258593354687e-05, + "loss": 0.0425, + "step": 16340 + }, + { + "epoch": 1.071217978117015, + "grad_norm": 0.26013123989105225, + "learning_rate": 9.651980098765591e-05, + "loss": 0.0636, + "step": 16350 + }, + { + "epoch": 1.0718731573085238, + "grad_norm": 0.6107398867607117, + "learning_rate": 9.651373755250765e-05, + "loss": 0.0351, + "step": 16360 + }, + { + "epoch": 1.0725283365000329, + "grad_norm": 0.36135777831077576, + "learning_rate": 9.650766903068697e-05, + "loss": 0.0406, + "step": 16370 + }, + { + "epoch": 1.0731835156915417, + "grad_norm": 0.8000341057777405, + "learning_rate": 9.650159542285753e-05, + "loss": 0.0599, + "step": 16380 + }, + { + "epoch": 1.0738386948830505, + "grad_norm": 0.49328136444091797, + "learning_rate": 9.649551672968353e-05, + "loss": 0.0703, + "step": 16390 + }, + { + "epoch": 1.0744938740745593, + "grad_norm": 1.0504891872406006, + "learning_rate": 9.648943295182973e-05, + "loss": 0.0613, + "step": 16400 + }, + { + "epoch": 1.0751490532660684, + "grad_norm": 0.7019011378288269, + "learning_rate": 9.648334408996144e-05, + "loss": 0.0491, + "step": 16410 + }, + { + "epoch": 1.0758042324575772, + "grad_norm": 1.0218405723571777, + "learning_rate": 9.647725014474452e-05, + "loss": 0.0664, + "step": 16420 + }, + { + "epoch": 1.076459411649086, + "grad_norm": 1.7036247253417969, + "learning_rate": 9.64711511168454e-05, + "loss": 0.0614, + "step": 16430 + }, + { + "epoch": 1.0771145908405948, + "grad_norm": 0.46196651458740234, + "learning_rate": 9.646504700693108e-05, + "loss": 0.0593, + "step": 16440 + }, + { + "epoch": 1.0777697700321038, + "grad_norm": 0.34031757712364197, + "learning_rate": 9.645893781566907e-05, + "loss": 0.0572, + "step": 16450 + }, + { + "epoch": 1.0784249492236126, + "grad_norm": 0.5915088653564453, + "learning_rate": 9.645282354372744e-05, + "loss": 0.0673, + "step": 16460 + }, + { + "epoch": 1.0790801284151215, + "grad_norm": 0.7095526456832886, + "learning_rate": 9.644670419177491e-05, + "loss": 0.0426, + "step": 16470 + }, + { + "epoch": 1.0797353076066305, + "grad_norm": 1.0091673135757446, + "learning_rate": 9.644057976048062e-05, + "loss": 0.0605, + "step": 16480 + }, + { + "epoch": 1.0803904867981393, + "grad_norm": 0.840774655342102, + "learning_rate": 9.643445025051435e-05, + "loss": 0.0553, + "step": 16490 + }, + { + "epoch": 1.0810456659896481, + "grad_norm": 0.458825945854187, + "learning_rate": 9.642831566254641e-05, + "loss": 0.0605, + "step": 16500 + }, + { + "epoch": 1.081700845181157, + "grad_norm": 0.4829000234603882, + "learning_rate": 9.642217599724769e-05, + "loss": 0.0676, + "step": 16510 + }, + { + "epoch": 1.082356024372666, + "grad_norm": 0.41944870352745056, + "learning_rate": 9.64160312552896e-05, + "loss": 0.0409, + "step": 16520 + }, + { + "epoch": 1.0830112035641748, + "grad_norm": 0.7343906760215759, + "learning_rate": 9.64098814373441e-05, + "loss": 0.0591, + "step": 16530 + }, + { + "epoch": 1.0836663827556836, + "grad_norm": 0.8431534171104431, + "learning_rate": 9.640372654408374e-05, + "loss": 0.0803, + "step": 16540 + }, + { + "epoch": 1.0843215619471926, + "grad_norm": 0.2539506256580353, + "learning_rate": 9.639756657618162e-05, + "loss": 0.0416, + "step": 16550 + }, + { + "epoch": 1.0849767411387015, + "grad_norm": 0.587068498134613, + "learning_rate": 9.639140153431138e-05, + "loss": 0.05, + "step": 16560 + }, + { + "epoch": 1.0856319203302103, + "grad_norm": 0.8803436160087585, + "learning_rate": 9.638523141914721e-05, + "loss": 0.0443, + "step": 16570 + }, + { + "epoch": 1.086287099521719, + "grad_norm": 0.31327107548713684, + "learning_rate": 9.637905623136388e-05, + "loss": 0.0593, + "step": 16580 + }, + { + "epoch": 1.0869422787132281, + "grad_norm": 0.2935597598552704, + "learning_rate": 9.637287597163669e-05, + "loss": 0.077, + "step": 16590 + }, + { + "epoch": 1.087597457904737, + "grad_norm": 0.590408980846405, + "learning_rate": 9.63666906406415e-05, + "loss": 0.0528, + "step": 16600 + }, + { + "epoch": 1.0882526370962458, + "grad_norm": 1.3374470472335815, + "learning_rate": 9.636050023905473e-05, + "loss": 0.0584, + "step": 16610 + }, + { + "epoch": 1.0889078162877548, + "grad_norm": 0.4571608006954193, + "learning_rate": 9.635430476755336e-05, + "loss": 0.0728, + "step": 16620 + }, + { + "epoch": 1.0895629954792636, + "grad_norm": 0.5179108381271362, + "learning_rate": 9.63481042268149e-05, + "loss": 0.0498, + "step": 16630 + }, + { + "epoch": 1.0902181746707724, + "grad_norm": 0.5101157426834106, + "learning_rate": 9.634189861751745e-05, + "loss": 0.0424, + "step": 16640 + }, + { + "epoch": 1.0908733538622812, + "grad_norm": 0.39547455310821533, + "learning_rate": 9.633568794033967e-05, + "loss": 0.0616, + "step": 16650 + }, + { + "epoch": 1.0915285330537903, + "grad_norm": 0.4182504415512085, + "learning_rate": 9.63294721959607e-05, + "loss": 0.0442, + "step": 16660 + }, + { + "epoch": 1.092183712245299, + "grad_norm": 0.9945369362831116, + "learning_rate": 9.63232513850603e-05, + "loss": 0.0623, + "step": 16670 + }, + { + "epoch": 1.092838891436808, + "grad_norm": 0.9508916139602661, + "learning_rate": 9.631702550831878e-05, + "loss": 0.0499, + "step": 16680 + }, + { + "epoch": 1.093494070628317, + "grad_norm": 0.5355792045593262, + "learning_rate": 9.631079456641698e-05, + "loss": 0.0424, + "step": 16690 + }, + { + "epoch": 1.0941492498198258, + "grad_norm": 0.4692195653915405, + "learning_rate": 9.630455856003632e-05, + "loss": 0.0578, + "step": 16700 + }, + { + "epoch": 1.0948044290113346, + "grad_norm": 0.21739773452281952, + "learning_rate": 9.629831748985876e-05, + "loss": 0.0334, + "step": 16710 + }, + { + "epoch": 1.0954596082028434, + "grad_norm": 0.3215731382369995, + "learning_rate": 9.629207135656679e-05, + "loss": 0.0576, + "step": 16720 + }, + { + "epoch": 1.0961147873943524, + "grad_norm": 0.4540056884288788, + "learning_rate": 9.628582016084353e-05, + "loss": 0.0537, + "step": 16730 + }, + { + "epoch": 1.0967699665858612, + "grad_norm": 0.5247315168380737, + "learning_rate": 9.627956390337254e-05, + "loss": 0.0656, + "step": 16740 + }, + { + "epoch": 1.09742514577737, + "grad_norm": 0.0761343315243721, + "learning_rate": 9.627330258483802e-05, + "loss": 0.0569, + "step": 16750 + }, + { + "epoch": 1.098080324968879, + "grad_norm": 0.24553784728050232, + "learning_rate": 9.62670362059247e-05, + "loss": 0.0555, + "step": 16760 + }, + { + "epoch": 1.098735504160388, + "grad_norm": 0.9259544014930725, + "learning_rate": 9.626076476731786e-05, + "loss": 0.0585, + "step": 16770 + }, + { + "epoch": 1.0993906833518967, + "grad_norm": 1.0552631616592407, + "learning_rate": 9.625448826970336e-05, + "loss": 0.0516, + "step": 16780 + }, + { + "epoch": 1.1000458625434055, + "grad_norm": 0.9471535682678223, + "learning_rate": 9.624820671376755e-05, + "loss": 0.0568, + "step": 16790 + }, + { + "epoch": 1.1007010417349146, + "grad_norm": 0.4132363200187683, + "learning_rate": 9.62419201001974e-05, + "loss": 0.0612, + "step": 16800 + }, + { + "epoch": 1.1013562209264234, + "grad_norm": 0.7691220641136169, + "learning_rate": 9.623562842968037e-05, + "loss": 0.0464, + "step": 16810 + }, + { + "epoch": 1.1020114001179322, + "grad_norm": 1.110168695449829, + "learning_rate": 9.622933170290454e-05, + "loss": 0.0496, + "step": 16820 + }, + { + "epoch": 1.1026665793094412, + "grad_norm": 0.3483980894088745, + "learning_rate": 9.622302992055849e-05, + "loss": 0.0564, + "step": 16830 + }, + { + "epoch": 1.10332175850095, + "grad_norm": 0.28381940722465515, + "learning_rate": 9.62167230833314e-05, + "loss": 0.0523, + "step": 16840 + }, + { + "epoch": 1.1039769376924589, + "grad_norm": 0.258341521024704, + "learning_rate": 9.621041119191295e-05, + "loss": 0.0595, + "step": 16850 + }, + { + "epoch": 1.1046321168839677, + "grad_norm": 1.3947983980178833, + "learning_rate": 9.620409424699342e-05, + "loss": 0.0715, + "step": 16860 + }, + { + "epoch": 1.1052872960754767, + "grad_norm": 0.35791313648223877, + "learning_rate": 9.619777224926359e-05, + "loss": 0.0764, + "step": 16870 + }, + { + "epoch": 1.1059424752669855, + "grad_norm": 0.5870449542999268, + "learning_rate": 9.619144519941485e-05, + "loss": 0.0459, + "step": 16880 + }, + { + "epoch": 1.1065976544584943, + "grad_norm": 0.9933807253837585, + "learning_rate": 9.618511309813912e-05, + "loss": 0.0673, + "step": 16890 + }, + { + "epoch": 1.1072528336500034, + "grad_norm": 0.43129727244377136, + "learning_rate": 9.617877594612886e-05, + "loss": 0.0556, + "step": 16900 + }, + { + "epoch": 1.1079080128415122, + "grad_norm": 0.7708754539489746, + "learning_rate": 9.617243374407707e-05, + "loss": 0.0459, + "step": 16910 + }, + { + "epoch": 1.108563192033021, + "grad_norm": 0.4140680432319641, + "learning_rate": 9.616608649267736e-05, + "loss": 0.1141, + "step": 16920 + }, + { + "epoch": 1.1092183712245298, + "grad_norm": 0.31490910053253174, + "learning_rate": 9.615973419262385e-05, + "loss": 0.0612, + "step": 16930 + }, + { + "epoch": 1.1098735504160389, + "grad_norm": 0.7970200777053833, + "learning_rate": 9.615337684461119e-05, + "loss": 0.0686, + "step": 16940 + }, + { + "epoch": 1.1105287296075477, + "grad_norm": 0.7935006022453308, + "learning_rate": 9.614701444933465e-05, + "loss": 0.0676, + "step": 16950 + }, + { + "epoch": 1.1111839087990565, + "grad_norm": 0.8026410937309265, + "learning_rate": 9.614064700748997e-05, + "loss": 0.0441, + "step": 16960 + }, + { + "epoch": 1.1118390879905655, + "grad_norm": 0.28777262568473816, + "learning_rate": 9.613427451977352e-05, + "loss": 0.0567, + "step": 16970 + }, + { + "epoch": 1.1124942671820743, + "grad_norm": 0.8483687043190002, + "learning_rate": 9.612789698688216e-05, + "loss": 0.0669, + "step": 16980 + }, + { + "epoch": 1.1131494463735832, + "grad_norm": 0.32392868399620056, + "learning_rate": 9.612151440951334e-05, + "loss": 0.0535, + "step": 16990 + }, + { + "epoch": 1.113804625565092, + "grad_norm": 0.9474427700042725, + "learning_rate": 9.611512678836506e-05, + "loss": 0.0663, + "step": 17000 + }, + { + "epoch": 1.114459804756601, + "grad_norm": 1.1792471408843994, + "learning_rate": 9.610873412413584e-05, + "loss": 0.047, + "step": 17010 + }, + { + "epoch": 1.1151149839481098, + "grad_norm": 0.41482996940612793, + "learning_rate": 9.610233641752476e-05, + "loss": 0.0594, + "step": 17020 + }, + { + "epoch": 1.1157701631396186, + "grad_norm": 0.5929920673370361, + "learning_rate": 9.609593366923151e-05, + "loss": 0.0532, + "step": 17030 + }, + { + "epoch": 1.1164253423311274, + "grad_norm": 0.6771838068962097, + "learning_rate": 9.608952587995625e-05, + "loss": 0.0742, + "step": 17040 + }, + { + "epoch": 1.1170805215226365, + "grad_norm": 0.5615236163139343, + "learning_rate": 9.608311305039972e-05, + "loss": 0.0665, + "step": 17050 + }, + { + "epoch": 1.1177357007141453, + "grad_norm": 0.7595090270042419, + "learning_rate": 9.607669518126326e-05, + "loss": 0.0516, + "step": 17060 + }, + { + "epoch": 1.1183908799056541, + "grad_norm": 0.2905902862548828, + "learning_rate": 9.607027227324866e-05, + "loss": 0.0596, + "step": 17070 + }, + { + "epoch": 1.1190460590971631, + "grad_norm": 0.5181713104248047, + "learning_rate": 9.606384432705837e-05, + "loss": 0.0454, + "step": 17080 + }, + { + "epoch": 1.119701238288672, + "grad_norm": 0.424075186252594, + "learning_rate": 9.60574113433953e-05, + "loss": 0.0509, + "step": 17090 + }, + { + "epoch": 1.1203564174801808, + "grad_norm": 0.17583155632019043, + "learning_rate": 9.6050973322963e-05, + "loss": 0.0856, + "step": 17100 + }, + { + "epoch": 1.1210115966716896, + "grad_norm": 0.5407838821411133, + "learning_rate": 9.604453026646547e-05, + "loss": 0.0573, + "step": 17110 + }, + { + "epoch": 1.1216667758631986, + "grad_norm": 0.5250619649887085, + "learning_rate": 9.603808217460735e-05, + "loss": 0.0467, + "step": 17120 + }, + { + "epoch": 1.1223219550547074, + "grad_norm": 0.6263852119445801, + "learning_rate": 9.603162904809377e-05, + "loss": 0.0558, + "step": 17130 + }, + { + "epoch": 1.1229771342462163, + "grad_norm": 0.4032648503780365, + "learning_rate": 9.602517088763045e-05, + "loss": 0.0689, + "step": 17140 + }, + { + "epoch": 1.1236323134377253, + "grad_norm": 0.21326816082000732, + "learning_rate": 9.601870769392365e-05, + "loss": 0.0527, + "step": 17150 + }, + { + "epoch": 1.124287492629234, + "grad_norm": 1.2267180681228638, + "learning_rate": 9.601223946768017e-05, + "loss": 0.0719, + "step": 17160 + }, + { + "epoch": 1.124942671820743, + "grad_norm": 1.0646016597747803, + "learning_rate": 9.600576620960734e-05, + "loss": 0.0543, + "step": 17170 + }, + { + "epoch": 1.1255978510122517, + "grad_norm": 0.43919625878334045, + "learning_rate": 9.599928792041308e-05, + "loss": 0.0463, + "step": 17180 + }, + { + "epoch": 1.1262530302037608, + "grad_norm": 0.6069203615188599, + "learning_rate": 9.599280460080587e-05, + "loss": 0.0253, + "step": 17190 + }, + { + "epoch": 1.1269082093952696, + "grad_norm": 0.35923925042152405, + "learning_rate": 9.59863162514947e-05, + "loss": 0.0384, + "step": 17200 + }, + { + "epoch": 1.1275633885867784, + "grad_norm": 0.7490515112876892, + "learning_rate": 9.597982287318911e-05, + "loss": 0.048, + "step": 17210 + }, + { + "epoch": 1.1282185677782874, + "grad_norm": 1.0550154447555542, + "learning_rate": 9.597332446659923e-05, + "loss": 0.0485, + "step": 17220 + }, + { + "epoch": 1.1288737469697963, + "grad_norm": 0.19657357037067413, + "learning_rate": 9.59668210324357e-05, + "loss": 0.0452, + "step": 17230 + }, + { + "epoch": 1.129528926161305, + "grad_norm": 0.42880991101264954, + "learning_rate": 9.596031257140974e-05, + "loss": 0.1037, + "step": 17240 + }, + { + "epoch": 1.1301841053528139, + "grad_norm": 0.3341609537601471, + "learning_rate": 9.59537990842331e-05, + "loss": 0.1061, + "step": 17250 + }, + { + "epoch": 1.130839284544323, + "grad_norm": 0.12193062901496887, + "learning_rate": 9.594728057161806e-05, + "loss": 0.0751, + "step": 17260 + }, + { + "epoch": 1.1314944637358317, + "grad_norm": 0.8789198398590088, + "learning_rate": 9.594075703427752e-05, + "loss": 0.0512, + "step": 17270 + }, + { + "epoch": 1.1321496429273405, + "grad_norm": 0.9001361131668091, + "learning_rate": 9.593422847292486e-05, + "loss": 0.0335, + "step": 17280 + }, + { + "epoch": 1.1328048221188496, + "grad_norm": 0.35644108057022095, + "learning_rate": 9.592769488827402e-05, + "loss": 0.041, + "step": 17290 + }, + { + "epoch": 1.1334600013103584, + "grad_norm": 0.29580044746398926, + "learning_rate": 9.592115628103952e-05, + "loss": 0.0436, + "step": 17300 + }, + { + "epoch": 1.1341151805018672, + "grad_norm": 0.24391017854213715, + "learning_rate": 9.591461265193643e-05, + "loss": 0.0514, + "step": 17310 + }, + { + "epoch": 1.134770359693376, + "grad_norm": 0.33320099115371704, + "learning_rate": 9.590806400168032e-05, + "loss": 0.0593, + "step": 17320 + }, + { + "epoch": 1.135425538884885, + "grad_norm": 0.535823404788971, + "learning_rate": 9.590151033098735e-05, + "loss": 0.0411, + "step": 17330 + }, + { + "epoch": 1.1360807180763939, + "grad_norm": 0.3491571247577667, + "learning_rate": 9.589495164057423e-05, + "loss": 0.0385, + "step": 17340 + }, + { + "epoch": 1.1367358972679027, + "grad_norm": 0.8274325728416443, + "learning_rate": 9.58883879311582e-05, + "loss": 0.0456, + "step": 17350 + }, + { + "epoch": 1.1373910764594117, + "grad_norm": 0.4404541254043579, + "learning_rate": 9.588181920345705e-05, + "loss": 0.0379, + "step": 17360 + }, + { + "epoch": 1.1380462556509205, + "grad_norm": 0.608910858631134, + "learning_rate": 9.587524545818913e-05, + "loss": 0.0751, + "step": 17370 + }, + { + "epoch": 1.1387014348424294, + "grad_norm": 0.19191625714302063, + "learning_rate": 9.586866669607335e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 1.1393566140339382, + "grad_norm": 0.20141072571277618, + "learning_rate": 9.586208291782915e-05, + "loss": 0.0341, + "step": 17390 + }, + { + "epoch": 1.1400117932254472, + "grad_norm": 0.5816513299942017, + "learning_rate": 9.58554941241765e-05, + "loss": 0.0498, + "step": 17400 + }, + { + "epoch": 1.140666972416956, + "grad_norm": 1.2082483768463135, + "learning_rate": 9.584890031583596e-05, + "loss": 0.059, + "step": 17410 + }, + { + "epoch": 1.1413221516084648, + "grad_norm": 0.8812469244003296, + "learning_rate": 9.584230149352861e-05, + "loss": 0.0534, + "step": 17420 + }, + { + "epoch": 1.1419773307999739, + "grad_norm": 0.6087281703948975, + "learning_rate": 9.58356976579761e-05, + "loss": 0.0459, + "step": 17430 + }, + { + "epoch": 1.1426325099914827, + "grad_norm": 0.387494295835495, + "learning_rate": 9.58290888099006e-05, + "loss": 0.0506, + "step": 17440 + }, + { + "epoch": 1.1432876891829915, + "grad_norm": 0.8119124174118042, + "learning_rate": 9.582247495002486e-05, + "loss": 0.0745, + "step": 17450 + }, + { + "epoch": 1.1439428683745003, + "grad_norm": 1.5662113428115845, + "learning_rate": 9.581585607907214e-05, + "loss": 0.0538, + "step": 17460 + }, + { + "epoch": 1.1445980475660094, + "grad_norm": 0.646527111530304, + "learning_rate": 9.580923219776628e-05, + "loss": 0.056, + "step": 17470 + }, + { + "epoch": 1.1452532267575182, + "grad_norm": 0.4142228662967682, + "learning_rate": 9.580260330683167e-05, + "loss": 0.0422, + "step": 17480 + }, + { + "epoch": 1.145908405949027, + "grad_norm": 1.3044791221618652, + "learning_rate": 9.579596940699322e-05, + "loss": 0.054, + "step": 17490 + }, + { + "epoch": 1.146563585140536, + "grad_norm": 0.8836984038352966, + "learning_rate": 9.578933049897643e-05, + "loss": 0.0555, + "step": 17500 + }, + { + "epoch": 1.1472187643320448, + "grad_norm": 0.22296738624572754, + "learning_rate": 9.578268658350728e-05, + "loss": 0.073, + "step": 17510 + }, + { + "epoch": 1.1478739435235537, + "grad_norm": 0.6839221119880676, + "learning_rate": 9.577603766131235e-05, + "loss": 0.0499, + "step": 17520 + }, + { + "epoch": 1.1485291227150625, + "grad_norm": 1.0560941696166992, + "learning_rate": 9.576938373311878e-05, + "loss": 0.0777, + "step": 17530 + }, + { + "epoch": 1.1491843019065715, + "grad_norm": 0.3226263225078583, + "learning_rate": 9.576272479965421e-05, + "loss": 0.0333, + "step": 17540 + }, + { + "epoch": 1.1498394810980803, + "grad_norm": 1.1226500272750854, + "learning_rate": 9.575606086164687e-05, + "loss": 0.0415, + "step": 17550 + }, + { + "epoch": 1.1504946602895891, + "grad_norm": 0.3743634521961212, + "learning_rate": 9.57493919198255e-05, + "loss": 0.0369, + "step": 17560 + }, + { + "epoch": 1.1511498394810982, + "grad_norm": 0.707072913646698, + "learning_rate": 9.57427179749194e-05, + "loss": 0.098, + "step": 17570 + }, + { + "epoch": 1.151805018672607, + "grad_norm": 0.3422117531299591, + "learning_rate": 9.573603902765846e-05, + "loss": 0.0623, + "step": 17580 + }, + { + "epoch": 1.1524601978641158, + "grad_norm": 0.4060186743736267, + "learning_rate": 9.572935507877304e-05, + "loss": 0.0749, + "step": 17590 + }, + { + "epoch": 1.1531153770556246, + "grad_norm": 0.26571348309516907, + "learning_rate": 9.57226661289941e-05, + "loss": 0.0381, + "step": 17600 + }, + { + "epoch": 1.1537705562471336, + "grad_norm": 1.0301426649093628, + "learning_rate": 9.571597217905315e-05, + "loss": 0.0622, + "step": 17610 + }, + { + "epoch": 1.1544257354386425, + "grad_norm": 1.001690149307251, + "learning_rate": 9.57092732296822e-05, + "loss": 0.0416, + "step": 17620 + }, + { + "epoch": 1.1550809146301513, + "grad_norm": 0.43857550621032715, + "learning_rate": 9.570256928161385e-05, + "loss": 0.0463, + "step": 17630 + }, + { + "epoch": 1.1557360938216603, + "grad_norm": 0.22045372426509857, + "learning_rate": 9.569586033558126e-05, + "loss": 0.0393, + "step": 17640 + }, + { + "epoch": 1.1563912730131691, + "grad_norm": 0.7208495140075684, + "learning_rate": 9.568914639231807e-05, + "loss": 0.0525, + "step": 17650 + }, + { + "epoch": 1.157046452204678, + "grad_norm": 0.12816758453845978, + "learning_rate": 9.568242745255852e-05, + "loss": 0.0865, + "step": 17660 + }, + { + "epoch": 1.1577016313961868, + "grad_norm": 0.4800626039505005, + "learning_rate": 9.567570351703739e-05, + "loss": 0.0657, + "step": 17670 + }, + { + "epoch": 1.1583568105876958, + "grad_norm": 0.3882232904434204, + "learning_rate": 9.566897458649001e-05, + "loss": 0.0551, + "step": 17680 + }, + { + "epoch": 1.1590119897792046, + "grad_norm": 0.6293248534202576, + "learning_rate": 9.566224066165221e-05, + "loss": 0.0895, + "step": 17690 + }, + { + "epoch": 1.1596671689707134, + "grad_norm": 0.5119550824165344, + "learning_rate": 9.565550174326043e-05, + "loss": 0.0435, + "step": 17700 + }, + { + "epoch": 1.1603223481622225, + "grad_norm": 0.42948970198631287, + "learning_rate": 9.564875783205162e-05, + "loss": 0.0457, + "step": 17710 + }, + { + "epoch": 1.1609775273537313, + "grad_norm": 0.4685109257698059, + "learning_rate": 9.564200892876328e-05, + "loss": 0.0437, + "step": 17720 + }, + { + "epoch": 1.16163270654524, + "grad_norm": 0.6774266958236694, + "learning_rate": 9.563525503413348e-05, + "loss": 0.0492, + "step": 17730 + }, + { + "epoch": 1.162287885736749, + "grad_norm": 0.7211990356445312, + "learning_rate": 9.562849614890079e-05, + "loss": 0.0658, + "step": 17740 + }, + { + "epoch": 1.162943064928258, + "grad_norm": 0.6713882088661194, + "learning_rate": 9.562173227380436e-05, + "loss": 0.0415, + "step": 17750 + }, + { + "epoch": 1.1635982441197668, + "grad_norm": 0.16038557887077332, + "learning_rate": 9.561496340958389e-05, + "loss": 0.0491, + "step": 17760 + }, + { + "epoch": 1.1642534233112756, + "grad_norm": 1.0927906036376953, + "learning_rate": 9.560818955697959e-05, + "loss": 0.0649, + "step": 17770 + }, + { + "epoch": 1.1649086025027846, + "grad_norm": 0.4227728247642517, + "learning_rate": 9.560141071673228e-05, + "loss": 0.0389, + "step": 17780 + }, + { + "epoch": 1.1655637816942934, + "grad_norm": 0.12926268577575684, + "learning_rate": 9.559462688958323e-05, + "loss": 0.0654, + "step": 17790 + }, + { + "epoch": 1.1662189608858022, + "grad_norm": 1.150467872619629, + "learning_rate": 9.558783807627434e-05, + "loss": 0.0457, + "step": 17800 + }, + { + "epoch": 1.166874140077311, + "grad_norm": 0.10359889268875122, + "learning_rate": 9.558104427754801e-05, + "loss": 0.0657, + "step": 17810 + }, + { + "epoch": 1.16752931926882, + "grad_norm": 0.4183116555213928, + "learning_rate": 9.557424549414722e-05, + "loss": 0.0562, + "step": 17820 + }, + { + "epoch": 1.168184498460329, + "grad_norm": 0.7978748083114624, + "learning_rate": 9.556744172681546e-05, + "loss": 0.0528, + "step": 17830 + }, + { + "epoch": 1.1688396776518377, + "grad_norm": 0.3138968050479889, + "learning_rate": 9.556063297629677e-05, + "loss": 0.0393, + "step": 17840 + }, + { + "epoch": 1.1694948568433468, + "grad_norm": 0.33899372816085815, + "learning_rate": 9.555381924333578e-05, + "loss": 0.0546, + "step": 17850 + }, + { + "epoch": 1.1701500360348556, + "grad_norm": 0.8486229777336121, + "learning_rate": 9.554700052867758e-05, + "loss": 0.0523, + "step": 17860 + }, + { + "epoch": 1.1708052152263644, + "grad_norm": 0.5689741969108582, + "learning_rate": 9.554017683306789e-05, + "loss": 0.0402, + "step": 17870 + }, + { + "epoch": 1.1714603944178732, + "grad_norm": 0.29318487644195557, + "learning_rate": 9.553334815725294e-05, + "loss": 0.0373, + "step": 17880 + }, + { + "epoch": 1.1721155736093822, + "grad_norm": 0.6952021718025208, + "learning_rate": 9.552651450197949e-05, + "loss": 0.0662, + "step": 17890 + }, + { + "epoch": 1.172770752800891, + "grad_norm": 0.526687502861023, + "learning_rate": 9.551967586799486e-05, + "loss": 0.0353, + "step": 17900 + }, + { + "epoch": 1.1734259319923999, + "grad_norm": 0.743867039680481, + "learning_rate": 9.551283225604692e-05, + "loss": 0.0494, + "step": 17910 + }, + { + "epoch": 1.174081111183909, + "grad_norm": 0.5776419043540955, + "learning_rate": 9.550598366688406e-05, + "loss": 0.0371, + "step": 17920 + }, + { + "epoch": 1.1747362903754177, + "grad_norm": 0.10640577226877213, + "learning_rate": 9.549913010125526e-05, + "loss": 0.0629, + "step": 17930 + }, + { + "epoch": 1.1753914695669265, + "grad_norm": 0.5995418429374695, + "learning_rate": 9.549227155990999e-05, + "loss": 0.0704, + "step": 17940 + }, + { + "epoch": 1.1760466487584353, + "grad_norm": 0.5292865037918091, + "learning_rate": 9.548540804359828e-05, + "loss": 0.0768, + "step": 17950 + }, + { + "epoch": 1.1767018279499444, + "grad_norm": 0.7603276968002319, + "learning_rate": 9.547853955307077e-05, + "loss": 0.0742, + "step": 17960 + }, + { + "epoch": 1.1773570071414532, + "grad_norm": 0.6927990317344666, + "learning_rate": 9.547166608907853e-05, + "loss": 0.065, + "step": 17970 + }, + { + "epoch": 1.178012186332962, + "grad_norm": 0.6408458948135376, + "learning_rate": 9.546478765237326e-05, + "loss": 0.0612, + "step": 17980 + }, + { + "epoch": 1.178667365524471, + "grad_norm": 0.45535513758659363, + "learning_rate": 9.545790424370715e-05, + "loss": 0.0551, + "step": 17990 + }, + { + "epoch": 1.1793225447159799, + "grad_norm": 0.7246459722518921, + "learning_rate": 9.5451015863833e-05, + "loss": 0.0546, + "step": 18000 + }, + { + "epoch": 1.1799777239074887, + "grad_norm": 0.8674584627151489, + "learning_rate": 9.544412251350408e-05, + "loss": 0.0752, + "step": 18010 + }, + { + "epoch": 1.1806329030989975, + "grad_norm": 0.6375929117202759, + "learning_rate": 9.543722419347422e-05, + "loss": 0.0772, + "step": 18020 + }, + { + "epoch": 1.1812880822905065, + "grad_norm": 0.4118753671646118, + "learning_rate": 9.543032090449788e-05, + "loss": 0.0988, + "step": 18030 + }, + { + "epoch": 1.1819432614820153, + "grad_norm": 0.4237883985042572, + "learning_rate": 9.542341264732992e-05, + "loss": 0.0431, + "step": 18040 + }, + { + "epoch": 1.1825984406735242, + "grad_norm": 0.25154343247413635, + "learning_rate": 9.541649942272585e-05, + "loss": 0.0582, + "step": 18050 + }, + { + "epoch": 1.1832536198650332, + "grad_norm": 0.5489751100540161, + "learning_rate": 9.54095812314417e-05, + "loss": 0.0413, + "step": 18060 + }, + { + "epoch": 1.183908799056542, + "grad_norm": 0.5093744397163391, + "learning_rate": 9.540265807423401e-05, + "loss": 0.0807, + "step": 18070 + }, + { + "epoch": 1.1845639782480508, + "grad_norm": 0.5402920842170715, + "learning_rate": 9.53957299518599e-05, + "loss": 0.0508, + "step": 18080 + }, + { + "epoch": 1.1852191574395596, + "grad_norm": 0.7429258823394775, + "learning_rate": 9.5388796865077e-05, + "loss": 0.0327, + "step": 18090 + }, + { + "epoch": 1.1858743366310687, + "grad_norm": 0.351461261510849, + "learning_rate": 9.538185881464353e-05, + "loss": 0.0593, + "step": 18100 + }, + { + "epoch": 1.1865295158225775, + "grad_norm": 0.300557404756546, + "learning_rate": 9.537491580131821e-05, + "loss": 0.054, + "step": 18110 + }, + { + "epoch": 1.1871846950140863, + "grad_norm": 0.6231177449226379, + "learning_rate": 9.53679678258603e-05, + "loss": 0.0509, + "step": 18120 + }, + { + "epoch": 1.1878398742055953, + "grad_norm": 0.6431884169578552, + "learning_rate": 9.536101488902966e-05, + "loss": 0.0697, + "step": 18130 + }, + { + "epoch": 1.1884950533971042, + "grad_norm": 1.100174903869629, + "learning_rate": 9.535405699158663e-05, + "loss": 0.0672, + "step": 18140 + }, + { + "epoch": 1.189150232588613, + "grad_norm": 0.6328405141830444, + "learning_rate": 9.53470941342921e-05, + "loss": 0.0407, + "step": 18150 + }, + { + "epoch": 1.1898054117801218, + "grad_norm": 1.0855684280395508, + "learning_rate": 9.534012631790756e-05, + "loss": 0.0618, + "step": 18160 + }, + { + "epoch": 1.1904605909716308, + "grad_norm": 0.5016604065895081, + "learning_rate": 9.533315354319494e-05, + "loss": 0.0826, + "step": 18170 + }, + { + "epoch": 1.1911157701631396, + "grad_norm": 0.581153154373169, + "learning_rate": 9.532617581091682e-05, + "loss": 0.0595, + "step": 18180 + }, + { + "epoch": 1.1917709493546484, + "grad_norm": 0.864215075969696, + "learning_rate": 9.531919312183629e-05, + "loss": 0.0397, + "step": 18190 + }, + { + "epoch": 1.1924261285461575, + "grad_norm": 0.22308997809886932, + "learning_rate": 9.531220547671688e-05, + "loss": 0.0262, + "step": 18200 + }, + { + "epoch": 1.1930813077376663, + "grad_norm": 0.344596803188324, + "learning_rate": 9.530521287632285e-05, + "loss": 0.038, + "step": 18210 + }, + { + "epoch": 1.1937364869291751, + "grad_norm": 0.16623881459236145, + "learning_rate": 9.529821532141884e-05, + "loss": 0.0498, + "step": 18220 + }, + { + "epoch": 1.194391666120684, + "grad_norm": 1.1714928150177002, + "learning_rate": 9.52912128127701e-05, + "loss": 0.0669, + "step": 18230 + }, + { + "epoch": 1.195046845312193, + "grad_norm": 0.8593022227287292, + "learning_rate": 9.528420535114244e-05, + "loss": 0.0585, + "step": 18240 + }, + { + "epoch": 1.1957020245037018, + "grad_norm": 0.49896135926246643, + "learning_rate": 9.527719293730215e-05, + "loss": 0.0556, + "step": 18250 + }, + { + "epoch": 1.1963572036952106, + "grad_norm": 0.3135271370410919, + "learning_rate": 9.527017557201611e-05, + "loss": 0.0482, + "step": 18260 + }, + { + "epoch": 1.1970123828867196, + "grad_norm": 0.6120530962944031, + "learning_rate": 9.526315325605176e-05, + "loss": 0.0421, + "step": 18270 + }, + { + "epoch": 1.1976675620782284, + "grad_norm": 0.6429656744003296, + "learning_rate": 9.525612599017699e-05, + "loss": 0.0767, + "step": 18280 + }, + { + "epoch": 1.1983227412697373, + "grad_norm": 1.1065306663513184, + "learning_rate": 9.524909377516033e-05, + "loss": 0.055, + "step": 18290 + }, + { + "epoch": 1.198977920461246, + "grad_norm": 0.4776744246482849, + "learning_rate": 9.524205661177081e-05, + "loss": 0.052, + "step": 18300 + }, + { + "epoch": 1.199633099652755, + "grad_norm": 0.2578316628932953, + "learning_rate": 9.523501450077801e-05, + "loss": 0.0472, + "step": 18310 + }, + { + "epoch": 1.200288278844264, + "grad_norm": 0.3128001391887665, + "learning_rate": 9.522796744295202e-05, + "loss": 0.0736, + "step": 18320 + }, + { + "epoch": 1.2009434580357727, + "grad_norm": 0.41342025995254517, + "learning_rate": 9.522091543906352e-05, + "loss": 0.0719, + "step": 18330 + }, + { + "epoch": 1.2015986372272818, + "grad_norm": 0.9653927683830261, + "learning_rate": 9.521385848988369e-05, + "loss": 0.0803, + "step": 18340 + }, + { + "epoch": 1.2022538164187906, + "grad_norm": 0.9606965184211731, + "learning_rate": 9.520679659618428e-05, + "loss": 0.0648, + "step": 18350 + }, + { + "epoch": 1.2029089956102994, + "grad_norm": 0.2532610297203064, + "learning_rate": 9.519972975873754e-05, + "loss": 0.0666, + "step": 18360 + }, + { + "epoch": 1.2035641748018082, + "grad_norm": 0.6106138229370117, + "learning_rate": 9.519265797831633e-05, + "loss": 0.0514, + "step": 18370 + }, + { + "epoch": 1.2042193539933173, + "grad_norm": 0.48028168082237244, + "learning_rate": 9.518558125569399e-05, + "loss": 0.1022, + "step": 18380 + }, + { + "epoch": 1.204874533184826, + "grad_norm": 0.8655226826667786, + "learning_rate": 9.517849959164442e-05, + "loss": 0.0605, + "step": 18390 + }, + { + "epoch": 1.2055297123763349, + "grad_norm": 0.7808778285980225, + "learning_rate": 9.517141298694205e-05, + "loss": 0.0491, + "step": 18400 + }, + { + "epoch": 1.206184891567844, + "grad_norm": 0.8048560619354248, + "learning_rate": 9.516432144236188e-05, + "loss": 0.0517, + "step": 18410 + }, + { + "epoch": 1.2068400707593527, + "grad_norm": 0.4281480610370636, + "learning_rate": 9.515722495867941e-05, + "loss": 0.0568, + "step": 18420 + }, + { + "epoch": 1.2074952499508615, + "grad_norm": 0.7680245041847229, + "learning_rate": 9.515012353667072e-05, + "loss": 0.0366, + "step": 18430 + }, + { + "epoch": 1.2081504291423704, + "grad_norm": 0.7772994041442871, + "learning_rate": 9.51430171771124e-05, + "loss": 0.0703, + "step": 18440 + }, + { + "epoch": 1.2088056083338794, + "grad_norm": 0.44369396567344666, + "learning_rate": 9.513590588078159e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 1.2094607875253882, + "grad_norm": 0.5413742065429688, + "learning_rate": 9.512878964845597e-05, + "loss": 0.0565, + "step": 18460 + }, + { + "epoch": 1.210115966716897, + "grad_norm": 0.7472937703132629, + "learning_rate": 9.512166848091377e-05, + "loss": 0.04, + "step": 18470 + }, + { + "epoch": 1.210771145908406, + "grad_norm": 0.4900820553302765, + "learning_rate": 9.511454237893376e-05, + "loss": 0.0688, + "step": 18480 + }, + { + "epoch": 1.2114263250999149, + "grad_norm": 0.7130264043807983, + "learning_rate": 9.51074113432952e-05, + "loss": 0.0579, + "step": 18490 + }, + { + "epoch": 1.2120815042914237, + "grad_norm": 0.7270646095275879, + "learning_rate": 9.510027537477797e-05, + "loss": 0.0545, + "step": 18500 + }, + { + "epoch": 1.2127366834829325, + "grad_norm": 0.7804653644561768, + "learning_rate": 9.509313447416242e-05, + "loss": 0.0421, + "step": 18510 + }, + { + "epoch": 1.2133918626744415, + "grad_norm": 0.3751300871372223, + "learning_rate": 9.508598864222949e-05, + "loss": 0.0651, + "step": 18520 + }, + { + "epoch": 1.2140470418659504, + "grad_norm": 1.0151700973510742, + "learning_rate": 9.507883787976062e-05, + "loss": 0.0803, + "step": 18530 + }, + { + "epoch": 1.2147022210574592, + "grad_norm": 0.15831777453422546, + "learning_rate": 9.507168218753781e-05, + "loss": 0.061, + "step": 18540 + }, + { + "epoch": 1.2153574002489682, + "grad_norm": 0.8682200908660889, + "learning_rate": 9.506452156634362e-05, + "loss": 0.0568, + "step": 18550 + }, + { + "epoch": 1.216012579440477, + "grad_norm": 0.3861425220966339, + "learning_rate": 9.505735601696109e-05, + "loss": 0.039, + "step": 18560 + }, + { + "epoch": 1.2166677586319858, + "grad_norm": 1.2164251804351807, + "learning_rate": 9.505018554017385e-05, + "loss": 0.0515, + "step": 18570 + }, + { + "epoch": 1.2173229378234947, + "grad_norm": 0.9301819801330566, + "learning_rate": 9.504301013676604e-05, + "loss": 0.0453, + "step": 18580 + }, + { + "epoch": 1.2179781170150037, + "grad_norm": 0.3616366386413574, + "learning_rate": 9.503582980752238e-05, + "loss": 0.0645, + "step": 18590 + }, + { + "epoch": 1.2186332962065125, + "grad_norm": 0.7752664685249329, + "learning_rate": 9.502864455322809e-05, + "loss": 0.0395, + "step": 18600 + }, + { + "epoch": 1.2192884753980213, + "grad_norm": 0.8780767917633057, + "learning_rate": 9.502145437466891e-05, + "loss": 0.0552, + "step": 18610 + }, + { + "epoch": 1.2199436545895304, + "grad_norm": 0.3680931329727173, + "learning_rate": 9.501425927263116e-05, + "loss": 0.0477, + "step": 18620 + }, + { + "epoch": 1.2205988337810392, + "grad_norm": 0.36180761456489563, + "learning_rate": 9.500705924790172e-05, + "loss": 0.0584, + "step": 18630 + }, + { + "epoch": 1.221254012972548, + "grad_norm": 0.8651975989341736, + "learning_rate": 9.499985430126794e-05, + "loss": 0.0571, + "step": 18640 + }, + { + "epoch": 1.2219091921640568, + "grad_norm": 0.24356558918952942, + "learning_rate": 9.499264443351775e-05, + "loss": 0.0685, + "step": 18650 + }, + { + "epoch": 1.2225643713555658, + "grad_norm": 0.5277144908905029, + "learning_rate": 9.498542964543961e-05, + "loss": 0.0486, + "step": 18660 + }, + { + "epoch": 1.2232195505470747, + "grad_norm": 0.2533093988895416, + "learning_rate": 9.497820993782252e-05, + "loss": 0.0414, + "step": 18670 + }, + { + "epoch": 1.2238747297385835, + "grad_norm": 0.5120894908905029, + "learning_rate": 9.497098531145601e-05, + "loss": 0.0856, + "step": 18680 + }, + { + "epoch": 1.2245299089300925, + "grad_norm": 0.4079952836036682, + "learning_rate": 9.496375576713017e-05, + "loss": 0.051, + "step": 18690 + }, + { + "epoch": 1.2251850881216013, + "grad_norm": 0.6839764714241028, + "learning_rate": 9.49565213056356e-05, + "loss": 0.0805, + "step": 18700 + }, + { + "epoch": 1.2258402673131101, + "grad_norm": 0.7666359543800354, + "learning_rate": 9.494928192776342e-05, + "loss": 0.0618, + "step": 18710 + }, + { + "epoch": 1.226495446504619, + "grad_norm": 0.24800927937030792, + "learning_rate": 9.494203763430538e-05, + "loss": 0.0464, + "step": 18720 + }, + { + "epoch": 1.227150625696128, + "grad_norm": 0.26284173130989075, + "learning_rate": 9.493478842605366e-05, + "loss": 0.0562, + "step": 18730 + }, + { + "epoch": 1.2278058048876368, + "grad_norm": 0.40329962968826294, + "learning_rate": 9.492753430380105e-05, + "loss": 0.0573, + "step": 18740 + }, + { + "epoch": 1.2284609840791456, + "grad_norm": 0.9101278185844421, + "learning_rate": 9.492027526834083e-05, + "loss": 0.0703, + "step": 18750 + }, + { + "epoch": 1.2291161632706544, + "grad_norm": 0.34702068567276, + "learning_rate": 9.491301132046684e-05, + "loss": 0.0801, + "step": 18760 + }, + { + "epoch": 1.2297713424621635, + "grad_norm": 0.5224130749702454, + "learning_rate": 9.490574246097345e-05, + "loss": 0.0564, + "step": 18770 + }, + { + "epoch": 1.2304265216536723, + "grad_norm": 1.7840898036956787, + "learning_rate": 9.48984686906556e-05, + "loss": 0.0578, + "step": 18780 + }, + { + "epoch": 1.231081700845181, + "grad_norm": 0.6435231566429138, + "learning_rate": 9.489119001030871e-05, + "loss": 0.0464, + "step": 18790 + }, + { + "epoch": 1.2317368800366901, + "grad_norm": 0.6522310376167297, + "learning_rate": 9.488390642072878e-05, + "loss": 0.0556, + "step": 18800 + }, + { + "epoch": 1.232392059228199, + "grad_norm": 0.591069221496582, + "learning_rate": 9.48766179227123e-05, + "loss": 0.0414, + "step": 18810 + }, + { + "epoch": 1.2330472384197078, + "grad_norm": 0.3335474729537964, + "learning_rate": 9.486932451705636e-05, + "loss": 0.0402, + "step": 18820 + }, + { + "epoch": 1.2337024176112166, + "grad_norm": 0.5566762685775757, + "learning_rate": 9.486202620455857e-05, + "loss": 0.0622, + "step": 18830 + }, + { + "epoch": 1.2343575968027256, + "grad_norm": 0.6435947418212891, + "learning_rate": 9.485472298601704e-05, + "loss": 0.0611, + "step": 18840 + }, + { + "epoch": 1.2350127759942344, + "grad_norm": 1.360075831413269, + "learning_rate": 9.484741486223043e-05, + "loss": 0.0344, + "step": 18850 + }, + { + "epoch": 1.2356679551857432, + "grad_norm": 0.5230138301849365, + "learning_rate": 9.484010183399797e-05, + "loss": 0.0506, + "step": 18860 + }, + { + "epoch": 1.2363231343772523, + "grad_norm": 0.23735514283180237, + "learning_rate": 9.483278390211938e-05, + "loss": 0.0518, + "step": 18870 + }, + { + "epoch": 1.236978313568761, + "grad_norm": 0.6140763759613037, + "learning_rate": 9.482546106739496e-05, + "loss": 0.0482, + "step": 18880 + }, + { + "epoch": 1.23763349276027, + "grad_norm": 0.6998968124389648, + "learning_rate": 9.48181333306255e-05, + "loss": 0.0641, + "step": 18890 + }, + { + "epoch": 1.2382886719517787, + "grad_norm": 0.21759581565856934, + "learning_rate": 9.481080069261237e-05, + "loss": 0.0373, + "step": 18900 + }, + { + "epoch": 1.2389438511432878, + "grad_norm": 0.4588048756122589, + "learning_rate": 9.480346315415745e-05, + "loss": 0.076, + "step": 18910 + }, + { + "epoch": 1.2395990303347966, + "grad_norm": 1.2115967273712158, + "learning_rate": 9.479612071606314e-05, + "loss": 0.0346, + "step": 18920 + }, + { + "epoch": 1.2402542095263054, + "grad_norm": 0.3989427387714386, + "learning_rate": 9.478877337913244e-05, + "loss": 0.06, + "step": 18930 + }, + { + "epoch": 1.2409093887178142, + "grad_norm": 0.7025251984596252, + "learning_rate": 9.478142114416881e-05, + "loss": 0.0465, + "step": 18940 + }, + { + "epoch": 1.2415645679093232, + "grad_norm": 0.5833274126052856, + "learning_rate": 9.47740640119763e-05, + "loss": 0.0503, + "step": 18950 + }, + { + "epoch": 1.242219747100832, + "grad_norm": 0.8541959524154663, + "learning_rate": 9.476670198335947e-05, + "loss": 0.0611, + "step": 18960 + }, + { + "epoch": 1.2428749262923409, + "grad_norm": 0.454820454120636, + "learning_rate": 9.47593350591234e-05, + "loss": 0.0256, + "step": 18970 + }, + { + "epoch": 1.24353010548385, + "grad_norm": 0.7754273414611816, + "learning_rate": 9.475196324007376e-05, + "loss": 0.0564, + "step": 18980 + }, + { + "epoch": 1.2441852846753587, + "grad_norm": 0.9708512425422668, + "learning_rate": 9.474458652701669e-05, + "loss": 0.0605, + "step": 18990 + }, + { + "epoch": 1.2448404638668675, + "grad_norm": 0.3638328015804291, + "learning_rate": 9.473720492075892e-05, + "loss": 0.0326, + "step": 19000 + }, + { + "epoch": 1.2454956430583763, + "grad_norm": 0.6891499757766724, + "learning_rate": 9.472981842210768e-05, + "loss": 0.0459, + "step": 19010 + }, + { + "epoch": 1.2461508222498854, + "grad_norm": 1.2563209533691406, + "learning_rate": 9.472242703187074e-05, + "loss": 0.0347, + "step": 19020 + }, + { + "epoch": 1.2468060014413942, + "grad_norm": 0.488187313079834, + "learning_rate": 9.471503075085643e-05, + "loss": 0.0521, + "step": 19030 + }, + { + "epoch": 1.247461180632903, + "grad_norm": 0.5654842853546143, + "learning_rate": 9.470762957987359e-05, + "loss": 0.0522, + "step": 19040 + }, + { + "epoch": 1.248116359824412, + "grad_norm": 0.6216372847557068, + "learning_rate": 9.470022351973158e-05, + "loss": 0.0941, + "step": 19050 + }, + { + "epoch": 1.2487715390159209, + "grad_norm": 0.8709917068481445, + "learning_rate": 9.469281257124034e-05, + "loss": 0.0458, + "step": 19060 + }, + { + "epoch": 1.2494267182074297, + "grad_norm": 0.8081860542297363, + "learning_rate": 9.46853967352103e-05, + "loss": 0.0302, + "step": 19070 + }, + { + "epoch": 1.2500818973989385, + "grad_norm": 0.6697636842727661, + "learning_rate": 9.467797601245246e-05, + "loss": 0.0504, + "step": 19080 + }, + { + "epoch": 1.2507370765904475, + "grad_norm": 0.6359090805053711, + "learning_rate": 9.467055040377834e-05, + "loss": 0.1145, + "step": 19090 + }, + { + "epoch": 1.2513922557819563, + "grad_norm": 0.2612255811691284, + "learning_rate": 9.466311990999999e-05, + "loss": 0.0635, + "step": 19100 + }, + { + "epoch": 1.2520474349734654, + "grad_norm": 1.000808596611023, + "learning_rate": 9.465568453193e-05, + "loss": 0.0477, + "step": 19110 + }, + { + "epoch": 1.2527026141649742, + "grad_norm": 0.9247007966041565, + "learning_rate": 9.464824427038148e-05, + "loss": 0.0793, + "step": 19120 + }, + { + "epoch": 1.253357793356483, + "grad_norm": 0.31677016615867615, + "learning_rate": 9.46407991261681e-05, + "loss": 0.0503, + "step": 19130 + }, + { + "epoch": 1.2540129725479918, + "grad_norm": 0.7644151449203491, + "learning_rate": 9.463334910010404e-05, + "loss": 0.0468, + "step": 19140 + }, + { + "epoch": 1.2546681517395006, + "grad_norm": 0.6586434841156006, + "learning_rate": 9.462589419300403e-05, + "loss": 0.0413, + "step": 19150 + }, + { + "epoch": 1.2553233309310097, + "grad_norm": 0.7901946902275085, + "learning_rate": 9.461843440568333e-05, + "loss": 0.048, + "step": 19160 + }, + { + "epoch": 1.2559785101225185, + "grad_norm": 0.24173937737941742, + "learning_rate": 9.461096973895773e-05, + "loss": 0.0353, + "step": 19170 + }, + { + "epoch": 1.2566336893140275, + "grad_norm": 0.6419515609741211, + "learning_rate": 9.460350019364355e-05, + "loss": 0.0382, + "step": 19180 + }, + { + "epoch": 1.2572888685055363, + "grad_norm": 0.491373747587204, + "learning_rate": 9.459602577055764e-05, + "loss": 0.0492, + "step": 19190 + }, + { + "epoch": 1.2579440476970452, + "grad_norm": 0.22024153172969818, + "learning_rate": 9.45885464705174e-05, + "loss": 0.0345, + "step": 19200 + }, + { + "epoch": 1.258599226888554, + "grad_norm": 0.4903078079223633, + "learning_rate": 9.458106229434076e-05, + "loss": 0.0531, + "step": 19210 + }, + { + "epoch": 1.2592544060800628, + "grad_norm": 0.274980753660202, + "learning_rate": 9.457357324284617e-05, + "loss": 0.048, + "step": 19220 + }, + { + "epoch": 1.2599095852715718, + "grad_norm": 0.7720242142677307, + "learning_rate": 9.456607931685262e-05, + "loss": 0.0554, + "step": 19230 + }, + { + "epoch": 1.2605647644630806, + "grad_norm": 0.4966922104358673, + "learning_rate": 9.455858051717965e-05, + "loss": 0.0665, + "step": 19240 + }, + { + "epoch": 1.2612199436545894, + "grad_norm": 0.5548005104064941, + "learning_rate": 9.45510768446473e-05, + "loss": 0.0815, + "step": 19250 + }, + { + "epoch": 1.2618751228460985, + "grad_norm": 1.1140377521514893, + "learning_rate": 9.454356830007618e-05, + "loss": 0.0584, + "step": 19260 + }, + { + "epoch": 1.2625303020376073, + "grad_norm": 0.23952271044254303, + "learning_rate": 9.45360548842874e-05, + "loss": 0.0638, + "step": 19270 + }, + { + "epoch": 1.2631854812291161, + "grad_norm": 0.43872544169425964, + "learning_rate": 9.452853659810261e-05, + "loss": 0.0462, + "step": 19280 + }, + { + "epoch": 1.263840660420625, + "grad_norm": 0.3569025993347168, + "learning_rate": 9.452101344234401e-05, + "loss": 0.0596, + "step": 19290 + }, + { + "epoch": 1.264495839612134, + "grad_norm": 0.5443289279937744, + "learning_rate": 9.451348541783431e-05, + "loss": 0.053, + "step": 19300 + }, + { + "epoch": 1.2651510188036428, + "grad_norm": 1.9003762006759644, + "learning_rate": 9.450595252539678e-05, + "loss": 0.072, + "step": 19310 + }, + { + "epoch": 1.2658061979951516, + "grad_norm": 0.3201926350593567, + "learning_rate": 9.449841476585518e-05, + "loss": 0.0564, + "step": 19320 + }, + { + "epoch": 1.2664613771866606, + "grad_norm": 0.7491788864135742, + "learning_rate": 9.449087214003384e-05, + "loss": 0.0599, + "step": 19330 + }, + { + "epoch": 1.2671165563781694, + "grad_norm": 0.9385274052619934, + "learning_rate": 9.448332464875765e-05, + "loss": 0.0378, + "step": 19340 + }, + { + "epoch": 1.2677717355696783, + "grad_norm": 0.7036600112915039, + "learning_rate": 9.447577229285192e-05, + "loss": 0.0469, + "step": 19350 + }, + { + "epoch": 1.268426914761187, + "grad_norm": 0.37911659479141235, + "learning_rate": 9.446821507314261e-05, + "loss": 0.0469, + "step": 19360 + }, + { + "epoch": 1.269082093952696, + "grad_norm": 0.5643960237503052, + "learning_rate": 9.446065299045617e-05, + "loss": 0.0851, + "step": 19370 + }, + { + "epoch": 1.269737273144205, + "grad_norm": 1.5498815774917603, + "learning_rate": 9.445308604561955e-05, + "loss": 0.0502, + "step": 19380 + }, + { + "epoch": 1.2703924523357137, + "grad_norm": 0.17481087148189545, + "learning_rate": 9.444551423946028e-05, + "loss": 0.0473, + "step": 19390 + }, + { + "epoch": 1.2710476315272228, + "grad_norm": 0.7445400357246399, + "learning_rate": 9.443793757280638e-05, + "loss": 0.0831, + "step": 19400 + }, + { + "epoch": 1.2717028107187316, + "grad_norm": 0.4605228006839752, + "learning_rate": 9.443035604648646e-05, + "loss": 0.0576, + "step": 19410 + }, + { + "epoch": 1.2723579899102404, + "grad_norm": 0.3387111723423004, + "learning_rate": 9.44227696613296e-05, + "loss": 0.0446, + "step": 19420 + }, + { + "epoch": 1.2730131691017492, + "grad_norm": 0.23827998340129852, + "learning_rate": 9.441517841816542e-05, + "loss": 0.0274, + "step": 19430 + }, + { + "epoch": 1.2736683482932583, + "grad_norm": 0.6560232043266296, + "learning_rate": 9.440758231782413e-05, + "loss": 0.0407, + "step": 19440 + }, + { + "epoch": 1.274323527484767, + "grad_norm": 0.17827126383781433, + "learning_rate": 9.439998136113639e-05, + "loss": 0.0451, + "step": 19450 + }, + { + "epoch": 1.2749787066762759, + "grad_norm": 0.4786401689052582, + "learning_rate": 9.439237554893344e-05, + "loss": 0.0651, + "step": 19460 + }, + { + "epoch": 1.275633885867785, + "grad_norm": 0.5630859136581421, + "learning_rate": 9.438476488204705e-05, + "loss": 0.0299, + "step": 19470 + }, + { + "epoch": 1.2762890650592937, + "grad_norm": 0.9268182516098022, + "learning_rate": 9.43771493613095e-05, + "loss": 0.0584, + "step": 19480 + }, + { + "epoch": 1.2769442442508026, + "grad_norm": 0.32697662711143494, + "learning_rate": 9.436952898755362e-05, + "loss": 0.0521, + "step": 19490 + }, + { + "epoch": 1.2775994234423114, + "grad_norm": 1.0736029148101807, + "learning_rate": 9.436190376161276e-05, + "loss": 0.0513, + "step": 19500 + }, + { + "epoch": 1.2782546026338204, + "grad_norm": 0.7951659560203552, + "learning_rate": 9.43542736843208e-05, + "loss": 0.0331, + "step": 19510 + }, + { + "epoch": 1.2789097818253292, + "grad_norm": 0.302251398563385, + "learning_rate": 9.434663875651216e-05, + "loss": 0.0771, + "step": 19520 + }, + { + "epoch": 1.279564961016838, + "grad_norm": 1.393570899963379, + "learning_rate": 9.433899897902177e-05, + "loss": 0.1154, + "step": 19530 + }, + { + "epoch": 1.280220140208347, + "grad_norm": 0.412194162607193, + "learning_rate": 9.433135435268511e-05, + "loss": 0.042, + "step": 19540 + }, + { + "epoch": 1.2808753193998559, + "grad_norm": 0.8005461692810059, + "learning_rate": 9.432370487833819e-05, + "loss": 0.0852, + "step": 19550 + }, + { + "epoch": 1.2815304985913647, + "grad_norm": 0.4183309078216553, + "learning_rate": 9.431605055681756e-05, + "loss": 0.0575, + "step": 19560 + }, + { + "epoch": 1.2821856777828735, + "grad_norm": 0.4394189417362213, + "learning_rate": 9.430839138896026e-05, + "loss": 0.0541, + "step": 19570 + }, + { + "epoch": 1.2828408569743825, + "grad_norm": 0.70897376537323, + "learning_rate": 9.43007273756039e-05, + "loss": 0.0301, + "step": 19580 + }, + { + "epoch": 1.2834960361658914, + "grad_norm": 0.9679563045501709, + "learning_rate": 9.429305851758658e-05, + "loss": 0.0227, + "step": 19590 + }, + { + "epoch": 1.2841512153574002, + "grad_norm": 1.0135886669158936, + "learning_rate": 9.428538481574699e-05, + "loss": 0.0352, + "step": 19600 + }, + { + "epoch": 1.2848063945489092, + "grad_norm": 0.15597258508205414, + "learning_rate": 9.42777062709243e-05, + "loss": 0.0346, + "step": 19610 + }, + { + "epoch": 1.285461573740418, + "grad_norm": 0.7457423806190491, + "learning_rate": 9.427002288395821e-05, + "loss": 0.0443, + "step": 19620 + }, + { + "epoch": 1.2861167529319268, + "grad_norm": 0.281803160905838, + "learning_rate": 9.426233465568898e-05, + "loss": 0.0768, + "step": 19630 + }, + { + "epoch": 1.2867719321234357, + "grad_norm": 0.33898767828941345, + "learning_rate": 9.42546415869574e-05, + "loss": 0.0445, + "step": 19640 + }, + { + "epoch": 1.2874271113149447, + "grad_norm": 0.5874878764152527, + "learning_rate": 9.424694367860473e-05, + "loss": 0.067, + "step": 19650 + }, + { + "epoch": 1.2880822905064535, + "grad_norm": 0.9820820689201355, + "learning_rate": 9.423924093147284e-05, + "loss": 0.099, + "step": 19660 + }, + { + "epoch": 1.2887374696979623, + "grad_norm": 0.6596928834915161, + "learning_rate": 9.423153334640407e-05, + "loss": 0.0549, + "step": 19670 + }, + { + "epoch": 1.2893926488894714, + "grad_norm": 0.7672058343887329, + "learning_rate": 9.42238209242413e-05, + "loss": 0.0568, + "step": 19680 + }, + { + "epoch": 1.2900478280809802, + "grad_norm": 0.7648168802261353, + "learning_rate": 9.421610366582798e-05, + "loss": 0.0484, + "step": 19690 + }, + { + "epoch": 1.290703007272489, + "grad_norm": 0.38706067204475403, + "learning_rate": 9.420838157200803e-05, + "loss": 0.0362, + "step": 19700 + }, + { + "epoch": 1.2913581864639978, + "grad_norm": 0.3060498833656311, + "learning_rate": 9.420065464362594e-05, + "loss": 0.0282, + "step": 19710 + }, + { + "epoch": 1.2920133656555068, + "grad_norm": 0.39192843437194824, + "learning_rate": 9.419292288152673e-05, + "loss": 0.066, + "step": 19720 + }, + { + "epoch": 1.2926685448470157, + "grad_norm": 0.3230993449687958, + "learning_rate": 9.418518628655588e-05, + "loss": 0.019, + "step": 19730 + }, + { + "epoch": 1.2933237240385245, + "grad_norm": 0.4923522472381592, + "learning_rate": 9.417744485955951e-05, + "loss": 0.0346, + "step": 19740 + }, + { + "epoch": 1.2939789032300335, + "grad_norm": 0.5484822988510132, + "learning_rate": 9.41696986013842e-05, + "loss": 0.0343, + "step": 19750 + }, + { + "epoch": 1.2946340824215423, + "grad_norm": 0.6803279519081116, + "learning_rate": 9.416194751287705e-05, + "loss": 0.0473, + "step": 19760 + }, + { + "epoch": 1.2952892616130511, + "grad_norm": 0.3231986165046692, + "learning_rate": 9.415419159488572e-05, + "loss": 0.0332, + "step": 19770 + }, + { + "epoch": 1.29594444080456, + "grad_norm": 0.6946686506271362, + "learning_rate": 9.414643084825837e-05, + "loss": 0.0829, + "step": 19780 + }, + { + "epoch": 1.296599619996069, + "grad_norm": 0.49025169014930725, + "learning_rate": 9.413866527384372e-05, + "loss": 0.0351, + "step": 19790 + }, + { + "epoch": 1.2972547991875778, + "grad_norm": 0.1349892020225525, + "learning_rate": 9.4130894872491e-05, + "loss": 0.0414, + "step": 19800 + }, + { + "epoch": 1.2979099783790866, + "grad_norm": 0.20277206599712372, + "learning_rate": 9.412311964504998e-05, + "loss": 0.0588, + "step": 19810 + }, + { + "epoch": 1.2985651575705957, + "grad_norm": 0.6998963356018066, + "learning_rate": 9.411533959237091e-05, + "loss": 0.0443, + "step": 19820 + }, + { + "epoch": 1.2992203367621045, + "grad_norm": 0.1758311241865158, + "learning_rate": 9.410755471530464e-05, + "loss": 0.044, + "step": 19830 + }, + { + "epoch": 1.2998755159536133, + "grad_norm": 0.10083567351102829, + "learning_rate": 9.40997650147025e-05, + "loss": 0.035, + "step": 19840 + }, + { + "epoch": 1.300530695145122, + "grad_norm": 0.5207780003547668, + "learning_rate": 9.409197049141637e-05, + "loss": 0.0614, + "step": 19850 + }, + { + "epoch": 1.3011858743366311, + "grad_norm": 0.5671939253807068, + "learning_rate": 9.408417114629863e-05, + "loss": 0.0534, + "step": 19860 + }, + { + "epoch": 1.30184105352814, + "grad_norm": 0.43330782651901245, + "learning_rate": 9.40763669802022e-05, + "loss": 0.0693, + "step": 19870 + }, + { + "epoch": 1.3024962327196488, + "grad_norm": 0.8098093271255493, + "learning_rate": 9.406855799398056e-05, + "loss": 0.0478, + "step": 19880 + }, + { + "epoch": 1.3031514119111578, + "grad_norm": 1.0734235048294067, + "learning_rate": 9.406074418848767e-05, + "loss": 0.0624, + "step": 19890 + }, + { + "epoch": 1.3038065911026666, + "grad_norm": 0.45437803864479065, + "learning_rate": 9.405292556457805e-05, + "loss": 0.0243, + "step": 19900 + }, + { + "epoch": 1.3044617702941754, + "grad_norm": 0.8875624537467957, + "learning_rate": 9.404510212310671e-05, + "loss": 0.0475, + "step": 19910 + }, + { + "epoch": 1.3051169494856842, + "grad_norm": 0.7182801365852356, + "learning_rate": 9.403727386492924e-05, + "loss": 0.0731, + "step": 19920 + }, + { + "epoch": 1.3057721286771933, + "grad_norm": 0.6282206773757935, + "learning_rate": 9.40294407909017e-05, + "loss": 0.0657, + "step": 19930 + }, + { + "epoch": 1.306427307868702, + "grad_norm": 0.25027623772621155, + "learning_rate": 9.40216029018807e-05, + "loss": 0.0473, + "step": 19940 + }, + { + "epoch": 1.307082487060211, + "grad_norm": 1.1925880908966064, + "learning_rate": 9.401376019872338e-05, + "loss": 0.0673, + "step": 19950 + }, + { + "epoch": 1.30773766625172, + "grad_norm": 1.12544584274292, + "learning_rate": 9.400591268228746e-05, + "loss": 0.0618, + "step": 19960 + }, + { + "epoch": 1.3083928454432288, + "grad_norm": 0.8403616547584534, + "learning_rate": 9.399806035343106e-05, + "loss": 0.0736, + "step": 19970 + }, + { + "epoch": 1.3090480246347376, + "grad_norm": 0.1993868052959442, + "learning_rate": 9.399020321301294e-05, + "loss": 0.0422, + "step": 19980 + }, + { + "epoch": 1.3097032038262464, + "grad_norm": 1.1330106258392334, + "learning_rate": 9.398234126189234e-05, + "loss": 0.1164, + "step": 19990 + }, + { + "epoch": 1.3103583830177554, + "grad_norm": 0.3538931906223297, + "learning_rate": 9.397447450092902e-05, + "loss": 0.0365, + "step": 20000 + }, + { + "epoch": 1.3110135622092642, + "grad_norm": 0.6399162411689758, + "learning_rate": 9.39666029309833e-05, + "loss": 0.0663, + "step": 20010 + }, + { + "epoch": 1.311668741400773, + "grad_norm": 0.4228290915489197, + "learning_rate": 9.395872655291596e-05, + "loss": 0.0401, + "step": 20020 + }, + { + "epoch": 1.312323920592282, + "grad_norm": 0.8269533514976501, + "learning_rate": 9.395084536758838e-05, + "loss": 0.0546, + "step": 20030 + }, + { + "epoch": 1.312979099783791, + "grad_norm": 0.3129405379295349, + "learning_rate": 9.394295937586243e-05, + "loss": 0.0469, + "step": 20040 + }, + { + "epoch": 1.3136342789752997, + "grad_norm": 0.795982837677002, + "learning_rate": 9.393506857860052e-05, + "loss": 0.0419, + "step": 20050 + }, + { + "epoch": 1.3142894581668085, + "grad_norm": 0.9886187314987183, + "learning_rate": 9.392717297666555e-05, + "loss": 0.069, + "step": 20060 + }, + { + "epoch": 1.3149446373583176, + "grad_norm": 0.6311957836151123, + "learning_rate": 9.391927257092101e-05, + "loss": 0.0483, + "step": 20070 + }, + { + "epoch": 1.3155998165498264, + "grad_norm": 0.9611292481422424, + "learning_rate": 9.391136736223085e-05, + "loss": 0.0579, + "step": 20080 + }, + { + "epoch": 1.3162549957413352, + "grad_norm": 0.35051944851875305, + "learning_rate": 9.390345735145956e-05, + "loss": 0.0479, + "step": 20090 + }, + { + "epoch": 1.3169101749328442, + "grad_norm": 0.6381497383117676, + "learning_rate": 9.389554253947219e-05, + "loss": 0.0679, + "step": 20100 + }, + { + "epoch": 1.317565354124353, + "grad_norm": 0.09431727975606918, + "learning_rate": 9.388762292713428e-05, + "loss": 0.0318, + "step": 20110 + }, + { + "epoch": 1.3182205333158619, + "grad_norm": 0.5207447409629822, + "learning_rate": 9.38796985153119e-05, + "loss": 0.0188, + "step": 20120 + }, + { + "epoch": 1.3188757125073707, + "grad_norm": 0.4290775954723358, + "learning_rate": 9.387176930487169e-05, + "loss": 0.0396, + "step": 20130 + }, + { + "epoch": 1.3195308916988797, + "grad_norm": 0.19680756330490112, + "learning_rate": 9.386383529668072e-05, + "loss": 0.0538, + "step": 20140 + }, + { + "epoch": 1.3201860708903885, + "grad_norm": 0.42529618740081787, + "learning_rate": 9.385589649160669e-05, + "loss": 0.0512, + "step": 20150 + }, + { + "epoch": 1.3208412500818973, + "grad_norm": 0.5154808759689331, + "learning_rate": 9.384795289051775e-05, + "loss": 0.0582, + "step": 20160 + }, + { + "epoch": 1.3214964292734064, + "grad_norm": 0.15776769816875458, + "learning_rate": 9.384000449428261e-05, + "loss": 0.0596, + "step": 20170 + }, + { + "epoch": 1.3221516084649152, + "grad_norm": 0.5740185379981995, + "learning_rate": 9.383205130377048e-05, + "loss": 0.0658, + "step": 20180 + }, + { + "epoch": 1.322806787656424, + "grad_norm": 0.7399234175682068, + "learning_rate": 9.382409331985114e-05, + "loss": 0.0536, + "step": 20190 + }, + { + "epoch": 1.3234619668479328, + "grad_norm": 0.13166338205337524, + "learning_rate": 9.381613054339482e-05, + "loss": 0.0604, + "step": 20200 + }, + { + "epoch": 1.3241171460394419, + "grad_norm": 0.4675890803337097, + "learning_rate": 9.380816297527235e-05, + "loss": 0.0246, + "step": 20210 + }, + { + "epoch": 1.3247723252309507, + "grad_norm": 0.31965723633766174, + "learning_rate": 9.380019061635506e-05, + "loss": 0.0372, + "step": 20220 + }, + { + "epoch": 1.3254275044224595, + "grad_norm": 0.2704122066497803, + "learning_rate": 9.379221346751474e-05, + "loss": 0.0319, + "step": 20230 + }, + { + "epoch": 1.3260826836139685, + "grad_norm": 0.1694469004869461, + "learning_rate": 9.378423152962382e-05, + "loss": 0.0453, + "step": 20240 + }, + { + "epoch": 1.3267378628054773, + "grad_norm": 0.6565530896186829, + "learning_rate": 9.377624480355517e-05, + "loss": 0.082, + "step": 20250 + }, + { + "epoch": 1.3273930419969862, + "grad_norm": 1.2430895566940308, + "learning_rate": 9.376825329018219e-05, + "loss": 0.0942, + "step": 20260 + }, + { + "epoch": 1.328048221188495, + "grad_norm": 1.0975923538208008, + "learning_rate": 9.376025699037884e-05, + "loss": 0.0675, + "step": 20270 + }, + { + "epoch": 1.328703400380004, + "grad_norm": 0.7981967926025391, + "learning_rate": 9.37522559050196e-05, + "loss": 0.0484, + "step": 20280 + }, + { + "epoch": 1.3293585795715128, + "grad_norm": 0.4645979404449463, + "learning_rate": 9.37442500349794e-05, + "loss": 0.0677, + "step": 20290 + }, + { + "epoch": 1.3300137587630216, + "grad_norm": 0.5420342683792114, + "learning_rate": 9.373623938113381e-05, + "loss": 0.053, + "step": 20300 + }, + { + "epoch": 1.3306689379545307, + "grad_norm": 0.2692251205444336, + "learning_rate": 9.372822394435883e-05, + "loss": 0.047, + "step": 20310 + }, + { + "epoch": 1.3313241171460395, + "grad_norm": 0.3344281017780304, + "learning_rate": 9.372020372553102e-05, + "loss": 0.0969, + "step": 20320 + }, + { + "epoch": 1.3319792963375483, + "grad_norm": 0.39262768626213074, + "learning_rate": 9.371217872552746e-05, + "loss": 0.0541, + "step": 20330 + }, + { + "epoch": 1.3326344755290571, + "grad_norm": 0.6192760467529297, + "learning_rate": 9.370414894522576e-05, + "loss": 0.0452, + "step": 20340 + }, + { + "epoch": 1.3332896547205662, + "grad_norm": 0.31124255061149597, + "learning_rate": 9.369611438550406e-05, + "loss": 0.0656, + "step": 20350 + }, + { + "epoch": 1.333944833912075, + "grad_norm": 1.0172785520553589, + "learning_rate": 9.368807504724095e-05, + "loss": 0.0596, + "step": 20360 + }, + { + "epoch": 1.3346000131035838, + "grad_norm": 0.39627912640571594, + "learning_rate": 9.368003093131565e-05, + "loss": 0.0661, + "step": 20370 + }, + { + "epoch": 1.3352551922950928, + "grad_norm": 1.2618649005889893, + "learning_rate": 9.367198203860785e-05, + "loss": 0.0482, + "step": 20380 + }, + { + "epoch": 1.3359103714866016, + "grad_norm": 0.7525333762168884, + "learning_rate": 9.366392836999774e-05, + "loss": 0.0747, + "step": 20390 + }, + { + "epoch": 1.3365655506781104, + "grad_norm": 0.5180660486221313, + "learning_rate": 9.365586992636607e-05, + "loss": 0.0529, + "step": 20400 + }, + { + "epoch": 1.3372207298696193, + "grad_norm": 0.4146157503128052, + "learning_rate": 9.364780670859412e-05, + "loss": 0.0412, + "step": 20410 + }, + { + "epoch": 1.3378759090611283, + "grad_norm": 0.47063422203063965, + "learning_rate": 9.363973871756364e-05, + "loss": 0.0645, + "step": 20420 + }, + { + "epoch": 1.3385310882526371, + "grad_norm": 0.3832770884037018, + "learning_rate": 9.363166595415696e-05, + "loss": 0.0658, + "step": 20430 + }, + { + "epoch": 1.339186267444146, + "grad_norm": 0.4016522467136383, + "learning_rate": 9.362358841925686e-05, + "loss": 0.0433, + "step": 20440 + }, + { + "epoch": 1.339841446635655, + "grad_norm": 0.7338289618492126, + "learning_rate": 9.361550611374674e-05, + "loss": 0.0628, + "step": 20450 + }, + { + "epoch": 1.3404966258271638, + "grad_norm": 0.18186014890670776, + "learning_rate": 9.360741903851043e-05, + "loss": 0.054, + "step": 20460 + }, + { + "epoch": 1.3411518050186726, + "grad_norm": 0.8552296161651611, + "learning_rate": 9.359932719443236e-05, + "loss": 0.0822, + "step": 20470 + }, + { + "epoch": 1.3418069842101814, + "grad_norm": 0.2419767528772354, + "learning_rate": 9.35912305823974e-05, + "loss": 0.0392, + "step": 20480 + }, + { + "epoch": 1.3424621634016904, + "grad_norm": 0.8288829922676086, + "learning_rate": 9.358312920329101e-05, + "loss": 0.0525, + "step": 20490 + }, + { + "epoch": 1.3431173425931993, + "grad_norm": 0.5043158531188965, + "learning_rate": 9.357502305799914e-05, + "loss": 0.038, + "step": 20500 + }, + { + "epoch": 1.343772521784708, + "grad_norm": 0.4335780441761017, + "learning_rate": 9.356691214740824e-05, + "loss": 0.0924, + "step": 20510 + }, + { + "epoch": 1.344427700976217, + "grad_norm": 0.29687032103538513, + "learning_rate": 9.355879647240535e-05, + "loss": 0.0392, + "step": 20520 + }, + { + "epoch": 1.345082880167726, + "grad_norm": 0.9341139793395996, + "learning_rate": 9.355067603387798e-05, + "loss": 0.0517, + "step": 20530 + }, + { + "epoch": 1.3457380593592347, + "grad_norm": 0.5551009178161621, + "learning_rate": 9.354255083271412e-05, + "loss": 0.0809, + "step": 20540 + }, + { + "epoch": 1.3463932385507436, + "grad_norm": 0.6693395972251892, + "learning_rate": 9.353442086980239e-05, + "loss": 0.0618, + "step": 20550 + }, + { + "epoch": 1.3470484177422524, + "grad_norm": 0.6483456492424011, + "learning_rate": 9.352628614603185e-05, + "loss": 0.0553, + "step": 20560 + }, + { + "epoch": 1.3477035969337614, + "grad_norm": 0.6476925611495972, + "learning_rate": 9.351814666229209e-05, + "loss": 0.0527, + "step": 20570 + }, + { + "epoch": 1.3483587761252702, + "grad_norm": 0.9748954772949219, + "learning_rate": 9.351000241947324e-05, + "loss": 0.0727, + "step": 20580 + }, + { + "epoch": 1.3490139553167793, + "grad_norm": 0.7786060571670532, + "learning_rate": 9.350185341846594e-05, + "loss": 0.0344, + "step": 20590 + }, + { + "epoch": 1.349669134508288, + "grad_norm": 0.5942990183830261, + "learning_rate": 9.349369966016134e-05, + "loss": 0.0455, + "step": 20600 + }, + { + "epoch": 1.3503243136997969, + "grad_norm": 0.6295963525772095, + "learning_rate": 9.348554114545117e-05, + "loss": 0.0971, + "step": 20610 + }, + { + "epoch": 1.3509794928913057, + "grad_norm": 1.8276087045669556, + "learning_rate": 9.347737787522758e-05, + "loss": 0.07, + "step": 20620 + }, + { + "epoch": 1.3516346720828145, + "grad_norm": 0.5030455589294434, + "learning_rate": 9.346920985038332e-05, + "loss": 0.0793, + "step": 20630 + }, + { + "epoch": 1.3522898512743236, + "grad_norm": 0.5137568116188049, + "learning_rate": 9.346103707181162e-05, + "loss": 0.0425, + "step": 20640 + }, + { + "epoch": 1.3529450304658324, + "grad_norm": 1.0221457481384277, + "learning_rate": 9.345285954040626e-05, + "loss": 0.0628, + "step": 20650 + }, + { + "epoch": 1.3536002096573414, + "grad_norm": 0.3952096402645111, + "learning_rate": 9.34446772570615e-05, + "loss": 0.0565, + "step": 20660 + }, + { + "epoch": 1.3542553888488502, + "grad_norm": 0.3056555688381195, + "learning_rate": 9.343649022267214e-05, + "loss": 0.0377, + "step": 20670 + }, + { + "epoch": 1.354910568040359, + "grad_norm": 1.2054407596588135, + "learning_rate": 9.342829843813353e-05, + "loss": 0.045, + "step": 20680 + }, + { + "epoch": 1.3555657472318678, + "grad_norm": 0.7362620830535889, + "learning_rate": 9.342010190434149e-05, + "loss": 0.0373, + "step": 20690 + }, + { + "epoch": 1.3562209264233767, + "grad_norm": 0.43572965264320374, + "learning_rate": 9.34119006221924e-05, + "loss": 0.0419, + "step": 20700 + }, + { + "epoch": 1.3568761056148857, + "grad_norm": 0.6159766912460327, + "learning_rate": 9.340369459258313e-05, + "loss": 0.0729, + "step": 20710 + }, + { + "epoch": 1.3575312848063945, + "grad_norm": 0.5805152058601379, + "learning_rate": 9.339548381641106e-05, + "loss": 0.0318, + "step": 20720 + }, + { + "epoch": 1.3581864639979035, + "grad_norm": 0.9670439958572388, + "learning_rate": 9.338726829457413e-05, + "loss": 0.0555, + "step": 20730 + }, + { + "epoch": 1.3588416431894124, + "grad_norm": 0.874390721321106, + "learning_rate": 9.337904802797078e-05, + "loss": 0.0699, + "step": 20740 + }, + { + "epoch": 1.3594968223809212, + "grad_norm": 0.44484761357307434, + "learning_rate": 9.337082301749993e-05, + "loss": 0.1081, + "step": 20750 + }, + { + "epoch": 1.36015200157243, + "grad_norm": 2.0392377376556396, + "learning_rate": 9.336259326406109e-05, + "loss": 0.0686, + "step": 20760 + }, + { + "epoch": 1.3608071807639388, + "grad_norm": 1.0793938636779785, + "learning_rate": 9.335435876855427e-05, + "loss": 0.0399, + "step": 20770 + }, + { + "epoch": 1.3614623599554478, + "grad_norm": 0.6239899396896362, + "learning_rate": 9.334611953187994e-05, + "loss": 0.0316, + "step": 20780 + }, + { + "epoch": 1.3621175391469567, + "grad_norm": 0.34225550293922424, + "learning_rate": 9.333787555493914e-05, + "loss": 0.0353, + "step": 20790 + }, + { + "epoch": 1.3627727183384657, + "grad_norm": 0.5385367274284363, + "learning_rate": 9.332962683863345e-05, + "loss": 0.0266, + "step": 20800 + }, + { + "epoch": 1.3634278975299745, + "grad_norm": 0.8148894309997559, + "learning_rate": 9.332137338386489e-05, + "loss": 0.0583, + "step": 20810 + }, + { + "epoch": 1.3640830767214833, + "grad_norm": 0.6740687489509583, + "learning_rate": 9.33131151915361e-05, + "loss": 0.0543, + "step": 20820 + }, + { + "epoch": 1.3647382559129921, + "grad_norm": 0.8461841344833374, + "learning_rate": 9.330485226255012e-05, + "loss": 0.0654, + "step": 20830 + }, + { + "epoch": 1.365393435104501, + "grad_norm": 0.6532135605812073, + "learning_rate": 9.329658459781061e-05, + "loss": 0.0516, + "step": 20840 + }, + { + "epoch": 1.36604861429601, + "grad_norm": 0.5001864433288574, + "learning_rate": 9.328831219822172e-05, + "loss": 0.0468, + "step": 20850 + }, + { + "epoch": 1.3667037934875188, + "grad_norm": 0.2622440755367279, + "learning_rate": 9.328003506468808e-05, + "loss": 0.0545, + "step": 20860 + }, + { + "epoch": 1.3673589726790278, + "grad_norm": 0.4977346360683441, + "learning_rate": 9.327175319811488e-05, + "loss": 0.0529, + "step": 20870 + }, + { + "epoch": 1.3680141518705367, + "grad_norm": 0.9524437189102173, + "learning_rate": 9.326346659940781e-05, + "loss": 0.0515, + "step": 20880 + }, + { + "epoch": 1.3686693310620455, + "grad_norm": 0.6952349543571472, + "learning_rate": 9.325517526947308e-05, + "loss": 0.0303, + "step": 20890 + }, + { + "epoch": 1.3693245102535543, + "grad_norm": 0.49954116344451904, + "learning_rate": 9.32468792092174e-05, + "loss": 0.0426, + "step": 20900 + }, + { + "epoch": 1.369979689445063, + "grad_norm": 0.341581791639328, + "learning_rate": 9.323857841954803e-05, + "loss": 0.0451, + "step": 20910 + }, + { + "epoch": 1.3706348686365721, + "grad_norm": 0.27639463543891907, + "learning_rate": 9.323027290137276e-05, + "loss": 0.0346, + "step": 20920 + }, + { + "epoch": 1.371290047828081, + "grad_norm": 1.1062184572219849, + "learning_rate": 9.322196265559981e-05, + "loss": 0.0537, + "step": 20930 + }, + { + "epoch": 1.37194522701959, + "grad_norm": 0.49044227600097656, + "learning_rate": 9.321364768313803e-05, + "loss": 0.0639, + "step": 20940 + }, + { + "epoch": 1.3726004062110988, + "grad_norm": 0.8042529225349426, + "learning_rate": 9.32053279848967e-05, + "loss": 0.047, + "step": 20950 + }, + { + "epoch": 1.3732555854026076, + "grad_norm": 0.3194376826286316, + "learning_rate": 9.319700356178567e-05, + "loss": 0.0325, + "step": 20960 + }, + { + "epoch": 1.3739107645941164, + "grad_norm": 0.6715950965881348, + "learning_rate": 9.318867441471527e-05, + "loss": 0.075, + "step": 20970 + }, + { + "epoch": 1.3745659437856252, + "grad_norm": 0.13095420598983765, + "learning_rate": 9.318034054459637e-05, + "loss": 0.0668, + "step": 20980 + }, + { + "epoch": 1.3752211229771343, + "grad_norm": 1.00981867313385, + "learning_rate": 9.317200195234034e-05, + "loss": 0.0584, + "step": 20990 + }, + { + "epoch": 1.375876302168643, + "grad_norm": 0.34238895773887634, + "learning_rate": 9.316365863885909e-05, + "loss": 0.038, + "step": 21000 + }, + { + "epoch": 1.3765314813601521, + "grad_norm": 0.8177923560142517, + "learning_rate": 9.315531060506502e-05, + "loss": 0.0574, + "step": 21010 + }, + { + "epoch": 1.377186660551661, + "grad_norm": 0.33728766441345215, + "learning_rate": 9.314695785187108e-05, + "loss": 0.0329, + "step": 21020 + }, + { + "epoch": 1.3778418397431698, + "grad_norm": 0.17811648547649384, + "learning_rate": 9.313860038019069e-05, + "loss": 0.0455, + "step": 21030 + }, + { + "epoch": 1.3784970189346786, + "grad_norm": 0.4920370578765869, + "learning_rate": 9.313023819093782e-05, + "loss": 0.0826, + "step": 21040 + }, + { + "epoch": 1.3791521981261874, + "grad_norm": 0.8315593600273132, + "learning_rate": 9.312187128502695e-05, + "loss": 0.0626, + "step": 21050 + }, + { + "epoch": 1.3798073773176964, + "grad_norm": 0.32068103551864624, + "learning_rate": 9.311349966337307e-05, + "loss": 0.0378, + "step": 21060 + }, + { + "epoch": 1.3804625565092052, + "grad_norm": 0.39171046018600464, + "learning_rate": 9.310512332689169e-05, + "loss": 0.0341, + "step": 21070 + }, + { + "epoch": 1.3811177357007143, + "grad_norm": 0.9401202201843262, + "learning_rate": 9.309674227649883e-05, + "loss": 0.0537, + "step": 21080 + }, + { + "epoch": 1.381772914892223, + "grad_norm": 0.647341787815094, + "learning_rate": 9.308835651311103e-05, + "loss": 0.0682, + "step": 21090 + }, + { + "epoch": 1.382428094083732, + "grad_norm": 0.45922040939331055, + "learning_rate": 9.307996603764533e-05, + "loss": 0.0458, + "step": 21100 + }, + { + "epoch": 1.3830832732752407, + "grad_norm": 0.564559817314148, + "learning_rate": 9.307157085101932e-05, + "loss": 0.0496, + "step": 21110 + }, + { + "epoch": 1.3837384524667495, + "grad_norm": 0.40287625789642334, + "learning_rate": 9.306317095415109e-05, + "loss": 0.0587, + "step": 21120 + }, + { + "epoch": 1.3843936316582586, + "grad_norm": 0.4668300151824951, + "learning_rate": 9.305476634795922e-05, + "loss": 0.0674, + "step": 21130 + }, + { + "epoch": 1.3850488108497674, + "grad_norm": 0.35643282532691956, + "learning_rate": 9.304635703336284e-05, + "loss": 0.0486, + "step": 21140 + }, + { + "epoch": 1.3857039900412764, + "grad_norm": 0.2629068195819855, + "learning_rate": 9.303794301128157e-05, + "loss": 0.0288, + "step": 21150 + }, + { + "epoch": 1.3863591692327852, + "grad_norm": 0.8747535943984985, + "learning_rate": 9.302952428263555e-05, + "loss": 0.0399, + "step": 21160 + }, + { + "epoch": 1.387014348424294, + "grad_norm": 0.17276091873645782, + "learning_rate": 9.302110084834545e-05, + "loss": 0.0542, + "step": 21170 + }, + { + "epoch": 1.3876695276158029, + "grad_norm": 0.3780987560749054, + "learning_rate": 9.301267270933245e-05, + "loss": 0.0395, + "step": 21180 + }, + { + "epoch": 1.3883247068073117, + "grad_norm": 0.9850565791130066, + "learning_rate": 9.300423986651823e-05, + "loss": 0.0549, + "step": 21190 + }, + { + "epoch": 1.3889798859988207, + "grad_norm": 0.7064600586891174, + "learning_rate": 9.299580232082501e-05, + "loss": 0.0381, + "step": 21200 + }, + { + "epoch": 1.3896350651903295, + "grad_norm": 0.7328970432281494, + "learning_rate": 9.298736007317547e-05, + "loss": 0.0481, + "step": 21210 + }, + { + "epoch": 1.3902902443818386, + "grad_norm": 0.7926015257835388, + "learning_rate": 9.297891312449288e-05, + "loss": 0.0741, + "step": 21220 + }, + { + "epoch": 1.3909454235733474, + "grad_norm": 1.2499384880065918, + "learning_rate": 9.297046147570094e-05, + "loss": 0.0556, + "step": 21230 + }, + { + "epoch": 1.3916006027648562, + "grad_norm": 0.13640707731246948, + "learning_rate": 9.296200512772396e-05, + "loss": 0.0393, + "step": 21240 + }, + { + "epoch": 1.392255781956365, + "grad_norm": 0.4069652259349823, + "learning_rate": 9.295354408148668e-05, + "loss": 0.0371, + "step": 21250 + }, + { + "epoch": 1.3929109611478738, + "grad_norm": 0.3836240768432617, + "learning_rate": 9.294507833791441e-05, + "loss": 0.0368, + "step": 21260 + }, + { + "epoch": 1.3935661403393829, + "grad_norm": 0.49075940251350403, + "learning_rate": 9.293660789793295e-05, + "loss": 0.077, + "step": 21270 + }, + { + "epoch": 1.3942213195308917, + "grad_norm": 0.4919561743736267, + "learning_rate": 9.292813276246858e-05, + "loss": 0.034, + "step": 21280 + }, + { + "epoch": 1.3948764987224007, + "grad_norm": 0.629605770111084, + "learning_rate": 9.291965293244816e-05, + "loss": 0.1093, + "step": 21290 + }, + { + "epoch": 1.3955316779139095, + "grad_norm": 0.23057830333709717, + "learning_rate": 9.291116840879904e-05, + "loss": 0.0505, + "step": 21300 + }, + { + "epoch": 1.3961868571054183, + "grad_norm": 0.19924475252628326, + "learning_rate": 9.290267919244904e-05, + "loss": 0.0514, + "step": 21310 + }, + { + "epoch": 1.3968420362969272, + "grad_norm": 0.19548480212688446, + "learning_rate": 9.289418528432655e-05, + "loss": 0.0444, + "step": 21320 + }, + { + "epoch": 1.397497215488436, + "grad_norm": 0.5142144560813904, + "learning_rate": 9.288568668536045e-05, + "loss": 0.0445, + "step": 21330 + }, + { + "epoch": 1.398152394679945, + "grad_norm": 0.8250845670700073, + "learning_rate": 9.287718339648013e-05, + "loss": 0.0661, + "step": 21340 + }, + { + "epoch": 1.3988075738714538, + "grad_norm": 0.42259010672569275, + "learning_rate": 9.28686754186155e-05, + "loss": 0.0438, + "step": 21350 + }, + { + "epoch": 1.3994627530629626, + "grad_norm": 0.33181071281433105, + "learning_rate": 9.286016275269698e-05, + "loss": 0.0616, + "step": 21360 + }, + { + "epoch": 1.4001179322544717, + "grad_norm": 0.4738413095474243, + "learning_rate": 9.285164539965551e-05, + "loss": 0.046, + "step": 21370 + }, + { + "epoch": 1.4007731114459805, + "grad_norm": 0.8650274872779846, + "learning_rate": 9.284312336042251e-05, + "loss": 0.047, + "step": 21380 + }, + { + "epoch": 1.4014282906374893, + "grad_norm": 0.18188515305519104, + "learning_rate": 9.283459663592996e-05, + "loss": 0.0343, + "step": 21390 + }, + { + "epoch": 1.4020834698289981, + "grad_norm": 0.807789146900177, + "learning_rate": 9.282606522711033e-05, + "loss": 0.05, + "step": 21400 + }, + { + "epoch": 1.4027386490205072, + "grad_norm": 0.289287805557251, + "learning_rate": 9.281752913489657e-05, + "loss": 0.0366, + "step": 21410 + }, + { + "epoch": 1.403393828212016, + "grad_norm": 0.15369510650634766, + "learning_rate": 9.280898836022222e-05, + "loss": 0.0372, + "step": 21420 + }, + { + "epoch": 1.4040490074035248, + "grad_norm": 0.4607982933521271, + "learning_rate": 9.280044290402126e-05, + "loss": 0.0674, + "step": 21430 + }, + { + "epoch": 1.4047041865950338, + "grad_norm": 0.2595629394054413, + "learning_rate": 9.279189276722821e-05, + "loss": 0.045, + "step": 21440 + }, + { + "epoch": 1.4053593657865426, + "grad_norm": 0.22568494081497192, + "learning_rate": 9.278333795077812e-05, + "loss": 0.0363, + "step": 21450 + }, + { + "epoch": 1.4060145449780515, + "grad_norm": 0.48246607184410095, + "learning_rate": 9.27747784556065e-05, + "loss": 0.0371, + "step": 21460 + }, + { + "epoch": 1.4066697241695603, + "grad_norm": 0.3627689480781555, + "learning_rate": 9.276621428264942e-05, + "loss": 0.0635, + "step": 21470 + }, + { + "epoch": 1.4073249033610693, + "grad_norm": 0.4813936650753021, + "learning_rate": 9.275764543284345e-05, + "loss": 0.058, + "step": 21480 + }, + { + "epoch": 1.4079800825525781, + "grad_norm": 0.331955224275589, + "learning_rate": 9.274907190712566e-05, + "loss": 0.056, + "step": 21490 + }, + { + "epoch": 1.408635261744087, + "grad_norm": 1.7889899015426636, + "learning_rate": 9.274049370643363e-05, + "loss": 0.0556, + "step": 21500 + }, + { + "epoch": 1.409290440935596, + "grad_norm": 0.6421029567718506, + "learning_rate": 9.273191083170547e-05, + "loss": 0.0526, + "step": 21510 + }, + { + "epoch": 1.4099456201271048, + "grad_norm": 0.6586095094680786, + "learning_rate": 9.27233232838798e-05, + "loss": 0.0245, + "step": 21520 + }, + { + "epoch": 1.4106007993186136, + "grad_norm": 0.1660095602273941, + "learning_rate": 9.27147310638957e-05, + "loss": 0.0526, + "step": 21530 + }, + { + "epoch": 1.4112559785101224, + "grad_norm": 0.2792741656303406, + "learning_rate": 9.270613417269286e-05, + "loss": 0.0616, + "step": 21540 + }, + { + "epoch": 1.4119111577016314, + "grad_norm": 0.7687212824821472, + "learning_rate": 9.269753261121138e-05, + "loss": 0.0576, + "step": 21550 + }, + { + "epoch": 1.4125663368931403, + "grad_norm": 0.19411921501159668, + "learning_rate": 9.268892638039194e-05, + "loss": 0.0516, + "step": 21560 + }, + { + "epoch": 1.413221516084649, + "grad_norm": 0.5769250392913818, + "learning_rate": 9.268031548117569e-05, + "loss": 0.0501, + "step": 21570 + }, + { + "epoch": 1.4138766952761581, + "grad_norm": 0.5659435391426086, + "learning_rate": 9.26716999145043e-05, + "loss": 0.05, + "step": 21580 + }, + { + "epoch": 1.414531874467667, + "grad_norm": 0.5248797535896301, + "learning_rate": 9.266307968131998e-05, + "loss": 0.0453, + "step": 21590 + }, + { + "epoch": 1.4151870536591757, + "grad_norm": 0.4260046184062958, + "learning_rate": 9.26544547825654e-05, + "loss": 0.0464, + "step": 21600 + }, + { + "epoch": 1.4158422328506846, + "grad_norm": 0.23428456485271454, + "learning_rate": 9.264582521918376e-05, + "loss": 0.034, + "step": 21610 + }, + { + "epoch": 1.4164974120421936, + "grad_norm": 0.36272919178009033, + "learning_rate": 9.263719099211881e-05, + "loss": 0.0517, + "step": 21620 + }, + { + "epoch": 1.4171525912337024, + "grad_norm": 0.6652068495750427, + "learning_rate": 9.262855210231476e-05, + "loss": 0.0395, + "step": 21630 + }, + { + "epoch": 1.4178077704252112, + "grad_norm": 0.6949242949485779, + "learning_rate": 9.261990855071633e-05, + "loss": 0.039, + "step": 21640 + }, + { + "epoch": 1.4184629496167203, + "grad_norm": 0.41002631187438965, + "learning_rate": 9.261126033826878e-05, + "loss": 0.0415, + "step": 21650 + }, + { + "epoch": 1.419118128808229, + "grad_norm": 0.36301133036613464, + "learning_rate": 9.260260746591786e-05, + "loss": 0.032, + "step": 21660 + }, + { + "epoch": 1.4197733079997379, + "grad_norm": 0.8988819122314453, + "learning_rate": 9.259394993460985e-05, + "loss": 0.0412, + "step": 21670 + }, + { + "epoch": 1.4204284871912467, + "grad_norm": 0.2767716646194458, + "learning_rate": 9.258528774529151e-05, + "loss": 0.0467, + "step": 21680 + }, + { + "epoch": 1.4210836663827557, + "grad_norm": 1.1702696084976196, + "learning_rate": 9.257662089891013e-05, + "loss": 0.0494, + "step": 21690 + }, + { + "epoch": 1.4217388455742646, + "grad_norm": 1.2974841594696045, + "learning_rate": 9.25679493964135e-05, + "loss": 0.0356, + "step": 21700 + }, + { + "epoch": 1.4223940247657734, + "grad_norm": 0.45153525471687317, + "learning_rate": 9.255927323874994e-05, + "loss": 0.0585, + "step": 21710 + }, + { + "epoch": 1.4230492039572824, + "grad_norm": 0.46345922350883484, + "learning_rate": 9.255059242686822e-05, + "loss": 0.0404, + "step": 21720 + }, + { + "epoch": 1.4237043831487912, + "grad_norm": 0.6315253376960754, + "learning_rate": 9.254190696171769e-05, + "loss": 0.0937, + "step": 21730 + }, + { + "epoch": 1.4243595623403, + "grad_norm": 0.8502766489982605, + "learning_rate": 9.25332168442482e-05, + "loss": 0.07, + "step": 21740 + }, + { + "epoch": 1.4250147415318088, + "grad_norm": 0.9661945700645447, + "learning_rate": 9.252452207541004e-05, + "loss": 0.0473, + "step": 21750 + }, + { + "epoch": 1.4256699207233179, + "grad_norm": 0.5983131527900696, + "learning_rate": 9.251582265615409e-05, + "loss": 0.045, + "step": 21760 + }, + { + "epoch": 1.4263250999148267, + "grad_norm": 0.8457139730453491, + "learning_rate": 9.250711858743169e-05, + "loss": 0.0383, + "step": 21770 + }, + { + "epoch": 1.4269802791063355, + "grad_norm": 0.7835913300514221, + "learning_rate": 9.24984098701947e-05, + "loss": 0.0337, + "step": 21780 + }, + { + "epoch": 1.4276354582978446, + "grad_norm": 0.394898921251297, + "learning_rate": 9.248969650539552e-05, + "loss": 0.03, + "step": 21790 + }, + { + "epoch": 1.4282906374893534, + "grad_norm": 0.35497236251831055, + "learning_rate": 9.2480978493987e-05, + "loss": 0.0531, + "step": 21800 + }, + { + "epoch": 1.4289458166808622, + "grad_norm": 0.381708025932312, + "learning_rate": 9.247225583692256e-05, + "loss": 0.0395, + "step": 21810 + }, + { + "epoch": 1.429600995872371, + "grad_norm": 0.7051113247871399, + "learning_rate": 9.246352853515607e-05, + "loss": 0.0476, + "step": 21820 + }, + { + "epoch": 1.43025617506388, + "grad_norm": 0.5554122924804688, + "learning_rate": 9.245479658964194e-05, + "loss": 0.0389, + "step": 21830 + }, + { + "epoch": 1.4309113542553888, + "grad_norm": 0.5428444147109985, + "learning_rate": 9.244606000133507e-05, + "loss": 0.0413, + "step": 21840 + }, + { + "epoch": 1.4315665334468977, + "grad_norm": 0.7312760353088379, + "learning_rate": 9.24373187711909e-05, + "loss": 0.0439, + "step": 21850 + }, + { + "epoch": 1.4322217126384067, + "grad_norm": 0.5303540825843811, + "learning_rate": 9.242857290016537e-05, + "loss": 0.0509, + "step": 21860 + }, + { + "epoch": 1.4328768918299155, + "grad_norm": 0.5781393051147461, + "learning_rate": 9.241982238921488e-05, + "loss": 0.0323, + "step": 21870 + }, + { + "epoch": 1.4335320710214243, + "grad_norm": 0.5753294229507446, + "learning_rate": 9.24110672392964e-05, + "loss": 0.0346, + "step": 21880 + }, + { + "epoch": 1.4341872502129331, + "grad_norm": 0.3768075704574585, + "learning_rate": 9.240230745136737e-05, + "loss": 0.0379, + "step": 21890 + }, + { + "epoch": 1.4348424294044422, + "grad_norm": 0.3979228436946869, + "learning_rate": 9.239354302638575e-05, + "loss": 0.039, + "step": 21900 + }, + { + "epoch": 1.435497608595951, + "grad_norm": 0.5259862542152405, + "learning_rate": 9.238477396531e-05, + "loss": 0.0391, + "step": 21910 + }, + { + "epoch": 1.4361527877874598, + "grad_norm": 0.7616268396377563, + "learning_rate": 9.23760002690991e-05, + "loss": 0.0813, + "step": 21920 + }, + { + "epoch": 1.4368079669789688, + "grad_norm": 0.5573756694793701, + "learning_rate": 9.236722193871252e-05, + "loss": 0.0357, + "step": 21930 + }, + { + "epoch": 1.4374631461704777, + "grad_norm": 0.7570632696151733, + "learning_rate": 9.235843897511023e-05, + "loss": 0.0505, + "step": 21940 + }, + { + "epoch": 1.4381183253619865, + "grad_norm": 0.742358922958374, + "learning_rate": 9.234965137925276e-05, + "loss": 0.0712, + "step": 21950 + }, + { + "epoch": 1.4387735045534953, + "grad_norm": 0.4658840000629425, + "learning_rate": 9.234085915210108e-05, + "loss": 0.044, + "step": 21960 + }, + { + "epoch": 1.4394286837450043, + "grad_norm": 0.5477244853973389, + "learning_rate": 9.23320622946167e-05, + "loss": 0.0371, + "step": 21970 + }, + { + "epoch": 1.4400838629365131, + "grad_norm": 0.6112015247344971, + "learning_rate": 9.232326080776163e-05, + "loss": 0.0666, + "step": 21980 + }, + { + "epoch": 1.440739042128022, + "grad_norm": 0.5408675670623779, + "learning_rate": 9.23144546924984e-05, + "loss": 0.0364, + "step": 21990 + }, + { + "epoch": 1.441394221319531, + "grad_norm": 0.7328258156776428, + "learning_rate": 9.230564394979e-05, + "loss": 0.0463, + "step": 22000 + }, + { + "epoch": 1.4420494005110398, + "grad_norm": 1.6438552141189575, + "learning_rate": 9.22968285806e-05, + "loss": 0.0713, + "step": 22010 + }, + { + "epoch": 1.4427045797025486, + "grad_norm": 0.7503129243850708, + "learning_rate": 9.228800858589242e-05, + "loss": 0.0353, + "step": 22020 + }, + { + "epoch": 1.4433597588940574, + "grad_norm": 0.8305317759513855, + "learning_rate": 9.227918396663179e-05, + "loss": 0.04, + "step": 22030 + }, + { + "epoch": 1.4440149380855665, + "grad_norm": 0.7216565012931824, + "learning_rate": 9.227035472378319e-05, + "loss": 0.0392, + "step": 22040 + }, + { + "epoch": 1.4446701172770753, + "grad_norm": 0.9367771148681641, + "learning_rate": 9.226152085831213e-05, + "loss": 0.0555, + "step": 22050 + }, + { + "epoch": 1.445325296468584, + "grad_norm": 0.5517463684082031, + "learning_rate": 9.22526823711847e-05, + "loss": 0.0631, + "step": 22060 + }, + { + "epoch": 1.4459804756600931, + "grad_norm": 0.3953673541545868, + "learning_rate": 9.224383926336745e-05, + "loss": 0.0535, + "step": 22070 + }, + { + "epoch": 1.446635654851602, + "grad_norm": 0.7760500311851501, + "learning_rate": 9.223499153582744e-05, + "loss": 0.0489, + "step": 22080 + }, + { + "epoch": 1.4472908340431108, + "grad_norm": 0.4552730619907379, + "learning_rate": 9.222613918953226e-05, + "loss": 0.0773, + "step": 22090 + }, + { + "epoch": 1.4479460132346196, + "grad_norm": 0.5019428730010986, + "learning_rate": 9.221728222544999e-05, + "loss": 0.0523, + "step": 22100 + }, + { + "epoch": 1.4486011924261286, + "grad_norm": 0.2063995599746704, + "learning_rate": 9.22084206445492e-05, + "loss": 0.0507, + "step": 22110 + }, + { + "epoch": 1.4492563716176374, + "grad_norm": 0.12322041392326355, + "learning_rate": 9.2199554447799e-05, + "loss": 0.0577, + "step": 22120 + }, + { + "epoch": 1.4499115508091462, + "grad_norm": 0.4888741075992584, + "learning_rate": 9.219068363616897e-05, + "loss": 0.0417, + "step": 22130 + }, + { + "epoch": 1.4505667300006553, + "grad_norm": 0.20596036314964294, + "learning_rate": 9.218180821062919e-05, + "loss": 0.0399, + "step": 22140 + }, + { + "epoch": 1.451221909192164, + "grad_norm": 0.7052624225616455, + "learning_rate": 9.21729281721503e-05, + "loss": 0.0706, + "step": 22150 + }, + { + "epoch": 1.451877088383673, + "grad_norm": 0.39777281880378723, + "learning_rate": 9.216404352170339e-05, + "loss": 0.0546, + "step": 22160 + }, + { + "epoch": 1.4525322675751817, + "grad_norm": 1.0840262174606323, + "learning_rate": 9.215515426026007e-05, + "loss": 0.0656, + "step": 22170 + }, + { + "epoch": 1.4531874467666908, + "grad_norm": 0.2599189877510071, + "learning_rate": 9.214626038879246e-05, + "loss": 0.0372, + "step": 22180 + }, + { + "epoch": 1.4538426259581996, + "grad_norm": 0.34312257170677185, + "learning_rate": 9.21373619082732e-05, + "loss": 0.0371, + "step": 22190 + }, + { + "epoch": 1.4544978051497084, + "grad_norm": 0.6389275193214417, + "learning_rate": 9.212845881967535e-05, + "loss": 0.048, + "step": 22200 + }, + { + "epoch": 1.4551529843412174, + "grad_norm": 0.6524154543876648, + "learning_rate": 9.211955112397262e-05, + "loss": 0.0563, + "step": 22210 + }, + { + "epoch": 1.4558081635327262, + "grad_norm": 0.6666520237922668, + "learning_rate": 9.211063882213909e-05, + "loss": 0.0561, + "step": 22220 + }, + { + "epoch": 1.456463342724235, + "grad_norm": 0.26732149720191956, + "learning_rate": 9.210172191514942e-05, + "loss": 0.0424, + "step": 22230 + }, + { + "epoch": 1.4571185219157439, + "grad_norm": 0.790603518486023, + "learning_rate": 9.209280040397874e-05, + "loss": 0.044, + "step": 22240 + }, + { + "epoch": 1.457773701107253, + "grad_norm": 0.4871838390827179, + "learning_rate": 9.208387428960268e-05, + "loss": 0.0516, + "step": 22250 + }, + { + "epoch": 1.4584288802987617, + "grad_norm": 0.18532903492450714, + "learning_rate": 9.20749435729974e-05, + "loss": 0.0649, + "step": 22260 + }, + { + "epoch": 1.4590840594902705, + "grad_norm": 0.9531288146972656, + "learning_rate": 9.206600825513957e-05, + "loss": 0.0484, + "step": 22270 + }, + { + "epoch": 1.4597392386817796, + "grad_norm": 0.47370675206184387, + "learning_rate": 9.20570683370063e-05, + "loss": 0.0472, + "step": 22280 + }, + { + "epoch": 1.4603944178732884, + "grad_norm": 0.37058156728744507, + "learning_rate": 9.204812381957528e-05, + "loss": 0.0396, + "step": 22290 + }, + { + "epoch": 1.4610495970647972, + "grad_norm": 0.3138844966888428, + "learning_rate": 9.203917470382465e-05, + "loss": 0.0469, + "step": 22300 + }, + { + "epoch": 1.461704776256306, + "grad_norm": 0.4129338562488556, + "learning_rate": 9.203022099073309e-05, + "loss": 0.0349, + "step": 22310 + }, + { + "epoch": 1.462359955447815, + "grad_norm": 0.8772969841957092, + "learning_rate": 9.202126268127976e-05, + "loss": 0.098, + "step": 22320 + }, + { + "epoch": 1.4630151346393239, + "grad_norm": 0.907609224319458, + "learning_rate": 9.20122997764443e-05, + "loss": 0.0429, + "step": 22330 + }, + { + "epoch": 1.4636703138308327, + "grad_norm": 0.42122146487236023, + "learning_rate": 9.200333227720692e-05, + "loss": 0.0801, + "step": 22340 + }, + { + "epoch": 1.4643254930223417, + "grad_norm": 0.5516082644462585, + "learning_rate": 9.199436018454826e-05, + "loss": 0.0451, + "step": 22350 + }, + { + "epoch": 1.4649806722138505, + "grad_norm": 1.024832010269165, + "learning_rate": 9.198538349944952e-05, + "loss": 0.0521, + "step": 22360 + }, + { + "epoch": 1.4656358514053593, + "grad_norm": 0.7088331580162048, + "learning_rate": 9.197640222289234e-05, + "loss": 0.0399, + "step": 22370 + }, + { + "epoch": 1.4662910305968682, + "grad_norm": 0.9267843961715698, + "learning_rate": 9.196741635585895e-05, + "loss": 0.0433, + "step": 22380 + }, + { + "epoch": 1.4669462097883772, + "grad_norm": 0.503842294216156, + "learning_rate": 9.195842589933199e-05, + "loss": 0.0556, + "step": 22390 + }, + { + "epoch": 1.467601388979886, + "grad_norm": 0.7977631688117981, + "learning_rate": 9.194943085429466e-05, + "loss": 0.0705, + "step": 22400 + }, + { + "epoch": 1.4682565681713948, + "grad_norm": 0.6208432912826538, + "learning_rate": 9.194043122173065e-05, + "loss": 0.0485, + "step": 22410 + }, + { + "epoch": 1.4689117473629039, + "grad_norm": 0.7421457171440125, + "learning_rate": 9.193142700262413e-05, + "loss": 0.0403, + "step": 22420 + }, + { + "epoch": 1.4695669265544127, + "grad_norm": 0.5618040561676025, + "learning_rate": 9.192241819795979e-05, + "loss": 0.0679, + "step": 22430 + }, + { + "epoch": 1.4702221057459215, + "grad_norm": 0.15768663585186005, + "learning_rate": 9.191340480872284e-05, + "loss": 0.0668, + "step": 22440 + }, + { + "epoch": 1.4708772849374303, + "grad_norm": 0.7350854277610779, + "learning_rate": 9.190438683589895e-05, + "loss": 0.0325, + "step": 22450 + }, + { + "epoch": 1.4715324641289393, + "grad_norm": 0.12172713875770569, + "learning_rate": 9.189536428047432e-05, + "loss": 0.0297, + "step": 22460 + }, + { + "epoch": 1.4721876433204482, + "grad_norm": 1.063808798789978, + "learning_rate": 9.188633714343564e-05, + "loss": 0.0632, + "step": 22470 + }, + { + "epoch": 1.472842822511957, + "grad_norm": 0.3062632977962494, + "learning_rate": 9.18773054257701e-05, + "loss": 0.0413, + "step": 22480 + }, + { + "epoch": 1.473498001703466, + "grad_norm": 0.478861540555954, + "learning_rate": 9.18682691284654e-05, + "loss": 0.0512, + "step": 22490 + }, + { + "epoch": 1.4741531808949748, + "grad_norm": 0.9499432444572449, + "learning_rate": 9.185922825250974e-05, + "loss": 0.0612, + "step": 22500 + }, + { + "epoch": 1.4748083600864836, + "grad_norm": 0.1659020185470581, + "learning_rate": 9.185018279889181e-05, + "loss": 0.0364, + "step": 22510 + }, + { + "epoch": 1.4754635392779925, + "grad_norm": 0.2550235092639923, + "learning_rate": 9.184113276860082e-05, + "loss": 0.0333, + "step": 22520 + }, + { + "epoch": 1.4761187184695015, + "grad_norm": 0.6656308770179749, + "learning_rate": 9.183207816262645e-05, + "loss": 0.0507, + "step": 22530 + }, + { + "epoch": 1.4767738976610103, + "grad_norm": 0.6929846405982971, + "learning_rate": 9.182301898195891e-05, + "loss": 0.0357, + "step": 22540 + }, + { + "epoch": 1.4774290768525191, + "grad_norm": 0.3787357807159424, + "learning_rate": 9.181395522758889e-05, + "loss": 0.0355, + "step": 22550 + }, + { + "epoch": 1.4780842560440282, + "grad_norm": 0.7073647379875183, + "learning_rate": 9.180488690050759e-05, + "loss": 0.0505, + "step": 22560 + }, + { + "epoch": 1.478739435235537, + "grad_norm": 0.6355065107345581, + "learning_rate": 9.179581400170671e-05, + "loss": 0.0366, + "step": 22570 + }, + { + "epoch": 1.4793946144270458, + "grad_norm": 0.5851870775222778, + "learning_rate": 9.178673653217845e-05, + "loss": 0.0501, + "step": 22580 + }, + { + "epoch": 1.4800497936185546, + "grad_norm": 0.4802267253398895, + "learning_rate": 9.177765449291551e-05, + "loss": 0.0993, + "step": 22590 + }, + { + "epoch": 1.4807049728100636, + "grad_norm": 0.5448585748672485, + "learning_rate": 9.176856788491109e-05, + "loss": 0.0386, + "step": 22600 + }, + { + "epoch": 1.4813601520015724, + "grad_norm": 0.1711624562740326, + "learning_rate": 9.175947670915887e-05, + "loss": 0.0373, + "step": 22610 + }, + { + "epoch": 1.4820153311930813, + "grad_norm": 0.2691153287887573, + "learning_rate": 9.175038096665309e-05, + "loss": 0.0373, + "step": 22620 + }, + { + "epoch": 1.4826705103845903, + "grad_norm": 0.7521606087684631, + "learning_rate": 9.17412806583884e-05, + "loss": 0.0373, + "step": 22630 + }, + { + "epoch": 1.4833256895760991, + "grad_norm": 0.20802512764930725, + "learning_rate": 9.173217578536002e-05, + "loss": 0.0589, + "step": 22640 + }, + { + "epoch": 1.483980868767608, + "grad_norm": 0.41859474778175354, + "learning_rate": 9.172306634856362e-05, + "loss": 0.0496, + "step": 22650 + }, + { + "epoch": 1.4846360479591167, + "grad_norm": 0.6892299056053162, + "learning_rate": 9.171395234899545e-05, + "loss": 0.0427, + "step": 22660 + }, + { + "epoch": 1.4852912271506256, + "grad_norm": 0.27950870990753174, + "learning_rate": 9.170483378765214e-05, + "loss": 0.0552, + "step": 22670 + }, + { + "epoch": 1.4859464063421346, + "grad_norm": 1.2762435674667358, + "learning_rate": 9.169571066553091e-05, + "loss": 0.0807, + "step": 22680 + }, + { + "epoch": 1.4866015855336434, + "grad_norm": 0.4809858202934265, + "learning_rate": 9.168658298362946e-05, + "loss": 0.0414, + "step": 22690 + }, + { + "epoch": 1.4872567647251524, + "grad_norm": 0.8804682493209839, + "learning_rate": 9.167745074294598e-05, + "loss": 0.0392, + "step": 22700 + }, + { + "epoch": 1.4879119439166613, + "grad_norm": 0.8898096680641174, + "learning_rate": 9.166831394447913e-05, + "loss": 0.0661, + "step": 22710 + }, + { + "epoch": 1.48856712310817, + "grad_norm": 0.33011502027511597, + "learning_rate": 9.165917258922812e-05, + "loss": 0.0424, + "step": 22720 + }, + { + "epoch": 1.489222302299679, + "grad_norm": 0.5121455788612366, + "learning_rate": 9.165002667819262e-05, + "loss": 0.0426, + "step": 22730 + }, + { + "epoch": 1.4898774814911877, + "grad_norm": 0.4288701117038727, + "learning_rate": 9.164087621237282e-05, + "loss": 0.0535, + "step": 22740 + }, + { + "epoch": 1.4905326606826967, + "grad_norm": 0.615239679813385, + "learning_rate": 9.163172119276942e-05, + "loss": 0.0513, + "step": 22750 + }, + { + "epoch": 1.4911878398742056, + "grad_norm": 0.8748026490211487, + "learning_rate": 9.162256162038358e-05, + "loss": 0.0477, + "step": 22760 + }, + { + "epoch": 1.4918430190657146, + "grad_norm": 0.41284704208374023, + "learning_rate": 9.161339749621698e-05, + "loss": 0.0488, + "step": 22770 + }, + { + "epoch": 1.4924981982572234, + "grad_norm": 0.13924846053123474, + "learning_rate": 9.160422882127177e-05, + "loss": 0.0468, + "step": 22780 + }, + { + "epoch": 1.4931533774487322, + "grad_norm": 0.7337716221809387, + "learning_rate": 9.159505559655069e-05, + "loss": 0.0791, + "step": 22790 + }, + { + "epoch": 1.493808556640241, + "grad_norm": 0.2260930985212326, + "learning_rate": 9.158587782305684e-05, + "loss": 0.0509, + "step": 22800 + }, + { + "epoch": 1.4944637358317499, + "grad_norm": 0.4256834089756012, + "learning_rate": 9.157669550179391e-05, + "loss": 0.0481, + "step": 22810 + }, + { + "epoch": 1.4951189150232589, + "grad_norm": 0.5304692387580872, + "learning_rate": 9.156750863376609e-05, + "loss": 0.0444, + "step": 22820 + }, + { + "epoch": 1.4957740942147677, + "grad_norm": 1.0151493549346924, + "learning_rate": 9.155831721997801e-05, + "loss": 0.0798, + "step": 22830 + }, + { + "epoch": 1.4964292734062767, + "grad_norm": 0.2620948553085327, + "learning_rate": 9.154912126143484e-05, + "loss": 0.0403, + "step": 22840 + }, + { + "epoch": 1.4970844525977856, + "grad_norm": 0.6577702164649963, + "learning_rate": 9.153992075914224e-05, + "loss": 0.0382, + "step": 22850 + }, + { + "epoch": 1.4977396317892944, + "grad_norm": 1.962684154510498, + "learning_rate": 9.153071571410635e-05, + "loss": 0.0594, + "step": 22860 + }, + { + "epoch": 1.4983948109808032, + "grad_norm": 0.19149509072303772, + "learning_rate": 9.152150612733384e-05, + "loss": 0.0616, + "step": 22870 + }, + { + "epoch": 1.499049990172312, + "grad_norm": 0.5745983123779297, + "learning_rate": 9.151229199983184e-05, + "loss": 0.0597, + "step": 22880 + }, + { + "epoch": 1.499705169363821, + "grad_norm": 0.799216091632843, + "learning_rate": 9.150307333260802e-05, + "loss": 0.0322, + "step": 22890 + }, + { + "epoch": 1.5003603485553298, + "grad_norm": 0.7821118235588074, + "learning_rate": 9.149385012667048e-05, + "loss": 0.0623, + "step": 22900 + }, + { + "epoch": 1.5010155277468389, + "grad_norm": 0.7783225178718567, + "learning_rate": 9.148462238302788e-05, + "loss": 0.0571, + "step": 22910 + }, + { + "epoch": 1.5016707069383477, + "grad_norm": 0.5730087757110596, + "learning_rate": 9.147539010268936e-05, + "loss": 0.0468, + "step": 22920 + }, + { + "epoch": 1.5023258861298565, + "grad_norm": 0.30091017484664917, + "learning_rate": 9.14661532866645e-05, + "loss": 0.0346, + "step": 22930 + }, + { + "epoch": 1.5029810653213653, + "grad_norm": 0.14296481013298035, + "learning_rate": 9.145691193596348e-05, + "loss": 0.035, + "step": 22940 + }, + { + "epoch": 1.5036362445128741, + "grad_norm": 0.7347245216369629, + "learning_rate": 9.144766605159691e-05, + "loss": 0.0617, + "step": 22950 + }, + { + "epoch": 1.5042914237043832, + "grad_norm": 1.4966819286346436, + "learning_rate": 9.14384156345759e-05, + "loss": 0.0805, + "step": 22960 + }, + { + "epoch": 1.504946602895892, + "grad_norm": 0.8648755550384521, + "learning_rate": 9.142916068591204e-05, + "loss": 0.0449, + "step": 22970 + }, + { + "epoch": 1.505601782087401, + "grad_norm": 0.4482906460762024, + "learning_rate": 9.141990120661746e-05, + "loss": 0.0428, + "step": 22980 + }, + { + "epoch": 1.5062569612789098, + "grad_norm": 0.23966048657894135, + "learning_rate": 9.141063719770475e-05, + "loss": 0.0485, + "step": 22990 + }, + { + "epoch": 1.5069121404704187, + "grad_norm": 0.18561117351055145, + "learning_rate": 9.140136866018704e-05, + "loss": 0.0239, + "step": 23000 + }, + { + "epoch": 1.5075673196619275, + "grad_norm": 0.5096158385276794, + "learning_rate": 9.139209559507788e-05, + "loss": 0.0532, + "step": 23010 + }, + { + "epoch": 1.5082224988534363, + "grad_norm": 0.48738017678260803, + "learning_rate": 9.13828180033914e-05, + "loss": 0.027, + "step": 23020 + }, + { + "epoch": 1.5088776780449453, + "grad_norm": 1.350507140159607, + "learning_rate": 9.137353588614212e-05, + "loss": 0.0611, + "step": 23030 + }, + { + "epoch": 1.5095328572364541, + "grad_norm": 0.2869527041912079, + "learning_rate": 9.136424924434519e-05, + "loss": 0.0186, + "step": 23040 + }, + { + "epoch": 1.5101880364279632, + "grad_norm": 0.29461273550987244, + "learning_rate": 9.135495807901615e-05, + "loss": 0.0798, + "step": 23050 + }, + { + "epoch": 1.510843215619472, + "grad_norm": 0.31841838359832764, + "learning_rate": 9.134566239117108e-05, + "loss": 0.0345, + "step": 23060 + }, + { + "epoch": 1.5114983948109808, + "grad_norm": 0.8371802568435669, + "learning_rate": 9.13363621818265e-05, + "loss": 0.066, + "step": 23070 + }, + { + "epoch": 1.5121535740024896, + "grad_norm": 0.4104580283164978, + "learning_rate": 9.132705745199953e-05, + "loss": 0.0341, + "step": 23080 + }, + { + "epoch": 1.5128087531939984, + "grad_norm": 1.1237342357635498, + "learning_rate": 9.131774820270768e-05, + "loss": 0.075, + "step": 23090 + }, + { + "epoch": 1.5134639323855075, + "grad_norm": 0.7499944567680359, + "learning_rate": 9.130843443496901e-05, + "loss": 0.0869, + "step": 23100 + }, + { + "epoch": 1.5141191115770163, + "grad_norm": 0.34935349225997925, + "learning_rate": 9.129911614980206e-05, + "loss": 0.0302, + "step": 23110 + }, + { + "epoch": 1.5147742907685253, + "grad_norm": 0.30936071276664734, + "learning_rate": 9.128979334822584e-05, + "loss": 0.081, + "step": 23120 + }, + { + "epoch": 1.5154294699600341, + "grad_norm": 0.44806969165802, + "learning_rate": 9.128046603125992e-05, + "loss": 0.0341, + "step": 23130 + }, + { + "epoch": 1.516084649151543, + "grad_norm": 0.6216328144073486, + "learning_rate": 9.12711341999243e-05, + "loss": 0.0696, + "step": 23140 + }, + { + "epoch": 1.5167398283430518, + "grad_norm": 0.2723061442375183, + "learning_rate": 9.12617978552395e-05, + "loss": 0.0649, + "step": 23150 + }, + { + "epoch": 1.5173950075345606, + "grad_norm": 1.4701049327850342, + "learning_rate": 9.12524569982265e-05, + "loss": 0.0517, + "step": 23160 + }, + { + "epoch": 1.5180501867260696, + "grad_norm": 0.5239159464836121, + "learning_rate": 9.124311162990684e-05, + "loss": 0.0537, + "step": 23170 + }, + { + "epoch": 1.5187053659175784, + "grad_norm": 0.3743836283683777, + "learning_rate": 9.12337617513025e-05, + "loss": 0.0665, + "step": 23180 + }, + { + "epoch": 1.5193605451090875, + "grad_norm": 0.40488678216934204, + "learning_rate": 9.122440736343596e-05, + "loss": 0.0311, + "step": 23190 + }, + { + "epoch": 1.5200157243005963, + "grad_norm": 0.5871321558952332, + "learning_rate": 9.12150484673302e-05, + "loss": 0.0474, + "step": 23200 + }, + { + "epoch": 1.520670903492105, + "grad_norm": 0.7310592532157898, + "learning_rate": 9.120568506400873e-05, + "loss": 0.0432, + "step": 23210 + }, + { + "epoch": 1.521326082683614, + "grad_norm": 0.5552395582199097, + "learning_rate": 9.119631715449548e-05, + "loss": 0.0582, + "step": 23220 + }, + { + "epoch": 1.5219812618751227, + "grad_norm": 0.14355522394180298, + "learning_rate": 9.118694473981493e-05, + "loss": 0.0511, + "step": 23230 + }, + { + "epoch": 1.5226364410666318, + "grad_norm": 0.5419275760650635, + "learning_rate": 9.117756782099203e-05, + "loss": 0.0335, + "step": 23240 + }, + { + "epoch": 1.5232916202581406, + "grad_norm": 0.6161248087882996, + "learning_rate": 9.11681863990522e-05, + "loss": 0.0366, + "step": 23250 + }, + { + "epoch": 1.5239467994496496, + "grad_norm": 0.36884352564811707, + "learning_rate": 9.115880047502142e-05, + "loss": 0.0354, + "step": 23260 + }, + { + "epoch": 1.5246019786411584, + "grad_norm": 1.4527316093444824, + "learning_rate": 9.114941004992609e-05, + "loss": 0.063, + "step": 23270 + }, + { + "epoch": 1.5252571578326672, + "grad_norm": 0.7699277400970459, + "learning_rate": 9.114001512479317e-05, + "loss": 0.0843, + "step": 23280 + }, + { + "epoch": 1.525912337024176, + "grad_norm": 0.3466300666332245, + "learning_rate": 9.113061570065003e-05, + "loss": 0.0489, + "step": 23290 + }, + { + "epoch": 1.5265675162156849, + "grad_norm": 0.18522466719150543, + "learning_rate": 9.112121177852459e-05, + "loss": 0.0538, + "step": 23300 + }, + { + "epoch": 1.527222695407194, + "grad_norm": 1.0917330980300903, + "learning_rate": 9.111180335944527e-05, + "loss": 0.0619, + "step": 23310 + }, + { + "epoch": 1.5278778745987027, + "grad_norm": 0.969989538192749, + "learning_rate": 9.110239044444093e-05, + "loss": 0.0509, + "step": 23320 + }, + { + "epoch": 1.5285330537902118, + "grad_norm": 0.5273010730743408, + "learning_rate": 9.109297303454099e-05, + "loss": 0.0508, + "step": 23330 + }, + { + "epoch": 1.5291882329817206, + "grad_norm": 1.119881510734558, + "learning_rate": 9.108355113077526e-05, + "loss": 0.0496, + "step": 23340 + }, + { + "epoch": 1.5298434121732294, + "grad_norm": 0.3655964434146881, + "learning_rate": 9.107412473417419e-05, + "loss": 0.0368, + "step": 23350 + }, + { + "epoch": 1.5304985913647382, + "grad_norm": 0.1789465993642807, + "learning_rate": 9.106469384576858e-05, + "loss": 0.0408, + "step": 23360 + }, + { + "epoch": 1.531153770556247, + "grad_norm": 0.36286887526512146, + "learning_rate": 9.105525846658978e-05, + "loss": 0.0611, + "step": 23370 + }, + { + "epoch": 1.531808949747756, + "grad_norm": 0.2229175865650177, + "learning_rate": 9.104581859766965e-05, + "loss": 0.0325, + "step": 23380 + }, + { + "epoch": 1.5324641289392649, + "grad_norm": 0.3101978003978729, + "learning_rate": 9.10363742400405e-05, + "loss": 0.0336, + "step": 23390 + }, + { + "epoch": 1.533119308130774, + "grad_norm": 0.979884684085846, + "learning_rate": 9.102692539473518e-05, + "loss": 0.0599, + "step": 23400 + }, + { + "epoch": 1.5337744873222827, + "grad_norm": 0.43575650453567505, + "learning_rate": 9.101747206278697e-05, + "loss": 0.0801, + "step": 23410 + }, + { + "epoch": 1.5344296665137915, + "grad_norm": 0.9095799922943115, + "learning_rate": 9.100801424522968e-05, + "loss": 0.0606, + "step": 23420 + }, + { + "epoch": 1.5350848457053003, + "grad_norm": 0.846137285232544, + "learning_rate": 9.099855194309762e-05, + "loss": 0.0711, + "step": 23430 + }, + { + "epoch": 1.5357400248968092, + "grad_norm": 0.4791010618209839, + "learning_rate": 9.098908515742554e-05, + "loss": 0.0301, + "step": 23440 + }, + { + "epoch": 1.5363952040883182, + "grad_norm": 0.7321571111679077, + "learning_rate": 9.097961388924873e-05, + "loss": 0.0634, + "step": 23450 + }, + { + "epoch": 1.537050383279827, + "grad_norm": 0.5872757434844971, + "learning_rate": 9.097013813960298e-05, + "loss": 0.0558, + "step": 23460 + }, + { + "epoch": 1.537705562471336, + "grad_norm": 1.0396829843521118, + "learning_rate": 9.09606579095245e-05, + "loss": 0.0514, + "step": 23470 + }, + { + "epoch": 1.5383607416628449, + "grad_norm": 0.850864589214325, + "learning_rate": 9.095117320005008e-05, + "loss": 0.048, + "step": 23480 + }, + { + "epoch": 1.5390159208543537, + "grad_norm": 1.1824549436569214, + "learning_rate": 9.094168401221691e-05, + "loss": 0.0424, + "step": 23490 + }, + { + "epoch": 1.5396711000458625, + "grad_norm": 0.16907621920108795, + "learning_rate": 9.093219034706273e-05, + "loss": 0.0624, + "step": 23500 + }, + { + "epoch": 1.5403262792373713, + "grad_norm": 0.9001042246818542, + "learning_rate": 9.092269220562577e-05, + "loss": 0.0506, + "step": 23510 + }, + { + "epoch": 1.5409814584288803, + "grad_norm": 0.7618017792701721, + "learning_rate": 9.09131895889447e-05, + "loss": 0.0688, + "step": 23520 + }, + { + "epoch": 1.5416366376203892, + "grad_norm": 0.4519120752811432, + "learning_rate": 9.090368249805873e-05, + "loss": 0.0773, + "step": 23530 + }, + { + "epoch": 1.5422918168118982, + "grad_norm": 0.5371301770210266, + "learning_rate": 9.089417093400754e-05, + "loss": 0.0413, + "step": 23540 + }, + { + "epoch": 1.542946996003407, + "grad_norm": 0.25322651863098145, + "learning_rate": 9.088465489783131e-05, + "loss": 0.0266, + "step": 23550 + }, + { + "epoch": 1.5436021751949158, + "grad_norm": 0.5879925489425659, + "learning_rate": 9.087513439057068e-05, + "loss": 0.0824, + "step": 23560 + }, + { + "epoch": 1.5442573543864246, + "grad_norm": 0.8594334125518799, + "learning_rate": 9.08656094132668e-05, + "loss": 0.05, + "step": 23570 + }, + { + "epoch": 1.5449125335779335, + "grad_norm": 0.16600556671619415, + "learning_rate": 9.085607996696134e-05, + "loss": 0.0255, + "step": 23580 + }, + { + "epoch": 1.5455677127694425, + "grad_norm": 1.1229708194732666, + "learning_rate": 9.084654605269639e-05, + "loss": 0.1394, + "step": 23590 + }, + { + "epoch": 1.5462228919609513, + "grad_norm": 0.5240740180015564, + "learning_rate": 9.083700767151457e-05, + "loss": 0.0325, + "step": 23600 + }, + { + "epoch": 1.5468780711524603, + "grad_norm": 0.40403810143470764, + "learning_rate": 9.082746482445898e-05, + "loss": 0.0598, + "step": 23610 + }, + { + "epoch": 1.5475332503439692, + "grad_norm": 0.9203404784202576, + "learning_rate": 9.081791751257325e-05, + "loss": 0.0451, + "step": 23620 + }, + { + "epoch": 1.548188429535478, + "grad_norm": 0.5234010815620422, + "learning_rate": 9.080836573690142e-05, + "loss": 0.0474, + "step": 23630 + }, + { + "epoch": 1.5488436087269868, + "grad_norm": 0.9479982852935791, + "learning_rate": 9.079880949848805e-05, + "loss": 0.0462, + "step": 23640 + }, + { + "epoch": 1.5494987879184956, + "grad_norm": 0.46660560369491577, + "learning_rate": 9.078924879837822e-05, + "loss": 0.0296, + "step": 23650 + }, + { + "epoch": 1.5501539671100046, + "grad_norm": 0.5690541863441467, + "learning_rate": 9.077968363761747e-05, + "loss": 0.0494, + "step": 23660 + }, + { + "epoch": 1.5508091463015135, + "grad_norm": 0.5170366764068604, + "learning_rate": 9.077011401725182e-05, + "loss": 0.0405, + "step": 23670 + }, + { + "epoch": 1.5514643254930225, + "grad_norm": 0.4256380498409271, + "learning_rate": 9.07605399383278e-05, + "loss": 0.0421, + "step": 23680 + }, + { + "epoch": 1.5521195046845313, + "grad_norm": 0.1323145031929016, + "learning_rate": 9.075096140189243e-05, + "loss": 0.0217, + "step": 23690 + }, + { + "epoch": 1.5527746838760401, + "grad_norm": 0.356315016746521, + "learning_rate": 9.074137840899318e-05, + "loss": 0.0336, + "step": 23700 + }, + { + "epoch": 1.553429863067549, + "grad_norm": 0.4701589345932007, + "learning_rate": 9.073179096067804e-05, + "loss": 0.0297, + "step": 23710 + }, + { + "epoch": 1.5540850422590577, + "grad_norm": 0.8257561922073364, + "learning_rate": 9.072219905799549e-05, + "loss": 0.0386, + "step": 23720 + }, + { + "epoch": 1.5547402214505668, + "grad_norm": 0.5331418514251709, + "learning_rate": 9.071260270199447e-05, + "loss": 0.037, + "step": 23730 + }, + { + "epoch": 1.5553954006420756, + "grad_norm": 0.5066092014312744, + "learning_rate": 9.070300189372441e-05, + "loss": 0.0438, + "step": 23740 + }, + { + "epoch": 1.5560505798335846, + "grad_norm": 0.31607764959335327, + "learning_rate": 9.069339663423528e-05, + "loss": 0.0456, + "step": 23750 + }, + { + "epoch": 1.5567057590250934, + "grad_norm": 0.2263612151145935, + "learning_rate": 9.068378692457747e-05, + "loss": 0.0491, + "step": 23760 + }, + { + "epoch": 1.5573609382166023, + "grad_norm": 1.398116111755371, + "learning_rate": 9.067417276580189e-05, + "loss": 0.0863, + "step": 23770 + }, + { + "epoch": 1.558016117408111, + "grad_norm": 0.3090342581272125, + "learning_rate": 9.066455415895993e-05, + "loss": 0.0274, + "step": 23780 + }, + { + "epoch": 1.55867129659962, + "grad_norm": 0.48877525329589844, + "learning_rate": 9.065493110510346e-05, + "loss": 0.0253, + "step": 23790 + }, + { + "epoch": 1.5593264757911287, + "grad_norm": 1.21426260471344, + "learning_rate": 9.064530360528484e-05, + "loss": 0.0599, + "step": 23800 + }, + { + "epoch": 1.5599816549826377, + "grad_norm": 0.3435448110103607, + "learning_rate": 9.063567166055695e-05, + "loss": 0.0299, + "step": 23810 + }, + { + "epoch": 1.5606368341741468, + "grad_norm": 0.8243012428283691, + "learning_rate": 9.062603527197308e-05, + "loss": 0.0549, + "step": 23820 + }, + { + "epoch": 1.5612920133656556, + "grad_norm": 0.739680826663971, + "learning_rate": 9.06163944405871e-05, + "loss": 0.0547, + "step": 23830 + }, + { + "epoch": 1.5619471925571644, + "grad_norm": 0.13872525095939636, + "learning_rate": 9.060674916745327e-05, + "loss": 0.0373, + "step": 23840 + }, + { + "epoch": 1.5626023717486732, + "grad_norm": 0.8693081736564636, + "learning_rate": 9.05970994536264e-05, + "loss": 0.0326, + "step": 23850 + }, + { + "epoch": 1.563257550940182, + "grad_norm": 0.37575799226760864, + "learning_rate": 9.05874453001618e-05, + "loss": 0.0364, + "step": 23860 + }, + { + "epoch": 1.5639127301316909, + "grad_norm": 0.2930029332637787, + "learning_rate": 9.057778670811517e-05, + "loss": 0.0279, + "step": 23870 + }, + { + "epoch": 1.5645679093232, + "grad_norm": 0.8923213481903076, + "learning_rate": 9.056812367854281e-05, + "loss": 0.082, + "step": 23880 + }, + { + "epoch": 1.565223088514709, + "grad_norm": 0.7879847884178162, + "learning_rate": 9.055845621250143e-05, + "loss": 0.0389, + "step": 23890 + }, + { + "epoch": 1.5658782677062177, + "grad_norm": 0.5827088356018066, + "learning_rate": 9.054878431104825e-05, + "loss": 0.0595, + "step": 23900 + }, + { + "epoch": 1.5665334468977266, + "grad_norm": 0.2418992668390274, + "learning_rate": 9.0539107975241e-05, + "loss": 0.041, + "step": 23910 + }, + { + "epoch": 1.5671886260892354, + "grad_norm": 0.3549078404903412, + "learning_rate": 9.052942720613784e-05, + "loss": 0.0574, + "step": 23920 + }, + { + "epoch": 1.5678438052807442, + "grad_norm": 0.4062061905860901, + "learning_rate": 9.051974200479745e-05, + "loss": 0.0637, + "step": 23930 + }, + { + "epoch": 1.568498984472253, + "grad_norm": 0.30342257022857666, + "learning_rate": 9.051005237227901e-05, + "loss": 0.0258, + "step": 23940 + }, + { + "epoch": 1.569154163663762, + "grad_norm": 1.3676732778549194, + "learning_rate": 9.050035830964215e-05, + "loss": 0.0618, + "step": 23950 + }, + { + "epoch": 1.569809342855271, + "grad_norm": 0.6575399041175842, + "learning_rate": 9.049065981794698e-05, + "loss": 0.0712, + "step": 23960 + }, + { + "epoch": 1.5704645220467799, + "grad_norm": 0.5742297172546387, + "learning_rate": 9.048095689825414e-05, + "loss": 0.0553, + "step": 23970 + }, + { + "epoch": 1.5711197012382887, + "grad_norm": 0.8058423399925232, + "learning_rate": 9.047124955162472e-05, + "loss": 0.0511, + "step": 23980 + }, + { + "epoch": 1.5717748804297975, + "grad_norm": 1.0323023796081543, + "learning_rate": 9.046153777912028e-05, + "loss": 0.0672, + "step": 23990 + }, + { + "epoch": 1.5724300596213063, + "grad_norm": 0.898012638092041, + "learning_rate": 9.045182158180292e-05, + "loss": 0.1111, + "step": 24000 + }, + { + "epoch": 1.5730852388128151, + "grad_norm": 0.3472115397453308, + "learning_rate": 9.044210096073516e-05, + "loss": 0.0292, + "step": 24010 + }, + { + "epoch": 1.5737404180043242, + "grad_norm": 0.7961719036102295, + "learning_rate": 9.043237591698004e-05, + "loss": 0.0405, + "step": 24020 + }, + { + "epoch": 1.5743955971958332, + "grad_norm": 0.539855420589447, + "learning_rate": 9.04226464516011e-05, + "loss": 0.0532, + "step": 24030 + }, + { + "epoch": 1.575050776387342, + "grad_norm": 0.8463633060455322, + "learning_rate": 9.041291256566229e-05, + "loss": 0.0445, + "step": 24040 + }, + { + "epoch": 1.5757059555788508, + "grad_norm": 0.63946932554245, + "learning_rate": 9.040317426022814e-05, + "loss": 0.0367, + "step": 24050 + }, + { + "epoch": 1.5763611347703597, + "grad_norm": 0.35942205786705017, + "learning_rate": 9.03934315363636e-05, + "loss": 0.0437, + "step": 24060 + }, + { + "epoch": 1.5770163139618685, + "grad_norm": 0.42907243967056274, + "learning_rate": 9.038368439513409e-05, + "loss": 0.0372, + "step": 24070 + }, + { + "epoch": 1.5776714931533773, + "grad_norm": 0.14895644783973694, + "learning_rate": 9.03739328376056e-05, + "loss": 0.0416, + "step": 24080 + }, + { + "epoch": 1.5783266723448863, + "grad_norm": 1.7793949842453003, + "learning_rate": 9.036417686484451e-05, + "loss": 0.053, + "step": 24090 + }, + { + "epoch": 1.5789818515363954, + "grad_norm": 0.37695667147636414, + "learning_rate": 9.035441647791773e-05, + "loss": 0.046, + "step": 24100 + }, + { + "epoch": 1.5796370307279042, + "grad_norm": 0.5303083658218384, + "learning_rate": 9.034465167789263e-05, + "loss": 0.0397, + "step": 24110 + }, + { + "epoch": 1.580292209919413, + "grad_norm": 0.17751863598823547, + "learning_rate": 9.033488246583706e-05, + "loss": 0.0484, + "step": 24120 + }, + { + "epoch": 1.5809473891109218, + "grad_norm": 0.5774486064910889, + "learning_rate": 9.032510884281941e-05, + "loss": 0.0512, + "step": 24130 + }, + { + "epoch": 1.5816025683024306, + "grad_norm": 0.28765177726745605, + "learning_rate": 9.031533080990848e-05, + "loss": 0.034, + "step": 24140 + }, + { + "epoch": 1.5822577474939394, + "grad_norm": 0.7285279035568237, + "learning_rate": 9.030554836817358e-05, + "loss": 0.0596, + "step": 24150 + }, + { + "epoch": 1.5829129266854485, + "grad_norm": 0.39904066920280457, + "learning_rate": 9.029576151868451e-05, + "loss": 0.0426, + "step": 24160 + }, + { + "epoch": 1.5835681058769575, + "grad_norm": 0.6243024468421936, + "learning_rate": 9.028597026251155e-05, + "loss": 0.0534, + "step": 24170 + }, + { + "epoch": 1.5842232850684663, + "grad_norm": 0.4650101363658905, + "learning_rate": 9.027617460072547e-05, + "loss": 0.0592, + "step": 24180 + }, + { + "epoch": 1.5848784642599751, + "grad_norm": 0.20118722319602966, + "learning_rate": 9.026637453439745e-05, + "loss": 0.0375, + "step": 24190 + }, + { + "epoch": 1.585533643451484, + "grad_norm": 0.4446292817592621, + "learning_rate": 9.025657006459927e-05, + "loss": 0.0606, + "step": 24200 + }, + { + "epoch": 1.5861888226429928, + "grad_norm": 0.18488425016403198, + "learning_rate": 9.024676119240311e-05, + "loss": 0.0599, + "step": 24210 + }, + { + "epoch": 1.5868440018345016, + "grad_norm": 0.540084183216095, + "learning_rate": 9.023694791888166e-05, + "loss": 0.0608, + "step": 24220 + }, + { + "epoch": 1.5874991810260106, + "grad_norm": 0.5396419763565063, + "learning_rate": 9.022713024510808e-05, + "loss": 0.0374, + "step": 24230 + }, + { + "epoch": 1.5881543602175197, + "grad_norm": 1.0908150672912598, + "learning_rate": 9.021730817215601e-05, + "loss": 0.0521, + "step": 24240 + }, + { + "epoch": 1.5888095394090285, + "grad_norm": 0.569530189037323, + "learning_rate": 9.02074817010996e-05, + "loss": 0.0519, + "step": 24250 + }, + { + "epoch": 1.5894647186005373, + "grad_norm": 0.892227292060852, + "learning_rate": 9.019765083301342e-05, + "loss": 0.0381, + "step": 24260 + }, + { + "epoch": 1.590119897792046, + "grad_norm": 0.16959717869758606, + "learning_rate": 9.01878155689726e-05, + "loss": 0.042, + "step": 24270 + }, + { + "epoch": 1.590775076983555, + "grad_norm": 0.2683074474334717, + "learning_rate": 9.017797591005268e-05, + "loss": 0.0499, + "step": 24280 + }, + { + "epoch": 1.5914302561750637, + "grad_norm": 0.6717470288276672, + "learning_rate": 9.016813185732972e-05, + "loss": 0.0431, + "step": 24290 + }, + { + "epoch": 1.5920854353665728, + "grad_norm": 0.4068327248096466, + "learning_rate": 9.015828341188027e-05, + "loss": 0.058, + "step": 24300 + }, + { + "epoch": 1.5927406145580816, + "grad_norm": 0.9833799600601196, + "learning_rate": 9.01484305747813e-05, + "loss": 0.0674, + "step": 24310 + }, + { + "epoch": 1.5933957937495906, + "grad_norm": 1.2266674041748047, + "learning_rate": 9.013857334711033e-05, + "loss": 0.0847, + "step": 24320 + }, + { + "epoch": 1.5940509729410994, + "grad_norm": 0.5754451751708984, + "learning_rate": 9.012871172994534e-05, + "loss": 0.0437, + "step": 24330 + }, + { + "epoch": 1.5947061521326082, + "grad_norm": 0.5976289510726929, + "learning_rate": 9.011884572436476e-05, + "loss": 0.0755, + "step": 24340 + }, + { + "epoch": 1.595361331324117, + "grad_norm": 0.5995612740516663, + "learning_rate": 9.010897533144754e-05, + "loss": 0.0483, + "step": 24350 + }, + { + "epoch": 1.5960165105156259, + "grad_norm": 0.7846173048019409, + "learning_rate": 9.009910055227306e-05, + "loss": 0.0591, + "step": 24360 + }, + { + "epoch": 1.596671689707135, + "grad_norm": 0.580546498298645, + "learning_rate": 9.008922138792124e-05, + "loss": 0.0509, + "step": 24370 + }, + { + "epoch": 1.5973268688986437, + "grad_norm": 0.47123485803604126, + "learning_rate": 9.007933783947244e-05, + "loss": 0.0617, + "step": 24380 + }, + { + "epoch": 1.5979820480901528, + "grad_norm": 0.20486000180244446, + "learning_rate": 9.006944990800752e-05, + "loss": 0.0527, + "step": 24390 + }, + { + "epoch": 1.5986372272816616, + "grad_norm": 0.8043825030326843, + "learning_rate": 9.005955759460779e-05, + "loss": 0.0557, + "step": 24400 + }, + { + "epoch": 1.5992924064731704, + "grad_norm": 0.5418539643287659, + "learning_rate": 9.004966090035508e-05, + "loss": 0.0414, + "step": 24410 + }, + { + "epoch": 1.5999475856646792, + "grad_norm": 0.5354282855987549, + "learning_rate": 9.003975982633166e-05, + "loss": 0.0367, + "step": 24420 + }, + { + "epoch": 1.600602764856188, + "grad_norm": 0.4579067528247833, + "learning_rate": 9.00298543736203e-05, + "loss": 0.0355, + "step": 24430 + }, + { + "epoch": 1.601257944047697, + "grad_norm": 0.4951806664466858, + "learning_rate": 9.001994454330427e-05, + "loss": 0.0466, + "step": 24440 + }, + { + "epoch": 1.6019131232392059, + "grad_norm": 0.5122501254081726, + "learning_rate": 9.001003033646727e-05, + "loss": 0.0308, + "step": 24450 + }, + { + "epoch": 1.602568302430715, + "grad_norm": 0.8021392226219177, + "learning_rate": 9.00001117541935e-05, + "loss": 0.0536, + "step": 24460 + }, + { + "epoch": 1.6032234816222237, + "grad_norm": 1.0840238332748413, + "learning_rate": 8.999018879756764e-05, + "loss": 0.0862, + "step": 24470 + }, + { + "epoch": 1.6038786608137325, + "grad_norm": 0.493176132440567, + "learning_rate": 8.998026146767487e-05, + "loss": 0.0357, + "step": 24480 + }, + { + "epoch": 1.6045338400052414, + "grad_norm": 0.3461529612541199, + "learning_rate": 8.99703297656008e-05, + "loss": 0.0282, + "step": 24490 + }, + { + "epoch": 1.6051890191967502, + "grad_norm": 0.4562426805496216, + "learning_rate": 8.996039369243156e-05, + "loss": 0.0518, + "step": 24500 + }, + { + "epoch": 1.6058441983882592, + "grad_norm": 0.43527933955192566, + "learning_rate": 8.995045324925378e-05, + "loss": 0.0489, + "step": 24510 + }, + { + "epoch": 1.606499377579768, + "grad_norm": 0.45631736516952515, + "learning_rate": 8.994050843715448e-05, + "loss": 0.0426, + "step": 24520 + }, + { + "epoch": 1.607154556771277, + "grad_norm": 0.7953872084617615, + "learning_rate": 8.993055925722121e-05, + "loss": 0.0517, + "step": 24530 + }, + { + "epoch": 1.6078097359627859, + "grad_norm": 0.2943131625652313, + "learning_rate": 8.992060571054202e-05, + "loss": 0.0513, + "step": 24540 + }, + { + "epoch": 1.6084649151542947, + "grad_norm": 0.3105776309967041, + "learning_rate": 8.991064779820542e-05, + "loss": 0.056, + "step": 24550 + }, + { + "epoch": 1.6091200943458035, + "grad_norm": 0.3735332190990448, + "learning_rate": 8.990068552130036e-05, + "loss": 0.0491, + "step": 24560 + }, + { + "epoch": 1.6097752735373123, + "grad_norm": 0.6518411636352539, + "learning_rate": 8.989071888091634e-05, + "loss": 0.0506, + "step": 24570 + }, + { + "epoch": 1.6104304527288213, + "grad_norm": 0.15591199696063995, + "learning_rate": 8.988074787814329e-05, + "loss": 0.0643, + "step": 24580 + }, + { + "epoch": 1.6110856319203302, + "grad_norm": 0.2164628505706787, + "learning_rate": 8.987077251407158e-05, + "loss": 0.0375, + "step": 24590 + }, + { + "epoch": 1.6117408111118392, + "grad_norm": 0.48362404108047485, + "learning_rate": 8.986079278979216e-05, + "loss": 0.0882, + "step": 24600 + }, + { + "epoch": 1.612395990303348, + "grad_norm": 0.3951072692871094, + "learning_rate": 8.985080870639635e-05, + "loss": 0.0509, + "step": 24610 + }, + { + "epoch": 1.6130511694948568, + "grad_norm": 0.5821871757507324, + "learning_rate": 8.984082026497603e-05, + "loss": 0.0611, + "step": 24620 + }, + { + "epoch": 1.6137063486863656, + "grad_norm": 0.3670029044151306, + "learning_rate": 8.98308274666235e-05, + "loss": 0.0414, + "step": 24630 + }, + { + "epoch": 1.6143615278778745, + "grad_norm": 0.3232022225856781, + "learning_rate": 8.982083031243155e-05, + "loss": 0.0656, + "step": 24640 + }, + { + "epoch": 1.6150167070693835, + "grad_norm": 0.3814889192581177, + "learning_rate": 8.98108288034935e-05, + "loss": 0.0552, + "step": 24650 + }, + { + "epoch": 1.6156718862608923, + "grad_norm": 0.6956449151039124, + "learning_rate": 8.980082294090305e-05, + "loss": 0.0706, + "step": 24660 + }, + { + "epoch": 1.6163270654524013, + "grad_norm": 0.12740753591060638, + "learning_rate": 8.979081272575443e-05, + "loss": 0.0354, + "step": 24670 + }, + { + "epoch": 1.6169822446439102, + "grad_norm": 0.5411584973335266, + "learning_rate": 8.978079815914236e-05, + "loss": 0.0587, + "step": 24680 + }, + { + "epoch": 1.617637423835419, + "grad_norm": 0.4110010862350464, + "learning_rate": 8.977077924216202e-05, + "loss": 0.0536, + "step": 24690 + }, + { + "epoch": 1.6182926030269278, + "grad_norm": 0.5621795654296875, + "learning_rate": 8.976075597590905e-05, + "loss": 0.0475, + "step": 24700 + }, + { + "epoch": 1.6189477822184366, + "grad_norm": 0.1622728407382965, + "learning_rate": 8.975072836147958e-05, + "loss": 0.0416, + "step": 24710 + }, + { + "epoch": 1.6196029614099456, + "grad_norm": 0.7235442399978638, + "learning_rate": 8.974069639997025e-05, + "loss": 0.0574, + "step": 24720 + }, + { + "epoch": 1.6202581406014545, + "grad_norm": 0.24239817261695862, + "learning_rate": 8.973066009247808e-05, + "loss": 0.0391, + "step": 24730 + }, + { + "epoch": 1.6209133197929635, + "grad_norm": 0.22500135004520416, + "learning_rate": 8.972061944010066e-05, + "loss": 0.0598, + "step": 24740 + }, + { + "epoch": 1.6215684989844723, + "grad_norm": 0.3400023281574249, + "learning_rate": 8.971057444393603e-05, + "loss": 0.0533, + "step": 24750 + }, + { + "epoch": 1.6222236781759811, + "grad_norm": 0.3078691363334656, + "learning_rate": 8.970052510508268e-05, + "loss": 0.0502, + "step": 24760 + }, + { + "epoch": 1.62287885736749, + "grad_norm": 0.5003771781921387, + "learning_rate": 8.969047142463959e-05, + "loss": 0.0385, + "step": 24770 + }, + { + "epoch": 1.6235340365589987, + "grad_norm": 0.7932572960853577, + "learning_rate": 8.968041340370621e-05, + "loss": 0.0515, + "step": 24780 + }, + { + "epoch": 1.6241892157505078, + "grad_norm": 0.38899344205856323, + "learning_rate": 8.96703510433825e-05, + "loss": 0.0337, + "step": 24790 + }, + { + "epoch": 1.6248443949420166, + "grad_norm": 0.23475874960422516, + "learning_rate": 8.966028434476883e-05, + "loss": 0.0406, + "step": 24800 + }, + { + "epoch": 1.6254995741335256, + "grad_norm": 0.7970697283744812, + "learning_rate": 8.96502133089661e-05, + "loss": 0.0662, + "step": 24810 + }, + { + "epoch": 1.6261547533250345, + "grad_norm": 0.06237906590104103, + "learning_rate": 8.964013793707564e-05, + "loss": 0.0402, + "step": 24820 + }, + { + "epoch": 1.6268099325165433, + "grad_norm": 0.47867071628570557, + "learning_rate": 8.963005823019932e-05, + "loss": 0.0537, + "step": 24830 + }, + { + "epoch": 1.627465111708052, + "grad_norm": 0.3819151222705841, + "learning_rate": 8.961997418943939e-05, + "loss": 0.0622, + "step": 24840 + }, + { + "epoch": 1.628120290899561, + "grad_norm": 0.15424461662769318, + "learning_rate": 8.960988581589865e-05, + "loss": 0.0495, + "step": 24850 + }, + { + "epoch": 1.62877547009107, + "grad_norm": 0.4978831112384796, + "learning_rate": 8.959979311068037e-05, + "loss": 0.0435, + "step": 24860 + }, + { + "epoch": 1.6294306492825787, + "grad_norm": 0.5021260380744934, + "learning_rate": 8.958969607488823e-05, + "loss": 0.0294, + "step": 24870 + }, + { + "epoch": 1.6300858284740878, + "grad_norm": 0.4989904463291168, + "learning_rate": 8.957959470962647e-05, + "loss": 0.0389, + "step": 24880 + }, + { + "epoch": 1.6307410076655966, + "grad_norm": 1.3951610326766968, + "learning_rate": 8.956948901599971e-05, + "loss": 0.0621, + "step": 24890 + }, + { + "epoch": 1.6313961868571054, + "grad_norm": 0.6683216094970703, + "learning_rate": 8.955937899511315e-05, + "loss": 0.0496, + "step": 24900 + }, + { + "epoch": 1.6320513660486142, + "grad_norm": 0.7939022779464722, + "learning_rate": 8.954926464807238e-05, + "loss": 0.0341, + "step": 24910 + }, + { + "epoch": 1.632706545240123, + "grad_norm": 1.7722734212875366, + "learning_rate": 8.953914597598347e-05, + "loss": 0.0615, + "step": 24920 + }, + { + "epoch": 1.633361724431632, + "grad_norm": 1.1170634031295776, + "learning_rate": 8.952902297995303e-05, + "loss": 0.0685, + "step": 24930 + }, + { + "epoch": 1.634016903623141, + "grad_norm": 0.2094375193119049, + "learning_rate": 8.951889566108804e-05, + "loss": 0.0453, + "step": 24940 + }, + { + "epoch": 1.63467208281465, + "grad_norm": 0.1546308696269989, + "learning_rate": 8.950876402049606e-05, + "loss": 0.0395, + "step": 24950 + }, + { + "epoch": 1.6353272620061587, + "grad_norm": 0.8169625997543335, + "learning_rate": 8.949862805928504e-05, + "loss": 0.06, + "step": 24960 + }, + { + "epoch": 1.6359824411976676, + "grad_norm": 0.8492131233215332, + "learning_rate": 8.948848777856343e-05, + "loss": 0.0559, + "step": 24970 + }, + { + "epoch": 1.6366376203891764, + "grad_norm": 0.6076658368110657, + "learning_rate": 8.947834317944017e-05, + "loss": 0.0567, + "step": 24980 + }, + { + "epoch": 1.6372927995806852, + "grad_norm": 2.4083163738250732, + "learning_rate": 8.946819426302466e-05, + "loss": 0.0882, + "step": 24990 + }, + { + "epoch": 1.6379479787721942, + "grad_norm": 0.4364594519138336, + "learning_rate": 8.945804103042676e-05, + "loss": 0.0683, + "step": 25000 + }, + { + "epoch": 1.638603157963703, + "grad_norm": 0.6898722052574158, + "learning_rate": 8.944788348275681e-05, + "loss": 0.0454, + "step": 25010 + }, + { + "epoch": 1.639258337155212, + "grad_norm": 0.8733898997306824, + "learning_rate": 8.943772162112565e-05, + "loss": 0.0421, + "step": 25020 + }, + { + "epoch": 1.639913516346721, + "grad_norm": 0.2245141714811325, + "learning_rate": 8.942755544664454e-05, + "loss": 0.0585, + "step": 25030 + }, + { + "epoch": 1.6405686955382297, + "grad_norm": 0.7462247610092163, + "learning_rate": 8.941738496042525e-05, + "loss": 0.0298, + "step": 25040 + }, + { + "epoch": 1.6412238747297385, + "grad_norm": 0.24558782577514648, + "learning_rate": 8.940721016357999e-05, + "loss": 0.0311, + "step": 25050 + }, + { + "epoch": 1.6418790539212473, + "grad_norm": 0.4976459741592407, + "learning_rate": 8.939703105722148e-05, + "loss": 0.0331, + "step": 25060 + }, + { + "epoch": 1.6425342331127564, + "grad_norm": 0.18907159566879272, + "learning_rate": 8.93868476424629e-05, + "loss": 0.0689, + "step": 25070 + }, + { + "epoch": 1.6431894123042652, + "grad_norm": 0.6522414088249207, + "learning_rate": 8.937665992041786e-05, + "loss": 0.0718, + "step": 25080 + }, + { + "epoch": 1.6438445914957742, + "grad_norm": 0.5404929518699646, + "learning_rate": 8.93664678922005e-05, + "loss": 0.0316, + "step": 25090 + }, + { + "epoch": 1.644499770687283, + "grad_norm": 0.631647527217865, + "learning_rate": 8.93562715589254e-05, + "loss": 0.062, + "step": 25100 + }, + { + "epoch": 1.6451549498787919, + "grad_norm": 0.5935980081558228, + "learning_rate": 8.934607092170762e-05, + "loss": 0.0691, + "step": 25110 + }, + { + "epoch": 1.6458101290703007, + "grad_norm": 1.0751606225967407, + "learning_rate": 8.933586598166266e-05, + "loss": 0.0387, + "step": 25120 + }, + { + "epoch": 1.6464653082618095, + "grad_norm": 0.254270076751709, + "learning_rate": 8.932565673990655e-05, + "loss": 0.0338, + "step": 25130 + }, + { + "epoch": 1.6471204874533185, + "grad_norm": 0.15609359741210938, + "learning_rate": 8.931544319755574e-05, + "loss": 0.0591, + "step": 25140 + }, + { + "epoch": 1.6477756666448273, + "grad_norm": 0.5510426163673401, + "learning_rate": 8.930522535572718e-05, + "loss": 0.0478, + "step": 25150 + }, + { + "epoch": 1.6484308458363364, + "grad_norm": 1.3062078952789307, + "learning_rate": 8.929500321553826e-05, + "loss": 0.0533, + "step": 25160 + }, + { + "epoch": 1.6490860250278452, + "grad_norm": 0.3473220467567444, + "learning_rate": 8.928477677810686e-05, + "loss": 0.032, + "step": 25170 + }, + { + "epoch": 1.649741204219354, + "grad_norm": 1.2635141611099243, + "learning_rate": 8.927454604455137e-05, + "loss": 0.0827, + "step": 25180 + }, + { + "epoch": 1.6503963834108628, + "grad_norm": 0.29023343324661255, + "learning_rate": 8.926431101599053e-05, + "loss": 0.0459, + "step": 25190 + }, + { + "epoch": 1.6510515626023716, + "grad_norm": 0.41266390681266785, + "learning_rate": 8.925407169354369e-05, + "loss": 0.0333, + "step": 25200 + }, + { + "epoch": 1.6517067417938807, + "grad_norm": 0.8294629454612732, + "learning_rate": 8.92438280783306e-05, + "loss": 0.0512, + "step": 25210 + }, + { + "epoch": 1.6523619209853895, + "grad_norm": 0.5743732452392578, + "learning_rate": 8.923358017147146e-05, + "loss": 0.0286, + "step": 25220 + }, + { + "epoch": 1.6530171001768985, + "grad_norm": 1.180319905281067, + "learning_rate": 8.922332797408697e-05, + "loss": 0.0587, + "step": 25230 + }, + { + "epoch": 1.6536722793684073, + "grad_norm": 0.3068559467792511, + "learning_rate": 8.921307148729831e-05, + "loss": 0.0364, + "step": 25240 + }, + { + "epoch": 1.6543274585599161, + "grad_norm": 0.7745224833488464, + "learning_rate": 8.920281071222712e-05, + "loss": 0.0572, + "step": 25250 + }, + { + "epoch": 1.654982637751425, + "grad_norm": 0.26337286829948425, + "learning_rate": 8.919254564999548e-05, + "loss": 0.0624, + "step": 25260 + }, + { + "epoch": 1.6556378169429338, + "grad_norm": 0.290642112493515, + "learning_rate": 8.918227630172598e-05, + "loss": 0.0357, + "step": 25270 + }, + { + "epoch": 1.6562929961344428, + "grad_norm": 0.7917060852050781, + "learning_rate": 8.917200266854165e-05, + "loss": 0.0261, + "step": 25280 + }, + { + "epoch": 1.6569481753259516, + "grad_norm": 0.2417306900024414, + "learning_rate": 8.9161724751566e-05, + "loss": 0.0593, + "step": 25290 + }, + { + "epoch": 1.6576033545174607, + "grad_norm": 0.309257835149765, + "learning_rate": 8.915144255192302e-05, + "loss": 0.0311, + "step": 25300 + }, + { + "epoch": 1.6582585337089695, + "grad_norm": 0.4454065263271332, + "learning_rate": 8.914115607073714e-05, + "loss": 0.0263, + "step": 25310 + }, + { + "epoch": 1.6589137129004783, + "grad_norm": 0.323953241109848, + "learning_rate": 8.913086530913327e-05, + "loss": 0.0207, + "step": 25320 + }, + { + "epoch": 1.659568892091987, + "grad_norm": 0.42462247610092163, + "learning_rate": 8.912057026823681e-05, + "loss": 0.0855, + "step": 25330 + }, + { + "epoch": 1.660224071283496, + "grad_norm": 0.47367870807647705, + "learning_rate": 8.91102709491736e-05, + "loss": 0.0277, + "step": 25340 + }, + { + "epoch": 1.660879250475005, + "grad_norm": 0.1562952995300293, + "learning_rate": 8.909996735306996e-05, + "loss": 0.0188, + "step": 25350 + }, + { + "epoch": 1.6615344296665138, + "grad_norm": 0.7993791699409485, + "learning_rate": 8.908965948105268e-05, + "loss": 0.0375, + "step": 25360 + }, + { + "epoch": 1.6621896088580228, + "grad_norm": 0.668444812297821, + "learning_rate": 8.907934733424901e-05, + "loss": 0.0357, + "step": 25370 + }, + { + "epoch": 1.6628447880495316, + "grad_norm": 0.264786958694458, + "learning_rate": 8.906903091378666e-05, + "loss": 0.0519, + "step": 25380 + }, + { + "epoch": 1.6634999672410404, + "grad_norm": 0.9090656638145447, + "learning_rate": 8.905871022079384e-05, + "loss": 0.0347, + "step": 25390 + }, + { + "epoch": 1.6641551464325492, + "grad_norm": 0.33704105019569397, + "learning_rate": 8.90483852563992e-05, + "loss": 0.0394, + "step": 25400 + }, + { + "epoch": 1.664810325624058, + "grad_norm": 0.6829872131347656, + "learning_rate": 8.903805602173185e-05, + "loss": 0.0423, + "step": 25410 + }, + { + "epoch": 1.665465504815567, + "grad_norm": 0.516741931438446, + "learning_rate": 8.902772251792137e-05, + "loss": 0.0413, + "step": 25420 + }, + { + "epoch": 1.666120684007076, + "grad_norm": 0.5224716067314148, + "learning_rate": 8.901738474609786e-05, + "loss": 0.0215, + "step": 25430 + }, + { + "epoch": 1.666775863198585, + "grad_norm": 0.3628625273704529, + "learning_rate": 8.900704270739179e-05, + "loss": 0.0674, + "step": 25440 + }, + { + "epoch": 1.6674310423900938, + "grad_norm": 0.5106024742126465, + "learning_rate": 8.89966964029342e-05, + "loss": 0.0494, + "step": 25450 + }, + { + "epoch": 1.6680862215816026, + "grad_norm": 1.4935643672943115, + "learning_rate": 8.898634583385652e-05, + "loss": 0.0323, + "step": 25460 + }, + { + "epoch": 1.6687414007731114, + "grad_norm": 0.6005637049674988, + "learning_rate": 8.897599100129065e-05, + "loss": 0.0368, + "step": 25470 + }, + { + "epoch": 1.6693965799646202, + "grad_norm": 0.6445067524909973, + "learning_rate": 8.896563190636903e-05, + "loss": 0.0475, + "step": 25480 + }, + { + "epoch": 1.6700517591561292, + "grad_norm": 0.7969626784324646, + "learning_rate": 8.895526855022448e-05, + "loss": 0.0572, + "step": 25490 + }, + { + "epoch": 1.670706938347638, + "grad_norm": 0.5151100158691406, + "learning_rate": 8.894490093399033e-05, + "loss": 0.039, + "step": 25500 + }, + { + "epoch": 1.671362117539147, + "grad_norm": 0.19778773188591003, + "learning_rate": 8.893452905880035e-05, + "loss": 0.038, + "step": 25510 + }, + { + "epoch": 1.672017296730656, + "grad_norm": 0.2269926369190216, + "learning_rate": 8.892415292578883e-05, + "loss": 0.0371, + "step": 25520 + }, + { + "epoch": 1.6726724759221647, + "grad_norm": 0.29562416672706604, + "learning_rate": 8.891377253609046e-05, + "loss": 0.0436, + "step": 25530 + }, + { + "epoch": 1.6733276551136735, + "grad_norm": 0.33947819471359253, + "learning_rate": 8.890338789084043e-05, + "loss": 0.0377, + "step": 25540 + }, + { + "epoch": 1.6739828343051824, + "grad_norm": 0.17300260066986084, + "learning_rate": 8.88929989911744e-05, + "loss": 0.0477, + "step": 25550 + }, + { + "epoch": 1.6746380134966914, + "grad_norm": 0.8916866779327393, + "learning_rate": 8.888260583822847e-05, + "loss": 0.0353, + "step": 25560 + }, + { + "epoch": 1.6752931926882002, + "grad_norm": 0.6344285011291504, + "learning_rate": 8.887220843313921e-05, + "loss": 0.0359, + "step": 25570 + }, + { + "epoch": 1.6759483718797092, + "grad_norm": 0.4270569384098053, + "learning_rate": 8.88618067770437e-05, + "loss": 0.039, + "step": 25580 + }, + { + "epoch": 1.676603551071218, + "grad_norm": 0.7264342904090881, + "learning_rate": 8.885140087107942e-05, + "loss": 0.0809, + "step": 25590 + }, + { + "epoch": 1.6772587302627269, + "grad_norm": 0.5743564963340759, + "learning_rate": 8.884099071638436e-05, + "loss": 0.0472, + "step": 25600 + }, + { + "epoch": 1.6779139094542357, + "grad_norm": 0.4188147783279419, + "learning_rate": 8.883057631409695e-05, + "loss": 0.0281, + "step": 25610 + }, + { + "epoch": 1.6785690886457445, + "grad_norm": 0.2421078234910965, + "learning_rate": 8.882015766535608e-05, + "loss": 0.0422, + "step": 25620 + }, + { + "epoch": 1.6792242678372535, + "grad_norm": 0.15515387058258057, + "learning_rate": 8.880973477130115e-05, + "loss": 0.0449, + "step": 25630 + }, + { + "epoch": 1.6798794470287624, + "grad_norm": 0.33254122734069824, + "learning_rate": 8.879930763307197e-05, + "loss": 0.0482, + "step": 25640 + }, + { + "epoch": 1.6805346262202714, + "grad_norm": 0.7220032215118408, + "learning_rate": 8.878887625180884e-05, + "loss": 0.0861, + "step": 25650 + }, + { + "epoch": 1.6811898054117802, + "grad_norm": 0.393711119890213, + "learning_rate": 8.877844062865253e-05, + "loss": 0.063, + "step": 25660 + }, + { + "epoch": 1.681844984603289, + "grad_norm": 0.26030856370925903, + "learning_rate": 8.876800076474424e-05, + "loss": 0.0256, + "step": 25670 + }, + { + "epoch": 1.6825001637947978, + "grad_norm": 0.485823392868042, + "learning_rate": 8.875755666122568e-05, + "loss": 0.0304, + "step": 25680 + }, + { + "epoch": 1.6831553429863066, + "grad_norm": 0.2881697118282318, + "learning_rate": 8.8747108319239e-05, + "loss": 0.0516, + "step": 25690 + }, + { + "epoch": 1.6838105221778157, + "grad_norm": 0.6706723570823669, + "learning_rate": 8.87366557399268e-05, + "loss": 0.0456, + "step": 25700 + }, + { + "epoch": 1.6844657013693245, + "grad_norm": 0.578077495098114, + "learning_rate": 8.872619892443217e-05, + "loss": 0.0372, + "step": 25710 + }, + { + "epoch": 1.6851208805608335, + "grad_norm": 0.5849544405937195, + "learning_rate": 8.871573787389865e-05, + "loss": 0.0315, + "step": 25720 + }, + { + "epoch": 1.6857760597523423, + "grad_norm": 0.39620721340179443, + "learning_rate": 8.870527258947024e-05, + "loss": 0.0463, + "step": 25730 + }, + { + "epoch": 1.6864312389438512, + "grad_norm": 0.3477281630039215, + "learning_rate": 8.869480307229143e-05, + "loss": 0.053, + "step": 25740 + }, + { + "epoch": 1.68708641813536, + "grad_norm": 0.9157842397689819, + "learning_rate": 8.868432932350712e-05, + "loss": 0.0371, + "step": 25750 + }, + { + "epoch": 1.6877415973268688, + "grad_norm": 0.5667171478271484, + "learning_rate": 8.867385134426272e-05, + "loss": 0.0236, + "step": 25760 + }, + { + "epoch": 1.6883967765183778, + "grad_norm": 0.7084688544273376, + "learning_rate": 8.866336913570407e-05, + "loss": 0.0588, + "step": 25770 + }, + { + "epoch": 1.6890519557098866, + "grad_norm": 0.31837213039398193, + "learning_rate": 8.865288269897751e-05, + "loss": 0.055, + "step": 25780 + }, + { + "epoch": 1.6897071349013957, + "grad_norm": 0.5735109448432922, + "learning_rate": 8.864239203522981e-05, + "loss": 0.0355, + "step": 25790 + }, + { + "epoch": 1.6903623140929045, + "grad_norm": 0.9971774816513062, + "learning_rate": 8.863189714560822e-05, + "loss": 0.0437, + "step": 25800 + }, + { + "epoch": 1.6910174932844133, + "grad_norm": 0.4097428619861603, + "learning_rate": 8.862139803126043e-05, + "loss": 0.0563, + "step": 25810 + }, + { + "epoch": 1.6916726724759221, + "grad_norm": 0.5752952098846436, + "learning_rate": 8.861089469333463e-05, + "loss": 0.0416, + "step": 25820 + }, + { + "epoch": 1.692327851667431, + "grad_norm": 0.1547429859638214, + "learning_rate": 8.860038713297944e-05, + "loss": 0.0517, + "step": 25830 + }, + { + "epoch": 1.69298303085894, + "grad_norm": 0.6191262602806091, + "learning_rate": 8.858987535134394e-05, + "loss": 0.0378, + "step": 25840 + }, + { + "epoch": 1.6936382100504488, + "grad_norm": 0.9757015705108643, + "learning_rate": 8.857935934957769e-05, + "loss": 0.0444, + "step": 25850 + }, + { + "epoch": 1.6942933892419578, + "grad_norm": 0.679523229598999, + "learning_rate": 8.856883912883071e-05, + "loss": 0.0332, + "step": 25860 + }, + { + "epoch": 1.6949485684334666, + "grad_norm": 0.11969691514968872, + "learning_rate": 8.855831469025346e-05, + "loss": 0.0368, + "step": 25870 + }, + { + "epoch": 1.6956037476249755, + "grad_norm": 0.6656765937805176, + "learning_rate": 8.854778603499689e-05, + "loss": 0.0582, + "step": 25880 + }, + { + "epoch": 1.6962589268164843, + "grad_norm": 0.14334005117416382, + "learning_rate": 8.85372531642124e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 1.696914106007993, + "grad_norm": 0.6305078268051147, + "learning_rate": 8.852671607905185e-05, + "loss": 0.0728, + "step": 25900 + }, + { + "epoch": 1.697569285199502, + "grad_norm": 0.5015376210212708, + "learning_rate": 8.851617478066754e-05, + "loss": 0.0465, + "step": 25910 + }, + { + "epoch": 1.698224464391011, + "grad_norm": 0.8222076296806335, + "learning_rate": 8.850562927021227e-05, + "loss": 0.0476, + "step": 25920 + }, + { + "epoch": 1.69887964358252, + "grad_norm": 0.7808675765991211, + "learning_rate": 8.849507954883928e-05, + "loss": 0.0584, + "step": 25930 + }, + { + "epoch": 1.6995348227740288, + "grad_norm": 0.3464774191379547, + "learning_rate": 8.848452561770226e-05, + "loss": 0.0354, + "step": 25940 + }, + { + "epoch": 1.7001900019655376, + "grad_norm": 0.5009920597076416, + "learning_rate": 8.847396747795538e-05, + "loss": 0.0655, + "step": 25950 + }, + { + "epoch": 1.7008451811570464, + "grad_norm": 0.3163636028766632, + "learning_rate": 8.846340513075327e-05, + "loss": 0.0539, + "step": 25960 + }, + { + "epoch": 1.7015003603485552, + "grad_norm": 0.6402565240859985, + "learning_rate": 8.845283857725099e-05, + "loss": 0.0388, + "step": 25970 + }, + { + "epoch": 1.702155539540064, + "grad_norm": 0.419752836227417, + "learning_rate": 8.844226781860409e-05, + "loss": 0.0409, + "step": 25980 + }, + { + "epoch": 1.702810718731573, + "grad_norm": 0.5477521419525146, + "learning_rate": 8.84316928559686e-05, + "loss": 0.0406, + "step": 25990 + }, + { + "epoch": 1.7034658979230821, + "grad_norm": 0.30531203746795654, + "learning_rate": 8.842111369050094e-05, + "loss": 0.0297, + "step": 26000 + }, + { + "epoch": 1.704121077114591, + "grad_norm": 0.22526751458644867, + "learning_rate": 8.841053032335808e-05, + "loss": 0.031, + "step": 26010 + }, + { + "epoch": 1.7047762563060997, + "grad_norm": 0.3303757905960083, + "learning_rate": 8.839994275569735e-05, + "loss": 0.032, + "step": 26020 + }, + { + "epoch": 1.7054314354976086, + "grad_norm": 0.28028029203414917, + "learning_rate": 8.838935098867662e-05, + "loss": 0.0508, + "step": 26030 + }, + { + "epoch": 1.7060866146891174, + "grad_norm": 0.3908112645149231, + "learning_rate": 8.837875502345418e-05, + "loss": 0.0195, + "step": 26040 + }, + { + "epoch": 1.7067417938806262, + "grad_norm": 0.5472773313522339, + "learning_rate": 8.83681548611888e-05, + "loss": 0.0766, + "step": 26050 + }, + { + "epoch": 1.7073969730721352, + "grad_norm": 0.8101149201393127, + "learning_rate": 8.835755050303969e-05, + "loss": 0.0419, + "step": 26060 + }, + { + "epoch": 1.7080521522636443, + "grad_norm": 0.6070889830589294, + "learning_rate": 8.834694195016653e-05, + "loss": 0.0548, + "step": 26070 + }, + { + "epoch": 1.708707331455153, + "grad_norm": 0.4509230852127075, + "learning_rate": 8.833632920372942e-05, + "loss": 0.0386, + "step": 26080 + }, + { + "epoch": 1.709362510646662, + "grad_norm": 0.5424062013626099, + "learning_rate": 8.832571226488903e-05, + "loss": 0.0435, + "step": 26090 + }, + { + "epoch": 1.7100176898381707, + "grad_norm": 0.7935674786567688, + "learning_rate": 8.831509113480634e-05, + "loss": 0.0589, + "step": 26100 + }, + { + "epoch": 1.7106728690296795, + "grad_norm": 0.8625946640968323, + "learning_rate": 8.83044658146429e-05, + "loss": 0.0518, + "step": 26110 + }, + { + "epoch": 1.7113280482211883, + "grad_norm": 0.7044157981872559, + "learning_rate": 8.829383630556067e-05, + "loss": 0.0362, + "step": 26120 + }, + { + "epoch": 1.7119832274126974, + "grad_norm": 0.14564527571201324, + "learning_rate": 8.828320260872207e-05, + "loss": 0.0873, + "step": 26130 + }, + { + "epoch": 1.7126384066042064, + "grad_norm": 0.8741104602813721, + "learning_rate": 8.827256472529e-05, + "loss": 0.0517, + "step": 26140 + }, + { + "epoch": 1.7132935857957152, + "grad_norm": 0.6987593173980713, + "learning_rate": 8.826192265642778e-05, + "loss": 0.0569, + "step": 26150 + }, + { + "epoch": 1.713948764987224, + "grad_norm": 0.1493770033121109, + "learning_rate": 8.825127640329923e-05, + "loss": 0.046, + "step": 26160 + }, + { + "epoch": 1.7146039441787329, + "grad_norm": 0.1333404928445816, + "learning_rate": 8.824062596706861e-05, + "loss": 0.0272, + "step": 26170 + }, + { + "epoch": 1.7152591233702417, + "grad_norm": 0.797353208065033, + "learning_rate": 8.822997134890062e-05, + "loss": 0.0592, + "step": 26180 + }, + { + "epoch": 1.7159143025617505, + "grad_norm": 0.3012152314186096, + "learning_rate": 8.821931254996044e-05, + "loss": 0.0385, + "step": 26190 + }, + { + "epoch": 1.7165694817532595, + "grad_norm": 0.7213609218597412, + "learning_rate": 8.82086495714137e-05, + "loss": 0.0418, + "step": 26200 + }, + { + "epoch": 1.7172246609447686, + "grad_norm": 0.36906784772872925, + "learning_rate": 8.81979824144265e-05, + "loss": 0.0573, + "step": 26210 + }, + { + "epoch": 1.7178798401362774, + "grad_norm": 0.3641807436943054, + "learning_rate": 8.818731108016536e-05, + "loss": 0.0549, + "step": 26220 + }, + { + "epoch": 1.7185350193277862, + "grad_norm": 0.6261157393455505, + "learning_rate": 8.81766355697973e-05, + "loss": 0.0641, + "step": 26230 + }, + { + "epoch": 1.719190198519295, + "grad_norm": 0.9930711388587952, + "learning_rate": 8.816595588448977e-05, + "loss": 0.0556, + "step": 26240 + }, + { + "epoch": 1.7198453777108038, + "grad_norm": 1.0809413194656372, + "learning_rate": 8.81552720254107e-05, + "loss": 0.0325, + "step": 26250 + }, + { + "epoch": 1.7205005569023126, + "grad_norm": 1.1800453662872314, + "learning_rate": 8.814458399372842e-05, + "loss": 0.0381, + "step": 26260 + }, + { + "epoch": 1.7211557360938217, + "grad_norm": 0.5067087411880493, + "learning_rate": 8.813389179061181e-05, + "loss": 0.0304, + "step": 26270 + }, + { + "epoch": 1.7218109152853307, + "grad_norm": 0.5212064385414124, + "learning_rate": 8.812319541723012e-05, + "loss": 0.0986, + "step": 26280 + }, + { + "epoch": 1.7224660944768395, + "grad_norm": 0.6746916770935059, + "learning_rate": 8.811249487475309e-05, + "loss": 0.0386, + "step": 26290 + }, + { + "epoch": 1.7231212736683483, + "grad_norm": 0.18075530230998993, + "learning_rate": 8.810179016435092e-05, + "loss": 0.0163, + "step": 26300 + }, + { + "epoch": 1.7237764528598571, + "grad_norm": 0.11763633042573929, + "learning_rate": 8.809108128719428e-05, + "loss": 0.033, + "step": 26310 + }, + { + "epoch": 1.724431632051366, + "grad_norm": 0.613986611366272, + "learning_rate": 8.808036824445424e-05, + "loss": 0.0486, + "step": 26320 + }, + { + "epoch": 1.7250868112428748, + "grad_norm": 0.22224591672420502, + "learning_rate": 8.806965103730238e-05, + "loss": 0.0448, + "step": 26330 + }, + { + "epoch": 1.7257419904343838, + "grad_norm": 0.4769037067890167, + "learning_rate": 8.805892966691074e-05, + "loss": 0.0685, + "step": 26340 + }, + { + "epoch": 1.7263971696258928, + "grad_norm": 1.024493932723999, + "learning_rate": 8.804820413445175e-05, + "loss": 0.0888, + "step": 26350 + }, + { + "epoch": 1.7270523488174017, + "grad_norm": 0.6182979941368103, + "learning_rate": 8.803747444109837e-05, + "loss": 0.0673, + "step": 26360 + }, + { + "epoch": 1.7277075280089105, + "grad_norm": 0.5353109240531921, + "learning_rate": 8.802674058802399e-05, + "loss": 0.0419, + "step": 26370 + }, + { + "epoch": 1.7283627072004193, + "grad_norm": 0.3621982932090759, + "learning_rate": 8.801600257640241e-05, + "loss": 0.049, + "step": 26380 + }, + { + "epoch": 1.729017886391928, + "grad_norm": 0.48130837082862854, + "learning_rate": 8.800526040740795e-05, + "loss": 0.0778, + "step": 26390 + }, + { + "epoch": 1.729673065583437, + "grad_norm": 0.2178499549627304, + "learning_rate": 8.799451408221535e-05, + "loss": 0.0257, + "step": 26400 + }, + { + "epoch": 1.730328244774946, + "grad_norm": 0.5193350911140442, + "learning_rate": 8.798376360199982e-05, + "loss": 0.0549, + "step": 26410 + }, + { + "epoch": 1.7309834239664548, + "grad_norm": 0.26243409514427185, + "learning_rate": 8.797300896793701e-05, + "loss": 0.0469, + "step": 26420 + }, + { + "epoch": 1.7316386031579638, + "grad_norm": 0.5536999106407166, + "learning_rate": 8.796225018120302e-05, + "loss": 0.038, + "step": 26430 + }, + { + "epoch": 1.7322937823494726, + "grad_norm": 0.5087193846702576, + "learning_rate": 8.795148724297444e-05, + "loss": 0.0443, + "step": 26440 + }, + { + "epoch": 1.7329489615409814, + "grad_norm": 0.184067040681839, + "learning_rate": 8.794072015442825e-05, + "loss": 0.0638, + "step": 26450 + }, + { + "epoch": 1.7336041407324903, + "grad_norm": 0.6275085210800171, + "learning_rate": 8.792994891674198e-05, + "loss": 0.0373, + "step": 26460 + }, + { + "epoch": 1.734259319923999, + "grad_norm": 0.44933441281318665, + "learning_rate": 8.79191735310935e-05, + "loss": 0.0293, + "step": 26470 + }, + { + "epoch": 1.734914499115508, + "grad_norm": 0.1963961273431778, + "learning_rate": 8.790839399866122e-05, + "loss": 0.0347, + "step": 26480 + }, + { + "epoch": 1.735569678307017, + "grad_norm": 0.8828932046890259, + "learning_rate": 8.789761032062397e-05, + "loss": 0.05, + "step": 26490 + }, + { + "epoch": 1.736224857498526, + "grad_norm": 0.30557337403297424, + "learning_rate": 8.788682249816103e-05, + "loss": 0.0353, + "step": 26500 + }, + { + "epoch": 1.7368800366900348, + "grad_norm": 0.5075646042823792, + "learning_rate": 8.787603053245215e-05, + "loss": 0.034, + "step": 26510 + }, + { + "epoch": 1.7375352158815436, + "grad_norm": 0.5467332005500793, + "learning_rate": 8.78652344246775e-05, + "loss": 0.0496, + "step": 26520 + }, + { + "epoch": 1.7381903950730524, + "grad_norm": 0.23429057002067566, + "learning_rate": 8.785443417601776e-05, + "loss": 0.0296, + "step": 26530 + }, + { + "epoch": 1.7388455742645612, + "grad_norm": 0.2322532832622528, + "learning_rate": 8.784362978765401e-05, + "loss": 0.044, + "step": 26540 + }, + { + "epoch": 1.7395007534560702, + "grad_norm": 0.31686869263648987, + "learning_rate": 8.783282126076779e-05, + "loss": 0.0285, + "step": 26550 + }, + { + "epoch": 1.740155932647579, + "grad_norm": 1.1817046403884888, + "learning_rate": 8.782200859654112e-05, + "loss": 0.0545, + "step": 26560 + }, + { + "epoch": 1.740811111839088, + "grad_norm": 0.19679544866085052, + "learning_rate": 8.781119179615646e-05, + "loss": 0.0232, + "step": 26570 + }, + { + "epoch": 1.741466291030597, + "grad_norm": 0.40239056944847107, + "learning_rate": 8.780037086079674e-05, + "loss": 0.0525, + "step": 26580 + }, + { + "epoch": 1.7421214702221057, + "grad_norm": 0.2584366202354431, + "learning_rate": 8.778954579164527e-05, + "loss": 0.1053, + "step": 26590 + }, + { + "epoch": 1.7427766494136145, + "grad_norm": 0.17184984683990479, + "learning_rate": 8.777871658988588e-05, + "loss": 0.0588, + "step": 26600 + }, + { + "epoch": 1.7434318286051234, + "grad_norm": 0.8906131982803345, + "learning_rate": 8.776788325670285e-05, + "loss": 0.0527, + "step": 26610 + }, + { + "epoch": 1.7440870077966324, + "grad_norm": 0.49503931403160095, + "learning_rate": 8.775704579328089e-05, + "loss": 0.0363, + "step": 26620 + }, + { + "epoch": 1.7447421869881412, + "grad_norm": 1.3211597204208374, + "learning_rate": 8.774620420080517e-05, + "loss": 0.0537, + "step": 26630 + }, + { + "epoch": 1.7453973661796502, + "grad_norm": 0.5622996687889099, + "learning_rate": 8.773535848046131e-05, + "loss": 0.0347, + "step": 26640 + }, + { + "epoch": 1.746052545371159, + "grad_norm": 0.3625056743621826, + "learning_rate": 8.772450863343538e-05, + "loss": 0.0458, + "step": 26650 + }, + { + "epoch": 1.7467077245626679, + "grad_norm": 0.6265057921409607, + "learning_rate": 8.77136546609139e-05, + "loss": 0.0351, + "step": 26660 + }, + { + "epoch": 1.7473629037541767, + "grad_norm": 0.1200837567448616, + "learning_rate": 8.770279656408385e-05, + "loss": 0.0316, + "step": 26670 + }, + { + "epoch": 1.7480180829456855, + "grad_norm": 0.2380172312259674, + "learning_rate": 8.769193434413265e-05, + "loss": 0.0602, + "step": 26680 + }, + { + "epoch": 1.7486732621371945, + "grad_norm": 0.3528609871864319, + "learning_rate": 8.76810680022482e-05, + "loss": 0.0634, + "step": 26690 + }, + { + "epoch": 1.7493284413287034, + "grad_norm": 0.5491473078727722, + "learning_rate": 8.767019753961878e-05, + "loss": 0.0628, + "step": 26700 + }, + { + "epoch": 1.7499836205202124, + "grad_norm": 1.1655220985412598, + "learning_rate": 8.765932295743321e-05, + "loss": 0.0662, + "step": 26710 + }, + { + "epoch": 1.7506387997117212, + "grad_norm": 0.3142126500606537, + "learning_rate": 8.764844425688068e-05, + "loss": 0.0542, + "step": 26720 + }, + { + "epoch": 1.75129397890323, + "grad_norm": 0.1646478772163391, + "learning_rate": 8.763756143915092e-05, + "loss": 0.0388, + "step": 26730 + }, + { + "epoch": 1.7519491580947388, + "grad_norm": 0.5043388605117798, + "learning_rate": 8.7626674505434e-05, + "loss": 0.0423, + "step": 26740 + }, + { + "epoch": 1.7526043372862476, + "grad_norm": 0.7295386791229248, + "learning_rate": 8.761578345692053e-05, + "loss": 0.0459, + "step": 26750 + }, + { + "epoch": 1.7532595164777567, + "grad_norm": 0.7014727592468262, + "learning_rate": 8.760488829480156e-05, + "loss": 0.0389, + "step": 26760 + }, + { + "epoch": 1.7539146956692655, + "grad_norm": 0.3080736994743347, + "learning_rate": 8.759398902026854e-05, + "loss": 0.041, + "step": 26770 + }, + { + "epoch": 1.7545698748607745, + "grad_norm": 0.19568537175655365, + "learning_rate": 8.758308563451339e-05, + "loss": 0.0534, + "step": 26780 + }, + { + "epoch": 1.7552250540522834, + "grad_norm": 0.18593378365039825, + "learning_rate": 8.75721781387285e-05, + "loss": 0.0629, + "step": 26790 + }, + { + "epoch": 1.7558802332437922, + "grad_norm": 0.5018649101257324, + "learning_rate": 8.75612665341067e-05, + "loss": 0.0525, + "step": 26800 + }, + { + "epoch": 1.756535412435301, + "grad_norm": 0.27048948407173157, + "learning_rate": 8.755035082184126e-05, + "loss": 0.0362, + "step": 26810 + }, + { + "epoch": 1.7571905916268098, + "grad_norm": 0.2582830488681793, + "learning_rate": 8.753943100312592e-05, + "loss": 0.0421, + "step": 26820 + }, + { + "epoch": 1.7578457708183188, + "grad_norm": 0.8917121887207031, + "learning_rate": 8.752850707915484e-05, + "loss": 0.0593, + "step": 26830 + }, + { + "epoch": 1.7585009500098276, + "grad_norm": 0.6190996170043945, + "learning_rate": 8.751757905112264e-05, + "loss": 0.0566, + "step": 26840 + }, + { + "epoch": 1.7591561292013367, + "grad_norm": 0.7590128183364868, + "learning_rate": 8.75066469202244e-05, + "loss": 0.0701, + "step": 26850 + }, + { + "epoch": 1.7598113083928455, + "grad_norm": 0.41397932171821594, + "learning_rate": 8.749571068765567e-05, + "loss": 0.0416, + "step": 26860 + }, + { + "epoch": 1.7604664875843543, + "grad_norm": 0.159251868724823, + "learning_rate": 8.748477035461238e-05, + "loss": 0.0695, + "step": 26870 + }, + { + "epoch": 1.7611216667758631, + "grad_norm": 0.5884260535240173, + "learning_rate": 8.747382592229095e-05, + "loss": 0.023, + "step": 26880 + }, + { + "epoch": 1.761776845967372, + "grad_norm": 0.5546339750289917, + "learning_rate": 8.746287739188828e-05, + "loss": 0.0489, + "step": 26890 + }, + { + "epoch": 1.762432025158881, + "grad_norm": 0.28155872225761414, + "learning_rate": 8.745192476460165e-05, + "loss": 0.0319, + "step": 26900 + }, + { + "epoch": 1.7630872043503898, + "grad_norm": 0.3581695556640625, + "learning_rate": 8.744096804162882e-05, + "loss": 0.044, + "step": 26910 + }, + { + "epoch": 1.7637423835418988, + "grad_norm": 0.2651240825653076, + "learning_rate": 8.743000722416804e-05, + "loss": 0.0278, + "step": 26920 + }, + { + "epoch": 1.7643975627334076, + "grad_norm": 0.8046073317527771, + "learning_rate": 8.741904231341793e-05, + "loss": 0.048, + "step": 26930 + }, + { + "epoch": 1.7650527419249165, + "grad_norm": 0.5522497892379761, + "learning_rate": 8.740807331057762e-05, + "loss": 0.0899, + "step": 26940 + }, + { + "epoch": 1.7657079211164253, + "grad_norm": 0.8537208437919617, + "learning_rate": 8.739710021684667e-05, + "loss": 0.0654, + "step": 26950 + }, + { + "epoch": 1.766363100307934, + "grad_norm": 0.23441606760025024, + "learning_rate": 8.738612303342503e-05, + "loss": 0.0504, + "step": 26960 + }, + { + "epoch": 1.7670182794994431, + "grad_norm": 0.1950870305299759, + "learning_rate": 8.73751417615132e-05, + "loss": 0.0288, + "step": 26970 + }, + { + "epoch": 1.767673458690952, + "grad_norm": 0.5760199427604675, + "learning_rate": 8.736415640231208e-05, + "loss": 0.0451, + "step": 26980 + }, + { + "epoch": 1.768328637882461, + "grad_norm": 0.9181889891624451, + "learning_rate": 8.735316695702297e-05, + "loss": 0.0689, + "step": 26990 + }, + { + "epoch": 1.7689838170739698, + "grad_norm": 0.19233965873718262, + "learning_rate": 8.734217342684769e-05, + "loss": 0.0274, + "step": 27000 + }, + { + "epoch": 1.7696389962654786, + "grad_norm": 0.41300690174102783, + "learning_rate": 8.733117581298847e-05, + "loss": 0.0499, + "step": 27010 + }, + { + "epoch": 1.7702941754569874, + "grad_norm": 0.14546352624893188, + "learning_rate": 8.732017411664796e-05, + "loss": 0.0388, + "step": 27020 + }, + { + "epoch": 1.7709493546484962, + "grad_norm": 0.2682909369468689, + "learning_rate": 8.730916833902936e-05, + "loss": 0.0325, + "step": 27030 + }, + { + "epoch": 1.7716045338400053, + "grad_norm": 0.6213412880897522, + "learning_rate": 8.729815848133618e-05, + "loss": 0.0421, + "step": 27040 + }, + { + "epoch": 1.772259713031514, + "grad_norm": 0.3720129430294037, + "learning_rate": 8.728714454477247e-05, + "loss": 0.0803, + "step": 27050 + }, + { + "epoch": 1.7729148922230231, + "grad_norm": 0.4819211959838867, + "learning_rate": 8.727612653054269e-05, + "loss": 0.0295, + "step": 27060 + }, + { + "epoch": 1.773570071414532, + "grad_norm": 0.7559753656387329, + "learning_rate": 8.726510443985176e-05, + "loss": 0.0339, + "step": 27070 + }, + { + "epoch": 1.7742252506060407, + "grad_norm": 0.6389660835266113, + "learning_rate": 8.725407827390503e-05, + "loss": 0.0565, + "step": 27080 + }, + { + "epoch": 1.7748804297975496, + "grad_norm": 0.34973251819610596, + "learning_rate": 8.724304803390833e-05, + "loss": 0.0369, + "step": 27090 + }, + { + "epoch": 1.7755356089890584, + "grad_norm": 0.14392729103565216, + "learning_rate": 8.723201372106788e-05, + "loss": 0.0356, + "step": 27100 + }, + { + "epoch": 1.7761907881805674, + "grad_norm": 0.15802399814128876, + "learning_rate": 8.722097533659038e-05, + "loss": 0.0245, + "step": 27110 + }, + { + "epoch": 1.7768459673720762, + "grad_norm": 0.49159812927246094, + "learning_rate": 8.720993288168299e-05, + "loss": 0.068, + "step": 27120 + }, + { + "epoch": 1.7775011465635853, + "grad_norm": 0.37556660175323486, + "learning_rate": 8.719888635755327e-05, + "loss": 0.0542, + "step": 27130 + }, + { + "epoch": 1.778156325755094, + "grad_norm": 0.8940541744232178, + "learning_rate": 8.718783576540928e-05, + "loss": 0.0562, + "step": 27140 + }, + { + "epoch": 1.778811504946603, + "grad_norm": 0.6084932684898376, + "learning_rate": 8.717678110645948e-05, + "loss": 0.052, + "step": 27150 + }, + { + "epoch": 1.7794666841381117, + "grad_norm": 0.263345867395401, + "learning_rate": 8.716572238191279e-05, + "loss": 0.036, + "step": 27160 + }, + { + "epoch": 1.7801218633296205, + "grad_norm": 0.38861846923828125, + "learning_rate": 8.715465959297857e-05, + "loss": 0.038, + "step": 27170 + }, + { + "epoch": 1.7807770425211296, + "grad_norm": 0.7838310599327087, + "learning_rate": 8.714359274086665e-05, + "loss": 0.0219, + "step": 27180 + }, + { + "epoch": 1.7814322217126384, + "grad_norm": 0.4637278914451599, + "learning_rate": 8.713252182678726e-05, + "loss": 0.0897, + "step": 27190 + }, + { + "epoch": 1.7820874009041474, + "grad_norm": 0.9645418524742126, + "learning_rate": 8.712144685195112e-05, + "loss": 0.0334, + "step": 27200 + }, + { + "epoch": 1.7827425800956562, + "grad_norm": 0.7344522476196289, + "learning_rate": 8.711036781756936e-05, + "loss": 0.0378, + "step": 27210 + }, + { + "epoch": 1.783397759287165, + "grad_norm": 0.6940813064575195, + "learning_rate": 8.709928472485357e-05, + "loss": 0.0527, + "step": 27220 + }, + { + "epoch": 1.7840529384786739, + "grad_norm": 0.22390116751194, + "learning_rate": 8.708819757501579e-05, + "loss": 0.0685, + "step": 27230 + }, + { + "epoch": 1.7847081176701827, + "grad_norm": 0.5016825199127197, + "learning_rate": 8.707710636926846e-05, + "loss": 0.0408, + "step": 27240 + }, + { + "epoch": 1.7853632968616917, + "grad_norm": 0.6874819397926331, + "learning_rate": 8.706601110882455e-05, + "loss": 0.0458, + "step": 27250 + }, + { + "epoch": 1.7860184760532005, + "grad_norm": 0.9251253008842468, + "learning_rate": 8.705491179489738e-05, + "loss": 0.0351, + "step": 27260 + }, + { + "epoch": 1.7866736552447096, + "grad_norm": 0.7352482080459595, + "learning_rate": 8.704380842870077e-05, + "loss": 0.0525, + "step": 27270 + }, + { + "epoch": 1.7873288344362184, + "grad_norm": 0.9313580393791199, + "learning_rate": 8.703270101144895e-05, + "loss": 0.0814, + "step": 27280 + }, + { + "epoch": 1.7879840136277272, + "grad_norm": 0.4386019706726074, + "learning_rate": 8.702158954435664e-05, + "loss": 0.0485, + "step": 27290 + }, + { + "epoch": 1.788639192819236, + "grad_norm": 0.3902246356010437, + "learning_rate": 8.701047402863896e-05, + "loss": 0.033, + "step": 27300 + }, + { + "epoch": 1.7892943720107448, + "grad_norm": 1.2750740051269531, + "learning_rate": 8.699935446551148e-05, + "loss": 0.0559, + "step": 27310 + }, + { + "epoch": 1.7899495512022539, + "grad_norm": 0.9336858987808228, + "learning_rate": 8.698823085619022e-05, + "loss": 0.0494, + "step": 27320 + }, + { + "epoch": 1.7906047303937627, + "grad_norm": 0.22155076265335083, + "learning_rate": 8.697710320189166e-05, + "loss": 0.0588, + "step": 27330 + }, + { + "epoch": 1.7912599095852717, + "grad_norm": 0.8721990585327148, + "learning_rate": 8.696597150383268e-05, + "loss": 0.0609, + "step": 27340 + }, + { + "epoch": 1.7919150887767805, + "grad_norm": 0.20095834136009216, + "learning_rate": 8.695483576323063e-05, + "loss": 0.0679, + "step": 27350 + }, + { + "epoch": 1.7925702679682893, + "grad_norm": 0.4471551477909088, + "learning_rate": 8.69436959813033e-05, + "loss": 0.0567, + "step": 27360 + }, + { + "epoch": 1.7932254471597981, + "grad_norm": 0.3261246085166931, + "learning_rate": 8.693255215926892e-05, + "loss": 0.0264, + "step": 27370 + }, + { + "epoch": 1.793880626351307, + "grad_norm": 0.5016831159591675, + "learning_rate": 8.692140429834617e-05, + "loss": 0.051, + "step": 27380 + }, + { + "epoch": 1.794535805542816, + "grad_norm": 0.8228007555007935, + "learning_rate": 8.691025239975415e-05, + "loss": 0.0336, + "step": 27390 + }, + { + "epoch": 1.7951909847343248, + "grad_norm": 0.19979189336299896, + "learning_rate": 8.689909646471243e-05, + "loss": 0.0558, + "step": 27400 + }, + { + "epoch": 1.7958461639258338, + "grad_norm": 0.25679585337638855, + "learning_rate": 8.688793649444099e-05, + "loss": 0.0405, + "step": 27410 + }, + { + "epoch": 1.7965013431173427, + "grad_norm": 1.3365586996078491, + "learning_rate": 8.687677249016029e-05, + "loss": 0.0481, + "step": 27420 + }, + { + "epoch": 1.7971565223088515, + "grad_norm": 0.310467004776001, + "learning_rate": 8.686560445309118e-05, + "loss": 0.0224, + "step": 27430 + }, + { + "epoch": 1.7978117015003603, + "grad_norm": 1.0986998081207275, + "learning_rate": 8.685443238445499e-05, + "loss": 0.0389, + "step": 27440 + }, + { + "epoch": 1.798466880691869, + "grad_norm": 0.6099084615707397, + "learning_rate": 8.68432562854735e-05, + "loss": 0.0531, + "step": 27450 + }, + { + "epoch": 1.7991220598833781, + "grad_norm": 0.4392401874065399, + "learning_rate": 8.683207615736887e-05, + "loss": 0.0615, + "step": 27460 + }, + { + "epoch": 1.799777239074887, + "grad_norm": 0.2471519261598587, + "learning_rate": 8.682089200136379e-05, + "loss": 0.0497, + "step": 27470 + }, + { + "epoch": 1.800432418266396, + "grad_norm": 0.48663130402565, + "learning_rate": 8.680970381868132e-05, + "loss": 0.0488, + "step": 27480 + }, + { + "epoch": 1.8010875974579048, + "grad_norm": 0.7093746662139893, + "learning_rate": 8.679851161054498e-05, + "loss": 0.0483, + "step": 27490 + }, + { + "epoch": 1.8017427766494136, + "grad_norm": 0.9247955679893494, + "learning_rate": 8.678731537817873e-05, + "loss": 0.0448, + "step": 27500 + }, + { + "epoch": 1.8023979558409224, + "grad_norm": 0.2200731486082077, + "learning_rate": 8.677611512280697e-05, + "loss": 0.0668, + "step": 27510 + }, + { + "epoch": 1.8030531350324313, + "grad_norm": 0.2710297703742981, + "learning_rate": 8.676491084565457e-05, + "loss": 0.042, + "step": 27520 + }, + { + "epoch": 1.8037083142239403, + "grad_norm": 0.11089368909597397, + "learning_rate": 8.675370254794678e-05, + "loss": 0.0321, + "step": 27530 + }, + { + "epoch": 1.804363493415449, + "grad_norm": 0.585163950920105, + "learning_rate": 8.674249023090935e-05, + "loss": 0.0386, + "step": 27540 + }, + { + "epoch": 1.8050186726069581, + "grad_norm": 0.07026571780443192, + "learning_rate": 8.673127389576843e-05, + "loss": 0.0513, + "step": 27550 + }, + { + "epoch": 1.805673851798467, + "grad_norm": 0.38210758566856384, + "learning_rate": 8.67200535437506e-05, + "loss": 0.0593, + "step": 27560 + }, + { + "epoch": 1.8063290309899758, + "grad_norm": 0.1579025834798813, + "learning_rate": 8.670882917608296e-05, + "loss": 0.0701, + "step": 27570 + }, + { + "epoch": 1.8069842101814846, + "grad_norm": 0.15045909583568573, + "learning_rate": 8.669760079399292e-05, + "loss": 0.0438, + "step": 27580 + }, + { + "epoch": 1.8076393893729934, + "grad_norm": 1.7688740491867065, + "learning_rate": 8.668636839870845e-05, + "loss": 0.0614, + "step": 27590 + }, + { + "epoch": 1.8082945685645024, + "grad_norm": 0.37010976672172546, + "learning_rate": 8.667513199145789e-05, + "loss": 0.0399, + "step": 27600 + }, + { + "epoch": 1.8089497477560113, + "grad_norm": 0.5306574702262878, + "learning_rate": 8.666389157347002e-05, + "loss": 0.0499, + "step": 27610 + }, + { + "epoch": 1.8096049269475203, + "grad_norm": 0.46109047532081604, + "learning_rate": 8.66526471459741e-05, + "loss": 0.0218, + "step": 27620 + }, + { + "epoch": 1.810260106139029, + "grad_norm": 0.7034527063369751, + "learning_rate": 8.66413987101998e-05, + "loss": 0.0424, + "step": 27630 + }, + { + "epoch": 1.810915285330538, + "grad_norm": 0.6412696242332458, + "learning_rate": 8.663014626737723e-05, + "loss": 0.0303, + "step": 27640 + }, + { + "epoch": 1.8115704645220467, + "grad_norm": 0.6212946176528931, + "learning_rate": 8.661888981873691e-05, + "loss": 0.0467, + "step": 27650 + }, + { + "epoch": 1.8122256437135555, + "grad_norm": 0.7446384429931641, + "learning_rate": 8.660762936550988e-05, + "loss": 0.0539, + "step": 27660 + }, + { + "epoch": 1.8128808229050646, + "grad_norm": 0.5088834762573242, + "learning_rate": 8.659636490892753e-05, + "loss": 0.0579, + "step": 27670 + }, + { + "epoch": 1.8135360020965734, + "grad_norm": 1.0471744537353516, + "learning_rate": 8.658509645022174e-05, + "loss": 0.061, + "step": 27680 + }, + { + "epoch": 1.8141911812880824, + "grad_norm": 0.39744317531585693, + "learning_rate": 8.657382399062481e-05, + "loss": 0.0569, + "step": 27690 + }, + { + "epoch": 1.8148463604795912, + "grad_norm": 0.6036180853843689, + "learning_rate": 8.656254753136946e-05, + "loss": 0.0396, + "step": 27700 + }, + { + "epoch": 1.8155015396711, + "grad_norm": 0.4552290141582489, + "learning_rate": 8.655126707368891e-05, + "loss": 0.0462, + "step": 27710 + }, + { + "epoch": 1.8161567188626089, + "grad_norm": 0.4597778022289276, + "learning_rate": 8.653998261881672e-05, + "loss": 0.0484, + "step": 27720 + }, + { + "epoch": 1.8168118980541177, + "grad_norm": 0.9999524354934692, + "learning_rate": 8.652869416798699e-05, + "loss": 0.0426, + "step": 27730 + }, + { + "epoch": 1.8174670772456267, + "grad_norm": 0.9828169941902161, + "learning_rate": 8.651740172243417e-05, + "loss": 0.042, + "step": 27740 + }, + { + "epoch": 1.8181222564371355, + "grad_norm": 0.44580069184303284, + "learning_rate": 8.65061052833932e-05, + "loss": 0.0262, + "step": 27750 + }, + { + "epoch": 1.8187774356286446, + "grad_norm": 0.788749635219574, + "learning_rate": 8.649480485209945e-05, + "loss": 0.0292, + "step": 27760 + }, + { + "epoch": 1.8194326148201534, + "grad_norm": 1.2181345224380493, + "learning_rate": 8.64835004297887e-05, + "loss": 0.0368, + "step": 27770 + }, + { + "epoch": 1.8200877940116622, + "grad_norm": 0.20330177247524261, + "learning_rate": 8.64721920176972e-05, + "loss": 0.0338, + "step": 27780 + }, + { + "epoch": 1.820742973203171, + "grad_norm": 0.8500066995620728, + "learning_rate": 8.646087961706164e-05, + "loss": 0.0403, + "step": 27790 + }, + { + "epoch": 1.8213981523946798, + "grad_norm": 0.49651122093200684, + "learning_rate": 8.644956322911908e-05, + "loss": 0.0339, + "step": 27800 + }, + { + "epoch": 1.8220533315861889, + "grad_norm": 0.47623661160469055, + "learning_rate": 8.643824285510709e-05, + "loss": 0.0547, + "step": 27810 + }, + { + "epoch": 1.8227085107776977, + "grad_norm": 0.5028743743896484, + "learning_rate": 8.642691849626364e-05, + "loss": 0.0421, + "step": 27820 + }, + { + "epoch": 1.8233636899692067, + "grad_norm": 0.08532645553350449, + "learning_rate": 8.641559015382717e-05, + "loss": 0.0493, + "step": 27830 + }, + { + "epoch": 1.8240188691607155, + "grad_norm": 0.6213110089302063, + "learning_rate": 8.640425782903649e-05, + "loss": 0.0413, + "step": 27840 + }, + { + "epoch": 1.8246740483522244, + "grad_norm": 0.6581875085830688, + "learning_rate": 8.639292152313091e-05, + "loss": 0.0559, + "step": 27850 + }, + { + "epoch": 1.8253292275437332, + "grad_norm": 0.3633226454257965, + "learning_rate": 8.638158123735015e-05, + "loss": 0.0248, + "step": 27860 + }, + { + "epoch": 1.825984406735242, + "grad_norm": 0.6352837085723877, + "learning_rate": 8.637023697293436e-05, + "loss": 0.0628, + "step": 27870 + }, + { + "epoch": 1.826639585926751, + "grad_norm": 0.2620892822742462, + "learning_rate": 8.635888873112414e-05, + "loss": 0.0359, + "step": 27880 + }, + { + "epoch": 1.8272947651182598, + "grad_norm": 0.43012306094169617, + "learning_rate": 8.634753651316052e-05, + "loss": 0.0289, + "step": 27890 + }, + { + "epoch": 1.8279499443097689, + "grad_norm": 0.23703867197036743, + "learning_rate": 8.633618032028496e-05, + "loss": 0.0526, + "step": 27900 + }, + { + "epoch": 1.8286051235012777, + "grad_norm": 0.2731086313724518, + "learning_rate": 8.632482015373934e-05, + "loss": 0.0655, + "step": 27910 + }, + { + "epoch": 1.8292603026927865, + "grad_norm": 0.4993983805179596, + "learning_rate": 8.6313456014766e-05, + "loss": 0.0902, + "step": 27920 + }, + { + "epoch": 1.8299154818842953, + "grad_norm": 0.4159434139728546, + "learning_rate": 8.630208790460771e-05, + "loss": 0.0415, + "step": 27930 + }, + { + "epoch": 1.8305706610758041, + "grad_norm": 0.2662709057331085, + "learning_rate": 8.629071582450768e-05, + "loss": 0.0381, + "step": 27940 + }, + { + "epoch": 1.8312258402673132, + "grad_norm": 0.2861038148403168, + "learning_rate": 8.62793397757095e-05, + "loss": 0.0405, + "step": 27950 + }, + { + "epoch": 1.831881019458822, + "grad_norm": 0.6431301236152649, + "learning_rate": 8.626795975945729e-05, + "loss": 0.0304, + "step": 27960 + }, + { + "epoch": 1.832536198650331, + "grad_norm": 0.4854452311992645, + "learning_rate": 8.625657577699551e-05, + "loss": 0.047, + "step": 27970 + }, + { + "epoch": 1.8331913778418398, + "grad_norm": 0.15965604782104492, + "learning_rate": 8.624518782956914e-05, + "loss": 0.0568, + "step": 27980 + }, + { + "epoch": 1.8338465570333486, + "grad_norm": 1.4321599006652832, + "learning_rate": 8.62337959184235e-05, + "loss": 0.0488, + "step": 27990 + }, + { + "epoch": 1.8345017362248575, + "grad_norm": 0.5611470937728882, + "learning_rate": 8.622240004480441e-05, + "loss": 0.0347, + "step": 28000 + }, + { + "epoch": 1.8351569154163663, + "grad_norm": 0.2882581353187561, + "learning_rate": 8.621100020995814e-05, + "loss": 0.0584, + "step": 28010 + }, + { + "epoch": 1.835812094607875, + "grad_norm": 0.6361557245254517, + "learning_rate": 8.619959641513132e-05, + "loss": 0.045, + "step": 28020 + }, + { + "epoch": 1.8364672737993841, + "grad_norm": 0.34489449858665466, + "learning_rate": 8.618818866157105e-05, + "loss": 0.0737, + "step": 28030 + }, + { + "epoch": 1.8371224529908932, + "grad_norm": 0.3059673011302948, + "learning_rate": 8.617677695052487e-05, + "loss": 0.0439, + "step": 28040 + }, + { + "epoch": 1.837777632182402, + "grad_norm": 0.7071667909622192, + "learning_rate": 8.616536128324078e-05, + "loss": 0.0525, + "step": 28050 + }, + { + "epoch": 1.8384328113739108, + "grad_norm": 0.7342749238014221, + "learning_rate": 8.615394166096712e-05, + "loss": 0.0563, + "step": 28060 + }, + { + "epoch": 1.8390879905654196, + "grad_norm": 0.5594747066497803, + "learning_rate": 8.614251808495279e-05, + "loss": 0.055, + "step": 28070 + }, + { + "epoch": 1.8397431697569284, + "grad_norm": 0.6261788010597229, + "learning_rate": 8.6131090556447e-05, + "loss": 0.0498, + "step": 28080 + }, + { + "epoch": 1.8403983489484372, + "grad_norm": 0.28439536690711975, + "learning_rate": 8.611965907669947e-05, + "loss": 0.0547, + "step": 28090 + }, + { + "epoch": 1.8410535281399463, + "grad_norm": 0.5645738244056702, + "learning_rate": 8.610822364696034e-05, + "loss": 0.0765, + "step": 28100 + }, + { + "epoch": 1.8417087073314553, + "grad_norm": 0.3332867920398712, + "learning_rate": 8.609678426848015e-05, + "loss": 0.0249, + "step": 28110 + }, + { + "epoch": 1.8423638865229641, + "grad_norm": 0.9189432263374329, + "learning_rate": 8.60853409425099e-05, + "loss": 0.0839, + "step": 28120 + }, + { + "epoch": 1.843019065714473, + "grad_norm": 0.967698335647583, + "learning_rate": 8.607389367030104e-05, + "loss": 0.0485, + "step": 28130 + }, + { + "epoch": 1.8436742449059818, + "grad_norm": 0.4838651716709137, + "learning_rate": 8.606244245310538e-05, + "loss": 0.0769, + "step": 28140 + }, + { + "epoch": 1.8443294240974906, + "grad_norm": 0.6417199969291687, + "learning_rate": 8.605098729217525e-05, + "loss": 0.0353, + "step": 28150 + }, + { + "epoch": 1.8449846032889994, + "grad_norm": 0.48066794872283936, + "learning_rate": 8.603952818876335e-05, + "loss": 0.0562, + "step": 28160 + }, + { + "epoch": 1.8456397824805084, + "grad_norm": 0.45528683066368103, + "learning_rate": 8.602806514412281e-05, + "loss": 0.0347, + "step": 28170 + }, + { + "epoch": 1.8462949616720175, + "grad_norm": 0.571972668170929, + "learning_rate": 8.601659815950726e-05, + "loss": 0.0414, + "step": 28180 + }, + { + "epoch": 1.8469501408635263, + "grad_norm": 0.18565733730793, + "learning_rate": 8.600512723617067e-05, + "loss": 0.0454, + "step": 28190 + }, + { + "epoch": 1.847605320055035, + "grad_norm": 1.1011298894882202, + "learning_rate": 8.59936523753675e-05, + "loss": 0.07, + "step": 28200 + }, + { + "epoch": 1.848260499246544, + "grad_norm": 0.4512563645839691, + "learning_rate": 8.598217357835264e-05, + "loss": 0.04, + "step": 28210 + }, + { + "epoch": 1.8489156784380527, + "grad_norm": 0.5030927062034607, + "learning_rate": 8.597069084638135e-05, + "loss": 0.0377, + "step": 28220 + }, + { + "epoch": 1.8495708576295615, + "grad_norm": 0.8436662554740906, + "learning_rate": 8.595920418070939e-05, + "loss": 0.0603, + "step": 28230 + }, + { + "epoch": 1.8502260368210706, + "grad_norm": 0.7222127318382263, + "learning_rate": 8.594771358259295e-05, + "loss": 0.0485, + "step": 28240 + }, + { + "epoch": 1.8508812160125796, + "grad_norm": 1.0013840198516846, + "learning_rate": 8.593621905328858e-05, + "loss": 0.0676, + "step": 28250 + }, + { + "epoch": 1.8515363952040884, + "grad_norm": 1.6965889930725098, + "learning_rate": 8.592472059405333e-05, + "loss": 0.0694, + "step": 28260 + }, + { + "epoch": 1.8521915743955972, + "grad_norm": 0.48904961347579956, + "learning_rate": 8.591321820614464e-05, + "loss": 0.042, + "step": 28270 + }, + { + "epoch": 1.852846753587106, + "grad_norm": 0.9992334246635437, + "learning_rate": 8.590171189082041e-05, + "loss": 0.0439, + "step": 28280 + }, + { + "epoch": 1.8535019327786149, + "grad_norm": 0.5383329391479492, + "learning_rate": 8.589020164933894e-05, + "loss": 0.0658, + "step": 28290 + }, + { + "epoch": 1.8541571119701237, + "grad_norm": 0.26421475410461426, + "learning_rate": 8.587868748295898e-05, + "loss": 0.0484, + "step": 28300 + }, + { + "epoch": 1.8548122911616327, + "grad_norm": 0.791275143623352, + "learning_rate": 8.586716939293971e-05, + "loss": 0.0544, + "step": 28310 + }, + { + "epoch": 1.8554674703531417, + "grad_norm": 0.5890633463859558, + "learning_rate": 8.58556473805407e-05, + "loss": 0.0706, + "step": 28320 + }, + { + "epoch": 1.8561226495446506, + "grad_norm": 0.6191546320915222, + "learning_rate": 8.584412144702202e-05, + "loss": 0.0396, + "step": 28330 + }, + { + "epoch": 1.8567778287361594, + "grad_norm": 0.527069091796875, + "learning_rate": 8.58325915936441e-05, + "loss": 0.0526, + "step": 28340 + }, + { + "epoch": 1.8574330079276682, + "grad_norm": 0.6749821305274963, + "learning_rate": 8.582105782166783e-05, + "loss": 0.0619, + "step": 28350 + }, + { + "epoch": 1.858088187119177, + "grad_norm": 0.5541948676109314, + "learning_rate": 8.580952013235455e-05, + "loss": 0.0482, + "step": 28360 + }, + { + "epoch": 1.8587433663106858, + "grad_norm": 0.11137406527996063, + "learning_rate": 8.579797852696596e-05, + "loss": 0.0244, + "step": 28370 + }, + { + "epoch": 1.8593985455021949, + "grad_norm": 0.10119855403900146, + "learning_rate": 8.578643300676428e-05, + "loss": 0.0209, + "step": 28380 + }, + { + "epoch": 1.860053724693704, + "grad_norm": 0.5354958176612854, + "learning_rate": 8.577488357301209e-05, + "loss": 0.0514, + "step": 28390 + }, + { + "epoch": 1.8607089038852127, + "grad_norm": 0.6983276605606079, + "learning_rate": 8.576333022697242e-05, + "loss": 0.0374, + "step": 28400 + }, + { + "epoch": 1.8613640830767215, + "grad_norm": 0.3364304006099701, + "learning_rate": 8.575177296990873e-05, + "loss": 0.048, + "step": 28410 + }, + { + "epoch": 1.8620192622682303, + "grad_norm": 0.8057951927185059, + "learning_rate": 8.574021180308489e-05, + "loss": 0.0895, + "step": 28420 + }, + { + "epoch": 1.8626744414597391, + "grad_norm": 0.7102183699607849, + "learning_rate": 8.572864672776523e-05, + "loss": 0.06, + "step": 28430 + }, + { + "epoch": 1.863329620651248, + "grad_norm": 0.7211599946022034, + "learning_rate": 8.571707774521447e-05, + "loss": 0.0368, + "step": 28440 + }, + { + "epoch": 1.863984799842757, + "grad_norm": 0.8372213244438171, + "learning_rate": 8.57055048566978e-05, + "loss": 0.0468, + "step": 28450 + }, + { + "epoch": 1.864639979034266, + "grad_norm": 0.48298028111457825, + "learning_rate": 8.569392806348078e-05, + "loss": 0.0313, + "step": 28460 + }, + { + "epoch": 1.8652951582257749, + "grad_norm": 0.3228246867656708, + "learning_rate": 8.568234736682947e-05, + "loss": 0.0451, + "step": 28470 + }, + { + "epoch": 1.8659503374172837, + "grad_norm": 0.27182596921920776, + "learning_rate": 8.567076276801029e-05, + "loss": 0.0215, + "step": 28480 + }, + { + "epoch": 1.8666055166087925, + "grad_norm": 0.5544021725654602, + "learning_rate": 8.565917426829013e-05, + "loss": 0.0435, + "step": 28490 + }, + { + "epoch": 1.8672606958003013, + "grad_norm": 0.5391684174537659, + "learning_rate": 8.564758186893628e-05, + "loss": 0.0434, + "step": 28500 + }, + { + "epoch": 1.86791587499181, + "grad_norm": 0.25956279039382935, + "learning_rate": 8.563598557121649e-05, + "loss": 0.0572, + "step": 28510 + }, + { + "epoch": 1.8685710541833191, + "grad_norm": 0.4976642429828644, + "learning_rate": 8.562438537639888e-05, + "loss": 0.0348, + "step": 28520 + }, + { + "epoch": 1.869226233374828, + "grad_norm": 0.20134486258029938, + "learning_rate": 8.561278128575206e-05, + "loss": 0.0672, + "step": 28530 + }, + { + "epoch": 1.869881412566337, + "grad_norm": 0.7408950328826904, + "learning_rate": 8.5601173300545e-05, + "loss": 0.0425, + "step": 28540 + }, + { + "epoch": 1.8705365917578458, + "grad_norm": 0.4266950190067291, + "learning_rate": 8.558956142204717e-05, + "loss": 0.09, + "step": 28550 + }, + { + "epoch": 1.8711917709493546, + "grad_norm": 0.4584607481956482, + "learning_rate": 8.55779456515284e-05, + "loss": 0.0382, + "step": 28560 + }, + { + "epoch": 1.8718469501408634, + "grad_norm": 0.3731119632720947, + "learning_rate": 8.556632599025898e-05, + "loss": 0.0527, + "step": 28570 + }, + { + "epoch": 1.8725021293323723, + "grad_norm": 0.5226166844367981, + "learning_rate": 8.555470243950964e-05, + "loss": 0.0558, + "step": 28580 + }, + { + "epoch": 1.8731573085238813, + "grad_norm": 0.6737975478172302, + "learning_rate": 8.554307500055148e-05, + "loss": 0.06, + "step": 28590 + }, + { + "epoch": 1.87381248771539, + "grad_norm": 0.4630531668663025, + "learning_rate": 8.553144367465609e-05, + "loss": 0.0335, + "step": 28600 + }, + { + "epoch": 1.8744676669068991, + "grad_norm": 0.23013246059417725, + "learning_rate": 8.551980846309544e-05, + "loss": 0.0464, + "step": 28610 + }, + { + "epoch": 1.875122846098408, + "grad_norm": 0.9662805199623108, + "learning_rate": 8.550816936714193e-05, + "loss": 0.0341, + "step": 28620 + }, + { + "epoch": 1.8757780252899168, + "grad_norm": 0.45222875475883484, + "learning_rate": 8.549652638806841e-05, + "loss": 0.0507, + "step": 28630 + }, + { + "epoch": 1.8764332044814256, + "grad_norm": 1.0925273895263672, + "learning_rate": 8.548487952714812e-05, + "loss": 0.0642, + "step": 28640 + }, + { + "epoch": 1.8770883836729344, + "grad_norm": 0.24297840893268585, + "learning_rate": 8.547322878565478e-05, + "loss": 0.0292, + "step": 28650 + }, + { + "epoch": 1.8777435628644434, + "grad_norm": 0.7069005966186523, + "learning_rate": 8.546157416486245e-05, + "loss": 0.046, + "step": 28660 + }, + { + "epoch": 1.8783987420559523, + "grad_norm": 0.39306673407554626, + "learning_rate": 8.54499156660457e-05, + "loss": 0.032, + "step": 28670 + }, + { + "epoch": 1.8790539212474613, + "grad_norm": 0.7311403751373291, + "learning_rate": 8.543825329047947e-05, + "loss": 0.0491, + "step": 28680 + }, + { + "epoch": 1.87970910043897, + "grad_norm": 0.7560368776321411, + "learning_rate": 8.542658703943913e-05, + "loss": 0.0595, + "step": 28690 + }, + { + "epoch": 1.880364279630479, + "grad_norm": 0.8756270408630371, + "learning_rate": 8.541491691420051e-05, + "loss": 0.0614, + "step": 28700 + }, + { + "epoch": 1.8810194588219877, + "grad_norm": 0.7380228042602539, + "learning_rate": 8.54032429160398e-05, + "loss": 0.0533, + "step": 28710 + }, + { + "epoch": 1.8816746380134965, + "grad_norm": 0.7271636724472046, + "learning_rate": 8.539156504623369e-05, + "loss": 0.0482, + "step": 28720 + }, + { + "epoch": 1.8823298172050056, + "grad_norm": 0.34476497769355774, + "learning_rate": 8.537988330605923e-05, + "loss": 0.0788, + "step": 28730 + }, + { + "epoch": 1.8829849963965144, + "grad_norm": 0.8847461938858032, + "learning_rate": 8.536819769679393e-05, + "loss": 0.0327, + "step": 28740 + }, + { + "epoch": 1.8836401755880234, + "grad_norm": 0.6283743381500244, + "learning_rate": 8.53565082197157e-05, + "loss": 0.0544, + "step": 28750 + }, + { + "epoch": 1.8842953547795322, + "grad_norm": 0.7788726687431335, + "learning_rate": 8.534481487610289e-05, + "loss": 0.0291, + "step": 28760 + }, + { + "epoch": 1.884950533971041, + "grad_norm": 0.6627737879753113, + "learning_rate": 8.533311766723428e-05, + "loss": 0.08, + "step": 28770 + }, + { + "epoch": 1.8856057131625499, + "grad_norm": 0.5261823534965515, + "learning_rate": 8.532141659438901e-05, + "loss": 0.0196, + "step": 28780 + }, + { + "epoch": 1.8862608923540587, + "grad_norm": 0.5295193791389465, + "learning_rate": 8.530971165884675e-05, + "loss": 0.0453, + "step": 28790 + }, + { + "epoch": 1.8869160715455677, + "grad_norm": 0.32038235664367676, + "learning_rate": 8.529800286188752e-05, + "loss": 0.0321, + "step": 28800 + }, + { + "epoch": 1.8875712507370765, + "grad_norm": 1.4864104986190796, + "learning_rate": 8.528629020479175e-05, + "loss": 0.0296, + "step": 28810 + }, + { + "epoch": 1.8882264299285856, + "grad_norm": 0.731037974357605, + "learning_rate": 8.527457368884033e-05, + "loss": 0.0332, + "step": 28820 + }, + { + "epoch": 1.8888816091200944, + "grad_norm": 0.37977075576782227, + "learning_rate": 8.526285331531458e-05, + "loss": 0.05, + "step": 28830 + }, + { + "epoch": 1.8895367883116032, + "grad_norm": 0.7624167203903198, + "learning_rate": 8.525112908549621e-05, + "loss": 0.0653, + "step": 28840 + }, + { + "epoch": 1.890191967503112, + "grad_norm": 0.7590762972831726, + "learning_rate": 8.523940100066735e-05, + "loss": 0.0437, + "step": 28850 + }, + { + "epoch": 1.8908471466946208, + "grad_norm": 0.5342332720756531, + "learning_rate": 8.52276690621106e-05, + "loss": 0.0464, + "step": 28860 + }, + { + "epoch": 1.8915023258861299, + "grad_norm": 0.34784841537475586, + "learning_rate": 8.521593327110889e-05, + "loss": 0.0665, + "step": 28870 + }, + { + "epoch": 1.8921575050776387, + "grad_norm": 0.33254534006118774, + "learning_rate": 8.520419362894569e-05, + "loss": 0.0526, + "step": 28880 + }, + { + "epoch": 1.8928126842691477, + "grad_norm": 0.1952946037054062, + "learning_rate": 8.51924501369048e-05, + "loss": 0.0329, + "step": 28890 + }, + { + "epoch": 1.8934678634606565, + "grad_norm": 0.4983533024787903, + "learning_rate": 8.518070279627047e-05, + "loss": 0.0288, + "step": 28900 + }, + { + "epoch": 1.8941230426521654, + "grad_norm": 0.19125628471374512, + "learning_rate": 8.516895160832737e-05, + "loss": 0.0554, + "step": 28910 + }, + { + "epoch": 1.8947782218436742, + "grad_norm": 0.4093112349510193, + "learning_rate": 8.515719657436061e-05, + "loss": 0.0339, + "step": 28920 + }, + { + "epoch": 1.895433401035183, + "grad_norm": 0.5537734627723694, + "learning_rate": 8.514543769565568e-05, + "loss": 0.04, + "step": 28930 + }, + { + "epoch": 1.896088580226692, + "grad_norm": 0.23680639266967773, + "learning_rate": 8.513367497349853e-05, + "loss": 0.0528, + "step": 28940 + }, + { + "epoch": 1.8967437594182008, + "grad_norm": 0.7784058451652527, + "learning_rate": 8.51219084091755e-05, + "loss": 0.0442, + "step": 28950 + }, + { + "epoch": 1.8973989386097099, + "grad_norm": 1.892896294593811, + "learning_rate": 8.511013800397338e-05, + "loss": 0.0652, + "step": 28960 + }, + { + "epoch": 1.8980541178012187, + "grad_norm": 0.6448855400085449, + "learning_rate": 8.509836375917937e-05, + "loss": 0.0616, + "step": 28970 + }, + { + "epoch": 1.8987092969927275, + "grad_norm": 0.8471712470054626, + "learning_rate": 8.508658567608104e-05, + "loss": 0.0737, + "step": 28980 + }, + { + "epoch": 1.8993644761842363, + "grad_norm": 0.977130115032196, + "learning_rate": 8.507480375596647e-05, + "loss": 0.0543, + "step": 28990 + }, + { + "epoch": 1.9000196553757451, + "grad_norm": 0.45970290899276733, + "learning_rate": 8.506301800012408e-05, + "loss": 0.0292, + "step": 29000 + }, + { + "epoch": 1.9006748345672542, + "grad_norm": 0.37910544872283936, + "learning_rate": 8.505122840984278e-05, + "loss": 0.0518, + "step": 29010 + }, + { + "epoch": 1.901330013758763, + "grad_norm": 0.3874974846839905, + "learning_rate": 8.503943498641182e-05, + "loss": 0.0443, + "step": 29020 + }, + { + "epoch": 1.901985192950272, + "grad_norm": 0.2624249756336212, + "learning_rate": 8.502763773112095e-05, + "loss": 0.0362, + "step": 29030 + }, + { + "epoch": 1.9026403721417808, + "grad_norm": 0.2884019911289215, + "learning_rate": 8.501583664526026e-05, + "loss": 0.0261, + "step": 29040 + }, + { + "epoch": 1.9032955513332896, + "grad_norm": 2.033770799636841, + "learning_rate": 8.500403173012032e-05, + "loss": 0.0694, + "step": 29050 + }, + { + "epoch": 1.9039507305247985, + "grad_norm": 0.6528716087341309, + "learning_rate": 8.499222298699211e-05, + "loss": 0.0337, + "step": 29060 + }, + { + "epoch": 1.9046059097163073, + "grad_norm": 0.1007140576839447, + "learning_rate": 8.498041041716701e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 1.9052610889078163, + "grad_norm": 0.1452118307352066, + "learning_rate": 8.496859402193681e-05, + "loss": 0.0483, + "step": 29080 + }, + { + "epoch": 1.9059162680993251, + "grad_norm": 0.5806809067726135, + "learning_rate": 8.495677380259374e-05, + "loss": 0.0357, + "step": 29090 + }, + { + "epoch": 1.9065714472908342, + "grad_norm": 0.4671275019645691, + "learning_rate": 8.494494976043045e-05, + "loss": 0.0215, + "step": 29100 + }, + { + "epoch": 1.907226626482343, + "grad_norm": 0.9083296656608582, + "learning_rate": 8.493312189673998e-05, + "loss": 0.0414, + "step": 29110 + }, + { + "epoch": 1.9078818056738518, + "grad_norm": 0.5067633390426636, + "learning_rate": 8.492129021281584e-05, + "loss": 0.0431, + "step": 29120 + }, + { + "epoch": 1.9085369848653606, + "grad_norm": 0.6226611733436584, + "learning_rate": 8.490945470995188e-05, + "loss": 0.0615, + "step": 29130 + }, + { + "epoch": 1.9091921640568694, + "grad_norm": 0.1874585747718811, + "learning_rate": 8.489761538944247e-05, + "loss": 0.0436, + "step": 29140 + }, + { + "epoch": 1.9098473432483785, + "grad_norm": 2.2674787044525146, + "learning_rate": 8.48857722525823e-05, + "loss": 0.0588, + "step": 29150 + }, + { + "epoch": 1.9105025224398873, + "grad_norm": 0.743145227432251, + "learning_rate": 8.487392530066652e-05, + "loss": 0.0502, + "step": 29160 + }, + { + "epoch": 1.9111577016313963, + "grad_norm": 0.33446723222732544, + "learning_rate": 8.486207453499069e-05, + "loss": 0.0396, + "step": 29170 + }, + { + "epoch": 1.9118128808229051, + "grad_norm": 0.36757805943489075, + "learning_rate": 8.485021995685082e-05, + "loss": 0.0256, + "step": 29180 + }, + { + "epoch": 1.912468060014414, + "grad_norm": 0.2253963053226471, + "learning_rate": 8.483836156754328e-05, + "loss": 0.0575, + "step": 29190 + }, + { + "epoch": 1.9131232392059228, + "grad_norm": 0.7223816514015198, + "learning_rate": 8.482649936836491e-05, + "loss": 0.0438, + "step": 29200 + }, + { + "epoch": 1.9137784183974316, + "grad_norm": 0.3312126696109772, + "learning_rate": 8.481463336061293e-05, + "loss": 0.0552, + "step": 29210 + }, + { + "epoch": 1.9144335975889406, + "grad_norm": 1.3530330657958984, + "learning_rate": 8.480276354558496e-05, + "loss": 0.0591, + "step": 29220 + }, + { + "epoch": 1.9150887767804494, + "grad_norm": 1.239295244216919, + "learning_rate": 8.479088992457913e-05, + "loss": 0.0464, + "step": 29230 + }, + { + "epoch": 1.9157439559719585, + "grad_norm": 0.3711037337779999, + "learning_rate": 8.477901249889387e-05, + "loss": 0.0436, + "step": 29240 + }, + { + "epoch": 1.9163991351634673, + "grad_norm": 1.153511881828308, + "learning_rate": 8.47671312698281e-05, + "loss": 0.0398, + "step": 29250 + }, + { + "epoch": 1.917054314354976, + "grad_norm": 0.3017680048942566, + "learning_rate": 8.475524623868112e-05, + "loss": 0.0278, + "step": 29260 + }, + { + "epoch": 1.917709493546485, + "grad_norm": 0.7987732291221619, + "learning_rate": 8.474335740675266e-05, + "loss": 0.0336, + "step": 29270 + }, + { + "epoch": 1.9183646727379937, + "grad_norm": 0.49109554290771484, + "learning_rate": 8.473146477534289e-05, + "loss": 0.0334, + "step": 29280 + }, + { + "epoch": 1.9190198519295028, + "grad_norm": 0.10834934562444687, + "learning_rate": 8.471956834575232e-05, + "loss": 0.058, + "step": 29290 + }, + { + "epoch": 1.9196750311210116, + "grad_norm": 0.6503281593322754, + "learning_rate": 8.470766811928197e-05, + "loss": 0.0654, + "step": 29300 + }, + { + "epoch": 1.9203302103125206, + "grad_norm": 0.7920076251029968, + "learning_rate": 8.469576409723323e-05, + "loss": 0.0378, + "step": 29310 + }, + { + "epoch": 1.9209853895040294, + "grad_norm": 0.47404563426971436, + "learning_rate": 8.468385628090788e-05, + "loss": 0.0418, + "step": 29320 + }, + { + "epoch": 1.9216405686955382, + "grad_norm": 0.5294764041900635, + "learning_rate": 8.467194467160815e-05, + "loss": 0.0703, + "step": 29330 + }, + { + "epoch": 1.922295747887047, + "grad_norm": 1.011920690536499, + "learning_rate": 8.466002927063667e-05, + "loss": 0.0438, + "step": 29340 + }, + { + "epoch": 1.9229509270785559, + "grad_norm": 0.32887423038482666, + "learning_rate": 8.464811007929651e-05, + "loss": 0.0273, + "step": 29350 + }, + { + "epoch": 1.923606106270065, + "grad_norm": 0.26105624437332153, + "learning_rate": 8.463618709889114e-05, + "loss": 0.0446, + "step": 29360 + }, + { + "epoch": 1.9242612854615737, + "grad_norm": 0.3886866867542267, + "learning_rate": 8.462426033072442e-05, + "loss": 0.0408, + "step": 29370 + }, + { + "epoch": 1.9249164646530827, + "grad_norm": 0.16238190233707428, + "learning_rate": 8.461232977610061e-05, + "loss": 0.0467, + "step": 29380 + }, + { + "epoch": 1.9255716438445916, + "grad_norm": 1.128002643585205, + "learning_rate": 8.46003954363245e-05, + "loss": 0.0507, + "step": 29390 + }, + { + "epoch": 1.9262268230361004, + "grad_norm": 0.3624087870121002, + "learning_rate": 8.458845731270115e-05, + "loss": 0.0341, + "step": 29400 + }, + { + "epoch": 1.9268820022276092, + "grad_norm": 0.23561225831508636, + "learning_rate": 8.45765154065361e-05, + "loss": 0.0717, + "step": 29410 + }, + { + "epoch": 1.927537181419118, + "grad_norm": 0.4155399799346924, + "learning_rate": 8.456456971913532e-05, + "loss": 0.0389, + "step": 29420 + }, + { + "epoch": 1.928192360610627, + "grad_norm": 0.684379518032074, + "learning_rate": 8.455262025180517e-05, + "loss": 0.0447, + "step": 29430 + }, + { + "epoch": 1.9288475398021359, + "grad_norm": 0.5286465287208557, + "learning_rate": 8.454066700585242e-05, + "loss": 0.0267, + "step": 29440 + }, + { + "epoch": 1.929502718993645, + "grad_norm": 0.6818398833274841, + "learning_rate": 8.452870998258423e-05, + "loss": 0.041, + "step": 29450 + }, + { + "epoch": 1.9301578981851537, + "grad_norm": 0.05451290309429169, + "learning_rate": 8.451674918330825e-05, + "loss": 0.0487, + "step": 29460 + }, + { + "epoch": 1.9308130773766625, + "grad_norm": 0.47031354904174805, + "learning_rate": 8.450478460933246e-05, + "loss": 0.0681, + "step": 29470 + }, + { + "epoch": 1.9314682565681713, + "grad_norm": 0.6959224939346313, + "learning_rate": 8.449281626196532e-05, + "loss": 0.0372, + "step": 29480 + }, + { + "epoch": 1.9321234357596802, + "grad_norm": 0.6589625477790833, + "learning_rate": 8.448084414251564e-05, + "loss": 0.0524, + "step": 29490 + }, + { + "epoch": 1.9327786149511892, + "grad_norm": 0.30548882484436035, + "learning_rate": 8.446886825229271e-05, + "loss": 0.0281, + "step": 29500 + }, + { + "epoch": 1.933433794142698, + "grad_norm": 0.7668283581733704, + "learning_rate": 8.445688859260615e-05, + "loss": 0.0874, + "step": 29510 + }, + { + "epoch": 1.934088973334207, + "grad_norm": 0.33746859431266785, + "learning_rate": 8.444490516476606e-05, + "loss": 0.0265, + "step": 29520 + }, + { + "epoch": 1.9347441525257159, + "grad_norm": 0.8766589760780334, + "learning_rate": 8.443291797008293e-05, + "loss": 0.056, + "step": 29530 + }, + { + "epoch": 1.9353993317172247, + "grad_norm": 0.17581449449062347, + "learning_rate": 8.442092700986765e-05, + "loss": 0.0399, + "step": 29540 + }, + { + "epoch": 1.9360545109087335, + "grad_norm": 0.4291774034500122, + "learning_rate": 8.440893228543156e-05, + "loss": 0.059, + "step": 29550 + }, + { + "epoch": 1.9367096901002423, + "grad_norm": 0.2594916820526123, + "learning_rate": 8.439693379808638e-05, + "loss": 0.0464, + "step": 29560 + }, + { + "epoch": 1.9373648692917513, + "grad_norm": 0.270337849855423, + "learning_rate": 8.43849315491442e-05, + "loss": 0.0383, + "step": 29570 + }, + { + "epoch": 1.9380200484832601, + "grad_norm": 0.9113141894340515, + "learning_rate": 8.437292553991763e-05, + "loss": 0.0475, + "step": 29580 + }, + { + "epoch": 1.9386752276747692, + "grad_norm": 0.16925999522209167, + "learning_rate": 8.436091577171959e-05, + "loss": 0.0227, + "step": 29590 + }, + { + "epoch": 1.939330406866278, + "grad_norm": 0.21366919577121735, + "learning_rate": 8.434890224586347e-05, + "loss": 0.0599, + "step": 29600 + }, + { + "epoch": 1.9399855860577868, + "grad_norm": 0.28948190808296204, + "learning_rate": 8.433688496366303e-05, + "loss": 0.0618, + "step": 29610 + }, + { + "epoch": 1.9406407652492956, + "grad_norm": 0.14228354394435883, + "learning_rate": 8.432486392643248e-05, + "loss": 0.0268, + "step": 29620 + }, + { + "epoch": 1.9412959444408044, + "grad_norm": 0.553494393825531, + "learning_rate": 8.431283913548643e-05, + "loss": 0.0571, + "step": 29630 + }, + { + "epoch": 1.9419511236323135, + "grad_norm": 0.22945690155029297, + "learning_rate": 8.430081059213985e-05, + "loss": 0.0437, + "step": 29640 + }, + { + "epoch": 1.9426063028238223, + "grad_norm": 0.28549641370773315, + "learning_rate": 8.428877829770823e-05, + "loss": 0.0472, + "step": 29650 + }, + { + "epoch": 1.9432614820153313, + "grad_norm": 0.9058216214179993, + "learning_rate": 8.427674225350735e-05, + "loss": 0.0522, + "step": 29660 + }, + { + "epoch": 1.9439166612068401, + "grad_norm": 0.559734046459198, + "learning_rate": 8.426470246085347e-05, + "loss": 0.0556, + "step": 29670 + }, + { + "epoch": 1.944571840398349, + "grad_norm": 0.4848386347293854, + "learning_rate": 8.425265892106324e-05, + "loss": 0.0494, + "step": 29680 + }, + { + "epoch": 1.9452270195898578, + "grad_norm": 0.6853020191192627, + "learning_rate": 8.424061163545374e-05, + "loss": 0.0617, + "step": 29690 + }, + { + "epoch": 1.9458821987813666, + "grad_norm": 0.37917113304138184, + "learning_rate": 8.422856060534243e-05, + "loss": 0.0492, + "step": 29700 + }, + { + "epoch": 1.9465373779728756, + "grad_norm": 0.34999629855155945, + "learning_rate": 8.421650583204718e-05, + "loss": 0.0546, + "step": 29710 + }, + { + "epoch": 1.9471925571643844, + "grad_norm": 0.7899240255355835, + "learning_rate": 8.420444731688633e-05, + "loss": 0.0476, + "step": 29720 + }, + { + "epoch": 1.9478477363558935, + "grad_norm": 0.12612394988536835, + "learning_rate": 8.419238506117852e-05, + "loss": 0.0524, + "step": 29730 + }, + { + "epoch": 1.9485029155474023, + "grad_norm": 0.31103599071502686, + "learning_rate": 8.418031906624289e-05, + "loss": 0.0515, + "step": 29740 + }, + { + "epoch": 1.949158094738911, + "grad_norm": 0.31307902932167053, + "learning_rate": 8.416824933339898e-05, + "loss": 0.0378, + "step": 29750 + }, + { + "epoch": 1.94981327393042, + "grad_norm": 0.3087299168109894, + "learning_rate": 8.415617586396667e-05, + "loss": 0.0451, + "step": 29760 + }, + { + "epoch": 1.9504684531219287, + "grad_norm": 0.47384268045425415, + "learning_rate": 8.414409865926632e-05, + "loss": 0.068, + "step": 29770 + }, + { + "epoch": 1.9511236323134378, + "grad_norm": 0.5524107217788696, + "learning_rate": 8.413201772061867e-05, + "loss": 0.0343, + "step": 29780 + }, + { + "epoch": 1.9517788115049466, + "grad_norm": 0.5009021759033203, + "learning_rate": 8.411993304934488e-05, + "loss": 0.0459, + "step": 29790 + }, + { + "epoch": 1.9524339906964556, + "grad_norm": 0.23673321306705475, + "learning_rate": 8.410784464676654e-05, + "loss": 0.0536, + "step": 29800 + }, + { + "epoch": 1.9530891698879644, + "grad_norm": 0.8111499547958374, + "learning_rate": 8.409575251420556e-05, + "loss": 0.046, + "step": 29810 + }, + { + "epoch": 1.9537443490794733, + "grad_norm": 0.8609753251075745, + "learning_rate": 8.408365665298435e-05, + "loss": 0.0529, + "step": 29820 + }, + { + "epoch": 1.954399528270982, + "grad_norm": 0.39833441376686096, + "learning_rate": 8.40715570644257e-05, + "loss": 0.0487, + "step": 29830 + }, + { + "epoch": 1.9550547074624909, + "grad_norm": 0.2611325681209564, + "learning_rate": 8.40594537498528e-05, + "loss": 0.0191, + "step": 29840 + }, + { + "epoch": 1.955709886654, + "grad_norm": 0.20162847638130188, + "learning_rate": 8.404734671058924e-05, + "loss": 0.0769, + "step": 29850 + }, + { + "epoch": 1.9563650658455087, + "grad_norm": 0.7174428105354309, + "learning_rate": 8.403523594795902e-05, + "loss": 0.0457, + "step": 29860 + }, + { + "epoch": 1.9570202450370178, + "grad_norm": 0.3796204924583435, + "learning_rate": 8.402312146328659e-05, + "loss": 0.0643, + "step": 29870 + }, + { + "epoch": 1.9576754242285266, + "grad_norm": 0.2791224420070648, + "learning_rate": 8.401100325789675e-05, + "loss": 0.0505, + "step": 29880 + }, + { + "epoch": 1.9583306034200354, + "grad_norm": 0.3971833884716034, + "learning_rate": 8.399888133311472e-05, + "loss": 0.0546, + "step": 29890 + }, + { + "epoch": 1.9589857826115442, + "grad_norm": 0.1854439526796341, + "learning_rate": 8.398675569026613e-05, + "loss": 0.0489, + "step": 29900 + }, + { + "epoch": 1.959640961803053, + "grad_norm": 0.24898239970207214, + "learning_rate": 8.397462633067705e-05, + "loss": 0.0396, + "step": 29910 + }, + { + "epoch": 1.960296140994562, + "grad_norm": 0.30391445755958557, + "learning_rate": 8.396249325567392e-05, + "loss": 0.0685, + "step": 29920 + }, + { + "epoch": 1.9609513201860709, + "grad_norm": 0.8943915963172913, + "learning_rate": 8.395035646658357e-05, + "loss": 0.0727, + "step": 29930 + }, + { + "epoch": 1.96160649937758, + "grad_norm": 0.4937747120857239, + "learning_rate": 8.39382159647333e-05, + "loss": 0.0549, + "step": 29940 + }, + { + "epoch": 1.9622616785690887, + "grad_norm": 0.32465171813964844, + "learning_rate": 8.392607175145075e-05, + "loss": 0.0583, + "step": 29950 + }, + { + "epoch": 1.9629168577605975, + "grad_norm": 0.2572081983089447, + "learning_rate": 8.3913923828064e-05, + "loss": 0.0264, + "step": 29960 + }, + { + "epoch": 1.9635720369521064, + "grad_norm": 0.22127892076969147, + "learning_rate": 8.390177219590152e-05, + "loss": 0.0382, + "step": 29970 + }, + { + "epoch": 1.9642272161436152, + "grad_norm": 0.577847957611084, + "learning_rate": 8.388961685629222e-05, + "loss": 0.0484, + "step": 29980 + }, + { + "epoch": 1.9648823953351242, + "grad_norm": 0.8205881118774414, + "learning_rate": 8.387745781056536e-05, + "loss": 0.0441, + "step": 29990 + }, + { + "epoch": 1.965537574526633, + "grad_norm": 0.35683855414390564, + "learning_rate": 8.386529506005065e-05, + "loss": 0.0422, + "step": 30000 + }, + { + "epoch": 1.966192753718142, + "grad_norm": 0.27028536796569824, + "learning_rate": 8.38531286060782e-05, + "loss": 0.0496, + "step": 30010 + }, + { + "epoch": 1.9668479329096509, + "grad_norm": 0.23547334969043732, + "learning_rate": 8.384095844997849e-05, + "loss": 0.0429, + "step": 30020 + }, + { + "epoch": 1.9675031121011597, + "grad_norm": 0.43062329292297363, + "learning_rate": 8.382878459308245e-05, + "loss": 0.0329, + "step": 30030 + }, + { + "epoch": 1.9681582912926685, + "grad_norm": 0.2821032702922821, + "learning_rate": 8.381660703672138e-05, + "loss": 0.0408, + "step": 30040 + }, + { + "epoch": 1.9688134704841773, + "grad_norm": 0.4329789876937866, + "learning_rate": 8.380442578222702e-05, + "loss": 0.0416, + "step": 30050 + }, + { + "epoch": 1.9694686496756861, + "grad_norm": 0.4651685357093811, + "learning_rate": 8.379224083093146e-05, + "loss": 0.0476, + "step": 30060 + }, + { + "epoch": 1.9701238288671952, + "grad_norm": 0.1867295503616333, + "learning_rate": 8.378005218416727e-05, + "loss": 0.0329, + "step": 30070 + }, + { + "epoch": 1.9707790080587042, + "grad_norm": 0.15423324704170227, + "learning_rate": 8.376785984326735e-05, + "loss": 0.0241, + "step": 30080 + }, + { + "epoch": 1.971434187250213, + "grad_norm": 1.1901437044143677, + "learning_rate": 8.375566380956506e-05, + "loss": 0.0451, + "step": 30090 + }, + { + "epoch": 1.9720893664417218, + "grad_norm": 0.08837107568979263, + "learning_rate": 8.374346408439411e-05, + "loss": 0.0271, + "step": 30100 + }, + { + "epoch": 1.9727445456332307, + "grad_norm": 0.3114491105079651, + "learning_rate": 8.373126066908868e-05, + "loss": 0.0264, + "step": 30110 + }, + { + "epoch": 1.9733997248247395, + "grad_norm": 0.22249098122119904, + "learning_rate": 8.371905356498326e-05, + "loss": 0.0311, + "step": 30120 + }, + { + "epoch": 1.9740549040162483, + "grad_norm": 0.2996640205383301, + "learning_rate": 8.370684277341288e-05, + "loss": 0.0336, + "step": 30130 + }, + { + "epoch": 1.9747100832077573, + "grad_norm": 0.6988824605941772, + "learning_rate": 8.369462829571282e-05, + "loss": 0.0425, + "step": 30140 + }, + { + "epoch": 1.9753652623992664, + "grad_norm": 0.3189854025840759, + "learning_rate": 8.36824101332189e-05, + "loss": 0.0602, + "step": 30150 + }, + { + "epoch": 1.9760204415907752, + "grad_norm": 0.3028342127799988, + "learning_rate": 8.367018828726721e-05, + "loss": 0.0737, + "step": 30160 + }, + { + "epoch": 1.976675620782284, + "grad_norm": 1.1844730377197266, + "learning_rate": 8.365796275919438e-05, + "loss": 0.0403, + "step": 30170 + }, + { + "epoch": 1.9773307999737928, + "grad_norm": 0.6372469067573547, + "learning_rate": 8.364573355033734e-05, + "loss": 0.0396, + "step": 30180 + }, + { + "epoch": 1.9779859791653016, + "grad_norm": 1.0356494188308716, + "learning_rate": 8.363350066203346e-05, + "loss": 0.0575, + "step": 30190 + }, + { + "epoch": 1.9786411583568104, + "grad_norm": 0.5703917741775513, + "learning_rate": 8.362126409562053e-05, + "loss": 0.0352, + "step": 30200 + }, + { + "epoch": 1.9792963375483195, + "grad_norm": 0.5894833207130432, + "learning_rate": 8.360902385243667e-05, + "loss": 0.0562, + "step": 30210 + }, + { + "epoch": 1.9799515167398285, + "grad_norm": 0.45806118845939636, + "learning_rate": 8.359677993382052e-05, + "loss": 0.029, + "step": 30220 + }, + { + "epoch": 1.9806066959313373, + "grad_norm": 0.3228779733181, + "learning_rate": 8.358453234111103e-05, + "loss": 0.0543, + "step": 30230 + }, + { + "epoch": 1.9812618751228461, + "grad_norm": 1.2202337980270386, + "learning_rate": 8.357228107564756e-05, + "loss": 0.0421, + "step": 30240 + }, + { + "epoch": 1.981917054314355, + "grad_norm": 0.16375912725925446, + "learning_rate": 8.356002613876993e-05, + "loss": 0.059, + "step": 30250 + }, + { + "epoch": 1.9825722335058638, + "grad_norm": 0.4283241629600525, + "learning_rate": 8.35477675318183e-05, + "loss": 0.0417, + "step": 30260 + }, + { + "epoch": 1.9832274126973726, + "grad_norm": 0.08359966427087784, + "learning_rate": 8.353550525613323e-05, + "loss": 0.0364, + "step": 30270 + }, + { + "epoch": 1.9838825918888816, + "grad_norm": 0.4855291545391083, + "learning_rate": 8.352323931305572e-05, + "loss": 0.032, + "step": 30280 + }, + { + "epoch": 1.9845377710803906, + "grad_norm": 0.3351413905620575, + "learning_rate": 8.351096970392717e-05, + "loss": 0.0339, + "step": 30290 + }, + { + "epoch": 1.9851929502718995, + "grad_norm": 0.15976303815841675, + "learning_rate": 8.349869643008937e-05, + "loss": 0.0381, + "step": 30300 + }, + { + "epoch": 1.9858481294634083, + "grad_norm": 0.25183290243148804, + "learning_rate": 8.348641949288449e-05, + "loss": 0.0623, + "step": 30310 + }, + { + "epoch": 1.986503308654917, + "grad_norm": 0.5947442054748535, + "learning_rate": 8.34741388936551e-05, + "loss": 0.0441, + "step": 30320 + }, + { + "epoch": 1.987158487846426, + "grad_norm": 0.8126329779624939, + "learning_rate": 8.346185463374423e-05, + "loss": 0.0473, + "step": 30330 + }, + { + "epoch": 1.9878136670379347, + "grad_norm": 0.7878909111022949, + "learning_rate": 8.344956671449524e-05, + "loss": 0.0846, + "step": 30340 + }, + { + "epoch": 1.9884688462294438, + "grad_norm": 0.31927385926246643, + "learning_rate": 8.343727513725192e-05, + "loss": 0.0448, + "step": 30350 + }, + { + "epoch": 1.9891240254209528, + "grad_norm": 0.1892528235912323, + "learning_rate": 8.342497990335847e-05, + "loss": 0.0238, + "step": 30360 + }, + { + "epoch": 1.9897792046124616, + "grad_norm": 0.1285087764263153, + "learning_rate": 8.341268101415946e-05, + "loss": 0.0629, + "step": 30370 + }, + { + "epoch": 1.9904343838039704, + "grad_norm": 0.7270863056182861, + "learning_rate": 8.34003784709999e-05, + "loss": 0.0605, + "step": 30380 + }, + { + "epoch": 1.9910895629954792, + "grad_norm": 0.12611769139766693, + "learning_rate": 8.338807227522516e-05, + "loss": 0.0385, + "step": 30390 + }, + { + "epoch": 1.991744742186988, + "grad_norm": 0.25696635246276855, + "learning_rate": 8.337576242818103e-05, + "loss": 0.0321, + "step": 30400 + }, + { + "epoch": 1.9923999213784969, + "grad_norm": 0.509185791015625, + "learning_rate": 8.336344893121372e-05, + "loss": 0.0492, + "step": 30410 + }, + { + "epoch": 1.993055100570006, + "grad_norm": 0.32662054896354675, + "learning_rate": 8.335113178566977e-05, + "loss": 0.0744, + "step": 30420 + }, + { + "epoch": 1.993710279761515, + "grad_norm": 0.20681032538414001, + "learning_rate": 8.33388109928962e-05, + "loss": 0.0561, + "step": 30430 + }, + { + "epoch": 1.9943654589530238, + "grad_norm": 0.8402205109596252, + "learning_rate": 8.33264865542404e-05, + "loss": 0.0301, + "step": 30440 + }, + { + "epoch": 1.9950206381445326, + "grad_norm": 0.2393728494644165, + "learning_rate": 8.331415847105013e-05, + "loss": 0.0402, + "step": 30450 + }, + { + "epoch": 1.9956758173360414, + "grad_norm": 0.4355563521385193, + "learning_rate": 8.330182674467357e-05, + "loss": 0.0765, + "step": 30460 + }, + { + "epoch": 1.9963309965275502, + "grad_norm": 0.5068885684013367, + "learning_rate": 8.32894913764593e-05, + "loss": 0.0413, + "step": 30470 + }, + { + "epoch": 1.996986175719059, + "grad_norm": 0.27245092391967773, + "learning_rate": 8.327715236775633e-05, + "loss": 0.0255, + "step": 30480 + }, + { + "epoch": 1.997641354910568, + "grad_norm": 0.7781189680099487, + "learning_rate": 8.326480971991398e-05, + "loss": 0.0514, + "step": 30490 + }, + { + "epoch": 1.998296534102077, + "grad_norm": 0.19766558706760406, + "learning_rate": 8.325246343428206e-05, + "loss": 0.0389, + "step": 30500 + }, + { + "epoch": 1.998951713293586, + "grad_norm": 0.24262695014476776, + "learning_rate": 8.324011351221072e-05, + "loss": 0.0412, + "step": 30510 + }, + { + "epoch": 1.9996068924850947, + "grad_norm": 0.4470628499984741, + "learning_rate": 8.322775995505057e-05, + "loss": 0.0608, + "step": 30520 + }, + { + "epoch": 2.0002620716766035, + "grad_norm": 0.1572585552930832, + "learning_rate": 8.321540276415254e-05, + "loss": 0.051, + "step": 30530 + }, + { + "epoch": 2.0009172508681123, + "grad_norm": 0.5480362772941589, + "learning_rate": 8.320304194086798e-05, + "loss": 0.0347, + "step": 30540 + }, + { + "epoch": 2.001572430059621, + "grad_norm": 0.5581886172294617, + "learning_rate": 8.31906774865487e-05, + "loss": 0.0458, + "step": 30550 + }, + { + "epoch": 2.00222760925113, + "grad_norm": 0.6459589004516602, + "learning_rate": 8.317830940254682e-05, + "loss": 0.0726, + "step": 30560 + }, + { + "epoch": 2.0028827884426392, + "grad_norm": 0.3580993413925171, + "learning_rate": 8.316593769021491e-05, + "loss": 0.0333, + "step": 30570 + }, + { + "epoch": 2.003537967634148, + "grad_norm": 0.29941022396087646, + "learning_rate": 8.315356235090592e-05, + "loss": 0.037, + "step": 30580 + }, + { + "epoch": 2.004193146825657, + "grad_norm": 1.273345708847046, + "learning_rate": 8.314118338597319e-05, + "loss": 0.0613, + "step": 30590 + }, + { + "epoch": 2.0048483260171657, + "grad_norm": 0.4646587073802948, + "learning_rate": 8.312880079677048e-05, + "loss": 0.0291, + "step": 30600 + }, + { + "epoch": 2.0055035052086745, + "grad_norm": 0.2982255220413208, + "learning_rate": 8.311641458465191e-05, + "loss": 0.0293, + "step": 30610 + }, + { + "epoch": 2.0061586844001833, + "grad_norm": 1.0015934705734253, + "learning_rate": 8.310402475097205e-05, + "loss": 0.0507, + "step": 30620 + }, + { + "epoch": 2.006813863591692, + "grad_norm": 0.21826860308647156, + "learning_rate": 8.309163129708581e-05, + "loss": 0.0566, + "step": 30630 + }, + { + "epoch": 2.0074690427832014, + "grad_norm": 0.32835066318511963, + "learning_rate": 8.307923422434852e-05, + "loss": 0.0235, + "step": 30640 + }, + { + "epoch": 2.00812422197471, + "grad_norm": 0.3905811607837677, + "learning_rate": 8.30668335341159e-05, + "loss": 0.0632, + "step": 30650 + }, + { + "epoch": 2.008779401166219, + "grad_norm": 1.3091797828674316, + "learning_rate": 8.305442922774408e-05, + "loss": 0.063, + "step": 30660 + }, + { + "epoch": 2.009434580357728, + "grad_norm": 0.4980609714984894, + "learning_rate": 8.304202130658959e-05, + "loss": 0.0374, + "step": 30670 + }, + { + "epoch": 2.0100897595492366, + "grad_norm": 0.4715918004512787, + "learning_rate": 8.302960977200931e-05, + "loss": 0.05, + "step": 30680 + }, + { + "epoch": 2.0107449387407454, + "grad_norm": 0.9432736039161682, + "learning_rate": 8.301719462536058e-05, + "loss": 0.0536, + "step": 30690 + }, + { + "epoch": 2.0114001179322543, + "grad_norm": 0.4063423275947571, + "learning_rate": 8.300477586800108e-05, + "loss": 0.0582, + "step": 30700 + }, + { + "epoch": 2.0120552971237635, + "grad_norm": 0.6831021308898926, + "learning_rate": 8.299235350128892e-05, + "loss": 0.0744, + "step": 30710 + }, + { + "epoch": 2.0127104763152723, + "grad_norm": 0.5090296268463135, + "learning_rate": 8.297992752658258e-05, + "loss": 0.0565, + "step": 30720 + }, + { + "epoch": 2.013365655506781, + "grad_norm": 0.3617841899394989, + "learning_rate": 8.296749794524094e-05, + "loss": 0.0571, + "step": 30730 + }, + { + "epoch": 2.01402083469829, + "grad_norm": 0.2904317378997803, + "learning_rate": 8.295506475862332e-05, + "loss": 0.0274, + "step": 30740 + }, + { + "epoch": 2.014676013889799, + "grad_norm": 0.44163650274276733, + "learning_rate": 8.294262796808933e-05, + "loss": 0.0605, + "step": 30750 + }, + { + "epoch": 2.0153311930813076, + "grad_norm": 0.14297616481781006, + "learning_rate": 8.293018757499909e-05, + "loss": 0.0546, + "step": 30760 + }, + { + "epoch": 2.0159863722728164, + "grad_norm": 0.4684356153011322, + "learning_rate": 8.291774358071305e-05, + "loss": 0.0454, + "step": 30770 + }, + { + "epoch": 2.0166415514643257, + "grad_norm": 0.26687368750572205, + "learning_rate": 8.290529598659205e-05, + "loss": 0.0473, + "step": 30780 + }, + { + "epoch": 2.0172967306558345, + "grad_norm": 0.08905215561389923, + "learning_rate": 8.289284479399738e-05, + "loss": 0.0431, + "step": 30790 + }, + { + "epoch": 2.0179519098473433, + "grad_norm": 0.34466105699539185, + "learning_rate": 8.288039000429064e-05, + "loss": 0.033, + "step": 30800 + }, + { + "epoch": 2.018607089038852, + "grad_norm": 0.1605990082025528, + "learning_rate": 8.286793161883388e-05, + "loss": 0.0297, + "step": 30810 + }, + { + "epoch": 2.019262268230361, + "grad_norm": 0.510213315486908, + "learning_rate": 8.285546963898954e-05, + "loss": 0.0447, + "step": 30820 + }, + { + "epoch": 2.0199174474218697, + "grad_norm": 0.13070067763328552, + "learning_rate": 8.284300406612044e-05, + "loss": 0.0287, + "step": 30830 + }, + { + "epoch": 2.0205726266133786, + "grad_norm": 0.47303423285484314, + "learning_rate": 8.283053490158978e-05, + "loss": 0.0745, + "step": 30840 + }, + { + "epoch": 2.021227805804888, + "grad_norm": 0.728404700756073, + "learning_rate": 8.28180621467612e-05, + "loss": 0.0408, + "step": 30850 + }, + { + "epoch": 2.0218829849963966, + "grad_norm": 0.448655903339386, + "learning_rate": 8.280558580299867e-05, + "loss": 0.0454, + "step": 30860 + }, + { + "epoch": 2.0225381641879054, + "grad_norm": 0.4863576889038086, + "learning_rate": 8.279310587166661e-05, + "loss": 0.0629, + "step": 30870 + }, + { + "epoch": 2.0231933433794143, + "grad_norm": 0.7946632504463196, + "learning_rate": 8.278062235412978e-05, + "loss": 0.0482, + "step": 30880 + }, + { + "epoch": 2.023848522570923, + "grad_norm": 0.7159155011177063, + "learning_rate": 8.276813525175339e-05, + "loss": 0.0548, + "step": 30890 + }, + { + "epoch": 2.024503701762432, + "grad_norm": 0.42323967814445496, + "learning_rate": 8.2755644565903e-05, + "loss": 0.056, + "step": 30900 + }, + { + "epoch": 2.0251588809539407, + "grad_norm": 0.5724169015884399, + "learning_rate": 8.274315029794454e-05, + "loss": 0.0505, + "step": 30910 + }, + { + "epoch": 2.02581406014545, + "grad_norm": 0.22084937989711761, + "learning_rate": 8.273065244924443e-05, + "loss": 0.0336, + "step": 30920 + }, + { + "epoch": 2.0264692393369588, + "grad_norm": 0.31041499972343445, + "learning_rate": 8.271815102116936e-05, + "loss": 0.038, + "step": 30930 + }, + { + "epoch": 2.0271244185284676, + "grad_norm": 0.31929364800453186, + "learning_rate": 8.270564601508648e-05, + "loss": 0.0621, + "step": 30940 + }, + { + "epoch": 2.0277795977199764, + "grad_norm": 0.47997474670410156, + "learning_rate": 8.269313743236333e-05, + "loss": 0.0572, + "step": 30950 + }, + { + "epoch": 2.028434776911485, + "grad_norm": 0.33799970149993896, + "learning_rate": 8.268062527436783e-05, + "loss": 0.0328, + "step": 30960 + }, + { + "epoch": 2.029089956102994, + "grad_norm": 0.3806191384792328, + "learning_rate": 8.266810954246828e-05, + "loss": 0.034, + "step": 30970 + }, + { + "epoch": 2.029745135294503, + "grad_norm": 0.7226218581199646, + "learning_rate": 8.265559023803338e-05, + "loss": 0.0413, + "step": 30980 + }, + { + "epoch": 2.030400314486012, + "grad_norm": 0.35653695464134216, + "learning_rate": 8.264306736243223e-05, + "loss": 0.0558, + "step": 30990 + }, + { + "epoch": 2.031055493677521, + "grad_norm": 1.3319001197814941, + "learning_rate": 8.263054091703432e-05, + "loss": 0.0367, + "step": 31000 + }, + { + "epoch": 2.0317106728690297, + "grad_norm": 0.7012376189231873, + "learning_rate": 8.26180109032095e-05, + "loss": 0.0282, + "step": 31010 + }, + { + "epoch": 2.0323658520605385, + "grad_norm": 0.5304422974586487, + "learning_rate": 8.260547732232807e-05, + "loss": 0.0552, + "step": 31020 + }, + { + "epoch": 2.0330210312520474, + "grad_norm": 0.49771392345428467, + "learning_rate": 8.259294017576063e-05, + "loss": 0.0611, + "step": 31030 + }, + { + "epoch": 2.033676210443556, + "grad_norm": 0.8332712054252625, + "learning_rate": 8.258039946487828e-05, + "loss": 0.0483, + "step": 31040 + }, + { + "epoch": 2.034331389635065, + "grad_norm": 1.0630971193313599, + "learning_rate": 8.256785519105241e-05, + "loss": 0.0331, + "step": 31050 + }, + { + "epoch": 2.0349865688265742, + "grad_norm": 0.08376873284578323, + "learning_rate": 8.255530735565488e-05, + "loss": 0.0556, + "step": 31060 + }, + { + "epoch": 2.035641748018083, + "grad_norm": 1.1733074188232422, + "learning_rate": 8.254275596005787e-05, + "loss": 0.0574, + "step": 31070 + }, + { + "epoch": 2.036296927209592, + "grad_norm": 0.6809641718864441, + "learning_rate": 8.2530201005634e-05, + "loss": 0.0688, + "step": 31080 + }, + { + "epoch": 2.0369521064011007, + "grad_norm": 0.48249852657318115, + "learning_rate": 8.251764249375626e-05, + "loss": 0.0567, + "step": 31090 + }, + { + "epoch": 2.0376072855926095, + "grad_norm": 0.45218196511268616, + "learning_rate": 8.250508042579803e-05, + "loss": 0.0349, + "step": 31100 + }, + { + "epoch": 2.0382624647841183, + "grad_norm": 0.25069013237953186, + "learning_rate": 8.249251480313307e-05, + "loss": 0.0456, + "step": 31110 + }, + { + "epoch": 2.038917643975627, + "grad_norm": 0.2384524792432785, + "learning_rate": 8.247994562713555e-05, + "loss": 0.0592, + "step": 31120 + }, + { + "epoch": 2.0395728231671364, + "grad_norm": 0.7831847667694092, + "learning_rate": 8.246737289917998e-05, + "loss": 0.0577, + "step": 31130 + }, + { + "epoch": 2.040228002358645, + "grad_norm": 0.9322441816329956, + "learning_rate": 8.245479662064135e-05, + "loss": 0.0327, + "step": 31140 + }, + { + "epoch": 2.040883181550154, + "grad_norm": 1.2223843336105347, + "learning_rate": 8.244221679289496e-05, + "loss": 0.0437, + "step": 31150 + }, + { + "epoch": 2.041538360741663, + "grad_norm": 0.6544502377510071, + "learning_rate": 8.242963341731652e-05, + "loss": 0.0805, + "step": 31160 + }, + { + "epoch": 2.0421935399331717, + "grad_norm": 0.25806382298469543, + "learning_rate": 8.24170464952821e-05, + "loss": 0.045, + "step": 31170 + }, + { + "epoch": 2.0428487191246805, + "grad_norm": 0.19014760851860046, + "learning_rate": 8.240445602816824e-05, + "loss": 0.0311, + "step": 31180 + }, + { + "epoch": 2.0435038983161893, + "grad_norm": 0.6444626450538635, + "learning_rate": 8.239186201735179e-05, + "loss": 0.0471, + "step": 31190 + }, + { + "epoch": 2.0441590775076985, + "grad_norm": 1.2293822765350342, + "learning_rate": 8.237926446420998e-05, + "loss": 0.0645, + "step": 31200 + }, + { + "epoch": 2.0448142566992074, + "grad_norm": 0.49713394045829773, + "learning_rate": 8.236666337012052e-05, + "loss": 0.0535, + "step": 31210 + }, + { + "epoch": 2.045469435890716, + "grad_norm": 1.0396710634231567, + "learning_rate": 8.23540587364614e-05, + "loss": 0.047, + "step": 31220 + }, + { + "epoch": 2.046124615082225, + "grad_norm": 0.1578870415687561, + "learning_rate": 8.234145056461107e-05, + "loss": 0.049, + "step": 31230 + }, + { + "epoch": 2.046779794273734, + "grad_norm": 1.2315393686294556, + "learning_rate": 8.232883885594832e-05, + "loss": 0.0506, + "step": 31240 + }, + { + "epoch": 2.0474349734652426, + "grad_norm": 0.37347909808158875, + "learning_rate": 8.231622361185236e-05, + "loss": 0.0365, + "step": 31250 + }, + { + "epoch": 2.0480901526567514, + "grad_norm": 0.7871219515800476, + "learning_rate": 8.230360483370278e-05, + "loss": 0.0366, + "step": 31260 + }, + { + "epoch": 2.0487453318482607, + "grad_norm": 0.13941942155361176, + "learning_rate": 8.229098252287953e-05, + "loss": 0.026, + "step": 31270 + }, + { + "epoch": 2.0494005110397695, + "grad_norm": 0.4371366798877716, + "learning_rate": 8.2278356680763e-05, + "loss": 0.0324, + "step": 31280 + }, + { + "epoch": 2.0500556902312783, + "grad_norm": 1.2517331838607788, + "learning_rate": 8.22657273087339e-05, + "loss": 0.0322, + "step": 31290 + }, + { + "epoch": 2.050710869422787, + "grad_norm": 0.733565092086792, + "learning_rate": 8.225309440817336e-05, + "loss": 0.0369, + "step": 31300 + }, + { + "epoch": 2.051366048614296, + "grad_norm": 0.2305726557970047, + "learning_rate": 8.224045798046293e-05, + "loss": 0.0368, + "step": 31310 + }, + { + "epoch": 2.0520212278058048, + "grad_norm": 0.9983813166618347, + "learning_rate": 8.22278180269845e-05, + "loss": 0.0455, + "step": 31320 + }, + { + "epoch": 2.0526764069973136, + "grad_norm": 0.49538373947143555, + "learning_rate": 8.221517454912031e-05, + "loss": 0.0353, + "step": 31330 + }, + { + "epoch": 2.053331586188823, + "grad_norm": 1.3324304819107056, + "learning_rate": 8.220252754825308e-05, + "loss": 0.058, + "step": 31340 + }, + { + "epoch": 2.0539867653803316, + "grad_norm": 0.19084733724594116, + "learning_rate": 8.218987702576586e-05, + "loss": 0.0276, + "step": 31350 + }, + { + "epoch": 2.0546419445718405, + "grad_norm": 0.9042876958847046, + "learning_rate": 8.217722298304207e-05, + "loss": 0.0519, + "step": 31360 + }, + { + "epoch": 2.0552971237633493, + "grad_norm": 1.2198535203933716, + "learning_rate": 8.216456542146557e-05, + "loss": 0.0486, + "step": 31370 + }, + { + "epoch": 2.055952302954858, + "grad_norm": 0.5461651682853699, + "learning_rate": 8.215190434242055e-05, + "loss": 0.0536, + "step": 31380 + }, + { + "epoch": 2.056607482146367, + "grad_norm": 0.3511994183063507, + "learning_rate": 8.213923974729161e-05, + "loss": 0.0674, + "step": 31390 + }, + { + "epoch": 2.0572626613378757, + "grad_norm": 0.553286612033844, + "learning_rate": 8.212657163746373e-05, + "loss": 0.0305, + "step": 31400 + }, + { + "epoch": 2.057917840529385, + "grad_norm": 0.5817808508872986, + "learning_rate": 8.211390001432227e-05, + "loss": 0.0636, + "step": 31410 + }, + { + "epoch": 2.058573019720894, + "grad_norm": 0.5797570943832397, + "learning_rate": 8.210122487925297e-05, + "loss": 0.0559, + "step": 31420 + }, + { + "epoch": 2.0592281989124026, + "grad_norm": 0.690660834312439, + "learning_rate": 8.208854623364202e-05, + "loss": 0.0535, + "step": 31430 + }, + { + "epoch": 2.0598833781039114, + "grad_norm": 0.9576364159584045, + "learning_rate": 8.207586407887589e-05, + "loss": 0.0695, + "step": 31440 + }, + { + "epoch": 2.0605385572954202, + "grad_norm": 0.8236555457115173, + "learning_rate": 8.206317841634148e-05, + "loss": 0.0664, + "step": 31450 + }, + { + "epoch": 2.061193736486929, + "grad_norm": 0.3039817810058594, + "learning_rate": 8.205048924742609e-05, + "loss": 0.0343, + "step": 31460 + }, + { + "epoch": 2.061848915678438, + "grad_norm": 0.5040038228034973, + "learning_rate": 8.203779657351738e-05, + "loss": 0.0721, + "step": 31470 + }, + { + "epoch": 2.062504094869947, + "grad_norm": 0.5454157590866089, + "learning_rate": 8.20251003960034e-05, + "loss": 0.0484, + "step": 31480 + }, + { + "epoch": 2.063159274061456, + "grad_norm": 0.4521949589252472, + "learning_rate": 8.201240071627258e-05, + "loss": 0.06, + "step": 31490 + }, + { + "epoch": 2.0638144532529648, + "grad_norm": 0.47048690915107727, + "learning_rate": 8.199969753571377e-05, + "loss": 0.0296, + "step": 31500 + }, + { + "epoch": 2.0644696324444736, + "grad_norm": 0.5525903701782227, + "learning_rate": 8.198699085571615e-05, + "loss": 0.0328, + "step": 31510 + }, + { + "epoch": 2.0651248116359824, + "grad_norm": 0.5595818161964417, + "learning_rate": 8.197428067766928e-05, + "loss": 0.0345, + "step": 31520 + }, + { + "epoch": 2.065779990827491, + "grad_norm": 0.19800841808319092, + "learning_rate": 8.196156700296316e-05, + "loss": 0.0413, + "step": 31530 + }, + { + "epoch": 2.066435170019, + "grad_norm": 0.862862765789032, + "learning_rate": 8.194884983298814e-05, + "loss": 0.0241, + "step": 31540 + }, + { + "epoch": 2.0670903492105093, + "grad_norm": 0.2704070508480072, + "learning_rate": 8.193612916913491e-05, + "loss": 0.0414, + "step": 31550 + }, + { + "epoch": 2.067745528402018, + "grad_norm": 0.2559894919395447, + "learning_rate": 8.192340501279463e-05, + "loss": 0.0178, + "step": 31560 + }, + { + "epoch": 2.068400707593527, + "grad_norm": 0.3108040392398834, + "learning_rate": 8.191067736535876e-05, + "loss": 0.0567, + "step": 31570 + }, + { + "epoch": 2.0690558867850357, + "grad_norm": 0.5180862545967102, + "learning_rate": 8.18979462282192e-05, + "loss": 0.058, + "step": 31580 + }, + { + "epoch": 2.0697110659765445, + "grad_norm": 0.5120975971221924, + "learning_rate": 8.188521160276819e-05, + "loss": 0.028, + "step": 31590 + }, + { + "epoch": 2.0703662451680533, + "grad_norm": 0.3273506760597229, + "learning_rate": 8.187247349039837e-05, + "loss": 0.0725, + "step": 31600 + }, + { + "epoch": 2.071021424359562, + "grad_norm": 0.7296558022499084, + "learning_rate": 8.185973189250278e-05, + "loss": 0.0604, + "step": 31610 + }, + { + "epoch": 2.0716766035510714, + "grad_norm": 0.409169465303421, + "learning_rate": 8.184698681047482e-05, + "loss": 0.0319, + "step": 31620 + }, + { + "epoch": 2.0723317827425802, + "grad_norm": 0.224674791097641, + "learning_rate": 8.183423824570827e-05, + "loss": 0.0449, + "step": 31630 + }, + { + "epoch": 2.072986961934089, + "grad_norm": 0.29576992988586426, + "learning_rate": 8.182148619959725e-05, + "loss": 0.0373, + "step": 31640 + }, + { + "epoch": 2.073642141125598, + "grad_norm": 0.5477611422538757, + "learning_rate": 8.180873067353636e-05, + "loss": 0.0667, + "step": 31650 + }, + { + "epoch": 2.0742973203171067, + "grad_norm": 0.7061733603477478, + "learning_rate": 8.179597166892052e-05, + "loss": 0.0383, + "step": 31660 + }, + { + "epoch": 2.0749524995086155, + "grad_norm": 0.34628310799598694, + "learning_rate": 8.178320918714501e-05, + "loss": 0.0691, + "step": 31670 + }, + { + "epoch": 2.0756076787001243, + "grad_norm": 0.572598397731781, + "learning_rate": 8.177044322960554e-05, + "loss": 0.0472, + "step": 31680 + }, + { + "epoch": 2.0762628578916336, + "grad_norm": 1.17498779296875, + "learning_rate": 8.175767379769816e-05, + "loss": 0.0527, + "step": 31690 + }, + { + "epoch": 2.0769180370831424, + "grad_norm": 0.5342634916305542, + "learning_rate": 8.174490089281932e-05, + "loss": 0.0477, + "step": 31700 + }, + { + "epoch": 2.077573216274651, + "grad_norm": 1.0694365501403809, + "learning_rate": 8.173212451636584e-05, + "loss": 0.0614, + "step": 31710 + }, + { + "epoch": 2.07822839546616, + "grad_norm": 0.5299078822135925, + "learning_rate": 8.171934466973493e-05, + "loss": 0.0369, + "step": 31720 + }, + { + "epoch": 2.078883574657669, + "grad_norm": 0.6304868459701538, + "learning_rate": 8.170656135432418e-05, + "loss": 0.056, + "step": 31730 + }, + { + "epoch": 2.0795387538491776, + "grad_norm": 0.38245636224746704, + "learning_rate": 8.169377457153155e-05, + "loss": 0.0467, + "step": 31740 + }, + { + "epoch": 2.0801939330406864, + "grad_norm": 0.37205806374549866, + "learning_rate": 8.168098432275539e-05, + "loss": 0.0638, + "step": 31750 + }, + { + "epoch": 2.0808491122321957, + "grad_norm": 0.5598516464233398, + "learning_rate": 8.166819060939442e-05, + "loss": 0.0532, + "step": 31760 + }, + { + "epoch": 2.0815042914237045, + "grad_norm": 0.5703648328781128, + "learning_rate": 8.165539343284772e-05, + "loss": 0.049, + "step": 31770 + }, + { + "epoch": 2.0821594706152133, + "grad_norm": 0.28167834877967834, + "learning_rate": 8.16425927945148e-05, + "loss": 0.0319, + "step": 31780 + }, + { + "epoch": 2.082814649806722, + "grad_norm": 0.38117241859436035, + "learning_rate": 8.162978869579551e-05, + "loss": 0.0239, + "step": 31790 + }, + { + "epoch": 2.083469828998231, + "grad_norm": 0.6443544626235962, + "learning_rate": 8.161698113809007e-05, + "loss": 0.0461, + "step": 31800 + }, + { + "epoch": 2.08412500818974, + "grad_norm": 0.27578070759773254, + "learning_rate": 8.160417012279911e-05, + "loss": 0.0321, + "step": 31810 + }, + { + "epoch": 2.0847801873812486, + "grad_norm": 0.508974015712738, + "learning_rate": 8.159135565132363e-05, + "loss": 0.0628, + "step": 31820 + }, + { + "epoch": 2.085435366572758, + "grad_norm": 1.1343613862991333, + "learning_rate": 8.157853772506498e-05, + "loss": 0.0538, + "step": 31830 + }, + { + "epoch": 2.0860905457642667, + "grad_norm": 0.6376755833625793, + "learning_rate": 8.156571634542494e-05, + "loss": 0.055, + "step": 31840 + }, + { + "epoch": 2.0867457249557755, + "grad_norm": 0.10265766084194183, + "learning_rate": 8.15528915138056e-05, + "loss": 0.0401, + "step": 31850 + }, + { + "epoch": 2.0874009041472843, + "grad_norm": 0.9096131324768066, + "learning_rate": 8.154006323160949e-05, + "loss": 0.0397, + "step": 31860 + }, + { + "epoch": 2.088056083338793, + "grad_norm": 0.7523145079612732, + "learning_rate": 8.152723150023949e-05, + "loss": 0.0416, + "step": 31870 + }, + { + "epoch": 2.088711262530302, + "grad_norm": 0.5393463969230652, + "learning_rate": 8.151439632109886e-05, + "loss": 0.0527, + "step": 31880 + }, + { + "epoch": 2.0893664417218107, + "grad_norm": 0.4121223986148834, + "learning_rate": 8.150155769559122e-05, + "loss": 0.0628, + "step": 31890 + }, + { + "epoch": 2.09002162091332, + "grad_norm": 0.8898232579231262, + "learning_rate": 8.148871562512058e-05, + "loss": 0.0208, + "step": 31900 + }, + { + "epoch": 2.090676800104829, + "grad_norm": 0.6432211995124817, + "learning_rate": 8.147587011109136e-05, + "loss": 0.0325, + "step": 31910 + }, + { + "epoch": 2.0913319792963376, + "grad_norm": 0.2392028421163559, + "learning_rate": 8.14630211549083e-05, + "loss": 0.0621, + "step": 31920 + }, + { + "epoch": 2.0919871584878464, + "grad_norm": 0.16446764767169952, + "learning_rate": 8.145016875797655e-05, + "loss": 0.0369, + "step": 31930 + }, + { + "epoch": 2.0926423376793553, + "grad_norm": 0.8163959383964539, + "learning_rate": 8.143731292170164e-05, + "loss": 0.0278, + "step": 31940 + }, + { + "epoch": 2.093297516870864, + "grad_norm": 1.0419610738754272, + "learning_rate": 8.142445364748944e-05, + "loss": 0.0563, + "step": 31950 + }, + { + "epoch": 2.093952696062373, + "grad_norm": 0.47233259677886963, + "learning_rate": 8.141159093674624e-05, + "loss": 0.0332, + "step": 31960 + }, + { + "epoch": 2.094607875253882, + "grad_norm": 0.3587673008441925, + "learning_rate": 8.139872479087869e-05, + "loss": 0.0468, + "step": 31970 + }, + { + "epoch": 2.095263054445391, + "grad_norm": 0.5063549876213074, + "learning_rate": 8.13858552112938e-05, + "loss": 0.0309, + "step": 31980 + }, + { + "epoch": 2.0959182336368998, + "grad_norm": 0.1891457736492157, + "learning_rate": 8.137298219939895e-05, + "loss": 0.035, + "step": 31990 + }, + { + "epoch": 2.0965734128284086, + "grad_norm": 0.7167643308639526, + "learning_rate": 8.136010575660196e-05, + "loss": 0.042, + "step": 32000 + }, + { + "epoch": 2.0972285920199174, + "grad_norm": 0.41006818413734436, + "learning_rate": 8.134722588431095e-05, + "loss": 0.0416, + "step": 32010 + }, + { + "epoch": 2.097883771211426, + "grad_norm": 0.30504554510116577, + "learning_rate": 8.133434258393444e-05, + "loss": 0.0431, + "step": 32020 + }, + { + "epoch": 2.098538950402935, + "grad_norm": 0.13800881803035736, + "learning_rate": 8.132145585688134e-05, + "loss": 0.0507, + "step": 32030 + }, + { + "epoch": 2.0991941295944443, + "grad_norm": 0.684779942035675, + "learning_rate": 8.130856570456093e-05, + "loss": 0.0342, + "step": 32040 + }, + { + "epoch": 2.099849308785953, + "grad_norm": 0.6679673194885254, + "learning_rate": 8.129567212838283e-05, + "loss": 0.0311, + "step": 32050 + }, + { + "epoch": 2.100504487977462, + "grad_norm": 0.40657883882522583, + "learning_rate": 8.128277512975708e-05, + "loss": 0.0514, + "step": 32060 + }, + { + "epoch": 2.1011596671689707, + "grad_norm": 0.8006504774093628, + "learning_rate": 8.126987471009408e-05, + "loss": 0.0493, + "step": 32070 + }, + { + "epoch": 2.1018148463604795, + "grad_norm": 0.11211194843053818, + "learning_rate": 8.125697087080459e-05, + "loss": 0.0418, + "step": 32080 + }, + { + "epoch": 2.1024700255519884, + "grad_norm": 0.3564334809780121, + "learning_rate": 8.124406361329976e-05, + "loss": 0.0328, + "step": 32090 + }, + { + "epoch": 2.103125204743497, + "grad_norm": 0.5231592059135437, + "learning_rate": 8.123115293899111e-05, + "loss": 0.0476, + "step": 32100 + }, + { + "epoch": 2.1037803839350064, + "grad_norm": 0.8646256923675537, + "learning_rate": 8.121823884929055e-05, + "loss": 0.0418, + "step": 32110 + }, + { + "epoch": 2.1044355631265153, + "grad_norm": 0.459825724363327, + "learning_rate": 8.12053213456103e-05, + "loss": 0.0324, + "step": 32120 + }, + { + "epoch": 2.105090742318024, + "grad_norm": 0.4017142653465271, + "learning_rate": 8.119240042936303e-05, + "loss": 0.0413, + "step": 32130 + }, + { + "epoch": 2.105745921509533, + "grad_norm": 0.2290257066488266, + "learning_rate": 8.117947610196175e-05, + "loss": 0.0327, + "step": 32140 + }, + { + "epoch": 2.1064011007010417, + "grad_norm": 0.741090714931488, + "learning_rate": 8.116654836481982e-05, + "loss": 0.0407, + "step": 32150 + }, + { + "epoch": 2.1070562798925505, + "grad_norm": 0.4654098451137543, + "learning_rate": 8.115361721935106e-05, + "loss": 0.0455, + "step": 32160 + }, + { + "epoch": 2.1077114590840593, + "grad_norm": 0.36063340306282043, + "learning_rate": 8.114068266696953e-05, + "loss": 0.0471, + "step": 32170 + }, + { + "epoch": 2.1083666382755686, + "grad_norm": 0.26175200939178467, + "learning_rate": 8.112774470908978e-05, + "loss": 0.0455, + "step": 32180 + }, + { + "epoch": 2.1090218174670774, + "grad_norm": 0.45549842715263367, + "learning_rate": 8.111480334712665e-05, + "loss": 0.0578, + "step": 32190 + }, + { + "epoch": 2.109676996658586, + "grad_norm": 0.4083307087421417, + "learning_rate": 8.110185858249542e-05, + "loss": 0.0342, + "step": 32200 + }, + { + "epoch": 2.110332175850095, + "grad_norm": 0.5065357685089111, + "learning_rate": 8.108891041661168e-05, + "loss": 0.0354, + "step": 32210 + }, + { + "epoch": 2.110987355041604, + "grad_norm": 0.4827863276004791, + "learning_rate": 8.107595885089146e-05, + "loss": 0.0489, + "step": 32220 + }, + { + "epoch": 2.1116425342331127, + "grad_norm": 0.419366717338562, + "learning_rate": 8.106300388675108e-05, + "loss": 0.0481, + "step": 32230 + }, + { + "epoch": 2.1122977134246215, + "grad_norm": 0.6025139689445496, + "learning_rate": 8.105004552560731e-05, + "loss": 0.0311, + "step": 32240 + }, + { + "epoch": 2.1129528926161303, + "grad_norm": 0.17870333790779114, + "learning_rate": 8.103708376887724e-05, + "loss": 0.0366, + "step": 32250 + }, + { + "epoch": 2.1136080718076395, + "grad_norm": 0.6224058866500854, + "learning_rate": 8.102411861797836e-05, + "loss": 0.0552, + "step": 32260 + }, + { + "epoch": 2.1142632509991484, + "grad_norm": 1.6606855392456055, + "learning_rate": 8.101115007432851e-05, + "loss": 0.0518, + "step": 32270 + }, + { + "epoch": 2.114918430190657, + "grad_norm": 0.7229071259498596, + "learning_rate": 8.09981781393459e-05, + "loss": 0.0263, + "step": 32280 + }, + { + "epoch": 2.115573609382166, + "grad_norm": 0.5242227911949158, + "learning_rate": 8.098520281444915e-05, + "loss": 0.0612, + "step": 32290 + }, + { + "epoch": 2.116228788573675, + "grad_norm": 0.6793241500854492, + "learning_rate": 8.09722241010572e-05, + "loss": 0.0386, + "step": 32300 + }, + { + "epoch": 2.1168839677651836, + "grad_norm": 1.0874340534210205, + "learning_rate": 8.095924200058939e-05, + "loss": 0.0452, + "step": 32310 + }, + { + "epoch": 2.117539146956693, + "grad_norm": 0.35210472345352173, + "learning_rate": 8.094625651446541e-05, + "loss": 0.0638, + "step": 32320 + }, + { + "epoch": 2.1181943261482017, + "grad_norm": 0.10503087937831879, + "learning_rate": 8.093326764410536e-05, + "loss": 0.0713, + "step": 32330 + }, + { + "epoch": 2.1188495053397105, + "grad_norm": 0.3353225290775299, + "learning_rate": 8.092027539092966e-05, + "loss": 0.0581, + "step": 32340 + }, + { + "epoch": 2.1195046845312193, + "grad_norm": 0.4372682571411133, + "learning_rate": 8.090727975635913e-05, + "loss": 0.0375, + "step": 32350 + }, + { + "epoch": 2.120159863722728, + "grad_norm": 1.0207314491271973, + "learning_rate": 8.089428074181497e-05, + "loss": 0.053, + "step": 32360 + }, + { + "epoch": 2.120815042914237, + "grad_norm": 0.68962562084198, + "learning_rate": 8.088127834871871e-05, + "loss": 0.0327, + "step": 32370 + }, + { + "epoch": 2.1214702221057458, + "grad_norm": 1.2369431257247925, + "learning_rate": 8.086827257849226e-05, + "loss": 0.0543, + "step": 32380 + }, + { + "epoch": 2.1221254012972546, + "grad_norm": 0.5353200435638428, + "learning_rate": 8.085526343255795e-05, + "loss": 0.0372, + "step": 32390 + }, + { + "epoch": 2.122780580488764, + "grad_norm": 0.5372066497802734, + "learning_rate": 8.084225091233842e-05, + "loss": 0.0412, + "step": 32400 + }, + { + "epoch": 2.1234357596802726, + "grad_norm": 0.6422219276428223, + "learning_rate": 8.082923501925668e-05, + "loss": 0.0885, + "step": 32410 + }, + { + "epoch": 2.1240909388717815, + "grad_norm": 0.1009458675980568, + "learning_rate": 8.081621575473617e-05, + "loss": 0.0353, + "step": 32420 + }, + { + "epoch": 2.1247461180632903, + "grad_norm": 0.26907190680503845, + "learning_rate": 8.080319312020064e-05, + "loss": 0.0473, + "step": 32430 + }, + { + "epoch": 2.125401297254799, + "grad_norm": 0.4631701111793518, + "learning_rate": 8.079016711707421e-05, + "loss": 0.0376, + "step": 32440 + }, + { + "epoch": 2.126056476446308, + "grad_norm": 1.0856350660324097, + "learning_rate": 8.077713774678139e-05, + "loss": 0.0452, + "step": 32450 + }, + { + "epoch": 2.126711655637817, + "grad_norm": 0.3118613064289093, + "learning_rate": 8.076410501074707e-05, + "loss": 0.0575, + "step": 32460 + }, + { + "epoch": 2.127366834829326, + "grad_norm": 0.12270630896091461, + "learning_rate": 8.075106891039647e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 2.128022014020835, + "grad_norm": 1.2869633436203003, + "learning_rate": 8.073802944715523e-05, + "loss": 0.0667, + "step": 32480 + }, + { + "epoch": 2.1286771932123436, + "grad_norm": 2.375849723815918, + "learning_rate": 8.072498662244929e-05, + "loss": 0.0817, + "step": 32490 + }, + { + "epoch": 2.1293323724038524, + "grad_norm": 0.5718629360198975, + "learning_rate": 8.0711940437705e-05, + "loss": 0.05, + "step": 32500 + }, + { + "epoch": 2.1299875515953612, + "grad_norm": 0.4532005786895752, + "learning_rate": 8.06988908943491e-05, + "loss": 0.0286, + "step": 32510 + }, + { + "epoch": 2.13064273078687, + "grad_norm": 0.21091057360172272, + "learning_rate": 8.068583799380863e-05, + "loss": 0.0922, + "step": 32520 + }, + { + "epoch": 2.131297909978379, + "grad_norm": 0.667627215385437, + "learning_rate": 8.067278173751104e-05, + "loss": 0.0503, + "step": 32530 + }, + { + "epoch": 2.131953089169888, + "grad_norm": 0.8270273208618164, + "learning_rate": 8.065972212688417e-05, + "loss": 0.0812, + "step": 32540 + }, + { + "epoch": 2.132608268361397, + "grad_norm": 0.28731000423431396, + "learning_rate": 8.064665916335618e-05, + "loss": 0.0281, + "step": 32550 + }, + { + "epoch": 2.1332634475529058, + "grad_norm": 0.5976918935775757, + "learning_rate": 8.063359284835564e-05, + "loss": 0.0455, + "step": 32560 + }, + { + "epoch": 2.1339186267444146, + "grad_norm": 0.6654462218284607, + "learning_rate": 8.062052318331142e-05, + "loss": 0.0252, + "step": 32570 + }, + { + "epoch": 2.1345738059359234, + "grad_norm": 0.4914608895778656, + "learning_rate": 8.060745016965283e-05, + "loss": 0.0386, + "step": 32580 + }, + { + "epoch": 2.135228985127432, + "grad_norm": 0.2948513329029083, + "learning_rate": 8.059437380880952e-05, + "loss": 0.0462, + "step": 32590 + }, + { + "epoch": 2.1358841643189415, + "grad_norm": 0.6913928985595703, + "learning_rate": 8.058129410221146e-05, + "loss": 0.0498, + "step": 32600 + }, + { + "epoch": 2.1365393435104503, + "grad_norm": 0.628453254699707, + "learning_rate": 8.056821105128908e-05, + "loss": 0.0575, + "step": 32610 + }, + { + "epoch": 2.137194522701959, + "grad_norm": 0.5820751190185547, + "learning_rate": 8.05551246574731e-05, + "loss": 0.0536, + "step": 32620 + }, + { + "epoch": 2.137849701893468, + "grad_norm": 0.24135315418243408, + "learning_rate": 8.05420349221946e-05, + "loss": 0.0256, + "step": 32630 + }, + { + "epoch": 2.1385048810849767, + "grad_norm": 0.7750040888786316, + "learning_rate": 8.05289418468851e-05, + "loss": 0.0436, + "step": 32640 + }, + { + "epoch": 2.1391600602764855, + "grad_norm": 0.6354228258132935, + "learning_rate": 8.051584543297642e-05, + "loss": 0.0772, + "step": 32650 + }, + { + "epoch": 2.1398152394679943, + "grad_norm": 0.10142135620117188, + "learning_rate": 8.050274568190074e-05, + "loss": 0.0489, + "step": 32660 + }, + { + "epoch": 2.140470418659503, + "grad_norm": 0.7942535877227783, + "learning_rate": 8.048964259509067e-05, + "loss": 0.056, + "step": 32670 + }, + { + "epoch": 2.1411255978510124, + "grad_norm": 0.6802778244018555, + "learning_rate": 8.047653617397914e-05, + "loss": 0.0473, + "step": 32680 + }, + { + "epoch": 2.1417807770425212, + "grad_norm": 0.6732683777809143, + "learning_rate": 8.046342641999941e-05, + "loss": 0.077, + "step": 32690 + }, + { + "epoch": 2.14243595623403, + "grad_norm": 0.7046927213668823, + "learning_rate": 8.045031333458517e-05, + "loss": 0.0582, + "step": 32700 + }, + { + "epoch": 2.143091135425539, + "grad_norm": 0.3808378279209137, + "learning_rate": 8.043719691917047e-05, + "loss": 0.0619, + "step": 32710 + }, + { + "epoch": 2.1437463146170477, + "grad_norm": 0.18721063435077667, + "learning_rate": 8.042407717518966e-05, + "loss": 0.0339, + "step": 32720 + }, + { + "epoch": 2.1444014938085565, + "grad_norm": 0.7544547915458679, + "learning_rate": 8.041095410407751e-05, + "loss": 0.0324, + "step": 32730 + }, + { + "epoch": 2.1450566730000658, + "grad_norm": 0.16630999743938446, + "learning_rate": 8.039782770726913e-05, + "loss": 0.03, + "step": 32740 + }, + { + "epoch": 2.1457118521915746, + "grad_norm": 0.9663761258125305, + "learning_rate": 8.038469798620004e-05, + "loss": 0.052, + "step": 32750 + }, + { + "epoch": 2.1463670313830834, + "grad_norm": 1.3002241849899292, + "learning_rate": 8.037156494230604e-05, + "loss": 0.0626, + "step": 32760 + }, + { + "epoch": 2.147022210574592, + "grad_norm": 0.8786734938621521, + "learning_rate": 8.035842857702338e-05, + "loss": 0.0354, + "step": 32770 + }, + { + "epoch": 2.147677389766101, + "grad_norm": 0.18586191534996033, + "learning_rate": 8.03452888917886e-05, + "loss": 0.0401, + "step": 32780 + }, + { + "epoch": 2.14833256895761, + "grad_norm": 0.4773220717906952, + "learning_rate": 8.033214588803866e-05, + "loss": 0.04, + "step": 32790 + }, + { + "epoch": 2.1489877481491186, + "grad_norm": 0.33412787318229675, + "learning_rate": 8.031899956721083e-05, + "loss": 0.0438, + "step": 32800 + }, + { + "epoch": 2.1496429273406275, + "grad_norm": 0.13760176301002502, + "learning_rate": 8.030584993074282e-05, + "loss": 0.0499, + "step": 32810 + }, + { + "epoch": 2.1502981065321367, + "grad_norm": 0.46564772725105286, + "learning_rate": 8.02926969800726e-05, + "loss": 0.0259, + "step": 32820 + }, + { + "epoch": 2.1509532857236455, + "grad_norm": 0.7883162498474121, + "learning_rate": 8.027954071663859e-05, + "loss": 0.0328, + "step": 32830 + }, + { + "epoch": 2.1516084649151543, + "grad_norm": 0.45223695039749146, + "learning_rate": 8.026638114187954e-05, + "loss": 0.0314, + "step": 32840 + }, + { + "epoch": 2.152263644106663, + "grad_norm": 1.413020133972168, + "learning_rate": 8.025321825723456e-05, + "loss": 0.0432, + "step": 32850 + }, + { + "epoch": 2.152918823298172, + "grad_norm": 0.6516719460487366, + "learning_rate": 8.02400520641431e-05, + "loss": 0.0499, + "step": 32860 + }, + { + "epoch": 2.153574002489681, + "grad_norm": 0.13077011704444885, + "learning_rate": 8.022688256404501e-05, + "loss": 0.0321, + "step": 32870 + }, + { + "epoch": 2.1542291816811896, + "grad_norm": 0.7414909601211548, + "learning_rate": 8.02137097583805e-05, + "loss": 0.0363, + "step": 32880 + }, + { + "epoch": 2.154884360872699, + "grad_norm": 0.31829214096069336, + "learning_rate": 8.02005336485901e-05, + "loss": 0.0484, + "step": 32890 + }, + { + "epoch": 2.1555395400642077, + "grad_norm": 0.18530985713005066, + "learning_rate": 8.018735423611476e-05, + "loss": 0.0296, + "step": 32900 + }, + { + "epoch": 2.1561947192557165, + "grad_norm": 0.385280579328537, + "learning_rate": 8.017417152239574e-05, + "loss": 0.0356, + "step": 32910 + }, + { + "epoch": 2.1568498984472253, + "grad_norm": 0.6479750275611877, + "learning_rate": 8.01609855088747e-05, + "loss": 0.0491, + "step": 32920 + }, + { + "epoch": 2.157505077638734, + "grad_norm": 0.2825072705745697, + "learning_rate": 8.014779619699362e-05, + "loss": 0.0309, + "step": 32930 + }, + { + "epoch": 2.158160256830243, + "grad_norm": 0.2383406162261963, + "learning_rate": 8.013460358819489e-05, + "loss": 0.0564, + "step": 32940 + }, + { + "epoch": 2.1588154360217517, + "grad_norm": 0.9116112589836121, + "learning_rate": 8.01214076839212e-05, + "loss": 0.0456, + "step": 32950 + }, + { + "epoch": 2.159470615213261, + "grad_norm": 0.09590993821620941, + "learning_rate": 8.010820848561565e-05, + "loss": 0.0607, + "step": 32960 + }, + { + "epoch": 2.16012579440477, + "grad_norm": 0.7929003238677979, + "learning_rate": 8.009500599472171e-05, + "loss": 0.0391, + "step": 32970 + }, + { + "epoch": 2.1607809735962786, + "grad_norm": 0.25853386521339417, + "learning_rate": 8.008180021268314e-05, + "loss": 0.032, + "step": 32980 + }, + { + "epoch": 2.1614361527877874, + "grad_norm": 0.5385251641273499, + "learning_rate": 8.006859114094414e-05, + "loss": 0.0546, + "step": 32990 + }, + { + "epoch": 2.1620913319792963, + "grad_norm": 0.16704261302947998, + "learning_rate": 8.005537878094921e-05, + "loss": 0.0615, + "step": 33000 + }, + { + "epoch": 2.162746511170805, + "grad_norm": 0.20783421397209167, + "learning_rate": 8.004216313414323e-05, + "loss": 0.0345, + "step": 33010 + }, + { + "epoch": 2.163401690362314, + "grad_norm": 0.2511434555053711, + "learning_rate": 8.002894420197149e-05, + "loss": 0.0198, + "step": 33020 + }, + { + "epoch": 2.164056869553823, + "grad_norm": 0.3373400866985321, + "learning_rate": 8.001572198587954e-05, + "loss": 0.0314, + "step": 33030 + }, + { + "epoch": 2.164712048745332, + "grad_norm": 0.25260505080223083, + "learning_rate": 8.000249648731338e-05, + "loss": 0.0361, + "step": 33040 + }, + { + "epoch": 2.1653672279368408, + "grad_norm": 0.6459565162658691, + "learning_rate": 7.998926770771928e-05, + "loss": 0.0393, + "step": 33050 + }, + { + "epoch": 2.1660224071283496, + "grad_norm": 0.4541533291339874, + "learning_rate": 7.997603564854397e-05, + "loss": 0.0369, + "step": 33060 + }, + { + "epoch": 2.1666775863198584, + "grad_norm": 0.8325661420822144, + "learning_rate": 7.996280031123448e-05, + "loss": 0.0398, + "step": 33070 + }, + { + "epoch": 2.167332765511367, + "grad_norm": 0.6187415719032288, + "learning_rate": 7.994956169723818e-05, + "loss": 0.0372, + "step": 33080 + }, + { + "epoch": 2.167987944702876, + "grad_norm": 0.11533107608556747, + "learning_rate": 7.993631980800285e-05, + "loss": 0.0801, + "step": 33090 + }, + { + "epoch": 2.1686431238943853, + "grad_norm": 0.14659100770950317, + "learning_rate": 7.992307464497659e-05, + "loss": 0.0569, + "step": 33100 + }, + { + "epoch": 2.169298303085894, + "grad_norm": 0.18715502321720123, + "learning_rate": 7.990982620960787e-05, + "loss": 0.0328, + "step": 33110 + }, + { + "epoch": 2.169953482277403, + "grad_norm": 1.079385757446289, + "learning_rate": 7.989657450334554e-05, + "loss": 0.0695, + "step": 33120 + }, + { + "epoch": 2.1706086614689117, + "grad_norm": 0.1737581342458725, + "learning_rate": 7.988331952763877e-05, + "loss": 0.0172, + "step": 33130 + }, + { + "epoch": 2.1712638406604206, + "grad_norm": 0.36679139733314514, + "learning_rate": 7.987006128393709e-05, + "loss": 0.0395, + "step": 33140 + }, + { + "epoch": 2.1719190198519294, + "grad_norm": 0.3463205099105835, + "learning_rate": 7.985679977369043e-05, + "loss": 0.0315, + "step": 33150 + }, + { + "epoch": 2.172574199043438, + "grad_norm": 0.378921777009964, + "learning_rate": 7.984353499834902e-05, + "loss": 0.0261, + "step": 33160 + }, + { + "epoch": 2.1732293782349474, + "grad_norm": 1.1245275735855103, + "learning_rate": 7.983026695936351e-05, + "loss": 0.0726, + "step": 33170 + }, + { + "epoch": 2.1738845574264563, + "grad_norm": 0.24016143381595612, + "learning_rate": 7.981699565818486e-05, + "loss": 0.0552, + "step": 33180 + }, + { + "epoch": 2.174539736617965, + "grad_norm": 0.6051830649375916, + "learning_rate": 7.980372109626437e-05, + "loss": 0.0397, + "step": 33190 + }, + { + "epoch": 2.175194915809474, + "grad_norm": 0.26389870047569275, + "learning_rate": 7.979044327505375e-05, + "loss": 0.0456, + "step": 33200 + }, + { + "epoch": 2.1758500950009827, + "grad_norm": 0.34468379616737366, + "learning_rate": 7.977716219600506e-05, + "loss": 0.0284, + "step": 33210 + }, + { + "epoch": 2.1765052741924915, + "grad_norm": 0.12796060740947723, + "learning_rate": 7.97638778605707e-05, + "loss": 0.0474, + "step": 33220 + }, + { + "epoch": 2.1771604533840003, + "grad_norm": 0.3375605642795563, + "learning_rate": 7.975059027020338e-05, + "loss": 0.0288, + "step": 33230 + }, + { + "epoch": 2.1778156325755096, + "grad_norm": 0.3062928020954132, + "learning_rate": 7.973729942635623e-05, + "loss": 0.0429, + "step": 33240 + }, + { + "epoch": 2.1784708117670184, + "grad_norm": 0.6573184132575989, + "learning_rate": 7.972400533048273e-05, + "loss": 0.0506, + "step": 33250 + }, + { + "epoch": 2.179125990958527, + "grad_norm": 0.20434722304344177, + "learning_rate": 7.97107079840367e-05, + "loss": 0.0447, + "step": 33260 + }, + { + "epoch": 2.179781170150036, + "grad_norm": 0.7377394437789917, + "learning_rate": 7.969740738847231e-05, + "loss": 0.0482, + "step": 33270 + }, + { + "epoch": 2.180436349341545, + "grad_norm": 0.30467137694358826, + "learning_rate": 7.968410354524411e-05, + "loss": 0.0276, + "step": 33280 + }, + { + "epoch": 2.1810915285330537, + "grad_norm": 0.9293165802955627, + "learning_rate": 7.967079645580697e-05, + "loss": 0.0455, + "step": 33290 + }, + { + "epoch": 2.1817467077245625, + "grad_norm": 0.24148523807525635, + "learning_rate": 7.965748612161612e-05, + "loss": 0.0201, + "step": 33300 + }, + { + "epoch": 2.1824018869160717, + "grad_norm": 1.6499481201171875, + "learning_rate": 7.96441725441272e-05, + "loss": 0.082, + "step": 33310 + }, + { + "epoch": 2.1830570661075805, + "grad_norm": 0.14973556995391846, + "learning_rate": 7.963085572479614e-05, + "loss": 0.0507, + "step": 33320 + }, + { + "epoch": 2.1837122452990894, + "grad_norm": 0.43793973326683044, + "learning_rate": 7.961753566507924e-05, + "loss": 0.0361, + "step": 33330 + }, + { + "epoch": 2.184367424490598, + "grad_norm": 1.766090750694275, + "learning_rate": 7.960421236643316e-05, + "loss": 0.029, + "step": 33340 + }, + { + "epoch": 2.185022603682107, + "grad_norm": 0.23271574079990387, + "learning_rate": 7.959088583031496e-05, + "loss": 0.0356, + "step": 33350 + }, + { + "epoch": 2.185677782873616, + "grad_norm": 0.3837582468986511, + "learning_rate": 7.957755605818194e-05, + "loss": 0.036, + "step": 33360 + }, + { + "epoch": 2.1863329620651246, + "grad_norm": 0.31816616654396057, + "learning_rate": 7.956422305149185e-05, + "loss": 0.0611, + "step": 33370 + }, + { + "epoch": 2.186988141256634, + "grad_norm": 0.7298951745033264, + "learning_rate": 7.95508868117028e-05, + "loss": 0.0484, + "step": 33380 + }, + { + "epoch": 2.1876433204481427, + "grad_norm": 0.5103100538253784, + "learning_rate": 7.953754734027318e-05, + "loss": 0.0511, + "step": 33390 + }, + { + "epoch": 2.1882984996396515, + "grad_norm": 0.9591511487960815, + "learning_rate": 7.952420463866182e-05, + "loss": 0.0444, + "step": 33400 + }, + { + "epoch": 2.1889536788311603, + "grad_norm": 0.3173859119415283, + "learning_rate": 7.951085870832782e-05, + "loss": 0.0417, + "step": 33410 + }, + { + "epoch": 2.189608858022669, + "grad_norm": 0.3800489008426666, + "learning_rate": 7.949750955073067e-05, + "loss": 0.0517, + "step": 33420 + }, + { + "epoch": 2.190264037214178, + "grad_norm": 0.2188979834318161, + "learning_rate": 7.948415716733022e-05, + "loss": 0.0465, + "step": 33430 + }, + { + "epoch": 2.1909192164056868, + "grad_norm": 0.12009687721729279, + "learning_rate": 7.947080155958669e-05, + "loss": 0.0366, + "step": 33440 + }, + { + "epoch": 2.191574395597196, + "grad_norm": 0.28546956181526184, + "learning_rate": 7.94574427289606e-05, + "loss": 0.0514, + "step": 33450 + }, + { + "epoch": 2.192229574788705, + "grad_norm": 0.45803868770599365, + "learning_rate": 7.944408067691284e-05, + "loss": 0.0299, + "step": 33460 + }, + { + "epoch": 2.1928847539802137, + "grad_norm": 0.22530408203601837, + "learning_rate": 7.943071540490473e-05, + "loss": 0.0378, + "step": 33470 + }, + { + "epoch": 2.1935399331717225, + "grad_norm": 1.0846627950668335, + "learning_rate": 7.94173469143978e-05, + "loss": 0.0533, + "step": 33480 + }, + { + "epoch": 2.1941951123632313, + "grad_norm": 1.0427979230880737, + "learning_rate": 7.940397520685406e-05, + "loss": 0.0336, + "step": 33490 + }, + { + "epoch": 2.19485029155474, + "grad_norm": 1.5182174444198608, + "learning_rate": 7.939060028373577e-05, + "loss": 0.0816, + "step": 33500 + }, + { + "epoch": 2.195505470746249, + "grad_norm": 0.23024596273899078, + "learning_rate": 7.937722214650565e-05, + "loss": 0.0619, + "step": 33510 + }, + { + "epoch": 2.196160649937758, + "grad_norm": 0.8173030614852905, + "learning_rate": 7.936384079662666e-05, + "loss": 0.0345, + "step": 33520 + }, + { + "epoch": 2.196815829129267, + "grad_norm": 0.4566074013710022, + "learning_rate": 7.93504562355622e-05, + "loss": 0.0353, + "step": 33530 + }, + { + "epoch": 2.197471008320776, + "grad_norm": 0.41708996891975403, + "learning_rate": 7.933706846477599e-05, + "loss": 0.0275, + "step": 33540 + }, + { + "epoch": 2.1981261875122846, + "grad_norm": 0.20413869619369507, + "learning_rate": 7.932367748573206e-05, + "loss": 0.0553, + "step": 33550 + }, + { + "epoch": 2.1987813667037934, + "grad_norm": 0.25420743227005005, + "learning_rate": 7.931028329989485e-05, + "loss": 0.0289, + "step": 33560 + }, + { + "epoch": 2.1994365458953022, + "grad_norm": 0.12338574230670929, + "learning_rate": 7.929688590872913e-05, + "loss": 0.0499, + "step": 33570 + }, + { + "epoch": 2.200091725086811, + "grad_norm": 0.3873741924762726, + "learning_rate": 7.928348531370003e-05, + "loss": 0.0385, + "step": 33580 + }, + { + "epoch": 2.2007469042783203, + "grad_norm": 0.553817629814148, + "learning_rate": 7.927008151627297e-05, + "loss": 0.0321, + "step": 33590 + }, + { + "epoch": 2.201402083469829, + "grad_norm": 0.19750496745109558, + "learning_rate": 7.925667451791383e-05, + "loss": 0.0246, + "step": 33600 + }, + { + "epoch": 2.202057262661338, + "grad_norm": 0.7668095231056213, + "learning_rate": 7.924326432008874e-05, + "loss": 0.0508, + "step": 33610 + }, + { + "epoch": 2.2027124418528468, + "grad_norm": 0.6824795007705688, + "learning_rate": 7.922985092426422e-05, + "loss": 0.0282, + "step": 33620 + }, + { + "epoch": 2.2033676210443556, + "grad_norm": 0.9650528430938721, + "learning_rate": 7.921643433190717e-05, + "loss": 0.0286, + "step": 33630 + }, + { + "epoch": 2.2040228002358644, + "grad_norm": 0.5695241689682007, + "learning_rate": 7.920301454448478e-05, + "loss": 0.0526, + "step": 33640 + }, + { + "epoch": 2.204677979427373, + "grad_norm": 0.4325770437717438, + "learning_rate": 7.918959156346461e-05, + "loss": 0.0249, + "step": 33650 + }, + { + "epoch": 2.2053331586188825, + "grad_norm": 0.6295238733291626, + "learning_rate": 7.91761653903146e-05, + "loss": 0.0427, + "step": 33660 + }, + { + "epoch": 2.2059883378103913, + "grad_norm": 0.1911918669939041, + "learning_rate": 7.916273602650302e-05, + "loss": 0.0205, + "step": 33670 + }, + { + "epoch": 2.2066435170019, + "grad_norm": 0.840073823928833, + "learning_rate": 7.914930347349847e-05, + "loss": 0.0378, + "step": 33680 + }, + { + "epoch": 2.207298696193409, + "grad_norm": 0.4899340569972992, + "learning_rate": 7.913586773276992e-05, + "loss": 0.0355, + "step": 33690 + }, + { + "epoch": 2.2079538753849177, + "grad_norm": 0.505113959312439, + "learning_rate": 7.912242880578667e-05, + "loss": 0.0453, + "step": 33700 + }, + { + "epoch": 2.2086090545764265, + "grad_norm": 0.13288632035255432, + "learning_rate": 7.910898669401839e-05, + "loss": 0.0804, + "step": 33710 + }, + { + "epoch": 2.2092642337679353, + "grad_norm": 0.5690181851387024, + "learning_rate": 7.909554139893511e-05, + "loss": 0.0587, + "step": 33720 + }, + { + "epoch": 2.2099194129594446, + "grad_norm": 0.8513482213020325, + "learning_rate": 7.908209292200715e-05, + "loss": 0.0316, + "step": 33730 + }, + { + "epoch": 2.2105745921509534, + "grad_norm": 1.1014230251312256, + "learning_rate": 7.906864126470523e-05, + "loss": 0.0591, + "step": 33740 + }, + { + "epoch": 2.2112297713424622, + "grad_norm": 0.6618552207946777, + "learning_rate": 7.905518642850041e-05, + "loss": 0.0338, + "step": 33750 + }, + { + "epoch": 2.211884950533971, + "grad_norm": 0.07658248394727707, + "learning_rate": 7.904172841486409e-05, + "loss": 0.0284, + "step": 33760 + }, + { + "epoch": 2.21254012972548, + "grad_norm": 1.0154571533203125, + "learning_rate": 7.902826722526801e-05, + "loss": 0.036, + "step": 33770 + }, + { + "epoch": 2.2131953089169887, + "grad_norm": 0.8953753709793091, + "learning_rate": 7.901480286118427e-05, + "loss": 0.0513, + "step": 33780 + }, + { + "epoch": 2.2138504881084975, + "grad_norm": 0.40812602639198303, + "learning_rate": 7.900133532408531e-05, + "loss": 0.0481, + "step": 33790 + }, + { + "epoch": 2.2145056673000068, + "grad_norm": 0.9011253714561462, + "learning_rate": 7.898786461544395e-05, + "loss": 0.0585, + "step": 33800 + }, + { + "epoch": 2.2151608464915156, + "grad_norm": 1.273047685623169, + "learning_rate": 7.897439073673325e-05, + "loss": 0.0329, + "step": 33810 + }, + { + "epoch": 2.2158160256830244, + "grad_norm": 0.730854868888855, + "learning_rate": 7.896091368942677e-05, + "loss": 0.0415, + "step": 33820 + }, + { + "epoch": 2.216471204874533, + "grad_norm": 0.4216933250427246, + "learning_rate": 7.894743347499832e-05, + "loss": 0.0345, + "step": 33830 + }, + { + "epoch": 2.217126384066042, + "grad_norm": 0.25007912516593933, + "learning_rate": 7.893395009492203e-05, + "loss": 0.0376, + "step": 33840 + }, + { + "epoch": 2.217781563257551, + "grad_norm": 0.96383136510849, + "learning_rate": 7.892046355067248e-05, + "loss": 0.0302, + "step": 33850 + }, + { + "epoch": 2.2184367424490596, + "grad_norm": 1.6511667966842651, + "learning_rate": 7.890697384372451e-05, + "loss": 0.0453, + "step": 33860 + }, + { + "epoch": 2.219091921640569, + "grad_norm": 0.3235699236392975, + "learning_rate": 7.889348097555336e-05, + "loss": 0.0371, + "step": 33870 + }, + { + "epoch": 2.2197471008320777, + "grad_norm": 0.9152939915657043, + "learning_rate": 7.887998494763455e-05, + "loss": 0.0465, + "step": 33880 + }, + { + "epoch": 2.2204022800235865, + "grad_norm": 0.8378221392631531, + "learning_rate": 7.886648576144404e-05, + "loss": 0.0252, + "step": 33890 + }, + { + "epoch": 2.2210574592150953, + "grad_norm": 0.5294920206069946, + "learning_rate": 7.885298341845802e-05, + "loss": 0.0345, + "step": 33900 + }, + { + "epoch": 2.221712638406604, + "grad_norm": 0.5941353440284729, + "learning_rate": 7.883947792015311e-05, + "loss": 0.0436, + "step": 33910 + }, + { + "epoch": 2.222367817598113, + "grad_norm": 0.7923770546913147, + "learning_rate": 7.882596926800628e-05, + "loss": 0.052, + "step": 33920 + }, + { + "epoch": 2.223022996789622, + "grad_norm": 0.7950312495231628, + "learning_rate": 7.881245746349477e-05, + "loss": 0.0386, + "step": 33930 + }, + { + "epoch": 2.223678175981131, + "grad_norm": 0.5196599960327148, + "learning_rate": 7.879894250809623e-05, + "loss": 0.0472, + "step": 33940 + }, + { + "epoch": 2.22433335517264, + "grad_norm": 0.5284286141395569, + "learning_rate": 7.878542440328865e-05, + "loss": 0.0326, + "step": 33950 + }, + { + "epoch": 2.2249885343641487, + "grad_norm": 0.42798224091529846, + "learning_rate": 7.877190315055031e-05, + "loss": 0.0664, + "step": 33960 + }, + { + "epoch": 2.2256437135556575, + "grad_norm": 0.7170315384864807, + "learning_rate": 7.875837875135991e-05, + "loss": 0.0476, + "step": 33970 + }, + { + "epoch": 2.2262988927471663, + "grad_norm": 0.6502120494842529, + "learning_rate": 7.874485120719646e-05, + "loss": 0.048, + "step": 33980 + }, + { + "epoch": 2.226954071938675, + "grad_norm": 0.5194541811943054, + "learning_rate": 7.873132051953928e-05, + "loss": 0.0416, + "step": 33990 + }, + { + "epoch": 2.227609251130184, + "grad_norm": 0.4274767339229584, + "learning_rate": 7.87177866898681e-05, + "loss": 0.0432, + "step": 34000 + }, + { + "epoch": 2.228264430321693, + "grad_norm": 0.7317407131195068, + "learning_rate": 7.870424971966294e-05, + "loss": 0.0503, + "step": 34010 + }, + { + "epoch": 2.228919609513202, + "grad_norm": 0.18372003734111786, + "learning_rate": 7.869070961040419e-05, + "loss": 0.0464, + "step": 34020 + }, + { + "epoch": 2.229574788704711, + "grad_norm": 0.33088618516921997, + "learning_rate": 7.867716636357257e-05, + "loss": 0.0545, + "step": 34030 + }, + { + "epoch": 2.2302299678962196, + "grad_norm": 0.3227626383304596, + "learning_rate": 7.866361998064915e-05, + "loss": 0.0428, + "step": 34040 + }, + { + "epoch": 2.2308851470877284, + "grad_norm": 0.5233379602432251, + "learning_rate": 7.865007046311534e-05, + "loss": 0.0448, + "step": 34050 + }, + { + "epoch": 2.2315403262792373, + "grad_norm": 0.30209118127822876, + "learning_rate": 7.86365178124529e-05, + "loss": 0.0298, + "step": 34060 + }, + { + "epoch": 2.232195505470746, + "grad_norm": 0.35280275344848633, + "learning_rate": 7.862296203014394e-05, + "loss": 0.0344, + "step": 34070 + }, + { + "epoch": 2.232850684662255, + "grad_norm": 0.3184216618537903, + "learning_rate": 7.860940311767088e-05, + "loss": 0.0292, + "step": 34080 + }, + { + "epoch": 2.233505863853764, + "grad_norm": 1.0584419965744019, + "learning_rate": 7.85958410765165e-05, + "loss": 0.0482, + "step": 34090 + }, + { + "epoch": 2.234161043045273, + "grad_norm": 0.31153857707977295, + "learning_rate": 7.858227590816394e-05, + "loss": 0.0289, + "step": 34100 + }, + { + "epoch": 2.234816222236782, + "grad_norm": 0.2286834865808487, + "learning_rate": 7.856870761409664e-05, + "loss": 0.0494, + "step": 34110 + }, + { + "epoch": 2.2354714014282906, + "grad_norm": 0.5709923505783081, + "learning_rate": 7.855513619579846e-05, + "loss": 0.0559, + "step": 34120 + }, + { + "epoch": 2.2361265806197994, + "grad_norm": 0.316517174243927, + "learning_rate": 7.85415616547535e-05, + "loss": 0.0297, + "step": 34130 + }, + { + "epoch": 2.2367817598113082, + "grad_norm": 0.38473862409591675, + "learning_rate": 7.852798399244627e-05, + "loss": 0.0297, + "step": 34140 + }, + { + "epoch": 2.2374369390028175, + "grad_norm": 0.71798175573349, + "learning_rate": 7.851440321036161e-05, + "loss": 0.0437, + "step": 34150 + }, + { + "epoch": 2.2380921181943263, + "grad_norm": 0.4884655475616455, + "learning_rate": 7.850081930998468e-05, + "loss": 0.0362, + "step": 34160 + }, + { + "epoch": 2.238747297385835, + "grad_norm": 0.2627866268157959, + "learning_rate": 7.8487232292801e-05, + "loss": 0.0249, + "step": 34170 + }, + { + "epoch": 2.239402476577344, + "grad_norm": 0.6401667594909668, + "learning_rate": 7.847364216029642e-05, + "loss": 0.0652, + "step": 34180 + }, + { + "epoch": 2.2400576557688527, + "grad_norm": 0.6308149695396423, + "learning_rate": 7.846004891395716e-05, + "loss": 0.0337, + "step": 34190 + }, + { + "epoch": 2.2407128349603616, + "grad_norm": 1.113707423210144, + "learning_rate": 7.844645255526972e-05, + "loss": 0.0607, + "step": 34200 + }, + { + "epoch": 2.2413680141518704, + "grad_norm": 0.14259959757328033, + "learning_rate": 7.843285308572101e-05, + "loss": 0.0429, + "step": 34210 + }, + { + "epoch": 2.242023193343379, + "grad_norm": 0.555375874042511, + "learning_rate": 7.841925050679823e-05, + "loss": 0.0461, + "step": 34220 + }, + { + "epoch": 2.2426783725348884, + "grad_norm": 0.25777778029441833, + "learning_rate": 7.840564481998895e-05, + "loss": 0.0502, + "step": 34230 + }, + { + "epoch": 2.2433335517263973, + "grad_norm": 1.2180055379867554, + "learning_rate": 7.839203602678105e-05, + "loss": 0.0469, + "step": 34240 + }, + { + "epoch": 2.243988730917906, + "grad_norm": 0.799246609210968, + "learning_rate": 7.837842412866279e-05, + "loss": 0.0999, + "step": 34250 + }, + { + "epoch": 2.244643910109415, + "grad_norm": 0.8744071125984192, + "learning_rate": 7.836480912712272e-05, + "loss": 0.0406, + "step": 34260 + }, + { + "epoch": 2.2452990893009237, + "grad_norm": 0.8204192519187927, + "learning_rate": 7.835119102364976e-05, + "loss": 0.0429, + "step": 34270 + }, + { + "epoch": 2.2459542684924325, + "grad_norm": 1.3456850051879883, + "learning_rate": 7.83375698197332e-05, + "loss": 0.0663, + "step": 34280 + }, + { + "epoch": 2.2466094476839418, + "grad_norm": 0.6817694902420044, + "learning_rate": 7.83239455168626e-05, + "loss": 0.0568, + "step": 34290 + }, + { + "epoch": 2.2472646268754506, + "grad_norm": 1.132073163986206, + "learning_rate": 7.83103181165279e-05, + "loss": 0.0345, + "step": 34300 + }, + { + "epoch": 2.2479198060669594, + "grad_norm": 0.4738680124282837, + "learning_rate": 7.829668762021937e-05, + "loss": 0.0415, + "step": 34310 + }, + { + "epoch": 2.248574985258468, + "grad_norm": 0.6535501480102539, + "learning_rate": 7.828305402942764e-05, + "loss": 0.062, + "step": 34320 + }, + { + "epoch": 2.249230164449977, + "grad_norm": 0.824967622756958, + "learning_rate": 7.826941734564363e-05, + "loss": 0.0316, + "step": 34330 + }, + { + "epoch": 2.249885343641486, + "grad_norm": 0.7078264355659485, + "learning_rate": 7.825577757035865e-05, + "loss": 0.0604, + "step": 34340 + }, + { + "epoch": 2.2505405228329947, + "grad_norm": 0.6532168984413147, + "learning_rate": 7.824213470506431e-05, + "loss": 0.0417, + "step": 34350 + }, + { + "epoch": 2.2511957020245035, + "grad_norm": 0.6972959041595459, + "learning_rate": 7.822848875125257e-05, + "loss": 0.029, + "step": 34360 + }, + { + "epoch": 2.2518508812160127, + "grad_norm": 0.22277112305164337, + "learning_rate": 7.821483971041576e-05, + "loss": 0.0382, + "step": 34370 + }, + { + "epoch": 2.2525060604075215, + "grad_norm": 0.22663764655590057, + "learning_rate": 7.820118758404649e-05, + "loss": 0.0223, + "step": 34380 + }, + { + "epoch": 2.2531612395990304, + "grad_norm": 0.2819330394268036, + "learning_rate": 7.818753237363776e-05, + "loss": 0.016, + "step": 34390 + }, + { + "epoch": 2.253816418790539, + "grad_norm": 0.33229461312294006, + "learning_rate": 7.817387408068286e-05, + "loss": 0.0417, + "step": 34400 + }, + { + "epoch": 2.254471597982048, + "grad_norm": 0.8851910829544067, + "learning_rate": 7.816021270667544e-05, + "loss": 0.0447, + "step": 34410 + }, + { + "epoch": 2.255126777173557, + "grad_norm": 0.21952928602695465, + "learning_rate": 7.81465482531095e-05, + "loss": 0.0345, + "step": 34420 + }, + { + "epoch": 2.255781956365066, + "grad_norm": 0.09932110458612442, + "learning_rate": 7.813288072147938e-05, + "loss": 0.0213, + "step": 34430 + }, + { + "epoch": 2.256437135556575, + "grad_norm": 0.9795455932617188, + "learning_rate": 7.811921011327972e-05, + "loss": 0.0435, + "step": 34440 + }, + { + "epoch": 2.2570923147480837, + "grad_norm": 0.12118110060691833, + "learning_rate": 7.810553643000549e-05, + "loss": 0.0334, + "step": 34450 + }, + { + "epoch": 2.2577474939395925, + "grad_norm": 0.26416802406311035, + "learning_rate": 7.809185967315206e-05, + "loss": 0.0403, + "step": 34460 + }, + { + "epoch": 2.2584026731311013, + "grad_norm": 0.6240950226783752, + "learning_rate": 7.80781798442151e-05, + "loss": 0.0476, + "step": 34470 + }, + { + "epoch": 2.25905785232261, + "grad_norm": 0.8647423386573792, + "learning_rate": 7.806449694469061e-05, + "loss": 0.0462, + "step": 34480 + }, + { + "epoch": 2.259713031514119, + "grad_norm": 0.3115162253379822, + "learning_rate": 7.805081097607492e-05, + "loss": 0.0327, + "step": 34490 + }, + { + "epoch": 2.2603682107056278, + "grad_norm": 0.11889001727104187, + "learning_rate": 7.803712193986474e-05, + "loss": 0.0281, + "step": 34500 + }, + { + "epoch": 2.261023389897137, + "grad_norm": 1.4714577198028564, + "learning_rate": 7.802342983755702e-05, + "loss": 0.0452, + "step": 34510 + }, + { + "epoch": 2.261678569088646, + "grad_norm": 0.6120941042900085, + "learning_rate": 7.800973467064918e-05, + "loss": 0.0335, + "step": 34520 + }, + { + "epoch": 2.2623337482801547, + "grad_norm": 0.9319802522659302, + "learning_rate": 7.799603644063884e-05, + "loss": 0.047, + "step": 34530 + }, + { + "epoch": 2.2629889274716635, + "grad_norm": 1.7136021852493286, + "learning_rate": 7.798233514902405e-05, + "loss": 0.0438, + "step": 34540 + }, + { + "epoch": 2.2636441066631723, + "grad_norm": 0.4114956855773926, + "learning_rate": 7.796863079730318e-05, + "loss": 0.0427, + "step": 34550 + }, + { + "epoch": 2.264299285854681, + "grad_norm": 0.9067822098731995, + "learning_rate": 7.795492338697488e-05, + "loss": 0.0592, + "step": 34560 + }, + { + "epoch": 2.2649544650461904, + "grad_norm": 1.135687232017517, + "learning_rate": 7.794121291953819e-05, + "loss": 0.0656, + "step": 34570 + }, + { + "epoch": 2.265609644237699, + "grad_norm": 0.2173110991716385, + "learning_rate": 7.792749939649246e-05, + "loss": 0.0422, + "step": 34580 + }, + { + "epoch": 2.266264823429208, + "grad_norm": 0.9982009530067444, + "learning_rate": 7.79137828193374e-05, + "loss": 0.0289, + "step": 34590 + }, + { + "epoch": 2.266920002620717, + "grad_norm": 0.7909042239189148, + "learning_rate": 7.790006318957301e-05, + "loss": 0.0959, + "step": 34600 + }, + { + "epoch": 2.2675751818122256, + "grad_norm": 0.6913414597511292, + "learning_rate": 7.788634050869965e-05, + "loss": 0.0216, + "step": 34610 + }, + { + "epoch": 2.2682303610037344, + "grad_norm": 0.22255130112171173, + "learning_rate": 7.787261477821803e-05, + "loss": 0.0435, + "step": 34620 + }, + { + "epoch": 2.2688855401952432, + "grad_norm": 0.8146640658378601, + "learning_rate": 7.785888599962916e-05, + "loss": 0.0455, + "step": 34630 + }, + { + "epoch": 2.269540719386752, + "grad_norm": 0.2813170254230499, + "learning_rate": 7.784515417443439e-05, + "loss": 0.0511, + "step": 34640 + }, + { + "epoch": 2.2701958985782613, + "grad_norm": 0.22992610931396484, + "learning_rate": 7.783141930413545e-05, + "loss": 0.0206, + "step": 34650 + }, + { + "epoch": 2.27085107776977, + "grad_norm": 0.15549501776695251, + "learning_rate": 7.78176813902343e-05, + "loss": 0.038, + "step": 34660 + }, + { + "epoch": 2.271506256961279, + "grad_norm": 0.31547877192497253, + "learning_rate": 7.780394043423336e-05, + "loss": 0.0312, + "step": 34670 + }, + { + "epoch": 2.2721614361527878, + "grad_norm": 0.660973072052002, + "learning_rate": 7.77901964376353e-05, + "loss": 0.0566, + "step": 34680 + }, + { + "epoch": 2.2728166153442966, + "grad_norm": 0.22564992308616638, + "learning_rate": 7.777644940194316e-05, + "loss": 0.0551, + "step": 34690 + }, + { + "epoch": 2.2734717945358054, + "grad_norm": 0.2898041605949402, + "learning_rate": 7.776269932866023e-05, + "loss": 0.0279, + "step": 34700 + }, + { + "epoch": 2.2741269737273146, + "grad_norm": 0.5237001180648804, + "learning_rate": 7.774894621929026e-05, + "loss": 0.0548, + "step": 34710 + }, + { + "epoch": 2.2747821529188235, + "grad_norm": 0.21277137100696564, + "learning_rate": 7.773519007533725e-05, + "loss": 0.0385, + "step": 34720 + }, + { + "epoch": 2.2754373321103323, + "grad_norm": 0.7137724757194519, + "learning_rate": 7.772143089830556e-05, + "loss": 0.0678, + "step": 34730 + }, + { + "epoch": 2.276092511301841, + "grad_norm": 0.9744530320167542, + "learning_rate": 7.770766868969985e-05, + "loss": 0.0394, + "step": 34740 + }, + { + "epoch": 2.27674769049335, + "grad_norm": 0.3658198118209839, + "learning_rate": 7.769390345102518e-05, + "loss": 0.0556, + "step": 34750 + }, + { + "epoch": 2.2774028696848587, + "grad_norm": 0.15878865122795105, + "learning_rate": 7.768013518378683e-05, + "loss": 0.0656, + "step": 34760 + }, + { + "epoch": 2.2780580488763675, + "grad_norm": 0.3464624285697937, + "learning_rate": 7.766636388949053e-05, + "loss": 0.0603, + "step": 34770 + }, + { + "epoch": 2.2787132280678764, + "grad_norm": 1.1741716861724854, + "learning_rate": 7.765258956964229e-05, + "loss": 0.0607, + "step": 34780 + }, + { + "epoch": 2.2793684072593856, + "grad_norm": 0.6219056248664856, + "learning_rate": 7.76388122257484e-05, + "loss": 0.0328, + "step": 34790 + }, + { + "epoch": 2.2800235864508944, + "grad_norm": 1.333649754524231, + "learning_rate": 7.762503185931558e-05, + "loss": 0.0314, + "step": 34800 + }, + { + "epoch": 2.2806787656424032, + "grad_norm": 0.7295699119567871, + "learning_rate": 7.76112484718508e-05, + "loss": 0.058, + "step": 34810 + }, + { + "epoch": 2.281333944833912, + "grad_norm": 0.1267637014389038, + "learning_rate": 7.75974620648614e-05, + "loss": 0.0499, + "step": 34820 + }, + { + "epoch": 2.281989124025421, + "grad_norm": 0.15008330345153809, + "learning_rate": 7.758367263985503e-05, + "loss": 0.0464, + "step": 34830 + }, + { + "epoch": 2.2826443032169297, + "grad_norm": 0.6104860901832581, + "learning_rate": 7.75698801983397e-05, + "loss": 0.0503, + "step": 34840 + }, + { + "epoch": 2.283299482408439, + "grad_norm": 0.923233151435852, + "learning_rate": 7.755608474182372e-05, + "loss": 0.0571, + "step": 34850 + }, + { + "epoch": 2.2839546615999478, + "grad_norm": 0.2674849033355713, + "learning_rate": 7.754228627181574e-05, + "loss": 0.0231, + "step": 34860 + }, + { + "epoch": 2.2846098407914566, + "grad_norm": 1.422049880027771, + "learning_rate": 7.752848478982476e-05, + "loss": 0.0774, + "step": 34870 + }, + { + "epoch": 2.2852650199829654, + "grad_norm": 1.0874768495559692, + "learning_rate": 7.751468029736006e-05, + "loss": 0.0773, + "step": 34880 + }, + { + "epoch": 2.285920199174474, + "grad_norm": 0.2708365321159363, + "learning_rate": 7.750087279593129e-05, + "loss": 0.0305, + "step": 34890 + }, + { + "epoch": 2.286575378365983, + "grad_norm": 0.8452282547950745, + "learning_rate": 7.748706228704843e-05, + "loss": 0.0503, + "step": 34900 + }, + { + "epoch": 2.287230557557492, + "grad_norm": 0.6809112429618835, + "learning_rate": 7.747324877222176e-05, + "loss": 0.0379, + "step": 34910 + }, + { + "epoch": 2.2878857367490006, + "grad_norm": 0.119744211435318, + "learning_rate": 7.745943225296188e-05, + "loss": 0.0232, + "step": 34920 + }, + { + "epoch": 2.28854091594051, + "grad_norm": 0.2791428565979004, + "learning_rate": 7.744561273077981e-05, + "loss": 0.0238, + "step": 34930 + }, + { + "epoch": 2.2891960951320187, + "grad_norm": 0.6926454305648804, + "learning_rate": 7.743179020718678e-05, + "loss": 0.0487, + "step": 34940 + }, + { + "epoch": 2.2898512743235275, + "grad_norm": 0.3323270380496979, + "learning_rate": 7.741796468369443e-05, + "loss": 0.0368, + "step": 34950 + }, + { + "epoch": 2.2905064535150363, + "grad_norm": 0.8575954437255859, + "learning_rate": 7.740413616181466e-05, + "loss": 0.0309, + "step": 34960 + }, + { + "epoch": 2.291161632706545, + "grad_norm": 0.25953409075737, + "learning_rate": 7.739030464305978e-05, + "loss": 0.0357, + "step": 34970 + }, + { + "epoch": 2.291816811898054, + "grad_norm": 0.21162794530391693, + "learning_rate": 7.737647012894235e-05, + "loss": 0.0278, + "step": 34980 + }, + { + "epoch": 2.2924719910895632, + "grad_norm": 1.613997220993042, + "learning_rate": 7.736263262097532e-05, + "loss": 0.0703, + "step": 34990 + }, + { + "epoch": 2.293127170281072, + "grad_norm": 0.1190279945731163, + "learning_rate": 7.734879212067192e-05, + "loss": 0.0411, + "step": 35000 + }, + { + "epoch": 2.293782349472581, + "grad_norm": 0.21847370266914368, + "learning_rate": 7.733494862954573e-05, + "loss": 0.0212, + "step": 35010 + }, + { + "epoch": 2.2944375286640897, + "grad_norm": 0.38193178176879883, + "learning_rate": 7.732110214911066e-05, + "loss": 0.0546, + "step": 35020 + }, + { + "epoch": 2.2950927078555985, + "grad_norm": 0.6726595163345337, + "learning_rate": 7.730725268088094e-05, + "loss": 0.0454, + "step": 35030 + }, + { + "epoch": 2.2957478870471073, + "grad_norm": 0.46231672167778015, + "learning_rate": 7.729340022637112e-05, + "loss": 0.0392, + "step": 35040 + }, + { + "epoch": 2.296403066238616, + "grad_norm": 0.9099710583686829, + "learning_rate": 7.727954478709607e-05, + "loss": 0.0427, + "step": 35050 + }, + { + "epoch": 2.297058245430125, + "grad_norm": 1.2968120574951172, + "learning_rate": 7.726568636457103e-05, + "loss": 0.0409, + "step": 35060 + }, + { + "epoch": 2.297713424621634, + "grad_norm": 0.7709367275238037, + "learning_rate": 7.725182496031153e-05, + "loss": 0.0627, + "step": 35070 + }, + { + "epoch": 2.298368603813143, + "grad_norm": 0.6780312657356262, + "learning_rate": 7.723796057583341e-05, + "loss": 0.0335, + "step": 35080 + }, + { + "epoch": 2.299023783004652, + "grad_norm": 0.5357990860939026, + "learning_rate": 7.722409321265287e-05, + "loss": 0.0303, + "step": 35090 + }, + { + "epoch": 2.2996789621961606, + "grad_norm": 0.49110323190689087, + "learning_rate": 7.721022287228645e-05, + "loss": 0.0395, + "step": 35100 + }, + { + "epoch": 2.3003341413876695, + "grad_norm": 0.749785304069519, + "learning_rate": 7.719634955625094e-05, + "loss": 0.0473, + "step": 35110 + }, + { + "epoch": 2.3009893205791783, + "grad_norm": 0.43581750988960266, + "learning_rate": 7.718247326606355e-05, + "loss": 0.0727, + "step": 35120 + }, + { + "epoch": 2.3016444997706875, + "grad_norm": 0.34085240960121155, + "learning_rate": 7.716859400324175e-05, + "loss": 0.0501, + "step": 35130 + }, + { + "epoch": 2.3022996789621963, + "grad_norm": 0.10455424338579178, + "learning_rate": 7.715471176930335e-05, + "loss": 0.0396, + "step": 35140 + }, + { + "epoch": 2.302954858153705, + "grad_norm": 0.5969371199607849, + "learning_rate": 7.714082656576651e-05, + "loss": 0.0584, + "step": 35150 + }, + { + "epoch": 2.303610037345214, + "grad_norm": 0.9016196727752686, + "learning_rate": 7.712693839414968e-05, + "loss": 0.0337, + "step": 35160 + }, + { + "epoch": 2.304265216536723, + "grad_norm": 0.9775313138961792, + "learning_rate": 7.711304725597164e-05, + "loss": 0.0203, + "step": 35170 + }, + { + "epoch": 2.3049203957282316, + "grad_norm": 0.5685505270957947, + "learning_rate": 7.709915315275151e-05, + "loss": 0.0329, + "step": 35180 + }, + { + "epoch": 2.3055755749197404, + "grad_norm": 0.28398820757865906, + "learning_rate": 7.708525608600876e-05, + "loss": 0.0246, + "step": 35190 + }, + { + "epoch": 2.3062307541112492, + "grad_norm": 0.13505150377750397, + "learning_rate": 7.707135605726311e-05, + "loss": 0.0341, + "step": 35200 + }, + { + "epoch": 2.3068859333027585, + "grad_norm": 0.31995293498039246, + "learning_rate": 7.705745306803466e-05, + "loss": 0.0327, + "step": 35210 + }, + { + "epoch": 2.3075411124942673, + "grad_norm": 0.6165083050727844, + "learning_rate": 7.704354711984383e-05, + "loss": 0.0364, + "step": 35220 + }, + { + "epoch": 2.308196291685776, + "grad_norm": 0.4731315076351166, + "learning_rate": 7.702963821421133e-05, + "loss": 0.0272, + "step": 35230 + }, + { + "epoch": 2.308851470877285, + "grad_norm": 0.23323650658130646, + "learning_rate": 7.701572635265826e-05, + "loss": 0.0176, + "step": 35240 + }, + { + "epoch": 2.3095066500687937, + "grad_norm": 0.419664204120636, + "learning_rate": 7.700181153670596e-05, + "loss": 0.0286, + "step": 35250 + }, + { + "epoch": 2.3101618292603026, + "grad_norm": 0.17964018881320953, + "learning_rate": 7.698789376787616e-05, + "loss": 0.0226, + "step": 35260 + }, + { + "epoch": 2.310817008451812, + "grad_norm": 0.5358154773712158, + "learning_rate": 7.697397304769085e-05, + "loss": 0.0392, + "step": 35270 + }, + { + "epoch": 2.3114721876433206, + "grad_norm": 0.8690844178199768, + "learning_rate": 7.696004937767241e-05, + "loss": 0.0456, + "step": 35280 + }, + { + "epoch": 2.3121273668348294, + "grad_norm": 1.0554530620574951, + "learning_rate": 7.694612275934352e-05, + "loss": 0.0471, + "step": 35290 + }, + { + "epoch": 2.3127825460263383, + "grad_norm": 0.15831124782562256, + "learning_rate": 7.693219319422714e-05, + "loss": 0.0365, + "step": 35300 + }, + { + "epoch": 2.313437725217847, + "grad_norm": 0.35542845726013184, + "learning_rate": 7.69182606838466e-05, + "loss": 0.0497, + "step": 35310 + }, + { + "epoch": 2.314092904409356, + "grad_norm": 0.39789795875549316, + "learning_rate": 7.690432522972558e-05, + "loss": 0.0322, + "step": 35320 + }, + { + "epoch": 2.3147480836008647, + "grad_norm": 1.8998972177505493, + "learning_rate": 7.689038683338796e-05, + "loss": 0.0406, + "step": 35330 + }, + { + "epoch": 2.3154032627923735, + "grad_norm": 0.5144146680831909, + "learning_rate": 7.687644549635808e-05, + "loss": 0.0529, + "step": 35340 + }, + { + "epoch": 2.3160584419838828, + "grad_norm": 0.11021237075328827, + "learning_rate": 7.686250122016053e-05, + "loss": 0.0245, + "step": 35350 + }, + { + "epoch": 2.3167136211753916, + "grad_norm": 1.115385890007019, + "learning_rate": 7.684855400632022e-05, + "loss": 0.0672, + "step": 35360 + }, + { + "epoch": 2.3173688003669004, + "grad_norm": 0.42163312435150146, + "learning_rate": 7.683460385636243e-05, + "loss": 0.026, + "step": 35370 + }, + { + "epoch": 2.318023979558409, + "grad_norm": 0.30344903469085693, + "learning_rate": 7.68206507718127e-05, + "loss": 0.0245, + "step": 35380 + }, + { + "epoch": 2.318679158749918, + "grad_norm": 0.3990662097930908, + "learning_rate": 7.680669475419693e-05, + "loss": 0.0662, + "step": 35390 + }, + { + "epoch": 2.319334337941427, + "grad_norm": 0.18451647460460663, + "learning_rate": 7.679273580504132e-05, + "loss": 0.0492, + "step": 35400 + }, + { + "epoch": 2.319989517132936, + "grad_norm": 0.7878129482269287, + "learning_rate": 7.67787739258724e-05, + "loss": 0.0189, + "step": 35410 + }, + { + "epoch": 2.320644696324445, + "grad_norm": 0.18366111814975739, + "learning_rate": 7.676480911821705e-05, + "loss": 0.0375, + "step": 35420 + }, + { + "epoch": 2.3212998755159537, + "grad_norm": 0.4960240125656128, + "learning_rate": 7.675084138360238e-05, + "loss": 0.0302, + "step": 35430 + }, + { + "epoch": 2.3219550547074626, + "grad_norm": 0.6110373139381409, + "learning_rate": 7.673687072355592e-05, + "loss": 0.0413, + "step": 35440 + }, + { + "epoch": 2.3226102338989714, + "grad_norm": 0.6232976913452148, + "learning_rate": 7.67228971396055e-05, + "loss": 0.0446, + "step": 35450 + }, + { + "epoch": 2.32326541309048, + "grad_norm": 0.5710333585739136, + "learning_rate": 7.670892063327922e-05, + "loss": 0.0354, + "step": 35460 + }, + { + "epoch": 2.323920592281989, + "grad_norm": 1.964115023612976, + "learning_rate": 7.669494120610552e-05, + "loss": 0.056, + "step": 35470 + }, + { + "epoch": 2.324575771473498, + "grad_norm": 0.4938349723815918, + "learning_rate": 7.66809588596132e-05, + "loss": 0.0425, + "step": 35480 + }, + { + "epoch": 2.325230950665007, + "grad_norm": 0.16024723649024963, + "learning_rate": 7.666697359533132e-05, + "loss": 0.0584, + "step": 35490 + }, + { + "epoch": 2.325886129856516, + "grad_norm": 0.4246369004249573, + "learning_rate": 7.665298541478932e-05, + "loss": 0.0597, + "step": 35500 + }, + { + "epoch": 2.3265413090480247, + "grad_norm": 0.216362863779068, + "learning_rate": 7.66389943195169e-05, + "loss": 0.0405, + "step": 35510 + }, + { + "epoch": 2.3271964882395335, + "grad_norm": 0.22048291563987732, + "learning_rate": 7.66250003110441e-05, + "loss": 0.0316, + "step": 35520 + }, + { + "epoch": 2.3278516674310423, + "grad_norm": 0.39827266335487366, + "learning_rate": 7.66110033909013e-05, + "loss": 0.05, + "step": 35530 + }, + { + "epoch": 2.328506846622551, + "grad_norm": 0.7405666708946228, + "learning_rate": 7.659700356061918e-05, + "loss": 0.053, + "step": 35540 + }, + { + "epoch": 2.32916202581406, + "grad_norm": 0.3718798756599426, + "learning_rate": 7.658300082172875e-05, + "loss": 0.0275, + "step": 35550 + }, + { + "epoch": 2.329817205005569, + "grad_norm": 0.4027998745441437, + "learning_rate": 7.65689951757613e-05, + "loss": 0.0193, + "step": 35560 + }, + { + "epoch": 2.330472384197078, + "grad_norm": 0.6014452576637268, + "learning_rate": 7.655498662424849e-05, + "loss": 0.0479, + "step": 35570 + }, + { + "epoch": 2.331127563388587, + "grad_norm": 0.602789044380188, + "learning_rate": 7.654097516872227e-05, + "loss": 0.0416, + "step": 35580 + }, + { + "epoch": 2.3317827425800957, + "grad_norm": 1.089496374130249, + "learning_rate": 7.65269608107149e-05, + "loss": 0.0459, + "step": 35590 + }, + { + "epoch": 2.3324379217716045, + "grad_norm": 0.5353690385818481, + "learning_rate": 7.6512943551759e-05, + "loss": 0.046, + "step": 35600 + }, + { + "epoch": 2.3330931009631133, + "grad_norm": 0.66877281665802, + "learning_rate": 7.649892339338743e-05, + "loss": 0.037, + "step": 35610 + }, + { + "epoch": 2.333748280154622, + "grad_norm": 0.7973138093948364, + "learning_rate": 7.648490033713344e-05, + "loss": 0.0371, + "step": 35620 + }, + { + "epoch": 2.3344034593461314, + "grad_norm": 0.4494097828865051, + "learning_rate": 7.647087438453058e-05, + "loss": 0.045, + "step": 35630 + }, + { + "epoch": 2.33505863853764, + "grad_norm": 0.45503437519073486, + "learning_rate": 7.64568455371127e-05, + "loss": 0.0371, + "step": 35640 + }, + { + "epoch": 2.335713817729149, + "grad_norm": 1.3507936000823975, + "learning_rate": 7.644281379641396e-05, + "loss": 0.0682, + "step": 35650 + }, + { + "epoch": 2.336368996920658, + "grad_norm": 0.7049915194511414, + "learning_rate": 7.642877916396887e-05, + "loss": 0.037, + "step": 35660 + }, + { + "epoch": 2.3370241761121666, + "grad_norm": 0.8600563406944275, + "learning_rate": 7.641474164131221e-05, + "loss": 0.0246, + "step": 35670 + }, + { + "epoch": 2.3376793553036754, + "grad_norm": 0.24918682873249054, + "learning_rate": 7.640070122997913e-05, + "loss": 0.0761, + "step": 35680 + }, + { + "epoch": 2.3383345344951842, + "grad_norm": 1.2760623693466187, + "learning_rate": 7.638665793150505e-05, + "loss": 0.0658, + "step": 35690 + }, + { + "epoch": 2.3389897136866935, + "grad_norm": 0.23980817198753357, + "learning_rate": 7.637261174742574e-05, + "loss": 0.0297, + "step": 35700 + }, + { + "epoch": 2.3396448928782023, + "grad_norm": 0.12437249720096588, + "learning_rate": 7.635856267927725e-05, + "loss": 0.031, + "step": 35710 + }, + { + "epoch": 2.340300072069711, + "grad_norm": 0.6515663266181946, + "learning_rate": 7.634451072859597e-05, + "loss": 0.0353, + "step": 35720 + }, + { + "epoch": 2.34095525126122, + "grad_norm": 0.19215181469917297, + "learning_rate": 7.633045589691863e-05, + "loss": 0.0513, + "step": 35730 + }, + { + "epoch": 2.3416104304527288, + "grad_norm": 0.3615119159221649, + "learning_rate": 7.63163981857822e-05, + "loss": 0.0427, + "step": 35740 + }, + { + "epoch": 2.3422656096442376, + "grad_norm": 1.9153635501861572, + "learning_rate": 7.630233759672403e-05, + "loss": 0.049, + "step": 35750 + }, + { + "epoch": 2.3429207888357464, + "grad_norm": 0.2979409694671631, + "learning_rate": 7.628827413128177e-05, + "loss": 0.0352, + "step": 35760 + }, + { + "epoch": 2.343575968027255, + "grad_norm": 0.8801999688148499, + "learning_rate": 7.627420779099338e-05, + "loss": 0.0628, + "step": 35770 + }, + { + "epoch": 2.3442311472187645, + "grad_norm": 0.1710706651210785, + "learning_rate": 7.626013857739711e-05, + "loss": 0.0511, + "step": 35780 + }, + { + "epoch": 2.3448863264102733, + "grad_norm": 0.6338013410568237, + "learning_rate": 7.624606649203158e-05, + "loss": 0.0646, + "step": 35790 + }, + { + "epoch": 2.345541505601782, + "grad_norm": 0.19688694179058075, + "learning_rate": 7.623199153643569e-05, + "loss": 0.0371, + "step": 35800 + }, + { + "epoch": 2.346196684793291, + "grad_norm": 0.6146577596664429, + "learning_rate": 7.621791371214863e-05, + "loss": 0.038, + "step": 35810 + }, + { + "epoch": 2.3468518639847997, + "grad_norm": 0.2138190120458603, + "learning_rate": 7.620383302070995e-05, + "loss": 0.0341, + "step": 35820 + }, + { + "epoch": 2.3475070431763085, + "grad_norm": 0.7401246428489685, + "learning_rate": 7.61897494636595e-05, + "loss": 0.0405, + "step": 35830 + }, + { + "epoch": 2.348162222367818, + "grad_norm": 0.17639324069023132, + "learning_rate": 7.617566304253739e-05, + "loss": 0.0352, + "step": 35840 + }, + { + "epoch": 2.3488174015593266, + "grad_norm": 0.18519841134548187, + "learning_rate": 7.616157375888416e-05, + "loss": 0.0322, + "step": 35850 + }, + { + "epoch": 2.3494725807508354, + "grad_norm": 0.5046973824501038, + "learning_rate": 7.614748161424053e-05, + "loss": 0.058, + "step": 35860 + }, + { + "epoch": 2.3501277599423442, + "grad_norm": 0.9135494828224182, + "learning_rate": 7.613338661014763e-05, + "loss": 0.0419, + "step": 35870 + }, + { + "epoch": 2.350782939133853, + "grad_norm": 0.40333691239356995, + "learning_rate": 7.611928874814686e-05, + "loss": 0.0435, + "step": 35880 + }, + { + "epoch": 2.351438118325362, + "grad_norm": 0.15514320135116577, + "learning_rate": 7.610518802977996e-05, + "loss": 0.05, + "step": 35890 + }, + { + "epoch": 2.3520932975168707, + "grad_norm": 0.7568789720535278, + "learning_rate": 7.609108445658893e-05, + "loss": 0.0535, + "step": 35900 + }, + { + "epoch": 2.3527484767083795, + "grad_norm": 0.3434429168701172, + "learning_rate": 7.607697803011612e-05, + "loss": 0.0321, + "step": 35910 + }, + { + "epoch": 2.3534036558998888, + "grad_norm": 0.424942284822464, + "learning_rate": 7.606286875190421e-05, + "loss": 0.0327, + "step": 35920 + }, + { + "epoch": 2.3540588350913976, + "grad_norm": 0.9281541705131531, + "learning_rate": 7.604875662349617e-05, + "loss": 0.0889, + "step": 35930 + }, + { + "epoch": 2.3547140142829064, + "grad_norm": 0.8310564160346985, + "learning_rate": 7.603464164643525e-05, + "loss": 0.038, + "step": 35940 + }, + { + "epoch": 2.355369193474415, + "grad_norm": 0.29481077194213867, + "learning_rate": 7.602052382226507e-05, + "loss": 0.0329, + "step": 35950 + }, + { + "epoch": 2.356024372665924, + "grad_norm": 0.20493371784687042, + "learning_rate": 7.600640315252954e-05, + "loss": 0.0745, + "step": 35960 + }, + { + "epoch": 2.356679551857433, + "grad_norm": 0.47461727261543274, + "learning_rate": 7.599227963877284e-05, + "loss": 0.0393, + "step": 35970 + }, + { + "epoch": 2.357334731048942, + "grad_norm": 0.3981775641441345, + "learning_rate": 7.597815328253954e-05, + "loss": 0.0614, + "step": 35980 + }, + { + "epoch": 2.357989910240451, + "grad_norm": 0.33058181405067444, + "learning_rate": 7.596402408537443e-05, + "loss": 0.0435, + "step": 35990 + }, + { + "epoch": 2.3586450894319597, + "grad_norm": 0.28197064995765686, + "learning_rate": 7.59498920488227e-05, + "loss": 0.0258, + "step": 36000 + }, + { + "epoch": 2.3593002686234685, + "grad_norm": 0.5187371969223022, + "learning_rate": 7.593575717442979e-05, + "loss": 0.0242, + "step": 36010 + }, + { + "epoch": 2.3599554478149773, + "grad_norm": 0.3822322189807892, + "learning_rate": 7.592161946374147e-05, + "loss": 0.0275, + "step": 36020 + }, + { + "epoch": 2.360610627006486, + "grad_norm": 0.7407003045082092, + "learning_rate": 7.590747891830381e-05, + "loss": 0.0452, + "step": 36030 + }, + { + "epoch": 2.361265806197995, + "grad_norm": 0.31278175115585327, + "learning_rate": 7.58933355396632e-05, + "loss": 0.0326, + "step": 36040 + }, + { + "epoch": 2.361920985389504, + "grad_norm": 0.2671273946762085, + "learning_rate": 7.587918932936636e-05, + "loss": 0.0497, + "step": 36050 + }, + { + "epoch": 2.362576164581013, + "grad_norm": 0.25656774640083313, + "learning_rate": 7.586504028896028e-05, + "loss": 0.029, + "step": 36060 + }, + { + "epoch": 2.363231343772522, + "grad_norm": 0.5824552774429321, + "learning_rate": 7.585088841999228e-05, + "loss": 0.0387, + "step": 36070 + }, + { + "epoch": 2.3638865229640307, + "grad_norm": 0.5428048968315125, + "learning_rate": 7.583673372400999e-05, + "loss": 0.0757, + "step": 36080 + }, + { + "epoch": 2.3645417021555395, + "grad_norm": 0.513920783996582, + "learning_rate": 7.582257620256134e-05, + "loss": 0.0985, + "step": 36090 + }, + { + "epoch": 2.3651968813470483, + "grad_norm": 0.684299886226654, + "learning_rate": 7.580841585719458e-05, + "loss": 0.0253, + "step": 36100 + }, + { + "epoch": 2.365852060538557, + "grad_norm": 0.38086840510368347, + "learning_rate": 7.579425268945825e-05, + "loss": 0.0508, + "step": 36110 + }, + { + "epoch": 2.3665072397300664, + "grad_norm": 0.3258211016654968, + "learning_rate": 7.578008670090127e-05, + "loss": 0.0236, + "step": 36120 + }, + { + "epoch": 2.367162418921575, + "grad_norm": 0.27307090163230896, + "learning_rate": 7.576591789307272e-05, + "loss": 0.0604, + "step": 36130 + }, + { + "epoch": 2.367817598113084, + "grad_norm": 0.1938316524028778, + "learning_rate": 7.575174626752216e-05, + "loss": 0.0415, + "step": 36140 + }, + { + "epoch": 2.368472777304593, + "grad_norm": 0.5472580790519714, + "learning_rate": 7.573757182579934e-05, + "loss": 0.0536, + "step": 36150 + }, + { + "epoch": 2.3691279564961016, + "grad_norm": 0.4524708390235901, + "learning_rate": 7.572339456945435e-05, + "loss": 0.0469, + "step": 36160 + }, + { + "epoch": 2.3697831356876105, + "grad_norm": 0.33024367690086365, + "learning_rate": 7.570921450003762e-05, + "loss": 0.0433, + "step": 36170 + }, + { + "epoch": 2.3704383148791193, + "grad_norm": 0.24188052117824554, + "learning_rate": 7.569503161909983e-05, + "loss": 0.0245, + "step": 36180 + }, + { + "epoch": 2.371093494070628, + "grad_norm": 0.45281746983528137, + "learning_rate": 7.568084592819202e-05, + "loss": 0.03, + "step": 36190 + }, + { + "epoch": 2.3717486732621373, + "grad_norm": 0.253168523311615, + "learning_rate": 7.566665742886551e-05, + "loss": 0.0326, + "step": 36200 + }, + { + "epoch": 2.372403852453646, + "grad_norm": 0.137232705950737, + "learning_rate": 7.565246612267194e-05, + "loss": 0.0579, + "step": 36210 + }, + { + "epoch": 2.373059031645155, + "grad_norm": 0.38880470395088196, + "learning_rate": 7.563827201116325e-05, + "loss": 0.044, + "step": 36220 + }, + { + "epoch": 2.373714210836664, + "grad_norm": 0.2600879967212677, + "learning_rate": 7.562407509589164e-05, + "loss": 0.0374, + "step": 36230 + }, + { + "epoch": 2.3743693900281726, + "grad_norm": 0.35894837975502014, + "learning_rate": 7.560987537840973e-05, + "loss": 0.0244, + "step": 36240 + }, + { + "epoch": 2.3750245692196814, + "grad_norm": 0.22342024743556976, + "learning_rate": 7.559567286027036e-05, + "loss": 0.0164, + "step": 36250 + }, + { + "epoch": 2.3756797484111907, + "grad_norm": 0.13678961992263794, + "learning_rate": 7.558146754302668e-05, + "loss": 0.0329, + "step": 36260 + }, + { + "epoch": 2.3763349276026995, + "grad_norm": 0.4955965578556061, + "learning_rate": 7.556725942823217e-05, + "loss": 0.0313, + "step": 36270 + }, + { + "epoch": 2.3769901067942083, + "grad_norm": 0.9587362408638, + "learning_rate": 7.555304851744061e-05, + "loss": 0.0317, + "step": 36280 + }, + { + "epoch": 2.377645285985717, + "grad_norm": 0.4325968623161316, + "learning_rate": 7.553883481220608e-05, + "loss": 0.0386, + "step": 36290 + }, + { + "epoch": 2.378300465177226, + "grad_norm": 0.6347614526748657, + "learning_rate": 7.552461831408298e-05, + "loss": 0.0281, + "step": 36300 + }, + { + "epoch": 2.3789556443687347, + "grad_norm": 0.8284971117973328, + "learning_rate": 7.551039902462599e-05, + "loss": 0.0454, + "step": 36310 + }, + { + "epoch": 2.3796108235602436, + "grad_norm": 0.8616347312927246, + "learning_rate": 7.549617694539011e-05, + "loss": 0.0326, + "step": 36320 + }, + { + "epoch": 2.3802660027517524, + "grad_norm": 0.09160186350345612, + "learning_rate": 7.548195207793066e-05, + "loss": 0.0632, + "step": 36330 + }, + { + "epoch": 2.3809211819432616, + "grad_norm": 0.6120185852050781, + "learning_rate": 7.546772442380323e-05, + "loss": 0.0261, + "step": 36340 + }, + { + "epoch": 2.3815763611347704, + "grad_norm": 0.3800303339958191, + "learning_rate": 7.545349398456376e-05, + "loss": 0.0535, + "step": 36350 + }, + { + "epoch": 2.3822315403262793, + "grad_norm": 0.9368495941162109, + "learning_rate": 7.543926076176845e-05, + "loss": 0.0315, + "step": 36360 + }, + { + "epoch": 2.382886719517788, + "grad_norm": 0.20774756371974945, + "learning_rate": 7.542502475697385e-05, + "loss": 0.0294, + "step": 36370 + }, + { + "epoch": 2.383541898709297, + "grad_norm": 0.6204390525817871, + "learning_rate": 7.541078597173675e-05, + "loss": 0.041, + "step": 36380 + }, + { + "epoch": 2.3841970779008057, + "grad_norm": 0.09030430763959885, + "learning_rate": 7.539654440761431e-05, + "loss": 0.0428, + "step": 36390 + }, + { + "epoch": 2.384852257092315, + "grad_norm": 0.36422044038772583, + "learning_rate": 7.538230006616395e-05, + "loss": 0.0497, + "step": 36400 + }, + { + "epoch": 2.385507436283824, + "grad_norm": 0.7696018218994141, + "learning_rate": 7.536805294894342e-05, + "loss": 0.0594, + "step": 36410 + }, + { + "epoch": 2.3861626154753326, + "grad_norm": 0.49521392583847046, + "learning_rate": 7.535380305751076e-05, + "loss": 0.0421, + "step": 36420 + }, + { + "epoch": 2.3868177946668414, + "grad_norm": 0.6070041656494141, + "learning_rate": 7.533955039342431e-05, + "loss": 0.0534, + "step": 36430 + }, + { + "epoch": 2.3874729738583502, + "grad_norm": 0.38203519582748413, + "learning_rate": 7.532529495824274e-05, + "loss": 0.0426, + "step": 36440 + }, + { + "epoch": 2.388128153049859, + "grad_norm": 0.36994171142578125, + "learning_rate": 7.531103675352497e-05, + "loss": 0.0355, + "step": 36450 + }, + { + "epoch": 2.388783332241368, + "grad_norm": 0.2837447226047516, + "learning_rate": 7.529677578083029e-05, + "loss": 0.0408, + "step": 36460 + }, + { + "epoch": 2.3894385114328767, + "grad_norm": 0.2717573046684265, + "learning_rate": 7.528251204171823e-05, + "loss": 0.0377, + "step": 36470 + }, + { + "epoch": 2.390093690624386, + "grad_norm": 0.6593257188796997, + "learning_rate": 7.526824553774867e-05, + "loss": 0.0408, + "step": 36480 + }, + { + "epoch": 2.3907488698158947, + "grad_norm": 0.7044612169265747, + "learning_rate": 7.525397627048178e-05, + "loss": 0.0752, + "step": 36490 + }, + { + "epoch": 2.3914040490074036, + "grad_norm": 0.17432036995887756, + "learning_rate": 7.523970424147802e-05, + "loss": 0.0287, + "step": 36500 + }, + { + "epoch": 2.3920592281989124, + "grad_norm": 0.7001562118530273, + "learning_rate": 7.522542945229813e-05, + "loss": 0.0297, + "step": 36510 + }, + { + "epoch": 2.392714407390421, + "grad_norm": 0.8529374599456787, + "learning_rate": 7.521115190450321e-05, + "loss": 0.04, + "step": 36520 + }, + { + "epoch": 2.39336958658193, + "grad_norm": 0.16559042036533356, + "learning_rate": 7.519687159965462e-05, + "loss": 0.0262, + "step": 36530 + }, + { + "epoch": 2.3940247657734393, + "grad_norm": 0.2480585277080536, + "learning_rate": 7.518258853931403e-05, + "loss": 0.0365, + "step": 36540 + }, + { + "epoch": 2.394679944964948, + "grad_norm": 0.8741594552993774, + "learning_rate": 7.516830272504342e-05, + "loss": 0.0359, + "step": 36550 + }, + { + "epoch": 2.395335124156457, + "grad_norm": 0.3426598012447357, + "learning_rate": 7.515401415840509e-05, + "loss": 0.0282, + "step": 36560 + }, + { + "epoch": 2.3959903033479657, + "grad_norm": 0.08944075554609299, + "learning_rate": 7.513972284096155e-05, + "loss": 0.0476, + "step": 36570 + }, + { + "epoch": 2.3966454825394745, + "grad_norm": 0.9214984774589539, + "learning_rate": 7.512542877427574e-05, + "loss": 0.0763, + "step": 36580 + }, + { + "epoch": 2.3973006617309833, + "grad_norm": 0.3921973705291748, + "learning_rate": 7.51111319599108e-05, + "loss": 0.0352, + "step": 36590 + }, + { + "epoch": 2.397955840922492, + "grad_norm": 0.34583762288093567, + "learning_rate": 7.509683239943024e-05, + "loss": 0.0384, + "step": 36600 + }, + { + "epoch": 2.398611020114001, + "grad_norm": 0.3506442904472351, + "learning_rate": 7.50825300943978e-05, + "loss": 0.0291, + "step": 36610 + }, + { + "epoch": 2.39926619930551, + "grad_norm": 1.1719391345977783, + "learning_rate": 7.50682250463776e-05, + "loss": 0.0294, + "step": 36620 + }, + { + "epoch": 2.399921378497019, + "grad_norm": 0.9736881256103516, + "learning_rate": 7.5053917256934e-05, + "loss": 0.0336, + "step": 36630 + }, + { + "epoch": 2.400576557688528, + "grad_norm": 0.2993719279766083, + "learning_rate": 7.503960672763166e-05, + "loss": 0.0227, + "step": 36640 + }, + { + "epoch": 2.4012317368800367, + "grad_norm": 0.33977431058883667, + "learning_rate": 7.502529346003559e-05, + "loss": 0.0304, + "step": 36650 + }, + { + "epoch": 2.4018869160715455, + "grad_norm": 0.5261876583099365, + "learning_rate": 7.501097745571106e-05, + "loss": 0.041, + "step": 36660 + }, + { + "epoch": 2.4025420952630543, + "grad_norm": 0.38502374291419983, + "learning_rate": 7.499665871622361e-05, + "loss": 0.0629, + "step": 36670 + }, + { + "epoch": 2.4031972744545635, + "grad_norm": 0.5576714873313904, + "learning_rate": 7.498233724313917e-05, + "loss": 0.031, + "step": 36680 + }, + { + "epoch": 2.4038524536460724, + "grad_norm": 0.38703691959381104, + "learning_rate": 7.496801303802387e-05, + "loss": 0.0385, + "step": 36690 + }, + { + "epoch": 2.404507632837581, + "grad_norm": 0.21179485321044922, + "learning_rate": 7.495368610244423e-05, + "loss": 0.0291, + "step": 36700 + }, + { + "epoch": 2.40516281202909, + "grad_norm": 0.6073002815246582, + "learning_rate": 7.493935643796697e-05, + "loss": 0.0346, + "step": 36710 + }, + { + "epoch": 2.405817991220599, + "grad_norm": 0.44524964690208435, + "learning_rate": 7.49250240461592e-05, + "loss": 0.0285, + "step": 36720 + }, + { + "epoch": 2.4064731704121076, + "grad_norm": 0.16568107903003693, + "learning_rate": 7.491068892858826e-05, + "loss": 0.031, + "step": 36730 + }, + { + "epoch": 2.4071283496036164, + "grad_norm": 0.15723636746406555, + "learning_rate": 7.489635108682184e-05, + "loss": 0.0353, + "step": 36740 + }, + { + "epoch": 2.4077835287951252, + "grad_norm": 0.2563248872756958, + "learning_rate": 7.48820105224279e-05, + "loss": 0.0206, + "step": 36750 + }, + { + "epoch": 2.4084387079866345, + "grad_norm": 0.45485010743141174, + "learning_rate": 7.48676672369747e-05, + "loss": 0.0364, + "step": 36760 + }, + { + "epoch": 2.4090938871781433, + "grad_norm": 0.5638884902000427, + "learning_rate": 7.485332123203079e-05, + "loss": 0.0436, + "step": 36770 + }, + { + "epoch": 2.409749066369652, + "grad_norm": 1.1281479597091675, + "learning_rate": 7.483897250916504e-05, + "loss": 0.0703, + "step": 36780 + }, + { + "epoch": 2.410404245561161, + "grad_norm": 0.4079635739326477, + "learning_rate": 7.48246210699466e-05, + "loss": 0.0493, + "step": 36790 + }, + { + "epoch": 2.4110594247526698, + "grad_norm": 0.8213363885879517, + "learning_rate": 7.481026691594492e-05, + "loss": 0.0371, + "step": 36800 + }, + { + "epoch": 2.4117146039441786, + "grad_norm": 0.5406926274299622, + "learning_rate": 7.479591004872975e-05, + "loss": 0.0352, + "step": 36810 + }, + { + "epoch": 2.412369783135688, + "grad_norm": 0.6900899410247803, + "learning_rate": 7.478155046987114e-05, + "loss": 0.038, + "step": 36820 + }, + { + "epoch": 2.4130249623271967, + "grad_norm": 0.8188183307647705, + "learning_rate": 7.476718818093944e-05, + "loss": 0.04, + "step": 36830 + }, + { + "epoch": 2.4136801415187055, + "grad_norm": 0.642259418964386, + "learning_rate": 7.475282318350527e-05, + "loss": 0.0715, + "step": 36840 + }, + { + "epoch": 2.4143353207102143, + "grad_norm": 0.43981775641441345, + "learning_rate": 7.473845547913957e-05, + "loss": 0.0657, + "step": 36850 + }, + { + "epoch": 2.414990499901723, + "grad_norm": 0.32457786798477173, + "learning_rate": 7.472408506941357e-05, + "loss": 0.0337, + "step": 36860 + }, + { + "epoch": 2.415645679093232, + "grad_norm": 0.2844753563404083, + "learning_rate": 7.470971195589881e-05, + "loss": 0.069, + "step": 36870 + }, + { + "epoch": 2.4163008582847407, + "grad_norm": 0.11459987610578537, + "learning_rate": 7.46953361401671e-05, + "loss": 0.0474, + "step": 36880 + }, + { + "epoch": 2.4169560374762495, + "grad_norm": 0.25585004687309265, + "learning_rate": 7.468095762379055e-05, + "loss": 0.0751, + "step": 36890 + }, + { + "epoch": 2.417611216667759, + "grad_norm": 0.2031533122062683, + "learning_rate": 7.466657640834158e-05, + "loss": 0.0413, + "step": 36900 + }, + { + "epoch": 2.4182663958592676, + "grad_norm": 0.3878383934497833, + "learning_rate": 7.46521924953929e-05, + "loss": 0.0376, + "step": 36910 + }, + { + "epoch": 2.4189215750507764, + "grad_norm": 0.44383642077445984, + "learning_rate": 7.463780588651752e-05, + "loss": 0.0469, + "step": 36920 + }, + { + "epoch": 2.4195767542422852, + "grad_norm": 0.10259666293859482, + "learning_rate": 7.462341658328873e-05, + "loss": 0.0301, + "step": 36930 + }, + { + "epoch": 2.420231933433794, + "grad_norm": 0.9191809296607971, + "learning_rate": 7.460902458728012e-05, + "loss": 0.0301, + "step": 36940 + }, + { + "epoch": 2.420887112625303, + "grad_norm": 0.2545137405395508, + "learning_rate": 7.459462990006558e-05, + "loss": 0.0855, + "step": 36950 + }, + { + "epoch": 2.421542291816812, + "grad_norm": 0.9227684140205383, + "learning_rate": 7.458023252321929e-05, + "loss": 0.0299, + "step": 36960 + }, + { + "epoch": 2.422197471008321, + "grad_norm": 0.6018545031547546, + "learning_rate": 7.456583245831573e-05, + "loss": 0.0238, + "step": 36970 + }, + { + "epoch": 2.4228526501998298, + "grad_norm": 0.0924191027879715, + "learning_rate": 7.455142970692968e-05, + "loss": 0.0366, + "step": 36980 + }, + { + "epoch": 2.4235078293913386, + "grad_norm": 1.7699028253555298, + "learning_rate": 7.453702427063615e-05, + "loss": 0.0498, + "step": 36990 + }, + { + "epoch": 2.4241630085828474, + "grad_norm": 0.8183063864707947, + "learning_rate": 7.452261615101057e-05, + "loss": 0.0523, + "step": 37000 + }, + { + "epoch": 2.424818187774356, + "grad_norm": 0.6487888097763062, + "learning_rate": 7.450820534962856e-05, + "loss": 0.0522, + "step": 37010 + }, + { + "epoch": 2.425473366965865, + "grad_norm": 0.3996940851211548, + "learning_rate": 7.449379186806604e-05, + "loss": 0.0375, + "step": 37020 + }, + { + "epoch": 2.426128546157374, + "grad_norm": 0.43436020612716675, + "learning_rate": 7.44793757078993e-05, + "loss": 0.0385, + "step": 37030 + }, + { + "epoch": 2.426783725348883, + "grad_norm": 0.6689348220825195, + "learning_rate": 7.44649568707048e-05, + "loss": 0.0213, + "step": 37040 + }, + { + "epoch": 2.427438904540392, + "grad_norm": 0.6591811776161194, + "learning_rate": 7.445053535805942e-05, + "loss": 0.0442, + "step": 37050 + }, + { + "epoch": 2.4280940837319007, + "grad_norm": 0.31921306252479553, + "learning_rate": 7.443611117154024e-05, + "loss": 0.0244, + "step": 37060 + }, + { + "epoch": 2.4287492629234095, + "grad_norm": 1.0032740831375122, + "learning_rate": 7.442168431272469e-05, + "loss": 0.0359, + "step": 37070 + }, + { + "epoch": 2.4294044421149183, + "grad_norm": 0.6544274091720581, + "learning_rate": 7.440725478319044e-05, + "loss": 0.0474, + "step": 37080 + }, + { + "epoch": 2.430059621306427, + "grad_norm": 0.491586297750473, + "learning_rate": 7.439282258451553e-05, + "loss": 0.0483, + "step": 37090 + }, + { + "epoch": 2.4307148004979364, + "grad_norm": 0.3403578996658325, + "learning_rate": 7.43783877182782e-05, + "loss": 0.0326, + "step": 37100 + }, + { + "epoch": 2.4313699796894452, + "grad_norm": 0.42625391483306885, + "learning_rate": 7.436395018605707e-05, + "loss": 0.0541, + "step": 37110 + }, + { + "epoch": 2.432025158880954, + "grad_norm": 0.6730939149856567, + "learning_rate": 7.434950998943094e-05, + "loss": 0.0329, + "step": 37120 + }, + { + "epoch": 2.432680338072463, + "grad_norm": 0.24695581197738647, + "learning_rate": 7.433506712997902e-05, + "loss": 0.0452, + "step": 37130 + }, + { + "epoch": 2.4333355172639717, + "grad_norm": 0.38608554005622864, + "learning_rate": 7.432062160928077e-05, + "loss": 0.0217, + "step": 37140 + }, + { + "epoch": 2.4339906964554805, + "grad_norm": 0.8962600231170654, + "learning_rate": 7.430617342891588e-05, + "loss": 0.0356, + "step": 37150 + }, + { + "epoch": 2.4346458756469893, + "grad_norm": 0.71773761510849, + "learning_rate": 7.429172259046442e-05, + "loss": 0.0475, + "step": 37160 + }, + { + "epoch": 2.435301054838498, + "grad_norm": 0.14589619636535645, + "learning_rate": 7.42772690955067e-05, + "loss": 0.0403, + "step": 37170 + }, + { + "epoch": 2.4359562340300074, + "grad_norm": 0.7099354863166809, + "learning_rate": 7.426281294562332e-05, + "loss": 0.0384, + "step": 37180 + }, + { + "epoch": 2.436611413221516, + "grad_norm": 1.0743292570114136, + "learning_rate": 7.424835414239519e-05, + "loss": 0.0397, + "step": 37190 + }, + { + "epoch": 2.437266592413025, + "grad_norm": 0.2807134687900543, + "learning_rate": 7.423389268740352e-05, + "loss": 0.0771, + "step": 37200 + }, + { + "epoch": 2.437921771604534, + "grad_norm": 0.34122875332832336, + "learning_rate": 7.42194285822298e-05, + "loss": 0.0451, + "step": 37210 + }, + { + "epoch": 2.4385769507960426, + "grad_norm": 0.49228277802467346, + "learning_rate": 7.420496182845576e-05, + "loss": 0.0361, + "step": 37220 + }, + { + "epoch": 2.4392321299875515, + "grad_norm": 0.6862915754318237, + "learning_rate": 7.419049242766348e-05, + "loss": 0.0286, + "step": 37230 + }, + { + "epoch": 2.4398873091790607, + "grad_norm": 0.42795121669769287, + "learning_rate": 7.417602038143533e-05, + "loss": 0.0486, + "step": 37240 + }, + { + "epoch": 2.4405424883705695, + "grad_norm": 0.5781195759773254, + "learning_rate": 7.416154569135393e-05, + "loss": 0.0469, + "step": 37250 + }, + { + "epoch": 2.4411976675620783, + "grad_norm": 0.28648653626441956, + "learning_rate": 7.414706835900222e-05, + "loss": 0.0528, + "step": 37260 + }, + { + "epoch": 2.441852846753587, + "grad_norm": 0.5633379220962524, + "learning_rate": 7.413258838596345e-05, + "loss": 0.0325, + "step": 37270 + }, + { + "epoch": 2.442508025945096, + "grad_norm": 0.11202720552682877, + "learning_rate": 7.411810577382106e-05, + "loss": 0.066, + "step": 37280 + }, + { + "epoch": 2.443163205136605, + "grad_norm": 0.6633576154708862, + "learning_rate": 7.410362052415889e-05, + "loss": 0.0484, + "step": 37290 + }, + { + "epoch": 2.4438183843281136, + "grad_norm": 0.6732990741729736, + "learning_rate": 7.408913263856102e-05, + "loss": 0.0287, + "step": 37300 + }, + { + "epoch": 2.4444735635196224, + "grad_norm": 0.7889837622642517, + "learning_rate": 7.407464211861183e-05, + "loss": 0.033, + "step": 37310 + }, + { + "epoch": 2.4451287427111317, + "grad_norm": 0.5164893865585327, + "learning_rate": 7.406014896589597e-05, + "loss": 0.0426, + "step": 37320 + }, + { + "epoch": 2.4457839219026405, + "grad_norm": 0.49471741914749146, + "learning_rate": 7.404565318199838e-05, + "loss": 0.0442, + "step": 37330 + }, + { + "epoch": 2.4464391010941493, + "grad_norm": 0.7359058856964111, + "learning_rate": 7.403115476850432e-05, + "loss": 0.0578, + "step": 37340 + }, + { + "epoch": 2.447094280285658, + "grad_norm": 0.5257327556610107, + "learning_rate": 7.401665372699932e-05, + "loss": 0.0518, + "step": 37350 + }, + { + "epoch": 2.447749459477167, + "grad_norm": 0.21417726576328278, + "learning_rate": 7.400215005906916e-05, + "loss": 0.0201, + "step": 37360 + }, + { + "epoch": 2.4484046386686757, + "grad_norm": 0.3136187195777893, + "learning_rate": 7.398764376629996e-05, + "loss": 0.0455, + "step": 37370 + }, + { + "epoch": 2.449059817860185, + "grad_norm": 0.409189909696579, + "learning_rate": 7.39731348502781e-05, + "loss": 0.021, + "step": 37380 + }, + { + "epoch": 2.449714997051694, + "grad_norm": 0.14981849491596222, + "learning_rate": 7.395862331259028e-05, + "loss": 0.0544, + "step": 37390 + }, + { + "epoch": 2.4503701762432026, + "grad_norm": 1.2520893812179565, + "learning_rate": 7.394410915482341e-05, + "loss": 0.0338, + "step": 37400 + }, + { + "epoch": 2.4510253554347115, + "grad_norm": 0.11103280633687973, + "learning_rate": 7.392959237856476e-05, + "loss": 0.0405, + "step": 37410 + }, + { + "epoch": 2.4516805346262203, + "grad_norm": 0.6159330010414124, + "learning_rate": 7.391507298540187e-05, + "loss": 0.0293, + "step": 37420 + }, + { + "epoch": 2.452335713817729, + "grad_norm": 0.4394742250442505, + "learning_rate": 7.390055097692257e-05, + "loss": 0.0396, + "step": 37430 + }, + { + "epoch": 2.452990893009238, + "grad_norm": 0.31724563241004944, + "learning_rate": 7.388602635471494e-05, + "loss": 0.0422, + "step": 37440 + }, + { + "epoch": 2.4536460722007467, + "grad_norm": 0.3385941684246063, + "learning_rate": 7.387149912036738e-05, + "loss": 0.0346, + "step": 37450 + }, + { + "epoch": 2.454301251392256, + "grad_norm": 0.9437400698661804, + "learning_rate": 7.385696927546854e-05, + "loss": 0.0529, + "step": 37460 + }, + { + "epoch": 2.454956430583765, + "grad_norm": 0.1203138455748558, + "learning_rate": 7.384243682160743e-05, + "loss": 0.08, + "step": 37470 + }, + { + "epoch": 2.4556116097752736, + "grad_norm": 0.36090075969696045, + "learning_rate": 7.382790176037329e-05, + "loss": 0.0361, + "step": 37480 + }, + { + "epoch": 2.4562667889667824, + "grad_norm": 0.4096725583076477, + "learning_rate": 7.381336409335562e-05, + "loss": 0.0467, + "step": 37490 + }, + { + "epoch": 2.4569219681582912, + "grad_norm": 0.20653633773326874, + "learning_rate": 7.379882382214426e-05, + "loss": 0.0267, + "step": 37500 + }, + { + "epoch": 2.4575771473498, + "grad_norm": 0.5941516757011414, + "learning_rate": 7.378428094832931e-05, + "loss": 0.0453, + "step": 37510 + }, + { + "epoch": 2.458232326541309, + "grad_norm": 0.36623072624206543, + "learning_rate": 7.376973547350114e-05, + "loss": 0.0168, + "step": 37520 + }, + { + "epoch": 2.458887505732818, + "grad_norm": 0.6241915225982666, + "learning_rate": 7.375518739925046e-05, + "loss": 0.0484, + "step": 37530 + }, + { + "epoch": 2.459542684924327, + "grad_norm": 1.3216255903244019, + "learning_rate": 7.374063672716818e-05, + "loss": 0.0642, + "step": 37540 + }, + { + "epoch": 2.4601978641158357, + "grad_norm": 0.7502813339233398, + "learning_rate": 7.372608345884558e-05, + "loss": 0.0462, + "step": 37550 + }, + { + "epoch": 2.4608530433073446, + "grad_norm": 0.31838521361351013, + "learning_rate": 7.371152759587415e-05, + "loss": 0.0237, + "step": 37560 + }, + { + "epoch": 2.4615082224988534, + "grad_norm": 0.2709442675113678, + "learning_rate": 7.369696913984572e-05, + "loss": 0.048, + "step": 37570 + }, + { + "epoch": 2.462163401690362, + "grad_norm": 0.2647649049758911, + "learning_rate": 7.368240809235237e-05, + "loss": 0.0328, + "step": 37580 + }, + { + "epoch": 2.462818580881871, + "grad_norm": 1.3320101499557495, + "learning_rate": 7.366784445498649e-05, + "loss": 0.0574, + "step": 37590 + }, + { + "epoch": 2.4634737600733803, + "grad_norm": 0.4530133008956909, + "learning_rate": 7.36532782293407e-05, + "loss": 0.0433, + "step": 37600 + }, + { + "epoch": 2.464128939264889, + "grad_norm": 1.0090546607971191, + "learning_rate": 7.363870941700797e-05, + "loss": 0.0767, + "step": 37610 + }, + { + "epoch": 2.464784118456398, + "grad_norm": 0.2893260717391968, + "learning_rate": 7.362413801958152e-05, + "loss": 0.0318, + "step": 37620 + }, + { + "epoch": 2.4654392976479067, + "grad_norm": 0.14789380133152008, + "learning_rate": 7.360956403865486e-05, + "loss": 0.0323, + "step": 37630 + }, + { + "epoch": 2.4660944768394155, + "grad_norm": 0.37132078409194946, + "learning_rate": 7.359498747582177e-05, + "loss": 0.0412, + "step": 37640 + }, + { + "epoch": 2.4667496560309243, + "grad_norm": 0.33049795031547546, + "learning_rate": 7.358040833267634e-05, + "loss": 0.0405, + "step": 37650 + }, + { + "epoch": 2.467404835222433, + "grad_norm": 0.37623170018196106, + "learning_rate": 7.356582661081288e-05, + "loss": 0.0262, + "step": 37660 + }, + { + "epoch": 2.4680600144139424, + "grad_norm": 0.6937607526779175, + "learning_rate": 7.355124231182607e-05, + "loss": 0.0371, + "step": 37670 + }, + { + "epoch": 2.468715193605451, + "grad_norm": 0.4358172118663788, + "learning_rate": 7.353665543731083e-05, + "loss": 0.0885, + "step": 37680 + }, + { + "epoch": 2.46937037279696, + "grad_norm": 0.16603899002075195, + "learning_rate": 7.352206598886229e-05, + "loss": 0.0277, + "step": 37690 + }, + { + "epoch": 2.470025551988469, + "grad_norm": 0.3565003573894501, + "learning_rate": 7.350747396807601e-05, + "loss": 0.034, + "step": 37700 + }, + { + "epoch": 2.4706807311799777, + "grad_norm": 0.5782313942909241, + "learning_rate": 7.349287937654772e-05, + "loss": 0.0427, + "step": 37710 + }, + { + "epoch": 2.4713359103714865, + "grad_norm": 0.4750462770462036, + "learning_rate": 7.347828221587345e-05, + "loss": 0.0265, + "step": 37720 + }, + { + "epoch": 2.4719910895629953, + "grad_norm": 0.4232843816280365, + "learning_rate": 7.346368248764955e-05, + "loss": 0.0569, + "step": 37730 + }, + { + "epoch": 2.4726462687545046, + "grad_norm": 0.8492642045021057, + "learning_rate": 7.34490801934726e-05, + "loss": 0.0484, + "step": 37740 + }, + { + "epoch": 2.4733014479460134, + "grad_norm": 0.1687699258327484, + "learning_rate": 7.343447533493947e-05, + "loss": 0.0378, + "step": 37750 + }, + { + "epoch": 2.473956627137522, + "grad_norm": 0.27135753631591797, + "learning_rate": 7.341986791364739e-05, + "loss": 0.0507, + "step": 37760 + }, + { + "epoch": 2.474611806329031, + "grad_norm": 0.6613906025886536, + "learning_rate": 7.340525793119374e-05, + "loss": 0.0366, + "step": 37770 + }, + { + "epoch": 2.47526698552054, + "grad_norm": 0.6949599981307983, + "learning_rate": 7.33906453891763e-05, + "loss": 0.0545, + "step": 37780 + }, + { + "epoch": 2.4759221647120486, + "grad_norm": 0.26894453167915344, + "learning_rate": 7.337603028919303e-05, + "loss": 0.0188, + "step": 37790 + }, + { + "epoch": 2.4765773439035574, + "grad_norm": 0.8416797518730164, + "learning_rate": 7.336141263284225e-05, + "loss": 0.0199, + "step": 37800 + }, + { + "epoch": 2.4772325230950667, + "grad_norm": 0.7649991512298584, + "learning_rate": 7.334679242172249e-05, + "loss": 0.0476, + "step": 37810 + }, + { + "epoch": 2.4778877022865755, + "grad_norm": 0.2116391807794571, + "learning_rate": 7.33321696574326e-05, + "loss": 0.044, + "step": 37820 + }, + { + "epoch": 2.4785428814780843, + "grad_norm": 0.29911574721336365, + "learning_rate": 7.331754434157174e-05, + "loss": 0.0297, + "step": 37830 + }, + { + "epoch": 2.479198060669593, + "grad_norm": 0.6229132413864136, + "learning_rate": 7.330291647573928e-05, + "loss": 0.0496, + "step": 37840 + }, + { + "epoch": 2.479853239861102, + "grad_norm": 0.5518564581871033, + "learning_rate": 7.32882860615349e-05, + "loss": 0.0519, + "step": 37850 + }, + { + "epoch": 2.4805084190526108, + "grad_norm": 0.06968197971582413, + "learning_rate": 7.327365310055858e-05, + "loss": 0.0385, + "step": 37860 + }, + { + "epoch": 2.4811635982441196, + "grad_norm": 0.18047532439231873, + "learning_rate": 7.325901759441055e-05, + "loss": 0.0928, + "step": 37870 + }, + { + "epoch": 2.4818187774356284, + "grad_norm": 0.28291431069374084, + "learning_rate": 7.324437954469132e-05, + "loss": 0.0256, + "step": 37880 + }, + { + "epoch": 2.4824739566271377, + "grad_norm": 0.7534339427947998, + "learning_rate": 7.32297389530017e-05, + "loss": 0.049, + "step": 37890 + }, + { + "epoch": 2.4831291358186465, + "grad_norm": 0.61979079246521, + "learning_rate": 7.321509582094274e-05, + "loss": 0.0432, + "step": 37900 + }, + { + "epoch": 2.4837843150101553, + "grad_norm": 0.4562670588493347, + "learning_rate": 7.320045015011582e-05, + "loss": 0.0413, + "step": 37910 + }, + { + "epoch": 2.484439494201664, + "grad_norm": 0.07486114650964737, + "learning_rate": 7.318580194212253e-05, + "loss": 0.0491, + "step": 37920 + }, + { + "epoch": 2.485094673393173, + "grad_norm": 0.5174303650856018, + "learning_rate": 7.317115119856483e-05, + "loss": 0.0289, + "step": 37930 + }, + { + "epoch": 2.4857498525846817, + "grad_norm": 0.33598053455352783, + "learning_rate": 7.315649792104485e-05, + "loss": 0.0217, + "step": 37940 + }, + { + "epoch": 2.486405031776191, + "grad_norm": 0.6547322273254395, + "learning_rate": 7.314184211116507e-05, + "loss": 0.0394, + "step": 37950 + }, + { + "epoch": 2.4870602109677, + "grad_norm": 0.924709677696228, + "learning_rate": 7.312718377052823e-05, + "loss": 0.0345, + "step": 37960 + }, + { + "epoch": 2.4877153901592086, + "grad_norm": 0.15535353124141693, + "learning_rate": 7.311252290073737e-05, + "loss": 0.043, + "step": 37970 + }, + { + "epoch": 2.4883705693507174, + "grad_norm": 0.12392053008079529, + "learning_rate": 7.309785950339572e-05, + "loss": 0.0312, + "step": 37980 + }, + { + "epoch": 2.4890257485422262, + "grad_norm": 1.3315540552139282, + "learning_rate": 7.308319358010692e-05, + "loss": 0.0645, + "step": 37990 + }, + { + "epoch": 2.489680927733735, + "grad_norm": 0.5617696642875671, + "learning_rate": 7.306852513247474e-05, + "loss": 0.0288, + "step": 38000 + }, + { + "epoch": 2.490336106925244, + "grad_norm": 0.2931419312953949, + "learning_rate": 7.305385416210336e-05, + "loss": 0.0511, + "step": 38010 + }, + { + "epoch": 2.4909912861167527, + "grad_norm": 0.962295413017273, + "learning_rate": 7.303918067059711e-05, + "loss": 0.0354, + "step": 38020 + }, + { + "epoch": 2.491646465308262, + "grad_norm": 0.16161181032657623, + "learning_rate": 7.302450465956075e-05, + "loss": 0.0581, + "step": 38030 + }, + { + "epoch": 2.4923016444997708, + "grad_norm": 0.5508818030357361, + "learning_rate": 7.300982613059914e-05, + "loss": 0.0336, + "step": 38040 + }, + { + "epoch": 2.4929568236912796, + "grad_norm": 1.2758580446243286, + "learning_rate": 7.299514508531757e-05, + "loss": 0.0383, + "step": 38050 + }, + { + "epoch": 2.4936120028827884, + "grad_norm": 0.5261476039886475, + "learning_rate": 7.298046152532148e-05, + "loss": 0.0421, + "step": 38060 + }, + { + "epoch": 2.494267182074297, + "grad_norm": 0.2953239679336548, + "learning_rate": 7.296577545221668e-05, + "loss": 0.0702, + "step": 38070 + }, + { + "epoch": 2.494922361265806, + "grad_norm": 0.8765751123428345, + "learning_rate": 7.29510868676092e-05, + "loss": 0.0397, + "step": 38080 + }, + { + "epoch": 2.4955775404573153, + "grad_norm": 0.8361433148384094, + "learning_rate": 7.293639577310538e-05, + "loss": 0.0344, + "step": 38090 + }, + { + "epoch": 2.496232719648824, + "grad_norm": 0.14607387781143188, + "learning_rate": 7.292170217031179e-05, + "loss": 0.0364, + "step": 38100 + }, + { + "epoch": 2.496887898840333, + "grad_norm": 0.21017104387283325, + "learning_rate": 7.290700606083532e-05, + "loss": 0.042, + "step": 38110 + }, + { + "epoch": 2.4975430780318417, + "grad_norm": 0.1964062601327896, + "learning_rate": 7.28923074462831e-05, + "loss": 0.0299, + "step": 38120 + }, + { + "epoch": 2.4981982572233505, + "grad_norm": 0.20153504610061646, + "learning_rate": 7.287760632826255e-05, + "loss": 0.0552, + "step": 38130 + }, + { + "epoch": 2.4988534364148594, + "grad_norm": 0.7230279445648193, + "learning_rate": 7.286290270838139e-05, + "loss": 0.0386, + "step": 38140 + }, + { + "epoch": 2.499508615606368, + "grad_norm": 0.5505397319793701, + "learning_rate": 7.284819658824756e-05, + "loss": 0.0307, + "step": 38150 + }, + { + "epoch": 2.500163794797877, + "grad_norm": 0.464507520198822, + "learning_rate": 7.283348796946929e-05, + "loss": 0.0647, + "step": 38160 + }, + { + "epoch": 2.5008189739893862, + "grad_norm": 1.035473108291626, + "learning_rate": 7.28187768536551e-05, + "loss": 0.0351, + "step": 38170 + }, + { + "epoch": 2.501474153180895, + "grad_norm": 1.8604813814163208, + "learning_rate": 7.280406324241379e-05, + "loss": 0.0574, + "step": 38180 + }, + { + "epoch": 2.502129332372404, + "grad_norm": 0.2913621962070465, + "learning_rate": 7.278934713735441e-05, + "loss": 0.0422, + "step": 38190 + }, + { + "epoch": 2.5027845115639127, + "grad_norm": 0.1915731430053711, + "learning_rate": 7.277462854008629e-05, + "loss": 0.0203, + "step": 38200 + }, + { + "epoch": 2.5034396907554215, + "grad_norm": 0.15182246267795563, + "learning_rate": 7.275990745221904e-05, + "loss": 0.0501, + "step": 38210 + }, + { + "epoch": 2.5040948699469308, + "grad_norm": 1.0429909229278564, + "learning_rate": 7.274518387536252e-05, + "loss": 0.0476, + "step": 38220 + }, + { + "epoch": 2.5047500491384396, + "grad_norm": 0.42609062790870667, + "learning_rate": 7.273045781112689e-05, + "loss": 0.0297, + "step": 38230 + }, + { + "epoch": 2.5054052283299484, + "grad_norm": 0.20785999298095703, + "learning_rate": 7.271572926112257e-05, + "loss": 0.0331, + "step": 38240 + }, + { + "epoch": 2.506060407521457, + "grad_norm": 0.2964385747909546, + "learning_rate": 7.270099822696024e-05, + "loss": 0.0465, + "step": 38250 + }, + { + "epoch": 2.506715586712966, + "grad_norm": 1.025687575340271, + "learning_rate": 7.26862647102509e-05, + "loss": 0.0563, + "step": 38260 + }, + { + "epoch": 2.507370765904475, + "grad_norm": 0.35313335061073303, + "learning_rate": 7.267152871260573e-05, + "loss": 0.034, + "step": 38270 + }, + { + "epoch": 2.5080259450959836, + "grad_norm": 0.17768354713916779, + "learning_rate": 7.265679023563629e-05, + "loss": 0.0718, + "step": 38280 + }, + { + "epoch": 2.5086811242874925, + "grad_norm": 0.33181697130203247, + "learning_rate": 7.264204928095433e-05, + "loss": 0.0408, + "step": 38290 + }, + { + "epoch": 2.5093363034790013, + "grad_norm": 0.2908321022987366, + "learning_rate": 7.262730585017188e-05, + "loss": 0.0339, + "step": 38300 + }, + { + "epoch": 2.5099914826705105, + "grad_norm": 0.395709365606308, + "learning_rate": 7.261255994490131e-05, + "loss": 0.0504, + "step": 38310 + }, + { + "epoch": 2.5106466618620193, + "grad_norm": 0.4397352635860443, + "learning_rate": 7.259781156675517e-05, + "loss": 0.0777, + "step": 38320 + }, + { + "epoch": 2.511301841053528, + "grad_norm": 0.24671579897403717, + "learning_rate": 7.258306071734633e-05, + "loss": 0.0207, + "step": 38330 + }, + { + "epoch": 2.511957020245037, + "grad_norm": 0.21292644739151, + "learning_rate": 7.256830739828792e-05, + "loss": 0.027, + "step": 38340 + }, + { + "epoch": 2.512612199436546, + "grad_norm": 0.12798844277858734, + "learning_rate": 7.255355161119336e-05, + "loss": 0.0252, + "step": 38350 + }, + { + "epoch": 2.513267378628055, + "grad_norm": 0.2200646996498108, + "learning_rate": 7.253879335767628e-05, + "loss": 0.0614, + "step": 38360 + }, + { + "epoch": 2.513922557819564, + "grad_norm": 0.1465122550725937, + "learning_rate": 7.252403263935067e-05, + "loss": 0.0389, + "step": 38370 + }, + { + "epoch": 2.5145777370110727, + "grad_norm": 1.0123085975646973, + "learning_rate": 7.25092694578307e-05, + "loss": 0.0453, + "step": 38380 + }, + { + "epoch": 2.5152329162025815, + "grad_norm": 0.7018684148788452, + "learning_rate": 7.249450381473087e-05, + "loss": 0.0578, + "step": 38390 + }, + { + "epoch": 2.5158880953940903, + "grad_norm": 0.8044062852859497, + "learning_rate": 7.247973571166593e-05, + "loss": 0.0592, + "step": 38400 + }, + { + "epoch": 2.516543274585599, + "grad_norm": 0.6363970041275024, + "learning_rate": 7.246496515025089e-05, + "loss": 0.035, + "step": 38410 + }, + { + "epoch": 2.517198453777108, + "grad_norm": 0.4290277063846588, + "learning_rate": 7.245019213210101e-05, + "loss": 0.0252, + "step": 38420 + }, + { + "epoch": 2.5178536329686168, + "grad_norm": 0.4528871178627014, + "learning_rate": 7.24354166588319e-05, + "loss": 0.0402, + "step": 38430 + }, + { + "epoch": 2.5185088121601256, + "grad_norm": 0.678098738193512, + "learning_rate": 7.242063873205935e-05, + "loss": 0.0468, + "step": 38440 + }, + { + "epoch": 2.519163991351635, + "grad_norm": 0.302834153175354, + "learning_rate": 7.240585835339946e-05, + "loss": 0.0536, + "step": 38450 + }, + { + "epoch": 2.5198191705431436, + "grad_norm": 0.09377434849739075, + "learning_rate": 7.239107552446857e-05, + "loss": 0.0439, + "step": 38460 + }, + { + "epoch": 2.5204743497346525, + "grad_norm": 1.0312049388885498, + "learning_rate": 7.237629024688334e-05, + "loss": 0.052, + "step": 38470 + }, + { + "epoch": 2.5211295289261613, + "grad_norm": 0.24221786856651306, + "learning_rate": 7.236150252226064e-05, + "loss": 0.0329, + "step": 38480 + }, + { + "epoch": 2.52178470811767, + "grad_norm": 0.13600507378578186, + "learning_rate": 7.234671235221765e-05, + "loss": 0.0503, + "step": 38490 + }, + { + "epoch": 2.522439887309179, + "grad_norm": 0.1898692399263382, + "learning_rate": 7.233191973837179e-05, + "loss": 0.054, + "step": 38500 + }, + { + "epoch": 2.523095066500688, + "grad_norm": 0.12774424254894257, + "learning_rate": 7.231712468234078e-05, + "loss": 0.0282, + "step": 38510 + }, + { + "epoch": 2.523750245692197, + "grad_norm": 0.22654519975185394, + "learning_rate": 7.230232718574254e-05, + "loss": 0.0264, + "step": 38520 + }, + { + "epoch": 2.524405424883706, + "grad_norm": 0.27575036883354187, + "learning_rate": 7.228752725019535e-05, + "loss": 0.0215, + "step": 38530 + }, + { + "epoch": 2.5250606040752146, + "grad_norm": 0.12049344927072525, + "learning_rate": 7.227272487731769e-05, + "loss": 0.0259, + "step": 38540 + }, + { + "epoch": 2.5257157832667234, + "grad_norm": 0.5662280917167664, + "learning_rate": 7.225792006872831e-05, + "loss": 0.0493, + "step": 38550 + }, + { + "epoch": 2.5263709624582322, + "grad_norm": 0.7266760468482971, + "learning_rate": 7.224311282604628e-05, + "loss": 0.0372, + "step": 38560 + }, + { + "epoch": 2.527026141649741, + "grad_norm": 0.3426131010055542, + "learning_rate": 7.222830315089085e-05, + "loss": 0.031, + "step": 38570 + }, + { + "epoch": 2.52768132084125, + "grad_norm": 0.7160592675209045, + "learning_rate": 7.22134910448816e-05, + "loss": 0.0438, + "step": 38580 + }, + { + "epoch": 2.528336500032759, + "grad_norm": 0.8232858180999756, + "learning_rate": 7.219867650963839e-05, + "loss": 0.032, + "step": 38590 + }, + { + "epoch": 2.528991679224268, + "grad_norm": 0.12188532203435898, + "learning_rate": 7.21838595467813e-05, + "loss": 0.0402, + "step": 38600 + }, + { + "epoch": 2.5296468584157767, + "grad_norm": 0.168903186917305, + "learning_rate": 7.216904015793067e-05, + "loss": 0.0431, + "step": 38610 + }, + { + "epoch": 2.5303020376072856, + "grad_norm": 0.5164052844047546, + "learning_rate": 7.215421834470713e-05, + "loss": 0.0582, + "step": 38620 + }, + { + "epoch": 2.5309572167987944, + "grad_norm": 0.315646767616272, + "learning_rate": 7.213939410873158e-05, + "loss": 0.0245, + "step": 38630 + }, + { + "epoch": 2.531612395990303, + "grad_norm": 0.2645866572856903, + "learning_rate": 7.212456745162519e-05, + "loss": 0.077, + "step": 38640 + }, + { + "epoch": 2.5322675751818124, + "grad_norm": 0.33142077922821045, + "learning_rate": 7.210973837500937e-05, + "loss": 0.037, + "step": 38650 + }, + { + "epoch": 2.5329227543733213, + "grad_norm": 0.1443328708410263, + "learning_rate": 7.209490688050578e-05, + "loss": 0.028, + "step": 38660 + }, + { + "epoch": 2.53357793356483, + "grad_norm": 0.6072419881820679, + "learning_rate": 7.208007296973641e-05, + "loss": 0.0417, + "step": 38670 + }, + { + "epoch": 2.534233112756339, + "grad_norm": 0.6306849122047424, + "learning_rate": 7.206523664432345e-05, + "loss": 0.045, + "step": 38680 + }, + { + "epoch": 2.5348882919478477, + "grad_norm": 0.42669278383255005, + "learning_rate": 7.205039790588939e-05, + "loss": 0.0327, + "step": 38690 + }, + { + "epoch": 2.5355434711393565, + "grad_norm": 0.2413020133972168, + "learning_rate": 7.203555675605697e-05, + "loss": 0.0315, + "step": 38700 + }, + { + "epoch": 2.5361986503308653, + "grad_norm": 0.8005293011665344, + "learning_rate": 7.202071319644917e-05, + "loss": 0.0421, + "step": 38710 + }, + { + "epoch": 2.536853829522374, + "grad_norm": 0.351998507976532, + "learning_rate": 7.200586722868932e-05, + "loss": 0.0146, + "step": 38720 + }, + { + "epoch": 2.5375090087138834, + "grad_norm": 0.37916630506515503, + "learning_rate": 7.199101885440088e-05, + "loss": 0.0635, + "step": 38730 + }, + { + "epoch": 2.538164187905392, + "grad_norm": 1.0798802375793457, + "learning_rate": 7.197616807520771e-05, + "loss": 0.0535, + "step": 38740 + }, + { + "epoch": 2.538819367096901, + "grad_norm": 0.788982093334198, + "learning_rate": 7.196131489273381e-05, + "loss": 0.0331, + "step": 38750 + }, + { + "epoch": 2.53947454628841, + "grad_norm": 0.21371883153915405, + "learning_rate": 7.194645930860355e-05, + "loss": 0.0528, + "step": 38760 + }, + { + "epoch": 2.5401297254799187, + "grad_norm": 0.501611053943634, + "learning_rate": 7.19316013244415e-05, + "loss": 0.0249, + "step": 38770 + }, + { + "epoch": 2.5407849046714275, + "grad_norm": 0.23957087099552155, + "learning_rate": 7.191674094187248e-05, + "loss": 0.0217, + "step": 38780 + }, + { + "epoch": 2.5414400838629367, + "grad_norm": 0.45943960547447205, + "learning_rate": 7.190187816252165e-05, + "loss": 0.028, + "step": 38790 + }, + { + "epoch": 2.5420952630544456, + "grad_norm": 0.09605368226766586, + "learning_rate": 7.188701298801435e-05, + "loss": 0.0316, + "step": 38800 + }, + { + "epoch": 2.5427504422459544, + "grad_norm": 0.18664661049842834, + "learning_rate": 7.18721454199762e-05, + "loss": 0.0487, + "step": 38810 + }, + { + "epoch": 2.543405621437463, + "grad_norm": 0.5153557658195496, + "learning_rate": 7.185727546003312e-05, + "loss": 0.0372, + "step": 38820 + }, + { + "epoch": 2.544060800628972, + "grad_norm": 0.3283784091472626, + "learning_rate": 7.184240310981126e-05, + "loss": 0.0207, + "step": 38830 + }, + { + "epoch": 2.544715979820481, + "grad_norm": 0.3771614730358124, + "learning_rate": 7.182752837093704e-05, + "loss": 0.0248, + "step": 38840 + }, + { + "epoch": 2.5453711590119896, + "grad_norm": 0.3786217272281647, + "learning_rate": 7.181265124503711e-05, + "loss": 0.0533, + "step": 38850 + }, + { + "epoch": 2.5460263382034984, + "grad_norm": 0.26660585403442383, + "learning_rate": 7.179777173373847e-05, + "loss": 0.027, + "step": 38860 + }, + { + "epoch": 2.5466815173950077, + "grad_norm": 1.289896011352539, + "learning_rate": 7.178288983866826e-05, + "loss": 0.0712, + "step": 38870 + }, + { + "epoch": 2.5473366965865165, + "grad_norm": 0.597817599773407, + "learning_rate": 7.176800556145397e-05, + "loss": 0.0461, + "step": 38880 + }, + { + "epoch": 2.5479918757780253, + "grad_norm": 0.5675580501556396, + "learning_rate": 7.175311890372334e-05, + "loss": 0.0546, + "step": 38890 + }, + { + "epoch": 2.548647054969534, + "grad_norm": 0.15688392519950867, + "learning_rate": 7.17382298671043e-05, + "loss": 0.0291, + "step": 38900 + }, + { + "epoch": 2.549302234161043, + "grad_norm": 0.3133058547973633, + "learning_rate": 7.172333845322515e-05, + "loss": 0.0491, + "step": 38910 + }, + { + "epoch": 2.5499574133525518, + "grad_norm": 0.9586992263793945, + "learning_rate": 7.170844466371436e-05, + "loss": 0.0587, + "step": 38920 + }, + { + "epoch": 2.550612592544061, + "grad_norm": 0.17616912722587585, + "learning_rate": 7.16935485002007e-05, + "loss": 0.0328, + "step": 38930 + }, + { + "epoch": 2.55126777173557, + "grad_norm": 2.4549670219421387, + "learning_rate": 7.167864996431319e-05, + "loss": 0.0437, + "step": 38940 + }, + { + "epoch": 2.5519229509270787, + "grad_norm": 0.42491522431373596, + "learning_rate": 7.166374905768111e-05, + "loss": 0.0322, + "step": 38950 + }, + { + "epoch": 2.5525781301185875, + "grad_norm": 0.34663620591163635, + "learning_rate": 7.164884578193401e-05, + "loss": 0.047, + "step": 38960 + }, + { + "epoch": 2.5532333093100963, + "grad_norm": 0.21371836960315704, + "learning_rate": 7.163394013870168e-05, + "loss": 0.0223, + "step": 38970 + }, + { + "epoch": 2.553888488501605, + "grad_norm": 0.3681418001651764, + "learning_rate": 7.161903212961419e-05, + "loss": 0.0415, + "step": 38980 + }, + { + "epoch": 2.554543667693114, + "grad_norm": 0.48255330324172974, + "learning_rate": 7.160412175630183e-05, + "loss": 0.0389, + "step": 38990 + }, + { + "epoch": 2.5551988468846227, + "grad_norm": 0.2927854359149933, + "learning_rate": 7.158920902039521e-05, + "loss": 0.0398, + "step": 39000 + }, + { + "epoch": 2.5558540260761315, + "grad_norm": 1.8284108638763428, + "learning_rate": 7.157429392352514e-05, + "loss": 0.0468, + "step": 39010 + }, + { + "epoch": 2.556509205267641, + "grad_norm": 0.17730411887168884, + "learning_rate": 7.155937646732275e-05, + "loss": 0.0433, + "step": 39020 + }, + { + "epoch": 2.5571643844591496, + "grad_norm": 0.5943756699562073, + "learning_rate": 7.154445665341933e-05, + "loss": 0.0556, + "step": 39030 + }, + { + "epoch": 2.5578195636506584, + "grad_norm": 0.32492050528526306, + "learning_rate": 7.152953448344654e-05, + "loss": 0.0245, + "step": 39040 + }, + { + "epoch": 2.5584747428421672, + "grad_norm": 0.582984983921051, + "learning_rate": 7.151460995903624e-05, + "loss": 0.0426, + "step": 39050 + }, + { + "epoch": 2.559129922033676, + "grad_norm": 0.6416367888450623, + "learning_rate": 7.149968308182052e-05, + "loss": 0.0461, + "step": 39060 + }, + { + "epoch": 2.5597851012251853, + "grad_norm": 0.7716245651245117, + "learning_rate": 7.14847538534318e-05, + "loss": 0.0439, + "step": 39070 + }, + { + "epoch": 2.560440280416694, + "grad_norm": 0.29831215739250183, + "learning_rate": 7.14698222755027e-05, + "loss": 0.0485, + "step": 39080 + }, + { + "epoch": 2.561095459608203, + "grad_norm": 0.31285929679870605, + "learning_rate": 7.14548883496661e-05, + "loss": 0.0473, + "step": 39090 + }, + { + "epoch": 2.5617506387997118, + "grad_norm": 0.8484737277030945, + "learning_rate": 7.143995207755517e-05, + "loss": 0.043, + "step": 39100 + }, + { + "epoch": 2.5624058179912206, + "grad_norm": 0.2928180992603302, + "learning_rate": 7.142501346080333e-05, + "loss": 0.0463, + "step": 39110 + }, + { + "epoch": 2.5630609971827294, + "grad_norm": 0.5283145904541016, + "learning_rate": 7.141007250104421e-05, + "loss": 0.0354, + "step": 39120 + }, + { + "epoch": 2.563716176374238, + "grad_norm": 0.2696027159690857, + "learning_rate": 7.139512919991176e-05, + "loss": 0.0285, + "step": 39130 + }, + { + "epoch": 2.564371355565747, + "grad_norm": 0.8339322209358215, + "learning_rate": 7.138018355904014e-05, + "loss": 0.0467, + "step": 39140 + }, + { + "epoch": 2.565026534757256, + "grad_norm": 0.5102400779724121, + "learning_rate": 7.13652355800638e-05, + "loss": 0.034, + "step": 39150 + }, + { + "epoch": 2.565681713948765, + "grad_norm": 0.3807276487350464, + "learning_rate": 7.135028526461743e-05, + "loss": 0.0515, + "step": 39160 + }, + { + "epoch": 2.566336893140274, + "grad_norm": 0.2038610428571701, + "learning_rate": 7.133533261433594e-05, + "loss": 0.0218, + "step": 39170 + }, + { + "epoch": 2.5669920723317827, + "grad_norm": 0.7247202396392822, + "learning_rate": 7.132037763085457e-05, + "loss": 0.0562, + "step": 39180 + }, + { + "epoch": 2.5676472515232915, + "grad_norm": 0.230916827917099, + "learning_rate": 7.130542031580875e-05, + "loss": 0.0421, + "step": 39190 + }, + { + "epoch": 2.5683024307148004, + "grad_norm": 0.19376550614833832, + "learning_rate": 7.12904606708342e-05, + "loss": 0.0305, + "step": 39200 + }, + { + "epoch": 2.5689576099063096, + "grad_norm": 0.6194016933441162, + "learning_rate": 7.127549869756687e-05, + "loss": 0.0343, + "step": 39210 + }, + { + "epoch": 2.5696127890978184, + "grad_norm": 0.8837741017341614, + "learning_rate": 7.126053439764299e-05, + "loss": 0.0334, + "step": 39220 + }, + { + "epoch": 2.5702679682893272, + "grad_norm": 0.4726470112800598, + "learning_rate": 7.124556777269904e-05, + "loss": 0.0256, + "step": 39230 + }, + { + "epoch": 2.570923147480836, + "grad_norm": 0.945478618144989, + "learning_rate": 7.123059882437174e-05, + "loss": 0.0497, + "step": 39240 + }, + { + "epoch": 2.571578326672345, + "grad_norm": 0.4175576865673065, + "learning_rate": 7.121562755429807e-05, + "loss": 0.0436, + "step": 39250 + }, + { + "epoch": 2.5722335058638537, + "grad_norm": 0.7760494947433472, + "learning_rate": 7.120065396411528e-05, + "loss": 0.0618, + "step": 39260 + }, + { + "epoch": 2.5728886850553625, + "grad_norm": 0.7194536924362183, + "learning_rate": 7.118567805546084e-05, + "loss": 0.033, + "step": 39270 + }, + { + "epoch": 2.5735438642468713, + "grad_norm": 0.17906557023525238, + "learning_rate": 7.117069982997248e-05, + "loss": 0.0362, + "step": 39280 + }, + { + "epoch": 2.57419904343838, + "grad_norm": 0.14603379368782043, + "learning_rate": 7.115571928928825e-05, + "loss": 0.0391, + "step": 39290 + }, + { + "epoch": 2.5748542226298894, + "grad_norm": 0.865169882774353, + "learning_rate": 7.114073643504635e-05, + "loss": 0.0457, + "step": 39300 + }, + { + "epoch": 2.575509401821398, + "grad_norm": 0.8125166296958923, + "learning_rate": 7.11257512688853e-05, + "loss": 0.0389, + "step": 39310 + }, + { + "epoch": 2.576164581012907, + "grad_norm": 0.17506586015224457, + "learning_rate": 7.111076379244384e-05, + "loss": 0.0346, + "step": 39320 + }, + { + "epoch": 2.576819760204416, + "grad_norm": 0.7300849556922913, + "learning_rate": 7.109577400736101e-05, + "loss": 0.0572, + "step": 39330 + }, + { + "epoch": 2.5774749393959246, + "grad_norm": 2.396970272064209, + "learning_rate": 7.108078191527605e-05, + "loss": 0.0717, + "step": 39340 + }, + { + "epoch": 2.578130118587434, + "grad_norm": 0.6558724641799927, + "learning_rate": 7.106578751782847e-05, + "loss": 0.0268, + "step": 39350 + }, + { + "epoch": 2.5787852977789427, + "grad_norm": 1.1773942708969116, + "learning_rate": 7.105079081665803e-05, + "loss": 0.0799, + "step": 39360 + }, + { + "epoch": 2.5794404769704515, + "grad_norm": 0.23099707067012787, + "learning_rate": 7.103579181340476e-05, + "loss": 0.0393, + "step": 39370 + }, + { + "epoch": 2.5800956561619603, + "grad_norm": 0.34675315022468567, + "learning_rate": 7.102079050970893e-05, + "loss": 0.0527, + "step": 39380 + }, + { + "epoch": 2.580750835353469, + "grad_norm": 0.7791340351104736, + "learning_rate": 7.100578690721105e-05, + "loss": 0.038, + "step": 39390 + }, + { + "epoch": 2.581406014544978, + "grad_norm": 0.35631856322288513, + "learning_rate": 7.09907810075519e-05, + "loss": 0.0212, + "step": 39400 + }, + { + "epoch": 2.582061193736487, + "grad_norm": 0.33646291494369507, + "learning_rate": 7.097577281237249e-05, + "loss": 0.0379, + "step": 39410 + }, + { + "epoch": 2.5827163729279956, + "grad_norm": 0.17377910017967224, + "learning_rate": 7.09607623233141e-05, + "loss": 0.0366, + "step": 39420 + }, + { + "epoch": 2.5833715521195044, + "grad_norm": 0.3006044924259186, + "learning_rate": 7.094574954201827e-05, + "loss": 0.0297, + "step": 39430 + }, + { + "epoch": 2.5840267313110137, + "grad_norm": 0.09696280211210251, + "learning_rate": 7.093073447012675e-05, + "loss": 0.0282, + "step": 39440 + }, + { + "epoch": 2.5846819105025225, + "grad_norm": 1.7133762836456299, + "learning_rate": 7.09157171092816e-05, + "loss": 0.0322, + "step": 39450 + }, + { + "epoch": 2.5853370896940313, + "grad_norm": 0.6065456867218018, + "learning_rate": 7.090069746112504e-05, + "loss": 0.0365, + "step": 39460 + }, + { + "epoch": 2.58599226888554, + "grad_norm": 0.4771261215209961, + "learning_rate": 7.088567552729965e-05, + "loss": 0.0432, + "step": 39470 + }, + { + "epoch": 2.586647448077049, + "grad_norm": 0.20970089733600616, + "learning_rate": 7.087065130944818e-05, + "loss": 0.035, + "step": 39480 + }, + { + "epoch": 2.587302627268558, + "grad_norm": 0.17776091396808624, + "learning_rate": 7.085562480921366e-05, + "loss": 0.0316, + "step": 39490 + }, + { + "epoch": 2.587957806460067, + "grad_norm": 0.2619125545024872, + "learning_rate": 7.084059602823937e-05, + "loss": 0.0321, + "step": 39500 + }, + { + "epoch": 2.588612985651576, + "grad_norm": 0.10934752225875854, + "learning_rate": 7.082556496816882e-05, + "loss": 0.0456, + "step": 39510 + }, + { + "epoch": 2.5892681648430846, + "grad_norm": 0.5431218147277832, + "learning_rate": 7.081053163064582e-05, + "loss": 0.0327, + "step": 39520 + }, + { + "epoch": 2.5899233440345935, + "grad_norm": 0.5373319983482361, + "learning_rate": 7.079549601731434e-05, + "loss": 0.0219, + "step": 39530 + }, + { + "epoch": 2.5905785232261023, + "grad_norm": 0.1933654546737671, + "learning_rate": 7.07804581298187e-05, + "loss": 0.0294, + "step": 39540 + }, + { + "epoch": 2.591233702417611, + "grad_norm": 0.4940401613712311, + "learning_rate": 7.07654179698034e-05, + "loss": 0.0352, + "step": 39550 + }, + { + "epoch": 2.59188888160912, + "grad_norm": 0.1585732400417328, + "learning_rate": 7.075037553891321e-05, + "loss": 0.0258, + "step": 39560 + }, + { + "epoch": 2.5925440608006287, + "grad_norm": 0.24843648076057434, + "learning_rate": 7.073533083879315e-05, + "loss": 0.0427, + "step": 39570 + }, + { + "epoch": 2.593199239992138, + "grad_norm": 0.23939253389835358, + "learning_rate": 7.072028387108849e-05, + "loss": 0.0264, + "step": 39580 + }, + { + "epoch": 2.593854419183647, + "grad_norm": 1.1327961683273315, + "learning_rate": 7.070523463744472e-05, + "loss": 0.0476, + "step": 39590 + }, + { + "epoch": 2.5945095983751556, + "grad_norm": 0.0713888555765152, + "learning_rate": 7.069018313950763e-05, + "loss": 0.0287, + "step": 39600 + }, + { + "epoch": 2.5951647775666644, + "grad_norm": 0.9816730618476868, + "learning_rate": 7.067512937892322e-05, + "loss": 0.0461, + "step": 39610 + }, + { + "epoch": 2.5958199567581732, + "grad_norm": 0.30386441946029663, + "learning_rate": 7.066007335733775e-05, + "loss": 0.0233, + "step": 39620 + }, + { + "epoch": 2.5964751359496825, + "grad_norm": 0.5608009099960327, + "learning_rate": 7.064501507639772e-05, + "loss": 0.0595, + "step": 39630 + }, + { + "epoch": 2.5971303151411913, + "grad_norm": 0.2214798480272293, + "learning_rate": 7.062995453774987e-05, + "loss": 0.0343, + "step": 39640 + }, + { + "epoch": 2.5977854943327, + "grad_norm": 0.8992705941200256, + "learning_rate": 7.061489174304121e-05, + "loss": 0.0435, + "step": 39650 + }, + { + "epoch": 2.598440673524209, + "grad_norm": 0.6957387328147888, + "learning_rate": 7.0599826693919e-05, + "loss": 0.0269, + "step": 39660 + }, + { + "epoch": 2.5990958527157177, + "grad_norm": 0.723651647567749, + "learning_rate": 7.05847593920307e-05, + "loss": 0.0406, + "step": 39670 + }, + { + "epoch": 2.5997510319072266, + "grad_norm": 0.4218110144138336, + "learning_rate": 7.056968983902406e-05, + "loss": 0.0772, + "step": 39680 + }, + { + "epoch": 2.6004062110987354, + "grad_norm": 0.8154491782188416, + "learning_rate": 7.055461803654706e-05, + "loss": 0.0636, + "step": 39690 + }, + { + "epoch": 2.601061390290244, + "grad_norm": 0.22859318554401398, + "learning_rate": 7.053954398624794e-05, + "loss": 0.0266, + "step": 39700 + }, + { + "epoch": 2.601716569481753, + "grad_norm": 0.3474915027618408, + "learning_rate": 7.052446768977518e-05, + "loss": 0.0435, + "step": 39710 + }, + { + "epoch": 2.6023717486732623, + "grad_norm": 0.8608506917953491, + "learning_rate": 7.050938914877748e-05, + "loss": 0.0331, + "step": 39720 + }, + { + "epoch": 2.603026927864771, + "grad_norm": 0.44972699880599976, + "learning_rate": 7.049430836490382e-05, + "loss": 0.0323, + "step": 39730 + }, + { + "epoch": 2.60368210705628, + "grad_norm": 0.1664380580186844, + "learning_rate": 7.047922533980342e-05, + "loss": 0.0188, + "step": 39740 + }, + { + "epoch": 2.6043372862477887, + "grad_norm": 1.1328258514404297, + "learning_rate": 7.046414007512571e-05, + "loss": 0.0503, + "step": 39750 + }, + { + "epoch": 2.6049924654392975, + "grad_norm": 0.5124687552452087, + "learning_rate": 7.044905257252042e-05, + "loss": 0.0272, + "step": 39760 + }, + { + "epoch": 2.605647644630807, + "grad_norm": 0.6538803577423096, + "learning_rate": 7.04339628336375e-05, + "loss": 0.0375, + "step": 39770 + }, + { + "epoch": 2.6063028238223156, + "grad_norm": 0.622025728225708, + "learning_rate": 7.041887086012711e-05, + "loss": 0.0268, + "step": 39780 + }, + { + "epoch": 2.6069580030138244, + "grad_norm": 0.8119916915893555, + "learning_rate": 7.040377665363969e-05, + "loss": 0.0233, + "step": 39790 + }, + { + "epoch": 2.6076131822053332, + "grad_norm": 0.7496776580810547, + "learning_rate": 7.038868021582594e-05, + "loss": 0.0374, + "step": 39800 + }, + { + "epoch": 2.608268361396842, + "grad_norm": 0.8010961413383484, + "learning_rate": 7.037358154833679e-05, + "loss": 0.0407, + "step": 39810 + }, + { + "epoch": 2.608923540588351, + "grad_norm": 0.16002638638019562, + "learning_rate": 7.035848065282339e-05, + "loss": 0.0208, + "step": 39820 + }, + { + "epoch": 2.6095787197798597, + "grad_norm": 0.5219075083732605, + "learning_rate": 7.034337753093714e-05, + "loss": 0.0307, + "step": 39830 + }, + { + "epoch": 2.6102338989713685, + "grad_norm": 0.25869762897491455, + "learning_rate": 7.032827218432972e-05, + "loss": 0.0321, + "step": 39840 + }, + { + "epoch": 2.6108890781628773, + "grad_norm": 0.07935555279254913, + "learning_rate": 7.031316461465302e-05, + "loss": 0.0234, + "step": 39850 + }, + { + "epoch": 2.6115442573543866, + "grad_norm": 0.5245959162712097, + "learning_rate": 7.029805482355915e-05, + "loss": 0.0617, + "step": 39860 + }, + { + "epoch": 2.6121994365458954, + "grad_norm": 1.3832545280456543, + "learning_rate": 7.028294281270055e-05, + "loss": 0.0689, + "step": 39870 + }, + { + "epoch": 2.612854615737404, + "grad_norm": 0.33623188734054565, + "learning_rate": 7.026782858372982e-05, + "loss": 0.0417, + "step": 39880 + }, + { + "epoch": 2.613509794928913, + "grad_norm": 0.351500928401947, + "learning_rate": 7.025271213829982e-05, + "loss": 0.0301, + "step": 39890 + }, + { + "epoch": 2.614164974120422, + "grad_norm": 0.5982696413993835, + "learning_rate": 7.023759347806366e-05, + "loss": 0.044, + "step": 39900 + }, + { + "epoch": 2.614820153311931, + "grad_norm": 0.7647559642791748, + "learning_rate": 7.022247260467472e-05, + "loss": 0.0395, + "step": 39910 + }, + { + "epoch": 2.61547533250344, + "grad_norm": 0.11290204524993896, + "learning_rate": 7.020734951978658e-05, + "loss": 0.0352, + "step": 39920 + }, + { + "epoch": 2.6161305116949487, + "grad_norm": 0.2210882306098938, + "learning_rate": 7.019222422505307e-05, + "loss": 0.0455, + "step": 39930 + }, + { + "epoch": 2.6167856908864575, + "grad_norm": 0.8463999032974243, + "learning_rate": 7.017709672212827e-05, + "loss": 0.0402, + "step": 39940 + }, + { + "epoch": 2.6174408700779663, + "grad_norm": 0.22693367302417755, + "learning_rate": 7.016196701266652e-05, + "loss": 0.0408, + "step": 39950 + }, + { + "epoch": 2.618096049269475, + "grad_norm": 0.37959203124046326, + "learning_rate": 7.014683509832238e-05, + "loss": 0.0427, + "step": 39960 + }, + { + "epoch": 2.618751228460984, + "grad_norm": 1.674920678138733, + "learning_rate": 7.013170098075063e-05, + "loss": 0.0642, + "step": 39970 + }, + { + "epoch": 2.6194064076524928, + "grad_norm": 0.3702631890773773, + "learning_rate": 7.011656466160632e-05, + "loss": 0.0386, + "step": 39980 + }, + { + "epoch": 2.6200615868440016, + "grad_norm": 1.2514886856079102, + "learning_rate": 7.010142614254475e-05, + "loss": 0.052, + "step": 39990 + }, + { + "epoch": 2.620716766035511, + "grad_norm": 0.5994771718978882, + "learning_rate": 7.008628542522147e-05, + "loss": 0.0501, + "step": 40000 + }, + { + "epoch": 2.6213719452270197, + "grad_norm": 0.6685777902603149, + "learning_rate": 7.007114251129218e-05, + "loss": 0.0541, + "step": 40010 + }, + { + "epoch": 2.6220271244185285, + "grad_norm": 0.27093833684921265, + "learning_rate": 7.005599740241293e-05, + "loss": 0.037, + "step": 40020 + }, + { + "epoch": 2.6226823036100373, + "grad_norm": 0.5917943120002747, + "learning_rate": 7.004085010023996e-05, + "loss": 0.0305, + "step": 40030 + }, + { + "epoch": 2.623337482801546, + "grad_norm": 0.5142666101455688, + "learning_rate": 7.002570060642976e-05, + "loss": 0.037, + "step": 40040 + }, + { + "epoch": 2.6239926619930554, + "grad_norm": 0.29519355297088623, + "learning_rate": 7.001054892263903e-05, + "loss": 0.0516, + "step": 40050 + }, + { + "epoch": 2.624647841184564, + "grad_norm": 0.34628117084503174, + "learning_rate": 6.999539505052477e-05, + "loss": 0.0286, + "step": 40060 + }, + { + "epoch": 2.625303020376073, + "grad_norm": 0.1077558770775795, + "learning_rate": 6.998023899174418e-05, + "loss": 0.0222, + "step": 40070 + }, + { + "epoch": 2.625958199567582, + "grad_norm": 0.21619637310504913, + "learning_rate": 6.996508074795467e-05, + "loss": 0.068, + "step": 40080 + }, + { + "epoch": 2.6266133787590906, + "grad_norm": 1.0414377450942993, + "learning_rate": 6.994992032081396e-05, + "loss": 0.0259, + "step": 40090 + }, + { + "epoch": 2.6272685579505994, + "grad_norm": 0.8062030076980591, + "learning_rate": 6.993475771197995e-05, + "loss": 0.0271, + "step": 40100 + }, + { + "epoch": 2.6279237371421083, + "grad_norm": 0.2517233192920685, + "learning_rate": 6.991959292311082e-05, + "loss": 0.0365, + "step": 40110 + }, + { + "epoch": 2.628578916333617, + "grad_norm": 0.15871410071849823, + "learning_rate": 6.990442595586495e-05, + "loss": 0.0501, + "step": 40120 + }, + { + "epoch": 2.629234095525126, + "grad_norm": 0.6024886965751648, + "learning_rate": 6.988925681190098e-05, + "loss": 0.0294, + "step": 40130 + }, + { + "epoch": 2.629889274716635, + "grad_norm": 0.28276780247688293, + "learning_rate": 6.987408549287778e-05, + "loss": 0.0662, + "step": 40140 + }, + { + "epoch": 2.630544453908144, + "grad_norm": 0.2995462119579315, + "learning_rate": 6.985891200045449e-05, + "loss": 0.051, + "step": 40150 + }, + { + "epoch": 2.6311996330996528, + "grad_norm": 0.3989596366882324, + "learning_rate": 6.984373633629045e-05, + "loss": 0.0311, + "step": 40160 + }, + { + "epoch": 2.6318548122911616, + "grad_norm": 0.3980557918548584, + "learning_rate": 6.98285585020452e-05, + "loss": 0.0589, + "step": 40170 + }, + { + "epoch": 2.6325099914826704, + "grad_norm": 0.4305124282836914, + "learning_rate": 6.981337849937864e-05, + "loss": 0.0283, + "step": 40180 + }, + { + "epoch": 2.6331651706741797, + "grad_norm": 0.0942913219332695, + "learning_rate": 6.979819632995078e-05, + "loss": 0.0439, + "step": 40190 + }, + { + "epoch": 2.6338203498656885, + "grad_norm": 0.235000878572464, + "learning_rate": 6.978301199542193e-05, + "loss": 0.0328, + "step": 40200 + }, + { + "epoch": 2.6344755290571973, + "grad_norm": 0.33211392164230347, + "learning_rate": 6.976782549745263e-05, + "loss": 0.0325, + "step": 40210 + }, + { + "epoch": 2.635130708248706, + "grad_norm": 0.42252933979034424, + "learning_rate": 6.975263683770367e-05, + "loss": 0.0619, + "step": 40220 + }, + { + "epoch": 2.635785887440215, + "grad_norm": 0.8747310042381287, + "learning_rate": 6.973744601783604e-05, + "loss": 0.0341, + "step": 40230 + }, + { + "epoch": 2.6364410666317237, + "grad_norm": 0.7407237887382507, + "learning_rate": 6.972225303951097e-05, + "loss": 0.0382, + "step": 40240 + }, + { + "epoch": 2.6370962458232325, + "grad_norm": 0.2914472818374634, + "learning_rate": 6.970705790438998e-05, + "loss": 0.0373, + "step": 40250 + }, + { + "epoch": 2.6377514250147414, + "grad_norm": 0.616797685623169, + "learning_rate": 6.969186061413477e-05, + "loss": 0.0222, + "step": 40260 + }, + { + "epoch": 2.63840660420625, + "grad_norm": 0.2691704332828522, + "learning_rate": 6.967666117040727e-05, + "loss": 0.0261, + "step": 40270 + }, + { + "epoch": 2.6390617833977594, + "grad_norm": 0.16708965599536896, + "learning_rate": 6.966145957486972e-05, + "loss": 0.0277, + "step": 40280 + }, + { + "epoch": 2.6397169625892682, + "grad_norm": 0.626811683177948, + "learning_rate": 6.964625582918449e-05, + "loss": 0.042, + "step": 40290 + }, + { + "epoch": 2.640372141780777, + "grad_norm": 0.2949146032333374, + "learning_rate": 6.963104993501425e-05, + "loss": 0.044, + "step": 40300 + }, + { + "epoch": 2.641027320972286, + "grad_norm": 1.028300404548645, + "learning_rate": 6.961584189402192e-05, + "loss": 0.0321, + "step": 40310 + }, + { + "epoch": 2.6416825001637947, + "grad_norm": 0.4741291403770447, + "learning_rate": 6.96006317078706e-05, + "loss": 0.0331, + "step": 40320 + }, + { + "epoch": 2.642337679355304, + "grad_norm": 0.44887977838516235, + "learning_rate": 6.958541937822367e-05, + "loss": 0.0441, + "step": 40330 + }, + { + "epoch": 2.6429928585468128, + "grad_norm": 0.19488392770290375, + "learning_rate": 6.957020490674473e-05, + "loss": 0.0288, + "step": 40340 + }, + { + "epoch": 2.6436480377383216, + "grad_norm": 0.9662615060806274, + "learning_rate": 6.95549882950976e-05, + "loss": 0.0415, + "step": 40350 + }, + { + "epoch": 2.6443032169298304, + "grad_norm": 0.08333837985992432, + "learning_rate": 6.953976954494635e-05, + "loss": 0.039, + "step": 40360 + }, + { + "epoch": 2.644958396121339, + "grad_norm": 0.30152225494384766, + "learning_rate": 6.952454865795528e-05, + "loss": 0.0227, + "step": 40370 + }, + { + "epoch": 2.645613575312848, + "grad_norm": 0.8712220788002014, + "learning_rate": 6.950932563578892e-05, + "loss": 0.0416, + "step": 40380 + }, + { + "epoch": 2.646268754504357, + "grad_norm": 1.0743310451507568, + "learning_rate": 6.949410048011206e-05, + "loss": 0.0533, + "step": 40390 + }, + { + "epoch": 2.6469239336958656, + "grad_norm": 0.24518699944019318, + "learning_rate": 6.947887319258966e-05, + "loss": 0.0446, + "step": 40400 + }, + { + "epoch": 2.6475791128873745, + "grad_norm": 0.5731245279312134, + "learning_rate": 6.946364377488696e-05, + "loss": 0.0238, + "step": 40410 + }, + { + "epoch": 2.6482342920788837, + "grad_norm": 0.14596547186374664, + "learning_rate": 6.944841222866947e-05, + "loss": 0.0268, + "step": 40420 + }, + { + "epoch": 2.6488894712703925, + "grad_norm": 0.22243355214595795, + "learning_rate": 6.943317855560284e-05, + "loss": 0.0319, + "step": 40430 + }, + { + "epoch": 2.6495446504619014, + "grad_norm": 0.14868539571762085, + "learning_rate": 6.941794275735302e-05, + "loss": 0.0321, + "step": 40440 + }, + { + "epoch": 2.65019982965341, + "grad_norm": 0.1605502963066101, + "learning_rate": 6.94027048355862e-05, + "loss": 0.0311, + "step": 40450 + }, + { + "epoch": 2.650855008844919, + "grad_norm": 0.4751152992248535, + "learning_rate": 6.938746479196871e-05, + "loss": 0.0217, + "step": 40460 + }, + { + "epoch": 2.6515101880364282, + "grad_norm": 0.8225913047790527, + "learning_rate": 6.937222262816724e-05, + "loss": 0.0387, + "step": 40470 + }, + { + "epoch": 2.652165367227937, + "grad_norm": 0.29666832089424133, + "learning_rate": 6.935697834584865e-05, + "loss": 0.0386, + "step": 40480 + }, + { + "epoch": 2.652820546419446, + "grad_norm": 1.9270795583724976, + "learning_rate": 6.934173194667998e-05, + "loss": 0.0284, + "step": 40490 + }, + { + "epoch": 2.6534757256109547, + "grad_norm": 0.09906996786594391, + "learning_rate": 6.93264834323286e-05, + "loss": 0.0435, + "step": 40500 + }, + { + "epoch": 2.6541309048024635, + "grad_norm": 0.6232884526252747, + "learning_rate": 6.931123280446206e-05, + "loss": 0.0349, + "step": 40510 + }, + { + "epoch": 2.6547860839939723, + "grad_norm": 0.6810073256492615, + "learning_rate": 6.929598006474811e-05, + "loss": 0.066, + "step": 40520 + }, + { + "epoch": 2.655441263185481, + "grad_norm": 0.32983988523483276, + "learning_rate": 6.92807252148548e-05, + "loss": 0.0459, + "step": 40530 + }, + { + "epoch": 2.65609644237699, + "grad_norm": 0.3525398373603821, + "learning_rate": 6.926546825645037e-05, + "loss": 0.0198, + "step": 40540 + }, + { + "epoch": 2.6567516215684988, + "grad_norm": 0.6682982444763184, + "learning_rate": 6.92502091912033e-05, + "loss": 0.0414, + "step": 40550 + }, + { + "epoch": 2.657406800760008, + "grad_norm": 0.734893798828125, + "learning_rate": 6.923494802078229e-05, + "loss": 0.0398, + "step": 40560 + }, + { + "epoch": 2.658061979951517, + "grad_norm": 0.14134326577186584, + "learning_rate": 6.921968474685628e-05, + "loss": 0.0374, + "step": 40570 + }, + { + "epoch": 2.6587171591430256, + "grad_norm": 0.125199556350708, + "learning_rate": 6.920441937109445e-05, + "loss": 0.0497, + "step": 40580 + }, + { + "epoch": 2.6593723383345345, + "grad_norm": 0.8894131183624268, + "learning_rate": 6.91891518951662e-05, + "loss": 0.0595, + "step": 40590 + }, + { + "epoch": 2.6600275175260433, + "grad_norm": 0.3124401569366455, + "learning_rate": 6.917388232074114e-05, + "loss": 0.0407, + "step": 40600 + }, + { + "epoch": 2.660682696717552, + "grad_norm": 0.32471394538879395, + "learning_rate": 6.915861064948914e-05, + "loss": 0.0216, + "step": 40610 + }, + { + "epoch": 2.6613378759090613, + "grad_norm": 0.5128856301307678, + "learning_rate": 6.914333688308025e-05, + "loss": 0.0399, + "step": 40620 + }, + { + "epoch": 2.66199305510057, + "grad_norm": 0.21146439015865326, + "learning_rate": 6.912806102318487e-05, + "loss": 0.0275, + "step": 40630 + }, + { + "epoch": 2.662648234292079, + "grad_norm": 0.45108461380004883, + "learning_rate": 6.911278307147347e-05, + "loss": 0.055, + "step": 40640 + }, + { + "epoch": 2.663303413483588, + "grad_norm": 0.3539188504219055, + "learning_rate": 6.909750302961684e-05, + "loss": 0.0423, + "step": 40650 + }, + { + "epoch": 2.6639585926750966, + "grad_norm": 0.35659903287887573, + "learning_rate": 6.9082220899286e-05, + "loss": 0.0244, + "step": 40660 + }, + { + "epoch": 2.6646137718666054, + "grad_norm": 0.3135240077972412, + "learning_rate": 6.906693668215219e-05, + "loss": 0.0525, + "step": 40670 + }, + { + "epoch": 2.6652689510581142, + "grad_norm": 0.8054590821266174, + "learning_rate": 6.905165037988683e-05, + "loss": 0.0428, + "step": 40680 + }, + { + "epoch": 2.665924130249623, + "grad_norm": 0.4097445011138916, + "learning_rate": 6.903636199416164e-05, + "loss": 0.0353, + "step": 40690 + }, + { + "epoch": 2.6665793094411323, + "grad_norm": 0.7074587941169739, + "learning_rate": 6.902107152664851e-05, + "loss": 0.0419, + "step": 40700 + }, + { + "epoch": 2.667234488632641, + "grad_norm": 0.1331399828195572, + "learning_rate": 6.90057789790196e-05, + "loss": 0.0484, + "step": 40710 + }, + { + "epoch": 2.66788966782415, + "grad_norm": 0.486510306596756, + "learning_rate": 6.899048435294728e-05, + "loss": 0.0557, + "step": 40720 + }, + { + "epoch": 2.6685448470156587, + "grad_norm": 0.5174886584281921, + "learning_rate": 6.897518765010415e-05, + "loss": 0.0442, + "step": 40730 + }, + { + "epoch": 2.6692000262071676, + "grad_norm": 0.4466499090194702, + "learning_rate": 6.895988887216302e-05, + "loss": 0.0392, + "step": 40740 + }, + { + "epoch": 2.6698552053986764, + "grad_norm": 0.19840465486049652, + "learning_rate": 6.894458802079694e-05, + "loss": 0.0405, + "step": 40750 + }, + { + "epoch": 2.6705103845901856, + "grad_norm": 0.43899306654930115, + "learning_rate": 6.892928509767922e-05, + "loss": 0.0601, + "step": 40760 + }, + { + "epoch": 2.6711655637816945, + "grad_norm": 1.0231369733810425, + "learning_rate": 6.891398010448333e-05, + "loss": 0.0401, + "step": 40770 + }, + { + "epoch": 2.6718207429732033, + "grad_norm": 0.2554596960544586, + "learning_rate": 6.8898673042883e-05, + "loss": 0.0448, + "step": 40780 + }, + { + "epoch": 2.672475922164712, + "grad_norm": 0.6584427952766418, + "learning_rate": 6.888336391455222e-05, + "loss": 0.0383, + "step": 40790 + }, + { + "epoch": 2.673131101356221, + "grad_norm": 1.1551753282546997, + "learning_rate": 6.886805272116513e-05, + "loss": 0.0426, + "step": 40800 + }, + { + "epoch": 2.6737862805477297, + "grad_norm": 0.4807511568069458, + "learning_rate": 6.885273946439617e-05, + "loss": 0.0354, + "step": 40810 + }, + { + "epoch": 2.6744414597392385, + "grad_norm": 0.3185623288154602, + "learning_rate": 6.883742414591998e-05, + "loss": 0.04, + "step": 40820 + }, + { + "epoch": 2.6750966389307473, + "grad_norm": 0.43629178404808044, + "learning_rate": 6.88221067674114e-05, + "loss": 0.0727, + "step": 40830 + }, + { + "epoch": 2.6757518181222566, + "grad_norm": 0.427524209022522, + "learning_rate": 6.88067873305455e-05, + "loss": 0.0373, + "step": 40840 + }, + { + "epoch": 2.6764069973137654, + "grad_norm": 0.31159836053848267, + "learning_rate": 6.879146583699765e-05, + "loss": 0.0396, + "step": 40850 + }, + { + "epoch": 2.6770621765052742, + "grad_norm": 0.42845121026039124, + "learning_rate": 6.877614228844334e-05, + "loss": 0.0301, + "step": 40860 + }, + { + "epoch": 2.677717355696783, + "grad_norm": 0.7973902821540833, + "learning_rate": 6.876081668655832e-05, + "loss": 0.0389, + "step": 40870 + }, + { + "epoch": 2.678372534888292, + "grad_norm": 0.34283027052879333, + "learning_rate": 6.874548903301861e-05, + "loss": 0.0363, + "step": 40880 + }, + { + "epoch": 2.6790277140798007, + "grad_norm": 0.19334733486175537, + "learning_rate": 6.873015932950039e-05, + "loss": 0.0451, + "step": 40890 + }, + { + "epoch": 2.67968289327131, + "grad_norm": 0.3769339621067047, + "learning_rate": 6.871482757768012e-05, + "loss": 0.0334, + "step": 40900 + }, + { + "epoch": 2.6803380724628187, + "grad_norm": 0.6600518822669983, + "learning_rate": 6.869949377923442e-05, + "loss": 0.0444, + "step": 40910 + }, + { + "epoch": 2.6809932516543276, + "grad_norm": 0.8293270468711853, + "learning_rate": 6.868415793584022e-05, + "loss": 0.0269, + "step": 40920 + }, + { + "epoch": 2.6816484308458364, + "grad_norm": 0.32277774810791016, + "learning_rate": 6.86688200491746e-05, + "loss": 0.0419, + "step": 40930 + }, + { + "epoch": 2.682303610037345, + "grad_norm": 0.16422846913337708, + "learning_rate": 6.865348012091488e-05, + "loss": 0.0481, + "step": 40940 + }, + { + "epoch": 2.682958789228854, + "grad_norm": 0.43151843547821045, + "learning_rate": 6.86381381527386e-05, + "loss": 0.0529, + "step": 40950 + }, + { + "epoch": 2.683613968420363, + "grad_norm": 0.1449091136455536, + "learning_rate": 6.862279414632359e-05, + "loss": 0.0228, + "step": 40960 + }, + { + "epoch": 2.6842691476118716, + "grad_norm": 0.13569092750549316, + "learning_rate": 6.860744810334778e-05, + "loss": 0.0491, + "step": 40970 + }, + { + "epoch": 2.684924326803381, + "grad_norm": 0.13453064858913422, + "learning_rate": 6.859210002548943e-05, + "loss": 0.0556, + "step": 40980 + }, + { + "epoch": 2.6855795059948897, + "grad_norm": 0.5059736967086792, + "learning_rate": 6.8576749914427e-05, + "loss": 0.0387, + "step": 40990 + }, + { + "epoch": 2.6862346851863985, + "grad_norm": 0.1967654526233673, + "learning_rate": 6.85613977718391e-05, + "loss": 0.0352, + "step": 41000 + }, + { + "epoch": 2.6868898643779073, + "grad_norm": 0.1778343915939331, + "learning_rate": 6.854604359940464e-05, + "loss": 0.023, + "step": 41010 + }, + { + "epoch": 2.687545043569416, + "grad_norm": 0.34532633423805237, + "learning_rate": 6.853068739880276e-05, + "loss": 0.0303, + "step": 41020 + }, + { + "epoch": 2.688200222760925, + "grad_norm": 0.12730489671230316, + "learning_rate": 6.851532917171276e-05, + "loss": 0.0423, + "step": 41030 + }, + { + "epoch": 2.688855401952434, + "grad_norm": 0.24381662905216217, + "learning_rate": 6.849996891981418e-05, + "loss": 0.033, + "step": 41040 + }, + { + "epoch": 2.689510581143943, + "grad_norm": 0.8503773808479309, + "learning_rate": 6.848460664478684e-05, + "loss": 0.0452, + "step": 41050 + }, + { + "epoch": 2.690165760335452, + "grad_norm": 0.2329435795545578, + "learning_rate": 6.846924234831069e-05, + "loss": 0.0181, + "step": 41060 + }, + { + "epoch": 2.6908209395269607, + "grad_norm": 0.08846666663885117, + "learning_rate": 6.845387603206597e-05, + "loss": 0.0481, + "step": 41070 + }, + { + "epoch": 2.6914761187184695, + "grad_norm": 0.3691621720790863, + "learning_rate": 6.843850769773312e-05, + "loss": 0.0264, + "step": 41080 + }, + { + "epoch": 2.6921312979099783, + "grad_norm": 0.2980443239212036, + "learning_rate": 6.842313734699278e-05, + "loss": 0.0662, + "step": 41090 + }, + { + "epoch": 2.692786477101487, + "grad_norm": 0.21739132702350616, + "learning_rate": 6.840776498152584e-05, + "loss": 0.029, + "step": 41100 + }, + { + "epoch": 2.693441656292996, + "grad_norm": 0.3263311982154846, + "learning_rate": 6.83923906030134e-05, + "loss": 0.0663, + "step": 41110 + }, + { + "epoch": 2.6940968354845047, + "grad_norm": 0.589346170425415, + "learning_rate": 6.837701421313677e-05, + "loss": 0.0192, + "step": 41120 + }, + { + "epoch": 2.694752014676014, + "grad_norm": 0.21804815530776978, + "learning_rate": 6.83616358135775e-05, + "loss": 0.0855, + "step": 41130 + }, + { + "epoch": 2.695407193867523, + "grad_norm": 0.2624220550060272, + "learning_rate": 6.834625540601734e-05, + "loss": 0.0429, + "step": 41140 + }, + { + "epoch": 2.6960623730590316, + "grad_norm": 1.0178886651992798, + "learning_rate": 6.833087299213829e-05, + "loss": 0.0342, + "step": 41150 + }, + { + "epoch": 2.6967175522505404, + "grad_norm": 0.16031302511692047, + "learning_rate": 6.83154885736225e-05, + "loss": 0.0271, + "step": 41160 + }, + { + "epoch": 2.6973727314420493, + "grad_norm": 0.12414176017045975, + "learning_rate": 6.830010215215244e-05, + "loss": 0.0245, + "step": 41170 + }, + { + "epoch": 2.6980279106335585, + "grad_norm": 0.3430415987968445, + "learning_rate": 6.828471372941072e-05, + "loss": 0.0334, + "step": 41180 + }, + { + "epoch": 2.6986830898250673, + "grad_norm": 1.020892858505249, + "learning_rate": 6.826932330708018e-05, + "loss": 0.048, + "step": 41190 + }, + { + "epoch": 2.699338269016576, + "grad_norm": 0.527275800704956, + "learning_rate": 6.825393088684393e-05, + "loss": 0.0517, + "step": 41200 + }, + { + "epoch": 2.699993448208085, + "grad_norm": 0.7996773719787598, + "learning_rate": 6.823853647038524e-05, + "loss": 0.0292, + "step": 41210 + }, + { + "epoch": 2.7006486273995938, + "grad_norm": 0.0908375158905983, + "learning_rate": 6.822314005938762e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 2.7013038065911026, + "grad_norm": 0.10477752238512039, + "learning_rate": 6.820774165553479e-05, + "loss": 0.0349, + "step": 41230 + }, + { + "epoch": 2.7019589857826114, + "grad_norm": 0.09551136195659637, + "learning_rate": 6.819234126051072e-05, + "loss": 0.0679, + "step": 41240 + }, + { + "epoch": 2.70261416497412, + "grad_norm": 0.4221500754356384, + "learning_rate": 6.817693887599956e-05, + "loss": 0.0514, + "step": 41250 + }, + { + "epoch": 2.703269344165629, + "grad_norm": 0.6023364067077637, + "learning_rate": 6.816153450368568e-05, + "loss": 0.034, + "step": 41260 + }, + { + "epoch": 2.7039245233571383, + "grad_norm": 0.32051292061805725, + "learning_rate": 6.814612814525372e-05, + "loss": 0.0441, + "step": 41270 + }, + { + "epoch": 2.704579702548647, + "grad_norm": 0.1188526451587677, + "learning_rate": 6.813071980238846e-05, + "loss": 0.0263, + "step": 41280 + }, + { + "epoch": 2.705234881740156, + "grad_norm": 0.6318060159683228, + "learning_rate": 6.811530947677492e-05, + "loss": 0.0388, + "step": 41290 + }, + { + "epoch": 2.7058900609316647, + "grad_norm": 0.14760079979896545, + "learning_rate": 6.809989717009839e-05, + "loss": 0.034, + "step": 41300 + }, + { + "epoch": 2.7065452401231735, + "grad_norm": 0.17313112318515778, + "learning_rate": 6.808448288404431e-05, + "loss": 0.0233, + "step": 41310 + }, + { + "epoch": 2.707200419314683, + "grad_norm": 0.06142563372850418, + "learning_rate": 6.806906662029838e-05, + "loss": 0.0376, + "step": 41320 + }, + { + "epoch": 2.7078555985061916, + "grad_norm": 0.398337721824646, + "learning_rate": 6.805364838054647e-05, + "loss": 0.0364, + "step": 41330 + }, + { + "epoch": 2.7085107776977004, + "grad_norm": 0.1914239078760147, + "learning_rate": 6.803822816647474e-05, + "loss": 0.0292, + "step": 41340 + }, + { + "epoch": 2.7091659568892092, + "grad_norm": 0.6446617841720581, + "learning_rate": 6.802280597976949e-05, + "loss": 0.0619, + "step": 41350 + }, + { + "epoch": 2.709821136080718, + "grad_norm": 0.49118512868881226, + "learning_rate": 6.800738182211727e-05, + "loss": 0.041, + "step": 41360 + }, + { + "epoch": 2.710476315272227, + "grad_norm": 0.5035281777381897, + "learning_rate": 6.799195569520485e-05, + "loss": 0.0454, + "step": 41370 + }, + { + "epoch": 2.7111314944637357, + "grad_norm": 0.6647999286651611, + "learning_rate": 6.79765276007192e-05, + "loss": 0.0502, + "step": 41380 + }, + { + "epoch": 2.7117866736552445, + "grad_norm": 0.5318121910095215, + "learning_rate": 6.796109754034753e-05, + "loss": 0.053, + "step": 41390 + }, + { + "epoch": 2.7124418528467533, + "grad_norm": 0.28492069244384766, + "learning_rate": 6.794566551577724e-05, + "loss": 0.0695, + "step": 41400 + }, + { + "epoch": 2.7130970320382626, + "grad_norm": 1.2063384056091309, + "learning_rate": 6.793023152869594e-05, + "loss": 0.0476, + "step": 41410 + }, + { + "epoch": 2.7137522112297714, + "grad_norm": 0.36765211820602417, + "learning_rate": 6.791479558079147e-05, + "loss": 0.0408, + "step": 41420 + }, + { + "epoch": 2.71440739042128, + "grad_norm": 0.3639516532421112, + "learning_rate": 6.789935767375191e-05, + "loss": 0.0545, + "step": 41430 + }, + { + "epoch": 2.715062569612789, + "grad_norm": 0.29531508684158325, + "learning_rate": 6.78839178092655e-05, + "loss": 0.0285, + "step": 41440 + }, + { + "epoch": 2.715717748804298, + "grad_norm": 0.5045745968818665, + "learning_rate": 6.786847598902072e-05, + "loss": 0.0506, + "step": 41450 + }, + { + "epoch": 2.716372927995807, + "grad_norm": 0.15497033298015594, + "learning_rate": 6.785303221470627e-05, + "loss": 0.0442, + "step": 41460 + }, + { + "epoch": 2.717028107187316, + "grad_norm": 0.18149587512016296, + "learning_rate": 6.783758648801108e-05, + "loss": 0.0334, + "step": 41470 + }, + { + "epoch": 2.7176832863788247, + "grad_norm": 0.42376264929771423, + "learning_rate": 6.782213881062422e-05, + "loss": 0.0259, + "step": 41480 + }, + { + "epoch": 2.7183384655703335, + "grad_norm": 0.4004310667514801, + "learning_rate": 6.780668918423508e-05, + "loss": 0.0419, + "step": 41490 + }, + { + "epoch": 2.7189936447618424, + "grad_norm": 1.545886516571045, + "learning_rate": 6.779123761053317e-05, + "loss": 0.0387, + "step": 41500 + }, + { + "epoch": 2.719648823953351, + "grad_norm": 0.12024514377117157, + "learning_rate": 6.777578409120826e-05, + "loss": 0.0411, + "step": 41510 + }, + { + "epoch": 2.72030400314486, + "grad_norm": 0.5247271060943604, + "learning_rate": 6.776032862795033e-05, + "loss": 0.0379, + "step": 41520 + }, + { + "epoch": 2.720959182336369, + "grad_norm": 1.0740355253219604, + "learning_rate": 6.774487122244956e-05, + "loss": 0.042, + "step": 41530 + }, + { + "epoch": 2.7216143615278776, + "grad_norm": 0.8139304518699646, + "learning_rate": 6.772941187639637e-05, + "loss": 0.0413, + "step": 41540 + }, + { + "epoch": 2.722269540719387, + "grad_norm": 0.21343772113323212, + "learning_rate": 6.771395059148134e-05, + "loss": 0.0636, + "step": 41550 + }, + { + "epoch": 2.7229247199108957, + "grad_norm": 0.5912080407142639, + "learning_rate": 6.76984873693953e-05, + "loss": 0.0331, + "step": 41560 + }, + { + "epoch": 2.7235798991024045, + "grad_norm": 0.35725393891334534, + "learning_rate": 6.768302221182931e-05, + "loss": 0.0406, + "step": 41570 + }, + { + "epoch": 2.7242350782939133, + "grad_norm": 1.1330331563949585, + "learning_rate": 6.766755512047457e-05, + "loss": 0.0487, + "step": 41580 + }, + { + "epoch": 2.724890257485422, + "grad_norm": 0.8283748030662537, + "learning_rate": 6.765208609702259e-05, + "loss": 0.0382, + "step": 41590 + }, + { + "epoch": 2.7255454366769314, + "grad_norm": 0.607173502445221, + "learning_rate": 6.763661514316499e-05, + "loss": 0.0383, + "step": 41600 + }, + { + "epoch": 2.72620061586844, + "grad_norm": 0.11240563541650772, + "learning_rate": 6.76211422605937e-05, + "loss": 0.0539, + "step": 41610 + }, + { + "epoch": 2.726855795059949, + "grad_norm": 0.7759227752685547, + "learning_rate": 6.760566745100076e-05, + "loss": 0.0468, + "step": 41620 + }, + { + "epoch": 2.727510974251458, + "grad_norm": 0.12550531327724457, + "learning_rate": 6.75901907160785e-05, + "loss": 0.0246, + "step": 41630 + }, + { + "epoch": 2.7281661534429666, + "grad_norm": 0.2928204834461212, + "learning_rate": 6.757471205751943e-05, + "loss": 0.0419, + "step": 41640 + }, + { + "epoch": 2.7288213326344755, + "grad_norm": 0.8206844925880432, + "learning_rate": 6.75592314770163e-05, + "loss": 0.0361, + "step": 41650 + }, + { + "epoch": 2.7294765118259843, + "grad_norm": 0.3453750014305115, + "learning_rate": 6.7543748976262e-05, + "loss": 0.0441, + "step": 41660 + }, + { + "epoch": 2.730131691017493, + "grad_norm": 0.5953071117401123, + "learning_rate": 6.752826455694968e-05, + "loss": 0.052, + "step": 41670 + }, + { + "epoch": 2.730786870209002, + "grad_norm": 0.18974347412586212, + "learning_rate": 6.751277822077271e-05, + "loss": 0.0268, + "step": 41680 + }, + { + "epoch": 2.731442049400511, + "grad_norm": 0.3796749711036682, + "learning_rate": 6.749728996942465e-05, + "loss": 0.0411, + "step": 41690 + }, + { + "epoch": 2.73209722859202, + "grad_norm": 0.3036820590496063, + "learning_rate": 6.748179980459924e-05, + "loss": 0.0263, + "step": 41700 + }, + { + "epoch": 2.732752407783529, + "grad_norm": 0.3401297330856323, + "learning_rate": 6.746630772799052e-05, + "loss": 0.0242, + "step": 41710 + }, + { + "epoch": 2.7334075869750376, + "grad_norm": 0.21938256919384003, + "learning_rate": 6.745081374129262e-05, + "loss": 0.032, + "step": 41720 + }, + { + "epoch": 2.7340627661665464, + "grad_norm": 0.7118943929672241, + "learning_rate": 6.743531784619998e-05, + "loss": 0.0363, + "step": 41730 + }, + { + "epoch": 2.7347179453580557, + "grad_norm": 0.4930303692817688, + "learning_rate": 6.741982004440719e-05, + "loss": 0.0341, + "step": 41740 + }, + { + "epoch": 2.7353731245495645, + "grad_norm": 0.1643223762512207, + "learning_rate": 6.740432033760907e-05, + "loss": 0.0249, + "step": 41750 + }, + { + "epoch": 2.7360283037410733, + "grad_norm": 0.9881799817085266, + "learning_rate": 6.738881872750066e-05, + "loss": 0.0606, + "step": 41760 + }, + { + "epoch": 2.736683482932582, + "grad_norm": 0.6968907117843628, + "learning_rate": 6.737331521577715e-05, + "loss": 0.0294, + "step": 41770 + }, + { + "epoch": 2.737338662124091, + "grad_norm": 1.0228337049484253, + "learning_rate": 6.735780980413403e-05, + "loss": 0.029, + "step": 41780 + }, + { + "epoch": 2.7379938413155998, + "grad_norm": 0.10273311287164688, + "learning_rate": 6.734230249426692e-05, + "loss": 0.0444, + "step": 41790 + }, + { + "epoch": 2.7386490205071086, + "grad_norm": 0.25334474444389343, + "learning_rate": 6.732679328787168e-05, + "loss": 0.0669, + "step": 41800 + }, + { + "epoch": 2.7393041996986174, + "grad_norm": 0.4605657458305359, + "learning_rate": 6.731128218664438e-05, + "loss": 0.0366, + "step": 41810 + }, + { + "epoch": 2.739959378890126, + "grad_norm": 0.5770133137702942, + "learning_rate": 6.729576919228129e-05, + "loss": 0.0249, + "step": 41820 + }, + { + "epoch": 2.7406145580816355, + "grad_norm": 1.0939728021621704, + "learning_rate": 6.728025430647888e-05, + "loss": 0.0229, + "step": 41830 + }, + { + "epoch": 2.7412697372731443, + "grad_norm": 1.1784013509750366, + "learning_rate": 6.726473753093383e-05, + "loss": 0.0538, + "step": 41840 + }, + { + "epoch": 2.741924916464653, + "grad_norm": 0.25005507469177246, + "learning_rate": 6.724921886734305e-05, + "loss": 0.0412, + "step": 41850 + }, + { + "epoch": 2.742580095656162, + "grad_norm": 0.2849162518978119, + "learning_rate": 6.723369831740363e-05, + "loss": 0.0325, + "step": 41860 + }, + { + "epoch": 2.7432352748476707, + "grad_norm": 0.5581310391426086, + "learning_rate": 6.721817588281288e-05, + "loss": 0.0516, + "step": 41870 + }, + { + "epoch": 2.74389045403918, + "grad_norm": 0.08343157917261124, + "learning_rate": 6.720265156526828e-05, + "loss": 0.033, + "step": 41880 + }, + { + "epoch": 2.744545633230689, + "grad_norm": 0.17157503962516785, + "learning_rate": 6.718712536646758e-05, + "loss": 0.0542, + "step": 41890 + }, + { + "epoch": 2.7452008124221976, + "grad_norm": 0.11397214978933334, + "learning_rate": 6.71715972881087e-05, + "loss": 0.0283, + "step": 41900 + }, + { + "epoch": 2.7458559916137064, + "grad_norm": 0.5116810202598572, + "learning_rate": 6.715606733188974e-05, + "loss": 0.0338, + "step": 41910 + }, + { + "epoch": 2.7465111708052152, + "grad_norm": 0.23985865712165833, + "learning_rate": 6.714053549950909e-05, + "loss": 0.0313, + "step": 41920 + }, + { + "epoch": 2.747166349996724, + "grad_norm": 2.6696979999542236, + "learning_rate": 6.71250017926652e-05, + "loss": 0.0349, + "step": 41930 + }, + { + "epoch": 2.747821529188233, + "grad_norm": 0.4364411532878876, + "learning_rate": 6.71094662130569e-05, + "loss": 0.0465, + "step": 41940 + }, + { + "epoch": 2.7484767083797417, + "grad_norm": 0.6911253333091736, + "learning_rate": 6.709392876238307e-05, + "loss": 0.0417, + "step": 41950 + }, + { + "epoch": 2.7491318875712505, + "grad_norm": 0.07875635474920273, + "learning_rate": 6.707838944234289e-05, + "loss": 0.0263, + "step": 41960 + }, + { + "epoch": 2.7497870667627597, + "grad_norm": 0.20749878883361816, + "learning_rate": 6.706284825463574e-05, + "loss": 0.0242, + "step": 41970 + }, + { + "epoch": 2.7504422459542686, + "grad_norm": 0.706190288066864, + "learning_rate": 6.704730520096114e-05, + "loss": 0.0249, + "step": 41980 + }, + { + "epoch": 2.7510974251457774, + "grad_norm": 0.2082131803035736, + "learning_rate": 6.703176028301888e-05, + "loss": 0.046, + "step": 41990 + }, + { + "epoch": 2.751752604337286, + "grad_norm": 0.1320333480834961, + "learning_rate": 6.701621350250892e-05, + "loss": 0.0481, + "step": 42000 + }, + { + "epoch": 2.752407783528795, + "grad_norm": 0.2922849953174591, + "learning_rate": 6.700066486113144e-05, + "loss": 0.0394, + "step": 42010 + }, + { + "epoch": 2.7530629627203043, + "grad_norm": 0.3293292820453644, + "learning_rate": 6.69851143605868e-05, + "loss": 0.052, + "step": 42020 + }, + { + "epoch": 2.753718141911813, + "grad_norm": 0.14844398200511932, + "learning_rate": 6.696956200257561e-05, + "loss": 0.0576, + "step": 42030 + }, + { + "epoch": 2.754373321103322, + "grad_norm": 0.4719153940677643, + "learning_rate": 6.695400778879863e-05, + "loss": 0.0577, + "step": 42040 + }, + { + "epoch": 2.7550285002948307, + "grad_norm": 0.21880705654621124, + "learning_rate": 6.693845172095683e-05, + "loss": 0.0419, + "step": 42050 + }, + { + "epoch": 2.7556836794863395, + "grad_norm": 0.8505707383155823, + "learning_rate": 6.692289380075142e-05, + "loss": 0.0478, + "step": 42060 + }, + { + "epoch": 2.7563388586778483, + "grad_norm": 0.219346284866333, + "learning_rate": 6.690733402988379e-05, + "loss": 0.0264, + "step": 42070 + }, + { + "epoch": 2.756994037869357, + "grad_norm": 0.4478501081466675, + "learning_rate": 6.689177241005553e-05, + "loss": 0.0293, + "step": 42080 + }, + { + "epoch": 2.757649217060866, + "grad_norm": 0.3517073094844818, + "learning_rate": 6.687620894296844e-05, + "loss": 0.0333, + "step": 42090 + }, + { + "epoch": 2.758304396252375, + "grad_norm": 0.13446994125843048, + "learning_rate": 6.686064363032451e-05, + "loss": 0.0461, + "step": 42100 + }, + { + "epoch": 2.758959575443884, + "grad_norm": 0.16623836755752563, + "learning_rate": 6.684507647382596e-05, + "loss": 0.0304, + "step": 42110 + }, + { + "epoch": 2.759614754635393, + "grad_norm": 0.4032348394393921, + "learning_rate": 6.682950747517513e-05, + "loss": 0.0436, + "step": 42120 + }, + { + "epoch": 2.7602699338269017, + "grad_norm": 0.4848393499851227, + "learning_rate": 6.68139366360747e-05, + "loss": 0.0296, + "step": 42130 + }, + { + "epoch": 2.7609251130184105, + "grad_norm": 0.17804661393165588, + "learning_rate": 6.679836395822744e-05, + "loss": 0.0559, + "step": 42140 + }, + { + "epoch": 2.7615802922099193, + "grad_norm": 0.6917383074760437, + "learning_rate": 6.678278944333633e-05, + "loss": 0.0389, + "step": 42150 + }, + { + "epoch": 2.7622354714014286, + "grad_norm": 0.4147471487522125, + "learning_rate": 6.676721309310462e-05, + "loss": 0.0336, + "step": 42160 + }, + { + "epoch": 2.7628906505929374, + "grad_norm": 0.3821680247783661, + "learning_rate": 6.67516349092357e-05, + "loss": 0.0389, + "step": 42170 + }, + { + "epoch": 2.763545829784446, + "grad_norm": 0.6794207096099854, + "learning_rate": 6.673605489343315e-05, + "loss": 0.025, + "step": 42180 + }, + { + "epoch": 2.764201008975955, + "grad_norm": 0.5409008264541626, + "learning_rate": 6.672047304740081e-05, + "loss": 0.0325, + "step": 42190 + }, + { + "epoch": 2.764856188167464, + "grad_norm": 0.5828886032104492, + "learning_rate": 6.670488937284268e-05, + "loss": 0.0295, + "step": 42200 + }, + { + "epoch": 2.7655113673589726, + "grad_norm": 0.8747169971466064, + "learning_rate": 6.668930387146294e-05, + "loss": 0.0321, + "step": 42210 + }, + { + "epoch": 2.7661665465504814, + "grad_norm": 0.382583349943161, + "learning_rate": 6.667371654496605e-05, + "loss": 0.027, + "step": 42220 + }, + { + "epoch": 2.7668217257419903, + "grad_norm": 0.3234272599220276, + "learning_rate": 6.665812739505659e-05, + "loss": 0.0324, + "step": 42230 + }, + { + "epoch": 2.767476904933499, + "grad_norm": 0.257742702960968, + "learning_rate": 6.664253642343935e-05, + "loss": 0.034, + "step": 42240 + }, + { + "epoch": 2.7681320841250083, + "grad_norm": 0.471204936504364, + "learning_rate": 6.662694363181935e-05, + "loss": 0.0346, + "step": 42250 + }, + { + "epoch": 2.768787263316517, + "grad_norm": 0.3342815041542053, + "learning_rate": 6.66113490219018e-05, + "loss": 0.0412, + "step": 42260 + }, + { + "epoch": 2.769442442508026, + "grad_norm": 0.367323637008667, + "learning_rate": 6.659575259539207e-05, + "loss": 0.0224, + "step": 42270 + }, + { + "epoch": 2.7700976216995348, + "grad_norm": 0.26990604400634766, + "learning_rate": 6.65801543539958e-05, + "loss": 0.0286, + "step": 42280 + }, + { + "epoch": 2.7707528008910436, + "grad_norm": 0.21984046697616577, + "learning_rate": 6.65645542994188e-05, + "loss": 0.0398, + "step": 42290 + }, + { + "epoch": 2.771407980082553, + "grad_norm": 1.0209803581237793, + "learning_rate": 6.654895243336702e-05, + "loss": 0.0315, + "step": 42300 + }, + { + "epoch": 2.7720631592740617, + "grad_norm": 1.0578950643539429, + "learning_rate": 6.65333487575467e-05, + "loss": 0.0426, + "step": 42310 + }, + { + "epoch": 2.7727183384655705, + "grad_norm": 0.11038186401128769, + "learning_rate": 6.651774327366422e-05, + "loss": 0.0299, + "step": 42320 + }, + { + "epoch": 2.7733735176570793, + "grad_norm": 0.387450248003006, + "learning_rate": 6.650213598342613e-05, + "loss": 0.0459, + "step": 42330 + }, + { + "epoch": 2.774028696848588, + "grad_norm": 0.08394956588745117, + "learning_rate": 6.648652688853931e-05, + "loss": 0.0345, + "step": 42340 + }, + { + "epoch": 2.774683876040097, + "grad_norm": 0.498238742351532, + "learning_rate": 6.647091599071066e-05, + "loss": 0.036, + "step": 42350 + }, + { + "epoch": 2.7753390552316057, + "grad_norm": 0.6306346654891968, + "learning_rate": 6.645530329164743e-05, + "loss": 0.0289, + "step": 42360 + }, + { + "epoch": 2.7759942344231145, + "grad_norm": 1.3050777912139893, + "learning_rate": 6.643968879305694e-05, + "loss": 0.0402, + "step": 42370 + }, + { + "epoch": 2.7766494136146234, + "grad_norm": 0.08103502541780472, + "learning_rate": 6.642407249664683e-05, + "loss": 0.0788, + "step": 42380 + }, + { + "epoch": 2.7773045928061326, + "grad_norm": 0.5750315189361572, + "learning_rate": 6.640845440412483e-05, + "loss": 0.047, + "step": 42390 + }, + { + "epoch": 2.7779597719976414, + "grad_norm": 0.7304487824440002, + "learning_rate": 6.639283451719893e-05, + "loss": 0.0641, + "step": 42400 + }, + { + "epoch": 2.7786149511891503, + "grad_norm": 0.4294642508029938, + "learning_rate": 6.637721283757729e-05, + "loss": 0.0695, + "step": 42410 + }, + { + "epoch": 2.779270130380659, + "grad_norm": 0.3097051680088043, + "learning_rate": 6.636158936696828e-05, + "loss": 0.0445, + "step": 42420 + }, + { + "epoch": 2.779925309572168, + "grad_norm": 0.28653475642204285, + "learning_rate": 6.634596410708047e-05, + "loss": 0.056, + "step": 42430 + }, + { + "epoch": 2.780580488763677, + "grad_norm": 0.44258883595466614, + "learning_rate": 6.633033705962257e-05, + "loss": 0.0387, + "step": 42440 + }, + { + "epoch": 2.781235667955186, + "grad_norm": 0.4154355227947235, + "learning_rate": 6.631470822630359e-05, + "loss": 0.0748, + "step": 42450 + }, + { + "epoch": 2.7818908471466948, + "grad_norm": 0.5547669529914856, + "learning_rate": 6.629907760883263e-05, + "loss": 0.0287, + "step": 42460 + }, + { + "epoch": 2.7825460263382036, + "grad_norm": 1.2200678586959839, + "learning_rate": 6.628344520891905e-05, + "loss": 0.0451, + "step": 42470 + }, + { + "epoch": 2.7832012055297124, + "grad_norm": 0.36746567487716675, + "learning_rate": 6.62678110282724e-05, + "loss": 0.069, + "step": 42480 + }, + { + "epoch": 2.783856384721221, + "grad_norm": 0.09135752171278, + "learning_rate": 6.625217506860237e-05, + "loss": 0.027, + "step": 42490 + }, + { + "epoch": 2.78451156391273, + "grad_norm": 0.17422354221343994, + "learning_rate": 6.623653733161892e-05, + "loss": 0.0493, + "step": 42500 + }, + { + "epoch": 2.785166743104239, + "grad_norm": 0.44330573081970215, + "learning_rate": 6.622089781903217e-05, + "loss": 0.0323, + "step": 42510 + }, + { + "epoch": 2.7858219222957477, + "grad_norm": 0.668449878692627, + "learning_rate": 6.620525653255241e-05, + "loss": 0.0308, + "step": 42520 + }, + { + "epoch": 2.786477101487257, + "grad_norm": 0.46383512020111084, + "learning_rate": 6.618961347389015e-05, + "loss": 0.0457, + "step": 42530 + }, + { + "epoch": 2.7871322806787657, + "grad_norm": 0.14176546037197113, + "learning_rate": 6.617396864475613e-05, + "loss": 0.0282, + "step": 42540 + }, + { + "epoch": 2.7877874598702745, + "grad_norm": 0.4117755889892578, + "learning_rate": 6.61583220468612e-05, + "loss": 0.0454, + "step": 42550 + }, + { + "epoch": 2.7884426390617834, + "grad_norm": 0.7623457312583923, + "learning_rate": 6.614267368191645e-05, + "loss": 0.0414, + "step": 42560 + }, + { + "epoch": 2.789097818253292, + "grad_norm": 0.4273803234100342, + "learning_rate": 6.61270235516332e-05, + "loss": 0.0591, + "step": 42570 + }, + { + "epoch": 2.7897529974448014, + "grad_norm": 0.30618932843208313, + "learning_rate": 6.611137165772287e-05, + "loss": 0.064, + "step": 42580 + }, + { + "epoch": 2.7904081766363102, + "grad_norm": 0.6000345945358276, + "learning_rate": 6.60957180018972e-05, + "loss": 0.0348, + "step": 42590 + }, + { + "epoch": 2.791063355827819, + "grad_norm": 1.1413694620132446, + "learning_rate": 6.608006258586797e-05, + "loss": 0.0343, + "step": 42600 + }, + { + "epoch": 2.791718535019328, + "grad_norm": 0.1316494345664978, + "learning_rate": 6.606440541134731e-05, + "loss": 0.0431, + "step": 42610 + }, + { + "epoch": 2.7923737142108367, + "grad_norm": 0.49728965759277344, + "learning_rate": 6.60487464800474e-05, + "loss": 0.0603, + "step": 42620 + }, + { + "epoch": 2.7930288934023455, + "grad_norm": 0.17008355259895325, + "learning_rate": 6.603308579368071e-05, + "loss": 0.0407, + "step": 42630 + }, + { + "epoch": 2.7936840725938543, + "grad_norm": 1.48375403881073, + "learning_rate": 6.601742335395987e-05, + "loss": 0.0426, + "step": 42640 + }, + { + "epoch": 2.794339251785363, + "grad_norm": 0.4978350102901459, + "learning_rate": 6.600175916259769e-05, + "loss": 0.0457, + "step": 42650 + }, + { + "epoch": 2.794994430976872, + "grad_norm": 0.0870542898774147, + "learning_rate": 6.598609322130718e-05, + "loss": 0.0496, + "step": 42660 + }, + { + "epoch": 2.795649610168381, + "grad_norm": 0.5436197519302368, + "learning_rate": 6.597042553180155e-05, + "loss": 0.0429, + "step": 42670 + }, + { + "epoch": 2.79630478935989, + "grad_norm": 0.5095956325531006, + "learning_rate": 6.59547560957942e-05, + "loss": 0.0356, + "step": 42680 + }, + { + "epoch": 2.796959968551399, + "grad_norm": 0.27647626399993896, + "learning_rate": 6.59390849149987e-05, + "loss": 0.0456, + "step": 42690 + }, + { + "epoch": 2.7976151477429076, + "grad_norm": 0.7744830846786499, + "learning_rate": 6.592341199112886e-05, + "loss": 0.0325, + "step": 42700 + }, + { + "epoch": 2.7982703269344165, + "grad_norm": 0.31989026069641113, + "learning_rate": 6.590773732589861e-05, + "loss": 0.0462, + "step": 42710 + }, + { + "epoch": 2.7989255061259253, + "grad_norm": 0.9872512817382812, + "learning_rate": 6.589206092102213e-05, + "loss": 0.0261, + "step": 42720 + }, + { + "epoch": 2.7995806853174345, + "grad_norm": 0.12461651861667633, + "learning_rate": 6.587638277821375e-05, + "loss": 0.027, + "step": 42730 + }, + { + "epoch": 2.8002358645089434, + "grad_norm": 0.2857239544391632, + "learning_rate": 6.586070289918803e-05, + "loss": 0.0465, + "step": 42740 + }, + { + "epoch": 2.800891043700452, + "grad_norm": 0.48541563749313354, + "learning_rate": 6.584502128565968e-05, + "loss": 0.0478, + "step": 42750 + }, + { + "epoch": 2.801546222891961, + "grad_norm": 0.8617855310440063, + "learning_rate": 6.582933793934363e-05, + "loss": 0.0421, + "step": 42760 + }, + { + "epoch": 2.80220140208347, + "grad_norm": 1.1637972593307495, + "learning_rate": 6.581365286195498e-05, + "loss": 0.031, + "step": 42770 + }, + { + "epoch": 2.8028565812749786, + "grad_norm": 0.42915382981300354, + "learning_rate": 6.579796605520903e-05, + "loss": 0.0393, + "step": 42780 + }, + { + "epoch": 2.8035117604664874, + "grad_norm": 0.4631800055503845, + "learning_rate": 6.578227752082127e-05, + "loss": 0.0629, + "step": 42790 + }, + { + "epoch": 2.8041669396579962, + "grad_norm": 0.47991743683815, + "learning_rate": 6.576658726050735e-05, + "loss": 0.0926, + "step": 42800 + }, + { + "epoch": 2.8048221188495055, + "grad_norm": 0.5393259525299072, + "learning_rate": 6.575089527598316e-05, + "loss": 0.0524, + "step": 42810 + }, + { + "epoch": 2.8054772980410143, + "grad_norm": 0.41576236486434937, + "learning_rate": 6.573520156896476e-05, + "loss": 0.021, + "step": 42820 + }, + { + "epoch": 2.806132477232523, + "grad_norm": 0.4299755394458771, + "learning_rate": 6.571950614116835e-05, + "loss": 0.0379, + "step": 42830 + }, + { + "epoch": 2.806787656424032, + "grad_norm": 0.06749375909566879, + "learning_rate": 6.570380899431039e-05, + "loss": 0.0338, + "step": 42840 + }, + { + "epoch": 2.8074428356155408, + "grad_norm": 0.7700912952423096, + "learning_rate": 6.568811013010749e-05, + "loss": 0.0496, + "step": 42850 + }, + { + "epoch": 2.8080980148070496, + "grad_norm": 0.356518030166626, + "learning_rate": 6.567240955027645e-05, + "loss": 0.0319, + "step": 42860 + }, + { + "epoch": 2.808753193998559, + "grad_norm": 0.9516153335571289, + "learning_rate": 6.565670725653427e-05, + "loss": 0.0424, + "step": 42870 + }, + { + "epoch": 2.8094083731900676, + "grad_norm": 0.6714251041412354, + "learning_rate": 6.564100325059812e-05, + "loss": 0.0421, + "step": 42880 + }, + { + "epoch": 2.8100635523815765, + "grad_norm": 0.2553376257419586, + "learning_rate": 6.562529753418538e-05, + "loss": 0.0468, + "step": 42890 + }, + { + "epoch": 2.8107187315730853, + "grad_norm": 0.5030432939529419, + "learning_rate": 6.56095901090136e-05, + "loss": 0.0321, + "step": 42900 + }, + { + "epoch": 2.811373910764594, + "grad_norm": 0.5667641758918762, + "learning_rate": 6.55938809768005e-05, + "loss": 0.0162, + "step": 42910 + }, + { + "epoch": 2.812029089956103, + "grad_norm": 0.10772863030433655, + "learning_rate": 6.557817013926402e-05, + "loss": 0.0304, + "step": 42920 + }, + { + "epoch": 2.8126842691476117, + "grad_norm": 1.4835498332977295, + "learning_rate": 6.556245759812227e-05, + "loss": 0.0267, + "step": 42930 + }, + { + "epoch": 2.8133394483391205, + "grad_norm": 0.40821677446365356, + "learning_rate": 6.554674335509357e-05, + "loss": 0.0554, + "step": 42940 + }, + { + "epoch": 2.81399462753063, + "grad_norm": 0.9862971305847168, + "learning_rate": 6.553102741189638e-05, + "loss": 0.0277, + "step": 42950 + }, + { + "epoch": 2.8146498067221386, + "grad_norm": 0.5298983454704285, + "learning_rate": 6.55153097702494e-05, + "loss": 0.0234, + "step": 42960 + }, + { + "epoch": 2.8153049859136474, + "grad_norm": 0.4202803075313568, + "learning_rate": 6.549959043187146e-05, + "loss": 0.029, + "step": 42970 + }, + { + "epoch": 2.8159601651051562, + "grad_norm": 0.912733256816864, + "learning_rate": 6.54838693984816e-05, + "loss": 0.0404, + "step": 42980 + }, + { + "epoch": 2.816615344296665, + "grad_norm": 0.8337812423706055, + "learning_rate": 6.546814667179909e-05, + "loss": 0.0655, + "step": 42990 + }, + { + "epoch": 2.817270523488174, + "grad_norm": 0.11095546185970306, + "learning_rate": 6.545242225354328e-05, + "loss": 0.061, + "step": 43000 + }, + { + "epoch": 2.817925702679683, + "grad_norm": 0.24316786229610443, + "learning_rate": 6.543669614543383e-05, + "loss": 0.0525, + "step": 43010 + }, + { + "epoch": 2.818580881871192, + "grad_norm": 0.49377572536468506, + "learning_rate": 6.542096834919049e-05, + "loss": 0.0465, + "step": 43020 + }, + { + "epoch": 2.8192360610627007, + "grad_norm": 0.8147760033607483, + "learning_rate": 6.540523886653324e-05, + "loss": 0.0376, + "step": 43030 + }, + { + "epoch": 2.8198912402542096, + "grad_norm": 0.21510100364685059, + "learning_rate": 6.538950769918222e-05, + "loss": 0.041, + "step": 43040 + }, + { + "epoch": 2.8205464194457184, + "grad_norm": 0.48751387000083923, + "learning_rate": 6.537377484885779e-05, + "loss": 0.043, + "step": 43050 + }, + { + "epoch": 2.821201598637227, + "grad_norm": 1.277023434638977, + "learning_rate": 6.535804031728045e-05, + "loss": 0.0307, + "step": 43060 + }, + { + "epoch": 2.821856777828736, + "grad_norm": 0.5093499422073364, + "learning_rate": 6.534230410617092e-05, + "loss": 0.0496, + "step": 43070 + }, + { + "epoch": 2.822511957020245, + "grad_norm": 0.7623153328895569, + "learning_rate": 6.532656621725007e-05, + "loss": 0.0427, + "step": 43080 + }, + { + "epoch": 2.823167136211754, + "grad_norm": 0.4773983061313629, + "learning_rate": 6.531082665223898e-05, + "loss": 0.0541, + "step": 43090 + }, + { + "epoch": 2.823822315403263, + "grad_norm": 0.6145754456520081, + "learning_rate": 6.529508541285889e-05, + "loss": 0.0398, + "step": 43100 + }, + { + "epoch": 2.8244774945947717, + "grad_norm": 0.28264814615249634, + "learning_rate": 6.527934250083128e-05, + "loss": 0.05, + "step": 43110 + }, + { + "epoch": 2.8251326737862805, + "grad_norm": 0.19514577090740204, + "learning_rate": 6.526359791787775e-05, + "loss": 0.0305, + "step": 43120 + }, + { + "epoch": 2.8257878529777893, + "grad_norm": 1.0150316953659058, + "learning_rate": 6.524785166572009e-05, + "loss": 0.0263, + "step": 43130 + }, + { + "epoch": 2.826443032169298, + "grad_norm": 1.1705621480941772, + "learning_rate": 6.523210374608029e-05, + "loss": 0.0364, + "step": 43140 + }, + { + "epoch": 2.8270982113608074, + "grad_norm": 0.6585761308670044, + "learning_rate": 6.521635416068054e-05, + "loss": 0.0458, + "step": 43150 + }, + { + "epoch": 2.8277533905523162, + "grad_norm": 2.0422048568725586, + "learning_rate": 6.520060291124317e-05, + "loss": 0.0397, + "step": 43160 + }, + { + "epoch": 2.828408569743825, + "grad_norm": 0.8009133338928223, + "learning_rate": 6.518484999949072e-05, + "loss": 0.0577, + "step": 43170 + }, + { + "epoch": 2.829063748935334, + "grad_norm": 0.28288766741752625, + "learning_rate": 6.516909542714591e-05, + "loss": 0.0764, + "step": 43180 + }, + { + "epoch": 2.8297189281268427, + "grad_norm": 0.4881785809993744, + "learning_rate": 6.515333919593161e-05, + "loss": 0.0365, + "step": 43190 + }, + { + "epoch": 2.8303741073183515, + "grad_norm": 0.9603593349456787, + "learning_rate": 6.513758130757094e-05, + "loss": 0.0407, + "step": 43200 + }, + { + "epoch": 2.8310292865098603, + "grad_norm": 2.2526004314422607, + "learning_rate": 6.512182176378713e-05, + "loss": 0.0589, + "step": 43210 + }, + { + "epoch": 2.831684465701369, + "grad_norm": 0.2728615701198578, + "learning_rate": 6.510606056630362e-05, + "loss": 0.0427, + "step": 43220 + }, + { + "epoch": 2.832339644892878, + "grad_norm": 0.11614683270454407, + "learning_rate": 6.509029771684405e-05, + "loss": 0.0208, + "step": 43230 + }, + { + "epoch": 2.832994824084387, + "grad_norm": 0.7838938236236572, + "learning_rate": 6.50745332171322e-05, + "loss": 0.0547, + "step": 43240 + }, + { + "epoch": 2.833650003275896, + "grad_norm": 0.24178630113601685, + "learning_rate": 6.505876706889207e-05, + "loss": 0.0474, + "step": 43250 + }, + { + "epoch": 2.834305182467405, + "grad_norm": 1.6498981714248657, + "learning_rate": 6.504299927384781e-05, + "loss": 0.0714, + "step": 43260 + }, + { + "epoch": 2.8349603616589136, + "grad_norm": 0.613288402557373, + "learning_rate": 6.502722983372378e-05, + "loss": 0.0397, + "step": 43270 + }, + { + "epoch": 2.8356155408504224, + "grad_norm": 0.25424036383628845, + "learning_rate": 6.501145875024446e-05, + "loss": 0.035, + "step": 43280 + }, + { + "epoch": 2.8362707200419317, + "grad_norm": 0.44653597474098206, + "learning_rate": 6.499568602513459e-05, + "loss": 0.0377, + "step": 43290 + }, + { + "epoch": 2.8369258992334405, + "grad_norm": 0.15916766226291656, + "learning_rate": 6.497991166011903e-05, + "loss": 0.054, + "step": 43300 + }, + { + "epoch": 2.8375810784249493, + "grad_norm": 0.5967651009559631, + "learning_rate": 6.49641356569229e-05, + "loss": 0.0394, + "step": 43310 + }, + { + "epoch": 2.838236257616458, + "grad_norm": 0.1975468099117279, + "learning_rate": 6.494835801727135e-05, + "loss": 0.0463, + "step": 43320 + }, + { + "epoch": 2.838891436807967, + "grad_norm": 0.49209555983543396, + "learning_rate": 6.493257874288986e-05, + "loss": 0.0501, + "step": 43330 + }, + { + "epoch": 2.8395466159994758, + "grad_norm": 0.7458394765853882, + "learning_rate": 6.491679783550399e-05, + "loss": 0.069, + "step": 43340 + }, + { + "epoch": 2.8402017951909846, + "grad_norm": 0.27676719427108765, + "learning_rate": 6.490101529683957e-05, + "loss": 0.0173, + "step": 43350 + }, + { + "epoch": 2.8408569743824934, + "grad_norm": 0.20798517763614655, + "learning_rate": 6.488523112862248e-05, + "loss": 0.0178, + "step": 43360 + }, + { + "epoch": 2.841512153574002, + "grad_norm": 0.5578407645225525, + "learning_rate": 6.486944533257891e-05, + "loss": 0.0376, + "step": 43370 + }, + { + "epoch": 2.8421673327655115, + "grad_norm": 0.18182240426540375, + "learning_rate": 6.485365791043517e-05, + "loss": 0.0298, + "step": 43380 + }, + { + "epoch": 2.8428225119570203, + "grad_norm": 0.6685062050819397, + "learning_rate": 6.483786886391772e-05, + "loss": 0.0412, + "step": 43390 + }, + { + "epoch": 2.843477691148529, + "grad_norm": 0.6897993683815002, + "learning_rate": 6.482207819475323e-05, + "loss": 0.0583, + "step": 43400 + }, + { + "epoch": 2.844132870340038, + "grad_norm": 0.7871944308280945, + "learning_rate": 6.480628590466857e-05, + "loss": 0.043, + "step": 43410 + }, + { + "epoch": 2.8447880495315467, + "grad_norm": 0.37321341037750244, + "learning_rate": 6.479049199539074e-05, + "loss": 0.0223, + "step": 43420 + }, + { + "epoch": 2.845443228723056, + "grad_norm": 0.8870698809623718, + "learning_rate": 6.477469646864693e-05, + "loss": 0.0472, + "step": 43430 + }, + { + "epoch": 2.846098407914565, + "grad_norm": 0.5048401355743408, + "learning_rate": 6.475889932616454e-05, + "loss": 0.0307, + "step": 43440 + }, + { + "epoch": 2.8467535871060736, + "grad_norm": 0.3311622142791748, + "learning_rate": 6.474310056967111e-05, + "loss": 0.0367, + "step": 43450 + }, + { + "epoch": 2.8474087662975824, + "grad_norm": 0.594916045665741, + "learning_rate": 6.472730020089437e-05, + "loss": 0.0325, + "step": 43460 + }, + { + "epoch": 2.8480639454890913, + "grad_norm": 0.2268133908510208, + "learning_rate": 6.471149822156222e-05, + "loss": 0.0409, + "step": 43470 + }, + { + "epoch": 2.8487191246806, + "grad_norm": 0.31303173303604126, + "learning_rate": 6.469569463340275e-05, + "loss": 0.0425, + "step": 43480 + }, + { + "epoch": 2.849374303872109, + "grad_norm": 0.9151239395141602, + "learning_rate": 6.467988943814418e-05, + "loss": 0.0577, + "step": 43490 + }, + { + "epoch": 2.8500294830636177, + "grad_norm": 0.4607435166835785, + "learning_rate": 6.4664082637515e-05, + "loss": 0.0417, + "step": 43500 + }, + { + "epoch": 2.8506846622551265, + "grad_norm": 0.6368753910064697, + "learning_rate": 6.464827423324377e-05, + "loss": 0.0325, + "step": 43510 + }, + { + "epoch": 2.8513398414466358, + "grad_norm": 0.34696993231773376, + "learning_rate": 6.463246422705931e-05, + "loss": 0.037, + "step": 43520 + }, + { + "epoch": 2.8519950206381446, + "grad_norm": 0.711469829082489, + "learning_rate": 6.461665262069055e-05, + "loss": 0.0317, + "step": 43530 + }, + { + "epoch": 2.8526501998296534, + "grad_norm": 0.7209115624427795, + "learning_rate": 6.460083941586665e-05, + "loss": 0.0555, + "step": 43540 + }, + { + "epoch": 2.853305379021162, + "grad_norm": 0.17492471635341644, + "learning_rate": 6.45850246143169e-05, + "loss": 0.0578, + "step": 43550 + }, + { + "epoch": 2.853960558212671, + "grad_norm": 0.7089681029319763, + "learning_rate": 6.456920821777077e-05, + "loss": 0.0233, + "step": 43560 + }, + { + "epoch": 2.8546157374041803, + "grad_norm": 0.695140540599823, + "learning_rate": 6.455339022795796e-05, + "loss": 0.0688, + "step": 43570 + }, + { + "epoch": 2.855270916595689, + "grad_norm": 0.459657222032547, + "learning_rate": 6.453757064660825e-05, + "loss": 0.0337, + "step": 43580 + }, + { + "epoch": 2.855926095787198, + "grad_norm": 1.2059966325759888, + "learning_rate": 6.452174947545169e-05, + "loss": 0.0475, + "step": 43590 + }, + { + "epoch": 2.8565812749787067, + "grad_norm": 0.5073481798171997, + "learning_rate": 6.450592671621842e-05, + "loss": 0.0296, + "step": 43600 + }, + { + "epoch": 2.8572364541702155, + "grad_norm": 0.49707943201065063, + "learning_rate": 6.449010237063882e-05, + "loss": 0.0384, + "step": 43610 + }, + { + "epoch": 2.8578916333617244, + "grad_norm": 0.7747435569763184, + "learning_rate": 6.447427644044343e-05, + "loss": 0.0476, + "step": 43620 + }, + { + "epoch": 2.858546812553233, + "grad_norm": 0.16217289865016937, + "learning_rate": 6.44584489273629e-05, + "loss": 0.0297, + "step": 43630 + }, + { + "epoch": 2.859201991744742, + "grad_norm": 0.32491418719291687, + "learning_rate": 6.444261983312817e-05, + "loss": 0.033, + "step": 43640 + }, + { + "epoch": 2.859857170936251, + "grad_norm": 0.6705611348152161, + "learning_rate": 6.442678915947023e-05, + "loss": 0.0316, + "step": 43650 + }, + { + "epoch": 2.86051235012776, + "grad_norm": 0.202991783618927, + "learning_rate": 6.441095690812032e-05, + "loss": 0.0404, + "step": 43660 + }, + { + "epoch": 2.861167529319269, + "grad_norm": 0.5967340469360352, + "learning_rate": 6.439512308080982e-05, + "loss": 0.0442, + "step": 43670 + }, + { + "epoch": 2.8618227085107777, + "grad_norm": 1.2070496082305908, + "learning_rate": 6.43792876792703e-05, + "loss": 0.0528, + "step": 43680 + }, + { + "epoch": 2.8624778877022865, + "grad_norm": 0.4004009962081909, + "learning_rate": 6.43634507052335e-05, + "loss": 0.0425, + "step": 43690 + }, + { + "epoch": 2.8631330668937953, + "grad_norm": 0.9759016036987305, + "learning_rate": 6.434761216043135e-05, + "loss": 0.0445, + "step": 43700 + }, + { + "epoch": 2.8637882460853046, + "grad_norm": 0.5966033935546875, + "learning_rate": 6.433177204659589e-05, + "loss": 0.0482, + "step": 43710 + }, + { + "epoch": 2.8644434252768134, + "grad_norm": 0.1647660732269287, + "learning_rate": 6.431593036545938e-05, + "loss": 0.0308, + "step": 43720 + }, + { + "epoch": 2.865098604468322, + "grad_norm": 0.23082295060157776, + "learning_rate": 6.430008711875426e-05, + "loss": 0.0338, + "step": 43730 + }, + { + "epoch": 2.865753783659831, + "grad_norm": 0.12979008257389069, + "learning_rate": 6.42842423082131e-05, + "loss": 0.0257, + "step": 43740 + }, + { + "epoch": 2.86640896285134, + "grad_norm": 0.2270967662334442, + "learning_rate": 6.426839593556868e-05, + "loss": 0.05, + "step": 43750 + }, + { + "epoch": 2.8670641420428487, + "grad_norm": 0.3351801931858063, + "learning_rate": 6.425254800255392e-05, + "loss": 0.0505, + "step": 43760 + }, + { + "epoch": 2.8677193212343575, + "grad_norm": 0.3714800477027893, + "learning_rate": 6.423669851090196e-05, + "loss": 0.0327, + "step": 43770 + }, + { + "epoch": 2.8683745004258663, + "grad_norm": 2.4447855949401855, + "learning_rate": 6.422084746234604e-05, + "loss": 0.0508, + "step": 43780 + }, + { + "epoch": 2.869029679617375, + "grad_norm": 0.3198043704032898, + "learning_rate": 6.420499485861965e-05, + "loss": 0.0157, + "step": 43790 + }, + { + "epoch": 2.8696848588088844, + "grad_norm": 0.5580177903175354, + "learning_rate": 6.418914070145636e-05, + "loss": 0.0488, + "step": 43800 + }, + { + "epoch": 2.870340038000393, + "grad_norm": 0.4653167128562927, + "learning_rate": 6.417328499258996e-05, + "loss": 0.0406, + "step": 43810 + }, + { + "epoch": 2.870995217191902, + "grad_norm": 0.7155889868736267, + "learning_rate": 6.415742773375446e-05, + "loss": 0.0565, + "step": 43820 + }, + { + "epoch": 2.871650396383411, + "grad_norm": 0.6416075229644775, + "learning_rate": 6.414156892668393e-05, + "loss": 0.0228, + "step": 43830 + }, + { + "epoch": 2.8723055755749196, + "grad_norm": 0.2383863776922226, + "learning_rate": 6.412570857311267e-05, + "loss": 0.037, + "step": 43840 + }, + { + "epoch": 2.872960754766429, + "grad_norm": 0.20894765853881836, + "learning_rate": 6.410984667477518e-05, + "loss": 0.0276, + "step": 43850 + }, + { + "epoch": 2.8736159339579377, + "grad_norm": 0.2953067421913147, + "learning_rate": 6.409398323340607e-05, + "loss": 0.0347, + "step": 43860 + }, + { + "epoch": 2.8742711131494465, + "grad_norm": 0.3845949172973633, + "learning_rate": 6.407811825074012e-05, + "loss": 0.0311, + "step": 43870 + }, + { + "epoch": 2.8749262923409553, + "grad_norm": 0.7075530886650085, + "learning_rate": 6.406225172851234e-05, + "loss": 0.0442, + "step": 43880 + }, + { + "epoch": 2.875581471532464, + "grad_norm": 0.34250420331954956, + "learning_rate": 6.404638366845786e-05, + "loss": 0.0376, + "step": 43890 + }, + { + "epoch": 2.876236650723973, + "grad_norm": 0.6791026592254639, + "learning_rate": 6.403051407231196e-05, + "loss": 0.0473, + "step": 43900 + }, + { + "epoch": 2.8768918299154818, + "grad_norm": 0.2847367227077484, + "learning_rate": 6.401464294181016e-05, + "loss": 0.0249, + "step": 43910 + }, + { + "epoch": 2.8775470091069906, + "grad_norm": 0.16411784291267395, + "learning_rate": 6.399877027868808e-05, + "loss": 0.0377, + "step": 43920 + }, + { + "epoch": 2.8782021882984994, + "grad_norm": 0.31119269132614136, + "learning_rate": 6.39828960846815e-05, + "loss": 0.0272, + "step": 43930 + }, + { + "epoch": 2.8788573674900086, + "grad_norm": 0.15032948553562164, + "learning_rate": 6.396702036152644e-05, + "loss": 0.0271, + "step": 43940 + }, + { + "epoch": 2.8795125466815175, + "grad_norm": 0.8362253308296204, + "learning_rate": 6.395114311095905e-05, + "loss": 0.0301, + "step": 43950 + }, + { + "epoch": 2.8801677258730263, + "grad_norm": 0.3729158639907837, + "learning_rate": 6.393526433471559e-05, + "loss": 0.0343, + "step": 43960 + }, + { + "epoch": 2.880822905064535, + "grad_norm": 1.7090661525726318, + "learning_rate": 6.391938403453262e-05, + "loss": 0.0324, + "step": 43970 + }, + { + "epoch": 2.881478084256044, + "grad_norm": 0.3000507056713104, + "learning_rate": 6.390350221214671e-05, + "loss": 0.0399, + "step": 43980 + }, + { + "epoch": 2.882133263447553, + "grad_norm": 0.25717610120773315, + "learning_rate": 6.38876188692947e-05, + "loss": 0.0325, + "step": 43990 + }, + { + "epoch": 2.882788442639062, + "grad_norm": 0.48823803663253784, + "learning_rate": 6.387173400771358e-05, + "loss": 0.0327, + "step": 44000 + }, + { + "epoch": 2.883443621830571, + "grad_norm": 0.4476128816604614, + "learning_rate": 6.385584762914047e-05, + "loss": 0.0371, + "step": 44010 + }, + { + "epoch": 2.8840988010220796, + "grad_norm": 0.7671319246292114, + "learning_rate": 6.38399597353127e-05, + "loss": 0.0377, + "step": 44020 + }, + { + "epoch": 2.8847539802135884, + "grad_norm": 0.1096438616514206, + "learning_rate": 6.382407032796775e-05, + "loss": 0.0355, + "step": 44030 + }, + { + "epoch": 2.8854091594050972, + "grad_norm": 0.13398700952529907, + "learning_rate": 6.380817940884325e-05, + "loss": 0.0199, + "step": 44040 + }, + { + "epoch": 2.886064338596606, + "grad_norm": 0.2110234946012497, + "learning_rate": 6.379228697967702e-05, + "loss": 0.0454, + "step": 44050 + }, + { + "epoch": 2.886719517788115, + "grad_norm": 0.46179622411727905, + "learning_rate": 6.3776393042207e-05, + "loss": 0.0343, + "step": 44060 + }, + { + "epoch": 2.8873746969796237, + "grad_norm": 0.19342082738876343, + "learning_rate": 6.376049759817137e-05, + "loss": 0.0329, + "step": 44070 + }, + { + "epoch": 2.888029876171133, + "grad_norm": 0.6914994120597839, + "learning_rate": 6.37446006493084e-05, + "loss": 0.0492, + "step": 44080 + }, + { + "epoch": 2.8886850553626418, + "grad_norm": 0.4700949788093567, + "learning_rate": 6.372870219735656e-05, + "loss": 0.0373, + "step": 44090 + }, + { + "epoch": 2.8893402345541506, + "grad_norm": 0.46820080280303955, + "learning_rate": 6.371280224405452e-05, + "loss": 0.0363, + "step": 44100 + }, + { + "epoch": 2.8899954137456594, + "grad_norm": 0.5384129881858826, + "learning_rate": 6.369690079114101e-05, + "loss": 0.0393, + "step": 44110 + }, + { + "epoch": 2.890650592937168, + "grad_norm": 0.5462536215782166, + "learning_rate": 6.368099784035504e-05, + "loss": 0.0528, + "step": 44120 + }, + { + "epoch": 2.8913057721286775, + "grad_norm": 1.5563873052597046, + "learning_rate": 6.366509339343572e-05, + "loss": 0.0342, + "step": 44130 + }, + { + "epoch": 2.8919609513201863, + "grad_norm": 0.4318990707397461, + "learning_rate": 6.364918745212232e-05, + "loss": 0.0317, + "step": 44140 + }, + { + "epoch": 2.892616130511695, + "grad_norm": 0.5486032366752625, + "learning_rate": 6.363328001815431e-05, + "loss": 0.0513, + "step": 44150 + }, + { + "epoch": 2.893271309703204, + "grad_norm": 0.17710067331790924, + "learning_rate": 6.361737109327128e-05, + "loss": 0.0348, + "step": 44160 + }, + { + "epoch": 2.8939264888947127, + "grad_norm": 0.36635780334472656, + "learning_rate": 6.360146067921301e-05, + "loss": 0.0322, + "step": 44170 + }, + { + "epoch": 2.8945816680862215, + "grad_norm": 0.46034759283065796, + "learning_rate": 6.35855487777195e-05, + "loss": 0.0285, + "step": 44180 + }, + { + "epoch": 2.8952368472777303, + "grad_norm": 0.1692354381084442, + "learning_rate": 6.356963539053076e-05, + "loss": 0.0376, + "step": 44190 + }, + { + "epoch": 2.895892026469239, + "grad_norm": 0.46992743015289307, + "learning_rate": 6.355372051938711e-05, + "loss": 0.0384, + "step": 44200 + }, + { + "epoch": 2.896547205660748, + "grad_norm": 0.10365484654903412, + "learning_rate": 6.353780416602894e-05, + "loss": 0.0305, + "step": 44210 + }, + { + "epoch": 2.8972023848522572, + "grad_norm": 1.267651915550232, + "learning_rate": 6.352188633219689e-05, + "loss": 0.0574, + "step": 44220 + }, + { + "epoch": 2.897857564043766, + "grad_norm": 0.8080046772956848, + "learning_rate": 6.350596701963166e-05, + "loss": 0.0529, + "step": 44230 + }, + { + "epoch": 2.898512743235275, + "grad_norm": 0.3522300124168396, + "learning_rate": 6.349004623007419e-05, + "loss": 0.0366, + "step": 44240 + }, + { + "epoch": 2.8991679224267837, + "grad_norm": 0.26124051213264465, + "learning_rate": 6.347412396526555e-05, + "loss": 0.0302, + "step": 44250 + }, + { + "epoch": 2.8998231016182925, + "grad_norm": 0.6794007420539856, + "learning_rate": 6.345820022694696e-05, + "loss": 0.0438, + "step": 44260 + }, + { + "epoch": 2.9004782808098017, + "grad_norm": 0.2248380184173584, + "learning_rate": 6.344227501685984e-05, + "loss": 0.0414, + "step": 44270 + }, + { + "epoch": 2.9011334600013106, + "grad_norm": 0.3353596031665802, + "learning_rate": 6.342634833674572e-05, + "loss": 0.0337, + "step": 44280 + }, + { + "epoch": 2.9017886391928194, + "grad_norm": 0.14752644300460815, + "learning_rate": 6.341042018834635e-05, + "loss": 0.024, + "step": 44290 + }, + { + "epoch": 2.902443818384328, + "grad_norm": 0.5393759608268738, + "learning_rate": 6.339449057340359e-05, + "loss": 0.0358, + "step": 44300 + }, + { + "epoch": 2.903098997575837, + "grad_norm": 0.2792373597621918, + "learning_rate": 6.337855949365945e-05, + "loss": 0.0295, + "step": 44310 + }, + { + "epoch": 2.903754176767346, + "grad_norm": 0.6084073185920715, + "learning_rate": 6.336262695085619e-05, + "loss": 0.0459, + "step": 44320 + }, + { + "epoch": 2.9044093559588546, + "grad_norm": 0.12124935537576675, + "learning_rate": 6.334669294673612e-05, + "loss": 0.0447, + "step": 44330 + }, + { + "epoch": 2.9050645351503634, + "grad_norm": 1.0569730997085571, + "learning_rate": 6.333075748304179e-05, + "loss": 0.0241, + "step": 44340 + }, + { + "epoch": 2.9057197143418723, + "grad_norm": 0.6409674286842346, + "learning_rate": 6.331482056151585e-05, + "loss": 0.0433, + "step": 44350 + }, + { + "epoch": 2.9063748935333815, + "grad_norm": 0.5294522643089294, + "learning_rate": 6.329888218390117e-05, + "loss": 0.027, + "step": 44360 + }, + { + "epoch": 2.9070300727248903, + "grad_norm": 0.19738803803920746, + "learning_rate": 6.328294235194072e-05, + "loss": 0.0205, + "step": 44370 + }, + { + "epoch": 2.907685251916399, + "grad_norm": 0.1650703400373459, + "learning_rate": 6.326700106737765e-05, + "loss": 0.041, + "step": 44380 + }, + { + "epoch": 2.908340431107908, + "grad_norm": 0.14969654381275177, + "learning_rate": 6.325105833195531e-05, + "loss": 0.0523, + "step": 44390 + }, + { + "epoch": 2.908995610299417, + "grad_norm": 0.509921133518219, + "learning_rate": 6.323511414741715e-05, + "loss": 0.0471, + "step": 44400 + }, + { + "epoch": 2.909650789490926, + "grad_norm": 0.42903682589530945, + "learning_rate": 6.321916851550678e-05, + "loss": 0.0172, + "step": 44410 + }, + { + "epoch": 2.910305968682435, + "grad_norm": 0.36666184663772583, + "learning_rate": 6.320322143796806e-05, + "loss": 0.0244, + "step": 44420 + }, + { + "epoch": 2.9109611478739437, + "grad_norm": 0.9097941517829895, + "learning_rate": 6.318727291654488e-05, + "loss": 0.0572, + "step": 44430 + }, + { + "epoch": 2.9116163270654525, + "grad_norm": 0.712209939956665, + "learning_rate": 6.317132295298134e-05, + "loss": 0.0488, + "step": 44440 + }, + { + "epoch": 2.9122715062569613, + "grad_norm": 0.4423668682575226, + "learning_rate": 6.315537154902173e-05, + "loss": 0.0314, + "step": 44450 + }, + { + "epoch": 2.91292668544847, + "grad_norm": 0.3453448712825775, + "learning_rate": 6.313941870641048e-05, + "loss": 0.043, + "step": 44460 + }, + { + "epoch": 2.913581864639979, + "grad_norm": 0.17827598750591278, + "learning_rate": 6.312346442689214e-05, + "loss": 0.0359, + "step": 44470 + }, + { + "epoch": 2.9142370438314877, + "grad_norm": 0.22858662903308868, + "learning_rate": 6.310750871221147e-05, + "loss": 0.0384, + "step": 44480 + }, + { + "epoch": 2.9148922230229966, + "grad_norm": 0.1939469575881958, + "learning_rate": 6.309155156411335e-05, + "loss": 0.0494, + "step": 44490 + }, + { + "epoch": 2.915547402214506, + "grad_norm": 0.3327619135379791, + "learning_rate": 6.307559298434284e-05, + "loss": 0.0501, + "step": 44500 + }, + { + "epoch": 2.9162025814060146, + "grad_norm": 0.31233370304107666, + "learning_rate": 6.305963297464512e-05, + "loss": 0.0343, + "step": 44510 + }, + { + "epoch": 2.9168577605975234, + "grad_norm": 0.26936545968055725, + "learning_rate": 6.304367153676561e-05, + "loss": 0.0325, + "step": 44520 + }, + { + "epoch": 2.9175129397890323, + "grad_norm": 0.26050257682800293, + "learning_rate": 6.302770867244978e-05, + "loss": 0.0452, + "step": 44530 + }, + { + "epoch": 2.918168118980541, + "grad_norm": 0.231293186545372, + "learning_rate": 6.301174438344328e-05, + "loss": 0.0511, + "step": 44540 + }, + { + "epoch": 2.9188232981720503, + "grad_norm": 0.988186776638031, + "learning_rate": 6.299577867149202e-05, + "loss": 0.0329, + "step": 44550 + }, + { + "epoch": 2.919478477363559, + "grad_norm": 0.32313334941864014, + "learning_rate": 6.297981153834192e-05, + "loss": 0.0284, + "step": 44560 + }, + { + "epoch": 2.920133656555068, + "grad_norm": 0.22273509204387665, + "learning_rate": 6.296384298573916e-05, + "loss": 0.0433, + "step": 44570 + }, + { + "epoch": 2.9207888357465768, + "grad_norm": 0.25581517815589905, + "learning_rate": 6.294787301543001e-05, + "loss": 0.0264, + "step": 44580 + }, + { + "epoch": 2.9214440149380856, + "grad_norm": 0.9012549519538879, + "learning_rate": 6.293190162916095e-05, + "loss": 0.0268, + "step": 44590 + }, + { + "epoch": 2.9220991941295944, + "grad_norm": 0.44765976071357727, + "learning_rate": 6.291592882867855e-05, + "loss": 0.0249, + "step": 44600 + }, + { + "epoch": 2.922754373321103, + "grad_norm": 0.4936361610889435, + "learning_rate": 6.28999546157296e-05, + "loss": 0.0283, + "step": 44610 + }, + { + "epoch": 2.923409552512612, + "grad_norm": 0.10514809936285019, + "learning_rate": 6.288397899206102e-05, + "loss": 0.0424, + "step": 44620 + }, + { + "epoch": 2.924064731704121, + "grad_norm": 0.26815474033355713, + "learning_rate": 6.286800195941984e-05, + "loss": 0.0372, + "step": 44630 + }, + { + "epoch": 2.92471991089563, + "grad_norm": 0.49138617515563965, + "learning_rate": 6.285202351955334e-05, + "loss": 0.0481, + "step": 44640 + }, + { + "epoch": 2.925375090087139, + "grad_norm": 0.520957350730896, + "learning_rate": 6.283604367420887e-05, + "loss": 0.0233, + "step": 44650 + }, + { + "epoch": 2.9260302692786477, + "grad_norm": 0.4042218327522278, + "learning_rate": 6.282006242513394e-05, + "loss": 0.0405, + "step": 44660 + }, + { + "epoch": 2.9266854484701565, + "grad_norm": 0.3240509331226349, + "learning_rate": 6.280407977407628e-05, + "loss": 0.0512, + "step": 44670 + }, + { + "epoch": 2.9273406276616654, + "grad_norm": 0.34032920002937317, + "learning_rate": 6.27880957227837e-05, + "loss": 0.0322, + "step": 44680 + }, + { + "epoch": 2.9279958068531746, + "grad_norm": 0.6578122973442078, + "learning_rate": 6.27721102730042e-05, + "loss": 0.0427, + "step": 44690 + }, + { + "epoch": 2.9286509860446834, + "grad_norm": 0.2644079327583313, + "learning_rate": 6.275612342648592e-05, + "loss": 0.045, + "step": 44700 + }, + { + "epoch": 2.9293061652361922, + "grad_norm": 0.7829111218452454, + "learning_rate": 6.274013518497716e-05, + "loss": 0.0231, + "step": 44710 + }, + { + "epoch": 2.929961344427701, + "grad_norm": 0.7779720425605774, + "learning_rate": 6.272414555022636e-05, + "loss": 0.0463, + "step": 44720 + }, + { + "epoch": 2.93061652361921, + "grad_norm": 0.37000560760498047, + "learning_rate": 6.270815452398215e-05, + "loss": 0.0613, + "step": 44730 + }, + { + "epoch": 2.9312717028107187, + "grad_norm": 0.5451076030731201, + "learning_rate": 6.269216210799326e-05, + "loss": 0.0434, + "step": 44740 + }, + { + "epoch": 2.9319268820022275, + "grad_norm": 0.7235223054885864, + "learning_rate": 6.26761683040086e-05, + "loss": 0.0404, + "step": 44750 + }, + { + "epoch": 2.9325820611937363, + "grad_norm": 0.3989701569080353, + "learning_rate": 6.266017311377723e-05, + "loss": 0.0282, + "step": 44760 + }, + { + "epoch": 2.933237240385245, + "grad_norm": 0.2637997567653656, + "learning_rate": 6.264417653904839e-05, + "loss": 0.0262, + "step": 44770 + }, + { + "epoch": 2.9338924195767544, + "grad_norm": 0.2157817929983139, + "learning_rate": 6.262817858157139e-05, + "loss": 0.0409, + "step": 44780 + }, + { + "epoch": 2.934547598768263, + "grad_norm": 0.6328967213630676, + "learning_rate": 6.261217924309576e-05, + "loss": 0.042, + "step": 44790 + }, + { + "epoch": 2.935202777959772, + "grad_norm": 0.4281482696533203, + "learning_rate": 6.259617852537118e-05, + "loss": 0.0376, + "step": 44800 + }, + { + "epoch": 2.935857957151281, + "grad_norm": 0.3039303421974182, + "learning_rate": 6.258017643014747e-05, + "loss": 0.0409, + "step": 44810 + }, + { + "epoch": 2.9365131363427897, + "grad_norm": 0.3501052260398865, + "learning_rate": 6.256417295917456e-05, + "loss": 0.0513, + "step": 44820 + }, + { + "epoch": 2.9371683155342985, + "grad_norm": 0.45181700587272644, + "learning_rate": 6.254816811420258e-05, + "loss": 0.0346, + "step": 44830 + }, + { + "epoch": 2.9378234947258077, + "grad_norm": 0.6705488562583923, + "learning_rate": 6.253216189698183e-05, + "loss": 0.0482, + "step": 44840 + }, + { + "epoch": 2.9384786739173165, + "grad_norm": 0.195684552192688, + "learning_rate": 6.251615430926267e-05, + "loss": 0.0234, + "step": 44850 + }, + { + "epoch": 2.9391338531088254, + "grad_norm": 0.4450741708278656, + "learning_rate": 6.25001453527957e-05, + "loss": 0.036, + "step": 44860 + }, + { + "epoch": 2.939789032300334, + "grad_norm": 0.4039517641067505, + "learning_rate": 6.248413502933164e-05, + "loss": 0.0497, + "step": 44870 + }, + { + "epoch": 2.940444211491843, + "grad_norm": 0.876694917678833, + "learning_rate": 6.246812334062133e-05, + "loss": 0.0465, + "step": 44880 + }, + { + "epoch": 2.941099390683352, + "grad_norm": 1.5153552293777466, + "learning_rate": 6.245211028841579e-05, + "loss": 0.049, + "step": 44890 + }, + { + "epoch": 2.9417545698748606, + "grad_norm": 0.5131750702857971, + "learning_rate": 6.24360958744662e-05, + "loss": 0.0389, + "step": 44900 + }, + { + "epoch": 2.9424097490663694, + "grad_norm": 0.14266285300254822, + "learning_rate": 6.242008010052387e-05, + "loss": 0.0245, + "step": 44910 + }, + { + "epoch": 2.9430649282578787, + "grad_norm": 0.8143408894538879, + "learning_rate": 6.240406296834024e-05, + "loss": 0.0337, + "step": 44920 + }, + { + "epoch": 2.9437201074493875, + "grad_norm": 0.19987869262695312, + "learning_rate": 6.238804447966694e-05, + "loss": 0.0283, + "step": 44930 + }, + { + "epoch": 2.9443752866408963, + "grad_norm": 1.2720227241516113, + "learning_rate": 6.237202463625573e-05, + "loss": 0.0565, + "step": 44940 + }, + { + "epoch": 2.945030465832405, + "grad_norm": 0.7930856347084045, + "learning_rate": 6.235600343985848e-05, + "loss": 0.0409, + "step": 44950 + }, + { + "epoch": 2.945685645023914, + "grad_norm": 0.22183804214000702, + "learning_rate": 6.233998089222729e-05, + "loss": 0.0374, + "step": 44960 + }, + { + "epoch": 2.9463408242154228, + "grad_norm": 0.39995431900024414, + "learning_rate": 6.232395699511433e-05, + "loss": 0.0248, + "step": 44970 + }, + { + "epoch": 2.946996003406932, + "grad_norm": 0.2911123037338257, + "learning_rate": 6.230793175027199e-05, + "loss": 0.0236, + "step": 44980 + }, + { + "epoch": 2.947651182598441, + "grad_norm": 0.2885056734085083, + "learning_rate": 6.22919051594527e-05, + "loss": 0.0161, + "step": 44990 + }, + { + "epoch": 2.9483063617899496, + "grad_norm": 0.3505008816719055, + "learning_rate": 6.227587722440917e-05, + "loss": 0.0483, + "step": 45000 + }, + { + "epoch": 2.9489615409814585, + "grad_norm": 0.5743699669837952, + "learning_rate": 6.225984794689414e-05, + "loss": 0.0454, + "step": 45010 + }, + { + "epoch": 2.9496167201729673, + "grad_norm": 0.29725515842437744, + "learning_rate": 6.22438173286606e-05, + "loss": 0.0252, + "step": 45020 + }, + { + "epoch": 2.950271899364476, + "grad_norm": 0.81437087059021, + "learning_rate": 6.222778537146159e-05, + "loss": 0.0341, + "step": 45030 + }, + { + "epoch": 2.950927078555985, + "grad_norm": 0.1841522753238678, + "learning_rate": 6.221175207705037e-05, + "loss": 0.0302, + "step": 45040 + }, + { + "epoch": 2.9515822577474937, + "grad_norm": 0.6383190751075745, + "learning_rate": 6.219571744718027e-05, + "loss": 0.0404, + "step": 45050 + }, + { + "epoch": 2.952237436939003, + "grad_norm": 0.624563455581665, + "learning_rate": 6.217968148360487e-05, + "loss": 0.0345, + "step": 45060 + }, + { + "epoch": 2.952892616130512, + "grad_norm": 0.5837278962135315, + "learning_rate": 6.216364418807782e-05, + "loss": 0.0354, + "step": 45070 + }, + { + "epoch": 2.9535477953220206, + "grad_norm": 0.3334425687789917, + "learning_rate": 6.214760556235292e-05, + "loss": 0.0264, + "step": 45080 + }, + { + "epoch": 2.9542029745135294, + "grad_norm": 0.5191294550895691, + "learning_rate": 6.213156560818414e-05, + "loss": 0.0382, + "step": 45090 + }, + { + "epoch": 2.9548581537050382, + "grad_norm": 0.24108052253723145, + "learning_rate": 6.211552432732559e-05, + "loss": 0.024, + "step": 45100 + }, + { + "epoch": 2.955513332896547, + "grad_norm": 0.48115038871765137, + "learning_rate": 6.20994817215315e-05, + "loss": 0.0446, + "step": 45110 + }, + { + "epoch": 2.9561685120880563, + "grad_norm": 0.6166666746139526, + "learning_rate": 6.208343779255628e-05, + "loss": 0.0424, + "step": 45120 + }, + { + "epoch": 2.956823691279565, + "grad_norm": 0.3735521137714386, + "learning_rate": 6.206739254215449e-05, + "loss": 0.035, + "step": 45130 + }, + { + "epoch": 2.957478870471074, + "grad_norm": 0.6600696444511414, + "learning_rate": 6.205134597208077e-05, + "loss": 0.0564, + "step": 45140 + }, + { + "epoch": 2.9581340496625828, + "grad_norm": 0.9039437174797058, + "learning_rate": 6.203529808409e-05, + "loss": 0.0269, + "step": 45150 + }, + { + "epoch": 2.9587892288540916, + "grad_norm": 1.2004449367523193, + "learning_rate": 6.201924887993712e-05, + "loss": 0.0791, + "step": 45160 + }, + { + "epoch": 2.9594444080456004, + "grad_norm": 0.1678592711687088, + "learning_rate": 6.200319836137725e-05, + "loss": 0.0364, + "step": 45170 + }, + { + "epoch": 2.960099587237109, + "grad_norm": 0.14430293440818787, + "learning_rate": 6.198714653016565e-05, + "loss": 0.0418, + "step": 45180 + }, + { + "epoch": 2.960754766428618, + "grad_norm": 0.47488078474998474, + "learning_rate": 6.197109338805774e-05, + "loss": 0.0499, + "step": 45190 + }, + { + "epoch": 2.9614099456201273, + "grad_norm": 1.0885816812515259, + "learning_rate": 6.195503893680903e-05, + "loss": 0.0362, + "step": 45200 + }, + { + "epoch": 2.962065124811636, + "grad_norm": 0.5433045625686646, + "learning_rate": 6.193898317817524e-05, + "loss": 0.0368, + "step": 45210 + }, + { + "epoch": 2.962720304003145, + "grad_norm": 0.780905544757843, + "learning_rate": 6.192292611391221e-05, + "loss": 0.0366, + "step": 45220 + }, + { + "epoch": 2.9633754831946537, + "grad_norm": 0.49520236253738403, + "learning_rate": 6.190686774577591e-05, + "loss": 0.0159, + "step": 45230 + }, + { + "epoch": 2.9640306623861625, + "grad_norm": 0.8601018786430359, + "learning_rate": 6.189080807552245e-05, + "loss": 0.0427, + "step": 45240 + }, + { + "epoch": 2.9646858415776713, + "grad_norm": 0.10738766938447952, + "learning_rate": 6.187474710490809e-05, + "loss": 0.0351, + "step": 45250 + }, + { + "epoch": 2.9653410207691806, + "grad_norm": 1.1430641412734985, + "learning_rate": 6.185868483568926e-05, + "loss": 0.0348, + "step": 45260 + }, + { + "epoch": 2.9659961999606894, + "grad_norm": 0.1469450294971466, + "learning_rate": 6.184262126962245e-05, + "loss": 0.0512, + "step": 45270 + }, + { + "epoch": 2.9666513791521982, + "grad_norm": 0.2838946282863617, + "learning_rate": 6.182655640846442e-05, + "loss": 0.025, + "step": 45280 + }, + { + "epoch": 2.967306558343707, + "grad_norm": 0.15252533555030823, + "learning_rate": 6.181049025397196e-05, + "loss": 0.0309, + "step": 45290 + }, + { + "epoch": 2.967961737535216, + "grad_norm": 0.9210895895957947, + "learning_rate": 6.179442280790202e-05, + "loss": 0.0535, + "step": 45300 + }, + { + "epoch": 2.9686169167267247, + "grad_norm": 0.3754172921180725, + "learning_rate": 6.177835407201174e-05, + "loss": 0.0269, + "step": 45310 + }, + { + "epoch": 2.9692720959182335, + "grad_norm": 0.6206186413764954, + "learning_rate": 6.176228404805839e-05, + "loss": 0.0424, + "step": 45320 + }, + { + "epoch": 2.9699272751097423, + "grad_norm": 0.33465349674224854, + "learning_rate": 6.174621273779932e-05, + "loss": 0.0355, + "step": 45330 + }, + { + "epoch": 2.970582454301251, + "grad_norm": 0.21117636561393738, + "learning_rate": 6.17301401429921e-05, + "loss": 0.0245, + "step": 45340 + }, + { + "epoch": 2.9712376334927604, + "grad_norm": 0.14490237832069397, + "learning_rate": 6.171406626539436e-05, + "loss": 0.032, + "step": 45350 + }, + { + "epoch": 2.971892812684269, + "grad_norm": 0.20845067501068115, + "learning_rate": 6.169799110676398e-05, + "loss": 0.0339, + "step": 45360 + }, + { + "epoch": 2.972547991875778, + "grad_norm": 0.8755435347557068, + "learning_rate": 6.168191466885885e-05, + "loss": 0.0672, + "step": 45370 + }, + { + "epoch": 2.973203171067287, + "grad_norm": 1.1066666841506958, + "learning_rate": 6.166583695343711e-05, + "loss": 0.045, + "step": 45380 + }, + { + "epoch": 2.9738583502587956, + "grad_norm": 0.3273071348667145, + "learning_rate": 6.164975796225698e-05, + "loss": 0.0445, + "step": 45390 + }, + { + "epoch": 2.974513529450305, + "grad_norm": 0.46960970759391785, + "learning_rate": 6.163367769707683e-05, + "loss": 0.0435, + "step": 45400 + }, + { + "epoch": 2.9751687086418137, + "grad_norm": 0.5397495627403259, + "learning_rate": 6.161759615965519e-05, + "loss": 0.0284, + "step": 45410 + }, + { + "epoch": 2.9758238878333225, + "grad_norm": 0.3718006908893585, + "learning_rate": 6.16015133517507e-05, + "loss": 0.0273, + "step": 45420 + }, + { + "epoch": 2.9764790670248313, + "grad_norm": 0.36153918504714966, + "learning_rate": 6.158542927512214e-05, + "loss": 0.0332, + "step": 45430 + }, + { + "epoch": 2.97713424621634, + "grad_norm": 1.6132193803787231, + "learning_rate": 6.156934393152846e-05, + "loss": 0.0424, + "step": 45440 + }, + { + "epoch": 2.977789425407849, + "grad_norm": 0.12489455193281174, + "learning_rate": 6.15532573227287e-05, + "loss": 0.0189, + "step": 45450 + }, + { + "epoch": 2.978444604599358, + "grad_norm": 0.6863153576850891, + "learning_rate": 6.153716945048212e-05, + "loss": 0.0272, + "step": 45460 + }, + { + "epoch": 2.9790997837908666, + "grad_norm": 0.6448436379432678, + "learning_rate": 6.152108031654802e-05, + "loss": 0.0349, + "step": 45470 + }, + { + "epoch": 2.9797549629823754, + "grad_norm": 0.7654843926429749, + "learning_rate": 6.15049899226859e-05, + "loss": 0.0402, + "step": 45480 + }, + { + "epoch": 2.9804101421738847, + "grad_norm": 0.6382696628570557, + "learning_rate": 6.148889827065537e-05, + "loss": 0.0361, + "step": 45490 + }, + { + "epoch": 2.9810653213653935, + "grad_norm": 0.6435016393661499, + "learning_rate": 6.147280536221622e-05, + "loss": 0.0369, + "step": 45500 + }, + { + "epoch": 2.9817205005569023, + "grad_norm": 1.064636468887329, + "learning_rate": 6.145671119912832e-05, + "loss": 0.0309, + "step": 45510 + }, + { + "epoch": 2.982375679748411, + "grad_norm": 0.47066977620124817, + "learning_rate": 6.144061578315169e-05, + "loss": 0.0379, + "step": 45520 + }, + { + "epoch": 2.98303085893992, + "grad_norm": 1.058772325515747, + "learning_rate": 6.142451911604654e-05, + "loss": 0.0635, + "step": 45530 + }, + { + "epoch": 2.983686038131429, + "grad_norm": 0.13730747997760773, + "learning_rate": 6.140842119957315e-05, + "loss": 0.0314, + "step": 45540 + }, + { + "epoch": 2.984341217322938, + "grad_norm": 0.6673192381858826, + "learning_rate": 6.139232203549197e-05, + "loss": 0.0255, + "step": 45550 + }, + { + "epoch": 2.984996396514447, + "grad_norm": 0.5093451738357544, + "learning_rate": 6.137622162556357e-05, + "loss": 0.0463, + "step": 45560 + }, + { + "epoch": 2.9856515757059556, + "grad_norm": 0.14582166075706482, + "learning_rate": 6.13601199715487e-05, + "loss": 0.0246, + "step": 45570 + }, + { + "epoch": 2.9863067548974644, + "grad_norm": 0.49050888419151306, + "learning_rate": 6.134401707520816e-05, + "loss": 0.0482, + "step": 45580 + }, + { + "epoch": 2.9869619340889733, + "grad_norm": 0.2965473234653473, + "learning_rate": 6.132791293830298e-05, + "loss": 0.0267, + "step": 45590 + }, + { + "epoch": 2.987617113280482, + "grad_norm": 0.278499573469162, + "learning_rate": 6.131180756259428e-05, + "loss": 0.0607, + "step": 45600 + }, + { + "epoch": 2.988272292471991, + "grad_norm": 0.26714324951171875, + "learning_rate": 6.129570094984331e-05, + "loss": 0.0287, + "step": 45610 + }, + { + "epoch": 2.9889274716634997, + "grad_norm": 0.2770504057407379, + "learning_rate": 6.127959310181145e-05, + "loss": 0.0316, + "step": 45620 + }, + { + "epoch": 2.989582650855009, + "grad_norm": 0.2911944091320038, + "learning_rate": 6.126348402026026e-05, + "loss": 0.0266, + "step": 45630 + }, + { + "epoch": 2.9902378300465178, + "grad_norm": 0.18166908621788025, + "learning_rate": 6.12473737069514e-05, + "loss": 0.0523, + "step": 45640 + }, + { + "epoch": 2.9908930092380266, + "grad_norm": 0.7276223301887512, + "learning_rate": 6.123126216364665e-05, + "loss": 0.0458, + "step": 45650 + }, + { + "epoch": 2.9915481884295354, + "grad_norm": 0.23916426301002502, + "learning_rate": 6.121514939210797e-05, + "loss": 0.0512, + "step": 45660 + }, + { + "epoch": 2.992203367621044, + "grad_norm": 0.7971847057342529, + "learning_rate": 6.119903539409741e-05, + "loss": 0.0702, + "step": 45670 + }, + { + "epoch": 2.9928585468125535, + "grad_norm": 0.5504698753356934, + "learning_rate": 6.118292017137716e-05, + "loss": 0.0177, + "step": 45680 + }, + { + "epoch": 2.9935137260040623, + "grad_norm": 0.4117599129676819, + "learning_rate": 6.116680372570959e-05, + "loss": 0.0247, + "step": 45690 + }, + { + "epoch": 2.994168905195571, + "grad_norm": 1.188327670097351, + "learning_rate": 6.115068605885713e-05, + "loss": 0.0274, + "step": 45700 + }, + { + "epoch": 2.99482408438708, + "grad_norm": 0.7294514775276184, + "learning_rate": 6.113456717258243e-05, + "loss": 0.024, + "step": 45710 + }, + { + "epoch": 2.9954792635785887, + "grad_norm": 0.7967405915260315, + "learning_rate": 6.11184470686482e-05, + "loss": 0.0327, + "step": 45720 + }, + { + "epoch": 2.9961344427700976, + "grad_norm": 0.5666521191596985, + "learning_rate": 6.110232574881731e-05, + "loss": 0.0427, + "step": 45730 + }, + { + "epoch": 2.9967896219616064, + "grad_norm": 0.37627744674682617, + "learning_rate": 6.108620321485277e-05, + "loss": 0.0346, + "step": 45740 + }, + { + "epoch": 2.997444801153115, + "grad_norm": 0.4183942377567291, + "learning_rate": 6.107007946851773e-05, + "loss": 0.0221, + "step": 45750 + }, + { + "epoch": 2.998099980344624, + "grad_norm": 0.20273339748382568, + "learning_rate": 6.105395451157542e-05, + "loss": 0.0513, + "step": 45760 + }, + { + "epoch": 2.9987551595361333, + "grad_norm": 0.308612197637558, + "learning_rate": 6.103782834578928e-05, + "loss": 0.0235, + "step": 45770 + }, + { + "epoch": 2.999410338727642, + "grad_norm": 0.8545552492141724, + "learning_rate": 6.102170097292281e-05, + "loss": 0.0239, + "step": 45780 + }, + { + "epoch": 3.000065517919151, + "grad_norm": 1.5302321910858154, + "learning_rate": 6.1005572394739716e-05, + "loss": 0.0436, + "step": 45790 + }, + { + "epoch": 3.0007206971106597, + "grad_norm": 0.20366130769252777, + "learning_rate": 6.0989442613003765e-05, + "loss": 0.0363, + "step": 45800 + }, + { + "epoch": 3.0013758763021685, + "grad_norm": 1.0823755264282227, + "learning_rate": 6.097331162947888e-05, + "loss": 0.0463, + "step": 45810 + }, + { + "epoch": 3.0020310554936773, + "grad_norm": 0.7097705602645874, + "learning_rate": 6.095717944592914e-05, + "loss": 0.0448, + "step": 45820 + }, + { + "epoch": 3.0026862346851866, + "grad_norm": 0.5116744041442871, + "learning_rate": 6.094104606411873e-05, + "loss": 0.0351, + "step": 45830 + }, + { + "epoch": 3.0033414138766954, + "grad_norm": 0.8368414044380188, + "learning_rate": 6.0924911485811966e-05, + "loss": 0.0419, + "step": 45840 + }, + { + "epoch": 3.003996593068204, + "grad_norm": 0.24387100338935852, + "learning_rate": 6.090877571277331e-05, + "loss": 0.0198, + "step": 45850 + }, + { + "epoch": 3.004651772259713, + "grad_norm": 0.21931913495063782, + "learning_rate": 6.089263874676736e-05, + "loss": 0.033, + "step": 45860 + }, + { + "epoch": 3.005306951451222, + "grad_norm": 1.06971275806427, + "learning_rate": 6.0876500589558796e-05, + "loss": 0.0351, + "step": 45870 + }, + { + "epoch": 3.0059621306427307, + "grad_norm": 0.9010985493659973, + "learning_rate": 6.086036124291248e-05, + "loss": 0.0348, + "step": 45880 + }, + { + "epoch": 3.0066173098342395, + "grad_norm": 0.5996993184089661, + "learning_rate": 6.084422070859339e-05, + "loss": 0.0427, + "step": 45890 + }, + { + "epoch": 3.0072724890257487, + "grad_norm": 0.3912021219730377, + "learning_rate": 6.082807898836663e-05, + "loss": 0.0298, + "step": 45900 + }, + { + "epoch": 3.0079276682172575, + "grad_norm": 0.5331597924232483, + "learning_rate": 6.081193608399742e-05, + "loss": 0.0496, + "step": 45910 + }, + { + "epoch": 3.0085828474087664, + "grad_norm": 0.2992759048938751, + "learning_rate": 6.0795791997251164e-05, + "loss": 0.0308, + "step": 45920 + }, + { + "epoch": 3.009238026600275, + "grad_norm": 0.8072065114974976, + "learning_rate": 6.0779646729893294e-05, + "loss": 0.0243, + "step": 45930 + }, + { + "epoch": 3.009893205791784, + "grad_norm": 0.25372305512428284, + "learning_rate": 6.0763500283689476e-05, + "loss": 0.0382, + "step": 45940 + }, + { + "epoch": 3.010548384983293, + "grad_norm": 0.7602366805076599, + "learning_rate": 6.0747352660405455e-05, + "loss": 0.0291, + "step": 45950 + }, + { + "epoch": 3.0112035641748016, + "grad_norm": 0.34957659244537354, + "learning_rate": 6.073120386180709e-05, + "loss": 0.0549, + "step": 45960 + }, + { + "epoch": 3.011858743366311, + "grad_norm": 1.0408477783203125, + "learning_rate": 6.0715053889660425e-05, + "loss": 0.0229, + "step": 45970 + }, + { + "epoch": 3.0125139225578197, + "grad_norm": 0.21484386920928955, + "learning_rate": 6.069890274573157e-05, + "loss": 0.0176, + "step": 45980 + }, + { + "epoch": 3.0131691017493285, + "grad_norm": 0.5123571753501892, + "learning_rate": 6.068275043178679e-05, + "loss": 0.0495, + "step": 45990 + }, + { + "epoch": 3.0138242809408373, + "grad_norm": 0.4342992603778839, + "learning_rate": 6.066659694959248e-05, + "loss": 0.0233, + "step": 46000 + }, + { + "epoch": 3.014479460132346, + "grad_norm": 0.24319718778133392, + "learning_rate": 6.0650442300915176e-05, + "loss": 0.0312, + "step": 46010 + }, + { + "epoch": 3.015134639323855, + "grad_norm": 0.3460446298122406, + "learning_rate": 6.063428648752152e-05, + "loss": 0.0239, + "step": 46020 + }, + { + "epoch": 3.0157898185153638, + "grad_norm": 0.15535616874694824, + "learning_rate": 6.0618129511178266e-05, + "loss": 0.0256, + "step": 46030 + }, + { + "epoch": 3.016444997706873, + "grad_norm": 0.2493705302476883, + "learning_rate": 6.060197137365234e-05, + "loss": 0.0343, + "step": 46040 + }, + { + "epoch": 3.017100176898382, + "grad_norm": 0.5880738496780396, + "learning_rate": 6.058581207671077e-05, + "loss": 0.0253, + "step": 46050 + }, + { + "epoch": 3.0177553560898907, + "grad_norm": 0.53461754322052, + "learning_rate": 6.056965162212072e-05, + "loss": 0.045, + "step": 46060 + }, + { + "epoch": 3.0184105352813995, + "grad_norm": 0.6182203888893127, + "learning_rate": 6.055349001164943e-05, + "loss": 0.0398, + "step": 46070 + }, + { + "epoch": 3.0190657144729083, + "grad_norm": 0.10476017743349075, + "learning_rate": 6.0537327247064347e-05, + "loss": 0.0166, + "step": 46080 + }, + { + "epoch": 3.019720893664417, + "grad_norm": 0.1550813466310501, + "learning_rate": 6.0521163330133014e-05, + "loss": 0.0347, + "step": 46090 + }, + { + "epoch": 3.020376072855926, + "grad_norm": 0.7690601348876953, + "learning_rate": 6.050499826262306e-05, + "loss": 0.0311, + "step": 46100 + }, + { + "epoch": 3.021031252047435, + "grad_norm": 0.11972104012966156, + "learning_rate": 6.0488832046302294e-05, + "loss": 0.0365, + "step": 46110 + }, + { + "epoch": 3.021686431238944, + "grad_norm": 0.6018096208572388, + "learning_rate": 6.0472664682938626e-05, + "loss": 0.0387, + "step": 46120 + }, + { + "epoch": 3.022341610430453, + "grad_norm": 0.7168620824813843, + "learning_rate": 6.045649617430009e-05, + "loss": 0.0441, + "step": 46130 + }, + { + "epoch": 3.0229967896219616, + "grad_norm": 0.25567924976348877, + "learning_rate": 6.0440326522154866e-05, + "loss": 0.0245, + "step": 46140 + }, + { + "epoch": 3.0236519688134704, + "grad_norm": 1.770220160484314, + "learning_rate": 6.0424155728271224e-05, + "loss": 0.0464, + "step": 46150 + }, + { + "epoch": 3.0243071480049792, + "grad_norm": 0.31524255871772766, + "learning_rate": 6.040798379441758e-05, + "loss": 0.0392, + "step": 46160 + }, + { + "epoch": 3.024962327196488, + "grad_norm": 0.28950121998786926, + "learning_rate": 6.0391810722362485e-05, + "loss": 0.0507, + "step": 46170 + }, + { + "epoch": 3.0256175063879973, + "grad_norm": 0.4341928958892822, + "learning_rate": 6.037563651387458e-05, + "loss": 0.0448, + "step": 46180 + }, + { + "epoch": 3.026272685579506, + "grad_norm": 0.2811853885650635, + "learning_rate": 6.0359461170722666e-05, + "loss": 0.0266, + "step": 46190 + }, + { + "epoch": 3.026927864771015, + "grad_norm": 1.1582633256912231, + "learning_rate": 6.034328469467566e-05, + "loss": 0.0577, + "step": 46200 + }, + { + "epoch": 3.0275830439625238, + "grad_norm": 0.6842330098152161, + "learning_rate": 6.0327107087502596e-05, + "loss": 0.0225, + "step": 46210 + }, + { + "epoch": 3.0282382231540326, + "grad_norm": 0.16618923842906952, + "learning_rate": 6.031092835097262e-05, + "loss": 0.0575, + "step": 46220 + }, + { + "epoch": 3.0288934023455414, + "grad_norm": 0.4319392442703247, + "learning_rate": 6.0294748486855024e-05, + "loss": 0.0346, + "step": 46230 + }, + { + "epoch": 3.02954858153705, + "grad_norm": 0.3615139126777649, + "learning_rate": 6.0278567496919216e-05, + "loss": 0.0297, + "step": 46240 + }, + { + "epoch": 3.0302037607285595, + "grad_norm": 0.10341528058052063, + "learning_rate": 6.026238538293472e-05, + "loss": 0.0257, + "step": 46250 + }, + { + "epoch": 3.0308589399200683, + "grad_norm": 0.2079574316740036, + "learning_rate": 6.024620214667118e-05, + "loss": 0.0295, + "step": 46260 + }, + { + "epoch": 3.031514119111577, + "grad_norm": 0.831721842288971, + "learning_rate": 6.0230017789898384e-05, + "loss": 0.0475, + "step": 46270 + }, + { + "epoch": 3.032169298303086, + "grad_norm": 0.6653355360031128, + "learning_rate": 6.021383231438622e-05, + "loss": 0.0349, + "step": 46280 + }, + { + "epoch": 3.0328244774945947, + "grad_norm": 0.7239329814910889, + "learning_rate": 6.0197645721904704e-05, + "loss": 0.0434, + "step": 46290 + }, + { + "epoch": 3.0334796566861035, + "grad_norm": 0.5548384785652161, + "learning_rate": 6.0181458014224e-05, + "loss": 0.0335, + "step": 46300 + }, + { + "epoch": 3.0341348358776123, + "grad_norm": 0.2954607903957367, + "learning_rate": 6.016526919311435e-05, + "loss": 0.0446, + "step": 46310 + }, + { + "epoch": 3.0347900150691216, + "grad_norm": 1.0853464603424072, + "learning_rate": 6.014907926034613e-05, + "loss": 0.0402, + "step": 46320 + }, + { + "epoch": 3.0354451942606304, + "grad_norm": 0.5173609852790833, + "learning_rate": 6.0132888217689875e-05, + "loss": 0.0513, + "step": 46330 + }, + { + "epoch": 3.0361003734521392, + "grad_norm": 0.1564335972070694, + "learning_rate": 6.01166960669162e-05, + "loss": 0.028, + "step": 46340 + }, + { + "epoch": 3.036755552643648, + "grad_norm": 0.9236352443695068, + "learning_rate": 6.0100502809795845e-05, + "loss": 0.0455, + "step": 46350 + }, + { + "epoch": 3.037410731835157, + "grad_norm": 0.29269641637802124, + "learning_rate": 6.00843084480997e-05, + "loss": 0.0393, + "step": 46360 + }, + { + "epoch": 3.0380659110266657, + "grad_norm": 0.44081926345825195, + "learning_rate": 6.0068112983598736e-05, + "loss": 0.0497, + "step": 46370 + }, + { + "epoch": 3.0387210902181745, + "grad_norm": 0.33059078454971313, + "learning_rate": 6.005191641806407e-05, + "loss": 0.0304, + "step": 46380 + }, + { + "epoch": 3.0393762694096838, + "grad_norm": 0.25995564460754395, + "learning_rate": 6.003571875326694e-05, + "loss": 0.039, + "step": 46390 + }, + { + "epoch": 3.0400314486011926, + "grad_norm": 1.000813364982605, + "learning_rate": 6.00195199909787e-05, + "loss": 0.0496, + "step": 46400 + }, + { + "epoch": 3.0406866277927014, + "grad_norm": 0.7275881171226501, + "learning_rate": 6.000332013297082e-05, + "loss": 0.0351, + "step": 46410 + }, + { + "epoch": 3.04134180698421, + "grad_norm": 0.30454370379447937, + "learning_rate": 5.998711918101487e-05, + "loss": 0.053, + "step": 46420 + }, + { + "epoch": 3.041996986175719, + "grad_norm": 0.46624812483787537, + "learning_rate": 5.997091713688261e-05, + "loss": 0.0395, + "step": 46430 + }, + { + "epoch": 3.042652165367228, + "grad_norm": 0.14701157808303833, + "learning_rate": 5.9954714002345836e-05, + "loss": 0.0431, + "step": 46440 + }, + { + "epoch": 3.0433073445587366, + "grad_norm": 0.1340833306312561, + "learning_rate": 5.993850977917649e-05, + "loss": 0.0519, + "step": 46450 + }, + { + "epoch": 3.043962523750246, + "grad_norm": 0.17461195588111877, + "learning_rate": 5.992230446914667e-05, + "loss": 0.0307, + "step": 46460 + }, + { + "epoch": 3.0446177029417547, + "grad_norm": 0.24359038472175598, + "learning_rate": 5.990609807402855e-05, + "loss": 0.0277, + "step": 46470 + }, + { + "epoch": 3.0452728821332635, + "grad_norm": 0.6463831067085266, + "learning_rate": 5.988989059559443e-05, + "loss": 0.06, + "step": 46480 + }, + { + "epoch": 3.0459280613247723, + "grad_norm": 0.6787002682685852, + "learning_rate": 5.987368203561675e-05, + "loss": 0.0292, + "step": 46490 + }, + { + "epoch": 3.046583240516281, + "grad_norm": 0.3826221525669098, + "learning_rate": 5.9857472395868055e-05, + "loss": 0.0489, + "step": 46500 + }, + { + "epoch": 3.04723841970779, + "grad_norm": 0.8574100136756897, + "learning_rate": 5.9841261678120983e-05, + "loss": 0.0781, + "step": 46510 + }, + { + "epoch": 3.047893598899299, + "grad_norm": 0.07920319586992264, + "learning_rate": 5.982504988414834e-05, + "loss": 0.0372, + "step": 46520 + }, + { + "epoch": 3.048548778090808, + "grad_norm": 0.24362704157829285, + "learning_rate": 5.9808837015723015e-05, + "loss": 0.0303, + "step": 46530 + }, + { + "epoch": 3.049203957282317, + "grad_norm": 0.5285578966140747, + "learning_rate": 5.9792623074618016e-05, + "loss": 0.0231, + "step": 46540 + }, + { + "epoch": 3.0498591364738257, + "grad_norm": 0.7632954120635986, + "learning_rate": 5.977640806260648e-05, + "loss": 0.0476, + "step": 46550 + }, + { + "epoch": 3.0505143156653345, + "grad_norm": 0.542673647403717, + "learning_rate": 5.976019198146166e-05, + "loss": 0.0228, + "step": 46560 + }, + { + "epoch": 3.0511694948568433, + "grad_norm": 0.25844213366508484, + "learning_rate": 5.974397483295692e-05, + "loss": 0.0195, + "step": 46570 + }, + { + "epoch": 3.051824674048352, + "grad_norm": 0.1412280648946762, + "learning_rate": 5.972775661886575e-05, + "loss": 0.0312, + "step": 46580 + }, + { + "epoch": 3.052479853239861, + "grad_norm": 0.4222421944141388, + "learning_rate": 5.971153734096173e-05, + "loss": 0.0365, + "step": 46590 + }, + { + "epoch": 3.05313503243137, + "grad_norm": 0.3711208999156952, + "learning_rate": 5.969531700101859e-05, + "loss": 0.0458, + "step": 46600 + }, + { + "epoch": 3.053790211622879, + "grad_norm": 0.15220893919467926, + "learning_rate": 5.9679095600810155e-05, + "loss": 0.0296, + "step": 46610 + }, + { + "epoch": 3.054445390814388, + "grad_norm": 0.28810182213783264, + "learning_rate": 5.9662873142110384e-05, + "loss": 0.0355, + "step": 46620 + }, + { + "epoch": 3.0551005700058966, + "grad_norm": 0.7097155451774597, + "learning_rate": 5.964664962669333e-05, + "loss": 0.0456, + "step": 46630 + }, + { + "epoch": 3.0557557491974054, + "grad_norm": 0.4313272535800934, + "learning_rate": 5.9630425056333186e-05, + "loss": 0.0465, + "step": 46640 + }, + { + "epoch": 3.0564109283889143, + "grad_norm": 0.5001418590545654, + "learning_rate": 5.961419943280422e-05, + "loss": 0.033, + "step": 46650 + }, + { + "epoch": 3.057066107580423, + "grad_norm": 0.5705667734146118, + "learning_rate": 5.959797275788087e-05, + "loss": 0.0317, + "step": 46660 + }, + { + "epoch": 3.0577212867719323, + "grad_norm": 0.0470898263156414, + "learning_rate": 5.958174503333765e-05, + "loss": 0.0224, + "step": 46670 + }, + { + "epoch": 3.058376465963441, + "grad_norm": 0.13553401827812195, + "learning_rate": 5.9565516260949195e-05, + "loss": 0.02, + "step": 46680 + }, + { + "epoch": 3.05903164515495, + "grad_norm": 0.5679879784584045, + "learning_rate": 5.9549286442490273e-05, + "loss": 0.0394, + "step": 46690 + }, + { + "epoch": 3.0596868243464588, + "grad_norm": 0.16296330094337463, + "learning_rate": 5.953305557973572e-05, + "loss": 0.0322, + "step": 46700 + }, + { + "epoch": 3.0603420035379676, + "grad_norm": 0.5003370046615601, + "learning_rate": 5.9516823674460564e-05, + "loss": 0.0379, + "step": 46710 + }, + { + "epoch": 3.0609971827294764, + "grad_norm": 0.25365734100341797, + "learning_rate": 5.9500590728439875e-05, + "loss": 0.0515, + "step": 46720 + }, + { + "epoch": 3.061652361920985, + "grad_norm": 0.27978992462158203, + "learning_rate": 5.948435674344886e-05, + "loss": 0.0304, + "step": 46730 + }, + { + "epoch": 3.0623075411124945, + "grad_norm": 0.8158338069915771, + "learning_rate": 5.946812172126285e-05, + "loss": 0.0309, + "step": 46740 + }, + { + "epoch": 3.0629627203040033, + "grad_norm": 0.41522327065467834, + "learning_rate": 5.945188566365729e-05, + "loss": 0.0297, + "step": 46750 + }, + { + "epoch": 3.063617899495512, + "grad_norm": 0.23943258821964264, + "learning_rate": 5.943564857240773e-05, + "loss": 0.0291, + "step": 46760 + }, + { + "epoch": 3.064273078687021, + "grad_norm": 0.5055270195007324, + "learning_rate": 5.941941044928981e-05, + "loss": 0.0571, + "step": 46770 + }, + { + "epoch": 3.0649282578785297, + "grad_norm": 0.5566695332527161, + "learning_rate": 5.940317129607935e-05, + "loss": 0.0373, + "step": 46780 + }, + { + "epoch": 3.0655834370700386, + "grad_norm": 0.49293985962867737, + "learning_rate": 5.9386931114552204e-05, + "loss": 0.052, + "step": 46790 + }, + { + "epoch": 3.0662386162615474, + "grad_norm": 0.19618555903434753, + "learning_rate": 5.937068990648438e-05, + "loss": 0.0257, + "step": 46800 + }, + { + "epoch": 3.0668937954530566, + "grad_norm": 0.7990381717681885, + "learning_rate": 5.935444767365199e-05, + "loss": 0.0324, + "step": 46810 + }, + { + "epoch": 3.0675489746445654, + "grad_norm": 0.20173408091068268, + "learning_rate": 5.933820441783129e-05, + "loss": 0.0341, + "step": 46820 + }, + { + "epoch": 3.0682041538360743, + "grad_norm": 0.2897605001926422, + "learning_rate": 5.932196014079857e-05, + "loss": 0.0255, + "step": 46830 + }, + { + "epoch": 3.068859333027583, + "grad_norm": 0.4367166757583618, + "learning_rate": 5.930571484433032e-05, + "loss": 0.0276, + "step": 46840 + }, + { + "epoch": 3.069514512219092, + "grad_norm": 0.9671338200569153, + "learning_rate": 5.928946853020309e-05, + "loss": 0.0541, + "step": 46850 + }, + { + "epoch": 3.0701696914106007, + "grad_norm": 0.15730537474155426, + "learning_rate": 5.927322120019352e-05, + "loss": 0.0253, + "step": 46860 + }, + { + "epoch": 3.0708248706021095, + "grad_norm": 0.5559119582176208, + "learning_rate": 5.925697285607844e-05, + "loss": 0.0288, + "step": 46870 + }, + { + "epoch": 3.0714800497936183, + "grad_norm": 0.4053433835506439, + "learning_rate": 5.924072349963472e-05, + "loss": 0.0423, + "step": 46880 + }, + { + "epoch": 3.0721352289851276, + "grad_norm": 0.6969658732414246, + "learning_rate": 5.922447313263938e-05, + "loss": 0.0278, + "step": 46890 + }, + { + "epoch": 3.0727904081766364, + "grad_norm": 1.6005762815475464, + "learning_rate": 5.920822175686952e-05, + "loss": 0.0275, + "step": 46900 + }, + { + "epoch": 3.073445587368145, + "grad_norm": 0.06831604242324829, + "learning_rate": 5.919196937410239e-05, + "loss": 0.0461, + "step": 46910 + }, + { + "epoch": 3.074100766559654, + "grad_norm": 0.26904556155204773, + "learning_rate": 5.917571598611529e-05, + "loss": 0.0287, + "step": 46920 + }, + { + "epoch": 3.074755945751163, + "grad_norm": 0.4005871117115021, + "learning_rate": 5.9159461594685686e-05, + "loss": 0.0324, + "step": 46930 + }, + { + "epoch": 3.0754111249426717, + "grad_norm": 0.25650379061698914, + "learning_rate": 5.914320620159114e-05, + "loss": 0.0141, + "step": 46940 + }, + { + "epoch": 3.076066304134181, + "grad_norm": 0.20298655331134796, + "learning_rate": 5.912694980860931e-05, + "loss": 0.0395, + "step": 46950 + }, + { + "epoch": 3.0767214833256897, + "grad_norm": 0.705146074295044, + "learning_rate": 5.9110692417517964e-05, + "loss": 0.0491, + "step": 46960 + }, + { + "epoch": 3.0773766625171985, + "grad_norm": 0.5113490223884583, + "learning_rate": 5.9094434030095e-05, + "loss": 0.0408, + "step": 46970 + }, + { + "epoch": 3.0780318417087074, + "grad_norm": 0.21633270382881165, + "learning_rate": 5.907817464811841e-05, + "loss": 0.0283, + "step": 46980 + }, + { + "epoch": 3.078687020900216, + "grad_norm": 0.17040112614631653, + "learning_rate": 5.906191427336627e-05, + "loss": 0.0883, + "step": 46990 + }, + { + "epoch": 3.079342200091725, + "grad_norm": 0.3382056653499603, + "learning_rate": 5.9045652907616835e-05, + "loss": 0.0258, + "step": 47000 + }, + { + "epoch": 3.079997379283234, + "grad_norm": 2.151987075805664, + "learning_rate": 5.902939055264838e-05, + "loss": 0.0621, + "step": 47010 + }, + { + "epoch": 3.0806525584747426, + "grad_norm": 0.19307619333267212, + "learning_rate": 5.901312721023935e-05, + "loss": 0.0347, + "step": 47020 + }, + { + "epoch": 3.081307737666252, + "grad_norm": 0.2823549211025238, + "learning_rate": 5.8996862882168294e-05, + "loss": 0.0283, + "step": 47030 + }, + { + "epoch": 3.0819629168577607, + "grad_norm": 0.08511023968458176, + "learning_rate": 5.8980597570213826e-05, + "loss": 0.0333, + "step": 47040 + }, + { + "epoch": 3.0826180960492695, + "grad_norm": 0.4239364266395569, + "learning_rate": 5.896433127615471e-05, + "loss": 0.0505, + "step": 47050 + }, + { + "epoch": 3.0832732752407783, + "grad_norm": 0.23000168800354004, + "learning_rate": 5.894806400176981e-05, + "loss": 0.0298, + "step": 47060 + }, + { + "epoch": 3.083928454432287, + "grad_norm": 0.8131154179573059, + "learning_rate": 5.893179574883808e-05, + "loss": 0.0209, + "step": 47070 + }, + { + "epoch": 3.084583633623796, + "grad_norm": 0.39119753241539, + "learning_rate": 5.8915526519138585e-05, + "loss": 0.0371, + "step": 47080 + }, + { + "epoch": 3.085238812815305, + "grad_norm": 0.3212124705314636, + "learning_rate": 5.889925631445053e-05, + "loss": 0.0206, + "step": 47090 + }, + { + "epoch": 3.085893992006814, + "grad_norm": 0.17704814672470093, + "learning_rate": 5.888298513655318e-05, + "loss": 0.0301, + "step": 47100 + }, + { + "epoch": 3.086549171198323, + "grad_norm": 0.4582884609699249, + "learning_rate": 5.886671298722592e-05, + "loss": 0.0319, + "step": 47110 + }, + { + "epoch": 3.0872043503898317, + "grad_norm": 0.4551258385181427, + "learning_rate": 5.885043986824828e-05, + "loss": 0.0297, + "step": 47120 + }, + { + "epoch": 3.0878595295813405, + "grad_norm": 0.671844482421875, + "learning_rate": 5.883416578139982e-05, + "loss": 0.0362, + "step": 47130 + }, + { + "epoch": 3.0885147087728493, + "grad_norm": 0.4757002890110016, + "learning_rate": 5.881789072846029e-05, + "loss": 0.0337, + "step": 47140 + }, + { + "epoch": 3.089169887964358, + "grad_norm": 0.18688848614692688, + "learning_rate": 5.8801614711209474e-05, + "loss": 0.0368, + "step": 47150 + }, + { + "epoch": 3.089825067155867, + "grad_norm": 0.15696488320827484, + "learning_rate": 5.878533773142732e-05, + "loss": 0.0389, + "step": 47160 + }, + { + "epoch": 3.090480246347376, + "grad_norm": 0.4244244396686554, + "learning_rate": 5.8769059790893834e-05, + "loss": 0.0347, + "step": 47170 + }, + { + "epoch": 3.091135425538885, + "grad_norm": 0.1950490027666092, + "learning_rate": 5.8752780891389145e-05, + "loss": 0.0448, + "step": 47180 + }, + { + "epoch": 3.091790604730394, + "grad_norm": 0.10467971116304398, + "learning_rate": 5.873650103469351e-05, + "loss": 0.03, + "step": 47190 + }, + { + "epoch": 3.0924457839219026, + "grad_norm": 0.4917903542518616, + "learning_rate": 5.872022022258725e-05, + "loss": 0.0388, + "step": 47200 + }, + { + "epoch": 3.0931009631134114, + "grad_norm": 0.81443852186203, + "learning_rate": 5.8703938456850814e-05, + "loss": 0.047, + "step": 47210 + }, + { + "epoch": 3.0937561423049202, + "grad_norm": 0.7761908769607544, + "learning_rate": 5.868765573926477e-05, + "loss": 0.0556, + "step": 47220 + }, + { + "epoch": 3.0944113214964295, + "grad_norm": 0.8396887183189392, + "learning_rate": 5.867137207160974e-05, + "loss": 0.0482, + "step": 47230 + }, + { + "epoch": 3.0950665006879383, + "grad_norm": 0.35780930519104004, + "learning_rate": 5.86550874556665e-05, + "loss": 0.0308, + "step": 47240 + }, + { + "epoch": 3.095721679879447, + "grad_norm": 0.2790321409702301, + "learning_rate": 5.863880189321592e-05, + "loss": 0.0376, + "step": 47250 + }, + { + "epoch": 3.096376859070956, + "grad_norm": 1.1050634384155273, + "learning_rate": 5.862251538603894e-05, + "loss": 0.0492, + "step": 47260 + }, + { + "epoch": 3.0970320382624648, + "grad_norm": 0.46163344383239746, + "learning_rate": 5.8606227935916656e-05, + "loss": 0.041, + "step": 47270 + }, + { + "epoch": 3.0976872174539736, + "grad_norm": 0.4563879072666168, + "learning_rate": 5.858993954463021e-05, + "loss": 0.0222, + "step": 47280 + }, + { + "epoch": 3.0983423966454824, + "grad_norm": 0.5637050867080688, + "learning_rate": 5.8573650213960896e-05, + "loss": 0.034, + "step": 47290 + }, + { + "epoch": 3.098997575836991, + "grad_norm": 1.4856325387954712, + "learning_rate": 5.8557359945690084e-05, + "loss": 0.0366, + "step": 47300 + }, + { + "epoch": 3.0996527550285005, + "grad_norm": 0.8817165493965149, + "learning_rate": 5.8541068741599246e-05, + "loss": 0.0779, + "step": 47310 + }, + { + "epoch": 3.1003079342200093, + "grad_norm": 1.283309817314148, + "learning_rate": 5.8524776603469985e-05, + "loss": 0.0507, + "step": 47320 + }, + { + "epoch": 3.100963113411518, + "grad_norm": 0.3301776945590973, + "learning_rate": 5.850848353308397e-05, + "loss": 0.0265, + "step": 47330 + }, + { + "epoch": 3.101618292603027, + "grad_norm": 0.13096071779727936, + "learning_rate": 5.849218953222297e-05, + "loss": 0.0441, + "step": 47340 + }, + { + "epoch": 3.1022734717945357, + "grad_norm": 0.646617591381073, + "learning_rate": 5.847589460266891e-05, + "loss": 0.0312, + "step": 47350 + }, + { + "epoch": 3.1029286509860445, + "grad_norm": 0.2308371365070343, + "learning_rate": 5.845959874620375e-05, + "loss": 0.0519, + "step": 47360 + }, + { + "epoch": 3.1035838301775533, + "grad_norm": 0.3463890254497528, + "learning_rate": 5.84433019646096e-05, + "loss": 0.0232, + "step": 47370 + }, + { + "epoch": 3.1042390093690626, + "grad_norm": 0.789427638053894, + "learning_rate": 5.842700425966863e-05, + "loss": 0.0337, + "step": 47380 + }, + { + "epoch": 3.1048941885605714, + "grad_norm": 0.3764660358428955, + "learning_rate": 5.841070563316315e-05, + "loss": 0.0525, + "step": 47390 + }, + { + "epoch": 3.1055493677520802, + "grad_norm": 2.2431130409240723, + "learning_rate": 5.839440608687554e-05, + "loss": 0.0475, + "step": 47400 + }, + { + "epoch": 3.106204546943589, + "grad_norm": 0.6944766044616699, + "learning_rate": 5.837810562258831e-05, + "loss": 0.0332, + "step": 47410 + }, + { + "epoch": 3.106859726135098, + "grad_norm": 1.4328041076660156, + "learning_rate": 5.836180424208405e-05, + "loss": 0.0641, + "step": 47420 + }, + { + "epoch": 3.1075149053266067, + "grad_norm": 0.6500887870788574, + "learning_rate": 5.834550194714543e-05, + "loss": 0.0432, + "step": 47430 + }, + { + "epoch": 3.1081700845181155, + "grad_norm": 0.25420984625816345, + "learning_rate": 5.832919873955528e-05, + "loss": 0.0397, + "step": 47440 + }, + { + "epoch": 3.1088252637096248, + "grad_norm": 0.6962030529975891, + "learning_rate": 5.831289462109648e-05, + "loss": 0.0403, + "step": 47450 + }, + { + "epoch": 3.1094804429011336, + "grad_norm": 0.23144927620887756, + "learning_rate": 5.829658959355202e-05, + "loss": 0.0225, + "step": 47460 + }, + { + "epoch": 3.1101356220926424, + "grad_norm": 0.16833139955997467, + "learning_rate": 5.8280283658704995e-05, + "loss": 0.0295, + "step": 47470 + }, + { + "epoch": 3.110790801284151, + "grad_norm": 0.17635543644428253, + "learning_rate": 5.826397681833859e-05, + "loss": 0.036, + "step": 47480 + }, + { + "epoch": 3.11144598047566, + "grad_norm": 0.23583176732063293, + "learning_rate": 5.824766907423612e-05, + "loss": 0.0618, + "step": 47490 + }, + { + "epoch": 3.112101159667169, + "grad_norm": 0.18731026351451874, + "learning_rate": 5.8231360428180945e-05, + "loss": 0.0416, + "step": 47500 + }, + { + "epoch": 3.1127563388586776, + "grad_norm": 0.20287178456783295, + "learning_rate": 5.821505088195658e-05, + "loss": 0.0214, + "step": 47510 + }, + { + "epoch": 3.113411518050187, + "grad_norm": 0.3526120185852051, + "learning_rate": 5.819874043734661e-05, + "loss": 0.0366, + "step": 47520 + }, + { + "epoch": 3.1140666972416957, + "grad_norm": 0.3719784617424011, + "learning_rate": 5.8182429096134695e-05, + "loss": 0.0453, + "step": 47530 + }, + { + "epoch": 3.1147218764332045, + "grad_norm": 0.40102308988571167, + "learning_rate": 5.816611686010465e-05, + "loss": 0.0455, + "step": 47540 + }, + { + "epoch": 3.1153770556247133, + "grad_norm": 0.593958854675293, + "learning_rate": 5.814980373104033e-05, + "loss": 0.0393, + "step": 47550 + }, + { + "epoch": 3.116032234816222, + "grad_norm": 0.5581513047218323, + "learning_rate": 5.813348971072572e-05, + "loss": 0.0333, + "step": 47560 + }, + { + "epoch": 3.116687414007731, + "grad_norm": 0.29884278774261475, + "learning_rate": 5.811717480094492e-05, + "loss": 0.0342, + "step": 47570 + }, + { + "epoch": 3.11734259319924, + "grad_norm": 1.0171430110931396, + "learning_rate": 5.810085900348209e-05, + "loss": 0.0573, + "step": 47580 + }, + { + "epoch": 3.117997772390749, + "grad_norm": 0.8626231551170349, + "learning_rate": 5.8084542320121483e-05, + "loss": 0.0317, + "step": 47590 + }, + { + "epoch": 3.118652951582258, + "grad_norm": 0.7746010422706604, + "learning_rate": 5.8068224752647497e-05, + "loss": 0.0462, + "step": 47600 + }, + { + "epoch": 3.1193081307737667, + "grad_norm": 0.16809755563735962, + "learning_rate": 5.805190630284457e-05, + "loss": 0.0545, + "step": 47610 + }, + { + "epoch": 3.1199633099652755, + "grad_norm": 0.9009044170379639, + "learning_rate": 5.803558697249729e-05, + "loss": 0.0382, + "step": 47620 + }, + { + "epoch": 3.1206184891567843, + "grad_norm": 0.2739253044128418, + "learning_rate": 5.8019266763390276e-05, + "loss": 0.0396, + "step": 47630 + }, + { + "epoch": 3.121273668348293, + "grad_norm": 0.21425017714500427, + "learning_rate": 5.800294567730833e-05, + "loss": 0.055, + "step": 47640 + }, + { + "epoch": 3.121928847539802, + "grad_norm": 0.6475667953491211, + "learning_rate": 5.798662371603628e-05, + "loss": 0.0238, + "step": 47650 + }, + { + "epoch": 3.122584026731311, + "grad_norm": 0.7883027195930481, + "learning_rate": 5.797030088135904e-05, + "loss": 0.0271, + "step": 47660 + }, + { + "epoch": 3.12323920592282, + "grad_norm": 0.31645143032073975, + "learning_rate": 5.7953977175061705e-05, + "loss": 0.0422, + "step": 47670 + }, + { + "epoch": 3.123894385114329, + "grad_norm": 0.7152258157730103, + "learning_rate": 5.793765259892938e-05, + "loss": 0.0336, + "step": 47680 + }, + { + "epoch": 3.1245495643058376, + "grad_norm": 0.336516797542572, + "learning_rate": 5.792132715474729e-05, + "loss": 0.0206, + "step": 47690 + }, + { + "epoch": 3.1252047434973464, + "grad_norm": 0.9498024582862854, + "learning_rate": 5.790500084430078e-05, + "loss": 0.0344, + "step": 47700 + }, + { + "epoch": 3.1258599226888553, + "grad_norm": 0.14113028347492218, + "learning_rate": 5.7888673669375274e-05, + "loss": 0.0378, + "step": 47710 + }, + { + "epoch": 3.126515101880364, + "grad_norm": 0.41914427280426025, + "learning_rate": 5.787234563175625e-05, + "loss": 0.02, + "step": 47720 + }, + { + "epoch": 3.1271702810718733, + "grad_norm": 0.804885745048523, + "learning_rate": 5.7856016733229355e-05, + "loss": 0.036, + "step": 47730 + }, + { + "epoch": 3.127825460263382, + "grad_norm": 0.5283040404319763, + "learning_rate": 5.7839686975580297e-05, + "loss": 0.044, + "step": 47740 + }, + { + "epoch": 3.128480639454891, + "grad_norm": 0.4162302315235138, + "learning_rate": 5.782335636059484e-05, + "loss": 0.0219, + "step": 47750 + }, + { + "epoch": 3.1291358186464, + "grad_norm": 0.8271237015724182, + "learning_rate": 5.780702489005889e-05, + "loss": 0.0398, + "step": 47760 + }, + { + "epoch": 3.1297909978379086, + "grad_norm": 0.13903376460075378, + "learning_rate": 5.779069256575845e-05, + "loss": 0.0343, + "step": 47770 + }, + { + "epoch": 3.1304461770294174, + "grad_norm": 0.3908008933067322, + "learning_rate": 5.7774359389479574e-05, + "loss": 0.0295, + "step": 47780 + }, + { + "epoch": 3.1311013562209262, + "grad_norm": 0.43140608072280884, + "learning_rate": 5.775802536300845e-05, + "loss": 0.0351, + "step": 47790 + }, + { + "epoch": 3.1317565354124355, + "grad_norm": 0.8621177077293396, + "learning_rate": 5.774169048813134e-05, + "loss": 0.0372, + "step": 47800 + }, + { + "epoch": 3.1324117146039443, + "grad_norm": 0.41341838240623474, + "learning_rate": 5.77253547666346e-05, + "loss": 0.0206, + "step": 47810 + }, + { + "epoch": 3.133066893795453, + "grad_norm": 0.33142170310020447, + "learning_rate": 5.770901820030465e-05, + "loss": 0.0319, + "step": 47820 + }, + { + "epoch": 3.133722072986962, + "grad_norm": 0.3318479061126709, + "learning_rate": 5.769268079092809e-05, + "loss": 0.0228, + "step": 47830 + }, + { + "epoch": 3.1343772521784707, + "grad_norm": 0.5515789985656738, + "learning_rate": 5.767634254029151e-05, + "loss": 0.0356, + "step": 47840 + }, + { + "epoch": 3.1350324313699796, + "grad_norm": 0.3447321355342865, + "learning_rate": 5.7660003450181655e-05, + "loss": 0.0382, + "step": 47850 + }, + { + "epoch": 3.1356876105614884, + "grad_norm": 0.9115360975265503, + "learning_rate": 5.764366352238534e-05, + "loss": 0.0293, + "step": 47860 + }, + { + "epoch": 3.1363427897529976, + "grad_norm": 0.32555466890335083, + "learning_rate": 5.7627322758689474e-05, + "loss": 0.0372, + "step": 47870 + }, + { + "epoch": 3.1369979689445064, + "grad_norm": 0.3260340988636017, + "learning_rate": 5.761098116088105e-05, + "loss": 0.0342, + "step": 47880 + }, + { + "epoch": 3.1376531481360153, + "grad_norm": 0.5753464698791504, + "learning_rate": 5.759463873074717e-05, + "loss": 0.0545, + "step": 47890 + }, + { + "epoch": 3.138308327327524, + "grad_norm": 0.15205338597297668, + "learning_rate": 5.757829547007504e-05, + "loss": 0.0367, + "step": 47900 + }, + { + "epoch": 3.138963506519033, + "grad_norm": 0.09592388570308685, + "learning_rate": 5.756195138065189e-05, + "loss": 0.0296, + "step": 47910 + }, + { + "epoch": 3.1396186857105417, + "grad_norm": 0.403197318315506, + "learning_rate": 5.754560646426511e-05, + "loss": 0.0376, + "step": 47920 + }, + { + "epoch": 3.1402738649020505, + "grad_norm": 0.48531484603881836, + "learning_rate": 5.752926072270216e-05, + "loss": 0.0464, + "step": 47930 + }, + { + "epoch": 3.1409290440935598, + "grad_norm": 0.16327442228794098, + "learning_rate": 5.7512914157750563e-05, + "loss": 0.0551, + "step": 47940 + }, + { + "epoch": 3.1415842232850686, + "grad_norm": 0.4218485355377197, + "learning_rate": 5.749656677119798e-05, + "loss": 0.0623, + "step": 47950 + }, + { + "epoch": 3.1422394024765774, + "grad_norm": 0.962084174156189, + "learning_rate": 5.7480218564832125e-05, + "loss": 0.0394, + "step": 47960 + }, + { + "epoch": 3.142894581668086, + "grad_norm": 0.14911256730556488, + "learning_rate": 5.746386954044082e-05, + "loss": 0.0412, + "step": 47970 + }, + { + "epoch": 3.143549760859595, + "grad_norm": 0.3765876889228821, + "learning_rate": 5.744751969981195e-05, + "loss": 0.0582, + "step": 47980 + }, + { + "epoch": 3.144204940051104, + "grad_norm": 0.08584430068731308, + "learning_rate": 5.7431169044733526e-05, + "loss": 0.0522, + "step": 47990 + }, + { + "epoch": 3.1448601192426127, + "grad_norm": 1.028907299041748, + "learning_rate": 5.741481757699364e-05, + "loss": 0.0528, + "step": 48000 + }, + { + "epoch": 3.145515298434122, + "grad_norm": 0.47365060448646545, + "learning_rate": 5.7398465298380434e-05, + "loss": 0.0585, + "step": 48010 + }, + { + "epoch": 3.1461704776256307, + "grad_norm": 0.5549283623695374, + "learning_rate": 5.7382112210682193e-05, + "loss": 0.0479, + "step": 48020 + }, + { + "epoch": 3.1468256568171395, + "grad_norm": 1.4726437330245972, + "learning_rate": 5.7365758315687266e-05, + "loss": 0.0407, + "step": 48030 + }, + { + "epoch": 3.1474808360086484, + "grad_norm": 0.24529758095741272, + "learning_rate": 5.734940361518407e-05, + "loss": 0.0261, + "step": 48040 + }, + { + "epoch": 3.148136015200157, + "grad_norm": 0.235584557056427, + "learning_rate": 5.733304811096116e-05, + "loss": 0.0419, + "step": 48050 + }, + { + "epoch": 3.148791194391666, + "grad_norm": 0.6252671480178833, + "learning_rate": 5.731669180480713e-05, + "loss": 0.0432, + "step": 48060 + }, + { + "epoch": 3.149446373583175, + "grad_norm": 0.2484399974346161, + "learning_rate": 5.730033469851067e-05, + "loss": 0.0233, + "step": 48070 + }, + { + "epoch": 3.150101552774684, + "grad_norm": 0.46576109528541565, + "learning_rate": 5.7283976793860607e-05, + "loss": 0.048, + "step": 48080 + }, + { + "epoch": 3.150756731966193, + "grad_norm": 0.3365814685821533, + "learning_rate": 5.726761809264577e-05, + "loss": 0.0323, + "step": 48090 + }, + { + "epoch": 3.1514119111577017, + "grad_norm": 0.544660210609436, + "learning_rate": 5.7251258596655155e-05, + "loss": 0.0355, + "step": 48100 + }, + { + "epoch": 3.1520670903492105, + "grad_norm": 0.4120131731033325, + "learning_rate": 5.72348983076778e-05, + "loss": 0.0188, + "step": 48110 + }, + { + "epoch": 3.1527222695407193, + "grad_norm": 0.10344790667295456, + "learning_rate": 5.7218537227502854e-05, + "loss": 0.0251, + "step": 48120 + }, + { + "epoch": 3.153377448732228, + "grad_norm": 0.15785710513591766, + "learning_rate": 5.720217535791951e-05, + "loss": 0.0321, + "step": 48130 + }, + { + "epoch": 3.154032627923737, + "grad_norm": 0.7153651118278503, + "learning_rate": 5.718581270071711e-05, + "loss": 0.0386, + "step": 48140 + }, + { + "epoch": 3.154687807115246, + "grad_norm": 0.2525821328163147, + "learning_rate": 5.716944925768505e-05, + "loss": 0.0318, + "step": 48150 + }, + { + "epoch": 3.155342986306755, + "grad_norm": 0.24410247802734375, + "learning_rate": 5.7153085030612786e-05, + "loss": 0.0531, + "step": 48160 + }, + { + "epoch": 3.155998165498264, + "grad_norm": 0.24953052401542664, + "learning_rate": 5.713672002128989e-05, + "loss": 0.0255, + "step": 48170 + }, + { + "epoch": 3.1566533446897727, + "grad_norm": 0.2948700189590454, + "learning_rate": 5.712035423150602e-05, + "loss": 0.0223, + "step": 48180 + }, + { + "epoch": 3.1573085238812815, + "grad_norm": 0.6163470149040222, + "learning_rate": 5.710398766305094e-05, + "loss": 0.046, + "step": 48190 + }, + { + "epoch": 3.1579637030727903, + "grad_norm": 0.2819417417049408, + "learning_rate": 5.708762031771442e-05, + "loss": 0.0376, + "step": 48200 + }, + { + "epoch": 3.158618882264299, + "grad_norm": 0.28916555643081665, + "learning_rate": 5.707125219728642e-05, + "loss": 0.0357, + "step": 48210 + }, + { + "epoch": 3.1592740614558084, + "grad_norm": 0.5603504180908203, + "learning_rate": 5.7054883303556905e-05, + "loss": 0.0314, + "step": 48220 + }, + { + "epoch": 3.159929240647317, + "grad_norm": 0.7216112017631531, + "learning_rate": 5.703851363831595e-05, + "loss": 0.0535, + "step": 48230 + }, + { + "epoch": 3.160584419838826, + "grad_norm": 0.5015674233436584, + "learning_rate": 5.702214320335374e-05, + "loss": 0.0206, + "step": 48240 + }, + { + "epoch": 3.161239599030335, + "grad_norm": 1.1458698511123657, + "learning_rate": 5.7005772000460514e-05, + "loss": 0.0513, + "step": 48250 + }, + { + "epoch": 3.1618947782218436, + "grad_norm": 0.16753964126110077, + "learning_rate": 5.698940003142656e-05, + "loss": 0.0736, + "step": 48260 + }, + { + "epoch": 3.1625499574133524, + "grad_norm": 0.44692668318748474, + "learning_rate": 5.697302729804236e-05, + "loss": 0.0596, + "step": 48270 + }, + { + "epoch": 3.1632051366048612, + "grad_norm": 0.8679671287536621, + "learning_rate": 5.695665380209837e-05, + "loss": 0.0401, + "step": 48280 + }, + { + "epoch": 3.1638603157963705, + "grad_norm": 0.11284174770116806, + "learning_rate": 5.6940279545385165e-05, + "loss": 0.0306, + "step": 48290 + }, + { + "epoch": 3.1645154949878793, + "grad_norm": 0.3144579231739044, + "learning_rate": 5.692390452969344e-05, + "loss": 0.0201, + "step": 48300 + }, + { + "epoch": 3.165170674179388, + "grad_norm": 0.54326331615448, + "learning_rate": 5.690752875681392e-05, + "loss": 0.0423, + "step": 48310 + }, + { + "epoch": 3.165825853370897, + "grad_norm": 0.09030642360448837, + "learning_rate": 5.6891152228537435e-05, + "loss": 0.0371, + "step": 48320 + }, + { + "epoch": 3.1664810325624058, + "grad_norm": 1.4295576810836792, + "learning_rate": 5.687477494665492e-05, + "loss": 0.0299, + "step": 48330 + }, + { + "epoch": 3.1671362117539146, + "grad_norm": 1.3488484621047974, + "learning_rate": 5.685839691295733e-05, + "loss": 0.0321, + "step": 48340 + }, + { + "epoch": 3.1677913909454234, + "grad_norm": 1.706958532333374, + "learning_rate": 5.6842018129235786e-05, + "loss": 0.0261, + "step": 48350 + }, + { + "epoch": 3.1684465701369326, + "grad_norm": 0.7445918321609497, + "learning_rate": 5.6825638597281404e-05, + "loss": 0.0475, + "step": 48360 + }, + { + "epoch": 3.1691017493284415, + "grad_norm": 0.43604451417922974, + "learning_rate": 5.680925831888546e-05, + "loss": 0.0243, + "step": 48370 + }, + { + "epoch": 3.1697569285199503, + "grad_norm": 0.752800464630127, + "learning_rate": 5.6792877295839274e-05, + "loss": 0.0501, + "step": 48380 + }, + { + "epoch": 3.170412107711459, + "grad_norm": 0.28219661116600037, + "learning_rate": 5.6776495529934224e-05, + "loss": 0.0364, + "step": 48390 + }, + { + "epoch": 3.171067286902968, + "grad_norm": 0.18879559636116028, + "learning_rate": 5.6760113022961824e-05, + "loss": 0.0403, + "step": 48400 + }, + { + "epoch": 3.1717224660944767, + "grad_norm": 0.6521719694137573, + "learning_rate": 5.6743729776713617e-05, + "loss": 0.0186, + "step": 48410 + }, + { + "epoch": 3.1723776452859855, + "grad_norm": 0.6481947302818298, + "learning_rate": 5.672734579298126e-05, + "loss": 0.0493, + "step": 48420 + }, + { + "epoch": 3.173032824477495, + "grad_norm": 0.30216121673583984, + "learning_rate": 5.671096107355649e-05, + "loss": 0.0213, + "step": 48430 + }, + { + "epoch": 3.1736880036690036, + "grad_norm": 2.288557767868042, + "learning_rate": 5.66945756202311e-05, + "loss": 0.0563, + "step": 48440 + }, + { + "epoch": 3.1743431828605124, + "grad_norm": 0.08084587752819061, + "learning_rate": 5.667818943479699e-05, + "loss": 0.0137, + "step": 48450 + }, + { + "epoch": 3.1749983620520212, + "grad_norm": 0.08864005655050278, + "learning_rate": 5.666180251904612e-05, + "loss": 0.0248, + "step": 48460 + }, + { + "epoch": 3.17565354124353, + "grad_norm": 1.1304699182510376, + "learning_rate": 5.6645414874770555e-05, + "loss": 0.0319, + "step": 48470 + }, + { + "epoch": 3.176308720435039, + "grad_norm": 0.44301745295524597, + "learning_rate": 5.66290265037624e-05, + "loss": 0.0427, + "step": 48480 + }, + { + "epoch": 3.1769638996265477, + "grad_norm": 0.14224684238433838, + "learning_rate": 5.661263740781386e-05, + "loss": 0.0371, + "step": 48490 + }, + { + "epoch": 3.177619078818057, + "grad_norm": 0.2773180305957794, + "learning_rate": 5.6596247588717254e-05, + "loss": 0.0522, + "step": 48500 + }, + { + "epoch": 3.1782742580095658, + "grad_norm": 1.0592803955078125, + "learning_rate": 5.6579857048264926e-05, + "loss": 0.0488, + "step": 48510 + }, + { + "epoch": 3.1789294372010746, + "grad_norm": 0.13364015519618988, + "learning_rate": 5.6563465788249314e-05, + "loss": 0.0316, + "step": 48520 + }, + { + "epoch": 3.1795846163925834, + "grad_norm": 0.643764317035675, + "learning_rate": 5.6547073810462956e-05, + "loss": 0.0375, + "step": 48530 + }, + { + "epoch": 3.180239795584092, + "grad_norm": 0.562703549861908, + "learning_rate": 5.653068111669846e-05, + "loss": 0.0292, + "step": 48540 + }, + { + "epoch": 3.180894974775601, + "grad_norm": 0.34931960701942444, + "learning_rate": 5.651428770874848e-05, + "loss": 0.031, + "step": 48550 + }, + { + "epoch": 3.18155015396711, + "grad_norm": 0.3564545214176178, + "learning_rate": 5.64978935884058e-05, + "loss": 0.0298, + "step": 48560 + }, + { + "epoch": 3.182205333158619, + "grad_norm": 0.8826240301132202, + "learning_rate": 5.6481498757463244e-05, + "loss": 0.0372, + "step": 48570 + }, + { + "epoch": 3.182860512350128, + "grad_norm": 0.3638269305229187, + "learning_rate": 5.646510321771373e-05, + "loss": 0.0208, + "step": 48580 + }, + { + "epoch": 3.1835156915416367, + "grad_norm": 0.20145674049854279, + "learning_rate": 5.644870697095024e-05, + "loss": 0.0477, + "step": 48590 + }, + { + "epoch": 3.1841708707331455, + "grad_norm": 3.1399660110473633, + "learning_rate": 5.643231001896586e-05, + "loss": 0.041, + "step": 48600 + }, + { + "epoch": 3.1848260499246543, + "grad_norm": 0.26053157448768616, + "learning_rate": 5.6415912363553726e-05, + "loss": 0.0225, + "step": 48610 + }, + { + "epoch": 3.185481229116163, + "grad_norm": 0.3953726589679718, + "learning_rate": 5.639951400650706e-05, + "loss": 0.0511, + "step": 48620 + }, + { + "epoch": 3.186136408307672, + "grad_norm": 0.339104026556015, + "learning_rate": 5.6383114949619165e-05, + "loss": 0.016, + "step": 48630 + }, + { + "epoch": 3.1867915874991812, + "grad_norm": 0.2702779173851013, + "learning_rate": 5.636671519468342e-05, + "loss": 0.0321, + "step": 48640 + }, + { + "epoch": 3.18744676669069, + "grad_norm": 0.37370067834854126, + "learning_rate": 5.635031474349327e-05, + "loss": 0.0279, + "step": 48650 + }, + { + "epoch": 3.188101945882199, + "grad_norm": 0.3812744915485382, + "learning_rate": 5.6333913597842246e-05, + "loss": 0.0226, + "step": 48660 + }, + { + "epoch": 3.1887571250737077, + "grad_norm": 0.4976571500301361, + "learning_rate": 5.6317511759523955e-05, + "loss": 0.0388, + "step": 48670 + }, + { + "epoch": 3.1894123042652165, + "grad_norm": 0.8036971688270569, + "learning_rate": 5.630110923033207e-05, + "loss": 0.0569, + "step": 48680 + }, + { + "epoch": 3.1900674834567253, + "grad_norm": 0.2704266309738159, + "learning_rate": 5.628470601206036e-05, + "loss": 0.0346, + "step": 48690 + }, + { + "epoch": 3.190722662648234, + "grad_norm": 0.25767782330513, + "learning_rate": 5.626830210650266e-05, + "loss": 0.035, + "step": 48700 + }, + { + "epoch": 3.191377841839743, + "grad_norm": 0.23826436698436737, + "learning_rate": 5.625189751545285e-05, + "loss": 0.035, + "step": 48710 + }, + { + "epoch": 3.192033021031252, + "grad_norm": 0.7597230076789856, + "learning_rate": 5.6235492240704936e-05, + "loss": 0.0499, + "step": 48720 + }, + { + "epoch": 3.192688200222761, + "grad_norm": 0.45083171129226685, + "learning_rate": 5.621908628405296e-05, + "loss": 0.0381, + "step": 48730 + }, + { + "epoch": 3.19334337941427, + "grad_norm": 0.8247312307357788, + "learning_rate": 5.620267964729106e-05, + "loss": 0.0233, + "step": 48740 + }, + { + "epoch": 3.1939985586057786, + "grad_norm": 0.7507661581039429, + "learning_rate": 5.618627233221344e-05, + "loss": 0.032, + "step": 48750 + }, + { + "epoch": 3.1946537377972875, + "grad_norm": 0.5793591141700745, + "learning_rate": 5.616986434061438e-05, + "loss": 0.0475, + "step": 48760 + }, + { + "epoch": 3.1953089169887963, + "grad_norm": 0.23782828450202942, + "learning_rate": 5.615345567428822e-05, + "loss": 0.0488, + "step": 48770 + }, + { + "epoch": 3.1959640961803055, + "grad_norm": 0.3702459931373596, + "learning_rate": 5.613704633502941e-05, + "loss": 0.06, + "step": 48780 + }, + { + "epoch": 3.1966192753718143, + "grad_norm": 0.695475161075592, + "learning_rate": 5.612063632463245e-05, + "loss": 0.0494, + "step": 48790 + }, + { + "epoch": 3.197274454563323, + "grad_norm": 0.09981531649827957, + "learning_rate": 5.610422564489188e-05, + "loss": 0.0388, + "step": 48800 + }, + { + "epoch": 3.197929633754832, + "grad_norm": 0.6582425236701965, + "learning_rate": 5.608781429760239e-05, + "loss": 0.0492, + "step": 48810 + }, + { + "epoch": 3.198584812946341, + "grad_norm": 0.9599102139472961, + "learning_rate": 5.607140228455866e-05, + "loss": 0.0737, + "step": 48820 + }, + { + "epoch": 3.1992399921378496, + "grad_norm": 0.7840073108673096, + "learning_rate": 5.605498960755553e-05, + "loss": 0.0264, + "step": 48830 + }, + { + "epoch": 3.1998951713293584, + "grad_norm": 0.19489216804504395, + "learning_rate": 5.603857626838782e-05, + "loss": 0.016, + "step": 48840 + }, + { + "epoch": 3.2005503505208672, + "grad_norm": 1.7255760431289673, + "learning_rate": 5.60221622688505e-05, + "loss": 0.0386, + "step": 48850 + }, + { + "epoch": 3.2012055297123765, + "grad_norm": 0.2722025513648987, + "learning_rate": 5.6005747610738565e-05, + "loss": 0.0385, + "step": 48860 + }, + { + "epoch": 3.2018607089038853, + "grad_norm": 0.2530507743358612, + "learning_rate": 5.59893322958471e-05, + "loss": 0.0485, + "step": 48870 + }, + { + "epoch": 3.202515888095394, + "grad_norm": 0.15675051510334015, + "learning_rate": 5.5972916325971256e-05, + "loss": 0.0319, + "step": 48880 + }, + { + "epoch": 3.203171067286903, + "grad_norm": 0.19868072867393494, + "learning_rate": 5.595649970290628e-05, + "loss": 0.0207, + "step": 48890 + }, + { + "epoch": 3.2038262464784117, + "grad_norm": 0.5321448445320129, + "learning_rate": 5.5940082428447426e-05, + "loss": 0.0398, + "step": 48900 + }, + { + "epoch": 3.2044814256699206, + "grad_norm": 0.6274491548538208, + "learning_rate": 5.592366450439012e-05, + "loss": 0.049, + "step": 48910 + }, + { + "epoch": 3.20513660486143, + "grad_norm": 0.6165081858634949, + "learning_rate": 5.590724593252975e-05, + "loss": 0.0294, + "step": 48920 + }, + { + "epoch": 3.2057917840529386, + "grad_norm": 0.6846511363983154, + "learning_rate": 5.589082671466184e-05, + "loss": 0.0277, + "step": 48930 + }, + { + "epoch": 3.2064469632444474, + "grad_norm": 0.19924670457839966, + "learning_rate": 5.587440685258199e-05, + "loss": 0.0299, + "step": 48940 + }, + { + "epoch": 3.2071021424359563, + "grad_norm": 0.8293700814247131, + "learning_rate": 5.585798634808583e-05, + "loss": 0.0261, + "step": 48950 + }, + { + "epoch": 3.207757321627465, + "grad_norm": 0.09523113071918488, + "learning_rate": 5.584156520296909e-05, + "loss": 0.0237, + "step": 48960 + }, + { + "epoch": 3.208412500818974, + "grad_norm": 0.7036954760551453, + "learning_rate": 5.582514341902757e-05, + "loss": 0.0435, + "step": 48970 + }, + { + "epoch": 3.2090676800104827, + "grad_norm": 0.6519138216972351, + "learning_rate": 5.580872099805713e-05, + "loss": 0.0516, + "step": 48980 + }, + { + "epoch": 3.2097228592019915, + "grad_norm": 0.6091548800468445, + "learning_rate": 5.5792297941853674e-05, + "loss": 0.0344, + "step": 48990 + }, + { + "epoch": 3.2103780383935008, + "grad_norm": 0.9115249514579773, + "learning_rate": 5.5775874252213247e-05, + "loss": 0.0185, + "step": 49000 + }, + { + "epoch": 3.2110332175850096, + "grad_norm": 0.835885226726532, + "learning_rate": 5.575944993093189e-05, + "loss": 0.0597, + "step": 49010 + }, + { + "epoch": 3.2116883967765184, + "grad_norm": 0.25563865900039673, + "learning_rate": 5.574302497980574e-05, + "loss": 0.0447, + "step": 49020 + }, + { + "epoch": 3.212343575968027, + "grad_norm": 0.5712826251983643, + "learning_rate": 5.5726599400631005e-05, + "loss": 0.0355, + "step": 49030 + }, + { + "epoch": 3.212998755159536, + "grad_norm": 0.27796468138694763, + "learning_rate": 5.571017319520399e-05, + "loss": 0.0375, + "step": 49040 + }, + { + "epoch": 3.213653934351045, + "grad_norm": 0.47192808985710144, + "learning_rate": 5.5693746365320984e-05, + "loss": 0.0261, + "step": 49050 + }, + { + "epoch": 3.214309113542554, + "grad_norm": 0.6391663551330566, + "learning_rate": 5.567731891277848e-05, + "loss": 0.0433, + "step": 49060 + }, + { + "epoch": 3.214964292734063, + "grad_norm": 1.3374487161636353, + "learning_rate": 5.566089083937288e-05, + "loss": 0.0717, + "step": 49070 + }, + { + "epoch": 3.2156194719255717, + "grad_norm": 0.2933475971221924, + "learning_rate": 5.564446214690079e-05, + "loss": 0.024, + "step": 49080 + }, + { + "epoch": 3.2162746511170806, + "grad_norm": 1.7274242639541626, + "learning_rate": 5.56280328371588e-05, + "loss": 0.0322, + "step": 49090 + }, + { + "epoch": 3.2169298303085894, + "grad_norm": 0.37281107902526855, + "learning_rate": 5.5611602911943596e-05, + "loss": 0.0371, + "step": 49100 + }, + { + "epoch": 3.217585009500098, + "grad_norm": 0.21515525877475739, + "learning_rate": 5.559517237305194e-05, + "loss": 0.0354, + "step": 49110 + }, + { + "epoch": 3.218240188691607, + "grad_norm": 0.17715129256248474, + "learning_rate": 5.557874122228064e-05, + "loss": 0.0355, + "step": 49120 + }, + { + "epoch": 3.218895367883116, + "grad_norm": 0.8490564823150635, + "learning_rate": 5.55623094614266e-05, + "loss": 0.0532, + "step": 49130 + }, + { + "epoch": 3.219550547074625, + "grad_norm": 0.587760865688324, + "learning_rate": 5.5545877092286743e-05, + "loss": 0.0372, + "step": 49140 + }, + { + "epoch": 3.220205726266134, + "grad_norm": 0.46045365929603577, + "learning_rate": 5.552944411665811e-05, + "loss": 0.0452, + "step": 49150 + }, + { + "epoch": 3.2208609054576427, + "grad_norm": 0.15452958643436432, + "learning_rate": 5.5513010536337795e-05, + "loss": 0.0389, + "step": 49160 + }, + { + "epoch": 3.2215160846491515, + "grad_norm": 0.1934976726770401, + "learning_rate": 5.5496576353122934e-05, + "loss": 0.0312, + "step": 49170 + }, + { + "epoch": 3.2221712638406603, + "grad_norm": 0.5334646105766296, + "learning_rate": 5.548014156881074e-05, + "loss": 0.0247, + "step": 49180 + }, + { + "epoch": 3.222826443032169, + "grad_norm": 0.1700599640607834, + "learning_rate": 5.5463706185198494e-05, + "loss": 0.0302, + "step": 49190 + }, + { + "epoch": 3.2234816222236784, + "grad_norm": 1.0080986022949219, + "learning_rate": 5.544727020408358e-05, + "loss": 0.0483, + "step": 49200 + }, + { + "epoch": 3.224136801415187, + "grad_norm": 0.4584976136684418, + "learning_rate": 5.5430833627263366e-05, + "loss": 0.043, + "step": 49210 + }, + { + "epoch": 3.224791980606696, + "grad_norm": 0.6804112195968628, + "learning_rate": 5.5414396456535354e-05, + "loss": 0.0288, + "step": 49220 + }, + { + "epoch": 3.225447159798205, + "grad_norm": 0.17207027971744537, + "learning_rate": 5.539795869369709e-05, + "loss": 0.0403, + "step": 49230 + }, + { + "epoch": 3.2261023389897137, + "grad_norm": 0.12570083141326904, + "learning_rate": 5.538152034054618e-05, + "loss": 0.0331, + "step": 49240 + }, + { + "epoch": 3.2267575181812225, + "grad_norm": 0.4923616051673889, + "learning_rate": 5.536508139888028e-05, + "loss": 0.0369, + "step": 49250 + }, + { + "epoch": 3.2274126973727313, + "grad_norm": 0.4551827013492584, + "learning_rate": 5.534864187049716e-05, + "loss": 0.0418, + "step": 49260 + }, + { + "epoch": 3.22806787656424, + "grad_norm": 0.25230056047439575, + "learning_rate": 5.5332201757194615e-05, + "loss": 0.046, + "step": 49270 + }, + { + "epoch": 3.2287230557557494, + "grad_norm": 0.5272762179374695, + "learning_rate": 5.531576106077049e-05, + "loss": 0.0516, + "step": 49280 + }, + { + "epoch": 3.229378234947258, + "grad_norm": 0.11693577468395233, + "learning_rate": 5.529931978302272e-05, + "loss": 0.0225, + "step": 49290 + }, + { + "epoch": 3.230033414138767, + "grad_norm": 0.9951595664024353, + "learning_rate": 5.528287792574932e-05, + "loss": 0.0326, + "step": 49300 + }, + { + "epoch": 3.230688593330276, + "grad_norm": 0.212760791182518, + "learning_rate": 5.5266435490748324e-05, + "loss": 0.0407, + "step": 49310 + }, + { + "epoch": 3.2313437725217846, + "grad_norm": 0.3329283595085144, + "learning_rate": 5.524999247981787e-05, + "loss": 0.0323, + "step": 49320 + }, + { + "epoch": 3.2319989517132934, + "grad_norm": 1.7229777574539185, + "learning_rate": 5.523354889475613e-05, + "loss": 0.0473, + "step": 49330 + }, + { + "epoch": 3.2326541309048027, + "grad_norm": 0.512793779373169, + "learning_rate": 5.521710473736134e-05, + "loss": 0.0544, + "step": 49340 + }, + { + "epoch": 3.2333093100963115, + "grad_norm": 0.4733777940273285, + "learning_rate": 5.5200660009431826e-05, + "loss": 0.0357, + "step": 49350 + }, + { + "epoch": 3.2339644892878203, + "grad_norm": 0.16993634402751923, + "learning_rate": 5.518421471276596e-05, + "loss": 0.0741, + "step": 49360 + }, + { + "epoch": 3.234619668479329, + "grad_norm": 0.8166833519935608, + "learning_rate": 5.5167768849162174e-05, + "loss": 0.0272, + "step": 49370 + }, + { + "epoch": 3.235274847670838, + "grad_norm": 0.6986278891563416, + "learning_rate": 5.515132242041893e-05, + "loss": 0.0429, + "step": 49380 + }, + { + "epoch": 3.2359300268623468, + "grad_norm": 0.6856933832168579, + "learning_rate": 5.513487542833483e-05, + "loss": 0.0551, + "step": 49390 + }, + { + "epoch": 3.2365852060538556, + "grad_norm": 0.1513512283563614, + "learning_rate": 5.5118427874708466e-05, + "loss": 0.0285, + "step": 49400 + }, + { + "epoch": 3.2372403852453644, + "grad_norm": 0.5253618955612183, + "learning_rate": 5.510197976133853e-05, + "loss": 0.0482, + "step": 49410 + }, + { + "epoch": 3.2378955644368737, + "grad_norm": 0.7392470240592957, + "learning_rate": 5.508553109002376e-05, + "loss": 0.0608, + "step": 49420 + }, + { + "epoch": 3.2385507436283825, + "grad_norm": 0.1667398065328598, + "learning_rate": 5.5069081862562957e-05, + "loss": 0.0241, + "step": 49430 + }, + { + "epoch": 3.2392059228198913, + "grad_norm": 0.3503401279449463, + "learning_rate": 5.5052632080754965e-05, + "loss": 0.0406, + "step": 49440 + }, + { + "epoch": 3.2398611020114, + "grad_norm": 0.14348019659519196, + "learning_rate": 5.503618174639874e-05, + "loss": 0.0402, + "step": 49450 + }, + { + "epoch": 3.240516281202909, + "grad_norm": 0.8173683881759644, + "learning_rate": 5.501973086129325e-05, + "loss": 0.0586, + "step": 49460 + }, + { + "epoch": 3.2411714603944177, + "grad_norm": 0.3738475441932678, + "learning_rate": 5.500327942723753e-05, + "loss": 0.0403, + "step": 49470 + }, + { + "epoch": 3.241826639585927, + "grad_norm": 0.1525246948003769, + "learning_rate": 5.498682744603071e-05, + "loss": 0.0199, + "step": 49480 + }, + { + "epoch": 3.242481818777436, + "grad_norm": 0.31419408321380615, + "learning_rate": 5.497037491947194e-05, + "loss": 0.0372, + "step": 49490 + }, + { + "epoch": 3.2431369979689446, + "grad_norm": 0.10095479339361191, + "learning_rate": 5.4953921849360424e-05, + "loss": 0.0235, + "step": 49500 + }, + { + "epoch": 3.2437921771604534, + "grad_norm": 0.6428582668304443, + "learning_rate": 5.493746823749547e-05, + "loss": 0.0376, + "step": 49510 + }, + { + "epoch": 3.2444473563519622, + "grad_norm": 1.1011059284210205, + "learning_rate": 5.49210140856764e-05, + "loss": 0.0354, + "step": 49520 + }, + { + "epoch": 3.245102535543471, + "grad_norm": 0.9384467601776123, + "learning_rate": 5.4904559395702647e-05, + "loss": 0.0571, + "step": 49530 + }, + { + "epoch": 3.24575771473498, + "grad_norm": 0.562870979309082, + "learning_rate": 5.488810416937364e-05, + "loss": 0.0295, + "step": 49540 + }, + { + "epoch": 3.2464128939264887, + "grad_norm": 0.4640067219734192, + "learning_rate": 5.4871648408488926e-05, + "loss": 0.0488, + "step": 49550 + }, + { + "epoch": 3.247068073117998, + "grad_norm": 0.5531608462333679, + "learning_rate": 5.485519211484807e-05, + "loss": 0.034, + "step": 49560 + }, + { + "epoch": 3.2477232523095068, + "grad_norm": 0.5743313431739807, + "learning_rate": 5.4838735290250676e-05, + "loss": 0.037, + "step": 49570 + }, + { + "epoch": 3.2483784315010156, + "grad_norm": 0.6776626110076904, + "learning_rate": 5.482227793649649e-05, + "loss": 0.0415, + "step": 49580 + }, + { + "epoch": 3.2490336106925244, + "grad_norm": 0.3249668478965759, + "learning_rate": 5.480582005538524e-05, + "loss": 0.0451, + "step": 49590 + }, + { + "epoch": 3.249688789884033, + "grad_norm": 0.7671938538551331, + "learning_rate": 5.478936164871671e-05, + "loss": 0.0396, + "step": 49600 + }, + { + "epoch": 3.250343969075542, + "grad_norm": 0.7992501854896545, + "learning_rate": 5.477290271829081e-05, + "loss": 0.0566, + "step": 49610 + }, + { + "epoch": 3.2509991482670513, + "grad_norm": 0.1215534508228302, + "learning_rate": 5.475644326590744e-05, + "loss": 0.0197, + "step": 49620 + }, + { + "epoch": 3.25165432745856, + "grad_norm": 0.47809121012687683, + "learning_rate": 5.473998329336658e-05, + "loss": 0.0295, + "step": 49630 + }, + { + "epoch": 3.252309506650069, + "grad_norm": 0.8555430769920349, + "learning_rate": 5.4723522802468286e-05, + "loss": 0.0655, + "step": 49640 + }, + { + "epoch": 3.2529646858415777, + "grad_norm": 0.6138927936553955, + "learning_rate": 5.470706179501264e-05, + "loss": 0.0312, + "step": 49650 + }, + { + "epoch": 3.2536198650330865, + "grad_norm": 0.1462133824825287, + "learning_rate": 5.4690600272799776e-05, + "loss": 0.0258, + "step": 49660 + }, + { + "epoch": 3.2542750442245953, + "grad_norm": 0.6766622066497803, + "learning_rate": 5.467413823762993e-05, + "loss": 0.0476, + "step": 49670 + }, + { + "epoch": 3.254930223416104, + "grad_norm": 0.3133887052536011, + "learning_rate": 5.465767569130335e-05, + "loss": 0.0469, + "step": 49680 + }, + { + "epoch": 3.255585402607613, + "grad_norm": 1.0222615003585815, + "learning_rate": 5.464121263562036e-05, + "loss": 0.0662, + "step": 49690 + }, + { + "epoch": 3.2562405817991222, + "grad_norm": 0.589434802532196, + "learning_rate": 5.4624749072381343e-05, + "loss": 0.038, + "step": 49700 + }, + { + "epoch": 3.256895760990631, + "grad_norm": 0.21138763427734375, + "learning_rate": 5.460828500338672e-05, + "loss": 0.0486, + "step": 49710 + }, + { + "epoch": 3.25755094018214, + "grad_norm": 1.2663841247558594, + "learning_rate": 5.459182043043698e-05, + "loss": 0.082, + "step": 49720 + }, + { + "epoch": 3.2582061193736487, + "grad_norm": 0.37464672327041626, + "learning_rate": 5.457535535533265e-05, + "loss": 0.0522, + "step": 49730 + }, + { + "epoch": 3.2588612985651575, + "grad_norm": 0.40468719601631165, + "learning_rate": 5.4558889779874334e-05, + "loss": 0.0632, + "step": 49740 + }, + { + "epoch": 3.2595164777566663, + "grad_norm": 0.09609721601009369, + "learning_rate": 5.454242370586269e-05, + "loss": 0.0373, + "step": 49750 + }, + { + "epoch": 3.2601716569481756, + "grad_norm": 0.5256710648536682, + "learning_rate": 5.452595713509843e-05, + "loss": 0.0262, + "step": 49760 + }, + { + "epoch": 3.2608268361396844, + "grad_norm": 1.2705637216567993, + "learning_rate": 5.4509490069382274e-05, + "loss": 0.0414, + "step": 49770 + }, + { + "epoch": 3.261482015331193, + "grad_norm": 0.15106996893882751, + "learning_rate": 5.449302251051509e-05, + "loss": 0.0329, + "step": 49780 + }, + { + "epoch": 3.262137194522702, + "grad_norm": 0.09991487860679626, + "learning_rate": 5.44765544602977e-05, + "loss": 0.0158, + "step": 49790 + }, + { + "epoch": 3.262792373714211, + "grad_norm": 0.40245547890663147, + "learning_rate": 5.446008592053107e-05, + "loss": 0.0483, + "step": 49800 + }, + { + "epoch": 3.2634475529057196, + "grad_norm": 0.41186630725860596, + "learning_rate": 5.4443616893016135e-05, + "loss": 0.0217, + "step": 49810 + }, + { + "epoch": 3.2641027320972285, + "grad_norm": 0.30595773458480835, + "learning_rate": 5.4427147379553934e-05, + "loss": 0.0313, + "step": 49820 + }, + { + "epoch": 3.2647579112887373, + "grad_norm": 0.7248753309249878, + "learning_rate": 5.441067738194556e-05, + "loss": 0.0426, + "step": 49830 + }, + { + "epoch": 3.2654130904802465, + "grad_norm": 0.5497925281524658, + "learning_rate": 5.439420690199214e-05, + "loss": 0.0647, + "step": 49840 + }, + { + "epoch": 3.2660682696717553, + "grad_norm": 0.7822244763374329, + "learning_rate": 5.437773594149484e-05, + "loss": 0.0356, + "step": 49850 + }, + { + "epoch": 3.266723448863264, + "grad_norm": 0.1652825027704239, + "learning_rate": 5.436126450225495e-05, + "loss": 0.0322, + "step": 49860 + }, + { + "epoch": 3.267378628054773, + "grad_norm": 0.17259295284748077, + "learning_rate": 5.434479258607371e-05, + "loss": 0.0346, + "step": 49870 + }, + { + "epoch": 3.268033807246282, + "grad_norm": 0.24813084304332733, + "learning_rate": 5.43283201947525e-05, + "loss": 0.0327, + "step": 49880 + }, + { + "epoch": 3.2686889864377906, + "grad_norm": 1.2294303178787231, + "learning_rate": 5.431184733009268e-05, + "loss": 0.0314, + "step": 49890 + }, + { + "epoch": 3.2693441656293, + "grad_norm": 0.6331045031547546, + "learning_rate": 5.4295373993895736e-05, + "loss": 0.031, + "step": 49900 + }, + { + "epoch": 3.2699993448208087, + "grad_norm": 0.27723103761672974, + "learning_rate": 5.4278900187963157e-05, + "loss": 0.0215, + "step": 49910 + }, + { + "epoch": 3.2706545240123175, + "grad_norm": 0.0960388109087944, + "learning_rate": 5.426242591409646e-05, + "loss": 0.0302, + "step": 49920 + }, + { + "epoch": 3.2713097032038263, + "grad_norm": 0.1289719194173813, + "learning_rate": 5.42459511740973e-05, + "loss": 0.045, + "step": 49930 + }, + { + "epoch": 3.271964882395335, + "grad_norm": 0.8870667219161987, + "learning_rate": 5.422947596976729e-05, + "loss": 0.0453, + "step": 49940 + }, + { + "epoch": 3.272620061586844, + "grad_norm": 1.1092649698257446, + "learning_rate": 5.4213000302908134e-05, + "loss": 0.0339, + "step": 49950 + }, + { + "epoch": 3.2732752407783527, + "grad_norm": 0.1700219362974167, + "learning_rate": 5.419652417532162e-05, + "loss": 0.0251, + "step": 49960 + }, + { + "epoch": 3.2739304199698616, + "grad_norm": 0.2334200143814087, + "learning_rate": 5.4180047588809534e-05, + "loss": 0.0322, + "step": 49970 + }, + { + "epoch": 3.274585599161371, + "grad_norm": 0.13719648122787476, + "learning_rate": 5.4163570545173704e-05, + "loss": 0.0421, + "step": 49980 + }, + { + "epoch": 3.2752407783528796, + "grad_norm": 0.18108634650707245, + "learning_rate": 5.414709304621608e-05, + "loss": 0.0357, + "step": 49990 + }, + { + "epoch": 3.2758959575443884, + "grad_norm": 0.5624659061431885, + "learning_rate": 5.413061509373858e-05, + "loss": 0.0202, + "step": 50000 + }, + { + "epoch": 3.2765511367358973, + "grad_norm": 0.18907424807548523, + "learning_rate": 5.411413668954324e-05, + "loss": 0.0203, + "step": 50010 + }, + { + "epoch": 3.277206315927406, + "grad_norm": 0.5973866581916809, + "learning_rate": 5.4097657835432083e-05, + "loss": 0.0435, + "step": 50020 + }, + { + "epoch": 3.277861495118915, + "grad_norm": 0.2911127805709839, + "learning_rate": 5.408117853320723e-05, + "loss": 0.0422, + "step": 50030 + }, + { + "epoch": 3.278516674310424, + "grad_norm": 0.3854619264602661, + "learning_rate": 5.406469878467084e-05, + "loss": 0.0851, + "step": 50040 + }, + { + "epoch": 3.279171853501933, + "grad_norm": 0.3685075640678406, + "learning_rate": 5.404821859162509e-05, + "loss": 0.0255, + "step": 50050 + }, + { + "epoch": 3.279827032693442, + "grad_norm": 0.11154581606388092, + "learning_rate": 5.403173795587225e-05, + "loss": 0.0421, + "step": 50060 + }, + { + "epoch": 3.2804822118849506, + "grad_norm": 0.37495097517967224, + "learning_rate": 5.4015256879214606e-05, + "loss": 0.0389, + "step": 50070 + }, + { + "epoch": 3.2811373910764594, + "grad_norm": 0.33153238892555237, + "learning_rate": 5.39987753634545e-05, + "loss": 0.0536, + "step": 50080 + }, + { + "epoch": 3.2817925702679682, + "grad_norm": 0.6714552640914917, + "learning_rate": 5.3982293410394335e-05, + "loss": 0.0389, + "step": 50090 + }, + { + "epoch": 3.282447749459477, + "grad_norm": 1.010091781616211, + "learning_rate": 5.396581102183655e-05, + "loss": 0.0355, + "step": 50100 + }, + { + "epoch": 3.283102928650986, + "grad_norm": 0.21405284106731415, + "learning_rate": 5.3949328199583634e-05, + "loss": 0.0223, + "step": 50110 + }, + { + "epoch": 3.283758107842495, + "grad_norm": 0.17148853838443756, + "learning_rate": 5.393284494543813e-05, + "loss": 0.0253, + "step": 50120 + }, + { + "epoch": 3.284413287034004, + "grad_norm": 0.6432908773422241, + "learning_rate": 5.391636126120262e-05, + "loss": 0.0354, + "step": 50130 + }, + { + "epoch": 3.2850684662255127, + "grad_norm": 0.3010774254798889, + "learning_rate": 5.389987714867971e-05, + "loss": 0.0228, + "step": 50140 + }, + { + "epoch": 3.2857236454170216, + "grad_norm": 0.14839890599250793, + "learning_rate": 5.3883392609672114e-05, + "loss": 0.0561, + "step": 50150 + }, + { + "epoch": 3.2863788246085304, + "grad_norm": 0.5521979928016663, + "learning_rate": 5.3866907645982545e-05, + "loss": 0.0362, + "step": 50160 + }, + { + "epoch": 3.287034003800039, + "grad_norm": 1.0003958940505981, + "learning_rate": 5.385042225941375e-05, + "loss": 0.0381, + "step": 50170 + }, + { + "epoch": 3.287689182991548, + "grad_norm": 0.7658777832984924, + "learning_rate": 5.3833936451768585e-05, + "loss": 0.0679, + "step": 50180 + }, + { + "epoch": 3.2883443621830573, + "grad_norm": 0.11985767632722855, + "learning_rate": 5.3817450224849884e-05, + "loss": 0.0323, + "step": 50190 + }, + { + "epoch": 3.288999541374566, + "grad_norm": 0.8329668045043945, + "learning_rate": 5.380096358046056e-05, + "loss": 0.0586, + "step": 50200 + }, + { + "epoch": 3.289654720566075, + "grad_norm": 0.18101812899112701, + "learning_rate": 5.378447652040359e-05, + "loss": 0.0283, + "step": 50210 + }, + { + "epoch": 3.2903098997575837, + "grad_norm": 0.5182900428771973, + "learning_rate": 5.376798904648195e-05, + "loss": 0.0374, + "step": 50220 + }, + { + "epoch": 3.2909650789490925, + "grad_norm": 0.4705137312412262, + "learning_rate": 5.375150116049869e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 3.2916202581406013, + "grad_norm": 0.6406623125076294, + "learning_rate": 5.373501286425691e-05, + "loss": 0.0402, + "step": 50240 + }, + { + "epoch": 3.29227543733211, + "grad_norm": 0.2413666844367981, + "learning_rate": 5.3718524159559726e-05, + "loss": 0.036, + "step": 50250 + }, + { + "epoch": 3.2929306165236194, + "grad_norm": 0.7513158321380615, + "learning_rate": 5.370203504821034e-05, + "loss": 0.0502, + "step": 50260 + }, + { + "epoch": 3.293585795715128, + "grad_norm": 0.1250956952571869, + "learning_rate": 5.368554553201196e-05, + "loss": 0.0201, + "step": 50270 + }, + { + "epoch": 3.294240974906637, + "grad_norm": 0.7338235378265381, + "learning_rate": 5.3669055612767874e-05, + "loss": 0.0527, + "step": 50280 + }, + { + "epoch": 3.294896154098146, + "grad_norm": 0.14889603853225708, + "learning_rate": 5.3652565292281376e-05, + "loss": 0.0356, + "step": 50290 + }, + { + "epoch": 3.2955513332896547, + "grad_norm": 0.44916877150535583, + "learning_rate": 5.3636074572355824e-05, + "loss": 0.0473, + "step": 50300 + }, + { + "epoch": 3.2962065124811635, + "grad_norm": 0.10829369723796844, + "learning_rate": 5.361958345479463e-05, + "loss": 0.0315, + "step": 50310 + }, + { + "epoch": 3.2968616916726723, + "grad_norm": 0.37395331263542175, + "learning_rate": 5.360309194140123e-05, + "loss": 0.0351, + "step": 50320 + }, + { + "epoch": 3.2975168708641815, + "grad_norm": 0.7544816136360168, + "learning_rate": 5.3586600033979106e-05, + "loss": 0.0525, + "step": 50330 + }, + { + "epoch": 3.2981720500556904, + "grad_norm": 0.3024815618991852, + "learning_rate": 5.357010773433181e-05, + "loss": 0.0595, + "step": 50340 + }, + { + "epoch": 3.298827229247199, + "grad_norm": 0.31206679344177246, + "learning_rate": 5.3553615044262895e-05, + "loss": 0.0247, + "step": 50350 + }, + { + "epoch": 3.299482408438708, + "grad_norm": 0.28602200746536255, + "learning_rate": 5.3537121965575964e-05, + "loss": 0.0259, + "step": 50360 + }, + { + "epoch": 3.300137587630217, + "grad_norm": 0.19463157653808594, + "learning_rate": 5.352062850007471e-05, + "loss": 0.0159, + "step": 50370 + }, + { + "epoch": 3.3007927668217256, + "grad_norm": 0.5497636795043945, + "learning_rate": 5.350413464956282e-05, + "loss": 0.0404, + "step": 50380 + }, + { + "epoch": 3.3014479460132344, + "grad_norm": 0.5170273184776306, + "learning_rate": 5.348764041584403e-05, + "loss": 0.0361, + "step": 50390 + }, + { + "epoch": 3.3021031252047433, + "grad_norm": 0.2894536256790161, + "learning_rate": 5.347114580072212e-05, + "loss": 0.0314, + "step": 50400 + }, + { + "epoch": 3.3027583043962525, + "grad_norm": 0.26027166843414307, + "learning_rate": 5.345465080600093e-05, + "loss": 0.0364, + "step": 50410 + }, + { + "epoch": 3.3034134835877613, + "grad_norm": 0.7386171817779541, + "learning_rate": 5.3438155433484326e-05, + "loss": 0.0332, + "step": 50420 + }, + { + "epoch": 3.30406866277927, + "grad_norm": 0.5612487196922302, + "learning_rate": 5.3421659684976197e-05, + "loss": 0.0586, + "step": 50430 + }, + { + "epoch": 3.304723841970779, + "grad_norm": 0.30591100454330444, + "learning_rate": 5.340516356228052e-05, + "loss": 0.0233, + "step": 50440 + }, + { + "epoch": 3.3053790211622878, + "grad_norm": 0.40567758679389954, + "learning_rate": 5.338866706720128e-05, + "loss": 0.0336, + "step": 50450 + }, + { + "epoch": 3.3060342003537966, + "grad_norm": 0.3806193470954895, + "learning_rate": 5.337217020154249e-05, + "loss": 0.0282, + "step": 50460 + }, + { + "epoch": 3.306689379545306, + "grad_norm": 0.5260007381439209, + "learning_rate": 5.335567296710825e-05, + "loss": 0.0317, + "step": 50470 + }, + { + "epoch": 3.3073445587368147, + "grad_norm": 0.9848410487174988, + "learning_rate": 5.333917536570265e-05, + "loss": 0.0466, + "step": 50480 + }, + { + "epoch": 3.3079997379283235, + "grad_norm": 0.07908214628696442, + "learning_rate": 5.332267739912986e-05, + "loss": 0.0325, + "step": 50490 + }, + { + "epoch": 3.3086549171198323, + "grad_norm": 0.4197009801864624, + "learning_rate": 5.330617906919405e-05, + "loss": 0.0154, + "step": 50500 + }, + { + "epoch": 3.309310096311341, + "grad_norm": 0.30505725741386414, + "learning_rate": 5.328968037769949e-05, + "loss": 0.0346, + "step": 50510 + }, + { + "epoch": 3.30996527550285, + "grad_norm": 1.813659906387329, + "learning_rate": 5.327318132645042e-05, + "loss": 0.0298, + "step": 50520 + }, + { + "epoch": 3.3106204546943587, + "grad_norm": 0.28483423590660095, + "learning_rate": 5.325668191725116e-05, + "loss": 0.024, + "step": 50530 + }, + { + "epoch": 3.3112756338858675, + "grad_norm": 0.48349976539611816, + "learning_rate": 5.324018215190606e-05, + "loss": 0.0353, + "step": 50540 + }, + { + "epoch": 3.311930813077377, + "grad_norm": 0.44559988379478455, + "learning_rate": 5.3223682032219515e-05, + "loss": 0.0487, + "step": 50550 + }, + { + "epoch": 3.3125859922688856, + "grad_norm": 1.4354023933410645, + "learning_rate": 5.320718155999595e-05, + "loss": 0.0473, + "step": 50560 + }, + { + "epoch": 3.3132411714603944, + "grad_norm": 0.14887042343616486, + "learning_rate": 5.3190680737039835e-05, + "loss": 0.0314, + "step": 50570 + }, + { + "epoch": 3.3138963506519032, + "grad_norm": 0.08655637502670288, + "learning_rate": 5.317417956515567e-05, + "loss": 0.0366, + "step": 50580 + }, + { + "epoch": 3.314551529843412, + "grad_norm": 0.5218527317047119, + "learning_rate": 5.315767804614798e-05, + "loss": 0.052, + "step": 50590 + }, + { + "epoch": 3.315206709034921, + "grad_norm": 0.8797337412834167, + "learning_rate": 5.3141176181821395e-05, + "loss": 0.0383, + "step": 50600 + }, + { + "epoch": 3.31586188822643, + "grad_norm": 0.17942439019680023, + "learning_rate": 5.31246739739805e-05, + "loss": 0.0264, + "step": 50610 + }, + { + "epoch": 3.316517067417939, + "grad_norm": 0.33727753162384033, + "learning_rate": 5.310817142442994e-05, + "loss": 0.0316, + "step": 50620 + }, + { + "epoch": 3.3171722466094478, + "grad_norm": 0.32063227891921997, + "learning_rate": 5.309166853497445e-05, + "loss": 0.0308, + "step": 50630 + }, + { + "epoch": 3.3178274258009566, + "grad_norm": 0.8089986443519592, + "learning_rate": 5.307516530741873e-05, + "loss": 0.0614, + "step": 50640 + }, + { + "epoch": 3.3184826049924654, + "grad_norm": 0.1478569358587265, + "learning_rate": 5.305866174356754e-05, + "loss": 0.0384, + "step": 50650 + }, + { + "epoch": 3.319137784183974, + "grad_norm": 0.4084159731864929, + "learning_rate": 5.304215784522571e-05, + "loss": 0.0217, + "step": 50660 + }, + { + "epoch": 3.319792963375483, + "grad_norm": 0.18430590629577637, + "learning_rate": 5.302565361419808e-05, + "loss": 0.0254, + "step": 50670 + }, + { + "epoch": 3.320448142566992, + "grad_norm": 0.22131547331809998, + "learning_rate": 5.3009149052289507e-05, + "loss": 0.0371, + "step": 50680 + }, + { + "epoch": 3.321103321758501, + "grad_norm": 0.22703351080417633, + "learning_rate": 5.299264416130493e-05, + "loss": 0.0352, + "step": 50690 + }, + { + "epoch": 3.32175850095001, + "grad_norm": 0.4801960587501526, + "learning_rate": 5.297613894304928e-05, + "loss": 0.04, + "step": 50700 + }, + { + "epoch": 3.3224136801415187, + "grad_norm": 0.3482247292995453, + "learning_rate": 5.2959633399327534e-05, + "loss": 0.0541, + "step": 50710 + }, + { + "epoch": 3.3230688593330275, + "grad_norm": 0.10933016240596771, + "learning_rate": 5.294312753194476e-05, + "loss": 0.0161, + "step": 50720 + }, + { + "epoch": 3.3237240385245364, + "grad_norm": 0.4993155896663666, + "learning_rate": 5.292662134270596e-05, + "loss": 0.0362, + "step": 50730 + }, + { + "epoch": 3.324379217716045, + "grad_norm": 0.29139700531959534, + "learning_rate": 5.291011483341626e-05, + "loss": 0.0304, + "step": 50740 + }, + { + "epoch": 3.3250343969075544, + "grad_norm": 0.6447833180427551, + "learning_rate": 5.2893608005880767e-05, + "loss": 0.0632, + "step": 50750 + }, + { + "epoch": 3.3256895760990632, + "grad_norm": 0.21433931589126587, + "learning_rate": 5.287710086190467e-05, + "loss": 0.02, + "step": 50760 + }, + { + "epoch": 3.326344755290572, + "grad_norm": 2.289822578430176, + "learning_rate": 5.286059340329314e-05, + "loss": 0.0318, + "step": 50770 + }, + { + "epoch": 3.326999934482081, + "grad_norm": 0.17651140689849854, + "learning_rate": 5.284408563185141e-05, + "loss": 0.0286, + "step": 50780 + }, + { + "epoch": 3.3276551136735897, + "grad_norm": 0.24352990090847015, + "learning_rate": 5.282757754938476e-05, + "loss": 0.0353, + "step": 50790 + }, + { + "epoch": 3.3283102928650985, + "grad_norm": 0.05741025507450104, + "learning_rate": 5.281106915769849e-05, + "loss": 0.0288, + "step": 50800 + }, + { + "epoch": 3.3289654720566073, + "grad_norm": 0.4081929326057434, + "learning_rate": 5.2794560458597897e-05, + "loss": 0.0204, + "step": 50810 + }, + { + "epoch": 3.329620651248116, + "grad_norm": 0.5280476808547974, + "learning_rate": 5.27780514538884e-05, + "loss": 0.0254, + "step": 50820 + }, + { + "epoch": 3.3302758304396254, + "grad_norm": 0.46944159269332886, + "learning_rate": 5.2761542145375365e-05, + "loss": 0.052, + "step": 50830 + }, + { + "epoch": 3.330931009631134, + "grad_norm": 0.5352941751480103, + "learning_rate": 5.274503253486421e-05, + "loss": 0.0257, + "step": 50840 + }, + { + "epoch": 3.331586188822643, + "grad_norm": 0.36506959795951843, + "learning_rate": 5.272852262416046e-05, + "loss": 0.0554, + "step": 50850 + }, + { + "epoch": 3.332241368014152, + "grad_norm": 0.2508940100669861, + "learning_rate": 5.2712012415069555e-05, + "loss": 0.0216, + "step": 50860 + }, + { + "epoch": 3.3328965472056606, + "grad_norm": 0.27231407165527344, + "learning_rate": 5.269550190939705e-05, + "loss": 0.0217, + "step": 50870 + }, + { + "epoch": 3.3335517263971695, + "grad_norm": 0.7503190040588379, + "learning_rate": 5.267899110894852e-05, + "loss": 0.0279, + "step": 50880 + }, + { + "epoch": 3.3342069055886787, + "grad_norm": 0.2759523391723633, + "learning_rate": 5.266248001552955e-05, + "loss": 0.0315, + "step": 50890 + }, + { + "epoch": 3.3348620847801875, + "grad_norm": 0.5927759408950806, + "learning_rate": 5.264596863094575e-05, + "loss": 0.0209, + "step": 50900 + }, + { + "epoch": 3.3355172639716963, + "grad_norm": 0.3263818323612213, + "learning_rate": 5.262945695700282e-05, + "loss": 0.0206, + "step": 50910 + }, + { + "epoch": 3.336172443163205, + "grad_norm": 0.4822939336299896, + "learning_rate": 5.261294499550643e-05, + "loss": 0.0439, + "step": 50920 + }, + { + "epoch": 3.336827622354714, + "grad_norm": 0.28172567486763, + "learning_rate": 5.25964327482623e-05, + "loss": 0.0402, + "step": 50930 + }, + { + "epoch": 3.337482801546223, + "grad_norm": 0.6730916500091553, + "learning_rate": 5.257992021707617e-05, + "loss": 0.0512, + "step": 50940 + }, + { + "epoch": 3.3381379807377316, + "grad_norm": 0.8108209371566772, + "learning_rate": 5.256340740375387e-05, + "loss": 0.0482, + "step": 50950 + }, + { + "epoch": 3.3387931599292404, + "grad_norm": 0.2094172090291977, + "learning_rate": 5.254689431010117e-05, + "loss": 0.0258, + "step": 50960 + }, + { + "epoch": 3.3394483391207497, + "grad_norm": 0.4535210430622101, + "learning_rate": 5.253038093792395e-05, + "loss": 0.0184, + "step": 50970 + }, + { + "epoch": 3.3401035183122585, + "grad_norm": 0.5396775007247925, + "learning_rate": 5.251386728902806e-05, + "loss": 0.0273, + "step": 50980 + }, + { + "epoch": 3.3407586975037673, + "grad_norm": 0.7790852189064026, + "learning_rate": 5.2497353365219446e-05, + "loss": 0.0328, + "step": 50990 + }, + { + "epoch": 3.341413876695276, + "grad_norm": 0.13284142315387726, + "learning_rate": 5.2480839168304e-05, + "loss": 0.0241, + "step": 51000 + }, + { + "epoch": 3.342069055886785, + "grad_norm": 0.6387774348258972, + "learning_rate": 5.246432470008773e-05, + "loss": 0.0426, + "step": 51010 + }, + { + "epoch": 3.3427242350782937, + "grad_norm": 0.1880325824022293, + "learning_rate": 5.2447809962376616e-05, + "loss": 0.0363, + "step": 51020 + }, + { + "epoch": 3.343379414269803, + "grad_norm": 0.08760130405426025, + "learning_rate": 5.243129495697666e-05, + "loss": 0.0429, + "step": 51030 + }, + { + "epoch": 3.344034593461312, + "grad_norm": 0.4119509756565094, + "learning_rate": 5.241477968569397e-05, + "loss": 0.0363, + "step": 51040 + }, + { + "epoch": 3.3446897726528206, + "grad_norm": 0.17496179044246674, + "learning_rate": 5.23982641503346e-05, + "loss": 0.0589, + "step": 51050 + }, + { + "epoch": 3.3453449518443295, + "grad_norm": 2.2755637168884277, + "learning_rate": 5.238174835270465e-05, + "loss": 0.0626, + "step": 51060 + }, + { + "epoch": 3.3460001310358383, + "grad_norm": 0.3912683129310608, + "learning_rate": 5.23652322946103e-05, + "loss": 0.0364, + "step": 51070 + }, + { + "epoch": 3.346655310227347, + "grad_norm": 0.6378939747810364, + "learning_rate": 5.234871597785771e-05, + "loss": 0.0475, + "step": 51080 + }, + { + "epoch": 3.347310489418856, + "grad_norm": 0.7041005492210388, + "learning_rate": 5.233219940425306e-05, + "loss": 0.053, + "step": 51090 + }, + { + "epoch": 3.3479656686103647, + "grad_norm": 0.14118912816047668, + "learning_rate": 5.2315682575602576e-05, + "loss": 0.0386, + "step": 51100 + }, + { + "epoch": 3.348620847801874, + "grad_norm": 1.0522663593292236, + "learning_rate": 5.229916549371254e-05, + "loss": 0.0415, + "step": 51110 + }, + { + "epoch": 3.349276026993383, + "grad_norm": 0.46872907876968384, + "learning_rate": 5.228264816038924e-05, + "loss": 0.0231, + "step": 51120 + }, + { + "epoch": 3.3499312061848916, + "grad_norm": 0.5581085681915283, + "learning_rate": 5.2266130577438935e-05, + "loss": 0.0505, + "step": 51130 + }, + { + "epoch": 3.3505863853764004, + "grad_norm": 0.754525899887085, + "learning_rate": 5.224961274666802e-05, + "loss": 0.0509, + "step": 51140 + }, + { + "epoch": 3.3512415645679092, + "grad_norm": 0.5580785274505615, + "learning_rate": 5.223309466988283e-05, + "loss": 0.0141, + "step": 51150 + }, + { + "epoch": 3.351896743759418, + "grad_norm": 0.24170467257499695, + "learning_rate": 5.221657634888976e-05, + "loss": 0.0433, + "step": 51160 + }, + { + "epoch": 3.3525519229509273, + "grad_norm": 0.29448646306991577, + "learning_rate": 5.2200057785495235e-05, + "loss": 0.0287, + "step": 51170 + }, + { + "epoch": 3.353207102142436, + "grad_norm": 0.3654685616493225, + "learning_rate": 5.2183538981505706e-05, + "loss": 0.0363, + "step": 51180 + }, + { + "epoch": 3.353862281333945, + "grad_norm": 0.3330886662006378, + "learning_rate": 5.216701993872762e-05, + "loss": 0.0462, + "step": 51190 + }, + { + "epoch": 3.3545174605254537, + "grad_norm": 0.696864128112793, + "learning_rate": 5.215050065896752e-05, + "loss": 0.0428, + "step": 51200 + }, + { + "epoch": 3.3551726397169626, + "grad_norm": 0.5647945404052734, + "learning_rate": 5.2133981144031875e-05, + "loss": 0.035, + "step": 51210 + }, + { + "epoch": 3.3558278189084714, + "grad_norm": 0.6677957773208618, + "learning_rate": 5.211746139572727e-05, + "loss": 0.0467, + "step": 51220 + }, + { + "epoch": 3.35648299809998, + "grad_norm": 0.8318150639533997, + "learning_rate": 5.210094141586026e-05, + "loss": 0.0333, + "step": 51230 + }, + { + "epoch": 3.357138177291489, + "grad_norm": 1.0494247674942017, + "learning_rate": 5.2084421206237476e-05, + "loss": 0.0338, + "step": 51240 + }, + { + "epoch": 3.3577933564829983, + "grad_norm": 0.38579538464546204, + "learning_rate": 5.20679007686655e-05, + "loss": 0.0319, + "step": 51250 + }, + { + "epoch": 3.358448535674507, + "grad_norm": 0.6592560410499573, + "learning_rate": 5.205138010495102e-05, + "loss": 0.042, + "step": 51260 + }, + { + "epoch": 3.359103714866016, + "grad_norm": 0.5121399164199829, + "learning_rate": 5.20348592169007e-05, + "loss": 0.0369, + "step": 51270 + }, + { + "epoch": 3.3597588940575247, + "grad_norm": 0.8222784399986267, + "learning_rate": 5.201833810632123e-05, + "loss": 0.0372, + "step": 51280 + }, + { + "epoch": 3.3604140732490335, + "grad_norm": 1.0143718719482422, + "learning_rate": 5.200181677501933e-05, + "loss": 0.0452, + "step": 51290 + }, + { + "epoch": 3.3610692524405423, + "grad_norm": 0.7603490948677063, + "learning_rate": 5.1985295224801777e-05, + "loss": 0.0411, + "step": 51300 + }, + { + "epoch": 3.3617244316320516, + "grad_norm": 0.2547018826007843, + "learning_rate": 5.196877345747531e-05, + "loss": 0.0209, + "step": 51310 + }, + { + "epoch": 3.3623796108235604, + "grad_norm": 0.4826333522796631, + "learning_rate": 5.195225147484674e-05, + "loss": 0.0528, + "step": 51320 + }, + { + "epoch": 3.363034790015069, + "grad_norm": 0.9372945427894592, + "learning_rate": 5.193572927872288e-05, + "loss": 0.0396, + "step": 51330 + }, + { + "epoch": 3.363689969206578, + "grad_norm": 0.19875206053256989, + "learning_rate": 5.191920687091059e-05, + "loss": 0.0322, + "step": 51340 + }, + { + "epoch": 3.364345148398087, + "grad_norm": 0.8403700590133667, + "learning_rate": 5.1902684253216695e-05, + "loss": 0.0432, + "step": 51350 + }, + { + "epoch": 3.3650003275895957, + "grad_norm": 0.739410936832428, + "learning_rate": 5.1886161427448134e-05, + "loss": 0.0162, + "step": 51360 + }, + { + "epoch": 3.3656555067811045, + "grad_norm": 0.4665006995201111, + "learning_rate": 5.186963839541179e-05, + "loss": 0.024, + "step": 51370 + }, + { + "epoch": 3.3663106859726133, + "grad_norm": 0.5626803040504456, + "learning_rate": 5.185311515891459e-05, + "loss": 0.0362, + "step": 51380 + }, + { + "epoch": 3.3669658651641226, + "grad_norm": 0.16391557455062866, + "learning_rate": 5.1836591719763516e-05, + "loss": 0.0261, + "step": 51390 + }, + { + "epoch": 3.3676210443556314, + "grad_norm": 0.152605801820755, + "learning_rate": 5.182006807976554e-05, + "loss": 0.0313, + "step": 51400 + }, + { + "epoch": 3.36827622354714, + "grad_norm": 0.8589211106300354, + "learning_rate": 5.180354424072763e-05, + "loss": 0.0549, + "step": 51410 + }, + { + "epoch": 3.368931402738649, + "grad_norm": 1.1207472085952759, + "learning_rate": 5.178702020445686e-05, + "loss": 0.0707, + "step": 51420 + }, + { + "epoch": 3.369586581930158, + "grad_norm": 0.7925529479980469, + "learning_rate": 5.177049597276024e-05, + "loss": 0.0441, + "step": 51430 + }, + { + "epoch": 3.3702417611216666, + "grad_norm": 0.1362379938364029, + "learning_rate": 5.175397154744482e-05, + "loss": 0.0238, + "step": 51440 + }, + { + "epoch": 3.370896940313176, + "grad_norm": 0.5651464462280273, + "learning_rate": 5.1737446930317736e-05, + "loss": 0.0337, + "step": 51450 + }, + { + "epoch": 3.3715521195046847, + "grad_norm": 0.2668374478816986, + "learning_rate": 5.172092212318604e-05, + "loss": 0.0259, + "step": 51460 + }, + { + "epoch": 3.3722072986961935, + "grad_norm": 0.19857609272003174, + "learning_rate": 5.170439712785692e-05, + "loss": 0.0409, + "step": 51470 + }, + { + "epoch": 3.3728624778877023, + "grad_norm": 0.9397009611129761, + "learning_rate": 5.1687871946137467e-05, + "loss": 0.0588, + "step": 51480 + }, + { + "epoch": 3.373517657079211, + "grad_norm": 0.29171591997146606, + "learning_rate": 5.16713465798349e-05, + "loss": 0.0298, + "step": 51490 + }, + { + "epoch": 3.37417283627072, + "grad_norm": 0.36214640736579895, + "learning_rate": 5.165482103075637e-05, + "loss": 0.0287, + "step": 51500 + }, + { + "epoch": 3.3748280154622288, + "grad_norm": 1.0015029907226562, + "learning_rate": 5.163829530070909e-05, + "loss": 0.0356, + "step": 51510 + }, + { + "epoch": 3.3754831946537376, + "grad_norm": 0.550583004951477, + "learning_rate": 5.162176939150032e-05, + "loss": 0.0418, + "step": 51520 + }, + { + "epoch": 3.376138373845247, + "grad_norm": 0.13419587910175323, + "learning_rate": 5.160524330493728e-05, + "loss": 0.0164, + "step": 51530 + }, + { + "epoch": 3.3767935530367557, + "grad_norm": 0.2884218096733093, + "learning_rate": 5.158871704282724e-05, + "loss": 0.0455, + "step": 51540 + }, + { + "epoch": 3.3774487322282645, + "grad_norm": 0.1688336730003357, + "learning_rate": 5.15721906069775e-05, + "loss": 0.0391, + "step": 51550 + }, + { + "epoch": 3.3781039114197733, + "grad_norm": 0.6498290300369263, + "learning_rate": 5.155566399919536e-05, + "loss": 0.0462, + "step": 51560 + }, + { + "epoch": 3.378759090611282, + "grad_norm": 0.6682944297790527, + "learning_rate": 5.153913722128813e-05, + "loss": 0.0629, + "step": 51570 + }, + { + "epoch": 3.379414269802791, + "grad_norm": 0.5325594544410706, + "learning_rate": 5.1522610275063176e-05, + "loss": 0.0377, + "step": 51580 + }, + { + "epoch": 3.3800694489943, + "grad_norm": 0.08978516608476639, + "learning_rate": 5.150608316232786e-05, + "loss": 0.0516, + "step": 51590 + }, + { + "epoch": 3.380724628185809, + "grad_norm": 0.46525663137435913, + "learning_rate": 5.148955588488955e-05, + "loss": 0.0307, + "step": 51600 + }, + { + "epoch": 3.381379807377318, + "grad_norm": 0.6454899311065674, + "learning_rate": 5.147302844455564e-05, + "loss": 0.0358, + "step": 51610 + }, + { + "epoch": 3.3820349865688266, + "grad_norm": 0.21770501136779785, + "learning_rate": 5.145650084313356e-05, + "loss": 0.0422, + "step": 51620 + }, + { + "epoch": 3.3826901657603354, + "grad_norm": 0.3952224552631378, + "learning_rate": 5.1439973082430734e-05, + "loss": 0.0258, + "step": 51630 + }, + { + "epoch": 3.3833453449518442, + "grad_norm": 0.36605072021484375, + "learning_rate": 5.142344516425461e-05, + "loss": 0.0223, + "step": 51640 + }, + { + "epoch": 3.384000524143353, + "grad_norm": 0.17887228727340698, + "learning_rate": 5.1406917090412673e-05, + "loss": 0.0262, + "step": 51650 + }, + { + "epoch": 3.384655703334862, + "grad_norm": 0.44787734746932983, + "learning_rate": 5.1390388862712404e-05, + "loss": 0.0395, + "step": 51660 + }, + { + "epoch": 3.385310882526371, + "grad_norm": 0.24209217727184296, + "learning_rate": 5.137386048296129e-05, + "loss": 0.0451, + "step": 51670 + }, + { + "epoch": 3.38596606171788, + "grad_norm": 0.1444845050573349, + "learning_rate": 5.135733195296686e-05, + "loss": 0.0281, + "step": 51680 + }, + { + "epoch": 3.3866212409093888, + "grad_norm": 0.06538397073745728, + "learning_rate": 5.134080327453664e-05, + "loss": 0.036, + "step": 51690 + }, + { + "epoch": 3.3872764201008976, + "grad_norm": 0.852232813835144, + "learning_rate": 5.1324274449478204e-05, + "loss": 0.0301, + "step": 51700 + }, + { + "epoch": 3.3879315992924064, + "grad_norm": 1.8396821022033691, + "learning_rate": 5.130774547959909e-05, + "loss": 0.0582, + "step": 51710 + }, + { + "epoch": 3.388586778483915, + "grad_norm": 0.2023461014032364, + "learning_rate": 5.129121636670692e-05, + "loss": 0.023, + "step": 51720 + }, + { + "epoch": 3.3892419576754245, + "grad_norm": 0.34013259410858154, + "learning_rate": 5.1274687112609246e-05, + "loss": 0.0449, + "step": 51730 + }, + { + "epoch": 3.3898971368669333, + "grad_norm": 0.400656521320343, + "learning_rate": 5.125815771911372e-05, + "loss": 0.0368, + "step": 51740 + }, + { + "epoch": 3.390552316058442, + "grad_norm": 0.4068857729434967, + "learning_rate": 5.1241628188027956e-05, + "loss": 0.027, + "step": 51750 + }, + { + "epoch": 3.391207495249951, + "grad_norm": 0.5897894501686096, + "learning_rate": 5.1225098521159585e-05, + "loss": 0.0325, + "step": 51760 + }, + { + "epoch": 3.3918626744414597, + "grad_norm": 0.5295219421386719, + "learning_rate": 5.12085687203163e-05, + "loss": 0.026, + "step": 51770 + }, + { + "epoch": 3.3925178536329685, + "grad_norm": 0.30276092886924744, + "learning_rate": 5.119203878730575e-05, + "loss": 0.0311, + "step": 51780 + }, + { + "epoch": 3.3931730328244774, + "grad_norm": 0.42032283544540405, + "learning_rate": 5.117550872393564e-05, + "loss": 0.02, + "step": 51790 + }, + { + "epoch": 3.393828212015986, + "grad_norm": 0.17042423784732819, + "learning_rate": 5.1158978532013645e-05, + "loss": 0.0262, + "step": 51800 + }, + { + "epoch": 3.3944833912074954, + "grad_norm": 1.116795539855957, + "learning_rate": 5.114244821334751e-05, + "loss": 0.0362, + "step": 51810 + }, + { + "epoch": 3.3951385703990042, + "grad_norm": 0.2879104018211365, + "learning_rate": 5.112591776974497e-05, + "loss": 0.0397, + "step": 51820 + }, + { + "epoch": 3.395793749590513, + "grad_norm": 0.15897822380065918, + "learning_rate": 5.110938720301374e-05, + "loss": 0.0427, + "step": 51830 + }, + { + "epoch": 3.396448928782022, + "grad_norm": 0.6270864009857178, + "learning_rate": 5.1092856514961595e-05, + "loss": 0.0355, + "step": 51840 + }, + { + "epoch": 3.3971041079735307, + "grad_norm": 0.4627437889575958, + "learning_rate": 5.107632570739631e-05, + "loss": 0.0432, + "step": 51850 + }, + { + "epoch": 3.3977592871650395, + "grad_norm": 2.4596405029296875, + "learning_rate": 5.1059794782125645e-05, + "loss": 0.0672, + "step": 51860 + }, + { + "epoch": 3.3984144663565488, + "grad_norm": 0.38938167691230774, + "learning_rate": 5.104326374095744e-05, + "loss": 0.0431, + "step": 51870 + }, + { + "epoch": 3.3990696455480576, + "grad_norm": 0.2285594344139099, + "learning_rate": 5.102673258569947e-05, + "loss": 0.0335, + "step": 51880 + }, + { + "epoch": 3.3997248247395664, + "grad_norm": 0.4048537313938141, + "learning_rate": 5.101020131815956e-05, + "loss": 0.0205, + "step": 51890 + }, + { + "epoch": 3.400380003931075, + "grad_norm": 0.4475633203983307, + "learning_rate": 5.099366994014556e-05, + "loss": 0.0336, + "step": 51900 + }, + { + "epoch": 3.401035183122584, + "grad_norm": 0.11846719682216644, + "learning_rate": 5.097713845346532e-05, + "loss": 0.0245, + "step": 51910 + }, + { + "epoch": 3.401690362314093, + "grad_norm": 0.45560088753700256, + "learning_rate": 5.096060685992668e-05, + "loss": 0.0323, + "step": 51920 + }, + { + "epoch": 3.4023455415056016, + "grad_norm": 0.14274610579013824, + "learning_rate": 5.0944075161337514e-05, + "loss": 0.0305, + "step": 51930 + }, + { + "epoch": 3.4030007206971105, + "grad_norm": 0.2015768438577652, + "learning_rate": 5.09275433595057e-05, + "loss": 0.0427, + "step": 51940 + }, + { + "epoch": 3.4036558998886197, + "grad_norm": 0.316655695438385, + "learning_rate": 5.0911011456239157e-05, + "loss": 0.0206, + "step": 51950 + }, + { + "epoch": 3.4043110790801285, + "grad_norm": 0.3301285207271576, + "learning_rate": 5.089447945334574e-05, + "loss": 0.0317, + "step": 51960 + }, + { + "epoch": 3.4049662582716373, + "grad_norm": 0.26029813289642334, + "learning_rate": 5.087794735263343e-05, + "loss": 0.0298, + "step": 51970 + }, + { + "epoch": 3.405621437463146, + "grad_norm": 0.11713292449712753, + "learning_rate": 5.086141515591011e-05, + "loss": 0.0444, + "step": 51980 + }, + { + "epoch": 3.406276616654655, + "grad_norm": 0.46636763215065, + "learning_rate": 5.084488286498371e-05, + "loss": 0.0392, + "step": 51990 + }, + { + "epoch": 3.406931795846164, + "grad_norm": 0.33229297399520874, + "learning_rate": 5.08283504816622e-05, + "loss": 0.0399, + "step": 52000 + }, + { + "epoch": 3.407586975037673, + "grad_norm": 0.25632816553115845, + "learning_rate": 5.081181800775353e-05, + "loss": 0.0263, + "step": 52010 + }, + { + "epoch": 3.408242154229182, + "grad_norm": 1.0927220582962036, + "learning_rate": 5.079528544506565e-05, + "loss": 0.0337, + "step": 52020 + }, + { + "epoch": 3.4088973334206907, + "grad_norm": 0.09551052004098892, + "learning_rate": 5.077875279540657e-05, + "loss": 0.0599, + "step": 52030 + }, + { + "epoch": 3.4095525126121995, + "grad_norm": 0.27561312913894653, + "learning_rate": 5.076222006058424e-05, + "loss": 0.0586, + "step": 52040 + }, + { + "epoch": 3.4102076918037083, + "grad_norm": 0.9389705657958984, + "learning_rate": 5.074568724240667e-05, + "loss": 0.0284, + "step": 52050 + }, + { + "epoch": 3.410862870995217, + "grad_norm": 0.8272232413291931, + "learning_rate": 5.072915434268187e-05, + "loss": 0.035, + "step": 52060 + }, + { + "epoch": 3.411518050186726, + "grad_norm": 0.0809953510761261, + "learning_rate": 5.071262136321785e-05, + "loss": 0.0337, + "step": 52070 + }, + { + "epoch": 3.4121732293782348, + "grad_norm": 0.7887351512908936, + "learning_rate": 5.069608830582262e-05, + "loss": 0.0806, + "step": 52080 + }, + { + "epoch": 3.412828408569744, + "grad_norm": 0.9865472316741943, + "learning_rate": 5.067955517230423e-05, + "loss": 0.0539, + "step": 52090 + }, + { + "epoch": 3.413483587761253, + "grad_norm": 1.042572021484375, + "learning_rate": 5.06630219644707e-05, + "loss": 0.049, + "step": 52100 + }, + { + "epoch": 3.4141387669527616, + "grad_norm": 1.0216041803359985, + "learning_rate": 5.064648868413008e-05, + "loss": 0.0462, + "step": 52110 + }, + { + "epoch": 3.4147939461442705, + "grad_norm": 0.36895647644996643, + "learning_rate": 5.062995533309044e-05, + "loss": 0.0511, + "step": 52120 + }, + { + "epoch": 3.4154491253357793, + "grad_norm": 0.6610296368598938, + "learning_rate": 5.061342191315983e-05, + "loss": 0.0333, + "step": 52130 + }, + { + "epoch": 3.416104304527288, + "grad_norm": 0.2618323266506195, + "learning_rate": 5.0596888426146325e-05, + "loss": 0.0252, + "step": 52140 + }, + { + "epoch": 3.416759483718797, + "grad_norm": 0.43409255146980286, + "learning_rate": 5.058035487385798e-05, + "loss": 0.0346, + "step": 52150 + }, + { + "epoch": 3.417414662910306, + "grad_norm": 0.5547110438346863, + "learning_rate": 5.0563821258102915e-05, + "loss": 0.0706, + "step": 52160 + }, + { + "epoch": 3.418069842101815, + "grad_norm": 0.4456480145454407, + "learning_rate": 5.05472875806892e-05, + "loss": 0.0362, + "step": 52170 + }, + { + "epoch": 3.418725021293324, + "grad_norm": 0.557694137096405, + "learning_rate": 5.053075384342494e-05, + "loss": 0.02, + "step": 52180 + }, + { + "epoch": 3.4193802004848326, + "grad_norm": 0.2824123203754425, + "learning_rate": 5.0514220048118234e-05, + "loss": 0.03, + "step": 52190 + }, + { + "epoch": 3.4200353796763414, + "grad_norm": 0.15897150337696075, + "learning_rate": 5.049768619657721e-05, + "loss": 0.0319, + "step": 52200 + }, + { + "epoch": 3.4206905588678502, + "grad_norm": 0.6645405292510986, + "learning_rate": 5.048115229060996e-05, + "loss": 0.0241, + "step": 52210 + }, + { + "epoch": 3.421345738059359, + "grad_norm": 0.40691158175468445, + "learning_rate": 5.046461833202462e-05, + "loss": 0.0205, + "step": 52220 + }, + { + "epoch": 3.4220009172508683, + "grad_norm": 0.6893669962882996, + "learning_rate": 5.044808432262932e-05, + "loss": 0.0409, + "step": 52230 + }, + { + "epoch": 3.422656096442377, + "grad_norm": 0.8580978512763977, + "learning_rate": 5.043155026423219e-05, + "loss": 0.0597, + "step": 52240 + }, + { + "epoch": 3.423311275633886, + "grad_norm": 1.2191637754440308, + "learning_rate": 5.041501615864137e-05, + "loss": 0.035, + "step": 52250 + }, + { + "epoch": 3.4239664548253947, + "grad_norm": 0.6336601376533508, + "learning_rate": 5.039848200766502e-05, + "loss": 0.0315, + "step": 52260 + }, + { + "epoch": 3.4246216340169036, + "grad_norm": 0.7681969404220581, + "learning_rate": 5.038194781311125e-05, + "loss": 0.0329, + "step": 52270 + }, + { + "epoch": 3.4252768132084124, + "grad_norm": 0.6958202719688416, + "learning_rate": 5.0365413576788264e-05, + "loss": 0.0723, + "step": 52280 + }, + { + "epoch": 3.425931992399921, + "grad_norm": 0.6897847056388855, + "learning_rate": 5.0348879300504194e-05, + "loss": 0.0258, + "step": 52290 + }, + { + "epoch": 3.4265871715914304, + "grad_norm": 1.100991129875183, + "learning_rate": 5.033234498606721e-05, + "loss": 0.0302, + "step": 52300 + }, + { + "epoch": 3.4272423507829393, + "grad_norm": 0.33315446972846985, + "learning_rate": 5.0315810635285476e-05, + "loss": 0.033, + "step": 52310 + }, + { + "epoch": 3.427897529974448, + "grad_norm": 1.8543405532836914, + "learning_rate": 5.0299276249967164e-05, + "loss": 0.0697, + "step": 52320 + }, + { + "epoch": 3.428552709165957, + "grad_norm": 0.6531810760498047, + "learning_rate": 5.0282741831920454e-05, + "loss": 0.0331, + "step": 52330 + }, + { + "epoch": 3.4292078883574657, + "grad_norm": 0.15889807045459747, + "learning_rate": 5.026620738295351e-05, + "loss": 0.0525, + "step": 52340 + }, + { + "epoch": 3.4298630675489745, + "grad_norm": 0.3211453855037689, + "learning_rate": 5.024967290487455e-05, + "loss": 0.0628, + "step": 52350 + }, + { + "epoch": 3.4305182467404833, + "grad_norm": 0.9977268576622009, + "learning_rate": 5.023313839949172e-05, + "loss": 0.0336, + "step": 52360 + }, + { + "epoch": 3.4311734259319926, + "grad_norm": 0.11870687454938889, + "learning_rate": 5.0216603868613224e-05, + "loss": 0.0151, + "step": 52370 + }, + { + "epoch": 3.4318286051235014, + "grad_norm": 0.1950482726097107, + "learning_rate": 5.020006931404726e-05, + "loss": 0.0492, + "step": 52380 + }, + { + "epoch": 3.4324837843150102, + "grad_norm": 0.1497829705476761, + "learning_rate": 5.0183534737602014e-05, + "loss": 0.0403, + "step": 52390 + }, + { + "epoch": 3.433138963506519, + "grad_norm": 0.7371841669082642, + "learning_rate": 5.0167000141085686e-05, + "loss": 0.0279, + "step": 52400 + }, + { + "epoch": 3.433794142698028, + "grad_norm": 0.2728865146636963, + "learning_rate": 5.015046552630647e-05, + "loss": 0.0484, + "step": 52410 + }, + { + "epoch": 3.4344493218895367, + "grad_norm": 0.20071576535701752, + "learning_rate": 5.013393089507258e-05, + "loss": 0.0174, + "step": 52420 + }, + { + "epoch": 3.4351045010810455, + "grad_norm": 0.8673457503318787, + "learning_rate": 5.011739624919219e-05, + "loss": 0.023, + "step": 52430 + }, + { + "epoch": 3.4357596802725547, + "grad_norm": 0.38495731353759766, + "learning_rate": 5.0100861590473545e-05, + "loss": 0.0298, + "step": 52440 + }, + { + "epoch": 3.4364148594640636, + "grad_norm": 0.5784446001052856, + "learning_rate": 5.008432692072482e-05, + "loss": 0.0367, + "step": 52450 + }, + { + "epoch": 3.4370700386555724, + "grad_norm": 0.8029409050941467, + "learning_rate": 5.006779224175423e-05, + "loss": 0.029, + "step": 52460 + }, + { + "epoch": 3.437725217847081, + "grad_norm": 0.20609347522258759, + "learning_rate": 5.005125755536998e-05, + "loss": 0.0261, + "step": 52470 + }, + { + "epoch": 3.43838039703859, + "grad_norm": 0.3139602839946747, + "learning_rate": 5.003472286338029e-05, + "loss": 0.0197, + "step": 52480 + }, + { + "epoch": 3.439035576230099, + "grad_norm": 0.6669353246688843, + "learning_rate": 5.001818816759335e-05, + "loss": 0.0385, + "step": 52490 + }, + { + "epoch": 3.4396907554216076, + "grad_norm": 0.2047605812549591, + "learning_rate": 5.000165346981738e-05, + "loss": 0.0569, + "step": 52500 + }, + { + "epoch": 3.4403459346131164, + "grad_norm": 0.8980453014373779, + "learning_rate": 4.9985118771860595e-05, + "loss": 0.0633, + "step": 52510 + }, + { + "epoch": 3.4410011138046257, + "grad_norm": 0.387442946434021, + "learning_rate": 4.996858407553119e-05, + "loss": 0.0181, + "step": 52520 + }, + { + "epoch": 3.4416562929961345, + "grad_norm": 0.3140571117401123, + "learning_rate": 4.99520493826374e-05, + "loss": 0.0412, + "step": 52530 + }, + { + "epoch": 3.4423114721876433, + "grad_norm": 0.28088778257369995, + "learning_rate": 4.9935514694987384e-05, + "loss": 0.0439, + "step": 52540 + }, + { + "epoch": 3.442966651379152, + "grad_norm": 0.3180124759674072, + "learning_rate": 4.991898001438941e-05, + "loss": 0.0484, + "step": 52550 + }, + { + "epoch": 3.443621830570661, + "grad_norm": 0.7487125396728516, + "learning_rate": 4.990244534265166e-05, + "loss": 0.0309, + "step": 52560 + }, + { + "epoch": 3.4442770097621698, + "grad_norm": 0.5533407330513, + "learning_rate": 4.988591068158234e-05, + "loss": 0.0167, + "step": 52570 + }, + { + "epoch": 3.444932188953679, + "grad_norm": 0.7166044116020203, + "learning_rate": 4.986937603298965e-05, + "loss": 0.053, + "step": 52580 + }, + { + "epoch": 3.445587368145188, + "grad_norm": 0.5802188515663147, + "learning_rate": 4.985284139868181e-05, + "loss": 0.0326, + "step": 52590 + }, + { + "epoch": 3.4462425473366967, + "grad_norm": 0.42574068903923035, + "learning_rate": 4.9836306780467e-05, + "loss": 0.0457, + "step": 52600 + }, + { + "epoch": 3.4468977265282055, + "grad_norm": 0.059336792677640915, + "learning_rate": 4.981977218015343e-05, + "loss": 0.0246, + "step": 52610 + }, + { + "epoch": 3.4475529057197143, + "grad_norm": 0.13705973327159882, + "learning_rate": 4.980323759954932e-05, + "loss": 0.0183, + "step": 52620 + }, + { + "epoch": 3.448208084911223, + "grad_norm": 0.5955187678337097, + "learning_rate": 4.978670304046284e-05, + "loss": 0.0306, + "step": 52630 + }, + { + "epoch": 3.448863264102732, + "grad_norm": 0.25143489241600037, + "learning_rate": 4.9770168504702195e-05, + "loss": 0.0223, + "step": 52640 + }, + { + "epoch": 3.4495184432942407, + "grad_norm": 0.4483775794506073, + "learning_rate": 4.975363399407559e-05, + "loss": 0.0355, + "step": 52650 + }, + { + "epoch": 3.45017362248575, + "grad_norm": 0.8077024817466736, + "learning_rate": 4.9737099510391194e-05, + "loss": 0.0382, + "step": 52660 + }, + { + "epoch": 3.450828801677259, + "grad_norm": 0.6916314363479614, + "learning_rate": 4.972056505545719e-05, + "loss": 0.0198, + "step": 52670 + }, + { + "epoch": 3.4514839808687676, + "grad_norm": 0.3069034218788147, + "learning_rate": 4.9704030631081775e-05, + "loss": 0.0251, + "step": 52680 + }, + { + "epoch": 3.4521391600602764, + "grad_norm": 1.0435155630111694, + "learning_rate": 4.968749623907313e-05, + "loss": 0.0357, + "step": 52690 + }, + { + "epoch": 3.4527943392517852, + "grad_norm": 0.20616617798805237, + "learning_rate": 4.967096188123943e-05, + "loss": 0.0247, + "step": 52700 + }, + { + "epoch": 3.453449518443294, + "grad_norm": 0.4969958961009979, + "learning_rate": 4.965442755938884e-05, + "loss": 0.0382, + "step": 52710 + }, + { + "epoch": 3.4541046976348033, + "grad_norm": 0.4299437999725342, + "learning_rate": 4.963789327532954e-05, + "loss": 0.0251, + "step": 52720 + }, + { + "epoch": 3.454759876826312, + "grad_norm": 0.3369503915309906, + "learning_rate": 4.962135903086968e-05, + "loss": 0.0391, + "step": 52730 + }, + { + "epoch": 3.455415056017821, + "grad_norm": 0.14033721387386322, + "learning_rate": 4.960482482781741e-05, + "loss": 0.034, + "step": 52740 + }, + { + "epoch": 3.4560702352093298, + "grad_norm": 0.2200000137090683, + "learning_rate": 4.958829066798092e-05, + "loss": 0.0214, + "step": 52750 + }, + { + "epoch": 3.4567254144008386, + "grad_norm": 0.9277251362800598, + "learning_rate": 4.957175655316835e-05, + "loss": 0.0184, + "step": 52760 + }, + { + "epoch": 3.4573805935923474, + "grad_norm": 0.4510621428489685, + "learning_rate": 4.9555222485187834e-05, + "loss": 0.0222, + "step": 52770 + }, + { + "epoch": 3.458035772783856, + "grad_norm": 0.23302531242370605, + "learning_rate": 4.953868846584752e-05, + "loss": 0.0186, + "step": 52780 + }, + { + "epoch": 3.458690951975365, + "grad_norm": 1.1831045150756836, + "learning_rate": 4.9522154496955544e-05, + "loss": 0.0413, + "step": 52790 + }, + { + "epoch": 3.4593461311668743, + "grad_norm": 0.8795732855796814, + "learning_rate": 4.950562058032001e-05, + "loss": 0.0538, + "step": 52800 + }, + { + "epoch": 3.460001310358383, + "grad_norm": 0.34702250361442566, + "learning_rate": 4.948908671774909e-05, + "loss": 0.0175, + "step": 52810 + }, + { + "epoch": 3.460656489549892, + "grad_norm": 0.4189860224723816, + "learning_rate": 4.947255291105087e-05, + "loss": 0.0234, + "step": 52820 + }, + { + "epoch": 3.4613116687414007, + "grad_norm": 0.2531570792198181, + "learning_rate": 4.945601916203348e-05, + "loss": 0.0284, + "step": 52830 + }, + { + "epoch": 3.4619668479329095, + "grad_norm": 0.5861910581588745, + "learning_rate": 4.9439485472505016e-05, + "loss": 0.0324, + "step": 52840 + }, + { + "epoch": 3.4626220271244184, + "grad_norm": 0.34751808643341064, + "learning_rate": 4.942295184427356e-05, + "loss": 0.0255, + "step": 52850 + }, + { + "epoch": 3.4632772063159276, + "grad_norm": 0.44686537981033325, + "learning_rate": 4.9406418279147234e-05, + "loss": 0.0339, + "step": 52860 + }, + { + "epoch": 3.4639323855074364, + "grad_norm": 0.295462965965271, + "learning_rate": 4.9389884778934116e-05, + "loss": 0.0314, + "step": 52870 + }, + { + "epoch": 3.4645875646989452, + "grad_norm": 0.6219167709350586, + "learning_rate": 4.937335134544228e-05, + "loss": 0.0315, + "step": 52880 + }, + { + "epoch": 3.465242743890454, + "grad_norm": 0.7555801868438721, + "learning_rate": 4.935681798047979e-05, + "loss": 0.0408, + "step": 52890 + }, + { + "epoch": 3.465897923081963, + "grad_norm": 0.17190328240394592, + "learning_rate": 4.934028468585473e-05, + "loss": 0.0188, + "step": 52900 + }, + { + "epoch": 3.4665531022734717, + "grad_norm": 0.6789793968200684, + "learning_rate": 4.932375146337514e-05, + "loss": 0.0416, + "step": 52910 + }, + { + "epoch": 3.4672082814649805, + "grad_norm": 0.1698969602584839, + "learning_rate": 4.930721831484908e-05, + "loss": 0.0221, + "step": 52920 + }, + { + "epoch": 3.4678634606564893, + "grad_norm": 0.39998096227645874, + "learning_rate": 4.929068524208456e-05, + "loss": 0.0733, + "step": 52930 + }, + { + "epoch": 3.4685186398479986, + "grad_norm": 0.6174997687339783, + "learning_rate": 4.927415224688965e-05, + "loss": 0.0323, + "step": 52940 + }, + { + "epoch": 3.4691738190395074, + "grad_norm": 0.21502085030078888, + "learning_rate": 4.9257619331072375e-05, + "loss": 0.0285, + "step": 52950 + }, + { + "epoch": 3.469828998231016, + "grad_norm": 0.6829662322998047, + "learning_rate": 4.924108649644072e-05, + "loss": 0.0346, + "step": 52960 + }, + { + "epoch": 3.470484177422525, + "grad_norm": 0.07225543260574341, + "learning_rate": 4.922455374480271e-05, + "loss": 0.0235, + "step": 52970 + }, + { + "epoch": 3.471139356614034, + "grad_norm": 0.12317764759063721, + "learning_rate": 4.920802107796634e-05, + "loss": 0.0332, + "step": 52980 + }, + { + "epoch": 3.4717945358055426, + "grad_norm": 0.2714029550552368, + "learning_rate": 4.9191488497739564e-05, + "loss": 0.0379, + "step": 52990 + }, + { + "epoch": 3.472449714997052, + "grad_norm": 0.24302709102630615, + "learning_rate": 4.917495600593042e-05, + "loss": 0.0251, + "step": 53000 + }, + { + "epoch": 3.4731048941885607, + "grad_norm": 0.5424205660820007, + "learning_rate": 4.915842360434684e-05, + "loss": 0.0335, + "step": 53010 + }, + { + "epoch": 3.4737600733800695, + "grad_norm": 0.3657538592815399, + "learning_rate": 4.914189129479679e-05, + "loss": 0.0255, + "step": 53020 + }, + { + "epoch": 3.4744152525715783, + "grad_norm": 0.47676533460617065, + "learning_rate": 4.912535907908822e-05, + "loss": 0.0342, + "step": 53030 + }, + { + "epoch": 3.475070431763087, + "grad_norm": 0.2404274344444275, + "learning_rate": 4.910882695902907e-05, + "loss": 0.0366, + "step": 53040 + }, + { + "epoch": 3.475725610954596, + "grad_norm": 0.2946668863296509, + "learning_rate": 4.9092294936427234e-05, + "loss": 0.049, + "step": 53050 + }, + { + "epoch": 3.476380790146105, + "grad_norm": 0.685038685798645, + "learning_rate": 4.9075763013090685e-05, + "loss": 0.0449, + "step": 53060 + }, + { + "epoch": 3.4770359693376136, + "grad_norm": 0.5582686066627502, + "learning_rate": 4.9059231190827305e-05, + "loss": 0.0205, + "step": 53070 + }, + { + "epoch": 3.477691148529123, + "grad_norm": 0.34372150897979736, + "learning_rate": 4.904269947144499e-05, + "loss": 0.0355, + "step": 53080 + }, + { + "epoch": 3.4783463277206317, + "grad_norm": 0.13498717546463013, + "learning_rate": 4.902616785675161e-05, + "loss": 0.0281, + "step": 53090 + }, + { + "epoch": 3.4790015069121405, + "grad_norm": 0.2948620617389679, + "learning_rate": 4.900963634855505e-05, + "loss": 0.0334, + "step": 53100 + }, + { + "epoch": 3.4796566861036493, + "grad_norm": 1.173396348953247, + "learning_rate": 4.8993104948663173e-05, + "loss": 0.0382, + "step": 53110 + }, + { + "epoch": 3.480311865295158, + "grad_norm": 0.5944495797157288, + "learning_rate": 4.8976573658883823e-05, + "loss": 0.0341, + "step": 53120 + }, + { + "epoch": 3.480967044486667, + "grad_norm": 0.18447181582450867, + "learning_rate": 4.896004248102484e-05, + "loss": 0.0225, + "step": 53130 + }, + { + "epoch": 3.481622223678176, + "grad_norm": 0.6944182515144348, + "learning_rate": 4.8943511416894045e-05, + "loss": 0.0221, + "step": 53140 + }, + { + "epoch": 3.482277402869685, + "grad_norm": 0.5683224201202393, + "learning_rate": 4.892698046829926e-05, + "loss": 0.06, + "step": 53150 + }, + { + "epoch": 3.482932582061194, + "grad_norm": 0.11115112155675888, + "learning_rate": 4.8910449637048275e-05, + "loss": 0.0318, + "step": 53160 + }, + { + "epoch": 3.4835877612527026, + "grad_norm": 0.42122170329093933, + "learning_rate": 4.8893918924948884e-05, + "loss": 0.0353, + "step": 53170 + }, + { + "epoch": 3.4842429404442115, + "grad_norm": 0.27244263887405396, + "learning_rate": 4.8877388333808835e-05, + "loss": 0.0179, + "step": 53180 + }, + { + "epoch": 3.4848981196357203, + "grad_norm": 0.6309353113174438, + "learning_rate": 4.8860857865435925e-05, + "loss": 0.0353, + "step": 53190 + }, + { + "epoch": 3.485553298827229, + "grad_norm": 0.34248480200767517, + "learning_rate": 4.884432752163789e-05, + "loss": 0.0221, + "step": 53200 + }, + { + "epoch": 3.486208478018738, + "grad_norm": 0.9294228553771973, + "learning_rate": 4.8827797304222466e-05, + "loss": 0.0481, + "step": 53210 + }, + { + "epoch": 3.486863657210247, + "grad_norm": 0.08046144992113113, + "learning_rate": 4.8811267214997366e-05, + "loss": 0.0246, + "step": 53220 + }, + { + "epoch": 3.487518836401756, + "grad_norm": 0.06225018575787544, + "learning_rate": 4.879473725577029e-05, + "loss": 0.0609, + "step": 53230 + }, + { + "epoch": 3.488174015593265, + "grad_norm": 1.29642915725708, + "learning_rate": 4.877820742834891e-05, + "loss": 0.0361, + "step": 53240 + }, + { + "epoch": 3.4888291947847736, + "grad_norm": 1.3313719034194946, + "learning_rate": 4.8761677734540956e-05, + "loss": 0.0464, + "step": 53250 + }, + { + "epoch": 3.4894843739762824, + "grad_norm": 0.8334248065948486, + "learning_rate": 4.874514817615407e-05, + "loss": 0.0316, + "step": 53260 + }, + { + "epoch": 3.4901395531677912, + "grad_norm": 0.12303437292575836, + "learning_rate": 4.872861875499588e-05, + "loss": 0.0189, + "step": 53270 + }, + { + "epoch": 3.4907947323593005, + "grad_norm": 0.6575003862380981, + "learning_rate": 4.871208947287404e-05, + "loss": 0.0774, + "step": 53280 + }, + { + "epoch": 3.4914499115508093, + "grad_norm": 0.671843409538269, + "learning_rate": 4.869556033159615e-05, + "loss": 0.0639, + "step": 53290 + }, + { + "epoch": 3.492105090742318, + "grad_norm": 0.5164141058921814, + "learning_rate": 4.8679031332969835e-05, + "loss": 0.0262, + "step": 53300 + }, + { + "epoch": 3.492760269933827, + "grad_norm": 0.2646293640136719, + "learning_rate": 4.8662502478802635e-05, + "loss": 0.028, + "step": 53310 + }, + { + "epoch": 3.4934154491253357, + "grad_norm": 1.0333983898162842, + "learning_rate": 4.8645973770902176e-05, + "loss": 0.0633, + "step": 53320 + }, + { + "epoch": 3.4940706283168446, + "grad_norm": 0.6401782631874084, + "learning_rate": 4.862944521107599e-05, + "loss": 0.0338, + "step": 53330 + }, + { + "epoch": 3.4947258075083534, + "grad_norm": 0.19079998135566711, + "learning_rate": 4.861291680113162e-05, + "loss": 0.0172, + "step": 53340 + }, + { + "epoch": 3.495380986699862, + "grad_norm": 1.0327544212341309, + "learning_rate": 4.8596388542876563e-05, + "loss": 0.035, + "step": 53350 + }, + { + "epoch": 3.4960361658913714, + "grad_norm": 0.4837036430835724, + "learning_rate": 4.857986043811836e-05, + "loss": 0.032, + "step": 53360 + }, + { + "epoch": 3.4966913450828803, + "grad_norm": 0.4095124900341034, + "learning_rate": 4.856333248866448e-05, + "loss": 0.0374, + "step": 53370 + }, + { + "epoch": 3.497346524274389, + "grad_norm": 0.779918909072876, + "learning_rate": 4.85468046963224e-05, + "loss": 0.0319, + "step": 53380 + }, + { + "epoch": 3.498001703465898, + "grad_norm": 0.33319827914237976, + "learning_rate": 4.853027706289957e-05, + "loss": 0.0347, + "step": 53390 + }, + { + "epoch": 3.4986568826574067, + "grad_norm": 0.4478291869163513, + "learning_rate": 4.851374959020344e-05, + "loss": 0.0288, + "step": 53400 + }, + { + "epoch": 3.4993120618489155, + "grad_norm": 0.34909898042678833, + "learning_rate": 4.849722228004141e-05, + "loss": 0.025, + "step": 53410 + }, + { + "epoch": 3.499967241040425, + "grad_norm": 0.27042055130004883, + "learning_rate": 4.848069513422091e-05, + "loss": 0.0553, + "step": 53420 + }, + { + "epoch": 3.5006224202319336, + "grad_norm": 0.739313542842865, + "learning_rate": 4.8464168154549294e-05, + "loss": 0.0426, + "step": 53430 + }, + { + "epoch": 3.5012775994234424, + "grad_norm": 0.82570880651474, + "learning_rate": 4.844764134283392e-05, + "loss": 0.0342, + "step": 53440 + }, + { + "epoch": 3.5019327786149512, + "grad_norm": 0.12761621177196503, + "learning_rate": 4.843111470088218e-05, + "loss": 0.0268, + "step": 53450 + }, + { + "epoch": 3.50258795780646, + "grad_norm": 0.40381523966789246, + "learning_rate": 4.841458823050137e-05, + "loss": 0.0291, + "step": 53460 + }, + { + "epoch": 3.503243136997969, + "grad_norm": 0.13999144732952118, + "learning_rate": 4.8398061933498816e-05, + "loss": 0.0207, + "step": 53470 + }, + { + "epoch": 3.5038983161894777, + "grad_norm": 2.2426693439483643, + "learning_rate": 4.83815358116818e-05, + "loss": 0.0415, + "step": 53480 + }, + { + "epoch": 3.5045534953809865, + "grad_norm": 0.17378798127174377, + "learning_rate": 4.8365009866857584e-05, + "loss": 0.0233, + "step": 53490 + }, + { + "epoch": 3.5052086745724957, + "grad_norm": 0.3764123022556305, + "learning_rate": 4.834848410083342e-05, + "loss": 0.0293, + "step": 53500 + }, + { + "epoch": 3.5058638537640046, + "grad_norm": 0.37144824862480164, + "learning_rate": 4.833195851541657e-05, + "loss": 0.0269, + "step": 53510 + }, + { + "epoch": 3.5065190329555134, + "grad_norm": 0.11514919251203537, + "learning_rate": 4.831543311241423e-05, + "loss": 0.0492, + "step": 53520 + }, + { + "epoch": 3.507174212147022, + "grad_norm": 0.08574477583169937, + "learning_rate": 4.829890789363359e-05, + "loss": 0.0296, + "step": 53530 + }, + { + "epoch": 3.507829391338531, + "grad_norm": 0.5507720112800598, + "learning_rate": 4.828238286088182e-05, + "loss": 0.0443, + "step": 53540 + }, + { + "epoch": 3.50848457053004, + "grad_norm": 0.806788444519043, + "learning_rate": 4.826585801596609e-05, + "loss": 0.0418, + "step": 53550 + }, + { + "epoch": 3.509139749721549, + "grad_norm": 0.5750271081924438, + "learning_rate": 4.8249333360693486e-05, + "loss": 0.0419, + "step": 53560 + }, + { + "epoch": 3.509794928913058, + "grad_norm": 0.20661881566047668, + "learning_rate": 4.8232808896871185e-05, + "loss": 0.0223, + "step": 53570 + }, + { + "epoch": 3.5104501081045667, + "grad_norm": 0.22696195542812347, + "learning_rate": 4.821628462630624e-05, + "loss": 0.0535, + "step": 53580 + }, + { + "epoch": 3.5111052872960755, + "grad_norm": 0.7732415199279785, + "learning_rate": 4.8199760550805735e-05, + "loss": 0.0554, + "step": 53590 + }, + { + "epoch": 3.5117604664875843, + "grad_norm": 0.13846231997013092, + "learning_rate": 4.818323667217669e-05, + "loss": 0.024, + "step": 53600 + }, + { + "epoch": 3.512415645679093, + "grad_norm": 0.8963651061058044, + "learning_rate": 4.8166712992226165e-05, + "loss": 0.0462, + "step": 53610 + }, + { + "epoch": 3.513070824870602, + "grad_norm": 0.5114948153495789, + "learning_rate": 4.8150189512761154e-05, + "loss": 0.0474, + "step": 53620 + }, + { + "epoch": 3.5137260040621108, + "grad_norm": 1.3400605916976929, + "learning_rate": 4.813366623558862e-05, + "loss": 0.0402, + "step": 53630 + }, + { + "epoch": 3.5143811832536196, + "grad_norm": 0.46251341700553894, + "learning_rate": 4.811714316251554e-05, + "loss": 0.0398, + "step": 53640 + }, + { + "epoch": 3.515036362445129, + "grad_norm": 0.5183666348457336, + "learning_rate": 4.810062029534886e-05, + "loss": 0.0185, + "step": 53650 + }, + { + "epoch": 3.5156915416366377, + "grad_norm": 0.5447751879692078, + "learning_rate": 4.8084097635895505e-05, + "loss": 0.0549, + "step": 53660 + }, + { + "epoch": 3.5163467208281465, + "grad_norm": 0.27376314997673035, + "learning_rate": 4.8067575185962335e-05, + "loss": 0.0332, + "step": 53670 + }, + { + "epoch": 3.5170019000196553, + "grad_norm": 0.18075039982795715, + "learning_rate": 4.805105294735625e-05, + "loss": 0.0204, + "step": 53680 + }, + { + "epoch": 3.517657079211164, + "grad_norm": 0.3927735984325409, + "learning_rate": 4.803453092188406e-05, + "loss": 0.0315, + "step": 53690 + }, + { + "epoch": 3.5183122584026734, + "grad_norm": 0.22009976208209991, + "learning_rate": 4.801800911135263e-05, + "loss": 0.0274, + "step": 53700 + }, + { + "epoch": 3.518967437594182, + "grad_norm": 0.1519327312707901, + "learning_rate": 4.8001487517568744e-05, + "loss": 0.0485, + "step": 53710 + }, + { + "epoch": 3.519622616785691, + "grad_norm": 0.8443809747695923, + "learning_rate": 4.798496614233918e-05, + "loss": 0.0479, + "step": 53720 + }, + { + "epoch": 3.5202777959772, + "grad_norm": 1.0219645500183105, + "learning_rate": 4.796844498747069e-05, + "loss": 0.0413, + "step": 53730 + }, + { + "epoch": 3.5209329751687086, + "grad_norm": 0.18417400121688843, + "learning_rate": 4.795192405477e-05, + "loss": 0.0485, + "step": 53740 + }, + { + "epoch": 3.5215881543602174, + "grad_norm": 0.7506529688835144, + "learning_rate": 4.79354033460438e-05, + "loss": 0.0339, + "step": 53750 + }, + { + "epoch": 3.5222433335517263, + "grad_norm": 0.7657853364944458, + "learning_rate": 4.79188828630988e-05, + "loss": 0.0359, + "step": 53760 + }, + { + "epoch": 3.522898512743235, + "grad_norm": 0.6010372042655945, + "learning_rate": 4.790236260774165e-05, + "loss": 0.0467, + "step": 53770 + }, + { + "epoch": 3.523553691934744, + "grad_norm": 0.9727668166160583, + "learning_rate": 4.788584258177896e-05, + "loss": 0.0634, + "step": 53780 + }, + { + "epoch": 3.524208871126253, + "grad_norm": 0.793401300907135, + "learning_rate": 4.7869322787017356e-05, + "loss": 0.0523, + "step": 53790 + }, + { + "epoch": 3.524864050317762, + "grad_norm": 0.6882519721984863, + "learning_rate": 4.78528032252634e-05, + "loss": 0.0478, + "step": 53800 + }, + { + "epoch": 3.5255192295092708, + "grad_norm": 0.25085312128067017, + "learning_rate": 4.783628389832366e-05, + "loss": 0.0526, + "step": 53810 + }, + { + "epoch": 3.5261744087007796, + "grad_norm": 0.5001278519630432, + "learning_rate": 4.7819764808004644e-05, + "loss": 0.0418, + "step": 53820 + }, + { + "epoch": 3.5268295878922884, + "grad_norm": 0.549870491027832, + "learning_rate": 4.780324595611289e-05, + "loss": 0.0206, + "step": 53830 + }, + { + "epoch": 3.5274847670837977, + "grad_norm": 0.3161821961402893, + "learning_rate": 4.7786727344454854e-05, + "loss": 0.0375, + "step": 53840 + }, + { + "epoch": 3.5281399462753065, + "grad_norm": 0.7666614055633545, + "learning_rate": 4.777020897483697e-05, + "loss": 0.0719, + "step": 53850 + }, + { + "epoch": 3.5287951254668153, + "grad_norm": 0.6675565838813782, + "learning_rate": 4.775369084906569e-05, + "loss": 0.0345, + "step": 53860 + }, + { + "epoch": 3.529450304658324, + "grad_norm": 1.110664963722229, + "learning_rate": 4.773717296894742e-05, + "loss": 0.0383, + "step": 53870 + }, + { + "epoch": 3.530105483849833, + "grad_norm": 0.26534897089004517, + "learning_rate": 4.772065533628848e-05, + "loss": 0.0448, + "step": 53880 + }, + { + "epoch": 3.5307606630413417, + "grad_norm": 0.6211462020874023, + "learning_rate": 4.7704137952895265e-05, + "loss": 0.0491, + "step": 53890 + }, + { + "epoch": 3.5314158422328505, + "grad_norm": 0.12905296683311462, + "learning_rate": 4.768762082057407e-05, + "loss": 0.0286, + "step": 53900 + }, + { + "epoch": 3.5320710214243594, + "grad_norm": 0.1816033273935318, + "learning_rate": 4.7671103941131186e-05, + "loss": 0.0181, + "step": 53910 + }, + { + "epoch": 3.532726200615868, + "grad_norm": 0.7982630133628845, + "learning_rate": 4.765458731637287e-05, + "loss": 0.0492, + "step": 53920 + }, + { + "epoch": 3.5333813798073774, + "grad_norm": 0.22183574736118317, + "learning_rate": 4.763807094810537e-05, + "loss": 0.0397, + "step": 53930 + }, + { + "epoch": 3.5340365589988862, + "grad_norm": 0.5612366199493408, + "learning_rate": 4.762155483813487e-05, + "loss": 0.0302, + "step": 53940 + }, + { + "epoch": 3.534691738190395, + "grad_norm": 0.19151674211025238, + "learning_rate": 4.760503898826754e-05, + "loss": 0.0166, + "step": 53950 + }, + { + "epoch": 3.535346917381904, + "grad_norm": 0.8612596988677979, + "learning_rate": 4.758852340030957e-05, + "loss": 0.0395, + "step": 53960 + }, + { + "epoch": 3.5360020965734127, + "grad_norm": 0.20002727210521698, + "learning_rate": 4.757200807606705e-05, + "loss": 0.0153, + "step": 53970 + }, + { + "epoch": 3.536657275764922, + "grad_norm": 2.7580864429473877, + "learning_rate": 4.755549301734607e-05, + "loss": 0.031, + "step": 53980 + }, + { + "epoch": 3.5373124549564308, + "grad_norm": 0.27209237217903137, + "learning_rate": 4.753897822595271e-05, + "loss": 0.0297, + "step": 53990 + }, + { + "epoch": 3.5379676341479396, + "grad_norm": 0.35864493250846863, + "learning_rate": 4.7522463703692994e-05, + "loss": 0.023, + "step": 54000 + }, + { + "epoch": 3.5386228133394484, + "grad_norm": 0.7459089159965515, + "learning_rate": 4.75059494523729e-05, + "loss": 0.0347, + "step": 54010 + }, + { + "epoch": 3.539277992530957, + "grad_norm": 0.7097464799880981, + "learning_rate": 4.748943547379843e-05, + "loss": 0.052, + "step": 54020 + }, + { + "epoch": 3.539933171722466, + "grad_norm": 0.39677244424819946, + "learning_rate": 4.7472921769775535e-05, + "loss": 0.036, + "step": 54030 + }, + { + "epoch": 3.540588350913975, + "grad_norm": 0.27478328347206116, + "learning_rate": 4.745640834211011e-05, + "loss": 0.0391, + "step": 54040 + }, + { + "epoch": 3.5412435301054837, + "grad_norm": 0.24896195530891418, + "learning_rate": 4.7439895192608054e-05, + "loss": 0.0217, + "step": 54050 + }, + { + "epoch": 3.5418987092969925, + "grad_norm": 0.4230068027973175, + "learning_rate": 4.7423382323075186e-05, + "loss": 0.0623, + "step": 54060 + }, + { + "epoch": 3.5425538884885017, + "grad_norm": 0.3307775855064392, + "learning_rate": 4.740686973531737e-05, + "loss": 0.0412, + "step": 54070 + }, + { + "epoch": 3.5432090676800105, + "grad_norm": 0.6425256133079529, + "learning_rate": 4.7390357431140375e-05, + "loss": 0.0342, + "step": 54080 + }, + { + "epoch": 3.5438642468715194, + "grad_norm": 0.26353514194488525, + "learning_rate": 4.737384541234998e-05, + "loss": 0.0469, + "step": 54090 + }, + { + "epoch": 3.544519426063028, + "grad_norm": 0.26378098130226135, + "learning_rate": 4.735733368075189e-05, + "loss": 0.0533, + "step": 54100 + }, + { + "epoch": 3.545174605254537, + "grad_norm": 0.38447973132133484, + "learning_rate": 4.734082223815182e-05, + "loss": 0.0306, + "step": 54110 + }, + { + "epoch": 3.5458297844460462, + "grad_norm": 0.9404019713401794, + "learning_rate": 4.7324311086355436e-05, + "loss": 0.028, + "step": 54120 + }, + { + "epoch": 3.546484963637555, + "grad_norm": 1.136273741722107, + "learning_rate": 4.730780022716836e-05, + "loss": 0.06, + "step": 54130 + }, + { + "epoch": 3.547140142829064, + "grad_norm": 0.7305817008018494, + "learning_rate": 4.729128966239619e-05, + "loss": 0.0543, + "step": 54140 + }, + { + "epoch": 3.5477953220205727, + "grad_norm": 0.8721815347671509, + "learning_rate": 4.7274779393844536e-05, + "loss": 0.0299, + "step": 54150 + }, + { + "epoch": 3.5484505012120815, + "grad_norm": 0.442929744720459, + "learning_rate": 4.725826942331891e-05, + "loss": 0.0255, + "step": 54160 + }, + { + "epoch": 3.5491056804035903, + "grad_norm": 0.4552248418331146, + "learning_rate": 4.724175975262481e-05, + "loss": 0.0386, + "step": 54170 + }, + { + "epoch": 3.549760859595099, + "grad_norm": 0.6098262667655945, + "learning_rate": 4.722525038356772e-05, + "loss": 0.0332, + "step": 54180 + }, + { + "epoch": 3.550416038786608, + "grad_norm": 0.9117936491966248, + "learning_rate": 4.7208741317953075e-05, + "loss": 0.0426, + "step": 54190 + }, + { + "epoch": 3.5510712179781168, + "grad_norm": 0.7278691530227661, + "learning_rate": 4.7192232557586264e-05, + "loss": 0.0249, + "step": 54200 + }, + { + "epoch": 3.551726397169626, + "grad_norm": 0.6346930861473083, + "learning_rate": 4.717572410427269e-05, + "loss": 0.0409, + "step": 54210 + }, + { + "epoch": 3.552381576361135, + "grad_norm": 0.21639324724674225, + "learning_rate": 4.71592159598177e-05, + "loss": 0.0376, + "step": 54220 + }, + { + "epoch": 3.5530367555526436, + "grad_norm": 0.5314701795578003, + "learning_rate": 4.714270812602657e-05, + "loss": 0.0411, + "step": 54230 + }, + { + "epoch": 3.5536919347441525, + "grad_norm": 0.5190209150314331, + "learning_rate": 4.712620060470458e-05, + "loss": 0.022, + "step": 54240 + }, + { + "epoch": 3.5543471139356613, + "grad_norm": 0.5751194357872009, + "learning_rate": 4.7109693397656986e-05, + "loss": 0.0294, + "step": 54250 + }, + { + "epoch": 3.5550022931271705, + "grad_norm": 0.40157049894332886, + "learning_rate": 4.709318650668894e-05, + "loss": 0.0183, + "step": 54260 + }, + { + "epoch": 3.5556574723186793, + "grad_norm": 0.23272421956062317, + "learning_rate": 4.7076679933605676e-05, + "loss": 0.0427, + "step": 54270 + }, + { + "epoch": 3.556312651510188, + "grad_norm": 0.24812853336334229, + "learning_rate": 4.70601736802123e-05, + "loss": 0.0159, + "step": 54280 + }, + { + "epoch": 3.556967830701697, + "grad_norm": 1.0295279026031494, + "learning_rate": 4.704366774831391e-05, + "loss": 0.0365, + "step": 54290 + }, + { + "epoch": 3.557623009893206, + "grad_norm": 0.29120492935180664, + "learning_rate": 4.702716213971557e-05, + "loss": 0.0285, + "step": 54300 + }, + { + "epoch": 3.5582781890847146, + "grad_norm": 0.5836547613143921, + "learning_rate": 4.70106568562223e-05, + "loss": 0.0868, + "step": 54310 + }, + { + "epoch": 3.5589333682762234, + "grad_norm": 0.4041314125061035, + "learning_rate": 4.6994151899639105e-05, + "loss": 0.0398, + "step": 54320 + }, + { + "epoch": 3.5595885474677322, + "grad_norm": 0.230789452791214, + "learning_rate": 4.697764727177093e-05, + "loss": 0.054, + "step": 54330 + }, + { + "epoch": 3.560243726659241, + "grad_norm": 0.5554624199867249, + "learning_rate": 4.696114297442272e-05, + "loss": 0.0221, + "step": 54340 + }, + { + "epoch": 3.5608989058507503, + "grad_norm": 0.646594226360321, + "learning_rate": 4.694463900939933e-05, + "loss": 0.0476, + "step": 54350 + }, + { + "epoch": 3.561554085042259, + "grad_norm": 0.23333962261676788, + "learning_rate": 4.692813537850564e-05, + "loss": 0.0506, + "step": 54360 + }, + { + "epoch": 3.562209264233768, + "grad_norm": 0.7469705939292908, + "learning_rate": 4.6911632083546445e-05, + "loss": 0.0926, + "step": 54370 + }, + { + "epoch": 3.5628644434252768, + "grad_norm": 0.4771120250225067, + "learning_rate": 4.6895129126326525e-05, + "loss": 0.0447, + "step": 54380 + }, + { + "epoch": 3.5635196226167856, + "grad_norm": 0.7001339793205261, + "learning_rate": 4.687862650865059e-05, + "loss": 0.0392, + "step": 54390 + }, + { + "epoch": 3.564174801808295, + "grad_norm": 0.5614458322525024, + "learning_rate": 4.686212423232339e-05, + "loss": 0.0524, + "step": 54400 + }, + { + "epoch": 3.5648299809998036, + "grad_norm": 1.1450564861297607, + "learning_rate": 4.6845622299149565e-05, + "loss": 0.0361, + "step": 54410 + }, + { + "epoch": 3.5654851601913125, + "grad_norm": 0.1497219055891037, + "learning_rate": 4.682912071093374e-05, + "loss": 0.0281, + "step": 54420 + }, + { + "epoch": 3.5661403393828213, + "grad_norm": 0.4896468222141266, + "learning_rate": 4.681261946948052e-05, + "loss": 0.0242, + "step": 54430 + }, + { + "epoch": 3.56679551857433, + "grad_norm": 0.6380288600921631, + "learning_rate": 4.679611857659443e-05, + "loss": 0.03, + "step": 54440 + }, + { + "epoch": 3.567450697765839, + "grad_norm": 0.9410753846168518, + "learning_rate": 4.6779618034079975e-05, + "loss": 0.0286, + "step": 54450 + }, + { + "epoch": 3.5681058769573477, + "grad_norm": 1.1756123304367065, + "learning_rate": 4.6763117843741675e-05, + "loss": 0.0247, + "step": 54460 + }, + { + "epoch": 3.5687610561488565, + "grad_norm": 0.368979811668396, + "learning_rate": 4.674661800738394e-05, + "loss": 0.0463, + "step": 54470 + }, + { + "epoch": 3.5694162353403653, + "grad_norm": 0.5903451442718506, + "learning_rate": 4.673011852681117e-05, + "loss": 0.0395, + "step": 54480 + }, + { + "epoch": 3.5700714145318746, + "grad_norm": 0.2191721498966217, + "learning_rate": 4.671361940382772e-05, + "loss": 0.04, + "step": 54490 + }, + { + "epoch": 3.5707265937233834, + "grad_norm": 0.3743134140968323, + "learning_rate": 4.669712064023791e-05, + "loss": 0.0326, + "step": 54500 + }, + { + "epoch": 3.5713817729148922, + "grad_norm": 0.13690447807312012, + "learning_rate": 4.668062223784602e-05, + "loss": 0.0361, + "step": 54510 + }, + { + "epoch": 3.572036952106401, + "grad_norm": 0.617144763469696, + "learning_rate": 4.6664124198456274e-05, + "loss": 0.0276, + "step": 54520 + }, + { + "epoch": 3.57269213129791, + "grad_norm": 0.6948131322860718, + "learning_rate": 4.6647626523872905e-05, + "loss": 0.0268, + "step": 54530 + }, + { + "epoch": 3.573347310489419, + "grad_norm": 0.20624174177646637, + "learning_rate": 4.6631129215900066e-05, + "loss": 0.0321, + "step": 54540 + }, + { + "epoch": 3.574002489680928, + "grad_norm": 0.09785109758377075, + "learning_rate": 4.661463227634187e-05, + "loss": 0.0285, + "step": 54550 + }, + { + "epoch": 3.5746576688724367, + "grad_norm": 0.24102027714252472, + "learning_rate": 4.659813570700238e-05, + "loss": 0.0259, + "step": 54560 + }, + { + "epoch": 3.5753128480639456, + "grad_norm": 0.18269091844558716, + "learning_rate": 4.658163950968568e-05, + "loss": 0.034, + "step": 54570 + }, + { + "epoch": 3.5759680272554544, + "grad_norm": 0.19956499338150024, + "learning_rate": 4.6565143686195726e-05, + "loss": 0.0293, + "step": 54580 + }, + { + "epoch": 3.576623206446963, + "grad_norm": 0.6083377599716187, + "learning_rate": 4.654864823833649e-05, + "loss": 0.0306, + "step": 54590 + }, + { + "epoch": 3.577278385638472, + "grad_norm": 0.16514572501182556, + "learning_rate": 4.65321531679119e-05, + "loss": 0.0649, + "step": 54600 + }, + { + "epoch": 3.577933564829981, + "grad_norm": 0.13849443197250366, + "learning_rate": 4.6515658476725834e-05, + "loss": 0.0366, + "step": 54610 + }, + { + "epoch": 3.5785887440214896, + "grad_norm": 0.4173848330974579, + "learning_rate": 4.649916416658212e-05, + "loss": 0.0396, + "step": 54620 + }, + { + "epoch": 3.579243923212999, + "grad_norm": 0.38196223974227905, + "learning_rate": 4.648267023928455e-05, + "loss": 0.0303, + "step": 54630 + }, + { + "epoch": 3.5798991024045077, + "grad_norm": 0.7236614227294922, + "learning_rate": 4.646617669663687e-05, + "loss": 0.046, + "step": 54640 + }, + { + "epoch": 3.5805542815960165, + "grad_norm": 0.31648409366607666, + "learning_rate": 4.644968354044277e-05, + "loss": 0.0344, + "step": 54650 + }, + { + "epoch": 3.5812094607875253, + "grad_norm": 0.11456429213285446, + "learning_rate": 4.643319077250596e-05, + "loss": 0.022, + "step": 54660 + }, + { + "epoch": 3.581864639979034, + "grad_norm": 0.33716705441474915, + "learning_rate": 4.6416698394630055e-05, + "loss": 0.0418, + "step": 54670 + }, + { + "epoch": 3.5825198191705434, + "grad_norm": 0.17629435658454895, + "learning_rate": 4.640020640861862e-05, + "loss": 0.0473, + "step": 54680 + }, + { + "epoch": 3.583174998362052, + "grad_norm": 0.47574493288993835, + "learning_rate": 4.6383714816275195e-05, + "loss": 0.0283, + "step": 54690 + }, + { + "epoch": 3.583830177553561, + "grad_norm": 0.39605170488357544, + "learning_rate": 4.636722361940329e-05, + "loss": 0.0372, + "step": 54700 + }, + { + "epoch": 3.58448535674507, + "grad_norm": 0.2257518321275711, + "learning_rate": 4.635073281980632e-05, + "loss": 0.0374, + "step": 54710 + }, + { + "epoch": 3.5851405359365787, + "grad_norm": 0.16165313124656677, + "learning_rate": 4.633424241928775e-05, + "loss": 0.023, + "step": 54720 + }, + { + "epoch": 3.5857957151280875, + "grad_norm": 0.17816674709320068, + "learning_rate": 4.6317752419650914e-05, + "loss": 0.0247, + "step": 54730 + }, + { + "epoch": 3.5864508943195963, + "grad_norm": 0.4549180567264557, + "learning_rate": 4.630126282269914e-05, + "loss": 0.026, + "step": 54740 + }, + { + "epoch": 3.587106073511105, + "grad_norm": 0.9518094062805176, + "learning_rate": 4.628477363023569e-05, + "loss": 0.0326, + "step": 54750 + }, + { + "epoch": 3.587761252702614, + "grad_norm": 0.4933185279369354, + "learning_rate": 4.626828484406383e-05, + "loss": 0.0439, + "step": 54760 + }, + { + "epoch": 3.588416431894123, + "grad_norm": 0.2799457907676697, + "learning_rate": 4.625179646598669e-05, + "loss": 0.0267, + "step": 54770 + }, + { + "epoch": 3.589071611085632, + "grad_norm": 0.6884332299232483, + "learning_rate": 4.6235308497807487e-05, + "loss": 0.0354, + "step": 54780 + }, + { + "epoch": 3.589726790277141, + "grad_norm": 0.1746470183134079, + "learning_rate": 4.6218820941329276e-05, + "loss": 0.0344, + "step": 54790 + }, + { + "epoch": 3.5903819694686496, + "grad_norm": 0.5721423625946045, + "learning_rate": 4.620233379835513e-05, + "loss": 0.0497, + "step": 54800 + }, + { + "epoch": 3.5910371486601584, + "grad_norm": 1.0694043636322021, + "learning_rate": 4.618584707068804e-05, + "loss": 0.0393, + "step": 54810 + }, + { + "epoch": 3.5916923278516677, + "grad_norm": 0.44754233956336975, + "learning_rate": 4.616936076013099e-05, + "loss": 0.0243, + "step": 54820 + }, + { + "epoch": 3.5923475070431765, + "grad_norm": 0.44993942975997925, + "learning_rate": 4.615287486848688e-05, + "loss": 0.0272, + "step": 54830 + }, + { + "epoch": 3.5930026862346853, + "grad_norm": 0.07680286467075348, + "learning_rate": 4.6136389397558584e-05, + "loss": 0.0473, + "step": 54840 + }, + { + "epoch": 3.593657865426194, + "grad_norm": 1.2256214618682861, + "learning_rate": 4.6119904349148935e-05, + "loss": 0.036, + "step": 54850 + }, + { + "epoch": 3.594313044617703, + "grad_norm": 0.2157352715730667, + "learning_rate": 4.6103419725060726e-05, + "loss": 0.022, + "step": 54860 + }, + { + "epoch": 3.5949682238092118, + "grad_norm": 0.5329359173774719, + "learning_rate": 4.608693552709667e-05, + "loss": 0.0485, + "step": 54870 + }, + { + "epoch": 3.5956234030007206, + "grad_norm": 1.1193726062774658, + "learning_rate": 4.607045175705947e-05, + "loss": 0.0216, + "step": 54880 + }, + { + "epoch": 3.5962785821922294, + "grad_norm": 0.2527482211589813, + "learning_rate": 4.605396841675176e-05, + "loss": 0.0306, + "step": 54890 + }, + { + "epoch": 3.596933761383738, + "grad_norm": 1.04144287109375, + "learning_rate": 4.6037485507976094e-05, + "loss": 0.0271, + "step": 54900 + }, + { + "epoch": 3.5975889405752475, + "grad_norm": 0.4049440324306488, + "learning_rate": 4.602100303253509e-05, + "loss": 0.0281, + "step": 54910 + }, + { + "epoch": 3.5982441197667563, + "grad_norm": 0.6023936867713928, + "learning_rate": 4.600452099223122e-05, + "loss": 0.0571, + "step": 54920 + }, + { + "epoch": 3.598899298958265, + "grad_norm": 0.5812992453575134, + "learning_rate": 4.598803938886691e-05, + "loss": 0.0441, + "step": 54930 + }, + { + "epoch": 3.599554478149774, + "grad_norm": 0.4885667562484741, + "learning_rate": 4.597155822424459e-05, + "loss": 0.0165, + "step": 54940 + }, + { + "epoch": 3.6002096573412827, + "grad_norm": 0.6655780673027039, + "learning_rate": 4.59550775001666e-05, + "loss": 0.0401, + "step": 54950 + }, + { + "epoch": 3.600864836532792, + "grad_norm": 0.3292076587677002, + "learning_rate": 4.5938597218435234e-05, + "loss": 0.0405, + "step": 54960 + }, + { + "epoch": 3.601520015724301, + "grad_norm": 0.6925805807113647, + "learning_rate": 4.592211738085279e-05, + "loss": 0.0291, + "step": 54970 + }, + { + "epoch": 3.6021751949158096, + "grad_norm": 0.5289504528045654, + "learning_rate": 4.590563798922146e-05, + "loss": 0.0326, + "step": 54980 + }, + { + "epoch": 3.6028303741073184, + "grad_norm": 0.6995851397514343, + "learning_rate": 4.5889159045343404e-05, + "loss": 0.013, + "step": 54990 + }, + { + "epoch": 3.6034855532988272, + "grad_norm": 0.2389146387577057, + "learning_rate": 4.5872680551020734e-05, + "loss": 0.0243, + "step": 55000 + }, + { + "epoch": 3.604140732490336, + "grad_norm": 0.3536697328090668, + "learning_rate": 4.585620250805552e-05, + "loss": 0.0251, + "step": 55010 + }, + { + "epoch": 3.604795911681845, + "grad_norm": 0.3869779109954834, + "learning_rate": 4.583972491824975e-05, + "loss": 0.042, + "step": 55020 + }, + { + "epoch": 3.6054510908733537, + "grad_norm": 1.2049752473831177, + "learning_rate": 4.582324778340541e-05, + "loss": 0.0664, + "step": 55030 + }, + { + "epoch": 3.6061062700648625, + "grad_norm": 0.8209589123725891, + "learning_rate": 4.5806771105324424e-05, + "loss": 0.0548, + "step": 55040 + }, + { + "epoch": 3.6067614492563718, + "grad_norm": 0.302094966173172, + "learning_rate": 4.579029488580865e-05, + "loss": 0.034, + "step": 55050 + }, + { + "epoch": 3.6074166284478806, + "grad_norm": 0.5508453845977783, + "learning_rate": 4.5773819126659876e-05, + "loss": 0.0298, + "step": 55060 + }, + { + "epoch": 3.6080718076393894, + "grad_norm": 0.9656165242195129, + "learning_rate": 4.575734382967991e-05, + "loss": 0.0374, + "step": 55070 + }, + { + "epoch": 3.608726986830898, + "grad_norm": 0.3732859194278717, + "learning_rate": 4.574086899667044e-05, + "loss": 0.0119, + "step": 55080 + }, + { + "epoch": 3.609382166022407, + "grad_norm": 0.8286269903182983, + "learning_rate": 4.5724394629433106e-05, + "loss": 0.032, + "step": 55090 + }, + { + "epoch": 3.6100373452139163, + "grad_norm": 0.9772318005561829, + "learning_rate": 4.570792072976958e-05, + "loss": 0.0282, + "step": 55100 + }, + { + "epoch": 3.610692524405425, + "grad_norm": 0.30898553133010864, + "learning_rate": 4.569144729948138e-05, + "loss": 0.0252, + "step": 55110 + }, + { + "epoch": 3.611347703596934, + "grad_norm": 0.8768417239189148, + "learning_rate": 4.5674974340370033e-05, + "loss": 0.0441, + "step": 55120 + }, + { + "epoch": 3.6120028827884427, + "grad_norm": 0.7223113179206848, + "learning_rate": 4.565850185423699e-05, + "loss": 0.0448, + "step": 55130 + }, + { + "epoch": 3.6126580619799515, + "grad_norm": 0.5428944230079651, + "learning_rate": 4.564202984288365e-05, + "loss": 0.0308, + "step": 55140 + }, + { + "epoch": 3.6133132411714604, + "grad_norm": 0.9661141037940979, + "learning_rate": 4.5625558308111356e-05, + "loss": 0.0409, + "step": 55150 + }, + { + "epoch": 3.613968420362969, + "grad_norm": 0.32180386781692505, + "learning_rate": 4.560908725172145e-05, + "loss": 0.0266, + "step": 55160 + }, + { + "epoch": 3.614623599554478, + "grad_norm": 0.7026591897010803, + "learning_rate": 4.559261667551516e-05, + "loss": 0.041, + "step": 55170 + }, + { + "epoch": 3.615278778745987, + "grad_norm": 0.10502950847148895, + "learning_rate": 4.5576146581293685e-05, + "loss": 0.0448, + "step": 55180 + }, + { + "epoch": 3.615933957937496, + "grad_norm": 0.13581550121307373, + "learning_rate": 4.5559676970858164e-05, + "loss": 0.0313, + "step": 55190 + }, + { + "epoch": 3.616589137129005, + "grad_norm": 0.4839932918548584, + "learning_rate": 4.5543207846009705e-05, + "loss": 0.0307, + "step": 55200 + }, + { + "epoch": 3.6172443163205137, + "grad_norm": 0.1388702243566513, + "learning_rate": 4.552673920854933e-05, + "loss": 0.0292, + "step": 55210 + }, + { + "epoch": 3.6178994955120225, + "grad_norm": 0.5035263895988464, + "learning_rate": 4.5510271060278006e-05, + "loss": 0.0725, + "step": 55220 + }, + { + "epoch": 3.6185546747035313, + "grad_norm": 0.6946473121643066, + "learning_rate": 4.5493803402996704e-05, + "loss": 0.0327, + "step": 55230 + }, + { + "epoch": 3.61920985389504, + "grad_norm": 0.5924214124679565, + "learning_rate": 4.5477336238506295e-05, + "loss": 0.0305, + "step": 55240 + }, + { + "epoch": 3.6198650330865494, + "grad_norm": 0.6916489601135254, + "learning_rate": 4.546086956860759e-05, + "loss": 0.0308, + "step": 55250 + }, + { + "epoch": 3.620520212278058, + "grad_norm": 0.20308035612106323, + "learning_rate": 4.5444403395101374e-05, + "loss": 0.0209, + "step": 55260 + }, + { + "epoch": 3.621175391469567, + "grad_norm": 0.701663613319397, + "learning_rate": 4.542793771978834e-05, + "loss": 0.0411, + "step": 55270 + }, + { + "epoch": 3.621830570661076, + "grad_norm": 0.12469707429409027, + "learning_rate": 4.541147254446916e-05, + "loss": 0.0275, + "step": 55280 + }, + { + "epoch": 3.6224857498525846, + "grad_norm": 0.17173463106155396, + "learning_rate": 4.539500787094445e-05, + "loss": 0.0345, + "step": 55290 + }, + { + "epoch": 3.6231409290440935, + "grad_norm": 0.17827726900577545, + "learning_rate": 4.537854370101475e-05, + "loss": 0.0263, + "step": 55300 + }, + { + "epoch": 3.6237961082356023, + "grad_norm": 0.318969190120697, + "learning_rate": 4.5362080036480566e-05, + "loss": 0.0573, + "step": 55310 + }, + { + "epoch": 3.624451287427111, + "grad_norm": 1.0854605436325073, + "learning_rate": 4.534561687914235e-05, + "loss": 0.0303, + "step": 55320 + }, + { + "epoch": 3.6251064666186203, + "grad_norm": 0.5393889546394348, + "learning_rate": 4.5329154230800466e-05, + "loss": 0.0349, + "step": 55330 + }, + { + "epoch": 3.625761645810129, + "grad_norm": 0.13046550750732422, + "learning_rate": 4.531269209325525e-05, + "loss": 0.0358, + "step": 55340 + }, + { + "epoch": 3.626416825001638, + "grad_norm": 0.20474793016910553, + "learning_rate": 4.529623046830696e-05, + "loss": 0.0246, + "step": 55350 + }, + { + "epoch": 3.627072004193147, + "grad_norm": 0.13399474322795868, + "learning_rate": 4.5279769357755866e-05, + "loss": 0.0377, + "step": 55360 + }, + { + "epoch": 3.6277271833846556, + "grad_norm": 0.16776061058044434, + "learning_rate": 4.5263308763402084e-05, + "loss": 0.027, + "step": 55370 + }, + { + "epoch": 3.6283823625761644, + "grad_norm": 0.5820937156677246, + "learning_rate": 4.524684868704574e-05, + "loss": 0.0254, + "step": 55380 + }, + { + "epoch": 3.6290375417676737, + "grad_norm": 0.07404876500368118, + "learning_rate": 4.523038913048688e-05, + "loss": 0.0415, + "step": 55390 + }, + { + "epoch": 3.6296927209591825, + "grad_norm": 0.5324252843856812, + "learning_rate": 4.521393009552549e-05, + "loss": 0.0253, + "step": 55400 + }, + { + "epoch": 3.6303479001506913, + "grad_norm": 0.4113834798336029, + "learning_rate": 4.5197471583961484e-05, + "loss": 0.0529, + "step": 55410 + }, + { + "epoch": 3.6310030793422, + "grad_norm": 1.0215591192245483, + "learning_rate": 4.51810135975948e-05, + "loss": 0.0316, + "step": 55420 + }, + { + "epoch": 3.631658258533709, + "grad_norm": 0.22385184466838837, + "learning_rate": 4.516455613822521e-05, + "loss": 0.0419, + "step": 55430 + }, + { + "epoch": 3.6323134377252178, + "grad_norm": 0.5803202390670776, + "learning_rate": 4.514809920765249e-05, + "loss": 0.0325, + "step": 55440 + }, + { + "epoch": 3.6329686169167266, + "grad_norm": 0.952775239944458, + "learning_rate": 4.5131642807676356e-05, + "loss": 0.0573, + "step": 55450 + }, + { + "epoch": 3.6336237961082354, + "grad_norm": 0.09950467199087143, + "learning_rate": 4.511518694009644e-05, + "loss": 0.0439, + "step": 55460 + }, + { + "epoch": 3.6342789752997446, + "grad_norm": 0.134229376912117, + "learning_rate": 4.509873160671231e-05, + "loss": 0.0259, + "step": 55470 + }, + { + "epoch": 3.6349341544912535, + "grad_norm": 0.1599532663822174, + "learning_rate": 4.508227680932356e-05, + "loss": 0.0138, + "step": 55480 + }, + { + "epoch": 3.6355893336827623, + "grad_norm": 0.4729720950126648, + "learning_rate": 4.5065822549729605e-05, + "loss": 0.0199, + "step": 55490 + }, + { + "epoch": 3.636244512874271, + "grad_norm": 0.1045105904340744, + "learning_rate": 4.504936882972989e-05, + "loss": 0.0301, + "step": 55500 + }, + { + "epoch": 3.63689969206578, + "grad_norm": 0.17590737342834473, + "learning_rate": 4.503291565112374e-05, + "loss": 0.0327, + "step": 55510 + }, + { + "epoch": 3.6375548712572887, + "grad_norm": 0.4784473478794098, + "learning_rate": 4.501646301571048e-05, + "loss": 0.0181, + "step": 55520 + }, + { + "epoch": 3.638210050448798, + "grad_norm": 2.6405293941497803, + "learning_rate": 4.500001092528933e-05, + "loss": 0.0297, + "step": 55530 + }, + { + "epoch": 3.638865229640307, + "grad_norm": 1.2127315998077393, + "learning_rate": 4.4983559381659455e-05, + "loss": 0.0398, + "step": 55540 + }, + { + "epoch": 3.6395204088318156, + "grad_norm": 0.2876402735710144, + "learning_rate": 4.496710838661999e-05, + "loss": 0.0327, + "step": 55550 + }, + { + "epoch": 3.6401755880233244, + "grad_norm": 0.6109623312950134, + "learning_rate": 4.495065794196999e-05, + "loss": 0.0585, + "step": 55560 + }, + { + "epoch": 3.6408307672148332, + "grad_norm": 0.2790171504020691, + "learning_rate": 4.4934208049508455e-05, + "loss": 0.0198, + "step": 55570 + }, + { + "epoch": 3.641485946406342, + "grad_norm": 0.27260318398475647, + "learning_rate": 4.4917758711034303e-05, + "loss": 0.0217, + "step": 55580 + }, + { + "epoch": 3.642141125597851, + "grad_norm": 0.5975915193557739, + "learning_rate": 4.490130992834642e-05, + "loss": 0.0299, + "step": 55590 + }, + { + "epoch": 3.6427963047893597, + "grad_norm": 0.15843932330608368, + "learning_rate": 4.488486170324359e-05, + "loss": 0.0239, + "step": 55600 + }, + { + "epoch": 3.643451483980869, + "grad_norm": 0.7922402620315552, + "learning_rate": 4.486841403752463e-05, + "loss": 0.041, + "step": 55610 + }, + { + "epoch": 3.6441066631723777, + "grad_norm": 1.117364764213562, + "learning_rate": 4.4851966932988174e-05, + "loss": 0.0401, + "step": 55620 + }, + { + "epoch": 3.6447618423638866, + "grad_norm": 0.24379657208919525, + "learning_rate": 4.4835520391432894e-05, + "loss": 0.0184, + "step": 55630 + }, + { + "epoch": 3.6454170215553954, + "grad_norm": 0.5688251256942749, + "learning_rate": 4.481907441465733e-05, + "loss": 0.0311, + "step": 55640 + }, + { + "epoch": 3.646072200746904, + "grad_norm": 0.5802160501480103, + "learning_rate": 4.480262900446e-05, + "loss": 0.0387, + "step": 55650 + }, + { + "epoch": 3.646727379938413, + "grad_norm": 0.2609640061855316, + "learning_rate": 4.478618416263933e-05, + "loss": 0.0435, + "step": 55660 + }, + { + "epoch": 3.6473825591299223, + "grad_norm": 0.19729453325271606, + "learning_rate": 4.476973989099373e-05, + "loss": 0.0341, + "step": 55670 + }, + { + "epoch": 3.648037738321431, + "grad_norm": 0.6062024235725403, + "learning_rate": 4.475329619132153e-05, + "loss": 0.033, + "step": 55680 + }, + { + "epoch": 3.64869291751294, + "grad_norm": 0.13001810014247894, + "learning_rate": 4.4736853065420956e-05, + "loss": 0.0208, + "step": 55690 + }, + { + "epoch": 3.6493480967044487, + "grad_norm": 0.35347604751586914, + "learning_rate": 4.472041051509023e-05, + "loss": 0.0259, + "step": 55700 + }, + { + "epoch": 3.6500032758959575, + "grad_norm": 0.4566890001296997, + "learning_rate": 4.470396854212746e-05, + "loss": 0.0749, + "step": 55710 + }, + { + "epoch": 3.6506584550874663, + "grad_norm": 0.7231694459915161, + "learning_rate": 4.468752714833072e-05, + "loss": 0.0302, + "step": 55720 + }, + { + "epoch": 3.651313634278975, + "grad_norm": 0.785065233707428, + "learning_rate": 4.467108633549803e-05, + "loss": 0.0293, + "step": 55730 + }, + { + "epoch": 3.651968813470484, + "grad_norm": 0.8215836882591248, + "learning_rate": 4.4654646105427335e-05, + "loss": 0.0456, + "step": 55740 + }, + { + "epoch": 3.652623992661993, + "grad_norm": 1.0972188711166382, + "learning_rate": 4.463820645991651e-05, + "loss": 0.0165, + "step": 55750 + }, + { + "epoch": 3.653279171853502, + "grad_norm": 0.23496338725090027, + "learning_rate": 4.462176740076334e-05, + "loss": 0.0445, + "step": 55760 + }, + { + "epoch": 3.653934351045011, + "grad_norm": 0.20309408009052277, + "learning_rate": 4.4605328929765624e-05, + "loss": 0.0274, + "step": 55770 + }, + { + "epoch": 3.6545895302365197, + "grad_norm": 0.34507423639297485, + "learning_rate": 4.458889104872102e-05, + "loss": 0.0234, + "step": 55780 + }, + { + "epoch": 3.6552447094280285, + "grad_norm": 0.4106156826019287, + "learning_rate": 4.457245375942715e-05, + "loss": 0.0385, + "step": 55790 + }, + { + "epoch": 3.6558998886195373, + "grad_norm": 0.1781245768070221, + "learning_rate": 4.455601706368157e-05, + "loss": 0.0372, + "step": 55800 + }, + { + "epoch": 3.6565550678110466, + "grad_norm": 0.23159146308898926, + "learning_rate": 4.453958096328179e-05, + "loss": 0.0381, + "step": 55810 + }, + { + "epoch": 3.6572102470025554, + "grad_norm": 0.2771592140197754, + "learning_rate": 4.452314546002523e-05, + "loss": 0.0207, + "step": 55820 + }, + { + "epoch": 3.657865426194064, + "grad_norm": 0.10772890597581863, + "learning_rate": 4.4506710555709246e-05, + "loss": 0.0266, + "step": 55830 + }, + { + "epoch": 3.658520605385573, + "grad_norm": 0.4687376618385315, + "learning_rate": 4.449027625213114e-05, + "loss": 0.0474, + "step": 55840 + }, + { + "epoch": 3.659175784577082, + "grad_norm": 0.917302131652832, + "learning_rate": 4.4473842551088135e-05, + "loss": 0.0347, + "step": 55850 + }, + { + "epoch": 3.6598309637685906, + "grad_norm": 0.470448762178421, + "learning_rate": 4.445740945437738e-05, + "loss": 0.0675, + "step": 55860 + }, + { + "epoch": 3.6604861429600994, + "grad_norm": 0.09711475670337677, + "learning_rate": 4.444097696379602e-05, + "loss": 0.0718, + "step": 55870 + }, + { + "epoch": 3.6611413221516083, + "grad_norm": 0.17200897634029388, + "learning_rate": 4.442454508114106e-05, + "loss": 0.0364, + "step": 55880 + }, + { + "epoch": 3.661796501343117, + "grad_norm": 0.2568037807941437, + "learning_rate": 4.4408113808209465e-05, + "loss": 0.0361, + "step": 55890 + }, + { + "epoch": 3.6624516805346263, + "grad_norm": 0.45164525508880615, + "learning_rate": 4.439168314679813e-05, + "loss": 0.0469, + "step": 55900 + }, + { + "epoch": 3.663106859726135, + "grad_norm": 0.8115699887275696, + "learning_rate": 4.43752530987039e-05, + "loss": 0.0455, + "step": 55910 + }, + { + "epoch": 3.663762038917644, + "grad_norm": 0.2340329885482788, + "learning_rate": 4.4358823665723515e-05, + "loss": 0.0507, + "step": 55920 + }, + { + "epoch": 3.6644172181091528, + "grad_norm": 0.27748867869377136, + "learning_rate": 4.434239484965371e-05, + "loss": 0.0633, + "step": 55930 + }, + { + "epoch": 3.6650723973006616, + "grad_norm": 0.5983583927154541, + "learning_rate": 4.4325966652291103e-05, + "loss": 0.0331, + "step": 55940 + }, + { + "epoch": 3.665727576492171, + "grad_norm": 0.714448094367981, + "learning_rate": 4.430953907543225e-05, + "loss": 0.0237, + "step": 55950 + }, + { + "epoch": 3.6663827556836797, + "grad_norm": 0.2850823402404785, + "learning_rate": 4.4293112120873645e-05, + "loss": 0.0206, + "step": 55960 + }, + { + "epoch": 3.6670379348751885, + "grad_norm": 0.6746061444282532, + "learning_rate": 4.427668579041172e-05, + "loss": 0.0283, + "step": 55970 + }, + { + "epoch": 3.6676931140666973, + "grad_norm": 0.26829472184181213, + "learning_rate": 4.4260260085842816e-05, + "loss": 0.0401, + "step": 55980 + }, + { + "epoch": 3.668348293258206, + "grad_norm": 0.46424803137779236, + "learning_rate": 4.4243835008963256e-05, + "loss": 0.0258, + "step": 55990 + }, + { + "epoch": 3.669003472449715, + "grad_norm": 0.3400178551673889, + "learning_rate": 4.422741056156925e-05, + "loss": 0.0302, + "step": 56000 + }, + { + "epoch": 3.6696586516412237, + "grad_norm": 0.72928386926651, + "learning_rate": 4.421098674545693e-05, + "loss": 0.0394, + "step": 56010 + }, + { + "epoch": 3.6703138308327325, + "grad_norm": 0.9702059626579285, + "learning_rate": 4.419456356242241e-05, + "loss": 0.0265, + "step": 56020 + }, + { + "epoch": 3.6709690100242414, + "grad_norm": 0.7102705240249634, + "learning_rate": 4.4178141014261685e-05, + "loss": 0.0222, + "step": 56030 + }, + { + "epoch": 3.6716241892157506, + "grad_norm": 0.2434845119714737, + "learning_rate": 4.416171910277071e-05, + "loss": 0.0476, + "step": 56040 + }, + { + "epoch": 3.6722793684072594, + "grad_norm": 0.17488647997379303, + "learning_rate": 4.414529782974533e-05, + "loss": 0.0293, + "step": 56050 + }, + { + "epoch": 3.6729345475987683, + "grad_norm": 1.091797947883606, + "learning_rate": 4.41288771969814e-05, + "loss": 0.0606, + "step": 56060 + }, + { + "epoch": 3.673589726790277, + "grad_norm": 0.4634553790092468, + "learning_rate": 4.411245720627464e-05, + "loss": 0.0493, + "step": 56070 + }, + { + "epoch": 3.674244905981786, + "grad_norm": 0.28600266575813293, + "learning_rate": 4.40960378594207e-05, + "loss": 0.0538, + "step": 56080 + }, + { + "epoch": 3.674900085173295, + "grad_norm": 0.4458741843700409, + "learning_rate": 4.407961915821519e-05, + "loss": 0.0353, + "step": 56090 + }, + { + "epoch": 3.675555264364804, + "grad_norm": 0.4646574556827545, + "learning_rate": 4.406320110445362e-05, + "loss": 0.0313, + "step": 56100 + }, + { + "epoch": 3.6762104435563128, + "grad_norm": 0.31035614013671875, + "learning_rate": 4.404678369993144e-05, + "loss": 0.0279, + "step": 56110 + }, + { + "epoch": 3.6768656227478216, + "grad_norm": 0.4638972580432892, + "learning_rate": 4.403036694644406e-05, + "loss": 0.0522, + "step": 56120 + }, + { + "epoch": 3.6775208019393304, + "grad_norm": 0.6201044917106628, + "learning_rate": 4.4013950845786764e-05, + "loss": 0.0314, + "step": 56130 + }, + { + "epoch": 3.678175981130839, + "grad_norm": 0.19976720213890076, + "learning_rate": 4.399753539975482e-05, + "loss": 0.0227, + "step": 56140 + }, + { + "epoch": 3.678831160322348, + "grad_norm": 0.4260309636592865, + "learning_rate": 4.398112061014337e-05, + "loss": 0.044, + "step": 56150 + }, + { + "epoch": 3.679486339513857, + "grad_norm": 0.20755918323993683, + "learning_rate": 4.396470647874753e-05, + "loss": 0.0277, + "step": 56160 + }, + { + "epoch": 3.6801415187053657, + "grad_norm": 0.975095272064209, + "learning_rate": 4.394829300736229e-05, + "loss": 0.0243, + "step": 56170 + }, + { + "epoch": 3.680796697896875, + "grad_norm": 0.3775423467159271, + "learning_rate": 4.393188019778265e-05, + "loss": 0.0339, + "step": 56180 + }, + { + "epoch": 3.6814518770883837, + "grad_norm": 0.6307576298713684, + "learning_rate": 4.391546805180347e-05, + "loss": 0.0518, + "step": 56190 + }, + { + "epoch": 3.6821070562798925, + "grad_norm": 0.2621191740036011, + "learning_rate": 4.389905657121955e-05, + "loss": 0.026, + "step": 56200 + }, + { + "epoch": 3.6827622354714014, + "grad_norm": 0.4801652431488037, + "learning_rate": 4.3882645757825635e-05, + "loss": 0.0256, + "step": 56210 + }, + { + "epoch": 3.68341741466291, + "grad_norm": 0.6482788324356079, + "learning_rate": 4.386623561341637e-05, + "loss": 0.0393, + "step": 56220 + }, + { + "epoch": 3.6840725938544194, + "grad_norm": 0.7133165001869202, + "learning_rate": 4.384982613978637e-05, + "loss": 0.0435, + "step": 56230 + }, + { + "epoch": 3.6847277730459282, + "grad_norm": 0.43648743629455566, + "learning_rate": 4.383341733873012e-05, + "loss": 0.0252, + "step": 56240 + }, + { + "epoch": 3.685382952237437, + "grad_norm": 1.0703321695327759, + "learning_rate": 4.381700921204209e-05, + "loss": 0.0397, + "step": 56250 + }, + { + "epoch": 3.686038131428946, + "grad_norm": 0.4967791736125946, + "learning_rate": 4.380060176151663e-05, + "loss": 0.0485, + "step": 56260 + }, + { + "epoch": 3.6866933106204547, + "grad_norm": 0.2870851755142212, + "learning_rate": 4.378419498894805e-05, + "loss": 0.0197, + "step": 56270 + }, + { + "epoch": 3.6873484898119635, + "grad_norm": 0.18434357643127441, + "learning_rate": 4.376778889613056e-05, + "loss": 0.0407, + "step": 56280 + }, + { + "epoch": 3.6880036690034723, + "grad_norm": 0.6579649448394775, + "learning_rate": 4.375138348485831e-05, + "loss": 0.0483, + "step": 56290 + }, + { + "epoch": 3.688658848194981, + "grad_norm": 0.12695224583148956, + "learning_rate": 4.3734978756925346e-05, + "loss": 0.0261, + "step": 56300 + }, + { + "epoch": 3.68931402738649, + "grad_norm": 0.4154002070426941, + "learning_rate": 4.3718574714125706e-05, + "loss": 0.0129, + "step": 56310 + }, + { + "epoch": 3.689969206577999, + "grad_norm": 0.1522371768951416, + "learning_rate": 4.370217135825329e-05, + "loss": 0.0302, + "step": 56320 + }, + { + "epoch": 3.690624385769508, + "grad_norm": 0.37436798214912415, + "learning_rate": 4.368576869110194e-05, + "loss": 0.0515, + "step": 56330 + }, + { + "epoch": 3.691279564961017, + "grad_norm": 0.2975999414920807, + "learning_rate": 4.366936671446544e-05, + "loss": 0.0254, + "step": 56340 + }, + { + "epoch": 3.6919347441525256, + "grad_norm": 0.10063428431749344, + "learning_rate": 4.3652965430137474e-05, + "loss": 0.022, + "step": 56350 + }, + { + "epoch": 3.6925899233440345, + "grad_norm": 0.3307547867298126, + "learning_rate": 4.3636564839911646e-05, + "loss": 0.0235, + "step": 56360 + }, + { + "epoch": 3.6932451025355437, + "grad_norm": 0.4942915439605713, + "learning_rate": 4.3620164945581545e-05, + "loss": 0.0276, + "step": 56370 + }, + { + "epoch": 3.6939002817270525, + "grad_norm": 0.4003697335720062, + "learning_rate": 4.360376574894061e-05, + "loss": 0.0486, + "step": 56380 + }, + { + "epoch": 3.6945554609185614, + "grad_norm": 0.21907545626163483, + "learning_rate": 4.358736725178224e-05, + "loss": 0.0403, + "step": 56390 + }, + { + "epoch": 3.69521064011007, + "grad_norm": 0.3794642686843872, + "learning_rate": 4.357096945589974e-05, + "loss": 0.0307, + "step": 56400 + }, + { + "epoch": 3.695865819301579, + "grad_norm": 0.3513697385787964, + "learning_rate": 4.355457236308636e-05, + "loss": 0.0196, + "step": 56410 + }, + { + "epoch": 3.696520998493088, + "grad_norm": 0.5424560308456421, + "learning_rate": 4.353817597513526e-05, + "loss": 0.0446, + "step": 56420 + }, + { + "epoch": 3.6971761776845966, + "grad_norm": 0.37166574597358704, + "learning_rate": 4.352178029383948e-05, + "loss": 0.0431, + "step": 56430 + }, + { + "epoch": 3.6978313568761054, + "grad_norm": 0.8091244697570801, + "learning_rate": 4.3505385320992105e-05, + "loss": 0.0629, + "step": 56440 + }, + { + "epoch": 3.6984865360676142, + "grad_norm": 0.2358521670103073, + "learning_rate": 4.348899105838602e-05, + "loss": 0.0458, + "step": 56450 + }, + { + "epoch": 3.6991417152591235, + "grad_norm": 0.4304322302341461, + "learning_rate": 4.3472597507814087e-05, + "loss": 0.0211, + "step": 56460 + }, + { + "epoch": 3.6997968944506323, + "grad_norm": 0.761355996131897, + "learning_rate": 4.3456204671069066e-05, + "loss": 0.0404, + "step": 56470 + }, + { + "epoch": 3.700452073642141, + "grad_norm": 0.45658913254737854, + "learning_rate": 4.343981254994367e-05, + "loss": 0.034, + "step": 56480 + }, + { + "epoch": 3.70110725283365, + "grad_norm": 0.47931063175201416, + "learning_rate": 4.34234211462305e-05, + "loss": 0.0219, + "step": 56490 + }, + { + "epoch": 3.7017624320251588, + "grad_norm": 0.16929680109024048, + "learning_rate": 4.340703046172213e-05, + "loss": 0.0332, + "step": 56500 + }, + { + "epoch": 3.702417611216668, + "grad_norm": 2.5885422229766846, + "learning_rate": 4.339064049821097e-05, + "loss": 0.0757, + "step": 56510 + }, + { + "epoch": 3.703072790408177, + "grad_norm": 0.2604769170284271, + "learning_rate": 4.3374251257489446e-05, + "loss": 0.0454, + "step": 56520 + }, + { + "epoch": 3.7037279695996856, + "grad_norm": 0.8073841333389282, + "learning_rate": 4.335786274134984e-05, + "loss": 0.0396, + "step": 56530 + }, + { + "epoch": 3.7043831487911945, + "grad_norm": 0.364461213350296, + "learning_rate": 4.334147495158438e-05, + "loss": 0.0441, + "step": 56540 + }, + { + "epoch": 3.7050383279827033, + "grad_norm": 0.6410499215126038, + "learning_rate": 4.33250878899852e-05, + "loss": 0.0448, + "step": 56550 + }, + { + "epoch": 3.705693507174212, + "grad_norm": 1.235711693763733, + "learning_rate": 4.330870155834437e-05, + "loss": 0.0293, + "step": 56560 + }, + { + "epoch": 3.706348686365721, + "grad_norm": 0.48108482360839844, + "learning_rate": 4.3292315958453884e-05, + "loss": 0.0331, + "step": 56570 + }, + { + "epoch": 3.7070038655572297, + "grad_norm": 0.18080934882164001, + "learning_rate": 4.3275931092105646e-05, + "loss": 0.025, + "step": 56580 + }, + { + "epoch": 3.7076590447487385, + "grad_norm": 0.5149590969085693, + "learning_rate": 4.325954696109147e-05, + "loss": 0.0434, + "step": 56590 + }, + { + "epoch": 3.708314223940248, + "grad_norm": 0.15049214661121368, + "learning_rate": 4.32431635672031e-05, + "loss": 0.0247, + "step": 56600 + }, + { + "epoch": 3.7089694031317566, + "grad_norm": 0.329201340675354, + "learning_rate": 4.3226780912232214e-05, + "loss": 0.0271, + "step": 56610 + }, + { + "epoch": 3.7096245823232654, + "grad_norm": 0.7613171339035034, + "learning_rate": 4.321039899797036e-05, + "loss": 0.0624, + "step": 56620 + }, + { + "epoch": 3.7102797615147742, + "grad_norm": 0.4738011062145233, + "learning_rate": 4.319401782620908e-05, + "loss": 0.0416, + "step": 56630 + }, + { + "epoch": 3.710934940706283, + "grad_norm": 0.4190782904624939, + "learning_rate": 4.317763739873978e-05, + "loss": 0.0401, + "step": 56640 + }, + { + "epoch": 3.7115901198977923, + "grad_norm": 0.6307849884033203, + "learning_rate": 4.31612577173538e-05, + "loss": 0.0453, + "step": 56650 + }, + { + "epoch": 3.712245299089301, + "grad_norm": 0.31511035561561584, + "learning_rate": 4.314487878384239e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 3.71290047828081, + "grad_norm": 0.467533141374588, + "learning_rate": 4.3128500599996736e-05, + "loss": 0.032, + "step": 56670 + }, + { + "epoch": 3.7135556574723187, + "grad_norm": 0.6367124915122986, + "learning_rate": 4.31121231676079e-05, + "loss": 0.0409, + "step": 56680 + }, + { + "epoch": 3.7142108366638276, + "grad_norm": 0.3923953175544739, + "learning_rate": 4.309574648846694e-05, + "loss": 0.0221, + "step": 56690 + }, + { + "epoch": 3.7148660158553364, + "grad_norm": 0.35882940888404846, + "learning_rate": 4.3079370564364755e-05, + "loss": 0.0249, + "step": 56700 + }, + { + "epoch": 3.715521195046845, + "grad_norm": 0.3155593276023865, + "learning_rate": 4.306299539709221e-05, + "loss": 0.0295, + "step": 56710 + }, + { + "epoch": 3.716176374238354, + "grad_norm": 0.22677601873874664, + "learning_rate": 4.3046620988440046e-05, + "loss": 0.0439, + "step": 56720 + }, + { + "epoch": 3.716831553429863, + "grad_norm": 0.6753522157669067, + "learning_rate": 4.303024734019897e-05, + "loss": 0.0305, + "step": 56730 + }, + { + "epoch": 3.717486732621372, + "grad_norm": 0.5652795433998108, + "learning_rate": 4.3013874454159565e-05, + "loss": 0.0431, + "step": 56740 + }, + { + "epoch": 3.718141911812881, + "grad_norm": 0.5361281037330627, + "learning_rate": 4.299750233211233e-05, + "loss": 0.0317, + "step": 56750 + }, + { + "epoch": 3.7187970910043897, + "grad_norm": 0.43683281540870667, + "learning_rate": 4.2981130975847715e-05, + "loss": 0.0288, + "step": 56760 + }, + { + "epoch": 3.7194522701958985, + "grad_norm": 0.23269696533679962, + "learning_rate": 4.2964760387156076e-05, + "loss": 0.0568, + "step": 56770 + }, + { + "epoch": 3.7201074493874073, + "grad_norm": 0.7163128852844238, + "learning_rate": 4.294839056782767e-05, + "loss": 0.034, + "step": 56780 + }, + { + "epoch": 3.7207626285789166, + "grad_norm": 0.7679603695869446, + "learning_rate": 4.293202151965266e-05, + "loss": 0.0427, + "step": 56790 + }, + { + "epoch": 3.7214178077704254, + "grad_norm": 0.9213759303092957, + "learning_rate": 4.291565324442115e-05, + "loss": 0.0532, + "step": 56800 + }, + { + "epoch": 3.7220729869619342, + "grad_norm": 0.8401661515235901, + "learning_rate": 4.2899285743923135e-05, + "loss": 0.04, + "step": 56810 + }, + { + "epoch": 3.722728166153443, + "grad_norm": 0.5595073699951172, + "learning_rate": 4.288291901994857e-05, + "loss": 0.0141, + "step": 56820 + }, + { + "epoch": 3.723383345344952, + "grad_norm": 0.4109090566635132, + "learning_rate": 4.286655307428729e-05, + "loss": 0.0379, + "step": 56830 + }, + { + "epoch": 3.7240385245364607, + "grad_norm": 0.3678019940853119, + "learning_rate": 4.285018790872903e-05, + "loss": 0.0219, + "step": 56840 + }, + { + "epoch": 3.7246937037279695, + "grad_norm": 0.2273455560207367, + "learning_rate": 4.283382352506347e-05, + "loss": 0.0367, + "step": 56850 + }, + { + "epoch": 3.7253488829194783, + "grad_norm": 0.1556978076696396, + "learning_rate": 4.2817459925080205e-05, + "loss": 0.0473, + "step": 56860 + }, + { + "epoch": 3.726004062110987, + "grad_norm": 0.9816267490386963, + "learning_rate": 4.28010971105687e-05, + "loss": 0.0249, + "step": 56870 + }, + { + "epoch": 3.7266592413024964, + "grad_norm": 0.6282848119735718, + "learning_rate": 4.278473508331841e-05, + "loss": 0.0235, + "step": 56880 + }, + { + "epoch": 3.727314420494005, + "grad_norm": 1.3715934753417969, + "learning_rate": 4.276837384511864e-05, + "loss": 0.0274, + "step": 56890 + }, + { + "epoch": 3.727969599685514, + "grad_norm": 0.4563915431499481, + "learning_rate": 4.275201339775864e-05, + "loss": 0.0487, + "step": 56900 + }, + { + "epoch": 3.728624778877023, + "grad_norm": 0.20218069851398468, + "learning_rate": 4.273565374302756e-05, + "loss": 0.0351, + "step": 56910 + }, + { + "epoch": 3.7292799580685316, + "grad_norm": 0.6136788129806519, + "learning_rate": 4.2719294882714454e-05, + "loss": 0.029, + "step": 56920 + }, + { + "epoch": 3.729935137260041, + "grad_norm": 0.47628554701805115, + "learning_rate": 4.270293681860831e-05, + "loss": 0.0219, + "step": 56930 + }, + { + "epoch": 3.7305903164515497, + "grad_norm": 0.4188166856765747, + "learning_rate": 4.268657955249802e-05, + "loss": 0.0366, + "step": 56940 + }, + { + "epoch": 3.7312454956430585, + "grad_norm": 0.19943565130233765, + "learning_rate": 4.267022308617241e-05, + "loss": 0.0344, + "step": 56950 + }, + { + "epoch": 3.7319006748345673, + "grad_norm": 0.7100774049758911, + "learning_rate": 4.265386742142018e-05, + "loss": 0.0526, + "step": 56960 + }, + { + "epoch": 3.732555854026076, + "grad_norm": 0.5283260345458984, + "learning_rate": 4.263751256002996e-05, + "loss": 0.0289, + "step": 56970 + }, + { + "epoch": 3.733211033217585, + "grad_norm": 0.2502785623073578, + "learning_rate": 4.26211585037903e-05, + "loss": 0.0692, + "step": 56980 + }, + { + "epoch": 3.7338662124090938, + "grad_norm": 0.13246606290340424, + "learning_rate": 4.2604805254489656e-05, + "loss": 0.0326, + "step": 56990 + }, + { + "epoch": 3.7345213916006026, + "grad_norm": 0.2039816677570343, + "learning_rate": 4.258845281391638e-05, + "loss": 0.03, + "step": 57000 + }, + { + "epoch": 3.7351765707921114, + "grad_norm": 1.998988389968872, + "learning_rate": 4.257210118385876e-05, + "loss": 0.0364, + "step": 57010 + }, + { + "epoch": 3.7358317499836207, + "grad_norm": 0.34245792031288147, + "learning_rate": 4.2555750366104994e-05, + "loss": 0.0282, + "step": 57020 + }, + { + "epoch": 3.7364869291751295, + "grad_norm": 0.5465847849845886, + "learning_rate": 4.2539400362443184e-05, + "loss": 0.0354, + "step": 57030 + }, + { + "epoch": 3.7371421083666383, + "grad_norm": 0.6574207544326782, + "learning_rate": 4.252305117466133e-05, + "loss": 0.021, + "step": 57040 + }, + { + "epoch": 3.737797287558147, + "grad_norm": 0.3160340487957001, + "learning_rate": 4.2506702804547356e-05, + "loss": 0.0458, + "step": 57050 + }, + { + "epoch": 3.738452466749656, + "grad_norm": 0.39827626943588257, + "learning_rate": 4.249035525388907e-05, + "loss": 0.0402, + "step": 57060 + }, + { + "epoch": 3.739107645941165, + "grad_norm": 0.3946293890476227, + "learning_rate": 4.247400852447427e-05, + "loss": 0.0381, + "step": 57070 + }, + { + "epoch": 3.739762825132674, + "grad_norm": 0.16735677421092987, + "learning_rate": 4.245766261809059e-05, + "loss": 0.0455, + "step": 57080 + }, + { + "epoch": 3.740418004324183, + "grad_norm": 0.31676045060157776, + "learning_rate": 4.244131753652558e-05, + "loss": 0.0316, + "step": 57090 + }, + { + "epoch": 3.7410731835156916, + "grad_norm": 0.6155538558959961, + "learning_rate": 4.242497328156672e-05, + "loss": 0.0329, + "step": 57100 + }, + { + "epoch": 3.7417283627072004, + "grad_norm": 0.14427347481250763, + "learning_rate": 4.240862985500139e-05, + "loss": 0.0322, + "step": 57110 + }, + { + "epoch": 3.7423835418987093, + "grad_norm": 0.46137484908103943, + "learning_rate": 4.2392287258616894e-05, + "loss": 0.055, + "step": 57120 + }, + { + "epoch": 3.743038721090218, + "grad_norm": 0.2836351990699768, + "learning_rate": 4.2375945494200396e-05, + "loss": 0.0217, + "step": 57130 + }, + { + "epoch": 3.743693900281727, + "grad_norm": 0.12356244027614594, + "learning_rate": 4.2359604563539054e-05, + "loss": 0.0269, + "step": 57140 + }, + { + "epoch": 3.7443490794732357, + "grad_norm": 0.4497629702091217, + "learning_rate": 4.2343264468419885e-05, + "loss": 0.0214, + "step": 57150 + }, + { + "epoch": 3.745004258664745, + "grad_norm": 0.2270260900259018, + "learning_rate": 4.232692521062979e-05, + "loss": 0.0408, + "step": 57160 + }, + { + "epoch": 3.7456594378562538, + "grad_norm": 0.20610705018043518, + "learning_rate": 4.231058679195562e-05, + "loss": 0.0213, + "step": 57170 + }, + { + "epoch": 3.7463146170477626, + "grad_norm": 0.18773896992206573, + "learning_rate": 4.229424921418411e-05, + "loss": 0.0343, + "step": 57180 + }, + { + "epoch": 3.7469697962392714, + "grad_norm": 0.11559700220823288, + "learning_rate": 4.2277912479101916e-05, + "loss": 0.0414, + "step": 57190 + }, + { + "epoch": 3.74762497543078, + "grad_norm": 3.1912450790405273, + "learning_rate": 4.226157658849562e-05, + "loss": 0.0445, + "step": 57200 + }, + { + "epoch": 3.7482801546222895, + "grad_norm": 0.28923341631889343, + "learning_rate": 4.2245241544151664e-05, + "loss": 0.0238, + "step": 57210 + }, + { + "epoch": 3.7489353338137983, + "grad_norm": 0.5654999017715454, + "learning_rate": 4.2228907347856425e-05, + "loss": 0.029, + "step": 57220 + }, + { + "epoch": 3.749590513005307, + "grad_norm": 0.6823712587356567, + "learning_rate": 4.2212574001396214e-05, + "loss": 0.0344, + "step": 57230 + }, + { + "epoch": 3.750245692196816, + "grad_norm": 0.3917354941368103, + "learning_rate": 4.21962415065572e-05, + "loss": 0.034, + "step": 57240 + }, + { + "epoch": 3.7509008713883247, + "grad_norm": 1.0566189289093018, + "learning_rate": 4.217990986512548e-05, + "loss": 0.0446, + "step": 57250 + }, + { + "epoch": 3.7515560505798335, + "grad_norm": 0.19116799533367157, + "learning_rate": 4.216357907888704e-05, + "loss": 0.0924, + "step": 57260 + }, + { + "epoch": 3.7522112297713424, + "grad_norm": 0.18659023940563202, + "learning_rate": 4.2147249149627824e-05, + "loss": 0.0322, + "step": 57270 + }, + { + "epoch": 3.752866408962851, + "grad_norm": 1.1273707151412964, + "learning_rate": 4.213092007913364e-05, + "loss": 0.0283, + "step": 57280 + }, + { + "epoch": 3.75352158815436, + "grad_norm": 0.4214595556259155, + "learning_rate": 4.2114591869190205e-05, + "loss": 0.0277, + "step": 57290 + }, + { + "epoch": 3.7541767673458692, + "grad_norm": 0.2771526277065277, + "learning_rate": 4.209826452158315e-05, + "loss": 0.0262, + "step": 57300 + }, + { + "epoch": 3.754831946537378, + "grad_norm": 0.46148359775543213, + "learning_rate": 4.2081938038098e-05, + "loss": 0.0137, + "step": 57310 + }, + { + "epoch": 3.755487125728887, + "grad_norm": 0.8454702496528625, + "learning_rate": 4.2065612420520185e-05, + "loss": 0.032, + "step": 57320 + }, + { + "epoch": 3.7561423049203957, + "grad_norm": 0.1750580072402954, + "learning_rate": 4.204928767063509e-05, + "loss": 0.047, + "step": 57330 + }, + { + "epoch": 3.7567974841119045, + "grad_norm": 0.19962632656097412, + "learning_rate": 4.2032963790227944e-05, + "loss": 0.0206, + "step": 57340 + }, + { + "epoch": 3.7574526633034133, + "grad_norm": 0.7295565009117126, + "learning_rate": 4.2016640781083887e-05, + "loss": 0.0404, + "step": 57350 + }, + { + "epoch": 3.7581078424949226, + "grad_norm": 0.08270327001810074, + "learning_rate": 4.2000318644988004e-05, + "loss": 0.0426, + "step": 57360 + }, + { + "epoch": 3.7587630216864314, + "grad_norm": 0.19299831986427307, + "learning_rate": 4.198399738372524e-05, + "loss": 0.0462, + "step": 57370 + }, + { + "epoch": 3.75941820087794, + "grad_norm": 1.5853363275527954, + "learning_rate": 4.196767699908045e-05, + "loss": 0.0335, + "step": 57380 + }, + { + "epoch": 3.760073380069449, + "grad_norm": 0.22753016650676727, + "learning_rate": 4.195135749283845e-05, + "loss": 0.0236, + "step": 57390 + }, + { + "epoch": 3.760728559260958, + "grad_norm": 0.9150092601776123, + "learning_rate": 4.1935038866783894e-05, + "loss": 0.0578, + "step": 57400 + }, + { + "epoch": 3.7613837384524667, + "grad_norm": 0.3580266535282135, + "learning_rate": 4.1918721122701355e-05, + "loss": 0.0216, + "step": 57410 + }, + { + "epoch": 3.7620389176439755, + "grad_norm": 0.4402984380722046, + "learning_rate": 4.190240426237533e-05, + "loss": 0.0341, + "step": 57420 + }, + { + "epoch": 3.7626940968354843, + "grad_norm": 0.8766916394233704, + "learning_rate": 4.188608828759019e-05, + "loss": 0.0368, + "step": 57430 + }, + { + "epoch": 3.7633492760269935, + "grad_norm": 0.25447946786880493, + "learning_rate": 4.1869773200130246e-05, + "loss": 0.0346, + "step": 57440 + }, + { + "epoch": 3.7640044552185024, + "grad_norm": 0.858099102973938, + "learning_rate": 4.185345900177967e-05, + "loss": 0.0544, + "step": 57450 + }, + { + "epoch": 3.764659634410011, + "grad_norm": 0.987051784992218, + "learning_rate": 4.183714569432258e-05, + "loss": 0.0196, + "step": 57460 + }, + { + "epoch": 3.76531481360152, + "grad_norm": 0.32689666748046875, + "learning_rate": 4.182083327954295e-05, + "loss": 0.0305, + "step": 57470 + }, + { + "epoch": 3.765969992793029, + "grad_norm": 0.943726658821106, + "learning_rate": 4.1804521759224714e-05, + "loss": 0.0472, + "step": 57480 + }, + { + "epoch": 3.7666251719845376, + "grad_norm": 0.5362997651100159, + "learning_rate": 4.178821113515165e-05, + "loss": 0.036, + "step": 57490 + }, + { + "epoch": 3.767280351176047, + "grad_norm": 0.4533127248287201, + "learning_rate": 4.1771901409107476e-05, + "loss": 0.0455, + "step": 57500 + }, + { + "epoch": 3.7679355303675557, + "grad_norm": 0.15930987894535065, + "learning_rate": 4.175559258287577e-05, + "loss": 0.0357, + "step": 57510 + }, + { + "epoch": 3.7685907095590645, + "grad_norm": 0.39270055294036865, + "learning_rate": 4.173928465824008e-05, + "loss": 0.0269, + "step": 57520 + }, + { + "epoch": 3.7692458887505733, + "grad_norm": 1.3215254545211792, + "learning_rate": 4.17229776369838e-05, + "loss": 0.0237, + "step": 57530 + }, + { + "epoch": 3.769901067942082, + "grad_norm": 0.5651760101318359, + "learning_rate": 4.170667152089025e-05, + "loss": 0.0413, + "step": 57540 + }, + { + "epoch": 3.770556247133591, + "grad_norm": 0.7599997520446777, + "learning_rate": 4.1690366311742626e-05, + "loss": 0.0452, + "step": 57550 + }, + { + "epoch": 3.7712114263250998, + "grad_norm": 0.6954110264778137, + "learning_rate": 4.1674062011324055e-05, + "loss": 0.0382, + "step": 57560 + }, + { + "epoch": 3.7718666055166086, + "grad_norm": 0.367190957069397, + "learning_rate": 4.165775862141752e-05, + "loss": 0.039, + "step": 57570 + }, + { + "epoch": 3.772521784708118, + "grad_norm": 0.13207495212554932, + "learning_rate": 4.164145614380598e-05, + "loss": 0.036, + "step": 57580 + }, + { + "epoch": 3.7731769638996266, + "grad_norm": 0.4407421052455902, + "learning_rate": 4.162515458027223e-05, + "loss": 0.0233, + "step": 57590 + }, + { + "epoch": 3.7738321430911355, + "grad_norm": 0.3301003873348236, + "learning_rate": 4.160885393259899e-05, + "loss": 0.0265, + "step": 57600 + }, + { + "epoch": 3.7744873222826443, + "grad_norm": 0.34011343121528625, + "learning_rate": 4.159255420256886e-05, + "loss": 0.0198, + "step": 57610 + }, + { + "epoch": 3.775142501474153, + "grad_norm": 0.253349632024765, + "learning_rate": 4.157625539196436e-05, + "loss": 0.029, + "step": 57620 + }, + { + "epoch": 3.775797680665662, + "grad_norm": 2.513761281967163, + "learning_rate": 4.155995750256791e-05, + "loss": 0.0525, + "step": 57630 + }, + { + "epoch": 3.776452859857171, + "grad_norm": 0.18436649441719055, + "learning_rate": 4.15436605361618e-05, + "loss": 0.0232, + "step": 57640 + }, + { + "epoch": 3.77710803904868, + "grad_norm": 0.07122519612312317, + "learning_rate": 4.152736449452827e-05, + "loss": 0.0443, + "step": 57650 + }, + { + "epoch": 3.777763218240189, + "grad_norm": 0.15284165740013123, + "learning_rate": 4.151106937944943e-05, + "loss": 0.0221, + "step": 57660 + }, + { + "epoch": 3.7784183974316976, + "grad_norm": 1.2220690250396729, + "learning_rate": 4.149477519270727e-05, + "loss": 0.0455, + "step": 57670 + }, + { + "epoch": 3.7790735766232064, + "grad_norm": 0.5864097476005554, + "learning_rate": 4.147848193608371e-05, + "loss": 0.0372, + "step": 57680 + }, + { + "epoch": 3.7797287558147152, + "grad_norm": 0.27653011679649353, + "learning_rate": 4.146218961136056e-05, + "loss": 0.024, + "step": 57690 + }, + { + "epoch": 3.780383935006224, + "grad_norm": 0.522344708442688, + "learning_rate": 4.14458982203195e-05, + "loss": 0.0378, + "step": 57700 + }, + { + "epoch": 3.781039114197733, + "grad_norm": 0.9272484183311462, + "learning_rate": 4.1429607764742165e-05, + "loss": 0.0266, + "step": 57710 + }, + { + "epoch": 3.781694293389242, + "grad_norm": 0.7162744998931885, + "learning_rate": 4.141331824641003e-05, + "loss": 0.036, + "step": 57720 + }, + { + "epoch": 3.782349472580751, + "grad_norm": 0.4177119731903076, + "learning_rate": 4.139702966710452e-05, + "loss": 0.0354, + "step": 57730 + }, + { + "epoch": 3.7830046517722598, + "grad_norm": 0.09114339202642441, + "learning_rate": 4.13807420286069e-05, + "loss": 0.0315, + "step": 57740 + }, + { + "epoch": 3.7836598309637686, + "grad_norm": 0.33126509189605713, + "learning_rate": 4.136445533269837e-05, + "loss": 0.024, + "step": 57750 + }, + { + "epoch": 3.7843150101552774, + "grad_norm": 0.48294374346733093, + "learning_rate": 4.134816958116003e-05, + "loss": 0.0297, + "step": 57760 + }, + { + "epoch": 3.784970189346786, + "grad_norm": 2.2416388988494873, + "learning_rate": 4.133188477577283e-05, + "loss": 0.0517, + "step": 57770 + }, + { + "epoch": 3.7856253685382955, + "grad_norm": 0.28659239411354065, + "learning_rate": 4.131560091831771e-05, + "loss": 0.0372, + "step": 57780 + }, + { + "epoch": 3.7862805477298043, + "grad_norm": 0.5629972815513611, + "learning_rate": 4.1299318010575405e-05, + "loss": 0.0316, + "step": 57790 + }, + { + "epoch": 3.786935726921313, + "grad_norm": 0.5204310417175293, + "learning_rate": 4.128303605432661e-05, + "loss": 0.0467, + "step": 57800 + }, + { + "epoch": 3.787590906112822, + "grad_norm": 0.18522705137729645, + "learning_rate": 4.126675505135188e-05, + "loss": 0.0351, + "step": 57810 + }, + { + "epoch": 3.7882460853043307, + "grad_norm": 0.2120252549648285, + "learning_rate": 4.125047500343169e-05, + "loss": 0.0307, + "step": 57820 + }, + { + "epoch": 3.7889012644958395, + "grad_norm": 0.27771902084350586, + "learning_rate": 4.1234195912346376e-05, + "loss": 0.0218, + "step": 57830 + }, + { + "epoch": 3.7895564436873483, + "grad_norm": 0.35490721464157104, + "learning_rate": 4.1217917779876235e-05, + "loss": 0.0573, + "step": 57840 + }, + { + "epoch": 3.790211622878857, + "grad_norm": 0.15895713865756989, + "learning_rate": 4.120164060780141e-05, + "loss": 0.0368, + "step": 57850 + }, + { + "epoch": 3.790866802070366, + "grad_norm": 0.14173898100852966, + "learning_rate": 4.1185364397901935e-05, + "loss": 0.0366, + "step": 57860 + }, + { + "epoch": 3.7915219812618752, + "grad_norm": 0.7365230321884155, + "learning_rate": 4.116908915195776e-05, + "loss": 0.0676, + "step": 57870 + }, + { + "epoch": 3.792177160453384, + "grad_norm": 0.6688982248306274, + "learning_rate": 4.115281487174872e-05, + "loss": 0.0207, + "step": 57880 + }, + { + "epoch": 3.792832339644893, + "grad_norm": 0.2830112874507904, + "learning_rate": 4.113654155905452e-05, + "loss": 0.0264, + "step": 57890 + }, + { + "epoch": 3.7934875188364017, + "grad_norm": 1.0653326511383057, + "learning_rate": 4.112026921565483e-05, + "loss": 0.0509, + "step": 57900 + }, + { + "epoch": 3.7941426980279105, + "grad_norm": 0.19694986939430237, + "learning_rate": 4.1103997843329147e-05, + "loss": 0.0339, + "step": 57910 + }, + { + "epoch": 3.7947978772194197, + "grad_norm": 0.39349815249443054, + "learning_rate": 4.108772744385689e-05, + "loss": 0.0257, + "step": 57920 + }, + { + "epoch": 3.7954530564109286, + "grad_norm": 1.0405471324920654, + "learning_rate": 4.1071458019017346e-05, + "loss": 0.0358, + "step": 57930 + }, + { + "epoch": 3.7961082356024374, + "grad_norm": 0.5150485634803772, + "learning_rate": 4.1055189570589734e-05, + "loss": 0.0394, + "step": 57940 + }, + { + "epoch": 3.796763414793946, + "grad_norm": 0.28646937012672424, + "learning_rate": 4.103892210035315e-05, + "loss": 0.0276, + "step": 57950 + }, + { + "epoch": 3.797418593985455, + "grad_norm": 0.12011610716581345, + "learning_rate": 4.102265561008656e-05, + "loss": 0.0346, + "step": 57960 + }, + { + "epoch": 3.798073773176964, + "grad_norm": 0.13283279538154602, + "learning_rate": 4.100639010156885e-05, + "loss": 0.049, + "step": 57970 + }, + { + "epoch": 3.7987289523684726, + "grad_norm": 0.7361465096473694, + "learning_rate": 4.099012557657882e-05, + "loss": 0.0573, + "step": 57980 + }, + { + "epoch": 3.7993841315599814, + "grad_norm": 0.1846204698085785, + "learning_rate": 4.09738620368951e-05, + "loss": 0.0182, + "step": 57990 + }, + { + "epoch": 3.8000393107514903, + "grad_norm": 0.2976301908493042, + "learning_rate": 4.095759948429626e-05, + "loss": 0.066, + "step": 58000 + }, + { + "epoch": 3.8006944899429995, + "grad_norm": 0.1847805231809616, + "learning_rate": 4.094133792056075e-05, + "loss": 0.0439, + "step": 58010 + }, + { + "epoch": 3.8013496691345083, + "grad_norm": 0.42352065443992615, + "learning_rate": 4.092507734746688e-05, + "loss": 0.0293, + "step": 58020 + }, + { + "epoch": 3.802004848326017, + "grad_norm": 0.48896265029907227, + "learning_rate": 4.090881776679293e-05, + "loss": 0.0324, + "step": 58030 + }, + { + "epoch": 3.802660027517526, + "grad_norm": 0.7922750115394592, + "learning_rate": 4.0892559180316994e-05, + "loss": 0.0456, + "step": 58040 + }, + { + "epoch": 3.803315206709035, + "grad_norm": 0.3786872625350952, + "learning_rate": 4.08763015898171e-05, + "loss": 0.0395, + "step": 58050 + }, + { + "epoch": 3.803970385900544, + "grad_norm": 0.2690292298793793, + "learning_rate": 4.0860044997071145e-05, + "loss": 0.0259, + "step": 58060 + }, + { + "epoch": 3.804625565092053, + "grad_norm": 0.9636183977127075, + "learning_rate": 4.084378940385693e-05, + "loss": 0.0398, + "step": 58070 + }, + { + "epoch": 3.8052807442835617, + "grad_norm": 0.23104460537433624, + "learning_rate": 4.082753481195211e-05, + "loss": 0.0281, + "step": 58080 + }, + { + "epoch": 3.8059359234750705, + "grad_norm": 0.6328854560852051, + "learning_rate": 4.0811281223134324e-05, + "loss": 0.0416, + "step": 58090 + }, + { + "epoch": 3.8065911026665793, + "grad_norm": 1.417441964149475, + "learning_rate": 4.0795028639181e-05, + "loss": 0.1083, + "step": 58100 + }, + { + "epoch": 3.807246281858088, + "grad_norm": 0.21786342561244965, + "learning_rate": 4.077877706186951e-05, + "loss": 0.0256, + "step": 58110 + }, + { + "epoch": 3.807901461049597, + "grad_norm": 0.8252599835395813, + "learning_rate": 4.076252649297709e-05, + "loss": 0.0283, + "step": 58120 + }, + { + "epoch": 3.8085566402411057, + "grad_norm": 0.23056170344352722, + "learning_rate": 4.07462769342809e-05, + "loss": 0.0294, + "step": 58130 + }, + { + "epoch": 3.8092118194326146, + "grad_norm": 0.902238130569458, + "learning_rate": 4.073002838755793e-05, + "loss": 0.0311, + "step": 58140 + }, + { + "epoch": 3.809866998624124, + "grad_norm": 0.7347462773323059, + "learning_rate": 4.0713780854585115e-05, + "loss": 0.0286, + "step": 58150 + }, + { + "epoch": 3.8105221778156326, + "grad_norm": 0.22673265635967255, + "learning_rate": 4.069753433713927e-05, + "loss": 0.0373, + "step": 58160 + }, + { + "epoch": 3.8111773570071414, + "grad_norm": 0.4525277614593506, + "learning_rate": 4.068128883699709e-05, + "loss": 0.0348, + "step": 58170 + }, + { + "epoch": 3.8118325361986503, + "grad_norm": 1.0241389274597168, + "learning_rate": 4.0665044355935155e-05, + "loss": 0.0399, + "step": 58180 + }, + { + "epoch": 3.812487715390159, + "grad_norm": 0.10615728050470352, + "learning_rate": 4.064880089572993e-05, + "loss": 0.0397, + "step": 58190 + }, + { + "epoch": 3.8131428945816683, + "grad_norm": 0.6040787100791931, + "learning_rate": 4.063255845815779e-05, + "loss": 0.0444, + "step": 58200 + }, + { + "epoch": 3.813798073773177, + "grad_norm": 0.696437418460846, + "learning_rate": 4.0616317044994945e-05, + "loss": 0.0277, + "step": 58210 + }, + { + "epoch": 3.814453252964686, + "grad_norm": 0.45948055386543274, + "learning_rate": 4.0600076658017585e-05, + "loss": 0.0386, + "step": 58220 + }, + { + "epoch": 3.8151084321561948, + "grad_norm": 0.11917272955179214, + "learning_rate": 4.0583837299001706e-05, + "loss": 0.0309, + "step": 58230 + }, + { + "epoch": 3.8157636113477036, + "grad_norm": 0.22158783674240112, + "learning_rate": 4.056759896972322e-05, + "loss": 0.0316, + "step": 58240 + }, + { + "epoch": 3.8164187905392124, + "grad_norm": 0.5172891020774841, + "learning_rate": 4.055136167195793e-05, + "loss": 0.0176, + "step": 58250 + }, + { + "epoch": 3.817073969730721, + "grad_norm": 0.5079054832458496, + "learning_rate": 4.053512540748152e-05, + "loss": 0.0259, + "step": 58260 + }, + { + "epoch": 3.81772914892223, + "grad_norm": 1.1437313556671143, + "learning_rate": 4.051889017806955e-05, + "loss": 0.0416, + "step": 58270 + }, + { + "epoch": 3.818384328113739, + "grad_norm": 0.1122591570019722, + "learning_rate": 4.0502655985497516e-05, + "loss": 0.0252, + "step": 58280 + }, + { + "epoch": 3.819039507305248, + "grad_norm": 0.9783211946487427, + "learning_rate": 4.048642283154074e-05, + "loss": 0.0496, + "step": 58290 + }, + { + "epoch": 3.819694686496757, + "grad_norm": 0.19981089234352112, + "learning_rate": 4.0470190717974464e-05, + "loss": 0.0192, + "step": 58300 + }, + { + "epoch": 3.8203498656882657, + "grad_norm": 0.11082793772220612, + "learning_rate": 4.045395964657381e-05, + "loss": 0.0359, + "step": 58310 + }, + { + "epoch": 3.8210050448797745, + "grad_norm": 0.1563756763935089, + "learning_rate": 4.043772961911376e-05, + "loss": 0.0504, + "step": 58320 + }, + { + "epoch": 3.8216602240712834, + "grad_norm": 0.20257805287837982, + "learning_rate": 4.042150063736924e-05, + "loss": 0.0316, + "step": 58330 + }, + { + "epoch": 3.8223154032627926, + "grad_norm": 0.12970362603664398, + "learning_rate": 4.0405272703114976e-05, + "loss": 0.022, + "step": 58340 + }, + { + "epoch": 3.8229705824543014, + "grad_norm": 0.9415095448493958, + "learning_rate": 4.038904581812569e-05, + "loss": 0.0488, + "step": 58350 + }, + { + "epoch": 3.8236257616458103, + "grad_norm": 0.17779791355133057, + "learning_rate": 4.03728199841759e-05, + "loss": 0.0443, + "step": 58360 + }, + { + "epoch": 3.824280940837319, + "grad_norm": 0.21296954154968262, + "learning_rate": 4.0356595203040044e-05, + "loss": 0.0413, + "step": 58370 + }, + { + "epoch": 3.824936120028828, + "grad_norm": 0.35659289360046387, + "learning_rate": 4.0340371476492426e-05, + "loss": 0.0181, + "step": 58380 + }, + { + "epoch": 3.8255912992203367, + "grad_norm": 0.5789397358894348, + "learning_rate": 4.032414880630727e-05, + "loss": 0.0558, + "step": 58390 + }, + { + "epoch": 3.8262464784118455, + "grad_norm": 0.09650544822216034, + "learning_rate": 4.0307927194258646e-05, + "loss": 0.0221, + "step": 58400 + }, + { + "epoch": 3.8269016576033543, + "grad_norm": 0.2813722789287567, + "learning_rate": 4.029170664212054e-05, + "loss": 0.0257, + "step": 58410 + }, + { + "epoch": 3.827556836794863, + "grad_norm": 0.18564485013484955, + "learning_rate": 4.0275487151666794e-05, + "loss": 0.0199, + "step": 58420 + }, + { + "epoch": 3.8282120159863724, + "grad_norm": 1.391582727432251, + "learning_rate": 4.025926872467116e-05, + "loss": 0.0331, + "step": 58430 + }, + { + "epoch": 3.828867195177881, + "grad_norm": 0.23857226967811584, + "learning_rate": 4.024305136290725e-05, + "loss": 0.0579, + "step": 58440 + }, + { + "epoch": 3.82952237436939, + "grad_norm": 0.25762128829956055, + "learning_rate": 4.022683506814858e-05, + "loss": 0.0283, + "step": 58450 + }, + { + "epoch": 3.830177553560899, + "grad_norm": 0.6160880327224731, + "learning_rate": 4.021061984216853e-05, + "loss": 0.046, + "step": 58460 + }, + { + "epoch": 3.8308327327524077, + "grad_norm": 0.831180989742279, + "learning_rate": 4.019440568674035e-05, + "loss": 0.0333, + "step": 58470 + }, + { + "epoch": 3.831487911943917, + "grad_norm": 1.01318359375, + "learning_rate": 4.017819260363724e-05, + "loss": 0.0509, + "step": 58480 + }, + { + "epoch": 3.8321430911354257, + "grad_norm": 0.2993878424167633, + "learning_rate": 4.016198059463223e-05, + "loss": 0.0256, + "step": 58490 + }, + { + "epoch": 3.8327982703269345, + "grad_norm": 0.6499392986297607, + "learning_rate": 4.0145769661498233e-05, + "loss": 0.0135, + "step": 58500 + }, + { + "epoch": 3.8334534495184434, + "grad_norm": 0.10017747431993484, + "learning_rate": 4.012955980600804e-05, + "loss": 0.0309, + "step": 58510 + }, + { + "epoch": 3.834108628709952, + "grad_norm": 0.41739895939826965, + "learning_rate": 4.0113351029934355e-05, + "loss": 0.0561, + "step": 58520 + }, + { + "epoch": 3.834763807901461, + "grad_norm": 0.7192800641059875, + "learning_rate": 4.0097143335049714e-05, + "loss": 0.0282, + "step": 58530 + }, + { + "epoch": 3.83541898709297, + "grad_norm": 1.0061441659927368, + "learning_rate": 4.0080936723126606e-05, + "loss": 0.0324, + "step": 58540 + }, + { + "epoch": 3.8360741662844786, + "grad_norm": 0.24537308514118195, + "learning_rate": 4.006473119593735e-05, + "loss": 0.02, + "step": 58550 + }, + { + "epoch": 3.8367293454759874, + "grad_norm": 0.1816800981760025, + "learning_rate": 4.004852675525415e-05, + "loss": 0.0316, + "step": 58560 + }, + { + "epoch": 3.8373845246674967, + "grad_norm": 0.3789042830467224, + "learning_rate": 4.00323234028491e-05, + "loss": 0.0232, + "step": 58570 + }, + { + "epoch": 3.8380397038590055, + "grad_norm": 0.17266301810741425, + "learning_rate": 4.0016121140494186e-05, + "loss": 0.0282, + "step": 58580 + }, + { + "epoch": 3.8386948830505143, + "grad_norm": 0.6307696104049683, + "learning_rate": 3.999991996996122e-05, + "loss": 0.0299, + "step": 58590 + }, + { + "epoch": 3.839350062242023, + "grad_norm": 1.7506178617477417, + "learning_rate": 3.9983719893022e-05, + "loss": 0.0351, + "step": 58600 + }, + { + "epoch": 3.840005241433532, + "grad_norm": 0.8638938069343567, + "learning_rate": 3.996752091144812e-05, + "loss": 0.0287, + "step": 58610 + }, + { + "epoch": 3.840660420625041, + "grad_norm": 0.750250518321991, + "learning_rate": 3.9951323027011055e-05, + "loss": 0.0273, + "step": 58620 + }, + { + "epoch": 3.84131559981655, + "grad_norm": 0.5646341443061829, + "learning_rate": 3.993512624148219e-05, + "loss": 0.0336, + "step": 58630 + }, + { + "epoch": 3.841970779008059, + "grad_norm": 0.6132603287696838, + "learning_rate": 3.99189305566328e-05, + "loss": 0.0347, + "step": 58640 + }, + { + "epoch": 3.8426259581995676, + "grad_norm": 0.6673178672790527, + "learning_rate": 3.990273597423401e-05, + "loss": 0.0498, + "step": 58650 + }, + { + "epoch": 3.8432811373910765, + "grad_norm": 0.3663729429244995, + "learning_rate": 3.988654249605681e-05, + "loss": 0.0349, + "step": 58660 + }, + { + "epoch": 3.8439363165825853, + "grad_norm": 0.33556148409843445, + "learning_rate": 3.987035012387213e-05, + "loss": 0.0374, + "step": 58670 + }, + { + "epoch": 3.844591495774094, + "grad_norm": 0.20747561752796173, + "learning_rate": 3.9854158859450735e-05, + "loss": 0.0389, + "step": 58680 + }, + { + "epoch": 3.845246674965603, + "grad_norm": 0.8891061544418335, + "learning_rate": 3.983796870456328e-05, + "loss": 0.0447, + "step": 58690 + }, + { + "epoch": 3.8459018541571117, + "grad_norm": 0.4693279266357422, + "learning_rate": 3.982177966098028e-05, + "loss": 0.0172, + "step": 58700 + }, + { + "epoch": 3.846557033348621, + "grad_norm": 0.05832390859723091, + "learning_rate": 3.980559173047216e-05, + "loss": 0.0234, + "step": 58710 + }, + { + "epoch": 3.84721221254013, + "grad_norm": 1.110649824142456, + "learning_rate": 3.9789404914809176e-05, + "loss": 0.0505, + "step": 58720 + }, + { + "epoch": 3.8478673917316386, + "grad_norm": 0.17194926738739014, + "learning_rate": 3.9773219215761544e-05, + "loss": 0.0221, + "step": 58730 + }, + { + "epoch": 3.8485225709231474, + "grad_norm": 0.4022351801395416, + "learning_rate": 3.975703463509928e-05, + "loss": 0.0288, + "step": 58740 + }, + { + "epoch": 3.8491777501146562, + "grad_norm": 0.6245572566986084, + "learning_rate": 3.974085117459231e-05, + "loss": 0.0271, + "step": 58750 + }, + { + "epoch": 3.8498329293061655, + "grad_norm": 0.13761502504348755, + "learning_rate": 3.972466883601043e-05, + "loss": 0.0381, + "step": 58760 + }, + { + "epoch": 3.8504881084976743, + "grad_norm": 0.4276558756828308, + "learning_rate": 3.9708487621123316e-05, + "loss": 0.0344, + "step": 58770 + }, + { + "epoch": 3.851143287689183, + "grad_norm": 0.3319995701313019, + "learning_rate": 3.969230753170051e-05, + "loss": 0.0351, + "step": 58780 + }, + { + "epoch": 3.851798466880692, + "grad_norm": 0.5609530210494995, + "learning_rate": 3.967612856951146e-05, + "loss": 0.0424, + "step": 58790 + }, + { + "epoch": 3.8524536460722008, + "grad_norm": 0.46494194865226746, + "learning_rate": 3.9659950736325476e-05, + "loss": 0.0562, + "step": 58800 + }, + { + "epoch": 3.8531088252637096, + "grad_norm": 0.10697679221630096, + "learning_rate": 3.964377403391174e-05, + "loss": 0.026, + "step": 58810 + }, + { + "epoch": 3.8537640044552184, + "grad_norm": 0.49375998973846436, + "learning_rate": 3.9627598464039303e-05, + "loss": 0.0293, + "step": 58820 + }, + { + "epoch": 3.854419183646727, + "grad_norm": 0.3553488850593567, + "learning_rate": 3.9611424028477096e-05, + "loss": 0.0473, + "step": 58830 + }, + { + "epoch": 3.855074362838236, + "grad_norm": 0.9093111753463745, + "learning_rate": 3.959525072899394e-05, + "loss": 0.0407, + "step": 58840 + }, + { + "epoch": 3.8557295420297453, + "grad_norm": 1.0244874954223633, + "learning_rate": 3.9579078567358506e-05, + "loss": 0.0453, + "step": 58850 + }, + { + "epoch": 3.856384721221254, + "grad_norm": 0.6839359402656555, + "learning_rate": 3.956290754533939e-05, + "loss": 0.0356, + "step": 58860 + }, + { + "epoch": 3.857039900412763, + "grad_norm": 0.615214467048645, + "learning_rate": 3.954673766470501e-05, + "loss": 0.0686, + "step": 58870 + }, + { + "epoch": 3.8576950796042717, + "grad_norm": 1.6693198680877686, + "learning_rate": 3.9530568927223674e-05, + "loss": 0.0315, + "step": 58880 + }, + { + "epoch": 3.8583502587957805, + "grad_norm": 0.555528998374939, + "learning_rate": 3.95144013346636e-05, + "loss": 0.0426, + "step": 58890 + }, + { + "epoch": 3.85900543798729, + "grad_norm": 0.09810708463191986, + "learning_rate": 3.949823488879282e-05, + "loss": 0.0299, + "step": 58900 + }, + { + "epoch": 3.8596606171787986, + "grad_norm": 0.6918981671333313, + "learning_rate": 3.9482069591379276e-05, + "loss": 0.0214, + "step": 58910 + }, + { + "epoch": 3.8603157963703074, + "grad_norm": 0.09839969128370285, + "learning_rate": 3.9465905444190784e-05, + "loss": 0.0443, + "step": 58920 + }, + { + "epoch": 3.8609709755618162, + "grad_norm": 0.9277000427246094, + "learning_rate": 3.944974244899505e-05, + "loss": 0.0531, + "step": 58930 + }, + { + "epoch": 3.861626154753325, + "grad_norm": 0.2230236828327179, + "learning_rate": 3.9433580607559616e-05, + "loss": 0.0395, + "step": 58940 + }, + { + "epoch": 3.862281333944834, + "grad_norm": 0.1613362431526184, + "learning_rate": 3.941741992165193e-05, + "loss": 0.0375, + "step": 58950 + }, + { + "epoch": 3.8629365131363427, + "grad_norm": 0.2833113372325897, + "learning_rate": 3.940126039303928e-05, + "loss": 0.0389, + "step": 58960 + }, + { + "epoch": 3.8635916923278515, + "grad_norm": 0.3667421340942383, + "learning_rate": 3.938510202348886e-05, + "loss": 0.0244, + "step": 58970 + }, + { + "epoch": 3.8642468715193603, + "grad_norm": 0.5650490522384644, + "learning_rate": 3.93689448147677e-05, + "loss": 0.0433, + "step": 58980 + }, + { + "epoch": 3.8649020507108696, + "grad_norm": 0.3119334280490875, + "learning_rate": 3.935278876864278e-05, + "loss": 0.0338, + "step": 58990 + }, + { + "epoch": 3.8655572299023784, + "grad_norm": 0.4638400375843048, + "learning_rate": 3.933663388688087e-05, + "loss": 0.0373, + "step": 59000 + }, + { + "epoch": 3.866212409093887, + "grad_norm": 0.5590919852256775, + "learning_rate": 3.9320480171248645e-05, + "loss": 0.0464, + "step": 59010 + }, + { + "epoch": 3.866867588285396, + "grad_norm": 0.32828015089035034, + "learning_rate": 3.930432762351265e-05, + "loss": 0.0348, + "step": 59020 + }, + { + "epoch": 3.867522767476905, + "grad_norm": 0.07931467890739441, + "learning_rate": 3.928817624543931e-05, + "loss": 0.0451, + "step": 59030 + }, + { + "epoch": 3.868177946668414, + "grad_norm": 0.32017290592193604, + "learning_rate": 3.9272026038794885e-05, + "loss": 0.0451, + "step": 59040 + }, + { + "epoch": 3.868833125859923, + "grad_norm": 0.6878113746643066, + "learning_rate": 3.9255877005345586e-05, + "loss": 0.0396, + "step": 59050 + }, + { + "epoch": 3.8694883050514317, + "grad_norm": 0.21222496032714844, + "learning_rate": 3.923972914685743e-05, + "loss": 0.0246, + "step": 59060 + }, + { + "epoch": 3.8701434842429405, + "grad_norm": 0.2601158916950226, + "learning_rate": 3.922358246509631e-05, + "loss": 0.0282, + "step": 59070 + }, + { + "epoch": 3.8707986634344493, + "grad_norm": 0.3762374222278595, + "learning_rate": 3.920743696182801e-05, + "loss": 0.0406, + "step": 59080 + }, + { + "epoch": 3.871453842625958, + "grad_norm": 2.501349687576294, + "learning_rate": 3.9191292638818164e-05, + "loss": 0.045, + "step": 59090 + }, + { + "epoch": 3.872109021817467, + "grad_norm": 0.4456925690174103, + "learning_rate": 3.9175149497832316e-05, + "loss": 0.0388, + "step": 59100 + }, + { + "epoch": 3.872764201008976, + "grad_norm": 0.5988945364952087, + "learning_rate": 3.915900754063584e-05, + "loss": 0.0411, + "step": 59110 + }, + { + "epoch": 3.8734193802004846, + "grad_norm": 0.7878968119621277, + "learning_rate": 3.9142866768994014e-05, + "loss": 0.0666, + "step": 59120 + }, + { + "epoch": 3.874074559391994, + "grad_norm": 0.3292161822319031, + "learning_rate": 3.912672718467193e-05, + "loss": 0.0233, + "step": 59130 + }, + { + "epoch": 3.8747297385835027, + "grad_norm": 0.3789961338043213, + "learning_rate": 3.911058878943463e-05, + "loss": 0.0365, + "step": 59140 + }, + { + "epoch": 3.8753849177750115, + "grad_norm": 0.7066669464111328, + "learning_rate": 3.9094451585046974e-05, + "loss": 0.0297, + "step": 59150 + }, + { + "epoch": 3.8760400969665203, + "grad_norm": 0.10434693098068237, + "learning_rate": 3.90783155732737e-05, + "loss": 0.0196, + "step": 59160 + }, + { + "epoch": 3.876695276158029, + "grad_norm": 0.6590308547019958, + "learning_rate": 3.9062180755879373e-05, + "loss": 0.0472, + "step": 59170 + }, + { + "epoch": 3.8773504553495384, + "grad_norm": 0.3113664984703064, + "learning_rate": 3.904604713462855e-05, + "loss": 0.037, + "step": 59180 + }, + { + "epoch": 3.878005634541047, + "grad_norm": 0.5923998355865479, + "learning_rate": 3.902991471128554e-05, + "loss": 0.0324, + "step": 59190 + }, + { + "epoch": 3.878660813732556, + "grad_norm": 0.7683789134025574, + "learning_rate": 3.901378348761456e-05, + "loss": 0.0401, + "step": 59200 + }, + { + "epoch": 3.879315992924065, + "grad_norm": 0.6472174525260925, + "learning_rate": 3.899765346537971e-05, + "loss": 0.036, + "step": 59210 + }, + { + "epoch": 3.8799711721155736, + "grad_norm": 0.3614961504936218, + "learning_rate": 3.898152464634493e-05, + "loss": 0.0333, + "step": 59220 + }, + { + "epoch": 3.8806263513070824, + "grad_norm": 0.3039950132369995, + "learning_rate": 3.896539703227402e-05, + "loss": 0.0267, + "step": 59230 + }, + { + "epoch": 3.8812815304985913, + "grad_norm": 0.2439994066953659, + "learning_rate": 3.894927062493073e-05, + "loss": 0.0391, + "step": 59240 + }, + { + "epoch": 3.8819367096901, + "grad_norm": 0.17503467202186584, + "learning_rate": 3.893314542607859e-05, + "loss": 0.0624, + "step": 59250 + }, + { + "epoch": 3.882591888881609, + "grad_norm": 0.5453868508338928, + "learning_rate": 3.891702143748103e-05, + "loss": 0.0288, + "step": 59260 + }, + { + "epoch": 3.883247068073118, + "grad_norm": 1.6817985773086548, + "learning_rate": 3.890089866090133e-05, + "loss": 0.024, + "step": 59270 + }, + { + "epoch": 3.883902247264627, + "grad_norm": 0.2584618628025055, + "learning_rate": 3.888477709810267e-05, + "loss": 0.0304, + "step": 59280 + }, + { + "epoch": 3.8845574264561358, + "grad_norm": 0.15565308928489685, + "learning_rate": 3.886865675084806e-05, + "loss": 0.0334, + "step": 59290 + }, + { + "epoch": 3.8852126056476446, + "grad_norm": 0.2494911104440689, + "learning_rate": 3.885253762090043e-05, + "loss": 0.0665, + "step": 59300 + }, + { + "epoch": 3.8858677848391534, + "grad_norm": 0.12327440083026886, + "learning_rate": 3.8836419710022534e-05, + "loss": 0.0223, + "step": 59310 + }, + { + "epoch": 3.8865229640306627, + "grad_norm": 0.14222992956638336, + "learning_rate": 3.882030301997698e-05, + "loss": 0.021, + "step": 59320 + }, + { + "epoch": 3.8871781432221715, + "grad_norm": 0.09821907430887222, + "learning_rate": 3.880418755252629e-05, + "loss": 0.1216, + "step": 59330 + }, + { + "epoch": 3.8878333224136803, + "grad_norm": 0.09537876397371292, + "learning_rate": 3.87880733094328e-05, + "loss": 0.0447, + "step": 59340 + }, + { + "epoch": 3.888488501605189, + "grad_norm": 0.40664902329444885, + "learning_rate": 3.877196029245877e-05, + "loss": 0.0232, + "step": 59350 + }, + { + "epoch": 3.889143680796698, + "grad_norm": 0.13087600469589233, + "learning_rate": 3.875584850336627e-05, + "loss": 0.024, + "step": 59360 + }, + { + "epoch": 3.8897988599882067, + "grad_norm": 0.07490125298500061, + "learning_rate": 3.873973794391728e-05, + "loss": 0.0259, + "step": 59370 + }, + { + "epoch": 3.8904540391797156, + "grad_norm": 0.2636473476886749, + "learning_rate": 3.872362861587361e-05, + "loss": 0.02, + "step": 59380 + }, + { + "epoch": 3.8911092183712244, + "grad_norm": 1.8338842391967773, + "learning_rate": 3.870752052099698e-05, + "loss": 0.0643, + "step": 59390 + }, + { + "epoch": 3.891764397562733, + "grad_norm": 0.26903828978538513, + "learning_rate": 3.8691413661048916e-05, + "loss": 0.0606, + "step": 59400 + }, + { + "epoch": 3.8924195767542424, + "grad_norm": 0.6559885144233704, + "learning_rate": 3.867530803779085e-05, + "loss": 0.0209, + "step": 59410 + }, + { + "epoch": 3.8930747559457513, + "grad_norm": 0.6611643433570862, + "learning_rate": 3.8659203652984054e-05, + "loss": 0.0322, + "step": 59420 + }, + { + "epoch": 3.89372993513726, + "grad_norm": 0.4459093511104584, + "learning_rate": 3.864310050838972e-05, + "loss": 0.0333, + "step": 59430 + }, + { + "epoch": 3.894385114328769, + "grad_norm": 0.6336097717285156, + "learning_rate": 3.8626998605768835e-05, + "loss": 0.0432, + "step": 59440 + }, + { + "epoch": 3.8950402935202777, + "grad_norm": 0.10812705755233765, + "learning_rate": 3.8610897946882294e-05, + "loss": 0.0263, + "step": 59450 + }, + { + "epoch": 3.8956954727117865, + "grad_norm": 0.3012797236442566, + "learning_rate": 3.859479853349082e-05, + "loss": 0.0321, + "step": 59460 + }, + { + "epoch": 3.8963506519032958, + "grad_norm": 0.041722577065229416, + "learning_rate": 3.857870036735503e-05, + "loss": 0.0381, + "step": 59470 + }, + { + "epoch": 3.8970058310948046, + "grad_norm": 0.47280460596084595, + "learning_rate": 3.856260345023538e-05, + "loss": 0.0224, + "step": 59480 + }, + { + "epoch": 3.8976610102863134, + "grad_norm": 0.23849551379680634, + "learning_rate": 3.854650778389224e-05, + "loss": 0.033, + "step": 59490 + }, + { + "epoch": 3.898316189477822, + "grad_norm": 0.3876032531261444, + "learning_rate": 3.8530413370085786e-05, + "loss": 0.0283, + "step": 59500 + }, + { + "epoch": 3.898971368669331, + "grad_norm": 0.36997702717781067, + "learning_rate": 3.851432021057608e-05, + "loss": 0.0595, + "step": 59510 + }, + { + "epoch": 3.89962654786084, + "grad_norm": 0.7113068103790283, + "learning_rate": 3.849822830712306e-05, + "loss": 0.0202, + "step": 59520 + }, + { + "epoch": 3.9002817270523487, + "grad_norm": 1.6209163665771484, + "learning_rate": 3.848213766148649e-05, + "loss": 0.0329, + "step": 59530 + }, + { + "epoch": 3.9009369062438575, + "grad_norm": 0.4952968955039978, + "learning_rate": 3.846604827542602e-05, + "loss": 0.0397, + "step": 59540 + }, + { + "epoch": 3.9015920854353667, + "grad_norm": 0.5593913197517395, + "learning_rate": 3.844996015070117e-05, + "loss": 0.0176, + "step": 59550 + }, + { + "epoch": 3.9022472646268755, + "grad_norm": 1.2370481491088867, + "learning_rate": 3.843387328907132e-05, + "loss": 0.0397, + "step": 59560 + }, + { + "epoch": 3.9029024438183844, + "grad_norm": 0.569191575050354, + "learning_rate": 3.8417787692295696e-05, + "loss": 0.0219, + "step": 59570 + }, + { + "epoch": 3.903557623009893, + "grad_norm": 0.6358357071876526, + "learning_rate": 3.84017033621334e-05, + "loss": 0.0252, + "step": 59580 + }, + { + "epoch": 3.904212802201402, + "grad_norm": 0.19361911714076996, + "learning_rate": 3.838562030034338e-05, + "loss": 0.0154, + "step": 59590 + }, + { + "epoch": 3.904867981392911, + "grad_norm": 0.768300473690033, + "learning_rate": 3.836953850868447e-05, + "loss": 0.0367, + "step": 59600 + }, + { + "epoch": 3.90552316058442, + "grad_norm": 0.9126607775688171, + "learning_rate": 3.835345798891533e-05, + "loss": 0.026, + "step": 59610 + }, + { + "epoch": 3.906178339775929, + "grad_norm": 0.3267541825771332, + "learning_rate": 3.833737874279452e-05, + "loss": 0.0383, + "step": 59620 + }, + { + "epoch": 3.9068335189674377, + "grad_norm": 0.5656448006629944, + "learning_rate": 3.8321300772080433e-05, + "loss": 0.0179, + "step": 59630 + }, + { + "epoch": 3.9074886981589465, + "grad_norm": 0.06482040882110596, + "learning_rate": 3.8305224078531345e-05, + "loss": 0.0288, + "step": 59640 + }, + { + "epoch": 3.9081438773504553, + "grad_norm": 0.5048551559448242, + "learning_rate": 3.828914866390537e-05, + "loss": 0.0292, + "step": 59650 + }, + { + "epoch": 3.908799056541964, + "grad_norm": 0.29938602447509766, + "learning_rate": 3.827307452996048e-05, + "loss": 0.0297, + "step": 59660 + }, + { + "epoch": 3.909454235733473, + "grad_norm": 0.5262569785118103, + "learning_rate": 3.825700167845454e-05, + "loss": 0.0301, + "step": 59670 + }, + { + "epoch": 3.9101094149249818, + "grad_norm": 1.1609773635864258, + "learning_rate": 3.8240930111145214e-05, + "loss": 0.0492, + "step": 59680 + }, + { + "epoch": 3.910764594116491, + "grad_norm": 0.41077688336372375, + "learning_rate": 3.8224859829790116e-05, + "loss": 0.0364, + "step": 59690 + }, + { + "epoch": 3.911419773308, + "grad_norm": 0.6418519616127014, + "learning_rate": 3.820879083614664e-05, + "loss": 0.0456, + "step": 59700 + }, + { + "epoch": 3.9120749524995087, + "grad_norm": 0.2230045050382614, + "learning_rate": 3.819272313197208e-05, + "loss": 0.0388, + "step": 59710 + }, + { + "epoch": 3.9127301316910175, + "grad_norm": 1.1271125078201294, + "learning_rate": 3.817665671902355e-05, + "loss": 0.0272, + "step": 59720 + }, + { + "epoch": 3.9133853108825263, + "grad_norm": 0.6024459004402161, + "learning_rate": 3.816059159905808e-05, + "loss": 0.0338, + "step": 59730 + }, + { + "epoch": 3.914040490074035, + "grad_norm": 0.3040097653865814, + "learning_rate": 3.8144527773832476e-05, + "loss": 0.0586, + "step": 59740 + }, + { + "epoch": 3.9146956692655444, + "grad_norm": 0.22171996533870697, + "learning_rate": 3.812846524510352e-05, + "loss": 0.0309, + "step": 59750 + }, + { + "epoch": 3.915350848457053, + "grad_norm": 0.9533942937850952, + "learning_rate": 3.811240401462775e-05, + "loss": 0.0264, + "step": 59760 + }, + { + "epoch": 3.916006027648562, + "grad_norm": 1.3836723566055298, + "learning_rate": 3.809634408416162e-05, + "loss": 0.0314, + "step": 59770 + }, + { + "epoch": 3.916661206840071, + "grad_norm": 0.15507692098617554, + "learning_rate": 3.8080285455461394e-05, + "loss": 0.0171, + "step": 59780 + }, + { + "epoch": 3.9173163860315796, + "grad_norm": 0.10435330122709274, + "learning_rate": 3.806422813028323e-05, + "loss": 0.0486, + "step": 59790 + }, + { + "epoch": 3.9179715652230884, + "grad_norm": 1.1782044172286987, + "learning_rate": 3.8048172110383114e-05, + "loss": 0.0591, + "step": 59800 + }, + { + "epoch": 3.9186267444145972, + "grad_norm": 0.21469752490520477, + "learning_rate": 3.8032117397516944e-05, + "loss": 0.021, + "step": 59810 + }, + { + "epoch": 3.919281923606106, + "grad_norm": 0.36876049637794495, + "learning_rate": 3.801606399344042e-05, + "loss": 0.0374, + "step": 59820 + }, + { + "epoch": 3.9199371027976153, + "grad_norm": 0.29706892371177673, + "learning_rate": 3.800001189990913e-05, + "loss": 0.021, + "step": 59830 + }, + { + "epoch": 3.920592281989124, + "grad_norm": 0.8905520439147949, + "learning_rate": 3.798396111867847e-05, + "loss": 0.0599, + "step": 59840 + }, + { + "epoch": 3.921247461180633, + "grad_norm": 0.11599646508693695, + "learning_rate": 3.796791165150378e-05, + "loss": 0.0131, + "step": 59850 + }, + { + "epoch": 3.9219026403721418, + "grad_norm": 0.7396306991577148, + "learning_rate": 3.795186350014018e-05, + "loss": 0.0257, + "step": 59860 + }, + { + "epoch": 3.9225578195636506, + "grad_norm": 0.14270643889904022, + "learning_rate": 3.793581666634266e-05, + "loss": 0.0519, + "step": 59870 + }, + { + "epoch": 3.9232129987551594, + "grad_norm": 0.9206165075302124, + "learning_rate": 3.791977115186609e-05, + "loss": 0.0315, + "step": 59880 + }, + { + "epoch": 3.9238681779466686, + "grad_norm": 0.09394565969705582, + "learning_rate": 3.79037269584652e-05, + "loss": 0.0339, + "step": 59890 + }, + { + "epoch": 3.9245233571381775, + "grad_norm": 0.3543013036251068, + "learning_rate": 3.7887684087894545e-05, + "loss": 0.0368, + "step": 59900 + }, + { + "epoch": 3.9251785363296863, + "grad_norm": 1.919386625289917, + "learning_rate": 3.787164254190854e-05, + "loss": 0.0182, + "step": 59910 + }, + { + "epoch": 3.925833715521195, + "grad_norm": 0.1429200917482376, + "learning_rate": 3.7855602322261473e-05, + "loss": 0.029, + "step": 59920 + }, + { + "epoch": 3.926488894712704, + "grad_norm": 0.6055293679237366, + "learning_rate": 3.783956343070746e-05, + "loss": 0.0278, + "step": 59930 + }, + { + "epoch": 3.9271440739042127, + "grad_norm": 0.2694997787475586, + "learning_rate": 3.7823525869000526e-05, + "loss": 0.0262, + "step": 59940 + }, + { + "epoch": 3.9277992530957215, + "grad_norm": 0.15771619975566864, + "learning_rate": 3.78074896388945e-05, + "loss": 0.0443, + "step": 59950 + }, + { + "epoch": 3.9284544322872303, + "grad_norm": 0.5024226307868958, + "learning_rate": 3.7791454742143075e-05, + "loss": 0.0308, + "step": 59960 + }, + { + "epoch": 3.929109611478739, + "grad_norm": 0.21680250763893127, + "learning_rate": 3.7775421180499795e-05, + "loss": 0.0297, + "step": 59970 + }, + { + "epoch": 3.9297647906702484, + "grad_norm": 0.10436999052762985, + "learning_rate": 3.7759388955718086e-05, + "loss": 0.0536, + "step": 59980 + }, + { + "epoch": 3.9304199698617572, + "grad_norm": 0.49394023418426514, + "learning_rate": 3.774335806955117e-05, + "loss": 0.0419, + "step": 59990 + }, + { + "epoch": 3.931075149053266, + "grad_norm": 1.0521951913833618, + "learning_rate": 3.7727328523752215e-05, + "loss": 0.0532, + "step": 60000 + }, + { + "epoch": 3.931730328244775, + "grad_norm": 0.3217744827270508, + "learning_rate": 3.771130032007415e-05, + "loss": 0.0221, + "step": 60010 + }, + { + "epoch": 3.9323855074362837, + "grad_norm": 0.6742395162582397, + "learning_rate": 3.769527346026982e-05, + "loss": 0.0412, + "step": 60020 + }, + { + "epoch": 3.933040686627793, + "grad_norm": 0.2416759878396988, + "learning_rate": 3.767924794609188e-05, + "loss": 0.024, + "step": 60030 + }, + { + "epoch": 3.9336958658193018, + "grad_norm": 0.11314300447702408, + "learning_rate": 3.7663223779292854e-05, + "loss": 0.033, + "step": 60040 + }, + { + "epoch": 3.9343510450108106, + "grad_norm": 0.20311768352985382, + "learning_rate": 3.764720096162512e-05, + "loss": 0.027, + "step": 60050 + }, + { + "epoch": 3.9350062242023194, + "grad_norm": 0.3236399292945862, + "learning_rate": 3.763117949484092e-05, + "loss": 0.0179, + "step": 60060 + }, + { + "epoch": 3.935661403393828, + "grad_norm": 0.34350475668907166, + "learning_rate": 3.7615159380692344e-05, + "loss": 0.024, + "step": 60070 + }, + { + "epoch": 3.936316582585337, + "grad_norm": 0.45559969544410706, + "learning_rate": 3.759914062093132e-05, + "loss": 0.0361, + "step": 60080 + }, + { + "epoch": 3.936971761776846, + "grad_norm": 0.6663540005683899, + "learning_rate": 3.7583123217309615e-05, + "loss": 0.0374, + "step": 60090 + }, + { + "epoch": 3.9376269409683546, + "grad_norm": 0.24770106375217438, + "learning_rate": 3.7567107171578904e-05, + "loss": 0.0333, + "step": 60100 + }, + { + "epoch": 3.9382821201598635, + "grad_norm": 0.4493127763271332, + "learning_rate": 3.755109248549066e-05, + "loss": 0.0275, + "step": 60110 + }, + { + "epoch": 3.9389372993513727, + "grad_norm": 0.8978562355041504, + "learning_rate": 3.75350791607962e-05, + "loss": 0.029, + "step": 60120 + }, + { + "epoch": 3.9395924785428815, + "grad_norm": 0.1553824543952942, + "learning_rate": 3.751906719924676e-05, + "loss": 0.0739, + "step": 60130 + }, + { + "epoch": 3.9402476577343903, + "grad_norm": 0.6665024161338806, + "learning_rate": 3.750305660259337e-05, + "loss": 0.0477, + "step": 60140 + }, + { + "epoch": 3.940902836925899, + "grad_norm": 0.08240238577127457, + "learning_rate": 3.748704737258693e-05, + "loss": 0.0284, + "step": 60150 + }, + { + "epoch": 3.941558016117408, + "grad_norm": 0.2096014767885208, + "learning_rate": 3.747103951097816e-05, + "loss": 0.0135, + "step": 60160 + }, + { + "epoch": 3.9422131953089172, + "grad_norm": 0.17751207947731018, + "learning_rate": 3.745503301951768e-05, + "loss": 0.0526, + "step": 60170 + }, + { + "epoch": 3.942868374500426, + "grad_norm": 0.41053107380867004, + "learning_rate": 3.743902789995592e-05, + "loss": 0.0298, + "step": 60180 + }, + { + "epoch": 3.943523553691935, + "grad_norm": 0.8723526000976562, + "learning_rate": 3.7423024154043183e-05, + "loss": 0.048, + "step": 60190 + }, + { + "epoch": 3.9441787328834437, + "grad_norm": 0.6515524387359619, + "learning_rate": 3.740702178352963e-05, + "loss": 0.0847, + "step": 60200 + }, + { + "epoch": 3.9448339120749525, + "grad_norm": 0.14314688742160797, + "learning_rate": 3.739102079016523e-05, + "loss": 0.0505, + "step": 60210 + }, + { + "epoch": 3.9454890912664613, + "grad_norm": 0.3199620246887207, + "learning_rate": 3.7375021175699846e-05, + "loss": 0.0296, + "step": 60220 + }, + { + "epoch": 3.94614427045797, + "grad_norm": 0.9384317398071289, + "learning_rate": 3.735902294188317e-05, + "loss": 0.0321, + "step": 60230 + }, + { + "epoch": 3.946799449649479, + "grad_norm": 0.28031405806541443, + "learning_rate": 3.734302609046473e-05, + "loss": 0.0437, + "step": 60240 + }, + { + "epoch": 3.9474546288409877, + "grad_norm": 0.20296624302864075, + "learning_rate": 3.732703062319391e-05, + "loss": 0.0214, + "step": 60250 + }, + { + "epoch": 3.948109808032497, + "grad_norm": 1.0303611755371094, + "learning_rate": 3.731103654181999e-05, + "loss": 0.0234, + "step": 60260 + }, + { + "epoch": 3.948764987224006, + "grad_norm": 0.26073166728019714, + "learning_rate": 3.729504384809203e-05, + "loss": 0.0448, + "step": 60270 + }, + { + "epoch": 3.9494201664155146, + "grad_norm": 0.11621750891208649, + "learning_rate": 3.727905254375898e-05, + "loss": 0.0153, + "step": 60280 + }, + { + "epoch": 3.9500753456070234, + "grad_norm": 0.13980385661125183, + "learning_rate": 3.7263062630569614e-05, + "loss": 0.0266, + "step": 60290 + }, + { + "epoch": 3.9507305247985323, + "grad_norm": 0.19534549117088318, + "learning_rate": 3.724707411027256e-05, + "loss": 0.0246, + "step": 60300 + }, + { + "epoch": 3.9513857039900415, + "grad_norm": 0.6750345826148987, + "learning_rate": 3.723108698461631e-05, + "loss": 0.0423, + "step": 60310 + }, + { + "epoch": 3.9520408831815503, + "grad_norm": 0.16162312030792236, + "learning_rate": 3.72151012553492e-05, + "loss": 0.0441, + "step": 60320 + }, + { + "epoch": 3.952696062373059, + "grad_norm": 0.5439394116401672, + "learning_rate": 3.719911692421939e-05, + "loss": 0.0435, + "step": 60330 + }, + { + "epoch": 3.953351241564568, + "grad_norm": 0.4763711392879486, + "learning_rate": 3.71831339929749e-05, + "loss": 0.0418, + "step": 60340 + }, + { + "epoch": 3.954006420756077, + "grad_norm": 0.6509830355644226, + "learning_rate": 3.716715246336362e-05, + "loss": 0.0311, + "step": 60350 + }, + { + "epoch": 3.9546615999475856, + "grad_norm": 0.3707696795463562, + "learning_rate": 3.715117233713324e-05, + "loss": 0.037, + "step": 60360 + }, + { + "epoch": 3.9553167791390944, + "grad_norm": 0.4483135938644409, + "learning_rate": 3.713519361603135e-05, + "loss": 0.0417, + "step": 60370 + }, + { + "epoch": 3.955971958330603, + "grad_norm": 0.4878486394882202, + "learning_rate": 3.711921630180532e-05, + "loss": 0.0404, + "step": 60380 + }, + { + "epoch": 3.956627137522112, + "grad_norm": 0.6739090085029602, + "learning_rate": 3.710324039620245e-05, + "loss": 0.0245, + "step": 60390 + }, + { + "epoch": 3.9572823167136213, + "grad_norm": 0.2847435772418976, + "learning_rate": 3.708726590096982e-05, + "loss": 0.0273, + "step": 60400 + }, + { + "epoch": 3.95793749590513, + "grad_norm": 0.7746772170066833, + "learning_rate": 3.707129281785437e-05, + "loss": 0.0321, + "step": 60410 + }, + { + "epoch": 3.958592675096639, + "grad_norm": 0.3324349522590637, + "learning_rate": 3.705532114860291e-05, + "loss": 0.0312, + "step": 60420 + }, + { + "epoch": 3.9592478542881477, + "grad_norm": 0.3956259787082672, + "learning_rate": 3.703935089496207e-05, + "loss": 0.0481, + "step": 60430 + }, + { + "epoch": 3.9599030334796566, + "grad_norm": 0.238908588886261, + "learning_rate": 3.702338205867829e-05, + "loss": 0.0616, + "step": 60440 + }, + { + "epoch": 3.960558212671166, + "grad_norm": 0.360522985458374, + "learning_rate": 3.7007414641497976e-05, + "loss": 0.0397, + "step": 60450 + }, + { + "epoch": 3.9612133918626746, + "grad_norm": 1.005061149597168, + "learning_rate": 3.699144864516726e-05, + "loss": 0.0222, + "step": 60460 + }, + { + "epoch": 3.9618685710541834, + "grad_norm": 0.8247829675674438, + "learning_rate": 3.6975484071432153e-05, + "loss": 0.027, + "step": 60470 + }, + { + "epoch": 3.9625237502456923, + "grad_norm": 0.8040796518325806, + "learning_rate": 3.695952092203854e-05, + "loss": 0.0202, + "step": 60480 + }, + { + "epoch": 3.963178929437201, + "grad_norm": 0.7505953907966614, + "learning_rate": 3.69435591987321e-05, + "loss": 0.0421, + "step": 60490 + }, + { + "epoch": 3.96383410862871, + "grad_norm": 0.1140637993812561, + "learning_rate": 3.6927598903258374e-05, + "loss": 0.0191, + "step": 60500 + }, + { + "epoch": 3.9644892878202187, + "grad_norm": 0.1484244465827942, + "learning_rate": 3.691164003736281e-05, + "loss": 0.0227, + "step": 60510 + }, + { + "epoch": 3.9651444670117275, + "grad_norm": 0.07766954600811005, + "learning_rate": 3.689568260279061e-05, + "loss": 0.0209, + "step": 60520 + }, + { + "epoch": 3.9657996462032363, + "grad_norm": 0.8090190887451172, + "learning_rate": 3.687972660128686e-05, + "loss": 0.0216, + "step": 60530 + }, + { + "epoch": 3.9664548253947456, + "grad_norm": 0.19625528156757355, + "learning_rate": 3.686377203459648e-05, + "loss": 0.023, + "step": 60540 + }, + { + "epoch": 3.9671100045862544, + "grad_norm": 0.3762214481830597, + "learning_rate": 3.6847818904464226e-05, + "loss": 0.0231, + "step": 60550 + }, + { + "epoch": 3.967765183777763, + "grad_norm": 0.9990289211273193, + "learning_rate": 3.683186721263474e-05, + "loss": 0.022, + "step": 60560 + }, + { + "epoch": 3.968420362969272, + "grad_norm": 0.3655785620212555, + "learning_rate": 3.6815916960852435e-05, + "loss": 0.0389, + "step": 60570 + }, + { + "epoch": 3.969075542160781, + "grad_norm": 0.16485317051410675, + "learning_rate": 3.679996815086165e-05, + "loss": 0.0663, + "step": 60580 + }, + { + "epoch": 3.96973072135229, + "grad_norm": 0.602466881275177, + "learning_rate": 3.678402078440649e-05, + "loss": 0.0486, + "step": 60590 + }, + { + "epoch": 3.970385900543799, + "grad_norm": 0.1565733104944229, + "learning_rate": 3.676807486323096e-05, + "loss": 0.0251, + "step": 60600 + }, + { + "epoch": 3.9710410797353077, + "grad_norm": 0.8421227335929871, + "learning_rate": 3.6752130389078864e-05, + "loss": 0.0414, + "step": 60610 + }, + { + "epoch": 3.9716962589268165, + "grad_norm": 0.31992778182029724, + "learning_rate": 3.673618736369388e-05, + "loss": 0.0352, + "step": 60620 + }, + { + "epoch": 3.9723514381183254, + "grad_norm": 0.9303898215293884, + "learning_rate": 3.6720245788819474e-05, + "loss": 0.0293, + "step": 60630 + }, + { + "epoch": 3.973006617309834, + "grad_norm": 0.17162057757377625, + "learning_rate": 3.6704305666199044e-05, + "loss": 0.0274, + "step": 60640 + }, + { + "epoch": 3.973661796501343, + "grad_norm": 0.5104685425758362, + "learning_rate": 3.668836699757576e-05, + "loss": 0.0276, + "step": 60650 + }, + { + "epoch": 3.974316975692852, + "grad_norm": 0.42246201634407043, + "learning_rate": 3.667242978469265e-05, + "loss": 0.0251, + "step": 60660 + }, + { + "epoch": 3.9749721548843606, + "grad_norm": 0.6348928213119507, + "learning_rate": 3.6656494029292575e-05, + "loss": 0.0263, + "step": 60670 + }, + { + "epoch": 3.97562733407587, + "grad_norm": 0.7395981550216675, + "learning_rate": 3.664055973311825e-05, + "loss": 0.0597, + "step": 60680 + }, + { + "epoch": 3.9762825132673787, + "grad_norm": 0.6879955530166626, + "learning_rate": 3.662462689791221e-05, + "loss": 0.0419, + "step": 60690 + }, + { + "epoch": 3.9769376924588875, + "grad_norm": 0.6394168138504028, + "learning_rate": 3.660869552541689e-05, + "loss": 0.0347, + "step": 60700 + }, + { + "epoch": 3.9775928716503963, + "grad_norm": 0.8641514778137207, + "learning_rate": 3.659276561737448e-05, + "loss": 0.0392, + "step": 60710 + }, + { + "epoch": 3.978248050841905, + "grad_norm": 0.5645419955253601, + "learning_rate": 3.657683717552708e-05, + "loss": 0.0253, + "step": 60720 + }, + { + "epoch": 3.9789032300334144, + "grad_norm": 1.3910099267959595, + "learning_rate": 3.6560910201616574e-05, + "loss": 0.041, + "step": 60730 + }, + { + "epoch": 3.979558409224923, + "grad_norm": 1.1603388786315918, + "learning_rate": 3.654498469738472e-05, + "loss": 0.0419, + "step": 60740 + }, + { + "epoch": 3.980213588416432, + "grad_norm": 0.5436002612113953, + "learning_rate": 3.6529060664573126e-05, + "loss": 0.0291, + "step": 60750 + }, + { + "epoch": 3.980868767607941, + "grad_norm": 0.7517393827438354, + "learning_rate": 3.6513138104923176e-05, + "loss": 0.0404, + "step": 60760 + }, + { + "epoch": 3.9815239467994497, + "grad_norm": 0.11641038209199905, + "learning_rate": 3.6497217020176176e-05, + "loss": 0.0253, + "step": 60770 + }, + { + "epoch": 3.9821791259909585, + "grad_norm": 0.1231447160243988, + "learning_rate": 3.648129741207323e-05, + "loss": 0.06, + "step": 60780 + }, + { + "epoch": 3.9828343051824673, + "grad_norm": 0.5038578510284424, + "learning_rate": 3.646537928235527e-05, + "loss": 0.0217, + "step": 60790 + }, + { + "epoch": 3.983489484373976, + "grad_norm": 1.6602742671966553, + "learning_rate": 3.6449462632763075e-05, + "loss": 0.0389, + "step": 60800 + }, + { + "epoch": 3.984144663565485, + "grad_norm": 1.2999110221862793, + "learning_rate": 3.643354746503729e-05, + "loss": 0.0389, + "step": 60810 + }, + { + "epoch": 3.984799842756994, + "grad_norm": 0.5182511806488037, + "learning_rate": 3.6417633780918345e-05, + "loss": 0.0341, + "step": 60820 + }, + { + "epoch": 3.985455021948503, + "grad_norm": 0.3671746253967285, + "learning_rate": 3.640172158214655e-05, + "loss": 0.0289, + "step": 60830 + }, + { + "epoch": 3.986110201140012, + "grad_norm": 0.6256510019302368, + "learning_rate": 3.6385810870462035e-05, + "loss": 0.0224, + "step": 60840 + }, + { + "epoch": 3.9867653803315206, + "grad_norm": 0.899115264415741, + "learning_rate": 3.636990164760479e-05, + "loss": 0.0195, + "step": 60850 + }, + { + "epoch": 3.9874205595230294, + "grad_norm": 0.4162883758544922, + "learning_rate": 3.6353993915314596e-05, + "loss": 0.0235, + "step": 60860 + }, + { + "epoch": 3.9880757387145387, + "grad_norm": 0.26196518540382385, + "learning_rate": 3.6338087675331124e-05, + "loss": 0.0302, + "step": 60870 + }, + { + "epoch": 3.9887309179060475, + "grad_norm": 0.6634824275970459, + "learning_rate": 3.632218292939383e-05, + "loss": 0.0481, + "step": 60880 + }, + { + "epoch": 3.9893860970975563, + "grad_norm": 1.781038522720337, + "learning_rate": 3.6306279679242033e-05, + "loss": 0.0417, + "step": 60890 + }, + { + "epoch": 3.990041276289065, + "grad_norm": 0.07303304970264435, + "learning_rate": 3.629037792661492e-05, + "loss": 0.035, + "step": 60900 + }, + { + "epoch": 3.990696455480574, + "grad_norm": 0.25880002975463867, + "learning_rate": 3.627447767325147e-05, + "loss": 0.0219, + "step": 60910 + }, + { + "epoch": 3.9913516346720828, + "grad_norm": 0.38155144453048706, + "learning_rate": 3.62585789208905e-05, + "loss": 0.0352, + "step": 60920 + }, + { + "epoch": 3.9920068138635916, + "grad_norm": 0.4502703845500946, + "learning_rate": 3.624268167127068e-05, + "loss": 0.0214, + "step": 60930 + }, + { + "epoch": 3.9926619930551004, + "grad_norm": 0.22969628870487213, + "learning_rate": 3.622678592613051e-05, + "loss": 0.0436, + "step": 60940 + }, + { + "epoch": 3.993317172246609, + "grad_norm": 0.15991297364234924, + "learning_rate": 3.6210891687208304e-05, + "loss": 0.0196, + "step": 60950 + }, + { + "epoch": 3.9939723514381185, + "grad_norm": 0.18027418851852417, + "learning_rate": 3.619499895624228e-05, + "loss": 0.0375, + "step": 60960 + }, + { + "epoch": 3.9946275306296273, + "grad_norm": 1.074341058731079, + "learning_rate": 3.61791077349704e-05, + "loss": 0.0212, + "step": 60970 + }, + { + "epoch": 3.995282709821136, + "grad_norm": 0.3387737572193146, + "learning_rate": 3.616321802513053e-05, + "loss": 0.0328, + "step": 60980 + }, + { + "epoch": 3.995937889012645, + "grad_norm": 0.47948595881462097, + "learning_rate": 3.6147329828460325e-05, + "loss": 0.064, + "step": 60990 + }, + { + "epoch": 3.9965930682041537, + "grad_norm": 0.7847294211387634, + "learning_rate": 3.6131443146697306e-05, + "loss": 0.0585, + "step": 61000 + }, + { + "epoch": 3.997248247395663, + "grad_norm": 0.19767212867736816, + "learning_rate": 3.611555798157879e-05, + "loss": 0.0388, + "step": 61010 + }, + { + "epoch": 3.997903426587172, + "grad_norm": 0.1535375714302063, + "learning_rate": 3.6099674334842e-05, + "loss": 0.0411, + "step": 61020 + }, + { + "epoch": 3.9985586057786806, + "grad_norm": 0.343163400888443, + "learning_rate": 3.6083792208223934e-05, + "loss": 0.0247, + "step": 61030 + }, + { + "epoch": 3.9992137849701894, + "grad_norm": 0.40029117465019226, + "learning_rate": 3.606791160346142e-05, + "loss": 0.0291, + "step": 61040 + }, + { + "epoch": 3.9998689641616982, + "grad_norm": 0.6034086346626282, + "learning_rate": 3.6052032522291134e-05, + "loss": 0.0312, + "step": 61050 + }, + { + "epoch": 4.000524143353207, + "grad_norm": 0.1754327118396759, + "learning_rate": 3.603615496644962e-05, + "loss": 0.0342, + "step": 61060 + }, + { + "epoch": 4.001179322544716, + "grad_norm": 0.9235950708389282, + "learning_rate": 3.60202789376732e-05, + "loss": 0.0486, + "step": 61070 + }, + { + "epoch": 4.001834501736225, + "grad_norm": 0.2010568380355835, + "learning_rate": 3.6004404437698036e-05, + "loss": 0.0254, + "step": 61080 + }, + { + "epoch": 4.0024896809277335, + "grad_norm": 0.11763820797204971, + "learning_rate": 3.5988531468260184e-05, + "loss": 0.0492, + "step": 61090 + }, + { + "epoch": 4.003144860119242, + "grad_norm": 0.596427857875824, + "learning_rate": 3.597266003109546e-05, + "loss": 0.0311, + "step": 61100 + }, + { + "epoch": 4.003800039310751, + "grad_norm": 0.17882274091243744, + "learning_rate": 3.5956790127939557e-05, + "loss": 0.0198, + "step": 61110 + }, + { + "epoch": 4.00445521850226, + "grad_norm": 0.4316486120223999, + "learning_rate": 3.594092176052796e-05, + "loss": 0.0147, + "step": 61120 + }, + { + "epoch": 4.00511039769377, + "grad_norm": 0.5402690172195435, + "learning_rate": 3.592505493059603e-05, + "loss": 0.0585, + "step": 61130 + }, + { + "epoch": 4.0057655768852785, + "grad_norm": 0.8039509654045105, + "learning_rate": 3.5909189639878905e-05, + "loss": 0.032, + "step": 61140 + }, + { + "epoch": 4.006420756076787, + "grad_norm": 0.2312329113483429, + "learning_rate": 3.5893325890111626e-05, + "loss": 0.034, + "step": 61150 + }, + { + "epoch": 4.007075935268296, + "grad_norm": 0.7293518781661987, + "learning_rate": 3.5877463683029034e-05, + "loss": 0.0549, + "step": 61160 + }, + { + "epoch": 4.007731114459805, + "grad_norm": 0.22212357819080353, + "learning_rate": 3.586160302036578e-05, + "loss": 0.0588, + "step": 61170 + }, + { + "epoch": 4.008386293651314, + "grad_norm": 0.9518961310386658, + "learning_rate": 3.584574390385636e-05, + "loss": 0.045, + "step": 61180 + }, + { + "epoch": 4.0090414728428225, + "grad_norm": 0.19942927360534668, + "learning_rate": 3.582988633523511e-05, + "loss": 0.0549, + "step": 61190 + }, + { + "epoch": 4.009696652034331, + "grad_norm": 0.42250874638557434, + "learning_rate": 3.581403031623616e-05, + "loss": 0.028, + "step": 61200 + }, + { + "epoch": 4.01035183122584, + "grad_norm": 0.43710070848464966, + "learning_rate": 3.579817584859354e-05, + "loss": 0.0356, + "step": 61210 + }, + { + "epoch": 4.011007010417349, + "grad_norm": 0.5097378492355347, + "learning_rate": 3.578232293404107e-05, + "loss": 0.0653, + "step": 61220 + }, + { + "epoch": 4.011662189608858, + "grad_norm": 0.26225924491882324, + "learning_rate": 3.5766471574312385e-05, + "loss": 0.0211, + "step": 61230 + }, + { + "epoch": 4.012317368800367, + "grad_norm": 0.4034614562988281, + "learning_rate": 3.5750621771140964e-05, + "loss": 0.0289, + "step": 61240 + }, + { + "epoch": 4.012972547991875, + "grad_norm": 0.14217223227024078, + "learning_rate": 3.5734773526260115e-05, + "loss": 0.0325, + "step": 61250 + }, + { + "epoch": 4.013627727183384, + "grad_norm": 0.4159696400165558, + "learning_rate": 3.5718926841402996e-05, + "loss": 0.0322, + "step": 61260 + }, + { + "epoch": 4.014282906374894, + "grad_norm": 0.23206128180027008, + "learning_rate": 3.5703081718302555e-05, + "loss": 0.0178, + "step": 61270 + }, + { + "epoch": 4.014938085566403, + "grad_norm": 0.11154823005199432, + "learning_rate": 3.568723815869162e-05, + "loss": 0.0141, + "step": 61280 + }, + { + "epoch": 4.015593264757912, + "grad_norm": 0.22084319591522217, + "learning_rate": 3.5671396164302777e-05, + "loss": 0.0168, + "step": 61290 + }, + { + "epoch": 4.01624844394942, + "grad_norm": 0.17641229927539825, + "learning_rate": 3.565555573686853e-05, + "loss": 0.0445, + "step": 61300 + }, + { + "epoch": 4.016903623140929, + "grad_norm": 0.0796196386218071, + "learning_rate": 3.563971687812113e-05, + "loss": 0.0493, + "step": 61310 + }, + { + "epoch": 4.017558802332438, + "grad_norm": 0.7251455187797546, + "learning_rate": 3.5623879589792706e-05, + "loss": 0.0338, + "step": 61320 + }, + { + "epoch": 4.018213981523947, + "grad_norm": 0.13402824103832245, + "learning_rate": 3.5608043873615185e-05, + "loss": 0.0249, + "step": 61330 + }, + { + "epoch": 4.018869160715456, + "grad_norm": 0.1717602014541626, + "learning_rate": 3.559220973132035e-05, + "loss": 0.0282, + "step": 61340 + }, + { + "epoch": 4.0195243399069644, + "grad_norm": 0.47256648540496826, + "learning_rate": 3.557637716463981e-05, + "loss": 0.0758, + "step": 61350 + }, + { + "epoch": 4.020179519098473, + "grad_norm": 0.29846516251564026, + "learning_rate": 3.556054617530497e-05, + "loss": 0.0332, + "step": 61360 + }, + { + "epoch": 4.020834698289982, + "grad_norm": 0.4753911793231964, + "learning_rate": 3.5544716765047106e-05, + "loss": 0.0365, + "step": 61370 + }, + { + "epoch": 4.021489877481491, + "grad_norm": 0.47348499298095703, + "learning_rate": 3.5528888935597274e-05, + "loss": 0.0321, + "step": 61380 + }, + { + "epoch": 4.022145056673, + "grad_norm": 0.7331682443618774, + "learning_rate": 3.551306268868636e-05, + "loss": 0.0349, + "step": 61390 + }, + { + "epoch": 4.0228002358645085, + "grad_norm": 0.21780124306678772, + "learning_rate": 3.5497238026045166e-05, + "loss": 0.0257, + "step": 61400 + }, + { + "epoch": 4.023455415056018, + "grad_norm": 0.37374719977378845, + "learning_rate": 3.548141494940421e-05, + "loss": 0.0214, + "step": 61410 + }, + { + "epoch": 4.024110594247527, + "grad_norm": 0.3257310688495636, + "learning_rate": 3.546559346049388e-05, + "loss": 0.0366, + "step": 61420 + }, + { + "epoch": 4.024765773439036, + "grad_norm": 0.11263278126716614, + "learning_rate": 3.544977356104441e-05, + "loss": 0.0249, + "step": 61430 + }, + { + "epoch": 4.025420952630545, + "grad_norm": 0.6603924632072449, + "learning_rate": 3.543395525278582e-05, + "loss": 0.0372, + "step": 61440 + }, + { + "epoch": 4.0260761318220535, + "grad_norm": 0.1651131957769394, + "learning_rate": 3.5418138537447985e-05, + "loss": 0.0361, + "step": 61450 + }, + { + "epoch": 4.026731311013562, + "grad_norm": 0.38225317001342773, + "learning_rate": 3.5402323416760574e-05, + "loss": 0.0507, + "step": 61460 + }, + { + "epoch": 4.027386490205071, + "grad_norm": 0.28028932213783264, + "learning_rate": 3.538650989245316e-05, + "loss": 0.0213, + "step": 61470 + }, + { + "epoch": 4.02804166939658, + "grad_norm": 0.11068711429834366, + "learning_rate": 3.537069796625504e-05, + "loss": 0.0216, + "step": 61480 + }, + { + "epoch": 4.028696848588089, + "grad_norm": 0.5893254280090332, + "learning_rate": 3.53548876398954e-05, + "loss": 0.0431, + "step": 61490 + }, + { + "epoch": 4.029352027779598, + "grad_norm": 0.3422461748123169, + "learning_rate": 3.5339078915103214e-05, + "loss": 0.0445, + "step": 61500 + }, + { + "epoch": 4.030007206971106, + "grad_norm": 0.39015671610832214, + "learning_rate": 3.532327179360733e-05, + "loss": 0.0305, + "step": 61510 + }, + { + "epoch": 4.030662386162615, + "grad_norm": 0.18569806218147278, + "learning_rate": 3.530746627713636e-05, + "loss": 0.0219, + "step": 61520 + }, + { + "epoch": 4.031317565354124, + "grad_norm": 0.16505511105060577, + "learning_rate": 3.5291662367418805e-05, + "loss": 0.02, + "step": 61530 + }, + { + "epoch": 4.031972744545633, + "grad_norm": 0.13114914298057556, + "learning_rate": 3.527586006618293e-05, + "loss": 0.0251, + "step": 61540 + }, + { + "epoch": 4.0326279237371425, + "grad_norm": 0.24210640788078308, + "learning_rate": 3.526005937515687e-05, + "loss": 0.0199, + "step": 61550 + }, + { + "epoch": 4.033283102928651, + "grad_norm": 0.5345771312713623, + "learning_rate": 3.524426029606856e-05, + "loss": 0.0397, + "step": 61560 + }, + { + "epoch": 4.03393828212016, + "grad_norm": 0.16310586035251617, + "learning_rate": 3.5228462830645746e-05, + "loss": 0.0363, + "step": 61570 + }, + { + "epoch": 4.034593461311669, + "grad_norm": 1.1673531532287598, + "learning_rate": 3.5212666980616036e-05, + "loss": 0.0371, + "step": 61580 + }, + { + "epoch": 4.035248640503178, + "grad_norm": 0.47335419058799744, + "learning_rate": 3.519687274770681e-05, + "loss": 0.025, + "step": 61590 + }, + { + "epoch": 4.035903819694687, + "grad_norm": 0.1929989755153656, + "learning_rate": 3.518108013364535e-05, + "loss": 0.0267, + "step": 61600 + }, + { + "epoch": 4.036558998886195, + "grad_norm": 0.06841916590929031, + "learning_rate": 3.516528914015868e-05, + "loss": 0.0144, + "step": 61610 + }, + { + "epoch": 4.037214178077704, + "grad_norm": 0.7654358744621277, + "learning_rate": 3.5149499768973694e-05, + "loss": 0.0278, + "step": 61620 + }, + { + "epoch": 4.037869357269213, + "grad_norm": 0.11803256720304489, + "learning_rate": 3.513371202181708e-05, + "loss": 0.0133, + "step": 61630 + }, + { + "epoch": 4.038524536460722, + "grad_norm": 0.38520005345344543, + "learning_rate": 3.511792590041537e-05, + "loss": 0.02, + "step": 61640 + }, + { + "epoch": 4.039179715652231, + "grad_norm": 0.8684030175209045, + "learning_rate": 3.510214140649488e-05, + "loss": 0.0429, + "step": 61650 + }, + { + "epoch": 4.0398348948437395, + "grad_norm": 0.5021288990974426, + "learning_rate": 3.508635854178184e-05, + "loss": 0.0269, + "step": 61660 + }, + { + "epoch": 4.040490074035248, + "grad_norm": 1.1217721700668335, + "learning_rate": 3.50705773080022e-05, + "loss": 0.0575, + "step": 61670 + }, + { + "epoch": 4.041145253226757, + "grad_norm": 0.4418889880180359, + "learning_rate": 3.505479770688178e-05, + "loss": 0.0392, + "step": 61680 + }, + { + "epoch": 4.041800432418267, + "grad_norm": 0.6823380589485168, + "learning_rate": 3.503901974014622e-05, + "loss": 0.0441, + "step": 61690 + }, + { + "epoch": 4.042455611609776, + "grad_norm": 1.2448673248291016, + "learning_rate": 3.502324340952095e-05, + "loss": 0.0574, + "step": 61700 + }, + { + "epoch": 4.043110790801284, + "grad_norm": 0.20072460174560547, + "learning_rate": 3.500746871673126e-05, + "loss": 0.0666, + "step": 61710 + }, + { + "epoch": 4.043765969992793, + "grad_norm": 0.8753123879432678, + "learning_rate": 3.499169566350227e-05, + "loss": 0.0359, + "step": 61720 + }, + { + "epoch": 4.044421149184302, + "grad_norm": 0.1264878660440445, + "learning_rate": 3.4975924251558876e-05, + "loss": 0.03, + "step": 61730 + }, + { + "epoch": 4.045076328375811, + "grad_norm": 0.12247342616319656, + "learning_rate": 3.496015448262582e-05, + "loss": 0.0397, + "step": 61740 + }, + { + "epoch": 4.04573150756732, + "grad_norm": 0.2100069522857666, + "learning_rate": 3.494438635842764e-05, + "loss": 0.026, + "step": 61750 + }, + { + "epoch": 4.0463866867588285, + "grad_norm": 0.8867641687393188, + "learning_rate": 3.492861988068874e-05, + "loss": 0.0298, + "step": 61760 + }, + { + "epoch": 4.047041865950337, + "grad_norm": 0.14484162628650665, + "learning_rate": 3.491285505113332e-05, + "loss": 0.0723, + "step": 61770 + }, + { + "epoch": 4.047697045141846, + "grad_norm": 0.4021467864513397, + "learning_rate": 3.4897091871485376e-05, + "loss": 0.035, + "step": 61780 + }, + { + "epoch": 4.048352224333355, + "grad_norm": 0.327938050031662, + "learning_rate": 3.488133034346875e-05, + "loss": 0.0499, + "step": 61790 + }, + { + "epoch": 4.049007403524864, + "grad_norm": 0.41986027359962463, + "learning_rate": 3.486557046880713e-05, + "loss": 0.0185, + "step": 61800 + }, + { + "epoch": 4.049662582716373, + "grad_norm": 0.9363787174224854, + "learning_rate": 3.484981224922395e-05, + "loss": 0.0319, + "step": 61810 + }, + { + "epoch": 4.050317761907881, + "grad_norm": 0.4899941682815552, + "learning_rate": 3.4834055686442544e-05, + "loss": 0.0239, + "step": 61820 + }, + { + "epoch": 4.050972941099391, + "grad_norm": 0.44776344299316406, + "learning_rate": 3.4818300782186e-05, + "loss": 0.0276, + "step": 61830 + }, + { + "epoch": 4.0516281202909, + "grad_norm": 0.26000866293907166, + "learning_rate": 3.4802547538177236e-05, + "loss": 0.058, + "step": 61840 + }, + { + "epoch": 4.052283299482409, + "grad_norm": 0.1697220802307129, + "learning_rate": 3.478679595613905e-05, + "loss": 0.0293, + "step": 61850 + }, + { + "epoch": 4.0529384786739175, + "grad_norm": 0.6499163508415222, + "learning_rate": 3.4771046037793975e-05, + "loss": 0.0124, + "step": 61860 + }, + { + "epoch": 4.053593657865426, + "grad_norm": 0.5408393740653992, + "learning_rate": 3.4755297784864416e-05, + "loss": 0.0186, + "step": 61870 + }, + { + "epoch": 4.054248837056935, + "grad_norm": 0.07204820960760117, + "learning_rate": 3.473955119907257e-05, + "loss": 0.0306, + "step": 61880 + }, + { + "epoch": 4.054904016248444, + "grad_norm": 0.4087597727775574, + "learning_rate": 3.472380628214046e-05, + "loss": 0.0325, + "step": 61890 + }, + { + "epoch": 4.055559195439953, + "grad_norm": 0.15758264064788818, + "learning_rate": 3.4708063035789906e-05, + "loss": 0.0299, + "step": 61900 + }, + { + "epoch": 4.056214374631462, + "grad_norm": 0.4350307881832123, + "learning_rate": 3.4692321461742614e-05, + "loss": 0.0487, + "step": 61910 + }, + { + "epoch": 4.05686955382297, + "grad_norm": 0.8714090585708618, + "learning_rate": 3.4676581561720023e-05, + "loss": 0.0433, + "step": 61920 + }, + { + "epoch": 4.057524733014479, + "grad_norm": 0.5430406928062439, + "learning_rate": 3.466084333744344e-05, + "loss": 0.0487, + "step": 61930 + }, + { + "epoch": 4.058179912205988, + "grad_norm": 0.28481152653694153, + "learning_rate": 3.4645106790633964e-05, + "loss": 0.0385, + "step": 61940 + }, + { + "epoch": 4.058835091397497, + "grad_norm": 0.6909318566322327, + "learning_rate": 3.4629371923012526e-05, + "loss": 0.0248, + "step": 61950 + }, + { + "epoch": 4.059490270589006, + "grad_norm": 1.0284786224365234, + "learning_rate": 3.4613638736299856e-05, + "loss": 0.0338, + "step": 61960 + }, + { + "epoch": 4.060145449780515, + "grad_norm": 0.3806392550468445, + "learning_rate": 3.459790723221652e-05, + "loss": 0.0388, + "step": 61970 + }, + { + "epoch": 4.060800628972024, + "grad_norm": 0.2338334172964096, + "learning_rate": 3.45821774124829e-05, + "loss": 0.0178, + "step": 61980 + }, + { + "epoch": 4.061455808163533, + "grad_norm": 0.22723153233528137, + "learning_rate": 3.456644927881918e-05, + "loss": 0.0298, + "step": 61990 + }, + { + "epoch": 4.062110987355042, + "grad_norm": 0.3413201570510864, + "learning_rate": 3.4550722832945354e-05, + "loss": 0.0407, + "step": 62000 + }, + { + "epoch": 4.062766166546551, + "grad_norm": 0.15267367660999298, + "learning_rate": 3.453499807658126e-05, + "loss": 0.0429, + "step": 62010 + }, + { + "epoch": 4.0634213457380595, + "grad_norm": 0.9268825054168701, + "learning_rate": 3.451927501144653e-05, + "loss": 0.042, + "step": 62020 + }, + { + "epoch": 4.064076524929568, + "grad_norm": 0.22595037519931793, + "learning_rate": 3.450355363926059e-05, + "loss": 0.0204, + "step": 62030 + }, + { + "epoch": 4.064731704121077, + "grad_norm": 1.3790684938430786, + "learning_rate": 3.448783396174273e-05, + "loss": 0.0301, + "step": 62040 + }, + { + "epoch": 4.065386883312586, + "grad_norm": 0.6941514015197754, + "learning_rate": 3.4472115980612034e-05, + "loss": 0.0364, + "step": 62050 + }, + { + "epoch": 4.066042062504095, + "grad_norm": 0.7246835231781006, + "learning_rate": 3.4456399697587396e-05, + "loss": 0.03, + "step": 62060 + }, + { + "epoch": 4.0666972416956035, + "grad_norm": 0.6434703469276428, + "learning_rate": 3.4440685114387506e-05, + "loss": 0.0199, + "step": 62070 + }, + { + "epoch": 4.067352420887112, + "grad_norm": 0.3957407474517822, + "learning_rate": 3.442497223273091e-05, + "loss": 0.0214, + "step": 62080 + }, + { + "epoch": 4.068007600078621, + "grad_norm": 0.569155752658844, + "learning_rate": 3.4409261054335926e-05, + "loss": 0.0333, + "step": 62090 + }, + { + "epoch": 4.06866277927013, + "grad_norm": 0.5479517579078674, + "learning_rate": 3.439355158092069e-05, + "loss": 0.0307, + "step": 62100 + }, + { + "epoch": 4.06931795846164, + "grad_norm": 0.6352693438529968, + "learning_rate": 3.437784381420322e-05, + "loss": 0.0375, + "step": 62110 + }, + { + "epoch": 4.0699731376531485, + "grad_norm": 1.0302172899246216, + "learning_rate": 3.436213775590126e-05, + "loss": 0.0284, + "step": 62120 + }, + { + "epoch": 4.070628316844657, + "grad_norm": 0.4025029242038727, + "learning_rate": 3.4346433407732405e-05, + "loss": 0.0291, + "step": 62130 + }, + { + "epoch": 4.071283496036166, + "grad_norm": 1.6073334217071533, + "learning_rate": 3.433073077141405e-05, + "loss": 0.0328, + "step": 62140 + }, + { + "epoch": 4.071938675227675, + "grad_norm": 0.16912876069545746, + "learning_rate": 3.431502984866343e-05, + "loss": 0.0273, + "step": 62150 + }, + { + "epoch": 4.072593854419184, + "grad_norm": 0.18465422093868256, + "learning_rate": 3.429933064119753e-05, + "loss": 0.0332, + "step": 62160 + }, + { + "epoch": 4.073249033610693, + "grad_norm": 0.2087400108575821, + "learning_rate": 3.4283633150733255e-05, + "loss": 0.0253, + "step": 62170 + }, + { + "epoch": 4.073904212802201, + "grad_norm": 0.36176159977912903, + "learning_rate": 3.4267937378987225e-05, + "loss": 0.0279, + "step": 62180 + }, + { + "epoch": 4.07455939199371, + "grad_norm": 0.23979206383228302, + "learning_rate": 3.425224332767592e-05, + "loss": 0.0289, + "step": 62190 + }, + { + "epoch": 4.075214571185219, + "grad_norm": 0.15725259482860565, + "learning_rate": 3.4236550998515595e-05, + "loss": 0.0227, + "step": 62200 + }, + { + "epoch": 4.075869750376728, + "grad_norm": 1.4770821332931519, + "learning_rate": 3.422086039322235e-05, + "loss": 0.037, + "step": 62210 + }, + { + "epoch": 4.076524929568237, + "grad_norm": 0.5184779167175293, + "learning_rate": 3.420517151351209e-05, + "loss": 0.0493, + "step": 62220 + }, + { + "epoch": 4.0771801087597455, + "grad_norm": 0.32775452733039856, + "learning_rate": 3.418948436110053e-05, + "loss": 0.0295, + "step": 62230 + }, + { + "epoch": 4.077835287951254, + "grad_norm": 0.5398567914962769, + "learning_rate": 3.417379893770319e-05, + "loss": 0.0275, + "step": 62240 + }, + { + "epoch": 4.078490467142764, + "grad_norm": 0.08791238814592361, + "learning_rate": 3.41581152450354e-05, + "loss": 0.0395, + "step": 62250 + }, + { + "epoch": 4.079145646334273, + "grad_norm": 0.18404467403888702, + "learning_rate": 3.414243328481231e-05, + "loss": 0.0214, + "step": 62260 + }, + { + "epoch": 4.079800825525782, + "grad_norm": 0.9423092007637024, + "learning_rate": 3.412675305874888e-05, + "loss": 0.0202, + "step": 62270 + }, + { + "epoch": 4.08045600471729, + "grad_norm": 0.7264878153800964, + "learning_rate": 3.4111074568559876e-05, + "loss": 0.0615, + "step": 62280 + }, + { + "epoch": 4.081111183908799, + "grad_norm": 0.5167020559310913, + "learning_rate": 3.4095397815959846e-05, + "loss": 0.0452, + "step": 62290 + }, + { + "epoch": 4.081766363100308, + "grad_norm": 0.4483635127544403, + "learning_rate": 3.407972280266322e-05, + "loss": 0.0448, + "step": 62300 + }, + { + "epoch": 4.082421542291817, + "grad_norm": 0.6121023297309875, + "learning_rate": 3.4064049530384166e-05, + "loss": 0.0614, + "step": 62310 + }, + { + "epoch": 4.083076721483326, + "grad_norm": 0.5425028800964355, + "learning_rate": 3.40483780008367e-05, + "loss": 0.0416, + "step": 62320 + }, + { + "epoch": 4.0837319006748345, + "grad_norm": 1.4437681436538696, + "learning_rate": 3.4032708215734635e-05, + "loss": 0.0418, + "step": 62330 + }, + { + "epoch": 4.084387079866343, + "grad_norm": 0.2083905190229416, + "learning_rate": 3.4017040176791584e-05, + "loss": 0.0241, + "step": 62340 + }, + { + "epoch": 4.085042259057852, + "grad_norm": 0.8945692181587219, + "learning_rate": 3.4001373885720976e-05, + "loss": 0.0415, + "step": 62350 + }, + { + "epoch": 4.085697438249361, + "grad_norm": 0.8508671522140503, + "learning_rate": 3.398570934423608e-05, + "loss": 0.0195, + "step": 62360 + }, + { + "epoch": 4.08635261744087, + "grad_norm": 0.12356555461883545, + "learning_rate": 3.397004655404994e-05, + "loss": 0.0328, + "step": 62370 + }, + { + "epoch": 4.087007796632379, + "grad_norm": 0.08921181410551071, + "learning_rate": 3.395438551687541e-05, + "loss": 0.0178, + "step": 62380 + }, + { + "epoch": 4.087662975823888, + "grad_norm": 0.6096433401107788, + "learning_rate": 3.393872623442515e-05, + "loss": 0.0244, + "step": 62390 + }, + { + "epoch": 4.088318155015397, + "grad_norm": 0.19384866952896118, + "learning_rate": 3.392306870841164e-05, + "loss": 0.0284, + "step": 62400 + }, + { + "epoch": 4.088973334206906, + "grad_norm": 0.6369197368621826, + "learning_rate": 3.390741294054715e-05, + "loss": 0.0348, + "step": 62410 + }, + { + "epoch": 4.089628513398415, + "grad_norm": 0.19120700657367706, + "learning_rate": 3.38917589325438e-05, + "loss": 0.019, + "step": 62420 + }, + { + "epoch": 4.0902836925899235, + "grad_norm": 0.568400502204895, + "learning_rate": 3.387610668611349e-05, + "loss": 0.0369, + "step": 62430 + }, + { + "epoch": 4.090938871781432, + "grad_norm": 0.40913230180740356, + "learning_rate": 3.38604562029679e-05, + "loss": 0.0352, + "step": 62440 + }, + { + "epoch": 4.091594050972941, + "grad_norm": 0.32786330580711365, + "learning_rate": 3.384480748481855e-05, + "loss": 0.0487, + "step": 62450 + }, + { + "epoch": 4.09224923016445, + "grad_norm": 0.3095281422138214, + "learning_rate": 3.382916053337676e-05, + "loss": 0.0398, + "step": 62460 + }, + { + "epoch": 4.092904409355959, + "grad_norm": 0.4072973132133484, + "learning_rate": 3.381351535035366e-05, + "loss": 0.0343, + "step": 62470 + }, + { + "epoch": 4.093559588547468, + "grad_norm": 0.801268994808197, + "learning_rate": 3.379787193746019e-05, + "loss": 0.0259, + "step": 62480 + }, + { + "epoch": 4.094214767738976, + "grad_norm": 0.44304516911506653, + "learning_rate": 3.378223029640709e-05, + "loss": 0.0306, + "step": 62490 + }, + { + "epoch": 4.094869946930485, + "grad_norm": 0.21751703321933746, + "learning_rate": 3.376659042890489e-05, + "loss": 0.0345, + "step": 62500 + }, + { + "epoch": 4.095525126121994, + "grad_norm": 1.3246276378631592, + "learning_rate": 3.375095233666397e-05, + "loss": 0.0489, + "step": 62510 + }, + { + "epoch": 4.096180305313503, + "grad_norm": 0.6078009009361267, + "learning_rate": 3.373531602139447e-05, + "loss": 0.0277, + "step": 62520 + }, + { + "epoch": 4.096835484505013, + "grad_norm": 0.16708099842071533, + "learning_rate": 3.371968148480635e-05, + "loss": 0.0278, + "step": 62530 + }, + { + "epoch": 4.097490663696521, + "grad_norm": 0.7474915385246277, + "learning_rate": 3.370404872860937e-05, + "loss": 0.0268, + "step": 62540 + }, + { + "epoch": 4.09814584288803, + "grad_norm": 0.4480190575122833, + "learning_rate": 3.3688417754513145e-05, + "loss": 0.027, + "step": 62550 + }, + { + "epoch": 4.098801022079539, + "grad_norm": 0.10562942922115326, + "learning_rate": 3.367278856422703e-05, + "loss": 0.0267, + "step": 62560 + }, + { + "epoch": 4.099456201271048, + "grad_norm": 0.7528059482574463, + "learning_rate": 3.365716115946021e-05, + "loss": 0.0414, + "step": 62570 + }, + { + "epoch": 4.100111380462557, + "grad_norm": 0.39124658703804016, + "learning_rate": 3.3641535541921666e-05, + "loss": 0.0204, + "step": 62580 + }, + { + "epoch": 4.1007665596540654, + "grad_norm": 0.34115511178970337, + "learning_rate": 3.36259117133202e-05, + "loss": 0.0415, + "step": 62590 + }, + { + "epoch": 4.101421738845574, + "grad_norm": 0.1430283933877945, + "learning_rate": 3.36102896753644e-05, + "loss": 0.0284, + "step": 62600 + }, + { + "epoch": 4.102076918037083, + "grad_norm": 0.21581995487213135, + "learning_rate": 3.35946694297627e-05, + "loss": 0.03, + "step": 62610 + }, + { + "epoch": 4.102732097228592, + "grad_norm": 0.36716413497924805, + "learning_rate": 3.357905097822328e-05, + "loss": 0.036, + "step": 62620 + }, + { + "epoch": 4.103387276420101, + "grad_norm": 0.2652049660682678, + "learning_rate": 3.356343432245416e-05, + "loss": 0.0171, + "step": 62630 + }, + { + "epoch": 4.1040424556116095, + "grad_norm": 0.6040903925895691, + "learning_rate": 3.354781946416315e-05, + "loss": 0.0545, + "step": 62640 + }, + { + "epoch": 4.104697634803118, + "grad_norm": 1.124964714050293, + "learning_rate": 3.3532206405057864e-05, + "loss": 0.028, + "step": 62650 + }, + { + "epoch": 4.105352813994627, + "grad_norm": 0.33986222743988037, + "learning_rate": 3.351659514684573e-05, + "loss": 0.0288, + "step": 62660 + }, + { + "epoch": 4.106007993186136, + "grad_norm": 0.4613879919052124, + "learning_rate": 3.350098569123395e-05, + "loss": 0.0364, + "step": 62670 + }, + { + "epoch": 4.106663172377646, + "grad_norm": 0.49391844868659973, + "learning_rate": 3.348537803992958e-05, + "loss": 0.0285, + "step": 62680 + }, + { + "epoch": 4.1073183515691545, + "grad_norm": 0.3275890350341797, + "learning_rate": 3.346977219463945e-05, + "loss": 0.0177, + "step": 62690 + }, + { + "epoch": 4.107973530760663, + "grad_norm": 0.25007322430610657, + "learning_rate": 3.345416815707017e-05, + "loss": 0.0684, + "step": 62700 + }, + { + "epoch": 4.108628709952172, + "grad_norm": 0.19383622705936432, + "learning_rate": 3.3438565928928185e-05, + "loss": 0.0198, + "step": 62710 + }, + { + "epoch": 4.109283889143681, + "grad_norm": 0.12175212800502777, + "learning_rate": 3.3422965511919736e-05, + "loss": 0.0257, + "step": 62720 + }, + { + "epoch": 4.10993906833519, + "grad_norm": 0.7735076546669006, + "learning_rate": 3.3407366907750846e-05, + "loss": 0.0345, + "step": 62730 + }, + { + "epoch": 4.1105942475266986, + "grad_norm": 0.3260372281074524, + "learning_rate": 3.339177011812737e-05, + "loss": 0.0311, + "step": 62740 + }, + { + "epoch": 4.111249426718207, + "grad_norm": 0.7689235806465149, + "learning_rate": 3.3376175144754935e-05, + "loss": 0.0376, + "step": 62750 + }, + { + "epoch": 4.111904605909716, + "grad_norm": 0.21786822378635406, + "learning_rate": 3.3360581989339004e-05, + "loss": 0.0347, + "step": 62760 + }, + { + "epoch": 4.112559785101225, + "grad_norm": 0.13949066400527954, + "learning_rate": 3.3344990653584804e-05, + "loss": 0.0177, + "step": 62770 + }, + { + "epoch": 4.113214964292734, + "grad_norm": 0.40168923139572144, + "learning_rate": 3.332940113919739e-05, + "loss": 0.0461, + "step": 62780 + }, + { + "epoch": 4.113870143484243, + "grad_norm": 1.2539387941360474, + "learning_rate": 3.331381344788161e-05, + "loss": 0.0452, + "step": 62790 + }, + { + "epoch": 4.114525322675751, + "grad_norm": 0.6860508322715759, + "learning_rate": 3.3298227581342076e-05, + "loss": 0.03, + "step": 62800 + }, + { + "epoch": 4.115180501867261, + "grad_norm": 0.38129350543022156, + "learning_rate": 3.328264354128329e-05, + "loss": 0.0431, + "step": 62810 + }, + { + "epoch": 4.11583568105877, + "grad_norm": 0.42740947008132935, + "learning_rate": 3.3267061329409465e-05, + "loss": 0.0335, + "step": 62820 + }, + { + "epoch": 4.116490860250279, + "grad_norm": 0.2985711693763733, + "learning_rate": 3.325148094742466e-05, + "loss": 0.0538, + "step": 62830 + }, + { + "epoch": 4.117146039441788, + "grad_norm": 0.10142099857330322, + "learning_rate": 3.323590239703271e-05, + "loss": 0.0339, + "step": 62840 + }, + { + "epoch": 4.117801218633296, + "grad_norm": 0.7306346297264099, + "learning_rate": 3.322032567993728e-05, + "loss": 0.0341, + "step": 62850 + }, + { + "epoch": 4.118456397824805, + "grad_norm": 0.20502403378486633, + "learning_rate": 3.3204750797841774e-05, + "loss": 0.0191, + "step": 62860 + }, + { + "epoch": 4.119111577016314, + "grad_norm": 0.5548287034034729, + "learning_rate": 3.31891777524495e-05, + "loss": 0.0349, + "step": 62870 + }, + { + "epoch": 4.119766756207823, + "grad_norm": 0.1892319917678833, + "learning_rate": 3.317360654546347e-05, + "loss": 0.0228, + "step": 62880 + }, + { + "epoch": 4.120421935399332, + "grad_norm": 0.2860310971736908, + "learning_rate": 3.3158037178586534e-05, + "loss": 0.0296, + "step": 62890 + }, + { + "epoch": 4.1210771145908405, + "grad_norm": 0.8054752349853516, + "learning_rate": 3.3142469653521316e-05, + "loss": 0.0425, + "step": 62900 + }, + { + "epoch": 4.121732293782349, + "grad_norm": 0.3024161159992218, + "learning_rate": 3.3126903971970284e-05, + "loss": 0.0222, + "step": 62910 + }, + { + "epoch": 4.122387472973858, + "grad_norm": 0.37320515513420105, + "learning_rate": 3.311134013563564e-05, + "loss": 0.0635, + "step": 62920 + }, + { + "epoch": 4.123042652165367, + "grad_norm": 0.2726922631263733, + "learning_rate": 3.309577814621947e-05, + "loss": 0.0394, + "step": 62930 + }, + { + "epoch": 4.123697831356876, + "grad_norm": 0.5093829035758972, + "learning_rate": 3.308021800542359e-05, + "loss": 0.0267, + "step": 62940 + }, + { + "epoch": 4.1243530105483845, + "grad_norm": 0.49026256799697876, + "learning_rate": 3.306465971494963e-05, + "loss": 0.0231, + "step": 62950 + }, + { + "epoch": 4.125008189739894, + "grad_norm": 0.09407589584589005, + "learning_rate": 3.304910327649901e-05, + "loss": 0.0283, + "step": 62960 + }, + { + "epoch": 4.125663368931403, + "grad_norm": 0.29148975014686584, + "learning_rate": 3.303354869177297e-05, + "loss": 0.0455, + "step": 62970 + }, + { + "epoch": 4.126318548122912, + "grad_norm": 0.4959986209869385, + "learning_rate": 3.301799596247255e-05, + "loss": 0.026, + "step": 62980 + }, + { + "epoch": 4.126973727314421, + "grad_norm": 0.3131123185157776, + "learning_rate": 3.300244509029854e-05, + "loss": 0.0308, + "step": 62990 + }, + { + "epoch": 4.1276289065059295, + "grad_norm": 0.47089964151382446, + "learning_rate": 3.2986896076951585e-05, + "loss": 0.0434, + "step": 63000 + }, + { + "epoch": 4.128284085697438, + "grad_norm": 0.1202099621295929, + "learning_rate": 3.297134892413211e-05, + "loss": 0.0431, + "step": 63010 + }, + { + "epoch": 4.128939264888947, + "grad_norm": 0.6789505481719971, + "learning_rate": 3.295580363354031e-05, + "loss": 0.0228, + "step": 63020 + }, + { + "epoch": 4.129594444080456, + "grad_norm": 0.7390956282615662, + "learning_rate": 3.2940260206876194e-05, + "loss": 0.0259, + "step": 63030 + }, + { + "epoch": 4.130249623271965, + "grad_norm": 0.6132130026817322, + "learning_rate": 3.292471864583958e-05, + "loss": 0.035, + "step": 63040 + }, + { + "epoch": 4.130904802463474, + "grad_norm": 0.22349205613136292, + "learning_rate": 3.290917895213004e-05, + "loss": 0.0242, + "step": 63050 + }, + { + "epoch": 4.131559981654982, + "grad_norm": 0.39744457602500916, + "learning_rate": 3.289364112744701e-05, + "loss": 0.0229, + "step": 63060 + }, + { + "epoch": 4.132215160846491, + "grad_norm": 0.4125292897224426, + "learning_rate": 3.287810517348967e-05, + "loss": 0.0554, + "step": 63070 + }, + { + "epoch": 4.132870340038, + "grad_norm": 0.5486242771148682, + "learning_rate": 3.2862571091957014e-05, + "loss": 0.0286, + "step": 63080 + }, + { + "epoch": 4.13352551922951, + "grad_norm": 0.6769044995307922, + "learning_rate": 3.284703888454781e-05, + "loss": 0.0466, + "step": 63090 + }, + { + "epoch": 4.1341806984210185, + "grad_norm": 0.22418181598186493, + "learning_rate": 3.2831508552960646e-05, + "loss": 0.0339, + "step": 63100 + }, + { + "epoch": 4.134835877612527, + "grad_norm": 0.6475116610527039, + "learning_rate": 3.281598009889387e-05, + "loss": 0.0319, + "step": 63110 + }, + { + "epoch": 4.135491056804036, + "grad_norm": 0.23966680467128754, + "learning_rate": 3.28004535240457e-05, + "loss": 0.0174, + "step": 63120 + }, + { + "epoch": 4.136146235995545, + "grad_norm": 0.12910157442092896, + "learning_rate": 3.278492883011407e-05, + "loss": 0.0332, + "step": 63130 + }, + { + "epoch": 4.136801415187054, + "grad_norm": 0.5355740189552307, + "learning_rate": 3.276940601879675e-05, + "loss": 0.0529, + "step": 63140 + }, + { + "epoch": 4.137456594378563, + "grad_norm": 0.6091853380203247, + "learning_rate": 3.275388509179127e-05, + "loss": 0.0485, + "step": 63150 + }, + { + "epoch": 4.138111773570071, + "grad_norm": 0.5584223866462708, + "learning_rate": 3.2738366050794985e-05, + "loss": 0.036, + "step": 63160 + }, + { + "epoch": 4.13876695276158, + "grad_norm": 0.2995356023311615, + "learning_rate": 3.2722848897505034e-05, + "loss": 0.0238, + "step": 63170 + }, + { + "epoch": 4.139422131953089, + "grad_norm": 0.9173963069915771, + "learning_rate": 3.2707333633618356e-05, + "loss": 0.0535, + "step": 63180 + }, + { + "epoch": 4.140077311144598, + "grad_norm": 0.46710333228111267, + "learning_rate": 3.269182026083167e-05, + "loss": 0.047, + "step": 63190 + }, + { + "epoch": 4.140732490336107, + "grad_norm": 0.5052980184555054, + "learning_rate": 3.2676308780841516e-05, + "loss": 0.0349, + "step": 63200 + }, + { + "epoch": 4.1413876695276155, + "grad_norm": 0.672664999961853, + "learning_rate": 3.266079919534416e-05, + "loss": 0.0336, + "step": 63210 + }, + { + "epoch": 4.142042848719124, + "grad_norm": 0.401718407869339, + "learning_rate": 3.264529150603575e-05, + "loss": 0.0338, + "step": 63220 + }, + { + "epoch": 4.142698027910633, + "grad_norm": 0.4488096833229065, + "learning_rate": 3.262978571461217e-05, + "loss": 0.0161, + "step": 63230 + }, + { + "epoch": 4.143353207102143, + "grad_norm": 0.41704732179641724, + "learning_rate": 3.26142818227691e-05, + "loss": 0.0399, + "step": 63240 + }, + { + "epoch": 4.144008386293652, + "grad_norm": 0.1298312097787857, + "learning_rate": 3.259877983220203e-05, + "loss": 0.0275, + "step": 63250 + }, + { + "epoch": 4.1446635654851605, + "grad_norm": 0.5140563249588013, + "learning_rate": 3.2583279744606246e-05, + "loss": 0.0332, + "step": 63260 + }, + { + "epoch": 4.145318744676669, + "grad_norm": 0.5031627416610718, + "learning_rate": 3.25677815616768e-05, + "loss": 0.0389, + "step": 63270 + }, + { + "epoch": 4.145973923868178, + "grad_norm": 0.3952106535434723, + "learning_rate": 3.255228528510856e-05, + "loss": 0.0524, + "step": 63280 + }, + { + "epoch": 4.146629103059687, + "grad_norm": 0.6185854077339172, + "learning_rate": 3.253679091659617e-05, + "loss": 0.0326, + "step": 63290 + }, + { + "epoch": 4.147284282251196, + "grad_norm": 0.32486748695373535, + "learning_rate": 3.252129845783404e-05, + "loss": 0.0347, + "step": 63300 + }, + { + "epoch": 4.1479394614427045, + "grad_norm": 0.5352954864501953, + "learning_rate": 3.2505807910516456e-05, + "loss": 0.0386, + "step": 63310 + }, + { + "epoch": 4.148594640634213, + "grad_norm": 0.9711044430732727, + "learning_rate": 3.249031927633742e-05, + "loss": 0.0324, + "step": 63320 + }, + { + "epoch": 4.149249819825722, + "grad_norm": 0.2938831150531769, + "learning_rate": 3.2474832556990737e-05, + "loss": 0.0368, + "step": 63330 + }, + { + "epoch": 4.149904999017231, + "grad_norm": 0.5938271880149841, + "learning_rate": 3.245934775417002e-05, + "loss": 0.0368, + "step": 63340 + }, + { + "epoch": 4.15056017820874, + "grad_norm": 0.3719237744808197, + "learning_rate": 3.244386486956866e-05, + "loss": 0.0382, + "step": 63350 + }, + { + "epoch": 4.151215357400249, + "grad_norm": 0.1445244997739792, + "learning_rate": 3.2428383904879846e-05, + "loss": 0.0277, + "step": 63360 + }, + { + "epoch": 4.151870536591757, + "grad_norm": 0.4297127425670624, + "learning_rate": 3.241290486179652e-05, + "loss": 0.0226, + "step": 63370 + }, + { + "epoch": 4.152525715783267, + "grad_norm": 0.1745194047689438, + "learning_rate": 3.239742774201151e-05, + "loss": 0.0305, + "step": 63380 + }, + { + "epoch": 4.153180894974776, + "grad_norm": 0.29599660634994507, + "learning_rate": 3.238195254721733e-05, + "loss": 0.0148, + "step": 63390 + }, + { + "epoch": 4.153836074166285, + "grad_norm": 0.9324469566345215, + "learning_rate": 3.236647927910633e-05, + "loss": 0.0365, + "step": 63400 + }, + { + "epoch": 4.154491253357794, + "grad_norm": 1.110753059387207, + "learning_rate": 3.2351007939370646e-05, + "loss": 0.0392, + "step": 63410 + }, + { + "epoch": 4.155146432549302, + "grad_norm": 0.18407195806503296, + "learning_rate": 3.2335538529702184e-05, + "loss": 0.0317, + "step": 63420 + }, + { + "epoch": 4.155801611740811, + "grad_norm": 0.3122023940086365, + "learning_rate": 3.232007105179268e-05, + "loss": 0.0283, + "step": 63430 + }, + { + "epoch": 4.15645679093232, + "grad_norm": 1.019441843032837, + "learning_rate": 3.230460550733363e-05, + "loss": 0.0257, + "step": 63440 + }, + { + "epoch": 4.157111970123829, + "grad_norm": 0.2568833827972412, + "learning_rate": 3.2289141898016315e-05, + "loss": 0.0348, + "step": 63450 + }, + { + "epoch": 4.157767149315338, + "grad_norm": 0.11917540431022644, + "learning_rate": 3.227368022553181e-05, + "loss": 0.0294, + "step": 63460 + }, + { + "epoch": 4.1584223285068465, + "grad_norm": 0.8352254629135132, + "learning_rate": 3.225822049157099e-05, + "loss": 0.0494, + "step": 63470 + }, + { + "epoch": 4.159077507698355, + "grad_norm": 1.0103485584259033, + "learning_rate": 3.2242762697824494e-05, + "loss": 0.0295, + "step": 63480 + }, + { + "epoch": 4.159732686889864, + "grad_norm": 0.9898157119750977, + "learning_rate": 3.222730684598279e-05, + "loss": 0.0503, + "step": 63490 + }, + { + "epoch": 4.160387866081373, + "grad_norm": 0.8274470567703247, + "learning_rate": 3.221185293773605e-05, + "loss": 0.0615, + "step": 63500 + }, + { + "epoch": 4.161043045272882, + "grad_norm": 0.5662248134613037, + "learning_rate": 3.2196400974774354e-05, + "loss": 0.0223, + "step": 63510 + }, + { + "epoch": 4.161698224464391, + "grad_norm": 0.5976825952529907, + "learning_rate": 3.2180950958787485e-05, + "loss": 0.0373, + "step": 63520 + }, + { + "epoch": 4.1623534036559, + "grad_norm": 0.265011727809906, + "learning_rate": 3.2165502891465024e-05, + "loss": 0.0413, + "step": 63530 + }, + { + "epoch": 4.163008582847409, + "grad_norm": 0.1327810436487198, + "learning_rate": 3.215005677449636e-05, + "loss": 0.0387, + "step": 63540 + }, + { + "epoch": 4.163663762038918, + "grad_norm": 1.5410518646240234, + "learning_rate": 3.2134612609570644e-05, + "loss": 0.0364, + "step": 63550 + }, + { + "epoch": 4.164318941230427, + "grad_norm": 1.1423230171203613, + "learning_rate": 3.211917039837682e-05, + "loss": 0.0374, + "step": 63560 + }, + { + "epoch": 4.1649741204219355, + "grad_norm": 0.13856813311576843, + "learning_rate": 3.210373014260365e-05, + "loss": 0.0258, + "step": 63570 + }, + { + "epoch": 4.165629299613444, + "grad_norm": 0.24064332246780396, + "learning_rate": 3.208829184393965e-05, + "loss": 0.0242, + "step": 63580 + }, + { + "epoch": 4.166284478804953, + "grad_norm": 0.6023169159889221, + "learning_rate": 3.207285550407312e-05, + "loss": 0.0277, + "step": 63590 + }, + { + "epoch": 4.166939657996462, + "grad_norm": 0.3926980197429657, + "learning_rate": 3.205742112469217e-05, + "loss": 0.023, + "step": 63600 + }, + { + "epoch": 4.167594837187971, + "grad_norm": 0.4889041781425476, + "learning_rate": 3.204198870748465e-05, + "loss": 0.0408, + "step": 63610 + }, + { + "epoch": 4.16825001637948, + "grad_norm": 0.11441830545663834, + "learning_rate": 3.202655825413824e-05, + "loss": 0.0347, + "step": 63620 + }, + { + "epoch": 4.168905195570988, + "grad_norm": 0.15624013543128967, + "learning_rate": 3.2011129766340416e-05, + "loss": 0.0402, + "step": 63630 + }, + { + "epoch": 4.169560374762497, + "grad_norm": 0.9613751769065857, + "learning_rate": 3.1995703245778384e-05, + "loss": 0.0522, + "step": 63640 + }, + { + "epoch": 4.170215553954006, + "grad_norm": 0.2038218379020691, + "learning_rate": 3.1980278694139186e-05, + "loss": 0.0369, + "step": 63650 + }, + { + "epoch": 4.170870733145516, + "grad_norm": 0.33200737833976746, + "learning_rate": 3.196485611310962e-05, + "loss": 0.0301, + "step": 63660 + }, + { + "epoch": 4.1715259123370245, + "grad_norm": 0.48920589685440063, + "learning_rate": 3.1949435504376257e-05, + "loss": 0.0383, + "step": 63670 + }, + { + "epoch": 4.172181091528533, + "grad_norm": 0.2681431174278259, + "learning_rate": 3.1934016869625505e-05, + "loss": 0.0433, + "step": 63680 + }, + { + "epoch": 4.172836270720042, + "grad_norm": 0.44305968284606934, + "learning_rate": 3.191860021054349e-05, + "loss": 0.0388, + "step": 63690 + }, + { + "epoch": 4.173491449911551, + "grad_norm": 0.8369107842445374, + "learning_rate": 3.190318552881618e-05, + "loss": 0.0346, + "step": 63700 + }, + { + "epoch": 4.17414662910306, + "grad_norm": 0.17707295715808868, + "learning_rate": 3.1887772826129285e-05, + "loss": 0.0337, + "step": 63710 + }, + { + "epoch": 4.174801808294569, + "grad_norm": 0.6748296618461609, + "learning_rate": 3.1872362104168335e-05, + "loss": 0.021, + "step": 63720 + }, + { + "epoch": 4.175456987486077, + "grad_norm": 0.9245247840881348, + "learning_rate": 3.185695336461861e-05, + "loss": 0.0465, + "step": 63730 + }, + { + "epoch": 4.176112166677586, + "grad_norm": 0.3637079894542694, + "learning_rate": 3.184154660916519e-05, + "loss": 0.0325, + "step": 63740 + }, + { + "epoch": 4.176767345869095, + "grad_norm": 0.7312350273132324, + "learning_rate": 3.1826141839492906e-05, + "loss": 0.0254, + "step": 63750 + }, + { + "epoch": 4.177422525060604, + "grad_norm": 0.5065305233001709, + "learning_rate": 3.181073905728644e-05, + "loss": 0.0307, + "step": 63760 + }, + { + "epoch": 4.178077704252113, + "grad_norm": 0.3809003233909607, + "learning_rate": 3.1795338264230214e-05, + "loss": 0.0299, + "step": 63770 + }, + { + "epoch": 4.1787328834436215, + "grad_norm": 0.23865871131420135, + "learning_rate": 3.1779939462008426e-05, + "loss": 0.0274, + "step": 63780 + }, + { + "epoch": 4.17938806263513, + "grad_norm": 0.5263676643371582, + "learning_rate": 3.1764542652305066e-05, + "loss": 0.0252, + "step": 63790 + }, + { + "epoch": 4.18004324182664, + "grad_norm": 0.34353816509246826, + "learning_rate": 3.17491478368039e-05, + "loss": 0.0182, + "step": 63800 + }, + { + "epoch": 4.180698421018149, + "grad_norm": 0.10046042501926422, + "learning_rate": 3.173375501718847e-05, + "loss": 0.0393, + "step": 63810 + }, + { + "epoch": 4.181353600209658, + "grad_norm": 0.19501733779907227, + "learning_rate": 3.171836419514215e-05, + "loss": 0.0358, + "step": 63820 + }, + { + "epoch": 4.182008779401166, + "grad_norm": 0.6688746213912964, + "learning_rate": 3.170297537234803e-05, + "loss": 0.0216, + "step": 63830 + }, + { + "epoch": 4.182663958592675, + "grad_norm": 0.5977782607078552, + "learning_rate": 3.168758855048901e-05, + "loss": 0.0278, + "step": 63840 + }, + { + "epoch": 4.183319137784184, + "grad_norm": 0.15185800194740295, + "learning_rate": 3.1672203731247785e-05, + "loss": 0.0402, + "step": 63850 + }, + { + "epoch": 4.183974316975693, + "grad_norm": 0.2546505630016327, + "learning_rate": 3.16568209163068e-05, + "loss": 0.0245, + "step": 63860 + }, + { + "epoch": 4.184629496167202, + "grad_norm": 0.4587727189064026, + "learning_rate": 3.1641440107348295e-05, + "loss": 0.0465, + "step": 63870 + }, + { + "epoch": 4.1852846753587105, + "grad_norm": 0.7621839046478271, + "learning_rate": 3.162606130605429e-05, + "loss": 0.0334, + "step": 63880 + }, + { + "epoch": 4.185939854550219, + "grad_norm": 0.3310460150241852, + "learning_rate": 3.16106845141066e-05, + "loss": 0.035, + "step": 63890 + }, + { + "epoch": 4.186595033741728, + "grad_norm": 0.5199667811393738, + "learning_rate": 3.159530973318682e-05, + "loss": 0.0358, + "step": 63900 + }, + { + "epoch": 4.187250212933237, + "grad_norm": 0.391729474067688, + "learning_rate": 3.1579936964976286e-05, + "loss": 0.0303, + "step": 63910 + }, + { + "epoch": 4.187905392124746, + "grad_norm": 0.6967337131500244, + "learning_rate": 3.156456621115615e-05, + "loss": 0.0215, + "step": 63920 + }, + { + "epoch": 4.188560571316255, + "grad_norm": 0.25258544087409973, + "learning_rate": 3.154919747340734e-05, + "loss": 0.0289, + "step": 63930 + }, + { + "epoch": 4.189215750507764, + "grad_norm": 0.23957523703575134, + "learning_rate": 3.153383075341054e-05, + "loss": 0.0235, + "step": 63940 + }, + { + "epoch": 4.189870929699273, + "grad_norm": 0.412675142288208, + "learning_rate": 3.1518466052846264e-05, + "loss": 0.0272, + "step": 63950 + }, + { + "epoch": 4.190526108890782, + "grad_norm": 0.30912330746650696, + "learning_rate": 3.150310337339474e-05, + "loss": 0.0491, + "step": 63960 + }, + { + "epoch": 4.191181288082291, + "grad_norm": 0.7821325659751892, + "learning_rate": 3.148774271673603e-05, + "loss": 0.0309, + "step": 63970 + }, + { + "epoch": 4.1918364672737995, + "grad_norm": 0.47467344999313354, + "learning_rate": 3.147238408454995e-05, + "loss": 0.0401, + "step": 63980 + }, + { + "epoch": 4.192491646465308, + "grad_norm": 0.45854851603507996, + "learning_rate": 3.1457027478516086e-05, + "loss": 0.0493, + "step": 63990 + }, + { + "epoch": 4.193146825656817, + "grad_norm": 0.6916714310646057, + "learning_rate": 3.144167290031382e-05, + "loss": 0.0425, + "step": 64000 + }, + { + "epoch": 4.193802004848326, + "grad_norm": 0.3138273060321808, + "learning_rate": 3.142632035162227e-05, + "loss": 0.0341, + "step": 64010 + }, + { + "epoch": 4.194457184039835, + "grad_norm": 0.4802565574645996, + "learning_rate": 3.141096983412044e-05, + "loss": 0.0257, + "step": 64020 + }, + { + "epoch": 4.195112363231344, + "grad_norm": 0.14838869869709015, + "learning_rate": 3.139562134948699e-05, + "loss": 0.0246, + "step": 64030 + }, + { + "epoch": 4.195767542422852, + "grad_norm": 0.3690524399280548, + "learning_rate": 3.1380274899400424e-05, + "loss": 0.0232, + "step": 64040 + }, + { + "epoch": 4.196422721614361, + "grad_norm": 0.1754065454006195, + "learning_rate": 3.1364930485539e-05, + "loss": 0.0438, + "step": 64050 + }, + { + "epoch": 4.19707790080587, + "grad_norm": 0.37461116909980774, + "learning_rate": 3.134958810958076e-05, + "loss": 0.0375, + "step": 64060 + }, + { + "epoch": 4.197733079997379, + "grad_norm": 0.30029433965682983, + "learning_rate": 3.13342477732035e-05, + "loss": 0.0256, + "step": 64070 + }, + { + "epoch": 4.198388259188889, + "grad_norm": 0.3778083324432373, + "learning_rate": 3.131890947808487e-05, + "loss": 0.0334, + "step": 64080 + }, + { + "epoch": 4.199043438380397, + "grad_norm": 0.6875403523445129, + "learning_rate": 3.13035732259022e-05, + "loss": 0.0482, + "step": 64090 + }, + { + "epoch": 4.199698617571906, + "grad_norm": 0.9109882116317749, + "learning_rate": 3.1288239018332656e-05, + "loss": 0.0329, + "step": 64100 + }, + { + "epoch": 4.200353796763415, + "grad_norm": 0.6024132966995239, + "learning_rate": 3.1272906857053164e-05, + "loss": 0.0287, + "step": 64110 + }, + { + "epoch": 4.201008975954924, + "grad_norm": 0.611677348613739, + "learning_rate": 3.125757674374042e-05, + "loss": 0.055, + "step": 64120 + }, + { + "epoch": 4.201664155146433, + "grad_norm": 0.3419587314128876, + "learning_rate": 3.124224868007088e-05, + "loss": 0.0435, + "step": 64130 + }, + { + "epoch": 4.2023193343379415, + "grad_norm": 0.6419224143028259, + "learning_rate": 3.1226922667720854e-05, + "loss": 0.025, + "step": 64140 + }, + { + "epoch": 4.20297451352945, + "grad_norm": 0.797681987285614, + "learning_rate": 3.121159870836633e-05, + "loss": 0.0291, + "step": 64150 + }, + { + "epoch": 4.203629692720959, + "grad_norm": 0.30184823274612427, + "learning_rate": 3.119627680368312e-05, + "loss": 0.0342, + "step": 64160 + }, + { + "epoch": 4.204284871912468, + "grad_norm": 0.88387531042099, + "learning_rate": 3.1180956955346816e-05, + "loss": 0.0333, + "step": 64170 + }, + { + "epoch": 4.204940051103977, + "grad_norm": 0.5023416876792908, + "learning_rate": 3.116563916503276e-05, + "loss": 0.0198, + "step": 64180 + }, + { + "epoch": 4.2055952302954855, + "grad_norm": 1.144201636314392, + "learning_rate": 3.1150323434416095e-05, + "loss": 0.0474, + "step": 64190 + }, + { + "epoch": 4.206250409486994, + "grad_norm": 0.37374550104141235, + "learning_rate": 3.1135009765171696e-05, + "loss": 0.0308, + "step": 64200 + }, + { + "epoch": 4.206905588678503, + "grad_norm": 0.09608368575572968, + "learning_rate": 3.111969815897429e-05, + "loss": 0.0224, + "step": 64210 + }, + { + "epoch": 4.207560767870013, + "grad_norm": 0.7938531637191772, + "learning_rate": 3.1104388617498295e-05, + "loss": 0.0378, + "step": 64220 + }, + { + "epoch": 4.208215947061522, + "grad_norm": 0.4992733597755432, + "learning_rate": 3.1089081142417966e-05, + "loss": 0.0241, + "step": 64230 + }, + { + "epoch": 4.2088711262530305, + "grad_norm": 0.5022972226142883, + "learning_rate": 3.1073775735407285e-05, + "loss": 0.0364, + "step": 64240 + }, + { + "epoch": 4.209526305444539, + "grad_norm": 0.2878110408782959, + "learning_rate": 3.1058472398140035e-05, + "loss": 0.0257, + "step": 64250 + }, + { + "epoch": 4.210181484636048, + "grad_norm": 0.6445035338401794, + "learning_rate": 3.1043171132289736e-05, + "loss": 0.038, + "step": 64260 + }, + { + "epoch": 4.210836663827557, + "grad_norm": 0.8636753559112549, + "learning_rate": 3.1027871939529774e-05, + "loss": 0.0311, + "step": 64270 + }, + { + "epoch": 4.211491843019066, + "grad_norm": 0.3294826149940491, + "learning_rate": 3.1012574821533214e-05, + "loss": 0.0234, + "step": 64280 + }, + { + "epoch": 4.212147022210575, + "grad_norm": 0.26422303915023804, + "learning_rate": 3.099727977997291e-05, + "loss": 0.0291, + "step": 64290 + }, + { + "epoch": 4.212802201402083, + "grad_norm": 0.9990756511688232, + "learning_rate": 3.098198681652154e-05, + "loss": 0.0226, + "step": 64300 + }, + { + "epoch": 4.213457380593592, + "grad_norm": 0.9208850264549255, + "learning_rate": 3.0966695932851475e-05, + "loss": 0.0248, + "step": 64310 + }, + { + "epoch": 4.214112559785101, + "grad_norm": 0.15846405923366547, + "learning_rate": 3.0951407130634926e-05, + "loss": 0.0391, + "step": 64320 + }, + { + "epoch": 4.21476773897661, + "grad_norm": 0.829221785068512, + "learning_rate": 3.0936120411543864e-05, + "loss": 0.0461, + "step": 64330 + }, + { + "epoch": 4.215422918168119, + "grad_norm": 0.33942681550979614, + "learning_rate": 3.0920835777250015e-05, + "loss": 0.0267, + "step": 64340 + }, + { + "epoch": 4.2160780973596275, + "grad_norm": 0.7755626440048218, + "learning_rate": 3.090555322942488e-05, + "loss": 0.0471, + "step": 64350 + }, + { + "epoch": 4.216733276551137, + "grad_norm": 1.2583540678024292, + "learning_rate": 3.089027276973973e-05, + "loss": 0.0295, + "step": 64360 + }, + { + "epoch": 4.217388455742646, + "grad_norm": 0.5069085955619812, + "learning_rate": 3.0874994399865615e-05, + "loss": 0.0299, + "step": 64370 + }, + { + "epoch": 4.218043634934155, + "grad_norm": 0.5242772698402405, + "learning_rate": 3.085971812147337e-05, + "loss": 0.0278, + "step": 64380 + }, + { + "epoch": 4.218698814125664, + "grad_norm": 0.4571276009082794, + "learning_rate": 3.0844443936233557e-05, + "loss": 0.034, + "step": 64390 + }, + { + "epoch": 4.219353993317172, + "grad_norm": 0.5006155967712402, + "learning_rate": 3.082917184581657e-05, + "loss": 0.0347, + "step": 64400 + }, + { + "epoch": 4.220009172508681, + "grad_norm": 0.4779432415962219, + "learning_rate": 3.0813901851892513e-05, + "loss": 0.0433, + "step": 64410 + }, + { + "epoch": 4.22066435170019, + "grad_norm": 0.40701600909233093, + "learning_rate": 3.0798633956131316e-05, + "loss": 0.0267, + "step": 64420 + }, + { + "epoch": 4.221319530891699, + "grad_norm": 0.1758584976196289, + "learning_rate": 3.078336816020264e-05, + "loss": 0.031, + "step": 64430 + }, + { + "epoch": 4.221974710083208, + "grad_norm": 0.4449167251586914, + "learning_rate": 3.0768104465775935e-05, + "loss": 0.025, + "step": 64440 + }, + { + "epoch": 4.2226298892747165, + "grad_norm": 0.3096228241920471, + "learning_rate": 3.0752842874520396e-05, + "loss": 0.0465, + "step": 64450 + }, + { + "epoch": 4.223285068466225, + "grad_norm": 0.5157759189605713, + "learning_rate": 3.073758338810503e-05, + "loss": 0.0352, + "step": 64460 + }, + { + "epoch": 4.223940247657734, + "grad_norm": 0.2939293682575226, + "learning_rate": 3.072232600819859e-05, + "loss": 0.0249, + "step": 64470 + }, + { + "epoch": 4.224595426849243, + "grad_norm": 0.8372147083282471, + "learning_rate": 3.070707073646959e-05, + "loss": 0.0464, + "step": 64480 + }, + { + "epoch": 4.225250606040752, + "grad_norm": 0.37345126271247864, + "learning_rate": 3.069181757458633e-05, + "loss": 0.0365, + "step": 64490 + }, + { + "epoch": 4.225905785232261, + "grad_norm": 0.9892610907554626, + "learning_rate": 3.067656652421687e-05, + "loss": 0.0232, + "step": 64500 + }, + { + "epoch": 4.22656096442377, + "grad_norm": 0.34616586565971375, + "learning_rate": 3.0661317587029025e-05, + "loss": 0.0322, + "step": 64510 + }, + { + "epoch": 4.227216143615279, + "grad_norm": 0.08253210783004761, + "learning_rate": 3.064607076469043e-05, + "loss": 0.0571, + "step": 64520 + }, + { + "epoch": 4.227871322806788, + "grad_norm": 0.43091633915901184, + "learning_rate": 3.063082605886845e-05, + "loss": 0.0312, + "step": 64530 + }, + { + "epoch": 4.228526501998297, + "grad_norm": 0.4726986289024353, + "learning_rate": 3.061558347123019e-05, + "loss": 0.0229, + "step": 64540 + }, + { + "epoch": 4.2291816811898055, + "grad_norm": 0.5582835674285889, + "learning_rate": 3.060034300344259e-05, + "loss": 0.045, + "step": 64550 + }, + { + "epoch": 4.229836860381314, + "grad_norm": 0.5842954516410828, + "learning_rate": 3.058510465717231e-05, + "loss": 0.035, + "step": 64560 + }, + { + "epoch": 4.230492039572823, + "grad_norm": 1.0239813327789307, + "learning_rate": 3.0569868434085804e-05, + "loss": 0.0353, + "step": 64570 + }, + { + "epoch": 4.231147218764332, + "grad_norm": 0.8829987049102783, + "learning_rate": 3.055463433584924e-05, + "loss": 0.042, + "step": 64580 + }, + { + "epoch": 4.231802397955841, + "grad_norm": 0.2327737808227539, + "learning_rate": 3.0539402364128656e-05, + "loss": 0.0567, + "step": 64590 + }, + { + "epoch": 4.23245757714735, + "grad_norm": 0.39652225375175476, + "learning_rate": 3.052417252058977e-05, + "loss": 0.0174, + "step": 64600 + }, + { + "epoch": 4.233112756338858, + "grad_norm": 0.47913870215415955, + "learning_rate": 3.0508944806898092e-05, + "loss": 0.0558, + "step": 64610 + }, + { + "epoch": 4.233767935530367, + "grad_norm": 0.5249766111373901, + "learning_rate": 3.04937192247189e-05, + "loss": 0.0313, + "step": 64620 + }, + { + "epoch": 4.234423114721876, + "grad_norm": 0.5302515625953674, + "learning_rate": 3.047849577571725e-05, + "loss": 0.0216, + "step": 64630 + }, + { + "epoch": 4.235078293913386, + "grad_norm": 0.31987208127975464, + "learning_rate": 3.046327446155795e-05, + "loss": 0.0152, + "step": 64640 + }, + { + "epoch": 4.235733473104895, + "grad_norm": 0.49952176213264465, + "learning_rate": 3.0448055283905587e-05, + "loss": 0.0254, + "step": 64650 + }, + { + "epoch": 4.236388652296403, + "grad_norm": 0.08562424033880234, + "learning_rate": 3.0432838244424494e-05, + "loss": 0.026, + "step": 64660 + }, + { + "epoch": 4.237043831487912, + "grad_norm": 0.5309025645256042, + "learning_rate": 3.04176233447788e-05, + "loss": 0.0453, + "step": 64670 + }, + { + "epoch": 4.237699010679421, + "grad_norm": 0.3635540306568146, + "learning_rate": 3.040241058663238e-05, + "loss": 0.0248, + "step": 64680 + }, + { + "epoch": 4.23835418987093, + "grad_norm": 0.14007815718650818, + "learning_rate": 3.038719997164887e-05, + "loss": 0.0216, + "step": 64690 + }, + { + "epoch": 4.239009369062439, + "grad_norm": 0.16307243704795837, + "learning_rate": 3.0371991501491683e-05, + "loss": 0.0228, + "step": 64700 + }, + { + "epoch": 4.2396645482539475, + "grad_norm": 0.3723461627960205, + "learning_rate": 3.0356785177823975e-05, + "loss": 0.0265, + "step": 64710 + }, + { + "epoch": 4.240319727445456, + "grad_norm": 0.11523589491844177, + "learning_rate": 3.0341581002308726e-05, + "loss": 0.0213, + "step": 64720 + }, + { + "epoch": 4.240974906636965, + "grad_norm": 0.4427684545516968, + "learning_rate": 3.0326378976608627e-05, + "loss": 0.0403, + "step": 64730 + }, + { + "epoch": 4.241630085828474, + "grad_norm": 0.30691924691200256, + "learning_rate": 3.0311179102386146e-05, + "loss": 0.0209, + "step": 64740 + }, + { + "epoch": 4.242285265019983, + "grad_norm": 0.5415634512901306, + "learning_rate": 3.0295981381303513e-05, + "loss": 0.0293, + "step": 64750 + }, + { + "epoch": 4.2429404442114915, + "grad_norm": 0.9476343393325806, + "learning_rate": 3.0280785815022727e-05, + "loss": 0.0434, + "step": 64760 + }, + { + "epoch": 4.243595623403, + "grad_norm": 0.32925254106521606, + "learning_rate": 3.0265592405205544e-05, + "loss": 0.0427, + "step": 64770 + }, + { + "epoch": 4.244250802594509, + "grad_norm": 0.6436139345169067, + "learning_rate": 3.0250401153513517e-05, + "loss": 0.0492, + "step": 64780 + }, + { + "epoch": 4.244905981786019, + "grad_norm": 0.20141254365444183, + "learning_rate": 3.0235212061607936e-05, + "loss": 0.0305, + "step": 64790 + }, + { + "epoch": 4.245561160977528, + "grad_norm": 0.710273027420044, + "learning_rate": 3.0220025131149842e-05, + "loss": 0.0218, + "step": 64800 + }, + { + "epoch": 4.2462163401690365, + "grad_norm": 0.5631672739982605, + "learning_rate": 3.020484036380007e-05, + "loss": 0.0494, + "step": 64810 + }, + { + "epoch": 4.246871519360545, + "grad_norm": 0.36401161551475525, + "learning_rate": 3.0189657761219188e-05, + "loss": 0.0286, + "step": 64820 + }, + { + "epoch": 4.247526698552054, + "grad_norm": 0.18356816470623016, + "learning_rate": 3.017447732506753e-05, + "loss": 0.0312, + "step": 64830 + }, + { + "epoch": 4.248181877743563, + "grad_norm": 0.121933214366436, + "learning_rate": 3.015929905700524e-05, + "loss": 0.0247, + "step": 64840 + }, + { + "epoch": 4.248837056935072, + "grad_norm": 0.7700938582420349, + "learning_rate": 3.014412295869218e-05, + "loss": 0.036, + "step": 64850 + }, + { + "epoch": 4.249492236126581, + "grad_norm": 0.553530216217041, + "learning_rate": 3.0128949031787977e-05, + "loss": 0.0632, + "step": 64860 + }, + { + "epoch": 4.250147415318089, + "grad_norm": 0.2102414220571518, + "learning_rate": 3.0113777277952022e-05, + "loss": 0.0288, + "step": 64870 + }, + { + "epoch": 4.250802594509598, + "grad_norm": 0.5150356292724609, + "learning_rate": 3.0098607698843496e-05, + "loss": 0.0144, + "step": 64880 + }, + { + "epoch": 4.251457773701107, + "grad_norm": 0.09945369511842728, + "learning_rate": 3.0083440296121312e-05, + "loss": 0.0346, + "step": 64890 + }, + { + "epoch": 4.252112952892616, + "grad_norm": 1.0296509265899658, + "learning_rate": 3.0068275071444142e-05, + "loss": 0.0444, + "step": 64900 + }, + { + "epoch": 4.252768132084125, + "grad_norm": 1.5020053386688232, + "learning_rate": 3.0053112026470442e-05, + "loss": 0.0522, + "step": 64910 + }, + { + "epoch": 4.253423311275634, + "grad_norm": 0.43224453926086426, + "learning_rate": 3.0037951162858434e-05, + "loss": 0.0257, + "step": 64920 + }, + { + "epoch": 4.254078490467143, + "grad_norm": 0.18158374726772308, + "learning_rate": 3.002279248226607e-05, + "loss": 0.0552, + "step": 64930 + }, + { + "epoch": 4.254733669658652, + "grad_norm": 0.13289549946784973, + "learning_rate": 3.000763598635109e-05, + "loss": 0.0373, + "step": 64940 + }, + { + "epoch": 4.255388848850161, + "grad_norm": 0.21604059636592865, + "learning_rate": 2.999248167677098e-05, + "loss": 0.0391, + "step": 64950 + }, + { + "epoch": 4.25604402804167, + "grad_norm": 0.5534622669219971, + "learning_rate": 2.9977329555182976e-05, + "loss": 0.0355, + "step": 64960 + }, + { + "epoch": 4.256699207233178, + "grad_norm": 0.49080514907836914, + "learning_rate": 2.9962179623244124e-05, + "loss": 0.0538, + "step": 64970 + }, + { + "epoch": 4.257354386424687, + "grad_norm": 0.8002592921257019, + "learning_rate": 2.9947031882611175e-05, + "loss": 0.0162, + "step": 64980 + }, + { + "epoch": 4.258009565616196, + "grad_norm": 0.4235740303993225, + "learning_rate": 2.9931886334940685e-05, + "loss": 0.0465, + "step": 64990 + }, + { + "epoch": 4.258664744807705, + "grad_norm": 0.6511480212211609, + "learning_rate": 2.9916742981888923e-05, + "loss": 0.0429, + "step": 65000 + }, + { + "epoch": 4.259319923999214, + "grad_norm": 0.7424668669700623, + "learning_rate": 2.990160182511196e-05, + "loss": 0.036, + "step": 65010 + }, + { + "epoch": 4.2599751031907225, + "grad_norm": 0.3212427496910095, + "learning_rate": 2.9886462866265574e-05, + "loss": 0.0219, + "step": 65020 + }, + { + "epoch": 4.260630282382231, + "grad_norm": 0.2697858512401581, + "learning_rate": 2.9871326107005392e-05, + "loss": 0.039, + "step": 65030 + }, + { + "epoch": 4.26128546157374, + "grad_norm": 0.808677613735199, + "learning_rate": 2.9856191548986735e-05, + "loss": 0.0372, + "step": 65040 + }, + { + "epoch": 4.261940640765249, + "grad_norm": 0.25712716579437256, + "learning_rate": 2.9841059193864672e-05, + "loss": 0.0218, + "step": 65050 + }, + { + "epoch": 4.262595819956758, + "grad_norm": 0.8354531526565552, + "learning_rate": 2.982592904329407e-05, + "loss": 0.0426, + "step": 65060 + }, + { + "epoch": 4.263250999148267, + "grad_norm": 0.6953337788581848, + "learning_rate": 2.981080109892954e-05, + "loss": 0.036, + "step": 65070 + }, + { + "epoch": 4.263906178339776, + "grad_norm": 0.5924437046051025, + "learning_rate": 2.9795675362425434e-05, + "loss": 0.0311, + "step": 65080 + }, + { + "epoch": 4.264561357531285, + "grad_norm": 0.20419302582740784, + "learning_rate": 2.97805518354359e-05, + "loss": 0.0369, + "step": 65090 + }, + { + "epoch": 4.265216536722794, + "grad_norm": 0.18716765940189362, + "learning_rate": 2.9765430519614825e-05, + "loss": 0.0284, + "step": 65100 + }, + { + "epoch": 4.265871715914303, + "grad_norm": 0.35182711482048035, + "learning_rate": 2.975031141661584e-05, + "loss": 0.0206, + "step": 65110 + }, + { + "epoch": 4.2665268951058115, + "grad_norm": 0.6879138350486755, + "learning_rate": 2.9735194528092347e-05, + "loss": 0.0198, + "step": 65120 + }, + { + "epoch": 4.26718207429732, + "grad_norm": 0.47283461689949036, + "learning_rate": 2.9720079855697525e-05, + "loss": 0.0247, + "step": 65130 + }, + { + "epoch": 4.267837253488829, + "grad_norm": 0.6075842976570129, + "learning_rate": 2.970496740108427e-05, + "loss": 0.0178, + "step": 65140 + }, + { + "epoch": 4.268492432680338, + "grad_norm": 0.46482157707214355, + "learning_rate": 2.968985716590527e-05, + "loss": 0.039, + "step": 65150 + }, + { + "epoch": 4.269147611871847, + "grad_norm": 1.0049189329147339, + "learning_rate": 2.9674749151812953e-05, + "loss": 0.0442, + "step": 65160 + }, + { + "epoch": 4.269802791063356, + "grad_norm": 0.6404855847358704, + "learning_rate": 2.965964336045951e-05, + "loss": 0.0215, + "step": 65170 + }, + { + "epoch": 4.270457970254864, + "grad_norm": 0.926144540309906, + "learning_rate": 2.9644539793496894e-05, + "loss": 0.0353, + "step": 65180 + }, + { + "epoch": 4.271113149446373, + "grad_norm": 0.8395259976387024, + "learning_rate": 2.9629438452576807e-05, + "loss": 0.0416, + "step": 65190 + }, + { + "epoch": 4.271768328637883, + "grad_norm": 0.8891525268554688, + "learning_rate": 2.9614339339350698e-05, + "loss": 0.0293, + "step": 65200 + }, + { + "epoch": 4.272423507829392, + "grad_norm": 0.5708650350570679, + "learning_rate": 2.959924245546979e-05, + "loss": 0.0301, + "step": 65210 + }, + { + "epoch": 4.2730786870209005, + "grad_norm": 1.0310691595077515, + "learning_rate": 2.9584147802585038e-05, + "loss": 0.0377, + "step": 65220 + }, + { + "epoch": 4.273733866212409, + "grad_norm": 0.3049702048301697, + "learning_rate": 2.956905538234721e-05, + "loss": 0.0313, + "step": 65230 + }, + { + "epoch": 4.274389045403918, + "grad_norm": 0.17784962058067322, + "learning_rate": 2.9553965196406764e-05, + "loss": 0.0405, + "step": 65240 + }, + { + "epoch": 4.275044224595427, + "grad_norm": 1.4964385032653809, + "learning_rate": 2.9538877246413943e-05, + "loss": 0.0594, + "step": 65250 + }, + { + "epoch": 4.275699403786936, + "grad_norm": 0.4522395133972168, + "learning_rate": 2.952379153401874e-05, + "loss": 0.0197, + "step": 65260 + }, + { + "epoch": 4.276354582978445, + "grad_norm": 0.3692321479320526, + "learning_rate": 2.9508708060870914e-05, + "loss": 0.0249, + "step": 65270 + }, + { + "epoch": 4.277009762169953, + "grad_norm": 0.9490233659744263, + "learning_rate": 2.9493626828619948e-05, + "loss": 0.0674, + "step": 65280 + }, + { + "epoch": 4.277664941361462, + "grad_norm": 0.26100701093673706, + "learning_rate": 2.9478547838915137e-05, + "loss": 0.051, + "step": 65290 + }, + { + "epoch": 4.278320120552971, + "grad_norm": 0.6818286180496216, + "learning_rate": 2.946347109340548e-05, + "loss": 0.0275, + "step": 65300 + }, + { + "epoch": 4.27897529974448, + "grad_norm": 0.3449108302593231, + "learning_rate": 2.944839659373976e-05, + "loss": 0.0388, + "step": 65310 + }, + { + "epoch": 4.279630478935989, + "grad_norm": 0.6551221609115601, + "learning_rate": 2.943332434156647e-05, + "loss": 0.0367, + "step": 65320 + }, + { + "epoch": 4.2802856581274975, + "grad_norm": 0.2039193958044052, + "learning_rate": 2.9418254338533912e-05, + "loss": 0.036, + "step": 65330 + }, + { + "epoch": 4.280940837319006, + "grad_norm": 0.30347341299057007, + "learning_rate": 2.940318658629011e-05, + "loss": 0.022, + "step": 65340 + }, + { + "epoch": 4.281596016510516, + "grad_norm": 0.32247844338417053, + "learning_rate": 2.938812108648287e-05, + "loss": 0.0362, + "step": 65350 + }, + { + "epoch": 4.282251195702025, + "grad_norm": 0.14035259187221527, + "learning_rate": 2.937305784075972e-05, + "loss": 0.0184, + "step": 65360 + }, + { + "epoch": 4.282906374893534, + "grad_norm": 0.8115250468254089, + "learning_rate": 2.9357996850767937e-05, + "loss": 0.04, + "step": 65370 + }, + { + "epoch": 4.2835615540850425, + "grad_norm": 0.2957543134689331, + "learning_rate": 2.9342938118154595e-05, + "loss": 0.0403, + "step": 65380 + }, + { + "epoch": 4.284216733276551, + "grad_norm": 0.18530920147895813, + "learning_rate": 2.9327881644566485e-05, + "loss": 0.0206, + "step": 65390 + }, + { + "epoch": 4.28487191246806, + "grad_norm": 0.19133706390857697, + "learning_rate": 2.9312827431650157e-05, + "loss": 0.014, + "step": 65400 + }, + { + "epoch": 4.285527091659569, + "grad_norm": 0.23558594286441803, + "learning_rate": 2.92977754810519e-05, + "loss": 0.0373, + "step": 65410 + }, + { + "epoch": 4.286182270851078, + "grad_norm": 0.5769022703170776, + "learning_rate": 2.928272579441781e-05, + "loss": 0.0359, + "step": 65420 + }, + { + "epoch": 4.2868374500425865, + "grad_norm": 0.15420213341712952, + "learning_rate": 2.9267678373393676e-05, + "loss": 0.026, + "step": 65430 + }, + { + "epoch": 4.287492629234095, + "grad_norm": 1.214676856994629, + "learning_rate": 2.925263321962507e-05, + "loss": 0.0316, + "step": 65440 + }, + { + "epoch": 4.288147808425604, + "grad_norm": 0.8309746384620667, + "learning_rate": 2.9237590334757292e-05, + "loss": 0.0364, + "step": 65450 + }, + { + "epoch": 4.288802987617113, + "grad_norm": 0.47000592947006226, + "learning_rate": 2.922254972043543e-05, + "loss": 0.0355, + "step": 65460 + }, + { + "epoch": 4.289458166808622, + "grad_norm": 0.3030545711517334, + "learning_rate": 2.9207511378304265e-05, + "loss": 0.0798, + "step": 65470 + }, + { + "epoch": 4.2901133460001315, + "grad_norm": 0.14393632113933563, + "learning_rate": 2.9192475310008428e-05, + "loss": 0.0444, + "step": 65480 + }, + { + "epoch": 4.29076852519164, + "grad_norm": 0.5152300596237183, + "learning_rate": 2.917744151719219e-05, + "loss": 0.0366, + "step": 65490 + }, + { + "epoch": 4.291423704383149, + "grad_norm": 0.43500834703445435, + "learning_rate": 2.916241000149964e-05, + "loss": 0.0501, + "step": 65500 + }, + { + "epoch": 4.292078883574658, + "grad_norm": 0.07881885021924973, + "learning_rate": 2.9147380764574622e-05, + "loss": 0.0228, + "step": 65510 + }, + { + "epoch": 4.292734062766167, + "grad_norm": 0.09570988267660141, + "learning_rate": 2.9132353808060674e-05, + "loss": 0.0243, + "step": 65520 + }, + { + "epoch": 4.293389241957676, + "grad_norm": 0.44808611273765564, + "learning_rate": 2.9117329133601134e-05, + "loss": 0.0346, + "step": 65530 + }, + { + "epoch": 4.294044421149184, + "grad_norm": 0.3333120346069336, + "learning_rate": 2.9102306742839093e-05, + "loss": 0.0341, + "step": 65540 + }, + { + "epoch": 4.294699600340693, + "grad_norm": 0.312394917011261, + "learning_rate": 2.9087286637417375e-05, + "loss": 0.028, + "step": 65550 + }, + { + "epoch": 4.295354779532202, + "grad_norm": 0.15852494537830353, + "learning_rate": 2.9072268818978537e-05, + "loss": 0.0498, + "step": 65560 + }, + { + "epoch": 4.296009958723711, + "grad_norm": 0.17922067642211914, + "learning_rate": 2.905725328916493e-05, + "loss": 0.0326, + "step": 65570 + }, + { + "epoch": 4.29666513791522, + "grad_norm": 0.37629446387290955, + "learning_rate": 2.9042240049618608e-05, + "loss": 0.0267, + "step": 65580 + }, + { + "epoch": 4.2973203171067285, + "grad_norm": 0.16252805292606354, + "learning_rate": 2.9027229101981413e-05, + "loss": 0.0261, + "step": 65590 + }, + { + "epoch": 4.297975496298237, + "grad_norm": 0.2653977572917938, + "learning_rate": 2.901222044789488e-05, + "loss": 0.0368, + "step": 65600 + }, + { + "epoch": 4.298630675489746, + "grad_norm": 0.5257033109664917, + "learning_rate": 2.89972140890004e-05, + "loss": 0.038, + "step": 65610 + }, + { + "epoch": 4.299285854681255, + "grad_norm": 0.8677962422370911, + "learning_rate": 2.8982210026938988e-05, + "loss": 0.0451, + "step": 65620 + }, + { + "epoch": 4.299941033872765, + "grad_norm": 0.12574225664138794, + "learning_rate": 2.896720826335151e-05, + "loss": 0.0177, + "step": 65630 + }, + { + "epoch": 4.300596213064273, + "grad_norm": 1.0718038082122803, + "learning_rate": 2.8952208799878493e-05, + "loss": 0.0439, + "step": 65640 + }, + { + "epoch": 4.301251392255782, + "grad_norm": 0.5529898405075073, + "learning_rate": 2.89372116381603e-05, + "loss": 0.0329, + "step": 65650 + }, + { + "epoch": 4.301906571447291, + "grad_norm": 0.32947465777397156, + "learning_rate": 2.8922216779836932e-05, + "loss": 0.0281, + "step": 65660 + }, + { + "epoch": 4.3025617506388, + "grad_norm": 0.3130796253681183, + "learning_rate": 2.8907224226548285e-05, + "loss": 0.0373, + "step": 65670 + }, + { + "epoch": 4.303216929830309, + "grad_norm": 0.9230312705039978, + "learning_rate": 2.8892233979933868e-05, + "loss": 0.0405, + "step": 65680 + }, + { + "epoch": 4.3038721090218175, + "grad_norm": 1.2508313655853271, + "learning_rate": 2.8877246041633026e-05, + "loss": 0.0322, + "step": 65690 + }, + { + "epoch": 4.304527288213326, + "grad_norm": 1.3778934478759766, + "learning_rate": 2.886226041328478e-05, + "loss": 0.0348, + "step": 65700 + }, + { + "epoch": 4.305182467404835, + "grad_norm": 0.8329241871833801, + "learning_rate": 2.8847277096527957e-05, + "loss": 0.0277, + "step": 65710 + }, + { + "epoch": 4.305837646596344, + "grad_norm": 0.34514138102531433, + "learning_rate": 2.8832296093001105e-05, + "loss": 0.0479, + "step": 65720 + }, + { + "epoch": 4.306492825787853, + "grad_norm": 0.10198367387056351, + "learning_rate": 2.8817317404342547e-05, + "loss": 0.0318, + "step": 65730 + }, + { + "epoch": 4.307148004979362, + "grad_norm": 0.13790501654148102, + "learning_rate": 2.880234103219029e-05, + "loss": 0.03, + "step": 65740 + }, + { + "epoch": 4.30780318417087, + "grad_norm": 1.010657787322998, + "learning_rate": 2.878736697818215e-05, + "loss": 0.0368, + "step": 65750 + }, + { + "epoch": 4.308458363362379, + "grad_norm": 0.35256439447402954, + "learning_rate": 2.877239524395568e-05, + "loss": 0.0227, + "step": 65760 + }, + { + "epoch": 4.309113542553889, + "grad_norm": 0.582562267780304, + "learning_rate": 2.8757425831148134e-05, + "loss": 0.0298, + "step": 65770 + }, + { + "epoch": 4.309768721745398, + "grad_norm": 0.5217640995979309, + "learning_rate": 2.874245874139658e-05, + "loss": 0.0334, + "step": 65780 + }, + { + "epoch": 4.3104239009369065, + "grad_norm": 0.09863989800214767, + "learning_rate": 2.8727493976337737e-05, + "loss": 0.0369, + "step": 65790 + }, + { + "epoch": 4.311079080128415, + "grad_norm": 0.41428038477897644, + "learning_rate": 2.8712531537608218e-05, + "loss": 0.0385, + "step": 65800 + }, + { + "epoch": 4.311734259319924, + "grad_norm": 1.081676721572876, + "learning_rate": 2.8697571426844217e-05, + "loss": 0.022, + "step": 65810 + }, + { + "epoch": 4.312389438511433, + "grad_norm": 0.23508845269680023, + "learning_rate": 2.86826136456818e-05, + "loss": 0.0169, + "step": 65820 + }, + { + "epoch": 4.313044617702942, + "grad_norm": 0.20111200213432312, + "learning_rate": 2.8667658195756685e-05, + "loss": 0.0248, + "step": 65830 + }, + { + "epoch": 4.313699796894451, + "grad_norm": 0.31846463680267334, + "learning_rate": 2.8652705078704412e-05, + "loss": 0.0611, + "step": 65840 + }, + { + "epoch": 4.314354976085959, + "grad_norm": 0.4210682213306427, + "learning_rate": 2.863775429616018e-05, + "loss": 0.0211, + "step": 65850 + }, + { + "epoch": 4.315010155277468, + "grad_norm": 2.3919050693511963, + "learning_rate": 2.862280584975906e-05, + "loss": 0.0387, + "step": 65860 + }, + { + "epoch": 4.315665334468977, + "grad_norm": 1.0696861743927002, + "learning_rate": 2.8607859741135735e-05, + "loss": 0.0518, + "step": 65870 + }, + { + "epoch": 4.316320513660486, + "grad_norm": 0.7864554524421692, + "learning_rate": 2.8592915971924723e-05, + "loss": 0.0227, + "step": 65880 + }, + { + "epoch": 4.316975692851995, + "grad_norm": 0.7400307655334473, + "learning_rate": 2.8577974543760212e-05, + "loss": 0.0374, + "step": 65890 + }, + { + "epoch": 4.3176308720435035, + "grad_norm": 0.13812004029750824, + "learning_rate": 2.8563035458276222e-05, + "loss": 0.0356, + "step": 65900 + }, + { + "epoch": 4.318286051235013, + "grad_norm": 0.7085509300231934, + "learning_rate": 2.8548098717106424e-05, + "loss": 0.0476, + "step": 65910 + }, + { + "epoch": 4.318941230426522, + "grad_norm": 0.628246009349823, + "learning_rate": 2.85331643218843e-05, + "loss": 0.0309, + "step": 65920 + }, + { + "epoch": 4.319596409618031, + "grad_norm": 0.16928589344024658, + "learning_rate": 2.8518232274243046e-05, + "loss": 0.0281, + "step": 65930 + }, + { + "epoch": 4.32025158880954, + "grad_norm": 0.22809214890003204, + "learning_rate": 2.8503302575815638e-05, + "loss": 0.0385, + "step": 65940 + }, + { + "epoch": 4.3209067680010484, + "grad_norm": 0.8110063672065735, + "learning_rate": 2.8488375228234716e-05, + "loss": 0.0559, + "step": 65950 + }, + { + "epoch": 4.321561947192557, + "grad_norm": 0.26837801933288574, + "learning_rate": 2.8473450233132738e-05, + "loss": 0.0501, + "step": 65960 + }, + { + "epoch": 4.322217126384066, + "grad_norm": 0.12235664576292038, + "learning_rate": 2.8458527592141898e-05, + "loss": 0.0212, + "step": 65970 + }, + { + "epoch": 4.322872305575575, + "grad_norm": 0.28622135519981384, + "learning_rate": 2.8443607306894073e-05, + "loss": 0.0225, + "step": 65980 + }, + { + "epoch": 4.323527484767084, + "grad_norm": 0.48323869705200195, + "learning_rate": 2.8428689379020946e-05, + "loss": 0.0274, + "step": 65990 + }, + { + "epoch": 4.3241826639585925, + "grad_norm": 0.6353256702423096, + "learning_rate": 2.841377381015391e-05, + "loss": 0.0329, + "step": 66000 + }, + { + "epoch": 4.324837843150101, + "grad_norm": 0.6967242360115051, + "learning_rate": 2.8398860601924144e-05, + "loss": 0.0503, + "step": 66010 + }, + { + "epoch": 4.32549302234161, + "grad_norm": 0.7405532598495483, + "learning_rate": 2.838394975596248e-05, + "loss": 0.0369, + "step": 66020 + }, + { + "epoch": 4.326148201533119, + "grad_norm": 0.8448024392127991, + "learning_rate": 2.83690412738996e-05, + "loss": 0.0273, + "step": 66030 + }, + { + "epoch": 4.326803380724628, + "grad_norm": 0.5249767303466797, + "learning_rate": 2.835413515736579e-05, + "loss": 0.0356, + "step": 66040 + }, + { + "epoch": 4.327458559916137, + "grad_norm": 1.1436232328414917, + "learning_rate": 2.8339231407991267e-05, + "loss": 0.0333, + "step": 66050 + }, + { + "epoch": 4.328113739107646, + "grad_norm": 0.37457308173179626, + "learning_rate": 2.832433002740581e-05, + "loss": 0.0327, + "step": 66060 + }, + { + "epoch": 4.328768918299155, + "grad_norm": 0.8046927452087402, + "learning_rate": 2.8309431017239047e-05, + "loss": 0.0308, + "step": 66070 + }, + { + "epoch": 4.329424097490664, + "grad_norm": 0.2606729567050934, + "learning_rate": 2.829453437912028e-05, + "loss": 0.0399, + "step": 66080 + }, + { + "epoch": 4.330079276682173, + "grad_norm": 0.2161693125963211, + "learning_rate": 2.8279640114678617e-05, + "loss": 0.0321, + "step": 66090 + }, + { + "epoch": 4.3307344558736816, + "grad_norm": 0.20352593064308167, + "learning_rate": 2.8264748225542836e-05, + "loss": 0.0271, + "step": 66100 + }, + { + "epoch": 4.33138963506519, + "grad_norm": 0.13794216513633728, + "learning_rate": 2.8249858713341516e-05, + "loss": 0.0294, + "step": 66110 + }, + { + "epoch": 4.332044814256699, + "grad_norm": 0.37225309014320374, + "learning_rate": 2.823497157970294e-05, + "loss": 0.0381, + "step": 66120 + }, + { + "epoch": 4.332699993448208, + "grad_norm": 0.45592641830444336, + "learning_rate": 2.8220086826255164e-05, + "loss": 0.0256, + "step": 66130 + }, + { + "epoch": 4.333355172639717, + "grad_norm": 0.20462985336780548, + "learning_rate": 2.8205204454625934e-05, + "loss": 0.0382, + "step": 66140 + }, + { + "epoch": 4.334010351831226, + "grad_norm": 0.6037302613258362, + "learning_rate": 2.8190324466442798e-05, + "loss": 0.04, + "step": 66150 + }, + { + "epoch": 4.334665531022734, + "grad_norm": 0.2437461018562317, + "learning_rate": 2.8175446863332965e-05, + "loss": 0.0434, + "step": 66160 + }, + { + "epoch": 4.335320710214243, + "grad_norm": 0.8256154656410217, + "learning_rate": 2.816057164692345e-05, + "loss": 0.0333, + "step": 66170 + }, + { + "epoch": 4.335975889405752, + "grad_norm": 0.22518685460090637, + "learning_rate": 2.8145698818840992e-05, + "loss": 0.0379, + "step": 66180 + }, + { + "epoch": 4.336631068597262, + "grad_norm": 0.2363658845424652, + "learning_rate": 2.8130828380712065e-05, + "loss": 0.0345, + "step": 66190 + }, + { + "epoch": 4.337286247788771, + "grad_norm": 0.07218612730503082, + "learning_rate": 2.811596033416285e-05, + "loss": 0.0322, + "step": 66200 + }, + { + "epoch": 4.337941426980279, + "grad_norm": 0.7276356220245361, + "learning_rate": 2.8101094680819307e-05, + "loss": 0.0305, + "step": 66210 + }, + { + "epoch": 4.338596606171788, + "grad_norm": 0.2674579918384552, + "learning_rate": 2.808623142230715e-05, + "loss": 0.0423, + "step": 66220 + }, + { + "epoch": 4.339251785363297, + "grad_norm": 0.7073169350624084, + "learning_rate": 2.8071370560251757e-05, + "loss": 0.04, + "step": 66230 + }, + { + "epoch": 4.339906964554806, + "grad_norm": 1.1266125440597534, + "learning_rate": 2.805651209627831e-05, + "loss": 0.0405, + "step": 66240 + }, + { + "epoch": 4.340562143746315, + "grad_norm": 0.8003162741661072, + "learning_rate": 2.804165603201171e-05, + "loss": 0.0395, + "step": 66250 + }, + { + "epoch": 4.3412173229378235, + "grad_norm": 0.24888354539871216, + "learning_rate": 2.802680236907661e-05, + "loss": 0.0359, + "step": 66260 + }, + { + "epoch": 4.341872502129332, + "grad_norm": 0.6626954674720764, + "learning_rate": 2.8011951109097344e-05, + "loss": 0.0507, + "step": 66270 + }, + { + "epoch": 4.342527681320841, + "grad_norm": 0.3301123082637787, + "learning_rate": 2.7997102253698072e-05, + "loss": 0.0369, + "step": 66280 + }, + { + "epoch": 4.34318286051235, + "grad_norm": 0.2414555549621582, + "learning_rate": 2.7982255804502588e-05, + "loss": 0.0403, + "step": 66290 + }, + { + "epoch": 4.343838039703859, + "grad_norm": 0.6264256238937378, + "learning_rate": 2.7967411763134516e-05, + "loss": 0.0333, + "step": 66300 + }, + { + "epoch": 4.3444932188953675, + "grad_norm": 0.6198685765266418, + "learning_rate": 2.7952570131217166e-05, + "loss": 0.0436, + "step": 66310 + }, + { + "epoch": 4.345148398086876, + "grad_norm": 0.24180985987186432, + "learning_rate": 2.7937730910373618e-05, + "loss": 0.0344, + "step": 66320 + }, + { + "epoch": 4.345803577278385, + "grad_norm": 0.7573059797286987, + "learning_rate": 2.7922894102226628e-05, + "loss": 0.0372, + "step": 66330 + }, + { + "epoch": 4.346458756469895, + "grad_norm": 0.4732860028743744, + "learning_rate": 2.790805970839876e-05, + "loss": 0.0425, + "step": 66340 + }, + { + "epoch": 4.347113935661404, + "grad_norm": 0.3428976833820343, + "learning_rate": 2.789322773051225e-05, + "loss": 0.0171, + "step": 66350 + }, + { + "epoch": 4.3477691148529125, + "grad_norm": 0.14886415004730225, + "learning_rate": 2.7878398170189124e-05, + "loss": 0.0213, + "step": 66360 + }, + { + "epoch": 4.348424294044421, + "grad_norm": 0.49001896381378174, + "learning_rate": 2.7863571029051112e-05, + "loss": 0.0171, + "step": 66370 + }, + { + "epoch": 4.34907947323593, + "grad_norm": 0.5310547947883606, + "learning_rate": 2.784874630871971e-05, + "loss": 0.0293, + "step": 66380 + }, + { + "epoch": 4.349734652427439, + "grad_norm": 0.6152661442756653, + "learning_rate": 2.7833924010816086e-05, + "loss": 0.0338, + "step": 66390 + }, + { + "epoch": 4.350389831618948, + "grad_norm": 0.12319200485944748, + "learning_rate": 2.7819104136961226e-05, + "loss": 0.0263, + "step": 66400 + }, + { + "epoch": 4.351045010810457, + "grad_norm": 0.13987693190574646, + "learning_rate": 2.780428668877577e-05, + "loss": 0.0226, + "step": 66410 + }, + { + "epoch": 4.351700190001965, + "grad_norm": 0.13419921696186066, + "learning_rate": 2.7789471667880152e-05, + "loss": 0.0214, + "step": 66420 + }, + { + "epoch": 4.352355369193474, + "grad_norm": 0.7743234038352966, + "learning_rate": 2.7774659075894523e-05, + "loss": 0.0209, + "step": 66430 + }, + { + "epoch": 4.353010548384983, + "grad_norm": 0.5745332837104797, + "learning_rate": 2.7759848914438778e-05, + "loss": 0.0299, + "step": 66440 + }, + { + "epoch": 4.353665727576492, + "grad_norm": 1.544848918914795, + "learning_rate": 2.7745041185132496e-05, + "loss": 0.0368, + "step": 66450 + }, + { + "epoch": 4.354320906768001, + "grad_norm": 0.29696202278137207, + "learning_rate": 2.773023588959506e-05, + "loss": 0.0369, + "step": 66460 + }, + { + "epoch": 4.35497608595951, + "grad_norm": 0.17040546238422394, + "learning_rate": 2.7715433029445555e-05, + "loss": 0.024, + "step": 66470 + }, + { + "epoch": 4.355631265151019, + "grad_norm": 0.6325590014457703, + "learning_rate": 2.7700632606302766e-05, + "loss": 0.03, + "step": 66480 + }, + { + "epoch": 4.356286444342528, + "grad_norm": 0.2676558196544647, + "learning_rate": 2.768583462178528e-05, + "loss": 0.0216, + "step": 66490 + }, + { + "epoch": 4.356941623534037, + "grad_norm": 0.3121233582496643, + "learning_rate": 2.767103907751136e-05, + "loss": 0.029, + "step": 66500 + }, + { + "epoch": 4.357596802725546, + "grad_norm": 0.20688751339912415, + "learning_rate": 2.7656245975099065e-05, + "loss": 0.0436, + "step": 66510 + }, + { + "epoch": 4.358251981917054, + "grad_norm": 0.39843279123306274, + "learning_rate": 2.7641455316166093e-05, + "loss": 0.0208, + "step": 66520 + }, + { + "epoch": 4.358907161108563, + "grad_norm": 0.1353054791688919, + "learning_rate": 2.7626667102329973e-05, + "loss": 0.0422, + "step": 66530 + }, + { + "epoch": 4.359562340300072, + "grad_norm": 0.1187821626663208, + "learning_rate": 2.761188133520788e-05, + "loss": 0.033, + "step": 66540 + }, + { + "epoch": 4.360217519491581, + "grad_norm": 0.5676360130310059, + "learning_rate": 2.7597098016416784e-05, + "loss": 0.0273, + "step": 66550 + }, + { + "epoch": 4.36087269868309, + "grad_norm": 0.6197702288627625, + "learning_rate": 2.7582317147573368e-05, + "loss": 0.0364, + "step": 66560 + }, + { + "epoch": 4.3615278778745985, + "grad_norm": 0.3400322198867798, + "learning_rate": 2.7567538730294064e-05, + "loss": 0.0488, + "step": 66570 + }, + { + "epoch": 4.362183057066107, + "grad_norm": 0.43596479296684265, + "learning_rate": 2.7552762766194973e-05, + "loss": 0.0243, + "step": 66580 + }, + { + "epoch": 4.362838236257616, + "grad_norm": 0.29464980959892273, + "learning_rate": 2.753798925689202e-05, + "loss": 0.0272, + "step": 66590 + }, + { + "epoch": 4.363493415449125, + "grad_norm": 0.478610098361969, + "learning_rate": 2.7523218204000774e-05, + "loss": 0.0434, + "step": 66600 + }, + { + "epoch": 4.364148594640634, + "grad_norm": 0.4029694199562073, + "learning_rate": 2.7508449609136612e-05, + "loss": 0.0275, + "step": 66610 + }, + { + "epoch": 4.3648037738321435, + "grad_norm": 0.6535963416099548, + "learning_rate": 2.749368347391454e-05, + "loss": 0.07, + "step": 66620 + }, + { + "epoch": 4.365458953023652, + "grad_norm": 0.6676713228225708, + "learning_rate": 2.7478919799949443e-05, + "loss": 0.0337, + "step": 66630 + }, + { + "epoch": 4.366114132215161, + "grad_norm": 0.14882776141166687, + "learning_rate": 2.7464158588855803e-05, + "loss": 0.0158, + "step": 66640 + }, + { + "epoch": 4.36676931140667, + "grad_norm": 0.49389535188674927, + "learning_rate": 2.7449399842247918e-05, + "loss": 0.0224, + "step": 66650 + }, + { + "epoch": 4.367424490598179, + "grad_norm": 0.22463931143283844, + "learning_rate": 2.7434643561739736e-05, + "loss": 0.0275, + "step": 66660 + }, + { + "epoch": 4.3680796697896875, + "grad_norm": 0.3660610616207123, + "learning_rate": 2.7419889748945016e-05, + "loss": 0.0607, + "step": 66670 + }, + { + "epoch": 4.368734848981196, + "grad_norm": 0.20273357629776, + "learning_rate": 2.7405138405477203e-05, + "loss": 0.0293, + "step": 66680 + }, + { + "epoch": 4.369390028172705, + "grad_norm": 0.4224242568016052, + "learning_rate": 2.73903895329495e-05, + "loss": 0.0336, + "step": 66690 + }, + { + "epoch": 4.370045207364214, + "grad_norm": 0.12818822264671326, + "learning_rate": 2.737564313297478e-05, + "loss": 0.0111, + "step": 66700 + }, + { + "epoch": 4.370700386555723, + "grad_norm": 0.5457078218460083, + "learning_rate": 2.736089920716572e-05, + "loss": 0.034, + "step": 66710 + }, + { + "epoch": 4.371355565747232, + "grad_norm": 0.45523637533187866, + "learning_rate": 2.7346157757134706e-05, + "loss": 0.0592, + "step": 66720 + }, + { + "epoch": 4.37201074493874, + "grad_norm": 0.19758877158164978, + "learning_rate": 2.733141878449379e-05, + "loss": 0.0511, + "step": 66730 + }, + { + "epoch": 4.372665924130249, + "grad_norm": 0.2667665481567383, + "learning_rate": 2.7316682290854844e-05, + "loss": 0.0349, + "step": 66740 + }, + { + "epoch": 4.373321103321759, + "grad_norm": 0.4189508259296417, + "learning_rate": 2.7301948277829414e-05, + "loss": 0.0374, + "step": 66750 + }, + { + "epoch": 4.373976282513268, + "grad_norm": 0.6640186309814453, + "learning_rate": 2.7287216747028814e-05, + "loss": 0.045, + "step": 66760 + }, + { + "epoch": 4.374631461704777, + "grad_norm": 0.17110620439052582, + "learning_rate": 2.7272487700064025e-05, + "loss": 0.0294, + "step": 66770 + }, + { + "epoch": 4.375286640896285, + "grad_norm": 0.19383810460567474, + "learning_rate": 2.725776113854583e-05, + "loss": 0.0408, + "step": 66780 + }, + { + "epoch": 4.375941820087794, + "grad_norm": 0.5295230150222778, + "learning_rate": 2.7243037064084664e-05, + "loss": 0.035, + "step": 66790 + }, + { + "epoch": 4.376596999279303, + "grad_norm": 0.5141541361808777, + "learning_rate": 2.7228315478290766e-05, + "loss": 0.0254, + "step": 66800 + }, + { + "epoch": 4.377252178470812, + "grad_norm": 0.7473584413528442, + "learning_rate": 2.721359638277402e-05, + "loss": 0.0449, + "step": 66810 + }, + { + "epoch": 4.377907357662321, + "grad_norm": 0.5130663514137268, + "learning_rate": 2.7198879779144148e-05, + "loss": 0.0293, + "step": 66820 + }, + { + "epoch": 4.3785625368538295, + "grad_norm": 0.21362394094467163, + "learning_rate": 2.7184165669010485e-05, + "loss": 0.0226, + "step": 66830 + }, + { + "epoch": 4.379217716045338, + "grad_norm": 0.6087156534194946, + "learning_rate": 2.716945405398218e-05, + "loss": 0.0312, + "step": 66840 + }, + { + "epoch": 4.379872895236847, + "grad_norm": 0.16436664760112762, + "learning_rate": 2.715474493566803e-05, + "loss": 0.0274, + "step": 66850 + }, + { + "epoch": 4.380528074428356, + "grad_norm": 0.30068036913871765, + "learning_rate": 2.7140038315676652e-05, + "loss": 0.0796, + "step": 66860 + }, + { + "epoch": 4.381183253619865, + "grad_norm": 0.0946308895945549, + "learning_rate": 2.7125334195616275e-05, + "loss": 0.0281, + "step": 66870 + }, + { + "epoch": 4.3818384328113735, + "grad_norm": 1.126017451286316, + "learning_rate": 2.7110632577094997e-05, + "loss": 0.0435, + "step": 66880 + }, + { + "epoch": 4.382493612002882, + "grad_norm": 0.22611482441425323, + "learning_rate": 2.709593346172051e-05, + "loss": 0.0312, + "step": 66890 + }, + { + "epoch": 4.383148791194392, + "grad_norm": 0.28604868054389954, + "learning_rate": 2.708123685110032e-05, + "loss": 0.0288, + "step": 66900 + }, + { + "epoch": 4.383803970385901, + "grad_norm": 0.47481995820999146, + "learning_rate": 2.7066542746841588e-05, + "loss": 0.0616, + "step": 66910 + }, + { + "epoch": 4.38445914957741, + "grad_norm": 0.6891726851463318, + "learning_rate": 2.7051851150551267e-05, + "loss": 0.0447, + "step": 66920 + }, + { + "epoch": 4.3851143287689185, + "grad_norm": 0.4267736077308655, + "learning_rate": 2.7037162063835996e-05, + "loss": 0.0437, + "step": 66930 + }, + { + "epoch": 4.385769507960427, + "grad_norm": 0.17084461450576782, + "learning_rate": 2.702247548830218e-05, + "loss": 0.0276, + "step": 66940 + }, + { + "epoch": 4.386424687151936, + "grad_norm": 0.7525601983070374, + "learning_rate": 2.7007791425555883e-05, + "loss": 0.0378, + "step": 66950 + }, + { + "epoch": 4.387079866343445, + "grad_norm": 0.3654661774635315, + "learning_rate": 2.6993109877202945e-05, + "loss": 0.0308, + "step": 66960 + }, + { + "epoch": 4.387735045534954, + "grad_norm": 0.45942047238349915, + "learning_rate": 2.6978430844848935e-05, + "loss": 0.0458, + "step": 66970 + }, + { + "epoch": 4.388390224726463, + "grad_norm": 0.2625426650047302, + "learning_rate": 2.69637543300991e-05, + "loss": 0.0364, + "step": 66980 + }, + { + "epoch": 4.389045403917971, + "grad_norm": 0.7013683915138245, + "learning_rate": 2.694908033455848e-05, + "loss": 0.0512, + "step": 66990 + }, + { + "epoch": 4.38970058310948, + "grad_norm": 0.8043395280838013, + "learning_rate": 2.6934408859831733e-05, + "loss": 0.0443, + "step": 67000 + }, + { + "epoch": 4.390355762300989, + "grad_norm": 0.10353200137615204, + "learning_rate": 2.69197399075234e-05, + "loss": 0.0151, + "step": 67010 + }, + { + "epoch": 4.391010941492498, + "grad_norm": 0.49519583582878113, + "learning_rate": 2.6905073479237584e-05, + "loss": 0.0293, + "step": 67020 + }, + { + "epoch": 4.3916661206840075, + "grad_norm": 0.12083671987056732, + "learning_rate": 2.6890409576578235e-05, + "loss": 0.0399, + "step": 67030 + }, + { + "epoch": 4.392321299875516, + "grad_norm": 0.6230376362800598, + "learning_rate": 2.687574820114892e-05, + "loss": 0.0272, + "step": 67040 + }, + { + "epoch": 4.392976479067025, + "grad_norm": 0.33145442605018616, + "learning_rate": 2.686108935455305e-05, + "loss": 0.0361, + "step": 67050 + }, + { + "epoch": 4.393631658258534, + "grad_norm": 0.3440625071525574, + "learning_rate": 2.6846433038393603e-05, + "loss": 0.022, + "step": 67060 + }, + { + "epoch": 4.394286837450043, + "grad_norm": 0.7388477325439453, + "learning_rate": 2.6831779254273482e-05, + "loss": 0.0244, + "step": 67070 + }, + { + "epoch": 4.394942016641552, + "grad_norm": 0.6086891889572144, + "learning_rate": 2.6817128003795122e-05, + "loss": 0.0356, + "step": 67080 + }, + { + "epoch": 4.39559719583306, + "grad_norm": 0.9090373516082764, + "learning_rate": 2.680247928856081e-05, + "loss": 0.0348, + "step": 67090 + }, + { + "epoch": 4.396252375024569, + "grad_norm": 0.241951584815979, + "learning_rate": 2.6787833110172466e-05, + "loss": 0.0292, + "step": 67100 + }, + { + "epoch": 4.396907554216078, + "grad_norm": 2.090036153793335, + "learning_rate": 2.6773189470231807e-05, + "loss": 0.03, + "step": 67110 + }, + { + "epoch": 4.397562733407587, + "grad_norm": 0.7588909268379211, + "learning_rate": 2.675854837034022e-05, + "loss": 0.0626, + "step": 67120 + }, + { + "epoch": 4.398217912599096, + "grad_norm": 0.10249993205070496, + "learning_rate": 2.6743909812098833e-05, + "loss": 0.0261, + "step": 67130 + }, + { + "epoch": 4.3988730917906045, + "grad_norm": 0.5037867426872253, + "learning_rate": 2.67292737971085e-05, + "loss": 0.0345, + "step": 67140 + }, + { + "epoch": 4.399528270982113, + "grad_norm": 0.2135668694972992, + "learning_rate": 2.671464032696982e-05, + "loss": 0.0241, + "step": 67150 + }, + { + "epoch": 4.400183450173622, + "grad_norm": 0.3609950840473175, + "learning_rate": 2.6700009403283055e-05, + "loss": 0.021, + "step": 67160 + }, + { + "epoch": 4.400838629365131, + "grad_norm": 0.10619278252124786, + "learning_rate": 2.6685381027648214e-05, + "loss": 0.0363, + "step": 67170 + }, + { + "epoch": 4.401493808556641, + "grad_norm": 0.3115551471710205, + "learning_rate": 2.6670755201665086e-05, + "loss": 0.0365, + "step": 67180 + }, + { + "epoch": 4.4021489877481494, + "grad_norm": 0.9213114976882935, + "learning_rate": 2.6656131926933066e-05, + "loss": 0.0297, + "step": 67190 + }, + { + "epoch": 4.402804166939658, + "grad_norm": 0.14701096713542938, + "learning_rate": 2.664151120505136e-05, + "loss": 0.0289, + "step": 67200 + }, + { + "epoch": 4.403459346131167, + "grad_norm": 0.21323572099208832, + "learning_rate": 2.6626893037618873e-05, + "loss": 0.0184, + "step": 67210 + }, + { + "epoch": 4.404114525322676, + "grad_norm": 0.1122533529996872, + "learning_rate": 2.6612277426234243e-05, + "loss": 0.0209, + "step": 67220 + }, + { + "epoch": 4.404769704514185, + "grad_norm": 0.9660581350326538, + "learning_rate": 2.659766437249577e-05, + "loss": 0.0458, + "step": 67230 + }, + { + "epoch": 4.4054248837056935, + "grad_norm": 0.7179709076881409, + "learning_rate": 2.6583053878001557e-05, + "loss": 0.0469, + "step": 67240 + }, + { + "epoch": 4.406080062897202, + "grad_norm": 0.17361681163311005, + "learning_rate": 2.6568445944349323e-05, + "loss": 0.029, + "step": 67250 + }, + { + "epoch": 4.406735242088711, + "grad_norm": 0.5454289317131042, + "learning_rate": 2.6553840573136657e-05, + "loss": 0.0369, + "step": 67260 + }, + { + "epoch": 4.40739042128022, + "grad_norm": 0.44495850801467896, + "learning_rate": 2.6539237765960706e-05, + "loss": 0.0264, + "step": 67270 + }, + { + "epoch": 4.408045600471729, + "grad_norm": 0.15608975291252136, + "learning_rate": 2.6524637524418465e-05, + "loss": 0.0264, + "step": 67280 + }, + { + "epoch": 4.408700779663238, + "grad_norm": 0.45799800753593445, + "learning_rate": 2.651003985010655e-05, + "loss": 0.0392, + "step": 67290 + }, + { + "epoch": 4.409355958854746, + "grad_norm": 1.0514979362487793, + "learning_rate": 2.649544474462138e-05, + "loss": 0.0274, + "step": 67300 + }, + { + "epoch": 4.410011138046256, + "grad_norm": 0.24385741353034973, + "learning_rate": 2.6480852209559015e-05, + "loss": 0.0356, + "step": 67310 + }, + { + "epoch": 4.410666317237765, + "grad_norm": 0.29192665219306946, + "learning_rate": 2.6466262246515296e-05, + "loss": 0.0183, + "step": 67320 + }, + { + "epoch": 4.411321496429274, + "grad_norm": 1.8894014358520508, + "learning_rate": 2.6451674857085746e-05, + "loss": 0.0227, + "step": 67330 + }, + { + "epoch": 4.4119766756207826, + "grad_norm": 0.5659967064857483, + "learning_rate": 2.6437090042865655e-05, + "loss": 0.0415, + "step": 67340 + }, + { + "epoch": 4.412631854812291, + "grad_norm": 0.11935118585824966, + "learning_rate": 2.6422507805449947e-05, + "loss": 0.0189, + "step": 67350 + }, + { + "epoch": 4.4132870340038, + "grad_norm": 0.5524082779884338, + "learning_rate": 2.6407928146433357e-05, + "loss": 0.0373, + "step": 67360 + }, + { + "epoch": 4.413942213195309, + "grad_norm": 0.30005043745040894, + "learning_rate": 2.639335106741026e-05, + "loss": 0.0257, + "step": 67370 + }, + { + "epoch": 4.414597392386818, + "grad_norm": 0.05548694729804993, + "learning_rate": 2.6378776569974796e-05, + "loss": 0.0379, + "step": 67380 + }, + { + "epoch": 4.415252571578327, + "grad_norm": 0.4953840970993042, + "learning_rate": 2.6364204655720814e-05, + "loss": 0.0316, + "step": 67390 + }, + { + "epoch": 4.415907750769835, + "grad_norm": 0.21590493619441986, + "learning_rate": 2.6349635326241894e-05, + "loss": 0.0431, + "step": 67400 + }, + { + "epoch": 4.416562929961344, + "grad_norm": 0.13456054031848907, + "learning_rate": 2.633506858313127e-05, + "loss": 0.0226, + "step": 67410 + }, + { + "epoch": 4.417218109152853, + "grad_norm": 0.4261609613895416, + "learning_rate": 2.6320504427981975e-05, + "loss": 0.0343, + "step": 67420 + }, + { + "epoch": 4.417873288344362, + "grad_norm": 0.20625342428684235, + "learning_rate": 2.6305942862386734e-05, + "loss": 0.0336, + "step": 67430 + }, + { + "epoch": 4.418528467535871, + "grad_norm": 0.8260899782180786, + "learning_rate": 2.629138388793793e-05, + "loss": 0.0561, + "step": 67440 + }, + { + "epoch": 4.4191836467273795, + "grad_norm": 0.29607775807380676, + "learning_rate": 2.6276827506227743e-05, + "loss": 0.0266, + "step": 67450 + }, + { + "epoch": 4.419838825918889, + "grad_norm": 0.19478265941143036, + "learning_rate": 2.6262273718848037e-05, + "loss": 0.0289, + "step": 67460 + }, + { + "epoch": 4.420494005110398, + "grad_norm": 0.4110245406627655, + "learning_rate": 2.6247722527390407e-05, + "loss": 0.0221, + "step": 67470 + }, + { + "epoch": 4.421149184301907, + "grad_norm": 0.7715611457824707, + "learning_rate": 2.6233173933446104e-05, + "loss": 0.0337, + "step": 67480 + }, + { + "epoch": 4.421804363493416, + "grad_norm": 0.2840760052204132, + "learning_rate": 2.6218627938606188e-05, + "loss": 0.0459, + "step": 67490 + }, + { + "epoch": 4.4224595426849245, + "grad_norm": 1.5183565616607666, + "learning_rate": 2.620408454446135e-05, + "loss": 0.0294, + "step": 67500 + }, + { + "epoch": 4.423114721876433, + "grad_norm": 0.7464296817779541, + "learning_rate": 2.6189543752602046e-05, + "loss": 0.0345, + "step": 67510 + }, + { + "epoch": 4.423769901067942, + "grad_norm": 0.3233148753643036, + "learning_rate": 2.6175005564618444e-05, + "loss": 0.0297, + "step": 67520 + }, + { + "epoch": 4.424425080259451, + "grad_norm": 0.14982911944389343, + "learning_rate": 2.6160469982100428e-05, + "loss": 0.0368, + "step": 67530 + }, + { + "epoch": 4.42508025945096, + "grad_norm": 0.423409640789032, + "learning_rate": 2.6145937006637555e-05, + "loss": 0.0175, + "step": 67540 + }, + { + "epoch": 4.4257354386424685, + "grad_norm": 0.16455744206905365, + "learning_rate": 2.6131406639819166e-05, + "loss": 0.0191, + "step": 67550 + }, + { + "epoch": 4.426390617833977, + "grad_norm": 0.8379818797111511, + "learning_rate": 2.611687888323424e-05, + "loss": 0.0433, + "step": 67560 + }, + { + "epoch": 4.427045797025486, + "grad_norm": 0.3954213261604309, + "learning_rate": 2.610235373847153e-05, + "loss": 0.0318, + "step": 67570 + }, + { + "epoch": 4.427700976216995, + "grad_norm": 0.1293572336435318, + "learning_rate": 2.6087831207119494e-05, + "loss": 0.0318, + "step": 67580 + }, + { + "epoch": 4.428356155408505, + "grad_norm": 0.19043786823749542, + "learning_rate": 2.60733112907663e-05, + "loss": 0.0123, + "step": 67590 + }, + { + "epoch": 4.4290113346000135, + "grad_norm": 0.08545953035354614, + "learning_rate": 2.6058793990999786e-05, + "loss": 0.0252, + "step": 67600 + }, + { + "epoch": 4.429666513791522, + "grad_norm": 0.8611860275268555, + "learning_rate": 2.6044279309407593e-05, + "loss": 0.0185, + "step": 67610 + }, + { + "epoch": 4.430321692983031, + "grad_norm": 0.17470654845237732, + "learning_rate": 2.6029767247576976e-05, + "loss": 0.0379, + "step": 67620 + }, + { + "epoch": 4.43097687217454, + "grad_norm": 0.24872876703739166, + "learning_rate": 2.601525780709497e-05, + "loss": 0.0202, + "step": 67630 + }, + { + "epoch": 4.431632051366049, + "grad_norm": 0.5313383936882019, + "learning_rate": 2.6000750989548318e-05, + "loss": 0.0335, + "step": 67640 + }, + { + "epoch": 4.432287230557558, + "grad_norm": 0.21663798391819, + "learning_rate": 2.5986246796523462e-05, + "loss": 0.0339, + "step": 67650 + }, + { + "epoch": 4.432942409749066, + "grad_norm": 0.21112385392189026, + "learning_rate": 2.5971745229606546e-05, + "loss": 0.0297, + "step": 67660 + }, + { + "epoch": 4.433597588940575, + "grad_norm": 0.2930945158004761, + "learning_rate": 2.5957246290383442e-05, + "loss": 0.0283, + "step": 67670 + }, + { + "epoch": 4.434252768132084, + "grad_norm": 0.5585311651229858, + "learning_rate": 2.594274998043975e-05, + "loss": 0.0211, + "step": 67680 + }, + { + "epoch": 4.434907947323593, + "grad_norm": 0.34526076912879944, + "learning_rate": 2.5928256301360733e-05, + "loss": 0.0306, + "step": 67690 + }, + { + "epoch": 4.435563126515102, + "grad_norm": 0.10247824341058731, + "learning_rate": 2.591376525473142e-05, + "loss": 0.0252, + "step": 67700 + }, + { + "epoch": 4.4362183057066105, + "grad_norm": 0.6475171446800232, + "learning_rate": 2.5899276842136523e-05, + "loss": 0.0216, + "step": 67710 + }, + { + "epoch": 4.436873484898119, + "grad_norm": 0.5517793893814087, + "learning_rate": 2.5884791065160495e-05, + "loss": 0.0336, + "step": 67720 + }, + { + "epoch": 4.437528664089628, + "grad_norm": 0.858226478099823, + "learning_rate": 2.5870307925387436e-05, + "loss": 0.0245, + "step": 67730 + }, + { + "epoch": 4.438183843281138, + "grad_norm": 0.9629794359207153, + "learning_rate": 2.585582742440125e-05, + "loss": 0.0616, + "step": 67740 + }, + { + "epoch": 4.438839022472647, + "grad_norm": 0.6380334496498108, + "learning_rate": 2.5841349563785455e-05, + "loss": 0.023, + "step": 67750 + }, + { + "epoch": 4.439494201664155, + "grad_norm": 0.10349663347005844, + "learning_rate": 2.5826874345123355e-05, + "loss": 0.0123, + "step": 67760 + }, + { + "epoch": 4.440149380855664, + "grad_norm": 0.2132169008255005, + "learning_rate": 2.581240176999792e-05, + "loss": 0.0381, + "step": 67770 + }, + { + "epoch": 4.440804560047173, + "grad_norm": 0.2612685561180115, + "learning_rate": 2.579793183999189e-05, + "loss": 0.0265, + "step": 67780 + }, + { + "epoch": 4.441459739238682, + "grad_norm": 0.9460596442222595, + "learning_rate": 2.5783464556687627e-05, + "loss": 0.0575, + "step": 67790 + }, + { + "epoch": 4.442114918430191, + "grad_norm": 0.7883065938949585, + "learning_rate": 2.576899992166729e-05, + "loss": 0.0449, + "step": 67800 + }, + { + "epoch": 4.4427700976216995, + "grad_norm": 0.19601905345916748, + "learning_rate": 2.5754537936512668e-05, + "loss": 0.0188, + "step": 67810 + }, + { + "epoch": 4.443425276813208, + "grad_norm": 0.24947553873062134, + "learning_rate": 2.574007860280535e-05, + "loss": 0.0392, + "step": 67820 + }, + { + "epoch": 4.444080456004717, + "grad_norm": 0.2125261127948761, + "learning_rate": 2.5725621922126518e-05, + "loss": 0.0321, + "step": 67830 + }, + { + "epoch": 4.444735635196226, + "grad_norm": 0.1263437420129776, + "learning_rate": 2.571116789605722e-05, + "loss": 0.0238, + "step": 67840 + }, + { + "epoch": 4.445390814387735, + "grad_norm": 0.3694451153278351, + "learning_rate": 2.5696716526178054e-05, + "loss": 0.0226, + "step": 67850 + }, + { + "epoch": 4.446045993579244, + "grad_norm": 0.9474707245826721, + "learning_rate": 2.568226781406946e-05, + "loss": 0.0399, + "step": 67860 + }, + { + "epoch": 4.446701172770752, + "grad_norm": 0.05911270156502724, + "learning_rate": 2.5667821761311474e-05, + "loss": 0.0202, + "step": 67870 + }, + { + "epoch": 4.447356351962262, + "grad_norm": 0.47288474440574646, + "learning_rate": 2.5653378369483917e-05, + "loss": 0.0287, + "step": 67880 + }, + { + "epoch": 4.448011531153771, + "grad_norm": 1.708572506904602, + "learning_rate": 2.5638937640166292e-05, + "loss": 0.0402, + "step": 67890 + }, + { + "epoch": 4.44866671034528, + "grad_norm": 0.21847203373908997, + "learning_rate": 2.5624499574937842e-05, + "loss": 0.0446, + "step": 67900 + }, + { + "epoch": 4.4493218895367885, + "grad_norm": 0.6938797831535339, + "learning_rate": 2.5610064175377456e-05, + "loss": 0.0266, + "step": 67910 + }, + { + "epoch": 4.449977068728297, + "grad_norm": 0.48619452118873596, + "learning_rate": 2.5595631443063782e-05, + "loss": 0.0313, + "step": 67920 + }, + { + "epoch": 4.450632247919806, + "grad_norm": 0.6182704567909241, + "learning_rate": 2.5581201379575177e-05, + "loss": 0.0499, + "step": 67930 + }, + { + "epoch": 4.451287427111315, + "grad_norm": 0.24063509702682495, + "learning_rate": 2.5566773986489655e-05, + "loss": 0.0365, + "step": 67940 + }, + { + "epoch": 4.451942606302824, + "grad_norm": 0.5031647086143494, + "learning_rate": 2.5552349265385e-05, + "loss": 0.0516, + "step": 67950 + }, + { + "epoch": 4.452597785494333, + "grad_norm": 0.5221884846687317, + "learning_rate": 2.5537927217838675e-05, + "loss": 0.0226, + "step": 67960 + }, + { + "epoch": 4.453252964685841, + "grad_norm": 0.5317400693893433, + "learning_rate": 2.5523507845427864e-05, + "loss": 0.0219, + "step": 67970 + }, + { + "epoch": 4.45390814387735, + "grad_norm": 0.07628556340932846, + "learning_rate": 2.5509091149729413e-05, + "loss": 0.0175, + "step": 67980 + }, + { + "epoch": 4.454563323068859, + "grad_norm": 0.40500620007514954, + "learning_rate": 2.5494677132319962e-05, + "loss": 0.0331, + "step": 67990 + }, + { + "epoch": 4.455218502260368, + "grad_norm": 0.7620886564254761, + "learning_rate": 2.548026579477575e-05, + "loss": 0.0411, + "step": 68000 + }, + { + "epoch": 4.455873681451877, + "grad_norm": 0.3817470371723175, + "learning_rate": 2.5465857138672834e-05, + "loss": 0.0371, + "step": 68010 + }, + { + "epoch": 4.456528860643386, + "grad_norm": 0.30520009994506836, + "learning_rate": 2.5451451165586848e-05, + "loss": 0.0211, + "step": 68020 + }, + { + "epoch": 4.457184039834895, + "grad_norm": 0.35867515206336975, + "learning_rate": 2.543704787709329e-05, + "loss": 0.0523, + "step": 68030 + }, + { + "epoch": 4.457839219026404, + "grad_norm": 0.7231789827346802, + "learning_rate": 2.542264727476723e-05, + "loss": 0.0771, + "step": 68040 + }, + { + "epoch": 4.458494398217913, + "grad_norm": 0.0307331345975399, + "learning_rate": 2.5408249360183535e-05, + "loss": 0.062, + "step": 68050 + }, + { + "epoch": 4.459149577409422, + "grad_norm": 0.5548707842826843, + "learning_rate": 2.5393854134916694e-05, + "loss": 0.0308, + "step": 68060 + }, + { + "epoch": 4.4598047566009305, + "grad_norm": 0.1442033350467682, + "learning_rate": 2.5379461600540993e-05, + "loss": 0.0235, + "step": 68070 + }, + { + "epoch": 4.460459935792439, + "grad_norm": 1.789230227470398, + "learning_rate": 2.5365071758630317e-05, + "loss": 0.061, + "step": 68080 + }, + { + "epoch": 4.461115114983948, + "grad_norm": 0.23988871276378632, + "learning_rate": 2.535068461075839e-05, + "loss": 0.0323, + "step": 68090 + }, + { + "epoch": 4.461770294175457, + "grad_norm": 0.16869333386421204, + "learning_rate": 2.5336300158498516e-05, + "loss": 0.0247, + "step": 68100 + }, + { + "epoch": 4.462425473366966, + "grad_norm": 0.32598772644996643, + "learning_rate": 2.5321918403423793e-05, + "loss": 0.0276, + "step": 68110 + }, + { + "epoch": 4.4630806525584745, + "grad_norm": 0.16842301189899445, + "learning_rate": 2.5307539347106957e-05, + "loss": 0.0272, + "step": 68120 + }, + { + "epoch": 4.463735831749983, + "grad_norm": 0.6617675423622131, + "learning_rate": 2.529316299112049e-05, + "loss": 0.029, + "step": 68130 + }, + { + "epoch": 4.464391010941492, + "grad_norm": 0.20330792665481567, + "learning_rate": 2.5278789337036567e-05, + "loss": 0.0313, + "step": 68140 + }, + { + "epoch": 4.465046190133001, + "grad_norm": 0.6220073103904724, + "learning_rate": 2.52644183864271e-05, + "loss": 0.0269, + "step": 68150 + }, + { + "epoch": 4.46570136932451, + "grad_norm": 0.7258641719818115, + "learning_rate": 2.525005014086363e-05, + "loss": 0.0233, + "step": 68160 + }, + { + "epoch": 4.4663565485160195, + "grad_norm": 0.2928757071495056, + "learning_rate": 2.5235684601917465e-05, + "loss": 0.0288, + "step": 68170 + }, + { + "epoch": 4.467011727707528, + "grad_norm": 0.6490404009819031, + "learning_rate": 2.5221321771159618e-05, + "loss": 0.023, + "step": 68180 + }, + { + "epoch": 4.467666906899037, + "grad_norm": 0.47122544050216675, + "learning_rate": 2.5206961650160743e-05, + "loss": 0.0489, + "step": 68190 + }, + { + "epoch": 4.468322086090546, + "grad_norm": 0.73904949426651, + "learning_rate": 2.519260424049129e-05, + "loss": 0.0224, + "step": 68200 + }, + { + "epoch": 4.468977265282055, + "grad_norm": 0.21407043933868408, + "learning_rate": 2.5178249543721304e-05, + "loss": 0.0154, + "step": 68210 + }, + { + "epoch": 4.469632444473564, + "grad_norm": 0.6343767046928406, + "learning_rate": 2.5163897561420668e-05, + "loss": 0.0275, + "step": 68220 + }, + { + "epoch": 4.470287623665072, + "grad_norm": 0.2346104085445404, + "learning_rate": 2.514954829515883e-05, + "loss": 0.0185, + "step": 68230 + }, + { + "epoch": 4.470942802856581, + "grad_norm": 1.2442455291748047, + "learning_rate": 2.513520174650505e-05, + "loss": 0.0744, + "step": 68240 + }, + { + "epoch": 4.47159798204809, + "grad_norm": 1.9991649389266968, + "learning_rate": 2.5120857917028207e-05, + "loss": 0.0194, + "step": 68250 + }, + { + "epoch": 4.472253161239599, + "grad_norm": 0.18024544417858124, + "learning_rate": 2.5106516808296953e-05, + "loss": 0.0196, + "step": 68260 + }, + { + "epoch": 4.472908340431108, + "grad_norm": 0.1920957863330841, + "learning_rate": 2.5092178421879552e-05, + "loss": 0.0305, + "step": 68270 + }, + { + "epoch": 4.4735635196226164, + "grad_norm": 0.7782745361328125, + "learning_rate": 2.5077842759344116e-05, + "loss": 0.0226, + "step": 68280 + }, + { + "epoch": 4.474218698814125, + "grad_norm": 0.2270762026309967, + "learning_rate": 2.5063509822258304e-05, + "loss": 0.0423, + "step": 68290 + }, + { + "epoch": 4.474873878005635, + "grad_norm": 0.16381363570690155, + "learning_rate": 2.5049179612189576e-05, + "loss": 0.0206, + "step": 68300 + }, + { + "epoch": 4.475529057197144, + "grad_norm": 0.48981615900993347, + "learning_rate": 2.5034852130705043e-05, + "loss": 0.0368, + "step": 68310 + }, + { + "epoch": 4.476184236388653, + "grad_norm": 1.2866857051849365, + "learning_rate": 2.5020527379371572e-05, + "loss": 0.0404, + "step": 68320 + }, + { + "epoch": 4.476839415580161, + "grad_norm": 0.212611123919487, + "learning_rate": 2.500620535975564e-05, + "loss": 0.0251, + "step": 68330 + }, + { + "epoch": 4.47749459477167, + "grad_norm": 0.21499744057655334, + "learning_rate": 2.4991886073423515e-05, + "loss": 0.0329, + "step": 68340 + }, + { + "epoch": 4.478149773963179, + "grad_norm": 0.3782069683074951, + "learning_rate": 2.4977569521941134e-05, + "loss": 0.0385, + "step": 68350 + }, + { + "epoch": 4.478804953154688, + "grad_norm": 0.187188059091568, + "learning_rate": 2.4963255706874143e-05, + "loss": 0.0198, + "step": 68360 + }, + { + "epoch": 4.479460132346197, + "grad_norm": 0.47441211342811584, + "learning_rate": 2.4948944629787853e-05, + "loss": 0.024, + "step": 68370 + }, + { + "epoch": 4.4801153115377055, + "grad_norm": 1.6790831089019775, + "learning_rate": 2.493463629224731e-05, + "loss": 0.0383, + "step": 68380 + }, + { + "epoch": 4.480770490729214, + "grad_norm": 0.8175102472305298, + "learning_rate": 2.492033069581728e-05, + "loss": 0.0381, + "step": 68390 + }, + { + "epoch": 4.481425669920723, + "grad_norm": 0.4740045666694641, + "learning_rate": 2.4906027842062137e-05, + "loss": 0.031, + "step": 68400 + }, + { + "epoch": 4.482080849112232, + "grad_norm": 0.9189004898071289, + "learning_rate": 2.48917277325461e-05, + "loss": 0.0384, + "step": 68410 + }, + { + "epoch": 4.482736028303741, + "grad_norm": 1.2374682426452637, + "learning_rate": 2.487743036883295e-05, + "loss": 0.0321, + "step": 68420 + }, + { + "epoch": 4.4833912074952496, + "grad_norm": 0.3788973391056061, + "learning_rate": 2.486313575248626e-05, + "loss": 0.0207, + "step": 68430 + }, + { + "epoch": 4.484046386686758, + "grad_norm": 0.6840598583221436, + "learning_rate": 2.4848843885069235e-05, + "loss": 0.0459, + "step": 68440 + }, + { + "epoch": 4.484701565878268, + "grad_norm": 0.6011937260627747, + "learning_rate": 2.483455476814484e-05, + "loss": 0.0242, + "step": 68450 + }, + { + "epoch": 4.485356745069777, + "grad_norm": 0.47130286693573, + "learning_rate": 2.4820268403275672e-05, + "loss": 0.0334, + "step": 68460 + }, + { + "epoch": 4.486011924261286, + "grad_norm": 0.3220856487751007, + "learning_rate": 2.480598479202413e-05, + "loss": 0.0299, + "step": 68470 + }, + { + "epoch": 4.4866671034527945, + "grad_norm": 0.6921897530555725, + "learning_rate": 2.4791703935952193e-05, + "loss": 0.0367, + "step": 68480 + }, + { + "epoch": 4.487322282644303, + "grad_norm": 0.18787673115730286, + "learning_rate": 2.477742583662163e-05, + "loss": 0.0224, + "step": 68490 + }, + { + "epoch": 4.487977461835812, + "grad_norm": 0.08877217024564743, + "learning_rate": 2.476315049559384e-05, + "loss": 0.0227, + "step": 68500 + }, + { + "epoch": 4.488632641027321, + "grad_norm": 0.4273446798324585, + "learning_rate": 2.4748877914429987e-05, + "loss": 0.0402, + "step": 68510 + }, + { + "epoch": 4.48928782021883, + "grad_norm": 0.42853209376335144, + "learning_rate": 2.4734608094690857e-05, + "loss": 0.0312, + "step": 68520 + }, + { + "epoch": 4.489942999410339, + "grad_norm": 0.2868867814540863, + "learning_rate": 2.4720341037937006e-05, + "loss": 0.0255, + "step": 68530 + }, + { + "epoch": 4.490598178601847, + "grad_norm": 0.8195233345031738, + "learning_rate": 2.4706076745728645e-05, + "loss": 0.0207, + "step": 68540 + }, + { + "epoch": 4.491253357793356, + "grad_norm": 1.610292673110962, + "learning_rate": 2.469181521962573e-05, + "loss": 0.0398, + "step": 68550 + }, + { + "epoch": 4.491908536984865, + "grad_norm": 0.32999709248542786, + "learning_rate": 2.467755646118783e-05, + "loss": 0.0283, + "step": 68560 + }, + { + "epoch": 4.492563716176374, + "grad_norm": 0.39405012130737305, + "learning_rate": 2.46633004719743e-05, + "loss": 0.0413, + "step": 68570 + }, + { + "epoch": 4.4932188953678835, + "grad_norm": 0.42464181780815125, + "learning_rate": 2.4649047253544123e-05, + "loss": 0.0225, + "step": 68580 + }, + { + "epoch": 4.493874074559392, + "grad_norm": 1.073311448097229, + "learning_rate": 2.4634796807456016e-05, + "loss": 0.0478, + "step": 68590 + }, + { + "epoch": 4.494529253750901, + "grad_norm": 0.3661409318447113, + "learning_rate": 2.46205491352684e-05, + "loss": 0.0261, + "step": 68600 + }, + { + "epoch": 4.49518443294241, + "grad_norm": 0.22669662535190582, + "learning_rate": 2.4606304238539375e-05, + "loss": 0.0218, + "step": 68610 + }, + { + "epoch": 4.495839612133919, + "grad_norm": 0.16275714337825775, + "learning_rate": 2.4592062118826753e-05, + "loss": 0.0267, + "step": 68620 + }, + { + "epoch": 4.496494791325428, + "grad_norm": 0.7632006406784058, + "learning_rate": 2.4577822777688004e-05, + "loss": 0.04, + "step": 68630 + }, + { + "epoch": 4.497149970516936, + "grad_norm": 0.13363026082515717, + "learning_rate": 2.4563586216680347e-05, + "loss": 0.0324, + "step": 68640 + }, + { + "epoch": 4.497805149708445, + "grad_norm": 0.9366847276687622, + "learning_rate": 2.4549352437360622e-05, + "loss": 0.0652, + "step": 68650 + }, + { + "epoch": 4.498460328899954, + "grad_norm": 0.29669052362442017, + "learning_rate": 2.4535121441285493e-05, + "loss": 0.0332, + "step": 68660 + }, + { + "epoch": 4.499115508091463, + "grad_norm": 0.15634547173976898, + "learning_rate": 2.4520893230011172e-05, + "loss": 0.0212, + "step": 68670 + }, + { + "epoch": 4.499770687282972, + "grad_norm": 1.220150351524353, + "learning_rate": 2.450666780509368e-05, + "loss": 0.0291, + "step": 68680 + }, + { + "epoch": 4.5004258664744805, + "grad_norm": 0.20737308263778687, + "learning_rate": 2.4492445168088656e-05, + "loss": 0.0204, + "step": 68690 + }, + { + "epoch": 4.501081045665989, + "grad_norm": 0.22454969584941864, + "learning_rate": 2.447822532055149e-05, + "loss": 0.0287, + "step": 68700 + }, + { + "epoch": 4.501736224857499, + "grad_norm": 0.29989588260650635, + "learning_rate": 2.4464008264037212e-05, + "loss": 0.029, + "step": 68710 + }, + { + "epoch": 4.502391404049007, + "grad_norm": 0.4106268882751465, + "learning_rate": 2.4449794000100605e-05, + "loss": 0.0275, + "step": 68720 + }, + { + "epoch": 4.503046583240517, + "grad_norm": 0.3042738437652588, + "learning_rate": 2.4435582530296108e-05, + "loss": 0.0567, + "step": 68730 + }, + { + "epoch": 4.5037017624320255, + "grad_norm": 0.4475792944431305, + "learning_rate": 2.4421373856177887e-05, + "loss": 0.0278, + "step": 68740 + }, + { + "epoch": 4.504356941623534, + "grad_norm": 0.14292724430561066, + "learning_rate": 2.440716797929975e-05, + "loss": 0.0438, + "step": 68750 + }, + { + "epoch": 4.505012120815043, + "grad_norm": 0.28947678208351135, + "learning_rate": 2.439296490121526e-05, + "loss": 0.0191, + "step": 68760 + }, + { + "epoch": 4.505667300006552, + "grad_norm": 0.9341912269592285, + "learning_rate": 2.4378764623477608e-05, + "loss": 0.0436, + "step": 68770 + }, + { + "epoch": 4.506322479198061, + "grad_norm": 0.34225139021873474, + "learning_rate": 2.436456714763974e-05, + "loss": 0.0325, + "step": 68780 + }, + { + "epoch": 4.5069776583895695, + "grad_norm": 0.595474123954773, + "learning_rate": 2.4350372475254262e-05, + "loss": 0.0223, + "step": 68790 + }, + { + "epoch": 4.507632837581078, + "grad_norm": 0.44050055742263794, + "learning_rate": 2.433618060787351e-05, + "loss": 0.0291, + "step": 68800 + }, + { + "epoch": 4.508288016772587, + "grad_norm": 0.5544010996818542, + "learning_rate": 2.432199154704944e-05, + "loss": 0.0316, + "step": 68810 + }, + { + "epoch": 4.508943195964096, + "grad_norm": 0.22689825296401978, + "learning_rate": 2.430780529433377e-05, + "loss": 0.0284, + "step": 68820 + }, + { + "epoch": 4.509598375155605, + "grad_norm": 0.2961742877960205, + "learning_rate": 2.42936218512779e-05, + "loss": 0.0335, + "step": 68830 + }, + { + "epoch": 4.510253554347114, + "grad_norm": 0.7500815391540527, + "learning_rate": 2.427944121943288e-05, + "loss": 0.0259, + "step": 68840 + }, + { + "epoch": 4.510908733538622, + "grad_norm": 0.32544732093811035, + "learning_rate": 2.4265263400349497e-05, + "loss": 0.0255, + "step": 68850 + }, + { + "epoch": 4.511563912730132, + "grad_norm": 0.14118434488773346, + "learning_rate": 2.4251088395578213e-05, + "loss": 0.0271, + "step": 68860 + }, + { + "epoch": 4.512219091921641, + "grad_norm": 0.3902077078819275, + "learning_rate": 2.4236916206669214e-05, + "loss": 0.0207, + "step": 68870 + }, + { + "epoch": 4.51287427111315, + "grad_norm": 0.22876659035682678, + "learning_rate": 2.4222746835172304e-05, + "loss": 0.032, + "step": 68880 + }, + { + "epoch": 4.513529450304659, + "grad_norm": 0.4442562758922577, + "learning_rate": 2.4208580282637065e-05, + "loss": 0.0256, + "step": 68890 + }, + { + "epoch": 4.514184629496167, + "grad_norm": 0.1265745759010315, + "learning_rate": 2.4194416550612687e-05, + "loss": 0.0237, + "step": 68900 + }, + { + "epoch": 4.514839808687676, + "grad_norm": 0.2213258147239685, + "learning_rate": 2.418025564064812e-05, + "loss": 0.0272, + "step": 68910 + }, + { + "epoch": 4.515494987879185, + "grad_norm": 0.5927059650421143, + "learning_rate": 2.4166097554291978e-05, + "loss": 0.0286, + "step": 68920 + }, + { + "epoch": 4.516150167070694, + "grad_norm": 0.38348886370658875, + "learning_rate": 2.4151942293092583e-05, + "loss": 0.0415, + "step": 68930 + }, + { + "epoch": 4.516805346262203, + "grad_norm": 0.393124520778656, + "learning_rate": 2.41377898585979e-05, + "loss": 0.0293, + "step": 68940 + }, + { + "epoch": 4.5174605254537115, + "grad_norm": 0.08768188953399658, + "learning_rate": 2.4123640252355652e-05, + "loss": 0.0246, + "step": 68950 + }, + { + "epoch": 4.51811570464522, + "grad_norm": 0.4613557755947113, + "learning_rate": 2.410949347591318e-05, + "loss": 0.0427, + "step": 68960 + }, + { + "epoch": 4.518770883836729, + "grad_norm": 0.5729779601097107, + "learning_rate": 2.4095349530817578e-05, + "loss": 0.0346, + "step": 68970 + }, + { + "epoch": 4.519426063028238, + "grad_norm": 0.6727737188339233, + "learning_rate": 2.4081208418615603e-05, + "loss": 0.0538, + "step": 68980 + }, + { + "epoch": 4.520081242219747, + "grad_norm": 0.498944491147995, + "learning_rate": 2.4067070140853725e-05, + "loss": 0.0221, + "step": 68990 + }, + { + "epoch": 4.5207364214112555, + "grad_norm": 0.470039963722229, + "learning_rate": 2.405293469907805e-05, + "loss": 0.0365, + "step": 69000 + }, + { + "epoch": 4.521391600602765, + "grad_norm": 0.8845431208610535, + "learning_rate": 2.4038802094834446e-05, + "loss": 0.0357, + "step": 69010 + }, + { + "epoch": 4.522046779794274, + "grad_norm": 0.2704201340675354, + "learning_rate": 2.4024672329668395e-05, + "loss": 0.0227, + "step": 69020 + }, + { + "epoch": 4.522701958985783, + "grad_norm": 1.0852001905441284, + "learning_rate": 2.4010545405125125e-05, + "loss": 0.0564, + "step": 69030 + }, + { + "epoch": 4.523357138177292, + "grad_norm": 0.21993814408779144, + "learning_rate": 2.3996421322749528e-05, + "loss": 0.0211, + "step": 69040 + }, + { + "epoch": 4.5240123173688005, + "grad_norm": 0.27819934487342834, + "learning_rate": 2.3982300084086224e-05, + "loss": 0.0271, + "step": 69050 + }, + { + "epoch": 4.524667496560309, + "grad_norm": 0.6975709795951843, + "learning_rate": 2.396818169067944e-05, + "loss": 0.0386, + "step": 69060 + }, + { + "epoch": 4.525322675751818, + "grad_norm": 0.587004542350769, + "learning_rate": 2.3954066144073178e-05, + "loss": 0.0213, + "step": 69070 + }, + { + "epoch": 4.525977854943327, + "grad_norm": 0.057401835918426514, + "learning_rate": 2.39399534458111e-05, + "loss": 0.0159, + "step": 69080 + }, + { + "epoch": 4.526633034134836, + "grad_norm": 0.44132524728775024, + "learning_rate": 2.392584359743651e-05, + "loss": 0.0228, + "step": 69090 + }, + { + "epoch": 4.527288213326345, + "grad_norm": 0.28979894518852234, + "learning_rate": 2.3911736600492463e-05, + "loss": 0.0349, + "step": 69100 + }, + { + "epoch": 4.527943392517853, + "grad_norm": 0.5232745409011841, + "learning_rate": 2.3897632456521672e-05, + "loss": 0.0231, + "step": 69110 + }, + { + "epoch": 4.528598571709362, + "grad_norm": 1.2491205930709839, + "learning_rate": 2.388353116706658e-05, + "loss": 0.027, + "step": 69120 + }, + { + "epoch": 4.529253750900871, + "grad_norm": 0.3849172592163086, + "learning_rate": 2.3869432733669232e-05, + "loss": 0.0305, + "step": 69130 + }, + { + "epoch": 4.529908930092381, + "grad_norm": 0.8072547912597656, + "learning_rate": 2.385533715787145e-05, + "loss": 0.0408, + "step": 69140 + }, + { + "epoch": 4.5305641092838895, + "grad_norm": 0.16134601831436157, + "learning_rate": 2.3841244441214673e-05, + "loss": 0.0461, + "step": 69150 + }, + { + "epoch": 4.531219288475398, + "grad_norm": 1.369869589805603, + "learning_rate": 2.3827154585240073e-05, + "loss": 0.0479, + "step": 69160 + }, + { + "epoch": 4.531874467666907, + "grad_norm": 0.38767483830451965, + "learning_rate": 2.3813067591488498e-05, + "loss": 0.0194, + "step": 69170 + }, + { + "epoch": 4.532529646858416, + "grad_norm": 0.7880676984786987, + "learning_rate": 2.3798983461500494e-05, + "loss": 0.0286, + "step": 69180 + }, + { + "epoch": 4.533184826049925, + "grad_norm": 0.2427201271057129, + "learning_rate": 2.3784902196816256e-05, + "loss": 0.0321, + "step": 69190 + }, + { + "epoch": 4.533840005241434, + "grad_norm": 0.22795045375823975, + "learning_rate": 2.3770823798975716e-05, + "loss": 0.0137, + "step": 69200 + }, + { + "epoch": 4.534495184432942, + "grad_norm": 1.9986108541488647, + "learning_rate": 2.3756748269518424e-05, + "loss": 0.0539, + "step": 69210 + }, + { + "epoch": 4.535150363624451, + "grad_norm": 0.32617008686065674, + "learning_rate": 2.374267560998371e-05, + "loss": 0.0228, + "step": 69220 + }, + { + "epoch": 4.53580554281596, + "grad_norm": 0.26084670424461365, + "learning_rate": 2.3728605821910477e-05, + "loss": 0.0205, + "step": 69230 + }, + { + "epoch": 4.536460722007469, + "grad_norm": 0.5637553930282593, + "learning_rate": 2.3714538906837453e-05, + "loss": 0.0428, + "step": 69240 + }, + { + "epoch": 4.537115901198978, + "grad_norm": 0.10174375027418137, + "learning_rate": 2.3700474866302913e-05, + "loss": 0.0171, + "step": 69250 + }, + { + "epoch": 4.5377710803904865, + "grad_norm": 0.2813628613948822, + "learning_rate": 2.3686413701844917e-05, + "loss": 0.0139, + "step": 69260 + }, + { + "epoch": 4.538426259581995, + "grad_norm": 0.5793545246124268, + "learning_rate": 2.367235541500114e-05, + "loss": 0.0357, + "step": 69270 + }, + { + "epoch": 4.539081438773504, + "grad_norm": 0.22757680714130402, + "learning_rate": 2.3658300007308993e-05, + "loss": 0.0264, + "step": 69280 + }, + { + "epoch": 4.539736617965014, + "grad_norm": 0.4029390811920166, + "learning_rate": 2.364424748030555e-05, + "loss": 0.0179, + "step": 69290 + }, + { + "epoch": 4.540391797156523, + "grad_norm": 0.38916462659835815, + "learning_rate": 2.3630197835527595e-05, + "loss": 0.0333, + "step": 69300 + }, + { + "epoch": 4.5410469763480314, + "grad_norm": 0.53097003698349, + "learning_rate": 2.3616151074511546e-05, + "loss": 0.0236, + "step": 69310 + }, + { + "epoch": 4.54170215553954, + "grad_norm": 1.4785224199295044, + "learning_rate": 2.3602107198793545e-05, + "loss": 0.0194, + "step": 69320 + }, + { + "epoch": 4.542357334731049, + "grad_norm": 0.5752726197242737, + "learning_rate": 2.3588066209909427e-05, + "loss": 0.0331, + "step": 69330 + }, + { + "epoch": 4.543012513922558, + "grad_norm": 0.34958815574645996, + "learning_rate": 2.3574028109394662e-05, + "loss": 0.0309, + "step": 69340 + }, + { + "epoch": 4.543667693114067, + "grad_norm": 0.38987091183662415, + "learning_rate": 2.3559992898784454e-05, + "loss": 0.0247, + "step": 69350 + }, + { + "epoch": 4.5443228723055755, + "grad_norm": 0.21123415231704712, + "learning_rate": 2.354596057961367e-05, + "loss": 0.0299, + "step": 69360 + }, + { + "epoch": 4.544978051497084, + "grad_norm": 0.12909694015979767, + "learning_rate": 2.353193115341688e-05, + "loss": 0.034, + "step": 69370 + }, + { + "epoch": 4.545633230688593, + "grad_norm": 0.5999157428741455, + "learning_rate": 2.3517904621728294e-05, + "loss": 0.0318, + "step": 69380 + }, + { + "epoch": 4.546288409880102, + "grad_norm": 0.2720537781715393, + "learning_rate": 2.3503880986081855e-05, + "loss": 0.0167, + "step": 69390 + }, + { + "epoch": 4.546943589071611, + "grad_norm": 0.10943720489740372, + "learning_rate": 2.3489860248011136e-05, + "loss": 0.0221, + "step": 69400 + }, + { + "epoch": 4.54759876826312, + "grad_norm": 0.13533294200897217, + "learning_rate": 2.3475842409049464e-05, + "loss": 0.0303, + "step": 69410 + }, + { + "epoch": 4.548253947454629, + "grad_norm": 0.6001183390617371, + "learning_rate": 2.346182747072975e-05, + "loss": 0.0393, + "step": 69420 + }, + { + "epoch": 4.548909126646137, + "grad_norm": 0.1503194272518158, + "learning_rate": 2.3447815434584718e-05, + "loss": 0.0388, + "step": 69430 + }, + { + "epoch": 4.549564305837647, + "grad_norm": 0.8213385343551636, + "learning_rate": 2.3433806302146656e-05, + "loss": 0.0338, + "step": 69440 + }, + { + "epoch": 4.550219485029156, + "grad_norm": 0.6067739129066467, + "learning_rate": 2.3419800074947612e-05, + "loss": 0.0222, + "step": 69450 + }, + { + "epoch": 4.550874664220665, + "grad_norm": 0.2927105724811554, + "learning_rate": 2.3405796754519248e-05, + "loss": 0.0227, + "step": 69460 + }, + { + "epoch": 4.551529843412173, + "grad_norm": 0.45872703194618225, + "learning_rate": 2.3391796342392992e-05, + "loss": 0.0241, + "step": 69470 + }, + { + "epoch": 4.552185022603682, + "grad_norm": 1.0155417919158936, + "learning_rate": 2.337779884009984e-05, + "loss": 0.0317, + "step": 69480 + }, + { + "epoch": 4.552840201795191, + "grad_norm": 0.0762157291173935, + "learning_rate": 2.3363804249170624e-05, + "loss": 0.041, + "step": 69490 + }, + { + "epoch": 4.5534953809867, + "grad_norm": 0.2359875738620758, + "learning_rate": 2.3349812571135703e-05, + "loss": 0.0368, + "step": 69500 + }, + { + "epoch": 4.554150560178209, + "grad_norm": 0.830368161201477, + "learning_rate": 2.333582380752523e-05, + "loss": 0.0222, + "step": 69510 + }, + { + "epoch": 4.554805739369717, + "grad_norm": 0.1765897423028946, + "learning_rate": 2.3321837959868963e-05, + "loss": 0.0346, + "step": 69520 + }, + { + "epoch": 4.555460918561226, + "grad_norm": 0.286818265914917, + "learning_rate": 2.3307855029696383e-05, + "loss": 0.0202, + "step": 69530 + }, + { + "epoch": 4.556116097752735, + "grad_norm": 0.949347734451294, + "learning_rate": 2.3293875018536648e-05, + "loss": 0.0363, + "step": 69540 + }, + { + "epoch": 4.556771276944244, + "grad_norm": 0.575324535369873, + "learning_rate": 2.327989792791861e-05, + "loss": 0.0247, + "step": 69550 + }, + { + "epoch": 4.557426456135753, + "grad_norm": 0.1086859405040741, + "learning_rate": 2.326592375937074e-05, + "loss": 0.0301, + "step": 69560 + }, + { + "epoch": 4.558081635327262, + "grad_norm": 0.13772396743297577, + "learning_rate": 2.3251952514421248e-05, + "loss": 0.0477, + "step": 69570 + }, + { + "epoch": 4.558736814518771, + "grad_norm": 0.31161633133888245, + "learning_rate": 2.323798419459804e-05, + "loss": 0.0273, + "step": 69580 + }, + { + "epoch": 4.55939199371028, + "grad_norm": 0.8972400426864624, + "learning_rate": 2.3224018801428625e-05, + "loss": 0.0622, + "step": 69590 + }, + { + "epoch": 4.560047172901789, + "grad_norm": 0.3974544405937195, + "learning_rate": 2.3210056336440273e-05, + "loss": 0.0352, + "step": 69600 + }, + { + "epoch": 4.560702352093298, + "grad_norm": 0.35040515661239624, + "learning_rate": 2.3196096801159844e-05, + "loss": 0.0284, + "step": 69610 + }, + { + "epoch": 4.5613575312848065, + "grad_norm": 1.1940137147903442, + "learning_rate": 2.3182140197114012e-05, + "loss": 0.0254, + "step": 69620 + }, + { + "epoch": 4.562012710476315, + "grad_norm": 0.1997143030166626, + "learning_rate": 2.3168186525828985e-05, + "loss": 0.0169, + "step": 69630 + }, + { + "epoch": 4.562667889667824, + "grad_norm": 0.40485283732414246, + "learning_rate": 2.3154235788830757e-05, + "loss": 0.0173, + "step": 69640 + }, + { + "epoch": 4.563323068859333, + "grad_norm": 0.26984187960624695, + "learning_rate": 2.3140287987644927e-05, + "loss": 0.0182, + "step": 69650 + }, + { + "epoch": 4.563978248050842, + "grad_norm": 2.062826156616211, + "learning_rate": 2.3126343123796833e-05, + "loss": 0.0338, + "step": 69660 + }, + { + "epoch": 4.5646334272423505, + "grad_norm": 0.4759272038936615, + "learning_rate": 2.3112401198811418e-05, + "loss": 0.0371, + "step": 69670 + }, + { + "epoch": 4.565288606433859, + "grad_norm": 0.6656922101974487, + "learning_rate": 2.3098462214213418e-05, + "loss": 0.0505, + "step": 69680 + }, + { + "epoch": 4.565943785625368, + "grad_norm": 1.007373332977295, + "learning_rate": 2.308452617152712e-05, + "loss": 0.0314, + "step": 69690 + }, + { + "epoch": 4.566598964816878, + "grad_norm": 0.627430260181427, + "learning_rate": 2.3070593072276603e-05, + "loss": 0.0212, + "step": 69700 + }, + { + "epoch": 4.567254144008386, + "grad_norm": 0.23345808684825897, + "learning_rate": 2.3056662917985518e-05, + "loss": 0.0208, + "step": 69710 + }, + { + "epoch": 4.5679093231998955, + "grad_norm": 0.5487989187240601, + "learning_rate": 2.3042735710177283e-05, + "loss": 0.0267, + "step": 69720 + }, + { + "epoch": 4.568564502391404, + "grad_norm": 0.7645478248596191, + "learning_rate": 2.302881145037492e-05, + "loss": 0.0393, + "step": 69730 + }, + { + "epoch": 4.569219681582913, + "grad_norm": 0.31229469180107117, + "learning_rate": 2.3014890140101192e-05, + "loss": 0.0152, + "step": 69740 + }, + { + "epoch": 4.569874860774422, + "grad_norm": 0.7275170683860779, + "learning_rate": 2.30009717808785e-05, + "loss": 0.0442, + "step": 69750 + }, + { + "epoch": 4.570530039965931, + "grad_norm": 0.5430474281311035, + "learning_rate": 2.2987056374228967e-05, + "loss": 0.0307, + "step": 69760 + }, + { + "epoch": 4.57118521915744, + "grad_norm": 3.6359751224517822, + "learning_rate": 2.297314392167431e-05, + "loss": 0.0494, + "step": 69770 + }, + { + "epoch": 4.571840398348948, + "grad_norm": 0.6589477062225342, + "learning_rate": 2.295923442473601e-05, + "loss": 0.0353, + "step": 69780 + }, + { + "epoch": 4.572495577540457, + "grad_norm": 0.40685200691223145, + "learning_rate": 2.294532788493519e-05, + "loss": 0.0519, + "step": 69790 + }, + { + "epoch": 4.573150756731966, + "grad_norm": 0.38400596380233765, + "learning_rate": 2.293142430379262e-05, + "loss": 0.0425, + "step": 69800 + }, + { + "epoch": 4.573805935923475, + "grad_norm": 0.3947737216949463, + "learning_rate": 2.291752368282879e-05, + "loss": 0.0333, + "step": 69810 + }, + { + "epoch": 4.574461115114984, + "grad_norm": 0.6741450428962708, + "learning_rate": 2.290362602356385e-05, + "loss": 0.0154, + "step": 69820 + }, + { + "epoch": 4.5751162943064925, + "grad_norm": 0.4177659749984741, + "learning_rate": 2.288973132751765e-05, + "loss": 0.0514, + "step": 69830 + }, + { + "epoch": 4.575771473498001, + "grad_norm": 0.40031659603118896, + "learning_rate": 2.287583959620965e-05, + "loss": 0.0353, + "step": 69840 + }, + { + "epoch": 4.576426652689511, + "grad_norm": 0.16420108079910278, + "learning_rate": 2.2861950831159074e-05, + "loss": 0.0415, + "step": 69850 + }, + { + "epoch": 4.57708183188102, + "grad_norm": 0.5439053773880005, + "learning_rate": 2.284806503388472e-05, + "loss": 0.0278, + "step": 69860 + }, + { + "epoch": 4.577737011072529, + "grad_norm": 0.4789089262485504, + "learning_rate": 2.283418220590518e-05, + "loss": 0.0232, + "step": 69870 + }, + { + "epoch": 4.578392190264037, + "grad_norm": 0.2074424922466278, + "learning_rate": 2.282030234873862e-05, + "loss": 0.0174, + "step": 69880 + }, + { + "epoch": 4.579047369455546, + "grad_norm": 0.4963652491569519, + "learning_rate": 2.2806425463902948e-05, + "loss": 0.0202, + "step": 69890 + }, + { + "epoch": 4.579702548647055, + "grad_norm": 0.5742167830467224, + "learning_rate": 2.279255155291568e-05, + "loss": 0.0391, + "step": 69900 + }, + { + "epoch": 4.580357727838564, + "grad_norm": 0.3227033019065857, + "learning_rate": 2.277868061729409e-05, + "loss": 0.0201, + "step": 69910 + }, + { + "epoch": 4.581012907030073, + "grad_norm": 0.10056763142347336, + "learning_rate": 2.276481265855504e-05, + "loss": 0.0382, + "step": 69920 + }, + { + "epoch": 4.5816680862215815, + "grad_norm": 0.18430295586585999, + "learning_rate": 2.2750947678215128e-05, + "loss": 0.0249, + "step": 69930 + }, + { + "epoch": 4.58232326541309, + "grad_norm": 0.22677671909332275, + "learning_rate": 2.2737085677790616e-05, + "loss": 0.0274, + "step": 69940 + }, + { + "epoch": 4.582978444604599, + "grad_norm": 0.1931498944759369, + "learning_rate": 2.2723226658797443e-05, + "loss": 0.0239, + "step": 69950 + }, + { + "epoch": 4.583633623796108, + "grad_norm": 0.6249955892562866, + "learning_rate": 2.270937062275117e-05, + "loss": 0.0191, + "step": 69960 + }, + { + "epoch": 4.584288802987617, + "grad_norm": 0.3601095676422119, + "learning_rate": 2.2695517571167113e-05, + "loss": 0.0337, + "step": 69970 + }, + { + "epoch": 4.5849439821791265, + "grad_norm": 0.7135555744171143, + "learning_rate": 2.268166750556019e-05, + "loss": 0.0262, + "step": 69980 + }, + { + "epoch": 4.585599161370634, + "grad_norm": 0.19496576488018036, + "learning_rate": 2.266782042744503e-05, + "loss": 0.0231, + "step": 69990 + }, + { + "epoch": 4.586254340562144, + "grad_norm": 0.2094307392835617, + "learning_rate": 2.2653976338335935e-05, + "loss": 0.0147, + "step": 70000 + }, + { + "epoch": 4.586909519753653, + "grad_norm": 0.5675320625305176, + "learning_rate": 2.2640135239746895e-05, + "loss": 0.0364, + "step": 70010 + }, + { + "epoch": 4.587564698945162, + "grad_norm": 0.20414413511753082, + "learning_rate": 2.2626297133191515e-05, + "loss": 0.0186, + "step": 70020 + }, + { + "epoch": 4.5882198781366705, + "grad_norm": 0.3011469841003418, + "learning_rate": 2.261246202018312e-05, + "loss": 0.0276, + "step": 70030 + }, + { + "epoch": 4.588875057328179, + "grad_norm": 0.7086028456687927, + "learning_rate": 2.2598629902234718e-05, + "loss": 0.0436, + "step": 70040 + }, + { + "epoch": 4.589530236519688, + "grad_norm": 0.34134915471076965, + "learning_rate": 2.258480078085894e-05, + "loss": 0.0318, + "step": 70050 + }, + { + "epoch": 4.590185415711197, + "grad_norm": 0.2925029993057251, + "learning_rate": 2.2570974657568126e-05, + "loss": 0.0324, + "step": 70060 + }, + { + "epoch": 4.590840594902706, + "grad_norm": 0.06404319405555725, + "learning_rate": 2.2557151533874286e-05, + "loss": 0.0351, + "step": 70070 + }, + { + "epoch": 4.591495774094215, + "grad_norm": 0.29923227429389954, + "learning_rate": 2.2543331411289116e-05, + "loss": 0.0457, + "step": 70080 + }, + { + "epoch": 4.592150953285723, + "grad_norm": 0.2719634473323822, + "learning_rate": 2.252951429132392e-05, + "loss": 0.0353, + "step": 70090 + }, + { + "epoch": 4.592806132477232, + "grad_norm": 0.9694145917892456, + "learning_rate": 2.2515700175489762e-05, + "loss": 0.0324, + "step": 70100 + }, + { + "epoch": 4.593461311668741, + "grad_norm": 0.3108251392841339, + "learning_rate": 2.2501889065297288e-05, + "loss": 0.028, + "step": 70110 + }, + { + "epoch": 4.59411649086025, + "grad_norm": 0.28701651096343994, + "learning_rate": 2.2488080962256887e-05, + "loss": 0.0357, + "step": 70120 + }, + { + "epoch": 4.59477167005176, + "grad_norm": 0.399080365896225, + "learning_rate": 2.2474275867878585e-05, + "loss": 0.0448, + "step": 70130 + }, + { + "epoch": 4.595426849243268, + "grad_norm": 0.652337908744812, + "learning_rate": 2.246047378367211e-05, + "loss": 0.0351, + "step": 70140 + }, + { + "epoch": 4.596082028434777, + "grad_norm": 0.18965129554271698, + "learning_rate": 2.2446674711146798e-05, + "loss": 0.0233, + "step": 70150 + }, + { + "epoch": 4.596737207626286, + "grad_norm": 0.15775224566459656, + "learning_rate": 2.2432878651811734e-05, + "loss": 0.0279, + "step": 70160 + }, + { + "epoch": 4.597392386817795, + "grad_norm": 0.6839014887809753, + "learning_rate": 2.2419085607175594e-05, + "loss": 0.0332, + "step": 70170 + }, + { + "epoch": 4.598047566009304, + "grad_norm": 0.37840303778648376, + "learning_rate": 2.240529557874678e-05, + "loss": 0.0274, + "step": 70180 + }, + { + "epoch": 4.5987027452008125, + "grad_norm": 0.4814601242542267, + "learning_rate": 2.2391508568033358e-05, + "loss": 0.0631, + "step": 70190 + }, + { + "epoch": 4.599357924392321, + "grad_norm": 0.213254913687706, + "learning_rate": 2.237772457654307e-05, + "loss": 0.0284, + "step": 70200 + }, + { + "epoch": 4.60001310358383, + "grad_norm": 0.37940505146980286, + "learning_rate": 2.236394360578327e-05, + "loss": 0.0764, + "step": 70210 + }, + { + "epoch": 4.600668282775339, + "grad_norm": 0.4013836979866028, + "learning_rate": 2.2350165657261073e-05, + "loss": 0.0217, + "step": 70220 + }, + { + "epoch": 4.601323461966848, + "grad_norm": 0.37872642278671265, + "learning_rate": 2.2336390732483164e-05, + "loss": 0.0224, + "step": 70230 + }, + { + "epoch": 4.6019786411583565, + "grad_norm": 0.16035297513008118, + "learning_rate": 2.2322618832955976e-05, + "loss": 0.0416, + "step": 70240 + }, + { + "epoch": 4.602633820349865, + "grad_norm": 1.0171011686325073, + "learning_rate": 2.2308849960185578e-05, + "loss": 0.04, + "step": 70250 + }, + { + "epoch": 4.603288999541375, + "grad_norm": 0.4971219301223755, + "learning_rate": 2.229508411567774e-05, + "loss": 0.0248, + "step": 70260 + }, + { + "epoch": 4.603944178732883, + "grad_norm": 0.6385331749916077, + "learning_rate": 2.2281321300937823e-05, + "loss": 0.0382, + "step": 70270 + }, + { + "epoch": 4.604599357924393, + "grad_norm": 0.8751692771911621, + "learning_rate": 2.2267561517470937e-05, + "loss": 0.0405, + "step": 70280 + }, + { + "epoch": 4.6052545371159015, + "grad_norm": 0.24554377794265747, + "learning_rate": 2.2253804766781845e-05, + "loss": 0.0261, + "step": 70290 + }, + { + "epoch": 4.60590971630741, + "grad_norm": 0.19568045437335968, + "learning_rate": 2.2240051050374933e-05, + "loss": 0.0424, + "step": 70300 + }, + { + "epoch": 4.606564895498919, + "grad_norm": 0.8161966800689697, + "learning_rate": 2.22263003697543e-05, + "loss": 0.0208, + "step": 70310 + }, + { + "epoch": 4.607220074690428, + "grad_norm": 0.6928130388259888, + "learning_rate": 2.2212552726423702e-05, + "loss": 0.0567, + "step": 70320 + }, + { + "epoch": 4.607875253881937, + "grad_norm": 0.47691407799720764, + "learning_rate": 2.2198808121886578e-05, + "loss": 0.0315, + "step": 70330 + }, + { + "epoch": 4.608530433073446, + "grad_norm": 1.1411855220794678, + "learning_rate": 2.218506655764598e-05, + "loss": 0.0312, + "step": 70340 + }, + { + "epoch": 4.609185612264954, + "grad_norm": 0.44233542680740356, + "learning_rate": 2.2171328035204704e-05, + "loss": 0.0375, + "step": 70350 + }, + { + "epoch": 4.609840791456463, + "grad_norm": 0.26465079188346863, + "learning_rate": 2.2157592556065133e-05, + "loss": 0.0339, + "step": 70360 + }, + { + "epoch": 4.610495970647972, + "grad_norm": 0.6872370839118958, + "learning_rate": 2.2143860121729382e-05, + "loss": 0.0331, + "step": 70370 + }, + { + "epoch": 4.611151149839481, + "grad_norm": 0.6987133026123047, + "learning_rate": 2.2130130733699206e-05, + "loss": 0.0247, + "step": 70380 + }, + { + "epoch": 4.61180632903099, + "grad_norm": 0.3746994733810425, + "learning_rate": 2.2116404393476052e-05, + "loss": 0.0507, + "step": 70390 + }, + { + "epoch": 4.6124615082224985, + "grad_norm": 0.1678469479084015, + "learning_rate": 2.2102681102560967e-05, + "loss": 0.0586, + "step": 70400 + }, + { + "epoch": 4.613116687414008, + "grad_norm": 0.31540173292160034, + "learning_rate": 2.208896086245476e-05, + "loss": 0.0289, + "step": 70410 + }, + { + "epoch": 4.613771866605517, + "grad_norm": 0.843906819820404, + "learning_rate": 2.2075243674657808e-05, + "loss": 0.0513, + "step": 70420 + }, + { + "epoch": 4.614427045797026, + "grad_norm": 0.38488829135894775, + "learning_rate": 2.206152954067024e-05, + "loss": 0.0283, + "step": 70430 + }, + { + "epoch": 4.615082224988535, + "grad_norm": 0.6217073202133179, + "learning_rate": 2.2047818461991758e-05, + "loss": 0.0448, + "step": 70440 + }, + { + "epoch": 4.615737404180043, + "grad_norm": 0.2589838206768036, + "learning_rate": 2.2034110440121864e-05, + "loss": 0.0258, + "step": 70450 + }, + { + "epoch": 4.616392583371552, + "grad_norm": 0.3330754041671753, + "learning_rate": 2.2020405476559585e-05, + "loss": 0.0428, + "step": 70460 + }, + { + "epoch": 4.617047762563061, + "grad_norm": 2.143439769744873, + "learning_rate": 2.2006703572803722e-05, + "loss": 0.0434, + "step": 70470 + }, + { + "epoch": 4.61770294175457, + "grad_norm": 0.3182678818702698, + "learning_rate": 2.1993004730352646e-05, + "loss": 0.0232, + "step": 70480 + }, + { + "epoch": 4.618358120946079, + "grad_norm": 0.38707539439201355, + "learning_rate": 2.1979308950704464e-05, + "loss": 0.024, + "step": 70490 + }, + { + "epoch": 4.6190133001375875, + "grad_norm": 0.2451363205909729, + "learning_rate": 2.1965616235356923e-05, + "loss": 0.0223, + "step": 70500 + }, + { + "epoch": 4.619668479329096, + "grad_norm": 0.5302656888961792, + "learning_rate": 2.1951926585807465e-05, + "loss": 0.0366, + "step": 70510 + }, + { + "epoch": 4.620323658520605, + "grad_norm": 0.34465745091438293, + "learning_rate": 2.193824000355313e-05, + "loss": 0.0324, + "step": 70520 + }, + { + "epoch": 4.620978837712114, + "grad_norm": 0.1540578305721283, + "learning_rate": 2.192455649009067e-05, + "loss": 0.0328, + "step": 70530 + }, + { + "epoch": 4.621634016903624, + "grad_norm": 0.3716970980167389, + "learning_rate": 2.1910876046916523e-05, + "loss": 0.0299, + "step": 70540 + }, + { + "epoch": 4.622289196095132, + "grad_norm": 0.5578243136405945, + "learning_rate": 2.1897198675526725e-05, + "loss": 0.0316, + "step": 70550 + }, + { + "epoch": 4.622944375286641, + "grad_norm": 0.5671855211257935, + "learning_rate": 2.1883524377417024e-05, + "loss": 0.0278, + "step": 70560 + }, + { + "epoch": 4.62359955447815, + "grad_norm": 0.15655848383903503, + "learning_rate": 2.1869853154082825e-05, + "loss": 0.0247, + "step": 70570 + }, + { + "epoch": 4.624254733669659, + "grad_norm": 0.20684079825878143, + "learning_rate": 2.1856185007019215e-05, + "loss": 0.031, + "step": 70580 + }, + { + "epoch": 4.624909912861168, + "grad_norm": 0.051057010889053345, + "learning_rate": 2.1842519937720874e-05, + "loss": 0.0205, + "step": 70590 + }, + { + "epoch": 4.6255650920526765, + "grad_norm": 0.3752821683883667, + "learning_rate": 2.182885794768224e-05, + "loss": 0.0128, + "step": 70600 + }, + { + "epoch": 4.626220271244185, + "grad_norm": 0.42695152759552, + "learning_rate": 2.1815199038397326e-05, + "loss": 0.0313, + "step": 70610 + }, + { + "epoch": 4.626875450435694, + "grad_norm": 0.5938395857810974, + "learning_rate": 2.1801543211359893e-05, + "loss": 0.0258, + "step": 70620 + }, + { + "epoch": 4.627530629627203, + "grad_norm": 0.5220646262168884, + "learning_rate": 2.178789046806326e-05, + "loss": 0.0215, + "step": 70630 + }, + { + "epoch": 4.628185808818712, + "grad_norm": 0.458327978849411, + "learning_rate": 2.1774240810000547e-05, + "loss": 0.0718, + "step": 70640 + }, + { + "epoch": 4.628840988010221, + "grad_norm": 0.24133636057376862, + "learning_rate": 2.17605942386644e-05, + "loss": 0.0219, + "step": 70650 + }, + { + "epoch": 4.629496167201729, + "grad_norm": 0.25226858258247375, + "learning_rate": 2.174695075554723e-05, + "loss": 0.0324, + "step": 70660 + }, + { + "epoch": 4.630151346393238, + "grad_norm": 0.09380562603473663, + "learning_rate": 2.173331036214103e-05, + "loss": 0.0309, + "step": 70670 + }, + { + "epoch": 4.630806525584747, + "grad_norm": 0.9296314716339111, + "learning_rate": 2.171967305993753e-05, + "loss": 0.0469, + "step": 70680 + }, + { + "epoch": 4.631461704776257, + "grad_norm": 0.29520511627197266, + "learning_rate": 2.1706038850428024e-05, + "loss": 0.0347, + "step": 70690 + }, + { + "epoch": 4.6321168839677656, + "grad_norm": 0.12982015311717987, + "learning_rate": 2.169240773510361e-05, + "loss": 0.0264, + "step": 70700 + }, + { + "epoch": 4.632772063159274, + "grad_norm": 0.6530637145042419, + "learning_rate": 2.1678779715454906e-05, + "loss": 0.0262, + "step": 70710 + }, + { + "epoch": 4.633427242350783, + "grad_norm": 0.3712545335292816, + "learning_rate": 2.166515479297229e-05, + "loss": 0.0458, + "step": 70720 + }, + { + "epoch": 4.634082421542292, + "grad_norm": 0.8728472590446472, + "learning_rate": 2.1651532969145732e-05, + "loss": 0.0485, + "step": 70730 + }, + { + "epoch": 4.634737600733801, + "grad_norm": 0.2341293841600418, + "learning_rate": 2.1637914245464903e-05, + "loss": 0.0298, + "step": 70740 + }, + { + "epoch": 4.63539277992531, + "grad_norm": 0.1432029902935028, + "learning_rate": 2.1624298623419133e-05, + "loss": 0.0262, + "step": 70750 + }, + { + "epoch": 4.636047959116818, + "grad_norm": 0.47969186305999756, + "learning_rate": 2.1610686104497413e-05, + "loss": 0.0197, + "step": 70760 + }, + { + "epoch": 4.636703138308327, + "grad_norm": 0.2505300045013428, + "learning_rate": 2.1597076690188366e-05, + "loss": 0.0582, + "step": 70770 + }, + { + "epoch": 4.637358317499836, + "grad_norm": 0.10230709612369537, + "learning_rate": 2.1583470381980315e-05, + "loss": 0.015, + "step": 70780 + }, + { + "epoch": 4.638013496691345, + "grad_norm": 0.20413053035736084, + "learning_rate": 2.156986718136123e-05, + "loss": 0.0288, + "step": 70790 + }, + { + "epoch": 4.638668675882854, + "grad_norm": 0.5399061441421509, + "learning_rate": 2.155626708981871e-05, + "loss": 0.0266, + "step": 70800 + }, + { + "epoch": 4.6393238550743625, + "grad_norm": 0.37970417737960815, + "learning_rate": 2.1542670108840075e-05, + "loss": 0.0351, + "step": 70810 + }, + { + "epoch": 4.639979034265872, + "grad_norm": 0.6463855504989624, + "learning_rate": 2.152907623991221e-05, + "loss": 0.022, + "step": 70820 + }, + { + "epoch": 4.64063421345738, + "grad_norm": 0.269035667181015, + "learning_rate": 2.1515485484521802e-05, + "loss": 0.0333, + "step": 70830 + }, + { + "epoch": 4.64128939264889, + "grad_norm": 1.0629498958587646, + "learning_rate": 2.150189784415506e-05, + "loss": 0.0319, + "step": 70840 + }, + { + "epoch": 4.641944571840399, + "grad_norm": 0.2564888894557953, + "learning_rate": 2.148831332029794e-05, + "loss": 0.0282, + "step": 70850 + }, + { + "epoch": 4.6425997510319075, + "grad_norm": 0.7774978876113892, + "learning_rate": 2.147473191443599e-05, + "loss": 0.032, + "step": 70860 + }, + { + "epoch": 4.643254930223416, + "grad_norm": 0.07665275782346725, + "learning_rate": 2.1461153628054492e-05, + "loss": 0.0489, + "step": 70870 + }, + { + "epoch": 4.643910109414925, + "grad_norm": 0.18334661424160004, + "learning_rate": 2.1447578462638284e-05, + "loss": 0.0212, + "step": 70880 + }, + { + "epoch": 4.644565288606434, + "grad_norm": 1.131174087524414, + "learning_rate": 2.143400641967201e-05, + "loss": 0.0268, + "step": 70890 + }, + { + "epoch": 4.645220467797943, + "grad_norm": 0.43094491958618164, + "learning_rate": 2.1420437500639823e-05, + "loss": 0.0327, + "step": 70900 + }, + { + "epoch": 4.6458756469894515, + "grad_norm": 0.3180173933506012, + "learning_rate": 2.1406871707025646e-05, + "loss": 0.0223, + "step": 70910 + }, + { + "epoch": 4.64653082618096, + "grad_norm": 0.5417888760566711, + "learning_rate": 2.1393309040312963e-05, + "loss": 0.0522, + "step": 70920 + }, + { + "epoch": 4.647186005372469, + "grad_norm": 0.20839941501617432, + "learning_rate": 2.137974950198502e-05, + "loss": 0.04, + "step": 70930 + }, + { + "epoch": 4.647841184563978, + "grad_norm": 0.12008541822433472, + "learning_rate": 2.1366193093524617e-05, + "loss": 0.0416, + "step": 70940 + }, + { + "epoch": 4.648496363755487, + "grad_norm": 0.9041624665260315, + "learning_rate": 2.1352639816414288e-05, + "loss": 0.0734, + "step": 70950 + }, + { + "epoch": 4.649151542946996, + "grad_norm": 0.17941273748874664, + "learning_rate": 2.1339089672136202e-05, + "loss": 0.0316, + "step": 70960 + }, + { + "epoch": 4.649806722138505, + "grad_norm": 0.759242594242096, + "learning_rate": 2.1325542662172193e-05, + "loss": 0.0357, + "step": 70970 + }, + { + "epoch": 4.650461901330014, + "grad_norm": 0.8060339689254761, + "learning_rate": 2.1311998788003712e-05, + "loss": 0.0335, + "step": 70980 + }, + { + "epoch": 4.651117080521523, + "grad_norm": 0.6073563694953918, + "learning_rate": 2.1298458051111907e-05, + "loss": 0.085, + "step": 70990 + }, + { + "epoch": 4.651772259713032, + "grad_norm": 0.13192833960056305, + "learning_rate": 2.1284920452977602e-05, + "loss": 0.0266, + "step": 71000 + }, + { + "epoch": 4.652427438904541, + "grad_norm": 0.8116374015808105, + "learning_rate": 2.12713859950812e-05, + "loss": 0.0201, + "step": 71010 + }, + { + "epoch": 4.653082618096049, + "grad_norm": 0.7303617596626282, + "learning_rate": 2.1257854678902832e-05, + "loss": 0.0322, + "step": 71020 + }, + { + "epoch": 4.653737797287558, + "grad_norm": 0.6864328980445862, + "learning_rate": 2.124432650592227e-05, + "loss": 0.0314, + "step": 71030 + }, + { + "epoch": 4.654392976479067, + "grad_norm": 4.196747303009033, + "learning_rate": 2.1230801477618937e-05, + "loss": 0.0544, + "step": 71040 + }, + { + "epoch": 4.655048155670576, + "grad_norm": 0.6779870986938477, + "learning_rate": 2.121727959547189e-05, + "loss": 0.0219, + "step": 71050 + }, + { + "epoch": 4.655703334862085, + "grad_norm": 0.16560739278793335, + "learning_rate": 2.1203760860959886e-05, + "loss": 0.0251, + "step": 71060 + }, + { + "epoch": 4.6563585140535935, + "grad_norm": 0.9901506900787354, + "learning_rate": 2.119024527556127e-05, + "loss": 0.0285, + "step": 71070 + }, + { + "epoch": 4.657013693245102, + "grad_norm": 0.06147978454828262, + "learning_rate": 2.117673284075415e-05, + "loss": 0.0303, + "step": 71080 + }, + { + "epoch": 4.657668872436611, + "grad_norm": 0.1708691567182541, + "learning_rate": 2.1163223558016165e-05, + "loss": 0.0243, + "step": 71090 + }, + { + "epoch": 4.65832405162812, + "grad_norm": 0.21976682543754578, + "learning_rate": 2.1149717428824722e-05, + "loss": 0.0205, + "step": 71100 + }, + { + "epoch": 4.658979230819629, + "grad_norm": 1.075168251991272, + "learning_rate": 2.1136214454656784e-05, + "loss": 0.0571, + "step": 71110 + }, + { + "epoch": 4.659634410011138, + "grad_norm": 0.39110031723976135, + "learning_rate": 2.1122714636989056e-05, + "loss": 0.0343, + "step": 71120 + }, + { + "epoch": 4.660289589202647, + "grad_norm": 0.5713983178138733, + "learning_rate": 2.1109217977297823e-05, + "loss": 0.0141, + "step": 71130 + }, + { + "epoch": 4.660944768394156, + "grad_norm": 1.1864259243011475, + "learning_rate": 2.1095724477059077e-05, + "loss": 0.0165, + "step": 71140 + }, + { + "epoch": 4.661599947585665, + "grad_norm": 0.16715821623802185, + "learning_rate": 2.1082234137748434e-05, + "loss": 0.0321, + "step": 71150 + }, + { + "epoch": 4.662255126777174, + "grad_norm": 0.16621512174606323, + "learning_rate": 2.106874696084122e-05, + "loss": 0.0181, + "step": 71160 + }, + { + "epoch": 4.6629103059686825, + "grad_norm": 0.19028490781784058, + "learning_rate": 2.1055262947812315e-05, + "loss": 0.0275, + "step": 71170 + }, + { + "epoch": 4.663565485160191, + "grad_norm": 0.5155711770057678, + "learning_rate": 2.104178210013636e-05, + "loss": 0.0284, + "step": 71180 + }, + { + "epoch": 4.6642206643517, + "grad_norm": 0.14312714338302612, + "learning_rate": 2.1028304419287554e-05, + "loss": 0.0211, + "step": 71190 + }, + { + "epoch": 4.664875843543209, + "grad_norm": 0.5477359294891357, + "learning_rate": 2.101482990673982e-05, + "loss": 0.0267, + "step": 71200 + }, + { + "epoch": 4.665531022734718, + "grad_norm": 0.6644006967544556, + "learning_rate": 2.1001358563966707e-05, + "loss": 0.0193, + "step": 71210 + }, + { + "epoch": 4.666186201926227, + "grad_norm": 0.32122671604156494, + "learning_rate": 2.098789039244144e-05, + "loss": 0.0311, + "step": 71220 + }, + { + "epoch": 4.666841381117735, + "grad_norm": 0.42378920316696167, + "learning_rate": 2.097442539363685e-05, + "loss": 0.0413, + "step": 71230 + }, + { + "epoch": 4.667496560309244, + "grad_norm": 0.4612860679626465, + "learning_rate": 2.0960963569025448e-05, + "loss": 0.0345, + "step": 71240 + }, + { + "epoch": 4.668151739500754, + "grad_norm": 0.3503812253475189, + "learning_rate": 2.0947504920079435e-05, + "loss": 0.0136, + "step": 71250 + }, + { + "epoch": 4.668806918692263, + "grad_norm": 0.490518182516098, + "learning_rate": 2.093404944827059e-05, + "loss": 0.0441, + "step": 71260 + }, + { + "epoch": 4.6694620978837715, + "grad_norm": 0.26982706785202026, + "learning_rate": 2.0920597155070394e-05, + "loss": 0.0255, + "step": 71270 + }, + { + "epoch": 4.67011727707528, + "grad_norm": 1.1268775463104248, + "learning_rate": 2.090714804194997e-05, + "loss": 0.0391, + "step": 71280 + }, + { + "epoch": 4.670772456266789, + "grad_norm": 0.38557711243629456, + "learning_rate": 2.0893702110380114e-05, + "loss": 0.0435, + "step": 71290 + }, + { + "epoch": 4.671427635458298, + "grad_norm": 0.2317410260438919, + "learning_rate": 2.0880259361831216e-05, + "loss": 0.0326, + "step": 71300 + }, + { + "epoch": 4.672082814649807, + "grad_norm": 0.179056778550148, + "learning_rate": 2.0866819797773397e-05, + "loss": 0.0285, + "step": 71310 + }, + { + "epoch": 4.672737993841316, + "grad_norm": 0.35017770528793335, + "learning_rate": 2.085338341967634e-05, + "loss": 0.0281, + "step": 71320 + }, + { + "epoch": 4.673393173032824, + "grad_norm": 0.6695926785469055, + "learning_rate": 2.083995022900946e-05, + "loss": 0.0381, + "step": 71330 + }, + { + "epoch": 4.674048352224333, + "grad_norm": 0.2986406087875366, + "learning_rate": 2.0826520227241768e-05, + "loss": 0.0284, + "step": 71340 + }, + { + "epoch": 4.674703531415842, + "grad_norm": 0.3817594051361084, + "learning_rate": 2.081309341584199e-05, + "loss": 0.0304, + "step": 71350 + }, + { + "epoch": 4.675358710607351, + "grad_norm": 0.1982140988111496, + "learning_rate": 2.0799669796278413e-05, + "loss": 0.0211, + "step": 71360 + }, + { + "epoch": 4.67601388979886, + "grad_norm": 0.23935075104236603, + "learning_rate": 2.078624937001906e-05, + "loss": 0.0337, + "step": 71370 + }, + { + "epoch": 4.6766690689903685, + "grad_norm": 0.6754233241081238, + "learning_rate": 2.0772832138531544e-05, + "loss": 0.0238, + "step": 71380 + }, + { + "epoch": 4.677324248181877, + "grad_norm": 0.22638842463493347, + "learning_rate": 2.0759418103283158e-05, + "loss": 0.0455, + "step": 71390 + }, + { + "epoch": 4.677979427373387, + "grad_norm": 0.35418686270713806, + "learning_rate": 2.0746007265740846e-05, + "loss": 0.0442, + "step": 71400 + }, + { + "epoch": 4.678634606564896, + "grad_norm": 0.3992256224155426, + "learning_rate": 2.0732599627371215e-05, + "loss": 0.041, + "step": 71410 + }, + { + "epoch": 4.679289785756405, + "grad_norm": 0.4896650016307831, + "learning_rate": 2.0719195189640468e-05, + "loss": 0.0164, + "step": 71420 + }, + { + "epoch": 4.6799449649479135, + "grad_norm": 0.6623049378395081, + "learning_rate": 2.0705793954014528e-05, + "loss": 0.0226, + "step": 71430 + }, + { + "epoch": 4.680600144139422, + "grad_norm": 0.7376546263694763, + "learning_rate": 2.06923959219589e-05, + "loss": 0.0369, + "step": 71440 + }, + { + "epoch": 4.681255323330931, + "grad_norm": 0.08853442221879959, + "learning_rate": 2.0679001094938793e-05, + "loss": 0.0303, + "step": 71450 + }, + { + "epoch": 4.68191050252244, + "grad_norm": 0.6997773051261902, + "learning_rate": 2.066560947441904e-05, + "loss": 0.0182, + "step": 71460 + }, + { + "epoch": 4.682565681713949, + "grad_norm": 0.545023500919342, + "learning_rate": 2.0652221061864152e-05, + "loss": 0.0438, + "step": 71470 + }, + { + "epoch": 4.6832208609054575, + "grad_norm": 0.09428088366985321, + "learning_rate": 2.0638835858738226e-05, + "loss": 0.0309, + "step": 71480 + }, + { + "epoch": 4.683876040096966, + "grad_norm": 0.19765977561473846, + "learning_rate": 2.0625453866505068e-05, + "loss": 0.0235, + "step": 71490 + }, + { + "epoch": 4.684531219288475, + "grad_norm": 0.42530548572540283, + "learning_rate": 2.0612075086628123e-05, + "loss": 0.0209, + "step": 71500 + }, + { + "epoch": 4.685186398479984, + "grad_norm": 0.27275189757347107, + "learning_rate": 2.059869952057045e-05, + "loss": 0.0501, + "step": 71510 + }, + { + "epoch": 4.685841577671493, + "grad_norm": 0.8227847218513489, + "learning_rate": 2.0585327169794793e-05, + "loss": 0.0351, + "step": 71520 + }, + { + "epoch": 4.6864967568630025, + "grad_norm": 0.2943783402442932, + "learning_rate": 2.0571958035763534e-05, + "loss": 0.0293, + "step": 71530 + }, + { + "epoch": 4.68715193605451, + "grad_norm": 0.32643863558769226, + "learning_rate": 2.0558592119938715e-05, + "loss": 0.0219, + "step": 71540 + }, + { + "epoch": 4.68780711524602, + "grad_norm": 3.074233293533325, + "learning_rate": 2.054522942378198e-05, + "loss": 0.0578, + "step": 71550 + }, + { + "epoch": 4.688462294437529, + "grad_norm": 0.2985226809978485, + "learning_rate": 2.053186994875469e-05, + "loss": 0.0512, + "step": 71560 + }, + { + "epoch": 4.689117473629038, + "grad_norm": 0.2998787760734558, + "learning_rate": 2.0518513696317775e-05, + "loss": 0.0217, + "step": 71570 + }, + { + "epoch": 4.689772652820547, + "grad_norm": 0.793239176273346, + "learning_rate": 2.050516066793188e-05, + "loss": 0.0273, + "step": 71580 + }, + { + "epoch": 4.690427832012055, + "grad_norm": 0.2982388436794281, + "learning_rate": 2.0491810865057276e-05, + "loss": 0.0202, + "step": 71590 + }, + { + "epoch": 4.691083011203564, + "grad_norm": 0.361141562461853, + "learning_rate": 2.0478464289153882e-05, + "loss": 0.0322, + "step": 71600 + }, + { + "epoch": 4.691738190395073, + "grad_norm": 0.9316269159317017, + "learning_rate": 2.0465120941681232e-05, + "loss": 0.0241, + "step": 71610 + }, + { + "epoch": 4.692393369586582, + "grad_norm": 0.5112412571907043, + "learning_rate": 2.0451780824098566e-05, + "loss": 0.0148, + "step": 71620 + }, + { + "epoch": 4.693048548778091, + "grad_norm": 0.8862060308456421, + "learning_rate": 2.043844393786471e-05, + "loss": 0.0427, + "step": 71630 + }, + { + "epoch": 4.6937037279695994, + "grad_norm": 0.2594476044178009, + "learning_rate": 2.0425110284438193e-05, + "loss": 0.0538, + "step": 71640 + }, + { + "epoch": 4.694358907161108, + "grad_norm": 0.7473689913749695, + "learning_rate": 2.041177986527712e-05, + "loss": 0.0194, + "step": 71650 + }, + { + "epoch": 4.695014086352617, + "grad_norm": 0.3239939510822296, + "learning_rate": 2.0398452681839347e-05, + "loss": 0.0383, + "step": 71660 + }, + { + "epoch": 4.695669265544126, + "grad_norm": 0.347699910402298, + "learning_rate": 2.0385128735582272e-05, + "loss": 0.0299, + "step": 71670 + }, + { + "epoch": 4.696324444735636, + "grad_norm": 0.11445202678442001, + "learning_rate": 2.0371808027963003e-05, + "loss": 0.0247, + "step": 71680 + }, + { + "epoch": 4.696979623927144, + "grad_norm": 0.36550161242485046, + "learning_rate": 2.035849056043825e-05, + "loss": 0.0516, + "step": 71690 + }, + { + "epoch": 4.697634803118653, + "grad_norm": 0.3529623746871948, + "learning_rate": 2.03451763344644e-05, + "loss": 0.0321, + "step": 71700 + }, + { + "epoch": 4.698289982310162, + "grad_norm": 1.1648730039596558, + "learning_rate": 2.0331865351497482e-05, + "loss": 0.0455, + "step": 71710 + }, + { + "epoch": 4.698945161501671, + "grad_norm": 0.13425768911838531, + "learning_rate": 2.0318557612993185e-05, + "loss": 0.0166, + "step": 71720 + }, + { + "epoch": 4.69960034069318, + "grad_norm": 0.18477413058280945, + "learning_rate": 2.0305253120406782e-05, + "loss": 0.0433, + "step": 71730 + }, + { + "epoch": 4.7002555198846885, + "grad_norm": 0.20523102581501007, + "learning_rate": 2.029195187519325e-05, + "loss": 0.0293, + "step": 71740 + }, + { + "epoch": 4.700910699076197, + "grad_norm": 0.1383301168680191, + "learning_rate": 2.0278653878807217e-05, + "loss": 0.0341, + "step": 71750 + }, + { + "epoch": 4.701565878267706, + "grad_norm": 0.144420325756073, + "learning_rate": 2.0265359132702892e-05, + "loss": 0.0273, + "step": 71760 + }, + { + "epoch": 4.702221057459215, + "grad_norm": 0.4237343966960907, + "learning_rate": 2.0252067638334182e-05, + "loss": 0.0667, + "step": 71770 + }, + { + "epoch": 4.702876236650724, + "grad_norm": 1.0733058452606201, + "learning_rate": 2.023877939715464e-05, + "loss": 0.0537, + "step": 71780 + }, + { + "epoch": 4.7035314158422326, + "grad_norm": 0.25091442465782166, + "learning_rate": 2.0225494410617448e-05, + "loss": 0.0197, + "step": 71790 + }, + { + "epoch": 4.704186595033741, + "grad_norm": 0.38256725668907166, + "learning_rate": 2.02122126801754e-05, + "loss": 0.024, + "step": 71800 + }, + { + "epoch": 4.704841774225251, + "grad_norm": 0.7459529042243958, + "learning_rate": 2.0198934207281018e-05, + "loss": 0.0331, + "step": 71810 + }, + { + "epoch": 4.705496953416759, + "grad_norm": 0.4071105718612671, + "learning_rate": 2.0185658993386357e-05, + "loss": 0.0386, + "step": 71820 + }, + { + "epoch": 4.706152132608269, + "grad_norm": 0.15279513597488403, + "learning_rate": 2.0172387039943226e-05, + "loss": 0.0329, + "step": 71830 + }, + { + "epoch": 4.7068073117997775, + "grad_norm": 0.40128326416015625, + "learning_rate": 2.0159118348402968e-05, + "loss": 0.03, + "step": 71840 + }, + { + "epoch": 4.707462490991286, + "grad_norm": 0.3673854172229767, + "learning_rate": 2.0145852920216697e-05, + "loss": 0.0477, + "step": 71850 + }, + { + "epoch": 4.708117670182795, + "grad_norm": 0.40182778239250183, + "learning_rate": 2.0132590756835045e-05, + "loss": 0.0287, + "step": 71860 + }, + { + "epoch": 4.708772849374304, + "grad_norm": 0.5026491284370422, + "learning_rate": 2.0119331859708386e-05, + "loss": 0.0203, + "step": 71870 + }, + { + "epoch": 4.709428028565813, + "grad_norm": 0.20500816404819489, + "learning_rate": 2.0106076230286654e-05, + "loss": 0.0231, + "step": 71880 + }, + { + "epoch": 4.710083207757322, + "grad_norm": 0.9279943704605103, + "learning_rate": 2.0092823870019488e-05, + "loss": 0.0382, + "step": 71890 + }, + { + "epoch": 4.71073838694883, + "grad_norm": 0.6027383804321289, + "learning_rate": 2.0079574780356115e-05, + "loss": 0.0466, + "step": 71900 + }, + { + "epoch": 4.711393566140339, + "grad_norm": 2.127298355102539, + "learning_rate": 2.0066328962745486e-05, + "loss": 0.0593, + "step": 71910 + }, + { + "epoch": 4.712048745331848, + "grad_norm": 0.8233283162117004, + "learning_rate": 2.00530864186361e-05, + "loss": 0.0268, + "step": 71920 + }, + { + "epoch": 4.712703924523357, + "grad_norm": 0.2806473672389984, + "learning_rate": 2.003984714947618e-05, + "loss": 0.0208, + "step": 71930 + }, + { + "epoch": 4.713359103714866, + "grad_norm": 0.3758818209171295, + "learning_rate": 2.0026611156713515e-05, + "loss": 0.0224, + "step": 71940 + }, + { + "epoch": 4.7140142829063745, + "grad_norm": 0.13690675795078278, + "learning_rate": 2.0013378441795584e-05, + "loss": 0.0171, + "step": 71950 + }, + { + "epoch": 4.714669462097884, + "grad_norm": 0.7077738046646118, + "learning_rate": 2.0000149006169506e-05, + "loss": 0.024, + "step": 71960 + }, + { + "epoch": 4.715324641289393, + "grad_norm": 1.342318058013916, + "learning_rate": 1.9986922851282048e-05, + "loss": 0.0549, + "step": 71970 + }, + { + "epoch": 4.715979820480902, + "grad_norm": 0.16562478244304657, + "learning_rate": 1.997369997857956e-05, + "loss": 0.045, + "step": 71980 + }, + { + "epoch": 4.716634999672411, + "grad_norm": 0.15116384625434875, + "learning_rate": 1.9960480389508103e-05, + "loss": 0.0171, + "step": 71990 + }, + { + "epoch": 4.717290178863919, + "grad_norm": 1.5545669794082642, + "learning_rate": 1.9947264085513363e-05, + "loss": 0.0201, + "step": 72000 + }, + { + "epoch": 4.717945358055428, + "grad_norm": 0.6170486807823181, + "learning_rate": 1.993405106804062e-05, + "loss": 0.027, + "step": 72010 + }, + { + "epoch": 4.718600537246937, + "grad_norm": 0.3123147785663605, + "learning_rate": 1.9920841338534867e-05, + "loss": 0.0346, + "step": 72020 + }, + { + "epoch": 4.719255716438446, + "grad_norm": 0.20489543676376343, + "learning_rate": 1.9907634898440646e-05, + "loss": 0.04, + "step": 72030 + }, + { + "epoch": 4.719910895629955, + "grad_norm": 0.16842781007289886, + "learning_rate": 1.989443174920227e-05, + "loss": 0.0406, + "step": 72040 + }, + { + "epoch": 4.7205660748214635, + "grad_norm": 0.6323316097259521, + "learning_rate": 1.988123189226355e-05, + "loss": 0.0398, + "step": 72050 + }, + { + "epoch": 4.721221254012972, + "grad_norm": 0.91071617603302, + "learning_rate": 1.986803532906805e-05, + "loss": 0.0303, + "step": 72060 + }, + { + "epoch": 4.721876433204481, + "grad_norm": 0.40083980560302734, + "learning_rate": 1.985484206105889e-05, + "loss": 0.0265, + "step": 72070 + }, + { + "epoch": 4.72253161239599, + "grad_norm": 0.11292162537574768, + "learning_rate": 1.9841652089678887e-05, + "loss": 0.0229, + "step": 72080 + }, + { + "epoch": 4.7231867915875, + "grad_norm": 0.6272091865539551, + "learning_rate": 1.9828465416370435e-05, + "loss": 0.0383, + "step": 72090 + }, + { + "epoch": 4.723841970779008, + "grad_norm": 0.3753099739551544, + "learning_rate": 1.9815282042575685e-05, + "loss": 0.0398, + "step": 72100 + }, + { + "epoch": 4.724497149970517, + "grad_norm": 0.32703760266304016, + "learning_rate": 1.9802101969736285e-05, + "loss": 0.0213, + "step": 72110 + }, + { + "epoch": 4.725152329162026, + "grad_norm": 0.3075619339942932, + "learning_rate": 1.9788925199293633e-05, + "loss": 0.0258, + "step": 72120 + }, + { + "epoch": 4.725807508353535, + "grad_norm": 0.5567365288734436, + "learning_rate": 1.977575173268868e-05, + "loss": 0.0403, + "step": 72130 + }, + { + "epoch": 4.726462687545044, + "grad_norm": 0.08960111439228058, + "learning_rate": 1.9762581571362094e-05, + "loss": 0.0342, + "step": 72140 + }, + { + "epoch": 4.7271178667365525, + "grad_norm": 0.41262733936309814, + "learning_rate": 1.9749414716754112e-05, + "loss": 0.0199, + "step": 72150 + }, + { + "epoch": 4.727773045928061, + "grad_norm": 0.8835704326629639, + "learning_rate": 1.973625117030465e-05, + "loss": 0.0479, + "step": 72160 + }, + { + "epoch": 4.72842822511957, + "grad_norm": 0.20088541507720947, + "learning_rate": 1.972309093345326e-05, + "loss": 0.0406, + "step": 72170 + }, + { + "epoch": 4.729083404311079, + "grad_norm": 0.4540637135505676, + "learning_rate": 1.970993400763915e-05, + "loss": 0.0448, + "step": 72180 + }, + { + "epoch": 4.729738583502588, + "grad_norm": 0.14991699159145355, + "learning_rate": 1.969678039430109e-05, + "loss": 0.0292, + "step": 72190 + }, + { + "epoch": 4.730393762694097, + "grad_norm": 0.29921144247055054, + "learning_rate": 1.9683630094877564e-05, + "loss": 0.0199, + "step": 72200 + }, + { + "epoch": 4.731048941885605, + "grad_norm": 0.10858375579118729, + "learning_rate": 1.9670483110806696e-05, + "loss": 0.0216, + "step": 72210 + }, + { + "epoch": 4.731704121077114, + "grad_norm": 0.181251659989357, + "learning_rate": 1.9657339443526172e-05, + "loss": 0.0281, + "step": 72220 + }, + { + "epoch": 4.732359300268623, + "grad_norm": 0.414849191904068, + "learning_rate": 1.964419909447339e-05, + "loss": 0.0251, + "step": 72230 + }, + { + "epoch": 4.733014479460133, + "grad_norm": 0.07046183943748474, + "learning_rate": 1.963106206508536e-05, + "loss": 0.0391, + "step": 72240 + }, + { + "epoch": 4.733669658651642, + "grad_norm": 1.8479856252670288, + "learning_rate": 1.9617928356798738e-05, + "loss": 0.0337, + "step": 72250 + }, + { + "epoch": 4.73432483784315, + "grad_norm": 0.42191970348358154, + "learning_rate": 1.960479797104977e-05, + "loss": 0.0507, + "step": 72260 + }, + { + "epoch": 4.734980017034659, + "grad_norm": 0.5974889993667603, + "learning_rate": 1.9591670909274418e-05, + "loss": 0.0214, + "step": 72270 + }, + { + "epoch": 4.735635196226168, + "grad_norm": 0.5019285082817078, + "learning_rate": 1.9578547172908184e-05, + "loss": 0.0337, + "step": 72280 + }, + { + "epoch": 4.736290375417677, + "grad_norm": 0.21505872905254364, + "learning_rate": 1.9565426763386325e-05, + "loss": 0.0315, + "step": 72290 + }, + { + "epoch": 4.736945554609186, + "grad_norm": 1.143625259399414, + "learning_rate": 1.9552309682143622e-05, + "loss": 0.0416, + "step": 72300 + }, + { + "epoch": 4.7376007338006945, + "grad_norm": 0.5814024806022644, + "learning_rate": 1.9539195930614574e-05, + "loss": 0.012, + "step": 72310 + }, + { + "epoch": 4.738255912992203, + "grad_norm": 0.3087814450263977, + "learning_rate": 1.952608551023324e-05, + "loss": 0.0249, + "step": 72320 + }, + { + "epoch": 4.738911092183712, + "grad_norm": 0.8483986854553223, + "learning_rate": 1.9512978422433397e-05, + "loss": 0.0468, + "step": 72330 + }, + { + "epoch": 4.739566271375221, + "grad_norm": 0.8895882368087769, + "learning_rate": 1.9499874668648376e-05, + "loss": 0.0258, + "step": 72340 + }, + { + "epoch": 4.74022145056673, + "grad_norm": 0.35617390275001526, + "learning_rate": 1.9486774250311213e-05, + "loss": 0.0243, + "step": 72350 + }, + { + "epoch": 4.7408766297582385, + "grad_norm": 0.28225255012512207, + "learning_rate": 1.9473677168854537e-05, + "loss": 0.029, + "step": 72360 + }, + { + "epoch": 4.741531808949748, + "grad_norm": 0.27998220920562744, + "learning_rate": 1.946058342571065e-05, + "loss": 0.0559, + "step": 72370 + }, + { + "epoch": 4.742186988141256, + "grad_norm": 0.6569376587867737, + "learning_rate": 1.9447493022311424e-05, + "loss": 0.044, + "step": 72380 + }, + { + "epoch": 4.742842167332766, + "grad_norm": 0.10365577042102814, + "learning_rate": 1.9434405960088442e-05, + "loss": 0.0365, + "step": 72390 + }, + { + "epoch": 4.743497346524275, + "grad_norm": 0.16571490466594696, + "learning_rate": 1.9421322240472857e-05, + "loss": 0.0307, + "step": 72400 + }, + { + "epoch": 4.7441525257157835, + "grad_norm": 0.7824513912200928, + "learning_rate": 1.9408241864895494e-05, + "loss": 0.0349, + "step": 72410 + }, + { + "epoch": 4.744807704907292, + "grad_norm": 0.08924952149391174, + "learning_rate": 1.939516483478681e-05, + "loss": 0.0371, + "step": 72420 + }, + { + "epoch": 4.745462884098801, + "grad_norm": 0.6770663261413574, + "learning_rate": 1.9382091151576902e-05, + "loss": 0.0462, + "step": 72430 + }, + { + "epoch": 4.74611806329031, + "grad_norm": 0.6257272958755493, + "learning_rate": 1.9369020816695454e-05, + "loss": 0.0248, + "step": 72440 + }, + { + "epoch": 4.746773242481819, + "grad_norm": 0.229884073138237, + "learning_rate": 1.935595383157184e-05, + "loss": 0.0307, + "step": 72450 + }, + { + "epoch": 4.747428421673328, + "grad_norm": 0.15116919577121735, + "learning_rate": 1.9342890197635057e-05, + "loss": 0.026, + "step": 72460 + }, + { + "epoch": 4.748083600864836, + "grad_norm": 0.47455471754074097, + "learning_rate": 1.9329829916313685e-05, + "loss": 0.0237, + "step": 72470 + }, + { + "epoch": 4.748738780056345, + "grad_norm": 0.47543469071388245, + "learning_rate": 1.9316772989036013e-05, + "loss": 0.0219, + "step": 72480 + }, + { + "epoch": 4.749393959247854, + "grad_norm": 0.18440069258213043, + "learning_rate": 1.930371941722991e-05, + "loss": 0.0145, + "step": 72490 + }, + { + "epoch": 4.750049138439363, + "grad_norm": 0.3682436943054199, + "learning_rate": 1.929066920232292e-05, + "loss": 0.0383, + "step": 72500 + }, + { + "epoch": 4.750704317630872, + "grad_norm": 0.190146803855896, + "learning_rate": 1.9277622345742152e-05, + "loss": 0.0222, + "step": 72510 + }, + { + "epoch": 4.751359496822381, + "grad_norm": 0.7000535130500793, + "learning_rate": 1.926457884891444e-05, + "loss": 0.0346, + "step": 72520 + }, + { + "epoch": 4.75201467601389, + "grad_norm": 0.19679249823093414, + "learning_rate": 1.9251538713266147e-05, + "loss": 0.0217, + "step": 72530 + }, + { + "epoch": 4.752669855205399, + "grad_norm": 0.17634259164333344, + "learning_rate": 1.9238501940223358e-05, + "loss": 0.032, + "step": 72540 + }, + { + "epoch": 4.753325034396908, + "grad_norm": 0.9784446358680725, + "learning_rate": 1.922546853121175e-05, + "loss": 0.0262, + "step": 72550 + }, + { + "epoch": 4.753980213588417, + "grad_norm": 0.42760443687438965, + "learning_rate": 1.9212438487656647e-05, + "loss": 0.0303, + "step": 72560 + }, + { + "epoch": 4.754635392779925, + "grad_norm": 0.1390870213508606, + "learning_rate": 1.9199411810982965e-05, + "loss": 0.0329, + "step": 72570 + }, + { + "epoch": 4.755290571971434, + "grad_norm": 0.1751706749200821, + "learning_rate": 1.918638850261532e-05, + "loss": 0.0193, + "step": 72580 + }, + { + "epoch": 4.755945751162943, + "grad_norm": 0.9020259976387024, + "learning_rate": 1.9173368563977885e-05, + "loss": 0.0316, + "step": 72590 + }, + { + "epoch": 4.756600930354452, + "grad_norm": 0.13470125198364258, + "learning_rate": 1.9160351996494513e-05, + "loss": 0.0232, + "step": 72600 + }, + { + "epoch": 4.757256109545961, + "grad_norm": 1.2737748622894287, + "learning_rate": 1.9147338801588677e-05, + "loss": 0.0343, + "step": 72610 + }, + { + "epoch": 4.7579112887374695, + "grad_norm": 0.670471727848053, + "learning_rate": 1.913432898068351e-05, + "loss": 0.0289, + "step": 72620 + }, + { + "epoch": 4.758566467928978, + "grad_norm": 0.6563805341720581, + "learning_rate": 1.912132253520169e-05, + "loss": 0.0213, + "step": 72630 + }, + { + "epoch": 4.759221647120487, + "grad_norm": 0.132045179605484, + "learning_rate": 1.910831946656564e-05, + "loss": 0.0285, + "step": 72640 + }, + { + "epoch": 4.759876826311997, + "grad_norm": 0.21459397673606873, + "learning_rate": 1.909531977619731e-05, + "loss": 0.0558, + "step": 72650 + }, + { + "epoch": 4.760532005503505, + "grad_norm": 1.060659646987915, + "learning_rate": 1.908232346551834e-05, + "loss": 0.057, + "step": 72660 + }, + { + "epoch": 4.7611871846950145, + "grad_norm": 0.42845413088798523, + "learning_rate": 1.9069330535949993e-05, + "loss": 0.0248, + "step": 72670 + }, + { + "epoch": 4.761842363886523, + "grad_norm": 0.6989598274230957, + "learning_rate": 1.9056340988913167e-05, + "loss": 0.0336, + "step": 72680 + }, + { + "epoch": 4.762497543078032, + "grad_norm": 0.4545632600784302, + "learning_rate": 1.9043354825828357e-05, + "loss": 0.0169, + "step": 72690 + }, + { + "epoch": 4.763152722269541, + "grad_norm": 1.8435852527618408, + "learning_rate": 1.9030372048115724e-05, + "loss": 0.0261, + "step": 72700 + }, + { + "epoch": 4.76380790146105, + "grad_norm": 1.1413952112197876, + "learning_rate": 1.9017392657195056e-05, + "loss": 0.0207, + "step": 72710 + }, + { + "epoch": 4.7644630806525585, + "grad_norm": 0.2507939040660858, + "learning_rate": 1.9004416654485724e-05, + "loss": 0.0276, + "step": 72720 + }, + { + "epoch": 4.765118259844067, + "grad_norm": 1.1754975318908691, + "learning_rate": 1.8991444041406785e-05, + "loss": 0.034, + "step": 72730 + }, + { + "epoch": 4.765773439035576, + "grad_norm": 0.8625389337539673, + "learning_rate": 1.8978474819376912e-05, + "loss": 0.0438, + "step": 72740 + }, + { + "epoch": 4.766428618227085, + "grad_norm": 0.18573541939258575, + "learning_rate": 1.8965508989814406e-05, + "loss": 0.0368, + "step": 72750 + }, + { + "epoch": 4.767083797418594, + "grad_norm": 0.6024415493011475, + "learning_rate": 1.8952546554137158e-05, + "loss": 0.03, + "step": 72760 + }, + { + "epoch": 4.767738976610103, + "grad_norm": 0.20353858172893524, + "learning_rate": 1.8939587513762757e-05, + "loss": 0.0264, + "step": 72770 + }, + { + "epoch": 4.768394155801611, + "grad_norm": 0.16036169230937958, + "learning_rate": 1.892663187010835e-05, + "loss": 0.0166, + "step": 72780 + }, + { + "epoch": 4.76904933499312, + "grad_norm": 0.3374277949333191, + "learning_rate": 1.891367962459077e-05, + "loss": 0.0215, + "step": 72790 + }, + { + "epoch": 4.76970451418463, + "grad_norm": 0.4296216666698456, + "learning_rate": 1.890073077862644e-05, + "loss": 0.0209, + "step": 72800 + }, + { + "epoch": 4.770359693376139, + "grad_norm": 0.20306839048862457, + "learning_rate": 1.888778533363146e-05, + "loss": 0.0293, + "step": 72810 + }, + { + "epoch": 4.771014872567648, + "grad_norm": 0.7065244317054749, + "learning_rate": 1.8874843291021478e-05, + "loss": 0.0275, + "step": 72820 + }, + { + "epoch": 4.771670051759156, + "grad_norm": 0.08472827821969986, + "learning_rate": 1.8861904652211853e-05, + "loss": 0.0199, + "step": 72830 + }, + { + "epoch": 4.772325230950665, + "grad_norm": 0.10924891382455826, + "learning_rate": 1.8848969418617507e-05, + "loss": 0.0356, + "step": 72840 + }, + { + "epoch": 4.772980410142174, + "grad_norm": 0.8564274907112122, + "learning_rate": 1.8836037591653044e-05, + "loss": 0.0226, + "step": 72850 + }, + { + "epoch": 4.773635589333683, + "grad_norm": 0.5482841730117798, + "learning_rate": 1.882310917273262e-05, + "loss": 0.0293, + "step": 72860 + }, + { + "epoch": 4.774290768525192, + "grad_norm": 0.17447710037231445, + "learning_rate": 1.881018416327014e-05, + "loss": 0.0238, + "step": 72870 + }, + { + "epoch": 4.7749459477167004, + "grad_norm": 0.6505233645439148, + "learning_rate": 1.8797262564679014e-05, + "loss": 0.0298, + "step": 72880 + }, + { + "epoch": 4.775601126908209, + "grad_norm": 0.5335474014282227, + "learning_rate": 1.8784344378372355e-05, + "loss": 0.0229, + "step": 72890 + }, + { + "epoch": 4.776256306099718, + "grad_norm": 0.3529595136642456, + "learning_rate": 1.8771429605762843e-05, + "loss": 0.0244, + "step": 72900 + }, + { + "epoch": 4.776911485291227, + "grad_norm": 0.6100154519081116, + "learning_rate": 1.8758518248262835e-05, + "loss": 0.0321, + "step": 72910 + }, + { + "epoch": 4.777566664482736, + "grad_norm": 0.864014744758606, + "learning_rate": 1.8745610307284305e-05, + "loss": 0.0224, + "step": 72920 + }, + { + "epoch": 4.7782218436742445, + "grad_norm": 0.17697066068649292, + "learning_rate": 1.8732705784238862e-05, + "loss": 0.0236, + "step": 72930 + }, + { + "epoch": 4.778877022865753, + "grad_norm": 0.7350736856460571, + "learning_rate": 1.8719804680537682e-05, + "loss": 0.0348, + "step": 72940 + }, + { + "epoch": 4.779532202057263, + "grad_norm": 0.24562111496925354, + "learning_rate": 1.870690699759164e-05, + "loss": 0.0234, + "step": 72950 + }, + { + "epoch": 4.780187381248772, + "grad_norm": 0.8306711316108704, + "learning_rate": 1.8694012736811213e-05, + "loss": 0.041, + "step": 72960 + }, + { + "epoch": 4.780842560440281, + "grad_norm": 0.6380283832550049, + "learning_rate": 1.8681121899606463e-05, + "loss": 0.0467, + "step": 72970 + }, + { + "epoch": 4.7814977396317895, + "grad_norm": 0.9783010482788086, + "learning_rate": 1.8668234487387144e-05, + "loss": 0.0246, + "step": 72980 + }, + { + "epoch": 4.782152918823298, + "grad_norm": 0.1292860060930252, + "learning_rate": 1.8655350501562592e-05, + "loss": 0.0227, + "step": 72990 + }, + { + "epoch": 4.782808098014807, + "grad_norm": 0.9861282110214233, + "learning_rate": 1.8642469943541796e-05, + "loss": 0.0454, + "step": 73000 + }, + { + "epoch": 4.783463277206316, + "grad_norm": 0.5449879765510559, + "learning_rate": 1.8629592814733328e-05, + "loss": 0.038, + "step": 73010 + }, + { + "epoch": 4.784118456397825, + "grad_norm": 0.26353660225868225, + "learning_rate": 1.8616719116545438e-05, + "loss": 0.0174, + "step": 73020 + }, + { + "epoch": 4.7847736355893336, + "grad_norm": 0.2647213935852051, + "learning_rate": 1.8603848850385952e-05, + "loss": 0.0242, + "step": 73030 + }, + { + "epoch": 4.785428814780842, + "grad_norm": 0.344488263130188, + "learning_rate": 1.859098201766236e-05, + "loss": 0.0295, + "step": 73040 + }, + { + "epoch": 4.786083993972351, + "grad_norm": 0.3276762068271637, + "learning_rate": 1.8578118619781726e-05, + "loss": 0.0369, + "step": 73050 + }, + { + "epoch": 4.78673917316386, + "grad_norm": 0.1444559097290039, + "learning_rate": 1.856525865815083e-05, + "loss": 0.0506, + "step": 73060 + }, + { + "epoch": 4.787394352355369, + "grad_norm": 0.31049391627311707, + "learning_rate": 1.8552402134175966e-05, + "loss": 0.0287, + "step": 73070 + }, + { + "epoch": 4.7880495315468785, + "grad_norm": 1.7357027530670166, + "learning_rate": 1.8539549049263143e-05, + "loss": 0.0286, + "step": 73080 + }, + { + "epoch": 4.788704710738387, + "grad_norm": 0.9313951730728149, + "learning_rate": 1.8526699404817917e-05, + "loss": 0.0209, + "step": 73090 + }, + { + "epoch": 4.789359889929896, + "grad_norm": 0.17115053534507751, + "learning_rate": 1.8513853202245546e-05, + "loss": 0.0204, + "step": 73100 + }, + { + "epoch": 4.790015069121405, + "grad_norm": 0.758599042892456, + "learning_rate": 1.850101044295081e-05, + "loss": 0.0291, + "step": 73110 + }, + { + "epoch": 4.790670248312914, + "grad_norm": 0.3026372194290161, + "learning_rate": 1.848817112833825e-05, + "loss": 0.0286, + "step": 73120 + }, + { + "epoch": 4.791325427504423, + "grad_norm": 0.2852189540863037, + "learning_rate": 1.8475335259811905e-05, + "loss": 0.0385, + "step": 73130 + }, + { + "epoch": 4.791980606695931, + "grad_norm": 0.25738799571990967, + "learning_rate": 1.8462502838775508e-05, + "loss": 0.0287, + "step": 73140 + }, + { + "epoch": 4.79263578588744, + "grad_norm": 0.5008986592292786, + "learning_rate": 1.8449673866632378e-05, + "loss": 0.0251, + "step": 73150 + }, + { + "epoch": 4.793290965078949, + "grad_norm": 1.263420820236206, + "learning_rate": 1.8436848344785473e-05, + "loss": 0.0361, + "step": 73160 + }, + { + "epoch": 4.793946144270458, + "grad_norm": 0.8993785381317139, + "learning_rate": 1.8424026274637378e-05, + "loss": 0.0241, + "step": 73170 + }, + { + "epoch": 4.794601323461967, + "grad_norm": 0.2527316212654114, + "learning_rate": 1.8411207657590312e-05, + "loss": 0.0163, + "step": 73180 + }, + { + "epoch": 4.7952565026534755, + "grad_norm": 0.38354817032814026, + "learning_rate": 1.839839249504607e-05, + "loss": 0.0305, + "step": 73190 + }, + { + "epoch": 4.795911681844984, + "grad_norm": 0.9228989481925964, + "learning_rate": 1.8385580788406108e-05, + "loss": 0.0427, + "step": 73200 + }, + { + "epoch": 4.796566861036493, + "grad_norm": 0.13189929723739624, + "learning_rate": 1.837277253907152e-05, + "loss": 0.0287, + "step": 73210 + }, + { + "epoch": 4.797222040228002, + "grad_norm": 0.6747673153877258, + "learning_rate": 1.8359967748442957e-05, + "loss": 0.0278, + "step": 73220 + }, + { + "epoch": 4.797877219419512, + "grad_norm": 0.23694023489952087, + "learning_rate": 1.8347166417920774e-05, + "loss": 0.0372, + "step": 73230 + }, + { + "epoch": 4.79853239861102, + "grad_norm": 0.6962522864341736, + "learning_rate": 1.833436854890484e-05, + "loss": 0.0261, + "step": 73240 + }, + { + "epoch": 4.799187577802529, + "grad_norm": 0.97385174036026, + "learning_rate": 1.8321574142794796e-05, + "loss": 0.025, + "step": 73250 + }, + { + "epoch": 4.799842756994038, + "grad_norm": 0.23474827408790588, + "learning_rate": 1.8308783200989748e-05, + "loss": 0.0342, + "step": 73260 + }, + { + "epoch": 4.800497936185547, + "grad_norm": 1.0917640924453735, + "learning_rate": 1.829599572488855e-05, + "loss": 0.0361, + "step": 73270 + }, + { + "epoch": 4.801153115377056, + "grad_norm": 0.5711306929588318, + "learning_rate": 1.828321171588957e-05, + "loss": 0.0359, + "step": 73280 + }, + { + "epoch": 4.8018082945685645, + "grad_norm": 0.19171550869941711, + "learning_rate": 1.827043117539089e-05, + "loss": 0.0374, + "step": 73290 + }, + { + "epoch": 4.802463473760073, + "grad_norm": 0.11125475913286209, + "learning_rate": 1.8257654104790116e-05, + "loss": 0.0443, + "step": 73300 + }, + { + "epoch": 4.803118652951582, + "grad_norm": 0.2514972686767578, + "learning_rate": 1.82448805054846e-05, + "loss": 0.0296, + "step": 73310 + }, + { + "epoch": 4.803773832143091, + "grad_norm": 0.14819712936878204, + "learning_rate": 1.823211037887119e-05, + "loss": 0.0222, + "step": 73320 + }, + { + "epoch": 4.8044290113346, + "grad_norm": 1.0101370811462402, + "learning_rate": 1.821934372634645e-05, + "loss": 0.0336, + "step": 73330 + }, + { + "epoch": 4.805084190526109, + "grad_norm": 0.9441937804222107, + "learning_rate": 1.8206580549306472e-05, + "loss": 0.03, + "step": 73340 + }, + { + "epoch": 4.805739369717617, + "grad_norm": 0.6141395568847656, + "learning_rate": 1.819382084914707e-05, + "loss": 0.0322, + "step": 73350 + }, + { + "epoch": 4.806394548909127, + "grad_norm": 1.0037055015563965, + "learning_rate": 1.8181064627263577e-05, + "loss": 0.0352, + "step": 73360 + }, + { + "epoch": 4.807049728100636, + "grad_norm": 0.3357381224632263, + "learning_rate": 1.8168311885051022e-05, + "loss": 0.0338, + "step": 73370 + }, + { + "epoch": 4.807704907292145, + "grad_norm": 0.679429829120636, + "learning_rate": 1.815556262390402e-05, + "loss": 0.026, + "step": 73380 + }, + { + "epoch": 4.8083600864836535, + "grad_norm": 0.48575690388679504, + "learning_rate": 1.8142816845216826e-05, + "loss": 0.0276, + "step": 73390 + }, + { + "epoch": 4.809015265675162, + "grad_norm": 0.8928332924842834, + "learning_rate": 1.8130074550383276e-05, + "loss": 0.0336, + "step": 73400 + }, + { + "epoch": 4.809670444866671, + "grad_norm": 0.760790228843689, + "learning_rate": 1.8117335740796848e-05, + "loss": 0.0372, + "step": 73410 + }, + { + "epoch": 4.81032562405818, + "grad_norm": 0.7157676815986633, + "learning_rate": 1.810460041785067e-05, + "loss": 0.0308, + "step": 73420 + }, + { + "epoch": 4.810980803249689, + "grad_norm": 0.8153489828109741, + "learning_rate": 1.809186858293742e-05, + "loss": 0.0277, + "step": 73430 + }, + { + "epoch": 4.811635982441198, + "grad_norm": 1.224806547164917, + "learning_rate": 1.8079140237449448e-05, + "loss": 0.0367, + "step": 73440 + }, + { + "epoch": 4.812291161632706, + "grad_norm": 0.2087937295436859, + "learning_rate": 1.8066415382778707e-05, + "loss": 0.0279, + "step": 73450 + }, + { + "epoch": 4.812946340824215, + "grad_norm": 0.11933579295873642, + "learning_rate": 1.8053694020316786e-05, + "loss": 0.0208, + "step": 73460 + }, + { + "epoch": 4.813601520015724, + "grad_norm": 0.15191064774990082, + "learning_rate": 1.8040976151454837e-05, + "loss": 0.0132, + "step": 73470 + }, + { + "epoch": 4.814256699207233, + "grad_norm": 0.6941729784011841, + "learning_rate": 1.8028261777583704e-05, + "loss": 0.0316, + "step": 73480 + }, + { + "epoch": 4.814911878398742, + "grad_norm": 0.1552269458770752, + "learning_rate": 1.8015550900093758e-05, + "loss": 0.0388, + "step": 73490 + }, + { + "epoch": 4.8155670575902505, + "grad_norm": 0.21512481570243835, + "learning_rate": 1.800284352037512e-05, + "loss": 0.0312, + "step": 73500 + }, + { + "epoch": 4.81622223678176, + "grad_norm": 0.5526420474052429, + "learning_rate": 1.799013963981738e-05, + "loss": 0.0239, + "step": 73510 + }, + { + "epoch": 4.816877415973269, + "grad_norm": 0.5391933917999268, + "learning_rate": 1.797743925980987e-05, + "loss": 0.0283, + "step": 73520 + }, + { + "epoch": 4.817532595164778, + "grad_norm": 0.3930329978466034, + "learning_rate": 1.7964742381741435e-05, + "loss": 0.027, + "step": 73530 + }, + { + "epoch": 4.818187774356287, + "grad_norm": 0.513643741607666, + "learning_rate": 1.7952049007000627e-05, + "loss": 0.0394, + "step": 73540 + }, + { + "epoch": 4.8188429535477955, + "grad_norm": 0.30774834752082825, + "learning_rate": 1.7939359136975535e-05, + "loss": 0.016, + "step": 73550 + }, + { + "epoch": 4.819498132739304, + "grad_norm": 0.2640838325023651, + "learning_rate": 1.7926672773053932e-05, + "loss": 0.0308, + "step": 73560 + }, + { + "epoch": 4.820153311930813, + "grad_norm": 1.25593101978302, + "learning_rate": 1.791398991662317e-05, + "loss": 0.0299, + "step": 73570 + }, + { + "epoch": 4.820808491122322, + "grad_norm": 1.2828328609466553, + "learning_rate": 1.7901310569070244e-05, + "loss": 0.023, + "step": 73580 + }, + { + "epoch": 4.821463670313831, + "grad_norm": 1.2204731702804565, + "learning_rate": 1.7888634731781724e-05, + "loss": 0.0478, + "step": 73590 + }, + { + "epoch": 4.8221188495053395, + "grad_norm": 0.09436891973018646, + "learning_rate": 1.7875962406143844e-05, + "loss": 0.0322, + "step": 73600 + }, + { + "epoch": 4.822774028696848, + "grad_norm": 0.5331740379333496, + "learning_rate": 1.7863293593542403e-05, + "loss": 0.0218, + "step": 73610 + }, + { + "epoch": 4.823429207888357, + "grad_norm": 0.44343313574790955, + "learning_rate": 1.7850628295362853e-05, + "loss": 0.0305, + "step": 73620 + }, + { + "epoch": 4.824084387079866, + "grad_norm": 0.9119307398796082, + "learning_rate": 1.7837966512990263e-05, + "loss": 0.0356, + "step": 73630 + }, + { + "epoch": 4.824739566271376, + "grad_norm": 0.5238821506500244, + "learning_rate": 1.782530824780931e-05, + "loss": 0.0463, + "step": 73640 + }, + { + "epoch": 4.825394745462884, + "grad_norm": 0.5758998990058899, + "learning_rate": 1.7812653501204263e-05, + "loss": 0.0402, + "step": 73650 + }, + { + "epoch": 4.826049924654393, + "grad_norm": 0.4741845428943634, + "learning_rate": 1.780000227455903e-05, + "loss": 0.0468, + "step": 73660 + }, + { + "epoch": 4.826705103845902, + "grad_norm": 0.16599075496196747, + "learning_rate": 1.778735456925716e-05, + "loss": 0.0466, + "step": 73670 + }, + { + "epoch": 4.827360283037411, + "grad_norm": 0.4529918134212494, + "learning_rate": 1.7774710386681737e-05, + "loss": 0.0352, + "step": 73680 + }, + { + "epoch": 4.82801546222892, + "grad_norm": 0.8110250234603882, + "learning_rate": 1.7762069728215545e-05, + "loss": 0.0419, + "step": 73690 + }, + { + "epoch": 4.828670641420429, + "grad_norm": 0.09521475434303284, + "learning_rate": 1.7749432595240932e-05, + "loss": 0.0181, + "step": 73700 + }, + { + "epoch": 4.829325820611937, + "grad_norm": 0.5274805426597595, + "learning_rate": 1.773679898913991e-05, + "loss": 0.0202, + "step": 73710 + }, + { + "epoch": 4.829980999803446, + "grad_norm": 0.22296655178070068, + "learning_rate": 1.7724168911294015e-05, + "loss": 0.0177, + "step": 73720 + }, + { + "epoch": 4.830636178994955, + "grad_norm": 0.6459769010543823, + "learning_rate": 1.7711542363084504e-05, + "loss": 0.0354, + "step": 73730 + }, + { + "epoch": 4.831291358186464, + "grad_norm": 0.1400393843650818, + "learning_rate": 1.7698919345892158e-05, + "loss": 0.0447, + "step": 73740 + }, + { + "epoch": 4.831946537377973, + "grad_norm": 1.0072364807128906, + "learning_rate": 1.768629986109743e-05, + "loss": 0.0248, + "step": 73750 + }, + { + "epoch": 4.8326017165694815, + "grad_norm": 0.9050938487052917, + "learning_rate": 1.7673683910080357e-05, + "loss": 0.0284, + "step": 73760 + }, + { + "epoch": 4.83325689576099, + "grad_norm": 0.38292360305786133, + "learning_rate": 1.7661071494220638e-05, + "loss": 0.0277, + "step": 73770 + }, + { + "epoch": 4.833912074952499, + "grad_norm": 0.46316981315612793, + "learning_rate": 1.7648462614897498e-05, + "loss": 0.0248, + "step": 73780 + }, + { + "epoch": 4.834567254144009, + "grad_norm": 0.7314793467521667, + "learning_rate": 1.7635857273489863e-05, + "loss": 0.0493, + "step": 73790 + }, + { + "epoch": 4.835222433335518, + "grad_norm": 0.5088319778442383, + "learning_rate": 1.7623255471376198e-05, + "loss": 0.0297, + "step": 73800 + }, + { + "epoch": 4.835877612527026, + "grad_norm": 0.11524926871061325, + "learning_rate": 1.7610657209934646e-05, + "loss": 0.0177, + "step": 73810 + }, + { + "epoch": 4.836532791718535, + "grad_norm": 1.0519139766693115, + "learning_rate": 1.7598062490542915e-05, + "loss": 0.0276, + "step": 73820 + }, + { + "epoch": 4.837187970910044, + "grad_norm": 0.9716769456863403, + "learning_rate": 1.7585471314578377e-05, + "loss": 0.0403, + "step": 73830 + }, + { + "epoch": 4.837843150101553, + "grad_norm": 1.0490587949752808, + "learning_rate": 1.7572883683417945e-05, + "loss": 0.0465, + "step": 73840 + }, + { + "epoch": 4.838498329293062, + "grad_norm": 0.6984431743621826, + "learning_rate": 1.7560299598438212e-05, + "loss": 0.0302, + "step": 73850 + }, + { + "epoch": 4.8391535084845705, + "grad_norm": 0.1276865303516388, + "learning_rate": 1.7547719061015334e-05, + "loss": 0.019, + "step": 73860 + }, + { + "epoch": 4.839808687676079, + "grad_norm": 0.16453738510608673, + "learning_rate": 1.75351420725251e-05, + "loss": 0.0199, + "step": 73870 + }, + { + "epoch": 4.840463866867588, + "grad_norm": 0.42911776900291443, + "learning_rate": 1.7522568634342923e-05, + "loss": 0.0559, + "step": 73880 + }, + { + "epoch": 4.841119046059097, + "grad_norm": 1.4660457372665405, + "learning_rate": 1.7509998747843825e-05, + "loss": 0.0641, + "step": 73890 + }, + { + "epoch": 4.841774225250606, + "grad_norm": 0.6774313449859619, + "learning_rate": 1.7497432414402403e-05, + "loss": 0.0301, + "step": 73900 + }, + { + "epoch": 4.842429404442115, + "grad_norm": 0.27003562450408936, + "learning_rate": 1.7484869635392903e-05, + "loss": 0.0216, + "step": 73910 + }, + { + "epoch": 4.843084583633624, + "grad_norm": 0.343196302652359, + "learning_rate": 1.747231041218919e-05, + "loss": 0.0337, + "step": 73920 + }, + { + "epoch": 4.843739762825132, + "grad_norm": 0.1766243577003479, + "learning_rate": 1.7459754746164692e-05, + "loss": 0.0278, + "step": 73930 + }, + { + "epoch": 4.844394942016642, + "grad_norm": 0.8423415422439575, + "learning_rate": 1.744720263869249e-05, + "loss": 0.0436, + "step": 73940 + }, + { + "epoch": 4.845050121208151, + "grad_norm": 0.39765334129333496, + "learning_rate": 1.7434654091145263e-05, + "loss": 0.0125, + "step": 73950 + }, + { + "epoch": 4.8457053003996595, + "grad_norm": 0.23874208331108093, + "learning_rate": 1.742210910489532e-05, + "loss": 0.0203, + "step": 73960 + }, + { + "epoch": 4.846360479591168, + "grad_norm": 1.3565889596939087, + "learning_rate": 1.7409567681314532e-05, + "loss": 0.0627, + "step": 73970 + }, + { + "epoch": 4.847015658782677, + "grad_norm": 1.203203558921814, + "learning_rate": 1.7397029821774434e-05, + "loss": 0.0269, + "step": 73980 + }, + { + "epoch": 4.847670837974186, + "grad_norm": 0.1761227399110794, + "learning_rate": 1.7384495527646126e-05, + "loss": 0.0355, + "step": 73990 + }, + { + "epoch": 4.848326017165695, + "grad_norm": 0.34623733162879944, + "learning_rate": 1.737196480030035e-05, + "loss": 0.0177, + "step": 74000 + }, + { + "epoch": 4.848981196357204, + "grad_norm": 0.09571487456560135, + "learning_rate": 1.7359437641107447e-05, + "loss": 0.031, + "step": 74010 + }, + { + "epoch": 4.849636375548712, + "grad_norm": 0.19834588468074799, + "learning_rate": 1.7346914051437387e-05, + "loss": 0.0423, + "step": 74020 + }, + { + "epoch": 4.850291554740221, + "grad_norm": 0.11983685940504074, + "learning_rate": 1.733439403265969e-05, + "loss": 0.024, + "step": 74030 + }, + { + "epoch": 4.85094673393173, + "grad_norm": 0.5210298299789429, + "learning_rate": 1.732187758614357e-05, + "loss": 0.0285, + "step": 74040 + }, + { + "epoch": 4.851601913123239, + "grad_norm": 0.5320227146148682, + "learning_rate": 1.730936471325777e-05, + "loss": 0.0266, + "step": 74050 + }, + { + "epoch": 4.852257092314748, + "grad_norm": 0.050944734364748, + "learning_rate": 1.7296855415370718e-05, + "loss": 0.0266, + "step": 74060 + }, + { + "epoch": 4.852912271506257, + "grad_norm": 0.43338677287101746, + "learning_rate": 1.728434969385035e-05, + "loss": 0.026, + "step": 74070 + }, + { + "epoch": 4.853567450697766, + "grad_norm": 0.9340763688087463, + "learning_rate": 1.7271847550064346e-05, + "loss": 0.0315, + "step": 74080 + }, + { + "epoch": 4.854222629889275, + "grad_norm": 1.0460237264633179, + "learning_rate": 1.725934898537987e-05, + "loss": 0.0532, + "step": 74090 + }, + { + "epoch": 4.854877809080784, + "grad_norm": 0.4013667404651642, + "learning_rate": 1.7246854001163783e-05, + "loss": 0.0258, + "step": 74100 + }, + { + "epoch": 4.855532988272293, + "grad_norm": 0.2059064507484436, + "learning_rate": 1.7234362598782476e-05, + "loss": 0.0283, + "step": 74110 + }, + { + "epoch": 4.856188167463801, + "grad_norm": 0.4066506028175354, + "learning_rate": 1.722187477960202e-05, + "loss": 0.0155, + "step": 74120 + }, + { + "epoch": 4.85684334665531, + "grad_norm": 0.2568848431110382, + "learning_rate": 1.7209390544988057e-05, + "loss": 0.0441, + "step": 74130 + }, + { + "epoch": 4.857498525846819, + "grad_norm": 0.22095561027526855, + "learning_rate": 1.7196909896305856e-05, + "loss": 0.0213, + "step": 74140 + }, + { + "epoch": 4.858153705038328, + "grad_norm": 0.3502658009529114, + "learning_rate": 1.718443283492026e-05, + "loss": 0.0265, + "step": 74150 + }, + { + "epoch": 4.858808884229837, + "grad_norm": 0.4001671075820923, + "learning_rate": 1.7171959362195746e-05, + "loss": 0.0172, + "step": 74160 + }, + { + "epoch": 4.8594640634213455, + "grad_norm": 0.18135499954223633, + "learning_rate": 1.715948947949642e-05, + "loss": 0.0476, + "step": 74170 + }, + { + "epoch": 4.860119242612854, + "grad_norm": 0.2797158360481262, + "learning_rate": 1.714702318818593e-05, + "loss": 0.0246, + "step": 74180 + }, + { + "epoch": 4.860774421804363, + "grad_norm": 0.9900391101837158, + "learning_rate": 1.7134560489627592e-05, + "loss": 0.0247, + "step": 74190 + }, + { + "epoch": 4.861429600995873, + "grad_norm": 0.2569568157196045, + "learning_rate": 1.712210138518431e-05, + "loss": 0.0254, + "step": 74200 + }, + { + "epoch": 4.862084780187381, + "grad_norm": 1.2789478302001953, + "learning_rate": 1.71096458762186e-05, + "loss": 0.0276, + "step": 74210 + }, + { + "epoch": 4.8627399593788905, + "grad_norm": 0.2162877470254898, + "learning_rate": 1.7097193964092552e-05, + "loss": 0.0539, + "step": 74220 + }, + { + "epoch": 4.863395138570399, + "grad_norm": 0.926478385925293, + "learning_rate": 1.708474565016792e-05, + "loss": 0.0877, + "step": 74230 + }, + { + "epoch": 4.864050317761908, + "grad_norm": 0.09471683204174042, + "learning_rate": 1.7072300935806e-05, + "loss": 0.0266, + "step": 74240 + }, + { + "epoch": 4.864705496953417, + "grad_norm": 0.2807711362838745, + "learning_rate": 1.705985982236776e-05, + "loss": 0.0524, + "step": 74250 + }, + { + "epoch": 4.865360676144926, + "grad_norm": 0.5379073023796082, + "learning_rate": 1.7047422311213697e-05, + "loss": 0.0211, + "step": 74260 + }, + { + "epoch": 4.8660158553364345, + "grad_norm": 1.3069629669189453, + "learning_rate": 1.703498840370401e-05, + "loss": 0.0367, + "step": 74270 + }, + { + "epoch": 4.866671034527943, + "grad_norm": 0.17240937054157257, + "learning_rate": 1.702255810119842e-05, + "loss": 0.0346, + "step": 74280 + }, + { + "epoch": 4.867326213719452, + "grad_norm": 0.14114393293857574, + "learning_rate": 1.7010131405056305e-05, + "loss": 0.024, + "step": 74290 + }, + { + "epoch": 4.867981392910961, + "grad_norm": 0.6062493324279785, + "learning_rate": 1.6997708316636613e-05, + "loss": 0.0247, + "step": 74300 + }, + { + "epoch": 4.86863657210247, + "grad_norm": 0.24413548409938812, + "learning_rate": 1.6985288837297932e-05, + "loss": 0.022, + "step": 74310 + }, + { + "epoch": 4.869291751293979, + "grad_norm": 0.8222270011901855, + "learning_rate": 1.6972872968398394e-05, + "loss": 0.0358, + "step": 74320 + }, + { + "epoch": 4.869946930485487, + "grad_norm": 0.49308010935783386, + "learning_rate": 1.6960460711295845e-05, + "loss": 0.0301, + "step": 74330 + }, + { + "epoch": 4.870602109676996, + "grad_norm": 0.13057412207126617, + "learning_rate": 1.6948052067347623e-05, + "loss": 0.0146, + "step": 74340 + }, + { + "epoch": 4.871257288868506, + "grad_norm": 0.9474713802337646, + "learning_rate": 1.693564703791075e-05, + "loss": 0.0458, + "step": 74350 + }, + { + "epoch": 4.871912468060015, + "grad_norm": 0.6111340522766113, + "learning_rate": 1.6923245624341793e-05, + "loss": 0.0342, + "step": 74360 + }, + { + "epoch": 4.872567647251524, + "grad_norm": 0.5738939642906189, + "learning_rate": 1.6910847827996962e-05, + "loss": 0.0169, + "step": 74370 + }, + { + "epoch": 4.873222826443032, + "grad_norm": 1.3638559579849243, + "learning_rate": 1.689845365023206e-05, + "loss": 0.0565, + "step": 74380 + }, + { + "epoch": 4.873878005634541, + "grad_norm": 0.7429587841033936, + "learning_rate": 1.6886063092402505e-05, + "loss": 0.0346, + "step": 74390 + }, + { + "epoch": 4.87453318482605, + "grad_norm": 0.21170823276042938, + "learning_rate": 1.6873676155863317e-05, + "loss": 0.046, + "step": 74400 + }, + { + "epoch": 4.875188364017559, + "grad_norm": 0.1406799703836441, + "learning_rate": 1.6861292841969085e-05, + "loss": 0.0259, + "step": 74410 + }, + { + "epoch": 4.875843543209068, + "grad_norm": 1.0956056118011475, + "learning_rate": 1.684891315207406e-05, + "loss": 0.0286, + "step": 74420 + }, + { + "epoch": 4.8764987224005765, + "grad_norm": 0.5625009536743164, + "learning_rate": 1.6836537087532034e-05, + "loss": 0.0167, + "step": 74430 + }, + { + "epoch": 4.877153901592085, + "grad_norm": 0.29563188552856445, + "learning_rate": 1.6824164649696465e-05, + "loss": 0.0265, + "step": 74440 + }, + { + "epoch": 4.877809080783594, + "grad_norm": 0.9163342118263245, + "learning_rate": 1.681179583992034e-05, + "loss": 0.0334, + "step": 74450 + }, + { + "epoch": 4.878464259975103, + "grad_norm": 0.6524119973182678, + "learning_rate": 1.679943065955636e-05, + "loss": 0.0239, + "step": 74460 + }, + { + "epoch": 4.879119439166612, + "grad_norm": 0.4667624235153198, + "learning_rate": 1.678706910995671e-05, + "loss": 0.0557, + "step": 74470 + }, + { + "epoch": 4.879774618358121, + "grad_norm": 0.21801859140396118, + "learning_rate": 1.6774711192473262e-05, + "loss": 0.0497, + "step": 74480 + }, + { + "epoch": 4.880429797549629, + "grad_norm": 0.8420490026473999, + "learning_rate": 1.676235690845743e-05, + "loss": 0.0922, + "step": 74490 + }, + { + "epoch": 4.881084976741139, + "grad_norm": 0.6534761786460876, + "learning_rate": 1.6750006259260285e-05, + "loss": 0.0451, + "step": 74500 + }, + { + "epoch": 4.881740155932648, + "grad_norm": 0.9327341318130493, + "learning_rate": 1.673765924623244e-05, + "loss": 0.0592, + "step": 74510 + }, + { + "epoch": 4.882395335124157, + "grad_norm": 0.6474465727806091, + "learning_rate": 1.67253158707242e-05, + "loss": 0.0423, + "step": 74520 + }, + { + "epoch": 4.8830505143156655, + "grad_norm": 0.7198342680931091, + "learning_rate": 1.671297613408537e-05, + "loss": 0.0326, + "step": 74530 + }, + { + "epoch": 4.883705693507174, + "grad_norm": 0.15387894213199615, + "learning_rate": 1.6700640037665444e-05, + "loss": 0.0332, + "step": 74540 + }, + { + "epoch": 4.884360872698683, + "grad_norm": 0.8346863985061646, + "learning_rate": 1.668830758281344e-05, + "loss": 0.0329, + "step": 74550 + }, + { + "epoch": 4.885016051890192, + "grad_norm": 0.09436961263418198, + "learning_rate": 1.6675978770878052e-05, + "loss": 0.0396, + "step": 74560 + }, + { + "epoch": 4.885671231081701, + "grad_norm": 0.37841370701789856, + "learning_rate": 1.6663653603207506e-05, + "loss": 0.0368, + "step": 74570 + }, + { + "epoch": 4.88632641027321, + "grad_norm": 0.582610011100769, + "learning_rate": 1.6651332081149685e-05, + "loss": 0.0386, + "step": 74580 + }, + { + "epoch": 4.886981589464718, + "grad_norm": 0.2348959594964981, + "learning_rate": 1.6639014206052046e-05, + "loss": 0.0201, + "step": 74590 + }, + { + "epoch": 4.887636768656227, + "grad_norm": 0.17975188791751862, + "learning_rate": 1.6626699979261657e-05, + "loss": 0.017, + "step": 74600 + }, + { + "epoch": 4.888291947847736, + "grad_norm": 0.04951559007167816, + "learning_rate": 1.6614389402125197e-05, + "loss": 0.0236, + "step": 74610 + }, + { + "epoch": 4.888947127039245, + "grad_norm": 0.2627407908439636, + "learning_rate": 1.66020824759889e-05, + "loss": 0.0393, + "step": 74620 + }, + { + "epoch": 4.8896023062307545, + "grad_norm": 0.5313182473182678, + "learning_rate": 1.6589779202198675e-05, + "loss": 0.0407, + "step": 74630 + }, + { + "epoch": 4.890257485422263, + "grad_norm": 0.2709634006023407, + "learning_rate": 1.6577479582099924e-05, + "loss": 0.0263, + "step": 74640 + }, + { + "epoch": 4.890912664613772, + "grad_norm": 0.1664564609527588, + "learning_rate": 1.6565183617037788e-05, + "loss": 0.0217, + "step": 74650 + }, + { + "epoch": 4.891567843805281, + "grad_norm": 0.6067521572113037, + "learning_rate": 1.6552891308356894e-05, + "loss": 0.045, + "step": 74660 + }, + { + "epoch": 4.89222302299679, + "grad_norm": 0.9778916239738464, + "learning_rate": 1.6540602657401526e-05, + "loss": 0.0426, + "step": 74670 + }, + { + "epoch": 4.892878202188299, + "grad_norm": 0.5709297060966492, + "learning_rate": 1.6528317665515537e-05, + "loss": 0.0606, + "step": 74680 + }, + { + "epoch": 4.893533381379807, + "grad_norm": 0.2998455762863159, + "learning_rate": 1.6516036334042416e-05, + "loss": 0.0171, + "step": 74690 + }, + { + "epoch": 4.894188560571316, + "grad_norm": 0.07690609246492386, + "learning_rate": 1.650375866432519e-05, + "loss": 0.0151, + "step": 74700 + }, + { + "epoch": 4.894843739762825, + "grad_norm": 0.6806520819664001, + "learning_rate": 1.649148465770659e-05, + "loss": 0.0397, + "step": 74710 + }, + { + "epoch": 4.895498918954334, + "grad_norm": 0.31622540950775146, + "learning_rate": 1.647921431552883e-05, + "loss": 0.0257, + "step": 74720 + }, + { + "epoch": 4.896154098145843, + "grad_norm": 0.0808551087975502, + "learning_rate": 1.646694763913381e-05, + "loss": 0.035, + "step": 74730 + }, + { + "epoch": 4.8968092773373515, + "grad_norm": 0.444078654050827, + "learning_rate": 1.6454684629862966e-05, + "loss": 0.0293, + "step": 74740 + }, + { + "epoch": 4.89746445652886, + "grad_norm": 0.21934139728546143, + "learning_rate": 1.6442425289057388e-05, + "loss": 0.0159, + "step": 74750 + }, + { + "epoch": 4.89811963572037, + "grad_norm": 0.6302344799041748, + "learning_rate": 1.6430169618057718e-05, + "loss": 0.0185, + "step": 74760 + }, + { + "epoch": 4.898774814911878, + "grad_norm": 0.15471941232681274, + "learning_rate": 1.641791761820422e-05, + "loss": 0.0148, + "step": 74770 + }, + { + "epoch": 4.899429994103388, + "grad_norm": 0.4121875464916229, + "learning_rate": 1.640566929083676e-05, + "loss": 0.0248, + "step": 74780 + }, + { + "epoch": 4.9000851732948965, + "grad_norm": 0.6732590198516846, + "learning_rate": 1.6393424637294823e-05, + "loss": 0.0346, + "step": 74790 + }, + { + "epoch": 4.900740352486405, + "grad_norm": 0.4354810416698456, + "learning_rate": 1.638118365891742e-05, + "loss": 0.0305, + "step": 74800 + }, + { + "epoch": 4.901395531677914, + "grad_norm": 0.2917622923851013, + "learning_rate": 1.636894635704323e-05, + "loss": 0.0221, + "step": 74810 + }, + { + "epoch": 4.902050710869423, + "grad_norm": 0.5392617583274841, + "learning_rate": 1.6356712733010517e-05, + "loss": 0.0282, + "step": 74820 + }, + { + "epoch": 4.902705890060932, + "grad_norm": 0.1497381180524826, + "learning_rate": 1.6344482788157096e-05, + "loss": 0.0108, + "step": 74830 + }, + { + "epoch": 4.9033610692524405, + "grad_norm": 0.12536677718162537, + "learning_rate": 1.6332256523820444e-05, + "loss": 0.0328, + "step": 74840 + }, + { + "epoch": 4.904016248443949, + "grad_norm": 0.9823005199432373, + "learning_rate": 1.6320033941337597e-05, + "loss": 0.035, + "step": 74850 + }, + { + "epoch": 4.904671427635458, + "grad_norm": 0.3865206241607666, + "learning_rate": 1.6307815042045216e-05, + "loss": 0.0214, + "step": 74860 + }, + { + "epoch": 4.905326606826967, + "grad_norm": 0.3752003014087677, + "learning_rate": 1.6295599827279513e-05, + "loss": 0.0199, + "step": 74870 + }, + { + "epoch": 4.905981786018476, + "grad_norm": 0.1399679332971573, + "learning_rate": 1.6283388298376356e-05, + "loss": 0.0142, + "step": 74880 + }, + { + "epoch": 4.906636965209985, + "grad_norm": 0.18468354642391205, + "learning_rate": 1.6271180456671125e-05, + "loss": 0.0334, + "step": 74890 + }, + { + "epoch": 4.907292144401493, + "grad_norm": 0.6274508237838745, + "learning_rate": 1.625897630349893e-05, + "loss": 0.0649, + "step": 74900 + }, + { + "epoch": 4.907947323593003, + "grad_norm": 0.7371634244918823, + "learning_rate": 1.6246775840194335e-05, + "loss": 0.0571, + "step": 74910 + }, + { + "epoch": 4.908602502784512, + "grad_norm": 0.3012319803237915, + "learning_rate": 1.6234579068091604e-05, + "loss": 0.03, + "step": 74920 + }, + { + "epoch": 4.909257681976021, + "grad_norm": 0.1644083559513092, + "learning_rate": 1.6222385988524524e-05, + "loss": 0.0136, + "step": 74930 + }, + { + "epoch": 4.90991286116753, + "grad_norm": 0.2957208752632141, + "learning_rate": 1.6210196602826544e-05, + "loss": 0.0264, + "step": 74940 + }, + { + "epoch": 4.910568040359038, + "grad_norm": 0.3552180230617523, + "learning_rate": 1.6198010912330646e-05, + "loss": 0.0234, + "step": 74950 + }, + { + "epoch": 4.911223219550547, + "grad_norm": 0.25138673186302185, + "learning_rate": 1.6185828918369456e-05, + "loss": 0.015, + "step": 74960 + }, + { + "epoch": 4.911878398742056, + "grad_norm": 0.5118651390075684, + "learning_rate": 1.617365062227517e-05, + "loss": 0.0222, + "step": 74970 + }, + { + "epoch": 4.912533577933565, + "grad_norm": 0.20856232941150665, + "learning_rate": 1.616147602537962e-05, + "loss": 0.0309, + "step": 74980 + }, + { + "epoch": 4.913188757125074, + "grad_norm": 0.4012579619884491, + "learning_rate": 1.614930512901415e-05, + "loss": 0.0314, + "step": 74990 + }, + { + "epoch": 4.9138439363165825, + "grad_norm": 0.2229444533586502, + "learning_rate": 1.6137137934509804e-05, + "loss": 0.0244, + "step": 75000 + }, + { + "epoch": 4.914499115508091, + "grad_norm": 0.3421940207481384, + "learning_rate": 1.6124974443197123e-05, + "loss": 0.0433, + "step": 75010 + }, + { + "epoch": 4.9151542946996, + "grad_norm": 0.590463399887085, + "learning_rate": 1.6112814656406306e-05, + "loss": 0.0313, + "step": 75020 + }, + { + "epoch": 4.915809473891109, + "grad_norm": 0.5006409883499146, + "learning_rate": 1.6100658575467125e-05, + "loss": 0.0267, + "step": 75030 + }, + { + "epoch": 4.916464653082618, + "grad_norm": 0.10046325623989105, + "learning_rate": 1.6088506201708974e-05, + "loss": 0.0373, + "step": 75040 + }, + { + "epoch": 4.9171198322741265, + "grad_norm": 0.5091058611869812, + "learning_rate": 1.6076357536460785e-05, + "loss": 0.0233, + "step": 75050 + }, + { + "epoch": 4.917775011465636, + "grad_norm": 0.12337964028120041, + "learning_rate": 1.606421258105113e-05, + "loss": 0.0671, + "step": 75060 + }, + { + "epoch": 4.918430190657145, + "grad_norm": 0.4294377267360687, + "learning_rate": 1.6052071336808183e-05, + "loss": 0.0821, + "step": 75070 + }, + { + "epoch": 4.919085369848654, + "grad_norm": 0.37597936391830444, + "learning_rate": 1.6039933805059658e-05, + "loss": 0.0308, + "step": 75080 + }, + { + "epoch": 4.919740549040163, + "grad_norm": 0.3852689266204834, + "learning_rate": 1.6027799987132915e-05, + "loss": 0.0364, + "step": 75090 + }, + { + "epoch": 4.9203957282316715, + "grad_norm": 0.669550895690918, + "learning_rate": 1.6015669884354888e-05, + "loss": 0.0293, + "step": 75100 + }, + { + "epoch": 4.92105090742318, + "grad_norm": 0.10605680197477341, + "learning_rate": 1.6003543498052125e-05, + "loss": 0.0245, + "step": 75110 + }, + { + "epoch": 4.921706086614689, + "grad_norm": 0.228699192404747, + "learning_rate": 1.5991420829550715e-05, + "loss": 0.0513, + "step": 75120 + }, + { + "epoch": 4.922361265806198, + "grad_norm": 0.8926640152931213, + "learning_rate": 1.5979301880176405e-05, + "loss": 0.0377, + "step": 75130 + }, + { + "epoch": 4.923016444997707, + "grad_norm": 0.09715364873409271, + "learning_rate": 1.5967186651254485e-05, + "loss": 0.023, + "step": 75140 + }, + { + "epoch": 4.923671624189216, + "grad_norm": 0.709206223487854, + "learning_rate": 1.5955075144109853e-05, + "loss": 0.0254, + "step": 75150 + }, + { + "epoch": 4.924326803380724, + "grad_norm": 0.2529566287994385, + "learning_rate": 1.594296736006703e-05, + "loss": 0.0192, + "step": 75160 + }, + { + "epoch": 4.924981982572233, + "grad_norm": 0.4432240128517151, + "learning_rate": 1.59308633004501e-05, + "loss": 0.0231, + "step": 75170 + }, + { + "epoch": 4.925637161763742, + "grad_norm": 0.28509533405303955, + "learning_rate": 1.5918762966582724e-05, + "loss": 0.019, + "step": 75180 + }, + { + "epoch": 4.926292340955252, + "grad_norm": 0.22602473199367523, + "learning_rate": 1.5906666359788203e-05, + "loss": 0.0342, + "step": 75190 + }, + { + "epoch": 4.9269475201467605, + "grad_norm": 3.083559989929199, + "learning_rate": 1.589457348138938e-05, + "loss": 0.0643, + "step": 75200 + }, + { + "epoch": 4.927602699338269, + "grad_norm": 0.3874005675315857, + "learning_rate": 1.588248433270872e-05, + "loss": 0.0401, + "step": 75210 + }, + { + "epoch": 4.928257878529778, + "grad_norm": 0.18164004385471344, + "learning_rate": 1.587039891506828e-05, + "loss": 0.036, + "step": 75220 + }, + { + "epoch": 4.928913057721287, + "grad_norm": 1.1551742553710938, + "learning_rate": 1.5858317229789715e-05, + "loss": 0.0293, + "step": 75230 + }, + { + "epoch": 4.929568236912796, + "grad_norm": 0.26157310605049133, + "learning_rate": 1.584623927819423e-05, + "loss": 0.0349, + "step": 75240 + }, + { + "epoch": 4.930223416104305, + "grad_norm": 0.5632299780845642, + "learning_rate": 1.583416506160269e-05, + "loss": 0.0286, + "step": 75250 + }, + { + "epoch": 4.930878595295813, + "grad_norm": 0.633072555065155, + "learning_rate": 1.5822094581335472e-05, + "loss": 0.0414, + "step": 75260 + }, + { + "epoch": 4.931533774487322, + "grad_norm": 0.3941613435745239, + "learning_rate": 1.5810027838712604e-05, + "loss": 0.0399, + "step": 75270 + }, + { + "epoch": 4.932188953678831, + "grad_norm": 0.2579978406429291, + "learning_rate": 1.5797964835053687e-05, + "loss": 0.0338, + "step": 75280 + }, + { + "epoch": 4.93284413287034, + "grad_norm": 1.247419834136963, + "learning_rate": 1.5785905571677933e-05, + "loss": 0.0445, + "step": 75290 + }, + { + "epoch": 4.933499312061849, + "grad_norm": 0.2460106760263443, + "learning_rate": 1.5773850049904086e-05, + "loss": 0.0304, + "step": 75300 + }, + { + "epoch": 4.9341544912533575, + "grad_norm": 0.8736498355865479, + "learning_rate": 1.5761798271050543e-05, + "loss": 0.0379, + "step": 75310 + }, + { + "epoch": 4.934809670444866, + "grad_norm": 0.3158479630947113, + "learning_rate": 1.5749750236435277e-05, + "loss": 0.0121, + "step": 75320 + }, + { + "epoch": 4.935464849636375, + "grad_norm": 0.8755660057067871, + "learning_rate": 1.5737705947375815e-05, + "loss": 0.0306, + "step": 75330 + }, + { + "epoch": 4.936120028827885, + "grad_norm": 0.8904792666435242, + "learning_rate": 1.5725665405189322e-05, + "loss": 0.036, + "step": 75340 + }, + { + "epoch": 4.936775208019394, + "grad_norm": 1.0152885913848877, + "learning_rate": 1.571362861119253e-05, + "loss": 0.0249, + "step": 75350 + }, + { + "epoch": 4.937430387210902, + "grad_norm": 0.18568375706672668, + "learning_rate": 1.570159556670177e-05, + "loss": 0.0305, + "step": 75360 + }, + { + "epoch": 4.938085566402411, + "grad_norm": 0.6092580556869507, + "learning_rate": 1.5689566273032946e-05, + "loss": 0.0181, + "step": 75370 + }, + { + "epoch": 4.93874074559392, + "grad_norm": 0.11930180341005325, + "learning_rate": 1.5677540731501583e-05, + "loss": 0.0361, + "step": 75380 + }, + { + "epoch": 4.939395924785429, + "grad_norm": 0.6111224293708801, + "learning_rate": 1.5665518943422746e-05, + "loss": 0.0287, + "step": 75390 + }, + { + "epoch": 4.940051103976938, + "grad_norm": 0.5580390691757202, + "learning_rate": 1.5653500910111134e-05, + "loss": 0.0176, + "step": 75400 + }, + { + "epoch": 4.9407062831684465, + "grad_norm": 1.0655443668365479, + "learning_rate": 1.5641486632881023e-05, + "loss": 0.0438, + "step": 75410 + }, + { + "epoch": 4.941361462359955, + "grad_norm": 0.40580350160598755, + "learning_rate": 1.56294761130463e-05, + "loss": 0.0207, + "step": 75420 + }, + { + "epoch": 4.942016641551464, + "grad_norm": 0.14530198276042938, + "learning_rate": 1.5617469351920373e-05, + "loss": 0.0442, + "step": 75430 + }, + { + "epoch": 4.942671820742973, + "grad_norm": 0.29546239972114563, + "learning_rate": 1.5605466350816315e-05, + "loss": 0.0157, + "step": 75440 + }, + { + "epoch": 4.943326999934482, + "grad_norm": 0.21707044541835785, + "learning_rate": 1.559346711104674e-05, + "loss": 0.029, + "step": 75450 + }, + { + "epoch": 4.943982179125991, + "grad_norm": 0.4003237187862396, + "learning_rate": 1.5581471633923883e-05, + "loss": 0.032, + "step": 75460 + }, + { + "epoch": 4.9446373583175, + "grad_norm": 0.6740651726722717, + "learning_rate": 1.5569479920759505e-05, + "loss": 0.0503, + "step": 75470 + }, + { + "epoch": 4.945292537509009, + "grad_norm": 0.15790805220603943, + "learning_rate": 1.5557491972865072e-05, + "loss": 0.0236, + "step": 75480 + }, + { + "epoch": 4.945947716700518, + "grad_norm": 0.24794220924377441, + "learning_rate": 1.554550779155152e-05, + "loss": 0.0209, + "step": 75490 + }, + { + "epoch": 4.946602895892027, + "grad_norm": 1.26435124874115, + "learning_rate": 1.5533527378129453e-05, + "loss": 0.0481, + "step": 75500 + }, + { + "epoch": 4.9472580750835355, + "grad_norm": 0.780264675617218, + "learning_rate": 1.552155073390899e-05, + "loss": 0.0355, + "step": 75510 + }, + { + "epoch": 4.947913254275044, + "grad_norm": 0.24702902138233185, + "learning_rate": 1.5509577860199903e-05, + "loss": 0.0224, + "step": 75520 + }, + { + "epoch": 4.948568433466553, + "grad_norm": 1.0076236724853516, + "learning_rate": 1.5497608758311525e-05, + "loss": 0.0236, + "step": 75530 + }, + { + "epoch": 4.949223612658062, + "grad_norm": 0.2565249502658844, + "learning_rate": 1.5485643429552797e-05, + "loss": 0.0224, + "step": 75540 + }, + { + "epoch": 4.949878791849571, + "grad_norm": 0.1686505526304245, + "learning_rate": 1.5473681875232197e-05, + "loss": 0.0319, + "step": 75550 + }, + { + "epoch": 4.95053397104108, + "grad_norm": 0.34087398648262024, + "learning_rate": 1.5461724096657837e-05, + "loss": 0.0328, + "step": 75560 + }, + { + "epoch": 4.951189150232588, + "grad_norm": 0.6081360578536987, + "learning_rate": 1.5449770095137416e-05, + "loss": 0.0265, + "step": 75570 + }, + { + "epoch": 4.951844329424097, + "grad_norm": 0.18654419481754303, + "learning_rate": 1.5437819871978178e-05, + "loss": 0.0475, + "step": 75580 + }, + { + "epoch": 4.952499508615606, + "grad_norm": 2.919269323348999, + "learning_rate": 1.5425873428486992e-05, + "loss": 0.0677, + "step": 75590 + }, + { + "epoch": 4.953154687807115, + "grad_norm": 0.4405401945114136, + "learning_rate": 1.5413930765970312e-05, + "loss": 0.0331, + "step": 75600 + }, + { + "epoch": 4.953809866998624, + "grad_norm": 0.7162681818008423, + "learning_rate": 1.5401991885734175e-05, + "loss": 0.0222, + "step": 75610 + }, + { + "epoch": 4.954465046190133, + "grad_norm": 0.10969268530607224, + "learning_rate": 1.5390056789084174e-05, + "loss": 0.0237, + "step": 75620 + }, + { + "epoch": 4.955120225381642, + "grad_norm": 0.16900849342346191, + "learning_rate": 1.5378125477325545e-05, + "loss": 0.039, + "step": 75630 + }, + { + "epoch": 4.955775404573151, + "grad_norm": 0.10111915320158005, + "learning_rate": 1.5366197951763046e-05, + "loss": 0.0196, + "step": 75640 + }, + { + "epoch": 4.95643058376466, + "grad_norm": 0.4191438853740692, + "learning_rate": 1.5354274213701075e-05, + "loss": 0.0452, + "step": 75650 + }, + { + "epoch": 4.957085762956169, + "grad_norm": 0.5334494709968567, + "learning_rate": 1.5342354264443566e-05, + "loss": 0.0228, + "step": 75660 + }, + { + "epoch": 4.9577409421476775, + "grad_norm": 0.25221043825149536, + "learning_rate": 1.533043810529411e-05, + "loss": 0.028, + "step": 75670 + }, + { + "epoch": 4.958396121339186, + "grad_norm": 0.37776806950569153, + "learning_rate": 1.5318525737555807e-05, + "loss": 0.0287, + "step": 75680 + }, + { + "epoch": 4.959051300530695, + "grad_norm": 0.8155661225318909, + "learning_rate": 1.53066171625314e-05, + "loss": 0.0357, + "step": 75690 + }, + { + "epoch": 4.959706479722204, + "grad_norm": 0.09462904185056686, + "learning_rate": 1.529471238152317e-05, + "loss": 0.0305, + "step": 75700 + }, + { + "epoch": 4.960361658913713, + "grad_norm": 0.5789101123809814, + "learning_rate": 1.5282811395833023e-05, + "loss": 0.0361, + "step": 75710 + }, + { + "epoch": 4.9610168381052215, + "grad_norm": 0.8160014152526855, + "learning_rate": 1.5270914206762397e-05, + "loss": 0.0201, + "step": 75720 + }, + { + "epoch": 4.96167201729673, + "grad_norm": 0.8754143118858337, + "learning_rate": 1.525902081561241e-05, + "loss": 0.0228, + "step": 75730 + }, + { + "epoch": 4.962327196488239, + "grad_norm": 0.3072887063026428, + "learning_rate": 1.5247131223683664e-05, + "loss": 0.0235, + "step": 75740 + }, + { + "epoch": 4.962982375679749, + "grad_norm": 0.21498288214206696, + "learning_rate": 1.5235245432276408e-05, + "loss": 0.0197, + "step": 75750 + }, + { + "epoch": 4.963637554871257, + "grad_norm": 0.3945547938346863, + "learning_rate": 1.5223363442690424e-05, + "loss": 0.0227, + "step": 75760 + }, + { + "epoch": 4.9642927340627665, + "grad_norm": 0.45819634199142456, + "learning_rate": 1.5211485256225133e-05, + "loss": 0.0422, + "step": 75770 + }, + { + "epoch": 4.964947913254275, + "grad_norm": 0.11742623150348663, + "learning_rate": 1.519961087417951e-05, + "loss": 0.0231, + "step": 75780 + }, + { + "epoch": 4.965603092445784, + "grad_norm": 0.09212666004896164, + "learning_rate": 1.5187740297852137e-05, + "loss": 0.0155, + "step": 75790 + }, + { + "epoch": 4.966258271637293, + "grad_norm": 0.6980666518211365, + "learning_rate": 1.517587352854113e-05, + "loss": 0.0298, + "step": 75800 + }, + { + "epoch": 4.966913450828802, + "grad_norm": 0.8075556755065918, + "learning_rate": 1.516401056754423e-05, + "loss": 0.034, + "step": 75810 + }, + { + "epoch": 4.967568630020311, + "grad_norm": 0.16120193898677826, + "learning_rate": 1.5152151416158783e-05, + "loss": 0.0276, + "step": 75820 + }, + { + "epoch": 4.968223809211819, + "grad_norm": 0.2296963632106781, + "learning_rate": 1.5140296075681642e-05, + "loss": 0.0375, + "step": 75830 + }, + { + "epoch": 4.968878988403328, + "grad_norm": 0.75144362449646, + "learning_rate": 1.5128444547409331e-05, + "loss": 0.025, + "step": 75840 + }, + { + "epoch": 4.969534167594837, + "grad_norm": 0.3507215678691864, + "learning_rate": 1.5116596832637864e-05, + "loss": 0.0254, + "step": 75850 + }, + { + "epoch": 4.970189346786346, + "grad_norm": 0.6652704477310181, + "learning_rate": 1.5104752932662952e-05, + "loss": 0.029, + "step": 75860 + }, + { + "epoch": 4.970844525977855, + "grad_norm": 2.26969051361084, + "learning_rate": 1.5092912848779778e-05, + "loss": 0.0391, + "step": 75870 + }, + { + "epoch": 4.9714997051693635, + "grad_norm": 0.3820829689502716, + "learning_rate": 1.5081076582283194e-05, + "loss": 0.0276, + "step": 75880 + }, + { + "epoch": 4.972154884360872, + "grad_norm": 0.506853461265564, + "learning_rate": 1.5069244134467553e-05, + "loss": 0.0422, + "step": 75890 + }, + { + "epoch": 4.972810063552382, + "grad_norm": 0.5658814311027527, + "learning_rate": 1.5057415506626876e-05, + "loss": 0.0395, + "step": 75900 + }, + { + "epoch": 4.973465242743891, + "grad_norm": 0.1011444553732872, + "learning_rate": 1.5045590700054668e-05, + "loss": 0.0322, + "step": 75910 + }, + { + "epoch": 4.9741204219354, + "grad_norm": 0.27967843413352966, + "learning_rate": 1.5033769716044144e-05, + "loss": 0.0321, + "step": 75920 + }, + { + "epoch": 4.974775601126908, + "grad_norm": 0.46433889865875244, + "learning_rate": 1.5021952555887975e-05, + "loss": 0.0343, + "step": 75930 + }, + { + "epoch": 4.975430780318417, + "grad_norm": 0.22051820158958435, + "learning_rate": 1.50101392208785e-05, + "loss": 0.0368, + "step": 75940 + }, + { + "epoch": 4.976085959509926, + "grad_norm": 2.014352321624756, + "learning_rate": 1.4998329712307579e-05, + "loss": 0.0484, + "step": 75950 + }, + { + "epoch": 4.976741138701435, + "grad_norm": 0.2322292923927307, + "learning_rate": 1.4986524031466708e-05, + "loss": 0.0279, + "step": 75960 + }, + { + "epoch": 4.977396317892944, + "grad_norm": 0.12043117731809616, + "learning_rate": 1.4974722179646916e-05, + "loss": 0.0334, + "step": 75970 + }, + { + "epoch": 4.9780514970844525, + "grad_norm": 0.771994948387146, + "learning_rate": 1.4962924158138847e-05, + "loss": 0.0365, + "step": 75980 + }, + { + "epoch": 4.978706676275961, + "grad_norm": 0.188833087682724, + "learning_rate": 1.4951129968232714e-05, + "loss": 0.0429, + "step": 75990 + }, + { + "epoch": 4.97936185546747, + "grad_norm": 0.3155241310596466, + "learning_rate": 1.4939339611218329e-05, + "loss": 0.0425, + "step": 76000 + }, + { + "epoch": 4.980017034658979, + "grad_norm": 0.36619800329208374, + "learning_rate": 1.4927553088385032e-05, + "loss": 0.0448, + "step": 76010 + }, + { + "epoch": 4.980672213850488, + "grad_norm": 0.17805609107017517, + "learning_rate": 1.4915770401021801e-05, + "loss": 0.0168, + "step": 76020 + }, + { + "epoch": 4.9813273930419975, + "grad_norm": 0.3934766352176666, + "learning_rate": 1.4903991550417185e-05, + "loss": 0.0185, + "step": 76030 + }, + { + "epoch": 4.981982572233505, + "grad_norm": 1.8528608083724976, + "learning_rate": 1.4892216537859277e-05, + "loss": 0.0421, + "step": 76040 + }, + { + "epoch": 4.982637751425015, + "grad_norm": 0.5264830589294434, + "learning_rate": 1.4880445364635775e-05, + "loss": 0.0222, + "step": 76050 + }, + { + "epoch": 4.983292930616524, + "grad_norm": 0.6323913335800171, + "learning_rate": 1.4868678032033973e-05, + "loss": 0.025, + "step": 76060 + }, + { + "epoch": 4.983948109808033, + "grad_norm": 0.09321557730436325, + "learning_rate": 1.4856914541340739e-05, + "loss": 0.0291, + "step": 76070 + }, + { + "epoch": 4.9846032889995415, + "grad_norm": 0.14431096613407135, + "learning_rate": 1.4845154893842472e-05, + "loss": 0.0193, + "step": 76080 + }, + { + "epoch": 4.98525846819105, + "grad_norm": 0.46126019954681396, + "learning_rate": 1.4833399090825234e-05, + "loss": 0.0304, + "step": 76090 + }, + { + "epoch": 4.985913647382559, + "grad_norm": 0.12482506781816483, + "learning_rate": 1.4821647133574562e-05, + "loss": 0.0253, + "step": 76100 + }, + { + "epoch": 4.986568826574068, + "grad_norm": 0.5702005624771118, + "learning_rate": 1.4809899023375706e-05, + "loss": 0.0303, + "step": 76110 + }, + { + "epoch": 4.987224005765577, + "grad_norm": 0.5308130383491516, + "learning_rate": 1.4798154761513366e-05, + "loss": 0.0204, + "step": 76120 + }, + { + "epoch": 4.987879184957086, + "grad_norm": 0.5980331301689148, + "learning_rate": 1.4786414349271916e-05, + "loss": 0.0291, + "step": 76130 + }, + { + "epoch": 4.988534364148594, + "grad_norm": 0.33170953392982483, + "learning_rate": 1.4774677787935226e-05, + "loss": 0.0471, + "step": 76140 + }, + { + "epoch": 4.989189543340103, + "grad_norm": 0.5678163766860962, + "learning_rate": 1.4762945078786838e-05, + "loss": 0.0266, + "step": 76150 + }, + { + "epoch": 4.989844722531612, + "grad_norm": 0.21428562700748444, + "learning_rate": 1.4751216223109777e-05, + "loss": 0.0282, + "step": 76160 + }, + { + "epoch": 4.990499901723121, + "grad_norm": 0.07201489806175232, + "learning_rate": 1.4739491222186714e-05, + "loss": 0.041, + "step": 76170 + }, + { + "epoch": 4.991155080914631, + "grad_norm": 0.5095187425613403, + "learning_rate": 1.4727770077299879e-05, + "loss": 0.0217, + "step": 76180 + }, + { + "epoch": 4.991810260106139, + "grad_norm": 0.4766782224178314, + "learning_rate": 1.4716052789731094e-05, + "loss": 0.0367, + "step": 76190 + }, + { + "epoch": 4.992465439297648, + "grad_norm": 0.21589769423007965, + "learning_rate": 1.470433936076171e-05, + "loss": 0.0369, + "step": 76200 + }, + { + "epoch": 4.993120618489157, + "grad_norm": 0.4663321077823639, + "learning_rate": 1.4692629791672719e-05, + "loss": 0.0284, + "step": 76210 + }, + { + "epoch": 4.993775797680666, + "grad_norm": 0.11619540303945541, + "learning_rate": 1.468092408374464e-05, + "loss": 0.0281, + "step": 76220 + }, + { + "epoch": 4.994430976872175, + "grad_norm": 0.1810365468263626, + "learning_rate": 1.4669222238257602e-05, + "loss": 0.0191, + "step": 76230 + }, + { + "epoch": 4.9950861560636834, + "grad_norm": 0.48193275928497314, + "learning_rate": 1.4657524256491301e-05, + "loss": 0.0412, + "step": 76240 + }, + { + "epoch": 4.995741335255192, + "grad_norm": 0.5716767907142639, + "learning_rate": 1.464583013972502e-05, + "loss": 0.0334, + "step": 76250 + }, + { + "epoch": 4.996396514446701, + "grad_norm": 0.28045928478240967, + "learning_rate": 1.4634139889237591e-05, + "loss": 0.031, + "step": 76260 + }, + { + "epoch": 4.99705169363821, + "grad_norm": 0.2571180760860443, + "learning_rate": 1.4622453506307448e-05, + "loss": 0.0248, + "step": 76270 + }, + { + "epoch": 4.997706872829719, + "grad_norm": 0.7449125051498413, + "learning_rate": 1.461077099221262e-05, + "loss": 0.0202, + "step": 76280 + }, + { + "epoch": 4.9983620520212275, + "grad_norm": 0.44058382511138916, + "learning_rate": 1.4599092348230652e-05, + "loss": 0.0248, + "step": 76290 + }, + { + "epoch": 4.999017231212736, + "grad_norm": 0.12094277888536453, + "learning_rate": 1.4587417575638724e-05, + "loss": 0.0114, + "step": 76300 + }, + { + "epoch": 4.999672410404246, + "grad_norm": 0.6526023745536804, + "learning_rate": 1.4575746675713563e-05, + "loss": 0.042, + "step": 76310 + }, + { + "epoch": 5.000327589595755, + "grad_norm": 0.20503103733062744, + "learning_rate": 1.4564079649731504e-05, + "loss": 0.0257, + "step": 76320 + }, + { + "epoch": 5.000982768787264, + "grad_norm": 0.921501636505127, + "learning_rate": 1.4552416498968402e-05, + "loss": 0.049, + "step": 76330 + }, + { + "epoch": 5.0016379479787725, + "grad_norm": 0.2319442331790924, + "learning_rate": 1.4540757224699747e-05, + "loss": 0.0371, + "step": 76340 + }, + { + "epoch": 5.002293127170281, + "grad_norm": 0.57429438829422, + "learning_rate": 1.452910182820056e-05, + "loss": 0.0317, + "step": 76350 + }, + { + "epoch": 5.00294830636179, + "grad_norm": 0.3816492259502411, + "learning_rate": 1.4517450310745468e-05, + "loss": 0.0241, + "step": 76360 + }, + { + "epoch": 5.003603485553299, + "grad_norm": 1.1696763038635254, + "learning_rate": 1.450580267360866e-05, + "loss": 0.0644, + "step": 76370 + }, + { + "epoch": 5.004258664744808, + "grad_norm": 0.4493890404701233, + "learning_rate": 1.4494158918063922e-05, + "loss": 0.029, + "step": 76380 + }, + { + "epoch": 5.0049138439363166, + "grad_norm": 0.45868587493896484, + "learning_rate": 1.4482519045384568e-05, + "loss": 0.02, + "step": 76390 + }, + { + "epoch": 5.005569023127825, + "grad_norm": 0.35337376594543457, + "learning_rate": 1.4470883056843548e-05, + "loss": 0.0237, + "step": 76400 + }, + { + "epoch": 5.006224202319334, + "grad_norm": 0.24546602368354797, + "learning_rate": 1.4459250953713322e-05, + "loss": 0.0404, + "step": 76410 + }, + { + "epoch": 5.006879381510843, + "grad_norm": 0.7259764671325684, + "learning_rate": 1.444762273726598e-05, + "loss": 0.035, + "step": 76420 + }, + { + "epoch": 5.007534560702352, + "grad_norm": 0.9831817150115967, + "learning_rate": 1.443599840877316e-05, + "loss": 0.035, + "step": 76430 + }, + { + "epoch": 5.008189739893861, + "grad_norm": 0.044037435203790665, + "learning_rate": 1.4424377969506103e-05, + "loss": 0.021, + "step": 76440 + }, + { + "epoch": 5.008844919085369, + "grad_norm": 0.4431988298892975, + "learning_rate": 1.4412761420735565e-05, + "loss": 0.03, + "step": 76450 + }, + { + "epoch": 5.009500098276879, + "grad_norm": 0.29002928733825684, + "learning_rate": 1.4401148763731953e-05, + "loss": 0.0189, + "step": 76460 + }, + { + "epoch": 5.010155277468388, + "grad_norm": 2.4912045001983643, + "learning_rate": 1.4389539999765173e-05, + "loss": 0.0414, + "step": 76470 + }, + { + "epoch": 5.010810456659897, + "grad_norm": 0.34019187092781067, + "learning_rate": 1.4377935130104758e-05, + "loss": 0.0231, + "step": 76480 + }, + { + "epoch": 5.011465635851406, + "grad_norm": 0.05982404202222824, + "learning_rate": 1.4366334156019801e-05, + "loss": 0.0483, + "step": 76490 + }, + { + "epoch": 5.012120815042914, + "grad_norm": 0.04959273710846901, + "learning_rate": 1.435473707877898e-05, + "loss": 0.0128, + "step": 76500 + }, + { + "epoch": 5.012775994234423, + "grad_norm": 0.254666268825531, + "learning_rate": 1.4343143899650508e-05, + "loss": 0.0531, + "step": 76510 + }, + { + "epoch": 5.013431173425932, + "grad_norm": 0.07907389104366302, + "learning_rate": 1.4331554619902204e-05, + "loss": 0.0366, + "step": 76520 + }, + { + "epoch": 5.014086352617441, + "grad_norm": 0.41710537672042847, + "learning_rate": 1.4319969240801484e-05, + "loss": 0.0212, + "step": 76530 + }, + { + "epoch": 5.01474153180895, + "grad_norm": 1.0130877494812012, + "learning_rate": 1.430838776361526e-05, + "loss": 0.029, + "step": 76540 + }, + { + "epoch": 5.0153967110004585, + "grad_norm": 0.4794992208480835, + "learning_rate": 1.4296810189610093e-05, + "loss": 0.0298, + "step": 76550 + }, + { + "epoch": 5.016051890191967, + "grad_norm": 0.07665941119194031, + "learning_rate": 1.4285236520052087e-05, + "loss": 0.0187, + "step": 76560 + }, + { + "epoch": 5.016707069383476, + "grad_norm": 1.066619873046875, + "learning_rate": 1.4273666756206933e-05, + "loss": 0.0209, + "step": 76570 + }, + { + "epoch": 5.017362248574985, + "grad_norm": 0.4314310550689697, + "learning_rate": 1.4262100899339848e-05, + "loss": 0.0155, + "step": 76580 + }, + { + "epoch": 5.018017427766494, + "grad_norm": 0.12917602062225342, + "learning_rate": 1.4250538950715696e-05, + "loss": 0.0373, + "step": 76590 + }, + { + "epoch": 5.018672606958003, + "grad_norm": 0.39977654814720154, + "learning_rate": 1.4238980911598842e-05, + "loss": 0.0212, + "step": 76600 + }, + { + "epoch": 5.019327786149512, + "grad_norm": 0.39600521326065063, + "learning_rate": 1.4227426783253267e-05, + "loss": 0.0302, + "step": 76610 + }, + { + "epoch": 5.019982965341021, + "grad_norm": 0.27024099230766296, + "learning_rate": 1.4215876566942521e-05, + "loss": 0.0466, + "step": 76620 + }, + { + "epoch": 5.02063814453253, + "grad_norm": 0.5979170203208923, + "learning_rate": 1.4204330263929721e-05, + "loss": 0.0317, + "step": 76630 + }, + { + "epoch": 5.021293323724039, + "grad_norm": 0.8275326490402222, + "learning_rate": 1.4192787875477537e-05, + "loss": 0.0272, + "step": 76640 + }, + { + "epoch": 5.0219485029155475, + "grad_norm": 0.22552178800106049, + "learning_rate": 1.4181249402848246e-05, + "loss": 0.0147, + "step": 76650 + }, + { + "epoch": 5.022603682107056, + "grad_norm": 0.15604589879512787, + "learning_rate": 1.4169714847303656e-05, + "loss": 0.0165, + "step": 76660 + }, + { + "epoch": 5.023258861298565, + "grad_norm": 0.1982191652059555, + "learning_rate": 1.4158184210105197e-05, + "loss": 0.0657, + "step": 76670 + }, + { + "epoch": 5.023914040490074, + "grad_norm": 0.8249514102935791, + "learning_rate": 1.4146657492513787e-05, + "loss": 0.0229, + "step": 76680 + }, + { + "epoch": 5.024569219681583, + "grad_norm": 0.9112637042999268, + "learning_rate": 1.4135134695790047e-05, + "loss": 0.0511, + "step": 76690 + }, + { + "epoch": 5.025224398873092, + "grad_norm": 1.0375980138778687, + "learning_rate": 1.4123615821194037e-05, + "loss": 0.0371, + "step": 76700 + }, + { + "epoch": 5.0258795780646, + "grad_norm": 0.5062288045883179, + "learning_rate": 1.4112100869985473e-05, + "loss": 0.0266, + "step": 76710 + }, + { + "epoch": 5.026534757256109, + "grad_norm": 0.19421665370464325, + "learning_rate": 1.4100589843423589e-05, + "loss": 0.0207, + "step": 76720 + }, + { + "epoch": 5.027189936447618, + "grad_norm": 0.2457108199596405, + "learning_rate": 1.408908274276723e-05, + "loss": 0.0257, + "step": 76730 + }, + { + "epoch": 5.027845115639128, + "grad_norm": 0.4787766933441162, + "learning_rate": 1.4077579569274786e-05, + "loss": 0.0771, + "step": 76740 + }, + { + "epoch": 5.0285002948306365, + "grad_norm": 0.2406752109527588, + "learning_rate": 1.4066080324204245e-05, + "loss": 0.0347, + "step": 76750 + }, + { + "epoch": 5.029155474022145, + "grad_norm": 1.1782634258270264, + "learning_rate": 1.405458500881312e-05, + "loss": 0.0462, + "step": 76760 + }, + { + "epoch": 5.029810653213654, + "grad_norm": 0.2356191873550415, + "learning_rate": 1.4043093624358538e-05, + "loss": 0.0248, + "step": 76770 + }, + { + "epoch": 5.030465832405163, + "grad_norm": 0.21616023778915405, + "learning_rate": 1.4031606172097189e-05, + "loss": 0.0178, + "step": 76780 + }, + { + "epoch": 5.031121011596672, + "grad_norm": 0.12496771663427353, + "learning_rate": 1.40201226532853e-05, + "loss": 0.0396, + "step": 76790 + }, + { + "epoch": 5.031776190788181, + "grad_norm": 0.3208090662956238, + "learning_rate": 1.4008643069178701e-05, + "loss": 0.0257, + "step": 76800 + }, + { + "epoch": 5.032431369979689, + "grad_norm": 0.3280695378780365, + "learning_rate": 1.399716742103278e-05, + "loss": 0.0339, + "step": 76810 + }, + { + "epoch": 5.033086549171198, + "grad_norm": 0.10632093250751495, + "learning_rate": 1.398569571010252e-05, + "loss": 0.0225, + "step": 76820 + }, + { + "epoch": 5.033741728362707, + "grad_norm": 0.27538010478019714, + "learning_rate": 1.3974227937642414e-05, + "loss": 0.0134, + "step": 76830 + }, + { + "epoch": 5.034396907554216, + "grad_norm": 0.5946484804153442, + "learning_rate": 1.3962764104906595e-05, + "loss": 0.0214, + "step": 76840 + }, + { + "epoch": 5.035052086745725, + "grad_norm": 0.0588311031460762, + "learning_rate": 1.3951304213148696e-05, + "loss": 0.0198, + "step": 76850 + }, + { + "epoch": 5.0357072659372335, + "grad_norm": 0.31748929619789124, + "learning_rate": 1.3939848263621979e-05, + "loss": 0.0172, + "step": 76860 + }, + { + "epoch": 5.036362445128742, + "grad_norm": 0.2851807177066803, + "learning_rate": 1.3928396257579219e-05, + "loss": 0.0336, + "step": 76870 + }, + { + "epoch": 5.037017624320252, + "grad_norm": 0.2286231368780136, + "learning_rate": 1.3916948196272833e-05, + "loss": 0.0396, + "step": 76880 + }, + { + "epoch": 5.037672803511761, + "grad_norm": 0.23141297698020935, + "learning_rate": 1.3905504080954734e-05, + "loss": 0.0266, + "step": 76890 + }, + { + "epoch": 5.03832798270327, + "grad_norm": 0.820406973361969, + "learning_rate": 1.3894063912876454e-05, + "loss": 0.0305, + "step": 76900 + }, + { + "epoch": 5.0389831618947785, + "grad_norm": 0.6918674111366272, + "learning_rate": 1.3882627693289046e-05, + "loss": 0.0308, + "step": 76910 + }, + { + "epoch": 5.039638341086287, + "grad_norm": 0.380700021982193, + "learning_rate": 1.3871195423443183e-05, + "loss": 0.0455, + "step": 76920 + }, + { + "epoch": 5.040293520277796, + "grad_norm": 0.5316529870033264, + "learning_rate": 1.3859767104589044e-05, + "loss": 0.0263, + "step": 76930 + }, + { + "epoch": 5.040948699469305, + "grad_norm": 0.31055355072021484, + "learning_rate": 1.3848342737976467e-05, + "loss": 0.0279, + "step": 76940 + }, + { + "epoch": 5.041603878660814, + "grad_norm": 0.08347731828689575, + "learning_rate": 1.3836922324854757e-05, + "loss": 0.0387, + "step": 76950 + }, + { + "epoch": 5.0422590578523225, + "grad_norm": 0.22860510647296906, + "learning_rate": 1.382550586647287e-05, + "loss": 0.0305, + "step": 76960 + }, + { + "epoch": 5.042914237043831, + "grad_norm": 0.15383721888065338, + "learning_rate": 1.3814093364079257e-05, + "loss": 0.0238, + "step": 76970 + }, + { + "epoch": 5.04356941623534, + "grad_norm": 1.0193833112716675, + "learning_rate": 1.3802684818921991e-05, + "loss": 0.0368, + "step": 76980 + }, + { + "epoch": 5.044224595426849, + "grad_norm": 0.5565757751464844, + "learning_rate": 1.3791280232248687e-05, + "loss": 0.0246, + "step": 76990 + }, + { + "epoch": 5.044879774618358, + "grad_norm": 0.35189664363861084, + "learning_rate": 1.3779879605306555e-05, + "loss": 0.0269, + "step": 77000 + }, + { + "epoch": 5.045534953809867, + "grad_norm": 0.2157721072435379, + "learning_rate": 1.376848293934232e-05, + "loss": 0.0169, + "step": 77010 + }, + { + "epoch": 5.046190133001376, + "grad_norm": 0.3612820506095886, + "learning_rate": 1.3757090235602316e-05, + "loss": 0.0324, + "step": 77020 + }, + { + "epoch": 5.046845312192885, + "grad_norm": 0.5431307554244995, + "learning_rate": 1.3745701495332447e-05, + "loss": 0.0323, + "step": 77030 + }, + { + "epoch": 5.047500491384394, + "grad_norm": 0.24092085659503937, + "learning_rate": 1.3734316719778145e-05, + "loss": 0.0361, + "step": 77040 + }, + { + "epoch": 5.048155670575903, + "grad_norm": 0.31228989362716675, + "learning_rate": 1.3722935910184454e-05, + "loss": 0.021, + "step": 77050 + }, + { + "epoch": 5.048810849767412, + "grad_norm": 0.43830329179763794, + "learning_rate": 1.3711559067795921e-05, + "loss": 0.0209, + "step": 77060 + }, + { + "epoch": 5.04946602895892, + "grad_norm": 0.47020575404167175, + "learning_rate": 1.3700186193856756e-05, + "loss": 0.0226, + "step": 77070 + }, + { + "epoch": 5.050121208150429, + "grad_norm": 0.1779983639717102, + "learning_rate": 1.368881728961064e-05, + "loss": 0.0182, + "step": 77080 + }, + { + "epoch": 5.050776387341938, + "grad_norm": 0.16966445744037628, + "learning_rate": 1.3677452356300885e-05, + "loss": 0.0304, + "step": 77090 + }, + { + "epoch": 5.051431566533447, + "grad_norm": 0.3416852653026581, + "learning_rate": 1.3666091395170321e-05, + "loss": 0.0306, + "step": 77100 + }, + { + "epoch": 5.052086745724956, + "grad_norm": 0.16920290887355804, + "learning_rate": 1.3654734407461384e-05, + "loss": 0.0335, + "step": 77110 + }, + { + "epoch": 5.0527419249164645, + "grad_norm": 0.18616604804992676, + "learning_rate": 1.3643381394416022e-05, + "loss": 0.0187, + "step": 77120 + }, + { + "epoch": 5.053397104107973, + "grad_norm": 0.09811052680015564, + "learning_rate": 1.3632032357275838e-05, + "loss": 0.0156, + "step": 77130 + }, + { + "epoch": 5.054052283299482, + "grad_norm": 0.11785566061735153, + "learning_rate": 1.3620687297281898e-05, + "loss": 0.0493, + "step": 77140 + }, + { + "epoch": 5.054707462490991, + "grad_norm": 0.2874045968055725, + "learning_rate": 1.3609346215674917e-05, + "loss": 0.0291, + "step": 77150 + }, + { + "epoch": 5.055362641682501, + "grad_norm": 1.135517954826355, + "learning_rate": 1.3598009113695109e-05, + "loss": 0.0603, + "step": 77160 + }, + { + "epoch": 5.056017820874009, + "grad_norm": 0.607120156288147, + "learning_rate": 1.358667599258231e-05, + "loss": 0.0463, + "step": 77170 + }, + { + "epoch": 5.056673000065518, + "grad_norm": 0.8712220191955566, + "learning_rate": 1.357534685357586e-05, + "loss": 0.0204, + "step": 77180 + }, + { + "epoch": 5.057328179257027, + "grad_norm": 1.1827903985977173, + "learning_rate": 1.3564021697914715e-05, + "loss": 0.0282, + "step": 77190 + }, + { + "epoch": 5.057983358448536, + "grad_norm": 0.6442124247550964, + "learning_rate": 1.3552700526837381e-05, + "loss": 0.056, + "step": 77200 + }, + { + "epoch": 5.058638537640045, + "grad_norm": 0.1291481852531433, + "learning_rate": 1.3541383341581925e-05, + "loss": 0.0427, + "step": 77210 + }, + { + "epoch": 5.0592937168315535, + "grad_norm": 0.29742828011512756, + "learning_rate": 1.3530070143385965e-05, + "loss": 0.0417, + "step": 77220 + }, + { + "epoch": 5.059948896023062, + "grad_norm": 0.36521655321121216, + "learning_rate": 1.3518760933486702e-05, + "loss": 0.0256, + "step": 77230 + }, + { + "epoch": 5.060604075214571, + "grad_norm": 0.9082655906677246, + "learning_rate": 1.3507455713120908e-05, + "loss": 0.0351, + "step": 77240 + }, + { + "epoch": 5.06125925440608, + "grad_norm": 0.7500807642936707, + "learning_rate": 1.3496154483524876e-05, + "loss": 0.0403, + "step": 77250 + }, + { + "epoch": 5.061914433597589, + "grad_norm": 0.24326881766319275, + "learning_rate": 1.3484857245934506e-05, + "loss": 0.0373, + "step": 77260 + }, + { + "epoch": 5.062569612789098, + "grad_norm": 0.3382115066051483, + "learning_rate": 1.3473564001585248e-05, + "loss": 0.0366, + "step": 77270 + }, + { + "epoch": 5.063224791980606, + "grad_norm": 0.21003849804401398, + "learning_rate": 1.3462274751712129e-05, + "loss": 0.031, + "step": 77280 + }, + { + "epoch": 5.063879971172115, + "grad_norm": 0.8222597241401672, + "learning_rate": 1.3450989497549698e-05, + "loss": 0.0329, + "step": 77290 + }, + { + "epoch": 5.064535150363625, + "grad_norm": 0.6505746841430664, + "learning_rate": 1.3439708240332115e-05, + "loss": 0.0344, + "step": 77300 + }, + { + "epoch": 5.065190329555134, + "grad_norm": 1.3168659210205078, + "learning_rate": 1.3428430981293043e-05, + "loss": 0.0869, + "step": 77310 + }, + { + "epoch": 5.0658455087466425, + "grad_norm": 0.401726633310318, + "learning_rate": 1.3417157721665808e-05, + "loss": 0.0538, + "step": 77320 + }, + { + "epoch": 5.066500687938151, + "grad_norm": 1.5507452487945557, + "learning_rate": 1.3405888462683186e-05, + "loss": 0.0332, + "step": 77330 + }, + { + "epoch": 5.06715586712966, + "grad_norm": 0.22751444578170776, + "learning_rate": 1.3394623205577595e-05, + "loss": 0.0281, + "step": 77340 + }, + { + "epoch": 5.067811046321169, + "grad_norm": 1.048150897026062, + "learning_rate": 1.3383361951580963e-05, + "loss": 0.085, + "step": 77350 + }, + { + "epoch": 5.068466225512678, + "grad_norm": 0.8138198256492615, + "learning_rate": 1.337210470192483e-05, + "loss": 0.0609, + "step": 77360 + }, + { + "epoch": 5.069121404704187, + "grad_norm": 0.3801536560058594, + "learning_rate": 1.3360851457840245e-05, + "loss": 0.0383, + "step": 77370 + }, + { + "epoch": 5.069776583895695, + "grad_norm": 1.0997684001922607, + "learning_rate": 1.3349602220557856e-05, + "loss": 0.0225, + "step": 77380 + }, + { + "epoch": 5.070431763087204, + "grad_norm": 1.3652974367141724, + "learning_rate": 1.3338356991307865e-05, + "loss": 0.062, + "step": 77390 + }, + { + "epoch": 5.071086942278713, + "grad_norm": 0.2801996171474457, + "learning_rate": 1.3327115771320048e-05, + "loss": 0.0178, + "step": 77400 + }, + { + "epoch": 5.071742121470222, + "grad_norm": 1.9171854257583618, + "learning_rate": 1.3315878561823697e-05, + "loss": 0.0157, + "step": 77410 + }, + { + "epoch": 5.072397300661731, + "grad_norm": 0.28130173683166504, + "learning_rate": 1.330464536404773e-05, + "loss": 0.0219, + "step": 77420 + }, + { + "epoch": 5.0730524798532395, + "grad_norm": 1.110909342765808, + "learning_rate": 1.3293416179220563e-05, + "loss": 0.0311, + "step": 77430 + }, + { + "epoch": 5.073707659044748, + "grad_norm": 0.33803266286849976, + "learning_rate": 1.328219100857021e-05, + "loss": 0.0172, + "step": 77440 + }, + { + "epoch": 5.074362838236258, + "grad_norm": 0.8637272119522095, + "learning_rate": 1.327096985332425e-05, + "loss": 0.0441, + "step": 77450 + }, + { + "epoch": 5.075018017427767, + "grad_norm": 0.9832139015197754, + "learning_rate": 1.3259752714709816e-05, + "loss": 0.023, + "step": 77460 + }, + { + "epoch": 5.075673196619276, + "grad_norm": 0.4098227620124817, + "learning_rate": 1.3248539593953569e-05, + "loss": 0.0229, + "step": 77470 + }, + { + "epoch": 5.076328375810784, + "grad_norm": 0.21260346472263336, + "learning_rate": 1.323733049228178e-05, + "loss": 0.0187, + "step": 77480 + }, + { + "epoch": 5.076983555002293, + "grad_norm": 0.2704125642776489, + "learning_rate": 1.3226125410920276e-05, + "loss": 0.0289, + "step": 77490 + }, + { + "epoch": 5.077638734193802, + "grad_norm": 0.3570564389228821, + "learning_rate": 1.3214924351094388e-05, + "loss": 0.0165, + "step": 77500 + }, + { + "epoch": 5.078293913385311, + "grad_norm": 0.4050026535987854, + "learning_rate": 1.3203727314029069e-05, + "loss": 0.0649, + "step": 77510 + }, + { + "epoch": 5.07894909257682, + "grad_norm": 0.5248226523399353, + "learning_rate": 1.3192534300948811e-05, + "loss": 0.0429, + "step": 77520 + }, + { + "epoch": 5.0796042717683285, + "grad_norm": 0.124229796230793, + "learning_rate": 1.3181345313077675e-05, + "loss": 0.0294, + "step": 77530 + }, + { + "epoch": 5.080259450959837, + "grad_norm": 0.8237001895904541, + "learning_rate": 1.3170160351639243e-05, + "loss": 0.0218, + "step": 77540 + }, + { + "epoch": 5.080914630151346, + "grad_norm": 0.23489601910114288, + "learning_rate": 1.3158979417856721e-05, + "loss": 0.0621, + "step": 77550 + }, + { + "epoch": 5.081569809342855, + "grad_norm": 0.595007598400116, + "learning_rate": 1.31478025129528e-05, + "loss": 0.0212, + "step": 77560 + }, + { + "epoch": 5.082224988534364, + "grad_norm": 0.4101603627204895, + "learning_rate": 1.3136629638149794e-05, + "loss": 0.0197, + "step": 77570 + }, + { + "epoch": 5.0828801677258735, + "grad_norm": 0.7338809370994568, + "learning_rate": 1.3125460794669547e-05, + "loss": 0.0185, + "step": 77580 + }, + { + "epoch": 5.083535346917382, + "grad_norm": 0.9891288876533508, + "learning_rate": 1.3114295983733483e-05, + "loss": 0.0385, + "step": 77590 + }, + { + "epoch": 5.084190526108891, + "grad_norm": 0.27937260270118713, + "learning_rate": 1.3103135206562534e-05, + "loss": 0.0233, + "step": 77600 + }, + { + "epoch": 5.0848457053004, + "grad_norm": 0.04863738641142845, + "learning_rate": 1.3091978464377264e-05, + "loss": 0.0138, + "step": 77610 + }, + { + "epoch": 5.085500884491909, + "grad_norm": 0.2236553281545639, + "learning_rate": 1.3080825758397724e-05, + "loss": 0.0197, + "step": 77620 + }, + { + "epoch": 5.0861560636834175, + "grad_norm": 0.29690682888031006, + "learning_rate": 1.3069677089843568e-05, + "loss": 0.0188, + "step": 77630 + }, + { + "epoch": 5.086811242874926, + "grad_norm": 0.18272320926189423, + "learning_rate": 1.305853245993401e-05, + "loss": 0.0235, + "step": 77640 + }, + { + "epoch": 5.087466422066435, + "grad_norm": 0.6514066457748413, + "learning_rate": 1.3047391869887809e-05, + "loss": 0.0315, + "step": 77650 + }, + { + "epoch": 5.088121601257944, + "grad_norm": 0.1993669718503952, + "learning_rate": 1.3036255320923263e-05, + "loss": 0.0111, + "step": 77660 + }, + { + "epoch": 5.088776780449453, + "grad_norm": 0.06486841291189194, + "learning_rate": 1.3025122814258273e-05, + "loss": 0.0212, + "step": 77670 + }, + { + "epoch": 5.089431959640962, + "grad_norm": 0.5346472263336182, + "learning_rate": 1.3013994351110249e-05, + "loss": 0.023, + "step": 77680 + }, + { + "epoch": 5.09008713883247, + "grad_norm": 0.5316944122314453, + "learning_rate": 1.3002869932696194e-05, + "loss": 0.0308, + "step": 77690 + }, + { + "epoch": 5.090742318023979, + "grad_norm": 0.25432273745536804, + "learning_rate": 1.2991749560232664e-05, + "loss": 0.0238, + "step": 77700 + }, + { + "epoch": 5.091397497215488, + "grad_norm": 0.2923281788825989, + "learning_rate": 1.2980633234935768e-05, + "loss": 0.0466, + "step": 77710 + }, + { + "epoch": 5.092052676406997, + "grad_norm": 0.7623829245567322, + "learning_rate": 1.2969520958021153e-05, + "loss": 0.0809, + "step": 77720 + }, + { + "epoch": 5.092707855598507, + "grad_norm": 0.2551504373550415, + "learning_rate": 1.295841273070405e-05, + "loss": 0.015, + "step": 77730 + }, + { + "epoch": 5.093363034790015, + "grad_norm": 0.2226567268371582, + "learning_rate": 1.2947308554199255e-05, + "loss": 0.0157, + "step": 77740 + }, + { + "epoch": 5.094018213981524, + "grad_norm": 0.5355130434036255, + "learning_rate": 1.2936208429721075e-05, + "loss": 0.0366, + "step": 77750 + }, + { + "epoch": 5.094673393173033, + "grad_norm": 0.13378779590129852, + "learning_rate": 1.2925112358483416e-05, + "loss": 0.0259, + "step": 77760 + }, + { + "epoch": 5.095328572364542, + "grad_norm": 0.13285169005393982, + "learning_rate": 1.2914020341699728e-05, + "loss": 0.034, + "step": 77770 + }, + { + "epoch": 5.095983751556051, + "grad_norm": 0.44592586159706116, + "learning_rate": 1.2902932380583032e-05, + "loss": 0.0182, + "step": 77780 + }, + { + "epoch": 5.0966389307475595, + "grad_norm": 0.14673751592636108, + "learning_rate": 1.2891848476345863e-05, + "loss": 0.0429, + "step": 77790 + }, + { + "epoch": 5.097294109939068, + "grad_norm": 0.13717974722385406, + "learning_rate": 1.2880768630200368e-05, + "loss": 0.0244, + "step": 77800 + }, + { + "epoch": 5.097949289130577, + "grad_norm": 0.12713854014873505, + "learning_rate": 1.286969284335819e-05, + "loss": 0.0234, + "step": 77810 + }, + { + "epoch": 5.098604468322086, + "grad_norm": 0.8441339135169983, + "learning_rate": 1.2858621117030578e-05, + "loss": 0.0456, + "step": 77820 + }, + { + "epoch": 5.099259647513595, + "grad_norm": 0.31729844212532043, + "learning_rate": 1.2847553452428324e-05, + "loss": 0.0125, + "step": 77830 + }, + { + "epoch": 5.0999148267051035, + "grad_norm": 0.5886583924293518, + "learning_rate": 1.283648985076178e-05, + "loss": 0.0245, + "step": 77840 + }, + { + "epoch": 5.100570005896612, + "grad_norm": 0.47667521238327026, + "learning_rate": 1.2825430313240817e-05, + "loss": 0.0321, + "step": 77850 + }, + { + "epoch": 5.101225185088121, + "grad_norm": 0.5465129613876343, + "learning_rate": 1.281437484107491e-05, + "loss": 0.0293, + "step": 77860 + }, + { + "epoch": 5.101880364279631, + "grad_norm": 0.14028531312942505, + "learning_rate": 1.2803323435473058e-05, + "loss": 0.0304, + "step": 77870 + }, + { + "epoch": 5.10253554347114, + "grad_norm": 0.27004289627075195, + "learning_rate": 1.2792276097643841e-05, + "loss": 0.0437, + "step": 77880 + }, + { + "epoch": 5.1031907226626485, + "grad_norm": 0.2578103244304657, + "learning_rate": 1.2781232828795342e-05, + "loss": 0.0248, + "step": 77890 + }, + { + "epoch": 5.103845901854157, + "grad_norm": 0.09515736997127533, + "learning_rate": 1.2770193630135292e-05, + "loss": 0.0196, + "step": 77900 + }, + { + "epoch": 5.104501081045666, + "grad_norm": 0.4247450828552246, + "learning_rate": 1.2759158502870877e-05, + "loss": 0.0387, + "step": 77910 + }, + { + "epoch": 5.105156260237175, + "grad_norm": 2.078338146209717, + "learning_rate": 1.2748127448208913e-05, + "loss": 0.0376, + "step": 77920 + }, + { + "epoch": 5.105811439428684, + "grad_norm": 0.09417929500341415, + "learning_rate": 1.2737100467355706e-05, + "loss": 0.0272, + "step": 77930 + }, + { + "epoch": 5.106466618620193, + "grad_norm": 0.2317447066307068, + "learning_rate": 1.2726077561517169e-05, + "loss": 0.0549, + "step": 77940 + }, + { + "epoch": 5.107121797811701, + "grad_norm": 0.2722310423851013, + "learning_rate": 1.2715058731898755e-05, + "loss": 0.0236, + "step": 77950 + }, + { + "epoch": 5.10777697700321, + "grad_norm": 0.27208852767944336, + "learning_rate": 1.2704043979705471e-05, + "loss": 0.0644, + "step": 77960 + }, + { + "epoch": 5.108432156194719, + "grad_norm": 0.2770887017250061, + "learning_rate": 1.269303330614185e-05, + "loss": 0.0232, + "step": 77970 + }, + { + "epoch": 5.109087335386228, + "grad_norm": 0.6748905181884766, + "learning_rate": 1.2682026712412015e-05, + "loss": 0.022, + "step": 77980 + }, + { + "epoch": 5.109742514577737, + "grad_norm": 0.8295252323150635, + "learning_rate": 1.2671024199719644e-05, + "loss": 0.0374, + "step": 77990 + }, + { + "epoch": 5.1103976937692455, + "grad_norm": 0.48939886689186096, + "learning_rate": 1.2660025769267936e-05, + "loss": 0.0286, + "step": 78000 + }, + { + "epoch": 5.111052872960755, + "grad_norm": 0.5052146911621094, + "learning_rate": 1.2649031422259667e-05, + "loss": 0.0175, + "step": 78010 + }, + { + "epoch": 5.111708052152264, + "grad_norm": 0.2863771617412567, + "learning_rate": 1.2638041159897163e-05, + "loss": 0.0366, + "step": 78020 + }, + { + "epoch": 5.112363231343773, + "grad_norm": 0.24909088015556335, + "learning_rate": 1.262705498338232e-05, + "loss": 0.0397, + "step": 78030 + }, + { + "epoch": 5.113018410535282, + "grad_norm": 0.6910372376441956, + "learning_rate": 1.2616072893916542e-05, + "loss": 0.0308, + "step": 78040 + }, + { + "epoch": 5.11367358972679, + "grad_norm": 0.6528547406196594, + "learning_rate": 1.2605094892700836e-05, + "loss": 0.021, + "step": 78050 + }, + { + "epoch": 5.114328768918299, + "grad_norm": 0.7137391567230225, + "learning_rate": 1.259412098093572e-05, + "loss": 0.0183, + "step": 78060 + }, + { + "epoch": 5.114983948109808, + "grad_norm": 0.5304438471794128, + "learning_rate": 1.258315115982131e-05, + "loss": 0.0232, + "step": 78070 + }, + { + "epoch": 5.115639127301317, + "grad_norm": 0.3106706738471985, + "learning_rate": 1.2572185430557203e-05, + "loss": 0.033, + "step": 78080 + }, + { + "epoch": 5.116294306492826, + "grad_norm": 0.09650702774524689, + "learning_rate": 1.2561223794342659e-05, + "loss": 0.0374, + "step": 78090 + }, + { + "epoch": 5.1169494856843345, + "grad_norm": 0.7944766283035278, + "learning_rate": 1.255026625237637e-05, + "loss": 0.0254, + "step": 78100 + }, + { + "epoch": 5.117604664875843, + "grad_norm": 0.48163726925849915, + "learning_rate": 1.2539312805856678e-05, + "loss": 0.0251, + "step": 78110 + }, + { + "epoch": 5.118259844067352, + "grad_norm": 0.35974356532096863, + "learning_rate": 1.25283634559814e-05, + "loss": 0.0318, + "step": 78120 + }, + { + "epoch": 5.118915023258861, + "grad_norm": 0.3427758514881134, + "learning_rate": 1.2517418203947972e-05, + "loss": 0.0348, + "step": 78130 + }, + { + "epoch": 5.11957020245037, + "grad_norm": 1.5844461917877197, + "learning_rate": 1.2506477050953308e-05, + "loss": 0.0305, + "step": 78140 + }, + { + "epoch": 5.1202253816418795, + "grad_norm": 0.29665470123291016, + "learning_rate": 1.2495539998193972e-05, + "loss": 0.0281, + "step": 78150 + }, + { + "epoch": 5.120880560833388, + "grad_norm": 0.10954322665929794, + "learning_rate": 1.2484607046865975e-05, + "loss": 0.0362, + "step": 78160 + }, + { + "epoch": 5.121535740024897, + "grad_norm": 0.8569085001945496, + "learning_rate": 1.2473678198164968e-05, + "loss": 0.0299, + "step": 78170 + }, + { + "epoch": 5.122190919216406, + "grad_norm": 0.26856887340545654, + "learning_rate": 1.2462753453286075e-05, + "loss": 0.0323, + "step": 78180 + }, + { + "epoch": 5.122846098407915, + "grad_norm": 0.21493178606033325, + "learning_rate": 1.2451832813424031e-05, + "loss": 0.0552, + "step": 78190 + }, + { + "epoch": 5.1235012775994235, + "grad_norm": 0.3906223177909851, + "learning_rate": 1.2440916279773095e-05, + "loss": 0.0199, + "step": 78200 + }, + { + "epoch": 5.124156456790932, + "grad_norm": 0.6884954571723938, + "learning_rate": 1.2430003853527101e-05, + "loss": 0.0468, + "step": 78210 + }, + { + "epoch": 5.124811635982441, + "grad_norm": 0.29568034410476685, + "learning_rate": 1.2419095535879383e-05, + "loss": 0.0263, + "step": 78220 + }, + { + "epoch": 5.12546681517395, + "grad_norm": 0.8595708012580872, + "learning_rate": 1.2408191328022878e-05, + "loss": 0.0214, + "step": 78230 + }, + { + "epoch": 5.126121994365459, + "grad_norm": 0.12613441050052643, + "learning_rate": 1.2397291231150066e-05, + "loss": 0.0495, + "step": 78240 + }, + { + "epoch": 5.126777173556968, + "grad_norm": 0.8587427139282227, + "learning_rate": 1.2386395246452937e-05, + "loss": 0.0251, + "step": 78250 + }, + { + "epoch": 5.127432352748476, + "grad_norm": 0.7359315156936646, + "learning_rate": 1.2375503375123083e-05, + "loss": 0.0536, + "step": 78260 + }, + { + "epoch": 5.128087531939985, + "grad_norm": 0.18624812364578247, + "learning_rate": 1.2364615618351583e-05, + "loss": 0.0202, + "step": 78270 + }, + { + "epoch": 5.128742711131494, + "grad_norm": 0.15061932802200317, + "learning_rate": 1.2353731977329169e-05, + "loss": 0.0264, + "step": 78280 + }, + { + "epoch": 5.129397890323004, + "grad_norm": 0.2897019386291504, + "learning_rate": 1.2342852453246007e-05, + "loss": 0.0312, + "step": 78290 + }, + { + "epoch": 5.130053069514513, + "grad_norm": 0.4219270944595337, + "learning_rate": 1.2331977047291898e-05, + "loss": 0.0444, + "step": 78300 + }, + { + "epoch": 5.130708248706021, + "grad_norm": 0.3304869830608368, + "learning_rate": 1.2321105760656137e-05, + "loss": 0.0253, + "step": 78310 + }, + { + "epoch": 5.13136342789753, + "grad_norm": 0.17860952019691467, + "learning_rate": 1.2310238594527613e-05, + "loss": 0.0221, + "step": 78320 + }, + { + "epoch": 5.132018607089039, + "grad_norm": 0.46096271276474, + "learning_rate": 1.2299375550094693e-05, + "loss": 0.0232, + "step": 78330 + }, + { + "epoch": 5.132673786280548, + "grad_norm": 0.5545713305473328, + "learning_rate": 1.2288516628545421e-05, + "loss": 0.0182, + "step": 78340 + }, + { + "epoch": 5.133328965472057, + "grad_norm": 0.17253239452838898, + "learning_rate": 1.2277661831067255e-05, + "loss": 0.0369, + "step": 78350 + }, + { + "epoch": 5.1339841446635655, + "grad_norm": 0.6921305060386658, + "learning_rate": 1.2266811158847286e-05, + "loss": 0.0358, + "step": 78360 + }, + { + "epoch": 5.134639323855074, + "grad_norm": 1.640457034111023, + "learning_rate": 1.2255964613072107e-05, + "loss": 0.0319, + "step": 78370 + }, + { + "epoch": 5.135294503046583, + "grad_norm": 0.22985711693763733, + "learning_rate": 1.2245122194927905e-05, + "loss": 0.0349, + "step": 78380 + }, + { + "epoch": 5.135949682238092, + "grad_norm": 0.19718365371227264, + "learning_rate": 1.2234283905600364e-05, + "loss": 0.0285, + "step": 78390 + }, + { + "epoch": 5.136604861429601, + "grad_norm": 0.8423188328742981, + "learning_rate": 1.222344974627475e-05, + "loss": 0.0482, + "step": 78400 + }, + { + "epoch": 5.1372600406211095, + "grad_norm": 0.8172035813331604, + "learning_rate": 1.2212619718135875e-05, + "loss": 0.0299, + "step": 78410 + }, + { + "epoch": 5.137915219812618, + "grad_norm": 2.7989094257354736, + "learning_rate": 1.2201793822368118e-05, + "loss": 0.0377, + "step": 78420 + }, + { + "epoch": 5.138570399004128, + "grad_norm": 0.2586400508880615, + "learning_rate": 1.2190972060155337e-05, + "loss": 0.0168, + "step": 78430 + }, + { + "epoch": 5.139225578195637, + "grad_norm": 0.46678096055984497, + "learning_rate": 1.2180154432681013e-05, + "loss": 0.0207, + "step": 78440 + }, + { + "epoch": 5.139880757387146, + "grad_norm": 0.43934980034828186, + "learning_rate": 1.2169340941128148e-05, + "loss": 0.0183, + "step": 78450 + }, + { + "epoch": 5.1405359365786545, + "grad_norm": 0.6761490106582642, + "learning_rate": 1.2158531586679267e-05, + "loss": 0.051, + "step": 78460 + }, + { + "epoch": 5.141191115770163, + "grad_norm": 0.5603345632553101, + "learning_rate": 1.2147726370516477e-05, + "loss": 0.0205, + "step": 78470 + }, + { + "epoch": 5.141846294961672, + "grad_norm": 0.40018728375434875, + "learning_rate": 1.2136925293821422e-05, + "loss": 0.0381, + "step": 78480 + }, + { + "epoch": 5.142501474153181, + "grad_norm": 0.6813912987709045, + "learning_rate": 1.2126128357775306e-05, + "loss": 0.0656, + "step": 78490 + }, + { + "epoch": 5.14315665334469, + "grad_norm": 0.6390243768692017, + "learning_rate": 1.2115335563558838e-05, + "loss": 0.031, + "step": 78500 + }, + { + "epoch": 5.143811832536199, + "grad_norm": 0.24828897416591644, + "learning_rate": 1.2104546912352327e-05, + "loss": 0.0167, + "step": 78510 + }, + { + "epoch": 5.144467011727707, + "grad_norm": 0.1725008189678192, + "learning_rate": 1.2093762405335557e-05, + "loss": 0.0219, + "step": 78520 + }, + { + "epoch": 5.145122190919216, + "grad_norm": 0.40327829122543335, + "learning_rate": 1.2082982043687974e-05, + "loss": 0.0352, + "step": 78530 + }, + { + "epoch": 5.145777370110725, + "grad_norm": 0.27604803442955017, + "learning_rate": 1.207220582858845e-05, + "loss": 0.0226, + "step": 78540 + }, + { + "epoch": 5.146432549302234, + "grad_norm": 0.953325629234314, + "learning_rate": 1.206143376121549e-05, + "loss": 0.0377, + "step": 78550 + }, + { + "epoch": 5.147087728493743, + "grad_norm": 0.22194065153598785, + "learning_rate": 1.2050665842747078e-05, + "loss": 0.0193, + "step": 78560 + }, + { + "epoch": 5.147742907685252, + "grad_norm": 1.1634529829025269, + "learning_rate": 1.2039902074360804e-05, + "loss": 0.0457, + "step": 78570 + }, + { + "epoch": 5.148398086876761, + "grad_norm": 0.9384582042694092, + "learning_rate": 1.2029142457233755e-05, + "loss": 0.0322, + "step": 78580 + }, + { + "epoch": 5.14905326606827, + "grad_norm": 0.13813284039497375, + "learning_rate": 1.2018386992542601e-05, + "loss": 0.0365, + "step": 78590 + }, + { + "epoch": 5.149708445259779, + "grad_norm": 0.24894604086875916, + "learning_rate": 1.2007635681463542e-05, + "loss": 0.0253, + "step": 78600 + }, + { + "epoch": 5.150363624451288, + "grad_norm": 0.5952584147453308, + "learning_rate": 1.1996888525172334e-05, + "loss": 0.0179, + "step": 78610 + }, + { + "epoch": 5.151018803642796, + "grad_norm": 0.5424257516860962, + "learning_rate": 1.198614552484425e-05, + "loss": 0.0259, + "step": 78620 + }, + { + "epoch": 5.151673982834305, + "grad_norm": 1.780663013458252, + "learning_rate": 1.197540668165415e-05, + "loss": 0.0344, + "step": 78630 + }, + { + "epoch": 5.152329162025814, + "grad_norm": 0.39464181661605835, + "learning_rate": 1.1964671996776395e-05, + "loss": 0.0258, + "step": 78640 + }, + { + "epoch": 5.152984341217323, + "grad_norm": 0.09751530736684799, + "learning_rate": 1.1953941471384922e-05, + "loss": 0.0141, + "step": 78650 + }, + { + "epoch": 5.153639520408832, + "grad_norm": 0.8995035886764526, + "learning_rate": 1.1943215106653205e-05, + "loss": 0.0428, + "step": 78660 + }, + { + "epoch": 5.1542946996003405, + "grad_norm": 0.4772951006889343, + "learning_rate": 1.193249290375429e-05, + "loss": 0.0343, + "step": 78670 + }, + { + "epoch": 5.154949878791849, + "grad_norm": 0.4113606810569763, + "learning_rate": 1.1921774863860696e-05, + "loss": 0.0244, + "step": 78680 + }, + { + "epoch": 5.155605057983358, + "grad_norm": 0.6321113109588623, + "learning_rate": 1.1911060988144556e-05, + "loss": 0.0355, + "step": 78690 + }, + { + "epoch": 5.156260237174867, + "grad_norm": 0.3779745399951935, + "learning_rate": 1.1900351277777533e-05, + "loss": 0.0338, + "step": 78700 + }, + { + "epoch": 5.156915416366377, + "grad_norm": 0.3093684911727905, + "learning_rate": 1.1889645733930798e-05, + "loss": 0.028, + "step": 78710 + }, + { + "epoch": 5.157570595557885, + "grad_norm": 0.30934998393058777, + "learning_rate": 1.1878944357775107e-05, + "loss": 0.021, + "step": 78720 + }, + { + "epoch": 5.158225774749394, + "grad_norm": 0.8088259100914001, + "learning_rate": 1.1868247150480749e-05, + "loss": 0.0622, + "step": 78730 + }, + { + "epoch": 5.158880953940903, + "grad_norm": 0.1957671046257019, + "learning_rate": 1.1857554113217567e-05, + "loss": 0.0188, + "step": 78740 + }, + { + "epoch": 5.159536133132412, + "grad_norm": 0.10916859656572342, + "learning_rate": 1.1846865247154903e-05, + "loss": 0.0583, + "step": 78750 + }, + { + "epoch": 5.160191312323921, + "grad_norm": 0.15940192341804504, + "learning_rate": 1.1836180553461711e-05, + "loss": 0.0215, + "step": 78760 + }, + { + "epoch": 5.1608464915154295, + "grad_norm": 0.2689131796360016, + "learning_rate": 1.1825500033306419e-05, + "loss": 0.0178, + "step": 78770 + }, + { + "epoch": 5.161501670706938, + "grad_norm": 0.40876540541648865, + "learning_rate": 1.1814823687857052e-05, + "loss": 0.038, + "step": 78780 + }, + { + "epoch": 5.162156849898447, + "grad_norm": 0.7485693693161011, + "learning_rate": 1.1804151518281153e-05, + "loss": 0.0271, + "step": 78790 + }, + { + "epoch": 5.162812029089956, + "grad_norm": 0.8059098124504089, + "learning_rate": 1.179348352574583e-05, + "loss": 0.036, + "step": 78800 + }, + { + "epoch": 5.163467208281465, + "grad_norm": 0.5104584097862244, + "learning_rate": 1.1782819711417697e-05, + "loss": 0.0225, + "step": 78810 + }, + { + "epoch": 5.164122387472974, + "grad_norm": 0.42157837748527527, + "learning_rate": 1.1772160076462951e-05, + "loss": 0.0441, + "step": 78820 + }, + { + "epoch": 5.164777566664482, + "grad_norm": 0.6171795725822449, + "learning_rate": 1.1761504622047293e-05, + "loss": 0.045, + "step": 78830 + }, + { + "epoch": 5.165432745855991, + "grad_norm": 0.22942790389060974, + "learning_rate": 1.1750853349335999e-05, + "loss": 0.0201, + "step": 78840 + }, + { + "epoch": 5.166087925047501, + "grad_norm": 0.5314213633537292, + "learning_rate": 1.1740206259493874e-05, + "loss": 0.0278, + "step": 78850 + }, + { + "epoch": 5.16674310423901, + "grad_norm": 0.9281693696975708, + "learning_rate": 1.1729563353685286e-05, + "loss": 0.0253, + "step": 78860 + }, + { + "epoch": 5.1673982834305185, + "grad_norm": 0.7634236216545105, + "learning_rate": 1.1718924633074091e-05, + "loss": 0.0347, + "step": 78870 + }, + { + "epoch": 5.168053462622027, + "grad_norm": 0.7567496299743652, + "learning_rate": 1.1708290098823765e-05, + "loss": 0.0441, + "step": 78880 + }, + { + "epoch": 5.168708641813536, + "grad_norm": 1.5551718473434448, + "learning_rate": 1.1697659752097245e-05, + "loss": 0.0238, + "step": 78890 + }, + { + "epoch": 5.169363821005045, + "grad_norm": 0.26762357354164124, + "learning_rate": 1.1687033594057068e-05, + "loss": 0.0354, + "step": 78900 + }, + { + "epoch": 5.170019000196554, + "grad_norm": 0.6749274134635925, + "learning_rate": 1.1676411625865297e-05, + "loss": 0.0255, + "step": 78910 + }, + { + "epoch": 5.170674179388063, + "grad_norm": 0.21289505064487457, + "learning_rate": 1.1665793848683542e-05, + "loss": 0.0324, + "step": 78920 + }, + { + "epoch": 5.171329358579571, + "grad_norm": 0.3019137382507324, + "learning_rate": 1.1655180263672927e-05, + "loss": 0.0315, + "step": 78930 + }, + { + "epoch": 5.17198453777108, + "grad_norm": 0.7292054295539856, + "learning_rate": 1.1644570871994142e-05, + "loss": 0.0278, + "step": 78940 + }, + { + "epoch": 5.172639716962589, + "grad_norm": 0.13557176291942596, + "learning_rate": 1.163396567480744e-05, + "loss": 0.0305, + "step": 78950 + }, + { + "epoch": 5.173294896154098, + "grad_norm": 0.2092612087726593, + "learning_rate": 1.1623364673272552e-05, + "loss": 0.0219, + "step": 78960 + }, + { + "epoch": 5.173950075345607, + "grad_norm": 0.541296660900116, + "learning_rate": 1.1612767868548807e-05, + "loss": 0.0154, + "step": 78970 + }, + { + "epoch": 5.1746052545371155, + "grad_norm": 0.2612437903881073, + "learning_rate": 1.1602175261795057e-05, + "loss": 0.0177, + "step": 78980 + }, + { + "epoch": 5.175260433728625, + "grad_norm": 0.2168593555688858, + "learning_rate": 1.1591586854169696e-05, + "loss": 0.0273, + "step": 78990 + }, + { + "epoch": 5.175915612920134, + "grad_norm": 0.05699237436056137, + "learning_rate": 1.158100264683064e-05, + "loss": 0.0275, + "step": 79000 + }, + { + "epoch": 5.176570792111643, + "grad_norm": 0.14464807510375977, + "learning_rate": 1.1570422640935386e-05, + "loss": 0.0313, + "step": 79010 + }, + { + "epoch": 5.177225971303152, + "grad_norm": 0.36577361822128296, + "learning_rate": 1.155984683764092e-05, + "loss": 0.0404, + "step": 79020 + }, + { + "epoch": 5.1778811504946605, + "grad_norm": 0.7764346599578857, + "learning_rate": 1.1549275238103812e-05, + "loss": 0.0541, + "step": 79030 + }, + { + "epoch": 5.178536329686169, + "grad_norm": 1.1296895742416382, + "learning_rate": 1.1538707843480152e-05, + "loss": 0.0322, + "step": 79040 + }, + { + "epoch": 5.179191508877678, + "grad_norm": 0.12305381894111633, + "learning_rate": 1.1528144654925593e-05, + "loss": 0.0169, + "step": 79050 + }, + { + "epoch": 5.179846688069187, + "grad_norm": 0.7390983700752258, + "learning_rate": 1.1517585673595277e-05, + "loss": 0.0318, + "step": 79060 + }, + { + "epoch": 5.180501867260696, + "grad_norm": 0.5584273934364319, + "learning_rate": 1.150703090064395e-05, + "loss": 0.0245, + "step": 79070 + }, + { + "epoch": 5.1811570464522045, + "grad_norm": 0.48015743494033813, + "learning_rate": 1.1496480337225835e-05, + "loss": 0.0202, + "step": 79080 + }, + { + "epoch": 5.181812225643713, + "grad_norm": 0.7129520773887634, + "learning_rate": 1.1485933984494751e-05, + "loss": 0.0457, + "step": 79090 + }, + { + "epoch": 5.182467404835222, + "grad_norm": 0.6582501530647278, + "learning_rate": 1.1475391843604e-05, + "loss": 0.0421, + "step": 79100 + }, + { + "epoch": 5.183122584026731, + "grad_norm": 0.17156095802783966, + "learning_rate": 1.1464853915706503e-05, + "loss": 0.018, + "step": 79110 + }, + { + "epoch": 5.18377776321824, + "grad_norm": 0.5480250716209412, + "learning_rate": 1.1454320201954626e-05, + "loss": 0.0256, + "step": 79120 + }, + { + "epoch": 5.1844329424097495, + "grad_norm": 0.354195773601532, + "learning_rate": 1.1443790703500356e-05, + "loss": 0.0353, + "step": 79130 + }, + { + "epoch": 5.185088121601258, + "grad_norm": 0.1934044361114502, + "learning_rate": 1.1433265421495154e-05, + "loss": 0.0252, + "step": 79140 + }, + { + "epoch": 5.185743300792767, + "grad_norm": 0.30314359068870544, + "learning_rate": 1.1422744357090065e-05, + "loss": 0.0413, + "step": 79150 + }, + { + "epoch": 5.186398479984276, + "grad_norm": 0.1564217507839203, + "learning_rate": 1.1412227511435647e-05, + "loss": 0.0225, + "step": 79160 + }, + { + "epoch": 5.187053659175785, + "grad_norm": 1.1892893314361572, + "learning_rate": 1.1401714885682025e-05, + "loss": 0.0423, + "step": 79170 + }, + { + "epoch": 5.187708838367294, + "grad_norm": 0.25325411558151245, + "learning_rate": 1.1391206480978823e-05, + "loss": 0.0258, + "step": 79180 + }, + { + "epoch": 5.188364017558802, + "grad_norm": 0.14366169273853302, + "learning_rate": 1.1380702298475227e-05, + "loss": 0.0215, + "step": 79190 + }, + { + "epoch": 5.189019196750311, + "grad_norm": 0.7274800539016724, + "learning_rate": 1.1370202339319985e-05, + "loss": 0.0416, + "step": 79200 + }, + { + "epoch": 5.18967437594182, + "grad_norm": 0.44280773401260376, + "learning_rate": 1.1359706604661319e-05, + "loss": 0.0484, + "step": 79210 + }, + { + "epoch": 5.190329555133329, + "grad_norm": 0.4326978921890259, + "learning_rate": 1.1349215095647042e-05, + "loss": 0.0377, + "step": 79220 + }, + { + "epoch": 5.190984734324838, + "grad_norm": 0.8571692109107971, + "learning_rate": 1.1338727813424493e-05, + "loss": 0.0446, + "step": 79230 + }, + { + "epoch": 5.1916399135163465, + "grad_norm": 0.37033790349960327, + "learning_rate": 1.132824475914056e-05, + "loss": 0.0246, + "step": 79240 + }, + { + "epoch": 5.192295092707855, + "grad_norm": 0.36456769704818726, + "learning_rate": 1.131776593394162e-05, + "loss": 0.0275, + "step": 79250 + }, + { + "epoch": 5.192950271899364, + "grad_norm": 0.29667165875434875, + "learning_rate": 1.1307291338973657e-05, + "loss": 0.0426, + "step": 79260 + }, + { + "epoch": 5.193605451090873, + "grad_norm": 0.5136242508888245, + "learning_rate": 1.1296820975382121e-05, + "loss": 0.0367, + "step": 79270 + }, + { + "epoch": 5.194260630282383, + "grad_norm": 0.15170477330684662, + "learning_rate": 1.1286354844312074e-05, + "loss": 0.0242, + "step": 79280 + }, + { + "epoch": 5.194915809473891, + "grad_norm": 0.354681134223938, + "learning_rate": 1.1275892946908023e-05, + "loss": 0.0251, + "step": 79290 + }, + { + "epoch": 5.1955709886654, + "grad_norm": 0.4772651791572571, + "learning_rate": 1.126543528431413e-05, + "loss": 0.0212, + "step": 79300 + }, + { + "epoch": 5.196226167856909, + "grad_norm": 2.3928134441375732, + "learning_rate": 1.125498185767398e-05, + "loss": 0.0265, + "step": 79310 + }, + { + "epoch": 5.196881347048418, + "grad_norm": 0.3057129681110382, + "learning_rate": 1.1244532668130781e-05, + "loss": 0.0168, + "step": 79320 + }, + { + "epoch": 5.197536526239927, + "grad_norm": 0.7369741201400757, + "learning_rate": 1.1234087716827203e-05, + "loss": 0.0307, + "step": 79330 + }, + { + "epoch": 5.1981917054314355, + "grad_norm": 0.3235260546207428, + "learning_rate": 1.1223647004905524e-05, + "loss": 0.0363, + "step": 79340 + }, + { + "epoch": 5.198846884622944, + "grad_norm": 0.40644437074661255, + "learning_rate": 1.1213210533507485e-05, + "loss": 0.0224, + "step": 79350 + }, + { + "epoch": 5.199502063814453, + "grad_norm": 0.4091549813747406, + "learning_rate": 1.1202778303774447e-05, + "loss": 0.0263, + "step": 79360 + }, + { + "epoch": 5.200157243005962, + "grad_norm": 0.420879989862442, + "learning_rate": 1.1192350316847234e-05, + "loss": 0.0245, + "step": 79370 + }, + { + "epoch": 5.200812422197471, + "grad_norm": 0.9123544096946716, + "learning_rate": 1.1181926573866258e-05, + "loss": 0.0385, + "step": 79380 + }, + { + "epoch": 5.20146760138898, + "grad_norm": 0.6927511096000671, + "learning_rate": 1.1171507075971416e-05, + "loss": 0.0475, + "step": 79390 + }, + { + "epoch": 5.202122780580488, + "grad_norm": 0.507431149482727, + "learning_rate": 1.1161091824302177e-05, + "loss": 0.0349, + "step": 79400 + }, + { + "epoch": 5.202777959771998, + "grad_norm": 0.21619580686092377, + "learning_rate": 1.115068081999755e-05, + "loss": 0.0253, + "step": 79410 + }, + { + "epoch": 5.203433138963507, + "grad_norm": 4.235117435455322, + "learning_rate": 1.1140274064196071e-05, + "loss": 0.0282, + "step": 79420 + }, + { + "epoch": 5.204088318155016, + "grad_norm": 0.15452370047569275, + "learning_rate": 1.112987155803578e-05, + "loss": 0.0267, + "step": 79430 + }, + { + "epoch": 5.2047434973465245, + "grad_norm": 1.0738614797592163, + "learning_rate": 1.1119473302654298e-05, + "loss": 0.0739, + "step": 79440 + }, + { + "epoch": 5.205398676538033, + "grad_norm": 0.1533908098936081, + "learning_rate": 1.1109079299188769e-05, + "loss": 0.023, + "step": 79450 + }, + { + "epoch": 5.206053855729542, + "grad_norm": 0.49861589074134827, + "learning_rate": 1.1098689548775847e-05, + "loss": 0.0361, + "step": 79460 + }, + { + "epoch": 5.206709034921051, + "grad_norm": 0.4156016707420349, + "learning_rate": 1.1088304052551762e-05, + "loss": 0.0239, + "step": 79470 + }, + { + "epoch": 5.20736421411256, + "grad_norm": 0.31533217430114746, + "learning_rate": 1.1077922811652215e-05, + "loss": 0.0242, + "step": 79480 + }, + { + "epoch": 5.208019393304069, + "grad_norm": 1.2042255401611328, + "learning_rate": 1.1067545827212534e-05, + "loss": 0.0409, + "step": 79490 + }, + { + "epoch": 5.208674572495577, + "grad_norm": 0.550544023513794, + "learning_rate": 1.1057173100367496e-05, + "loss": 0.0501, + "step": 79500 + }, + { + "epoch": 5.209329751687086, + "grad_norm": 0.457980751991272, + "learning_rate": 1.1046804632251474e-05, + "loss": 0.0399, + "step": 79510 + }, + { + "epoch": 5.209984930878595, + "grad_norm": 0.5172178745269775, + "learning_rate": 1.1036440423998318e-05, + "loss": 0.0198, + "step": 79520 + }, + { + "epoch": 5.210640110070104, + "grad_norm": 0.5973979830741882, + "learning_rate": 1.1026080476741469e-05, + "loss": 0.0164, + "step": 79530 + }, + { + "epoch": 5.211295289261613, + "grad_norm": 1.136711835861206, + "learning_rate": 1.1015724791613836e-05, + "loss": 0.0687, + "step": 79540 + }, + { + "epoch": 5.2119504684531215, + "grad_norm": 0.12444864213466644, + "learning_rate": 1.1005373369747951e-05, + "loss": 0.0171, + "step": 79550 + }, + { + "epoch": 5.212605647644631, + "grad_norm": 0.24180784821510315, + "learning_rate": 1.09950262122758e-05, + "loss": 0.0238, + "step": 79560 + }, + { + "epoch": 5.21326082683614, + "grad_norm": 0.3051927983760834, + "learning_rate": 1.0984683320328948e-05, + "loss": 0.0285, + "step": 79570 + }, + { + "epoch": 5.213916006027649, + "grad_norm": 0.8463698029518127, + "learning_rate": 1.0974344695038458e-05, + "loss": 0.0276, + "step": 79580 + }, + { + "epoch": 5.214571185219158, + "grad_norm": 0.1721671223640442, + "learning_rate": 1.0964010337534968e-05, + "loss": 0.0297, + "step": 79590 + }, + { + "epoch": 5.2152263644106664, + "grad_norm": 0.1264435350894928, + "learning_rate": 1.0953680248948611e-05, + "loss": 0.0913, + "step": 79600 + }, + { + "epoch": 5.215881543602175, + "grad_norm": 0.7886635661125183, + "learning_rate": 1.0943354430409069e-05, + "loss": 0.0412, + "step": 79610 + }, + { + "epoch": 5.216536722793684, + "grad_norm": 1.175963044166565, + "learning_rate": 1.0933032883045574e-05, + "loss": 0.037, + "step": 79620 + }, + { + "epoch": 5.217191901985193, + "grad_norm": 0.861547589302063, + "learning_rate": 1.0922715607986872e-05, + "loss": 0.0305, + "step": 79630 + }, + { + "epoch": 5.217847081176702, + "grad_norm": 0.6628041863441467, + "learning_rate": 1.0912402606361234e-05, + "loss": 0.0189, + "step": 79640 + }, + { + "epoch": 5.2185022603682105, + "grad_norm": 0.7129495143890381, + "learning_rate": 1.090209387929647e-05, + "loss": 0.0263, + "step": 79650 + }, + { + "epoch": 5.219157439559719, + "grad_norm": 0.3584224283695221, + "learning_rate": 1.0891789427919957e-05, + "loss": 0.0191, + "step": 79660 + }, + { + "epoch": 5.219812618751228, + "grad_norm": 0.30039745569229126, + "learning_rate": 1.088148925335854e-05, + "loss": 0.026, + "step": 79670 + }, + { + "epoch": 5.220467797942737, + "grad_norm": 0.29913410544395447, + "learning_rate": 1.087119335673864e-05, + "loss": 0.0322, + "step": 79680 + }, + { + "epoch": 5.221122977134247, + "grad_norm": 0.3941560387611389, + "learning_rate": 1.0860901739186208e-05, + "loss": 0.0256, + "step": 79690 + }, + { + "epoch": 5.2217781563257555, + "grad_norm": 0.7447000741958618, + "learning_rate": 1.085061440182673e-05, + "loss": 0.0456, + "step": 79700 + }, + { + "epoch": 5.222433335517264, + "grad_norm": 0.796229362487793, + "learning_rate": 1.0840331345785193e-05, + "loss": 0.0286, + "step": 79710 + }, + { + "epoch": 5.223088514708773, + "grad_norm": 0.15628233551979065, + "learning_rate": 1.083005257218615e-05, + "loss": 0.0308, + "step": 79720 + }, + { + "epoch": 5.223743693900282, + "grad_norm": 0.4074920117855072, + "learning_rate": 1.0819778082153642e-05, + "loss": 0.0213, + "step": 79730 + }, + { + "epoch": 5.224398873091791, + "grad_norm": 1.9171100854873657, + "learning_rate": 1.0809507876811325e-05, + "loss": 0.0315, + "step": 79740 + }, + { + "epoch": 5.2250540522832996, + "grad_norm": 0.1371224820613861, + "learning_rate": 1.0799241957282292e-05, + "loss": 0.0222, + "step": 79750 + }, + { + "epoch": 5.225709231474808, + "grad_norm": 0.8351614475250244, + "learning_rate": 1.078898032468923e-05, + "loss": 0.031, + "step": 79760 + }, + { + "epoch": 5.226364410666317, + "grad_norm": 0.21190783381462097, + "learning_rate": 1.0778722980154315e-05, + "loss": 0.0137, + "step": 79770 + }, + { + "epoch": 5.227019589857826, + "grad_norm": 0.7807899117469788, + "learning_rate": 1.0768469924799302e-05, + "loss": 0.0226, + "step": 79780 + }, + { + "epoch": 5.227674769049335, + "grad_norm": 0.2811185121536255, + "learning_rate": 1.0758221159745419e-05, + "loss": 0.0406, + "step": 79790 + }, + { + "epoch": 5.228329948240844, + "grad_norm": 0.6211202144622803, + "learning_rate": 1.074797668611347e-05, + "loss": 0.02, + "step": 79800 + }, + { + "epoch": 5.228985127432352, + "grad_norm": 0.7979344129562378, + "learning_rate": 1.0737736505023777e-05, + "loss": 0.0304, + "step": 79810 + }, + { + "epoch": 5.229640306623861, + "grad_norm": 0.5540692210197449, + "learning_rate": 1.0727500617596203e-05, + "loss": 0.0266, + "step": 79820 + }, + { + "epoch": 5.23029548581537, + "grad_norm": 0.29484841227531433, + "learning_rate": 1.0717269024950104e-05, + "loss": 0.0262, + "step": 79830 + }, + { + "epoch": 5.23095066500688, + "grad_norm": 0.5131467580795288, + "learning_rate": 1.070704172820442e-05, + "loss": 0.0378, + "step": 79840 + }, + { + "epoch": 5.231605844198389, + "grad_norm": 0.16300620138645172, + "learning_rate": 1.0696818728477559e-05, + "loss": 0.0248, + "step": 79850 + }, + { + "epoch": 5.232261023389897, + "grad_norm": 1.1179102659225464, + "learning_rate": 1.0686600026887511e-05, + "loss": 0.0345, + "step": 79860 + }, + { + "epoch": 5.232916202581406, + "grad_norm": 0.18750424683094025, + "learning_rate": 1.0676385624551777e-05, + "loss": 0.0342, + "step": 79870 + }, + { + "epoch": 5.233571381772915, + "grad_norm": 0.13193953037261963, + "learning_rate": 1.0666175522587402e-05, + "loss": 0.0258, + "step": 79880 + }, + { + "epoch": 5.234226560964424, + "grad_norm": 0.7602090835571289, + "learning_rate": 1.0655969722110925e-05, + "loss": 0.0309, + "step": 79890 + }, + { + "epoch": 5.234881740155933, + "grad_norm": 0.5273078083992004, + "learning_rate": 1.0645768224238439e-05, + "loss": 0.0237, + "step": 79900 + }, + { + "epoch": 5.2355369193474415, + "grad_norm": 0.3811976909637451, + "learning_rate": 1.063557103008559e-05, + "loss": 0.0213, + "step": 79910 + }, + { + "epoch": 5.23619209853895, + "grad_norm": 0.8999372720718384, + "learning_rate": 1.0625378140767494e-05, + "loss": 0.0427, + "step": 79920 + }, + { + "epoch": 5.236847277730459, + "grad_norm": 0.7901005744934082, + "learning_rate": 1.0615189557398846e-05, + "loss": 0.0304, + "step": 79930 + }, + { + "epoch": 5.237502456921968, + "grad_norm": 0.14459243416786194, + "learning_rate": 1.060500528109386e-05, + "loss": 0.0258, + "step": 79940 + }, + { + "epoch": 5.238157636113477, + "grad_norm": 0.19945582747459412, + "learning_rate": 1.0594825312966272e-05, + "loss": 0.0321, + "step": 79950 + }, + { + "epoch": 5.2388128153049855, + "grad_norm": 0.14049673080444336, + "learning_rate": 1.0584649654129336e-05, + "loss": 0.0192, + "step": 79960 + }, + { + "epoch": 5.239467994496494, + "grad_norm": 0.8352777361869812, + "learning_rate": 1.0574478305695863e-05, + "loss": 0.0316, + "step": 79970 + }, + { + "epoch": 5.240123173688004, + "grad_norm": 0.2914498448371887, + "learning_rate": 1.0564311268778154e-05, + "loss": 0.0388, + "step": 79980 + }, + { + "epoch": 5.240778352879513, + "grad_norm": 0.10589442402124405, + "learning_rate": 1.0554148544488073e-05, + "loss": 0.0192, + "step": 79990 + }, + { + "epoch": 5.241433532071022, + "grad_norm": 0.3419724404811859, + "learning_rate": 1.0543990133937004e-05, + "loss": 0.054, + "step": 80000 + }, + { + "epoch": 5.2420887112625305, + "grad_norm": 0.10939750075340271, + "learning_rate": 1.0533836038235868e-05, + "loss": 0.0267, + "step": 80010 + }, + { + "epoch": 5.242743890454039, + "grad_norm": 0.1919105499982834, + "learning_rate": 1.052368625849507e-05, + "loss": 0.0229, + "step": 80020 + }, + { + "epoch": 5.243399069645548, + "grad_norm": 1.17485773563385, + "learning_rate": 1.05135407958246e-05, + "loss": 0.0281, + "step": 80030 + }, + { + "epoch": 5.244054248837057, + "grad_norm": 0.5258260369300842, + "learning_rate": 1.0503399651333934e-05, + "loss": 0.0196, + "step": 80040 + }, + { + "epoch": 5.244709428028566, + "grad_norm": 1.4188501834869385, + "learning_rate": 1.04932628261321e-05, + "loss": 0.028, + "step": 80050 + }, + { + "epoch": 5.245364607220075, + "grad_norm": 0.5436235666275024, + "learning_rate": 1.0483130321327644e-05, + "loss": 0.0206, + "step": 80060 + }, + { + "epoch": 5.246019786411583, + "grad_norm": 0.22889626026153564, + "learning_rate": 1.0473002138028654e-05, + "loss": 0.0183, + "step": 80070 + }, + { + "epoch": 5.246674965603092, + "grad_norm": 0.19026176631450653, + "learning_rate": 1.0462878277342713e-05, + "loss": 0.0355, + "step": 80080 + }, + { + "epoch": 5.247330144794601, + "grad_norm": 0.4545835554599762, + "learning_rate": 1.0452758740376972e-05, + "loss": 0.0225, + "step": 80090 + }, + { + "epoch": 5.24798532398611, + "grad_norm": 0.6991850137710571, + "learning_rate": 1.0442643528238067e-05, + "loss": 0.0413, + "step": 80100 + }, + { + "epoch": 5.248640503177619, + "grad_norm": 1.7240607738494873, + "learning_rate": 1.0432532642032188e-05, + "loss": 0.0231, + "step": 80110 + }, + { + "epoch": 5.249295682369128, + "grad_norm": 0.4302958548069, + "learning_rate": 1.042242608286505e-05, + "loss": 0.0179, + "step": 80120 + }, + { + "epoch": 5.249950861560637, + "grad_norm": 0.13797588646411896, + "learning_rate": 1.041232385184191e-05, + "loss": 0.0832, + "step": 80130 + }, + { + "epoch": 5.250606040752146, + "grad_norm": 0.21400602161884308, + "learning_rate": 1.0402225950067495e-05, + "loss": 0.0431, + "step": 80140 + }, + { + "epoch": 5.251261219943655, + "grad_norm": 1.0358623266220093, + "learning_rate": 1.0392132378646119e-05, + "loss": 0.0293, + "step": 80150 + }, + { + "epoch": 5.251916399135164, + "grad_norm": 0.6536313891410828, + "learning_rate": 1.0382043138681607e-05, + "loss": 0.0369, + "step": 80160 + }, + { + "epoch": 5.252571578326672, + "grad_norm": 0.3417986333370209, + "learning_rate": 1.0371958231277284e-05, + "loss": 0.0264, + "step": 80170 + }, + { + "epoch": 5.253226757518181, + "grad_norm": 0.5052886009216309, + "learning_rate": 1.0361877657536024e-05, + "loss": 0.0407, + "step": 80180 + }, + { + "epoch": 5.25388193670969, + "grad_norm": 0.5626409649848938, + "learning_rate": 1.035180141856023e-05, + "loss": 0.0387, + "step": 80190 + }, + { + "epoch": 5.254537115901199, + "grad_norm": 0.209828183054924, + "learning_rate": 1.0341729515451836e-05, + "loss": 0.0306, + "step": 80200 + }, + { + "epoch": 5.255192295092708, + "grad_norm": 0.7186260223388672, + "learning_rate": 1.0331661949312265e-05, + "loss": 0.0356, + "step": 80210 + }, + { + "epoch": 5.2558474742842165, + "grad_norm": 0.2485426515340805, + "learning_rate": 1.0321598721242514e-05, + "loss": 0.0294, + "step": 80220 + }, + { + "epoch": 5.256502653475725, + "grad_norm": 0.377454936504364, + "learning_rate": 1.0311539832343064e-05, + "loss": 0.0318, + "step": 80230 + }, + { + "epoch": 5.257157832667234, + "grad_norm": 0.14225418865680695, + "learning_rate": 1.0301485283713942e-05, + "loss": 0.0219, + "step": 80240 + }, + { + "epoch": 5.257813011858744, + "grad_norm": 0.08949122577905655, + "learning_rate": 1.0291435076454703e-05, + "loss": 0.021, + "step": 80250 + }, + { + "epoch": 5.258468191050253, + "grad_norm": 0.1678074598312378, + "learning_rate": 1.028138921166444e-05, + "loss": 0.0192, + "step": 80260 + }, + { + "epoch": 5.2591233702417615, + "grad_norm": 0.5485626459121704, + "learning_rate": 1.0271347690441719e-05, + "loss": 0.021, + "step": 80270 + }, + { + "epoch": 5.25977854943327, + "grad_norm": 0.8785804510116577, + "learning_rate": 1.0261310513884698e-05, + "loss": 0.0509, + "step": 80280 + }, + { + "epoch": 5.260433728624779, + "grad_norm": 0.16988301277160645, + "learning_rate": 1.0251277683090998e-05, + "loss": 0.0388, + "step": 80290 + }, + { + "epoch": 5.261088907816288, + "grad_norm": 0.4204905331134796, + "learning_rate": 1.0241249199157826e-05, + "loss": 0.0267, + "step": 80300 + }, + { + "epoch": 5.261744087007797, + "grad_norm": 0.44504767656326294, + "learning_rate": 1.0231225063181837e-05, + "loss": 0.024, + "step": 80310 + }, + { + "epoch": 5.2623992661993055, + "grad_norm": 0.40399712324142456, + "learning_rate": 1.0221205276259316e-05, + "loss": 0.0268, + "step": 80320 + }, + { + "epoch": 5.263054445390814, + "grad_norm": 0.16258499026298523, + "learning_rate": 1.0211189839485958e-05, + "loss": 0.0365, + "step": 80330 + }, + { + "epoch": 5.263709624582323, + "grad_norm": 0.592038094997406, + "learning_rate": 1.0201178753957075e-05, + "loss": 0.0196, + "step": 80340 + }, + { + "epoch": 5.264364803773832, + "grad_norm": 0.4281509816646576, + "learning_rate": 1.0191172020767432e-05, + "loss": 0.0286, + "step": 80350 + }, + { + "epoch": 5.265019982965341, + "grad_norm": 0.29743269085884094, + "learning_rate": 1.0181169641011362e-05, + "loss": 0.0406, + "step": 80360 + }, + { + "epoch": 5.26567516215685, + "grad_norm": 0.35249826312065125, + "learning_rate": 1.0171171615782721e-05, + "loss": 0.0151, + "step": 80370 + }, + { + "epoch": 5.266330341348358, + "grad_norm": 0.32544225454330444, + "learning_rate": 1.0161177946174866e-05, + "loss": 0.03, + "step": 80380 + }, + { + "epoch": 5.266985520539867, + "grad_norm": 0.6663718819618225, + "learning_rate": 1.0151188633280706e-05, + "loss": 0.0243, + "step": 80390 + }, + { + "epoch": 5.267640699731377, + "grad_norm": 0.3521246314048767, + "learning_rate": 1.0141203678192635e-05, + "loss": 0.0416, + "step": 80400 + }, + { + "epoch": 5.268295878922886, + "grad_norm": 0.6721802949905396, + "learning_rate": 1.0131223082002612e-05, + "loss": 0.0243, + "step": 80410 + }, + { + "epoch": 5.268951058114395, + "grad_norm": 0.8584481477737427, + "learning_rate": 1.012124684580208e-05, + "loss": 0.0326, + "step": 80420 + }, + { + "epoch": 5.269606237305903, + "grad_norm": 0.481913298368454, + "learning_rate": 1.0111274970682039e-05, + "loss": 0.0259, + "step": 80430 + }, + { + "epoch": 5.270261416497412, + "grad_norm": 0.8909264206886292, + "learning_rate": 1.010130745773299e-05, + "loss": 0.028, + "step": 80440 + }, + { + "epoch": 5.270916595688921, + "grad_norm": 0.35502052307128906, + "learning_rate": 1.0091344308044986e-05, + "loss": 0.0353, + "step": 80450 + }, + { + "epoch": 5.27157177488043, + "grad_norm": 0.3203582167625427, + "learning_rate": 1.0081385522707554e-05, + "loss": 0.0219, + "step": 80460 + }, + { + "epoch": 5.272226954071939, + "grad_norm": 0.4036918878555298, + "learning_rate": 1.0071431102809803e-05, + "loss": 0.0302, + "step": 80470 + }, + { + "epoch": 5.2728821332634475, + "grad_norm": 0.8423210978507996, + "learning_rate": 1.0061481049440297e-05, + "loss": 0.0339, + "step": 80480 + }, + { + "epoch": 5.273537312454956, + "grad_norm": 0.07153789699077606, + "learning_rate": 1.0051535363687187e-05, + "loss": 0.0168, + "step": 80490 + }, + { + "epoch": 5.274192491646465, + "grad_norm": 0.1330695003271103, + "learning_rate": 1.0041594046638087e-05, + "loss": 0.0374, + "step": 80500 + }, + { + "epoch": 5.274847670837974, + "grad_norm": 0.5550451278686523, + "learning_rate": 1.0031657099380204e-05, + "loss": 0.0375, + "step": 80510 + }, + { + "epoch": 5.275502850029483, + "grad_norm": 0.12371572852134705, + "learning_rate": 1.0021724523000204e-05, + "loss": 0.0135, + "step": 80520 + }, + { + "epoch": 5.2761580292209915, + "grad_norm": 0.14492429792881012, + "learning_rate": 1.001179631858431e-05, + "loss": 0.0368, + "step": 80530 + }, + { + "epoch": 5.276813208412501, + "grad_norm": 0.594606876373291, + "learning_rate": 1.0001872487218245e-05, + "loss": 0.0103, + "step": 80540 + }, + { + "epoch": 5.27746838760401, + "grad_norm": 0.91281658411026, + "learning_rate": 9.991953029987272e-06, + "loss": 0.022, + "step": 80550 + }, + { + "epoch": 5.278123566795519, + "grad_norm": 0.3408931791782379, + "learning_rate": 9.982037947976147e-06, + "loss": 0.0419, + "step": 80560 + }, + { + "epoch": 5.278778745987028, + "grad_norm": 0.9755566716194153, + "learning_rate": 9.972127242269208e-06, + "loss": 0.0477, + "step": 80570 + }, + { + "epoch": 5.2794339251785365, + "grad_norm": 0.684299647808075, + "learning_rate": 9.962220913950238e-06, + "loss": 0.0655, + "step": 80580 + }, + { + "epoch": 5.280089104370045, + "grad_norm": 0.4212726354598999, + "learning_rate": 9.952318964102591e-06, + "loss": 0.0201, + "step": 80590 + }, + { + "epoch": 5.280744283561554, + "grad_norm": 0.7804506421089172, + "learning_rate": 9.942421393809148e-06, + "loss": 0.026, + "step": 80600 + }, + { + "epoch": 5.281399462753063, + "grad_norm": 0.8412820100784302, + "learning_rate": 9.932528204152264e-06, + "loss": 0.0307, + "step": 80610 + }, + { + "epoch": 5.282054641944572, + "grad_norm": 0.6219547986984253, + "learning_rate": 9.922639396213856e-06, + "loss": 0.0376, + "step": 80620 + }, + { + "epoch": 5.282709821136081, + "grad_norm": 0.16182534396648407, + "learning_rate": 9.912754971075344e-06, + "loss": 0.025, + "step": 80630 + }, + { + "epoch": 5.283365000327589, + "grad_norm": 0.46245506405830383, + "learning_rate": 9.902874929817696e-06, + "loss": 0.022, + "step": 80640 + }, + { + "epoch": 5.284020179519098, + "grad_norm": 0.15263570845127106, + "learning_rate": 9.892999273521342e-06, + "loss": 0.018, + "step": 80650 + }, + { + "epoch": 5.284675358710607, + "grad_norm": 0.16309796273708344, + "learning_rate": 9.883128003266302e-06, + "loss": 0.0465, + "step": 80660 + }, + { + "epoch": 5.285330537902116, + "grad_norm": 0.335114061832428, + "learning_rate": 9.87326112013206e-06, + "loss": 0.033, + "step": 80670 + }, + { + "epoch": 5.2859857170936255, + "grad_norm": 0.8021431565284729, + "learning_rate": 9.863398625197662e-06, + "loss": 0.0364, + "step": 80680 + }, + { + "epoch": 5.286640896285134, + "grad_norm": 0.13537107408046722, + "learning_rate": 9.853540519541615e-06, + "loss": 0.0249, + "step": 80690 + }, + { + "epoch": 5.287296075476643, + "grad_norm": 0.3327806293964386, + "learning_rate": 9.843686804242053e-06, + "loss": 0.0391, + "step": 80700 + }, + { + "epoch": 5.287951254668152, + "grad_norm": 0.2095082849264145, + "learning_rate": 9.833837480376506e-06, + "loss": 0.0097, + "step": 80710 + }, + { + "epoch": 5.288606433859661, + "grad_norm": 4.193116664886475, + "learning_rate": 9.82399254902212e-06, + "loss": 0.0558, + "step": 80720 + }, + { + "epoch": 5.28926161305117, + "grad_norm": 0.32875120639801025, + "learning_rate": 9.814152011255494e-06, + "loss": 0.0161, + "step": 80730 + }, + { + "epoch": 5.289916792242678, + "grad_norm": 0.26633813977241516, + "learning_rate": 9.804315868152797e-06, + "loss": 0.0318, + "step": 80740 + }, + { + "epoch": 5.290571971434187, + "grad_norm": 0.22836466133594513, + "learning_rate": 9.79448412078966e-06, + "loss": 0.0188, + "step": 80750 + }, + { + "epoch": 5.291227150625696, + "grad_norm": 0.4515672028064728, + "learning_rate": 9.784656770241312e-06, + "loss": 0.0266, + "step": 80760 + }, + { + "epoch": 5.291882329817205, + "grad_norm": 0.7629488110542297, + "learning_rate": 9.77483381758243e-06, + "loss": 0.0252, + "step": 80770 + }, + { + "epoch": 5.292537509008714, + "grad_norm": 0.31731852889060974, + "learning_rate": 9.76501526388725e-06, + "loss": 0.0232, + "step": 80780 + }, + { + "epoch": 5.2931926882002225, + "grad_norm": 0.5569823980331421, + "learning_rate": 9.7552011102295e-06, + "loss": 0.03, + "step": 80790 + }, + { + "epoch": 5.293847867391731, + "grad_norm": 0.7356587648391724, + "learning_rate": 9.745391357682449e-06, + "loss": 0.0393, + "step": 80800 + }, + { + "epoch": 5.29450304658324, + "grad_norm": 0.34845995903015137, + "learning_rate": 9.735586007318875e-06, + "loss": 0.0252, + "step": 80810 + }, + { + "epoch": 5.295158225774749, + "grad_norm": 0.19529123604297638, + "learning_rate": 9.725785060211096e-06, + "loss": 0.0342, + "step": 80820 + }, + { + "epoch": 5.295813404966259, + "grad_norm": 0.7705422639846802, + "learning_rate": 9.715988517430896e-06, + "loss": 0.0327, + "step": 80830 + }, + { + "epoch": 5.2964685841577674, + "grad_norm": 0.5588611960411072, + "learning_rate": 9.706196380049625e-06, + "loss": 0.0234, + "step": 80840 + }, + { + "epoch": 5.297123763349276, + "grad_norm": 0.698694109916687, + "learning_rate": 9.69640864913815e-06, + "loss": 0.0474, + "step": 80850 + }, + { + "epoch": 5.297778942540785, + "grad_norm": 0.14868715405464172, + "learning_rate": 9.686625325766818e-06, + "loss": 0.0359, + "step": 80860 + }, + { + "epoch": 5.298434121732294, + "grad_norm": 0.11116308718919754, + "learning_rate": 9.67684641100554e-06, + "loss": 0.0106, + "step": 80870 + }, + { + "epoch": 5.299089300923803, + "grad_norm": 0.21169234812259674, + "learning_rate": 9.667071905923692e-06, + "loss": 0.0251, + "step": 80880 + }, + { + "epoch": 5.2997444801153115, + "grad_norm": 0.16971874237060547, + "learning_rate": 9.657301811590242e-06, + "loss": 0.0168, + "step": 80890 + }, + { + "epoch": 5.30039965930682, + "grad_norm": 0.26228052377700806, + "learning_rate": 9.647536129073603e-06, + "loss": 0.0239, + "step": 80900 + }, + { + "epoch": 5.301054838498329, + "grad_norm": 0.08567851036787033, + "learning_rate": 9.637774859441751e-06, + "loss": 0.0416, + "step": 80910 + }, + { + "epoch": 5.301710017689838, + "grad_norm": 1.0102177858352661, + "learning_rate": 9.628018003762152e-06, + "loss": 0.0611, + "step": 80920 + }, + { + "epoch": 5.302365196881347, + "grad_norm": 0.5599446296691895, + "learning_rate": 9.618265563101813e-06, + "loss": 0.0261, + "step": 80930 + }, + { + "epoch": 5.303020376072856, + "grad_norm": 1.058807611465454, + "learning_rate": 9.608517538527211e-06, + "loss": 0.0278, + "step": 80940 + }, + { + "epoch": 5.303675555264364, + "grad_norm": 0.5755307674407959, + "learning_rate": 9.598773931104431e-06, + "loss": 0.0425, + "step": 80950 + }, + { + "epoch": 5.304330734455874, + "grad_norm": 0.5827378034591675, + "learning_rate": 9.589034741898983e-06, + "loss": 0.042, + "step": 80960 + }, + { + "epoch": 5.304985913647383, + "grad_norm": 1.1382063627243042, + "learning_rate": 9.579299971975947e-06, + "loss": 0.0359, + "step": 80970 + }, + { + "epoch": 5.305641092838892, + "grad_norm": 1.388720989227295, + "learning_rate": 9.569569622399882e-06, + "loss": 0.0297, + "step": 80980 + }, + { + "epoch": 5.3062962720304006, + "grad_norm": 0.6391516923904419, + "learning_rate": 9.559843694234916e-06, + "loss": 0.0146, + "step": 80990 + }, + { + "epoch": 5.306951451221909, + "grad_norm": 0.24284440279006958, + "learning_rate": 9.550122188544625e-06, + "loss": 0.0352, + "step": 81000 + }, + { + "epoch": 5.307606630413418, + "grad_norm": 0.11145425587892532, + "learning_rate": 9.540405106392152e-06, + "loss": 0.0356, + "step": 81010 + }, + { + "epoch": 5.308261809604927, + "grad_norm": 0.13432705402374268, + "learning_rate": 9.53069244884015e-06, + "loss": 0.0166, + "step": 81020 + }, + { + "epoch": 5.308916988796436, + "grad_norm": 1.217342495918274, + "learning_rate": 9.52098421695079e-06, + "loss": 0.031, + "step": 81030 + }, + { + "epoch": 5.309572167987945, + "grad_norm": 0.7718533873558044, + "learning_rate": 9.511280411785728e-06, + "loss": 0.028, + "step": 81040 + }, + { + "epoch": 5.310227347179453, + "grad_norm": 0.2973255217075348, + "learning_rate": 9.50158103440616e-06, + "loss": 0.0169, + "step": 81050 + }, + { + "epoch": 5.310882526370962, + "grad_norm": 0.22358861565589905, + "learning_rate": 9.491886085872815e-06, + "loss": 0.0425, + "step": 81060 + }, + { + "epoch": 5.311537705562471, + "grad_norm": 0.33070486783981323, + "learning_rate": 9.482195567245895e-06, + "loss": 0.0325, + "step": 81070 + }, + { + "epoch": 5.31219288475398, + "grad_norm": 0.10374753177165985, + "learning_rate": 9.472509479585151e-06, + "loss": 0.0229, + "step": 81080 + }, + { + "epoch": 5.312848063945489, + "grad_norm": 0.41636398434638977, + "learning_rate": 9.462827823949833e-06, + "loss": 0.0367, + "step": 81090 + }, + { + "epoch": 5.3135032431369975, + "grad_norm": 0.3519321084022522, + "learning_rate": 9.453150601398735e-06, + "loss": 0.0163, + "step": 81100 + }, + { + "epoch": 5.314158422328507, + "grad_norm": 0.7728390097618103, + "learning_rate": 9.44347781299011e-06, + "loss": 0.0196, + "step": 81110 + }, + { + "epoch": 5.314813601520016, + "grad_norm": 1.5497082471847534, + "learning_rate": 9.43380945978179e-06, + "loss": 0.046, + "step": 81120 + }, + { + "epoch": 5.315468780711525, + "grad_norm": 0.3499756455421448, + "learning_rate": 9.424145542831047e-06, + "loss": 0.036, + "step": 81130 + }, + { + "epoch": 5.316123959903034, + "grad_norm": 0.27613741159439087, + "learning_rate": 9.414486063194778e-06, + "loss": 0.0357, + "step": 81140 + }, + { + "epoch": 5.3167791390945425, + "grad_norm": 0.3586326837539673, + "learning_rate": 9.404831021929273e-06, + "loss": 0.0282, + "step": 81150 + }, + { + "epoch": 5.317434318286051, + "grad_norm": 0.06291992217302322, + "learning_rate": 9.395180420090427e-06, + "loss": 0.0321, + "step": 81160 + }, + { + "epoch": 5.31808949747756, + "grad_norm": 0.180702805519104, + "learning_rate": 9.385534258733591e-06, + "loss": 0.0283, + "step": 81170 + }, + { + "epoch": 5.318744676669069, + "grad_norm": 0.35866779088974, + "learning_rate": 9.375892538913677e-06, + "loss": 0.0337, + "step": 81180 + }, + { + "epoch": 5.319399855860578, + "grad_norm": 0.1159864217042923, + "learning_rate": 9.36625526168507e-06, + "loss": 0.0294, + "step": 81190 + }, + { + "epoch": 5.3200550350520865, + "grad_norm": 0.5420145988464355, + "learning_rate": 9.356622428101686e-06, + "loss": 0.0109, + "step": 81200 + }, + { + "epoch": 5.320710214243595, + "grad_norm": 0.29888835549354553, + "learning_rate": 9.346994039216972e-06, + "loss": 0.0311, + "step": 81210 + }, + { + "epoch": 5.321365393435104, + "grad_norm": 0.6304943561553955, + "learning_rate": 9.337370096083886e-06, + "loss": 0.0503, + "step": 81220 + }, + { + "epoch": 5.322020572626613, + "grad_norm": 0.8279045224189758, + "learning_rate": 9.327750599754854e-06, + "loss": 0.0267, + "step": 81230 + }, + { + "epoch": 5.322675751818123, + "grad_norm": 0.6335012912750244, + "learning_rate": 9.318135551281875e-06, + "loss": 0.025, + "step": 81240 + }, + { + "epoch": 5.3233309310096315, + "grad_norm": 0.1734188348054886, + "learning_rate": 9.308524951716424e-06, + "loss": 0.0296, + "step": 81250 + }, + { + "epoch": 5.32398611020114, + "grad_norm": 0.12829646468162537, + "learning_rate": 9.298918802109497e-06, + "loss": 0.0191, + "step": 81260 + }, + { + "epoch": 5.324641289392649, + "grad_norm": 0.7865809202194214, + "learning_rate": 9.289317103511625e-06, + "loss": 0.0301, + "step": 81270 + }, + { + "epoch": 5.325296468584158, + "grad_norm": 0.0869721844792366, + "learning_rate": 9.279719856972835e-06, + "loss": 0.0225, + "step": 81280 + }, + { + "epoch": 5.325951647775667, + "grad_norm": 2.7002201080322266, + "learning_rate": 9.270127063542645e-06, + "loss": 0.0386, + "step": 81290 + }, + { + "epoch": 5.326606826967176, + "grad_norm": 0.1687779575586319, + "learning_rate": 9.260538724270124e-06, + "loss": 0.0222, + "step": 81300 + }, + { + "epoch": 5.327262006158684, + "grad_norm": 0.2030700296163559, + "learning_rate": 9.250954840203852e-06, + "loss": 0.0183, + "step": 81310 + }, + { + "epoch": 5.327917185350193, + "grad_norm": 0.7333410978317261, + "learning_rate": 9.241375412391878e-06, + "loss": 0.0186, + "step": 81320 + }, + { + "epoch": 5.328572364541702, + "grad_norm": 0.4919206202030182, + "learning_rate": 9.23180044188181e-06, + "loss": 0.0325, + "step": 81330 + }, + { + "epoch": 5.329227543733211, + "grad_norm": 0.5558984875679016, + "learning_rate": 9.222229929720755e-06, + "loss": 0.0467, + "step": 81340 + }, + { + "epoch": 5.32988272292472, + "grad_norm": 0.13815593719482422, + "learning_rate": 9.212663876955336e-06, + "loss": 0.0346, + "step": 81350 + }, + { + "epoch": 5.3305379021162285, + "grad_norm": 0.1108478233218193, + "learning_rate": 9.20310228463166e-06, + "loss": 0.0261, + "step": 81360 + }, + { + "epoch": 5.331193081307737, + "grad_norm": 0.8500959873199463, + "learning_rate": 9.19354515379539e-06, + "loss": 0.046, + "step": 81370 + }, + { + "epoch": 5.331848260499246, + "grad_norm": 0.14112737774848938, + "learning_rate": 9.183992485491666e-06, + "loss": 0.0238, + "step": 81380 + }, + { + "epoch": 5.332503439690756, + "grad_norm": 1.0340276956558228, + "learning_rate": 9.174444280765154e-06, + "loss": 0.0293, + "step": 81390 + }, + { + "epoch": 5.333158618882265, + "grad_norm": 0.3759150803089142, + "learning_rate": 9.164900540660032e-06, + "loss": 0.0358, + "step": 81400 + }, + { + "epoch": 5.333813798073773, + "grad_norm": 0.8737431764602661, + "learning_rate": 9.155361266220008e-06, + "loss": 0.0298, + "step": 81410 + }, + { + "epoch": 5.334468977265282, + "grad_norm": 0.22946617007255554, + "learning_rate": 9.145826458488254e-06, + "loss": 0.0391, + "step": 81420 + }, + { + "epoch": 5.335124156456791, + "grad_norm": 0.14046230912208557, + "learning_rate": 9.136296118507509e-06, + "loss": 0.0268, + "step": 81430 + }, + { + "epoch": 5.3357793356483, + "grad_norm": 0.14102846384048462, + "learning_rate": 9.126770247319966e-06, + "loss": 0.0259, + "step": 81440 + }, + { + "epoch": 5.336434514839809, + "grad_norm": 0.21316803991794586, + "learning_rate": 9.11724884596738e-06, + "loss": 0.0475, + "step": 81450 + }, + { + "epoch": 5.3370896940313175, + "grad_norm": 0.9528684020042419, + "learning_rate": 9.107731915490996e-06, + "loss": 0.0207, + "step": 81460 + }, + { + "epoch": 5.337744873222826, + "grad_norm": 2.0678508281707764, + "learning_rate": 9.098219456931579e-06, + "loss": 0.0273, + "step": 81470 + }, + { + "epoch": 5.338400052414335, + "grad_norm": 0.6666977405548096, + "learning_rate": 9.088711471329376e-06, + "loss": 0.0455, + "step": 81480 + }, + { + "epoch": 5.339055231605844, + "grad_norm": 0.15688253939151764, + "learning_rate": 9.079207959724184e-06, + "loss": 0.0238, + "step": 81490 + }, + { + "epoch": 5.339710410797353, + "grad_norm": 0.24976062774658203, + "learning_rate": 9.069708923155279e-06, + "loss": 0.0342, + "step": 81500 + }, + { + "epoch": 5.340365589988862, + "grad_norm": 0.6936963796615601, + "learning_rate": 9.060214362661468e-06, + "loss": 0.0903, + "step": 81510 + }, + { + "epoch": 5.341020769180371, + "grad_norm": 0.19100980460643768, + "learning_rate": 9.050724279281064e-06, + "loss": 0.0248, + "step": 81520 + }, + { + "epoch": 5.34167594837188, + "grad_norm": 0.16450762748718262, + "learning_rate": 9.041238674051894e-06, + "loss": 0.0174, + "step": 81530 + }, + { + "epoch": 5.342331127563389, + "grad_norm": 0.25442764163017273, + "learning_rate": 9.031757548011277e-06, + "loss": 0.0354, + "step": 81540 + }, + { + "epoch": 5.342986306754898, + "grad_norm": 0.7504900097846985, + "learning_rate": 9.022280902196051e-06, + "loss": 0.0261, + "step": 81550 + }, + { + "epoch": 5.3436414859464065, + "grad_norm": 0.23610006272792816, + "learning_rate": 9.012808737642592e-06, + "loss": 0.032, + "step": 81560 + }, + { + "epoch": 5.344296665137915, + "grad_norm": 0.4836288392543793, + "learning_rate": 9.003341055386739e-06, + "loss": 0.0219, + "step": 81570 + }, + { + "epoch": 5.344951844329424, + "grad_norm": 0.23606131970882416, + "learning_rate": 8.993877856463862e-06, + "loss": 0.0196, + "step": 81580 + }, + { + "epoch": 5.345607023520933, + "grad_norm": 1.8735253810882568, + "learning_rate": 8.984419141908856e-06, + "loss": 0.0418, + "step": 81590 + }, + { + "epoch": 5.346262202712442, + "grad_norm": 0.19645938277244568, + "learning_rate": 8.974964912756117e-06, + "loss": 0.0302, + "step": 81600 + }, + { + "epoch": 5.346917381903951, + "grad_norm": 0.6489204168319702, + "learning_rate": 8.965515170039523e-06, + "loss": 0.046, + "step": 81610 + }, + { + "epoch": 5.347572561095459, + "grad_norm": 0.46043217182159424, + "learning_rate": 8.956069914792514e-06, + "loss": 0.0356, + "step": 81620 + }, + { + "epoch": 5.348227740286968, + "grad_norm": 0.19083376228809357, + "learning_rate": 8.946629148047969e-06, + "loss": 0.0267, + "step": 81630 + }, + { + "epoch": 5.348882919478477, + "grad_norm": 0.8683565855026245, + "learning_rate": 8.93719287083834e-06, + "loss": 0.0409, + "step": 81640 + }, + { + "epoch": 5.349538098669986, + "grad_norm": 0.16947023570537567, + "learning_rate": 8.927761084195569e-06, + "loss": 0.0275, + "step": 81650 + }, + { + "epoch": 5.350193277861495, + "grad_norm": 0.2116069793701172, + "learning_rate": 8.918333789151096e-06, + "loss": 0.0331, + "step": 81660 + }, + { + "epoch": 5.350848457053004, + "grad_norm": 0.1840563863515854, + "learning_rate": 8.90891098673587e-06, + "loss": 0.0169, + "step": 81670 + }, + { + "epoch": 5.351503636244513, + "grad_norm": 0.5648795366287231, + "learning_rate": 8.89949267798037e-06, + "loss": 0.031, + "step": 81680 + }, + { + "epoch": 5.352158815436022, + "grad_norm": 0.34995749592781067, + "learning_rate": 8.890078863914542e-06, + "loss": 0.0204, + "step": 81690 + }, + { + "epoch": 5.352813994627531, + "grad_norm": 0.5357644557952881, + "learning_rate": 8.880669545567894e-06, + "loss": 0.0211, + "step": 81700 + }, + { + "epoch": 5.35346917381904, + "grad_norm": 1.0568920373916626, + "learning_rate": 8.871264723969381e-06, + "loss": 0.0253, + "step": 81710 + }, + { + "epoch": 5.3541243530105485, + "grad_norm": 0.08769823610782623, + "learning_rate": 8.86186440014754e-06, + "loss": 0.0297, + "step": 81720 + }, + { + "epoch": 5.354779532202057, + "grad_norm": 0.8533346652984619, + "learning_rate": 8.852468575130351e-06, + "loss": 0.0237, + "step": 81730 + }, + { + "epoch": 5.355434711393566, + "grad_norm": 1.8745123147964478, + "learning_rate": 8.843077249945347e-06, + "loss": 0.0405, + "step": 81740 + }, + { + "epoch": 5.356089890585075, + "grad_norm": 0.6663625240325928, + "learning_rate": 8.833690425619523e-06, + "loss": 0.0391, + "step": 81750 + }, + { + "epoch": 5.356745069776584, + "grad_norm": 0.2421121448278427, + "learning_rate": 8.82430810317942e-06, + "loss": 0.0226, + "step": 81760 + }, + { + "epoch": 5.3574002489680925, + "grad_norm": 0.24464841187000275, + "learning_rate": 8.814930283651074e-06, + "loss": 0.0342, + "step": 81770 + }, + { + "epoch": 5.358055428159601, + "grad_norm": 0.31576719880104065, + "learning_rate": 8.805556968060048e-06, + "loss": 0.0411, + "step": 81780 + }, + { + "epoch": 5.35871060735111, + "grad_norm": 0.144292950630188, + "learning_rate": 8.796188157431362e-06, + "loss": 0.0222, + "step": 81790 + }, + { + "epoch": 5.35936578654262, + "grad_norm": 0.8070035576820374, + "learning_rate": 8.786823852789595e-06, + "loss": 0.0285, + "step": 81800 + }, + { + "epoch": 5.360020965734129, + "grad_norm": 0.5422377586364746, + "learning_rate": 8.777464055158813e-06, + "loss": 0.0258, + "step": 81810 + }, + { + "epoch": 5.3606761449256375, + "grad_norm": 0.19810523092746735, + "learning_rate": 8.768108765562578e-06, + "loss": 0.079, + "step": 81820 + }, + { + "epoch": 5.361331324117146, + "grad_norm": 0.9004135727882385, + "learning_rate": 8.758757985023975e-06, + "loss": 0.0238, + "step": 81830 + }, + { + "epoch": 5.361986503308655, + "grad_norm": 0.1203421875834465, + "learning_rate": 8.749411714565598e-06, + "loss": 0.0329, + "step": 81840 + }, + { + "epoch": 5.362641682500164, + "grad_norm": 0.430002361536026, + "learning_rate": 8.74006995520954e-06, + "loss": 0.0269, + "step": 81850 + }, + { + "epoch": 5.363296861691673, + "grad_norm": 0.7346829771995544, + "learning_rate": 8.730732707977397e-06, + "loss": 0.0465, + "step": 81860 + }, + { + "epoch": 5.363952040883182, + "grad_norm": 0.5724815726280212, + "learning_rate": 8.721399973890281e-06, + "loss": 0.0486, + "step": 81870 + }, + { + "epoch": 5.36460722007469, + "grad_norm": 0.231783926486969, + "learning_rate": 8.712071753968792e-06, + "loss": 0.0159, + "step": 81880 + }, + { + "epoch": 5.365262399266199, + "grad_norm": 0.473805695772171, + "learning_rate": 8.702748049233072e-06, + "loss": 0.0472, + "step": 81890 + }, + { + "epoch": 5.365917578457708, + "grad_norm": 0.7409064769744873, + "learning_rate": 8.693428860702718e-06, + "loss": 0.0266, + "step": 81900 + }, + { + "epoch": 5.366572757649217, + "grad_norm": 0.2247837781906128, + "learning_rate": 8.684114189396898e-06, + "loss": 0.0292, + "step": 81910 + }, + { + "epoch": 5.367227936840726, + "grad_norm": 0.8489463925361633, + "learning_rate": 8.674804036334222e-06, + "loss": 0.0385, + "step": 81920 + }, + { + "epoch": 5.3678831160322344, + "grad_norm": 0.19468148052692413, + "learning_rate": 8.665498402532857e-06, + "loss": 0.0234, + "step": 81930 + }, + { + "epoch": 5.368538295223743, + "grad_norm": 0.1114361584186554, + "learning_rate": 8.656197289010432e-06, + "loss": 0.021, + "step": 81940 + }, + { + "epoch": 5.369193474415253, + "grad_norm": 0.5713685154914856, + "learning_rate": 8.646900696784122e-06, + "loss": 0.036, + "step": 81950 + }, + { + "epoch": 5.369848653606762, + "grad_norm": 0.3245597779750824, + "learning_rate": 8.637608626870553e-06, + "loss": 0.0226, + "step": 81960 + }, + { + "epoch": 5.370503832798271, + "grad_norm": 0.8224713802337646, + "learning_rate": 8.628321080285944e-06, + "loss": 0.0333, + "step": 81970 + }, + { + "epoch": 5.371159011989779, + "grad_norm": 0.20734460651874542, + "learning_rate": 8.619038058045925e-06, + "loss": 0.0137, + "step": 81980 + }, + { + "epoch": 5.371814191181288, + "grad_norm": 0.16217416524887085, + "learning_rate": 8.609759561165704e-06, + "loss": 0.0385, + "step": 81990 + }, + { + "epoch": 5.372469370372797, + "grad_norm": 0.9152446985244751, + "learning_rate": 8.600485590659934e-06, + "loss": 0.0334, + "step": 82000 + }, + { + "epoch": 5.373124549564306, + "grad_norm": 0.5258349180221558, + "learning_rate": 8.591216147542813e-06, + "loss": 0.0278, + "step": 82010 + }, + { + "epoch": 5.373779728755815, + "grad_norm": 0.4014248847961426, + "learning_rate": 8.581951232828045e-06, + "loss": 0.0264, + "step": 82020 + }, + { + "epoch": 5.3744349079473235, + "grad_norm": 0.2759817838668823, + "learning_rate": 8.572690847528825e-06, + "loss": 0.0397, + "step": 82030 + }, + { + "epoch": 5.375090087138832, + "grad_norm": 0.7434698343276978, + "learning_rate": 8.563434992657838e-06, + "loss": 0.0171, + "step": 82040 + }, + { + "epoch": 5.375745266330341, + "grad_norm": 0.37047263979911804, + "learning_rate": 8.554183669227307e-06, + "loss": 0.024, + "step": 82050 + }, + { + "epoch": 5.37640044552185, + "grad_norm": 0.1453297883272171, + "learning_rate": 8.544936878248937e-06, + "loss": 0.0585, + "step": 82060 + }, + { + "epoch": 5.377055624713359, + "grad_norm": 0.83172607421875, + "learning_rate": 8.535694620733942e-06, + "loss": 0.0215, + "step": 82070 + }, + { + "epoch": 5.377710803904868, + "grad_norm": 0.06753363460302353, + "learning_rate": 8.526456897693053e-06, + "loss": 0.0317, + "step": 82080 + }, + { + "epoch": 5.378365983096377, + "grad_norm": 0.13142968714237213, + "learning_rate": 8.517223710136456e-06, + "loss": 0.0359, + "step": 82090 + }, + { + "epoch": 5.379021162287886, + "grad_norm": 0.12720663845539093, + "learning_rate": 8.507995059073931e-06, + "loss": 0.0365, + "step": 82100 + }, + { + "epoch": 5.379676341479395, + "grad_norm": 0.24976378679275513, + "learning_rate": 8.498770945514673e-06, + "loss": 0.0472, + "step": 82110 + }, + { + "epoch": 5.380331520670904, + "grad_norm": 0.4513603150844574, + "learning_rate": 8.489551370467436e-06, + "loss": 0.0199, + "step": 82120 + }, + { + "epoch": 5.3809866998624125, + "grad_norm": 0.3706549108028412, + "learning_rate": 8.480336334940442e-06, + "loss": 0.0314, + "step": 82130 + }, + { + "epoch": 5.381641879053921, + "grad_norm": 0.26104190945625305, + "learning_rate": 8.471125839941458e-06, + "loss": 0.0116, + "step": 82140 + }, + { + "epoch": 5.38229705824543, + "grad_norm": 0.3426010012626648, + "learning_rate": 8.461919886477682e-06, + "loss": 0.0219, + "step": 82150 + }, + { + "epoch": 5.382952237436939, + "grad_norm": 0.09841502457857132, + "learning_rate": 8.452718475555927e-06, + "loss": 0.0378, + "step": 82160 + }, + { + "epoch": 5.383607416628448, + "grad_norm": 0.5937157273292542, + "learning_rate": 8.443521608182398e-06, + "loss": 0.0297, + "step": 82170 + }, + { + "epoch": 5.384262595819957, + "grad_norm": 0.3117114007472992, + "learning_rate": 8.43432928536288e-06, + "loss": 0.0385, + "step": 82180 + }, + { + "epoch": 5.384917775011465, + "grad_norm": 0.14120979607105255, + "learning_rate": 8.425141508102607e-06, + "loss": 0.0228, + "step": 82190 + }, + { + "epoch": 5.385572954202974, + "grad_norm": 0.41565442085266113, + "learning_rate": 8.415958277406365e-06, + "loss": 0.0234, + "step": 82200 + }, + { + "epoch": 5.386228133394483, + "grad_norm": 0.16178737580776215, + "learning_rate": 8.406779594278392e-06, + "loss": 0.0232, + "step": 82210 + }, + { + "epoch": 5.386883312585992, + "grad_norm": 0.6610947251319885, + "learning_rate": 8.397605459722468e-06, + "loss": 0.0374, + "step": 82220 + }, + { + "epoch": 5.3875384917775015, + "grad_norm": 0.3172730803489685, + "learning_rate": 8.388435874741863e-06, + "loss": 0.0313, + "step": 82230 + }, + { + "epoch": 5.38819367096901, + "grad_norm": 0.7530319690704346, + "learning_rate": 8.37927084033936e-06, + "loss": 0.0456, + "step": 82240 + }, + { + "epoch": 5.388848850160519, + "grad_norm": 0.6406867504119873, + "learning_rate": 8.370110357517208e-06, + "loss": 0.0416, + "step": 82250 + }, + { + "epoch": 5.389504029352028, + "grad_norm": 0.593625545501709, + "learning_rate": 8.360954427277201e-06, + "loss": 0.0289, + "step": 82260 + }, + { + "epoch": 5.390159208543537, + "grad_norm": 0.542180061340332, + "learning_rate": 8.351803050620627e-06, + "loss": 0.0186, + "step": 82270 + }, + { + "epoch": 5.390814387735046, + "grad_norm": 1.009901762008667, + "learning_rate": 8.34265622854824e-06, + "loss": 0.0218, + "step": 82280 + }, + { + "epoch": 5.391469566926554, + "grad_norm": 0.15439742803573608, + "learning_rate": 8.333513962060335e-06, + "loss": 0.0158, + "step": 82290 + }, + { + "epoch": 5.392124746118063, + "grad_norm": 0.6109400987625122, + "learning_rate": 8.324376252156701e-06, + "loss": 0.0393, + "step": 82300 + }, + { + "epoch": 5.392779925309572, + "grad_norm": 0.14682945609092712, + "learning_rate": 8.315243099836633e-06, + "loss": 0.012, + "step": 82310 + }, + { + "epoch": 5.393435104501081, + "grad_norm": 0.5470443964004517, + "learning_rate": 8.3061145060989e-06, + "loss": 0.0268, + "step": 82320 + }, + { + "epoch": 5.39409028369259, + "grad_norm": 0.9569965600967407, + "learning_rate": 8.296990471941806e-06, + "loss": 0.0426, + "step": 82330 + }, + { + "epoch": 5.3947454628840985, + "grad_norm": 0.7599746584892273, + "learning_rate": 8.287870998363112e-06, + "loss": 0.037, + "step": 82340 + }, + { + "epoch": 5.395400642075607, + "grad_norm": 1.1017029285430908, + "learning_rate": 8.278756086360156e-06, + "loss": 0.0163, + "step": 82350 + }, + { + "epoch": 5.396055821267117, + "grad_norm": 0.28188738226890564, + "learning_rate": 8.269645736929693e-06, + "loss": 0.0182, + "step": 82360 + }, + { + "epoch": 5.396711000458626, + "grad_norm": 1.4792684316635132, + "learning_rate": 8.260539951068048e-06, + "loss": 0.0495, + "step": 82370 + }, + { + "epoch": 5.397366179650135, + "grad_norm": 0.16210457682609558, + "learning_rate": 8.251438729770983e-06, + "loss": 0.0331, + "step": 82380 + }, + { + "epoch": 5.3980213588416435, + "grad_norm": 0.12054276466369629, + "learning_rate": 8.242342074033826e-06, + "loss": 0.0355, + "step": 82390 + }, + { + "epoch": 5.398676538033152, + "grad_norm": 0.3295758366584778, + "learning_rate": 8.233249984851344e-06, + "loss": 0.0227, + "step": 82400 + }, + { + "epoch": 5.399331717224661, + "grad_norm": 0.20856733620166779, + "learning_rate": 8.224162463217849e-06, + "loss": 0.0139, + "step": 82410 + }, + { + "epoch": 5.39998689641617, + "grad_norm": 0.21139755845069885, + "learning_rate": 8.215079510127139e-06, + "loss": 0.0399, + "step": 82420 + }, + { + "epoch": 5.400642075607679, + "grad_norm": 0.31770527362823486, + "learning_rate": 8.206001126572521e-06, + "loss": 0.0405, + "step": 82430 + }, + { + "epoch": 5.4012972547991875, + "grad_norm": 0.32758456468582153, + "learning_rate": 8.196927313546776e-06, + "loss": 0.0503, + "step": 82440 + }, + { + "epoch": 5.401952433990696, + "grad_norm": 0.5716198086738586, + "learning_rate": 8.18785807204222e-06, + "loss": 0.0134, + "step": 82450 + }, + { + "epoch": 5.402607613182205, + "grad_norm": 0.37567856907844543, + "learning_rate": 8.178793403050627e-06, + "loss": 0.0175, + "step": 82460 + }, + { + "epoch": 5.403262792373714, + "grad_norm": 0.6173544526100159, + "learning_rate": 8.16973330756331e-06, + "loss": 0.0415, + "step": 82470 + }, + { + "epoch": 5.403917971565223, + "grad_norm": 0.620593786239624, + "learning_rate": 8.16067778657107e-06, + "loss": 0.0295, + "step": 82480 + }, + { + "epoch": 5.404573150756732, + "grad_norm": 0.10489927232265472, + "learning_rate": 8.151626841064214e-06, + "loss": 0.0185, + "step": 82490 + }, + { + "epoch": 5.40522832994824, + "grad_norm": 0.7900774478912354, + "learning_rate": 8.142580472032518e-06, + "loss": 0.0395, + "step": 82500 + }, + { + "epoch": 5.40588350913975, + "grad_norm": 0.17418110370635986, + "learning_rate": 8.133538680465285e-06, + "loss": 0.0293, + "step": 82510 + }, + { + "epoch": 5.406538688331259, + "grad_norm": 0.2606140971183777, + "learning_rate": 8.124501467351336e-06, + "loss": 0.0314, + "step": 82520 + }, + { + "epoch": 5.407193867522768, + "grad_norm": 0.9871073365211487, + "learning_rate": 8.11546883367893e-06, + "loss": 0.0173, + "step": 82530 + }, + { + "epoch": 5.407849046714277, + "grad_norm": 0.6637263298034668, + "learning_rate": 8.106440780435882e-06, + "loss": 0.0262, + "step": 82540 + }, + { + "epoch": 5.408504225905785, + "grad_norm": 1.3170180320739746, + "learning_rate": 8.097417308609485e-06, + "loss": 0.0515, + "step": 82550 + }, + { + "epoch": 5.409159405097294, + "grad_norm": 1.1438372135162354, + "learning_rate": 8.088398419186544e-06, + "loss": 0.0537, + "step": 82560 + }, + { + "epoch": 5.409814584288803, + "grad_norm": 0.13885171711444855, + "learning_rate": 8.079384113153321e-06, + "loss": 0.0173, + "step": 82570 + }, + { + "epoch": 5.410469763480312, + "grad_norm": 0.1769644021987915, + "learning_rate": 8.070374391495644e-06, + "loss": 0.018, + "step": 82580 + }, + { + "epoch": 5.411124942671821, + "grad_norm": 0.14836791157722473, + "learning_rate": 8.061369255198764e-06, + "loss": 0.0217, + "step": 82590 + }, + { + "epoch": 5.4117801218633295, + "grad_norm": 1.1079483032226562, + "learning_rate": 8.052368705247493e-06, + "loss": 0.0283, + "step": 82600 + }, + { + "epoch": 5.412435301054838, + "grad_norm": 0.5209329724311829, + "learning_rate": 8.043372742626114e-06, + "loss": 0.0286, + "step": 82610 + }, + { + "epoch": 5.413090480246347, + "grad_norm": 0.48524072766304016, + "learning_rate": 8.034381368318422e-06, + "loss": 0.0396, + "step": 82620 + }, + { + "epoch": 5.413745659437856, + "grad_norm": 0.12426851689815521, + "learning_rate": 8.025394583307677e-06, + "loss": 0.0274, + "step": 82630 + }, + { + "epoch": 5.414400838629365, + "grad_norm": 0.18052686750888824, + "learning_rate": 8.016412388576683e-06, + "loss": 0.0196, + "step": 82640 + }, + { + "epoch": 5.415056017820874, + "grad_norm": 0.15492503345012665, + "learning_rate": 8.007434785107699e-06, + "loss": 0.0287, + "step": 82650 + }, + { + "epoch": 5.415711197012383, + "grad_norm": 0.4175020158290863, + "learning_rate": 7.998461773882509e-06, + "loss": 0.0161, + "step": 82660 + }, + { + "epoch": 5.416366376203892, + "grad_norm": 0.11877617239952087, + "learning_rate": 7.989493355882393e-06, + "loss": 0.0221, + "step": 82670 + }, + { + "epoch": 5.417021555395401, + "grad_norm": 0.45284339785575867, + "learning_rate": 7.980529532088137e-06, + "loss": 0.0169, + "step": 82680 + }, + { + "epoch": 5.41767673458691, + "grad_norm": 0.8649369478225708, + "learning_rate": 7.97157030347998e-06, + "loss": 0.0517, + "step": 82690 + }, + { + "epoch": 5.4183319137784185, + "grad_norm": 0.1310017704963684, + "learning_rate": 7.96261567103772e-06, + "loss": 0.0375, + "step": 82700 + }, + { + "epoch": 5.418987092969927, + "grad_norm": 0.3141894042491913, + "learning_rate": 7.953665635740593e-06, + "loss": 0.0253, + "step": 82710 + }, + { + "epoch": 5.419642272161436, + "grad_norm": 0.28744956851005554, + "learning_rate": 7.94472019856738e-06, + "loss": 0.017, + "step": 82720 + }, + { + "epoch": 5.420297451352945, + "grad_norm": 0.27560073137283325, + "learning_rate": 7.935779360496337e-06, + "loss": 0.0149, + "step": 82730 + }, + { + "epoch": 5.420952630544454, + "grad_norm": 0.5891409516334534, + "learning_rate": 7.92684312250523e-06, + "loss": 0.0149, + "step": 82740 + }, + { + "epoch": 5.421607809735963, + "grad_norm": 0.3971507251262665, + "learning_rate": 7.917911485571289e-06, + "loss": 0.0356, + "step": 82750 + }, + { + "epoch": 5.422262988927471, + "grad_norm": 0.6586543321609497, + "learning_rate": 7.90898445067128e-06, + "loss": 0.0426, + "step": 82760 + }, + { + "epoch": 5.42291816811898, + "grad_norm": 0.22345151007175446, + "learning_rate": 7.900062018781462e-06, + "loss": 0.0153, + "step": 82770 + }, + { + "epoch": 5.423573347310489, + "grad_norm": 0.12430208921432495, + "learning_rate": 7.891144190877553e-06, + "loss": 0.0339, + "step": 82780 + }, + { + "epoch": 5.424228526501999, + "grad_norm": 0.9102693200111389, + "learning_rate": 7.882230967934806e-06, + "loss": 0.0391, + "step": 82790 + }, + { + "epoch": 5.4248837056935075, + "grad_norm": 0.2433486133813858, + "learning_rate": 7.873322350927959e-06, + "loss": 0.018, + "step": 82800 + }, + { + "epoch": 5.425538884885016, + "grad_norm": 0.3027121126651764, + "learning_rate": 7.864418340831253e-06, + "loss": 0.0201, + "step": 82810 + }, + { + "epoch": 5.426194064076525, + "grad_norm": 0.7443525195121765, + "learning_rate": 7.855518938618395e-06, + "loss": 0.0283, + "step": 82820 + }, + { + "epoch": 5.426849243268034, + "grad_norm": 0.6198862791061401, + "learning_rate": 7.846624145262633e-06, + "loss": 0.0195, + "step": 82830 + }, + { + "epoch": 5.427504422459543, + "grad_norm": 0.2874487042427063, + "learning_rate": 7.837733961736672e-06, + "loss": 0.0154, + "step": 82840 + }, + { + "epoch": 5.428159601651052, + "grad_norm": 0.47381383180618286, + "learning_rate": 7.828848389012727e-06, + "loss": 0.0439, + "step": 82850 + }, + { + "epoch": 5.42881478084256, + "grad_norm": 0.1691436618566513, + "learning_rate": 7.819967428062524e-06, + "loss": 0.0472, + "step": 82860 + }, + { + "epoch": 5.429469960034069, + "grad_norm": 0.8214083909988403, + "learning_rate": 7.811091079857274e-06, + "loss": 0.0292, + "step": 82870 + }, + { + "epoch": 5.430125139225578, + "grad_norm": 0.36138832569122314, + "learning_rate": 7.802219345367667e-06, + "loss": 0.0195, + "step": 82880 + }, + { + "epoch": 5.430780318417087, + "grad_norm": 0.1856774091720581, + "learning_rate": 7.793352225563916e-06, + "loss": 0.0145, + "step": 82890 + }, + { + "epoch": 5.431435497608596, + "grad_norm": 0.6508703827857971, + "learning_rate": 7.7844897214157e-06, + "loss": 0.0311, + "step": 82900 + }, + { + "epoch": 5.4320906768001045, + "grad_norm": 0.2799055576324463, + "learning_rate": 7.775631833892228e-06, + "loss": 0.0661, + "step": 82910 + }, + { + "epoch": 5.432745855991613, + "grad_norm": 0.46868351101875305, + "learning_rate": 7.76677856396215e-06, + "loss": 0.0194, + "step": 82920 + }, + { + "epoch": 5.433401035183122, + "grad_norm": 0.2051856964826584, + "learning_rate": 7.757929912593697e-06, + "loss": 0.0253, + "step": 82930 + }, + { + "epoch": 5.434056214374632, + "grad_norm": 0.3572387099266052, + "learning_rate": 7.749085880754509e-06, + "loss": 0.0228, + "step": 82940 + }, + { + "epoch": 5.434711393566141, + "grad_norm": 0.9502637982368469, + "learning_rate": 7.740246469411777e-06, + "loss": 0.0167, + "step": 82950 + }, + { + "epoch": 5.4353665727576495, + "grad_norm": 0.14256179332733154, + "learning_rate": 7.731411679532142e-06, + "loss": 0.031, + "step": 82960 + }, + { + "epoch": 5.436021751949158, + "grad_norm": 0.7347416281700134, + "learning_rate": 7.722581512081779e-06, + "loss": 0.0411, + "step": 82970 + }, + { + "epoch": 5.436676931140667, + "grad_norm": 0.38667431473731995, + "learning_rate": 7.713755968026331e-06, + "loss": 0.0577, + "step": 82980 + }, + { + "epoch": 5.437332110332176, + "grad_norm": 0.712617039680481, + "learning_rate": 7.704935048330969e-06, + "loss": 0.0203, + "step": 82990 + }, + { + "epoch": 5.437987289523685, + "grad_norm": 0.5259793996810913, + "learning_rate": 7.696118753960313e-06, + "loss": 0.0311, + "step": 83000 + }, + { + "epoch": 5.4386424687151935, + "grad_norm": 0.24857108294963837, + "learning_rate": 7.687307085878503e-06, + "loss": 0.0333, + "step": 83010 + }, + { + "epoch": 5.439297647906702, + "grad_norm": 0.6466138958930969, + "learning_rate": 7.678500045049186e-06, + "loss": 0.0164, + "step": 83020 + }, + { + "epoch": 5.439952827098211, + "grad_norm": 0.3964802324771881, + "learning_rate": 7.669697632435463e-06, + "loss": 0.0347, + "step": 83030 + }, + { + "epoch": 5.44060800628972, + "grad_norm": 0.43828773498535156, + "learning_rate": 7.660899848999963e-06, + "loss": 0.032, + "step": 83040 + }, + { + "epoch": 5.441263185481229, + "grad_norm": 1.2886180877685547, + "learning_rate": 7.652106695704803e-06, + "loss": 0.0464, + "step": 83050 + }, + { + "epoch": 5.441918364672738, + "grad_norm": 0.818295955657959, + "learning_rate": 7.643318173511598e-06, + "loss": 0.0197, + "step": 83060 + }, + { + "epoch": 5.442573543864247, + "grad_norm": 0.4169715642929077, + "learning_rate": 7.634534283381423e-06, + "loss": 0.0281, + "step": 83070 + }, + { + "epoch": 5.443228723055756, + "grad_norm": 0.13025817275047302, + "learning_rate": 7.625755026274889e-06, + "loss": 0.0546, + "step": 83080 + }, + { + "epoch": 5.443883902247265, + "grad_norm": 0.6040236949920654, + "learning_rate": 7.616980403152069e-06, + "loss": 0.0417, + "step": 83090 + }, + { + "epoch": 5.444539081438774, + "grad_norm": 0.3154393434524536, + "learning_rate": 7.608210414972561e-06, + "loss": 0.0283, + "step": 83100 + }, + { + "epoch": 5.445194260630283, + "grad_norm": 0.1938558965921402, + "learning_rate": 7.599445062695404e-06, + "loss": 0.0496, + "step": 83110 + }, + { + "epoch": 5.445849439821791, + "grad_norm": 0.2775847613811493, + "learning_rate": 7.590684347279214e-06, + "loss": 0.0348, + "step": 83120 + }, + { + "epoch": 5.4465046190133, + "grad_norm": 0.13663209974765778, + "learning_rate": 7.581928269682004e-06, + "loss": 0.0281, + "step": 83130 + }, + { + "epoch": 5.447159798204809, + "grad_norm": 1.0376499891281128, + "learning_rate": 7.573176830861361e-06, + "loss": 0.0214, + "step": 83140 + }, + { + "epoch": 5.447814977396318, + "grad_norm": 0.3663599491119385, + "learning_rate": 7.5644300317743e-06, + "loss": 0.03, + "step": 83150 + }, + { + "epoch": 5.448470156587827, + "grad_norm": 0.12323389947414398, + "learning_rate": 7.555687873377376e-06, + "loss": 0.017, + "step": 83160 + }, + { + "epoch": 5.449125335779335, + "grad_norm": 0.481161892414093, + "learning_rate": 7.546950356626592e-06, + "loss": 0.0431, + "step": 83170 + }, + { + "epoch": 5.449780514970844, + "grad_norm": 0.2982211709022522, + "learning_rate": 7.538217482477516e-06, + "loss": 0.0326, + "step": 83180 + }, + { + "epoch": 5.450435694162353, + "grad_norm": 0.41327691078186035, + "learning_rate": 7.529489251885119e-06, + "loss": 0.0276, + "step": 83190 + }, + { + "epoch": 5.451090873353862, + "grad_norm": 0.24146831035614014, + "learning_rate": 7.520765665803941e-06, + "loss": 0.0408, + "step": 83200 + }, + { + "epoch": 5.451746052545371, + "grad_norm": 0.1645025759935379, + "learning_rate": 7.512046725187949e-06, + "loss": 0.0187, + "step": 83210 + }, + { + "epoch": 5.45240123173688, + "grad_norm": 0.7431649565696716, + "learning_rate": 7.503332430990645e-06, + "loss": 0.0244, + "step": 83220 + }, + { + "epoch": 5.453056410928389, + "grad_norm": 0.35510092973709106, + "learning_rate": 7.494622784165017e-06, + "loss": 0.0334, + "step": 83230 + }, + { + "epoch": 5.453711590119898, + "grad_norm": 0.3607746660709381, + "learning_rate": 7.485917785663543e-06, + "loss": 0.025, + "step": 83240 + }, + { + "epoch": 5.454366769311407, + "grad_norm": 0.21011443436145782, + "learning_rate": 7.477217436438172e-06, + "loss": 0.023, + "step": 83250 + }, + { + "epoch": 5.455021948502916, + "grad_norm": 0.33258599042892456, + "learning_rate": 7.46852173744037e-06, + "loss": 0.0327, + "step": 83260 + }, + { + "epoch": 5.4556771276944245, + "grad_norm": 1.4310277700424194, + "learning_rate": 7.4598306896211004e-06, + "loss": 0.053, + "step": 83270 + }, + { + "epoch": 5.456332306885933, + "grad_norm": 0.26611822843551636, + "learning_rate": 7.451144293930773e-06, + "loss": 0.0358, + "step": 83280 + }, + { + "epoch": 5.456987486077442, + "grad_norm": 0.22963649034500122, + "learning_rate": 7.4424625513193455e-06, + "loss": 0.032, + "step": 83290 + }, + { + "epoch": 5.457642665268951, + "grad_norm": 1.1921919584274292, + "learning_rate": 7.433785462736209e-06, + "loss": 0.0325, + "step": 83300 + }, + { + "epoch": 5.45829784446046, + "grad_norm": 0.3246151804924011, + "learning_rate": 7.425113029130315e-06, + "loss": 0.0254, + "step": 83310 + }, + { + "epoch": 5.4589530236519686, + "grad_norm": 1.174556016921997, + "learning_rate": 7.416445251450044e-06, + "loss": 0.0311, + "step": 83320 + }, + { + "epoch": 5.459608202843477, + "grad_norm": 0.4723874628543854, + "learning_rate": 7.4077821306433035e-06, + "loss": 0.0369, + "step": 83330 + }, + { + "epoch": 5.460263382034986, + "grad_norm": 0.22886508703231812, + "learning_rate": 7.399123667657465e-06, + "loss": 0.0512, + "step": 83340 + }, + { + "epoch": 5.460918561226496, + "grad_norm": 0.7507201433181763, + "learning_rate": 7.390469863439426e-06, + "loss": 0.0443, + "step": 83350 + }, + { + "epoch": 5.461573740418005, + "grad_norm": 0.4875238537788391, + "learning_rate": 7.381820718935512e-06, + "loss": 0.0535, + "step": 83360 + }, + { + "epoch": 5.4622289196095135, + "grad_norm": 0.9425344467163086, + "learning_rate": 7.373176235091633e-06, + "loss": 0.0527, + "step": 83370 + }, + { + "epoch": 5.462884098801022, + "grad_norm": 0.8322470784187317, + "learning_rate": 7.364536412853101e-06, + "loss": 0.028, + "step": 83380 + }, + { + "epoch": 5.463539277992531, + "grad_norm": 0.5701047778129578, + "learning_rate": 7.355901253164776e-06, + "loss": 0.0343, + "step": 83390 + }, + { + "epoch": 5.46419445718404, + "grad_norm": 0.14558479189872742, + "learning_rate": 7.34727075697097e-06, + "loss": 0.0149, + "step": 83400 + }, + { + "epoch": 5.464849636375549, + "grad_norm": 0.2915809154510498, + "learning_rate": 7.3386449252155095e-06, + "loss": 0.0292, + "step": 83410 + }, + { + "epoch": 5.465504815567058, + "grad_norm": 0.18232621252536774, + "learning_rate": 7.330023758841692e-06, + "loss": 0.038, + "step": 83420 + }, + { + "epoch": 5.466159994758566, + "grad_norm": 0.5785380005836487, + "learning_rate": 7.321407258792323e-06, + "loss": 0.0328, + "step": 83430 + }, + { + "epoch": 5.466815173950075, + "grad_norm": 0.5948190093040466, + "learning_rate": 7.312795426009694e-06, + "loss": 0.0293, + "step": 83440 + }, + { + "epoch": 5.467470353141584, + "grad_norm": 0.4840984642505646, + "learning_rate": 7.304188261435585e-06, + "loss": 0.0256, + "step": 83450 + }, + { + "epoch": 5.468125532333093, + "grad_norm": 0.554757297039032, + "learning_rate": 7.295585766011242e-06, + "loss": 0.0315, + "step": 83460 + }, + { + "epoch": 5.468780711524602, + "grad_norm": 0.6805905103683472, + "learning_rate": 7.286987940677436e-06, + "loss": 0.0234, + "step": 83470 + }, + { + "epoch": 5.4694358907161105, + "grad_norm": 0.7258414626121521, + "learning_rate": 7.278394786374426e-06, + "loss": 0.0271, + "step": 83480 + }, + { + "epoch": 5.470091069907619, + "grad_norm": 0.32465261220932007, + "learning_rate": 7.269806304041915e-06, + "loss": 0.0319, + "step": 83490 + }, + { + "epoch": 5.470746249099129, + "grad_norm": 0.10992717742919922, + "learning_rate": 7.261222494619147e-06, + "loss": 0.0401, + "step": 83500 + }, + { + "epoch": 5.471401428290638, + "grad_norm": 0.5841556191444397, + "learning_rate": 7.252643359044825e-06, + "loss": 0.0322, + "step": 83510 + }, + { + "epoch": 5.472056607482147, + "grad_norm": 0.2778913974761963, + "learning_rate": 7.244068898257172e-06, + "loss": 0.0224, + "step": 83520 + }, + { + "epoch": 5.472711786673655, + "grad_norm": 0.2748878300189972, + "learning_rate": 7.2354991131938495e-06, + "loss": 0.045, + "step": 83530 + }, + { + "epoch": 5.473366965865164, + "grad_norm": 0.5894374251365662, + "learning_rate": 7.2269340047920574e-06, + "loss": 0.0395, + "step": 83540 + }, + { + "epoch": 5.474022145056673, + "grad_norm": 0.2481045126914978, + "learning_rate": 7.218373573988435e-06, + "loss": 0.0378, + "step": 83550 + }, + { + "epoch": 5.474677324248182, + "grad_norm": 0.7042310237884521, + "learning_rate": 7.209817821719178e-06, + "loss": 0.0348, + "step": 83560 + }, + { + "epoch": 5.475332503439691, + "grad_norm": 0.6610979437828064, + "learning_rate": 7.201266748919894e-06, + "loss": 0.0373, + "step": 83570 + }, + { + "epoch": 5.4759876826311995, + "grad_norm": 1.4785988330841064, + "learning_rate": 7.19272035652575e-06, + "loss": 0.0336, + "step": 83580 + }, + { + "epoch": 5.476642861822708, + "grad_norm": 0.2295139729976654, + "learning_rate": 7.184178645471334e-06, + "loss": 0.0154, + "step": 83590 + }, + { + "epoch": 5.477298041014217, + "grad_norm": 0.5932558178901672, + "learning_rate": 7.175641616690775e-06, + "loss": 0.0415, + "step": 83600 + }, + { + "epoch": 5.477953220205726, + "grad_norm": 1.272296667098999, + "learning_rate": 7.167109271117661e-06, + "loss": 0.0144, + "step": 83610 + }, + { + "epoch": 5.478608399397235, + "grad_norm": 0.7223988771438599, + "learning_rate": 7.158581609685072e-06, + "loss": 0.0247, + "step": 83620 + }, + { + "epoch": 5.4792635785887445, + "grad_norm": 0.245682492852211, + "learning_rate": 7.150058633325591e-06, + "loss": 0.0458, + "step": 83630 + }, + { + "epoch": 5.479918757780253, + "grad_norm": 0.9257249236106873, + "learning_rate": 7.141540342971287e-06, + "loss": 0.0357, + "step": 83640 + }, + { + "epoch": 5.480573936971762, + "grad_norm": 1.0501682758331299, + "learning_rate": 7.133026739553677e-06, + "loss": 0.0438, + "step": 83650 + }, + { + "epoch": 5.481229116163271, + "grad_norm": 0.32330408692359924, + "learning_rate": 7.124517824003829e-06, + "loss": 0.0145, + "step": 83660 + }, + { + "epoch": 5.48188429535478, + "grad_norm": 0.7643765211105347, + "learning_rate": 7.116013597252236e-06, + "loss": 0.0424, + "step": 83670 + }, + { + "epoch": 5.4825394745462885, + "grad_norm": 0.651641309261322, + "learning_rate": 7.107514060228921e-06, + "loss": 0.0268, + "step": 83680 + }, + { + "epoch": 5.483194653737797, + "grad_norm": 0.12362263351678848, + "learning_rate": 7.099019213863378e-06, + "loss": 0.032, + "step": 83690 + }, + { + "epoch": 5.483849832929306, + "grad_norm": 0.359731525182724, + "learning_rate": 7.090529059084611e-06, + "loss": 0.0264, + "step": 83700 + }, + { + "epoch": 5.484505012120815, + "grad_norm": 0.5441693067550659, + "learning_rate": 7.082043596821058e-06, + "loss": 0.0165, + "step": 83710 + }, + { + "epoch": 5.485160191312324, + "grad_norm": 0.30002522468566895, + "learning_rate": 7.07356282800069e-06, + "loss": 0.0284, + "step": 83720 + }, + { + "epoch": 5.485815370503833, + "grad_norm": 0.7269915342330933, + "learning_rate": 7.0650867535509645e-06, + "loss": 0.0223, + "step": 83730 + }, + { + "epoch": 5.486470549695341, + "grad_norm": 0.8752853870391846, + "learning_rate": 7.0566153743987895e-06, + "loss": 0.0415, + "step": 83740 + }, + { + "epoch": 5.48712572888685, + "grad_norm": 0.55778968334198, + "learning_rate": 7.04814869147059e-06, + "loss": 0.0339, + "step": 83750 + }, + { + "epoch": 5.487780908078359, + "grad_norm": 0.2701091766357422, + "learning_rate": 7.0396867056922725e-06, + "loss": 0.0275, + "step": 83760 + }, + { + "epoch": 5.488436087269868, + "grad_norm": 0.2891935706138611, + "learning_rate": 7.031229417989244e-06, + "loss": 0.0251, + "step": 83770 + }, + { + "epoch": 5.489091266461378, + "grad_norm": 0.44959700107574463, + "learning_rate": 7.022776829286348e-06, + "loss": 0.0232, + "step": 83780 + }, + { + "epoch": 5.489746445652886, + "grad_norm": 0.8741547465324402, + "learning_rate": 7.014328940507975e-06, + "loss": 0.0225, + "step": 83790 + }, + { + "epoch": 5.490401624844395, + "grad_norm": 0.754081130027771, + "learning_rate": 7.005885752577945e-06, + "loss": 0.06, + "step": 83800 + }, + { + "epoch": 5.491056804035904, + "grad_norm": 0.3037523329257965, + "learning_rate": 6.997447266419615e-06, + "loss": 0.0256, + "step": 83810 + }, + { + "epoch": 5.491711983227413, + "grad_norm": 0.2640494406223297, + "learning_rate": 6.989013482955797e-06, + "loss": 0.0268, + "step": 83820 + }, + { + "epoch": 5.492367162418922, + "grad_norm": 0.2667750120162964, + "learning_rate": 6.980584403108808e-06, + "loss": 0.0217, + "step": 83830 + }, + { + "epoch": 5.4930223416104305, + "grad_norm": 0.16732245683670044, + "learning_rate": 6.972160027800417e-06, + "loss": 0.0187, + "step": 83840 + }, + { + "epoch": 5.493677520801939, + "grad_norm": 0.14988090097904205, + "learning_rate": 6.963740357951931e-06, + "loss": 0.019, + "step": 83850 + }, + { + "epoch": 5.494332699993448, + "grad_norm": 1.2653391361236572, + "learning_rate": 6.95532539448408e-06, + "loss": 0.0337, + "step": 83860 + }, + { + "epoch": 5.494987879184957, + "grad_norm": 0.22448024153709412, + "learning_rate": 6.946915138317128e-06, + "loss": 0.0316, + "step": 83870 + }, + { + "epoch": 5.495643058376466, + "grad_norm": 0.5337898135185242, + "learning_rate": 6.9385095903708065e-06, + "loss": 0.0165, + "step": 83880 + }, + { + "epoch": 5.4962982375679745, + "grad_norm": 0.5344067215919495, + "learning_rate": 6.930108751564346e-06, + "loss": 0.0158, + "step": 83890 + }, + { + "epoch": 5.496953416759483, + "grad_norm": 0.6050282716751099, + "learning_rate": 6.921712622816428e-06, + "loss": 0.0343, + "step": 83900 + }, + { + "epoch": 5.497608595950993, + "grad_norm": 0.7292431592941284, + "learning_rate": 6.913321205045259e-06, + "loss": 0.0284, + "step": 83910 + }, + { + "epoch": 5.498263775142502, + "grad_norm": 0.48054438829421997, + "learning_rate": 6.904934499168491e-06, + "loss": 0.029, + "step": 83920 + }, + { + "epoch": 5.498918954334011, + "grad_norm": 0.7779446244239807, + "learning_rate": 6.896552506103299e-06, + "loss": 0.0474, + "step": 83930 + }, + { + "epoch": 5.4995741335255195, + "grad_norm": 0.21829916536808014, + "learning_rate": 6.888175226766313e-06, + "loss": 0.011, + "step": 83940 + }, + { + "epoch": 5.500229312717028, + "grad_norm": 0.6637679934501648, + "learning_rate": 6.879802662073686e-06, + "loss": 0.0287, + "step": 83950 + }, + { + "epoch": 5.500884491908537, + "grad_norm": 0.7366859912872314, + "learning_rate": 6.871434812940991e-06, + "loss": 0.0307, + "step": 83960 + }, + { + "epoch": 5.501539671100046, + "grad_norm": 0.4354690909385681, + "learning_rate": 6.863071680283345e-06, + "loss": 0.0227, + "step": 83970 + }, + { + "epoch": 5.502194850291555, + "grad_norm": 0.18402047455310822, + "learning_rate": 6.854713265015339e-06, + "loss": 0.0225, + "step": 83980 + }, + { + "epoch": 5.502850029483064, + "grad_norm": 0.5053089261054993, + "learning_rate": 6.8463595680510025e-06, + "loss": 0.0359, + "step": 83990 + }, + { + "epoch": 5.503505208674572, + "grad_norm": 0.05302373692393303, + "learning_rate": 6.838010590303911e-06, + "loss": 0.0361, + "step": 84000 + }, + { + "epoch": 5.504160387866081, + "grad_norm": 0.26268526911735535, + "learning_rate": 6.829666332687085e-06, + "loss": 0.024, + "step": 84010 + }, + { + "epoch": 5.50481556705759, + "grad_norm": 0.32237210869789124, + "learning_rate": 6.821326796113059e-06, + "loss": 0.0439, + "step": 84020 + }, + { + "epoch": 5.505470746249099, + "grad_norm": 0.10845739394426346, + "learning_rate": 6.812991981493799e-06, + "loss": 0.0453, + "step": 84030 + }, + { + "epoch": 5.506125925440608, + "grad_norm": 0.843927264213562, + "learning_rate": 6.804661889740821e-06, + "loss": 0.0289, + "step": 84040 + }, + { + "epoch": 5.5067811046321165, + "grad_norm": 0.2354244738817215, + "learning_rate": 6.7963365217650625e-06, + "loss": 0.0277, + "step": 84050 + }, + { + "epoch": 5.507436283823626, + "grad_norm": 0.9995350241661072, + "learning_rate": 6.788015878476983e-06, + "loss": 0.0482, + "step": 84060 + }, + { + "epoch": 5.508091463015135, + "grad_norm": 0.31280118227005005, + "learning_rate": 6.779699960786523e-06, + "loss": 0.0222, + "step": 84070 + }, + { + "epoch": 5.508746642206644, + "grad_norm": 0.7301998138427734, + "learning_rate": 6.771388769603104e-06, + "loss": 0.0289, + "step": 84080 + }, + { + "epoch": 5.509401821398153, + "grad_norm": 0.902916669845581, + "learning_rate": 6.763082305835605e-06, + "loss": 0.0408, + "step": 84090 + }, + { + "epoch": 5.510057000589661, + "grad_norm": 0.50876784324646, + "learning_rate": 6.754780570392433e-06, + "loss": 0.0336, + "step": 84100 + }, + { + "epoch": 5.51071217978117, + "grad_norm": 0.28943389654159546, + "learning_rate": 6.7464835641814215e-06, + "loss": 0.0323, + "step": 84110 + }, + { + "epoch": 5.511367358972679, + "grad_norm": 0.20997591316699982, + "learning_rate": 6.7381912881099515e-06, + "loss": 0.0435, + "step": 84120 + }, + { + "epoch": 5.512022538164188, + "grad_norm": 1.4836313724517822, + "learning_rate": 6.72990374308482e-06, + "loss": 0.0598, + "step": 84130 + }, + { + "epoch": 5.512677717355697, + "grad_norm": 1.733152985572815, + "learning_rate": 6.721620930012379e-06, + "loss": 0.0332, + "step": 84140 + }, + { + "epoch": 5.5133328965472055, + "grad_norm": 0.8775144219398499, + "learning_rate": 6.713342849798393e-06, + "loss": 0.0271, + "step": 84150 + }, + { + "epoch": 5.513988075738714, + "grad_norm": 0.6889039278030396, + "learning_rate": 6.705069503348166e-06, + "loss": 0.0276, + "step": 84160 + }, + { + "epoch": 5.514643254930223, + "grad_norm": 0.8176788091659546, + "learning_rate": 6.696800891566435e-06, + "loss": 0.0474, + "step": 84170 + }, + { + "epoch": 5.515298434121732, + "grad_norm": 1.613385558128357, + "learning_rate": 6.688537015357449e-06, + "loss": 0.0432, + "step": 84180 + }, + { + "epoch": 5.515953613313242, + "grad_norm": 0.7401576638221741, + "learning_rate": 6.680277875624941e-06, + "loss": 0.0218, + "step": 84190 + }, + { + "epoch": 5.51660879250475, + "grad_norm": 0.6381421089172363, + "learning_rate": 6.672023473272121e-06, + "loss": 0.0263, + "step": 84200 + }, + { + "epoch": 5.517263971696259, + "grad_norm": 0.23040910065174103, + "learning_rate": 6.663773809201667e-06, + "loss": 0.0319, + "step": 84210 + }, + { + "epoch": 5.517919150887768, + "grad_norm": 0.12076546251773834, + "learning_rate": 6.655528884315748e-06, + "loss": 0.03, + "step": 84220 + }, + { + "epoch": 5.518574330079277, + "grad_norm": 0.5173049569129944, + "learning_rate": 6.647288699516036e-06, + "loss": 0.0559, + "step": 84230 + }, + { + "epoch": 5.519229509270786, + "grad_norm": 0.5028664469718933, + "learning_rate": 6.639053255703642e-06, + "loss": 0.0338, + "step": 84240 + }, + { + "epoch": 5.5198846884622945, + "grad_norm": 0.5829650163650513, + "learning_rate": 6.6308225537791925e-06, + "loss": 0.044, + "step": 84250 + }, + { + "epoch": 5.520539867653803, + "grad_norm": 0.0696604922413826, + "learning_rate": 6.622596594642777e-06, + "loss": 0.0153, + "step": 84260 + }, + { + "epoch": 5.521195046845312, + "grad_norm": 2.1276557445526123, + "learning_rate": 6.614375379193999e-06, + "loss": 0.0496, + "step": 84270 + }, + { + "epoch": 5.521850226036821, + "grad_norm": 0.3450150191783905, + "learning_rate": 6.606158908331889e-06, + "loss": 0.0106, + "step": 84280 + }, + { + "epoch": 5.52250540522833, + "grad_norm": 0.09661705791950226, + "learning_rate": 6.597947182955005e-06, + "loss": 0.0273, + "step": 84290 + }, + { + "epoch": 5.523160584419839, + "grad_norm": 0.6932510733604431, + "learning_rate": 6.589740203961359e-06, + "loss": 0.0303, + "step": 84300 + }, + { + "epoch": 5.523815763611347, + "grad_norm": 0.22906282544136047, + "learning_rate": 6.581537972248464e-06, + "loss": 0.0254, + "step": 84310 + }, + { + "epoch": 5.524470942802856, + "grad_norm": 0.2434324473142624, + "learning_rate": 6.573340488713275e-06, + "loss": 0.012, + "step": 84320 + }, + { + "epoch": 5.525126121994365, + "grad_norm": 1.0487607717514038, + "learning_rate": 6.565147754252304e-06, + "loss": 0.029, + "step": 84330 + }, + { + "epoch": 5.525781301185875, + "grad_norm": 0.3171430826187134, + "learning_rate": 6.556959769761462e-06, + "loss": 0.0333, + "step": 84340 + }, + { + "epoch": 5.5264364803773836, + "grad_norm": 0.5708022117614746, + "learning_rate": 6.548776536136192e-06, + "loss": 0.0288, + "step": 84350 + }, + { + "epoch": 5.527091659568892, + "grad_norm": 0.7202521562576294, + "learning_rate": 6.54059805427138e-06, + "loss": 0.031, + "step": 84360 + }, + { + "epoch": 5.527746838760401, + "grad_norm": 0.5152254104614258, + "learning_rate": 6.532424325061437e-06, + "loss": 0.029, + "step": 84370 + }, + { + "epoch": 5.52840201795191, + "grad_norm": 0.3626581132411957, + "learning_rate": 6.5242553494001935e-06, + "loss": 0.0585, + "step": 84380 + }, + { + "epoch": 5.529057197143419, + "grad_norm": 0.8775494694709778, + "learning_rate": 6.516091128181046e-06, + "loss": 0.0249, + "step": 84390 + }, + { + "epoch": 5.529712376334928, + "grad_norm": 1.5630563497543335, + "learning_rate": 6.507931662296785e-06, + "loss": 0.0285, + "step": 84400 + }, + { + "epoch": 5.530367555526436, + "grad_norm": 0.5709458589553833, + "learning_rate": 6.4997769526397354e-06, + "loss": 0.0168, + "step": 84410 + }, + { + "epoch": 5.531022734717945, + "grad_norm": 0.8769680261611938, + "learning_rate": 6.491627000101663e-06, + "loss": 0.0205, + "step": 84420 + }, + { + "epoch": 5.531677913909454, + "grad_norm": 0.3131439685821533, + "learning_rate": 6.4834818055738545e-06, + "loss": 0.0212, + "step": 84430 + }, + { + "epoch": 5.532333093100963, + "grad_norm": 0.17634792625904083, + "learning_rate": 6.4753413699470465e-06, + "loss": 0.0372, + "step": 84440 + }, + { + "epoch": 5.532988272292472, + "grad_norm": 1.2100284099578857, + "learning_rate": 6.467205694111478e-06, + "loss": 0.0271, + "step": 84450 + }, + { + "epoch": 5.5336434514839805, + "grad_norm": 0.27866193652153015, + "learning_rate": 6.459074778956836e-06, + "loss": 0.0337, + "step": 84460 + }, + { + "epoch": 5.53429863067549, + "grad_norm": 0.9047338962554932, + "learning_rate": 6.450948625372311e-06, + "loss": 0.0199, + "step": 84470 + }, + { + "epoch": 5.534953809866998, + "grad_norm": 0.3302791118621826, + "learning_rate": 6.442827234246585e-06, + "loss": 0.0281, + "step": 84480 + }, + { + "epoch": 5.535608989058508, + "grad_norm": 0.2061413824558258, + "learning_rate": 6.43471060646777e-06, + "loss": 0.0286, + "step": 84490 + }, + { + "epoch": 5.536264168250017, + "grad_norm": 0.03651725873351097, + "learning_rate": 6.426598742923512e-06, + "loss": 0.0355, + "step": 84500 + }, + { + "epoch": 5.5369193474415255, + "grad_norm": 0.23845632374286652, + "learning_rate": 6.418491644500885e-06, + "loss": 0.0555, + "step": 84510 + }, + { + "epoch": 5.537574526633034, + "grad_norm": 0.1347522884607315, + "learning_rate": 6.410389312086512e-06, + "loss": 0.0248, + "step": 84520 + }, + { + "epoch": 5.538229705824543, + "grad_norm": 0.46901005506515503, + "learning_rate": 6.4022917465664135e-06, + "loss": 0.0244, + "step": 84530 + }, + { + "epoch": 5.538884885016052, + "grad_norm": 1.5142303705215454, + "learning_rate": 6.3941989488261575e-06, + "loss": 0.0461, + "step": 84540 + }, + { + "epoch": 5.539540064207561, + "grad_norm": 0.7391736507415771, + "learning_rate": 6.386110919750726e-06, + "loss": 0.0259, + "step": 84550 + }, + { + "epoch": 5.5401952433990695, + "grad_norm": 0.5583258271217346, + "learning_rate": 6.378027660224651e-06, + "loss": 0.0284, + "step": 84560 + }, + { + "epoch": 5.540850422590578, + "grad_norm": 0.948661744594574, + "learning_rate": 6.369949171131856e-06, + "loss": 0.0337, + "step": 84570 + }, + { + "epoch": 5.541505601782087, + "grad_norm": 0.27530306577682495, + "learning_rate": 6.361875453355848e-06, + "loss": 0.027, + "step": 84580 + }, + { + "epoch": 5.542160780973596, + "grad_norm": 0.2788335680961609, + "learning_rate": 6.353806507779525e-06, + "loss": 0.0218, + "step": 84590 + }, + { + "epoch": 5.542815960165105, + "grad_norm": 0.22271600365638733, + "learning_rate": 6.3457423352853044e-06, + "loss": 0.0297, + "step": 84600 + }, + { + "epoch": 5.543471139356614, + "grad_norm": 0.32235243916511536, + "learning_rate": 6.337682936755062e-06, + "loss": 0.0236, + "step": 84610 + }, + { + "epoch": 5.544126318548123, + "grad_norm": 0.21803492307662964, + "learning_rate": 6.329628313070179e-06, + "loss": 0.0175, + "step": 84620 + }, + { + "epoch": 5.544781497739632, + "grad_norm": 0.0835493728518486, + "learning_rate": 6.321578465111477e-06, + "loss": 0.0262, + "step": 84630 + }, + { + "epoch": 5.545436676931141, + "grad_norm": 0.25792568922042847, + "learning_rate": 6.31353339375928e-06, + "loss": 0.0238, + "step": 84640 + }, + { + "epoch": 5.54609185612265, + "grad_norm": 0.04551071673631668, + "learning_rate": 6.305493099893395e-06, + "loss": 0.0247, + "step": 84650 + }, + { + "epoch": 5.546747035314159, + "grad_norm": 0.6214466691017151, + "learning_rate": 6.297457584393096e-06, + "loss": 0.0323, + "step": 84660 + }, + { + "epoch": 5.547402214505667, + "grad_norm": 1.1996901035308838, + "learning_rate": 6.289426848137126e-06, + "loss": 0.0268, + "step": 84670 + }, + { + "epoch": 5.548057393697176, + "grad_norm": 0.29997172951698303, + "learning_rate": 6.281400892003713e-06, + "loss": 0.0308, + "step": 84680 + }, + { + "epoch": 5.548712572888685, + "grad_norm": 0.5096113681793213, + "learning_rate": 6.273379716870581e-06, + "loss": 0.029, + "step": 84690 + }, + { + "epoch": 5.549367752080194, + "grad_norm": 0.31194689869880676, + "learning_rate": 6.265363323614887e-06, + "loss": 0.0257, + "step": 84700 + }, + { + "epoch": 5.550022931271703, + "grad_norm": 0.35041549801826477, + "learning_rate": 6.257351713113308e-06, + "loss": 0.0257, + "step": 84710 + }, + { + "epoch": 5.5506781104632115, + "grad_norm": 0.6575756669044495, + "learning_rate": 6.249344886241981e-06, + "loss": 0.0313, + "step": 84720 + }, + { + "epoch": 5.55133328965472, + "grad_norm": 1.1566565036773682, + "learning_rate": 6.2413428438765295e-06, + "loss": 0.0392, + "step": 84730 + }, + { + "epoch": 5.551988468846229, + "grad_norm": 0.7483593821525574, + "learning_rate": 6.233345586892026e-06, + "loss": 0.0233, + "step": 84740 + }, + { + "epoch": 5.552643648037739, + "grad_norm": 0.5489878058433533, + "learning_rate": 6.225353116163052e-06, + "loss": 0.0246, + "step": 84750 + }, + { + "epoch": 5.553298827229247, + "grad_norm": 1.013559341430664, + "learning_rate": 6.217365432563627e-06, + "loss": 0.0405, + "step": 84760 + }, + { + "epoch": 5.553954006420756, + "grad_norm": 0.10437938570976257, + "learning_rate": 6.209382536967318e-06, + "loss": 0.0133, + "step": 84770 + }, + { + "epoch": 5.554609185612265, + "grad_norm": 0.30212998390197754, + "learning_rate": 6.201404430247082e-06, + "loss": 0.0384, + "step": 84780 + }, + { + "epoch": 5.555264364803774, + "grad_norm": 0.1463378220796585, + "learning_rate": 6.1934311132754165e-06, + "loss": 0.0249, + "step": 84790 + }, + { + "epoch": 5.555919543995283, + "grad_norm": 1.2814148664474487, + "learning_rate": 6.185462586924257e-06, + "loss": 0.0415, + "step": 84800 + }, + { + "epoch": 5.556574723186792, + "grad_norm": 0.9584314227104187, + "learning_rate": 6.177498852065039e-06, + "loss": 0.0448, + "step": 84810 + }, + { + "epoch": 5.5572299023783005, + "grad_norm": 0.2448907345533371, + "learning_rate": 6.169539909568655e-06, + "loss": 0.0417, + "step": 84820 + }, + { + "epoch": 5.557885081569809, + "grad_norm": 0.18818619847297668, + "learning_rate": 6.161585760305488e-06, + "loss": 0.0203, + "step": 84830 + }, + { + "epoch": 5.558540260761318, + "grad_norm": 0.6650078892707825, + "learning_rate": 6.153636405145391e-06, + "loss": 0.02, + "step": 84840 + }, + { + "epoch": 5.559195439952827, + "grad_norm": 0.7819424271583557, + "learning_rate": 6.145691844957702e-06, + "loss": 0.0326, + "step": 84850 + }, + { + "epoch": 5.559850619144336, + "grad_norm": 0.2076312154531479, + "learning_rate": 6.1377520806112145e-06, + "loss": 0.0316, + "step": 84860 + }, + { + "epoch": 5.560505798335845, + "grad_norm": 0.45930206775665283, + "learning_rate": 6.129817112974218e-06, + "loss": 0.0323, + "step": 84870 + }, + { + "epoch": 5.561160977527353, + "grad_norm": 0.16445904970169067, + "learning_rate": 6.121886942914457e-06, + "loss": 0.0354, + "step": 84880 + }, + { + "epoch": 5.561816156718862, + "grad_norm": 0.8653443455696106, + "learning_rate": 6.11396157129917e-06, + "loss": 0.0433, + "step": 84890 + }, + { + "epoch": 5.562471335910372, + "grad_norm": 0.979309618473053, + "learning_rate": 6.106040998995066e-06, + "loss": 0.0625, + "step": 84900 + }, + { + "epoch": 5.563126515101881, + "grad_norm": 0.18302705883979797, + "learning_rate": 6.098125226868329e-06, + "loss": 0.025, + "step": 84910 + }, + { + "epoch": 5.5637816942933895, + "grad_norm": 0.7636023163795471, + "learning_rate": 6.090214255784604e-06, + "loss": 0.03, + "step": 84920 + }, + { + "epoch": 5.564436873484898, + "grad_norm": 0.8827638626098633, + "learning_rate": 6.0823080866090346e-06, + "loss": 0.0347, + "step": 84930 + }, + { + "epoch": 5.565092052676407, + "grad_norm": 0.2926449477672577, + "learning_rate": 6.074406720206233e-06, + "loss": 0.0215, + "step": 84940 + }, + { + "epoch": 5.565747231867916, + "grad_norm": 0.14975838363170624, + "learning_rate": 6.066510157440259e-06, + "loss": 0.0257, + "step": 84950 + }, + { + "epoch": 5.566402411059425, + "grad_norm": 0.6425701975822449, + "learning_rate": 6.058618399174676e-06, + "loss": 0.0259, + "step": 84960 + }, + { + "epoch": 5.567057590250934, + "grad_norm": 0.9024213552474976, + "learning_rate": 6.050731446272528e-06, + "loss": 0.0452, + "step": 84970 + }, + { + "epoch": 5.567712769442442, + "grad_norm": 0.40766263008117676, + "learning_rate": 6.042849299596321e-06, + "loss": 0.0473, + "step": 84980 + }, + { + "epoch": 5.568367948633951, + "grad_norm": 0.8570975661277771, + "learning_rate": 6.034971960008013e-06, + "loss": 0.0164, + "step": 84990 + }, + { + "epoch": 5.56902312782546, + "grad_norm": 0.8023928999900818, + "learning_rate": 6.027099428369082e-06, + "loss": 0.0299, + "step": 85000 + }, + { + "epoch": 5.569678307016969, + "grad_norm": 0.44330576062202454, + "learning_rate": 6.019231705540435e-06, + "loss": 0.0455, + "step": 85010 + }, + { + "epoch": 5.570333486208478, + "grad_norm": 0.6982066035270691, + "learning_rate": 6.011368792382488e-06, + "loss": 0.0383, + "step": 85020 + }, + { + "epoch": 5.570988665399987, + "grad_norm": 0.4070359468460083, + "learning_rate": 6.00351068975511e-06, + "loss": 0.0376, + "step": 85030 + }, + { + "epoch": 5.571643844591495, + "grad_norm": 0.5921809673309326, + "learning_rate": 5.9956573985176674e-06, + "loss": 0.0244, + "step": 85040 + }, + { + "epoch": 5.572299023783005, + "grad_norm": 0.6675366163253784, + "learning_rate": 5.987808919528959e-06, + "loss": 0.0389, + "step": 85050 + }, + { + "epoch": 5.572954202974514, + "grad_norm": 0.9555246233940125, + "learning_rate": 5.979965253647307e-06, + "loss": 0.019, + "step": 85060 + }, + { + "epoch": 5.573609382166023, + "grad_norm": 0.4306662380695343, + "learning_rate": 5.9721264017304655e-06, + "loss": 0.0289, + "step": 85070 + }, + { + "epoch": 5.5742645613575315, + "grad_norm": 0.192369744181633, + "learning_rate": 5.964292364635682e-06, + "loss": 0.0125, + "step": 85080 + }, + { + "epoch": 5.57491974054904, + "grad_norm": 0.6354840397834778, + "learning_rate": 5.956463143219682e-06, + "loss": 0.0206, + "step": 85090 + }, + { + "epoch": 5.575574919740549, + "grad_norm": 0.13725580275058746, + "learning_rate": 5.948638738338663e-06, + "loss": 0.0279, + "step": 85100 + }, + { + "epoch": 5.576230098932058, + "grad_norm": 0.43551480770111084, + "learning_rate": 5.940819150848271e-06, + "loss": 0.0247, + "step": 85110 + }, + { + "epoch": 5.576885278123567, + "grad_norm": 0.34231001138687134, + "learning_rate": 5.933004381603663e-06, + "loss": 0.0534, + "step": 85120 + }, + { + "epoch": 5.5775404573150755, + "grad_norm": 0.5011656880378723, + "learning_rate": 5.92519443145943e-06, + "loss": 0.032, + "step": 85130 + }, + { + "epoch": 5.578195636506584, + "grad_norm": 0.5348939299583435, + "learning_rate": 5.917389301269671e-06, + "loss": 0.023, + "step": 85140 + }, + { + "epoch": 5.578850815698093, + "grad_norm": 0.47383758425712585, + "learning_rate": 5.9095889918879435e-06, + "loss": 0.0181, + "step": 85150 + }, + { + "epoch": 5.579505994889602, + "grad_norm": 0.2740821838378906, + "learning_rate": 5.901793504167286e-06, + "loss": 0.0404, + "step": 85160 + }, + { + "epoch": 5.580161174081111, + "grad_norm": 0.36358317732810974, + "learning_rate": 5.894002838960172e-06, + "loss": 0.0211, + "step": 85170 + }, + { + "epoch": 5.5808163532726205, + "grad_norm": 0.2581685185432434, + "learning_rate": 5.886216997118604e-06, + "loss": 0.026, + "step": 85180 + }, + { + "epoch": 5.581471532464129, + "grad_norm": 0.2264532744884491, + "learning_rate": 5.878435979494023e-06, + "loss": 0.0312, + "step": 85190 + }, + { + "epoch": 5.582126711655638, + "grad_norm": 0.22009259462356567, + "learning_rate": 5.870659786937344e-06, + "loss": 0.0206, + "step": 85200 + }, + { + "epoch": 5.582781890847147, + "grad_norm": 0.0698859840631485, + "learning_rate": 5.862888420298962e-06, + "loss": 0.0288, + "step": 85210 + }, + { + "epoch": 5.583437070038656, + "grad_norm": 0.33872535824775696, + "learning_rate": 5.855121880428743e-06, + "loss": 0.0398, + "step": 85220 + }, + { + "epoch": 5.584092249230165, + "grad_norm": 1.1379462480545044, + "learning_rate": 5.84736016817603e-06, + "loss": 0.0281, + "step": 85230 + }, + { + "epoch": 5.584747428421673, + "grad_norm": 0.9391049146652222, + "learning_rate": 5.839603284389622e-06, + "loss": 0.0296, + "step": 85240 + }, + { + "epoch": 5.585402607613182, + "grad_norm": 0.29479554295539856, + "learning_rate": 5.831851229917812e-06, + "loss": 0.042, + "step": 85250 + }, + { + "epoch": 5.586057786804691, + "grad_norm": 0.15059269964694977, + "learning_rate": 5.824104005608338e-06, + "loss": 0.0224, + "step": 85260 + }, + { + "epoch": 5.5867129659962, + "grad_norm": 0.6409042477607727, + "learning_rate": 5.81636161230843e-06, + "loss": 0.0325, + "step": 85270 + }, + { + "epoch": 5.587368145187709, + "grad_norm": 1.7973190546035767, + "learning_rate": 5.808624050864791e-06, + "loss": 0.0306, + "step": 85280 + }, + { + "epoch": 5.5880233243792174, + "grad_norm": 1.388026237487793, + "learning_rate": 5.800891322123592e-06, + "loss": 0.0372, + "step": 85290 + }, + { + "epoch": 5.588678503570726, + "grad_norm": 0.6113359928131104, + "learning_rate": 5.793163426930454e-06, + "loss": 0.0403, + "step": 85300 + }, + { + "epoch": 5.589333682762236, + "grad_norm": 0.36089497804641724, + "learning_rate": 5.785440366130512e-06, + "loss": 0.0434, + "step": 85310 + }, + { + "epoch": 5.589988861953744, + "grad_norm": 0.21964912116527557, + "learning_rate": 5.777722140568326e-06, + "loss": 0.0282, + "step": 85320 + }, + { + "epoch": 5.590644041145254, + "grad_norm": 0.8978848457336426, + "learning_rate": 5.7700087510879685e-06, + "loss": 0.0337, + "step": 85330 + }, + { + "epoch": 5.591299220336762, + "grad_norm": 1.1181199550628662, + "learning_rate": 5.762300198532933e-06, + "loss": 0.0433, + "step": 85340 + }, + { + "epoch": 5.591954399528271, + "grad_norm": 0.4006832540035248, + "learning_rate": 5.7545964837462606e-06, + "loss": 0.0144, + "step": 85350 + }, + { + "epoch": 5.59260957871978, + "grad_norm": 0.19875407218933105, + "learning_rate": 5.746897607570384e-06, + "loss": 0.0267, + "step": 85360 + }, + { + "epoch": 5.593264757911289, + "grad_norm": 0.11446386575698853, + "learning_rate": 5.7392035708472626e-06, + "loss": 0.0525, + "step": 85370 + }, + { + "epoch": 5.593919937102798, + "grad_norm": 0.7475147247314453, + "learning_rate": 5.731514374418278e-06, + "loss": 0.027, + "step": 85380 + }, + { + "epoch": 5.5945751162943065, + "grad_norm": 0.1538623869419098, + "learning_rate": 5.7238300191243325e-06, + "loss": 0.0435, + "step": 85390 + }, + { + "epoch": 5.595230295485815, + "grad_norm": 0.24127402901649475, + "learning_rate": 5.716150505805762e-06, + "loss": 0.0246, + "step": 85400 + }, + { + "epoch": 5.595885474677324, + "grad_norm": 1.6690104007720947, + "learning_rate": 5.708475835302407e-06, + "loss": 0.0317, + "step": 85410 + }, + { + "epoch": 5.596540653868833, + "grad_norm": 1.420758605003357, + "learning_rate": 5.70080600845353e-06, + "loss": 0.0227, + "step": 85420 + }, + { + "epoch": 5.597195833060342, + "grad_norm": 0.2971959114074707, + "learning_rate": 5.693141026097909e-06, + "loss": 0.0303, + "step": 85430 + }, + { + "epoch": 5.5978510122518506, + "grad_norm": 0.5827977061271667, + "learning_rate": 5.6854808890737745e-06, + "loss": 0.0489, + "step": 85440 + }, + { + "epoch": 5.598506191443359, + "grad_norm": 0.4231318235397339, + "learning_rate": 5.6778255982188225e-06, + "loss": 0.0286, + "step": 85450 + }, + { + "epoch": 5.599161370634869, + "grad_norm": 1.339799404144287, + "learning_rate": 5.670175154370217e-06, + "loss": 0.0504, + "step": 85460 + }, + { + "epoch": 5.599816549826378, + "grad_norm": 0.24142426252365112, + "learning_rate": 5.662529558364616e-06, + "loss": 0.0151, + "step": 85470 + }, + { + "epoch": 5.600471729017887, + "grad_norm": 2.0625, + "learning_rate": 5.654888811038128e-06, + "loss": 0.0316, + "step": 85480 + }, + { + "epoch": 5.6011269082093955, + "grad_norm": 0.15155765414237976, + "learning_rate": 5.647252913226314e-06, + "loss": 0.0239, + "step": 85490 + }, + { + "epoch": 5.601782087400904, + "grad_norm": 0.9490312337875366, + "learning_rate": 5.639621865764255e-06, + "loss": 0.0249, + "step": 85500 + }, + { + "epoch": 5.602437266592413, + "grad_norm": 0.9239888191223145, + "learning_rate": 5.63199566948644e-06, + "loss": 0.0373, + "step": 85510 + }, + { + "epoch": 5.603092445783922, + "grad_norm": 0.33175793290138245, + "learning_rate": 5.6243743252268835e-06, + "loss": 0.0282, + "step": 85520 + }, + { + "epoch": 5.603747624975431, + "grad_norm": 0.22574172914028168, + "learning_rate": 5.616757833819009e-06, + "loss": 0.0225, + "step": 85530 + }, + { + "epoch": 5.60440280416694, + "grad_norm": 0.2778749465942383, + "learning_rate": 5.609146196095794e-06, + "loss": 0.0227, + "step": 85540 + }, + { + "epoch": 5.605057983358448, + "grad_norm": 0.35948196053504944, + "learning_rate": 5.6015394128896005e-06, + "loss": 0.0263, + "step": 85550 + }, + { + "epoch": 5.605713162549957, + "grad_norm": 0.6876695156097412, + "learning_rate": 5.593937485032308e-06, + "loss": 0.0333, + "step": 85560 + }, + { + "epoch": 5.606368341741466, + "grad_norm": 0.9021486043930054, + "learning_rate": 5.586340413355246e-06, + "loss": 0.0454, + "step": 85570 + }, + { + "epoch": 5.607023520932975, + "grad_norm": 0.7274705767631531, + "learning_rate": 5.578748198689227e-06, + "loss": 0.0449, + "step": 85580 + }, + { + "epoch": 5.607678700124484, + "grad_norm": 0.14086033403873444, + "learning_rate": 5.571160841864504e-06, + "loss": 0.0178, + "step": 85590 + }, + { + "epoch": 5.6083338793159925, + "grad_norm": 0.3138097822666168, + "learning_rate": 5.563578343710846e-06, + "loss": 0.0459, + "step": 85600 + }, + { + "epoch": 5.608989058507502, + "grad_norm": 0.21751539409160614, + "learning_rate": 5.556000705057446e-06, + "loss": 0.034, + "step": 85610 + }, + { + "epoch": 5.609644237699011, + "grad_norm": 0.3155394196510315, + "learning_rate": 5.548427926732996e-06, + "loss": 0.0267, + "step": 85620 + }, + { + "epoch": 5.61029941689052, + "grad_norm": 0.39123567938804626, + "learning_rate": 5.540860009565629e-06, + "loss": 0.046, + "step": 85630 + }, + { + "epoch": 5.610954596082029, + "grad_norm": 0.08803008496761322, + "learning_rate": 5.533296954382966e-06, + "loss": 0.0529, + "step": 85640 + }, + { + "epoch": 5.611609775273537, + "grad_norm": 0.33604151010513306, + "learning_rate": 5.525738762012095e-06, + "loss": 0.0274, + "step": 85650 + }, + { + "epoch": 5.612264954465046, + "grad_norm": 0.9296669363975525, + "learning_rate": 5.518185433279577e-06, + "loss": 0.0214, + "step": 85660 + }, + { + "epoch": 5.612920133656555, + "grad_norm": 0.8024226427078247, + "learning_rate": 5.510636969011407e-06, + "loss": 0.0288, + "step": 85670 + }, + { + "epoch": 5.613575312848064, + "grad_norm": 0.6123316884040833, + "learning_rate": 5.503093370033096e-06, + "loss": 0.0288, + "step": 85680 + }, + { + "epoch": 5.614230492039573, + "grad_norm": 0.07507157325744629, + "learning_rate": 5.495554637169603e-06, + "loss": 0.0199, + "step": 85690 + }, + { + "epoch": 5.6148856712310815, + "grad_norm": 0.8572289347648621, + "learning_rate": 5.488020771245328e-06, + "loss": 0.0386, + "step": 85700 + }, + { + "epoch": 5.61554085042259, + "grad_norm": 0.15820439159870148, + "learning_rate": 5.480491773084195e-06, + "loss": 0.02, + "step": 85710 + }, + { + "epoch": 5.616196029614099, + "grad_norm": 0.4196818172931671, + "learning_rate": 5.472967643509525e-06, + "loss": 0.0174, + "step": 85720 + }, + { + "epoch": 5.616851208805608, + "grad_norm": 0.2837105691432953, + "learning_rate": 5.46544838334419e-06, + "loss": 0.029, + "step": 85730 + }, + { + "epoch": 5.617506387997118, + "grad_norm": 0.2518863081932068, + "learning_rate": 5.4579339934104525e-06, + "loss": 0.0284, + "step": 85740 + }, + { + "epoch": 5.6181615671886265, + "grad_norm": 0.9305800795555115, + "learning_rate": 5.450424474530091e-06, + "loss": 0.0431, + "step": 85750 + }, + { + "epoch": 5.618816746380135, + "grad_norm": 0.3289091885089874, + "learning_rate": 5.442919827524323e-06, + "loss": 0.0153, + "step": 85760 + }, + { + "epoch": 5.619471925571644, + "grad_norm": 0.17634157836437225, + "learning_rate": 5.435420053213863e-06, + "loss": 0.05, + "step": 85770 + }, + { + "epoch": 5.620127104763153, + "grad_norm": 0.9054521918296814, + "learning_rate": 5.427925152418845e-06, + "loss": 0.0426, + "step": 85780 + }, + { + "epoch": 5.620782283954662, + "grad_norm": 0.6224214434623718, + "learning_rate": 5.420435125958934e-06, + "loss": 0.0323, + "step": 85790 + }, + { + "epoch": 5.6214374631461705, + "grad_norm": 0.5484628081321716, + "learning_rate": 5.412949974653209e-06, + "loss": 0.0442, + "step": 85800 + }, + { + "epoch": 5.622092642337679, + "grad_norm": 0.4326564073562622, + "learning_rate": 5.405469699320248e-06, + "loss": 0.0307, + "step": 85810 + }, + { + "epoch": 5.622747821529188, + "grad_norm": 2.171391487121582, + "learning_rate": 5.397994300778064e-06, + "loss": 0.0544, + "step": 85820 + }, + { + "epoch": 5.623403000720697, + "grad_norm": 0.6911959648132324, + "learning_rate": 5.39052377984417e-06, + "loss": 0.0301, + "step": 85830 + }, + { + "epoch": 5.624058179912206, + "grad_norm": 0.4733719229698181, + "learning_rate": 5.383058137335523e-06, + "loss": 0.016, + "step": 85840 + }, + { + "epoch": 5.624713359103715, + "grad_norm": 0.6916825771331787, + "learning_rate": 5.375597374068553e-06, + "loss": 0.0317, + "step": 85850 + }, + { + "epoch": 5.625368538295223, + "grad_norm": 0.08719589561223984, + "learning_rate": 5.368141490859163e-06, + "loss": 0.0609, + "step": 85860 + }, + { + "epoch": 5.626023717486732, + "grad_norm": 0.2758108079433441, + "learning_rate": 5.360690488522729e-06, + "loss": 0.0382, + "step": 85870 + }, + { + "epoch": 5.626678896678241, + "grad_norm": 0.6982370018959045, + "learning_rate": 5.353244367874055e-06, + "loss": 0.0448, + "step": 85880 + }, + { + "epoch": 5.627334075869751, + "grad_norm": 0.07095380127429962, + "learning_rate": 5.345803129727456e-06, + "loss": 0.0146, + "step": 85890 + }, + { + "epoch": 5.62798925506126, + "grad_norm": 0.21258337795734406, + "learning_rate": 5.338366774896697e-06, + "loss": 0.0338, + "step": 85900 + }, + { + "epoch": 5.628644434252768, + "grad_norm": 0.3288947641849518, + "learning_rate": 5.330935304194995e-06, + "loss": 0.0306, + "step": 85910 + }, + { + "epoch": 5.629299613444277, + "grad_norm": 0.1660422384738922, + "learning_rate": 5.323508718435044e-06, + "loss": 0.0154, + "step": 85920 + }, + { + "epoch": 5.629954792635786, + "grad_norm": 0.7223925590515137, + "learning_rate": 5.316087018429011e-06, + "loss": 0.0279, + "step": 85930 + }, + { + "epoch": 5.630609971827295, + "grad_norm": 0.5621246695518494, + "learning_rate": 5.30867020498853e-06, + "loss": 0.0424, + "step": 85940 + }, + { + "epoch": 5.631265151018804, + "grad_norm": 0.2763294279575348, + "learning_rate": 5.30125827892467e-06, + "loss": 0.0364, + "step": 85950 + }, + { + "epoch": 5.6319203302103125, + "grad_norm": 1.910145878791809, + "learning_rate": 5.293851241048015e-06, + "loss": 0.0471, + "step": 85960 + }, + { + "epoch": 5.632575509401821, + "grad_norm": 0.15865060687065125, + "learning_rate": 5.286449092168555e-06, + "loss": 0.0295, + "step": 85970 + }, + { + "epoch": 5.63323068859333, + "grad_norm": 1.752842664718628, + "learning_rate": 5.2790518330958105e-06, + "loss": 0.0272, + "step": 85980 + }, + { + "epoch": 5.633885867784839, + "grad_norm": 0.5864384174346924, + "learning_rate": 5.271659464638717e-06, + "loss": 0.0255, + "step": 85990 + }, + { + "epoch": 5.634541046976348, + "grad_norm": 0.3551011085510254, + "learning_rate": 5.264271987605701e-06, + "loss": 0.0353, + "step": 86000 + }, + { + "epoch": 5.6351962261678565, + "grad_norm": 0.8267326354980469, + "learning_rate": 5.256889402804632e-06, + "loss": 0.02, + "step": 86010 + }, + { + "epoch": 5.635851405359366, + "grad_norm": 0.74532550573349, + "learning_rate": 5.249511711042881e-06, + "loss": 0.0568, + "step": 86020 + }, + { + "epoch": 5.636506584550874, + "grad_norm": 0.2632356286048889, + "learning_rate": 5.242138913127237e-06, + "loss": 0.0149, + "step": 86030 + }, + { + "epoch": 5.637161763742384, + "grad_norm": 0.21590059995651245, + "learning_rate": 5.234771009863987e-06, + "loss": 0.0324, + "step": 86040 + }, + { + "epoch": 5.637816942933893, + "grad_norm": 0.08108416944742203, + "learning_rate": 5.227408002058881e-06, + "loss": 0.0277, + "step": 86050 + }, + { + "epoch": 5.6384721221254015, + "grad_norm": 0.058859679847955704, + "learning_rate": 5.220049890517126e-06, + "loss": 0.0149, + "step": 86060 + }, + { + "epoch": 5.63912730131691, + "grad_norm": 0.19560450315475464, + "learning_rate": 5.212696676043377e-06, + "loss": 0.0268, + "step": 86070 + }, + { + "epoch": 5.639782480508419, + "grad_norm": 0.40965673327445984, + "learning_rate": 5.205348359441798e-06, + "loss": 0.0389, + "step": 86080 + }, + { + "epoch": 5.640437659699928, + "grad_norm": 0.47935572266578674, + "learning_rate": 5.198004941515966e-06, + "loss": 0.0232, + "step": 86090 + }, + { + "epoch": 5.641092838891437, + "grad_norm": 0.21191613376140594, + "learning_rate": 5.190666423068952e-06, + "loss": 0.0283, + "step": 86100 + }, + { + "epoch": 5.641748018082946, + "grad_norm": 0.34002500772476196, + "learning_rate": 5.183332804903284e-06, + "loss": 0.0305, + "step": 86110 + }, + { + "epoch": 5.642403197274454, + "grad_norm": 0.909256100654602, + "learning_rate": 5.176004087820979e-06, + "loss": 0.0222, + "step": 86120 + }, + { + "epoch": 5.643058376465963, + "grad_norm": 0.5787336230278015, + "learning_rate": 5.168680272623455e-06, + "loss": 0.0213, + "step": 86130 + }, + { + "epoch": 5.643713555657472, + "grad_norm": 0.07856592535972595, + "learning_rate": 5.161361360111661e-06, + "loss": 0.0277, + "step": 86140 + }, + { + "epoch": 5.644368734848981, + "grad_norm": 0.27709174156188965, + "learning_rate": 5.154047351085984e-06, + "loss": 0.0463, + "step": 86150 + }, + { + "epoch": 5.64502391404049, + "grad_norm": 0.4343813955783844, + "learning_rate": 5.1467382463462385e-06, + "loss": 0.0274, + "step": 86160 + }, + { + "epoch": 5.645679093231999, + "grad_norm": 0.2831363379955292, + "learning_rate": 5.139434046691788e-06, + "loss": 0.0191, + "step": 86170 + }, + { + "epoch": 5.646334272423508, + "grad_norm": 0.24474884569644928, + "learning_rate": 5.13213475292137e-06, + "loss": 0.0232, + "step": 86180 + }, + { + "epoch": 5.646989451615017, + "grad_norm": 0.53865647315979, + "learning_rate": 5.12484036583325e-06, + "loss": 0.0492, + "step": 86190 + }, + { + "epoch": 5.647644630806526, + "grad_norm": 0.1948166787624359, + "learning_rate": 5.117550886225109e-06, + "loss": 0.0383, + "step": 86200 + }, + { + "epoch": 5.648299809998035, + "grad_norm": 1.1594544649124146, + "learning_rate": 5.11026631489413e-06, + "loss": 0.0304, + "step": 86210 + }, + { + "epoch": 5.648954989189543, + "grad_norm": 0.9201173782348633, + "learning_rate": 5.102986652636926e-06, + "loss": 0.0247, + "step": 86220 + }, + { + "epoch": 5.649610168381052, + "grad_norm": 0.19098861515522003, + "learning_rate": 5.095711900249606e-06, + "loss": 0.0264, + "step": 86230 + }, + { + "epoch": 5.650265347572561, + "grad_norm": 0.4044479727745056, + "learning_rate": 5.088442058527715e-06, + "loss": 0.0146, + "step": 86240 + }, + { + "epoch": 5.65092052676407, + "grad_norm": 0.2394205629825592, + "learning_rate": 5.08117712826629e-06, + "loss": 0.0175, + "step": 86250 + }, + { + "epoch": 5.651575705955579, + "grad_norm": 1.063957929611206, + "learning_rate": 5.073917110259785e-06, + "loss": 0.053, + "step": 86260 + }, + { + "epoch": 5.6522308851470875, + "grad_norm": 0.8265037536621094, + "learning_rate": 5.066662005302175e-06, + "loss": 0.0374, + "step": 86270 + }, + { + "epoch": 5.652886064338596, + "grad_norm": 0.48249927163124084, + "learning_rate": 5.059411814186838e-06, + "loss": 0.0248, + "step": 86280 + }, + { + "epoch": 5.653541243530105, + "grad_norm": 0.3568542003631592, + "learning_rate": 5.05216653770666e-06, + "loss": 0.0255, + "step": 86290 + }, + { + "epoch": 5.654196422721615, + "grad_norm": 0.481319785118103, + "learning_rate": 5.044926176653969e-06, + "loss": 0.025, + "step": 86300 + }, + { + "epoch": 5.654851601913123, + "grad_norm": 0.7378474473953247, + "learning_rate": 5.037690731820571e-06, + "loss": 0.0331, + "step": 86310 + }, + { + "epoch": 5.6555067811046325, + "grad_norm": 0.5720626711845398, + "learning_rate": 5.030460203997711e-06, + "loss": 0.0362, + "step": 86320 + }, + { + "epoch": 5.656161960296141, + "grad_norm": 0.6370004415512085, + "learning_rate": 5.0232345939761174e-06, + "loss": 0.037, + "step": 86330 + }, + { + "epoch": 5.65681713948765, + "grad_norm": 0.5337473154067993, + "learning_rate": 5.016013902545957e-06, + "loss": 0.0201, + "step": 86340 + }, + { + "epoch": 5.657472318679159, + "grad_norm": 0.6324433088302612, + "learning_rate": 5.008798130496889e-06, + "loss": 0.0292, + "step": 86350 + }, + { + "epoch": 5.658127497870668, + "grad_norm": 0.3450187146663666, + "learning_rate": 5.001587278618008e-06, + "loss": 0.0472, + "step": 86360 + }, + { + "epoch": 5.6587826770621765, + "grad_norm": 0.5534144639968872, + "learning_rate": 4.994381347697891e-06, + "loss": 0.024, + "step": 86370 + }, + { + "epoch": 5.659437856253685, + "grad_norm": 0.6514527201652527, + "learning_rate": 4.987180338524572e-06, + "loss": 0.0464, + "step": 86380 + }, + { + "epoch": 5.660093035445194, + "grad_norm": 0.07820023596286774, + "learning_rate": 4.9799842518855275e-06, + "loss": 0.0242, + "step": 86390 + }, + { + "epoch": 5.660748214636703, + "grad_norm": 0.11026445776224136, + "learning_rate": 4.9727930885677285e-06, + "loss": 0.0202, + "step": 86400 + }, + { + "epoch": 5.661403393828212, + "grad_norm": 0.3800837993621826, + "learning_rate": 4.965606849357573e-06, + "loss": 0.037, + "step": 86410 + }, + { + "epoch": 5.662058573019721, + "grad_norm": 1.1449939012527466, + "learning_rate": 4.9584255350409445e-06, + "loss": 0.0245, + "step": 86420 + }, + { + "epoch": 5.662713752211229, + "grad_norm": 0.7511789202690125, + "learning_rate": 4.951249146403175e-06, + "loss": 0.0324, + "step": 86430 + }, + { + "epoch": 5.663368931402738, + "grad_norm": 0.7771415114402771, + "learning_rate": 4.9440776842290815e-06, + "loss": 0.0534, + "step": 86440 + }, + { + "epoch": 5.664024110594248, + "grad_norm": 0.7756445407867432, + "learning_rate": 4.936911149302903e-06, + "loss": 0.0312, + "step": 86450 + }, + { + "epoch": 5.664679289785757, + "grad_norm": 0.7171912789344788, + "learning_rate": 4.929749542408379e-06, + "loss": 0.0231, + "step": 86460 + }, + { + "epoch": 5.665334468977266, + "grad_norm": 0.19514018297195435, + "learning_rate": 4.922592864328673e-06, + "loss": 0.0227, + "step": 86470 + }, + { + "epoch": 5.665989648168774, + "grad_norm": 0.21535709500312805, + "learning_rate": 4.915441115846436e-06, + "loss": 0.0169, + "step": 86480 + }, + { + "epoch": 5.666644827360283, + "grad_norm": 0.6318908333778381, + "learning_rate": 4.908294297743782e-06, + "loss": 0.0398, + "step": 86490 + }, + { + "epoch": 5.667300006551792, + "grad_norm": 0.18926803767681122, + "learning_rate": 4.901152410802273e-06, + "loss": 0.0605, + "step": 86500 + }, + { + "epoch": 5.667955185743301, + "grad_norm": 0.10272018611431122, + "learning_rate": 4.894015455802919e-06, + "loss": 0.0194, + "step": 86510 + }, + { + "epoch": 5.66861036493481, + "grad_norm": 0.18254779279232025, + "learning_rate": 4.8868834335262346e-06, + "loss": 0.0269, + "step": 86520 + }, + { + "epoch": 5.6692655441263184, + "grad_norm": 0.2851342260837555, + "learning_rate": 4.87975634475214e-06, + "loss": 0.0204, + "step": 86530 + }, + { + "epoch": 5.669920723317827, + "grad_norm": 0.47618043422698975, + "learning_rate": 4.8726341902600614e-06, + "loss": 0.0216, + "step": 86540 + }, + { + "epoch": 5.670575902509336, + "grad_norm": 0.11802388727664948, + "learning_rate": 4.865516970828843e-06, + "loss": 0.0326, + "step": 86550 + }, + { + "epoch": 5.671231081700845, + "grad_norm": 0.16740131378173828, + "learning_rate": 4.85840468723685e-06, + "loss": 0.0294, + "step": 86560 + }, + { + "epoch": 5.671886260892354, + "grad_norm": 0.16615208983421326, + "learning_rate": 4.85129734026184e-06, + "loss": 0.0253, + "step": 86570 + }, + { + "epoch": 5.672541440083863, + "grad_norm": 0.18417784571647644, + "learning_rate": 4.844194930681073e-06, + "loss": 0.0351, + "step": 86580 + }, + { + "epoch": 5.673196619275371, + "grad_norm": 0.5951090455055237, + "learning_rate": 4.837097459271267e-06, + "loss": 0.0216, + "step": 86590 + }, + { + "epoch": 5.673851798466881, + "grad_norm": 0.1116471141576767, + "learning_rate": 4.830004926808573e-06, + "loss": 0.0169, + "step": 86600 + }, + { + "epoch": 5.67450697765839, + "grad_norm": 0.6572790145874023, + "learning_rate": 4.822917334068627e-06, + "loss": 0.034, + "step": 86610 + }, + { + "epoch": 5.675162156849899, + "grad_norm": 0.14171907305717468, + "learning_rate": 4.81583468182652e-06, + "loss": 0.0629, + "step": 86620 + }, + { + "epoch": 5.6758173360414075, + "grad_norm": 0.8348695039749146, + "learning_rate": 4.808756970856809e-06, + "loss": 0.0273, + "step": 86630 + }, + { + "epoch": 5.676472515232916, + "grad_norm": 0.8304874897003174, + "learning_rate": 4.801684201933476e-06, + "loss": 0.0283, + "step": 86640 + }, + { + "epoch": 5.677127694424425, + "grad_norm": 0.17467743158340454, + "learning_rate": 4.794616375830019e-06, + "loss": 0.0344, + "step": 86650 + }, + { + "epoch": 5.677782873615934, + "grad_norm": 0.429024338722229, + "learning_rate": 4.787553493319341e-06, + "loss": 0.0324, + "step": 86660 + }, + { + "epoch": 5.678438052807443, + "grad_norm": 0.22411276400089264, + "learning_rate": 4.780495555173836e-06, + "loss": 0.0127, + "step": 86670 + }, + { + "epoch": 5.6790932319989516, + "grad_norm": 0.05391024053096771, + "learning_rate": 4.7734425621653475e-06, + "loss": 0.0226, + "step": 86680 + }, + { + "epoch": 5.67974841119046, + "grad_norm": 0.43421390652656555, + "learning_rate": 4.766394515065198e-06, + "loss": 0.0318, + "step": 86690 + }, + { + "epoch": 5.680403590381969, + "grad_norm": 0.3875715434551239, + "learning_rate": 4.75935141464412e-06, + "loss": 0.0371, + "step": 86700 + }, + { + "epoch": 5.681058769573478, + "grad_norm": 0.7553514242172241, + "learning_rate": 4.75231326167237e-06, + "loss": 0.025, + "step": 86710 + }, + { + "epoch": 5.681713948764987, + "grad_norm": 0.4134896695613861, + "learning_rate": 4.745280056919599e-06, + "loss": 0.0302, + "step": 86720 + }, + { + "epoch": 5.6823691279564965, + "grad_norm": 0.1848945915699005, + "learning_rate": 4.738251801154975e-06, + "loss": 0.0245, + "step": 86730 + }, + { + "epoch": 5.683024307148005, + "grad_norm": 0.8416123986244202, + "learning_rate": 4.731228495147061e-06, + "loss": 0.0329, + "step": 86740 + }, + { + "epoch": 5.683679486339514, + "grad_norm": 0.1938173919916153, + "learning_rate": 4.7242101396639595e-06, + "loss": 0.0277, + "step": 86750 + }, + { + "epoch": 5.684334665531023, + "grad_norm": 0.2238098531961441, + "learning_rate": 4.717196735473151e-06, + "loss": 0.0383, + "step": 86760 + }, + { + "epoch": 5.684989844722532, + "grad_norm": 0.18421462178230286, + "learning_rate": 4.710188283341643e-06, + "loss": 0.0212, + "step": 86770 + }, + { + "epoch": 5.685645023914041, + "grad_norm": 0.22822633385658264, + "learning_rate": 4.703184784035835e-06, + "loss": 0.0148, + "step": 86780 + }, + { + "epoch": 5.686300203105549, + "grad_norm": 0.7969011068344116, + "learning_rate": 4.696186238321642e-06, + "loss": 0.0406, + "step": 86790 + }, + { + "epoch": 5.686955382297058, + "grad_norm": 0.31154385209083557, + "learning_rate": 4.689192646964408e-06, + "loss": 0.0191, + "step": 86800 + }, + { + "epoch": 5.687610561488567, + "grad_norm": 0.7841305136680603, + "learning_rate": 4.682204010728952e-06, + "loss": 0.0279, + "step": 86810 + }, + { + "epoch": 5.688265740680076, + "grad_norm": 0.05584254115819931, + "learning_rate": 4.67522033037952e-06, + "loss": 0.0125, + "step": 86820 + }, + { + "epoch": 5.688920919871585, + "grad_norm": 0.5818349719047546, + "learning_rate": 4.668241606679846e-06, + "loss": 0.022, + "step": 86830 + }, + { + "epoch": 5.6895760990630935, + "grad_norm": 0.07811884582042694, + "learning_rate": 4.66126784039313e-06, + "loss": 0.0288, + "step": 86840 + }, + { + "epoch": 5.690231278254602, + "grad_norm": 0.307883083820343, + "learning_rate": 4.6542990322819836e-06, + "loss": 0.0335, + "step": 86850 + }, + { + "epoch": 5.690886457446112, + "grad_norm": 0.34550175070762634, + "learning_rate": 4.64733518310852e-06, + "loss": 0.0236, + "step": 86860 + }, + { + "epoch": 5.69154163663762, + "grad_norm": 0.08821115642786026, + "learning_rate": 4.640376293634291e-06, + "loss": 0.0342, + "step": 86870 + }, + { + "epoch": 5.69219681582913, + "grad_norm": 0.6098953485488892, + "learning_rate": 4.633422364620327e-06, + "loss": 0.0554, + "step": 86880 + }, + { + "epoch": 5.692851995020638, + "grad_norm": 0.41254374384880066, + "learning_rate": 4.6264733968270745e-06, + "loss": 0.0319, + "step": 86890 + }, + { + "epoch": 5.693507174212147, + "grad_norm": 0.13440614938735962, + "learning_rate": 4.619529391014482e-06, + "loss": 0.0136, + "step": 86900 + }, + { + "epoch": 5.694162353403656, + "grad_norm": 0.42861634492874146, + "learning_rate": 4.61259034794192e-06, + "loss": 0.0235, + "step": 86910 + }, + { + "epoch": 5.694817532595165, + "grad_norm": 0.7753854393959045, + "learning_rate": 4.605656268368247e-06, + "loss": 0.035, + "step": 86920 + }, + { + "epoch": 5.695472711786674, + "grad_norm": 0.9608535766601562, + "learning_rate": 4.598727153051735e-06, + "loss": 0.0184, + "step": 86930 + }, + { + "epoch": 5.6961278909781825, + "grad_norm": 0.6410377621650696, + "learning_rate": 4.591803002750178e-06, + "loss": 0.0338, + "step": 86940 + }, + { + "epoch": 5.696783070169691, + "grad_norm": 0.3770337700843811, + "learning_rate": 4.58488381822077e-06, + "loss": 0.0372, + "step": 86950 + }, + { + "epoch": 5.6974382493612, + "grad_norm": 1.2979836463928223, + "learning_rate": 4.5779696002201946e-06, + "loss": 0.0231, + "step": 86960 + }, + { + "epoch": 5.698093428552709, + "grad_norm": 0.557316780090332, + "learning_rate": 4.571060349504563e-06, + "loss": 0.0247, + "step": 86970 + }, + { + "epoch": 5.698748607744218, + "grad_norm": 0.45232439041137695, + "learning_rate": 4.564156066829478e-06, + "loss": 0.0218, + "step": 86980 + }, + { + "epoch": 5.699403786935727, + "grad_norm": 0.3179748058319092, + "learning_rate": 4.557256752949957e-06, + "loss": 0.0316, + "step": 86990 + }, + { + "epoch": 5.700058966127235, + "grad_norm": 0.3712714612483978, + "learning_rate": 4.550362408620534e-06, + "loss": 0.0236, + "step": 87000 + }, + { + "epoch": 5.700714145318745, + "grad_norm": 1.020263433456421, + "learning_rate": 4.543473034595136e-06, + "loss": 0.0445, + "step": 87010 + }, + { + "epoch": 5.701369324510254, + "grad_norm": 0.2716124355792999, + "learning_rate": 4.536588631627192e-06, + "loss": 0.0295, + "step": 87020 + }, + { + "epoch": 5.702024503701763, + "grad_norm": 0.13439424335956573, + "learning_rate": 4.5297092004695565e-06, + "loss": 0.0541, + "step": 87030 + }, + { + "epoch": 5.7026796828932715, + "grad_norm": 0.8255612850189209, + "learning_rate": 4.522834741874555e-06, + "loss": 0.0435, + "step": 87040 + }, + { + "epoch": 5.70333486208478, + "grad_norm": 0.07305286824703217, + "learning_rate": 4.515965256593974e-06, + "loss": 0.0261, + "step": 87050 + }, + { + "epoch": 5.703990041276289, + "grad_norm": 0.7166330814361572, + "learning_rate": 4.509100745379058e-06, + "loss": 0.0553, + "step": 87060 + }, + { + "epoch": 5.704645220467798, + "grad_norm": 0.5899561643600464, + "learning_rate": 4.502241208980479e-06, + "loss": 0.0259, + "step": 87070 + }, + { + "epoch": 5.705300399659307, + "grad_norm": 0.5532768964767456, + "learning_rate": 4.495386648148398e-06, + "loss": 0.0299, + "step": 87080 + }, + { + "epoch": 5.705955578850816, + "grad_norm": 0.6806175112724304, + "learning_rate": 4.488537063632431e-06, + "loss": 0.0422, + "step": 87090 + }, + { + "epoch": 5.706610758042324, + "grad_norm": 0.13019855320453644, + "learning_rate": 4.481692456181608e-06, + "loss": 0.016, + "step": 87100 + }, + { + "epoch": 5.707265937233833, + "grad_norm": 0.9216971397399902, + "learning_rate": 4.474852826544479e-06, + "loss": 0.0368, + "step": 87110 + }, + { + "epoch": 5.707921116425342, + "grad_norm": 0.767454206943512, + "learning_rate": 4.468018175468974e-06, + "loss": 0.0657, + "step": 87120 + }, + { + "epoch": 5.708576295616851, + "grad_norm": 0.15707267820835114, + "learning_rate": 4.461188503702568e-06, + "loss": 0.0177, + "step": 87130 + }, + { + "epoch": 5.709231474808361, + "grad_norm": 0.3213093876838684, + "learning_rate": 4.454363811992107e-06, + "loss": 0.018, + "step": 87140 + }, + { + "epoch": 5.7098866539998685, + "grad_norm": 0.3898251950740814, + "learning_rate": 4.447544101083956e-06, + "loss": 0.0443, + "step": 87150 + }, + { + "epoch": 5.710541833191378, + "grad_norm": 0.7770704627037048, + "learning_rate": 4.440729371723884e-06, + "loss": 0.022, + "step": 87160 + }, + { + "epoch": 5.711197012382887, + "grad_norm": 0.4605502784252167, + "learning_rate": 4.433919624657162e-06, + "loss": 0.0274, + "step": 87170 + }, + { + "epoch": 5.711852191574396, + "grad_norm": 0.10863636434078217, + "learning_rate": 4.427114860628462e-06, + "loss": 0.0186, + "step": 87180 + }, + { + "epoch": 5.712507370765905, + "grad_norm": 0.6499249339103699, + "learning_rate": 4.420315080381982e-06, + "loss": 0.0345, + "step": 87190 + }, + { + "epoch": 5.7131625499574135, + "grad_norm": 0.32175910472869873, + "learning_rate": 4.4135202846613045e-06, + "loss": 0.0722, + "step": 87200 + }, + { + "epoch": 5.713817729148922, + "grad_norm": 0.13983210921287537, + "learning_rate": 4.406730474209525e-06, + "loss": 0.0293, + "step": 87210 + }, + { + "epoch": 5.714472908340431, + "grad_norm": 0.19670702517032623, + "learning_rate": 4.399945649769138e-06, + "loss": 0.0163, + "step": 87220 + }, + { + "epoch": 5.71512808753194, + "grad_norm": 0.17038245499134064, + "learning_rate": 4.393165812082151e-06, + "loss": 0.0186, + "step": 87230 + }, + { + "epoch": 5.715783266723449, + "grad_norm": 0.11179404705762863, + "learning_rate": 4.386390961889969e-06, + "loss": 0.0235, + "step": 87240 + }, + { + "epoch": 5.7164384459149575, + "grad_norm": 0.5398284196853638, + "learning_rate": 4.3796210999334895e-06, + "loss": 0.0258, + "step": 87250 + }, + { + "epoch": 5.717093625106466, + "grad_norm": 0.9057984352111816, + "learning_rate": 4.372856226953065e-06, + "loss": 0.016, + "step": 87260 + }, + { + "epoch": 5.717748804297975, + "grad_norm": 0.37944260239601135, + "learning_rate": 4.366096343688486e-06, + "loss": 0.0487, + "step": 87270 + }, + { + "epoch": 5.718403983489484, + "grad_norm": 0.2516416311264038, + "learning_rate": 4.359341450878995e-06, + "loss": 0.0295, + "step": 87280 + }, + { + "epoch": 5.719059162680994, + "grad_norm": 0.5838364958763123, + "learning_rate": 4.352591549263302e-06, + "loss": 0.0318, + "step": 87290 + }, + { + "epoch": 5.7197143418725025, + "grad_norm": 0.8216543793678284, + "learning_rate": 4.345846639579582e-06, + "loss": 0.0353, + "step": 87300 + }, + { + "epoch": 5.720369521064011, + "grad_norm": 0.5833924412727356, + "learning_rate": 4.339106722565417e-06, + "loss": 0.035, + "step": 87310 + }, + { + "epoch": 5.72102470025552, + "grad_norm": 0.8884563446044922, + "learning_rate": 4.332371798957896e-06, + "loss": 0.0383, + "step": 87320 + }, + { + "epoch": 5.721679879447029, + "grad_norm": 1.034454107284546, + "learning_rate": 4.3256418694935295e-06, + "loss": 0.0569, + "step": 87330 + }, + { + "epoch": 5.722335058638538, + "grad_norm": 0.9945911765098572, + "learning_rate": 4.3189169349083125e-06, + "loss": 0.025, + "step": 87340 + }, + { + "epoch": 5.722990237830047, + "grad_norm": 0.16773739457130432, + "learning_rate": 4.312196995937645e-06, + "loss": 0.0221, + "step": 87350 + }, + { + "epoch": 5.723645417021555, + "grad_norm": 0.24713604152202606, + "learning_rate": 4.305482053316434e-06, + "loss": 0.0225, + "step": 87360 + }, + { + "epoch": 5.724300596213064, + "grad_norm": 0.1588675081729889, + "learning_rate": 4.2987721077789925e-06, + "loss": 0.0344, + "step": 87370 + }, + { + "epoch": 5.724955775404573, + "grad_norm": 0.473019003868103, + "learning_rate": 4.2920671600591325e-06, + "loss": 0.034, + "step": 87380 + }, + { + "epoch": 5.725610954596082, + "grad_norm": 0.16534925997257233, + "learning_rate": 4.285367210890084e-06, + "loss": 0.028, + "step": 87390 + }, + { + "epoch": 5.726266133787591, + "grad_norm": 0.18376865983009338, + "learning_rate": 4.278672261004551e-06, + "loss": 0.027, + "step": 87400 + }, + { + "epoch": 5.7269213129790995, + "grad_norm": 0.3703210651874542, + "learning_rate": 4.2719823111346704e-06, + "loss": 0.0301, + "step": 87410 + }, + { + "epoch": 5.727576492170609, + "grad_norm": 1.0861948728561401, + "learning_rate": 4.265297362012061e-06, + "loss": 0.0496, + "step": 87420 + }, + { + "epoch": 5.728231671362117, + "grad_norm": 0.5996478199958801, + "learning_rate": 4.258617414367766e-06, + "loss": 0.0174, + "step": 87430 + }, + { + "epoch": 5.728886850553627, + "grad_norm": 0.22053812444210052, + "learning_rate": 4.251942468932296e-06, + "loss": 0.0221, + "step": 87440 + }, + { + "epoch": 5.729542029745136, + "grad_norm": 0.17311634123325348, + "learning_rate": 4.2452725264356175e-06, + "loss": 0.0386, + "step": 87450 + }, + { + "epoch": 5.730197208936644, + "grad_norm": 0.05326778069138527, + "learning_rate": 4.2386075876071575e-06, + "loss": 0.0425, + "step": 87460 + }, + { + "epoch": 5.730852388128153, + "grad_norm": 0.5937438607215881, + "learning_rate": 4.231947653175755e-06, + "loss": 0.0115, + "step": 87470 + }, + { + "epoch": 5.731507567319662, + "grad_norm": 0.1284918338060379, + "learning_rate": 4.225292723869761e-06, + "loss": 0.0246, + "step": 87480 + }, + { + "epoch": 5.732162746511171, + "grad_norm": 0.17132499814033508, + "learning_rate": 4.218642800416916e-06, + "loss": 0.0236, + "step": 87490 + }, + { + "epoch": 5.73281792570268, + "grad_norm": 0.2379889339208603, + "learning_rate": 4.211997883544472e-06, + "loss": 0.0217, + "step": 87500 + }, + { + "epoch": 5.7334731048941885, + "grad_norm": 0.21678797900676727, + "learning_rate": 4.205357973979096e-06, + "loss": 0.0239, + "step": 87510 + }, + { + "epoch": 5.734128284085697, + "grad_norm": 0.24181444942951202, + "learning_rate": 4.198723072446925e-06, + "loss": 0.0209, + "step": 87520 + }, + { + "epoch": 5.734783463277206, + "grad_norm": 0.8309301137924194, + "learning_rate": 4.192093179673534e-06, + "loss": 0.0317, + "step": 87530 + }, + { + "epoch": 5.735438642468715, + "grad_norm": 0.42369866371154785, + "learning_rate": 4.185468296383959e-06, + "loss": 0.0345, + "step": 87540 + }, + { + "epoch": 5.736093821660224, + "grad_norm": 0.4007505774497986, + "learning_rate": 4.1788484233027005e-06, + "loss": 0.0237, + "step": 87550 + }, + { + "epoch": 5.736749000851733, + "grad_norm": 0.7454981803894043, + "learning_rate": 4.172233561153677e-06, + "loss": 0.019, + "step": 87560 + }, + { + "epoch": 5.737404180043242, + "grad_norm": 0.08304090052843094, + "learning_rate": 4.165623710660288e-06, + "loss": 0.0185, + "step": 87570 + }, + { + "epoch": 5.738059359234751, + "grad_norm": 0.3263293206691742, + "learning_rate": 4.1590188725453835e-06, + "loss": 0.0171, + "step": 87580 + }, + { + "epoch": 5.73871453842626, + "grad_norm": 1.5317823886871338, + "learning_rate": 4.152419047531259e-06, + "loss": 0.0424, + "step": 87590 + }, + { + "epoch": 5.739369717617769, + "grad_norm": 0.4647505283355713, + "learning_rate": 4.145824236339651e-06, + "loss": 0.0264, + "step": 87600 + }, + { + "epoch": 5.7400248968092775, + "grad_norm": 0.19098182022571564, + "learning_rate": 4.1392344396917695e-06, + "loss": 0.0287, + "step": 87610 + }, + { + "epoch": 5.740680076000786, + "grad_norm": 0.6031236052513123, + "learning_rate": 4.132649658308252e-06, + "loss": 0.024, + "step": 87620 + }, + { + "epoch": 5.741335255192295, + "grad_norm": 0.3483636975288391, + "learning_rate": 4.126069892909213e-06, + "loss": 0.0188, + "step": 87630 + }, + { + "epoch": 5.741990434383804, + "grad_norm": 1.3568052053451538, + "learning_rate": 4.119495144214191e-06, + "loss": 0.0541, + "step": 87640 + }, + { + "epoch": 5.742645613575313, + "grad_norm": 0.23224100470542908, + "learning_rate": 4.112925412942214e-06, + "loss": 0.0147, + "step": 87650 + }, + { + "epoch": 5.743300792766822, + "grad_norm": 0.2032221257686615, + "learning_rate": 4.106360699811712e-06, + "loss": 0.04, + "step": 87660 + }, + { + "epoch": 5.74395597195833, + "grad_norm": 0.33991408348083496, + "learning_rate": 4.099801005540616e-06, + "loss": 0.0169, + "step": 87670 + }, + { + "epoch": 5.744611151149839, + "grad_norm": 0.42095500230789185, + "learning_rate": 4.093246330846256e-06, + "loss": 0.0551, + "step": 87680 + }, + { + "epoch": 5.745266330341348, + "grad_norm": 0.14008954167366028, + "learning_rate": 4.086696676445467e-06, + "loss": 0.0256, + "step": 87690 + }, + { + "epoch": 5.745921509532857, + "grad_norm": 1.433282732963562, + "learning_rate": 4.0801520430544914e-06, + "loss": 0.0289, + "step": 87700 + }, + { + "epoch": 5.746576688724366, + "grad_norm": 0.5082205533981323, + "learning_rate": 4.073612431389062e-06, + "loss": 0.0178, + "step": 87710 + }, + { + "epoch": 5.747231867915875, + "grad_norm": 0.6639462113380432, + "learning_rate": 4.067077842164319e-06, + "loss": 0.0259, + "step": 87720 + }, + { + "epoch": 5.747887047107384, + "grad_norm": 0.15529650449752808, + "learning_rate": 4.0605482760948945e-06, + "loss": 0.0399, + "step": 87730 + }, + { + "epoch": 5.748542226298893, + "grad_norm": 0.33691325783729553, + "learning_rate": 4.054023733894829e-06, + "loss": 0.0177, + "step": 87740 + }, + { + "epoch": 5.749197405490402, + "grad_norm": 1.5033124685287476, + "learning_rate": 4.047504216277653e-06, + "loss": 0.0356, + "step": 87750 + }, + { + "epoch": 5.749852584681911, + "grad_norm": 0.7390540838241577, + "learning_rate": 4.040989723956323e-06, + "loss": 0.0352, + "step": 87760 + }, + { + "epoch": 5.750507763873419, + "grad_norm": 0.7225953936576843, + "learning_rate": 4.03448025764327e-06, + "loss": 0.0164, + "step": 87770 + }, + { + "epoch": 5.751162943064928, + "grad_norm": 0.20714877545833588, + "learning_rate": 4.027975818050339e-06, + "loss": 0.0217, + "step": 87780 + }, + { + "epoch": 5.751818122256437, + "grad_norm": 0.30041754245758057, + "learning_rate": 4.021476405888858e-06, + "loss": 0.0225, + "step": 87790 + }, + { + "epoch": 5.752473301447946, + "grad_norm": 0.13617205619812012, + "learning_rate": 4.01498202186959e-06, + "loss": 0.0177, + "step": 87800 + }, + { + "epoch": 5.753128480639455, + "grad_norm": 0.6828560829162598, + "learning_rate": 4.008492666702751e-06, + "loss": 0.0548, + "step": 87810 + }, + { + "epoch": 5.7537836598309635, + "grad_norm": 0.4164265990257263, + "learning_rate": 4.002008341098001e-06, + "loss": 0.023, + "step": 87820 + }, + { + "epoch": 5.754438839022472, + "grad_norm": 0.3064253330230713, + "learning_rate": 3.995529045764462e-06, + "loss": 0.0199, + "step": 87830 + }, + { + "epoch": 5.755094018213981, + "grad_norm": 0.25497692823410034, + "learning_rate": 3.989054781410712e-06, + "loss": 0.0359, + "step": 87840 + }, + { + "epoch": 5.755749197405491, + "grad_norm": 0.5781642198562622, + "learning_rate": 3.982585548744744e-06, + "loss": 0.0221, + "step": 87850 + }, + { + "epoch": 5.756404376597, + "grad_norm": 0.7034024596214294, + "learning_rate": 3.9761213484740435e-06, + "loss": 0.0445, + "step": 87860 + }, + { + "epoch": 5.7570595557885085, + "grad_norm": 0.4315215051174164, + "learning_rate": 3.9696621813055055e-06, + "loss": 0.0249, + "step": 87870 + }, + { + "epoch": 5.757714734980017, + "grad_norm": 0.40324416756629944, + "learning_rate": 3.9632080479455044e-06, + "loss": 0.0294, + "step": 87880 + }, + { + "epoch": 5.758369914171526, + "grad_norm": 0.16500189900398254, + "learning_rate": 3.956758949099859e-06, + "loss": 0.0171, + "step": 87890 + }, + { + "epoch": 5.759025093363035, + "grad_norm": 0.215603306889534, + "learning_rate": 3.9503148854738385e-06, + "loss": 0.0231, + "step": 87900 + }, + { + "epoch": 5.759680272554544, + "grad_norm": 0.16895122826099396, + "learning_rate": 3.943875857772139e-06, + "loss": 0.0479, + "step": 87910 + }, + { + "epoch": 5.7603354517460525, + "grad_norm": 0.6485070586204529, + "learning_rate": 3.9374418666989375e-06, + "loss": 0.0266, + "step": 87920 + }, + { + "epoch": 5.760990630937561, + "grad_norm": 0.524212658405304, + "learning_rate": 3.931012912957838e-06, + "loss": 0.0155, + "step": 87930 + }, + { + "epoch": 5.76164581012907, + "grad_norm": 0.3931739926338196, + "learning_rate": 3.9245889972519045e-06, + "loss": 0.0089, + "step": 87940 + }, + { + "epoch": 5.762300989320579, + "grad_norm": 0.43657928705215454, + "learning_rate": 3.9181701202836265e-06, + "loss": 0.0349, + "step": 87950 + }, + { + "epoch": 5.762956168512088, + "grad_norm": 0.5756843686103821, + "learning_rate": 3.911756282755003e-06, + "loss": 0.0366, + "step": 87960 + }, + { + "epoch": 5.763611347703597, + "grad_norm": 0.6474247574806213, + "learning_rate": 3.905347485367411e-06, + "loss": 0.0365, + "step": 87970 + }, + { + "epoch": 5.764266526895105, + "grad_norm": 0.6056151390075684, + "learning_rate": 3.898943728821725e-06, + "loss": 0.037, + "step": 87980 + }, + { + "epoch": 5.764921706086614, + "grad_norm": 0.24939711391925812, + "learning_rate": 3.892545013818227e-06, + "loss": 0.0309, + "step": 87990 + }, + { + "epoch": 5.765576885278124, + "grad_norm": 0.24122141301631927, + "learning_rate": 3.886151341056687e-06, + "loss": 0.0226, + "step": 88000 + }, + { + "epoch": 5.766232064469633, + "grad_norm": 0.2907430827617645, + "learning_rate": 3.8797627112363066e-06, + "loss": 0.0274, + "step": 88010 + }, + { + "epoch": 5.766887243661142, + "grad_norm": 0.8032187819480896, + "learning_rate": 3.873379125055743e-06, + "loss": 0.0296, + "step": 88020 + }, + { + "epoch": 5.76754242285265, + "grad_norm": 0.626226007938385, + "learning_rate": 3.867000583213082e-06, + "loss": 0.0262, + "step": 88030 + }, + { + "epoch": 5.768197602044159, + "grad_norm": 0.8464924693107605, + "learning_rate": 3.860627086405871e-06, + "loss": 0.031, + "step": 88040 + }, + { + "epoch": 5.768852781235668, + "grad_norm": 0.8903023600578308, + "learning_rate": 3.854258635331126e-06, + "loss": 0.0498, + "step": 88050 + }, + { + "epoch": 5.769507960427177, + "grad_norm": 0.323115736246109, + "learning_rate": 3.8478952306852665e-06, + "loss": 0.0311, + "step": 88060 + }, + { + "epoch": 5.770163139618686, + "grad_norm": 0.44335588812828064, + "learning_rate": 3.841536873164198e-06, + "loss": 0.0291, + "step": 88070 + }, + { + "epoch": 5.7708183188101945, + "grad_norm": 0.4220777451992035, + "learning_rate": 3.8351835634632626e-06, + "loss": 0.023, + "step": 88080 + }, + { + "epoch": 5.771473498001703, + "grad_norm": 0.8223716616630554, + "learning_rate": 3.828835302277245e-06, + "loss": 0.0412, + "step": 88090 + }, + { + "epoch": 5.772128677193212, + "grad_norm": 1.0508373975753784, + "learning_rate": 3.822492090300378e-06, + "loss": 0.0385, + "step": 88100 + }, + { + "epoch": 5.772783856384721, + "grad_norm": 0.19881883263587952, + "learning_rate": 3.8161539282263566e-06, + "loss": 0.0146, + "step": 88110 + }, + { + "epoch": 5.77343903557623, + "grad_norm": 1.7986910343170166, + "learning_rate": 3.8098208167482995e-06, + "loss": 0.0348, + "step": 88120 + }, + { + "epoch": 5.774094214767739, + "grad_norm": 0.23797498643398285, + "learning_rate": 3.803492756558796e-06, + "loss": 0.0267, + "step": 88130 + }, + { + "epoch": 5.774749393959247, + "grad_norm": 0.15345561504364014, + "learning_rate": 3.797169748349855e-06, + "loss": 0.051, + "step": 88140 + }, + { + "epoch": 5.775404573150757, + "grad_norm": 0.720916748046875, + "learning_rate": 3.790851792812977e-06, + "loss": 0.0185, + "step": 88150 + }, + { + "epoch": 5.776059752342266, + "grad_norm": 0.17936253547668457, + "learning_rate": 3.784538890639072e-06, + "loss": 0.033, + "step": 88160 + }, + { + "epoch": 5.776714931533775, + "grad_norm": 0.5200068950653076, + "learning_rate": 3.7782310425185153e-06, + "loss": 0.0238, + "step": 88170 + }, + { + "epoch": 5.7773701107252835, + "grad_norm": 0.2528349757194519, + "learning_rate": 3.77192824914111e-06, + "loss": 0.0217, + "step": 88180 + }, + { + "epoch": 5.778025289916792, + "grad_norm": 0.5231308937072754, + "learning_rate": 3.7656305111961333e-06, + "loss": 0.0343, + "step": 88190 + }, + { + "epoch": 5.778680469108301, + "grad_norm": 0.35134726762771606, + "learning_rate": 3.7593378293722782e-06, + "loss": 0.0104, + "step": 88200 + }, + { + "epoch": 5.77933564829981, + "grad_norm": 2.0078976154327393, + "learning_rate": 3.7530502043577333e-06, + "loss": 0.036, + "step": 88210 + }, + { + "epoch": 5.779990827491319, + "grad_norm": 0.1346105933189392, + "learning_rate": 3.746767636840076e-06, + "loss": 0.0173, + "step": 88220 + }, + { + "epoch": 5.780646006682828, + "grad_norm": 0.20153795182704926, + "learning_rate": 3.74049012750638e-06, + "loss": 0.0217, + "step": 88230 + }, + { + "epoch": 5.781301185874336, + "grad_norm": 1.38457453250885, + "learning_rate": 3.734217677043128e-06, + "loss": 0.0332, + "step": 88240 + }, + { + "epoch": 5.781956365065845, + "grad_norm": 0.33574992418289185, + "learning_rate": 3.727950286136267e-06, + "loss": 0.0164, + "step": 88250 + }, + { + "epoch": 5.782611544257354, + "grad_norm": 0.38485264778137207, + "learning_rate": 3.7216879554711925e-06, + "loss": 0.0579, + "step": 88260 + }, + { + "epoch": 5.783266723448863, + "grad_norm": 0.8588561415672302, + "learning_rate": 3.715430685732757e-06, + "loss": 0.0185, + "step": 88270 + }, + { + "epoch": 5.7839219026403725, + "grad_norm": 0.4179079532623291, + "learning_rate": 3.7091784776052243e-06, + "loss": 0.0145, + "step": 88280 + }, + { + "epoch": 5.784577081831881, + "grad_norm": 0.2879675328731537, + "learning_rate": 3.7029313317723356e-06, + "loss": 0.0271, + "step": 88290 + }, + { + "epoch": 5.78523226102339, + "grad_norm": 0.8108915686607361, + "learning_rate": 3.6966892489172787e-06, + "loss": 0.0309, + "step": 88300 + }, + { + "epoch": 5.785887440214899, + "grad_norm": 0.24774199724197388, + "learning_rate": 3.6904522297226575e-06, + "loss": 0.0247, + "step": 88310 + }, + { + "epoch": 5.786542619406408, + "grad_norm": 0.181335911154747, + "learning_rate": 3.6842202748705647e-06, + "loss": 0.0174, + "step": 88320 + }, + { + "epoch": 5.787197798597917, + "grad_norm": 0.24581760168075562, + "learning_rate": 3.677993385042494e-06, + "loss": 0.0262, + "step": 88330 + }, + { + "epoch": 5.787852977789425, + "grad_norm": 0.5878925323486328, + "learning_rate": 3.67177156091944e-06, + "loss": 0.0472, + "step": 88340 + }, + { + "epoch": 5.788508156980934, + "grad_norm": 0.49430975317955017, + "learning_rate": 3.6655548031817866e-06, + "loss": 0.0622, + "step": 88350 + }, + { + "epoch": 5.789163336172443, + "grad_norm": 0.48300084471702576, + "learning_rate": 3.6593431125094e-06, + "loss": 0.0132, + "step": 88360 + }, + { + "epoch": 5.789818515363952, + "grad_norm": 0.11927792429924011, + "learning_rate": 3.6531364895815766e-06, + "loss": 0.0294, + "step": 88370 + }, + { + "epoch": 5.790473694555461, + "grad_norm": 1.8901561498641968, + "learning_rate": 3.6469349350770677e-06, + "loss": 0.0418, + "step": 88380 + }, + { + "epoch": 5.7911288737469695, + "grad_norm": 0.3950968086719513, + "learning_rate": 3.6407384496740525e-06, + "loss": 0.0252, + "step": 88390 + }, + { + "epoch": 5.791784052938478, + "grad_norm": 0.5800145864486694, + "learning_rate": 3.6345470340501954e-06, + "loss": 0.0261, + "step": 88400 + }, + { + "epoch": 5.792439232129988, + "grad_norm": 0.28910917043685913, + "learning_rate": 3.6283606888825537e-06, + "loss": 0.0146, + "step": 88410 + }, + { + "epoch": 5.793094411321496, + "grad_norm": 0.20751632750034332, + "learning_rate": 3.622179414847676e-06, + "loss": 0.0095, + "step": 88420 + }, + { + "epoch": 5.793749590513006, + "grad_norm": 0.9224219918251038, + "learning_rate": 3.616003212621527e-06, + "loss": 0.0281, + "step": 88430 + }, + { + "epoch": 5.7944047697045145, + "grad_norm": 0.8389871716499329, + "learning_rate": 3.609832082879533e-06, + "loss": 0.0849, + "step": 88440 + }, + { + "epoch": 5.795059948896023, + "grad_norm": 0.6927467584609985, + "learning_rate": 3.6036660262965485e-06, + "loss": 0.0661, + "step": 88450 + }, + { + "epoch": 5.795715128087532, + "grad_norm": 0.6054608225822449, + "learning_rate": 3.597505043546895e-06, + "loss": 0.0379, + "step": 88460 + }, + { + "epoch": 5.796370307279041, + "grad_norm": 0.795386791229248, + "learning_rate": 3.5913491353043227e-06, + "loss": 0.0392, + "step": 88470 + }, + { + "epoch": 5.79702548647055, + "grad_norm": 0.32276636362075806, + "learning_rate": 3.5851983022420375e-06, + "loss": 0.0241, + "step": 88480 + }, + { + "epoch": 5.7976806656620585, + "grad_norm": 0.4488815665245056, + "learning_rate": 3.5790525450326797e-06, + "loss": 0.0314, + "step": 88490 + }, + { + "epoch": 5.798335844853567, + "grad_norm": 0.7737470269203186, + "learning_rate": 3.572911864348344e-06, + "loss": 0.0524, + "step": 88500 + }, + { + "epoch": 5.798991024045076, + "grad_norm": 0.3407561480998993, + "learning_rate": 3.5667762608605717e-06, + "loss": 0.0245, + "step": 88510 + }, + { + "epoch": 5.799646203236585, + "grad_norm": 0.7884722352027893, + "learning_rate": 3.5606457352403255e-06, + "loss": 0.0384, + "step": 88520 + }, + { + "epoch": 5.800301382428094, + "grad_norm": 0.19316469132900238, + "learning_rate": 3.554520288158042e-06, + "loss": 0.0176, + "step": 88530 + }, + { + "epoch": 5.800956561619603, + "grad_norm": 0.7461273074150085, + "learning_rate": 3.5483999202835903e-06, + "loss": 0.0183, + "step": 88540 + }, + { + "epoch": 5.801611740811111, + "grad_norm": 0.6512607336044312, + "learning_rate": 3.5422846322862913e-06, + "loss": 0.034, + "step": 88550 + }, + { + "epoch": 5.802266920002621, + "grad_norm": 0.3549014925956726, + "learning_rate": 3.5361744248348925e-06, + "loss": 0.0319, + "step": 88560 + }, + { + "epoch": 5.80292209919413, + "grad_norm": 0.3261853754520416, + "learning_rate": 3.5300692985976047e-06, + "loss": 0.0205, + "step": 88570 + }, + { + "epoch": 5.803577278385639, + "grad_norm": 0.43456560373306274, + "learning_rate": 3.52396925424206e-06, + "loss": 0.0183, + "step": 88580 + }, + { + "epoch": 5.804232457577148, + "grad_norm": 0.7287803292274475, + "learning_rate": 3.5178742924353747e-06, + "loss": 0.0456, + "step": 88590 + }, + { + "epoch": 5.804887636768656, + "grad_norm": 0.4645825922489166, + "learning_rate": 3.511784413844066e-06, + "loss": 0.0174, + "step": 88600 + }, + { + "epoch": 5.805542815960165, + "grad_norm": 0.3097866177558899, + "learning_rate": 3.5056996191341287e-06, + "loss": 0.048, + "step": 88610 + }, + { + "epoch": 5.806197995151674, + "grad_norm": 0.2610132396221161, + "learning_rate": 3.4996199089709692e-06, + "loss": 0.0228, + "step": 88620 + }, + { + "epoch": 5.806853174343183, + "grad_norm": 0.22880633175373077, + "learning_rate": 3.4935452840194725e-06, + "loss": 0.0231, + "step": 88630 + }, + { + "epoch": 5.807508353534692, + "grad_norm": 0.13787618279457092, + "learning_rate": 3.48747574494393e-06, + "loss": 0.0191, + "step": 88640 + }, + { + "epoch": 5.8081635327262005, + "grad_norm": 0.3425597548484802, + "learning_rate": 3.4814112924081154e-06, + "loss": 0.0198, + "step": 88650 + }, + { + "epoch": 5.808818711917709, + "grad_norm": 0.40163499116897583, + "learning_rate": 3.475351927075221e-06, + "loss": 0.0182, + "step": 88660 + }, + { + "epoch": 5.809473891109218, + "grad_norm": 0.7631982564926147, + "learning_rate": 3.469297649607894e-06, + "loss": 0.0223, + "step": 88670 + }, + { + "epoch": 5.810129070300727, + "grad_norm": 0.2591356933116913, + "learning_rate": 3.463248460668217e-06, + "loss": 0.019, + "step": 88680 + }, + { + "epoch": 5.810784249492237, + "grad_norm": 0.42296600341796875, + "learning_rate": 3.457204360917726e-06, + "loss": 0.0294, + "step": 88690 + }, + { + "epoch": 5.8114394286837445, + "grad_norm": 0.6403422951698303, + "learning_rate": 3.4511653510173827e-06, + "loss": 0.0304, + "step": 88700 + }, + { + "epoch": 5.812094607875254, + "grad_norm": 0.18065929412841797, + "learning_rate": 3.4451314316276074e-06, + "loss": 0.0213, + "step": 88710 + }, + { + "epoch": 5.812749787066763, + "grad_norm": 0.3401201367378235, + "learning_rate": 3.4391026034082674e-06, + "loss": 0.0266, + "step": 88720 + }, + { + "epoch": 5.813404966258272, + "grad_norm": 0.45781785249710083, + "learning_rate": 3.433078867018674e-06, + "loss": 0.0296, + "step": 88730 + }, + { + "epoch": 5.814060145449781, + "grad_norm": 0.3661835193634033, + "learning_rate": 3.427060223117551e-06, + "loss": 0.024, + "step": 88740 + }, + { + "epoch": 5.8147153246412895, + "grad_norm": 0.7625076174736023, + "learning_rate": 3.4210466723631042e-06, + "loss": 0.0573, + "step": 88750 + }, + { + "epoch": 5.815370503832798, + "grad_norm": 0.2497660368680954, + "learning_rate": 3.415038215412969e-06, + "loss": 0.0246, + "step": 88760 + }, + { + "epoch": 5.816025683024307, + "grad_norm": 0.06883016973733902, + "learning_rate": 3.4090348529242088e-06, + "loss": 0.0274, + "step": 88770 + }, + { + "epoch": 5.816680862215816, + "grad_norm": 0.48149797320365906, + "learning_rate": 3.4030365855533487e-06, + "loss": 0.0155, + "step": 88780 + }, + { + "epoch": 5.817336041407325, + "grad_norm": 0.8960673213005066, + "learning_rate": 3.397043413956347e-06, + "loss": 0.0403, + "step": 88790 + }, + { + "epoch": 5.817991220598834, + "grad_norm": 0.12680794298648834, + "learning_rate": 3.3910553387886237e-06, + "loss": 0.0151, + "step": 88800 + }, + { + "epoch": 5.818646399790342, + "grad_norm": 0.5470655560493469, + "learning_rate": 3.3850723607049996e-06, + "loss": 0.0268, + "step": 88810 + }, + { + "epoch": 5.819301578981851, + "grad_norm": 0.3320864737033844, + "learning_rate": 3.3790944803597902e-06, + "loss": 0.0503, + "step": 88820 + }, + { + "epoch": 5.81995675817336, + "grad_norm": 0.22844946384429932, + "learning_rate": 3.373121698406706e-06, + "loss": 0.0359, + "step": 88830 + }, + { + "epoch": 5.82061193736487, + "grad_norm": 0.7766441106796265, + "learning_rate": 3.3671540154989302e-06, + "loss": 0.0387, + "step": 88840 + }, + { + "epoch": 5.8212671165563785, + "grad_norm": 0.3280091881752014, + "learning_rate": 3.361191432289079e-06, + "loss": 0.0217, + "step": 88850 + }, + { + "epoch": 5.821922295747887, + "grad_norm": 0.06386388838291168, + "learning_rate": 3.3552339494292207e-06, + "loss": 0.0356, + "step": 88860 + }, + { + "epoch": 5.822577474939396, + "grad_norm": 0.353922963142395, + "learning_rate": 3.349281567570839e-06, + "loss": 0.0255, + "step": 88870 + }, + { + "epoch": 5.823232654130905, + "grad_norm": 0.4109230935573578, + "learning_rate": 3.3433342873648965e-06, + "loss": 0.018, + "step": 88880 + }, + { + "epoch": 5.823887833322414, + "grad_norm": 0.23587770760059357, + "learning_rate": 3.337392109461762e-06, + "loss": 0.0156, + "step": 88890 + }, + { + "epoch": 5.824543012513923, + "grad_norm": 0.7331915497779846, + "learning_rate": 3.331455034511266e-06, + "loss": 0.019, + "step": 88900 + }, + { + "epoch": 5.825198191705431, + "grad_norm": 0.6365292072296143, + "learning_rate": 3.325523063162689e-06, + "loss": 0.0254, + "step": 88910 + }, + { + "epoch": 5.82585337089694, + "grad_norm": 0.259625643491745, + "learning_rate": 3.3195961960647393e-06, + "loss": 0.0233, + "step": 88920 + }, + { + "epoch": 5.826508550088449, + "grad_norm": 0.3310423195362091, + "learning_rate": 3.31367443386556e-06, + "loss": 0.047, + "step": 88930 + }, + { + "epoch": 5.827163729279958, + "grad_norm": 0.12283723801374435, + "learning_rate": 3.3077577772127556e-06, + "loss": 0.0259, + "step": 88940 + }, + { + "epoch": 5.827818908471467, + "grad_norm": 0.37766802310943604, + "learning_rate": 3.301846226753358e-06, + "loss": 0.0264, + "step": 88950 + }, + { + "epoch": 5.8284740876629755, + "grad_norm": 0.2543095052242279, + "learning_rate": 3.2959397831338444e-06, + "loss": 0.0492, + "step": 88960 + }, + { + "epoch": 5.829129266854485, + "grad_norm": 0.45100119709968567, + "learning_rate": 3.2900384470001432e-06, + "loss": 0.0219, + "step": 88970 + }, + { + "epoch": 5.829784446045993, + "grad_norm": 0.7802789211273193, + "learning_rate": 3.28414221899761e-06, + "loss": 0.0168, + "step": 88980 + }, + { + "epoch": 5.830439625237503, + "grad_norm": 0.993627667427063, + "learning_rate": 3.278251099771046e-06, + "loss": 0.0341, + "step": 88990 + }, + { + "epoch": 5.831094804429012, + "grad_norm": 0.6358469724655151, + "learning_rate": 3.2723650899646906e-06, + "loss": 0.0186, + "step": 89000 + }, + { + "epoch": 5.83174998362052, + "grad_norm": 0.31693679094314575, + "learning_rate": 3.2664841902222466e-06, + "loss": 0.0205, + "step": 89010 + }, + { + "epoch": 5.832405162812029, + "grad_norm": 0.14165587723255157, + "learning_rate": 3.2606084011868208e-06, + "loss": 0.0206, + "step": 89020 + }, + { + "epoch": 5.833060342003538, + "grad_norm": 0.18016570806503296, + "learning_rate": 3.254737723500989e-06, + "loss": 0.0291, + "step": 89030 + }, + { + "epoch": 5.833715521195047, + "grad_norm": 0.82978355884552, + "learning_rate": 3.2488721578067595e-06, + "loss": 0.0468, + "step": 89040 + }, + { + "epoch": 5.834370700386556, + "grad_norm": 0.2590144872665405, + "learning_rate": 3.243011704745591e-06, + "loss": 0.0183, + "step": 89050 + }, + { + "epoch": 5.8350258795780645, + "grad_norm": 0.23606093227863312, + "learning_rate": 3.237156364958355e-06, + "loss": 0.0246, + "step": 89060 + }, + { + "epoch": 5.835681058769573, + "grad_norm": 0.39531823992729187, + "learning_rate": 3.231306139085394e-06, + "loss": 0.0176, + "step": 89070 + }, + { + "epoch": 5.836336237961082, + "grad_norm": 0.43509334325790405, + "learning_rate": 3.2254610277664797e-06, + "loss": 0.0198, + "step": 89080 + }, + { + "epoch": 5.836991417152591, + "grad_norm": 0.8960873484611511, + "learning_rate": 3.2196210316408183e-06, + "loss": 0.0263, + "step": 89090 + }, + { + "epoch": 5.8376465963441, + "grad_norm": 0.4446945786476135, + "learning_rate": 3.2137861513470644e-06, + "loss": 0.0242, + "step": 89100 + }, + { + "epoch": 5.838301775535609, + "grad_norm": 1.0465295314788818, + "learning_rate": 3.2079563875233252e-06, + "loss": 0.0319, + "step": 89110 + }, + { + "epoch": 5.838956954727118, + "grad_norm": 0.1122356504201889, + "learning_rate": 3.2021317408071184e-06, + "loss": 0.0272, + "step": 89120 + }, + { + "epoch": 5.839612133918627, + "grad_norm": 0.7417239546775818, + "learning_rate": 3.196312211835434e-06, + "loss": 0.0215, + "step": 89130 + }, + { + "epoch": 5.840267313110136, + "grad_norm": 0.4113447368144989, + "learning_rate": 3.1904978012446686e-06, + "loss": 0.0302, + "step": 89140 + }, + { + "epoch": 5.840922492301645, + "grad_norm": 0.18490315973758698, + "learning_rate": 3.184688509670691e-06, + "loss": 0.0184, + "step": 89150 + }, + { + "epoch": 5.8415776714931535, + "grad_norm": 0.08087257295846939, + "learning_rate": 3.178884337748783e-06, + "loss": 0.022, + "step": 89160 + }, + { + "epoch": 5.842232850684662, + "grad_norm": 0.3816160559654236, + "learning_rate": 3.173085286113697e-06, + "loss": 0.0179, + "step": 89170 + }, + { + "epoch": 5.842888029876171, + "grad_norm": 0.974229097366333, + "learning_rate": 3.1672913553995986e-06, + "loss": 0.0365, + "step": 89180 + }, + { + "epoch": 5.84354320906768, + "grad_norm": 0.14724008738994598, + "learning_rate": 3.1615025462401138e-06, + "loss": 0.019, + "step": 89190 + }, + { + "epoch": 5.844198388259189, + "grad_norm": 0.29898348450660706, + "learning_rate": 3.1557188592682808e-06, + "loss": 0.0317, + "step": 89200 + }, + { + "epoch": 5.844853567450698, + "grad_norm": 1.5655975341796875, + "learning_rate": 3.1499402951166045e-06, + "loss": 0.0198, + "step": 89210 + }, + { + "epoch": 5.845508746642206, + "grad_norm": 0.4036600589752197, + "learning_rate": 3.144166854417019e-06, + "loss": 0.0452, + "step": 89220 + }, + { + "epoch": 5.846163925833715, + "grad_norm": 0.058597587049007416, + "learning_rate": 3.1383985378009073e-06, + "loss": 0.0161, + "step": 89230 + }, + { + "epoch": 5.846819105025224, + "grad_norm": 0.8826005458831787, + "learning_rate": 3.1326353458990654e-06, + "loss": 0.0312, + "step": 89240 + }, + { + "epoch": 5.847474284216734, + "grad_norm": 0.05607189983129501, + "learning_rate": 3.126877279341761e-06, + "loss": 0.0301, + "step": 89250 + }, + { + "epoch": 5.848129463408242, + "grad_norm": 0.24596333503723145, + "learning_rate": 3.1211243387586964e-06, + "loss": 0.0434, + "step": 89260 + }, + { + "epoch": 5.848784642599751, + "grad_norm": 1.1294968128204346, + "learning_rate": 3.115376524778979e-06, + "loss": 0.0351, + "step": 89270 + }, + { + "epoch": 5.84943982179126, + "grad_norm": 0.4081003665924072, + "learning_rate": 3.109633838031195e-06, + "loss": 0.0445, + "step": 89280 + }, + { + "epoch": 5.850095000982769, + "grad_norm": 0.6161164045333862, + "learning_rate": 3.103896279143359e-06, + "loss": 0.0157, + "step": 89290 + }, + { + "epoch": 5.850750180174278, + "grad_norm": 0.6675835251808167, + "learning_rate": 3.098163848742924e-06, + "loss": 0.0214, + "step": 89300 + }, + { + "epoch": 5.851405359365787, + "grad_norm": 0.7806201577186584, + "learning_rate": 3.0924365474567675e-06, + "loss": 0.0422, + "step": 89310 + }, + { + "epoch": 5.8520605385572955, + "grad_norm": 0.5566790699958801, + "learning_rate": 3.086714375911237e-06, + "loss": 0.0348, + "step": 89320 + }, + { + "epoch": 5.852715717748804, + "grad_norm": 0.2244226634502411, + "learning_rate": 3.0809973347320775e-06, + "loss": 0.017, + "step": 89330 + }, + { + "epoch": 5.853370896940313, + "grad_norm": 0.29395171999931335, + "learning_rate": 3.075285424544516e-06, + "loss": 0.0398, + "step": 89340 + }, + { + "epoch": 5.854026076131822, + "grad_norm": 0.09275885671377182, + "learning_rate": 3.069578645973176e-06, + "loss": 0.0355, + "step": 89350 + }, + { + "epoch": 5.854681255323331, + "grad_norm": 0.07538006454706192, + "learning_rate": 3.0638769996421735e-06, + "loss": 0.0172, + "step": 89360 + }, + { + "epoch": 5.8553364345148395, + "grad_norm": 0.30893954634666443, + "learning_rate": 3.058180486175005e-06, + "loss": 0.0274, + "step": 89370 + }, + { + "epoch": 5.855991613706348, + "grad_norm": 0.6475958228111267, + "learning_rate": 3.052489106194645e-06, + "loss": 0.0255, + "step": 89380 + }, + { + "epoch": 5.856646792897857, + "grad_norm": 0.2792145907878876, + "learning_rate": 3.0468028603234946e-06, + "loss": 0.0294, + "step": 89390 + }, + { + "epoch": 5.857301972089367, + "grad_norm": 0.8344174027442932, + "learning_rate": 3.04112174918339e-06, + "loss": 0.0224, + "step": 89400 + }, + { + "epoch": 5.857957151280876, + "grad_norm": 0.2168542593717575, + "learning_rate": 3.0354457733955954e-06, + "loss": 0.03, + "step": 89410 + }, + { + "epoch": 5.8586123304723845, + "grad_norm": 0.8190056085586548, + "learning_rate": 3.029774933580859e-06, + "loss": 0.0291, + "step": 89420 + }, + { + "epoch": 5.859267509663893, + "grad_norm": 0.17669035494327545, + "learning_rate": 3.0241092303593065e-06, + "loss": 0.0216, + "step": 89430 + }, + { + "epoch": 5.859922688855402, + "grad_norm": 0.4967392683029175, + "learning_rate": 3.018448664350554e-06, + "loss": 0.0281, + "step": 89440 + }, + { + "epoch": 5.860577868046911, + "grad_norm": 0.6914746761322021, + "learning_rate": 3.0127932361736055e-06, + "loss": 0.018, + "step": 89450 + }, + { + "epoch": 5.86123304723842, + "grad_norm": 0.5953497886657715, + "learning_rate": 3.00714294644695e-06, + "loss": 0.0289, + "step": 89460 + }, + { + "epoch": 5.861888226429929, + "grad_norm": 0.4318140745162964, + "learning_rate": 3.0014977957884828e-06, + "loss": 0.0483, + "step": 89470 + }, + { + "epoch": 5.862543405621437, + "grad_norm": 0.3540530800819397, + "learning_rate": 2.99585778481557e-06, + "loss": 0.0548, + "step": 89480 + }, + { + "epoch": 5.863198584812946, + "grad_norm": 0.44826406240463257, + "learning_rate": 2.990222914144969e-06, + "loss": 0.0274, + "step": 89490 + }, + { + "epoch": 5.863853764004455, + "grad_norm": 0.24511520564556122, + "learning_rate": 2.984593184392914e-06, + "loss": 0.0165, + "step": 89500 + }, + { + "epoch": 5.864508943195964, + "grad_norm": 0.33008652925491333, + "learning_rate": 2.978968596175069e-06, + "loss": 0.033, + "step": 89510 + }, + { + "epoch": 5.865164122387473, + "grad_norm": 0.6889456510543823, + "learning_rate": 2.9733491501065137e-06, + "loss": 0.0231, + "step": 89520 + }, + { + "epoch": 5.865819301578982, + "grad_norm": 0.3198104798793793, + "learning_rate": 2.9677348468018005e-06, + "loss": 0.0137, + "step": 89530 + }, + { + "epoch": 5.86647448077049, + "grad_norm": 0.05643859878182411, + "learning_rate": 2.9621256868748726e-06, + "loss": 0.0224, + "step": 89540 + }, + { + "epoch": 5.867129659962, + "grad_norm": 0.14055310189723969, + "learning_rate": 2.956521670939183e-06, + "loss": 0.033, + "step": 89550 + }, + { + "epoch": 5.867784839153509, + "grad_norm": 0.4152703285217285, + "learning_rate": 2.950922799607542e-06, + "loss": 0.0215, + "step": 89560 + }, + { + "epoch": 5.868440018345018, + "grad_norm": 1.1010143756866455, + "learning_rate": 2.9453290734922535e-06, + "loss": 0.018, + "step": 89570 + }, + { + "epoch": 5.869095197536526, + "grad_norm": 0.49835968017578125, + "learning_rate": 2.939740493205029e-06, + "loss": 0.0404, + "step": 89580 + }, + { + "epoch": 5.869750376728035, + "grad_norm": 1.1714627742767334, + "learning_rate": 2.9341570593570345e-06, + "loss": 0.0374, + "step": 89590 + }, + { + "epoch": 5.870405555919544, + "grad_norm": 0.1778874695301056, + "learning_rate": 2.928578772558843e-06, + "loss": 0.0398, + "step": 89600 + }, + { + "epoch": 5.871060735111053, + "grad_norm": 0.6145445108413696, + "learning_rate": 2.923005633420528e-06, + "loss": 0.0365, + "step": 89610 + }, + { + "epoch": 5.871715914302562, + "grad_norm": 0.7572212815284729, + "learning_rate": 2.9174376425515237e-06, + "loss": 0.0738, + "step": 89620 + }, + { + "epoch": 5.8723710934940705, + "grad_norm": 1.4490509033203125, + "learning_rate": 2.91187480056076e-06, + "loss": 0.054, + "step": 89630 + }, + { + "epoch": 5.873026272685579, + "grad_norm": 0.2340116649866104, + "learning_rate": 2.9063171080565664e-06, + "loss": 0.031, + "step": 89640 + }, + { + "epoch": 5.873681451877088, + "grad_norm": 0.6460694074630737, + "learning_rate": 2.900764565646741e-06, + "loss": 0.036, + "step": 89650 + }, + { + "epoch": 5.874336631068597, + "grad_norm": 0.5760231018066406, + "learning_rate": 2.895217173938475e-06, + "loss": 0.0281, + "step": 89660 + }, + { + "epoch": 5.874991810260106, + "grad_norm": 1.5361363887786865, + "learning_rate": 2.8896749335384445e-06, + "loss": 0.0322, + "step": 89670 + }, + { + "epoch": 5.8756469894516155, + "grad_norm": 0.4170433580875397, + "learning_rate": 2.884137845052731e-06, + "loss": 0.0432, + "step": 89680 + }, + { + "epoch": 5.876302168643124, + "grad_norm": 0.10292268544435501, + "learning_rate": 2.8786059090868734e-06, + "loss": 0.0322, + "step": 89690 + }, + { + "epoch": 5.876957347834633, + "grad_norm": 0.42392656207084656, + "learning_rate": 2.87307912624582e-06, + "loss": 0.0292, + "step": 89700 + }, + { + "epoch": 5.877612527026142, + "grad_norm": 0.2801700830459595, + "learning_rate": 2.867557497133977e-06, + "loss": 0.0261, + "step": 89710 + }, + { + "epoch": 5.878267706217651, + "grad_norm": 0.19576925039291382, + "learning_rate": 2.862041022355194e-06, + "loss": 0.0218, + "step": 89720 + }, + { + "epoch": 5.8789228854091595, + "grad_norm": 0.22071073949337006, + "learning_rate": 2.8565297025127226e-06, + "loss": 0.0186, + "step": 89730 + }, + { + "epoch": 5.879578064600668, + "grad_norm": 0.4633987247943878, + "learning_rate": 2.8510235382092853e-06, + "loss": 0.0426, + "step": 89740 + }, + { + "epoch": 5.880233243792177, + "grad_norm": 0.5770448446273804, + "learning_rate": 2.845522530047029e-06, + "loss": 0.0262, + "step": 89750 + }, + { + "epoch": 5.880888422983686, + "grad_norm": 1.1908749341964722, + "learning_rate": 2.8400266786275387e-06, + "loss": 0.0309, + "step": 89760 + }, + { + "epoch": 5.881543602175195, + "grad_norm": 0.4576699137687683, + "learning_rate": 2.834535984551817e-06, + "loss": 0.0274, + "step": 89770 + }, + { + "epoch": 5.882198781366704, + "grad_norm": 0.4451866149902344, + "learning_rate": 2.8290504484203397e-06, + "loss": 0.0385, + "step": 89780 + }, + { + "epoch": 5.882853960558212, + "grad_norm": 0.2434200793504715, + "learning_rate": 2.8235700708329705e-06, + "loss": 0.0318, + "step": 89790 + }, + { + "epoch": 5.883509139749721, + "grad_norm": 0.8151198029518127, + "learning_rate": 2.818094852389058e-06, + "loss": 0.0481, + "step": 89800 + }, + { + "epoch": 5.88416431894123, + "grad_norm": 0.2135932892560959, + "learning_rate": 2.8126247936873516e-06, + "loss": 0.0258, + "step": 89810 + }, + { + "epoch": 5.884819498132739, + "grad_norm": 0.4346754550933838, + "learning_rate": 2.8071598953260614e-06, + "loss": 0.0248, + "step": 89820 + }, + { + "epoch": 5.885474677324249, + "grad_norm": 0.6884034872055054, + "learning_rate": 2.8017001579028035e-06, + "loss": 0.0322, + "step": 89830 + }, + { + "epoch": 5.886129856515757, + "grad_norm": 0.3366396427154541, + "learning_rate": 2.7962455820146617e-06, + "loss": 0.0477, + "step": 89840 + }, + { + "epoch": 5.886785035707266, + "grad_norm": 0.5943257212638855, + "learning_rate": 2.7907961682581253e-06, + "loss": 0.0187, + "step": 89850 + }, + { + "epoch": 5.887440214898775, + "grad_norm": 0.7894384264945984, + "learning_rate": 2.7853519172291453e-06, + "loss": 0.0182, + "step": 89860 + }, + { + "epoch": 5.888095394090284, + "grad_norm": 2.393200397491455, + "learning_rate": 2.7799128295230904e-06, + "loss": 0.0316, + "step": 89870 + }, + { + "epoch": 5.888750573281793, + "grad_norm": 0.25505995750427246, + "learning_rate": 2.774478905734784e-06, + "loss": 0.0388, + "step": 89880 + }, + { + "epoch": 5.8894057524733014, + "grad_norm": 1.1352142095565796, + "learning_rate": 2.7690501464584516e-06, + "loss": 0.0405, + "step": 89890 + }, + { + "epoch": 5.89006093166481, + "grad_norm": 0.5951191186904907, + "learning_rate": 2.763626552287796e-06, + "loss": 0.0313, + "step": 89900 + }, + { + "epoch": 5.890716110856319, + "grad_norm": 1.416003704071045, + "learning_rate": 2.7582081238159083e-06, + "loss": 0.043, + "step": 89910 + }, + { + "epoch": 5.891371290047828, + "grad_norm": 0.5614339113235474, + "learning_rate": 2.7527948616353548e-06, + "loss": 0.0314, + "step": 89920 + }, + { + "epoch": 5.892026469239337, + "grad_norm": 0.4687212407588959, + "learning_rate": 2.7473867663381226e-06, + "loss": 0.0361, + "step": 89930 + }, + { + "epoch": 5.8926816484308455, + "grad_norm": 0.45235154032707214, + "learning_rate": 2.7419838385156383e-06, + "loss": 0.0196, + "step": 89940 + }, + { + "epoch": 5.893336827622354, + "grad_norm": 0.08903801441192627, + "learning_rate": 2.7365860787587407e-06, + "loss": 0.0273, + "step": 89950 + }, + { + "epoch": 5.893992006813864, + "grad_norm": 0.43190887570381165, + "learning_rate": 2.7311934876577295e-06, + "loss": 0.0408, + "step": 89960 + }, + { + "epoch": 5.894647186005373, + "grad_norm": 0.49290701746940613, + "learning_rate": 2.725806065802339e-06, + "loss": 0.02, + "step": 89970 + }, + { + "epoch": 5.895302365196882, + "grad_norm": 0.22777864336967468, + "learning_rate": 2.720423813781714e-06, + "loss": 0.0179, + "step": 89980 + }, + { + "epoch": 5.8959575443883905, + "grad_norm": 0.6542510390281677, + "learning_rate": 2.715046732184462e-06, + "loss": 0.0243, + "step": 89990 + }, + { + "epoch": 5.896612723579899, + "grad_norm": 0.15996623039245605, + "learning_rate": 2.7096748215986013e-06, + "loss": 0.026, + "step": 90000 + }, + { + "epoch": 5.897267902771408, + "grad_norm": 0.6539825797080994, + "learning_rate": 2.704308082611612e-06, + "loss": 0.0215, + "step": 90010 + }, + { + "epoch": 5.897923081962917, + "grad_norm": 0.2078641653060913, + "learning_rate": 2.698946515810374e-06, + "loss": 0.0288, + "step": 90020 + }, + { + "epoch": 5.898578261154426, + "grad_norm": 0.0815335363149643, + "learning_rate": 2.6935901217812367e-06, + "loss": 0.0227, + "step": 90030 + }, + { + "epoch": 5.8992334403459346, + "grad_norm": 0.2641143202781677, + "learning_rate": 2.688238901109952e-06, + "loss": 0.0166, + "step": 90040 + }, + { + "epoch": 5.899888619537443, + "grad_norm": 0.11945336312055588, + "learning_rate": 2.68289285438173e-06, + "loss": 0.0216, + "step": 90050 + }, + { + "epoch": 5.900543798728952, + "grad_norm": 0.1371966451406479, + "learning_rate": 2.6775519821812033e-06, + "loss": 0.0162, + "step": 90060 + }, + { + "epoch": 5.901198977920461, + "grad_norm": 0.41430580615997314, + "learning_rate": 2.672216285092449e-06, + "loss": 0.0532, + "step": 90070 + }, + { + "epoch": 5.90185415711197, + "grad_norm": 0.29158923029899597, + "learning_rate": 2.666885763698962e-06, + "loss": 0.0133, + "step": 90080 + }, + { + "epoch": 5.902509336303479, + "grad_norm": 1.9070937633514404, + "learning_rate": 2.661560418583692e-06, + "loss": 0.0576, + "step": 90090 + }, + { + "epoch": 5.903164515494987, + "grad_norm": 0.51564621925354, + "learning_rate": 2.656240250328995e-06, + "loss": 0.0249, + "step": 90100 + }, + { + "epoch": 5.903819694686497, + "grad_norm": 0.519036591053009, + "learning_rate": 2.650925259516679e-06, + "loss": 0.0281, + "step": 90110 + }, + { + "epoch": 5.904474873878006, + "grad_norm": 1.1354601383209229, + "learning_rate": 2.645615446727995e-06, + "loss": 0.0289, + "step": 90120 + }, + { + "epoch": 5.905130053069515, + "grad_norm": 0.7260355949401855, + "learning_rate": 2.6403108125436115e-06, + "loss": 0.0256, + "step": 90130 + }, + { + "epoch": 5.905785232261024, + "grad_norm": 0.142106831073761, + "learning_rate": 2.6350113575436264e-06, + "loss": 0.0188, + "step": 90140 + }, + { + "epoch": 5.906440411452532, + "grad_norm": 0.5008075833320618, + "learning_rate": 2.629717082307598e-06, + "loss": 0.0196, + "step": 90150 + }, + { + "epoch": 5.907095590644041, + "grad_norm": 0.3817977011203766, + "learning_rate": 2.6244279874144793e-06, + "loss": 0.0128, + "step": 90160 + }, + { + "epoch": 5.90775076983555, + "grad_norm": 0.7320970296859741, + "learning_rate": 2.6191440734426862e-06, + "loss": 0.0289, + "step": 90170 + }, + { + "epoch": 5.908405949027059, + "grad_norm": 0.19006185233592987, + "learning_rate": 2.613865340970062e-06, + "loss": 0.0544, + "step": 90180 + }, + { + "epoch": 5.909061128218568, + "grad_norm": 0.7208732962608337, + "learning_rate": 2.608591790573889e-06, + "loss": 0.0488, + "step": 90190 + }, + { + "epoch": 5.9097163074100765, + "grad_norm": 0.3908263146877289, + "learning_rate": 2.603323422830861e-06, + "loss": 0.0121, + "step": 90200 + }, + { + "epoch": 5.910371486601585, + "grad_norm": 0.6146987676620483, + "learning_rate": 2.5980602383171183e-06, + "loss": 0.0206, + "step": 90210 + }, + { + "epoch": 5.911026665793094, + "grad_norm": 0.27049389481544495, + "learning_rate": 2.5928022376082496e-06, + "loss": 0.0336, + "step": 90220 + }, + { + "epoch": 5.911681844984603, + "grad_norm": 0.5470695495605469, + "learning_rate": 2.5875494212792505e-06, + "loss": 0.0247, + "step": 90230 + }, + { + "epoch": 5.912337024176113, + "grad_norm": 1.4837613105773926, + "learning_rate": 2.5823017899045565e-06, + "loss": 0.0691, + "step": 90240 + }, + { + "epoch": 5.9129922033676205, + "grad_norm": 0.3242878019809723, + "learning_rate": 2.577059344058047e-06, + "loss": 0.02, + "step": 90250 + }, + { + "epoch": 5.91364738255913, + "grad_norm": 0.17133721709251404, + "learning_rate": 2.5718220843130415e-06, + "loss": 0.0243, + "step": 90260 + }, + { + "epoch": 5.914302561750639, + "grad_norm": 0.3130173683166504, + "learning_rate": 2.5665900112422535e-06, + "loss": 0.0338, + "step": 90270 + }, + { + "epoch": 5.914957740942148, + "grad_norm": 0.5221499800682068, + "learning_rate": 2.561363125417876e-06, + "loss": 0.0326, + "step": 90280 + }, + { + "epoch": 5.915612920133657, + "grad_norm": 0.13833312690258026, + "learning_rate": 2.556141427411496e-06, + "loss": 0.0235, + "step": 90290 + }, + { + "epoch": 5.9162680993251655, + "grad_norm": 0.16930252313613892, + "learning_rate": 2.550924917794156e-06, + "loss": 0.0148, + "step": 90300 + }, + { + "epoch": 5.916923278516674, + "grad_norm": 0.21543461084365845, + "learning_rate": 2.545713597136329e-06, + "loss": 0.0377, + "step": 90310 + }, + { + "epoch": 5.917578457708183, + "grad_norm": 0.3635423183441162, + "learning_rate": 2.5405074660079298e-06, + "loss": 0.025, + "step": 90320 + }, + { + "epoch": 5.918233636899692, + "grad_norm": 0.20557048916816711, + "learning_rate": 2.5353065249782647e-06, + "loss": 0.0238, + "step": 90330 + }, + { + "epoch": 5.918888816091201, + "grad_norm": 0.42595306038856506, + "learning_rate": 2.530110774616129e-06, + "loss": 0.0296, + "step": 90340 + }, + { + "epoch": 5.91954399528271, + "grad_norm": 0.6344985365867615, + "learning_rate": 2.524920215489701e-06, + "loss": 0.0174, + "step": 90350 + }, + { + "epoch": 5.920199174474218, + "grad_norm": 0.07537141442298889, + "learning_rate": 2.519734848166627e-06, + "loss": 0.0227, + "step": 90360 + }, + { + "epoch": 5.920854353665727, + "grad_norm": 0.22769182920455933, + "learning_rate": 2.514554673213948e-06, + "loss": 0.0175, + "step": 90370 + }, + { + "epoch": 5.921509532857236, + "grad_norm": 1.0348001718521118, + "learning_rate": 2.5093796911981936e-06, + "loss": 0.0249, + "step": 90380 + }, + { + "epoch": 5.922164712048746, + "grad_norm": 0.5076555609703064, + "learning_rate": 2.5042099026852672e-06, + "loss": 0.0291, + "step": 90390 + }, + { + "epoch": 5.9228198912402545, + "grad_norm": 0.6591706275939941, + "learning_rate": 2.499045308240544e-06, + "loss": 0.0216, + "step": 90400 + }, + { + "epoch": 5.923475070431763, + "grad_norm": 0.37684208154678345, + "learning_rate": 2.4938859084288003e-06, + "loss": 0.0213, + "step": 90410 + }, + { + "epoch": 5.924130249623272, + "grad_norm": 0.28492966294288635, + "learning_rate": 2.4887317038142734e-06, + "loss": 0.0215, + "step": 90420 + }, + { + "epoch": 5.924785428814781, + "grad_norm": 0.4303241968154907, + "learning_rate": 2.4835826949606124e-06, + "loss": 0.0255, + "step": 90430 + }, + { + "epoch": 5.92544060800629, + "grad_norm": 0.7039520740509033, + "learning_rate": 2.4784388824309167e-06, + "loss": 0.0283, + "step": 90440 + }, + { + "epoch": 5.926095787197799, + "grad_norm": 0.7625243067741394, + "learning_rate": 2.473300266787687e-06, + "loss": 0.0311, + "step": 90450 + }, + { + "epoch": 5.926750966389307, + "grad_norm": 0.7022971510887146, + "learning_rate": 2.46816684859289e-06, + "loss": 0.0288, + "step": 90460 + }, + { + "epoch": 5.927406145580816, + "grad_norm": 0.5182335376739502, + "learning_rate": 2.4630386284079043e-06, + "loss": 0.0288, + "step": 90470 + }, + { + "epoch": 5.928061324772325, + "grad_norm": 0.3230717182159424, + "learning_rate": 2.4579156067935427e-06, + "loss": 0.0171, + "step": 90480 + }, + { + "epoch": 5.928716503963834, + "grad_norm": 0.16918402910232544, + "learning_rate": 2.452797784310046e-06, + "loss": 0.0207, + "step": 90490 + }, + { + "epoch": 5.929371683155343, + "grad_norm": 0.28433167934417725, + "learning_rate": 2.447685161517105e-06, + "loss": 0.0184, + "step": 90500 + }, + { + "epoch": 5.9300268623468515, + "grad_norm": 0.22962941229343414, + "learning_rate": 2.4425777389738224e-06, + "loss": 0.0351, + "step": 90510 + }, + { + "epoch": 5.930682041538361, + "grad_norm": 0.6661422252655029, + "learning_rate": 2.43747551723873e-06, + "loss": 0.0519, + "step": 90520 + }, + { + "epoch": 5.931337220729869, + "grad_norm": 0.14990508556365967, + "learning_rate": 2.4323784968698136e-06, + "loss": 0.0531, + "step": 90530 + }, + { + "epoch": 5.931992399921379, + "grad_norm": 0.5341704487800598, + "learning_rate": 2.4272866784244565e-06, + "loss": 0.0196, + "step": 90540 + }, + { + "epoch": 5.932647579112888, + "grad_norm": 0.34167200326919556, + "learning_rate": 2.422200062459518e-06, + "loss": 0.0295, + "step": 90550 + }, + { + "epoch": 5.9333027583043965, + "grad_norm": 0.9091720581054688, + "learning_rate": 2.41711864953123e-06, + "loss": 0.0267, + "step": 90560 + }, + { + "epoch": 5.933957937495905, + "grad_norm": 1.1764942407608032, + "learning_rate": 2.412042440195322e-06, + "loss": 0.0361, + "step": 90570 + }, + { + "epoch": 5.934613116687414, + "grad_norm": 0.575988233089447, + "learning_rate": 2.4069714350068993e-06, + "loss": 0.0301, + "step": 90580 + }, + { + "epoch": 5.935268295878923, + "grad_norm": 0.915009081363678, + "learning_rate": 2.4019056345205294e-06, + "loss": 0.0258, + "step": 90590 + }, + { + "epoch": 5.935923475070432, + "grad_norm": 0.7488800883293152, + "learning_rate": 2.396845039290191e-06, + "loss": 0.0526, + "step": 90600 + }, + { + "epoch": 5.9365786542619405, + "grad_norm": 1.119473934173584, + "learning_rate": 2.3917896498693194e-06, + "loss": 0.0228, + "step": 90610 + }, + { + "epoch": 5.937233833453449, + "grad_norm": 1.2374197244644165, + "learning_rate": 2.3867394668107335e-06, + "loss": 0.0532, + "step": 90620 + }, + { + "epoch": 5.937889012644958, + "grad_norm": 0.19195958971977234, + "learning_rate": 2.3816944906667527e-06, + "loss": 0.0259, + "step": 90630 + }, + { + "epoch": 5.938544191836467, + "grad_norm": 1.0884184837341309, + "learning_rate": 2.3766547219890634e-06, + "loss": 0.0487, + "step": 90640 + }, + { + "epoch": 5.939199371027976, + "grad_norm": 0.39645424485206604, + "learning_rate": 2.371620161328819e-06, + "loss": 0.0179, + "step": 90650 + }, + { + "epoch": 5.939854550219485, + "grad_norm": 0.9726547002792358, + "learning_rate": 2.366590809236574e-06, + "loss": 0.0259, + "step": 90660 + }, + { + "epoch": 5.940509729410994, + "grad_norm": 0.7121244072914124, + "learning_rate": 2.3615666662623494e-06, + "loss": 0.0349, + "step": 90670 + }, + { + "epoch": 5.941164908602503, + "grad_norm": 0.44517627358436584, + "learning_rate": 2.3565477329555664e-06, + "loss": 0.0182, + "step": 90680 + }, + { + "epoch": 5.941820087794012, + "grad_norm": 0.8601646423339844, + "learning_rate": 2.3515340098650975e-06, + "loss": 0.0363, + "step": 90690 + }, + { + "epoch": 5.942475266985521, + "grad_norm": 0.41367530822753906, + "learning_rate": 2.3465254975392258e-06, + "loss": 0.0305, + "step": 90700 + }, + { + "epoch": 5.94313044617703, + "grad_norm": 0.9844794273376465, + "learning_rate": 2.3415221965256805e-06, + "loss": 0.0335, + "step": 90710 + }, + { + "epoch": 5.943785625368538, + "grad_norm": 0.6284162402153015, + "learning_rate": 2.3365241073716237e-06, + "loss": 0.0244, + "step": 90720 + }, + { + "epoch": 5.944440804560047, + "grad_norm": 0.7016556859016418, + "learning_rate": 2.3315312306236183e-06, + "loss": 0.0219, + "step": 90730 + }, + { + "epoch": 5.945095983751556, + "grad_norm": 0.7999117374420166, + "learning_rate": 2.326543566827699e-06, + "loss": 0.0373, + "step": 90740 + }, + { + "epoch": 5.945751162943065, + "grad_norm": 0.6455729603767395, + "learning_rate": 2.3215611165292862e-06, + "loss": 0.0179, + "step": 90750 + }, + { + "epoch": 5.946406342134574, + "grad_norm": 0.8355702757835388, + "learning_rate": 2.3165838802732765e-06, + "loss": 0.0343, + "step": 90760 + }, + { + "epoch": 5.9470615213260825, + "grad_norm": 0.45217758417129517, + "learning_rate": 2.311611858603957e-06, + "loss": 0.0263, + "step": 90770 + }, + { + "epoch": 5.947716700517591, + "grad_norm": 0.966896653175354, + "learning_rate": 2.3066450520650763e-06, + "loss": 0.0305, + "step": 90780 + }, + { + "epoch": 5.9483718797091, + "grad_norm": 0.3814871609210968, + "learning_rate": 2.3016834611997773e-06, + "loss": 0.0402, + "step": 90790 + }, + { + "epoch": 5.94902705890061, + "grad_norm": 0.6488766074180603, + "learning_rate": 2.296727086550665e-06, + "loss": 0.0327, + "step": 90800 + }, + { + "epoch": 5.949682238092118, + "grad_norm": 0.2654576301574707, + "learning_rate": 2.2917759286597507e-06, + "loss": 0.0228, + "step": 90810 + }, + { + "epoch": 5.950337417283627, + "grad_norm": 0.0675271600484848, + "learning_rate": 2.2868299880685006e-06, + "loss": 0.0282, + "step": 90820 + }, + { + "epoch": 5.950992596475136, + "grad_norm": 0.2713579535484314, + "learning_rate": 2.2818892653177827e-06, + "loss": 0.0236, + "step": 90830 + }, + { + "epoch": 5.951647775666645, + "grad_norm": 0.3643115758895874, + "learning_rate": 2.2769537609479196e-06, + "loss": 0.077, + "step": 90840 + }, + { + "epoch": 5.952302954858154, + "grad_norm": 0.22309845685958862, + "learning_rate": 2.2720234754986357e-06, + "loss": 0.016, + "step": 90850 + }, + { + "epoch": 5.952958134049663, + "grad_norm": 0.18188640475273132, + "learning_rate": 2.2670984095091108e-06, + "loss": 0.019, + "step": 90860 + }, + { + "epoch": 5.9536133132411715, + "grad_norm": 0.34127289056777954, + "learning_rate": 2.262178563517936e-06, + "loss": 0.0252, + "step": 90870 + }, + { + "epoch": 5.95426849243268, + "grad_norm": 0.41627663373947144, + "learning_rate": 2.257263938063137e-06, + "loss": 0.0252, + "step": 90880 + }, + { + "epoch": 5.954923671624189, + "grad_norm": 0.12935373187065125, + "learning_rate": 2.2523545336821715e-06, + "loss": 0.0163, + "step": 90890 + }, + { + "epoch": 5.955578850815698, + "grad_norm": 0.07984542101621628, + "learning_rate": 2.247450350911939e-06, + "loss": 0.0339, + "step": 90900 + }, + { + "epoch": 5.956234030007207, + "grad_norm": 0.37867504358291626, + "learning_rate": 2.2425513902887328e-06, + "loss": 0.0402, + "step": 90910 + }, + { + "epoch": 5.956889209198716, + "grad_norm": 0.5668832063674927, + "learning_rate": 2.2376576523483015e-06, + "loss": 0.0383, + "step": 90920 + }, + { + "epoch": 5.957544388390224, + "grad_norm": 0.5368494391441345, + "learning_rate": 2.232769137625829e-06, + "loss": 0.0298, + "step": 90930 + }, + { + "epoch": 5.958199567581733, + "grad_norm": 0.7190997004508972, + "learning_rate": 2.227885846655903e-06, + "loss": 0.0235, + "step": 90940 + }, + { + "epoch": 5.958854746773243, + "grad_norm": 0.22133953869342804, + "learning_rate": 2.2230077799725523e-06, + "loss": 0.0299, + "step": 90950 + }, + { + "epoch": 5.959509925964752, + "grad_norm": 0.19012819230556488, + "learning_rate": 2.2181349381092396e-06, + "loss": 0.0317, + "step": 90960 + }, + { + "epoch": 5.9601651051562605, + "grad_norm": 0.6096690893173218, + "learning_rate": 2.2132673215988552e-06, + "loss": 0.0166, + "step": 90970 + }, + { + "epoch": 5.960820284347769, + "grad_norm": 0.1369543969631195, + "learning_rate": 2.2084049309737065e-06, + "loss": 0.0425, + "step": 90980 + }, + { + "epoch": 5.961475463539278, + "grad_norm": 0.7636327147483826, + "learning_rate": 2.2035477667655513e-06, + "loss": 0.0673, + "step": 90990 + }, + { + "epoch": 5.962130642730787, + "grad_norm": 0.7566077709197998, + "learning_rate": 2.1986958295055315e-06, + "loss": 0.0431, + "step": 91000 + }, + { + "epoch": 5.962785821922296, + "grad_norm": 0.25427183508872986, + "learning_rate": 2.193849119724284e-06, + "loss": 0.0326, + "step": 91010 + }, + { + "epoch": 5.963441001113805, + "grad_norm": 1.3639317750930786, + "learning_rate": 2.1890076379518175e-06, + "loss": 0.0445, + "step": 91020 + }, + { + "epoch": 5.964096180305313, + "grad_norm": 0.2204178273677826, + "learning_rate": 2.1841713847175925e-06, + "loss": 0.0274, + "step": 91030 + }, + { + "epoch": 5.964751359496822, + "grad_norm": 0.22641222178936005, + "learning_rate": 2.1793403605504968e-06, + "loss": 0.0484, + "step": 91040 + }, + { + "epoch": 5.965406538688331, + "grad_norm": 0.3744733929634094, + "learning_rate": 2.1745145659788413e-06, + "loss": 0.0341, + "step": 91050 + }, + { + "epoch": 5.96606171787984, + "grad_norm": 0.4841466248035431, + "learning_rate": 2.1696940015303645e-06, + "loss": 0.0271, + "step": 91060 + }, + { + "epoch": 5.966716897071349, + "grad_norm": 0.19782935082912445, + "learning_rate": 2.1648786677322397e-06, + "loss": 0.0103, + "step": 91070 + }, + { + "epoch": 5.967372076262858, + "grad_norm": 0.1290571093559265, + "learning_rate": 2.1600685651110673e-06, + "loss": 0.0333, + "step": 91080 + }, + { + "epoch": 5.968027255454366, + "grad_norm": 0.09424621611833572, + "learning_rate": 2.1552636941928717e-06, + "loss": 0.0109, + "step": 91090 + }, + { + "epoch": 5.968682434645876, + "grad_norm": 0.27166488766670227, + "learning_rate": 2.1504640555031043e-06, + "loss": 0.0366, + "step": 91100 + }, + { + "epoch": 5.969337613837385, + "grad_norm": 0.9257627129554749, + "learning_rate": 2.1456696495666506e-06, + "loss": 0.0204, + "step": 91110 + }, + { + "epoch": 5.969992793028894, + "grad_norm": 0.07065921276807785, + "learning_rate": 2.1408804769078084e-06, + "loss": 0.0194, + "step": 91120 + }, + { + "epoch": 5.970647972220402, + "grad_norm": 0.42005595564842224, + "learning_rate": 2.136096538050325e-06, + "loss": 0.0209, + "step": 91130 + }, + { + "epoch": 5.971303151411911, + "grad_norm": 0.15502072870731354, + "learning_rate": 2.13131783351736e-06, + "loss": 0.0251, + "step": 91140 + }, + { + "epoch": 5.97195833060342, + "grad_norm": 0.11996231228113174, + "learning_rate": 2.126544363831512e-06, + "loss": 0.0217, + "step": 91150 + }, + { + "epoch": 5.972613509794929, + "grad_norm": 0.2991523742675781, + "learning_rate": 2.121776129514791e-06, + "loss": 0.03, + "step": 91160 + }, + { + "epoch": 5.973268688986438, + "grad_norm": 0.21863000094890594, + "learning_rate": 2.1170131310886523e-06, + "loss": 0.0192, + "step": 91170 + }, + { + "epoch": 5.9739238681779465, + "grad_norm": 1.1284195184707642, + "learning_rate": 2.112255369073968e-06, + "loss": 0.0319, + "step": 91180 + }, + { + "epoch": 5.974579047369455, + "grad_norm": 0.18206311762332916, + "learning_rate": 2.1075028439910328e-06, + "loss": 0.0264, + "step": 91190 + }, + { + "epoch": 5.975234226560964, + "grad_norm": 0.10259415954351425, + "learning_rate": 2.1027555563595813e-06, + "loss": 0.0248, + "step": 91200 + }, + { + "epoch": 5.975889405752473, + "grad_norm": 0.31303083896636963, + "learning_rate": 2.0980135066987706e-06, + "loss": 0.017, + "step": 91210 + }, + { + "epoch": 5.976544584943982, + "grad_norm": 1.0888572931289673, + "learning_rate": 2.093276695527191e-06, + "loss": 0.03, + "step": 91220 + }, + { + "epoch": 5.9771997641354915, + "grad_norm": 0.5069630146026611, + "learning_rate": 2.0885451233628396e-06, + "loss": 0.0502, + "step": 91230 + }, + { + "epoch": 5.977854943327, + "grad_norm": 0.8863465189933777, + "learning_rate": 2.0838187907231633e-06, + "loss": 0.0301, + "step": 91240 + }, + { + "epoch": 5.978510122518509, + "grad_norm": 0.20493485033512115, + "learning_rate": 2.0790976981250153e-06, + "loss": 0.0489, + "step": 91250 + }, + { + "epoch": 5.979165301710018, + "grad_norm": 0.2513875663280487, + "learning_rate": 2.0743818460847e-06, + "loss": 0.0431, + "step": 91260 + }, + { + "epoch": 5.979820480901527, + "grad_norm": 0.27297183871269226, + "learning_rate": 2.069671235117926e-06, + "loss": 0.0477, + "step": 91270 + }, + { + "epoch": 5.9804756600930356, + "grad_norm": 0.5548529028892517, + "learning_rate": 2.0649658657398484e-06, + "loss": 0.0662, + "step": 91280 + }, + { + "epoch": 5.981130839284544, + "grad_norm": 0.20194190740585327, + "learning_rate": 2.060265738465034e-06, + "loss": 0.0208, + "step": 91290 + }, + { + "epoch": 5.981786018476053, + "grad_norm": 0.6909294128417969, + "learning_rate": 2.055570853807487e-06, + "loss": 0.0337, + "step": 91300 + }, + { + "epoch": 5.982441197667562, + "grad_norm": 0.22040200233459473, + "learning_rate": 2.0508812122806196e-06, + "loss": 0.0502, + "step": 91310 + }, + { + "epoch": 5.983096376859071, + "grad_norm": 1.9898004531860352, + "learning_rate": 2.046196814397294e-06, + "loss": 0.0248, + "step": 91320 + }, + { + "epoch": 5.98375155605058, + "grad_norm": 0.8370474576950073, + "learning_rate": 2.0415176606697896e-06, + "loss": 0.0332, + "step": 91330 + }, + { + "epoch": 5.984406735242088, + "grad_norm": 0.3414883613586426, + "learning_rate": 2.0368437516098126e-06, + "loss": 0.0277, + "step": 91340 + }, + { + "epoch": 5.985061914433597, + "grad_norm": 0.7309499979019165, + "learning_rate": 2.032175087728483e-06, + "loss": 0.0302, + "step": 91350 + }, + { + "epoch": 5.985717093625107, + "grad_norm": 0.6564101576805115, + "learning_rate": 2.027511669536375e-06, + "loss": 0.0547, + "step": 91360 + }, + { + "epoch": 5.986372272816615, + "grad_norm": 0.47627031803131104, + "learning_rate": 2.0228534975434643e-06, + "loss": 0.0343, + "step": 91370 + }, + { + "epoch": 5.987027452008125, + "grad_norm": 1.4242007732391357, + "learning_rate": 2.0182005722591548e-06, + "loss": 0.076, + "step": 91380 + }, + { + "epoch": 5.987682631199633, + "grad_norm": 0.6044320464134216, + "learning_rate": 2.0135528941922944e-06, + "loss": 0.0393, + "step": 91390 + }, + { + "epoch": 5.988337810391142, + "grad_norm": 0.6391236782073975, + "learning_rate": 2.008910463851149e-06, + "loss": 0.0441, + "step": 91400 + }, + { + "epoch": 5.988992989582651, + "grad_norm": 0.1356344074010849, + "learning_rate": 2.0042732817433963e-06, + "loss": 0.0354, + "step": 91410 + }, + { + "epoch": 5.98964816877416, + "grad_norm": 0.08415236324071884, + "learning_rate": 1.999641348376158e-06, + "loss": 0.0201, + "step": 91420 + }, + { + "epoch": 5.990303347965669, + "grad_norm": 0.09109601378440857, + "learning_rate": 1.9950146642559787e-06, + "loss": 0.0263, + "step": 91430 + }, + { + "epoch": 5.9909585271571775, + "grad_norm": 1.1000856161117554, + "learning_rate": 1.990393229888815e-06, + "loss": 0.025, + "step": 91440 + }, + { + "epoch": 5.991613706348686, + "grad_norm": 0.3896036744117737, + "learning_rate": 1.9857770457800684e-06, + "loss": 0.0235, + "step": 91450 + }, + { + "epoch": 5.992268885540195, + "grad_norm": 0.7391855120658875, + "learning_rate": 1.9811661124345517e-06, + "loss": 0.015, + "step": 91460 + }, + { + "epoch": 5.992924064731704, + "grad_norm": 0.23673787713050842, + "learning_rate": 1.976560430356522e-06, + "loss": 0.0193, + "step": 91470 + }, + { + "epoch": 5.993579243923213, + "grad_norm": 0.6467635631561279, + "learning_rate": 1.9719600000496383e-06, + "loss": 0.0371, + "step": 91480 + }, + { + "epoch": 5.9942344231147215, + "grad_norm": 0.08966309577226639, + "learning_rate": 1.967364822017004e-06, + "loss": 0.0375, + "step": 91490 + }, + { + "epoch": 5.99488960230623, + "grad_norm": 0.15109893679618835, + "learning_rate": 1.962774896761127e-06, + "loss": 0.0446, + "step": 91500 + }, + { + "epoch": 5.99554478149774, + "grad_norm": 0.31270959973335266, + "learning_rate": 1.958190224783968e-06, + "loss": 0.0245, + "step": 91510 + }, + { + "epoch": 5.996199960689249, + "grad_norm": 0.45211392641067505, + "learning_rate": 1.9536108065868975e-06, + "loss": 0.0519, + "step": 91520 + }, + { + "epoch": 5.996855139880758, + "grad_norm": 0.3514050543308258, + "learning_rate": 1.949036642670715e-06, + "loss": 0.0406, + "step": 91530 + }, + { + "epoch": 5.9975103190722665, + "grad_norm": 1.0405356884002686, + "learning_rate": 1.9444677335356377e-06, + "loss": 0.0201, + "step": 91540 + }, + { + "epoch": 5.998165498263775, + "grad_norm": 0.12399162352085114, + "learning_rate": 1.939904079681326e-06, + "loss": 0.0291, + "step": 91550 + }, + { + "epoch": 5.998820677455284, + "grad_norm": 0.33782052993774414, + "learning_rate": 1.935345681606837e-06, + "loss": 0.0319, + "step": 91560 + }, + { + "epoch": 5.999475856646793, + "grad_norm": 0.3364892899990082, + "learning_rate": 1.930792539810694e-06, + "loss": 0.0364, + "step": 91570 + }, + { + "epoch": 6.000131035838302, + "grad_norm": 0.15649165213108063, + "learning_rate": 1.9262446547907877e-06, + "loss": 0.0505, + "step": 91580 + }, + { + "epoch": 6.000786215029811, + "grad_norm": 0.622900128364563, + "learning_rate": 1.921702027044503e-06, + "loss": 0.0207, + "step": 91590 + }, + { + "epoch": 6.001441394221319, + "grad_norm": 0.9078409671783447, + "learning_rate": 1.917164657068593e-06, + "loss": 0.0434, + "step": 91600 + }, + { + "epoch": 6.002096573412828, + "grad_norm": 0.9378566145896912, + "learning_rate": 1.912632545359272e-06, + "loss": 0.0498, + "step": 91610 + }, + { + "epoch": 6.002751752604337, + "grad_norm": 0.14544491469860077, + "learning_rate": 1.9081056924121488e-06, + "loss": 0.0138, + "step": 91620 + }, + { + "epoch": 6.003406931795846, + "grad_norm": 1.2220262289047241, + "learning_rate": 1.9035840987222887e-06, + "loss": 0.0232, + "step": 91630 + }, + { + "epoch": 6.004062110987355, + "grad_norm": 1.0094246864318848, + "learning_rate": 1.8990677647841515e-06, + "loss": 0.0525, + "step": 91640 + }, + { + "epoch": 6.004717290178864, + "grad_norm": 0.2626262903213501, + "learning_rate": 1.8945566910916534e-06, + "loss": 0.021, + "step": 91650 + }, + { + "epoch": 6.005372469370373, + "grad_norm": 0.19241881370544434, + "learning_rate": 1.8900508781381054e-06, + "loss": 0.0448, + "step": 91660 + }, + { + "epoch": 6.006027648561882, + "grad_norm": 0.4814234972000122, + "learning_rate": 1.8855503264162633e-06, + "loss": 0.0322, + "step": 91670 + }, + { + "epoch": 6.006682827753391, + "grad_norm": 1.3240758180618286, + "learning_rate": 1.8810550364182998e-06, + "loss": 0.0341, + "step": 91680 + }, + { + "epoch": 6.0073380069449, + "grad_norm": 1.0488985776901245, + "learning_rate": 1.8765650086358112e-06, + "loss": 0.0241, + "step": 91690 + }, + { + "epoch": 6.007993186136408, + "grad_norm": 0.3021407425403595, + "learning_rate": 1.8720802435598151e-06, + "loss": 0.033, + "step": 91700 + }, + { + "epoch": 6.008648365327917, + "grad_norm": 0.16907329857349396, + "learning_rate": 1.8676007416807694e-06, + "loss": 0.0171, + "step": 91710 + }, + { + "epoch": 6.009303544519426, + "grad_norm": 0.8404820561408997, + "learning_rate": 1.8631265034885437e-06, + "loss": 0.0419, + "step": 91720 + }, + { + "epoch": 6.009958723710935, + "grad_norm": 0.5962705016136169, + "learning_rate": 1.858657529472424e-06, + "loss": 0.0349, + "step": 91730 + }, + { + "epoch": 6.010613902902444, + "grad_norm": 0.47190725803375244, + "learning_rate": 1.8541938201211473e-06, + "loss": 0.0189, + "step": 91740 + }, + { + "epoch": 6.0112690820939525, + "grad_norm": 0.5304276347160339, + "learning_rate": 1.8497353759228398e-06, + "loss": 0.029, + "step": 91750 + }, + { + "epoch": 6.011924261285461, + "grad_norm": 1.0818222761154175, + "learning_rate": 1.8452821973650835e-06, + "loss": 0.0165, + "step": 91760 + }, + { + "epoch": 6.01257944047697, + "grad_norm": 0.20692627131938934, + "learning_rate": 1.840834284934856e-06, + "loss": 0.0276, + "step": 91770 + }, + { + "epoch": 6.013234619668479, + "grad_norm": 0.9945173263549805, + "learning_rate": 1.8363916391185954e-06, + "loss": 0.0204, + "step": 91780 + }, + { + "epoch": 6.013889798859989, + "grad_norm": 0.35558536648750305, + "learning_rate": 1.8319542604021245e-06, + "loss": 0.043, + "step": 91790 + }, + { + "epoch": 6.0145449780514975, + "grad_norm": 0.7652702927589417, + "learning_rate": 1.8275221492707273e-06, + "loss": 0.0577, + "step": 91800 + }, + { + "epoch": 6.015200157243006, + "grad_norm": 0.3940947651863098, + "learning_rate": 1.8230953062090717e-06, + "loss": 0.0179, + "step": 91810 + }, + { + "epoch": 6.015855336434515, + "grad_norm": 0.6468888521194458, + "learning_rate": 1.8186737317012926e-06, + "loss": 0.0311, + "step": 91820 + }, + { + "epoch": 6.016510515626024, + "grad_norm": 1.7487236261367798, + "learning_rate": 1.8142574262308977e-06, + "loss": 0.0277, + "step": 91830 + }, + { + "epoch": 6.017165694817533, + "grad_norm": 0.3217769265174866, + "learning_rate": 1.8098463902808727e-06, + "loss": 0.0236, + "step": 91840 + }, + { + "epoch": 6.0178208740090415, + "grad_norm": 0.8316143751144409, + "learning_rate": 1.805440624333593e-06, + "loss": 0.0284, + "step": 91850 + }, + { + "epoch": 6.01847605320055, + "grad_norm": 0.25615644454956055, + "learning_rate": 1.8010401288708734e-06, + "loss": 0.0155, + "step": 91860 + }, + { + "epoch": 6.019131232392059, + "grad_norm": 0.5413250923156738, + "learning_rate": 1.7966449043739343e-06, + "loss": 0.0207, + "step": 91870 + }, + { + "epoch": 6.019786411583568, + "grad_norm": 0.8675181269645691, + "learning_rate": 1.79225495132343e-06, + "loss": 0.0404, + "step": 91880 + }, + { + "epoch": 6.020441590775077, + "grad_norm": 0.33421337604522705, + "learning_rate": 1.7878702701994487e-06, + "loss": 0.0186, + "step": 91890 + }, + { + "epoch": 6.021096769966586, + "grad_norm": 0.16367676854133606, + "learning_rate": 1.783490861481496e-06, + "loss": 0.039, + "step": 91900 + }, + { + "epoch": 6.021751949158094, + "grad_norm": 0.4865973889827728, + "learning_rate": 1.7791167256484832e-06, + "loss": 0.0172, + "step": 91910 + }, + { + "epoch": 6.022407128349603, + "grad_norm": 1.078665852546692, + "learning_rate": 1.774747863178766e-06, + "loss": 0.0777, + "step": 91920 + }, + { + "epoch": 6.023062307541113, + "grad_norm": 0.29245615005493164, + "learning_rate": 1.7703842745501186e-06, + "loss": 0.0235, + "step": 91930 + }, + { + "epoch": 6.023717486732622, + "grad_norm": 0.7222318649291992, + "learning_rate": 1.7660259602397366e-06, + "loss": 0.0306, + "step": 91940 + }, + { + "epoch": 6.024372665924131, + "grad_norm": 0.10513530671596527, + "learning_rate": 1.7616729207242388e-06, + "loss": 0.035, + "step": 91950 + }, + { + "epoch": 6.025027845115639, + "grad_norm": 0.38612493872642517, + "learning_rate": 1.7573251564796556e-06, + "loss": 0.0357, + "step": 91960 + }, + { + "epoch": 6.025683024307148, + "grad_norm": 0.21932445466518402, + "learning_rate": 1.7529826679814676e-06, + "loss": 0.0363, + "step": 91970 + }, + { + "epoch": 6.026338203498657, + "grad_norm": 0.7902799248695374, + "learning_rate": 1.7486454557045562e-06, + "loss": 0.0444, + "step": 91980 + }, + { + "epoch": 6.026993382690166, + "grad_norm": 0.531852126121521, + "learning_rate": 1.7443135201232364e-06, + "loss": 0.0164, + "step": 91990 + }, + { + "epoch": 6.027648561881675, + "grad_norm": 0.8255022764205933, + "learning_rate": 1.7399868617112347e-06, + "loss": 0.0244, + "step": 92000 + }, + { + "epoch": 6.0283037410731835, + "grad_norm": 0.4618980586528778, + "learning_rate": 1.7356654809417171e-06, + "loss": 0.0363, + "step": 92010 + }, + { + "epoch": 6.028958920264692, + "grad_norm": 0.20629416406154633, + "learning_rate": 1.7313493782872448e-06, + "loss": 0.0279, + "step": 92020 + }, + { + "epoch": 6.029614099456201, + "grad_norm": 0.3304322063922882, + "learning_rate": 1.7270385542198453e-06, + "loss": 0.0165, + "step": 92030 + }, + { + "epoch": 6.03026927864771, + "grad_norm": 0.1086028441786766, + "learning_rate": 1.7227330092109307e-06, + "loss": 0.0397, + "step": 92040 + }, + { + "epoch": 6.030924457839219, + "grad_norm": 0.2330922782421112, + "learning_rate": 1.718432743731352e-06, + "loss": 0.0281, + "step": 92050 + }, + { + "epoch": 6.0315796370307275, + "grad_norm": 0.2216382920742035, + "learning_rate": 1.7141377582513718e-06, + "loss": 0.0312, + "step": 92060 + }, + { + "epoch": 6.032234816222237, + "grad_norm": 0.3841842710971832, + "learning_rate": 1.709848053240698e-06, + "loss": 0.0268, + "step": 92070 + }, + { + "epoch": 6.032889995413746, + "grad_norm": 0.5668713450431824, + "learning_rate": 1.7055636291684328e-06, + "loss": 0.0225, + "step": 92080 + }, + { + "epoch": 6.033545174605255, + "grad_norm": 1.4545365571975708, + "learning_rate": 1.7012844865031185e-06, + "loss": 0.0544, + "step": 92090 + }, + { + "epoch": 6.034200353796764, + "grad_norm": 0.25465086102485657, + "learning_rate": 1.6970106257127138e-06, + "loss": 0.0272, + "step": 92100 + }, + { + "epoch": 6.0348555329882725, + "grad_norm": 1.9110933542251587, + "learning_rate": 1.6927420472646117e-06, + "loss": 0.0318, + "step": 92110 + }, + { + "epoch": 6.035510712179781, + "grad_norm": 1.3342527151107788, + "learning_rate": 1.6884787516256106e-06, + "loss": 0.0395, + "step": 92120 + }, + { + "epoch": 6.03616589137129, + "grad_norm": 0.2612208127975464, + "learning_rate": 1.6842207392619325e-06, + "loss": 0.0458, + "step": 92130 + }, + { + "epoch": 6.036821070562799, + "grad_norm": 0.23202545940876007, + "learning_rate": 1.679968010639238e-06, + "loss": 0.0118, + "step": 92140 + }, + { + "epoch": 6.037476249754308, + "grad_norm": 1.06582772731781, + "learning_rate": 1.6757205662225829e-06, + "loss": 0.0284, + "step": 92150 + }, + { + "epoch": 6.038131428945817, + "grad_norm": 0.7253896594047546, + "learning_rate": 1.6714784064764844e-06, + "loss": 0.0242, + "step": 92160 + }, + { + "epoch": 6.038786608137325, + "grad_norm": 0.48443248867988586, + "learning_rate": 1.6672415318648383e-06, + "loss": 0.0411, + "step": 92170 + }, + { + "epoch": 6.039441787328834, + "grad_norm": 0.36455070972442627, + "learning_rate": 1.6630099428510016e-06, + "loss": 0.0572, + "step": 92180 + }, + { + "epoch": 6.040096966520343, + "grad_norm": 0.15781891345977783, + "learning_rate": 1.6587836398977153e-06, + "loss": 0.0223, + "step": 92190 + }, + { + "epoch": 6.040752145711852, + "grad_norm": 0.9970755577087402, + "learning_rate": 1.6545626234671762e-06, + "loss": 0.0474, + "step": 92200 + }, + { + "epoch": 6.041407324903361, + "grad_norm": 0.805385410785675, + "learning_rate": 1.650346894020971e-06, + "loss": 0.0311, + "step": 92210 + }, + { + "epoch": 6.04206250409487, + "grad_norm": 0.3012971580028534, + "learning_rate": 1.6461364520201472e-06, + "loss": 0.0235, + "step": 92220 + }, + { + "epoch": 6.042717683286379, + "grad_norm": 0.38665199279785156, + "learning_rate": 1.6419312979251366e-06, + "loss": 0.0269, + "step": 92230 + }, + { + "epoch": 6.043372862477888, + "grad_norm": 0.30529114603996277, + "learning_rate": 1.6377314321958214e-06, + "loss": 0.0203, + "step": 92240 + }, + { + "epoch": 6.044028041669397, + "grad_norm": 0.7312307357788086, + "learning_rate": 1.6335368552914787e-06, + "loss": 0.023, + "step": 92250 + }, + { + "epoch": 6.044683220860906, + "grad_norm": 0.43789857625961304, + "learning_rate": 1.6293475676708358e-06, + "loss": 0.0255, + "step": 92260 + }, + { + "epoch": 6.045338400052414, + "grad_norm": 1.8862483501434326, + "learning_rate": 1.6251635697920097e-06, + "loss": 0.0383, + "step": 92270 + }, + { + "epoch": 6.045993579243923, + "grad_norm": 0.4681839644908905, + "learning_rate": 1.6209848621125624e-06, + "loss": 0.0317, + "step": 92280 + }, + { + "epoch": 6.046648758435432, + "grad_norm": 0.8216774463653564, + "learning_rate": 1.6168114450894778e-06, + "loss": 0.0315, + "step": 92290 + }, + { + "epoch": 6.047303937626941, + "grad_norm": 1.2242127656936646, + "learning_rate": 1.612643319179158e-06, + "loss": 0.0206, + "step": 92300 + }, + { + "epoch": 6.04795911681845, + "grad_norm": 0.2813614308834076, + "learning_rate": 1.6084804848374046e-06, + "loss": 0.014, + "step": 92310 + }, + { + "epoch": 6.0486142960099585, + "grad_norm": 0.8464245200157166, + "learning_rate": 1.6043229425194817e-06, + "loss": 0.0398, + "step": 92320 + }, + { + "epoch": 6.049269475201467, + "grad_norm": 0.6298983693122864, + "learning_rate": 1.6001706926800308e-06, + "loss": 0.0334, + "step": 92330 + }, + { + "epoch": 6.049924654392976, + "grad_norm": 0.10295037925243378, + "learning_rate": 1.5960237357731444e-06, + "loss": 0.0243, + "step": 92340 + }, + { + "epoch": 6.050579833584485, + "grad_norm": 0.32806679606437683, + "learning_rate": 1.5918820722523264e-06, + "loss": 0.0293, + "step": 92350 + }, + { + "epoch": 6.051235012775995, + "grad_norm": 0.3083593249320984, + "learning_rate": 1.587745702570509e-06, + "loss": 0.0221, + "step": 92360 + }, + { + "epoch": 6.051890191967503, + "grad_norm": 0.22596286237239838, + "learning_rate": 1.5836146271800356e-06, + "loss": 0.0359, + "step": 92370 + }, + { + "epoch": 6.052545371159012, + "grad_norm": 0.5513815879821777, + "learning_rate": 1.5794888465326728e-06, + "loss": 0.0324, + "step": 92380 + }, + { + "epoch": 6.053200550350521, + "grad_norm": 0.14125698804855347, + "learning_rate": 1.5753683610796155e-06, + "loss": 0.0273, + "step": 92390 + }, + { + "epoch": 6.05385572954203, + "grad_norm": 0.808043897151947, + "learning_rate": 1.5712531712714528e-06, + "loss": 0.0328, + "step": 92400 + }, + { + "epoch": 6.054510908733539, + "grad_norm": 0.6514552235603333, + "learning_rate": 1.5671432775582474e-06, + "loss": 0.0457, + "step": 92410 + }, + { + "epoch": 6.0551660879250475, + "grad_norm": 0.2788650393486023, + "learning_rate": 1.563038680389428e-06, + "loss": 0.0292, + "step": 92420 + }, + { + "epoch": 6.055821267116556, + "grad_norm": 0.7894077897071838, + "learning_rate": 1.5589393802138808e-06, + "loss": 0.0377, + "step": 92430 + }, + { + "epoch": 6.056476446308065, + "grad_norm": 0.3357158899307251, + "learning_rate": 1.554845377479891e-06, + "loss": 0.0283, + "step": 92440 + }, + { + "epoch": 6.057131625499574, + "grad_norm": 0.844900369644165, + "learning_rate": 1.550756672635173e-06, + "loss": 0.0346, + "step": 92450 + }, + { + "epoch": 6.057786804691083, + "grad_norm": 0.2523321807384491, + "learning_rate": 1.5466732661268635e-06, + "loss": 0.0263, + "step": 92460 + }, + { + "epoch": 6.058441983882592, + "grad_norm": 0.23931458592414856, + "learning_rate": 1.5425951584015163e-06, + "loss": 0.06, + "step": 92470 + }, + { + "epoch": 6.0590971630741, + "grad_norm": 1.027948260307312, + "learning_rate": 1.5385223499051026e-06, + "loss": 0.0254, + "step": 92480 + }, + { + "epoch": 6.059752342265609, + "grad_norm": 0.4777188301086426, + "learning_rate": 1.5344548410830328e-06, + "loss": 0.0285, + "step": 92490 + }, + { + "epoch": 6.060407521457119, + "grad_norm": 0.7800927758216858, + "learning_rate": 1.5303926323801121e-06, + "loss": 0.0272, + "step": 92500 + }, + { + "epoch": 6.061062700648628, + "grad_norm": 0.27573302388191223, + "learning_rate": 1.5263357242405795e-06, + "loss": 0.0318, + "step": 92510 + }, + { + "epoch": 6.0617178798401365, + "grad_norm": 0.17664968967437744, + "learning_rate": 1.5222841171080915e-06, + "loss": 0.0184, + "step": 92520 + }, + { + "epoch": 6.062373059031645, + "grad_norm": 0.4265410006046295, + "learning_rate": 1.5182378114257268e-06, + "loss": 0.0182, + "step": 92530 + }, + { + "epoch": 6.063028238223154, + "grad_norm": 0.6084480285644531, + "learning_rate": 1.5141968076359869e-06, + "loss": 0.026, + "step": 92540 + }, + { + "epoch": 6.063683417414663, + "grad_norm": 0.3512929677963257, + "learning_rate": 1.5101611061807908e-06, + "loss": 0.0228, + "step": 92550 + }, + { + "epoch": 6.064338596606172, + "grad_norm": 0.9465004205703735, + "learning_rate": 1.506130707501463e-06, + "loss": 0.0282, + "step": 92560 + }, + { + "epoch": 6.064993775797681, + "grad_norm": 0.403349369764328, + "learning_rate": 1.5021056120387734e-06, + "loss": 0.0559, + "step": 92570 + }, + { + "epoch": 6.065648954989189, + "grad_norm": 1.0841121673583984, + "learning_rate": 1.4980858202329084e-06, + "loss": 0.0756, + "step": 92580 + }, + { + "epoch": 6.066304134180698, + "grad_norm": 0.6392319202423096, + "learning_rate": 1.4940713325234445e-06, + "loss": 0.0113, + "step": 92590 + }, + { + "epoch": 6.066959313372207, + "grad_norm": 0.3358542323112488, + "learning_rate": 1.4900621493494137e-06, + "loss": 0.0343, + "step": 92600 + }, + { + "epoch": 6.067614492563716, + "grad_norm": 1.1501914262771606, + "learning_rate": 1.4860582711492544e-06, + "loss": 0.0317, + "step": 92610 + }, + { + "epoch": 6.068269671755225, + "grad_norm": 0.20030245184898376, + "learning_rate": 1.4820596983608325e-06, + "loss": 0.0249, + "step": 92620 + }, + { + "epoch": 6.0689248509467335, + "grad_norm": 0.07033828645944595, + "learning_rate": 1.4780664314214044e-06, + "loss": 0.0117, + "step": 92630 + }, + { + "epoch": 6.069580030138243, + "grad_norm": 0.7970473766326904, + "learning_rate": 1.474078470767687e-06, + "loss": 0.0293, + "step": 92640 + }, + { + "epoch": 6.070235209329752, + "grad_norm": 0.3490099310874939, + "learning_rate": 1.470095816835787e-06, + "loss": 0.0209, + "step": 92650 + }, + { + "epoch": 6.070890388521261, + "grad_norm": 0.6057643890380859, + "learning_rate": 1.4661184700612395e-06, + "loss": 0.0202, + "step": 92660 + }, + { + "epoch": 6.07154556771277, + "grad_norm": 0.5311042666435242, + "learning_rate": 1.4621464308790133e-06, + "loss": 0.0382, + "step": 92670 + }, + { + "epoch": 6.0722007469042785, + "grad_norm": 0.34993597865104675, + "learning_rate": 1.4581796997234775e-06, + "loss": 0.0242, + "step": 92680 + }, + { + "epoch": 6.072855926095787, + "grad_norm": 0.2856265902519226, + "learning_rate": 1.4542182770284297e-06, + "loss": 0.0313, + "step": 92690 + }, + { + "epoch": 6.073511105287296, + "grad_norm": 0.6761451363563538, + "learning_rate": 1.450262163227084e-06, + "loss": 0.0243, + "step": 92700 + }, + { + "epoch": 6.074166284478805, + "grad_norm": 0.9251343011856079, + "learning_rate": 1.4463113587520671e-06, + "loss": 0.0328, + "step": 92710 + }, + { + "epoch": 6.074821463670314, + "grad_norm": 0.25004640221595764, + "learning_rate": 1.4423658640354443e-06, + "loss": 0.0242, + "step": 92720 + }, + { + "epoch": 6.0754766428618225, + "grad_norm": 0.620094358921051, + "learning_rate": 1.438425679508687e-06, + "loss": 0.038, + "step": 92730 + }, + { + "epoch": 6.076131822053331, + "grad_norm": 0.6321430206298828, + "learning_rate": 1.4344908056026895e-06, + "loss": 0.0335, + "step": 92740 + }, + { + "epoch": 6.07678700124484, + "grad_norm": 1.773182988166809, + "learning_rate": 1.4305612427477521e-06, + "loss": 0.0169, + "step": 92750 + }, + { + "epoch": 6.077442180436349, + "grad_norm": 0.40506991744041443, + "learning_rate": 1.42663699137362e-06, + "loss": 0.0313, + "step": 92760 + }, + { + "epoch": 6.078097359627858, + "grad_norm": 0.5896541476249695, + "learning_rate": 1.4227180519094386e-06, + "loss": 0.0265, + "step": 92770 + }, + { + "epoch": 6.0787525388193675, + "grad_norm": 0.4388304054737091, + "learning_rate": 1.4188044247837706e-06, + "loss": 0.0147, + "step": 92780 + }, + { + "epoch": 6.079407718010876, + "grad_norm": 0.4507558345794678, + "learning_rate": 1.4148961104246128e-06, + "loss": 0.0316, + "step": 92790 + }, + { + "epoch": 6.080062897202385, + "grad_norm": 0.4970085620880127, + "learning_rate": 1.4109931092593731e-06, + "loss": 0.0318, + "step": 92800 + }, + { + "epoch": 6.080718076393894, + "grad_norm": 0.4158879816532135, + "learning_rate": 1.4070954217148658e-06, + "loss": 0.019, + "step": 92810 + }, + { + "epoch": 6.081373255585403, + "grad_norm": 0.1101650521159172, + "learning_rate": 1.4032030482173497e-06, + "loss": 0.02, + "step": 92820 + }, + { + "epoch": 6.082028434776912, + "grad_norm": 0.4452461302280426, + "learning_rate": 1.3993159891924901e-06, + "loss": 0.0246, + "step": 92830 + }, + { + "epoch": 6.08268361396842, + "grad_norm": 0.3194720447063446, + "learning_rate": 1.3954342450653524e-06, + "loss": 0.0199, + "step": 92840 + }, + { + "epoch": 6.083338793159929, + "grad_norm": 0.46862828731536865, + "learning_rate": 1.3915578162604582e-06, + "loss": 0.0195, + "step": 92850 + }, + { + "epoch": 6.083993972351438, + "grad_norm": 0.2374069094657898, + "learning_rate": 1.3876867032017126e-06, + "loss": 0.0219, + "step": 92860 + }, + { + "epoch": 6.084649151542947, + "grad_norm": 0.17564623057842255, + "learning_rate": 1.383820906312472e-06, + "loss": 0.0281, + "step": 92870 + }, + { + "epoch": 6.085304330734456, + "grad_norm": 0.5607072710990906, + "learning_rate": 1.3799604260154752e-06, + "loss": 0.0331, + "step": 92880 + }, + { + "epoch": 6.0859595099259645, + "grad_norm": 0.17764176428318024, + "learning_rate": 1.376105262732913e-06, + "loss": 0.0295, + "step": 92890 + }, + { + "epoch": 6.086614689117473, + "grad_norm": 0.3563988208770752, + "learning_rate": 1.37225541688637e-06, + "loss": 0.0185, + "step": 92900 + }, + { + "epoch": 6.087269868308982, + "grad_norm": 0.5337191820144653, + "learning_rate": 1.3684108888968594e-06, + "loss": 0.0242, + "step": 92910 + }, + { + "epoch": 6.087925047500492, + "grad_norm": 0.20622415840625763, + "learning_rate": 1.364571679184823e-06, + "loss": 0.0397, + "step": 92920 + }, + { + "epoch": 6.088580226692001, + "grad_norm": 1.0572715997695923, + "learning_rate": 1.3607377881701077e-06, + "loss": 0.0272, + "step": 92930 + }, + { + "epoch": 6.089235405883509, + "grad_norm": 0.7551186084747314, + "learning_rate": 1.3569092162719787e-06, + "loss": 0.0359, + "step": 92940 + }, + { + "epoch": 6.089890585075018, + "grad_norm": 0.29744282364845276, + "learning_rate": 1.353085963909123e-06, + "loss": 0.0206, + "step": 92950 + }, + { + "epoch": 6.090545764266527, + "grad_norm": 0.22721756994724274, + "learning_rate": 1.3492680314996454e-06, + "loss": 0.0358, + "step": 92960 + }, + { + "epoch": 6.091200943458036, + "grad_norm": 0.09489724785089493, + "learning_rate": 1.345455419461078e-06, + "loss": 0.0293, + "step": 92970 + }, + { + "epoch": 6.091856122649545, + "grad_norm": 0.3256237506866455, + "learning_rate": 1.3416481282103431e-06, + "loss": 0.0375, + "step": 92980 + }, + { + "epoch": 6.0925113018410535, + "grad_norm": 0.9793704152107239, + "learning_rate": 1.337846158163819e-06, + "loss": 0.0109, + "step": 92990 + }, + { + "epoch": 6.093166481032562, + "grad_norm": 0.2780138850212097, + "learning_rate": 1.334049509737273e-06, + "loss": 0.016, + "step": 93000 + }, + { + "epoch": 6.093821660224071, + "grad_norm": 0.6005891561508179, + "learning_rate": 1.3302581833459116e-06, + "loss": 0.019, + "step": 93010 + }, + { + "epoch": 6.09447683941558, + "grad_norm": 0.31295061111450195, + "learning_rate": 1.3264721794043366e-06, + "loss": 0.0427, + "step": 93020 + }, + { + "epoch": 6.095132018607089, + "grad_norm": 0.3039318025112152, + "learning_rate": 1.3226914983265836e-06, + "loss": 0.0183, + "step": 93030 + }, + { + "epoch": 6.095787197798598, + "grad_norm": 0.1998998373746872, + "learning_rate": 1.3189161405260997e-06, + "loss": 0.0172, + "step": 93040 + }, + { + "epoch": 6.096442376990106, + "grad_norm": 0.22414325177669525, + "learning_rate": 1.3151461064157656e-06, + "loss": 0.0167, + "step": 93050 + }, + { + "epoch": 6.097097556181616, + "grad_norm": 0.3385004997253418, + "learning_rate": 1.3113813964078459e-06, + "loss": 0.0196, + "step": 93060 + }, + { + "epoch": 6.097752735373125, + "grad_norm": 0.5438088774681091, + "learning_rate": 1.3076220109140614e-06, + "loss": 0.0205, + "step": 93070 + }, + { + "epoch": 6.098407914564634, + "grad_norm": 0.617360532283783, + "learning_rate": 1.3038679503455275e-06, + "loss": 0.0302, + "step": 93080 + }, + { + "epoch": 6.0990630937561425, + "grad_norm": 0.07352149486541748, + "learning_rate": 1.3001192151127773e-06, + "loss": 0.0128, + "step": 93090 + }, + { + "epoch": 6.099718272947651, + "grad_norm": 0.7426431179046631, + "learning_rate": 1.2963758056257657e-06, + "loss": 0.0323, + "step": 93100 + }, + { + "epoch": 6.10037345213916, + "grad_norm": 0.23046189546585083, + "learning_rate": 1.2926377222938768e-06, + "loss": 0.0229, + "step": 93110 + }, + { + "epoch": 6.101028631330669, + "grad_norm": 0.21610227227210999, + "learning_rate": 1.2889049655259e-06, + "loss": 0.0497, + "step": 93120 + }, + { + "epoch": 6.101683810522178, + "grad_norm": 0.4827043414115906, + "learning_rate": 1.2851775357300312e-06, + "loss": 0.0282, + "step": 93130 + }, + { + "epoch": 6.102338989713687, + "grad_norm": 1.6437174081802368, + "learning_rate": 1.281455433313916e-06, + "loss": 0.0607, + "step": 93140 + }, + { + "epoch": 6.102994168905195, + "grad_norm": 0.1425211876630783, + "learning_rate": 1.2777386586845796e-06, + "loss": 0.0228, + "step": 93150 + }, + { + "epoch": 6.103649348096704, + "grad_norm": 0.8335294127464294, + "learning_rate": 1.2740272122484964e-06, + "loss": 0.0233, + "step": 93160 + }, + { + "epoch": 6.104304527288213, + "grad_norm": 0.3346331715583801, + "learning_rate": 1.2703210944115308e-06, + "loss": 0.0158, + "step": 93170 + }, + { + "epoch": 6.104959706479722, + "grad_norm": 0.44995129108428955, + "learning_rate": 1.2666203055789916e-06, + "loss": 0.0236, + "step": 93180 + }, + { + "epoch": 6.105614885671231, + "grad_norm": 0.98499596118927, + "learning_rate": 1.2629248461555832e-06, + "loss": 0.0334, + "step": 93190 + }, + { + "epoch": 6.10627006486274, + "grad_norm": 0.5737144947052002, + "learning_rate": 1.2592347165454431e-06, + "loss": 0.0488, + "step": 93200 + }, + { + "epoch": 6.106925244054249, + "grad_norm": 0.3024933636188507, + "learning_rate": 1.2555499171521102e-06, + "loss": 0.0267, + "step": 93210 + }, + { + "epoch": 6.107580423245758, + "grad_norm": 0.551342785358429, + "learning_rate": 1.251870448378556e-06, + "loss": 0.0365, + "step": 93220 + }, + { + "epoch": 6.108235602437267, + "grad_norm": 0.5646647214889526, + "learning_rate": 1.2481963106271478e-06, + "loss": 0.032, + "step": 93230 + }, + { + "epoch": 6.108890781628776, + "grad_norm": 0.9741244912147522, + "learning_rate": 1.2445275042997084e-06, + "loss": 0.0387, + "step": 93240 + }, + { + "epoch": 6.1095459608202844, + "grad_norm": 2.142883062362671, + "learning_rate": 1.2408640297974283e-06, + "loss": 0.0303, + "step": 93250 + }, + { + "epoch": 6.110201140011793, + "grad_norm": 1.2530969381332397, + "learning_rate": 1.2372058875209535e-06, + "loss": 0.0247, + "step": 93260 + }, + { + "epoch": 6.110856319203302, + "grad_norm": 0.15142413973808289, + "learning_rate": 1.2335530778703252e-06, + "loss": 0.0194, + "step": 93270 + }, + { + "epoch": 6.111511498394811, + "grad_norm": 0.6829878687858582, + "learning_rate": 1.2299056012450182e-06, + "loss": 0.0541, + "step": 93280 + }, + { + "epoch": 6.11216667758632, + "grad_norm": 0.5570991039276123, + "learning_rate": 1.2262634580439026e-06, + "loss": 0.023, + "step": 93290 + }, + { + "epoch": 6.1128218567778285, + "grad_norm": 0.517997145652771, + "learning_rate": 1.2226266486652926e-06, + "loss": 0.0248, + "step": 93300 + }, + { + "epoch": 6.113477035969337, + "grad_norm": 0.642400860786438, + "learning_rate": 1.2189951735068927e-06, + "loss": 0.0492, + "step": 93310 + }, + { + "epoch": 6.114132215160846, + "grad_norm": 1.0812700986862183, + "learning_rate": 1.2153690329658407e-06, + "loss": 0.0347, + "step": 93320 + }, + { + "epoch": 6.114787394352355, + "grad_norm": 0.3879023790359497, + "learning_rate": 1.2117482274386859e-06, + "loss": 0.0532, + "step": 93330 + }, + { + "epoch": 6.115442573543865, + "grad_norm": 0.968626081943512, + "learning_rate": 1.2081327573213953e-06, + "loss": 0.0319, + "step": 93340 + }, + { + "epoch": 6.1160977527353735, + "grad_norm": 0.13411621749401093, + "learning_rate": 1.2045226230093465e-06, + "loss": 0.0272, + "step": 93350 + }, + { + "epoch": 6.116752931926882, + "grad_norm": 0.6621155142784119, + "learning_rate": 1.2009178248973296e-06, + "loss": 0.0366, + "step": 93360 + }, + { + "epoch": 6.117408111118391, + "grad_norm": 0.6917613744735718, + "learning_rate": 1.1973183633795849e-06, + "loss": 0.0229, + "step": 93370 + }, + { + "epoch": 6.1180632903099, + "grad_norm": 1.0242022275924683, + "learning_rate": 1.193724238849725e-06, + "loss": 0.0251, + "step": 93380 + }, + { + "epoch": 6.118718469501409, + "grad_norm": 0.5684864521026611, + "learning_rate": 1.190135451700808e-06, + "loss": 0.0443, + "step": 93390 + }, + { + "epoch": 6.1193736486929176, + "grad_norm": 0.11470380425453186, + "learning_rate": 1.186552002325292e-06, + "loss": 0.0362, + "step": 93400 + }, + { + "epoch": 6.120028827884426, + "grad_norm": 0.22299803793430328, + "learning_rate": 1.1829738911150578e-06, + "loss": 0.0329, + "step": 93410 + }, + { + "epoch": 6.120684007075935, + "grad_norm": 0.3756260275840759, + "learning_rate": 1.1794011184613985e-06, + "loss": 0.0115, + "step": 93420 + }, + { + "epoch": 6.121339186267444, + "grad_norm": 1.4920318126678467, + "learning_rate": 1.17583368475504e-06, + "loss": 0.0326, + "step": 93430 + }, + { + "epoch": 6.121994365458953, + "grad_norm": 0.39889562129974365, + "learning_rate": 1.1722715903860982e-06, + "loss": 0.0243, + "step": 93440 + }, + { + "epoch": 6.122649544650462, + "grad_norm": 0.18460208177566528, + "learning_rate": 1.168714835744128e-06, + "loss": 0.0359, + "step": 93450 + }, + { + "epoch": 6.12330472384197, + "grad_norm": 1.0994300842285156, + "learning_rate": 1.165163421218085e-06, + "loss": 0.0305, + "step": 93460 + }, + { + "epoch": 6.123959903033479, + "grad_norm": 0.060153529047966, + "learning_rate": 1.1616173471963477e-06, + "loss": 0.0351, + "step": 93470 + }, + { + "epoch": 6.124615082224989, + "grad_norm": 0.10895717144012451, + "learning_rate": 1.1580766140667109e-06, + "loss": 0.0373, + "step": 93480 + }, + { + "epoch": 6.125270261416498, + "grad_norm": 0.1380683034658432, + "learning_rate": 1.1545412222163764e-06, + "loss": 0.0247, + "step": 93490 + }, + { + "epoch": 6.125925440608007, + "grad_norm": 0.9394170641899109, + "learning_rate": 1.1510111720319795e-06, + "loss": 0.0468, + "step": 93500 + }, + { + "epoch": 6.126580619799515, + "grad_norm": 1.2983677387237549, + "learning_rate": 1.1474864638995663e-06, + "loss": 0.0564, + "step": 93510 + }, + { + "epoch": 6.127235798991024, + "grad_norm": 0.4773406684398651, + "learning_rate": 1.1439670982045735e-06, + "loss": 0.062, + "step": 93520 + }, + { + "epoch": 6.127890978182533, + "grad_norm": 0.5815320611000061, + "learning_rate": 1.1404530753318877e-06, + "loss": 0.0304, + "step": 93530 + }, + { + "epoch": 6.128546157374042, + "grad_norm": 0.2630961835384369, + "learning_rate": 1.136944395665801e-06, + "loss": 0.0304, + "step": 93540 + }, + { + "epoch": 6.129201336565551, + "grad_norm": 0.650616466999054, + "learning_rate": 1.1334410595900014e-06, + "loss": 0.0498, + "step": 93550 + }, + { + "epoch": 6.1298565157570595, + "grad_norm": 0.518256425857544, + "learning_rate": 1.129943067487621e-06, + "loss": 0.0226, + "step": 93560 + }, + { + "epoch": 6.130511694948568, + "grad_norm": 0.2457280457019806, + "learning_rate": 1.126450419741193e-06, + "loss": 0.0422, + "step": 93570 + }, + { + "epoch": 6.131166874140077, + "grad_norm": 0.07596541941165924, + "learning_rate": 1.1229631167326671e-06, + "loss": 0.0267, + "step": 93580 + }, + { + "epoch": 6.131822053331586, + "grad_norm": 0.11409083753824234, + "learning_rate": 1.1194811588434052e-06, + "loss": 0.0675, + "step": 93590 + }, + { + "epoch": 6.132477232523095, + "grad_norm": 0.13007290661334991, + "learning_rate": 1.1160045464542023e-06, + "loss": 0.0257, + "step": 93600 + }, + { + "epoch": 6.1331324117146035, + "grad_norm": 0.5279057025909424, + "learning_rate": 1.1125332799452325e-06, + "loss": 0.0388, + "step": 93610 + }, + { + "epoch": 6.133787590906113, + "grad_norm": 0.6203641891479492, + "learning_rate": 1.109067359696131e-06, + "loss": 0.0288, + "step": 93620 + }, + { + "epoch": 6.134442770097622, + "grad_norm": 0.15858764946460724, + "learning_rate": 1.1056067860859166e-06, + "loss": 0.0355, + "step": 93630 + }, + { + "epoch": 6.135097949289131, + "grad_norm": 0.4525991976261139, + "learning_rate": 1.102151559493031e-06, + "loss": 0.0321, + "step": 93640 + }, + { + "epoch": 6.13575312848064, + "grad_norm": 0.07001493871212006, + "learning_rate": 1.0987016802953332e-06, + "loss": 0.0215, + "step": 93650 + }, + { + "epoch": 6.1364083076721485, + "grad_norm": 1.2847312688827515, + "learning_rate": 1.0952571488701048e-06, + "loss": 0.0315, + "step": 93660 + }, + { + "epoch": 6.137063486863657, + "grad_norm": 0.7294890880584717, + "learning_rate": 1.0918179655940163e-06, + "loss": 0.0297, + "step": 93670 + }, + { + "epoch": 6.137718666055166, + "grad_norm": 0.1774882674217224, + "learning_rate": 1.088384130843184e-06, + "loss": 0.0163, + "step": 93680 + }, + { + "epoch": 6.138373845246675, + "grad_norm": 0.7167577743530273, + "learning_rate": 1.0849556449931296e-06, + "loss": 0.0326, + "step": 93690 + }, + { + "epoch": 6.139029024438184, + "grad_norm": 0.3902640640735626, + "learning_rate": 1.0815325084187866e-06, + "loss": 0.0282, + "step": 93700 + }, + { + "epoch": 6.139684203629693, + "grad_norm": 0.22460588812828064, + "learning_rate": 1.0781147214944943e-06, + "loss": 0.0172, + "step": 93710 + }, + { + "epoch": 6.140339382821201, + "grad_norm": 0.6353907585144043, + "learning_rate": 1.0747022845940313e-06, + "loss": 0.041, + "step": 93720 + }, + { + "epoch": 6.14099456201271, + "grad_norm": 0.19703409075737, + "learning_rate": 1.0712951980905605e-06, + "loss": 0.0155, + "step": 93730 + }, + { + "epoch": 6.141649741204219, + "grad_norm": 1.0307292938232422, + "learning_rate": 1.0678934623566839e-06, + "loss": 0.0368, + "step": 93740 + }, + { + "epoch": 6.142304920395728, + "grad_norm": 0.5690317749977112, + "learning_rate": 1.0644970777644091e-06, + "loss": 0.0244, + "step": 93750 + }, + { + "epoch": 6.142960099587237, + "grad_norm": 0.8406808376312256, + "learning_rate": 1.0611060446851673e-06, + "loss": 0.0375, + "step": 93760 + }, + { + "epoch": 6.143615278778746, + "grad_norm": 0.40178585052490234, + "learning_rate": 1.0577203634897836e-06, + "loss": 0.0291, + "step": 93770 + }, + { + "epoch": 6.144270457970255, + "grad_norm": 0.22732573747634888, + "learning_rate": 1.0543400345485178e-06, + "loss": 0.018, + "step": 93780 + }, + { + "epoch": 6.144925637161764, + "grad_norm": 1.1922719478607178, + "learning_rate": 1.0509650582310404e-06, + "loss": 0.0225, + "step": 93790 + }, + { + "epoch": 6.145580816353273, + "grad_norm": 0.6041253209114075, + "learning_rate": 1.0475954349064287e-06, + "loss": 0.046, + "step": 93800 + }, + { + "epoch": 6.146235995544782, + "grad_norm": 0.2321402132511139, + "learning_rate": 1.0442311649431825e-06, + "loss": 0.0205, + "step": 93810 + }, + { + "epoch": 6.14689117473629, + "grad_norm": 0.37433937191963196, + "learning_rate": 1.0408722487092071e-06, + "loss": 0.0302, + "step": 93820 + }, + { + "epoch": 6.147546353927799, + "grad_norm": 0.77745521068573, + "learning_rate": 1.0375186865718424e-06, + "loss": 0.0258, + "step": 93830 + }, + { + "epoch": 6.148201533119308, + "grad_norm": 0.33701521158218384, + "learning_rate": 1.0341704788978113e-06, + "loss": 0.0172, + "step": 93840 + }, + { + "epoch": 6.148856712310817, + "grad_norm": 0.2897767126560211, + "learning_rate": 1.0308276260532879e-06, + "loss": 0.0265, + "step": 93850 + }, + { + "epoch": 6.149511891502326, + "grad_norm": 0.46896639466285706, + "learning_rate": 1.0274901284038241e-06, + "loss": 0.027, + "step": 93860 + }, + { + "epoch": 6.1501670706938345, + "grad_norm": 0.334940642118454, + "learning_rate": 1.0241579863144114e-06, + "loss": 0.0382, + "step": 93870 + }, + { + "epoch": 6.150822249885343, + "grad_norm": 0.5180766582489014, + "learning_rate": 1.0208312001494413e-06, + "loss": 0.0237, + "step": 93880 + }, + { + "epoch": 6.151477429076852, + "grad_norm": 1.315159559249878, + "learning_rate": 1.0175097702727453e-06, + "loss": 0.0421, + "step": 93890 + }, + { + "epoch": 6.152132608268362, + "grad_norm": 0.21428722143173218, + "learning_rate": 1.0141936970475275e-06, + "loss": 0.0319, + "step": 93900 + }, + { + "epoch": 6.152787787459871, + "grad_norm": 0.26352113485336304, + "learning_rate": 1.0108829808364417e-06, + "loss": 0.0165, + "step": 93910 + }, + { + "epoch": 6.1534429666513795, + "grad_norm": 0.6964882016181946, + "learning_rate": 1.0075776220015376e-06, + "loss": 0.0242, + "step": 93920 + }, + { + "epoch": 6.154098145842888, + "grad_norm": 1.0969775915145874, + "learning_rate": 1.0042776209042816e-06, + "loss": 0.0285, + "step": 93930 + }, + { + "epoch": 6.154753325034397, + "grad_norm": 0.1744416356086731, + "learning_rate": 1.000982977905568e-06, + "loss": 0.0258, + "step": 93940 + }, + { + "epoch": 6.155408504225906, + "grad_norm": 0.4095724821090698, + "learning_rate": 9.976936933656867e-07, + "loss": 0.046, + "step": 93950 + }, + { + "epoch": 6.156063683417415, + "grad_norm": 0.19301621615886688, + "learning_rate": 9.944097676443442e-07, + "loss": 0.035, + "step": 93960 + }, + { + "epoch": 6.1567188626089235, + "grad_norm": 0.38317301869392395, + "learning_rate": 9.911312011006757e-07, + "loss": 0.0158, + "step": 93970 + }, + { + "epoch": 6.157374041800432, + "grad_norm": 0.6968376636505127, + "learning_rate": 9.878579940932109e-07, + "loss": 0.0418, + "step": 93980 + }, + { + "epoch": 6.158029220991941, + "grad_norm": 0.6327629685401917, + "learning_rate": 9.84590146979908e-07, + "loss": 0.0857, + "step": 93990 + }, + { + "epoch": 6.15868440018345, + "grad_norm": 0.9004532694816589, + "learning_rate": 9.813276601181365e-07, + "loss": 0.0392, + "step": 94000 + }, + { + "epoch": 6.159339579374959, + "grad_norm": 0.12939296662807465, + "learning_rate": 9.780705338646724e-07, + "loss": 0.0419, + "step": 94010 + }, + { + "epoch": 6.159994758566468, + "grad_norm": 0.17851139605045319, + "learning_rate": 9.748187685757083e-07, + "loss": 0.031, + "step": 94020 + }, + { + "epoch": 6.160649937757976, + "grad_norm": 0.2850800156593323, + "learning_rate": 9.715723646068543e-07, + "loss": 0.0315, + "step": 94030 + }, + { + "epoch": 6.161305116949485, + "grad_norm": 0.14638769626617432, + "learning_rate": 9.68331322313143e-07, + "loss": 0.0423, + "step": 94040 + }, + { + "epoch": 6.161960296140995, + "grad_norm": 0.3958643078804016, + "learning_rate": 9.650956420489909e-07, + "loss": 0.0302, + "step": 94050 + }, + { + "epoch": 6.162615475332504, + "grad_norm": 1.2824077606201172, + "learning_rate": 9.618653241682595e-07, + "loss": 0.0839, + "step": 94060 + }, + { + "epoch": 6.163270654524013, + "grad_norm": 0.12417373806238174, + "learning_rate": 9.586403690242053e-07, + "loss": 0.0625, + "step": 94070 + }, + { + "epoch": 6.163925833715521, + "grad_norm": 1.2650865316390991, + "learning_rate": 9.554207769695122e-07, + "loss": 0.0311, + "step": 94080 + }, + { + "epoch": 6.16458101290703, + "grad_norm": 0.14849936962127686, + "learning_rate": 9.522065483562659e-07, + "loss": 0.0374, + "step": 94090 + }, + { + "epoch": 6.165236192098539, + "grad_norm": 1.3308302164077759, + "learning_rate": 9.489976835359681e-07, + "loss": 0.0246, + "step": 94100 + }, + { + "epoch": 6.165891371290048, + "grad_norm": 0.566381573677063, + "learning_rate": 9.457941828595329e-07, + "loss": 0.0374, + "step": 94110 + }, + { + "epoch": 6.166546550481557, + "grad_norm": 0.2965049147605896, + "learning_rate": 9.425960466772965e-07, + "loss": 0.0266, + "step": 94120 + }, + { + "epoch": 6.1672017296730655, + "grad_norm": 0.16186627745628357, + "learning_rate": 9.394032753390014e-07, + "loss": 0.0248, + "step": 94130 + }, + { + "epoch": 6.167856908864574, + "grad_norm": 0.480387806892395, + "learning_rate": 9.362158691938017e-07, + "loss": 0.0355, + "step": 94140 + }, + { + "epoch": 6.168512088056083, + "grad_norm": 0.2587251663208008, + "learning_rate": 9.330338285902685e-07, + "loss": 0.0213, + "step": 94150 + }, + { + "epoch": 6.169167267247592, + "grad_norm": 0.21440178155899048, + "learning_rate": 9.298571538763845e-07, + "loss": 0.0409, + "step": 94160 + }, + { + "epoch": 6.169822446439101, + "grad_norm": 0.6914598345756531, + "learning_rate": 9.266858453995441e-07, + "loss": 0.0258, + "step": 94170 + }, + { + "epoch": 6.17047762563061, + "grad_norm": 0.4625341296195984, + "learning_rate": 9.235199035065645e-07, + "loss": 0.0446, + "step": 94180 + }, + { + "epoch": 6.171132804822119, + "grad_norm": 0.15643338859081268, + "learning_rate": 9.203593285436574e-07, + "loss": 0.0136, + "step": 94190 + }, + { + "epoch": 6.171787984013628, + "grad_norm": 0.7852720618247986, + "learning_rate": 9.172041208564686e-07, + "loss": 0.0162, + "step": 94200 + }, + { + "epoch": 6.172443163205137, + "grad_norm": 0.32178324460983276, + "learning_rate": 9.140542807900332e-07, + "loss": 0.0332, + "step": 94210 + }, + { + "epoch": 6.173098342396646, + "grad_norm": 0.6415852308273315, + "learning_rate": 9.109098086888313e-07, + "loss": 0.0198, + "step": 94220 + }, + { + "epoch": 6.1737535215881545, + "grad_norm": 0.9994800686836243, + "learning_rate": 9.077707048967265e-07, + "loss": 0.0316, + "step": 94230 + }, + { + "epoch": 6.174408700779663, + "grad_norm": 0.48356059193611145, + "learning_rate": 9.046369697570057e-07, + "loss": 0.0277, + "step": 94240 + }, + { + "epoch": 6.175063879971172, + "grad_norm": 0.2314886450767517, + "learning_rate": 9.015086036123721e-07, + "loss": 0.0185, + "step": 94250 + }, + { + "epoch": 6.175719059162681, + "grad_norm": 0.3914130926132202, + "learning_rate": 8.983856068049413e-07, + "loss": 0.0306, + "step": 94260 + }, + { + "epoch": 6.17637423835419, + "grad_norm": 0.8585995435714722, + "learning_rate": 8.952679796762398e-07, + "loss": 0.024, + "step": 94270 + }, + { + "epoch": 6.177029417545699, + "grad_norm": 1.146770715713501, + "learning_rate": 8.921557225671951e-07, + "loss": 0.0337, + "step": 94280 + }, + { + "epoch": 6.177684596737207, + "grad_norm": 1.0364633798599243, + "learning_rate": 8.890488358181792e-07, + "loss": 0.0232, + "step": 94290 + }, + { + "epoch": 6.178339775928716, + "grad_norm": 0.3242431581020355, + "learning_rate": 8.859473197689372e-07, + "loss": 0.0221, + "step": 94300 + }, + { + "epoch": 6.178994955120225, + "grad_norm": 0.13159453868865967, + "learning_rate": 8.828511747586588e-07, + "loss": 0.0272, + "step": 94310 + }, + { + "epoch": 6.179650134311734, + "grad_norm": 0.2517247200012207, + "learning_rate": 8.797604011259287e-07, + "loss": 0.034, + "step": 94320 + }, + { + "epoch": 6.1803053135032435, + "grad_norm": 0.19577106833457947, + "learning_rate": 8.766749992087487e-07, + "loss": 0.0293, + "step": 94330 + }, + { + "epoch": 6.180960492694752, + "grad_norm": 0.8213270306587219, + "learning_rate": 8.735949693445378e-07, + "loss": 0.0375, + "step": 94340 + }, + { + "epoch": 6.181615671886261, + "grad_norm": 0.3931225538253784, + "learning_rate": 8.705203118701211e-07, + "loss": 0.0558, + "step": 94350 + }, + { + "epoch": 6.18227085107777, + "grad_norm": 0.14032931625843048, + "learning_rate": 8.674510271217295e-07, + "loss": 0.0195, + "step": 94360 + }, + { + "epoch": 6.182926030269279, + "grad_norm": 0.23759952187538147, + "learning_rate": 8.643871154350336e-07, + "loss": 0.0215, + "step": 94370 + }, + { + "epoch": 6.183581209460788, + "grad_norm": 0.6650974154472351, + "learning_rate": 8.613285771450818e-07, + "loss": 0.0645, + "step": 94380 + }, + { + "epoch": 6.184236388652296, + "grad_norm": 0.6501094102859497, + "learning_rate": 8.582754125863623e-07, + "loss": 0.0356, + "step": 94390 + }, + { + "epoch": 6.184891567843805, + "grad_norm": 1.124473214149475, + "learning_rate": 8.55227622092758e-07, + "loss": 0.0221, + "step": 94400 + }, + { + "epoch": 6.185546747035314, + "grad_norm": 1.519795298576355, + "learning_rate": 8.521852059975743e-07, + "loss": 0.0479, + "step": 94410 + }, + { + "epoch": 6.186201926226823, + "grad_norm": 0.8019671440124512, + "learning_rate": 8.491481646335231e-07, + "loss": 0.037, + "step": 94420 + }, + { + "epoch": 6.186857105418332, + "grad_norm": 0.9876717925071716, + "learning_rate": 8.46116498332733e-07, + "loss": 0.0202, + "step": 94430 + }, + { + "epoch": 6.1875122846098405, + "grad_norm": 0.08269915729761124, + "learning_rate": 8.430902074267333e-07, + "loss": 0.0128, + "step": 94440 + }, + { + "epoch": 6.188167463801349, + "grad_norm": 0.1356395184993744, + "learning_rate": 8.400692922464925e-07, + "loss": 0.0194, + "step": 94450 + }, + { + "epoch": 6.188822642992859, + "grad_norm": 1.0023226737976074, + "learning_rate": 8.370537531223577e-07, + "loss": 0.0404, + "step": 94460 + }, + { + "epoch": 6.189477822184368, + "grad_norm": 0.34760648012161255, + "learning_rate": 8.340435903841092e-07, + "loss": 0.0176, + "step": 94470 + }, + { + "epoch": 6.190133001375877, + "grad_norm": 0.6710610389709473, + "learning_rate": 8.310388043609341e-07, + "loss": 0.0232, + "step": 94480 + }, + { + "epoch": 6.1907881805673854, + "grad_norm": 2.0809059143066406, + "learning_rate": 8.280393953814303e-07, + "loss": 0.0264, + "step": 94490 + }, + { + "epoch": 6.191443359758894, + "grad_norm": 0.24059654772281647, + "learning_rate": 8.250453637736077e-07, + "loss": 0.0217, + "step": 94500 + }, + { + "epoch": 6.192098538950403, + "grad_norm": 0.7028981447219849, + "learning_rate": 8.22056709864899e-07, + "loss": 0.0409, + "step": 94510 + }, + { + "epoch": 6.192753718141912, + "grad_norm": 0.7850958704948425, + "learning_rate": 8.190734339821204e-07, + "loss": 0.014, + "step": 94520 + }, + { + "epoch": 6.193408897333421, + "grad_norm": 0.398425817489624, + "learning_rate": 8.160955364515333e-07, + "loss": 0.0185, + "step": 94530 + }, + { + "epoch": 6.1940640765249295, + "grad_norm": 0.18115265667438507, + "learning_rate": 8.131230175987936e-07, + "loss": 0.0287, + "step": 94540 + }, + { + "epoch": 6.194719255716438, + "grad_norm": 0.8679574728012085, + "learning_rate": 8.101558777489693e-07, + "loss": 0.0408, + "step": 94550 + }, + { + "epoch": 6.195374434907947, + "grad_norm": 0.2937382161617279, + "learning_rate": 8.071941172265396e-07, + "loss": 0.0219, + "step": 94560 + }, + { + "epoch": 6.196029614099456, + "grad_norm": 0.1935468167066574, + "learning_rate": 8.042377363553955e-07, + "loss": 0.0254, + "step": 94570 + }, + { + "epoch": 6.196684793290965, + "grad_norm": 0.561191976070404, + "learning_rate": 8.01286735458856e-07, + "loss": 0.0308, + "step": 94580 + }, + { + "epoch": 6.197339972482474, + "grad_norm": 0.4415982961654663, + "learning_rate": 7.983411148596298e-07, + "loss": 0.0369, + "step": 94590 + }, + { + "epoch": 6.197995151673982, + "grad_norm": 0.6662211418151855, + "learning_rate": 7.954008748798425e-07, + "loss": 0.0243, + "step": 94600 + }, + { + "epoch": 6.198650330865492, + "grad_norm": 0.19315217435359955, + "learning_rate": 7.924660158410369e-07, + "loss": 0.0266, + "step": 94610 + }, + { + "epoch": 6.199305510057001, + "grad_norm": 0.3940088152885437, + "learning_rate": 7.895365380641728e-07, + "loss": 0.0314, + "step": 94620 + }, + { + "epoch": 6.19996068924851, + "grad_norm": 0.2699145972728729, + "learning_rate": 7.866124418695941e-07, + "loss": 0.0345, + "step": 94630 + }, + { + "epoch": 6.2006158684400186, + "grad_norm": 0.11872562766075134, + "learning_rate": 7.83693727577095e-07, + "loss": 0.0118, + "step": 94640 + }, + { + "epoch": 6.201271047631527, + "grad_norm": 0.13122890889644623, + "learning_rate": 7.807803955058535e-07, + "loss": 0.0215, + "step": 94650 + }, + { + "epoch": 6.201926226823036, + "grad_norm": 1.0789984464645386, + "learning_rate": 7.778724459744702e-07, + "loss": 0.0307, + "step": 94660 + }, + { + "epoch": 6.202581406014545, + "grad_norm": 0.5960435271263123, + "learning_rate": 7.749698793009464e-07, + "loss": 0.0194, + "step": 94670 + }, + { + "epoch": 6.203236585206054, + "grad_norm": 0.1552848219871521, + "learning_rate": 7.720726958027169e-07, + "loss": 0.0225, + "step": 94680 + }, + { + "epoch": 6.203891764397563, + "grad_norm": 0.16611476242542267, + "learning_rate": 7.691808957965951e-07, + "loss": 0.0341, + "step": 94690 + }, + { + "epoch": 6.204546943589071, + "grad_norm": 0.2123579978942871, + "learning_rate": 7.662944795988336e-07, + "loss": 0.0234, + "step": 94700 + }, + { + "epoch": 6.20520212278058, + "grad_norm": 0.5086084008216858, + "learning_rate": 7.634134475250854e-07, + "loss": 0.0401, + "step": 94710 + }, + { + "epoch": 6.205857301972089, + "grad_norm": 0.17993111908435822, + "learning_rate": 7.605377998904262e-07, + "loss": 0.017, + "step": 94720 + }, + { + "epoch": 6.206512481163598, + "grad_norm": 0.4279100000858307, + "learning_rate": 7.5766753700931e-07, + "loss": 0.0135, + "step": 94730 + }, + { + "epoch": 6.207167660355107, + "grad_norm": 1.1803762912750244, + "learning_rate": 7.54802659195647e-07, + "loss": 0.0603, + "step": 94740 + }, + { + "epoch": 6.207822839546616, + "grad_norm": 0.3167823255062103, + "learning_rate": 7.519431667627253e-07, + "loss": 0.025, + "step": 94750 + }, + { + "epoch": 6.208478018738125, + "grad_norm": 0.19501715898513794, + "learning_rate": 7.490890600232503e-07, + "loss": 0.0227, + "step": 94760 + }, + { + "epoch": 6.209133197929634, + "grad_norm": 0.5848727822303772, + "learning_rate": 7.462403392893502e-07, + "loss": 0.0374, + "step": 94770 + }, + { + "epoch": 6.209788377121143, + "grad_norm": 0.43558749556541443, + "learning_rate": 7.433970048725536e-07, + "loss": 0.0301, + "step": 94780 + }, + { + "epoch": 6.210443556312652, + "grad_norm": 0.10358642786741257, + "learning_rate": 7.405590570838062e-07, + "loss": 0.0216, + "step": 94790 + }, + { + "epoch": 6.2110987355041605, + "grad_norm": 0.5238473415374756, + "learning_rate": 7.377264962334596e-07, + "loss": 0.0232, + "step": 94800 + }, + { + "epoch": 6.211753914695669, + "grad_norm": 0.118538998067379, + "learning_rate": 7.348993226312773e-07, + "loss": 0.0274, + "step": 94810 + }, + { + "epoch": 6.212409093887178, + "grad_norm": 1.5156933069229126, + "learning_rate": 7.32077536586434e-07, + "loss": 0.0268, + "step": 94820 + }, + { + "epoch": 6.213064273078687, + "grad_norm": 0.802013635635376, + "learning_rate": 7.292611384075221e-07, + "loss": 0.0363, + "step": 94830 + }, + { + "epoch": 6.213719452270196, + "grad_norm": 0.8468459248542786, + "learning_rate": 7.264501284025338e-07, + "loss": 0.0348, + "step": 94840 + }, + { + "epoch": 6.2143746314617045, + "grad_norm": 0.17279554903507233, + "learning_rate": 7.236445068788789e-07, + "loss": 0.0249, + "step": 94850 + }, + { + "epoch": 6.215029810653213, + "grad_norm": 0.2158946543931961, + "learning_rate": 7.20844274143373e-07, + "loss": 0.0238, + "step": 94860 + }, + { + "epoch": 6.215684989844722, + "grad_norm": 1.4322460889816284, + "learning_rate": 7.180494305022489e-07, + "loss": 0.0536, + "step": 94870 + }, + { + "epoch": 6.216340169036231, + "grad_norm": 0.3512232303619385, + "learning_rate": 7.152599762611456e-07, + "loss": 0.0259, + "step": 94880 + }, + { + "epoch": 6.216995348227741, + "grad_norm": 0.47330018877983093, + "learning_rate": 7.124759117251079e-07, + "loss": 0.0363, + "step": 94890 + }, + { + "epoch": 6.2176505274192495, + "grad_norm": 0.3770882189273834, + "learning_rate": 7.096972371986033e-07, + "loss": 0.0145, + "step": 94900 + }, + { + "epoch": 6.218305706610758, + "grad_norm": 0.7752100825309753, + "learning_rate": 7.069239529855054e-07, + "loss": 0.0324, + "step": 94910 + }, + { + "epoch": 6.218960885802267, + "grad_norm": 0.9694914221763611, + "learning_rate": 7.041560593890939e-07, + "loss": 0.0284, + "step": 94920 + }, + { + "epoch": 6.219616064993776, + "grad_norm": 0.2231239527463913, + "learning_rate": 7.013935567120599e-07, + "loss": 0.024, + "step": 94930 + }, + { + "epoch": 6.220271244185285, + "grad_norm": 0.13762766122817993, + "learning_rate": 6.986364452565009e-07, + "loss": 0.0284, + "step": 94940 + }, + { + "epoch": 6.220926423376794, + "grad_norm": 0.12706837058067322, + "learning_rate": 6.95884725323942e-07, + "loss": 0.0197, + "step": 94950 + }, + { + "epoch": 6.221581602568302, + "grad_norm": 0.29093676805496216, + "learning_rate": 6.931383972152982e-07, + "loss": 0.0158, + "step": 94960 + }, + { + "epoch": 6.222236781759811, + "grad_norm": 0.9319371581077576, + "learning_rate": 6.903974612309127e-07, + "loss": 0.0376, + "step": 94970 + }, + { + "epoch": 6.22289196095132, + "grad_norm": 0.2647685110569, + "learning_rate": 6.876619176705179e-07, + "loss": 0.0228, + "step": 94980 + }, + { + "epoch": 6.223547140142829, + "grad_norm": 0.7026460766792297, + "learning_rate": 6.8493176683328e-07, + "loss": 0.0211, + "step": 94990 + }, + { + "epoch": 6.224202319334338, + "grad_norm": 0.0970192700624466, + "learning_rate": 6.822070090177657e-07, + "loss": 0.0075, + "step": 95000 + }, + { + "epoch": 6.2248574985258465, + "grad_norm": 0.7688113451004028, + "learning_rate": 6.794876445219367e-07, + "loss": 0.0323, + "step": 95010 + }, + { + "epoch": 6.225512677717355, + "grad_norm": 0.13697555661201477, + "learning_rate": 6.767736736431885e-07, + "loss": 0.0291, + "step": 95020 + }, + { + "epoch": 6.226167856908865, + "grad_norm": 1.3178611993789673, + "learning_rate": 6.740650966783113e-07, + "loss": 0.0256, + "step": 95030 + }, + { + "epoch": 6.226823036100374, + "grad_norm": 0.6875267624855042, + "learning_rate": 6.713619139235239e-07, + "loss": 0.0213, + "step": 95040 + }, + { + "epoch": 6.227478215291883, + "grad_norm": 0.3172498345375061, + "learning_rate": 6.686641256744286e-07, + "loss": 0.0244, + "step": 95050 + }, + { + "epoch": 6.228133394483391, + "grad_norm": 0.3452130854129791, + "learning_rate": 6.659717322260562e-07, + "loss": 0.0237, + "step": 95060 + }, + { + "epoch": 6.2287885736749, + "grad_norm": 0.7000442147254944, + "learning_rate": 6.632847338728377e-07, + "loss": 0.0412, + "step": 95070 + }, + { + "epoch": 6.229443752866409, + "grad_norm": 0.15369252860546112, + "learning_rate": 6.60603130908627e-07, + "loss": 0.0268, + "step": 95080 + }, + { + "epoch": 6.230098932057918, + "grad_norm": 0.6334192156791687, + "learning_rate": 6.579269236266783e-07, + "loss": 0.0165, + "step": 95090 + }, + { + "epoch": 6.230754111249427, + "grad_norm": 0.5574760437011719, + "learning_rate": 6.552561123196632e-07, + "loss": 0.0229, + "step": 95100 + }, + { + "epoch": 6.2314092904409355, + "grad_norm": 0.6278159022331238, + "learning_rate": 6.525906972796425e-07, + "loss": 0.0271, + "step": 95110 + }, + { + "epoch": 6.232064469632444, + "grad_norm": 0.5958238840103149, + "learning_rate": 6.499306787981108e-07, + "loss": 0.0329, + "step": 95120 + }, + { + "epoch": 6.232719648823953, + "grad_norm": 0.8363333344459534, + "learning_rate": 6.472760571659631e-07, + "loss": 0.0139, + "step": 95130 + }, + { + "epoch": 6.233374828015462, + "grad_norm": 0.3270891606807709, + "learning_rate": 6.446268326735061e-07, + "loss": 0.0425, + "step": 95140 + }, + { + "epoch": 6.234030007206971, + "grad_norm": 0.9805633425712585, + "learning_rate": 6.41983005610458e-07, + "loss": 0.031, + "step": 95150 + }, + { + "epoch": 6.23468518639848, + "grad_norm": 0.2488611340522766, + "learning_rate": 6.393445762659378e-07, + "loss": 0.0278, + "step": 95160 + }, + { + "epoch": 6.235340365589989, + "grad_norm": 0.2286245971918106, + "learning_rate": 6.36711544928481e-07, + "loss": 0.0377, + "step": 95170 + }, + { + "epoch": 6.235995544781498, + "grad_norm": 0.7172495722770691, + "learning_rate": 6.340839118860298e-07, + "loss": 0.0258, + "step": 95180 + }, + { + "epoch": 6.236650723973007, + "grad_norm": 0.47542262077331543, + "learning_rate": 6.314616774259485e-07, + "loss": 0.0107, + "step": 95190 + }, + { + "epoch": 6.237305903164516, + "grad_norm": 0.5875977277755737, + "learning_rate": 6.288448418349858e-07, + "loss": 0.0286, + "step": 95200 + }, + { + "epoch": 6.2379610823560245, + "grad_norm": 0.4322850704193115, + "learning_rate": 6.262334053993235e-07, + "loss": 0.0343, + "step": 95210 + }, + { + "epoch": 6.238616261547533, + "grad_norm": 0.796550989151001, + "learning_rate": 6.236273684045501e-07, + "loss": 0.0144, + "step": 95220 + }, + { + "epoch": 6.239271440739042, + "grad_norm": 0.46793773770332336, + "learning_rate": 6.210267311356488e-07, + "loss": 0.0262, + "step": 95230 + }, + { + "epoch": 6.239926619930551, + "grad_norm": 0.7673544883728027, + "learning_rate": 6.184314938770197e-07, + "loss": 0.0443, + "step": 95240 + }, + { + "epoch": 6.24058179912206, + "grad_norm": 0.10476833581924438, + "learning_rate": 6.158416569124858e-07, + "loss": 0.0298, + "step": 95250 + }, + { + "epoch": 6.241236978313569, + "grad_norm": 0.0966269001364708, + "learning_rate": 6.132572205252541e-07, + "loss": 0.037, + "step": 95260 + }, + { + "epoch": 6.241892157505077, + "grad_norm": 0.24471215903759003, + "learning_rate": 6.106781849979648e-07, + "loss": 0.0149, + "step": 95270 + }, + { + "epoch": 6.242547336696586, + "grad_norm": 1.5312788486480713, + "learning_rate": 6.081045506126482e-07, + "loss": 0.0408, + "step": 95280 + }, + { + "epoch": 6.243202515888095, + "grad_norm": 0.7791761159896851, + "learning_rate": 6.055363176507678e-07, + "loss": 0.0185, + "step": 95290 + }, + { + "epoch": 6.243857695079604, + "grad_norm": 0.09497278928756714, + "learning_rate": 6.029734863931602e-07, + "loss": 0.0269, + "step": 95300 + }, + { + "epoch": 6.244512874271114, + "grad_norm": 0.8491337895393372, + "learning_rate": 6.004160571201179e-07, + "loss": 0.0233, + "step": 95310 + }, + { + "epoch": 6.245168053462622, + "grad_norm": 0.2260824739933014, + "learning_rate": 5.978640301112947e-07, + "loss": 0.02, + "step": 95320 + }, + { + "epoch": 6.245823232654131, + "grad_norm": 1.3905367851257324, + "learning_rate": 5.9531740564579e-07, + "loss": 0.0238, + "step": 95330 + }, + { + "epoch": 6.24647841184564, + "grad_norm": 0.7936580181121826, + "learning_rate": 5.927761840020918e-07, + "loss": 0.0331, + "step": 95340 + }, + { + "epoch": 6.247133591037149, + "grad_norm": 0.7308593988418579, + "learning_rate": 5.902403654581113e-07, + "loss": 0.0273, + "step": 95350 + }, + { + "epoch": 6.247788770228658, + "grad_norm": 0.7838078737258911, + "learning_rate": 5.877099502911542e-07, + "loss": 0.0411, + "step": 95360 + }, + { + "epoch": 6.2484439494201665, + "grad_norm": 0.4651951789855957, + "learning_rate": 5.851849387779496e-07, + "loss": 0.024, + "step": 95370 + }, + { + "epoch": 6.249099128611675, + "grad_norm": 0.4124661982059479, + "learning_rate": 5.826653311946262e-07, + "loss": 0.04, + "step": 95380 + }, + { + "epoch": 6.249754307803184, + "grad_norm": 0.11973555386066437, + "learning_rate": 5.801511278167249e-07, + "loss": 0.0223, + "step": 95390 + }, + { + "epoch": 6.250409486994693, + "grad_norm": 0.29500433802604675, + "learning_rate": 5.776423289191868e-07, + "loss": 0.0179, + "step": 95400 + }, + { + "epoch": 6.251064666186202, + "grad_norm": 1.042051911354065, + "learning_rate": 5.751389347763869e-07, + "loss": 0.0312, + "step": 95410 + }, + { + "epoch": 6.2517198453777105, + "grad_norm": 0.537077784538269, + "learning_rate": 5.726409456620841e-07, + "loss": 0.0424, + "step": 95420 + }, + { + "epoch": 6.252375024569219, + "grad_norm": 0.46694162487983704, + "learning_rate": 5.701483618494541e-07, + "loss": 0.0253, + "step": 95430 + }, + { + "epoch": 6.253030203760728, + "grad_norm": 0.21287338435649872, + "learning_rate": 5.67661183611079e-07, + "loss": 0.0211, + "step": 95440 + }, + { + "epoch": 6.253685382952238, + "grad_norm": 0.08022672683000565, + "learning_rate": 5.651794112189579e-07, + "loss": 0.029, + "step": 95450 + }, + { + "epoch": 6.254340562143747, + "grad_norm": 0.4592949151992798, + "learning_rate": 5.627030449444903e-07, + "loss": 0.0434, + "step": 95460 + }, + { + "epoch": 6.2549957413352555, + "grad_norm": 0.23659709095954895, + "learning_rate": 5.60232085058493e-07, + "loss": 0.0194, + "step": 95470 + }, + { + "epoch": 6.255650920526764, + "grad_norm": 0.20322181284427643, + "learning_rate": 5.577665318311776e-07, + "loss": 0.0316, + "step": 95480 + }, + { + "epoch": 6.256306099718273, + "grad_norm": 0.13989003002643585, + "learning_rate": 5.553063855321838e-07, + "loss": 0.0272, + "step": 95490 + }, + { + "epoch": 6.256961278909782, + "grad_norm": 0.6894847750663757, + "learning_rate": 5.528516464305467e-07, + "loss": 0.0205, + "step": 95500 + }, + { + "epoch": 6.257616458101291, + "grad_norm": 0.10993259400129318, + "learning_rate": 5.504023147947068e-07, + "loss": 0.0256, + "step": 95510 + }, + { + "epoch": 6.2582716372928, + "grad_norm": 0.48991677165031433, + "learning_rate": 5.479583908925279e-07, + "loss": 0.029, + "step": 95520 + }, + { + "epoch": 6.258926816484308, + "grad_norm": 0.4842332899570465, + "learning_rate": 5.455198749912626e-07, + "loss": 0.0176, + "step": 95530 + }, + { + "epoch": 6.259581995675817, + "grad_norm": 0.16015799343585968, + "learning_rate": 5.430867673575979e-07, + "loss": 0.0158, + "step": 95540 + }, + { + "epoch": 6.260237174867326, + "grad_norm": 0.45715630054473877, + "learning_rate": 5.406590682575985e-07, + "loss": 0.0218, + "step": 95550 + }, + { + "epoch": 6.260892354058835, + "grad_norm": 0.3433401882648468, + "learning_rate": 5.382367779567687e-07, + "loss": 0.0501, + "step": 95560 + }, + { + "epoch": 6.261547533250344, + "grad_norm": 0.2870153784751892, + "learning_rate": 5.358198967200023e-07, + "loss": 0.0121, + "step": 95570 + }, + { + "epoch": 6.2622027124418524, + "grad_norm": 0.08653496205806732, + "learning_rate": 5.334084248115989e-07, + "loss": 0.0266, + "step": 95580 + }, + { + "epoch": 6.262857891633361, + "grad_norm": 0.11378616094589233, + "learning_rate": 5.310023624952754e-07, + "loss": 0.0135, + "step": 95590 + }, + { + "epoch": 6.263513070824871, + "grad_norm": 0.06964272260665894, + "learning_rate": 5.286017100341656e-07, + "loss": 0.0313, + "step": 95600 + }, + { + "epoch": 6.26416825001638, + "grad_norm": 0.19702522456645966, + "learning_rate": 5.262064676907874e-07, + "loss": 0.033, + "step": 95610 + }, + { + "epoch": 6.264823429207889, + "grad_norm": 0.8684076070785522, + "learning_rate": 5.238166357270924e-07, + "loss": 0.0185, + "step": 95620 + }, + { + "epoch": 6.265478608399397, + "grad_norm": 0.34219637513160706, + "learning_rate": 5.214322144044215e-07, + "loss": 0.0152, + "step": 95630 + }, + { + "epoch": 6.266133787590906, + "grad_norm": 0.2871100604534149, + "learning_rate": 5.190532039835327e-07, + "loss": 0.0271, + "step": 95640 + }, + { + "epoch": 6.266788966782415, + "grad_norm": 0.6489977836608887, + "learning_rate": 5.166796047245903e-07, + "loss": 0.0176, + "step": 95650 + }, + { + "epoch": 6.267444145973924, + "grad_norm": 0.6139306426048279, + "learning_rate": 5.143114168871699e-07, + "loss": 0.0344, + "step": 95660 + }, + { + "epoch": 6.268099325165433, + "grad_norm": 0.5335203409194946, + "learning_rate": 5.119486407302532e-07, + "loss": 0.043, + "step": 95670 + }, + { + "epoch": 6.2687545043569415, + "grad_norm": 0.27327656745910645, + "learning_rate": 5.095912765122335e-07, + "loss": 0.0494, + "step": 95680 + }, + { + "epoch": 6.26940968354845, + "grad_norm": 0.21140645444393158, + "learning_rate": 5.07239324490899e-07, + "loss": 0.026, + "step": 95690 + }, + { + "epoch": 6.270064862739959, + "grad_norm": 1.7631337642669678, + "learning_rate": 5.048927849234608e-07, + "loss": 0.0398, + "step": 95700 + }, + { + "epoch": 6.270720041931468, + "grad_norm": 0.7617247104644775, + "learning_rate": 5.025516580665358e-07, + "loss": 0.0305, + "step": 95710 + }, + { + "epoch": 6.271375221122977, + "grad_norm": 0.48332810401916504, + "learning_rate": 5.002159441761412e-07, + "loss": 0.0263, + "step": 95720 + }, + { + "epoch": 6.272030400314486, + "grad_norm": 0.5514057874679565, + "learning_rate": 4.978856435077062e-07, + "loss": 0.0411, + "step": 95730 + }, + { + "epoch": 6.272685579505995, + "grad_norm": 0.15229403972625732, + "learning_rate": 4.955607563160769e-07, + "loss": 0.0258, + "step": 95740 + }, + { + "epoch": 6.273340758697504, + "grad_norm": 0.29665228724479675, + "learning_rate": 4.932412828554944e-07, + "loss": 0.0192, + "step": 95750 + }, + { + "epoch": 6.273995937889013, + "grad_norm": 0.15646962821483612, + "learning_rate": 4.909272233796114e-07, + "loss": 0.0238, + "step": 95760 + }, + { + "epoch": 6.274651117080522, + "grad_norm": 0.5466071963310242, + "learning_rate": 4.88618578141492e-07, + "loss": 0.0188, + "step": 95770 + }, + { + "epoch": 6.2753062962720305, + "grad_norm": 0.47456666827201843, + "learning_rate": 4.863153473936011e-07, + "loss": 0.027, + "step": 95780 + }, + { + "epoch": 6.275961475463539, + "grad_norm": 0.6114237904548645, + "learning_rate": 4.840175313878259e-07, + "loss": 0.0258, + "step": 95790 + }, + { + "epoch": 6.276616654655048, + "grad_norm": 0.12753495573997498, + "learning_rate": 4.817251303754433e-07, + "loss": 0.0107, + "step": 95800 + }, + { + "epoch": 6.277271833846557, + "grad_norm": 0.27207860350608826, + "learning_rate": 4.794381446071583e-07, + "loss": 0.0252, + "step": 95810 + }, + { + "epoch": 6.277927013038066, + "grad_norm": 0.7108606696128845, + "learning_rate": 4.771565743330597e-07, + "loss": 0.0455, + "step": 95820 + }, + { + "epoch": 6.278582192229575, + "grad_norm": 0.10321065783500671, + "learning_rate": 4.7488041980266463e-07, + "loss": 0.0218, + "step": 95830 + }, + { + "epoch": 6.279237371421083, + "grad_norm": 0.3075089156627655, + "learning_rate": 4.726096812648795e-07, + "loss": 0.0437, + "step": 95840 + }, + { + "epoch": 6.279892550612592, + "grad_norm": 0.1524832397699356, + "learning_rate": 4.703443589680445e-07, + "loss": 0.0375, + "step": 95850 + }, + { + "epoch": 6.280547729804101, + "grad_norm": 0.15848581492900848, + "learning_rate": 4.6808445315987826e-07, + "loss": 0.0304, + "step": 95860 + }, + { + "epoch": 6.28120290899561, + "grad_norm": 0.08621928095817566, + "learning_rate": 4.658299640875274e-07, + "loss": 0.0382, + "step": 95870 + }, + { + "epoch": 6.2818580881871195, + "grad_norm": 0.17785969376564026, + "learning_rate": 4.635808919975393e-07, + "loss": 0.0297, + "step": 95880 + }, + { + "epoch": 6.282513267378628, + "grad_norm": 0.6341018080711365, + "learning_rate": 4.6133723713586706e-07, + "loss": 0.0443, + "step": 95890 + }, + { + "epoch": 6.283168446570137, + "grad_norm": 1.1637498140335083, + "learning_rate": 4.590989997478756e-07, + "loss": 0.024, + "step": 95900 + }, + { + "epoch": 6.283823625761646, + "grad_norm": 0.619429886341095, + "learning_rate": 4.5686618007833027e-07, + "loss": 0.0341, + "step": 95910 + }, + { + "epoch": 6.284478804953155, + "grad_norm": 0.2129313200712204, + "learning_rate": 4.54638778371419e-07, + "loss": 0.0527, + "step": 95920 + }, + { + "epoch": 6.285133984144664, + "grad_norm": 0.7558925747871399, + "learning_rate": 4.524167948707192e-07, + "loss": 0.0477, + "step": 95930 + }, + { + "epoch": 6.285789163336172, + "grad_norm": 0.9038824439048767, + "learning_rate": 4.5020022981921986e-07, + "loss": 0.0358, + "step": 95940 + }, + { + "epoch": 6.286444342527681, + "grad_norm": 0.3354431092739105, + "learning_rate": 4.47989083459327e-07, + "loss": 0.0358, + "step": 95950 + }, + { + "epoch": 6.28709952171919, + "grad_norm": 0.3903663158416748, + "learning_rate": 4.457833560328528e-07, + "loss": 0.0224, + "step": 95960 + }, + { + "epoch": 6.287754700910699, + "grad_norm": 0.3496050536632538, + "learning_rate": 4.4358304778100434e-07, + "loss": 0.0238, + "step": 95970 + }, + { + "epoch": 6.288409880102208, + "grad_norm": 0.07022199779748917, + "learning_rate": 4.413881589444058e-07, + "loss": 0.0229, + "step": 95980 + }, + { + "epoch": 6.2890650592937165, + "grad_norm": 0.974515974521637, + "learning_rate": 4.391986897630873e-07, + "loss": 0.0318, + "step": 95990 + }, + { + "epoch": 6.289720238485225, + "grad_norm": 0.8432802557945251, + "learning_rate": 4.370146404764908e-07, + "loss": 0.0292, + "step": 96000 + }, + { + "epoch": 6.290375417676735, + "grad_norm": 0.4696885943412781, + "learning_rate": 4.3483601132345285e-07, + "loss": 0.0221, + "step": 96010 + }, + { + "epoch": 6.291030596868244, + "grad_norm": 0.48561209440231323, + "learning_rate": 4.32662802542233e-07, + "loss": 0.022, + "step": 96020 + }, + { + "epoch": 6.291685776059753, + "grad_norm": 0.1411409229040146, + "learning_rate": 4.3049501437047444e-07, + "loss": 0.0276, + "step": 96030 + }, + { + "epoch": 6.2923409552512615, + "grad_norm": 1.4298981428146362, + "learning_rate": 4.283326470452653e-07, + "loss": 0.0217, + "step": 96040 + }, + { + "epoch": 6.29299613444277, + "grad_norm": 0.997920036315918, + "learning_rate": 4.2617570080306647e-07, + "loss": 0.0227, + "step": 96050 + }, + { + "epoch": 6.293651313634279, + "grad_norm": 0.5641460418701172, + "learning_rate": 4.2402417587975585e-07, + "loss": 0.0331, + "step": 96060 + }, + { + "epoch": 6.294306492825788, + "grad_norm": 0.12020494043827057, + "learning_rate": 4.218780725106286e-07, + "loss": 0.0199, + "step": 96070 + }, + { + "epoch": 6.294961672017297, + "grad_norm": 0.2967088222503662, + "learning_rate": 4.197373909303748e-07, + "loss": 0.0187, + "step": 96080 + }, + { + "epoch": 6.2956168512088055, + "grad_norm": 0.11655636876821518, + "learning_rate": 4.17602131373096e-07, + "loss": 0.0329, + "step": 96090 + }, + { + "epoch": 6.296272030400314, + "grad_norm": 0.1160365492105484, + "learning_rate": 4.154722940723055e-07, + "loss": 0.033, + "step": 96100 + }, + { + "epoch": 6.296927209591823, + "grad_norm": 0.5629313588142395, + "learning_rate": 4.1334787926091136e-07, + "loss": 0.0296, + "step": 96110 + }, + { + "epoch": 6.297582388783332, + "grad_norm": 0.5083618760108948, + "learning_rate": 4.1122888717124444e-07, + "loss": 0.0275, + "step": 96120 + }, + { + "epoch": 6.298237567974841, + "grad_norm": 0.22072777152061462, + "learning_rate": 4.091153180350249e-07, + "loss": 0.0129, + "step": 96130 + }, + { + "epoch": 6.29889274716635, + "grad_norm": 0.13747480511665344, + "learning_rate": 4.0700717208340125e-07, + "loss": 0.0129, + "step": 96140 + }, + { + "epoch": 6.299547926357858, + "grad_norm": 0.4307585060596466, + "learning_rate": 4.049044495469112e-07, + "loss": 0.0287, + "step": 96150 + }, + { + "epoch": 6.300203105549368, + "grad_norm": 0.6516761779785156, + "learning_rate": 4.0280715065549866e-07, + "loss": 0.0202, + "step": 96160 + }, + { + "epoch": 6.300858284740877, + "grad_norm": 0.39756134152412415, + "learning_rate": 4.0071527563853575e-07, + "loss": 0.0254, + "step": 96170 + }, + { + "epoch": 6.301513463932386, + "grad_norm": 1.1805990934371948, + "learning_rate": 3.9862882472477823e-07, + "loss": 0.0337, + "step": 96180 + }, + { + "epoch": 6.302168643123895, + "grad_norm": 0.13409298658370972, + "learning_rate": 3.9654779814239375e-07, + "loss": 0.0441, + "step": 96190 + }, + { + "epoch": 6.302823822315403, + "grad_norm": 0.44046667218208313, + "learning_rate": 3.944721961189668e-07, + "loss": 0.0163, + "step": 96200 + }, + { + "epoch": 6.303479001506912, + "grad_norm": 1.4474058151245117, + "learning_rate": 3.9240201888147697e-07, + "loss": 0.0218, + "step": 96210 + }, + { + "epoch": 6.304134180698421, + "grad_norm": 0.3767666816711426, + "learning_rate": 3.9033726665632096e-07, + "loss": 0.0197, + "step": 96220 + }, + { + "epoch": 6.30478935988993, + "grad_norm": 0.5915700793266296, + "learning_rate": 3.882779396692959e-07, + "loss": 0.0315, + "step": 96230 + }, + { + "epoch": 6.305444539081439, + "grad_norm": 0.19629064202308655, + "learning_rate": 3.8622403814559947e-07, + "loss": 0.0362, + "step": 96240 + }, + { + "epoch": 6.3060997182729475, + "grad_norm": 0.5651209950447083, + "learning_rate": 3.841755623098575e-07, + "loss": 0.0307, + "step": 96250 + }, + { + "epoch": 6.306754897464456, + "grad_norm": 0.6846607327461243, + "learning_rate": 3.821325123860742e-07, + "loss": 0.0154, + "step": 96260 + }, + { + "epoch": 6.307410076655965, + "grad_norm": 2.250288724899292, + "learning_rate": 3.800948885976818e-07, + "loss": 0.016, + "step": 96270 + }, + { + "epoch": 6.308065255847474, + "grad_norm": 0.2334141731262207, + "learning_rate": 3.780626911675078e-07, + "loss": 0.0216, + "step": 96280 + }, + { + "epoch": 6.308720435038984, + "grad_norm": 0.24188697338104248, + "learning_rate": 3.76035920317791e-07, + "loss": 0.0381, + "step": 96290 + }, + { + "epoch": 6.309375614230492, + "grad_norm": 0.3962734639644623, + "learning_rate": 3.740145762701819e-07, + "loss": 0.0707, + "step": 96300 + }, + { + "epoch": 6.310030793422001, + "grad_norm": 0.13523994386196136, + "learning_rate": 3.7199865924572585e-07, + "loss": 0.022, + "step": 96310 + }, + { + "epoch": 6.31068597261351, + "grad_norm": 0.23679430782794952, + "learning_rate": 3.699881694648799e-07, + "loss": 0.0235, + "step": 96320 + }, + { + "epoch": 6.311341151805019, + "grad_norm": 0.15899673104286194, + "learning_rate": 3.679831071475126e-07, + "loss": 0.0171, + "step": 96330 + }, + { + "epoch": 6.311996330996528, + "grad_norm": 0.09320506453514099, + "learning_rate": 3.65983472512893e-07, + "loss": 0.0294, + "step": 96340 + }, + { + "epoch": 6.3126515101880365, + "grad_norm": 0.2891136407852173, + "learning_rate": 3.6398926577969616e-07, + "loss": 0.0219, + "step": 96350 + }, + { + "epoch": 6.313306689379545, + "grad_norm": 0.7368913888931274, + "learning_rate": 3.620004871660032e-07, + "loss": 0.0242, + "step": 96360 + }, + { + "epoch": 6.313961868571054, + "grad_norm": 0.8984843492507935, + "learning_rate": 3.600171368893124e-07, + "loss": 0.0426, + "step": 96370 + }, + { + "epoch": 6.314617047762563, + "grad_norm": 0.47181960940361023, + "learning_rate": 3.580392151665113e-07, + "loss": 0.0298, + "step": 96380 + }, + { + "epoch": 6.315272226954072, + "grad_norm": 1.0920445919036865, + "learning_rate": 3.5606672221391024e-07, + "loss": 0.0301, + "step": 96390 + }, + { + "epoch": 6.315927406145581, + "grad_norm": 0.5562773942947388, + "learning_rate": 3.540996582472089e-07, + "loss": 0.0286, + "step": 96400 + }, + { + "epoch": 6.316582585337089, + "grad_norm": 0.3432355523109436, + "learning_rate": 3.521380234815297e-07, + "loss": 0.0197, + "step": 96410 + }, + { + "epoch": 6.317237764528598, + "grad_norm": 0.14878961443901062, + "learning_rate": 3.5018181813138985e-07, + "loss": 0.034, + "step": 96420 + }, + { + "epoch": 6.317892943720107, + "grad_norm": 0.5491803884506226, + "learning_rate": 3.482310424107238e-07, + "loss": 0.0294, + "step": 96430 + }, + { + "epoch": 6.318548122911617, + "grad_norm": 0.15430156886577606, + "learning_rate": 3.4628569653285537e-07, + "loss": 0.0174, + "step": 96440 + }, + { + "epoch": 6.3192033021031255, + "grad_norm": 0.45477020740509033, + "learning_rate": 3.44345780710531e-07, + "loss": 0.0344, + "step": 96450 + }, + { + "epoch": 6.319858481294634, + "grad_norm": 0.12727181613445282, + "learning_rate": 3.424112951558978e-07, + "loss": 0.0161, + "step": 96460 + }, + { + "epoch": 6.320513660486143, + "grad_norm": 0.8903011083602905, + "learning_rate": 3.40482240080503e-07, + "loss": 0.0293, + "step": 96470 + }, + { + "epoch": 6.321168839677652, + "grad_norm": 0.8360950350761414, + "learning_rate": 3.3855861569531133e-07, + "loss": 0.021, + "step": 96480 + }, + { + "epoch": 6.321824018869161, + "grad_norm": 0.5224394798278809, + "learning_rate": 3.3664042221068226e-07, + "loss": 0.039, + "step": 96490 + }, + { + "epoch": 6.32247919806067, + "grad_norm": 0.19482924044132233, + "learning_rate": 3.347276598363924e-07, + "loss": 0.0417, + "step": 96500 + }, + { + "epoch": 6.323134377252178, + "grad_norm": 0.7352383732795715, + "learning_rate": 3.328203287816134e-07, + "loss": 0.042, + "step": 96510 + }, + { + "epoch": 6.323789556443687, + "grad_norm": 0.23505999147891998, + "learning_rate": 3.3091842925493386e-07, + "loss": 0.0336, + "step": 96520 + }, + { + "epoch": 6.324444735635196, + "grad_norm": 0.1621103733778, + "learning_rate": 3.2902196146433196e-07, + "loss": 0.0263, + "step": 96530 + }, + { + "epoch": 6.325099914826705, + "grad_norm": 0.20176559686660767, + "learning_rate": 3.2713092561720836e-07, + "loss": 0.0354, + "step": 96540 + }, + { + "epoch": 6.325755094018214, + "grad_norm": 0.13536803424358368, + "learning_rate": 3.2524532192036996e-07, + "loss": 0.0446, + "step": 96550 + }, + { + "epoch": 6.3264102732097225, + "grad_norm": 0.6984161138534546, + "learning_rate": 3.2336515058001836e-07, + "loss": 0.026, + "step": 96560 + }, + { + "epoch": 6.327065452401232, + "grad_norm": 0.35515546798706055, + "learning_rate": 3.2149041180176143e-07, + "loss": 0.0324, + "step": 96570 + }, + { + "epoch": 6.327720631592741, + "grad_norm": 0.7469133734703064, + "learning_rate": 3.1962110579062954e-07, + "loss": 0.0242, + "step": 96580 + }, + { + "epoch": 6.32837581078425, + "grad_norm": 0.1350974291563034, + "learning_rate": 3.177572327510314e-07, + "loss": 0.0266, + "step": 96590 + }, + { + "epoch": 6.329030989975759, + "grad_norm": 0.23416180908679962, + "learning_rate": 3.158987928868151e-07, + "loss": 0.0367, + "step": 96600 + }, + { + "epoch": 6.3296861691672675, + "grad_norm": 0.20927812159061432, + "learning_rate": 3.1404578640120144e-07, + "loss": 0.037, + "step": 96610 + }, + { + "epoch": 6.330341348358776, + "grad_norm": 0.18040503561496735, + "learning_rate": 3.1219821349683933e-07, + "loss": 0.0279, + "step": 96620 + }, + { + "epoch": 6.330996527550285, + "grad_norm": 0.41571173071861267, + "learning_rate": 3.103560743757783e-07, + "loss": 0.0291, + "step": 96630 + }, + { + "epoch": 6.331651706741794, + "grad_norm": 0.35964521765708923, + "learning_rate": 3.0851936923946834e-07, + "loss": 0.0186, + "step": 96640 + }, + { + "epoch": 6.332306885933303, + "grad_norm": 0.5995782017707825, + "learning_rate": 3.06688098288771e-07, + "loss": 0.0267, + "step": 96650 + }, + { + "epoch": 6.3329620651248115, + "grad_norm": 0.11906292289495468, + "learning_rate": 3.0486226172394825e-07, + "loss": 0.0133, + "step": 96660 + }, + { + "epoch": 6.33361724431632, + "grad_norm": 0.23614682257175446, + "learning_rate": 3.030418597446738e-07, + "loss": 0.0266, + "step": 96670 + }, + { + "epoch": 6.334272423507829, + "grad_norm": 0.8534631133079529, + "learning_rate": 3.012268925500272e-07, + "loss": 0.04, + "step": 96680 + }, + { + "epoch": 6.334927602699338, + "grad_norm": 0.18400847911834717, + "learning_rate": 2.994173603384831e-07, + "loss": 0.019, + "step": 96690 + }, + { + "epoch": 6.335582781890847, + "grad_norm": 0.3296569883823395, + "learning_rate": 2.9761326330793317e-07, + "loss": 0.0175, + "step": 96700 + }, + { + "epoch": 6.336237961082356, + "grad_norm": 0.23962543904781342, + "learning_rate": 2.958146016556751e-07, + "loss": 0.0291, + "step": 96710 + }, + { + "epoch": 6.336893140273865, + "grad_norm": 0.8625067472457886, + "learning_rate": 2.94021375578396e-07, + "loss": 0.0315, + "step": 96720 + }, + { + "epoch": 6.337548319465374, + "grad_norm": 0.09961409866809845, + "learning_rate": 2.9223358527221135e-07, + "loss": 0.0123, + "step": 96730 + }, + { + "epoch": 6.338203498656883, + "grad_norm": 0.1605876386165619, + "learning_rate": 2.904512309326257e-07, + "loss": 0.0282, + "step": 96740 + }, + { + "epoch": 6.338858677848392, + "grad_norm": 0.12692134082317352, + "learning_rate": 2.8867431275455545e-07, + "loss": 0.0235, + "step": 96750 + }, + { + "epoch": 6.339513857039901, + "grad_norm": 0.5895891189575195, + "learning_rate": 2.86902830932323e-07, + "loss": 0.0238, + "step": 96760 + }, + { + "epoch": 6.340169036231409, + "grad_norm": 0.21635164320468903, + "learning_rate": 2.851367856596565e-07, + "loss": 0.0239, + "step": 96770 + }, + { + "epoch": 6.340824215422918, + "grad_norm": 0.24752283096313477, + "learning_rate": 2.83376177129685e-07, + "loss": 0.0175, + "step": 96780 + }, + { + "epoch": 6.341479394614427, + "grad_norm": 0.09221428632736206, + "learning_rate": 2.8162100553494884e-07, + "loss": 0.0285, + "step": 96790 + }, + { + "epoch": 6.342134573805936, + "grad_norm": 0.5196470022201538, + "learning_rate": 2.798712710673834e-07, + "loss": 0.0316, + "step": 96800 + }, + { + "epoch": 6.342789752997445, + "grad_norm": 0.6564390659332275, + "learning_rate": 2.7812697391834676e-07, + "loss": 0.0154, + "step": 96810 + }, + { + "epoch": 6.3434449321889534, + "grad_norm": 0.38596493005752563, + "learning_rate": 2.7638811427858627e-07, + "loss": 0.024, + "step": 96820 + }, + { + "epoch": 6.344100111380462, + "grad_norm": 0.5503352880477905, + "learning_rate": 2.746546923382665e-07, + "loss": 0.0326, + "step": 96830 + }, + { + "epoch": 6.344755290571971, + "grad_norm": 0.4511525630950928, + "learning_rate": 2.729267082869469e-07, + "loss": 0.0227, + "step": 96840 + }, + { + "epoch": 6.345410469763481, + "grad_norm": 0.15646813809871674, + "learning_rate": 2.712041623135986e-07, + "loss": 0.0419, + "step": 96850 + }, + { + "epoch": 6.34606564895499, + "grad_norm": 0.12060427665710449, + "learning_rate": 2.694870546065931e-07, + "loss": 0.032, + "step": 96860 + }, + { + "epoch": 6.346720828146498, + "grad_norm": 1.0822561979293823, + "learning_rate": 2.677753853537246e-07, + "loss": 0.0185, + "step": 96870 + }, + { + "epoch": 6.347376007338007, + "grad_norm": 1.238101840019226, + "learning_rate": 2.660691547421601e-07, + "loss": 0.0552, + "step": 96880 + }, + { + "epoch": 6.348031186529516, + "grad_norm": 0.3993348777294159, + "learning_rate": 2.6436836295850033e-07, + "loss": 0.0234, + "step": 96890 + }, + { + "epoch": 6.348686365721025, + "grad_norm": 0.379391610622406, + "learning_rate": 2.6267301018874093e-07, + "loss": 0.0308, + "step": 96900 + }, + { + "epoch": 6.349341544912534, + "grad_norm": 0.7614040970802307, + "learning_rate": 2.6098309661828355e-07, + "loss": 0.0368, + "step": 96910 + }, + { + "epoch": 6.3499967241040425, + "grad_norm": 0.6431868076324463, + "learning_rate": 2.592986224319305e-07, + "loss": 0.0273, + "step": 96920 + }, + { + "epoch": 6.350651903295551, + "grad_norm": 0.16390255093574524, + "learning_rate": 2.576195878139009e-07, + "loss": 0.0241, + "step": 96930 + }, + { + "epoch": 6.35130708248706, + "grad_norm": 0.2689073383808136, + "learning_rate": 2.5594599294780364e-07, + "loss": 0.0207, + "step": 96940 + }, + { + "epoch": 6.351962261678569, + "grad_norm": 0.3082652688026428, + "learning_rate": 2.542778380166644e-07, + "loss": 0.037, + "step": 96950 + }, + { + "epoch": 6.352617440870078, + "grad_norm": 0.3081272840499878, + "learning_rate": 2.526151232029095e-07, + "loss": 0.0242, + "step": 96960 + }, + { + "epoch": 6.3532726200615866, + "grad_norm": 0.18050946295261383, + "learning_rate": 2.5095784868836567e-07, + "loss": 0.0269, + "step": 96970 + }, + { + "epoch": 6.353927799253095, + "grad_norm": 1.1396154165267944, + "learning_rate": 2.493060146542825e-07, + "loss": 0.0276, + "step": 96980 + }, + { + "epoch": 6.354582978444604, + "grad_norm": 0.1844458132982254, + "learning_rate": 2.476596212812876e-07, + "loss": 0.0168, + "step": 96990 + }, + { + "epoch": 6.355238157636114, + "grad_norm": 0.4072859585285187, + "learning_rate": 2.4601866874943703e-07, + "loss": 0.0537, + "step": 97000 + }, + { + "epoch": 6.355893336827623, + "grad_norm": 0.9520748853683472, + "learning_rate": 2.4438315723818163e-07, + "loss": 0.0291, + "step": 97010 + }, + { + "epoch": 6.3565485160191315, + "grad_norm": 0.21627384424209595, + "learning_rate": 2.4275308692637833e-07, + "loss": 0.0136, + "step": 97020 + }, + { + "epoch": 6.35720369521064, + "grad_norm": 0.27462583780288696, + "learning_rate": 2.411284579922901e-07, + "loss": 0.0349, + "step": 97030 + }, + { + "epoch": 6.357858874402149, + "grad_norm": 1.2952829599380493, + "learning_rate": 2.395092706135804e-07, + "loss": 0.0425, + "step": 97040 + }, + { + "epoch": 6.358514053593658, + "grad_norm": 0.1555529534816742, + "learning_rate": 2.3789552496731872e-07, + "loss": 0.0256, + "step": 97050 + }, + { + "epoch": 6.359169232785167, + "grad_norm": 0.12589649856090546, + "learning_rate": 2.3628722122999158e-07, + "loss": 0.0275, + "step": 97060 + }, + { + "epoch": 6.359824411976676, + "grad_norm": 0.13468952476978302, + "learning_rate": 2.3468435957747503e-07, + "loss": 0.014, + "step": 97070 + }, + { + "epoch": 6.360479591168184, + "grad_norm": 0.9211287498474121, + "learning_rate": 2.330869401850566e-07, + "loss": 0.0421, + "step": 97080 + }, + { + "epoch": 6.361134770359693, + "grad_norm": 0.5368008613586426, + "learning_rate": 2.3149496322742437e-07, + "loss": 0.0233, + "step": 97090 + }, + { + "epoch": 6.361789949551202, + "grad_norm": 1.00409996509552, + "learning_rate": 2.2990842887868346e-07, + "loss": 0.0231, + "step": 97100 + }, + { + "epoch": 6.362445128742711, + "grad_norm": 0.3719261586666107, + "learning_rate": 2.2832733731232292e-07, + "loss": 0.02, + "step": 97110 + }, + { + "epoch": 6.36310030793422, + "grad_norm": 0.6794707179069519, + "learning_rate": 2.2675168870124886e-07, + "loss": 0.0384, + "step": 97120 + }, + { + "epoch": 6.3637554871257285, + "grad_norm": 0.23698173463344574, + "learning_rate": 2.2518148321778455e-07, + "loss": 0.0192, + "step": 97130 + }, + { + "epoch": 6.364410666317238, + "grad_norm": 0.30903351306915283, + "learning_rate": 2.2361672103363706e-07, + "loss": 0.0261, + "step": 97140 + }, + { + "epoch": 6.365065845508747, + "grad_norm": 1.1996841430664062, + "learning_rate": 2.220574023199251e-07, + "loss": 0.0729, + "step": 97150 + }, + { + "epoch": 6.365721024700256, + "grad_norm": 1.0694071054458618, + "learning_rate": 2.2050352724717894e-07, + "loss": 0.0295, + "step": 97160 + }, + { + "epoch": 6.366376203891765, + "grad_norm": 0.08153841644525528, + "learning_rate": 2.1895509598532372e-07, + "loss": 0.0559, + "step": 97170 + }, + { + "epoch": 6.367031383083273, + "grad_norm": 0.45068156719207764, + "learning_rate": 2.1741210870369066e-07, + "loss": 0.0302, + "step": 97180 + }, + { + "epoch": 6.367686562274782, + "grad_norm": 0.27167463302612305, + "learning_rate": 2.1587456557102814e-07, + "loss": 0.016, + "step": 97190 + }, + { + "epoch": 6.368341741466291, + "grad_norm": 0.3964502811431885, + "learning_rate": 2.143424667554683e-07, + "loss": 0.0227, + "step": 97200 + }, + { + "epoch": 6.3689969206578, + "grad_norm": 0.13134893774986267, + "learning_rate": 2.128158124245716e-07, + "loss": 0.0213, + "step": 97210 + }, + { + "epoch": 6.369652099849309, + "grad_norm": 0.27131643891334534, + "learning_rate": 2.1129460274527668e-07, + "loss": 0.0235, + "step": 97220 + }, + { + "epoch": 6.3703072790408175, + "grad_norm": 0.6018393039703369, + "learning_rate": 2.0977883788395046e-07, + "loss": 0.0225, + "step": 97230 + }, + { + "epoch": 6.370962458232326, + "grad_norm": 0.13164305686950684, + "learning_rate": 2.0826851800634928e-07, + "loss": 0.0428, + "step": 97240 + }, + { + "epoch": 6.371617637423835, + "grad_norm": 0.6707331538200378, + "learning_rate": 2.0676364327764653e-07, + "loss": 0.0191, + "step": 97250 + }, + { + "epoch": 6.372272816615344, + "grad_norm": 0.33386099338531494, + "learning_rate": 2.0526421386240502e-07, + "loss": 0.0129, + "step": 97260 + }, + { + "epoch": 6.372927995806853, + "grad_norm": 0.2746676206588745, + "learning_rate": 2.0377022992461025e-07, + "loss": 0.0391, + "step": 97270 + }, + { + "epoch": 6.3735831749983625, + "grad_norm": 0.7053083777427673, + "learning_rate": 2.0228169162763156e-07, + "loss": 0.023, + "step": 97280 + }, + { + "epoch": 6.374238354189871, + "grad_norm": 0.5530886650085449, + "learning_rate": 2.0079859913425536e-07, + "loss": 0.0295, + "step": 97290 + }, + { + "epoch": 6.37489353338138, + "grad_norm": 0.4951317608356476, + "learning_rate": 1.9932095260667417e-07, + "loss": 0.0159, + "step": 97300 + }, + { + "epoch": 6.375548712572889, + "grad_norm": 0.5370819568634033, + "learning_rate": 1.9784875220648092e-07, + "loss": 0.0191, + "step": 97310 + }, + { + "epoch": 6.376203891764398, + "grad_norm": 0.2663523256778717, + "learning_rate": 1.9638199809466907e-07, + "loss": 0.0139, + "step": 97320 + }, + { + "epoch": 6.3768590709559065, + "grad_norm": 1.6931859254837036, + "learning_rate": 1.9492069043164364e-07, + "loss": 0.0344, + "step": 97330 + }, + { + "epoch": 6.377514250147415, + "grad_norm": 0.6040779948234558, + "learning_rate": 1.9346482937721567e-07, + "loss": 0.0332, + "step": 97340 + }, + { + "epoch": 6.378169429338924, + "grad_norm": 0.29015207290649414, + "learning_rate": 1.920144150905856e-07, + "loss": 0.0228, + "step": 97350 + }, + { + "epoch": 6.378824608530433, + "grad_norm": 0.42091161012649536, + "learning_rate": 1.9056944773037655e-07, + "loss": 0.0331, + "step": 97360 + }, + { + "epoch": 6.379479787721942, + "grad_norm": 2.088839530944824, + "learning_rate": 1.8912992745460102e-07, + "loss": 0.0319, + "step": 97370 + }, + { + "epoch": 6.380134966913451, + "grad_norm": 0.5839114189147949, + "learning_rate": 1.876958544206886e-07, + "loss": 0.0459, + "step": 97380 + }, + { + "epoch": 6.380790146104959, + "grad_norm": 0.3430182635784149, + "learning_rate": 1.8626722878546942e-07, + "loss": 0.0617, + "step": 97390 + }, + { + "epoch": 6.381445325296468, + "grad_norm": 0.2307797223329544, + "learning_rate": 1.8484405070516851e-07, + "loss": 0.0433, + "step": 97400 + }, + { + "epoch": 6.382100504487977, + "grad_norm": 0.4222326874732971, + "learning_rate": 1.8342632033542805e-07, + "loss": 0.0412, + "step": 97410 + }, + { + "epoch": 6.382755683679486, + "grad_norm": 0.4515330493450165, + "learning_rate": 1.8201403783129066e-07, + "loss": 0.0172, + "step": 97420 + }, + { + "epoch": 6.383410862870996, + "grad_norm": 0.3356415331363678, + "learning_rate": 1.806072033471884e-07, + "loss": 0.0164, + "step": 97430 + }, + { + "epoch": 6.384066042062504, + "grad_norm": 0.176569864153862, + "learning_rate": 1.79205817036987e-07, + "loss": 0.0185, + "step": 97440 + }, + { + "epoch": 6.384721221254013, + "grad_norm": 0.18350109457969666, + "learning_rate": 1.7780987905393065e-07, + "loss": 0.0282, + "step": 97450 + }, + { + "epoch": 6.385376400445522, + "grad_norm": 0.05261107534170151, + "learning_rate": 1.764193895506805e-07, + "loss": 0.0222, + "step": 97460 + }, + { + "epoch": 6.386031579637031, + "grad_norm": 0.1426117718219757, + "learning_rate": 1.7503434867929824e-07, + "loss": 0.0285, + "step": 97470 + }, + { + "epoch": 6.38668675882854, + "grad_norm": 0.1381606161594391, + "learning_rate": 1.7365475659125165e-07, + "loss": 0.0195, + "step": 97480 + }, + { + "epoch": 6.3873419380200485, + "grad_norm": 0.2977893054485321, + "learning_rate": 1.7228061343740332e-07, + "loss": 0.0271, + "step": 97490 + }, + { + "epoch": 6.387997117211557, + "grad_norm": 0.3564405143260956, + "learning_rate": 1.7091191936803308e-07, + "loss": 0.016, + "step": 97500 + }, + { + "epoch": 6.388652296403066, + "grad_norm": 0.13249385356903076, + "learning_rate": 1.6954867453281563e-07, + "loss": 0.0143, + "step": 97510 + }, + { + "epoch": 6.389307475594575, + "grad_norm": 0.16702409088611603, + "learning_rate": 1.6819087908084285e-07, + "loss": 0.0195, + "step": 97520 + }, + { + "epoch": 6.389962654786084, + "grad_norm": 0.1761711984872818, + "learning_rate": 1.6683853316058485e-07, + "loss": 0.0231, + "step": 97530 + }, + { + "epoch": 6.3906178339775925, + "grad_norm": 0.6849392652511597, + "learning_rate": 1.6549163691995106e-07, + "loss": 0.0316, + "step": 97540 + }, + { + "epoch": 6.391273013169101, + "grad_norm": 0.1340857893228531, + "learning_rate": 1.641501905062237e-07, + "loss": 0.0209, + "step": 97550 + }, + { + "epoch": 6.391928192360611, + "grad_norm": 0.14913518726825714, + "learning_rate": 1.628141940661021e-07, + "loss": 0.0203, + "step": 97560 + }, + { + "epoch": 6.39258337155212, + "grad_norm": 0.7605581283569336, + "learning_rate": 1.6148364774569158e-07, + "loss": 0.0268, + "step": 97570 + }, + { + "epoch": 6.393238550743629, + "grad_norm": 0.08422813564538956, + "learning_rate": 1.6015855169050355e-07, + "loss": 0.0153, + "step": 97580 + }, + { + "epoch": 6.3938937299351375, + "grad_norm": 0.5716274976730347, + "learning_rate": 1.5883890604543873e-07, + "loss": 0.0226, + "step": 97590 + }, + { + "epoch": 6.394548909126646, + "grad_norm": 0.305794358253479, + "learning_rate": 1.5752471095482057e-07, + "loss": 0.0433, + "step": 97600 + }, + { + "epoch": 6.395204088318155, + "grad_norm": 0.8087901473045349, + "learning_rate": 1.5621596656235638e-07, + "loss": 0.0173, + "step": 97610 + }, + { + "epoch": 6.395859267509664, + "grad_norm": 0.4291023015975952, + "learning_rate": 1.5491267301117608e-07, + "loss": 0.0203, + "step": 97620 + }, + { + "epoch": 6.396514446701173, + "grad_norm": 0.8856654167175293, + "learning_rate": 1.536148304438101e-07, + "loss": 0.0312, + "step": 97630 + }, + { + "epoch": 6.397169625892682, + "grad_norm": 0.12205275148153305, + "learning_rate": 1.5232243900217823e-07, + "loss": 0.0229, + "step": 97640 + }, + { + "epoch": 6.39782480508419, + "grad_norm": 0.10950011014938354, + "learning_rate": 1.51035498827623e-07, + "loss": 0.0282, + "step": 97650 + }, + { + "epoch": 6.398479984275699, + "grad_norm": 0.3549857437610626, + "learning_rate": 1.497540100608763e-07, + "loss": 0.0234, + "step": 97660 + }, + { + "epoch": 6.399135163467208, + "grad_norm": 0.7198334336280823, + "learning_rate": 1.4847797284208153e-07, + "loss": 0.0402, + "step": 97670 + }, + { + "epoch": 6.399790342658717, + "grad_norm": 0.5656124949455261, + "learning_rate": 1.472073873107882e-07, + "loss": 0.0219, + "step": 97680 + }, + { + "epoch": 6.400445521850226, + "grad_norm": 0.45170098543167114, + "learning_rate": 1.459422536059407e-07, + "loss": 0.0421, + "step": 97690 + }, + { + "epoch": 6.4011007010417345, + "grad_norm": 1.5602607727050781, + "learning_rate": 1.4468257186589508e-07, + "loss": 0.0539, + "step": 97700 + }, + { + "epoch": 6.401755880233244, + "grad_norm": 0.6346864104270935, + "learning_rate": 1.4342834222840217e-07, + "loss": 0.0191, + "step": 97710 + }, + { + "epoch": 6.402411059424753, + "grad_norm": 0.1629628986120224, + "learning_rate": 1.4217956483063566e-07, + "loss": 0.0316, + "step": 97720 + }, + { + "epoch": 6.403066238616262, + "grad_norm": 0.04273862764239311, + "learning_rate": 1.409362398091474e-07, + "loss": 0.0285, + "step": 97730 + }, + { + "epoch": 6.403721417807771, + "grad_norm": 0.3155580461025238, + "learning_rate": 1.3969836729990638e-07, + "loss": 0.0212, + "step": 97740 + }, + { + "epoch": 6.404376596999279, + "grad_norm": 0.4678639769554138, + "learning_rate": 1.384659474382932e-07, + "loss": 0.0264, + "step": 97750 + }, + { + "epoch": 6.405031776190788, + "grad_norm": 0.9000741839408875, + "learning_rate": 1.372389803590779e-07, + "loss": 0.0169, + "step": 97760 + }, + { + "epoch": 6.405686955382297, + "grad_norm": 0.3557407557964325, + "learning_rate": 1.360174661964364e-07, + "loss": 0.0313, + "step": 97770 + }, + { + "epoch": 6.406342134573806, + "grad_norm": 1.0255427360534668, + "learning_rate": 1.3480140508396188e-07, + "loss": 0.0228, + "step": 97780 + }, + { + "epoch": 6.406997313765315, + "grad_norm": 0.1896803230047226, + "learning_rate": 1.335907971546313e-07, + "loss": 0.0147, + "step": 97790 + }, + { + "epoch": 6.4076524929568235, + "grad_norm": 0.6563982367515564, + "learning_rate": 1.3238564254083875e-07, + "loss": 0.0313, + "step": 97800 + }, + { + "epoch": 6.408307672148332, + "grad_norm": 0.17765763401985168, + "learning_rate": 1.3118594137437323e-07, + "loss": 0.0244, + "step": 97810 + }, + { + "epoch": 6.408962851339841, + "grad_norm": 0.2964567542076111, + "learning_rate": 1.2999169378644094e-07, + "loss": 0.0197, + "step": 97820 + }, + { + "epoch": 6.40961803053135, + "grad_norm": 0.9107770919799805, + "learning_rate": 1.2880289990763738e-07, + "loss": 0.035, + "step": 97830 + }, + { + "epoch": 6.41027320972286, + "grad_norm": 0.23877087235450745, + "learning_rate": 1.2761955986796968e-07, + "loss": 0.0213, + "step": 97840 + }, + { + "epoch": 6.4109283889143684, + "grad_norm": 0.5443270802497864, + "learning_rate": 1.264416737968399e-07, + "loss": 0.0383, + "step": 97850 + }, + { + "epoch": 6.411583568105877, + "grad_norm": 0.5325798392295837, + "learning_rate": 1.2526924182307275e-07, + "loss": 0.0269, + "step": 97860 + }, + { + "epoch": 6.412238747297386, + "grad_norm": 0.3292677700519562, + "learning_rate": 1.2410226407487124e-07, + "loss": 0.0302, + "step": 97870 + }, + { + "epoch": 6.412893926488895, + "grad_norm": 0.3555963635444641, + "learning_rate": 1.229407406798555e-07, + "loss": 0.0355, + "step": 97880 + }, + { + "epoch": 6.413549105680404, + "grad_norm": 0.1849622279405594, + "learning_rate": 1.2178467176505726e-07, + "loss": 0.0193, + "step": 97890 + }, + { + "epoch": 6.4142042848719125, + "grad_norm": 0.12093784660100937, + "learning_rate": 1.2063405745689205e-07, + "loss": 0.0204, + "step": 97900 + }, + { + "epoch": 6.414859464063421, + "grad_norm": 0.4714623987674713, + "learning_rate": 1.1948889788119812e-07, + "loss": 0.0377, + "step": 97910 + }, + { + "epoch": 6.41551464325493, + "grad_norm": 0.34522247314453125, + "learning_rate": 1.1834919316320303e-07, + "loss": 0.0291, + "step": 97920 + }, + { + "epoch": 6.416169822446439, + "grad_norm": 0.42902952432632446, + "learning_rate": 1.1721494342754048e-07, + "loss": 0.0338, + "step": 97930 + }, + { + "epoch": 6.416825001637948, + "grad_norm": 0.35712897777557373, + "learning_rate": 1.1608614879825563e-07, + "loss": 0.0231, + "step": 97940 + }, + { + "epoch": 6.417480180829457, + "grad_norm": 0.4967474937438965, + "learning_rate": 1.1496280939879422e-07, + "loss": 0.0154, + "step": 97950 + }, + { + "epoch": 6.418135360020965, + "grad_norm": 0.3808959722518921, + "learning_rate": 1.1384492535199687e-07, + "loss": 0.0353, + "step": 97960 + }, + { + "epoch": 6.418790539212474, + "grad_norm": 0.39988845586776733, + "learning_rate": 1.1273249678011578e-07, + "loss": 0.0183, + "step": 97970 + }, + { + "epoch": 6.419445718403983, + "grad_norm": 0.7424031496047974, + "learning_rate": 1.1162552380480362e-07, + "loss": 0.0272, + "step": 97980 + }, + { + "epoch": 6.420100897595493, + "grad_norm": 0.0624440535902977, + "learning_rate": 1.1052400654711359e-07, + "loss": 0.0279, + "step": 97990 + }, + { + "epoch": 6.4207560767870016, + "grad_norm": 0.9729273319244385, + "learning_rate": 1.0942794512751597e-07, + "loss": 0.0367, + "step": 98000 + }, + { + "epoch": 6.42141125597851, + "grad_norm": 1.3120473623275757, + "learning_rate": 1.0833733966587045e-07, + "loss": 0.0233, + "step": 98010 + }, + { + "epoch": 6.422066435170019, + "grad_norm": 0.9352582693099976, + "learning_rate": 1.0725219028143718e-07, + "loss": 0.0245, + "step": 98020 + }, + { + "epoch": 6.422721614361528, + "grad_norm": 0.12535522878170013, + "learning_rate": 1.0617249709289345e-07, + "loss": 0.0207, + "step": 98030 + }, + { + "epoch": 6.423376793553037, + "grad_norm": 0.4464684724807739, + "learning_rate": 1.0509826021831703e-07, + "loss": 0.0174, + "step": 98040 + }, + { + "epoch": 6.424031972744546, + "grad_norm": 0.08217024803161621, + "learning_rate": 1.0402947977517508e-07, + "loss": 0.0627, + "step": 98050 + }, + { + "epoch": 6.424687151936054, + "grad_norm": 0.34110572934150696, + "learning_rate": 1.0296615588035185e-07, + "loss": 0.0513, + "step": 98060 + }, + { + "epoch": 6.425342331127563, + "grad_norm": 1.9283145666122437, + "learning_rate": 1.0190828865012659e-07, + "loss": 0.0382, + "step": 98070 + }, + { + "epoch": 6.425997510319072, + "grad_norm": 0.48748263716697693, + "learning_rate": 1.008558782001956e-07, + "loss": 0.0213, + "step": 98080 + }, + { + "epoch": 6.426652689510581, + "grad_norm": 0.07687485218048096, + "learning_rate": 9.980892464563906e-08, + "loss": 0.0245, + "step": 98090 + }, + { + "epoch": 6.42730786870209, + "grad_norm": 0.3739432990550995, + "learning_rate": 9.876742810095985e-08, + "loss": 0.0168, + "step": 98100 + }, + { + "epoch": 6.4279630478935985, + "grad_norm": 0.6961866021156311, + "learning_rate": 9.77313886800446e-08, + "loss": 0.0161, + "step": 98110 + }, + { + "epoch": 6.428618227085108, + "grad_norm": 0.503907322883606, + "learning_rate": 9.670080649619717e-08, + "loss": 0.0285, + "step": 98120 + }, + { + "epoch": 6.429273406276617, + "grad_norm": 1.038221836090088, + "learning_rate": 9.567568166212737e-08, + "loss": 0.0305, + "step": 98130 + }, + { + "epoch": 6.429928585468126, + "grad_norm": 1.0313483476638794, + "learning_rate": 9.465601428992888e-08, + "loss": 0.0162, + "step": 98140 + }, + { + "epoch": 6.430583764659635, + "grad_norm": 0.12810292840003967, + "learning_rate": 9.364180449111803e-08, + "loss": 0.0304, + "step": 98150 + }, + { + "epoch": 6.4312389438511435, + "grad_norm": 0.6832948327064514, + "learning_rate": 9.263305237661169e-08, + "loss": 0.0301, + "step": 98160 + }, + { + "epoch": 6.431894123042652, + "grad_norm": 0.30534884333610535, + "learning_rate": 9.162975805671603e-08, + "loss": 0.0301, + "step": 98170 + }, + { + "epoch": 6.432549302234161, + "grad_norm": 1.2251535654067993, + "learning_rate": 9.063192164115442e-08, + "loss": 0.039, + "step": 98180 + }, + { + "epoch": 6.43320448142567, + "grad_norm": 1.1052488088607788, + "learning_rate": 8.963954323904511e-08, + "loss": 0.0358, + "step": 98190 + }, + { + "epoch": 6.433859660617179, + "grad_norm": 0.618790328502655, + "learning_rate": 8.865262295891796e-08, + "loss": 0.0232, + "step": 98200 + }, + { + "epoch": 6.4345148398086875, + "grad_norm": 0.4578779935836792, + "learning_rate": 8.767116090870331e-08, + "loss": 0.0213, + "step": 98210 + }, + { + "epoch": 6.435170019000196, + "grad_norm": 0.6519488096237183, + "learning_rate": 8.669515719572086e-08, + "loss": 0.0285, + "step": 98220 + }, + { + "epoch": 6.435825198191705, + "grad_norm": 0.10137177258729935, + "learning_rate": 8.572461192671855e-08, + "loss": 0.0153, + "step": 98230 + }, + { + "epoch": 6.436480377383214, + "grad_norm": 1.2065284252166748, + "learning_rate": 8.475952520782815e-08, + "loss": 0.0222, + "step": 98240 + }, + { + "epoch": 6.437135556574723, + "grad_norm": 0.40897777676582336, + "learning_rate": 8.379989714458747e-08, + "loss": 0.0311, + "step": 98250 + }, + { + "epoch": 6.437790735766232, + "grad_norm": 0.3132737874984741, + "learning_rate": 8.284572784194034e-08, + "loss": 0.0274, + "step": 98260 + }, + { + "epoch": 6.438445914957741, + "grad_norm": 0.3072531521320343, + "learning_rate": 8.189701740424216e-08, + "loss": 0.0126, + "step": 98270 + }, + { + "epoch": 6.43910109414925, + "grad_norm": 0.22205345332622528, + "learning_rate": 8.095376593522663e-08, + "loss": 0.0266, + "step": 98280 + }, + { + "epoch": 6.439756273340759, + "grad_norm": 0.8203412294387817, + "learning_rate": 8.001597353806123e-08, + "loss": 0.0236, + "step": 98290 + }, + { + "epoch": 6.440411452532268, + "grad_norm": 0.15510351955890656, + "learning_rate": 7.90836403152917e-08, + "loss": 0.0263, + "step": 98300 + }, + { + "epoch": 6.441066631723777, + "grad_norm": 0.1887347251176834, + "learning_rate": 7.815676636888092e-08, + "loss": 0.0161, + "step": 98310 + }, + { + "epoch": 6.441721810915285, + "grad_norm": 0.5301778316497803, + "learning_rate": 7.723535180019226e-08, + "loss": 0.0461, + "step": 98320 + }, + { + "epoch": 6.442376990106794, + "grad_norm": 0.08084898442029953, + "learning_rate": 7.6319396709984e-08, + "loss": 0.0258, + "step": 98330 + }, + { + "epoch": 6.443032169298303, + "grad_norm": 0.810594916343689, + "learning_rate": 7.540890119843158e-08, + "loss": 0.0387, + "step": 98340 + }, + { + "epoch": 6.443687348489812, + "grad_norm": 0.6539959907531738, + "learning_rate": 7.450386536509979e-08, + "loss": 0.019, + "step": 98350 + }, + { + "epoch": 6.444342527681321, + "grad_norm": 0.4069642126560211, + "learning_rate": 7.360428930895947e-08, + "loss": 0.024, + "step": 98360 + }, + { + "epoch": 6.4449977068728295, + "grad_norm": 0.39110907912254333, + "learning_rate": 7.271017312839302e-08, + "loss": 0.0201, + "step": 98370 + }, + { + "epoch": 6.445652886064338, + "grad_norm": 1.2633816003799438, + "learning_rate": 7.182151692118333e-08, + "loss": 0.0248, + "step": 98380 + }, + { + "epoch": 6.446308065255847, + "grad_norm": 0.3129615783691406, + "learning_rate": 7.093832078449713e-08, + "loss": 0.0194, + "step": 98390 + }, + { + "epoch": 6.446963244447357, + "grad_norm": 0.11790487915277481, + "learning_rate": 7.006058481493493e-08, + "loss": 0.0246, + "step": 98400 + }, + { + "epoch": 6.447618423638866, + "grad_norm": 0.4457651674747467, + "learning_rate": 6.918830910847552e-08, + "loss": 0.0174, + "step": 98410 + }, + { + "epoch": 6.448273602830374, + "grad_norm": 0.9508020281791687, + "learning_rate": 6.832149376052033e-08, + "loss": 0.0641, + "step": 98420 + }, + { + "epoch": 6.448928782021883, + "grad_norm": 0.46701323986053467, + "learning_rate": 6.746013886584912e-08, + "loss": 0.0291, + "step": 98430 + }, + { + "epoch": 6.449583961213392, + "grad_norm": 0.09857923537492752, + "learning_rate": 6.660424451866431e-08, + "loss": 0.0387, + "step": 98440 + }, + { + "epoch": 6.450239140404901, + "grad_norm": 0.7975029349327087, + "learning_rate": 6.575381081256327e-08, + "loss": 0.0257, + "step": 98450 + }, + { + "epoch": 6.45089431959641, + "grad_norm": 0.38058531284332275, + "learning_rate": 6.49088378405549e-08, + "loss": 0.0215, + "step": 98460 + }, + { + "epoch": 6.4515494987879185, + "grad_norm": 0.3878282904624939, + "learning_rate": 6.406932569503755e-08, + "loss": 0.0306, + "step": 98470 + }, + { + "epoch": 6.452204677979427, + "grad_norm": 0.060727838426828384, + "learning_rate": 6.323527446782662e-08, + "loss": 0.0236, + "step": 98480 + }, + { + "epoch": 6.452859857170936, + "grad_norm": 0.3978220224380493, + "learning_rate": 6.240668425012142e-08, + "loss": 0.0469, + "step": 98490 + }, + { + "epoch": 6.453515036362445, + "grad_norm": 0.2666739821434021, + "learning_rate": 6.158355513254388e-08, + "loss": 0.0262, + "step": 98500 + }, + { + "epoch": 6.454170215553954, + "grad_norm": 1.4337356090545654, + "learning_rate": 6.076588720510534e-08, + "loss": 0.0285, + "step": 98510 + }, + { + "epoch": 6.454825394745463, + "grad_norm": 1.639498233795166, + "learning_rate": 5.995368055722316e-08, + "loss": 0.0418, + "step": 98520 + }, + { + "epoch": 6.455480573936971, + "grad_norm": 0.6285361647605896, + "learning_rate": 5.914693527773185e-08, + "loss": 0.0354, + "step": 98530 + }, + { + "epoch": 6.45613575312848, + "grad_norm": 0.19706687331199646, + "learning_rate": 5.834565145483861e-08, + "loss": 0.0239, + "step": 98540 + }, + { + "epoch": 6.45679093231999, + "grad_norm": 0.27305155992507935, + "learning_rate": 5.754982917618446e-08, + "loss": 0.0381, + "step": 98550 + }, + { + "epoch": 6.457446111511499, + "grad_norm": 0.9228254556655884, + "learning_rate": 5.675946852878866e-08, + "loss": 0.0209, + "step": 98560 + }, + { + "epoch": 6.4581012907030075, + "grad_norm": 0.16998232901096344, + "learning_rate": 5.59745695990932e-08, + "loss": 0.0276, + "step": 98570 + }, + { + "epoch": 6.458756469894516, + "grad_norm": 1.031331181526184, + "learning_rate": 5.51951324729294e-08, + "loss": 0.0283, + "step": 98580 + }, + { + "epoch": 6.459411649086025, + "grad_norm": 0.9018298387527466, + "learning_rate": 5.4421157235529093e-08, + "loss": 0.0367, + "step": 98590 + }, + { + "epoch": 6.460066828277534, + "grad_norm": 0.3209682106971741, + "learning_rate": 5.365264397154679e-08, + "loss": 0.0371, + "step": 98600 + }, + { + "epoch": 6.460722007469043, + "grad_norm": 0.3356700837612152, + "learning_rate": 5.288959276501526e-08, + "loss": 0.0211, + "step": 98610 + }, + { + "epoch": 6.461377186660552, + "grad_norm": 0.1788172423839569, + "learning_rate": 5.2132003699378876e-08, + "loss": 0.0292, + "step": 98620 + }, + { + "epoch": 6.46203236585206, + "grad_norm": 0.4469262957572937, + "learning_rate": 5.137987685749912e-08, + "loss": 0.0466, + "step": 98630 + }, + { + "epoch": 6.462687545043569, + "grad_norm": 0.6807342171669006, + "learning_rate": 5.0633212321610227e-08, + "loss": 0.0205, + "step": 98640 + }, + { + "epoch": 6.463342724235078, + "grad_norm": 0.4482610523700714, + "learning_rate": 4.989201017338019e-08, + "loss": 0.0215, + "step": 98650 + }, + { + "epoch": 6.463997903426587, + "grad_norm": 0.44817373156547546, + "learning_rate": 4.9156270493860846e-08, + "loss": 0.0433, + "step": 98660 + }, + { + "epoch": 6.464653082618096, + "grad_norm": 0.1343693882226944, + "learning_rate": 4.842599336351561e-08, + "loss": 0.044, + "step": 98670 + }, + { + "epoch": 6.465308261809605, + "grad_norm": 0.5993519425392151, + "learning_rate": 4.7701178862197274e-08, + "loss": 0.0271, + "step": 98680 + }, + { + "epoch": 6.465963441001114, + "grad_norm": 0.5440029501914978, + "learning_rate": 4.698182706918131e-08, + "loss": 0.0365, + "step": 98690 + }, + { + "epoch": 6.466618620192623, + "grad_norm": 0.5174829959869385, + "learning_rate": 4.6267938063121466e-08, + "loss": 0.0216, + "step": 98700 + }, + { + "epoch": 6.467273799384132, + "grad_norm": 0.6058070659637451, + "learning_rate": 4.555951192209973e-08, + "loss": 0.0356, + "step": 98710 + }, + { + "epoch": 6.467928978575641, + "grad_norm": 0.7090508937835693, + "learning_rate": 4.485654872358747e-08, + "loss": 0.0284, + "step": 98720 + }, + { + "epoch": 6.4685841577671495, + "grad_norm": 0.5449764728546143, + "learning_rate": 4.415904854445097e-08, + "loss": 0.0507, + "step": 98730 + }, + { + "epoch": 6.469239336958658, + "grad_norm": 0.1690400093793869, + "learning_rate": 4.34670114609792e-08, + "loss": 0.0154, + "step": 98740 + }, + { + "epoch": 6.469894516150167, + "grad_norm": 0.4391169548034668, + "learning_rate": 4.278043754884498e-08, + "loss": 0.0327, + "step": 98750 + }, + { + "epoch": 6.470549695341676, + "grad_norm": 0.5238998532295227, + "learning_rate": 4.209932688312712e-08, + "loss": 0.0331, + "step": 98760 + }, + { + "epoch": 6.471204874533185, + "grad_norm": 0.3549748957157135, + "learning_rate": 4.14236795383216e-08, + "loss": 0.0341, + "step": 98770 + }, + { + "epoch": 6.4718600537246935, + "grad_norm": 0.35203292965888977, + "learning_rate": 4.075349558830821e-08, + "loss": 0.0163, + "step": 98780 + }, + { + "epoch": 6.472515232916202, + "grad_norm": 1.8837957382202148, + "learning_rate": 4.008877510638387e-08, + "loss": 0.0646, + "step": 98790 + }, + { + "epoch": 6.473170412107711, + "grad_norm": 0.4626079797744751, + "learning_rate": 3.942951816523488e-08, + "loss": 0.0252, + "step": 98800 + }, + { + "epoch": 6.47382559129922, + "grad_norm": 0.479135662317276, + "learning_rate": 3.8775724836959125e-08, + "loss": 0.0331, + "step": 98810 + }, + { + "epoch": 6.474480770490729, + "grad_norm": 0.18098968267440796, + "learning_rate": 3.812739519305497e-08, + "loss": 0.0323, + "step": 98820 + }, + { + "epoch": 6.4751359496822385, + "grad_norm": 0.8058443069458008, + "learning_rate": 3.748452930442126e-08, + "loss": 0.0151, + "step": 98830 + }, + { + "epoch": 6.475791128873747, + "grad_norm": 2.984621286392212, + "learning_rate": 3.6847127241362855e-08, + "loss": 0.0365, + "step": 98840 + }, + { + "epoch": 6.476446308065256, + "grad_norm": 0.24944289028644562, + "learning_rate": 3.621518907358512e-08, + "loss": 0.0259, + "step": 98850 + }, + { + "epoch": 6.477101487256765, + "grad_norm": 0.3516736328601837, + "learning_rate": 3.558871487019388e-08, + "loss": 0.0323, + "step": 98860 + }, + { + "epoch": 6.477756666448274, + "grad_norm": 1.505363941192627, + "learning_rate": 3.4967704699701007e-08, + "loss": 0.0346, + "step": 98870 + }, + { + "epoch": 6.478411845639783, + "grad_norm": 1.2667739391326904, + "learning_rate": 3.435215863001884e-08, + "loss": 0.0517, + "step": 98880 + }, + { + "epoch": 6.479067024831291, + "grad_norm": 0.09078451991081238, + "learning_rate": 3.374207672846019e-08, + "loss": 0.0443, + "step": 98890 + }, + { + "epoch": 6.4797222040228, + "grad_norm": 0.8132898211479187, + "learning_rate": 3.313745906174392e-08, + "loss": 0.0412, + "step": 98900 + }, + { + "epoch": 6.480377383214309, + "grad_norm": 0.7901007533073425, + "learning_rate": 3.253830569599492e-08, + "loss": 0.0367, + "step": 98910 + }, + { + "epoch": 6.481032562405818, + "grad_norm": 0.3777218163013458, + "learning_rate": 3.1944616696727436e-08, + "loss": 0.0343, + "step": 98920 + }, + { + "epoch": 6.481687741597327, + "grad_norm": 0.46492400765419006, + "learning_rate": 3.135639212887287e-08, + "loss": 0.0268, + "step": 98930 + }, + { + "epoch": 6.4823429207888354, + "grad_norm": 0.5895746350288391, + "learning_rate": 3.077363205675754e-08, + "loss": 0.0353, + "step": 98940 + }, + { + "epoch": 6.482998099980344, + "grad_norm": 0.5215312242507935, + "learning_rate": 3.0196336544113804e-08, + "loss": 0.015, + "step": 98950 + }, + { + "epoch": 6.483653279171854, + "grad_norm": 0.3583359718322754, + "learning_rate": 2.9624505654063383e-08, + "loss": 0.0278, + "step": 98960 + }, + { + "epoch": 6.484308458363363, + "grad_norm": 0.2187940627336502, + "learning_rate": 2.90581394491507e-08, + "loss": 0.0404, + "step": 98970 + }, + { + "epoch": 6.484963637554872, + "grad_norm": 0.23390017449855804, + "learning_rate": 2.8497237991309545e-08, + "loss": 0.0217, + "step": 98980 + }, + { + "epoch": 6.48561881674638, + "grad_norm": 0.47215232253074646, + "learning_rate": 2.7941801341879735e-08, + "loss": 0.0244, + "step": 98990 + }, + { + "epoch": 6.486273995937889, + "grad_norm": 0.13957789540290833, + "learning_rate": 2.7391829561601578e-08, + "loss": 0.0349, + "step": 99000 + }, + { + "epoch": 6.486929175129398, + "grad_norm": 0.3392374515533447, + "learning_rate": 2.6847322710621403e-08, + "loss": 0.0302, + "step": 99010 + }, + { + "epoch": 6.487584354320907, + "grad_norm": 0.42659908533096313, + "learning_rate": 2.6308280848486022e-08, + "loss": 0.0419, + "step": 99020 + }, + { + "epoch": 6.488239533512416, + "grad_norm": 0.5919580459594727, + "learning_rate": 2.5774704034137177e-08, + "loss": 0.0483, + "step": 99030 + }, + { + "epoch": 6.4888947127039245, + "grad_norm": 0.4295472800731659, + "learning_rate": 2.524659232593929e-08, + "loss": 0.0274, + "step": 99040 + }, + { + "epoch": 6.489549891895433, + "grad_norm": 0.11383741348981857, + "learning_rate": 2.472394578163506e-08, + "loss": 0.0323, + "step": 99050 + }, + { + "epoch": 6.490205071086942, + "grad_norm": 0.49023517966270447, + "learning_rate": 2.4206764458378772e-08, + "loss": 0.0169, + "step": 99060 + }, + { + "epoch": 6.490860250278451, + "grad_norm": 0.3228576183319092, + "learning_rate": 2.3695048412736285e-08, + "loss": 0.0252, + "step": 99070 + }, + { + "epoch": 6.49151542946996, + "grad_norm": 0.3412066698074341, + "learning_rate": 2.3188797700668395e-08, + "loss": 0.038, + "step": 99080 + }, + { + "epoch": 6.492170608661469, + "grad_norm": 0.44869908690452576, + "learning_rate": 2.268801237753082e-08, + "loss": 0.0428, + "step": 99090 + }, + { + "epoch": 6.492825787852977, + "grad_norm": 0.23574714362621307, + "learning_rate": 2.2192692498090862e-08, + "loss": 0.0245, + "step": 99100 + }, + { + "epoch": 6.493480967044487, + "grad_norm": 0.7763553857803345, + "learning_rate": 2.170283811652185e-08, + "loss": 0.0471, + "step": 99110 + }, + { + "epoch": 6.494136146235996, + "grad_norm": 0.841831624507904, + "learning_rate": 2.12184492863865e-08, + "loss": 0.0572, + "step": 99120 + }, + { + "epoch": 6.494791325427505, + "grad_norm": 0.6307092308998108, + "learning_rate": 2.073952606066465e-08, + "loss": 0.0179, + "step": 99130 + }, + { + "epoch": 6.4954465046190135, + "grad_norm": 0.18250197172164917, + "learning_rate": 2.026606849171997e-08, + "loss": 0.0223, + "step": 99140 + }, + { + "epoch": 6.496101683810522, + "grad_norm": 0.09277075529098511, + "learning_rate": 1.9798076631333262e-08, + "loss": 0.0334, + "step": 99150 + }, + { + "epoch": 6.496756863002031, + "grad_norm": 1.0260674953460693, + "learning_rate": 1.933555053069136e-08, + "loss": 0.0447, + "step": 99160 + }, + { + "epoch": 6.49741204219354, + "grad_norm": 0.10007373243570328, + "learning_rate": 1.887849024036492e-08, + "loss": 0.0096, + "step": 99170 + }, + { + "epoch": 6.498067221385049, + "grad_norm": 0.1393079161643982, + "learning_rate": 1.8426895810341736e-08, + "loss": 0.021, + "step": 99180 + }, + { + "epoch": 6.498722400576558, + "grad_norm": 0.34568870067596436, + "learning_rate": 1.798076729000453e-08, + "loss": 0.0289, + "step": 99190 + }, + { + "epoch": 6.499377579768066, + "grad_norm": 0.5821666717529297, + "learning_rate": 1.75401047281476e-08, + "loss": 0.0153, + "step": 99200 + }, + { + "epoch": 6.500032758959575, + "grad_norm": 0.5447115898132324, + "learning_rate": 1.7104908172954626e-08, + "loss": 0.0288, + "step": 99210 + }, + { + "epoch": 6.500687938151084, + "grad_norm": 0.3532538414001465, + "learning_rate": 1.6675177672020868e-08, + "loss": 0.0178, + "step": 99220 + }, + { + "epoch": 6.501343117342593, + "grad_norm": 0.7720922231674194, + "learning_rate": 1.6250913272342073e-08, + "loss": 0.0441, + "step": 99230 + }, + { + "epoch": 6.5019982965341026, + "grad_norm": 0.1911877691745758, + "learning_rate": 1.5832115020314453e-08, + "loss": 0.0218, + "step": 99240 + }, + { + "epoch": 6.5026534757256105, + "grad_norm": 0.6136397123336792, + "learning_rate": 1.5418782961734712e-08, + "loss": 0.0292, + "step": 99250 + }, + { + "epoch": 6.50330865491712, + "grad_norm": 0.18925383687019348, + "learning_rate": 1.501091714181113e-08, + "loss": 0.0198, + "step": 99260 + }, + { + "epoch": 6.503963834108629, + "grad_norm": 0.733181357383728, + "learning_rate": 1.460851760513582e-08, + "loss": 0.0286, + "step": 99270 + }, + { + "epoch": 6.504619013300138, + "grad_norm": 0.4728134870529175, + "learning_rate": 1.4211584395723566e-08, + "loss": 0.0205, + "step": 99280 + }, + { + "epoch": 6.505274192491647, + "grad_norm": 0.24326598644256592, + "learning_rate": 1.382011755697854e-08, + "loss": 0.0453, + "step": 99290 + }, + { + "epoch": 6.505929371683155, + "grad_norm": 1.3394826650619507, + "learning_rate": 1.3434117131716494e-08, + "loss": 0.0389, + "step": 99300 + }, + { + "epoch": 6.506584550874664, + "grad_norm": 0.18864792585372925, + "learning_rate": 1.3053583162142557e-08, + "loss": 0.0282, + "step": 99310 + }, + { + "epoch": 6.507239730066173, + "grad_norm": 0.21952655911445618, + "learning_rate": 1.2678515689873438e-08, + "loss": 0.0294, + "step": 99320 + }, + { + "epoch": 6.507894909257682, + "grad_norm": 0.2692555785179138, + "learning_rate": 1.2308914755931877e-08, + "loss": 0.0223, + "step": 99330 + }, + { + "epoch": 6.508550088449191, + "grad_norm": 0.18139973282814026, + "learning_rate": 1.1944780400729993e-08, + "loss": 0.0314, + "step": 99340 + }, + { + "epoch": 6.5092052676406995, + "grad_norm": 0.21939006447792053, + "learning_rate": 1.1586112664085935e-08, + "loss": 0.0169, + "step": 99350 + }, + { + "epoch": 6.509860446832208, + "grad_norm": 0.3155728578567505, + "learning_rate": 1.1232911585234984e-08, + "loss": 0.0359, + "step": 99360 + }, + { + "epoch": 6.510515626023717, + "grad_norm": 0.05681530758738518, + "learning_rate": 1.0885177202790697e-08, + "loss": 0.0456, + "step": 99370 + }, + { + "epoch": 6.511170805215226, + "grad_norm": 0.7608529925346375, + "learning_rate": 1.0542909554789315e-08, + "loss": 0.0181, + "step": 99380 + }, + { + "epoch": 6.511825984406736, + "grad_norm": 1.9225462675094604, + "learning_rate": 1.0206108678656456e-08, + "loss": 0.0218, + "step": 99390 + }, + { + "epoch": 6.5124811635982445, + "grad_norm": 0.487578809261322, + "learning_rate": 9.874774611223769e-09, + "loss": 0.0205, + "step": 99400 + }, + { + "epoch": 6.513136342789753, + "grad_norm": 0.8369139432907104, + "learning_rate": 9.548907388728933e-09, + "loss": 0.0231, + "step": 99410 + }, + { + "epoch": 6.513791521981262, + "grad_norm": 0.25892579555511475, + "learning_rate": 9.228507046804557e-09, + "loss": 0.0403, + "step": 99420 + }, + { + "epoch": 6.514446701172771, + "grad_norm": 0.5666705369949341, + "learning_rate": 8.91357362048928e-09, + "loss": 0.0137, + "step": 99430 + }, + { + "epoch": 6.51510188036428, + "grad_norm": 0.4682944416999817, + "learning_rate": 8.604107144227769e-09, + "loss": 0.0336, + "step": 99440 + }, + { + "epoch": 6.5157570595557885, + "grad_norm": 0.22846122086048126, + "learning_rate": 8.300107651859623e-09, + "loss": 0.0215, + "step": 99450 + }, + { + "epoch": 6.516412238747297, + "grad_norm": 0.10048479586839676, + "learning_rate": 8.001575176630472e-09, + "loss": 0.0182, + "step": 99460 + }, + { + "epoch": 6.517067417938806, + "grad_norm": 0.12490063160657883, + "learning_rate": 7.708509751186422e-09, + "loss": 0.0274, + "step": 99470 + }, + { + "epoch": 6.517722597130315, + "grad_norm": 0.11986622959375381, + "learning_rate": 7.420911407579611e-09, + "loss": 0.0422, + "step": 99480 + }, + { + "epoch": 6.518377776321824, + "grad_norm": 1.7540315389633179, + "learning_rate": 7.13878017725711e-09, + "loss": 0.0337, + "step": 99490 + }, + { + "epoch": 6.519032955513333, + "grad_norm": 0.4103158414363861, + "learning_rate": 6.8621160910720125e-09, + "loss": 0.0363, + "step": 99500 + }, + { + "epoch": 6.519688134704841, + "grad_norm": 0.5410376191139221, + "learning_rate": 6.59091917928345e-09, + "loss": 0.0226, + "step": 99510 + }, + { + "epoch": 6.520343313896351, + "grad_norm": 0.5081993937492371, + "learning_rate": 6.32518947155103e-09, + "loss": 0.0228, + "step": 99520 + }, + { + "epoch": 6.520998493087859, + "grad_norm": 0.8015349507331848, + "learning_rate": 6.064926996929288e-09, + "loss": 0.0404, + "step": 99530 + }, + { + "epoch": 6.521653672279369, + "grad_norm": 0.3602796792984009, + "learning_rate": 5.810131783884343e-09, + "loss": 0.0214, + "step": 99540 + }, + { + "epoch": 6.522308851470878, + "grad_norm": 0.4212692677974701, + "learning_rate": 5.5608038602772415e-09, + "loss": 0.0171, + "step": 99550 + }, + { + "epoch": 6.522964030662386, + "grad_norm": 0.09020109474658966, + "learning_rate": 5.316943253375062e-09, + "loss": 0.0361, + "step": 99560 + }, + { + "epoch": 6.523619209853895, + "grad_norm": 0.7865017652511597, + "learning_rate": 5.07854998984536e-09, + "loss": 0.033, + "step": 99570 + }, + { + "epoch": 6.524274389045404, + "grad_norm": 0.9018840789794922, + "learning_rate": 4.845624095756174e-09, + "loss": 0.0258, + "step": 99580 + }, + { + "epoch": 6.524929568236913, + "grad_norm": 0.4175656735897064, + "learning_rate": 4.618165596587121e-09, + "loss": 0.0506, + "step": 99590 + }, + { + "epoch": 6.525584747428422, + "grad_norm": 0.6200273633003235, + "learning_rate": 4.396174517207197e-09, + "loss": 0.0334, + "step": 99600 + }, + { + "epoch": 6.5262399266199305, + "grad_norm": 0.7959036231040955, + "learning_rate": 4.179650881896979e-09, + "loss": 0.0281, + "step": 99610 + }, + { + "epoch": 6.526895105811439, + "grad_norm": 0.21271374821662903, + "learning_rate": 3.968594714331975e-09, + "loss": 0.0206, + "step": 99620 + }, + { + "epoch": 6.527550285002948, + "grad_norm": 0.8227105140686035, + "learning_rate": 3.763006037593719e-09, + "loss": 0.0339, + "step": 99630 + }, + { + "epoch": 6.528205464194457, + "grad_norm": 0.4332076609134674, + "learning_rate": 3.5628848741586786e-09, + "loss": 0.028, + "step": 99640 + }, + { + "epoch": 6.528860643385966, + "grad_norm": 0.4118333160877228, + "learning_rate": 3.368231245926001e-09, + "loss": 0.0343, + "step": 99650 + }, + { + "epoch": 6.5295158225774745, + "grad_norm": 0.14641620218753815, + "learning_rate": 3.179045174167561e-09, + "loss": 0.0293, + "step": 99660 + }, + { + "epoch": 6.530171001768984, + "grad_norm": 0.09711211919784546, + "learning_rate": 2.9953266795834657e-09, + "loss": 0.0187, + "step": 99670 + }, + { + "epoch": 6.530826180960493, + "grad_norm": 0.06425721198320389, + "learning_rate": 2.817075782263201e-09, + "loss": 0.0177, + "step": 99680 + }, + { + "epoch": 6.531481360152002, + "grad_norm": 0.09364182502031326, + "learning_rate": 2.6442925016967325e-09, + "loss": 0.0227, + "step": 99690 + }, + { + "epoch": 6.532136539343511, + "grad_norm": 0.16567817330360413, + "learning_rate": 2.4769768567800555e-09, + "loss": 0.0302, + "step": 99700 + }, + { + "epoch": 6.5327917185350195, + "grad_norm": 0.30951228737831116, + "learning_rate": 2.315128865809646e-09, + "loss": 0.0229, + "step": 99710 + }, + { + "epoch": 6.533446897726528, + "grad_norm": 0.3630356788635254, + "learning_rate": 2.1587485464880097e-09, + "loss": 0.0227, + "step": 99720 + }, + { + "epoch": 6.534102076918037, + "grad_norm": 0.10984452068805695, + "learning_rate": 2.0078359159125813e-09, + "loss": 0.0232, + "step": 99730 + }, + { + "epoch": 6.534757256109546, + "grad_norm": 0.0895906612277031, + "learning_rate": 1.8623909905923776e-09, + "loss": 0.0192, + "step": 99740 + }, + { + "epoch": 6.535412435301055, + "grad_norm": 0.19151043891906738, + "learning_rate": 1.7224137864257916e-09, + "loss": 0.0234, + "step": 99750 + }, + { + "epoch": 6.536067614492564, + "grad_norm": 0.2200065553188324, + "learning_rate": 1.5879043187283505e-09, + "loss": 0.0357, + "step": 99760 + }, + { + "epoch": 6.536722793684072, + "grad_norm": 0.5262331366539001, + "learning_rate": 1.458862602204958e-09, + "loss": 0.0219, + "step": 99770 + }, + { + "epoch": 6.537377972875581, + "grad_norm": 0.2227221578359604, + "learning_rate": 1.3352886509720996e-09, + "loss": 0.0245, + "step": 99780 + }, + { + "epoch": 6.53803315206709, + "grad_norm": 0.7358943819999695, + "learning_rate": 1.2171824785356389e-09, + "loss": 0.0563, + "step": 99790 + }, + { + "epoch": 6.5386883312586, + "grad_norm": 0.23304486274719238, + "learning_rate": 1.1045440978185718e-09, + "loss": 0.0293, + "step": 99800 + }, + { + "epoch": 6.539343510450108, + "grad_norm": 0.1932375431060791, + "learning_rate": 9.973735211332714e-10, + "loss": 0.0333, + "step": 99810 + }, + { + "epoch": 6.539998689641617, + "grad_norm": 0.2550124526023865, + "learning_rate": 8.956707602036929e-10, + "loss": 0.0283, + "step": 99820 + }, + { + "epoch": 6.540653868833126, + "grad_norm": 0.3215198218822479, + "learning_rate": 7.994358261542712e-10, + "loss": 0.0284, + "step": 99830 + }, + { + "epoch": 6.541309048024635, + "grad_norm": 0.42553797364234924, + "learning_rate": 7.086687295043692e-10, + "loss": 0.0377, + "step": 99840 + }, + { + "epoch": 6.541964227216144, + "grad_norm": 0.18670785427093506, + "learning_rate": 6.23369480179381e-10, + "loss": 0.0261, + "step": 99850 + }, + { + "epoch": 6.542619406407653, + "grad_norm": 0.3316664397716522, + "learning_rate": 5.43538087510731e-10, + "loss": 0.0313, + "step": 99860 + }, + { + "epoch": 6.543274585599161, + "grad_norm": 0.34571054577827454, + "learning_rate": 4.691745602303233e-10, + "loss": 0.0474, + "step": 99870 + }, + { + "epoch": 6.54392976479067, + "grad_norm": 0.2959955930709839, + "learning_rate": 4.0027890646499034e-10, + "loss": 0.0235, + "step": 99880 + }, + { + "epoch": 6.544584943982179, + "grad_norm": 0.22268769145011902, + "learning_rate": 3.368511337531466e-10, + "loss": 0.0243, + "step": 99890 + }, + { + "epoch": 6.545240123173688, + "grad_norm": 1.2102831602096558, + "learning_rate": 2.788912490281348e-10, + "loss": 0.0309, + "step": 99900 + }, + { + "epoch": 6.545895302365197, + "grad_norm": 0.1376136988401413, + "learning_rate": 2.2639925863487954e-10, + "loss": 0.0186, + "step": 99910 + }, + { + "epoch": 6.5465504815567055, + "grad_norm": 0.5734387636184692, + "learning_rate": 1.7937516830213163e-10, + "loss": 0.0421, + "step": 99920 + }, + { + "epoch": 6.547205660748214, + "grad_norm": 0.11110133677721024, + "learning_rate": 1.3781898318687702e-10, + "loss": 0.0243, + "step": 99930 + }, + { + "epoch": 6.547860839939723, + "grad_norm": 1.146498203277588, + "learning_rate": 1.0173070781882566e-10, + "loss": 0.0306, + "step": 99940 + }, + { + "epoch": 6.548516019131233, + "grad_norm": 0.7719282507896423, + "learning_rate": 7.111034615592261e-11, + "loss": 0.0972, + "step": 99950 + }, + { + "epoch": 6.549171198322742, + "grad_norm": 0.13613876700401306, + "learning_rate": 4.595790153993918e-11, + "loss": 0.0154, + "step": 99960 + }, + { + "epoch": 6.5498263775142505, + "grad_norm": 0.6258400678634644, + "learning_rate": 2.6273376729779585e-11, + "loss": 0.0335, + "step": 99970 + }, + { + "epoch": 6.550481556705759, + "grad_norm": 0.4196563959121704, + "learning_rate": 1.2056773868174276e-11, + "loss": 0.0193, + "step": 99980 + }, + { + "epoch": 6.551136735897268, + "grad_norm": 0.6382051706314087, + "learning_rate": 3.3080945149865926e-12, + "loss": 0.0424, + "step": 99990 + }, + { + "epoch": 6.551791915088777, + "grad_norm": 0.63372802734375, + "learning_rate": 2.7339625008337977e-14, + "loss": 0.0358, + "step": 100000 + } + ], + "logging_steps": 10, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}