{ "best_global_step": 6000, "best_metric": 0.5057546679788949, "best_model_checkpoint": "./SALAMA_NEW8/checkpoint-6000", "epoch": 3.0, "eval_steps": 2000, "global_step": 7512, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003994407829039345, "grad_norm": 3.46816349029541, "learning_rate": 1.8e-07, "loss": 0.0307, "step": 10 }, { "epoch": 0.00798881565807869, "grad_norm": 3.722654104232788, "learning_rate": 3.8e-07, "loss": 0.0243, "step": 20 }, { "epoch": 0.011983223487118035, "grad_norm": 2.703404664993286, "learning_rate": 5.800000000000001e-07, "loss": 0.0376, "step": 30 }, { "epoch": 0.01597763131615738, "grad_norm": 2.4501798152923584, "learning_rate": 7.8e-07, "loss": 0.0264, "step": 40 }, { "epoch": 0.019972039145196723, "grad_norm": 2.6027143001556396, "learning_rate": 9.800000000000001e-07, "loss": 0.0369, "step": 50 }, { "epoch": 0.02396644697423607, "grad_norm": 2.276334047317505, "learning_rate": 1.1800000000000001e-06, "loss": 0.0192, "step": 60 }, { "epoch": 0.027960854803275415, "grad_norm": 1.9809595346450806, "learning_rate": 1.3800000000000001e-06, "loss": 0.0227, "step": 70 }, { "epoch": 0.03195526263231476, "grad_norm": 3.635103702545166, "learning_rate": 1.5800000000000001e-06, "loss": 0.022, "step": 80 }, { "epoch": 0.035949670461354104, "grad_norm": 2.506479263305664, "learning_rate": 1.7800000000000001e-06, "loss": 0.0241, "step": 90 }, { "epoch": 0.039944078290393446, "grad_norm": 1.891008734703064, "learning_rate": 1.98e-06, "loss": 0.0218, "step": 100 }, { "epoch": 0.043938486119432796, "grad_norm": 4.943603992462158, "learning_rate": 2.1800000000000003e-06, "loss": 0.0229, "step": 110 }, { "epoch": 0.04793289394847214, "grad_norm": 0.3047073781490326, "learning_rate": 2.38e-06, "loss": 0.0369, "step": 120 }, { "epoch": 0.05192730177751148, "grad_norm": 2.3492813110351562, "learning_rate": 2.5800000000000003e-06, "loss": 0.0251, "step": 130 }, { "epoch": 0.05592170960655083, "grad_norm": 1.2943974733352661, "learning_rate": 2.7800000000000005e-06, "loss": 0.0216, "step": 140 }, { "epoch": 0.05991611743559017, "grad_norm": 3.320284843444824, "learning_rate": 2.9800000000000003e-06, "loss": 0.0248, "step": 150 }, { "epoch": 0.06391052526462952, "grad_norm": 3.313554048538208, "learning_rate": 3.1800000000000005e-06, "loss": 0.035, "step": 160 }, { "epoch": 0.06790493309366886, "grad_norm": 2.999142646789551, "learning_rate": 3.3800000000000007e-06, "loss": 0.02, "step": 170 }, { "epoch": 0.07189934092270821, "grad_norm": 2.4451065063476562, "learning_rate": 3.58e-06, "loss": 0.0183, "step": 180 }, { "epoch": 0.07589374875174755, "grad_norm": 2.6344587802886963, "learning_rate": 3.7800000000000002e-06, "loss": 0.0232, "step": 190 }, { "epoch": 0.07988815658078689, "grad_norm": 1.9555765390396118, "learning_rate": 3.980000000000001e-06, "loss": 0.019, "step": 200 }, { "epoch": 0.08388256440982625, "grad_norm": 1.7121124267578125, "learning_rate": 4.18e-06, "loss": 0.0194, "step": 210 }, { "epoch": 0.08787697223886559, "grad_norm": 3.0031440258026123, "learning_rate": 4.38e-06, "loss": 0.0385, "step": 220 }, { "epoch": 0.09187138006790493, "grad_norm": 2.69048810005188, "learning_rate": 4.58e-06, "loss": 0.0283, "step": 230 }, { "epoch": 0.09586578789694428, "grad_norm": 1.9045894145965576, "learning_rate": 4.78e-06, "loss": 0.04, "step": 240 }, { "epoch": 0.09986019572598362, "grad_norm": 2.2926864624023438, "learning_rate": 4.980000000000001e-06, "loss": 0.0303, "step": 250 }, { "epoch": 0.10385460355502296, "grad_norm": 1.7242863178253174, "learning_rate": 5.18e-06, "loss": 0.0203, "step": 260 }, { "epoch": 0.10784901138406232, "grad_norm": 1.9205068349838257, "learning_rate": 5.380000000000001e-06, "loss": 0.0332, "step": 270 }, { "epoch": 0.11184341921310166, "grad_norm": 2.80607008934021, "learning_rate": 5.580000000000001e-06, "loss": 0.0281, "step": 280 }, { "epoch": 0.115837827042141, "grad_norm": 3.7204532623291016, "learning_rate": 5.78e-06, "loss": 0.0187, "step": 290 }, { "epoch": 0.11983223487118035, "grad_norm": 2.965312957763672, "learning_rate": 5.98e-06, "loss": 0.0329, "step": 300 }, { "epoch": 0.12382664270021969, "grad_norm": 4.226632118225098, "learning_rate": 6.18e-06, "loss": 0.0211, "step": 310 }, { "epoch": 0.12782105052925904, "grad_norm": 1.7873696088790894, "learning_rate": 6.380000000000001e-06, "loss": 0.0183, "step": 320 }, { "epoch": 0.1318154583582984, "grad_norm": 2.310504674911499, "learning_rate": 6.5800000000000005e-06, "loss": 0.0199, "step": 330 }, { "epoch": 0.13580986618733773, "grad_norm": 2.269408702850342, "learning_rate": 6.780000000000001e-06, "loss": 0.0258, "step": 340 }, { "epoch": 0.13980427401637707, "grad_norm": 2.605928421020508, "learning_rate": 6.98e-06, "loss": 0.0273, "step": 350 }, { "epoch": 0.14379868184541642, "grad_norm": 2.4324028491973877, "learning_rate": 7.180000000000001e-06, "loss": 0.0183, "step": 360 }, { "epoch": 0.14779308967445576, "grad_norm": 1.8827005624771118, "learning_rate": 7.3800000000000005e-06, "loss": 0.0231, "step": 370 }, { "epoch": 0.1517874975034951, "grad_norm": 1.3971014022827148, "learning_rate": 7.58e-06, "loss": 0.0232, "step": 380 }, { "epoch": 0.15578190533253444, "grad_norm": 2.3092613220214844, "learning_rate": 7.78e-06, "loss": 0.0268, "step": 390 }, { "epoch": 0.15977631316157379, "grad_norm": 2.469114065170288, "learning_rate": 7.980000000000002e-06, "loss": 0.0207, "step": 400 }, { "epoch": 0.16377072099061313, "grad_norm": 3.1286327838897705, "learning_rate": 8.18e-06, "loss": 0.031, "step": 410 }, { "epoch": 0.1677651288196525, "grad_norm": 3.00813627243042, "learning_rate": 8.380000000000001e-06, "loss": 0.0285, "step": 420 }, { "epoch": 0.17175953664869184, "grad_norm": 4.581786155700684, "learning_rate": 8.580000000000001e-06, "loss": 0.0434, "step": 430 }, { "epoch": 0.17575394447773118, "grad_norm": 1.9707911014556885, "learning_rate": 8.78e-06, "loss": 0.0294, "step": 440 }, { "epoch": 0.17974835230677053, "grad_norm": 2.7084460258483887, "learning_rate": 8.98e-06, "loss": 0.0287, "step": 450 }, { "epoch": 0.18374276013580987, "grad_norm": 5.108406066894531, "learning_rate": 9.180000000000002e-06, "loss": 0.037, "step": 460 }, { "epoch": 0.1877371679648492, "grad_norm": 1.6236001253128052, "learning_rate": 9.38e-06, "loss": 0.0192, "step": 470 }, { "epoch": 0.19173157579388855, "grad_norm": 3.1745877265930176, "learning_rate": 9.58e-06, "loss": 0.0287, "step": 480 }, { "epoch": 0.1957259836229279, "grad_norm": 3.5089573860168457, "learning_rate": 9.780000000000001e-06, "loss": 0.0226, "step": 490 }, { "epoch": 0.19972039145196724, "grad_norm": 1.9109275341033936, "learning_rate": 9.980000000000001e-06, "loss": 0.0365, "step": 500 }, { "epoch": 0.20371479928100658, "grad_norm": 3.1366279125213623, "learning_rate": 9.98716486023959e-06, "loss": 0.0333, "step": 510 }, { "epoch": 0.20770920711004592, "grad_norm": 2.936997175216675, "learning_rate": 9.972903593839133e-06, "loss": 0.0265, "step": 520 }, { "epoch": 0.2117036149390853, "grad_norm": 2.2245354652404785, "learning_rate": 9.958642327438678e-06, "loss": 0.0246, "step": 530 }, { "epoch": 0.21569802276812464, "grad_norm": 2.7430694103240967, "learning_rate": 9.944381061038221e-06, "loss": 0.0302, "step": 540 }, { "epoch": 0.21969243059716398, "grad_norm": 2.499619483947754, "learning_rate": 9.930119794637765e-06, "loss": 0.0407, "step": 550 }, { "epoch": 0.22368683842620332, "grad_norm": 3.148715019226074, "learning_rate": 9.91585852823731e-06, "loss": 0.0254, "step": 560 }, { "epoch": 0.22768124625524266, "grad_norm": 2.967156171798706, "learning_rate": 9.901597261836851e-06, "loss": 0.0274, "step": 570 }, { "epoch": 0.231675654084282, "grad_norm": 2.920797824859619, "learning_rate": 9.887335995436396e-06, "loss": 0.0339, "step": 580 }, { "epoch": 0.23567006191332135, "grad_norm": 2.7116758823394775, "learning_rate": 9.87307472903594e-06, "loss": 0.0339, "step": 590 }, { "epoch": 0.2396644697423607, "grad_norm": 2.9919016361236572, "learning_rate": 9.858813462635483e-06, "loss": 0.034, "step": 600 }, { "epoch": 0.24365887757140003, "grad_norm": 2.4491732120513916, "learning_rate": 9.844552196235026e-06, "loss": 0.0313, "step": 610 }, { "epoch": 0.24765328540043938, "grad_norm": 2.839397668838501, "learning_rate": 9.83029092983457e-06, "loss": 0.0336, "step": 620 }, { "epoch": 0.2516476932294787, "grad_norm": 3.5191802978515625, "learning_rate": 9.816029663434114e-06, "loss": 0.0326, "step": 630 }, { "epoch": 0.2556421010585181, "grad_norm": 3.609706163406372, "learning_rate": 9.801768397033657e-06, "loss": 0.0282, "step": 640 }, { "epoch": 0.2596365088875574, "grad_norm": 2.35837459564209, "learning_rate": 9.787507130633202e-06, "loss": 0.0247, "step": 650 }, { "epoch": 0.2636309167165968, "grad_norm": 2.5145423412323, "learning_rate": 9.773245864232744e-06, "loss": 0.0298, "step": 660 }, { "epoch": 0.2676253245456361, "grad_norm": 4.194463729858398, "learning_rate": 9.758984597832289e-06, "loss": 0.0336, "step": 670 }, { "epoch": 0.27161973237467546, "grad_norm": 3.227799892425537, "learning_rate": 9.744723331431832e-06, "loss": 0.028, "step": 680 }, { "epoch": 0.2756141402037148, "grad_norm": 2.315329074859619, "learning_rate": 9.730462065031375e-06, "loss": 0.0353, "step": 690 }, { "epoch": 0.27960854803275415, "grad_norm": 3.3833742141723633, "learning_rate": 9.71620079863092e-06, "loss": 0.0263, "step": 700 }, { "epoch": 0.2836029558617935, "grad_norm": 2.724463701248169, "learning_rate": 9.701939532230463e-06, "loss": 0.0292, "step": 710 }, { "epoch": 0.28759736369083283, "grad_norm": 3.384976863861084, "learning_rate": 9.687678265830007e-06, "loss": 0.0355, "step": 720 }, { "epoch": 0.2915917715198722, "grad_norm": 2.455484390258789, "learning_rate": 9.67341699942955e-06, "loss": 0.0344, "step": 730 }, { "epoch": 0.2955861793489115, "grad_norm": 2.9067740440368652, "learning_rate": 9.659155733029095e-06, "loss": 0.0372, "step": 740 }, { "epoch": 0.2995805871779509, "grad_norm": 3.116934061050415, "learning_rate": 9.644894466628636e-06, "loss": 0.0291, "step": 750 }, { "epoch": 0.3035749950069902, "grad_norm": 3.9219629764556885, "learning_rate": 9.630633200228181e-06, "loss": 0.0296, "step": 760 }, { "epoch": 0.30756940283602957, "grad_norm": 2.3769092559814453, "learning_rate": 9.616371933827725e-06, "loss": 0.0364, "step": 770 }, { "epoch": 0.3115638106650689, "grad_norm": 2.7386128902435303, "learning_rate": 9.602110667427268e-06, "loss": 0.0368, "step": 780 }, { "epoch": 0.31555821849410826, "grad_norm": 2.2549479007720947, "learning_rate": 9.587849401026813e-06, "loss": 0.042, "step": 790 }, { "epoch": 0.31955262632314757, "grad_norm": 2.8821749687194824, "learning_rate": 9.573588134626356e-06, "loss": 0.0291, "step": 800 }, { "epoch": 0.32354703415218694, "grad_norm": 2.2567620277404785, "learning_rate": 9.5593268682259e-06, "loss": 0.0287, "step": 810 }, { "epoch": 0.32754144198122626, "grad_norm": 2.1578598022460938, "learning_rate": 9.545065601825442e-06, "loss": 0.0387, "step": 820 }, { "epoch": 0.3315358498102656, "grad_norm": 3.6625397205352783, "learning_rate": 9.530804335424987e-06, "loss": 0.0288, "step": 830 }, { "epoch": 0.335530257639305, "grad_norm": 2.7808022499084473, "learning_rate": 9.516543069024529e-06, "loss": 0.0376, "step": 840 }, { "epoch": 0.3395246654683443, "grad_norm": 3.6855411529541016, "learning_rate": 9.502281802624074e-06, "loss": 0.0372, "step": 850 }, { "epoch": 0.3435190732973837, "grad_norm": 2.3804266452789307, "learning_rate": 9.488020536223617e-06, "loss": 0.0262, "step": 860 }, { "epoch": 0.347513481126423, "grad_norm": 2.1207637786865234, "learning_rate": 9.47375926982316e-06, "loss": 0.0249, "step": 870 }, { "epoch": 0.35150788895546237, "grad_norm": 2.7773971557617188, "learning_rate": 9.459498003422705e-06, "loss": 0.0329, "step": 880 }, { "epoch": 0.3555022967845017, "grad_norm": 1.848315954208374, "learning_rate": 9.445236737022249e-06, "loss": 0.0334, "step": 890 }, { "epoch": 0.35949670461354105, "grad_norm": 2.8464772701263428, "learning_rate": 9.430975470621792e-06, "loss": 0.0276, "step": 900 }, { "epoch": 0.36349111244258037, "grad_norm": 2.721813678741455, "learning_rate": 9.416714204221335e-06, "loss": 0.0302, "step": 910 }, { "epoch": 0.36748552027161974, "grad_norm": 2.085888147354126, "learning_rate": 9.40245293782088e-06, "loss": 0.0271, "step": 920 }, { "epoch": 0.37147992810065905, "grad_norm": 3.2898266315460205, "learning_rate": 9.388191671420423e-06, "loss": 0.0393, "step": 930 }, { "epoch": 0.3754743359296984, "grad_norm": 2.3367676734924316, "learning_rate": 9.373930405019966e-06, "loss": 0.0327, "step": 940 }, { "epoch": 0.3794687437587378, "grad_norm": 2.8437299728393555, "learning_rate": 9.35966913861951e-06, "loss": 0.0242, "step": 950 }, { "epoch": 0.3834631515877771, "grad_norm": 3.629429817199707, "learning_rate": 9.345407872219053e-06, "loss": 0.0317, "step": 960 }, { "epoch": 0.3874575594168165, "grad_norm": 1.7099779844284058, "learning_rate": 9.331146605818598e-06, "loss": 0.0366, "step": 970 }, { "epoch": 0.3914519672458558, "grad_norm": 3.043698310852051, "learning_rate": 9.316885339418141e-06, "loss": 0.0351, "step": 980 }, { "epoch": 0.39544637507489516, "grad_norm": 4.258053302764893, "learning_rate": 9.302624073017684e-06, "loss": 0.0381, "step": 990 }, { "epoch": 0.3994407829039345, "grad_norm": 1.7449215650558472, "learning_rate": 9.288362806617228e-06, "loss": 0.0434, "step": 1000 }, { "epoch": 0.40343519073297385, "grad_norm": 2.3904478549957275, "learning_rate": 9.274101540216773e-06, "loss": 0.0356, "step": 1010 }, { "epoch": 0.40742959856201316, "grad_norm": 4.740054130554199, "learning_rate": 9.259840273816316e-06, "loss": 0.0353, "step": 1020 }, { "epoch": 0.41142400639105253, "grad_norm": 3.9382550716400146, "learning_rate": 9.245579007415859e-06, "loss": 0.0452, "step": 1030 }, { "epoch": 0.41541841422009185, "grad_norm": 3.4149060249328613, "learning_rate": 9.231317741015402e-06, "loss": 0.0312, "step": 1040 }, { "epoch": 0.4194128220491312, "grad_norm": 1.7139674425125122, "learning_rate": 9.217056474614946e-06, "loss": 0.0277, "step": 1050 }, { "epoch": 0.4234072298781706, "grad_norm": 2.9577815532684326, "learning_rate": 9.20279520821449e-06, "loss": 0.0266, "step": 1060 }, { "epoch": 0.4274016377072099, "grad_norm": 2.4664392471313477, "learning_rate": 9.188533941814034e-06, "loss": 0.0237, "step": 1070 }, { "epoch": 0.4313960455362493, "grad_norm": 2.966825246810913, "learning_rate": 9.174272675413579e-06, "loss": 0.0193, "step": 1080 }, { "epoch": 0.4353904533652886, "grad_norm": 2.5782456398010254, "learning_rate": 9.16001140901312e-06, "loss": 0.0334, "step": 1090 }, { "epoch": 0.43938486119432796, "grad_norm": 3.482473134994507, "learning_rate": 9.145750142612665e-06, "loss": 0.0458, "step": 1100 }, { "epoch": 0.4433792690233673, "grad_norm": 2.6434431076049805, "learning_rate": 9.131488876212208e-06, "loss": 0.0385, "step": 1110 }, { "epoch": 0.44737367685240664, "grad_norm": 3.8204150199890137, "learning_rate": 9.117227609811752e-06, "loss": 0.0275, "step": 1120 }, { "epoch": 0.45136808468144596, "grad_norm": 2.0705342292785645, "learning_rate": 9.102966343411297e-06, "loss": 0.0332, "step": 1130 }, { "epoch": 0.45536249251048533, "grad_norm": 4.166182994842529, "learning_rate": 9.08870507701084e-06, "loss": 0.0301, "step": 1140 }, { "epoch": 0.45935690033952464, "grad_norm": 3.771206855773926, "learning_rate": 9.074443810610383e-06, "loss": 0.0285, "step": 1150 }, { "epoch": 0.463351308168564, "grad_norm": 3.3919098377227783, "learning_rate": 9.060182544209926e-06, "loss": 0.0333, "step": 1160 }, { "epoch": 0.46734571599760333, "grad_norm": 3.0169448852539062, "learning_rate": 9.045921277809471e-06, "loss": 0.0376, "step": 1170 }, { "epoch": 0.4713401238266427, "grad_norm": 3.0113446712493896, "learning_rate": 9.031660011409013e-06, "loss": 0.0367, "step": 1180 }, { "epoch": 0.47533453165568207, "grad_norm": 3.8403584957122803, "learning_rate": 9.017398745008558e-06, "loss": 0.0403, "step": 1190 }, { "epoch": 0.4793289394847214, "grad_norm": 1.8323266506195068, "learning_rate": 9.003137478608101e-06, "loss": 0.0254, "step": 1200 }, { "epoch": 0.48332334731376075, "grad_norm": 3.447676420211792, "learning_rate": 8.988876212207644e-06, "loss": 0.0309, "step": 1210 }, { "epoch": 0.48731775514280007, "grad_norm": 3.089911937713623, "learning_rate": 8.97461494580719e-06, "loss": 0.0368, "step": 1220 }, { "epoch": 0.49131216297183944, "grad_norm": 2.221872091293335, "learning_rate": 8.960353679406733e-06, "loss": 0.0225, "step": 1230 }, { "epoch": 0.49530657080087875, "grad_norm": 1.7740514278411865, "learning_rate": 8.946092413006276e-06, "loss": 0.03, "step": 1240 }, { "epoch": 0.4993009786299181, "grad_norm": 2.922903299331665, "learning_rate": 8.931831146605819e-06, "loss": 0.0234, "step": 1250 }, { "epoch": 0.5032953864589574, "grad_norm": 2.8122918605804443, "learning_rate": 8.917569880205364e-06, "loss": 0.0453, "step": 1260 }, { "epoch": 0.5072897942879968, "grad_norm": 2.1735727787017822, "learning_rate": 8.903308613804906e-06, "loss": 0.0295, "step": 1270 }, { "epoch": 0.5112842021170362, "grad_norm": 2.4211249351501465, "learning_rate": 8.88904734740445e-06, "loss": 0.0292, "step": 1280 }, { "epoch": 0.5152786099460755, "grad_norm": 5.627583026885986, "learning_rate": 8.874786081003994e-06, "loss": 0.0438, "step": 1290 }, { "epoch": 0.5192730177751148, "grad_norm": 2.746601104736328, "learning_rate": 8.860524814603537e-06, "loss": 0.0361, "step": 1300 }, { "epoch": 0.5232674256041542, "grad_norm": 3.283212661743164, "learning_rate": 8.846263548203082e-06, "loss": 0.0389, "step": 1310 }, { "epoch": 0.5272618334331935, "grad_norm": 2.8431429862976074, "learning_rate": 8.832002281802625e-06, "loss": 0.0344, "step": 1320 }, { "epoch": 0.5312562412622329, "grad_norm": 1.620734453201294, "learning_rate": 8.817741015402168e-06, "loss": 0.0281, "step": 1330 }, { "epoch": 0.5352506490912722, "grad_norm": 3.2256009578704834, "learning_rate": 8.803479749001712e-06, "loss": 0.0351, "step": 1340 }, { "epoch": 0.5392450569203115, "grad_norm": 3.3469367027282715, "learning_rate": 8.789218482601257e-06, "loss": 0.0334, "step": 1350 }, { "epoch": 0.5432394647493509, "grad_norm": 4.458150386810303, "learning_rate": 8.7749572162008e-06, "loss": 0.0226, "step": 1360 }, { "epoch": 0.5472338725783903, "grad_norm": 3.770509719848633, "learning_rate": 8.760695949800343e-06, "loss": 0.027, "step": 1370 }, { "epoch": 0.5512282804074295, "grad_norm": 1.770533561706543, "learning_rate": 8.746434683399886e-06, "loss": 0.0387, "step": 1380 }, { "epoch": 0.5552226882364689, "grad_norm": 3.1090400218963623, "learning_rate": 8.73217341699943e-06, "loss": 0.0335, "step": 1390 }, { "epoch": 0.5592170960655083, "grad_norm": 3.3928542137145996, "learning_rate": 8.717912150598975e-06, "loss": 0.0405, "step": 1400 }, { "epoch": 0.5632115038945477, "grad_norm": 2.3857524394989014, "learning_rate": 8.703650884198518e-06, "loss": 0.0356, "step": 1410 }, { "epoch": 0.567205911723587, "grad_norm": 2.3248329162597656, "learning_rate": 8.689389617798061e-06, "loss": 0.0366, "step": 1420 }, { "epoch": 0.5712003195526263, "grad_norm": 2.1319854259490967, "learning_rate": 8.675128351397604e-06, "loss": 0.036, "step": 1430 }, { "epoch": 0.5751947273816657, "grad_norm": 1.5602308511734009, "learning_rate": 8.66086708499715e-06, "loss": 0.0286, "step": 1440 }, { "epoch": 0.579189135210705, "grad_norm": 2.8112518787384033, "learning_rate": 8.646605818596692e-06, "loss": 0.0409, "step": 1450 }, { "epoch": 0.5831835430397444, "grad_norm": 3.130460739135742, "learning_rate": 8.632344552196236e-06, "loss": 0.0295, "step": 1460 }, { "epoch": 0.5871779508687837, "grad_norm": 4.073850154876709, "learning_rate": 8.618083285795779e-06, "loss": 0.0433, "step": 1470 }, { "epoch": 0.591172358697823, "grad_norm": 2.9691765308380127, "learning_rate": 8.603822019395322e-06, "loss": 0.0367, "step": 1480 }, { "epoch": 0.5951667665268624, "grad_norm": 2.8310015201568604, "learning_rate": 8.589560752994867e-06, "loss": 0.0286, "step": 1490 }, { "epoch": 0.5991611743559018, "grad_norm": 4.259763717651367, "learning_rate": 8.57529948659441e-06, "loss": 0.0374, "step": 1500 }, { "epoch": 0.603155582184941, "grad_norm": 4.179434299468994, "learning_rate": 8.561038220193954e-06, "loss": 0.0363, "step": 1510 }, { "epoch": 0.6071499900139804, "grad_norm": 1.6247011423110962, "learning_rate": 8.546776953793497e-06, "loss": 0.0225, "step": 1520 }, { "epoch": 0.6111443978430198, "grad_norm": 2.520911693572998, "learning_rate": 8.532515687393042e-06, "loss": 0.0331, "step": 1530 }, { "epoch": 0.6151388056720591, "grad_norm": 2.6778786182403564, "learning_rate": 8.518254420992585e-06, "loss": 0.0335, "step": 1540 }, { "epoch": 0.6191332135010985, "grad_norm": 2.9722695350646973, "learning_rate": 8.503993154592128e-06, "loss": 0.0419, "step": 1550 }, { "epoch": 0.6231276213301378, "grad_norm": 2.7999324798583984, "learning_rate": 8.489731888191672e-06, "loss": 0.0323, "step": 1560 }, { "epoch": 0.6271220291591771, "grad_norm": 2.474074363708496, "learning_rate": 8.475470621791215e-06, "loss": 0.0371, "step": 1570 }, { "epoch": 0.6311164369882165, "grad_norm": 2.244399309158325, "learning_rate": 8.46120935539076e-06, "loss": 0.0247, "step": 1580 }, { "epoch": 0.6351108448172559, "grad_norm": 2.866180896759033, "learning_rate": 8.446948088990303e-06, "loss": 0.039, "step": 1590 }, { "epoch": 0.6391052526462951, "grad_norm": 2.729217529296875, "learning_rate": 8.432686822589846e-06, "loss": 0.0378, "step": 1600 }, { "epoch": 0.6430996604753345, "grad_norm": 2.707044839859009, "learning_rate": 8.41842555618939e-06, "loss": 0.0246, "step": 1610 }, { "epoch": 0.6470940683043739, "grad_norm": 4.565586566925049, "learning_rate": 8.404164289788934e-06, "loss": 0.0452, "step": 1620 }, { "epoch": 0.6510884761334133, "grad_norm": 2.372103214263916, "learning_rate": 8.389903023388478e-06, "loss": 0.0343, "step": 1630 }, { "epoch": 0.6550828839624525, "grad_norm": 3.3515093326568604, "learning_rate": 8.375641756988021e-06, "loss": 0.0341, "step": 1640 }, { "epoch": 0.6590772917914919, "grad_norm": 2.545055866241455, "learning_rate": 8.361380490587566e-06, "loss": 0.028, "step": 1650 }, { "epoch": 0.6630716996205313, "grad_norm": 3.0382556915283203, "learning_rate": 8.347119224187107e-06, "loss": 0.0283, "step": 1660 }, { "epoch": 0.6670661074495706, "grad_norm": 3.1847259998321533, "learning_rate": 8.332857957786652e-06, "loss": 0.0304, "step": 1670 }, { "epoch": 0.67106051527861, "grad_norm": 4.07405424118042, "learning_rate": 8.318596691386196e-06, "loss": 0.0321, "step": 1680 }, { "epoch": 0.6750549231076493, "grad_norm": 2.5733461380004883, "learning_rate": 8.304335424985739e-06, "loss": 0.0335, "step": 1690 }, { "epoch": 0.6790493309366886, "grad_norm": 4.693726062774658, "learning_rate": 8.290074158585282e-06, "loss": 0.0345, "step": 1700 }, { "epoch": 0.683043738765728, "grad_norm": 3.111382007598877, "learning_rate": 8.275812892184827e-06, "loss": 0.0423, "step": 1710 }, { "epoch": 0.6870381465947674, "grad_norm": 3.0038416385650635, "learning_rate": 8.26155162578437e-06, "loss": 0.0272, "step": 1720 }, { "epoch": 0.6910325544238066, "grad_norm": 2.7561848163604736, "learning_rate": 8.247290359383914e-06, "loss": 0.0325, "step": 1730 }, { "epoch": 0.695026962252846, "grad_norm": 2.5607120990753174, "learning_rate": 8.233029092983458e-06, "loss": 0.027, "step": 1740 }, { "epoch": 0.6990213700818854, "grad_norm": 1.522266149520874, "learning_rate": 8.218767826583002e-06, "loss": 0.0391, "step": 1750 }, { "epoch": 0.7030157779109247, "grad_norm": 3.8156557083129883, "learning_rate": 8.204506560182545e-06, "loss": 0.0354, "step": 1760 }, { "epoch": 0.7070101857399641, "grad_norm": 2.6884236335754395, "learning_rate": 8.190245293782088e-06, "loss": 0.037, "step": 1770 }, { "epoch": 0.7110045935690034, "grad_norm": 2.2173361778259277, "learning_rate": 8.175984027381633e-06, "loss": 0.0394, "step": 1780 }, { "epoch": 0.7149990013980427, "grad_norm": 1.7859878540039062, "learning_rate": 8.161722760981175e-06, "loss": 0.0295, "step": 1790 }, { "epoch": 0.7189934092270821, "grad_norm": 3.347618341445923, "learning_rate": 8.14746149458072e-06, "loss": 0.0332, "step": 1800 }, { "epoch": 0.7229878170561215, "grad_norm": 3.5312180519104004, "learning_rate": 8.133200228180263e-06, "loss": 0.0331, "step": 1810 }, { "epoch": 0.7269822248851607, "grad_norm": 3.1601924896240234, "learning_rate": 8.118938961779806e-06, "loss": 0.0273, "step": 1820 }, { "epoch": 0.7309766327142001, "grad_norm": 2.3679819107055664, "learning_rate": 8.104677695379351e-06, "loss": 0.0306, "step": 1830 }, { "epoch": 0.7349710405432395, "grad_norm": 3.013176441192627, "learning_rate": 8.090416428978894e-06, "loss": 0.0477, "step": 1840 }, { "epoch": 0.7389654483722788, "grad_norm": 2.3330905437469482, "learning_rate": 8.076155162578438e-06, "loss": 0.0295, "step": 1850 }, { "epoch": 0.7429598562013181, "grad_norm": 3.5097014904022217, "learning_rate": 8.06189389617798e-06, "loss": 0.0367, "step": 1860 }, { "epoch": 0.7469542640303575, "grad_norm": 2.2575650215148926, "learning_rate": 8.047632629777526e-06, "loss": 0.0315, "step": 1870 }, { "epoch": 0.7509486718593968, "grad_norm": 2.637472152709961, "learning_rate": 8.033371363377069e-06, "loss": 0.0428, "step": 1880 }, { "epoch": 0.7549430796884362, "grad_norm": 2.459928274154663, "learning_rate": 8.019110096976612e-06, "loss": 0.025, "step": 1890 }, { "epoch": 0.7589374875174756, "grad_norm": 2.7636241912841797, "learning_rate": 8.004848830576156e-06, "loss": 0.0259, "step": 1900 }, { "epoch": 0.7629318953465148, "grad_norm": 3.4332215785980225, "learning_rate": 7.990587564175699e-06, "loss": 0.0287, "step": 1910 }, { "epoch": 0.7669263031755542, "grad_norm": 3.5681216716766357, "learning_rate": 7.976326297775244e-06, "loss": 0.0314, "step": 1920 }, { "epoch": 0.7709207110045936, "grad_norm": 4.455057144165039, "learning_rate": 7.962065031374787e-06, "loss": 0.0381, "step": 1930 }, { "epoch": 0.774915118833633, "grad_norm": 3.433385133743286, "learning_rate": 7.94780376497433e-06, "loss": 0.035, "step": 1940 }, { "epoch": 0.7789095266626722, "grad_norm": 2.730055809020996, "learning_rate": 7.933542498573873e-06, "loss": 0.0269, "step": 1950 }, { "epoch": 0.7829039344917116, "grad_norm": 2.8840601444244385, "learning_rate": 7.919281232173418e-06, "loss": 0.0399, "step": 1960 }, { "epoch": 0.786898342320751, "grad_norm": 3.8547487258911133, "learning_rate": 7.905019965772962e-06, "loss": 0.0322, "step": 1970 }, { "epoch": 0.7908927501497903, "grad_norm": 2.3967716693878174, "learning_rate": 7.890758699372505e-06, "loss": 0.0272, "step": 1980 }, { "epoch": 0.7948871579788296, "grad_norm": 3.3462061882019043, "learning_rate": 7.876497432972048e-06, "loss": 0.0284, "step": 1990 }, { "epoch": 0.798881565807869, "grad_norm": 2.8046774864196777, "learning_rate": 7.862236166571591e-06, "loss": 0.0358, "step": 2000 }, { "epoch": 0.798881565807869, "eval_loss": 0.01868388056755066, "eval_runtime": 7338.2585, "eval_samples_per_second": 2.729, "eval_steps_per_second": 0.341, "eval_wer": 1.722299680144345, "step": 2000 }, { "epoch": 0.8028759736369083, "grad_norm": 2.368256092071533, "learning_rate": 7.847974900171136e-06, "loss": 0.0278, "step": 2010 }, { "epoch": 0.8068703814659477, "grad_norm": 2.8456475734710693, "learning_rate": 7.83371363377068e-06, "loss": 0.0339, "step": 2020 }, { "epoch": 0.8108647892949871, "grad_norm": 1.9898338317871094, "learning_rate": 7.819452367370223e-06, "loss": 0.0289, "step": 2030 }, { "epoch": 0.8148591971240263, "grad_norm": 3.1095328330993652, "learning_rate": 7.805191100969766e-06, "loss": 0.0291, "step": 2040 }, { "epoch": 0.8188536049530657, "grad_norm": 2.8899197578430176, "learning_rate": 7.790929834569311e-06, "loss": 0.0375, "step": 2050 }, { "epoch": 0.8228480127821051, "grad_norm": 3.9213523864746094, "learning_rate": 7.776668568168854e-06, "loss": 0.0444, "step": 2060 }, { "epoch": 0.8268424206111444, "grad_norm": 2.5879340171813965, "learning_rate": 7.762407301768397e-06, "loss": 0.0282, "step": 2070 }, { "epoch": 0.8308368284401837, "grad_norm": 1.8168540000915527, "learning_rate": 7.748146035367942e-06, "loss": 0.025, "step": 2080 }, { "epoch": 0.8348312362692231, "grad_norm": 3.3461949825286865, "learning_rate": 7.733884768967484e-06, "loss": 0.0273, "step": 2090 }, { "epoch": 0.8388256440982624, "grad_norm": 2.239011764526367, "learning_rate": 7.719623502567029e-06, "loss": 0.0315, "step": 2100 }, { "epoch": 0.8428200519273018, "grad_norm": 1.676645278930664, "learning_rate": 7.705362236166572e-06, "loss": 0.036, "step": 2110 }, { "epoch": 0.8468144597563412, "grad_norm": 3.113101005554199, "learning_rate": 7.691100969766115e-06, "loss": 0.0332, "step": 2120 }, { "epoch": 0.8508088675853804, "grad_norm": 3.9707107543945312, "learning_rate": 7.676839703365659e-06, "loss": 0.0247, "step": 2130 }, { "epoch": 0.8548032754144198, "grad_norm": 2.901353120803833, "learning_rate": 7.662578436965204e-06, "loss": 0.0304, "step": 2140 }, { "epoch": 0.8587976832434592, "grad_norm": 2.0267584323883057, "learning_rate": 7.648317170564747e-06, "loss": 0.0256, "step": 2150 }, { "epoch": 0.8627920910724985, "grad_norm": 2.433629274368286, "learning_rate": 7.63405590416429e-06, "loss": 0.0267, "step": 2160 }, { "epoch": 0.8667864989015378, "grad_norm": 2.8645429611206055, "learning_rate": 7.619794637763834e-06, "loss": 0.0263, "step": 2170 }, { "epoch": 0.8707809067305772, "grad_norm": 2.485597610473633, "learning_rate": 7.6055333713633774e-06, "loss": 0.0285, "step": 2180 }, { "epoch": 0.8747753145596165, "grad_norm": 2.311027765274048, "learning_rate": 7.5912721049629215e-06, "loss": 0.0237, "step": 2190 }, { "epoch": 0.8787697223886559, "grad_norm": 2.102123260498047, "learning_rate": 7.577010838562466e-06, "loss": 0.0251, "step": 2200 }, { "epoch": 0.8827641302176952, "grad_norm": 1.7563374042510986, "learning_rate": 7.562749572162008e-06, "loss": 0.0331, "step": 2210 }, { "epoch": 0.8867585380467345, "grad_norm": 4.571224689483643, "learning_rate": 7.548488305761552e-06, "loss": 0.0411, "step": 2220 }, { "epoch": 0.8907529458757739, "grad_norm": 2.5844295024871826, "learning_rate": 7.534227039361096e-06, "loss": 0.0233, "step": 2230 }, { "epoch": 0.8947473537048133, "grad_norm": 2.9091975688934326, "learning_rate": 7.519965772960639e-06, "loss": 0.019, "step": 2240 }, { "epoch": 0.8987417615338527, "grad_norm": 3.633951425552368, "learning_rate": 7.505704506560183e-06, "loss": 0.0322, "step": 2250 }, { "epoch": 0.9027361693628919, "grad_norm": 2.073758602142334, "learning_rate": 7.491443240159727e-06, "loss": 0.0213, "step": 2260 }, { "epoch": 0.9067305771919313, "grad_norm": 1.943109154701233, "learning_rate": 7.47718197375927e-06, "loss": 0.0309, "step": 2270 }, { "epoch": 0.9107249850209707, "grad_norm": 2.816091775894165, "learning_rate": 7.462920707358814e-06, "loss": 0.0342, "step": 2280 }, { "epoch": 0.91471939285001, "grad_norm": 2.2993264198303223, "learning_rate": 7.448659440958358e-06, "loss": 0.0284, "step": 2290 }, { "epoch": 0.9187138006790493, "grad_norm": 2.5072145462036133, "learning_rate": 7.434398174557901e-06, "loss": 0.0261, "step": 2300 }, { "epoch": 0.9227082085080887, "grad_norm": 2.1535396575927734, "learning_rate": 7.420136908157445e-06, "loss": 0.0318, "step": 2310 }, { "epoch": 0.926702616337128, "grad_norm": 3.5091567039489746, "learning_rate": 7.405875641756989e-06, "loss": 0.0286, "step": 2320 }, { "epoch": 0.9306970241661674, "grad_norm": 3.785184144973755, "learning_rate": 7.391614375356533e-06, "loss": 0.0409, "step": 2330 }, { "epoch": 0.9346914319952067, "grad_norm": 3.4833099842071533, "learning_rate": 7.377353108956075e-06, "loss": 0.0334, "step": 2340 }, { "epoch": 0.938685839824246, "grad_norm": 2.384570598602295, "learning_rate": 7.3630918425556194e-06, "loss": 0.0311, "step": 2350 }, { "epoch": 0.9426802476532854, "grad_norm": 3.203986406326294, "learning_rate": 7.3488305761551635e-06, "loss": 0.0269, "step": 2360 }, { "epoch": 0.9466746554823248, "grad_norm": 2.0542373657226562, "learning_rate": 7.334569309754707e-06, "loss": 0.027, "step": 2370 }, { "epoch": 0.9506690633113641, "grad_norm": 4.261587142944336, "learning_rate": 7.320308043354251e-06, "loss": 0.0264, "step": 2380 }, { "epoch": 0.9546634711404034, "grad_norm": 3.3129560947418213, "learning_rate": 7.306046776953795e-06, "loss": 0.037, "step": 2390 }, { "epoch": 0.9586578789694428, "grad_norm": 2.2495546340942383, "learning_rate": 7.291785510553337e-06, "loss": 0.034, "step": 2400 }, { "epoch": 0.9626522867984821, "grad_norm": 3.4507882595062256, "learning_rate": 7.2775242441528815e-06, "loss": 0.0258, "step": 2410 }, { "epoch": 0.9666466946275215, "grad_norm": 3.949570417404175, "learning_rate": 7.2632629777524256e-06, "loss": 0.0315, "step": 2420 }, { "epoch": 0.9706411024565608, "grad_norm": 1.7982138395309448, "learning_rate": 7.249001711351969e-06, "loss": 0.0259, "step": 2430 }, { "epoch": 0.9746355102856001, "grad_norm": 2.3690760135650635, "learning_rate": 7.234740444951512e-06, "loss": 0.0351, "step": 2440 }, { "epoch": 0.9786299181146395, "grad_norm": 3.11702823638916, "learning_rate": 7.220479178551056e-06, "loss": 0.0258, "step": 2450 }, { "epoch": 0.9826243259436789, "grad_norm": 2.2427687644958496, "learning_rate": 7.206217912150599e-06, "loss": 0.0328, "step": 2460 }, { "epoch": 0.9866187337727182, "grad_norm": 1.585270881652832, "learning_rate": 7.1919566457501435e-06, "loss": 0.0337, "step": 2470 }, { "epoch": 0.9906131416017575, "grad_norm": 3.1885695457458496, "learning_rate": 7.1776953793496876e-06, "loss": 0.0299, "step": 2480 }, { "epoch": 0.9946075494307969, "grad_norm": 3.44587779045105, "learning_rate": 7.16343411294923e-06, "loss": 0.0293, "step": 2490 }, { "epoch": 0.9986019572598362, "grad_norm": 2.4740006923675537, "learning_rate": 7.149172846548774e-06, "loss": 0.0317, "step": 2500 }, { "epoch": 1.0023966446974235, "grad_norm": 1.7924280166625977, "learning_rate": 7.134911580148318e-06, "loss": 0.0203, "step": 2510 }, { "epoch": 1.006391052526463, "grad_norm": 0.6403665542602539, "learning_rate": 7.120650313747861e-06, "loss": 0.0077, "step": 2520 }, { "epoch": 1.0103854603555023, "grad_norm": 1.2908625602722168, "learning_rate": 7.1063890473474055e-06, "loss": 0.0099, "step": 2530 }, { "epoch": 1.0143798681845417, "grad_norm": 0.9998202323913574, "learning_rate": 7.092127780946949e-06, "loss": 0.0079, "step": 2540 }, { "epoch": 1.018374276013581, "grad_norm": 0.6649907827377319, "learning_rate": 7.077866514546492e-06, "loss": 0.0098, "step": 2550 }, { "epoch": 1.0223686838426203, "grad_norm": 1.456899642944336, "learning_rate": 7.063605248146036e-06, "loss": 0.0095, "step": 2560 }, { "epoch": 1.0263630916716597, "grad_norm": 1.0050690174102783, "learning_rate": 7.04934398174558e-06, "loss": 0.0082, "step": 2570 }, { "epoch": 1.030357499500699, "grad_norm": 1.0253567695617676, "learning_rate": 7.035082715345123e-06, "loss": 0.011, "step": 2580 }, { "epoch": 1.0343519073297385, "grad_norm": 1.203191876411438, "learning_rate": 7.020821448944667e-06, "loss": 0.0086, "step": 2590 }, { "epoch": 1.0383463151587777, "grad_norm": 1.5093960762023926, "learning_rate": 7.006560182544211e-06, "loss": 0.0102, "step": 2600 }, { "epoch": 1.042340722987817, "grad_norm": 1.8297464847564697, "learning_rate": 6.992298916143754e-06, "loss": 0.0098, "step": 2610 }, { "epoch": 1.0463351308168565, "grad_norm": 1.1694895029067993, "learning_rate": 6.978037649743298e-06, "loss": 0.0084, "step": 2620 }, { "epoch": 1.0503295386458957, "grad_norm": 1.4883347749710083, "learning_rate": 6.963776383342841e-06, "loss": 0.0069, "step": 2630 }, { "epoch": 1.054323946474935, "grad_norm": 1.6328678131103516, "learning_rate": 6.949515116942385e-06, "loss": 0.0072, "step": 2640 }, { "epoch": 1.0583183543039745, "grad_norm": 1.665246605873108, "learning_rate": 6.935253850541929e-06, "loss": 0.0184, "step": 2650 }, { "epoch": 1.0623127621330137, "grad_norm": 0.8198590874671936, "learning_rate": 6.920992584141473e-06, "loss": 0.0075, "step": 2660 }, { "epoch": 1.0663071699620532, "grad_norm": 0.7312160730361938, "learning_rate": 6.906731317741015e-06, "loss": 0.0106, "step": 2670 }, { "epoch": 1.0703015777910925, "grad_norm": 1.2935587167739868, "learning_rate": 6.892470051340559e-06, "loss": 0.0121, "step": 2680 }, { "epoch": 1.0742959856201317, "grad_norm": 0.6566735506057739, "learning_rate": 6.878208784940103e-06, "loss": 0.0071, "step": 2690 }, { "epoch": 1.0782903934491712, "grad_norm": 1.318433403968811, "learning_rate": 6.863947518539647e-06, "loss": 0.0099, "step": 2700 }, { "epoch": 1.0822848012782105, "grad_norm": 2.248878240585327, "learning_rate": 6.849686252139191e-06, "loss": 0.0133, "step": 2710 }, { "epoch": 1.08627920910725, "grad_norm": 0.9404581785202026, "learning_rate": 6.835424985738735e-06, "loss": 0.0087, "step": 2720 }, { "epoch": 1.0902736169362892, "grad_norm": 1.51760733127594, "learning_rate": 6.821163719338277e-06, "loss": 0.0089, "step": 2730 }, { "epoch": 1.0942680247653285, "grad_norm": 1.5794305801391602, "learning_rate": 6.806902452937821e-06, "loss": 0.0096, "step": 2740 }, { "epoch": 1.098262432594368, "grad_norm": 1.9199167490005493, "learning_rate": 6.792641186537365e-06, "loss": 0.0091, "step": 2750 }, { "epoch": 1.1022568404234072, "grad_norm": 1.6599887609481812, "learning_rate": 6.778379920136909e-06, "loss": 0.0106, "step": 2760 }, { "epoch": 1.1062512482524465, "grad_norm": 0.8282005190849304, "learning_rate": 6.764118653736452e-06, "loss": 0.0118, "step": 2770 }, { "epoch": 1.110245656081486, "grad_norm": 2.2222962379455566, "learning_rate": 6.749857387335996e-06, "loss": 0.0075, "step": 2780 }, { "epoch": 1.1142400639105252, "grad_norm": 0.39625316858291626, "learning_rate": 6.735596120935539e-06, "loss": 0.011, "step": 2790 }, { "epoch": 1.1182344717395647, "grad_norm": 1.2594791650772095, "learning_rate": 6.721334854535083e-06, "loss": 0.0059, "step": 2800 }, { "epoch": 1.122228879568604, "grad_norm": 1.6027462482452393, "learning_rate": 6.7070735881346274e-06, "loss": 0.0098, "step": 2810 }, { "epoch": 1.1262232873976432, "grad_norm": 1.6769353151321411, "learning_rate": 6.69281232173417e-06, "loss": 0.008, "step": 2820 }, { "epoch": 1.1302176952266827, "grad_norm": 0.7512936592102051, "learning_rate": 6.678551055333714e-06, "loss": 0.0141, "step": 2830 }, { "epoch": 1.134212103055722, "grad_norm": 1.700231671333313, "learning_rate": 6.664289788933258e-06, "loss": 0.0121, "step": 2840 }, { "epoch": 1.1382065108847614, "grad_norm": 1.984163761138916, "learning_rate": 6.650028522532801e-06, "loss": 0.0133, "step": 2850 }, { "epoch": 1.1422009187138007, "grad_norm": 2.2872514724731445, "learning_rate": 6.635767256132345e-06, "loss": 0.0094, "step": 2860 }, { "epoch": 1.14619532654284, "grad_norm": 1.120602011680603, "learning_rate": 6.621505989731889e-06, "loss": 0.0076, "step": 2870 }, { "epoch": 1.1501897343718794, "grad_norm": 0.8228394389152527, "learning_rate": 6.607244723331432e-06, "loss": 0.0086, "step": 2880 }, { "epoch": 1.1541841422009187, "grad_norm": 1.6478750705718994, "learning_rate": 6.592983456930976e-06, "loss": 0.0145, "step": 2890 }, { "epoch": 1.158178550029958, "grad_norm": 2.5362820625305176, "learning_rate": 6.57872219053052e-06, "loss": 0.0097, "step": 2900 }, { "epoch": 1.1621729578589974, "grad_norm": 1.6985121965408325, "learning_rate": 6.5644609241300625e-06, "loss": 0.0114, "step": 2910 }, { "epoch": 1.1661673656880367, "grad_norm": 0.9929540753364563, "learning_rate": 6.5501996577296066e-06, "loss": 0.0107, "step": 2920 }, { "epoch": 1.1701617735170762, "grad_norm": 1.1043400764465332, "learning_rate": 6.535938391329151e-06, "loss": 0.0119, "step": 2930 }, { "epoch": 1.1741561813461154, "grad_norm": 2.4200518131256104, "learning_rate": 6.521677124928694e-06, "loss": 0.0111, "step": 2940 }, { "epoch": 1.1781505891751547, "grad_norm": 1.107531189918518, "learning_rate": 6.507415858528238e-06, "loss": 0.009, "step": 2950 }, { "epoch": 1.1821449970041942, "grad_norm": 1.9044643640518188, "learning_rate": 6.493154592127781e-06, "loss": 0.007, "step": 2960 }, { "epoch": 1.1861394048332334, "grad_norm": 1.1096996068954468, "learning_rate": 6.478893325727325e-06, "loss": 0.0076, "step": 2970 }, { "epoch": 1.190133812662273, "grad_norm": 2.1665825843811035, "learning_rate": 6.464632059326869e-06, "loss": 0.0097, "step": 2980 }, { "epoch": 1.1941282204913122, "grad_norm": 1.3429949283599854, "learning_rate": 6.450370792926413e-06, "loss": 0.0069, "step": 2990 }, { "epoch": 1.1981226283203514, "grad_norm": 1.5667604207992554, "learning_rate": 6.436109526525957e-06, "loss": 0.0091, "step": 3000 }, { "epoch": 1.202117036149391, "grad_norm": 1.0686432123184204, "learning_rate": 6.421848260125499e-06, "loss": 0.0073, "step": 3010 }, { "epoch": 1.2061114439784302, "grad_norm": 1.4043235778808594, "learning_rate": 6.407586993725043e-06, "loss": 0.0089, "step": 3020 }, { "epoch": 1.2101058518074694, "grad_norm": 1.6781501770019531, "learning_rate": 6.393325727324587e-06, "loss": 0.0129, "step": 3030 }, { "epoch": 1.214100259636509, "grad_norm": 0.7282060384750366, "learning_rate": 6.379064460924131e-06, "loss": 0.0109, "step": 3040 }, { "epoch": 1.2180946674655482, "grad_norm": 1.3478143215179443, "learning_rate": 6.364803194523675e-06, "loss": 0.0096, "step": 3050 }, { "epoch": 1.2220890752945877, "grad_norm": 1.9877102375030518, "learning_rate": 6.350541928123218e-06, "loss": 0.0083, "step": 3060 }, { "epoch": 1.226083483123627, "grad_norm": 1.2017593383789062, "learning_rate": 6.336280661722761e-06, "loss": 0.0067, "step": 3070 }, { "epoch": 1.2300778909526662, "grad_norm": 0.6694231629371643, "learning_rate": 6.322019395322305e-06, "loss": 0.0074, "step": 3080 }, { "epoch": 1.2340722987817057, "grad_norm": 0.8456215262413025, "learning_rate": 6.307758128921849e-06, "loss": 0.0083, "step": 3090 }, { "epoch": 1.238066706610745, "grad_norm": 0.8469221591949463, "learning_rate": 6.293496862521392e-06, "loss": 0.0111, "step": 3100 }, { "epoch": 1.2420611144397844, "grad_norm": 2.0267868041992188, "learning_rate": 6.279235596120936e-06, "loss": 0.0088, "step": 3110 }, { "epoch": 1.2460555222688237, "grad_norm": 1.6765652894973755, "learning_rate": 6.26497432972048e-06, "loss": 0.0101, "step": 3120 }, { "epoch": 1.250049930097863, "grad_norm": 0.974431037902832, "learning_rate": 6.250713063320023e-06, "loss": 0.0118, "step": 3130 }, { "epoch": 1.2540443379269024, "grad_norm": 1.4480584859848022, "learning_rate": 6.236451796919567e-06, "loss": 0.0096, "step": 3140 }, { "epoch": 1.2580387457559417, "grad_norm": 0.9492830038070679, "learning_rate": 6.222190530519111e-06, "loss": 0.01, "step": 3150 }, { "epoch": 1.262033153584981, "grad_norm": 2.0500473976135254, "learning_rate": 6.207929264118654e-06, "loss": 0.0121, "step": 3160 }, { "epoch": 1.2660275614140204, "grad_norm": 2.319540500640869, "learning_rate": 6.193667997718198e-06, "loss": 0.013, "step": 3170 }, { "epoch": 1.2700219692430597, "grad_norm": 1.5995023250579834, "learning_rate": 6.179406731317742e-06, "loss": 0.014, "step": 3180 }, { "epoch": 1.2740163770720991, "grad_norm": 1.7551332712173462, "learning_rate": 6.165145464917285e-06, "loss": 0.0103, "step": 3190 }, { "epoch": 1.2780107849011384, "grad_norm": 2.8841307163238525, "learning_rate": 6.1508841985168285e-06, "loss": 0.0085, "step": 3200 }, { "epoch": 1.2820051927301779, "grad_norm": 1.5693269968032837, "learning_rate": 6.136622932116373e-06, "loss": 0.013, "step": 3210 }, { "epoch": 1.2859996005592171, "grad_norm": 1.059067726135254, "learning_rate": 6.122361665715916e-06, "loss": 0.0122, "step": 3220 }, { "epoch": 1.2899940083882564, "grad_norm": 1.7539544105529785, "learning_rate": 6.10810039931546e-06, "loss": 0.0092, "step": 3230 }, { "epoch": 1.2939884162172959, "grad_norm": 1.1549718379974365, "learning_rate": 6.093839132915004e-06, "loss": 0.0084, "step": 3240 }, { "epoch": 1.2979828240463351, "grad_norm": 1.7076776027679443, "learning_rate": 6.0795778665145464e-06, "loss": 0.0084, "step": 3250 }, { "epoch": 1.3019772318753744, "grad_norm": 0.9175742268562317, "learning_rate": 6.0653166001140905e-06, "loss": 0.0069, "step": 3260 }, { "epoch": 1.3059716397044139, "grad_norm": 1.440813660621643, "learning_rate": 6.051055333713635e-06, "loss": 0.0115, "step": 3270 }, { "epoch": 1.3099660475334531, "grad_norm": 0.5184957981109619, "learning_rate": 6.036794067313178e-06, "loss": 0.0072, "step": 3280 }, { "epoch": 1.3139604553624924, "grad_norm": 0.3342525362968445, "learning_rate": 6.022532800912721e-06, "loss": 0.014, "step": 3290 }, { "epoch": 1.3179548631915319, "grad_norm": 2.238618850708008, "learning_rate": 6.008271534512265e-06, "loss": 0.0082, "step": 3300 }, { "epoch": 1.3219492710205711, "grad_norm": 7.149013996124268, "learning_rate": 5.9940102681118085e-06, "loss": 0.0127, "step": 3310 }, { "epoch": 1.3259436788496106, "grad_norm": 1.1425492763519287, "learning_rate": 5.9797490017113525e-06, "loss": 0.0115, "step": 3320 }, { "epoch": 1.3299380866786499, "grad_norm": 1.9267038106918335, "learning_rate": 5.965487735310897e-06, "loss": 0.0094, "step": 3330 }, { "epoch": 1.3339324945076894, "grad_norm": 0.8218601942062378, "learning_rate": 5.951226468910439e-06, "loss": 0.008, "step": 3340 }, { "epoch": 1.3379269023367286, "grad_norm": 1.620312213897705, "learning_rate": 5.936965202509983e-06, "loss": 0.0067, "step": 3350 }, { "epoch": 1.3419213101657679, "grad_norm": 0.7273368835449219, "learning_rate": 5.922703936109527e-06, "loss": 0.0096, "step": 3360 }, { "epoch": 1.3459157179948074, "grad_norm": 3.9472219944000244, "learning_rate": 5.9084426697090705e-06, "loss": 0.0102, "step": 3370 }, { "epoch": 1.3499101258238466, "grad_norm": 2.4310262203216553, "learning_rate": 5.8941814033086146e-06, "loss": 0.0154, "step": 3380 }, { "epoch": 1.3539045336528859, "grad_norm": 1.8209127187728882, "learning_rate": 5.879920136908158e-06, "loss": 0.0094, "step": 3390 }, { "epoch": 1.3578989414819254, "grad_norm": 1.022280216217041, "learning_rate": 5.865658870507701e-06, "loss": 0.0062, "step": 3400 }, { "epoch": 1.3618933493109646, "grad_norm": 1.8226642608642578, "learning_rate": 5.851397604107245e-06, "loss": 0.0113, "step": 3410 }, { "epoch": 1.3658877571400039, "grad_norm": 2.004735231399536, "learning_rate": 5.837136337706789e-06, "loss": 0.0121, "step": 3420 }, { "epoch": 1.3698821649690434, "grad_norm": 3.744194507598877, "learning_rate": 5.822875071306332e-06, "loss": 0.0096, "step": 3430 }, { "epoch": 1.3738765727980826, "grad_norm": 1.8464494943618774, "learning_rate": 5.808613804905876e-06, "loss": 0.009, "step": 3440 }, { "epoch": 1.377870980627122, "grad_norm": 1.8000208139419556, "learning_rate": 5.79435253850542e-06, "loss": 0.0112, "step": 3450 }, { "epoch": 1.3818653884561614, "grad_norm": 3.974322557449341, "learning_rate": 5.780091272104963e-06, "loss": 0.0114, "step": 3460 }, { "epoch": 1.3858597962852008, "grad_norm": 0.7045836448669434, "learning_rate": 5.765830005704507e-06, "loss": 0.0091, "step": 3470 }, { "epoch": 1.38985420411424, "grad_norm": 0.4988465905189514, "learning_rate": 5.751568739304051e-06, "loss": 0.0069, "step": 3480 }, { "epoch": 1.3938486119432794, "grad_norm": 0.8017894625663757, "learning_rate": 5.737307472903594e-06, "loss": 0.0096, "step": 3490 }, { "epoch": 1.3978430197723188, "grad_norm": 1.61961030960083, "learning_rate": 5.723046206503138e-06, "loss": 0.0081, "step": 3500 }, { "epoch": 1.401837427601358, "grad_norm": 2.5354208946228027, "learning_rate": 5.708784940102682e-06, "loss": 0.0091, "step": 3510 }, { "epoch": 1.4058318354303974, "grad_norm": 0.8456679582595825, "learning_rate": 5.694523673702225e-06, "loss": 0.012, "step": 3520 }, { "epoch": 1.4098262432594368, "grad_norm": 2.605341911315918, "learning_rate": 5.680262407301768e-06, "loss": 0.0088, "step": 3530 }, { "epoch": 1.413820651088476, "grad_norm": 1.4951987266540527, "learning_rate": 5.6660011409013125e-06, "loss": 0.013, "step": 3540 }, { "epoch": 1.4178150589175154, "grad_norm": 2.0663983821868896, "learning_rate": 5.651739874500856e-06, "loss": 0.0096, "step": 3550 }, { "epoch": 1.4218094667465548, "grad_norm": 0.9899235963821411, "learning_rate": 5.6374786081004e-06, "loss": 0.0079, "step": 3560 }, { "epoch": 1.425803874575594, "grad_norm": 0.7591453194618225, "learning_rate": 5.623217341699944e-06, "loss": 0.0067, "step": 3570 }, { "epoch": 1.4297982824046336, "grad_norm": 1.2285091876983643, "learning_rate": 5.608956075299488e-06, "loss": 0.0062, "step": 3580 }, { "epoch": 1.4337926902336728, "grad_norm": 1.8616957664489746, "learning_rate": 5.59469480889903e-06, "loss": 0.0092, "step": 3590 }, { "epoch": 1.4377870980627123, "grad_norm": 1.0690120458602905, "learning_rate": 5.5804335424985745e-06, "loss": 0.0066, "step": 3600 }, { "epoch": 1.4417815058917516, "grad_norm": 1.20885169506073, "learning_rate": 5.566172276098119e-06, "loss": 0.0107, "step": 3610 }, { "epoch": 1.4457759137207908, "grad_norm": 1.5971421003341675, "learning_rate": 5.551911009697661e-06, "loss": 0.0079, "step": 3620 }, { "epoch": 1.4497703215498303, "grad_norm": 1.570513129234314, "learning_rate": 5.537649743297205e-06, "loss": 0.0087, "step": 3630 }, { "epoch": 1.4537647293788696, "grad_norm": 0.9042888879776001, "learning_rate": 5.523388476896749e-06, "loss": 0.0127, "step": 3640 }, { "epoch": 1.4577591372079088, "grad_norm": 1.7177059650421143, "learning_rate": 5.509127210496292e-06, "loss": 0.008, "step": 3650 }, { "epoch": 1.4617535450369483, "grad_norm": 1.1265398263931274, "learning_rate": 5.4948659440958365e-06, "loss": 0.0066, "step": 3660 }, { "epoch": 1.4657479528659876, "grad_norm": 2.9831197261810303, "learning_rate": 5.480604677695381e-06, "loss": 0.0072, "step": 3670 }, { "epoch": 1.4697423606950268, "grad_norm": 1.4246543645858765, "learning_rate": 5.466343411294923e-06, "loss": 0.0074, "step": 3680 }, { "epoch": 1.4737367685240663, "grad_norm": 1.62272310256958, "learning_rate": 5.452082144894467e-06, "loss": 0.0077, "step": 3690 }, { "epoch": 1.4777311763531056, "grad_norm": 1.5177655220031738, "learning_rate": 5.437820878494011e-06, "loss": 0.0121, "step": 3700 }, { "epoch": 1.481725584182145, "grad_norm": 2.8331351280212402, "learning_rate": 5.4235596120935544e-06, "loss": 0.0061, "step": 3710 }, { "epoch": 1.4857199920111843, "grad_norm": 1.5248603820800781, "learning_rate": 5.409298345693098e-06, "loss": 0.0109, "step": 3720 }, { "epoch": 1.4897143998402238, "grad_norm": 1.143097996711731, "learning_rate": 5.395037079292642e-06, "loss": 0.01, "step": 3730 }, { "epoch": 1.493708807669263, "grad_norm": 1.9222826957702637, "learning_rate": 5.380775812892185e-06, "loss": 0.0076, "step": 3740 }, { "epoch": 1.4977032154983023, "grad_norm": 1.4354894161224365, "learning_rate": 5.366514546491729e-06, "loss": 0.0101, "step": 3750 }, { "epoch": 1.5016976233273418, "grad_norm": 0.6621835231781006, "learning_rate": 5.352253280091273e-06, "loss": 0.0098, "step": 3760 }, { "epoch": 1.505692031156381, "grad_norm": 1.5087827444076538, "learning_rate": 5.337992013690816e-06, "loss": 0.0114, "step": 3770 }, { "epoch": 1.5096864389854203, "grad_norm": 1.2942672967910767, "learning_rate": 5.32373074729036e-06, "loss": 0.0069, "step": 3780 }, { "epoch": 1.5136808468144598, "grad_norm": 1.9707920551300049, "learning_rate": 5.309469480889904e-06, "loss": 0.0098, "step": 3790 }, { "epoch": 1.517675254643499, "grad_norm": 0.8530262112617493, "learning_rate": 5.295208214489447e-06, "loss": 0.0083, "step": 3800 }, { "epoch": 1.5216696624725383, "grad_norm": 1.9188873767852783, "learning_rate": 5.280946948088991e-06, "loss": 0.0081, "step": 3810 }, { "epoch": 1.5256640703015778, "grad_norm": 1.7006468772888184, "learning_rate": 5.266685681688534e-06, "loss": 0.0136, "step": 3820 }, { "epoch": 1.5296584781306173, "grad_norm": 1.3003968000411987, "learning_rate": 5.252424415288078e-06, "loss": 0.0071, "step": 3830 }, { "epoch": 1.5336528859596563, "grad_norm": 1.1423379182815552, "learning_rate": 5.238163148887622e-06, "loss": 0.0095, "step": 3840 }, { "epoch": 1.5376472937886958, "grad_norm": 0.8597344756126404, "learning_rate": 5.223901882487166e-06, "loss": 0.0084, "step": 3850 }, { "epoch": 1.5416417016177353, "grad_norm": 1.0262315273284912, "learning_rate": 5.209640616086708e-06, "loss": 0.0104, "step": 3860 }, { "epoch": 1.5456361094467745, "grad_norm": 1.3405098915100098, "learning_rate": 5.195379349686252e-06, "loss": 0.0172, "step": 3870 }, { "epoch": 1.5496305172758138, "grad_norm": 0.7442639470100403, "learning_rate": 5.1811180832857964e-06, "loss": 0.0077, "step": 3880 }, { "epoch": 1.5536249251048533, "grad_norm": 0.856499433517456, "learning_rate": 5.16685681688534e-06, "loss": 0.0089, "step": 3890 }, { "epoch": 1.5576193329338925, "grad_norm": 2.6278645992279053, "learning_rate": 5.152595550484884e-06, "loss": 0.0095, "step": 3900 }, { "epoch": 1.5616137407629318, "grad_norm": 1.86880624294281, "learning_rate": 5.138334284084428e-06, "loss": 0.0124, "step": 3910 }, { "epoch": 1.5656081485919713, "grad_norm": 0.583057165145874, "learning_rate": 5.12407301768397e-06, "loss": 0.0062, "step": 3920 }, { "epoch": 1.5696025564210105, "grad_norm": 1.2156856060028076, "learning_rate": 5.109811751283514e-06, "loss": 0.0081, "step": 3930 }, { "epoch": 1.5735969642500498, "grad_norm": 0.9274953603744507, "learning_rate": 5.0955504848830584e-06, "loss": 0.01, "step": 3940 }, { "epoch": 1.5775913720790893, "grad_norm": 1.0414108037948608, "learning_rate": 5.081289218482601e-06, "loss": 0.0088, "step": 3950 }, { "epoch": 1.5815857799081288, "grad_norm": 1.3919917345046997, "learning_rate": 5.067027952082145e-06, "loss": 0.0082, "step": 3960 }, { "epoch": 1.5855801877371678, "grad_norm": 0.8071143627166748, "learning_rate": 5.052766685681689e-06, "loss": 0.0073, "step": 3970 }, { "epoch": 1.5895745955662073, "grad_norm": 1.5692224502563477, "learning_rate": 5.038505419281232e-06, "loss": 0.0074, "step": 3980 }, { "epoch": 1.5935690033952468, "grad_norm": 1.0880498886108398, "learning_rate": 5.024244152880776e-06, "loss": 0.0127, "step": 3990 }, { "epoch": 1.597563411224286, "grad_norm": 1.7401431798934937, "learning_rate": 5.0099828864803205e-06, "loss": 0.015, "step": 4000 }, { "epoch": 1.597563411224286, "eval_loss": 0.01016959734261036, "eval_runtime": 7385.3342, "eval_samples_per_second": 2.711, "eval_steps_per_second": 0.339, "eval_wer": 0.8737253615462425, "step": 4000 }, { "epoch": 1.6015578190533253, "grad_norm": 0.680465817451477, "learning_rate": 4.995721620079864e-06, "loss": 0.009, "step": 4010 }, { "epoch": 1.6055522268823648, "grad_norm": 0.3950757086277008, "learning_rate": 4.981460353679407e-06, "loss": 0.0091, "step": 4020 }, { "epoch": 1.609546634711404, "grad_norm": 2.6939499378204346, "learning_rate": 4.96719908727895e-06, "loss": 0.0106, "step": 4030 }, { "epoch": 1.6135410425404433, "grad_norm": 1.401220679283142, "learning_rate": 4.952937820878494e-06, "loss": 0.0112, "step": 4040 }, { "epoch": 1.6175354503694828, "grad_norm": 1.7981312274932861, "learning_rate": 4.9386765544780376e-06, "loss": 0.0061, "step": 4050 }, { "epoch": 1.621529858198522, "grad_norm": 2.05911922454834, "learning_rate": 4.924415288077582e-06, "loss": 0.0088, "step": 4060 }, { "epoch": 1.6255242660275613, "grad_norm": 0.5581210851669312, "learning_rate": 4.910154021677126e-06, "loss": 0.0095, "step": 4070 }, { "epoch": 1.6295186738566008, "grad_norm": 1.671968936920166, "learning_rate": 4.895892755276669e-06, "loss": 0.0102, "step": 4080 }, { "epoch": 1.6335130816856402, "grad_norm": 1.0555320978164673, "learning_rate": 4.881631488876213e-06, "loss": 0.012, "step": 4090 }, { "epoch": 1.6375074895146795, "grad_norm": 0.5234736204147339, "learning_rate": 4.867370222475756e-06, "loss": 0.0092, "step": 4100 }, { "epoch": 1.6415018973437188, "grad_norm": 1.0109747648239136, "learning_rate": 4.8531089560753e-06, "loss": 0.0086, "step": 4110 }, { "epoch": 1.6454963051727582, "grad_norm": 2.0059330463409424, "learning_rate": 4.838847689674844e-06, "loss": 0.0119, "step": 4120 }, { "epoch": 1.6494907130017975, "grad_norm": 1.5553911924362183, "learning_rate": 4.824586423274387e-06, "loss": 0.0083, "step": 4130 }, { "epoch": 1.6534851208308368, "grad_norm": 1.573845624923706, "learning_rate": 4.810325156873931e-06, "loss": 0.0093, "step": 4140 }, { "epoch": 1.6574795286598762, "grad_norm": 0.5471251606941223, "learning_rate": 4.796063890473474e-06, "loss": 0.0067, "step": 4150 }, { "epoch": 1.6614739364889155, "grad_norm": 2.054495096206665, "learning_rate": 4.781802624073018e-06, "loss": 0.016, "step": 4160 }, { "epoch": 1.6654683443179548, "grad_norm": 1.1892727613449097, "learning_rate": 4.767541357672562e-06, "loss": 0.0089, "step": 4170 }, { "epoch": 1.6694627521469942, "grad_norm": 0.7050298452377319, "learning_rate": 4.753280091272106e-06, "loss": 0.0068, "step": 4180 }, { "epoch": 1.6734571599760335, "grad_norm": 1.208492398262024, "learning_rate": 4.739018824871649e-06, "loss": 0.0064, "step": 4190 }, { "epoch": 1.6774515678050728, "grad_norm": 0.864264965057373, "learning_rate": 4.724757558471192e-06, "loss": 0.0084, "step": 4200 }, { "epoch": 1.6814459756341122, "grad_norm": 1.3503026962280273, "learning_rate": 4.710496292070736e-06, "loss": 0.0082, "step": 4210 }, { "epoch": 1.6854403834631517, "grad_norm": 1.3305470943450928, "learning_rate": 4.6962350256702795e-06, "loss": 0.0108, "step": 4220 }, { "epoch": 1.689434791292191, "grad_norm": 1.7467732429504395, "learning_rate": 4.681973759269824e-06, "loss": 0.0081, "step": 4230 }, { "epoch": 1.6934291991212302, "grad_norm": 2.1052684783935547, "learning_rate": 4.667712492869368e-06, "loss": 0.0157, "step": 4240 }, { "epoch": 1.6974236069502697, "grad_norm": 1.0252904891967773, "learning_rate": 4.653451226468911e-06, "loss": 0.0066, "step": 4250 }, { "epoch": 1.701418014779309, "grad_norm": 1.6374844312667847, "learning_rate": 4.639189960068454e-06, "loss": 0.0096, "step": 4260 }, { "epoch": 1.7054124226083482, "grad_norm": 1.1286324262619019, "learning_rate": 4.624928693667998e-06, "loss": 0.0076, "step": 4270 }, { "epoch": 1.7094068304373877, "grad_norm": 2.2228775024414062, "learning_rate": 4.6106674272675416e-06, "loss": 0.0092, "step": 4280 }, { "epoch": 1.713401238266427, "grad_norm": 1.5603349208831787, "learning_rate": 4.596406160867085e-06, "loss": 0.0065, "step": 4290 }, { "epoch": 1.7173956460954662, "grad_norm": 1.7231379747390747, "learning_rate": 4.582144894466629e-06, "loss": 0.0079, "step": 4300 }, { "epoch": 1.7213900539245057, "grad_norm": 2.4461798667907715, "learning_rate": 4.567883628066173e-06, "loss": 0.0091, "step": 4310 }, { "epoch": 1.725384461753545, "grad_norm": 1.6841260194778442, "learning_rate": 4.553622361665716e-06, "loss": 0.0102, "step": 4320 }, { "epoch": 1.7293788695825842, "grad_norm": 1.3950663805007935, "learning_rate": 4.53936109526526e-06, "loss": 0.0063, "step": 4330 }, { "epoch": 1.7333732774116237, "grad_norm": 1.364123821258545, "learning_rate": 4.525099828864804e-06, "loss": 0.0088, "step": 4340 }, { "epoch": 1.7373676852406632, "grad_norm": 1.0013163089752197, "learning_rate": 4.510838562464347e-06, "loss": 0.0083, "step": 4350 }, { "epoch": 1.7413620930697025, "grad_norm": 3.0454695224761963, "learning_rate": 4.496577296063891e-06, "loss": 0.007, "step": 4360 }, { "epoch": 1.7453565008987417, "grad_norm": 1.0976226329803467, "learning_rate": 4.482316029663434e-06, "loss": 0.0065, "step": 4370 }, { "epoch": 1.7493509087277812, "grad_norm": 1.0366853475570679, "learning_rate": 4.468054763262978e-06, "loss": 0.0072, "step": 4380 }, { "epoch": 1.7533453165568205, "grad_norm": 1.7006181478500366, "learning_rate": 4.4537934968625215e-06, "loss": 0.0079, "step": 4390 }, { "epoch": 1.7573397243858597, "grad_norm": 2.483431100845337, "learning_rate": 4.439532230462066e-06, "loss": 0.0097, "step": 4400 }, { "epoch": 1.7613341322148992, "grad_norm": 0.44049394130706787, "learning_rate": 4.425270964061609e-06, "loss": 0.011, "step": 4410 }, { "epoch": 1.7653285400439385, "grad_norm": 1.3942147493362427, "learning_rate": 4.411009697661153e-06, "loss": 0.0076, "step": 4420 }, { "epoch": 1.7693229478729777, "grad_norm": 1.5767748355865479, "learning_rate": 4.396748431260696e-06, "loss": 0.0114, "step": 4430 }, { "epoch": 1.7733173557020172, "grad_norm": 2.5710384845733643, "learning_rate": 4.38248716486024e-06, "loss": 0.0123, "step": 4440 }, { "epoch": 1.7773117635310565, "grad_norm": 0.730352520942688, "learning_rate": 4.3682258984597835e-06, "loss": 0.0048, "step": 4450 }, { "epoch": 1.7813061713600957, "grad_norm": 0.7298102974891663, "learning_rate": 4.353964632059327e-06, "loss": 0.0135, "step": 4460 }, { "epoch": 1.7853005791891352, "grad_norm": 0.6274979114532471, "learning_rate": 4.339703365658871e-06, "loss": 0.0091, "step": 4470 }, { "epoch": 1.7892949870181747, "grad_norm": 3.2221016883850098, "learning_rate": 4.325442099258414e-06, "loss": 0.0136, "step": 4480 }, { "epoch": 1.793289394847214, "grad_norm": 0.959096372127533, "learning_rate": 4.311180832857958e-06, "loss": 0.0083, "step": 4490 }, { "epoch": 1.7972838026762532, "grad_norm": 1.216997504234314, "learning_rate": 4.296919566457502e-06, "loss": 0.0064, "step": 4500 }, { "epoch": 1.8012782105052927, "grad_norm": 1.628796935081482, "learning_rate": 4.2826583000570456e-06, "loss": 0.0082, "step": 4510 }, { "epoch": 1.805272618334332, "grad_norm": 0.5727189183235168, "learning_rate": 4.268397033656589e-06, "loss": 0.0102, "step": 4520 }, { "epoch": 1.8092670261633712, "grad_norm": 2.1575019359588623, "learning_rate": 4.254135767256133e-06, "loss": 0.0072, "step": 4530 }, { "epoch": 1.8132614339924107, "grad_norm": 1.3159669637680054, "learning_rate": 4.239874500855676e-06, "loss": 0.0115, "step": 4540 }, { "epoch": 1.81725584182145, "grad_norm": 1.2897024154663086, "learning_rate": 4.225613234455219e-06, "loss": 0.007, "step": 4550 }, { "epoch": 1.8212502496504892, "grad_norm": 1.1860707998275757, "learning_rate": 4.2113519680547635e-06, "loss": 0.0071, "step": 4560 }, { "epoch": 1.8252446574795287, "grad_norm": 1.6072814464569092, "learning_rate": 4.197090701654308e-06, "loss": 0.0072, "step": 4570 }, { "epoch": 1.8292390653085682, "grad_norm": 0.8610338568687439, "learning_rate": 4.182829435253851e-06, "loss": 0.0079, "step": 4580 }, { "epoch": 1.8332334731376072, "grad_norm": 1.0036977529525757, "learning_rate": 4.168568168853395e-06, "loss": 0.0053, "step": 4590 }, { "epoch": 1.8372278809666467, "grad_norm": 0.8917168974876404, "learning_rate": 4.154306902452938e-06, "loss": 0.0065, "step": 4600 }, { "epoch": 1.8412222887956862, "grad_norm": 2.3281478881835938, "learning_rate": 4.1400456360524814e-06, "loss": 0.0067, "step": 4610 }, { "epoch": 1.8452166966247254, "grad_norm": 1.1472433805465698, "learning_rate": 4.1257843696520255e-06, "loss": 0.0116, "step": 4620 }, { "epoch": 1.8492111044537647, "grad_norm": 1.0347813367843628, "learning_rate": 4.111523103251569e-06, "loss": 0.0057, "step": 4630 }, { "epoch": 1.8532055122828042, "grad_norm": 0.8118449449539185, "learning_rate": 4.097261836851113e-06, "loss": 0.0067, "step": 4640 }, { "epoch": 1.8571999201118434, "grad_norm": 0.8567641973495483, "learning_rate": 4.083000570450656e-06, "loss": 0.0099, "step": 4650 }, { "epoch": 1.8611943279408827, "grad_norm": 1.2339649200439453, "learning_rate": 4.0687393040502e-06, "loss": 0.0057, "step": 4660 }, { "epoch": 1.8651887357699222, "grad_norm": 3.1318016052246094, "learning_rate": 4.054478037649744e-06, "loss": 0.0083, "step": 4670 }, { "epoch": 1.8691831435989614, "grad_norm": 1.414455771446228, "learning_rate": 4.0402167712492876e-06, "loss": 0.0082, "step": 4680 }, { "epoch": 1.8731775514280007, "grad_norm": 1.44307279586792, "learning_rate": 4.025955504848831e-06, "loss": 0.0109, "step": 4690 }, { "epoch": 1.8771719592570402, "grad_norm": 1.7109726667404175, "learning_rate": 4.011694238448375e-06, "loss": 0.0083, "step": 4700 }, { "epoch": 1.8811663670860796, "grad_norm": 0.5837492942810059, "learning_rate": 3.997432972047918e-06, "loss": 0.0093, "step": 4710 }, { "epoch": 1.8851607749151187, "grad_norm": 1.2008476257324219, "learning_rate": 3.983171705647461e-06, "loss": 0.0084, "step": 4720 }, { "epoch": 1.8891551827441582, "grad_norm": 0.686278760433197, "learning_rate": 3.9689104392470055e-06, "loss": 0.0081, "step": 4730 }, { "epoch": 1.8931495905731976, "grad_norm": 1.5602636337280273, "learning_rate": 3.954649172846549e-06, "loss": 0.0063, "step": 4740 }, { "epoch": 1.897143998402237, "grad_norm": 2.410867214202881, "learning_rate": 3.940387906446093e-06, "loss": 0.0109, "step": 4750 }, { "epoch": 1.9011384062312762, "grad_norm": 1.423948049545288, "learning_rate": 3.926126640045637e-06, "loss": 0.0082, "step": 4760 }, { "epoch": 1.9051328140603156, "grad_norm": 1.2718310356140137, "learning_rate": 3.91186537364518e-06, "loss": 0.0093, "step": 4770 }, { "epoch": 1.909127221889355, "grad_norm": 1.6484652757644653, "learning_rate": 3.897604107244723e-06, "loss": 0.0053, "step": 4780 }, { "epoch": 1.9131216297183942, "grad_norm": 1.5819053649902344, "learning_rate": 3.8833428408442675e-06, "loss": 0.007, "step": 4790 }, { "epoch": 1.9171160375474336, "grad_norm": 0.7520307898521423, "learning_rate": 3.869081574443811e-06, "loss": 0.0077, "step": 4800 }, { "epoch": 1.921110445376473, "grad_norm": 1.954355001449585, "learning_rate": 3.854820308043354e-06, "loss": 0.006, "step": 4810 }, { "epoch": 1.9251048532055122, "grad_norm": 1.853141188621521, "learning_rate": 3.840559041642898e-06, "loss": 0.0061, "step": 4820 }, { "epoch": 1.9290992610345516, "grad_norm": 1.5008714199066162, "learning_rate": 3.826297775242442e-06, "loss": 0.0074, "step": 4830 }, { "epoch": 1.9330936688635911, "grad_norm": 1.2396436929702759, "learning_rate": 3.8120365088419854e-06, "loss": 0.0068, "step": 4840 }, { "epoch": 1.9370880766926302, "grad_norm": 1.0389924049377441, "learning_rate": 3.797775242441529e-06, "loss": 0.0103, "step": 4850 }, { "epoch": 1.9410824845216696, "grad_norm": 0.9938933253288269, "learning_rate": 3.7835139760410728e-06, "loss": 0.0092, "step": 4860 }, { "epoch": 1.9450768923507091, "grad_norm": 2.3726730346679688, "learning_rate": 3.769252709640616e-06, "loss": 0.0059, "step": 4870 }, { "epoch": 1.9490713001797484, "grad_norm": 1.4253814220428467, "learning_rate": 3.75499144324016e-06, "loss": 0.0071, "step": 4880 }, { "epoch": 1.9530657080087876, "grad_norm": 0.777284562587738, "learning_rate": 3.740730176839704e-06, "loss": 0.0083, "step": 4890 }, { "epoch": 1.9570601158378271, "grad_norm": 1.6175594329833984, "learning_rate": 3.726468910439247e-06, "loss": 0.0081, "step": 4900 }, { "epoch": 1.9610545236668664, "grad_norm": 1.2781018018722534, "learning_rate": 3.712207644038791e-06, "loss": 0.0078, "step": 4910 }, { "epoch": 1.9650489314959056, "grad_norm": 1.007778525352478, "learning_rate": 3.6979463776383344e-06, "loss": 0.0073, "step": 4920 }, { "epoch": 1.9690433393249451, "grad_norm": 2.0084667205810547, "learning_rate": 3.683685111237878e-06, "loss": 0.0089, "step": 4930 }, { "epoch": 1.9730377471539844, "grad_norm": 1.049784541130066, "learning_rate": 3.669423844837422e-06, "loss": 0.0064, "step": 4940 }, { "epoch": 1.9770321549830236, "grad_norm": 1.5071396827697754, "learning_rate": 3.6551625784369654e-06, "loss": 0.0059, "step": 4950 }, { "epoch": 1.9810265628120631, "grad_norm": 1.4314241409301758, "learning_rate": 3.640901312036509e-06, "loss": 0.0069, "step": 4960 }, { "epoch": 1.9850209706411026, "grad_norm": 1.3423244953155518, "learning_rate": 3.6266400456360527e-06, "loss": 0.0084, "step": 4970 }, { "epoch": 1.9890153784701416, "grad_norm": 0.7151647806167603, "learning_rate": 3.6123787792355964e-06, "loss": 0.009, "step": 4980 }, { "epoch": 1.9930097862991811, "grad_norm": 0.9428077936172485, "learning_rate": 3.5981175128351405e-06, "loss": 0.0047, "step": 4990 }, { "epoch": 1.9970041941282206, "grad_norm": 1.7378363609313965, "learning_rate": 3.5838562464346838e-06, "loss": 0.0073, "step": 5000 }, { "epoch": 2.000798881565808, "grad_norm": 0.26533937454223633, "learning_rate": 3.5695949800342274e-06, "loss": 0.004, "step": 5010 }, { "epoch": 2.004793289394847, "grad_norm": 0.6320909261703491, "learning_rate": 3.555333713633771e-06, "loss": 0.0028, "step": 5020 }, { "epoch": 2.0087876972238865, "grad_norm": 0.3733059763908386, "learning_rate": 3.5410724472333148e-06, "loss": 0.0024, "step": 5030 }, { "epoch": 2.012782105052926, "grad_norm": 0.3936101794242859, "learning_rate": 3.526811180832858e-06, "loss": 0.0015, "step": 5040 }, { "epoch": 2.0167765128819655, "grad_norm": 0.34495025873184204, "learning_rate": 3.512549914432402e-06, "loss": 0.0017, "step": 5050 }, { "epoch": 2.0207709207110045, "grad_norm": 0.19415637850761414, "learning_rate": 3.4982886480319454e-06, "loss": 0.0023, "step": 5060 }, { "epoch": 2.024765328540044, "grad_norm": 0.3872200548648834, "learning_rate": 3.484027381631489e-06, "loss": 0.0016, "step": 5070 }, { "epoch": 2.0287597363690835, "grad_norm": 0.3008989691734314, "learning_rate": 3.469766115231033e-06, "loss": 0.0013, "step": 5080 }, { "epoch": 2.0327541441981225, "grad_norm": 0.2125740796327591, "learning_rate": 3.4555048488305764e-06, "loss": 0.0017, "step": 5090 }, { "epoch": 2.036748552027162, "grad_norm": 0.5899261236190796, "learning_rate": 3.44124358243012e-06, "loss": 0.0021, "step": 5100 }, { "epoch": 2.0407429598562015, "grad_norm": 0.5872080326080322, "learning_rate": 3.4269823160296637e-06, "loss": 0.0021, "step": 5110 }, { "epoch": 2.0447373676852405, "grad_norm": 1.2929030656814575, "learning_rate": 3.4127210496292074e-06, "loss": 0.0028, "step": 5120 }, { "epoch": 2.04873177551428, "grad_norm": 0.15616489946842194, "learning_rate": 3.3984597832287506e-06, "loss": 0.0023, "step": 5130 }, { "epoch": 2.0527261833433195, "grad_norm": 0.25699567794799805, "learning_rate": 3.3841985168282947e-06, "loss": 0.0033, "step": 5140 }, { "epoch": 2.0567205911723585, "grad_norm": 1.5154634714126587, "learning_rate": 3.3699372504278384e-06, "loss": 0.0017, "step": 5150 }, { "epoch": 2.060714999001398, "grad_norm": 0.952170729637146, "learning_rate": 3.3556759840273816e-06, "loss": 0.0029, "step": 5160 }, { "epoch": 2.0647094068304375, "grad_norm": 0.32178279757499695, "learning_rate": 3.3414147176269257e-06, "loss": 0.0026, "step": 5170 }, { "epoch": 2.068703814659477, "grad_norm": 0.3404058516025543, "learning_rate": 3.327153451226469e-06, "loss": 0.0022, "step": 5180 }, { "epoch": 2.072698222488516, "grad_norm": 0.47805553674697876, "learning_rate": 3.3128921848260127e-06, "loss": 0.0044, "step": 5190 }, { "epoch": 2.0766926303175555, "grad_norm": 0.15118145942687988, "learning_rate": 3.2986309184255568e-06, "loss": 0.0015, "step": 5200 }, { "epoch": 2.080687038146595, "grad_norm": 1.853371262550354, "learning_rate": 3.2843696520251e-06, "loss": 0.0057, "step": 5210 }, { "epoch": 2.084681445975634, "grad_norm": 0.37700751423835754, "learning_rate": 3.2701083856246437e-06, "loss": 0.0051, "step": 5220 }, { "epoch": 2.0886758538046735, "grad_norm": 0.3602418601512909, "learning_rate": 3.2558471192241873e-06, "loss": 0.0019, "step": 5230 }, { "epoch": 2.092670261633713, "grad_norm": 0.20452773571014404, "learning_rate": 3.241585852823731e-06, "loss": 0.0028, "step": 5240 }, { "epoch": 2.096664669462752, "grad_norm": 0.23050379753112793, "learning_rate": 3.2273245864232743e-06, "loss": 0.0013, "step": 5250 }, { "epoch": 2.1006590772917915, "grad_norm": 1.7879595756530762, "learning_rate": 3.2130633200228184e-06, "loss": 0.0054, "step": 5260 }, { "epoch": 2.104653485120831, "grad_norm": 0.29012367129325867, "learning_rate": 3.198802053622362e-06, "loss": 0.0018, "step": 5270 }, { "epoch": 2.10864789294987, "grad_norm": 0.08294820040464401, "learning_rate": 3.1845407872219053e-06, "loss": 0.0021, "step": 5280 }, { "epoch": 2.1126423007789095, "grad_norm": 0.23702895641326904, "learning_rate": 3.1702795208214494e-06, "loss": 0.0017, "step": 5290 }, { "epoch": 2.116636708607949, "grad_norm": 0.1386740505695343, "learning_rate": 3.1560182544209926e-06, "loss": 0.0014, "step": 5300 }, { "epoch": 2.1206311164369884, "grad_norm": 0.18367265164852142, "learning_rate": 3.1417569880205367e-06, "loss": 0.0027, "step": 5310 }, { "epoch": 2.1246255242660275, "grad_norm": 0.40612369775772095, "learning_rate": 3.1274957216200804e-06, "loss": 0.0031, "step": 5320 }, { "epoch": 2.128619932095067, "grad_norm": 1.8997491598129272, "learning_rate": 3.1132344552196236e-06, "loss": 0.0016, "step": 5330 }, { "epoch": 2.1326143399241064, "grad_norm": 0.16036643087863922, "learning_rate": 3.0989731888191677e-06, "loss": 0.0019, "step": 5340 }, { "epoch": 2.1366087477531455, "grad_norm": 0.1557471603155136, "learning_rate": 3.084711922418711e-06, "loss": 0.0022, "step": 5350 }, { "epoch": 2.140603155582185, "grad_norm": 0.12167658656835556, "learning_rate": 3.0704506560182546e-06, "loss": 0.0028, "step": 5360 }, { "epoch": 2.1445975634112244, "grad_norm": 0.21444401144981384, "learning_rate": 3.0561893896177987e-06, "loss": 0.0033, "step": 5370 }, { "epoch": 2.1485919712402635, "grad_norm": 0.588347852230072, "learning_rate": 3.041928123217342e-06, "loss": 0.0019, "step": 5380 }, { "epoch": 2.152586379069303, "grad_norm": 0.1496289074420929, "learning_rate": 3.0276668568168852e-06, "loss": 0.0022, "step": 5390 }, { "epoch": 2.1565807868983424, "grad_norm": 0.11246706545352936, "learning_rate": 3.0134055904164293e-06, "loss": 0.0019, "step": 5400 }, { "epoch": 2.1605751947273815, "grad_norm": 0.6423232555389404, "learning_rate": 2.999144324015973e-06, "loss": 0.0012, "step": 5410 }, { "epoch": 2.164569602556421, "grad_norm": 0.17058268189430237, "learning_rate": 2.9848830576155162e-06, "loss": 0.0016, "step": 5420 }, { "epoch": 2.1685640103854604, "grad_norm": 0.4728565216064453, "learning_rate": 2.9706217912150603e-06, "loss": 0.0052, "step": 5430 }, { "epoch": 2.1725584182145, "grad_norm": 0.19660444557666779, "learning_rate": 2.9563605248146036e-06, "loss": 0.003, "step": 5440 }, { "epoch": 2.176552826043539, "grad_norm": 0.35805192589759827, "learning_rate": 2.9420992584141473e-06, "loss": 0.0013, "step": 5450 }, { "epoch": 2.1805472338725784, "grad_norm": 0.5961002111434937, "learning_rate": 2.9278379920136913e-06, "loss": 0.0016, "step": 5460 }, { "epoch": 2.184541641701618, "grad_norm": 0.40820205211639404, "learning_rate": 2.9135767256132346e-06, "loss": 0.0012, "step": 5470 }, { "epoch": 2.188536049530657, "grad_norm": 0.22801029682159424, "learning_rate": 2.8993154592127783e-06, "loss": 0.0019, "step": 5480 }, { "epoch": 2.1925304573596964, "grad_norm": 0.18611891567707062, "learning_rate": 2.885054192812322e-06, "loss": 0.0023, "step": 5490 }, { "epoch": 2.196524865188736, "grad_norm": 0.2685863673686981, "learning_rate": 2.8707929264118656e-06, "loss": 0.0024, "step": 5500 }, { "epoch": 2.200519273017775, "grad_norm": 0.4682328402996063, "learning_rate": 2.856531660011409e-06, "loss": 0.0018, "step": 5510 }, { "epoch": 2.2045136808468144, "grad_norm": 0.11273177713155746, "learning_rate": 2.842270393610953e-06, "loss": 0.0013, "step": 5520 }, { "epoch": 2.208508088675854, "grad_norm": 0.3277330994606018, "learning_rate": 2.8280091272104966e-06, "loss": 0.0014, "step": 5530 }, { "epoch": 2.212502496504893, "grad_norm": 0.12008125334978104, "learning_rate": 2.81374786081004e-06, "loss": 0.0014, "step": 5540 }, { "epoch": 2.2164969043339324, "grad_norm": 0.07059912383556366, "learning_rate": 2.799486594409584e-06, "loss": 0.0038, "step": 5550 }, { "epoch": 2.220491312162972, "grad_norm": 0.33425989747047424, "learning_rate": 2.785225328009127e-06, "loss": 0.004, "step": 5560 }, { "epoch": 2.2244857199920114, "grad_norm": 0.7122307419776917, "learning_rate": 2.770964061608671e-06, "loss": 0.0022, "step": 5570 }, { "epoch": 2.2284801278210504, "grad_norm": 0.598380982875824, "learning_rate": 2.756702795208215e-06, "loss": 0.0019, "step": 5580 }, { "epoch": 2.23247453565009, "grad_norm": 0.10900934785604477, "learning_rate": 2.7424415288077582e-06, "loss": 0.0014, "step": 5590 }, { "epoch": 2.2364689434791294, "grad_norm": 0.11831914633512497, "learning_rate": 2.728180262407302e-06, "loss": 0.0032, "step": 5600 }, { "epoch": 2.2404633513081684, "grad_norm": 0.15863606333732605, "learning_rate": 2.7139189960068456e-06, "loss": 0.0021, "step": 5610 }, { "epoch": 2.244457759137208, "grad_norm": 0.8150619864463806, "learning_rate": 2.6996577296063892e-06, "loss": 0.0025, "step": 5620 }, { "epoch": 2.2484521669662474, "grad_norm": 0.16623784601688385, "learning_rate": 2.6853964632059333e-06, "loss": 0.0016, "step": 5630 }, { "epoch": 2.2524465747952864, "grad_norm": 0.48935920000076294, "learning_rate": 2.6711351968054766e-06, "loss": 0.0013, "step": 5640 }, { "epoch": 2.256440982624326, "grad_norm": 0.13353575766086578, "learning_rate": 2.6568739304050202e-06, "loss": 0.0014, "step": 5650 }, { "epoch": 2.2604353904533654, "grad_norm": 0.10841682553291321, "learning_rate": 2.642612664004564e-06, "loss": 0.003, "step": 5660 }, { "epoch": 2.2644297982824044, "grad_norm": 0.18046899139881134, "learning_rate": 2.6283513976041076e-06, "loss": 0.0018, "step": 5670 }, { "epoch": 2.268424206111444, "grad_norm": 0.16014991700649261, "learning_rate": 2.614090131203651e-06, "loss": 0.0023, "step": 5680 }, { "epoch": 2.2724186139404834, "grad_norm": 0.19527053833007812, "learning_rate": 2.599828864803195e-06, "loss": 0.0019, "step": 5690 }, { "epoch": 2.276413021769523, "grad_norm": 0.2142389565706253, "learning_rate": 2.5855675984027386e-06, "loss": 0.0016, "step": 5700 }, { "epoch": 2.280407429598562, "grad_norm": 0.2971132695674896, "learning_rate": 2.571306332002282e-06, "loss": 0.0024, "step": 5710 }, { "epoch": 2.2844018374276014, "grad_norm": 0.29556071758270264, "learning_rate": 2.557045065601826e-06, "loss": 0.003, "step": 5720 }, { "epoch": 2.288396245256641, "grad_norm": 0.1596396416425705, "learning_rate": 2.542783799201369e-06, "loss": 0.0027, "step": 5730 }, { "epoch": 2.29239065308568, "grad_norm": 0.13799555599689484, "learning_rate": 2.528522532800913e-06, "loss": 0.0011, "step": 5740 }, { "epoch": 2.2963850609147194, "grad_norm": 0.28748640418052673, "learning_rate": 2.514261266400457e-06, "loss": 0.0025, "step": 5750 }, { "epoch": 2.300379468743759, "grad_norm": 0.06540987640619278, "learning_rate": 2.5e-06, "loss": 0.0027, "step": 5760 }, { "epoch": 2.304373876572798, "grad_norm": 0.2653571665287018, "learning_rate": 2.485738733599544e-06, "loss": 0.0008, "step": 5770 }, { "epoch": 2.3083682844018374, "grad_norm": 0.1664719134569168, "learning_rate": 2.4714774671990875e-06, "loss": 0.0018, "step": 5780 }, { "epoch": 2.312362692230877, "grad_norm": 0.041943565011024475, "learning_rate": 2.4572162007986312e-06, "loss": 0.0013, "step": 5790 }, { "epoch": 2.316357100059916, "grad_norm": 0.23101578652858734, "learning_rate": 2.442954934398175e-06, "loss": 0.0024, "step": 5800 }, { "epoch": 2.3203515078889554, "grad_norm": 0.5547912120819092, "learning_rate": 2.428693667997718e-06, "loss": 0.0019, "step": 5810 }, { "epoch": 2.324345915717995, "grad_norm": 1.9931162595748901, "learning_rate": 2.414432401597262e-06, "loss": 0.003, "step": 5820 }, { "epoch": 2.3283403235470344, "grad_norm": 0.2455291897058487, "learning_rate": 2.400171135196806e-06, "loss": 0.0013, "step": 5830 }, { "epoch": 2.3323347313760734, "grad_norm": 0.18468691408634186, "learning_rate": 2.3859098687963496e-06, "loss": 0.0038, "step": 5840 }, { "epoch": 2.336329139205113, "grad_norm": 0.2789057493209839, "learning_rate": 2.371648602395893e-06, "loss": 0.002, "step": 5850 }, { "epoch": 2.3403235470341524, "grad_norm": 0.3302832543849945, "learning_rate": 2.3573873359954365e-06, "loss": 0.0021, "step": 5860 }, { "epoch": 2.3443179548631914, "grad_norm": 0.3988155722618103, "learning_rate": 2.34312606959498e-06, "loss": 0.0018, "step": 5870 }, { "epoch": 2.348312362692231, "grad_norm": 0.5014130473136902, "learning_rate": 2.328864803194524e-06, "loss": 0.0017, "step": 5880 }, { "epoch": 2.3523067705212704, "grad_norm": 0.31072625517845154, "learning_rate": 2.3146035367940675e-06, "loss": 0.0023, "step": 5890 }, { "epoch": 2.3563011783503094, "grad_norm": 0.13283780217170715, "learning_rate": 2.300342270393611e-06, "loss": 0.0015, "step": 5900 }, { "epoch": 2.360295586179349, "grad_norm": 0.48449787497520447, "learning_rate": 2.286081003993155e-06, "loss": 0.0018, "step": 5910 }, { "epoch": 2.3642899940083884, "grad_norm": 0.6114631295204163, "learning_rate": 2.2718197375926985e-06, "loss": 0.0018, "step": 5920 }, { "epoch": 2.3682844018374274, "grad_norm": 0.3139302730560303, "learning_rate": 2.257558471192242e-06, "loss": 0.0011, "step": 5930 }, { "epoch": 2.372278809666467, "grad_norm": 0.6530600786209106, "learning_rate": 2.2432972047917854e-06, "loss": 0.0027, "step": 5940 }, { "epoch": 2.3762732174955064, "grad_norm": 0.30340439081192017, "learning_rate": 2.229035938391329e-06, "loss": 0.0013, "step": 5950 }, { "epoch": 2.380267625324546, "grad_norm": 0.19712309539318085, "learning_rate": 2.214774671990873e-06, "loss": 0.0014, "step": 5960 }, { "epoch": 2.384262033153585, "grad_norm": 0.4613218307495117, "learning_rate": 2.2005134055904164e-06, "loss": 0.0019, "step": 5970 }, { "epoch": 2.3882564409826244, "grad_norm": 0.3004595637321472, "learning_rate": 2.18625213918996e-06, "loss": 0.0021, "step": 5980 }, { "epoch": 2.392250848811664, "grad_norm": 0.07407541573047638, "learning_rate": 2.171990872789504e-06, "loss": 0.0019, "step": 5990 }, { "epoch": 2.396245256640703, "grad_norm": 0.16345179080963135, "learning_rate": 2.1577296063890475e-06, "loss": 0.0012, "step": 6000 }, { "epoch": 2.396245256640703, "eval_loss": 0.006617086939513683, "eval_runtime": 7437.099, "eval_samples_per_second": 2.693, "eval_steps_per_second": 0.337, "eval_wer": 0.5057546679788949, "step": 6000 }, { "epoch": 2.4002396644697424, "grad_norm": 0.5397946834564209, "learning_rate": 2.143468339988591e-06, "loss": 0.0032, "step": 6010 }, { "epoch": 2.404234072298782, "grad_norm": 0.1278027594089508, "learning_rate": 2.129207073588135e-06, "loss": 0.0011, "step": 6020 }, { "epoch": 2.408228480127821, "grad_norm": 0.34275180101394653, "learning_rate": 2.1149458071876785e-06, "loss": 0.0011, "step": 6030 }, { "epoch": 2.4122228879568604, "grad_norm": 0.25335317850112915, "learning_rate": 2.100684540787222e-06, "loss": 0.0024, "step": 6040 }, { "epoch": 2.4162172957859, "grad_norm": 0.1778995245695114, "learning_rate": 2.086423274386766e-06, "loss": 0.0024, "step": 6050 }, { "epoch": 2.420211703614939, "grad_norm": 0.10461229830980301, "learning_rate": 2.0721620079863095e-06, "loss": 0.0011, "step": 6060 }, { "epoch": 2.4242061114439784, "grad_norm": 0.10788627713918686, "learning_rate": 2.0579007415858527e-06, "loss": 0.0022, "step": 6070 }, { "epoch": 2.428200519273018, "grad_norm": 0.5851423740386963, "learning_rate": 2.043639475185397e-06, "loss": 0.0017, "step": 6080 }, { "epoch": 2.4321949271020573, "grad_norm": 0.11152782291173935, "learning_rate": 2.0293782087849405e-06, "loss": 0.0013, "step": 6090 }, { "epoch": 2.4361893349310964, "grad_norm": 0.9274498224258423, "learning_rate": 2.0151169423844837e-06, "loss": 0.0024, "step": 6100 }, { "epoch": 2.440183742760136, "grad_norm": 0.10081840306520462, "learning_rate": 2.0008556759840274e-06, "loss": 0.0023, "step": 6110 }, { "epoch": 2.4441781505891753, "grad_norm": 0.08972252160310745, "learning_rate": 1.986594409583571e-06, "loss": 0.0012, "step": 6120 }, { "epoch": 2.4481725584182144, "grad_norm": 0.32298198342323303, "learning_rate": 1.9723331431831148e-06, "loss": 0.0021, "step": 6130 }, { "epoch": 2.452166966247254, "grad_norm": 0.10888329893350601, "learning_rate": 1.9580718767826584e-06, "loss": 0.0016, "step": 6140 }, { "epoch": 2.4561613740762933, "grad_norm": 9.623656272888184, "learning_rate": 1.943810610382202e-06, "loss": 0.0059, "step": 6150 }, { "epoch": 2.4601557819053324, "grad_norm": 0.10817322880029678, "learning_rate": 1.9295493439817458e-06, "loss": 0.001, "step": 6160 }, { "epoch": 2.464150189734372, "grad_norm": 2.2221720218658447, "learning_rate": 1.9152880775812894e-06, "loss": 0.0018, "step": 6170 }, { "epoch": 2.4681445975634113, "grad_norm": 0.6658632159233093, "learning_rate": 1.9010268111808331e-06, "loss": 0.0009, "step": 6180 }, { "epoch": 2.4721390053924504, "grad_norm": 0.45321881771087646, "learning_rate": 1.8867655447803768e-06, "loss": 0.0019, "step": 6190 }, { "epoch": 2.47613341322149, "grad_norm": 0.3559039533138275, "learning_rate": 1.8725042783799202e-06, "loss": 0.0009, "step": 6200 }, { "epoch": 2.4801278210505293, "grad_norm": 0.1388717144727707, "learning_rate": 1.858243011979464e-06, "loss": 0.0018, "step": 6210 }, { "epoch": 2.484122228879569, "grad_norm": 1.1052217483520508, "learning_rate": 1.8439817455790076e-06, "loss": 0.0019, "step": 6220 }, { "epoch": 2.488116636708608, "grad_norm": 0.2527538239955902, "learning_rate": 1.829720479178551e-06, "loss": 0.0017, "step": 6230 }, { "epoch": 2.4921110445376473, "grad_norm": 0.21638405323028564, "learning_rate": 1.815459212778095e-06, "loss": 0.0016, "step": 6240 }, { "epoch": 2.496105452366687, "grad_norm": 0.2525235414505005, "learning_rate": 1.8011979463776386e-06, "loss": 0.0011, "step": 6250 }, { "epoch": 2.500099860195726, "grad_norm": 0.204837828874588, "learning_rate": 1.786936679977182e-06, "loss": 0.002, "step": 6260 }, { "epoch": 2.5040942680247653, "grad_norm": 0.3697673976421356, "learning_rate": 1.7726754135767257e-06, "loss": 0.0035, "step": 6270 }, { "epoch": 2.508088675853805, "grad_norm": 1.4454883337020874, "learning_rate": 1.7584141471762694e-06, "loss": 0.0019, "step": 6280 }, { "epoch": 2.5120830836828443, "grad_norm": 0.2247629016637802, "learning_rate": 1.7441528807758129e-06, "loss": 0.0011, "step": 6290 }, { "epoch": 2.5160774915118833, "grad_norm": 0.12582184374332428, "learning_rate": 1.7298916143753565e-06, "loss": 0.0011, "step": 6300 }, { "epoch": 2.520071899340923, "grad_norm": 0.09849651902914047, "learning_rate": 1.7156303479749004e-06, "loss": 0.0015, "step": 6310 }, { "epoch": 2.524066307169962, "grad_norm": 0.1161109134554863, "learning_rate": 1.701369081574444e-06, "loss": 0.0024, "step": 6320 }, { "epoch": 2.5280607149990013, "grad_norm": 0.13831503689289093, "learning_rate": 1.6871078151739875e-06, "loss": 0.0017, "step": 6330 }, { "epoch": 2.532055122828041, "grad_norm": 0.19472339749336243, "learning_rate": 1.6728465487735312e-06, "loss": 0.0033, "step": 6340 }, { "epoch": 2.5360495306570803, "grad_norm": 0.4595012664794922, "learning_rate": 1.6585852823730749e-06, "loss": 0.0015, "step": 6350 }, { "epoch": 2.5400439384861193, "grad_norm": 0.09621255099773407, "learning_rate": 1.6443240159726183e-06, "loss": 0.0033, "step": 6360 }, { "epoch": 2.544038346315159, "grad_norm": 1.4646393060684204, "learning_rate": 1.6300627495721622e-06, "loss": 0.0017, "step": 6370 }, { "epoch": 2.5480327541441983, "grad_norm": 0.16254696249961853, "learning_rate": 1.615801483171706e-06, "loss": 0.0016, "step": 6380 }, { "epoch": 2.5520271619732373, "grad_norm": 0.16140027344226837, "learning_rate": 1.6015402167712494e-06, "loss": 0.0013, "step": 6390 }, { "epoch": 2.556021569802277, "grad_norm": 0.14379070699214935, "learning_rate": 1.587278950370793e-06, "loss": 0.0027, "step": 6400 }, { "epoch": 2.5600159776313163, "grad_norm": 0.18138810992240906, "learning_rate": 1.5730176839703367e-06, "loss": 0.0008, "step": 6410 }, { "epoch": 2.5640103854603558, "grad_norm": 0.048751723021268845, "learning_rate": 1.5587564175698802e-06, "loss": 0.004, "step": 6420 }, { "epoch": 2.568004793289395, "grad_norm": 0.16989260911941528, "learning_rate": 1.544495151169424e-06, "loss": 0.0014, "step": 6430 }, { "epoch": 2.5719992011184343, "grad_norm": 0.29287806153297424, "learning_rate": 1.5302338847689677e-06, "loss": 0.0022, "step": 6440 }, { "epoch": 2.5759936089474733, "grad_norm": 0.12781338393688202, "learning_rate": 1.5159726183685112e-06, "loss": 0.0013, "step": 6450 }, { "epoch": 2.579988016776513, "grad_norm": 0.21234439313411713, "learning_rate": 1.5017113519680548e-06, "loss": 0.0013, "step": 6460 }, { "epoch": 2.5839824246055523, "grad_norm": 0.5938289761543274, "learning_rate": 1.4874500855675985e-06, "loss": 0.0012, "step": 6470 }, { "epoch": 2.5879768324345918, "grad_norm": 0.5471766591072083, "learning_rate": 1.4731888191671424e-06, "loss": 0.0026, "step": 6480 }, { "epoch": 2.591971240263631, "grad_norm": 0.23329491913318634, "learning_rate": 1.4589275527666856e-06, "loss": 0.0015, "step": 6490 }, { "epoch": 2.5959656480926703, "grad_norm": 1.038935899734497, "learning_rate": 1.4446662863662295e-06, "loss": 0.0015, "step": 6500 }, { "epoch": 2.5999600559217098, "grad_norm": 0.16698375344276428, "learning_rate": 1.4304050199657732e-06, "loss": 0.0013, "step": 6510 }, { "epoch": 2.603954463750749, "grad_norm": 0.10930318385362625, "learning_rate": 1.4161437535653167e-06, "loss": 0.0024, "step": 6520 }, { "epoch": 2.6079488715797883, "grad_norm": 0.24214453995227814, "learning_rate": 1.4018824871648603e-06, "loss": 0.0026, "step": 6530 }, { "epoch": 2.6119432794088278, "grad_norm": 0.1356784552335739, "learning_rate": 1.387621220764404e-06, "loss": 0.002, "step": 6540 }, { "epoch": 2.6159376872378672, "grad_norm": 0.10308735817670822, "learning_rate": 1.3733599543639475e-06, "loss": 0.0031, "step": 6550 }, { "epoch": 2.6199320950669063, "grad_norm": 0.5427199006080627, "learning_rate": 1.3590986879634913e-06, "loss": 0.0033, "step": 6560 }, { "epoch": 2.6239265028959458, "grad_norm": 0.4210723340511322, "learning_rate": 1.344837421563035e-06, "loss": 0.001, "step": 6570 }, { "epoch": 2.627920910724985, "grad_norm": 0.14636996388435364, "learning_rate": 1.3305761551625785e-06, "loss": 0.0026, "step": 6580 }, { "epoch": 2.6319153185540243, "grad_norm": 0.14456219971179962, "learning_rate": 1.3163148887621221e-06, "loss": 0.0014, "step": 6590 }, { "epoch": 2.6359097263830638, "grad_norm": 0.206597238779068, "learning_rate": 1.3020536223616658e-06, "loss": 0.0012, "step": 6600 }, { "epoch": 2.6399041342121032, "grad_norm": 0.11497008800506592, "learning_rate": 1.2877923559612093e-06, "loss": 0.0016, "step": 6610 }, { "epoch": 2.6438985420411423, "grad_norm": 0.18552027642726898, "learning_rate": 1.2735310895607532e-06, "loss": 0.0037, "step": 6620 }, { "epoch": 2.6478929498701818, "grad_norm": 0.35586246848106384, "learning_rate": 1.2592698231602968e-06, "loss": 0.0018, "step": 6630 }, { "epoch": 2.6518873576992212, "grad_norm": 0.39738965034484863, "learning_rate": 1.2450085567598403e-06, "loss": 0.0026, "step": 6640 }, { "epoch": 2.6558817655282603, "grad_norm": 0.4074229300022125, "learning_rate": 1.230747290359384e-06, "loss": 0.0012, "step": 6650 }, { "epoch": 2.6598761733572998, "grad_norm": 0.08064166456460953, "learning_rate": 1.2164860239589276e-06, "loss": 0.0025, "step": 6660 }, { "epoch": 2.6638705811863392, "grad_norm": 0.8385900855064392, "learning_rate": 1.2022247575584713e-06, "loss": 0.0014, "step": 6670 }, { "epoch": 2.6678649890153787, "grad_norm": 0.24224676191806793, "learning_rate": 1.187963491158015e-06, "loss": 0.0024, "step": 6680 }, { "epoch": 2.6718593968444178, "grad_norm": 0.11100082844495773, "learning_rate": 1.1737022247575586e-06, "loss": 0.0015, "step": 6690 }, { "epoch": 2.6758538046734572, "grad_norm": 0.49067580699920654, "learning_rate": 1.159440958357102e-06, "loss": 0.0017, "step": 6700 }, { "epoch": 2.6798482125024963, "grad_norm": 0.17083971202373505, "learning_rate": 1.145179691956646e-06, "loss": 0.0034, "step": 6710 }, { "epoch": 2.6838426203315358, "grad_norm": 0.16969099640846252, "learning_rate": 1.1309184255561894e-06, "loss": 0.0014, "step": 6720 }, { "epoch": 2.6878370281605752, "grad_norm": 0.9469853043556213, "learning_rate": 1.1166571591557331e-06, "loss": 0.0024, "step": 6730 }, { "epoch": 2.6918314359896147, "grad_norm": 0.0797346830368042, "learning_rate": 1.1023958927552768e-06, "loss": 0.0026, "step": 6740 }, { "epoch": 2.6958258438186538, "grad_norm": 0.1065623015165329, "learning_rate": 1.0881346263548204e-06, "loss": 0.0012, "step": 6750 }, { "epoch": 2.6998202516476932, "grad_norm": 1.0359104871749878, "learning_rate": 1.0738733599543641e-06, "loss": 0.0027, "step": 6760 }, { "epoch": 2.7038146594767327, "grad_norm": 0.16578173637390137, "learning_rate": 1.0596120935539076e-06, "loss": 0.003, "step": 6770 }, { "epoch": 2.7078090673057718, "grad_norm": 0.04761478304862976, "learning_rate": 1.0453508271534513e-06, "loss": 0.0024, "step": 6780 }, { "epoch": 2.7118034751348112, "grad_norm": 1.9014323949813843, "learning_rate": 1.031089560752995e-06, "loss": 0.0023, "step": 6790 }, { "epoch": 2.7157978829638507, "grad_norm": 0.3101843297481537, "learning_rate": 1.0168282943525386e-06, "loss": 0.0021, "step": 6800 }, { "epoch": 2.71979229079289, "grad_norm": 0.282959520816803, "learning_rate": 1.0025670279520823e-06, "loss": 0.0018, "step": 6810 }, { "epoch": 2.7237866986219292, "grad_norm": 0.06943622976541519, "learning_rate": 9.88305761551626e-07, "loss": 0.0012, "step": 6820 }, { "epoch": 2.7277811064509687, "grad_norm": 0.3266574740409851, "learning_rate": 9.740444951511694e-07, "loss": 0.0019, "step": 6830 }, { "epoch": 2.7317755142800078, "grad_norm": 0.3858787715435028, "learning_rate": 9.597832287507133e-07, "loss": 0.0019, "step": 6840 }, { "epoch": 2.7357699221090472, "grad_norm": 0.28762754797935486, "learning_rate": 9.455219623502567e-07, "loss": 0.0013, "step": 6850 }, { "epoch": 2.7397643299380867, "grad_norm": 0.1022636890411377, "learning_rate": 9.312606959498004e-07, "loss": 0.0015, "step": 6860 }, { "epoch": 2.743758737767126, "grad_norm": 0.8769612908363342, "learning_rate": 9.169994295493441e-07, "loss": 0.0018, "step": 6870 }, { "epoch": 2.7477531455961652, "grad_norm": 0.20930863916873932, "learning_rate": 9.027381631488876e-07, "loss": 0.0018, "step": 6880 }, { "epoch": 2.7517475534252047, "grad_norm": 0.08737757056951523, "learning_rate": 8.884768967484313e-07, "loss": 0.0012, "step": 6890 }, { "epoch": 2.755741961254244, "grad_norm": 0.20543090999126434, "learning_rate": 8.74215630347975e-07, "loss": 0.001, "step": 6900 }, { "epoch": 2.7597363690832832, "grad_norm": 0.21611656248569489, "learning_rate": 8.599543639475185e-07, "loss": 0.0038, "step": 6910 }, { "epoch": 2.7637307769123227, "grad_norm": 0.1640675812959671, "learning_rate": 8.456930975470623e-07, "loss": 0.0027, "step": 6920 }, { "epoch": 2.767725184741362, "grad_norm": 0.31564095616340637, "learning_rate": 8.314318311466059e-07, "loss": 0.001, "step": 6930 }, { "epoch": 2.7717195925704017, "grad_norm": 1.0901868343353271, "learning_rate": 8.171705647461495e-07, "loss": 0.0017, "step": 6940 }, { "epoch": 2.7757140003994407, "grad_norm": 0.3010713458061218, "learning_rate": 8.029092983456932e-07, "loss": 0.001, "step": 6950 }, { "epoch": 2.77970840822848, "grad_norm": 0.04908519238233566, "learning_rate": 7.886480319452368e-07, "loss": 0.0034, "step": 6960 }, { "epoch": 2.7837028160575192, "grad_norm": 0.07861047983169556, "learning_rate": 7.743867655447804e-07, "loss": 0.003, "step": 6970 }, { "epoch": 2.7876972238865587, "grad_norm": 0.9499324560165405, "learning_rate": 7.601254991443241e-07, "loss": 0.0016, "step": 6980 }, { "epoch": 2.791691631715598, "grad_norm": 0.11067607998847961, "learning_rate": 7.458642327438677e-07, "loss": 0.002, "step": 6990 }, { "epoch": 2.7956860395446377, "grad_norm": 0.23731118440628052, "learning_rate": 7.316029663434114e-07, "loss": 0.0021, "step": 7000 }, { "epoch": 2.7996804473736767, "grad_norm": 0.5374778509140015, "learning_rate": 7.17341699942955e-07, "loss": 0.0013, "step": 7010 }, { "epoch": 2.803674855202716, "grad_norm": 0.13841059803962708, "learning_rate": 7.030804335424986e-07, "loss": 0.0036, "step": 7020 }, { "epoch": 2.8076692630317557, "grad_norm": 0.13221804797649384, "learning_rate": 6.888191671420423e-07, "loss": 0.0013, "step": 7030 }, { "epoch": 2.8116636708607947, "grad_norm": 0.20090872049331665, "learning_rate": 6.745579007415858e-07, "loss": 0.0017, "step": 7040 }, { "epoch": 2.815658078689834, "grad_norm": 0.06092801317572594, "learning_rate": 6.602966343411295e-07, "loss": 0.0015, "step": 7050 }, { "epoch": 2.8196524865188737, "grad_norm": 0.7219746112823486, "learning_rate": 6.460353679406732e-07, "loss": 0.0014, "step": 7060 }, { "epoch": 2.823646894347913, "grad_norm": 0.11110103875398636, "learning_rate": 6.317741015402168e-07, "loss": 0.001, "step": 7070 }, { "epoch": 2.827641302176952, "grad_norm": 0.09267082810401917, "learning_rate": 6.175128351397604e-07, "loss": 0.0013, "step": 7080 }, { "epoch": 2.8316357100059917, "grad_norm": 0.19514226913452148, "learning_rate": 6.032515687393041e-07, "loss": 0.0027, "step": 7090 }, { "epoch": 2.8356301178350307, "grad_norm": 0.09698819369077682, "learning_rate": 5.889903023388478e-07, "loss": 0.0034, "step": 7100 }, { "epoch": 2.83962452566407, "grad_norm": 1.0999977588653564, "learning_rate": 5.747290359383914e-07, "loss": 0.0016, "step": 7110 }, { "epoch": 2.8436189334931097, "grad_norm": 0.27861669659614563, "learning_rate": 5.60467769537935e-07, "loss": 0.0021, "step": 7120 }, { "epoch": 2.847613341322149, "grad_norm": 0.6540320515632629, "learning_rate": 5.462065031374787e-07, "loss": 0.0027, "step": 7130 }, { "epoch": 2.851607749151188, "grad_norm": 0.719205379486084, "learning_rate": 5.319452367370223e-07, "loss": 0.0009, "step": 7140 }, { "epoch": 2.8556021569802277, "grad_norm": 0.08803220093250275, "learning_rate": 5.17683970336566e-07, "loss": 0.0016, "step": 7150 }, { "epoch": 2.859596564809267, "grad_norm": 0.4269472658634186, "learning_rate": 5.034227039361096e-07, "loss": 0.0024, "step": 7160 }, { "epoch": 2.863590972638306, "grad_norm": 1.9801479578018188, "learning_rate": 4.891614375356533e-07, "loss": 0.0021, "step": 7170 }, { "epoch": 2.8675853804673457, "grad_norm": 0.14175833761692047, "learning_rate": 4.7490017113519687e-07, "loss": 0.001, "step": 7180 }, { "epoch": 2.871579788296385, "grad_norm": 0.3901681900024414, "learning_rate": 4.6063890473474054e-07, "loss": 0.0024, "step": 7190 }, { "epoch": 2.8755741961254246, "grad_norm": 0.20490513741970062, "learning_rate": 4.463776383342841e-07, "loss": 0.002, "step": 7200 }, { "epoch": 2.8795686039544637, "grad_norm": 0.06986338645219803, "learning_rate": 4.321163719338278e-07, "loss": 0.0011, "step": 7210 }, { "epoch": 2.883563011783503, "grad_norm": 0.2736944258213043, "learning_rate": 4.178551055333714e-07, "loss": 0.0015, "step": 7220 }, { "epoch": 2.887557419612542, "grad_norm": 0.39138877391815186, "learning_rate": 4.0359383913291507e-07, "loss": 0.0048, "step": 7230 }, { "epoch": 2.8915518274415817, "grad_norm": 0.3493710160255432, "learning_rate": 3.8933257273245863e-07, "loss": 0.0014, "step": 7240 }, { "epoch": 2.895546235270621, "grad_norm": 0.47019582986831665, "learning_rate": 3.750713063320023e-07, "loss": 0.0025, "step": 7250 }, { "epoch": 2.8995406430996606, "grad_norm": 0.09903858602046967, "learning_rate": 3.6081003993154597e-07, "loss": 0.0018, "step": 7260 }, { "epoch": 2.9035350509286997, "grad_norm": 0.13944830000400543, "learning_rate": 3.4654877353108964e-07, "loss": 0.0013, "step": 7270 }, { "epoch": 2.907529458757739, "grad_norm": 0.6667160987854004, "learning_rate": 3.322875071306332e-07, "loss": 0.0013, "step": 7280 }, { "epoch": 2.9115238665867786, "grad_norm": 0.22438709437847137, "learning_rate": 3.180262407301769e-07, "loss": 0.0011, "step": 7290 }, { "epoch": 2.9155182744158177, "grad_norm": 0.07875990122556686, "learning_rate": 3.037649743297205e-07, "loss": 0.0019, "step": 7300 }, { "epoch": 2.919512682244857, "grad_norm": 1.9263901710510254, "learning_rate": 2.895037079292641e-07, "loss": 0.0019, "step": 7310 }, { "epoch": 2.9235070900738966, "grad_norm": 0.09846038371324539, "learning_rate": 2.752424415288078e-07, "loss": 0.0019, "step": 7320 }, { "epoch": 2.927501497902936, "grad_norm": 0.10731885582208633, "learning_rate": 2.609811751283514e-07, "loss": 0.0012, "step": 7330 }, { "epoch": 2.931495905731975, "grad_norm": 0.0784291997551918, "learning_rate": 2.467199087278951e-07, "loss": 0.0017, "step": 7340 }, { "epoch": 2.9354903135610146, "grad_norm": 0.05892806872725487, "learning_rate": 2.324586423274387e-07, "loss": 0.0007, "step": 7350 }, { "epoch": 2.9394847213900537, "grad_norm": 1.340722918510437, "learning_rate": 2.1819737592698234e-07, "loss": 0.0037, "step": 7360 }, { "epoch": 2.943479129219093, "grad_norm": 0.2646709382534027, "learning_rate": 2.0393610952652596e-07, "loss": 0.0015, "step": 7370 }, { "epoch": 2.9474735370481326, "grad_norm": 0.35851484537124634, "learning_rate": 1.8967484312606963e-07, "loss": 0.0009, "step": 7380 }, { "epoch": 2.951467944877172, "grad_norm": 0.664960503578186, "learning_rate": 1.7541357672561324e-07, "loss": 0.0009, "step": 7390 }, { "epoch": 2.955462352706211, "grad_norm": 0.14400076866149902, "learning_rate": 1.611523103251569e-07, "loss": 0.0012, "step": 7400 }, { "epoch": 2.9594567605352506, "grad_norm": 0.0934806764125824, "learning_rate": 1.4689104392470053e-07, "loss": 0.0018, "step": 7410 }, { "epoch": 2.96345116836429, "grad_norm": 0.12735722959041595, "learning_rate": 1.3262977752424415e-07, "loss": 0.0025, "step": 7420 }, { "epoch": 2.967445576193329, "grad_norm": 0.819703221321106, "learning_rate": 1.183685111237878e-07, "loss": 0.0014, "step": 7430 }, { "epoch": 2.9714399840223686, "grad_norm": 0.10716011375188828, "learning_rate": 1.0410724472333144e-07, "loss": 0.0012, "step": 7440 }, { "epoch": 2.975434391851408, "grad_norm": 0.9824792742729187, "learning_rate": 8.984597832287507e-08, "loss": 0.0023, "step": 7450 }, { "epoch": 2.9794287996804476, "grad_norm": 0.20575283467769623, "learning_rate": 7.558471192241872e-08, "loss": 0.0019, "step": 7460 }, { "epoch": 2.9834232075094866, "grad_norm": 0.10139793157577515, "learning_rate": 6.132344552196236e-08, "loss": 0.004, "step": 7470 }, { "epoch": 2.987417615338526, "grad_norm": 0.4164970815181732, "learning_rate": 4.7062179121506e-08, "loss": 0.0012, "step": 7480 }, { "epoch": 2.991412023167565, "grad_norm": 0.10282018035650253, "learning_rate": 3.280091272104963e-08, "loss": 0.0019, "step": 7490 }, { "epoch": 2.9954064309966046, "grad_norm": 0.08646287769079208, "learning_rate": 1.853964632059327e-08, "loss": 0.001, "step": 7500 }, { "epoch": 2.999400838825644, "grad_norm": 0.25076162815093994, "learning_rate": 4.278379920136909e-09, "loss": 0.0015, "step": 7510 } ], "logging_steps": 10, "max_steps": 7512, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 6.9347071696896e+19, "train_batch_size": 16, "trial_name": null, "trial_params": null }