error577's picture
Training in progress, step 1600, checkpoint
87fb642 verified
Raw History Blame
283 kB
{
"best_global_step": 1600,
"best_metric": 0.0074059851467609406,
"best_model_checkpoint": "miner_id_24/checkpoint-1600",
"epoch": 0.3352364988738149,
"eval_steps": 200,
"global_step": 1600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0002095228117961343,
"grad_norm": 6.6040472984313965,
"learning_rate": 2.0000000000000003e-06,
"loss": 12.5068,
"step": 1
},
{
"epoch": 0.0002095228117961343,
"eval_loss": 12.532599449157715,
"eval_runtime": 89.0598,
"eval_samples_per_second": 8.623,
"eval_steps_per_second": 4.312,
"step": 1
},
{
"epoch": 0.0004190456235922686,
"grad_norm": 10.516180038452148,
"learning_rate": 4.000000000000001e-06,
"loss": 12.4956,
"step": 2
},
{
"epoch": 0.0006285684353884029,
"grad_norm": 5.9784955978393555,
"learning_rate": 6e-06,
"loss": 12.4652,
"step": 3
},
{
"epoch": 0.0008380912471845372,
"grad_norm": 4.773026943206787,
"learning_rate": 8.000000000000001e-06,
"loss": 12.5635,
"step": 4
},
{
"epoch": 0.0010476140589806716,
"grad_norm": 4.345443248748779,
"learning_rate": 1e-05,
"loss": 12.4502,
"step": 5
},
{
"epoch": 0.0012571368707768058,
"grad_norm": 7.477049350738525,
"learning_rate": 1.2e-05,
"loss": 12.5031,
"step": 6
},
{
"epoch": 0.0014666596825729402,
"grad_norm": 6.29019832611084,
"learning_rate": 1.4000000000000001e-05,
"loss": 12.5355,
"step": 7
},
{
"epoch": 0.0016761824943690743,
"grad_norm": 7.278188228607178,
"learning_rate": 1.6000000000000003e-05,
"loss": 12.382,
"step": 8
},
{
"epoch": 0.0018857053061652087,
"grad_norm": 11.482406616210938,
"learning_rate": 1.8e-05,
"loss": 12.224,
"step": 9
},
{
"epoch": 0.002095228117961343,
"grad_norm": 76.79756927490234,
"learning_rate": 2e-05,
"loss": 12.1469,
"step": 10
},
{
"epoch": 0.0023047509297574775,
"grad_norm": 3.686225414276123,
"learning_rate": 2.2000000000000003e-05,
"loss": 12.0783,
"step": 11
},
{
"epoch": 0.0025142737415536115,
"grad_norm": 6.081927299499512,
"learning_rate": 2.4e-05,
"loss": 11.9893,
"step": 12
},
{
"epoch": 0.002723796553349746,
"grad_norm": 3.427201747894287,
"learning_rate": 2.6000000000000002e-05,
"loss": 11.9078,
"step": 13
},
{
"epoch": 0.0029333193651458803,
"grad_norm": 16.03325080871582,
"learning_rate": 2.8000000000000003e-05,
"loss": 11.8939,
"step": 14
},
{
"epoch": 0.0031428421769420147,
"grad_norm": 4.581820487976074,
"learning_rate": 3e-05,
"loss": 11.8592,
"step": 15
},
{
"epoch": 0.0033523649887381487,
"grad_norm": 6.6432342529296875,
"learning_rate": 3.2000000000000005e-05,
"loss": 11.8373,
"step": 16
},
{
"epoch": 0.003561887800534283,
"grad_norm": 61.577571868896484,
"learning_rate": 3.4000000000000007e-05,
"loss": 11.7898,
"step": 17
},
{
"epoch": 0.0037714106123304175,
"grad_norm": 9.466367721557617,
"learning_rate": 3.6e-05,
"loss": 11.7693,
"step": 18
},
{
"epoch": 0.0039809334241265515,
"grad_norm": 1.1112542152404785,
"learning_rate": 3.8e-05,
"loss": 11.7571,
"step": 19
},
{
"epoch": 0.004190456235922686,
"grad_norm": 1.6447243690490723,
"learning_rate": 4e-05,
"loss": 11.7253,
"step": 20
},
{
"epoch": 0.00439997904771882,
"grad_norm": 14.649271965026855,
"learning_rate": 4.2e-05,
"loss": 11.7103,
"step": 21
},
{
"epoch": 0.004609501859514955,
"grad_norm": 1.7011709213256836,
"learning_rate": 4.4000000000000006e-05,
"loss": 11.6808,
"step": 22
},
{
"epoch": 0.004819024671311089,
"grad_norm": 3.6001105308532715,
"learning_rate": 4.600000000000001e-05,
"loss": 11.631,
"step": 23
},
{
"epoch": 0.005028547483107223,
"grad_norm": 12.185075759887695,
"learning_rate": 4.8e-05,
"loss": 11.5698,
"step": 24
},
{
"epoch": 0.005238070294903358,
"grad_norm": 5.6634955406188965,
"learning_rate": 5e-05,
"loss": 11.4747,
"step": 25
},
{
"epoch": 0.005447593106699492,
"grad_norm": 3.705836057662964,
"learning_rate": 5.2000000000000004e-05,
"loss": 11.5002,
"step": 26
},
{
"epoch": 0.005657115918495626,
"grad_norm": 66.66976928710938,
"learning_rate": 5.4000000000000005e-05,
"loss": 11.3843,
"step": 27
},
{
"epoch": 0.005866638730291761,
"grad_norm": 4.706162452697754,
"learning_rate": 5.6000000000000006e-05,
"loss": 11.209,
"step": 28
},
{
"epoch": 0.006076161542087895,
"grad_norm": 4.608543872833252,
"learning_rate": 5.8e-05,
"loss": 11.0922,
"step": 29
},
{
"epoch": 0.0062856843538840294,
"grad_norm": 21.13187599182129,
"learning_rate": 6e-05,
"loss": 10.78,
"step": 30
},
{
"epoch": 0.006495207165680163,
"grad_norm": 5.859175682067871,
"learning_rate": 6.2e-05,
"loss": 10.3626,
"step": 31
},
{
"epoch": 0.006704729977476297,
"grad_norm": 10.68819808959961,
"learning_rate": 6.400000000000001e-05,
"loss": 10.3803,
"step": 32
},
{
"epoch": 0.006914252789272432,
"grad_norm": 4.8769941329956055,
"learning_rate": 6.6e-05,
"loss": 10.0995,
"step": 33
},
{
"epoch": 0.007123775601068566,
"grad_norm": 6.648291110992432,
"learning_rate": 6.800000000000001e-05,
"loss": 9.928,
"step": 34
},
{
"epoch": 0.007333298412864701,
"grad_norm": 50.16209411621094,
"learning_rate": 7e-05,
"loss": 9.8941,
"step": 35
},
{
"epoch": 0.007542821224660835,
"grad_norm": 6.54470682144165,
"learning_rate": 7.2e-05,
"loss": 9.8453,
"step": 36
},
{
"epoch": 0.007752344036456969,
"grad_norm": 8.670441627502441,
"learning_rate": 7.4e-05,
"loss": 9.5793,
"step": 37
},
{
"epoch": 0.007961866848253103,
"grad_norm": 123.02629852294922,
"learning_rate": 7.6e-05,
"loss": 9.3425,
"step": 38
},
{
"epoch": 0.008171389660049239,
"grad_norm": 14.480856895446777,
"learning_rate": 7.800000000000001e-05,
"loss": 9.3416,
"step": 39
},
{
"epoch": 0.008380912471845373,
"grad_norm": 21.085674285888672,
"learning_rate": 8e-05,
"loss": 9.0526,
"step": 40
},
{
"epoch": 0.008590435283641507,
"grad_norm": 19.711458206176758,
"learning_rate": 8.2e-05,
"loss": 9.0394,
"step": 41
},
{
"epoch": 0.00879995809543764,
"grad_norm": 4.5215630531311035,
"learning_rate": 8.4e-05,
"loss": 8.7826,
"step": 42
},
{
"epoch": 0.009009480907233774,
"grad_norm": 21.183956146240234,
"learning_rate": 8.6e-05,
"loss": 8.6163,
"step": 43
},
{
"epoch": 0.00921900371902991,
"grad_norm": 39.6319465637207,
"learning_rate": 8.800000000000001e-05,
"loss": 8.6326,
"step": 44
},
{
"epoch": 0.009428526530826044,
"grad_norm": 13.957026481628418,
"learning_rate": 9e-05,
"loss": 8.8564,
"step": 45
},
{
"epoch": 0.009638049342622178,
"grad_norm": 35.4068489074707,
"learning_rate": 9.200000000000001e-05,
"loss": 8.5426,
"step": 46
},
{
"epoch": 0.009847572154418312,
"grad_norm": 8.321797370910645,
"learning_rate": 9.4e-05,
"loss": 8.4767,
"step": 47
},
{
"epoch": 0.010057094966214446,
"grad_norm": 49.62624740600586,
"learning_rate": 9.6e-05,
"loss": 8.2416,
"step": 48
},
{
"epoch": 0.010266617778010582,
"grad_norm": 62.95499038696289,
"learning_rate": 9.8e-05,
"loss": 7.942,
"step": 49
},
{
"epoch": 0.010476140589806716,
"grad_norm": 38.04218292236328,
"learning_rate": 0.0001,
"loss": 7.7725,
"step": 50
},
{
"epoch": 0.01068566340160285,
"grad_norm": 12.888092994689941,
"learning_rate": 0.00010200000000000001,
"loss": 8.1733,
"step": 51
},
{
"epoch": 0.010895186213398984,
"grad_norm": 8.064024925231934,
"learning_rate": 0.00010400000000000001,
"loss": 7.6733,
"step": 52
},
{
"epoch": 0.011104709025195118,
"grad_norm": 37.64444351196289,
"learning_rate": 0.00010600000000000002,
"loss": 6.8757,
"step": 53
},
{
"epoch": 0.011314231836991252,
"grad_norm": 12.460424423217773,
"learning_rate": 0.00010800000000000001,
"loss": 6.8061,
"step": 54
},
{
"epoch": 0.011523754648787387,
"grad_norm": 4.761670112609863,
"learning_rate": 0.00011000000000000002,
"loss": 6.3257,
"step": 55
},
{
"epoch": 0.011733277460583521,
"grad_norm": 10.131464958190918,
"learning_rate": 0.00011200000000000001,
"loss": 6.1174,
"step": 56
},
{
"epoch": 0.011942800272379655,
"grad_norm": 4.439296722412109,
"learning_rate": 0.00011399999999999999,
"loss": 6.1032,
"step": 57
},
{
"epoch": 0.01215232308417579,
"grad_norm": 102.8046875,
"learning_rate": 0.000116,
"loss": 5.5948,
"step": 58
},
{
"epoch": 0.012361845895971923,
"grad_norm": 9.46191692352295,
"learning_rate": 0.000118,
"loss": 5.7645,
"step": 59
},
{
"epoch": 0.012571368707768059,
"grad_norm": 10.870979309082031,
"learning_rate": 0.00012,
"loss": 5.0875,
"step": 60
},
{
"epoch": 0.012780891519564193,
"grad_norm": 23.203157424926758,
"learning_rate": 0.000122,
"loss": 5.6791,
"step": 61
},
{
"epoch": 0.012990414331360327,
"grad_norm": 4.249388217926025,
"learning_rate": 0.000124,
"loss": 5.1933,
"step": 62
},
{
"epoch": 0.01319993714315646,
"grad_norm": 19.030372619628906,
"learning_rate": 0.000126,
"loss": 4.7766,
"step": 63
},
{
"epoch": 0.013409459954952595,
"grad_norm": 4.8991217613220215,
"learning_rate": 0.00012800000000000002,
"loss": 3.941,
"step": 64
},
{
"epoch": 0.01361898276674873,
"grad_norm": 6.8061933517456055,
"learning_rate": 0.00013000000000000002,
"loss": 3.8998,
"step": 65
},
{
"epoch": 0.013828505578544864,
"grad_norm": 4.572474002838135,
"learning_rate": 0.000132,
"loss": 3.7389,
"step": 66
},
{
"epoch": 0.014038028390340998,
"grad_norm": 5.179816246032715,
"learning_rate": 0.000134,
"loss": 3.5838,
"step": 67
},
{
"epoch": 0.014247551202137132,
"grad_norm": 4.74563455581665,
"learning_rate": 0.00013600000000000003,
"loss": 3.4106,
"step": 68
},
{
"epoch": 0.014457074013933266,
"grad_norm": 3.925071954727173,
"learning_rate": 0.000138,
"loss": 3.4364,
"step": 69
},
{
"epoch": 0.014666596825729402,
"grad_norm": 3.9857025146484375,
"learning_rate": 0.00014,
"loss": 2.8864,
"step": 70
},
{
"epoch": 0.014876119637525536,
"grad_norm": 3.004260778427124,
"learning_rate": 0.000142,
"loss": 3.3081,
"step": 71
},
{
"epoch": 0.01508564244932167,
"grad_norm": 3.27565598487854,
"learning_rate": 0.000144,
"loss": 3.302,
"step": 72
},
{
"epoch": 0.015295165261117804,
"grad_norm": 3.430575132369995,
"learning_rate": 0.000146,
"loss": 2.8037,
"step": 73
},
{
"epoch": 0.015504688072913938,
"grad_norm": 3.178133964538574,
"learning_rate": 0.000148,
"loss": 2.5663,
"step": 74
},
{
"epoch": 0.015714210884710072,
"grad_norm": 2.6925830841064453,
"learning_rate": 0.00015000000000000001,
"loss": 2.0117,
"step": 75
},
{
"epoch": 0.015923733696506206,
"grad_norm": 3.363239049911499,
"learning_rate": 0.000152,
"loss": 2.235,
"step": 76
},
{
"epoch": 0.01613325650830234,
"grad_norm": 3.3394203186035156,
"learning_rate": 0.000154,
"loss": 1.8811,
"step": 77
},
{
"epoch": 0.016342779320098477,
"grad_norm": 2.964437484741211,
"learning_rate": 0.00015600000000000002,
"loss": 2.1085,
"step": 78
},
{
"epoch": 0.01655230213189461,
"grad_norm": 5.256375789642334,
"learning_rate": 0.00015800000000000002,
"loss": 2.2984,
"step": 79
},
{
"epoch": 0.016761824943690745,
"grad_norm": 4.152324199676514,
"learning_rate": 0.00016,
"loss": 1.754,
"step": 80
},
{
"epoch": 0.01697134775548688,
"grad_norm": 2.9110007286071777,
"learning_rate": 0.000162,
"loss": 1.9343,
"step": 81
},
{
"epoch": 0.017180870567283013,
"grad_norm": 3.011155605316162,
"learning_rate": 0.000164,
"loss": 1.7661,
"step": 82
},
{
"epoch": 0.017390393379079147,
"grad_norm": 2.766448736190796,
"learning_rate": 0.000166,
"loss": 1.6191,
"step": 83
},
{
"epoch": 0.01759991619087528,
"grad_norm": 2.8244659900665283,
"learning_rate": 0.000168,
"loss": 1.2594,
"step": 84
},
{
"epoch": 0.017809439002671415,
"grad_norm": 3.7324302196502686,
"learning_rate": 0.00017,
"loss": 1.9778,
"step": 85
},
{
"epoch": 0.01801896181446755,
"grad_norm": 4.044966220855713,
"learning_rate": 0.000172,
"loss": 1.8828,
"step": 86
},
{
"epoch": 0.018228484626263683,
"grad_norm": 3.119807481765747,
"learning_rate": 0.000174,
"loss": 1.5972,
"step": 87
},
{
"epoch": 0.01843800743805982,
"grad_norm": 2.9005658626556396,
"learning_rate": 0.00017600000000000002,
"loss": 1.2696,
"step": 88
},
{
"epoch": 0.018647530249855954,
"grad_norm": 2.5967376232147217,
"learning_rate": 0.00017800000000000002,
"loss": 1.7584,
"step": 89
},
{
"epoch": 0.01885705306165209,
"grad_norm": 2.364287853240967,
"learning_rate": 0.00018,
"loss": 1.2882,
"step": 90
},
{
"epoch": 0.019066575873448222,
"grad_norm": 3.401552438735962,
"learning_rate": 0.000182,
"loss": 1.184,
"step": 91
},
{
"epoch": 0.019276098685244356,
"grad_norm": 3.9743735790252686,
"learning_rate": 0.00018400000000000003,
"loss": 1.431,
"step": 92
},
{
"epoch": 0.01948562149704049,
"grad_norm": 3.0031299591064453,
"learning_rate": 0.00018600000000000002,
"loss": 1.2836,
"step": 93
},
{
"epoch": 0.019695144308836624,
"grad_norm": 7.433412551879883,
"learning_rate": 0.000188,
"loss": 1.0807,
"step": 94
},
{
"epoch": 0.019904667120632758,
"grad_norm": 2.799851417541504,
"learning_rate": 0.00019,
"loss": 1.1394,
"step": 95
},
{
"epoch": 0.020114189932428892,
"grad_norm": 2.473200798034668,
"learning_rate": 0.000192,
"loss": 1.1627,
"step": 96
},
{
"epoch": 0.020323712744225026,
"grad_norm": 2.2027812004089355,
"learning_rate": 0.000194,
"loss": 0.9715,
"step": 97
},
{
"epoch": 0.020533235556021164,
"grad_norm": 32.15480041503906,
"learning_rate": 0.000196,
"loss": 1.1629,
"step": 98
},
{
"epoch": 0.020742758367817297,
"grad_norm": 3.503061532974243,
"learning_rate": 0.00019800000000000002,
"loss": 0.8654,
"step": 99
},
{
"epoch": 0.02095228117961343,
"grad_norm": 2.616403102874756,
"learning_rate": 0.0002,
"loss": 0.916,
"step": 100
},
{
"epoch": 0.021161803991409565,
"grad_norm": 2.435654640197754,
"learning_rate": 0.00019999999755817286,
"loss": 1.0219,
"step": 101
},
{
"epoch": 0.0213713268032057,
"grad_norm": 2.061518430709839,
"learning_rate": 0.00019999999023269155,
"loss": 0.8254,
"step": 102
},
{
"epoch": 0.021580849615001833,
"grad_norm": 3.721984386444092,
"learning_rate": 0.00019999997802355637,
"loss": 0.7386,
"step": 103
},
{
"epoch": 0.021790372426797967,
"grad_norm": 2.62754225730896,
"learning_rate": 0.00019999996093076797,
"loss": 0.8236,
"step": 104
},
{
"epoch": 0.0219998952385941,
"grad_norm": 2.1241190433502197,
"learning_rate": 0.0001999999389543272,
"loss": 1.2671,
"step": 105
},
{
"epoch": 0.022209418050390235,
"grad_norm": 1.9771560430526733,
"learning_rate": 0.0001999999120942351,
"loss": 0.9647,
"step": 106
},
{
"epoch": 0.02241894086218637,
"grad_norm": 2.057161331176758,
"learning_rate": 0.000199999880350493,
"loss": 0.6585,
"step": 107
},
{
"epoch": 0.022628463673982503,
"grad_norm": 1.6255578994750977,
"learning_rate": 0.00019999984372310242,
"loss": 0.5253,
"step": 108
},
{
"epoch": 0.02283798648577864,
"grad_norm": 2.0477192401885986,
"learning_rate": 0.0001999998022120652,
"loss": 0.5285,
"step": 109
},
{
"epoch": 0.023047509297574775,
"grad_norm": 2.287067413330078,
"learning_rate": 0.0001999997558173833,
"loss": 0.8685,
"step": 110
},
{
"epoch": 0.02325703210937091,
"grad_norm": 1.9694267511367798,
"learning_rate": 0.00019999970453905906,
"loss": 0.7955,
"step": 111
},
{
"epoch": 0.023466554921167043,
"grad_norm": 1.9603341817855835,
"learning_rate": 0.00019999964837709493,
"loss": 0.6248,
"step": 112
},
{
"epoch": 0.023676077732963176,
"grad_norm": 6.513728618621826,
"learning_rate": 0.00019999958733149366,
"loss": 1.0181,
"step": 113
},
{
"epoch": 0.02388560054475931,
"grad_norm": 3.4034464359283447,
"learning_rate": 0.00019999952140225825,
"loss": 0.6049,
"step": 114
},
{
"epoch": 0.024095123356555444,
"grad_norm": 1.8157330751419067,
"learning_rate": 0.00019999945058939194,
"loss": 0.3287,
"step": 115
},
{
"epoch": 0.02430464616835158,
"grad_norm": 1.7763786315917969,
"learning_rate": 0.00019999937489289816,
"loss": 0.6354,
"step": 116
},
{
"epoch": 0.024514168980147712,
"grad_norm": 2.1037404537200928,
"learning_rate": 0.00019999929431278055,
"loss": 0.5555,
"step": 117
},
{
"epoch": 0.024723691791943846,
"grad_norm": 1.833885908126831,
"learning_rate": 0.00019999920884904314,
"loss": 0.9263,
"step": 118
},
{
"epoch": 0.024933214603739984,
"grad_norm": 1.3898894786834717,
"learning_rate": 0.00019999911850169006,
"loss": 0.4843,
"step": 119
},
{
"epoch": 0.025142737415536118,
"grad_norm": 1.4807199239730835,
"learning_rate": 0.00019999902327072573,
"loss": 0.4379,
"step": 120
},
{
"epoch": 0.02535226022733225,
"grad_norm": 2.069584608078003,
"learning_rate": 0.0001999989231561548,
"loss": 0.5131,
"step": 121
},
{
"epoch": 0.025561783039128386,
"grad_norm": 1.5789371728897095,
"learning_rate": 0.00019999881815798215,
"loss": 0.5341,
"step": 122
},
{
"epoch": 0.02577130585092452,
"grad_norm": 3.203845500946045,
"learning_rate": 0.0001999987082762129,
"loss": 0.3562,
"step": 123
},
{
"epoch": 0.025980828662720654,
"grad_norm": 2.1900198459625244,
"learning_rate": 0.00019999859351085245,
"loss": 0.5325,
"step": 124
},
{
"epoch": 0.026190351474516788,
"grad_norm": 1.3382114171981812,
"learning_rate": 0.00019999847386190642,
"loss": 0.345,
"step": 125
},
{
"epoch": 0.02639987428631292,
"grad_norm": 5.412052631378174,
"learning_rate": 0.00019999834932938059,
"loss": 0.3673,
"step": 126
},
{
"epoch": 0.026609397098109056,
"grad_norm": 1.5586938858032227,
"learning_rate": 0.00019999821991328106,
"loss": 0.4643,
"step": 127
},
{
"epoch": 0.02681891990990519,
"grad_norm": 1.665566086769104,
"learning_rate": 0.00019999808561361418,
"loss": 0.5068,
"step": 128
},
{
"epoch": 0.027028442721701327,
"grad_norm": 1.1209849119186401,
"learning_rate": 0.00019999794643038647,
"loss": 0.2948,
"step": 129
},
{
"epoch": 0.02723796553349746,
"grad_norm": 1.27447509765625,
"learning_rate": 0.00019999780236360475,
"loss": 0.3542,
"step": 130
},
{
"epoch": 0.027447488345293595,
"grad_norm": 1.072536587715149,
"learning_rate": 0.00019999765341327609,
"loss": 0.276,
"step": 131
},
{
"epoch": 0.02765701115708973,
"grad_norm": 1.3301913738250732,
"learning_rate": 0.0001999974995794077,
"loss": 0.3438,
"step": 132
},
{
"epoch": 0.027866533968885863,
"grad_norm": 1.4693162441253662,
"learning_rate": 0.0001999973408620071,
"loss": 0.4075,
"step": 133
},
{
"epoch": 0.028076056780681997,
"grad_norm": 0.9863178133964539,
"learning_rate": 0.0001999971772610821,
"loss": 0.1655,
"step": 134
},
{
"epoch": 0.02828557959247813,
"grad_norm": 1.4921284914016724,
"learning_rate": 0.00019999700877664062,
"loss": 0.4577,
"step": 135
},
{
"epoch": 0.028495102404274265,
"grad_norm": 2.7712841033935547,
"learning_rate": 0.00019999683540869093,
"loss": 1.004,
"step": 136
},
{
"epoch": 0.0287046252160704,
"grad_norm": 2.392787218093872,
"learning_rate": 0.0001999966571572415,
"loss": 0.3403,
"step": 137
},
{
"epoch": 0.028914148027866533,
"grad_norm": 1.188810110092163,
"learning_rate": 0.00019999647402230099,
"loss": 0.349,
"step": 138
},
{
"epoch": 0.029123670839662667,
"grad_norm": 1.2173994779586792,
"learning_rate": 0.00019999628600387843,
"loss": 0.4411,
"step": 139
},
{
"epoch": 0.029333193651458804,
"grad_norm": 0.8118991851806641,
"learning_rate": 0.0001999960931019829,
"loss": 0.1753,
"step": 140
},
{
"epoch": 0.029542716463254938,
"grad_norm": 2.2631447315216064,
"learning_rate": 0.00019999589531662388,
"loss": 0.5814,
"step": 141
},
{
"epoch": 0.029752239275051072,
"grad_norm": 1.0578008890151978,
"learning_rate": 0.00019999569264781103,
"loss": 0.2053,
"step": 142
},
{
"epoch": 0.029961762086847206,
"grad_norm": 1.2986905574798584,
"learning_rate": 0.0001999954850955542,
"loss": 0.3395,
"step": 143
},
{
"epoch": 0.03017128489864334,
"grad_norm": 0.8881211280822754,
"learning_rate": 0.00019999527265986358,
"loss": 0.2499,
"step": 144
},
{
"epoch": 0.030380807710439474,
"grad_norm": 0.6569392085075378,
"learning_rate": 0.00019999505534074955,
"loss": 0.0921,
"step": 145
},
{
"epoch": 0.030590330522235608,
"grad_norm": 2.994436264038086,
"learning_rate": 0.00019999483313822266,
"loss": 0.2138,
"step": 146
},
{
"epoch": 0.030799853334031742,
"grad_norm": 0.7886744141578674,
"learning_rate": 0.00019999460605229382,
"loss": 0.1777,
"step": 147
},
{
"epoch": 0.031009376145827876,
"grad_norm": 1.0376852750778198,
"learning_rate": 0.0001999943740829741,
"loss": 0.3462,
"step": 148
},
{
"epoch": 0.03121889895762401,
"grad_norm": 2.2212657928466797,
"learning_rate": 0.00019999413723027483,
"loss": 0.3019,
"step": 149
},
{
"epoch": 0.031428421769420144,
"grad_norm": 0.9935262799263,
"learning_rate": 0.00019999389549420758,
"loss": 0.3099,
"step": 150
},
{
"epoch": 0.03163794458121628,
"grad_norm": 1.0019413232803345,
"learning_rate": 0.00019999364887478413,
"loss": 0.3573,
"step": 151
},
{
"epoch": 0.03184746739301241,
"grad_norm": 0.8304846882820129,
"learning_rate": 0.00019999339737201657,
"loss": 0.2236,
"step": 152
},
{
"epoch": 0.032056990204808546,
"grad_norm": 0.864344596862793,
"learning_rate": 0.00019999314098591712,
"loss": 0.2807,
"step": 153
},
{
"epoch": 0.03226651301660468,
"grad_norm": 0.7545657157897949,
"learning_rate": 0.0001999928797164984,
"loss": 0.2469,
"step": 154
},
{
"epoch": 0.03247603582840082,
"grad_norm": 0.8980641961097717,
"learning_rate": 0.00019999261356377308,
"loss": 0.2295,
"step": 155
},
{
"epoch": 0.032685558640196954,
"grad_norm": 1.1351686716079712,
"learning_rate": 0.00019999234252775417,
"loss": 0.2729,
"step": 156
},
{
"epoch": 0.03289508145199309,
"grad_norm": 0.8582344055175781,
"learning_rate": 0.00019999206660845494,
"loss": 0.092,
"step": 157
},
{
"epoch": 0.03310460426378922,
"grad_norm": 0.9123420715332031,
"learning_rate": 0.00019999178580588884,
"loss": 0.3193,
"step": 158
},
{
"epoch": 0.033314127075585356,
"grad_norm": 0.7440950274467468,
"learning_rate": 0.00019999150012006957,
"loss": 0.1619,
"step": 159
},
{
"epoch": 0.03352364988738149,
"grad_norm": 0.9979766607284546,
"learning_rate": 0.00019999120955101115,
"loss": 0.4078,
"step": 160
},
{
"epoch": 0.033733172699177624,
"grad_norm": 0.8199712038040161,
"learning_rate": 0.0001999909140987277,
"loss": 0.2446,
"step": 161
},
{
"epoch": 0.03394269551097376,
"grad_norm": 0.7930268049240112,
"learning_rate": 0.00019999061376323364,
"loss": 0.1743,
"step": 162
},
{
"epoch": 0.03415221832276989,
"grad_norm": 0.8929171562194824,
"learning_rate": 0.00019999030854454372,
"loss": 0.2667,
"step": 163
},
{
"epoch": 0.034361741134566026,
"grad_norm": 0.7461153864860535,
"learning_rate": 0.00019998999844267278,
"loss": 0.2506,
"step": 164
},
{
"epoch": 0.03457126394636216,
"grad_norm": 0.8262781500816345,
"learning_rate": 0.00019998968345763597,
"loss": 0.1328,
"step": 165
},
{
"epoch": 0.034780786758158294,
"grad_norm": 86.18844604492188,
"learning_rate": 0.0001999893635894487,
"loss": 0.1376,
"step": 166
},
{
"epoch": 0.03499030956995443,
"grad_norm": 1.760168194770813,
"learning_rate": 0.00019998903883812656,
"loss": 0.3097,
"step": 167
},
{
"epoch": 0.03519983238175056,
"grad_norm": 0.7841578125953674,
"learning_rate": 0.00019998870920368545,
"loss": 0.1192,
"step": 168
},
{
"epoch": 0.035409355193546696,
"grad_norm": 1.00667405128479,
"learning_rate": 0.00019998837468614142,
"loss": 0.2442,
"step": 169
},
{
"epoch": 0.03561887800534283,
"grad_norm": 0.703844428062439,
"learning_rate": 0.0001999880352855108,
"loss": 0.0965,
"step": 170
},
{
"epoch": 0.035828400817138964,
"grad_norm": 0.7780567407608032,
"learning_rate": 0.00019998769100181024,
"loss": 0.2322,
"step": 171
},
{
"epoch": 0.0360379236289351,
"grad_norm": 1.236464023590088,
"learning_rate": 0.00019998734183505649,
"loss": 0.4349,
"step": 172
},
{
"epoch": 0.03624744644073123,
"grad_norm": 0.555381715297699,
"learning_rate": 0.00019998698778526662,
"loss": 0.1172,
"step": 173
},
{
"epoch": 0.036456969252527366,
"grad_norm": 10.66753101348877,
"learning_rate": 0.0001999866288524579,
"loss": 0.1629,
"step": 174
},
{
"epoch": 0.0366664920643235,
"grad_norm": 0.7637689113616943,
"learning_rate": 0.00019998626503664793,
"loss": 0.1148,
"step": 175
},
{
"epoch": 0.03687601487611964,
"grad_norm": 1.3090569972991943,
"learning_rate": 0.00019998589633785438,
"loss": 0.2716,
"step": 176
},
{
"epoch": 0.037085537687915775,
"grad_norm": 0.7185351848602295,
"learning_rate": 0.00019998552275609532,
"loss": 0.2852,
"step": 177
},
{
"epoch": 0.03729506049971191,
"grad_norm": 0.5750662684440613,
"learning_rate": 0.00019998514429138897,
"loss": 0.1041,
"step": 178
},
{
"epoch": 0.03750458331150804,
"grad_norm": 0.8854494094848633,
"learning_rate": 0.0001999847609437538,
"loss": 0.1748,
"step": 179
},
{
"epoch": 0.03771410612330418,
"grad_norm": 0.4940667450428009,
"learning_rate": 0.0001999843727132086,
"loss": 0.0824,
"step": 180
},
{
"epoch": 0.03792362893510031,
"grad_norm": 0.423206090927124,
"learning_rate": 0.00019998397959977227,
"loss": 0.0339,
"step": 181
},
{
"epoch": 0.038133151746896445,
"grad_norm": 0.6689960956573486,
"learning_rate": 0.000199983581603464,
"loss": 0.0825,
"step": 182
},
{
"epoch": 0.03834267455869258,
"grad_norm": 1.6195799112319946,
"learning_rate": 0.00019998317872430325,
"loss": 0.4116,
"step": 183
},
{
"epoch": 0.03855219737048871,
"grad_norm": 0.6361613869667053,
"learning_rate": 0.0001999827709623097,
"loss": 0.1088,
"step": 184
},
{
"epoch": 0.038761720182284846,
"grad_norm": 0.46394994854927063,
"learning_rate": 0.00019998235831750324,
"loss": 0.0392,
"step": 185
},
{
"epoch": 0.03897124299408098,
"grad_norm": 2.3084888458251953,
"learning_rate": 0.00019998194078990405,
"loss": 0.1127,
"step": 186
},
{
"epoch": 0.039180765805877114,
"grad_norm": 1.0078133344650269,
"learning_rate": 0.0001999815183795325,
"loss": 0.2206,
"step": 187
},
{
"epoch": 0.03939028861767325,
"grad_norm": 9.654485702514648,
"learning_rate": 0.00019998109108640923,
"loss": 0.0456,
"step": 188
},
{
"epoch": 0.03959981142946938,
"grad_norm": 1.100779414176941,
"learning_rate": 0.0001999806589105551,
"loss": 0.0901,
"step": 189
},
{
"epoch": 0.039809334241265516,
"grad_norm": 0.7839688062667847,
"learning_rate": 0.00019998022185199122,
"loss": 0.1717,
"step": 190
},
{
"epoch": 0.04001885705306165,
"grad_norm": 0.7395772933959961,
"learning_rate": 0.00019997977991073897,
"loss": 0.1766,
"step": 191
},
{
"epoch": 0.040228379864857784,
"grad_norm": 0.8579478859901428,
"learning_rate": 0.00019997933308681988,
"loss": 0.2444,
"step": 192
},
{
"epoch": 0.04043790267665392,
"grad_norm": 0.7338067293167114,
"learning_rate": 0.00019997888138025577,
"loss": 0.0912,
"step": 193
},
{
"epoch": 0.04064742548845005,
"grad_norm": 0.5419819951057434,
"learning_rate": 0.0001999784247910687,
"loss": 0.14,
"step": 194
},
{
"epoch": 0.040856948300246186,
"grad_norm": 0.6646466255187988,
"learning_rate": 0.00019997796331928103,
"loss": 0.2105,
"step": 195
},
{
"epoch": 0.04106647111204233,
"grad_norm": 0.6302490830421448,
"learning_rate": 0.00019997749696491524,
"loss": 0.1134,
"step": 196
},
{
"epoch": 0.04127599392383846,
"grad_norm": 0.4890528619289398,
"learning_rate": 0.00019997702572799412,
"loss": 0.1081,
"step": 197
},
{
"epoch": 0.041485516735634595,
"grad_norm": 0.8870651125907898,
"learning_rate": 0.00019997654960854066,
"loss": 0.4391,
"step": 198
},
{
"epoch": 0.04169503954743073,
"grad_norm": 0.36530014872550964,
"learning_rate": 0.00019997606860657817,
"loss": 0.0521,
"step": 199
},
{
"epoch": 0.04190456235922686,
"grad_norm": 0.543272078037262,
"learning_rate": 0.0001999755827221301,
"loss": 0.0804,
"step": 200
},
{
"epoch": 0.04190456235922686,
"eval_loss": 0.10344672948122025,
"eval_runtime": 89.1847,
"eval_samples_per_second": 8.611,
"eval_steps_per_second": 4.306,
"step": 200
},
{
"epoch": 0.042114085171023,
"grad_norm": 0.5941206812858582,
"learning_rate": 0.00019997509195522015,
"loss": 0.1788,
"step": 201
},
{
"epoch": 0.04232360798281913,
"grad_norm": 0.6516295075416565,
"learning_rate": 0.00019997459630587234,
"loss": 0.1104,
"step": 202
},
{
"epoch": 0.042533130794615265,
"grad_norm": 0.5159923434257507,
"learning_rate": 0.00019997409577411087,
"loss": 0.065,
"step": 203
},
{
"epoch": 0.0427426536064114,
"grad_norm": 0.26800480484962463,
"learning_rate": 0.00019997359035996018,
"loss": 0.0315,
"step": 204
},
{
"epoch": 0.04295217641820753,
"grad_norm": 0.490592896938324,
"learning_rate": 0.0001999730800634449,
"loss": 0.0737,
"step": 205
},
{
"epoch": 0.04316169923000367,
"grad_norm": 0.6091493964195251,
"learning_rate": 0.00019997256488459003,
"loss": 0.1075,
"step": 206
},
{
"epoch": 0.0433712220417998,
"grad_norm": 0.6991119980812073,
"learning_rate": 0.00019997204482342067,
"loss": 0.1717,
"step": 207
},
{
"epoch": 0.043580744853595935,
"grad_norm": 0.7354519963264465,
"learning_rate": 0.00019997151987996228,
"loss": 0.1065,
"step": 208
},
{
"epoch": 0.04379026766539207,
"grad_norm": 0.5201641917228699,
"learning_rate": 0.00019997099005424044,
"loss": 0.0715,
"step": 209
},
{
"epoch": 0.0439997904771882,
"grad_norm": 0.8263329267501831,
"learning_rate": 0.00019997045534628102,
"loss": 0.0924,
"step": 210
},
{
"epoch": 0.04420931328898434,
"grad_norm": 0.566967785358429,
"learning_rate": 0.00019996991575611017,
"loss": 0.0791,
"step": 211
},
{
"epoch": 0.04441883610078047,
"grad_norm": 0.5574502348899841,
"learning_rate": 0.00019996937128375427,
"loss": 0.1083,
"step": 212
},
{
"epoch": 0.044628358912576604,
"grad_norm": 0.6977449059486389,
"learning_rate": 0.00019996882192923983,
"loss": 0.1085,
"step": 213
},
{
"epoch": 0.04483788172437274,
"grad_norm": 0.633960485458374,
"learning_rate": 0.00019996826769259373,
"loss": 0.0676,
"step": 214
},
{
"epoch": 0.04504740453616887,
"grad_norm": 0.5775350332260132,
"learning_rate": 0.000199967708573843,
"loss": 0.0677,
"step": 215
},
{
"epoch": 0.045256927347965006,
"grad_norm": 0.4216829836368561,
"learning_rate": 0.000199967144573015,
"loss": 0.0267,
"step": 216
},
{
"epoch": 0.04546645015976115,
"grad_norm": 137.67056274414062,
"learning_rate": 0.0001999665756901372,
"loss": 0.1274,
"step": 217
},
{
"epoch": 0.04567597297155728,
"grad_norm": 0.4533352553844452,
"learning_rate": 0.0001999660019252375,
"loss": 0.036,
"step": 218
},
{
"epoch": 0.045885495783353415,
"grad_norm": 0.6027405858039856,
"learning_rate": 0.0001999654232783438,
"loss": 0.0529,
"step": 219
},
{
"epoch": 0.04609501859514955,
"grad_norm": 11.28525161743164,
"learning_rate": 0.0001999648397494844,
"loss": 0.1955,
"step": 220
},
{
"epoch": 0.04630454140694568,
"grad_norm": 0.8258134722709656,
"learning_rate": 0.0001999642513386878,
"loss": 0.0888,
"step": 221
},
{
"epoch": 0.04651406421874182,
"grad_norm": 1.017187237739563,
"learning_rate": 0.00019996365804598277,
"loss": 0.3099,
"step": 222
},
{
"epoch": 0.04672358703053795,
"grad_norm": 0.6367133259773254,
"learning_rate": 0.00019996305987139823,
"loss": 0.1296,
"step": 223
},
{
"epoch": 0.046933109842334085,
"grad_norm": 7.626189231872559,
"learning_rate": 0.00019996245681496344,
"loss": 0.0709,
"step": 224
},
{
"epoch": 0.04714263265413022,
"grad_norm": 0.47820284962654114,
"learning_rate": 0.00019996184887670782,
"loss": 0.1181,
"step": 225
},
{
"epoch": 0.04735215546592635,
"grad_norm": 0.6066839694976807,
"learning_rate": 0.00019996123605666103,
"loss": 0.1149,
"step": 226
},
{
"epoch": 0.04756167827772249,
"grad_norm": 2.130101442337036,
"learning_rate": 0.00019996061835485308,
"loss": 0.1298,
"step": 227
},
{
"epoch": 0.04777120108951862,
"grad_norm": 0.5670686960220337,
"learning_rate": 0.00019995999577131408,
"loss": 0.2267,
"step": 228
},
{
"epoch": 0.047980723901314755,
"grad_norm": 0.7000769972801208,
"learning_rate": 0.00019995936830607446,
"loss": 0.1806,
"step": 229
},
{
"epoch": 0.04819024671311089,
"grad_norm": 0.7342044711112976,
"learning_rate": 0.0001999587359591648,
"loss": 0.0655,
"step": 230
},
{
"epoch": 0.04839976952490702,
"grad_norm": 0.5491848587989807,
"learning_rate": 0.00019995809873061608,
"loss": 0.1275,
"step": 231
},
{
"epoch": 0.04860929233670316,
"grad_norm": 0.3583279848098755,
"learning_rate": 0.00019995745662045936,
"loss": 0.0391,
"step": 232
},
{
"epoch": 0.04881881514849929,
"grad_norm": 2.7566843032836914,
"learning_rate": 0.000199956809628726,
"loss": 0.0951,
"step": 233
},
{
"epoch": 0.049028337960295425,
"grad_norm": 2.23818039894104,
"learning_rate": 0.00019995615775544762,
"loss": 0.2119,
"step": 234
},
{
"epoch": 0.04923786077209156,
"grad_norm": 0.6462116837501526,
"learning_rate": 0.00019995550100065606,
"loss": 0.1219,
"step": 235
},
{
"epoch": 0.04944738358388769,
"grad_norm": 0.8119645118713379,
"learning_rate": 0.00019995483936438335,
"loss": 0.2064,
"step": 236
},
{
"epoch": 0.04965690639568383,
"grad_norm": 0.4794083535671234,
"learning_rate": 0.00019995417284666183,
"loss": 0.0229,
"step": 237
},
{
"epoch": 0.04986642920747997,
"grad_norm": 0.6471462845802307,
"learning_rate": 0.00019995350144752405,
"loss": 0.1775,
"step": 238
},
{
"epoch": 0.0500759520192761,
"grad_norm": 0.5981686115264893,
"learning_rate": 0.00019995282516700277,
"loss": 0.0767,
"step": 239
},
{
"epoch": 0.050285474831072235,
"grad_norm": 0.3105001449584961,
"learning_rate": 0.00019995214400513107,
"loss": 0.0408,
"step": 240
},
{
"epoch": 0.05049499764286837,
"grad_norm": 1.2822221517562866,
"learning_rate": 0.00019995145796194218,
"loss": 0.045,
"step": 241
},
{
"epoch": 0.0507045204546645,
"grad_norm": 0.561123251914978,
"learning_rate": 0.00019995076703746965,
"loss": 0.1002,
"step": 242
},
{
"epoch": 0.05091404326646064,
"grad_norm": 0.33375391364097595,
"learning_rate": 0.00019995007123174715,
"loss": 0.0283,
"step": 243
},
{
"epoch": 0.05112356607825677,
"grad_norm": 0.34149792790412903,
"learning_rate": 0.00019994937054480872,
"loss": 0.0484,
"step": 244
},
{
"epoch": 0.051333088890052905,
"grad_norm": 0.6636136770248413,
"learning_rate": 0.00019994866497668855,
"loss": 0.1882,
"step": 245
},
{
"epoch": 0.05154261170184904,
"grad_norm": 5.894417762756348,
"learning_rate": 0.00019994795452742107,
"loss": 0.1755,
"step": 246
},
{
"epoch": 0.05175213451364517,
"grad_norm": 0.34999197721481323,
"learning_rate": 0.00019994723919704104,
"loss": 0.0353,
"step": 247
},
{
"epoch": 0.05196165732544131,
"grad_norm": 0.750867486000061,
"learning_rate": 0.00019994651898558338,
"loss": 0.1368,
"step": 248
},
{
"epoch": 0.05217118013723744,
"grad_norm": 0.35057973861694336,
"learning_rate": 0.0001999457938930832,
"loss": 0.0414,
"step": 249
},
{
"epoch": 0.052380702949033575,
"grad_norm": 0.6619654893875122,
"learning_rate": 0.000199945063919576,
"loss": 0.0132,
"step": 250
},
{
"epoch": 0.05259022576082971,
"grad_norm": 0.6549753546714783,
"learning_rate": 0.00019994432906509738,
"loss": 0.1807,
"step": 251
},
{
"epoch": 0.05279974857262584,
"grad_norm": 0.567844808101654,
"learning_rate": 0.00019994358932968323,
"loss": 0.0272,
"step": 252
},
{
"epoch": 0.05300927138442198,
"grad_norm": 0.49006056785583496,
"learning_rate": 0.00019994284471336966,
"loss": 0.0698,
"step": 253
},
{
"epoch": 0.05321879419621811,
"grad_norm": 0.655574381351471,
"learning_rate": 0.00019994209521619305,
"loss": 0.1171,
"step": 254
},
{
"epoch": 0.053428317008014245,
"grad_norm": 0.36629414558410645,
"learning_rate": 0.00019994134083819006,
"loss": 0.0883,
"step": 255
},
{
"epoch": 0.05363783981981038,
"grad_norm": 0.5820656418800354,
"learning_rate": 0.00019994058157939745,
"loss": 0.0661,
"step": 256
},
{
"epoch": 0.05384736263160651,
"grad_norm": 0.5091819763183594,
"learning_rate": 0.0001999398174398523,
"loss": 0.0839,
"step": 257
},
{
"epoch": 0.054056885443402654,
"grad_norm": 0.2602584660053253,
"learning_rate": 0.00019993904841959198,
"loss": 0.024,
"step": 258
},
{
"epoch": 0.05426640825519879,
"grad_norm": 0.5362564921379089,
"learning_rate": 0.00019993827451865405,
"loss": 0.0908,
"step": 259
},
{
"epoch": 0.05447593106699492,
"grad_norm": 0.5144950747489929,
"learning_rate": 0.00019993749573707622,
"loss": 0.0897,
"step": 260
},
{
"epoch": 0.054685453878791056,
"grad_norm": 0.4115906357765198,
"learning_rate": 0.00019993671207489664,
"loss": 0.0795,
"step": 261
},
{
"epoch": 0.05489497669058719,
"grad_norm": 0.286371111869812,
"learning_rate": 0.0001999359235321535,
"loss": 0.0123,
"step": 262
},
{
"epoch": 0.055104499502383324,
"grad_norm": 0.46544694900512695,
"learning_rate": 0.00019993513010888532,
"loss": 0.076,
"step": 263
},
{
"epoch": 0.05531402231417946,
"grad_norm": 1.89289128780365,
"learning_rate": 0.00019993433180513087,
"loss": 0.1722,
"step": 264
},
{
"epoch": 0.05552354512597559,
"grad_norm": 0.37900757789611816,
"learning_rate": 0.0001999335286209291,
"loss": 0.0712,
"step": 265
},
{
"epoch": 0.055733067937771726,
"grad_norm": 0.5412926077842712,
"learning_rate": 0.00019993272055631928,
"loss": 0.1188,
"step": 266
},
{
"epoch": 0.05594259074956786,
"grad_norm": 0.37314048409461975,
"learning_rate": 0.00019993190761134088,
"loss": 0.0592,
"step": 267
},
{
"epoch": 0.056152113561363994,
"grad_norm": 0.3859734535217285,
"learning_rate": 0.00019993108978603354,
"loss": 0.0591,
"step": 268
},
{
"epoch": 0.05636163637316013,
"grad_norm": 0.7167041301727295,
"learning_rate": 0.00019993026708043724,
"loss": 0.1676,
"step": 269
},
{
"epoch": 0.05657115918495626,
"grad_norm": 0.557241678237915,
"learning_rate": 0.00019992943949459217,
"loss": 0.106,
"step": 270
},
{
"epoch": 0.056780681996752395,
"grad_norm": 0.5354751944541931,
"learning_rate": 0.00019992860702853872,
"loss": 0.0828,
"step": 271
},
{
"epoch": 0.05699020480854853,
"grad_norm": 0.6553617119789124,
"learning_rate": 0.00019992776968231755,
"loss": 0.0578,
"step": 272
},
{
"epoch": 0.05719972762034466,
"grad_norm": 0.45751243829727173,
"learning_rate": 0.00019992692745596956,
"loss": 0.0476,
"step": 273
},
{
"epoch": 0.0574092504321408,
"grad_norm": 0.5948287844657898,
"learning_rate": 0.00019992608034953587,
"loss": 0.077,
"step": 274
},
{
"epoch": 0.05761877324393693,
"grad_norm": 0.4865845739841461,
"learning_rate": 0.0001999252283630579,
"loss": 0.0771,
"step": 275
},
{
"epoch": 0.057828296055733065,
"grad_norm": 0.5067867636680603,
"learning_rate": 0.00019992437149657719,
"loss": 0.1177,
"step": 276
},
{
"epoch": 0.0580378188675292,
"grad_norm": 1.10435152053833,
"learning_rate": 0.00019992350975013562,
"loss": 0.1476,
"step": 277
},
{
"epoch": 0.05824734167932533,
"grad_norm": 0.40550732612609863,
"learning_rate": 0.00019992264312377524,
"loss": 0.0382,
"step": 278
},
{
"epoch": 0.058456864491121474,
"grad_norm": 0.23179441690444946,
"learning_rate": 0.00019992177161753843,
"loss": 0.0308,
"step": 279
},
{
"epoch": 0.05866638730291761,
"grad_norm": 0.37846237421035767,
"learning_rate": 0.00019992089523146768,
"loss": 0.0857,
"step": 280
},
{
"epoch": 0.05887591011471374,
"grad_norm": 0.3958498239517212,
"learning_rate": 0.00019992001396560588,
"loss": 0.0874,
"step": 281
},
{
"epoch": 0.059085432926509876,
"grad_norm": 5.02588415145874,
"learning_rate": 0.000199919127819996,
"loss": 0.1182,
"step": 282
},
{
"epoch": 0.05929495573830601,
"grad_norm": 0.32736262679100037,
"learning_rate": 0.00019991823679468132,
"loss": 0.0429,
"step": 283
},
{
"epoch": 0.059504478550102144,
"grad_norm": 0.43987375497817993,
"learning_rate": 0.00019991734088970538,
"loss": 0.0614,
"step": 284
},
{
"epoch": 0.05971400136189828,
"grad_norm": 0.5660803318023682,
"learning_rate": 0.00019991644010511196,
"loss": 0.1227,
"step": 285
},
{
"epoch": 0.05992352417369441,
"grad_norm": 1.1450581550598145,
"learning_rate": 0.000199915534440945,
"loss": 0.0406,
"step": 286
},
{
"epoch": 0.060133046985490546,
"grad_norm": 0.27143630385398865,
"learning_rate": 0.00019991462389724873,
"loss": 0.029,
"step": 287
},
{
"epoch": 0.06034256979728668,
"grad_norm": 0.3402729332447052,
"learning_rate": 0.00019991370847406763,
"loss": 0.064,
"step": 288
},
{
"epoch": 0.060552092609082814,
"grad_norm": 0.39539551734924316,
"learning_rate": 0.0001999127881714464,
"loss": 0.0554,
"step": 289
},
{
"epoch": 0.06076161542087895,
"grad_norm": 0.26591384410858154,
"learning_rate": 0.00019991186298943002,
"loss": 0.0314,
"step": 290
},
{
"epoch": 0.06097113823267508,
"grad_norm": 0.3784986138343811,
"learning_rate": 0.00019991093292806363,
"loss": 0.1253,
"step": 291
},
{
"epoch": 0.061180661044471216,
"grad_norm": 0.4088621437549591,
"learning_rate": 0.00019990999798739266,
"loss": 0.0606,
"step": 292
},
{
"epoch": 0.06139018385626735,
"grad_norm": 0.4378616213798523,
"learning_rate": 0.0001999090581674628,
"loss": 0.0108,
"step": 293
},
{
"epoch": 0.061599706668063484,
"grad_norm": 0.43905991315841675,
"learning_rate": 0.0001999081134683199,
"loss": 0.0577,
"step": 294
},
{
"epoch": 0.06180922947985962,
"grad_norm": 0.2687093913555145,
"learning_rate": 0.0001999071638900101,
"loss": 0.0123,
"step": 295
},
{
"epoch": 0.06201875229165575,
"grad_norm": 0.3932202458381653,
"learning_rate": 0.00019990620943257983,
"loss": 0.0295,
"step": 296
},
{
"epoch": 0.062228275103451886,
"grad_norm": 0.4472673833370209,
"learning_rate": 0.00019990525009607566,
"loss": 0.0741,
"step": 297
},
{
"epoch": 0.06243779791524802,
"grad_norm": 0.40663135051727295,
"learning_rate": 0.00019990428588054443,
"loss": 0.0779,
"step": 298
},
{
"epoch": 0.06264732072704415,
"grad_norm": 0.3737165927886963,
"learning_rate": 0.00019990331678603328,
"loss": 0.0292,
"step": 299
},
{
"epoch": 0.06285684353884029,
"grad_norm": 0.5894246697425842,
"learning_rate": 0.00019990234281258947,
"loss": 0.0945,
"step": 300
},
{
"epoch": 0.06306636635063642,
"grad_norm": 0.5821501016616821,
"learning_rate": 0.0001999013639602606,
"loss": 0.1296,
"step": 301
},
{
"epoch": 0.06327588916243256,
"grad_norm": 0.30660367012023926,
"learning_rate": 0.00019990038022909447,
"loss": 0.0321,
"step": 302
},
{
"epoch": 0.06348541197422869,
"grad_norm": 0.42192137241363525,
"learning_rate": 0.00019989939161913913,
"loss": 0.0532,
"step": 303
},
{
"epoch": 0.06369493478602482,
"grad_norm": 0.5271754264831543,
"learning_rate": 0.00019989839813044287,
"loss": 0.0716,
"step": 304
},
{
"epoch": 0.06390445759782096,
"grad_norm": 0.20992733538150787,
"learning_rate": 0.00019989739976305416,
"loss": 0.0104,
"step": 305
},
{
"epoch": 0.06411398040961709,
"grad_norm": 0.245597243309021,
"learning_rate": 0.0001998963965170218,
"loss": 0.0175,
"step": 306
},
{
"epoch": 0.06432350322141323,
"grad_norm": 0.4208592176437378,
"learning_rate": 0.0001998953883923948,
"loss": 0.0774,
"step": 307
},
{
"epoch": 0.06453302603320936,
"grad_norm": 0.4989780783653259,
"learning_rate": 0.00019989437538922236,
"loss": 0.0687,
"step": 308
},
{
"epoch": 0.0647425488450055,
"grad_norm": 0.4769832193851471,
"learning_rate": 0.00019989335750755395,
"loss": 0.091,
"step": 309
},
{
"epoch": 0.06495207165680164,
"grad_norm": 0.3996768891811371,
"learning_rate": 0.00019989233474743926,
"loss": 0.039,
"step": 310
},
{
"epoch": 0.06516159446859777,
"grad_norm": 3.7684757709503174,
"learning_rate": 0.0001998913071089283,
"loss": 0.0485,
"step": 311
},
{
"epoch": 0.06537111728039391,
"grad_norm": 7.5791425704956055,
"learning_rate": 0.0001998902745920712,
"loss": 0.0234,
"step": 312
},
{
"epoch": 0.06558064009219004,
"grad_norm": 0.3686707019805908,
"learning_rate": 0.00019988923719691846,
"loss": 0.0449,
"step": 313
},
{
"epoch": 0.06579016290398618,
"grad_norm": 0.2311185598373413,
"learning_rate": 0.00019988819492352064,
"loss": 0.0242,
"step": 314
},
{
"epoch": 0.06599968571578231,
"grad_norm": 0.27226027846336365,
"learning_rate": 0.00019988714777192872,
"loss": 0.0281,
"step": 315
},
{
"epoch": 0.06620920852757844,
"grad_norm": 0.7749564051628113,
"learning_rate": 0.0001998860957421938,
"loss": 0.1284,
"step": 316
},
{
"epoch": 0.06641873133937458,
"grad_norm": 0.4700535237789154,
"learning_rate": 0.00019988503883436727,
"loss": 0.0723,
"step": 317
},
{
"epoch": 0.06662825415117071,
"grad_norm": 1.5472089052200317,
"learning_rate": 0.00019988397704850074,
"loss": 0.041,
"step": 318
},
{
"epoch": 0.06683777696296685,
"grad_norm": 0.99204421043396,
"learning_rate": 0.00019988291038464605,
"loss": 0.0464,
"step": 319
},
{
"epoch": 0.06704729977476298,
"grad_norm": 0.43524572253227234,
"learning_rate": 0.00019988183884285535,
"loss": 0.0346,
"step": 320
},
{
"epoch": 0.06725682258655911,
"grad_norm": 0.539294421672821,
"learning_rate": 0.00019988076242318092,
"loss": 0.1489,
"step": 321
},
{
"epoch": 0.06746634539835525,
"grad_norm": 0.3940819203853607,
"learning_rate": 0.0001998796811256753,
"loss": 0.0618,
"step": 322
},
{
"epoch": 0.06767586821015138,
"grad_norm": 0.1761884242296219,
"learning_rate": 0.0001998785949503914,
"loss": 0.0109,
"step": 323
},
{
"epoch": 0.06788539102194752,
"grad_norm": 0.4405103027820587,
"learning_rate": 0.00019987750389738215,
"loss": 0.0874,
"step": 324
},
{
"epoch": 0.06809491383374365,
"grad_norm": 0.34329459071159363,
"learning_rate": 0.0001998764079667009,
"loss": 0.0229,
"step": 325
},
{
"epoch": 0.06830443664553978,
"grad_norm": 0.26604002714157104,
"learning_rate": 0.0001998753071584012,
"loss": 0.0187,
"step": 326
},
{
"epoch": 0.06851395945733592,
"grad_norm": 0.762101411819458,
"learning_rate": 0.0001998742014725367,
"loss": 0.1381,
"step": 327
},
{
"epoch": 0.06872348226913205,
"grad_norm": 0.4048270881175995,
"learning_rate": 0.0001998730909091615,
"loss": 0.0635,
"step": 328
},
{
"epoch": 0.06893300508092819,
"grad_norm": 0.2603943645954132,
"learning_rate": 0.0001998719754683298,
"loss": 0.0073,
"step": 329
},
{
"epoch": 0.06914252789272432,
"grad_norm": 0.36697712540626526,
"learning_rate": 0.00019987085515009605,
"loss": 0.0498,
"step": 330
},
{
"epoch": 0.06935205070452045,
"grad_norm": 0.3427172601222992,
"learning_rate": 0.000199869729954515,
"loss": 0.051,
"step": 331
},
{
"epoch": 0.06956157351631659,
"grad_norm": 0.5265083909034729,
"learning_rate": 0.00019986859988164159,
"loss": 0.1354,
"step": 332
},
{
"epoch": 0.06977109632811272,
"grad_norm": 0.24376115202903748,
"learning_rate": 0.000199867464931531,
"loss": 0.0258,
"step": 333
},
{
"epoch": 0.06998061913990886,
"grad_norm": 0.39624571800231934,
"learning_rate": 0.00019986632510423866,
"loss": 0.0468,
"step": 334
},
{
"epoch": 0.07019014195170499,
"grad_norm": 0.3155399560928345,
"learning_rate": 0.00019986518039982024,
"loss": 0.0426,
"step": 335
},
{
"epoch": 0.07039966476350112,
"grad_norm": 0.3613176643848419,
"learning_rate": 0.00019986403081833167,
"loss": 0.0499,
"step": 336
},
{
"epoch": 0.07060918757529726,
"grad_norm": 0.37132036685943604,
"learning_rate": 0.000199862876359829,
"loss": 0.0516,
"step": 337
},
{
"epoch": 0.07081871038709339,
"grad_norm": 0.15536396205425262,
"learning_rate": 0.00019986171702436874,
"loss": 0.0114,
"step": 338
},
{
"epoch": 0.07102823319888953,
"grad_norm": 0.3651553690433502,
"learning_rate": 0.0001998605528120074,
"loss": 0.0383,
"step": 339
},
{
"epoch": 0.07123775601068566,
"grad_norm": 0.795184850692749,
"learning_rate": 0.00019985938372280192,
"loss": 0.0248,
"step": 340
},
{
"epoch": 0.0714472788224818,
"grad_norm": 0.28453385829925537,
"learning_rate": 0.0001998582097568093,
"loss": 0.0101,
"step": 341
},
{
"epoch": 0.07165680163427793,
"grad_norm": 0.4126911461353302,
"learning_rate": 0.00019985703091408694,
"loss": 0.0156,
"step": 342
},
{
"epoch": 0.07186632444607406,
"grad_norm": 0.3952067792415619,
"learning_rate": 0.00019985584719469241,
"loss": 0.0614,
"step": 343
},
{
"epoch": 0.0720758472578702,
"grad_norm": 0.3046548664569855,
"learning_rate": 0.0001998546585986835,
"loss": 0.0098,
"step": 344
},
{
"epoch": 0.07228537006966633,
"grad_norm": 0.46837934851646423,
"learning_rate": 0.00019985346512611826,
"loss": 0.1381,
"step": 345
},
{
"epoch": 0.07249489288146246,
"grad_norm": 0.3552689552307129,
"learning_rate": 0.00019985226677705497,
"loss": 0.0574,
"step": 346
},
{
"epoch": 0.0727044156932586,
"grad_norm": 0.39785584807395935,
"learning_rate": 0.00019985106355155214,
"loss": 0.1023,
"step": 347
},
{
"epoch": 0.07291393850505473,
"grad_norm": 0.4071919322013855,
"learning_rate": 0.00019984985544966856,
"loss": 0.0505,
"step": 348
},
{
"epoch": 0.07312346131685087,
"grad_norm": 0.4632468521595001,
"learning_rate": 0.0001998486424714632,
"loss": 0.0877,
"step": 349
},
{
"epoch": 0.073332984128647,
"grad_norm": 0.49063947796821594,
"learning_rate": 0.00019984742461699536,
"loss": 0.1152,
"step": 350
},
{
"epoch": 0.07354250694044315,
"grad_norm": 0.3042297959327698,
"learning_rate": 0.00019984620188632447,
"loss": 0.0318,
"step": 351
},
{
"epoch": 0.07375202975223928,
"grad_norm": 0.15186455845832825,
"learning_rate": 0.00019984497427951022,
"loss": 0.019,
"step": 352
},
{
"epoch": 0.07396155256403542,
"grad_norm": 0.4884166121482849,
"learning_rate": 0.00019984374179661264,
"loss": 0.0994,
"step": 353
},
{
"epoch": 0.07417107537583155,
"grad_norm": 0.5241116881370544,
"learning_rate": 0.0001998425044376918,
"loss": 0.083,
"step": 354
},
{
"epoch": 0.07438059818762768,
"grad_norm": 0.3421548306941986,
"learning_rate": 0.00019984126220280824,
"loss": 0.0552,
"step": 355
},
{
"epoch": 0.07459012099942382,
"grad_norm": 0.5838769674301147,
"learning_rate": 0.00019984001509202256,
"loss": 0.115,
"step": 356
},
{
"epoch": 0.07479964381121995,
"grad_norm": 0.27096959948539734,
"learning_rate": 0.0001998387631053957,
"loss": 0.0133,
"step": 357
},
{
"epoch": 0.07500916662301609,
"grad_norm": 0.22428373992443085,
"learning_rate": 0.0001998375062429888,
"loss": 0.0149,
"step": 358
},
{
"epoch": 0.07521868943481222,
"grad_norm": 0.4558626115322113,
"learning_rate": 0.0001998362445048632,
"loss": 0.1142,
"step": 359
},
{
"epoch": 0.07542821224660835,
"grad_norm": 0.17673128843307495,
"learning_rate": 0.00019983497789108058,
"loss": 0.023,
"step": 360
},
{
"epoch": 0.07563773505840449,
"grad_norm": 0.2917342185974121,
"learning_rate": 0.00019983370640170277,
"loss": 0.0132,
"step": 361
},
{
"epoch": 0.07584725787020062,
"grad_norm": 0.5691831707954407,
"learning_rate": 0.00019983243003679184,
"loss": 0.117,
"step": 362
},
{
"epoch": 0.07605678068199676,
"grad_norm": 0.3779132068157196,
"learning_rate": 0.00019983114879641018,
"loss": 0.0424,
"step": 363
},
{
"epoch": 0.07626630349379289,
"grad_norm": 0.1933259516954422,
"learning_rate": 0.0001998298626806203,
"loss": 0.0068,
"step": 364
},
{
"epoch": 0.07647582630558902,
"grad_norm": 0.3409664034843445,
"learning_rate": 0.00019982857168948504,
"loss": 0.0299,
"step": 365
},
{
"epoch": 0.07668534911738516,
"grad_norm": 0.21561767160892487,
"learning_rate": 0.00019982727582306743,
"loss": 0.014,
"step": 366
},
{
"epoch": 0.07689487192918129,
"grad_norm": 0.5830832123756409,
"learning_rate": 0.00019982597508143082,
"loss": 0.0851,
"step": 367
},
{
"epoch": 0.07710439474097742,
"grad_norm": 0.33271265029907227,
"learning_rate": 0.00019982466946463862,
"loss": 0.0806,
"step": 368
},
{
"epoch": 0.07731391755277356,
"grad_norm": 0.3000403642654419,
"learning_rate": 0.00019982335897275472,
"loss": 0.0338,
"step": 369
},
{
"epoch": 0.07752344036456969,
"grad_norm": 0.686974823474884,
"learning_rate": 0.00019982204360584303,
"loss": 0.082,
"step": 370
},
{
"epoch": 0.07773296317636583,
"grad_norm": 0.22148525714874268,
"learning_rate": 0.00019982072336396779,
"loss": 0.0049,
"step": 371
},
{
"epoch": 0.07794248598816196,
"grad_norm": 0.35142630338668823,
"learning_rate": 0.0001998193982471935,
"loss": 0.0228,
"step": 372
},
{
"epoch": 0.0781520087999581,
"grad_norm": 0.42207226157188416,
"learning_rate": 0.0001998180682555849,
"loss": 0.0082,
"step": 373
},
{
"epoch": 0.07836153161175423,
"grad_norm": 0.15880601108074188,
"learning_rate": 0.00019981673338920693,
"loss": 0.0209,
"step": 374
},
{
"epoch": 0.07857105442355036,
"grad_norm": 0.24623283743858337,
"learning_rate": 0.00019981539364812473,
"loss": 0.0443,
"step": 375
},
{
"epoch": 0.0787805772353465,
"grad_norm": 0.467542439699173,
"learning_rate": 0.00019981404903240377,
"loss": 0.0494,
"step": 376
},
{
"epoch": 0.07899010004714263,
"grad_norm": 0.3720758855342865,
"learning_rate": 0.00019981269954210974,
"loss": 0.0221,
"step": 377
},
{
"epoch": 0.07919962285893876,
"grad_norm": 0.3619315028190613,
"learning_rate": 0.0001998113451773085,
"loss": 0.0585,
"step": 378
},
{
"epoch": 0.0794091456707349,
"grad_norm": 0.26087716221809387,
"learning_rate": 0.00019980998593806617,
"loss": 0.0153,
"step": 379
},
{
"epoch": 0.07961866848253103,
"grad_norm": 0.4933098256587982,
"learning_rate": 0.0001998086218244492,
"loss": 0.0653,
"step": 380
},
{
"epoch": 0.07982819129432717,
"grad_norm": 0.27362197637557983,
"learning_rate": 0.00019980725283652418,
"loss": 0.0094,
"step": 381
},
{
"epoch": 0.0800377141061233,
"grad_norm": 0.39937883615493774,
"learning_rate": 0.00019980587897435795,
"loss": 0.0496,
"step": 382
},
{
"epoch": 0.08024723691791943,
"grad_norm": 0.22025969624519348,
"learning_rate": 0.00019980450023801763,
"loss": 0.0184,
"step": 383
},
{
"epoch": 0.08045675972971557,
"grad_norm": 5.886214733123779,
"learning_rate": 0.00019980311662757055,
"loss": 0.0031,
"step": 384
},
{
"epoch": 0.0806662825415117,
"grad_norm": 0.43273064494132996,
"learning_rate": 0.00019980172814308426,
"loss": 0.0363,
"step": 385
},
{
"epoch": 0.08087580535330784,
"grad_norm": 0.23640966415405273,
"learning_rate": 0.0001998003347846266,
"loss": 0.0313,
"step": 386
},
{
"epoch": 0.08108532816510397,
"grad_norm": 0.31189092993736267,
"learning_rate": 0.0001997989365522656,
"loss": 0.0372,
"step": 387
},
{
"epoch": 0.0812948509769001,
"grad_norm": 0.5833081007003784,
"learning_rate": 0.00019979753344606947,
"loss": 0.125,
"step": 388
},
{
"epoch": 0.08150437378869624,
"grad_norm": 0.2126295268535614,
"learning_rate": 0.00019979612546610687,
"loss": 0.0166,
"step": 389
},
{
"epoch": 0.08171389660049237,
"grad_norm": 0.4186943471431732,
"learning_rate": 0.0001997947126124465,
"loss": 0.0833,
"step": 390
},
{
"epoch": 0.0819234194122885,
"grad_norm": 0.143530935049057,
"learning_rate": 0.0001997932948851573,
"loss": 0.013,
"step": 391
},
{
"epoch": 0.08213294222408465,
"grad_norm": 0.31951332092285156,
"learning_rate": 0.0001997918722843086,
"loss": 0.0214,
"step": 392
},
{
"epoch": 0.08234246503588079,
"grad_norm": 0.32597488164901733,
"learning_rate": 0.00019979044480996985,
"loss": 0.0501,
"step": 393
},
{
"epoch": 0.08255198784767692,
"grad_norm": 0.2582109868526459,
"learning_rate": 0.00019978901246221073,
"loss": 0.0358,
"step": 394
},
{
"epoch": 0.08276151065947306,
"grad_norm": 0.5526602864265442,
"learning_rate": 0.00019978757524110124,
"loss": 0.0851,
"step": 395
},
{
"epoch": 0.08297103347126919,
"grad_norm": 0.26851099729537964,
"learning_rate": 0.0001997861331467115,
"loss": 0.0301,
"step": 396
},
{
"epoch": 0.08318055628306532,
"grad_norm": 0.3494580388069153,
"learning_rate": 0.00019978468617911198,
"loss": 0.0483,
"step": 397
},
{
"epoch": 0.08339007909486146,
"grad_norm": 0.062350254505872726,
"learning_rate": 0.0001997832343383734,
"loss": 0.0026,
"step": 398
},
{
"epoch": 0.08359960190665759,
"grad_norm": 0.28618085384368896,
"learning_rate": 0.00019978177762456652,
"loss": 0.0472,
"step": 399
},
{
"epoch": 0.08380912471845373,
"grad_norm": 0.4362439513206482,
"learning_rate": 0.00019978031603776263,
"loss": 0.0664,
"step": 400
},
{
"epoch": 0.08380912471845373,
"eval_loss": 0.03672989457845688,
"eval_runtime": 89.1437,
"eval_samples_per_second": 8.615,
"eval_steps_per_second": 4.308,
"step": 400
},
{
"epoch": 0.08401864753024986,
"grad_norm": 0.264435350894928,
"learning_rate": 0.000199778849578033,
"loss": 0.0699,
"step": 401
},
{
"epoch": 0.084228170342046,
"grad_norm": 0.27643996477127075,
"learning_rate": 0.00019977737824544928,
"loss": 0.0324,
"step": 402
},
{
"epoch": 0.08443769315384213,
"grad_norm": 0.22616292536258698,
"learning_rate": 0.00019977590204008338,
"loss": 0.0119,
"step": 403
},
{
"epoch": 0.08464721596563826,
"grad_norm": 0.3187446594238281,
"learning_rate": 0.00019977442096200733,
"loss": 0.0118,
"step": 404
},
{
"epoch": 0.0848567387774344,
"grad_norm": 0.42286938428878784,
"learning_rate": 0.00019977293501129349,
"loss": 0.0548,
"step": 405
},
{
"epoch": 0.08506626158923053,
"grad_norm": 0.3166376054286957,
"learning_rate": 0.0001997714441880144,
"loss": 0.0521,
"step": 406
},
{
"epoch": 0.08527578440102666,
"grad_norm": 17.92824935913086,
"learning_rate": 0.0001997699484922429,
"loss": 0.0271,
"step": 407
},
{
"epoch": 0.0854853072128228,
"grad_norm": 0.4461129307746887,
"learning_rate": 0.00019976844792405198,
"loss": 0.0275,
"step": 408
},
{
"epoch": 0.08569483002461893,
"grad_norm": 0.27550551295280457,
"learning_rate": 0.000199766942483515,
"loss": 0.048,
"step": 409
},
{
"epoch": 0.08590435283641507,
"grad_norm": 0.24863511323928833,
"learning_rate": 0.00019976543217070541,
"loss": 0.0251,
"step": 410
},
{
"epoch": 0.0861138756482112,
"grad_norm": 0.16819795966148376,
"learning_rate": 0.00019976391698569703,
"loss": 0.0053,
"step": 411
},
{
"epoch": 0.08632339846000733,
"grad_norm": 0.4115440547466278,
"learning_rate": 0.00019976239692856383,
"loss": 0.0581,
"step": 412
},
{
"epoch": 0.08653292127180347,
"grad_norm": 0.26209789514541626,
"learning_rate": 0.00019976087199938004,
"loss": 0.0267,
"step": 413
},
{
"epoch": 0.0867424440835996,
"grad_norm": 0.2739785313606262,
"learning_rate": 0.0001997593421982201,
"loss": 0.042,
"step": 414
},
{
"epoch": 0.08695196689539574,
"grad_norm": 0.343146413564682,
"learning_rate": 0.0001997578075251588,
"loss": 0.0505,
"step": 415
},
{
"epoch": 0.08716148970719187,
"grad_norm": 0.4430124759674072,
"learning_rate": 0.000199756267980271,
"loss": 0.0837,
"step": 416
},
{
"epoch": 0.087371012518988,
"grad_norm": 0.2285781353712082,
"learning_rate": 0.00019975472356363193,
"loss": 0.0207,
"step": 417
},
{
"epoch": 0.08758053533078414,
"grad_norm": 0.4867056608200073,
"learning_rate": 0.00019975317427531705,
"loss": 0.068,
"step": 418
},
{
"epoch": 0.08779005814258027,
"grad_norm": 0.2114226073026657,
"learning_rate": 0.00019975162011540197,
"loss": 0.0037,
"step": 419
},
{
"epoch": 0.0879995809543764,
"grad_norm": 0.3507726788520813,
"learning_rate": 0.00019975006108396257,
"loss": 0.0473,
"step": 420
},
{
"epoch": 0.08820910376617254,
"grad_norm": 0.346073716878891,
"learning_rate": 0.00019974849718107506,
"loss": 0.0297,
"step": 421
},
{
"epoch": 0.08841862657796867,
"grad_norm": 0.1917935311794281,
"learning_rate": 0.00019974692840681578,
"loss": 0.0312,
"step": 422
},
{
"epoch": 0.08862814938976481,
"grad_norm": 0.2887978255748749,
"learning_rate": 0.0001997453547612613,
"loss": 0.0281,
"step": 423
},
{
"epoch": 0.08883767220156094,
"grad_norm": 0.268467515707016,
"learning_rate": 0.00019974377624448853,
"loss": 0.024,
"step": 424
},
{
"epoch": 0.08904719501335707,
"grad_norm": 0.24345600605010986,
"learning_rate": 0.00019974219285657453,
"loss": 0.0136,
"step": 425
},
{
"epoch": 0.08925671782515321,
"grad_norm": 0.3125511705875397,
"learning_rate": 0.0001997406045975967,
"loss": 0.0234,
"step": 426
},
{
"epoch": 0.08946624063694934,
"grad_norm": 0.13774999976158142,
"learning_rate": 0.00019973901146763247,
"loss": 0.0119,
"step": 427
},
{
"epoch": 0.08967576344874548,
"grad_norm": 0.3603483736515045,
"learning_rate": 0.00019973741346675973,
"loss": 0.0682,
"step": 428
},
{
"epoch": 0.08988528626054161,
"grad_norm": 0.3651808500289917,
"learning_rate": 0.00019973581059505656,
"loss": 0.0447,
"step": 429
},
{
"epoch": 0.09009480907233774,
"grad_norm": 0.4079362154006958,
"learning_rate": 0.0001997342028526011,
"loss": 0.0625,
"step": 430
},
{
"epoch": 0.09030433188413388,
"grad_norm": 0.2838890254497528,
"learning_rate": 0.000199732590239472,
"loss": 0.056,
"step": 431
},
{
"epoch": 0.09051385469593001,
"grad_norm": 0.5139966011047363,
"learning_rate": 0.00019973097275574798,
"loss": 0.1074,
"step": 432
},
{
"epoch": 0.09072337750772616,
"grad_norm": 0.04335456341505051,
"learning_rate": 0.000199729350401508,
"loss": 0.001,
"step": 433
},
{
"epoch": 0.0909329003195223,
"grad_norm": 0.22099369764328003,
"learning_rate": 0.0001997277231768313,
"loss": 0.0229,
"step": 434
},
{
"epoch": 0.09114242313131843,
"grad_norm": 0.32649365067481995,
"learning_rate": 0.00019972609108179738,
"loss": 0.0403,
"step": 435
},
{
"epoch": 0.09135194594311456,
"grad_norm": 0.09762894362211227,
"learning_rate": 0.00019972445411648593,
"loss": 0.0063,
"step": 436
},
{
"epoch": 0.0915614687549107,
"grad_norm": 0.45374247431755066,
"learning_rate": 0.00019972281228097685,
"loss": 0.073,
"step": 437
},
{
"epoch": 0.09177099156670683,
"grad_norm": 0.5042026042938232,
"learning_rate": 0.0001997211655753504,
"loss": 0.0784,
"step": 438
},
{
"epoch": 0.09198051437850296,
"grad_norm": 0.2711179852485657,
"learning_rate": 0.00019971951399968695,
"loss": 0.0391,
"step": 439
},
{
"epoch": 0.0921900371902991,
"grad_norm": 0.3778410851955414,
"learning_rate": 0.00019971785755406712,
"loss": 0.0598,
"step": 440
},
{
"epoch": 0.09239956000209523,
"grad_norm": 0.2895963191986084,
"learning_rate": 0.0001997161962385719,
"loss": 0.0362,
"step": 441
},
{
"epoch": 0.09260908281389137,
"grad_norm": 0.2975994944572449,
"learning_rate": 0.00019971453005328233,
"loss": 0.0398,
"step": 442
},
{
"epoch": 0.0928186056256875,
"grad_norm": 0.3291097581386566,
"learning_rate": 0.00019971285899827987,
"loss": 0.0506,
"step": 443
},
{
"epoch": 0.09302812843748363,
"grad_norm": 0.20877154171466827,
"learning_rate": 0.00019971118307364605,
"loss": 0.0232,
"step": 444
},
{
"epoch": 0.09323765124927977,
"grad_norm": 0.2314630150794983,
"learning_rate": 0.00019970950227946274,
"loss": 0.0342,
"step": 445
},
{
"epoch": 0.0934471740610759,
"grad_norm": 0.30340278148651123,
"learning_rate": 0.00019970781661581208,
"loss": 0.0384,
"step": 446
},
{
"epoch": 0.09365669687287204,
"grad_norm": 0.17421570420265198,
"learning_rate": 0.00019970612608277631,
"loss": 0.0079,
"step": 447
},
{
"epoch": 0.09386621968466817,
"grad_norm": 0.1339452713727951,
"learning_rate": 0.00019970443068043805,
"loss": 0.0036,
"step": 448
},
{
"epoch": 0.0940757424964643,
"grad_norm": 0.10046014189720154,
"learning_rate": 0.00019970273040888003,
"loss": 0.0022,
"step": 449
},
{
"epoch": 0.09428526530826044,
"grad_norm": 0.21752049028873444,
"learning_rate": 0.00019970102526818538,
"loss": 0.0028,
"step": 450
},
{
"epoch": 0.09449478812005657,
"grad_norm": 0.38837727904319763,
"learning_rate": 0.00019969931525843726,
"loss": 0.0535,
"step": 451
},
{
"epoch": 0.0947043109318527,
"grad_norm": 0.15972809493541718,
"learning_rate": 0.00019969760037971928,
"loss": 0.0041,
"step": 452
},
{
"epoch": 0.09491383374364884,
"grad_norm": 0.08293309062719345,
"learning_rate": 0.00019969588063211514,
"loss": 0.0013,
"step": 453
},
{
"epoch": 0.09512335655544497,
"grad_norm": 0.3711838722229004,
"learning_rate": 0.00019969415601570883,
"loss": 0.0566,
"step": 454
},
{
"epoch": 0.09533287936724111,
"grad_norm": 0.21283899247646332,
"learning_rate": 0.00019969242653058463,
"loss": 0.0134,
"step": 455
},
{
"epoch": 0.09554240217903724,
"grad_norm": 0.4132600724697113,
"learning_rate": 0.0001996906921768269,
"loss": 0.0615,
"step": 456
},
{
"epoch": 0.09575192499083338,
"grad_norm": 0.37747833132743835,
"learning_rate": 0.00019968895295452042,
"loss": 0.0832,
"step": 457
},
{
"epoch": 0.09596144780262951,
"grad_norm": 0.23868821561336517,
"learning_rate": 0.00019968720886375007,
"loss": 0.0133,
"step": 458
},
{
"epoch": 0.09617097061442564,
"grad_norm": 0.23426426947116852,
"learning_rate": 0.0001996854599046011,
"loss": 0.0075,
"step": 459
},
{
"epoch": 0.09638049342622178,
"grad_norm": 0.5458897352218628,
"learning_rate": 0.00019968370607715885,
"loss": 0.151,
"step": 460
},
{
"epoch": 0.09659001623801791,
"grad_norm": 0.12708933651447296,
"learning_rate": 0.00019968194738150902,
"loss": 0.0037,
"step": 461
},
{
"epoch": 0.09679953904981405,
"grad_norm": 0.19058242440223694,
"learning_rate": 0.00019968018381773744,
"loss": 0.0158,
"step": 462
},
{
"epoch": 0.09700906186161018,
"grad_norm": 0.42510750889778137,
"learning_rate": 0.0001996784153859303,
"loss": 0.0818,
"step": 463
},
{
"epoch": 0.09721858467340631,
"grad_norm": 0.17952972650527954,
"learning_rate": 0.00019967664208617395,
"loss": 0.004,
"step": 464
},
{
"epoch": 0.09742810748520245,
"grad_norm": 0.255454421043396,
"learning_rate": 0.00019967486391855497,
"loss": 0.0224,
"step": 465
},
{
"epoch": 0.09763763029699858,
"grad_norm": 0.07083744555711746,
"learning_rate": 0.0001996730808831602,
"loss": 0.001,
"step": 466
},
{
"epoch": 0.09784715310879472,
"grad_norm": 0.1643357276916504,
"learning_rate": 0.00019967129298007676,
"loss": 0.0042,
"step": 467
},
{
"epoch": 0.09805667592059085,
"grad_norm": 0.3091591000556946,
"learning_rate": 0.0001996695002093919,
"loss": 0.0279,
"step": 468
},
{
"epoch": 0.09826619873238698,
"grad_norm": 0.1364019513130188,
"learning_rate": 0.00019966770257119322,
"loss": 0.0031,
"step": 469
},
{
"epoch": 0.09847572154418312,
"grad_norm": 0.12853386998176575,
"learning_rate": 0.00019966590006556848,
"loss": 0.0032,
"step": 470
},
{
"epoch": 0.09868524435597925,
"grad_norm": 0.39151233434677124,
"learning_rate": 0.00019966409269260577,
"loss": 0.0614,
"step": 471
},
{
"epoch": 0.09889476716777539,
"grad_norm": 0.013196398504078388,
"learning_rate": 0.00019966228045239327,
"loss": 0.0002,
"step": 472
},
{
"epoch": 0.09910428997957152,
"grad_norm": 0.1299104243516922,
"learning_rate": 0.00019966046334501952,
"loss": 0.0029,
"step": 473
},
{
"epoch": 0.09931381279136765,
"grad_norm": 0.4250582456588745,
"learning_rate": 0.0001996586413705733,
"loss": 0.0176,
"step": 474
},
{
"epoch": 0.0995233356031638,
"grad_norm": 0.1893891841173172,
"learning_rate": 0.0001996568145291435,
"loss": 0.0283,
"step": 475
},
{
"epoch": 0.09973285841495994,
"grad_norm": 0.4147910475730896,
"learning_rate": 0.00019965498282081945,
"loss": 0.048,
"step": 476
},
{
"epoch": 0.09994238122675607,
"grad_norm": 0.28522536158561707,
"learning_rate": 0.0001996531462456905,
"loss": 0.0333,
"step": 477
},
{
"epoch": 0.1001519040385522,
"grad_norm": 0.3188202977180481,
"learning_rate": 0.0001996513048038464,
"loss": 0.0127,
"step": 478
},
{
"epoch": 0.10036142685034834,
"grad_norm": 0.44609835743904114,
"learning_rate": 0.00019964945849537707,
"loss": 0.0579,
"step": 479
},
{
"epoch": 0.10057094966214447,
"grad_norm": 0.21903985738754272,
"learning_rate": 0.00019964760732037266,
"loss": 0.0105,
"step": 480
},
{
"epoch": 0.1007804724739406,
"grad_norm": 0.4808081388473511,
"learning_rate": 0.00019964575127892362,
"loss": 0.0644,
"step": 481
},
{
"epoch": 0.10098999528573674,
"grad_norm": 0.32994869351387024,
"learning_rate": 0.00019964389037112055,
"loss": 0.0401,
"step": 482
},
{
"epoch": 0.10119951809753287,
"grad_norm": 0.32692381739616394,
"learning_rate": 0.00019964202459705433,
"loss": 0.0316,
"step": 483
},
{
"epoch": 0.101409040909329,
"grad_norm": 0.24021996557712555,
"learning_rate": 0.0001996401539568161,
"loss": 0.0122,
"step": 484
},
{
"epoch": 0.10161856372112514,
"grad_norm": 0.4242231845855713,
"learning_rate": 0.00019963827845049717,
"loss": 0.0038,
"step": 485
},
{
"epoch": 0.10182808653292127,
"grad_norm": 0.028105545789003372,
"learning_rate": 0.00019963639807818922,
"loss": 0.0005,
"step": 486
},
{
"epoch": 0.10203760934471741,
"grad_norm": 0.45947086811065674,
"learning_rate": 0.000199634512839984,
"loss": 0.0611,
"step": 487
},
{
"epoch": 0.10224713215651354,
"grad_norm": 0.14129653573036194,
"learning_rate": 0.00019963262273597362,
"loss": 0.0093,
"step": 488
},
{
"epoch": 0.10245665496830968,
"grad_norm": 0.2705402970314026,
"learning_rate": 0.00019963072776625035,
"loss": 0.0236,
"step": 489
},
{
"epoch": 0.10266617778010581,
"grad_norm": 0.24441330134868622,
"learning_rate": 0.0001996288279309068,
"loss": 0.0318,
"step": 490
},
{
"epoch": 0.10287570059190194,
"grad_norm": 0.2647246718406677,
"learning_rate": 0.0001996269232300357,
"loss": 0.0095,
"step": 491
},
{
"epoch": 0.10308522340369808,
"grad_norm": 0.06135563924908638,
"learning_rate": 0.00019962501366373005,
"loss": 0.0012,
"step": 492
},
{
"epoch": 0.10329474621549421,
"grad_norm": 0.4516589045524597,
"learning_rate": 0.00019962309923208315,
"loss": 0.0518,
"step": 493
},
{
"epoch": 0.10350426902729035,
"grad_norm": 0.18813203275203705,
"learning_rate": 0.0001996211799351885,
"loss": 0.0054,
"step": 494
},
{
"epoch": 0.10371379183908648,
"grad_norm": 0.48847323656082153,
"learning_rate": 0.00019961925577313978,
"loss": 0.0173,
"step": 495
},
{
"epoch": 0.10392331465088261,
"grad_norm": 0.012579197995364666,
"learning_rate": 0.000199617326746031,
"loss": 0.0002,
"step": 496
},
{
"epoch": 0.10413283746267875,
"grad_norm": 0.6042206287384033,
"learning_rate": 0.00019961539285395636,
"loss": 0.0135,
"step": 497
},
{
"epoch": 0.10434236027447488,
"grad_norm": 0.05990588292479515,
"learning_rate": 0.00019961345409701028,
"loss": 0.001,
"step": 498
},
{
"epoch": 0.10455188308627102,
"grad_norm": 0.41477465629577637,
"learning_rate": 0.0001996115104752875,
"loss": 0.0703,
"step": 499
},
{
"epoch": 0.10476140589806715,
"grad_norm": 0.06293825805187225,
"learning_rate": 0.00019960956198888287,
"loss": 0.001,
"step": 500
},
{
"epoch": 0.10497092870986328,
"grad_norm": 0.36649006605148315,
"learning_rate": 0.0001996076086378916,
"loss": 0.0534,
"step": 501
},
{
"epoch": 0.10518045152165942,
"grad_norm": 0.29327359795570374,
"learning_rate": 0.00019960565042240907,
"loss": 0.022,
"step": 502
},
{
"epoch": 0.10538997433345555,
"grad_norm": 0.3998446762561798,
"learning_rate": 0.0001996036873425309,
"loss": 0.0819,
"step": 503
},
{
"epoch": 0.10559949714525169,
"grad_norm": 0.43492838740348816,
"learning_rate": 0.00019960171939835297,
"loss": 0.0237,
"step": 504
},
{
"epoch": 0.10580901995704782,
"grad_norm": 0.3007138967514038,
"learning_rate": 0.00019959974658997142,
"loss": 0.015,
"step": 505
},
{
"epoch": 0.10601854276884395,
"grad_norm": 0.10651643574237823,
"learning_rate": 0.0001995977689174825,
"loss": 0.0054,
"step": 506
},
{
"epoch": 0.10622806558064009,
"grad_norm": 0.3926455080509186,
"learning_rate": 0.00019959578638098288,
"loss": 0.0468,
"step": 507
},
{
"epoch": 0.10643758839243622,
"grad_norm": 0.4921119511127472,
"learning_rate": 0.00019959379898056937,
"loss": 0.0352,
"step": 508
},
{
"epoch": 0.10664711120423236,
"grad_norm": 0.15699435770511627,
"learning_rate": 0.000199591806716339,
"loss": 0.0058,
"step": 509
},
{
"epoch": 0.10685663401602849,
"grad_norm": 0.1476970762014389,
"learning_rate": 0.00019958980958838907,
"loss": 0.0176,
"step": 510
},
{
"epoch": 0.10706615682782462,
"grad_norm": 0.24107927083969116,
"learning_rate": 0.00019958780759681712,
"loss": 0.015,
"step": 511
},
{
"epoch": 0.10727567963962076,
"grad_norm": 0.27523073554039,
"learning_rate": 0.00019958580074172093,
"loss": 0.0332,
"step": 512
},
{
"epoch": 0.10748520245141689,
"grad_norm": 0.32143285870552063,
"learning_rate": 0.0001995837890231985,
"loss": 0.0221,
"step": 513
},
{
"epoch": 0.10769472526321303,
"grad_norm": 0.12730306386947632,
"learning_rate": 0.00019958177244134805,
"loss": 0.0035,
"step": 514
},
{
"epoch": 0.10790424807500916,
"grad_norm": 0.09283745288848877,
"learning_rate": 0.00019957975099626812,
"loss": 0.0067,
"step": 515
},
{
"epoch": 0.10811377088680531,
"grad_norm": 0.20224162936210632,
"learning_rate": 0.00019957772468805738,
"loss": 0.0178,
"step": 516
},
{
"epoch": 0.10832329369860144,
"grad_norm": 0.5251257419586182,
"learning_rate": 0.0001995756935168148,
"loss": 0.1334,
"step": 517
},
{
"epoch": 0.10853281651039758,
"grad_norm": 0.21088866889476776,
"learning_rate": 0.00019957365748263958,
"loss": 0.0055,
"step": 518
},
{
"epoch": 0.10874233932219371,
"grad_norm": 0.22367249429225922,
"learning_rate": 0.00019957161658563116,
"loss": 0.0097,
"step": 519
},
{
"epoch": 0.10895186213398984,
"grad_norm": 0.2710024118423462,
"learning_rate": 0.00019956957082588924,
"loss": 0.0435,
"step": 520
},
{
"epoch": 0.10916138494578598,
"grad_norm": 0.021475620567798615,
"learning_rate": 0.00019956752020351367,
"loss": 0.0004,
"step": 521
},
{
"epoch": 0.10937090775758211,
"grad_norm": 0.2687474489212036,
"learning_rate": 0.00019956546471860457,
"loss": 0.0292,
"step": 522
},
{
"epoch": 0.10958043056937825,
"grad_norm": 0.21540652215480804,
"learning_rate": 0.00019956340437126243,
"loss": 0.0088,
"step": 523
},
{
"epoch": 0.10978995338117438,
"grad_norm": 0.009567869827151299,
"learning_rate": 0.00019956133916158777,
"loss": 0.0002,
"step": 524
},
{
"epoch": 0.10999947619297051,
"grad_norm": 0.1552845984697342,
"learning_rate": 0.00019955926908968148,
"loss": 0.0216,
"step": 525
},
{
"epoch": 0.11020899900476665,
"grad_norm": 0.21615681052207947,
"learning_rate": 0.0001995571941556447,
"loss": 0.025,
"step": 526
},
{
"epoch": 0.11041852181656278,
"grad_norm": 0.08468720316886902,
"learning_rate": 0.0001995551143595787,
"loss": 0.0012,
"step": 527
},
{
"epoch": 0.11062804462835892,
"grad_norm": 0.4749722480773926,
"learning_rate": 0.00019955302970158507,
"loss": 0.0218,
"step": 528
},
{
"epoch": 0.11083756744015505,
"grad_norm": 0.2639084458351135,
"learning_rate": 0.00019955094018176564,
"loss": 0.0293,
"step": 529
},
{
"epoch": 0.11104709025195118,
"grad_norm": 0.01069062389433384,
"learning_rate": 0.00019954884580022244,
"loss": 0.0003,
"step": 530
},
{
"epoch": 0.11125661306374732,
"grad_norm": 0.2846812903881073,
"learning_rate": 0.00019954674655705773,
"loss": 0.0147,
"step": 531
},
{
"epoch": 0.11146613587554345,
"grad_norm": 0.21964187920093536,
"learning_rate": 0.00019954464245237408,
"loss": 0.0174,
"step": 532
},
{
"epoch": 0.11167565868733959,
"grad_norm": 0.19246606528759003,
"learning_rate": 0.00019954253348627418,
"loss": 0.0037,
"step": 533
},
{
"epoch": 0.11188518149913572,
"grad_norm": 0.20292885601520538,
"learning_rate": 0.0001995404196588611,
"loss": 0.0133,
"step": 534
},
{
"epoch": 0.11209470431093185,
"grad_norm": 0.11336927860975266,
"learning_rate": 0.000199538300970238,
"loss": 0.0019,
"step": 535
},
{
"epoch": 0.11230422712272799,
"grad_norm": 0.32474613189697266,
"learning_rate": 0.0001995361774205084,
"loss": 0.034,
"step": 536
},
{
"epoch": 0.11251374993452412,
"grad_norm": 0.21186456084251404,
"learning_rate": 0.000199534049009776,
"loss": 0.0139,
"step": 537
},
{
"epoch": 0.11272327274632025,
"grad_norm": 0.33177071809768677,
"learning_rate": 0.0001995319157381447,
"loss": 0.0333,
"step": 538
},
{
"epoch": 0.11293279555811639,
"grad_norm": 0.4328499734401703,
"learning_rate": 0.00019952977760571874,
"loss": 0.0529,
"step": 539
},
{
"epoch": 0.11314231836991252,
"grad_norm": 0.2981981933116913,
"learning_rate": 0.00019952763461260253,
"loss": 0.0319,
"step": 540
},
{
"epoch": 0.11335184118170866,
"grad_norm": 0.15183664858341217,
"learning_rate": 0.0001995254867589007,
"loss": 0.0021,
"step": 541
},
{
"epoch": 0.11356136399350479,
"grad_norm": 0.42127111554145813,
"learning_rate": 0.00019952333404471814,
"loss": 0.0519,
"step": 542
},
{
"epoch": 0.11377088680530092,
"grad_norm": 0.3105916976928711,
"learning_rate": 0.00019952117647015997,
"loss": 0.0263,
"step": 543
},
{
"epoch": 0.11398040961709706,
"grad_norm": 0.39538949728012085,
"learning_rate": 0.00019951901403533165,
"loss": 0.0662,
"step": 544
},
{
"epoch": 0.11418993242889319,
"grad_norm": 0.09913661330938339,
"learning_rate": 0.00019951684674033868,
"loss": 0.0017,
"step": 545
},
{
"epoch": 0.11439945524068933,
"grad_norm": 0.11607985198497772,
"learning_rate": 0.00019951467458528694,
"loss": 0.0046,
"step": 546
},
{
"epoch": 0.11460897805248546,
"grad_norm": 0.018681658431887627,
"learning_rate": 0.00019951249757028255,
"loss": 0.0002,
"step": 547
},
{
"epoch": 0.1148185008642816,
"grad_norm": 0.250102162361145,
"learning_rate": 0.00019951031569543176,
"loss": 0.0149,
"step": 548
},
{
"epoch": 0.11502802367607773,
"grad_norm": 0.08178752660751343,
"learning_rate": 0.00019950812896084115,
"loss": 0.0015,
"step": 549
},
{
"epoch": 0.11523754648787386,
"grad_norm": 0.12305411696434021,
"learning_rate": 0.00019950593736661753,
"loss": 0.0101,
"step": 550
},
{
"epoch": 0.11544706929967,
"grad_norm": 0.2741960287094116,
"learning_rate": 0.00019950374091286794,
"loss": 0.0319,
"step": 551
},
{
"epoch": 0.11565659211146613,
"grad_norm": 0.18848703801631927,
"learning_rate": 0.0001995015395996996,
"loss": 0.0123,
"step": 552
},
{
"epoch": 0.11586611492326226,
"grad_norm": 0.04150771722197533,
"learning_rate": 0.00019949933342722006,
"loss": 0.0007,
"step": 553
},
{
"epoch": 0.1160756377350584,
"grad_norm": 0.4060817062854767,
"learning_rate": 0.00019949712239553702,
"loss": 0.0568,
"step": 554
},
{
"epoch": 0.11628516054685453,
"grad_norm": 0.2881668210029602,
"learning_rate": 0.00019949490650475848,
"loss": 0.0156,
"step": 555
},
{
"epoch": 0.11649468335865067,
"grad_norm": 0.23062264919281006,
"learning_rate": 0.0001994926857549927,
"loss": 0.0144,
"step": 556
},
{
"epoch": 0.11670420617044681,
"grad_norm": 0.43525660037994385,
"learning_rate": 0.00019949046014634806,
"loss": 0.0678,
"step": 557
},
{
"epoch": 0.11691372898224295,
"grad_norm": 0.5896109938621521,
"learning_rate": 0.0001994882296789333,
"loss": 0.0418,
"step": 558
},
{
"epoch": 0.11712325179403908,
"grad_norm": 0.37245413661003113,
"learning_rate": 0.0001994859943528573,
"loss": 0.0527,
"step": 559
},
{
"epoch": 0.11733277460583522,
"grad_norm": 0.18692028522491455,
"learning_rate": 0.00019948375416822928,
"loss": 0.0029,
"step": 560
},
{
"epoch": 0.11754229741763135,
"grad_norm": 0.15593203902244568,
"learning_rate": 0.00019948150912515864,
"loss": 0.0068,
"step": 561
},
{
"epoch": 0.11775182022942748,
"grad_norm": 0.18825027346611023,
"learning_rate": 0.00019947925922375498,
"loss": 0.0112,
"step": 562
},
{
"epoch": 0.11796134304122362,
"grad_norm": 0.2613517940044403,
"learning_rate": 0.0001994770044641282,
"loss": 0.0226,
"step": 563
},
{
"epoch": 0.11817086585301975,
"grad_norm": 0.27340763807296753,
"learning_rate": 0.00019947474484638842,
"loss": 0.0384,
"step": 564
},
{
"epoch": 0.11838038866481589,
"grad_norm": 0.16672497987747192,
"learning_rate": 0.00019947248037064595,
"loss": 0.022,
"step": 565
},
{
"epoch": 0.11858991147661202,
"grad_norm": 0.35599300265312195,
"learning_rate": 0.00019947021103701145,
"loss": 0.0811,
"step": 566
},
{
"epoch": 0.11879943428840815,
"grad_norm": 0.26652926206588745,
"learning_rate": 0.0001994679368455957,
"loss": 0.0316,
"step": 567
},
{
"epoch": 0.11900895710020429,
"grad_norm": 0.4590820074081421,
"learning_rate": 0.00019946565779650977,
"loss": 0.0976,
"step": 568
},
{
"epoch": 0.11921847991200042,
"grad_norm": 0.2231590449810028,
"learning_rate": 0.00019946337388986496,
"loss": 0.0525,
"step": 569
},
{
"epoch": 0.11942800272379656,
"grad_norm": 0.2221345603466034,
"learning_rate": 0.0001994610851257728,
"loss": 0.0087,
"step": 570
},
{
"epoch": 0.11963752553559269,
"grad_norm": 0.25754231214523315,
"learning_rate": 0.00019945879150434513,
"loss": 0.0286,
"step": 571
},
{
"epoch": 0.11984704834738882,
"grad_norm": 0.24685294926166534,
"learning_rate": 0.00019945649302569387,
"loss": 0.0473,
"step": 572
},
{
"epoch": 0.12005657115918496,
"grad_norm": 0.2519691586494446,
"learning_rate": 0.0001994541896899313,
"loss": 0.0497,
"step": 573
},
{
"epoch": 0.12026609397098109,
"grad_norm": 0.1403161734342575,
"learning_rate": 0.00019945188149716995,
"loss": 0.0114,
"step": 574
},
{
"epoch": 0.12047561678277723,
"grad_norm": 0.2621317505836487,
"learning_rate": 0.00019944956844752249,
"loss": 0.0507,
"step": 575
},
{
"epoch": 0.12068513959457336,
"grad_norm": 0.21445804834365845,
"learning_rate": 0.00019944725054110192,
"loss": 0.0143,
"step": 576
},
{
"epoch": 0.1208946624063695,
"grad_norm": 0.14356960356235504,
"learning_rate": 0.00019944492777802138,
"loss": 0.0041,
"step": 577
},
{
"epoch": 0.12110418521816563,
"grad_norm": 0.17768289148807526,
"learning_rate": 0.0001994426001583944,
"loss": 0.0113,
"step": 578
},
{
"epoch": 0.12131370802996176,
"grad_norm": 0.08187161386013031,
"learning_rate": 0.00019944026768233458,
"loss": 0.001,
"step": 579
},
{
"epoch": 0.1215232308417579,
"grad_norm": 0.16168196499347687,
"learning_rate": 0.00019943793034995583,
"loss": 0.0049,
"step": 580
},
{
"epoch": 0.12173275365355403,
"grad_norm": 0.7370941042900085,
"learning_rate": 0.00019943558816137237,
"loss": 0.0387,
"step": 581
},
{
"epoch": 0.12194227646535016,
"grad_norm": 0.01981206052005291,
"learning_rate": 0.00019943324111669847,
"loss": 0.0006,
"step": 582
},
{
"epoch": 0.1221517992771463,
"grad_norm": 0.3816307485103607,
"learning_rate": 0.00019943088921604884,
"loss": 0.0461,
"step": 583
},
{
"epoch": 0.12236132208894243,
"grad_norm": 0.3173670768737793,
"learning_rate": 0.00019942853245953832,
"loss": 0.0471,
"step": 584
},
{
"epoch": 0.12257084490073857,
"grad_norm": 0.006391249131411314,
"learning_rate": 0.00019942617084728198,
"loss": 0.0001,
"step": 585
},
{
"epoch": 0.1227803677125347,
"grad_norm": 0.07457860559225082,
"learning_rate": 0.00019942380437939518,
"loss": 0.0007,
"step": 586
},
{
"epoch": 0.12298989052433083,
"grad_norm": 0.12297150492668152,
"learning_rate": 0.00019942143305599346,
"loss": 0.0016,
"step": 587
},
{
"epoch": 0.12319941333612697,
"grad_norm": 0.28252649307250977,
"learning_rate": 0.00019941905687719268,
"loss": 0.0339,
"step": 588
},
{
"epoch": 0.1234089361479231,
"grad_norm": 0.29715150594711304,
"learning_rate": 0.00019941667584310883,
"loss": 0.0297,
"step": 589
},
{
"epoch": 0.12361845895971924,
"grad_norm": 0.09914545714855194,
"learning_rate": 0.00019941428995385823,
"loss": 0.0057,
"step": 590
},
{
"epoch": 0.12382798177151537,
"grad_norm": 0.26020947098731995,
"learning_rate": 0.00019941189920955735,
"loss": 0.0249,
"step": 591
},
{
"epoch": 0.1240375045833115,
"grad_norm": 0.32188501954078674,
"learning_rate": 0.00019940950361032301,
"loss": 0.0487,
"step": 592
},
{
"epoch": 0.12424702739510764,
"grad_norm": 0.18735957145690918,
"learning_rate": 0.00019940710315627218,
"loss": 0.012,
"step": 593
},
{
"epoch": 0.12445655020690377,
"grad_norm": 0.38900870084762573,
"learning_rate": 0.00019940469784752207,
"loss": 0.0914,
"step": 594
},
{
"epoch": 0.1246660730186999,
"grad_norm": 0.6044458150863647,
"learning_rate": 0.00019940228768419016,
"loss": 0.0383,
"step": 595
},
{
"epoch": 0.12487559583049604,
"grad_norm": 0.1374080628156662,
"learning_rate": 0.00019939987266639412,
"loss": 0.002,
"step": 596
},
{
"epoch": 0.1250851186422922,
"grad_norm": 0.2042607069015503,
"learning_rate": 0.00019939745279425196,
"loss": 0.004,
"step": 597
},
{
"epoch": 0.1252946414540883,
"grad_norm": 0.2692388594150543,
"learning_rate": 0.00019939502806788178,
"loss": 0.0081,
"step": 598
},
{
"epoch": 0.12550416426588445,
"grad_norm": 0.3025107681751251,
"learning_rate": 0.0001993925984874021,
"loss": 0.032,
"step": 599
},
{
"epoch": 0.12571368707768057,
"grad_norm": 0.5294435024261475,
"learning_rate": 0.00019939016405293148,
"loss": 0.0305,
"step": 600
},
{
"epoch": 0.12571368707768057,
"eval_loss": 0.016291819512844086,
"eval_runtime": 89.0596,
"eval_samples_per_second": 8.623,
"eval_steps_per_second": 4.312,
"step": 600
},
{
"epoch": 0.12592320988947672,
"grad_norm": 0.07965458184480667,
"learning_rate": 0.00019938772476458882,
"loss": 0.0013,
"step": 601
},
{
"epoch": 0.12613273270127284,
"grad_norm": 0.161927729845047,
"learning_rate": 0.00019938528062249326,
"loss": 0.0051,
"step": 602
},
{
"epoch": 0.126342255513069,
"grad_norm": 0.2723260819911957,
"learning_rate": 0.00019938283162676418,
"loss": 0.0124,
"step": 603
},
{
"epoch": 0.1265517783248651,
"grad_norm": 0.33464378118515015,
"learning_rate": 0.00019938037777752116,
"loss": 0.0118,
"step": 604
},
{
"epoch": 0.12676130113666126,
"grad_norm": 0.21107061207294464,
"learning_rate": 0.00019937791907488408,
"loss": 0.0127,
"step": 605
},
{
"epoch": 0.12697082394845738,
"grad_norm": 0.3803098499774933,
"learning_rate": 0.00019937545551897295,
"loss": 0.0138,
"step": 606
},
{
"epoch": 0.12718034676025353,
"grad_norm": 0.2563781440258026,
"learning_rate": 0.0001993729871099081,
"loss": 0.0048,
"step": 607
},
{
"epoch": 0.12738986957204965,
"grad_norm": 0.15731534361839294,
"learning_rate": 0.00019937051384781012,
"loss": 0.0014,
"step": 608
},
{
"epoch": 0.1275993923838458,
"grad_norm": 0.07711096107959747,
"learning_rate": 0.00019936803573279975,
"loss": 0.0013,
"step": 609
},
{
"epoch": 0.12780891519564191,
"grad_norm": 0.07940837740898132,
"learning_rate": 0.00019936555276499804,
"loss": 0.0011,
"step": 610
},
{
"epoch": 0.12801843800743806,
"grad_norm": 0.26590949296951294,
"learning_rate": 0.00019936306494452625,
"loss": 0.0189,
"step": 611
},
{
"epoch": 0.12822796081923418,
"grad_norm": 0.3016117811203003,
"learning_rate": 0.00019936057227150583,
"loss": 0.0182,
"step": 612
},
{
"epoch": 0.12843748363103033,
"grad_norm": 0.2093367725610733,
"learning_rate": 0.00019935807474605856,
"loss": 0.0143,
"step": 613
},
{
"epoch": 0.12864700644282645,
"grad_norm": 0.22494176030158997,
"learning_rate": 0.0001993555723683064,
"loss": 0.0178,
"step": 614
},
{
"epoch": 0.1288565292546226,
"grad_norm": 0.1732041835784912,
"learning_rate": 0.00019935306513837157,
"loss": 0.014,
"step": 615
},
{
"epoch": 0.12906605206641872,
"grad_norm": 0.04634670913219452,
"learning_rate": 0.0001993505530563765,
"loss": 0.0004,
"step": 616
},
{
"epoch": 0.12927557487821487,
"grad_norm": 0.24240362644195557,
"learning_rate": 0.00019934803612244387,
"loss": 0.0055,
"step": 617
},
{
"epoch": 0.129485097690011,
"grad_norm": 0.3496238887310028,
"learning_rate": 0.0001993455143366966,
"loss": 0.0463,
"step": 618
},
{
"epoch": 0.12969462050180713,
"grad_norm": 0.3036729097366333,
"learning_rate": 0.0001993429876992578,
"loss": 0.028,
"step": 619
},
{
"epoch": 0.12990414331360328,
"grad_norm": 0.19408352673053741,
"learning_rate": 0.00019934045621025097,
"loss": 0.0051,
"step": 620
},
{
"epoch": 0.1301136661253994,
"grad_norm": 0.4805002808570862,
"learning_rate": 0.00019933791986979964,
"loss": 0.0933,
"step": 621
},
{
"epoch": 0.13032318893719555,
"grad_norm": 0.06430470943450928,
"learning_rate": 0.00019933537867802774,
"loss": 0.0002,
"step": 622
},
{
"epoch": 0.13053271174899167,
"grad_norm": 0.23753847181797028,
"learning_rate": 0.00019933283263505932,
"loss": 0.023,
"step": 623
},
{
"epoch": 0.13074223456078782,
"grad_norm": 0.07812167704105377,
"learning_rate": 0.00019933028174101876,
"loss": 0.0008,
"step": 624
},
{
"epoch": 0.13095175737258394,
"grad_norm": 0.4537438750267029,
"learning_rate": 0.0001993277259960306,
"loss": 0.0196,
"step": 625
},
{
"epoch": 0.13116128018438009,
"grad_norm": 0.11305277794599533,
"learning_rate": 0.00019932516540021973,
"loss": 0.0136,
"step": 626
},
{
"epoch": 0.1313708029961762,
"grad_norm": 0.13304772973060608,
"learning_rate": 0.00019932259995371113,
"loss": 0.0013,
"step": 627
},
{
"epoch": 0.13158032580797235,
"grad_norm": 0.1756788194179535,
"learning_rate": 0.00019932002965663008,
"loss": 0.0033,
"step": 628
},
{
"epoch": 0.13178984861976847,
"grad_norm": 0.11411143094301224,
"learning_rate": 0.00019931745450910216,
"loss": 0.0011,
"step": 629
},
{
"epoch": 0.13199937143156462,
"grad_norm": 0.21732713282108307,
"learning_rate": 0.00019931487451125307,
"loss": 0.0138,
"step": 630
},
{
"epoch": 0.13220889424336074,
"grad_norm": 0.011272534728050232,
"learning_rate": 0.00019931228966320883,
"loss": 0.0001,
"step": 631
},
{
"epoch": 0.1324184170551569,
"grad_norm": 0.2807530462741852,
"learning_rate": 0.00019930969996509573,
"loss": 0.0142,
"step": 632
},
{
"epoch": 0.132627939866953,
"grad_norm": 0.24529138207435608,
"learning_rate": 0.00019930710541704016,
"loss": 0.0096,
"step": 633
},
{
"epoch": 0.13283746267874916,
"grad_norm": 0.2031984180212021,
"learning_rate": 0.00019930450601916887,
"loss": 0.0169,
"step": 634
},
{
"epoch": 0.13304698549054528,
"grad_norm": 0.06990045309066772,
"learning_rate": 0.0001993019017716088,
"loss": 0.0008,
"step": 635
},
{
"epoch": 0.13325650830234143,
"grad_norm": 0.3571747839450836,
"learning_rate": 0.00019929929267448713,
"loss": 0.0603,
"step": 636
},
{
"epoch": 0.13346603111413755,
"grad_norm": 0.28505340218544006,
"learning_rate": 0.0001992966787279313,
"loss": 0.0057,
"step": 637
},
{
"epoch": 0.1336755539259337,
"grad_norm": 0.5852503776550293,
"learning_rate": 0.00019929405993206894,
"loss": 0.0892,
"step": 638
},
{
"epoch": 0.1338850767377298,
"grad_norm": 0.11871087551116943,
"learning_rate": 0.00019929143628702795,
"loss": 0.0019,
"step": 639
},
{
"epoch": 0.13409459954952596,
"grad_norm": 0.09931576997041702,
"learning_rate": 0.00019928880779293646,
"loss": 0.0015,
"step": 640
},
{
"epoch": 0.13430412236132208,
"grad_norm": 0.08426062762737274,
"learning_rate": 0.0001992861744499228,
"loss": 0.0086,
"step": 641
},
{
"epoch": 0.13451364517311823,
"grad_norm": 0.20797474682331085,
"learning_rate": 0.00019928353625811565,
"loss": 0.0136,
"step": 642
},
{
"epoch": 0.13472316798491435,
"grad_norm": 0.1903754323720932,
"learning_rate": 0.00019928089321764384,
"loss": 0.0294,
"step": 643
},
{
"epoch": 0.1349326907967105,
"grad_norm": 0.19812871515750885,
"learning_rate": 0.00019927824532863636,
"loss": 0.0296,
"step": 644
},
{
"epoch": 0.13514221360850662,
"grad_norm": 0.3847843408584595,
"learning_rate": 0.00019927559259122263,
"loss": 0.0503,
"step": 645
},
{
"epoch": 0.13535173642030277,
"grad_norm": 0.2451527714729309,
"learning_rate": 0.00019927293500553213,
"loss": 0.0196,
"step": 646
},
{
"epoch": 0.13556125923209889,
"grad_norm": 0.4702102243900299,
"learning_rate": 0.00019927027257169466,
"loss": 0.0913,
"step": 647
},
{
"epoch": 0.13577078204389503,
"grad_norm": 0.10364199429750443,
"learning_rate": 0.00019926760528984026,
"loss": 0.002,
"step": 648
},
{
"epoch": 0.13598030485569115,
"grad_norm": 0.19361010193824768,
"learning_rate": 0.00019926493316009918,
"loss": 0.0167,
"step": 649
},
{
"epoch": 0.1361898276674873,
"grad_norm": 0.11378200352191925,
"learning_rate": 0.00019926225618260195,
"loss": 0.017,
"step": 650
},
{
"epoch": 0.13639935047928342,
"grad_norm": 0.25813087821006775,
"learning_rate": 0.00019925957435747925,
"loss": 0.0089,
"step": 651
},
{
"epoch": 0.13660887329107957,
"grad_norm": 13.186869621276855,
"learning_rate": 0.0001992568876848621,
"loss": 0.0159,
"step": 652
},
{
"epoch": 0.1368183961028757,
"grad_norm": 0.28748589754104614,
"learning_rate": 0.00019925419616488165,
"loss": 0.0278,
"step": 653
},
{
"epoch": 0.13702791891467184,
"grad_norm": 0.21648818254470825,
"learning_rate": 0.00019925149979766938,
"loss": 0.0133,
"step": 654
},
{
"epoch": 0.13723744172646796,
"grad_norm": 0.08441877365112305,
"learning_rate": 0.00019924879858335698,
"loss": 0.0149,
"step": 655
},
{
"epoch": 0.1374469645382641,
"grad_norm": 0.15462300181388855,
"learning_rate": 0.00019924609252207638,
"loss": 0.0065,
"step": 656
},
{
"epoch": 0.13765648735006022,
"grad_norm": 0.2789871394634247,
"learning_rate": 0.00019924338161395971,
"loss": 0.018,
"step": 657
},
{
"epoch": 0.13786601016185637,
"grad_norm": 0.019855428487062454,
"learning_rate": 0.00019924066585913933,
"loss": 0.0003,
"step": 658
},
{
"epoch": 0.1380755329736525,
"grad_norm": 0.2729547619819641,
"learning_rate": 0.00019923794525774792,
"loss": 0.0315,
"step": 659
},
{
"epoch": 0.13828505578544864,
"grad_norm": 0.18892449140548706,
"learning_rate": 0.00019923521980991834,
"loss": 0.0203,
"step": 660
},
{
"epoch": 0.1384945785972448,
"grad_norm": 0.02755853161215782,
"learning_rate": 0.00019923248951578368,
"loss": 0.0003,
"step": 661
},
{
"epoch": 0.1387041014090409,
"grad_norm": 0.016459867358207703,
"learning_rate": 0.00019922975437547727,
"loss": 0.0003,
"step": 662
},
{
"epoch": 0.13891362422083706,
"grad_norm": 0.29161739349365234,
"learning_rate": 0.0001992270143891327,
"loss": 0.021,
"step": 663
},
{
"epoch": 0.13912314703263318,
"grad_norm": 0.38459256291389465,
"learning_rate": 0.00019922426955688375,
"loss": 0.013,
"step": 664
},
{
"epoch": 0.13933266984442932,
"grad_norm": 0.30727267265319824,
"learning_rate": 0.0001992215198788645,
"loss": 0.0163,
"step": 665
},
{
"epoch": 0.13954219265622544,
"grad_norm": 0.4134649336338043,
"learning_rate": 0.00019921876535520925,
"loss": 0.029,
"step": 666
},
{
"epoch": 0.1397517154680216,
"grad_norm": 0.08295508474111557,
"learning_rate": 0.00019921600598605247,
"loss": 0.0013,
"step": 667
},
{
"epoch": 0.1399612382798177,
"grad_norm": 3.298372983932495,
"learning_rate": 0.000199213241771529,
"loss": 0.019,
"step": 668
},
{
"epoch": 0.14017076109161386,
"grad_norm": 0.2583758234977722,
"learning_rate": 0.0001992104727117737,
"loss": 0.0384,
"step": 669
},
{
"epoch": 0.14038028390340998,
"grad_norm": 0.14796248078346252,
"learning_rate": 0.00019920769880692193,
"loss": 0.0035,
"step": 670
},
{
"epoch": 0.14058980671520613,
"grad_norm": 0.006577362772077322,
"learning_rate": 0.0001992049200571091,
"loss": 0.0002,
"step": 671
},
{
"epoch": 0.14079932952700225,
"grad_norm": 0.3680857717990875,
"learning_rate": 0.00019920213646247093,
"loss": 0.0824,
"step": 672
},
{
"epoch": 0.1410088523387984,
"grad_norm": 0.07552894949913025,
"learning_rate": 0.00019919934802314335,
"loss": 0.0008,
"step": 673
},
{
"epoch": 0.14121837515059452,
"grad_norm": 0.19062870740890503,
"learning_rate": 0.00019919655473926256,
"loss": 0.0081,
"step": 674
},
{
"epoch": 0.14142789796239066,
"grad_norm": 0.3882509171962738,
"learning_rate": 0.00019919375661096492,
"loss": 0.0785,
"step": 675
},
{
"epoch": 0.14163742077418678,
"grad_norm": 0.22886379063129425,
"learning_rate": 0.00019919095363838714,
"loss": 0.0291,
"step": 676
},
{
"epoch": 0.14184694358598293,
"grad_norm": 0.5058407783508301,
"learning_rate": 0.0001991881458216661,
"loss": 0.0812,
"step": 677
},
{
"epoch": 0.14205646639777905,
"grad_norm": 0.19598716497421265,
"learning_rate": 0.00019918533316093888,
"loss": 0.0166,
"step": 678
},
{
"epoch": 0.1422659892095752,
"grad_norm": 0.3560320734977722,
"learning_rate": 0.0001991825156563429,
"loss": 0.0452,
"step": 679
},
{
"epoch": 0.14247551202137132,
"grad_norm": 0.23111455142498016,
"learning_rate": 0.0001991796933080157,
"loss": 0.0225,
"step": 680
},
{
"epoch": 0.14268503483316747,
"grad_norm": 0.18612909317016602,
"learning_rate": 0.00019917686611609515,
"loss": 0.005,
"step": 681
},
{
"epoch": 0.1428945576449636,
"grad_norm": 0.31149572134017944,
"learning_rate": 0.00019917403408071933,
"loss": 0.0557,
"step": 682
},
{
"epoch": 0.14310408045675974,
"grad_norm": 0.43545660376548767,
"learning_rate": 0.00019917119720202649,
"loss": 0.0246,
"step": 683
},
{
"epoch": 0.14331360326855586,
"grad_norm": 0.20558680593967438,
"learning_rate": 0.00019916835548015524,
"loss": 0.0144,
"step": 684
},
{
"epoch": 0.143523126080352,
"grad_norm": 0.19289907813072205,
"learning_rate": 0.00019916550891524435,
"loss": 0.004,
"step": 685
},
{
"epoch": 0.14373264889214812,
"grad_norm": 0.2448616772890091,
"learning_rate": 0.0001991626575074328,
"loss": 0.0114,
"step": 686
},
{
"epoch": 0.14394217170394427,
"grad_norm": 0.1556064635515213,
"learning_rate": 0.00019915980125685985,
"loss": 0.0031,
"step": 687
},
{
"epoch": 0.1441516945157404,
"grad_norm": 0.17058494687080383,
"learning_rate": 0.000199156940163665,
"loss": 0.008,
"step": 688
},
{
"epoch": 0.14436121732753654,
"grad_norm": 0.32099649310112,
"learning_rate": 0.00019915407422798798,
"loss": 0.0548,
"step": 689
},
{
"epoch": 0.14457074013933266,
"grad_norm": 0.11824580281972885,
"learning_rate": 0.00019915120344996876,
"loss": 0.0019,
"step": 690
},
{
"epoch": 0.1447802629511288,
"grad_norm": 0.13612987101078033,
"learning_rate": 0.00019914832782974752,
"loss": 0.0076,
"step": 691
},
{
"epoch": 0.14498978576292493,
"grad_norm": 0.3279030919075012,
"learning_rate": 0.0001991454473674647,
"loss": 0.0282,
"step": 692
},
{
"epoch": 0.14519930857472108,
"grad_norm": 0.22970880568027496,
"learning_rate": 0.00019914256206326098,
"loss": 0.0157,
"step": 693
},
{
"epoch": 0.1454088313865172,
"grad_norm": 0.01073368638753891,
"learning_rate": 0.00019913967191727725,
"loss": 0.0002,
"step": 694
},
{
"epoch": 0.14561835419831334,
"grad_norm": 0.20023581385612488,
"learning_rate": 0.00019913677692965468,
"loss": 0.0091,
"step": 695
},
{
"epoch": 0.14582787701010946,
"grad_norm": 0.08236939460039139,
"learning_rate": 0.00019913387710053466,
"loss": 0.0008,
"step": 696
},
{
"epoch": 0.1460373998219056,
"grad_norm": 0.17614483833312988,
"learning_rate": 0.00019913097243005876,
"loss": 0.0149,
"step": 697
},
{
"epoch": 0.14624692263370173,
"grad_norm": 0.4621315598487854,
"learning_rate": 0.00019912806291836886,
"loss": 0.0858,
"step": 698
},
{
"epoch": 0.14645644544549788,
"grad_norm": 4.610383987426758,
"learning_rate": 0.00019912514856560706,
"loss": 0.0345,
"step": 699
},
{
"epoch": 0.146665968257294,
"grad_norm": 0.02050645463168621,
"learning_rate": 0.00019912222937191568,
"loss": 0.0003,
"step": 700
},
{
"epoch": 0.14687549106909015,
"grad_norm": 0.10539437085390091,
"learning_rate": 0.0001991193053374373,
"loss": 0.0012,
"step": 701
},
{
"epoch": 0.1470850138808863,
"grad_norm": 0.2869635820388794,
"learning_rate": 0.00019911637646231466,
"loss": 0.0303,
"step": 702
},
{
"epoch": 0.14729453669268242,
"grad_norm": 0.29913896322250366,
"learning_rate": 0.00019911344274669085,
"loss": 0.0083,
"step": 703
},
{
"epoch": 0.14750405950447856,
"grad_norm": 0.21330636739730835,
"learning_rate": 0.00019911050419070913,
"loss": 0.0246,
"step": 704
},
{
"epoch": 0.14771358231627468,
"grad_norm": 0.010840353555977345,
"learning_rate": 0.00019910756079451303,
"loss": 0.0002,
"step": 705
},
{
"epoch": 0.14792310512807083,
"grad_norm": 0.31384751200675964,
"learning_rate": 0.00019910461255824626,
"loss": 0.0356,
"step": 706
},
{
"epoch": 0.14813262793986695,
"grad_norm": 0.20688189566135406,
"learning_rate": 0.00019910165948205282,
"loss": 0.022,
"step": 707
},
{
"epoch": 0.1483421507516631,
"grad_norm": 0.37733882665634155,
"learning_rate": 0.00019909870156607692,
"loss": 0.0762,
"step": 708
},
{
"epoch": 0.14855167356345922,
"grad_norm": 0.3063774108886719,
"learning_rate": 0.000199095738810463,
"loss": 0.0408,
"step": 709
},
{
"epoch": 0.14876119637525537,
"grad_norm": 0.38248753547668457,
"learning_rate": 0.0001990927712153558,
"loss": 0.0192,
"step": 710
},
{
"epoch": 0.1489707191870515,
"grad_norm": 0.1782226413488388,
"learning_rate": 0.00019908979878090018,
"loss": 0.0052,
"step": 711
},
{
"epoch": 0.14918024199884763,
"grad_norm": 0.34892570972442627,
"learning_rate": 0.00019908682150724137,
"loss": 0.0729,
"step": 712
},
{
"epoch": 0.14938976481064375,
"grad_norm": 0.266937792301178,
"learning_rate": 0.00019908383939452472,
"loss": 0.0168,
"step": 713
},
{
"epoch": 0.1495992876224399,
"grad_norm": 0.06860234588384628,
"learning_rate": 0.0001990808524428959,
"loss": 0.0009,
"step": 714
},
{
"epoch": 0.14980881043423602,
"grad_norm": 0.029732417315244675,
"learning_rate": 0.00019907786065250077,
"loss": 0.0005,
"step": 715
},
{
"epoch": 0.15001833324603217,
"grad_norm": 0.36950287222862244,
"learning_rate": 0.00019907486402348542,
"loss": 0.0457,
"step": 716
},
{
"epoch": 0.1502278560578283,
"grad_norm": 0.29846426844596863,
"learning_rate": 0.00019907186255599624,
"loss": 0.0389,
"step": 717
},
{
"epoch": 0.15043737886962444,
"grad_norm": 0.25788208842277527,
"learning_rate": 0.00019906885625017978,
"loss": 0.0322,
"step": 718
},
{
"epoch": 0.15064690168142056,
"grad_norm": 1.249671459197998,
"learning_rate": 0.00019906584510618282,
"loss": 0.0354,
"step": 719
},
{
"epoch": 0.1508564244932167,
"grad_norm": 0.02440783567726612,
"learning_rate": 0.00019906282912415249,
"loss": 0.0005,
"step": 720
},
{
"epoch": 0.15106594730501283,
"grad_norm": 0.22200000286102295,
"learning_rate": 0.000199059808304236,
"loss": 0.0232,
"step": 721
},
{
"epoch": 0.15127547011680897,
"grad_norm": 0.3790658116340637,
"learning_rate": 0.00019905678264658095,
"loss": 0.0429,
"step": 722
},
{
"epoch": 0.1514849929286051,
"grad_norm": 0.19107124209403992,
"learning_rate": 0.0001990537521513351,
"loss": 0.0303,
"step": 723
},
{
"epoch": 0.15169451574040124,
"grad_norm": 0.1712956577539444,
"learning_rate": 0.0001990507168186464,
"loss": 0.0097,
"step": 724
},
{
"epoch": 0.15190403855219736,
"grad_norm": 0.17346502840518951,
"learning_rate": 0.00019904767664866312,
"loss": 0.0111,
"step": 725
},
{
"epoch": 0.1521135613639935,
"grad_norm": 0.19370456039905548,
"learning_rate": 0.00019904463164153367,
"loss": 0.0049,
"step": 726
},
{
"epoch": 0.15232308417578963,
"grad_norm": 0.02195919305086136,
"learning_rate": 0.00019904158179740686,
"loss": 0.0007,
"step": 727
},
{
"epoch": 0.15253260698758578,
"grad_norm": 0.39460641145706177,
"learning_rate": 0.00019903852711643157,
"loss": 0.0444,
"step": 728
},
{
"epoch": 0.1527421297993819,
"grad_norm": 0.1139521673321724,
"learning_rate": 0.00019903546759875696,
"loss": 0.0103,
"step": 729
},
{
"epoch": 0.15295165261117805,
"grad_norm": 0.07739255577325821,
"learning_rate": 0.0001990324032445325,
"loss": 0.0053,
"step": 730
},
{
"epoch": 0.15316117542297417,
"grad_norm": 0.3823677599430084,
"learning_rate": 0.00019902933405390783,
"loss": 0.04,
"step": 731
},
{
"epoch": 0.15337069823477031,
"grad_norm": 0.24092869460582733,
"learning_rate": 0.00019902626002703283,
"loss": 0.0482,
"step": 732
},
{
"epoch": 0.15358022104656643,
"grad_norm": 0.027126243337988853,
"learning_rate": 0.00019902318116405757,
"loss": 0.0008,
"step": 733
},
{
"epoch": 0.15378974385836258,
"grad_norm": 0.12171046435832977,
"learning_rate": 0.00019902009746513254,
"loss": 0.0058,
"step": 734
},
{
"epoch": 0.1539992666701587,
"grad_norm": 0.10180404037237167,
"learning_rate": 0.0001990170089304082,
"loss": 0.0046,
"step": 735
},
{
"epoch": 0.15420878948195485,
"grad_norm": 0.26162436604499817,
"learning_rate": 0.00019901391556003545,
"loss": 0.0375,
"step": 736
},
{
"epoch": 0.15441831229375097,
"grad_norm": 0.19409051537513733,
"learning_rate": 0.00019901081735416537,
"loss": 0.0063,
"step": 737
},
{
"epoch": 0.15462783510554712,
"grad_norm": 0.17046910524368286,
"learning_rate": 0.00019900771431294927,
"loss": 0.0151,
"step": 738
},
{
"epoch": 0.15483735791734324,
"grad_norm": 0.16491179168224335,
"learning_rate": 0.00019900460643653865,
"loss": 0.0228,
"step": 739
},
{
"epoch": 0.15504688072913939,
"grad_norm": 0.4564284384250641,
"learning_rate": 0.00019900149372508528,
"loss": 0.0516,
"step": 740
},
{
"epoch": 0.1552564035409355,
"grad_norm": 0.2507207989692688,
"learning_rate": 0.00019899837617874124,
"loss": 0.0186,
"step": 741
},
{
"epoch": 0.15546592635273165,
"grad_norm": 0.18009963631629944,
"learning_rate": 0.00019899525379765877,
"loss": 0.0056,
"step": 742
},
{
"epoch": 0.1556754491645278,
"grad_norm": 0.2279609739780426,
"learning_rate": 0.0001989921265819903,
"loss": 0.0183,
"step": 743
},
{
"epoch": 0.15588497197632392,
"grad_norm": 0.1303996741771698,
"learning_rate": 0.00019898899453188855,
"loss": 0.0041,
"step": 744
},
{
"epoch": 0.15609449478812007,
"grad_norm": 0.2132912129163742,
"learning_rate": 0.00019898585764750655,
"loss": 0.0169,
"step": 745
},
{
"epoch": 0.1563040175999162,
"grad_norm": 0.01601547561585903,
"learning_rate": 0.00019898271592899743,
"loss": 0.0003,
"step": 746
},
{
"epoch": 0.15651354041171234,
"grad_norm": 0.09785155951976776,
"learning_rate": 0.00019897956937651467,
"loss": 0.0018,
"step": 747
},
{
"epoch": 0.15672306322350846,
"grad_norm": 1.353312611579895,
"learning_rate": 0.0001989764179902119,
"loss": 0.0169,
"step": 748
},
{
"epoch": 0.1569325860353046,
"grad_norm": 0.2503106892108917,
"learning_rate": 0.00019897326177024304,
"loss": 0.0186,
"step": 749
},
{
"epoch": 0.15714210884710073,
"grad_norm": 0.23809774219989777,
"learning_rate": 0.00019897010071676222,
"loss": 0.0317,
"step": 750
},
{
"epoch": 0.15735163165889687,
"grad_norm": 0.332171767950058,
"learning_rate": 0.0001989669348299238,
"loss": 0.0446,
"step": 751
},
{
"epoch": 0.157561154470693,
"grad_norm": 0.21880975365638733,
"learning_rate": 0.00019896376410988244,
"loss": 0.0172,
"step": 752
},
{
"epoch": 0.15777067728248914,
"grad_norm": 0.056077443063259125,
"learning_rate": 0.00019896058855679293,
"loss": 0.0008,
"step": 753
},
{
"epoch": 0.15798020009428526,
"grad_norm": 0.38360893726348877,
"learning_rate": 0.00019895740817081042,
"loss": 0.0377,
"step": 754
},
{
"epoch": 0.1581897229060814,
"grad_norm": 0.2242017686367035,
"learning_rate": 0.00019895422295209016,
"loss": 0.0023,
"step": 755
},
{
"epoch": 0.15839924571787753,
"grad_norm": 0.4162478744983673,
"learning_rate": 0.00019895103290078773,
"loss": 0.0054,
"step": 756
},
{
"epoch": 0.15860876852967368,
"grad_norm": 0.12590032815933228,
"learning_rate": 0.00019894783801705895,
"loss": 0.007,
"step": 757
},
{
"epoch": 0.1588182913414698,
"grad_norm": 0.10577323287725449,
"learning_rate": 0.0001989446383010598,
"loss": 0.0109,
"step": 758
},
{
"epoch": 0.15902781415326595,
"grad_norm": 0.1796669065952301,
"learning_rate": 0.00019894143375294657,
"loss": 0.0109,
"step": 759
},
{
"epoch": 0.15923733696506207,
"grad_norm": 0.19568434357643127,
"learning_rate": 0.00019893822437287576,
"loss": 0.0059,
"step": 760
},
{
"epoch": 0.1594468597768582,
"grad_norm": 0.22991542518138885,
"learning_rate": 0.00019893501016100408,
"loss": 0.0173,
"step": 761
},
{
"epoch": 0.15965638258865433,
"grad_norm": 0.19536703824996948,
"learning_rate": 0.00019893179111748856,
"loss": 0.022,
"step": 762
},
{
"epoch": 0.15986590540045048,
"grad_norm": 0.18410378694534302,
"learning_rate": 0.00019892856724248634,
"loss": 0.016,
"step": 763
},
{
"epoch": 0.1600754282122466,
"grad_norm": 0.0290357768535614,
"learning_rate": 0.00019892533853615492,
"loss": 0.0002,
"step": 764
},
{
"epoch": 0.16028495102404275,
"grad_norm": 0.23183897137641907,
"learning_rate": 0.00019892210499865192,
"loss": 0.0047,
"step": 765
},
{
"epoch": 0.16049447383583887,
"grad_norm": 0.2753854990005493,
"learning_rate": 0.00019891886663013527,
"loss": 0.0173,
"step": 766
},
{
"epoch": 0.16070399664763502,
"grad_norm": 0.11987002938985825,
"learning_rate": 0.00019891562343076315,
"loss": 0.0028,
"step": 767
},
{
"epoch": 0.16091351945943114,
"grad_norm": 0.15542055666446686,
"learning_rate": 0.00019891237540069394,
"loss": 0.0242,
"step": 768
},
{
"epoch": 0.16112304227122728,
"grad_norm": 0.2240624576807022,
"learning_rate": 0.00019890912254008628,
"loss": 0.0149,
"step": 769
},
{
"epoch": 0.1613325650830234,
"grad_norm": 0.14880992472171783,
"learning_rate": 0.00019890586484909895,
"loss": 0.0105,
"step": 770
},
{
"epoch": 0.16154208789481955,
"grad_norm": 0.18625305593013763,
"learning_rate": 0.00019890260232789114,
"loss": 0.0144,
"step": 771
},
{
"epoch": 0.16175161070661567,
"grad_norm": 0.03506407141685486,
"learning_rate": 0.00019889933497662206,
"loss": 0.0004,
"step": 772
},
{
"epoch": 0.16196113351841182,
"grad_norm": 0.2667176127433777,
"learning_rate": 0.00019889606279545143,
"loss": 0.0252,
"step": 773
},
{
"epoch": 0.16217065633020794,
"grad_norm": 0.32680612802505493,
"learning_rate": 0.00019889278578453896,
"loss": 0.0284,
"step": 774
},
{
"epoch": 0.1623801791420041,
"grad_norm": 3.199450731277466,
"learning_rate": 0.0001988895039440447,
"loss": 0.0165,
"step": 775
},
{
"epoch": 0.1625897019538002,
"grad_norm": 0.04759670048952103,
"learning_rate": 0.00019888621727412892,
"loss": 0.0006,
"step": 776
},
{
"epoch": 0.16279922476559636,
"grad_norm": 0.21335595846176147,
"learning_rate": 0.00019888292577495214,
"loss": 0.0064,
"step": 777
},
{
"epoch": 0.16300874757739248,
"grad_norm": 0.01456568855792284,
"learning_rate": 0.0001988796294466751,
"loss": 0.0003,
"step": 778
},
{
"epoch": 0.16321827038918862,
"grad_norm": 0.12444102019071579,
"learning_rate": 0.00019887632828945878,
"loss": 0.0105,
"step": 779
},
{
"epoch": 0.16342779320098474,
"grad_norm": 0.3305560350418091,
"learning_rate": 0.0001988730223034644,
"loss": 0.0722,
"step": 780
},
{
"epoch": 0.1636373160127809,
"grad_norm": 0.2740437686443329,
"learning_rate": 0.00019886971148885344,
"loss": 0.0282,
"step": 781
},
{
"epoch": 0.163846838824577,
"grad_norm": 0.34872737526893616,
"learning_rate": 0.00019886639584578753,
"loss": 0.0281,
"step": 782
},
{
"epoch": 0.16405636163637316,
"grad_norm": 0.03285228833556175,
"learning_rate": 0.00019886307537442865,
"loss": 0.0003,
"step": 783
},
{
"epoch": 0.1642658844481693,
"grad_norm": 0.15017957985401154,
"learning_rate": 0.00019885975007493892,
"loss": 0.004,
"step": 784
},
{
"epoch": 0.16447540725996543,
"grad_norm": 0.14671967923641205,
"learning_rate": 0.00019885641994748076,
"loss": 0.0045,
"step": 785
},
{
"epoch": 0.16468493007176158,
"grad_norm": 0.009040172211825848,
"learning_rate": 0.0001988530849922168,
"loss": 0.0002,
"step": 786
},
{
"epoch": 0.1648944528835577,
"grad_norm": 0.1735561639070511,
"learning_rate": 0.0001988497452093099,
"loss": 0.0139,
"step": 787
},
{
"epoch": 0.16510397569535384,
"grad_norm": 0.1698157787322998,
"learning_rate": 0.00019884640059892315,
"loss": 0.0077,
"step": 788
},
{
"epoch": 0.16531349850714996,
"grad_norm": 0.0034829252399504185,
"learning_rate": 0.0001988430511612199,
"loss": 0.0001,
"step": 789
},
{
"epoch": 0.1655230213189461,
"grad_norm": 0.01388998981565237,
"learning_rate": 0.00019883969689636378,
"loss": 0.0001,
"step": 790
},
{
"epoch": 0.16573254413074223,
"grad_norm": 0.19312739372253418,
"learning_rate": 0.00019883633780451852,
"loss": 0.0028,
"step": 791
},
{
"epoch": 0.16594206694253838,
"grad_norm": 0.31580138206481934,
"learning_rate": 0.00019883297388584818,
"loss": 0.0243,
"step": 792
},
{
"epoch": 0.1661515897543345,
"grad_norm": 0.017013167962431908,
"learning_rate": 0.00019882960514051707,
"loss": 0.0002,
"step": 793
},
{
"epoch": 0.16636111256613065,
"grad_norm": 0.34026068449020386,
"learning_rate": 0.0001988262315686897,
"loss": 0.0391,
"step": 794
},
{
"epoch": 0.16657063537792677,
"grad_norm": 0.2697843313217163,
"learning_rate": 0.00019882285317053083,
"loss": 0.0469,
"step": 795
},
{
"epoch": 0.16678015818972292,
"grad_norm": 0.28238460421562195,
"learning_rate": 0.0001988194699462054,
"loss": 0.0299,
"step": 796
},
{
"epoch": 0.16698968100151904,
"grad_norm": 0.18721148371696472,
"learning_rate": 0.00019881608189587874,
"loss": 0.0051,
"step": 797
},
{
"epoch": 0.16719920381331518,
"grad_norm": 0.37057632207870483,
"learning_rate": 0.0001988126890197162,
"loss": 0.0385,
"step": 798
},
{
"epoch": 0.1674087266251113,
"grad_norm": 0.15513499081134796,
"learning_rate": 0.00019880929131788354,
"loss": 0.002,
"step": 799
},
{
"epoch": 0.16761824943690745,
"grad_norm": 0.16879650950431824,
"learning_rate": 0.00019880588879054667,
"loss": 0.0015,
"step": 800
},
{
"epoch": 0.16761824943690745,
"eval_loss": 0.019750364124774933,
"eval_runtime": 89.1527,
"eval_samples_per_second": 8.614,
"eval_steps_per_second": 4.307,
"step": 800
},
{
"epoch": 0.16782777224870357,
"grad_norm": 0.024027680978178978,
"learning_rate": 0.00019880248143787176,
"loss": 0.0003,
"step": 801
},
{
"epoch": 0.16803729506049972,
"grad_norm": 0.5421876907348633,
"learning_rate": 0.00019879906926002522,
"loss": 0.0324,
"step": 802
},
{
"epoch": 0.16824681787229584,
"grad_norm": 0.16251395642757416,
"learning_rate": 0.00019879565225717366,
"loss": 0.003,
"step": 803
},
{
"epoch": 0.168456340684092,
"grad_norm": 0.1223423108458519,
"learning_rate": 0.00019879223042948402,
"loss": 0.0101,
"step": 804
},
{
"epoch": 0.1686658634958881,
"grad_norm": 0.21323543787002563,
"learning_rate": 0.00019878880377712337,
"loss": 0.0149,
"step": 805
},
{
"epoch": 0.16887538630768426,
"grad_norm": 0.2671852707862854,
"learning_rate": 0.00019878537230025902,
"loss": 0.0441,
"step": 806
},
{
"epoch": 0.16908490911948038,
"grad_norm": 0.46291500329971313,
"learning_rate": 0.0001987819359990586,
"loss": 0.026,
"step": 807
},
{
"epoch": 0.16929443193127652,
"grad_norm": 0.011554433964192867,
"learning_rate": 0.00019877849487368992,
"loss": 0.0002,
"step": 808
},
{
"epoch": 0.16950395474307264,
"grad_norm": 0.16879846155643463,
"learning_rate": 0.00019877504892432102,
"loss": 0.0032,
"step": 809
},
{
"epoch": 0.1697134775548688,
"grad_norm": 0.13901455700397491,
"learning_rate": 0.0001987715981511202,
"loss": 0.0015,
"step": 810
},
{
"epoch": 0.1699230003666649,
"grad_norm": 0.17090804874897003,
"learning_rate": 0.000198768142554256,
"loss": 0.0182,
"step": 811
},
{
"epoch": 0.17013252317846106,
"grad_norm": 0.2233785092830658,
"learning_rate": 0.0001987646821338971,
"loss": 0.0218,
"step": 812
},
{
"epoch": 0.17034204599025718,
"grad_norm": 0.05967244133353233,
"learning_rate": 0.00019876121689021257,
"loss": 0.0007,
"step": 813
},
{
"epoch": 0.17055156880205333,
"grad_norm": 0.16300848126411438,
"learning_rate": 0.00019875774682337163,
"loss": 0.0034,
"step": 814
},
{
"epoch": 0.17076109161384945,
"grad_norm": 0.11895183473825455,
"learning_rate": 0.00019875427193354373,
"loss": 0.0016,
"step": 815
},
{
"epoch": 0.1709706144256456,
"grad_norm": 0.17483823001384735,
"learning_rate": 0.00019875079222089862,
"loss": 0.0046,
"step": 816
},
{
"epoch": 0.17118013723744172,
"grad_norm": 0.16585186123847961,
"learning_rate": 0.00019874730768560617,
"loss": 0.0069,
"step": 817
},
{
"epoch": 0.17138966004923786,
"grad_norm": 0.017436591908335686,
"learning_rate": 0.0001987438183278366,
"loss": 0.0002,
"step": 818
},
{
"epoch": 0.17159918286103398,
"grad_norm": 0.004752448294311762,
"learning_rate": 0.00019874032414776027,
"loss": 0.0001,
"step": 819
},
{
"epoch": 0.17180870567283013,
"grad_norm": 0.1625145524740219,
"learning_rate": 0.00019873682514554787,
"loss": 0.0122,
"step": 820
},
{
"epoch": 0.17201822848462625,
"grad_norm": 0.16962960362434387,
"learning_rate": 0.00019873332132137026,
"loss": 0.0187,
"step": 821
},
{
"epoch": 0.1722277512964224,
"grad_norm": 0.20390696823596954,
"learning_rate": 0.00019872981267539858,
"loss": 0.0171,
"step": 822
},
{
"epoch": 0.17243727410821852,
"grad_norm": 0.36607325077056885,
"learning_rate": 0.0001987262992078041,
"loss": 0.0434,
"step": 823
},
{
"epoch": 0.17264679692001467,
"grad_norm": 0.24736827611923218,
"learning_rate": 0.00019872278091875848,
"loss": 0.0229,
"step": 824
},
{
"epoch": 0.17285631973181081,
"grad_norm": 0.09645625203847885,
"learning_rate": 0.00019871925780843354,
"loss": 0.0014,
"step": 825
},
{
"epoch": 0.17306584254360693,
"grad_norm": 0.17568598687648773,
"learning_rate": 0.0001987157298770013,
"loss": 0.0157,
"step": 826
},
{
"epoch": 0.17327536535540308,
"grad_norm": 0.11099965870380402,
"learning_rate": 0.00019871219712463408,
"loss": 0.0009,
"step": 827
},
{
"epoch": 0.1734848881671992,
"grad_norm": 0.2161271870136261,
"learning_rate": 0.00019870865955150442,
"loss": 0.0186,
"step": 828
},
{
"epoch": 0.17369441097899535,
"grad_norm": 0.3481244742870331,
"learning_rate": 0.00019870511715778505,
"loss": 0.0241,
"step": 829
},
{
"epoch": 0.17390393379079147,
"grad_norm": 0.30809855461120605,
"learning_rate": 0.000198701569943649,
"loss": 0.0445,
"step": 830
},
{
"epoch": 0.17411345660258762,
"grad_norm": 0.07733512669801712,
"learning_rate": 0.00019869801790926945,
"loss": 0.0008,
"step": 831
},
{
"epoch": 0.17432297941438374,
"grad_norm": 0.207721546292305,
"learning_rate": 0.0001986944610548199,
"loss": 0.0131,
"step": 832
},
{
"epoch": 0.1745325022261799,
"grad_norm": 0.20278921723365784,
"learning_rate": 0.0001986908993804741,
"loss": 0.0237,
"step": 833
},
{
"epoch": 0.174742025037976,
"grad_norm": 0.28272855281829834,
"learning_rate": 0.00019868733288640588,
"loss": 0.0262,
"step": 834
},
{
"epoch": 0.17495154784977215,
"grad_norm": 0.35000142455101013,
"learning_rate": 0.00019868376157278954,
"loss": 0.0208,
"step": 835
},
{
"epoch": 0.17516107066156827,
"grad_norm": 0.2046387940645218,
"learning_rate": 0.0001986801854397994,
"loss": 0.0075,
"step": 836
},
{
"epoch": 0.17537059347336442,
"grad_norm": 0.0028248566668480635,
"learning_rate": 0.00019867660448761017,
"loss": 0.0001,
"step": 837
},
{
"epoch": 0.17558011628516054,
"grad_norm": 0.26853737235069275,
"learning_rate": 0.00019867301871639666,
"loss": 0.0161,
"step": 838
},
{
"epoch": 0.1757896390969567,
"grad_norm": 0.18713527917861938,
"learning_rate": 0.00019866942812633406,
"loss": 0.0167,
"step": 839
},
{
"epoch": 0.1759991619087528,
"grad_norm": 0.12379568815231323,
"learning_rate": 0.0001986658327175977,
"loss": 0.0114,
"step": 840
},
{
"epoch": 0.17620868472054896,
"grad_norm": 0.04250200465321541,
"learning_rate": 0.00019866223249036312,
"loss": 0.0008,
"step": 841
},
{
"epoch": 0.17641820753234508,
"grad_norm": 0.1923532485961914,
"learning_rate": 0.0001986586274448062,
"loss": 0.0127,
"step": 842
},
{
"epoch": 0.17662773034414123,
"grad_norm": 0.1902576982975006,
"learning_rate": 0.00019865501758110297,
"loss": 0.0052,
"step": 843
},
{
"epoch": 0.17683725315593735,
"grad_norm": 0.27607104182243347,
"learning_rate": 0.00019865140289942974,
"loss": 0.0629,
"step": 844
},
{
"epoch": 0.1770467759677335,
"grad_norm": 0.286934494972229,
"learning_rate": 0.00019864778339996302,
"loss": 0.0253,
"step": 845
},
{
"epoch": 0.17725629877952961,
"grad_norm": 0.19526824355125427,
"learning_rate": 0.00019864415908287956,
"loss": 0.0053,
"step": 846
},
{
"epoch": 0.17746582159132576,
"grad_norm": 0.21397161483764648,
"learning_rate": 0.00019864052994835642,
"loss": 0.0219,
"step": 847
},
{
"epoch": 0.17767534440312188,
"grad_norm": 0.2589309811592102,
"learning_rate": 0.00019863689599657077,
"loss": 0.0239,
"step": 848
},
{
"epoch": 0.17788486721491803,
"grad_norm": 4.077011585235596,
"learning_rate": 0.0001986332572277001,
"loss": 0.0057,
"step": 849
},
{
"epoch": 0.17809439002671415,
"grad_norm": 0.3254491090774536,
"learning_rate": 0.00019862961364192216,
"loss": 0.0259,
"step": 850
},
{
"epoch": 0.1783039128385103,
"grad_norm": 0.00824171956628561,
"learning_rate": 0.0001986259652394148,
"loss": 0.0001,
"step": 851
},
{
"epoch": 0.17851343565030642,
"grad_norm": 0.1307021528482437,
"learning_rate": 0.0001986223120203563,
"loss": 0.0033,
"step": 852
},
{
"epoch": 0.17872295846210257,
"grad_norm": 0.040325507521629333,
"learning_rate": 0.00019861865398492499,
"loss": 0.0005,
"step": 853
},
{
"epoch": 0.17893248127389869,
"grad_norm": 0.2458508163690567,
"learning_rate": 0.00019861499113329954,
"loss": 0.0339,
"step": 854
},
{
"epoch": 0.17914200408569483,
"grad_norm": 0.4500856101512909,
"learning_rate": 0.00019861132346565884,
"loss": 0.0344,
"step": 855
},
{
"epoch": 0.17935152689749095,
"grad_norm": 0.09408875554800034,
"learning_rate": 0.000198607650982182,
"loss": 0.0028,
"step": 856
},
{
"epoch": 0.1795610497092871,
"grad_norm": 0.00868634507060051,
"learning_rate": 0.00019860397368304838,
"loss": 0.0001,
"step": 857
},
{
"epoch": 0.17977057252108322,
"grad_norm": 0.042256616055965424,
"learning_rate": 0.00019860029156843753,
"loss": 0.0006,
"step": 858
},
{
"epoch": 0.17998009533287937,
"grad_norm": 0.1415722370147705,
"learning_rate": 0.00019859660463852932,
"loss": 0.0025,
"step": 859
},
{
"epoch": 0.1801896181446755,
"grad_norm": 0.26225030422210693,
"learning_rate": 0.0001985929128935038,
"loss": 0.0264,
"step": 860
},
{
"epoch": 0.18039914095647164,
"grad_norm": 0.4579645097255707,
"learning_rate": 0.00019858921633354122,
"loss": 0.0695,
"step": 861
},
{
"epoch": 0.18060866376826776,
"grad_norm": 0.10728690028190613,
"learning_rate": 0.00019858551495882215,
"loss": 0.0097,
"step": 862
},
{
"epoch": 0.1808181865800639,
"grad_norm": 0.01471271738409996,
"learning_rate": 0.0001985818087695273,
"loss": 0.0002,
"step": 863
},
{
"epoch": 0.18102770939186003,
"grad_norm": 0.107490174472332,
"learning_rate": 0.00019857809776583775,
"loss": 0.0113,
"step": 864
},
{
"epoch": 0.18123723220365617,
"grad_norm": 0.13018178939819336,
"learning_rate": 0.0001985743819479347,
"loss": 0.0014,
"step": 865
},
{
"epoch": 0.18144675501545232,
"grad_norm": 0.3460003137588501,
"learning_rate": 0.00019857066131599957,
"loss": 0.0389,
"step": 866
},
{
"epoch": 0.18165627782724844,
"grad_norm": 0.3483485281467438,
"learning_rate": 0.0001985669358702141,
"loss": 0.0466,
"step": 867
},
{
"epoch": 0.1818658006390446,
"grad_norm": 0.13328337669372559,
"learning_rate": 0.00019856320561076026,
"loss": 0.001,
"step": 868
},
{
"epoch": 0.1820753234508407,
"grad_norm": 0.24165037274360657,
"learning_rate": 0.00019855947053782017,
"loss": 0.0193,
"step": 869
},
{
"epoch": 0.18228484626263686,
"grad_norm": 0.036382876336574554,
"learning_rate": 0.00019855573065157621,
"loss": 0.0004,
"step": 870
},
{
"epoch": 0.18249436907443298,
"grad_norm": 0.234747052192688,
"learning_rate": 0.00019855198595221114,
"loss": 0.012,
"step": 871
},
{
"epoch": 0.18270389188622913,
"grad_norm": 0.19845280051231384,
"learning_rate": 0.00019854823643990774,
"loss": 0.0138,
"step": 872
},
{
"epoch": 0.18291341469802525,
"grad_norm": 0.05412408709526062,
"learning_rate": 0.00019854448211484917,
"loss": 0.0006,
"step": 873
},
{
"epoch": 0.1831229375098214,
"grad_norm": 0.15850593149662018,
"learning_rate": 0.00019854072297721876,
"loss": 0.012,
"step": 874
},
{
"epoch": 0.1833324603216175,
"grad_norm": 0.535932183265686,
"learning_rate": 0.00019853695902720007,
"loss": 0.041,
"step": 875
},
{
"epoch": 0.18354198313341366,
"grad_norm": 0.4136875569820404,
"learning_rate": 0.00019853319026497696,
"loss": 0.0517,
"step": 876
},
{
"epoch": 0.18375150594520978,
"grad_norm": 0.1701594591140747,
"learning_rate": 0.00019852941669073343,
"loss": 0.0016,
"step": 877
},
{
"epoch": 0.18396102875700593,
"grad_norm": 0.3235994875431061,
"learning_rate": 0.00019852563830465383,
"loss": 0.013,
"step": 878
},
{
"epoch": 0.18417055156880205,
"grad_norm": 0.13185402750968933,
"learning_rate": 0.00019852185510692264,
"loss": 0.0011,
"step": 879
},
{
"epoch": 0.1843800743805982,
"grad_norm": 0.4098699986934662,
"learning_rate": 0.0001985180670977247,
"loss": 0.015,
"step": 880
},
{
"epoch": 0.18458959719239432,
"grad_norm": 0.048046402633190155,
"learning_rate": 0.00019851427427724488,
"loss": 0.0004,
"step": 881
},
{
"epoch": 0.18479912000419046,
"grad_norm": 0.3164336383342743,
"learning_rate": 0.00019851047664566846,
"loss": 0.0679,
"step": 882
},
{
"epoch": 0.18500864281598658,
"grad_norm": 0.41321858763694763,
"learning_rate": 0.00019850667420318093,
"loss": 0.0139,
"step": 883
},
{
"epoch": 0.18521816562778273,
"grad_norm": 0.07493787258863449,
"learning_rate": 0.00019850286694996796,
"loss": 0.0162,
"step": 884
},
{
"epoch": 0.18542768843957885,
"grad_norm": 0.08418352156877518,
"learning_rate": 0.0001984990548862155,
"loss": 0.001,
"step": 885
},
{
"epoch": 0.185637211251375,
"grad_norm": 0.19334886968135834,
"learning_rate": 0.00019849523801210973,
"loss": 0.0291,
"step": 886
},
{
"epoch": 0.18584673406317112,
"grad_norm": 0.05708039924502373,
"learning_rate": 0.000198491416327837,
"loss": 0.0006,
"step": 887
},
{
"epoch": 0.18605625687496727,
"grad_norm": 0.1267891228199005,
"learning_rate": 0.00019848758983358398,
"loss": 0.0124,
"step": 888
},
{
"epoch": 0.1862657796867634,
"grad_norm": 0.3097265362739563,
"learning_rate": 0.00019848375852953755,
"loss": 0.02,
"step": 889
},
{
"epoch": 0.18647530249855954,
"grad_norm": 0.19478236138820648,
"learning_rate": 0.00019847992241588483,
"loss": 0.0292,
"step": 890
},
{
"epoch": 0.18668482531035566,
"grad_norm": 0.13891831040382385,
"learning_rate": 0.0001984760814928131,
"loss": 0.0112,
"step": 891
},
{
"epoch": 0.1868943481221518,
"grad_norm": 0.09047941863536835,
"learning_rate": 0.00019847223576050998,
"loss": 0.0055,
"step": 892
},
{
"epoch": 0.18710387093394792,
"grad_norm": 0.42029663920402527,
"learning_rate": 0.00019846838521916333,
"loss": 0.0313,
"step": 893
},
{
"epoch": 0.18731339374574407,
"grad_norm": 0.15446271002292633,
"learning_rate": 0.00019846452986896113,
"loss": 0.0016,
"step": 894
},
{
"epoch": 0.1875229165575402,
"grad_norm": 0.21927279233932495,
"learning_rate": 0.00019846066971009168,
"loss": 0.0187,
"step": 895
},
{
"epoch": 0.18773243936933634,
"grad_norm": 0.20879912376403809,
"learning_rate": 0.00019845680474274348,
"loss": 0.0158,
"step": 896
},
{
"epoch": 0.18794196218113246,
"grad_norm": 0.13854913413524628,
"learning_rate": 0.00019845293496710534,
"loss": 0.011,
"step": 897
},
{
"epoch": 0.1881514849929286,
"grad_norm": 0.16847798228263855,
"learning_rate": 0.00019844906038336615,
"loss": 0.0111,
"step": 898
},
{
"epoch": 0.18836100780472473,
"grad_norm": 0.07531256973743439,
"learning_rate": 0.00019844518099171524,
"loss": 0.0009,
"step": 899
},
{
"epoch": 0.18857053061652088,
"grad_norm": 0.10310601443052292,
"learning_rate": 0.000198441296792342,
"loss": 0.0027,
"step": 900
},
{
"epoch": 0.188780053428317,
"grad_norm": 0.2742733955383301,
"learning_rate": 0.00019843740778543613,
"loss": 0.0088,
"step": 901
},
{
"epoch": 0.18898957624011314,
"grad_norm": 0.06786957383155823,
"learning_rate": 0.00019843351397118755,
"loss": 0.0009,
"step": 902
},
{
"epoch": 0.18919909905190926,
"grad_norm": 0.208217591047287,
"learning_rate": 0.00019842961534978645,
"loss": 0.0268,
"step": 903
},
{
"epoch": 0.1894086218637054,
"grad_norm": 0.028746692463755608,
"learning_rate": 0.0001984257119214232,
"loss": 0.0004,
"step": 904
},
{
"epoch": 0.18961814467550153,
"grad_norm": 0.27264735102653503,
"learning_rate": 0.00019842180368628842,
"loss": 0.0179,
"step": 905
},
{
"epoch": 0.18982766748729768,
"grad_norm": 0.023868918418884277,
"learning_rate": 0.00019841789064457298,
"loss": 0.0003,
"step": 906
},
{
"epoch": 0.1900371902990938,
"grad_norm": 0.030625615268945694,
"learning_rate": 0.000198413972796468,
"loss": 0.0004,
"step": 907
},
{
"epoch": 0.19024671311088995,
"grad_norm": 0.12535369396209717,
"learning_rate": 0.00019841005014216485,
"loss": 0.0052,
"step": 908
},
{
"epoch": 0.1904562359226861,
"grad_norm": 0.08116110414266586,
"learning_rate": 0.000198406122681855,
"loss": 0.0003,
"step": 909
},
{
"epoch": 0.19066575873448222,
"grad_norm": 0.22774437069892883,
"learning_rate": 0.00019840219041573033,
"loss": 0.0151,
"step": 910
},
{
"epoch": 0.19087528154627836,
"grad_norm": 0.22736741602420807,
"learning_rate": 0.00019839825334398286,
"loss": 0.0197,
"step": 911
},
{
"epoch": 0.19108480435807448,
"grad_norm": 0.012811636552214622,
"learning_rate": 0.00019839431146680486,
"loss": 0.0002,
"step": 912
},
{
"epoch": 0.19129432716987063,
"grad_norm": 0.23712413012981415,
"learning_rate": 0.00019839036478438882,
"loss": 0.0169,
"step": 913
},
{
"epoch": 0.19150384998166675,
"grad_norm": 0.08614616841077805,
"learning_rate": 0.0001983864132969275,
"loss": 0.0013,
"step": 914
},
{
"epoch": 0.1917133727934629,
"grad_norm": 0.23105621337890625,
"learning_rate": 0.00019838245700461385,
"loss": 0.0118,
"step": 915
},
{
"epoch": 0.19192289560525902,
"grad_norm": 0.14425182342529297,
"learning_rate": 0.00019837849590764115,
"loss": 0.0145,
"step": 916
},
{
"epoch": 0.19213241841705517,
"grad_norm": 0.24712929129600525,
"learning_rate": 0.0001983745300062028,
"loss": 0.0194,
"step": 917
},
{
"epoch": 0.1923419412288513,
"grad_norm": 0.32611796259880066,
"learning_rate": 0.00019837055930049243,
"loss": 0.0407,
"step": 918
},
{
"epoch": 0.19255146404064744,
"grad_norm": 0.22980710864067078,
"learning_rate": 0.00019836658379070407,
"loss": 0.0424,
"step": 919
},
{
"epoch": 0.19276098685244356,
"grad_norm": 0.35468605160713196,
"learning_rate": 0.0001983626034770318,
"loss": 0.019,
"step": 920
},
{
"epoch": 0.1929705096642397,
"grad_norm": 0.006418622564524412,
"learning_rate": 0.00019835861835967001,
"loss": 0.0001,
"step": 921
},
{
"epoch": 0.19318003247603582,
"grad_norm": 0.34932658076286316,
"learning_rate": 0.0001983546284388133,
"loss": 0.084,
"step": 922
},
{
"epoch": 0.19338955528783197,
"grad_norm": 0.034760601818561554,
"learning_rate": 0.00019835063371465657,
"loss": 0.0004,
"step": 923
},
{
"epoch": 0.1935990780996281,
"grad_norm": 0.2593575716018677,
"learning_rate": 0.0001983466341873949,
"loss": 0.0178,
"step": 924
},
{
"epoch": 0.19380860091142424,
"grad_norm": 0.05930793285369873,
"learning_rate": 0.00019834262985722357,
"loss": 0.0013,
"step": 925
},
{
"epoch": 0.19401812372322036,
"grad_norm": 0.17973913252353668,
"learning_rate": 0.00019833862072433816,
"loss": 0.0159,
"step": 926
},
{
"epoch": 0.1942276465350165,
"grad_norm": 1.3181376457214355,
"learning_rate": 0.0001983346067889345,
"loss": 0.0079,
"step": 927
},
{
"epoch": 0.19443716934681263,
"grad_norm": 0.01102768536657095,
"learning_rate": 0.00019833058805120856,
"loss": 0.0001,
"step": 928
},
{
"epoch": 0.19464669215860878,
"grad_norm": 0.3218340575695038,
"learning_rate": 0.00019832656451135666,
"loss": 0.0682,
"step": 929
},
{
"epoch": 0.1948562149704049,
"grad_norm": 0.029258154332637787,
"learning_rate": 0.00019832253616957523,
"loss": 0.0005,
"step": 930
},
{
"epoch": 0.19506573778220104,
"grad_norm": 0.029073555022478104,
"learning_rate": 0.00019831850302606103,
"loss": 0.0003,
"step": 931
},
{
"epoch": 0.19527526059399716,
"grad_norm": 0.2904273271560669,
"learning_rate": 0.00019831446508101103,
"loss": 0.0159,
"step": 932
},
{
"epoch": 0.1954847834057933,
"grad_norm": 0.1123453825712204,
"learning_rate": 0.00019831042233462245,
"loss": 0.0049,
"step": 933
},
{
"epoch": 0.19569430621758943,
"grad_norm": 0.11524815112352371,
"learning_rate": 0.00019830637478709267,
"loss": 0.0157,
"step": 934
},
{
"epoch": 0.19590382902938558,
"grad_norm": 0.23699933290481567,
"learning_rate": 0.00019830232243861942,
"loss": 0.0071,
"step": 935
},
{
"epoch": 0.1961133518411817,
"grad_norm": 0.1821090430021286,
"learning_rate": 0.00019829826528940057,
"loss": 0.0137,
"step": 936
},
{
"epoch": 0.19632287465297785,
"grad_norm": 0.35947421193122864,
"learning_rate": 0.00019829420333963426,
"loss": 0.0265,
"step": 937
},
{
"epoch": 0.19653239746477397,
"grad_norm": 0.22742979228496552,
"learning_rate": 0.00019829013658951883,
"loss": 0.017,
"step": 938
},
{
"epoch": 0.19674192027657011,
"grad_norm": 0.06752747297286987,
"learning_rate": 0.00019828606503925292,
"loss": 0.0013,
"step": 939
},
{
"epoch": 0.19695144308836623,
"grad_norm": 0.03469230234622955,
"learning_rate": 0.0001982819886890354,
"loss": 0.0003,
"step": 940
},
{
"epoch": 0.19716096590016238,
"grad_norm": 0.19584067165851593,
"learning_rate": 0.0001982779075390653,
"loss": 0.0072,
"step": 941
},
{
"epoch": 0.1973704887119585,
"grad_norm": 0.08708450198173523,
"learning_rate": 0.00019827382158954193,
"loss": 0.0091,
"step": 942
},
{
"epoch": 0.19758001152375465,
"grad_norm": 0.22899965941905975,
"learning_rate": 0.00019826973084066484,
"loss": 0.0155,
"step": 943
},
{
"epoch": 0.19778953433555077,
"grad_norm": 0.2735210061073303,
"learning_rate": 0.00019826563529263383,
"loss": 0.0062,
"step": 944
},
{
"epoch": 0.19799905714734692,
"grad_norm": 0.13284745812416077,
"learning_rate": 0.00019826153494564887,
"loss": 0.0024,
"step": 945
},
{
"epoch": 0.19820857995914304,
"grad_norm": 0.24890956282615662,
"learning_rate": 0.00019825742979991026,
"loss": 0.027,
"step": 946
},
{
"epoch": 0.1984181027709392,
"grad_norm": 0.2778416574001312,
"learning_rate": 0.00019825331985561846,
"loss": 0.0321,
"step": 947
},
{
"epoch": 0.1986276255827353,
"grad_norm": 0.17397484183311462,
"learning_rate": 0.00019824920511297416,
"loss": 0.0154,
"step": 948
},
{
"epoch": 0.19883714839453145,
"grad_norm": 0.08111430704593658,
"learning_rate": 0.0001982450855721783,
"loss": 0.0011,
"step": 949
},
{
"epoch": 0.1990466712063276,
"grad_norm": 0.11453194171190262,
"learning_rate": 0.00019824096123343212,
"loss": 0.0059,
"step": 950
},
{
"epoch": 0.19925619401812372,
"grad_norm": 0.17547598481178284,
"learning_rate": 0.00019823683209693705,
"loss": 0.0062,
"step": 951
},
{
"epoch": 0.19946571682991987,
"grad_norm": 0.02609160542488098,
"learning_rate": 0.00019823269816289464,
"loss": 0.0002,
"step": 952
},
{
"epoch": 0.199675239641716,
"grad_norm": 0.41189810633659363,
"learning_rate": 0.00019822855943150687,
"loss": 0.091,
"step": 953
},
{
"epoch": 0.19988476245351214,
"grad_norm": 0.1962253600358963,
"learning_rate": 0.00019822441590297581,
"loss": 0.0239,
"step": 954
},
{
"epoch": 0.20009428526530826,
"grad_norm": 0.0019026676891371608,
"learning_rate": 0.00019822026757750386,
"loss": 0.0,
"step": 955
},
{
"epoch": 0.2003038080771044,
"grad_norm": 0.29771414399147034,
"learning_rate": 0.0001982161144552936,
"loss": 0.0503,
"step": 956
},
{
"epoch": 0.20051333088890053,
"grad_norm": 0.12233370542526245,
"learning_rate": 0.00019821195653654784,
"loss": 0.0018,
"step": 957
},
{
"epoch": 0.20072285370069667,
"grad_norm": 0.022733421996235847,
"learning_rate": 0.0001982077938214696,
"loss": 0.0002,
"step": 958
},
{
"epoch": 0.2009323765124928,
"grad_norm": 0.30897054076194763,
"learning_rate": 0.00019820362631026224,
"loss": 0.0256,
"step": 959
},
{
"epoch": 0.20114189932428894,
"grad_norm": 0.16434000432491302,
"learning_rate": 0.00019819945400312926,
"loss": 0.0209,
"step": 960
},
{
"epoch": 0.20135142213608506,
"grad_norm": 0.14610372483730316,
"learning_rate": 0.0001981952769002744,
"loss": 0.0011,
"step": 961
},
{
"epoch": 0.2015609449478812,
"grad_norm": 0.27273574471473694,
"learning_rate": 0.0001981910950019017,
"loss": 0.0284,
"step": 962
},
{
"epoch": 0.20177046775967733,
"grad_norm": 0.1361648440361023,
"learning_rate": 0.0001981869083082154,
"loss": 0.0065,
"step": 963
},
{
"epoch": 0.20197999057147348,
"grad_norm": 0.1183745265007019,
"learning_rate": 0.00019818271681941986,
"loss": 0.0034,
"step": 964
},
{
"epoch": 0.2021895133832696,
"grad_norm": 0.2849377989768982,
"learning_rate": 0.0001981785205357199,
"loss": 0.0119,
"step": 965
},
{
"epoch": 0.20239903619506575,
"grad_norm": 0.013914634473621845,
"learning_rate": 0.00019817431945732034,
"loss": 0.0001,
"step": 966
},
{
"epoch": 0.20260855900686187,
"grad_norm": 0.17100666463375092,
"learning_rate": 0.00019817011358442646,
"loss": 0.0134,
"step": 967
},
{
"epoch": 0.202818081818658,
"grad_norm": 0.3089207112789154,
"learning_rate": 0.00019816590291724356,
"loss": 0.0347,
"step": 968
},
{
"epoch": 0.20302760463045413,
"grad_norm": 0.08947692811489105,
"learning_rate": 0.00019816168745597735,
"loss": 0.0006,
"step": 969
},
{
"epoch": 0.20323712744225028,
"grad_norm": 0.3102778196334839,
"learning_rate": 0.00019815746720083365,
"loss": 0.0385,
"step": 970
},
{
"epoch": 0.2034466502540464,
"grad_norm": 0.3651212751865387,
"learning_rate": 0.0001981532421520186,
"loss": 0.0431,
"step": 971
},
{
"epoch": 0.20365617306584255,
"grad_norm": 0.14294616878032684,
"learning_rate": 0.0001981490123097385,
"loss": 0.0048,
"step": 972
},
{
"epoch": 0.20386569587763867,
"grad_norm": 0.2009478658437729,
"learning_rate": 0.00019814477767419994,
"loss": 0.0031,
"step": 973
},
{
"epoch": 0.20407521868943482,
"grad_norm": 0.19270165264606476,
"learning_rate": 0.00019814053824560975,
"loss": 0.0074,
"step": 974
},
{
"epoch": 0.20428474150123094,
"grad_norm": 0.006634276360273361,
"learning_rate": 0.00019813629402417492,
"loss": 0.0001,
"step": 975
},
{
"epoch": 0.20449426431302709,
"grad_norm": 0.17981410026550293,
"learning_rate": 0.00019813204501010275,
"loss": 0.0054,
"step": 976
},
{
"epoch": 0.2047037871248232,
"grad_norm": 0.03605171665549278,
"learning_rate": 0.00019812779120360072,
"loss": 0.0004,
"step": 977
},
{
"epoch": 0.20491330993661935,
"grad_norm": 0.09814287722110748,
"learning_rate": 0.00019812353260487662,
"loss": 0.0005,
"step": 978
},
{
"epoch": 0.20512283274841547,
"grad_norm": 0.41026777029037476,
"learning_rate": 0.0001981192692141384,
"loss": 0.0149,
"step": 979
},
{
"epoch": 0.20533235556021162,
"grad_norm": 0.2844783067703247,
"learning_rate": 0.00019811500103159428,
"loss": 0.039,
"step": 980
},
{
"epoch": 0.20554187837200774,
"grad_norm": 0.03213176503777504,
"learning_rate": 0.00019811072805745266,
"loss": 0.0003,
"step": 981
},
{
"epoch": 0.2057514011838039,
"grad_norm": 0.012578954920172691,
"learning_rate": 0.00019810645029192227,
"loss": 0.0002,
"step": 982
},
{
"epoch": 0.2059609239956,
"grad_norm": 0.18747162818908691,
"learning_rate": 0.000198102167735212,
"loss": 0.0018,
"step": 983
},
{
"epoch": 0.20617044680739616,
"grad_norm": 0.20987571775913239,
"learning_rate": 0.000198097880387531,
"loss": 0.0023,
"step": 984
},
{
"epoch": 0.20637996961919228,
"grad_norm": 0.16509518027305603,
"learning_rate": 0.00019809358824908862,
"loss": 0.0026,
"step": 985
},
{
"epoch": 0.20658949243098843,
"grad_norm": 0.12439261376857758,
"learning_rate": 0.0001980892913200945,
"loss": 0.0133,
"step": 986
},
{
"epoch": 0.20679901524278455,
"grad_norm": 0.2837631404399872,
"learning_rate": 0.0001980849896007585,
"loss": 0.0058,
"step": 987
},
{
"epoch": 0.2070085380545807,
"grad_norm": 0.22140386700630188,
"learning_rate": 0.0001980806830912907,
"loss": 0.0194,
"step": 988
},
{
"epoch": 0.2072180608663768,
"grad_norm": 0.19810140132904053,
"learning_rate": 0.00019807637179190138,
"loss": 0.0074,
"step": 989
},
{
"epoch": 0.20742758367817296,
"grad_norm": 0.1639014035463333,
"learning_rate": 0.0001980720557028011,
"loss": 0.0039,
"step": 990
},
{
"epoch": 0.2076371064899691,
"grad_norm": 0.2784484624862671,
"learning_rate": 0.00019806773482420066,
"loss": 0.0019,
"step": 991
},
{
"epoch": 0.20784662930176523,
"grad_norm": 0.060419704765081406,
"learning_rate": 0.00019806340915631108,
"loss": 0.0006,
"step": 992
},
{
"epoch": 0.20805615211356138,
"grad_norm": 0.048128314316272736,
"learning_rate": 0.00019805907869934361,
"loss": 0.0005,
"step": 993
},
{
"epoch": 0.2082656749253575,
"grad_norm": 0.005089947488158941,
"learning_rate": 0.0001980547434535097,
"loss": 0.0001,
"step": 994
},
{
"epoch": 0.20847519773715364,
"grad_norm": 0.430545836687088,
"learning_rate": 0.0001980504034190211,
"loss": 0.0808,
"step": 995
},
{
"epoch": 0.20868472054894976,
"grad_norm": 0.023343857377767563,
"learning_rate": 0.00019804605859608977,
"loss": 0.0004,
"step": 996
},
{
"epoch": 0.2088942433607459,
"grad_norm": 0.2382965236902237,
"learning_rate": 0.00019804170898492788,
"loss": 0.0454,
"step": 997
},
{
"epoch": 0.20910376617254203,
"grad_norm": 0.12488018721342087,
"learning_rate": 0.00019803735458574784,
"loss": 0.0089,
"step": 998
},
{
"epoch": 0.20931328898433818,
"grad_norm": 0.005318134557455778,
"learning_rate": 0.00019803299539876233,
"loss": 0.0001,
"step": 999
},
{
"epoch": 0.2095228117961343,
"grad_norm": 0.1782492995262146,
"learning_rate": 0.0001980286314241842,
"loss": 0.0178,
"step": 1000
},
{
"epoch": 0.2095228117961343,
"eval_loss": 0.015503130853176117,
"eval_runtime": 89.0527,
"eval_samples_per_second": 8.624,
"eval_steps_per_second": 4.312,
"step": 1000
},
{
"epoch": 0.20973233460793045,
"grad_norm": 0.24667583405971527,
"learning_rate": 0.00019802426266222662,
"loss": 0.024,
"step": 1001
},
{
"epoch": 0.20994185741972657,
"grad_norm": 0.0044623855501413345,
"learning_rate": 0.0001980198891131029,
"loss": 0.0001,
"step": 1002
},
{
"epoch": 0.21015138023152272,
"grad_norm": 0.2540886402130127,
"learning_rate": 0.00019801551077702664,
"loss": 0.0351,
"step": 1003
},
{
"epoch": 0.21036090304331884,
"grad_norm": 0.014486823230981827,
"learning_rate": 0.00019801112765421172,
"loss": 0.0001,
"step": 1004
},
{
"epoch": 0.21057042585511498,
"grad_norm": 0.19617009162902832,
"learning_rate": 0.0001980067397448721,
"loss": 0.011,
"step": 1005
},
{
"epoch": 0.2107799486669111,
"grad_norm": 0.16247636079788208,
"learning_rate": 0.00019800234704922214,
"loss": 0.0134,
"step": 1006
},
{
"epoch": 0.21098947147870725,
"grad_norm": 0.06774534285068512,
"learning_rate": 0.00019799794956747635,
"loss": 0.0008,
"step": 1007
},
{
"epoch": 0.21119899429050337,
"grad_norm": 0.1491418480873108,
"learning_rate": 0.00019799354729984947,
"loss": 0.0085,
"step": 1008
},
{
"epoch": 0.21140851710229952,
"grad_norm": 0.18873615562915802,
"learning_rate": 0.00019798914024655648,
"loss": 0.011,
"step": 1009
},
{
"epoch": 0.21161803991409564,
"grad_norm": 0.17292487621307373,
"learning_rate": 0.00019798472840781267,
"loss": 0.0201,
"step": 1010
},
{
"epoch": 0.2118275627258918,
"grad_norm": 0.24127565324306488,
"learning_rate": 0.00019798031178383342,
"loss": 0.0448,
"step": 1011
},
{
"epoch": 0.2120370855376879,
"grad_norm": 0.026395272463560104,
"learning_rate": 0.0001979758903748345,
"loss": 0.0006,
"step": 1012
},
{
"epoch": 0.21224660834948406,
"grad_norm": 0.2279701977968216,
"learning_rate": 0.00019797146418103174,
"loss": 0.0087,
"step": 1013
},
{
"epoch": 0.21245613116128018,
"grad_norm": 0.18590952455997467,
"learning_rate": 0.0001979670332026414,
"loss": 0.0011,
"step": 1014
},
{
"epoch": 0.21266565397307632,
"grad_norm": 0.2261616587638855,
"learning_rate": 0.0001979625974398798,
"loss": 0.0333,
"step": 1015
},
{
"epoch": 0.21287517678487244,
"grad_norm": 0.22065399587154388,
"learning_rate": 0.00019795815689296363,
"loss": 0.0042,
"step": 1016
},
{
"epoch": 0.2130846995966686,
"grad_norm": 0.24057383835315704,
"learning_rate": 0.0001979537115621097,
"loss": 0.0184,
"step": 1017
},
{
"epoch": 0.2132942224084647,
"grad_norm": 0.19768278300762177,
"learning_rate": 0.00019794926144753511,
"loss": 0.0179,
"step": 1018
},
{
"epoch": 0.21350374522026086,
"grad_norm": 0.18621546030044556,
"learning_rate": 0.00019794480654945721,
"loss": 0.0066,
"step": 1019
},
{
"epoch": 0.21371326803205698,
"grad_norm": 0.13579484820365906,
"learning_rate": 0.00019794034686809354,
"loss": 0.0081,
"step": 1020
},
{
"epoch": 0.21392279084385313,
"grad_norm": 0.08393269777297974,
"learning_rate": 0.00019793588240366193,
"loss": 0.001,
"step": 1021
},
{
"epoch": 0.21413231365564925,
"grad_norm": 0.12567999958992004,
"learning_rate": 0.00019793141315638038,
"loss": 0.008,
"step": 1022
},
{
"epoch": 0.2143418364674454,
"grad_norm": 0.3008198142051697,
"learning_rate": 0.00019792693912646715,
"loss": 0.0248,
"step": 1023
},
{
"epoch": 0.21455135927924152,
"grad_norm": 0.3744416832923889,
"learning_rate": 0.00019792246031414073,
"loss": 0.025,
"step": 1024
},
{
"epoch": 0.21476088209103766,
"grad_norm": 0.3006001114845276,
"learning_rate": 0.00019791797671961988,
"loss": 0.0275,
"step": 1025
},
{
"epoch": 0.21497040490283378,
"grad_norm": 0.12467708438634872,
"learning_rate": 0.00019791348834312356,
"loss": 0.0128,
"step": 1026
},
{
"epoch": 0.21517992771462993,
"grad_norm": 0.17428621649742126,
"learning_rate": 0.00019790899518487096,
"loss": 0.0269,
"step": 1027
},
{
"epoch": 0.21538945052642605,
"grad_norm": 0.20888212323188782,
"learning_rate": 0.00019790449724508148,
"loss": 0.0226,
"step": 1028
},
{
"epoch": 0.2155989733382222,
"grad_norm": 0.1774069368839264,
"learning_rate": 0.00019789999452397483,
"loss": 0.0029,
"step": 1029
},
{
"epoch": 0.21580849615001832,
"grad_norm": 0.17364779114723206,
"learning_rate": 0.00019789548702177088,
"loss": 0.0065,
"step": 1030
},
{
"epoch": 0.21601801896181447,
"grad_norm": 0.0968272015452385,
"learning_rate": 0.00019789097473868976,
"loss": 0.0008,
"step": 1031
},
{
"epoch": 0.21622754177361062,
"grad_norm": 0.2362106889486313,
"learning_rate": 0.00019788645767495185,
"loss": 0.0329,
"step": 1032
},
{
"epoch": 0.21643706458540674,
"grad_norm": 0.11848565191030502,
"learning_rate": 0.0001978819358307777,
"loss": 0.0126,
"step": 1033
},
{
"epoch": 0.21664658739720288,
"grad_norm": 0.24092267453670502,
"learning_rate": 0.00019787740920638823,
"loss": 0.0094,
"step": 1034
},
{
"epoch": 0.216856110208999,
"grad_norm": 0.030452342703938484,
"learning_rate": 0.00019787287780200443,
"loss": 0.0086,
"step": 1035
},
{
"epoch": 0.21706563302079515,
"grad_norm": 0.21786944568157196,
"learning_rate": 0.00019786834161784764,
"loss": 0.0353,
"step": 1036
},
{
"epoch": 0.21727515583259127,
"grad_norm": 0.14589764177799225,
"learning_rate": 0.0001978638006541394,
"loss": 0.0027,
"step": 1037
},
{
"epoch": 0.21748467864438742,
"grad_norm": 0.22719940543174744,
"learning_rate": 0.0001978592549111014,
"loss": 0.0278,
"step": 1038
},
{
"epoch": 0.21769420145618354,
"grad_norm": 0.2133379429578781,
"learning_rate": 0.00019785470438895569,
"loss": 0.0163,
"step": 1039
},
{
"epoch": 0.2179037242679797,
"grad_norm": 0.025479169562458992,
"learning_rate": 0.00019785014908792453,
"loss": 0.0002,
"step": 1040
},
{
"epoch": 0.2181132470797758,
"grad_norm": 0.2815213203430176,
"learning_rate": 0.00019784558900823033,
"loss": 0.0176,
"step": 1041
},
{
"epoch": 0.21832276989157196,
"grad_norm": 0.018222063779830933,
"learning_rate": 0.00019784102415009583,
"loss": 0.0003,
"step": 1042
},
{
"epoch": 0.21853229270336808,
"grad_norm": 0.12288132309913635,
"learning_rate": 0.00019783645451374392,
"loss": 0.0143,
"step": 1043
},
{
"epoch": 0.21874181551516422,
"grad_norm": 0.2956022620201111,
"learning_rate": 0.0001978318800993978,
"loss": 0.0416,
"step": 1044
},
{
"epoch": 0.21895133832696034,
"grad_norm": 0.024944905191659927,
"learning_rate": 0.0001978273009072809,
"loss": 0.0003,
"step": 1045
},
{
"epoch": 0.2191608611387565,
"grad_norm": 0.17367038130760193,
"learning_rate": 0.00019782271693761678,
"loss": 0.0176,
"step": 1046
},
{
"epoch": 0.2193703839505526,
"grad_norm": 0.05597973242402077,
"learning_rate": 0.00019781812819062933,
"loss": 0.0083,
"step": 1047
},
{
"epoch": 0.21957990676234876,
"grad_norm": 0.12071727216243744,
"learning_rate": 0.00019781353466654267,
"loss": 0.0023,
"step": 1048
},
{
"epoch": 0.21978942957414488,
"grad_norm": 0.08094103634357452,
"learning_rate": 0.00019780893636558107,
"loss": 0.001,
"step": 1049
},
{
"epoch": 0.21999895238594103,
"grad_norm": 0.2887085974216461,
"learning_rate": 0.00019780433328796918,
"loss": 0.0249,
"step": 1050
},
{
"epoch": 0.22020847519773715,
"grad_norm": 0.10958587378263474,
"learning_rate": 0.0001977997254339318,
"loss": 0.0075,
"step": 1051
},
{
"epoch": 0.2204179980095333,
"grad_norm": 0.027714794501662254,
"learning_rate": 0.00019779511280369386,
"loss": 0.0003,
"step": 1052
},
{
"epoch": 0.22062752082132941,
"grad_norm": 0.11167047172784805,
"learning_rate": 0.0001977904953974807,
"loss": 0.009,
"step": 1053
},
{
"epoch": 0.22083704363312556,
"grad_norm": 0.0651421770453453,
"learning_rate": 0.00019778587321551782,
"loss": 0.0009,
"step": 1054
},
{
"epoch": 0.22104656644492168,
"grad_norm": 0.021489938721060753,
"learning_rate": 0.00019778124625803093,
"loss": 0.0002,
"step": 1055
},
{
"epoch": 0.22125608925671783,
"grad_norm": 0.22498519718647003,
"learning_rate": 0.000197776614525246,
"loss": 0.023,
"step": 1056
},
{
"epoch": 0.22146561206851395,
"grad_norm": 0.1699984073638916,
"learning_rate": 0.00019777197801738923,
"loss": 0.0209,
"step": 1057
},
{
"epoch": 0.2216751348803101,
"grad_norm": 0.09820838272571564,
"learning_rate": 0.00019776733673468707,
"loss": 0.0044,
"step": 1058
},
{
"epoch": 0.22188465769210622,
"grad_norm": 0.01764521189033985,
"learning_rate": 0.00019776269067736614,
"loss": 0.0003,
"step": 1059
},
{
"epoch": 0.22209418050390237,
"grad_norm": 0.1787995547056198,
"learning_rate": 0.00019775803984565336,
"loss": 0.0134,
"step": 1060
},
{
"epoch": 0.2223037033156985,
"grad_norm": 0.10098960250616074,
"learning_rate": 0.00019775338423977585,
"loss": 0.0006,
"step": 1061
},
{
"epoch": 0.22251322612749463,
"grad_norm": 0.40593600273132324,
"learning_rate": 0.00019774872385996102,
"loss": 0.0599,
"step": 1062
},
{
"epoch": 0.22272274893929075,
"grad_norm": 0.27995312213897705,
"learning_rate": 0.0001977440587064364,
"loss": 0.0427,
"step": 1063
},
{
"epoch": 0.2229322717510869,
"grad_norm": 0.317348450422287,
"learning_rate": 0.00019773938877942987,
"loss": 0.0325,
"step": 1064
},
{
"epoch": 0.22314179456288302,
"grad_norm": 0.3249301016330719,
"learning_rate": 0.0001977347140791695,
"loss": 0.0173,
"step": 1065
},
{
"epoch": 0.22335131737467917,
"grad_norm": 0.16467350721359253,
"learning_rate": 0.0001977300346058835,
"loss": 0.0184,
"step": 1066
},
{
"epoch": 0.2235608401864753,
"grad_norm": 0.16043822467327118,
"learning_rate": 0.00019772535035980052,
"loss": 0.0012,
"step": 1067
},
{
"epoch": 0.22377036299827144,
"grad_norm": 0.16195471584796906,
"learning_rate": 0.0001977206613411492,
"loss": 0.0099,
"step": 1068
},
{
"epoch": 0.22397988581006756,
"grad_norm": 0.01258161198347807,
"learning_rate": 0.00019771596755015867,
"loss": 0.0002,
"step": 1069
},
{
"epoch": 0.2241894086218637,
"grad_norm": 0.0009052444947883487,
"learning_rate": 0.00019771126898705804,
"loss": 0.0,
"step": 1070
},
{
"epoch": 0.22439893143365983,
"grad_norm": 0.09450815618038177,
"learning_rate": 0.00019770656565207685,
"loss": 0.0018,
"step": 1071
},
{
"epoch": 0.22460845424545597,
"grad_norm": 0.2263779193162918,
"learning_rate": 0.00019770185754544475,
"loss": 0.0388,
"step": 1072
},
{
"epoch": 0.22481797705725212,
"grad_norm": 0.0903795063495636,
"learning_rate": 0.0001976971446673917,
"loss": 0.0123,
"step": 1073
},
{
"epoch": 0.22502749986904824,
"grad_norm": 0.002591166878119111,
"learning_rate": 0.0001976924270181478,
"loss": 0.0001,
"step": 1074
},
{
"epoch": 0.2252370226808444,
"grad_norm": 0.17566534876823425,
"learning_rate": 0.00019768770459794351,
"loss": 0.0189,
"step": 1075
},
{
"epoch": 0.2254465454926405,
"grad_norm": 0.1893072873353958,
"learning_rate": 0.00019768297740700944,
"loss": 0.002,
"step": 1076
},
{
"epoch": 0.22565606830443666,
"grad_norm": 0.40008923411369324,
"learning_rate": 0.00019767824544557642,
"loss": 0.0479,
"step": 1077
},
{
"epoch": 0.22586559111623278,
"grad_norm": 0.0161714069545269,
"learning_rate": 0.00019767350871387557,
"loss": 0.0002,
"step": 1078
},
{
"epoch": 0.22607511392802893,
"grad_norm": 0.2905687391757965,
"learning_rate": 0.00019766876721213824,
"loss": 0.0301,
"step": 1079
},
{
"epoch": 0.22628463673982505,
"grad_norm": 0.1825425624847412,
"learning_rate": 0.0001976640209405959,
"loss": 0.0258,
"step": 1080
},
{
"epoch": 0.2264941595516212,
"grad_norm": 0.2943969964981079,
"learning_rate": 0.00019765926989948044,
"loss": 0.0253,
"step": 1081
},
{
"epoch": 0.2267036823634173,
"grad_norm": 0.03229305148124695,
"learning_rate": 0.00019765451408902383,
"loss": 0.0004,
"step": 1082
},
{
"epoch": 0.22691320517521346,
"grad_norm": 0.07214413583278656,
"learning_rate": 0.00019764975350945838,
"loss": 0.0026,
"step": 1083
},
{
"epoch": 0.22712272798700958,
"grad_norm": 0.18299317359924316,
"learning_rate": 0.00019764498816101652,
"loss": 0.0153,
"step": 1084
},
{
"epoch": 0.22733225079880573,
"grad_norm": 0.28171518445014954,
"learning_rate": 0.00019764021804393099,
"loss": 0.0329,
"step": 1085
},
{
"epoch": 0.22754177361060185,
"grad_norm": 0.18641382455825806,
"learning_rate": 0.00019763544315843474,
"loss": 0.0147,
"step": 1086
},
{
"epoch": 0.227751296422398,
"grad_norm": 0.16016316413879395,
"learning_rate": 0.00019763066350476097,
"loss": 0.004,
"step": 1087
},
{
"epoch": 0.22796081923419412,
"grad_norm": 0.15008705854415894,
"learning_rate": 0.0001976258790831431,
"loss": 0.0235,
"step": 1088
},
{
"epoch": 0.22817034204599027,
"grad_norm": 0.24476243555545807,
"learning_rate": 0.00019762108989381481,
"loss": 0.0299,
"step": 1089
},
{
"epoch": 0.22837986485778639,
"grad_norm": 0.027294384315609932,
"learning_rate": 0.00019761629593701,
"loss": 0.0003,
"step": 1090
},
{
"epoch": 0.22858938766958253,
"grad_norm": 0.20721612870693207,
"learning_rate": 0.00019761149721296269,
"loss": 0.0046,
"step": 1091
},
{
"epoch": 0.22879891048137865,
"grad_norm": 0.34984588623046875,
"learning_rate": 0.0001976066937219073,
"loss": 0.0551,
"step": 1092
},
{
"epoch": 0.2290084332931748,
"grad_norm": 0.08879242092370987,
"learning_rate": 0.00019760188546407844,
"loss": 0.0005,
"step": 1093
},
{
"epoch": 0.22921795610497092,
"grad_norm": 0.08779024332761765,
"learning_rate": 0.00019759707243971087,
"loss": 0.0009,
"step": 1094
},
{
"epoch": 0.22942747891676707,
"grad_norm": 0.22871573269367218,
"learning_rate": 0.00019759225464903972,
"loss": 0.0155,
"step": 1095
},
{
"epoch": 0.2296370017285632,
"grad_norm": 0.005597609095275402,
"learning_rate": 0.00019758743209230022,
"loss": 0.0001,
"step": 1096
},
{
"epoch": 0.22984652454035934,
"grad_norm": 0.180862158536911,
"learning_rate": 0.0001975826047697279,
"loss": 0.0159,
"step": 1097
},
{
"epoch": 0.23005604735215546,
"grad_norm": 0.011714598163962364,
"learning_rate": 0.00019757777268155846,
"loss": 0.0001,
"step": 1098
},
{
"epoch": 0.2302655701639516,
"grad_norm": 0.16202227771282196,
"learning_rate": 0.00019757293582802794,
"loss": 0.0225,
"step": 1099
},
{
"epoch": 0.23047509297574773,
"grad_norm": 0.030815673992037773,
"learning_rate": 0.00019756809420937255,
"loss": 0.0002,
"step": 1100
},
{
"epoch": 0.23068461578754387,
"grad_norm": 0.1840132772922516,
"learning_rate": 0.00019756324782582873,
"loss": 0.0099,
"step": 1101
},
{
"epoch": 0.23089413859934,
"grad_norm": 0.16079594194889069,
"learning_rate": 0.00019755839667763317,
"loss": 0.006,
"step": 1102
},
{
"epoch": 0.23110366141113614,
"grad_norm": 0.30695217847824097,
"learning_rate": 0.00019755354076502276,
"loss": 0.0175,
"step": 1103
},
{
"epoch": 0.23131318422293226,
"grad_norm": 0.011515875346958637,
"learning_rate": 0.00019754868008823467,
"loss": 0.0001,
"step": 1104
},
{
"epoch": 0.2315227070347284,
"grad_norm": 0.14559753239154816,
"learning_rate": 0.00019754381464750628,
"loss": 0.0078,
"step": 1105
},
{
"epoch": 0.23173222984652453,
"grad_norm": 0.007846977561712265,
"learning_rate": 0.00019753894444307516,
"loss": 0.0001,
"step": 1106
},
{
"epoch": 0.23194175265832068,
"grad_norm": 0.03170756250619888,
"learning_rate": 0.00019753406947517924,
"loss": 0.0003,
"step": 1107
},
{
"epoch": 0.2321512754701168,
"grad_norm": 0.0651014968752861,
"learning_rate": 0.00019752918974405648,
"loss": 0.0031,
"step": 1108
},
{
"epoch": 0.23236079828191294,
"grad_norm": 0.26344966888427734,
"learning_rate": 0.00019752430524994528,
"loss": 0.0232,
"step": 1109
},
{
"epoch": 0.23257032109370906,
"grad_norm": 0.2215491235256195,
"learning_rate": 0.0001975194159930841,
"loss": 0.002,
"step": 1110
},
{
"epoch": 0.2327798439055052,
"grad_norm": 0.00715980026870966,
"learning_rate": 0.00019751452197371182,
"loss": 0.0001,
"step": 1111
},
{
"epoch": 0.23298936671730133,
"grad_norm": 0.01364982221275568,
"learning_rate": 0.0001975096231920674,
"loss": 0.0002,
"step": 1112
},
{
"epoch": 0.23319888952909748,
"grad_norm": 0.17877884209156036,
"learning_rate": 0.00019750471964839005,
"loss": 0.0027,
"step": 1113
},
{
"epoch": 0.23340841234089363,
"grad_norm": 0.31934672594070435,
"learning_rate": 0.0001974998113429193,
"loss": 0.0443,
"step": 1114
},
{
"epoch": 0.23361793515268975,
"grad_norm": 0.41927701234817505,
"learning_rate": 0.00019749489827589477,
"loss": 0.0651,
"step": 1115
},
{
"epoch": 0.2338274579644859,
"grad_norm": 0.004164255689829588,
"learning_rate": 0.00019748998044755647,
"loss": 0.0001,
"step": 1116
},
{
"epoch": 0.23403698077628202,
"grad_norm": 0.003149515949189663,
"learning_rate": 0.00019748505785814456,
"loss": 0.0001,
"step": 1117
},
{
"epoch": 0.23424650358807816,
"grad_norm": 0.22820650041103363,
"learning_rate": 0.0001974801305078994,
"loss": 0.0291,
"step": 1118
},
{
"epoch": 0.23445602639987428,
"grad_norm": 0.06346362829208374,
"learning_rate": 0.00019747519839706165,
"loss": 0.0006,
"step": 1119
},
{
"epoch": 0.23466554921167043,
"grad_norm": 0.17930179834365845,
"learning_rate": 0.0001974702615258722,
"loss": 0.0085,
"step": 1120
},
{
"epoch": 0.23487507202346655,
"grad_norm": 0.03995645418763161,
"learning_rate": 0.00019746531989457214,
"loss": 0.0004,
"step": 1121
},
{
"epoch": 0.2350845948352627,
"grad_norm": 0.16451582312583923,
"learning_rate": 0.00019746037350340276,
"loss": 0.0111,
"step": 1122
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.1479364037513733,
"learning_rate": 0.0001974554223526057,
"loss": 0.0243,
"step": 1123
},
{
"epoch": 0.23550364045885497,
"grad_norm": 0.04975096508860588,
"learning_rate": 0.00019745046644242267,
"loss": 0.0005,
"step": 1124
},
{
"epoch": 0.2357131632706511,
"grad_norm": 0.10589946806430817,
"learning_rate": 0.00019744550577309575,
"loss": 0.0018,
"step": 1125
},
{
"epoch": 0.23592268608244724,
"grad_norm": 0.07141809910535812,
"learning_rate": 0.00019744054034486722,
"loss": 0.0009,
"step": 1126
},
{
"epoch": 0.23613220889424336,
"grad_norm": 0.11263751983642578,
"learning_rate": 0.00019743557015797954,
"loss": 0.0129,
"step": 1127
},
{
"epoch": 0.2363417317060395,
"grad_norm": 0.16582761704921722,
"learning_rate": 0.0001974305952126754,
"loss": 0.013,
"step": 1128
},
{
"epoch": 0.23655125451783562,
"grad_norm": 0.19919754564762115,
"learning_rate": 0.0001974256155091979,
"loss": 0.0019,
"step": 1129
},
{
"epoch": 0.23676077732963177,
"grad_norm": 0.052698615938425064,
"learning_rate": 0.00019742063104779008,
"loss": 0.0023,
"step": 1130
},
{
"epoch": 0.2369703001414279,
"grad_norm": 0.14383988082408905,
"learning_rate": 0.00019741564182869544,
"loss": 0.0095,
"step": 1131
},
{
"epoch": 0.23717982295322404,
"grad_norm": 0.34627342224121094,
"learning_rate": 0.00019741064785215758,
"loss": 0.0533,
"step": 1132
},
{
"epoch": 0.23738934576502016,
"grad_norm": 0.3608137369155884,
"learning_rate": 0.00019740564911842044,
"loss": 0.0742,
"step": 1133
},
{
"epoch": 0.2375988685768163,
"grad_norm": 0.3834199905395508,
"learning_rate": 0.00019740064562772815,
"loss": 0.0685,
"step": 1134
},
{
"epoch": 0.23780839138861243,
"grad_norm": 0.19751276075839996,
"learning_rate": 0.00019739563738032504,
"loss": 0.0296,
"step": 1135
},
{
"epoch": 0.23801791420040858,
"grad_norm": 0.2406461536884308,
"learning_rate": 0.00019739062437645568,
"loss": 0.032,
"step": 1136
},
{
"epoch": 0.2382274370122047,
"grad_norm": 0.036680351942777634,
"learning_rate": 0.00019738560661636492,
"loss": 0.0006,
"step": 1137
},
{
"epoch": 0.23843695982400084,
"grad_norm": 0.12333112210035324,
"learning_rate": 0.00019738058410029776,
"loss": 0.027,
"step": 1138
},
{
"epoch": 0.23864648263579696,
"grad_norm": 0.12504327297210693,
"learning_rate": 0.00019737555682849956,
"loss": 0.0077,
"step": 1139
},
{
"epoch": 0.2388560054475931,
"grad_norm": 0.26424646377563477,
"learning_rate": 0.00019737052480121577,
"loss": 0.0243,
"step": 1140
},
{
"epoch": 0.23906552825938923,
"grad_norm": 0.05994394049048424,
"learning_rate": 0.00019736548801869215,
"loss": 0.0008,
"step": 1141
},
{
"epoch": 0.23927505107118538,
"grad_norm": 0.15409626066684723,
"learning_rate": 0.0001973604464811747,
"loss": 0.0101,
"step": 1142
},
{
"epoch": 0.2394845738829815,
"grad_norm": 0.011877711862325668,
"learning_rate": 0.0001973554001889096,
"loss": 0.0002,
"step": 1143
},
{
"epoch": 0.23969409669477765,
"grad_norm": 0.14174003899097443,
"learning_rate": 0.0001973503491421433,
"loss": 0.0019,
"step": 1144
},
{
"epoch": 0.23990361950657377,
"grad_norm": 0.13701941072940826,
"learning_rate": 0.00019734529334112253,
"loss": 0.014,
"step": 1145
},
{
"epoch": 0.24011314231836992,
"grad_norm": 0.19319787621498108,
"learning_rate": 0.00019734023278609412,
"loss": 0.0064,
"step": 1146
},
{
"epoch": 0.24032266513016604,
"grad_norm": 0.09388478845357895,
"learning_rate": 0.00019733516747730528,
"loss": 0.0058,
"step": 1147
},
{
"epoch": 0.24053218794196218,
"grad_norm": 0.00462870579212904,
"learning_rate": 0.0001973300974150033,
"loss": 0.0001,
"step": 1148
},
{
"epoch": 0.2407417107537583,
"grad_norm": 0.21105362474918365,
"learning_rate": 0.00019732502259943587,
"loss": 0.0435,
"step": 1149
},
{
"epoch": 0.24095123356555445,
"grad_norm": 0.21553511917591095,
"learning_rate": 0.00019731994303085075,
"loss": 0.0065,
"step": 1150
},
{
"epoch": 0.24116075637735057,
"grad_norm": 0.1303602159023285,
"learning_rate": 0.00019731485870949606,
"loss": 0.0086,
"step": 1151
},
{
"epoch": 0.24137027918914672,
"grad_norm": 0.009067162871360779,
"learning_rate": 0.00019730976963562008,
"loss": 0.0001,
"step": 1152
},
{
"epoch": 0.24157980200094284,
"grad_norm": 0.30989691615104675,
"learning_rate": 0.00019730467580947138,
"loss": 0.0328,
"step": 1153
},
{
"epoch": 0.241789324812739,
"grad_norm": 0.05775981396436691,
"learning_rate": 0.00019729957723129866,
"loss": 0.0005,
"step": 1154
},
{
"epoch": 0.24199884762453513,
"grad_norm": 0.14925909042358398,
"learning_rate": 0.00019729447390135098,
"loss": 0.0147,
"step": 1155
},
{
"epoch": 0.24220837043633126,
"grad_norm": 0.1444745808839798,
"learning_rate": 0.00019728936581987752,
"loss": 0.0148,
"step": 1156
},
{
"epoch": 0.2424178932481274,
"grad_norm": 0.004791347309947014,
"learning_rate": 0.00019728425298712776,
"loss": 0.0001,
"step": 1157
},
{
"epoch": 0.24262741605992352,
"grad_norm": 0.12054464221000671,
"learning_rate": 0.00019727913540335142,
"loss": 0.0099,
"step": 1158
},
{
"epoch": 0.24283693887171967,
"grad_norm": 0.02685241401195526,
"learning_rate": 0.00019727401306879837,
"loss": 0.0002,
"step": 1159
},
{
"epoch": 0.2430464616835158,
"grad_norm": 0.1325118988752365,
"learning_rate": 0.00019726888598371883,
"loss": 0.0129,
"step": 1160
},
{
"epoch": 0.24325598449531194,
"grad_norm": 0.010920906439423561,
"learning_rate": 0.0001972637541483631,
"loss": 0.0001,
"step": 1161
},
{
"epoch": 0.24346550730710806,
"grad_norm": 0.1611892580986023,
"learning_rate": 0.00019725861756298191,
"loss": 0.0025,
"step": 1162
},
{
"epoch": 0.2436750301189042,
"grad_norm": 0.0020086602307856083,
"learning_rate": 0.00019725347622782604,
"loss": 0.0,
"step": 1163
},
{
"epoch": 0.24388455293070033,
"grad_norm": 0.20563958585262299,
"learning_rate": 0.00019724833014314662,
"loss": 0.009,
"step": 1164
},
{
"epoch": 0.24409407574249647,
"grad_norm": 0.21522867679595947,
"learning_rate": 0.0001972431793091949,
"loss": 0.0177,
"step": 1165
},
{
"epoch": 0.2443035985542926,
"grad_norm": 0.25818154215812683,
"learning_rate": 0.00019723802372622248,
"loss": 0.0223,
"step": 1166
},
{
"epoch": 0.24451312136608874,
"grad_norm": 0.2856491208076477,
"learning_rate": 0.00019723286339448114,
"loss": 0.0353,
"step": 1167
},
{
"epoch": 0.24472264417788486,
"grad_norm": 0.12360674142837524,
"learning_rate": 0.0001972276983142229,
"loss": 0.0035,
"step": 1168
},
{
"epoch": 0.244932166989681,
"grad_norm": 0.07090656459331512,
"learning_rate": 0.0001972225284857,
"loss": 0.0007,
"step": 1169
},
{
"epoch": 0.24514168980147713,
"grad_norm": 0.23631127178668976,
"learning_rate": 0.00019721735390916485,
"loss": 0.0039,
"step": 1170
},
{
"epoch": 0.24535121261327328,
"grad_norm": 0.005582223180681467,
"learning_rate": 0.00019721217458487024,
"loss": 0.0001,
"step": 1171
},
{
"epoch": 0.2455607354250694,
"grad_norm": 0.11760203540325165,
"learning_rate": 0.0001972069905130691,
"loss": 0.0187,
"step": 1172
},
{
"epoch": 0.24577025823686555,
"grad_norm": 0.0623784065246582,
"learning_rate": 0.00019720180169401457,
"loss": 0.0026,
"step": 1173
},
{
"epoch": 0.24597978104866167,
"grad_norm": 0.06554267555475235,
"learning_rate": 0.00019719660812796006,
"loss": 0.0008,
"step": 1174
},
{
"epoch": 0.24618930386045781,
"grad_norm": 0.15499131381511688,
"learning_rate": 0.00019719140981515923,
"loss": 0.0155,
"step": 1175
},
{
"epoch": 0.24639882667225393,
"grad_norm": 0.36670204997062683,
"learning_rate": 0.00019718620675586592,
"loss": 0.0826,
"step": 1176
},
{
"epoch": 0.24660834948405008,
"grad_norm": 0.09148335456848145,
"learning_rate": 0.00019718099895033427,
"loss": 0.0008,
"step": 1177
},
{
"epoch": 0.2468178722958462,
"grad_norm": 0.0192439965903759,
"learning_rate": 0.00019717578639881858,
"loss": 0.0003,
"step": 1178
},
{
"epoch": 0.24702739510764235,
"grad_norm": 0.17157042026519775,
"learning_rate": 0.0001971705691015734,
"loss": 0.0329,
"step": 1179
},
{
"epoch": 0.24723691791943847,
"grad_norm": 0.26223111152648926,
"learning_rate": 0.00019716534705885354,
"loss": 0.0499,
"step": 1180
},
{
"epoch": 0.24744644073123462,
"grad_norm": 0.07771110534667969,
"learning_rate": 0.00019716012027091404,
"loss": 0.001,
"step": 1181
},
{
"epoch": 0.24765596354303074,
"grad_norm": 0.16626375913619995,
"learning_rate": 0.00019715488873801014,
"loss": 0.0301,
"step": 1182
},
{
"epoch": 0.24786548635482689,
"grad_norm": 0.3291771113872528,
"learning_rate": 0.00019714965246039735,
"loss": 0.0383,
"step": 1183
},
{
"epoch": 0.248075009166623,
"grad_norm": 0.1012611836194992,
"learning_rate": 0.00019714441143833138,
"loss": 0.003,
"step": 1184
},
{
"epoch": 0.24828453197841915,
"grad_norm": 0.03600292280316353,
"learning_rate": 0.00019713916567206817,
"loss": 0.0003,
"step": 1185
},
{
"epoch": 0.24849405479021527,
"grad_norm": 0.14185172319412231,
"learning_rate": 0.00019713391516186393,
"loss": 0.0117,
"step": 1186
},
{
"epoch": 0.24870357760201142,
"grad_norm": 0.17416037619113922,
"learning_rate": 0.00019712865990797505,
"loss": 0.028,
"step": 1187
},
{
"epoch": 0.24891310041380754,
"grad_norm": 0.2895718216896057,
"learning_rate": 0.00019712339991065817,
"loss": 0.0307,
"step": 1188
},
{
"epoch": 0.2491226232256037,
"grad_norm": 0.09385382384061813,
"learning_rate": 0.00019711813517017026,
"loss": 0.0074,
"step": 1189
},
{
"epoch": 0.2493321460373998,
"grad_norm": 0.1465684324502945,
"learning_rate": 0.00019711286568676833,
"loss": 0.0107,
"step": 1190
},
{
"epoch": 0.24954166884919596,
"grad_norm": 0.019048716872930527,
"learning_rate": 0.00019710759146070972,
"loss": 0.0005,
"step": 1191
},
{
"epoch": 0.24975119166099208,
"grad_norm": 0.10628386586904526,
"learning_rate": 0.00019710231249225207,
"loss": 0.0018,
"step": 1192
},
{
"epoch": 0.24996071447278823,
"grad_norm": 0.0259134192019701,
"learning_rate": 0.0001970970287816532,
"loss": 0.0095,
"step": 1193
},
{
"epoch": 0.2501702372845844,
"grad_norm": 0.07972554117441177,
"learning_rate": 0.00019709174032917103,
"loss": 0.0011,
"step": 1194
},
{
"epoch": 0.25037976009638047,
"grad_norm": 0.006588938180357218,
"learning_rate": 0.00019708644713506396,
"loss": 0.0001,
"step": 1195
},
{
"epoch": 0.2505892829081766,
"grad_norm": 0.2775128483772278,
"learning_rate": 0.00019708114919959045,
"loss": 0.0128,
"step": 1196
},
{
"epoch": 0.25079880571997276,
"grad_norm": 0.004361490719020367,
"learning_rate": 0.0001970758465230092,
"loss": 0.0001,
"step": 1197
},
{
"epoch": 0.2510083285317689,
"grad_norm": 0.11160224676132202,
"learning_rate": 0.00019707053910557917,
"loss": 0.0017,
"step": 1198
},
{
"epoch": 0.25121785134356506,
"grad_norm": 0.013135232962667942,
"learning_rate": 0.0001970652269475596,
"loss": 0.0001,
"step": 1199
},
{
"epoch": 0.25142737415536115,
"grad_norm": 0.2182655930519104,
"learning_rate": 0.00019705991004920993,
"loss": 0.0159,
"step": 1200
},
{
"epoch": 0.25142737415536115,
"eval_loss": 0.01089566946029663,
"eval_runtime": 89.1143,
"eval_samples_per_second": 8.618,
"eval_steps_per_second": 4.309,
"step": 1200
},
{
"epoch": 0.2516368969671573,
"grad_norm": 0.24273119866847992,
"learning_rate": 0.00019705458841078974,
"loss": 0.0178,
"step": 1201
},
{
"epoch": 0.25184641977895345,
"grad_norm": 0.036558933556079865,
"learning_rate": 0.00019704926203255898,
"loss": 0.0029,
"step": 1202
},
{
"epoch": 0.2520559425907496,
"grad_norm": 0.005220042075961828,
"learning_rate": 0.0001970439309147778,
"loss": 0.0001,
"step": 1203
},
{
"epoch": 0.2522654654025457,
"grad_norm": 0.1721489429473877,
"learning_rate": 0.00019703859505770644,
"loss": 0.0113,
"step": 1204
},
{
"epoch": 0.25247498821434183,
"grad_norm": 0.18657046556472778,
"learning_rate": 0.00019703325446160564,
"loss": 0.0133,
"step": 1205
},
{
"epoch": 0.252684511026138,
"grad_norm": 0.15162359178066254,
"learning_rate": 0.0001970279091267361,
"loss": 0.0045,
"step": 1206
},
{
"epoch": 0.25289403383793413,
"grad_norm": 0.16691045463085175,
"learning_rate": 0.0001970225590533589,
"loss": 0.0052,
"step": 1207
},
{
"epoch": 0.2531035566497302,
"grad_norm": 0.17631007730960846,
"learning_rate": 0.0001970172042417353,
"loss": 0.0079,
"step": 1208
},
{
"epoch": 0.25331307946152637,
"grad_norm": 0.23643368482589722,
"learning_rate": 0.00019701184469212685,
"loss": 0.0239,
"step": 1209
},
{
"epoch": 0.2535226022733225,
"grad_norm": 0.18658863008022308,
"learning_rate": 0.00019700648040479527,
"loss": 0.0333,
"step": 1210
},
{
"epoch": 0.25373212508511866,
"grad_norm": 0.17524613440036774,
"learning_rate": 0.00019700111138000256,
"loss": 0.0039,
"step": 1211
},
{
"epoch": 0.25394164789691476,
"grad_norm": 0.0618131160736084,
"learning_rate": 0.00019699573761801085,
"loss": 0.009,
"step": 1212
},
{
"epoch": 0.2541511707087109,
"grad_norm": 0.3288923501968384,
"learning_rate": 0.00019699035911908267,
"loss": 0.0353,
"step": 1213
},
{
"epoch": 0.25436069352050705,
"grad_norm": 0.1573517769575119,
"learning_rate": 0.00019698497588348067,
"loss": 0.013,
"step": 1214
},
{
"epoch": 0.2545702163323032,
"grad_norm": 0.1708369106054306,
"learning_rate": 0.00019697958791146768,
"loss": 0.0132,
"step": 1215
},
{
"epoch": 0.2547797391440993,
"grad_norm": 0.25349169969558716,
"learning_rate": 0.00019697419520330688,
"loss": 0.0068,
"step": 1216
},
{
"epoch": 0.25498926195589544,
"grad_norm": 0.12924553453922272,
"learning_rate": 0.00019696879775926162,
"loss": 0.0084,
"step": 1217
},
{
"epoch": 0.2551987847676916,
"grad_norm": 0.12420932948589325,
"learning_rate": 0.00019696339557959548,
"loss": 0.001,
"step": 1218
},
{
"epoch": 0.25540830757948774,
"grad_norm": 0.05618319287896156,
"learning_rate": 0.00019695798866457235,
"loss": 0.0004,
"step": 1219
},
{
"epoch": 0.25561783039128383,
"grad_norm": 0.06126785650849342,
"learning_rate": 0.0001969525770144562,
"loss": 0.001,
"step": 1220
},
{
"epoch": 0.25582735320308,
"grad_norm": 0.10019136220216751,
"learning_rate": 0.00019694716062951137,
"loss": 0.0021,
"step": 1221
},
{
"epoch": 0.2560368760148761,
"grad_norm": 0.021487459540367126,
"learning_rate": 0.00019694173951000234,
"loss": 0.0002,
"step": 1222
},
{
"epoch": 0.2562463988266723,
"grad_norm": 0.13679708540439606,
"learning_rate": 0.00019693631365619389,
"loss": 0.0114,
"step": 1223
},
{
"epoch": 0.25645592163846836,
"grad_norm": 0.001410324708558619,
"learning_rate": 0.000196930883068351,
"loss": 0.0,
"step": 1224
},
{
"epoch": 0.2566654444502645,
"grad_norm": 0.0035328457597643137,
"learning_rate": 0.00019692544774673886,
"loss": 0.0001,
"step": 1225
},
{
"epoch": 0.25687496726206066,
"grad_norm": 0.3505251705646515,
"learning_rate": 0.00019692000769162292,
"loss": 0.0147,
"step": 1226
},
{
"epoch": 0.2570844900738568,
"grad_norm": 0.14239129424095154,
"learning_rate": 0.00019691456290326883,
"loss": 0.0117,
"step": 1227
},
{
"epoch": 0.2572940128856529,
"grad_norm": 0.30475863814353943,
"learning_rate": 0.00019690911338194256,
"loss": 0.0572,
"step": 1228
},
{
"epoch": 0.25750353569744905,
"grad_norm": 0.1941041797399521,
"learning_rate": 0.0001969036591279102,
"loss": 0.0134,
"step": 1229
},
{
"epoch": 0.2577130585092452,
"grad_norm": 0.019271748140454292,
"learning_rate": 0.00019689820014143813,
"loss": 0.0001,
"step": 1230
},
{
"epoch": 0.25792258132104134,
"grad_norm": 0.11813981086015701,
"learning_rate": 0.00019689273642279292,
"loss": 0.0023,
"step": 1231
},
{
"epoch": 0.25813210413283744,
"grad_norm": 0.057555362582206726,
"learning_rate": 0.00019688726797224143,
"loss": 0.0008,
"step": 1232
},
{
"epoch": 0.2583416269446336,
"grad_norm": 0.24133259057998657,
"learning_rate": 0.00019688179479005068,
"loss": 0.0049,
"step": 1233
},
{
"epoch": 0.25855114975642973,
"grad_norm": 0.24271616339683533,
"learning_rate": 0.00019687631687648804,
"loss": 0.0231,
"step": 1234
},
{
"epoch": 0.2587606725682259,
"grad_norm": 0.15305756032466888,
"learning_rate": 0.00019687083423182094,
"loss": 0.0102,
"step": 1235
},
{
"epoch": 0.258970195380022,
"grad_norm": 0.04598066210746765,
"learning_rate": 0.00019686534685631722,
"loss": 0.0007,
"step": 1236
},
{
"epoch": 0.2591797181918181,
"grad_norm": 0.015225352719426155,
"learning_rate": 0.0001968598547502448,
"loss": 0.0002,
"step": 1237
},
{
"epoch": 0.25938924100361427,
"grad_norm": 0.18596412241458893,
"learning_rate": 0.0001968543579138719,
"loss": 0.0282,
"step": 1238
},
{
"epoch": 0.2595987638154104,
"grad_norm": 0.004785776603966951,
"learning_rate": 0.000196848856347467,
"loss": 0.0001,
"step": 1239
},
{
"epoch": 0.25980828662720656,
"grad_norm": 0.1279231756925583,
"learning_rate": 0.00019684335005129878,
"loss": 0.0023,
"step": 1240
},
{
"epoch": 0.26001780943900266,
"grad_norm": 0.2739182114601135,
"learning_rate": 0.00019683783902563608,
"loss": 0.0029,
"step": 1241
},
{
"epoch": 0.2602273322507988,
"grad_norm": 0.15039706230163574,
"learning_rate": 0.00019683232327074812,
"loss": 0.0053,
"step": 1242
},
{
"epoch": 0.26043685506259495,
"grad_norm": 0.0031035866122692823,
"learning_rate": 0.00019682680278690424,
"loss": 0.0,
"step": 1243
},
{
"epoch": 0.2606463778743911,
"grad_norm": 0.11631850898265839,
"learning_rate": 0.00019682127757437404,
"loss": 0.0024,
"step": 1244
},
{
"epoch": 0.2608559006861872,
"grad_norm": 0.005499638617038727,
"learning_rate": 0.00019681574763342737,
"loss": 0.0001,
"step": 1245
},
{
"epoch": 0.26106542349798334,
"grad_norm": 0.24378150701522827,
"learning_rate": 0.00019681021296433425,
"loss": 0.0046,
"step": 1246
},
{
"epoch": 0.2612749463097795,
"grad_norm": 0.22416795790195465,
"learning_rate": 0.00019680467356736502,
"loss": 0.0242,
"step": 1247
},
{
"epoch": 0.26148446912157564,
"grad_norm": 0.31286659836769104,
"learning_rate": 0.00019679912944279014,
"loss": 0.0414,
"step": 1248
},
{
"epoch": 0.26169399193337173,
"grad_norm": 0.22582346200942993,
"learning_rate": 0.00019679358059088045,
"loss": 0.013,
"step": 1249
},
{
"epoch": 0.2619035147451679,
"grad_norm": 0.10074607282876968,
"learning_rate": 0.0001967880270119069,
"loss": 0.0037,
"step": 1250
},
{
"epoch": 0.262113037556964,
"grad_norm": 0.20810967683792114,
"learning_rate": 0.00019678246870614069,
"loss": 0.004,
"step": 1251
},
{
"epoch": 0.26232256036876017,
"grad_norm": 0.18476144969463348,
"learning_rate": 0.00019677690567385328,
"loss": 0.0212,
"step": 1252
},
{
"epoch": 0.26253208318055626,
"grad_norm": 0.3872232735157013,
"learning_rate": 0.00019677133791531638,
"loss": 0.067,
"step": 1253
},
{
"epoch": 0.2627416059923524,
"grad_norm": 0.28218770027160645,
"learning_rate": 0.00019676576543080186,
"loss": 0.0365,
"step": 1254
},
{
"epoch": 0.26295112880414856,
"grad_norm": 0.2109004259109497,
"learning_rate": 0.0001967601882205819,
"loss": 0.0353,
"step": 1255
},
{
"epoch": 0.2631606516159447,
"grad_norm": 0.11097536981105804,
"learning_rate": 0.00019675460628492878,
"loss": 0.0016,
"step": 1256
},
{
"epoch": 0.2633701744277408,
"grad_norm": 0.2835911810398102,
"learning_rate": 0.0001967490196241152,
"loss": 0.0228,
"step": 1257
},
{
"epoch": 0.26357969723953695,
"grad_norm": 0.03125814348459244,
"learning_rate": 0.00019674342823841398,
"loss": 0.0005,
"step": 1258
},
{
"epoch": 0.2637892200513331,
"grad_norm": 0.2683068811893463,
"learning_rate": 0.00019673783212809814,
"loss": 0.0115,
"step": 1259
},
{
"epoch": 0.26399874286312924,
"grad_norm": 0.06683268398046494,
"learning_rate": 0.00019673223129344102,
"loss": 0.0011,
"step": 1260
},
{
"epoch": 0.26420826567492534,
"grad_norm": 0.15417706966400146,
"learning_rate": 0.00019672662573471614,
"loss": 0.0055,
"step": 1261
},
{
"epoch": 0.2644177884867215,
"grad_norm": 0.22664068639278412,
"learning_rate": 0.0001967210154521972,
"loss": 0.0324,
"step": 1262
},
{
"epoch": 0.26462731129851763,
"grad_norm": 0.15802893042564392,
"learning_rate": 0.00019671540044615825,
"loss": 0.0013,
"step": 1263
},
{
"epoch": 0.2648368341103138,
"grad_norm": 0.20269808173179626,
"learning_rate": 0.0001967097807168735,
"loss": 0.0179,
"step": 1264
},
{
"epoch": 0.26504635692210987,
"grad_norm": 0.20458929240703583,
"learning_rate": 0.00019670415626461737,
"loss": 0.0248,
"step": 1265
},
{
"epoch": 0.265255879733906,
"grad_norm": 0.05246388539671898,
"learning_rate": 0.00019669852708966454,
"loss": 0.0006,
"step": 1266
},
{
"epoch": 0.26546540254570217,
"grad_norm": 0.02725144289433956,
"learning_rate": 0.00019669289319228995,
"loss": 0.0004,
"step": 1267
},
{
"epoch": 0.2656749253574983,
"grad_norm": 0.07908505201339722,
"learning_rate": 0.0001966872545727687,
"loss": 0.001,
"step": 1268
},
{
"epoch": 0.2658844481692944,
"grad_norm": 0.26351892948150635,
"learning_rate": 0.0001966816112313762,
"loss": 0.0317,
"step": 1269
},
{
"epoch": 0.26609397098109056,
"grad_norm": 0.09313903748989105,
"learning_rate": 0.00019667596316838803,
"loss": 0.0014,
"step": 1270
},
{
"epoch": 0.2663034937928867,
"grad_norm": 0.1861148625612259,
"learning_rate": 0.00019667031038408004,
"loss": 0.0376,
"step": 1271
},
{
"epoch": 0.26651301660468285,
"grad_norm": 0.12870298326015472,
"learning_rate": 0.00019666465287872825,
"loss": 0.0155,
"step": 1272
},
{
"epoch": 0.26672253941647894,
"grad_norm": 0.3237362205982208,
"learning_rate": 0.00019665899065260902,
"loss": 0.0644,
"step": 1273
},
{
"epoch": 0.2669320622282751,
"grad_norm": 0.19791316986083984,
"learning_rate": 0.0001966533237059988,
"loss": 0.0144,
"step": 1274
},
{
"epoch": 0.26714158504007124,
"grad_norm": 0.08102938532829285,
"learning_rate": 0.0001966476520391744,
"loss": 0.0004,
"step": 1275
},
{
"epoch": 0.2673511078518674,
"grad_norm": 0.2837512493133545,
"learning_rate": 0.00019664197565241279,
"loss": 0.0317,
"step": 1276
},
{
"epoch": 0.2675606306636635,
"grad_norm": 0.09621527045965195,
"learning_rate": 0.00019663629454599115,
"loss": 0.0019,
"step": 1277
},
{
"epoch": 0.2677701534754596,
"grad_norm": 0.2916226387023926,
"learning_rate": 0.00019663060872018696,
"loss": 0.0294,
"step": 1278
},
{
"epoch": 0.2679796762872558,
"grad_norm": 0.07113649696111679,
"learning_rate": 0.0001966249181752779,
"loss": 0.0049,
"step": 1279
},
{
"epoch": 0.2681891990990519,
"grad_norm": 0.027167638763785362,
"learning_rate": 0.00019661922291154188,
"loss": 0.0005,
"step": 1280
},
{
"epoch": 0.26839872191084807,
"grad_norm": 0.033556949347257614,
"learning_rate": 0.00019661352292925702,
"loss": 0.0002,
"step": 1281
},
{
"epoch": 0.26860824472264416,
"grad_norm": 0.15087144076824188,
"learning_rate": 0.00019660781822870165,
"loss": 0.0019,
"step": 1282
},
{
"epoch": 0.2688177675344403,
"grad_norm": 0.1311134546995163,
"learning_rate": 0.00019660210881015448,
"loss": 0.003,
"step": 1283
},
{
"epoch": 0.26902729034623646,
"grad_norm": 0.10054054856300354,
"learning_rate": 0.00019659639467389423,
"loss": 0.0084,
"step": 1284
},
{
"epoch": 0.2692368131580326,
"grad_norm": 0.11179205030202866,
"learning_rate": 0.0001965906758202,
"loss": 0.0014,
"step": 1285
},
{
"epoch": 0.2694463359698287,
"grad_norm": 0.11406797170639038,
"learning_rate": 0.00019658495224935107,
"loss": 0.0036,
"step": 1286
},
{
"epoch": 0.26965585878162485,
"grad_norm": 0.16709235310554504,
"learning_rate": 0.00019657922396162697,
"loss": 0.0183,
"step": 1287
},
{
"epoch": 0.269865381593421,
"grad_norm": 0.02597794309258461,
"learning_rate": 0.00019657349095730743,
"loss": 0.0002,
"step": 1288
},
{
"epoch": 0.27007490440521714,
"grad_norm": 0.10171519219875336,
"learning_rate": 0.00019656775323667247,
"loss": 0.0044,
"step": 1289
},
{
"epoch": 0.27028442721701323,
"grad_norm": 0.293571799993515,
"learning_rate": 0.00019656201080000228,
"loss": 0.0385,
"step": 1290
},
{
"epoch": 0.2704939500288094,
"grad_norm": 0.005718114785850048,
"learning_rate": 0.00019655626364757726,
"loss": 0.0001,
"step": 1291
},
{
"epoch": 0.27070347284060553,
"grad_norm": 0.03530212864279747,
"learning_rate": 0.00019655051177967813,
"loss": 0.0005,
"step": 1292
},
{
"epoch": 0.2709129956524017,
"grad_norm": 0.31564685702323914,
"learning_rate": 0.0001965447551965858,
"loss": 0.0443,
"step": 1293
},
{
"epoch": 0.27112251846419777,
"grad_norm": 0.05824049189686775,
"learning_rate": 0.00019653899389858136,
"loss": 0.0004,
"step": 1294
},
{
"epoch": 0.2713320412759939,
"grad_norm": 0.06219063326716423,
"learning_rate": 0.00019653322788594618,
"loss": 0.0005,
"step": 1295
},
{
"epoch": 0.27154156408779007,
"grad_norm": 0.0012616126332432032,
"learning_rate": 0.0001965274571589619,
"loss": 0.0,
"step": 1296
},
{
"epoch": 0.2717510868995862,
"grad_norm": 0.10456488281488419,
"learning_rate": 0.00019652168171791027,
"loss": 0.0123,
"step": 1297
},
{
"epoch": 0.2719606097113823,
"grad_norm": 0.027907853946089745,
"learning_rate": 0.0001965159015630734,
"loss": 0.0004,
"step": 1298
},
{
"epoch": 0.27217013252317845,
"grad_norm": 0.1813032478094101,
"learning_rate": 0.00019651011669473355,
"loss": 0.0189,
"step": 1299
},
{
"epoch": 0.2723796553349746,
"grad_norm": 0.1199437752366066,
"learning_rate": 0.00019650432711317323,
"loss": 0.002,
"step": 1300
},
{
"epoch": 0.27258917814677075,
"grad_norm": 0.009131750091910362,
"learning_rate": 0.0001964985328186752,
"loss": 0.0002,
"step": 1301
},
{
"epoch": 0.27279870095856684,
"grad_norm": 0.13628257811069489,
"learning_rate": 0.0001964927338115224,
"loss": 0.0211,
"step": 1302
},
{
"epoch": 0.273008223770363,
"grad_norm": 0.16198435425758362,
"learning_rate": 0.00019648693009199808,
"loss": 0.0293,
"step": 1303
},
{
"epoch": 0.27321774658215914,
"grad_norm": 0.09153685718774796,
"learning_rate": 0.0001964811216603856,
"loss": 0.0123,
"step": 1304
},
{
"epoch": 0.2734272693939553,
"grad_norm": 0.13093078136444092,
"learning_rate": 0.00019647530851696875,
"loss": 0.0074,
"step": 1305
},
{
"epoch": 0.2736367922057514,
"grad_norm": 0.02477131597697735,
"learning_rate": 0.00019646949066203127,
"loss": 0.0004,
"step": 1306
},
{
"epoch": 0.2738463150175475,
"grad_norm": 0.17570193111896515,
"learning_rate": 0.00019646366809585738,
"loss": 0.0128,
"step": 1307
},
{
"epoch": 0.2740558378293437,
"grad_norm": 0.03512417525053024,
"learning_rate": 0.00019645784081873143,
"loss": 0.0017,
"step": 1308
},
{
"epoch": 0.2742653606411398,
"grad_norm": 0.03261154517531395,
"learning_rate": 0.00019645200883093797,
"loss": 0.0004,
"step": 1309
},
{
"epoch": 0.2744748834529359,
"grad_norm": 0.200796976685524,
"learning_rate": 0.00019644617213276187,
"loss": 0.0433,
"step": 1310
},
{
"epoch": 0.27468440626473206,
"grad_norm": 0.18961785733699799,
"learning_rate": 0.00019644033072448812,
"loss": 0.0051,
"step": 1311
},
{
"epoch": 0.2748939290765282,
"grad_norm": 0.0028431613463908434,
"learning_rate": 0.000196434484606402,
"loss": 0.0001,
"step": 1312
},
{
"epoch": 0.27510345188832436,
"grad_norm": 0.0011763208312913775,
"learning_rate": 0.000196428633778789,
"loss": 0.0,
"step": 1313
},
{
"epoch": 0.27531297470012045,
"grad_norm": 0.28432533144950867,
"learning_rate": 0.00019642277824193492,
"loss": 0.0259,
"step": 1314
},
{
"epoch": 0.2755224975119166,
"grad_norm": 0.09515893459320068,
"learning_rate": 0.00019641691799612566,
"loss": 0.0082,
"step": 1315
},
{
"epoch": 0.27573202032371275,
"grad_norm": 0.16080771386623383,
"learning_rate": 0.00019641105304164742,
"loss": 0.0087,
"step": 1316
},
{
"epoch": 0.2759415431355089,
"grad_norm": 0.14703740179538727,
"learning_rate": 0.00019640518337878668,
"loss": 0.0037,
"step": 1317
},
{
"epoch": 0.276151065947305,
"grad_norm": 0.21766169369220734,
"learning_rate": 0.00019639930900783003,
"loss": 0.0155,
"step": 1318
},
{
"epoch": 0.27636058875910113,
"grad_norm": 0.12979577481746674,
"learning_rate": 0.0001963934299290644,
"loss": 0.0168,
"step": 1319
},
{
"epoch": 0.2765701115708973,
"grad_norm": 0.20718175172805786,
"learning_rate": 0.00019638754614277685,
"loss": 0.0152,
"step": 1320
},
{
"epoch": 0.27677963438269343,
"grad_norm": 0.19059571623802185,
"learning_rate": 0.00019638165764925477,
"loss": 0.0206,
"step": 1321
},
{
"epoch": 0.2769891571944896,
"grad_norm": 0.07117602229118347,
"learning_rate": 0.00019637576444878573,
"loss": 0.0007,
"step": 1322
},
{
"epoch": 0.27719868000628567,
"grad_norm": 0.014720194973051548,
"learning_rate": 0.00019636986654165752,
"loss": 0.0002,
"step": 1323
},
{
"epoch": 0.2774082028180818,
"grad_norm": 0.08522726595401764,
"learning_rate": 0.0001963639639281582,
"loss": 0.0023,
"step": 1324
},
{
"epoch": 0.27761772562987796,
"grad_norm": 0.24886921048164368,
"learning_rate": 0.00019635805660857595,
"loss": 0.0334,
"step": 1325
},
{
"epoch": 0.2778272484416741,
"grad_norm": 0.10195689648389816,
"learning_rate": 0.00019635214458319937,
"loss": 0.0209,
"step": 1326
},
{
"epoch": 0.2780367712534702,
"grad_norm": 0.13332587480545044,
"learning_rate": 0.00019634622785231717,
"loss": 0.0012,
"step": 1327
},
{
"epoch": 0.27824629406526635,
"grad_norm": 0.19609646499156952,
"learning_rate": 0.0001963403064162182,
"loss": 0.0038,
"step": 1328
},
{
"epoch": 0.2784558168770625,
"grad_norm": 0.23733653128147125,
"learning_rate": 0.00019633438027519176,
"loss": 0.007,
"step": 1329
},
{
"epoch": 0.27866533968885865,
"grad_norm": 0.10206770151853561,
"learning_rate": 0.00019632844942952719,
"loss": 0.0131,
"step": 1330
},
{
"epoch": 0.27887486250065474,
"grad_norm": 0.080831378698349,
"learning_rate": 0.00019632251387951413,
"loss": 0.0032,
"step": 1331
},
{
"epoch": 0.2790843853124509,
"grad_norm": 0.14923220872879028,
"learning_rate": 0.0001963165736254425,
"loss": 0.0066,
"step": 1332
},
{
"epoch": 0.27929390812424704,
"grad_norm": 0.12396574765443802,
"learning_rate": 0.0001963106286676024,
"loss": 0.0093,
"step": 1333
},
{
"epoch": 0.2795034309360432,
"grad_norm": 0.19563056528568268,
"learning_rate": 0.00019630467900628412,
"loss": 0.0277,
"step": 1334
},
{
"epoch": 0.2797129537478393,
"grad_norm": 0.038371458649635315,
"learning_rate": 0.00019629872464177825,
"loss": 0.0004,
"step": 1335
},
{
"epoch": 0.2799224765596354,
"grad_norm": 0.26363372802734375,
"learning_rate": 0.00019629276557437558,
"loss": 0.0464,
"step": 1336
},
{
"epoch": 0.2801319993714316,
"grad_norm": 0.03453537076711655,
"learning_rate": 0.0001962868018043671,
"loss": 0.0005,
"step": 1337
},
{
"epoch": 0.2803415221832277,
"grad_norm": 0.0300081018358469,
"learning_rate": 0.0001962808333320441,
"loss": 0.0004,
"step": 1338
},
{
"epoch": 0.2805510449950238,
"grad_norm": 0.13338404893875122,
"learning_rate": 0.00019627486015769802,
"loss": 0.0076,
"step": 1339
},
{
"epoch": 0.28076056780681996,
"grad_norm": 0.1329384744167328,
"learning_rate": 0.0001962688822816206,
"loss": 0.0009,
"step": 1340
},
{
"epoch": 0.2809700906186161,
"grad_norm": 0.18859624862670898,
"learning_rate": 0.00019626289970410378,
"loss": 0.0167,
"step": 1341
},
{
"epoch": 0.28117961343041226,
"grad_norm": 0.13575711846351624,
"learning_rate": 0.00019625691242543971,
"loss": 0.019,
"step": 1342
},
{
"epoch": 0.28138913624220835,
"grad_norm": 0.010435916483402252,
"learning_rate": 0.00019625092044592082,
"loss": 0.0001,
"step": 1343
},
{
"epoch": 0.2815986590540045,
"grad_norm": 0.1329658329486847,
"learning_rate": 0.00019624492376583966,
"loss": 0.0081,
"step": 1344
},
{
"epoch": 0.28180818186580064,
"grad_norm": 0.09858644008636475,
"learning_rate": 0.00019623892238548917,
"loss": 0.0025,
"step": 1345
},
{
"epoch": 0.2820177046775968,
"grad_norm": 0.3315856456756592,
"learning_rate": 0.00019623291630516243,
"loss": 0.0332,
"step": 1346
},
{
"epoch": 0.2822272274893929,
"grad_norm": 0.11820532381534576,
"learning_rate": 0.0001962269055251527,
"loss": 0.0014,
"step": 1347
},
{
"epoch": 0.28243675030118903,
"grad_norm": 0.17505581676959991,
"learning_rate": 0.00019622089004575356,
"loss": 0.0208,
"step": 1348
},
{
"epoch": 0.2826462731129852,
"grad_norm": 0.0583079494535923,
"learning_rate": 0.00019621486986725882,
"loss": 0.0005,
"step": 1349
},
{
"epoch": 0.28285579592478133,
"grad_norm": 0.1627739816904068,
"learning_rate": 0.0001962088449899624,
"loss": 0.0179,
"step": 1350
},
{
"epoch": 0.2830653187365774,
"grad_norm": 0.22073708474636078,
"learning_rate": 0.00019620281541415862,
"loss": 0.0338,
"step": 1351
},
{
"epoch": 0.28327484154837357,
"grad_norm": 0.02409421280026436,
"learning_rate": 0.0001961967811401419,
"loss": 0.0002,
"step": 1352
},
{
"epoch": 0.2834843643601697,
"grad_norm": 0.09562385827302933,
"learning_rate": 0.0001961907421682069,
"loss": 0.0016,
"step": 1353
},
{
"epoch": 0.28369388717196586,
"grad_norm": 0.07804533839225769,
"learning_rate": 0.00019618469849864864,
"loss": 0.0008,
"step": 1354
},
{
"epoch": 0.28390340998376196,
"grad_norm": 0.23339737951755524,
"learning_rate": 0.00019617865013176218,
"loss": 0.0061,
"step": 1355
},
{
"epoch": 0.2841129327955581,
"grad_norm": 0.12158526480197906,
"learning_rate": 0.00019617259706784297,
"loss": 0.0032,
"step": 1356
},
{
"epoch": 0.28432245560735425,
"grad_norm": 0.056320130825042725,
"learning_rate": 0.00019616653930718653,
"loss": 0.0011,
"step": 1357
},
{
"epoch": 0.2845319784191504,
"grad_norm": 0.1585444062948227,
"learning_rate": 0.0001961604768500888,
"loss": 0.0091,
"step": 1358
},
{
"epoch": 0.2847415012309465,
"grad_norm": 0.11987308412790298,
"learning_rate": 0.0001961544096968458,
"loss": 0.0052,
"step": 1359
},
{
"epoch": 0.28495102404274264,
"grad_norm": 0.3789823353290558,
"learning_rate": 0.00019614833784775383,
"loss": 0.0714,
"step": 1360
},
{
"epoch": 0.2851605468545388,
"grad_norm": 0.2178063690662384,
"learning_rate": 0.00019614226130310942,
"loss": 0.015,
"step": 1361
},
{
"epoch": 0.28537006966633494,
"grad_norm": 0.17697137594223022,
"learning_rate": 0.00019613618006320932,
"loss": 0.026,
"step": 1362
},
{
"epoch": 0.2855795924781311,
"grad_norm": 0.1952069103717804,
"learning_rate": 0.00019613009412835055,
"loss": 0.0192,
"step": 1363
},
{
"epoch": 0.2857891152899272,
"grad_norm": 0.24872924387454987,
"learning_rate": 0.00019612400349883032,
"loss": 0.023,
"step": 1364
},
{
"epoch": 0.2859986381017233,
"grad_norm": 0.0086398059502244,
"learning_rate": 0.00019611790817494602,
"loss": 0.0001,
"step": 1365
},
{
"epoch": 0.28620816091351947,
"grad_norm": 0.3663384020328522,
"learning_rate": 0.00019611180815699538,
"loss": 0.039,
"step": 1366
},
{
"epoch": 0.2864176837253156,
"grad_norm": 0.03998996317386627,
"learning_rate": 0.00019610570344527628,
"loss": 0.0005,
"step": 1367
},
{
"epoch": 0.2866272065371117,
"grad_norm": 0.04309941083192825,
"learning_rate": 0.00019609959404008685,
"loss": 0.0005,
"step": 1368
},
{
"epoch": 0.28683672934890786,
"grad_norm": 0.0018193083815276623,
"learning_rate": 0.00019609347994172546,
"loss": 0.0,
"step": 1369
},
{
"epoch": 0.287046252160704,
"grad_norm": 0.2627018690109253,
"learning_rate": 0.0001960873611504907,
"loss": 0.0188,
"step": 1370
},
{
"epoch": 0.28725577497250016,
"grad_norm": 0.08187177032232285,
"learning_rate": 0.00019608123766668144,
"loss": 0.0025,
"step": 1371
},
{
"epoch": 0.28746529778429625,
"grad_norm": 0.015533563680946827,
"learning_rate": 0.00019607510949059664,
"loss": 0.0002,
"step": 1372
},
{
"epoch": 0.2876748205960924,
"grad_norm": 0.001533325994387269,
"learning_rate": 0.00019606897662253564,
"loss": 0.0,
"step": 1373
},
{
"epoch": 0.28788434340788854,
"grad_norm": 0.014341753907501698,
"learning_rate": 0.00019606283906279791,
"loss": 0.0002,
"step": 1374
},
{
"epoch": 0.2880938662196847,
"grad_norm": 0.01508825272321701,
"learning_rate": 0.00019605669681168323,
"loss": 0.0002,
"step": 1375
},
{
"epoch": 0.2883033890314808,
"grad_norm": 0.11747533828020096,
"learning_rate": 0.00019605054986949157,
"loss": 0.001,
"step": 1376
},
{
"epoch": 0.28851291184327693,
"grad_norm": 0.14469660818576813,
"learning_rate": 0.00019604439823652306,
"loss": 0.0178,
"step": 1377
},
{
"epoch": 0.2887224346550731,
"grad_norm": 0.10356391221284866,
"learning_rate": 0.00019603824191307816,
"loss": 0.0014,
"step": 1378
},
{
"epoch": 0.2889319574668692,
"grad_norm": 0.0018149783136323094,
"learning_rate": 0.00019603208089945755,
"loss": 0.0,
"step": 1379
},
{
"epoch": 0.2891414802786653,
"grad_norm": 0.06800977885723114,
"learning_rate": 0.00019602591519596207,
"loss": 0.0007,
"step": 1380
},
{
"epoch": 0.28935100309046147,
"grad_norm": 0.05669765546917915,
"learning_rate": 0.00019601974480289286,
"loss": 0.0043,
"step": 1381
},
{
"epoch": 0.2895605259022576,
"grad_norm": 0.1664445847272873,
"learning_rate": 0.00019601356972055126,
"loss": 0.0083,
"step": 1382
},
{
"epoch": 0.28977004871405376,
"grad_norm": 0.2715985178947449,
"learning_rate": 0.00019600738994923883,
"loss": 0.0166,
"step": 1383
},
{
"epoch": 0.28997957152584986,
"grad_norm": 0.011416422203183174,
"learning_rate": 0.00019600120548925735,
"loss": 0.0001,
"step": 1384
},
{
"epoch": 0.290189094337646,
"grad_norm": 0.11940954625606537,
"learning_rate": 0.0001959950163409089,
"loss": 0.0051,
"step": 1385
},
{
"epoch": 0.29039861714944215,
"grad_norm": 0.20202504098415375,
"learning_rate": 0.00019598882250449569,
"loss": 0.0281,
"step": 1386
},
{
"epoch": 0.2906081399612383,
"grad_norm": 0.2038898915052414,
"learning_rate": 0.00019598262398032022,
"loss": 0.0102,
"step": 1387
},
{
"epoch": 0.2908176627730344,
"grad_norm": 0.0836292952299118,
"learning_rate": 0.00019597642076868522,
"loss": 0.0012,
"step": 1388
},
{
"epoch": 0.29102718558483054,
"grad_norm": 0.2007703334093094,
"learning_rate": 0.00019597021286989362,
"loss": 0.0381,
"step": 1389
},
{
"epoch": 0.2912367083966267,
"grad_norm": 0.04771614074707031,
"learning_rate": 0.00019596400028424856,
"loss": 0.0024,
"step": 1390
},
{
"epoch": 0.29144623120842283,
"grad_norm": 0.16608139872550964,
"learning_rate": 0.0001959577830120535,
"loss": 0.0029,
"step": 1391
},
{
"epoch": 0.2916557540202189,
"grad_norm": 0.06598784774541855,
"learning_rate": 0.000195951561053612,
"loss": 0.0006,
"step": 1392
},
{
"epoch": 0.2918652768320151,
"grad_norm": 0.001551535096950829,
"learning_rate": 0.00019594533440922803,
"loss": 0.0,
"step": 1393
},
{
"epoch": 0.2920747996438112,
"grad_norm": 0.005216945894062519,
"learning_rate": 0.00019593910307920555,
"loss": 0.0001,
"step": 1394
},
{
"epoch": 0.29228432245560737,
"grad_norm": 0.21973158419132233,
"learning_rate": 0.00019593286706384898,
"loss": 0.0147,
"step": 1395
},
{
"epoch": 0.29249384526740346,
"grad_norm": 0.24168424308300018,
"learning_rate": 0.0001959266263634628,
"loss": 0.0376,
"step": 1396
},
{
"epoch": 0.2927033680791996,
"grad_norm": 0.03945903480052948,
"learning_rate": 0.0001959203809783518,
"loss": 0.0003,
"step": 1397
},
{
"epoch": 0.29291289089099576,
"grad_norm": 0.006036645267158747,
"learning_rate": 0.00019591413090882097,
"loss": 0.0001,
"step": 1398
},
{
"epoch": 0.2931224137027919,
"grad_norm": 0.16238859295845032,
"learning_rate": 0.00019590787615517558,
"loss": 0.0151,
"step": 1399
},
{
"epoch": 0.293331936514588,
"grad_norm": 0.024547500535845757,
"learning_rate": 0.0001959016167177211,
"loss": 0.0,
"step": 1400
},
{
"epoch": 0.293331936514588,
"eval_loss": 0.009350336156785488,
"eval_runtime": 89.0528,
"eval_samples_per_second": 8.624,
"eval_steps_per_second": 4.312,
"step": 1400
},
{
"epoch": 0.29354145932638415,
"grad_norm": 0.1304274946451187,
"learning_rate": 0.00019589535259676314,
"loss": 0.0031,
"step": 1401
},
{
"epoch": 0.2937509821381803,
"grad_norm": 0.14638224244117737,
"learning_rate": 0.0001958890837926077,
"loss": 0.0062,
"step": 1402
},
{
"epoch": 0.29396050494997644,
"grad_norm": 0.19852378964424133,
"learning_rate": 0.0001958828103055609,
"loss": 0.0349,
"step": 1403
},
{
"epoch": 0.2941700277617726,
"grad_norm": 0.1417485922574997,
"learning_rate": 0.0001958765321359291,
"loss": 0.0071,
"step": 1404
},
{
"epoch": 0.2943795505735687,
"grad_norm": 0.08396097272634506,
"learning_rate": 0.0001958702492840189,
"loss": 0.0027,
"step": 1405
},
{
"epoch": 0.29458907338536483,
"grad_norm": 0.005264434032142162,
"learning_rate": 0.00019586396175013714,
"loss": 0.0001,
"step": 1406
},
{
"epoch": 0.294798596197161,
"grad_norm": 0.005200863350182772,
"learning_rate": 0.00019585766953459093,
"loss": 0.0001,
"step": 1407
},
{
"epoch": 0.2950081190089571,
"grad_norm": 0.08635362237691879,
"learning_rate": 0.00019585137263768748,
"loss": 0.0092,
"step": 1408
},
{
"epoch": 0.2952176418207532,
"grad_norm": 0.030663859099149704,
"learning_rate": 0.00019584507105973437,
"loss": 0.0004,
"step": 1409
},
{
"epoch": 0.29542716463254937,
"grad_norm": 0.2146034836769104,
"learning_rate": 0.0001958387648010393,
"loss": 0.0028,
"step": 1410
},
{
"epoch": 0.2956366874443455,
"grad_norm": 0.008759552612900734,
"learning_rate": 0.0001958324538619103,
"loss": 0.0001,
"step": 1411
},
{
"epoch": 0.29584621025614166,
"grad_norm": 0.29600149393081665,
"learning_rate": 0.00019582613824265551,
"loss": 0.0148,
"step": 1412
},
{
"epoch": 0.29605573306793775,
"grad_norm": 0.14502890408039093,
"learning_rate": 0.00019581981794358343,
"loss": 0.0051,
"step": 1413
},
{
"epoch": 0.2962652558797339,
"grad_norm": 0.16055640578269958,
"learning_rate": 0.00019581349296500267,
"loss": 0.0112,
"step": 1414
},
{
"epoch": 0.29647477869153005,
"grad_norm": 0.0015173458959907293,
"learning_rate": 0.00019580716330722215,
"loss": 0.0,
"step": 1415
},
{
"epoch": 0.2966843015033262,
"grad_norm": 0.204542875289917,
"learning_rate": 0.000195800828970551,
"loss": 0.0219,
"step": 1416
},
{
"epoch": 0.2968938243151223,
"grad_norm": 0.039979007095098495,
"learning_rate": 0.0001957944899552985,
"loss": 0.0004,
"step": 1417
},
{
"epoch": 0.29710334712691844,
"grad_norm": 0.011981209740042686,
"learning_rate": 0.00019578814626177427,
"loss": 0.0002,
"step": 1418
},
{
"epoch": 0.2973128699387146,
"grad_norm": 0.12856760621070862,
"learning_rate": 0.00019578179789028814,
"loss": 0.0028,
"step": 1419
},
{
"epoch": 0.29752239275051073,
"grad_norm": 0.2889501452445984,
"learning_rate": 0.00019577544484115014,
"loss": 0.0528,
"step": 1420
},
{
"epoch": 0.2977319155623068,
"grad_norm": 0.10836676508188248,
"learning_rate": 0.00019576908711467045,
"loss": 0.0012,
"step": 1421
},
{
"epoch": 0.297941438374103,
"grad_norm": 0.0011059404350817204,
"learning_rate": 0.00019576272471115967,
"loss": 0.0,
"step": 1422
},
{
"epoch": 0.2981509611858991,
"grad_norm": 0.09822409600019455,
"learning_rate": 0.00019575635763092843,
"loss": 0.0021,
"step": 1423
},
{
"epoch": 0.29836048399769527,
"grad_norm": 0.1385686993598938,
"learning_rate": 0.00019574998587428772,
"loss": 0.0184,
"step": 1424
},
{
"epoch": 0.29857000680949136,
"grad_norm": 0.01703231781721115,
"learning_rate": 0.00019574360944154873,
"loss": 0.0002,
"step": 1425
},
{
"epoch": 0.2987795296212875,
"grad_norm": 0.192989781498909,
"learning_rate": 0.00019573722833302282,
"loss": 0.025,
"step": 1426
},
{
"epoch": 0.29898905243308366,
"grad_norm": 0.15308697521686554,
"learning_rate": 0.00019573084254902162,
"loss": 0.014,
"step": 1427
},
{
"epoch": 0.2991985752448798,
"grad_norm": 0.017074331641197205,
"learning_rate": 0.00019572445208985704,
"loss": 0.0002,
"step": 1428
},
{
"epoch": 0.2994080980566759,
"grad_norm": 0.12004858255386353,
"learning_rate": 0.00019571805695584114,
"loss": 0.0045,
"step": 1429
},
{
"epoch": 0.29961762086847205,
"grad_norm": 0.13323943316936493,
"learning_rate": 0.0001957116571472862,
"loss": 0.0019,
"step": 1430
},
{
"epoch": 0.2998271436802682,
"grad_norm": 0.2345772534608841,
"learning_rate": 0.00019570525266450486,
"loss": 0.0306,
"step": 1431
},
{
"epoch": 0.30003666649206434,
"grad_norm": 0.1890985071659088,
"learning_rate": 0.00019569884350780975,
"loss": 0.0197,
"step": 1432
},
{
"epoch": 0.30024618930386043,
"grad_norm": 0.0014320454793050885,
"learning_rate": 0.00019569242967751402,
"loss": 0.0,
"step": 1433
},
{
"epoch": 0.3004557121156566,
"grad_norm": 0.10571502149105072,
"learning_rate": 0.0001956860111739308,
"loss": 0.0114,
"step": 1434
},
{
"epoch": 0.30066523492745273,
"grad_norm": 0.08180177211761475,
"learning_rate": 0.0001956795879973736,
"loss": 0.0013,
"step": 1435
},
{
"epoch": 0.3008747577392489,
"grad_norm": 0.0072664557956159115,
"learning_rate": 0.00019567316014815607,
"loss": 0.0001,
"step": 1436
},
{
"epoch": 0.30108428055104497,
"grad_norm": 0.13646478950977325,
"learning_rate": 0.00019566672762659214,
"loss": 0.0027,
"step": 1437
},
{
"epoch": 0.3012938033628411,
"grad_norm": 0.0891512855887413,
"learning_rate": 0.00019566029043299594,
"loss": 0.0023,
"step": 1438
},
{
"epoch": 0.30150332617463727,
"grad_norm": 0.08809952437877655,
"learning_rate": 0.00019565384856768187,
"loss": 0.002,
"step": 1439
},
{
"epoch": 0.3017128489864334,
"grad_norm": 0.10799501836299896,
"learning_rate": 0.0001956474020309645,
"loss": 0.0227,
"step": 1440
},
{
"epoch": 0.3019223717982295,
"grad_norm": 0.007641882169991732,
"learning_rate": 0.00019564095082315864,
"loss": 0.0001,
"step": 1441
},
{
"epoch": 0.30213189461002565,
"grad_norm": 0.3086688816547394,
"learning_rate": 0.0001956344949445794,
"loss": 0.0412,
"step": 1442
},
{
"epoch": 0.3023414174218218,
"grad_norm": 26.637779235839844,
"learning_rate": 0.00019562803439554207,
"loss": 0.0136,
"step": 1443
},
{
"epoch": 0.30255094023361795,
"grad_norm": 0.1183343455195427,
"learning_rate": 0.0001956215691763621,
"loss": 0.0037,
"step": 1444
},
{
"epoch": 0.3027604630454141,
"grad_norm": 0.0019468015525490046,
"learning_rate": 0.00019561509928735525,
"loss": 0.0,
"step": 1445
},
{
"epoch": 0.3029699858572102,
"grad_norm": 0.02043256163597107,
"learning_rate": 0.00019560862472883748,
"loss": 0.0002,
"step": 1446
},
{
"epoch": 0.30317950866900634,
"grad_norm": 0.07333844900131226,
"learning_rate": 0.000195602145501125,
"loss": 0.0087,
"step": 1447
},
{
"epoch": 0.3033890314808025,
"grad_norm": 0.27351680397987366,
"learning_rate": 0.0001955956616045342,
"loss": 0.0139,
"step": 1448
},
{
"epoch": 0.30359855429259863,
"grad_norm": 0.11866743117570877,
"learning_rate": 0.0001955891730393818,
"loss": 0.0069,
"step": 1449
},
{
"epoch": 0.3038080771043947,
"grad_norm": 0.16603772342205048,
"learning_rate": 0.00019558267980598464,
"loss": 0.0185,
"step": 1450
},
{
"epoch": 0.3040175999161909,
"grad_norm": 0.17106042802333832,
"learning_rate": 0.0001955761819046598,
"loss": 0.0068,
"step": 1451
},
{
"epoch": 0.304227122727987,
"grad_norm": 0.13401584327220917,
"learning_rate": 0.00019556967933572465,
"loss": 0.0081,
"step": 1452
},
{
"epoch": 0.30443664553978317,
"grad_norm": 0.2146846204996109,
"learning_rate": 0.00019556317209949676,
"loss": 0.0317,
"step": 1453
},
{
"epoch": 0.30464616835157926,
"grad_norm": 0.18583877384662628,
"learning_rate": 0.0001955566601962939,
"loss": 0.0255,
"step": 1454
},
{
"epoch": 0.3048556911633754,
"grad_norm": 0.0639459639787674,
"learning_rate": 0.00019555014362643408,
"loss": 0.0006,
"step": 1455
},
{
"epoch": 0.30506521397517156,
"grad_norm": 0.07844740152359009,
"learning_rate": 0.00019554362239023555,
"loss": 0.0004,
"step": 1456
},
{
"epoch": 0.3052747367869677,
"grad_norm": 0.04650677740573883,
"learning_rate": 0.00019553709648801682,
"loss": 0.0005,
"step": 1457
},
{
"epoch": 0.3054842595987638,
"grad_norm": 0.22158075869083405,
"learning_rate": 0.00019553056592009656,
"loss": 0.026,
"step": 1458
},
{
"epoch": 0.30569378241055994,
"grad_norm": 0.0060492088086903095,
"learning_rate": 0.00019552403068679372,
"loss": 0.0001,
"step": 1459
},
{
"epoch": 0.3059033052223561,
"grad_norm": 0.07878894358873367,
"learning_rate": 0.00019551749078842743,
"loss": 0.0009,
"step": 1460
},
{
"epoch": 0.30611282803415224,
"grad_norm": 0.06819705665111542,
"learning_rate": 0.0001955109462253171,
"loss": 0.0109,
"step": 1461
},
{
"epoch": 0.30632235084594833,
"grad_norm": 0.06026251241564751,
"learning_rate": 0.00019550439699778232,
"loss": 0.0007,
"step": 1462
},
{
"epoch": 0.3065318736577445,
"grad_norm": 0.6910133361816406,
"learning_rate": 0.000195497843106143,
"loss": 0.03,
"step": 1463
},
{
"epoch": 0.30674139646954063,
"grad_norm": 0.0064484793692827225,
"learning_rate": 0.0001954912845507191,
"loss": 0.0001,
"step": 1464
},
{
"epoch": 0.3069509192813368,
"grad_norm": 0.38042137026786804,
"learning_rate": 0.000195484721331831,
"loss": 0.0715,
"step": 1465
},
{
"epoch": 0.30716044209313287,
"grad_norm": 0.07180149853229523,
"learning_rate": 0.0001954781534497992,
"loss": 0.0074,
"step": 1466
},
{
"epoch": 0.307369964904929,
"grad_norm": 0.12332933396100998,
"learning_rate": 0.0001954715809049444,
"loss": 0.0123,
"step": 1467
},
{
"epoch": 0.30757948771672516,
"grad_norm": 0.12823301553726196,
"learning_rate": 0.0001954650036975877,
"loss": 0.0067,
"step": 1468
},
{
"epoch": 0.3077890105285213,
"grad_norm": 0.12684230506420135,
"learning_rate": 0.00019545842182805024,
"loss": 0.0029,
"step": 1469
},
{
"epoch": 0.3079985333403174,
"grad_norm": 0.29130974411964417,
"learning_rate": 0.0001954518352966534,
"loss": 0.0129,
"step": 1470
},
{
"epoch": 0.30820805615211355,
"grad_norm": 0.029337555170059204,
"learning_rate": 0.00019544524410371894,
"loss": 0.0002,
"step": 1471
},
{
"epoch": 0.3084175789639097,
"grad_norm": 0.3716062307357788,
"learning_rate": 0.0001954386482495687,
"loss": 0.0503,
"step": 1472
},
{
"epoch": 0.30862710177570585,
"grad_norm": 0.15721499919891357,
"learning_rate": 0.00019543204773452482,
"loss": 0.0231,
"step": 1473
},
{
"epoch": 0.30883662458750194,
"grad_norm": 0.1504199206829071,
"learning_rate": 0.00019542544255890962,
"loss": 0.0046,
"step": 1474
},
{
"epoch": 0.3090461473992981,
"grad_norm": 0.18560700118541718,
"learning_rate": 0.00019541883272304573,
"loss": 0.0251,
"step": 1475
},
{
"epoch": 0.30925567021109424,
"grad_norm": 0.22650940716266632,
"learning_rate": 0.00019541221822725585,
"loss": 0.0067,
"step": 1476
},
{
"epoch": 0.3094651930228904,
"grad_norm": 0.17605389654636383,
"learning_rate": 0.0001954055990718631,
"loss": 0.0119,
"step": 1477
},
{
"epoch": 0.3096747158346865,
"grad_norm": 0.010634154081344604,
"learning_rate": 0.00019539897525719073,
"loss": 0.0001,
"step": 1478
},
{
"epoch": 0.3098842386464826,
"grad_norm": 0.17018188536167145,
"learning_rate": 0.00019539234678356217,
"loss": 0.0182,
"step": 1479
},
{
"epoch": 0.31009376145827877,
"grad_norm": 0.009977507404983044,
"learning_rate": 0.00019538571365130118,
"loss": 0.0001,
"step": 1480
},
{
"epoch": 0.3103032842700749,
"grad_norm": 0.018772488459944725,
"learning_rate": 0.00019537907586073168,
"loss": 0.0003,
"step": 1481
},
{
"epoch": 0.310512807081871,
"grad_norm": 0.32882487773895264,
"learning_rate": 0.0001953724334121778,
"loss": 0.008,
"step": 1482
},
{
"epoch": 0.31072232989366716,
"grad_norm": 0.17600393295288086,
"learning_rate": 0.00019536578630596405,
"loss": 0.008,
"step": 1483
},
{
"epoch": 0.3109318527054633,
"grad_norm": 0.18962863087654114,
"learning_rate": 0.00019535913454241493,
"loss": 0.0134,
"step": 1484
},
{
"epoch": 0.31114137551725946,
"grad_norm": 0.0949021726846695,
"learning_rate": 0.00019535247812185532,
"loss": 0.011,
"step": 1485
},
{
"epoch": 0.3113508983290556,
"grad_norm": 0.0024588159285485744,
"learning_rate": 0.00019534581704461033,
"loss": 0.0,
"step": 1486
},
{
"epoch": 0.3115604211408517,
"grad_norm": 0.22583766281604767,
"learning_rate": 0.00019533915131100525,
"loss": 0.0037,
"step": 1487
},
{
"epoch": 0.31176994395264784,
"grad_norm": 0.13111519813537598,
"learning_rate": 0.0001953324809213656,
"loss": 0.0106,
"step": 1488
},
{
"epoch": 0.311979466764444,
"grad_norm": 0.1254248470067978,
"learning_rate": 0.00019532580587601714,
"loss": 0.0156,
"step": 1489
},
{
"epoch": 0.31218898957624014,
"grad_norm": 0.2393532544374466,
"learning_rate": 0.00019531912617528585,
"loss": 0.0111,
"step": 1490
},
{
"epoch": 0.31239851238803623,
"grad_norm": 0.12270370125770569,
"learning_rate": 0.00019531244181949796,
"loss": 0.0076,
"step": 1491
},
{
"epoch": 0.3126080351998324,
"grad_norm": 0.2543415129184723,
"learning_rate": 0.00019530575280897992,
"loss": 0.0179,
"step": 1492
},
{
"epoch": 0.3128175580116285,
"grad_norm": 0.1431514471769333,
"learning_rate": 0.00019529905914405837,
"loss": 0.0011,
"step": 1493
},
{
"epoch": 0.3130270808234247,
"grad_norm": 0.20022700726985931,
"learning_rate": 0.00019529236082506022,
"loss": 0.0294,
"step": 1494
},
{
"epoch": 0.31323660363522077,
"grad_norm": 0.0343584343791008,
"learning_rate": 0.00019528565785231262,
"loss": 0.0005,
"step": 1495
},
{
"epoch": 0.3134461264470169,
"grad_norm": 0.0010648473398759961,
"learning_rate": 0.00019527895022614287,
"loss": 0.0,
"step": 1496
},
{
"epoch": 0.31365564925881306,
"grad_norm": 0.20229478180408478,
"learning_rate": 0.00019527223794687853,
"loss": 0.0246,
"step": 1497
},
{
"epoch": 0.3138651720706092,
"grad_norm": 0.014464381150901318,
"learning_rate": 0.0001952655210148475,
"loss": 0.0001,
"step": 1498
},
{
"epoch": 0.3140746948824053,
"grad_norm": 0.1936783790588379,
"learning_rate": 0.00019525879943037774,
"loss": 0.0045,
"step": 1499
},
{
"epoch": 0.31428421769420145,
"grad_norm": 0.012530984356999397,
"learning_rate": 0.0001952520731937975,
"loss": 0.0001,
"step": 1500
},
{
"epoch": 0.3144937405059976,
"grad_norm": 0.1337125450372696,
"learning_rate": 0.00019524534230543533,
"loss": 0.0089,
"step": 1501
},
{
"epoch": 0.31470326331779375,
"grad_norm": 0.22295622527599335,
"learning_rate": 0.00019523860676561988,
"loss": 0.0114,
"step": 1502
},
{
"epoch": 0.31491278612958984,
"grad_norm": 0.0035251390654593706,
"learning_rate": 0.0001952318665746801,
"loss": 0.0,
"step": 1503
},
{
"epoch": 0.315122308941386,
"grad_norm": 0.004611232317984104,
"learning_rate": 0.0001952251217329452,
"loss": 0.0001,
"step": 1504
},
{
"epoch": 0.31533183175318213,
"grad_norm": 0.002356649376451969,
"learning_rate": 0.00019521837224074454,
"loss": 0.0,
"step": 1505
},
{
"epoch": 0.3155413545649783,
"grad_norm": 0.027044709771871567,
"learning_rate": 0.00019521161809840773,
"loss": 0.0002,
"step": 1506
},
{
"epoch": 0.3157508773767744,
"grad_norm": 0.07268866151571274,
"learning_rate": 0.00019520485930626464,
"loss": 0.0004,
"step": 1507
},
{
"epoch": 0.3159604001885705,
"grad_norm": 0.010898386128246784,
"learning_rate": 0.00019519809586464537,
"loss": 0.0001,
"step": 1508
},
{
"epoch": 0.31616992300036667,
"grad_norm": 0.03874179720878601,
"learning_rate": 0.00019519132777388016,
"loss": 0.0006,
"step": 1509
},
{
"epoch": 0.3163794458121628,
"grad_norm": 0.005732359830290079,
"learning_rate": 0.0001951845550342996,
"loss": 0.0,
"step": 1510
},
{
"epoch": 0.3165889686239589,
"grad_norm": 0.2763909697532654,
"learning_rate": 0.00019517777764623443,
"loss": 0.0179,
"step": 1511
},
{
"epoch": 0.31679849143575506,
"grad_norm": 0.05047501623630524,
"learning_rate": 0.00019517099561001562,
"loss": 0.0005,
"step": 1512
},
{
"epoch": 0.3170080142475512,
"grad_norm": 0.0008512980421073735,
"learning_rate": 0.00019516420892597437,
"loss": 0.0,
"step": 1513
},
{
"epoch": 0.31721753705934735,
"grad_norm": 0.13507002592086792,
"learning_rate": 0.00019515741759444218,
"loss": 0.0166,
"step": 1514
},
{
"epoch": 0.31742705987114345,
"grad_norm": 0.18085655570030212,
"learning_rate": 0.00019515062161575063,
"loss": 0.0019,
"step": 1515
},
{
"epoch": 0.3176365826829396,
"grad_norm": 0.06925208121538162,
"learning_rate": 0.00019514382099023166,
"loss": 0.0008,
"step": 1516
},
{
"epoch": 0.31784610549473574,
"grad_norm": 0.17617951333522797,
"learning_rate": 0.00019513701571821738,
"loss": 0.0075,
"step": 1517
},
{
"epoch": 0.3180556283065319,
"grad_norm": 0.27964097261428833,
"learning_rate": 0.00019513020580004016,
"loss": 0.0222,
"step": 1518
},
{
"epoch": 0.318265151118328,
"grad_norm": 0.10818011313676834,
"learning_rate": 0.00019512339123603257,
"loss": 0.0009,
"step": 1519
},
{
"epoch": 0.31847467393012413,
"grad_norm": 0.2028188854455948,
"learning_rate": 0.00019511657202652736,
"loss": 0.0187,
"step": 1520
},
{
"epoch": 0.3186841967419203,
"grad_norm": 0.15811379253864288,
"learning_rate": 0.0001951097481718576,
"loss": 0.0138,
"step": 1521
},
{
"epoch": 0.3188937195537164,
"grad_norm": 0.18826134502887726,
"learning_rate": 0.00019510291967235653,
"loss": 0.0339,
"step": 1522
},
{
"epoch": 0.3191032423655125,
"grad_norm": 0.0027802777476608753,
"learning_rate": 0.00019509608652835763,
"loss": 0.0001,
"step": 1523
},
{
"epoch": 0.31931276517730867,
"grad_norm": 0.027045130729675293,
"learning_rate": 0.0001950892487401946,
"loss": 0.0018,
"step": 1524
},
{
"epoch": 0.3195222879891048,
"grad_norm": 0.05583294481039047,
"learning_rate": 0.00019508240630820137,
"loss": 0.0008,
"step": 1525
},
{
"epoch": 0.31973181080090096,
"grad_norm": 0.04633820801973343,
"learning_rate": 0.00019507555923271218,
"loss": 0.0004,
"step": 1526
},
{
"epoch": 0.3199413336126971,
"grad_norm": 0.23768991231918335,
"learning_rate": 0.0001950687075140613,
"loss": 0.0131,
"step": 1527
},
{
"epoch": 0.3201508564244932,
"grad_norm": 0.24667467176914215,
"learning_rate": 0.00019506185115258343,
"loss": 0.0397,
"step": 1528
},
{
"epoch": 0.32036037923628935,
"grad_norm": 0.04483865574002266,
"learning_rate": 0.00019505499014861333,
"loss": 0.0005,
"step": 1529
},
{
"epoch": 0.3205699020480855,
"grad_norm": 0.030646564438939095,
"learning_rate": 0.00019504812450248614,
"loss": 0.0005,
"step": 1530
},
{
"epoch": 0.32077942485988165,
"grad_norm": 0.04976831376552582,
"learning_rate": 0.0001950412542145371,
"loss": 0.0006,
"step": 1531
},
{
"epoch": 0.32098894767167774,
"grad_norm": 0.1266006976366043,
"learning_rate": 0.0001950343792851018,
"loss": 0.0027,
"step": 1532
},
{
"epoch": 0.3211984704834739,
"grad_norm": 0.07971259206533432,
"learning_rate": 0.00019502749971451596,
"loss": 0.0013,
"step": 1533
},
{
"epoch": 0.32140799329527003,
"grad_norm": 0.002387990476563573,
"learning_rate": 0.0001950206155031155,
"loss": 0.0,
"step": 1534
},
{
"epoch": 0.3216175161070662,
"grad_norm": 0.00598552543669939,
"learning_rate": 0.00019501372665123667,
"loss": 0.0,
"step": 1535
},
{
"epoch": 0.3218270389188623,
"grad_norm": 0.018055522814393044,
"learning_rate": 0.00019500683315921592,
"loss": 0.0002,
"step": 1536
},
{
"epoch": 0.3220365617306584,
"grad_norm": 0.29587745666503906,
"learning_rate": 0.00019499993502738986,
"loss": 0.0266,
"step": 1537
},
{
"epoch": 0.32224608454245457,
"grad_norm": 0.33568182587623596,
"learning_rate": 0.00019499303225609539,
"loss": 0.0128,
"step": 1538
},
{
"epoch": 0.3224556073542507,
"grad_norm": 0.020032409578561783,
"learning_rate": 0.0001949861248456696,
"loss": 0.0002,
"step": 1539
},
{
"epoch": 0.3226651301660468,
"grad_norm": 0.006906419992446899,
"learning_rate": 0.00019497921279644986,
"loss": 0.0001,
"step": 1540
},
{
"epoch": 0.32287465297784296,
"grad_norm": 0.01124792080372572,
"learning_rate": 0.0001949722961087737,
"loss": 0.0001,
"step": 1541
},
{
"epoch": 0.3230841757896391,
"grad_norm": 0.011126923374831676,
"learning_rate": 0.0001949653747829789,
"loss": 0.0001,
"step": 1542
},
{
"epoch": 0.32329369860143525,
"grad_norm": 0.0007546422421000898,
"learning_rate": 0.0001949584488194035,
"loss": 0.0,
"step": 1543
},
{
"epoch": 0.32350322141323135,
"grad_norm": 0.12543243169784546,
"learning_rate": 0.00019495151821838572,
"loss": 0.0123,
"step": 1544
},
{
"epoch": 0.3237127442250275,
"grad_norm": 0.014699449762701988,
"learning_rate": 0.0001949445829802641,
"loss": 0.0002,
"step": 1545
},
{
"epoch": 0.32392226703682364,
"grad_norm": 0.0024809150490909815,
"learning_rate": 0.0001949376431053772,
"loss": 0.0,
"step": 1546
},
{
"epoch": 0.3241317898486198,
"grad_norm": 0.22279804944992065,
"learning_rate": 0.000194930698594064,
"loss": 0.0312,
"step": 1547
},
{
"epoch": 0.3243413126604159,
"grad_norm": 0.11176104843616486,
"learning_rate": 0.00019492374944666367,
"loss": 0.015,
"step": 1548
},
{
"epoch": 0.32455083547221203,
"grad_norm": 0.09751889854669571,
"learning_rate": 0.0001949167956635156,
"loss": 0.0007,
"step": 1549
},
{
"epoch": 0.3247603582840082,
"grad_norm": 0.15919186174869537,
"learning_rate": 0.00019490983724495932,
"loss": 0.0017,
"step": 1550
},
{
"epoch": 0.3249698810958043,
"grad_norm": 0.13658560812473297,
"learning_rate": 0.00019490287419133472,
"loss": 0.0022,
"step": 1551
},
{
"epoch": 0.3251794039076004,
"grad_norm": 0.21580643951892853,
"learning_rate": 0.0001948959065029818,
"loss": 0.0233,
"step": 1552
},
{
"epoch": 0.32538892671939657,
"grad_norm": 0.009592859074473381,
"learning_rate": 0.00019488893418024086,
"loss": 0.0001,
"step": 1553
},
{
"epoch": 0.3255984495311927,
"grad_norm": 0.2529331147670746,
"learning_rate": 0.0001948819572234524,
"loss": 0.0189,
"step": 1554
},
{
"epoch": 0.32580797234298886,
"grad_norm": 0.009520837105810642,
"learning_rate": 0.00019487497563295717,
"loss": 0.0001,
"step": 1555
},
{
"epoch": 0.32601749515478495,
"grad_norm": 0.0009123004856519401,
"learning_rate": 0.00019486798940909612,
"loss": 0.0,
"step": 1556
},
{
"epoch": 0.3262270179665811,
"grad_norm": 0.33604323863983154,
"learning_rate": 0.00019486099855221042,
"loss": 0.0302,
"step": 1557
},
{
"epoch": 0.32643654077837725,
"grad_norm": 0.20807093381881714,
"learning_rate": 0.00019485400306264148,
"loss": 0.0235,
"step": 1558
},
{
"epoch": 0.3266460635901734,
"grad_norm": 0.012143682688474655,
"learning_rate": 0.0001948470029407309,
"loss": 0.0001,
"step": 1559
},
{
"epoch": 0.3268555864019695,
"grad_norm": 0.04122312366962433,
"learning_rate": 0.00019483999818682062,
"loss": 0.0006,
"step": 1560
},
{
"epoch": 0.32706510921376564,
"grad_norm": 0.20180264115333557,
"learning_rate": 0.0001948329888012527,
"loss": 0.036,
"step": 1561
},
{
"epoch": 0.3272746320255618,
"grad_norm": 0.17562663555145264,
"learning_rate": 0.00019482597478436944,
"loss": 0.0174,
"step": 1562
},
{
"epoch": 0.32748415483735793,
"grad_norm": 0.01567312516272068,
"learning_rate": 0.00019481895613651337,
"loss": 0.0003,
"step": 1563
},
{
"epoch": 0.327693677649154,
"grad_norm": 0.0051555633544921875,
"learning_rate": 0.00019481193285802726,
"loss": 0.0001,
"step": 1564
},
{
"epoch": 0.3279032004609502,
"grad_norm": 0.052859075367450714,
"learning_rate": 0.00019480490494925412,
"loss": 0.0005,
"step": 1565
},
{
"epoch": 0.3281127232727463,
"grad_norm": 0.17717738449573517,
"learning_rate": 0.00019479787241053717,
"loss": 0.0222,
"step": 1566
},
{
"epoch": 0.32832224608454247,
"grad_norm": 0.17727364599704742,
"learning_rate": 0.0001947908352422198,
"loss": 0.0135,
"step": 1567
},
{
"epoch": 0.3285317688963386,
"grad_norm": 0.07884063571691513,
"learning_rate": 0.00019478379344464578,
"loss": 0.0013,
"step": 1568
},
{
"epoch": 0.3287412917081347,
"grad_norm": 0.056564390659332275,
"learning_rate": 0.00019477674701815897,
"loss": 0.001,
"step": 1569
},
{
"epoch": 0.32895081451993086,
"grad_norm": 0.0061117722652852535,
"learning_rate": 0.00019476969596310344,
"loss": 0.0001,
"step": 1570
},
{
"epoch": 0.329160337331727,
"grad_norm": 0.16239264607429504,
"learning_rate": 0.0001947626402798236,
"loss": 0.0027,
"step": 1571
},
{
"epoch": 0.32936986014352315,
"grad_norm": 0.24103544652462006,
"learning_rate": 0.00019475557996866396,
"loss": 0.0426,
"step": 1572
},
{
"epoch": 0.32957938295531924,
"grad_norm": 0.24487897753715515,
"learning_rate": 0.0001947485150299694,
"loss": 0.0215,
"step": 1573
},
{
"epoch": 0.3297889057671154,
"grad_norm": 0.19202744960784912,
"learning_rate": 0.0001947414454640849,
"loss": 0.0027,
"step": 1574
},
{
"epoch": 0.32999842857891154,
"grad_norm": 0.07547135651111603,
"learning_rate": 0.0001947343712713557,
"loss": 0.0106,
"step": 1575
},
{
"epoch": 0.3302079513907077,
"grad_norm": 0.07378017157316208,
"learning_rate": 0.00019472729245212736,
"loss": 0.0053,
"step": 1576
},
{
"epoch": 0.3304174742025038,
"grad_norm": 0.048381540924310684,
"learning_rate": 0.00019472020900674548,
"loss": 0.0007,
"step": 1577
},
{
"epoch": 0.33062699701429993,
"grad_norm": 0.09767217189073563,
"learning_rate": 0.00019471312093555606,
"loss": 0.0067,
"step": 1578
},
{
"epoch": 0.3308365198260961,
"grad_norm": 0.0035071636084467173,
"learning_rate": 0.00019470602823890522,
"loss": 0.0001,
"step": 1579
},
{
"epoch": 0.3310460426378922,
"grad_norm": 0.17983557283878326,
"learning_rate": 0.00019469893091713937,
"loss": 0.0138,
"step": 1580
},
{
"epoch": 0.3312555654496883,
"grad_norm": 0.20754225552082062,
"learning_rate": 0.0001946918289706051,
"loss": 0.0089,
"step": 1581
},
{
"epoch": 0.33146508826148446,
"grad_norm": 0.09018977731466293,
"learning_rate": 0.00019468472239964925,
"loss": 0.0017,
"step": 1582
},
{
"epoch": 0.3316746110732806,
"grad_norm": 0.003919263835996389,
"learning_rate": 0.0001946776112046189,
"loss": 0.0001,
"step": 1583
},
{
"epoch": 0.33188413388507676,
"grad_norm": 0.009066829457879066,
"learning_rate": 0.0001946704953858613,
"loss": 0.0001,
"step": 1584
},
{
"epoch": 0.33209365669687285,
"grad_norm": 0.05145661532878876,
"learning_rate": 0.00019466337494372397,
"loss": 0.0006,
"step": 1585
},
{
"epoch": 0.332303179508669,
"grad_norm": 0.017742574214935303,
"learning_rate": 0.0001946562498785547,
"loss": 0.0002,
"step": 1586
},
{
"epoch": 0.33251270232046515,
"grad_norm": 0.05322175845503807,
"learning_rate": 0.00019464912019070136,
"loss": 0.0005,
"step": 1587
},
{
"epoch": 0.3327222251322613,
"grad_norm": 0.24456927180290222,
"learning_rate": 0.00019464198588051225,
"loss": 0.0292,
"step": 1588
},
{
"epoch": 0.3329317479440574,
"grad_norm": 0.23175880312919617,
"learning_rate": 0.00019463484694833566,
"loss": 0.0208,
"step": 1589
},
{
"epoch": 0.33314127075585354,
"grad_norm": 0.14457044005393982,
"learning_rate": 0.00019462770339452031,
"loss": 0.004,
"step": 1590
},
{
"epoch": 0.3333507935676497,
"grad_norm": 0.11340788006782532,
"learning_rate": 0.0001946205552194151,
"loss": 0.0142,
"step": 1591
},
{
"epoch": 0.33356031637944583,
"grad_norm": 0.06547866761684418,
"learning_rate": 0.00019461340242336902,
"loss": 0.001,
"step": 1592
},
{
"epoch": 0.3337698391912419,
"grad_norm": 0.3663977384567261,
"learning_rate": 0.00019460624500673146,
"loss": 0.0423,
"step": 1593
},
{
"epoch": 0.33397936200303807,
"grad_norm": 0.0014640424633398652,
"learning_rate": 0.00019459908296985193,
"loss": 0.0,
"step": 1594
},
{
"epoch": 0.3341888848148342,
"grad_norm": 0.014392894692718983,
"learning_rate": 0.00019459191631308027,
"loss": 0.0003,
"step": 1595
},
{
"epoch": 0.33439840762663037,
"grad_norm": 0.10644684731960297,
"learning_rate": 0.0001945847450367664,
"loss": 0.0048,
"step": 1596
},
{
"epoch": 0.33460793043842646,
"grad_norm": 0.028983639553189278,
"learning_rate": 0.00019457756914126052,
"loss": 0.0119,
"step": 1597
},
{
"epoch": 0.3348174532502226,
"grad_norm": 0.027979666367173195,
"learning_rate": 0.00019457038862691315,
"loss": 0.0002,
"step": 1598
},
{
"epoch": 0.33502697606201876,
"grad_norm": 0.21817360818386078,
"learning_rate": 0.00019456320349407492,
"loss": 0.0402,
"step": 1599
},
{
"epoch": 0.3352364988738149,
"grad_norm": 0.0055143400095403194,
"learning_rate": 0.00019455601374309672,
"loss": 0.0001,
"step": 1600
},
{
"epoch": 0.3352364988738149,
"eval_loss": 0.0074059851467609406,
"eval_runtime": 89.0563,
"eval_samples_per_second": 8.624,
"eval_steps_per_second": 4.312,
"step": 1600
}
],
"logging_steps": 1,
"max_steps": 14316,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 200,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 3,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.087044627857408e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}