0505_refseg_finetune / trainer_state.json
insomnia7's picture
Upload folder using huggingface_hub
56049ee verified
Raw
History Blame Contribute Delete
173 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 997,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010030090270812437,
"grad_norm": 59.89604187011719,
"learning_rate": 0.0,
"loss": 6.5574,
"step": 1
},
{
"epoch": 0.0020060180541624875,
"grad_norm": 59.900455474853516,
"learning_rate": 3.3333333333333335e-07,
"loss": 7.0704,
"step": 2
},
{
"epoch": 0.003009027081243731,
"grad_norm": 55.409217834472656,
"learning_rate": 6.666666666666667e-07,
"loss": 6.3999,
"step": 3
},
{
"epoch": 0.004012036108324975,
"grad_norm": 64.29638671875,
"learning_rate": 1.0000000000000002e-06,
"loss": 6.7734,
"step": 4
},
{
"epoch": 0.0050150451354062184,
"grad_norm": 63.90745162963867,
"learning_rate": 1.3333333333333334e-06,
"loss": 6.9351,
"step": 5
},
{
"epoch": 0.006018054162487462,
"grad_norm": 84.24751281738281,
"learning_rate": 1.6666666666666667e-06,
"loss": 6.5649,
"step": 6
},
{
"epoch": 0.007021063189568706,
"grad_norm": 57.16598129272461,
"learning_rate": 2.0000000000000003e-06,
"loss": 6.2622,
"step": 7
},
{
"epoch": 0.00802407221664995,
"grad_norm": 55.82274627685547,
"learning_rate": 2.3333333333333336e-06,
"loss": 6.3501,
"step": 8
},
{
"epoch": 0.009027081243731194,
"grad_norm": 50.48621368408203,
"learning_rate": 2.666666666666667e-06,
"loss": 5.5747,
"step": 9
},
{
"epoch": 0.010030090270812437,
"grad_norm": 48.029598236083984,
"learning_rate": 3e-06,
"loss": 5.1629,
"step": 10
},
{
"epoch": 0.011033099297893681,
"grad_norm": 59.219242095947266,
"learning_rate": 3.3333333333333333e-06,
"loss": 5.2439,
"step": 11
},
{
"epoch": 0.012036108324974924,
"grad_norm": 34.82276153564453,
"learning_rate": 3.6666666666666666e-06,
"loss": 4.5952,
"step": 12
},
{
"epoch": 0.013039117352056168,
"grad_norm": 25.146026611328125,
"learning_rate": 4.000000000000001e-06,
"loss": 4.101,
"step": 13
},
{
"epoch": 0.014042126379137413,
"grad_norm": 22.085208892822266,
"learning_rate": 4.333333333333334e-06,
"loss": 3.7766,
"step": 14
},
{
"epoch": 0.015045135406218655,
"grad_norm": 24.058496475219727,
"learning_rate": 4.666666666666667e-06,
"loss": 3.7508,
"step": 15
},
{
"epoch": 0.0160481444332999,
"grad_norm": 20.09467887878418,
"learning_rate": 5e-06,
"loss": 3.3125,
"step": 16
},
{
"epoch": 0.017051153460381142,
"grad_norm": 19.153560638427734,
"learning_rate": 5.333333333333334e-06,
"loss": 2.989,
"step": 17
},
{
"epoch": 0.01805416248746239,
"grad_norm": 15.94387435913086,
"learning_rate": 5.666666666666667e-06,
"loss": 2.6406,
"step": 18
},
{
"epoch": 0.01905717151454363,
"grad_norm": 17.28965950012207,
"learning_rate": 6e-06,
"loss": 2.4188,
"step": 19
},
{
"epoch": 0.020060180541624874,
"grad_norm": 14.005041122436523,
"learning_rate": 6.333333333333333e-06,
"loss": 2.6113,
"step": 20
},
{
"epoch": 0.02106318956870612,
"grad_norm": 17.092670440673828,
"learning_rate": 6.666666666666667e-06,
"loss": 2.4192,
"step": 21
},
{
"epoch": 0.022066198595787363,
"grad_norm": 10.753110885620117,
"learning_rate": 7e-06,
"loss": 2.3551,
"step": 22
},
{
"epoch": 0.023069207622868605,
"grad_norm": 8.637234687805176,
"learning_rate": 7.333333333333333e-06,
"loss": 2.0748,
"step": 23
},
{
"epoch": 0.024072216649949848,
"grad_norm": 8.181758880615234,
"learning_rate": 7.666666666666667e-06,
"loss": 2.2,
"step": 24
},
{
"epoch": 0.025075225677031094,
"grad_norm": 6.329337120056152,
"learning_rate": 8.000000000000001e-06,
"loss": 1.927,
"step": 25
},
{
"epoch": 0.026078234704112337,
"grad_norm": 7.261748790740967,
"learning_rate": 8.333333333333334e-06,
"loss": 1.734,
"step": 26
},
{
"epoch": 0.02708124373119358,
"grad_norm": 6.460800647735596,
"learning_rate": 8.666666666666668e-06,
"loss": 1.6001,
"step": 27
},
{
"epoch": 0.028084252758274825,
"grad_norm": 5.904067516326904,
"learning_rate": 9e-06,
"loss": 1.7785,
"step": 28
},
{
"epoch": 0.029087261785356068,
"grad_norm": 5.706021308898926,
"learning_rate": 9.333333333333334e-06,
"loss": 1.8143,
"step": 29
},
{
"epoch": 0.03009027081243731,
"grad_norm": 6.162593364715576,
"learning_rate": 9.666666666666667e-06,
"loss": 1.8583,
"step": 30
},
{
"epoch": 0.031093279839518557,
"grad_norm": 5.3871941566467285,
"learning_rate": 1e-05,
"loss": 1.6588,
"step": 31
},
{
"epoch": 0.0320962888665998,
"grad_norm": 5.04168176651001,
"learning_rate": 9.999973613218315e-06,
"loss": 1.6822,
"step": 32
},
{
"epoch": 0.033099297893681046,
"grad_norm": 5.57396936416626,
"learning_rate": 9.99989445315176e-06,
"loss": 1.5525,
"step": 33
},
{
"epoch": 0.034102306920762285,
"grad_norm": 4.78289794921875,
"learning_rate": 9.99976252063585e-06,
"loss": 1.6572,
"step": 34
},
{
"epoch": 0.03510531594784353,
"grad_norm": 4.6783647537231445,
"learning_rate": 9.99957781706309e-06,
"loss": 1.5081,
"step": 35
},
{
"epoch": 0.03610832497492478,
"grad_norm": 4.941147804260254,
"learning_rate": 9.99934034438298e-06,
"loss": 1.4998,
"step": 36
},
{
"epoch": 0.037111334002006016,
"grad_norm": 3.8590803146362305,
"learning_rate": 9.999050105101971e-06,
"loss": 1.6779,
"step": 37
},
{
"epoch": 0.03811434302908726,
"grad_norm": 4.6746697425842285,
"learning_rate": 9.998707102283457e-06,
"loss": 1.485,
"step": 38
},
{
"epoch": 0.03911735205616851,
"grad_norm": 4.716154098510742,
"learning_rate": 9.998311339547735e-06,
"loss": 1.3541,
"step": 39
},
{
"epoch": 0.04012036108324975,
"grad_norm": 4.6193461418151855,
"learning_rate": 9.997862821071966e-06,
"loss": 1.3988,
"step": 40
},
{
"epoch": 0.041123370110330994,
"grad_norm": 6.171404838562012,
"learning_rate": 9.997361551590132e-06,
"loss": 1.6146,
"step": 41
},
{
"epoch": 0.04212637913741224,
"grad_norm": 4.036905765533447,
"learning_rate": 9.99680753639299e-06,
"loss": 1.5718,
"step": 42
},
{
"epoch": 0.04312938816449348,
"grad_norm": 4.592777729034424,
"learning_rate": 9.996200781328011e-06,
"loss": 1.475,
"step": 43
},
{
"epoch": 0.044132397191574725,
"grad_norm": 4.167215347290039,
"learning_rate": 9.995541292799322e-06,
"loss": 1.5253,
"step": 44
},
{
"epoch": 0.04513540621865597,
"grad_norm": 4.920252799987793,
"learning_rate": 9.994829077767632e-06,
"loss": 1.3514,
"step": 45
},
{
"epoch": 0.04613841524573721,
"grad_norm": 4.500365734100342,
"learning_rate": 9.994064143750168e-06,
"loss": 1.498,
"step": 46
},
{
"epoch": 0.04714142427281846,
"grad_norm": 3.473527431488037,
"learning_rate": 9.993246498820588e-06,
"loss": 1.4883,
"step": 47
},
{
"epoch": 0.048144433299899696,
"grad_norm": 4.3483686447143555,
"learning_rate": 9.992376151608898e-06,
"loss": 1.3003,
"step": 48
},
{
"epoch": 0.04914744232698094,
"grad_norm": 3.9893996715545654,
"learning_rate": 9.991453111301367e-06,
"loss": 1.3952,
"step": 49
},
{
"epoch": 0.05015045135406219,
"grad_norm": 3.776866912841797,
"learning_rate": 9.990477387640416e-06,
"loss": 1.3315,
"step": 50
},
{
"epoch": 0.05115346038114343,
"grad_norm": 5.393939018249512,
"learning_rate": 9.989448990924528e-06,
"loss": 1.3376,
"step": 51
},
{
"epoch": 0.05215646940822467,
"grad_norm": 5.019289493560791,
"learning_rate": 9.988367932008138e-06,
"loss": 1.1864,
"step": 52
},
{
"epoch": 0.05315947843530592,
"grad_norm": 4.224878311157227,
"learning_rate": 9.987234222301511e-06,
"loss": 1.3287,
"step": 53
},
{
"epoch": 0.05416248746238716,
"grad_norm": 4.485996246337891,
"learning_rate": 9.986047873770626e-06,
"loss": 1.3246,
"step": 54
},
{
"epoch": 0.055165496489468405,
"grad_norm": 3.9437544345855713,
"learning_rate": 9.984808898937051e-06,
"loss": 1.3481,
"step": 55
},
{
"epoch": 0.05616850551654965,
"grad_norm": 4.048461437225342,
"learning_rate": 9.98351731087781e-06,
"loss": 1.265,
"step": 56
},
{
"epoch": 0.05717151454363089,
"grad_norm": 4.810393333435059,
"learning_rate": 9.982173123225244e-06,
"loss": 1.3715,
"step": 57
},
{
"epoch": 0.058174523570712136,
"grad_norm": 4.251540660858154,
"learning_rate": 9.980776350166868e-06,
"loss": 1.2398,
"step": 58
},
{
"epoch": 0.05917753259779338,
"grad_norm": 4.765919208526611,
"learning_rate": 9.979327006445218e-06,
"loss": 1.3958,
"step": 59
},
{
"epoch": 0.06018054162487462,
"grad_norm": 4.104883670806885,
"learning_rate": 9.977825107357702e-06,
"loss": 1.2801,
"step": 60
},
{
"epoch": 0.06118355065195587,
"grad_norm": 4.080819606781006,
"learning_rate": 9.976270668756435e-06,
"loss": 1.2813,
"step": 61
},
{
"epoch": 0.062186559679037114,
"grad_norm": 4.3112688064575195,
"learning_rate": 9.974663707048068e-06,
"loss": 1.4004,
"step": 62
},
{
"epoch": 0.06318956870611836,
"grad_norm": 5.164820671081543,
"learning_rate": 9.973004239193619e-06,
"loss": 1.257,
"step": 63
},
{
"epoch": 0.0641925777331996,
"grad_norm": 3.600616693496704,
"learning_rate": 9.971292282708295e-06,
"loss": 1.5004,
"step": 64
},
{
"epoch": 0.06519558676028084,
"grad_norm": 3.622041940689087,
"learning_rate": 9.969527855661308e-06,
"loss": 1.3264,
"step": 65
},
{
"epoch": 0.06619859578736209,
"grad_norm": 3.9386863708496094,
"learning_rate": 9.967710976675675e-06,
"loss": 1.4459,
"step": 66
},
{
"epoch": 0.06720160481444333,
"grad_norm": 5.236454486846924,
"learning_rate": 9.965841664928034e-06,
"loss": 1.2096,
"step": 67
},
{
"epoch": 0.06820461384152457,
"grad_norm": 4.4768805503845215,
"learning_rate": 9.96391994014843e-06,
"loss": 1.2606,
"step": 68
},
{
"epoch": 0.06920762286860582,
"grad_norm": 4.6142988204956055,
"learning_rate": 9.96194582262012e-06,
"loss": 1.2989,
"step": 69
},
{
"epoch": 0.07021063189568706,
"grad_norm": 3.9170420169830322,
"learning_rate": 9.959919333179344e-06,
"loss": 1.2367,
"step": 70
},
{
"epoch": 0.0712136409227683,
"grad_norm": 4.681827068328857,
"learning_rate": 9.957840493215116e-06,
"loss": 1.2789,
"step": 71
},
{
"epoch": 0.07221664994984955,
"grad_norm": 3.701741933822632,
"learning_rate": 9.955709324668999e-06,
"loss": 1.4477,
"step": 72
},
{
"epoch": 0.0732196589769308,
"grad_norm": 3.5081491470336914,
"learning_rate": 9.953525850034858e-06,
"loss": 1.2995,
"step": 73
},
{
"epoch": 0.07422266800401203,
"grad_norm": 4.213584899902344,
"learning_rate": 9.951290092358645e-06,
"loss": 1.2631,
"step": 74
},
{
"epoch": 0.07522567703109329,
"grad_norm": 3.3419406414031982,
"learning_rate": 9.949002075238138e-06,
"loss": 1.262,
"step": 75
},
{
"epoch": 0.07622868605817452,
"grad_norm": 4.627947807312012,
"learning_rate": 9.946661822822702e-06,
"loss": 1.1869,
"step": 76
},
{
"epoch": 0.07723169508525576,
"grad_norm": 3.837960958480835,
"learning_rate": 9.944269359813027e-06,
"loss": 1.2523,
"step": 77
},
{
"epoch": 0.07823470411233702,
"grad_norm": 3.5142035484313965,
"learning_rate": 9.941824711460873e-06,
"loss": 1.2541,
"step": 78
},
{
"epoch": 0.07923771313941826,
"grad_norm": 3.524188995361328,
"learning_rate": 9.939327903568802e-06,
"loss": 1.2416,
"step": 79
},
{
"epoch": 0.0802407221664995,
"grad_norm": 3.9529268741607666,
"learning_rate": 9.936778962489902e-06,
"loss": 1.2728,
"step": 80
},
{
"epoch": 0.08124373119358075,
"grad_norm": 4.3076910972595215,
"learning_rate": 9.934177915127516e-06,
"loss": 1.2245,
"step": 81
},
{
"epoch": 0.08224674022066199,
"grad_norm": 3.0643045902252197,
"learning_rate": 9.93152478893495e-06,
"loss": 1.3398,
"step": 82
},
{
"epoch": 0.08324974924774323,
"grad_norm": 5.804976940155029,
"learning_rate": 9.92881961191519e-06,
"loss": 1.3076,
"step": 83
},
{
"epoch": 0.08425275827482448,
"grad_norm": 5.293807029724121,
"learning_rate": 9.9260624126206e-06,
"loss": 1.2459,
"step": 84
},
{
"epoch": 0.08525576730190572,
"grad_norm": 5.097753047943115,
"learning_rate": 9.923253220152628e-06,
"loss": 1.2719,
"step": 85
},
{
"epoch": 0.08625877632898696,
"grad_norm": 5.01963472366333,
"learning_rate": 9.920392064161493e-06,
"loss": 1.1038,
"step": 86
},
{
"epoch": 0.08726178535606821,
"grad_norm": 4.5692877769470215,
"learning_rate": 9.917478974845875e-06,
"loss": 1.1541,
"step": 87
},
{
"epoch": 0.08826479438314945,
"grad_norm": 3.702141284942627,
"learning_rate": 9.914513982952594e-06,
"loss": 1.2372,
"step": 88
},
{
"epoch": 0.08926780341023069,
"grad_norm": 4.985547065734863,
"learning_rate": 9.911497119776286e-06,
"loss": 1.1056,
"step": 89
},
{
"epoch": 0.09027081243731194,
"grad_norm": 3.845374584197998,
"learning_rate": 9.908428417159078e-06,
"loss": 1.3137,
"step": 90
},
{
"epoch": 0.09127382146439318,
"grad_norm": 4.062560558319092,
"learning_rate": 9.905307907490243e-06,
"loss": 1.3115,
"step": 91
},
{
"epoch": 0.09227683049147442,
"grad_norm": 3.845418691635132,
"learning_rate": 9.902135623705864e-06,
"loss": 1.1755,
"step": 92
},
{
"epoch": 0.09327983951855567,
"grad_norm": 4.479650974273682,
"learning_rate": 9.898911599288485e-06,
"loss": 1.2274,
"step": 93
},
{
"epoch": 0.09428284854563691,
"grad_norm": 4.4514479637146,
"learning_rate": 9.895635868266757e-06,
"loss": 1.2045,
"step": 94
},
{
"epoch": 0.09528585757271815,
"grad_norm": 3.8463187217712402,
"learning_rate": 9.89230846521508e-06,
"loss": 1.3114,
"step": 95
},
{
"epoch": 0.09628886659979939,
"grad_norm": 4.823773384094238,
"learning_rate": 9.888929425253237e-06,
"loss": 1.0594,
"step": 96
},
{
"epoch": 0.09729187562688064,
"grad_norm": 5.140885829925537,
"learning_rate": 9.885498784046024e-06,
"loss": 1.2702,
"step": 97
},
{
"epoch": 0.09829488465396188,
"grad_norm": 3.95483660697937,
"learning_rate": 9.882016577802873e-06,
"loss": 1.292,
"step": 98
},
{
"epoch": 0.09929789368104312,
"grad_norm": 4.888804912567139,
"learning_rate": 9.87848284327747e-06,
"loss": 1.1088,
"step": 99
},
{
"epoch": 0.10030090270812438,
"grad_norm": 3.426290273666382,
"learning_rate": 9.874897617767369e-06,
"loss": 1.2179,
"step": 100
},
{
"epoch": 0.10130391173520562,
"grad_norm": 4.530692100524902,
"learning_rate": 9.871260939113596e-06,
"loss": 1.1874,
"step": 101
},
{
"epoch": 0.10230692076228685,
"grad_norm": 4.689167499542236,
"learning_rate": 9.867572845700246e-06,
"loss": 1.2866,
"step": 102
},
{
"epoch": 0.10330992978936811,
"grad_norm": 6.877061367034912,
"learning_rate": 9.863833376454087e-06,
"loss": 1.2948,
"step": 103
},
{
"epoch": 0.10431293881644935,
"grad_norm": 3.0505056381225586,
"learning_rate": 9.860042570844143e-06,
"loss": 1.1882,
"step": 104
},
{
"epoch": 0.10531594784353059,
"grad_norm": 5.552443504333496,
"learning_rate": 9.856200468881277e-06,
"loss": 1.0689,
"step": 105
},
{
"epoch": 0.10631895687061184,
"grad_norm": 3.9618895053863525,
"learning_rate": 9.852307111117771e-06,
"loss": 1.243,
"step": 106
},
{
"epoch": 0.10732196589769308,
"grad_norm": 3.764258861541748,
"learning_rate": 9.848362538646898e-06,
"loss": 1.2543,
"step": 107
},
{
"epoch": 0.10832497492477432,
"grad_norm": 4.093452453613281,
"learning_rate": 9.844366793102489e-06,
"loss": 1.2918,
"step": 108
},
{
"epoch": 0.10932798395185557,
"grad_norm": 4.288122177124023,
"learning_rate": 9.840319916658488e-06,
"loss": 1.2791,
"step": 109
},
{
"epoch": 0.11033099297893681,
"grad_norm": 4.908104419708252,
"learning_rate": 9.836221952028514e-06,
"loss": 1.2009,
"step": 110
},
{
"epoch": 0.11133400200601805,
"grad_norm": 4.961916923522949,
"learning_rate": 9.832072942465404e-06,
"loss": 1.1072,
"step": 111
},
{
"epoch": 0.1123370110330993,
"grad_norm": 4.8414692878723145,
"learning_rate": 9.827872931760765e-06,
"loss": 1.1958,
"step": 112
},
{
"epoch": 0.11334002006018054,
"grad_norm": 4.421910762786865,
"learning_rate": 9.8236219642445e-06,
"loss": 1.1948,
"step": 113
},
{
"epoch": 0.11434302908726178,
"grad_norm": 4.096850395202637,
"learning_rate": 9.819320084784351e-06,
"loss": 1.2064,
"step": 114
},
{
"epoch": 0.11534603811434303,
"grad_norm": 4.6389360427856445,
"learning_rate": 9.814967338785423e-06,
"loss": 1.1042,
"step": 115
},
{
"epoch": 0.11634904714142427,
"grad_norm": 3.4901745319366455,
"learning_rate": 9.810563772189696e-06,
"loss": 1.2238,
"step": 116
},
{
"epoch": 0.11735205616850551,
"grad_norm": 4.026661396026611,
"learning_rate": 9.80610943147555e-06,
"loss": 1.2069,
"step": 117
},
{
"epoch": 0.11835506519558676,
"grad_norm": 4.937434196472168,
"learning_rate": 9.80160436365727e-06,
"loss": 1.1554,
"step": 118
},
{
"epoch": 0.119358074222668,
"grad_norm": 5.201080322265625,
"learning_rate": 9.79704861628456e-06,
"loss": 1.1287,
"step": 119
},
{
"epoch": 0.12036108324974924,
"grad_norm": 4.528622627258301,
"learning_rate": 9.792442237442013e-06,
"loss": 1.159,
"step": 120
},
{
"epoch": 0.1213640922768305,
"grad_norm": 3.706096649169922,
"learning_rate": 9.787785275748644e-06,
"loss": 1.1591,
"step": 121
},
{
"epoch": 0.12236710130391174,
"grad_norm": 4.256808280944824,
"learning_rate": 9.78307778035734e-06,
"loss": 1.1385,
"step": 122
},
{
"epoch": 0.12337011033099297,
"grad_norm": 3.357220411300659,
"learning_rate": 9.778319800954365e-06,
"loss": 1.2346,
"step": 123
},
{
"epoch": 0.12437311935807423,
"grad_norm": 3.560046434402466,
"learning_rate": 9.773511387758823e-06,
"loss": 1.1831,
"step": 124
},
{
"epoch": 0.12537612838515547,
"grad_norm": 5.149918079376221,
"learning_rate": 9.768652591522134e-06,
"loss": 1.2174,
"step": 125
},
{
"epoch": 0.12637913741223672,
"grad_norm": 3.8315470218658447,
"learning_rate": 9.763743463527498e-06,
"loss": 1.1958,
"step": 126
},
{
"epoch": 0.12738214643931794,
"grad_norm": 4.7747063636779785,
"learning_rate": 9.758784055589346e-06,
"loss": 1.1702,
"step": 127
},
{
"epoch": 0.1283851554663992,
"grad_norm": 5.382303237915039,
"learning_rate": 9.753774420052809e-06,
"loss": 1.0555,
"step": 128
},
{
"epoch": 0.12938816449348045,
"grad_norm": 3.541020393371582,
"learning_rate": 9.74871460979315e-06,
"loss": 1.1857,
"step": 129
},
{
"epoch": 0.13039117352056168,
"grad_norm": 4.483475208282471,
"learning_rate": 9.743604678215207e-06,
"loss": 1.1574,
"step": 130
},
{
"epoch": 0.13139418254764293,
"grad_norm": 4.273180961608887,
"learning_rate": 9.738444679252846e-06,
"loss": 1.238,
"step": 131
},
{
"epoch": 0.13239719157472418,
"grad_norm": 4.408812999725342,
"learning_rate": 9.73323466736837e-06,
"loss": 1.1097,
"step": 132
},
{
"epoch": 0.1334002006018054,
"grad_norm": 3.6852378845214844,
"learning_rate": 9.727974697551959e-06,
"loss": 1.2133,
"step": 133
},
{
"epoch": 0.13440320962888666,
"grad_norm": 3.772181987762451,
"learning_rate": 9.722664825321081e-06,
"loss": 1.272,
"step": 134
},
{
"epoch": 0.1354062186559679,
"grad_norm": 4.381350994110107,
"learning_rate": 9.717305106719917e-06,
"loss": 1.1259,
"step": 135
},
{
"epoch": 0.13640922768304914,
"grad_norm": 3.9863905906677246,
"learning_rate": 9.71189559831875e-06,
"loss": 1.0537,
"step": 136
},
{
"epoch": 0.1374122367101304,
"grad_norm": 3.549950361251831,
"learning_rate": 9.706436357213392e-06,
"loss": 1.234,
"step": 137
},
{
"epoch": 0.13841524573721165,
"grad_norm": 2.6753225326538086,
"learning_rate": 9.700927441024561e-06,
"loss": 1.1318,
"step": 138
},
{
"epoch": 0.13941825476429287,
"grad_norm": 3.756247043609619,
"learning_rate": 9.695368907897287e-06,
"loss": 1.2711,
"step": 139
},
{
"epoch": 0.14042126379137412,
"grad_norm": 3.9676246643066406,
"learning_rate": 9.689760816500287e-06,
"loss": 1.1864,
"step": 140
},
{
"epoch": 0.14142427281845538,
"grad_norm": 4.058771133422852,
"learning_rate": 9.684103226025356e-06,
"loss": 1.126,
"step": 141
},
{
"epoch": 0.1424272818455366,
"grad_norm": 3.8615100383758545,
"learning_rate": 9.678396196186738e-06,
"loss": 1.2091,
"step": 142
},
{
"epoch": 0.14343029087261785,
"grad_norm": 4.725363254547119,
"learning_rate": 9.672639787220491e-06,
"loss": 1.092,
"step": 143
},
{
"epoch": 0.1444332998996991,
"grad_norm": 4.6195220947265625,
"learning_rate": 9.666834059883859e-06,
"loss": 1.1948,
"step": 144
},
{
"epoch": 0.14543630892678033,
"grad_norm": 3.5028111934661865,
"learning_rate": 9.660979075454622e-06,
"loss": 1.1475,
"step": 145
},
{
"epoch": 0.1464393179538616,
"grad_norm": 3.256493330001831,
"learning_rate": 9.655074895730463e-06,
"loss": 1.2352,
"step": 146
},
{
"epoch": 0.14744232698094284,
"grad_norm": 4.629305839538574,
"learning_rate": 9.6491215830283e-06,
"loss": 1.0479,
"step": 147
},
{
"epoch": 0.14844533600802406,
"grad_norm": 5.180238723754883,
"learning_rate": 9.64311920018364e-06,
"loss": 1.2001,
"step": 148
},
{
"epoch": 0.14944834503510532,
"grad_norm": 4.484172821044922,
"learning_rate": 9.637067810549908e-06,
"loss": 1.2128,
"step": 149
},
{
"epoch": 0.15045135406218657,
"grad_norm": 3.683945655822754,
"learning_rate": 9.630967477997781e-06,
"loss": 1.2427,
"step": 150
},
{
"epoch": 0.1514543630892678,
"grad_norm": 4.569667816162109,
"learning_rate": 9.624818266914521e-06,
"loss": 1.073,
"step": 151
},
{
"epoch": 0.15245737211634905,
"grad_norm": 4.574268341064453,
"learning_rate": 9.61862024220328e-06,
"loss": 1.1375,
"step": 152
},
{
"epoch": 0.1534603811434303,
"grad_norm": 3.6187503337860107,
"learning_rate": 9.61237346928243e-06,
"loss": 1.1958,
"step": 153
},
{
"epoch": 0.15446339017051153,
"grad_norm": 4.539130687713623,
"learning_rate": 9.606078014084863e-06,
"loss": 1.139,
"step": 154
},
{
"epoch": 0.15546639919759278,
"grad_norm": 3.2087390422821045,
"learning_rate": 9.5997339430573e-06,
"loss": 1.0685,
"step": 155
},
{
"epoch": 0.15646940822467403,
"grad_norm": 4.436512470245361,
"learning_rate": 9.59334132315959e-06,
"loss": 1.0861,
"step": 156
},
{
"epoch": 0.15747241725175526,
"grad_norm": 4.206517696380615,
"learning_rate": 9.586900221863997e-06,
"loss": 1.1604,
"step": 157
},
{
"epoch": 0.1584754262788365,
"grad_norm": 4.249381065368652,
"learning_rate": 9.580410707154495e-06,
"loss": 1.1249,
"step": 158
},
{
"epoch": 0.15947843530591777,
"grad_norm": 3.336700201034546,
"learning_rate": 9.57387284752605e-06,
"loss": 1.2373,
"step": 159
},
{
"epoch": 0.160481444332999,
"grad_norm": 3.6848483085632324,
"learning_rate": 9.567286711983885e-06,
"loss": 1.1696,
"step": 160
},
{
"epoch": 0.16148445336008024,
"grad_norm": 4.119887828826904,
"learning_rate": 9.560652370042774e-06,
"loss": 1.1537,
"step": 161
},
{
"epoch": 0.1624874623871615,
"grad_norm": 5.4756598472595215,
"learning_rate": 9.553969891726289e-06,
"loss": 1.2289,
"step": 162
},
{
"epoch": 0.16349047141424272,
"grad_norm": 3.5389654636383057,
"learning_rate": 9.547239347566069e-06,
"loss": 1.2609,
"step": 163
},
{
"epoch": 0.16449348044132397,
"grad_norm": 3.4587771892547607,
"learning_rate": 9.54046080860107e-06,
"loss": 1.1411,
"step": 164
},
{
"epoch": 0.16549648946840523,
"grad_norm": 6.294703483581543,
"learning_rate": 9.533634346376828e-06,
"loss": 1.2175,
"step": 165
},
{
"epoch": 0.16649949849548645,
"grad_norm": 3.4253592491149902,
"learning_rate": 9.526760032944688e-06,
"loss": 1.1948,
"step": 166
},
{
"epoch": 0.1675025075225677,
"grad_norm": 4.520028114318848,
"learning_rate": 9.519837940861051e-06,
"loss": 1.0899,
"step": 167
},
{
"epoch": 0.16850551654964896,
"grad_norm": 3.001176595687866,
"learning_rate": 9.512868143186616e-06,
"loss": 1.0331,
"step": 168
},
{
"epoch": 0.16950852557673018,
"grad_norm": 2.793478012084961,
"learning_rate": 9.505850713485589e-06,
"loss": 1.1404,
"step": 169
},
{
"epoch": 0.17051153460381144,
"grad_norm": 2.9990012645721436,
"learning_rate": 9.498785725824928e-06,
"loss": 1.1628,
"step": 170
},
{
"epoch": 0.1715145436308927,
"grad_norm": 4.268702507019043,
"learning_rate": 9.491673254773545e-06,
"loss": 1.1353,
"step": 171
},
{
"epoch": 0.17251755265797392,
"grad_norm": 3.4793057441711426,
"learning_rate": 9.484513375401532e-06,
"loss": 1.073,
"step": 172
},
{
"epoch": 0.17352056168505517,
"grad_norm": 4.264081001281738,
"learning_rate": 9.477306163279354e-06,
"loss": 1.1496,
"step": 173
},
{
"epoch": 0.17452357071213642,
"grad_norm": 3.2441599369049072,
"learning_rate": 9.470051694477067e-06,
"loss": 1.2058,
"step": 174
},
{
"epoch": 0.17552657973921765,
"grad_norm": 4.85215425491333,
"learning_rate": 9.462750045563503e-06,
"loss": 1.1191,
"step": 175
},
{
"epoch": 0.1765295887662989,
"grad_norm": 4.243691921234131,
"learning_rate": 9.45540129360547e-06,
"loss": 1.0968,
"step": 176
},
{
"epoch": 0.17753259779338015,
"grad_norm": 3.19276762008667,
"learning_rate": 9.448005516166935e-06,
"loss": 1.1181,
"step": 177
},
{
"epoch": 0.17853560682046138,
"grad_norm": 3.7177531719207764,
"learning_rate": 9.440562791308202e-06,
"loss": 1.1119,
"step": 178
},
{
"epoch": 0.17953861584754263,
"grad_norm": 4.204746246337891,
"learning_rate": 9.43307319758509e-06,
"loss": 1.1201,
"step": 179
},
{
"epoch": 0.18054162487462388,
"grad_norm": 3.252166986465454,
"learning_rate": 9.425536814048114e-06,
"loss": 1.1436,
"step": 180
},
{
"epoch": 0.1815446339017051,
"grad_norm": 4.400845050811768,
"learning_rate": 9.417953720241634e-06,
"loss": 1.1902,
"step": 181
},
{
"epoch": 0.18254764292878636,
"grad_norm": 4.592411041259766,
"learning_rate": 9.410323996203028e-06,
"loss": 0.9777,
"step": 182
},
{
"epoch": 0.18355065195586762,
"grad_norm": 3.59598970413208,
"learning_rate": 9.40264772246184e-06,
"loss": 1.0643,
"step": 183
},
{
"epoch": 0.18455366098294884,
"grad_norm": 3.0653820037841797,
"learning_rate": 9.394924980038934e-06,
"loss": 1.2124,
"step": 184
},
{
"epoch": 0.1855566700100301,
"grad_norm": 4.089035511016846,
"learning_rate": 9.387155850445635e-06,
"loss": 1.1255,
"step": 185
},
{
"epoch": 0.18655967903711135,
"grad_norm": 3.0382211208343506,
"learning_rate": 9.379340415682879e-06,
"loss": 1.1521,
"step": 186
},
{
"epoch": 0.18756268806419257,
"grad_norm": 5.777576446533203,
"learning_rate": 9.371478758240328e-06,
"loss": 0.9333,
"step": 187
},
{
"epoch": 0.18856569709127383,
"grad_norm": 4.489172458648682,
"learning_rate": 9.363570961095525e-06,
"loss": 1.0658,
"step": 188
},
{
"epoch": 0.18956870611835505,
"grad_norm": 4.035121440887451,
"learning_rate": 9.35561710771299e-06,
"loss": 1.1754,
"step": 189
},
{
"epoch": 0.1905717151454363,
"grad_norm": 4.10087251663208,
"learning_rate": 9.347617282043362e-06,
"loss": 1.0799,
"step": 190
},
{
"epoch": 0.19157472417251756,
"grad_norm": 3.615675449371338,
"learning_rate": 9.339571568522504e-06,
"loss": 1.2365,
"step": 191
},
{
"epoch": 0.19257773319959878,
"grad_norm": 4.021636486053467,
"learning_rate": 9.331480052070607e-06,
"loss": 1.1007,
"step": 192
},
{
"epoch": 0.19358074222668004,
"grad_norm": 4.580084800720215,
"learning_rate": 9.323342818091308e-06,
"loss": 1.0488,
"step": 193
},
{
"epoch": 0.1945837512537613,
"grad_norm": 4.915612697601318,
"learning_rate": 9.315159952470766e-06,
"loss": 1.0841,
"step": 194
},
{
"epoch": 0.19558676028084251,
"grad_norm": 3.621744394302368,
"learning_rate": 9.306931541576785e-06,
"loss": 1.1874,
"step": 195
},
{
"epoch": 0.19658976930792377,
"grad_norm": 4.198000431060791,
"learning_rate": 9.298657672257872e-06,
"loss": 1.0966,
"step": 196
},
{
"epoch": 0.19759277833500502,
"grad_norm": 3.9390480518341064,
"learning_rate": 9.29033843184234e-06,
"loss": 1.1392,
"step": 197
},
{
"epoch": 0.19859578736208625,
"grad_norm": 4.119678974151611,
"learning_rate": 9.281973908137385e-06,
"loss": 1.1948,
"step": 198
},
{
"epoch": 0.1995987963891675,
"grad_norm": 2.6715545654296875,
"learning_rate": 9.27356418942815e-06,
"loss": 1.1067,
"step": 199
},
{
"epoch": 0.20060180541624875,
"grad_norm": 3.901723861694336,
"learning_rate": 9.265109364476798e-06,
"loss": 0.9084,
"step": 200
},
{
"epoch": 0.20160481444332998,
"grad_norm": 4.000180244445801,
"learning_rate": 9.25660952252158e-06,
"loss": 1.1422,
"step": 201
},
{
"epoch": 0.20260782347041123,
"grad_norm": 3.146162509918213,
"learning_rate": 9.248064753275882e-06,
"loss": 1.1526,
"step": 202
},
{
"epoch": 0.20361083249749248,
"grad_norm": 4.950468063354492,
"learning_rate": 9.23947514692729e-06,
"loss": 1.0996,
"step": 203
},
{
"epoch": 0.2046138415245737,
"grad_norm": 4.1964616775512695,
"learning_rate": 9.23084079413663e-06,
"loss": 1.0187,
"step": 204
},
{
"epoch": 0.20561685055165496,
"grad_norm": 2.879345178604126,
"learning_rate": 9.222161786037018e-06,
"loss": 1.0951,
"step": 205
},
{
"epoch": 0.20661985957873621,
"grad_norm": 4.137180805206299,
"learning_rate": 9.213438214232889e-06,
"loss": 1.102,
"step": 206
},
{
"epoch": 0.20762286860581744,
"grad_norm": 3.8605384826660156,
"learning_rate": 9.204670170799035e-06,
"loss": 1.1286,
"step": 207
},
{
"epoch": 0.2086258776328987,
"grad_norm": 4.290992259979248,
"learning_rate": 9.195857748279638e-06,
"loss": 1.0697,
"step": 208
},
{
"epoch": 0.20962888665997995,
"grad_norm": 6.566559791564941,
"learning_rate": 9.187001039687285e-06,
"loss": 1.123,
"step": 209
},
{
"epoch": 0.21063189568706117,
"grad_norm": 2.973170518875122,
"learning_rate": 9.178100138501987e-06,
"loss": 1.1779,
"step": 210
},
{
"epoch": 0.21163490471414242,
"grad_norm": 4.0626349449157715,
"learning_rate": 9.169155138670202e-06,
"loss": 1.0474,
"step": 211
},
{
"epoch": 0.21263791374122368,
"grad_norm": 4.198647975921631,
"learning_rate": 9.160166134603835e-06,
"loss": 1.0445,
"step": 212
},
{
"epoch": 0.2136409227683049,
"grad_norm": 3.7403082847595215,
"learning_rate": 9.151133221179237e-06,
"loss": 0.9615,
"step": 213
},
{
"epoch": 0.21464393179538616,
"grad_norm": 3.183431625366211,
"learning_rate": 9.142056493736216e-06,
"loss": 0.9464,
"step": 214
},
{
"epoch": 0.2156469408224674,
"grad_norm": 3.731672763824463,
"learning_rate": 9.132936048077021e-06,
"loss": 1.0952,
"step": 215
},
{
"epoch": 0.21664994984954863,
"grad_norm": 5.230870246887207,
"learning_rate": 9.123771980465337e-06,
"loss": 1.139,
"step": 216
},
{
"epoch": 0.2176529588766299,
"grad_norm": 4.951679229736328,
"learning_rate": 9.114564387625262e-06,
"loss": 1.0951,
"step": 217
},
{
"epoch": 0.21865596790371114,
"grad_norm": 4.416994571685791,
"learning_rate": 9.105313366740295e-06,
"loss": 1.104,
"step": 218
},
{
"epoch": 0.21965897693079237,
"grad_norm": 4.5483622550964355,
"learning_rate": 9.096019015452304e-06,
"loss": 1.0747,
"step": 219
},
{
"epoch": 0.22066198595787362,
"grad_norm": 4.986255645751953,
"learning_rate": 9.086681431860494e-06,
"loss": 1.199,
"step": 220
},
{
"epoch": 0.22166499498495487,
"grad_norm": 4.599176406860352,
"learning_rate": 9.077300714520379e-06,
"loss": 1.031,
"step": 221
},
{
"epoch": 0.2226680040120361,
"grad_norm": 3.861804962158203,
"learning_rate": 9.067876962442732e-06,
"loss": 1.0702,
"step": 222
},
{
"epoch": 0.22367101303911735,
"grad_norm": 6.953867435455322,
"learning_rate": 9.058410275092554e-06,
"loss": 1.0166,
"step": 223
},
{
"epoch": 0.2246740220661986,
"grad_norm": 4.319597244262695,
"learning_rate": 9.048900752388006e-06,
"loss": 1.0857,
"step": 224
},
{
"epoch": 0.22567703109327983,
"grad_norm": 2.7114877700805664,
"learning_rate": 9.039348494699367e-06,
"loss": 1.0882,
"step": 225
},
{
"epoch": 0.22668004012036108,
"grad_norm": 3.8347630500793457,
"learning_rate": 9.029753602847974e-06,
"loss": 1.0609,
"step": 226
},
{
"epoch": 0.22768304914744233,
"grad_norm": 4.422910213470459,
"learning_rate": 9.020116178105154e-06,
"loss": 1.0111,
"step": 227
},
{
"epoch": 0.22868605817452356,
"grad_norm": 2.614464044570923,
"learning_rate": 9.010436322191156e-06,
"loss": 1.1646,
"step": 228
},
{
"epoch": 0.2296890672016048,
"grad_norm": 4.739869594573975,
"learning_rate": 9.000714137274078e-06,
"loss": 1.075,
"step": 229
},
{
"epoch": 0.23069207622868607,
"grad_norm": 3.141099214553833,
"learning_rate": 8.990949725968788e-06,
"loss": 1.0802,
"step": 230
},
{
"epoch": 0.2316950852557673,
"grad_norm": 4.554986000061035,
"learning_rate": 8.98114319133584e-06,
"loss": 0.9099,
"step": 231
},
{
"epoch": 0.23269809428284854,
"grad_norm": 4.7924933433532715,
"learning_rate": 8.971294636880392e-06,
"loss": 1.119,
"step": 232
},
{
"epoch": 0.2337011033099298,
"grad_norm": 3.946301221847534,
"learning_rate": 8.961404166551104e-06,
"loss": 1.0026,
"step": 233
},
{
"epoch": 0.23470411233701102,
"grad_norm": 3.5471274852752686,
"learning_rate": 8.951471884739052e-06,
"loss": 1.045,
"step": 234
},
{
"epoch": 0.23570712136409228,
"grad_norm": 4.546420097351074,
"learning_rate": 8.941497896276615e-06,
"loss": 1.0892,
"step": 235
},
{
"epoch": 0.23671013039117353,
"grad_norm": 3.528237819671631,
"learning_rate": 8.931482306436374e-06,
"loss": 1.036,
"step": 236
},
{
"epoch": 0.23771313941825475,
"grad_norm": 3.4267489910125732,
"learning_rate": 8.921425220930002e-06,
"loss": 1.032,
"step": 237
},
{
"epoch": 0.238716148445336,
"grad_norm": 4.36963415145874,
"learning_rate": 8.911326745907152e-06,
"loss": 1.0906,
"step": 238
},
{
"epoch": 0.23971915747241726,
"grad_norm": 3.686178207397461,
"learning_rate": 8.90118698795432e-06,
"loss": 1.1173,
"step": 239
},
{
"epoch": 0.24072216649949849,
"grad_norm": 2.928985834121704,
"learning_rate": 8.89100605409374e-06,
"loss": 1.0762,
"step": 240
},
{
"epoch": 0.24172517552657974,
"grad_norm": 4.694427013397217,
"learning_rate": 8.880784051782244e-06,
"loss": 1.0551,
"step": 241
},
{
"epoch": 0.242728184553661,
"grad_norm": 2.930530071258545,
"learning_rate": 8.87052108891013e-06,
"loss": 1.1197,
"step": 242
},
{
"epoch": 0.24373119358074222,
"grad_norm": 3.972811222076416,
"learning_rate": 8.86021727380002e-06,
"loss": 0.9655,
"step": 243
},
{
"epoch": 0.24473420260782347,
"grad_norm": 3.6450603008270264,
"learning_rate": 8.849872715205726e-06,
"loss": 1.0934,
"step": 244
},
{
"epoch": 0.24573721163490472,
"grad_norm": 3.210270881652832,
"learning_rate": 8.839487522311088e-06,
"loss": 1.1259,
"step": 245
},
{
"epoch": 0.24674022066198595,
"grad_norm": 4.2320990562438965,
"learning_rate": 8.829061804728835e-06,
"loss": 1.0088,
"step": 246
},
{
"epoch": 0.2477432296890672,
"grad_norm": 3.296393394470215,
"learning_rate": 8.81859567249942e-06,
"loss": 1.0348,
"step": 247
},
{
"epoch": 0.24874623871614845,
"grad_norm": 4.077010631561279,
"learning_rate": 8.808089236089859e-06,
"loss": 1.0902,
"step": 248
},
{
"epoch": 0.24974924774322968,
"grad_norm": 2.6496405601501465,
"learning_rate": 8.797542606392572e-06,
"loss": 0.9961,
"step": 249
},
{
"epoch": 0.25075225677031093,
"grad_norm": 3.7086265087127686,
"learning_rate": 8.786955894724207e-06,
"loss": 1.0769,
"step": 250
},
{
"epoch": 0.25175526579739216,
"grad_norm": 4.005990505218506,
"learning_rate": 8.776329212824462e-06,
"loss": 1.104,
"step": 251
},
{
"epoch": 0.25275827482447344,
"grad_norm": 3.4448018074035645,
"learning_rate": 8.765662672854909e-06,
"loss": 1.0084,
"step": 252
},
{
"epoch": 0.25376128385155466,
"grad_norm": 3.686387062072754,
"learning_rate": 8.754956387397814e-06,
"loss": 1.1048,
"step": 253
},
{
"epoch": 0.2547642928786359,
"grad_norm": 2.537383794784546,
"learning_rate": 8.744210469454947e-06,
"loss": 1.0322,
"step": 254
},
{
"epoch": 0.25576730190571717,
"grad_norm": 3.093122720718384,
"learning_rate": 8.733425032446382e-06,
"loss": 1.1029,
"step": 255
},
{
"epoch": 0.2567703109327984,
"grad_norm": 4.30277156829834,
"learning_rate": 8.722600190209304e-06,
"loss": 0.9876,
"step": 256
},
{
"epoch": 0.2577733199598796,
"grad_norm": 3.175452470779419,
"learning_rate": 8.711736056996818e-06,
"loss": 0.9799,
"step": 257
},
{
"epoch": 0.2587763289869609,
"grad_norm": 4.154230117797852,
"learning_rate": 8.700832747476725e-06,
"loss": 0.8639,
"step": 258
},
{
"epoch": 0.2597793380140421,
"grad_norm": 4.742602825164795,
"learning_rate": 8.689890376730328e-06,
"loss": 1.0602,
"step": 259
},
{
"epoch": 0.26078234704112335,
"grad_norm": 3.2317545413970947,
"learning_rate": 8.678909060251202e-06,
"loss": 1.1419,
"step": 260
},
{
"epoch": 0.26178535606820463,
"grad_norm": 3.7621326446533203,
"learning_rate": 8.667888913943989e-06,
"loss": 1.0314,
"step": 261
},
{
"epoch": 0.26278836509528586,
"grad_norm": 5.490180015563965,
"learning_rate": 8.65683005412317e-06,
"loss": 1.0427,
"step": 262
},
{
"epoch": 0.2637913741223671,
"grad_norm": 2.488373041152954,
"learning_rate": 8.645732597511827e-06,
"loss": 1.0587,
"step": 263
},
{
"epoch": 0.26479438314944836,
"grad_norm": 4.172122001647949,
"learning_rate": 8.634596661240429e-06,
"loss": 1.0129,
"step": 264
},
{
"epoch": 0.2657973921765296,
"grad_norm": 3.6035194396972656,
"learning_rate": 8.623422362845582e-06,
"loss": 0.9723,
"step": 265
},
{
"epoch": 0.2668004012036108,
"grad_norm": 3.12439227104187,
"learning_rate": 8.6122098202688e-06,
"loss": 1.1359,
"step": 266
},
{
"epoch": 0.2678034102306921,
"grad_norm": 4.614977836608887,
"learning_rate": 8.60095915185524e-06,
"loss": 0.95,
"step": 267
},
{
"epoch": 0.2688064192577733,
"grad_norm": 4.308961868286133,
"learning_rate": 8.589670476352484e-06,
"loss": 0.9132,
"step": 268
},
{
"epoch": 0.26980942828485455,
"grad_norm": 4.492208003997803,
"learning_rate": 8.578343912909252e-06,
"loss": 0.9235,
"step": 269
},
{
"epoch": 0.2708124373119358,
"grad_norm": 3.372297763824463,
"learning_rate": 8.566979581074169e-06,
"loss": 1.1106,
"step": 270
},
{
"epoch": 0.27181544633901705,
"grad_norm": 5.646235466003418,
"learning_rate": 8.55557760079449e-06,
"loss": 0.913,
"step": 271
},
{
"epoch": 0.2728184553660983,
"grad_norm": 4.03857421875,
"learning_rate": 8.544138092414843e-06,
"loss": 1.0796,
"step": 272
},
{
"epoch": 0.27382146439317956,
"grad_norm": 4.554337978363037,
"learning_rate": 8.532661176675952e-06,
"loss": 1.0698,
"step": 273
},
{
"epoch": 0.2748244734202608,
"grad_norm": 3.750927686691284,
"learning_rate": 8.521146974713363e-06,
"loss": 1.1378,
"step": 274
},
{
"epoch": 0.275827482447342,
"grad_norm": 7.945387840270996,
"learning_rate": 8.509595608056171e-06,
"loss": 0.9929,
"step": 275
},
{
"epoch": 0.2768304914744233,
"grad_norm": 4.650017738342285,
"learning_rate": 8.498007198625733e-06,
"loss": 1.0443,
"step": 276
},
{
"epoch": 0.2778335005015045,
"grad_norm": 2.8968331813812256,
"learning_rate": 8.486381868734379e-06,
"loss": 1.0171,
"step": 277
},
{
"epoch": 0.27883650952858574,
"grad_norm": 4.479628086090088,
"learning_rate": 8.474719741084127e-06,
"loss": 1.0674,
"step": 278
},
{
"epoch": 0.279839518555667,
"grad_norm": 3.172144651412964,
"learning_rate": 8.463020938765385e-06,
"loss": 1.0884,
"step": 279
},
{
"epoch": 0.28084252758274825,
"grad_norm": 3.7971673011779785,
"learning_rate": 8.451285585255647e-06,
"loss": 0.975,
"step": 280
},
{
"epoch": 0.2818455366098295,
"grad_norm": 3.386007308959961,
"learning_rate": 8.439513804418196e-06,
"loss": 1.0113,
"step": 281
},
{
"epoch": 0.28284854563691075,
"grad_norm": 3.7247607707977295,
"learning_rate": 8.4277057205008e-06,
"loss": 1.0052,
"step": 282
},
{
"epoch": 0.283851554663992,
"grad_norm": 4.630113124847412,
"learning_rate": 8.415861458134394e-06,
"loss": 1.1261,
"step": 283
},
{
"epoch": 0.2848545636910732,
"grad_norm": 3.96730375289917,
"learning_rate": 8.403981142331758e-06,
"loss": 1.0327,
"step": 284
},
{
"epoch": 0.2858575727181545,
"grad_norm": 4.412420749664307,
"learning_rate": 8.392064898486216e-06,
"loss": 1.0673,
"step": 285
},
{
"epoch": 0.2868605817452357,
"grad_norm": 4.324704647064209,
"learning_rate": 8.380112852370297e-06,
"loss": 1.0105,
"step": 286
},
{
"epoch": 0.28786359077231694,
"grad_norm": 3.071364402770996,
"learning_rate": 8.368125130134414e-06,
"loss": 1.0496,
"step": 287
},
{
"epoch": 0.2888665997993982,
"grad_norm": 4.933933734893799,
"learning_rate": 8.35610185830553e-06,
"loss": 1.0681,
"step": 288
},
{
"epoch": 0.28986960882647944,
"grad_norm": 3.603890895843506,
"learning_rate": 8.344043163785824e-06,
"loss": 0.9834,
"step": 289
},
{
"epoch": 0.29087261785356067,
"grad_norm": 4.808909893035889,
"learning_rate": 8.331949173851356e-06,
"loss": 0.9344,
"step": 290
},
{
"epoch": 0.29187562688064195,
"grad_norm": 3.681326150894165,
"learning_rate": 8.319820016150707e-06,
"loss": 1.0474,
"step": 291
},
{
"epoch": 0.2928786359077232,
"grad_norm": 2.417637586593628,
"learning_rate": 8.307655818703658e-06,
"loss": 1.1327,
"step": 292
},
{
"epoch": 0.2938816449348044,
"grad_norm": 4.81406307220459,
"learning_rate": 8.295456709899818e-06,
"loss": 1.0092,
"step": 293
},
{
"epoch": 0.2948846539618857,
"grad_norm": 3.891993284225464,
"learning_rate": 8.28322281849727e-06,
"loss": 1.0242,
"step": 294
},
{
"epoch": 0.2958876629889669,
"grad_norm": 4.634517669677734,
"learning_rate": 8.270954273621229e-06,
"loss": 0.9521,
"step": 295
},
{
"epoch": 0.29689067201604813,
"grad_norm": 3.0400383472442627,
"learning_rate": 8.258651204762658e-06,
"loss": 1.0752,
"step": 296
},
{
"epoch": 0.2978936810431294,
"grad_norm": 4.023530006408691,
"learning_rate": 8.246313741776911e-06,
"loss": 0.9428,
"step": 297
},
{
"epoch": 0.29889669007021064,
"grad_norm": 3.9196455478668213,
"learning_rate": 8.23394201488237e-06,
"loss": 1.1396,
"step": 298
},
{
"epoch": 0.29989969909729186,
"grad_norm": 5.009520053863525,
"learning_rate": 8.221536154659056e-06,
"loss": 1.0324,
"step": 299
},
{
"epoch": 0.30090270812437314,
"grad_norm": 3.679565906524658,
"learning_rate": 8.209096292047258e-06,
"loss": 1.0606,
"step": 300
},
{
"epoch": 0.30190571715145437,
"grad_norm": 3.043320655822754,
"learning_rate": 8.196622558346152e-06,
"loss": 1.0344,
"step": 301
},
{
"epoch": 0.3029087261785356,
"grad_norm": 3.6096246242523193,
"learning_rate": 8.184115085212414e-06,
"loss": 0.9679,
"step": 302
},
{
"epoch": 0.3039117352056169,
"grad_norm": 4.0385284423828125,
"learning_rate": 8.171574004658829e-06,
"loss": 1.0448,
"step": 303
},
{
"epoch": 0.3049147442326981,
"grad_norm": 4.102316856384277,
"learning_rate": 8.158999449052899e-06,
"loss": 1.0945,
"step": 304
},
{
"epoch": 0.3059177532597793,
"grad_norm": 3.1016600131988525,
"learning_rate": 8.146391551115444e-06,
"loss": 0.9971,
"step": 305
},
{
"epoch": 0.3069207622868606,
"grad_norm": 4.629376411437988,
"learning_rate": 8.133750443919207e-06,
"loss": 1.0537,
"step": 306
},
{
"epoch": 0.30792377131394183,
"grad_norm": 4.228681564331055,
"learning_rate": 8.121076260887437e-06,
"loss": 1.0373,
"step": 307
},
{
"epoch": 0.30892678034102306,
"grad_norm": 3.6265642642974854,
"learning_rate": 8.1083691357925e-06,
"loss": 1.134,
"step": 308
},
{
"epoch": 0.30992978936810434,
"grad_norm": 4.494879245758057,
"learning_rate": 8.095629202754448e-06,
"loss": 0.9954,
"step": 309
},
{
"epoch": 0.31093279839518556,
"grad_norm": 3.3157126903533936,
"learning_rate": 8.082856596239614e-06,
"loss": 1.029,
"step": 310
},
{
"epoch": 0.3119358074222668,
"grad_norm": 4.181100368499756,
"learning_rate": 8.070051451059188e-06,
"loss": 0.9642,
"step": 311
},
{
"epoch": 0.31293881644934807,
"grad_norm": 3.6057820320129395,
"learning_rate": 8.057213902367802e-06,
"loss": 0.9791,
"step": 312
},
{
"epoch": 0.3139418254764293,
"grad_norm": 3.7801015377044678,
"learning_rate": 8.044344085662093e-06,
"loss": 1.0596,
"step": 313
},
{
"epoch": 0.3149448345035105,
"grad_norm": 4.987183094024658,
"learning_rate": 8.031442136779272e-06,
"loss": 1.122,
"step": 314
},
{
"epoch": 0.3159478435305918,
"grad_norm": 3.5851123332977295,
"learning_rate": 8.018508191895714e-06,
"loss": 1.0318,
"step": 315
},
{
"epoch": 0.316950852557673,
"grad_norm": 2.953979969024658,
"learning_rate": 8.00554238752548e-06,
"loss": 1.0467,
"step": 316
},
{
"epoch": 0.31795386158475425,
"grad_norm": 5.37313175201416,
"learning_rate": 7.992544860518917e-06,
"loss": 0.912,
"step": 317
},
{
"epoch": 0.31895687061183553,
"grad_norm": 3.2147107124328613,
"learning_rate": 7.979515748061182e-06,
"loss": 1.0973,
"step": 318
},
{
"epoch": 0.31995987963891676,
"grad_norm": 4.388527870178223,
"learning_rate": 7.96645518767082e-06,
"loss": 1.0496,
"step": 319
},
{
"epoch": 0.320962888665998,
"grad_norm": 4.035599231719971,
"learning_rate": 7.953363317198288e-06,
"loss": 1.0008,
"step": 320
},
{
"epoch": 0.32196589769307926,
"grad_norm": 4.014002323150635,
"learning_rate": 7.94024027482452e-06,
"loss": 1.0156,
"step": 321
},
{
"epoch": 0.3229689067201605,
"grad_norm": 4.31812047958374,
"learning_rate": 7.927086199059457e-06,
"loss": 0.9369,
"step": 322
},
{
"epoch": 0.3239719157472417,
"grad_norm": 4.76963472366333,
"learning_rate": 7.91390122874059e-06,
"loss": 0.8067,
"step": 323
},
{
"epoch": 0.324974924774323,
"grad_norm": 6.049552917480469,
"learning_rate": 7.900685503031491e-06,
"loss": 1.0889,
"step": 324
},
{
"epoch": 0.3259779338014042,
"grad_norm": 4.714635848999023,
"learning_rate": 7.887439161420348e-06,
"loss": 0.9993,
"step": 325
},
{
"epoch": 0.32698094282848544,
"grad_norm": 4.099154472351074,
"learning_rate": 7.87416234371849e-06,
"loss": 1.1027,
"step": 326
},
{
"epoch": 0.3279839518555667,
"grad_norm": 3.5714809894561768,
"learning_rate": 7.860855190058914e-06,
"loss": 0.9963,
"step": 327
},
{
"epoch": 0.32898696088264795,
"grad_norm": 3.500962734222412,
"learning_rate": 7.847517840894805e-06,
"loss": 0.8997,
"step": 328
},
{
"epoch": 0.3299899699097292,
"grad_norm": 3.778474807739258,
"learning_rate": 7.834150436998047e-06,
"loss": 0.9888,
"step": 329
},
{
"epoch": 0.33099297893681046,
"grad_norm": 5.302026748657227,
"learning_rate": 7.82075311945775e-06,
"loss": 1.006,
"step": 330
},
{
"epoch": 0.3319959879638917,
"grad_norm": 3.919344902038574,
"learning_rate": 7.807326029678754e-06,
"loss": 1.0011,
"step": 331
},
{
"epoch": 0.3329989969909729,
"grad_norm": 3.4691314697265625,
"learning_rate": 7.79386930938013e-06,
"loss": 1.0495,
"step": 332
},
{
"epoch": 0.3340020060180542,
"grad_norm": 3.729933977127075,
"learning_rate": 7.780383100593693e-06,
"loss": 1.0036,
"step": 333
},
{
"epoch": 0.3350050150451354,
"grad_norm": 4.451638221740723,
"learning_rate": 7.766867545662507e-06,
"loss": 0.9716,
"step": 334
},
{
"epoch": 0.33600802407221664,
"grad_norm": 2.9692766666412354,
"learning_rate": 7.753322787239366e-06,
"loss": 0.9777,
"step": 335
},
{
"epoch": 0.3370110330992979,
"grad_norm": 4.90089225769043,
"learning_rate": 7.739748968285305e-06,
"loss": 0.9698,
"step": 336
},
{
"epoch": 0.33801404212637914,
"grad_norm": 4.4351301193237305,
"learning_rate": 7.726146232068084e-06,
"loss": 1.0134,
"step": 337
},
{
"epoch": 0.33901705115346037,
"grad_norm": 4.581515789031982,
"learning_rate": 7.712514722160673e-06,
"loss": 1.1292,
"step": 338
},
{
"epoch": 0.34002006018054165,
"grad_norm": 3.3677995204925537,
"learning_rate": 7.698854582439746e-06,
"loss": 1.0064,
"step": 339
},
{
"epoch": 0.3410230692076229,
"grad_norm": 4.035177707672119,
"learning_rate": 7.685165957084148e-06,
"loss": 0.9143,
"step": 340
},
{
"epoch": 0.3420260782347041,
"grad_norm": 4.151326656341553,
"learning_rate": 7.671448990573392e-06,
"loss": 1.0979,
"step": 341
},
{
"epoch": 0.3430290872617854,
"grad_norm": 3.808765411376953,
"learning_rate": 7.657703827686116e-06,
"loss": 0.9445,
"step": 342
},
{
"epoch": 0.3440320962888666,
"grad_norm": 5.776411533355713,
"learning_rate": 7.643930613498563e-06,
"loss": 0.9331,
"step": 343
},
{
"epoch": 0.34503510531594783,
"grad_norm": 4.4571428298950195,
"learning_rate": 7.630129493383052e-06,
"loss": 1.0198,
"step": 344
},
{
"epoch": 0.3460381143430291,
"grad_norm": 4.128350257873535,
"learning_rate": 7.616300613006443e-06,
"loss": 0.9384,
"step": 345
},
{
"epoch": 0.34704112337011034,
"grad_norm": 3.3957364559173584,
"learning_rate": 7.602444118328593e-06,
"loss": 0.947,
"step": 346
},
{
"epoch": 0.34804413239719156,
"grad_norm": 3.9797606468200684,
"learning_rate": 7.588560155600823e-06,
"loss": 0.957,
"step": 347
},
{
"epoch": 0.34904714142427284,
"grad_norm": 4.085581302642822,
"learning_rate": 7.574648871364369e-06,
"loss": 0.9014,
"step": 348
},
{
"epoch": 0.35005015045135407,
"grad_norm": 4.2193379402160645,
"learning_rate": 7.56071041244884e-06,
"loss": 0.9814,
"step": 349
},
{
"epoch": 0.3510531594784353,
"grad_norm": 3.9035120010375977,
"learning_rate": 7.5467449259706645e-06,
"loss": 1.023,
"step": 350
},
{
"epoch": 0.3520561685055166,
"grad_norm": 3.4941298961639404,
"learning_rate": 7.53275255933154e-06,
"loss": 0.9744,
"step": 351
},
{
"epoch": 0.3530591775325978,
"grad_norm": 3.543471097946167,
"learning_rate": 7.518733460216876e-06,
"loss": 0.9373,
"step": 352
},
{
"epoch": 0.354062186559679,
"grad_norm": 4.236612319946289,
"learning_rate": 7.504687776594235e-06,
"loss": 0.9922,
"step": 353
},
{
"epoch": 0.3550651955867603,
"grad_norm": 5.047423362731934,
"learning_rate": 7.490615656711773e-06,
"loss": 1.0178,
"step": 354
},
{
"epoch": 0.35606820461384153,
"grad_norm": 3.6122939586639404,
"learning_rate": 7.476517249096671e-06,
"loss": 0.9952,
"step": 355
},
{
"epoch": 0.35707121364092276,
"grad_norm": 4.43414306640625,
"learning_rate": 7.4623927025535715e-06,
"loss": 0.939,
"step": 356
},
{
"epoch": 0.35807422266800404,
"grad_norm": 3.8483691215515137,
"learning_rate": 7.448242166163005e-06,
"loss": 0.828,
"step": 357
},
{
"epoch": 0.35907723169508526,
"grad_norm": 3.749648332595825,
"learning_rate": 7.4340657892798165e-06,
"loss": 0.9452,
"step": 358
},
{
"epoch": 0.3600802407221665,
"grad_norm": 3.7156455516815186,
"learning_rate": 7.41986372153159e-06,
"loss": 0.8323,
"step": 359
},
{
"epoch": 0.36108324974924777,
"grad_norm": 4.360509872436523,
"learning_rate": 7.405636112817073e-06,
"loss": 0.924,
"step": 360
},
{
"epoch": 0.362086258776329,
"grad_norm": 5.340662479400635,
"learning_rate": 7.391383113304583e-06,
"loss": 0.9794,
"step": 361
},
{
"epoch": 0.3630892678034102,
"grad_norm": 4.195555210113525,
"learning_rate": 7.377104873430439e-06,
"loss": 1.0092,
"step": 362
},
{
"epoch": 0.3640922768304915,
"grad_norm": 2.8496367931365967,
"learning_rate": 7.362801543897358e-06,
"loss": 0.9695,
"step": 363
},
{
"epoch": 0.3650952858575727,
"grad_norm": 4.262087821960449,
"learning_rate": 7.348473275672873e-06,
"loss": 0.9992,
"step": 364
},
{
"epoch": 0.36609829488465395,
"grad_norm": 5.186934471130371,
"learning_rate": 7.334120219987742e-06,
"loss": 1.0151,
"step": 365
},
{
"epoch": 0.36710130391173523,
"grad_norm": 3.989398241043091,
"learning_rate": 7.31974252833434e-06,
"loss": 1.004,
"step": 366
},
{
"epoch": 0.36810431293881646,
"grad_norm": 4.705235004425049,
"learning_rate": 7.305340352465071e-06,
"loss": 0.9415,
"step": 367
},
{
"epoch": 0.3691073219658977,
"grad_norm": 5.499981880187988,
"learning_rate": 7.290913844390766e-06,
"loss": 0.9193,
"step": 368
},
{
"epoch": 0.37011033099297896,
"grad_norm": 3.84904146194458,
"learning_rate": 7.2764631563790696e-06,
"loss": 1.045,
"step": 369
},
{
"epoch": 0.3711133400200602,
"grad_norm": 4.123418807983398,
"learning_rate": 7.261988440952845e-06,
"loss": 0.9799,
"step": 370
},
{
"epoch": 0.3721163490471414,
"grad_norm": 3.4615368843078613,
"learning_rate": 7.247489850888552e-06,
"loss": 1.0958,
"step": 371
},
{
"epoch": 0.3731193580742227,
"grad_norm": 4.5892510414123535,
"learning_rate": 7.232967539214643e-06,
"loss": 0.9843,
"step": 372
},
{
"epoch": 0.3741223671013039,
"grad_norm": 3.608722686767578,
"learning_rate": 7.218421659209949e-06,
"loss": 1.0282,
"step": 373
},
{
"epoch": 0.37512537612838515,
"grad_norm": 3.9482369422912598,
"learning_rate": 7.203852364402049e-06,
"loss": 1.035,
"step": 374
},
{
"epoch": 0.37612838515546637,
"grad_norm": 4.018014907836914,
"learning_rate": 7.189259808565665e-06,
"loss": 0.9008,
"step": 375
},
{
"epoch": 0.37713139418254765,
"grad_norm": 3.3777265548706055,
"learning_rate": 7.174644145721031e-06,
"loss": 0.9633,
"step": 376
},
{
"epoch": 0.3781344032096289,
"grad_norm": 3.4682204723358154,
"learning_rate": 7.1600055301322705e-06,
"loss": 1.0273,
"step": 377
},
{
"epoch": 0.3791374122367101,
"grad_norm": 4.235317707061768,
"learning_rate": 7.145344116305763e-06,
"loss": 0.9975,
"step": 378
},
{
"epoch": 0.3801404212637914,
"grad_norm": 3.3626129627227783,
"learning_rate": 7.130660058988521e-06,
"loss": 0.9486,
"step": 379
},
{
"epoch": 0.3811434302908726,
"grad_norm": 4.25431489944458,
"learning_rate": 7.115953513166549e-06,
"loss": 1.0036,
"step": 380
},
{
"epoch": 0.38214643931795383,
"grad_norm": 3.6873717308044434,
"learning_rate": 7.101224634063213e-06,
"loss": 1.0069,
"step": 381
},
{
"epoch": 0.3831494483450351,
"grad_norm": 4.648606777191162,
"learning_rate": 7.086473577137598e-06,
"loss": 0.8974,
"step": 382
},
{
"epoch": 0.38415245737211634,
"grad_norm": 2.658944845199585,
"learning_rate": 7.071700498082874e-06,
"loss": 0.9258,
"step": 383
},
{
"epoch": 0.38515546639919757,
"grad_norm": 4.260404586791992,
"learning_rate": 7.056905552824646e-06,
"loss": 0.861,
"step": 384
},
{
"epoch": 0.38615847542627885,
"grad_norm": 3.9974656105041504,
"learning_rate": 7.042088897519307e-06,
"loss": 0.9252,
"step": 385
},
{
"epoch": 0.38716148445336007,
"grad_norm": 4.013027191162109,
"learning_rate": 7.0272506885524e-06,
"loss": 0.9941,
"step": 386
},
{
"epoch": 0.3881644934804413,
"grad_norm": 3.6313085556030273,
"learning_rate": 7.012391082536956e-06,
"loss": 1.1019,
"step": 387
},
{
"epoch": 0.3891675025075226,
"grad_norm": 3.9201042652130127,
"learning_rate": 6.997510236311846e-06,
"loss": 0.8283,
"step": 388
},
{
"epoch": 0.3901705115346038,
"grad_norm": 3.4184234142303467,
"learning_rate": 6.982608306940129e-06,
"loss": 0.9634,
"step": 389
},
{
"epoch": 0.39117352056168503,
"grad_norm": 2.9435737133026123,
"learning_rate": 6.967685451707384e-06,
"loss": 0.9725,
"step": 390
},
{
"epoch": 0.3921765295887663,
"grad_norm": 3.0873332023620605,
"learning_rate": 6.952741828120063e-06,
"loss": 0.9439,
"step": 391
},
{
"epoch": 0.39317953861584753,
"grad_norm": 3.547354221343994,
"learning_rate": 6.937777593903818e-06,
"loss": 0.9484,
"step": 392
},
{
"epoch": 0.39418254764292876,
"grad_norm": 3.1705925464630127,
"learning_rate": 6.922792907001843e-06,
"loss": 1.0267,
"step": 393
},
{
"epoch": 0.39518555667001004,
"grad_norm": 3.89007568359375,
"learning_rate": 6.9077879255732e-06,
"loss": 0.8934,
"step": 394
},
{
"epoch": 0.39618856569709127,
"grad_norm": 4.767911434173584,
"learning_rate": 6.8927628079911604e-06,
"loss": 1.0084,
"step": 395
},
{
"epoch": 0.3971915747241725,
"grad_norm": 3.4053468704223633,
"learning_rate": 6.8777177128415206e-06,
"loss": 1.0329,
"step": 396
},
{
"epoch": 0.3981945837512538,
"grad_norm": 4.725776672363281,
"learning_rate": 6.862652798920939e-06,
"loss": 0.9935,
"step": 397
},
{
"epoch": 0.399197592778335,
"grad_norm": 4.242211818695068,
"learning_rate": 6.84756822523525e-06,
"loss": 0.9998,
"step": 398
},
{
"epoch": 0.4002006018054162,
"grad_norm": 2.7795650959014893,
"learning_rate": 6.8324641509978e-06,
"loss": 0.9704,
"step": 399
},
{
"epoch": 0.4012036108324975,
"grad_norm": 4.177364826202393,
"learning_rate": 6.817340735627746e-06,
"loss": 0.9587,
"step": 400
},
{
"epoch": 0.40220661985957873,
"grad_norm": 3.8208518028259277,
"learning_rate": 6.802198138748398e-06,
"loss": 0.8334,
"step": 401
},
{
"epoch": 0.40320962888665995,
"grad_norm": 4.306139945983887,
"learning_rate": 6.787036520185511e-06,
"loss": 0.9828,
"step": 402
},
{
"epoch": 0.40421263791374124,
"grad_norm": 4.775547027587891,
"learning_rate": 6.7718560399656165e-06,
"loss": 0.9807,
"step": 403
},
{
"epoch": 0.40521564694082246,
"grad_norm": 5.744744777679443,
"learning_rate": 6.756656858314319e-06,
"loss": 0.987,
"step": 404
},
{
"epoch": 0.4062186559679037,
"grad_norm": 2.946840286254883,
"learning_rate": 6.741439135654613e-06,
"loss": 0.9597,
"step": 405
},
{
"epoch": 0.40722166499498497,
"grad_norm": 3.476158618927002,
"learning_rate": 6.726203032605191e-06,
"loss": 0.9739,
"step": 406
},
{
"epoch": 0.4082246740220662,
"grad_norm": 3.910567045211792,
"learning_rate": 6.7109487099787415e-06,
"loss": 0.9612,
"step": 407
},
{
"epoch": 0.4092276830491474,
"grad_norm": 4.358532428741455,
"learning_rate": 6.695676328780256e-06,
"loss": 0.888,
"step": 408
},
{
"epoch": 0.4102306920762287,
"grad_norm": 3.15893292427063,
"learning_rate": 6.680386050205333e-06,
"loss": 0.9255,
"step": 409
},
{
"epoch": 0.4112337011033099,
"grad_norm": 3.3531334400177,
"learning_rate": 6.665078035638466e-06,
"loss": 0.8755,
"step": 410
},
{
"epoch": 0.41223671013039115,
"grad_norm": 5.0254950523376465,
"learning_rate": 6.649752446651353e-06,
"loss": 1.1064,
"step": 411
},
{
"epoch": 0.41323971915747243,
"grad_norm": 4.246181488037109,
"learning_rate": 6.6344094450011806e-06,
"loss": 0.9382,
"step": 412
},
{
"epoch": 0.41424272818455365,
"grad_norm": 5.631673336029053,
"learning_rate": 6.619049192628925e-06,
"loss": 0.9564,
"step": 413
},
{
"epoch": 0.4152457372116349,
"grad_norm": 4.842679023742676,
"learning_rate": 6.603671851657635e-06,
"loss": 0.832,
"step": 414
},
{
"epoch": 0.41624874623871616,
"grad_norm": 4.033610820770264,
"learning_rate": 6.588277584390725e-06,
"loss": 0.8331,
"step": 415
},
{
"epoch": 0.4172517552657974,
"grad_norm": 3.797995090484619,
"learning_rate": 6.572866553310267e-06,
"loss": 0.9334,
"step": 416
},
{
"epoch": 0.4182547642928786,
"grad_norm": 4.544320106506348,
"learning_rate": 6.557438921075259e-06,
"loss": 0.8599,
"step": 417
},
{
"epoch": 0.4192577733199599,
"grad_norm": 4.663610935211182,
"learning_rate": 6.541994850519934e-06,
"loss": 0.9168,
"step": 418
},
{
"epoch": 0.4202607823470411,
"grad_norm": 3.8332691192626953,
"learning_rate": 6.5265345046520145e-06,
"loss": 0.8969,
"step": 419
},
{
"epoch": 0.42126379137412234,
"grad_norm": 3.1332781314849854,
"learning_rate": 6.511058046651011e-06,
"loss": 1.0156,
"step": 420
},
{
"epoch": 0.4222668004012036,
"grad_norm": 2.8914341926574707,
"learning_rate": 6.4955656398664915e-06,
"loss": 0.9074,
"step": 421
},
{
"epoch": 0.42326980942828485,
"grad_norm": 3.263746738433838,
"learning_rate": 6.4800574478163554e-06,
"loss": 1.0018,
"step": 422
},
{
"epoch": 0.4242728184553661,
"grad_norm": 3.428309440612793,
"learning_rate": 6.464533634185117e-06,
"loss": 0.9724,
"step": 423
},
{
"epoch": 0.42527582748244735,
"grad_norm": 4.138224124908447,
"learning_rate": 6.4489943628221675e-06,
"loss": 0.9949,
"step": 424
},
{
"epoch": 0.4262788365095286,
"grad_norm": 5.183621883392334,
"learning_rate": 6.43343979774005e-06,
"loss": 0.9026,
"step": 425
},
{
"epoch": 0.4272818455366098,
"grad_norm": 3.35832142829895,
"learning_rate": 6.417870103112732e-06,
"loss": 0.8482,
"step": 426
},
{
"epoch": 0.4282848545636911,
"grad_norm": 4.023225784301758,
"learning_rate": 6.4022854432738656e-06,
"loss": 0.7483,
"step": 427
},
{
"epoch": 0.4292878635907723,
"grad_norm": 4.992588520050049,
"learning_rate": 6.386685982715057e-06,
"loss": 0.9827,
"step": 428
},
{
"epoch": 0.43029087261785354,
"grad_norm": 4.83533239364624,
"learning_rate": 6.371071886084132e-06,
"loss": 0.9184,
"step": 429
},
{
"epoch": 0.4312938816449348,
"grad_norm": 3.8867318630218506,
"learning_rate": 6.3554433181833945e-06,
"loss": 0.8487,
"step": 430
},
{
"epoch": 0.43229689067201604,
"grad_norm": 4.48527717590332,
"learning_rate": 6.339800443967884e-06,
"loss": 0.928,
"step": 431
},
{
"epoch": 0.43329989969909727,
"grad_norm": 3.2990341186523438,
"learning_rate": 6.3241434285436475e-06,
"loss": 0.8856,
"step": 432
},
{
"epoch": 0.43430290872617855,
"grad_norm": 3.5551490783691406,
"learning_rate": 6.3084724371659826e-06,
"loss": 0.9216,
"step": 433
},
{
"epoch": 0.4353059177532598,
"grad_norm": 4.301858901977539,
"learning_rate": 6.292787635237701e-06,
"loss": 0.8705,
"step": 434
},
{
"epoch": 0.436308926780341,
"grad_norm": 4.350210189819336,
"learning_rate": 6.2770891883073795e-06,
"loss": 0.8991,
"step": 435
},
{
"epoch": 0.4373119358074223,
"grad_norm": 3.3270864486694336,
"learning_rate": 6.261377262067615e-06,
"loss": 0.9284,
"step": 436
},
{
"epoch": 0.4383149448345035,
"grad_norm": 4.551054000854492,
"learning_rate": 6.245652022353277e-06,
"loss": 0.9681,
"step": 437
},
{
"epoch": 0.43931795386158473,
"grad_norm": 3.319453716278076,
"learning_rate": 6.229913635139749e-06,
"loss": 0.8287,
"step": 438
},
{
"epoch": 0.440320962888666,
"grad_norm": 4.894309997558594,
"learning_rate": 6.214162266541188e-06,
"loss": 0.9911,
"step": 439
},
{
"epoch": 0.44132397191574724,
"grad_norm": 4.263638496398926,
"learning_rate": 6.198398082808764e-06,
"loss": 0.9962,
"step": 440
},
{
"epoch": 0.44232698094282846,
"grad_norm": 5.57863712310791,
"learning_rate": 6.182621250328906e-06,
"loss": 0.8979,
"step": 441
},
{
"epoch": 0.44332998996990974,
"grad_norm": 4.514066696166992,
"learning_rate": 6.166831935621546e-06,
"loss": 0.9628,
"step": 442
},
{
"epoch": 0.44433299899699097,
"grad_norm": 4.59286642074585,
"learning_rate": 6.151030305338368e-06,
"loss": 0.9762,
"step": 443
},
{
"epoch": 0.4453360080240722,
"grad_norm": 3.5197913646698,
"learning_rate": 6.135216526261036e-06,
"loss": 0.9293,
"step": 444
},
{
"epoch": 0.4463390170511535,
"grad_norm": 3.1230807304382324,
"learning_rate": 6.119390765299448e-06,
"loss": 0.8991,
"step": 445
},
{
"epoch": 0.4473420260782347,
"grad_norm": 4.823525428771973,
"learning_rate": 6.10355318948996e-06,
"loss": 0.9769,
"step": 446
},
{
"epoch": 0.4483450351053159,
"grad_norm": 5.179238796234131,
"learning_rate": 6.087703965993637e-06,
"loss": 0.9412,
"step": 447
},
{
"epoch": 0.4493480441323972,
"grad_norm": 4.475254535675049,
"learning_rate": 6.071843262094477e-06,
"loss": 0.9225,
"step": 448
},
{
"epoch": 0.45035105315947843,
"grad_norm": 4.747657299041748,
"learning_rate": 6.0559712451976525e-06,
"loss": 0.8836,
"step": 449
},
{
"epoch": 0.45135406218655966,
"grad_norm": 3.4025394916534424,
"learning_rate": 6.040088082827744e-06,
"loss": 0.9982,
"step": 450
},
{
"epoch": 0.45235707121364094,
"grad_norm": 4.142173767089844,
"learning_rate": 6.024193942626962e-06,
"loss": 0.9264,
"step": 451
},
{
"epoch": 0.45336008024072216,
"grad_norm": 4.095114231109619,
"learning_rate": 6.008288992353396e-06,
"loss": 0.7056,
"step": 452
},
{
"epoch": 0.4543630892678034,
"grad_norm": 4.608190536499023,
"learning_rate": 5.992373399879221e-06,
"loss": 0.9666,
"step": 453
},
{
"epoch": 0.45536609829488467,
"grad_norm": 3.399383306503296,
"learning_rate": 5.976447333188945e-06,
"loss": 1.0767,
"step": 454
},
{
"epoch": 0.4563691073219659,
"grad_norm": 4.751845359802246,
"learning_rate": 5.960510960377627e-06,
"loss": 0.9283,
"step": 455
},
{
"epoch": 0.4573721163490471,
"grad_norm": 4.31138801574707,
"learning_rate": 5.9445644496491e-06,
"loss": 0.9523,
"step": 456
},
{
"epoch": 0.4583751253761284,
"grad_norm": 3.711291551589966,
"learning_rate": 5.928607969314202e-06,
"loss": 0.9982,
"step": 457
},
{
"epoch": 0.4593781344032096,
"grad_norm": 4.085208892822266,
"learning_rate": 5.912641687789003e-06,
"loss": 0.9742,
"step": 458
},
{
"epoch": 0.46038114343029085,
"grad_norm": 5.210893630981445,
"learning_rate": 5.896665773593013e-06,
"loss": 0.9096,
"step": 459
},
{
"epoch": 0.46138415245737213,
"grad_norm": 3.2804415225982666,
"learning_rate": 5.880680395347419e-06,
"loss": 0.9013,
"step": 460
},
{
"epoch": 0.46238716148445336,
"grad_norm": 4.84675407409668,
"learning_rate": 5.864685721773293e-06,
"loss": 0.8392,
"step": 461
},
{
"epoch": 0.4633901705115346,
"grad_norm": 2.984220266342163,
"learning_rate": 5.848681921689821e-06,
"loss": 0.8971,
"step": 462
},
{
"epoch": 0.46439317953861586,
"grad_norm": 3.941147565841675,
"learning_rate": 5.832669164012514e-06,
"loss": 0.7912,
"step": 463
},
{
"epoch": 0.4653961885656971,
"grad_norm": 3.821648359298706,
"learning_rate": 5.8166476177514256e-06,
"loss": 0.9779,
"step": 464
},
{
"epoch": 0.4663991975927783,
"grad_norm": 4.071436405181885,
"learning_rate": 5.800617452009376e-06,
"loss": 0.8164,
"step": 465
},
{
"epoch": 0.4674022066198596,
"grad_norm": 4.478545665740967,
"learning_rate": 5.784578835980158e-06,
"loss": 1.0268,
"step": 466
},
{
"epoch": 0.4684052156469408,
"grad_norm": 4.578827381134033,
"learning_rate": 5.768531938946756e-06,
"loss": 1.0158,
"step": 467
},
{
"epoch": 0.46940822467402205,
"grad_norm": 4.314323425292969,
"learning_rate": 5.752476930279558e-06,
"loss": 0.8999,
"step": 468
},
{
"epoch": 0.4704112337011033,
"grad_norm": 3.880741596221924,
"learning_rate": 5.736413979434567e-06,
"loss": 0.9864,
"step": 469
},
{
"epoch": 0.47141424272818455,
"grad_norm": 3.10372257232666,
"learning_rate": 5.720343255951612e-06,
"loss": 0.8713,
"step": 470
},
{
"epoch": 0.4724172517552658,
"grad_norm": 4.399007797241211,
"learning_rate": 5.704264929452563e-06,
"loss": 0.884,
"step": 471
},
{
"epoch": 0.47342026078234706,
"grad_norm": 5.845457077026367,
"learning_rate": 5.688179169639538e-06,
"loss": 0.9062,
"step": 472
},
{
"epoch": 0.4744232698094283,
"grad_norm": 3.394307851791382,
"learning_rate": 5.672086146293109e-06,
"loss": 0.8997,
"step": 473
},
{
"epoch": 0.4754262788365095,
"grad_norm": 5.672798156738281,
"learning_rate": 5.655986029270512e-06,
"loss": 0.8948,
"step": 474
},
{
"epoch": 0.4764292878635908,
"grad_norm": 3.985313892364502,
"learning_rate": 5.639878988503859e-06,
"loss": 1.0173,
"step": 475
},
{
"epoch": 0.477432296890672,
"grad_norm": 4.498374938964844,
"learning_rate": 5.623765193998334e-06,
"loss": 0.8865,
"step": 476
},
{
"epoch": 0.47843530591775324,
"grad_norm": 4.075979232788086,
"learning_rate": 5.607644815830412e-06,
"loss": 0.8853,
"step": 477
},
{
"epoch": 0.4794383149448345,
"grad_norm": 3.184473991394043,
"learning_rate": 5.5915180241460496e-06,
"loss": 0.8411,
"step": 478
},
{
"epoch": 0.48044132397191575,
"grad_norm": 5.224008083343506,
"learning_rate": 5.5753849891589005e-06,
"loss": 1.027,
"step": 479
},
{
"epoch": 0.48144433299899697,
"grad_norm": 3.7843403816223145,
"learning_rate": 5.559245881148514e-06,
"loss": 0.9375,
"step": 480
},
{
"epoch": 0.48244734202607825,
"grad_norm": 5.324986457824707,
"learning_rate": 5.543100870458537e-06,
"loss": 0.9576,
"step": 481
},
{
"epoch": 0.4834503510531595,
"grad_norm": 4.161240100860596,
"learning_rate": 5.526950127494919e-06,
"loss": 0.9537,
"step": 482
},
{
"epoch": 0.4844533600802407,
"grad_norm": 4.03006649017334,
"learning_rate": 5.510793822724111e-06,
"loss": 0.8328,
"step": 483
},
{
"epoch": 0.485456369107322,
"grad_norm": 2.5849175453186035,
"learning_rate": 5.494632126671269e-06,
"loss": 0.9159,
"step": 484
},
{
"epoch": 0.4864593781344032,
"grad_norm": 3.324787139892578,
"learning_rate": 5.47846520991845e-06,
"loss": 0.9209,
"step": 485
},
{
"epoch": 0.48746238716148443,
"grad_norm": 4.150218486785889,
"learning_rate": 5.462293243102817e-06,
"loss": 0.8648,
"step": 486
},
{
"epoch": 0.4884653961885657,
"grad_norm": 4.268919944763184,
"learning_rate": 5.44611639691483e-06,
"loss": 0.858,
"step": 487
},
{
"epoch": 0.48946840521564694,
"grad_norm": 4.742982864379883,
"learning_rate": 5.429934842096454e-06,
"loss": 0.9289,
"step": 488
},
{
"epoch": 0.49047141424272817,
"grad_norm": 4.175072193145752,
"learning_rate": 5.413748749439352e-06,
"loss": 0.9581,
"step": 489
},
{
"epoch": 0.49147442326980945,
"grad_norm": 4.109638214111328,
"learning_rate": 5.397558289783079e-06,
"loss": 0.9035,
"step": 490
},
{
"epoch": 0.49247743229689067,
"grad_norm": 4.354793071746826,
"learning_rate": 5.3813636340132856e-06,
"loss": 0.9675,
"step": 491
},
{
"epoch": 0.4934804413239719,
"grad_norm": 4.927343368530273,
"learning_rate": 5.365164953059912e-06,
"loss": 0.8699,
"step": 492
},
{
"epoch": 0.4944834503510532,
"grad_norm": 3.839871883392334,
"learning_rate": 5.348962417895379e-06,
"loss": 0.8888,
"step": 493
},
{
"epoch": 0.4954864593781344,
"grad_norm": 4.1549201011657715,
"learning_rate": 5.332756199532791e-06,
"loss": 0.8023,
"step": 494
},
{
"epoch": 0.49648946840521563,
"grad_norm": 4.788138389587402,
"learning_rate": 5.316546469024128e-06,
"loss": 0.9228,
"step": 495
},
{
"epoch": 0.4974924774322969,
"grad_norm": 4.4567975997924805,
"learning_rate": 5.300333397458436e-06,
"loss": 0.9113,
"step": 496
},
{
"epoch": 0.49849548645937813,
"grad_norm": 5.03196382522583,
"learning_rate": 5.284117155960026e-06,
"loss": 0.9923,
"step": 497
},
{
"epoch": 0.49949849548645936,
"grad_norm": 5.067019462585449,
"learning_rate": 5.267897915686669e-06,
"loss": 0.9563,
"step": 498
},
{
"epoch": 0.5005015045135406,
"grad_norm": 4.418056011199951,
"learning_rate": 5.251675847827785e-06,
"loss": 0.9137,
"step": 499
},
{
"epoch": 0.5015045135406219,
"grad_norm": 3.900459051132202,
"learning_rate": 5.235451123602642e-06,
"loss": 1.0351,
"step": 500
},
{
"epoch": 0.5025075225677031,
"grad_norm": 3.014345169067383,
"learning_rate": 5.219223914258539e-06,
"loss": 0.8897,
"step": 501
},
{
"epoch": 0.5035105315947843,
"grad_norm": 4.5792975425720215,
"learning_rate": 5.202994391069009e-06,
"loss": 0.8987,
"step": 502
},
{
"epoch": 0.5045135406218656,
"grad_norm": 4.500517845153809,
"learning_rate": 5.186762725332008e-06,
"loss": 0.909,
"step": 503
},
{
"epoch": 0.5055165496489469,
"grad_norm": 3.911684274673462,
"learning_rate": 5.170529088368104e-06,
"loss": 0.9866,
"step": 504
},
{
"epoch": 0.506519558676028,
"grad_norm": 4.596591949462891,
"learning_rate": 5.154293651518667e-06,
"loss": 0.9133,
"step": 505
},
{
"epoch": 0.5075225677031093,
"grad_norm": 4.134907245635986,
"learning_rate": 5.1380565861440725e-06,
"loss": 0.9349,
"step": 506
},
{
"epoch": 0.5085255767301906,
"grad_norm": 4.103421211242676,
"learning_rate": 5.121818063621877e-06,
"loss": 0.9844,
"step": 507
},
{
"epoch": 0.5095285857572718,
"grad_norm": 3.4258902072906494,
"learning_rate": 5.105578255345021e-06,
"loss": 0.9471,
"step": 508
},
{
"epoch": 0.5105315947843531,
"grad_norm": 4.5036163330078125,
"learning_rate": 5.089337332720017e-06,
"loss": 0.9218,
"step": 509
},
{
"epoch": 0.5115346038114343,
"grad_norm": 4.1147661209106445,
"learning_rate": 5.0730954671651345e-06,
"loss": 0.9016,
"step": 510
},
{
"epoch": 0.5125376128385155,
"grad_norm": 2.9174067974090576,
"learning_rate": 5.056852830108598e-06,
"loss": 0.823,
"step": 511
},
{
"epoch": 0.5135406218655968,
"grad_norm": 3.397568464279175,
"learning_rate": 5.040609592986776e-06,
"loss": 0.9438,
"step": 512
},
{
"epoch": 0.5145436308926781,
"grad_norm": 5.088010787963867,
"learning_rate": 5.024365927242367e-06,
"loss": 0.7205,
"step": 513
},
{
"epoch": 0.5155466399197592,
"grad_norm": 4.969124794006348,
"learning_rate": 5.008122004322598e-06,
"loss": 0.8343,
"step": 514
},
{
"epoch": 0.5165496489468405,
"grad_norm": 5.184592247009277,
"learning_rate": 4.991877995677405e-06,
"loss": 0.8023,
"step": 515
},
{
"epoch": 0.5175526579739218,
"grad_norm": 3.7894270420074463,
"learning_rate": 4.975634072757635e-06,
"loss": 0.8951,
"step": 516
},
{
"epoch": 0.518555667001003,
"grad_norm": 4.564579963684082,
"learning_rate": 4.959390407013226e-06,
"loss": 1.0316,
"step": 517
},
{
"epoch": 0.5195586760280843,
"grad_norm": 3.4184181690216064,
"learning_rate": 4.943147169891403e-06,
"loss": 0.9163,
"step": 518
},
{
"epoch": 0.5205616850551655,
"grad_norm": 5.040964126586914,
"learning_rate": 4.926904532834867e-06,
"loss": 0.8271,
"step": 519
},
{
"epoch": 0.5215646940822467,
"grad_norm": 4.553344249725342,
"learning_rate": 4.910662667279984e-06,
"loss": 0.9705,
"step": 520
},
{
"epoch": 0.522567703109328,
"grad_norm": 3.099721908569336,
"learning_rate": 4.8944217446549794e-06,
"loss": 0.956,
"step": 521
},
{
"epoch": 0.5235707121364093,
"grad_norm": 5.947968482971191,
"learning_rate": 4.878181936378125e-06,
"loss": 0.8834,
"step": 522
},
{
"epoch": 0.5245737211634904,
"grad_norm": 3.4968111515045166,
"learning_rate": 4.861943413855929e-06,
"loss": 0.9785,
"step": 523
},
{
"epoch": 0.5255767301905717,
"grad_norm": 5.2941389083862305,
"learning_rate": 4.845706348481334e-06,
"loss": 0.8478,
"step": 524
},
{
"epoch": 0.526579739217653,
"grad_norm": 3.7787928581237793,
"learning_rate": 4.8294709116318986e-06,
"loss": 0.9362,
"step": 525
},
{
"epoch": 0.5275827482447342,
"grad_norm": 3.890321969985962,
"learning_rate": 4.8132372746679925e-06,
"loss": 0.8353,
"step": 526
},
{
"epoch": 0.5285857572718154,
"grad_norm": 4.203886985778809,
"learning_rate": 4.797005608930992e-06,
"loss": 0.8221,
"step": 527
},
{
"epoch": 0.5295887662988967,
"grad_norm": 4.37759256362915,
"learning_rate": 4.7807760857414626e-06,
"loss": 0.9212,
"step": 528
},
{
"epoch": 0.5305917753259779,
"grad_norm": 3.307236671447754,
"learning_rate": 4.76454887639736e-06,
"loss": 0.8242,
"step": 529
},
{
"epoch": 0.5315947843530592,
"grad_norm": 4.246848106384277,
"learning_rate": 4.7483241521722165e-06,
"loss": 0.8465,
"step": 530
},
{
"epoch": 0.5325977933801405,
"grad_norm": 4.23176383972168,
"learning_rate": 4.732102084313333e-06,
"loss": 0.7919,
"step": 531
},
{
"epoch": 0.5336008024072216,
"grad_norm": 4.603852272033691,
"learning_rate": 4.715882844039975e-06,
"loss": 0.9187,
"step": 532
},
{
"epoch": 0.5346038114343029,
"grad_norm": 3.9758505821228027,
"learning_rate": 4.699666602541566e-06,
"loss": 0.6982,
"step": 533
},
{
"epoch": 0.5356068204613842,
"grad_norm": 3.7466206550598145,
"learning_rate": 4.6834535309758725e-06,
"loss": 0.8021,
"step": 534
},
{
"epoch": 0.5366098294884654,
"grad_norm": 4.444018840789795,
"learning_rate": 4.667243800467209e-06,
"loss": 1.0597,
"step": 535
},
{
"epoch": 0.5376128385155466,
"grad_norm": 3.1866021156311035,
"learning_rate": 4.6510375821046215e-06,
"loss": 0.7176,
"step": 536
},
{
"epoch": 0.5386158475426279,
"grad_norm": 5.325089931488037,
"learning_rate": 4.634835046940089e-06,
"loss": 0.8456,
"step": 537
},
{
"epoch": 0.5396188565697091,
"grad_norm": 3.6649844646453857,
"learning_rate": 4.618636365986714e-06,
"loss": 0.8887,
"step": 538
},
{
"epoch": 0.5406218655967904,
"grad_norm": 3.809011220932007,
"learning_rate": 4.602441710216922e-06,
"loss": 0.8757,
"step": 539
},
{
"epoch": 0.5416248746238717,
"grad_norm": 4.5428948402404785,
"learning_rate": 4.586251250560649e-06,
"loss": 0.8886,
"step": 540
},
{
"epoch": 0.5426278836509528,
"grad_norm": 3.85677170753479,
"learning_rate": 4.570065157903546e-06,
"loss": 0.862,
"step": 541
},
{
"epoch": 0.5436308926780341,
"grad_norm": 2.898193836212158,
"learning_rate": 4.5538836030851705e-06,
"loss": 0.8122,
"step": 542
},
{
"epoch": 0.5446339017051154,
"grad_norm": 4.943717002868652,
"learning_rate": 4.537706756897184e-06,
"loss": 0.9281,
"step": 543
},
{
"epoch": 0.5456369107321966,
"grad_norm": 4.602163791656494,
"learning_rate": 4.52153479008155e-06,
"loss": 0.8995,
"step": 544
},
{
"epoch": 0.5466399197592778,
"grad_norm": 3.8922247886657715,
"learning_rate": 4.505367873328732e-06,
"loss": 0.8791,
"step": 545
},
{
"epoch": 0.5476429287863591,
"grad_norm": 5.127221584320068,
"learning_rate": 4.48920617727589e-06,
"loss": 0.7215,
"step": 546
},
{
"epoch": 0.5486459378134403,
"grad_norm": 3.9078500270843506,
"learning_rate": 4.473049872505082e-06,
"loss": 0.8896,
"step": 547
},
{
"epoch": 0.5496489468405216,
"grad_norm": 4.704128742218018,
"learning_rate": 4.456899129541464e-06,
"loss": 1.0129,
"step": 548
},
{
"epoch": 0.5506519558676028,
"grad_norm": 3.71657657623291,
"learning_rate": 4.440754118851486e-06,
"loss": 0.8578,
"step": 549
},
{
"epoch": 0.551654964894684,
"grad_norm": 3.771890878677368,
"learning_rate": 4.4246150108410995e-06,
"loss": 0.8269,
"step": 550
},
{
"epoch": 0.5526579739217653,
"grad_norm": 4.481631278991699,
"learning_rate": 4.408481975853951e-06,
"loss": 0.7112,
"step": 551
},
{
"epoch": 0.5536609829488466,
"grad_norm": 4.309824466705322,
"learning_rate": 4.392355184169588e-06,
"loss": 0.8766,
"step": 552
},
{
"epoch": 0.5546639919759278,
"grad_norm": 4.343716144561768,
"learning_rate": 4.376234806001666e-06,
"loss": 0.8649,
"step": 553
},
{
"epoch": 0.555667001003009,
"grad_norm": 3.9397213459014893,
"learning_rate": 4.360121011496142e-06,
"loss": 0.9902,
"step": 554
},
{
"epoch": 0.5566700100300903,
"grad_norm": 3.5650391578674316,
"learning_rate": 4.34401397072949e-06,
"loss": 0.9286,
"step": 555
},
{
"epoch": 0.5576730190571715,
"grad_norm": 4.326221466064453,
"learning_rate": 4.327913853706894e-06,
"loss": 0.737,
"step": 556
},
{
"epoch": 0.5586760280842528,
"grad_norm": 3.1165454387664795,
"learning_rate": 4.3118208303604645e-06,
"loss": 0.8537,
"step": 557
},
{
"epoch": 0.559679037111334,
"grad_norm": 3.2974557876586914,
"learning_rate": 4.295735070547439e-06,
"loss": 0.7883,
"step": 558
},
{
"epoch": 0.5606820461384152,
"grad_norm": 4.610488414764404,
"learning_rate": 4.279656744048391e-06,
"loss": 0.8679,
"step": 559
},
{
"epoch": 0.5616850551654965,
"grad_norm": 4.936162948608398,
"learning_rate": 4.263586020565437e-06,
"loss": 0.964,
"step": 560
},
{
"epoch": 0.5626880641925778,
"grad_norm": 5.205172538757324,
"learning_rate": 4.247523069720445e-06,
"loss": 0.8294,
"step": 561
},
{
"epoch": 0.563691073219659,
"grad_norm": 3.9387030601501465,
"learning_rate": 4.231468061053245e-06,
"loss": 0.9558,
"step": 562
},
{
"epoch": 0.5646940822467402,
"grad_norm": 4.719061374664307,
"learning_rate": 4.215421164019843e-06,
"loss": 0.7299,
"step": 563
},
{
"epoch": 0.5656970912738215,
"grad_norm": 4.1843767166137695,
"learning_rate": 4.199382547990626e-06,
"loss": 0.8695,
"step": 564
},
{
"epoch": 0.5667001003009027,
"grad_norm": 5.512943267822266,
"learning_rate": 4.183352382248577e-06,
"loss": 1.0263,
"step": 565
},
{
"epoch": 0.567703109327984,
"grad_norm": 3.520388126373291,
"learning_rate": 4.16733083598749e-06,
"loss": 0.9238,
"step": 566
},
{
"epoch": 0.5687061183550652,
"grad_norm": 4.626349449157715,
"learning_rate": 4.151318078310181e-06,
"loss": 0.8939,
"step": 567
},
{
"epoch": 0.5697091273821464,
"grad_norm": 5.5067362785339355,
"learning_rate": 4.135314278226709e-06,
"loss": 1.0198,
"step": 568
},
{
"epoch": 0.5707121364092277,
"grad_norm": 4.588747978210449,
"learning_rate": 4.119319604652584e-06,
"loss": 1.0281,
"step": 569
},
{
"epoch": 0.571715145436309,
"grad_norm": 4.2089104652404785,
"learning_rate": 4.10333422640699e-06,
"loss": 0.7827,
"step": 570
},
{
"epoch": 0.5727181544633901,
"grad_norm": 4.466251850128174,
"learning_rate": 4.087358312210999e-06,
"loss": 0.9088,
"step": 571
},
{
"epoch": 0.5737211634904714,
"grad_norm": 3.7433619499206543,
"learning_rate": 4.071392030685799e-06,
"loss": 0.8721,
"step": 572
},
{
"epoch": 0.5747241725175527,
"grad_norm": 3.044492483139038,
"learning_rate": 4.0554355503509035e-06,
"loss": 0.8949,
"step": 573
},
{
"epoch": 0.5757271815446339,
"grad_norm": 4.252336502075195,
"learning_rate": 4.039489039622376e-06,
"loss": 0.7801,
"step": 574
},
{
"epoch": 0.5767301905717152,
"grad_norm": 4.381480693817139,
"learning_rate": 4.023552666811056e-06,
"loss": 0.7232,
"step": 575
},
{
"epoch": 0.5777331995987964,
"grad_norm": 3.7939393520355225,
"learning_rate": 4.0076266001207804e-06,
"loss": 0.9669,
"step": 576
},
{
"epoch": 0.5787362086258776,
"grad_norm": 3.9904580116271973,
"learning_rate": 3.9917110076466055e-06,
"loss": 0.8936,
"step": 577
},
{
"epoch": 0.5797392176529589,
"grad_norm": 4.3604350090026855,
"learning_rate": 3.975806057373039e-06,
"loss": 0.7821,
"step": 578
},
{
"epoch": 0.5807422266800402,
"grad_norm": 3.206235885620117,
"learning_rate": 3.959911917172258e-06,
"loss": 0.8831,
"step": 579
},
{
"epoch": 0.5817452357071213,
"grad_norm": 3.629844903945923,
"learning_rate": 3.944028754802348e-06,
"loss": 0.9616,
"step": 580
},
{
"epoch": 0.5827482447342026,
"grad_norm": 4.511115074157715,
"learning_rate": 3.928156737905525e-06,
"loss": 0.8332,
"step": 581
},
{
"epoch": 0.5837512537612839,
"grad_norm": 5.118964195251465,
"learning_rate": 3.912296034006365e-06,
"loss": 0.9782,
"step": 582
},
{
"epoch": 0.5847542627883651,
"grad_norm": 4.1819915771484375,
"learning_rate": 3.896446810510042e-06,
"loss": 0.8465,
"step": 583
},
{
"epoch": 0.5857572718154463,
"grad_norm": 4.242429733276367,
"learning_rate": 3.880609234700554e-06,
"loss": 0.9946,
"step": 584
},
{
"epoch": 0.5867602808425276,
"grad_norm": 3.031345844268799,
"learning_rate": 3.864783473738965e-06,
"loss": 0.8462,
"step": 585
},
{
"epoch": 0.5877632898696088,
"grad_norm": 4.662555694580078,
"learning_rate": 3.848969694661634e-06,
"loss": 0.8552,
"step": 586
},
{
"epoch": 0.5887662988966901,
"grad_norm": 5.042914390563965,
"learning_rate": 3.833168064378455e-06,
"loss": 0.8229,
"step": 587
},
{
"epoch": 0.5897693079237714,
"grad_norm": 3.2561464309692383,
"learning_rate": 3.817378749671096e-06,
"loss": 0.8248,
"step": 588
},
{
"epoch": 0.5907723169508525,
"grad_norm": 5.049379348754883,
"learning_rate": 3.801601917191238e-06,
"loss": 0.7746,
"step": 589
},
{
"epoch": 0.5917753259779338,
"grad_norm": 4.368361473083496,
"learning_rate": 3.785837733458813e-06,
"loss": 0.8582,
"step": 590
},
{
"epoch": 0.5927783350050151,
"grad_norm": 4.155791759490967,
"learning_rate": 3.770086364860252e-06,
"loss": 0.8452,
"step": 591
},
{
"epoch": 0.5937813440320963,
"grad_norm": 4.592344284057617,
"learning_rate": 3.754347977646725e-06,
"loss": 0.8946,
"step": 592
},
{
"epoch": 0.5947843530591775,
"grad_norm": 3.998250722885132,
"learning_rate": 3.738622737932386e-06,
"loss": 0.8501,
"step": 593
},
{
"epoch": 0.5957873620862588,
"grad_norm": 4.634448528289795,
"learning_rate": 3.7229108116926226e-06,
"loss": 0.9039,
"step": 594
},
{
"epoch": 0.59679037111334,
"grad_norm": 5.307826519012451,
"learning_rate": 3.7072123647623015e-06,
"loss": 0.9555,
"step": 595
},
{
"epoch": 0.5977933801404213,
"grad_norm": 3.186934471130371,
"learning_rate": 3.6915275628340187e-06,
"loss": 0.6823,
"step": 596
},
{
"epoch": 0.5987963891675026,
"grad_norm": 3.9275457859039307,
"learning_rate": 3.6758565714563538e-06,
"loss": 0.9286,
"step": 597
},
{
"epoch": 0.5997993981945837,
"grad_norm": 3.847639799118042,
"learning_rate": 3.6601995560321167e-06,
"loss": 0.7823,
"step": 598
},
{
"epoch": 0.600802407221665,
"grad_norm": 3.12260103225708,
"learning_rate": 3.6445566818166076e-06,
"loss": 0.8017,
"step": 599
},
{
"epoch": 0.6018054162487463,
"grad_norm": 4.4961748123168945,
"learning_rate": 3.6289281139158687e-06,
"loss": 0.762,
"step": 600
},
{
"epoch": 0.6028084252758275,
"grad_norm": 3.9709415435791016,
"learning_rate": 3.6133140172849433e-06,
"loss": 0.793,
"step": 601
},
{
"epoch": 0.6038114343029087,
"grad_norm": 3.9396042823791504,
"learning_rate": 3.5977145567261357e-06,
"loss": 0.7384,
"step": 602
},
{
"epoch": 0.60481444332999,
"grad_norm": 4.9546332359313965,
"learning_rate": 3.5821298968872696e-06,
"loss": 0.9495,
"step": 603
},
{
"epoch": 0.6058174523570712,
"grad_norm": 4.9068708419799805,
"learning_rate": 3.5665602022599516e-06,
"loss": 0.7842,
"step": 604
},
{
"epoch": 0.6068204613841525,
"grad_norm": 4.662585258483887,
"learning_rate": 3.551005637177834e-06,
"loss": 0.7383,
"step": 605
},
{
"epoch": 0.6078234704112337,
"grad_norm": 4.0315961837768555,
"learning_rate": 3.535466365814884e-06,
"loss": 0.9327,
"step": 606
},
{
"epoch": 0.6088264794383149,
"grad_norm": 3.3967230319976807,
"learning_rate": 3.519942552183645e-06,
"loss": 0.8332,
"step": 607
},
{
"epoch": 0.6098294884653962,
"grad_norm": 4.561408042907715,
"learning_rate": 3.5044343601335106e-06,
"loss": 0.8087,
"step": 608
},
{
"epoch": 0.6108324974924775,
"grad_norm": 3.941023826599121,
"learning_rate": 3.48894195334899e-06,
"loss": 0.8227,
"step": 609
},
{
"epoch": 0.6118355065195586,
"grad_norm": 4.427003860473633,
"learning_rate": 3.4734654953479867e-06,
"loss": 0.81,
"step": 610
},
{
"epoch": 0.6128385155466399,
"grad_norm": 4.276743412017822,
"learning_rate": 3.4580051494800683e-06,
"loss": 0.7365,
"step": 611
},
{
"epoch": 0.6138415245737212,
"grad_norm": 3.4753129482269287,
"learning_rate": 3.4425610789247422e-06,
"loss": 0.8708,
"step": 612
},
{
"epoch": 0.6148445336008024,
"grad_norm": 3.8165876865386963,
"learning_rate": 3.4271334466897353e-06,
"loss": 0.6912,
"step": 613
},
{
"epoch": 0.6158475426278837,
"grad_norm": 4.05587911605835,
"learning_rate": 3.4117224156092754e-06,
"loss": 0.8261,
"step": 614
},
{
"epoch": 0.6168505516549649,
"grad_norm": 3.390885353088379,
"learning_rate": 3.396328148342366e-06,
"loss": 0.7844,
"step": 615
},
{
"epoch": 0.6178535606820461,
"grad_norm": 3.5539562702178955,
"learning_rate": 3.3809508073710757e-06,
"loss": 0.7445,
"step": 616
},
{
"epoch": 0.6188565697091274,
"grad_norm": 3.823821544647217,
"learning_rate": 3.3655905549988194e-06,
"loss": 0.6729,
"step": 617
},
{
"epoch": 0.6198595787362087,
"grad_norm": 2.9422364234924316,
"learning_rate": 3.350247553348648e-06,
"loss": 0.8044,
"step": 618
},
{
"epoch": 0.6208625877632898,
"grad_norm": 4.469389915466309,
"learning_rate": 3.334921964361535e-06,
"loss": 0.7135,
"step": 619
},
{
"epoch": 0.6218655967903711,
"grad_norm": 5.434567928314209,
"learning_rate": 3.319613949794668e-06,
"loss": 0.8655,
"step": 620
},
{
"epoch": 0.6228686058174524,
"grad_norm": 5.1121625900268555,
"learning_rate": 3.3043236712197448e-06,
"loss": 0.8928,
"step": 621
},
{
"epoch": 0.6238716148445336,
"grad_norm": 4.20275354385376,
"learning_rate": 3.289051290021259e-06,
"loss": 0.7245,
"step": 622
},
{
"epoch": 0.6248746238716149,
"grad_norm": 4.040639400482178,
"learning_rate": 3.2737969673948093e-06,
"loss": 0.8814,
"step": 623
},
{
"epoch": 0.6258776328986961,
"grad_norm": 5.009786128997803,
"learning_rate": 3.258560864345387e-06,
"loss": 0.8277,
"step": 624
},
{
"epoch": 0.6268806419257773,
"grad_norm": 4.480382442474365,
"learning_rate": 3.243343141685682e-06,
"loss": 0.8679,
"step": 625
},
{
"epoch": 0.6278836509528586,
"grad_norm": 4.348001480102539,
"learning_rate": 3.228143960034384e-06,
"loss": 0.8935,
"step": 626
},
{
"epoch": 0.6288866599799399,
"grad_norm": 3.9595484733581543,
"learning_rate": 3.212963479814489e-06,
"loss": 0.629,
"step": 627
},
{
"epoch": 0.629889669007021,
"grad_norm": 5.169403076171875,
"learning_rate": 3.1978018612516026e-06,
"loss": 0.9004,
"step": 628
},
{
"epoch": 0.6308926780341023,
"grad_norm": 5.387321472167969,
"learning_rate": 3.182659264372254e-06,
"loss": 0.8772,
"step": 629
},
{
"epoch": 0.6318956870611836,
"grad_norm": 4.459048748016357,
"learning_rate": 3.167535849002201e-06,
"loss": 0.7771,
"step": 630
},
{
"epoch": 0.6328986960882648,
"grad_norm": 4.168999671936035,
"learning_rate": 3.152431774764749e-06,
"loss": 0.6934,
"step": 631
},
{
"epoch": 0.633901705115346,
"grad_norm": 3.6566436290740967,
"learning_rate": 3.1373472010790617e-06,
"loss": 0.8692,
"step": 632
},
{
"epoch": 0.6349047141424273,
"grad_norm": 4.426352500915527,
"learning_rate": 3.122282287158479e-06,
"loss": 0.9844,
"step": 633
},
{
"epoch": 0.6359077231695085,
"grad_norm": 4.211833477020264,
"learning_rate": 3.1072371920088395e-06,
"loss": 0.7831,
"step": 634
},
{
"epoch": 0.6369107321965898,
"grad_norm": 3.8359768390655518,
"learning_rate": 3.0922120744267993e-06,
"loss": 0.6705,
"step": 635
},
{
"epoch": 0.6379137412236711,
"grad_norm": 4.391419410705566,
"learning_rate": 3.077207092998159e-06,
"loss": 0.8854,
"step": 636
},
{
"epoch": 0.6389167502507522,
"grad_norm": 4.663003444671631,
"learning_rate": 3.0622224060961834e-06,
"loss": 0.5812,
"step": 637
},
{
"epoch": 0.6399197592778335,
"grad_norm": 3.665712594985962,
"learning_rate": 3.0472581718799394e-06,
"loss": 0.7576,
"step": 638
},
{
"epoch": 0.6409227683049148,
"grad_norm": 4.79156494140625,
"learning_rate": 3.0323145482926185e-06,
"loss": 0.9444,
"step": 639
},
{
"epoch": 0.641925777331996,
"grad_norm": 3.7670211791992188,
"learning_rate": 3.0173916930598746e-06,
"loss": 0.8783,
"step": 640
},
{
"epoch": 0.6429287863590772,
"grad_norm": 5.021609783172607,
"learning_rate": 3.002489763688156e-06,
"loss": 0.7133,
"step": 641
},
{
"epoch": 0.6439317953861585,
"grad_norm": 5.50873327255249,
"learning_rate": 2.9876089174630467e-06,
"loss": 0.9178,
"step": 642
},
{
"epoch": 0.6449348044132397,
"grad_norm": 4.057301998138428,
"learning_rate": 2.9727493114476025e-06,
"loss": 0.9236,
"step": 643
},
{
"epoch": 0.645937813440321,
"grad_norm": 4.566405296325684,
"learning_rate": 2.9579111024806945e-06,
"loss": 0.7752,
"step": 644
},
{
"epoch": 0.6469408224674023,
"grad_norm": 3.470957040786743,
"learning_rate": 2.9430944471753563e-06,
"loss": 0.7844,
"step": 645
},
{
"epoch": 0.6479438314944834,
"grad_norm": 4.3673553466796875,
"learning_rate": 2.928299501917128e-06,
"loss": 0.7213,
"step": 646
},
{
"epoch": 0.6489468405215647,
"grad_norm": 4.8384528160095215,
"learning_rate": 2.9135264228624037e-06,
"loss": 0.7932,
"step": 647
},
{
"epoch": 0.649949849548646,
"grad_norm": 4.3337249755859375,
"learning_rate": 2.8987753659367888e-06,
"loss": 0.8125,
"step": 648
},
{
"epoch": 0.6509528585757272,
"grad_norm": 5.572671413421631,
"learning_rate": 2.8840464868334534e-06,
"loss": 0.9675,
"step": 649
},
{
"epoch": 0.6519558676028084,
"grad_norm": 4.2997660636901855,
"learning_rate": 2.8693399410114796e-06,
"loss": 0.8282,
"step": 650
},
{
"epoch": 0.6529588766298897,
"grad_norm": 4.765951633453369,
"learning_rate": 2.8546558836942385e-06,
"loss": 0.7906,
"step": 651
},
{
"epoch": 0.6539618856569709,
"grad_norm": 4.505593299865723,
"learning_rate": 2.8399944698677307e-06,
"loss": 0.7407,
"step": 652
},
{
"epoch": 0.6549648946840522,
"grad_norm": 4.665410995483398,
"learning_rate": 2.8253558542789706e-06,
"loss": 0.7051,
"step": 653
},
{
"epoch": 0.6559679037111334,
"grad_norm": 4.7228569984436035,
"learning_rate": 2.8107401914343367e-06,
"loss": 0.7442,
"step": 654
},
{
"epoch": 0.6569709127382146,
"grad_norm": 4.746094703674316,
"learning_rate": 2.7961476355979543e-06,
"loss": 0.8766,
"step": 655
},
{
"epoch": 0.6579739217652959,
"grad_norm": 4.242105960845947,
"learning_rate": 2.781578340790053e-06,
"loss": 0.8118,
"step": 656
},
{
"epoch": 0.6589769307923772,
"grad_norm": 5.309788227081299,
"learning_rate": 2.767032460785356e-06,
"loss": 0.836,
"step": 657
},
{
"epoch": 0.6599799398194583,
"grad_norm": 3.7402758598327637,
"learning_rate": 2.7525101491114494e-06,
"loss": 0.8005,
"step": 658
},
{
"epoch": 0.6609829488465396,
"grad_norm": 5.357933044433594,
"learning_rate": 2.7380115590471556e-06,
"loss": 0.8661,
"step": 659
},
{
"epoch": 0.6619859578736209,
"grad_norm": 4.294086933135986,
"learning_rate": 2.7235368436209312e-06,
"loss": 0.7023,
"step": 660
},
{
"epoch": 0.6629889669007021,
"grad_norm": 4.667852878570557,
"learning_rate": 2.709086155609235e-06,
"loss": 0.6972,
"step": 661
},
{
"epoch": 0.6639919759277834,
"grad_norm": 3.9336674213409424,
"learning_rate": 2.6946596475349306e-06,
"loss": 0.583,
"step": 662
},
{
"epoch": 0.6649949849548646,
"grad_norm": 4.823256969451904,
"learning_rate": 2.6802574716656616e-06,
"loss": 0.9382,
"step": 663
},
{
"epoch": 0.6659979939819458,
"grad_norm": 3.940246105194092,
"learning_rate": 2.6658797800122605e-06,
"loss": 0.7325,
"step": 664
},
{
"epoch": 0.6670010030090271,
"grad_norm": 4.826223850250244,
"learning_rate": 2.651526724327127e-06,
"loss": 0.759,
"step": 665
},
{
"epoch": 0.6680040120361084,
"grad_norm": 3.3941001892089844,
"learning_rate": 2.637198456102642e-06,
"loss": 0.6132,
"step": 666
},
{
"epoch": 0.6690070210631895,
"grad_norm": 4.705685138702393,
"learning_rate": 2.6228951265695625e-06,
"loss": 0.8048,
"step": 667
},
{
"epoch": 0.6700100300902708,
"grad_norm": 5.192498683929443,
"learning_rate": 2.6086168866954174e-06,
"loss": 0.8658,
"step": 668
},
{
"epoch": 0.6710130391173521,
"grad_norm": 4.6966776847839355,
"learning_rate": 2.5943638871829304e-06,
"loss": 0.8422,
"step": 669
},
{
"epoch": 0.6720160481444333,
"grad_norm": 5.429586887359619,
"learning_rate": 2.5801362784684107e-06,
"loss": 0.9351,
"step": 670
},
{
"epoch": 0.6730190571715146,
"grad_norm": 3.5969576835632324,
"learning_rate": 2.565934210720186e-06,
"loss": 0.7316,
"step": 671
},
{
"epoch": 0.6740220661985958,
"grad_norm": 3.8780598640441895,
"learning_rate": 2.5517578338369966e-06,
"loss": 0.757,
"step": 672
},
{
"epoch": 0.675025075225677,
"grad_norm": 4.866453647613525,
"learning_rate": 2.5376072974464285e-06,
"loss": 0.8405,
"step": 673
},
{
"epoch": 0.6760280842527583,
"grad_norm": 3.916318893432617,
"learning_rate": 2.52348275090333e-06,
"loss": 0.8406,
"step": 674
},
{
"epoch": 0.6770310932798396,
"grad_norm": 4.055075645446777,
"learning_rate": 2.5093843432882276e-06,
"loss": 0.759,
"step": 675
},
{
"epoch": 0.6780341023069207,
"grad_norm": 4.514216423034668,
"learning_rate": 2.4953122234057665e-06,
"loss": 0.7549,
"step": 676
},
{
"epoch": 0.679037111334002,
"grad_norm": 3.8270721435546875,
"learning_rate": 2.481266539783125e-06,
"loss": 0.8207,
"step": 677
},
{
"epoch": 0.6800401203610833,
"grad_norm": 4.5839691162109375,
"learning_rate": 2.4672474406684623e-06,
"loss": 0.9116,
"step": 678
},
{
"epoch": 0.6810431293881645,
"grad_norm": 3.7903006076812744,
"learning_rate": 2.4532550740293363e-06,
"loss": 0.7932,
"step": 679
},
{
"epoch": 0.6820461384152458,
"grad_norm": 4.376543045043945,
"learning_rate": 2.439289587551162e-06,
"loss": 0.7433,
"step": 680
},
{
"epoch": 0.683049147442327,
"grad_norm": 4.350039958953857,
"learning_rate": 2.4253511286356324e-06,
"loss": 0.8579,
"step": 681
},
{
"epoch": 0.6840521564694082,
"grad_norm": 3.196552276611328,
"learning_rate": 2.411439844399177e-06,
"loss": 0.6942,
"step": 682
},
{
"epoch": 0.6850551654964895,
"grad_norm": 3.100128173828125,
"learning_rate": 2.397555881671408e-06,
"loss": 0.6888,
"step": 683
},
{
"epoch": 0.6860581745235708,
"grad_norm": 5.0521159172058105,
"learning_rate": 2.383699386993557e-06,
"loss": 0.691,
"step": 684
},
{
"epoch": 0.6870611835506519,
"grad_norm": 5.03604793548584,
"learning_rate": 2.3698705066169486e-06,
"loss": 0.7578,
"step": 685
},
{
"epoch": 0.6880641925777332,
"grad_norm": 5.0749006271362305,
"learning_rate": 2.3560693865014383e-06,
"loss": 0.8756,
"step": 686
},
{
"epoch": 0.6890672016048145,
"grad_norm": 3.5390970706939697,
"learning_rate": 2.3422961723138864e-06,
"loss": 0.7515,
"step": 687
},
{
"epoch": 0.6900702106318957,
"grad_norm": 4.655844211578369,
"learning_rate": 2.328551009426609e-06,
"loss": 0.7575,
"step": 688
},
{
"epoch": 0.691073219658977,
"grad_norm": 4.955108165740967,
"learning_rate": 2.314834042915853e-06,
"loss": 0.8792,
"step": 689
},
{
"epoch": 0.6920762286860582,
"grad_norm": 4.956750392913818,
"learning_rate": 2.301145417560256e-06,
"loss": 0.7986,
"step": 690
},
{
"epoch": 0.6930792377131394,
"grad_norm": 4.120952129364014,
"learning_rate": 2.287485277839327e-06,
"loss": 0.6563,
"step": 691
},
{
"epoch": 0.6940822467402207,
"grad_norm": 3.3682332038879395,
"learning_rate": 2.2738537679319184e-06,
"loss": 0.677,
"step": 692
},
{
"epoch": 0.695085255767302,
"grad_norm": 4.446828365325928,
"learning_rate": 2.260251031714696e-06,
"loss": 0.7787,
"step": 693
},
{
"epoch": 0.6960882647943831,
"grad_norm": 3.3289477825164795,
"learning_rate": 2.2466772127606364e-06,
"loss": 0.7592,
"step": 694
},
{
"epoch": 0.6970912738214644,
"grad_norm": 4.985026836395264,
"learning_rate": 2.2331324543374944e-06,
"loss": 0.7962,
"step": 695
},
{
"epoch": 0.6980942828485457,
"grad_norm": 3.8351457118988037,
"learning_rate": 2.2196168994063076e-06,
"loss": 0.5227,
"step": 696
},
{
"epoch": 0.6990972918756269,
"grad_norm": 4.575567245483398,
"learning_rate": 2.206130690619871e-06,
"loss": 0.8213,
"step": 697
},
{
"epoch": 0.7001003009027081,
"grad_norm": 3.6639046669006348,
"learning_rate": 2.1926739703212474e-06,
"loss": 0.7516,
"step": 698
},
{
"epoch": 0.7011033099297894,
"grad_norm": 4.742379188537598,
"learning_rate": 2.179246880542249e-06,
"loss": 0.897,
"step": 699
},
{
"epoch": 0.7021063189568706,
"grad_norm": 3.5077478885650635,
"learning_rate": 2.165849563001952e-06,
"loss": 0.7371,
"step": 700
},
{
"epoch": 0.7031093279839519,
"grad_norm": 4.919702529907227,
"learning_rate": 2.1524821591051963e-06,
"loss": 0.8393,
"step": 701
},
{
"epoch": 0.7041123370110332,
"grad_norm": 4.482843399047852,
"learning_rate": 2.1391448099410856e-06,
"loss": 0.6463,
"step": 702
},
{
"epoch": 0.7051153460381143,
"grad_norm": 3.620887041091919,
"learning_rate": 2.1258376562815112e-06,
"loss": 0.7534,
"step": 703
},
{
"epoch": 0.7061183550651956,
"grad_norm": 4.8982977867126465,
"learning_rate": 2.112560838579653e-06,
"loss": 0.909,
"step": 704
},
{
"epoch": 0.7071213640922769,
"grad_norm": 4.648558139801025,
"learning_rate": 2.099314496968511e-06,
"loss": 0.6678,
"step": 705
},
{
"epoch": 0.708124373119358,
"grad_norm": 4.6663031578063965,
"learning_rate": 2.0860987712594103e-06,
"loss": 0.7577,
"step": 706
},
{
"epoch": 0.7091273821464393,
"grad_norm": 4.495550632476807,
"learning_rate": 2.0729138009405417e-06,
"loss": 0.9459,
"step": 707
},
{
"epoch": 0.7101303911735206,
"grad_norm": 4.677994728088379,
"learning_rate": 2.0597597251754803e-06,
"loss": 0.8014,
"step": 708
},
{
"epoch": 0.7111334002006018,
"grad_norm": 4.385025978088379,
"learning_rate": 2.0466366828017115e-06,
"loss": 0.7715,
"step": 709
},
{
"epoch": 0.7121364092276831,
"grad_norm": 4.049365520477295,
"learning_rate": 2.033544812329181e-06,
"loss": 0.8148,
"step": 710
},
{
"epoch": 0.7131394182547643,
"grad_norm": 4.608902454376221,
"learning_rate": 2.0204842519388176e-06,
"loss": 0.7244,
"step": 711
},
{
"epoch": 0.7141424272818455,
"grad_norm": 4.603305339813232,
"learning_rate": 2.0074551394810853e-06,
"loss": 0.7097,
"step": 712
},
{
"epoch": 0.7151454363089268,
"grad_norm": 6.252641677856445,
"learning_rate": 1.9944576124745207e-06,
"loss": 0.6376,
"step": 713
},
{
"epoch": 0.7161484453360081,
"grad_norm": 4.792734146118164,
"learning_rate": 1.9814918081042887e-06,
"loss": 0.9104,
"step": 714
},
{
"epoch": 0.7171514543630892,
"grad_norm": 4.855016708374023,
"learning_rate": 1.968557863220727e-06,
"loss": 0.7701,
"step": 715
},
{
"epoch": 0.7181544633901705,
"grad_norm": 3.6340222358703613,
"learning_rate": 1.95565591433791e-06,
"loss": 0.7832,
"step": 716
},
{
"epoch": 0.7191574724172518,
"grad_norm": 3.9203600883483887,
"learning_rate": 1.9427860976322e-06,
"loss": 0.613,
"step": 717
},
{
"epoch": 0.720160481444333,
"grad_norm": 4.4278178215026855,
"learning_rate": 1.9299485489408125e-06,
"loss": 0.7788,
"step": 718
},
{
"epoch": 0.7211634904714143,
"grad_norm": 3.5168228149414062,
"learning_rate": 1.9171434037603885e-06,
"loss": 0.7084,
"step": 719
},
{
"epoch": 0.7221664994984955,
"grad_norm": 4.843722343444824,
"learning_rate": 1.9043707972455538e-06,
"loss": 0.7999,
"step": 720
},
{
"epoch": 0.7231695085255767,
"grad_norm": 3.9402291774749756,
"learning_rate": 1.891630864207501e-06,
"loss": 0.8285,
"step": 721
},
{
"epoch": 0.724172517552658,
"grad_norm": 3.990156888961792,
"learning_rate": 1.8789237391125647e-06,
"loss": 0.9103,
"step": 722
},
{
"epoch": 0.7251755265797393,
"grad_norm": 5.624143600463867,
"learning_rate": 1.866249556080796e-06,
"loss": 0.7957,
"step": 723
},
{
"epoch": 0.7261785356068204,
"grad_norm": 5.01121187210083,
"learning_rate": 1.8536084488845585e-06,
"loss": 0.7271,
"step": 724
},
{
"epoch": 0.7271815446339017,
"grad_norm": 5.025772571563721,
"learning_rate": 1.841000550947103e-06,
"loss": 0.8014,
"step": 725
},
{
"epoch": 0.728184553660983,
"grad_norm": 4.405618667602539,
"learning_rate": 1.8284259953411731e-06,
"loss": 0.829,
"step": 726
},
{
"epoch": 0.7291875626880642,
"grad_norm": 4.663231372833252,
"learning_rate": 1.8158849147875873e-06,
"loss": 0.8276,
"step": 727
},
{
"epoch": 0.7301905717151455,
"grad_norm": 4.1873650550842285,
"learning_rate": 1.8033774416538501e-06,
"loss": 0.8987,
"step": 728
},
{
"epoch": 0.7311935807422267,
"grad_norm": 5.60460901260376,
"learning_rate": 1.7909037079527436e-06,
"loss": 0.6632,
"step": 729
},
{
"epoch": 0.7321965897693079,
"grad_norm": 4.5685577392578125,
"learning_rate": 1.7784638453409452e-06,
"loss": 0.8197,
"step": 730
},
{
"epoch": 0.7331995987963892,
"grad_norm": 4.601626873016357,
"learning_rate": 1.7660579851176318e-06,
"loss": 0.8307,
"step": 731
},
{
"epoch": 0.7342026078234705,
"grad_norm": 3.853898286819458,
"learning_rate": 1.7536862582230896e-06,
"loss": 0.6466,
"step": 732
},
{
"epoch": 0.7352056168505516,
"grad_norm": 4.068875789642334,
"learning_rate": 1.7413487952373459e-06,
"loss": 0.8667,
"step": 733
},
{
"epoch": 0.7362086258776329,
"grad_norm": 4.8334455490112305,
"learning_rate": 1.729045726378773e-06,
"loss": 0.7005,
"step": 734
},
{
"epoch": 0.7372116349047142,
"grad_norm": 4.624211311340332,
"learning_rate": 1.7167771815027318e-06,
"loss": 0.7219,
"step": 735
},
{
"epoch": 0.7382146439317954,
"grad_norm": 4.838943004608154,
"learning_rate": 1.7045432901001846e-06,
"loss": 0.7754,
"step": 736
},
{
"epoch": 0.7392176529588766,
"grad_norm": 4.927722930908203,
"learning_rate": 1.6923441812963437e-06,
"loss": 0.8565,
"step": 737
},
{
"epoch": 0.7402206619859579,
"grad_norm": 4.852552890777588,
"learning_rate": 1.6801799838492945e-06,
"loss": 0.7327,
"step": 738
},
{
"epoch": 0.7412236710130391,
"grad_norm": 3.834653377532959,
"learning_rate": 1.6680508261486468e-06,
"loss": 0.6826,
"step": 739
},
{
"epoch": 0.7422266800401204,
"grad_norm": 5.076817512512207,
"learning_rate": 1.655956836214177e-06,
"loss": 0.8856,
"step": 740
},
{
"epoch": 0.7432296890672017,
"grad_norm": 4.383863925933838,
"learning_rate": 1.643898141694471e-06,
"loss": 0.7258,
"step": 741
},
{
"epoch": 0.7442326980942828,
"grad_norm": 4.276555061340332,
"learning_rate": 1.6318748698655872e-06,
"loss": 0.7015,
"step": 742
},
{
"epoch": 0.7452357071213641,
"grad_norm": 3.9618241786956787,
"learning_rate": 1.6198871476297034e-06,
"loss": 0.7138,
"step": 743
},
{
"epoch": 0.7462387161484454,
"grad_norm": 3.8278191089630127,
"learning_rate": 1.607935101513785e-06,
"loss": 0.6459,
"step": 744
},
{
"epoch": 0.7472417251755266,
"grad_norm": 4.899794101715088,
"learning_rate": 1.5960188576682422e-06,
"loss": 0.6454,
"step": 745
},
{
"epoch": 0.7482447342026078,
"grad_norm": 4.544212818145752,
"learning_rate": 1.5841385418656068e-06,
"loss": 0.7824,
"step": 746
},
{
"epoch": 0.7492477432296891,
"grad_norm": 5.481677532196045,
"learning_rate": 1.5722942794991996e-06,
"loss": 0.8462,
"step": 747
},
{
"epoch": 0.7502507522567703,
"grad_norm": 5.115010738372803,
"learning_rate": 1.560486195581804e-06,
"loss": 0.7292,
"step": 748
},
{
"epoch": 0.7512537612838516,
"grad_norm": 4.429975986480713,
"learning_rate": 1.5487144147443561e-06,
"loss": 0.806,
"step": 749
},
{
"epoch": 0.7522567703109327,
"grad_norm": 4.784054279327393,
"learning_rate": 1.536979061234617e-06,
"loss": 0.7697,
"step": 750
},
{
"epoch": 0.753259779338014,
"grad_norm": 4.078995227813721,
"learning_rate": 1.5252802589158738e-06,
"loss": 0.5987,
"step": 751
},
{
"epoch": 0.7542627883650953,
"grad_norm": 4.898708343505859,
"learning_rate": 1.513618131265621e-06,
"loss": 0.7273,
"step": 752
},
{
"epoch": 0.7552657973921765,
"grad_norm": 4.614200592041016,
"learning_rate": 1.5019928013742685e-06,
"loss": 0.637,
"step": 753
},
{
"epoch": 0.7562688064192578,
"grad_norm": 4.564877510070801,
"learning_rate": 1.4904043919438294e-06,
"loss": 0.7045,
"step": 754
},
{
"epoch": 0.757271815446339,
"grad_norm": 5.43159294128418,
"learning_rate": 1.4788530252866373e-06,
"loss": 0.7572,
"step": 755
},
{
"epoch": 0.7582748244734202,
"grad_norm": 5.415429592132568,
"learning_rate": 1.4673388233240505e-06,
"loss": 0.7423,
"step": 756
},
{
"epoch": 0.7592778335005015,
"grad_norm": 4.873239517211914,
"learning_rate": 1.4558619075851582e-06,
"loss": 0.7772,
"step": 757
},
{
"epoch": 0.7602808425275828,
"grad_norm": 4.5504841804504395,
"learning_rate": 1.4444223992055117e-06,
"loss": 0.7693,
"step": 758
},
{
"epoch": 0.7612838515546639,
"grad_norm": 4.655095100402832,
"learning_rate": 1.4330204189258329e-06,
"loss": 0.7821,
"step": 759
},
{
"epoch": 0.7622868605817452,
"grad_norm": 4.574293613433838,
"learning_rate": 1.4216560870907498e-06,
"loss": 0.7757,
"step": 760
},
{
"epoch": 0.7632898696088265,
"grad_norm": 5.1897807121276855,
"learning_rate": 1.4103295236475168e-06,
"loss": 0.6772,
"step": 761
},
{
"epoch": 0.7642928786359077,
"grad_norm": 4.52416467666626,
"learning_rate": 1.3990408481447599e-06,
"loss": 0.7775,
"step": 762
},
{
"epoch": 0.765295887662989,
"grad_norm": 5.417598724365234,
"learning_rate": 1.3877901797312022e-06,
"loss": 0.9068,
"step": 763
},
{
"epoch": 0.7662988966900702,
"grad_norm": 5.372793197631836,
"learning_rate": 1.3765776371544177e-06,
"loss": 0.8809,
"step": 764
},
{
"epoch": 0.7673019057171514,
"grad_norm": 4.353141784667969,
"learning_rate": 1.3654033387595733e-06,
"loss": 0.5866,
"step": 765
},
{
"epoch": 0.7683049147442327,
"grad_norm": 5.083629608154297,
"learning_rate": 1.3542674024881747e-06,
"loss": 0.7205,
"step": 766
},
{
"epoch": 0.769307923771314,
"grad_norm": 4.177886009216309,
"learning_rate": 1.3431699458768333e-06,
"loss": 0.7128,
"step": 767
},
{
"epoch": 0.7703109327983951,
"grad_norm": 4.836990833282471,
"learning_rate": 1.332111086056011e-06,
"loss": 0.7707,
"step": 768
},
{
"epoch": 0.7713139418254764,
"grad_norm": 4.888391971588135,
"learning_rate": 1.3210909397487998e-06,
"loss": 0.6584,
"step": 769
},
{
"epoch": 0.7723169508525577,
"grad_norm": 5.298000812530518,
"learning_rate": 1.3101096232696735e-06,
"loss": 0.8403,
"step": 770
},
{
"epoch": 0.7733199598796389,
"grad_norm": 5.193215370178223,
"learning_rate": 1.2991672525232758e-06,
"loss": 0.8047,
"step": 771
},
{
"epoch": 0.7743229689067201,
"grad_norm": 3.786947250366211,
"learning_rate": 1.2882639430031836e-06,
"loss": 0.7377,
"step": 772
},
{
"epoch": 0.7753259779338014,
"grad_norm": 4.563846111297607,
"learning_rate": 1.2773998097906965e-06,
"loss": 0.5632,
"step": 773
},
{
"epoch": 0.7763289869608826,
"grad_norm": 4.023874282836914,
"learning_rate": 1.266574967553621e-06,
"loss": 0.6381,
"step": 774
},
{
"epoch": 0.7773319959879639,
"grad_norm": 5.078170299530029,
"learning_rate": 1.2557895305450535e-06,
"loss": 0.7863,
"step": 775
},
{
"epoch": 0.7783350050150452,
"grad_norm": 4.149288177490234,
"learning_rate": 1.2450436126021864e-06,
"loss": 0.6536,
"step": 776
},
{
"epoch": 0.7793380140421263,
"grad_norm": 3.8331847190856934,
"learning_rate": 1.234337327145092e-06,
"loss": 0.6756,
"step": 777
},
{
"epoch": 0.7803410230692076,
"grad_norm": 4.096665382385254,
"learning_rate": 1.2236707871755404e-06,
"loss": 0.7581,
"step": 778
},
{
"epoch": 0.7813440320962889,
"grad_norm": 4.041867733001709,
"learning_rate": 1.213044105275794e-06,
"loss": 0.7642,
"step": 779
},
{
"epoch": 0.7823470411233701,
"grad_norm": 4.151028633117676,
"learning_rate": 1.2024573936074275e-06,
"loss": 0.738,
"step": 780
},
{
"epoch": 0.7833500501504513,
"grad_norm": 4.884620666503906,
"learning_rate": 1.1919107639101424e-06,
"loss": 0.6238,
"step": 781
},
{
"epoch": 0.7843530591775326,
"grad_norm": 5.395772457122803,
"learning_rate": 1.181404327500582e-06,
"loss": 0.7566,
"step": 782
},
{
"epoch": 0.7853560682046138,
"grad_norm": 4.006275177001953,
"learning_rate": 1.1709381952711668e-06,
"loss": 0.7842,
"step": 783
},
{
"epoch": 0.7863590772316951,
"grad_norm": 3.6466565132141113,
"learning_rate": 1.1605124776889126e-06,
"loss": 0.5598,
"step": 784
},
{
"epoch": 0.7873620862587764,
"grad_norm": 3.977656126022339,
"learning_rate": 1.1501272847942752e-06,
"loss": 0.6004,
"step": 785
},
{
"epoch": 0.7883650952858575,
"grad_norm": 5.086451530456543,
"learning_rate": 1.1397827261999795e-06,
"loss": 0.7513,
"step": 786
},
{
"epoch": 0.7893681043129388,
"grad_norm": 3.921558380126953,
"learning_rate": 1.1294789110898712e-06,
"loss": 0.5705,
"step": 787
},
{
"epoch": 0.7903711133400201,
"grad_norm": 5.067392349243164,
"learning_rate": 1.1192159482177561e-06,
"loss": 0.7763,
"step": 788
},
{
"epoch": 0.7913741223671013,
"grad_norm": 5.236445426940918,
"learning_rate": 1.1089939459062605e-06,
"loss": 0.8821,
"step": 789
},
{
"epoch": 0.7923771313941825,
"grad_norm": 4.675759315490723,
"learning_rate": 1.0988130120456813e-06,
"loss": 0.7719,
"step": 790
},
{
"epoch": 0.7933801404212638,
"grad_norm": 4.499670505523682,
"learning_rate": 1.088673254092849e-06,
"loss": 0.8,
"step": 791
},
{
"epoch": 0.794383149448345,
"grad_norm": 4.458972930908203,
"learning_rate": 1.078574779069998e-06,
"loss": 0.7312,
"step": 792
},
{
"epoch": 0.7953861584754263,
"grad_norm": 5.155145168304443,
"learning_rate": 1.0685176935636265e-06,
"loss": 0.6534,
"step": 793
},
{
"epoch": 0.7963891675025075,
"grad_norm": 5.616570472717285,
"learning_rate": 1.0585021037233873e-06,
"loss": 0.8878,
"step": 794
},
{
"epoch": 0.7973921765295887,
"grad_norm": 4.5944037437438965,
"learning_rate": 1.0485281152609484e-06,
"loss": 0.8386,
"step": 795
},
{
"epoch": 0.79839518555667,
"grad_norm": 4.263227939605713,
"learning_rate": 1.0385958334488967e-06,
"loss": 0.7942,
"step": 796
},
{
"epoch": 0.7993981945837513,
"grad_norm": 4.048926830291748,
"learning_rate": 1.0287053631196109e-06,
"loss": 0.7017,
"step": 797
},
{
"epoch": 0.8004012036108324,
"grad_norm": 4.2994303703308105,
"learning_rate": 1.0188568086641614e-06,
"loss": 0.7319,
"step": 798
},
{
"epoch": 0.8014042126379137,
"grad_norm": 4.350577354431152,
"learning_rate": 1.0090502740312153e-06,
"loss": 0.8326,
"step": 799
},
{
"epoch": 0.802407221664995,
"grad_norm": 4.654552936553955,
"learning_rate": 9.992858627259238e-07,
"loss": 0.7457,
"step": 800
},
{
"epoch": 0.8034102306920762,
"grad_norm": 4.828591346740723,
"learning_rate": 9.895636778088458e-07,
"loss": 0.5425,
"step": 801
},
{
"epoch": 0.8044132397191575,
"grad_norm": 4.602827548980713,
"learning_rate": 9.79883821894847e-07,
"loss": 0.7285,
"step": 802
},
{
"epoch": 0.8054162487462387,
"grad_norm": 4.667062759399414,
"learning_rate": 9.702463971520264e-07,
"loss": 0.5778,
"step": 803
},
{
"epoch": 0.8064192577733199,
"grad_norm": 4.593809604644775,
"learning_rate": 9.60651505300635e-07,
"loss": 0.7539,
"step": 804
},
{
"epoch": 0.8074222668004012,
"grad_norm": 4.234933853149414,
"learning_rate": 9.510992476119962e-07,
"loss": 0.7307,
"step": 805
},
{
"epoch": 0.8084252758274825,
"grad_norm": 4.887782096862793,
"learning_rate": 9.415897249074479e-07,
"loss": 0.728,
"step": 806
},
{
"epoch": 0.8094282848545636,
"grad_norm": 4.268975257873535,
"learning_rate": 9.321230375572682e-07,
"loss": 0.8332,
"step": 807
},
{
"epoch": 0.8104312938816449,
"grad_norm": 4.051448822021484,
"learning_rate": 9.226992854796235e-07,
"loss": 0.7643,
"step": 808
},
{
"epoch": 0.8114343029087262,
"grad_norm": 4.242846488952637,
"learning_rate": 9.133185681395074e-07,
"loss": 0.8009,
"step": 809
},
{
"epoch": 0.8124373119358074,
"grad_norm": 4.693063735961914,
"learning_rate": 9.039809845476971e-07,
"loss": 0.7375,
"step": 810
},
{
"epoch": 0.8134403209628887,
"grad_norm": 4.185790538787842,
"learning_rate": 8.946866332597065e-07,
"loss": 0.7547,
"step": 811
},
{
"epoch": 0.8144433299899699,
"grad_norm": 4.585914611816406,
"learning_rate": 8.854356123747393e-07,
"loss": 0.6169,
"step": 812
},
{
"epoch": 0.8154463390170511,
"grad_norm": 4.708098411560059,
"learning_rate": 8.762280195346656e-07,
"loss": 0.7005,
"step": 813
},
{
"epoch": 0.8164493480441324,
"grad_norm": 4.51668119430542,
"learning_rate": 8.670639519229801e-07,
"loss": 0.7734,
"step": 814
},
{
"epoch": 0.8174523570712137,
"grad_norm": 4.6272149085998535,
"learning_rate": 8.579435062637864e-07,
"loss": 0.8573,
"step": 815
},
{
"epoch": 0.8184553660982948,
"grad_norm": 4.618356227874756,
"learning_rate": 8.488667788207644e-07,
"loss": 0.6907,
"step": 816
},
{
"epoch": 0.8194583751253761,
"grad_norm": 5.047024726867676,
"learning_rate": 8.398338653961674e-07,
"loss": 0.9222,
"step": 817
},
{
"epoch": 0.8204613841524574,
"grad_norm": 4.793258190155029,
"learning_rate": 8.308448613297982e-07,
"loss": 0.6165,
"step": 818
},
{
"epoch": 0.8214643931795386,
"grad_norm": 5.409603118896484,
"learning_rate": 8.218998614980134e-07,
"loss": 0.6569,
"step": 819
},
{
"epoch": 0.8224674022066198,
"grad_norm": 4.52425479888916,
"learning_rate": 8.12998960312718e-07,
"loss": 0.7587,
"step": 820
},
{
"epoch": 0.8234704112337011,
"grad_norm": 4.566806793212891,
"learning_rate": 8.041422517203629e-07,
"loss": 0.8022,
"step": 821
},
{
"epoch": 0.8244734202607823,
"grad_norm": 5.383869171142578,
"learning_rate": 7.953298292009659e-07,
"loss": 0.8917,
"step": 822
},
{
"epoch": 0.8254764292878636,
"grad_norm": 5.402740478515625,
"learning_rate": 7.865617857671121e-07,
"loss": 0.6883,
"step": 823
},
{
"epoch": 0.8264794383149449,
"grad_norm": 4.737725734710693,
"learning_rate": 7.778382139629831e-07,
"loss": 0.8098,
"step": 824
},
{
"epoch": 0.827482447342026,
"grad_norm": 4.177610874176025,
"learning_rate": 7.691592058633696e-07,
"loss": 0.6841,
"step": 825
},
{
"epoch": 0.8284854563691073,
"grad_norm": 4.095300197601318,
"learning_rate": 7.605248530727116e-07,
"loss": 0.6517,
"step": 826
},
{
"epoch": 0.8294884653961886,
"grad_norm": 4.254879951477051,
"learning_rate": 7.519352467241198e-07,
"loss": 0.7027,
"step": 827
},
{
"epoch": 0.8304914744232698,
"grad_norm": 3.8898143768310547,
"learning_rate": 7.433904774784217e-07,
"loss": 0.7203,
"step": 828
},
{
"epoch": 0.831494483450351,
"grad_norm": 5.161689281463623,
"learning_rate": 7.348906355232027e-07,
"loss": 0.8591,
"step": 829
},
{
"epoch": 0.8324974924774323,
"grad_norm": 4.899776935577393,
"learning_rate": 7.264358105718505e-07,
"loss": 0.8149,
"step": 830
},
{
"epoch": 0.8335005015045135,
"grad_norm": 4.75742769241333,
"learning_rate": 7.180260918626153e-07,
"loss": 0.7591,
"step": 831
},
{
"epoch": 0.8345035105315948,
"grad_norm": 3.808969736099243,
"learning_rate": 7.096615681576596e-07,
"loss": 0.6284,
"step": 832
},
{
"epoch": 0.835506519558676,
"grad_norm": 4.761439323425293,
"learning_rate": 7.0134232774213e-07,
"loss": 0.8803,
"step": 833
},
{
"epoch": 0.8365095285857572,
"grad_norm": 4.2167744636535645,
"learning_rate": 6.930684584232161e-07,
"loss": 0.8008,
"step": 834
},
{
"epoch": 0.8375125376128385,
"grad_norm": 4.578309535980225,
"learning_rate": 6.848400475292344e-07,
"loss": 0.7168,
"step": 835
},
{
"epoch": 0.8385155466399198,
"grad_norm": 4.427273750305176,
"learning_rate": 6.766571819086942e-07,
"loss": 0.7554,
"step": 836
},
{
"epoch": 0.839518555667001,
"grad_norm": 3.3490822315216064,
"learning_rate": 6.68519947929393e-07,
"loss": 0.6519,
"step": 837
},
{
"epoch": 0.8405215646940822,
"grad_norm": 3.7455391883850098,
"learning_rate": 6.604284314774984e-07,
"loss": 0.7362,
"step": 838
},
{
"epoch": 0.8415245737211635,
"grad_norm": 5.733519554138184,
"learning_rate": 6.523827179566394e-07,
"loss": 0.6435,
"step": 839
},
{
"epoch": 0.8425275827482447,
"grad_norm": 5.544914722442627,
"learning_rate": 6.443828922870127e-07,
"loss": 0.8534,
"step": 840
},
{
"epoch": 0.843530591775326,
"grad_norm": 4.3752593994140625,
"learning_rate": 6.36429038904477e-07,
"loss": 0.7725,
"step": 841
},
{
"epoch": 0.8445336008024072,
"grad_norm": 4.2751994132995605,
"learning_rate": 6.28521241759672e-07,
"loss": 0.6116,
"step": 842
},
{
"epoch": 0.8455366098294884,
"grad_norm": 4.379554748535156,
"learning_rate": 6.206595843171225e-07,
"loss": 0.5035,
"step": 843
},
{
"epoch": 0.8465396188565697,
"grad_norm": 4.365800380706787,
"learning_rate": 6.128441495543647e-07,
"loss": 0.7656,
"step": 844
},
{
"epoch": 0.847542627883651,
"grad_norm": 4.838818073272705,
"learning_rate": 6.050750199610683e-07,
"loss": 0.8122,
"step": 845
},
{
"epoch": 0.8485456369107321,
"grad_norm": 4.719117164611816,
"learning_rate": 5.973522775381618e-07,
"loss": 0.915,
"step": 846
},
{
"epoch": 0.8495486459378134,
"grad_norm": 4.398984909057617,
"learning_rate": 5.89676003796974e-07,
"loss": 0.5796,
"step": 847
},
{
"epoch": 0.8505516549648947,
"grad_norm": 4.586338520050049,
"learning_rate": 5.820462797583671e-07,
"loss": 0.615,
"step": 848
},
{
"epoch": 0.8515546639919759,
"grad_norm": 4.437315464019775,
"learning_rate": 5.744631859518879e-07,
"loss": 0.6241,
"step": 849
},
{
"epoch": 0.8525576730190572,
"grad_norm": 3.477703809738159,
"learning_rate": 5.669268024149111e-07,
"loss": 0.7041,
"step": 850
},
{
"epoch": 0.8535606820461384,
"grad_norm": 5.6405348777771,
"learning_rate": 5.59437208691801e-07,
"loss": 0.8992,
"step": 851
},
{
"epoch": 0.8545636910732196,
"grad_norm": 4.2000250816345215,
"learning_rate": 5.51994483833066e-07,
"loss": 0.7213,
"step": 852
},
{
"epoch": 0.8555667001003009,
"grad_norm": 4.124228000640869,
"learning_rate": 5.44598706394529e-07,
"loss": 0.7059,
"step": 853
},
{
"epoch": 0.8565697091273822,
"grad_norm": 5.095402717590332,
"learning_rate": 5.372499544364973e-07,
"loss": 0.7081,
"step": 854
},
{
"epoch": 0.8575727181544633,
"grad_norm": 5.275686740875244,
"learning_rate": 5.299483055229338e-07,
"loss": 0.9325,
"step": 855
},
{
"epoch": 0.8585757271815446,
"grad_norm": 5.121447563171387,
"learning_rate": 5.226938367206474e-07,
"loss": 0.717,
"step": 856
},
{
"epoch": 0.8595787362086259,
"grad_norm": 3.1360950469970703,
"learning_rate": 5.154866245984697e-07,
"loss": 0.6586,
"step": 857
},
{
"epoch": 0.8605817452357071,
"grad_norm": 5.618362903594971,
"learning_rate": 5.083267452264556e-07,
"loss": 0.7734,
"step": 858
},
{
"epoch": 0.8615847542627884,
"grad_norm": 4.624363422393799,
"learning_rate": 5.012142741750725e-07,
"loss": 0.6275,
"step": 859
},
{
"epoch": 0.8625877632898696,
"grad_norm": 3.7502849102020264,
"learning_rate": 4.941492865144115e-07,
"loss": 0.6578,
"step": 860
},
{
"epoch": 0.8635907723169508,
"grad_norm": 4.694582939147949,
"learning_rate": 4.871318568133848e-07,
"loss": 0.812,
"step": 861
},
{
"epoch": 0.8645937813440321,
"grad_norm": 4.233691215515137,
"learning_rate": 4.801620591389478e-07,
"loss": 0.6501,
"step": 862
},
{
"epoch": 0.8655967903711134,
"grad_norm": 4.722357273101807,
"learning_rate": 4.732399670553134e-07,
"loss": 0.5827,
"step": 863
},
{
"epoch": 0.8665997993981945,
"grad_norm": 4.257192611694336,
"learning_rate": 4.663656536231731e-07,
"loss": 0.8428,
"step": 864
},
{
"epoch": 0.8676028084252758,
"grad_norm": 4.021536827087402,
"learning_rate": 4.595391913989311e-07,
"loss": 0.6767,
"step": 865
},
{
"epoch": 0.8686058174523571,
"grad_norm": 5.024959564208984,
"learning_rate": 4.527606524339329e-07,
"loss": 0.681,
"step": 866
},
{
"epoch": 0.8696088264794383,
"grad_norm": 6.4967193603515625,
"learning_rate": 4.460301082737123e-07,
"loss": 0.8774,
"step": 867
},
{
"epoch": 0.8706118355065195,
"grad_norm": 4.467845439910889,
"learning_rate": 4.393476299572264e-07,
"loss": 0.6123,
"step": 868
},
{
"epoch": 0.8716148445336008,
"grad_norm": 4.03832483291626,
"learning_rate": 4.327132880161161e-07,
"loss": 0.709,
"step": 869
},
{
"epoch": 0.872617853560682,
"grad_norm": 4.846073150634766,
"learning_rate": 4.261271524739524e-07,
"loss": 0.799,
"step": 870
},
{
"epoch": 0.8736208625877633,
"grad_norm": 4.465785503387451,
"learning_rate": 4.195892928455048e-07,
"loss": 0.7433,
"step": 871
},
{
"epoch": 0.8746238716148446,
"grad_norm": 4.872369766235352,
"learning_rate": 4.1309977813600353e-07,
"loss": 0.7538,
"step": 872
},
{
"epoch": 0.8756268806419257,
"grad_norm": 3.6774091720581055,
"learning_rate": 4.066586768404102e-07,
"loss": 0.654,
"step": 873
},
{
"epoch": 0.876629889669007,
"grad_norm": 4.146250247955322,
"learning_rate": 4.002660569426997e-07,
"loss": 0.5712,
"step": 874
},
{
"epoch": 0.8776328986960883,
"grad_norm": 3.8843162059783936,
"learning_rate": 3.9392198591513775e-07,
"loss": 0.6435,
"step": 875
},
{
"epoch": 0.8786359077231695,
"grad_norm": 4.3653059005737305,
"learning_rate": 3.8762653071757137e-07,
"loss": 0.7175,
"step": 876
},
{
"epoch": 0.8796389167502507,
"grad_norm": 5.461905002593994,
"learning_rate": 3.81379757796721e-07,
"loss": 0.7422,
"step": 877
},
{
"epoch": 0.880641925777332,
"grad_norm": 3.634816884994507,
"learning_rate": 3.7518173308548065e-07,
"loss": 0.7174,
"step": 878
},
{
"epoch": 0.8816449348044132,
"grad_norm": 4.787415504455566,
"learning_rate": 3.690325220022201e-07,
"loss": 0.7121,
"step": 879
},
{
"epoch": 0.8826479438314945,
"grad_norm": 3.988579273223877,
"learning_rate": 3.6293218945009365e-07,
"loss": 0.6696,
"step": 880
},
{
"epoch": 0.8836509528585758,
"grad_norm": 4.939713478088379,
"learning_rate": 3.5688079981636203e-07,
"loss": 0.7605,
"step": 881
},
{
"epoch": 0.8846539618856569,
"grad_norm": 5.217735290527344,
"learning_rate": 3.508784169717011e-07,
"loss": 0.8241,
"step": 882
},
{
"epoch": 0.8856569709127382,
"grad_norm": 4.151040077209473,
"learning_rate": 3.4492510426953783e-07,
"loss": 0.6344,
"step": 883
},
{
"epoch": 0.8866599799398195,
"grad_norm": 4.423863887786865,
"learning_rate": 3.3902092454537904e-07,
"loss": 0.5726,
"step": 884
},
{
"epoch": 0.8876629889669007,
"grad_norm": 5.126745700836182,
"learning_rate": 3.3316594011614357e-07,
"loss": 0.744,
"step": 885
},
{
"epoch": 0.8886659979939819,
"grad_norm": 4.856997013092041,
"learning_rate": 3.2736021277951057e-07,
"loss": 0.634,
"step": 886
},
{
"epoch": 0.8896690070210632,
"grad_norm": 3.5816633701324463,
"learning_rate": 3.216038038132624e-07,
"loss": 0.7042,
"step": 887
},
{
"epoch": 0.8906720160481444,
"grad_norm": 5.054128646850586,
"learning_rate": 3.1589677397464435e-07,
"loss": 0.8196,
"step": 888
},
{
"epoch": 0.8916750250752257,
"grad_norm": 6.017844200134277,
"learning_rate": 3.102391834997143e-07,
"loss": 0.9218,
"step": 889
},
{
"epoch": 0.892678034102307,
"grad_norm": 4.536251068115234,
"learning_rate": 3.046310921027157e-07,
"loss": 0.6836,
"step": 890
},
{
"epoch": 0.8936810431293881,
"grad_norm": 4.6317596435546875,
"learning_rate": 2.990725589754406e-07,
"loss": 0.6691,
"step": 891
},
{
"epoch": 0.8946840521564694,
"grad_norm": 5.083393573760986,
"learning_rate": 2.9356364278660955e-07,
"loss": 0.7181,
"step": 892
},
{
"epoch": 0.8956870611835507,
"grad_norm": 4.133002281188965,
"learning_rate": 2.881044016812506e-07,
"loss": 0.734,
"step": 893
},
{
"epoch": 0.8966900702106319,
"grad_norm": 4.024208068847656,
"learning_rate": 2.826948932800844e-07,
"loss": 0.7544,
"step": 894
},
{
"epoch": 0.8976930792377131,
"grad_norm": 5.048941612243652,
"learning_rate": 2.773351746789182e-07,
"loss": 0.7579,
"step": 895
},
{
"epoch": 0.8986960882647944,
"grad_norm": 4.465607643127441,
"learning_rate": 2.720253024480418e-07,
"loss": 0.6304,
"step": 896
},
{
"epoch": 0.8996990972918756,
"grad_norm": 4.029648303985596,
"learning_rate": 2.6676533263163105e-07,
"loss": 0.7058,
"step": 897
},
{
"epoch": 0.9007021063189569,
"grad_norm": 5.086363315582275,
"learning_rate": 2.615553207471555e-07,
"loss": 0.686,
"step": 898
},
{
"epoch": 0.9017051153460381,
"grad_norm": 6.115987777709961,
"learning_rate": 2.563953217847942e-07,
"loss": 0.8888,
"step": 899
},
{
"epoch": 0.9027081243731193,
"grad_norm": 4.454765319824219,
"learning_rate": 2.5128539020685294e-07,
"loss": 0.7863,
"step": 900
},
{
"epoch": 0.9037111334002006,
"grad_norm": 4.63258171081543,
"learning_rate": 2.462255799471913e-07,
"loss": 0.6753,
"step": 901
},
{
"epoch": 0.9047141424272819,
"grad_norm": 3.8224151134490967,
"learning_rate": 2.412159444106543e-07,
"loss": 0.5568,
"step": 902
},
{
"epoch": 0.905717151454363,
"grad_norm": 4.1036601066589355,
"learning_rate": 2.3625653647250392e-07,
"loss": 0.7821,
"step": 903
},
{
"epoch": 0.9067201604814443,
"grad_norm": 8.105584144592285,
"learning_rate": 2.3134740847786717e-07,
"loss": 0.6691,
"step": 904
},
{
"epoch": 0.9077231695085256,
"grad_norm": 4.451644420623779,
"learning_rate": 2.2648861224117858e-07,
"loss": 0.7858,
"step": 905
},
{
"epoch": 0.9087261785356068,
"grad_norm": 5.970065116882324,
"learning_rate": 2.2168019904563686e-07,
"loss": 0.7558,
"step": 906
},
{
"epoch": 0.9097291875626881,
"grad_norm": 4.9913249015808105,
"learning_rate": 2.1692221964266125e-07,
"loss": 0.6912,
"step": 907
},
{
"epoch": 0.9107321965897693,
"grad_norm": 5.410460948944092,
"learning_rate": 2.122147242513578e-07,
"loss": 0.7068,
"step": 908
},
{
"epoch": 0.9117352056168505,
"grad_norm": 4.701066970825195,
"learning_rate": 2.075577625579872e-07,
"loss": 0.6389,
"step": 909
},
{
"epoch": 0.9127382146439318,
"grad_norm": 4.227931976318359,
"learning_rate": 2.0295138371544232e-07,
"loss": 0.6387,
"step": 910
},
{
"epoch": 0.9137412236710131,
"grad_norm": 3.9191296100616455,
"learning_rate": 1.9839563634272973e-07,
"loss": 0.7389,
"step": 911
},
{
"epoch": 0.9147442326980942,
"grad_norm": 2.9703845977783203,
"learning_rate": 1.9389056852445132e-07,
"loss": 0.7662,
"step": 912
},
{
"epoch": 0.9157472417251755,
"grad_norm": 4.562188148498535,
"learning_rate": 1.8943622781030568e-07,
"loss": 0.7359,
"step": 913
},
{
"epoch": 0.9167502507522568,
"grad_norm": 4.0336594581604,
"learning_rate": 1.8503266121457753e-07,
"loss": 0.7228,
"step": 914
},
{
"epoch": 0.917753259779338,
"grad_norm": 4.453738212585449,
"learning_rate": 1.8067991521564854e-07,
"loss": 0.6569,
"step": 915
},
{
"epoch": 0.9187562688064193,
"grad_norm": 4.996483325958252,
"learning_rate": 1.7637803575550115e-07,
"loss": 0.7717,
"step": 916
},
{
"epoch": 0.9197592778335005,
"grad_norm": 4.743129730224609,
"learning_rate": 1.7212706823923675e-07,
"loss": 0.8788,
"step": 917
},
{
"epoch": 0.9207622868605817,
"grad_norm": 5.228214263916016,
"learning_rate": 1.679270575345976e-07,
"loss": 0.7314,
"step": 918
},
{
"epoch": 0.921765295887663,
"grad_norm": 4.503567695617676,
"learning_rate": 1.6377804797148788e-07,
"loss": 0.6119,
"step": 919
},
{
"epoch": 0.9227683049147443,
"grad_norm": 4.443480491638184,
"learning_rate": 1.5968008334151351e-07,
"loss": 0.8009,
"step": 920
},
{
"epoch": 0.9237713139418254,
"grad_norm": 3.952268362045288,
"learning_rate": 1.5563320689751194e-07,
"loss": 0.6124,
"step": 921
},
{
"epoch": 0.9247743229689067,
"grad_norm": 5.096187591552734,
"learning_rate": 1.5163746135310186e-07,
"loss": 0.7673,
"step": 922
},
{
"epoch": 0.925777331995988,
"grad_norm": 4.789193153381348,
"learning_rate": 1.4769288888222988e-07,
"loss": 0.7439,
"step": 923
},
{
"epoch": 0.9267803410230692,
"grad_norm": 3.701528310775757,
"learning_rate": 1.4379953111872458e-07,
"loss": 0.7515,
"step": 924
},
{
"epoch": 0.9277833500501504,
"grad_norm": 6.08350133895874,
"learning_rate": 1.399574291558581e-07,
"loss": 0.6835,
"step": 925
},
{
"epoch": 0.9287863590772317,
"grad_norm": 4.917857646942139,
"learning_rate": 1.361666235459136e-07,
"loss": 0.7786,
"step": 926
},
{
"epoch": 0.9297893681043129,
"grad_norm": 3.9800169467926025,
"learning_rate": 1.3242715429975517e-07,
"loss": 0.7483,
"step": 927
},
{
"epoch": 0.9307923771313942,
"grad_norm": 5.71744966506958,
"learning_rate": 1.2873906088640476e-07,
"loss": 0.8735,
"step": 928
},
{
"epoch": 0.9317953861584755,
"grad_norm": 4.791073799133301,
"learning_rate": 1.2510238223263084e-07,
"loss": 0.7336,
"step": 929
},
{
"epoch": 0.9327983951855566,
"grad_norm": 3.646848678588867,
"learning_rate": 1.2151715672252985e-07,
"loss": 0.8121,
"step": 930
},
{
"epoch": 0.9338014042126379,
"grad_norm": 4.681507587432861,
"learning_rate": 1.1798342219712822e-07,
"loss": 0.7552,
"step": 931
},
{
"epoch": 0.9348044132397192,
"grad_norm": 4.455327987670898,
"learning_rate": 1.1450121595397712e-07,
"loss": 0.8217,
"step": 932
},
{
"epoch": 0.9358074222668004,
"grad_norm": 4.647828102111816,
"learning_rate": 1.1107057474676441e-07,
"loss": 0.6471,
"step": 933
},
{
"epoch": 0.9368104312938816,
"grad_norm": 4.541031837463379,
"learning_rate": 1.0769153478492112e-07,
"loss": 0.6941,
"step": 934
},
{
"epoch": 0.9378134403209629,
"grad_norm": 4.167369842529297,
"learning_rate": 1.0436413173324388e-07,
"loss": 0.7649,
"step": 935
},
{
"epoch": 0.9388164493480441,
"grad_norm": 4.509149074554443,
"learning_rate": 1.0108840071151648e-07,
"loss": 0.7033,
"step": 936
},
{
"epoch": 0.9398194583751254,
"grad_norm": 4.177834510803223,
"learning_rate": 9.78643762941367e-08,
"loss": 0.8545,
"step": 937
},
{
"epoch": 0.9408224674022067,
"grad_norm": 4.97866678237915,
"learning_rate": 9.469209250975774e-08,
"loss": 0.6118,
"step": 938
},
{
"epoch": 0.9418254764292878,
"grad_norm": 3.9421169757843018,
"learning_rate": 9.157158284092249e-08,
"loss": 0.803,
"step": 939
},
{
"epoch": 0.9428284854563691,
"grad_norm": 3.787537097930908,
"learning_rate": 8.850288022371478e-08,
"loss": 0.6872,
"step": 940
},
{
"epoch": 0.9438314944834504,
"grad_norm": 5.575892925262451,
"learning_rate": 8.548601704740756e-08,
"loss": 0.7377,
"step": 941
},
{
"epoch": 0.9448345035105316,
"grad_norm": 4.359327793121338,
"learning_rate": 8.25210251541264e-08,
"loss": 0.735,
"step": 942
},
{
"epoch": 0.9458375125376128,
"grad_norm": 4.343613624572754,
"learning_rate": 7.960793583850768e-08,
"loss": 0.7444,
"step": 943
},
{
"epoch": 0.9468405215646941,
"grad_norm": 4.57480525970459,
"learning_rate": 7.674677984737255e-08,
"loss": 0.6987,
"step": 944
},
{
"epoch": 0.9478435305917753,
"grad_norm": 5.964742183685303,
"learning_rate": 7.393758737940127e-08,
"loss": 0.8617,
"step": 945
},
{
"epoch": 0.9488465396188566,
"grad_norm": 4.881872177124023,
"learning_rate": 7.118038808481165e-08,
"loss": 0.6689,
"step": 946
},
{
"epoch": 0.9498495486459378,
"grad_norm": 5.220876216888428,
"learning_rate": 6.847521106505107e-08,
"loss": 0.7263,
"step": 947
},
{
"epoch": 0.950852557673019,
"grad_norm": 3.770507335662842,
"learning_rate": 6.582208487248498e-08,
"loss": 0.6386,
"step": 948
},
{
"epoch": 0.9518555667001003,
"grad_norm": 4.965149402618408,
"learning_rate": 6.322103751009834e-08,
"loss": 0.6863,
"step": 949
},
{
"epoch": 0.9528585757271816,
"grad_norm": 4.452226161956787,
"learning_rate": 6.06720964311991e-08,
"loss": 0.6324,
"step": 950
},
{
"epoch": 0.9538615847542627,
"grad_norm": 4.892546653747559,
"learning_rate": 5.817528853912735e-08,
"loss": 0.8314,
"step": 951
},
{
"epoch": 0.954864593781344,
"grad_norm": 3.985069513320923,
"learning_rate": 5.5730640186973936e-08,
"loss": 0.6095,
"step": 952
},
{
"epoch": 0.9558676028084253,
"grad_norm": 5.226090908050537,
"learning_rate": 5.333817717729894e-08,
"loss": 0.8446,
"step": 953
},
{
"epoch": 0.9568706118355065,
"grad_norm": 3.97580885887146,
"learning_rate": 5.0997924761862494e-08,
"loss": 0.5942,
"step": 954
},
{
"epoch": 0.9578736208625878,
"grad_norm": 5.360856533050537,
"learning_rate": 4.8709907641355524e-08,
"loss": 0.757,
"step": 955
},
{
"epoch": 0.958876629889669,
"grad_norm": 5.256042957305908,
"learning_rate": 4.6474149965142766e-08,
"loss": 0.6407,
"step": 956
},
{
"epoch": 0.9598796389167502,
"grad_norm": 3.9611828327178955,
"learning_rate": 4.4290675331002954e-08,
"loss": 0.8301,
"step": 957
},
{
"epoch": 0.9608826479438315,
"grad_norm": 5.468574047088623,
"learning_rate": 4.2159506784884005e-08,
"loss": 0.8779,
"step": 958
},
{
"epoch": 0.9618856569709128,
"grad_norm": 3.9516358375549316,
"learning_rate": 4.008066682065714e-08,
"loss": 0.7016,
"step": 959
},
{
"epoch": 0.9628886659979939,
"grad_norm": 4.945855140686035,
"learning_rate": 3.8054177379881483e-08,
"loss": 0.7165,
"step": 960
},
{
"epoch": 0.9638916750250752,
"grad_norm": 5.862647533416748,
"learning_rate": 3.608005985157037e-08,
"loss": 0.8532,
"step": 961
},
{
"epoch": 0.9648946840521565,
"grad_norm": 4.712942600250244,
"learning_rate": 3.415833507196764e-08,
"loss": 0.8651,
"step": 962
},
{
"epoch": 0.9658976930792377,
"grad_norm": 4.585020065307617,
"learning_rate": 3.2289023324325596e-08,
"loss": 0.5557,
"step": 963
},
{
"epoch": 0.966900702106319,
"grad_norm": 3.7760512828826904,
"learning_rate": 3.047214433869239e-08,
"loss": 0.8373,
"step": 964
},
{
"epoch": 0.9679037111334002,
"grad_norm": 4.538245677947998,
"learning_rate": 2.8707717291704406e-08,
"loss": 0.6343,
"step": 965
},
{
"epoch": 0.9689067201604814,
"grad_norm": 4.196023941040039,
"learning_rate": 2.6995760806382e-08,
"loss": 0.735,
"step": 966
},
{
"epoch": 0.9699097291875627,
"grad_norm": 4.290002346038818,
"learning_rate": 2.5336292951933517e-08,
"loss": 0.7128,
"step": 967
},
{
"epoch": 0.970912738214644,
"grad_norm": 5.1577043533325195,
"learning_rate": 2.372933124356602e-08,
"loss": 0.745,
"step": 968
},
{
"epoch": 0.9719157472417251,
"grad_norm": 5.1396989822387695,
"learning_rate": 2.217489264229822e-08,
"loss": 0.8372,
"step": 969
},
{
"epoch": 0.9729187562688064,
"grad_norm": 4.005364894866943,
"learning_rate": 2.067299355478336e-08,
"loss": 0.7098,
"step": 970
},
{
"epoch": 0.9739217652958877,
"grad_norm": 4.569650650024414,
"learning_rate": 1.922364983313385e-08,
"loss": 0.7347,
"step": 971
},
{
"epoch": 0.9749247743229689,
"grad_norm": 3.6062047481536865,
"learning_rate": 1.7826876774757474e-08,
"loss": 0.8476,
"step": 972
},
{
"epoch": 0.9759277833500501,
"grad_norm": 5.83583927154541,
"learning_rate": 1.648268912219142e-08,
"loss": 0.6966,
"step": 973
},
{
"epoch": 0.9769307923771314,
"grad_norm": 3.9248313903808594,
"learning_rate": 1.5191101062950188e-08,
"loss": 0.7599,
"step": 974
},
{
"epoch": 0.9779338014042126,
"grad_norm": 3.60282301902771,
"learning_rate": 1.3952126229375695e-08,
"loss": 0.6102,
"step": 975
},
{
"epoch": 0.9789368104312939,
"grad_norm": 4.640085697174072,
"learning_rate": 1.2765777698490189e-08,
"loss": 0.6826,
"step": 976
},
{
"epoch": 0.9799398194583752,
"grad_norm": 4.5401291847229,
"learning_rate": 1.163206799186245e-08,
"loss": 0.7868,
"step": 977
},
{
"epoch": 0.9809428284854563,
"grad_norm": 4.223124980926514,
"learning_rate": 1.0551009075471796e-08,
"loss": 0.6972,
"step": 978
},
{
"epoch": 0.9819458375125376,
"grad_norm": 4.554103851318359,
"learning_rate": 9.522612359585404e-09,
"loss": 0.7147,
"step": 979
},
{
"epoch": 0.9829488465396189,
"grad_norm": 4.4233622550964355,
"learning_rate": 8.546888698634515e-09,
"loss": 0.769,
"step": 980
},
{
"epoch": 0.9839518555667001,
"grad_norm": 3.9840776920318604,
"learning_rate": 7.623848391102307e-09,
"loss": 0.67,
"step": 981
},
{
"epoch": 0.9849548645937813,
"grad_norm": 4.4340500831604,
"learning_rate": 6.7535011794134245e-09,
"loss": 0.7422,
"step": 982
},
{
"epoch": 0.9859578736208626,
"grad_norm": 5.257681846618652,
"learning_rate": 5.935856249833505e-09,
"loss": 0.7816,
"step": 983
},
{
"epoch": 0.9869608826479438,
"grad_norm": 4.62564754486084,
"learning_rate": 5.170922232369258e-09,
"loss": 0.6533,
"step": 984
},
{
"epoch": 0.9879638916750251,
"grad_norm": 5.749030590057373,
"learning_rate": 4.458707200679646e-09,
"loss": 0.8785,
"step": 985
},
{
"epoch": 0.9889669007021064,
"grad_norm": 4.4205756187438965,
"learning_rate": 3.799218671989291e-09,
"loss": 0.68,
"step": 986
},
{
"epoch": 0.9899699097291875,
"grad_norm": 4.471177101135254,
"learning_rate": 3.192463607010754e-09,
"loss": 0.7755,
"step": 987
},
{
"epoch": 0.9909729187562688,
"grad_norm": 3.8128039836883545,
"learning_rate": 2.638448409869043e-09,
"loss": 0.6275,
"step": 988
},
{
"epoch": 0.9919759277833501,
"grad_norm": 4.73145055770874,
"learning_rate": 2.1371789280355547e-09,
"loss": 0.5856,
"step": 989
},
{
"epoch": 0.9929789368104313,
"grad_norm": 4.149558067321777,
"learning_rate": 1.6886604522659e-09,
"loss": 0.8192,
"step": 990
},
{
"epoch": 0.9939819458375125,
"grad_norm": 5.041884422302246,
"learning_rate": 1.2928977165427293e-09,
"loss": 0.8046,
"step": 991
},
{
"epoch": 0.9949849548645938,
"grad_norm": 5.230462551116943,
"learning_rate": 9.498948980291022e-10,
"loss": 0.8453,
"step": 992
},
{
"epoch": 0.995987963891675,
"grad_norm": 4.824460983276367,
"learning_rate": 6.596556170207491e-10,
"loss": 0.7198,
"step": 993
},
{
"epoch": 0.9969909729187563,
"grad_norm": 4.745485305786133,
"learning_rate": 4.221829369094321e-10,
"loss": 0.6268,
"step": 994
},
{
"epoch": 0.9979939819458375,
"grad_norm": 4.974123477935791,
"learning_rate": 2.3747936415186024e-10,
"loss": 0.8355,
"step": 995
},
{
"epoch": 0.9989969909729187,
"grad_norm": 5.097357273101807,
"learning_rate": 1.0554684824137795e-10,
"loss": 0.6945,
"step": 996
},
{
"epoch": 1.0,
"grad_norm": 5.223090171813965,
"learning_rate": 2.6386781686316144e-11,
"loss": 0.584,
"step": 997
},
{
"epoch": 1.0,
"step": 997,
"total_flos": 2.3571596710531564e+18,
"train_loss": 1.012767364172902,
"train_runtime": 25597.8291,
"train_samples_per_second": 0.623,
"train_steps_per_second": 0.039
}
],
"logging_steps": 1.0,
"max_steps": 997,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.3571596710531564e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}