openorca / trainer_state.json
ciyazzzk
First model version
2efa347
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9938573003622615,
"eval_steps": 500,
"global_step": 297,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.010080327610647345,
"grad_norm": 20.48233413696289,
"learning_rate": 6.666666666666667e-07,
"loss": 1.318,
"step": 1
},
{
"epoch": 0.02016065522129469,
"grad_norm": 22.18445587158203,
"learning_rate": 1.3333333333333334e-06,
"loss": 1.2034,
"step": 2
},
{
"epoch": 0.03024098283194204,
"grad_norm": 20.34792137145996,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.0455,
"step": 3
},
{
"epoch": 0.04032131044258938,
"grad_norm": 8.893705368041992,
"learning_rate": 2.666666666666667e-06,
"loss": 2.0515,
"step": 4
},
{
"epoch": 0.05040163805323673,
"grad_norm": 6.274827003479004,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.0793,
"step": 5
},
{
"epoch": 0.06048196566388408,
"grad_norm": 6.051919937133789,
"learning_rate": 4.000000000000001e-06,
"loss": 0.9434,
"step": 6
},
{
"epoch": 0.07056229327453142,
"grad_norm": 7.502919673919678,
"learning_rate": 4.666666666666667e-06,
"loss": 0.9741,
"step": 7
},
{
"epoch": 0.08064262088517876,
"grad_norm": 6.437217712402344,
"learning_rate": 5.333333333333334e-06,
"loss": 1.1227,
"step": 8
},
{
"epoch": 0.09072294849582611,
"grad_norm": 5.169360160827637,
"learning_rate": 6e-06,
"loss": 0.9156,
"step": 9
},
{
"epoch": 0.10080327610647347,
"grad_norm": 5.1474432945251465,
"learning_rate": 6.666666666666667e-06,
"loss": 1.0925,
"step": 10
},
{
"epoch": 0.1108836037171208,
"grad_norm": 6.759896755218506,
"learning_rate": 7.333333333333333e-06,
"loss": 0.8985,
"step": 11
},
{
"epoch": 0.12096393132776816,
"grad_norm": 5.231770992279053,
"learning_rate": 8.000000000000001e-06,
"loss": 1.6954,
"step": 12
},
{
"epoch": 0.1310442589384155,
"grad_norm": 5.028665542602539,
"learning_rate": 8.666666666666668e-06,
"loss": 0.9763,
"step": 13
},
{
"epoch": 0.14112458654906285,
"grad_norm": 7.008236408233643,
"learning_rate": 9.333333333333334e-06,
"loss": 0.9969,
"step": 14
},
{
"epoch": 0.15120491415971019,
"grad_norm": 4.675139904022217,
"learning_rate": 1e-05,
"loss": 0.8619,
"step": 15
},
{
"epoch": 0.16128524177035752,
"grad_norm": 5.249491214752197,
"learning_rate": 1.0666666666666667e-05,
"loss": 0.9176,
"step": 16
},
{
"epoch": 0.1713655693810049,
"grad_norm": 4.5402092933654785,
"learning_rate": 1.1333333333333334e-05,
"loss": 0.8809,
"step": 17
},
{
"epoch": 0.18144589699165223,
"grad_norm": 4.799923896789551,
"learning_rate": 1.2e-05,
"loss": 1.005,
"step": 18
},
{
"epoch": 0.19152622460229957,
"grad_norm": 3.9302682876586914,
"learning_rate": 1.2666666666666667e-05,
"loss": 0.8784,
"step": 19
},
{
"epoch": 0.20160655221294693,
"grad_norm": 4.542870044708252,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.0378,
"step": 20
},
{
"epoch": 0.21168687982359427,
"grad_norm": 10.4898042678833,
"learning_rate": 1.4e-05,
"loss": 0.8458,
"step": 21
},
{
"epoch": 0.2217672074342416,
"grad_norm": 4.678144454956055,
"learning_rate": 1.4666666666666666e-05,
"loss": 0.9237,
"step": 22
},
{
"epoch": 0.23184753504488895,
"grad_norm": 4.897671222686768,
"learning_rate": 1.5333333333333334e-05,
"loss": 0.9661,
"step": 23
},
{
"epoch": 0.2419278626555363,
"grad_norm": 5.067351818084717,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.9975,
"step": 24
},
{
"epoch": 0.25200819026618365,
"grad_norm": 3.7107274532318115,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.9316,
"step": 25
},
{
"epoch": 0.262088517876831,
"grad_norm": 4.553698539733887,
"learning_rate": 1.7333333333333336e-05,
"loss": 0.871,
"step": 26
},
{
"epoch": 0.27216884548747833,
"grad_norm": 5.10447359085083,
"learning_rate": 1.8e-05,
"loss": 0.8687,
"step": 27
},
{
"epoch": 0.2822491730981257,
"grad_norm": 3.7116832733154297,
"learning_rate": 1.866666666666667e-05,
"loss": 0.8586,
"step": 28
},
{
"epoch": 0.29232950070877306,
"grad_norm": 5.299854755401611,
"learning_rate": 1.9333333333333333e-05,
"loss": 0.8595,
"step": 29
},
{
"epoch": 0.30240982831942037,
"grad_norm": 3.825899600982666,
"learning_rate": 2e-05,
"loss": 0.9615,
"step": 30
},
{
"epoch": 0.31249015593006774,
"grad_norm": 4.806526184082031,
"learning_rate": 1.999930778307066e-05,
"loss": 0.8595,
"step": 31
},
{
"epoch": 0.32257048354071505,
"grad_norm": 3.5444633960723877,
"learning_rate": 1.9997231228115487e-05,
"loss": 0.9748,
"step": 32
},
{
"epoch": 0.3326508111513624,
"grad_norm": 4.36836051940918,
"learning_rate": 1.9993770622619784e-05,
"loss": 0.8577,
"step": 33
},
{
"epoch": 0.3427311387620098,
"grad_norm": 3.0740671157836914,
"learning_rate": 1.9988926445681495e-05,
"loss": 0.8407,
"step": 34
},
{
"epoch": 0.3528114663726571,
"grad_norm": 3.8326187133789062,
"learning_rate": 1.998269936794487e-05,
"loss": 0.8997,
"step": 35
},
{
"epoch": 0.36289179398330446,
"grad_norm": 3.6133008003234863,
"learning_rate": 1.9975090251507637e-05,
"loss": 0.9572,
"step": 36
},
{
"epoch": 0.3729721215939518,
"grad_norm": 4.111402988433838,
"learning_rate": 1.9966100149801648e-05,
"loss": 0.8465,
"step": 37
},
{
"epoch": 0.38305244920459913,
"grad_norm": 3.105464220046997,
"learning_rate": 1.9955730307447015e-05,
"loss": 0.84,
"step": 38
},
{
"epoch": 0.3931327768152465,
"grad_norm": 3.377089738845825,
"learning_rate": 1.9943982160079823e-05,
"loss": 0.977,
"step": 39
},
{
"epoch": 0.40321310442589386,
"grad_norm": 3.674912214279175,
"learning_rate": 1.9930857334153374e-05,
"loss": 0.9114,
"step": 40
},
{
"epoch": 0.4132934320365412,
"grad_norm": 3.491791248321533,
"learning_rate": 1.9916357646713006e-05,
"loss": 0.8507,
"step": 41
},
{
"epoch": 0.42337375964718854,
"grad_norm": 3.5988316535949707,
"learning_rate": 1.9900485105144544e-05,
"loss": 0.8459,
"step": 42
},
{
"epoch": 0.4334540872578359,
"grad_norm": 3.147287368774414,
"learning_rate": 1.988324190689639e-05,
"loss": 0.9254,
"step": 43
},
{
"epoch": 0.4435344148684832,
"grad_norm": 3.4546704292297363,
"learning_rate": 1.9864630439175282e-05,
"loss": 0.9388,
"step": 44
},
{
"epoch": 0.4536147424791306,
"grad_norm": 3.39437198638916,
"learning_rate": 1.9844653278615836e-05,
"loss": 0.8751,
"step": 45
},
{
"epoch": 0.4636950700897779,
"grad_norm": 2.966585159301758,
"learning_rate": 1.9823313190923797e-05,
"loss": 0.833,
"step": 46
},
{
"epoch": 0.47377539770042526,
"grad_norm": 6.668085098266602,
"learning_rate": 1.9800613130493158e-05,
"loss": 0.9399,
"step": 47
},
{
"epoch": 0.4838557253110726,
"grad_norm": 4.198956489562988,
"learning_rate": 1.9776556239997146e-05,
"loss": 0.8604,
"step": 48
},
{
"epoch": 0.49393605292171994,
"grad_norm": 3.0325896739959717,
"learning_rate": 1.9751145849953135e-05,
"loss": 0.8399,
"step": 49
},
{
"epoch": 0.5040163805323673,
"grad_norm": 3.0284290313720703,
"learning_rate": 1.972438547826156e-05,
"loss": 0.962,
"step": 50
},
{
"epoch": 0.5140967081430147,
"grad_norm": 3.7126681804656982,
"learning_rate": 1.9696278829718882e-05,
"loss": 0.8381,
"step": 51
},
{
"epoch": 0.524177035753662,
"grad_norm": 3.0753326416015625,
"learning_rate": 1.9666829795504693e-05,
"loss": 1.2808,
"step": 52
},
{
"epoch": 0.5342573633643093,
"grad_norm": 3.3041043281555176,
"learning_rate": 1.9636042452643004e-05,
"loss": 1.0921,
"step": 53
},
{
"epoch": 0.5443376909749567,
"grad_norm": 2.974684953689575,
"learning_rate": 1.9603921063437795e-05,
"loss": 1.1766,
"step": 54
},
{
"epoch": 0.554418018585604,
"grad_norm": 3.104491710662842,
"learning_rate": 1.9570470074882947e-05,
"loss": 0.8245,
"step": 55
},
{
"epoch": 0.5644983461962514,
"grad_norm": 2.8233213424682617,
"learning_rate": 1.9535694118046584e-05,
"loss": 0.8327,
"step": 56
},
{
"epoch": 0.5745786738068988,
"grad_norm": 3.0855748653411865,
"learning_rate": 1.949959800742991e-05,
"loss": 0.8333,
"step": 57
},
{
"epoch": 0.5846590014175461,
"grad_norm": 3.309098482131958,
"learning_rate": 1.9462186740300697e-05,
"loss": 0.8437,
"step": 58
},
{
"epoch": 0.5947393290281934,
"grad_norm": 3.7956173419952393,
"learning_rate": 1.942346549600144e-05,
"loss": 0.8533,
"step": 59
},
{
"epoch": 0.6048196566388407,
"grad_norm": 2.885507106781006,
"learning_rate": 1.9383439635232296e-05,
"loss": 0.9791,
"step": 60
},
{
"epoch": 0.6148999842494881,
"grad_norm": 2.976921319961548,
"learning_rate": 1.9342114699308962e-05,
"loss": 0.9537,
"step": 61
},
{
"epoch": 0.6249803118601355,
"grad_norm": 3.9822583198547363,
"learning_rate": 1.9299496409395482e-05,
"loss": 0.8513,
"step": 62
},
{
"epoch": 0.6350606394707828,
"grad_norm": 3.2742626667022705,
"learning_rate": 1.9255590665712214e-05,
"loss": 0.8606,
"step": 63
},
{
"epoch": 0.6451409670814301,
"grad_norm": 2.989588975906372,
"learning_rate": 1.921040354671897e-05,
"loss": 1.0236,
"step": 64
},
{
"epoch": 0.6552212946920775,
"grad_norm": 3.3849992752075195,
"learning_rate": 1.9163941308273504e-05,
"loss": 0.8274,
"step": 65
},
{
"epoch": 0.6653016223027248,
"grad_norm": 2.9485199451446533,
"learning_rate": 1.911621038276542e-05,
"loss": 0.8366,
"step": 66
},
{
"epoch": 0.6753819499133722,
"grad_norm": 2.8090484142303467,
"learning_rate": 1.9067217378225655e-05,
"loss": 1.0603,
"step": 67
},
{
"epoch": 0.6854622775240196,
"grad_norm": 2.795466899871826,
"learning_rate": 1.9016969077411645e-05,
"loss": 1.0391,
"step": 68
},
{
"epoch": 0.6955426051346669,
"grad_norm": 3.0513088703155518,
"learning_rate": 1.8965472436868288e-05,
"loss": 0.9469,
"step": 69
},
{
"epoch": 0.7056229327453142,
"grad_norm": 2.89764404296875,
"learning_rate": 1.891273458596486e-05,
"loss": 0.8358,
"step": 70
},
{
"epoch": 0.7157032603559615,
"grad_norm": 2.940281629562378,
"learning_rate": 1.8858762825908e-05,
"loss": 0.9117,
"step": 71
},
{
"epoch": 0.7257835879666089,
"grad_norm": 3.0357506275177,
"learning_rate": 1.8803564628730916e-05,
"loss": 0.8441,
"step": 72
},
{
"epoch": 0.7358639155772563,
"grad_norm": 3.0957512855529785,
"learning_rate": 1.874714763625892e-05,
"loss": 0.8154,
"step": 73
},
{
"epoch": 0.7459442431879036,
"grad_norm": 3.3382112979888916,
"learning_rate": 1.8689519659051467e-05,
"loss": 1.0091,
"step": 74
},
{
"epoch": 0.7560245707985509,
"grad_norm": 2.7359678745269775,
"learning_rate": 1.8630688675320844e-05,
"loss": 0.9901,
"step": 75
},
{
"epoch": 0.7661048984091983,
"grad_norm": 2.9160029888153076,
"learning_rate": 1.8570662829827632e-05,
"loss": 1.0645,
"step": 76
},
{
"epoch": 0.7761852260198456,
"grad_norm": 3.7096657752990723,
"learning_rate": 1.8509450432753123e-05,
"loss": 0.8458,
"step": 77
},
{
"epoch": 0.786265553630493,
"grad_norm": 2.9605114459991455,
"learning_rate": 1.8447059958548822e-05,
"loss": 0.8315,
"step": 78
},
{
"epoch": 0.7963458812411404,
"grad_norm": 3.1716909408569336,
"learning_rate": 1.8383500044763226e-05,
"loss": 0.8427,
"step": 79
},
{
"epoch": 0.8064262088517877,
"grad_norm": 4.014035224914551,
"learning_rate": 1.8318779490846005e-05,
"loss": 0.8391,
"step": 80
},
{
"epoch": 0.816506536462435,
"grad_norm": 2.6693358421325684,
"learning_rate": 1.8252907256929777e-05,
"loss": 1.0444,
"step": 81
},
{
"epoch": 0.8265868640730824,
"grad_norm": 3.824836254119873,
"learning_rate": 1.818589246258964e-05,
"loss": 0.9738,
"step": 82
},
{
"epoch": 0.8366671916837297,
"grad_norm": 3.0832369327545166,
"learning_rate": 1.8117744385580627e-05,
"loss": 0.9109,
"step": 83
},
{
"epoch": 0.8467475192943771,
"grad_norm": 3.707331657409668,
"learning_rate": 1.804847246055326e-05,
"loss": 0.8324,
"step": 84
},
{
"epoch": 0.8568278469050244,
"grad_norm": 3.3008220195770264,
"learning_rate": 1.797808627774738e-05,
"loss": 0.9383,
"step": 85
},
{
"epoch": 0.8669081745156718,
"grad_norm": 3.1887435913085938,
"learning_rate": 1.7906595581664462e-05,
"loss": 0.8441,
"step": 86
},
{
"epoch": 0.8769885021263191,
"grad_norm": 2.8672823905944824,
"learning_rate": 1.7834010269718526e-05,
"loss": 0.8353,
"step": 87
},
{
"epoch": 0.8870688297369664,
"grad_norm": 3.0346055030822754,
"learning_rate": 1.776034039086592e-05,
"loss": 1.348,
"step": 88
},
{
"epoch": 0.8971491573476138,
"grad_norm": 3.2000229358673096,
"learning_rate": 1.768559614421411e-05,
"loss": 0.8544,
"step": 89
},
{
"epoch": 0.9072294849582612,
"grad_norm": 3.007753610610962,
"learning_rate": 1.7609787877609678e-05,
"loss": 0.8505,
"step": 90
},
{
"epoch": 0.9173098125689085,
"grad_norm": 2.9321534633636475,
"learning_rate": 1.753292608620573e-05,
"loss": 0.8402,
"step": 91
},
{
"epoch": 0.9273901401795558,
"grad_norm": 2.6721303462982178,
"learning_rate": 1.7455021411008906e-05,
"loss": 0.8421,
"step": 92
},
{
"epoch": 0.9374704677902032,
"grad_norm": 3.0125327110290527,
"learning_rate": 1.7376084637406222e-05,
"loss": 0.8443,
"step": 93
},
{
"epoch": 0.9475507954008505,
"grad_norm": 2.6522045135498047,
"learning_rate": 1.7296126693671886e-05,
"loss": 1.2249,
"step": 94
},
{
"epoch": 0.9576311230114979,
"grad_norm": 2.6742281913757324,
"learning_rate": 1.721515864945435e-05,
"loss": 0.846,
"step": 95
},
{
"epoch": 0.9677114506221453,
"grad_norm": 2.9093592166900635,
"learning_rate": 1.7133191714243805e-05,
"loss": 0.9391,
"step": 96
},
{
"epoch": 0.9777917782327926,
"grad_norm": 3.675670623779297,
"learning_rate": 1.7050237235820287e-05,
"loss": 0.8723,
"step": 97
},
{
"epoch": 0.9878721058434399,
"grad_norm": 2.698991298675537,
"learning_rate": 1.6966306698682672e-05,
"loss": 0.8491,
"step": 98
},
{
"epoch": 0.9979524334540872,
"grad_norm": 2.8708646297454834,
"learning_rate": 1.6881411722458688e-05,
"loss": 1.1059,
"step": 99
},
{
"epoch": 1.0080327610647346,
"grad_norm": 3.9162817001342773,
"learning_rate": 1.6795564060296295e-05,
"loss": 0.711,
"step": 100
},
{
"epoch": 1.018113088675382,
"grad_norm": 3.2829103469848633,
"learning_rate": 1.6708775597236507e-05,
"loss": 0.7179,
"step": 101
},
{
"epoch": 1.0281934162860293,
"grad_norm": 4.782021999359131,
"learning_rate": 1.6621058348568008e-05,
"loss": 0.759,
"step": 102
},
{
"epoch": 1.0382737438966767,
"grad_norm": 4.203643798828125,
"learning_rate": 1.6532424458163692e-05,
"loss": 0.7717,
"step": 103
},
{
"epoch": 1.048354071507324,
"grad_norm": 4.421259880065918,
"learning_rate": 1.6442886196799465e-05,
"loss": 0.651,
"step": 104
},
{
"epoch": 1.0584343991179714,
"grad_norm": 3.6463119983673096,
"learning_rate": 1.6352455960455385e-05,
"loss": 0.7719,
"step": 105
},
{
"epoch": 1.0685147267286186,
"grad_norm": 3.408778429031372,
"learning_rate": 1.6261146268599564e-05,
"loss": 0.6591,
"step": 106
},
{
"epoch": 1.078595054339266,
"grad_norm": 3.7105519771575928,
"learning_rate": 1.6168969762454897e-05,
"loss": 0.6645,
"step": 107
},
{
"epoch": 1.0886753819499133,
"grad_norm": 3.8667266368865967,
"learning_rate": 1.607593920324899e-05,
"loss": 0.7159,
"step": 108
},
{
"epoch": 1.0987557095605607,
"grad_norm": 3.4244542121887207,
"learning_rate": 1.598206747044746e-05,
"loss": 0.8331,
"step": 109
},
{
"epoch": 1.108836037171208,
"grad_norm": 3.326638698577881,
"learning_rate": 1.5887367559970825e-05,
"loss": 0.6831,
"step": 110
},
{
"epoch": 1.1189163647818554,
"grad_norm": 3.7026546001434326,
"learning_rate": 1.5791852582395334e-05,
"loss": 0.6642,
"step": 111
},
{
"epoch": 1.1289966923925028,
"grad_norm": 3.1158483028411865,
"learning_rate": 1.569553576113789e-05,
"loss": 0.6531,
"step": 112
},
{
"epoch": 1.1390770200031501,
"grad_norm": 3.1516995429992676,
"learning_rate": 1.5598430430625335e-05,
"loss": 0.6734,
"step": 113
},
{
"epoch": 1.1491573476137975,
"grad_norm": 3.048830270767212,
"learning_rate": 1.5500550034448415e-05,
"loss": 0.6412,
"step": 114
},
{
"epoch": 1.1592376752244449,
"grad_norm": 2.9057071208953857,
"learning_rate": 1.540190812350059e-05,
"loss": 0.6441,
"step": 115
},
{
"epoch": 1.1693180028350922,
"grad_norm": 4.930371284484863,
"learning_rate": 1.5302518354101992e-05,
"loss": 0.6499,
"step": 116
},
{
"epoch": 1.1793983304457396,
"grad_norm": 3.2132606506347656,
"learning_rate": 1.5202394486108823e-05,
"loss": 0.7648,
"step": 117
},
{
"epoch": 1.1894786580563868,
"grad_norm": 3.247512102127075,
"learning_rate": 1.5101550381008377e-05,
"loss": 0.6341,
"step": 118
},
{
"epoch": 1.1995589856670341,
"grad_norm": 2.8837008476257324,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.6785,
"step": 119
},
{
"epoch": 1.2096393132776815,
"grad_norm": 3.362884998321533,
"learning_rate": 1.4897757402062285e-05,
"loss": 0.6433,
"step": 120
},
{
"epoch": 1.2197196408883288,
"grad_norm": 3.0538456439971924,
"learning_rate": 1.4794836742006667e-05,
"loss": 0.7454,
"step": 121
},
{
"epoch": 1.2297999684989762,
"grad_norm": 3.1768131256103516,
"learning_rate": 1.4691252268517794e-05,
"loss": 0.7879,
"step": 122
},
{
"epoch": 1.2398802961096236,
"grad_norm": 3.0788283348083496,
"learning_rate": 1.4587018322180906e-05,
"loss": 0.7103,
"step": 123
},
{
"epoch": 1.249960623720271,
"grad_norm": 2.7563843727111816,
"learning_rate": 1.4482149333496455e-05,
"loss": 0.6592,
"step": 124
},
{
"epoch": 1.2600409513309183,
"grad_norm": 3.421998977661133,
"learning_rate": 1.4376659820882308e-05,
"loss": 0.7862,
"step": 125
},
{
"epoch": 1.2701212789415657,
"grad_norm": 3.0229763984680176,
"learning_rate": 1.4270564388663761e-05,
"loss": 0.6586,
"step": 126
},
{
"epoch": 1.2802016065522128,
"grad_norm": 2.8393478393554688,
"learning_rate": 1.4163877725051677e-05,
"loss": 0.7359,
"step": 127
},
{
"epoch": 1.2902819341628602,
"grad_norm": 3.009399175643921,
"learning_rate": 1.4056614600108998e-05,
"loss": 0.6755,
"step": 128
},
{
"epoch": 1.3003622617735076,
"grad_norm": 2.9821226596832275,
"learning_rate": 1.3948789863705914e-05,
"loss": 0.629,
"step": 129
},
{
"epoch": 1.310442589384155,
"grad_norm": 3.1034419536590576,
"learning_rate": 1.3840418443464015e-05,
"loss": 0.6466,
"step": 130
},
{
"epoch": 1.3205229169948023,
"grad_norm": 3.318528413772583,
"learning_rate": 1.3731515342689654e-05,
"loss": 1.0047,
"step": 131
},
{
"epoch": 1.3306032446054497,
"grad_norm": 2.9240915775299072,
"learning_rate": 1.3622095638296827e-05,
"loss": 0.668,
"step": 132
},
{
"epoch": 1.340683572216097,
"grad_norm": 3.9342963695526123,
"learning_rate": 1.3512174478719896e-05,
"loss": 0.6465,
"step": 133
},
{
"epoch": 1.3507638998267444,
"grad_norm": 2.6329188346862793,
"learning_rate": 1.340176708181637e-05,
"loss": 0.8036,
"step": 134
},
{
"epoch": 1.3608442274373918,
"grad_norm": 3.2634246349334717,
"learning_rate": 1.32908887327601e-05,
"loss": 0.6348,
"step": 135
},
{
"epoch": 1.3709245550480391,
"grad_norm": 2.8796093463897705,
"learning_rate": 1.317955478192515e-05,
"loss": 0.7079,
"step": 136
},
{
"epoch": 1.3810048826586865,
"grad_norm": 3.15555477142334,
"learning_rate": 1.306778064276064e-05,
"loss": 0.7197,
"step": 137
},
{
"epoch": 1.3910852102693338,
"grad_norm": 2.8696324825286865,
"learning_rate": 1.2955581789656844e-05,
"loss": 0.6422,
"step": 138
},
{
"epoch": 1.4011655378799812,
"grad_norm": 2.841829538345337,
"learning_rate": 1.2842973755802872e-05,
"loss": 0.6522,
"step": 139
},
{
"epoch": 1.4112458654906286,
"grad_norm": 2.869424819946289,
"learning_rate": 1.2729972131036212e-05,
"loss": 0.7462,
"step": 140
},
{
"epoch": 1.4213261931012757,
"grad_norm": 3.1612401008605957,
"learning_rate": 1.2616592559684408e-05,
"loss": 0.6471,
"step": 141
},
{
"epoch": 1.431406520711923,
"grad_norm": 2.7609212398529053,
"learning_rate": 1.25028507383992e-05,
"loss": 0.7627,
"step": 142
},
{
"epoch": 1.4414868483225705,
"grad_norm": 2.678645610809326,
"learning_rate": 1.2388762413983447e-05,
"loss": 0.6729,
"step": 143
},
{
"epoch": 1.4515671759332178,
"grad_norm": 2.8031179904937744,
"learning_rate": 1.2274343381211067e-05,
"loss": 0.6497,
"step": 144
},
{
"epoch": 1.4616475035438652,
"grad_norm": 2.735318660736084,
"learning_rate": 1.2159609480640361e-05,
"loss": 0.6786,
"step": 145
},
{
"epoch": 1.4717278311545126,
"grad_norm": 2.9970738887786865,
"learning_rate": 1.2044576596421003e-05,
"loss": 0.6498,
"step": 146
},
{
"epoch": 1.48180815876516,
"grad_norm": 2.8578624725341797,
"learning_rate": 1.192926065409497e-05,
"loss": 0.6432,
"step": 147
},
{
"epoch": 1.4918884863758073,
"grad_norm": 2.9433505535125732,
"learning_rate": 1.1813677618391759e-05,
"loss": 0.6274,
"step": 148
},
{
"epoch": 1.5019688139864544,
"grad_norm": 2.799851417541504,
"learning_rate": 1.1697843491018189e-05,
"loss": 0.6507,
"step": 149
},
{
"epoch": 1.5120491415971018,
"grad_norm": 2.589261770248413,
"learning_rate": 1.1581774308443042e-05,
"loss": 0.8801,
"step": 150
},
{
"epoch": 1.5221294692077492,
"grad_norm": 2.7499136924743652,
"learning_rate": 1.1465486139676955e-05,
"loss": 0.6428,
"step": 151
},
{
"epoch": 1.5322097968183965,
"grad_norm": 3.6993484497070312,
"learning_rate": 1.134899508404775e-05,
"loss": 0.6641,
"step": 152
},
{
"epoch": 1.542290124429044,
"grad_norm": 5.174093723297119,
"learning_rate": 1.1232317268971586e-05,
"loss": 0.7828,
"step": 153
},
{
"epoch": 1.5523704520396913,
"grad_norm": 2.734003782272339,
"learning_rate": 1.1115468847720245e-05,
"loss": 0.7631,
"step": 154
},
{
"epoch": 1.5624507796503386,
"grad_norm": 3.996946334838867,
"learning_rate": 1.0998465997184798e-05,
"loss": 0.6416,
"step": 155
},
{
"epoch": 1.572531107260986,
"grad_norm": 3.302497386932373,
"learning_rate": 1.088132491563602e-05,
"loss": 0.6543,
"step": 156
},
{
"epoch": 1.5826114348716334,
"grad_norm": 2.9945664405822754,
"learning_rate": 1.0764061820481872e-05,
"loss": 0.6902,
"step": 157
},
{
"epoch": 1.5926917624822807,
"grad_norm": 2.6038448810577393,
"learning_rate": 1.0646692946022285e-05,
"loss": 0.6289,
"step": 158
},
{
"epoch": 1.602772090092928,
"grad_norm": 2.6396548748016357,
"learning_rate": 1.0529234541201631e-05,
"loss": 0.8164,
"step": 159
},
{
"epoch": 1.6128524177035755,
"grad_norm": 2.770799160003662,
"learning_rate": 1.041170286735918e-05,
"loss": 0.6438,
"step": 160
},
{
"epoch": 1.6229327453142228,
"grad_norm": 2.666429042816162,
"learning_rate": 1.0294114195977796e-05,
"loss": 0.6912,
"step": 161
},
{
"epoch": 1.6330130729248702,
"grad_norm": 2.9112017154693604,
"learning_rate": 1.0176484806431288e-05,
"loss": 0.7345,
"step": 162
},
{
"epoch": 1.6430934005355176,
"grad_norm": 3.0811562538146973,
"learning_rate": 1.0058830983730622e-05,
"loss": 0.7558,
"step": 163
},
{
"epoch": 1.6531737281461647,
"grad_norm": 2.8588948249816895,
"learning_rate": 9.94116901626938e-06,
"loss": 0.648,
"step": 164
},
{
"epoch": 1.663254055756812,
"grad_norm": 2.7404801845550537,
"learning_rate": 9.823515193568715e-06,
"loss": 0.695,
"step": 165
},
{
"epoch": 1.6733343833674594,
"grad_norm": 2.9604055881500244,
"learning_rate": 9.705885804022207e-06,
"loss": 0.6304,
"step": 166
},
{
"epoch": 1.6834147109781068,
"grad_norm": 3.1369986534118652,
"learning_rate": 9.588297132640824e-06,
"loss": 0.8216,
"step": 167
},
{
"epoch": 1.6934950385887542,
"grad_norm": 2.912094831466675,
"learning_rate": 9.470765458798369e-06,
"loss": 0.6653,
"step": 168
},
{
"epoch": 1.7035753661994015,
"grad_norm": 2.6910438537597656,
"learning_rate": 9.353307053977717e-06,
"loss": 0.6645,
"step": 169
},
{
"epoch": 1.7136556938100487,
"grad_norm": 2.958939790725708,
"learning_rate": 9.235938179518131e-06,
"loss": 0.6222,
"step": 170
},
{
"epoch": 1.723736021420696,
"grad_norm": 2.7143542766571045,
"learning_rate": 9.118675084363986e-06,
"loss": 0.7051,
"step": 171
},
{
"epoch": 1.7338163490313434,
"grad_norm": 2.6610677242279053,
"learning_rate": 9.001534002815209e-06,
"loss": 0.6333,
"step": 172
},
{
"epoch": 1.7438966766419908,
"grad_norm": 2.9667537212371826,
"learning_rate": 8.884531152279757e-06,
"loss": 0.6832,
"step": 173
},
{
"epoch": 1.7539770042526381,
"grad_norm": 2.64609956741333,
"learning_rate": 8.767682731028415e-06,
"loss": 0.9484,
"step": 174
},
{
"epoch": 1.7640573318632855,
"grad_norm": 2.682523012161255,
"learning_rate": 8.651004915952252e-06,
"loss": 0.8721,
"step": 175
},
{
"epoch": 1.7741376594739329,
"grad_norm": 2.5906975269317627,
"learning_rate": 8.534513860323047e-06,
"loss": 0.9793,
"step": 176
},
{
"epoch": 1.7842179870845802,
"grad_norm": 2.636467456817627,
"learning_rate": 8.418225691556962e-06,
"loss": 0.9016,
"step": 177
},
{
"epoch": 1.7942983146952276,
"grad_norm": 3.5005948543548584,
"learning_rate": 8.302156508981816e-06,
"loss": 0.738,
"step": 178
},
{
"epoch": 1.804378642305875,
"grad_norm": 2.7986643314361572,
"learning_rate": 8.18632238160824e-06,
"loss": 0.6635,
"step": 179
},
{
"epoch": 1.8144589699165223,
"grad_norm": 2.8597512245178223,
"learning_rate": 8.070739345905032e-06,
"loss": 0.7473,
"step": 180
},
{
"epoch": 1.8245392975271697,
"grad_norm": 2.7487239837646484,
"learning_rate": 7.955423403578998e-06,
"loss": 0.7526,
"step": 181
},
{
"epoch": 1.834619625137817,
"grad_norm": 2.68874454498291,
"learning_rate": 7.840390519359644e-06,
"loss": 0.6491,
"step": 182
},
{
"epoch": 1.8446999527484644,
"grad_norm": 2.8393709659576416,
"learning_rate": 7.725656618788938e-06,
"loss": 0.6401,
"step": 183
},
{
"epoch": 1.8547802803591118,
"grad_norm": 2.8322646617889404,
"learning_rate": 7.611237586016558e-06,
"loss": 0.7692,
"step": 184
},
{
"epoch": 1.8648606079697592,
"grad_norm": 2.760575771331787,
"learning_rate": 7.497149261600803e-06,
"loss": 0.926,
"step": 185
},
{
"epoch": 1.8749409355804063,
"grad_norm": 2.6379311084747314,
"learning_rate": 7.383407440315595e-06,
"loss": 0.654,
"step": 186
},
{
"epoch": 1.8850212631910537,
"grad_norm": 2.6411261558532715,
"learning_rate": 7.27002786896379e-06,
"loss": 0.7753,
"step": 187
},
{
"epoch": 1.895101590801701,
"grad_norm": 2.6866044998168945,
"learning_rate": 7.157026244197132e-06,
"loss": 0.6479,
"step": 188
},
{
"epoch": 1.9051819184123484,
"grad_norm": 2.743093252182007,
"learning_rate": 7.044418210343161e-06,
"loss": 0.7825,
"step": 189
},
{
"epoch": 1.9152622460229958,
"grad_norm": 2.7608628273010254,
"learning_rate": 6.932219357239362e-06,
"loss": 0.6497,
"step": 190
},
{
"epoch": 1.9253425736336431,
"grad_norm": 2.581033706665039,
"learning_rate": 6.820445218074849e-06,
"loss": 0.658,
"step": 191
},
{
"epoch": 1.9354229012442903,
"grad_norm": 2.8341994285583496,
"learning_rate": 6.7091112672399e-06,
"loss": 0.8367,
"step": 192
},
{
"epoch": 1.9455032288549376,
"grad_norm": 2.712247133255005,
"learning_rate": 6.5982329181836325e-06,
"loss": 0.647,
"step": 193
},
{
"epoch": 1.955583556465585,
"grad_norm": 2.683356761932373,
"learning_rate": 6.487825521280109e-06,
"loss": 0.7316,
"step": 194
},
{
"epoch": 1.9656638840762324,
"grad_norm": 2.6433842182159424,
"learning_rate": 6.3779043617031775e-06,
"loss": 0.84,
"step": 195
},
{
"epoch": 1.9757442116868797,
"grad_norm": 4.231180667877197,
"learning_rate": 6.268484657310351e-06,
"loss": 0.7416,
"step": 196
},
{
"epoch": 1.9858245392975271,
"grad_norm": 3.0023813247680664,
"learning_rate": 6.159581556535989e-06,
"loss": 0.8632,
"step": 197
},
{
"epoch": 1.9959048669081745,
"grad_norm": 2.6306092739105225,
"learning_rate": 6.051210136294089e-06,
"loss": 0.6557,
"step": 198
},
{
"epoch": 2.005985194518822,
"grad_norm": 3.3288090229034424,
"learning_rate": 5.943385399891004e-06,
"loss": 0.5327,
"step": 199
},
{
"epoch": 2.016065522129469,
"grad_norm": 3.394890069961548,
"learning_rate": 5.8361222749483246e-06,
"loss": 0.5682,
"step": 200
},
{
"epoch": 2.0261458497401166,
"grad_norm": 4.5706658363342285,
"learning_rate": 5.729435611336239e-06,
"loss": 0.521,
"step": 201
},
{
"epoch": 2.036226177350764,
"grad_norm": 3.595043420791626,
"learning_rate": 5.6233401791176946e-06,
"loss": 0.4973,
"step": 202
},
{
"epoch": 2.0463065049614113,
"grad_norm": 3.381319761276245,
"learning_rate": 5.517850666503547e-06,
"loss": 0.7273,
"step": 203
},
{
"epoch": 2.0563868325720587,
"grad_norm": 3.813019037246704,
"learning_rate": 5.412981677819094e-06,
"loss": 0.5748,
"step": 204
},
{
"epoch": 2.066467160182706,
"grad_norm": 2.9697418212890625,
"learning_rate": 5.308747731482207e-06,
"loss": 0.6197,
"step": 205
},
{
"epoch": 2.0765474877933534,
"grad_norm": 4.898626804351807,
"learning_rate": 5.205163257993341e-06,
"loss": 0.4839,
"step": 206
},
{
"epoch": 2.0866278154040008,
"grad_norm": 4.427509784698486,
"learning_rate": 5.1022425979377174e-06,
"loss": 0.455,
"step": 207
},
{
"epoch": 2.096708143014648,
"grad_norm": 3.7625739574432373,
"learning_rate": 5.000000000000003e-06,
"loss": 0.538,
"step": 208
},
{
"epoch": 2.1067884706252955,
"grad_norm": 2.9984402656555176,
"learning_rate": 4.89844961899163e-06,
"loss": 0.5426,
"step": 209
},
{
"epoch": 2.116868798235943,
"grad_norm": 3.2791342735290527,
"learning_rate": 4.797605513891179e-06,
"loss": 0.5505,
"step": 210
},
{
"epoch": 2.12694912584659,
"grad_norm": 3.1224496364593506,
"learning_rate": 4.697481645898012e-06,
"loss": 0.5466,
"step": 211
},
{
"epoch": 2.137029453457237,
"grad_norm": 2.9908924102783203,
"learning_rate": 4.598091876499417e-06,
"loss": 0.4739,
"step": 212
},
{
"epoch": 2.1471097810678845,
"grad_norm": 3.272909164428711,
"learning_rate": 4.4994499655515865e-06,
"loss": 0.4773,
"step": 213
},
{
"epoch": 2.157190108678532,
"grad_norm": 3.1660666465759277,
"learning_rate": 4.4015695693746685e-06,
"loss": 0.6012,
"step": 214
},
{
"epoch": 2.1672704362891793,
"grad_norm": 2.9826879501342773,
"learning_rate": 4.304464238862115e-06,
"loss": 0.559,
"step": 215
},
{
"epoch": 2.1773507638998266,
"grad_norm": 3.156632900238037,
"learning_rate": 4.208147417604665e-06,
"loss": 0.4658,
"step": 216
},
{
"epoch": 2.187431091510474,
"grad_norm": 2.7207696437835693,
"learning_rate": 4.112632440029176e-06,
"loss": 0.4746,
"step": 217
},
{
"epoch": 2.1975114191211214,
"grad_norm": 3.170917272567749,
"learning_rate": 4.017932529552543e-06,
"loss": 0.4555,
"step": 218
},
{
"epoch": 2.2075917467317687,
"grad_norm": 2.873971939086914,
"learning_rate": 3.924060796751012e-06,
"loss": 0.4927,
"step": 219
},
{
"epoch": 2.217672074342416,
"grad_norm": 3.0037410259246826,
"learning_rate": 3.83103023754511e-06,
"loss": 0.5199,
"step": 220
},
{
"epoch": 2.2277524019530635,
"grad_norm": 3.2414352893829346,
"learning_rate": 3.7388537314004394e-06,
"loss": 0.4665,
"step": 221
},
{
"epoch": 2.237832729563711,
"grad_norm": 2.9535632133483887,
"learning_rate": 3.647544039544615e-06,
"loss": 0.4625,
"step": 222
},
{
"epoch": 2.247913057174358,
"grad_norm": 2.874563455581665,
"learning_rate": 3.557113803200537e-06,
"loss": 0.4651,
"step": 223
},
{
"epoch": 2.2579933847850056,
"grad_norm": 2.8400771617889404,
"learning_rate": 3.4675755418363054e-06,
"loss": 0.4741,
"step": 224
},
{
"epoch": 2.268073712395653,
"grad_norm": 3.162914752960205,
"learning_rate": 3.378941651431996e-06,
"loss": 0.5043,
"step": 225
},
{
"epoch": 2.2781540400063003,
"grad_norm": 3.108367681503296,
"learning_rate": 3.2912244027634953e-06,
"loss": 0.4612,
"step": 226
},
{
"epoch": 2.2882343676169477,
"grad_norm": 2.9453072547912598,
"learning_rate": 3.204435939703705e-06,
"loss": 0.5951,
"step": 227
},
{
"epoch": 2.298314695227595,
"grad_norm": 2.926748752593994,
"learning_rate": 3.1185882775413123e-06,
"loss": 0.4727,
"step": 228
},
{
"epoch": 2.3083950228382424,
"grad_norm": 2.7505362033843994,
"learning_rate": 3.0336933013173307e-06,
"loss": 0.4771,
"step": 229
},
{
"epoch": 2.3184753504488897,
"grad_norm": 3.1303627490997314,
"learning_rate": 2.949762764179711e-06,
"loss": 0.5534,
"step": 230
},
{
"epoch": 2.328555678059537,
"grad_norm": 2.6740784645080566,
"learning_rate": 2.8668082857562006e-06,
"loss": 0.4713,
"step": 231
},
{
"epoch": 2.3386360056701845,
"grad_norm": 2.6228513717651367,
"learning_rate": 2.7848413505456564e-06,
"loss": 0.532,
"step": 232
},
{
"epoch": 2.348716333280832,
"grad_norm": 3.3503799438476562,
"learning_rate": 2.7038733063281177e-06,
"loss": 0.5022,
"step": 233
},
{
"epoch": 2.358796660891479,
"grad_norm": 2.798093557357788,
"learning_rate": 2.6239153625937786e-06,
"loss": 0.4674,
"step": 234
},
{
"epoch": 2.368876988502126,
"grad_norm": 2.9283840656280518,
"learning_rate": 2.544978588991096e-06,
"loss": 0.5145,
"step": 235
},
{
"epoch": 2.3789573161127735,
"grad_norm": 2.9294168949127197,
"learning_rate": 2.4670739137942723e-06,
"loss": 0.5262,
"step": 236
},
{
"epoch": 2.389037643723421,
"grad_norm": 2.9340579509735107,
"learning_rate": 2.390212122390323e-06,
"loss": 0.4654,
"step": 237
},
{
"epoch": 2.3991179713340682,
"grad_norm": 2.9282586574554443,
"learning_rate": 2.3144038557858915e-06,
"loss": 0.7147,
"step": 238
},
{
"epoch": 2.4091982989447156,
"grad_norm": 2.9570140838623047,
"learning_rate": 2.2396596091340805e-06,
"loss": 0.4643,
"step": 239
},
{
"epoch": 2.419278626555363,
"grad_norm": 2.7537012100219727,
"learning_rate": 2.165989730281475e-06,
"loss": 0.4467,
"step": 240
},
{
"epoch": 2.4293589541660103,
"grad_norm": 2.764420986175537,
"learning_rate": 2.0934044183355384e-06,
"loss": 0.4774,
"step": 241
},
{
"epoch": 2.4394392817766577,
"grad_norm": 3.11476993560791,
"learning_rate": 2.0219137222526188e-06,
"loss": 0.5792,
"step": 242
},
{
"epoch": 2.449519609387305,
"grad_norm": 4.960219860076904,
"learning_rate": 1.9515275394467446e-06,
"loss": 0.457,
"step": 243
},
{
"epoch": 2.4595999369979524,
"grad_norm": 3.0883636474609375,
"learning_rate": 1.882255614419376e-06,
"loss": 0.6268,
"step": 244
},
{
"epoch": 2.4696802646086,
"grad_norm": 2.9060003757476807,
"learning_rate": 1.8141075374103634e-06,
"loss": 0.5785,
"step": 245
},
{
"epoch": 2.479760592219247,
"grad_norm": 2.7196030616760254,
"learning_rate": 1.7470927430702277e-06,
"loss": 0.4658,
"step": 246
},
{
"epoch": 2.4898409198298945,
"grad_norm": 2.8663458824157715,
"learning_rate": 1.6812205091539979e-06,
"loss": 0.4635,
"step": 247
},
{
"epoch": 2.499921247440542,
"grad_norm": 2.7674198150634766,
"learning_rate": 1.6164999552367767e-06,
"loss": 0.475,
"step": 248
},
{
"epoch": 2.5100015750511893,
"grad_norm": 2.971801280975342,
"learning_rate": 1.5529400414511809e-06,
"loss": 0.5657,
"step": 249
},
{
"epoch": 2.5200819026618366,
"grad_norm": 3.0309677124023438,
"learning_rate": 1.4905495672468784e-06,
"loss": 0.4609,
"step": 250
},
{
"epoch": 2.530162230272484,
"grad_norm": 2.9311161041259766,
"learning_rate": 1.4293371701723701e-06,
"loss": 0.5184,
"step": 251
},
{
"epoch": 2.5402425578831314,
"grad_norm": 2.813347578048706,
"learning_rate": 1.369311324679159e-06,
"loss": 0.4675,
"step": 252
},
{
"epoch": 2.5503228854937783,
"grad_norm": 2.8366470336914062,
"learning_rate": 1.3104803409485357e-06,
"loss": 0.5566,
"step": 253
},
{
"epoch": 2.5604032131044256,
"grad_norm": 2.9654905796051025,
"learning_rate": 1.252852363741084e-06,
"loss": 0.4739,
"step": 254
},
{
"epoch": 2.570483540715073,
"grad_norm": 2.7376723289489746,
"learning_rate": 1.196435371269089e-06,
"loss": 0.4755,
"step": 255
},
{
"epoch": 2.5805638683257204,
"grad_norm": 2.603374719619751,
"learning_rate": 1.1412371740920036e-06,
"loss": 0.45,
"step": 256
},
{
"epoch": 2.5906441959363677,
"grad_norm": 2.7828543186187744,
"learning_rate": 1.0872654140351458e-06,
"loss": 0.7043,
"step": 257
},
{
"epoch": 2.600724523547015,
"grad_norm": 2.6468091011047363,
"learning_rate": 1.0345275631317165e-06,
"loss": 0.6008,
"step": 258
},
{
"epoch": 2.6108048511576625,
"grad_norm": 2.7856605052948,
"learning_rate": 9.830309225883562e-07,
"loss": 0.55,
"step": 259
},
{
"epoch": 2.62088517876831,
"grad_norm": 2.949894428253174,
"learning_rate": 9.327826217743452e-07,
"loss": 0.4517,
"step": 260
},
{
"epoch": 2.630965506378957,
"grad_norm": 3.2041678428649902,
"learning_rate": 8.837896172345827e-07,
"loss": 0.5641,
"step": 261
},
{
"epoch": 2.6410458339896046,
"grad_norm": 3.0084447860717773,
"learning_rate": 8.360586917264979e-07,
"loss": 0.4635,
"step": 262
},
{
"epoch": 2.651126161600252,
"grad_norm": 2.7008798122406006,
"learning_rate": 7.895964532810318e-07,
"loss": 0.5861,
"step": 263
},
{
"epoch": 2.6612064892108993,
"grad_norm": 2.8974366188049316,
"learning_rate": 7.4440933428779e-07,
"loss": 0.4619,
"step": 264
},
{
"epoch": 2.6712868168215467,
"grad_norm": 2.954524040222168,
"learning_rate": 7.005035906045199e-07,
"loss": 0.4819,
"step": 265
},
{
"epoch": 2.681367144432194,
"grad_norm": 2.6953723430633545,
"learning_rate": 6.578853006910402e-07,
"loss": 0.4843,
"step": 266
},
{
"epoch": 2.6914474720428414,
"grad_norm": 3.108698844909668,
"learning_rate": 6.165603647677054e-07,
"loss": 0.4586,
"step": 267
},
{
"epoch": 2.7015277996534888,
"grad_norm": 2.9222874641418457,
"learning_rate": 5.765345039985648e-07,
"loss": 0.4737,
"step": 268
},
{
"epoch": 2.711608127264136,
"grad_norm": 2.904561996459961,
"learning_rate": 5.378132596993047e-07,
"loss": 0.4641,
"step": 269
},
{
"epoch": 2.7216884548747835,
"grad_norm": 3.0697550773620605,
"learning_rate": 5.004019925700921e-07,
"loss": 0.5555,
"step": 270
},
{
"epoch": 2.731768782485431,
"grad_norm": 2.980802536010742,
"learning_rate": 4.6430588195341853e-07,
"loss": 0.5908,
"step": 271
},
{
"epoch": 2.7418491100960782,
"grad_norm": 3.690682888031006,
"learning_rate": 4.295299251170537e-07,
"loss": 0.5399,
"step": 272
},
{
"epoch": 2.7519294377067256,
"grad_norm": 2.920565128326416,
"learning_rate": 3.960789365622075e-07,
"loss": 0.7304,
"step": 273
},
{
"epoch": 2.762009765317373,
"grad_norm": 2.853963851928711,
"learning_rate": 3.6395754735699896e-07,
"loss": 0.571,
"step": 274
},
{
"epoch": 2.7720900929280203,
"grad_norm": 2.8439598083496094,
"learning_rate": 3.3317020449530666e-07,
"loss": 0.5728,
"step": 275
},
{
"epoch": 2.7821704205386677,
"grad_norm": 2.6904118061065674,
"learning_rate": 3.0372117028111825e-07,
"loss": 0.4688,
"step": 276
},
{
"epoch": 2.792250748149315,
"grad_norm": 2.8385207653045654,
"learning_rate": 2.7561452173844206e-07,
"loss": 0.5202,
"step": 277
},
{
"epoch": 2.8023310757599624,
"grad_norm": 2.739987373352051,
"learning_rate": 2.488541500468666e-07,
"loss": 0.586,
"step": 278
},
{
"epoch": 2.81241140337061,
"grad_norm": 2.9050772190093994,
"learning_rate": 2.2344376000285606e-07,
"loss": 0.4698,
"step": 279
},
{
"epoch": 2.822491730981257,
"grad_norm": 2.7418956756591797,
"learning_rate": 1.993868695068457e-07,
"loss": 0.4432,
"step": 280
},
{
"epoch": 2.832572058591904,
"grad_norm": 2.746795892715454,
"learning_rate": 1.766868090762075e-07,
"loss": 0.5331,
"step": 281
},
{
"epoch": 2.8426523862025515,
"grad_norm": 2.7366440296173096,
"learning_rate": 1.553467213841664e-07,
"loss": 0.4813,
"step": 282
},
{
"epoch": 2.852732713813199,
"grad_norm": 2.755558729171753,
"learning_rate": 1.3536956082472074e-07,
"loss": 0.4859,
"step": 283
},
{
"epoch": 2.862813041423846,
"grad_norm": 2.7609097957611084,
"learning_rate": 1.1675809310361497e-07,
"loss": 0.4655,
"step": 284
},
{
"epoch": 2.8728933690344935,
"grad_norm": 3.0042688846588135,
"learning_rate": 9.951489485545696e-08,
"loss": 0.4557,
"step": 285
},
{
"epoch": 2.882973696645141,
"grad_norm": 2.9627745151519775,
"learning_rate": 8.364235328699566e-08,
"loss": 0.4968,
"step": 286
},
{
"epoch": 2.8930540242557883,
"grad_norm": 2.8256924152374268,
"learning_rate": 6.914266584662988e-08,
"loss": 0.4728,
"step": 287
},
{
"epoch": 2.9031343518664356,
"grad_norm": 2.7654502391815186,
"learning_rate": 5.6017839920180506e-08,
"loss": 0.4594,
"step": 288
},
{
"epoch": 2.913214679477083,
"grad_norm": 5.463405609130859,
"learning_rate": 4.426969255298841e-08,
"loss": 0.4786,
"step": 289
},
{
"epoch": 2.9232950070877304,
"grad_norm": 2.978433132171631,
"learning_rate": 3.38998501983534e-08,
"loss": 0.4687,
"step": 290
},
{
"epoch": 2.9333753346983777,
"grad_norm": 2.887697458267212,
"learning_rate": 2.4909748492362162e-08,
"loss": 0.5466,
"step": 291
},
{
"epoch": 2.943455662309025,
"grad_norm": 2.823009729385376,
"learning_rate": 1.730063205513277e-08,
"loss": 0.5068,
"step": 292
},
{
"epoch": 2.9535359899196725,
"grad_norm": 2.9089882373809814,
"learning_rate": 1.1073554318509206e-08,
"loss": 0.4521,
"step": 293
},
{
"epoch": 2.96361631753032,
"grad_norm": 2.763700246810913,
"learning_rate": 6.229377380218005e-09,
"loss": 0.456,
"step": 294
},
{
"epoch": 2.973696645140967,
"grad_norm": 2.9146535396575928,
"learning_rate": 2.7687718845148538e-09,
"loss": 0.4586,
"step": 295
},
{
"epoch": 2.9837769727516146,
"grad_norm": 2.999981641769409,
"learning_rate": 6.922169293421821e-10,
"loss": 0.464,
"step": 296
},
{
"epoch": 2.9938573003622615,
"grad_norm": 4.130459785461426,
"learning_rate": 0.0,
"loss": 0.4729,
"step": 297
},
{
"epoch": 2.9938573003622615,
"step": 297,
"total_flos": 1.3286712322107113e+19,
"train_loss": 0.7266113918638389,
"train_runtime": 26017.2965,
"train_samples_per_second": 5.856,
"train_steps_per_second": 0.011
}
],
"logging_steps": 1.0,
"max_steps": 297,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.3286712322107113e+19,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}