adaption_fim_event_classification / trainer_state.json
dowdryan's picture
Add 11 files
395835f verified
Raw History Blame Contribute Delete
43.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 47,
"global_step": 235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02127659574468085,
"grad_norm": 0.3123471736907959,
"learning_rate": 0.0,
"loss": 1.396484375,
"step": 1
},
{
"epoch": 0.0425531914893617,
"grad_norm": 0.37419506907463074,
"learning_rate": 8.333333333333334e-06,
"loss": 1.4912109375,
"step": 2
},
{
"epoch": 0.06382978723404255,
"grad_norm": 0.40808141231536865,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.5693359375,
"step": 3
},
{
"epoch": 0.0851063829787234,
"grad_norm": 0.36097270250320435,
"learning_rate": 2.5e-05,
"loss": 1.46826171875,
"step": 4
},
{
"epoch": 0.10638297872340426,
"grad_norm": 0.3010517358779907,
"learning_rate": 3.3333333333333335e-05,
"loss": 1.3857421875,
"step": 5
},
{
"epoch": 0.1276595744680851,
"grad_norm": 0.35135418176651,
"learning_rate": 4.166666666666667e-05,
"loss": 1.6064453125,
"step": 6
},
{
"epoch": 0.14893617021276595,
"grad_norm": 0.2829645276069641,
"learning_rate": 5e-05,
"loss": 1.5078125,
"step": 7
},
{
"epoch": 0.1702127659574468,
"grad_norm": 0.2620680630207062,
"learning_rate": 5.833333333333334e-05,
"loss": 1.515625,
"step": 8
},
{
"epoch": 0.19148936170212766,
"grad_norm": 0.24931293725967407,
"learning_rate": 6.666666666666667e-05,
"loss": 1.443359375,
"step": 9
},
{
"epoch": 0.2127659574468085,
"grad_norm": 0.28700199723243713,
"learning_rate": 7.500000000000001e-05,
"loss": 1.3603515625,
"step": 10
},
{
"epoch": 0.23404255319148937,
"grad_norm": 0.26376253366470337,
"learning_rate": 8.333333333333334e-05,
"loss": 1.265625,
"step": 11
},
{
"epoch": 0.2553191489361702,
"grad_norm": 0.2816177010536194,
"learning_rate": 9.166666666666667e-05,
"loss": 1.365234375,
"step": 12
},
{
"epoch": 0.2765957446808511,
"grad_norm": 0.24004390835762024,
"learning_rate": 0.0001,
"loss": 1.31201171875,
"step": 13
},
{
"epoch": 0.2978723404255319,
"grad_norm": 0.18498215079307556,
"learning_rate": 9.999603070998052e-05,
"loss": 1.255859375,
"step": 14
},
{
"epoch": 0.3191489361702128,
"grad_norm": 0.1898963451385498,
"learning_rate": 9.998412362768521e-05,
"loss": 1.208984375,
"step": 15
},
{
"epoch": 0.3404255319148936,
"grad_norm": 0.16993463039398193,
"learning_rate": 9.996428111624723e-05,
"loss": 1.291015625,
"step": 16
},
{
"epoch": 0.3617021276595745,
"grad_norm": 0.1416313499212265,
"learning_rate": 9.993650711370072e-05,
"loss": 1.3310546875,
"step": 17
},
{
"epoch": 0.3829787234042553,
"grad_norm": 0.13824915885925293,
"learning_rate": 9.990080713219923e-05,
"loss": 1.33203125,
"step": 18
},
{
"epoch": 0.40425531914893614,
"grad_norm": 0.1501912921667099,
"learning_rate": 9.985718825692175e-05,
"loss": 1.3603515625,
"step": 19
},
{
"epoch": 0.425531914893617,
"grad_norm": 0.12451738119125366,
"learning_rate": 9.980565914466662e-05,
"loss": 1.3720703125,
"step": 20
},
{
"epoch": 0.44680851063829785,
"grad_norm": 0.10955026000738144,
"learning_rate": 9.97462300221334e-05,
"loss": 1.3642578125,
"step": 21
},
{
"epoch": 0.46808510638297873,
"grad_norm": 0.09165201336145401,
"learning_rate": 9.967891268389317e-05,
"loss": 1.1064453125,
"step": 22
},
{
"epoch": 0.48936170212765956,
"grad_norm": 0.09895090013742447,
"learning_rate": 9.960372049004798e-05,
"loss": 1.25439453125,
"step": 23
},
{
"epoch": 0.5106382978723404,
"grad_norm": 0.09168048202991486,
"learning_rate": 9.952066836357896e-05,
"loss": 1.2314453125,
"step": 24
},
{
"epoch": 0.5319148936170213,
"grad_norm": 0.09091522544622421,
"learning_rate": 9.942977278738501e-05,
"loss": 1.3154296875,
"step": 25
},
{
"epoch": 0.5531914893617021,
"grad_norm": 0.08179040253162384,
"learning_rate": 9.933105180101125e-05,
"loss": 1.169921875,
"step": 26
},
{
"epoch": 0.574468085106383,
"grad_norm": 0.0734533965587616,
"learning_rate": 9.922452499706902e-05,
"loss": 1.15576171875,
"step": 27
},
{
"epoch": 0.5957446808510638,
"grad_norm": 0.0753985196352005,
"learning_rate": 9.911021351734728e-05,
"loss": 1.2431640625,
"step": 28
},
{
"epoch": 0.6170212765957447,
"grad_norm": 0.07138530910015106,
"learning_rate": 9.898814004861681e-05,
"loss": 1.173828125,
"step": 29
},
{
"epoch": 0.6382978723404256,
"grad_norm": 0.08894495666027069,
"learning_rate": 9.885832881812768e-05,
"loss": 1.283203125,
"step": 30
},
{
"epoch": 0.6595744680851063,
"grad_norm": 0.0702628493309021,
"learning_rate": 9.872080558880095e-05,
"loss": 1.32421875,
"step": 31
},
{
"epoch": 0.6808510638297872,
"grad_norm": 0.07132714986801147,
"learning_rate": 9.85755976541157e-05,
"loss": 1.2841796875,
"step": 32
},
{
"epoch": 0.7021276595744681,
"grad_norm": 0.06748408079147339,
"learning_rate": 9.842273383269224e-05,
"loss": 1.1552734375,
"step": 33
},
{
"epoch": 0.723404255319149,
"grad_norm": 0.07180913537740707,
"learning_rate": 9.826224446257261e-05,
"loss": 1.2578125,
"step": 34
},
{
"epoch": 0.7446808510638298,
"grad_norm": 0.07629711925983429,
"learning_rate": 9.809416139519953e-05,
"loss": 1.2958984375,
"step": 35
},
{
"epoch": 0.7659574468085106,
"grad_norm": 0.07060649991035461,
"learning_rate": 9.791851798909513e-05,
"loss": 1.263671875,
"step": 36
},
{
"epoch": 0.7872340425531915,
"grad_norm": 0.06965949386358261,
"learning_rate": 9.773534910324033e-05,
"loss": 1.2705078125,
"step": 37
},
{
"epoch": 0.8085106382978723,
"grad_norm": 0.06155743449926376,
"learning_rate": 9.754469109015665e-05,
"loss": 1.20654296875,
"step": 38
},
{
"epoch": 0.8297872340425532,
"grad_norm": 0.07107419520616531,
"learning_rate": 9.734658178869154e-05,
"loss": 1.2216796875,
"step": 39
},
{
"epoch": 0.851063829787234,
"grad_norm": 0.06438836455345154,
"learning_rate": 9.71410605165086e-05,
"loss": 1.19580078125,
"step": 40
},
{
"epoch": 0.8723404255319149,
"grad_norm": 0.06160025671124458,
"learning_rate": 9.692816806228462e-05,
"loss": 1.20654296875,
"step": 41
},
{
"epoch": 0.8936170212765957,
"grad_norm": 0.06318897008895874,
"learning_rate": 9.670794667761422e-05,
"loss": 1.1572265625,
"step": 42
},
{
"epoch": 0.9148936170212766,
"grad_norm": 0.06440263986587524,
"learning_rate": 9.648044006862469e-05,
"loss": 1.14697265625,
"step": 43
},
{
"epoch": 0.9361702127659575,
"grad_norm": 0.06616906076669693,
"learning_rate": 9.624569338730158e-05,
"loss": 1.2890625,
"step": 44
},
{
"epoch": 0.9574468085106383,
"grad_norm": 0.07043126970529556,
"learning_rate": 9.60037532225279e-05,
"loss": 1.2490234375,
"step": 45
},
{
"epoch": 0.9787234042553191,
"grad_norm": 0.05934222415089607,
"learning_rate": 9.57546675908377e-05,
"loss": 1.14208984375,
"step": 46
},
{
"epoch": 1.0,
"grad_norm": 0.08394598215818405,
"learning_rate": 9.549848592688658e-05,
"loss": 1.1787109375,
"step": 47
},
{
"epoch": 1.0,
"eval_loss": 1.23828125,
"eval_runtime": 0.7094,
"eval_samples_per_second": 7.048,
"eval_steps_per_second": 1.41,
"step": 47
},
{
"epoch": 1.0212765957446808,
"grad_norm": 0.08404431492090225,
"learning_rate": 9.523525907364064e-05,
"loss": 1.16650390625,
"step": 48
},
{
"epoch": 1.0425531914893618,
"grad_norm": 0.06203487515449524,
"learning_rate": 9.496503927228593e-05,
"loss": 1.212890625,
"step": 49
},
{
"epoch": 1.0638297872340425,
"grad_norm": 0.06887223571538925,
"learning_rate": 9.468788015186048e-05,
"loss": 1.26171875,
"step": 50
},
{
"epoch": 1.0851063829787233,
"grad_norm": 0.06444814056158066,
"learning_rate": 9.440383671861084e-05,
"loss": 1.181640625,
"step": 51
},
{
"epoch": 1.1063829787234043,
"grad_norm": 0.06496715545654297,
"learning_rate": 9.411296534507524e-05,
"loss": 1.16015625,
"step": 52
},
{
"epoch": 1.127659574468085,
"grad_norm": 0.07317639142274857,
"learning_rate": 9.381532375889565e-05,
"loss": 1.3076171875,
"step": 53
},
{
"epoch": 1.148936170212766,
"grad_norm": 0.0656491219997406,
"learning_rate": 9.351097103136097e-05,
"loss": 1.2529296875,
"step": 54
},
{
"epoch": 1.1702127659574468,
"grad_norm": 0.06748198717832565,
"learning_rate": 9.319996756568339e-05,
"loss": 1.2900390625,
"step": 55
},
{
"epoch": 1.1914893617021276,
"grad_norm": 0.06838950514793396,
"learning_rate": 9.288237508501052e-05,
"loss": 1.2578125,
"step": 56
},
{
"epoch": 1.2127659574468086,
"grad_norm": 0.08372005820274353,
"learning_rate": 9.255825662017554e-05,
"loss": 1.177734375,
"step": 57
},
{
"epoch": 1.2340425531914894,
"grad_norm": 0.06518688052892685,
"learning_rate": 9.222767649718784e-05,
"loss": 1.11669921875,
"step": 58
},
{
"epoch": 1.2553191489361701,
"grad_norm": 0.06871097534894943,
"learning_rate": 9.189070032446657e-05,
"loss": 1.21875,
"step": 59
},
{
"epoch": 1.2765957446808511,
"grad_norm": 0.06764581799507141,
"learning_rate": 9.154739497981964e-05,
"loss": 1.16259765625,
"step": 60
},
{
"epoch": 1.297872340425532,
"grad_norm": 0.06686984747648239,
"learning_rate": 9.119782859717101e-05,
"loss": 1.1259765625,
"step": 61
},
{
"epoch": 1.3191489361702127,
"grad_norm": 0.0660318061709404,
"learning_rate": 9.084207055303837e-05,
"loss": 1.08935546875,
"step": 62
},
{
"epoch": 1.3404255319148937,
"grad_norm": 0.07046317309141159,
"learning_rate": 9.048019145276441e-05,
"loss": 1.17626953125,
"step": 63
},
{
"epoch": 1.3617021276595744,
"grad_norm": 0.07238190621137619,
"learning_rate": 9.011226311650414e-05,
"loss": 1.21484375,
"step": 64
},
{
"epoch": 1.3829787234042552,
"grad_norm": 0.07361134886741638,
"learning_rate": 8.973835856497125e-05,
"loss": 1.2197265625,
"step": 65
},
{
"epoch": 1.4042553191489362,
"grad_norm": 0.07109193503856659,
"learning_rate": 8.935855200494598e-05,
"loss": 1.244140625,
"step": 66
},
{
"epoch": 1.425531914893617,
"grad_norm": 0.07751573622226715,
"learning_rate": 8.89729188145477e-05,
"loss": 1.2568359375,
"step": 67
},
{
"epoch": 1.4468085106382977,
"grad_norm": 0.07774888724088669,
"learning_rate": 8.858153552827512e-05,
"loss": 1.2529296875,
"step": 68
},
{
"epoch": 1.4680851063829787,
"grad_norm": 0.0669010654091835,
"learning_rate": 8.818447982181684e-05,
"loss": 1.01708984375,
"step": 69
},
{
"epoch": 1.4893617021276595,
"grad_norm": 0.07133079320192337,
"learning_rate": 8.778183049663548e-05,
"loss": 1.1572265625,
"step": 70
},
{
"epoch": 1.5106382978723403,
"grad_norm": 0.0725470781326294,
"learning_rate": 8.737366746432847e-05,
"loss": 1.1376953125,
"step": 71
},
{
"epoch": 1.5319148936170213,
"grad_norm": 0.07313201576471329,
"learning_rate": 8.69600717307683e-05,
"loss": 1.2236328125,
"step": 72
},
{
"epoch": 1.5531914893617023,
"grad_norm": 0.06871186196804047,
"learning_rate": 8.654112538002586e-05,
"loss": 1.0810546875,
"step": 73
},
{
"epoch": 1.574468085106383,
"grad_norm": 0.06845533847808838,
"learning_rate": 8.611691155807955e-05,
"loss": 1.07421875,
"step": 74
},
{
"epoch": 1.5957446808510638,
"grad_norm": 0.07805629074573517,
"learning_rate": 8.568751445631386e-05,
"loss": 1.1572265625,
"step": 75
},
{
"epoch": 1.6170212765957448,
"grad_norm": 0.07151457667350769,
"learning_rate": 8.525301929481037e-05,
"loss": 1.0908203125,
"step": 76
},
{
"epoch": 1.6382978723404256,
"grad_norm": 0.0800897628068924,
"learning_rate": 8.481351230543439e-05,
"loss": 1.1943359375,
"step": 77
},
{
"epoch": 1.6595744680851063,
"grad_norm": 0.07139340043067932,
"learning_rate": 8.436908071472132e-05,
"loss": 1.2412109375,
"step": 78
},
{
"epoch": 1.6808510638297873,
"grad_norm": 0.07299311459064484,
"learning_rate": 8.3919812726565e-05,
"loss": 1.208984375,
"step": 79
},
{
"epoch": 1.702127659574468,
"grad_norm": 0.07305039465427399,
"learning_rate": 8.346579750471246e-05,
"loss": 1.08251953125,
"step": 80
},
{
"epoch": 1.7234042553191489,
"grad_norm": 0.0772218331694603,
"learning_rate": 8.300712515506821e-05,
"loss": 1.1796875,
"step": 81
},
{
"epoch": 1.7446808510638299,
"grad_norm": 0.07679948955774307,
"learning_rate": 8.254388670781119e-05,
"loss": 1.2197265625,
"step": 82
},
{
"epoch": 1.7659574468085106,
"grad_norm": 0.08456248044967651,
"learning_rate": 8.207617409932868e-05,
"loss": 1.1845703125,
"step": 83
},
{
"epoch": 1.7872340425531914,
"grad_norm": 0.08139702677726746,
"learning_rate": 8.160408015397012e-05,
"loss": 1.1962890625,
"step": 84
},
{
"epoch": 1.8085106382978724,
"grad_norm": 0.07431219518184662,
"learning_rate": 8.11276985656248e-05,
"loss": 1.14013671875,
"step": 85
},
{
"epoch": 1.8297872340425532,
"grad_norm": 0.09725148975849152,
"learning_rate": 8.064712387912692e-05,
"loss": 1.14990234375,
"step": 86
},
{
"epoch": 1.851063829787234,
"grad_norm": 0.08682967722415924,
"learning_rate": 8.016245147149179e-05,
"loss": 1.12890625,
"step": 87
},
{
"epoch": 1.872340425531915,
"grad_norm": 0.07889049500226974,
"learning_rate": 7.967377753298695e-05,
"loss": 1.14111328125,
"step": 88
},
{
"epoch": 1.8936170212765957,
"grad_norm": 0.07676804065704346,
"learning_rate": 7.918119904804175e-05,
"loss": 1.0927734375,
"step": 89
},
{
"epoch": 1.9148936170212765,
"grad_norm": 0.08221984654664993,
"learning_rate": 7.86848137759994e-05,
"loss": 1.08740234375,
"step": 90
},
{
"epoch": 1.9361702127659575,
"grad_norm": 0.09177003800868988,
"learning_rate": 7.818472023171516e-05,
"loss": 1.220703125,
"step": 91
},
{
"epoch": 1.9574468085106385,
"grad_norm": 0.08985753357410431,
"learning_rate": 7.76810176660048e-05,
"loss": 1.1826171875,
"step": 92
},
{
"epoch": 1.978723404255319,
"grad_norm": 0.10842283815145493,
"learning_rate": 7.717380604594664e-05,
"loss": 1.08251953125,
"step": 93
},
{
"epoch": 2.0,
"grad_norm": 0.11710213124752045,
"learning_rate": 7.666318603504172e-05,
"loss": 1.10888671875,
"step": 94
},
{
"epoch": 2.0,
"eval_loss": 1.1904296875,
"eval_runtime": 0.706,
"eval_samples_per_second": 7.082,
"eval_steps_per_second": 1.416,
"step": 94
},
{
"epoch": 2.021276595744681,
"grad_norm": 0.0820951983332634,
"learning_rate": 7.614925897323572e-05,
"loss": 1.1064453125,
"step": 95
},
{
"epoch": 2.0425531914893615,
"grad_norm": 0.0934935212135315,
"learning_rate": 7.56321268568065e-05,
"loss": 1.1552734375,
"step": 96
},
{
"epoch": 2.0638297872340425,
"grad_norm": 0.09159503877162933,
"learning_rate": 7.511189231812146e-05,
"loss": 1.1943359375,
"step": 97
},
{
"epoch": 2.0851063829787235,
"grad_norm": 0.09112660586833954,
"learning_rate": 7.458865860526871e-05,
"loss": 1.12109375,
"step": 98
},
{
"epoch": 2.106382978723404,
"grad_norm": 0.08333063870668411,
"learning_rate": 7.406252956156601e-05,
"loss": 1.1064453125,
"step": 99
},
{
"epoch": 2.127659574468085,
"grad_norm": 0.10156647861003876,
"learning_rate": 7.353360960495141e-05,
"loss": 1.24609375,
"step": 100
},
{
"epoch": 2.148936170212766,
"grad_norm": 0.09100986272096634,
"learning_rate": 7.300200370726018e-05,
"loss": 1.193359375,
"step": 101
},
{
"epoch": 2.1702127659574466,
"grad_norm": 0.08234874904155731,
"learning_rate": 7.246781737339154e-05,
"loss": 1.234375,
"step": 102
},
{
"epoch": 2.1914893617021276,
"grad_norm": 0.08503551036119461,
"learning_rate": 7.193115662036964e-05,
"loss": 1.19921875,
"step": 103
},
{
"epoch": 2.2127659574468086,
"grad_norm": 0.08407233655452728,
"learning_rate": 7.139212795630301e-05,
"loss": 1.12744140625,
"step": 104
},
{
"epoch": 2.2340425531914896,
"grad_norm": 0.07890662550926208,
"learning_rate": 7.085083835924653e-05,
"loss": 1.0673828125,
"step": 105
},
{
"epoch": 2.25531914893617,
"grad_norm": 0.08644261211156845,
"learning_rate": 7.03073952559699e-05,
"loss": 1.1669921875,
"step": 106
},
{
"epoch": 2.276595744680851,
"grad_norm": 0.08978094160556793,
"learning_rate": 6.976190650063745e-05,
"loss": 1.111328125,
"step": 107
},
{
"epoch": 2.297872340425532,
"grad_norm": 0.07937248796224594,
"learning_rate": 6.921448035340281e-05,
"loss": 1.07763671875,
"step": 108
},
{
"epoch": 2.3191489361702127,
"grad_norm": 0.08262552320957184,
"learning_rate": 6.866522545892306e-05,
"loss": 1.044921875,
"step": 109
},
{
"epoch": 2.3404255319148937,
"grad_norm": 0.09170478582382202,
"learning_rate": 6.811425082479681e-05,
"loss": 1.12841796875,
"step": 110
},
{
"epoch": 2.3617021276595747,
"grad_norm": 0.0882730484008789,
"learning_rate": 6.756166579992981e-05,
"loss": 1.1640625,
"step": 111
},
{
"epoch": 2.382978723404255,
"grad_norm": 0.08479703217744827,
"learning_rate": 6.70075800528333e-05,
"loss": 1.171875,
"step": 112
},
{
"epoch": 2.404255319148936,
"grad_norm": 0.09456633031368256,
"learning_rate": 6.645210354985857e-05,
"loss": 1.19140625,
"step": 113
},
{
"epoch": 2.425531914893617,
"grad_norm": 0.09943220764398575,
"learning_rate": 6.589534653337259e-05,
"loss": 1.2080078125,
"step": 114
},
{
"epoch": 2.4468085106382977,
"grad_norm": 0.09369660168886185,
"learning_rate": 6.533741949987876e-05,
"loss": 1.2041015625,
"step": 115
},
{
"epoch": 2.4680851063829787,
"grad_norm": 0.07937616109848022,
"learning_rate": 6.477843317808742e-05,
"loss": 0.97607421875,
"step": 116
},
{
"epoch": 2.4893617021276597,
"grad_norm": 0.10216047614812851,
"learning_rate": 6.421849850693994e-05,
"loss": 1.11279296875,
"step": 117
},
{
"epoch": 2.5106382978723403,
"grad_norm": 0.08806140720844269,
"learning_rate": 6.365772661359143e-05,
"loss": 1.09521484375,
"step": 118
},
{
"epoch": 2.5319148936170213,
"grad_norm": 0.09562306106090546,
"learning_rate": 6.309622879135586e-05,
"loss": 1.17578125,
"step": 119
},
{
"epoch": 2.5531914893617023,
"grad_norm": 0.08352405577898026,
"learning_rate": 6.253411647761829e-05,
"loss": 1.03955078125,
"step": 120
},
{
"epoch": 2.574468085106383,
"grad_norm": 0.08265415579080582,
"learning_rate": 6.197150123171857e-05,
"loss": 1.0361328125,
"step": 121
},
{
"epoch": 2.595744680851064,
"grad_norm": 0.09541942179203033,
"learning_rate": 6.140849471281073e-05,
"loss": 1.115234375,
"step": 122
},
{
"epoch": 2.617021276595745,
"grad_norm": 0.0888490378856659,
"learning_rate": 6.0845208657702625e-05,
"loss": 1.05029296875,
"step": 123
},
{
"epoch": 2.6382978723404253,
"grad_norm": 0.08975188434123993,
"learning_rate": 6.028175485868002e-05,
"loss": 1.1494140625,
"step": 124
},
{
"epoch": 2.6595744680851063,
"grad_norm": 0.11306588351726532,
"learning_rate": 5.971824514132e-05,
"loss": 1.19921875,
"step": 125
},
{
"epoch": 2.6808510638297873,
"grad_norm": 0.08459223061800003,
"learning_rate": 5.91547913422974e-05,
"loss": 1.1669921875,
"step": 126
},
{
"epoch": 2.702127659574468,
"grad_norm": 0.0851873978972435,
"learning_rate": 5.859150528718926e-05,
"loss": 1.04541015625,
"step": 127
},
{
"epoch": 2.723404255319149,
"grad_norm": 0.09753607958555222,
"learning_rate": 5.8028498768281424e-05,
"loss": 1.1357421875,
"step": 128
},
{
"epoch": 2.74468085106383,
"grad_norm": 0.09338207542896271,
"learning_rate": 5.7465883522381734e-05,
"loss": 1.1796875,
"step": 129
},
{
"epoch": 2.7659574468085104,
"grad_norm": 0.10513012111186981,
"learning_rate": 5.690377120864416e-05,
"loss": 1.146484375,
"step": 130
},
{
"epoch": 2.7872340425531914,
"grad_norm": 0.10612934082746506,
"learning_rate": 5.634227338640858e-05,
"loss": 1.15576171875,
"step": 131
},
{
"epoch": 2.8085106382978724,
"grad_norm": 0.08152056485414505,
"learning_rate": 5.578150149306007e-05,
"loss": 1.1025390625,
"step": 132
},
{
"epoch": 2.829787234042553,
"grad_norm": 0.09891179949045181,
"learning_rate": 5.522156682191259e-05,
"loss": 1.11181640625,
"step": 133
},
{
"epoch": 2.851063829787234,
"grad_norm": 0.10099881142377853,
"learning_rate": 5.466258050012125e-05,
"loss": 1.09521484375,
"step": 134
},
{
"epoch": 2.872340425531915,
"grad_norm": 0.08331961184740067,
"learning_rate": 5.410465346662743e-05,
"loss": 1.10595703125,
"step": 135
},
{
"epoch": 2.8936170212765955,
"grad_norm": 0.08690949529409409,
"learning_rate": 5.354789645014144e-05,
"loss": 1.05712890625,
"step": 136
},
{
"epoch": 2.9148936170212765,
"grad_norm": 0.09597815573215485,
"learning_rate": 5.2992419947166705e-05,
"loss": 1.05224609375,
"step": 137
},
{
"epoch": 2.9361702127659575,
"grad_norm": 0.10523609071969986,
"learning_rate": 5.243833420007021e-05,
"loss": 1.1865234375,
"step": 138
},
{
"epoch": 2.9574468085106385,
"grad_norm": 0.10435938835144043,
"learning_rate": 5.188574917520321e-05,
"loss": 1.1484375,
"step": 139
},
{
"epoch": 2.978723404255319,
"grad_norm": 0.09967571496963501,
"learning_rate": 5.1334774541076946e-05,
"loss": 1.05322265625,
"step": 140
},
{
"epoch": 3.0,
"grad_norm": 0.13476115465164185,
"learning_rate": 5.078551964659721e-05,
"loss": 1.06396484375,
"step": 141
},
{
"epoch": 3.0,
"eval_loss": 1.16943359375,
"eval_runtime": 0.7052,
"eval_samples_per_second": 7.091,
"eval_steps_per_second": 1.418,
"step": 141
},
{
"epoch": 3.021276595744681,
"grad_norm": 0.09152151644229889,
"learning_rate": 5.023809349936256e-05,
"loss": 1.07373046875,
"step": 142
},
{
"epoch": 3.0425531914893615,
"grad_norm": 0.11027701199054718,
"learning_rate": 4.9692604744030106e-05,
"loss": 1.12255859375,
"step": 143
},
{
"epoch": 3.0638297872340425,
"grad_norm": 0.11380738765001297,
"learning_rate": 4.9149161640753496e-05,
"loss": 1.16015625,
"step": 144
},
{
"epoch": 3.0851063829787235,
"grad_norm": 0.10427799075841904,
"learning_rate": 4.8607872043696996e-05,
"loss": 1.08837890625,
"step": 145
},
{
"epoch": 3.106382978723404,
"grad_norm": 0.09149792045354843,
"learning_rate": 4.8068843379630396e-05,
"loss": 1.076171875,
"step": 146
},
{
"epoch": 3.127659574468085,
"grad_norm": 0.10543116927146912,
"learning_rate": 4.753218262660849e-05,
"loss": 1.2119140625,
"step": 147
},
{
"epoch": 3.148936170212766,
"grad_norm": 0.09941239655017853,
"learning_rate": 4.6997996292739833e-05,
"loss": 1.16015625,
"step": 148
},
{
"epoch": 3.1702127659574466,
"grad_norm": 0.09674631804227829,
"learning_rate": 4.646639039504861e-05,
"loss": 1.201171875,
"step": 149
},
{
"epoch": 3.1914893617021276,
"grad_norm": 0.10010098665952682,
"learning_rate": 4.5937470438434015e-05,
"loss": 1.166015625,
"step": 150
},
{
"epoch": 3.2127659574468086,
"grad_norm": 0.1016378104686737,
"learning_rate": 4.54113413947313e-05,
"loss": 1.10107421875,
"step": 151
},
{
"epoch": 3.2340425531914896,
"grad_norm": 0.08743297308683395,
"learning_rate": 4.4888107681878565e-05,
"loss": 1.04052734375,
"step": 152
},
{
"epoch": 3.25531914893617,
"grad_norm": 0.10094916075468063,
"learning_rate": 4.436787314319352e-05,
"loss": 1.1376953125,
"step": 153
},
{
"epoch": 3.276595744680851,
"grad_norm": 0.1199064627289772,
"learning_rate": 4.3850741026764285e-05,
"loss": 1.0830078125,
"step": 154
},
{
"epoch": 3.297872340425532,
"grad_norm": 0.09346320480108261,
"learning_rate": 4.33368139649583e-05,
"loss": 1.05126953125,
"step": 155
},
{
"epoch": 3.3191489361702127,
"grad_norm": 0.09531646221876144,
"learning_rate": 4.282619395405339e-05,
"loss": 1.02197265625,
"step": 156
},
{
"epoch": 3.3404255319148937,
"grad_norm": 0.09398731589317322,
"learning_rate": 4.2318982333995205e-05,
"loss": 1.10107421875,
"step": 157
},
{
"epoch": 3.3617021276595747,
"grad_norm": 0.10401123017072678,
"learning_rate": 4.181527976828485e-05,
"loss": 1.1357421875,
"step": 158
},
{
"epoch": 3.382978723404255,
"grad_norm": 0.10376442968845367,
"learning_rate": 4.131518622400061e-05,
"loss": 1.146484375,
"step": 159
},
{
"epoch": 3.404255319148936,
"grad_norm": 0.10089726001024246,
"learning_rate": 4.0818800951958255e-05,
"loss": 1.1630859375,
"step": 160
},
{
"epoch": 3.425531914893617,
"grad_norm": 0.09647805988788605,
"learning_rate": 4.032622246701306e-05,
"loss": 1.1787109375,
"step": 161
},
{
"epoch": 3.4468085106382977,
"grad_norm": 0.10889782011508942,
"learning_rate": 3.983754852850822e-05,
"loss": 1.177734375,
"step": 162
},
{
"epoch": 3.4680851063829787,
"grad_norm": 0.091889388859272,
"learning_rate": 3.93528761208731e-05,
"loss": 0.95361328125,
"step": 163
},
{
"epoch": 3.4893617021276597,
"grad_norm": 0.09745949506759644,
"learning_rate": 3.887230143437521e-05,
"loss": 1.0869140625,
"step": 164
},
{
"epoch": 3.5106382978723403,
"grad_norm": 0.09781906753778458,
"learning_rate": 3.8395919846029886e-05,
"loss": 1.068359375,
"step": 165
},
{
"epoch": 3.5319148936170213,
"grad_norm": 0.09253612905740738,
"learning_rate": 3.792382590067132e-05,
"loss": 1.15185546875,
"step": 166
},
{
"epoch": 3.5531914893617023,
"grad_norm": 0.09065967798233032,
"learning_rate": 3.7456113292188824e-05,
"loss": 1.015625,
"step": 167
},
{
"epoch": 3.574468085106383,
"grad_norm": 0.09022831171751022,
"learning_rate": 3.6992874844931816e-05,
"loss": 1.01318359375,
"step": 168
},
{
"epoch": 3.595744680851064,
"grad_norm": 0.10448946803808212,
"learning_rate": 3.653420249528754e-05,
"loss": 1.08984375,
"step": 169
},
{
"epoch": 3.617021276595745,
"grad_norm": 0.08936721831560135,
"learning_rate": 3.608018727343503e-05,
"loss": 1.02685546875,
"step": 170
},
{
"epoch": 3.6382978723404253,
"grad_norm": 0.09918684512376785,
"learning_rate": 3.563091928527869e-05,
"loss": 1.12548828125,
"step": 171
},
{
"epoch": 3.6595744680851063,
"grad_norm": 0.10508580505847931,
"learning_rate": 3.518648769456562e-05,
"loss": 1.1748046875,
"step": 172
},
{
"epoch": 3.6808510638297873,
"grad_norm": 0.09724324941635132,
"learning_rate": 3.4746980705189647e-05,
"loss": 1.142578125,
"step": 173
},
{
"epoch": 3.702127659574468,
"grad_norm": 0.10831834375858307,
"learning_rate": 3.4312485543686135e-05,
"loss": 1.02490234375,
"step": 174
},
{
"epoch": 3.723404255319149,
"grad_norm": 0.10373376309871674,
"learning_rate": 3.3883088441920475e-05,
"loss": 1.1103515625,
"step": 175
},
{
"epoch": 3.74468085106383,
"grad_norm": 0.10523442178964615,
"learning_rate": 3.3458874619974166e-05,
"loss": 1.15625,
"step": 176
},
{
"epoch": 3.7659574468085104,
"grad_norm": 0.09964374452829361,
"learning_rate": 3.303992826923171e-05,
"loss": 1.123046875,
"step": 177
},
{
"epoch": 3.7872340425531914,
"grad_norm": 0.1167880967259407,
"learning_rate": 3.262633253567154e-05,
"loss": 1.1337890625,
"step": 178
},
{
"epoch": 3.8085106382978724,
"grad_norm": 0.12453437596559525,
"learning_rate": 3.221816950336452e-05,
"loss": 1.0810546875,
"step": 179
},
{
"epoch": 3.829787234042553,
"grad_norm": 0.11341120302677155,
"learning_rate": 3.181552017818318e-05,
"loss": 1.09326171875,
"step": 180
},
{
"epoch": 3.851063829787234,
"grad_norm": 0.10341572761535645,
"learning_rate": 3.14184644717249e-05,
"loss": 1.07470703125,
"step": 181
},
{
"epoch": 3.872340425531915,
"grad_norm": 0.09896044433116913,
"learning_rate": 3.102708118545232e-05,
"loss": 1.0859375,
"step": 182
},
{
"epoch": 3.8936170212765955,
"grad_norm": 0.11353881657123566,
"learning_rate": 3.064144799505404e-05,
"loss": 1.03955078125,
"step": 183
},
{
"epoch": 3.9148936170212765,
"grad_norm": 0.10180618613958359,
"learning_rate": 3.026164143502876e-05,
"loss": 1.03369140625,
"step": 184
},
{
"epoch": 3.9361702127659575,
"grad_norm": 0.10199908912181854,
"learning_rate": 2.9887736883495883e-05,
"loss": 1.1640625,
"step": 185
},
{
"epoch": 3.9574468085106385,
"grad_norm": 0.12163249403238297,
"learning_rate": 2.9519808547235617e-05,
"loss": 1.126953125,
"step": 186
},
{
"epoch": 3.978723404255319,
"grad_norm": 0.10170035809278488,
"learning_rate": 2.9157929446961644e-05,
"loss": 1.0341796875,
"step": 187
},
{
"epoch": 4.0,
"grad_norm": 0.13411518931388855,
"learning_rate": 2.8802171402828992e-05,
"loss": 1.03759765625,
"step": 188
},
{
"epoch": 4.0,
"eval_loss": 1.1630859375,
"eval_runtime": 0.7016,
"eval_samples_per_second": 7.126,
"eval_steps_per_second": 1.425,
"step": 188
},
{
"epoch": 4.0212765957446805,
"grad_norm": 0.09304608404636383,
"learning_rate": 2.8452605020180383e-05,
"loss": 1.05712890625,
"step": 189
},
{
"epoch": 4.042553191489362,
"grad_norm": 0.10317014902830124,
"learning_rate": 2.8109299675533464e-05,
"loss": 1.10400390625,
"step": 190
},
{
"epoch": 4.0638297872340425,
"grad_norm": 0.1193184182047844,
"learning_rate": 2.777232350281218e-05,
"loss": 1.1396484375,
"step": 191
},
{
"epoch": 4.085106382978723,
"grad_norm": 0.1019568219780922,
"learning_rate": 2.7441743379824488e-05,
"loss": 1.0703125,
"step": 192
},
{
"epoch": 4.1063829787234045,
"grad_norm": 0.09876756370067596,
"learning_rate": 2.7117624914989498e-05,
"loss": 1.05908203125,
"step": 193
},
{
"epoch": 4.127659574468085,
"grad_norm": 0.11759065091609955,
"learning_rate": 2.680003243431662e-05,
"loss": 1.193359375,
"step": 194
},
{
"epoch": 4.148936170212766,
"grad_norm": 0.10515289008617401,
"learning_rate": 2.6489028968639028e-05,
"loss": 1.1435546875,
"step": 195
},
{
"epoch": 4.170212765957447,
"grad_norm": 0.1038760095834732,
"learning_rate": 2.6184676241104355e-05,
"loss": 1.1845703125,
"step": 196
},
{
"epoch": 4.191489361702128,
"grad_norm": 0.1032799556851387,
"learning_rate": 2.588703465492478e-05,
"loss": 1.1474609375,
"step": 197
},
{
"epoch": 4.212765957446808,
"grad_norm": 0.10190946608781815,
"learning_rate": 2.559616328138918e-05,
"loss": 1.08447265625,
"step": 198
},
{
"epoch": 4.23404255319149,
"grad_norm": 0.10085217654705048,
"learning_rate": 2.531211984813954e-05,
"loss": 1.0263671875,
"step": 199
},
{
"epoch": 4.25531914893617,
"grad_norm": 0.10069141536951065,
"learning_rate": 2.5034960727714097e-05,
"loss": 1.1201171875,
"step": 200
},
{
"epoch": 4.276595744680851,
"grad_norm": 0.10138094425201416,
"learning_rate": 2.4764740926359385e-05,
"loss": 1.06640625,
"step": 201
},
{
"epoch": 4.297872340425532,
"grad_norm": 0.09249741584062576,
"learning_rate": 2.450151407311343e-05,
"loss": 1.03564453125,
"step": 202
},
{
"epoch": 4.319148936170213,
"grad_norm": 0.10047057271003723,
"learning_rate": 2.4245332409162305e-05,
"loss": 1.0087890625,
"step": 203
},
{
"epoch": 4.340425531914893,
"grad_norm": 0.09843046218156815,
"learning_rate": 2.399624677747211e-05,
"loss": 1.08642578125,
"step": 204
},
{
"epoch": 4.361702127659575,
"grad_norm": 0.098610520362854,
"learning_rate": 2.3754306612698423e-05,
"loss": 1.12109375,
"step": 205
},
{
"epoch": 4.382978723404255,
"grad_norm": 0.10245000571012497,
"learning_rate": 2.3519559931375334e-05,
"loss": 1.1279296875,
"step": 206
},
{
"epoch": 4.404255319148936,
"grad_norm": 0.10176238417625427,
"learning_rate": 2.3292053322385783e-05,
"loss": 1.146484375,
"step": 207
},
{
"epoch": 4.425531914893617,
"grad_norm": 0.09995201975107193,
"learning_rate": 2.3071831937715403e-05,
"loss": 1.162109375,
"step": 208
},
{
"epoch": 4.446808510638298,
"grad_norm": 0.10793449729681015,
"learning_rate": 2.2858939483491403e-05,
"loss": 1.1611328125,
"step": 209
},
{
"epoch": 4.468085106382979,
"grad_norm": 0.1183137595653534,
"learning_rate": 2.2653418211308484e-05,
"loss": 0.93896484375,
"step": 210
},
{
"epoch": 4.48936170212766,
"grad_norm": 0.09787902981042862,
"learning_rate": 2.2455308909843364e-05,
"loss": 1.0732421875,
"step": 211
},
{
"epoch": 4.51063829787234,
"grad_norm": 0.09635757654905319,
"learning_rate": 2.2264650896759677e-05,
"loss": 1.0556640625,
"step": 212
},
{
"epoch": 4.531914893617021,
"grad_norm": 0.10069479048252106,
"learning_rate": 2.2081482010904874e-05,
"loss": 1.13623046875,
"step": 213
},
{
"epoch": 4.553191489361702,
"grad_norm": 0.09444998949766159,
"learning_rate": 2.1905838604800468e-05,
"loss": 1.001953125,
"step": 214
},
{
"epoch": 4.574468085106383,
"grad_norm": 0.09090348333120346,
"learning_rate": 2.1737755537427403e-05,
"loss": 1.0,
"step": 215
},
{
"epoch": 4.595744680851064,
"grad_norm": 0.10889911651611328,
"learning_rate": 2.1577266167307765e-05,
"loss": 1.076171875,
"step": 216
},
{
"epoch": 4.617021276595745,
"grad_norm": 0.0926787406206131,
"learning_rate": 2.1424402345884305e-05,
"loss": 1.015625,
"step": 217
},
{
"epoch": 4.638297872340425,
"grad_norm": 0.11637773364782333,
"learning_rate": 2.1279194411199068e-05,
"loss": 1.1103515625,
"step": 218
},
{
"epoch": 4.659574468085106,
"grad_norm": 0.09567286819219589,
"learning_rate": 2.114167118187233e-05,
"loss": 1.16259765625,
"step": 219
},
{
"epoch": 4.680851063829787,
"grad_norm": 0.12339469790458679,
"learning_rate": 2.1011859951383194e-05,
"loss": 1.130859375,
"step": 220
},
{
"epoch": 4.702127659574468,
"grad_norm": 0.09166810661554337,
"learning_rate": 2.088978648265273e-05,
"loss": 1.01318359375,
"step": 221
},
{
"epoch": 4.723404255319149,
"grad_norm": 0.09855899214744568,
"learning_rate": 2.0775475002930988e-05,
"loss": 1.0986328125,
"step": 222
},
{
"epoch": 4.74468085106383,
"grad_norm": 0.10451032221317291,
"learning_rate": 2.0668948198988754e-05,
"loss": 1.14404296875,
"step": 223
},
{
"epoch": 4.76595744680851,
"grad_norm": 0.10801032930612564,
"learning_rate": 2.057022721261501e-05,
"loss": 1.10888671875,
"step": 224
},
{
"epoch": 4.787234042553192,
"grad_norm": 0.10492988675832748,
"learning_rate": 2.0479331636421045e-05,
"loss": 1.12060546875,
"step": 225
},
{
"epoch": 4.808510638297872,
"grad_norm": 0.09453857690095901,
"learning_rate": 2.039627950995204e-05,
"loss": 1.0712890625,
"step": 226
},
{
"epoch": 4.829787234042553,
"grad_norm": 0.10679788887500763,
"learning_rate": 2.0321087316106823e-05,
"loss": 1.08154296875,
"step": 227
},
{
"epoch": 4.851063829787234,
"grad_norm": 0.0981895849108696,
"learning_rate": 2.025376997786663e-05,
"loss": 1.060546875,
"step": 228
},
{
"epoch": 4.872340425531915,
"grad_norm": 0.09476146847009659,
"learning_rate": 2.0194340855333396e-05,
"loss": 1.0732421875,
"step": 229
},
{
"epoch": 4.8936170212765955,
"grad_norm": 0.09872428327798843,
"learning_rate": 2.0142811743078267e-05,
"loss": 1.02783203125,
"step": 230
},
{
"epoch": 4.914893617021277,
"grad_norm": 0.10203375667333603,
"learning_rate": 2.0099192867800796e-05,
"loss": 1.0234375,
"step": 231
},
{
"epoch": 4.9361702127659575,
"grad_norm": 0.10611573606729507,
"learning_rate": 2.0063492886299294e-05,
"loss": 1.1513671875,
"step": 232
},
{
"epoch": 4.957446808510638,
"grad_norm": 0.11534586548805237,
"learning_rate": 2.0035718883752775e-05,
"loss": 1.11572265625,
"step": 233
},
{
"epoch": 4.9787234042553195,
"grad_norm": 0.09584429860115051,
"learning_rate": 2.0015876372314803e-05,
"loss": 1.0234375,
"step": 234
},
{
"epoch": 5.0,
"grad_norm": 0.14444786310195923,
"learning_rate": 2.0003969290019492e-05,
"loss": 1.0234375,
"step": 235
},
{
"epoch": 5.0,
"eval_loss": 1.15625,
"eval_runtime": 0.705,
"eval_samples_per_second": 7.092,
"eval_steps_per_second": 1.418,
"step": 235
}
],
"logging_steps": 1.0,
"max_steps": 235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 0,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.704890317986857e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}