arcwarden46's picture
Training in progress, step 150, checkpoint
f93d778 verified
raw
history blame
27.9 kB
{
"best_metric": 10.302956581115723,
"best_model_checkpoint": "miner_id_24/checkpoint-150",
"epoch": 0.03545051698670606,
"eval_steps": 50,
"global_step": 150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0002363367799113737,
"grad_norm": 0.6855791211128235,
"learning_rate": 1e-05,
"loss": 41.5264,
"step": 1
},
{
"epoch": 0.0002363367799113737,
"eval_loss": 10.377434730529785,
"eval_runtime": 14.0654,
"eval_samples_per_second": 506.703,
"eval_steps_per_second": 126.693,
"step": 1
},
{
"epoch": 0.0004726735598227474,
"grad_norm": 0.7774749994277954,
"learning_rate": 2e-05,
"loss": 41.5087,
"step": 2
},
{
"epoch": 0.0007090103397341211,
"grad_norm": 0.7203883528709412,
"learning_rate": 3e-05,
"loss": 41.5092,
"step": 3
},
{
"epoch": 0.0009453471196454948,
"grad_norm": 0.9194475412368774,
"learning_rate": 4e-05,
"loss": 41.4963,
"step": 4
},
{
"epoch": 0.0011816838995568684,
"grad_norm": 0.8236255645751953,
"learning_rate": 5e-05,
"loss": 41.5004,
"step": 5
},
{
"epoch": 0.0014180206794682422,
"grad_norm": 0.8352706432342529,
"learning_rate": 6e-05,
"loss": 41.5035,
"step": 6
},
{
"epoch": 0.001654357459379616,
"grad_norm": 0.783135712146759,
"learning_rate": 7e-05,
"loss": 41.4943,
"step": 7
},
{
"epoch": 0.0018906942392909896,
"grad_norm": 0.8593385219573975,
"learning_rate": 8e-05,
"loss": 41.4956,
"step": 8
},
{
"epoch": 0.0021270310192023634,
"grad_norm": 0.7282395362854004,
"learning_rate": 9e-05,
"loss": 41.5,
"step": 9
},
{
"epoch": 0.002363367799113737,
"grad_norm": 0.7706413865089417,
"learning_rate": 0.0001,
"loss": 41.4767,
"step": 10
},
{
"epoch": 0.002599704579025111,
"grad_norm": 0.831785261631012,
"learning_rate": 9.999316524962345e-05,
"loss": 41.4771,
"step": 11
},
{
"epoch": 0.0028360413589364844,
"grad_norm": 0.8828088045120239,
"learning_rate": 9.997266286704631e-05,
"loss": 41.4981,
"step": 12
},
{
"epoch": 0.0030723781388478583,
"grad_norm": 0.9311338067054749,
"learning_rate": 9.993849845741524e-05,
"loss": 41.4736,
"step": 13
},
{
"epoch": 0.003308714918759232,
"grad_norm": 0.9374275803565979,
"learning_rate": 9.989068136093873e-05,
"loss": 41.4564,
"step": 14
},
{
"epoch": 0.0035450516986706058,
"grad_norm": 1.2535253763198853,
"learning_rate": 9.98292246503335e-05,
"loss": 41.4439,
"step": 15
},
{
"epoch": 0.0037813884785819793,
"grad_norm": 1.1903295516967773,
"learning_rate": 9.975414512725057e-05,
"loss": 41.4627,
"step": 16
},
{
"epoch": 0.004017725258493353,
"grad_norm": 1.0873545408248901,
"learning_rate": 9.966546331768191e-05,
"loss": 41.4695,
"step": 17
},
{
"epoch": 0.004254062038404727,
"grad_norm": 1.307550072669983,
"learning_rate": 9.956320346634876e-05,
"loss": 41.4553,
"step": 18
},
{
"epoch": 0.0044903988183161,
"grad_norm": 1.0924787521362305,
"learning_rate": 9.944739353007344e-05,
"loss": 41.4305,
"step": 19
},
{
"epoch": 0.004726735598227474,
"grad_norm": 1.4635841846466064,
"learning_rate": 9.931806517013612e-05,
"loss": 41.4392,
"step": 20
},
{
"epoch": 0.004963072378138848,
"grad_norm": 1.5859185457229614,
"learning_rate": 9.917525374361912e-05,
"loss": 41.4098,
"step": 21
},
{
"epoch": 0.005199409158050222,
"grad_norm": 1.5835672616958618,
"learning_rate": 9.901899829374047e-05,
"loss": 41.4195,
"step": 22
},
{
"epoch": 0.005435745937961595,
"grad_norm": 1.5444989204406738,
"learning_rate": 9.884934153917997e-05,
"loss": 41.4233,
"step": 23
},
{
"epoch": 0.005672082717872969,
"grad_norm": 1.6059609651565552,
"learning_rate": 9.86663298624003e-05,
"loss": 41.4173,
"step": 24
},
{
"epoch": 0.005908419497784343,
"grad_norm": 1.460196614265442,
"learning_rate": 9.847001329696653e-05,
"loss": 41.4259,
"step": 25
},
{
"epoch": 0.006144756277695717,
"grad_norm": 1.8095245361328125,
"learning_rate": 9.826044551386744e-05,
"loss": 41.3885,
"step": 26
},
{
"epoch": 0.00638109305760709,
"grad_norm": 1.8408820629119873,
"learning_rate": 9.803768380684242e-05,
"loss": 41.3851,
"step": 27
},
{
"epoch": 0.006617429837518464,
"grad_norm": 1.849540114402771,
"learning_rate": 9.780178907671789e-05,
"loss": 41.3981,
"step": 28
},
{
"epoch": 0.006853766617429837,
"grad_norm": 2.062612533569336,
"learning_rate": 9.755282581475769e-05,
"loss": 41.401,
"step": 29
},
{
"epoch": 0.0070901033973412115,
"grad_norm": 1.838928461074829,
"learning_rate": 9.729086208503174e-05,
"loss": 41.3591,
"step": 30
},
{
"epoch": 0.007326440177252585,
"grad_norm": 1.7389370203018188,
"learning_rate": 9.701596950580806e-05,
"loss": 41.3756,
"step": 31
},
{
"epoch": 0.0075627769571639586,
"grad_norm": 1.8914003372192383,
"learning_rate": 9.672822322997305e-05,
"loss": 41.3296,
"step": 32
},
{
"epoch": 0.007799113737075332,
"grad_norm": 1.6221506595611572,
"learning_rate": 9.642770192448536e-05,
"loss": 41.3345,
"step": 33
},
{
"epoch": 0.008035450516986706,
"grad_norm": 1.8179858922958374,
"learning_rate": 9.611448774886924e-05,
"loss": 41.324,
"step": 34
},
{
"epoch": 0.00827178729689808,
"grad_norm": 1.6034808158874512,
"learning_rate": 9.578866633275288e-05,
"loss": 41.3101,
"step": 35
},
{
"epoch": 0.008508124076809453,
"grad_norm": 1.5293911695480347,
"learning_rate": 9.545032675245813e-05,
"loss": 41.3234,
"step": 36
},
{
"epoch": 0.008744460856720827,
"grad_norm": 1.5270508527755737,
"learning_rate": 9.509956150664796e-05,
"loss": 41.2781,
"step": 37
},
{
"epoch": 0.0089807976366322,
"grad_norm": 1.3063689470291138,
"learning_rate": 9.473646649103818e-05,
"loss": 41.302,
"step": 38
},
{
"epoch": 0.009217134416543574,
"grad_norm": 1.4343252182006836,
"learning_rate": 9.43611409721806e-05,
"loss": 41.2907,
"step": 39
},
{
"epoch": 0.009453471196454948,
"grad_norm": 1.2640676498413086,
"learning_rate": 9.397368756032445e-05,
"loss": 41.3028,
"step": 40
},
{
"epoch": 0.009689807976366323,
"grad_norm": 1.1832501888275146,
"learning_rate": 9.357421218136386e-05,
"loss": 41.299,
"step": 41
},
{
"epoch": 0.009926144756277696,
"grad_norm": 1.1388680934906006,
"learning_rate": 9.316282404787871e-05,
"loss": 41.2568,
"step": 42
},
{
"epoch": 0.01016248153618907,
"grad_norm": 1.110303521156311,
"learning_rate": 9.273963562927695e-05,
"loss": 41.2381,
"step": 43
},
{
"epoch": 0.010398818316100443,
"grad_norm": 1.0398728847503662,
"learning_rate": 9.230476262104677e-05,
"loss": 41.2495,
"step": 44
},
{
"epoch": 0.010635155096011817,
"grad_norm": 0.9440134167671204,
"learning_rate": 9.185832391312644e-05,
"loss": 41.2392,
"step": 45
},
{
"epoch": 0.01087149187592319,
"grad_norm": 0.9111349582672119,
"learning_rate": 9.140044155740101e-05,
"loss": 41.2446,
"step": 46
},
{
"epoch": 0.011107828655834564,
"grad_norm": 0.9278951287269592,
"learning_rate": 9.093124073433463e-05,
"loss": 41.237,
"step": 47
},
{
"epoch": 0.011344165435745937,
"grad_norm": 0.7140603065490723,
"learning_rate": 9.045084971874738e-05,
"loss": 41.2872,
"step": 48
},
{
"epoch": 0.011580502215657311,
"grad_norm": 0.8889676332473755,
"learning_rate": 8.995939984474624e-05,
"loss": 41.2582,
"step": 49
},
{
"epoch": 0.011816838995568686,
"grad_norm": 1.084532380104065,
"learning_rate": 8.945702546981969e-05,
"loss": 41.2216,
"step": 50
},
{
"epoch": 0.011816838995568686,
"eval_loss": 10.309220314025879,
"eval_runtime": 14.0346,
"eval_samples_per_second": 507.815,
"eval_steps_per_second": 126.972,
"step": 50
},
{
"epoch": 0.01205317577548006,
"grad_norm": 0.8177053332328796,
"learning_rate": 8.894386393810563e-05,
"loss": 41.2913,
"step": 51
},
{
"epoch": 0.012289512555391433,
"grad_norm": 0.7033486366271973,
"learning_rate": 8.842005554284296e-05,
"loss": 41.2543,
"step": 52
},
{
"epoch": 0.012525849335302807,
"grad_norm": 1.0284385681152344,
"learning_rate": 8.788574348801675e-05,
"loss": 41.2326,
"step": 53
},
{
"epoch": 0.01276218611521418,
"grad_norm": 0.772871732711792,
"learning_rate": 8.73410738492077e-05,
"loss": 41.2613,
"step": 54
},
{
"epoch": 0.012998522895125554,
"grad_norm": 0.7103416919708252,
"learning_rate": 8.678619553365659e-05,
"loss": 41.2451,
"step": 55
},
{
"epoch": 0.013234859675036927,
"grad_norm": 0.9658076167106628,
"learning_rate": 8.622126023955446e-05,
"loss": 41.228,
"step": 56
},
{
"epoch": 0.0134711964549483,
"grad_norm": 1.05716872215271,
"learning_rate": 8.564642241456986e-05,
"loss": 41.2597,
"step": 57
},
{
"epoch": 0.013707533234859674,
"grad_norm": 0.61740642786026,
"learning_rate": 8.506183921362443e-05,
"loss": 41.235,
"step": 58
},
{
"epoch": 0.01394387001477105,
"grad_norm": 0.5879736542701721,
"learning_rate": 8.44676704559283e-05,
"loss": 41.2239,
"step": 59
},
{
"epoch": 0.014180206794682423,
"grad_norm": 0.7688696384429932,
"learning_rate": 8.386407858128706e-05,
"loss": 41.2473,
"step": 60
},
{
"epoch": 0.014416543574593797,
"grad_norm": 0.5945973992347717,
"learning_rate": 8.32512286056924e-05,
"loss": 41.2215,
"step": 61
},
{
"epoch": 0.01465288035450517,
"grad_norm": 0.5151475667953491,
"learning_rate": 8.262928807620843e-05,
"loss": 41.208,
"step": 62
},
{
"epoch": 0.014889217134416544,
"grad_norm": 0.7458901405334473,
"learning_rate": 8.199842702516583e-05,
"loss": 41.243,
"step": 63
},
{
"epoch": 0.015125553914327917,
"grad_norm": 0.6170288324356079,
"learning_rate": 8.135881792367686e-05,
"loss": 41.2167,
"step": 64
},
{
"epoch": 0.01536189069423929,
"grad_norm": 0.7434762120246887,
"learning_rate": 8.07106356344834e-05,
"loss": 41.2176,
"step": 65
},
{
"epoch": 0.015598227474150664,
"grad_norm": 0.5235951542854309,
"learning_rate": 8.005405736415126e-05,
"loss": 41.2029,
"step": 66
},
{
"epoch": 0.015834564254062038,
"grad_norm": 0.8569870591163635,
"learning_rate": 7.938926261462366e-05,
"loss": 41.2486,
"step": 67
},
{
"epoch": 0.016070901033973413,
"grad_norm": 0.7099596858024597,
"learning_rate": 7.871643313414718e-05,
"loss": 41.2315,
"step": 68
},
{
"epoch": 0.016307237813884785,
"grad_norm": 0.5294151902198792,
"learning_rate": 7.803575286758364e-05,
"loss": 41.2379,
"step": 69
},
{
"epoch": 0.01654357459379616,
"grad_norm": 0.6385646462440491,
"learning_rate": 7.734740790612136e-05,
"loss": 41.2244,
"step": 70
},
{
"epoch": 0.01677991137370753,
"grad_norm": 0.5439596772193909,
"learning_rate": 7.66515864363997e-05,
"loss": 41.19,
"step": 71
},
{
"epoch": 0.017016248153618907,
"grad_norm": 0.690276563167572,
"learning_rate": 7.594847868906076e-05,
"loss": 41.1941,
"step": 72
},
{
"epoch": 0.017252584933530282,
"grad_norm": 0.7714937329292297,
"learning_rate": 7.52382768867422e-05,
"loss": 41.1957,
"step": 73
},
{
"epoch": 0.017488921713441654,
"grad_norm": 0.4136044383049011,
"learning_rate": 7.452117519152542e-05,
"loss": 41.2347,
"step": 74
},
{
"epoch": 0.01772525849335303,
"grad_norm": 0.6972559094429016,
"learning_rate": 7.379736965185368e-05,
"loss": 41.2204,
"step": 75
},
{
"epoch": 0.0179615952732644,
"grad_norm": 0.49755576252937317,
"learning_rate": 7.30670581489344e-05,
"loss": 41.2282,
"step": 76
},
{
"epoch": 0.018197932053175776,
"grad_norm": 0.4795960783958435,
"learning_rate": 7.233044034264034e-05,
"loss": 41.2386,
"step": 77
},
{
"epoch": 0.018434268833087148,
"grad_norm": 0.856358528137207,
"learning_rate": 7.158771761692464e-05,
"loss": 41.2183,
"step": 78
},
{
"epoch": 0.018670605612998523,
"grad_norm": 0.6066249012947083,
"learning_rate": 7.083909302476453e-05,
"loss": 41.1974,
"step": 79
},
{
"epoch": 0.018906942392909895,
"grad_norm": 0.4587751030921936,
"learning_rate": 7.008477123264848e-05,
"loss": 41.2179,
"step": 80
},
{
"epoch": 0.01914327917282127,
"grad_norm": 0.6782584190368652,
"learning_rate": 6.932495846462261e-05,
"loss": 41.2085,
"step": 81
},
{
"epoch": 0.019379615952732646,
"grad_norm": 0.43092355132102966,
"learning_rate": 6.855986244591104e-05,
"loss": 41.2086,
"step": 82
},
{
"epoch": 0.019615952732644017,
"grad_norm": 0.5649749040603638,
"learning_rate": 6.778969234612584e-05,
"loss": 41.1887,
"step": 83
},
{
"epoch": 0.019852289512555393,
"grad_norm": 0.5035609602928162,
"learning_rate": 6.701465872208216e-05,
"loss": 41.2125,
"step": 84
},
{
"epoch": 0.020088626292466764,
"grad_norm": 0.5236511826515198,
"learning_rate": 6.623497346023418e-05,
"loss": 41.2211,
"step": 85
},
{
"epoch": 0.02032496307237814,
"grad_norm": 0.42666125297546387,
"learning_rate": 6.545084971874738e-05,
"loss": 41.2114,
"step": 86
},
{
"epoch": 0.02056129985228951,
"grad_norm": 0.5870954990386963,
"learning_rate": 6.466250186922325e-05,
"loss": 41.1914,
"step": 87
},
{
"epoch": 0.020797636632200887,
"grad_norm": 0.4540001153945923,
"learning_rate": 6.387014543809223e-05,
"loss": 41.2223,
"step": 88
},
{
"epoch": 0.02103397341211226,
"grad_norm": 0.47619494795799255,
"learning_rate": 6.307399704769099e-05,
"loss": 41.2383,
"step": 89
},
{
"epoch": 0.021270310192023634,
"grad_norm": 0.4194251596927643,
"learning_rate": 6.227427435703997e-05,
"loss": 41.2252,
"step": 90
},
{
"epoch": 0.02150664697193501,
"grad_norm": 0.5211244821548462,
"learning_rate": 6.147119600233758e-05,
"loss": 41.2258,
"step": 91
},
{
"epoch": 0.02174298375184638,
"grad_norm": 0.5141422748565674,
"learning_rate": 6.066498153718735e-05,
"loss": 41.2133,
"step": 92
},
{
"epoch": 0.021979320531757756,
"grad_norm": 0.4102741479873657,
"learning_rate": 5.985585137257401e-05,
"loss": 41.235,
"step": 93
},
{
"epoch": 0.022215657311669128,
"grad_norm": 0.5932872891426086,
"learning_rate": 5.90440267166055e-05,
"loss": 41.2422,
"step": 94
},
{
"epoch": 0.022451994091580503,
"grad_norm": 0.7138688564300537,
"learning_rate": 5.8229729514036705e-05,
"loss": 41.1864,
"step": 95
},
{
"epoch": 0.022688330871491875,
"grad_norm": 0.5744657516479492,
"learning_rate": 5.74131823855921e-05,
"loss": 41.2133,
"step": 96
},
{
"epoch": 0.02292466765140325,
"grad_norm": 0.7336076498031616,
"learning_rate": 5.6594608567103456e-05,
"loss": 41.2483,
"step": 97
},
{
"epoch": 0.023161004431314622,
"grad_norm": 0.6511675715446472,
"learning_rate": 5.577423184847932e-05,
"loss": 41.2562,
"step": 98
},
{
"epoch": 0.023397341211225997,
"grad_norm": 0.8956031799316406,
"learning_rate": 5.495227651252315e-05,
"loss": 41.2185,
"step": 99
},
{
"epoch": 0.023633677991137372,
"grad_norm": 1.2400425672531128,
"learning_rate": 5.4128967273616625e-05,
"loss": 41.1847,
"step": 100
},
{
"epoch": 0.023633677991137372,
"eval_loss": 10.304642677307129,
"eval_runtime": 14.0917,
"eval_samples_per_second": 505.759,
"eval_steps_per_second": 126.458,
"step": 100
},
{
"epoch": 0.023870014771048744,
"grad_norm": 0.6042443513870239,
"learning_rate": 5.330452921628497e-05,
"loss": 41.2659,
"step": 101
},
{
"epoch": 0.02410635155096012,
"grad_norm": 0.6423136591911316,
"learning_rate": 5.247918773366112e-05,
"loss": 41.2584,
"step": 102
},
{
"epoch": 0.02434268833087149,
"grad_norm": 0.8272608518600464,
"learning_rate": 5.165316846586541e-05,
"loss": 41.264,
"step": 103
},
{
"epoch": 0.024579025110782866,
"grad_norm": 0.5923576951026917,
"learning_rate": 5.0826697238317935e-05,
"loss": 41.2572,
"step": 104
},
{
"epoch": 0.024815361890694238,
"grad_norm": 0.5755631327629089,
"learning_rate": 5e-05,
"loss": 41.2527,
"step": 105
},
{
"epoch": 0.025051698670605613,
"grad_norm": 0.8759294152259827,
"learning_rate": 4.917330276168208e-05,
"loss": 41.231,
"step": 106
},
{
"epoch": 0.025288035450516985,
"grad_norm": 0.7109207510948181,
"learning_rate": 4.834683153413459e-05,
"loss": 41.2625,
"step": 107
},
{
"epoch": 0.02552437223042836,
"grad_norm": 0.9061951637268066,
"learning_rate": 4.7520812266338885e-05,
"loss": 41.2426,
"step": 108
},
{
"epoch": 0.025760709010339736,
"grad_norm": 0.5387170910835266,
"learning_rate": 4.669547078371504e-05,
"loss": 41.2334,
"step": 109
},
{
"epoch": 0.025997045790251107,
"grad_norm": 0.6056496500968933,
"learning_rate": 4.5871032726383386e-05,
"loss": 41.222,
"step": 110
},
{
"epoch": 0.026233382570162483,
"grad_norm": 0.5027456283569336,
"learning_rate": 4.504772348747687e-05,
"loss": 41.229,
"step": 111
},
{
"epoch": 0.026469719350073855,
"grad_norm": 0.4441814422607422,
"learning_rate": 4.4225768151520694e-05,
"loss": 41.2496,
"step": 112
},
{
"epoch": 0.02670605612998523,
"grad_norm": 0.5234106183052063,
"learning_rate": 4.3405391432896555e-05,
"loss": 41.1787,
"step": 113
},
{
"epoch": 0.0269423929098966,
"grad_norm": 0.6285588145256042,
"learning_rate": 4.2586817614407895e-05,
"loss": 41.2019,
"step": 114
},
{
"epoch": 0.027178729689807977,
"grad_norm": 0.7189363837242126,
"learning_rate": 4.17702704859633e-05,
"loss": 41.2289,
"step": 115
},
{
"epoch": 0.02741506646971935,
"grad_norm": 0.4671284854412079,
"learning_rate": 4.095597328339452e-05,
"loss": 41.2176,
"step": 116
},
{
"epoch": 0.027651403249630724,
"grad_norm": 0.7947689294815063,
"learning_rate": 4.0144148627425993e-05,
"loss": 41.2367,
"step": 117
},
{
"epoch": 0.0278877400295421,
"grad_norm": 0.7902439832687378,
"learning_rate": 3.933501846281267e-05,
"loss": 41.2048,
"step": 118
},
{
"epoch": 0.02812407680945347,
"grad_norm": 0.4855709969997406,
"learning_rate": 3.852880399766243e-05,
"loss": 41.2063,
"step": 119
},
{
"epoch": 0.028360413589364846,
"grad_norm": 0.5833091139793396,
"learning_rate": 3.772572564296005e-05,
"loss": 41.2372,
"step": 120
},
{
"epoch": 0.028596750369276218,
"grad_norm": 0.7160181999206543,
"learning_rate": 3.6926002952309016e-05,
"loss": 41.2106,
"step": 121
},
{
"epoch": 0.028833087149187593,
"grad_norm": 0.6571805477142334,
"learning_rate": 3.612985456190778e-05,
"loss": 41.2137,
"step": 122
},
{
"epoch": 0.029069423929098965,
"grad_norm": 0.5013795495033264,
"learning_rate": 3.533749813077677e-05,
"loss": 41.1875,
"step": 123
},
{
"epoch": 0.02930576070901034,
"grad_norm": 0.5372549295425415,
"learning_rate": 3.4549150281252636e-05,
"loss": 41.197,
"step": 124
},
{
"epoch": 0.029542097488921712,
"grad_norm": 0.3943471908569336,
"learning_rate": 3.3765026539765834e-05,
"loss": 41.2067,
"step": 125
},
{
"epoch": 0.029778434268833087,
"grad_norm": 0.5975308418273926,
"learning_rate": 3.298534127791785e-05,
"loss": 41.2444,
"step": 126
},
{
"epoch": 0.030014771048744462,
"grad_norm": 0.5731347799301147,
"learning_rate": 3.221030765387417e-05,
"loss": 41.2314,
"step": 127
},
{
"epoch": 0.030251107828655834,
"grad_norm": 0.67473304271698,
"learning_rate": 3.144013755408895e-05,
"loss": 41.2306,
"step": 128
},
{
"epoch": 0.03048744460856721,
"grad_norm": 0.5370274782180786,
"learning_rate": 3.0675041535377405e-05,
"loss": 41.1631,
"step": 129
},
{
"epoch": 0.03072378138847858,
"grad_norm": 0.5157413482666016,
"learning_rate": 2.991522876735154e-05,
"loss": 41.1876,
"step": 130
},
{
"epoch": 0.030960118168389957,
"grad_norm": 0.4812129735946655,
"learning_rate": 2.916090697523549e-05,
"loss": 41.2219,
"step": 131
},
{
"epoch": 0.03119645494830133,
"grad_norm": 0.7001417875289917,
"learning_rate": 2.8412282383075363e-05,
"loss": 41.2037,
"step": 132
},
{
"epoch": 0.0314327917282127,
"grad_norm": 0.5039445161819458,
"learning_rate": 2.766955965735968e-05,
"loss": 41.2263,
"step": 133
},
{
"epoch": 0.031669128508124075,
"grad_norm": 0.5022205114364624,
"learning_rate": 2.693294185106562e-05,
"loss": 41.2047,
"step": 134
},
{
"epoch": 0.03190546528803545,
"grad_norm": 0.4550672173500061,
"learning_rate": 2.6202630348146324e-05,
"loss": 41.2202,
"step": 135
},
{
"epoch": 0.032141802067946826,
"grad_norm": 0.5127445459365845,
"learning_rate": 2.547882480847461e-05,
"loss": 41.212,
"step": 136
},
{
"epoch": 0.0323781388478582,
"grad_norm": 0.43551337718963623,
"learning_rate": 2.476172311325783e-05,
"loss": 41.2019,
"step": 137
},
{
"epoch": 0.03261447562776957,
"grad_norm": 0.5555402636528015,
"learning_rate": 2.405152131093926e-05,
"loss": 41.1991,
"step": 138
},
{
"epoch": 0.032850812407680945,
"grad_norm": 0.5227714776992798,
"learning_rate": 2.3348413563600325e-05,
"loss": 41.2078,
"step": 139
},
{
"epoch": 0.03308714918759232,
"grad_norm": 0.46548959612846375,
"learning_rate": 2.2652592093878666e-05,
"loss": 41.2231,
"step": 140
},
{
"epoch": 0.033323485967503695,
"grad_norm": 0.4519064426422119,
"learning_rate": 2.196424713241637e-05,
"loss": 41.2182,
"step": 141
},
{
"epoch": 0.03355982274741506,
"grad_norm": 0.5990817546844482,
"learning_rate": 2.128356686585282e-05,
"loss": 41.1755,
"step": 142
},
{
"epoch": 0.03379615952732644,
"grad_norm": 0.43311795592308044,
"learning_rate": 2.061073738537635e-05,
"loss": 41.2137,
"step": 143
},
{
"epoch": 0.034032496307237814,
"grad_norm": 0.44968536496162415,
"learning_rate": 1.9945942635848748e-05,
"loss": 41.2196,
"step": 144
},
{
"epoch": 0.03426883308714919,
"grad_norm": 0.5897616147994995,
"learning_rate": 1.928936436551661e-05,
"loss": 41.2188,
"step": 145
},
{
"epoch": 0.034505169867060564,
"grad_norm": 0.5212418437004089,
"learning_rate": 1.8641182076323148e-05,
"loss": 41.2309,
"step": 146
},
{
"epoch": 0.03474150664697193,
"grad_norm": 0.5393649339675903,
"learning_rate": 1.800157297483417e-05,
"loss": 41.217,
"step": 147
},
{
"epoch": 0.03497784342688331,
"grad_norm": 0.5517672300338745,
"learning_rate": 1.7370711923791567e-05,
"loss": 41.1982,
"step": 148
},
{
"epoch": 0.03521418020679468,
"grad_norm": 0.7273964285850525,
"learning_rate": 1.6748771394307585e-05,
"loss": 41.1897,
"step": 149
},
{
"epoch": 0.03545051698670606,
"grad_norm": 1.0823817253112793,
"learning_rate": 1.6135921418712956e-05,
"loss": 41.1503,
"step": 150
},
{
"epoch": 0.03545051698670606,
"eval_loss": 10.302956581115723,
"eval_runtime": 14.0592,
"eval_samples_per_second": 506.928,
"eval_steps_per_second": 126.75,
"step": 150
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 32597293400064.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}