abaddon182's picture
Training in progress, step 200, checkpoint
acb066f verified
{
"best_metric": 0.1947915256023407,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.032298437563082885,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00016149218781541443,
"grad_norm": 66.04085540771484,
"learning_rate": 1e-05,
"loss": 6.8126,
"step": 1
},
{
"epoch": 0.00016149218781541443,
"eval_loss": 2.066232681274414,
"eval_runtime": 764.8511,
"eval_samples_per_second": 13.635,
"eval_steps_per_second": 3.41,
"step": 1
},
{
"epoch": 0.00032298437563082886,
"grad_norm": 58.66199493408203,
"learning_rate": 2e-05,
"loss": 5.9843,
"step": 2
},
{
"epoch": 0.0004844765634462433,
"grad_norm": 49.33218002319336,
"learning_rate": 3e-05,
"loss": 5.3096,
"step": 3
},
{
"epoch": 0.0006459687512616577,
"grad_norm": 53.177345275878906,
"learning_rate": 4e-05,
"loss": 4.4682,
"step": 4
},
{
"epoch": 0.0008074609390770722,
"grad_norm": 63.67516326904297,
"learning_rate": 5e-05,
"loss": 4.1167,
"step": 5
},
{
"epoch": 0.0009689531268924866,
"grad_norm": 27.747201919555664,
"learning_rate": 6e-05,
"loss": 3.2396,
"step": 6
},
{
"epoch": 0.001130445314707901,
"grad_norm": 52.03746795654297,
"learning_rate": 7e-05,
"loss": 2.5648,
"step": 7
},
{
"epoch": 0.0012919375025233154,
"grad_norm": 19.467458724975586,
"learning_rate": 8e-05,
"loss": 2.2277,
"step": 8
},
{
"epoch": 0.0014534296903387298,
"grad_norm": 18.454164505004883,
"learning_rate": 9e-05,
"loss": 2.119,
"step": 9
},
{
"epoch": 0.0016149218781541443,
"grad_norm": 35.559288024902344,
"learning_rate": 0.0001,
"loss": 1.6841,
"step": 10
},
{
"epoch": 0.0017764140659695586,
"grad_norm": 94.37145233154297,
"learning_rate": 9.999316524962345e-05,
"loss": 2.2293,
"step": 11
},
{
"epoch": 0.0019379062537849732,
"grad_norm": 16.54827880859375,
"learning_rate": 9.997266286704631e-05,
"loss": 1.5729,
"step": 12
},
{
"epoch": 0.0020993984416003875,
"grad_norm": 15.243042945861816,
"learning_rate": 9.993849845741524e-05,
"loss": 1.4742,
"step": 13
},
{
"epoch": 0.002260890629415802,
"grad_norm": 19.226024627685547,
"learning_rate": 9.989068136093873e-05,
"loss": 1.168,
"step": 14
},
{
"epoch": 0.0024223828172312166,
"grad_norm": 13.779980659484863,
"learning_rate": 9.98292246503335e-05,
"loss": 1.4476,
"step": 15
},
{
"epoch": 0.002583875005046631,
"grad_norm": 14.464970588684082,
"learning_rate": 9.975414512725057e-05,
"loss": 1.6151,
"step": 16
},
{
"epoch": 0.002745367192862045,
"grad_norm": 8.296271324157715,
"learning_rate": 9.966546331768191e-05,
"loss": 0.8688,
"step": 17
},
{
"epoch": 0.0029068593806774595,
"grad_norm": 11.498894691467285,
"learning_rate": 9.956320346634876e-05,
"loss": 1.0009,
"step": 18
},
{
"epoch": 0.0030683515684928743,
"grad_norm": 16.902612686157227,
"learning_rate": 9.944739353007344e-05,
"loss": 1.1112,
"step": 19
},
{
"epoch": 0.0032298437563082886,
"grad_norm": 7.251413822174072,
"learning_rate": 9.931806517013612e-05,
"loss": 0.9666,
"step": 20
},
{
"epoch": 0.003391335944123703,
"grad_norm": 27.787395477294922,
"learning_rate": 9.917525374361912e-05,
"loss": 1.344,
"step": 21
},
{
"epoch": 0.0035528281319391172,
"grad_norm": 11.747908592224121,
"learning_rate": 9.901899829374047e-05,
"loss": 0.9903,
"step": 22
},
{
"epoch": 0.003714320319754532,
"grad_norm": 10.102913856506348,
"learning_rate": 9.884934153917997e-05,
"loss": 1.2567,
"step": 23
},
{
"epoch": 0.0038758125075699463,
"grad_norm": 11.134997367858887,
"learning_rate": 9.86663298624003e-05,
"loss": 0.9622,
"step": 24
},
{
"epoch": 0.004037304695385361,
"grad_norm": 9.873003959655762,
"learning_rate": 9.847001329696653e-05,
"loss": 1.3336,
"step": 25
},
{
"epoch": 0.004198796883200775,
"grad_norm": 7.550684928894043,
"learning_rate": 9.826044551386744e-05,
"loss": 1.0543,
"step": 26
},
{
"epoch": 0.004360289071016189,
"grad_norm": 6.65726375579834,
"learning_rate": 9.803768380684242e-05,
"loss": 0.6565,
"step": 27
},
{
"epoch": 0.004521781258831604,
"grad_norm": 9.558565139770508,
"learning_rate": 9.780178907671789e-05,
"loss": 1.2839,
"step": 28
},
{
"epoch": 0.004683273446647019,
"grad_norm": 12.047572135925293,
"learning_rate": 9.755282581475769e-05,
"loss": 1.0033,
"step": 29
},
{
"epoch": 0.004844765634462433,
"grad_norm": 11.137338638305664,
"learning_rate": 9.729086208503174e-05,
"loss": 0.9529,
"step": 30
},
{
"epoch": 0.0050062578222778474,
"grad_norm": 12.697957992553711,
"learning_rate": 9.701596950580806e-05,
"loss": 1.3462,
"step": 31
},
{
"epoch": 0.005167750010093262,
"grad_norm": 8.739521980285645,
"learning_rate": 9.672822322997305e-05,
"loss": 1.0512,
"step": 32
},
{
"epoch": 0.005329242197908676,
"grad_norm": 19.00428009033203,
"learning_rate": 9.642770192448536e-05,
"loss": 1.3543,
"step": 33
},
{
"epoch": 0.00549073438572409,
"grad_norm": 13.473799705505371,
"learning_rate": 9.611448774886924e-05,
"loss": 1.4515,
"step": 34
},
{
"epoch": 0.005652226573539505,
"grad_norm": 9.601152420043945,
"learning_rate": 9.578866633275288e-05,
"loss": 1.1191,
"step": 35
},
{
"epoch": 0.005813718761354919,
"grad_norm": 10.229377746582031,
"learning_rate": 9.545032675245813e-05,
"loss": 0.9022,
"step": 36
},
{
"epoch": 0.005975210949170334,
"grad_norm": 8.150893211364746,
"learning_rate": 9.509956150664796e-05,
"loss": 1.1132,
"step": 37
},
{
"epoch": 0.0061367031369857486,
"grad_norm": 7.512841701507568,
"learning_rate": 9.473646649103818e-05,
"loss": 1.1505,
"step": 38
},
{
"epoch": 0.006298195324801163,
"grad_norm": 5.389281272888184,
"learning_rate": 9.43611409721806e-05,
"loss": 0.6357,
"step": 39
},
{
"epoch": 0.006459687512616577,
"grad_norm": 10.969232559204102,
"learning_rate": 9.397368756032445e-05,
"loss": 1.2097,
"step": 40
},
{
"epoch": 0.0066211797004319915,
"grad_norm": 7.593698501586914,
"learning_rate": 9.357421218136386e-05,
"loss": 1.1604,
"step": 41
},
{
"epoch": 0.006782671888247406,
"grad_norm": 6.020129680633545,
"learning_rate": 9.316282404787871e-05,
"loss": 0.7548,
"step": 42
},
{
"epoch": 0.00694416407606282,
"grad_norm": 8.169340133666992,
"learning_rate": 9.273963562927695e-05,
"loss": 1.0628,
"step": 43
},
{
"epoch": 0.0071056562638782345,
"grad_norm": 7.304530143737793,
"learning_rate": 9.230476262104677e-05,
"loss": 1.0682,
"step": 44
},
{
"epoch": 0.00726714845169365,
"grad_norm": 7.301529884338379,
"learning_rate": 9.185832391312644e-05,
"loss": 1.132,
"step": 45
},
{
"epoch": 0.007428640639509064,
"grad_norm": 8.042936325073242,
"learning_rate": 9.140044155740101e-05,
"loss": 0.8787,
"step": 46
},
{
"epoch": 0.007590132827324478,
"grad_norm": 8.325237274169922,
"learning_rate": 9.093124073433463e-05,
"loss": 0.9612,
"step": 47
},
{
"epoch": 0.007751625015139893,
"grad_norm": 8.121849060058594,
"learning_rate": 9.045084971874738e-05,
"loss": 1.2475,
"step": 48
},
{
"epoch": 0.007913117202955308,
"grad_norm": 7.383357048034668,
"learning_rate": 8.995939984474624e-05,
"loss": 1.2338,
"step": 49
},
{
"epoch": 0.008074609390770721,
"grad_norm": 8.456700325012207,
"learning_rate": 8.945702546981969e-05,
"loss": 1.2756,
"step": 50
},
{
"epoch": 0.008074609390770721,
"eval_loss": 0.2579116225242615,
"eval_runtime": 769.3851,
"eval_samples_per_second": 13.555,
"eval_steps_per_second": 3.39,
"step": 50
},
{
"epoch": 0.008236101578586136,
"grad_norm": 12.224557876586914,
"learning_rate": 8.894386393810563e-05,
"loss": 1.1957,
"step": 51
},
{
"epoch": 0.00839759376640155,
"grad_norm": 8.004744529724121,
"learning_rate": 8.842005554284296e-05,
"loss": 0.9128,
"step": 52
},
{
"epoch": 0.008559085954216965,
"grad_norm": 5.253309726715088,
"learning_rate": 8.788574348801675e-05,
"loss": 0.9792,
"step": 53
},
{
"epoch": 0.008720578142032379,
"grad_norm": 3.9836456775665283,
"learning_rate": 8.73410738492077e-05,
"loss": 0.8306,
"step": 54
},
{
"epoch": 0.008882070329847794,
"grad_norm": 9.755916595458984,
"learning_rate": 8.678619553365659e-05,
"loss": 1.2643,
"step": 55
},
{
"epoch": 0.009043562517663207,
"grad_norm": 3.1687428951263428,
"learning_rate": 8.622126023955446e-05,
"loss": 0.4911,
"step": 56
},
{
"epoch": 0.009205054705478622,
"grad_norm": 4.697276592254639,
"learning_rate": 8.564642241456986e-05,
"loss": 0.8635,
"step": 57
},
{
"epoch": 0.009366546893294038,
"grad_norm": 5.923514366149902,
"learning_rate": 8.506183921362443e-05,
"loss": 0.7305,
"step": 58
},
{
"epoch": 0.009528039081109451,
"grad_norm": 4.387604236602783,
"learning_rate": 8.44676704559283e-05,
"loss": 0.8195,
"step": 59
},
{
"epoch": 0.009689531268924866,
"grad_norm": 7.801131248474121,
"learning_rate": 8.386407858128706e-05,
"loss": 1.091,
"step": 60
},
{
"epoch": 0.00985102345674028,
"grad_norm": 4.7785964012146,
"learning_rate": 8.32512286056924e-05,
"loss": 0.7809,
"step": 61
},
{
"epoch": 0.010012515644555695,
"grad_norm": 6.387250900268555,
"learning_rate": 8.262928807620843e-05,
"loss": 0.8659,
"step": 62
},
{
"epoch": 0.010174007832371108,
"grad_norm": 7.59951114654541,
"learning_rate": 8.199842702516583e-05,
"loss": 0.6494,
"step": 63
},
{
"epoch": 0.010335500020186524,
"grad_norm": 8.04648208618164,
"learning_rate": 8.135881792367686e-05,
"loss": 0.8663,
"step": 64
},
{
"epoch": 0.010496992208001939,
"grad_norm": 8.109950065612793,
"learning_rate": 8.07106356344834e-05,
"loss": 1.0831,
"step": 65
},
{
"epoch": 0.010658484395817352,
"grad_norm": 6.599982261657715,
"learning_rate": 8.005405736415126e-05,
"loss": 1.1722,
"step": 66
},
{
"epoch": 0.010819976583632767,
"grad_norm": 5.933468818664551,
"learning_rate": 7.938926261462366e-05,
"loss": 1.1879,
"step": 67
},
{
"epoch": 0.01098146877144818,
"grad_norm": 6.831328868865967,
"learning_rate": 7.871643313414718e-05,
"loss": 0.7366,
"step": 68
},
{
"epoch": 0.011142960959263596,
"grad_norm": 3.481276273727417,
"learning_rate": 7.803575286758364e-05,
"loss": 0.6217,
"step": 69
},
{
"epoch": 0.01130445314707901,
"grad_norm": 5.323179721832275,
"learning_rate": 7.734740790612136e-05,
"loss": 0.9114,
"step": 70
},
{
"epoch": 0.011465945334894425,
"grad_norm": 5.245665550231934,
"learning_rate": 7.66515864363997e-05,
"loss": 0.9737,
"step": 71
},
{
"epoch": 0.011627437522709838,
"grad_norm": 5.951338768005371,
"learning_rate": 7.594847868906076e-05,
"loss": 0.9656,
"step": 72
},
{
"epoch": 0.011788929710525253,
"grad_norm": 10.116493225097656,
"learning_rate": 7.52382768867422e-05,
"loss": 1.49,
"step": 73
},
{
"epoch": 0.011950421898340668,
"grad_norm": 4.708444118499756,
"learning_rate": 7.452117519152542e-05,
"loss": 0.8565,
"step": 74
},
{
"epoch": 0.012111914086156082,
"grad_norm": 5.650876045227051,
"learning_rate": 7.379736965185368e-05,
"loss": 0.728,
"step": 75
},
{
"epoch": 0.012273406273971497,
"grad_norm": 4.68143367767334,
"learning_rate": 7.30670581489344e-05,
"loss": 0.9765,
"step": 76
},
{
"epoch": 0.01243489846178691,
"grad_norm": 5.334131240844727,
"learning_rate": 7.233044034264034e-05,
"loss": 1.3298,
"step": 77
},
{
"epoch": 0.012596390649602326,
"grad_norm": 6.564842700958252,
"learning_rate": 7.158771761692464e-05,
"loss": 1.0345,
"step": 78
},
{
"epoch": 0.01275788283741774,
"grad_norm": 4.808290958404541,
"learning_rate": 7.083909302476453e-05,
"loss": 0.8163,
"step": 79
},
{
"epoch": 0.012919375025233154,
"grad_norm": 3.600611925125122,
"learning_rate": 7.008477123264848e-05,
"loss": 0.8536,
"step": 80
},
{
"epoch": 0.01308086721304857,
"grad_norm": 5.961524486541748,
"learning_rate": 6.932495846462261e-05,
"loss": 0.8228,
"step": 81
},
{
"epoch": 0.013242359400863983,
"grad_norm": 6.8201093673706055,
"learning_rate": 6.855986244591104e-05,
"loss": 1.046,
"step": 82
},
{
"epoch": 0.013403851588679398,
"grad_norm": 5.5537028312683105,
"learning_rate": 6.778969234612584e-05,
"loss": 0.5573,
"step": 83
},
{
"epoch": 0.013565343776494812,
"grad_norm": 4.81182336807251,
"learning_rate": 6.701465872208216e-05,
"loss": 0.8525,
"step": 84
},
{
"epoch": 0.013726835964310227,
"grad_norm": 6.7080841064453125,
"learning_rate": 6.623497346023418e-05,
"loss": 1.3358,
"step": 85
},
{
"epoch": 0.01388832815212564,
"grad_norm": 4.635632038116455,
"learning_rate": 6.545084971874738e-05,
"loss": 1.0484,
"step": 86
},
{
"epoch": 0.014049820339941056,
"grad_norm": 5.049192428588867,
"learning_rate": 6.466250186922325e-05,
"loss": 0.8898,
"step": 87
},
{
"epoch": 0.014211312527756469,
"grad_norm": 6.034664154052734,
"learning_rate": 6.387014543809223e-05,
"loss": 1.0545,
"step": 88
},
{
"epoch": 0.014372804715571884,
"grad_norm": 5.578686714172363,
"learning_rate": 6.307399704769099e-05,
"loss": 0.8689,
"step": 89
},
{
"epoch": 0.0145342969033873,
"grad_norm": 5.399660110473633,
"learning_rate": 6.227427435703997e-05,
"loss": 0.9705,
"step": 90
},
{
"epoch": 0.014695789091202713,
"grad_norm": 3.986812114715576,
"learning_rate": 6.147119600233758e-05,
"loss": 0.959,
"step": 91
},
{
"epoch": 0.014857281279018128,
"grad_norm": 3.3498876094818115,
"learning_rate": 6.066498153718735e-05,
"loss": 0.6344,
"step": 92
},
{
"epoch": 0.015018773466833541,
"grad_norm": 4.480000972747803,
"learning_rate": 5.985585137257401e-05,
"loss": 0.675,
"step": 93
},
{
"epoch": 0.015180265654648957,
"grad_norm": 5.378062725067139,
"learning_rate": 5.90440267166055e-05,
"loss": 0.9703,
"step": 94
},
{
"epoch": 0.01534175784246437,
"grad_norm": 6.625367164611816,
"learning_rate": 5.8229729514036705e-05,
"loss": 1.2045,
"step": 95
},
{
"epoch": 0.015503250030279785,
"grad_norm": 7.684246063232422,
"learning_rate": 5.74131823855921e-05,
"loss": 1.2175,
"step": 96
},
{
"epoch": 0.0156647422180952,
"grad_norm": 8.043671607971191,
"learning_rate": 5.6594608567103456e-05,
"loss": 1.1995,
"step": 97
},
{
"epoch": 0.015826234405910616,
"grad_norm": 6.534055709838867,
"learning_rate": 5.577423184847932e-05,
"loss": 0.786,
"step": 98
},
{
"epoch": 0.015987726593726027,
"grad_norm": 6.658209323883057,
"learning_rate": 5.495227651252315e-05,
"loss": 0.9529,
"step": 99
},
{
"epoch": 0.016149218781541443,
"grad_norm": 6.959057331085205,
"learning_rate": 5.4128967273616625e-05,
"loss": 1.1648,
"step": 100
},
{
"epoch": 0.016149218781541443,
"eval_loss": 0.24134224653244019,
"eval_runtime": 769.3709,
"eval_samples_per_second": 13.555,
"eval_steps_per_second": 3.39,
"step": 100
},
{
"epoch": 0.016310710969356858,
"grad_norm": 4.9987311363220215,
"learning_rate": 5.330452921628497e-05,
"loss": 0.6489,
"step": 101
},
{
"epoch": 0.016472203157172273,
"grad_norm": 4.759182929992676,
"learning_rate": 5.247918773366112e-05,
"loss": 0.6947,
"step": 102
},
{
"epoch": 0.016633695344987685,
"grad_norm": 3.8671958446502686,
"learning_rate": 5.165316846586541e-05,
"loss": 0.6864,
"step": 103
},
{
"epoch": 0.0167951875328031,
"grad_norm": 4.098058700561523,
"learning_rate": 5.0826697238317935e-05,
"loss": 0.7082,
"step": 104
},
{
"epoch": 0.016956679720618515,
"grad_norm": 6.332775592803955,
"learning_rate": 5e-05,
"loss": 0.9335,
"step": 105
},
{
"epoch": 0.01711817190843393,
"grad_norm": 5.647943019866943,
"learning_rate": 4.917330276168208e-05,
"loss": 1.0795,
"step": 106
},
{
"epoch": 0.017279664096249345,
"grad_norm": 6.842463970184326,
"learning_rate": 4.834683153413459e-05,
"loss": 0.8744,
"step": 107
},
{
"epoch": 0.017441156284064757,
"grad_norm": 3.7888684272766113,
"learning_rate": 4.7520812266338885e-05,
"loss": 0.9048,
"step": 108
},
{
"epoch": 0.017602648471880172,
"grad_norm": 4.186934471130371,
"learning_rate": 4.669547078371504e-05,
"loss": 0.9536,
"step": 109
},
{
"epoch": 0.017764140659695588,
"grad_norm": 3.675734758377075,
"learning_rate": 4.5871032726383386e-05,
"loss": 0.7804,
"step": 110
},
{
"epoch": 0.017925632847511003,
"grad_norm": 10.108420372009277,
"learning_rate": 4.504772348747687e-05,
"loss": 1.2615,
"step": 111
},
{
"epoch": 0.018087125035326414,
"grad_norm": 4.677420139312744,
"learning_rate": 4.4225768151520694e-05,
"loss": 0.7271,
"step": 112
},
{
"epoch": 0.01824861722314183,
"grad_norm": 6.296616077423096,
"learning_rate": 4.3405391432896555e-05,
"loss": 1.4184,
"step": 113
},
{
"epoch": 0.018410109410957245,
"grad_norm": 3.8865997791290283,
"learning_rate": 4.2586817614407895e-05,
"loss": 0.6349,
"step": 114
},
{
"epoch": 0.01857160159877266,
"grad_norm": 4.712935447692871,
"learning_rate": 4.17702704859633e-05,
"loss": 0.8963,
"step": 115
},
{
"epoch": 0.018733093786588075,
"grad_norm": 4.991330146789551,
"learning_rate": 4.095597328339452e-05,
"loss": 0.8648,
"step": 116
},
{
"epoch": 0.018894585974403487,
"grad_norm": 4.175543308258057,
"learning_rate": 4.0144148627425993e-05,
"loss": 0.7846,
"step": 117
},
{
"epoch": 0.019056078162218902,
"grad_norm": 3.322171926498413,
"learning_rate": 3.933501846281267e-05,
"loss": 0.7698,
"step": 118
},
{
"epoch": 0.019217570350034317,
"grad_norm": 3.053602457046509,
"learning_rate": 3.852880399766243e-05,
"loss": 0.6644,
"step": 119
},
{
"epoch": 0.019379062537849732,
"grad_norm": 4.954108715057373,
"learning_rate": 3.772572564296005e-05,
"loss": 1.1093,
"step": 120
},
{
"epoch": 0.019540554725665148,
"grad_norm": 3.911067485809326,
"learning_rate": 3.6926002952309016e-05,
"loss": 0.6895,
"step": 121
},
{
"epoch": 0.01970204691348056,
"grad_norm": 5.648387432098389,
"learning_rate": 3.612985456190778e-05,
"loss": 1.2158,
"step": 122
},
{
"epoch": 0.019863539101295975,
"grad_norm": 4.296334266662598,
"learning_rate": 3.533749813077677e-05,
"loss": 0.7483,
"step": 123
},
{
"epoch": 0.02002503128911139,
"grad_norm": 4.768459320068359,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.9063,
"step": 124
},
{
"epoch": 0.020186523476926805,
"grad_norm": 4.852877616882324,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.6464,
"step": 125
},
{
"epoch": 0.020348015664742217,
"grad_norm": 3.614675521850586,
"learning_rate": 3.298534127791785e-05,
"loss": 0.714,
"step": 126
},
{
"epoch": 0.020509507852557632,
"grad_norm": 3.900026798248291,
"learning_rate": 3.221030765387417e-05,
"loss": 0.7799,
"step": 127
},
{
"epoch": 0.020671000040373047,
"grad_norm": 4.189906120300293,
"learning_rate": 3.144013755408895e-05,
"loss": 0.7379,
"step": 128
},
{
"epoch": 0.020832492228188462,
"grad_norm": 3.848201274871826,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.6732,
"step": 129
},
{
"epoch": 0.020993984416003877,
"grad_norm": 5.243552207946777,
"learning_rate": 2.991522876735154e-05,
"loss": 0.8563,
"step": 130
},
{
"epoch": 0.02115547660381929,
"grad_norm": 5.969999313354492,
"learning_rate": 2.916090697523549e-05,
"loss": 1.2854,
"step": 131
},
{
"epoch": 0.021316968791634704,
"grad_norm": 5.380910873413086,
"learning_rate": 2.8412282383075363e-05,
"loss": 1.0254,
"step": 132
},
{
"epoch": 0.02147846097945012,
"grad_norm": 4.808969974517822,
"learning_rate": 2.766955965735968e-05,
"loss": 1.0162,
"step": 133
},
{
"epoch": 0.021639953167265535,
"grad_norm": 4.541471481323242,
"learning_rate": 2.693294185106562e-05,
"loss": 1.1123,
"step": 134
},
{
"epoch": 0.021801445355080946,
"grad_norm": 3.330897331237793,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.6584,
"step": 135
},
{
"epoch": 0.02196293754289636,
"grad_norm": 3.167905807495117,
"learning_rate": 2.547882480847461e-05,
"loss": 0.618,
"step": 136
},
{
"epoch": 0.022124429730711777,
"grad_norm": 6.538463115692139,
"learning_rate": 2.476172311325783e-05,
"loss": 1.2219,
"step": 137
},
{
"epoch": 0.022285921918527192,
"grad_norm": 3.817915678024292,
"learning_rate": 2.405152131093926e-05,
"loss": 0.7884,
"step": 138
},
{
"epoch": 0.022447414106342607,
"grad_norm": 3.567233085632324,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.6219,
"step": 139
},
{
"epoch": 0.02260890629415802,
"grad_norm": 7.849123001098633,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.7381,
"step": 140
},
{
"epoch": 0.022770398481973434,
"grad_norm": 4.571730136871338,
"learning_rate": 2.196424713241637e-05,
"loss": 0.7234,
"step": 141
},
{
"epoch": 0.02293189066978885,
"grad_norm": 4.714327335357666,
"learning_rate": 2.128356686585282e-05,
"loss": 0.56,
"step": 142
},
{
"epoch": 0.023093382857604264,
"grad_norm": 3.729198455810547,
"learning_rate": 2.061073738537635e-05,
"loss": 0.7249,
"step": 143
},
{
"epoch": 0.023254875045419676,
"grad_norm": 3.5794527530670166,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.6373,
"step": 144
},
{
"epoch": 0.02341636723323509,
"grad_norm": 7.21630859375,
"learning_rate": 1.928936436551661e-05,
"loss": 0.9469,
"step": 145
},
{
"epoch": 0.023577859421050507,
"grad_norm": 5.132265090942383,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.8685,
"step": 146
},
{
"epoch": 0.023739351608865922,
"grad_norm": 6.083048343658447,
"learning_rate": 1.800157297483417e-05,
"loss": 0.8698,
"step": 147
},
{
"epoch": 0.023900843796681337,
"grad_norm": 4.920403003692627,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.7564,
"step": 148
},
{
"epoch": 0.02406233598449675,
"grad_norm": 10.307737350463867,
"learning_rate": 1.6748771394307585e-05,
"loss": 1.1976,
"step": 149
},
{
"epoch": 0.024223828172312164,
"grad_norm": 9.064608573913574,
"learning_rate": 1.6135921418712956e-05,
"loss": 1.8906,
"step": 150
},
{
"epoch": 0.024223828172312164,
"eval_loss": 0.2087429314851761,
"eval_runtime": 769.5887,
"eval_samples_per_second": 13.551,
"eval_steps_per_second": 3.389,
"step": 150
},
{
"epoch": 0.02438532036012758,
"grad_norm": 3.442979097366333,
"learning_rate": 1.553232954407171e-05,
"loss": 0.5201,
"step": 151
},
{
"epoch": 0.024546812547942994,
"grad_norm": 5.38673210144043,
"learning_rate": 1.4938160786375572e-05,
"loss": 0.9532,
"step": 152
},
{
"epoch": 0.024708304735758406,
"grad_norm": 4.110217571258545,
"learning_rate": 1.435357758543015e-05,
"loss": 0.6819,
"step": 153
},
{
"epoch": 0.02486979692357382,
"grad_norm": 5.627509593963623,
"learning_rate": 1.3778739760445552e-05,
"loss": 1.3099,
"step": 154
},
{
"epoch": 0.025031289111389236,
"grad_norm": 4.468465328216553,
"learning_rate": 1.3213804466343421e-05,
"loss": 0.6302,
"step": 155
},
{
"epoch": 0.02519278129920465,
"grad_norm": 5.513147830963135,
"learning_rate": 1.2658926150792322e-05,
"loss": 0.7853,
"step": 156
},
{
"epoch": 0.025354273487020067,
"grad_norm": 5.395639896392822,
"learning_rate": 1.2114256511983274e-05,
"loss": 0.9788,
"step": 157
},
{
"epoch": 0.02551576567483548,
"grad_norm": 4.572993278503418,
"learning_rate": 1.157994445715706e-05,
"loss": 1.0696,
"step": 158
},
{
"epoch": 0.025677257862650894,
"grad_norm": 3.8697474002838135,
"learning_rate": 1.1056136061894384e-05,
"loss": 0.6712,
"step": 159
},
{
"epoch": 0.02583875005046631,
"grad_norm": 3.169379711151123,
"learning_rate": 1.0542974530180327e-05,
"loss": 0.6972,
"step": 160
},
{
"epoch": 0.026000242238281724,
"grad_norm": 2.9400551319122314,
"learning_rate": 1.0040600155253765e-05,
"loss": 0.5203,
"step": 161
},
{
"epoch": 0.02616173442609714,
"grad_norm": 4.197710037231445,
"learning_rate": 9.549150281252633e-06,
"loss": 0.5885,
"step": 162
},
{
"epoch": 0.02632322661391255,
"grad_norm": 4.708157539367676,
"learning_rate": 9.068759265665384e-06,
"loss": 1.189,
"step": 163
},
{
"epoch": 0.026484718801727966,
"grad_norm": 3.5737617015838623,
"learning_rate": 8.599558442598998e-06,
"loss": 0.7108,
"step": 164
},
{
"epoch": 0.02664621098954338,
"grad_norm": 4.992034435272217,
"learning_rate": 8.141676086873572e-06,
"loss": 1.0945,
"step": 165
},
{
"epoch": 0.026807703177358796,
"grad_norm": 5.485217094421387,
"learning_rate": 7.695237378953223e-06,
"loss": 1.0337,
"step": 166
},
{
"epoch": 0.026969195365174208,
"grad_norm": 5.019131183624268,
"learning_rate": 7.260364370723044e-06,
"loss": 1.0977,
"step": 167
},
{
"epoch": 0.027130687552989623,
"grad_norm": 3.8975744247436523,
"learning_rate": 6.837175952121306e-06,
"loss": 0.8933,
"step": 168
},
{
"epoch": 0.02729217974080504,
"grad_norm": 2.8902029991149902,
"learning_rate": 6.425787818636131e-06,
"loss": 0.4471,
"step": 169
},
{
"epoch": 0.027453671928620454,
"grad_norm": 3.9361510276794434,
"learning_rate": 6.026312439675552e-06,
"loss": 0.7946,
"step": 170
},
{
"epoch": 0.02761516411643587,
"grad_norm": 4.5258049964904785,
"learning_rate": 5.6388590278194096e-06,
"loss": 1.0502,
"step": 171
},
{
"epoch": 0.02777665630425128,
"grad_norm": 5.8037943840026855,
"learning_rate": 5.263533508961827e-06,
"loss": 1.2896,
"step": 172
},
{
"epoch": 0.027938148492066696,
"grad_norm": 3.9754068851470947,
"learning_rate": 4.900438493352055e-06,
"loss": 0.771,
"step": 173
},
{
"epoch": 0.02809964067988211,
"grad_norm": 4.887168884277344,
"learning_rate": 4.549673247541875e-06,
"loss": 0.8007,
"step": 174
},
{
"epoch": 0.028261132867697526,
"grad_norm": 4.25801420211792,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.8883,
"step": 175
},
{
"epoch": 0.028422625055512938,
"grad_norm": 3.7918171882629395,
"learning_rate": 3.885512251130763e-06,
"loss": 0.7367,
"step": 176
},
{
"epoch": 0.028584117243328353,
"grad_norm": 5.017009735107422,
"learning_rate": 3.5722980755146517e-06,
"loss": 1.0301,
"step": 177
},
{
"epoch": 0.02874560943114377,
"grad_norm": 3.8177950382232666,
"learning_rate": 3.271776770026963e-06,
"loss": 0.858,
"step": 178
},
{
"epoch": 0.028907101618959184,
"grad_norm": 3.8096659183502197,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.6779,
"step": 179
},
{
"epoch": 0.0290685938067746,
"grad_norm": 3.4850761890411377,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.7676,
"step": 180
},
{
"epoch": 0.02923008599459001,
"grad_norm": 5.1498494148254395,
"learning_rate": 2.4471741852423237e-06,
"loss": 1.237,
"step": 181
},
{
"epoch": 0.029391578182405426,
"grad_norm": 3.281991720199585,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.7704,
"step": 182
},
{
"epoch": 0.02955307037022084,
"grad_norm": 3.717332601547241,
"learning_rate": 1.962316193157593e-06,
"loss": 0.7846,
"step": 183
},
{
"epoch": 0.029714562558036256,
"grad_norm": 3.9455020427703857,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.7002,
"step": 184
},
{
"epoch": 0.029876054745851668,
"grad_norm": 3.668330430984497,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.4983,
"step": 185
},
{
"epoch": 0.030037546933667083,
"grad_norm": 3.9461002349853516,
"learning_rate": 1.333670137599713e-06,
"loss": 0.9362,
"step": 186
},
{
"epoch": 0.030199039121482498,
"grad_norm": 4.843723773956299,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.9576,
"step": 187
},
{
"epoch": 0.030360531309297913,
"grad_norm": 3.4709835052490234,
"learning_rate": 9.810017062595322e-07,
"loss": 0.5871,
"step": 188
},
{
"epoch": 0.03052202349711333,
"grad_norm": 4.323427200317383,
"learning_rate": 8.247462563808817e-07,
"loss": 0.9341,
"step": 189
},
{
"epoch": 0.03068351568492874,
"grad_norm": 7.534937858581543,
"learning_rate": 6.819348298638839e-07,
"loss": 1.0517,
"step": 190
},
{
"epoch": 0.030845007872744155,
"grad_norm": 9.260086059570312,
"learning_rate": 5.526064699265753e-07,
"loss": 1.5213,
"step": 191
},
{
"epoch": 0.03100650006055957,
"grad_norm": 4.949000358581543,
"learning_rate": 4.367965336512403e-07,
"loss": 1.0168,
"step": 192
},
{
"epoch": 0.031167992248374986,
"grad_norm": 5.339089870452881,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.9305,
"step": 193
},
{
"epoch": 0.0313294844361904,
"grad_norm": 4.739263534545898,
"learning_rate": 2.458548727494292e-07,
"loss": 0.6322,
"step": 194
},
{
"epoch": 0.031490976624005816,
"grad_norm": 5.8488335609436035,
"learning_rate": 1.7077534966650766e-07,
"loss": 1.1369,
"step": 195
},
{
"epoch": 0.03165246881182123,
"grad_norm": 3.8751072883605957,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.8728,
"step": 196
},
{
"epoch": 0.03181396099963664,
"grad_norm": 5.9032745361328125,
"learning_rate": 6.150154258476315e-08,
"loss": 1.215,
"step": 197
},
{
"epoch": 0.031975453187452055,
"grad_norm": 3.801668643951416,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.6435,
"step": 198
},
{
"epoch": 0.03213694537526747,
"grad_norm": 3.6923794746398926,
"learning_rate": 6.834750376549792e-09,
"loss": 0.6713,
"step": 199
},
{
"epoch": 0.032298437563082885,
"grad_norm": 8.89438533782959,
"learning_rate": 0.0,
"loss": 1.257,
"step": 200
},
{
"epoch": 0.032298437563082885,
"eval_loss": 0.1947915256023407,
"eval_runtime": 769.4243,
"eval_samples_per_second": 13.554,
"eval_steps_per_second": 3.39,
"step": 200
}
],
"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": true
},
"attributes": {}
}
},
"total_flos": 2.861999239200768e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}