Chwerb-AI-300 / trainer_state.json
haithem1234's picture
Upload folder using huggingface_hub
804707d verified
Raw History Blame Contribute Delete
56.1 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.1885310290652003,
"eval_steps": 500,
"global_step": 300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006284367635506677,
"grad_norm": 6.582937717437744,
"learning_rate": 0.0,
"loss": 4.603842735290527,
"step": 1
},
{
"epoch": 0.0012568735271013354,
"grad_norm": 6.695369243621826,
"learning_rate": 8.333333333333333e-07,
"loss": 4.715240955352783,
"step": 2
},
{
"epoch": 0.0018853102906520032,
"grad_norm": 6.19420862197876,
"learning_rate": 1.6666666666666667e-06,
"loss": 4.704594135284424,
"step": 3
},
{
"epoch": 0.0025137470542026708,
"grad_norm": 6.287356853485107,
"learning_rate": 2.5e-06,
"loss": 4.682320594787598,
"step": 4
},
{
"epoch": 0.003142183817753339,
"grad_norm": 5.510616302490234,
"learning_rate": 3.3333333333333333e-06,
"loss": 4.330413341522217,
"step": 5
},
{
"epoch": 0.0037706205813040064,
"grad_norm": 6.080907344818115,
"learning_rate": 4.166666666666667e-06,
"loss": 4.201181411743164,
"step": 6
},
{
"epoch": 0.004399057344854674,
"grad_norm": 5.887509822845459,
"learning_rate": 5e-06,
"loss": 4.35393762588501,
"step": 7
},
{
"epoch": 0.0050274941084053415,
"grad_norm": 5.863635063171387,
"learning_rate": 5.833333333333334e-06,
"loss": 4.161603927612305,
"step": 8
},
{
"epoch": 0.005655930871956009,
"grad_norm": 5.592325687408447,
"learning_rate": 6.666666666666667e-06,
"loss": 4.1693806648254395,
"step": 9
},
{
"epoch": 0.006284367635506678,
"grad_norm": 5.076478481292725,
"learning_rate": 7.5e-06,
"loss": 3.8860297203063965,
"step": 10
},
{
"epoch": 0.006912804399057345,
"grad_norm": 5.2288818359375,
"learning_rate": 8.333333333333334e-06,
"loss": 3.8872628211975098,
"step": 11
},
{
"epoch": 0.007541241162608013,
"grad_norm": 5.319539546966553,
"learning_rate": 9.166666666666666e-06,
"loss": 3.8444418907165527,
"step": 12
},
{
"epoch": 0.00816967792615868,
"grad_norm": 4.9492387771606445,
"learning_rate": 1e-05,
"loss": 3.6013431549072266,
"step": 13
},
{
"epoch": 0.008798114689709348,
"grad_norm": 4.738637924194336,
"learning_rate": 1.0833333333333334e-05,
"loss": 3.4792113304138184,
"step": 14
},
{
"epoch": 0.009426551453260016,
"grad_norm": 4.255862236022949,
"learning_rate": 1.1666666666666668e-05,
"loss": 3.426337718963623,
"step": 15
},
{
"epoch": 0.010054988216810683,
"grad_norm": 3.833632469177246,
"learning_rate": 1.25e-05,
"loss": 3.473579168319702,
"step": 16
},
{
"epoch": 0.010683424980361352,
"grad_norm": 3.6634793281555176,
"learning_rate": 1.3333333333333333e-05,
"loss": 3.1470890045166016,
"step": 17
},
{
"epoch": 0.011311861743912018,
"grad_norm": 3.9700350761413574,
"learning_rate": 1.4166666666666668e-05,
"loss": 2.8047187328338623,
"step": 18
},
{
"epoch": 0.011940298507462687,
"grad_norm": 3.912999153137207,
"learning_rate": 1.5e-05,
"loss": 3.0094175338745117,
"step": 19
},
{
"epoch": 0.012568735271013355,
"grad_norm": 4.881673812866211,
"learning_rate": 1.5833333333333333e-05,
"loss": 2.488555431365967,
"step": 20
},
{
"epoch": 0.013197172034564022,
"grad_norm": 4.43076753616333,
"learning_rate": 1.6666666666666667e-05,
"loss": 2.4106087684631348,
"step": 21
},
{
"epoch": 0.01382560879811469,
"grad_norm": 3.91679310798645,
"learning_rate": 1.75e-05,
"loss": 2.5461294651031494,
"step": 22
},
{
"epoch": 0.014454045561665357,
"grad_norm": 4.173760414123535,
"learning_rate": 1.8333333333333333e-05,
"loss": 2.4555070400238037,
"step": 23
},
{
"epoch": 0.015082482325216025,
"grad_norm": 4.746689319610596,
"learning_rate": 1.9166666666666667e-05,
"loss": 1.9969418048858643,
"step": 24
},
{
"epoch": 0.015710919088766692,
"grad_norm": 5.101377010345459,
"learning_rate": 2e-05,
"loss": 1.8463797569274902,
"step": 25
},
{
"epoch": 0.01633935585231736,
"grad_norm": 4.012807369232178,
"learning_rate": 2.0833333333333336e-05,
"loss": 1.6978498697280884,
"step": 26
},
{
"epoch": 0.01696779261586803,
"grad_norm": 3.5796396732330322,
"learning_rate": 2.1666666666666667e-05,
"loss": 1.6642639636993408,
"step": 27
},
{
"epoch": 0.017596229379418696,
"grad_norm": 3.8680806159973145,
"learning_rate": 2.25e-05,
"loss": 1.9199894666671753,
"step": 28
},
{
"epoch": 0.018224666142969363,
"grad_norm": 4.025764465332031,
"learning_rate": 2.3333333333333336e-05,
"loss": 1.8153927326202393,
"step": 29
},
{
"epoch": 0.018853102906520033,
"grad_norm": 4.699272632598877,
"learning_rate": 2.4166666666666667e-05,
"loss": 1.5831806659698486,
"step": 30
},
{
"epoch": 0.0194815396700707,
"grad_norm": 4.271459579467773,
"learning_rate": 2.5e-05,
"loss": 1.8813101053237915,
"step": 31
},
{
"epoch": 0.020109976433621366,
"grad_norm": 4.671807765960693,
"learning_rate": 2.5833333333333336e-05,
"loss": 1.712389349937439,
"step": 32
},
{
"epoch": 0.020738413197172033,
"grad_norm": 4.068957328796387,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.906123399734497,
"step": 33
},
{
"epoch": 0.021366849960722703,
"grad_norm": 4.866921424865723,
"learning_rate": 2.7500000000000004e-05,
"loss": 1.5461045503616333,
"step": 34
},
{
"epoch": 0.02199528672427337,
"grad_norm": 4.347884178161621,
"learning_rate": 2.8333333333333335e-05,
"loss": 1.5146393775939941,
"step": 35
},
{
"epoch": 0.022623723487824036,
"grad_norm": 3.437556266784668,
"learning_rate": 2.916666666666667e-05,
"loss": 1.5056453943252563,
"step": 36
},
{
"epoch": 0.023252160251374707,
"grad_norm": 3.0069897174835205,
"learning_rate": 3e-05,
"loss": 1.3426285982131958,
"step": 37
},
{
"epoch": 0.023880597014925373,
"grad_norm": 2.147547483444214,
"learning_rate": 3.0833333333333335e-05,
"loss": 1.4989075660705566,
"step": 38
},
{
"epoch": 0.02450903377847604,
"grad_norm": 1.5224647521972656,
"learning_rate": 3.1666666666666666e-05,
"loss": 1.3997061252593994,
"step": 39
},
{
"epoch": 0.02513747054202671,
"grad_norm": 0.9915205836296082,
"learning_rate": 3.2500000000000004e-05,
"loss": 1.2027744054794312,
"step": 40
},
{
"epoch": 0.025765907305577377,
"grad_norm": 0.903066873550415,
"learning_rate": 3.3333333333333335e-05,
"loss": 1.6686099767684937,
"step": 41
},
{
"epoch": 0.026394344069128044,
"grad_norm": 0.7829668521881104,
"learning_rate": 3.4166666666666666e-05,
"loss": 1.4906165599822998,
"step": 42
},
{
"epoch": 0.02702278083267871,
"grad_norm": 0.7567051649093628,
"learning_rate": 3.5e-05,
"loss": 1.6501740217208862,
"step": 43
},
{
"epoch": 0.02765121759622938,
"grad_norm": 0.8439779877662659,
"learning_rate": 3.5833333333333335e-05,
"loss": 1.43964684009552,
"step": 44
},
{
"epoch": 0.028279654359780047,
"grad_norm": 0.6391922831535339,
"learning_rate": 3.6666666666666666e-05,
"loss": 1.0846329927444458,
"step": 45
},
{
"epoch": 0.028908091123330714,
"grad_norm": 1.2203965187072754,
"learning_rate": 3.7500000000000003e-05,
"loss": 1.7158523797988892,
"step": 46
},
{
"epoch": 0.029536527886881384,
"grad_norm": 3.8870856761932373,
"learning_rate": 3.8333333333333334e-05,
"loss": 1.308109998703003,
"step": 47
},
{
"epoch": 0.03016496465043205,
"grad_norm": 0.7830280065536499,
"learning_rate": 3.9166666666666665e-05,
"loss": 1.1710951328277588,
"step": 48
},
{
"epoch": 0.030793401413982718,
"grad_norm": 0.8194233179092407,
"learning_rate": 4e-05,
"loss": 1.4279857873916626,
"step": 49
},
{
"epoch": 0.031421838177533384,
"grad_norm": 0.8869267106056213,
"learning_rate": 4.0833333333333334e-05,
"loss": 1.3380258083343506,
"step": 50
},
{
"epoch": 0.032050274941084055,
"grad_norm": 0.848308265209198,
"learning_rate": 4.166666666666667e-05,
"loss": 1.5400422811508179,
"step": 51
},
{
"epoch": 0.03267871170463472,
"grad_norm": 0.9038040637969971,
"learning_rate": 4.25e-05,
"loss": 1.3563334941864014,
"step": 52
},
{
"epoch": 0.03330714846818539,
"grad_norm": 0.7090128064155579,
"learning_rate": 4.3333333333333334e-05,
"loss": 1.0491596460342407,
"step": 53
},
{
"epoch": 0.03393558523173606,
"grad_norm": 0.7986202239990234,
"learning_rate": 4.4166666666666665e-05,
"loss": 1.1293988227844238,
"step": 54
},
{
"epoch": 0.03456402199528672,
"grad_norm": 0.7759880423545837,
"learning_rate": 4.5e-05,
"loss": 1.1359784603118896,
"step": 55
},
{
"epoch": 0.03519245875883739,
"grad_norm": 0.9220878481864929,
"learning_rate": 4.5833333333333334e-05,
"loss": 1.8005272150039673,
"step": 56
},
{
"epoch": 0.03582089552238806,
"grad_norm": 0.9730035662651062,
"learning_rate": 4.666666666666667e-05,
"loss": 1.770522117614746,
"step": 57
},
{
"epoch": 0.036449332285938725,
"grad_norm": 0.8269259929656982,
"learning_rate": 4.75e-05,
"loss": 1.4446560144424438,
"step": 58
},
{
"epoch": 0.037077769049489395,
"grad_norm": 0.8967049717903137,
"learning_rate": 4.8333333333333334e-05,
"loss": 1.3438148498535156,
"step": 59
},
{
"epoch": 0.037706205813040065,
"grad_norm": 0.9631702899932861,
"learning_rate": 4.9166666666666665e-05,
"loss": 1.2305524349212646,
"step": 60
},
{
"epoch": 0.03833464257659073,
"grad_norm": 0.847324788570404,
"learning_rate": 5e-05,
"loss": 1.112710952758789,
"step": 61
},
{
"epoch": 0.0389630793401414,
"grad_norm": 0.9951688647270203,
"learning_rate": 4.9956140350877195e-05,
"loss": 1.4946831464767456,
"step": 62
},
{
"epoch": 0.03959151610369207,
"grad_norm": 0.9058189392089844,
"learning_rate": 4.991228070175439e-05,
"loss": 1.2028950452804565,
"step": 63
},
{
"epoch": 0.04021995286724273,
"grad_norm": 1.0516862869262695,
"learning_rate": 4.986842105263158e-05,
"loss": 1.5654267072677612,
"step": 64
},
{
"epoch": 0.0408483896307934,
"grad_norm": 0.9898484945297241,
"learning_rate": 4.9824561403508773e-05,
"loss": 1.4516839981079102,
"step": 65
},
{
"epoch": 0.041476826394344066,
"grad_norm": 0.8785293102264404,
"learning_rate": 4.9780701754385966e-05,
"loss": 1.2362587451934814,
"step": 66
},
{
"epoch": 0.042105263157894736,
"grad_norm": 0.9406870007514954,
"learning_rate": 4.973684210526316e-05,
"loss": 1.283206820487976,
"step": 67
},
{
"epoch": 0.042733699921445406,
"grad_norm": 0.9507713317871094,
"learning_rate": 4.969298245614035e-05,
"loss": 1.2595372200012207,
"step": 68
},
{
"epoch": 0.04336213668499607,
"grad_norm": 0.9840788245201111,
"learning_rate": 4.9649122807017544e-05,
"loss": 1.2669472694396973,
"step": 69
},
{
"epoch": 0.04399057344854674,
"grad_norm": 1.0010781288146973,
"learning_rate": 4.960526315789474e-05,
"loss": 1.4899603128433228,
"step": 70
},
{
"epoch": 0.04461901021209741,
"grad_norm": 0.9709664583206177,
"learning_rate": 4.956140350877193e-05,
"loss": 1.2075190544128418,
"step": 71
},
{
"epoch": 0.04524744697564807,
"grad_norm": 0.9632940292358398,
"learning_rate": 4.951754385964912e-05,
"loss": 1.3631963729858398,
"step": 72
},
{
"epoch": 0.04587588373919874,
"grad_norm": 0.8987375497817993,
"learning_rate": 4.9473684210526315e-05,
"loss": 1.0288957357406616,
"step": 73
},
{
"epoch": 0.04650432050274941,
"grad_norm": 0.9523040056228638,
"learning_rate": 4.942982456140351e-05,
"loss": 1.4955958127975464,
"step": 74
},
{
"epoch": 0.04713275726630008,
"grad_norm": 1.038588047027588,
"learning_rate": 4.93859649122807e-05,
"loss": 1.312672734260559,
"step": 75
},
{
"epoch": 0.04776119402985075,
"grad_norm": 1.606860876083374,
"learning_rate": 4.9342105263157894e-05,
"loss": 1.4430378675460815,
"step": 76
},
{
"epoch": 0.04838963079340142,
"grad_norm": 0.937546968460083,
"learning_rate": 4.9298245614035086e-05,
"loss": 1.2572695016860962,
"step": 77
},
{
"epoch": 0.04901806755695208,
"grad_norm": 0.8281884789466858,
"learning_rate": 4.925438596491228e-05,
"loss": 1.0899391174316406,
"step": 78
},
{
"epoch": 0.04964650432050275,
"grad_norm": 0.9356880187988281,
"learning_rate": 4.921052631578947e-05,
"loss": 1.4328938722610474,
"step": 79
},
{
"epoch": 0.05027494108405342,
"grad_norm": 1.0862209796905518,
"learning_rate": 4.9166666666666665e-05,
"loss": 1.259037971496582,
"step": 80
},
{
"epoch": 0.050903377847604084,
"grad_norm": 0.8265055418014526,
"learning_rate": 4.912280701754386e-05,
"loss": 0.9618508219718933,
"step": 81
},
{
"epoch": 0.051531814611154754,
"grad_norm": 0.9510069489479065,
"learning_rate": 4.907894736842106e-05,
"loss": 1.2391194105148315,
"step": 82
},
{
"epoch": 0.05216025137470542,
"grad_norm": 0.979525089263916,
"learning_rate": 4.903508771929825e-05,
"loss": 1.4140366315841675,
"step": 83
},
{
"epoch": 0.05278868813825609,
"grad_norm": 1.0451817512512207,
"learning_rate": 4.899122807017544e-05,
"loss": 1.330329179763794,
"step": 84
},
{
"epoch": 0.05341712490180676,
"grad_norm": 1.1743742227554321,
"learning_rate": 4.8947368421052635e-05,
"loss": 1.5114189386367798,
"step": 85
},
{
"epoch": 0.05404556166535742,
"grad_norm": 0.9567965269088745,
"learning_rate": 4.890350877192983e-05,
"loss": 1.094409465789795,
"step": 86
},
{
"epoch": 0.05467399842890809,
"grad_norm": 1.0258677005767822,
"learning_rate": 4.885964912280702e-05,
"loss": 1.13893723487854,
"step": 87
},
{
"epoch": 0.05530243519245876,
"grad_norm": 1.1723560094833374,
"learning_rate": 4.881578947368421e-05,
"loss": 1.5450183153152466,
"step": 88
},
{
"epoch": 0.055930871956009424,
"grad_norm": 1.119633436203003,
"learning_rate": 4.8771929824561406e-05,
"loss": 1.1542819738388062,
"step": 89
},
{
"epoch": 0.056559308719560095,
"grad_norm": 1.0176407098770142,
"learning_rate": 4.87280701754386e-05,
"loss": 1.139894962310791,
"step": 90
},
{
"epoch": 0.057187745483110765,
"grad_norm": 1.015350341796875,
"learning_rate": 4.868421052631579e-05,
"loss": 1.2419655323028564,
"step": 91
},
{
"epoch": 0.05781618224666143,
"grad_norm": 1.2359726428985596,
"learning_rate": 4.8640350877192984e-05,
"loss": 1.658208966255188,
"step": 92
},
{
"epoch": 0.0584446190102121,
"grad_norm": 1.3616856336593628,
"learning_rate": 4.859649122807018e-05,
"loss": 1.4880285263061523,
"step": 93
},
{
"epoch": 0.05907305577376277,
"grad_norm": 1.1564569473266602,
"learning_rate": 4.855263157894737e-05,
"loss": 1.4070013761520386,
"step": 94
},
{
"epoch": 0.05970149253731343,
"grad_norm": 1.372467041015625,
"learning_rate": 4.850877192982456e-05,
"loss": 1.4177615642547607,
"step": 95
},
{
"epoch": 0.0603299293008641,
"grad_norm": 1.1938430070877075,
"learning_rate": 4.8464912280701755e-05,
"loss": 1.5566543340682983,
"step": 96
},
{
"epoch": 0.060958366064414765,
"grad_norm": 1.1628551483154297,
"learning_rate": 4.842105263157895e-05,
"loss": 1.238955020904541,
"step": 97
},
{
"epoch": 0.061586802827965435,
"grad_norm": 1.0895317792892456,
"learning_rate": 4.837719298245614e-05,
"loss": 1.0830061435699463,
"step": 98
},
{
"epoch": 0.062215239591516105,
"grad_norm": 1.1284940242767334,
"learning_rate": 4.8333333333333334e-05,
"loss": 1.3517670631408691,
"step": 99
},
{
"epoch": 0.06284367635506677,
"grad_norm": 1.072861909866333,
"learning_rate": 4.8289473684210526e-05,
"loss": 1.4782456159591675,
"step": 100
},
{
"epoch": 0.06347211311861745,
"grad_norm": 1.126695990562439,
"learning_rate": 4.824561403508772e-05,
"loss": 1.4386738538742065,
"step": 101
},
{
"epoch": 0.06410054988216811,
"grad_norm": 0.9199826121330261,
"learning_rate": 4.820175438596491e-05,
"loss": 1.0082112550735474,
"step": 102
},
{
"epoch": 0.06472898664571877,
"grad_norm": 0.9667104482650757,
"learning_rate": 4.8157894736842105e-05,
"loss": 1.1209678649902344,
"step": 103
},
{
"epoch": 0.06535742340926944,
"grad_norm": 1.1905107498168945,
"learning_rate": 4.81140350877193e-05,
"loss": 1.1586198806762695,
"step": 104
},
{
"epoch": 0.06598586017282011,
"grad_norm": 1.0409283638000488,
"learning_rate": 4.807017543859649e-05,
"loss": 1.2226606607437134,
"step": 105
},
{
"epoch": 0.06661429693637078,
"grad_norm": 0.9099530577659607,
"learning_rate": 4.802631578947368e-05,
"loss": 0.9255151748657227,
"step": 106
},
{
"epoch": 0.06724273369992144,
"grad_norm": 1.0770549774169922,
"learning_rate": 4.7982456140350876e-05,
"loss": 1.4956984519958496,
"step": 107
},
{
"epoch": 0.06787117046347212,
"grad_norm": 1.2913087606430054,
"learning_rate": 4.793859649122807e-05,
"loss": 1.203768253326416,
"step": 108
},
{
"epoch": 0.06849960722702278,
"grad_norm": 1.0569708347320557,
"learning_rate": 4.789473684210526e-05,
"loss": 1.18813955783844,
"step": 109
},
{
"epoch": 0.06912804399057344,
"grad_norm": 1.2576522827148438,
"learning_rate": 4.785087719298246e-05,
"loss": 1.2618587017059326,
"step": 110
},
{
"epoch": 0.06975648075412412,
"grad_norm": 1.0505105257034302,
"learning_rate": 4.780701754385965e-05,
"loss": 1.1373695135116577,
"step": 111
},
{
"epoch": 0.07038491751767478,
"grad_norm": 1.1691850423812866,
"learning_rate": 4.7763157894736846e-05,
"loss": 1.3569108247756958,
"step": 112
},
{
"epoch": 0.07101335428122545,
"grad_norm": 1.1929399967193604,
"learning_rate": 4.771929824561404e-05,
"loss": 1.0983715057373047,
"step": 113
},
{
"epoch": 0.07164179104477612,
"grad_norm": 1.3630225658416748,
"learning_rate": 4.767543859649123e-05,
"loss": 1.2036861181259155,
"step": 114
},
{
"epoch": 0.07227022780832679,
"grad_norm": 1.340999960899353,
"learning_rate": 4.7631578947368424e-05,
"loss": 1.6164659261703491,
"step": 115
},
{
"epoch": 0.07289866457187745,
"grad_norm": 1.5692455768585205,
"learning_rate": 4.758771929824562e-05,
"loss": 1.370465874671936,
"step": 116
},
{
"epoch": 0.07352710133542813,
"grad_norm": 1.2575063705444336,
"learning_rate": 4.754385964912281e-05,
"loss": 1.294209361076355,
"step": 117
},
{
"epoch": 0.07415553809897879,
"grad_norm": 1.3092509508132935,
"learning_rate": 4.75e-05,
"loss": 1.2222367525100708,
"step": 118
},
{
"epoch": 0.07478397486252945,
"grad_norm": 1.4458547830581665,
"learning_rate": 4.7456140350877195e-05,
"loss": 1.3677136898040771,
"step": 119
},
{
"epoch": 0.07541241162608013,
"grad_norm": 1.3372570276260376,
"learning_rate": 4.741228070175439e-05,
"loss": 1.271146297454834,
"step": 120
},
{
"epoch": 0.0760408483896308,
"grad_norm": 1.367821455001831,
"learning_rate": 4.736842105263158e-05,
"loss": 1.2260288000106812,
"step": 121
},
{
"epoch": 0.07666928515318146,
"grad_norm": 1.0059062242507935,
"learning_rate": 4.7324561403508774e-05,
"loss": 0.8689628839492798,
"step": 122
},
{
"epoch": 0.07729772191673213,
"grad_norm": 1.1224220991134644,
"learning_rate": 4.728070175438597e-05,
"loss": 1.0949623584747314,
"step": 123
},
{
"epoch": 0.0779261586802828,
"grad_norm": 1.3654863834381104,
"learning_rate": 4.7236842105263166e-05,
"loss": 1.4401837587356567,
"step": 124
},
{
"epoch": 0.07855459544383346,
"grad_norm": 1.2437621355056763,
"learning_rate": 4.719298245614036e-05,
"loss": 1.2469260692596436,
"step": 125
},
{
"epoch": 0.07918303220738414,
"grad_norm": 1.1593888998031616,
"learning_rate": 4.714912280701755e-05,
"loss": 1.1807864904403687,
"step": 126
},
{
"epoch": 0.0798114689709348,
"grad_norm": 1.1882439851760864,
"learning_rate": 4.7105263157894744e-05,
"loss": 1.1621779203414917,
"step": 127
},
{
"epoch": 0.08043990573448546,
"grad_norm": 1.0895029306411743,
"learning_rate": 4.706140350877194e-05,
"loss": 1.1161084175109863,
"step": 128
},
{
"epoch": 0.08106834249803614,
"grad_norm": 1.1587965488433838,
"learning_rate": 4.701754385964913e-05,
"loss": 1.2836101055145264,
"step": 129
},
{
"epoch": 0.0816967792615868,
"grad_norm": 1.1367735862731934,
"learning_rate": 4.697368421052632e-05,
"loss": 1.2266199588775635,
"step": 130
},
{
"epoch": 0.08232521602513747,
"grad_norm": 1.0789273977279663,
"learning_rate": 4.6929824561403515e-05,
"loss": 1.0686663389205933,
"step": 131
},
{
"epoch": 0.08295365278868813,
"grad_norm": 1.2545297145843506,
"learning_rate": 4.688596491228071e-05,
"loss": 1.2312991619110107,
"step": 132
},
{
"epoch": 0.08358208955223881,
"grad_norm": 1.220672607421875,
"learning_rate": 4.68421052631579e-05,
"loss": 1.166621446609497,
"step": 133
},
{
"epoch": 0.08421052631578947,
"grad_norm": 1.210046410560608,
"learning_rate": 4.679824561403509e-05,
"loss": 1.0885021686553955,
"step": 134
},
{
"epoch": 0.08483896307934014,
"grad_norm": 1.3004860877990723,
"learning_rate": 4.6754385964912286e-05,
"loss": 1.510931372642517,
"step": 135
},
{
"epoch": 0.08546739984289081,
"grad_norm": 1.2907251119613647,
"learning_rate": 4.671052631578948e-05,
"loss": 1.238710880279541,
"step": 136
},
{
"epoch": 0.08609583660644148,
"grad_norm": 1.1571893692016602,
"learning_rate": 4.666666666666667e-05,
"loss": 1.1060110330581665,
"step": 137
},
{
"epoch": 0.08672427336999214,
"grad_norm": 1.3332407474517822,
"learning_rate": 4.6622807017543864e-05,
"loss": 1.340341329574585,
"step": 138
},
{
"epoch": 0.08735271013354282,
"grad_norm": 1.2119009494781494,
"learning_rate": 4.657894736842106e-05,
"loss": 1.3405832052230835,
"step": 139
},
{
"epoch": 0.08798114689709348,
"grad_norm": 1.1208232641220093,
"learning_rate": 4.653508771929825e-05,
"loss": 0.9515865445137024,
"step": 140
},
{
"epoch": 0.08860958366064414,
"grad_norm": 1.2222814559936523,
"learning_rate": 4.649122807017544e-05,
"loss": 1.2669777870178223,
"step": 141
},
{
"epoch": 0.08923802042419482,
"grad_norm": 1.2387925386428833,
"learning_rate": 4.6447368421052635e-05,
"loss": 1.2471401691436768,
"step": 142
},
{
"epoch": 0.08986645718774548,
"grad_norm": 1.324230432510376,
"learning_rate": 4.640350877192983e-05,
"loss": 1.17795991897583,
"step": 143
},
{
"epoch": 0.09049489395129615,
"grad_norm": 1.3605045080184937,
"learning_rate": 4.635964912280702e-05,
"loss": 1.2591118812561035,
"step": 144
},
{
"epoch": 0.09112333071484682,
"grad_norm": 1.2737178802490234,
"learning_rate": 4.6315789473684214e-05,
"loss": 1.190453290939331,
"step": 145
},
{
"epoch": 0.09175176747839749,
"grad_norm": 1.2495275735855103,
"learning_rate": 4.6271929824561406e-05,
"loss": 1.2909812927246094,
"step": 146
},
{
"epoch": 0.09238020424194815,
"grad_norm": 1.15914785861969,
"learning_rate": 4.62280701754386e-05,
"loss": 1.1968251466751099,
"step": 147
},
{
"epoch": 0.09300864100549883,
"grad_norm": 1.1547869443893433,
"learning_rate": 4.618421052631579e-05,
"loss": 1.2185707092285156,
"step": 148
},
{
"epoch": 0.09363707776904949,
"grad_norm": 1.2578293085098267,
"learning_rate": 4.6140350877192985e-05,
"loss": 1.001325249671936,
"step": 149
},
{
"epoch": 0.09426551453260015,
"grad_norm": 1.35823655128479,
"learning_rate": 4.609649122807018e-05,
"loss": 1.4483919143676758,
"step": 150
},
{
"epoch": 0.09489395129615083,
"grad_norm": 1.308301568031311,
"learning_rate": 4.605263157894737e-05,
"loss": 1.2364054918289185,
"step": 151
},
{
"epoch": 0.0955223880597015,
"grad_norm": 1.263966679573059,
"learning_rate": 4.600877192982456e-05,
"loss": 1.3925046920776367,
"step": 152
},
{
"epoch": 0.09615082482325216,
"grad_norm": 1.1889537572860718,
"learning_rate": 4.5964912280701756e-05,
"loss": 1.4211318492889404,
"step": 153
},
{
"epoch": 0.09677926158680283,
"grad_norm": 1.1588282585144043,
"learning_rate": 4.592105263157895e-05,
"loss": 1.2376643419265747,
"step": 154
},
{
"epoch": 0.0974076983503535,
"grad_norm": 1.329275369644165,
"learning_rate": 4.587719298245614e-05,
"loss": 1.1794888973236084,
"step": 155
},
{
"epoch": 0.09803613511390416,
"grad_norm": 1.3373125791549683,
"learning_rate": 4.5833333333333334e-05,
"loss": 1.2727038860321045,
"step": 156
},
{
"epoch": 0.09866457187745484,
"grad_norm": 1.2136530876159668,
"learning_rate": 4.5789473684210527e-05,
"loss": 1.2274446487426758,
"step": 157
},
{
"epoch": 0.0992930086410055,
"grad_norm": 1.2008614540100098,
"learning_rate": 4.574561403508772e-05,
"loss": 0.9441820383071899,
"step": 158
},
{
"epoch": 0.09992144540455616,
"grad_norm": 1.2223150730133057,
"learning_rate": 4.570175438596491e-05,
"loss": 1.4267957210540771,
"step": 159
},
{
"epoch": 0.10054988216810684,
"grad_norm": 1.221267819404602,
"learning_rate": 4.5657894736842105e-05,
"loss": 1.186667799949646,
"step": 160
},
{
"epoch": 0.1011783189316575,
"grad_norm": 1.1329443454742432,
"learning_rate": 4.56140350877193e-05,
"loss": 1.0301666259765625,
"step": 161
},
{
"epoch": 0.10180675569520817,
"grad_norm": 1.1673669815063477,
"learning_rate": 4.557017543859649e-05,
"loss": 1.1144086122512817,
"step": 162
},
{
"epoch": 0.10243519245875883,
"grad_norm": 1.2523666620254517,
"learning_rate": 4.552631578947369e-05,
"loss": 1.191396713256836,
"step": 163
},
{
"epoch": 0.10306362922230951,
"grad_norm": 1.2631977796554565,
"learning_rate": 4.548245614035088e-05,
"loss": 1.0433307886123657,
"step": 164
},
{
"epoch": 0.10369206598586017,
"grad_norm": 1.1773271560668945,
"learning_rate": 4.5438596491228075e-05,
"loss": 1.1910687685012817,
"step": 165
},
{
"epoch": 0.10432050274941083,
"grad_norm": 1.259174108505249,
"learning_rate": 4.539473684210527e-05,
"loss": 1.342336893081665,
"step": 166
},
{
"epoch": 0.10494893951296151,
"grad_norm": 1.2281540632247925,
"learning_rate": 4.535087719298246e-05,
"loss": 1.2856794595718384,
"step": 167
},
{
"epoch": 0.10557737627651217,
"grad_norm": 1.2372385263442993,
"learning_rate": 4.5307017543859654e-05,
"loss": 1.3405929803848267,
"step": 168
},
{
"epoch": 0.10620581304006284,
"grad_norm": 1.2017349004745483,
"learning_rate": 4.5263157894736846e-05,
"loss": 1.0045437812805176,
"step": 169
},
{
"epoch": 0.10683424980361352,
"grad_norm": 1.186673879623413,
"learning_rate": 4.521929824561404e-05,
"loss": 1.0637688636779785,
"step": 170
},
{
"epoch": 0.10746268656716418,
"grad_norm": 1.2139995098114014,
"learning_rate": 4.517543859649123e-05,
"loss": 1.0720140933990479,
"step": 171
},
{
"epoch": 0.10809112333071484,
"grad_norm": 1.2609187364578247,
"learning_rate": 4.5131578947368425e-05,
"loss": 1.0572521686553955,
"step": 172
},
{
"epoch": 0.10871956009426552,
"grad_norm": 1.2682764530181885,
"learning_rate": 4.508771929824562e-05,
"loss": 1.2119594812393188,
"step": 173
},
{
"epoch": 0.10934799685781618,
"grad_norm": 1.3460633754730225,
"learning_rate": 4.504385964912281e-05,
"loss": 1.3025211095809937,
"step": 174
},
{
"epoch": 0.10997643362136685,
"grad_norm": 1.2264760732650757,
"learning_rate": 4.5e-05,
"loss": 1.2492730617523193,
"step": 175
},
{
"epoch": 0.11060487038491752,
"grad_norm": 1.1876654624938965,
"learning_rate": 4.4956140350877196e-05,
"loss": 1.1845881938934326,
"step": 176
},
{
"epoch": 0.11123330714846819,
"grad_norm": 1.4287419319152832,
"learning_rate": 4.491228070175439e-05,
"loss": 1.1898674964904785,
"step": 177
},
{
"epoch": 0.11186174391201885,
"grad_norm": 1.4143961668014526,
"learning_rate": 4.486842105263158e-05,
"loss": 1.231255054473877,
"step": 178
},
{
"epoch": 0.11249018067556953,
"grad_norm": 1.3170647621154785,
"learning_rate": 4.4824561403508774e-05,
"loss": 1.2355223894119263,
"step": 179
},
{
"epoch": 0.11311861743912019,
"grad_norm": 1.2761123180389404,
"learning_rate": 4.4780701754385967e-05,
"loss": 1.2717682123184204,
"step": 180
},
{
"epoch": 0.11374705420267085,
"grad_norm": 1.3970870971679688,
"learning_rate": 4.473684210526316e-05,
"loss": 1.3677964210510254,
"step": 181
},
{
"epoch": 0.11437549096622153,
"grad_norm": 1.3235961198806763,
"learning_rate": 4.469298245614035e-05,
"loss": 0.9941667914390564,
"step": 182
},
{
"epoch": 0.11500392772977219,
"grad_norm": 1.5064845085144043,
"learning_rate": 4.4649122807017545e-05,
"loss": 1.3989945650100708,
"step": 183
},
{
"epoch": 0.11563236449332286,
"grad_norm": 1.3286075592041016,
"learning_rate": 4.460526315789474e-05,
"loss": 1.102576494216919,
"step": 184
},
{
"epoch": 0.11626080125687353,
"grad_norm": 1.1819852590560913,
"learning_rate": 4.456140350877193e-05,
"loss": 0.984578549861908,
"step": 185
},
{
"epoch": 0.1168892380204242,
"grad_norm": 1.4942841529846191,
"learning_rate": 4.451754385964912e-05,
"loss": 1.1566405296325684,
"step": 186
},
{
"epoch": 0.11751767478397486,
"grad_norm": 1.3385881185531616,
"learning_rate": 4.4473684210526316e-05,
"loss": 1.17711341381073,
"step": 187
},
{
"epoch": 0.11814611154752554,
"grad_norm": 1.3535542488098145,
"learning_rate": 4.442982456140351e-05,
"loss": 1.1837196350097656,
"step": 188
},
{
"epoch": 0.1187745483110762,
"grad_norm": 1.461076021194458,
"learning_rate": 4.43859649122807e-05,
"loss": 1.388396143913269,
"step": 189
},
{
"epoch": 0.11940298507462686,
"grad_norm": 1.2865407466888428,
"learning_rate": 4.4342105263157894e-05,
"loss": 1.0031954050064087,
"step": 190
},
{
"epoch": 0.12003142183817753,
"grad_norm": 1.427472472190857,
"learning_rate": 4.429824561403509e-05,
"loss": 1.2914338111877441,
"step": 191
},
{
"epoch": 0.1206598586017282,
"grad_norm": 1.3917181491851807,
"learning_rate": 4.425438596491228e-05,
"loss": 1.1631019115447998,
"step": 192
},
{
"epoch": 0.12128829536527887,
"grad_norm": 1.4971448183059692,
"learning_rate": 4.421052631578947e-05,
"loss": 1.4322973489761353,
"step": 193
},
{
"epoch": 0.12191673212882953,
"grad_norm": 1.5286405086517334,
"learning_rate": 4.4166666666666665e-05,
"loss": 1.4280428886413574,
"step": 194
},
{
"epoch": 0.12254516889238021,
"grad_norm": 1.3017997741699219,
"learning_rate": 4.412280701754386e-05,
"loss": 1.2430918216705322,
"step": 195
},
{
"epoch": 0.12317360565593087,
"grad_norm": 1.4457416534423828,
"learning_rate": 4.407894736842105e-05,
"loss": 1.3773921728134155,
"step": 196
},
{
"epoch": 0.12380204241948153,
"grad_norm": 1.2650363445281982,
"learning_rate": 4.403508771929824e-05,
"loss": 1.2715158462524414,
"step": 197
},
{
"epoch": 0.12443047918303221,
"grad_norm": 1.2442690134048462,
"learning_rate": 4.3991228070175436e-05,
"loss": 1.180819034576416,
"step": 198
},
{
"epoch": 0.1250589159465829,
"grad_norm": 1.348306655883789,
"learning_rate": 4.394736842105263e-05,
"loss": 1.3506646156311035,
"step": 199
},
{
"epoch": 0.12568735271013354,
"grad_norm": 1.1927528381347656,
"learning_rate": 4.390350877192982e-05,
"loss": 1.0852235555648804,
"step": 200
},
{
"epoch": 0.12631578947368421,
"grad_norm": 1.4049670696258545,
"learning_rate": 4.3859649122807014e-05,
"loss": 1.3031636476516724,
"step": 201
},
{
"epoch": 0.1269442262372349,
"grad_norm": 1.2193639278411865,
"learning_rate": 4.381578947368421e-05,
"loss": 1.1517620086669922,
"step": 202
},
{
"epoch": 0.12757266300078554,
"grad_norm": 1.153424859046936,
"learning_rate": 4.37719298245614e-05,
"loss": 1.1903609037399292,
"step": 203
},
{
"epoch": 0.12820109976433622,
"grad_norm": 1.2179418802261353,
"learning_rate": 4.37280701754386e-05,
"loss": 1.1280285120010376,
"step": 204
},
{
"epoch": 0.12882953652788687,
"grad_norm": 1.1544826030731201,
"learning_rate": 4.368421052631579e-05,
"loss": 1.1437346935272217,
"step": 205
},
{
"epoch": 0.12945797329143754,
"grad_norm": 1.0923442840576172,
"learning_rate": 4.3640350877192985e-05,
"loss": 0.9591341614723206,
"step": 206
},
{
"epoch": 0.13008641005498822,
"grad_norm": 1.3880579471588135,
"learning_rate": 4.359649122807018e-05,
"loss": 1.226274847984314,
"step": 207
},
{
"epoch": 0.13071484681853887,
"grad_norm": 1.3133132457733154,
"learning_rate": 4.355263157894737e-05,
"loss": 1.1430646181106567,
"step": 208
},
{
"epoch": 0.13134328358208955,
"grad_norm": 1.5252676010131836,
"learning_rate": 4.350877192982456e-05,
"loss": 1.3102521896362305,
"step": 209
},
{
"epoch": 0.13197172034564023,
"grad_norm": 1.348258376121521,
"learning_rate": 4.3464912280701756e-05,
"loss": 0.9769993424415588,
"step": 210
},
{
"epoch": 0.13260015710919087,
"grad_norm": 1.5256011486053467,
"learning_rate": 4.342105263157895e-05,
"loss": 1.3732298612594604,
"step": 211
},
{
"epoch": 0.13322859387274155,
"grad_norm": 1.2226386070251465,
"learning_rate": 4.337719298245614e-05,
"loss": 1.1497232913970947,
"step": 212
},
{
"epoch": 0.13385703063629223,
"grad_norm": 1.8030542135238647,
"learning_rate": 4.3333333333333334e-05,
"loss": 1.2828925848007202,
"step": 213
},
{
"epoch": 0.13448546739984288,
"grad_norm": 1.0925843715667725,
"learning_rate": 4.328947368421053e-05,
"loss": 0.9497895240783691,
"step": 214
},
{
"epoch": 0.13511390416339356,
"grad_norm": 1.2209727764129639,
"learning_rate": 4.324561403508772e-05,
"loss": 1.1838711500167847,
"step": 215
},
{
"epoch": 0.13574234092694423,
"grad_norm": 1.064263105392456,
"learning_rate": 4.320175438596491e-05,
"loss": 0.8520067930221558,
"step": 216
},
{
"epoch": 0.13637077769049488,
"grad_norm": 1.1862550973892212,
"learning_rate": 4.3157894736842105e-05,
"loss": 1.0145241022109985,
"step": 217
},
{
"epoch": 0.13699921445404556,
"grad_norm": 1.1630593538284302,
"learning_rate": 4.31140350877193e-05,
"loss": 1.0773402452468872,
"step": 218
},
{
"epoch": 0.13762765121759624,
"grad_norm": 1.119510531425476,
"learning_rate": 4.30701754385965e-05,
"loss": 0.9022130370140076,
"step": 219
},
{
"epoch": 0.13825608798114689,
"grad_norm": 1.3166377544403076,
"learning_rate": 4.302631578947369e-05,
"loss": 1.128119707107544,
"step": 220
},
{
"epoch": 0.13888452474469756,
"grad_norm": 1.254410743713379,
"learning_rate": 4.298245614035088e-05,
"loss": 1.1960126161575317,
"step": 221
},
{
"epoch": 0.13951296150824824,
"grad_norm": 1.5011787414550781,
"learning_rate": 4.2938596491228076e-05,
"loss": 1.309230089187622,
"step": 222
},
{
"epoch": 0.1401413982717989,
"grad_norm": 1.3256782293319702,
"learning_rate": 4.289473684210527e-05,
"loss": 1.0180084705352783,
"step": 223
},
{
"epoch": 0.14076983503534957,
"grad_norm": 1.4042634963989258,
"learning_rate": 4.285087719298246e-05,
"loss": 1.1306205987930298,
"step": 224
},
{
"epoch": 0.14139827179890024,
"grad_norm": 1.5284535884857178,
"learning_rate": 4.2807017543859654e-05,
"loss": 1.762579321861267,
"step": 225
},
{
"epoch": 0.1420267085624509,
"grad_norm": 1.277335286140442,
"learning_rate": 4.2763157894736847e-05,
"loss": 1.2047964334487915,
"step": 226
},
{
"epoch": 0.14265514532600157,
"grad_norm": 1.2712452411651611,
"learning_rate": 4.271929824561404e-05,
"loss": 1.0682660341262817,
"step": 227
},
{
"epoch": 0.14328358208955225,
"grad_norm": 1.3352336883544922,
"learning_rate": 4.267543859649123e-05,
"loss": 1.281740427017212,
"step": 228
},
{
"epoch": 0.1439120188531029,
"grad_norm": 1.2175366878509521,
"learning_rate": 4.2631578947368425e-05,
"loss": 1.0105313062667847,
"step": 229
},
{
"epoch": 0.14454045561665357,
"grad_norm": 1.2264503240585327,
"learning_rate": 4.258771929824562e-05,
"loss": 1.157004475593567,
"step": 230
},
{
"epoch": 0.14516889238020425,
"grad_norm": 1.509265661239624,
"learning_rate": 4.254385964912281e-05,
"loss": 1.3766828775405884,
"step": 231
},
{
"epoch": 0.1457973291437549,
"grad_norm": 1.4066628217697144,
"learning_rate": 4.25e-05,
"loss": 1.239786982536316,
"step": 232
},
{
"epoch": 0.14642576590730558,
"grad_norm": 1.3470476865768433,
"learning_rate": 4.2456140350877196e-05,
"loss": 1.0104211568832397,
"step": 233
},
{
"epoch": 0.14705420267085625,
"grad_norm": 1.3273537158966064,
"learning_rate": 4.241228070175439e-05,
"loss": 1.229073405265808,
"step": 234
},
{
"epoch": 0.1476826394344069,
"grad_norm": 1.4133408069610596,
"learning_rate": 4.236842105263158e-05,
"loss": 1.2357122898101807,
"step": 235
},
{
"epoch": 0.14831107619795758,
"grad_norm": 1.3120609521865845,
"learning_rate": 4.2324561403508774e-05,
"loss": 1.2164379358291626,
"step": 236
},
{
"epoch": 0.14893951296150826,
"grad_norm": 1.3264360427856445,
"learning_rate": 4.228070175438597e-05,
"loss": 1.2308655977249146,
"step": 237
},
{
"epoch": 0.1495679497250589,
"grad_norm": 1.8172162771224976,
"learning_rate": 4.223684210526316e-05,
"loss": 1.437404751777649,
"step": 238
},
{
"epoch": 0.15019638648860958,
"grad_norm": 1.023013710975647,
"learning_rate": 4.219298245614035e-05,
"loss": 0.7946437001228333,
"step": 239
},
{
"epoch": 0.15082482325216026,
"grad_norm": 1.2326797246932983,
"learning_rate": 4.2149122807017545e-05,
"loss": 1.063747763633728,
"step": 240
},
{
"epoch": 0.1514532600157109,
"grad_norm": 1.1943098306655884,
"learning_rate": 4.210526315789474e-05,
"loss": 1.0022391080856323,
"step": 241
},
{
"epoch": 0.1520816967792616,
"grad_norm": 1.2937514781951904,
"learning_rate": 4.206140350877193e-05,
"loss": 1.1166528463363647,
"step": 242
},
{
"epoch": 0.15271013354281227,
"grad_norm": 1.1338832378387451,
"learning_rate": 4.201754385964912e-05,
"loss": 0.9130970239639282,
"step": 243
},
{
"epoch": 0.15333857030636291,
"grad_norm": 1.1659705638885498,
"learning_rate": 4.1973684210526316e-05,
"loss": 1.0536553859710693,
"step": 244
},
{
"epoch": 0.1539670070699136,
"grad_norm": 1.2916209697723389,
"learning_rate": 4.1929824561403516e-05,
"loss": 1.0721904039382935,
"step": 245
},
{
"epoch": 0.15459544383346427,
"grad_norm": 1.3614411354064941,
"learning_rate": 4.188596491228071e-05,
"loss": 1.2267605066299438,
"step": 246
},
{
"epoch": 0.15522388059701492,
"grad_norm": 1.3184254169464111,
"learning_rate": 4.18421052631579e-05,
"loss": 1.0888007879257202,
"step": 247
},
{
"epoch": 0.1558523173605656,
"grad_norm": 1.3329331874847412,
"learning_rate": 4.1798245614035094e-05,
"loss": 1.2176873683929443,
"step": 248
},
{
"epoch": 0.15648075412411627,
"grad_norm": 1.421427607536316,
"learning_rate": 4.1754385964912287e-05,
"loss": 1.3885694742202759,
"step": 249
},
{
"epoch": 0.15710919088766692,
"grad_norm": 1.5203689336776733,
"learning_rate": 4.171052631578948e-05,
"loss": 1.5506526231765747,
"step": 250
},
{
"epoch": 0.1577376276512176,
"grad_norm": 1.442990779876709,
"learning_rate": 4.166666666666667e-05,
"loss": 1.136791467666626,
"step": 251
},
{
"epoch": 0.15836606441476828,
"grad_norm": 1.4820572137832642,
"learning_rate": 4.1622807017543865e-05,
"loss": 1.3279308080673218,
"step": 252
},
{
"epoch": 0.15899450117831893,
"grad_norm": 1.3196053504943848,
"learning_rate": 4.157894736842106e-05,
"loss": 1.1103812456130981,
"step": 253
},
{
"epoch": 0.1596229379418696,
"grad_norm": 1.6428762674331665,
"learning_rate": 4.153508771929825e-05,
"loss": 1.1857703924179077,
"step": 254
},
{
"epoch": 0.16025137470542028,
"grad_norm": 1.6189777851104736,
"learning_rate": 4.149122807017544e-05,
"loss": 1.2028125524520874,
"step": 255
},
{
"epoch": 0.16087981146897093,
"grad_norm": 1.182461142539978,
"learning_rate": 4.1447368421052636e-05,
"loss": 0.8788681626319885,
"step": 256
},
{
"epoch": 0.1615082482325216,
"grad_norm": 1.3860394954681396,
"learning_rate": 4.140350877192983e-05,
"loss": 1.0995696783065796,
"step": 257
},
{
"epoch": 0.16213668499607228,
"grad_norm": 1.5398815870285034,
"learning_rate": 4.135964912280702e-05,
"loss": 1.1756093502044678,
"step": 258
},
{
"epoch": 0.16276512175962293,
"grad_norm": 1.4708590507507324,
"learning_rate": 4.1315789473684214e-05,
"loss": 1.2203091382980347,
"step": 259
},
{
"epoch": 0.1633935585231736,
"grad_norm": 1.3383499383926392,
"learning_rate": 4.127192982456141e-05,
"loss": 1.1289970874786377,
"step": 260
},
{
"epoch": 0.1640219952867243,
"grad_norm": 1.3741295337677002,
"learning_rate": 4.12280701754386e-05,
"loss": 1.086601972579956,
"step": 261
},
{
"epoch": 0.16465043205027494,
"grad_norm": 1.3370002508163452,
"learning_rate": 4.118421052631579e-05,
"loss": 1.0841387510299683,
"step": 262
},
{
"epoch": 0.1652788688138256,
"grad_norm": 1.4351061582565308,
"learning_rate": 4.1140350877192985e-05,
"loss": 1.2324360609054565,
"step": 263
},
{
"epoch": 0.16590730557737626,
"grad_norm": 1.5790796279907227,
"learning_rate": 4.109649122807018e-05,
"loss": 1.3853118419647217,
"step": 264
},
{
"epoch": 0.16653574234092694,
"grad_norm": 1.3597822189331055,
"learning_rate": 4.105263157894737e-05,
"loss": 1.1459277868270874,
"step": 265
},
{
"epoch": 0.16716417910447762,
"grad_norm": 1.1724517345428467,
"learning_rate": 4.100877192982456e-05,
"loss": 0.981572687625885,
"step": 266
},
{
"epoch": 0.16779261586802827,
"grad_norm": 1.5677584409713745,
"learning_rate": 4.0964912280701756e-05,
"loss": 1.4467698335647583,
"step": 267
},
{
"epoch": 0.16842105263157894,
"grad_norm": 1.3887449502944946,
"learning_rate": 4.092105263157895e-05,
"loss": 1.0908503532409668,
"step": 268
},
{
"epoch": 0.16904948939512962,
"grad_norm": 1.3477891683578491,
"learning_rate": 4.087719298245614e-05,
"loss": 1.0803351402282715,
"step": 269
},
{
"epoch": 0.16967792615868027,
"grad_norm": 1.3207802772521973,
"learning_rate": 4.0833333333333334e-05,
"loss": 1.1803001165390015,
"step": 270
},
{
"epoch": 0.17030636292223095,
"grad_norm": 1.2993263006210327,
"learning_rate": 4.078947368421053e-05,
"loss": 1.2406667470932007,
"step": 271
},
{
"epoch": 0.17093479968578162,
"grad_norm": 1.6038602590560913,
"learning_rate": 4.074561403508772e-05,
"loss": 1.2316172122955322,
"step": 272
},
{
"epoch": 0.17156323644933227,
"grad_norm": 1.3902779817581177,
"learning_rate": 4.070175438596491e-05,
"loss": 1.1208603382110596,
"step": 273
},
{
"epoch": 0.17219167321288295,
"grad_norm": 1.2620736360549927,
"learning_rate": 4.0657894736842105e-05,
"loss": 0.9888361692428589,
"step": 274
},
{
"epoch": 0.17282010997643363,
"grad_norm": 1.32010018825531,
"learning_rate": 4.06140350877193e-05,
"loss": 1.1029255390167236,
"step": 275
},
{
"epoch": 0.17344854673998428,
"grad_norm": 1.215519905090332,
"learning_rate": 4.057017543859649e-05,
"loss": 1.024904727935791,
"step": 276
},
{
"epoch": 0.17407698350353495,
"grad_norm": 1.2227493524551392,
"learning_rate": 4.0526315789473684e-05,
"loss": 1.1528581380844116,
"step": 277
},
{
"epoch": 0.17470542026708563,
"grad_norm": 1.288981556892395,
"learning_rate": 4.0482456140350876e-05,
"loss": 1.316483974456787,
"step": 278
},
{
"epoch": 0.17533385703063628,
"grad_norm": 1.2461531162261963,
"learning_rate": 4.043859649122807e-05,
"loss": 1.0285229682922363,
"step": 279
},
{
"epoch": 0.17596229379418696,
"grad_norm": 1.2361931800842285,
"learning_rate": 4.039473684210526e-05,
"loss": 1.1215217113494873,
"step": 280
},
{
"epoch": 0.17659073055773764,
"grad_norm": 1.5122343301773071,
"learning_rate": 4.0350877192982455e-05,
"loss": 1.2658298015594482,
"step": 281
},
{
"epoch": 0.17721916732128828,
"grad_norm": 1.3544495105743408,
"learning_rate": 4.030701754385965e-05,
"loss": 1.1877720355987549,
"step": 282
},
{
"epoch": 0.17784760408483896,
"grad_norm": 1.3129796981811523,
"learning_rate": 4.026315789473684e-05,
"loss": 1.2332338094711304,
"step": 283
},
{
"epoch": 0.17847604084838964,
"grad_norm": 1.3286563158035278,
"learning_rate": 4.021929824561403e-05,
"loss": 1.0553640127182007,
"step": 284
},
{
"epoch": 0.1791044776119403,
"grad_norm": 1.3582968711853027,
"learning_rate": 4.017543859649123e-05,
"loss": 1.267555594444275,
"step": 285
},
{
"epoch": 0.17973291437549097,
"grad_norm": 1.3733668327331543,
"learning_rate": 4.0131578947368425e-05,
"loss": 1.164658784866333,
"step": 286
},
{
"epoch": 0.18036135113904164,
"grad_norm": 1.2795850038528442,
"learning_rate": 4.008771929824562e-05,
"loss": 1.057339072227478,
"step": 287
},
{
"epoch": 0.1809897879025923,
"grad_norm": 1.3459197282791138,
"learning_rate": 4.004385964912281e-05,
"loss": 1.14511239528656,
"step": 288
},
{
"epoch": 0.18161822466614297,
"grad_norm": 1.4577739238739014,
"learning_rate": 4e-05,
"loss": 1.1687101125717163,
"step": 289
},
{
"epoch": 0.18224666142969365,
"grad_norm": 1.4675325155258179,
"learning_rate": 3.9956140350877196e-05,
"loss": 1.1327722072601318,
"step": 290
},
{
"epoch": 0.1828750981932443,
"grad_norm": 1.5075569152832031,
"learning_rate": 3.991228070175439e-05,
"loss": 1.2157615423202515,
"step": 291
},
{
"epoch": 0.18350353495679497,
"grad_norm": 1.4438762664794922,
"learning_rate": 3.986842105263158e-05,
"loss": 1.0715789794921875,
"step": 292
},
{
"epoch": 0.18413197172034565,
"grad_norm": 1.3650764226913452,
"learning_rate": 3.9824561403508774e-05,
"loss": 0.9346863031387329,
"step": 293
},
{
"epoch": 0.1847604084838963,
"grad_norm": 1.8095093965530396,
"learning_rate": 3.978070175438597e-05,
"loss": 1.3191410303115845,
"step": 294
},
{
"epoch": 0.18538884524744698,
"grad_norm": 1.24713134765625,
"learning_rate": 3.973684210526316e-05,
"loss": 0.7988866567611694,
"step": 295
},
{
"epoch": 0.18601728201099765,
"grad_norm": 1.5450276136398315,
"learning_rate": 3.969298245614035e-05,
"loss": 1.101773977279663,
"step": 296
},
{
"epoch": 0.1866457187745483,
"grad_norm": 1.3477085828781128,
"learning_rate": 3.9649122807017545e-05,
"loss": 1.0699836015701294,
"step": 297
},
{
"epoch": 0.18727415553809898,
"grad_norm": 1.278314232826233,
"learning_rate": 3.960526315789474e-05,
"loss": 1.0960936546325684,
"step": 298
},
{
"epoch": 0.18790259230164966,
"grad_norm": 1.4135265350341797,
"learning_rate": 3.956140350877193e-05,
"loss": 1.205112338066101,
"step": 299
},
{
"epoch": 0.1885310290652003,
"grad_norm": 1.5187957286834717,
"learning_rate": 3.9517543859649124e-05,
"loss": 1.256544828414917,
"step": 300
}
],
"logging_steps": 1,
"max_steps": 1200,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.041709664305152e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}