Spark2Scale_v3 / checkpoint-200 /trainer_state.json
Dohahemdann's picture
Checkpoint at step 200
4c53fe6 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.1383365035448729,
"eval_steps": 500,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006916825177243646,
"grad_norm": 0.2954893708229065,
"learning_rate": 0.0,
"loss": 2.3812,
"step": 1
},
{
"epoch": 0.0013833650354487291,
"grad_norm": 0.3372870683670044,
"learning_rate": 1.3698630136986302e-06,
"loss": 2.6562,
"step": 2
},
{
"epoch": 0.0020750475531730937,
"grad_norm": 0.41708794236183167,
"learning_rate": 2.7397260273972604e-06,
"loss": 2.779,
"step": 3
},
{
"epoch": 0.0027667300708974583,
"grad_norm": 0.48284393548965454,
"learning_rate": 4.10958904109589e-06,
"loss": 2.83,
"step": 4
},
{
"epoch": 0.0034584125886218224,
"grad_norm": 0.328521728515625,
"learning_rate": 5.479452054794521e-06,
"loss": 2.3607,
"step": 5
},
{
"epoch": 0.004150095106346187,
"grad_norm": 0.3413754403591156,
"learning_rate": 6.849315068493151e-06,
"loss": 1.9744,
"step": 6
},
{
"epoch": 0.0048417776240705515,
"grad_norm": 0.43425995111465454,
"learning_rate": 8.21917808219178e-06,
"loss": 1.915,
"step": 7
},
{
"epoch": 0.0055334601417949165,
"grad_norm": 0.45184195041656494,
"learning_rate": 9.589041095890411e-06,
"loss": 2.0002,
"step": 8
},
{
"epoch": 0.006225142659519281,
"grad_norm": 0.3451146185398102,
"learning_rate": 1.0958904109589042e-05,
"loss": 2.5739,
"step": 9
},
{
"epoch": 0.006916825177243645,
"grad_norm": 0.39400503039360046,
"learning_rate": 1.2328767123287671e-05,
"loss": 2.6139,
"step": 10
},
{
"epoch": 0.00760850769496801,
"grad_norm": 0.4555888772010803,
"learning_rate": 1.3698630136986302e-05,
"loss": 2.4346,
"step": 11
},
{
"epoch": 0.008300190212692375,
"grad_norm": 0.5475619435310364,
"learning_rate": 1.5068493150684931e-05,
"loss": 1.485,
"step": 12
},
{
"epoch": 0.008991872730416739,
"grad_norm": 1.219076156616211,
"learning_rate": 1.643835616438356e-05,
"loss": 2.0555,
"step": 13
},
{
"epoch": 0.009683555248141103,
"grad_norm": 0.3271433711051941,
"learning_rate": 1.780821917808219e-05,
"loss": 2.8418,
"step": 14
},
{
"epoch": 0.010375237765865467,
"grad_norm": 0.3261394500732422,
"learning_rate": 1.9178082191780822e-05,
"loss": 2.287,
"step": 15
},
{
"epoch": 0.011066920283589833,
"grad_norm": 0.39713796973228455,
"learning_rate": 2.0547945205479453e-05,
"loss": 2.7331,
"step": 16
},
{
"epoch": 0.011758602801314197,
"grad_norm": 0.3226444125175476,
"learning_rate": 2.1917808219178083e-05,
"loss": 2.7307,
"step": 17
},
{
"epoch": 0.012450285319038561,
"grad_norm": 0.4153122007846832,
"learning_rate": 2.328767123287671e-05,
"loss": 2.3229,
"step": 18
},
{
"epoch": 0.013141967836762926,
"grad_norm": 0.29302462935447693,
"learning_rate": 2.4657534246575342e-05,
"loss": 2.2093,
"step": 19
},
{
"epoch": 0.01383365035448729,
"grad_norm": 0.42535701394081116,
"learning_rate": 2.6027397260273973e-05,
"loss": 2.3698,
"step": 20
},
{
"epoch": 0.014525332872211656,
"grad_norm": 0.33567023277282715,
"learning_rate": 2.7397260273972603e-05,
"loss": 2.6456,
"step": 21
},
{
"epoch": 0.01521701538993602,
"grad_norm": 0.37375620007514954,
"learning_rate": 2.8767123287671234e-05,
"loss": 2.4148,
"step": 22
},
{
"epoch": 0.015908697907660384,
"grad_norm": 0.2704203128814697,
"learning_rate": 3.0136986301369862e-05,
"loss": 1.6722,
"step": 23
},
{
"epoch": 0.01660038042538475,
"grad_norm": 0.3946782946586609,
"learning_rate": 3.1506849315068496e-05,
"loss": 2.6759,
"step": 24
},
{
"epoch": 0.017292062943109112,
"grad_norm": 0.368335098028183,
"learning_rate": 3.287671232876712e-05,
"loss": 2.6882,
"step": 25
},
{
"epoch": 0.017983745460833478,
"grad_norm": 0.38029783964157104,
"learning_rate": 3.424657534246575e-05,
"loss": 2.6915,
"step": 26
},
{
"epoch": 0.01867542797855784,
"grad_norm": 0.36253222823143005,
"learning_rate": 3.561643835616438e-05,
"loss": 2.1426,
"step": 27
},
{
"epoch": 0.019367110496282206,
"grad_norm": 0.35769322514533997,
"learning_rate": 3.698630136986301e-05,
"loss": 2.6634,
"step": 28
},
{
"epoch": 0.020058793014006572,
"grad_norm": 0.44577229022979736,
"learning_rate": 3.8356164383561644e-05,
"loss": 2.5996,
"step": 29
},
{
"epoch": 0.020750475531730934,
"grad_norm": 0.5225628614425659,
"learning_rate": 3.9726027397260274e-05,
"loss": 2.5993,
"step": 30
},
{
"epoch": 0.0214421580494553,
"grad_norm": 0.45649707317352295,
"learning_rate": 4.1095890410958905e-05,
"loss": 2.3344,
"step": 31
},
{
"epoch": 0.022133840567179666,
"grad_norm": 0.31408172845840454,
"learning_rate": 4.2465753424657536e-05,
"loss": 2.343,
"step": 32
},
{
"epoch": 0.02282552308490403,
"grad_norm": 0.3498693108558655,
"learning_rate": 4.383561643835617e-05,
"loss": 1.7196,
"step": 33
},
{
"epoch": 0.023517205602628394,
"grad_norm": 0.5255292654037476,
"learning_rate": 4.520547945205479e-05,
"loss": 1.5354,
"step": 34
},
{
"epoch": 0.024208888120352757,
"grad_norm": 0.3460487425327301,
"learning_rate": 4.657534246575342e-05,
"loss": 2.1435,
"step": 35
},
{
"epoch": 0.024900570638077123,
"grad_norm": 0.47466233372688293,
"learning_rate": 4.794520547945205e-05,
"loss": 2.7025,
"step": 36
},
{
"epoch": 0.02559225315580149,
"grad_norm": 0.39667466282844543,
"learning_rate": 4.9315068493150684e-05,
"loss": 2.2736,
"step": 37
},
{
"epoch": 0.02628393567352585,
"grad_norm": 0.40798133611679077,
"learning_rate": 5.068493150684932e-05,
"loss": 1.4074,
"step": 38
},
{
"epoch": 0.026975618191250217,
"grad_norm": 0.39603471755981445,
"learning_rate": 5.2054794520547945e-05,
"loss": 2.4354,
"step": 39
},
{
"epoch": 0.02766730070897458,
"grad_norm": 0.5650593042373657,
"learning_rate": 5.342465753424658e-05,
"loss": 1.7307,
"step": 40
},
{
"epoch": 0.028358983226698945,
"grad_norm": 0.7711919546127319,
"learning_rate": 5.479452054794521e-05,
"loss": 1.2067,
"step": 41
},
{
"epoch": 0.02905066574442331,
"grad_norm": 0.46603119373321533,
"learning_rate": 5.616438356164384e-05,
"loss": 1.0539,
"step": 42
},
{
"epoch": 0.029742348262147673,
"grad_norm": 0.2997550666332245,
"learning_rate": 5.753424657534247e-05,
"loss": 1.9997,
"step": 43
},
{
"epoch": 0.03043403077987204,
"grad_norm": 0.4058617651462555,
"learning_rate": 5.89041095890411e-05,
"loss": 2.3323,
"step": 44
},
{
"epoch": 0.0311257132975964,
"grad_norm": 0.3763635754585266,
"learning_rate": 6.0273972602739724e-05,
"loss": 1.8749,
"step": 45
},
{
"epoch": 0.03181739581532077,
"grad_norm": 0.3507993221282959,
"learning_rate": 6.164383561643835e-05,
"loss": 2.3828,
"step": 46
},
{
"epoch": 0.03250907833304513,
"grad_norm": 0.34240859746932983,
"learning_rate": 6.301369863013699e-05,
"loss": 2.3809,
"step": 47
},
{
"epoch": 0.0332007608507695,
"grad_norm": 0.4183844029903412,
"learning_rate": 6.438356164383562e-05,
"loss": 2.2065,
"step": 48
},
{
"epoch": 0.03389244336849386,
"grad_norm": 0.5209120512008667,
"learning_rate": 6.575342465753424e-05,
"loss": 2.061,
"step": 49
},
{
"epoch": 0.034584125886218224,
"grad_norm": 0.6568111181259155,
"learning_rate": 6.712328767123288e-05,
"loss": 2.1576,
"step": 50
},
{
"epoch": 0.03527580840394259,
"grad_norm": 0.5327390432357788,
"learning_rate": 6.84931506849315e-05,
"loss": 1.5562,
"step": 51
},
{
"epoch": 0.035967490921666956,
"grad_norm": 0.6389997005462646,
"learning_rate": 6.986301369863014e-05,
"loss": 2.3966,
"step": 52
},
{
"epoch": 0.03665917343939132,
"grad_norm": 0.6374388933181763,
"learning_rate": 7.123287671232876e-05,
"loss": 2.4013,
"step": 53
},
{
"epoch": 0.03735085595711568,
"grad_norm": 0.4383719563484192,
"learning_rate": 7.26027397260274e-05,
"loss": 2.061,
"step": 54
},
{
"epoch": 0.03804253847484005,
"grad_norm": 0.4306814670562744,
"learning_rate": 7.397260273972603e-05,
"loss": 2.4931,
"step": 55
},
{
"epoch": 0.03873422099256441,
"grad_norm": 0.5693032145500183,
"learning_rate": 7.534246575342466e-05,
"loss": 2.2842,
"step": 56
},
{
"epoch": 0.039425903510288775,
"grad_norm": 0.7792916297912598,
"learning_rate": 7.671232876712329e-05,
"loss": 1.8452,
"step": 57
},
{
"epoch": 0.040117586028013144,
"grad_norm": 0.4936494827270508,
"learning_rate": 7.808219178082192e-05,
"loss": 1.8668,
"step": 58
},
{
"epoch": 0.040809268545737507,
"grad_norm": 0.49626269936561584,
"learning_rate": 7.945205479452055e-05,
"loss": 1.9564,
"step": 59
},
{
"epoch": 0.04150095106346187,
"grad_norm": 0.4702766537666321,
"learning_rate": 8.082191780821919e-05,
"loss": 2.419,
"step": 60
},
{
"epoch": 0.04219263358118624,
"grad_norm": 0.6959900259971619,
"learning_rate": 8.219178082191781e-05,
"loss": 1.5835,
"step": 61
},
{
"epoch": 0.0428843160989106,
"grad_norm": 0.5963976979255676,
"learning_rate": 8.356164383561645e-05,
"loss": 2.2932,
"step": 62
},
{
"epoch": 0.04357599861663496,
"grad_norm": 0.4605785310268402,
"learning_rate": 8.493150684931507e-05,
"loss": 2.397,
"step": 63
},
{
"epoch": 0.04426768113435933,
"grad_norm": 0.4479796588420868,
"learning_rate": 8.630136986301371e-05,
"loss": 2.0257,
"step": 64
},
{
"epoch": 0.044959363652083695,
"grad_norm": 0.4342099726200104,
"learning_rate": 8.767123287671233e-05,
"loss": 1.9476,
"step": 65
},
{
"epoch": 0.04565104616980806,
"grad_norm": 0.44643279910087585,
"learning_rate": 8.904109589041096e-05,
"loss": 2.4669,
"step": 66
},
{
"epoch": 0.04634272868753242,
"grad_norm": 0.513254702091217,
"learning_rate": 9.041095890410958e-05,
"loss": 2.2581,
"step": 67
},
{
"epoch": 0.04703441120525679,
"grad_norm": 0.6621958613395691,
"learning_rate": 9.178082191780822e-05,
"loss": 1.3013,
"step": 68
},
{
"epoch": 0.04772609372298115,
"grad_norm": 0.43388229608535767,
"learning_rate": 9.315068493150684e-05,
"loss": 2.5325,
"step": 69
},
{
"epoch": 0.048417776240705514,
"grad_norm": 0.539757490158081,
"learning_rate": 9.452054794520548e-05,
"loss": 1.76,
"step": 70
},
{
"epoch": 0.04910945875842988,
"grad_norm": 0.33598029613494873,
"learning_rate": 9.58904109589041e-05,
"loss": 1.6931,
"step": 71
},
{
"epoch": 0.049801141276154245,
"grad_norm": 0.8661749958992004,
"learning_rate": 9.726027397260274e-05,
"loss": 2.1326,
"step": 72
},
{
"epoch": 0.05049282379387861,
"grad_norm": 0.5956202149391174,
"learning_rate": 9.863013698630137e-05,
"loss": 2.0554,
"step": 73
},
{
"epoch": 0.05118450631160298,
"grad_norm": 0.7723128199577332,
"learning_rate": 0.0001,
"loss": 1.4837,
"step": 74
},
{
"epoch": 0.05187618882932734,
"grad_norm": 0.5357879996299744,
"learning_rate": 9.992716678805537e-05,
"loss": 2.1688,
"step": 75
},
{
"epoch": 0.0525678713470517,
"grad_norm": 0.48623108863830566,
"learning_rate": 9.985433357611071e-05,
"loss": 2.1567,
"step": 76
},
{
"epoch": 0.053259553864776064,
"grad_norm": 0.4244164228439331,
"learning_rate": 9.978150036416607e-05,
"loss": 1.8235,
"step": 77
},
{
"epoch": 0.053951236382500434,
"grad_norm": 0.6056337356567383,
"learning_rate": 9.970866715222141e-05,
"loss": 2.0154,
"step": 78
},
{
"epoch": 0.054642918900224796,
"grad_norm": 0.5280382037162781,
"learning_rate": 9.963583394027677e-05,
"loss": 1.9959,
"step": 79
},
{
"epoch": 0.05533460141794916,
"grad_norm": 0.44745901226997375,
"learning_rate": 9.956300072833212e-05,
"loss": 1.2171,
"step": 80
},
{
"epoch": 0.05602628393567353,
"grad_norm": 0.7180289030075073,
"learning_rate": 9.949016751638748e-05,
"loss": 1.7036,
"step": 81
},
{
"epoch": 0.05671796645339789,
"grad_norm": 0.6465222239494324,
"learning_rate": 9.941733430444284e-05,
"loss": 1.8992,
"step": 82
},
{
"epoch": 0.05740964897112225,
"grad_norm": 0.4792153239250183,
"learning_rate": 9.934450109249819e-05,
"loss": 2.0712,
"step": 83
},
{
"epoch": 0.05810133148884662,
"grad_norm": 0.47839808464050293,
"learning_rate": 9.927166788055353e-05,
"loss": 1.7548,
"step": 84
},
{
"epoch": 0.058793014006570984,
"grad_norm": 0.7118557095527649,
"learning_rate": 9.919883466860888e-05,
"loss": 1.5835,
"step": 85
},
{
"epoch": 0.05948469652429535,
"grad_norm": 0.5179592370986938,
"learning_rate": 9.912600145666424e-05,
"loss": 1.621,
"step": 86
},
{
"epoch": 0.060176379042019716,
"grad_norm": 0.9235703349113464,
"learning_rate": 9.905316824471959e-05,
"loss": 0.8581,
"step": 87
},
{
"epoch": 0.06086806155974408,
"grad_norm": 0.8734163045883179,
"learning_rate": 9.898033503277495e-05,
"loss": 2.2878,
"step": 88
},
{
"epoch": 0.06155974407746844,
"grad_norm": 0.45049625635147095,
"learning_rate": 9.890750182083031e-05,
"loss": 2.2186,
"step": 89
},
{
"epoch": 0.0622514265951928,
"grad_norm": 0.6119429469108582,
"learning_rate": 9.883466860888566e-05,
"loss": 2.2461,
"step": 90
},
{
"epoch": 0.06294310911291717,
"grad_norm": 0.5095340013504028,
"learning_rate": 9.876183539694101e-05,
"loss": 2.2954,
"step": 91
},
{
"epoch": 0.06363479163064154,
"grad_norm": 0.6094918251037598,
"learning_rate": 9.868900218499635e-05,
"loss": 2.3478,
"step": 92
},
{
"epoch": 0.0643264741483659,
"grad_norm": 1.8420301675796509,
"learning_rate": 9.861616897305172e-05,
"loss": 2.213,
"step": 93
},
{
"epoch": 0.06501815666609026,
"grad_norm": 0.6151532530784607,
"learning_rate": 9.854333576110706e-05,
"loss": 1.8737,
"step": 94
},
{
"epoch": 0.06570983918381462,
"grad_norm": 0.5210903286933899,
"learning_rate": 9.847050254916242e-05,
"loss": 2.019,
"step": 95
},
{
"epoch": 0.066401521701539,
"grad_norm": 0.535746693611145,
"learning_rate": 9.839766933721779e-05,
"loss": 2.399,
"step": 96
},
{
"epoch": 0.06709320421926336,
"grad_norm": 0.8154505491256714,
"learning_rate": 9.832483612527313e-05,
"loss": 1.9442,
"step": 97
},
{
"epoch": 0.06778488673698772,
"grad_norm": 0.521247923374176,
"learning_rate": 9.825200291332848e-05,
"loss": 2.5264,
"step": 98
},
{
"epoch": 0.06847656925471209,
"grad_norm": 0.41544124484062195,
"learning_rate": 9.817916970138383e-05,
"loss": 1.8267,
"step": 99
},
{
"epoch": 0.06916825177243645,
"grad_norm": 0.538253128528595,
"learning_rate": 9.810633648943919e-05,
"loss": 1.6463,
"step": 100
},
{
"epoch": 0.06985993429016081,
"grad_norm": 0.8509485125541687,
"learning_rate": 9.803350327749454e-05,
"loss": 1.5877,
"step": 101
},
{
"epoch": 0.07055161680788519,
"grad_norm": 0.47714293003082275,
"learning_rate": 9.79606700655499e-05,
"loss": 1.5043,
"step": 102
},
{
"epoch": 0.07124329932560955,
"grad_norm": 0.5433914065361023,
"learning_rate": 9.788783685360526e-05,
"loss": 1.6454,
"step": 103
},
{
"epoch": 0.07193498184333391,
"grad_norm": 0.6400073766708374,
"learning_rate": 9.781500364166059e-05,
"loss": 1.6994,
"step": 104
},
{
"epoch": 0.07262666436105827,
"grad_norm": 0.6944612860679626,
"learning_rate": 9.774217042971595e-05,
"loss": 2.2063,
"step": 105
},
{
"epoch": 0.07331834687878264,
"grad_norm": 0.6701168417930603,
"learning_rate": 9.76693372177713e-05,
"loss": 1.2487,
"step": 106
},
{
"epoch": 0.074010029396507,
"grad_norm": 0.7888538837432861,
"learning_rate": 9.759650400582666e-05,
"loss": 1.8784,
"step": 107
},
{
"epoch": 0.07470171191423136,
"grad_norm": 0.7106865048408508,
"learning_rate": 9.752367079388201e-05,
"loss": 2.3673,
"step": 108
},
{
"epoch": 0.07539339443195574,
"grad_norm": 0.49588626623153687,
"learning_rate": 9.745083758193737e-05,
"loss": 1.8638,
"step": 109
},
{
"epoch": 0.0760850769496801,
"grad_norm": 0.48046427965164185,
"learning_rate": 9.737800436999273e-05,
"loss": 2.333,
"step": 110
},
{
"epoch": 0.07677675946740446,
"grad_norm": 0.6239060163497925,
"learning_rate": 9.730517115804807e-05,
"loss": 1.8378,
"step": 111
},
{
"epoch": 0.07746844198512882,
"grad_norm": 0.6083034873008728,
"learning_rate": 9.723233794610343e-05,
"loss": 1.9826,
"step": 112
},
{
"epoch": 0.07816012450285319,
"grad_norm": 0.8679235577583313,
"learning_rate": 9.715950473415877e-05,
"loss": 1.3375,
"step": 113
},
{
"epoch": 0.07885180702057755,
"grad_norm": 0.542757511138916,
"learning_rate": 9.708667152221414e-05,
"loss": 1.6349,
"step": 114
},
{
"epoch": 0.07954348953830193,
"grad_norm": 0.5968992114067078,
"learning_rate": 9.701383831026948e-05,
"loss": 1.3598,
"step": 115
},
{
"epoch": 0.08023517205602629,
"grad_norm": 0.5693714022636414,
"learning_rate": 9.694100509832484e-05,
"loss": 2.1527,
"step": 116
},
{
"epoch": 0.08092685457375065,
"grad_norm": 0.6307690143585205,
"learning_rate": 9.68681718863802e-05,
"loss": 1.7178,
"step": 117
},
{
"epoch": 0.08161853709147501,
"grad_norm": 0.5192516446113586,
"learning_rate": 9.679533867443554e-05,
"loss": 1.8115,
"step": 118
},
{
"epoch": 0.08231021960919938,
"grad_norm": 1.036036729812622,
"learning_rate": 9.67225054624909e-05,
"loss": 2.2423,
"step": 119
},
{
"epoch": 0.08300190212692374,
"grad_norm": 0.6450679302215576,
"learning_rate": 9.664967225054625e-05,
"loss": 1.9298,
"step": 120
},
{
"epoch": 0.0836935846446481,
"grad_norm": 0.8683953881263733,
"learning_rate": 9.657683903860161e-05,
"loss": 2.0791,
"step": 121
},
{
"epoch": 0.08438526716237248,
"grad_norm": 0.7013575434684753,
"learning_rate": 9.650400582665696e-05,
"loss": 2.0814,
"step": 122
},
{
"epoch": 0.08507694968009684,
"grad_norm": 0.6849325299263,
"learning_rate": 9.643117261471232e-05,
"loss": 1.83,
"step": 123
},
{
"epoch": 0.0857686321978212,
"grad_norm": 0.9820392727851868,
"learning_rate": 9.635833940276767e-05,
"loss": 1.5665,
"step": 124
},
{
"epoch": 0.08646031471554556,
"grad_norm": 0.5546866655349731,
"learning_rate": 9.628550619082301e-05,
"loss": 1.7215,
"step": 125
},
{
"epoch": 0.08715199723326993,
"grad_norm": 0.6748363971710205,
"learning_rate": 9.621267297887837e-05,
"loss": 1.6532,
"step": 126
},
{
"epoch": 0.08784367975099429,
"grad_norm": 0.4760434627532959,
"learning_rate": 9.613983976693372e-05,
"loss": 1.3476,
"step": 127
},
{
"epoch": 0.08853536226871866,
"grad_norm": 0.5710906386375427,
"learning_rate": 9.606700655498908e-05,
"loss": 2.025,
"step": 128
},
{
"epoch": 0.08922704478644303,
"grad_norm": 0.5950302481651306,
"learning_rate": 9.599417334304443e-05,
"loss": 2.6324,
"step": 129
},
{
"epoch": 0.08991872730416739,
"grad_norm": 0.4994860887527466,
"learning_rate": 9.592134013109979e-05,
"loss": 2.0315,
"step": 130
},
{
"epoch": 0.09061040982189175,
"grad_norm": 0.6400249004364014,
"learning_rate": 9.584850691915514e-05,
"loss": 2.3628,
"step": 131
},
{
"epoch": 0.09130209233961611,
"grad_norm": 1.0873314142227173,
"learning_rate": 9.577567370721049e-05,
"loss": 1.2664,
"step": 132
},
{
"epoch": 0.09199377485734048,
"grad_norm": 0.5430288314819336,
"learning_rate": 9.570284049526585e-05,
"loss": 1.7676,
"step": 133
},
{
"epoch": 0.09268545737506484,
"grad_norm": 0.5969283580780029,
"learning_rate": 9.56300072833212e-05,
"loss": 2.0824,
"step": 134
},
{
"epoch": 0.09337713989278922,
"grad_norm": 0.6228020191192627,
"learning_rate": 9.555717407137656e-05,
"loss": 2.0687,
"step": 135
},
{
"epoch": 0.09406882241051358,
"grad_norm": 0.6905536651611328,
"learning_rate": 9.54843408594319e-05,
"loss": 2.2948,
"step": 136
},
{
"epoch": 0.09476050492823794,
"grad_norm": 0.6397948861122131,
"learning_rate": 9.541150764748726e-05,
"loss": 1.4147,
"step": 137
},
{
"epoch": 0.0954521874459623,
"grad_norm": 0.5489908456802368,
"learning_rate": 9.533867443554261e-05,
"loss": 2.0582,
"step": 138
},
{
"epoch": 0.09614386996368667,
"grad_norm": 0.5713220834732056,
"learning_rate": 9.526584122359796e-05,
"loss": 2.1708,
"step": 139
},
{
"epoch": 0.09683555248141103,
"grad_norm": 0.6736558079719543,
"learning_rate": 9.519300801165332e-05,
"loss": 1.6808,
"step": 140
},
{
"epoch": 0.0975272349991354,
"grad_norm": 1.1222553253173828,
"learning_rate": 9.512017479970867e-05,
"loss": 1.563,
"step": 141
},
{
"epoch": 0.09821891751685977,
"grad_norm": 0.6231582760810852,
"learning_rate": 9.504734158776403e-05,
"loss": 1.6458,
"step": 142
},
{
"epoch": 0.09891060003458413,
"grad_norm": 0.716089129447937,
"learning_rate": 9.497450837581938e-05,
"loss": 1.7055,
"step": 143
},
{
"epoch": 0.09960228255230849,
"grad_norm": 0.49629953503608704,
"learning_rate": 9.490167516387472e-05,
"loss": 1.2985,
"step": 144
},
{
"epoch": 0.10029396507003285,
"grad_norm": 0.8354843258857727,
"learning_rate": 9.482884195193008e-05,
"loss": 1.8557,
"step": 145
},
{
"epoch": 0.10098564758775722,
"grad_norm": 0.7507709860801697,
"learning_rate": 9.475600873998543e-05,
"loss": 1.1816,
"step": 146
},
{
"epoch": 0.10167733010548158,
"grad_norm": 0.8973211050033569,
"learning_rate": 9.46831755280408e-05,
"loss": 1.4495,
"step": 147
},
{
"epoch": 0.10236901262320595,
"grad_norm": 0.7720927000045776,
"learning_rate": 9.461034231609614e-05,
"loss": 2.1743,
"step": 148
},
{
"epoch": 0.10306069514093032,
"grad_norm": 0.511075496673584,
"learning_rate": 9.45375091041515e-05,
"loss": 1.8267,
"step": 149
},
{
"epoch": 0.10375237765865468,
"grad_norm": 0.6523205041885376,
"learning_rate": 9.446467589220685e-05,
"loss": 1.9682,
"step": 150
},
{
"epoch": 0.10444406017637904,
"grad_norm": 0.703565239906311,
"learning_rate": 9.43918426802622e-05,
"loss": 2.3272,
"step": 151
},
{
"epoch": 0.1051357426941034,
"grad_norm": 0.5773376226425171,
"learning_rate": 9.431900946831756e-05,
"loss": 1.9753,
"step": 152
},
{
"epoch": 0.10582742521182777,
"grad_norm": 0.6234054565429688,
"learning_rate": 9.42461762563729e-05,
"loss": 2.5085,
"step": 153
},
{
"epoch": 0.10651910772955213,
"grad_norm": 0.7076170444488525,
"learning_rate": 9.417334304442827e-05,
"loss": 1.9139,
"step": 154
},
{
"epoch": 0.1072107902472765,
"grad_norm": 0.5035070776939392,
"learning_rate": 9.410050983248361e-05,
"loss": 2.0164,
"step": 155
},
{
"epoch": 0.10790247276500087,
"grad_norm": 1.046075701713562,
"learning_rate": 9.402767662053898e-05,
"loss": 1.2478,
"step": 156
},
{
"epoch": 0.10859415528272523,
"grad_norm": 0.5620921850204468,
"learning_rate": 9.395484340859432e-05,
"loss": 1.565,
"step": 157
},
{
"epoch": 0.10928583780044959,
"grad_norm": 0.8284108638763428,
"learning_rate": 9.388201019664967e-05,
"loss": 2.0823,
"step": 158
},
{
"epoch": 0.10997752031817395,
"grad_norm": 0.584373950958252,
"learning_rate": 9.380917698470503e-05,
"loss": 1.6352,
"step": 159
},
{
"epoch": 0.11066920283589832,
"grad_norm": 0.7779399752616882,
"learning_rate": 9.373634377276038e-05,
"loss": 1.83,
"step": 160
},
{
"epoch": 0.1113608853536227,
"grad_norm": 0.6635437607765198,
"learning_rate": 9.366351056081574e-05,
"loss": 1.8935,
"step": 161
},
{
"epoch": 0.11205256787134706,
"grad_norm": 0.6441919207572937,
"learning_rate": 9.359067734887109e-05,
"loss": 1.9566,
"step": 162
},
{
"epoch": 0.11274425038907142,
"grad_norm": 1.20408296585083,
"learning_rate": 9.351784413692645e-05,
"loss": 1.2529,
"step": 163
},
{
"epoch": 0.11343593290679578,
"grad_norm": 0.6320855021476746,
"learning_rate": 9.34450109249818e-05,
"loss": 1.8071,
"step": 164
},
{
"epoch": 0.11412761542452014,
"grad_norm": 0.5197457075119019,
"learning_rate": 9.337217771303714e-05,
"loss": 1.7303,
"step": 165
},
{
"epoch": 0.1148192979422445,
"grad_norm": 0.6230719089508057,
"learning_rate": 9.32993445010925e-05,
"loss": 2.0685,
"step": 166
},
{
"epoch": 0.11551098045996887,
"grad_norm": 0.6185227036476135,
"learning_rate": 9.322651128914785e-05,
"loss": 1.9485,
"step": 167
},
{
"epoch": 0.11620266297769324,
"grad_norm": 0.7290086150169373,
"learning_rate": 9.315367807720321e-05,
"loss": 1.6923,
"step": 168
},
{
"epoch": 0.1168943454954176,
"grad_norm": 0.9122477769851685,
"learning_rate": 9.308084486525856e-05,
"loss": 1.3395,
"step": 169
},
{
"epoch": 0.11758602801314197,
"grad_norm": 0.5755578875541687,
"learning_rate": 9.300801165331392e-05,
"loss": 1.0775,
"step": 170
},
{
"epoch": 0.11827771053086633,
"grad_norm": 0.6516758799552917,
"learning_rate": 9.293517844136927e-05,
"loss": 1.6174,
"step": 171
},
{
"epoch": 0.1189693930485907,
"grad_norm": 0.5685858726501465,
"learning_rate": 9.286234522942462e-05,
"loss": 1.851,
"step": 172
},
{
"epoch": 0.11966107556631506,
"grad_norm": 0.5784198641777039,
"learning_rate": 9.278951201747998e-05,
"loss": 1.712,
"step": 173
},
{
"epoch": 0.12035275808403943,
"grad_norm": 0.7357056736946106,
"learning_rate": 9.271667880553533e-05,
"loss": 2.2286,
"step": 174
},
{
"epoch": 0.1210444406017638,
"grad_norm": 0.5122838616371155,
"learning_rate": 9.264384559359069e-05,
"loss": 1.5581,
"step": 175
},
{
"epoch": 0.12173612311948816,
"grad_norm": 0.6864873170852661,
"learning_rate": 9.257101238164603e-05,
"loss": 1.2737,
"step": 176
},
{
"epoch": 0.12242780563721252,
"grad_norm": 0.441771000623703,
"learning_rate": 9.24981791697014e-05,
"loss": 1.2291,
"step": 177
},
{
"epoch": 0.12311948815493688,
"grad_norm": 0.634203314781189,
"learning_rate": 9.242534595775674e-05,
"loss": 2.1797,
"step": 178
},
{
"epoch": 0.12381117067266124,
"grad_norm": 0.6232409477233887,
"learning_rate": 9.235251274581209e-05,
"loss": 1.5954,
"step": 179
},
{
"epoch": 0.1245028531903856,
"grad_norm": 0.7475627660751343,
"learning_rate": 9.227967953386745e-05,
"loss": 2.3014,
"step": 180
},
{
"epoch": 0.12519453570810998,
"grad_norm": 0.7436168789863586,
"learning_rate": 9.22068463219228e-05,
"loss": 2.1053,
"step": 181
},
{
"epoch": 0.12588621822583435,
"grad_norm": 0.754830002784729,
"learning_rate": 9.213401310997816e-05,
"loss": 2.2968,
"step": 182
},
{
"epoch": 0.1265779007435587,
"grad_norm": 0.7968791723251343,
"learning_rate": 9.206117989803351e-05,
"loss": 1.9433,
"step": 183
},
{
"epoch": 0.12726958326128307,
"grad_norm": 0.7031124234199524,
"learning_rate": 9.198834668608885e-05,
"loss": 1.6162,
"step": 184
},
{
"epoch": 0.12796126577900743,
"grad_norm": 1.5919851064682007,
"learning_rate": 9.191551347414422e-05,
"loss": 2.3995,
"step": 185
},
{
"epoch": 0.1286529482967318,
"grad_norm": 0.8990558385848999,
"learning_rate": 9.184268026219956e-05,
"loss": 2.1597,
"step": 186
},
{
"epoch": 0.12934463081445616,
"grad_norm": 0.6514426469802856,
"learning_rate": 9.176984705025492e-05,
"loss": 2.2172,
"step": 187
},
{
"epoch": 0.13003631333218052,
"grad_norm": 0.5041804313659668,
"learning_rate": 9.169701383831027e-05,
"loss": 1.6619,
"step": 188
},
{
"epoch": 0.13072799584990488,
"grad_norm": 0.5292882323265076,
"learning_rate": 9.162418062636563e-05,
"loss": 1.3204,
"step": 189
},
{
"epoch": 0.13141967836762924,
"grad_norm": 0.9279568791389465,
"learning_rate": 9.155134741442098e-05,
"loss": 1.2023,
"step": 190
},
{
"epoch": 0.13211136088535363,
"grad_norm": 0.8143125772476196,
"learning_rate": 9.147851420247633e-05,
"loss": 1.5768,
"step": 191
},
{
"epoch": 0.132803043403078,
"grad_norm": 0.9918676614761353,
"learning_rate": 9.140568099053169e-05,
"loss": 1.3813,
"step": 192
},
{
"epoch": 0.13349472592080236,
"grad_norm": 0.6452721357345581,
"learning_rate": 9.133284777858704e-05,
"loss": 2.4017,
"step": 193
},
{
"epoch": 0.13418640843852672,
"grad_norm": 0.7932286858558655,
"learning_rate": 9.12600145666424e-05,
"loss": 1.1203,
"step": 194
},
{
"epoch": 0.13487809095625108,
"grad_norm": 0.5378614068031311,
"learning_rate": 9.118718135469774e-05,
"loss": 1.9103,
"step": 195
},
{
"epoch": 0.13556977347397545,
"grad_norm": 0.6193660497665405,
"learning_rate": 9.11143481427531e-05,
"loss": 2.0729,
"step": 196
},
{
"epoch": 0.1362614559916998,
"grad_norm": 0.8437771797180176,
"learning_rate": 9.104151493080845e-05,
"loss": 2.4214,
"step": 197
},
{
"epoch": 0.13695313850942417,
"grad_norm": 0.8080397248268127,
"learning_rate": 9.09686817188638e-05,
"loss": 1.2323,
"step": 198
},
{
"epoch": 0.13764482102714853,
"grad_norm": 0.5907486081123352,
"learning_rate": 9.089584850691916e-05,
"loss": 1.6866,
"step": 199
},
{
"epoch": 0.1383365035448729,
"grad_norm": 0.6559591293334961,
"learning_rate": 9.082301529497451e-05,
"loss": 2.1268,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 1446,
"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": 6421740837765120.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}