9b-16 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
47bf0ee verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2055,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00291970802919708,
"grad_norm": 2.453125,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.3209164142608643,
"step": 2
},
{
"epoch": 0.00583941605839416,
"grad_norm": 12.125,
"learning_rate": 9.67741935483871e-07,
"loss": 2.3325037956237793,
"step": 4
},
{
"epoch": 0.008759124087591242,
"grad_norm": 2.796875,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.9444684982299805,
"step": 6
},
{
"epoch": 0.01167883211678832,
"grad_norm": 1.2109375,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.6828501224517822,
"step": 8
},
{
"epoch": 0.014598540145985401,
"grad_norm": 1.671875,
"learning_rate": 2.903225806451613e-06,
"loss": 1.8027360439300537,
"step": 10
},
{
"epoch": 0.017518248175182483,
"grad_norm": 2.5,
"learning_rate": 3.548387096774194e-06,
"loss": 1.9229137897491455,
"step": 12
},
{
"epoch": 0.020437956204379562,
"grad_norm": 1.265625,
"learning_rate": 4.193548387096774e-06,
"loss": 1.5638442039489746,
"step": 14
},
{
"epoch": 0.02335766423357664,
"grad_norm": 4.90625,
"learning_rate": 4.838709677419355e-06,
"loss": 1.7475087642669678,
"step": 16
},
{
"epoch": 0.026277372262773723,
"grad_norm": 2.296875,
"learning_rate": 5.483870967741935e-06,
"loss": 1.8932218551635742,
"step": 18
},
{
"epoch": 0.029197080291970802,
"grad_norm": 2.40625,
"learning_rate": 6.129032258064517e-06,
"loss": 1.8747022151947021,
"step": 20
},
{
"epoch": 0.032116788321167884,
"grad_norm": 3.140625,
"learning_rate": 6.774193548387097e-06,
"loss": 1.9435057640075684,
"step": 22
},
{
"epoch": 0.035036496350364967,
"grad_norm": 2.171875,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.8176038265228271,
"step": 24
},
{
"epoch": 0.03795620437956204,
"grad_norm": 6.4375,
"learning_rate": 8.064516129032258e-06,
"loss": 1.910407304763794,
"step": 26
},
{
"epoch": 0.040875912408759124,
"grad_norm": 5.71875,
"learning_rate": 8.70967741935484e-06,
"loss": 2.2947583198547363,
"step": 28
},
{
"epoch": 0.043795620437956206,
"grad_norm": 2.421875,
"learning_rate": 9.35483870967742e-06,
"loss": 1.426323413848877,
"step": 30
},
{
"epoch": 0.04671532846715328,
"grad_norm": 2.9375,
"learning_rate": 1e-05,
"loss": 1.81662917137146,
"step": 32
},
{
"epoch": 0.049635036496350364,
"grad_norm": 2.984375,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.9644252061843872,
"step": 34
},
{
"epoch": 0.052554744525547446,
"grad_norm": 15.125,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.9643396139144897,
"step": 36
},
{
"epoch": 0.05547445255474453,
"grad_norm": 2.25,
"learning_rate": 1.1935483870967743e-05,
"loss": 2.181769371032715,
"step": 38
},
{
"epoch": 0.058394160583941604,
"grad_norm": 1.46875,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.5383245944976807,
"step": 40
},
{
"epoch": 0.061313868613138686,
"grad_norm": 5.84375,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.7693108320236206,
"step": 42
},
{
"epoch": 0.06423357664233577,
"grad_norm": 1.640625,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.5715991258621216,
"step": 44
},
{
"epoch": 0.06715328467153285,
"grad_norm": 5.59375,
"learning_rate": 1.4516129032258066e-05,
"loss": 2.1106648445129395,
"step": 46
},
{
"epoch": 0.07007299270072993,
"grad_norm": 4.34375,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.659260869026184,
"step": 48
},
{
"epoch": 0.072992700729927,
"grad_norm": 2.5625,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.5844422578811646,
"step": 50
},
{
"epoch": 0.07591240875912408,
"grad_norm": 0.73046875,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.2334352731704712,
"step": 52
},
{
"epoch": 0.07883211678832117,
"grad_norm": 1.6015625,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.6998990774154663,
"step": 54
},
{
"epoch": 0.08175182481751825,
"grad_norm": 2.984375,
"learning_rate": 1.774193548387097e-05,
"loss": 1.7014799118041992,
"step": 56
},
{
"epoch": 0.08467153284671533,
"grad_norm": 1.34375,
"learning_rate": 1.838709677419355e-05,
"loss": 1.5540595054626465,
"step": 58
},
{
"epoch": 0.08759124087591241,
"grad_norm": 2.046875,
"learning_rate": 1.903225806451613e-05,
"loss": 1.6097842454910278,
"step": 60
},
{
"epoch": 0.0905109489051095,
"grad_norm": 2.90625,
"learning_rate": 1.967741935483871e-05,
"loss": 1.6974210739135742,
"step": 62
},
{
"epoch": 0.09343065693430656,
"grad_norm": 1.5234375,
"learning_rate": 1.9999988818564274e-05,
"loss": 1.5296247005462646,
"step": 64
},
{
"epoch": 0.09635036496350365,
"grad_norm": 0.82421875,
"learning_rate": 1.9999899367245136e-05,
"loss": 1.3334264755249023,
"step": 66
},
{
"epoch": 0.09927007299270073,
"grad_norm": 1.6015625,
"learning_rate": 1.9999720465495917e-05,
"loss": 1.5143179893493652,
"step": 68
},
{
"epoch": 0.10218978102189781,
"grad_norm": 1.3984375,
"learning_rate": 1.999945211509474e-05,
"loss": 1.4130449295043945,
"step": 70
},
{
"epoch": 0.10510948905109489,
"grad_norm": 6.03125,
"learning_rate": 1.9999094318708743e-05,
"loss": 0.9752007126808167,
"step": 72
},
{
"epoch": 0.10802919708029197,
"grad_norm": 3.171875,
"learning_rate": 1.9998647079894086e-05,
"loss": 1.258436679840088,
"step": 74
},
{
"epoch": 0.11094890510948906,
"grad_norm": 0.65234375,
"learning_rate": 1.99981104030959e-05,
"loss": 1.276557445526123,
"step": 76
},
{
"epoch": 0.11386861313868613,
"grad_norm": 2.34375,
"learning_rate": 1.999748429364823e-05,
"loss": 1.4276189804077148,
"step": 78
},
{
"epoch": 0.11678832116788321,
"grad_norm": 0.875,
"learning_rate": 1.999676875777402e-05,
"loss": 0.998885452747345,
"step": 80
},
{
"epoch": 0.11970802919708029,
"grad_norm": 1.125,
"learning_rate": 1.9995963802585007e-05,
"loss": 1.5045676231384277,
"step": 82
},
{
"epoch": 0.12262773722627737,
"grad_norm": 2.5625,
"learning_rate": 1.9995069436081685e-05,
"loss": 1.3018450736999512,
"step": 84
},
{
"epoch": 0.12554744525547445,
"grad_norm": 2.0625,
"learning_rate": 1.9994085667153203e-05,
"loss": 1.4471237659454346,
"step": 86
},
{
"epoch": 0.12846715328467154,
"grad_norm": 1.8203125,
"learning_rate": 1.9993012505577282e-05,
"loss": 1.3635817766189575,
"step": 88
},
{
"epoch": 0.13138686131386862,
"grad_norm": 1.0234375,
"learning_rate": 1.9991849962020138e-05,
"loss": 1.1354179382324219,
"step": 90
},
{
"epoch": 0.1343065693430657,
"grad_norm": 1.3125,
"learning_rate": 1.999059804803634e-05,
"loss": 1.4238762855529785,
"step": 92
},
{
"epoch": 0.13722627737226278,
"grad_norm": 1.1796875,
"learning_rate": 1.998925677606871e-05,
"loss": 1.2135484218597412,
"step": 94
},
{
"epoch": 0.14014598540145987,
"grad_norm": 4.75,
"learning_rate": 1.998782615944824e-05,
"loss": 1.5475749969482422,
"step": 96
},
{
"epoch": 0.14306569343065692,
"grad_norm": 1.7421875,
"learning_rate": 1.9986306212393876e-05,
"loss": 1.3037400245666504,
"step": 98
},
{
"epoch": 0.145985401459854,
"grad_norm": 4.1875,
"learning_rate": 1.998469695001245e-05,
"loss": 1.2510803937911987,
"step": 100
},
{
"epoch": 0.14890510948905109,
"grad_norm": 1.2734375,
"learning_rate": 1.99829983882985e-05,
"loss": 1.3476972579956055,
"step": 102
},
{
"epoch": 0.15182481751824817,
"grad_norm": 1.3671875,
"learning_rate": 1.9981210544134113e-05,
"loss": 1.304163932800293,
"step": 104
},
{
"epoch": 0.15474452554744525,
"grad_norm": 5.375,
"learning_rate": 1.997933343528875e-05,
"loss": 1.4027568101882935,
"step": 106
},
{
"epoch": 0.15766423357664233,
"grad_norm": 5.0,
"learning_rate": 1.9977367080419096e-05,
"loss": 1.4859360456466675,
"step": 108
},
{
"epoch": 0.16058394160583941,
"grad_norm": 2.8125,
"learning_rate": 1.9975311499068836e-05,
"loss": 1.1744787693023682,
"step": 110
},
{
"epoch": 0.1635036496350365,
"grad_norm": 1.5234375,
"learning_rate": 1.99731667116685e-05,
"loss": 1.3460320234298706,
"step": 112
},
{
"epoch": 0.16642335766423358,
"grad_norm": 1.453125,
"learning_rate": 1.997093273953523e-05,
"loss": 1.2084546089172363,
"step": 114
},
{
"epoch": 0.16934306569343066,
"grad_norm": 1.359375,
"learning_rate": 1.996860960487258e-05,
"loss": 1.2922112941741943,
"step": 116
},
{
"epoch": 0.17226277372262774,
"grad_norm": 3.6875,
"learning_rate": 1.9966197330770293e-05,
"loss": 1.2132439613342285,
"step": 118
},
{
"epoch": 0.17518248175182483,
"grad_norm": 2.046875,
"learning_rate": 1.9963695941204085e-05,
"loss": 1.5152297019958496,
"step": 120
},
{
"epoch": 0.1781021897810219,
"grad_norm": 3.46875,
"learning_rate": 1.996110546103538e-05,
"loss": 1.4819408655166626,
"step": 122
},
{
"epoch": 0.181021897810219,
"grad_norm": 2.703125,
"learning_rate": 1.9958425916011086e-05,
"loss": 1.4910974502563477,
"step": 124
},
{
"epoch": 0.18394160583941604,
"grad_norm": 7.8125,
"learning_rate": 1.9955657332763333e-05,
"loss": 1.3679836988449097,
"step": 126
},
{
"epoch": 0.18686131386861313,
"grad_norm": 1.6171875,
"learning_rate": 1.9952799738809195e-05,
"loss": 1.5538718700408936,
"step": 128
},
{
"epoch": 0.1897810218978102,
"grad_norm": 1.59375,
"learning_rate": 1.994985316255044e-05,
"loss": 1.480999231338501,
"step": 130
},
{
"epoch": 0.1927007299270073,
"grad_norm": 6.90625,
"learning_rate": 1.9946817633273226e-05,
"loss": 1.2054316997528076,
"step": 132
},
{
"epoch": 0.19562043795620437,
"grad_norm": 3.125,
"learning_rate": 1.9943693181147833e-05,
"loss": 1.4222183227539062,
"step": 134
},
{
"epoch": 0.19854014598540146,
"grad_norm": 1.4921875,
"learning_rate": 1.9940479837228327e-05,
"loss": 1.3593757152557373,
"step": 136
},
{
"epoch": 0.20145985401459854,
"grad_norm": 0.66015625,
"learning_rate": 1.9937177633452302e-05,
"loss": 1.2240228652954102,
"step": 138
},
{
"epoch": 0.20437956204379562,
"grad_norm": 2.46875,
"learning_rate": 1.99337866026405e-05,
"loss": 1.3149917125701904,
"step": 140
},
{
"epoch": 0.2072992700729927,
"grad_norm": 1.5234375,
"learning_rate": 1.9930306778496558e-05,
"loss": 1.5356433391571045,
"step": 142
},
{
"epoch": 0.21021897810218979,
"grad_norm": 1.78125,
"learning_rate": 1.9926738195606603e-05,
"loss": 1.298794150352478,
"step": 144
},
{
"epoch": 0.21313868613138687,
"grad_norm": 10.375,
"learning_rate": 1.992308088943895e-05,
"loss": 1.5843662023544312,
"step": 146
},
{
"epoch": 0.21605839416058395,
"grad_norm": 1.3125,
"learning_rate": 1.9919334896343747e-05,
"loss": 1.2190661430358887,
"step": 148
},
{
"epoch": 0.21897810218978103,
"grad_norm": 1.5390625,
"learning_rate": 1.9915500253552597e-05,
"loss": 1.337175965309143,
"step": 150
},
{
"epoch": 0.22189781021897811,
"grad_norm": 1.90625,
"learning_rate": 1.991157699917819e-05,
"loss": 0.8397458791732788,
"step": 152
},
{
"epoch": 0.22481751824817517,
"grad_norm": 1.03125,
"learning_rate": 1.9907565172213948e-05,
"loss": 0.962209939956665,
"step": 154
},
{
"epoch": 0.22773722627737225,
"grad_norm": 3.578125,
"learning_rate": 1.990346481253361e-05,
"loss": 0.3393441140651703,
"step": 156
},
{
"epoch": 0.23065693430656933,
"grad_norm": 1.6875,
"learning_rate": 1.989927596089084e-05,
"loss": 1.3953263759613037,
"step": 158
},
{
"epoch": 0.23357664233576642,
"grad_norm": 4.0625,
"learning_rate": 1.9894998658918842e-05,
"loss": 1.164942979812622,
"step": 160
},
{
"epoch": 0.2364963503649635,
"grad_norm": 1.703125,
"learning_rate": 1.989063294912992e-05,
"loss": 1.3989766836166382,
"step": 162
},
{
"epoch": 0.23941605839416058,
"grad_norm": 1.3828125,
"learning_rate": 1.988617887491508e-05,
"loss": 1.115430235862732,
"step": 164
},
{
"epoch": 0.24233576642335766,
"grad_norm": 5.15625,
"learning_rate": 1.9881636480543568e-05,
"loss": 1.3270823955535889,
"step": 166
},
{
"epoch": 0.24525547445255474,
"grad_norm": 0.69921875,
"learning_rate": 1.9877005811162464e-05,
"loss": 1.1645622253417969,
"step": 168
},
{
"epoch": 0.24817518248175183,
"grad_norm": 11.0,
"learning_rate": 1.9872286912796214e-05,
"loss": 1.2632205486297607,
"step": 170
},
{
"epoch": 0.2510948905109489,
"grad_norm": 1.5546875,
"learning_rate": 1.986747983234617e-05,
"loss": 1.208564043045044,
"step": 172
},
{
"epoch": 0.25401459854014596,
"grad_norm": 4.28125,
"learning_rate": 1.9862584617590138e-05,
"loss": 1.3891022205352783,
"step": 174
},
{
"epoch": 0.2569343065693431,
"grad_norm": 1.2265625,
"learning_rate": 1.9857601317181893e-05,
"loss": 1.0644850730895996,
"step": 176
},
{
"epoch": 0.25985401459854013,
"grad_norm": 3.765625,
"learning_rate": 1.985252998065069e-05,
"loss": 1.0803289413452148,
"step": 178
},
{
"epoch": 0.26277372262773724,
"grad_norm": 2.84375,
"learning_rate": 1.984737065840079e-05,
"loss": 1.4185163974761963,
"step": 180
},
{
"epoch": 0.2656934306569343,
"grad_norm": 3.203125,
"learning_rate": 1.9842123401710946e-05,
"loss": 1.2833259105682373,
"step": 182
},
{
"epoch": 0.2686131386861314,
"grad_norm": 0.4921875,
"learning_rate": 1.9836788262733882e-05,
"loss": 1.0898754596710205,
"step": 184
},
{
"epoch": 0.27153284671532846,
"grad_norm": 2.046875,
"learning_rate": 1.983136529449581e-05,
"loss": 1.2598240375518799,
"step": 186
},
{
"epoch": 0.27445255474452557,
"grad_norm": 1.859375,
"learning_rate": 1.982585455089587e-05,
"loss": 1.1179606914520264,
"step": 188
},
{
"epoch": 0.2773722627737226,
"grad_norm": 1.3046875,
"learning_rate": 1.9820256086705603e-05,
"loss": 1.2375895977020264,
"step": 190
},
{
"epoch": 0.28029197080291973,
"grad_norm": 2.734375,
"learning_rate": 1.981456995756841e-05,
"loss": 1.3714609146118164,
"step": 192
},
{
"epoch": 0.2832116788321168,
"grad_norm": 1.7890625,
"learning_rate": 1.9808796219999005e-05,
"loss": 1.5060479640960693,
"step": 194
},
{
"epoch": 0.28613138686131384,
"grad_norm": 1.5703125,
"learning_rate": 1.9802934931382836e-05,
"loss": 1.305965781211853,
"step": 196
},
{
"epoch": 0.28905109489051095,
"grad_norm": 1.421875,
"learning_rate": 1.9796986149975537e-05,
"loss": 1.3187477588653564,
"step": 198
},
{
"epoch": 0.291970802919708,
"grad_norm": 1.3046875,
"learning_rate": 1.9790949934902327e-05,
"loss": 1.518209457397461,
"step": 200
},
{
"epoch": 0.2948905109489051,
"grad_norm": 1.546875,
"learning_rate": 1.978482634615744e-05,
"loss": 1.3524653911590576,
"step": 202
},
{
"epoch": 0.29781021897810217,
"grad_norm": 1.4609375,
"learning_rate": 1.9778615444603516e-05,
"loss": 1.5233702659606934,
"step": 204
},
{
"epoch": 0.3007299270072993,
"grad_norm": 1.578125,
"learning_rate": 1.9772317291970993e-05,
"loss": 1.2738330364227295,
"step": 206
},
{
"epoch": 0.30364963503649633,
"grad_norm": 1.8203125,
"learning_rate": 1.9765931950857528e-05,
"loss": 1.5982074737548828,
"step": 208
},
{
"epoch": 0.30656934306569344,
"grad_norm": 4.40625,
"learning_rate": 1.975945948472732e-05,
"loss": 1.4199731349945068,
"step": 210
},
{
"epoch": 0.3094890510948905,
"grad_norm": 0.66796875,
"learning_rate": 1.975289995791052e-05,
"loss": 1.0434203147888184,
"step": 212
},
{
"epoch": 0.3124087591240876,
"grad_norm": 4.21875,
"learning_rate": 1.9746253435602586e-05,
"loss": 0.9944186210632324,
"step": 214
},
{
"epoch": 0.31532846715328466,
"grad_norm": 1.671875,
"learning_rate": 1.9739519983863613e-05,
"loss": 1.0899577140808105,
"step": 216
},
{
"epoch": 0.3182481751824818,
"grad_norm": 8.4375,
"learning_rate": 1.97326996696177e-05,
"loss": 0.8038634657859802,
"step": 218
},
{
"epoch": 0.32116788321167883,
"grad_norm": 4.15625,
"learning_rate": 1.972579256065228e-05,
"loss": 1.4780833721160889,
"step": 220
},
{
"epoch": 0.32408759124087594,
"grad_norm": 2.875,
"learning_rate": 1.971879872561744e-05,
"loss": 1.3284614086151123,
"step": 222
},
{
"epoch": 0.327007299270073,
"grad_norm": 1.0234375,
"learning_rate": 1.971171823402524e-05,
"loss": 1.107306957244873,
"step": 224
},
{
"epoch": 0.32992700729927005,
"grad_norm": 2.890625,
"learning_rate": 1.9704551156249024e-05,
"loss": 1.305222749710083,
"step": 226
},
{
"epoch": 0.33284671532846716,
"grad_norm": 0.8671875,
"learning_rate": 1.9697297563522716e-05,
"loss": 1.1090219020843506,
"step": 228
},
{
"epoch": 0.3357664233576642,
"grad_norm": 2.640625,
"learning_rate": 1.968995752794013e-05,
"loss": 1.1776087284088135,
"step": 230
},
{
"epoch": 0.3386861313868613,
"grad_norm": 3.375,
"learning_rate": 1.968253112245422e-05,
"loss": 0.8213838338851929,
"step": 232
},
{
"epoch": 0.3416058394160584,
"grad_norm": 4.75,
"learning_rate": 1.9675018420876394e-05,
"loss": 1.3054924011230469,
"step": 234
},
{
"epoch": 0.3445255474452555,
"grad_norm": 5.8125,
"learning_rate": 1.9667419497875747e-05,
"loss": 1.3970417976379395,
"step": 236
},
{
"epoch": 0.34744525547445254,
"grad_norm": 1.4765625,
"learning_rate": 1.9659734428978342e-05,
"loss": 1.2723561525344849,
"step": 238
},
{
"epoch": 0.35036496350364965,
"grad_norm": 4.90625,
"learning_rate": 1.965196329056644e-05,
"loss": 1.798250675201416,
"step": 240
},
{
"epoch": 0.3532846715328467,
"grad_norm": 1.265625,
"learning_rate": 1.9644106159877764e-05,
"loss": 1.2417988777160645,
"step": 242
},
{
"epoch": 0.3562043795620438,
"grad_norm": 1.015625,
"learning_rate": 1.9636163115004708e-05,
"loss": 1.3138728141784668,
"step": 244
},
{
"epoch": 0.35912408759124087,
"grad_norm": 0.64453125,
"learning_rate": 1.9628134234893575e-05,
"loss": 1.0861350297927856,
"step": 246
},
{
"epoch": 0.362043795620438,
"grad_norm": 0.98828125,
"learning_rate": 1.962001959934379e-05,
"loss": 1.4500981569290161,
"step": 248
},
{
"epoch": 0.36496350364963503,
"grad_norm": 0.7890625,
"learning_rate": 1.9611819289007102e-05,
"loss": 1.0715610980987549,
"step": 250
},
{
"epoch": 0.3678832116788321,
"grad_norm": 2.921875,
"learning_rate": 1.9603533385386787e-05,
"loss": 1.319586992263794,
"step": 252
},
{
"epoch": 0.3708029197080292,
"grad_norm": 4.25,
"learning_rate": 1.959516197083684e-05,
"loss": 1.5193085670471191,
"step": 254
},
{
"epoch": 0.37372262773722625,
"grad_norm": 1.1640625,
"learning_rate": 1.9586705128561147e-05,
"loss": 1.2118489742279053,
"step": 256
},
{
"epoch": 0.37664233576642336,
"grad_norm": 2.765625,
"learning_rate": 1.9578162942612668e-05,
"loss": 1.3003623485565186,
"step": 258
},
{
"epoch": 0.3795620437956204,
"grad_norm": 2.84375,
"learning_rate": 1.9569535497892602e-05,
"loss": 1.3232567310333252,
"step": 260
},
{
"epoch": 0.38248175182481753,
"grad_norm": 1.6953125,
"learning_rate": 1.9560822880149528e-05,
"loss": 1.2432191371917725,
"step": 262
},
{
"epoch": 0.3854014598540146,
"grad_norm": 0.8984375,
"learning_rate": 1.9552025175978576e-05,
"loss": 1.197983741760254,
"step": 264
},
{
"epoch": 0.3883211678832117,
"grad_norm": 2.3125,
"learning_rate": 1.9543142472820548e-05,
"loss": 1.380842924118042,
"step": 266
},
{
"epoch": 0.39124087591240875,
"grad_norm": 1.4375,
"learning_rate": 1.9534174858961047e-05,
"loss": 1.1659390926361084,
"step": 268
},
{
"epoch": 0.39416058394160586,
"grad_norm": 3.390625,
"learning_rate": 1.9525122423529627e-05,
"loss": 1.3639929294586182,
"step": 270
},
{
"epoch": 0.3970802919708029,
"grad_norm": 3.546875,
"learning_rate": 1.951598525649887e-05,
"loss": 1.127337098121643,
"step": 272
},
{
"epoch": 0.4,
"grad_norm": 2.46875,
"learning_rate": 1.9506763448683516e-05,
"loss": 1.5703895092010498,
"step": 274
},
{
"epoch": 0.4029197080291971,
"grad_norm": 1.21875,
"learning_rate": 1.9497457091739557e-05,
"loss": 1.2389445304870605,
"step": 276
},
{
"epoch": 0.4058394160583942,
"grad_norm": 3.0,
"learning_rate": 1.9488066278163317e-05,
"loss": 1.139906406402588,
"step": 278
},
{
"epoch": 0.40875912408759124,
"grad_norm": 1.59375,
"learning_rate": 1.9478591101290543e-05,
"loss": 1.4885528087615967,
"step": 280
},
{
"epoch": 0.4116788321167883,
"grad_norm": 1.9921875,
"learning_rate": 1.946903165529547e-05,
"loss": 1.2521553039550781,
"step": 282
},
{
"epoch": 0.4145985401459854,
"grad_norm": 1.6171875,
"learning_rate": 1.9459388035189888e-05,
"loss": 1.2449872493743896,
"step": 284
},
{
"epoch": 0.41751824817518246,
"grad_norm": 1.2578125,
"learning_rate": 1.94496603368222e-05,
"loss": 1.1515569686889648,
"step": 286
},
{
"epoch": 0.42043795620437957,
"grad_norm": 1.140625,
"learning_rate": 1.9439848656876464e-05,
"loss": 1.2705414295196533,
"step": 288
},
{
"epoch": 0.4233576642335766,
"grad_norm": 1.21875,
"learning_rate": 1.9429953092871433e-05,
"loss": 1.2827699184417725,
"step": 290
},
{
"epoch": 0.42627737226277373,
"grad_norm": 1.5859375,
"learning_rate": 1.9419973743159596e-05,
"loss": 1.1295063495635986,
"step": 292
},
{
"epoch": 0.4291970802919708,
"grad_norm": 1.75,
"learning_rate": 1.9409910706926186e-05,
"loss": 1.1715033054351807,
"step": 294
},
{
"epoch": 0.4321167883211679,
"grad_norm": 2.796875,
"learning_rate": 1.93997640841882e-05,
"loss": 1.4130077362060547,
"step": 296
},
{
"epoch": 0.43503649635036495,
"grad_norm": 3.515625,
"learning_rate": 1.9389533975793405e-05,
"loss": 1.4112361669540405,
"step": 298
},
{
"epoch": 0.43795620437956206,
"grad_norm": 1.765625,
"learning_rate": 1.9379220483419342e-05,
"loss": 1.0633643865585327,
"step": 300
},
{
"epoch": 0.4408759124087591,
"grad_norm": 1.4140625,
"learning_rate": 1.9368823709572296e-05,
"loss": 1.2538483142852783,
"step": 302
},
{
"epoch": 0.44379562043795623,
"grad_norm": 1.34375,
"learning_rate": 1.935834375758631e-05,
"loss": 1.294917106628418,
"step": 304
},
{
"epoch": 0.4467153284671533,
"grad_norm": 4.09375,
"learning_rate": 1.934778073162212e-05,
"loss": 1.0700111389160156,
"step": 306
},
{
"epoch": 0.44963503649635034,
"grad_norm": 1.03125,
"learning_rate": 1.933713473666615e-05,
"loss": 0.9852163791656494,
"step": 308
},
{
"epoch": 0.45255474452554745,
"grad_norm": 2.921875,
"learning_rate": 1.9326405878529453e-05,
"loss": 1.5578174591064453,
"step": 310
},
{
"epoch": 0.4554744525547445,
"grad_norm": 1.96875,
"learning_rate": 1.931559426384666e-05,
"loss": 1.2433103322982788,
"step": 312
},
{
"epoch": 0.4583941605839416,
"grad_norm": 1.203125,
"learning_rate": 1.9304700000074927e-05,
"loss": 1.2242311239242554,
"step": 314
},
{
"epoch": 0.46131386861313867,
"grad_norm": 1.296875,
"learning_rate": 1.9293723195492864e-05,
"loss": 1.2348496913909912,
"step": 316
},
{
"epoch": 0.4642335766423358,
"grad_norm": 3.15625,
"learning_rate": 1.9282663959199455e-05,
"loss": 1.2458059787750244,
"step": 318
},
{
"epoch": 0.46715328467153283,
"grad_norm": 1.53125,
"learning_rate": 1.9271522401112978e-05,
"loss": 1.479885220527649,
"step": 320
},
{
"epoch": 0.47007299270072994,
"grad_norm": 1.8515625,
"learning_rate": 1.9260298631969917e-05,
"loss": 1.2687187194824219,
"step": 322
},
{
"epoch": 0.472992700729927,
"grad_norm": 1.78125,
"learning_rate": 1.9248992763323845e-05,
"loss": 1.2376818656921387,
"step": 324
},
{
"epoch": 0.4759124087591241,
"grad_norm": 0.859375,
"learning_rate": 1.9237604907544325e-05,
"loss": 1.2510634660720825,
"step": 326
},
{
"epoch": 0.47883211678832116,
"grad_norm": 2.171875,
"learning_rate": 1.9226135177815812e-05,
"loss": 1.365638017654419,
"step": 328
},
{
"epoch": 0.48175182481751827,
"grad_norm": 1.390625,
"learning_rate": 1.9214583688136493e-05,
"loss": 1.1146392822265625,
"step": 330
},
{
"epoch": 0.4846715328467153,
"grad_norm": 1.640625,
"learning_rate": 1.9202950553317173e-05,
"loss": 1.2945137023925781,
"step": 332
},
{
"epoch": 0.48759124087591244,
"grad_norm": 2.890625,
"learning_rate": 1.9191235888980142e-05,
"loss": 1.4794032573699951,
"step": 334
},
{
"epoch": 0.4905109489051095,
"grad_norm": 0.59375,
"learning_rate": 1.9179439811558e-05,
"loss": 1.0064594745635986,
"step": 336
},
{
"epoch": 0.49343065693430654,
"grad_norm": 3.46875,
"learning_rate": 1.9167562438292535e-05,
"loss": 1.498932123184204,
"step": 338
},
{
"epoch": 0.49635036496350365,
"grad_norm": 2.125,
"learning_rate": 1.915560388723352e-05,
"loss": 1.2725715637207031,
"step": 340
},
{
"epoch": 0.4992700729927007,
"grad_norm": 4.8125,
"learning_rate": 1.914356427723757e-05,
"loss": 1.3230655193328857,
"step": 342
},
{
"epoch": 0.5021897810218978,
"grad_norm": 1.4140625,
"learning_rate": 1.9131443727966934e-05,
"loss": 1.2821331024169922,
"step": 344
},
{
"epoch": 0.5051094890510949,
"grad_norm": 1.984375,
"learning_rate": 1.9119242359888344e-05,
"loss": 1.255110263824463,
"step": 346
},
{
"epoch": 0.5080291970802919,
"grad_norm": 1.3671875,
"learning_rate": 1.9106960294271774e-05,
"loss": 1.29063081741333,
"step": 348
},
{
"epoch": 0.5109489051094891,
"grad_norm": 2.21875,
"learning_rate": 1.9094597653189268e-05,
"loss": 1.0841729640960693,
"step": 350
},
{
"epoch": 0.5138686131386861,
"grad_norm": 2.375,
"learning_rate": 1.9082154559513694e-05,
"loss": 1.3983030319213867,
"step": 352
},
{
"epoch": 0.5167883211678832,
"grad_norm": 0.640625,
"learning_rate": 1.9069631136917566e-05,
"loss": 1.1833513975143433,
"step": 354
},
{
"epoch": 0.5197080291970803,
"grad_norm": 4.1875,
"learning_rate": 1.905702750987178e-05,
"loss": 1.2047162055969238,
"step": 356
},
{
"epoch": 0.5226277372262774,
"grad_norm": 2.015625,
"learning_rate": 1.9044343803644377e-05,
"loss": 1.3358124494552612,
"step": 358
},
{
"epoch": 0.5255474452554745,
"grad_norm": 1.5703125,
"learning_rate": 1.903158014429933e-05,
"loss": 1.316916823387146,
"step": 360
},
{
"epoch": 0.5284671532846715,
"grad_norm": 2.8125,
"learning_rate": 1.9018736658695256e-05,
"loss": 1.3399839401245117,
"step": 362
},
{
"epoch": 0.5313868613138686,
"grad_norm": 1.3046875,
"learning_rate": 1.9005813474484173e-05,
"loss": 1.0654617547988892,
"step": 364
},
{
"epoch": 0.5343065693430656,
"grad_norm": 3.671875,
"learning_rate": 1.8992810720110227e-05,
"loss": 1.374320387840271,
"step": 366
},
{
"epoch": 0.5372262773722628,
"grad_norm": 1.7109375,
"learning_rate": 1.8979728524808414e-05,
"loss": 1.1677433252334595,
"step": 368
},
{
"epoch": 0.5401459854014599,
"grad_norm": 3.953125,
"learning_rate": 1.8966567018603295e-05,
"loss": 1.1534755229949951,
"step": 370
},
{
"epoch": 0.5430656934306569,
"grad_norm": 3.46875,
"learning_rate": 1.8953326332307723e-05,
"loss": 1.060274600982666,
"step": 372
},
{
"epoch": 0.545985401459854,
"grad_norm": 1.828125,
"learning_rate": 1.894000659752149e-05,
"loss": 1.2867351770401,
"step": 374
},
{
"epoch": 0.5489051094890511,
"grad_norm": 1.265625,
"learning_rate": 1.8926607946630102e-05,
"loss": 1.4288403987884521,
"step": 376
},
{
"epoch": 0.5518248175182482,
"grad_norm": 6.03125,
"learning_rate": 1.891313051280337e-05,
"loss": 1.3884141445159912,
"step": 378
},
{
"epoch": 0.5547445255474452,
"grad_norm": 0.8671875,
"learning_rate": 1.8899574429994165e-05,
"loss": 1.1666433811187744,
"step": 380
},
{
"epoch": 0.5576642335766423,
"grad_norm": 1.2890625,
"learning_rate": 1.8885939832937037e-05,
"loss": 1.2260024547576904,
"step": 382
},
{
"epoch": 0.5605839416058395,
"grad_norm": 3.75,
"learning_rate": 1.8872226857146903e-05,
"loss": 1.1594487428665161,
"step": 384
},
{
"epoch": 0.5635036496350365,
"grad_norm": 2.1875,
"learning_rate": 1.8858435638917675e-05,
"loss": 1.2140047550201416,
"step": 386
},
{
"epoch": 0.5664233576642336,
"grad_norm": 1.9765625,
"learning_rate": 1.884456631532094e-05,
"loss": 1.4128262996673584,
"step": 388
},
{
"epoch": 0.5693430656934306,
"grad_norm": 1.28125,
"learning_rate": 1.8830619024204567e-05,
"loss": 1.2183442115783691,
"step": 390
},
{
"epoch": 0.5722627737226277,
"grad_norm": 1.625,
"learning_rate": 1.8816593904191343e-05,
"loss": 1.0716540813446045,
"step": 392
},
{
"epoch": 0.5751824817518248,
"grad_norm": 2.515625,
"learning_rate": 1.8802491094677616e-05,
"loss": 1.3082427978515625,
"step": 394
},
{
"epoch": 0.5781021897810219,
"grad_norm": 3.03125,
"learning_rate": 1.8788310735831875e-05,
"loss": 1.3730230331420898,
"step": 396
},
{
"epoch": 0.581021897810219,
"grad_norm": 2.8125,
"learning_rate": 1.8774052968593386e-05,
"loss": 1.3711814880371094,
"step": 398
},
{
"epoch": 0.583941605839416,
"grad_norm": 0.5078125,
"learning_rate": 1.8759717934670787e-05,
"loss": 1.0740855932235718,
"step": 400
},
{
"epoch": 0.5868613138686132,
"grad_norm": 1.828125,
"learning_rate": 1.8745305776540655e-05,
"loss": 1.3020991086959839,
"step": 402
},
{
"epoch": 0.5897810218978102,
"grad_norm": 2.265625,
"learning_rate": 1.8730816637446124e-05,
"loss": 1.2509665489196777,
"step": 404
},
{
"epoch": 0.5927007299270073,
"grad_norm": 0.83203125,
"learning_rate": 1.8716250661395444e-05,
"loss": 1.1264358758926392,
"step": 406
},
{
"epoch": 0.5956204379562043,
"grad_norm": 3.609375,
"learning_rate": 1.8701607993160537e-05,
"loss": 1.2399773597717285,
"step": 408
},
{
"epoch": 0.5985401459854015,
"grad_norm": 1.6953125,
"learning_rate": 1.8686888778275593e-05,
"loss": 1.2281595468521118,
"step": 410
},
{
"epoch": 0.6014598540145986,
"grad_norm": 0.71484375,
"learning_rate": 1.8672093163035585e-05,
"loss": 1.1488683223724365,
"step": 412
},
{
"epoch": 0.6043795620437956,
"grad_norm": 0.73828125,
"learning_rate": 1.8657221294494845e-05,
"loss": 1.2081717252731323,
"step": 414
},
{
"epoch": 0.6072992700729927,
"grad_norm": 1.40625,
"learning_rate": 1.8642273320465584e-05,
"loss": 1.2371435165405273,
"step": 416
},
{
"epoch": 0.6102189781021898,
"grad_norm": 3.546875,
"learning_rate": 1.8627249389516426e-05,
"loss": 1.0430543422698975,
"step": 418
},
{
"epoch": 0.6131386861313869,
"grad_norm": 4.84375,
"learning_rate": 1.8612149650970947e-05,
"loss": 1.475630521774292,
"step": 420
},
{
"epoch": 0.6160583941605839,
"grad_norm": 1.7578125,
"learning_rate": 1.8596974254906167e-05,
"loss": 1.204869270324707,
"step": 422
},
{
"epoch": 0.618978102189781,
"grad_norm": 2.921875,
"learning_rate": 1.858172335215107e-05,
"loss": 1.3133511543273926,
"step": 424
},
{
"epoch": 0.621897810218978,
"grad_norm": 2.15625,
"learning_rate": 1.856639709428511e-05,
"loss": 1.1244256496429443,
"step": 426
},
{
"epoch": 0.6248175182481752,
"grad_norm": 103.0,
"learning_rate": 1.85509956336367e-05,
"loss": 1.2631075382232666,
"step": 428
},
{
"epoch": 0.6277372262773723,
"grad_norm": 5.1875,
"learning_rate": 1.8535519123281686e-05,
"loss": 1.1989383697509766,
"step": 430
},
{
"epoch": 0.6306569343065693,
"grad_norm": 3.34375,
"learning_rate": 1.8519967717041852e-05,
"loss": 1.3323516845703125,
"step": 432
},
{
"epoch": 0.6335766423357664,
"grad_norm": 6.03125,
"learning_rate": 1.850434156948336e-05,
"loss": 1.2115147113800049,
"step": 434
},
{
"epoch": 0.6364963503649635,
"grad_norm": 3.796875,
"learning_rate": 1.8488640835915247e-05,
"loss": 1.1399335861206055,
"step": 436
},
{
"epoch": 0.6394160583941606,
"grad_norm": 4.6875,
"learning_rate": 1.8472865672387847e-05,
"loss": 1.0190486907958984,
"step": 438
},
{
"epoch": 0.6423357664233577,
"grad_norm": 0.75390625,
"learning_rate": 1.8457016235691268e-05,
"loss": 1.1691625118255615,
"step": 440
},
{
"epoch": 0.6452554744525547,
"grad_norm": 2.34375,
"learning_rate": 1.8441092683353822e-05,
"loss": 1.2145811319351196,
"step": 442
},
{
"epoch": 0.6481751824817519,
"grad_norm": 3.046875,
"learning_rate": 1.842509517364045e-05,
"loss": 1.3449723720550537,
"step": 444
},
{
"epoch": 0.6510948905109489,
"grad_norm": 2.546875,
"learning_rate": 1.840902386555117e-05,
"loss": 1.215829849243164,
"step": 446
},
{
"epoch": 0.654014598540146,
"grad_norm": 3.25,
"learning_rate": 1.839287891881948e-05,
"loss": 1.2495944499969482,
"step": 448
},
{
"epoch": 0.656934306569343,
"grad_norm": 1.1015625,
"learning_rate": 1.837666049391078e-05,
"loss": 1.3823912143707275,
"step": 450
},
{
"epoch": 0.6598540145985401,
"grad_norm": 1.6015625,
"learning_rate": 1.8360368752020764e-05,
"loss": 1.383428692817688,
"step": 452
},
{
"epoch": 0.6627737226277373,
"grad_norm": 0.71484375,
"learning_rate": 1.8344003855073842e-05,
"loss": 1.1197372674942017,
"step": 454
},
{
"epoch": 0.6656934306569343,
"grad_norm": 1.7109375,
"learning_rate": 1.8327565965721502e-05,
"loss": 1.2648000717163086,
"step": 456
},
{
"epoch": 0.6686131386861314,
"grad_norm": 7.28125,
"learning_rate": 1.8311055247340714e-05,
"loss": 0.9837288856506348,
"step": 458
},
{
"epoch": 0.6715328467153284,
"grad_norm": 1.7578125,
"learning_rate": 1.82944718640323e-05,
"loss": 1.2219568490982056,
"step": 460
},
{
"epoch": 0.6744525547445256,
"grad_norm": 2.046875,
"learning_rate": 1.827781598061929e-05,
"loss": 1.189807653427124,
"step": 462
},
{
"epoch": 0.6773722627737226,
"grad_norm": 1.8125,
"learning_rate": 1.8261087762645316e-05,
"loss": 0.8583734631538391,
"step": 464
},
{
"epoch": 0.6802919708029197,
"grad_norm": 3.0,
"learning_rate": 1.8244287376372932e-05,
"loss": 1.1929311752319336,
"step": 466
},
{
"epoch": 0.6832116788321168,
"grad_norm": 1.59375,
"learning_rate": 1.822741498878198e-05,
"loss": 1.255252480506897,
"step": 468
},
{
"epoch": 0.6861313868613139,
"grad_norm": 4.125,
"learning_rate": 1.8210470767567924e-05,
"loss": 1.1243486404418945,
"step": 470
},
{
"epoch": 0.689051094890511,
"grad_norm": 2.078125,
"learning_rate": 1.8193454881140195e-05,
"loss": 1.2001137733459473,
"step": 472
},
{
"epoch": 0.691970802919708,
"grad_norm": 1.6796875,
"learning_rate": 1.8176367498620497e-05,
"loss": 1.1785471439361572,
"step": 474
},
{
"epoch": 0.6948905109489051,
"grad_norm": 1.9140625,
"learning_rate": 1.8159208789841143e-05,
"loss": 1.1160335540771484,
"step": 476
},
{
"epoch": 0.6978102189781021,
"grad_norm": 2.75,
"learning_rate": 1.8141978925343362e-05,
"loss": 1.2501084804534912,
"step": 478
},
{
"epoch": 0.7007299270072993,
"grad_norm": 6.9375,
"learning_rate": 1.8124678076375598e-05,
"loss": 1.0928605794906616,
"step": 480
},
{
"epoch": 0.7036496350364964,
"grad_norm": 1.421875,
"learning_rate": 1.810730641489182e-05,
"loss": 1.272300124168396,
"step": 482
},
{
"epoch": 0.7065693430656934,
"grad_norm": 1.8984375,
"learning_rate": 1.8089864113549794e-05,
"loss": 1.201921820640564,
"step": 484
},
{
"epoch": 0.7094890510948905,
"grad_norm": 2.75,
"learning_rate": 1.8072351345709395e-05,
"loss": 1.1742405891418457,
"step": 486
},
{
"epoch": 0.7124087591240876,
"grad_norm": 1.1875,
"learning_rate": 1.805476828543086e-05,
"loss": 1.1028424501419067,
"step": 488
},
{
"epoch": 0.7153284671532847,
"grad_norm": 4.6875,
"learning_rate": 1.803711510747306e-05,
"loss": 1.251287817955017,
"step": 490
},
{
"epoch": 0.7182481751824817,
"grad_norm": 2.015625,
"learning_rate": 1.8019391987291775e-05,
"loss": 1.045281171798706,
"step": 492
},
{
"epoch": 0.7211678832116788,
"grad_norm": 1.875,
"learning_rate": 1.800159910103795e-05,
"loss": 1.1656675338745117,
"step": 494
},
{
"epoch": 0.724087591240876,
"grad_norm": 0.8125,
"learning_rate": 1.7983736625555926e-05,
"loss": 1.1865403652191162,
"step": 496
},
{
"epoch": 0.727007299270073,
"grad_norm": 1.9296875,
"learning_rate": 1.79658047383817e-05,
"loss": 1.105522871017456,
"step": 498
},
{
"epoch": 0.7299270072992701,
"grad_norm": 1.3828125,
"learning_rate": 1.794780361774117e-05,
"loss": 1.2371889352798462,
"step": 500
},
{
"epoch": 0.7328467153284671,
"grad_norm": 0.60546875,
"learning_rate": 1.7929733442548314e-05,
"loss": 1.2281248569488525,
"step": 502
},
{
"epoch": 0.7357664233576642,
"grad_norm": 1.53125,
"learning_rate": 1.7911594392403476e-05,
"loss": 1.4537439346313477,
"step": 504
},
{
"epoch": 0.7386861313868613,
"grad_norm": 0.984375,
"learning_rate": 1.789338664759155e-05,
"loss": 1.1025795936584473,
"step": 506
},
{
"epoch": 0.7416058394160584,
"grad_norm": 1.6015625,
"learning_rate": 1.787511038908017e-05,
"loss": 0.9628173112869263,
"step": 508
},
{
"epoch": 0.7445255474452555,
"grad_norm": 1.3984375,
"learning_rate": 1.7856765798517952e-05,
"loss": 0.7550945281982422,
"step": 510
},
{
"epoch": 0.7474452554744525,
"grad_norm": 1.6171875,
"learning_rate": 1.7838353058232654e-05,
"loss": 1.124826431274414,
"step": 512
},
{
"epoch": 0.7503649635036497,
"grad_norm": 2.8125,
"learning_rate": 1.7819872351229383e-05,
"loss": 1.1538705825805664,
"step": 514
},
{
"epoch": 0.7532846715328467,
"grad_norm": 1.4375,
"learning_rate": 1.780132386118877e-05,
"loss": 1.2035133838653564,
"step": 516
},
{
"epoch": 0.7562043795620438,
"grad_norm": 4.375,
"learning_rate": 1.7782707772465145e-05,
"loss": 1.1414223909378052,
"step": 518
},
{
"epoch": 0.7591240875912408,
"grad_norm": 2.265625,
"learning_rate": 1.77640242700847e-05,
"loss": 0.739942729473114,
"step": 520
},
{
"epoch": 0.762043795620438,
"grad_norm": 1.9765625,
"learning_rate": 1.7745273539743662e-05,
"loss": 1.208895206451416,
"step": 522
},
{
"epoch": 0.7649635036496351,
"grad_norm": 2.296875,
"learning_rate": 1.7726455767806436e-05,
"loss": 1.447777509689331,
"step": 524
},
{
"epoch": 0.7678832116788321,
"grad_norm": 2.796875,
"learning_rate": 1.7707571141303754e-05,
"loss": 1.4376122951507568,
"step": 526
},
{
"epoch": 0.7708029197080292,
"grad_norm": 0.90234375,
"learning_rate": 1.7688619847930827e-05,
"loss": 1.1304607391357422,
"step": 528
},
{
"epoch": 0.7737226277372263,
"grad_norm": 1.8046875,
"learning_rate": 1.7669602076045462e-05,
"loss": 1.0774279832839966,
"step": 530
},
{
"epoch": 0.7766423357664234,
"grad_norm": 9.0625,
"learning_rate": 1.7650518014666202e-05,
"loss": 1.4045021533966064,
"step": 532
},
{
"epoch": 0.7795620437956204,
"grad_norm": 26.25,
"learning_rate": 1.763136785347045e-05,
"loss": 1.4213430881500244,
"step": 534
},
{
"epoch": 0.7824817518248175,
"grad_norm": 1.8984375,
"learning_rate": 1.761215178279257e-05,
"loss": 1.2604089975357056,
"step": 536
},
{
"epoch": 0.7854014598540145,
"grad_norm": 1.3984375,
"learning_rate": 1.7592869993622005e-05,
"loss": 1.2845478057861328,
"step": 538
},
{
"epoch": 0.7883211678832117,
"grad_norm": 1.5859375,
"learning_rate": 1.757352267760138e-05,
"loss": 1.2007977962493896,
"step": 540
},
{
"epoch": 0.7912408759124088,
"grad_norm": 3.359375,
"learning_rate": 1.7554110027024606e-05,
"loss": 1.3627820014953613,
"step": 542
},
{
"epoch": 0.7941605839416058,
"grad_norm": 3.015625,
"learning_rate": 1.753463223483493e-05,
"loss": 1.5062451362609863,
"step": 544
},
{
"epoch": 0.7970802919708029,
"grad_norm": 1.5859375,
"learning_rate": 1.7515089494623064e-05,
"loss": 1.4260886907577515,
"step": 546
},
{
"epoch": 0.8,
"grad_norm": 2.140625,
"learning_rate": 1.749548200062524e-05,
"loss": 1.3138827085494995,
"step": 548
},
{
"epoch": 0.8029197080291971,
"grad_norm": 2.296875,
"learning_rate": 1.7475809947721276e-05,
"loss": 1.4285473823547363,
"step": 550
},
{
"epoch": 0.8058394160583942,
"grad_norm": 0.9765625,
"learning_rate": 1.745607353143265e-05,
"loss": 1.3441548347473145,
"step": 552
},
{
"epoch": 0.8087591240875912,
"grad_norm": 3.390625,
"learning_rate": 1.743627294792054e-05,
"loss": 1.2827033996582031,
"step": 554
},
{
"epoch": 0.8116788321167884,
"grad_norm": 1.703125,
"learning_rate": 1.7416408393983897e-05,
"loss": 1.1730741262435913,
"step": 556
},
{
"epoch": 0.8145985401459854,
"grad_norm": 1.6953125,
"learning_rate": 1.739648006705747e-05,
"loss": 1.2399193048477173,
"step": 558
},
{
"epoch": 0.8175182481751825,
"grad_norm": 3.75,
"learning_rate": 1.7376488165209856e-05,
"loss": 1.3772962093353271,
"step": 560
},
{
"epoch": 0.8204379562043795,
"grad_norm": 3.46875,
"learning_rate": 1.7356432887141522e-05,
"loss": 1.3898119926452637,
"step": 562
},
{
"epoch": 0.8233576642335766,
"grad_norm": 1.4609375,
"learning_rate": 1.7336314432182834e-05,
"loss": 1.200927495956421,
"step": 564
},
{
"epoch": 0.8262773722627738,
"grad_norm": 1.5625,
"learning_rate": 1.7316133000292083e-05,
"loss": 1.105010747909546,
"step": 566
},
{
"epoch": 0.8291970802919708,
"grad_norm": 3.96875,
"learning_rate": 1.7295888792053477e-05,
"loss": 1.3687732219696045,
"step": 568
},
{
"epoch": 0.8321167883211679,
"grad_norm": 1.65625,
"learning_rate": 1.7275582008675174e-05,
"loss": 1.1783493757247925,
"step": 570
},
{
"epoch": 0.8350364963503649,
"grad_norm": 4.875,
"learning_rate": 1.725521285198727e-05,
"loss": 1.2445366382598877,
"step": 572
},
{
"epoch": 0.8379562043795621,
"grad_norm": 2.734375,
"learning_rate": 1.7234781524439773e-05,
"loss": 1.1870619058609009,
"step": 574
},
{
"epoch": 0.8408759124087591,
"grad_norm": 2.203125,
"learning_rate": 1.7214288229100636e-05,
"loss": 1.3939554691314697,
"step": 576
},
{
"epoch": 0.8437956204379562,
"grad_norm": 1.96875,
"learning_rate": 1.71937331696537e-05,
"loss": 1.3431785106658936,
"step": 578
},
{
"epoch": 0.8467153284671532,
"grad_norm": 1.2265625,
"learning_rate": 1.7173116550396683e-05,
"loss": 1.2495307922363281,
"step": 580
},
{
"epoch": 0.8496350364963504,
"grad_norm": 1.7421875,
"learning_rate": 1.7152438576239146e-05,
"loss": 1.173033356666565,
"step": 582
},
{
"epoch": 0.8525547445255475,
"grad_norm": 0.859375,
"learning_rate": 1.7131699452700467e-05,
"loss": 1.092018961906433,
"step": 584
},
{
"epoch": 0.8554744525547445,
"grad_norm": 2.015625,
"learning_rate": 1.7110899385907784e-05,
"loss": 1.2871640920639038,
"step": 586
},
{
"epoch": 0.8583941605839416,
"grad_norm": 1.4453125,
"learning_rate": 1.709003858259395e-05,
"loss": 1.2312414646148682,
"step": 588
},
{
"epoch": 0.8613138686131386,
"grad_norm": 0.58984375,
"learning_rate": 1.7069117250095496e-05,
"loss": 1.1266227960586548,
"step": 590
},
{
"epoch": 0.8642335766423358,
"grad_norm": 1.4296875,
"learning_rate": 1.7048135596350544e-05,
"loss": 1.1233997344970703,
"step": 592
},
{
"epoch": 0.8671532846715329,
"grad_norm": 1.390625,
"learning_rate": 1.7027093829896748e-05,
"loss": 1.1032161712646484,
"step": 594
},
{
"epoch": 0.8700729927007299,
"grad_norm": 2.0,
"learning_rate": 1.7005992159869237e-05,
"loss": 1.1408822536468506,
"step": 596
},
{
"epoch": 0.872992700729927,
"grad_norm": 0.765625,
"learning_rate": 1.6984830795998513e-05,
"loss": 1.007914423942566,
"step": 598
},
{
"epoch": 0.8759124087591241,
"grad_norm": 0.83984375,
"learning_rate": 1.6963609948608382e-05,
"loss": 1.100301742553711,
"step": 600
},
{
"epoch": 0.8788321167883212,
"grad_norm": 0.69140625,
"learning_rate": 1.6942329828613868e-05,
"loss": 1.1425342559814453,
"step": 602
},
{
"epoch": 0.8817518248175182,
"grad_norm": 3.4375,
"learning_rate": 1.69209906475191e-05,
"loss": 1.2450380325317383,
"step": 604
},
{
"epoch": 0.8846715328467153,
"grad_norm": 1.6640625,
"learning_rate": 1.6899592617415212e-05,
"loss": 1.1817412376403809,
"step": 606
},
{
"epoch": 0.8875912408759125,
"grad_norm": 7.9375,
"learning_rate": 1.6878135950978255e-05,
"loss": 1.0453073978424072,
"step": 608
},
{
"epoch": 0.8905109489051095,
"grad_norm": 2.5625,
"learning_rate": 1.6856620861467058e-05,
"loss": 1.2505488395690918,
"step": 610
},
{
"epoch": 0.8934306569343066,
"grad_norm": 1.6953125,
"learning_rate": 1.6835047562721127e-05,
"loss": 1.1939127445220947,
"step": 612
},
{
"epoch": 0.8963503649635036,
"grad_norm": 2.515625,
"learning_rate": 1.6813416269158513e-05,
"loss": 1.4930554628372192,
"step": 614
},
{
"epoch": 0.8992700729927007,
"grad_norm": 0.67578125,
"learning_rate": 1.679172719577367e-05,
"loss": 1.073207974433899,
"step": 616
},
{
"epoch": 0.9021897810218978,
"grad_norm": 1.8125,
"learning_rate": 1.6769980558135334e-05,
"loss": 0.7536028623580933,
"step": 618
},
{
"epoch": 0.9051094890510949,
"grad_norm": 12.5,
"learning_rate": 1.6748176572384385e-05,
"loss": 1.0931081771850586,
"step": 620
},
{
"epoch": 0.908029197080292,
"grad_norm": 1.9765625,
"learning_rate": 1.6726315455231678e-05,
"loss": 1.282623291015625,
"step": 622
},
{
"epoch": 0.910948905109489,
"grad_norm": 2.09375,
"learning_rate": 1.670439742395589e-05,
"loss": 1.187821626663208,
"step": 624
},
{
"epoch": 0.9138686131386862,
"grad_norm": 1.28125,
"learning_rate": 1.6682422696401397e-05,
"loss": 1.1983062028884888,
"step": 626
},
{
"epoch": 0.9167883211678832,
"grad_norm": 3.65625,
"learning_rate": 1.6660391490976057e-05,
"loss": 1.4403834342956543,
"step": 628
},
{
"epoch": 0.9197080291970803,
"grad_norm": 2.21875,
"learning_rate": 1.6638304026649077e-05,
"loss": 1.2000679969787598,
"step": 630
},
{
"epoch": 0.9226277372262773,
"grad_norm": 1.109375,
"learning_rate": 1.661616052294882e-05,
"loss": 1.2217278480529785,
"step": 632
},
{
"epoch": 0.9255474452554745,
"grad_norm": 1.46875,
"learning_rate": 1.6593961199960634e-05,
"loss": 1.0947747230529785,
"step": 634
},
{
"epoch": 0.9284671532846716,
"grad_norm": 0.78515625,
"learning_rate": 1.657170627832464e-05,
"loss": 0.9176870584487915,
"step": 636
},
{
"epoch": 0.9313868613138686,
"grad_norm": 1.71875,
"learning_rate": 1.6549395979233578e-05,
"loss": 1.2252191305160522,
"step": 638
},
{
"epoch": 0.9343065693430657,
"grad_norm": 1.203125,
"learning_rate": 1.652703052443057e-05,
"loss": 1.0950853824615479,
"step": 640
},
{
"epoch": 0.9372262773722628,
"grad_norm": 1.703125,
"learning_rate": 1.6504610136206945e-05,
"loss": 1.1910455226898193,
"step": 642
},
{
"epoch": 0.9401459854014599,
"grad_norm": 0.90625,
"learning_rate": 1.6482135037400016e-05,
"loss": 1.1781953573226929,
"step": 644
},
{
"epoch": 0.9430656934306569,
"grad_norm": 1.359375,
"learning_rate": 1.645960545139086e-05,
"loss": 1.1620197296142578,
"step": 646
},
{
"epoch": 0.945985401459854,
"grad_norm": 37.75,
"learning_rate": 1.643702160210211e-05,
"loss": 0.6202739477157593,
"step": 648
},
{
"epoch": 0.948905109489051,
"grad_norm": 1.484375,
"learning_rate": 1.6414383713995723e-05,
"loss": 1.2171499729156494,
"step": 650
},
{
"epoch": 0.9518248175182482,
"grad_norm": 3.453125,
"learning_rate": 1.6391692012070754e-05,
"loss": 1.1996110677719116,
"step": 652
},
{
"epoch": 0.9547445255474453,
"grad_norm": 7.25,
"learning_rate": 1.6368946721861113e-05,
"loss": 1.2185020446777344,
"step": 654
},
{
"epoch": 0.9576642335766423,
"grad_norm": 4.375,
"learning_rate": 1.6346148069433324e-05,
"loss": 1.2642889022827148,
"step": 656
},
{
"epoch": 0.9605839416058394,
"grad_norm": 1.1171875,
"learning_rate": 1.6323296281384283e-05,
"loss": 1.1335656642913818,
"step": 658
},
{
"epoch": 0.9635036496350365,
"grad_norm": 1.6796875,
"learning_rate": 1.6300391584839007e-05,
"loss": 1.1738358736038208,
"step": 660
},
{
"epoch": 0.9664233576642336,
"grad_norm": 0.72265625,
"learning_rate": 1.627743420744837e-05,
"loss": 1.09047532081604,
"step": 662
},
{
"epoch": 0.9693430656934306,
"grad_norm": 1.390625,
"learning_rate": 1.6254424377386843e-05,
"loss": 1.1291742324829102,
"step": 664
},
{
"epoch": 0.9722627737226277,
"grad_norm": 6.0,
"learning_rate": 1.6231362323350224e-05,
"loss": 1.341015100479126,
"step": 666
},
{
"epoch": 0.9751824817518249,
"grad_norm": 4.71875,
"learning_rate": 1.6208248274553376e-05,
"loss": 1.4208297729492188,
"step": 668
},
{
"epoch": 0.9781021897810219,
"grad_norm": 4.78125,
"learning_rate": 1.618508246072793e-05,
"loss": 1.3850116729736328,
"step": 670
},
{
"epoch": 0.981021897810219,
"grad_norm": 0.8828125,
"learning_rate": 1.616186511212002e-05,
"loss": 1.1544121503829956,
"step": 672
},
{
"epoch": 0.983941605839416,
"grad_norm": 2.765625,
"learning_rate": 1.613859645948798e-05,
"loss": 1.299857497215271,
"step": 674
},
{
"epoch": 0.9868613138686131,
"grad_norm": 3.3125,
"learning_rate": 1.611527673410007e-05,
"loss": 0.34064072370529175,
"step": 676
},
{
"epoch": 0.9897810218978103,
"grad_norm": 1.171875,
"learning_rate": 1.609190616773214e-05,
"loss": 1.1542079448699951,
"step": 678
},
{
"epoch": 0.9927007299270073,
"grad_norm": 2.015625,
"learning_rate": 1.6068484992665385e-05,
"loss": 1.1845728158950806,
"step": 680
},
{
"epoch": 0.9956204379562044,
"grad_norm": 2.59375,
"learning_rate": 1.604501344168397e-05,
"loss": 1.1579151153564453,
"step": 682
},
{
"epoch": 0.9985401459854014,
"grad_norm": 3.640625,
"learning_rate": 1.6021491748072768e-05,
"loss": 1.417609453201294,
"step": 684
},
{
"epoch": 1.0014598540145985,
"grad_norm": 0.79296875,
"learning_rate": 1.5997920145615012e-05,
"loss": 1.2141101360321045,
"step": 686
},
{
"epoch": 1.0043795620437956,
"grad_norm": 2.171875,
"learning_rate": 1.5974298868589992e-05,
"loss": 1.1417958736419678,
"step": 688
},
{
"epoch": 1.0072992700729928,
"grad_norm": 3.125,
"learning_rate": 1.5950628151770705e-05,
"loss": 1.2956650257110596,
"step": 690
},
{
"epoch": 1.0102189781021897,
"grad_norm": 3.515625,
"learning_rate": 1.592690823042154e-05,
"loss": 1.1093413829803467,
"step": 692
},
{
"epoch": 1.013138686131387,
"grad_norm": 3.046875,
"learning_rate": 1.5903139340295924e-05,
"loss": 1.0855910778045654,
"step": 694
},
{
"epoch": 1.0160583941605839,
"grad_norm": 2.015625,
"learning_rate": 1.5879321717634004e-05,
"loss": 1.0838496685028076,
"step": 696
},
{
"epoch": 1.018978102189781,
"grad_norm": 1.75,
"learning_rate": 1.5855455599160256e-05,
"loss": 1.2246263027191162,
"step": 698
},
{
"epoch": 1.0218978102189782,
"grad_norm": 1.453125,
"learning_rate": 1.5831541222081178e-05,
"loss": 1.0423314571380615,
"step": 700
},
{
"epoch": 1.0248175182481751,
"grad_norm": 1.7109375,
"learning_rate": 1.5807578824082916e-05,
"loss": 1.1862300634384155,
"step": 702
},
{
"epoch": 1.0277372262773723,
"grad_norm": 1.734375,
"learning_rate": 1.5783568643328874e-05,
"loss": 1.153383493423462,
"step": 704
},
{
"epoch": 1.0306569343065692,
"grad_norm": 3.921875,
"learning_rate": 1.5759510918457396e-05,
"loss": 1.1981422901153564,
"step": 706
},
{
"epoch": 1.0335766423357664,
"grad_norm": 2.015625,
"learning_rate": 1.5735405888579356e-05,
"loss": 1.156768798828125,
"step": 708
},
{
"epoch": 1.0364963503649636,
"grad_norm": 2.46875,
"learning_rate": 1.5711253793275806e-05,
"loss": 1.3564773797988892,
"step": 710
},
{
"epoch": 1.0394160583941605,
"grad_norm": 3.015625,
"learning_rate": 1.568705487259556e-05,
"loss": 1.1167981624603271,
"step": 712
},
{
"epoch": 1.0423357664233577,
"grad_norm": 1.0625,
"learning_rate": 1.5662809367052864e-05,
"loss": 1.086395025253296,
"step": 714
},
{
"epoch": 1.0452554744525548,
"grad_norm": 2.140625,
"learning_rate": 1.5638517517624947e-05,
"loss": 1.1450870037078857,
"step": 716
},
{
"epoch": 1.0481751824817518,
"grad_norm": 3.671875,
"learning_rate": 1.5614179565749666e-05,
"loss": 0.9007392525672913,
"step": 718
},
{
"epoch": 1.051094890510949,
"grad_norm": 2.40625,
"learning_rate": 1.5589795753323084e-05,
"loss": 1.0877478122711182,
"step": 720
},
{
"epoch": 1.054014598540146,
"grad_norm": 8.5,
"learning_rate": 1.556536632269709e-05,
"loss": 0.7627002000808716,
"step": 722
},
{
"epoch": 1.056934306569343,
"grad_norm": 3.421875,
"learning_rate": 1.5540891516676943e-05,
"loss": 1.2761828899383545,
"step": 724
},
{
"epoch": 1.0598540145985402,
"grad_norm": 1.5625,
"learning_rate": 1.5516371578518926e-05,
"loss": 1.0273706912994385,
"step": 726
},
{
"epoch": 1.0627737226277372,
"grad_norm": 1.3359375,
"learning_rate": 1.5491806751927877e-05,
"loss": 0.8984286189079285,
"step": 728
},
{
"epoch": 1.0656934306569343,
"grad_norm": 3.515625,
"learning_rate": 1.546719728105477e-05,
"loss": 1.1889033317565918,
"step": 730
},
{
"epoch": 1.0686131386861315,
"grad_norm": 2.421875,
"learning_rate": 1.5442543410494326e-05,
"loss": 1.0481176376342773,
"step": 732
},
{
"epoch": 1.0715328467153284,
"grad_norm": 4.1875,
"learning_rate": 1.5417845385282528e-05,
"loss": 0.7755412459373474,
"step": 734
},
{
"epoch": 1.0744525547445256,
"grad_norm": 1.203125,
"learning_rate": 1.5393103450894244e-05,
"loss": 1.0950218439102173,
"step": 736
},
{
"epoch": 1.0773722627737226,
"grad_norm": 3.765625,
"learning_rate": 1.5368317853240733e-05,
"loss": 1.052286982536316,
"step": 738
},
{
"epoch": 1.0802919708029197,
"grad_norm": 1.75,
"learning_rate": 1.5343488838667244e-05,
"loss": 0.8312653303146362,
"step": 740
},
{
"epoch": 1.0832116788321169,
"grad_norm": 1.21875,
"learning_rate": 1.531861665395053e-05,
"loss": 1.089158535003662,
"step": 742
},
{
"epoch": 1.0861313868613138,
"grad_norm": 4.03125,
"learning_rate": 1.529370154629643e-05,
"loss": 1.0753670930862427,
"step": 744
},
{
"epoch": 1.089051094890511,
"grad_norm": 1.7578125,
"learning_rate": 1.5268743763337387e-05,
"loss": 1.0492002964019775,
"step": 746
},
{
"epoch": 1.091970802919708,
"grad_norm": 3.796875,
"learning_rate": 1.5243743553130004e-05,
"loss": 1.1771512031555176,
"step": 748
},
{
"epoch": 1.094890510948905,
"grad_norm": 1.2109375,
"learning_rate": 1.521870116415256e-05,
"loss": 1.0497887134552002,
"step": 750
},
{
"epoch": 1.0978102189781023,
"grad_norm": 1.3984375,
"learning_rate": 1.5193616845302559e-05,
"loss": 0.8820348978042603,
"step": 752
},
{
"epoch": 1.1007299270072992,
"grad_norm": 1.5390625,
"learning_rate": 1.5168490845894242e-05,
"loss": 1.0606813430786133,
"step": 754
},
{
"epoch": 1.1036496350364964,
"grad_norm": 1.46875,
"learning_rate": 1.5143323415656124e-05,
"loss": 0.9001914858818054,
"step": 756
},
{
"epoch": 1.1065693430656935,
"grad_norm": 1.2265625,
"learning_rate": 1.5118114804728495e-05,
"loss": 0.6586837768554688,
"step": 758
},
{
"epoch": 1.1094890510948905,
"grad_norm": 2.359375,
"learning_rate": 1.5092865263660943e-05,
"loss": 1.0042301416397095,
"step": 760
},
{
"epoch": 1.1124087591240877,
"grad_norm": 2.015625,
"learning_rate": 1.506757504340986e-05,
"loss": 1.0140289068222046,
"step": 762
},
{
"epoch": 1.1153284671532846,
"grad_norm": 0.84765625,
"learning_rate": 1.5042244395335964e-05,
"loss": 1.0186865329742432,
"step": 764
},
{
"epoch": 1.1182481751824818,
"grad_norm": 5.5625,
"learning_rate": 1.5016873571201768e-05,
"loss": 0.9784363508224487,
"step": 766
},
{
"epoch": 1.121167883211679,
"grad_norm": 0.76953125,
"learning_rate": 1.4991462823169113e-05,
"loss": 0.7668118476867676,
"step": 768
},
{
"epoch": 1.1240875912408759,
"grad_norm": 1.6015625,
"learning_rate": 1.4966012403796635e-05,
"loss": 1.0627202987670898,
"step": 770
},
{
"epoch": 1.127007299270073,
"grad_norm": 2.1875,
"learning_rate": 1.4940522566037265e-05,
"loss": 1.0267741680145264,
"step": 772
},
{
"epoch": 1.12992700729927,
"grad_norm": 1.875,
"learning_rate": 1.4914993563235728e-05,
"loss": 0.9466987252235413,
"step": 774
},
{
"epoch": 1.1328467153284671,
"grad_norm": 1.1328125,
"learning_rate": 1.4889425649125995e-05,
"loss": 0.966658353805542,
"step": 776
},
{
"epoch": 1.1357664233576643,
"grad_norm": 1.5859375,
"learning_rate": 1.4863819077828792e-05,
"loss": 1.0004557371139526,
"step": 778
},
{
"epoch": 1.1386861313868613,
"grad_norm": 1.3359375,
"learning_rate": 1.4838174103849052e-05,
"loss": 0.9684220552444458,
"step": 780
},
{
"epoch": 1.1416058394160584,
"grad_norm": 1.421875,
"learning_rate": 1.4812490982073403e-05,
"loss": 1.0590606927871704,
"step": 782
},
{
"epoch": 1.1445255474452556,
"grad_norm": 2.421875,
"learning_rate": 1.478676996776761e-05,
"loss": 1.0327863693237305,
"step": 784
},
{
"epoch": 1.1474452554744525,
"grad_norm": 1.890625,
"learning_rate": 1.4761011316574075e-05,
"loss": 0.9912806749343872,
"step": 786
},
{
"epoch": 1.1503649635036497,
"grad_norm": 1.5234375,
"learning_rate": 1.4735215284509255e-05,
"loss": 0.8372938632965088,
"step": 788
},
{
"epoch": 1.1532846715328466,
"grad_norm": 1.5,
"learning_rate": 1.4709382127961153e-05,
"loss": 1.014420747756958,
"step": 790
},
{
"epoch": 1.1562043795620438,
"grad_norm": 3.171875,
"learning_rate": 1.4683512103686742e-05,
"loss": 0.8318605422973633,
"step": 792
},
{
"epoch": 1.159124087591241,
"grad_norm": 2.359375,
"learning_rate": 1.465760546880943e-05,
"loss": 1.0174920558929443,
"step": 794
},
{
"epoch": 1.162043795620438,
"grad_norm": 2.0625,
"learning_rate": 1.4631662480816504e-05,
"loss": 1.0595332384109497,
"step": 796
},
{
"epoch": 1.164963503649635,
"grad_norm": 2.296875,
"learning_rate": 1.4605683397556559e-05,
"loss": 0.9171582460403442,
"step": 798
},
{
"epoch": 1.167883211678832,
"grad_norm": 1.84375,
"learning_rate": 1.4579668477236944e-05,
"loss": 0.9063014388084412,
"step": 800
},
{
"epoch": 1.1708029197080292,
"grad_norm": 2.375,
"learning_rate": 1.4553617978421197e-05,
"loss": 0.8801769018173218,
"step": 802
},
{
"epoch": 1.1737226277372264,
"grad_norm": 2.859375,
"learning_rate": 1.4527532160026472e-05,
"loss": 1.0798723697662354,
"step": 804
},
{
"epoch": 1.1766423357664233,
"grad_norm": 6.40625,
"learning_rate": 1.4501411281320965e-05,
"loss": 0.9010332822799683,
"step": 806
},
{
"epoch": 1.1795620437956205,
"grad_norm": 3.921875,
"learning_rate": 1.4475255601921337e-05,
"loss": 1.2930481433868408,
"step": 808
},
{
"epoch": 1.1824817518248176,
"grad_norm": 1.484375,
"learning_rate": 1.444906538179014e-05,
"loss": 0.9960885047912598,
"step": 810
},
{
"epoch": 1.1854014598540146,
"grad_norm": 2.828125,
"learning_rate": 1.4422840881233222e-05,
"loss": 1.1361961364746094,
"step": 812
},
{
"epoch": 1.1883211678832117,
"grad_norm": 4.21875,
"learning_rate": 1.439658236089714e-05,
"loss": 1.0548433065414429,
"step": 814
},
{
"epoch": 1.1912408759124087,
"grad_norm": 0.828125,
"learning_rate": 1.4370290081766596e-05,
"loss": 0.9712170362472534,
"step": 816
},
{
"epoch": 1.1941605839416058,
"grad_norm": 0.9296875,
"learning_rate": 1.43439643051618e-05,
"loss": 0.7919751405715942,
"step": 818
},
{
"epoch": 1.197080291970803,
"grad_norm": 1.625,
"learning_rate": 1.4317605292735905e-05,
"loss": 1.25571870803833,
"step": 820
},
{
"epoch": 1.2,
"grad_norm": 1.4140625,
"learning_rate": 1.4291213306472402e-05,
"loss": 0.9223061800003052,
"step": 822
},
{
"epoch": 1.2029197080291971,
"grad_norm": 2.203125,
"learning_rate": 1.4264788608682499e-05,
"loss": 0.91511070728302,
"step": 824
},
{
"epoch": 1.205839416058394,
"grad_norm": 6.375,
"learning_rate": 1.423833146200253e-05,
"loss": 0.6899645328521729,
"step": 826
},
{
"epoch": 1.2087591240875912,
"grad_norm": 2.328125,
"learning_rate": 1.4211842129391345e-05,
"loss": 0.991096019744873,
"step": 828
},
{
"epoch": 1.2116788321167884,
"grad_norm": 3.546875,
"learning_rate": 1.4185320874127684e-05,
"loss": 0.6836661100387573,
"step": 830
},
{
"epoch": 1.2145985401459853,
"grad_norm": 4.3125,
"learning_rate": 1.4158767959807572e-05,
"loss": 1.1839122772216797,
"step": 832
},
{
"epoch": 1.2175182481751825,
"grad_norm": 1.4375,
"learning_rate": 1.4132183650341689e-05,
"loss": 0.9149377346038818,
"step": 834
},
{
"epoch": 1.2204379562043797,
"grad_norm": 1.6484375,
"learning_rate": 1.4105568209952768e-05,
"loss": 0.5033729076385498,
"step": 836
},
{
"epoch": 1.2233576642335766,
"grad_norm": 1.7734375,
"learning_rate": 1.4078921903172933e-05,
"loss": 0.4498552978038788,
"step": 838
},
{
"epoch": 1.2262773722627738,
"grad_norm": 4.21875,
"learning_rate": 1.4052244994841109e-05,
"loss": 0.5252397060394287,
"step": 840
},
{
"epoch": 1.2291970802919707,
"grad_norm": 4.0625,
"learning_rate": 1.402553775010036e-05,
"loss": 0.4305902123451233,
"step": 842
},
{
"epoch": 1.2321167883211679,
"grad_norm": 1.5,
"learning_rate": 1.3998800434395266e-05,
"loss": 0.8443759679794312,
"step": 844
},
{
"epoch": 1.235036496350365,
"grad_norm": 2.484375,
"learning_rate": 1.397203331346929e-05,
"loss": 0.9912139177322388,
"step": 846
},
{
"epoch": 1.237956204379562,
"grad_norm": 1.5078125,
"learning_rate": 1.3945236653362125e-05,
"loss": 0.9120661020278931,
"step": 848
},
{
"epoch": 1.2408759124087592,
"grad_norm": 2.140625,
"learning_rate": 1.3918410720407059e-05,
"loss": 0.9410139918327332,
"step": 850
},
{
"epoch": 1.243795620437956,
"grad_norm": 1.5234375,
"learning_rate": 1.3891555781228314e-05,
"loss": 0.836361289024353,
"step": 852
},
{
"epoch": 1.2467153284671533,
"grad_norm": 1.3046875,
"learning_rate": 1.3864672102738423e-05,
"loss": 0.9195941686630249,
"step": 854
},
{
"epoch": 1.2496350364963504,
"grad_norm": 1.1484375,
"learning_rate": 1.383775995213555e-05,
"loss": 0.8999115824699402,
"step": 856
},
{
"epoch": 1.2525547445255474,
"grad_norm": 5.1875,
"learning_rate": 1.3810819596900847e-05,
"loss": 0.7722897529602051,
"step": 858
},
{
"epoch": 1.2554744525547445,
"grad_norm": 2.015625,
"learning_rate": 1.3783851304795786e-05,
"loss": 0.7147636413574219,
"step": 860
},
{
"epoch": 1.2583941605839417,
"grad_norm": 1.0625,
"learning_rate": 1.375685534385952e-05,
"loss": 0.8630459308624268,
"step": 862
},
{
"epoch": 1.2613138686131387,
"grad_norm": 1.625,
"learning_rate": 1.3729831982406192e-05,
"loss": 0.9289191365242004,
"step": 864
},
{
"epoch": 1.2642335766423358,
"grad_norm": 2.75,
"learning_rate": 1.3702781489022285e-05,
"loss": 0.6711573004722595,
"step": 866
},
{
"epoch": 1.2671532846715328,
"grad_norm": 1.296875,
"learning_rate": 1.367570413256394e-05,
"loss": 1.0273014307022095,
"step": 868
},
{
"epoch": 1.27007299270073,
"grad_norm": 1.9140625,
"learning_rate": 1.3648600182154305e-05,
"loss": 0.8584736585617065,
"step": 870
},
{
"epoch": 1.2729927007299269,
"grad_norm": 0.953125,
"learning_rate": 1.3621469907180835e-05,
"loss": 0.8978360891342163,
"step": 872
},
{
"epoch": 1.275912408759124,
"grad_norm": 1.953125,
"learning_rate": 1.3594313577292632e-05,
"loss": 0.9334902763366699,
"step": 874
},
{
"epoch": 1.2788321167883212,
"grad_norm": 1.1875,
"learning_rate": 1.3567131462397753e-05,
"loss": 0.9140644073486328,
"step": 876
},
{
"epoch": 1.2817518248175181,
"grad_norm": 3.09375,
"learning_rate": 1.3539923832660541e-05,
"loss": 0.8537070155143738,
"step": 878
},
{
"epoch": 1.2846715328467153,
"grad_norm": 1.3828125,
"learning_rate": 1.3512690958498929e-05,
"loss": 0.9819170832633972,
"step": 880
},
{
"epoch": 1.2875912408759125,
"grad_norm": 2.28125,
"learning_rate": 1.3485433110581754e-05,
"loss": 0.9769294261932373,
"step": 882
},
{
"epoch": 1.2905109489051094,
"grad_norm": 2.96875,
"learning_rate": 1.3458150559826066e-05,
"loss": 1.1729984283447266,
"step": 884
},
{
"epoch": 1.2934306569343066,
"grad_norm": 1.609375,
"learning_rate": 1.3430843577394452e-05,
"loss": 0.7712341547012329,
"step": 886
},
{
"epoch": 1.2963503649635038,
"grad_norm": 2.453125,
"learning_rate": 1.3403512434692303e-05,
"loss": 1.0945894718170166,
"step": 888
},
{
"epoch": 1.2992700729927007,
"grad_norm": 1.609375,
"learning_rate": 1.3376157403365162e-05,
"loss": 1.0084092617034912,
"step": 890
},
{
"epoch": 1.3021897810218979,
"grad_norm": 2.96875,
"learning_rate": 1.3348778755295986e-05,
"loss": 0.9818345308303833,
"step": 892
},
{
"epoch": 1.305109489051095,
"grad_norm": 2.625,
"learning_rate": 1.3321376762602475e-05,
"loss": 0.9358723163604736,
"step": 894
},
{
"epoch": 1.308029197080292,
"grad_norm": 2.4375,
"learning_rate": 1.329395169763434e-05,
"loss": 1.0269190073013306,
"step": 896
},
{
"epoch": 1.310948905109489,
"grad_norm": 1.328125,
"learning_rate": 1.3266503832970612e-05,
"loss": 0.9094824194908142,
"step": 898
},
{
"epoch": 1.313868613138686,
"grad_norm": 1.328125,
"learning_rate": 1.3239033441416933e-05,
"loss": 0.9367675185203552,
"step": 900
},
{
"epoch": 1.3167883211678832,
"grad_norm": 0.953125,
"learning_rate": 1.3211540796002832e-05,
"loss": 0.8247213363647461,
"step": 902
},
{
"epoch": 1.3197080291970802,
"grad_norm": 2.671875,
"learning_rate": 1.3184026169979034e-05,
"loss": 0.6118776798248291,
"step": 904
},
{
"epoch": 1.3226277372262774,
"grad_norm": 3.4375,
"learning_rate": 1.3156489836814718e-05,
"loss": 0.7799744009971619,
"step": 906
},
{
"epoch": 1.3255474452554745,
"grad_norm": 0.96484375,
"learning_rate": 1.312893207019482e-05,
"loss": 0.9586873650550842,
"step": 908
},
{
"epoch": 1.3284671532846715,
"grad_norm": 1.65625,
"learning_rate": 1.310135314401729e-05,
"loss": 0.9315115809440613,
"step": 910
},
{
"epoch": 1.3313868613138686,
"grad_norm": 0.84375,
"learning_rate": 1.3073753332390404e-05,
"loss": 1.0107285976409912,
"step": 912
},
{
"epoch": 1.3343065693430658,
"grad_norm": 1.0546875,
"learning_rate": 1.3046132909630006e-05,
"loss": 0.7950001955032349,
"step": 914
},
{
"epoch": 1.3372262773722627,
"grad_norm": 6.75,
"learning_rate": 1.3018492150256801e-05,
"loss": 1.0125911235809326,
"step": 916
},
{
"epoch": 1.34014598540146,
"grad_norm": 3.734375,
"learning_rate": 1.299083132899361e-05,
"loss": 0.5360926389694214,
"step": 918
},
{
"epoch": 1.343065693430657,
"grad_norm": 1.4140625,
"learning_rate": 1.2963150720762662e-05,
"loss": 1.1542425155639648,
"step": 920
},
{
"epoch": 1.345985401459854,
"grad_norm": 1.1640625,
"learning_rate": 1.293545060068285e-05,
"loss": 0.783527135848999,
"step": 922
},
{
"epoch": 1.348905109489051,
"grad_norm": 2.875,
"learning_rate": 1.2907731244066987e-05,
"loss": 1.0478227138519287,
"step": 924
},
{
"epoch": 1.3518248175182481,
"grad_norm": 0.765625,
"learning_rate": 1.2879992926419088e-05,
"loss": 1.1474854946136475,
"step": 926
},
{
"epoch": 1.3547445255474453,
"grad_norm": 1.7421875,
"learning_rate": 1.2852235923431609e-05,
"loss": 0.7717668414115906,
"step": 928
},
{
"epoch": 1.3576642335766422,
"grad_norm": 1.140625,
"learning_rate": 1.2824460510982737e-05,
"loss": 0.7952892184257507,
"step": 930
},
{
"epoch": 1.3605839416058394,
"grad_norm": 2.234375,
"learning_rate": 1.2796666965133622e-05,
"loss": 1.0578279495239258,
"step": 932
},
{
"epoch": 1.3635036496350366,
"grad_norm": 1.0703125,
"learning_rate": 1.276885556212565e-05,
"loss": 0.8681128025054932,
"step": 934
},
{
"epoch": 1.3664233576642335,
"grad_norm": 1.046875,
"learning_rate": 1.2741026578377676e-05,
"loss": 0.6660016775131226,
"step": 936
},
{
"epoch": 1.3693430656934307,
"grad_norm": 3.546875,
"learning_rate": 1.271318029048331e-05,
"loss": 0.8460044860839844,
"step": 938
},
{
"epoch": 1.3722627737226278,
"grad_norm": 1.4140625,
"learning_rate": 1.2685316975208128e-05,
"loss": 1.2607700824737549,
"step": 940
},
{
"epoch": 1.3751824817518248,
"grad_norm": 1.9140625,
"learning_rate": 1.265743690948696e-05,
"loss": 0.9621885418891907,
"step": 942
},
{
"epoch": 1.378102189781022,
"grad_norm": 1.7890625,
"learning_rate": 1.2629540370421112e-05,
"loss": 0.8699158430099487,
"step": 944
},
{
"epoch": 1.3810218978102191,
"grad_norm": 1.546875,
"learning_rate": 1.2601627635275615e-05,
"loss": 0.9138844013214111,
"step": 946
},
{
"epoch": 1.383941605839416,
"grad_norm": 1.7734375,
"learning_rate": 1.257369898147648e-05,
"loss": 1.0127596855163574,
"step": 948
},
{
"epoch": 1.3868613138686132,
"grad_norm": 3.890625,
"learning_rate": 1.2545754686607936e-05,
"loss": 0.8778249621391296,
"step": 950
},
{
"epoch": 1.3897810218978102,
"grad_norm": 1.9765625,
"learning_rate": 1.251779502840966e-05,
"loss": 0.989855170249939,
"step": 952
},
{
"epoch": 1.3927007299270073,
"grad_norm": 1.375,
"learning_rate": 1.2489820284774037e-05,
"loss": 0.9114617109298706,
"step": 954
},
{
"epoch": 1.3956204379562043,
"grad_norm": 1.53125,
"learning_rate": 1.2461830733743377e-05,
"loss": 1.0407018661499023,
"step": 956
},
{
"epoch": 1.3985401459854014,
"grad_norm": 5.4375,
"learning_rate": 1.2433826653507165e-05,
"loss": 0.9131146669387817,
"step": 958
},
{
"epoch": 1.4014598540145986,
"grad_norm": 2.90625,
"learning_rate": 1.2405808322399297e-05,
"loss": 1.0465645790100098,
"step": 960
},
{
"epoch": 1.4043795620437955,
"grad_norm": 1.09375,
"learning_rate": 1.2377776018895304e-05,
"loss": 0.8719322681427002,
"step": 962
},
{
"epoch": 1.4072992700729927,
"grad_norm": 2.65625,
"learning_rate": 1.2349730021609591e-05,
"loss": 0.9842914938926697,
"step": 964
},
{
"epoch": 1.4102189781021899,
"grad_norm": 1.7109375,
"learning_rate": 1.2321670609292667e-05,
"loss": 0.9857591390609741,
"step": 966
},
{
"epoch": 1.4131386861313868,
"grad_norm": 1.0859375,
"learning_rate": 1.2293598060828368e-05,
"loss": 0.8483753204345703,
"step": 968
},
{
"epoch": 1.416058394160584,
"grad_norm": 0.5703125,
"learning_rate": 1.2265512655231102e-05,
"loss": 0.7562495470046997,
"step": 970
},
{
"epoch": 1.4189781021897812,
"grad_norm": 3.203125,
"learning_rate": 1.2237414671643052e-05,
"loss": 0.9674968123435974,
"step": 972
},
{
"epoch": 1.421897810218978,
"grad_norm": 2.171875,
"learning_rate": 1.220930438933142e-05,
"loss": 1.0779824256896973,
"step": 974
},
{
"epoch": 1.4248175182481753,
"grad_norm": 8.4375,
"learning_rate": 1.2181182087685652e-05,
"loss": 0.9298703670501709,
"step": 976
},
{
"epoch": 1.4277372262773722,
"grad_norm": 1.109375,
"learning_rate": 1.2153048046214644e-05,
"loss": 0.9324536323547363,
"step": 978
},
{
"epoch": 1.4306569343065694,
"grad_norm": 1.8359375,
"learning_rate": 1.2124902544543977e-05,
"loss": 0.9824227094650269,
"step": 980
},
{
"epoch": 1.4335766423357663,
"grad_norm": 3.359375,
"learning_rate": 1.2096745862413141e-05,
"loss": 0.8633835911750793,
"step": 982
},
{
"epoch": 1.4364963503649635,
"grad_norm": 1.046875,
"learning_rate": 1.2068578279672752e-05,
"loss": 0.9417979121208191,
"step": 984
},
{
"epoch": 1.4394160583941606,
"grad_norm": 1.1875,
"learning_rate": 1.2040400076281747e-05,
"loss": 0.8707695007324219,
"step": 986
},
{
"epoch": 1.4423357664233576,
"grad_norm": 1.8046875,
"learning_rate": 1.2012211532304649e-05,
"loss": 0.9090181589126587,
"step": 988
},
{
"epoch": 1.4452554744525548,
"grad_norm": 0.71875,
"learning_rate": 1.198401292790874e-05,
"loss": 0.6352782249450684,
"step": 990
},
{
"epoch": 1.448175182481752,
"grad_norm": 4.0625,
"learning_rate": 1.1955804543361296e-05,
"loss": 0.9220297336578369,
"step": 992
},
{
"epoch": 1.4510948905109489,
"grad_norm": 2.84375,
"learning_rate": 1.1927586659026798e-05,
"loss": 0.8673491477966309,
"step": 994
},
{
"epoch": 1.454014598540146,
"grad_norm": 3.59375,
"learning_rate": 1.1899359555364151e-05,
"loss": 0.9454919099807739,
"step": 996
},
{
"epoch": 1.4569343065693432,
"grad_norm": 1.25,
"learning_rate": 1.1871123512923888e-05,
"loss": 0.9542334675788879,
"step": 998
},
{
"epoch": 1.4598540145985401,
"grad_norm": 1.7734375,
"learning_rate": 1.1842878812345384e-05,
"loss": 0.9865251779556274,
"step": 1000
},
{
"epoch": 1.4627737226277373,
"grad_norm": 3.953125,
"learning_rate": 1.1814625734354066e-05,
"loss": 0.9096630811691284,
"step": 1002
},
{
"epoch": 1.4656934306569342,
"grad_norm": 5.125,
"learning_rate": 1.1786364559758633e-05,
"loss": 0.9506045579910278,
"step": 1004
},
{
"epoch": 1.4686131386861314,
"grad_norm": 2.328125,
"learning_rate": 1.1758095569448241e-05,
"loss": 0.9175775051116943,
"step": 1006
},
{
"epoch": 1.4715328467153284,
"grad_norm": 2.328125,
"learning_rate": 1.1729819044389744e-05,
"loss": 0.9773789644241333,
"step": 1008
},
{
"epoch": 1.4744525547445255,
"grad_norm": 1.359375,
"learning_rate": 1.1701535265624872e-05,
"loss": 1.0058449506759644,
"step": 1010
},
{
"epoch": 1.4773722627737227,
"grad_norm": 2.84375,
"learning_rate": 1.1673244514267457e-05,
"loss": 0.9778100252151489,
"step": 1012
},
{
"epoch": 1.4802919708029196,
"grad_norm": 3.46875,
"learning_rate": 1.1644947071500629e-05,
"loss": 0.9185872077941895,
"step": 1014
},
{
"epoch": 1.4832116788321168,
"grad_norm": 1.6640625,
"learning_rate": 1.1616643218574022e-05,
"loss": 0.9796271324157715,
"step": 1016
},
{
"epoch": 1.486131386861314,
"grad_norm": 1.5390625,
"learning_rate": 1.1588333236800989e-05,
"loss": 0.8974401950836182,
"step": 1018
},
{
"epoch": 1.489051094890511,
"grad_norm": 0.703125,
"learning_rate": 1.1560017407555784e-05,
"loss": 0.9837539196014404,
"step": 1020
},
{
"epoch": 1.491970802919708,
"grad_norm": 1.765625,
"learning_rate": 1.1531696012270794e-05,
"loss": 0.9453492760658264,
"step": 1022
},
{
"epoch": 1.4948905109489052,
"grad_norm": 2.15625,
"learning_rate": 1.1503369332433719e-05,
"loss": 1.1556532382965088,
"step": 1024
},
{
"epoch": 1.4978102189781022,
"grad_norm": 1.1484375,
"learning_rate": 1.1475037649584784e-05,
"loss": 0.6949787139892578,
"step": 1026
},
{
"epoch": 1.5007299270072991,
"grad_norm": 1.3984375,
"learning_rate": 1.1446701245313938e-05,
"loss": 0.736177921295166,
"step": 1028
},
{
"epoch": 1.5036496350364965,
"grad_norm": 1.4453125,
"learning_rate": 1.1418360401258062e-05,
"loss": 0.8855810165405273,
"step": 1030
},
{
"epoch": 1.5065693430656935,
"grad_norm": 3.0625,
"learning_rate": 1.1390015399098154e-05,
"loss": 1.0128211975097656,
"step": 1032
},
{
"epoch": 1.5094890510948904,
"grad_norm": 1.4609375,
"learning_rate": 1.136166652055655e-05,
"loss": 0.882917582988739,
"step": 1034
},
{
"epoch": 1.5124087591240876,
"grad_norm": 3.015625,
"learning_rate": 1.1333314047394104e-05,
"loss": 0.9593569040298462,
"step": 1036
},
{
"epoch": 1.5153284671532847,
"grad_norm": 2.09375,
"learning_rate": 1.1304958261407411e-05,
"loss": 0.8776706457138062,
"step": 1038
},
{
"epoch": 1.5182481751824817,
"grad_norm": 1.390625,
"learning_rate": 1.1276599444425974e-05,
"loss": 0.8630977272987366,
"step": 1040
},
{
"epoch": 1.5211678832116788,
"grad_norm": 0.59765625,
"learning_rate": 1.124823787830944e-05,
"loss": 1.0511329174041748,
"step": 1042
},
{
"epoch": 1.524087591240876,
"grad_norm": 3.421875,
"learning_rate": 1.1219873844944762e-05,
"loss": 1.0726768970489502,
"step": 1044
},
{
"epoch": 1.527007299270073,
"grad_norm": 2.859375,
"learning_rate": 1.1191507626243427e-05,
"loss": 0.9741050004959106,
"step": 1046
},
{
"epoch": 1.5299270072992701,
"grad_norm": 1.25,
"learning_rate": 1.1163139504138643e-05,
"loss": 0.8808518052101135,
"step": 1048
},
{
"epoch": 1.5328467153284673,
"grad_norm": 2.9375,
"learning_rate": 1.1134769760582524e-05,
"loss": 0.7224781513214111,
"step": 1050
},
{
"epoch": 1.5357664233576642,
"grad_norm": 3.421875,
"learning_rate": 1.1106398677543315e-05,
"loss": 0.691923201084137,
"step": 1052
},
{
"epoch": 1.5386861313868612,
"grad_norm": 0.671875,
"learning_rate": 1.1078026537002562e-05,
"loss": 0.929024338722229,
"step": 1054
},
{
"epoch": 1.5416058394160586,
"grad_norm": 1.234375,
"learning_rate": 1.1049653620952327e-05,
"loss": 1.0192841291427612,
"step": 1056
},
{
"epoch": 1.5445255474452555,
"grad_norm": 2.984375,
"learning_rate": 1.1021280211392377e-05,
"loss": 0.5925525426864624,
"step": 1058
},
{
"epoch": 1.5474452554744524,
"grad_norm": 5.1875,
"learning_rate": 1.0992906590327394e-05,
"loss": 0.7697658538818359,
"step": 1060
},
{
"epoch": 1.5503649635036496,
"grad_norm": 2.046875,
"learning_rate": 1.0964533039764146e-05,
"loss": 1.038510799407959,
"step": 1062
},
{
"epoch": 1.5532846715328468,
"grad_norm": 1.453125,
"learning_rate": 1.0936159841708712e-05,
"loss": 1.0497779846191406,
"step": 1064
},
{
"epoch": 1.5562043795620437,
"grad_norm": 4.40625,
"learning_rate": 1.0907787278163663e-05,
"loss": 0.9128830432891846,
"step": 1066
},
{
"epoch": 1.5591240875912409,
"grad_norm": 3.75,
"learning_rate": 1.0879415631125266e-05,
"loss": 0.8196215629577637,
"step": 1068
},
{
"epoch": 1.562043795620438,
"grad_norm": 1.515625,
"learning_rate": 1.0851045182580676e-05,
"loss": 0.9469729661941528,
"step": 1070
},
{
"epoch": 1.564963503649635,
"grad_norm": 8.0,
"learning_rate": 1.0822676214505134e-05,
"loss": 1.0648167133331299,
"step": 1072
},
{
"epoch": 1.5678832116788322,
"grad_norm": 1.2265625,
"learning_rate": 1.0794309008859176e-05,
"loss": 0.8813024759292603,
"step": 1074
},
{
"epoch": 1.5708029197080293,
"grad_norm": 5.96875,
"learning_rate": 1.0765943847585813e-05,
"loss": 0.5665163397789001,
"step": 1076
},
{
"epoch": 1.5737226277372263,
"grad_norm": 0.72265625,
"learning_rate": 1.0737581012607736e-05,
"loss": 0.9029617309570312,
"step": 1078
},
{
"epoch": 1.5766423357664232,
"grad_norm": 3.0,
"learning_rate": 1.0709220785824524e-05,
"loss": 1.0868520736694336,
"step": 1080
},
{
"epoch": 1.5795620437956206,
"grad_norm": 3.546875,
"learning_rate": 1.068086344910982e-05,
"loss": 0.7701568603515625,
"step": 1082
},
{
"epoch": 1.5824817518248175,
"grad_norm": 1.7421875,
"learning_rate": 1.0652509284308557e-05,
"loss": 1.0572642087936401,
"step": 1084
},
{
"epoch": 1.5854014598540145,
"grad_norm": 2.109375,
"learning_rate": 1.0624158573234125e-05,
"loss": 0.7475051879882812,
"step": 1086
},
{
"epoch": 1.5883211678832116,
"grad_norm": 2.265625,
"learning_rate": 1.0595811597665603e-05,
"loss": 1.0147130489349365,
"step": 1088
},
{
"epoch": 1.5912408759124088,
"grad_norm": 1.953125,
"learning_rate": 1.0567468639344938e-05,
"loss": 0.9254045486450195,
"step": 1090
},
{
"epoch": 1.5941605839416058,
"grad_norm": 0.64453125,
"learning_rate": 1.0539129979974143e-05,
"loss": 0.7282826900482178,
"step": 1092
},
{
"epoch": 1.597080291970803,
"grad_norm": 1.1171875,
"learning_rate": 1.051079590121251e-05,
"loss": 0.8476436138153076,
"step": 1094
},
{
"epoch": 1.6,
"grad_norm": 1.203125,
"learning_rate": 1.0482466684673804e-05,
"loss": 0.8929286003112793,
"step": 1096
},
{
"epoch": 1.602919708029197,
"grad_norm": 1.140625,
"learning_rate": 1.0454142611923459e-05,
"loss": 0.8924300670623779,
"step": 1098
},
{
"epoch": 1.6058394160583942,
"grad_norm": 1.328125,
"learning_rate": 1.0425823964475792e-05,
"loss": 1.0469520092010498,
"step": 1100
},
{
"epoch": 1.6087591240875914,
"grad_norm": 3.125,
"learning_rate": 1.0397511023791192e-05,
"loss": 0.8471413254737854,
"step": 1102
},
{
"epoch": 1.6116788321167883,
"grad_norm": 0.87109375,
"learning_rate": 1.0369204071273328e-05,
"loss": 0.48374485969543457,
"step": 1104
},
{
"epoch": 1.6145985401459853,
"grad_norm": 1.6171875,
"learning_rate": 1.0340903388266362e-05,
"loss": 1.0457675457000732,
"step": 1106
},
{
"epoch": 1.6175182481751826,
"grad_norm": 2.578125,
"learning_rate": 1.0312609256052125e-05,
"loss": 0.8381612300872803,
"step": 1108
},
{
"epoch": 1.6204379562043796,
"grad_norm": 1.09375,
"learning_rate": 1.0284321955847357e-05,
"loss": 0.7611550092697144,
"step": 1110
},
{
"epoch": 1.6233576642335765,
"grad_norm": 1.375,
"learning_rate": 1.0256041768800886e-05,
"loss": 0.9808844327926636,
"step": 1112
},
{
"epoch": 1.6262773722627737,
"grad_norm": 1.1328125,
"learning_rate": 1.0227768975990837e-05,
"loss": 1.1369640827178955,
"step": 1114
},
{
"epoch": 1.6291970802919709,
"grad_norm": 1.828125,
"learning_rate": 1.019950385842186e-05,
"loss": 0.9222966432571411,
"step": 1116
},
{
"epoch": 1.6321167883211678,
"grad_norm": 3.984375,
"learning_rate": 1.0171246697022302e-05,
"loss": 0.8137625455856323,
"step": 1118
},
{
"epoch": 1.635036496350365,
"grad_norm": 1.203125,
"learning_rate": 1.0142997772641443e-05,
"loss": 0.8257233500480652,
"step": 1120
},
{
"epoch": 1.6379562043795621,
"grad_norm": 1.0625,
"learning_rate": 1.0114757366046692e-05,
"loss": 1.0450224876403809,
"step": 1122
},
{
"epoch": 1.640875912408759,
"grad_norm": 1.34375,
"learning_rate": 1.0086525757920803e-05,
"loss": 0.41819870471954346,
"step": 1124
},
{
"epoch": 1.6437956204379562,
"grad_norm": 1.1015625,
"learning_rate": 1.0058303228859076e-05,
"loss": 0.9673611521720886,
"step": 1126
},
{
"epoch": 1.6467153284671534,
"grad_norm": 2.703125,
"learning_rate": 1.003009005936658e-05,
"loss": 0.9797380566596985,
"step": 1128
},
{
"epoch": 1.6496350364963503,
"grad_norm": 5.90625,
"learning_rate": 1.0001886529855351e-05,
"loss": 0.9807537794113159,
"step": 1130
},
{
"epoch": 1.6525547445255473,
"grad_norm": 1.296875,
"learning_rate": 9.97369292064162e-06,
"loss": 0.8815328478813171,
"step": 1132
},
{
"epoch": 1.6554744525547447,
"grad_norm": 3.984375,
"learning_rate": 9.945509511943016e-06,
"loss": 0.9661548733711243,
"step": 1134
},
{
"epoch": 1.6583941605839416,
"grad_norm": 8.125,
"learning_rate": 9.917336583875788e-06,
"loss": 0.877112627029419,
"step": 1136
},
{
"epoch": 1.6613138686131386,
"grad_norm": 1.15625,
"learning_rate": 9.889174416452011e-06,
"loss": 0.949897289276123,
"step": 1138
},
{
"epoch": 1.6642335766423357,
"grad_norm": 1.421875,
"learning_rate": 9.861023289576816e-06,
"loss": 0.7110643982887268,
"step": 1140
},
{
"epoch": 1.667153284671533,
"grad_norm": 6.09375,
"learning_rate": 9.832883483045597e-06,
"loss": 1.0274568796157837,
"step": 1142
},
{
"epoch": 1.6700729927007298,
"grad_norm": 1.234375,
"learning_rate": 9.804755276541238e-06,
"loss": 0.5871222019195557,
"step": 1144
},
{
"epoch": 1.672992700729927,
"grad_norm": 1.0625,
"learning_rate": 9.776638949631325e-06,
"loss": 0.89629727602005,
"step": 1146
},
{
"epoch": 1.6759124087591242,
"grad_norm": 2.0,
"learning_rate": 9.74853478176538e-06,
"loss": 0.5665899515151978,
"step": 1148
},
{
"epoch": 1.6788321167883211,
"grad_norm": 1.65625,
"learning_rate": 9.72044305227207e-06,
"loss": 0.9088619947433472,
"step": 1150
},
{
"epoch": 1.6817518248175183,
"grad_norm": 3.765625,
"learning_rate": 9.692364040356435e-06,
"loss": 0.9651062488555908,
"step": 1152
},
{
"epoch": 1.6846715328467154,
"grad_norm": 2.46875,
"learning_rate": 9.664298025097118e-06,
"loss": 0.9715033769607544,
"step": 1154
},
{
"epoch": 1.6875912408759124,
"grad_norm": 0.9296875,
"learning_rate": 9.636245285443587e-06,
"loss": 0.7159008383750916,
"step": 1156
},
{
"epoch": 1.6905109489051093,
"grad_norm": 1.0078125,
"learning_rate": 9.608206100213364e-06,
"loss": 1.0738484859466553,
"step": 1158
},
{
"epoch": 1.6934306569343067,
"grad_norm": 0.95703125,
"learning_rate": 9.580180748089245e-06,
"loss": 1.040489673614502,
"step": 1160
},
{
"epoch": 1.6963503649635037,
"grad_norm": 0.76953125,
"learning_rate": 9.552169507616554e-06,
"loss": 0.9367128610610962,
"step": 1162
},
{
"epoch": 1.6992700729927006,
"grad_norm": 2.484375,
"learning_rate": 9.524172657200342e-06,
"loss": 1.1005165576934814,
"step": 1164
},
{
"epoch": 1.7021897810218978,
"grad_norm": 2.46875,
"learning_rate": 9.496190475102649e-06,
"loss": 0.5636100769042969,
"step": 1166
},
{
"epoch": 1.705109489051095,
"grad_norm": 1.2734375,
"learning_rate": 9.468223239439714e-06,
"loss": 0.7754777073860168,
"step": 1168
},
{
"epoch": 1.7080291970802919,
"grad_norm": 3.09375,
"learning_rate": 9.440271228179237e-06,
"loss": 0.875159740447998,
"step": 1170
},
{
"epoch": 1.710948905109489,
"grad_norm": 1.109375,
"learning_rate": 9.412334719137587e-06,
"loss": 0.9143173694610596,
"step": 1172
},
{
"epoch": 1.7138686131386862,
"grad_norm": 2.609375,
"learning_rate": 9.384413989977069e-06,
"loss": 0.7207679748535156,
"step": 1174
},
{
"epoch": 1.7167883211678832,
"grad_norm": 1.5859375,
"learning_rate": 9.35650931820314e-06,
"loss": 0.9489813446998596,
"step": 1176
},
{
"epoch": 1.7197080291970803,
"grad_norm": 2.984375,
"learning_rate": 9.328620981161674e-06,
"loss": 0.9765996336936951,
"step": 1178
},
{
"epoch": 1.7226277372262775,
"grad_norm": 2.109375,
"learning_rate": 9.300749256036174e-06,
"loss": 0.8786243796348572,
"step": 1180
},
{
"epoch": 1.7255474452554744,
"grad_norm": 1.1015625,
"learning_rate": 9.272894419845063e-06,
"loss": 0.766198992729187,
"step": 1182
},
{
"epoch": 1.7284671532846714,
"grad_norm": 1.2265625,
"learning_rate": 9.245056749438877e-06,
"loss": 0.969190239906311,
"step": 1184
},
{
"epoch": 1.7313868613138688,
"grad_norm": 3.671875,
"learning_rate": 9.217236521497563e-06,
"loss": 0.9522278308868408,
"step": 1186
},
{
"epoch": 1.7343065693430657,
"grad_norm": 3.046875,
"learning_rate": 9.189434012527685e-06,
"loss": 1.0335571765899658,
"step": 1188
},
{
"epoch": 1.7372262773722627,
"grad_norm": 4.1875,
"learning_rate": 9.161649498859723e-06,
"loss": 0.9262440204620361,
"step": 1190
},
{
"epoch": 1.7401459854014598,
"grad_norm": 1.8515625,
"learning_rate": 9.133883256645273e-06,
"loss": 0.8710659742355347,
"step": 1192
},
{
"epoch": 1.743065693430657,
"grad_norm": 1.109375,
"learning_rate": 9.106135561854357e-06,
"loss": 1.0118987560272217,
"step": 1194
},
{
"epoch": 1.745985401459854,
"grad_norm": 5.75,
"learning_rate": 9.078406690272628e-06,
"loss": 0.4204956591129303,
"step": 1196
},
{
"epoch": 1.748905109489051,
"grad_norm": 0.6953125,
"learning_rate": 9.050696917498676e-06,
"loss": 0.7705672979354858,
"step": 1198
},
{
"epoch": 1.7518248175182483,
"grad_norm": 3.8125,
"learning_rate": 9.023006518941247e-06,
"loss": 0.7334332466125488,
"step": 1200
},
{
"epoch": 1.7547445255474452,
"grad_norm": 1.8828125,
"learning_rate": 8.99533576981655e-06,
"loss": 0.9681108593940735,
"step": 1202
},
{
"epoch": 1.7576642335766424,
"grad_norm": 3.5,
"learning_rate": 8.967684945145467e-06,
"loss": 0.6784968376159668,
"step": 1204
},
{
"epoch": 1.7605839416058395,
"grad_norm": 1.5,
"learning_rate": 8.940054319750882e-06,
"loss": 0.5141807794570923,
"step": 1206
},
{
"epoch": 1.7635036496350365,
"grad_norm": 2.609375,
"learning_rate": 8.91244416825489e-06,
"loss": 0.9204733371734619,
"step": 1208
},
{
"epoch": 1.7664233576642334,
"grad_norm": 1.9140625,
"learning_rate": 8.884854765076116e-06,
"loss": 1.1004142761230469,
"step": 1210
},
{
"epoch": 1.7693430656934308,
"grad_norm": 1.375,
"learning_rate": 8.857286384426947e-06,
"loss": 1.0370523929595947,
"step": 1212
},
{
"epoch": 1.7722627737226277,
"grad_norm": 1.1953125,
"learning_rate": 8.829739300310847e-06,
"loss": 0.9668702483177185,
"step": 1214
},
{
"epoch": 1.7751824817518247,
"grad_norm": 2.625,
"learning_rate": 8.802213786519593e-06,
"loss": 0.880242109298706,
"step": 1216
},
{
"epoch": 1.7781021897810219,
"grad_norm": 1.2421875,
"learning_rate": 8.774710116630584e-06,
"loss": 0.7876850366592407,
"step": 1218
},
{
"epoch": 1.781021897810219,
"grad_norm": 2.09375,
"learning_rate": 8.747228564004118e-06,
"loss": 1.0225201845169067,
"step": 1220
},
{
"epoch": 1.783941605839416,
"grad_norm": 3.734375,
"learning_rate": 8.71976940178065e-06,
"loss": 0.9184777736663818,
"step": 1222
},
{
"epoch": 1.7868613138686131,
"grad_norm": 1.8203125,
"learning_rate": 8.692332902878116e-06,
"loss": 0.9042671918869019,
"step": 1224
},
{
"epoch": 1.7897810218978103,
"grad_norm": 4.53125,
"learning_rate": 8.664919339989177e-06,
"loss": 0.938407838344574,
"step": 1226
},
{
"epoch": 1.7927007299270072,
"grad_norm": 2.25,
"learning_rate": 8.637528985578562e-06,
"loss": 1.018449068069458,
"step": 1228
},
{
"epoch": 1.7956204379562044,
"grad_norm": 3.203125,
"learning_rate": 8.610162111880296e-06,
"loss": 0.9853857755661011,
"step": 1230
},
{
"epoch": 1.7985401459854016,
"grad_norm": 6.875,
"learning_rate": 8.58281899089506e-06,
"loss": 0.8306851387023926,
"step": 1232
},
{
"epoch": 1.8014598540145985,
"grad_norm": 1.3671875,
"learning_rate": 8.555499894387427e-06,
"loss": 1.0020043849945068,
"step": 1234
},
{
"epoch": 1.8043795620437955,
"grad_norm": 2.1875,
"learning_rate": 8.528205093883217e-06,
"loss": 1.1009316444396973,
"step": 1236
},
{
"epoch": 1.8072992700729928,
"grad_norm": 1.8046875,
"learning_rate": 8.500934860666744e-06,
"loss": 0.7341406345367432,
"step": 1238
},
{
"epoch": 1.8102189781021898,
"grad_norm": 1.2890625,
"learning_rate": 8.473689465778171e-06,
"loss": 0.997475802898407,
"step": 1240
},
{
"epoch": 1.8131386861313867,
"grad_norm": 1.703125,
"learning_rate": 8.446469180010773e-06,
"loss": 1.0310256481170654,
"step": 1242
},
{
"epoch": 1.816058394160584,
"grad_norm": 1.75,
"learning_rate": 8.419274273908273e-06,
"loss": 0.9313318133354187,
"step": 1244
},
{
"epoch": 1.818978102189781,
"grad_norm": 2.5625,
"learning_rate": 8.392105017762143e-06,
"loss": 1.0743803977966309,
"step": 1246
},
{
"epoch": 1.821897810218978,
"grad_norm": 1.5234375,
"learning_rate": 8.364961681608922e-06,
"loss": 1.0000810623168945,
"step": 1248
},
{
"epoch": 1.8248175182481752,
"grad_norm": 1.4765625,
"learning_rate": 8.33784453522752e-06,
"loss": 0.9280273914337158,
"step": 1250
},
{
"epoch": 1.8277372262773723,
"grad_norm": 2.421875,
"learning_rate": 8.310753848136557e-06,
"loss": 0.9379640817642212,
"step": 1252
},
{
"epoch": 1.8306569343065693,
"grad_norm": 1.3203125,
"learning_rate": 8.283689889591662e-06,
"loss": 0.9863895177841187,
"step": 1254
},
{
"epoch": 1.8335766423357664,
"grad_norm": 2.171875,
"learning_rate": 8.256652928582817e-06,
"loss": 0.9451329112052917,
"step": 1256
},
{
"epoch": 1.8364963503649636,
"grad_norm": 2.734375,
"learning_rate": 8.229643233831667e-06,
"loss": 0.9604065418243408,
"step": 1258
},
{
"epoch": 1.8394160583941606,
"grad_norm": 2.421875,
"learning_rate": 8.202661073788864e-06,
"loss": 0.8217023015022278,
"step": 1260
},
{
"epoch": 1.8423357664233575,
"grad_norm": 4.15625,
"learning_rate": 8.175706716631378e-06,
"loss": 1.1534137725830078,
"step": 1262
},
{
"epoch": 1.845255474452555,
"grad_norm": 3.140625,
"learning_rate": 8.148780430259863e-06,
"loss": 1.0975501537322998,
"step": 1264
},
{
"epoch": 1.8481751824817518,
"grad_norm": 1.328125,
"learning_rate": 8.12188248229596e-06,
"loss": 0.8799439668655396,
"step": 1266
},
{
"epoch": 1.8510948905109488,
"grad_norm": 1.6015625,
"learning_rate": 8.095013140079666e-06,
"loss": 0.8988907337188721,
"step": 1268
},
{
"epoch": 1.854014598540146,
"grad_norm": 1.5546875,
"learning_rate": 8.068172670666649e-06,
"loss": 0.9588282108306885,
"step": 1270
},
{
"epoch": 1.856934306569343,
"grad_norm": 1.4140625,
"learning_rate": 8.041361340825626e-06,
"loss": 0.9363412857055664,
"step": 1272
},
{
"epoch": 1.85985401459854,
"grad_norm": 4.125,
"learning_rate": 8.014579417035681e-06,
"loss": 0.9633190631866455,
"step": 1274
},
{
"epoch": 1.8627737226277372,
"grad_norm": 1.2265625,
"learning_rate": 7.98782716548364e-06,
"loss": 0.9566249251365662,
"step": 1276
},
{
"epoch": 1.8656934306569344,
"grad_norm": 1.15625,
"learning_rate": 7.961104852061402e-06,
"loss": 0.9364519119262695,
"step": 1278
},
{
"epoch": 1.8686131386861313,
"grad_norm": 2.484375,
"learning_rate": 7.934412742363327e-06,
"loss": 0.9307054281234741,
"step": 1280
},
{
"epoch": 1.8715328467153285,
"grad_norm": 1.1015625,
"learning_rate": 7.907751101683561e-06,
"loss": 0.9123919010162354,
"step": 1282
},
{
"epoch": 1.8744525547445257,
"grad_norm": 1.46875,
"learning_rate": 7.881120195013432e-06,
"loss": 0.7690815925598145,
"step": 1284
},
{
"epoch": 1.8773722627737226,
"grad_norm": 1.28125,
"learning_rate": 7.85452028703879e-06,
"loss": 0.9250288009643555,
"step": 1286
},
{
"epoch": 1.8802919708029195,
"grad_norm": 1.125,
"learning_rate": 7.827951642137393e-06,
"loss": 0.9514033794403076,
"step": 1288
},
{
"epoch": 1.883211678832117,
"grad_norm": 1.890625,
"learning_rate": 7.801414524376276e-06,
"loss": 1.0991572141647339,
"step": 1290
},
{
"epoch": 1.8861313868613139,
"grad_norm": 2.703125,
"learning_rate": 7.774909197509117e-06,
"loss": 0.7432125806808472,
"step": 1292
},
{
"epoch": 1.8890510948905108,
"grad_norm": 1.3125,
"learning_rate": 7.74843592497363e-06,
"loss": 0.6110070943832397,
"step": 1294
},
{
"epoch": 1.891970802919708,
"grad_norm": 4.1875,
"learning_rate": 7.721994969888934e-06,
"loss": 0.8845908641815186,
"step": 1296
},
{
"epoch": 1.8948905109489051,
"grad_norm": 5.21875,
"learning_rate": 7.695586595052947e-06,
"loss": 0.9912600517272949,
"step": 1298
},
{
"epoch": 1.897810218978102,
"grad_norm": 3.78125,
"learning_rate": 7.669211062939769e-06,
"loss": 0.9381648302078247,
"step": 1300
},
{
"epoch": 1.9007299270072993,
"grad_norm": 1.53125,
"learning_rate": 7.642868635697075e-06,
"loss": 0.9803470373153687,
"step": 1302
},
{
"epoch": 1.9036496350364964,
"grad_norm": 0.953125,
"learning_rate": 7.6165595751435065e-06,
"loss": 0.27998918294906616,
"step": 1304
},
{
"epoch": 1.9065693430656934,
"grad_norm": 3.375,
"learning_rate": 7.590284142766076e-06,
"loss": 0.899682879447937,
"step": 1306
},
{
"epoch": 1.9094890510948905,
"grad_norm": 6.0,
"learning_rate": 7.564042599717561e-06,
"loss": 0.9667043089866638,
"step": 1308
},
{
"epoch": 1.9124087591240877,
"grad_norm": 1.1953125,
"learning_rate": 7.5378352068139174e-06,
"loss": 0.9081999063491821,
"step": 1310
},
{
"epoch": 1.9153284671532846,
"grad_norm": 2.28125,
"learning_rate": 7.511662224531668e-06,
"loss": 1.0329896211624146,
"step": 1312
},
{
"epoch": 1.9182481751824818,
"grad_norm": 1.34375,
"learning_rate": 7.485523913005339e-06,
"loss": 0.8019554615020752,
"step": 1314
},
{
"epoch": 1.921167883211679,
"grad_norm": 1.84375,
"learning_rate": 7.45942053202486e-06,
"loss": 0.9977414011955261,
"step": 1316
},
{
"epoch": 1.924087591240876,
"grad_norm": 1.5546875,
"learning_rate": 7.433352341032981e-06,
"loss": 0.8954068422317505,
"step": 1318
},
{
"epoch": 1.9270072992700729,
"grad_norm": 1.7421875,
"learning_rate": 7.407319599122695e-06,
"loss": 0.8836451768875122,
"step": 1320
},
{
"epoch": 1.92992700729927,
"grad_norm": 2.8125,
"learning_rate": 7.381322565034675e-06,
"loss": 0.9566698670387268,
"step": 1322
},
{
"epoch": 1.9328467153284672,
"grad_norm": 0.58203125,
"learning_rate": 7.35536149715468e-06,
"loss": 0.738754391670227,
"step": 1324
},
{
"epoch": 1.9357664233576641,
"grad_norm": 1.8203125,
"learning_rate": 7.32943665351101e-06,
"loss": 0.910930871963501,
"step": 1326
},
{
"epoch": 1.9386861313868613,
"grad_norm": 2.625,
"learning_rate": 7.30354829177192e-06,
"loss": 0.9571232795715332,
"step": 1328
},
{
"epoch": 1.9416058394160585,
"grad_norm": 1.40625,
"learning_rate": 7.277696669243081e-06,
"loss": 0.9391112923622131,
"step": 1330
},
{
"epoch": 1.9445255474452554,
"grad_norm": 1.875,
"learning_rate": 7.251882042864998e-06,
"loss": 0.8907726407051086,
"step": 1332
},
{
"epoch": 1.9474452554744526,
"grad_norm": 4.0,
"learning_rate": 7.226104669210483e-06,
"loss": 0.4091779589653015,
"step": 1334
},
{
"epoch": 1.9503649635036497,
"grad_norm": 1.46875,
"learning_rate": 7.200364804482083e-06,
"loss": 1.086578130722046,
"step": 1336
},
{
"epoch": 1.9532846715328467,
"grad_norm": 5.09375,
"learning_rate": 7.1746627045095415e-06,
"loss": 1.0218377113342285,
"step": 1338
},
{
"epoch": 1.9562043795620438,
"grad_norm": 0.87109375,
"learning_rate": 7.14899862474726e-06,
"loss": 0.9351223111152649,
"step": 1340
},
{
"epoch": 1.959124087591241,
"grad_norm": 1.59375,
"learning_rate": 7.123372820271758e-06,
"loss": 1.1845860481262207,
"step": 1342
},
{
"epoch": 1.962043795620438,
"grad_norm": 2.953125,
"learning_rate": 7.097785545779129e-06,
"loss": 0.6177266836166382,
"step": 1344
},
{
"epoch": 1.964963503649635,
"grad_norm": 1.7734375,
"learning_rate": 7.072237055582516e-06,
"loss": 0.911766529083252,
"step": 1346
},
{
"epoch": 1.967883211678832,
"grad_norm": 2.375,
"learning_rate": 7.0467276036095834e-06,
"loss": 0.9010423421859741,
"step": 1348
},
{
"epoch": 1.9708029197080292,
"grad_norm": 2.125,
"learning_rate": 7.021257443399998e-06,
"loss": 0.9431062936782837,
"step": 1350
},
{
"epoch": 1.9737226277372262,
"grad_norm": 1.3046875,
"learning_rate": 6.995826828102899e-06,
"loss": 0.9261475801467896,
"step": 1352
},
{
"epoch": 1.9766423357664233,
"grad_norm": 1.53125,
"learning_rate": 6.970436010474386e-06,
"loss": 0.9187385439872742,
"step": 1354
},
{
"epoch": 1.9795620437956205,
"grad_norm": 1.140625,
"learning_rate": 6.945085242875007e-06,
"loss": 1.0751965045928955,
"step": 1356
},
{
"epoch": 1.9824817518248175,
"grad_norm": 0.734375,
"learning_rate": 6.919774777267259e-06,
"loss": 1.0283749103546143,
"step": 1358
},
{
"epoch": 1.9854014598540146,
"grad_norm": 2.71875,
"learning_rate": 6.894504865213065e-06,
"loss": 0.5450443625450134,
"step": 1360
},
{
"epoch": 1.9883211678832118,
"grad_norm": 0.859375,
"learning_rate": 6.869275757871289e-06,
"loss": 0.5233281850814819,
"step": 1362
},
{
"epoch": 1.9912408759124087,
"grad_norm": 1.8359375,
"learning_rate": 6.844087705995238e-06,
"loss": 1.0329816341400146,
"step": 1364
},
{
"epoch": 1.994160583941606,
"grad_norm": 1.6328125,
"learning_rate": 6.818940959930165e-06,
"loss": 0.9405212998390198,
"step": 1366
},
{
"epoch": 1.997080291970803,
"grad_norm": 5.375,
"learning_rate": 6.793835769610778e-06,
"loss": 1.1037087440490723,
"step": 1368
},
{
"epoch": 2.0,
"grad_norm": 2.21875,
"learning_rate": 6.768772384558761e-06,
"loss": 1.0288667678833008,
"step": 1370
},
{
"epoch": 2.002919708029197,
"grad_norm": 0.921875,
"learning_rate": 6.743751053880302e-06,
"loss": 1.0435996055603027,
"step": 1372
},
{
"epoch": 2.0058394160583943,
"grad_norm": 4.75,
"learning_rate": 6.718772026263602e-06,
"loss": 0.8969197273254395,
"step": 1374
},
{
"epoch": 2.0087591240875913,
"grad_norm": 1.3125,
"learning_rate": 6.69383554997641e-06,
"loss": 0.8929958343505859,
"step": 1376
},
{
"epoch": 2.011678832116788,
"grad_norm": 1.15625,
"learning_rate": 6.66894187286355e-06,
"loss": 0.7274095416069031,
"step": 1378
},
{
"epoch": 2.0145985401459856,
"grad_norm": 1.171875,
"learning_rate": 6.644091242344478e-06,
"loss": 0.7429904341697693,
"step": 1380
},
{
"epoch": 2.0175182481751825,
"grad_norm": 2.203125,
"learning_rate": 6.619283905410796e-06,
"loss": 0.9532324075698853,
"step": 1382
},
{
"epoch": 2.0204379562043795,
"grad_norm": 0.76171875,
"learning_rate": 6.594520108623811e-06,
"loss": 0.9710752964019775,
"step": 1384
},
{
"epoch": 2.0233576642335764,
"grad_norm": 2.890625,
"learning_rate": 6.569800098112079e-06,
"loss": 0.8583542108535767,
"step": 1386
},
{
"epoch": 2.026277372262774,
"grad_norm": 1.34375,
"learning_rate": 6.545124119568974e-06,
"loss": 0.9491511583328247,
"step": 1388
},
{
"epoch": 2.0291970802919708,
"grad_norm": 1.3203125,
"learning_rate": 6.520492418250223e-06,
"loss": 0.9267318248748779,
"step": 1390
},
{
"epoch": 2.0321167883211677,
"grad_norm": 1.203125,
"learning_rate": 6.495905238971481e-06,
"loss": 1.0519380569458008,
"step": 1392
},
{
"epoch": 2.035036496350365,
"grad_norm": 10.8125,
"learning_rate": 6.471362826105898e-06,
"loss": 0.9691998362541199,
"step": 1394
},
{
"epoch": 2.037956204379562,
"grad_norm": 6.4375,
"learning_rate": 6.4468654235816945e-06,
"loss": 0.7549631595611572,
"step": 1396
},
{
"epoch": 2.040875912408759,
"grad_norm": 3.375,
"learning_rate": 6.422413274879719e-06,
"loss": 0.7525750398635864,
"step": 1398
},
{
"epoch": 2.0437956204379564,
"grad_norm": 1.40625,
"learning_rate": 6.3980066230310535e-06,
"loss": 0.8843213319778442,
"step": 1400
},
{
"epoch": 2.0467153284671533,
"grad_norm": 2.09375,
"learning_rate": 6.3736457106145676e-06,
"loss": 0.9530847668647766,
"step": 1402
},
{
"epoch": 2.0496350364963503,
"grad_norm": 0.875,
"learning_rate": 6.349330779754542e-06,
"loss": 0.6189695596694946,
"step": 1404
},
{
"epoch": 2.0525547445255476,
"grad_norm": 1.109375,
"learning_rate": 6.325062072118238e-06,
"loss": 0.4646834135055542,
"step": 1406
},
{
"epoch": 2.0554744525547446,
"grad_norm": 1.546875,
"learning_rate": 6.300839828913495e-06,
"loss": 0.890283465385437,
"step": 1408
},
{
"epoch": 2.0583941605839415,
"grad_norm": 1.2890625,
"learning_rate": 6.276664290886347e-06,
"loss": 1.0025547742843628,
"step": 1410
},
{
"epoch": 2.0613138686131385,
"grad_norm": 4.375,
"learning_rate": 6.252535698318629e-06,
"loss": 0.630736231803894,
"step": 1412
},
{
"epoch": 2.064233576642336,
"grad_norm": 3.203125,
"learning_rate": 6.2284542910255694e-06,
"loss": 0.7758012413978577,
"step": 1414
},
{
"epoch": 2.067153284671533,
"grad_norm": 5.53125,
"learning_rate": 6.204420308353432e-06,
"loss": 0.5808634757995605,
"step": 1416
},
{
"epoch": 2.0700729927007298,
"grad_norm": 3.28125,
"learning_rate": 6.18043398917711e-06,
"loss": 0.6857596635818481,
"step": 1418
},
{
"epoch": 2.072992700729927,
"grad_norm": 1.453125,
"learning_rate": 6.156495571897786e-06,
"loss": 0.5333114862442017,
"step": 1420
},
{
"epoch": 2.075912408759124,
"grad_norm": 0.8828125,
"learning_rate": 6.132605294440532e-06,
"loss": 1.025094747543335,
"step": 1422
},
{
"epoch": 2.078832116788321,
"grad_norm": 1.546875,
"learning_rate": 6.108763394251956e-06,
"loss": 0.7176234722137451,
"step": 1424
},
{
"epoch": 2.0817518248175184,
"grad_norm": 1.8203125,
"learning_rate": 6.084970108297841e-06,
"loss": 0.5737876296043396,
"step": 1426
},
{
"epoch": 2.0846715328467154,
"grad_norm": 1.1328125,
"learning_rate": 6.061225673060799e-06,
"loss": 0.7959654331207275,
"step": 1428
},
{
"epoch": 2.0875912408759123,
"grad_norm": 1.65625,
"learning_rate": 6.037530324537905e-06,
"loss": 0.8338675498962402,
"step": 1430
},
{
"epoch": 2.0905109489051097,
"grad_norm": 3.890625,
"learning_rate": 6.013884298238358e-06,
"loss": 0.8504579067230225,
"step": 1432
},
{
"epoch": 2.0934306569343066,
"grad_norm": 1.5,
"learning_rate": 5.990287829181148e-06,
"loss": 0.8194718360900879,
"step": 1434
},
{
"epoch": 2.0963503649635036,
"grad_norm": 0.97265625,
"learning_rate": 5.96674115189271e-06,
"loss": 0.6566216945648193,
"step": 1436
},
{
"epoch": 2.0992700729927005,
"grad_norm": 1.5625,
"learning_rate": 5.943244500404591e-06,
"loss": 0.8156367540359497,
"step": 1438
},
{
"epoch": 2.102189781021898,
"grad_norm": 1.53125,
"learning_rate": 5.919798108251132e-06,
"loss": 0.7687311768531799,
"step": 1440
},
{
"epoch": 2.105109489051095,
"grad_norm": 1.9609375,
"learning_rate": 5.896402208467153e-06,
"loss": 0.46858537197113037,
"step": 1442
},
{
"epoch": 2.108029197080292,
"grad_norm": 1.78125,
"learning_rate": 5.873057033585615e-06,
"loss": 0.9006612300872803,
"step": 1444
},
{
"epoch": 2.110948905109489,
"grad_norm": 2.015625,
"learning_rate": 5.849762815635328e-06,
"loss": 0.8107215762138367,
"step": 1446
},
{
"epoch": 2.113868613138686,
"grad_norm": 2.015625,
"learning_rate": 5.826519786138636e-06,
"loss": 0.8245384693145752,
"step": 1448
},
{
"epoch": 2.116788321167883,
"grad_norm": 1.1640625,
"learning_rate": 5.8033281761091255e-06,
"loss": 0.8235875368118286,
"step": 1450
},
{
"epoch": 2.1197080291970805,
"grad_norm": 1.171875,
"learning_rate": 5.780188216049315e-06,
"loss": 0.7649439573287964,
"step": 1452
},
{
"epoch": 2.1226277372262774,
"grad_norm": 1.328125,
"learning_rate": 5.7571001359483725e-06,
"loss": 0.6783745288848877,
"step": 1454
},
{
"epoch": 2.1255474452554743,
"grad_norm": 2.296875,
"learning_rate": 5.734064165279828e-06,
"loss": 0.8276048898696899,
"step": 1456
},
{
"epoch": 2.1284671532846717,
"grad_norm": 1.71875,
"learning_rate": 5.7110805329993e-06,
"loss": 0.7815314531326294,
"step": 1458
},
{
"epoch": 2.1313868613138687,
"grad_norm": 0.99609375,
"learning_rate": 5.688149467542208e-06,
"loss": 0.6708962917327881,
"step": 1460
},
{
"epoch": 2.1343065693430656,
"grad_norm": 1.859375,
"learning_rate": 5.665271196821506e-06,
"loss": 0.8067224025726318,
"step": 1462
},
{
"epoch": 2.137226277372263,
"grad_norm": 1.09375,
"learning_rate": 5.642445948225416e-06,
"loss": 0.8288531303405762,
"step": 1464
},
{
"epoch": 2.14014598540146,
"grad_norm": 2.8125,
"learning_rate": 5.619673948615189e-06,
"loss": 0.8746398091316223,
"step": 1466
},
{
"epoch": 2.143065693430657,
"grad_norm": 1.8828125,
"learning_rate": 5.5969554243228e-06,
"loss": 0.6434183120727539,
"step": 1468
},
{
"epoch": 2.145985401459854,
"grad_norm": 1.40625,
"learning_rate": 5.57429060114876e-06,
"loss": 0.6681355237960815,
"step": 1470
},
{
"epoch": 2.1489051094890512,
"grad_norm": 2.40625,
"learning_rate": 5.5516797043598245e-06,
"loss": 0.7707985639572144,
"step": 1472
},
{
"epoch": 2.151824817518248,
"grad_norm": 2.078125,
"learning_rate": 5.529122958686786e-06,
"loss": 0.6538320779800415,
"step": 1474
},
{
"epoch": 2.154744525547445,
"grad_norm": 3.1875,
"learning_rate": 5.506620588322206e-06,
"loss": 0.561845064163208,
"step": 1476
},
{
"epoch": 2.1576642335766425,
"grad_norm": 4.46875,
"learning_rate": 5.484172816918227e-06,
"loss": 0.6652007102966309,
"step": 1478
},
{
"epoch": 2.1605839416058394,
"grad_norm": 1.3515625,
"learning_rate": 5.46177986758432e-06,
"loss": 0.8560731410980225,
"step": 1480
},
{
"epoch": 2.1635036496350364,
"grad_norm": 1.5,
"learning_rate": 5.439441962885084e-06,
"loss": 0.8291876316070557,
"step": 1482
},
{
"epoch": 2.1664233576642338,
"grad_norm": 0.93359375,
"learning_rate": 5.417159324838012e-06,
"loss": 0.8320481777191162,
"step": 1484
},
{
"epoch": 2.1693430656934307,
"grad_norm": 1.96875,
"learning_rate": 5.394932174911316e-06,
"loss": 0.6385836601257324,
"step": 1486
},
{
"epoch": 2.1722627737226277,
"grad_norm": 2.640625,
"learning_rate": 5.372760734021697e-06,
"loss": 0.7288645505905151,
"step": 1488
},
{
"epoch": 2.1751824817518246,
"grad_norm": 1.578125,
"learning_rate": 5.3506452225321756e-06,
"loss": 0.7009255886077881,
"step": 1490
},
{
"epoch": 2.178102189781022,
"grad_norm": 1.90625,
"learning_rate": 5.3285858602498655e-06,
"loss": 0.7125614881515503,
"step": 1492
},
{
"epoch": 2.181021897810219,
"grad_norm": 1.640625,
"learning_rate": 5.30658286642383e-06,
"loss": 0.895104169845581,
"step": 1494
},
{
"epoch": 2.183941605839416,
"grad_norm": 2.328125,
"learning_rate": 5.284636459742878e-06,
"loss": 0.7410358190536499,
"step": 1496
},
{
"epoch": 2.1868613138686133,
"grad_norm": 1.40625,
"learning_rate": 5.262746858333386e-06,
"loss": 0.7033236026763916,
"step": 1498
},
{
"epoch": 2.18978102189781,
"grad_norm": 0.66015625,
"learning_rate": 5.24091427975716e-06,
"loss": 0.8284467458724976,
"step": 1500
},
{
"epoch": 2.192700729927007,
"grad_norm": 3.359375,
"learning_rate": 5.219138941009232e-06,
"loss": 0.6668428182601929,
"step": 1502
},
{
"epoch": 2.1956204379562045,
"grad_norm": 3.484375,
"learning_rate": 5.197421058515747e-06,
"loss": 0.8068520426750183,
"step": 1504
},
{
"epoch": 2.1985401459854015,
"grad_norm": 1.890625,
"learning_rate": 5.175760848131764e-06,
"loss": 0.8340035080909729,
"step": 1506
},
{
"epoch": 2.2014598540145984,
"grad_norm": 4.96875,
"learning_rate": 5.15415852513916e-06,
"loss": 0.7390716075897217,
"step": 1508
},
{
"epoch": 2.204379562043796,
"grad_norm": 2.546875,
"learning_rate": 5.132614304244448e-06,
"loss": 0.5261499881744385,
"step": 1510
},
{
"epoch": 2.2072992700729928,
"grad_norm": 1.6328125,
"learning_rate": 5.111128399576679e-06,
"loss": 0.5483267307281494,
"step": 1512
},
{
"epoch": 2.2102189781021897,
"grad_norm": 1.59375,
"learning_rate": 5.089701024685276e-06,
"loss": 0.6743176579475403,
"step": 1514
},
{
"epoch": 2.213138686131387,
"grad_norm": 4.59375,
"learning_rate": 5.068332392537949e-06,
"loss": 0.5492575168609619,
"step": 1516
},
{
"epoch": 2.216058394160584,
"grad_norm": 4.5,
"learning_rate": 5.047022715518552e-06,
"loss": 0.6473349332809448,
"step": 1518
},
{
"epoch": 2.218978102189781,
"grad_norm": 1.2265625,
"learning_rate": 5.025772205424989e-06,
"loss": 0.6930059194564819,
"step": 1520
},
{
"epoch": 2.221897810218978,
"grad_norm": 1.5703125,
"learning_rate": 5.004581073467086e-06,
"loss": 0.27295994758605957,
"step": 1522
},
{
"epoch": 2.2248175182481753,
"grad_norm": 1.984375,
"learning_rate": 4.9834495302645245e-06,
"loss": 0.4621976912021637,
"step": 1524
},
{
"epoch": 2.2277372262773723,
"grad_norm": 1.0,
"learning_rate": 4.96237778584472e-06,
"loss": 0.02943926304578781,
"step": 1526
},
{
"epoch": 2.230656934306569,
"grad_norm": 1.4140625,
"learning_rate": 4.941366049640749e-06,
"loss": 0.46130675077438354,
"step": 1528
},
{
"epoch": 2.2335766423357666,
"grad_norm": 1.5,
"learning_rate": 4.920414530489257e-06,
"loss": 0.838486909866333,
"step": 1530
},
{
"epoch": 2.2364963503649635,
"grad_norm": 1.96875,
"learning_rate": 4.8995234366284e-06,
"loss": 0.5507088899612427,
"step": 1532
},
{
"epoch": 2.2394160583941605,
"grad_norm": 0.75390625,
"learning_rate": 4.8786929756957605e-06,
"loss": 0.7234805822372437,
"step": 1534
},
{
"epoch": 2.242335766423358,
"grad_norm": 5.15625,
"learning_rate": 4.857923354726279e-06,
"loss": 0.5796624422073364,
"step": 1536
},
{
"epoch": 2.245255474452555,
"grad_norm": 0.98046875,
"learning_rate": 4.837214780150214e-06,
"loss": 0.8029043674468994,
"step": 1538
},
{
"epoch": 2.2481751824817517,
"grad_norm": 4.90625,
"learning_rate": 4.81656745779108e-06,
"loss": 0.525973916053772,
"step": 1540
},
{
"epoch": 2.2510948905109487,
"grad_norm": 1.6328125,
"learning_rate": 4.7959815928636e-06,
"loss": 0.8653099536895752,
"step": 1542
},
{
"epoch": 2.254014598540146,
"grad_norm": 3.109375,
"learning_rate": 4.77545738997167e-06,
"loss": 0.6535752415657043,
"step": 1544
},
{
"epoch": 2.256934306569343,
"grad_norm": 1.75,
"learning_rate": 4.7549950531063195e-06,
"loss": 0.35478323698043823,
"step": 1546
},
{
"epoch": 2.25985401459854,
"grad_norm": 0.88671875,
"learning_rate": 4.7345947856437006e-06,
"loss": 0.697108268737793,
"step": 1548
},
{
"epoch": 2.2627737226277373,
"grad_norm": 3.546875,
"learning_rate": 4.7142567903430425e-06,
"loss": 0.7007002234458923,
"step": 1550
},
{
"epoch": 2.2656934306569343,
"grad_norm": 2.09375,
"learning_rate": 4.693981269344652e-06,
"loss": 0.5467470288276672,
"step": 1552
},
{
"epoch": 2.2686131386861312,
"grad_norm": 0.6484375,
"learning_rate": 4.6737684241679e-06,
"loss": 0.680496335029602,
"step": 1554
},
{
"epoch": 2.2715328467153286,
"grad_norm": 2.0,
"learning_rate": 4.653618455709228e-06,
"loss": 0.7767674922943115,
"step": 1556
},
{
"epoch": 2.2744525547445256,
"grad_norm": 1.8203125,
"learning_rate": 4.633531564240132e-06,
"loss": 0.7562911510467529,
"step": 1558
},
{
"epoch": 2.2773722627737225,
"grad_norm": 1.1171875,
"learning_rate": 4.613507949405185e-06,
"loss": 0.6934938430786133,
"step": 1560
},
{
"epoch": 2.28029197080292,
"grad_norm": 8.25,
"learning_rate": 4.59354781022005e-06,
"loss": 0.5609025359153748,
"step": 1562
},
{
"epoch": 2.283211678832117,
"grad_norm": 1.5234375,
"learning_rate": 4.573651345069512e-06,
"loss": 0.66295325756073,
"step": 1564
},
{
"epoch": 2.286131386861314,
"grad_norm": 1.4921875,
"learning_rate": 4.553818751705485e-06,
"loss": 0.684248685836792,
"step": 1566
},
{
"epoch": 2.289051094890511,
"grad_norm": 2.125,
"learning_rate": 4.5340502272450605e-06,
"loss": 0.8481276035308838,
"step": 1568
},
{
"epoch": 2.291970802919708,
"grad_norm": 1.265625,
"learning_rate": 4.514345968168553e-06,
"loss": 0.7585752010345459,
"step": 1570
},
{
"epoch": 2.294890510948905,
"grad_norm": 1.8984375,
"learning_rate": 4.494706170317533e-06,
"loss": 0.6100776791572571,
"step": 1572
},
{
"epoch": 2.297810218978102,
"grad_norm": 1.671875,
"learning_rate": 4.475131028892891e-06,
"loss": 0.8301199674606323,
"step": 1574
},
{
"epoch": 2.3007299270072994,
"grad_norm": 1.5546875,
"learning_rate": 4.4556207384528905e-06,
"loss": 0.7175470590591431,
"step": 1576
},
{
"epoch": 2.3036496350364963,
"grad_norm": 1.484375,
"learning_rate": 4.436175492911244e-06,
"loss": 0.6281868815422058,
"step": 1578
},
{
"epoch": 2.3065693430656933,
"grad_norm": 1.6484375,
"learning_rate": 4.416795485535178e-06,
"loss": 0.7572029829025269,
"step": 1580
},
{
"epoch": 2.3094890510948907,
"grad_norm": 0.98046875,
"learning_rate": 4.39748090894351e-06,
"loss": 0.9321935176849365,
"step": 1582
},
{
"epoch": 2.3124087591240876,
"grad_norm": 2.109375,
"learning_rate": 4.378231955104737e-06,
"loss": 0.8666127920150757,
"step": 1584
},
{
"epoch": 2.3153284671532846,
"grad_norm": 1.1796875,
"learning_rate": 4.359048815335138e-06,
"loss": 0.661139190196991,
"step": 1586
},
{
"epoch": 2.318248175182482,
"grad_norm": 5.1875,
"learning_rate": 4.339931680296853e-06,
"loss": 0.449010968208313,
"step": 1588
},
{
"epoch": 2.321167883211679,
"grad_norm": 4.8125,
"learning_rate": 4.320880739996001e-06,
"loss": 0.8577573299407959,
"step": 1590
},
{
"epoch": 2.324087591240876,
"grad_norm": 4.28125,
"learning_rate": 4.301896183780786e-06,
"loss": 0.31971099972724915,
"step": 1592
},
{
"epoch": 2.3270072992700728,
"grad_norm": 1.046875,
"learning_rate": 4.282978200339625e-06,
"loss": 0.9564056396484375,
"step": 1594
},
{
"epoch": 2.32992700729927,
"grad_norm": 1.6875,
"learning_rate": 4.264126977699257e-06,
"loss": 0.7362431287765503,
"step": 1596
},
{
"epoch": 2.332846715328467,
"grad_norm": 0.69140625,
"learning_rate": 4.245342703222886e-06,
"loss": 0.7109739780426025,
"step": 1598
},
{
"epoch": 2.335766423357664,
"grad_norm": 1.28125,
"learning_rate": 4.226625563608309e-06,
"loss": 0.8343263864517212,
"step": 1600
},
{
"epoch": 2.3386861313868614,
"grad_norm": 1.015625,
"learning_rate": 4.2079757448860815e-06,
"loss": 0.40778160095214844,
"step": 1602
},
{
"epoch": 2.3416058394160584,
"grad_norm": 1.6171875,
"learning_rate": 4.189393432417636e-06,
"loss": 0.8560945391654968,
"step": 1604
},
{
"epoch": 2.3445255474452553,
"grad_norm": 4.5,
"learning_rate": 4.170878810893466e-06,
"loss": 0.6990410089492798,
"step": 1606
},
{
"epoch": 2.3474452554744527,
"grad_norm": 1.5859375,
"learning_rate": 4.1524320643312785e-06,
"loss": 0.6028338670730591,
"step": 1608
},
{
"epoch": 2.3503649635036497,
"grad_norm": 3.65625,
"learning_rate": 4.134053376074173e-06,
"loss": 0.9524053335189819,
"step": 1610
},
{
"epoch": 2.3532846715328466,
"grad_norm": 1.703125,
"learning_rate": 4.1157429287888106e-06,
"loss": 0.7160009145736694,
"step": 1612
},
{
"epoch": 2.356204379562044,
"grad_norm": 2.21875,
"learning_rate": 4.097500904463599e-06,
"loss": 0.5132364630699158,
"step": 1614
},
{
"epoch": 2.359124087591241,
"grad_norm": 0.58203125,
"learning_rate": 4.0793274844068894e-06,
"loss": 0.7284754514694214,
"step": 1616
},
{
"epoch": 2.362043795620438,
"grad_norm": 2.203125,
"learning_rate": 4.0612228492451774e-06,
"loss": 0.7883840799331665,
"step": 1618
},
{
"epoch": 2.3649635036496353,
"grad_norm": 0.75390625,
"learning_rate": 4.043187178921295e-06,
"loss": 0.6828069686889648,
"step": 1620
},
{
"epoch": 2.367883211678832,
"grad_norm": 6.125,
"learning_rate": 4.025220652692632e-06,
"loss": 0.5448120832443237,
"step": 1622
},
{
"epoch": 2.370802919708029,
"grad_norm": 3.171875,
"learning_rate": 4.007323449129348e-06,
"loss": 0.6537898778915405,
"step": 1624
},
{
"epoch": 2.373722627737226,
"grad_norm": 0.79296875,
"learning_rate": 3.9894957461126115e-06,
"loss": 0.9079203605651855,
"step": 1626
},
{
"epoch": 2.3766423357664235,
"grad_norm": 2.078125,
"learning_rate": 3.971737720832811e-06,
"loss": 0.6240258812904358,
"step": 1628
},
{
"epoch": 2.3795620437956204,
"grad_norm": 5.25,
"learning_rate": 3.95404954978781e-06,
"loss": 0.7859882712364197,
"step": 1630
},
{
"epoch": 2.3824817518248174,
"grad_norm": 2.1875,
"learning_rate": 3.936431408781182e-06,
"loss": 0.8363872766494751,
"step": 1632
},
{
"epoch": 2.3854014598540147,
"grad_norm": 1.3125,
"learning_rate": 3.918883472920482e-06,
"loss": 0.6829777956008911,
"step": 1634
},
{
"epoch": 2.3883211678832117,
"grad_norm": 2.28125,
"learning_rate": 3.901405916615472e-06,
"loss": 0.731148362159729,
"step": 1636
},
{
"epoch": 2.3912408759124086,
"grad_norm": 1.2421875,
"learning_rate": 3.883998913576426e-06,
"loss": 0.8181612491607666,
"step": 1638
},
{
"epoch": 2.394160583941606,
"grad_norm": 2.90625,
"learning_rate": 3.866662636812382e-06,
"loss": 0.7890921831130981,
"step": 1640
},
{
"epoch": 2.397080291970803,
"grad_norm": 2.609375,
"learning_rate": 3.849397258629418e-06,
"loss": 0.9046710729598999,
"step": 1642
},
{
"epoch": 2.4,
"grad_norm": 3.03125,
"learning_rate": 3.832202950628957e-06,
"loss": 0.5645934343338013,
"step": 1644
},
{
"epoch": 2.402919708029197,
"grad_norm": 1.2578125,
"learning_rate": 3.815079883706044e-06,
"loss": 0.7521035671234131,
"step": 1646
},
{
"epoch": 2.4058394160583942,
"grad_norm": 1.78125,
"learning_rate": 3.798028228047666e-06,
"loss": 0.9176548719406128,
"step": 1648
},
{
"epoch": 2.408759124087591,
"grad_norm": 1.3046875,
"learning_rate": 3.7810481531310393e-06,
"loss": 0.7428127527236938,
"step": 1650
},
{
"epoch": 2.411678832116788,
"grad_norm": 1.6484375,
"learning_rate": 3.764139827721942e-06,
"loss": 0.7727073431015015,
"step": 1652
},
{
"epoch": 2.4145985401459855,
"grad_norm": 1.4609375,
"learning_rate": 3.7473034198730253e-06,
"loss": 0.5838227272033691,
"step": 1654
},
{
"epoch": 2.4175182481751825,
"grad_norm": 1.3515625,
"learning_rate": 3.7305390969221568e-06,
"loss": 0.4641542434692383,
"step": 1656
},
{
"epoch": 2.4204379562043794,
"grad_norm": 0.9921875,
"learning_rate": 3.713847025490741e-06,
"loss": 0.9994175434112549,
"step": 1658
},
{
"epoch": 2.423357664233577,
"grad_norm": 1.1796875,
"learning_rate": 3.697227371482075e-06,
"loss": 0.8180762529373169,
"step": 1660
},
{
"epoch": 2.4262773722627737,
"grad_norm": 1.546875,
"learning_rate": 3.680680300079691e-06,
"loss": 0.8322299718856812,
"step": 1662
},
{
"epoch": 2.4291970802919707,
"grad_norm": 1.3984375,
"learning_rate": 3.6642059757457305e-06,
"loss": 0.8561440706253052,
"step": 1664
},
{
"epoch": 2.432116788321168,
"grad_norm": 2.5,
"learning_rate": 3.64780456221928e-06,
"loss": 0.7428604364395142,
"step": 1666
},
{
"epoch": 2.435036496350365,
"grad_norm": 1.6015625,
"learning_rate": 3.631476222514783e-06,
"loss": 0.5356122255325317,
"step": 1668
},
{
"epoch": 2.437956204379562,
"grad_norm": 2.40625,
"learning_rate": 3.6152211189203795e-06,
"loss": 0.7833819389343262,
"step": 1670
},
{
"epoch": 2.4408759124087593,
"grad_norm": 2.984375,
"learning_rate": 3.59903941299633e-06,
"loss": 0.6892082095146179,
"step": 1672
},
{
"epoch": 2.4437956204379563,
"grad_norm": 1.4296875,
"learning_rate": 3.5829312655733755e-06,
"loss": 0.664943277835846,
"step": 1674
},
{
"epoch": 2.4467153284671532,
"grad_norm": 1.84375,
"learning_rate": 3.5668968367511688e-06,
"loss": 0.43954727053642273,
"step": 1676
},
{
"epoch": 2.44963503649635,
"grad_norm": 4.65625,
"learning_rate": 3.55093628589666e-06,
"loss": 0.8540928959846497,
"step": 1678
},
{
"epoch": 2.4525547445255476,
"grad_norm": 3.671875,
"learning_rate": 3.5350497716425336e-06,
"loss": 0.5841000080108643,
"step": 1680
},
{
"epoch": 2.4554744525547445,
"grad_norm": 1.609375,
"learning_rate": 3.5192374518856053e-06,
"loss": 0.8527902364730835,
"step": 1682
},
{
"epoch": 2.4583941605839414,
"grad_norm": 1.1171875,
"learning_rate": 3.503499483785286e-06,
"loss": 0.8345115780830383,
"step": 1684
},
{
"epoch": 2.461313868613139,
"grad_norm": 1.7421875,
"learning_rate": 3.487836023761985e-06,
"loss": 0.7301859259605408,
"step": 1686
},
{
"epoch": 2.4642335766423358,
"grad_norm": 1.2109375,
"learning_rate": 3.47224722749559e-06,
"loss": 0.8187237977981567,
"step": 1688
},
{
"epoch": 2.4671532846715327,
"grad_norm": 1.328125,
"learning_rate": 3.4567332499238814e-06,
"loss": 0.6591298580169678,
"step": 1690
},
{
"epoch": 2.47007299270073,
"grad_norm": 1.84375,
"learning_rate": 3.4412942452410315e-06,
"loss": 0.8038393259048462,
"step": 1692
},
{
"epoch": 2.472992700729927,
"grad_norm": 1.6796875,
"learning_rate": 3.4259303668960447e-06,
"loss": 0.8325489163398743,
"step": 1694
},
{
"epoch": 2.475912408759124,
"grad_norm": 0.9765625,
"learning_rate": 3.410641767591241e-06,
"loss": 0.8613182902336121,
"step": 1696
},
{
"epoch": 2.478832116788321,
"grad_norm": 1.34375,
"learning_rate": 3.395428599280738e-06,
"loss": 0.7140976190567017,
"step": 1698
},
{
"epoch": 2.4817518248175183,
"grad_norm": 5.71875,
"learning_rate": 3.380291013168947e-06,
"loss": 0.7978304028511047,
"step": 1700
},
{
"epoch": 2.4846715328467153,
"grad_norm": 2.046875,
"learning_rate": 3.365229159709058e-06,
"loss": 0.7424026727676392,
"step": 1702
},
{
"epoch": 2.487591240875912,
"grad_norm": 4.0625,
"learning_rate": 3.350243188601553e-06,
"loss": 0.7429658770561218,
"step": 1704
},
{
"epoch": 2.4905109489051096,
"grad_norm": 4.1875,
"learning_rate": 3.335333248792713e-06,
"loss": 0.9073584079742432,
"step": 1706
},
{
"epoch": 2.4934306569343065,
"grad_norm": 4.0,
"learning_rate": 3.3204994884731438e-06,
"loss": 0.8908743858337402,
"step": 1708
},
{
"epoch": 2.4963503649635035,
"grad_norm": 3.015625,
"learning_rate": 3.305742055076305e-06,
"loss": 0.5691918730735779,
"step": 1710
},
{
"epoch": 2.499270072992701,
"grad_norm": 6.125,
"learning_rate": 3.291061095277022e-06,
"loss": 0.5017768740653992,
"step": 1712
},
{
"epoch": 2.502189781021898,
"grad_norm": 1.59375,
"learning_rate": 3.276456754990063e-06,
"loss": 0.6777371168136597,
"step": 1714
},
{
"epoch": 2.5051094890510948,
"grad_norm": 1.5,
"learning_rate": 3.26192917936866e-06,
"loss": 0.7224640846252441,
"step": 1716
},
{
"epoch": 2.508029197080292,
"grad_norm": 1.3828125,
"learning_rate": 3.2474785128030874e-06,
"loss": 0.7180345058441162,
"step": 1718
},
{
"epoch": 2.510948905109489,
"grad_norm": 1.671875,
"learning_rate": 3.233104898919204e-06,
"loss": 0.7574636936187744,
"step": 1720
},
{
"epoch": 2.513868613138686,
"grad_norm": 2.015625,
"learning_rate": 3.2188084805770516e-06,
"loss": 0.8155980110168457,
"step": 1722
},
{
"epoch": 2.5167883211678834,
"grad_norm": 0.62109375,
"learning_rate": 3.204589399869413e-06,
"loss": 0.633004903793335,
"step": 1724
},
{
"epoch": 2.5197080291970804,
"grad_norm": 1.5390625,
"learning_rate": 3.1904477981204095e-06,
"loss": 0.8904657363891602,
"step": 1726
},
{
"epoch": 2.5226277372262773,
"grad_norm": 3.6875,
"learning_rate": 3.1763838158840975e-06,
"loss": 0.9445555806159973,
"step": 1728
},
{
"epoch": 2.5255474452554747,
"grad_norm": 1.84375,
"learning_rate": 3.162397592943072e-06,
"loss": 0.7310269474983215,
"step": 1730
},
{
"epoch": 2.5284671532846716,
"grad_norm": 2.8125,
"learning_rate": 3.148489268307069e-06,
"loss": 0.8217026591300964,
"step": 1732
},
{
"epoch": 2.5313868613138686,
"grad_norm": 0.671875,
"learning_rate": 3.134658980211592e-06,
"loss": 0.757714033126831,
"step": 1734
},
{
"epoch": 2.5343065693430655,
"grad_norm": 3.1875,
"learning_rate": 3.120906866116534e-06,
"loss": 0.33906611800193787,
"step": 1736
},
{
"epoch": 2.537226277372263,
"grad_norm": 1.578125,
"learning_rate": 3.1072330627048186e-06,
"loss": 0.6817200183868408,
"step": 1738
},
{
"epoch": 2.54014598540146,
"grad_norm": 2.046875,
"learning_rate": 3.0936377058810297e-06,
"loss": 0.9685699343681335,
"step": 1740
},
{
"epoch": 2.543065693430657,
"grad_norm": 1.6875,
"learning_rate": 3.0801209307700667e-06,
"loss": 0.44337669014930725,
"step": 1742
},
{
"epoch": 2.5459854014598537,
"grad_norm": 1.5234375,
"learning_rate": 3.0666828717158047e-06,
"loss": 0.7047996520996094,
"step": 1744
},
{
"epoch": 2.548905109489051,
"grad_norm": 1.171875,
"learning_rate": 3.0533236622797573e-06,
"loss": 0.6539790630340576,
"step": 1746
},
{
"epoch": 2.551824817518248,
"grad_norm": 6.65625,
"learning_rate": 3.0400434352397444e-06,
"loss": 0.6270900964736938,
"step": 1748
},
{
"epoch": 2.554744525547445,
"grad_norm": 0.859375,
"learning_rate": 3.026842322588577e-06,
"loss": 1.0059078931808472,
"step": 1750
},
{
"epoch": 2.5576642335766424,
"grad_norm": 1.3046875,
"learning_rate": 3.0137204555327433e-06,
"loss": 0.7666376829147339,
"step": 1752
},
{
"epoch": 2.5605839416058394,
"grad_norm": 1.6171875,
"learning_rate": 3.0006779644911106e-06,
"loss": 0.7863413095474243,
"step": 1754
},
{
"epoch": 2.5635036496350363,
"grad_norm": 2.4375,
"learning_rate": 2.9877149790936196e-06,
"loss": 0.8101619482040405,
"step": 1756
},
{
"epoch": 2.5664233576642337,
"grad_norm": 3.46875,
"learning_rate": 2.9748316281800015e-06,
"loss": 0.8166474103927612,
"step": 1758
},
{
"epoch": 2.5693430656934306,
"grad_norm": 1.359375,
"learning_rate": 2.9620280397984958e-06,
"loss": 0.6388765573501587,
"step": 1760
},
{
"epoch": 2.5722627737226276,
"grad_norm": 1.578125,
"learning_rate": 2.949304341204582e-06,
"loss": 0.4045102894306183,
"step": 1762
},
{
"epoch": 2.575182481751825,
"grad_norm": 2.703125,
"learning_rate": 2.9366606588597067e-06,
"loss": 0.857252836227417,
"step": 1764
},
{
"epoch": 2.578102189781022,
"grad_norm": 4.4375,
"learning_rate": 2.9240971184300353e-06,
"loss": 0.7178918719291687,
"step": 1766
},
{
"epoch": 2.581021897810219,
"grad_norm": 1.8046875,
"learning_rate": 2.9116138447851932e-06,
"loss": 0.6988078355789185,
"step": 1768
},
{
"epoch": 2.5839416058394162,
"grad_norm": 0.59375,
"learning_rate": 2.8992109619970386e-06,
"loss": 0.9602130651473999,
"step": 1770
},
{
"epoch": 2.586861313868613,
"grad_norm": 1.8515625,
"learning_rate": 2.8868885933384137e-06,
"loss": 0.6232843995094299,
"step": 1772
},
{
"epoch": 2.58978102189781,
"grad_norm": 1.875,
"learning_rate": 2.87464686128193e-06,
"loss": 0.9189915060997009,
"step": 1774
},
{
"epoch": 2.5927007299270075,
"grad_norm": 1.90625,
"learning_rate": 2.8624858874987464e-06,
"loss": 0.7828019857406616,
"step": 1776
},
{
"epoch": 2.5956204379562045,
"grad_norm": 2.109375,
"learning_rate": 2.850405792857365e-06,
"loss": 0.6011842489242554,
"step": 1778
},
{
"epoch": 2.5985401459854014,
"grad_norm": 1.5546875,
"learning_rate": 2.8384066974224245e-06,
"loss": 0.5729674100875854,
"step": 1780
},
{
"epoch": 2.601459854014599,
"grad_norm": 1.0390625,
"learning_rate": 2.826488720453504e-06,
"loss": 0.840190052986145,
"step": 1782
},
{
"epoch": 2.6043795620437957,
"grad_norm": 1.671875,
"learning_rate": 2.814651980403951e-06,
"loss": 0.8621047735214233,
"step": 1784
},
{
"epoch": 2.6072992700729927,
"grad_norm": 1.6484375,
"learning_rate": 2.8028965949196898e-06,
"loss": 0.7542301416397095,
"step": 1786
},
{
"epoch": 2.61021897810219,
"grad_norm": 7.71875,
"learning_rate": 2.791222680838056e-06,
"loss": 0.265644371509552,
"step": 1788
},
{
"epoch": 2.613138686131387,
"grad_norm": 5.15625,
"learning_rate": 2.779630354186641e-06,
"loss": 0.6778677701950073,
"step": 1790
},
{
"epoch": 2.616058394160584,
"grad_norm": 2.546875,
"learning_rate": 2.768119730182135e-06,
"loss": 0.657376766204834,
"step": 1792
},
{
"epoch": 2.618978102189781,
"grad_norm": 2.125,
"learning_rate": 2.7566909232291796e-06,
"loss": 0.5387110114097595,
"step": 1794
},
{
"epoch": 2.621897810218978,
"grad_norm": 4.375,
"learning_rate": 2.7453440469192337e-06,
"loss": 0.885540246963501,
"step": 1796
},
{
"epoch": 2.624817518248175,
"grad_norm": 2.34375,
"learning_rate": 2.7340792140294403e-06,
"loss": 0.9935919046401978,
"step": 1798
},
{
"epoch": 2.627737226277372,
"grad_norm": 1.8359375,
"learning_rate": 2.7228965365215177e-06,
"loss": 0.8067992925643921,
"step": 1800
},
{
"epoch": 2.630656934306569,
"grad_norm": 3.359375,
"learning_rate": 2.711796125540628e-06,
"loss": 0.5137391090393066,
"step": 1802
},
{
"epoch": 2.6335766423357665,
"grad_norm": 4.96875,
"learning_rate": 2.7007780914142883e-06,
"loss": 0.9413221478462219,
"step": 1804
},
{
"epoch": 2.6364963503649634,
"grad_norm": 2.3125,
"learning_rate": 2.6898425436512645e-06,
"loss": 0.6748592853546143,
"step": 1806
},
{
"epoch": 2.6394160583941604,
"grad_norm": 1.3515625,
"learning_rate": 2.678989590940492e-06,
"loss": 0.4462352693080902,
"step": 1808
},
{
"epoch": 2.6423357664233578,
"grad_norm": 3.5,
"learning_rate": 2.6682193411499844e-06,
"loss": 0.7442806363105774,
"step": 1810
},
{
"epoch": 2.6452554744525547,
"grad_norm": 2.234375,
"learning_rate": 2.657531901325771e-06,
"loss": 0.8577563762664795,
"step": 1812
},
{
"epoch": 2.6481751824817517,
"grad_norm": 0.71875,
"learning_rate": 2.646927377690823e-06,
"loss": 0.848198413848877,
"step": 1814
},
{
"epoch": 2.651094890510949,
"grad_norm": 1.515625,
"learning_rate": 2.6364058756440132e-06,
"loss": 0.9002335071563721,
"step": 1816
},
{
"epoch": 2.654014598540146,
"grad_norm": 3.171875,
"learning_rate": 2.6259674997590502e-06,
"loss": 0.46745210886001587,
"step": 1818
},
{
"epoch": 2.656934306569343,
"grad_norm": 0.9609375,
"learning_rate": 2.6156123537834515e-06,
"loss": 0.7870053648948669,
"step": 1820
},
{
"epoch": 2.6598540145985403,
"grad_norm": 1.2109375,
"learning_rate": 2.6053405406375074e-06,
"loss": 0.9401460886001587,
"step": 1822
},
{
"epoch": 2.6627737226277373,
"grad_norm": 0.48828125,
"learning_rate": 2.5951521624132623e-06,
"loss": 0.5525987148284912,
"step": 1824
},
{
"epoch": 2.665693430656934,
"grad_norm": 1.75,
"learning_rate": 2.5850473203734923e-06,
"loss": 0.8264008164405823,
"step": 1826
},
{
"epoch": 2.6686131386861316,
"grad_norm": 1.3125,
"learning_rate": 2.575026114950706e-06,
"loss": 0.37122803926467896,
"step": 1828
},
{
"epoch": 2.6715328467153285,
"grad_norm": 1.3046875,
"learning_rate": 2.565088645746142e-06,
"loss": 0.833563506603241,
"step": 1830
},
{
"epoch": 2.6744525547445255,
"grad_norm": 2.234375,
"learning_rate": 2.555235011528786e-06,
"loss": 0.7572871446609497,
"step": 1832
},
{
"epoch": 2.677372262773723,
"grad_norm": 1.390625,
"learning_rate": 2.5454653102343734e-06,
"loss": 0.40431341528892517,
"step": 1834
},
{
"epoch": 2.68029197080292,
"grad_norm": 2.671875,
"learning_rate": 2.535779638964437e-06,
"loss": 0.8050529956817627,
"step": 1836
},
{
"epoch": 2.6832116788321168,
"grad_norm": 1.3125,
"learning_rate": 2.5261780939853247e-06,
"loss": 0.8251926898956299,
"step": 1838
},
{
"epoch": 2.686131386861314,
"grad_norm": 2.71875,
"learning_rate": 2.5166607707272546e-06,
"loss": 0.6571418046951294,
"step": 1840
},
{
"epoch": 2.689051094890511,
"grad_norm": 2.828125,
"learning_rate": 2.507227763783351e-06,
"loss": 0.9095054268836975,
"step": 1842
},
{
"epoch": 2.691970802919708,
"grad_norm": 0.86328125,
"learning_rate": 2.4978791669087256e-06,
"loss": 1.0508363246917725,
"step": 1844
},
{
"epoch": 2.694890510948905,
"grad_norm": 2.421875,
"learning_rate": 2.488615073019527e-06,
"loss": 0.8601441979408264,
"step": 1846
},
{
"epoch": 2.697810218978102,
"grad_norm": 2.796875,
"learning_rate": 2.4794355741920313e-06,
"loss": 0.9534690380096436,
"step": 1848
},
{
"epoch": 2.7007299270072993,
"grad_norm": 3.828125,
"learning_rate": 2.4703407616617126e-06,
"loss": 0.5424234867095947,
"step": 1850
},
{
"epoch": 2.7036496350364962,
"grad_norm": 1.8984375,
"learning_rate": 2.461330725822348e-06,
"loss": 0.7102502584457397,
"step": 1852
},
{
"epoch": 2.706569343065693,
"grad_norm": 24.375,
"learning_rate": 2.4524055562251162e-06,
"loss": 0.6451407670974731,
"step": 1854
},
{
"epoch": 2.7094890510948906,
"grad_norm": 1.3984375,
"learning_rate": 2.4435653415777033e-06,
"loss": 0.8459779024124146,
"step": 1856
},
{
"epoch": 2.7124087591240875,
"grad_norm": 0.8203125,
"learning_rate": 2.434810169743425e-06,
"loss": 0.7882039546966553,
"step": 1858
},
{
"epoch": 2.7153284671532845,
"grad_norm": 1.765625,
"learning_rate": 2.426140127740351e-06,
"loss": 0.6165429353713989,
"step": 1860
},
{
"epoch": 2.718248175182482,
"grad_norm": 0.95703125,
"learning_rate": 2.4175553017404494e-06,
"loss": 0.7575932145118713,
"step": 1862
},
{
"epoch": 2.721167883211679,
"grad_norm": 1.6015625,
"learning_rate": 2.4090557770687095e-06,
"loss": 0.7993899583816528,
"step": 1864
},
{
"epoch": 2.7240875912408757,
"grad_norm": 1.0859375,
"learning_rate": 2.4006416382023168e-06,
"loss": 0.6624000668525696,
"step": 1866
},
{
"epoch": 2.727007299270073,
"grad_norm": 1.375,
"learning_rate": 2.392312968769797e-06,
"loss": 0.8750714063644409,
"step": 1868
},
{
"epoch": 2.72992700729927,
"grad_norm": 1.3515625,
"learning_rate": 2.3840698515501998e-06,
"loss": 0.6998022794723511,
"step": 1870
},
{
"epoch": 2.732846715328467,
"grad_norm": 0.76953125,
"learning_rate": 2.375912368472254e-06,
"loss": 0.8878059983253479,
"step": 1872
},
{
"epoch": 2.7357664233576644,
"grad_norm": 1.578125,
"learning_rate": 2.3678406006135774e-06,
"loss": 0.8704814910888672,
"step": 1874
},
{
"epoch": 2.7386861313868613,
"grad_norm": 0.78515625,
"learning_rate": 2.359854628199857e-06,
"loss": 0.8811229467391968,
"step": 1876
},
{
"epoch": 2.7416058394160583,
"grad_norm": 1.3984375,
"learning_rate": 2.3519545306040557e-06,
"loss": 0.8793524503707886,
"step": 1878
},
{
"epoch": 2.7445255474452557,
"grad_norm": 0.56640625,
"learning_rate": 2.344140386345616e-06,
"loss": 0.5257948637008667,
"step": 1880
},
{
"epoch": 2.7474452554744526,
"grad_norm": 0.890625,
"learning_rate": 2.3364122730896963e-06,
"loss": 0.6838935017585754,
"step": 1882
},
{
"epoch": 2.7503649635036496,
"grad_norm": 3.09375,
"learning_rate": 2.3287702676463817e-06,
"loss": 0.7133142948150635,
"step": 1884
},
{
"epoch": 2.753284671532847,
"grad_norm": 1.5078125,
"learning_rate": 2.321214445969931e-06,
"loss": 0.5482426285743713,
"step": 1886
},
{
"epoch": 2.756204379562044,
"grad_norm": 2.765625,
"learning_rate": 2.313744883158014e-06,
"loss": 0.4747272729873657,
"step": 1888
},
{
"epoch": 2.759124087591241,
"grad_norm": 0.30859375,
"learning_rate": 2.3063616534509766e-06,
"loss": 0.5745252370834351,
"step": 1890
},
{
"epoch": 2.7620437956204382,
"grad_norm": 1.0546875,
"learning_rate": 2.299064830231089e-06,
"loss": 0.7720372080802917,
"step": 1892
},
{
"epoch": 2.764963503649635,
"grad_norm": 2.546875,
"learning_rate": 2.2918544860218262e-06,
"loss": 0.9943041205406189,
"step": 1894
},
{
"epoch": 2.767883211678832,
"grad_norm": 4.75,
"learning_rate": 2.2847306924871454e-06,
"loss": 0.9632670879364014,
"step": 1896
},
{
"epoch": 2.770802919708029,
"grad_norm": 0.88671875,
"learning_rate": 2.2776935204307705e-06,
"loss": 0.8723521828651428,
"step": 1898
},
{
"epoch": 2.7737226277372264,
"grad_norm": 1.5234375,
"learning_rate": 2.2707430397954907e-06,
"loss": 0.8441874980926514,
"step": 1900
},
{
"epoch": 2.7766423357664234,
"grad_norm": 3.59375,
"learning_rate": 2.263879319662466e-06,
"loss": 0.5843279361724854,
"step": 1902
},
{
"epoch": 2.7795620437956203,
"grad_norm": 6.53125,
"learning_rate": 2.2571024282505374e-06,
"loss": 0.805155873298645,
"step": 1904
},
{
"epoch": 2.7824817518248173,
"grad_norm": 1.7109375,
"learning_rate": 2.250412432915554e-06,
"loss": 0.8528783321380615,
"step": 1906
},
{
"epoch": 2.7854014598540147,
"grad_norm": 1.28125,
"learning_rate": 2.2438094001497e-06,
"loss": 0.7224323749542236,
"step": 1908
},
{
"epoch": 2.7883211678832116,
"grad_norm": 1.40625,
"learning_rate": 2.2372933955808325e-06,
"loss": 0.8677688837051392,
"step": 1910
},
{
"epoch": 2.7912408759124085,
"grad_norm": 3.640625,
"learning_rate": 2.2308644839718307e-06,
"loss": 0.8146109580993652,
"step": 1912
},
{
"epoch": 2.794160583941606,
"grad_norm": 3.859375,
"learning_rate": 2.224522729219958e-06,
"loss": 0.7773847579956055,
"step": 1914
},
{
"epoch": 2.797080291970803,
"grad_norm": 1.6171875,
"learning_rate": 2.2182681943562164e-06,
"loss": 0.8882875442504883,
"step": 1916
},
{
"epoch": 2.8,
"grad_norm": 6.9375,
"learning_rate": 2.212100941544726e-06,
"loss": 0.6702533960342407,
"step": 1918
},
{
"epoch": 2.802919708029197,
"grad_norm": 3.046875,
"learning_rate": 2.2060210320821086e-06,
"loss": 0.9423357844352722,
"step": 1920
},
{
"epoch": 2.805839416058394,
"grad_norm": 1.7265625,
"learning_rate": 2.200028526396876e-06,
"loss": 0.7287482023239136,
"step": 1922
},
{
"epoch": 2.808759124087591,
"grad_norm": 4.21875,
"learning_rate": 2.1941234840488263e-06,
"loss": 0.7009254693984985,
"step": 1924
},
{
"epoch": 2.8116788321167885,
"grad_norm": 4.15625,
"learning_rate": 2.1883059637284585e-06,
"loss": 0.9232114553451538,
"step": 1926
},
{
"epoch": 2.8145985401459854,
"grad_norm": 2.125,
"learning_rate": 2.1825760232563855e-06,
"loss": 0.9375965595245361,
"step": 1928
},
{
"epoch": 2.8175182481751824,
"grad_norm": 3.515625,
"learning_rate": 2.1769337195827607e-06,
"loss": 0.7196213006973267,
"step": 1930
},
{
"epoch": 2.8204379562043798,
"grad_norm": 1.5390625,
"learning_rate": 2.1713791087867082e-06,
"loss": 0.9744350910186768,
"step": 1932
},
{
"epoch": 2.8233576642335767,
"grad_norm": 4.09375,
"learning_rate": 2.16591224607577e-06,
"loss": 0.830897331237793,
"step": 1934
},
{
"epoch": 2.8262773722627736,
"grad_norm": 1.7109375,
"learning_rate": 2.160533185785359e-06,
"loss": 0.8744158744812012,
"step": 1936
},
{
"epoch": 2.829197080291971,
"grad_norm": 3.328125,
"learning_rate": 2.1552419813782094e-06,
"loss": 0.8899505138397217,
"step": 1938
},
{
"epoch": 2.832116788321168,
"grad_norm": 2.453125,
"learning_rate": 2.1500386854438556e-06,
"loss": 0.7949346303939819,
"step": 1940
},
{
"epoch": 2.835036496350365,
"grad_norm": 1.6796875,
"learning_rate": 2.144923349698105e-06,
"loss": 0.8629809021949768,
"step": 1942
},
{
"epoch": 2.8379562043795623,
"grad_norm": 2.46875,
"learning_rate": 2.139896024982525e-06,
"loss": 0.8164567947387695,
"step": 1944
},
{
"epoch": 2.8408759124087593,
"grad_norm": 2.203125,
"learning_rate": 2.1349567612639347e-06,
"loss": 0.7339398860931396,
"step": 1946
},
{
"epoch": 2.843795620437956,
"grad_norm": 1.8671875,
"learning_rate": 2.1301056076339134e-06,
"loss": 0.9814774990081787,
"step": 1948
},
{
"epoch": 2.846715328467153,
"grad_norm": 1.4296875,
"learning_rate": 2.1253426123083075e-06,
"loss": 0.9599044322967529,
"step": 1950
},
{
"epoch": 2.8496350364963505,
"grad_norm": 2.40625,
"learning_rate": 2.1206678226267568e-06,
"loss": 0.781501829624176,
"step": 1952
},
{
"epoch": 2.8525547445255475,
"grad_norm": 0.875,
"learning_rate": 2.1160812850522197e-06,
"loss": 0.8890868425369263,
"step": 1954
},
{
"epoch": 2.8554744525547444,
"grad_norm": 4.21875,
"learning_rate": 2.1115830451705103e-06,
"loss": 0.8284950256347656,
"step": 1956
},
{
"epoch": 2.8583941605839414,
"grad_norm": 1.6328125,
"learning_rate": 2.1071731476898525e-06,
"loss": 0.8606098890304565,
"step": 1958
},
{
"epoch": 2.8613138686131387,
"grad_norm": 0.546875,
"learning_rate": 2.1028516364404255e-06,
"loss": 0.919540286064148,
"step": 1960
},
{
"epoch": 2.8642335766423357,
"grad_norm": 1.4609375,
"learning_rate": 2.0986185543739387e-06,
"loss": 0.9102145433425903,
"step": 1962
},
{
"epoch": 2.8671532846715326,
"grad_norm": 1.65625,
"learning_rate": 2.0944739435631965e-06,
"loss": 0.90533447265625,
"step": 1964
},
{
"epoch": 2.87007299270073,
"grad_norm": 1.25,
"learning_rate": 2.090417845201682e-06,
"loss": 0.9714438319206238,
"step": 1966
},
{
"epoch": 2.872992700729927,
"grad_norm": 0.6796875,
"learning_rate": 2.086450299603153e-06,
"loss": 0.5790750980377197,
"step": 1968
},
{
"epoch": 2.875912408759124,
"grad_norm": 1.1640625,
"learning_rate": 2.082571346201234e-06,
"loss": 0.8377712965011597,
"step": 1970
},
{
"epoch": 2.8788321167883213,
"grad_norm": 1.4921875,
"learning_rate": 2.078781023549028e-06,
"loss": 0.867621898651123,
"step": 1972
},
{
"epoch": 2.8817518248175182,
"grad_norm": 1.1328125,
"learning_rate": 2.075079369318733e-06,
"loss": 1.032489538192749,
"step": 1974
},
{
"epoch": 2.884671532846715,
"grad_norm": 1.3046875,
"learning_rate": 2.071466420301268e-06,
"loss": 0.8769449591636658,
"step": 1976
},
{
"epoch": 2.8875912408759126,
"grad_norm": 3.59375,
"learning_rate": 2.067942212405907e-06,
"loss": 0.255039781332016,
"step": 1978
},
{
"epoch": 2.8905109489051095,
"grad_norm": 2.171875,
"learning_rate": 2.064506780659918e-06,
"loss": 0.6561243534088135,
"step": 1980
},
{
"epoch": 2.8934306569343065,
"grad_norm": 1.5546875,
"learning_rate": 2.061160159208224e-06,
"loss": 0.9538788795471191,
"step": 1982
},
{
"epoch": 2.896350364963504,
"grad_norm": 3.40625,
"learning_rate": 2.0579023813130535e-06,
"loss": 0.9035916328430176,
"step": 1984
},
{
"epoch": 2.899270072992701,
"grad_norm": 0.625,
"learning_rate": 2.054733479353618e-06,
"loss": 0.8100244998931885,
"step": 1986
},
{
"epoch": 2.9021897810218977,
"grad_norm": 0.3125,
"learning_rate": 2.0516534848257856e-06,
"loss": 0.49608534574508667,
"step": 1988
},
{
"epoch": 2.905109489051095,
"grad_norm": 1.953125,
"learning_rate": 2.0486624283417685e-06,
"loss": 0.749464750289917,
"step": 1990
},
{
"epoch": 2.908029197080292,
"grad_norm": 2.234375,
"learning_rate": 2.0457603396298204e-06,
"loss": 0.667478084564209,
"step": 1992
},
{
"epoch": 2.910948905109489,
"grad_norm": 1.96875,
"learning_rate": 2.0429472475339403e-06,
"loss": 0.8862441778182983,
"step": 1994
},
{
"epoch": 2.9138686131386864,
"grad_norm": 1.3125,
"learning_rate": 2.040223180013584e-06,
"loss": 0.8121696710586548,
"step": 1996
},
{
"epoch": 2.9167883211678833,
"grad_norm": 2.8125,
"learning_rate": 2.0375881641433896e-06,
"loss": 0.7258896827697754,
"step": 1998
},
{
"epoch": 2.9197080291970803,
"grad_norm": 3.015625,
"learning_rate": 2.0350422261129058e-06,
"loss": 0.8913263082504272,
"step": 2000
},
{
"epoch": 2.9226277372262772,
"grad_norm": 1.5546875,
"learning_rate": 2.0325853912263317e-06,
"loss": 0.831497848033905,
"step": 2002
},
{
"epoch": 2.9255474452554746,
"grad_norm": 1.40625,
"learning_rate": 2.0302176839022675e-06,
"loss": 0.8855400085449219,
"step": 2004
},
{
"epoch": 2.9284671532846716,
"grad_norm": 5.4375,
"learning_rate": 2.0279391276734694e-06,
"loss": 0.8521960973739624,
"step": 2006
},
{
"epoch": 2.9313868613138685,
"grad_norm": 1.296875,
"learning_rate": 2.0257497451866165e-06,
"loss": 0.9489337205886841,
"step": 2008
},
{
"epoch": 2.9343065693430654,
"grad_norm": 1.28125,
"learning_rate": 2.0236495582020855e-06,
"loss": 0.5924053192138672,
"step": 2010
},
{
"epoch": 2.937226277372263,
"grad_norm": 3.828125,
"learning_rate": 2.021638587593735e-06,
"loss": 0.8434425592422485,
"step": 2012
},
{
"epoch": 2.9401459854014598,
"grad_norm": 1.1796875,
"learning_rate": 2.0197168533486997e-06,
"loss": 0.9456912279129028,
"step": 2014
},
{
"epoch": 2.9430656934306567,
"grad_norm": 2.125,
"learning_rate": 2.017884374567187e-06,
"loss": 0.823800802230835,
"step": 2016
},
{
"epoch": 2.945985401459854,
"grad_norm": 1.4921875,
"learning_rate": 2.016141169462291e-06,
"loss": 0.46315228939056396,
"step": 2018
},
{
"epoch": 2.948905109489051,
"grad_norm": 1.171875,
"learning_rate": 2.0144872553598107e-06,
"loss": 0.7524667978286743,
"step": 2020
},
{
"epoch": 2.951824817518248,
"grad_norm": 2.578125,
"learning_rate": 2.0129226486980786e-06,
"loss": 0.9795228242874146,
"step": 2022
},
{
"epoch": 2.9547445255474454,
"grad_norm": 6.21875,
"learning_rate": 2.0114473650277925e-06,
"loss": 0.9093575477600098,
"step": 2024
},
{
"epoch": 2.9576642335766423,
"grad_norm": 2.625,
"learning_rate": 2.01006141901187e-06,
"loss": 1.015989065170288,
"step": 2026
},
{
"epoch": 2.9605839416058393,
"grad_norm": 0.92578125,
"learning_rate": 2.0087648244252936e-06,
"loss": 0.8345716595649719,
"step": 2028
},
{
"epoch": 2.9635036496350367,
"grad_norm": 1.8828125,
"learning_rate": 2.0075575941549793e-06,
"loss": 0.6250303983688354,
"step": 2030
},
{
"epoch": 2.9664233576642336,
"grad_norm": 0.88671875,
"learning_rate": 2.006439740199645e-06,
"loss": 0.8835278153419495,
"step": 2032
},
{
"epoch": 2.9693430656934305,
"grad_norm": 1.390625,
"learning_rate": 2.005411273669696e-06,
"loss": 0.7776151299476624,
"step": 2034
},
{
"epoch": 2.972262773722628,
"grad_norm": 4.3125,
"learning_rate": 2.0044722047871113e-06,
"loss": 0.8826329708099365,
"step": 2036
},
{
"epoch": 2.975182481751825,
"grad_norm": 6.6875,
"learning_rate": 2.003622542885339e-06,
"loss": 0.7542710304260254,
"step": 2038
},
{
"epoch": 2.978102189781022,
"grad_norm": 3.25,
"learning_rate": 2.0028622964092113e-06,
"loss": 1.0535058975219727,
"step": 2040
},
{
"epoch": 2.981021897810219,
"grad_norm": 1.7109375,
"learning_rate": 2.0021914729148532e-06,
"loss": 0.9216081500053406,
"step": 2042
},
{
"epoch": 2.983941605839416,
"grad_norm": 2.875,
"learning_rate": 2.0016100790696116e-06,
"loss": 0.9473830461502075,
"step": 2044
},
{
"epoch": 2.986861313868613,
"grad_norm": 1.0234375,
"learning_rate": 2.0011181206519882e-06,
"loss": 0.03890227898955345,
"step": 2046
},
{
"epoch": 2.9897810218978105,
"grad_norm": 1.375,
"learning_rate": 2.0007156025515813e-06,
"loss": 0.9145385026931763,
"step": 2048
},
{
"epoch": 2.9927007299270074,
"grad_norm": 2.703125,
"learning_rate": 2.0004025287690367e-06,
"loss": 0.9966373443603516,
"step": 2050
},
{
"epoch": 2.9956204379562044,
"grad_norm": 0.96875,
"learning_rate": 2.0001789024160103e-06,
"loss": 0.898999810218811,
"step": 2052
},
{
"epoch": 2.9985401459854013,
"grad_norm": 2.546875,
"learning_rate": 2.0000447257151354e-06,
"loss": 1.0698673725128174,
"step": 2054
},
{
"epoch": 3.0,
"step": 2055,
"total_flos": 2.976816812190073e+18,
"train_loss": 0.9887462435448837,
"train_runtime": 18114.3764,
"train_samples_per_second": 1.815,
"train_steps_per_second": 0.113
}
],
"logging_steps": 2,
"max_steps": 2055,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 9999999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.976816812190073e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}