martimfasantos's picture
Model save
ae55cbd verified
raw
history blame contribute delete
No virus
49.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9996613613274636,
"eval_steps": 500,
"global_step": 1476,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006772773450728074,
"grad_norm": 11.0625,
"learning_rate": 6.756756756756757e-08,
"loss": 2.4477,
"step": 1
},
{
"epoch": 0.0033863867253640365,
"grad_norm": 11.8125,
"learning_rate": 3.378378378378379e-07,
"loss": 2.4744,
"step": 5
},
{
"epoch": 0.006772773450728073,
"grad_norm": 12.625,
"learning_rate": 6.756756756756758e-07,
"loss": 2.4671,
"step": 10
},
{
"epoch": 0.01015916017609211,
"grad_norm": 11.8125,
"learning_rate": 1.0135135135135136e-06,
"loss": 2.4589,
"step": 15
},
{
"epoch": 0.013545546901456146,
"grad_norm": 11.6875,
"learning_rate": 1.3513513513513515e-06,
"loss": 2.4356,
"step": 20
},
{
"epoch": 0.016931933626820182,
"grad_norm": 10.5625,
"learning_rate": 1.6891891891891894e-06,
"loss": 2.4538,
"step": 25
},
{
"epoch": 0.02031832035218422,
"grad_norm": 8.3125,
"learning_rate": 2.0270270270270273e-06,
"loss": 2.4135,
"step": 30
},
{
"epoch": 0.023704707077548254,
"grad_norm": 5.0625,
"learning_rate": 2.364864864864865e-06,
"loss": 2.4153,
"step": 35
},
{
"epoch": 0.027091093802912292,
"grad_norm": 3.40625,
"learning_rate": 2.702702702702703e-06,
"loss": 2.409,
"step": 40
},
{
"epoch": 0.03047748052827633,
"grad_norm": 2.96875,
"learning_rate": 3.040540540540541e-06,
"loss": 2.3663,
"step": 45
},
{
"epoch": 0.033863867253640365,
"grad_norm": 2.375,
"learning_rate": 3.3783783783783788e-06,
"loss": 2.348,
"step": 50
},
{
"epoch": 0.0372502539790044,
"grad_norm": 2.25,
"learning_rate": 3.7162162162162162e-06,
"loss": 2.3493,
"step": 55
},
{
"epoch": 0.04063664070436844,
"grad_norm": 1.8359375,
"learning_rate": 4.0540540540540545e-06,
"loss": 2.338,
"step": 60
},
{
"epoch": 0.04402302742973248,
"grad_norm": 1.765625,
"learning_rate": 4.391891891891892e-06,
"loss": 2.3151,
"step": 65
},
{
"epoch": 0.04740941415509651,
"grad_norm": 1.6953125,
"learning_rate": 4.72972972972973e-06,
"loss": 2.2869,
"step": 70
},
{
"epoch": 0.05079580088046055,
"grad_norm": 1.7109375,
"learning_rate": 5.067567567567568e-06,
"loss": 2.2964,
"step": 75
},
{
"epoch": 0.054182187605824585,
"grad_norm": 1.59375,
"learning_rate": 5.405405405405406e-06,
"loss": 2.2719,
"step": 80
},
{
"epoch": 0.05756857433118862,
"grad_norm": 1.5234375,
"learning_rate": 5.743243243243244e-06,
"loss": 2.2315,
"step": 85
},
{
"epoch": 0.06095496105655266,
"grad_norm": 1.515625,
"learning_rate": 6.081081081081082e-06,
"loss": 2.255,
"step": 90
},
{
"epoch": 0.06434134778191669,
"grad_norm": 1.46875,
"learning_rate": 6.41891891891892e-06,
"loss": 2.2319,
"step": 95
},
{
"epoch": 0.06772773450728073,
"grad_norm": 1.421875,
"learning_rate": 6.7567567567567575e-06,
"loss": 2.224,
"step": 100
},
{
"epoch": 0.07111412123264477,
"grad_norm": 1.3359375,
"learning_rate": 7.0945945945945946e-06,
"loss": 2.2143,
"step": 105
},
{
"epoch": 0.0745005079580088,
"grad_norm": 1.3671875,
"learning_rate": 7.4324324324324324e-06,
"loss": 2.2221,
"step": 110
},
{
"epoch": 0.07788689468337284,
"grad_norm": 1.3984375,
"learning_rate": 7.77027027027027e-06,
"loss": 2.2246,
"step": 115
},
{
"epoch": 0.08127328140873688,
"grad_norm": 1.359375,
"learning_rate": 8.108108108108109e-06,
"loss": 2.2054,
"step": 120
},
{
"epoch": 0.08465966813410092,
"grad_norm": 1.3125,
"learning_rate": 8.445945945945948e-06,
"loss": 2.189,
"step": 125
},
{
"epoch": 0.08804605485946496,
"grad_norm": 1.296875,
"learning_rate": 8.783783783783785e-06,
"loss": 2.2255,
"step": 130
},
{
"epoch": 0.091432441584829,
"grad_norm": 1.3359375,
"learning_rate": 9.121621621621622e-06,
"loss": 2.2229,
"step": 135
},
{
"epoch": 0.09481882831019302,
"grad_norm": 1.3359375,
"learning_rate": 9.45945945945946e-06,
"loss": 2.1797,
"step": 140
},
{
"epoch": 0.09820521503555706,
"grad_norm": 1.265625,
"learning_rate": 9.797297297297298e-06,
"loss": 2.18,
"step": 145
},
{
"epoch": 0.1015916017609211,
"grad_norm": 1.28125,
"learning_rate": 9.999944036768366e-06,
"loss": 2.1779,
"step": 150
},
{
"epoch": 0.10497798848628513,
"grad_norm": 1.3046875,
"learning_rate": 9.999314464799478e-06,
"loss": 2.1901,
"step": 155
},
{
"epoch": 0.10836437521164917,
"grad_norm": 1.296875,
"learning_rate": 9.997985455197114e-06,
"loss": 2.2083,
"step": 160
},
{
"epoch": 0.11175076193701321,
"grad_norm": 1.3203125,
"learning_rate": 9.995957193898633e-06,
"loss": 2.1738,
"step": 165
},
{
"epoch": 0.11513714866237724,
"grad_norm": 1.265625,
"learning_rate": 9.9932299646714e-06,
"loss": 2.1735,
"step": 170
},
{
"epoch": 0.11852353538774128,
"grad_norm": 1.3203125,
"learning_rate": 9.98980414907308e-06,
"loss": 2.1897,
"step": 175
},
{
"epoch": 0.12190992211310532,
"grad_norm": 1.296875,
"learning_rate": 9.985680226398261e-06,
"loss": 2.1894,
"step": 180
},
{
"epoch": 0.12529630883846934,
"grad_norm": 1.28125,
"learning_rate": 9.980858773611387e-06,
"loss": 2.187,
"step": 185
},
{
"epoch": 0.12868269556383338,
"grad_norm": 1.2734375,
"learning_rate": 9.975340465266054e-06,
"loss": 2.1754,
"step": 190
},
{
"epoch": 0.13206908228919742,
"grad_norm": 1.3046875,
"learning_rate": 9.969126073410618e-06,
"loss": 2.2011,
"step": 195
},
{
"epoch": 0.13545546901456146,
"grad_norm": 1.28125,
"learning_rate": 9.962216467480192e-06,
"loss": 2.1805,
"step": 200
},
{
"epoch": 0.1388418557399255,
"grad_norm": 1.2265625,
"learning_rate": 9.954612614175004e-06,
"loss": 2.1972,
"step": 205
},
{
"epoch": 0.14222824246528953,
"grad_norm": 1.2734375,
"learning_rate": 9.94631557732514e-06,
"loss": 2.1828,
"step": 210
},
{
"epoch": 0.14561462919065357,
"grad_norm": 1.265625,
"learning_rate": 9.937326517741725e-06,
"loss": 2.1778,
"step": 215
},
{
"epoch": 0.1490010159160176,
"grad_norm": 1.2265625,
"learning_rate": 9.927646693054498e-06,
"loss": 2.1709,
"step": 220
},
{
"epoch": 0.15238740264138165,
"grad_norm": 1.3203125,
"learning_rate": 9.917277457535873e-06,
"loss": 2.1903,
"step": 225
},
{
"epoch": 0.15577378936674569,
"grad_norm": 1.265625,
"learning_rate": 9.906220261911466e-06,
"loss": 2.1777,
"step": 230
},
{
"epoch": 0.15916017609210972,
"grad_norm": 1.265625,
"learning_rate": 9.894476653157125e-06,
"loss": 2.1703,
"step": 235
},
{
"epoch": 0.16254656281747376,
"grad_norm": 1.2890625,
"learning_rate": 9.882048274282505e-06,
"loss": 2.1767,
"step": 240
},
{
"epoch": 0.1659329495428378,
"grad_norm": 1.2421875,
"learning_rate": 9.868936864101188e-06,
"loss": 2.1766,
"step": 245
},
{
"epoch": 0.16931933626820184,
"grad_norm": 1.2421875,
"learning_rate": 9.855144256987423e-06,
"loss": 2.1813,
"step": 250
},
{
"epoch": 0.17270572299356587,
"grad_norm": 1.2265625,
"learning_rate": 9.840672382619479e-06,
"loss": 2.1853,
"step": 255
},
{
"epoch": 0.1760921097189299,
"grad_norm": 1.296875,
"learning_rate": 9.825523265709667e-06,
"loss": 2.1928,
"step": 260
},
{
"epoch": 0.17947849644429395,
"grad_norm": 1.2578125,
"learning_rate": 9.80969902572108e-06,
"loss": 2.1909,
"step": 265
},
{
"epoch": 0.182864883169658,
"grad_norm": 1.28125,
"learning_rate": 9.793201876571052e-06,
"loss": 2.1736,
"step": 270
},
{
"epoch": 0.186251269895022,
"grad_norm": 1.265625,
"learning_rate": 9.776034126321429e-06,
"loss": 2.1925,
"step": 275
},
{
"epoch": 0.18963765662038604,
"grad_norm": 1.25,
"learning_rate": 9.758198176855648e-06,
"loss": 2.1663,
"step": 280
},
{
"epoch": 0.19302404334575007,
"grad_norm": 1.234375,
"learning_rate": 9.739696523542696e-06,
"loss": 2.1875,
"step": 285
},
{
"epoch": 0.1964104300711141,
"grad_norm": 1.2734375,
"learning_rate": 9.720531754888002e-06,
"loss": 2.1673,
"step": 290
},
{
"epoch": 0.19979681679647815,
"grad_norm": 1.234375,
"learning_rate": 9.700706552171268e-06,
"loss": 2.1534,
"step": 295
},
{
"epoch": 0.2031832035218422,
"grad_norm": 1.2734375,
"learning_rate": 9.680223689071364e-06,
"loss": 2.1819,
"step": 300
},
{
"epoch": 0.20656959024720623,
"grad_norm": 1.2578125,
"learning_rate": 9.659086031278256e-06,
"loss": 2.1523,
"step": 305
},
{
"epoch": 0.20995597697257026,
"grad_norm": 1.234375,
"learning_rate": 9.637296536092076e-06,
"loss": 2.2039,
"step": 310
},
{
"epoch": 0.2133423636979343,
"grad_norm": 1.234375,
"learning_rate": 9.614858252009385e-06,
"loss": 2.1527,
"step": 315
},
{
"epoch": 0.21672875042329834,
"grad_norm": 1.2421875,
"learning_rate": 9.591774318296661e-06,
"loss": 2.1698,
"step": 320
},
{
"epoch": 0.22011513714866238,
"grad_norm": 1.28125,
"learning_rate": 9.568047964551102e-06,
"loss": 2.1633,
"step": 325
},
{
"epoch": 0.22350152387402641,
"grad_norm": 1.25,
"learning_rate": 9.54368251024877e-06,
"loss": 2.1671,
"step": 330
},
{
"epoch": 0.22688791059939045,
"grad_norm": 1.25,
"learning_rate": 9.51868136428019e-06,
"loss": 2.1933,
"step": 335
},
{
"epoch": 0.2302742973247545,
"grad_norm": 1.2890625,
"learning_rate": 9.493048024473413e-06,
"loss": 2.1617,
"step": 340
},
{
"epoch": 0.23366068405011853,
"grad_norm": 1.2265625,
"learning_rate": 9.466786077104646e-06,
"loss": 2.1816,
"step": 345
},
{
"epoch": 0.23704707077548257,
"grad_norm": 1.2734375,
"learning_rate": 9.439899196396517e-06,
"loss": 2.1776,
"step": 350
},
{
"epoch": 0.2404334575008466,
"grad_norm": 1.21875,
"learning_rate": 9.412391144004019e-06,
"loss": 2.1749,
"step": 355
},
{
"epoch": 0.24381984422621064,
"grad_norm": 1.265625,
"learning_rate": 9.384265768488226e-06,
"loss": 2.1701,
"step": 360
},
{
"epoch": 0.24720623095157468,
"grad_norm": 1.2578125,
"learning_rate": 9.355527004777868e-06,
"loss": 2.1888,
"step": 365
},
{
"epoch": 0.2505926176769387,
"grad_norm": 1.2421875,
"learning_rate": 9.32617887361879e-06,
"loss": 2.1692,
"step": 370
},
{
"epoch": 0.25397900440230275,
"grad_norm": 1.21875,
"learning_rate": 9.296225481011436e-06,
"loss": 2.165,
"step": 375
},
{
"epoch": 0.25736539112766676,
"grad_norm": 1.2109375,
"learning_rate": 9.265671017636384e-06,
"loss": 2.1893,
"step": 380
},
{
"epoch": 0.26075177785303083,
"grad_norm": 1.265625,
"learning_rate": 9.23451975826805e-06,
"loss": 2.1684,
"step": 385
},
{
"epoch": 0.26413816457839484,
"grad_norm": 1.2421875,
"learning_rate": 9.202776061176606e-06,
"loss": 2.1823,
"step": 390
},
{
"epoch": 0.2675245513037589,
"grad_norm": 1.2578125,
"learning_rate": 9.170444367518243e-06,
"loss": 2.1515,
"step": 395
},
{
"epoch": 0.2709109380291229,
"grad_norm": 1.25,
"learning_rate": 9.137529200713811e-06,
"loss": 2.1754,
"step": 400
},
{
"epoch": 0.274297324754487,
"grad_norm": 1.2109375,
"learning_rate": 9.104035165815971e-06,
"loss": 2.1664,
"step": 405
},
{
"epoch": 0.277683711479851,
"grad_norm": 1.28125,
"learning_rate": 9.069966948864917e-06,
"loss": 2.1783,
"step": 410
},
{
"epoch": 0.28107009820521506,
"grad_norm": 1.2265625,
"learning_rate": 9.035329316232755e-06,
"loss": 2.1555,
"step": 415
},
{
"epoch": 0.28445648493057907,
"grad_norm": 1.234375,
"learning_rate": 9.000127113956673e-06,
"loss": 2.1668,
"step": 420
},
{
"epoch": 0.28784287165594313,
"grad_norm": 1.2109375,
"learning_rate": 8.964365267060936e-06,
"loss": 2.1706,
"step": 425
},
{
"epoch": 0.29122925838130714,
"grad_norm": 1.2265625,
"learning_rate": 8.928048778867848e-06,
"loss": 2.1535,
"step": 430
},
{
"epoch": 0.2946156451066712,
"grad_norm": 1.2578125,
"learning_rate": 8.89118273029775e-06,
"loss": 2.1704,
"step": 435
},
{
"epoch": 0.2980020318320352,
"grad_norm": 1.2578125,
"learning_rate": 8.853772279158166e-06,
"loss": 2.1643,
"step": 440
},
{
"epoch": 0.30138841855739923,
"grad_norm": 1.21875,
"learning_rate": 8.815822659422195e-06,
"loss": 2.1575,
"step": 445
},
{
"epoch": 0.3047748052827633,
"grad_norm": 1.1953125,
"learning_rate": 8.777339180496238e-06,
"loss": 2.1684,
"step": 450
},
{
"epoch": 0.3081611920081273,
"grad_norm": 1.2734375,
"learning_rate": 8.738327226477177e-06,
"loss": 2.1451,
"step": 455
},
{
"epoch": 0.31154757873349137,
"grad_norm": 1.234375,
"learning_rate": 8.698792255399104e-06,
"loss": 2.155,
"step": 460
},
{
"epoch": 0.3149339654588554,
"grad_norm": 1.234375,
"learning_rate": 8.658739798469713e-06,
"loss": 2.1683,
"step": 465
},
{
"epoch": 0.31832035218421945,
"grad_norm": 1.296875,
"learning_rate": 8.618175459296434e-06,
"loss": 2.1671,
"step": 470
},
{
"epoch": 0.32170673890958346,
"grad_norm": 1.265625,
"learning_rate": 8.577104913102458e-06,
"loss": 2.1756,
"step": 475
},
{
"epoch": 0.3250931256349475,
"grad_norm": 1.21875,
"learning_rate": 8.535533905932739e-06,
"loss": 2.1726,
"step": 480
},
{
"epoch": 0.32847951236031153,
"grad_norm": 1.25,
"learning_rate": 8.49346825385007e-06,
"loss": 2.1656,
"step": 485
},
{
"epoch": 0.3318658990856756,
"grad_norm": 1.2421875,
"learning_rate": 8.450913842121396e-06,
"loss": 2.1694,
"step": 490
},
{
"epoch": 0.3352522858110396,
"grad_norm": 1.171875,
"learning_rate": 8.407876624394407e-06,
"loss": 2.1842,
"step": 495
},
{
"epoch": 0.3386386725364037,
"grad_norm": 1.2578125,
"learning_rate": 8.364362621864595e-06,
"loss": 2.1546,
"step": 500
},
{
"epoch": 0.3420250592617677,
"grad_norm": 1.2890625,
"learning_rate": 8.320377922432847e-06,
"loss": 2.1503,
"step": 505
},
{
"epoch": 0.34541144598713175,
"grad_norm": 1.21875,
"learning_rate": 8.275928679853704e-06,
"loss": 2.1594,
"step": 510
},
{
"epoch": 0.34879783271249576,
"grad_norm": 1.234375,
"learning_rate": 8.231021112874401e-06,
"loss": 2.1611,
"step": 515
},
{
"epoch": 0.3521842194378598,
"grad_norm": 1.25,
"learning_rate": 8.185661504364845e-06,
"loss": 2.1375,
"step": 520
},
{
"epoch": 0.35557060616322383,
"grad_norm": 1.28125,
"learning_rate": 8.139856200438574e-06,
"loss": 2.141,
"step": 525
},
{
"epoch": 0.3589569928885879,
"grad_norm": 1.234375,
"learning_rate": 8.093611609564913e-06,
"loss": 2.1684,
"step": 530
},
{
"epoch": 0.3623433796139519,
"grad_norm": 1.1875,
"learning_rate": 8.046934201672375e-06,
"loss": 2.1704,
"step": 535
},
{
"epoch": 0.365729766339316,
"grad_norm": 1.2421875,
"learning_rate": 7.999830507243478e-06,
"loss": 2.1675,
"step": 540
},
{
"epoch": 0.36911615306468,
"grad_norm": 1.265625,
"learning_rate": 7.952307116401086e-06,
"loss": 2.1924,
"step": 545
},
{
"epoch": 0.372502539790044,
"grad_norm": 1.2109375,
"learning_rate": 7.904370677986404e-06,
"loss": 2.1652,
"step": 550
},
{
"epoch": 0.37588892651540806,
"grad_norm": 1.2109375,
"learning_rate": 7.856027898628771e-06,
"loss": 2.1754,
"step": 555
},
{
"epoch": 0.37927531324077207,
"grad_norm": 1.2265625,
"learning_rate": 7.807285541807342e-06,
"loss": 2.177,
"step": 560
},
{
"epoch": 0.38266169996613614,
"grad_norm": 1.234375,
"learning_rate": 7.758150426904844e-06,
"loss": 2.1673,
"step": 565
},
{
"epoch": 0.38604808669150015,
"grad_norm": 1.1875,
"learning_rate": 7.708629428253497e-06,
"loss": 2.1743,
"step": 570
},
{
"epoch": 0.3894344734168642,
"grad_norm": 1.234375,
"learning_rate": 7.658729474173241e-06,
"loss": 2.1706,
"step": 575
},
{
"epoch": 0.3928208601422282,
"grad_norm": 1.28125,
"learning_rate": 7.608457546002423e-06,
"loss": 2.1593,
"step": 580
},
{
"epoch": 0.3962072468675923,
"grad_norm": 1.25,
"learning_rate": 7.5578206771210675e-06,
"loss": 2.1696,
"step": 585
},
{
"epoch": 0.3995936335929563,
"grad_norm": 1.21875,
"learning_rate": 7.506825951966843e-06,
"loss": 2.1469,
"step": 590
},
{
"epoch": 0.40298002031832036,
"grad_norm": 1.234375,
"learning_rate": 7.45548050504392e-06,
"loss": 2.1627,
"step": 595
},
{
"epoch": 0.4063664070436844,
"grad_norm": 1.2578125,
"learning_rate": 7.403791519924794e-06,
"loss": 2.1599,
"step": 600
},
{
"epoch": 0.40975279376904844,
"grad_norm": 1.2421875,
"learning_rate": 7.351766228245259e-06,
"loss": 2.172,
"step": 605
},
{
"epoch": 0.41313918049441245,
"grad_norm": 1.21875,
"learning_rate": 7.29941190869265e-06,
"loss": 2.1532,
"step": 610
},
{
"epoch": 0.4165255672197765,
"grad_norm": 1.203125,
"learning_rate": 7.246735885987515e-06,
"loss": 2.1545,
"step": 615
},
{
"epoch": 0.4199119539451405,
"grad_norm": 1.234375,
"learning_rate": 7.193745529858827e-06,
"loss": 2.1588,
"step": 620
},
{
"epoch": 0.4232983406705046,
"grad_norm": 1.21875,
"learning_rate": 7.140448254012912e-06,
"loss": 2.1717,
"step": 625
},
{
"epoch": 0.4266847273958686,
"grad_norm": 1.2265625,
"learning_rate": 7.086851515096233e-06,
"loss": 2.1679,
"step": 630
},
{
"epoch": 0.43007111412123267,
"grad_norm": 1.2578125,
"learning_rate": 7.032962811652133e-06,
"loss": 2.1703,
"step": 635
},
{
"epoch": 0.4334575008465967,
"grad_norm": 1.2734375,
"learning_rate": 6.978789683071761e-06,
"loss": 2.1786,
"step": 640
},
{
"epoch": 0.43684388757196074,
"grad_norm": 1.1953125,
"learning_rate": 6.924339708539244e-06,
"loss": 2.1673,
"step": 645
},
{
"epoch": 0.44023027429732475,
"grad_norm": 1.234375,
"learning_rate": 6.86962050597132e-06,
"loss": 2.152,
"step": 650
},
{
"epoch": 0.4436166610226888,
"grad_norm": 1.234375,
"learning_rate": 6.814639730951532e-06,
"loss": 2.1769,
"step": 655
},
{
"epoch": 0.44700304774805283,
"grad_norm": 1.234375,
"learning_rate": 6.759405075659165e-06,
"loss": 2.1577,
"step": 660
},
{
"epoch": 0.45038943447341684,
"grad_norm": 1.1953125,
"learning_rate": 6.703924267793061e-06,
"loss": 2.1549,
"step": 665
},
{
"epoch": 0.4537758211987809,
"grad_norm": 1.3046875,
"learning_rate": 6.648205069490451e-06,
"loss": 2.1482,
"step": 670
},
{
"epoch": 0.4571622079241449,
"grad_norm": 1.2421875,
"learning_rate": 6.592255276240994e-06,
"loss": 2.1745,
"step": 675
},
{
"epoch": 0.460548594649509,
"grad_norm": 1.2265625,
"learning_rate": 6.536082715796125e-06,
"loss": 2.1522,
"step": 680
},
{
"epoch": 0.463934981374873,
"grad_norm": 1.1875,
"learning_rate": 6.479695247073907e-06,
"loss": 2.1512,
"step": 685
},
{
"epoch": 0.46732136810023706,
"grad_norm": 1.21875,
"learning_rate": 6.4231007590595094e-06,
"loss": 2.1681,
"step": 690
},
{
"epoch": 0.47070775482560107,
"grad_norm": 1.2109375,
"learning_rate": 6.366307169701495e-06,
"loss": 2.1872,
"step": 695
},
{
"epoch": 0.47409414155096513,
"grad_norm": 1.21875,
"learning_rate": 6.309322424804034e-06,
"loss": 2.1545,
"step": 700
},
{
"epoch": 0.47748052827632914,
"grad_norm": 1.2265625,
"learning_rate": 6.252154496915244e-06,
"loss": 2.1832,
"step": 705
},
{
"epoch": 0.4808669150016932,
"grad_norm": 1.265625,
"learning_rate": 6.194811384211769e-06,
"loss": 2.1544,
"step": 710
},
{
"epoch": 0.4842533017270572,
"grad_norm": 1.1953125,
"learning_rate": 6.137301109379784e-06,
"loss": 2.164,
"step": 715
},
{
"epoch": 0.4876396884524213,
"grad_norm": 1.2265625,
"learning_rate": 6.079631718492569e-06,
"loss": 2.1528,
"step": 720
},
{
"epoch": 0.4910260751777853,
"grad_norm": 1.234375,
"learning_rate": 6.021811279884807e-06,
"loss": 2.1659,
"step": 725
},
{
"epoch": 0.49441246190314936,
"grad_norm": 1.25,
"learning_rate": 5.96384788302377e-06,
"loss": 2.1603,
"step": 730
},
{
"epoch": 0.49779884862851337,
"grad_norm": 1.3046875,
"learning_rate": 5.905749637377549e-06,
"loss": 2.1584,
"step": 735
},
{
"epoch": 0.5011852353538774,
"grad_norm": 1.2421875,
"learning_rate": 5.8475246712804845e-06,
"loss": 2.1734,
"step": 740
},
{
"epoch": 0.5045716220792414,
"grad_norm": 1.21875,
"learning_rate": 5.789181130795958e-06,
"loss": 2.1692,
"step": 745
},
{
"epoch": 0.5079580088046055,
"grad_norm": 1.234375,
"learning_rate": 5.730727178576703e-06,
"loss": 2.1903,
"step": 750
},
{
"epoch": 0.5113443955299696,
"grad_norm": 1.1875,
"learning_rate": 5.672170992722797e-06,
"loss": 2.1359,
"step": 755
},
{
"epoch": 0.5147307822553335,
"grad_norm": 1.234375,
"learning_rate": 5.613520765637489e-06,
"loss": 2.1573,
"step": 760
},
{
"epoch": 0.5181171689806976,
"grad_norm": 1.2109375,
"learning_rate": 5.554784702881026e-06,
"loss": 2.161,
"step": 765
},
{
"epoch": 0.5215035557060617,
"grad_norm": 1.2265625,
"learning_rate": 5.495971022022638e-06,
"loss": 2.171,
"step": 770
},
{
"epoch": 0.5248899424314256,
"grad_norm": 1.2421875,
"learning_rate": 5.437087951490856e-06,
"loss": 2.153,
"step": 775
},
{
"epoch": 0.5282763291567897,
"grad_norm": 1.2421875,
"learning_rate": 5.378143729422285e-06,
"loss": 2.1706,
"step": 780
},
{
"epoch": 0.5316627158821537,
"grad_norm": 1.203125,
"learning_rate": 5.319146602509042e-06,
"loss": 2.1715,
"step": 785
},
{
"epoch": 0.5350491026075178,
"grad_norm": 1.2109375,
"learning_rate": 5.2601048248449894e-06,
"loss": 2.1625,
"step": 790
},
{
"epoch": 0.5384354893328818,
"grad_norm": 1.1953125,
"learning_rate": 5.201026656770927e-06,
"loss": 2.152,
"step": 795
},
{
"epoch": 0.5418218760582458,
"grad_norm": 1.2265625,
"learning_rate": 5.141920363718916e-06,
"loss": 2.1642,
"step": 800
},
{
"epoch": 0.5452082627836099,
"grad_norm": 1.25,
"learning_rate": 5.082794215055894e-06,
"loss": 2.1624,
"step": 805
},
{
"epoch": 0.548594649508974,
"grad_norm": 1.21875,
"learning_rate": 5.023656482926728e-06,
"loss": 2.1619,
"step": 810
},
{
"epoch": 0.5519810362343379,
"grad_norm": 1.2421875,
"learning_rate": 4.96451544109689e-06,
"loss": 2.1742,
"step": 815
},
{
"epoch": 0.555367422959702,
"grad_norm": 1.2265625,
"learning_rate": 4.905379363794907e-06,
"loss": 2.1768,
"step": 820
},
{
"epoch": 0.558753809685066,
"grad_norm": 1.265625,
"learning_rate": 4.846256524554725e-06,
"loss": 2.1551,
"step": 825
},
{
"epoch": 0.5621401964104301,
"grad_norm": 1.2421875,
"learning_rate": 4.7871551950582105e-06,
"loss": 2.1588,
"step": 830
},
{
"epoch": 0.5655265831357941,
"grad_norm": 1.2578125,
"learning_rate": 4.728083643977855e-06,
"loss": 2.1643,
"step": 835
},
{
"epoch": 0.5689129698611581,
"grad_norm": 1.21875,
"learning_rate": 4.669050135819966e-06,
"loss": 2.153,
"step": 840
},
{
"epoch": 0.5722993565865222,
"grad_norm": 1.21875,
"learning_rate": 4.610062929768383e-06,
"loss": 2.1578,
"step": 845
},
{
"epoch": 0.5756857433118863,
"grad_norm": 1.2265625,
"learning_rate": 4.551130278528968e-06,
"loss": 2.1646,
"step": 850
},
{
"epoch": 0.5790721300372502,
"grad_norm": 1.1953125,
"learning_rate": 4.492260427175007e-06,
"loss": 2.1355,
"step": 855
},
{
"epoch": 0.5824585167626143,
"grad_norm": 1.25,
"learning_rate": 4.4334616119936516e-06,
"loss": 2.1638,
"step": 860
},
{
"epoch": 0.5858449034879784,
"grad_norm": 1.2578125,
"learning_rate": 4.374742059333621e-06,
"loss": 2.1568,
"step": 865
},
{
"epoch": 0.5892312902133424,
"grad_norm": 1.2578125,
"learning_rate": 4.316109984454278e-06,
"loss": 2.1672,
"step": 870
},
{
"epoch": 0.5926176769387064,
"grad_norm": 1.234375,
"learning_rate": 4.257573590376252e-06,
"loss": 2.1671,
"step": 875
},
{
"epoch": 0.5960040636640704,
"grad_norm": 1.25,
"learning_rate": 4.1991410667337896e-06,
"loss": 2.1578,
"step": 880
},
{
"epoch": 0.5993904503894345,
"grad_norm": 1.1953125,
"learning_rate": 4.140820588628964e-06,
"loss": 2.1538,
"step": 885
},
{
"epoch": 0.6027768371147985,
"grad_norm": 1.2109375,
"learning_rate": 4.082620315487931e-06,
"loss": 2.1407,
"step": 890
},
{
"epoch": 0.6061632238401625,
"grad_norm": 1.25,
"learning_rate": 4.02454838991936e-06,
"loss": 2.194,
"step": 895
},
{
"epoch": 0.6095496105655266,
"grad_norm": 1.2265625,
"learning_rate": 3.966612936575235e-06,
"loss": 2.1406,
"step": 900
},
{
"epoch": 0.6129359972908907,
"grad_norm": 1.21875,
"learning_rate": 3.908822061014165e-06,
"loss": 2.157,
"step": 905
},
{
"epoch": 0.6163223840162546,
"grad_norm": 1.1953125,
"learning_rate": 3.851183848567351e-06,
"loss": 2.1663,
"step": 910
},
{
"epoch": 0.6197087707416187,
"grad_norm": 1.203125,
"learning_rate": 3.7937063632074037e-06,
"loss": 2.1706,
"step": 915
},
{
"epoch": 0.6230951574669827,
"grad_norm": 1.2109375,
"learning_rate": 3.7363976464201348e-06,
"loss": 2.1575,
"step": 920
},
{
"epoch": 0.6264815441923468,
"grad_norm": 1.203125,
"learning_rate": 3.679265716079501e-06,
"loss": 2.1488,
"step": 925
},
{
"epoch": 0.6298679309177108,
"grad_norm": 1.21875,
"learning_rate": 3.6223185653258467e-06,
"loss": 2.1537,
"step": 930
},
{
"epoch": 0.6332543176430748,
"grad_norm": 1.2109375,
"learning_rate": 3.5655641614476172e-06,
"loss": 2.164,
"step": 935
},
{
"epoch": 0.6366407043684389,
"grad_norm": 1.25,
"learning_rate": 3.509010444766674e-06,
"loss": 2.1682,
"step": 940
},
{
"epoch": 0.640027091093803,
"grad_norm": 1.1953125,
"learning_rate": 3.452665327527391e-06,
"loss": 2.1675,
"step": 945
},
{
"epoch": 0.6434134778191669,
"grad_norm": 1.234375,
"learning_rate": 3.3965366927896864e-06,
"loss": 2.1635,
"step": 950
},
{
"epoch": 0.646799864544531,
"grad_norm": 1.2734375,
"learning_rate": 3.340632393326118e-06,
"loss": 2.1483,
"step": 955
},
{
"epoch": 0.650186251269895,
"grad_norm": 1.2421875,
"learning_rate": 3.284960250523237e-06,
"loss": 2.1621,
"step": 960
},
{
"epoch": 0.6535726379952591,
"grad_norm": 1.234375,
"learning_rate": 3.229528053287323e-06,
"loss": 2.1462,
"step": 965
},
{
"epoch": 0.6569590247206231,
"grad_norm": 1.2265625,
"learning_rate": 3.174343556954652e-06,
"loss": 2.1514,
"step": 970
},
{
"epoch": 0.6603454114459871,
"grad_norm": 1.25,
"learning_rate": 3.1194144822064943e-06,
"loss": 2.1728,
"step": 975
},
{
"epoch": 0.6637317981713512,
"grad_norm": 1.2578125,
"learning_rate": 3.0647485139889145e-06,
"loss": 2.1422,
"step": 980
},
{
"epoch": 0.6671181848967151,
"grad_norm": 1.1953125,
"learning_rate": 3.0103533004376183e-06,
"loss": 2.1335,
"step": 985
},
{
"epoch": 0.6705045716220792,
"grad_norm": 1.1953125,
"learning_rate": 2.9562364518079106e-06,
"loss": 2.1755,
"step": 990
},
{
"epoch": 0.6738909583474433,
"grad_norm": 1.2265625,
"learning_rate": 2.902405539409978e-06,
"loss": 2.1664,
"step": 995
},
{
"epoch": 0.6772773450728073,
"grad_norm": 1.25,
"learning_rate": 2.848868094549615e-06,
"loss": 2.1624,
"step": 1000
},
{
"epoch": 0.6806637317981713,
"grad_norm": 1.234375,
"learning_rate": 2.7956316074745294e-06,
"loss": 2.1761,
"step": 1005
},
{
"epoch": 0.6840501185235354,
"grad_norm": 1.21875,
"learning_rate": 2.7427035263264222e-06,
"loss": 2.1679,
"step": 1010
},
{
"epoch": 0.6874365052488994,
"grad_norm": 1.21875,
"learning_rate": 2.6900912560989363e-06,
"loss": 2.1499,
"step": 1015
},
{
"epoch": 0.6908228919742635,
"grad_norm": 1.2421875,
"learning_rate": 2.6378021576016467e-06,
"loss": 2.1401,
"step": 1020
},
{
"epoch": 0.6942092786996275,
"grad_norm": 1.2265625,
"learning_rate": 2.585843546430232e-06,
"loss": 2.1699,
"step": 1025
},
{
"epoch": 0.6975956654249915,
"grad_norm": 1.21875,
"learning_rate": 2.5342226919429807e-06,
"loss": 2.1637,
"step": 1030
},
{
"epoch": 0.7009820521503556,
"grad_norm": 1.234375,
"learning_rate": 2.4829468162437553e-06,
"loss": 2.1638,
"step": 1035
},
{
"epoch": 0.7043684388757196,
"grad_norm": 1.21875,
"learning_rate": 2.43202309317157e-06,
"loss": 2.1662,
"step": 1040
},
{
"epoch": 0.7077548256010836,
"grad_norm": 1.2109375,
"learning_rate": 2.3814586472969253e-06,
"loss": 2.1535,
"step": 1045
},
{
"epoch": 0.7111412123264477,
"grad_norm": 1.2265625,
"learning_rate": 2.331260552925028e-06,
"loss": 2.1592,
"step": 1050
},
{
"epoch": 0.7145275990518117,
"grad_norm": 1.2109375,
"learning_rate": 2.2814358331060533e-06,
"loss": 2.1723,
"step": 1055
},
{
"epoch": 0.7179139857771758,
"grad_norm": 1.2109375,
"learning_rate": 2.2319914586525776e-06,
"loss": 2.1504,
"step": 1060
},
{
"epoch": 0.7213003725025398,
"grad_norm": 1.2421875,
"learning_rate": 2.1829343471642995e-06,
"loss": 2.1636,
"step": 1065
},
{
"epoch": 0.7246867592279038,
"grad_norm": 1.234375,
"learning_rate": 2.1342713620602377e-06,
"loss": 2.1696,
"step": 1070
},
{
"epoch": 0.7280731459532679,
"grad_norm": 1.2578125,
"learning_rate": 2.0860093116184798e-06,
"loss": 2.1625,
"step": 1075
},
{
"epoch": 0.731459532678632,
"grad_norm": 1.2734375,
"learning_rate": 2.0381549480236685e-06,
"loss": 2.1559,
"step": 1080
},
{
"epoch": 0.7348459194039959,
"grad_norm": 1.2421875,
"learning_rate": 1.9907149664223206e-06,
"loss": 2.1921,
"step": 1085
},
{
"epoch": 0.73823230612936,
"grad_norm": 1.28125,
"learning_rate": 1.9436960039861324e-06,
"loss": 2.1845,
"step": 1090
},
{
"epoch": 0.741618692854724,
"grad_norm": 1.2421875,
"learning_rate": 1.8971046389833952e-06,
"loss": 2.1518,
"step": 1095
},
{
"epoch": 0.745005079580088,
"grad_norm": 1.2421875,
"learning_rate": 1.8509473898586432e-06,
"loss": 2.1557,
"step": 1100
},
{
"epoch": 0.7483914663054521,
"grad_norm": 1.2109375,
"learning_rate": 1.805230714320701e-06,
"loss": 2.1408,
"step": 1105
},
{
"epoch": 0.7517778530308161,
"grad_norm": 1.2421875,
"learning_rate": 1.7599610084391782e-06,
"loss": 2.1408,
"step": 1110
},
{
"epoch": 0.7551642397561802,
"grad_norm": 1.2578125,
"learning_rate": 1.7151446057496407e-06,
"loss": 2.1682,
"step": 1115
},
{
"epoch": 0.7585506264815441,
"grad_norm": 1.2578125,
"learning_rate": 1.670787776367489e-06,
"loss": 2.1714,
"step": 1120
},
{
"epoch": 0.7619370132069082,
"grad_norm": 1.25,
"learning_rate": 1.6268967261107426e-06,
"loss": 2.1632,
"step": 1125
},
{
"epoch": 0.7653233999322723,
"grad_norm": 1.25,
"learning_rate": 1.5834775956317937e-06,
"loss": 2.1885,
"step": 1130
},
{
"epoch": 0.7687097866576363,
"grad_norm": 1.1875,
"learning_rate": 1.5405364595582861e-06,
"loss": 2.1512,
"step": 1135
},
{
"epoch": 0.7720961733830003,
"grad_norm": 1.21875,
"learning_rate": 1.4980793256432474e-06,
"loss": 2.1597,
"step": 1140
},
{
"epoch": 0.7754825601083644,
"grad_norm": 1.75,
"learning_rate": 1.4561121339245487e-06,
"loss": 2.146,
"step": 1145
},
{
"epoch": 0.7788689468337284,
"grad_norm": 1.2734375,
"learning_rate": 1.4146407558938695e-06,
"loss": 2.1775,
"step": 1150
},
{
"epoch": 0.7822553335590925,
"grad_norm": 1.21875,
"learning_rate": 1.3736709936752195e-06,
"loss": 2.1759,
"step": 1155
},
{
"epoch": 0.7856417202844564,
"grad_norm": 1.2109375,
"learning_rate": 1.3332085792131966e-06,
"loss": 2.1523,
"step": 1160
},
{
"epoch": 0.7890281070098205,
"grad_norm": 1.2578125,
"learning_rate": 1.293259173471041e-06,
"loss": 2.1617,
"step": 1165
},
{
"epoch": 0.7924144937351846,
"grad_norm": 1.25,
"learning_rate": 1.2538283656386318e-06,
"loss": 2.1611,
"step": 1170
},
{
"epoch": 0.7958008804605486,
"grad_norm": 1.28125,
"learning_rate": 1.2149216723505247e-06,
"loss": 2.1594,
"step": 1175
},
{
"epoch": 0.7991872671859126,
"grad_norm": 1.2578125,
"learning_rate": 1.1765445369141276e-06,
"loss": 2.1871,
"step": 1180
},
{
"epoch": 0.8025736539112767,
"grad_norm": 1.2265625,
"learning_rate": 1.1387023285481573e-06,
"loss": 2.1724,
"step": 1185
},
{
"epoch": 0.8059600406366407,
"grad_norm": 1.2734375,
"learning_rate": 1.101400341631444e-06,
"loss": 2.1572,
"step": 1190
},
{
"epoch": 0.8093464273620048,
"grad_norm": 1.2421875,
"learning_rate": 1.0646437949622118e-06,
"loss": 2.1488,
"step": 1195
},
{
"epoch": 0.8127328140873687,
"grad_norm": 1.2109375,
"learning_rate": 1.028437831027937e-06,
"loss": 2.1736,
"step": 1200
},
{
"epoch": 0.8161192008127328,
"grad_norm": 1.1875,
"learning_rate": 9.92787515285873e-07,
"loss": 2.1676,
"step": 1205
},
{
"epoch": 0.8195055875380969,
"grad_norm": 1.2265625,
"learning_rate": 9.57697835454367e-07,
"loss": 2.173,
"step": 1210
},
{
"epoch": 0.8228919742634608,
"grad_norm": 1.25,
"learning_rate": 9.231737008150416e-07,
"loss": 2.1729,
"step": 1215
},
{
"epoch": 0.8262783609888249,
"grad_norm": 1.1875,
"learning_rate": 8.892199415259501e-07,
"loss": 2.1597,
"step": 1220
},
{
"epoch": 0.829664747714189,
"grad_norm": 1.234375,
"learning_rate": 8.558413079458106e-07,
"loss": 2.1688,
"step": 1225
},
{
"epoch": 0.833051134439553,
"grad_norm": 1.234375,
"learning_rate": 8.230424699693923e-07,
"loss": 2.1591,
"step": 1230
},
{
"epoch": 0.836437521164917,
"grad_norm": 1.234375,
"learning_rate": 7.908280163741733e-07,
"loss": 2.1761,
"step": 1235
},
{
"epoch": 0.839823907890281,
"grad_norm": 1.21875,
"learning_rate": 7.592024541783344e-07,
"loss": 2.1578,
"step": 1240
},
{
"epoch": 0.8432102946156451,
"grad_norm": 1.2109375,
"learning_rate": 7.281702080102004e-07,
"loss": 2.1533,
"step": 1245
},
{
"epoch": 0.8465966813410092,
"grad_norm": 1.1953125,
"learning_rate": 6.977356194891999e-07,
"loss": 2.1389,
"step": 1250
},
{
"epoch": 0.8499830680663731,
"grad_norm": 1.234375,
"learning_rate": 6.679029466184506e-07,
"loss": 2.1738,
"step": 1255
},
{
"epoch": 0.8533694547917372,
"grad_norm": 1.265625,
"learning_rate": 6.386763631890313e-07,
"loss": 2.1579,
"step": 1260
},
{
"epoch": 0.8567558415171013,
"grad_norm": 1.203125,
"learning_rate": 6.100599581960414e-07,
"loss": 2.1672,
"step": 1265
},
{
"epoch": 0.8601422282424653,
"grad_norm": 1.2578125,
"learning_rate": 5.820577352665252e-07,
"loss": 2.1668,
"step": 1270
},
{
"epoch": 0.8635286149678293,
"grad_norm": 1.21875,
"learning_rate": 5.546736120993318e-07,
"loss": 2.1421,
"step": 1275
},
{
"epoch": 0.8669150016931934,
"grad_norm": 1.25,
"learning_rate": 5.279114199170094e-07,
"loss": 2.1483,
"step": 1280
},
{
"epoch": 0.8703013884185574,
"grad_norm": 1.21875,
"learning_rate": 5.017749029297919e-07,
"loss": 2.1677,
"step": 1285
},
{
"epoch": 0.8736877751439215,
"grad_norm": 1.2421875,
"learning_rate": 4.762677178117503e-07,
"loss": 2.1594,
"step": 1290
},
{
"epoch": 0.8770741618692854,
"grad_norm": 1.2578125,
"learning_rate": 4.5139343318920946e-07,
"loss": 2.1631,
"step": 1295
},
{
"epoch": 0.8804605485946495,
"grad_norm": 1.203125,
"learning_rate": 4.271555291414636e-07,
"loss": 2.154,
"step": 1300
},
{
"epoch": 0.8838469353200136,
"grad_norm": 1.2421875,
"learning_rate": 4.0355739671390226e-07,
"loss": 2.1637,
"step": 1305
},
{
"epoch": 0.8872333220453776,
"grad_norm": 1.25,
"learning_rate": 3.8060233744356634e-07,
"loss": 2.1729,
"step": 1310
},
{
"epoch": 0.8906197087707416,
"grad_norm": 1.25,
"learning_rate": 3.582935628972523e-07,
"loss": 2.1615,
"step": 1315
},
{
"epoch": 0.8940060954961057,
"grad_norm": 1.2109375,
"learning_rate": 3.366341942221868e-07,
"loss": 2.1647,
"step": 1320
},
{
"epoch": 0.8973924822214697,
"grad_norm": 1.1875,
"learning_rate": 3.156272617093553e-07,
"loss": 2.1601,
"step": 1325
},
{
"epoch": 0.9007788689468337,
"grad_norm": 1.2421875,
"learning_rate": 2.9527570436955255e-07,
"loss": 2.1758,
"step": 1330
},
{
"epoch": 0.9041652556721977,
"grad_norm": 1.1796875,
"learning_rate": 2.7558236952218485e-07,
"loss": 2.1537,
"step": 1335
},
{
"epoch": 0.9075516423975618,
"grad_norm": 1.2421875,
"learning_rate": 2.5655001239691836e-07,
"loss": 2.1707,
"step": 1340
},
{
"epoch": 0.9109380291229259,
"grad_norm": 1.2109375,
"learning_rate": 2.3818129574819992e-07,
"loss": 2.1601,
"step": 1345
},
{
"epoch": 0.9143244158482898,
"grad_norm": 1.234375,
"learning_rate": 2.2047878948272373e-07,
"loss": 2.1648,
"step": 1350
},
{
"epoch": 0.9177108025736539,
"grad_norm": 1.1953125,
"learning_rate": 2.0344497029988087e-07,
"loss": 2.1615,
"step": 1355
},
{
"epoch": 0.921097189299018,
"grad_norm": 1.234375,
"learning_rate": 1.8708222134525168e-07,
"loss": 2.1707,
"step": 1360
},
{
"epoch": 0.924483576024382,
"grad_norm": 1.2421875,
"learning_rate": 1.7139283187719124e-07,
"loss": 2.175,
"step": 1365
},
{
"epoch": 0.927869962749746,
"grad_norm": 1.234375,
"learning_rate": 1.5637899694654456e-07,
"loss": 2.1695,
"step": 1370
},
{
"epoch": 0.93125634947511,
"grad_norm": 1.25,
"learning_rate": 1.4204281708954437e-07,
"loss": 2.1624,
"step": 1375
},
{
"epoch": 0.9346427362004741,
"grad_norm": 1.234375,
"learning_rate": 1.2838629803393343e-07,
"loss": 2.1503,
"step": 1380
},
{
"epoch": 0.9380291229258382,
"grad_norm": 1.2578125,
"learning_rate": 1.1541135041834628e-07,
"loss": 2.1846,
"step": 1385
},
{
"epoch": 0.9414155096512021,
"grad_norm": 1.2265625,
"learning_rate": 1.0311978952500301e-07,
"loss": 2.163,
"step": 1390
},
{
"epoch": 0.9448018963765662,
"grad_norm": 1.21875,
"learning_rate": 9.151333502573468e-08,
"loss": 2.1635,
"step": 1395
},
{
"epoch": 0.9481882831019303,
"grad_norm": 1.2578125,
"learning_rate": 8.059361074139293e-08,
"loss": 2.1687,
"step": 1400
},
{
"epoch": 0.9515746698272943,
"grad_norm": 1.2265625,
"learning_rate": 7.03621444146635e-08,
"loss": 2.1606,
"step": 1405
},
{
"epoch": 0.9549610565526583,
"grad_norm": 1.2265625,
"learning_rate": 6.082036749632702e-08,
"loss": 2.1683,
"step": 1410
},
{
"epoch": 0.9583474432780223,
"grad_norm": 1.21875,
"learning_rate": 5.1969614944989976e-08,
"loss": 2.1599,
"step": 1415
},
{
"epoch": 0.9617338300033864,
"grad_norm": 1.21875,
"learning_rate": 4.381112504031337e-08,
"loss": 2.1777,
"step": 1420
},
{
"epoch": 0.9651202167287504,
"grad_norm": 1.203125,
"learning_rate": 3.634603920976809e-08,
"loss": 2.1465,
"step": 1425
},
{
"epoch": 0.9685066034541144,
"grad_norm": 1.2109375,
"learning_rate": 2.957540186894481e-08,
"loss": 2.1506,
"step": 1430
},
{
"epoch": 0.9718929901794785,
"grad_norm": 1.3359375,
"learning_rate": 2.3500160275430895e-08,
"loss": 2.1696,
"step": 1435
},
{
"epoch": 0.9752793769048426,
"grad_norm": 1.25,
"learning_rate": 1.8121164396283643e-08,
"loss": 2.1463,
"step": 1440
},
{
"epoch": 0.9786657636302065,
"grad_norm": 1.1953125,
"learning_rate": 1.3439166789113723e-08,
"loss": 2.1605,
"step": 1445
},
{
"epoch": 0.9820521503555706,
"grad_norm": 1.2109375,
"learning_rate": 9.454822496796635e-09,
"loss": 2.1646,
"step": 1450
},
{
"epoch": 0.9854385370809347,
"grad_norm": 1.2265625,
"learning_rate": 6.1686889558304485e-09,
"loss": 2.1537,
"step": 1455
},
{
"epoch": 0.9888249238062987,
"grad_norm": 1.21875,
"learning_rate": 3.5812259183426457e-09,
"loss": 2.1711,
"step": 1460
},
{
"epoch": 0.9922113105316627,
"grad_norm": 1.1953125,
"learning_rate": 1.692795387772117e-09,
"loss": 2.159,
"step": 1465
},
{
"epoch": 0.9955976972570267,
"grad_norm": 1.2265625,
"learning_rate": 5.03661568219127e-10,
"loss": 2.1677,
"step": 1470
},
{
"epoch": 0.9989840839823908,
"grad_norm": 1.203125,
"learning_rate": 1.3990827483212343e-11,
"loss": 2.1562,
"step": 1475
},
{
"epoch": 0.9996613613274636,
"eval_loss": 2.1607718467712402,
"eval_runtime": 85.3315,
"eval_samples_per_second": 15.293,
"eval_steps_per_second": 1.922,
"step": 1476
},
{
"epoch": 0.9996613613274636,
"step": 1476,
"total_flos": 3.002086743311647e+17,
"train_loss": 2.1796100780892824,
"train_runtime": 6613.0875,
"train_samples_per_second": 3.572,
"train_steps_per_second": 0.223
}
],
"logging_steps": 5,
"max_steps": 1476,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.002086743311647e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}