flyingbugs's picture
Model save
278f54f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1611,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00931098696461825,
"grad_norm": 16.336199791449793,
"learning_rate": 2.469135802469136e-06,
"loss": 1.9327,
"step": 5
},
{
"epoch": 0.0186219739292365,
"grad_norm": 9.22734324139331,
"learning_rate": 5.555555555555556e-06,
"loss": 1.6983,
"step": 10
},
{
"epoch": 0.027932960893854747,
"grad_norm": 2.4572991614653783,
"learning_rate": 8.641975308641975e-06,
"loss": 1.1843,
"step": 15
},
{
"epoch": 0.037243947858473,
"grad_norm": 0.8356668158228655,
"learning_rate": 1.1728395061728396e-05,
"loss": 0.9297,
"step": 20
},
{
"epoch": 0.04655493482309125,
"grad_norm": 0.4936127329444582,
"learning_rate": 1.4814814814814815e-05,
"loss": 0.8284,
"step": 25
},
{
"epoch": 0.055865921787709494,
"grad_norm": 0.3478722983428711,
"learning_rate": 1.7901234567901236e-05,
"loss": 0.7489,
"step": 30
},
{
"epoch": 0.06517690875232775,
"grad_norm": 0.25513496703130056,
"learning_rate": 2.0987654320987655e-05,
"loss": 0.7016,
"step": 35
},
{
"epoch": 0.074487895716946,
"grad_norm": 0.2403829000286689,
"learning_rate": 2.4074074074074074e-05,
"loss": 0.673,
"step": 40
},
{
"epoch": 0.08379888268156424,
"grad_norm": 0.38912791976591726,
"learning_rate": 2.7160493827160493e-05,
"loss": 0.6583,
"step": 45
},
{
"epoch": 0.0931098696461825,
"grad_norm": 0.1953193028806376,
"learning_rate": 3.0246913580246916e-05,
"loss": 0.6318,
"step": 50
},
{
"epoch": 0.10242085661080075,
"grad_norm": 0.17012378672445239,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.615,
"step": 55
},
{
"epoch": 0.11173184357541899,
"grad_norm": 0.19550839394534458,
"learning_rate": 3.6419753086419754e-05,
"loss": 0.6027,
"step": 60
},
{
"epoch": 0.12104283054003724,
"grad_norm": 0.1475381137651654,
"learning_rate": 3.950617283950617e-05,
"loss": 0.5917,
"step": 65
},
{
"epoch": 0.1303538175046555,
"grad_norm": 0.14017992287056805,
"learning_rate": 4.259259259259259e-05,
"loss": 0.5858,
"step": 70
},
{
"epoch": 0.13966480446927373,
"grad_norm": 0.17019572604849684,
"learning_rate": 4.567901234567901e-05,
"loss": 0.5905,
"step": 75
},
{
"epoch": 0.148975791433892,
"grad_norm": 0.17525567903422554,
"learning_rate": 4.876543209876544e-05,
"loss": 0.5734,
"step": 80
},
{
"epoch": 0.15828677839851024,
"grad_norm": 0.14195461289074676,
"learning_rate": 4.999957311534635e-05,
"loss": 0.561,
"step": 85
},
{
"epoch": 0.16759776536312848,
"grad_norm": 0.15882225996967175,
"learning_rate": 4.9996964434456464e-05,
"loss": 0.5562,
"step": 90
},
{
"epoch": 0.17690875232774675,
"grad_norm": 0.15508701484518203,
"learning_rate": 4.999198450544716e-05,
"loss": 0.5615,
"step": 95
},
{
"epoch": 0.186219739292365,
"grad_norm": 0.15282408576751363,
"learning_rate": 4.998463385321802e-05,
"loss": 0.5566,
"step": 100
},
{
"epoch": 0.19553072625698323,
"grad_norm": 0.1441039498919609,
"learning_rate": 4.997491325255003e-05,
"loss": 0.5504,
"step": 105
},
{
"epoch": 0.2048417132216015,
"grad_norm": 0.1414977005402374,
"learning_rate": 4.9962823728023915e-05,
"loss": 0.5419,
"step": 110
},
{
"epoch": 0.21415270018621974,
"grad_norm": 0.14874293035506805,
"learning_rate": 4.9948366553912146e-05,
"loss": 0.5307,
"step": 115
},
{
"epoch": 0.22346368715083798,
"grad_norm": 0.16062863369880248,
"learning_rate": 4.99315432540446e-05,
"loss": 0.534,
"step": 120
},
{
"epoch": 0.23277467411545624,
"grad_norm": 0.15097433303561048,
"learning_rate": 4.9912355601648e-05,
"loss": 0.5345,
"step": 125
},
{
"epoch": 0.24208566108007448,
"grad_norm": 0.17442330409969137,
"learning_rate": 4.9890805619158954e-05,
"loss": 0.5375,
"step": 130
},
{
"epoch": 0.25139664804469275,
"grad_norm": 0.21332022009684865,
"learning_rate": 4.9866895578010826e-05,
"loss": 0.5306,
"step": 135
},
{
"epoch": 0.260707635009311,
"grad_norm": 0.1652894070341978,
"learning_rate": 4.9840627998394285e-05,
"loss": 0.5318,
"step": 140
},
{
"epoch": 0.27001862197392923,
"grad_norm": 0.1637036108907819,
"learning_rate": 4.9812005648991715e-05,
"loss": 0.5324,
"step": 145
},
{
"epoch": 0.27932960893854747,
"grad_norm": 0.2654815231921947,
"learning_rate": 4.978103154668534e-05,
"loss": 0.5382,
"step": 150
},
{
"epoch": 0.2886405959031657,
"grad_norm": 0.1647627509019107,
"learning_rate": 4.9747708956239245e-05,
"loss": 0.5217,
"step": 155
},
{
"epoch": 0.297951582867784,
"grad_norm": 0.15907215366520921,
"learning_rate": 4.971204138995531e-05,
"loss": 0.5191,
"step": 160
},
{
"epoch": 0.30726256983240224,
"grad_norm": 0.15147466492817646,
"learning_rate": 4.967403260730291e-05,
"loss": 0.5134,
"step": 165
},
{
"epoch": 0.3165735567970205,
"grad_norm": 0.1405938594476101,
"learning_rate": 4.963368661452277e-05,
"loss": 0.5195,
"step": 170
},
{
"epoch": 0.3258845437616387,
"grad_norm": 0.1595878244498165,
"learning_rate": 4.959100766420458e-05,
"loss": 0.5232,
"step": 175
},
{
"epoch": 0.33519553072625696,
"grad_norm": 0.15951772828099983,
"learning_rate": 4.954600025483885e-05,
"loss": 0.5141,
"step": 180
},
{
"epoch": 0.34450651769087526,
"grad_norm": 0.16382396726117066,
"learning_rate": 4.9498669130342686e-05,
"loss": 0.516,
"step": 185
},
{
"epoch": 0.3538175046554935,
"grad_norm": 0.17573565863431748,
"learning_rate": 4.944901927955983e-05,
"loss": 0.5161,
"step": 190
},
{
"epoch": 0.36312849162011174,
"grad_norm": 0.1574125159399186,
"learning_rate": 4.939705593573477e-05,
"loss": 0.5185,
"step": 195
},
{
"epoch": 0.37243947858473,
"grad_norm": 0.15370896499046016,
"learning_rate": 4.9342784575961175e-05,
"loss": 0.5082,
"step": 200
},
{
"epoch": 0.3817504655493482,
"grad_norm": 0.14999931215152523,
"learning_rate": 4.928621092060457e-05,
"loss": 0.5177,
"step": 205
},
{
"epoch": 0.39106145251396646,
"grad_norm": 0.15559155877009215,
"learning_rate": 4.922734093269941e-05,
"loss": 0.5118,
"step": 210
},
{
"epoch": 0.40037243947858475,
"grad_norm": 0.14927855222097725,
"learning_rate": 4.916618081732053e-05,
"loss": 0.5123,
"step": 215
},
{
"epoch": 0.409683426443203,
"grad_norm": 0.17618505030960954,
"learning_rate": 4.9102737020929135e-05,
"loss": 0.5166,
"step": 220
},
{
"epoch": 0.41899441340782123,
"grad_norm": 0.15485079003135388,
"learning_rate": 4.903701623069334e-05,
"loss": 0.5031,
"step": 225
},
{
"epoch": 0.42830540037243947,
"grad_norm": 0.1516909848278948,
"learning_rate": 4.8969025373783266e-05,
"loss": 0.5025,
"step": 230
},
{
"epoch": 0.4376163873370577,
"grad_norm": 0.16374748063322728,
"learning_rate": 4.889877161664096e-05,
"loss": 0.5056,
"step": 235
},
{
"epoch": 0.44692737430167595,
"grad_norm": 0.20591850992077587,
"learning_rate": 4.882626236422498e-05,
"loss": 0.5083,
"step": 240
},
{
"epoch": 0.45623836126629425,
"grad_norm": 0.1556213381298468,
"learning_rate": 4.875150525922994e-05,
"loss": 0.5031,
"step": 245
},
{
"epoch": 0.4655493482309125,
"grad_norm": 0.16267277069927719,
"learning_rate": 4.8674508181280865e-05,
"loss": 0.5074,
"step": 250
},
{
"epoch": 0.4748603351955307,
"grad_norm": 0.15425490006295214,
"learning_rate": 4.8595279246102754e-05,
"loss": 0.4995,
"step": 255
},
{
"epoch": 0.48417132216014896,
"grad_norm": 0.13465996416398995,
"learning_rate": 4.85138268046651e-05,
"loss": 0.5046,
"step": 260
},
{
"epoch": 0.4934823091247672,
"grad_norm": 0.17570069468472505,
"learning_rate": 4.843015944230166e-05,
"loss": 0.5042,
"step": 265
},
{
"epoch": 0.5027932960893855,
"grad_norm": 0.15339985049013902,
"learning_rate": 4.8344285977805594e-05,
"loss": 0.4972,
"step": 270
},
{
"epoch": 0.5121042830540037,
"grad_norm": 0.14216267176397104,
"learning_rate": 4.8256215462499866e-05,
"loss": 0.4977,
"step": 275
},
{
"epoch": 0.521415270018622,
"grad_norm": 0.1442544392520683,
"learning_rate": 4.816595717928327e-05,
"loss": 0.5045,
"step": 280
},
{
"epoch": 0.5307262569832403,
"grad_norm": 0.14073656875745008,
"learning_rate": 4.807352064165195e-05,
"loss": 0.4899,
"step": 285
},
{
"epoch": 0.5400372439478585,
"grad_norm": 0.18295662717507738,
"learning_rate": 4.7978915592696644e-05,
"loss": 0.4968,
"step": 290
},
{
"epoch": 0.5493482309124768,
"grad_norm": 0.1804436580847151,
"learning_rate": 4.788215200407576e-05,
"loss": 0.4921,
"step": 295
},
{
"epoch": 0.5586592178770949,
"grad_norm": 0.14254600145362717,
"learning_rate": 4.7783240074964316e-05,
"loss": 0.4924,
"step": 300
},
{
"epoch": 0.5679702048417132,
"grad_norm": 0.1510375274736302,
"learning_rate": 4.76821902309789e-05,
"loss": 0.4969,
"step": 305
},
{
"epoch": 0.5772811918063314,
"grad_norm": 0.14322180938059342,
"learning_rate": 4.757901312307882e-05,
"loss": 0.4867,
"step": 310
},
{
"epoch": 0.5865921787709497,
"grad_norm": 0.15778761438406094,
"learning_rate": 4.747371962644344e-05,
"loss": 0.5062,
"step": 315
},
{
"epoch": 0.595903165735568,
"grad_norm": 0.15689388105498725,
"learning_rate": 4.7366320839325864e-05,
"loss": 0.4992,
"step": 320
},
{
"epoch": 0.6052141527001862,
"grad_norm": 0.14383076831723837,
"learning_rate": 4.7256828081883205e-05,
"loss": 0.4989,
"step": 325
},
{
"epoch": 0.6145251396648045,
"grad_norm": 0.21080496453629785,
"learning_rate": 4.71452528949834e-05,
"loss": 0.4898,
"step": 330
},
{
"epoch": 0.6238361266294227,
"grad_norm": 0.14254271015618503,
"learning_rate": 4.7031607038988735e-05,
"loss": 0.4802,
"step": 335
},
{
"epoch": 0.633147113594041,
"grad_norm": 0.15811041126836908,
"learning_rate": 4.691590249251627e-05,
"loss": 0.4907,
"step": 340
},
{
"epoch": 0.6424581005586593,
"grad_norm": 0.13753498439384687,
"learning_rate": 4.67981514511753e-05,
"loss": 0.494,
"step": 345
},
{
"epoch": 0.6517690875232774,
"grad_norm": 0.1381443189067191,
"learning_rate": 4.667836632628185e-05,
"loss": 0.4929,
"step": 350
},
{
"epoch": 0.6610800744878957,
"grad_norm": 0.14535188259873358,
"learning_rate": 4.655655974355051e-05,
"loss": 0.492,
"step": 355
},
{
"epoch": 0.6703910614525139,
"grad_norm": 0.16265320769415334,
"learning_rate": 4.643274454176365e-05,
"loss": 0.4854,
"step": 360
},
{
"epoch": 0.6797020484171322,
"grad_norm": 0.1426180701421254,
"learning_rate": 4.630693377141816e-05,
"loss": 0.4865,
"step": 365
},
{
"epoch": 0.6890130353817505,
"grad_norm": 0.1422570068389307,
"learning_rate": 4.6179140693349894e-05,
"loss": 0.4848,
"step": 370
},
{
"epoch": 0.6983240223463687,
"grad_norm": 0.14960885172487687,
"learning_rate": 4.6049378777335925e-05,
"loss": 0.478,
"step": 375
},
{
"epoch": 0.707635009310987,
"grad_norm": 0.1686075463208643,
"learning_rate": 4.591766170067484e-05,
"loss": 0.4856,
"step": 380
},
{
"epoch": 0.7169459962756052,
"grad_norm": 0.1696479397319911,
"learning_rate": 4.578400334674503e-05,
"loss": 0.4738,
"step": 385
},
{
"epoch": 0.7262569832402235,
"grad_norm": 0.12636437658397243,
"learning_rate": 4.5648417803541405e-05,
"loss": 0.4861,
"step": 390
},
{
"epoch": 0.7355679702048417,
"grad_norm": 0.14823499403718474,
"learning_rate": 4.5510919362190465e-05,
"loss": 0.4874,
"step": 395
},
{
"epoch": 0.74487895716946,
"grad_norm": 0.13876911181183,
"learning_rate": 4.5371522515443934e-05,
"loss": 0.481,
"step": 400
},
{
"epoch": 0.7541899441340782,
"grad_norm": 0.14555324320118293,
"learning_rate": 4.523024195615124e-05,
"loss": 0.4917,
"step": 405
},
{
"epoch": 0.7635009310986964,
"grad_norm": 0.1483421827835732,
"learning_rate": 4.508709257571081e-05,
"loss": 0.4818,
"step": 410
},
{
"epoch": 0.7728119180633147,
"grad_norm": 0.1419680809174011,
"learning_rate": 4.494208946250042e-05,
"loss": 0.4772,
"step": 415
},
{
"epoch": 0.7821229050279329,
"grad_norm": 0.16513221694643368,
"learning_rate": 4.4795247900286975e-05,
"loss": 0.4845,
"step": 420
},
{
"epoch": 0.7914338919925512,
"grad_norm": 0.14036938615466898,
"learning_rate": 4.46465833666154e-05,
"loss": 0.4904,
"step": 425
},
{
"epoch": 0.8007448789571695,
"grad_norm": 0.14632734868202635,
"learning_rate": 4.449611153117736e-05,
"loss": 0.4916,
"step": 430
},
{
"epoch": 0.8100558659217877,
"grad_norm": 0.20473257481525392,
"learning_rate": 4.4343848254159587e-05,
"loss": 0.4805,
"step": 435
},
{
"epoch": 0.819366852886406,
"grad_norm": 0.12821369541497668,
"learning_rate": 4.418980958457217e-05,
"loss": 0.4778,
"step": 440
},
{
"epoch": 0.8286778398510242,
"grad_norm": 0.14169275068446857,
"learning_rate": 4.4034011758556957e-05,
"loss": 0.4824,
"step": 445
},
{
"epoch": 0.8379888268156425,
"grad_norm": 0.13905298509358965,
"learning_rate": 4.387647119767619e-05,
"loss": 0.4879,
"step": 450
},
{
"epoch": 0.8472998137802608,
"grad_norm": 0.14408464936999882,
"learning_rate": 4.3717204507181665e-05,
"loss": 0.4826,
"step": 455
},
{
"epoch": 0.8566108007448789,
"grad_norm": 0.1408096706025395,
"learning_rate": 4.355622847426443e-05,
"loss": 0.481,
"step": 460
},
{
"epoch": 0.8659217877094972,
"grad_norm": 0.18158777593456274,
"learning_rate": 4.339356006628542e-05,
"loss": 0.4688,
"step": 465
},
{
"epoch": 0.8752327746741154,
"grad_norm": 0.12760922940847635,
"learning_rate": 4.322921642898701e-05,
"loss": 0.4802,
"step": 470
},
{
"epoch": 0.8845437616387337,
"grad_norm": 0.11771870402744473,
"learning_rate": 4.3063214884685835e-05,
"loss": 0.4856,
"step": 475
},
{
"epoch": 0.8938547486033519,
"grad_norm": 0.12319800101836006,
"learning_rate": 4.2895572930446895e-05,
"loss": 0.4864,
"step": 480
},
{
"epoch": 0.9031657355679702,
"grad_norm": 0.1266489948893168,
"learning_rate": 4.272630823623942e-05,
"loss": 0.4718,
"step": 485
},
{
"epoch": 0.9124767225325885,
"grad_norm": 0.1539105039445935,
"learning_rate": 4.255543864307431e-05,
"loss": 0.4852,
"step": 490
},
{
"epoch": 0.9217877094972067,
"grad_norm": 0.14332478118043573,
"learning_rate": 4.238298216112367e-05,
"loss": 0.4773,
"step": 495
},
{
"epoch": 0.931098696461825,
"grad_norm": 0.12617484788483693,
"learning_rate": 4.220895696782251e-05,
"loss": 0.4727,
"step": 500
},
{
"epoch": 0.9404096834264432,
"grad_norm": 0.1360094821692902,
"learning_rate": 4.2033381405952715e-05,
"loss": 0.4765,
"step": 505
},
{
"epoch": 0.9497206703910615,
"grad_norm": 0.1283299871000522,
"learning_rate": 4.1856273981709757e-05,
"loss": 0.4785,
"step": 510
},
{
"epoch": 0.9590316573556797,
"grad_norm": 0.1456300012973035,
"learning_rate": 4.167765336275201e-05,
"loss": 0.4836,
"step": 515
},
{
"epoch": 0.9683426443202979,
"grad_norm": 0.12631639520637056,
"learning_rate": 4.149753837623317e-05,
"loss": 0.4742,
"step": 520
},
{
"epoch": 0.9776536312849162,
"grad_norm": 0.12595076088478818,
"learning_rate": 4.1315948006817755e-05,
"loss": 0.4808,
"step": 525
},
{
"epoch": 0.9869646182495344,
"grad_norm": 0.12055722184602191,
"learning_rate": 4.113290139468015e-05,
"loss": 0.4725,
"step": 530
},
{
"epoch": 0.9962756052141527,
"grad_norm": 0.13737170916427302,
"learning_rate": 4.0948417833487104e-05,
"loss": 0.4771,
"step": 535
},
{
"epoch": 1.005586592178771,
"grad_norm": 0.1402038369271413,
"learning_rate": 4.0762516768364164e-05,
"loss": 0.4531,
"step": 540
},
{
"epoch": 1.0148975791433892,
"grad_norm": 0.1389386280489692,
"learning_rate": 4.0575217793846045e-05,
"loss": 0.4618,
"step": 545
},
{
"epoch": 1.0242085661080074,
"grad_norm": 0.1301676008479176,
"learning_rate": 4.038654065181137e-05,
"loss": 0.4444,
"step": 550
},
{
"epoch": 1.0335195530726258,
"grad_norm": 0.12917021441633983,
"learning_rate": 4.0196505229401785e-05,
"loss": 0.4478,
"step": 555
},
{
"epoch": 1.042830540037244,
"grad_norm": 0.12625994093795317,
"learning_rate": 4.0005131556925795e-05,
"loss": 0.447,
"step": 560
},
{
"epoch": 1.0521415270018621,
"grad_norm": 0.1407545892773896,
"learning_rate": 3.98124398057475e-05,
"loss": 0.4511,
"step": 565
},
{
"epoch": 1.0614525139664805,
"grad_norm": 0.1392139399442254,
"learning_rate": 3.961845028616052e-05,
"loss": 0.4581,
"step": 570
},
{
"epoch": 1.0707635009310987,
"grad_norm": 0.11692992722859218,
"learning_rate": 3.9423183445247117e-05,
"loss": 0.4466,
"step": 575
},
{
"epoch": 1.080074487895717,
"grad_norm": 0.11866718899104756,
"learning_rate": 3.9226659864723166e-05,
"loss": 0.4505,
"step": 580
},
{
"epoch": 1.089385474860335,
"grad_norm": 0.1301769972288239,
"learning_rate": 3.902890025876867e-05,
"loss": 0.4568,
"step": 585
},
{
"epoch": 1.0986964618249535,
"grad_norm": 0.12988708729398774,
"learning_rate": 3.8829925471844434e-05,
"loss": 0.4486,
"step": 590
},
{
"epoch": 1.1080074487895717,
"grad_norm": 0.14294084168875013,
"learning_rate": 3.8629756476495024e-05,
"loss": 0.4519,
"step": 595
},
{
"epoch": 1.1173184357541899,
"grad_norm": 0.12461053171632061,
"learning_rate": 3.842841437113821e-05,
"loss": 0.4551,
"step": 600
},
{
"epoch": 1.1266294227188083,
"grad_norm": 0.12220034734319962,
"learning_rate": 3.822592037784109e-05,
"loss": 0.4445,
"step": 605
},
{
"epoch": 1.1359404096834265,
"grad_norm": 0.11924325065263018,
"learning_rate": 3.802229584008321e-05,
"loss": 0.4493,
"step": 610
},
{
"epoch": 1.1452513966480447,
"grad_norm": 0.1549692686985166,
"learning_rate": 3.781756222050696e-05,
"loss": 0.4549,
"step": 615
},
{
"epoch": 1.1545623836126628,
"grad_norm": 0.11567955790271696,
"learning_rate": 3.761174109865529e-05,
"loss": 0.4484,
"step": 620
},
{
"epoch": 1.1638733705772812,
"grad_norm": 0.11742160828357712,
"learning_rate": 3.7404854168697215e-05,
"loss": 0.4567,
"step": 625
},
{
"epoch": 1.1731843575418994,
"grad_norm": 0.12900092793504658,
"learning_rate": 3.719692323714109e-05,
"loss": 0.4352,
"step": 630
},
{
"epoch": 1.1824953445065176,
"grad_norm": 0.14140700517521818,
"learning_rate": 3.6987970220536246e-05,
"loss": 0.4488,
"step": 635
},
{
"epoch": 1.191806331471136,
"grad_norm": 0.1293764570562897,
"learning_rate": 3.677801714316283e-05,
"loss": 0.4512,
"step": 640
},
{
"epoch": 1.2011173184357542,
"grad_norm": 0.1312966218053429,
"learning_rate": 3.6567086134710466e-05,
"loss": 0.4484,
"step": 645
},
{
"epoch": 1.2104283054003724,
"grad_norm": 0.11960370887697533,
"learning_rate": 3.6355199427945634e-05,
"loss": 0.4499,
"step": 650
},
{
"epoch": 1.2197392923649906,
"grad_norm": 0.12886802374746667,
"learning_rate": 3.6142379356368334e-05,
"loss": 0.4508,
"step": 655
},
{
"epoch": 1.229050279329609,
"grad_norm": 0.12443137480159507,
"learning_rate": 3.5928648351858027e-05,
"loss": 0.4531,
"step": 660
},
{
"epoch": 1.2383612662942272,
"grad_norm": 0.1235305037487046,
"learning_rate": 3.571402894230926e-05,
"loss": 0.4535,
"step": 665
},
{
"epoch": 1.2476722532588453,
"grad_norm": 0.12201354012941291,
"learning_rate": 3.549854374925716e-05,
"loss": 0.4442,
"step": 670
},
{
"epoch": 1.2569832402234637,
"grad_norm": 0.11281012174298978,
"learning_rate": 3.528221548549307e-05,
"loss": 0.4427,
"step": 675
},
{
"epoch": 1.266294227188082,
"grad_norm": 0.12838733714308004,
"learning_rate": 3.506506695267052e-05,
"loss": 0.4529,
"step": 680
},
{
"epoch": 1.2756052141527001,
"grad_norm": 0.15014600853193164,
"learning_rate": 3.4847121038901877e-05,
"loss": 0.4477,
"step": 685
},
{
"epoch": 1.2849162011173183,
"grad_norm": 0.1302974794693121,
"learning_rate": 3.4628400716345896e-05,
"loss": 0.4426,
"step": 690
},
{
"epoch": 1.2942271880819367,
"grad_norm": 0.14919901945518166,
"learning_rate": 3.440892903878634e-05,
"loss": 0.4548,
"step": 695
},
{
"epoch": 1.303538175046555,
"grad_norm": 0.1440080046929404,
"learning_rate": 3.4188729139202065e-05,
"loss": 0.4468,
"step": 700
},
{
"epoch": 1.3128491620111733,
"grad_norm": 0.13277438244879466,
"learning_rate": 3.396782422732874e-05,
"loss": 0.4477,
"step": 705
},
{
"epoch": 1.3221601489757915,
"grad_norm": 0.12813547908714812,
"learning_rate": 3.374623758721246e-05,
"loss": 0.443,
"step": 710
},
{
"epoch": 1.3314711359404097,
"grad_norm": 0.1142868678900171,
"learning_rate": 3.352399257475553e-05,
"loss": 0.4476,
"step": 715
},
{
"epoch": 1.3407821229050279,
"grad_norm": 0.12281619965751127,
"learning_rate": 3.330111261525471e-05,
"loss": 0.4509,
"step": 720
},
{
"epoch": 1.3500931098696463,
"grad_norm": 0.12546558627371193,
"learning_rate": 3.307762120093208e-05,
"loss": 0.4467,
"step": 725
},
{
"epoch": 1.3594040968342644,
"grad_norm": 0.12160846175906896,
"learning_rate": 3.285354188845892e-05,
"loss": 0.4559,
"step": 730
},
{
"epoch": 1.3687150837988826,
"grad_norm": 0.1327867446422573,
"learning_rate": 3.262889829647276e-05,
"loss": 0.4461,
"step": 735
},
{
"epoch": 1.378026070763501,
"grad_norm": 0.11851775176530409,
"learning_rate": 3.2403714103087875e-05,
"loss": 0.4376,
"step": 740
},
{
"epoch": 1.3873370577281192,
"grad_norm": 0.13354798789102165,
"learning_rate": 3.217801304339959e-05,
"loss": 0.4483,
"step": 745
},
{
"epoch": 1.3966480446927374,
"grad_norm": 0.1219907416413629,
"learning_rate": 3.1951818906982474e-05,
"loss": 0.4574,
"step": 750
},
{
"epoch": 1.4059590316573556,
"grad_norm": 0.1341034895178442,
"learning_rate": 3.1725155535382905e-05,
"loss": 0.4557,
"step": 755
},
{
"epoch": 1.415270018621974,
"grad_norm": 0.12347048681951657,
"learning_rate": 3.1498046819606056e-05,
"loss": 0.447,
"step": 760
},
{
"epoch": 1.4245810055865922,
"grad_norm": 0.11917886931538982,
"learning_rate": 3.127051669759769e-05,
"loss": 0.4454,
"step": 765
},
{
"epoch": 1.4338919925512104,
"grad_norm": 0.11857820273802168,
"learning_rate": 3.1042589151721134e-05,
"loss": 0.449,
"step": 770
},
{
"epoch": 1.4432029795158288,
"grad_norm": 0.11694287535715704,
"learning_rate": 3.081428820622935e-05,
"loss": 0.4384,
"step": 775
},
{
"epoch": 1.452513966480447,
"grad_norm": 0.1155522561144092,
"learning_rate": 3.058563792473273e-05,
"loss": 0.4391,
"step": 780
},
{
"epoch": 1.4618249534450651,
"grad_norm": 0.12724124843039097,
"learning_rate": 3.0356662407662768e-05,
"loss": 0.4464,
"step": 785
},
{
"epoch": 1.4711359404096833,
"grad_norm": 0.12271179738453024,
"learning_rate": 3.0127385789731766e-05,
"loss": 0.447,
"step": 790
},
{
"epoch": 1.4804469273743017,
"grad_norm": 0.13153662517383716,
"learning_rate": 2.9897832237388977e-05,
"loss": 0.4522,
"step": 795
},
{
"epoch": 1.48975791433892,
"grad_norm": 0.13079604733987632,
"learning_rate": 2.9668025946273353e-05,
"loss": 0.4466,
"step": 800
},
{
"epoch": 1.499068901303538,
"grad_norm": 0.12483287404044502,
"learning_rate": 2.943799113866329e-05,
"loss": 0.4535,
"step": 805
},
{
"epoch": 1.5083798882681565,
"grad_norm": 0.16043696471672628,
"learning_rate": 2.9207752060923526e-05,
"loss": 0.4451,
"step": 810
},
{
"epoch": 1.5176908752327747,
"grad_norm": 0.1237851353311521,
"learning_rate": 2.8977332980949468e-05,
"loss": 0.4311,
"step": 815
},
{
"epoch": 1.5270018621973929,
"grad_norm": 0.11667518885636707,
"learning_rate": 2.874675818560933e-05,
"loss": 0.439,
"step": 820
},
{
"epoch": 1.536312849162011,
"grad_norm": 0.1165443133197676,
"learning_rate": 2.851605197818416e-05,
"loss": 0.4533,
"step": 825
},
{
"epoch": 1.5456238361266295,
"grad_norm": 0.11852825430527979,
"learning_rate": 2.828523867580628e-05,
"loss": 0.4478,
"step": 830
},
{
"epoch": 1.5549348230912476,
"grad_norm": 0.11263242835984934,
"learning_rate": 2.80543426068961e-05,
"loss": 0.4413,
"step": 835
},
{
"epoch": 1.564245810055866,
"grad_norm": 0.11792947875462449,
"learning_rate": 2.7823388108597904e-05,
"loss": 0.4419,
"step": 840
},
{
"epoch": 1.5735567970204842,
"grad_norm": 0.11536152090272359,
"learning_rate": 2.75923995242146e-05,
"loss": 0.4441,
"step": 845
},
{
"epoch": 1.5828677839851024,
"grad_norm": 0.11948166926439772,
"learning_rate": 2.7361401200641883e-05,
"loss": 0.4481,
"step": 850
},
{
"epoch": 1.5921787709497206,
"grad_norm": 0.13523357634130967,
"learning_rate": 2.7130417485801968e-05,
"loss": 0.446,
"step": 855
},
{
"epoch": 1.6014897579143388,
"grad_norm": 0.12128750858126454,
"learning_rate": 2.6899472726077275e-05,
"loss": 0.4425,
"step": 860
},
{
"epoch": 1.6108007448789572,
"grad_norm": 0.11455173358368062,
"learning_rate": 2.6668591263744246e-05,
"loss": 0.4441,
"step": 865
},
{
"epoch": 1.6201117318435754,
"grad_norm": 0.13042675513137278,
"learning_rate": 2.6437797434407574e-05,
"loss": 0.4398,
"step": 870
},
{
"epoch": 1.6294227188081938,
"grad_norm": 0.12385390445885733,
"learning_rate": 2.620711556443516e-05,
"loss": 0.4511,
"step": 875
},
{
"epoch": 1.638733705772812,
"grad_norm": 0.11517915674628858,
"learning_rate": 2.5976569968394083e-05,
"loss": 0.4443,
"step": 880
},
{
"epoch": 1.6480446927374302,
"grad_norm": 0.10395291968134747,
"learning_rate": 2.574618494648771e-05,
"loss": 0.4442,
"step": 885
},
{
"epoch": 1.6573556797020483,
"grad_norm": 0.11431662054546529,
"learning_rate": 2.5515984781994417e-05,
"loss": 0.4433,
"step": 890
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.11997395296814937,
"learning_rate": 2.528599373870804e-05,
"loss": 0.4412,
"step": 895
},
{
"epoch": 1.675977653631285,
"grad_norm": 0.11251548912326118,
"learning_rate": 2.5056236058380422e-05,
"loss": 0.4341,
"step": 900
},
{
"epoch": 1.6852886405959033,
"grad_norm": 0.12200163922402081,
"learning_rate": 2.4826735958166238e-05,
"loss": 0.4391,
"step": 905
},
{
"epoch": 1.6945996275605215,
"grad_norm": 0.12515447326859908,
"learning_rate": 2.4597517628070434e-05,
"loss": 0.4335,
"step": 910
},
{
"epoch": 1.7039106145251397,
"grad_norm": 0.11900007427268738,
"learning_rate": 2.4368605228398534e-05,
"loss": 0.4442,
"step": 915
},
{
"epoch": 1.7132216014897579,
"grad_norm": 0.10327616314867456,
"learning_rate": 2.4140022887210074e-05,
"loss": 0.442,
"step": 920
},
{
"epoch": 1.722532588454376,
"grad_norm": 0.11195530093482621,
"learning_rate": 2.391179469777543e-05,
"loss": 0.4335,
"step": 925
},
{
"epoch": 1.7318435754189943,
"grad_norm": 0.10733741642428904,
"learning_rate": 2.3683944716036316e-05,
"loss": 0.4379,
"step": 930
},
{
"epoch": 1.7411545623836127,
"grad_norm": 0.10722035609207795,
"learning_rate": 2.345649695807019e-05,
"loss": 0.4367,
"step": 935
},
{
"epoch": 1.750465549348231,
"grad_norm": 0.11022965957105299,
"learning_rate": 2.322947539755892e-05,
"loss": 0.4433,
"step": 940
},
{
"epoch": 1.7597765363128492,
"grad_norm": 0.11737086019377437,
"learning_rate": 2.3002903963261884e-05,
"loss": 0.4403,
"step": 945
},
{
"epoch": 1.7690875232774674,
"grad_norm": 0.10855004430890172,
"learning_rate": 2.2776806536493774e-05,
"loss": 0.4391,
"step": 950
},
{
"epoch": 1.7783985102420856,
"grad_norm": 0.11126097809718666,
"learning_rate": 2.2551206948607466e-05,
"loss": 0.4344,
"step": 955
},
{
"epoch": 1.7877094972067038,
"grad_norm": 0.11482383919676267,
"learning_rate": 2.2326128978482103e-05,
"loss": 0.4369,
"step": 960
},
{
"epoch": 1.7970204841713222,
"grad_norm": 0.10817387409350918,
"learning_rate": 2.2101596350016756e-05,
"loss": 0.4384,
"step": 965
},
{
"epoch": 1.8063314711359404,
"grad_norm": 0.10278656697485476,
"learning_rate": 2.1877632729629803e-05,
"loss": 0.4363,
"step": 970
},
{
"epoch": 1.8156424581005588,
"grad_norm": 0.10003238891948513,
"learning_rate": 2.16542617237645e-05,
"loss": 0.4386,
"step": 975
},
{
"epoch": 1.824953445065177,
"grad_norm": 0.12056952514610235,
"learning_rate": 2.1431506876400694e-05,
"loss": 0.4324,
"step": 980
},
{
"epoch": 1.8342644320297952,
"grad_norm": 0.11514887779350329,
"learning_rate": 2.1209391666573265e-05,
"loss": 0.4431,
"step": 985
},
{
"epoch": 1.8435754189944134,
"grad_norm": 0.11972084038390686,
"learning_rate": 2.0987939505897393e-05,
"loss": 0.4452,
"step": 990
},
{
"epoch": 1.8528864059590315,
"grad_norm": 0.11700688868223293,
"learning_rate": 2.076717373610083e-05,
"loss": 0.4423,
"step": 995
},
{
"epoch": 1.86219739292365,
"grad_norm": 0.111956011270972,
"learning_rate": 2.054711762656369e-05,
"loss": 0.4377,
"step": 1000
},
{
"epoch": 1.8715083798882681,
"grad_norm": 0.10113399321289508,
"learning_rate": 2.032779437186572e-05,
"loss": 0.4386,
"step": 1005
},
{
"epoch": 1.8808193668528865,
"grad_norm": 0.1058333563087994,
"learning_rate": 2.010922708934157e-05,
"loss": 0.4317,
"step": 1010
},
{
"epoch": 1.8901303538175047,
"grad_norm": 0.10370579219821685,
"learning_rate": 1.9891438816644114e-05,
"loss": 0.4349,
"step": 1015
},
{
"epoch": 1.899441340782123,
"grad_norm": 0.1206689535928644,
"learning_rate": 1.967445250931627e-05,
"loss": 0.4365,
"step": 1020
},
{
"epoch": 1.908752327746741,
"grad_norm": 0.1202510464984422,
"learning_rate": 1.945829103837133e-05,
"loss": 0.4479,
"step": 1025
},
{
"epoch": 1.9180633147113593,
"grad_norm": 0.10822437459168585,
"learning_rate": 1.9242977187882383e-05,
"loss": 0.4438,
"step": 1030
},
{
"epoch": 1.9273743016759777,
"grad_norm": 0.11150240856764543,
"learning_rate": 1.902853365258072e-05,
"loss": 0.4366,
"step": 1035
},
{
"epoch": 1.9366852886405959,
"grad_norm": 0.10570123944866744,
"learning_rate": 1.8814983035463794e-05,
"loss": 0.4433,
"step": 1040
},
{
"epoch": 1.9459962756052143,
"grad_norm": 0.10602330715260787,
"learning_rate": 1.86023478454128e-05,
"loss": 0.4388,
"step": 1045
},
{
"epoch": 1.9553072625698324,
"grad_norm": 0.10899590289579546,
"learning_rate": 1.8390650494820107e-05,
"loss": 0.4362,
"step": 1050
},
{
"epoch": 1.9646182495344506,
"grad_norm": 0.11765129534980089,
"learning_rate": 1.8179913297226993e-05,
"loss": 0.4408,
"step": 1055
},
{
"epoch": 1.9739292364990688,
"grad_norm": 0.11030605011675014,
"learning_rate": 1.7970158464971714e-05,
"loss": 0.4467,
"step": 1060
},
{
"epoch": 1.983240223463687,
"grad_norm": 0.10612432103217494,
"learning_rate": 1.7761408106848217e-05,
"loss": 0.4346,
"step": 1065
},
{
"epoch": 1.9925512104283054,
"grad_norm": 0.10754649072396187,
"learning_rate": 1.7553684225775867e-05,
"loss": 0.429,
"step": 1070
},
{
"epoch": 2.001862197392924,
"grad_norm": 0.10691719626323028,
"learning_rate": 1.734700871648024e-05,
"loss": 0.4264,
"step": 1075
},
{
"epoch": 2.011173184357542,
"grad_norm": 0.11506657646587219,
"learning_rate": 1.7141403363185325e-05,
"loss": 0.4226,
"step": 1080
},
{
"epoch": 2.02048417132216,
"grad_norm": 0.10752371863340741,
"learning_rate": 1.6936889837317466e-05,
"loss": 0.421,
"step": 1085
},
{
"epoch": 2.0297951582867784,
"grad_norm": 0.10258005127471478,
"learning_rate": 1.6733489695221056e-05,
"loss": 0.4148,
"step": 1090
},
{
"epoch": 2.0391061452513966,
"grad_norm": 0.10283082864546561,
"learning_rate": 1.6531224375886507e-05,
"loss": 0.4202,
"step": 1095
},
{
"epoch": 2.0484171322160147,
"grad_norm": 0.10708122228606631,
"learning_rate": 1.6330115198690462e-05,
"loss": 0.4287,
"step": 1100
},
{
"epoch": 2.0577281191806334,
"grad_norm": 0.10210619581682416,
"learning_rate": 1.6130183361148672e-05,
"loss": 0.4212,
"step": 1105
},
{
"epoch": 2.0670391061452515,
"grad_norm": 0.09909955572450115,
"learning_rate": 1.593144993668175e-05,
"loss": 0.4128,
"step": 1110
},
{
"epoch": 2.0763500931098697,
"grad_norm": 0.09903027868569465,
"learning_rate": 1.5733935872393924e-05,
"loss": 0.4102,
"step": 1115
},
{
"epoch": 2.085661080074488,
"grad_norm": 0.11465913565196915,
"learning_rate": 1.5537661986865187e-05,
"loss": 0.4196,
"step": 1120
},
{
"epoch": 2.094972067039106,
"grad_norm": 0.10431445935320442,
"learning_rate": 1.5342648967956935e-05,
"loss": 0.4211,
"step": 1125
},
{
"epoch": 2.1042830540037243,
"grad_norm": 0.10867828847006396,
"learning_rate": 1.5148917370631389e-05,
"loss": 0.4273,
"step": 1130
},
{
"epoch": 2.1135940409683425,
"grad_norm": 0.10564151475662023,
"learning_rate": 1.4956487614785076e-05,
"loss": 0.4177,
"step": 1135
},
{
"epoch": 2.122905027932961,
"grad_norm": 0.10252126454298957,
"learning_rate": 1.47653799830965e-05,
"loss": 0.4256,
"step": 1140
},
{
"epoch": 2.1322160148975793,
"grad_norm": 0.10583448098931744,
"learning_rate": 1.4575614618888248e-05,
"loss": 0.4162,
"step": 1145
},
{
"epoch": 2.1415270018621975,
"grad_norm": 0.11129869429457806,
"learning_rate": 1.438721152400388e-05,
"loss": 0.4214,
"step": 1150
},
{
"epoch": 2.1508379888268156,
"grad_norm": 0.10661908698656541,
"learning_rate": 1.4200190556699653e-05,
"loss": 0.4171,
"step": 1155
},
{
"epoch": 2.160148975791434,
"grad_norm": 0.10172941781846247,
"learning_rate": 1.4014571429551348e-05,
"loss": 0.4198,
"step": 1160
},
{
"epoch": 2.169459962756052,
"grad_norm": 0.10129345871926003,
"learning_rate": 1.3830373707376623e-05,
"loss": 0.4207,
"step": 1165
},
{
"epoch": 2.17877094972067,
"grad_norm": 0.1039002300974548,
"learning_rate": 1.364761680517269e-05,
"loss": 0.414,
"step": 1170
},
{
"epoch": 2.188081936685289,
"grad_norm": 0.09557941393888705,
"learning_rate": 1.3466319986070008e-05,
"loss": 0.419,
"step": 1175
},
{
"epoch": 2.197392923649907,
"grad_norm": 0.10379899436807119,
"learning_rate": 1.3286502359301862e-05,
"loss": 0.423,
"step": 1180
},
{
"epoch": 2.206703910614525,
"grad_norm": 0.09519170293308159,
"learning_rate": 1.3108182878190156e-05,
"loss": 0.4071,
"step": 1185
},
{
"epoch": 2.2160148975791434,
"grad_norm": 0.09880897724698645,
"learning_rate": 1.2931380338147745e-05,
"loss": 0.4137,
"step": 1190
},
{
"epoch": 2.2253258845437616,
"grad_norm": 0.10001625336759871,
"learning_rate": 1.2756113374697301e-05,
"loss": 0.4141,
"step": 1195
},
{
"epoch": 2.2346368715083798,
"grad_norm": 0.0996637237587361,
"learning_rate": 1.2582400461507055e-05,
"loss": 0.4203,
"step": 1200
},
{
"epoch": 2.243947858472998,
"grad_norm": 0.10454993690871738,
"learning_rate": 1.2410259908443657e-05,
"loss": 0.4102,
"step": 1205
},
{
"epoch": 2.2532588454376166,
"grad_norm": 0.1003379879509651,
"learning_rate": 1.2239709859642236e-05,
"loss": 0.4223,
"step": 1210
},
{
"epoch": 2.2625698324022347,
"grad_norm": 0.09802095097599599,
"learning_rate": 1.2070768291593964e-05,
"loss": 0.4151,
"step": 1215
},
{
"epoch": 2.271880819366853,
"grad_norm": 0.09783634576245234,
"learning_rate": 1.1903453011251281e-05,
"loss": 0.4171,
"step": 1220
},
{
"epoch": 2.281191806331471,
"grad_norm": 0.10651764573200316,
"learning_rate": 1.1737781654150954e-05,
"loss": 0.4171,
"step": 1225
},
{
"epoch": 2.2905027932960893,
"grad_norm": 0.10998450255021262,
"learning_rate": 1.1573771682555298e-05,
"loss": 0.4212,
"step": 1230
},
{
"epoch": 2.2998137802607075,
"grad_norm": 0.09653384858037176,
"learning_rate": 1.1411440383611559e-05,
"loss": 0.4149,
"step": 1235
},
{
"epoch": 2.3091247672253257,
"grad_norm": 0.10197498165183494,
"learning_rate": 1.1250804867529793e-05,
"loss": 0.4174,
"step": 1240
},
{
"epoch": 2.3184357541899443,
"grad_norm": 0.10064617044586674,
"learning_rate": 1.1091882065779429e-05,
"loss": 0.4142,
"step": 1245
},
{
"epoch": 2.3277467411545625,
"grad_norm": 0.09997648870136769,
"learning_rate": 1.093468872930462e-05,
"loss": 0.4248,
"step": 1250
},
{
"epoch": 2.3370577281191807,
"grad_norm": 0.10126015276405256,
"learning_rate": 1.0779241426758628e-05,
"loss": 0.4187,
"step": 1255
},
{
"epoch": 2.346368715083799,
"grad_norm": 0.10248882729986777,
"learning_rate": 1.0625556542757494e-05,
"loss": 0.4256,
"step": 1260
},
{
"epoch": 2.355679702048417,
"grad_norm": 0.09629566154715391,
"learning_rate": 1.0473650276152972e-05,
"loss": 0.4215,
"step": 1265
},
{
"epoch": 2.364990689013035,
"grad_norm": 0.09893527097271672,
"learning_rate": 1.0323538638325184e-05,
"loss": 0.4188,
"step": 1270
},
{
"epoch": 2.3743016759776534,
"grad_norm": 0.09302606623894093,
"learning_rate": 1.0175237451494946e-05,
"loss": 0.4145,
"step": 1275
},
{
"epoch": 2.383612662942272,
"grad_norm": 0.10187238212320712,
"learning_rate": 1.0028762347056049e-05,
"loss": 0.4087,
"step": 1280
},
{
"epoch": 2.39292364990689,
"grad_norm": 0.09943852275889314,
"learning_rate": 9.88412876392769e-06,
"loss": 0.4234,
"step": 1285
},
{
"epoch": 2.4022346368715084,
"grad_norm": 0.09579290986291013,
"learning_rate": 9.741351946927141e-06,
"loss": 0.4209,
"step": 1290
},
{
"epoch": 2.4115456238361266,
"grad_norm": 0.10121357816296453,
"learning_rate": 9.600446945162908e-06,
"loss": 0.4221,
"step": 1295
},
{
"epoch": 2.4208566108007448,
"grad_norm": 0.09927071713182525,
"learning_rate": 9.461428610448502e-06,
"loss": 0.4227,
"step": 1300
},
{
"epoch": 2.430167597765363,
"grad_norm": 0.10159715499322029,
"learning_rate": 9.324311595737025e-06,
"loss": 0.4169,
"step": 1305
},
{
"epoch": 2.439478584729981,
"grad_norm": 0.10043193606766965,
"learning_rate": 9.189110353576703e-06,
"loss": 0.4117,
"step": 1310
},
{
"epoch": 2.4487895716945998,
"grad_norm": 0.09912578603874618,
"learning_rate": 9.055839134587527e-06,
"loss": 0.4151,
"step": 1315
},
{
"epoch": 2.458100558659218,
"grad_norm": 0.09297775914893847,
"learning_rate": 8.92451198595921e-06,
"loss": 0.4099,
"step": 1320
},
{
"epoch": 2.467411545623836,
"grad_norm": 0.09984530224126298,
"learning_rate": 8.795142749970564e-06,
"loss": 0.4123,
"step": 1325
},
{
"epoch": 2.4767225325884543,
"grad_norm": 0.09120975321788215,
"learning_rate": 8.667745062530503e-06,
"loss": 0.4143,
"step": 1330
},
{
"epoch": 2.4860335195530725,
"grad_norm": 0.09774167066305631,
"learning_rate": 8.542332351740727e-06,
"loss": 0.4189,
"step": 1335
},
{
"epoch": 2.4953445065176907,
"grad_norm": 0.09827792678481405,
"learning_rate": 8.418917836480417e-06,
"loss": 0.4194,
"step": 1340
},
{
"epoch": 2.504655493482309,
"grad_norm": 0.09390126470923525,
"learning_rate": 8.297514525012894e-06,
"loss": 0.422,
"step": 1345
},
{
"epoch": 2.5139664804469275,
"grad_norm": 0.09346706560041741,
"learning_rate": 8.178135213614499e-06,
"loss": 0.4198,
"step": 1350
},
{
"epoch": 2.5232774674115457,
"grad_norm": 0.09400692518566071,
"learning_rate": 8.060792485225877e-06,
"loss": 0.4153,
"step": 1355
},
{
"epoch": 2.532588454376164,
"grad_norm": 0.09935167204099171,
"learning_rate": 7.945498708125612e-06,
"loss": 0.419,
"step": 1360
},
{
"epoch": 2.541899441340782,
"grad_norm": 0.09255945749166977,
"learning_rate": 7.832266034626668e-06,
"loss": 0.4077,
"step": 1365
},
{
"epoch": 2.5512104283054002,
"grad_norm": 0.09429134332541908,
"learning_rate": 7.721106399795427e-06,
"loss": 0.4111,
"step": 1370
},
{
"epoch": 2.560521415270019,
"grad_norm": 0.09593431408689819,
"learning_rate": 7.612031520193731e-06,
"loss": 0.4144,
"step": 1375
},
{
"epoch": 2.5698324022346366,
"grad_norm": 0.10216201759504691,
"learning_rate": 7.505052892643915e-06,
"loss": 0.4147,
"step": 1380
},
{
"epoch": 2.5791433891992552,
"grad_norm": 0.09195284463580435,
"learning_rate": 7.4001817930170036e-06,
"loss": 0.42,
"step": 1385
},
{
"epoch": 2.5884543761638734,
"grad_norm": 0.09617799061198681,
"learning_rate": 7.297429275044194e-06,
"loss": 0.4093,
"step": 1390
},
{
"epoch": 2.5977653631284916,
"grad_norm": 0.10099179291339958,
"learning_rate": 7.196806169151786e-06,
"loss": 0.4233,
"step": 1395
},
{
"epoch": 2.60707635009311,
"grad_norm": 0.09641311151569762,
"learning_rate": 7.098323081319576e-06,
"loss": 0.4155,
"step": 1400
},
{
"epoch": 2.616387337057728,
"grad_norm": 0.09465528720326888,
"learning_rate": 7.001990391963011e-06,
"loss": 0.4179,
"step": 1405
},
{
"epoch": 2.6256983240223466,
"grad_norm": 0.09927128319879645,
"learning_rate": 6.907818254839017e-06,
"loss": 0.4203,
"step": 1410
},
{
"epoch": 2.635009310986965,
"grad_norm": 0.09942785015352072,
"learning_rate": 6.815816595975771e-06,
"loss": 0.4216,
"step": 1415
},
{
"epoch": 2.644320297951583,
"grad_norm": 0.1002604796284761,
"learning_rate": 6.725995112626489e-06,
"loss": 0.4214,
"step": 1420
},
{
"epoch": 2.653631284916201,
"grad_norm": 0.09509870809273925,
"learning_rate": 6.638363272247286e-06,
"loss": 0.4149,
"step": 1425
},
{
"epoch": 2.6629422718808193,
"grad_norm": 0.10038637331627898,
"learning_rate": 6.552930311499275e-06,
"loss": 0.4243,
"step": 1430
},
{
"epoch": 2.6722532588454375,
"grad_norm": 0.09849966019071536,
"learning_rate": 6.469705235275016e-06,
"loss": 0.423,
"step": 1435
},
{
"epoch": 2.6815642458100557,
"grad_norm": 0.09583945495890936,
"learning_rate": 6.388696815749346e-06,
"loss": 0.4297,
"step": 1440
},
{
"epoch": 2.6908752327746743,
"grad_norm": 0.09818740944151227,
"learning_rate": 6.309913591454788e-06,
"loss": 0.4175,
"step": 1445
},
{
"epoch": 2.7001862197392925,
"grad_norm": 0.09297545599110695,
"learning_rate": 6.233363866381562e-06,
"loss": 0.4151,
"step": 1450
},
{
"epoch": 2.7094972067039107,
"grad_norm": 0.09347294049524463,
"learning_rate": 6.159055709102293e-06,
"loss": 0.4211,
"step": 1455
},
{
"epoch": 2.718808193668529,
"grad_norm": 0.09063495788855541,
"learning_rate": 6.086996951921601e-06,
"loss": 0.4169,
"step": 1460
},
{
"epoch": 2.728119180633147,
"grad_norm": 0.09576157520224456,
"learning_rate": 6.017195190050504e-06,
"loss": 0.4178,
"step": 1465
},
{
"epoch": 2.7374301675977653,
"grad_norm": 0.09821892447570103,
"learning_rate": 5.949657780805906e-06,
"loss": 0.4196,
"step": 1470
},
{
"epoch": 2.7467411545623834,
"grad_norm": 0.09403307674636713,
"learning_rate": 5.88439184283509e-06,
"loss": 0.4191,
"step": 1475
},
{
"epoch": 2.756052141527002,
"grad_norm": 0.09844473844449389,
"learning_rate": 5.821404255365381e-06,
"loss": 0.4167,
"step": 1480
},
{
"epoch": 2.7653631284916202,
"grad_norm": 0.1049290391027304,
"learning_rate": 5.7607016574790845e-06,
"loss": 0.4158,
"step": 1485
},
{
"epoch": 2.7746741154562384,
"grad_norm": 0.09142057289478683,
"learning_rate": 5.702290447413681e-06,
"loss": 0.4195,
"step": 1490
},
{
"epoch": 2.7839851024208566,
"grad_norm": 0.09437947535885335,
"learning_rate": 5.646176781887437e-06,
"loss": 0.422,
"step": 1495
},
{
"epoch": 2.793296089385475,
"grad_norm": 0.09658988683610302,
"learning_rate": 5.592366575450494e-06,
"loss": 0.4132,
"step": 1500
},
{
"epoch": 2.802607076350093,
"grad_norm": 0.10021332194018327,
"learning_rate": 5.5408654998614055e-06,
"loss": 0.4285,
"step": 1505
},
{
"epoch": 2.811918063314711,
"grad_norm": 0.09395338001674504,
"learning_rate": 5.491678983489372e-06,
"loss": 0.4056,
"step": 1510
},
{
"epoch": 2.82122905027933,
"grad_norm": 0.10069367341531493,
"learning_rate": 5.4448122107420415e-06,
"loss": 0.4289,
"step": 1515
},
{
"epoch": 2.830540037243948,
"grad_norm": 0.0977726987894797,
"learning_rate": 5.40027012151906e-06,
"loss": 0.4178,
"step": 1520
},
{
"epoch": 2.839851024208566,
"grad_norm": 0.09024612259420411,
"learning_rate": 5.3580574106914015e-06,
"loss": 0.4118,
"step": 1525
},
{
"epoch": 2.8491620111731844,
"grad_norm": 0.09342541886916174,
"learning_rate": 5.318178527606511e-06,
"loss": 0.4205,
"step": 1530
},
{
"epoch": 2.8584729981378025,
"grad_norm": 0.09438754600698024,
"learning_rate": 5.2806376756193194e-06,
"loss": 0.4129,
"step": 1535
},
{
"epoch": 2.8677839851024207,
"grad_norm": 0.09314142674041269,
"learning_rate": 5.245438811649216e-06,
"loss": 0.4174,
"step": 1540
},
{
"epoch": 2.877094972067039,
"grad_norm": 0.09920083405177257,
"learning_rate": 5.212585645762963e-06,
"loss": 0.413,
"step": 1545
},
{
"epoch": 2.8864059590316575,
"grad_norm": 0.10054514046549108,
"learning_rate": 5.182081640783647e-06,
"loss": 0.4201,
"step": 1550
},
{
"epoch": 2.8957169459962757,
"grad_norm": 0.09375029466655158,
"learning_rate": 5.153930011925686e-06,
"loss": 0.4117,
"step": 1555
},
{
"epoch": 2.905027932960894,
"grad_norm": 0.1187162272510011,
"learning_rate": 5.128133726455941e-06,
"loss": 0.4147,
"step": 1560
},
{
"epoch": 2.914338919925512,
"grad_norm": 0.09097957682917539,
"learning_rate": 5.104695503380948e-06,
"loss": 0.406,
"step": 1565
},
{
"epoch": 2.9236499068901303,
"grad_norm": 0.08917608409363194,
"learning_rate": 5.083617813160338e-06,
"loss": 0.4139,
"step": 1570
},
{
"epoch": 2.9329608938547485,
"grad_norm": 0.09357577576462056,
"learning_rate": 5.064902877446432e-06,
"loss": 0.4198,
"step": 1575
},
{
"epoch": 2.9422718808193666,
"grad_norm": 0.09162535893385207,
"learning_rate": 5.048552668850077e-06,
"loss": 0.4211,
"step": 1580
},
{
"epoch": 2.9515828677839853,
"grad_norm": 0.09497386963890858,
"learning_rate": 5.034568910732737e-06,
"loss": 0.4143,
"step": 1585
},
{
"epoch": 2.9608938547486034,
"grad_norm": 0.09474281729138122,
"learning_rate": 5.022953077024818e-06,
"loss": 0.4194,
"step": 1590
},
{
"epoch": 2.9702048417132216,
"grad_norm": 0.09634825165067597,
"learning_rate": 5.0137063920703455e-06,
"loss": 0.4142,
"step": 1595
},
{
"epoch": 2.97951582867784,
"grad_norm": 0.09441014556129712,
"learning_rate": 5.0068298304978894e-06,
"loss": 0.4174,
"step": 1600
},
{
"epoch": 2.988826815642458,
"grad_norm": 0.09449692277995367,
"learning_rate": 5.002324117117843e-06,
"loss": 0.4177,
"step": 1605
},
{
"epoch": 2.998137802607076,
"grad_norm": 0.09204215488513472,
"learning_rate": 5.000189726846031e-06,
"loss": 0.4193,
"step": 1610
},
{
"epoch": 3.0,
"step": 1611,
"total_flos": 1468598568615936.0,
"train_loss": 0.4715816771147933,
"train_runtime": 19678.9404,
"train_samples_per_second": 2.617,
"train_steps_per_second": 0.082
}
],
"logging_steps": 5,
"max_steps": 1611,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1468598568615936.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}