gemma3-1b-it_eqa_lora_sft / trainer_state.json
hexuan21's picture
Upload folder using huggingface_hub
af3bfe2 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1795,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0027868738242876052,
"grad_norm": 4.701911449432373,
"learning_rate": 2.2222222222222225e-06,
"loss": 3.5703,
"step": 5
},
{
"epoch": 0.0055737476485752105,
"grad_norm": 4.517332077026367,
"learning_rate": 5e-06,
"loss": 3.5228,
"step": 10
},
{
"epoch": 0.008360621472862817,
"grad_norm": 4.366923809051514,
"learning_rate": 7.777777777777777e-06,
"loss": 3.4713,
"step": 15
},
{
"epoch": 0.011147495297150421,
"grad_norm": 4.379341125488281,
"learning_rate": 1.0555555555555555e-05,
"loss": 3.5622,
"step": 20
},
{
"epoch": 0.013934369121438027,
"grad_norm": 5.151899814605713,
"learning_rate": 1.3333333333333333e-05,
"loss": 3.4422,
"step": 25
},
{
"epoch": 0.016721242945725634,
"grad_norm": 4.80206298828125,
"learning_rate": 1.6111111111111115e-05,
"loss": 3.3216,
"step": 30
},
{
"epoch": 0.019508116770013236,
"grad_norm": 3.687617778778076,
"learning_rate": 1.888888888888889e-05,
"loss": 2.942,
"step": 35
},
{
"epoch": 0.022294990594300842,
"grad_norm": 3.1759581565856934,
"learning_rate": 2.1666666666666667e-05,
"loss": 2.6761,
"step": 40
},
{
"epoch": 0.025081864418588447,
"grad_norm": 2.588658332824707,
"learning_rate": 2.4444444444444445e-05,
"loss": 2.6441,
"step": 45
},
{
"epoch": 0.027868738242876053,
"grad_norm": 2.1114368438720703,
"learning_rate": 2.7222222222222223e-05,
"loss": 2.4923,
"step": 50
},
{
"epoch": 0.03065561206716366,
"grad_norm": 2.0633840560913086,
"learning_rate": 3e-05,
"loss": 2.2875,
"step": 55
},
{
"epoch": 0.03344248589145127,
"grad_norm": 1.8759537935256958,
"learning_rate": 3.277777777777778e-05,
"loss": 2.408,
"step": 60
},
{
"epoch": 0.03622935971573887,
"grad_norm": 1.9867842197418213,
"learning_rate": 3.555555555555556e-05,
"loss": 2.3362,
"step": 65
},
{
"epoch": 0.03901623354002647,
"grad_norm": 1.781015157699585,
"learning_rate": 3.8333333333333334e-05,
"loss": 2.1793,
"step": 70
},
{
"epoch": 0.04180310736431408,
"grad_norm": 2.1073060035705566,
"learning_rate": 4.111111111111111e-05,
"loss": 2.2342,
"step": 75
},
{
"epoch": 0.044589981188601684,
"grad_norm": 1.8073269128799438,
"learning_rate": 4.388888888888889e-05,
"loss": 2.1447,
"step": 80
},
{
"epoch": 0.04737685501288929,
"grad_norm": 1.8754196166992188,
"learning_rate": 4.666666666666667e-05,
"loss": 2.3169,
"step": 85
},
{
"epoch": 0.050163728837176895,
"grad_norm": 1.9674766063690186,
"learning_rate": 4.9444444444444446e-05,
"loss": 2.1966,
"step": 90
},
{
"epoch": 0.0529506026614645,
"grad_norm": 2.0507538318634033,
"learning_rate": 5.222222222222223e-05,
"loss": 2.1436,
"step": 95
},
{
"epoch": 0.055737476485752106,
"grad_norm": 2.1478066444396973,
"learning_rate": 5.500000000000001e-05,
"loss": 2.0947,
"step": 100
},
{
"epoch": 0.05852435031003971,
"grad_norm": 1.8122990131378174,
"learning_rate": 5.7777777777777776e-05,
"loss": 2.1266,
"step": 105
},
{
"epoch": 0.06131122413432732,
"grad_norm": 2.30039119720459,
"learning_rate": 6.055555555555555e-05,
"loss": 2.175,
"step": 110
},
{
"epoch": 0.06409809795861493,
"grad_norm": 2.1259849071502686,
"learning_rate": 6.333333333333333e-05,
"loss": 2.0622,
"step": 115
},
{
"epoch": 0.06688497178290254,
"grad_norm": 2.159100294113159,
"learning_rate": 6.611111111111111e-05,
"loss": 2.078,
"step": 120
},
{
"epoch": 0.06967184560719013,
"grad_norm": 2.2886006832122803,
"learning_rate": 6.88888888888889e-05,
"loss": 2.1515,
"step": 125
},
{
"epoch": 0.07245871943147773,
"grad_norm": 2.1963438987731934,
"learning_rate": 7.166666666666667e-05,
"loss": 2.0886,
"step": 130
},
{
"epoch": 0.07524559325576534,
"grad_norm": 1.9774861335754395,
"learning_rate": 7.444444444444444e-05,
"loss": 2.0426,
"step": 135
},
{
"epoch": 0.07803246708005294,
"grad_norm": 2.2124135494232178,
"learning_rate": 7.722222222222223e-05,
"loss": 2.0686,
"step": 140
},
{
"epoch": 0.08081934090434055,
"grad_norm": 2.0250704288482666,
"learning_rate": 8e-05,
"loss": 2.1701,
"step": 145
},
{
"epoch": 0.08360621472862816,
"grad_norm": 2.2717087268829346,
"learning_rate": 8.277777777777778e-05,
"loss": 2.1313,
"step": 150
},
{
"epoch": 0.08639308855291576,
"grad_norm": 2.090191602706909,
"learning_rate": 8.555555555555556e-05,
"loss": 2.0572,
"step": 155
},
{
"epoch": 0.08917996237720337,
"grad_norm": 2.09201979637146,
"learning_rate": 8.833333333333333e-05,
"loss": 1.982,
"step": 160
},
{
"epoch": 0.09196683620149097,
"grad_norm": 2.206078290939331,
"learning_rate": 9.111111111111112e-05,
"loss": 2.0063,
"step": 165
},
{
"epoch": 0.09475371002577858,
"grad_norm": 2.1212692260742188,
"learning_rate": 9.388888888888889e-05,
"loss": 2.0597,
"step": 170
},
{
"epoch": 0.09754058385006618,
"grad_norm": 2.215811252593994,
"learning_rate": 9.666666666666667e-05,
"loss": 2.0566,
"step": 175
},
{
"epoch": 0.10032745767435379,
"grad_norm": 2.1642961502075195,
"learning_rate": 9.944444444444446e-05,
"loss": 2.0843,
"step": 180
},
{
"epoch": 0.1031143314986414,
"grad_norm": 2.315004348754883,
"learning_rate": 9.999848639521432e-05,
"loss": 2.0011,
"step": 185
},
{
"epoch": 0.105901205322929,
"grad_norm": 2.2057459354400635,
"learning_rate": 9.999233753283091e-05,
"loss": 1.9906,
"step": 190
},
{
"epoch": 0.1086880791472166,
"grad_norm": 2.516157865524292,
"learning_rate": 9.998145939378577e-05,
"loss": 1.9941,
"step": 195
},
{
"epoch": 0.11147495297150421,
"grad_norm": 2.2229058742523193,
"learning_rate": 9.996585300715116e-05,
"loss": 1.9935,
"step": 200
},
{
"epoch": 0.11426182679579182,
"grad_norm": 2.1316332817077637,
"learning_rate": 9.994551984929175e-05,
"loss": 1.9692,
"step": 205
},
{
"epoch": 0.11704870062007942,
"grad_norm": 1.8723924160003662,
"learning_rate": 9.992046184372492e-05,
"loss": 2.0208,
"step": 210
},
{
"epoch": 0.11983557444436703,
"grad_norm": 2.1844356060028076,
"learning_rate": 9.989068136093873e-05,
"loss": 1.9584,
"step": 215
},
{
"epoch": 0.12262244826865464,
"grad_norm": 2.2411093711853027,
"learning_rate": 9.985618121816779e-05,
"loss": 2.0796,
"step": 220
},
{
"epoch": 0.12540932209294225,
"grad_norm": 2.0734896659851074,
"learning_rate": 9.981696467912664e-05,
"loss": 1.9712,
"step": 225
},
{
"epoch": 0.12819619591722986,
"grad_norm": 2.1242079734802246,
"learning_rate": 9.97730354537011e-05,
"loss": 1.9637,
"step": 230
},
{
"epoch": 0.13098306974151747,
"grad_norm": 2.026093006134033,
"learning_rate": 9.972439769759722e-05,
"loss": 1.8615,
"step": 235
},
{
"epoch": 0.13376994356580507,
"grad_norm": 2.029449701309204,
"learning_rate": 9.967105601194823e-05,
"loss": 2.11,
"step": 240
},
{
"epoch": 0.13655681739009268,
"grad_norm": 2.0343263149261475,
"learning_rate": 9.961301544287922e-05,
"loss": 1.9325,
"step": 245
},
{
"epoch": 0.13934369121438026,
"grad_norm": 2.2704412937164307,
"learning_rate": 9.955028148102979e-05,
"loss": 1.8849,
"step": 250
},
{
"epoch": 0.14213056503866786,
"grad_norm": 2.563328742980957,
"learning_rate": 9.948286006103466e-05,
"loss": 2.1103,
"step": 255
},
{
"epoch": 0.14491743886295547,
"grad_norm": 2.4252612590789795,
"learning_rate": 9.941075756096226e-05,
"loss": 1.9867,
"step": 260
},
{
"epoch": 0.14770431268724307,
"grad_norm": 2.1968955993652344,
"learning_rate": 9.933398080171123e-05,
"loss": 1.998,
"step": 265
},
{
"epoch": 0.15049118651153068,
"grad_norm": 2.1625869274139404,
"learning_rate": 9.925253704636543e-05,
"loss": 1.9043,
"step": 270
},
{
"epoch": 0.15327806033581828,
"grad_norm": 2.349252700805664,
"learning_rate": 9.916643399950656e-05,
"loss": 1.9854,
"step": 275
},
{
"epoch": 0.1560649341601059,
"grad_norm": 2.2877767086029053,
"learning_rate": 9.907567980648549e-05,
"loss": 1.9119,
"step": 280
},
{
"epoch": 0.1588518079843935,
"grad_norm": 2.0312581062316895,
"learning_rate": 9.898028305265169e-05,
"loss": 2.0732,
"step": 285
},
{
"epoch": 0.1616386818086811,
"grad_norm": 2.036317825317383,
"learning_rate": 9.888025276254096e-05,
"loss": 2.0147,
"step": 290
},
{
"epoch": 0.1644255556329687,
"grad_norm": 2.345585823059082,
"learning_rate": 9.877559839902184e-05,
"loss": 1.9635,
"step": 295
},
{
"epoch": 0.1672124294572563,
"grad_norm": 2.371412992477417,
"learning_rate": 9.86663298624003e-05,
"loss": 1.9034,
"step": 300
},
{
"epoch": 0.16999930328154392,
"grad_norm": 2.3761579990386963,
"learning_rate": 9.855245748948326e-05,
"loss": 1.9687,
"step": 305
},
{
"epoch": 0.17278617710583152,
"grad_norm": 2.1475725173950195,
"learning_rate": 9.843399205260068e-05,
"loss": 1.9903,
"step": 310
},
{
"epoch": 0.17557305093011913,
"grad_norm": 2.40266752243042,
"learning_rate": 9.831094475858652e-05,
"loss": 1.9729,
"step": 315
},
{
"epoch": 0.17835992475440673,
"grad_norm": 2.187901258468628,
"learning_rate": 9.818332724771857e-05,
"loss": 2.0002,
"step": 320
},
{
"epoch": 0.18114679857869434,
"grad_norm": 2.374246597290039,
"learning_rate": 9.805115159261726e-05,
"loss": 1.8938,
"step": 325
},
{
"epoch": 0.18393367240298195,
"grad_norm": 2.2676525115966797,
"learning_rate": 9.791443029710361e-05,
"loss": 1.8728,
"step": 330
},
{
"epoch": 0.18672054622726955,
"grad_norm": 2.175401449203491,
"learning_rate": 9.777317629501636e-05,
"loss": 1.9195,
"step": 335
},
{
"epoch": 0.18950742005155716,
"grad_norm": 2.104386329650879,
"learning_rate": 9.762740294898846e-05,
"loss": 1.8618,
"step": 340
},
{
"epoch": 0.19229429387584476,
"grad_norm": 2.2644734382629395,
"learning_rate": 9.747712404918286e-05,
"loss": 1.8384,
"step": 345
},
{
"epoch": 0.19508116770013237,
"grad_norm": 2.6918885707855225,
"learning_rate": 9.732235381198813e-05,
"loss": 1.9305,
"step": 350
},
{
"epoch": 0.19786804152441997,
"grad_norm": 2.374101161956787,
"learning_rate": 9.716310687867342e-05,
"loss": 1.7964,
"step": 355
},
{
"epoch": 0.20065491534870758,
"grad_norm": 2.401925802230835,
"learning_rate": 9.699939831400351e-05,
"loss": 2.0372,
"step": 360
},
{
"epoch": 0.20344178917299519,
"grad_norm": 2.3377015590667725,
"learning_rate": 9.683124360481364e-05,
"loss": 1.9354,
"step": 365
},
{
"epoch": 0.2062286629972828,
"grad_norm": 2.669875144958496,
"learning_rate": 9.665865865854445e-05,
"loss": 1.9243,
"step": 370
},
{
"epoch": 0.2090155368215704,
"grad_norm": 2.251633882522583,
"learning_rate": 9.648165980173712e-05,
"loss": 1.9379,
"step": 375
},
{
"epoch": 0.211802410645858,
"grad_norm": 2.413219928741455,
"learning_rate": 9.630026377848892e-05,
"loss": 1.9232,
"step": 380
},
{
"epoch": 0.2145892844701456,
"grad_norm": 2.4033939838409424,
"learning_rate": 9.611448774886924e-05,
"loss": 1.9206,
"step": 385
},
{
"epoch": 0.2173761582944332,
"grad_norm": 2.7085165977478027,
"learning_rate": 9.592434928729616e-05,
"loss": 1.965,
"step": 390
},
{
"epoch": 0.22016303211872082,
"grad_norm": 2.440049409866333,
"learning_rate": 9.572986638087396e-05,
"loss": 1.938,
"step": 395
},
{
"epoch": 0.22294990594300843,
"grad_norm": 2.2908005714416504,
"learning_rate": 9.553105742769154e-05,
"loss": 1.8841,
"step": 400
},
{
"epoch": 0.22573677976729603,
"grad_norm": 2.0851612091064453,
"learning_rate": 9.532794123508197e-05,
"loss": 1.7699,
"step": 405
},
{
"epoch": 0.22852365359158364,
"grad_norm": 2.466432809829712,
"learning_rate": 9.512053701784329e-05,
"loss": 1.9162,
"step": 410
},
{
"epoch": 0.23131052741587124,
"grad_norm": 2.130993127822876,
"learning_rate": 9.490886439642081e-05,
"loss": 1.9023,
"step": 415
},
{
"epoch": 0.23409740124015885,
"grad_norm": 2.220752716064453,
"learning_rate": 9.469294339505098e-05,
"loss": 1.975,
"step": 420
},
{
"epoch": 0.23688427506444645,
"grad_norm": 2.1913466453552246,
"learning_rate": 9.447279443986716e-05,
"loss": 1.8632,
"step": 425
},
{
"epoch": 0.23967114888873406,
"grad_norm": 2.321277379989624,
"learning_rate": 9.424843835696724e-05,
"loss": 1.9294,
"step": 430
},
{
"epoch": 0.24245802271302166,
"grad_norm": 2.286395788192749,
"learning_rate": 9.401989637044355e-05,
"loss": 1.8428,
"step": 435
},
{
"epoch": 0.24524489653730927,
"grad_norm": 2.3919999599456787,
"learning_rate": 9.3787190100375e-05,
"loss": 1.8813,
"step": 440
},
{
"epoch": 0.24803177036159688,
"grad_norm": 2.1720130443573,
"learning_rate": 9.355034156078188e-05,
"loss": 1.8569,
"step": 445
},
{
"epoch": 0.2508186441858845,
"grad_norm": 2.086071014404297,
"learning_rate": 9.330937315754329e-05,
"loss": 1.8796,
"step": 450
},
{
"epoch": 0.2536055180101721,
"grad_norm": 2.447908639907837,
"learning_rate": 9.306430768627753e-05,
"loss": 1.851,
"step": 455
},
{
"epoch": 0.2563923918344597,
"grad_norm": 2.631868362426758,
"learning_rate": 9.281516833018571e-05,
"loss": 1.8706,
"step": 460
},
{
"epoch": 0.2591792656587473,
"grad_norm": 2.198009729385376,
"learning_rate": 9.256197865785854e-05,
"loss": 1.9986,
"step": 465
},
{
"epoch": 0.26196613948303493,
"grad_norm": 2.151010513305664,
"learning_rate": 9.230476262104677e-05,
"loss": 1.8895,
"step": 470
},
{
"epoch": 0.26475301330732254,
"grad_norm": 2.335463047027588,
"learning_rate": 9.204354455239539e-05,
"loss": 1.9415,
"step": 475
},
{
"epoch": 0.26753988713161014,
"grad_norm": 2.4173576831817627,
"learning_rate": 9.177834916314165e-05,
"loss": 1.8676,
"step": 480
},
{
"epoch": 0.27032676095589775,
"grad_norm": 2.409341812133789,
"learning_rate": 9.150920154077754e-05,
"loss": 1.879,
"step": 485
},
{
"epoch": 0.27311363478018535,
"grad_norm": 2.496391534805298,
"learning_rate": 9.123612714667634e-05,
"loss": 1.9162,
"step": 490
},
{
"epoch": 0.2759005086044729,
"grad_norm": 2.2236454486846924,
"learning_rate": 9.095915181368412e-05,
"loss": 1.9825,
"step": 495
},
{
"epoch": 0.2786873824287605,
"grad_norm": 2.4536592960357666,
"learning_rate": 9.067830174367586e-05,
"loss": 1.933,
"step": 500
},
{
"epoch": 0.2814742562530481,
"grad_norm": 2.130772829055786,
"learning_rate": 9.039360350507679e-05,
"loss": 1.901,
"step": 505
},
{
"epoch": 0.2842611300773357,
"grad_norm": 2.2794065475463867,
"learning_rate": 9.010508403034898e-05,
"loss": 1.8599,
"step": 510
},
{
"epoch": 0.2870480039016233,
"grad_norm": 2.342435598373413,
"learning_rate": 8.98127706134436e-05,
"loss": 1.8981,
"step": 515
},
{
"epoch": 0.28983487772591093,
"grad_norm": 2.3479907512664795,
"learning_rate": 8.951669090721881e-05,
"loss": 1.9312,
"step": 520
},
{
"epoch": 0.29262175155019854,
"grad_norm": 2.5833990573883057,
"learning_rate": 8.921687292082393e-05,
"loss": 1.8288,
"step": 525
},
{
"epoch": 0.29540862537448614,
"grad_norm": 2.2673983573913574,
"learning_rate": 8.891334501704962e-05,
"loss": 1.9026,
"step": 530
},
{
"epoch": 0.29819549919877375,
"grad_norm": 2.6991782188415527,
"learning_rate": 8.86061359096449e-05,
"loss": 1.8369,
"step": 535
},
{
"epoch": 0.30098237302306136,
"grad_norm": 2.9269566535949707,
"learning_rate": 8.829527466060072e-05,
"loss": 1.7069,
"step": 540
},
{
"epoch": 0.30376924684734896,
"grad_norm": 2.2337591648101807,
"learning_rate": 8.798079067740077e-05,
"loss": 1.994,
"step": 545
},
{
"epoch": 0.30655612067163657,
"grad_norm": 2.3603134155273438,
"learning_rate": 8.766271371023948e-05,
"loss": 1.8576,
"step": 550
},
{
"epoch": 0.3093429944959242,
"grad_norm": 2.4910824298858643,
"learning_rate": 8.73410738492077e-05,
"loss": 1.8134,
"step": 555
},
{
"epoch": 0.3121298683202118,
"grad_norm": 2.3701236248016357,
"learning_rate": 8.701590152144612e-05,
"loss": 1.8629,
"step": 560
},
{
"epoch": 0.3149167421444994,
"grad_norm": 2.594458818435669,
"learning_rate": 8.668722748826693e-05,
"loss": 1.8344,
"step": 565
},
{
"epoch": 0.317703615968787,
"grad_norm": 2.4443206787109375,
"learning_rate": 8.635508284224371e-05,
"loss": 2.0025,
"step": 570
},
{
"epoch": 0.3204904897930746,
"grad_norm": 2.4144582748413086,
"learning_rate": 8.601949900427016e-05,
"loss": 1.8186,
"step": 575
},
{
"epoch": 0.3232773636173622,
"grad_norm": 2.4722774028778076,
"learning_rate": 8.568050772058762e-05,
"loss": 1.8776,
"step": 580
},
{
"epoch": 0.3260642374416498,
"grad_norm": 2.769221544265747,
"learning_rate": 8.533814105978191e-05,
"loss": 1.86,
"step": 585
},
{
"epoch": 0.3288511112659374,
"grad_norm": 2.3548312187194824,
"learning_rate": 8.499243140974966e-05,
"loss": 1.8534,
"step": 590
},
{
"epoch": 0.331637985090225,
"grad_norm": 2.4437191486358643,
"learning_rate": 8.464341147463431e-05,
"loss": 1.865,
"step": 595
},
{
"epoch": 0.3344248589145126,
"grad_norm": 2.4781646728515625,
"learning_rate": 8.429111427173241e-05,
"loss": 1.7592,
"step": 600
},
{
"epoch": 0.33721173273880023,
"grad_norm": 2.605130910873413,
"learning_rate": 8.393557312837018e-05,
"loss": 1.7903,
"step": 605
},
{
"epoch": 0.33999860656308784,
"grad_norm": 2.6218454837799072,
"learning_rate": 8.357682167875062e-05,
"loss": 1.8664,
"step": 610
},
{
"epoch": 0.34278548038737544,
"grad_norm": 2.7091429233551025,
"learning_rate": 8.321489386077192e-05,
"loss": 1.87,
"step": 615
},
{
"epoch": 0.34557235421166305,
"grad_norm": 2.571959972381592,
"learning_rate": 8.28498239128167e-05,
"loss": 1.852,
"step": 620
},
{
"epoch": 0.34835922803595065,
"grad_norm": 2.72074031829834,
"learning_rate": 8.248164637051321e-05,
"loss": 1.8048,
"step": 625
},
{
"epoch": 0.35114610186023826,
"grad_norm": 2.7408368587493896,
"learning_rate": 8.211039606346826e-05,
"loss": 1.9272,
"step": 630
},
{
"epoch": 0.35393297568452586,
"grad_norm": 2.6123206615448,
"learning_rate": 8.173610811197226e-05,
"loss": 1.8601,
"step": 635
},
{
"epoch": 0.35671984950881347,
"grad_norm": 2.283332347869873,
"learning_rate": 8.135881792367686e-05,
"loss": 1.8755,
"step": 640
},
{
"epoch": 0.3595067233331011,
"grad_norm": 2.5847232341766357,
"learning_rate": 8.097856119024545e-05,
"loss": 1.9465,
"step": 645
},
{
"epoch": 0.3622935971573887,
"grad_norm": 2.7280638217926025,
"learning_rate": 8.059537388397665e-05,
"loss": 1.8577,
"step": 650
},
{
"epoch": 0.3650804709816763,
"grad_norm": 2.4492626190185547,
"learning_rate": 8.020929225440137e-05,
"loss": 1.8689,
"step": 655
},
{
"epoch": 0.3678673448059639,
"grad_norm": 2.558868169784546,
"learning_rate": 7.98203528248536e-05,
"loss": 1.8553,
"step": 660
},
{
"epoch": 0.3706542186302515,
"grad_norm": 2.6071197986602783,
"learning_rate": 7.942859238901528e-05,
"loss": 1.9035,
"step": 665
},
{
"epoch": 0.3734410924545391,
"grad_norm": 2.6189825534820557,
"learning_rate": 7.903404800743564e-05,
"loss": 1.9362,
"step": 670
},
{
"epoch": 0.3762279662788267,
"grad_norm": 2.8328700065612793,
"learning_rate": 7.863675700402526e-05,
"loss": 1.8079,
"step": 675
},
{
"epoch": 0.3790148401031143,
"grad_norm": 2.458878755569458,
"learning_rate": 7.823675696252524e-05,
"loss": 1.9774,
"step": 680
},
{
"epoch": 0.3818017139274019,
"grad_norm": 2.524718761444092,
"learning_rate": 7.783408572295174e-05,
"loss": 1.9575,
"step": 685
},
{
"epoch": 0.3845885877516895,
"grad_norm": 2.4117393493652344,
"learning_rate": 7.742878137801639e-05,
"loss": 1.8446,
"step": 690
},
{
"epoch": 0.38737546157597713,
"grad_norm": 2.6879658699035645,
"learning_rate": 7.702088226952258e-05,
"loss": 1.8512,
"step": 695
},
{
"epoch": 0.39016233540026474,
"grad_norm": 2.5234720706939697,
"learning_rate": 7.661042698473853e-05,
"loss": 1.778,
"step": 700
},
{
"epoch": 0.39294920922455234,
"grad_norm": 2.4943764209747314,
"learning_rate": 7.619745435274667e-05,
"loss": 1.8087,
"step": 705
},
{
"epoch": 0.39573608304883995,
"grad_norm": 2.897895574569702,
"learning_rate": 7.578200344077073e-05,
"loss": 1.8764,
"step": 710
},
{
"epoch": 0.39852295687312755,
"grad_norm": 2.5875868797302246,
"learning_rate": 7.536411355047964e-05,
"loss": 1.8782,
"step": 715
},
{
"epoch": 0.40130983069741516,
"grad_norm": 2.7180755138397217,
"learning_rate": 7.494382421426984e-05,
"loss": 1.8987,
"step": 720
},
{
"epoch": 0.40409670452170277,
"grad_norm": 2.798145294189453,
"learning_rate": 7.452117519152542e-05,
"loss": 1.8426,
"step": 725
},
{
"epoch": 0.40688357834599037,
"grad_norm": 3.262134552001953,
"learning_rate": 7.409620646485685e-05,
"loss": 1.7929,
"step": 730
},
{
"epoch": 0.409670452170278,
"grad_norm": 2.6406302452087402,
"learning_rate": 7.36689582363187e-05,
"loss": 1.876,
"step": 735
},
{
"epoch": 0.4124573259945656,
"grad_norm": 2.56366229057312,
"learning_rate": 7.323947092360649e-05,
"loss": 1.7674,
"step": 740
},
{
"epoch": 0.4152441998188532,
"grad_norm": 2.5193862915039062,
"learning_rate": 7.280778515623314e-05,
"loss": 1.9506,
"step": 745
},
{
"epoch": 0.4180310736431408,
"grad_norm": 2.9591546058654785,
"learning_rate": 7.237394177168548e-05,
"loss": 1.8831,
"step": 750
},
{
"epoch": 0.4208179474674284,
"grad_norm": 2.7385146617889404,
"learning_rate": 7.193798181156095e-05,
"loss": 1.9302,
"step": 755
},
{
"epoch": 0.423604821291716,
"grad_norm": 2.6676039695739746,
"learning_rate": 7.149994651768514e-05,
"loss": 1.7349,
"step": 760
},
{
"epoch": 0.4263916951160036,
"grad_norm": 3.0209970474243164,
"learning_rate": 7.10598773282103e-05,
"loss": 1.8751,
"step": 765
},
{
"epoch": 0.4291785689402912,
"grad_norm": 2.7455992698669434,
"learning_rate": 7.061781587369519e-05,
"loss": 1.9342,
"step": 770
},
{
"epoch": 0.4319654427645788,
"grad_norm": 2.842349052429199,
"learning_rate": 7.017380397316695e-05,
"loss": 1.8249,
"step": 775
},
{
"epoch": 0.4347523165888664,
"grad_norm": 2.6208105087280273,
"learning_rate": 6.972788363016497e-05,
"loss": 1.7323,
"step": 780
},
{
"epoch": 0.43753919041315403,
"grad_norm": 2.757404088973999,
"learning_rate": 6.92800970287674e-05,
"loss": 1.8447,
"step": 785
},
{
"epoch": 0.44032606423744164,
"grad_norm": 2.525618553161621,
"learning_rate": 6.883048652960038e-05,
"loss": 1.8332,
"step": 790
},
{
"epoch": 0.44311293806172924,
"grad_norm": 2.5981028079986572,
"learning_rate": 6.837909466583095e-05,
"loss": 1.8783,
"step": 795
},
{
"epoch": 0.44589981188601685,
"grad_norm": 2.546555280685425,
"learning_rate": 6.792596413914324e-05,
"loss": 1.7139,
"step": 800
},
{
"epoch": 0.44868668571030446,
"grad_norm": 2.530876398086548,
"learning_rate": 6.747113781569892e-05,
"loss": 1.9029,
"step": 805
},
{
"epoch": 0.45147355953459206,
"grad_norm": 2.799030303955078,
"learning_rate": 6.701465872208216e-05,
"loss": 1.9114,
"step": 810
},
{
"epoch": 0.45426043335887967,
"grad_norm": 2.816466808319092,
"learning_rate": 6.655657004122916e-05,
"loss": 1.9019,
"step": 815
},
{
"epoch": 0.4570473071831673,
"grad_norm": 3.0766985416412354,
"learning_rate": 6.60969151083432e-05,
"loss": 1.809,
"step": 820
},
{
"epoch": 0.4598341810074549,
"grad_norm": 2.733771800994873,
"learning_rate": 6.563573740679496e-05,
"loss": 1.8111,
"step": 825
},
{
"epoch": 0.4626210548317425,
"grad_norm": 2.653470516204834,
"learning_rate": 6.517308056400917e-05,
"loss": 1.8637,
"step": 830
},
{
"epoch": 0.4654079286560301,
"grad_norm": 2.7409327030181885,
"learning_rate": 6.470898834733731e-05,
"loss": 1.8205,
"step": 835
},
{
"epoch": 0.4681948024803177,
"grad_norm": 2.6908485889434814,
"learning_rate": 6.42435046599173e-05,
"loss": 1.8235,
"step": 840
},
{
"epoch": 0.4709816763046053,
"grad_norm": 2.7610902786254883,
"learning_rate": 6.377667353652022e-05,
"loss": 1.8789,
"step": 845
},
{
"epoch": 0.4737685501288929,
"grad_norm": 2.7466371059417725,
"learning_rate": 6.330853913938466e-05,
"loss": 1.8,
"step": 850
},
{
"epoch": 0.4765554239531805,
"grad_norm": 2.6152100563049316,
"learning_rate": 6.283914575403888e-05,
"loss": 1.9228,
"step": 855
},
{
"epoch": 0.4793422977774681,
"grad_norm": 2.903139114379883,
"learning_rate": 6.236853778511156e-05,
"loss": 1.8079,
"step": 860
},
{
"epoch": 0.4821291716017557,
"grad_norm": 2.941880941390991,
"learning_rate": 6.189675975213094e-05,
"loss": 1.8818,
"step": 865
},
{
"epoch": 0.48491604542604333,
"grad_norm": 2.928757429122925,
"learning_rate": 6.142385628531342e-05,
"loss": 1.883,
"step": 870
},
{
"epoch": 0.48770291925033094,
"grad_norm": 3.0800089836120605,
"learning_rate": 6.09498721213414e-05,
"loss": 1.7964,
"step": 875
},
{
"epoch": 0.49048979307461854,
"grad_norm": 2.787975549697876,
"learning_rate": 6.047485209913137e-05,
"loss": 1.9406,
"step": 880
},
{
"epoch": 0.49327666689890615,
"grad_norm": 2.5201809406280518,
"learning_rate": 5.999884115559192e-05,
"loss": 1.8103,
"step": 885
},
{
"epoch": 0.49606354072319375,
"grad_norm": 2.659198760986328,
"learning_rate": 5.952188432137293e-05,
"loss": 1.879,
"step": 890
},
{
"epoch": 0.49885041454748136,
"grad_norm": 2.3848142623901367,
"learning_rate": 5.90440267166055e-05,
"loss": 1.7912,
"step": 895
},
{
"epoch": 0.501637288371769,
"grad_norm": 3.1244029998779297,
"learning_rate": 5.8565313546633684e-05,
"loss": 1.8467,
"step": 900
},
{
"epoch": 0.5044241621960566,
"grad_norm": 2.36189866065979,
"learning_rate": 5.8085790097738025e-05,
"loss": 1.8748,
"step": 905
},
{
"epoch": 0.5072110360203442,
"grad_norm": 2.6590969562530518,
"learning_rate": 5.7605501732851475e-05,
"loss": 1.8892,
"step": 910
},
{
"epoch": 0.5099979098446318,
"grad_norm": 2.493211030960083,
"learning_rate": 5.712449388726807e-05,
"loss": 1.9295,
"step": 915
},
{
"epoch": 0.5127847836689194,
"grad_norm": 2.8434841632843018,
"learning_rate": 5.664281206434472e-05,
"loss": 1.7492,
"step": 920
},
{
"epoch": 0.515571657493207,
"grad_norm": 3.066373348236084,
"learning_rate": 5.616050183119663e-05,
"loss": 1.9097,
"step": 925
},
{
"epoch": 0.5183585313174947,
"grad_norm": 2.5913686752319336,
"learning_rate": 5.5677608814386616e-05,
"loss": 1.7054,
"step": 930
},
{
"epoch": 0.5211454051417822,
"grad_norm": 2.7032017707824707,
"learning_rate": 5.519417869560889e-05,
"loss": 1.816,
"step": 935
},
{
"epoch": 0.5239322789660699,
"grad_norm": 3.0254099369049072,
"learning_rate": 5.471025720736747e-05,
"loss": 1.8939,
"step": 940
},
{
"epoch": 0.5267191527903574,
"grad_norm": 3.013810873031616,
"learning_rate": 5.422589012864996e-05,
"loss": 1.8953,
"step": 945
},
{
"epoch": 0.5295060266146451,
"grad_norm": 2.9430110454559326,
"learning_rate": 5.3741123280596864e-05,
"loss": 1.8232,
"step": 950
},
{
"epoch": 0.5322929004389326,
"grad_norm": 2.7250967025756836,
"learning_rate": 5.325600252216685e-05,
"loss": 1.8462,
"step": 955
},
{
"epoch": 0.5350797742632203,
"grad_norm": 2.7945637702941895,
"learning_rate": 5.27705737457985e-05,
"loss": 1.7108,
"step": 960
},
{
"epoch": 0.5378666480875078,
"grad_norm": 2.6481263637542725,
"learning_rate": 5.228488287306896e-05,
"loss": 1.8026,
"step": 965
},
{
"epoch": 0.5406535219117955,
"grad_norm": 2.819915771484375,
"learning_rate": 5.179897585034963e-05,
"loss": 1.8175,
"step": 970
},
{
"epoch": 0.543440395736083,
"grad_norm": 3.0015969276428223,
"learning_rate": 5.1312898644459776e-05,
"loss": 1.7393,
"step": 975
},
{
"epoch": 0.5462272695603707,
"grad_norm": 3.3672678470611572,
"learning_rate": 5.0826697238317935e-05,
"loss": 1.8442,
"step": 980
},
{
"epoch": 0.5490141433846583,
"grad_norm": 2.963168144226074,
"learning_rate": 5.0340417626592016e-05,
"loss": 1.8713,
"step": 985
},
{
"epoch": 0.5518010172089458,
"grad_norm": 3.0407772064208984,
"learning_rate": 4.9854105811348216e-05,
"loss": 1.8686,
"step": 990
},
{
"epoch": 0.5545878910332335,
"grad_norm": 3.218909978866577,
"learning_rate": 4.936780779769913e-05,
"loss": 1.8486,
"step": 995
},
{
"epoch": 0.557374764857521,
"grad_norm": 2.7832529544830322,
"learning_rate": 4.888156958945174e-05,
"loss": 1.8146,
"step": 1000
},
{
"epoch": 0.5601616386818087,
"grad_norm": 2.852426290512085,
"learning_rate": 4.839543718475543e-05,
"loss": 1.8197,
"step": 1005
},
{
"epoch": 0.5629485125060962,
"grad_norm": 3.1005451679229736,
"learning_rate": 4.790945657175061e-05,
"loss": 1.8336,
"step": 1010
},
{
"epoch": 0.5657353863303839,
"grad_norm": 2.834667921066284,
"learning_rate": 4.742367372421811e-05,
"loss": 1.909,
"step": 1015
},
{
"epoch": 0.5685222601546714,
"grad_norm": 3.0797226428985596,
"learning_rate": 4.69381345972302e-05,
"loss": 1.7971,
"step": 1020
},
{
"epoch": 0.5713091339789591,
"grad_norm": 3.210432291030884,
"learning_rate": 4.6452885122803205e-05,
"loss": 1.7896,
"step": 1025
},
{
"epoch": 0.5740960078032467,
"grad_norm": 2.669483184814453,
"learning_rate": 4.5967971205552194e-05,
"loss": 1.8121,
"step": 1030
},
{
"epoch": 0.5768828816275343,
"grad_norm": 2.850985288619995,
"learning_rate": 4.548343871834864e-05,
"loss": 1.8144,
"step": 1035
},
{
"epoch": 0.5796697554518219,
"grad_norm": 2.8982150554656982,
"learning_rate": 4.499933349798067e-05,
"loss": 1.8813,
"step": 1040
},
{
"epoch": 0.5824566292761095,
"grad_norm": 2.7808709144592285,
"learning_rate": 4.451570134081694e-05,
"loss": 1.9348,
"step": 1045
},
{
"epoch": 0.5852435031003971,
"grad_norm": 2.872607946395874,
"learning_rate": 4.403258799847433e-05,
"loss": 1.7817,
"step": 1050
},
{
"epoch": 0.5880303769246847,
"grad_norm": 3.2354702949523926,
"learning_rate": 4.3550039173489845e-05,
"loss": 1.7557,
"step": 1055
},
{
"epoch": 0.5908172507489723,
"grad_norm": 3.2129600048065186,
"learning_rate": 4.306810051499708e-05,
"loss": 1.7631,
"step": 1060
},
{
"epoch": 0.59360412457326,
"grad_norm": 3.0379066467285156,
"learning_rate": 4.2586817614407895e-05,
"loss": 1.8084,
"step": 1065
},
{
"epoch": 0.5963909983975475,
"grad_norm": 2.7816271781921387,
"learning_rate": 4.210623600109946e-05,
"loss": 1.6655,
"step": 1070
},
{
"epoch": 0.5991778722218352,
"grad_norm": 3.1788153648376465,
"learning_rate": 4.162640113810706e-05,
"loss": 1.8399,
"step": 1075
},
{
"epoch": 0.6019647460461227,
"grad_norm": 2.919539213180542,
"learning_rate": 4.114735841782347e-05,
"loss": 1.8972,
"step": 1080
},
{
"epoch": 0.6047516198704104,
"grad_norm": 2.948122978210449,
"learning_rate": 4.06691531577047e-05,
"loss": 1.8159,
"step": 1085
},
{
"epoch": 0.6075384936946979,
"grad_norm": 2.851313591003418,
"learning_rate": 4.019183059598296e-05,
"loss": 1.7902,
"step": 1090
},
{
"epoch": 0.6103253675189856,
"grad_norm": 2.989866256713867,
"learning_rate": 3.971543588738724e-05,
"loss": 1.8125,
"step": 1095
},
{
"epoch": 0.6131122413432731,
"grad_norm": 2.9144411087036133,
"learning_rate": 3.924001409887158e-05,
"loss": 1.7389,
"step": 1100
},
{
"epoch": 0.6158991151675608,
"grad_norm": 3.15948224067688,
"learning_rate": 3.87656102053517e-05,
"loss": 1.8098,
"step": 1105
},
{
"epoch": 0.6186859889918483,
"grad_norm": 3.0768699645996094,
"learning_rate": 3.8292269085450474e-05,
"loss": 1.7795,
"step": 1110
},
{
"epoch": 0.621472862816136,
"grad_norm": 2.7446067333221436,
"learning_rate": 3.782003551725236e-05,
"loss": 1.8634,
"step": 1115
},
{
"epoch": 0.6242597366404236,
"grad_norm": 2.785001277923584,
"learning_rate": 3.734895417406734e-05,
"loss": 1.8444,
"step": 1120
},
{
"epoch": 0.6270466104647112,
"grad_norm": 2.594785690307617,
"learning_rate": 3.687906962020491e-05,
"loss": 1.8379,
"step": 1125
},
{
"epoch": 0.6298334842889988,
"grad_norm": 3.196511745452881,
"learning_rate": 3.641042630675829e-05,
"loss": 1.8797,
"step": 1130
},
{
"epoch": 0.6326203581132864,
"grad_norm": 2.786207437515259,
"learning_rate": 3.594306856739924e-05,
"loss": 1.7519,
"step": 1135
},
{
"epoch": 0.635407231937574,
"grad_norm": 3.124060869216919,
"learning_rate": 3.547704061418424e-05,
"loss": 1.783,
"step": 1140
},
{
"epoch": 0.6381941057618616,
"grad_norm": 3.125051498413086,
"learning_rate": 3.501238653337194e-05,
"loss": 1.8377,
"step": 1145
},
{
"epoch": 0.6409809795861492,
"grad_norm": 3.672471284866333,
"learning_rate": 3.4549150281252636e-05,
"loss": 1.886,
"step": 1150
},
{
"epoch": 0.6437678534104369,
"grad_norm": 2.9537386894226074,
"learning_rate": 3.408737567998993e-05,
"loss": 1.7076,
"step": 1155
},
{
"epoch": 0.6465547272347244,
"grad_norm": 2.8255364894866943,
"learning_rate": 3.362710641347524e-05,
"loss": 1.8263,
"step": 1160
},
{
"epoch": 0.6493416010590121,
"grad_norm": 2.87804913520813,
"learning_rate": 3.316838602319532e-05,
"loss": 1.8526,
"step": 1165
},
{
"epoch": 0.6521284748832996,
"grad_norm": 3.53425669670105,
"learning_rate": 3.271125790411309e-05,
"loss": 1.6949,
"step": 1170
},
{
"epoch": 0.6549153487075873,
"grad_norm": 3.1044154167175293,
"learning_rate": 3.225576530056264e-05,
"loss": 1.8605,
"step": 1175
},
{
"epoch": 0.6577022225318748,
"grad_norm": 2.7356083393096924,
"learning_rate": 3.180195130215824e-05,
"loss": 1.8447,
"step": 1180
},
{
"epoch": 0.6604890963561625,
"grad_norm": 2.868706703186035,
"learning_rate": 3.1349858839717986e-05,
"loss": 1.7386,
"step": 1185
},
{
"epoch": 0.66327597018045,
"grad_norm": 2.992685556411743,
"learning_rate": 3.089953068120271e-05,
"loss": 1.7356,
"step": 1190
},
{
"epoch": 0.6660628440047377,
"grad_norm": 3.2466917037963867,
"learning_rate": 3.0451009427669986e-05,
"loss": 1.7546,
"step": 1195
},
{
"epoch": 0.6688497178290252,
"grad_norm": 3.205050230026245,
"learning_rate": 3.000433750924414e-05,
"loss": 1.844,
"step": 1200
},
{
"epoch": 0.6716365916533129,
"grad_norm": 2.86040997505188,
"learning_rate": 2.9559557181102315e-05,
"loss": 1.823,
"step": 1205
},
{
"epoch": 0.6744234654776005,
"grad_norm": 3.5542171001434326,
"learning_rate": 2.911671051947722e-05,
"loss": 1.741,
"step": 1210
},
{
"epoch": 0.6772103393018881,
"grad_norm": 3.017136812210083,
"learning_rate": 2.867583941767657e-05,
"loss": 1.7154,
"step": 1215
},
{
"epoch": 0.6799972131261757,
"grad_norm": 2.9082391262054443,
"learning_rate": 2.823698558212009e-05,
"loss": 1.8659,
"step": 1220
},
{
"epoch": 0.6827840869504633,
"grad_norm": 2.9137349128723145,
"learning_rate": 2.7800190528394122e-05,
"loss": 1.78,
"step": 1225
},
{
"epoch": 0.6855709607747509,
"grad_norm": 2.919658899307251,
"learning_rate": 2.736549557732405e-05,
"loss": 1.7336,
"step": 1230
},
{
"epoch": 0.6883578345990385,
"grad_norm": 2.902566909790039,
"learning_rate": 2.693294185106562e-05,
"loss": 2.0199,
"step": 1235
},
{
"epoch": 0.6911447084233261,
"grad_norm": 3.1913843154907227,
"learning_rate": 2.650257026921455e-05,
"loss": 1.8186,
"step": 1240
},
{
"epoch": 0.6939315822476138,
"grad_norm": 3.2584824562072754,
"learning_rate": 2.607442154493568e-05,
"loss": 1.8856,
"step": 1245
},
{
"epoch": 0.6967184560719013,
"grad_norm": 3.0244908332824707,
"learning_rate": 2.5648536181111438e-05,
"loss": 1.7954,
"step": 1250
},
{
"epoch": 0.699505329896189,
"grad_norm": 3.0413196086883545,
"learning_rate": 2.5224954466510274e-05,
"loss": 1.7916,
"step": 1255
},
{
"epoch": 0.7022922037204765,
"grad_norm": 3.007747173309326,
"learning_rate": 2.480371647197538e-05,
"loss": 1.8692,
"step": 1260
},
{
"epoch": 0.7050790775447642,
"grad_norm": 3.2990870475769043,
"learning_rate": 2.438486204663391e-05,
"loss": 1.8834,
"step": 1265
},
{
"epoch": 0.7078659513690517,
"grad_norm": 2.8733415603637695,
"learning_rate": 2.3968430814127385e-05,
"loss": 1.8423,
"step": 1270
},
{
"epoch": 0.7106528251933394,
"grad_norm": 3.1287338733673096,
"learning_rate": 2.3554462168863085e-05,
"loss": 1.8361,
"step": 1275
},
{
"epoch": 0.7134396990176269,
"grad_norm": 3.009936809539795,
"learning_rate": 2.314299527228759e-05,
"loss": 1.8482,
"step": 1280
},
{
"epoch": 0.7162265728419146,
"grad_norm": 3.00722074508667,
"learning_rate": 2.2734069049181882e-05,
"loss": 1.7856,
"step": 1285
},
{
"epoch": 0.7190134466662021,
"grad_norm": 3.005175828933716,
"learning_rate": 2.2327722183979212e-05,
"loss": 1.8073,
"step": 1290
},
{
"epoch": 0.7218003204904898,
"grad_norm": 3.111478090286255,
"learning_rate": 2.1923993117105462e-05,
"loss": 1.7625,
"step": 1295
},
{
"epoch": 0.7245871943147774,
"grad_norm": 3.3276472091674805,
"learning_rate": 2.1522920041342704e-05,
"loss": 1.7659,
"step": 1300
},
{
"epoch": 0.727374068139065,
"grad_norm": 3.3914248943328857,
"learning_rate": 2.1124540898216248e-05,
"loss": 1.8393,
"step": 1305
},
{
"epoch": 0.7301609419633526,
"grad_norm": 2.898927927017212,
"learning_rate": 2.0728893374405166e-05,
"loss": 1.6879,
"step": 1310
},
{
"epoch": 0.7329478157876402,
"grad_norm": 3.7136518955230713,
"learning_rate": 2.033601489817738e-05,
"loss": 1.9492,
"step": 1315
},
{
"epoch": 0.7357346896119278,
"grad_norm": 2.906088352203369,
"learning_rate": 1.9945942635848748e-05,
"loss": 1.8337,
"step": 1320
},
{
"epoch": 0.7385215634362154,
"grad_norm": 2.833622694015503,
"learning_rate": 1.9558713488267238e-05,
"loss": 1.8229,
"step": 1325
},
{
"epoch": 0.741308437260503,
"grad_norm": 3.2317726612091064,
"learning_rate": 1.917436408732208e-05,
"loss": 1.831,
"step": 1330
},
{
"epoch": 0.7440953110847907,
"grad_norm": 3.6513025760650635,
"learning_rate": 1.8792930792478357e-05,
"loss": 1.7859,
"step": 1335
},
{
"epoch": 0.7468821849090782,
"grad_norm": 3.0361874103546143,
"learning_rate": 1.8414449687337464e-05,
"loss": 1.714,
"step": 1340
},
{
"epoch": 0.7496690587333659,
"grad_norm": 2.797312021255493,
"learning_rate": 1.8038956576223504e-05,
"loss": 1.6919,
"step": 1345
},
{
"epoch": 0.7524559325576534,
"grad_norm": 2.9055614471435547,
"learning_rate": 1.766648698079635e-05,
"loss": 1.7895,
"step": 1350
},
{
"epoch": 0.7552428063819411,
"grad_norm": 3.578946828842163,
"learning_rate": 1.7297076136691072e-05,
"loss": 1.9271,
"step": 1355
},
{
"epoch": 0.7580296802062286,
"grad_norm": 3.3223307132720947,
"learning_rate": 1.6930758990184875e-05,
"loss": 1.8167,
"step": 1360
},
{
"epoch": 0.7608165540305163,
"grad_norm": 3.483105182647705,
"learning_rate": 1.6567570194891024e-05,
"loss": 1.7878,
"step": 1365
},
{
"epoch": 0.7636034278548038,
"grad_norm": 3.632483720779419,
"learning_rate": 1.620754410848069e-05,
"loss": 1.8081,
"step": 1370
},
{
"epoch": 0.7663903016790915,
"grad_norm": 2.8937482833862305,
"learning_rate": 1.5850714789432663e-05,
"loss": 1.8768,
"step": 1375
},
{
"epoch": 0.769177175503379,
"grad_norm": 3.270364284515381,
"learning_rate": 1.549711599381145e-05,
"loss": 1.8446,
"step": 1380
},
{
"epoch": 0.7719640493276667,
"grad_norm": 3.1240315437316895,
"learning_rate": 1.5146781172073959e-05,
"loss": 1.7562,
"step": 1385
},
{
"epoch": 0.7747509231519543,
"grad_norm": 3.0998189449310303,
"learning_rate": 1.479974346590503e-05,
"loss": 1.8045,
"step": 1390
},
{
"epoch": 0.7775377969762419,
"grad_norm": 2.866483688354492,
"learning_rate": 1.4456035705082349e-05,
"loss": 1.773,
"step": 1395
},
{
"epoch": 0.7803246708005295,
"grad_norm": 3.3193795680999756,
"learning_rate": 1.4115690404370551e-05,
"loss": 1.781,
"step": 1400
},
{
"epoch": 0.7831115446248171,
"grad_norm": 3.3579161167144775,
"learning_rate": 1.3778739760445552e-05,
"loss": 1.765,
"step": 1405
},
{
"epoch": 0.7858984184491047,
"grad_norm": 2.840357542037964,
"learning_rate": 1.344521564884858e-05,
"loss": 1.8955,
"step": 1410
},
{
"epoch": 0.7886852922733923,
"grad_norm": 2.9612786769866943,
"learning_rate": 1.3115149620970795e-05,
"loss": 1.8964,
"step": 1415
},
{
"epoch": 0.7914721660976799,
"grad_norm": 3.50901460647583,
"learning_rate": 1.2788572901068552e-05,
"loss": 1.7923,
"step": 1420
},
{
"epoch": 0.7942590399219676,
"grad_norm": 3.2638587951660156,
"learning_rate": 1.2465516383309551e-05,
"loss": 1.8114,
"step": 1425
},
{
"epoch": 0.7970459137462551,
"grad_norm": 3.109443187713623,
"learning_rate": 1.2146010628850268e-05,
"loss": 1.8885,
"step": 1430
},
{
"epoch": 0.7998327875705428,
"grad_norm": 2.9360194206237793,
"learning_rate": 1.183008586294485e-05,
"loss": 1.7705,
"step": 1435
},
{
"epoch": 0.8026196613948303,
"grad_norm": 3.056124448776245,
"learning_rate": 1.151777197208585e-05,
"loss": 1.7987,
"step": 1440
},
{
"epoch": 0.805406535219118,
"grad_norm": 3.191528558731079,
"learning_rate": 1.1209098501176896e-05,
"loss": 1.752,
"step": 1445
},
{
"epoch": 0.8081934090434055,
"grad_norm": 3.37919020652771,
"learning_rate": 1.0904094650737795e-05,
"loss": 1.7828,
"step": 1450
},
{
"epoch": 0.8109802828676932,
"grad_norm": 3.4256997108459473,
"learning_rate": 1.0602789274142133e-05,
"loss": 1.8676,
"step": 1455
},
{
"epoch": 0.8137671566919807,
"grad_norm": 3.1404809951782227,
"learning_rate": 1.0305210874887766e-05,
"loss": 1.6705,
"step": 1460
},
{
"epoch": 0.8165540305162684,
"grad_norm": 3.3049511909484863,
"learning_rate": 1.0011387603900385e-05,
"loss": 1.7632,
"step": 1465
},
{
"epoch": 0.819340904340556,
"grad_norm": 3.3583621978759766,
"learning_rate": 9.7213472568704e-06,
"loss": 1.7868,
"step": 1470
},
{
"epoch": 0.8221277781648436,
"grad_norm": 3.3916678428649902,
"learning_rate": 9.435117271623566e-06,
"loss": 1.6243,
"step": 1475
},
{
"epoch": 0.8249146519891312,
"grad_norm": 3.367898464202881,
"learning_rate": 9.152724725525202e-06,
"loss": 1.8333,
"step": 1480
},
{
"epoch": 0.8277015258134188,
"grad_norm": 2.8452632427215576,
"learning_rate": 8.87419633291886e-06,
"loss": 1.8044,
"step": 1485
},
{
"epoch": 0.8304883996377064,
"grad_norm": 3.5473368167877197,
"learning_rate": 8.599558442598998e-06,
"loss": 1.7415,
"step": 1490
},
{
"epoch": 0.833275273461994,
"grad_norm": 3.542752504348755,
"learning_rate": 8.328837035318448e-06,
"loss": 1.8796,
"step": 1495
},
{
"epoch": 0.8360621472862816,
"grad_norm": 2.9325203895568848,
"learning_rate": 8.06205772133063e-06,
"loss": 1.7187,
"step": 1500
},
{
"epoch": 0.8388490211105692,
"grad_norm": 3.436312198638916,
"learning_rate": 7.799245737966821e-06,
"loss": 1.7763,
"step": 1505
},
{
"epoch": 0.8416358949348568,
"grad_norm": 3.262385129928589,
"learning_rate": 7.540425947248697e-06,
"loss": 1.791,
"step": 1510
},
{
"epoch": 0.8444227687591445,
"grad_norm": 3.1413116455078125,
"learning_rate": 7.28562283353637e-06,
"loss": 1.7876,
"step": 1515
},
{
"epoch": 0.847209642583432,
"grad_norm": 2.9553627967834473,
"learning_rate": 7.034860501212243e-06,
"loss": 1.8214,
"step": 1520
},
{
"epoch": 0.8499965164077197,
"grad_norm": 3.232959032058716,
"learning_rate": 6.788162672400583e-06,
"loss": 1.8821,
"step": 1525
},
{
"epoch": 0.8527833902320072,
"grad_norm": 2.9750733375549316,
"learning_rate": 6.5455526847235825e-06,
"loss": 1.8333,
"step": 1530
},
{
"epoch": 0.8555702640562949,
"grad_norm": 3.1739635467529297,
"learning_rate": 6.307053489093506e-06,
"loss": 1.7966,
"step": 1535
},
{
"epoch": 0.8583571378805824,
"grad_norm": 2.911165237426758,
"learning_rate": 6.072687647541553e-06,
"loss": 1.8175,
"step": 1540
},
{
"epoch": 0.8611440117048701,
"grad_norm": 3.130584955215454,
"learning_rate": 5.842477331083518e-06,
"loss": 1.8233,
"step": 1545
},
{
"epoch": 0.8639308855291576,
"grad_norm": 2.979978084564209,
"learning_rate": 5.616444317622388e-06,
"loss": 1.7002,
"step": 1550
},
{
"epoch": 0.8667177593534453,
"grad_norm": 3.0265040397644043,
"learning_rate": 5.394609989888161e-06,
"loss": 1.7969,
"step": 1555
},
{
"epoch": 0.8695046331777329,
"grad_norm": 4.560696601867676,
"learning_rate": 5.176995333415019e-06,
"loss": 1.8274,
"step": 1560
},
{
"epoch": 0.8722915070020205,
"grad_norm": 3.179011344909668,
"learning_rate": 4.963620934556168e-06,
"loss": 1.8228,
"step": 1565
},
{
"epoch": 0.8750783808263081,
"grad_norm": 2.9435315132141113,
"learning_rate": 4.754506978536227e-06,
"loss": 1.7943,
"step": 1570
},
{
"epoch": 0.8778652546505957,
"grad_norm": 2.801589250564575,
"learning_rate": 4.549673247541875e-06,
"loss": 1.804,
"step": 1575
},
{
"epoch": 0.8806521284748833,
"grad_norm": 3.0648934841156006,
"learning_rate": 4.3491391188503264e-06,
"loss": 1.7944,
"step": 1580
},
{
"epoch": 0.8834390022991709,
"grad_norm": 3.4953231811523438,
"learning_rate": 4.152923562996297e-06,
"loss": 1.7327,
"step": 1585
},
{
"epoch": 0.8862258761234585,
"grad_norm": 3.209007740020752,
"learning_rate": 3.961045141977376e-06,
"loss": 1.8297,
"step": 1590
},
{
"epoch": 0.8890127499477462,
"grad_norm": 3.367311954498291,
"learning_rate": 3.773522007498065e-06,
"loss": 1.8334,
"step": 1595
},
{
"epoch": 0.8917996237720337,
"grad_norm": 3.161616563796997,
"learning_rate": 3.590371899252659e-06,
"loss": 1.7933,
"step": 1600
},
{
"epoch": 0.8945864975963214,
"grad_norm": 2.933220386505127,
"learning_rate": 3.4116121432469615e-06,
"loss": 1.7989,
"step": 1605
},
{
"epoch": 0.8973733714206089,
"grad_norm": 3.294678211212158,
"learning_rate": 3.237259650159402e-06,
"loss": 1.7387,
"step": 1610
},
{
"epoch": 0.9001602452448966,
"grad_norm": 3.03759765625,
"learning_rate": 3.0673309137411564e-06,
"loss": 1.8489,
"step": 1615
},
{
"epoch": 0.9029471190691841,
"grad_norm": 3.226689338684082,
"learning_rate": 2.9018420092558786e-06,
"loss": 1.7449,
"step": 1620
},
{
"epoch": 0.9057339928934718,
"grad_norm": 3.118697166442871,
"learning_rate": 2.7408085919590264e-06,
"loss": 1.9021,
"step": 1625
},
{
"epoch": 0.9085208667177593,
"grad_norm": 3.2170827388763428,
"learning_rate": 2.584245895616788e-06,
"loss": 1.8544,
"step": 1630
},
{
"epoch": 0.911307740542047,
"grad_norm": 3.2920360565185547,
"learning_rate": 2.4321687310650487e-06,
"loss": 1.8315,
"step": 1635
},
{
"epoch": 0.9140946143663345,
"grad_norm": 2.9630749225616455,
"learning_rate": 2.2845914848082127e-06,
"loss": 1.8697,
"step": 1640
},
{
"epoch": 0.9168814881906222,
"grad_norm": 3.2118113040924072,
"learning_rate": 2.1415281176583203e-06,
"loss": 1.7885,
"step": 1645
},
{
"epoch": 0.9196683620149098,
"grad_norm": 3.565486192703247,
"learning_rate": 2.0029921634142632e-06,
"loss": 1.9092,
"step": 1650
},
{
"epoch": 0.9224552358391974,
"grad_norm": 3.0646369457244873,
"learning_rate": 1.8689967275815679e-06,
"loss": 1.8476,
"step": 1655
},
{
"epoch": 0.925242109663485,
"grad_norm": 3.4350948333740234,
"learning_rate": 1.7395544861325718e-06,
"loss": 1.8642,
"step": 1660
},
{
"epoch": 0.9280289834877726,
"grad_norm": 3.2032108306884766,
"learning_rate": 1.614677684307264e-06,
"loss": 1.8052,
"step": 1665
},
{
"epoch": 0.9308158573120602,
"grad_norm": 3.4371564388275146,
"learning_rate": 1.494378135454938e-06,
"loss": 1.8034,
"step": 1670
},
{
"epoch": 0.9336027311363478,
"grad_norm": 2.9803051948547363,
"learning_rate": 1.3786672199165962e-06,
"loss": 1.85,
"step": 1675
},
{
"epoch": 0.9363896049606354,
"grad_norm": 3.113248348236084,
"learning_rate": 1.2675558839483848e-06,
"loss": 1.7443,
"step": 1680
},
{
"epoch": 0.939176478784923,
"grad_norm": 2.9999353885650635,
"learning_rate": 1.1610546386860988e-06,
"loss": 1.7635,
"step": 1685
},
{
"epoch": 0.9419633526092106,
"grad_norm": 3.1148786544799805,
"learning_rate": 1.0591735591507946e-06,
"loss": 1.7791,
"step": 1690
},
{
"epoch": 0.9447502264334983,
"grad_norm": 3.0757246017456055,
"learning_rate": 9.619222832957243e-07,
"loss": 1.7563,
"step": 1695
},
{
"epoch": 0.9475371002577858,
"grad_norm": 2.859780788421631,
"learning_rate": 8.693100110945484e-07,
"loss": 1.7649,
"step": 1700
},
{
"epoch": 0.9503239740820735,
"grad_norm": 3.228289842605591,
"learning_rate": 7.813455036710715e-07,
"loss": 1.7731,
"step": 1705
},
{
"epoch": 0.953110847906361,
"grad_norm": 3.275424003601074,
"learning_rate": 6.980370824703763e-07,
"loss": 1.8506,
"step": 1710
},
{
"epoch": 0.9558977217306487,
"grad_norm": 3.517732858657837,
"learning_rate": 6.193926284716711e-07,
"loss": 1.7351,
"step": 1715
},
{
"epoch": 0.9586845955549362,
"grad_norm": 3.136549711227417,
"learning_rate": 5.454195814427021e-07,
"loss": 1.924,
"step": 1720
},
{
"epoch": 0.9614714693792239,
"grad_norm": 3.0273568630218506,
"learning_rate": 4.76124939235989e-07,
"loss": 1.7315,
"step": 1725
},
{
"epoch": 0.9642583432035114,
"grad_norm": 2.9298269748687744,
"learning_rate": 4.1151525712680996e-07,
"loss": 1.7149,
"step": 1730
},
{
"epoch": 0.9670452170277991,
"grad_norm": 3.315765857696533,
"learning_rate": 3.515966471930643e-07,
"loss": 1.897,
"step": 1735
},
{
"epoch": 0.9698320908520867,
"grad_norm": 3.05407452583313,
"learning_rate": 2.963747777370907e-07,
"loss": 1.8239,
"step": 1740
},
{
"epoch": 0.9726189646763743,
"grad_norm": 3.4511966705322266,
"learning_rate": 2.458548727494292e-07,
"loss": 1.6959,
"step": 1745
},
{
"epoch": 0.9754058385006619,
"grad_norm": 2.9159231185913086,
"learning_rate": 2.0004171141464467e-07,
"loss": 1.8196,
"step": 1750
},
{
"epoch": 0.9781927123249495,
"grad_norm": 3.1625008583068848,
"learning_rate": 1.589396276591937e-07,
"loss": 1.7935,
"step": 1755
},
{
"epoch": 0.9809795861492371,
"grad_norm": 3.055556058883667,
"learning_rate": 1.225525097414637e-07,
"loss": 1.7227,
"step": 1760
},
{
"epoch": 0.9837664599735247,
"grad_norm": 3.7383105754852295,
"learning_rate": 9.088379988392848e-08,
"loss": 1.8226,
"step": 1765
},
{
"epoch": 0.9865533337978123,
"grad_norm": 3.2375192642211914,
"learning_rate": 6.393649394749734e-08,
"loss": 1.7506,
"step": 1770
},
{
"epoch": 0.9893402076221,
"grad_norm": 3.6092934608459473,
"learning_rate": 4.171314114815306e-08,
"loss": 1.8767,
"step": 1775
},
{
"epoch": 0.9921270814463875,
"grad_norm": 2.9383866786956787,
"learning_rate": 2.4215843815733607e-08,
"loss": 1.7004,
"step": 1780
},
{
"epoch": 0.9949139552706752,
"grad_norm": 2.7816483974456787,
"learning_rate": 1.1446257195119048e-08,
"loss": 1.8521,
"step": 1785
},
{
"epoch": 0.9977008290949627,
"grad_norm": 3.2277441024780273,
"learning_rate": 3.4055892895901167e-09,
"loss": 1.8238,
"step": 1790
},
{
"epoch": 1.0,
"grad_norm": 12.097465515136719,
"learning_rate": 9.460074656963969e-11,
"loss": 1.9539,
"step": 1795
},
{
"epoch": 1.0,
"step": 1795,
"total_flos": 1.07378806147584e+16,
"train_loss": 1.90524691950976,
"train_runtime": 3624.8086,
"train_samples_per_second": 3.96,
"train_steps_per_second": 0.495
}
],
"logging_steps": 5,
"max_steps": 1795,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.07378806147584e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}