Files
SmolLM2-MagPiePro/trainer_state.json

2998 lines
73 KiB
JSON
Raw Normal View History

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9997629770087698,
"eval_steps": 500,
"global_step": 2109,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002370229912301493,
"grad_norm": 37.87957407825341,
"learning_rate": 7.109004739336492e-06,
"loss": 3.2574,
"step": 5
},
{
"epoch": 0.004740459824602986,
"grad_norm": 18.177249571303804,
"learning_rate": 1.4218009478672984e-05,
"loss": 2.7918,
"step": 10
},
{
"epoch": 0.00711068973690448,
"grad_norm": 5.001343795126334,
"learning_rate": 2.1327014218009475e-05,
"loss": 1.7693,
"step": 15
},
{
"epoch": 0.009480919649205973,
"grad_norm": 2.325367997984495,
"learning_rate": 2.8436018957345967e-05,
"loss": 1.2881,
"step": 20
},
{
"epoch": 0.011851149561507466,
"grad_norm": 1.0235836184491791,
"learning_rate": 3.554502369668246e-05,
"loss": 1.0407,
"step": 25
},
{
"epoch": 0.01422137947380896,
"grad_norm": 0.6279074046022562,
"learning_rate": 4.265402843601895e-05,
"loss": 0.9531,
"step": 30
},
{
"epoch": 0.016591609386110453,
"grad_norm": 0.6025558774407773,
"learning_rate": 4.9763033175355446e-05,
"loss": 0.8999,
"step": 35
},
{
"epoch": 0.018961839298411946,
"grad_norm": 0.4763893378885808,
"learning_rate": 5.6872037914691935e-05,
"loss": 0.8687,
"step": 40
},
{
"epoch": 0.02133206921071344,
"grad_norm": 0.37024070517127505,
"learning_rate": 6.398104265402843e-05,
"loss": 0.8436,
"step": 45
},
{
"epoch": 0.02370229912301493,
"grad_norm": 0.3630769483112416,
"learning_rate": 7.109004739336492e-05,
"loss": 0.8374,
"step": 50
},
{
"epoch": 0.026072529035316427,
"grad_norm": 0.20827819545344178,
"learning_rate": 7.819905213270142e-05,
"loss": 0.8075,
"step": 55
},
{
"epoch": 0.02844275894761792,
"grad_norm": 0.1601696569471494,
"learning_rate": 8.53080568720379e-05,
"loss": 0.7993,
"step": 60
},
{
"epoch": 0.030812988859919413,
"grad_norm": 0.1395108840386357,
"learning_rate": 9.24170616113744e-05,
"loss": 0.7844,
"step": 65
},
{
"epoch": 0.033183218772220906,
"grad_norm": 0.10908293522762928,
"learning_rate": 9.952606635071089e-05,
"loss": 0.7878,
"step": 70
},
{
"epoch": 0.0355534486845224,
"grad_norm": 0.09936997217620527,
"learning_rate": 0.00010663507109004738,
"loss": 0.7492,
"step": 75
},
{
"epoch": 0.03792367859682389,
"grad_norm": 0.09353265095940279,
"learning_rate": 0.00011374407582938387,
"loss": 0.7694,
"step": 80
},
{
"epoch": 0.040293908509125384,
"grad_norm": 0.10028988988229223,
"learning_rate": 0.00012085308056872037,
"loss": 0.7645,
"step": 85
},
{
"epoch": 0.04266413842142688,
"grad_norm": 0.08708652067409801,
"learning_rate": 0.00012796208530805686,
"loss": 0.7673,
"step": 90
},
{
"epoch": 0.04503436833372837,
"grad_norm": 0.07995303225473299,
"learning_rate": 0.00013507109004739335,
"loss": 0.7535,
"step": 95
},
{
"epoch": 0.04740459824602986,
"grad_norm": 0.07979293896147231,
"learning_rate": 0.00014218009478672984,
"loss": 0.7689,
"step": 100
},
{
"epoch": 0.049774828158331355,
"grad_norm": 0.08476243740502522,
"learning_rate": 0.00014928909952606633,
"loss": 0.7534,
"step": 105
},
{
"epoch": 0.052145058070632855,
"grad_norm": 0.07851254869100184,
"learning_rate": 0.00015639810426540285,
"loss": 0.7535,
"step": 110
},
{
"epoch": 0.05451528798293435,
"grad_norm": 0.08598983089548484,
"learning_rate": 0.0001635071090047393,
"loss": 0.7501,
"step": 115
},
{
"epoch": 0.05688551789523584,
"grad_norm": 0.08580450529497156,
"learning_rate": 0.0001706161137440758,
"loss": 0.7547,
"step": 120
},
{
"epoch": 0.05925574780753733,
"grad_norm": 0.08276852095286981,
"learning_rate": 0.00017772511848341231,
"loss": 0.7481,
"step": 125
},
{
"epoch": 0.061625977719838826,
"grad_norm": 0.0925763549038801,
"learning_rate": 0.0001848341232227488,
"loss": 0.7511,
"step": 130
},
{
"epoch": 0.06399620763214031,
"grad_norm": 0.07917042236946292,
"learning_rate": 0.00019194312796208527,
"loss": 0.752,
"step": 135
},
{
"epoch": 0.06636643754444181,
"grad_norm": 0.09521113296814705,
"learning_rate": 0.00019905213270142178,
"loss": 0.7525,
"step": 140
},
{
"epoch": 0.0687366674567433,
"grad_norm": 0.08361670279214221,
"learning_rate": 0.00020616113744075827,
"loss": 0.7499,
"step": 145
},
{
"epoch": 0.0711068973690448,
"grad_norm": 0.0808939093486955,
"learning_rate": 0.00021327014218009476,
"loss": 0.7564,
"step": 150
},
{
"epoch": 0.0734771272813463,
"grad_norm": 0.07840457880380175,
"learning_rate": 0.00022037914691943128,
"loss": 0.7456,
"step": 155
},
{
"epoch": 0.07584735719364778,
"grad_norm": 0.08285428720448265,
"learning_rate": 0.00022748815165876774,
"loss": 0.7573,
"step": 160
},
{
"epoch": 0.07821758710594928,
"grad_norm": 0.08834398679552397,
"learning_rate": 0.00023459715639810423,
"loss": 0.7528,
"step": 165
},
{
"epoch": 0.08058781701825077,
"grad_norm": 0.07707799033964063,
"learning_rate": 0.00024170616113744075,
"loss": 0.7555,
"step": 170
},
{
"epoch": 0.08295804693055227,
"grad_norm": 0.08232085673512189,
"learning_rate": 0.0002488151658767772,
"loss": 0.7428,
"step": 175
},
{
"epoch": 0.08532827684285375,
"grad_norm": 0.08887042007982729,
"learning_rate": 0.0002559241706161137,
"loss": 0.7624,
"step": 180
},
{
"epoch": 0.08769850675515525,
"grad_norm": 0.09470799851879333,
"learning_rate": 0.0002630331753554502,
"loss": 0.7528,
"step": 185
},
{
"epoch": 0.09006873666745674,
"grad_norm": 0.10657490417564752,
"learning_rate": 0.0002701421800947867,
"loss": 0.7557,
"step": 190
},
{
"epoch": 0.09243896657975824,
"grad_norm": 0.08182579598618632,
"learning_rate": 0.0002772511848341232,
"loss": 0.7364,
"step": 195
},
{
"epoch": 0.09480919649205972,
"grad_norm": 0.07272030097431592,
"learning_rate": 0.0002843601895734597,
"loss": 0.737,
"step": 200
},
{
"epoch": 0.09717942640436122,
"grad_norm": 0.0764754646270827,
"learning_rate": 0.0002914691943127962,
"loss": 0.7458,
"step": 205
},
{
"epoch": 0.09954965631666271,
"grad_norm": 0.08027094645738686,
"learning_rate": 0.00029857819905213266,
"loss": 0.7563,
"step": 210
},
{
"epoch": 0.10191988622896421,
"grad_norm": 0.08015350192584438,
"learning_rate": 0.0002999967123393108,
"loss": 0.7471,
"step": 215
},
{
"epoch": 0.10429011614126571,
"grad_norm": 0.07996649533232081,
"learning_rate": 0.0002999833564647566,
"loss": 0.7593,
"step": 220
},
{
"epoch": 0.1066603460535672,
"grad_norm": 0.07683862115921537,
"learning_rate": 0.00029995972781162177,
"loss": 0.7563,
"step": 225
},
{
"epoch": 0.1090305759658687,
"grad_norm": 0.08428065421678084,
"learning_rate": 0.0002999258279982986,
"loss": 0.7542,
"step": 230
},
{
"epoch": 0.11140080587817018,
"grad_norm": 0.09166260522899684,
"learning_rate": 0.00029988165934668,
"loss": 0.7422,
"step": 235
},
{
"epoch": 0.11377103579047168,
"grad_norm": 0.08294402646546879,
"learning_rate": 0.00029982722488200013,
"loss": 0.7464,
"step": 240
},
{
"epoch": 0.11614126570277317,
"grad_norm": 0.07426872103025435,
"learning_rate": 0.00029976252833262734,
"loss": 0.7568,
"step": 245
},
{
"epoch": 0.11851149561507467,
"grad_norm": 0.07034251693444554,
"learning_rate": 0.00029968757412980883,
"loss": 0.7515,
"step": 250
},
{
"epoch": 0.12088172552737615,
"grad_norm": 0.07550642047266844,
"learning_rate": 0.000299602367407367,
"loss": 0.7499,
"step": 255
},
{
"epoch": 0.12325195543967765,
"grad_norm": 0.07047433670873945,
"learning_rate": 0.00029950691400134824,
"loss": 0.7496,
"step": 260
},
{
"epoch": 0.12562218535197914,
"grad_norm": 0.0857398982188067,
"learning_rate": 0.0002994012204496225,
"loss": 0.7422,
"step": 265
},
{
"epoch": 0.12799241526428062,
"grad_norm": 0.08049042276394297,
"learning_rate": 0.00029928529399143606,
"loss": 0.7554,
"step": 270
},
{
"epoch": 0.13036264517658214,
"grad_norm": 0.08078197460014627,
"learning_rate": 0.0002991591425669156,
"loss": 0.7455,
"step": 275
},
{
"epoch": 0.13273287508888362,
"grad_norm": 0.07709419757715143,
"learning_rate": 0.0002990227748165241,
"loss": 0.7475,
"step": 280
},
{
"epoch": 0.1351031050011851,
"grad_norm": 0.07552451990431377,
"learning_rate": 0.00029887620008046934,
"loss": 0.7386,
"step": 285
},
{
"epoch": 0.1374733349134866,
"grad_norm": 0.07498431824773105,
"learning_rate": 0.00029871942839806396,
"loss": 0.7263,
"step": 290
},
{
"epoch": 0.1398435648257881,
"grad_norm": 44.58672438459057,
"learning_rate": 0.0002985524705070379,
"loss": 1.4646,
"step": 295
},
{
"epoch": 0.1422137947380896,
"grad_norm": 34.95300570734307,
"learning_rate": 0.0002983753378428031,
"loss": 3.0175,
"step": 300
},
{
"epoch": 0.14458402465039108,
"grad_norm": 1.898833557773242,
"learning_rate": 0.00029818804253766974,
"loss": 1.2782,
"step": 305
},
{
"epoch": 0.1469542545626926,
"grad_norm": 2.4268359613555375,
"learning_rate": 0.00029799059742001606,
"loss": 2.2389,
"step": 310
},
{
"epoch": 0.14932448447499408,
"grad_norm": 1.0640566578307415,
"learning_rate": 0.00029778301601340906,
"loss": 1.0058,
"step": 315
},
{
"epoch": 0.15169471438729557,
"grad_norm": 0.18597050864762904,
"learning_rate": 0.0002975653125356785,
"loss": 0.8664,
"step": 320
},
{
"epoch": 0.15406494429959705,
"grad_norm": 0.14018854346808488,
"learning_rate": 0.000297337501897943,
"loss": 0.8061,
"step": 325
},
{
"epoch": 0.15643517421189856,
"grad_norm": 0.1064736567889867,
"learning_rate": 0.0002970995997035888,
"loss": 0.7942,
"step": 330
},
{
"epoch": 0.15880540412420005,
"grad_norm": 0.09820103029046393,
"learning_rate": 0.0002968516222472011,
"loss": 0.7866,
"step": 335
},
{
"epoch": 0.16117563403650154,
"grad_norm": 0.10051042892221068,
"learning_rate": 0.0002965935865134476,
"loss": 0.7917,
"step": 340
},
{
"epoch": 0.16354586394880302,
"grad_norm": 0.07991174466381477,
"learning_rate": 0.0002963255101759159,
"loss": 0.7776,
"step": 345
},
{
"epoch": 0.16591609386110454,
"grad_norm": 0.07973784335269311,
"learning_rate": 0.0002960474115959023,
"loss": 0.7684,
"step": 350
},
{
"epoch": 0.16828632377340602,
"grad_norm": 3.2883719479596585,
"learning_rate": 0.00029575930982115445,
"loss": 0.7581,
"step": 355
},
{
"epoch": 0.1706565536857075,
"grad_norm": 0.31783313849101436,
"learning_rate": 0.0002954612245845669,
"loss": 1.0093,
"step": 360
},
{
"epoch": 0.17302678359800902,
"grad_norm": 0.14168083007400084,
"learning_rate": 0.0002951531763028292,
"loss": 0.8543,
"step": 365
},
{
"epoch": 0.1753970135103105,
"grad_norm": 0.11564628343820599,
"learning_rate": 0.0002948351860750277,
"loss": 0.8189,
"step": 370
},
{
"epoch": 0.177767243422612,
"grad_norm": 5.805978160293609,
"learning_rate": 0.00029450727568120037,
"loss": 0.8103,
"step": 375
},
{
"epoch": 0.18013747333491348,
"grad_norm": 0.08117200321663696,
"learning_rate": 0.0002941694675808451,
"loss": 0.7753,
"step": 380
},
{
"epoch": 0.182507703247215,
"grad_norm": 0.07607198738704968,
"learning_rate": 0.0002938217849113812,
"loss": 0.7603,
"step": 385
},
{
"epoch": 0.18487793315951648,
"grad_norm": 0.07113347917796854,
"learning_rate": 0.00029346425148656487,
"loss": 0.754,
"step": 390
},
{
"epoch": 0.18724816307181796,
"grad_norm": 0.06638185085999807,
"learning_rate": 0.0002930968917948581,
"loss": 0.7561,
"step": 395
},
{
"epoch": 0.18961839298411945,
"grad_norm": 0.0636288319721399,
"learning_rate": 0.000292719730997751,
"loss": 0.7431,
"step": 400
},
{
"epoch": 0.19198862289642096,
"grad_norm": 0.06951776019726015,
"learning_rate": 0.0002923327949280391,
"loss": 0.731,
"step": 405
},
{
"epoch": 0.19435885280872245,
"grad_norm": 0.0628405894324144,
"learning_rate": 0.0002919361100880534,
"loss": 0.735,
"step": 410
},
{
"epoch": 0.19672908272102393,
"grad_norm": 0.06251446911308753,
"learning_rate": 0.0002915297036478455,
"loss": 0.7277,
"step": 415
},
{
"epoch": 0.19909931263332542,
"grad_norm": 0.0606660234416891,
"learning_rate": 0.00029111360344332647,
"loss": 0.7324,
"step": 420
},
{
"epoch": 0.20146954254562693,
"grad_norm": 0.060370913004170065,
"learning_rate": 0.0002906878379743604,
"loss": 0.7234,
"step": 425
},
{
"epoch": 0.20383977245792842,
"grad_norm": 0.06265360740334025,
"learning_rate": 0.00029025243640281223,
"loss": 0.7357,
"step": 430
},
{
"epoch": 0.2062100023702299,
"grad_norm": 0.06358474472495813,
"learning_rate": 0.0002898074285505504,
"loss": 0.7188,
"step": 435
},
{
"epoch": 0.20858023228253142,
"grad_norm": 0.059826146661527135,
"learning_rate": 0.0002893528448974046,
"loss": 0.7183,
"step": 440
},
{
"epoch": 0.2109504621948329,
"grad_norm": 0.0601231927656965,
"learning_rate": 0.00028888871657907744,
"loss": 0.7189,
"step": 445
},
{
"epoch": 0.2133206921071344,
"grad_norm": 0.8526105639659928,
"learning_rate": 0.00028841507538501267,
"loss": 0.7202,
"step": 450
},
{
"epoch": 0.21569092201943588,
"grad_norm": 0.06361964303367311,
"learning_rate": 0.0002879319537562172,
"loss": 0.7191,
"step": 455
},
{
"epoch": 0.2180611519317374,
"grad_norm": 0.061558109294327805,
"learning_rate": 0.00028743938478303937,
"loss": 0.7202,
"step": 460
},
{
"epoch": 0.22043138184403888,
"grad_norm": 0.06228127012654291,
"learning_rate": 0.0002869374022029026,
"loss": 0.7269,
"step": 465
},
{
"epoch": 0.22280161175634036,
"grad_norm": 0.06239279055248459,
"learning_rate": 0.00028642604039799457,
"loss": 0.7211,
"step": 470
},
{
"epoch": 0.22517184166864185,
"grad_norm": 0.08371349247174907,
"learning_rate": 0.00028590533439291205,
"loss": 0.7132,
"step": 475
},
{
"epoch": 0.22754207158094336,
"grad_norm": 0.0630124916842619,
"learning_rate": 0.0002853753198522623,
"loss": 0.7216,
"step": 480
},
{
"epoch": 0.22991230149324485,
"grad_norm": 0.06231078679066273,
"learning_rate": 0.00028483603307822,
"loss": 0.7255,
"step": 485
},
{
"epoch": 0.23228253140554633,
"grad_norm": 0.06370990821920278,
"learning_rate": 0.00028428751100804115,
"loss": 0.7058,
"step": 490
},
{
"epoch": 0.23465276131784782,
"grad_norm": 0.05998852944044534,
"learning_rate": 0.0002837297912115329,
"loss": 0.7075,
"step": 495
},
{
"epoch": 0.23702299123014933,
"grad_norm": 0.05934170652874789,
"learning_rate": 0.00028316291188848025,
"loss": 0.6999,
"step": 500
},
{
"epoch": 0.23939322114245082,
"grad_norm": 0.06375843749865423,
"learning_rate": 0.0002825869118660298,
"loss": 0.713,
"step": 505
},
{
"epoch": 0.2417634510547523,
"grad_norm": 0.06159522236341869,
"learning_rate": 0.0002820018305960303,
"loss": 0.7088,
"step": 510
},
{
"epoch": 0.24413368096705382,
"grad_norm": 0.061758194634036964,
"learning_rate": 0.0002814077081523305,
"loss": 0.7042,
"step": 515
},
{
"epoch": 0.2465039108793553,
"grad_norm": 0.06170727020552632,
"learning_rate": 0.0002808045852280343,
"loss": 0.7103,
"step": 520
},
{
"epoch": 0.2488741407916568,
"grad_norm": 0.05840561215234994,
"learning_rate": 0.0002801925031327137,
"loss": 0.6987,
"step": 525
},
{
"epoch": 0.2512443707039583,
"grad_norm": 0.058175029378156634,
"learning_rate": 0.0002795715037895792,
"loss": 0.6909,
"step": 530
},
{
"epoch": 0.2536146006162598,
"grad_norm": 0.06101870086518658,
"learning_rate": 0.0002789416297326088,
"loss": 0.7184,
"step": 535
},
{
"epoch": 0.25598483052856125,
"grad_norm": 0.06279243933970817,
"learning_rate": 0.0002783029241036339,
"loss": 0.7123,
"step": 540
},
{
"epoch": 0.25835506044086276,
"grad_norm": 0.06346784165112655,
"learning_rate": 0.0002776554306493854,
"loss": 0.7196,
"step": 545
},
{
"epoch": 0.2607252903531643,
"grad_norm": 0.05748578268561143,
"learning_rate": 0.00027699919371849664,
"loss": 0.7084,
"step": 550
},
{
"epoch": 0.26309552026546573,
"grad_norm": 0.06563979843305176,
"learning_rate": 0.0002763342582584661,
"loss": 0.7048,
"step": 555
},
{
"epoch": 0.26546575017776725,
"grad_norm": 0.06177493929500722,
"learning_rate": 0.0002756606698125787,
"loss": 0.7058,
"step": 560
},
{
"epoch": 0.26783598009006876,
"grad_norm": 0.06067602378089436,
"learning_rate": 0.0002749784745167868,
"loss": 0.6934,
"step": 565
},
{
"epoch": 0.2702062100023702,
"grad_norm": 0.062834571657062,
"learning_rate": 0.0002742877190965495,
"loss": 0.6963,
"step": 570
},
{
"epoch": 0.27257643991467173,
"grad_norm": 0.059163687863750675,
"learning_rate": 0.0002735884508636331,
"loss": 0.7028,
"step": 575
},
{
"epoch": 0.2749466698269732,
"grad_norm": 0.05983910867748256,
"learning_rate": 0.00027288071771286996,
"loss": 0.6889,
"step": 580
},
{
"epoch": 0.2773168997392747,
"grad_norm": 0.05836025064363057,
"learning_rate": 0.0002721645681188784,
"loss": 0.6973,
"step": 585
},
{
"epoch": 0.2796871296515762,
"grad_norm": 0.06355277332945615,
"learning_rate": 0.00027144005113274237,
"loss": 0.6949,
"step": 590
},
{
"epoch": 0.2820573595638777,
"grad_norm": 0.060849753041127434,
"learning_rate": 0.0002707072163786518,
"loss": 0.69,
"step": 595
},
{
"epoch": 0.2844275894761792,
"grad_norm": 0.06170856816087761,
"learning_rate": 0.00026996611405050386,
"loss": 0.6976,
"step": 600
},
{
"epoch": 0.2867978193884807,
"grad_norm": 0.06086967149436421,
"learning_rate": 0.0002692167949084649,
"loss": 0.7081,
"step": 605
},
{
"epoch": 0.28916804930078216,
"grad_norm": 0.059663431709652645,
"learning_rate": 0.00026845931027549375,
"loss": 0.7034,
"step": 610
},
{
"epoch": 0.2915382792130837,
"grad_norm": 0.05768729651844958,
"learning_rate": 0.0002676937120338265,
"loss": 0.6988,
"step": 615
},
{
"epoch": 0.2939085091253852,
"grad_norm": 0.059449464935387684,
"learning_rate": 0.00026692005262142303,
"loss": 0.6911,
"step": 620
},
{
"epoch": 0.29627873903768664,
"grad_norm": 0.05908588685261821,
"learning_rate": 0.00026613838502837514,
"loss": 0.7012,
"step": 625
},
{
"epoch": 0.29864896894998816,
"grad_norm": 0.05823031421787097,
"learning_rate": 0.0002653487627932775,
"loss": 0.6815,
"step": 630
},
{
"epoch": 0.3010191988622896,
"grad_norm": 0.06346993001531062,
"learning_rate": 0.00026455123999956,
"loss": 0.6778,
"step": 635
},
{
"epoch": 0.30338942877459113,
"grad_norm": 0.05766157585577824,
"learning_rate": 0.00026374587127178434,
"loss": 0.6872,
"step": 640
},
{
"epoch": 0.30575965868689264,
"grad_norm": 0.060699498460503425,
"learning_rate": 0.00026293271177190184,
"loss": 0.698,
"step": 645
},
{
"epoch": 0.3081298885991941,
"grad_norm": 0.06470838061776032,
"learning_rate": 0.0002621118171954755,
"loss": 0.6863,
"step": 650
},
{
"epoch": 0.3105001185114956,
"grad_norm": 0.05809323133547399,
"learning_rate": 0.00026128324376786555,
"loss": 0.6853,
"step": 655
},
{
"epoch": 0.31287034842379713,
"grad_norm": 0.060605773782501364,
"learning_rate": 0.00026044704824037786,
"loss": 0.706,
"step": 660
},
{
"epoch": 0.3152405783360986,
"grad_norm": 0.05891479933624849,
"learning_rate": 0.0002596032878863774,
"loss": 0.6696,
"step": 665
},
{
"epoch": 0.3176108082484001,
"grad_norm": 0.0625427319419523,
"learning_rate": 0.0002587520204973651,
"loss": 0.6827,
"step": 670
},
{
"epoch": 0.3199810381607016,
"grad_norm": 0.06168949595593282,
"learning_rate": 0.00025789330437901974,
"loss": 0.6798,
"step": 675
},
{
"epoch": 0.32235126807300307,
"grad_norm": 0.05715106171580865,
"learning_rate": 0.00025702719834720433,
"loss": 0.6757,
"step": 680
},
{
"epoch": 0.3247214979853046,
"grad_norm": 0.05793783101869966,
"learning_rate": 0.00025615376172393774,
"loss": 0.6824,
"step": 685
},
{
"epoch": 0.32709172789760604,
"grad_norm": 0.05937949955225907,
"learning_rate": 0.00025527305433333127,
"loss": 0.6855,
"step": 690
},
{
"epoch": 0.32946195780990756,
"grad_norm": 0.06672193394439148,
"learning_rate": 0.00025438513649749173,
"loss": 0.6878,
"step": 695
},
{
"epoch": 0.33183218772220907,
"grad_norm": 0.06046441135071207,
"learning_rate": 0.00025349006903238906,
"loss": 0.6702,
"step": 700
},
{
"epoch": 0.33420241763451053,
"grad_norm": 0.06234393756385736,
"learning_rate": 0.0002525879132436916,
"loss": 0.6885,
"step": 705
},
{
"epoch": 0.33657264754681204,
"grad_norm": 0.06442431249968464,
"learning_rate": 0.0002516787309225664,
"loss": 0.6824,
"step": 710
},
{
"epoch": 0.33894287745911356,
"grad_norm": 0.057449230900877835,
"learning_rate": 0.0002507625843414477,
"loss": 0.6825,
"step": 715
},
{
"epoch": 0.341313107371415,
"grad_norm": 0.06647075866637231,
"learning_rate": 0.000249839536249771,
"loss": 0.6715,
"step": 720
},
{
"epoch": 0.34368333728371653,
"grad_norm": 0.06516829971838099,
"learning_rate": 0.0002489096498696758,
"loss": 0.6618,
"step": 725
},
{
"epoch": 0.34605356719601804,
"grad_norm": 0.06010008501643476,
"learning_rate": 0.0002479729888916749,
"loss": 0.6791,
"step": 730
},
{
"epoch": 0.3484237971083195,
"grad_norm": 0.05742770698758986,
"learning_rate": 0.0002470296174702921,
"loss": 0.679,
"step": 735
},
{
"epoch": 0.350794027020621,
"grad_norm": 0.056906839817947766,
"learning_rate": 0.0002460796002196684,
"loss": 0.6667,
"step": 740
},
{
"epoch": 0.35316425693292247,
"grad_norm": 0.05683684204049849,
"learning_rate": 0.00024512300220913617,
"loss": 0.6771,
"step": 745
},
{
"epoch": 0.355534486845224,
"grad_norm": 0.05936901663811349,
"learning_rate": 0.0002441598889587624,
"loss": 0.6694,
"step": 750
},
{
"epoch": 0.3579047167575255,
"grad_norm": 0.056734740760335606,
"learning_rate": 0.0002431903264348607,
"loss": 0.6656,
"step": 755
},
{
"epoch": 0.36027494666982696,
"grad_norm": 0.05711271475645182,
"learning_rate": 0.00024221438104547394,
"loss": 0.6777,
"step": 760
},
{
"epoch": 0.36264517658212847,
"grad_norm": 0.06484350452592977,
"learning_rate": 0.00024123211963582493,
"loss": 0.6712,
"step": 765
},
{
"epoch": 0.36501540649443,
"grad_norm": 0.06427694084052618,
"learning_rate": 0.00024024360948373834,
"loss": 0.6655,
"step": 770
},
{
"epoch": 0.36738563640673144,
"grad_norm": 0.06329540165136863,
"learning_rate": 0.0002392489182950328,
"loss": 0.6649,
"step": 775
},
{
"epoch": 0.36975586631903296,
"grad_norm": 0.05829453649974294,
"learning_rate": 0.0002382481141988833,
"loss": 0.6825,
"step": 780
},
{
"epoch": 0.3721260962313344,
"grad_norm": 0.05785471731045562,
"learning_rate": 0.00023724126574315484,
"loss": 0.673,
"step": 785
},
{
"epoch": 0.3744963261436359,
"grad_norm": 0.0561890074327068,
"learning_rate": 0.0002362284418897076,
"loss": 0.6674,
"step": 790
},
{
"epoch": 0.37686655605593744,
"grad_norm": 0.05695140762918065,
"learning_rate": 0.00023520971200967334,
"loss": 0.6699,
"step": 795
},
{
"epoch": 0.3792367859682389,
"grad_norm": 0.056382588072258655,
"learning_rate": 0.00023418514587870414,
"loss": 0.6669,
"step": 800
},
{
"epoch": 0.3816070158805404,
"grad_norm": 0.05699605706668468,
"learning_rate": 0.00023315481367219306,
"loss": 0.6646,
"step": 805
},
{
"epoch": 0.3839772457928419,
"grad_norm": 0.05971904780219733,
"learning_rate": 0.00023211878596046795,
"loss": 0.6694,
"step": 810
},
{
"epoch": 0.3863474757051434,
"grad_norm": 0.05872202455985711,
"learning_rate": 0.00023107713370395752,
"loss": 0.6564,
"step": 815
},
{
"epoch": 0.3887177056174449,
"grad_norm": 0.056267010708901014,
"learning_rate": 0.00023002992824833134,
"loss": 0.6619,
"step": 820
},
{
"epoch": 0.3910879355297464,
"grad_norm": 0.05858819926921814,
"learning_rate": 0.00022897724131961314,
"loss": 0.6555,
"step": 825
},
{
"epoch": 0.39345816544204787,
"grad_norm": 0.058908341888103316,
"learning_rate": 0.00022791914501926785,
"loss": 0.6587,
"step": 830
},
{
"epoch": 0.3958283953543494,
"grad_norm": 0.05581026158289953,
"learning_rate": 0.00022685571181926354,
"loss": 0.6725,
"step": 835
},
{
"epoch": 0.39819862526665084,
"grad_norm": 0.05851912322437669,
"learning_rate": 0.00022578701455710734,
"loss": 0.6627,
"step": 840
},
{
"epoch": 0.40056885517895235,
"grad_norm": 0.05512523024478408,
"learning_rate": 0.00022471312643085667,
"loss": 0.6597,
"step": 845
},
{
"epoch": 0.40293908509125387,
"grad_norm": 0.05488965997035471,
"learning_rate": 0.00022363412099410585,
"loss": 0.6546,
"step": 850
},
{
"epoch": 0.4053093150035553,
"grad_norm": 0.05590348015503589,
"learning_rate": 0.0002225500721509479,
"loss": 0.6676,
"step": 855
},
{
"epoch": 0.40767954491585684,
"grad_norm": 0.0557504280408415,
"learning_rate": 0.0002214610541509129,
"loss": 0.6559,
"step": 860
},
{
"epoch": 0.41004977482815835,
"grad_norm": 0.060424783129262576,
"learning_rate": 0.00022036714158388235,
"loss": 0.6712,
"step": 865
},
{
"epoch": 0.4124200047404598,
"grad_norm": 0.05481897236135071,
"learning_rate": 0.00021926840937498034,
"loss": 0.6506,
"step": 870
},
{
"epoch": 0.4147902346527613,
"grad_norm": 0.05656177938481429,
"learning_rate": 0.00021816493277944157,
"loss": 0.6746,
"step": 875
},
{
"epoch": 0.41716046456506284,
"grad_norm": 0.05625948208637603,
"learning_rate": 0.00021705678737745713,
"loss": 0.6378,
"step": 880
},
{
"epoch": 0.4195306944773643,
"grad_norm": 0.05903464854176921,
"learning_rate": 0.0002159440490689976,
"loss": 0.6606,
"step": 885
},
{
"epoch": 0.4219009243896658,
"grad_norm": 0.05683800007835736,
"learning_rate": 0.0002148267940686146,
"loss": 0.6621,
"step": 890
},
{
"epoch": 0.42427115430196727,
"grad_norm": 0.05931947846319525,
"learning_rate": 0.0002137050989002206,
"loss": 0.6539,
"step": 895
},
{
"epoch": 0.4266413842142688,
"grad_norm": 0.05335168668707516,
"learning_rate": 0.0002125790403918475,
"loss": 0.656,
"step": 900
},
{
"epoch": 0.4290116141265703,
"grad_norm": 0.0544878383415537,
"learning_rate": 0.0002114486956703846,
"loss": 0.6438,
"step": 905
},
{
"epoch": 0.43138184403887175,
"grad_norm": 0.06212920385348046,
"learning_rate": 0.00021031414215629576,
"loss": 0.6519,
"step": 910
},
{
"epoch": 0.43375207395117327,
"grad_norm": 0.053801804211920747,
"learning_rate": 0.000209175457558317,
"loss": 0.6534,
"step": 915
},
{
"epoch": 0.4361223038634748,
"grad_norm": 0.055676827319867436,
"learning_rate": 0.00020803271986813364,
"loss": 0.6434,
"step": 920
},
{
"epoch": 0.43849253377577624,
"grad_norm": 0.06221867524670911,
"learning_rate": 0.00020688600735503863,
"loss": 0.6642,
"step": 925
},
{
"epoch": 0.44086276368807775,
"grad_norm": 0.05719745436620134,
"learning_rate": 0.0002057353985605717,
"loss": 0.6654,
"step": 930
},
{
"epoch": 0.4432329936003792,
"grad_norm": 0.05636204685752076,
"learning_rate": 0.00020458097229313963,
"loss": 0.6469,
"step": 935
},
{
"epoch": 0.4456032235126807,
"grad_norm": 0.05949021529775284,
"learning_rate": 0.00020342280762261874,
"loss": 0.6514,
"step": 940
},
{
"epoch": 0.44797345342498224,
"grad_norm": 0.05819024554018189,
"learning_rate": 0.0002022609838749389,
"loss": 0.6601,
"step": 945
},
{
"epoch": 0.4503436833372837,
"grad_norm": 0.054298063737541105,
"learning_rate": 0.00020109558062665037,
"loss": 0.6561,
"step": 950
},
{
"epoch": 0.4527139132495852,
"grad_norm": 0.05325573283656968,
"learning_rate": 0.00019992667769947348,
"loss": 0.6556,
"step": 955
},
{
"epoch": 0.4550841431618867,
"grad_norm": 0.055557610793003705,
"learning_rate": 0.00019875435515483126,
"loss": 0.6598,
"step": 960
},
{
"epoch": 0.4574543730741882,
"grad_norm": 0.05410922961948251,
"learning_rate": 0.0001975786932883659,
"loss": 0.6494,
"step": 965
},
{
"epoch": 0.4598246029864897,
"grad_norm": 0.05513130388145586,
"learning_rate": 0.00019639977262443908,
"loss": 0.6458,
"step": 970
},
{
"epoch": 0.4621948328987912,
"grad_norm": 0.05722322905223257,
"learning_rate": 0.00019521767391061663,
"loss": 0.6708,
"step": 975
},
{
"epoch": 0.46456506281109267,
"grad_norm": 0.06292159556421471,
"learning_rate": 0.00019403247811213776,
"loss": 0.6482,
"step": 980
},
{
"epoch": 0.4669352927233942,
"grad_norm": 0.061013794703927336,
"learning_rate": 0.00019284426640636982,
"loss": 0.6407,
"step": 985
},
{
"epoch": 0.46930552263569564,
"grad_norm": 0.05521810364010896,
"learning_rate": 0.000191653120177248,
"loss": 0.6433,
"step": 990
},
{
"epoch": 0.47167575254799715,
"grad_norm": 0.05464230308163921,
"learning_rate": 0.00019045912100970112,
"loss": 0.6408,
"step": 995
},
{
"epoch": 0.47404598246029866,
"grad_norm": 0.05534936230890589,
"learning_rate": 0.00018926235068406378,
"loss": 0.6406,
"step": 1000
},
{
"epoch": 0.4764162123726001,
"grad_norm": 0.05541859380358046,
"learning_rate": 0.00018806289117047494,
"loss": 0.6453,
"step": 1005
},
{
"epoch": 0.47878644228490164,
"grad_norm": 0.057019399656362324,
"learning_rate": 0.00018686082462326346,
"loss": 0.6424,
"step": 1010
},
{
"epoch": 0.48115667219720315,
"grad_norm": 0.05678834933291726,
"learning_rate": 0.00018565623337532137,
"loss": 0.6372,
"step": 1015
},
{
"epoch": 0.4835269021095046,
"grad_norm": 0.0626356752316691,
"learning_rate": 0.00018444919993246427,
"loss": 0.6536,
"step": 1020
},
{
"epoch": 0.4858971320218061,
"grad_norm": 0.05533208427640317,
"learning_rate": 0.00018323980696778074,
"loss": 0.6392,
"step": 1025
},
{
"epoch": 0.48826736193410764,
"grad_norm": 0.1583885749203545,
"learning_rate": 0.0001820281373159694,
"loss": 0.6377,
"step": 1030
},
{
"epoch": 0.4906375918464091,
"grad_norm": 0.05696942950681657,
"learning_rate": 0.00018081427396766565,
"loss": 0.632,
"step": 1035
},
{
"epoch": 0.4930078217587106,
"grad_norm": 0.06885427735472549,
"learning_rate": 0.0001795983000637572,
"loss": 0.6513,
"step": 1040
},
{
"epoch": 0.49537805167101207,
"grad_norm": 0.055666281836575904,
"learning_rate": 0.00017838029888968973,
"loss": 0.64,
"step": 1045
},
{
"epoch": 0.4977482815833136,
"grad_norm": 0.05590039370646873,
"learning_rate": 0.00017716035386976215,
"loss": 0.6363,
"step": 1050
},
{
"epoch": 0.500118511495615,
"grad_norm": 0.05747509683831957,
"learning_rate": 0.00017593854856141287,
"loss": 0.622,
"step": 1055
},
{
"epoch": 0.5024887414079166,
"grad_norm": 0.05599406050839353,
"learning_rate": 0.00017471496664949666,
"loss": 0.6352,
"step": 1060
},
{
"epoch": 0.5048589713202181,
"grad_norm": 0.056394714786088526,
"learning_rate": 0.00017348969194055285,
"loss": 0.628,
"step": 1065
},
{
"epoch": 0.5072292012325196,
"grad_norm": 0.05379996817708946,
"learning_rate": 0.00017226280835706503,
"loss": 0.6253,
"step": 1070
},
{
"epoch": 0.5095994311448211,
"grad_norm": 0.052644081064009085,
"learning_rate": 0.00017103439993171326,
"loss": 0.635,
"step": 1075
},
{
"epoch": 0.5119696610571225,
"grad_norm": 0.06183603107070903,
"learning_rate": 0.00016980455080161817,
"loss": 0.6247,
"step": 1080
},
{
"epoch": 0.514339890969424,
"grad_norm": 0.055776407065979555,
"learning_rate": 0.00016857334520257837,
"loss": 0.6356,
"step": 1085
},
{
"epoch": 0.5167101208817255,
"grad_norm": 0.057457278964487705,
"learning_rate": 0.00016734086746330066,
"loss": 0.6253,
"step": 1090
},
{
"epoch": 0.519080350794027,
"grad_norm": 0.054238874177121164,
"learning_rate": 0.00016610720199962444,
"loss": 0.6394,
"step": 1095
},
{
"epoch": 0.5214505807063285,
"grad_norm": 0.05556319159032643,
"learning_rate": 0.00016487243330873957,
"loss": 0.647,
"step": 1100
},
{
"epoch": 0.5238208106186301,
"grad_norm": 0.05533410752493418,
"learning_rate": 0.00016363664596339905,
"loss": 0.6301,
"step": 1105
},
{
"epoch": 0.5261910405309315,
"grad_norm": 0.057938893164981804,
"learning_rate": 0.00016239992460612636,
"loss": 0.6229,
"step": 1110
},
{
"epoch": 0.528561270443233,
"grad_norm": 0.05173600564638707,
"learning_rate": 0.00016116235394341807,
"loss": 0.6277,
"step": 1115
},
{
"epoch": 0.5309315003555345,
"grad_norm": 0.055079698685841325,
"learning_rate": 0.00015992401873994206,
"loss": 0.6169,
"step": 1120
},
{
"epoch": 0.533301730267836,
"grad_norm": 0.05549655687980898,
"learning_rate": 0.00015868500381273167,
"loss": 0.6389,
"step": 1125
},
{
"epoch": 0.5356719601801375,
"grad_norm": 0.059015018470196176,
"learning_rate": 0.0001574453940253765,
"loss": 0.6406,
"step": 1130
},
{
"epoch": 0.5380421900924389,
"grad_norm": 0.05635084046958829,
"learning_rate": 0.0001562052742822098,
"loss": 0.6245,
"step": 1135
},
{
"epoch": 0.5404124200047404,
"grad_norm": 0.056720749371614865,
"learning_rate": 0.00015496472952249298,
"loss": 0.6322,
"step": 1140
},
{
"epoch": 0.542782649917042,
"grad_norm": 0.055644623371899404,
"learning_rate": 0.00015372384471459816,
"loss": 0.6338,
"step": 1145
},
{
"epoch": 0.5451528798293435,
"grad_norm": 0.056962336114346236,
"learning_rate": 0.0001524827048501883,
"loss": 0.6363,
"step": 1150
},
{
"epoch": 0.547523109741645,
"grad_norm": 3.4539173382130626,
"learning_rate": 0.00015124139493839575,
"loss": 0.6514,
"step": 1155
},
{
"epoch": 0.5498933396539464,
"grad_norm": 0.0592035753126842,
"learning_rate": 0.00015,
"loss": 0.6354,
"step": 1160
},
{
"epoch": 0.5522635695662479,
"grad_norm": 0.05848608337431566,
"learning_rate": 0.0001487586050616043,
"loss": 0.6291,
"step": 1165
},
{
"epoch": 0.5546337994785494,
"grad_norm": 0.056837303715594346,
"learning_rate": 0.00014751729514981169,
"loss": 0.6186,
"step": 1170
},
{
"epoch": 0.5570040293908509,
"grad_norm": 0.05430228685183931,
"learning_rate": 0.0001462761552854018,
"loss": 0.6353,
"step": 1175
},
{
"epoch": 0.5593742593031524,
"grad_norm": 0.05561193103292618,
"learning_rate": 0.00014503527047750702,
"loss": 0.621,
"step": 1180
},
{
"epoch": 0.561744489215454,
"grad_norm": 0.054843289485073124,
"learning_rate": 0.00014379472571779024,
"loss": 0.6142,
"step": 1185
},
{
"epoch": 0.5641147191277553,
"grad_norm": 0.05315905080273884,
"learning_rate": 0.00014255460597462346,
"loss": 0.6161,
"step": 1190
},
{
"epoch": 0.5664849490400569,
"grad_norm": 0.05493581701875387,
"learning_rate": 0.00014131499618726833,
"loss": 0.6285,
"step": 1195
},
{
"epoch": 0.5688551789523584,
"grad_norm": 0.05700162567210758,
"learning_rate": 0.00014007598126005797,
"loss": 0.6257,
"step": 1200
},
{
"epoch": 0.5712254088646599,
"grad_norm": 0.052067584151166996,
"learning_rate": 0.0001388376460565819,
"loss": 0.6141,
"step": 1205
},
{
"epoch": 0.5735956387769614,
"grad_norm": 0.057269236575854914,
"learning_rate": 0.00013760007539387364,
"loss": 0.626,
"step": 1210
},
{
"epoch": 0.5759658686892628,
"grad_norm": 0.05865861625327605,
"learning_rate": 0.00013636335403660092,
"loss": 0.6275,
"step": 1215
},
{
"epoch": 0.5783360986015643,
"grad_norm": 0.055532429030743166,
"learning_rate": 0.00013512756669126043,
"loss": 0.6205,
"step": 1220
},
{
"epoch": 0.5807063285138658,
"grad_norm": 0.06249705139064605,
"learning_rate": 0.00013389279800037553,
"loss": 0.6343,
"step": 1225
},
{
"epoch": 0.5830765584261673,
"grad_norm": 0.053614563419465064,
"learning_rate": 0.0001326591325366993,
"loss": 0.6113,
"step": 1230
},
{
"epoch": 0.5854467883384689,
"grad_norm": 0.05424894433907924,
"learning_rate": 0.00013142665479742163,
"loss": 0.623,
"step": 1235
},
{
"epoch": 0.5878170182507704,
"grad_norm": 0.0542014909020612,
"learning_rate": 0.00013019544919838177,
"loss": 0.6186,
"step": 1240
},
{
"epoch": 0.5901872481630718,
"grad_norm": 0.053382059431365064,
"learning_rate": 0.0001289656000682867,
"loss": 0.607,
"step": 1245
},
{
"epoch": 0.5925574780753733,
"grad_norm": 0.05601301326761515,
"learning_rate": 0.00012773719164293497,
"loss": 0.6185,
"step": 1250
},
{
"epoch": 0.5949277079876748,
"grad_norm": 0.05319350407228568,
"learning_rate": 0.00012651030805944713,
"loss": 0.6093,
"step": 1255
},
{
"epoch": 0.5972979378999763,
"grad_norm": 0.05115446334747276,
"learning_rate": 0.0001252850333505033,
"loss": 0.615,
"step": 1260
},
{
"epoch": 0.5996681678122778,
"grad_norm": 0.05192578571658311,
"learning_rate": 0.00012406145143858713,
"loss": 0.6136,
"step": 1265
},
{
"epoch": 0.6020383977245792,
"grad_norm": 0.054594901119558795,
"learning_rate": 0.00012283964613023788,
"loss": 0.6053,
"step": 1270
},
{
"epoch": 0.6044086276368807,
"grad_norm": 0.055180387588156125,
"learning_rate": 0.00012161970111031023,
"loss": 0.6235,
"step": 1275
},
{
"epoch": 0.6067788575491823,
"grad_norm": 0.05162312778472747,
"learning_rate": 0.00012040169993624276,
"loss": 0.6224,
"step": 1280
},
{
"epoch": 0.6091490874614838,
"grad_norm": 0.05862655622704016,
"learning_rate": 0.00011918572603233436,
"loss": 0.6208,
"step": 1285
},
{
"epoch": 0.6115193173737853,
"grad_norm": 0.05577369996114947,
"learning_rate": 0.00011797186268403057,
"loss": 0.6176,
"step": 1290
},
{
"epoch": 0.6138895472860868,
"grad_norm": 0.05380234343377544,
"learning_rate": 0.00011676019303221925,
"loss": 0.6133,
"step": 1295
},
{
"epoch": 0.6162597771983882,
"grad_norm": 0.058884284873815654,
"learning_rate": 0.0001155508000675357,
"loss": 0.6053,
"step": 1300
},
{
"epoch": 0.6186300071106897,
"grad_norm": 0.053535760742643955,
"learning_rate": 0.00011434376662467867,
"loss": 0.6142,
"step": 1305
},
{
"epoch": 0.6210002370229912,
"grad_norm": 0.05339335441771012,
"learning_rate": 0.00011313917537673648,
"loss": 0.6069,
"step": 1310
},
{
"epoch": 0.6233704669352927,
"grad_norm": 0.05379638473884841,
"learning_rate": 0.00011193710882952506,
"loss": 0.5951,
"step": 1315
},
{
"epoch": 0.6257406968475943,
"grad_norm": 0.05461329438051225,
"learning_rate": 0.00011073764931593622,
"loss": 0.6236,
"step": 1320
},
{
"epoch": 0.6281109267598957,
"grad_norm": 0.056457300525490715,
"learning_rate": 0.00010954087899029885,
"loss": 0.6056,
"step": 1325
},
{
"epoch": 0.6304811566721972,
"grad_norm": 0.05737998956682243,
"learning_rate": 0.00010834687982275199,
"loss": 0.6023,
"step": 1330
},
{
"epoch": 0.6328513865844987,
"grad_norm": 0.05393272686617499,
"learning_rate": 0.00010715573359363018,
"loss": 0.607,
"step": 1335
},
{
"epoch": 0.6352216164968002,
"grad_norm": 0.058430747172485496,
"learning_rate": 0.00010596752188786223,
"loss": 0.6145,
"step": 1340
},
{
"epoch": 0.6375918464091017,
"grad_norm": 0.05380330186102191,
"learning_rate": 0.00010478232608938339,
"loss": 0.6038,
"step": 1345
},
{
"epoch": 0.6399620763214032,
"grad_norm": 0.057322372930785785,
"learning_rate": 0.0001036002273755609,
"loss": 0.6081,
"step": 1350
},
{
"epoch": 0.6423323062337046,
"grad_norm": 0.059264371354861206,
"learning_rate": 0.0001024213067116341,
"loss": 0.6007,
"step": 1355
},
{
"epoch": 0.6447025361460061,
"grad_norm": 0.05436742717666459,
"learning_rate": 0.00010124564484516874,
"loss": 0.614,
"step": 1360
},
{
"epoch": 0.6470727660583077,
"grad_norm": 0.055704527406812185,
"learning_rate": 0.00010007332230052652,
"loss": 0.611,
"step": 1365
},
{
"epoch": 0.6494429959706092,
"grad_norm": 0.05458191321816212,
"learning_rate": 9.890441937334963e-05,
"loss": 0.6018,
"step": 1370
},
{
"epoch": 0.6518132258829107,
"grad_norm": 0.05555570625108052,
"learning_rate": 9.773901612506113e-05,
"loss": 0.5986,
"step": 1375
},
{
"epoch": 0.6541834557952121,
"grad_norm": 0.05694735645213942,
"learning_rate": 9.657719237738122e-05,
"loss": 0.6161,
"step": 1380
},
{
"epoch": 0.6565536857075136,
"grad_norm": 0.05412760351903458,
"learning_rate": 9.541902770686034e-05,
"loss": 0.5933,
"step": 1385
},
{
"epoch": 0.6589239156198151,
"grad_norm": 0.05413679325940457,
"learning_rate": 9.426460143942832e-05,
"loss": 0.6037,
"step": 1390
},
{
"epoch": 0.6612941455321166,
"grad_norm": 0.05174840431523283,
"learning_rate": 9.311399264496135e-05,
"loss": 0.5947,
"step": 1395
},
{
"epoch": 0.6636643754444181,
"grad_norm": 0.0532643825199599,
"learning_rate": 9.196728013186636e-05,
"loss": 0.5998,
"step": 1400
},
{
"epoch": 0.6660346053567197,
"grad_norm": 0.055311945512636886,
"learning_rate": 9.082454244168301e-05,
"loss": 0.5961,
"step": 1405
},
{
"epoch": 0.6684048352690211,
"grad_norm": 0.05305987298812406,
"learning_rate": 8.968585784370424e-05,
"loss": 0.6057,
"step": 1410
},
{
"epoch": 0.6707750651813226,
"grad_norm": 0.051852358998889965,
"learning_rate": 8.855130432961541e-05,
"loss": 0.5847,
"step": 1415
},
{
"epoch": 0.6731452950936241,
"grad_norm": 0.05695676096470251,
"learning_rate": 8.74209596081525e-05,
"loss": 0.6027,
"step": 1420
},
{
"epoch": 0.6755155250059256,
"grad_norm": 0.052927987042953824,
"learning_rate": 8.629490109977937e-05,
"loss": 0.6149,
"step": 1425
},
{
"epoch": 0.6778857549182271,
"grad_norm": 0.05418803856885555,
"learning_rate": 8.517320593138534e-05,
"loss": 0.6056,
"step": 1430
},
{
"epoch": 0.6802559848305285,
"grad_norm": 0.05515959054343932,
"learning_rate": 8.405595093100234e-05,
"loss": 0.6057,
"step": 1435
},
{
"epoch": 0.68262621474283,
"grad_norm": 0.054880537220212015,
"learning_rate": 8.294321262254287e-05,
"loss": 0.5905,
"step": 1440
},
{
"epoch": 0.6849964446551315,
"grad_norm": 0.054943521074427744,
"learning_rate": 8.183506722055836e-05,
"loss": 0.5945,
"step": 1445
},
{
"epoch": 0.6873666745674331,
"grad_norm": 0.05472729142594704,
"learning_rate": 8.073159062501963e-05,
"loss": 0.5926,
"step": 1450
},
{
"epoch": 0.6897369044797346,
"grad_norm": 0.058850442027131224,
"learning_rate": 7.963285841611758e-05,
"loss": 0.6008,
"step": 1455
},
{
"epoch": 0.6921071343920361,
"grad_norm": 0.05707369096484443,
"learning_rate": 7.853894584908711e-05,
"loss": 0.6039,
"step": 1460
},
{
"epoch": 0.6944773643043375,
"grad_norm": 0.0523587623400406,
"learning_rate": 7.744992784905206e-05,
"loss": 0.6007,
"step": 1465
},
{
"epoch": 0.696847594216639,
"grad_norm": 0.053321883103446634,
"learning_rate": 7.636587900589415e-05,
"loss": 0.5856,
"step": 1470
},
{
"epoch": 0.6992178241289405,
"grad_norm": 0.05442178077056066,
"learning_rate": 7.52868735691433e-05,
"loss": 0.6024,
"step": 1475
},
{
"epoch": 0.701588054041242,
"grad_norm": 0.05378553098177081,
"learning_rate": 7.421298544289267e-05,
"loss": 0.6041,
"step": 1480
},
{
"epoch": 0.7039582839535435,
"grad_norm": 0.05107025701826348,
"learning_rate": 7.314428818073645e-05,
"loss": 0.5978,
"step": 1485
},
{
"epoch": 0.7063285138658449,
"grad_norm": 0.056885586353233,
"learning_rate": 7.208085498073218e-05,
"loss": 0.5982,
"step": 1490
},
{
"epoch": 0.7086987437781465,
"grad_norm": 0.05367334237421415,
"learning_rate": 7.102275868038689e-05,
"loss": 0.5938,
"step": 1495
},
{
"epoch": 0.711068973690448,
"grad_norm": 0.049920747527540395,
"learning_rate": 6.997007175166862e-05,
"loss": 0.5801,
"step": 1500
},
{
"epoch": 0.7134392036027495,
"grad_norm": 0.053158788899790074,
"learning_rate": 6.892286629604244e-05,
"loss": 0.5867,
"step": 1505
},
{
"epoch": 0.715809433515051,
"grad_norm": 0.05623068287636579,
"learning_rate": 6.788121403953207e-05,
"loss": 0.5958,
"step": 1510
},
{
"epoch": 0.7181796634273524,
"grad_norm": 0.05310094430866351,
"learning_rate": 6.684518632780688e-05,
"loss": 0.5936,
"step": 1515
},
{
"epoch": 0.7205498933396539,
"grad_norm": 0.051694922971736335,
"learning_rate": 6.581485412129587e-05,
"loss": 0.5893,
"step": 1520
},
{
"epoch": 0.7229201232519554,
"grad_norm": 0.05405965263345783,
"learning_rate": 6.479028799032664e-05,
"loss": 0.603,
"step": 1525
},
{
"epoch": 0.7252903531642569,
"grad_norm": 0.05541132610068022,
"learning_rate": 6.37715581102924e-05,
"loss": 0.5887,
"step": 1530
},
{
"epoch": 0.7276605830765585,
"grad_norm": 0.05286163558669179,
"learning_rate": 6.275873425684514e-05,
"loss": 0.5881,
"step": 1535
},
{
"epoch": 0.73003081298886,
"grad_norm": 0.052024799630321195,
"learning_rate": 6.175188580111667e-05,
"loss": 0.6045,
"step": 1540
},
{
"epoch": 0.7324010429011614,
"grad_norm": 0.055373729325846666,
"learning_rate": 6.075108170496718e-05,
"loss": 0.5832,
"step": 1545
},
{
"epoch": 0.7347712728134629,
"grad_norm": 0.05516378115033556,
"learning_rate": 5.9756390516261576e-05,
"loss": 0.5934,
"step": 1550
},
{
"epoch": 0.7371415027257644,
"grad_norm": 0.05458895386019013,
"learning_rate": 5.8767880364175076e-05,
"loss": 0.6055,
"step": 1555
},
{
"epoch": 0.7395117326380659,
"grad_norm": 0.054173922532646177,
"learning_rate": 5.778561895452602e-05,
"loss": 0.5912,
"step": 1560
},
{
"epoch": 0.7418819625503674,
"grad_norm": 0.0535683841018571,
"learning_rate": 5.680967356513931e-05,
"loss": 0.5846,
"step": 1565
},
{
"epoch": 0.7442521924626688,
"grad_norm": 0.05450496516039284,
"learning_rate": 5.5840111041237614e-05,
"loss": 0.5813,
"step": 1570
},
{
"epoch": 0.7466224223749703,
"grad_norm": 0.054578000158860905,
"learning_rate": 5.48769977908638e-05,
"loss": 0.593,
"step": 1575
},
{
"epoch": 0.7489926522872719,
"grad_norm": 0.0515712714221276,
"learning_rate": 5.392039978033157e-05,
"loss": 0.5927,
"step": 1580
},
{
"epoch": 0.7513628821995734,
"grad_norm": 0.05321518991804319,
"learning_rate": 5.29703825297079e-05,
"loss": 0.5831,
"step": 1585
},
{
"epoch": 0.7537331121118749,
"grad_norm": 0.05298981424430984,
"learning_rate": 5.202701110832513e-05,
"loss": 0.5892,
"step": 1590
},
{
"epoch": 0.7561033420241764,
"grad_norm": 0.052907289573775475,
"learning_rate": 5.1090350130324196e-05,
"loss": 0.5943,
"step": 1595
},
{
"epoch": 0.7584735719364778,
"grad_norm": 0.05190800487566414,
"learning_rate": 5.016046375022892e-05,
"loss": 0.5802,
"step": 1600
},
{
"epoch": 0.7608438018487793,
"grad_norm": 0.0538931538504174,
"learning_rate": 4.923741565855231e-05,
"loss": 0.5818,
"step": 1605
},
{
"epoch": 0.7632140317610808,
"grad_norm": 0.052494006253503485,
"learning_rate": 4.832126907743356e-05,
"loss": 0.5861,
"step": 1610
},
{
"epoch": 0.7655842616733823,
"grad_norm": 0.05335836205095365,
"learning_rate": 4.7412086756308456e-05,
"loss": 0.5834,
"step": 1615
},
{
"epoch": 0.7679544915856839,
"grad_norm": 0.05509114267028979,
"learning_rate": 4.650993096761088e-05,
"loss": 0.5869,
"step": 1620
},
{
"epoch": 0.7703247214979853,
"grad_norm": 0.052187260630405116,
"learning_rate": 4.5614863502508274e-05,
"loss": 0.6069,
"step": 1625
},
{
"epoch": 0.7726949514102868,
"grad_norm": 0.05431858623792003,
"learning_rate": 4.4726945666668675e-05,
"loss": 0.5946,
"step": 1630
},
{
"epoch": 0.7750651813225883,
"grad_norm": 0.052723752996540384,
"learning_rate": 4.384623827606225e-05,
"loss": 0.5789,
"step": 1635
},
{
"epoch": 0.7774354112348898,
"grad_norm": 0.05256027199703368,
"learning_rate": 4.297280165279562e-05,
"loss": 0.5716,
"step": 1640
},
{
"epoch": 0.7798056411471913,
"grad_norm": 0.054172042789112265,
"learning_rate": 4.210669562098021e-05,
"loss": 0.5873,
"step": 1645
},
{
"epoch": 0.7821758710594928,
"grad_norm": 0.05964267895580253,
"learning_rate": 4.124797950263491e-05,
"loss": 0.5883,
"step": 1650
},
{
"epoch": 0.7845461009717942,
"grad_norm": 0.052843311666916505,
"learning_rate": 4.039671211362262e-05,
"loss": 0.5912,
"step": 1655
},
{
"epoch": 0.7869163308840957,
"grad_norm": 0.05799110456913064,
"learning_rate": 3.9552951759622134e-05,
"loss": 0.5836,
"step": 1660
},
{
"epoch": 0.7892865607963973,
"grad_norm": 0.053733688999035426,
"learning_rate": 3.8716756232134436e-05,
"loss": 0.5873,
"step": 1665
},
{
"epoch": 0.7916567907086988,
"grad_norm": 0.05503743591767426,
"learning_rate": 3.788818280452443e-05,
"loss": 0.5806,
"step": 1670
},
{
"epoch": 0.7940270206210003,
"grad_norm": 0.05314889782998826,
"learning_rate": 3.706728822809813e-05,
"loss": 0.5859,
"step": 1675
},
{
"epoch": 0.7963972505333017,
"grad_norm": 0.05263806721046257,
"learning_rate": 3.625412872821565e-05,
"loss": 0.5701,
"step": 1680
},
{
"epoch": 0.7987674804456032,
"grad_norm": 0.053056629300361355,
"learning_rate": 3.544876000043997e-05,
"loss": 0.5689,
"step": 1685
},
{
"epoch": 0.8011377103579047,
"grad_norm": 0.05166101520125829,
"learning_rate": 3.465123720672255e-05,
"loss": 0.5624,
"step": 1690
},
{
"epoch": 0.8035079402702062,
"grad_norm": 0.05341566401567655,
"learning_rate": 3.3861614971624825e-05,
"loss": 0.597,
"step": 1695
},
{
"epoch": 0.8058781701825077,
"grad_norm": 0.05184611062114667,
"learning_rate": 3.3079947378577e-05,
"loss": 0.589,
"step": 1700
},
{
"epoch": 0.8082484000948092,
"grad_norm": 0.05652651621528694,
"learning_rate": 3.230628796617345e-05,
"loss": 0.5656,
"step": 1705
},
{
"epoch": 0.8106186300071107,
"grad_norm": 0.05185062445049837,
"learning_rate": 3.154068972450626e-05,
"loss": 0.5803,
"step": 1710
},
{
"epoch": 0.8129888599194122,
"grad_norm": 0.05236594549663459,
"learning_rate": 3.07832050915351e-05,
"loss": 0.5687,
"step": 1715
},
{
"epoch": 0.8153590898317137,
"grad_norm": 0.052981202994182464,
"learning_rate": 3.0033885949496157e-05,
"loss": 0.5845,
"step": 1720
},
{
"epoch": 0.8177293197440152,
"grad_norm": 0.05119910241673355,
"learning_rate": 2.9292783621348176e-05,
"loss": 0.5881,
"step": 1725
},
{
"epoch": 0.8200995496563167,
"grad_norm": 0.05198844865138733,
"learning_rate": 2.8559948867257643e-05,
"loss": 0.5733,
"step": 1730
},
{
"epoch": 0.8224697795686181,
"grad_norm": 0.0664972668514182,
"learning_rate": 2.783543188112156e-05,
"loss": 0.5748,
"step": 1735
},
{
"epoch": 0.8248400094809196,
"grad_norm": 0.05135611458630472,
"learning_rate": 2.711928228712999e-05,
"loss": 0.5649,
"step": 1740
},
{
"epoch": 0.8272102393932211,
"grad_norm": 0.05397408303013281,
"learning_rate": 2.6411549136366874e-05,
"loss": 0.5889,
"step": 1745
},
{
"epoch": 0.8295804693055226,
"grad_norm": 0.052454038077671426,
"learning_rate": 2.5712280903450495e-05,
"loss": 0.6,
"step": 1750
},
{
"epoch": 0.8319506992178242,
"grad_norm": 0.05123417132112601,
"learning_rate": 2.502152548321324e-05,
"loss": 0.5707,
"step": 1755
},
{
"epoch": 0.8343209291301257,
"grad_norm": 0.05251617509320771,
"learning_rate": 2.4339330187421264e-05,
"loss": 0.5769,
"step": 1760
},
{
"epoch": 0.8366911590424271,
"grad_norm": 0.06318258557079645,
"learning_rate": 2.3665741741533918e-05,
"loss": 0.5727,
"step": 1765
},
{
"epoch": 0.8390613889547286,
"grad_norm": 0.05419644476733012,
"learning_rate": 2.3000806281503324e-05,
"loss": 0.5639,
"step": 1770
},
{
"epoch": 0.8414316188670301,
"grad_norm": 0.0529717737403971,
"learning_rate": 2.2344569350614548e-05,
"loss": 0.5747,
"step": 1775
},
{
"epoch": 0.8438018487793316,
"grad_norm": 0.053183308485240915,
"learning_rate": 2.169707589636606e-05,
"loss": 0.5756,
"step": 1780
},
{
"epoch": 0.8461720786916331,
"grad_norm": 0.048981198328497996,
"learning_rate": 2.1058370267391245e-05,
"loss": 0.5594,
"step": 1785
},
{
"epoch": 0.8485423086039345,
"grad_norm": 0.047709819338697244,
"learning_rate": 2.0428496210420753e-05,
"loss": 0.5814,
"step": 1790
},
{
"epoch": 0.850912538516236,
"grad_norm": 0.051806025941571895,
"learning_rate": 1.980749686728631e-05,
"loss": 0.5853,
"step": 1795
},
{
"epoch": 0.8532827684285376,
"grad_norm": 0.05238919786303432,
"learning_rate": 1.919541477196569e-05,
"loss": 0.5793,
"step": 1800
},
{
"epoch": 0.8556529983408391,
"grad_norm": 0.050957261348074434,
"learning_rate": 1.859229184766951e-05,
"loss": 0.5724,
"step": 1805
},
{
"epoch": 0.8580232282531406,
"grad_norm": 0.052439653700969896,
"learning_rate": 1.7998169403969652e-05,
"loss": 0.5794,
"step": 1810
},
{
"epoch": 0.860393458165442,
"grad_norm": 0.05158899579880472,
"learning_rate": 1.741308813397017e-05,
"loss": 0.587,
"step": 1815
},
{
"epoch": 0.8627636880777435,
"grad_norm": 0.05347972640720942,
"learning_rate": 1.683708811151972e-05,
"loss": 0.5737,
"step": 1820
},
{
"epoch": 0.865133917990045,
"grad_norm": 0.05251163235177713,
"learning_rate": 1.6270208788467054e-05,
"loss": 0.5851,
"step": 1825
},
{
"epoch": 0.8675041479023465,
"grad_norm": 0.05130891776047808,
"learning_rate": 1.5712488991958773e-05,
"loss": 0.5653,
"step": 1830
},
{
"epoch": 0.869874377814648,
"grad_norm": 0.051624628347411626,
"learning_rate": 1.5163966921779974e-05,
"loss": 0.5631,
"step": 1835
},
{
"epoch": 0.8722446077269496,
"grad_norm": 0.05192898094836235,
"learning_rate": 1.4624680147737694e-05,
"loss": 0.5701,
"step": 1840
},
{
"epoch": 0.874614837639251,
"grad_norm": 0.05273604182501612,
"learning_rate": 1.4094665607087919e-05,
"loss": 0.5739,
"step": 1845
},
{
"epoch": 0.8769850675515525,
"grad_norm": 0.05064520739634672,
"learning_rate": 1.357395960200538e-05,
"loss": 0.5692,
"step": 1850
},
{
"epoch": 0.879355297463854,
"grad_norm": 0.050839315021426905,
"learning_rate": 1.3062597797097346e-05,
"loss": 0.5704,
"step": 1855
},
{
"epoch": 0.8817255273761555,
"grad_norm": 0.051292554412449225,
"learning_rate": 1.2560615216960607e-05,
"loss": 0.5695,
"step": 1860
},
{
"epoch": 0.884095757288457,
"grad_norm": 0.05149899783415366,
"learning_rate": 1.2068046243782826e-05,
"loss": 0.5708,
"step": 1865
},
{
"epoch": 0.8864659872007584,
"grad_norm": 0.05155770982273923,
"learning_rate": 1.1584924614987323e-05,
"loss": 0.5682,
"step": 1870
},
{
"epoch": 0.8888362171130599,
"grad_norm": 0.05263051860972721,
"learning_rate": 1.1111283420922528e-05,
"loss": 0.5788,
"step": 1875
},
{
"epoch": 0.8912064470253614,
"grad_norm": 0.05226957394199282,
"learning_rate": 1.0647155102595406e-05,
"loss": 0.5665,
"step": 1880
},
{
"epoch": 0.893576676937663,
"grad_norm": 0.05393679935274463,
"learning_rate": 1.0192571449449567e-05,
"loss": 0.5782,
"step": 1885
},
{
"epoch": 0.8959469068499645,
"grad_norm": 0.05290726883653969,
"learning_rate": 9.747563597187791e-06,
"loss": 0.5725,
"step": 1890
},
{
"epoch": 0.898317136762266,
"grad_norm": 0.052272530064695,
"learning_rate": 9.312162025639575e-06,
"loss": 0.5696,
"step": 1895
},
{
"epoch": 0.9006873666745674,
"grad_norm": 0.052434933736767136,
"learning_rate": 8.886396556673476e-06,
"loss": 0.5799,
"step": 1900
},
{
"epoch": 0.9030575965868689,
"grad_norm": 0.049740509342076265,
"learning_rate": 8.470296352154466e-06,
"loss": 0.5586,
"step": 1905
},
{
"epoch": 0.9054278264991704,
"grad_norm": 0.05205499325333267,
"learning_rate": 8.063889911946597e-06,
"loss": 0.5715,
"step": 1910
},
{
"epoch": 0.9077980564114719,
"grad_norm": 0.05138219725318278,
"learning_rate": 7.667205071960897e-06,
"loss": 0.5706,
"step": 1915
},
{
"epoch": 0.9101682863237734,
"grad_norm": 0.053660634763309435,
"learning_rate": 7.280269002249006e-06,
"loss": 0.5761,
"step": 1920
},
{
"epoch": 0.9125385162360748,
"grad_norm": 0.05161198079700086,
"learning_rate": 6.903108205141922e-06,
"loss": 0.5676,
"step": 1925
},
{
"epoch": 0.9149087461483764,
"grad_norm": 0.05035126816965241,
"learning_rate": 6.535748513435063e-06,
"loss": 0.5661,
"step": 1930
},
{
"epoch": 0.9172789760606779,
"grad_norm": 0.05307717275910678,
"learning_rate": 6.178215088618754e-06,
"loss": 0.5679,
"step": 1935
},
{
"epoch": 0.9196492059729794,
"grad_norm": 0.05151414534729879,
"learning_rate": 5.830532419154904e-06,
"loss": 0.5655,
"step": 1940
},
{
"epoch": 0.9220194358852809,
"grad_norm": 0.05138066400564488,
"learning_rate": 5.492724318799613e-06,
"loss": 0.5745,
"step": 1945
},
{
"epoch": 0.9243896657975824,
"grad_norm": 0.053677080886407436,
"learning_rate": 5.164813924972294e-06,
"loss": 0.5845,
"step": 1950
},
{
"epoch": 0.9267598957098838,
"grad_norm": 0.052774180712557975,
"learning_rate": 4.846823697170793e-06,
"loss": 0.5766,
"step": 1955
},
{
"epoch": 0.9291301256221853,
"grad_norm": 0.0521259881298222,
"learning_rate": 4.538775415433099e-06,
"loss": 0.5817,
"step": 1960
},
{
"epoch": 0.9315003555344868,
"grad_norm": 0.05114820813668919,
"learning_rate": 4.240690178845507e-06,
"loss": 0.586,
"step": 1965
},
{
"epoch": 0.9338705854467884,
"grad_norm": 0.054406572055921386,
"learning_rate": 3.952588404097706e-06,
"loss": 0.5763,
"step": 1970
},
{
"epoch": 0.9362408153590899,
"grad_norm": 0.051364642860872965,
"learning_rate": 3.6744898240840816e-06,
"loss": 0.5724,
"step": 1975
},
{
"epoch": 0.9386110452713913,
"grad_norm": 0.051346137904923476,
"learning_rate": 3.4064134865523617e-06,
"loss": 0.5682,
"step": 1980
},
{
"epoch": 0.9409812751836928,
"grad_norm": 0.051110496158009784,
"learning_rate": 3.148377752798914e-06,
"loss": 0.5693,
"step": 1985
},
{
"epoch": 0.9433515050959943,
"grad_norm": 0.05315034654966591,
"learning_rate": 2.9004002964111485e-06,
"loss": 0.5755,
"step": 1990
},
{
"epoch": 0.9457217350082958,
"grad_norm": 0.052170597166830944,
"learning_rate": 2.66249810205697e-06,
"loss": 0.5796,
"step": 1995
},
{
"epoch": 0.9480919649205973,
"grad_norm": 0.05200440928257493,
"learning_rate": 2.4346874643214774e-06,
"loss": 0.5717,
"step": 2000
},
{
"epoch": 0.9504621948328988,
"grad_norm": 0.05163936837645852,
"learning_rate": 2.2169839865909036e-06,
"loss": 0.5631,
"step": 2005
},
{
"epoch": 0.9528324247452002,
"grad_norm": 0.049675359269013855,
"learning_rate": 2.009402579983921e-06,
"loss": 0.5716,
"step": 2010
},
{
"epoch": 0.9552026546575018,
"grad_norm": 0.051088496480251495,
"learning_rate": 1.8119574623302436e-06,
"loss": 0.561,
"step": 2015
},
{
"epoch": 0.9575728845698033,
"grad_norm": 0.05018933256487338,
"learning_rate": 1.6246621571969376e-06,
"loss": 0.5686,
"step": 2020
},
{
"epoch": 0.9599431144821048,
"grad_norm": 0.05409872852152585,
"learning_rate": 1.447529492962063e-06,
"loss": 0.5658,
"step": 2025
},
{
"epoch": 0.9623133443944063,
"grad_norm": 0.05217850594057917,
"learning_rate": 1.280571601936009e-06,
"loss": 0.5643,
"step": 2030
},
{
"epoch": 0.9646835743067077,
"grad_norm": 0.052394586476799676,
"learning_rate": 1.1237999195306102e-06,
"loss": 0.575,
"step": 2035
},
{
"epoch": 0.9670538042190092,
"grad_norm": 0.05288849803820924,
"learning_rate": 9.77225183475855e-07,
"loss": 0.5753,
"step": 2040
},
{
"epoch": 0.9694240341313107,
"grad_norm": 0.05313372067559112,
"learning_rate": 8.408574330843576e-07,
"loss": 0.5745,
"step": 2045
},
{
"epoch": 0.9717942640436122,
"grad_norm": 0.0529192880219954,
"learning_rate": 7.147060085638757e-07,
"loss": 0.5749,
"step": 2050
},
{
"epoch": 0.9741644939559138,
"grad_norm": 0.05401698190339256,
"learning_rate": 5.987795503774706e-07,
"loss": 0.5743,
"step": 2055
},
{
"epoch": 0.9765347238682153,
"grad_norm": 0.05269598387655656,
"learning_rate": 4.930859986517322e-07,
"loss": 0.5696,
"step": 2060
},
{
"epoch": 0.9789049537805167,
"grad_norm": 0.05202660809884899,
"learning_rate": 3.9763259263292936e-07,
"loss": 0.5596,
"step": 2065
},
{
"epoch": 0.9812751836928182,
"grad_norm": 0.06891661422461348,
"learning_rate": 3.1242587019117417e-07,
"loss": 0.5855,
"step": 2070
},
{
"epoch": 0.9836454136051197,
"grad_norm": 0.04973359674018789,
"learning_rate": 2.3747166737264622e-07,
"loss": 0.5651,
"step": 2075
},
{
"epoch": 0.9860156435174212,
"grad_norm": 0.051060019252012756,
"learning_rate": 1.7277511799984601e-07,
"loss": 0.564,
"step": 2080
},
{
"epoch": 0.9883858734297227,
"grad_norm": 0.05056552339235572,
"learning_rate": 1.1834065331995935e-07,
"loss": 0.5585,
"step": 2085
},
{
"epoch": 0.9907561033420241,
"grad_norm": 0.05126255919954496,
"learning_rate": 7.41720017013503e-08,
"loss": 0.5649,
"step": 2090
},
{
"epoch": 0.9931263332543256,
"grad_norm": 0.05801752774330538,
"learning_rate": 4.027218837821533e-08,
"loss": 0.5712,
"step": 2095
},
{
"epoch": 0.9954965631666272,
"grad_norm": 0.05065216789680233,
"learning_rate": 1.6643535243365815e-08,
"loss": 0.5745,
"step": 2100
},
{
"epoch": 0.9978667930789287,
"grad_norm": 0.05241312893170103,
"learning_rate": 3.287660689171856e-09,
"loss": 0.5783,
"step": 2105
},
{
"epoch": 0.9997629770087698,
"eval_loss": 0.5719642639160156,
"eval_runtime": 101.3361,
"eval_samples_per_second": 296.045,
"eval_steps_per_second": 9.256,
"step": 2109
},
{
"epoch": 0.9997629770087698,
"step": 2109,
"total_flos": 141388749045760.0,
"train_loss": 0.6815844999777304,
"train_runtime": 3835.2311,
"train_samples_per_second": 70.4,
"train_steps_per_second": 0.55
}
],
"logging_steps": 5,
"max_steps": 2109,
"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": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 141388749045760.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}