nothink_8b_200k / trainer_state.json
modrill's picture
Upload folder using huggingface_hub
a8888ad verified
Raw
History Blame Contribute Delete
111 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 6250,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0032,
"grad_norm": 3.4761534528703826,
"learning_rate": 9.574468085106384e-07,
"loss": 0.5382460594177246,
"step": 10
},
{
"epoch": 0.0064,
"grad_norm": 1.6969362338939127,
"learning_rate": 2.021276595744681e-06,
"loss": 0.4980194091796875,
"step": 20
},
{
"epoch": 0.0096,
"grad_norm": 1.5291612326138313,
"learning_rate": 3.0851063829787237e-06,
"loss": 0.43514370918273926,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 1.2077494143077712,
"learning_rate": 4.148936170212766e-06,
"loss": 0.4033975601196289,
"step": 40
},
{
"epoch": 0.016,
"grad_norm": 1.2213130190976706,
"learning_rate": 5.212765957446809e-06,
"loss": 0.38650567531585694,
"step": 50
},
{
"epoch": 0.0192,
"grad_norm": 1.270184532343321,
"learning_rate": 6.276595744680851e-06,
"loss": 0.38349549770355223,
"step": 60
},
{
"epoch": 0.0224,
"grad_norm": 1.2083859946854583,
"learning_rate": 7.340425531914894e-06,
"loss": 0.3722045421600342,
"step": 70
},
{
"epoch": 0.0256,
"grad_norm": 1.174299356926062,
"learning_rate": 8.404255319148937e-06,
"loss": 0.34796991348266604,
"step": 80
},
{
"epoch": 0.0288,
"grad_norm": 1.1731280203577552,
"learning_rate": 9.46808510638298e-06,
"loss": 0.3427796602249146,
"step": 90
},
{
"epoch": 0.032,
"grad_norm": 1.0966961972086104,
"learning_rate": 1.0531914893617022e-05,
"loss": 0.34212379455566405,
"step": 100
},
{
"epoch": 0.0352,
"grad_norm": 1.167216653195505,
"learning_rate": 1.1595744680851065e-05,
"loss": 0.3391818284988403,
"step": 110
},
{
"epoch": 0.0384,
"grad_norm": 1.0580126635168647,
"learning_rate": 1.2659574468085108e-05,
"loss": 0.3281343221664429,
"step": 120
},
{
"epoch": 0.0416,
"grad_norm": 1.274987116104605,
"learning_rate": 1.372340425531915e-05,
"loss": 0.3355553388595581,
"step": 130
},
{
"epoch": 0.0448,
"grad_norm": 0.9729281885233478,
"learning_rate": 1.4787234042553193e-05,
"loss": 0.3120665788650513,
"step": 140
},
{
"epoch": 0.048,
"grad_norm": 1.1234676191413073,
"learning_rate": 1.5851063829787235e-05,
"loss": 0.32906901836395264,
"step": 150
},
{
"epoch": 0.0512,
"grad_norm": 1.0253515819885477,
"learning_rate": 1.6914893617021276e-05,
"loss": 0.31854383945465087,
"step": 160
},
{
"epoch": 0.0544,
"grad_norm": 1.0110118751725905,
"learning_rate": 1.797872340425532e-05,
"loss": 0.31460957527160643,
"step": 170
},
{
"epoch": 0.0576,
"grad_norm": 0.7998154487362293,
"learning_rate": 1.9042553191489363e-05,
"loss": 0.3108847141265869,
"step": 180
},
{
"epoch": 0.0608,
"grad_norm": 0.8571090253498624,
"learning_rate": 1.9999998657117928e-05,
"loss": 0.32082979679107665,
"step": 190
},
{
"epoch": 0.064,
"grad_norm": 0.8479730296902297,
"learning_rate": 1.9999837511705667e-05,
"loss": 0.30427350997924807,
"step": 200
},
{
"epoch": 0.0672,
"grad_norm": 0.8760108811001399,
"learning_rate": 1.9999407794838095e-05,
"loss": 0.31640825271606443,
"step": 210
},
{
"epoch": 0.0704,
"grad_norm": 0.8007164876314884,
"learning_rate": 1.999870951805638e-05,
"loss": 0.3111215114593506,
"step": 220
},
{
"epoch": 0.0736,
"grad_norm": 0.9161712382711664,
"learning_rate": 1.999774270011454e-05,
"loss": 0.30509908199310304,
"step": 230
},
{
"epoch": 0.0768,
"grad_norm": 0.7471036542504835,
"learning_rate": 1.9996507366978966e-05,
"loss": 0.3062256097793579,
"step": 240
},
{
"epoch": 0.08,
"grad_norm": 0.8137152336077607,
"learning_rate": 1.999500355182772e-05,
"loss": 0.3118621826171875,
"step": 250
},
{
"epoch": 0.0832,
"grad_norm": 0.7499071366260794,
"learning_rate": 1.9993231295049648e-05,
"loss": 0.30167379379272463,
"step": 260
},
{
"epoch": 0.0864,
"grad_norm": 0.7169239854749703,
"learning_rate": 1.9991190644243267e-05,
"loss": 0.30208702087402345,
"step": 270
},
{
"epoch": 0.0896,
"grad_norm": 0.7374882704576025,
"learning_rate": 1.998888165421554e-05,
"loss": 0.3055035352706909,
"step": 280
},
{
"epoch": 0.0928,
"grad_norm": 0.7879676401361891,
"learning_rate": 1.998630438698034e-05,
"loss": 0.29742021560668946,
"step": 290
},
{
"epoch": 0.096,
"grad_norm": 0.7057839394598209,
"learning_rate": 1.9983458911756847e-05,
"loss": 0.3093631505966187,
"step": 300
},
{
"epoch": 0.0992,
"grad_norm": 0.6954721509404341,
"learning_rate": 1.9980345304967635e-05,
"loss": 0.29180073738098145,
"step": 310
},
{
"epoch": 0.1024,
"grad_norm": 0.7492378339689201,
"learning_rate": 1.9976963650236657e-05,
"loss": 0.29152491092681887,
"step": 320
},
{
"epoch": 0.1056,
"grad_norm": 0.6704933595023297,
"learning_rate": 1.9973314038386983e-05,
"loss": 0.2992133617401123,
"step": 330
},
{
"epoch": 0.1088,
"grad_norm": 0.5977670382643498,
"learning_rate": 1.996939656743836e-05,
"loss": 0.28931488990783694,
"step": 340
},
{
"epoch": 0.112,
"grad_norm": 0.5904456666286032,
"learning_rate": 1.9965211342604586e-05,
"loss": 0.2889702796936035,
"step": 350
},
{
"epoch": 0.1152,
"grad_norm": 0.6560078865860166,
"learning_rate": 1.9960758476290684e-05,
"loss": 0.2911426544189453,
"step": 360
},
{
"epoch": 0.1184,
"grad_norm": 0.6500562277646021,
"learning_rate": 1.9956038088089865e-05,
"loss": 0.2970966577529907,
"step": 370
},
{
"epoch": 0.1216,
"grad_norm": 0.648181642485819,
"learning_rate": 1.9951050304780357e-05,
"loss": 0.2893670558929443,
"step": 380
},
{
"epoch": 0.1248,
"grad_norm": 0.6193783967918901,
"learning_rate": 1.9945795260321944e-05,
"loss": 0.29062709808349607,
"step": 390
},
{
"epoch": 0.128,
"grad_norm": 0.6356935820653824,
"learning_rate": 1.994027309585242e-05,
"loss": 0.28874099254608154,
"step": 400
},
{
"epoch": 0.1312,
"grad_norm": 0.6386947730890384,
"learning_rate": 1.9934483959683773e-05,
"loss": 0.27861876487731935,
"step": 410
},
{
"epoch": 0.1344,
"grad_norm": 0.5931217059762242,
"learning_rate": 1.99284280072982e-05,
"loss": 0.2877832889556885,
"step": 420
},
{
"epoch": 0.1376,
"grad_norm": 0.7258232974872497,
"learning_rate": 1.9922105401343933e-05,
"loss": 0.27797112464904783,
"step": 430
},
{
"epoch": 0.1408,
"grad_norm": 0.6260089195159241,
"learning_rate": 1.9915516311630885e-05,
"loss": 0.2892880439758301,
"step": 440
},
{
"epoch": 0.144,
"grad_norm": 0.646692568847579,
"learning_rate": 1.990866091512607e-05,
"loss": 0.2822598457336426,
"step": 450
},
{
"epoch": 0.1472,
"grad_norm": 0.6530455991172327,
"learning_rate": 1.9901539395948864e-05,
"loss": 0.2818694829940796,
"step": 460
},
{
"epoch": 0.1504,
"grad_norm": 0.6750198663602146,
"learning_rate": 1.989415194536605e-05,
"loss": 0.28128292560577395,
"step": 470
},
{
"epoch": 0.1536,
"grad_norm": 0.6828529904560057,
"learning_rate": 1.9886498761786685e-05,
"loss": 0.2903521776199341,
"step": 480
},
{
"epoch": 0.1568,
"grad_norm": 0.5724952516230734,
"learning_rate": 1.987858005075678e-05,
"loss": 0.2736215114593506,
"step": 490
},
{
"epoch": 0.16,
"grad_norm": 0.6049468863582851,
"learning_rate": 1.987039602495376e-05,
"loss": 0.2764644384384155,
"step": 500
},
{
"epoch": 0.1632,
"grad_norm": 0.5657503603897129,
"learning_rate": 1.9861946904180772e-05,
"loss": 0.2722026824951172,
"step": 510
},
{
"epoch": 0.1664,
"grad_norm": 0.6559347227240493,
"learning_rate": 1.985323291536077e-05,
"loss": 0.2760221004486084,
"step": 520
},
{
"epoch": 0.1696,
"grad_norm": 0.5724398365673927,
"learning_rate": 1.9844254292530418e-05,
"loss": 0.2836419105529785,
"step": 530
},
{
"epoch": 0.1728,
"grad_norm": 0.5364089510917117,
"learning_rate": 1.983501127683382e-05,
"loss": 0.2790844440460205,
"step": 540
},
{
"epoch": 0.176,
"grad_norm": 0.5824389123000472,
"learning_rate": 1.982550411651602e-05,
"loss": 0.28430187702178955,
"step": 550
},
{
"epoch": 0.1792,
"grad_norm": 0.5989456956659608,
"learning_rate": 1.981573306691636e-05,
"loss": 0.2710455894470215,
"step": 560
},
{
"epoch": 0.1824,
"grad_norm": 0.5795466763717597,
"learning_rate": 1.9805698390461606e-05,
"loss": 0.2826793670654297,
"step": 570
},
{
"epoch": 0.1856,
"grad_norm": 0.5805465878317628,
"learning_rate": 1.9795400356658906e-05,
"loss": 0.2649863243103027,
"step": 580
},
{
"epoch": 0.1888,
"grad_norm": 0.563042233372079,
"learning_rate": 1.978483924208854e-05,
"loss": 0.26342689990997314,
"step": 590
},
{
"epoch": 0.192,
"grad_norm": 0.5840344311346182,
"learning_rate": 1.977401533039651e-05,
"loss": 0.2620288133621216,
"step": 600
},
{
"epoch": 0.1952,
"grad_norm": 0.5647153834400541,
"learning_rate": 1.976292891228691e-05,
"loss": 0.284657883644104,
"step": 610
},
{
"epoch": 0.1984,
"grad_norm": 0.5687696568893406,
"learning_rate": 1.9751580285514123e-05,
"loss": 0.26548752784729,
"step": 620
},
{
"epoch": 0.2016,
"grad_norm": 0.6091240198963148,
"learning_rate": 1.9739969754874827e-05,
"loss": 0.2517454385757446,
"step": 630
},
{
"epoch": 0.2048,
"grad_norm": 0.6200887397540381,
"learning_rate": 1.9728097632199795e-05,
"loss": 0.26299190521240234,
"step": 640
},
{
"epoch": 0.208,
"grad_norm": 0.6451516814867,
"learning_rate": 1.9715964236345537e-05,
"loss": 0.27175164222717285,
"step": 650
},
{
"epoch": 0.2112,
"grad_norm": 0.5326090728563391,
"learning_rate": 1.9703569893185723e-05,
"loss": 0.27169122695922854,
"step": 660
},
{
"epoch": 0.2144,
"grad_norm": 0.5668695642822408,
"learning_rate": 1.9690914935602446e-05,
"loss": 0.26070384979248046,
"step": 670
},
{
"epoch": 0.2176,
"grad_norm": 0.5212043757960159,
"learning_rate": 1.9677999703477262e-05,
"loss": 0.2644060134887695,
"step": 680
},
{
"epoch": 0.2208,
"grad_norm": 0.5873639247758615,
"learning_rate": 1.9664824543682074e-05,
"loss": 0.2679416656494141,
"step": 690
},
{
"epoch": 0.224,
"grad_norm": 0.5985639925649104,
"learning_rate": 1.965138981006982e-05,
"loss": 0.2744985818862915,
"step": 700
},
{
"epoch": 0.2272,
"grad_norm": 0.5343657335107097,
"learning_rate": 1.963769586346496e-05,
"loss": 0.26201353073120115,
"step": 710
},
{
"epoch": 0.2304,
"grad_norm": 0.5821694229101441,
"learning_rate": 1.962374307165378e-05,
"loss": 0.2570213317871094,
"step": 720
},
{
"epoch": 0.2336,
"grad_norm": 0.5644403425910355,
"learning_rate": 1.9609531809374537e-05,
"loss": 0.2660036087036133,
"step": 730
},
{
"epoch": 0.2368,
"grad_norm": 0.5772939814482504,
"learning_rate": 1.9595062458307363e-05,
"loss": 0.26155648231506345,
"step": 740
},
{
"epoch": 0.24,
"grad_norm": 0.6116595759992614,
"learning_rate": 1.958033540706405e-05,
"loss": 0.2499995708465576,
"step": 750
},
{
"epoch": 0.2432,
"grad_norm": 0.5886232735686205,
"learning_rate": 1.9565351051177575e-05,
"loss": 0.26529579162597655,
"step": 760
},
{
"epoch": 0.2464,
"grad_norm": 0.558618488824251,
"learning_rate": 1.955010979309151e-05,
"loss": 0.25505809783935546,
"step": 770
},
{
"epoch": 0.2496,
"grad_norm": 0.6123776265814855,
"learning_rate": 1.9534612042149187e-05,
"loss": 0.2727694272994995,
"step": 780
},
{
"epoch": 0.2528,
"grad_norm": 0.5477526048856636,
"learning_rate": 1.951885821458273e-05,
"loss": 0.2434636354446411,
"step": 790
},
{
"epoch": 0.256,
"grad_norm": 0.5239755378033735,
"learning_rate": 1.9502848733501845e-05,
"loss": 0.2573689460754395,
"step": 800
},
{
"epoch": 0.2592,
"grad_norm": 0.5468554090686162,
"learning_rate": 1.9486584028882486e-05,
"loss": 0.25816586017608645,
"step": 810
},
{
"epoch": 0.2624,
"grad_norm": 0.6022591326954797,
"learning_rate": 1.9470064537555294e-05,
"loss": 0.25444707870483396,
"step": 820
},
{
"epoch": 0.2656,
"grad_norm": 0.609651486638484,
"learning_rate": 1.9453290703193852e-05,
"loss": 0.2577912092208862,
"step": 830
},
{
"epoch": 0.2688,
"grad_norm": 0.5515372419634454,
"learning_rate": 1.9436262976302797e-05,
"loss": 0.25690488815307616,
"step": 840
},
{
"epoch": 0.272,
"grad_norm": 0.5273411346926319,
"learning_rate": 1.9418981814205697e-05,
"loss": 0.264233922958374,
"step": 850
},
{
"epoch": 0.2752,
"grad_norm": 0.5230867978500603,
"learning_rate": 1.940144768103278e-05,
"loss": 0.24478046894073485,
"step": 860
},
{
"epoch": 0.2784,
"grad_norm": 0.5689502990590942,
"learning_rate": 1.9383661047708462e-05,
"loss": 0.25587031841278074,
"step": 870
},
{
"epoch": 0.2816,
"grad_norm": 0.46953986024477196,
"learning_rate": 1.936562239193871e-05,
"loss": 0.24489195346832277,
"step": 880
},
{
"epoch": 0.2848,
"grad_norm": 0.5530054727758975,
"learning_rate": 1.934733219819819e-05,
"loss": 0.2520304679870605,
"step": 890
},
{
"epoch": 0.288,
"grad_norm": 0.74048996090797,
"learning_rate": 1.932879095771729e-05,
"loss": 0.24458229541778564,
"step": 900
},
{
"epoch": 0.2912,
"grad_norm": 0.4910430125614687,
"learning_rate": 1.9309999168468887e-05,
"loss": 0.24846596717834474,
"step": 910
},
{
"epoch": 0.2944,
"grad_norm": 0.5021401960451934,
"learning_rate": 1.9290957335155004e-05,
"loss": 0.2496645450592041,
"step": 920
},
{
"epoch": 0.2976,
"grad_norm": 0.5839254046786111,
"learning_rate": 1.927166596919324e-05,
"loss": 0.2453162908554077,
"step": 930
},
{
"epoch": 0.3008,
"grad_norm": 0.5059837845975429,
"learning_rate": 1.9252125588703036e-05,
"loss": 0.2580821752548218,
"step": 940
},
{
"epoch": 0.304,
"grad_norm": 0.5099365623548049,
"learning_rate": 1.923233671849176e-05,
"loss": 0.2529602527618408,
"step": 950
},
{
"epoch": 0.3072,
"grad_norm": 0.5471459701834757,
"learning_rate": 1.921229989004062e-05,
"loss": 0.25374755859375,
"step": 960
},
{
"epoch": 0.3104,
"grad_norm": 0.5378204495307715,
"learning_rate": 1.9192015641490377e-05,
"loss": 0.25465896129608157,
"step": 970
},
{
"epoch": 0.3136,
"grad_norm": 0.4865512231368037,
"learning_rate": 1.9171484517626893e-05,
"loss": 0.24098854064941405,
"step": 980
},
{
"epoch": 0.3168,
"grad_norm": 0.6813880068730167,
"learning_rate": 1.9150707069866512e-05,
"loss": 0.2437004804611206,
"step": 990
},
{
"epoch": 0.32,
"grad_norm": 0.49793671494929714,
"learning_rate": 1.912968385624124e-05,
"loss": 0.25569858551025393,
"step": 1000
},
{
"epoch": 0.3232,
"grad_norm": 0.48200331238296507,
"learning_rate": 1.9108415441383762e-05,
"loss": 0.2474271297454834,
"step": 1010
},
{
"epoch": 0.3264,
"grad_norm": 0.5513495640732348,
"learning_rate": 1.908690239651227e-05,
"loss": 0.2525751829147339,
"step": 1020
},
{
"epoch": 0.3296,
"grad_norm": 0.5260359911423702,
"learning_rate": 1.906514529941513e-05,
"loss": 0.24476206302642822,
"step": 1030
},
{
"epoch": 0.3328,
"grad_norm": 0.5223760379669316,
"learning_rate": 1.904314473443536e-05,
"loss": 0.2423255443572998,
"step": 1040
},
{
"epoch": 0.336,
"grad_norm": 0.5324893422229405,
"learning_rate": 1.902090129245494e-05,
"loss": 0.25180060863494874,
"step": 1050
},
{
"epoch": 0.3392,
"grad_norm": 0.5533803845369737,
"learning_rate": 1.8998415570878924e-05,
"loss": 0.25927422046661375,
"step": 1060
},
{
"epoch": 0.3424,
"grad_norm": 0.5898511531326375,
"learning_rate": 1.8975688173619436e-05,
"loss": 0.2446223735809326,
"step": 1070
},
{
"epoch": 0.3456,
"grad_norm": 0.5309892938599098,
"learning_rate": 1.89527197110794e-05,
"loss": 0.249619197845459,
"step": 1080
},
{
"epoch": 0.3488,
"grad_norm": 0.5493895699411476,
"learning_rate": 1.8929510800136185e-05,
"loss": 0.25883936882019043,
"step": 1090
},
{
"epoch": 0.352,
"grad_norm": 0.5085105686504151,
"learning_rate": 1.890606206412502e-05,
"loss": 0.239257550239563,
"step": 1100
},
{
"epoch": 0.3552,
"grad_norm": 0.6140153059040666,
"learning_rate": 1.8882374132822247e-05,
"loss": 0.24560930728912353,
"step": 1110
},
{
"epoch": 0.3584,
"grad_norm": 0.5800869633821086,
"learning_rate": 1.8858447642428426e-05,
"loss": 0.25429868698120117,
"step": 1120
},
{
"epoch": 0.3616,
"grad_norm": 0.49224263274214997,
"learning_rate": 1.8834283235551237e-05,
"loss": 0.24312000274658202,
"step": 1130
},
{
"epoch": 0.3648,
"grad_norm": 0.544606497577084,
"learning_rate": 1.880988156118821e-05,
"loss": 0.24494631290435792,
"step": 1140
},
{
"epoch": 0.368,
"grad_norm": 0.5305635975075,
"learning_rate": 1.878524327470932e-05,
"loss": 0.2393354892730713,
"step": 1150
},
{
"epoch": 0.3712,
"grad_norm": 0.5272956424247851,
"learning_rate": 1.876036903783936e-05,
"loss": 0.24124324321746826,
"step": 1160
},
{
"epoch": 0.3744,
"grad_norm": 0.5299817729233058,
"learning_rate": 1.8735259518640186e-05,
"loss": 0.23604679107666016,
"step": 1170
},
{
"epoch": 0.3776,
"grad_norm": 0.49447109755210217,
"learning_rate": 1.870991539149277e-05,
"loss": 0.2502137660980225,
"step": 1180
},
{
"epoch": 0.3808,
"grad_norm": 0.48041771840318376,
"learning_rate": 1.868433733707908e-05,
"loss": 0.25159125328063964,
"step": 1190
},
{
"epoch": 0.384,
"grad_norm": 0.5126220398880542,
"learning_rate": 1.8658526042363813e-05,
"loss": 0.2372525930404663,
"step": 1200
},
{
"epoch": 0.3872,
"grad_norm": 0.5708502563559649,
"learning_rate": 1.8632482200575924e-05,
"loss": 0.2469635248184204,
"step": 1210
},
{
"epoch": 0.3904,
"grad_norm": 0.5393640054446197,
"learning_rate": 1.8606206511190033e-05,
"loss": 0.23446245193481446,
"step": 1220
},
{
"epoch": 0.3936,
"grad_norm": 0.4851190014030714,
"learning_rate": 1.8579699679907622e-05,
"loss": 0.238480544090271,
"step": 1230
},
{
"epoch": 0.3968,
"grad_norm": 0.5002760709833548,
"learning_rate": 1.855296241863808e-05,
"loss": 0.23645896911621095,
"step": 1240
},
{
"epoch": 0.4,
"grad_norm": 0.5188378170166971,
"learning_rate": 1.8525995445479596e-05,
"loss": 0.24882338047027588,
"step": 1250
},
{
"epoch": 0.4032,
"grad_norm": 0.5260536100487226,
"learning_rate": 1.8498799484699856e-05,
"loss": 0.25458498001098634,
"step": 1260
},
{
"epoch": 0.4064,
"grad_norm": 0.4925259973954508,
"learning_rate": 1.847137526671661e-05,
"loss": 0.2500078916549683,
"step": 1270
},
{
"epoch": 0.4096,
"grad_norm": 0.45653168601168276,
"learning_rate": 1.8443723528078036e-05,
"loss": 0.2439669132232666,
"step": 1280
},
{
"epoch": 0.4128,
"grad_norm": 0.5150530641408225,
"learning_rate": 1.8415845011442972e-05,
"loss": 0.241209077835083,
"step": 1290
},
{
"epoch": 0.416,
"grad_norm": 0.4651357986611782,
"learning_rate": 1.8387740465560963e-05,
"loss": 0.24701554775238038,
"step": 1300
},
{
"epoch": 0.4192,
"grad_norm": 0.5294681328173586,
"learning_rate": 1.8359410645252152e-05,
"loss": 0.24441919326782227,
"step": 1310
},
{
"epoch": 0.4224,
"grad_norm": 0.5092407050516744,
"learning_rate": 1.8330856311387008e-05,
"loss": 0.24915692806243897,
"step": 1320
},
{
"epoch": 0.4256,
"grad_norm": 0.4529623785928456,
"learning_rate": 1.8302078230865893e-05,
"loss": 0.24472570419311523,
"step": 1330
},
{
"epoch": 0.4288,
"grad_norm": 0.45190707315055384,
"learning_rate": 1.8273077176598465e-05,
"loss": 0.2396528959274292,
"step": 1340
},
{
"epoch": 0.432,
"grad_norm": 0.4918132971853266,
"learning_rate": 1.8243853927482915e-05,
"loss": 0.24474334716796875,
"step": 1350
},
{
"epoch": 0.4352,
"grad_norm": 0.5131263451430461,
"learning_rate": 1.8214409268385047e-05,
"loss": 0.23907217979431153,
"step": 1360
},
{
"epoch": 0.4384,
"grad_norm": 0.5119258458876889,
"learning_rate": 1.818474399011721e-05,
"loss": 0.2347952127456665,
"step": 1370
},
{
"epoch": 0.4416,
"grad_norm": 0.5107865849291362,
"learning_rate": 1.815485888941704e-05,
"loss": 0.2456207752227783,
"step": 1380
},
{
"epoch": 0.4448,
"grad_norm": 0.4909950247265926,
"learning_rate": 1.8124754768926078e-05,
"loss": 0.23840250968933105,
"step": 1390
},
{
"epoch": 0.448,
"grad_norm": 0.5148126576854635,
"learning_rate": 1.809443243716821e-05,
"loss": 0.22835490703582764,
"step": 1400
},
{
"epoch": 0.4512,
"grad_norm": 0.5339124403003681,
"learning_rate": 1.806389270852795e-05,
"loss": 0.2322523593902588,
"step": 1410
},
{
"epoch": 0.4544,
"grad_norm": 0.49451019785851624,
"learning_rate": 1.803313640322855e-05,
"loss": 0.2312631130218506,
"step": 1420
},
{
"epoch": 0.4576,
"grad_norm": 0.48742564627319424,
"learning_rate": 1.800216434731001e-05,
"loss": 0.23309907913208008,
"step": 1430
},
{
"epoch": 0.4608,
"grad_norm": 0.47538763051327393,
"learning_rate": 1.7970977372606854e-05,
"loss": 0.24499197006225587,
"step": 1440
},
{
"epoch": 0.464,
"grad_norm": 0.518466349726495,
"learning_rate": 1.7939576316725817e-05,
"loss": 0.23887057304382325,
"step": 1450
},
{
"epoch": 0.4672,
"grad_norm": 0.646984813075625,
"learning_rate": 1.7907962023023323e-05,
"loss": 0.2410289764404297,
"step": 1460
},
{
"epoch": 0.4704,
"grad_norm": 0.4692201078656916,
"learning_rate": 1.787613534058286e-05,
"loss": 0.24536819458007814,
"step": 1470
},
{
"epoch": 0.4736,
"grad_norm": 0.46717482673320465,
"learning_rate": 1.7844097124192153e-05,
"loss": 0.238448429107666,
"step": 1480
},
{
"epoch": 0.4768,
"grad_norm": 0.4945525229623021,
"learning_rate": 1.7811848234320232e-05,
"loss": 0.24734117984771728,
"step": 1490
},
{
"epoch": 0.48,
"grad_norm": 0.5129873631143006,
"learning_rate": 1.7779389537094296e-05,
"loss": 0.22828893661499022,
"step": 1500
},
{
"epoch": 0.4832,
"grad_norm": 0.5578512462612871,
"learning_rate": 1.7746721904276462e-05,
"loss": 0.22841358184814453,
"step": 1510
},
{
"epoch": 0.4864,
"grad_norm": 0.4833011366287632,
"learning_rate": 1.7713846213240357e-05,
"loss": 0.24015889167785645,
"step": 1520
},
{
"epoch": 0.4896,
"grad_norm": 0.46535358913511693,
"learning_rate": 1.768076334694755e-05,
"loss": 0.23306748867034913,
"step": 1530
},
{
"epoch": 0.4928,
"grad_norm": 0.4632550256884276,
"learning_rate": 1.764747419392383e-05,
"loss": 0.227636981010437,
"step": 1540
},
{
"epoch": 0.496,
"grad_norm": 0.4747806511979542,
"learning_rate": 1.761397964823535e-05,
"loss": 0.23008160591125487,
"step": 1550
},
{
"epoch": 0.4992,
"grad_norm": 0.46006860023135016,
"learning_rate": 1.758028060946462e-05,
"loss": 0.23051168918609619,
"step": 1560
},
{
"epoch": 0.5024,
"grad_norm": 0.5297844615664509,
"learning_rate": 1.754637798268633e-05,
"loss": 0.23695905208587648,
"step": 1570
},
{
"epoch": 0.5056,
"grad_norm": 0.450046278899454,
"learning_rate": 1.7512272678443062e-05,
"loss": 0.2251911163330078,
"step": 1580
},
{
"epoch": 0.5088,
"grad_norm": 0.48552476645482967,
"learning_rate": 1.747796561272081e-05,
"loss": 0.2302877426147461,
"step": 1590
},
{
"epoch": 0.512,
"grad_norm": 0.4574189912662348,
"learning_rate": 1.7443457706924413e-05,
"loss": 0.2232276439666748,
"step": 1600
},
{
"epoch": 0.5152,
"grad_norm": 0.5241821726532029,
"learning_rate": 1.740874988785277e-05,
"loss": 0.23505861759185792,
"step": 1610
},
{
"epoch": 0.5184,
"grad_norm": 0.4651850408646502,
"learning_rate": 1.7373843087673986e-05,
"loss": 0.23003506660461426,
"step": 1620
},
{
"epoch": 0.5216,
"grad_norm": 0.45785716968617546,
"learning_rate": 1.7338738243900296e-05,
"loss": 0.22193725109100343,
"step": 1630
},
{
"epoch": 0.5248,
"grad_norm": 0.5052163496935622,
"learning_rate": 1.730343629936292e-05,
"loss": 0.22255406379699708,
"step": 1640
},
{
"epoch": 0.528,
"grad_norm": 0.5107615530446142,
"learning_rate": 1.7267938202186732e-05,
"loss": 0.23032815456390382,
"step": 1650
},
{
"epoch": 0.5312,
"grad_norm": 0.4515860293987711,
"learning_rate": 1.723224490576478e-05,
"loss": 0.23593826293945314,
"step": 1660
},
{
"epoch": 0.5344,
"grad_norm": 0.4809653855693939,
"learning_rate": 1.71963573687327e-05,
"loss": 0.23353569507598876,
"step": 1670
},
{
"epoch": 0.5376,
"grad_norm": 0.4080248483710274,
"learning_rate": 1.7160276554942953e-05,
"loss": 0.23344860076904297,
"step": 1680
},
{
"epoch": 0.5408,
"grad_norm": 0.4491334154824311,
"learning_rate": 1.7124003433438962e-05,
"loss": 0.22191376686096193,
"step": 1690
},
{
"epoch": 0.544,
"grad_norm": 0.45129704270341864,
"learning_rate": 1.7087538978429048e-05,
"loss": 0.22902898788452147,
"step": 1700
},
{
"epoch": 0.5472,
"grad_norm": 0.5177167279818674,
"learning_rate": 1.7050884169260306e-05,
"loss": 0.2173527717590332,
"step": 1710
},
{
"epoch": 0.5504,
"grad_norm": 0.4508850270616855,
"learning_rate": 1.7014039990392274e-05,
"loss": 0.23019766807556152,
"step": 1720
},
{
"epoch": 0.5536,
"grad_norm": 0.46240400356243727,
"learning_rate": 1.6977007431370508e-05,
"loss": 0.23034143447875977,
"step": 1730
},
{
"epoch": 0.5568,
"grad_norm": 0.4733906447507717,
"learning_rate": 1.6939787486799994e-05,
"loss": 0.22877271175384523,
"step": 1740
},
{
"epoch": 0.56,
"grad_norm": 0.5127943588620476,
"learning_rate": 1.690238115631844e-05,
"loss": 0.22471845149993896,
"step": 1750
},
{
"epoch": 0.5632,
"grad_norm": 0.47017162211903196,
"learning_rate": 1.686478944456944e-05,
"loss": 0.23125965595245362,
"step": 1760
},
{
"epoch": 0.5664,
"grad_norm": 0.43356217544567177,
"learning_rate": 1.682701336117546e-05,
"loss": 0.23335847854614258,
"step": 1770
},
{
"epoch": 0.5696,
"grad_norm": 0.46976912912472235,
"learning_rate": 1.678905392071077e-05,
"loss": 0.22588520050048827,
"step": 1780
},
{
"epoch": 0.5728,
"grad_norm": 0.46937606827798445,
"learning_rate": 1.6750912142674137e-05,
"loss": 0.22791750431060792,
"step": 1790
},
{
"epoch": 0.576,
"grad_norm": 0.5014267089147832,
"learning_rate": 1.6712589051461507e-05,
"loss": 0.22334072589874268,
"step": 1800
},
{
"epoch": 0.5792,
"grad_norm": 0.4646105664950186,
"learning_rate": 1.6674085676338423e-05,
"loss": 0.23307304382324218,
"step": 1810
},
{
"epoch": 0.5824,
"grad_norm": 0.4693051052799574,
"learning_rate": 1.6635403051412452e-05,
"loss": 0.228350830078125,
"step": 1820
},
{
"epoch": 0.5856,
"grad_norm": 0.4837541806902605,
"learning_rate": 1.6596542215605353e-05,
"loss": 0.22374842166900635,
"step": 1830
},
{
"epoch": 0.5888,
"grad_norm": 0.46205210323151985,
"learning_rate": 1.6557504212625212e-05,
"loss": 0.23145170211791993,
"step": 1840
},
{
"epoch": 0.592,
"grad_norm": 0.47806138132269654,
"learning_rate": 1.6518290090938396e-05,
"loss": 0.23109970092773438,
"step": 1850
},
{
"epoch": 0.5952,
"grad_norm": 0.535783462031045,
"learning_rate": 1.6478900903741385e-05,
"loss": 0.23631663322448732,
"step": 1860
},
{
"epoch": 0.5984,
"grad_norm": 0.4721307006012522,
"learning_rate": 1.6439337708932507e-05,
"loss": 0.2285689353942871,
"step": 1870
},
{
"epoch": 0.6016,
"grad_norm": 0.45542793097188045,
"learning_rate": 1.6399601569083504e-05,
"loss": 0.23337695598602295,
"step": 1880
},
{
"epoch": 0.6048,
"grad_norm": 0.4312483310139632,
"learning_rate": 1.6359693551411008e-05,
"loss": 0.23037366867065429,
"step": 1890
},
{
"epoch": 0.608,
"grad_norm": 0.4416301879424995,
"learning_rate": 1.6319614727747876e-05,
"loss": 0.2173173427581787,
"step": 1900
},
{
"epoch": 0.6112,
"grad_norm": 0.4702227320661915,
"learning_rate": 1.6279366174514394e-05,
"loss": 0.2244588851928711,
"step": 1910
},
{
"epoch": 0.6144,
"grad_norm": 0.4790796626786329,
"learning_rate": 1.6238948972689383e-05,
"loss": 0.22464096546173096,
"step": 1920
},
{
"epoch": 0.6176,
"grad_norm": 0.46759559501534365,
"learning_rate": 1.6198364207781148e-05,
"loss": 0.24331374168395997,
"step": 1930
},
{
"epoch": 0.6208,
"grad_norm": 0.47419793878092914,
"learning_rate": 1.6157612969798347e-05,
"loss": 0.2228463649749756,
"step": 1940
},
{
"epoch": 0.624,
"grad_norm": 0.4879136585575402,
"learning_rate": 1.611669635322068e-05,
"loss": 0.22599225044250487,
"step": 1950
},
{
"epoch": 0.6272,
"grad_norm": 0.44396387385901565,
"learning_rate": 1.607561545696954e-05,
"loss": 0.22489728927612304,
"step": 1960
},
{
"epoch": 0.6304,
"grad_norm": 0.4605042997867878,
"learning_rate": 1.603437138437846e-05,
"loss": 0.21850183010101318,
"step": 1970
},
{
"epoch": 0.6336,
"grad_norm": 0.5137911465963019,
"learning_rate": 1.59929652431635e-05,
"loss": 0.23050074577331542,
"step": 1980
},
{
"epoch": 0.6368,
"grad_norm": 0.459549153502152,
"learning_rate": 1.5951398145393486e-05,
"loss": 0.22538766860961915,
"step": 1990
},
{
"epoch": 0.64,
"grad_norm": 0.5000960425172363,
"learning_rate": 1.5909671207460156e-05,
"loss": 0.23207569122314453,
"step": 2000
},
{
"epoch": 0.6432,
"grad_norm": 0.4666560311376051,
"learning_rate": 1.5867785550048163e-05,
"loss": 0.2227254867553711,
"step": 2010
},
{
"epoch": 0.6464,
"grad_norm": 0.44287994107528383,
"learning_rate": 1.5825742298104984e-05,
"loss": 0.22267618179321289,
"step": 2020
},
{
"epoch": 0.6496,
"grad_norm": 0.4737899603721493,
"learning_rate": 1.5783542580810702e-05,
"loss": 0.22628979682922362,
"step": 2030
},
{
"epoch": 0.6528,
"grad_norm": 0.5207143997270841,
"learning_rate": 1.5741187531547685e-05,
"loss": 0.22457518577575683,
"step": 2040
},
{
"epoch": 0.656,
"grad_norm": 0.5146030700059351,
"learning_rate": 1.5698678287870136e-05,
"loss": 0.23781847953796387,
"step": 2050
},
{
"epoch": 0.6592,
"grad_norm": 0.4581158208263549,
"learning_rate": 1.5656015991473547e-05,
"loss": 0.21890289783477784,
"step": 2060
},
{
"epoch": 0.6624,
"grad_norm": 0.4841544982383807,
"learning_rate": 1.5613201788164043e-05,
"loss": 0.21993796825408934,
"step": 2070
},
{
"epoch": 0.6656,
"grad_norm": 0.5155659177918491,
"learning_rate": 1.5570236827827597e-05,
"loss": 0.22598867416381835,
"step": 2080
},
{
"epoch": 0.6688,
"grad_norm": 0.49408946842053203,
"learning_rate": 1.5527122264399152e-05,
"loss": 0.22067124843597413,
"step": 2090
},
{
"epoch": 0.672,
"grad_norm": 0.4165960504043829,
"learning_rate": 1.5483859255831628e-05,
"loss": 0.2252934455871582,
"step": 2100
},
{
"epoch": 0.6752,
"grad_norm": 0.42105671031868686,
"learning_rate": 1.544044896406482e-05,
"loss": 0.21734967231750488,
"step": 2110
},
{
"epoch": 0.6784,
"grad_norm": 0.4617462925802655,
"learning_rate": 1.5396892554994195e-05,
"loss": 0.22457587718963623,
"step": 2120
},
{
"epoch": 0.6816,
"grad_norm": 0.49035666686617185,
"learning_rate": 1.535319119843959e-05,
"loss": 0.22814135551452636,
"step": 2130
},
{
"epoch": 0.6848,
"grad_norm": 0.43254807682337243,
"learning_rate": 1.530934606811375e-05,
"loss": 0.22064077854156494,
"step": 2140
},
{
"epoch": 0.688,
"grad_norm": 0.41346378865396444,
"learning_rate": 1.5265358341590867e-05,
"loss": 0.2305276870727539,
"step": 2150
},
{
"epoch": 0.6912,
"grad_norm": 0.4330515284614143,
"learning_rate": 1.5221229200274911e-05,
"loss": 0.2309799909591675,
"step": 2160
},
{
"epoch": 0.6944,
"grad_norm": 0.4566135764432998,
"learning_rate": 1.5176959829367907e-05,
"loss": 0.2205212116241455,
"step": 2170
},
{
"epoch": 0.6976,
"grad_norm": 0.4610907834905202,
"learning_rate": 1.5132551417838115e-05,
"loss": 0.22324662208557128,
"step": 2180
},
{
"epoch": 0.7008,
"grad_norm": 0.4454592081887688,
"learning_rate": 1.5088005158388082e-05,
"loss": 0.22123246192932128,
"step": 2190
},
{
"epoch": 0.704,
"grad_norm": 0.49546144737300474,
"learning_rate": 1.504332224742262e-05,
"loss": 0.2316317081451416,
"step": 2200
},
{
"epoch": 0.7072,
"grad_norm": 0.4143856424510617,
"learning_rate": 1.4998503885016678e-05,
"loss": 0.2110152244567871,
"step": 2210
},
{
"epoch": 0.7104,
"grad_norm": 0.4078340838171134,
"learning_rate": 1.4953551274883089e-05,
"loss": 0.21802306175231934,
"step": 2220
},
{
"epoch": 0.7136,
"grad_norm": 0.4808521283621004,
"learning_rate": 1.490846562434027e-05,
"loss": 0.22784788608551027,
"step": 2230
},
{
"epoch": 0.7168,
"grad_norm": 0.4637872678416109,
"learning_rate": 1.4863248144279769e-05,
"loss": 0.22210421562194824,
"step": 2240
},
{
"epoch": 0.72,
"grad_norm": 0.4600241908304365,
"learning_rate": 1.4817900049133761e-05,
"loss": 0.23263945579528808,
"step": 2250
},
{
"epoch": 0.7232,
"grad_norm": 0.4875002299355316,
"learning_rate": 1.4772422556842434e-05,
"loss": 0.21289660930633544,
"step": 2260
},
{
"epoch": 0.7264,
"grad_norm": 0.5278053929605496,
"learning_rate": 1.4726816888821258e-05,
"loss": 0.220017147064209,
"step": 2270
},
{
"epoch": 0.7296,
"grad_norm": 0.40621075540467777,
"learning_rate": 1.46810842699282e-05,
"loss": 0.2199113368988037,
"step": 2280
},
{
"epoch": 0.7328,
"grad_norm": 0.5000633126672952,
"learning_rate": 1.4635225928430816e-05,
"loss": 0.22381181716918946,
"step": 2290
},
{
"epoch": 0.736,
"grad_norm": 0.4429925870620894,
"learning_rate": 1.4589243095973275e-05,
"loss": 0.22666068077087403,
"step": 2300
},
{
"epoch": 0.7392,
"grad_norm": 0.5285059643667018,
"learning_rate": 1.4543137007543265e-05,
"loss": 0.22487318515777588,
"step": 2310
},
{
"epoch": 0.7424,
"grad_norm": 0.4967799932967407,
"learning_rate": 1.4496908901438833e-05,
"loss": 0.22217817306518556,
"step": 2320
},
{
"epoch": 0.7456,
"grad_norm": 0.49914503284787376,
"learning_rate": 1.4450560019235126e-05,
"loss": 0.21865057945251465,
"step": 2330
},
{
"epoch": 0.7488,
"grad_norm": 0.4507081691454913,
"learning_rate": 1.4404091605751047e-05,
"loss": 0.2127352237701416,
"step": 2340
},
{
"epoch": 0.752,
"grad_norm": 0.5208424407476361,
"learning_rate": 1.4357504909015822e-05,
"loss": 0.2290294885635376,
"step": 2350
},
{
"epoch": 0.7552,
"grad_norm": 0.47384662182954906,
"learning_rate": 1.4310801180235465e-05,
"loss": 0.22156953811645508,
"step": 2360
},
{
"epoch": 0.7584,
"grad_norm": 0.4354043725213946,
"learning_rate": 1.4263981673759211e-05,
"loss": 0.2247826099395752,
"step": 2370
},
{
"epoch": 0.7616,
"grad_norm": 0.44200353570423373,
"learning_rate": 1.4217047647045781e-05,
"loss": 0.22562670707702637,
"step": 2380
},
{
"epoch": 0.7648,
"grad_norm": 0.49530417752061345,
"learning_rate": 1.417000036062965e-05,
"loss": 0.21812112331390382,
"step": 2390
},
{
"epoch": 0.768,
"grad_norm": 0.4488865761192983,
"learning_rate": 1.4122841078087165e-05,
"loss": 0.2214284896850586,
"step": 2400
},
{
"epoch": 0.7712,
"grad_norm": 0.4321613220968368,
"learning_rate": 1.407557106600262e-05,
"loss": 0.21662631034851074,
"step": 2410
},
{
"epoch": 0.7744,
"grad_norm": 0.43914352518718497,
"learning_rate": 1.402819159393424e-05,
"loss": 0.20961742401123046,
"step": 2420
},
{
"epoch": 0.7776,
"grad_norm": 0.465936531093921,
"learning_rate": 1.3980703934380074e-05,
"loss": 0.2129192590713501,
"step": 2430
},
{
"epoch": 0.7808,
"grad_norm": 0.4882180345453942,
"learning_rate": 1.3933109362743833e-05,
"loss": 0.22881345748901366,
"step": 2440
},
{
"epoch": 0.784,
"grad_norm": 0.4210074365902586,
"learning_rate": 1.3885409157300624e-05,
"loss": 0.21948463916778566,
"step": 2450
},
{
"epoch": 0.7872,
"grad_norm": 0.43738183055017865,
"learning_rate": 1.3837604599162623e-05,
"loss": 0.21794588565826417,
"step": 2460
},
{
"epoch": 0.7904,
"grad_norm": 0.44837604530586583,
"learning_rate": 1.3789696972244666e-05,
"loss": 0.20748381614685057,
"step": 2470
},
{
"epoch": 0.7936,
"grad_norm": 0.4350573307716704,
"learning_rate": 1.3741687563229768e-05,
"loss": 0.21708590984344484,
"step": 2480
},
{
"epoch": 0.7968,
"grad_norm": 0.47042583489847606,
"learning_rate": 1.3693577661534569e-05,
"loss": 0.21381185054779053,
"step": 2490
},
{
"epoch": 0.8,
"grad_norm": 0.3885392296831242,
"learning_rate": 1.364536855927469e-05,
"loss": 0.21558475494384766,
"step": 2500
},
{
"epoch": 0.8032,
"grad_norm": 0.4713749196166673,
"learning_rate": 1.3597061551230049e-05,
"loss": 0.20907931327819823,
"step": 2510
},
{
"epoch": 0.8064,
"grad_norm": 0.4348549948415516,
"learning_rate": 1.354865793481007e-05,
"loss": 0.22130863666534423,
"step": 2520
},
{
"epoch": 0.8096,
"grad_norm": 0.4597506002475158,
"learning_rate": 1.3500159010018847e-05,
"loss": 0.22623491287231445,
"step": 2530
},
{
"epoch": 0.8128,
"grad_norm": 0.47040030900152735,
"learning_rate": 1.345156607942023e-05,
"loss": 0.22732398509979249,
"step": 2540
},
{
"epoch": 0.816,
"grad_norm": 0.4480420790659535,
"learning_rate": 1.340288044810283e-05,
"loss": 0.21516098976135253,
"step": 2550
},
{
"epoch": 0.8192,
"grad_norm": 0.4061866310554717,
"learning_rate": 1.3354103423644982e-05,
"loss": 0.20598292350769043,
"step": 2560
},
{
"epoch": 0.8224,
"grad_norm": 0.4356231506435344,
"learning_rate": 1.3305236316079615e-05,
"loss": 0.2088902473449707,
"step": 2570
},
{
"epoch": 0.8256,
"grad_norm": 0.45438035936444127,
"learning_rate": 1.3256280437859075e-05,
"loss": 0.21653916835784912,
"step": 2580
},
{
"epoch": 0.8288,
"grad_norm": 0.4224917112067228,
"learning_rate": 1.320723710381987e-05,
"loss": 0.2043593406677246,
"step": 2590
},
{
"epoch": 0.832,
"grad_norm": 0.4298149169704579,
"learning_rate": 1.3158107631147367e-05,
"loss": 0.21864793300628663,
"step": 2600
},
{
"epoch": 0.8352,
"grad_norm": 0.4730028273473233,
"learning_rate": 1.3108893339340403e-05,
"loss": 0.22418289184570311,
"step": 2610
},
{
"epoch": 0.8384,
"grad_norm": 0.410194239876692,
"learning_rate": 1.3059595550175847e-05,
"loss": 0.21612415313720704,
"step": 2620
},
{
"epoch": 0.8416,
"grad_norm": 0.41353252828391834,
"learning_rate": 1.3010215587673121e-05,
"loss": 0.20649394989013672,
"step": 2630
},
{
"epoch": 0.8448,
"grad_norm": 0.45768379185501146,
"learning_rate": 1.2960754778058603e-05,
"loss": 0.22061245441436766,
"step": 2640
},
{
"epoch": 0.848,
"grad_norm": 0.48022107881187487,
"learning_rate": 1.2911214449730044e-05,
"loss": 0.20851058959960939,
"step": 2650
},
{
"epoch": 0.8512,
"grad_norm": 0.4226594029508972,
"learning_rate": 1.2861595933220873e-05,
"loss": 0.21881065368652344,
"step": 2660
},
{
"epoch": 0.8544,
"grad_norm": 0.4759210428559727,
"learning_rate": 1.2811900561164459e-05,
"loss": 0.21680545806884766,
"step": 2670
},
{
"epoch": 0.8576,
"grad_norm": 0.44152077578188165,
"learning_rate": 1.2762129668258335e-05,
"loss": 0.21884400844573976,
"step": 2680
},
{
"epoch": 0.8608,
"grad_norm": 0.4371091168079885,
"learning_rate": 1.2712284591228326e-05,
"loss": 0.21911420822143554,
"step": 2690
},
{
"epoch": 0.864,
"grad_norm": 0.4755990552834391,
"learning_rate": 1.2662366668792681e-05,
"loss": 0.21593255996704103,
"step": 2700
},
{
"epoch": 0.8672,
"grad_norm": 0.4193225750204993,
"learning_rate": 1.2612377241626087e-05,
"loss": 0.2154374361038208,
"step": 2710
},
{
"epoch": 0.8704,
"grad_norm": 0.4657709120189377,
"learning_rate": 1.2562317652323681e-05,
"loss": 0.21304264068603515,
"step": 2720
},
{
"epoch": 0.8736,
"grad_norm": 0.40630470242327155,
"learning_rate": 1.2512189245364986e-05,
"loss": 0.21559524536132812,
"step": 2730
},
{
"epoch": 0.8768,
"grad_norm": 0.4413935692066605,
"learning_rate": 1.2461993367077795e-05,
"loss": 0.2167214870452881,
"step": 2740
},
{
"epoch": 0.88,
"grad_norm": 0.4560124290390928,
"learning_rate": 1.2411731365602027e-05,
"loss": 0.2166179895401001,
"step": 2750
},
{
"epoch": 0.8832,
"grad_norm": 0.475836593056793,
"learning_rate": 1.2361404590853503e-05,
"loss": 0.21170167922973632,
"step": 2760
},
{
"epoch": 0.8864,
"grad_norm": 0.4493282523252224,
"learning_rate": 1.2311014394487699e-05,
"loss": 0.21825058460235597,
"step": 2770
},
{
"epoch": 0.8896,
"grad_norm": 0.43608601365652616,
"learning_rate": 1.2260562129863443e-05,
"loss": 0.22325677871704103,
"step": 2780
},
{
"epoch": 0.8928,
"grad_norm": 0.4184433136907479,
"learning_rate": 1.2210049152006562e-05,
"loss": 0.20934054851531983,
"step": 2790
},
{
"epoch": 0.896,
"grad_norm": 0.45846229639126757,
"learning_rate": 1.2159476817573506e-05,
"loss": 0.19943830966949463,
"step": 2800
},
{
"epoch": 0.8992,
"grad_norm": 0.47343808673350296,
"learning_rate": 1.210884648481488e-05,
"loss": 0.2045361042022705,
"step": 2810
},
{
"epoch": 0.9024,
"grad_norm": 0.4297176879254465,
"learning_rate": 1.2058159513539005e-05,
"loss": 0.2150510787963867,
"step": 2820
},
{
"epoch": 0.9056,
"grad_norm": 0.48924986935190085,
"learning_rate": 1.2007417265075357e-05,
"loss": 0.21673121452331542,
"step": 2830
},
{
"epoch": 0.9088,
"grad_norm": 0.4118209956292729,
"learning_rate": 1.1956621102238035e-05,
"loss": 0.20586814880371093,
"step": 2840
},
{
"epoch": 0.912,
"grad_norm": 0.490941881583275,
"learning_rate": 1.1905772389289147e-05,
"loss": 0.20310707092285157,
"step": 2850
},
{
"epoch": 0.9152,
"grad_norm": 0.4418649425267541,
"learning_rate": 1.1854872491902162e-05,
"loss": 0.20629138946533204,
"step": 2860
},
{
"epoch": 0.9184,
"grad_norm": 0.4794522588070875,
"learning_rate": 1.1803922777125247e-05,
"loss": 0.21506185531616212,
"step": 2870
},
{
"epoch": 0.9216,
"grad_norm": 0.39231670686280146,
"learning_rate": 1.1752924613344546e-05,
"loss": 0.20521464347839355,
"step": 2880
},
{
"epoch": 0.9248,
"grad_norm": 0.417896792522618,
"learning_rate": 1.1701879370247408e-05,
"loss": 0.21295480728149413,
"step": 2890
},
{
"epoch": 0.928,
"grad_norm": 0.45191505684405564,
"learning_rate": 1.1650788418785646e-05,
"loss": 0.21235859394073486,
"step": 2900
},
{
"epoch": 0.9312,
"grad_norm": 0.45739995881129936,
"learning_rate": 1.1599653131138669e-05,
"loss": 0.21502935886383057,
"step": 2910
},
{
"epoch": 0.9344,
"grad_norm": 0.411222693547838,
"learning_rate": 1.1548474880676654e-05,
"loss": 0.2119450807571411,
"step": 2920
},
{
"epoch": 0.9376,
"grad_norm": 0.4089068813880911,
"learning_rate": 1.1497255041923658e-05,
"loss": 0.21014490127563476,
"step": 2930
},
{
"epoch": 0.9408,
"grad_norm": 0.4855460499959178,
"learning_rate": 1.1445994990520697e-05,
"loss": 0.20929114818572997,
"step": 2940
},
{
"epoch": 0.944,
"grad_norm": 0.4229094274962923,
"learning_rate": 1.1394696103188798e-05,
"loss": 0.2090162754058838,
"step": 2950
},
{
"epoch": 0.9472,
"grad_norm": 0.4352966302843019,
"learning_rate": 1.1343359757692034e-05,
"loss": 0.22288289070129394,
"step": 2960
},
{
"epoch": 0.9504,
"grad_norm": 0.4218717951548965,
"learning_rate": 1.1291987332800511e-05,
"loss": 0.21245884895324707,
"step": 2970
},
{
"epoch": 0.9536,
"grad_norm": 0.41563547119186744,
"learning_rate": 1.1240580208253337e-05,
"loss": 0.21266646385192872,
"step": 2980
},
{
"epoch": 0.9568,
"grad_norm": 0.4300362659664126,
"learning_rate": 1.1189139764721576e-05,
"loss": 0.2064434051513672,
"step": 2990
},
{
"epoch": 0.96,
"grad_norm": 0.4493925856421068,
"learning_rate": 1.1137667383771152e-05,
"loss": 0.21611130237579346,
"step": 3000
},
{
"epoch": 0.9632,
"grad_norm": 0.4193518477278948,
"learning_rate": 1.108616444782575e-05,
"loss": 0.21042509078979493,
"step": 3010
},
{
"epoch": 0.9664,
"grad_norm": 0.44238369674667316,
"learning_rate": 1.10346323401297e-05,
"loss": 0.2128499984741211,
"step": 3020
},
{
"epoch": 0.9696,
"grad_norm": 0.43610381809849075,
"learning_rate": 1.09830724447108e-05,
"loss": 0.2186016321182251,
"step": 3030
},
{
"epoch": 0.9728,
"grad_norm": 0.43457480324184106,
"learning_rate": 1.0931486146343168e-05,
"loss": 0.21048007011413575,
"step": 3040
},
{
"epoch": 0.976,
"grad_norm": 0.41650132073473073,
"learning_rate": 1.0879874830510043e-05,
"loss": 0.20714735984802246,
"step": 3050
},
{
"epoch": 0.9792,
"grad_norm": 0.38422318663267513,
"learning_rate": 1.082823988336656e-05,
"loss": 0.21143741607666017,
"step": 3060
},
{
"epoch": 0.9824,
"grad_norm": 0.4931446496754532,
"learning_rate": 1.0776582691702553e-05,
"loss": 0.20889267921447754,
"step": 3070
},
{
"epoch": 0.9856,
"grad_norm": 0.42413877752068785,
"learning_rate": 1.0724904642905272e-05,
"loss": 0.2180039405822754,
"step": 3080
},
{
"epoch": 0.9888,
"grad_norm": 0.4062979797254314,
"learning_rate": 1.0673207124922148e-05,
"loss": 0.2040691375732422,
"step": 3090
},
{
"epoch": 0.992,
"grad_norm": 0.41395420086674883,
"learning_rate": 1.0621491526223508e-05,
"loss": 0.22020745277404785,
"step": 3100
},
{
"epoch": 0.9952,
"grad_norm": 0.4028308707660964,
"learning_rate": 1.056975923576528e-05,
"loss": 0.21360325813293457,
"step": 3110
},
{
"epoch": 0.9984,
"grad_norm": 0.4108098723915902,
"learning_rate": 1.0518011642951694e-05,
"loss": 0.2199021577835083,
"step": 3120
},
{
"epoch": 1.0016,
"grad_norm": 0.4000157932165563,
"learning_rate": 1.0466250137597966e-05,
"loss": 0.18031301498413085,
"step": 3130
},
{
"epoch": 1.0048,
"grad_norm": 0.4609316862502506,
"learning_rate": 1.0414476109892966e-05,
"loss": 0.15385096073150634,
"step": 3140
},
{
"epoch": 1.008,
"grad_norm": 0.4234100558984651,
"learning_rate": 1.0362690950361882e-05,
"loss": 0.15555551052093505,
"step": 3150
},
{
"epoch": 1.0112,
"grad_norm": 0.4189930362335922,
"learning_rate": 1.0310896049828884e-05,
"loss": 0.1488593339920044,
"step": 3160
},
{
"epoch": 1.0144,
"grad_norm": 0.4820698446456614,
"learning_rate": 1.0259092799379756e-05,
"loss": 0.146068012714386,
"step": 3170
},
{
"epoch": 1.0176,
"grad_norm": 0.5034373010726465,
"learning_rate": 1.0207282590324534e-05,
"loss": 0.15055253505706787,
"step": 3180
},
{
"epoch": 1.0208,
"grad_norm": 0.4526022388010309,
"learning_rate": 1.0155466814160163e-05,
"loss": 0.15590319633483887,
"step": 3190
},
{
"epoch": 1.024,
"grad_norm": 0.4567093017038738,
"learning_rate": 1.0103646862533091e-05,
"loss": 0.15103107690811157,
"step": 3200
},
{
"epoch": 1.0272,
"grad_norm": 0.43181928780771217,
"learning_rate": 1.0051824127201919e-05,
"loss": 0.1524778723716736,
"step": 3210
},
{
"epoch": 1.0304,
"grad_norm": 0.43405623588227665,
"learning_rate": 1e-05,
"loss": 0.14858064651489258,
"step": 3220
},
{
"epoch": 1.0336,
"grad_norm": 0.43623088323890313,
"learning_rate": 9.948175872798084e-06,
"loss": 0.14743958711624144,
"step": 3230
},
{
"epoch": 1.0368,
"grad_norm": 0.42714630781910634,
"learning_rate": 9.896353137466912e-06,
"loss": 0.15115349292755126,
"step": 3240
},
{
"epoch": 1.04,
"grad_norm": 0.42473683250520694,
"learning_rate": 9.844533185839839e-06,
"loss": 0.15899989604949952,
"step": 3250
},
{
"epoch": 1.0432,
"grad_norm": 0.4352262906073268,
"learning_rate": 9.792717409675467e-06,
"loss": 0.14718787670135497,
"step": 3260
},
{
"epoch": 1.0464,
"grad_norm": 0.41878404412985504,
"learning_rate": 9.740907200620251e-06,
"loss": 0.14776992797851562,
"step": 3270
},
{
"epoch": 1.0496,
"grad_norm": 0.4064186962582049,
"learning_rate": 9.68910395017112e-06,
"loss": 0.14597526788711548,
"step": 3280
},
{
"epoch": 1.0528,
"grad_norm": 0.41225988218913745,
"learning_rate": 9.637309049638121e-06,
"loss": 0.15010814666748046,
"step": 3290
},
{
"epoch": 1.056,
"grad_norm": 0.4268549040770501,
"learning_rate": 9.585523890107035e-06,
"loss": 0.15626006126403807,
"step": 3300
},
{
"epoch": 1.0592,
"grad_norm": 0.4970359514206216,
"learning_rate": 9.533749862402034e-06,
"loss": 0.15661698579788208,
"step": 3310
},
{
"epoch": 1.0624,
"grad_norm": 0.4119567820390883,
"learning_rate": 9.481988357048306e-06,
"loss": 0.15095505714416504,
"step": 3320
},
{
"epoch": 1.0656,
"grad_norm": 0.45980487839863327,
"learning_rate": 9.430240764234722e-06,
"loss": 0.1474763870239258,
"step": 3330
},
{
"epoch": 1.0688,
"grad_norm": 0.3940767816510516,
"learning_rate": 9.378508473776494e-06,
"loss": 0.1548082113265991,
"step": 3340
},
{
"epoch": 1.072,
"grad_norm": 0.45177749754027824,
"learning_rate": 9.326792875077856e-06,
"loss": 0.13937206268310548,
"step": 3350
},
{
"epoch": 1.0752,
"grad_norm": 0.46269817785896816,
"learning_rate": 9.275095357094732e-06,
"loss": 0.14653103351593016,
"step": 3360
},
{
"epoch": 1.0784,
"grad_norm": 0.44710672071083446,
"learning_rate": 9.22341730829745e-06,
"loss": 0.1488661766052246,
"step": 3370
},
{
"epoch": 1.0816,
"grad_norm": 0.4060364188373684,
"learning_rate": 9.171760116633443e-06,
"loss": 0.14752568006515504,
"step": 3380
},
{
"epoch": 1.0848,
"grad_norm": 0.40164458611346504,
"learning_rate": 9.120125169489962e-06,
"loss": 0.14859724044799805,
"step": 3390
},
{
"epoch": 1.088,
"grad_norm": 0.4500638821173802,
"learning_rate": 9.068513853656833e-06,
"loss": 0.1452120065689087,
"step": 3400
},
{
"epoch": 1.0912,
"grad_norm": 0.4752962319624515,
"learning_rate": 9.016927555289205e-06,
"loss": 0.1434119701385498,
"step": 3410
},
{
"epoch": 1.0944,
"grad_norm": 0.4576015166651795,
"learning_rate": 8.965367659870305e-06,
"loss": 0.14795585870742797,
"step": 3420
},
{
"epoch": 1.0976,
"grad_norm": 0.4201692095047699,
"learning_rate": 8.913835552174251e-06,
"loss": 0.15459375381469725,
"step": 3430
},
{
"epoch": 1.1008,
"grad_norm": 0.4721066946951407,
"learning_rate": 8.862332616228853e-06,
"loss": 0.1551951289176941,
"step": 3440
},
{
"epoch": 1.104,
"grad_norm": 0.4853394677188302,
"learning_rate": 8.810860235278424e-06,
"loss": 0.14358808994293212,
"step": 3450
},
{
"epoch": 1.1072,
"grad_norm": 0.4333918422110321,
"learning_rate": 8.759419791746663e-06,
"loss": 0.14327778816223144,
"step": 3460
},
{
"epoch": 1.1104,
"grad_norm": 0.41119486357107543,
"learning_rate": 8.708012667199489e-06,
"loss": 0.1394709825515747,
"step": 3470
},
{
"epoch": 1.1136,
"grad_norm": 0.42944759689740497,
"learning_rate": 8.656640242307965e-06,
"loss": 0.14728530645370483,
"step": 3480
},
{
"epoch": 1.1168,
"grad_norm": 0.4485656444362996,
"learning_rate": 8.605303896811205e-06,
"loss": 0.1442720413208008,
"step": 3490
},
{
"epoch": 1.12,
"grad_norm": 0.5498391911152251,
"learning_rate": 8.554005009479307e-06,
"loss": 0.1475757837295532,
"step": 3500
},
{
"epoch": 1.1232,
"grad_norm": 0.4670442104926288,
"learning_rate": 8.502744958076344e-06,
"loss": 0.1499885082244873,
"step": 3510
},
{
"epoch": 1.1264,
"grad_norm": 0.45281368068303746,
"learning_rate": 8.45152511932335e-06,
"loss": 0.1545062780380249,
"step": 3520
},
{
"epoch": 1.1296,
"grad_norm": 0.4371890853989195,
"learning_rate": 8.400346868861334e-06,
"loss": 0.14784979820251465,
"step": 3530
},
{
"epoch": 1.1328,
"grad_norm": 0.46523685183268504,
"learning_rate": 8.349211581214355e-06,
"loss": 0.1493913412094116,
"step": 3540
},
{
"epoch": 1.1360000000000001,
"grad_norm": 0.44271226316314916,
"learning_rate": 8.298120629752595e-06,
"loss": 0.14975399971008302,
"step": 3550
},
{
"epoch": 1.1392,
"grad_norm": 0.39885329135000175,
"learning_rate": 8.24707538665546e-06,
"loss": 0.13927403688430787,
"step": 3560
},
{
"epoch": 1.1424,
"grad_norm": 0.4348204720948393,
"learning_rate": 8.196077222874755e-06,
"loss": 0.15003037452697754,
"step": 3570
},
{
"epoch": 1.1456,
"grad_norm": 0.36492369609904335,
"learning_rate": 8.145127508097843e-06,
"loss": 0.1406249523162842,
"step": 3580
},
{
"epoch": 1.1488,
"grad_norm": 0.4350260787673257,
"learning_rate": 8.094227610710857e-06,
"loss": 0.1464400053024292,
"step": 3590
},
{
"epoch": 1.152,
"grad_norm": 0.46135233459988695,
"learning_rate": 8.043378897761965e-06,
"loss": 0.13928555250167846,
"step": 3600
},
{
"epoch": 1.1552,
"grad_norm": 0.47991479247774826,
"learning_rate": 7.992582734924643e-06,
"loss": 0.1482172727584839,
"step": 3610
},
{
"epoch": 1.1584,
"grad_norm": 0.44791488211850955,
"learning_rate": 7.941840486460998e-06,
"loss": 0.14748493432998658,
"step": 3620
},
{
"epoch": 1.1616,
"grad_norm": 0.4566052512491053,
"learning_rate": 7.891153515185122e-06,
"loss": 0.15098787546157838,
"step": 3630
},
{
"epoch": 1.1648,
"grad_norm": 0.43671442763198143,
"learning_rate": 7.840523182426497e-06,
"loss": 0.14221494197845458,
"step": 3640
},
{
"epoch": 1.168,
"grad_norm": 0.4788117568041109,
"learning_rate": 7.78995084799344e-06,
"loss": 0.14516299962997437,
"step": 3650
},
{
"epoch": 1.1712,
"grad_norm": 0.4112340910971337,
"learning_rate": 7.73943787013656e-06,
"loss": 0.14898358583450316,
"step": 3660
},
{
"epoch": 1.1743999999999999,
"grad_norm": 0.4242798737576548,
"learning_rate": 7.688985605512303e-06,
"loss": 0.14956285953521728,
"step": 3670
},
{
"epoch": 1.1776,
"grad_norm": 0.4506364145963579,
"learning_rate": 7.6385954091465e-06,
"loss": 0.14612977504730223,
"step": 3680
},
{
"epoch": 1.1808,
"grad_norm": 0.42745313266849205,
"learning_rate": 7.588268634397975e-06,
"loss": 0.1503207802772522,
"step": 3690
},
{
"epoch": 1.184,
"grad_norm": 0.43042522162824437,
"learning_rate": 7.538006632922207e-06,
"loss": 0.1493225336074829,
"step": 3700
},
{
"epoch": 1.1872,
"grad_norm": 0.46683069899544616,
"learning_rate": 7.4878107546350184e-06,
"loss": 0.14673912525177002,
"step": 3710
},
{
"epoch": 1.1904,
"grad_norm": 0.4523609168534499,
"learning_rate": 7.437682347676322e-06,
"loss": 0.14425115585327147,
"step": 3720
},
{
"epoch": 1.1936,
"grad_norm": 0.468589792542915,
"learning_rate": 7.387622758373918e-06,
"loss": 0.14940693378448486,
"step": 3730
},
{
"epoch": 1.1968,
"grad_norm": 0.45409829064458057,
"learning_rate": 7.3376333312073254e-06,
"loss": 0.15014851093292236,
"step": 3740
},
{
"epoch": 1.2,
"grad_norm": 0.4456230990472778,
"learning_rate": 7.287715408771675e-06,
"loss": 0.14813472032546998,
"step": 3750
},
{
"epoch": 1.2032,
"grad_norm": 0.43022014288157456,
"learning_rate": 7.2378703317416695e-06,
"loss": 0.14652148485183716,
"step": 3760
},
{
"epoch": 1.2064,
"grad_norm": 0.4307313551897378,
"learning_rate": 7.188099438835543e-06,
"loss": 0.144598126411438,
"step": 3770
},
{
"epoch": 1.2096,
"grad_norm": 0.4559627092125971,
"learning_rate": 7.13840406677913e-06,
"loss": 0.14601032733917235,
"step": 3780
},
{
"epoch": 1.2128,
"grad_norm": 0.5001634717378843,
"learning_rate": 7.08878555026996e-06,
"loss": 0.15028810501098633,
"step": 3790
},
{
"epoch": 1.216,
"grad_norm": 0.4297286825932178,
"learning_rate": 7.039245221941402e-06,
"loss": 0.14278573989868165,
"step": 3800
},
{
"epoch": 1.2192,
"grad_norm": 0.5171534935896881,
"learning_rate": 6.989784412326883e-06,
"loss": 0.13963335752487183,
"step": 3810
},
{
"epoch": 1.2224,
"grad_norm": 0.4408362657985811,
"learning_rate": 6.9404044498241554e-06,
"loss": 0.14114012718200683,
"step": 3820
},
{
"epoch": 1.2256,
"grad_norm": 0.40807401968745716,
"learning_rate": 6.8911066606596015e-06,
"loss": 0.15458221435546876,
"step": 3830
},
{
"epoch": 1.2288000000000001,
"grad_norm": 0.48461114389102444,
"learning_rate": 6.841892368852634e-06,
"loss": 0.14744759798049928,
"step": 3840
},
{
"epoch": 1.232,
"grad_norm": 0.47057712254142486,
"learning_rate": 6.792762896180133e-06,
"loss": 0.1442049741744995,
"step": 3850
},
{
"epoch": 1.2352,
"grad_norm": 0.4164332655845991,
"learning_rate": 6.743719562140929e-06,
"loss": 0.14831459522247314,
"step": 3860
},
{
"epoch": 1.2384,
"grad_norm": 0.41101157253125653,
"learning_rate": 6.694763683920389e-06,
"loss": 0.14746562242507935,
"step": 3870
},
{
"epoch": 1.2416,
"grad_norm": 0.4547876403447785,
"learning_rate": 6.645896576355023e-06,
"loss": 0.14349616765975953,
"step": 3880
},
{
"epoch": 1.2448,
"grad_norm": 0.40718181703023704,
"learning_rate": 6.597119551897173e-06,
"loss": 0.14700911045074463,
"step": 3890
},
{
"epoch": 1.248,
"grad_norm": 0.4716336816544168,
"learning_rate": 6.548433920579772e-06,
"loss": 0.14113694429397583,
"step": 3900
},
{
"epoch": 1.2511999999999999,
"grad_norm": 0.4099009630185907,
"learning_rate": 6.499840989981153e-06,
"loss": 0.13905718326568603,
"step": 3910
},
{
"epoch": 1.2544,
"grad_norm": 0.48376988970660606,
"learning_rate": 6.451342065189932e-06,
"loss": 0.1491898775100708,
"step": 3920
},
{
"epoch": 1.2576,
"grad_norm": 0.44840084401999347,
"learning_rate": 6.4029384487699545e-06,
"loss": 0.1401118516921997,
"step": 3930
},
{
"epoch": 1.2608,
"grad_norm": 0.43484064180320947,
"learning_rate": 6.354631440725312e-06,
"loss": 0.1413504958152771,
"step": 3940
},
{
"epoch": 1.264,
"grad_norm": 0.41172481053076787,
"learning_rate": 6.306422338465433e-06,
"loss": 0.14377684593200685,
"step": 3950
},
{
"epoch": 1.2671999999999999,
"grad_norm": 0.459122324768509,
"learning_rate": 6.258312436770233e-06,
"loss": 0.14535202980041503,
"step": 3960
},
{
"epoch": 1.2704,
"grad_norm": 0.4731697150942664,
"learning_rate": 6.210303027755335e-06,
"loss": 0.15567996501922607,
"step": 3970
},
{
"epoch": 1.2736,
"grad_norm": 0.4092287659589343,
"learning_rate": 6.162395400837378e-06,
"loss": 0.14346708059310914,
"step": 3980
},
{
"epoch": 1.2768,
"grad_norm": 0.465320180567289,
"learning_rate": 6.114590842699379e-06,
"loss": 0.1483353853225708,
"step": 3990
},
{
"epoch": 1.28,
"grad_norm": 0.4281616701033715,
"learning_rate": 6.06689063725617e-06,
"loss": 0.1407594084739685,
"step": 4000
},
{
"epoch": 1.2832,
"grad_norm": 0.4242588043461909,
"learning_rate": 6.019296065619929e-06,
"loss": 0.14471921920776368,
"step": 4010
},
{
"epoch": 1.2864,
"grad_norm": 0.4669753246520701,
"learning_rate": 5.9718084060657665e-06,
"loss": 0.14049596786499025,
"step": 4020
},
{
"epoch": 1.2896,
"grad_norm": 0.4620096589358672,
"learning_rate": 5.924428933997385e-06,
"loss": 0.1442751407623291,
"step": 4030
},
{
"epoch": 1.2928,
"grad_norm": 0.44289430877012187,
"learning_rate": 5.87715892191284e-06,
"loss": 0.14725582599639891,
"step": 4040
},
{
"epoch": 1.296,
"grad_norm": 0.40474534048333993,
"learning_rate": 5.829999639370352e-06,
"loss": 0.13810265064239502,
"step": 4050
},
{
"epoch": 1.2992,
"grad_norm": 0.4665294490766403,
"learning_rate": 5.7829523529542185e-06,
"loss": 0.144439172744751,
"step": 4060
},
{
"epoch": 1.3024,
"grad_norm": 0.4746701265533749,
"learning_rate": 5.736018326240791e-06,
"loss": 0.14861757755279542,
"step": 4070
},
{
"epoch": 1.3056,
"grad_norm": 0.46542565257152224,
"learning_rate": 5.689198819764538e-06,
"loss": 0.14305856227874755,
"step": 4080
},
{
"epoch": 1.3088,
"grad_norm": 0.49016318971378436,
"learning_rate": 5.642495090984183e-06,
"loss": 0.1409945845603943,
"step": 4090
},
{
"epoch": 1.312,
"grad_norm": 0.4075307294315194,
"learning_rate": 5.595908394248955e-06,
"loss": 0.14292657375335693,
"step": 4100
},
{
"epoch": 1.3152,
"grad_norm": 0.437839616444408,
"learning_rate": 5.549439980764876e-06,
"loss": 0.14641907215118408,
"step": 4110
},
{
"epoch": 1.3184,
"grad_norm": 0.4253229105901944,
"learning_rate": 5.50309109856117e-06,
"loss": 0.14037476778030394,
"step": 4120
},
{
"epoch": 1.3216,
"grad_norm": 0.5383669090167431,
"learning_rate": 5.456862992456738e-06,
"loss": 0.13926520347595214,
"step": 4130
},
{
"epoch": 1.3248,
"grad_norm": 0.4425109632384422,
"learning_rate": 5.41075690402673e-06,
"loss": 0.14616999626159669,
"step": 4140
},
{
"epoch": 1.328,
"grad_norm": 0.45702874353732925,
"learning_rate": 5.364774071569188e-06,
"loss": 0.14451576471328736,
"step": 4150
},
{
"epoch": 1.3312,
"grad_norm": 0.44287340111471485,
"learning_rate": 5.318915730071806e-06,
"loss": 0.1453162908554077,
"step": 4160
},
{
"epoch": 1.3344,
"grad_norm": 0.4103259694310813,
"learning_rate": 5.273183111178746e-06,
"loss": 0.1465613603591919,
"step": 4170
},
{
"epoch": 1.3376000000000001,
"grad_norm": 0.42517411011244094,
"learning_rate": 5.227577443157568e-06,
"loss": 0.14057021141052245,
"step": 4180
},
{
"epoch": 1.3408,
"grad_norm": 0.4375166843607463,
"learning_rate": 5.182099950866239e-06,
"loss": 0.15020095109939574,
"step": 4190
},
{
"epoch": 1.3439999999999999,
"grad_norm": 0.416291462805907,
"learning_rate": 5.136751855720235e-06,
"loss": 0.13850533962249756,
"step": 4200
},
{
"epoch": 1.3472,
"grad_norm": 0.47070203521728027,
"learning_rate": 5.091534375659732e-06,
"loss": 0.1416234016418457,
"step": 4210
},
{
"epoch": 1.3504,
"grad_norm": 0.4889172059496099,
"learning_rate": 5.046448725116911e-06,
"loss": 0.14958680868148805,
"step": 4220
},
{
"epoch": 1.3536000000000001,
"grad_norm": 0.4519004948149354,
"learning_rate": 5.001496114983324e-06,
"loss": 0.1389833450317383,
"step": 4230
},
{
"epoch": 1.3568,
"grad_norm": 0.4574524631418956,
"learning_rate": 4.95667775257738e-06,
"loss": 0.14382555484771728,
"step": 4240
},
{
"epoch": 1.3599999999999999,
"grad_norm": 0.4342184251058471,
"learning_rate": 4.911994841611924e-06,
"loss": 0.14533275365829468,
"step": 4250
},
{
"epoch": 1.3632,
"grad_norm": 0.45747364937481483,
"learning_rate": 4.86744858216189e-06,
"loss": 0.14123116731643676,
"step": 4260
},
{
"epoch": 1.3664,
"grad_norm": 0.4514856726756779,
"learning_rate": 4.823040170632095e-06,
"loss": 0.1405439257621765,
"step": 4270
},
{
"epoch": 1.3696,
"grad_norm": 0.4132222466516193,
"learning_rate": 4.778770799725091e-06,
"loss": 0.14514267444610596,
"step": 4280
},
{
"epoch": 1.3728,
"grad_norm": 0.43302874997771584,
"learning_rate": 4.7346416584091335e-06,
"loss": 0.14771273136138915,
"step": 4290
},
{
"epoch": 1.376,
"grad_norm": 0.42717123197649837,
"learning_rate": 4.690653931886252e-06,
"loss": 0.13940255641937255,
"step": 4300
},
{
"epoch": 1.3792,
"grad_norm": 0.4723842337728928,
"learning_rate": 4.646808801560419e-06,
"loss": 0.14405333995819092,
"step": 4310
},
{
"epoch": 1.3824,
"grad_norm": 0.47729687745290117,
"learning_rate": 4.603107445005806e-06,
"loss": 0.14326223134994506,
"step": 4320
},
{
"epoch": 1.3856,
"grad_norm": 0.4473211198460003,
"learning_rate": 4.559551035935184e-06,
"loss": 0.1440434455871582,
"step": 4330
},
{
"epoch": 1.3888,
"grad_norm": 0.46275108214351707,
"learning_rate": 4.516140744168376e-06,
"loss": 0.13933472633361815,
"step": 4340
},
{
"epoch": 1.392,
"grad_norm": 0.4741659716320151,
"learning_rate": 4.4728777356008464e-06,
"loss": 0.14344491958618164,
"step": 4350
},
{
"epoch": 1.3952,
"grad_norm": 0.42077756412222833,
"learning_rate": 4.429763172172405e-06,
"loss": 0.1480068564414978,
"step": 4360
},
{
"epoch": 1.3984,
"grad_norm": 0.41018320173982264,
"learning_rate": 4.38679821183596e-06,
"loss": 0.14648473262786865,
"step": 4370
},
{
"epoch": 1.4016,
"grad_norm": 0.4790430737149808,
"learning_rate": 4.343984008526456e-06,
"loss": 0.14415454864501953,
"step": 4380
},
{
"epoch": 1.4048,
"grad_norm": 0.4728038469264328,
"learning_rate": 4.301321712129868e-06,
"loss": 0.141951322555542,
"step": 4390
},
{
"epoch": 1.408,
"grad_norm": 0.47426297713696836,
"learning_rate": 4.258812468452317e-06,
"loss": 0.14557615518569947,
"step": 4400
},
{
"epoch": 1.4112,
"grad_norm": 0.46607917091935247,
"learning_rate": 4.216457419189298e-06,
"loss": 0.14407466650009154,
"step": 4410
},
{
"epoch": 1.4144,
"grad_norm": 0.4704024214191625,
"learning_rate": 4.1742577018950205e-06,
"loss": 0.1447770357131958,
"step": 4420
},
{
"epoch": 1.4176,
"grad_norm": 0.41646527190589205,
"learning_rate": 4.132214449951842e-06,
"loss": 0.1404719114303589,
"step": 4430
},
{
"epoch": 1.4208,
"grad_norm": 0.430577942578432,
"learning_rate": 4.090328792539848e-06,
"loss": 0.1420907497406006,
"step": 4440
},
{
"epoch": 1.424,
"grad_norm": 0.4565625212409375,
"learning_rate": 4.048601854606519e-06,
"loss": 0.14833756685256957,
"step": 4450
},
{
"epoch": 1.4272,
"grad_norm": 0.43429612904548576,
"learning_rate": 4.007034756836505e-06,
"loss": 0.1458716869354248,
"step": 4460
},
{
"epoch": 1.4304000000000001,
"grad_norm": 0.4549040695992988,
"learning_rate": 3.965628615621542e-06,
"loss": 0.13687772750854493,
"step": 4470
},
{
"epoch": 1.4336,
"grad_norm": 0.4241157059563243,
"learning_rate": 3.924384543030464e-06,
"loss": 0.1489526152610779,
"step": 4480
},
{
"epoch": 1.4368,
"grad_norm": 0.4194255280681186,
"learning_rate": 3.8833036467793224e-06,
"loss": 0.1386746883392334,
"step": 4490
},
{
"epoch": 1.44,
"grad_norm": 0.3782602629635505,
"learning_rate": 3.842387030201655e-06,
"loss": 0.1376037359237671,
"step": 4500
},
{
"epoch": 1.4432,
"grad_norm": 0.4153807526811226,
"learning_rate": 3.8016357922188496e-06,
"loss": 0.1373946785926819,
"step": 4510
},
{
"epoch": 1.4464000000000001,
"grad_norm": 0.45629103722813863,
"learning_rate": 3.761051027310617e-06,
"loss": 0.1372739315032959,
"step": 4520
},
{
"epoch": 1.4496,
"grad_norm": 0.40430778404566603,
"learning_rate": 3.72063382548561e-06,
"loss": 0.14090638160705565,
"step": 4530
},
{
"epoch": 1.4527999999999999,
"grad_norm": 0.4437840627529129,
"learning_rate": 3.6803852722521294e-06,
"loss": 0.14170970916748046,
"step": 4540
},
{
"epoch": 1.456,
"grad_norm": 0.44580819697797336,
"learning_rate": 3.640306448588996e-06,
"loss": 0.14584769010543824,
"step": 4550
},
{
"epoch": 1.4592,
"grad_norm": 0.41229577462439376,
"learning_rate": 3.6003984309164997e-06,
"loss": 0.1420591115951538,
"step": 4560
},
{
"epoch": 1.4624,
"grad_norm": 0.5235658173208154,
"learning_rate": 3.560662291067496e-06,
"loss": 0.15003039836883544,
"step": 4570
},
{
"epoch": 1.4656,
"grad_norm": 0.4150493582594386,
"learning_rate": 3.5210990962586154e-06,
"loss": 0.14279818534851074,
"step": 4580
},
{
"epoch": 1.4687999999999999,
"grad_norm": 0.45846599888609124,
"learning_rate": 3.481709909061608e-06,
"loss": 0.14115376472473146,
"step": 4590
},
{
"epoch": 1.472,
"grad_norm": 0.5069176213144664,
"learning_rate": 3.4424957873747898e-06,
"loss": 0.14473507404327393,
"step": 4600
},
{
"epoch": 1.4752,
"grad_norm": 0.4086970447753847,
"learning_rate": 3.4034577843946503e-06,
"loss": 0.14712631702423096,
"step": 4610
},
{
"epoch": 1.4784,
"grad_norm": 0.44335521431457986,
"learning_rate": 3.3645969485875528e-06,
"loss": 0.14121794700622559,
"step": 4620
},
{
"epoch": 1.4816,
"grad_norm": 0.4096646163299698,
"learning_rate": 3.32591432366158e-06,
"loss": 0.14097131490707399,
"step": 4630
},
{
"epoch": 1.4848,
"grad_norm": 0.4738878565930835,
"learning_rate": 3.2874109485384986e-06,
"loss": 0.14501018524169923,
"step": 4640
},
{
"epoch": 1.488,
"grad_norm": 0.46665484431929005,
"learning_rate": 3.2490878573258623e-06,
"loss": 0.13956862688064575,
"step": 4650
},
{
"epoch": 1.4912,
"grad_norm": 0.46057793559551674,
"learning_rate": 3.210946079289232e-06,
"loss": 0.13989123106002807,
"step": 4660
},
{
"epoch": 1.4944,
"grad_norm": 0.4093464688850494,
"learning_rate": 3.172986638824538e-06,
"loss": 0.14247305393218995,
"step": 4670
},
{
"epoch": 1.4976,
"grad_norm": 0.41699165029616675,
"learning_rate": 3.135210555430561e-06,
"loss": 0.1417543649673462,
"step": 4680
},
{
"epoch": 1.5008,
"grad_norm": 0.44782157921471477,
"learning_rate": 3.097618843681558e-06,
"loss": 0.14267433881759645,
"step": 4690
},
{
"epoch": 1.504,
"grad_norm": 0.4159318334264574,
"learning_rate": 3.0602125132000094e-06,
"loss": 0.1427731990814209,
"step": 4700
},
{
"epoch": 1.5072,
"grad_norm": 0.48485333808986114,
"learning_rate": 3.0229925686294935e-06,
"loss": 0.14570138454437256,
"step": 4710
},
{
"epoch": 1.5104,
"grad_norm": 0.4218617495100559,
"learning_rate": 2.9859600096077266e-06,
"loss": 0.146050763130188,
"step": 4720
},
{
"epoch": 1.5135999999999998,
"grad_norm": 0.46160012223248365,
"learning_rate": 2.949115830739696e-06,
"loss": 0.14328941106796264,
"step": 4730
},
{
"epoch": 1.5168,
"grad_norm": 0.42327675878618,
"learning_rate": 2.9124610215709536e-06,
"loss": 0.1439073085784912,
"step": 4740
},
{
"epoch": 1.52,
"grad_norm": 0.4362647125078029,
"learning_rate": 2.8759965665610413e-06,
"loss": 0.13419511318206787,
"step": 4750
},
{
"epoch": 1.5232,
"grad_norm": 0.4724776525869113,
"learning_rate": 2.839723445057049e-06,
"loss": 0.13762248754501344,
"step": 4760
},
{
"epoch": 1.5264,
"grad_norm": 0.4425061488213635,
"learning_rate": 2.8036426312673036e-06,
"loss": 0.1396605849266052,
"step": 4770
},
{
"epoch": 1.5295999999999998,
"grad_norm": 0.42813152437597185,
"learning_rate": 2.7677550942352227e-06,
"loss": 0.13956373929977417,
"step": 4780
},
{
"epoch": 1.5328,
"grad_norm": 0.4773255268965,
"learning_rate": 2.73206179781327e-06,
"loss": 0.1380013108253479,
"step": 4790
},
{
"epoch": 1.536,
"grad_norm": 0.4771949764129806,
"learning_rate": 2.6965637006370803e-06,
"loss": 0.1415919065475464,
"step": 4800
},
{
"epoch": 1.5392000000000001,
"grad_norm": 0.4247560084798639,
"learning_rate": 2.661261756099707e-06,
"loss": 0.13790349960327147,
"step": 4810
},
{
"epoch": 1.5424,
"grad_norm": 0.4251374231471267,
"learning_rate": 2.6261569123260165e-06,
"loss": 0.13341727256774902,
"step": 4820
},
{
"epoch": 1.5455999999999999,
"grad_norm": 0.4634708649512159,
"learning_rate": 2.5912501121472287e-06,
"loss": 0.1372056484222412,
"step": 4830
},
{
"epoch": 1.5488,
"grad_norm": 0.43564269280454887,
"learning_rate": 2.556542293075588e-06,
"loss": 0.1422506332397461,
"step": 4840
},
{
"epoch": 1.552,
"grad_norm": 0.4389171780201917,
"learning_rate": 2.5220343872791898e-06,
"loss": 0.1365510940551758,
"step": 4850
},
{
"epoch": 1.5552000000000001,
"grad_norm": 0.43725025717343674,
"learning_rate": 2.4877273215569443e-06,
"loss": 0.13687204122543334,
"step": 4860
},
{
"epoch": 1.5584,
"grad_norm": 0.42299546017617295,
"learning_rate": 2.453622017313674e-06,
"loss": 0.14183000326156617,
"step": 4870
},
{
"epoch": 1.5615999999999999,
"grad_norm": 0.4333528010588871,
"learning_rate": 2.4197193905353842e-06,
"loss": 0.13642750978469848,
"step": 4880
},
{
"epoch": 1.5648,
"grad_norm": 0.4362457931543249,
"learning_rate": 2.3860203517646517e-06,
"loss": 0.13224549293518068,
"step": 4890
},
{
"epoch": 1.568,
"grad_norm": 0.4319734529200874,
"learning_rate": 2.3525258060761734e-06,
"loss": 0.1371970295906067,
"step": 4900
},
{
"epoch": 1.5712000000000002,
"grad_norm": 0.4099475519452622,
"learning_rate": 2.3192366530524513e-06,
"loss": 0.13916906118392944,
"step": 4910
},
{
"epoch": 1.5744,
"grad_norm": 0.4935910437733969,
"learning_rate": 2.286153786759643e-06,
"loss": 0.13969881534576417,
"step": 4920
},
{
"epoch": 1.5776,
"grad_norm": 0.4343731618899666,
"learning_rate": 2.2532780957235423e-06,
"loss": 0.14086977243423462,
"step": 4930
},
{
"epoch": 1.5808,
"grad_norm": 0.4450724443635755,
"learning_rate": 2.220610462905709e-06,
"loss": 0.13837013244628907,
"step": 4940
},
{
"epoch": 1.584,
"grad_norm": 0.4446260432739019,
"learning_rate": 2.188151765679768e-06,
"loss": 0.14388253688812255,
"step": 4950
},
{
"epoch": 1.5872000000000002,
"grad_norm": 0.4678158235163482,
"learning_rate": 2.155902875807846e-06,
"loss": 0.1334782838821411,
"step": 4960
},
{
"epoch": 1.5904,
"grad_norm": 0.4724040337896473,
"learning_rate": 2.123864659417144e-06,
"loss": 0.14168214797973633,
"step": 4970
},
{
"epoch": 1.5936,
"grad_norm": 0.4300022512980927,
"learning_rate": 2.0920379769766787e-06,
"loss": 0.14135398864746093,
"step": 4980
},
{
"epoch": 1.5968,
"grad_norm": 0.47922705000758936,
"learning_rate": 2.0604236832741852e-06,
"loss": 0.1382269024848938,
"step": 4990
},
{
"epoch": 1.6,
"grad_norm": 0.4508427087497204,
"learning_rate": 2.0290226273931457e-06,
"loss": 0.13884921073913575,
"step": 5000
},
{
"epoch": 1.6032,
"grad_norm": 0.4161572388594623,
"learning_rate": 1.9978356526899912e-06,
"loss": 0.13653182983398438,
"step": 5010
},
{
"epoch": 1.6064,
"grad_norm": 0.46313899130075753,
"learning_rate": 1.966863596771451e-06,
"loss": 0.14188729524612426,
"step": 5020
},
{
"epoch": 1.6096,
"grad_norm": 0.4722692335225561,
"learning_rate": 1.936107291472057e-06,
"loss": 0.14433767795562744,
"step": 5030
},
{
"epoch": 1.6128,
"grad_norm": 0.42756557875825413,
"learning_rate": 1.9055675628317926e-06,
"loss": 0.1385686993598938,
"step": 5040
},
{
"epoch": 1.616,
"grad_norm": 0.46054145836890564,
"learning_rate": 1.8752452310739256e-06,
"loss": 0.14349981546401977,
"step": 5050
},
{
"epoch": 1.6192,
"grad_norm": 0.4836820660223868,
"learning_rate": 1.8451411105829653e-06,
"loss": 0.13743669986724855,
"step": 5060
},
{
"epoch": 1.6223999999999998,
"grad_norm": 0.4442414773401163,
"learning_rate": 1.8152560098827953e-06,
"loss": 0.13747843503952026,
"step": 5070
},
{
"epoch": 1.6256,
"grad_norm": 0.49633450773782095,
"learning_rate": 1.7855907316149546e-06,
"loss": 0.13720426559448243,
"step": 5080
},
{
"epoch": 1.6288,
"grad_norm": 0.41017005089538033,
"learning_rate": 1.7561460725170887e-06,
"loss": 0.14144107103347778,
"step": 5090
},
{
"epoch": 1.6320000000000001,
"grad_norm": 0.42328112922160144,
"learning_rate": 1.7269228234015345e-06,
"loss": 0.13741862773895264,
"step": 5100
},
{
"epoch": 1.6352,
"grad_norm": 0.42990591915446885,
"learning_rate": 1.6979217691341054e-06,
"loss": 0.13247919082641602,
"step": 5110
},
{
"epoch": 1.6383999999999999,
"grad_norm": 0.46728040991273373,
"learning_rate": 1.669143688612992e-06,
"loss": 0.14041290283203126,
"step": 5120
},
{
"epoch": 1.6416,
"grad_norm": 0.4516163275937175,
"learning_rate": 1.6405893547478491e-06,
"loss": 0.1399753451347351,
"step": 5130
},
{
"epoch": 1.6448,
"grad_norm": 0.4636892946415084,
"learning_rate": 1.6122595344390402e-06,
"loss": 0.14033472537994385,
"step": 5140
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.4467388794841459,
"learning_rate": 1.5841549885570307e-06,
"loss": 0.14232743978500367,
"step": 5150
},
{
"epoch": 1.6512,
"grad_norm": 0.4580726235379971,
"learning_rate": 1.5562764719219658e-06,
"loss": 0.14039382934570313,
"step": 5160
},
{
"epoch": 1.6543999999999999,
"grad_norm": 0.4069774634987291,
"learning_rate": 1.5286247332833926e-06,
"loss": 0.13857524394989013,
"step": 5170
},
{
"epoch": 1.6576,
"grad_norm": 0.4360341162685081,
"learning_rate": 1.5012005153001463e-06,
"loss": 0.14505903720855712,
"step": 5180
},
{
"epoch": 1.6608,
"grad_norm": 0.47930150294738827,
"learning_rate": 1.4740045545204073e-06,
"loss": 0.1332105278968811,
"step": 5190
},
{
"epoch": 1.6640000000000001,
"grad_norm": 0.44899479746425036,
"learning_rate": 1.4470375813619231e-06,
"loss": 0.14088051319122313,
"step": 5200
},
{
"epoch": 1.6672,
"grad_norm": 0.4423494298958755,
"learning_rate": 1.4203003200923815e-06,
"loss": 0.14016900062561036,
"step": 5210
},
{
"epoch": 1.6703999999999999,
"grad_norm": 0.4421418158037591,
"learning_rate": 1.393793488809969e-06,
"loss": 0.13580211400985717,
"step": 5220
},
{
"epoch": 1.6736,
"grad_norm": 0.40701385888299624,
"learning_rate": 1.3675177994240784e-06,
"loss": 0.13472111225128175,
"step": 5230
},
{
"epoch": 1.6768,
"grad_norm": 0.45815081953046,
"learning_rate": 1.3414739576361913e-06,
"loss": 0.13832969665527345,
"step": 5240
},
{
"epoch": 1.6800000000000002,
"grad_norm": 0.5071195521294466,
"learning_rate": 1.3156626629209224e-06,
"loss": 0.13420993089675903,
"step": 5250
},
{
"epoch": 1.6832,
"grad_norm": 0.4127840405266335,
"learning_rate": 1.2900846085072328e-06,
"loss": 0.14366337060928344,
"step": 5260
},
{
"epoch": 1.6864,
"grad_norm": 0.47511453648289,
"learning_rate": 1.2647404813598163e-06,
"loss": 0.14288046360015869,
"step": 5270
},
{
"epoch": 1.6896,
"grad_norm": 0.4642677094830462,
"learning_rate": 1.2396309621606428e-06,
"loss": 0.1446826457977295,
"step": 5280
},
{
"epoch": 1.6928,
"grad_norm": 0.4815421824805429,
"learning_rate": 1.2147567252906833e-06,
"loss": 0.1393401503562927,
"step": 5290
},
{
"epoch": 1.696,
"grad_norm": 0.46411081098461066,
"learning_rate": 1.1901184388117915e-06,
"loss": 0.13378889560699464,
"step": 5300
},
{
"epoch": 1.6992,
"grad_norm": 0.45456214213330326,
"learning_rate": 1.1657167644487677e-06,
"loss": 0.1344744086265564,
"step": 5310
},
{
"epoch": 1.7024,
"grad_norm": 0.45711413497200987,
"learning_rate": 1.1415523575715758e-06,
"loss": 0.13973560333251953,
"step": 5320
},
{
"epoch": 1.7056,
"grad_norm": 0.4775944401931014,
"learning_rate": 1.1176258671777562e-06,
"loss": 0.1370839834213257,
"step": 5330
},
{
"epoch": 1.7088,
"grad_norm": 0.4585631515862284,
"learning_rate": 1.0939379358749847e-06,
"loss": 0.1421511173248291,
"step": 5340
},
{
"epoch": 1.712,
"grad_norm": 0.44422507290687796,
"learning_rate": 1.0704891998638168e-06,
"loss": 0.13729255199432372,
"step": 5350
},
{
"epoch": 1.7151999999999998,
"grad_norm": 0.41764656895801,
"learning_rate": 1.0472802889206013e-06,
"loss": 0.14359140396118164,
"step": 5360
},
{
"epoch": 1.7184,
"grad_norm": 0.4426566449536881,
"learning_rate": 1.0243118263805685e-06,
"loss": 0.13689072132110597,
"step": 5370
},
{
"epoch": 1.7216,
"grad_norm": 0.4179586529401397,
"learning_rate": 1.0015844291210775e-06,
"loss": 0.13186384439468385,
"step": 5380
},
{
"epoch": 1.7248,
"grad_norm": 0.446510659582666,
"learning_rate": 9.790987075450652e-07,
"loss": 0.13834744691848755,
"step": 5390
},
{
"epoch": 1.728,
"grad_norm": 0.5126496112923015,
"learning_rate": 9.56855265564639e-07,
"loss": 0.14210819005966185,
"step": 5400
},
{
"epoch": 1.7311999999999999,
"grad_norm": 0.4464790298992936,
"learning_rate": 9.348547005848696e-07,
"loss": 0.13760073184967042,
"step": 5410
},
{
"epoch": 1.7344,
"grad_norm": 0.48147891500546974,
"learning_rate": 9.130976034877303e-07,
"loss": 0.13761359453201294,
"step": 5420
},
{
"epoch": 1.7376,
"grad_norm": 0.5225201468019522,
"learning_rate": 8.915845586162386e-07,
"loss": 0.13882596492767335,
"step": 5430
},
{
"epoch": 1.7408000000000001,
"grad_norm": 0.39148554754230247,
"learning_rate": 8.703161437587604e-07,
"loss": 0.12863160371780397,
"step": 5440
},
{
"epoch": 1.744,
"grad_norm": 0.4409802567795446,
"learning_rate": 8.492929301334896e-07,
"loss": 0.1338444709777832,
"step": 5450
},
{
"epoch": 1.7471999999999999,
"grad_norm": 0.4541676130570632,
"learning_rate": 8.285154823731101e-07,
"loss": 0.13422996997833253,
"step": 5460
},
{
"epoch": 1.7504,
"grad_norm": 0.42809913585936676,
"learning_rate": 8.079843585096259e-07,
"loss": 0.13907593488693237,
"step": 5470
},
{
"epoch": 1.7536,
"grad_norm": 0.4328261547821427,
"learning_rate": 7.877001099593818e-07,
"loss": 0.14067745208740234,
"step": 5480
},
{
"epoch": 1.7568000000000001,
"grad_norm": 0.456079329026488,
"learning_rate": 7.676632815082408e-07,
"loss": 0.13802475929260255,
"step": 5490
},
{
"epoch": 1.76,
"grad_norm": 0.440742597392519,
"learning_rate": 7.47874411296966e-07,
"loss": 0.13189576864242553,
"step": 5500
},
{
"epoch": 1.7631999999999999,
"grad_norm": 0.43343116893965217,
"learning_rate": 7.283340308067621e-07,
"loss": 0.1389593482017517,
"step": 5510
},
{
"epoch": 1.7664,
"grad_norm": 0.4369296090155059,
"learning_rate": 7.090426648449977e-07,
"loss": 0.1361026406288147,
"step": 5520
},
{
"epoch": 1.7696,
"grad_norm": 0.44463646756456265,
"learning_rate": 6.900008315311147e-07,
"loss": 0.1334540843963623,
"step": 5530
},
{
"epoch": 1.7728000000000002,
"grad_norm": 0.4347479300182727,
"learning_rate": 6.712090422827144e-07,
"loss": 0.1351664423942566,
"step": 5540
},
{
"epoch": 1.776,
"grad_norm": 0.4252524201661553,
"learning_rate": 6.526678018018107e-07,
"loss": 0.13711977005004883,
"step": 5550
},
{
"epoch": 1.7792,
"grad_norm": 0.44161084874301865,
"learning_rate": 6.343776080612941e-07,
"loss": 0.1323324680328369,
"step": 5560
},
{
"epoch": 1.7824,
"grad_norm": 0.38860040321784506,
"learning_rate": 6.163389522915387e-07,
"loss": 0.1394704818725586,
"step": 5570
},
{
"epoch": 1.7856,
"grad_norm": 0.4136779998451306,
"learning_rate": 5.985523189672215e-07,
"loss": 0.13335754871368408,
"step": 5580
},
{
"epoch": 1.7888,
"grad_norm": 0.4323523832298023,
"learning_rate": 5.810181857943054e-07,
"loss": 0.1335891842842102,
"step": 5590
},
{
"epoch": 1.792,
"grad_norm": 0.3977869584829257,
"learning_rate": 5.63737023697205e-07,
"loss": 0.13949548006057738,
"step": 5600
},
{
"epoch": 1.7952,
"grad_norm": 0.4053507509038296,
"learning_rate": 5.467092968061494e-07,
"loss": 0.13953766822814942,
"step": 5610
},
{
"epoch": 1.7984,
"grad_norm": 2.3159376802413103,
"learning_rate": 5.299354624447084e-07,
"loss": 0.13574401140213013,
"step": 5620
},
{
"epoch": 1.8016,
"grad_norm": 0.44618272798844455,
"learning_rate": 5.134159711175135e-07,
"loss": 0.13270254135131837,
"step": 5630
},
{
"epoch": 1.8048,
"grad_norm": 0.44181894607134586,
"learning_rate": 4.971512664981548e-07,
"loss": 0.13275084495544434,
"step": 5640
},
{
"epoch": 1.808,
"grad_norm": 0.4818852107908557,
"learning_rate": 4.811417854172729e-07,
"loss": 0.14589433670043944,
"step": 5650
},
{
"epoch": 1.8112,
"grad_norm": 0.46715654398775347,
"learning_rate": 4.653879578508136e-07,
"loss": 0.13999476432800292,
"step": 5660
},
{
"epoch": 1.8144,
"grad_norm": 0.4483410244408988,
"learning_rate": 4.4989020690849315e-07,
"loss": 0.1353672981262207,
"step": 5670
},
{
"epoch": 1.8176,
"grad_norm": 0.49502628267447824,
"learning_rate": 4.3464894882242705e-07,
"loss": 0.13813743591308594,
"step": 5680
},
{
"epoch": 1.8208,
"grad_norm": 0.3991057089987906,
"learning_rate": 4.196645929359533e-07,
"loss": 0.13932307958602905,
"step": 5690
},
{
"epoch": 1.8239999999999998,
"grad_norm": 0.414166568313499,
"learning_rate": 4.049375416926382e-07,
"loss": 0.1340416193008423,
"step": 5700
},
{
"epoch": 1.8272,
"grad_norm": 0.42576097286815373,
"learning_rate": 3.9046819062546705e-07,
"loss": 0.1395948886871338,
"step": 5710
},
{
"epoch": 1.8304,
"grad_norm": 0.4032147777900584,
"learning_rate": 3.762569283462214e-07,
"loss": 0.1376317858695984,
"step": 5720
},
{
"epoch": 1.8336000000000001,
"grad_norm": 0.44989195218742856,
"learning_rate": 3.6230413653504327e-07,
"loss": 0.13669462203979493,
"step": 5730
},
{
"epoch": 1.8368,
"grad_norm": 0.4438644111136516,
"learning_rate": 3.486101899301797e-07,
"loss": 0.1317627787590027,
"step": 5740
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.41880717043654114,
"learning_rate": 3.3517545631792767e-07,
"loss": 0.13299596309661865,
"step": 5750
},
{
"epoch": 1.8432,
"grad_norm": 0.4756221632929248,
"learning_rate": 3.220002965227409e-07,
"loss": 0.1396946430206299,
"step": 5760
},
{
"epoch": 1.8464,
"grad_norm": 0.4437499812814737,
"learning_rate": 3.0908506439755625e-07,
"loss": 0.13375253677368165,
"step": 5770
},
{
"epoch": 1.8496000000000001,
"grad_norm": 0.3878361397413755,
"learning_rate": 2.964301068142772e-07,
"loss": 0.12817233800888062,
"step": 5780
},
{
"epoch": 1.8528,
"grad_norm": 0.4651498686383848,
"learning_rate": 2.840357636544655e-07,
"loss": 0.14180712699890136,
"step": 5790
},
{
"epoch": 1.8559999999999999,
"grad_norm": 0.5480141328338131,
"learning_rate": 2.719023678002075e-07,
"loss": 0.13765851259231568,
"step": 5800
},
{
"epoch": 1.8592,
"grad_norm": 0.4906033655243011,
"learning_rate": 2.6003024512517683e-07,
"loss": 0.13171169757843018,
"step": 5810
},
{
"epoch": 1.8624,
"grad_norm": 0.4856076540728153,
"learning_rate": 2.484197144858802e-07,
"loss": 0.1396042823791504,
"step": 5820
},
{
"epoch": 1.8656000000000001,
"grad_norm": 0.4874848183833614,
"learning_rate": 2.370710877130955e-07,
"loss": 0.13632986545562745,
"step": 5830
},
{
"epoch": 1.8688,
"grad_norm": 0.41663687564556867,
"learning_rate": 2.259846696034962e-07,
"loss": 0.12913339138031005,
"step": 5840
},
{
"epoch": 1.8719999999999999,
"grad_norm": 0.49688262365950053,
"learning_rate": 2.1516075791146452e-07,
"loss": 0.1359020471572876,
"step": 5850
},
{
"epoch": 1.8752,
"grad_norm": 0.4757621929961804,
"learning_rate": 2.0459964334109794e-07,
"loss": 0.13454790115356446,
"step": 5860
},
{
"epoch": 1.8784,
"grad_norm": 0.4149727628472437,
"learning_rate": 1.9430160953839428e-07,
"loss": 0.13284122943878174,
"step": 5870
},
{
"epoch": 1.8816000000000002,
"grad_norm": 0.46826675476722257,
"learning_rate": 1.8426693308364108e-07,
"loss": 0.1376336097717285,
"step": 5880
},
{
"epoch": 1.8848,
"grad_norm": 0.45866753455612513,
"learning_rate": 1.7449588348398272e-07,
"loss": 0.14035621881484986,
"step": 5890
},
{
"epoch": 1.888,
"grad_norm": 0.434147661219072,
"learning_rate": 1.6498872316618397e-07,
"loss": 0.13828661441802978,
"step": 5900
},
{
"epoch": 1.8912,
"grad_norm": 0.45418794472882734,
"learning_rate": 1.5574570746958339e-07,
"loss": 0.13735527992248536,
"step": 5910
},
{
"epoch": 1.8944,
"grad_norm": 0.40855381115386324,
"learning_rate": 1.4676708463923329e-07,
"loss": 0.1430106282234192,
"step": 5920
},
{
"epoch": 1.8976,
"grad_norm": 0.45089245918606846,
"learning_rate": 1.3805309581922832e-07,
"loss": 0.13202996253967286,
"step": 5930
},
{
"epoch": 1.9008,
"grad_norm": 0.4385085502238853,
"learning_rate": 1.2960397504623968e-07,
"loss": 0.13440369367599486,
"step": 5940
},
{
"epoch": 1.904,
"grad_norm": 0.3997870824729913,
"learning_rate": 1.214199492432211e-07,
"loss": 0.1342543125152588,
"step": 5950
},
{
"epoch": 1.9072,
"grad_norm": 0.46917598645018627,
"learning_rate": 1.1350123821331493e-07,
"loss": 0.13450114727020263,
"step": 5960
},
{
"epoch": 1.9104,
"grad_norm": 0.40749771807525076,
"learning_rate": 1.0584805463395242e-07,
"loss": 0.13663610219955444,
"step": 5970
},
{
"epoch": 1.9136,
"grad_norm": 0.47811929874068837,
"learning_rate": 9.846060405113821e-08,
"loss": 0.13826658725738525,
"step": 5980
},
{
"epoch": 1.9167999999999998,
"grad_norm": 0.48860395938414863,
"learning_rate": 9.133908487393262e-08,
"loss": 0.13649516105651854,
"step": 5990
},
{
"epoch": 1.92,
"grad_norm": 0.4668638499416011,
"learning_rate": 8.448368836911692e-08,
"loss": 0.13388367891311645,
"step": 6000
},
{
"epoch": 1.9232,
"grad_norm": 0.4618223030858044,
"learning_rate": 7.789459865606864e-08,
"loss": 0.1364944338798523,
"step": 6010
},
{
"epoch": 1.9264000000000001,
"grad_norm": 0.4871755675798615,
"learning_rate": 7.157199270180326e-08,
"loss": 0.13289897441864013,
"step": 6020
},
{
"epoch": 1.9296,
"grad_norm": 0.4637376607300702,
"learning_rate": 6.551604031622805e-08,
"loss": 0.13801203966140746,
"step": 6030
},
{
"epoch": 1.9327999999999999,
"grad_norm": 0.4473914125162625,
"learning_rate": 5.972690414758008e-08,
"loss": 0.135065758228302,
"step": 6040
},
{
"epoch": 1.936,
"grad_norm": 0.44311936776321703,
"learning_rate": 5.420473967805762e-08,
"loss": 0.1335088014602661,
"step": 6050
},
{
"epoch": 1.9392,
"grad_norm": 0.46182194980520064,
"learning_rate": 4.8949695219647763e-08,
"loss": 0.14293944835662842,
"step": 6060
},
{
"epoch": 1.9424000000000001,
"grad_norm": 0.5226455149331062,
"learning_rate": 4.396191191013532e-08,
"loss": 0.1381105899810791,
"step": 6070
},
{
"epoch": 1.9456,
"grad_norm": 0.47124584687681537,
"learning_rate": 3.924152370932022e-08,
"loss": 0.13668475151062012,
"step": 6080
},
{
"epoch": 1.9487999999999999,
"grad_norm": 0.444679809855597,
"learning_rate": 3.4788657395414813e-08,
"loss": 0.13323181867599487,
"step": 6090
},
{
"epoch": 1.952,
"grad_norm": 0.5258053555724833,
"learning_rate": 3.060343256164222e-08,
"loss": 0.13993802070617675,
"step": 6100
},
{
"epoch": 1.9552,
"grad_norm": 0.44867967593793734,
"learning_rate": 2.668596161301995e-08,
"loss": 0.1421555757522583,
"step": 6110
},
{
"epoch": 1.9584000000000001,
"grad_norm": 0.44233188918218574,
"learning_rate": 2.3036349763344567e-08,
"loss": 0.14029024839401244,
"step": 6120
},
{
"epoch": 1.9616,
"grad_norm": 0.4514789852845869,
"learning_rate": 1.9654695032367278e-08,
"loss": 0.13499693870544432,
"step": 6130
},
{
"epoch": 1.9647999999999999,
"grad_norm": 0.4594976709436296,
"learning_rate": 1.6541088243156035e-08,
"loss": 0.1375267744064331,
"step": 6140
},
{
"epoch": 1.968,
"grad_norm": 0.430422483066298,
"learning_rate": 1.3695613019659716e-08,
"loss": 0.13967254161834716,
"step": 6150
},
{
"epoch": 1.9712,
"grad_norm": 0.4536587569169835,
"learning_rate": 1.1118345784463247e-08,
"loss": 0.13422807455062866,
"step": 6160
},
{
"epoch": 1.9744000000000002,
"grad_norm": 0.4610181498669583,
"learning_rate": 8.809355756732585e-09,
"loss": 0.13575806617736816,
"step": 6170
},
{
"epoch": 1.9776,
"grad_norm": 0.46487435443295355,
"learning_rate": 6.7687049503561975e-09,
"loss": 0.14011634588241578,
"step": 6180
},
{
"epoch": 1.9808,
"grad_norm": 0.45486168002394717,
"learning_rate": 4.996448172279733e-09,
"loss": 0.13615384101867675,
"step": 6190
},
{
"epoch": 1.984,
"grad_norm": 0.4857202159763611,
"learning_rate": 3.4926330210360846e-09,
"loss": 0.14173336029052735,
"step": 6200
},
{
"epoch": 1.9872,
"grad_norm": 0.4364865514843521,
"learning_rate": 2.257299885461972e-09,
"loss": 0.1347902774810791,
"step": 6210
},
{
"epoch": 1.9904,
"grad_norm": 0.4811459059430188,
"learning_rate": 1.2904819436210247e-09,
"loss": 0.13855082988739015,
"step": 6220
},
{
"epoch": 1.9936,
"grad_norm": 0.45909428175159994,
"learning_rate": 5.922051619033919e-10,
"loss": 0.12805842161178588,
"step": 6230
},
{
"epoch": 1.9968,
"grad_norm": 0.44711524032559186,
"learning_rate": 1.6248829433518355e-10,
"loss": 0.13427029848098754,
"step": 6240
},
{
"epoch": 2.0,
"grad_norm": 0.5591263504728373,
"learning_rate": 1.3428820722083402e-12,
"loss": 0.1432194948196411,
"step": 6250
},
{
"epoch": 2.0,
"step": 6250,
"total_flos": 866286168834048.0,
"train_loss": 0.04097769989013672,
"train_runtime": 29064.5154,
"train_samples_per_second": 13.762,
"train_steps_per_second": 0.215
}
],
"logging_steps": 10,
"max_steps": 6250,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 866286168834048.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}