SQSD / initial-state /dolly_ckpt_5850 /trainer_state.json
Wxxxx's picture
Upload folder using huggingface_hub
c83515c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 9.36,
"eval_steps": 500,
"global_step": 5850,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.016,
"grad_norm": 0.7532607316970825,
"learning_rate": 7.200000000000001e-08,
"loss": 2.4579,
"step": 10
},
{
"epoch": 0.032,
"grad_norm": 1.1350858211517334,
"learning_rate": 1.52e-07,
"loss": 2.5284,
"step": 20
},
{
"epoch": 0.048,
"grad_norm": 0.7984673380851746,
"learning_rate": 2.32e-07,
"loss": 2.4452,
"step": 30
},
{
"epoch": 0.064,
"grad_norm": 1.093232274055481,
"learning_rate": 3.12e-07,
"loss": 2.4453,
"step": 40
},
{
"epoch": 0.08,
"grad_norm": 1.177712321281433,
"learning_rate": 3.92e-07,
"loss": 2.9928,
"step": 50
},
{
"epoch": 0.096,
"grad_norm": 1.0691078901290894,
"learning_rate": 4.7200000000000004e-07,
"loss": 2.7053,
"step": 60
},
{
"epoch": 0.112,
"grad_norm": 1.2344281673431396,
"learning_rate": 5.520000000000001e-07,
"loss": 2.8079,
"step": 70
},
{
"epoch": 0.128,
"grad_norm": 0.6605137586593628,
"learning_rate": 6.320000000000002e-07,
"loss": 2.4089,
"step": 80
},
{
"epoch": 0.144,
"grad_norm": 1.0095362663269043,
"learning_rate": 7.12e-07,
"loss": 2.5198,
"step": 90
},
{
"epoch": 0.16,
"grad_norm": 1.014433741569519,
"learning_rate": 7.920000000000001e-07,
"loss": 3.1162,
"step": 100
},
{
"epoch": 0.176,
"grad_norm": 1.0782325267791748,
"learning_rate": 8.720000000000001e-07,
"loss": 2.6057,
"step": 110
},
{
"epoch": 0.192,
"grad_norm": 1.3253213167190552,
"learning_rate": 9.520000000000002e-07,
"loss": 2.6288,
"step": 120
},
{
"epoch": 0.208,
"grad_norm": 1.0826811790466309,
"learning_rate": 1.032e-06,
"loss": 2.8531,
"step": 130
},
{
"epoch": 0.224,
"grad_norm": 1.3482102155685425,
"learning_rate": 1.1120000000000001e-06,
"loss": 2.6991,
"step": 140
},
{
"epoch": 0.24,
"grad_norm": 1.6016103029251099,
"learning_rate": 1.1920000000000002e-06,
"loss": 2.5683,
"step": 150
},
{
"epoch": 0.256,
"grad_norm": 0.8933637738227844,
"learning_rate": 1.2720000000000003e-06,
"loss": 2.5362,
"step": 160
},
{
"epoch": 0.272,
"grad_norm": 0.804311215877533,
"learning_rate": 1.352e-06,
"loss": 2.6641,
"step": 170
},
{
"epoch": 0.288,
"grad_norm": 3.0151922702789307,
"learning_rate": 1.432e-06,
"loss": 2.744,
"step": 180
},
{
"epoch": 0.304,
"grad_norm": 3.0294179916381836,
"learning_rate": 1.512e-06,
"loss": 2.6508,
"step": 190
},
{
"epoch": 0.32,
"grad_norm": 1.329832673072815,
"learning_rate": 1.5920000000000002e-06,
"loss": 2.4969,
"step": 200
},
{
"epoch": 0.336,
"grad_norm": 1.018463134765625,
"learning_rate": 1.672e-06,
"loss": 2.7223,
"step": 210
},
{
"epoch": 0.352,
"grad_norm": 3.8947651386260986,
"learning_rate": 1.7520000000000001e-06,
"loss": 2.7512,
"step": 220
},
{
"epoch": 0.368,
"grad_norm": 1.932680606842041,
"learning_rate": 1.8320000000000002e-06,
"loss": 2.1643,
"step": 230
},
{
"epoch": 0.384,
"grad_norm": 1.0789552927017212,
"learning_rate": 1.912e-06,
"loss": 2.2978,
"step": 240
},
{
"epoch": 0.4,
"grad_norm": 1.4427675008773804,
"learning_rate": 1.992e-06,
"loss": 2.5924,
"step": 250
},
{
"epoch": 0.416,
"grad_norm": 2.1163601875305176,
"learning_rate": 2.0720000000000002e-06,
"loss": 2.4019,
"step": 260
},
{
"epoch": 0.432,
"grad_norm": 1.1988381147384644,
"learning_rate": 2.1520000000000003e-06,
"loss": 2.4047,
"step": 270
},
{
"epoch": 0.448,
"grad_norm": 1.1772013902664185,
"learning_rate": 2.2320000000000004e-06,
"loss": 2.4253,
"step": 280
},
{
"epoch": 0.464,
"grad_norm": 1.5872080326080322,
"learning_rate": 2.312e-06,
"loss": 1.9848,
"step": 290
},
{
"epoch": 0.48,
"grad_norm": 1.0806490182876587,
"learning_rate": 2.392e-06,
"loss": 2.1602,
"step": 300
},
{
"epoch": 0.496,
"grad_norm": 1.6096185445785522,
"learning_rate": 2.4720000000000002e-06,
"loss": 2.0775,
"step": 310
},
{
"epoch": 0.512,
"grad_norm": 1.667353868484497,
"learning_rate": 2.552e-06,
"loss": 2.0046,
"step": 320
},
{
"epoch": 0.528,
"grad_norm": 1.1516858339309692,
"learning_rate": 2.632e-06,
"loss": 2.2142,
"step": 330
},
{
"epoch": 0.544,
"grad_norm": 1.5734412670135498,
"learning_rate": 2.712e-06,
"loss": 2.3045,
"step": 340
},
{
"epoch": 0.56,
"grad_norm": 1.2886998653411865,
"learning_rate": 2.792e-06,
"loss": 2.1157,
"step": 350
},
{
"epoch": 0.576,
"grad_norm": 1.6008777618408203,
"learning_rate": 2.872e-06,
"loss": 2.1303,
"step": 360
},
{
"epoch": 0.592,
"grad_norm": 0.9446455240249634,
"learning_rate": 2.9520000000000003e-06,
"loss": 1.9004,
"step": 370
},
{
"epoch": 0.608,
"grad_norm": 1.0258949995040894,
"learning_rate": 3.0320000000000004e-06,
"loss": 1.7588,
"step": 380
},
{
"epoch": 0.624,
"grad_norm": 3.327923536300659,
"learning_rate": 3.112e-06,
"loss": 1.7562,
"step": 390
},
{
"epoch": 0.64,
"grad_norm": 0.7848489880561829,
"learning_rate": 3.192e-06,
"loss": 2.0036,
"step": 400
},
{
"epoch": 0.656,
"grad_norm": 0.901010274887085,
"learning_rate": 3.272e-06,
"loss": 1.8734,
"step": 410
},
{
"epoch": 0.672,
"grad_norm": 1.4475135803222656,
"learning_rate": 3.3520000000000003e-06,
"loss": 1.8156,
"step": 420
},
{
"epoch": 0.688,
"grad_norm": 1.1953306198120117,
"learning_rate": 3.4320000000000003e-06,
"loss": 1.7434,
"step": 430
},
{
"epoch": 0.704,
"grad_norm": 1.0600852966308594,
"learning_rate": 3.5120000000000004e-06,
"loss": 1.792,
"step": 440
},
{
"epoch": 0.72,
"grad_norm": 0.8356530070304871,
"learning_rate": 3.5920000000000005e-06,
"loss": 1.9139,
"step": 450
},
{
"epoch": 0.736,
"grad_norm": 1.1501522064208984,
"learning_rate": 3.6720000000000006e-06,
"loss": 1.7984,
"step": 460
},
{
"epoch": 0.752,
"grad_norm": 1.3865025043487549,
"learning_rate": 3.7520000000000002e-06,
"loss": 1.6949,
"step": 470
},
{
"epoch": 0.768,
"grad_norm": 0.41009730100631714,
"learning_rate": 3.832e-06,
"loss": 1.4455,
"step": 480
},
{
"epoch": 0.784,
"grad_norm": 1.4735100269317627,
"learning_rate": 3.912e-06,
"loss": 1.9003,
"step": 490
},
{
"epoch": 0.8,
"grad_norm": 1.0598748922348022,
"learning_rate": 3.992e-06,
"loss": 1.7243,
"step": 500
},
{
"epoch": 0.816,
"grad_norm": 0.7431689500808716,
"learning_rate": 4.072e-06,
"loss": 1.4202,
"step": 510
},
{
"epoch": 0.832,
"grad_norm": 1.8781498670578003,
"learning_rate": 4.152e-06,
"loss": 1.7335,
"step": 520
},
{
"epoch": 0.848,
"grad_norm": 1.2483923435211182,
"learning_rate": 4.232e-06,
"loss": 1.6562,
"step": 530
},
{
"epoch": 0.864,
"grad_norm": 0.9352028965950012,
"learning_rate": 4.312e-06,
"loss": 1.6151,
"step": 540
},
{
"epoch": 0.88,
"grad_norm": 0.8558632731437683,
"learning_rate": 4.3920000000000005e-06,
"loss": 1.5366,
"step": 550
},
{
"epoch": 0.896,
"grad_norm": 1.0936564207077026,
"learning_rate": 4.4720000000000006e-06,
"loss": 1.6026,
"step": 560
},
{
"epoch": 0.912,
"grad_norm": 0.6884616613388062,
"learning_rate": 4.552000000000001e-06,
"loss": 1.597,
"step": 570
},
{
"epoch": 0.928,
"grad_norm": 0.917011022567749,
"learning_rate": 4.632000000000001e-06,
"loss": 1.7136,
"step": 580
},
{
"epoch": 0.944,
"grad_norm": 0.6565596461296082,
"learning_rate": 4.712000000000001e-06,
"loss": 1.5694,
"step": 590
},
{
"epoch": 0.96,
"grad_norm": 0.7469630241394043,
"learning_rate": 4.792000000000001e-06,
"loss": 1.7206,
"step": 600
},
{
"epoch": 0.976,
"grad_norm": 0.7540640234947205,
"learning_rate": 4.872000000000001e-06,
"loss": 1.5857,
"step": 610
},
{
"epoch": 0.992,
"grad_norm": 0.6710789203643799,
"learning_rate": 4.952e-06,
"loss": 1.702,
"step": 620
},
{
"epoch": 1.008,
"grad_norm": 0.8187215328216553,
"learning_rate": 4.999993761437838e-06,
"loss": 1.6127,
"step": 630
},
{
"epoch": 1.024,
"grad_norm": 1.805862545967102,
"learning_rate": 4.999923577971082e-06,
"loss": 1.7038,
"step": 640
},
{
"epoch": 1.04,
"grad_norm": 1.2114150524139404,
"learning_rate": 4.999775415031381e-06,
"loss": 1.5832,
"step": 650
},
{
"epoch": 1.056,
"grad_norm": 0.8302497267723083,
"learning_rate": 4.999549277240342e-06,
"loss": 1.4196,
"step": 660
},
{
"epoch": 1.072,
"grad_norm": 0.41772493720054626,
"learning_rate": 4.999245171651823e-06,
"loss": 1.8191,
"step": 670
},
{
"epoch": 1.088,
"grad_norm": 1.5382171869277954,
"learning_rate": 4.998863107751711e-06,
"loss": 1.6041,
"step": 680
},
{
"epoch": 1.104,
"grad_norm": 0.9522147178649902,
"learning_rate": 4.998403097457629e-06,
"loss": 1.8859,
"step": 690
},
{
"epoch": 1.12,
"grad_norm": 0.48818162083625793,
"learning_rate": 4.997865155118557e-06,
"loss": 1.7237,
"step": 700
},
{
"epoch": 1.1360000000000001,
"grad_norm": 0.5890428423881531,
"learning_rate": 4.9972492975143936e-06,
"loss": 1.6564,
"step": 710
},
{
"epoch": 1.152,
"grad_norm": 0.7820872664451599,
"learning_rate": 4.996555543855426e-06,
"loss": 1.6438,
"step": 720
},
{
"epoch": 1.168,
"grad_norm": 0.8783853650093079,
"learning_rate": 4.995783915781734e-06,
"loss": 1.634,
"step": 730
},
{
"epoch": 1.184,
"grad_norm": 1.5447999238967896,
"learning_rate": 4.994934437362513e-06,
"loss": 1.4191,
"step": 740
},
{
"epoch": 1.2,
"grad_norm": 0.6656126379966736,
"learning_rate": 4.994007135095326e-06,
"loss": 1.5044,
"step": 750
},
{
"epoch": 1.216,
"grad_norm": 0.8769685626029968,
"learning_rate": 4.9930020379052715e-06,
"loss": 1.3899,
"step": 760
},
{
"epoch": 1.232,
"grad_norm": 1.2364633083343506,
"learning_rate": 4.9919191771440905e-06,
"loss": 1.4277,
"step": 770
},
{
"epoch": 1.248,
"grad_norm": 0.983772337436676,
"learning_rate": 4.990758586589178e-06,
"loss": 1.2793,
"step": 780
},
{
"epoch": 1.264,
"grad_norm": 0.914150595664978,
"learning_rate": 4.989520302442539e-06,
"loss": 1.7752,
"step": 790
},
{
"epoch": 1.28,
"grad_norm": 0.5487367510795593,
"learning_rate": 4.988204363329648e-06,
"loss": 1.5862,
"step": 800
},
{
"epoch": 1.296,
"grad_norm": 0.8448557257652283,
"learning_rate": 4.98681081029826e-06,
"loss": 1.5009,
"step": 810
},
{
"epoch": 1.312,
"grad_norm": 0.6510465145111084,
"learning_rate": 4.985339686817114e-06,
"loss": 1.5672,
"step": 820
},
{
"epoch": 1.328,
"grad_norm": 1.03178870677948,
"learning_rate": 4.983791038774585e-06,
"loss": 1.4901,
"step": 830
},
{
"epoch": 1.3439999999999999,
"grad_norm": 0.6161647439002991,
"learning_rate": 4.982164914477255e-06,
"loss": 1.5713,
"step": 840
},
{
"epoch": 1.3599999999999999,
"grad_norm": 1.0519344806671143,
"learning_rate": 4.980461364648399e-06,
"loss": 1.6525,
"step": 850
},
{
"epoch": 1.376,
"grad_norm": 0.5218610167503357,
"learning_rate": 4.978680442426409e-06,
"loss": 1.6173,
"step": 860
},
{
"epoch": 1.392,
"grad_norm": 0.6652182936668396,
"learning_rate": 4.976822203363135e-06,
"loss": 1.3168,
"step": 870
},
{
"epoch": 1.408,
"grad_norm": 0.9998406767845154,
"learning_rate": 4.974886705422149e-06,
"loss": 1.5201,
"step": 880
},
{
"epoch": 1.424,
"grad_norm": 0.6580556631088257,
"learning_rate": 4.97287400897694e-06,
"loss": 1.6578,
"step": 890
},
{
"epoch": 1.44,
"grad_norm": 1.8045531511306763,
"learning_rate": 4.970784176809032e-06,
"loss": 1.7959,
"step": 900
},
{
"epoch": 1.456,
"grad_norm": 0.7481666803359985,
"learning_rate": 4.968617274106019e-06,
"loss": 1.6719,
"step": 910
},
{
"epoch": 1.472,
"grad_norm": 0.4933053255081177,
"learning_rate": 4.966373368459542e-06,
"loss": 1.6534,
"step": 920
},
{
"epoch": 1.488,
"grad_norm": 0.768468976020813,
"learning_rate": 4.964052529863171e-06,
"loss": 1.685,
"step": 930
},
{
"epoch": 1.504,
"grad_norm": 0.648497462272644,
"learning_rate": 4.961654830710229e-06,
"loss": 1.6674,
"step": 940
},
{
"epoch": 1.52,
"grad_norm": 0.6459643244743347,
"learning_rate": 4.959180345791528e-06,
"loss": 1.7712,
"step": 950
},
{
"epoch": 1.536,
"grad_norm": 0.9094835519790649,
"learning_rate": 4.956629152293038e-06,
"loss": 1.8576,
"step": 960
},
{
"epoch": 1.552,
"grad_norm": 0.764054536819458,
"learning_rate": 4.954001329793483e-06,
"loss": 1.729,
"step": 970
},
{
"epoch": 1.568,
"grad_norm": 1.0334659814834595,
"learning_rate": 4.951296960261853e-06,
"loss": 1.6599,
"step": 980
},
{
"epoch": 1.584,
"grad_norm": 0.8426533341407776,
"learning_rate": 4.948516128054852e-06,
"loss": 1.4876,
"step": 990
},
{
"epoch": 1.6,
"grad_norm": 0.7043594121932983,
"learning_rate": 4.945658919914263e-06,
"loss": 1.5253,
"step": 1000
},
{
"epoch": 1.616,
"grad_norm": 0.7252524495124817,
"learning_rate": 4.9427254249642445e-06,
"loss": 1.5219,
"step": 1010
},
{
"epoch": 1.6320000000000001,
"grad_norm": 0.8069806098937988,
"learning_rate": 4.93971573470855e-06,
"loss": 1.5841,
"step": 1020
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.5051997900009155,
"learning_rate": 4.936629943027672e-06,
"loss": 1.5054,
"step": 1030
},
{
"epoch": 1.6640000000000001,
"grad_norm": 0.7083950638771057,
"learning_rate": 4.933468146175918e-06,
"loss": 1.616,
"step": 1040
},
{
"epoch": 1.6800000000000002,
"grad_norm": 0.6450363993644714,
"learning_rate": 4.930230442778403e-06,
"loss": 1.5149,
"step": 1050
},
{
"epoch": 1.696,
"grad_norm": 1.0520232915878296,
"learning_rate": 4.926916933827977e-06,
"loss": 1.6154,
"step": 1060
},
{
"epoch": 1.712,
"grad_norm": 0.7227002382278442,
"learning_rate": 4.92352772268207e-06,
"loss": 1.6732,
"step": 1070
},
{
"epoch": 1.728,
"grad_norm": 0.6014836430549622,
"learning_rate": 4.920062915059474e-06,
"loss": 1.6007,
"step": 1080
},
{
"epoch": 1.744,
"grad_norm": 0.9546307325363159,
"learning_rate": 4.916522619037043e-06,
"loss": 1.4207,
"step": 1090
},
{
"epoch": 1.76,
"grad_norm": 0.7327417135238647,
"learning_rate": 4.912906945046319e-06,
"loss": 1.6769,
"step": 1100
},
{
"epoch": 1.776,
"grad_norm": 0.8192687630653381,
"learning_rate": 4.909216005870091e-06,
"loss": 1.6177,
"step": 1110
},
{
"epoch": 1.792,
"grad_norm": 0.6791466474533081,
"learning_rate": 4.905449916638874e-06,
"loss": 1.7395,
"step": 1120
},
{
"epoch": 1.808,
"grad_norm": 0.5381625890731812,
"learning_rate": 4.901608794827321e-06,
"loss": 1.6157,
"step": 1130
},
{
"epoch": 1.8239999999999998,
"grad_norm": 0.6079979538917542,
"learning_rate": 4.897692760250556e-06,
"loss": 1.3609,
"step": 1140
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.8681022524833679,
"learning_rate": 4.8937019350604385e-06,
"loss": 1.4606,
"step": 1150
},
{
"epoch": 1.8559999999999999,
"grad_norm": 0.6992301940917969,
"learning_rate": 4.889636443741752e-06,
"loss": 1.6518,
"step": 1160
},
{
"epoch": 1.8719999999999999,
"grad_norm": 0.990447998046875,
"learning_rate": 4.885496413108321e-06,
"loss": 1.5659,
"step": 1170
},
{
"epoch": 1.888,
"grad_norm": 0.9825424551963806,
"learning_rate": 4.881281972299055e-06,
"loss": 1.5427,
"step": 1180
},
{
"epoch": 1.904,
"grad_norm": 0.7673924565315247,
"learning_rate": 4.876993252773923e-06,
"loss": 1.641,
"step": 1190
},
{
"epoch": 1.92,
"grad_norm": 0.796088457107544,
"learning_rate": 4.872630388309849e-06,
"loss": 1.6838,
"step": 1200
},
{
"epoch": 1.936,
"grad_norm": 0.9166289567947388,
"learning_rate": 4.868193514996542e-06,
"loss": 1.1927,
"step": 1210
},
{
"epoch": 1.952,
"grad_norm": 0.6708204746246338,
"learning_rate": 4.863682771232249e-06,
"loss": 1.6665,
"step": 1220
},
{
"epoch": 1.968,
"grad_norm": 0.5752835869789124,
"learning_rate": 4.85909829771944e-06,
"loss": 1.6575,
"step": 1230
},
{
"epoch": 1.984,
"grad_norm": 0.4450138509273529,
"learning_rate": 4.854440237460418e-06,
"loss": 1.5765,
"step": 1240
},
{
"epoch": 2.0,
"grad_norm": 0.932624876499176,
"learning_rate": 4.8497087357528585e-06,
"loss": 1.5759,
"step": 1250
},
{
"epoch": 2.016,
"grad_norm": 0.9572265148162842,
"learning_rate": 4.844903940185277e-06,
"loss": 1.6021,
"step": 1260
},
{
"epoch": 2.032,
"grad_norm": 0.5924295783042908,
"learning_rate": 4.840026000632423e-06,
"loss": 1.6385,
"step": 1270
},
{
"epoch": 2.048,
"grad_norm": 0.6063358783721924,
"learning_rate": 4.835075069250613e-06,
"loss": 1.57,
"step": 1280
},
{
"epoch": 2.064,
"grad_norm": 0.7654261589050293,
"learning_rate": 4.8300513004729735e-06,
"loss": 1.4305,
"step": 1290
},
{
"epoch": 2.08,
"grad_norm": 0.7088677883148193,
"learning_rate": 4.824954851004633e-06,
"loss": 1.5811,
"step": 1300
},
{
"epoch": 2.096,
"grad_norm": 1.081394076347351,
"learning_rate": 4.819785879817827e-06,
"loss": 1.5196,
"step": 1310
},
{
"epoch": 2.112,
"grad_norm": 0.7218335866928101,
"learning_rate": 4.814544548146945e-06,
"loss": 1.4357,
"step": 1320
},
{
"epoch": 2.128,
"grad_norm": 0.5773430466651917,
"learning_rate": 4.809231019483497e-06,
"loss": 1.6223,
"step": 1330
},
{
"epoch": 2.144,
"grad_norm": 0.8747012615203857,
"learning_rate": 4.803845459571014e-06,
"loss": 1.5586,
"step": 1340
},
{
"epoch": 2.16,
"grad_norm": 0.833926796913147,
"learning_rate": 4.798388036399883e-06,
"loss": 1.5551,
"step": 1350
},
{
"epoch": 2.176,
"grad_norm": 0.6463444828987122,
"learning_rate": 4.792858920202099e-06,
"loss": 1.5776,
"step": 1360
},
{
"epoch": 2.192,
"grad_norm": 0.5678787231445312,
"learning_rate": 4.787258283445962e-06,
"loss": 1.6474,
"step": 1370
},
{
"epoch": 2.208,
"grad_norm": 0.5224615335464478,
"learning_rate": 4.781586300830693e-06,
"loss": 1.4153,
"step": 1380
},
{
"epoch": 2.224,
"grad_norm": 0.6795189380645752,
"learning_rate": 4.775843149280986e-06,
"loss": 1.3767,
"step": 1390
},
{
"epoch": 2.24,
"grad_norm": 0.6834966540336609,
"learning_rate": 4.77002900794149e-06,
"loss": 1.3661,
"step": 1400
},
{
"epoch": 2.2560000000000002,
"grad_norm": 0.8537456393241882,
"learning_rate": 4.7641440581712194e-06,
"loss": 1.5483,
"step": 1410
},
{
"epoch": 2.2720000000000002,
"grad_norm": 0.3871168792247772,
"learning_rate": 4.758188483537898e-06,
"loss": 1.5351,
"step": 1420
},
{
"epoch": 2.288,
"grad_norm": 1.1125162839889526,
"learning_rate": 4.752162469812234e-06,
"loss": 1.4494,
"step": 1430
},
{
"epoch": 2.304,
"grad_norm": 1.2243694067001343,
"learning_rate": 4.7460662049621235e-06,
"loss": 1.4805,
"step": 1440
},
{
"epoch": 2.32,
"grad_norm": 0.8491847515106201,
"learning_rate": 4.739899879146785e-06,
"loss": 1.7852,
"step": 1450
},
{
"epoch": 2.336,
"grad_norm": 0.8607208728790283,
"learning_rate": 4.733663684710835e-06,
"loss": 1.4335,
"step": 1460
},
{
"epoch": 2.352,
"grad_norm": 1.0052345991134644,
"learning_rate": 4.727357816178282e-06,
"loss": 1.601,
"step": 1470
},
{
"epoch": 2.368,
"grad_norm": 0.6858043074607849,
"learning_rate": 4.72098247024646e-06,
"loss": 1.553,
"step": 1480
},
{
"epoch": 2.384,
"grad_norm": 0.7939265966415405,
"learning_rate": 4.714537845779894e-06,
"loss": 1.7166,
"step": 1490
},
{
"epoch": 2.4,
"grad_norm": 0.6755343079566956,
"learning_rate": 4.708024143804097e-06,
"loss": 1.3968,
"step": 1500
},
{
"epoch": 2.416,
"grad_norm": 0.7775924801826477,
"learning_rate": 4.701441567499301e-06,
"loss": 1.6513,
"step": 1510
},
{
"epoch": 2.432,
"grad_norm": 1.3676393032073975,
"learning_rate": 4.694790322194111e-06,
"loss": 1.2937,
"step": 1520
},
{
"epoch": 2.448,
"grad_norm": 0.7261756062507629,
"learning_rate": 4.688070615359114e-06,
"loss": 1.8156,
"step": 1530
},
{
"epoch": 2.464,
"grad_norm": 1.1282861232757568,
"learning_rate": 4.681282656600393e-06,
"loss": 1.6019,
"step": 1540
},
{
"epoch": 2.48,
"grad_norm": 1.0394994020462036,
"learning_rate": 4.674426657653003e-06,
"loss": 1.7059,
"step": 1550
},
{
"epoch": 2.496,
"grad_norm": 0.6906456351280212,
"learning_rate": 4.667502832374352e-06,
"loss": 1.6788,
"step": 1560
},
{
"epoch": 2.512,
"grad_norm": 0.4713647961616516,
"learning_rate": 4.660511396737541e-06,
"loss": 1.6816,
"step": 1570
},
{
"epoch": 2.528,
"grad_norm": 0.7387018203735352,
"learning_rate": 4.653452568824625e-06,
"loss": 1.8899,
"step": 1580
},
{
"epoch": 2.544,
"grad_norm": 0.7270022630691528,
"learning_rate": 4.646326568819805e-06,
"loss": 1.3103,
"step": 1590
},
{
"epoch": 2.56,
"grad_norm": 0.787842333316803,
"learning_rate": 4.639133619002565e-06,
"loss": 1.2932,
"step": 1600
},
{
"epoch": 2.576,
"grad_norm": 0.4514208137989044,
"learning_rate": 4.63187394374074e-06,
"loss": 1.516,
"step": 1610
},
{
"epoch": 2.592,
"grad_norm": 0.8605286478996277,
"learning_rate": 4.6245477694835104e-06,
"loss": 1.5961,
"step": 1620
},
{
"epoch": 2.608,
"grad_norm": 0.5745855569839478,
"learning_rate": 4.617155324754347e-06,
"loss": 1.4135,
"step": 1630
},
{
"epoch": 2.624,
"grad_norm": 0.6500321626663208,
"learning_rate": 4.609696840143875e-06,
"loss": 1.6172,
"step": 1640
},
{
"epoch": 2.64,
"grad_norm": 0.6863396167755127,
"learning_rate": 4.602172548302684e-06,
"loss": 1.5619,
"step": 1650
},
{
"epoch": 2.656,
"grad_norm": 0.7934362292289734,
"learning_rate": 4.594582683934078e-06,
"loss": 1.5107,
"step": 1660
},
{
"epoch": 2.672,
"grad_norm": 0.9190770983695984,
"learning_rate": 4.586927483786739e-06,
"loss": 1.5431,
"step": 1670
},
{
"epoch": 2.6879999999999997,
"grad_norm": 0.6435351967811584,
"learning_rate": 4.579207186647357e-06,
"loss": 1.6868,
"step": 1680
},
{
"epoch": 2.7039999999999997,
"grad_norm": 0.9924895763397217,
"learning_rate": 4.571422033333176e-06,
"loss": 1.5309,
"step": 1690
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.7778785228729248,
"learning_rate": 4.563572266684478e-06,
"loss": 1.5349,
"step": 1700
},
{
"epoch": 2.7359999999999998,
"grad_norm": 0.9731547236442566,
"learning_rate": 4.555658131557015e-06,
"loss": 1.5695,
"step": 1710
},
{
"epoch": 2.752,
"grad_norm": 0.7542728185653687,
"learning_rate": 4.547679874814368e-06,
"loss": 1.6187,
"step": 1720
},
{
"epoch": 2.768,
"grad_norm": 1.1981087923049927,
"learning_rate": 4.539637745320247e-06,
"loss": 1.2666,
"step": 1730
},
{
"epoch": 2.784,
"grad_norm": 0.5126321315765381,
"learning_rate": 4.5315319939307275e-06,
"loss": 1.699,
"step": 1740
},
{
"epoch": 2.8,
"grad_norm": 0.8448460698127747,
"learning_rate": 4.523362873486427e-06,
"loss": 1.5705,
"step": 1750
},
{
"epoch": 2.816,
"grad_norm": 0.7451621294021606,
"learning_rate": 4.5151306388046175e-06,
"loss": 1.7746,
"step": 1760
},
{
"epoch": 2.832,
"grad_norm": 0.9366666078567505,
"learning_rate": 4.5068355466712785e-06,
"loss": 1.644,
"step": 1770
},
{
"epoch": 2.848,
"grad_norm": 0.6129621267318726,
"learning_rate": 4.498477855833084e-06,
"loss": 1.5837,
"step": 1780
},
{
"epoch": 2.864,
"grad_norm": 0.7466766238212585,
"learning_rate": 4.490057826989333e-06,
"loss": 1.5324,
"step": 1790
},
{
"epoch": 2.88,
"grad_norm": 0.5993590950965881,
"learning_rate": 4.481575722783821e-06,
"loss": 1.4446,
"step": 1800
},
{
"epoch": 2.896,
"grad_norm": 0.9300441145896912,
"learning_rate": 4.47303180779664e-06,
"loss": 1.6557,
"step": 1810
},
{
"epoch": 2.912,
"grad_norm": 0.7721385359764099,
"learning_rate": 4.4644263485359316e-06,
"loss": 1.8196,
"step": 1820
},
{
"epoch": 2.928,
"grad_norm": 0.7181217670440674,
"learning_rate": 4.455759613429573e-06,
"loss": 1.5393,
"step": 1830
},
{
"epoch": 2.944,
"grad_norm": 0.9731667637825012,
"learning_rate": 4.4470318728168036e-06,
"loss": 1.5022,
"step": 1840
},
{
"epoch": 2.96,
"grad_norm": 0.4396623373031616,
"learning_rate": 4.43824339893979e-06,
"loss": 1.7437,
"step": 1850
},
{
"epoch": 2.976,
"grad_norm": 0.7844721078872681,
"learning_rate": 4.429394465935135e-06,
"loss": 1.8313,
"step": 1860
},
{
"epoch": 2.992,
"grad_norm": 1.1775206327438354,
"learning_rate": 4.420485349825332e-06,
"loss": 1.4683,
"step": 1870
},
{
"epoch": 3.008,
"grad_norm": 1.0284100770950317,
"learning_rate": 4.411516328510145e-06,
"loss": 1.4946,
"step": 1880
},
{
"epoch": 3.024,
"grad_norm": 0.8102648854255676,
"learning_rate": 4.40248768175795e-06,
"loss": 1.4036,
"step": 1890
},
{
"epoch": 3.04,
"grad_norm": 0.8291281461715698,
"learning_rate": 4.393399691197e-06,
"loss": 1.5588,
"step": 1900
},
{
"epoch": 3.056,
"grad_norm": 0.8688443899154663,
"learning_rate": 4.384252640306649e-06,
"loss": 1.343,
"step": 1910
},
{
"epoch": 3.072,
"grad_norm": 1.4661997556686401,
"learning_rate": 4.3750468144085e-06,
"loss": 1.6441,
"step": 1920
},
{
"epoch": 3.088,
"grad_norm": 0.5696677565574646,
"learning_rate": 4.36578250065751e-06,
"loss": 1.5555,
"step": 1930
},
{
"epoch": 3.104,
"grad_norm": 0.5522274971008301,
"learning_rate": 4.356459988033039e-06,
"loss": 1.6707,
"step": 1940
},
{
"epoch": 3.12,
"grad_norm": 0.5719353556632996,
"learning_rate": 4.347079567329821e-06,
"loss": 1.536,
"step": 1950
},
{
"epoch": 3.136,
"grad_norm": 0.9399852752685547,
"learning_rate": 4.337641531148906e-06,
"loss": 1.7248,
"step": 1960
},
{
"epoch": 3.152,
"grad_norm": 0.7337310910224915,
"learning_rate": 4.328146173888528e-06,
"loss": 1.4025,
"step": 1970
},
{
"epoch": 3.168,
"grad_norm": 0.48429638147354126,
"learning_rate": 4.318593791734924e-06,
"loss": 1.5781,
"step": 1980
},
{
"epoch": 3.184,
"grad_norm": 1.1470130681991577,
"learning_rate": 4.308984682653092e-06,
"loss": 1.4868,
"step": 1990
},
{
"epoch": 3.2,
"grad_norm": 0.8633829355239868,
"learning_rate": 4.2993191463775e-06,
"loss": 1.6465,
"step": 2000
},
{
"epoch": 3.216,
"grad_norm": 0.6459351181983948,
"learning_rate": 4.289597484402732e-06,
"loss": 1.4083,
"step": 2010
},
{
"epoch": 3.232,
"grad_norm": 0.8190066814422607,
"learning_rate": 4.279819999974091e-06,
"loss": 1.5047,
"step": 2020
},
{
"epoch": 3.248,
"grad_norm": 0.7891396284103394,
"learning_rate": 4.269986998078132e-06,
"loss": 1.4592,
"step": 2030
},
{
"epoch": 3.2640000000000002,
"grad_norm": 0.760246753692627,
"learning_rate": 4.260098785433153e-06,
"loss": 1.467,
"step": 2040
},
{
"epoch": 3.2800000000000002,
"grad_norm": 1.1161779165267944,
"learning_rate": 4.250155670479628e-06,
"loss": 1.4729,
"step": 2050
},
{
"epoch": 3.296,
"grad_norm": 0.8143382668495178,
"learning_rate": 4.240157963370583e-06,
"loss": 1.6712,
"step": 2060
},
{
"epoch": 3.312,
"grad_norm": 1.0454487800598145,
"learning_rate": 4.230105975961921e-06,
"loss": 1.4567,
"step": 2070
},
{
"epoch": 3.328,
"grad_norm": 1.1491258144378662,
"learning_rate": 4.220000021802703e-06,
"loss": 1.6005,
"step": 2080
},
{
"epoch": 3.344,
"grad_norm": 0.9478589296340942,
"learning_rate": 4.209840416125353e-06,
"loss": 1.56,
"step": 2090
},
{
"epoch": 3.36,
"grad_norm": 0.8324580192565918,
"learning_rate": 4.1996274758358405e-06,
"loss": 1.5804,
"step": 2100
},
{
"epoch": 3.376,
"grad_norm": 0.8769364953041077,
"learning_rate": 4.18936151950378e-06,
"loss": 1.2801,
"step": 2110
},
{
"epoch": 3.392,
"grad_norm": 0.607882022857666,
"learning_rate": 4.1790428673525104e-06,
"loss": 1.604,
"step": 2120
},
{
"epoch": 3.408,
"grad_norm": 0.893365204334259,
"learning_rate": 4.168671841249091e-06,
"loss": 1.4616,
"step": 2130
},
{
"epoch": 3.424,
"grad_norm": 0.5159714818000793,
"learning_rate": 4.15824876469427e-06,
"loss": 1.5749,
"step": 2140
},
{
"epoch": 3.44,
"grad_norm": 0.4347703456878662,
"learning_rate": 4.147773962812393e-06,
"loss": 1.5695,
"step": 2150
},
{
"epoch": 3.456,
"grad_norm": 0.6824937462806702,
"learning_rate": 4.137247762341262e-06,
"loss": 1.6246,
"step": 2160
},
{
"epoch": 3.472,
"grad_norm": 0.9752215147018433,
"learning_rate": 4.126670491621938e-06,
"loss": 1.7512,
"step": 2170
},
{
"epoch": 3.488,
"grad_norm": 1.5154378414154053,
"learning_rate": 4.116042480588505e-06,
"loss": 1.5603,
"step": 2180
},
{
"epoch": 3.504,
"grad_norm": 1.0059340000152588,
"learning_rate": 4.105364060757776e-06,
"loss": 1.5112,
"step": 2190
},
{
"epoch": 3.52,
"grad_norm": 0.5788377523422241,
"learning_rate": 4.094635565218955e-06,
"loss": 1.4864,
"step": 2200
},
{
"epoch": 3.536,
"grad_norm": 1.1504844427108765,
"learning_rate": 4.083857328623243e-06,
"loss": 1.3071,
"step": 2210
},
{
"epoch": 3.552,
"grad_norm": 0.8494038581848145,
"learning_rate": 4.073029687173399e-06,
"loss": 1.513,
"step": 2220
},
{
"epoch": 3.568,
"grad_norm": 0.8015080094337463,
"learning_rate": 4.062152978613259e-06,
"loss": 1.424,
"step": 2230
},
{
"epoch": 3.584,
"grad_norm": 0.6550512909889221,
"learning_rate": 4.051227542217192e-06,
"loss": 1.7912,
"step": 2240
},
{
"epoch": 3.6,
"grad_norm": 0.837598979473114,
"learning_rate": 4.040253718779528e-06,
"loss": 1.6095,
"step": 2250
},
{
"epoch": 3.616,
"grad_norm": 0.8367552757263184,
"learning_rate": 4.029231850603914e-06,
"loss": 1.6545,
"step": 2260
},
{
"epoch": 3.632,
"grad_norm": 0.5127103924751282,
"learning_rate": 4.018162281492651e-06,
"loss": 1.4162,
"step": 2270
},
{
"epoch": 3.648,
"grad_norm": 0.6145904064178467,
"learning_rate": 4.007045356735959e-06,
"loss": 1.671,
"step": 2280
},
{
"epoch": 3.664,
"grad_norm": 1.1353576183319092,
"learning_rate": 3.995881423101211e-06,
"loss": 1.5599,
"step": 2290
},
{
"epoch": 3.68,
"grad_norm": 0.9976987838745117,
"learning_rate": 3.984670828822118e-06,
"loss": 1.6871,
"step": 2300
},
{
"epoch": 3.6959999999999997,
"grad_norm": 0.7358677983283997,
"learning_rate": 3.973413923587862e-06,
"loss": 1.7595,
"step": 2310
},
{
"epoch": 3.7119999999999997,
"grad_norm": 0.7533473372459412,
"learning_rate": 3.962111058532192e-06,
"loss": 1.461,
"step": 2320
},
{
"epoch": 3.7279999999999998,
"grad_norm": 0.8930823802947998,
"learning_rate": 3.950762586222469e-06,
"loss": 1.6558,
"step": 2330
},
{
"epoch": 3.7439999999999998,
"grad_norm": 1.2027279138565063,
"learning_rate": 3.93936886064867e-06,
"loss": 1.4123,
"step": 2340
},
{
"epoch": 3.76,
"grad_norm": 0.6071750521659851,
"learning_rate": 3.927930237212345e-06,
"loss": 1.5771,
"step": 2350
},
{
"epoch": 3.776,
"grad_norm": 0.6392596960067749,
"learning_rate": 3.916447072715531e-06,
"loss": 1.5848,
"step": 2360
},
{
"epoch": 3.792,
"grad_norm": 0.8910986185073853,
"learning_rate": 3.904919725349626e-06,
"loss": 1.4619,
"step": 2370
},
{
"epoch": 3.808,
"grad_norm": 0.7371443510055542,
"learning_rate": 3.893348554684209e-06,
"loss": 1.6317,
"step": 2380
},
{
"epoch": 3.824,
"grad_norm": 1.6092065572738647,
"learning_rate": 3.881733921655829e-06,
"loss": 1.3639,
"step": 2390
},
{
"epoch": 3.84,
"grad_norm": 1.204833984375,
"learning_rate": 3.870076188556746e-06,
"loss": 1.59,
"step": 2400
},
{
"epoch": 3.856,
"grad_norm": 0.7493409514427185,
"learning_rate": 3.8583757190236296e-06,
"loss": 1.4722,
"step": 2410
},
{
"epoch": 3.872,
"grad_norm": 0.6909350752830505,
"learning_rate": 3.846632878026214e-06,
"loss": 1.6397,
"step": 2420
},
{
"epoch": 3.888,
"grad_norm": 1.029907464981079,
"learning_rate": 3.834848031855919e-06,
"loss": 1.4298,
"step": 2430
},
{
"epoch": 3.904,
"grad_norm": 1.067639708518982,
"learning_rate": 3.8230215481144175e-06,
"loss": 1.5334,
"step": 2440
},
{
"epoch": 3.92,
"grad_norm": 0.6574661731719971,
"learning_rate": 3.8111537957021736e-06,
"loss": 1.6364,
"step": 2450
},
{
"epoch": 3.936,
"grad_norm": 0.8869898915290833,
"learning_rate": 3.7992451448069368e-06,
"loss": 1.6119,
"step": 2460
},
{
"epoch": 3.952,
"grad_norm": 0.8028042316436768,
"learning_rate": 3.787295966892189e-06,
"loss": 1.5989,
"step": 2470
},
{
"epoch": 3.968,
"grad_norm": 0.9179840683937073,
"learning_rate": 3.775306634685562e-06,
"loss": 1.5097,
"step": 2480
},
{
"epoch": 3.984,
"grad_norm": 1.3384634256362915,
"learning_rate": 3.763277522167212e-06,
"loss": 1.446,
"step": 2490
},
{
"epoch": 4.0,
"grad_norm": 0.8146370053291321,
"learning_rate": 3.7512090045581488e-06,
"loss": 1.5723,
"step": 2500
},
{
"epoch": 4.016,
"grad_norm": 0.6829802393913269,
"learning_rate": 3.7391014583085384e-06,
"loss": 1.5526,
"step": 2510
},
{
"epoch": 4.032,
"grad_norm": 0.6764781475067139,
"learning_rate": 3.726955261085956e-06,
"loss": 1.5369,
"step": 2520
},
{
"epoch": 4.048,
"grad_norm": 1.2887903451919556,
"learning_rate": 3.714770791763605e-06,
"loss": 1.5666,
"step": 2530
},
{
"epoch": 4.064,
"grad_norm": 0.9172888398170471,
"learning_rate": 3.7025484304085035e-06,
"loss": 1.3176,
"step": 2540
},
{
"epoch": 4.08,
"grad_norm": 1.1424694061279297,
"learning_rate": 3.6902885582696236e-06,
"loss": 1.6824,
"step": 2550
},
{
"epoch": 4.096,
"grad_norm": 0.5998602509498596,
"learning_rate": 3.677991557766002e-06,
"loss": 1.5053,
"step": 2560
},
{
"epoch": 4.112,
"grad_norm": 0.7892935872077942,
"learning_rate": 3.665657812474812e-06,
"loss": 1.4574,
"step": 2570
},
{
"epoch": 4.128,
"grad_norm": 0.6210141181945801,
"learning_rate": 3.653287707119397e-06,
"loss": 1.5496,
"step": 2580
},
{
"epoch": 4.144,
"grad_norm": 0.8353777527809143,
"learning_rate": 3.640881627557271e-06,
"loss": 1.6421,
"step": 2590
},
{
"epoch": 4.16,
"grad_norm": 0.8043954372406006,
"learning_rate": 3.628439960768082e-06,
"loss": 1.5182,
"step": 2600
},
{
"epoch": 4.176,
"grad_norm": 0.6764125823974609,
"learning_rate": 3.61596309484154e-06,
"loss": 1.6664,
"step": 2610
},
{
"epoch": 4.192,
"grad_norm": 0.7379188537597656,
"learning_rate": 3.603451418965313e-06,
"loss": 1.4409,
"step": 2620
},
{
"epoch": 4.208,
"grad_norm": 1.0923508405685425,
"learning_rate": 3.5909053234128893e-06,
"loss": 1.3932,
"step": 2630
},
{
"epoch": 4.224,
"grad_norm": 1.2085895538330078,
"learning_rate": 3.5783251995313985e-06,
"loss": 1.4844,
"step": 2640
},
{
"epoch": 4.24,
"grad_norm": 0.7708174586296082,
"learning_rate": 3.5657114397294077e-06,
"loss": 1.7488,
"step": 2650
},
{
"epoch": 4.256,
"grad_norm": 0.6567518711090088,
"learning_rate": 3.553064437464682e-06,
"loss": 1.4351,
"step": 2660
},
{
"epoch": 4.272,
"grad_norm": 1.4016966819763184,
"learning_rate": 3.540384587231906e-06,
"loss": 1.6682,
"step": 2670
},
{
"epoch": 4.288,
"grad_norm": 0.9759495258331299,
"learning_rate": 3.5276722845503893e-06,
"loss": 1.5559,
"step": 2680
},
{
"epoch": 4.304,
"grad_norm": 1.0950008630752563,
"learning_rate": 3.5149279259517165e-06,
"loss": 1.6631,
"step": 2690
},
{
"epoch": 4.32,
"grad_norm": 0.7884267568588257,
"learning_rate": 3.5021519089673876e-06,
"loss": 1.5031,
"step": 2700
},
{
"epoch": 4.336,
"grad_norm": 0.89117032289505,
"learning_rate": 3.489344632116412e-06,
"loss": 1.734,
"step": 2710
},
{
"epoch": 4.352,
"grad_norm": 0.88322913646698,
"learning_rate": 3.4765064948928813e-06,
"loss": 1.6414,
"step": 2720
},
{
"epoch": 4.368,
"grad_norm": 1.0615278482437134,
"learning_rate": 3.4636378977535073e-06,
"loss": 1.4443,
"step": 2730
},
{
"epoch": 4.384,
"grad_norm": 0.8732738494873047,
"learning_rate": 3.45073924210513e-06,
"loss": 1.4998,
"step": 2740
},
{
"epoch": 4.4,
"grad_norm": 0.9869545102119446,
"learning_rate": 3.4378109302921946e-06,
"loss": 1.6065,
"step": 2750
},
{
"epoch": 4.416,
"grad_norm": 1.3117120265960693,
"learning_rate": 3.4248533655842064e-06,
"loss": 1.4688,
"step": 2760
},
{
"epoch": 4.432,
"grad_norm": 0.8496220111846924,
"learning_rate": 3.4118669521631465e-06,
"loss": 1.5675,
"step": 2770
},
{
"epoch": 4.448,
"grad_norm": 0.642301619052887,
"learning_rate": 3.3988520951108683e-06,
"loss": 1.4467,
"step": 2780
},
{
"epoch": 4.464,
"grad_norm": 1.203744649887085,
"learning_rate": 3.3858092003964595e-06,
"loss": 1.5239,
"step": 2790
},
{
"epoch": 4.48,
"grad_norm": 0.8205372095108032,
"learning_rate": 3.372738674863577e-06,
"loss": 1.5656,
"step": 2800
},
{
"epoch": 4.496,
"grad_norm": 1.3869353532791138,
"learning_rate": 3.3596409262177633e-06,
"loss": 1.585,
"step": 2810
},
{
"epoch": 4.5120000000000005,
"grad_norm": 1.4306498765945435,
"learning_rate": 3.346516363013719e-06,
"loss": 1.3659,
"step": 2820
},
{
"epoch": 4.5280000000000005,
"grad_norm": 1.2702066898345947,
"learning_rate": 3.33336539464257e-06,
"loss": 1.5395,
"step": 2830
},
{
"epoch": 4.5440000000000005,
"grad_norm": 0.893438994884491,
"learning_rate": 3.320188431319088e-06,
"loss": 1.5555,
"step": 2840
},
{
"epoch": 4.5600000000000005,
"grad_norm": 0.5918306112289429,
"learning_rate": 3.3069858840688997e-06,
"loss": 1.5422,
"step": 2850
},
{
"epoch": 4.576,
"grad_norm": 0.8166379332542419,
"learning_rate": 3.293758164715663e-06,
"loss": 1.3793,
"step": 2860
},
{
"epoch": 4.592,
"grad_norm": 0.8559311032295227,
"learning_rate": 3.280505685868226e-06,
"loss": 1.6722,
"step": 2870
},
{
"epoch": 4.608,
"grad_norm": 1.6434208154678345,
"learning_rate": 3.267228860907751e-06,
"loss": 1.6425,
"step": 2880
},
{
"epoch": 4.624,
"grad_norm": 1.1215018033981323,
"learning_rate": 3.2539281039748234e-06,
"loss": 1.6401,
"step": 2890
},
{
"epoch": 4.64,
"grad_norm": 1.3066354990005493,
"learning_rate": 3.240603829956531e-06,
"loss": 1.5596,
"step": 2900
},
{
"epoch": 4.656,
"grad_norm": 1.1937469244003296,
"learning_rate": 3.227256454473526e-06,
"loss": 1.4965,
"step": 2910
},
{
"epoch": 4.672,
"grad_norm": 1.349308729171753,
"learning_rate": 3.213886393867057e-06,
"loss": 1.61,
"step": 2920
},
{
"epoch": 4.688,
"grad_norm": 0.5878828763961792,
"learning_rate": 3.2004940651859844e-06,
"loss": 1.5297,
"step": 2930
},
{
"epoch": 4.704,
"grad_norm": 1.5151363611221313,
"learning_rate": 3.1870798861737705e-06,
"loss": 1.4417,
"step": 2940
},
{
"epoch": 4.72,
"grad_norm": 0.7873701453208923,
"learning_rate": 3.1736442752554514e-06,
"loss": 1.4677,
"step": 2950
},
{
"epoch": 4.736,
"grad_norm": 1.0918796062469482,
"learning_rate": 3.16018765152458e-06,
"loss": 1.4907,
"step": 2960
},
{
"epoch": 4.752,
"grad_norm": 0.8090459108352661,
"learning_rate": 3.146710434730159e-06,
"loss": 1.597,
"step": 2970
},
{
"epoch": 4.768,
"grad_norm": 0.4874608814716339,
"learning_rate": 3.1332130452635434e-06,
"loss": 1.4013,
"step": 2980
},
{
"epoch": 4.784,
"grad_norm": 1.1235930919647217,
"learning_rate": 3.11969590414533e-06,
"loss": 1.4681,
"step": 2990
},
{
"epoch": 4.8,
"grad_norm": 1.1741480827331543,
"learning_rate": 3.1061594330122245e-06,
"loss": 1.4493,
"step": 3000
},
{
"epoch": 4.816,
"grad_norm": 1.1874006986618042,
"learning_rate": 3.092604054103888e-06,
"loss": 1.4582,
"step": 3010
},
{
"epoch": 4.832,
"grad_norm": 0.9722936749458313,
"learning_rate": 3.0790301902497664e-06,
"loss": 1.6464,
"step": 3020
},
{
"epoch": 4.848,
"grad_norm": 0.8346335887908936,
"learning_rate": 3.065438264855903e-06,
"loss": 1.4819,
"step": 3030
},
{
"epoch": 4.864,
"grad_norm": 1.137216567993164,
"learning_rate": 3.051828701891729e-06,
"loss": 1.3768,
"step": 3040
},
{
"epoch": 4.88,
"grad_norm": 1.2007477283477783,
"learning_rate": 3.03820192587684e-06,
"loss": 1.464,
"step": 3050
},
{
"epoch": 4.896,
"grad_norm": 0.9882920384407043,
"learning_rate": 3.024558361867756e-06,
"loss": 1.3728,
"step": 3060
},
{
"epoch": 4.912,
"grad_norm": 1.1882582902908325,
"learning_rate": 3.0108984354446556e-06,
"loss": 1.557,
"step": 3070
},
{
"epoch": 4.928,
"grad_norm": 0.5336953997612,
"learning_rate": 2.9972225726981114e-06,
"loss": 1.4592,
"step": 3080
},
{
"epoch": 4.944,
"grad_norm": 0.9478628039360046,
"learning_rate": 2.9835312002157914e-06,
"loss": 1.633,
"step": 3090
},
{
"epoch": 4.96,
"grad_norm": 0.8075321912765503,
"learning_rate": 2.9698247450691524e-06,
"loss": 1.622,
"step": 3100
},
{
"epoch": 4.976,
"grad_norm": 0.8833026885986328,
"learning_rate": 2.956103634800126e-06,
"loss": 1.4935,
"step": 3110
},
{
"epoch": 4.992,
"grad_norm": 0.8639602661132812,
"learning_rate": 2.9423682974077723e-06,
"loss": 1.4142,
"step": 3120
},
{
"epoch": 5.008,
"grad_norm": 0.6490864753723145,
"learning_rate": 2.9286191613349373e-06,
"loss": 1.6463,
"step": 3130
},
{
"epoch": 5.024,
"grad_norm": 0.9260121583938599,
"learning_rate": 2.9148566554548857e-06,
"loss": 1.4281,
"step": 3140
},
{
"epoch": 5.04,
"grad_norm": 1.3701825141906738,
"learning_rate": 2.901081209057921e-06,
"loss": 1.7251,
"step": 3150
},
{
"epoch": 5.056,
"grad_norm": 0.5848241448402405,
"learning_rate": 2.887293251838e-06,
"loss": 1.7117,
"step": 3160
},
{
"epoch": 5.072,
"grad_norm": 1.0574699640274048,
"learning_rate": 2.8734932138793226e-06,
"loss": 1.5759,
"step": 3170
},
{
"epoch": 5.088,
"grad_norm": 0.6507989764213562,
"learning_rate": 2.8596815256429235e-06,
"loss": 1.2764,
"step": 3180
},
{
"epoch": 5.104,
"grad_norm": 0.8501136898994446,
"learning_rate": 2.8458586179532393e-06,
"loss": 1.5876,
"step": 3190
},
{
"epoch": 5.12,
"grad_norm": 0.9756034016609192,
"learning_rate": 2.832024921984674e-06,
"loss": 1.5839,
"step": 3200
},
{
"epoch": 5.136,
"grad_norm": 0.8798894882202148,
"learning_rate": 2.8181808692481454e-06,
"loss": 1.2936,
"step": 3210
},
{
"epoch": 5.152,
"grad_norm": 0.9644187092781067,
"learning_rate": 2.80432689157763e-06,
"loss": 1.6215,
"step": 3220
},
{
"epoch": 5.168,
"grad_norm": 1.244094967842102,
"learning_rate": 2.7904634211166877e-06,
"loss": 1.5676,
"step": 3230
},
{
"epoch": 5.184,
"grad_norm": 0.9972564578056335,
"learning_rate": 2.776590890304985e-06,
"loss": 1.6059,
"step": 3240
},
{
"epoch": 5.2,
"grad_norm": 0.5297699570655823,
"learning_rate": 2.7627097318648076e-06,
"loss": 1.4291,
"step": 3250
},
{
"epoch": 5.216,
"grad_norm": 1.0355069637298584,
"learning_rate": 2.748820378787558e-06,
"loss": 1.5404,
"step": 3260
},
{
"epoch": 5.232,
"grad_norm": 1.3221070766448975,
"learning_rate": 2.734923264320254e-06,
"loss": 1.5406,
"step": 3270
},
{
"epoch": 5.248,
"grad_norm": 0.9608879089355469,
"learning_rate": 2.721018821952011e-06,
"loss": 1.3568,
"step": 3280
},
{
"epoch": 5.264,
"grad_norm": 1.1379098892211914,
"learning_rate": 2.7071074854005206e-06,
"loss": 1.8331,
"step": 3290
},
{
"epoch": 5.28,
"grad_norm": 0.9943151473999023,
"learning_rate": 2.693189688598528e-06,
"loss": 1.374,
"step": 3300
},
{
"epoch": 5.296,
"grad_norm": 1.6240702867507935,
"learning_rate": 2.6792658656802857e-06,
"loss": 1.5203,
"step": 3310
},
{
"epoch": 5.312,
"grad_norm": 1.1698226928710938,
"learning_rate": 2.665336450968019e-06,
"loss": 1.5716,
"step": 3320
},
{
"epoch": 5.328,
"grad_norm": 0.7153477668762207,
"learning_rate": 2.6514018789583784e-06,
"loss": 1.458,
"step": 3330
},
{
"epoch": 5.344,
"grad_norm": 1.190158486366272,
"learning_rate": 2.6374625843088815e-06,
"loss": 1.4229,
"step": 3340
},
{
"epoch": 5.36,
"grad_norm": 1.0464446544647217,
"learning_rate": 2.6235190018243623e-06,
"loss": 1.5747,
"step": 3350
},
{
"epoch": 5.376,
"grad_norm": 0.8236386179924011,
"learning_rate": 2.6095715664433997e-06,
"loss": 1.3094,
"step": 3360
},
{
"epoch": 5.392,
"grad_norm": 0.6790027022361755,
"learning_rate": 2.5956207132247566e-06,
"loss": 1.3048,
"step": 3370
},
{
"epoch": 5.408,
"grad_norm": 1.0304784774780273,
"learning_rate": 2.58166687733381e-06,
"loss": 1.4594,
"step": 3380
},
{
"epoch": 5.424,
"grad_norm": 1.907761573791504,
"learning_rate": 2.5677104940289704e-06,
"loss": 1.5877,
"step": 3390
},
{
"epoch": 5.44,
"grad_norm": 1.12175714969635,
"learning_rate": 2.5537519986481125e-06,
"loss": 1.5274,
"step": 3400
},
{
"epoch": 5.456,
"grad_norm": 1.022863507270813,
"learning_rate": 2.539791826594991e-06,
"loss": 1.5628,
"step": 3410
},
{
"epoch": 5.4719999999999995,
"grad_norm": 1.6051089763641357,
"learning_rate": 2.525830413325661e-06,
"loss": 1.4258,
"step": 3420
},
{
"epoch": 5.4879999999999995,
"grad_norm": 0.7874528169631958,
"learning_rate": 2.5118681943348945e-06,
"loss": 1.4244,
"step": 3430
},
{
"epoch": 5.504,
"grad_norm": 1.5209931135177612,
"learning_rate": 2.497905605142595e-06,
"loss": 1.4659,
"step": 3440
},
{
"epoch": 5.52,
"grad_norm": 1.3238812685012817,
"learning_rate": 2.483943081280216e-06,
"loss": 1.5904,
"step": 3450
},
{
"epoch": 5.536,
"grad_norm": 1.0516330003738403,
"learning_rate": 2.4699810582771714e-06,
"loss": 1.3692,
"step": 3460
},
{
"epoch": 5.552,
"grad_norm": 0.753239095211029,
"learning_rate": 2.456019971647251e-06,
"loss": 1.5688,
"step": 3470
},
{
"epoch": 5.568,
"grad_norm": 1.348121166229248,
"learning_rate": 2.4420602568750394e-06,
"loss": 1.6268,
"step": 3480
},
{
"epoch": 5.584,
"grad_norm": 0.5859225988388062,
"learning_rate": 2.428102349402328e-06,
"loss": 1.4437,
"step": 3490
},
{
"epoch": 5.6,
"grad_norm": 2.787992000579834,
"learning_rate": 2.4141466846145332e-06,
"loss": 1.366,
"step": 3500
},
{
"epoch": 5.616,
"grad_norm": 0.7856543064117432,
"learning_rate": 2.4001936978271144e-06,
"loss": 1.5804,
"step": 3510
},
{
"epoch": 5.632,
"grad_norm": 0.8130792379379272,
"learning_rate": 2.386243824272e-06,
"loss": 1.459,
"step": 3520
},
{
"epoch": 5.648,
"grad_norm": 1.173449993133545,
"learning_rate": 2.372297499084006e-06,
"loss": 1.2318,
"step": 3530
},
{
"epoch": 5.664,
"grad_norm": 1.3133184909820557,
"learning_rate": 2.3583551572872657e-06,
"loss": 1.3609,
"step": 3540
},
{
"epoch": 5.68,
"grad_norm": 1.0046403408050537,
"learning_rate": 2.344417233781659e-06,
"loss": 1.5925,
"step": 3550
},
{
"epoch": 5.696,
"grad_norm": 1.0500761270523071,
"learning_rate": 2.3304841633292487e-06,
"loss": 1.4706,
"step": 3560
},
{
"epoch": 5.712,
"grad_norm": 1.2255127429962158,
"learning_rate": 2.316556380540715e-06,
"loss": 1.4721,
"step": 3570
},
{
"epoch": 5.728,
"grad_norm": 1.082068920135498,
"learning_rate": 2.3026343198618042e-06,
"loss": 1.7957,
"step": 3580
},
{
"epoch": 5.744,
"grad_norm": 1.0235259532928467,
"learning_rate": 2.2887184155597725e-06,
"loss": 1.6842,
"step": 3590
},
{
"epoch": 5.76,
"grad_norm": 0.9329107403755188,
"learning_rate": 2.274809101709842e-06,
"loss": 1.2767,
"step": 3600
},
{
"epoch": 5.776,
"grad_norm": 0.8458839058876038,
"learning_rate": 2.2609068121816614e-06,
"loss": 1.1407,
"step": 3610
},
{
"epoch": 5.792,
"grad_norm": 1.359635829925537,
"learning_rate": 2.247011980625771e-06,
"loss": 1.6127,
"step": 3620
},
{
"epoch": 5.808,
"grad_norm": 0.6604207158088684,
"learning_rate": 2.2331250404600757e-06,
"loss": 1.6394,
"step": 3630
},
{
"epoch": 5.824,
"grad_norm": 0.9887331128120422,
"learning_rate": 2.2192464248563267e-06,
"loss": 1.6715,
"step": 3640
},
{
"epoch": 5.84,
"grad_norm": 1.3716261386871338,
"learning_rate": 2.205376566726611e-06,
"loss": 1.7019,
"step": 3650
},
{
"epoch": 5.856,
"grad_norm": 0.9047254920005798,
"learning_rate": 2.1915158987098432e-06,
"loss": 1.7054,
"step": 3660
},
{
"epoch": 5.872,
"grad_norm": 1.3385180234909058,
"learning_rate": 2.177664853158276e-06,
"loss": 1.6999,
"step": 3670
},
{
"epoch": 5.888,
"grad_norm": 1.3120460510253906,
"learning_rate": 2.163823862124007e-06,
"loss": 1.5676,
"step": 3680
},
{
"epoch": 5.904,
"grad_norm": 0.927678108215332,
"learning_rate": 2.149993357345511e-06,
"loss": 1.6347,
"step": 3690
},
{
"epoch": 5.92,
"grad_norm": 1.3440815210342407,
"learning_rate": 2.1361737702341634e-06,
"loss": 1.6197,
"step": 3700
},
{
"epoch": 5.936,
"grad_norm": 0.873028576374054,
"learning_rate": 2.1223655318607907e-06,
"loss": 1.4831,
"step": 3710
},
{
"epoch": 5.952,
"grad_norm": 0.7099441289901733,
"learning_rate": 2.1085690729422175e-06,
"loss": 1.5792,
"step": 3720
},
{
"epoch": 5.968,
"grad_norm": 0.8678792715072632,
"learning_rate": 2.0947848238278387e-06,
"loss": 1.6523,
"step": 3730
},
{
"epoch": 5.984,
"grad_norm": 1.2380930185317993,
"learning_rate": 2.08101321448619e-06,
"loss": 1.6203,
"step": 3740
},
{
"epoch": 6.0,
"grad_norm": 1.4631417989730835,
"learning_rate": 2.0672546744915377e-06,
"loss": 1.3443,
"step": 3750
},
{
"epoch": 6.016,
"grad_norm": 0.8544397950172424,
"learning_rate": 2.0535096330104804e-06,
"loss": 1.2683,
"step": 3760
},
{
"epoch": 6.032,
"grad_norm": 1.2243201732635498,
"learning_rate": 2.03977851878856e-06,
"loss": 1.4203,
"step": 3770
},
{
"epoch": 6.048,
"grad_norm": 1.218111515045166,
"learning_rate": 2.0260617601368887e-06,
"loss": 1.4817,
"step": 3780
},
{
"epoch": 6.064,
"grad_norm": 0.7969552278518677,
"learning_rate": 2.012359784918792e-06,
"loss": 1.4691,
"step": 3790
},
{
"epoch": 6.08,
"grad_norm": 1.2019294500350952,
"learning_rate": 1.998673020536456e-06,
"loss": 1.4709,
"step": 3800
},
{
"epoch": 6.096,
"grad_norm": 0.8056063055992126,
"learning_rate": 1.9850018939176014e-06,
"loss": 1.6325,
"step": 3810
},
{
"epoch": 6.112,
"grad_norm": 1.419593095779419,
"learning_rate": 1.971346831502162e-06,
"loss": 1.3406,
"step": 3820
},
{
"epoch": 6.128,
"grad_norm": 1.7819221019744873,
"learning_rate": 1.957708259228987e-06,
"loss": 1.5992,
"step": 3830
},
{
"epoch": 6.144,
"grad_norm": 0.6604801416397095,
"learning_rate": 1.9440866025225525e-06,
"loss": 1.7421,
"step": 3840
},
{
"epoch": 6.16,
"grad_norm": 1.2192350625991821,
"learning_rate": 1.9304822862796906e-06,
"loss": 1.5088,
"step": 3850
},
{
"epoch": 6.176,
"grad_norm": 0.49949243664741516,
"learning_rate": 1.916895734856338e-06,
"loss": 1.6165,
"step": 3860
},
{
"epoch": 6.192,
"grad_norm": 3.4333205223083496,
"learning_rate": 1.9033273720542978e-06,
"loss": 1.5777,
"step": 3870
},
{
"epoch": 6.208,
"grad_norm": 0.515190064907074,
"learning_rate": 1.8897776211080182e-06,
"loss": 1.481,
"step": 3880
},
{
"epoch": 6.224,
"grad_norm": 0.7310590147972107,
"learning_rate": 1.8762469046713954e-06,
"loss": 1.4441,
"step": 3890
},
{
"epoch": 6.24,
"grad_norm": 0.8631799817085266,
"learning_rate": 1.862735644804587e-06,
"loss": 1.5025,
"step": 3900
},
{
"epoch": 6.256,
"grad_norm": 0.9323322176933289,
"learning_rate": 1.8492442629608433e-06,
"loss": 1.5654,
"step": 3910
},
{
"epoch": 6.272,
"grad_norm": 1.3464089632034302,
"learning_rate": 1.8357731799733686e-06,
"loss": 1.4824,
"step": 3920
},
{
"epoch": 6.288,
"grad_norm": 1.3562898635864258,
"learning_rate": 1.822322816042188e-06,
"loss": 1.7292,
"step": 3930
},
{
"epoch": 6.304,
"grad_norm": 1.107529640197754,
"learning_rate": 1.808893590721042e-06,
"loss": 1.4697,
"step": 3940
},
{
"epoch": 6.32,
"grad_norm": 0.9843621253967285,
"learning_rate": 1.7954859229043017e-06,
"loss": 1.4186,
"step": 3950
},
{
"epoch": 6.336,
"grad_norm": 1.065913200378418,
"learning_rate": 1.782100230813899e-06,
"loss": 1.5471,
"step": 3960
},
{
"epoch": 6.352,
"grad_norm": 1.0334281921386719,
"learning_rate": 1.768736931986284e-06,
"loss": 1.6078,
"step": 3970
},
{
"epoch": 6.368,
"grad_norm": 1.085191249847412,
"learning_rate": 1.7553964432593976e-06,
"loss": 1.4757,
"step": 3980
},
{
"epoch": 6.384,
"grad_norm": 1.6427229642868042,
"learning_rate": 1.7420791807596722e-06,
"loss": 1.6809,
"step": 3990
},
{
"epoch": 6.4,
"grad_norm": 1.117626428604126,
"learning_rate": 1.728785559889052e-06,
"loss": 1.5255,
"step": 4000
},
{
"epoch": 6.416,
"grad_norm": 0.9383686780929565,
"learning_rate": 1.7155159953120315e-06,
"loss": 1.3656,
"step": 4010
},
{
"epoch": 6.432,
"grad_norm": 1.320457935333252,
"learning_rate": 1.702270900942724e-06,
"loss": 1.4838,
"step": 4020
},
{
"epoch": 6.448,
"grad_norm": 1.8928338289260864,
"learning_rate": 1.689050689931951e-06,
"loss": 1.4088,
"step": 4030
},
{
"epoch": 6.464,
"grad_norm": 1.0204055309295654,
"learning_rate": 1.6758557746543518e-06,
"loss": 1.5274,
"step": 4040
},
{
"epoch": 6.48,
"grad_norm": 1.039493203163147,
"learning_rate": 1.662686566695525e-06,
"loss": 1.4647,
"step": 4050
},
{
"epoch": 6.496,
"grad_norm": 1.0225248336791992,
"learning_rate": 1.6495434768391854e-06,
"loss": 1.2431,
"step": 4060
},
{
"epoch": 6.5120000000000005,
"grad_norm": 0.5082935690879822,
"learning_rate": 1.6364269150543533e-06,
"loss": 1.6238,
"step": 4070
},
{
"epoch": 6.5280000000000005,
"grad_norm": 1.7832340002059937,
"learning_rate": 1.6233372904825656e-06,
"loss": 1.5233,
"step": 4080
},
{
"epoch": 6.5440000000000005,
"grad_norm": 1.5844208002090454,
"learning_rate": 1.6102750114251129e-06,
"loss": 1.5317,
"step": 4090
},
{
"epoch": 6.5600000000000005,
"grad_norm": 1.0076955556869507,
"learning_rate": 1.5972404853303061e-06,
"loss": 1.3785,
"step": 4100
},
{
"epoch": 6.576,
"grad_norm": 1.2678745985031128,
"learning_rate": 1.5842341187807642e-06,
"loss": 1.5139,
"step": 4110
},
{
"epoch": 6.592,
"grad_norm": 1.051327109336853,
"learning_rate": 1.5712563174807318e-06,
"loss": 1.5818,
"step": 4120
},
{
"epoch": 6.608,
"grad_norm": 0.5343210697174072,
"learning_rate": 1.5583074862434254e-06,
"loss": 1.4759,
"step": 4130
},
{
"epoch": 6.624,
"grad_norm": 0.9687218070030212,
"learning_rate": 1.5453880289784067e-06,
"loss": 1.5957,
"step": 4140
},
{
"epoch": 6.64,
"grad_norm": 0.853032648563385,
"learning_rate": 1.532498348678982e-06,
"loss": 1.6676,
"step": 4150
},
{
"epoch": 6.656,
"grad_norm": 1.3790019750595093,
"learning_rate": 1.519638847409632e-06,
"loss": 1.3662,
"step": 4160
},
{
"epoch": 6.672,
"grad_norm": 1.1166746616363525,
"learning_rate": 1.506809926293471e-06,
"loss": 1.4482,
"step": 4170
},
{
"epoch": 6.688,
"grad_norm": 1.1632003784179688,
"learning_rate": 1.4940119854997354e-06,
"loss": 1.5254,
"step": 4180
},
{
"epoch": 6.704,
"grad_norm": 1.0707969665527344,
"learning_rate": 1.481245424231298e-06,
"loss": 1.2793,
"step": 4190
},
{
"epoch": 6.72,
"grad_norm": 1.1206421852111816,
"learning_rate": 1.4685106407122218e-06,
"loss": 1.7606,
"step": 4200
},
{
"epoch": 6.736,
"grad_norm": 0.6367834210395813,
"learning_rate": 1.455808032175331e-06,
"loss": 1.4479,
"step": 4210
},
{
"epoch": 6.752,
"grad_norm": 1.693575143814087,
"learning_rate": 1.4431379948498254e-06,
"loss": 1.3899,
"step": 4220
},
{
"epoch": 6.768,
"grad_norm": 1.2394471168518066,
"learning_rate": 1.4305009239489193e-06,
"loss": 1.7582,
"step": 4230
},
{
"epoch": 6.784,
"grad_norm": 1.8645274639129639,
"learning_rate": 1.4178972136575152e-06,
"loss": 1.4129,
"step": 4240
},
{
"epoch": 6.8,
"grad_norm": 0.9838801622390747,
"learning_rate": 1.4053272571199037e-06,
"loss": 1.6496,
"step": 4250
},
{
"epoch": 6.816,
"grad_norm": 0.8779441118240356,
"learning_rate": 1.392791446427505e-06,
"loss": 1.7329,
"step": 4260
},
{
"epoch": 6.832,
"grad_norm": 1.0405304431915283,
"learning_rate": 1.3802901726066377e-06,
"loss": 1.3688,
"step": 4270
},
{
"epoch": 6.848,
"grad_norm": 1.0179954767227173,
"learning_rate": 1.3678238256063193e-06,
"loss": 1.4024,
"step": 4280
},
{
"epoch": 6.864,
"grad_norm": 1.1966112852096558,
"learning_rate": 1.3553927942861018e-06,
"loss": 1.4992,
"step": 4290
},
{
"epoch": 6.88,
"grad_norm": 2.0822248458862305,
"learning_rate": 1.3429974664039502e-06,
"loss": 1.4744,
"step": 4300
},
{
"epoch": 6.896,
"grad_norm": 1.085398554801941,
"learning_rate": 1.330638228604137e-06,
"loss": 1.658,
"step": 4310
},
{
"epoch": 6.912,
"grad_norm": 1.6239418983459473,
"learning_rate": 1.3183154664051855e-06,
"loss": 1.2885,
"step": 4320
},
{
"epoch": 6.928,
"grad_norm": 1.1970949172973633,
"learning_rate": 1.3060295641878473e-06,
"loss": 1.3509,
"step": 4330
},
{
"epoch": 6.944,
"grad_norm": 1.0090463161468506,
"learning_rate": 1.2937809051831102e-06,
"loss": 1.5188,
"step": 4340
},
{
"epoch": 6.96,
"grad_norm": 0.8989217877388,
"learning_rate": 1.2815698714602404e-06,
"loss": 1.3904,
"step": 4350
},
{
"epoch": 6.976,
"grad_norm": 0.9415572285652161,
"learning_rate": 1.2693968439148726e-06,
"loss": 1.551,
"step": 4360
},
{
"epoch": 6.992,
"grad_norm": 1.224729061126709,
"learning_rate": 1.257262202257124e-06,
"loss": 1.7128,
"step": 4370
},
{
"epoch": 7.008,
"grad_norm": 0.9762874245643616,
"learning_rate": 1.245166324999747e-06,
"loss": 1.5733,
"step": 4380
},
{
"epoch": 7.024,
"grad_norm": 1.44058358669281,
"learning_rate": 1.23310958944633e-06,
"loss": 1.6368,
"step": 4390
},
{
"epoch": 7.04,
"grad_norm": 0.8687445521354675,
"learning_rate": 1.2210923716795233e-06,
"loss": 1.5624,
"step": 4400
},
{
"epoch": 7.056,
"grad_norm": 0.7125824689865112,
"learning_rate": 1.2091150465493084e-06,
"loss": 1.5261,
"step": 4410
},
{
"epoch": 7.072,
"grad_norm": 1.331190824508667,
"learning_rate": 1.1971779876613062e-06,
"loss": 1.4998,
"step": 4420
},
{
"epoch": 7.088,
"grad_norm": 1.2545830011367798,
"learning_rate": 1.1852815673651246e-06,
"loss": 1.5449,
"step": 4430
},
{
"epoch": 7.104,
"grad_norm": 0.9906517863273621,
"learning_rate": 1.1734261567427385e-06,
"loss": 1.6022,
"step": 4440
},
{
"epoch": 7.12,
"grad_norm": 2.724125385284424,
"learning_rate": 1.1616121255969226e-06,
"loss": 1.5236,
"step": 4450
},
{
"epoch": 7.136,
"grad_norm": 1.118022084236145,
"learning_rate": 1.1498398424397106e-06,
"loss": 1.7033,
"step": 4460
},
{
"epoch": 7.152,
"grad_norm": 1.2464755773544312,
"learning_rate": 1.1381096744809029e-06,
"loss": 1.6036,
"step": 4470
},
{
"epoch": 7.168,
"grad_norm": 1.2054471969604492,
"learning_rate": 1.1264219876166086e-06,
"loss": 1.5169,
"step": 4480
},
{
"epoch": 7.184,
"grad_norm": 1.4373327493667603,
"learning_rate": 1.1147771464178378e-06,
"loss": 1.4206,
"step": 4490
},
{
"epoch": 7.2,
"grad_norm": 0.8014916181564331,
"learning_rate": 1.1031755141191259e-06,
"loss": 1.7048,
"step": 4500
},
{
"epoch": 7.216,
"grad_norm": 1.7281831502914429,
"learning_rate": 1.091617452607204e-06,
"loss": 1.3917,
"step": 4510
},
{
"epoch": 7.232,
"grad_norm": 1.2847610712051392,
"learning_rate": 1.080103322409711e-06,
"loss": 1.5944,
"step": 4520
},
{
"epoch": 7.248,
"grad_norm": 1.6191173791885376,
"learning_rate": 1.068633482683949e-06,
"loss": 1.7186,
"step": 4530
},
{
"epoch": 7.264,
"grad_norm": 0.8263424038887024,
"learning_rate": 1.057208291205675e-06,
"loss": 1.5196,
"step": 4540
},
{
"epoch": 7.28,
"grad_norm": 1.3846720457077026,
"learning_rate": 1.0458281043579482e-06,
"loss": 1.485,
"step": 4550
},
{
"epoch": 7.296,
"grad_norm": 1.7968668937683105,
"learning_rate": 1.0344932771200098e-06,
"loss": 1.4884,
"step": 4560
},
{
"epoch": 7.312,
"grad_norm": 1.0439798831939697,
"learning_rate": 1.0232041630562078e-06,
"loss": 1.5332,
"step": 4570
},
{
"epoch": 7.328,
"grad_norm": 1.4442821741104126,
"learning_rate": 1.0119611143049731e-06,
"loss": 1.7714,
"step": 4580
},
{
"epoch": 7.344,
"grad_norm": 1.526407241821289,
"learning_rate": 1.0007644815678327e-06,
"loss": 1.6068,
"step": 4590
},
{
"epoch": 7.36,
"grad_norm": 1.316821575164795,
"learning_rate": 9.89614614098471e-07,
"loss": 1.4775,
"step": 4600
},
{
"epoch": 7.376,
"grad_norm": 1.1869702339172363,
"learning_rate": 9.78511859691835e-07,
"loss": 1.5314,
"step": 4610
},
{
"epoch": 7.392,
"grad_norm": 0.7795867919921875,
"learning_rate": 9.674565646732866e-07,
"loss": 1.488,
"step": 4620
},
{
"epoch": 7.408,
"grad_norm": 1.4521294832229614,
"learning_rate": 9.56449073887798e-07,
"loss": 1.4497,
"step": 4630
},
{
"epoch": 7.424,
"grad_norm": 1.3386362791061401,
"learning_rate": 9.454897306891972e-07,
"loss": 1.6415,
"step": 4640
},
{
"epoch": 7.44,
"grad_norm": 1.7994210720062256,
"learning_rate": 9.34578876929457e-07,
"loss": 1.4092,
"step": 4650
},
{
"epoch": 7.456,
"grad_norm": 1.65791654586792,
"learning_rate": 9.237168529480323e-07,
"loss": 1.4772,
"step": 4660
},
{
"epoch": 7.4719999999999995,
"grad_norm": 1.3225324153900146,
"learning_rate": 9.129039975612408e-07,
"loss": 1.5664,
"step": 4670
},
{
"epoch": 7.4879999999999995,
"grad_norm": 0.5595719814300537,
"learning_rate": 9.021406480516992e-07,
"loss": 1.517,
"step": 4680
},
{
"epoch": 7.504,
"grad_norm": 0.99941486120224,
"learning_rate": 8.914271401578004e-07,
"loss": 1.252,
"step": 4690
},
{
"epoch": 7.52,
"grad_norm": 0.6022624969482422,
"learning_rate": 8.807638080632375e-07,
"loss": 1.4753,
"step": 4700
},
{
"epoch": 7.536,
"grad_norm": 0.983218252658844,
"learning_rate": 8.701509843865852e-07,
"loss": 1.4677,
"step": 4710
},
{
"epoch": 7.552,
"grad_norm": 1.096902847290039,
"learning_rate": 8.595890001709243e-07,
"loss": 1.3219,
"step": 4720
},
{
"epoch": 7.568,
"grad_norm": 0.9014173150062561,
"learning_rate": 8.49078184873508e-07,
"loss": 1.5366,
"step": 4730
},
{
"epoch": 7.584,
"grad_norm": 2.083513021469116,
"learning_rate": 8.386188663554948e-07,
"loss": 1.6013,
"step": 4740
},
{
"epoch": 7.6,
"grad_norm": 1.6989493370056152,
"learning_rate": 8.282113708717168e-07,
"loss": 1.6007,
"step": 4750
},
{
"epoch": 7.616,
"grad_norm": 1.1784847974777222,
"learning_rate": 8.178560230605012e-07,
"loss": 1.6399,
"step": 4760
},
{
"epoch": 7.632,
"grad_norm": 1.3991700410842896,
"learning_rate": 8.075531459335487e-07,
"loss": 1.3951,
"step": 4770
},
{
"epoch": 7.648,
"grad_norm": 1.9881689548492432,
"learning_rate": 7.973030608658541e-07,
"loss": 1.6573,
"step": 4780
},
{
"epoch": 7.664,
"grad_norm": 1.1137332916259766,
"learning_rate": 7.871060875856854e-07,
"loss": 1.6251,
"step": 4790
},
{
"epoch": 7.68,
"grad_norm": 1.2128005027770996,
"learning_rate": 7.769625441646039e-07,
"loss": 1.531,
"step": 4800
},
{
"epoch": 7.696,
"grad_norm": 0.7743262648582458,
"learning_rate": 7.66872747007551e-07,
"loss": 1.5373,
"step": 4810
},
{
"epoch": 7.712,
"grad_norm": 1.0655006170272827,
"learning_rate": 7.568370108429732e-07,
"loss": 1.5439,
"step": 4820
},
{
"epoch": 7.728,
"grad_norm": 1.0778528451919556,
"learning_rate": 7.468556487130065e-07,
"loss": 1.5511,
"step": 4830
},
{
"epoch": 7.744,
"grad_norm": 1.0320682525634766,
"learning_rate": 7.369289719637118e-07,
"loss": 1.5978,
"step": 4840
},
{
"epoch": 7.76,
"grad_norm": 0.6691346168518066,
"learning_rate": 7.270572902353634e-07,
"loss": 1.4303,
"step": 4850
},
{
"epoch": 7.776,
"grad_norm": 1.355306625366211,
"learning_rate": 7.172409114527881e-07,
"loss": 1.6586,
"step": 4860
},
{
"epoch": 7.792,
"grad_norm": 0.8070526123046875,
"learning_rate": 7.074801418157643e-07,
"loss": 1.4508,
"step": 4870
},
{
"epoch": 7.808,
"grad_norm": 0.8805339336395264,
"learning_rate": 6.977752857894684e-07,
"loss": 1.5362,
"step": 4880
},
{
"epoch": 7.824,
"grad_norm": 0.9314810633659363,
"learning_rate": 6.881266460949764e-07,
"loss": 1.353,
"step": 4890
},
{
"epoch": 7.84,
"grad_norm": 0.990050196647644,
"learning_rate": 6.785345236998242e-07,
"loss": 1.6943,
"step": 4900
},
{
"epoch": 7.856,
"grad_norm": 1.3979194164276123,
"learning_rate": 6.689992178086174e-07,
"loss": 1.6162,
"step": 4910
},
{
"epoch": 7.872,
"grad_norm": 1.0376416444778442,
"learning_rate": 6.595210258536997e-07,
"loss": 1.5699,
"step": 4920
},
{
"epoch": 7.888,
"grad_norm": 1.1278938055038452,
"learning_rate": 6.501002434858736e-07,
"loss": 1.3961,
"step": 4930
},
{
"epoch": 7.904,
"grad_norm": 2.2884480953216553,
"learning_rate": 6.407371645651808e-07,
"loss": 1.4523,
"step": 4940
},
{
"epoch": 7.92,
"grad_norm": 1.2708555459976196,
"learning_rate": 6.314320811517314e-07,
"loss": 1.1415,
"step": 4950
},
{
"epoch": 7.936,
"grad_norm": 1.3077367544174194,
"learning_rate": 6.221852834965983e-07,
"loss": 1.3157,
"step": 4960
},
{
"epoch": 7.952,
"grad_norm": 1.0594474077224731,
"learning_rate": 6.129970600327623e-07,
"loss": 1.4966,
"step": 4970
},
{
"epoch": 7.968,
"grad_norm": 2.269719362258911,
"learning_rate": 6.038676973661142e-07,
"loss": 1.4233,
"step": 4980
},
{
"epoch": 7.984,
"grad_norm": 1.891919732093811,
"learning_rate": 5.947974802665135e-07,
"loss": 1.4827,
"step": 4990
},
{
"epoch": 8.0,
"grad_norm": 1.3045769929885864,
"learning_rate": 5.857866916589089e-07,
"loss": 1.4586,
"step": 5000
},
{
"epoch": 8.016,
"grad_norm": 1.204778790473938,
"learning_rate": 5.76835612614512e-07,
"loss": 1.3929,
"step": 5010
},
{
"epoch": 8.032,
"grad_norm": 0.8054949641227722,
"learning_rate": 5.679445223420257e-07,
"loss": 1.4817,
"step": 5020
},
{
"epoch": 8.048,
"grad_norm": 1.2945849895477295,
"learning_rate": 5.591136981789439e-07,
"loss": 1.524,
"step": 5030
},
{
"epoch": 8.064,
"grad_norm": 0.6468287110328674,
"learning_rate": 5.503434155828924e-07,
"loss": 1.483,
"step": 5040
},
{
"epoch": 8.08,
"grad_norm": 1.2295925617218018,
"learning_rate": 5.416339481230379e-07,
"loss": 1.4483,
"step": 5050
},
{
"epoch": 8.096,
"grad_norm": 1.4299932718276978,
"learning_rate": 5.329855674715592e-07,
"loss": 1.5119,
"step": 5060
},
{
"epoch": 8.112,
"grad_norm": 1.1725962162017822,
"learning_rate": 5.243985433951693e-07,
"loss": 1.4884,
"step": 5070
},
{
"epoch": 8.128,
"grad_norm": 1.4912844896316528,
"learning_rate": 5.158731437466993e-07,
"loss": 1.3456,
"step": 5080
},
{
"epoch": 8.144,
"grad_norm": 1.4730242490768433,
"learning_rate": 5.074096344567475e-07,
"loss": 1.3254,
"step": 5090
},
{
"epoch": 8.16,
"grad_norm": 1.6218539476394653,
"learning_rate": 4.990082795253825e-07,
"loss": 1.5614,
"step": 5100
},
{
"epoch": 8.176,
"grad_norm": 0.7668495178222656,
"learning_rate": 4.906693410139057e-07,
"loss": 1.4146,
"step": 5110
},
{
"epoch": 8.192,
"grad_norm": 1.1908650398254395,
"learning_rate": 4.823930790366801e-07,
"loss": 1.534,
"step": 5120
},
{
"epoch": 8.208,
"grad_norm": 2.078003168106079,
"learning_rate": 4.7417975175301835e-07,
"loss": 1.3527,
"step": 5130
},
{
"epoch": 8.224,
"grad_norm": 1.5037908554077148,
"learning_rate": 4.6602961535912315e-07,
"loss": 1.4449,
"step": 5140
},
{
"epoch": 8.24,
"grad_norm": 1.4006104469299316,
"learning_rate": 4.5794292408010285e-07,
"loss": 1.553,
"step": 5150
},
{
"epoch": 8.256,
"grad_norm": 0.6962419748306274,
"learning_rate": 4.499199301620377e-07,
"loss": 1.5909,
"step": 5160
},
{
"epoch": 8.272,
"grad_norm": 0.8190912008285522,
"learning_rate": 4.419608838641132e-07,
"loss": 1.6939,
"step": 5170
},
{
"epoch": 8.288,
"grad_norm": 1.5709136724472046,
"learning_rate": 4.340660334508115e-07,
"loss": 1.6655,
"step": 5180
},
{
"epoch": 8.304,
"grad_norm": 0.6704534888267517,
"learning_rate": 4.2623562518417095e-07,
"loss": 1.5961,
"step": 5190
},
{
"epoch": 8.32,
"grad_norm": 1.3276532888412476,
"learning_rate": 4.1846990331610246e-07,
"loss": 1.6921,
"step": 5200
},
{
"epoch": 8.336,
"grad_norm": 1.118941307067871,
"learning_rate": 4.1076911008076895e-07,
"loss": 1.4401,
"step": 5210
},
{
"epoch": 8.352,
"grad_norm": 1.010671615600586,
"learning_rate": 4.031334856870325e-07,
"loss": 1.6014,
"step": 5220
},
{
"epoch": 8.368,
"grad_norm": 0.868981122970581,
"learning_rate": 3.955632683109617e-07,
"loss": 1.4861,
"step": 5230
},
{
"epoch": 8.384,
"grad_norm": 1.936972975730896,
"learning_rate": 3.880586940883979e-07,
"loss": 1.2846,
"step": 5240
},
{
"epoch": 8.4,
"grad_norm": 0.7472865581512451,
"learning_rate": 3.806199971075947e-07,
"loss": 1.3225,
"step": 5250
},
{
"epoch": 8.416,
"grad_norm": 0.8359613418579102,
"learning_rate": 3.732474094019131e-07,
"loss": 1.4802,
"step": 5260
},
{
"epoch": 8.432,
"grad_norm": 1.1112358570098877,
"learning_rate": 3.659411609425834e-07,
"loss": 1.7094,
"step": 5270
},
{
"epoch": 8.448,
"grad_norm": 1.2264232635498047,
"learning_rate": 3.58701479631535e-07,
"loss": 1.2799,
"step": 5280
},
{
"epoch": 8.464,
"grad_norm": 0.952431857585907,
"learning_rate": 3.5152859129428423e-07,
"loss": 1.721,
"step": 5290
},
{
"epoch": 8.48,
"grad_norm": 0.6120615601539612,
"learning_rate": 3.4442271967289083e-07,
"loss": 1.6105,
"step": 5300
},
{
"epoch": 8.496,
"grad_norm": 0.5835082530975342,
"learning_rate": 3.3738408641898005e-07,
"loss": 1.69,
"step": 5310
},
{
"epoch": 8.512,
"grad_norm": 1.4935648441314697,
"learning_rate": 3.304129110868284e-07,
"loss": 1.4912,
"step": 5320
},
{
"epoch": 8.528,
"grad_norm": 1.0045180320739746,
"learning_rate": 3.235094111265141e-07,
"loss": 1.6642,
"step": 5330
},
{
"epoch": 8.544,
"grad_norm": 1.1329545974731445,
"learning_rate": 3.1667380187713535e-07,
"loss": 1.7507,
"step": 5340
},
{
"epoch": 8.56,
"grad_norm": 1.0779494047164917,
"learning_rate": 3.099062965600924e-07,
"loss": 1.6032,
"step": 5350
},
{
"epoch": 8.576,
"grad_norm": 1.9241889715194702,
"learning_rate": 3.0320710627243815e-07,
"loss": 1.4284,
"step": 5360
},
{
"epoch": 8.592,
"grad_norm": 1.4561920166015625,
"learning_rate": 2.9657643998029063e-07,
"loss": 1.5227,
"step": 5370
},
{
"epoch": 8.608,
"grad_norm": 1.7750341892242432,
"learning_rate": 2.900145045123173e-07,
"loss": 1.2836,
"step": 5380
},
{
"epoch": 8.624,
"grad_norm": 1.3114453554153442,
"learning_rate": 2.835215045532841e-07,
"loss": 1.5601,
"step": 5390
},
{
"epoch": 8.64,
"grad_norm": 0.6750730872154236,
"learning_rate": 2.7709764263766703e-07,
"loss": 1.5809,
"step": 5400
},
{
"epoch": 8.656,
"grad_norm": 1.6933780908584595,
"learning_rate": 2.7074311914333797e-07,
"loss": 1.4322,
"step": 5410
},
{
"epoch": 8.672,
"grad_norm": 1.8922215700149536,
"learning_rate": 2.64458132285314e-07,
"loss": 1.3897,
"step": 5420
},
{
"epoch": 8.688,
"grad_norm": 1.5831923484802246,
"learning_rate": 2.5824287810957194e-07,
"loss": 1.5138,
"step": 5430
},
{
"epoch": 8.704,
"grad_norm": 1.5072382688522339,
"learning_rate": 2.520975504869375e-07,
"loss": 1.6559,
"step": 5440
},
{
"epoch": 8.72,
"grad_norm": 1.1533654928207397,
"learning_rate": 2.460223411070337e-07,
"loss": 1.3763,
"step": 5450
},
{
"epoch": 8.736,
"grad_norm": 0.6706821322441101,
"learning_rate": 2.400174394723029e-07,
"loss": 1.5283,
"step": 5460
},
{
"epoch": 8.752,
"grad_norm": 1.1693994998931885,
"learning_rate": 2.3408303289209706e-07,
"loss": 1.637,
"step": 5470
},
{
"epoch": 8.768,
"grad_norm": 1.6314516067504883,
"learning_rate": 2.2821930647683427e-07,
"loss": 1.7647,
"step": 5480
},
{
"epoch": 8.784,
"grad_norm": 1.4832297563552856,
"learning_rate": 2.224264431322229e-07,
"loss": 1.4667,
"step": 5490
},
{
"epoch": 8.8,
"grad_norm": 1.2678664922714233,
"learning_rate": 2.167046235535597e-07,
"loss": 1.5611,
"step": 5500
},
{
"epoch": 8.816,
"grad_norm": 0.8833269476890564,
"learning_rate": 2.1105402622008996e-07,
"loss": 1.3795,
"step": 5510
},
{
"epoch": 8.832,
"grad_norm": 1.2408838272094727,
"learning_rate": 2.0547482738944302e-07,
"loss": 1.4189,
"step": 5520
},
{
"epoch": 8.848,
"grad_norm": 2.066340446472168,
"learning_rate": 1.9996720109213134e-07,
"loss": 1.6006,
"step": 5530
},
{
"epoch": 8.864,
"grad_norm": 1.897200584411621,
"learning_rate": 1.9453131912612694e-07,
"loss": 1.3799,
"step": 5540
},
{
"epoch": 8.88,
"grad_norm": 1.138106107711792,
"learning_rate": 1.891673510514952e-07,
"loss": 1.4437,
"step": 5550
},
{
"epoch": 8.896,
"grad_norm": 2.373166084289551,
"learning_rate": 1.8387546418511343e-07,
"loss": 1.3654,
"step": 5560
},
{
"epoch": 8.912,
"grad_norm": 1.4822841882705688,
"learning_rate": 1.7865582359544664e-07,
"loss": 1.4131,
"step": 5570
},
{
"epoch": 8.928,
"grad_norm": 0.9219613075256348,
"learning_rate": 1.735085920974011e-07,
"loss": 1.5291,
"step": 5580
},
{
"epoch": 8.943999999999999,
"grad_norm": 0.986291229724884,
"learning_rate": 1.684339302472443e-07,
"loss": 1.5373,
"step": 5590
},
{
"epoch": 8.96,
"grad_norm": 0.7983741760253906,
"learning_rate": 1.6343199633759715e-07,
"loss": 1.542,
"step": 5600
},
{
"epoch": 8.975999999999999,
"grad_norm": 2.3131961822509766,
"learning_rate": 1.5850294639249835e-07,
"loss": 1.2797,
"step": 5610
},
{
"epoch": 8.992,
"grad_norm": 1.8482275009155273,
"learning_rate": 1.5364693416253323e-07,
"loss": 1.4034,
"step": 5620
},
{
"epoch": 9.008,
"grad_norm": 1.6701992750167847,
"learning_rate": 1.4886411112004258e-07,
"loss": 1.6857,
"step": 5630
},
{
"epoch": 9.024,
"grad_norm": 1.170453429222107,
"learning_rate": 1.4415462645439498e-07,
"loss": 1.5003,
"step": 5640
},
{
"epoch": 9.04,
"grad_norm": 1.3455228805541992,
"learning_rate": 1.395186270673335e-07,
"loss": 1.3894,
"step": 5650
},
{
"epoch": 9.056,
"grad_norm": 1.3283231258392334,
"learning_rate": 1.3495625756839464e-07,
"loss": 1.4429,
"step": 5660
},
{
"epoch": 9.072,
"grad_norm": 1.189936637878418,
"learning_rate": 1.3046766027039638e-07,
"loss": 1.464,
"step": 5670
},
{
"epoch": 9.088,
"grad_norm": 1.3344308137893677,
"learning_rate": 1.260529751849987e-07,
"loss": 1.5315,
"step": 5680
},
{
"epoch": 9.104,
"grad_norm": 1.8056098222732544,
"learning_rate": 1.2171234001833788e-07,
"loss": 1.3657,
"step": 5690
},
{
"epoch": 9.12,
"grad_norm": 0.6753764748573303,
"learning_rate": 1.1744589016672997e-07,
"loss": 1.2953,
"step": 5700
},
{
"epoch": 9.136,
"grad_norm": 0.8828732371330261,
"learning_rate": 1.1325375871244771e-07,
"loss": 1.4204,
"step": 5710
},
{
"epoch": 9.152,
"grad_norm": 1.75863778591156,
"learning_rate": 1.0913607641956842e-07,
"loss": 1.3356,
"step": 5720
},
{
"epoch": 9.168,
"grad_norm": 1.3419684171676636,
"learning_rate": 1.0509297172989652e-07,
"loss": 1.4037,
"step": 5730
},
{
"epoch": 9.184,
"grad_norm": 0.8731301426887512,
"learning_rate": 1.0112457075895643e-07,
"loss": 1.4981,
"step": 5740
},
{
"epoch": 9.2,
"grad_norm": 1.4133774042129517,
"learning_rate": 9.72309972920582e-08,
"loss": 1.5368,
"step": 5750
},
{
"epoch": 9.216,
"grad_norm": 1.106041669845581,
"learning_rate": 9.341237278043786e-08,
"loss": 1.5814,
"step": 5760
},
{
"epoch": 9.232,
"grad_norm": 1.1448427438735962,
"learning_rate": 8.966881633746732e-08,
"loss": 1.3992,
"step": 5770
},
{
"epoch": 9.248,
"grad_norm": 1.489063024520874,
"learning_rate": 8.600044473493856e-08,
"loss": 1.7416,
"step": 5780
},
{
"epoch": 9.264,
"grad_norm": 2.715735673904419,
"learning_rate": 8.24073723994237e-08,
"loss": 1.4689,
"step": 5790
},
{
"epoch": 9.28,
"grad_norm": 1.1526888608932495,
"learning_rate": 7.888971140870394e-08,
"loss": 1.5287,
"step": 5800
},
{
"epoch": 9.296,
"grad_norm": 1.4548875093460083,
"learning_rate": 7.544757148827297e-08,
"loss": 1.6217,
"step": 5810
},
{
"epoch": 9.312,
"grad_norm": 1.181388020515442,
"learning_rate": 7.208106000791582e-08,
"loss": 1.4709,
"step": 5820
},
{
"epoch": 9.328,
"grad_norm": 1.2242498397827148,
"learning_rate": 6.879028197835869e-08,
"loss": 1.4985,
"step": 5830
},
{
"epoch": 9.344,
"grad_norm": 0.9628410935401917,
"learning_rate": 6.557534004799443e-08,
"loss": 1.5612,
"step": 5840
},
{
"epoch": 9.36,
"grad_norm": 0.8261465430259705,
"learning_rate": 6.243633449967923e-08,
"loss": 1.7772,
"step": 5850
}
],
"logging_steps": 10,
"max_steps": 6250,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 150,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.1565293851378975e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}