14034 lines
316 KiB
JSON
14034 lines
316 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 10000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0005,
|
|
"grad_norm": 1.9079082012176514,
|
|
"learning_rate": 0.0001,
|
|
"loss": 5.5891,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.001,
|
|
"grad_norm": 0.7927042841911316,
|
|
"learning_rate": 0.0002,
|
|
"loss": 5.1812,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.0015,
|
|
"grad_norm": 0.6806358098983765,
|
|
"learning_rate": 0.0001998998998998999,
|
|
"loss": 4.9562,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.002,
|
|
"grad_norm": 1.6273939609527588,
|
|
"learning_rate": 0.0001997997997997998,
|
|
"loss": 4.7403,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.0025,
|
|
"grad_norm": 0.7797424793243408,
|
|
"learning_rate": 0.0001996996996996997,
|
|
"loss": 4.4994,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.003,
|
|
"grad_norm": 0.6761952638626099,
|
|
"learning_rate": 0.0001995995995995996,
|
|
"loss": 4.315,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.0035,
|
|
"grad_norm": 0.6525298953056335,
|
|
"learning_rate": 0.0001994994994994995,
|
|
"loss": 4.1599,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.004,
|
|
"grad_norm": 0.8141308426856995,
|
|
"learning_rate": 0.0001993993993993994,
|
|
"loss": 4.0421,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.0045,
|
|
"grad_norm": 0.7103850841522217,
|
|
"learning_rate": 0.00019929929929929932,
|
|
"loss": 3.9519,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.005,
|
|
"grad_norm": 0.4970089793205261,
|
|
"learning_rate": 0.0001991991991991992,
|
|
"loss": 3.8788,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.0055,
|
|
"grad_norm": 1.0156959295272827,
|
|
"learning_rate": 0.00019909909909909912,
|
|
"loss": 3.8267,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.006,
|
|
"grad_norm": 0.6486234664916992,
|
|
"learning_rate": 0.000198998998998999,
|
|
"loss": 3.7721,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.0065,
|
|
"grad_norm": 0.560275673866272,
|
|
"learning_rate": 0.0001988988988988989,
|
|
"loss": 3.7394,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.007,
|
|
"grad_norm": 0.6026498675346375,
|
|
"learning_rate": 0.0001987987987987988,
|
|
"loss": 3.7272,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.0075,
|
|
"grad_norm": 0.502984881401062,
|
|
"learning_rate": 0.0001986986986986987,
|
|
"loss": 3.6899,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.008,
|
|
"grad_norm": 0.45901086926460266,
|
|
"learning_rate": 0.0001985985985985986,
|
|
"loss": 3.664,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.0085,
|
|
"grad_norm": 0.6574872732162476,
|
|
"learning_rate": 0.0001984984984984985,
|
|
"loss": 3.6586,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.009,
|
|
"grad_norm": 0.5188643932342529,
|
|
"learning_rate": 0.0001983983983983984,
|
|
"loss": 3.6381,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.0095,
|
|
"grad_norm": 0.46534261107444763,
|
|
"learning_rate": 0.00019829829829829833,
|
|
"loss": 3.6189,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"grad_norm": 0.5883278250694275,
|
|
"learning_rate": 0.0001981981981981982,
|
|
"loss": 3.5719,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.0105,
|
|
"grad_norm": 0.5875305533409119,
|
|
"learning_rate": 0.00019809809809809813,
|
|
"loss": 3.5682,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.011,
|
|
"grad_norm": 0.6514595746994019,
|
|
"learning_rate": 0.000197997997997998,
|
|
"loss": 3.5454,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.0115,
|
|
"grad_norm": 0.7901559472084045,
|
|
"learning_rate": 0.0001978978978978979,
|
|
"loss": 3.535,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.012,
|
|
"grad_norm": 0.8417282700538635,
|
|
"learning_rate": 0.0001977977977977978,
|
|
"loss": 3.499,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.0125,
|
|
"grad_norm": 0.5768285393714905,
|
|
"learning_rate": 0.0001976976976976977,
|
|
"loss": 3.4973,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.013,
|
|
"grad_norm": 0.5534257292747498,
|
|
"learning_rate": 0.0001975975975975976,
|
|
"loss": 3.443,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.0135,
|
|
"grad_norm": 0.6167763471603394,
|
|
"learning_rate": 0.0001974974974974975,
|
|
"loss": 3.4109,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.014,
|
|
"grad_norm": 0.7287874221801758,
|
|
"learning_rate": 0.00019739739739739739,
|
|
"loss": 3.3787,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.0145,
|
|
"grad_norm": 0.5176478028297424,
|
|
"learning_rate": 0.0001972972972972973,
|
|
"loss": 3.3451,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.015,
|
|
"grad_norm": 0.5850136280059814,
|
|
"learning_rate": 0.0001971971971971972,
|
|
"loss": 3.3066,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.0155,
|
|
"grad_norm": 0.7310737371444702,
|
|
"learning_rate": 0.00019709709709709713,
|
|
"loss": 3.2662,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.016,
|
|
"grad_norm": 0.6415804028511047,
|
|
"learning_rate": 0.00019699699699699701,
|
|
"loss": 3.2275,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.0165,
|
|
"grad_norm": 0.6757485270500183,
|
|
"learning_rate": 0.0001968968968968969,
|
|
"loss": 3.1872,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.017,
|
|
"grad_norm": 0.6382904052734375,
|
|
"learning_rate": 0.00019679679679679681,
|
|
"loss": 3.1226,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.0175,
|
|
"grad_norm": 0.754895806312561,
|
|
"learning_rate": 0.0001966966966966967,
|
|
"loss": 3.0844,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.018,
|
|
"grad_norm": 0.6877201795578003,
|
|
"learning_rate": 0.00019659659659659661,
|
|
"loss": 3.0516,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.0185,
|
|
"grad_norm": 0.7147930264472961,
|
|
"learning_rate": 0.0001964964964964965,
|
|
"loss": 3.0131,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.019,
|
|
"grad_norm": 0.6776050329208374,
|
|
"learning_rate": 0.0001963963963963964,
|
|
"loss": 2.9869,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.0195,
|
|
"grad_norm": 0.7239454388618469,
|
|
"learning_rate": 0.0001962962962962963,
|
|
"loss": 2.9589,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"grad_norm": 0.6836283802986145,
|
|
"learning_rate": 0.00019619619619619621,
|
|
"loss": 2.8852,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.0205,
|
|
"grad_norm": 0.9422793984413147,
|
|
"learning_rate": 0.00019609609609609613,
|
|
"loss": 2.879,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.021,
|
|
"grad_norm": 0.6841469407081604,
|
|
"learning_rate": 0.00019599599599599602,
|
|
"loss": 2.8571,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.0215,
|
|
"grad_norm": 0.6966745257377625,
|
|
"learning_rate": 0.0001958958958958959,
|
|
"loss": 2.7985,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.022,
|
|
"grad_norm": 0.8714805841445923,
|
|
"learning_rate": 0.00019579579579579582,
|
|
"loss": 2.7878,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.0225,
|
|
"grad_norm": 0.7409490346908569,
|
|
"learning_rate": 0.0001956956956956957,
|
|
"loss": 2.7447,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.023,
|
|
"grad_norm": 0.7043179273605347,
|
|
"learning_rate": 0.00019559559559559562,
|
|
"loss": 2.7465,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.0235,
|
|
"grad_norm": 0.756272554397583,
|
|
"learning_rate": 0.0001954954954954955,
|
|
"loss": 2.7175,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.024,
|
|
"grad_norm": 0.7182954549789429,
|
|
"learning_rate": 0.0001953953953953954,
|
|
"loss": 2.6932,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.0245,
|
|
"grad_norm": 0.8014600276947021,
|
|
"learning_rate": 0.0001952952952952953,
|
|
"loss": 2.6721,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.025,
|
|
"grad_norm": 0.7439653277397156,
|
|
"learning_rate": 0.0001951951951951952,
|
|
"loss": 2.6526,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.0255,
|
|
"grad_norm": 0.7318114042282104,
|
|
"learning_rate": 0.0001950950950950951,
|
|
"loss": 2.6535,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.026,
|
|
"grad_norm": 0.7556585073471069,
|
|
"learning_rate": 0.00019499499499499502,
|
|
"loss": 2.6466,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.0265,
|
|
"grad_norm": 0.7585588693618774,
|
|
"learning_rate": 0.0001948948948948949,
|
|
"loss": 2.6062,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.027,
|
|
"grad_norm": 0.7325690388679504,
|
|
"learning_rate": 0.00019479479479479482,
|
|
"loss": 2.604,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.0275,
|
|
"grad_norm": 0.7965534925460815,
|
|
"learning_rate": 0.0001946946946946947,
|
|
"loss": 2.5905,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.028,
|
|
"grad_norm": 0.8178598880767822,
|
|
"learning_rate": 0.00019459459459459462,
|
|
"loss": 2.5676,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.0285,
|
|
"grad_norm": 0.750317394733429,
|
|
"learning_rate": 0.0001944944944944945,
|
|
"loss": 2.562,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 0.029,
|
|
"grad_norm": 0.7719892263412476,
|
|
"learning_rate": 0.0001943943943943944,
|
|
"loss": 2.562,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.0295,
|
|
"grad_norm": 0.8082327246665955,
|
|
"learning_rate": 0.0001942942942942943,
|
|
"loss": 2.5188,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"grad_norm": 0.8357940912246704,
|
|
"learning_rate": 0.0001941941941941942,
|
|
"loss": 2.5437,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.0305,
|
|
"grad_norm": 0.789130687713623,
|
|
"learning_rate": 0.0001940940940940941,
|
|
"loss": 2.5006,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 0.031,
|
|
"grad_norm": 0.8761134743690491,
|
|
"learning_rate": 0.00019399399399399402,
|
|
"loss": 2.5121,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.0315,
|
|
"grad_norm": 0.7823195457458496,
|
|
"learning_rate": 0.0001938938938938939,
|
|
"loss": 2.52,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 0.032,
|
|
"grad_norm": 0.7430358529090881,
|
|
"learning_rate": 0.00019379379379379382,
|
|
"loss": 2.5108,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.0325,
|
|
"grad_norm": 0.7893998026847839,
|
|
"learning_rate": 0.0001936936936936937,
|
|
"loss": 2.4943,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 0.033,
|
|
"grad_norm": 0.79250568151474,
|
|
"learning_rate": 0.0001935935935935936,
|
|
"loss": 2.5005,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.0335,
|
|
"grad_norm": 0.9912872910499573,
|
|
"learning_rate": 0.0001934934934934935,
|
|
"loss": 2.4595,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 0.034,
|
|
"grad_norm": 0.7901522517204285,
|
|
"learning_rate": 0.0001933933933933934,
|
|
"loss": 2.4501,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.0345,
|
|
"grad_norm": 0.7800329327583313,
|
|
"learning_rate": 0.0001932932932932933,
|
|
"loss": 2.46,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 0.035,
|
|
"grad_norm": 0.838046669960022,
|
|
"learning_rate": 0.0001931931931931932,
|
|
"loss": 2.4495,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.0355,
|
|
"grad_norm": 0.7794272899627686,
|
|
"learning_rate": 0.00019309309309309308,
|
|
"loss": 2.4537,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 0.036,
|
|
"grad_norm": 0.8007591962814331,
|
|
"learning_rate": 0.000192992992992993,
|
|
"loss": 2.4339,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.0365,
|
|
"grad_norm": 0.8491222262382507,
|
|
"learning_rate": 0.0001928928928928929,
|
|
"loss": 2.4461,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 0.037,
|
|
"grad_norm": 0.8045406937599182,
|
|
"learning_rate": 0.00019279279279279282,
|
|
"loss": 2.4192,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.0375,
|
|
"grad_norm": 0.8075376152992249,
|
|
"learning_rate": 0.0001926926926926927,
|
|
"loss": 2.4254,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 0.038,
|
|
"grad_norm": 0.7965284585952759,
|
|
"learning_rate": 0.0001925925925925926,
|
|
"loss": 2.4136,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.0385,
|
|
"grad_norm": 0.8071419596672058,
|
|
"learning_rate": 0.0001924924924924925,
|
|
"loss": 2.4018,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 0.039,
|
|
"grad_norm": 0.7852412462234497,
|
|
"learning_rate": 0.0001923923923923924,
|
|
"loss": 2.3948,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.0395,
|
|
"grad_norm": 0.7773271799087524,
|
|
"learning_rate": 0.0001922922922922923,
|
|
"loss": 2.3731,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"grad_norm": 0.943693220615387,
|
|
"learning_rate": 0.0001921921921921922,
|
|
"loss": 2.4004,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.0405,
|
|
"grad_norm": 0.8425918817520142,
|
|
"learning_rate": 0.00019209209209209208,
|
|
"loss": 2.3765,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 0.041,
|
|
"grad_norm": 0.7713614702224731,
|
|
"learning_rate": 0.000191991991991992,
|
|
"loss": 2.3505,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.0415,
|
|
"grad_norm": 0.9050373435020447,
|
|
"learning_rate": 0.0001918918918918919,
|
|
"loss": 2.3851,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 0.042,
|
|
"grad_norm": 0.9341697096824646,
|
|
"learning_rate": 0.00019179179179179182,
|
|
"loss": 2.393,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.0425,
|
|
"grad_norm": 0.9182080626487732,
|
|
"learning_rate": 0.0001916916916916917,
|
|
"loss": 2.3528,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 0.043,
|
|
"grad_norm": 0.8569957613945007,
|
|
"learning_rate": 0.0001915915915915916,
|
|
"loss": 2.3723,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.0435,
|
|
"grad_norm": 0.8590387105941772,
|
|
"learning_rate": 0.0001914914914914915,
|
|
"loss": 2.3591,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 0.044,
|
|
"grad_norm": 0.8434947729110718,
|
|
"learning_rate": 0.0001913913913913914,
|
|
"loss": 2.3316,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.0445,
|
|
"grad_norm": 0.8210241794586182,
|
|
"learning_rate": 0.0001912912912912913,
|
|
"loss": 2.3346,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 0.045,
|
|
"grad_norm": 0.815951943397522,
|
|
"learning_rate": 0.0001911911911911912,
|
|
"loss": 2.3335,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.0455,
|
|
"grad_norm": 0.8260611891746521,
|
|
"learning_rate": 0.00019109109109109108,
|
|
"loss": 2.3345,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 0.046,
|
|
"grad_norm": 0.8018535375595093,
|
|
"learning_rate": 0.000190990990990991,
|
|
"loss": 2.3162,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.0465,
|
|
"grad_norm": 0.7817215919494629,
|
|
"learning_rate": 0.00019089089089089088,
|
|
"loss": 2.3227,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 0.047,
|
|
"grad_norm": 0.8590435981750488,
|
|
"learning_rate": 0.0001907907907907908,
|
|
"loss": 2.3195,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 0.0475,
|
|
"grad_norm": 0.834904670715332,
|
|
"learning_rate": 0.0001906906906906907,
|
|
"loss": 2.3172,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 0.048,
|
|
"grad_norm": 0.7928949594497681,
|
|
"learning_rate": 0.0001905905905905906,
|
|
"loss": 2.3117,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.0485,
|
|
"grad_norm": 0.7825335264205933,
|
|
"learning_rate": 0.0001904904904904905,
|
|
"loss": 2.2934,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 0.049,
|
|
"grad_norm": 0.8034011125564575,
|
|
"learning_rate": 0.0001903903903903904,
|
|
"loss": 2.2964,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 0.0495,
|
|
"grad_norm": 0.8105394840240479,
|
|
"learning_rate": 0.0001902902902902903,
|
|
"loss": 2.3114,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"grad_norm": 0.7957478165626526,
|
|
"learning_rate": 0.0001901901901901902,
|
|
"loss": 2.3039,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.0505,
|
|
"grad_norm": 0.8164071440696716,
|
|
"learning_rate": 0.00019009009009009008,
|
|
"loss": 2.2803,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 0.051,
|
|
"grad_norm": 0.8089556097984314,
|
|
"learning_rate": 0.00018998998998999,
|
|
"loss": 2.2965,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 0.0515,
|
|
"grad_norm": 0.7860431671142578,
|
|
"learning_rate": 0.00018988988988988988,
|
|
"loss": 2.2686,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 0.052,
|
|
"grad_norm": 0.859445333480835,
|
|
"learning_rate": 0.0001897897897897898,
|
|
"loss": 2.2996,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.0525,
|
|
"grad_norm": 0.8138173818588257,
|
|
"learning_rate": 0.0001896896896896897,
|
|
"loss": 2.3295,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 0.053,
|
|
"grad_norm": 0.7922393679618835,
|
|
"learning_rate": 0.0001895895895895896,
|
|
"loss": 2.2671,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 0.0535,
|
|
"grad_norm": 0.8085952401161194,
|
|
"learning_rate": 0.0001894894894894895,
|
|
"loss": 2.2899,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 0.054,
|
|
"grad_norm": 0.8285341858863831,
|
|
"learning_rate": 0.0001893893893893894,
|
|
"loss": 2.2911,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.0545,
|
|
"grad_norm": 0.801395833492279,
|
|
"learning_rate": 0.0001892892892892893,
|
|
"loss": 2.2729,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 0.055,
|
|
"grad_norm": 0.8205108642578125,
|
|
"learning_rate": 0.0001891891891891892,
|
|
"loss": 2.2709,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.0555,
|
|
"grad_norm": 0.821998119354248,
|
|
"learning_rate": 0.00018908908908908908,
|
|
"loss": 2.2847,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 0.056,
|
|
"grad_norm": 0.8000075221061707,
|
|
"learning_rate": 0.000188988988988989,
|
|
"loss": 2.2663,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.0565,
|
|
"grad_norm": 0.8233977556228638,
|
|
"learning_rate": 0.00018888888888888888,
|
|
"loss": 2.261,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 0.057,
|
|
"grad_norm": 0.8421242833137512,
|
|
"learning_rate": 0.0001887887887887888,
|
|
"loss": 2.2626,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 0.0575,
|
|
"grad_norm": 0.8450160026550293,
|
|
"learning_rate": 0.00018868868868868868,
|
|
"loss": 2.2755,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 0.058,
|
|
"grad_norm": 0.8315795660018921,
|
|
"learning_rate": 0.0001885885885885886,
|
|
"loss": 2.2381,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.0585,
|
|
"grad_norm": 0.8966110348701477,
|
|
"learning_rate": 0.0001884884884884885,
|
|
"loss": 2.2517,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 0.059,
|
|
"grad_norm": 0.8162801861763,
|
|
"learning_rate": 0.0001883883883883884,
|
|
"loss": 2.2629,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 0.0595,
|
|
"grad_norm": 0.7899118661880493,
|
|
"learning_rate": 0.0001882882882882883,
|
|
"loss": 2.2304,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"grad_norm": 0.8371149301528931,
|
|
"learning_rate": 0.0001881881881881882,
|
|
"loss": 2.2232,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.0605,
|
|
"grad_norm": 0.8292490243911743,
|
|
"learning_rate": 0.00018808808808808809,
|
|
"loss": 2.2416,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 0.061,
|
|
"grad_norm": 0.775081217288971,
|
|
"learning_rate": 0.000187987987987988,
|
|
"loss": 2.2547,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 0.0615,
|
|
"grad_norm": 0.8445130586624146,
|
|
"learning_rate": 0.00018788788788788789,
|
|
"loss": 2.2243,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 0.062,
|
|
"grad_norm": 0.9159201383590698,
|
|
"learning_rate": 0.0001877877877877878,
|
|
"loss": 2.2423,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.0625,
|
|
"grad_norm": 0.8316602110862732,
|
|
"learning_rate": 0.00018768768768768769,
|
|
"loss": 2.2489,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 0.063,
|
|
"grad_norm": 0.8371139764785767,
|
|
"learning_rate": 0.0001875875875875876,
|
|
"loss": 2.2255,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 0.0635,
|
|
"grad_norm": 0.7829753756523132,
|
|
"learning_rate": 0.0001874874874874875,
|
|
"loss": 2.2279,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 0.064,
|
|
"grad_norm": 0.8447110652923584,
|
|
"learning_rate": 0.0001873873873873874,
|
|
"loss": 2.2392,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.0645,
|
|
"grad_norm": 0.8826359510421753,
|
|
"learning_rate": 0.00018728728728728731,
|
|
"loss": 2.2343,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 0.065,
|
|
"grad_norm": 0.9065181016921997,
|
|
"learning_rate": 0.0001871871871871872,
|
|
"loss": 2.2295,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.0655,
|
|
"grad_norm": 0.849250316619873,
|
|
"learning_rate": 0.0001870870870870871,
|
|
"loss": 2.2214,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 0.066,
|
|
"grad_norm": 0.7951939105987549,
|
|
"learning_rate": 0.000186986986986987,
|
|
"loss": 2.2028,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.0665,
|
|
"grad_norm": 0.8047727942466736,
|
|
"learning_rate": 0.0001868868868868869,
|
|
"loss": 2.2092,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 0.067,
|
|
"grad_norm": 0.8044347167015076,
|
|
"learning_rate": 0.0001867867867867868,
|
|
"loss": 2.214,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 0.0675,
|
|
"grad_norm": 0.831087052822113,
|
|
"learning_rate": 0.0001866866866866867,
|
|
"loss": 2.1976,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 0.068,
|
|
"grad_norm": 0.818250298500061,
|
|
"learning_rate": 0.00018658658658658657,
|
|
"loss": 2.211,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.0685,
|
|
"grad_norm": 0.8615385890007019,
|
|
"learning_rate": 0.0001864864864864865,
|
|
"loss": 2.2073,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 0.069,
|
|
"grad_norm": 0.7918171882629395,
|
|
"learning_rate": 0.0001863863863863864,
|
|
"loss": 2.222,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 0.0695,
|
|
"grad_norm": 0.7995744347572327,
|
|
"learning_rate": 0.00018628628628628632,
|
|
"loss": 2.1958,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"grad_norm": 0.8070845603942871,
|
|
"learning_rate": 0.0001861861861861862,
|
|
"loss": 2.2098,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.0705,
|
|
"grad_norm": 0.7857322692871094,
|
|
"learning_rate": 0.0001860860860860861,
|
|
"loss": 2.195,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 0.071,
|
|
"grad_norm": 0.8171130418777466,
|
|
"learning_rate": 0.000185985985985986,
|
|
"loss": 2.1882,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 0.0715,
|
|
"grad_norm": 0.8484584093093872,
|
|
"learning_rate": 0.0001858858858858859,
|
|
"loss": 2.1875,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 0.072,
|
|
"grad_norm": 0.7893686890602112,
|
|
"learning_rate": 0.0001857857857857858,
|
|
"loss": 2.2256,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.0725,
|
|
"grad_norm": 0.8339685797691345,
|
|
"learning_rate": 0.0001856856856856857,
|
|
"loss": 2.2197,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 0.073,
|
|
"grad_norm": 0.8456217646598816,
|
|
"learning_rate": 0.00018558558558558558,
|
|
"loss": 2.1971,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 0.0735,
|
|
"grad_norm": 0.8281363844871521,
|
|
"learning_rate": 0.0001854854854854855,
|
|
"loss": 2.1801,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 0.074,
|
|
"grad_norm": 0.8099832534790039,
|
|
"learning_rate": 0.0001853853853853854,
|
|
"loss": 2.1826,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.0745,
|
|
"grad_norm": 0.8337610960006714,
|
|
"learning_rate": 0.00018528528528528532,
|
|
"loss": 2.185,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 0.075,
|
|
"grad_norm": 0.8083080053329468,
|
|
"learning_rate": 0.0001851851851851852,
|
|
"loss": 2.1803,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.0755,
|
|
"grad_norm": 0.7984271049499512,
|
|
"learning_rate": 0.0001850850850850851,
|
|
"loss": 2.203,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 0.076,
|
|
"grad_norm": 0.8373366594314575,
|
|
"learning_rate": 0.000184984984984985,
|
|
"loss": 2.1887,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.0765,
|
|
"grad_norm": 0.8333912491798401,
|
|
"learning_rate": 0.0001848848848848849,
|
|
"loss": 2.1863,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 0.077,
|
|
"grad_norm": 0.8119728565216064,
|
|
"learning_rate": 0.0001847847847847848,
|
|
"loss": 2.2026,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 0.0775,
|
|
"grad_norm": 0.8200430274009705,
|
|
"learning_rate": 0.0001846846846846847,
|
|
"loss": 2.1644,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 0.078,
|
|
"grad_norm": 0.797855019569397,
|
|
"learning_rate": 0.00018458458458458458,
|
|
"loss": 2.1739,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.0785,
|
|
"grad_norm": 0.8249688148498535,
|
|
"learning_rate": 0.0001844844844844845,
|
|
"loss": 2.1716,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 0.079,
|
|
"grad_norm": 0.8358808159828186,
|
|
"learning_rate": 0.00018438438438438438,
|
|
"loss": 2.1849,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 0.0795,
|
|
"grad_norm": 0.8711438179016113,
|
|
"learning_rate": 0.0001842842842842843,
|
|
"loss": 2.1695,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"grad_norm": 0.9179810881614685,
|
|
"learning_rate": 0.0001841841841841842,
|
|
"loss": 2.1597,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.0805,
|
|
"grad_norm": 0.8647555708885193,
|
|
"learning_rate": 0.0001840840840840841,
|
|
"loss": 2.1918,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 0.081,
|
|
"grad_norm": 0.8114230036735535,
|
|
"learning_rate": 0.000183983983983984,
|
|
"loss": 2.163,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 0.0815,
|
|
"grad_norm": 0.8165040612220764,
|
|
"learning_rate": 0.0001838838838838839,
|
|
"loss": 2.1833,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 0.082,
|
|
"grad_norm": 0.8150405287742615,
|
|
"learning_rate": 0.0001837837837837838,
|
|
"loss": 2.1452,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.0825,
|
|
"grad_norm": 0.8312757015228271,
|
|
"learning_rate": 0.0001836836836836837,
|
|
"loss": 2.1535,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 0.083,
|
|
"grad_norm": 0.7952617406845093,
|
|
"learning_rate": 0.00018358358358358358,
|
|
"loss": 2.1463,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 0.0835,
|
|
"grad_norm": 0.855327308177948,
|
|
"learning_rate": 0.0001834834834834835,
|
|
"loss": 2.1622,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 0.084,
|
|
"grad_norm": 0.8281525373458862,
|
|
"learning_rate": 0.00018338338338338338,
|
|
"loss": 2.149,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.0845,
|
|
"grad_norm": 0.8069710731506348,
|
|
"learning_rate": 0.0001832832832832833,
|
|
"loss": 2.1548,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 0.085,
|
|
"grad_norm": 0.8265529870986938,
|
|
"learning_rate": 0.0001831831831831832,
|
|
"loss": 2.1588,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.0855,
|
|
"grad_norm": 0.8836376667022705,
|
|
"learning_rate": 0.0001830830830830831,
|
|
"loss": 2.1503,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 0.086,
|
|
"grad_norm": 0.8916524648666382,
|
|
"learning_rate": 0.000182982982982983,
|
|
"loss": 2.1771,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.0865,
|
|
"grad_norm": 0.9221314787864685,
|
|
"learning_rate": 0.0001828828828828829,
|
|
"loss": 2.1462,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 0.087,
|
|
"grad_norm": 0.9381101727485657,
|
|
"learning_rate": 0.0001827827827827828,
|
|
"loss": 2.1593,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 0.0875,
|
|
"grad_norm": 0.8455514311790466,
|
|
"learning_rate": 0.0001826826826826827,
|
|
"loss": 2.151,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 0.088,
|
|
"grad_norm": 0.89353346824646,
|
|
"learning_rate": 0.00018258258258258258,
|
|
"loss": 2.1447,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.0885,
|
|
"grad_norm": 0.9089289903640747,
|
|
"learning_rate": 0.0001824824824824825,
|
|
"loss": 2.1531,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 0.089,
|
|
"grad_norm": 0.8870664238929749,
|
|
"learning_rate": 0.00018238238238238238,
|
|
"loss": 2.1541,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 0.0895,
|
|
"grad_norm": 0.9005898237228394,
|
|
"learning_rate": 0.0001822822822822823,
|
|
"loss": 2.1582,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"grad_norm": 0.8735866546630859,
|
|
"learning_rate": 0.00018218218218218218,
|
|
"loss": 2.1419,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.0905,
|
|
"grad_norm": 0.8308618664741516,
|
|
"learning_rate": 0.0001820820820820821,
|
|
"loss": 2.1606,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 0.091,
|
|
"grad_norm": 0.7997368574142456,
|
|
"learning_rate": 0.000181981981981982,
|
|
"loss": 2.1242,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 0.0915,
|
|
"grad_norm": 0.804722785949707,
|
|
"learning_rate": 0.0001818818818818819,
|
|
"loss": 2.1264,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 0.092,
|
|
"grad_norm": 0.8133739829063416,
|
|
"learning_rate": 0.0001817817817817818,
|
|
"loss": 2.1403,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.0925,
|
|
"grad_norm": 0.8166353702545166,
|
|
"learning_rate": 0.0001816816816816817,
|
|
"loss": 2.159,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 0.093,
|
|
"grad_norm": 0.8658384084701538,
|
|
"learning_rate": 0.00018158158158158158,
|
|
"loss": 2.1383,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 0.0935,
|
|
"grad_norm": 0.8393406867980957,
|
|
"learning_rate": 0.0001814814814814815,
|
|
"loss": 2.1362,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 0.094,
|
|
"grad_norm": 0.8342397809028625,
|
|
"learning_rate": 0.00018138138138138138,
|
|
"loss": 2.1218,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.0945,
|
|
"grad_norm": 0.8732867240905762,
|
|
"learning_rate": 0.0001812812812812813,
|
|
"loss": 2.1378,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 0.095,
|
|
"grad_norm": 0.8951284289360046,
|
|
"learning_rate": 0.00018118118118118118,
|
|
"loss": 2.141,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.0955,
|
|
"grad_norm": 0.8014113903045654,
|
|
"learning_rate": 0.0001810810810810811,
|
|
"loss": 2.1341,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 0.096,
|
|
"grad_norm": 0.8125712871551514,
|
|
"learning_rate": 0.000180980980980981,
|
|
"loss": 2.1527,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.0965,
|
|
"grad_norm": 0.8115226030349731,
|
|
"learning_rate": 0.0001808808808808809,
|
|
"loss": 2.1237,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 0.097,
|
|
"grad_norm": 0.7778534889221191,
|
|
"learning_rate": 0.00018078078078078078,
|
|
"loss": 2.133,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 0.0975,
|
|
"grad_norm": 0.8221558928489685,
|
|
"learning_rate": 0.0001806806806806807,
|
|
"loss": 2.1301,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 0.098,
|
|
"grad_norm": 0.8114945888519287,
|
|
"learning_rate": 0.00018058058058058058,
|
|
"loss": 2.1219,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.0985,
|
|
"grad_norm": 0.8403328657150269,
|
|
"learning_rate": 0.0001804804804804805,
|
|
"loss": 2.1286,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 0.099,
|
|
"grad_norm": 0.8078803420066833,
|
|
"learning_rate": 0.00018038038038038038,
|
|
"loss": 2.1298,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 0.0995,
|
|
"grad_norm": 0.7954163551330566,
|
|
"learning_rate": 0.00018028028028028027,
|
|
"loss": 2.1243,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"grad_norm": 0.8172447681427002,
|
|
"learning_rate": 0.00018018018018018018,
|
|
"loss": 2.1422,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.1005,
|
|
"grad_norm": 0.8081322312355042,
|
|
"learning_rate": 0.00018008008008008007,
|
|
"loss": 2.1293,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 0.101,
|
|
"grad_norm": 0.8215644955635071,
|
|
"learning_rate": 0.00017997997997997998,
|
|
"loss": 2.1281,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 0.1015,
|
|
"grad_norm": 0.8085120916366577,
|
|
"learning_rate": 0.0001798798798798799,
|
|
"loss": 2.1144,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 0.102,
|
|
"grad_norm": 0.8848192691802979,
|
|
"learning_rate": 0.00017977977977977978,
|
|
"loss": 2.1076,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.1025,
|
|
"grad_norm": 0.8614477515220642,
|
|
"learning_rate": 0.0001796796796796797,
|
|
"loss": 2.1271,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 0.103,
|
|
"grad_norm": 0.7808970808982849,
|
|
"learning_rate": 0.00017957957957957958,
|
|
"loss": 2.132,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 0.1035,
|
|
"grad_norm": 0.8185530304908752,
|
|
"learning_rate": 0.0001794794794794795,
|
|
"loss": 2.1138,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 0.104,
|
|
"grad_norm": 0.8249826431274414,
|
|
"learning_rate": 0.00017937937937937938,
|
|
"loss": 2.1203,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.1045,
|
|
"grad_norm": 0.8197057247161865,
|
|
"learning_rate": 0.00017927927927927927,
|
|
"loss": 2.1301,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 0.105,
|
|
"grad_norm": 0.8169847130775452,
|
|
"learning_rate": 0.00017917917917917918,
|
|
"loss": 2.1199,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.1055,
|
|
"grad_norm": 0.8283096551895142,
|
|
"learning_rate": 0.00017907907907907907,
|
|
"loss": 2.1329,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 0.106,
|
|
"grad_norm": 0.8258969783782959,
|
|
"learning_rate": 0.00017897897897897898,
|
|
"loss": 2.1345,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.1065,
|
|
"grad_norm": 0.8351008296012878,
|
|
"learning_rate": 0.0001788788788788789,
|
|
"loss": 2.1043,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 0.107,
|
|
"grad_norm": 0.8097892999649048,
|
|
"learning_rate": 0.00017877877877877878,
|
|
"loss": 2.115,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 0.1075,
|
|
"grad_norm": 0.8576760292053223,
|
|
"learning_rate": 0.0001786786786786787,
|
|
"loss": 2.1142,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 0.108,
|
|
"grad_norm": 0.8387484550476074,
|
|
"learning_rate": 0.00017857857857857859,
|
|
"loss": 2.1179,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.1085,
|
|
"grad_norm": 0.8194589614868164,
|
|
"learning_rate": 0.0001784784784784785,
|
|
"loss": 2.096,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 0.109,
|
|
"grad_norm": 0.8305883407592773,
|
|
"learning_rate": 0.00017837837837837839,
|
|
"loss": 2.1312,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 0.1095,
|
|
"grad_norm": 0.7996582984924316,
|
|
"learning_rate": 0.00017827827827827827,
|
|
"loss": 2.1102,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"grad_norm": 0.7716205716133118,
|
|
"learning_rate": 0.00017817817817817819,
|
|
"loss": 2.1136,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.1105,
|
|
"grad_norm": 0.8277180194854736,
|
|
"learning_rate": 0.00017807807807807807,
|
|
"loss": 2.0968,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 0.111,
|
|
"grad_norm": 0.8593622446060181,
|
|
"learning_rate": 0.00017797797797797799,
|
|
"loss": 2.1082,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 0.1115,
|
|
"grad_norm": 0.8498421311378479,
|
|
"learning_rate": 0.00017787787787787787,
|
|
"loss": 2.1008,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 0.112,
|
|
"grad_norm": 0.7923771739006042,
|
|
"learning_rate": 0.00017777777777777779,
|
|
"loss": 2.124,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 0.1125,
|
|
"grad_norm": 0.7583004832267761,
|
|
"learning_rate": 0.0001776776776776777,
|
|
"loss": 2.1046,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 0.113,
|
|
"grad_norm": 0.7722444534301758,
|
|
"learning_rate": 0.0001775775775775776,
|
|
"loss": 2.0927,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 0.1135,
|
|
"grad_norm": 0.8124234676361084,
|
|
"learning_rate": 0.0001774774774774775,
|
|
"loss": 2.0971,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 0.114,
|
|
"grad_norm": 0.815171480178833,
|
|
"learning_rate": 0.0001773773773773774,
|
|
"loss": 2.109,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 0.1145,
|
|
"grad_norm": 0.8697826266288757,
|
|
"learning_rate": 0.00017727727727727727,
|
|
"loss": 2.1232,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 0.115,
|
|
"grad_norm": 0.7589493989944458,
|
|
"learning_rate": 0.0001771771771771772,
|
|
"loss": 2.0896,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.1155,
|
|
"grad_norm": 0.7957839965820312,
|
|
"learning_rate": 0.00017707707707707707,
|
|
"loss": 2.0935,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 0.116,
|
|
"grad_norm": 0.7996811866760254,
|
|
"learning_rate": 0.000176976976976977,
|
|
"loss": 2.0873,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 0.1165,
|
|
"grad_norm": 0.9038351774215698,
|
|
"learning_rate": 0.00017687687687687687,
|
|
"loss": 2.1061,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 0.117,
|
|
"grad_norm": 0.8256264328956604,
|
|
"learning_rate": 0.0001767767767767768,
|
|
"loss": 2.101,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 0.1175,
|
|
"grad_norm": 0.8364511728286743,
|
|
"learning_rate": 0.0001766766766766767,
|
|
"loss": 2.0703,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 0.118,
|
|
"grad_norm": 0.8705337643623352,
|
|
"learning_rate": 0.0001765765765765766,
|
|
"loss": 2.0993,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 0.1185,
|
|
"grad_norm": 0.817215621471405,
|
|
"learning_rate": 0.0001764764764764765,
|
|
"loss": 2.0877,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 0.119,
|
|
"grad_norm": 0.8533849716186523,
|
|
"learning_rate": 0.0001763763763763764,
|
|
"loss": 2.1219,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 0.1195,
|
|
"grad_norm": 0.8578528761863708,
|
|
"learning_rate": 0.00017627627627627627,
|
|
"loss": 2.0859,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"grad_norm": 0.9291893243789673,
|
|
"learning_rate": 0.0001761761761761762,
|
|
"loss": 2.0715,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.1205,
|
|
"grad_norm": 0.8820558786392212,
|
|
"learning_rate": 0.00017607607607607608,
|
|
"loss": 2.0945,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 0.121,
|
|
"grad_norm": 0.8163352012634277,
|
|
"learning_rate": 0.000175975975975976,
|
|
"loss": 2.084,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 0.1215,
|
|
"grad_norm": 0.8108770847320557,
|
|
"learning_rate": 0.00017587587587587588,
|
|
"loss": 2.103,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 0.122,
|
|
"grad_norm": 0.7584842443466187,
|
|
"learning_rate": 0.00017577577577577576,
|
|
"loss": 2.1035,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 0.1225,
|
|
"grad_norm": 0.8116704225540161,
|
|
"learning_rate": 0.00017567567567567568,
|
|
"loss": 2.0872,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 0.123,
|
|
"grad_norm": 0.8035491108894348,
|
|
"learning_rate": 0.0001755755755755756,
|
|
"loss": 2.0806,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 0.1235,
|
|
"grad_norm": 0.8007566332817078,
|
|
"learning_rate": 0.0001754754754754755,
|
|
"loss": 2.0872,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 0.124,
|
|
"grad_norm": 0.8226689100265503,
|
|
"learning_rate": 0.0001753753753753754,
|
|
"loss": 2.0863,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 0.1245,
|
|
"grad_norm": 0.7925471663475037,
|
|
"learning_rate": 0.00017527527527527528,
|
|
"loss": 2.0818,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 0.125,
|
|
"grad_norm": 0.8135493397712708,
|
|
"learning_rate": 0.0001751751751751752,
|
|
"loss": 2.1068,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.1255,
|
|
"grad_norm": 0.7937769889831543,
|
|
"learning_rate": 0.00017507507507507508,
|
|
"loss": 2.076,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 0.126,
|
|
"grad_norm": 0.774671733379364,
|
|
"learning_rate": 0.000174974974974975,
|
|
"loss": 2.0856,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 0.1265,
|
|
"grad_norm": 0.772739589214325,
|
|
"learning_rate": 0.00017487487487487488,
|
|
"loss": 2.0804,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 0.127,
|
|
"grad_norm": 0.7890838980674744,
|
|
"learning_rate": 0.00017477477477477476,
|
|
"loss": 2.074,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 0.1275,
|
|
"grad_norm": 0.8244984149932861,
|
|
"learning_rate": 0.00017467467467467468,
|
|
"loss": 2.0829,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 0.128,
|
|
"grad_norm": 0.8121048808097839,
|
|
"learning_rate": 0.0001745745745745746,
|
|
"loss": 2.0686,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 0.1285,
|
|
"grad_norm": 0.8105731010437012,
|
|
"learning_rate": 0.0001744744744744745,
|
|
"loss": 2.0791,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 0.129,
|
|
"grad_norm": 0.8011786937713623,
|
|
"learning_rate": 0.0001743743743743744,
|
|
"loss": 2.0656,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 0.1295,
|
|
"grad_norm": 0.8228626251220703,
|
|
"learning_rate": 0.00017427427427427428,
|
|
"loss": 2.0847,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"grad_norm": 0.8151708245277405,
|
|
"learning_rate": 0.0001741741741741742,
|
|
"loss": 2.0597,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.1305,
|
|
"grad_norm": 0.771611213684082,
|
|
"learning_rate": 0.00017407407407407408,
|
|
"loss": 2.0463,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 0.131,
|
|
"grad_norm": 0.8352901339530945,
|
|
"learning_rate": 0.000173973973973974,
|
|
"loss": 2.063,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 0.1315,
|
|
"grad_norm": 0.8361049294471741,
|
|
"learning_rate": 0.00017387387387387388,
|
|
"loss": 2.064,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 0.132,
|
|
"grad_norm": 0.8306949138641357,
|
|
"learning_rate": 0.00017377377377377376,
|
|
"loss": 2.0955,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 0.1325,
|
|
"grad_norm": 0.7998603582382202,
|
|
"learning_rate": 0.00017367367367367368,
|
|
"loss": 2.0794,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 0.133,
|
|
"grad_norm": 0.7653375864028931,
|
|
"learning_rate": 0.00017357357357357357,
|
|
"loss": 2.0688,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 0.1335,
|
|
"grad_norm": 0.8033472895622253,
|
|
"learning_rate": 0.00017347347347347348,
|
|
"loss": 2.0889,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 0.134,
|
|
"grad_norm": 0.8225494623184204,
|
|
"learning_rate": 0.0001733733733733734,
|
|
"loss": 2.092,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 0.1345,
|
|
"grad_norm": 0.8260511159896851,
|
|
"learning_rate": 0.00017327327327327328,
|
|
"loss": 2.071,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 0.135,
|
|
"grad_norm": 0.9667144417762756,
|
|
"learning_rate": 0.0001731731731731732,
|
|
"loss": 2.0685,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.1355,
|
|
"grad_norm": 0.8435398936271667,
|
|
"learning_rate": 0.00017307307307307308,
|
|
"loss": 2.0781,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 0.136,
|
|
"grad_norm": 0.8424886465072632,
|
|
"learning_rate": 0.000172972972972973,
|
|
"loss": 2.0757,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 0.1365,
|
|
"grad_norm": 0.8169295787811279,
|
|
"learning_rate": 0.00017287287287287288,
|
|
"loss": 2.0734,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 0.137,
|
|
"grad_norm": 0.8099573254585266,
|
|
"learning_rate": 0.00017277277277277277,
|
|
"loss": 2.0714,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 0.1375,
|
|
"grad_norm": 0.8306437730789185,
|
|
"learning_rate": 0.00017267267267267268,
|
|
"loss": 2.0735,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 0.138,
|
|
"grad_norm": 0.7802367210388184,
|
|
"learning_rate": 0.00017257257257257257,
|
|
"loss": 2.0712,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 0.1385,
|
|
"grad_norm": 0.8200580477714539,
|
|
"learning_rate": 0.00017247247247247248,
|
|
"loss": 2.056,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 0.139,
|
|
"grad_norm": 0.8281895518302917,
|
|
"learning_rate": 0.0001723723723723724,
|
|
"loss": 2.0632,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 0.1395,
|
|
"grad_norm": 0.7939669489860535,
|
|
"learning_rate": 0.00017227227227227228,
|
|
"loss": 2.0628,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"grad_norm": 0.7863283157348633,
|
|
"learning_rate": 0.0001721721721721722,
|
|
"loss": 2.0593,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.1405,
|
|
"grad_norm": 0.7744026184082031,
|
|
"learning_rate": 0.00017207207207207208,
|
|
"loss": 2.0734,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 0.141,
|
|
"grad_norm": 0.7912557721138,
|
|
"learning_rate": 0.000171971971971972,
|
|
"loss": 2.045,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 0.1415,
|
|
"grad_norm": 0.8050516843795776,
|
|
"learning_rate": 0.00017187187187187188,
|
|
"loss": 2.0546,
|
|
"step": 1415
|
|
},
|
|
{
|
|
"epoch": 0.142,
|
|
"grad_norm": 0.7703220248222351,
|
|
"learning_rate": 0.00017177177177177177,
|
|
"loss": 2.0647,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 0.1425,
|
|
"grad_norm": 0.8010295033454895,
|
|
"learning_rate": 0.00017167167167167168,
|
|
"loss": 2.0556,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 0.143,
|
|
"grad_norm": 0.8166594505310059,
|
|
"learning_rate": 0.00017157157157157157,
|
|
"loss": 2.0691,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 0.1435,
|
|
"grad_norm": 0.7984005212783813,
|
|
"learning_rate": 0.00017147147147147148,
|
|
"loss": 2.0826,
|
|
"step": 1435
|
|
},
|
|
{
|
|
"epoch": 0.144,
|
|
"grad_norm": 0.837327778339386,
|
|
"learning_rate": 0.00017137137137137137,
|
|
"loss": 2.0631,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 0.1445,
|
|
"grad_norm": 0.815968930721283,
|
|
"learning_rate": 0.00017127127127127128,
|
|
"loss": 2.0457,
|
|
"step": 1445
|
|
},
|
|
{
|
|
"epoch": 0.145,
|
|
"grad_norm": 0.8031236529350281,
|
|
"learning_rate": 0.0001711711711711712,
|
|
"loss": 2.0685,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.1455,
|
|
"grad_norm": 0.7846900224685669,
|
|
"learning_rate": 0.00017107107107107108,
|
|
"loss": 2.0528,
|
|
"step": 1455
|
|
},
|
|
{
|
|
"epoch": 0.146,
|
|
"grad_norm": 0.7788130044937134,
|
|
"learning_rate": 0.000170970970970971,
|
|
"loss": 2.0667,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 0.1465,
|
|
"grad_norm": 0.7862707376480103,
|
|
"learning_rate": 0.00017087087087087088,
|
|
"loss": 2.068,
|
|
"step": 1465
|
|
},
|
|
{
|
|
"epoch": 0.147,
|
|
"grad_norm": 0.8099005222320557,
|
|
"learning_rate": 0.00017077077077077077,
|
|
"loss": 2.0689,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 0.1475,
|
|
"grad_norm": 0.8044168949127197,
|
|
"learning_rate": 0.00017067067067067068,
|
|
"loss": 2.07,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 0.148,
|
|
"grad_norm": 0.7925140261650085,
|
|
"learning_rate": 0.00017057057057057057,
|
|
"loss": 2.0628,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 0.1485,
|
|
"grad_norm": 0.8262330293655396,
|
|
"learning_rate": 0.00017047047047047048,
|
|
"loss": 2.0525,
|
|
"step": 1485
|
|
},
|
|
{
|
|
"epoch": 0.149,
|
|
"grad_norm": 0.8038750886917114,
|
|
"learning_rate": 0.00017037037037037037,
|
|
"loss": 2.057,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 0.1495,
|
|
"grad_norm": 0.778914213180542,
|
|
"learning_rate": 0.00017027027027027028,
|
|
"loss": 2.0714,
|
|
"step": 1495
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"grad_norm": 0.7874584794044495,
|
|
"learning_rate": 0.0001701701701701702,
|
|
"loss": 2.0799,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.1505,
|
|
"grad_norm": 0.7739863991737366,
|
|
"learning_rate": 0.00017007007007007008,
|
|
"loss": 2.03,
|
|
"step": 1505
|
|
},
|
|
{
|
|
"epoch": 0.151,
|
|
"grad_norm": 0.8992478251457214,
|
|
"learning_rate": 0.00016996996996997,
|
|
"loss": 2.06,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 0.1515,
|
|
"grad_norm": 0.7923077344894409,
|
|
"learning_rate": 0.00016986986986986988,
|
|
"loss": 2.0449,
|
|
"step": 1515
|
|
},
|
|
{
|
|
"epoch": 0.152,
|
|
"grad_norm": 0.8039366006851196,
|
|
"learning_rate": 0.00016976976976976977,
|
|
"loss": 2.0581,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 0.1525,
|
|
"grad_norm": 0.7828361392021179,
|
|
"learning_rate": 0.00016966966966966968,
|
|
"loss": 2.0525,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 0.153,
|
|
"grad_norm": 0.7761391401290894,
|
|
"learning_rate": 0.00016956956956956957,
|
|
"loss": 2.0407,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 0.1535,
|
|
"grad_norm": 0.8189055323600769,
|
|
"learning_rate": 0.00016946946946946948,
|
|
"loss": 2.0552,
|
|
"step": 1535
|
|
},
|
|
{
|
|
"epoch": 0.154,
|
|
"grad_norm": 0.8271364569664001,
|
|
"learning_rate": 0.00016936936936936937,
|
|
"loss": 2.07,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 0.1545,
|
|
"grad_norm": 0.7701451778411865,
|
|
"learning_rate": 0.00016926926926926926,
|
|
"loss": 2.0252,
|
|
"step": 1545
|
|
},
|
|
{
|
|
"epoch": 0.155,
|
|
"grad_norm": 0.8270406723022461,
|
|
"learning_rate": 0.00016916916916916917,
|
|
"loss": 2.0465,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.1555,
|
|
"grad_norm": 0.8067330121994019,
|
|
"learning_rate": 0.00016906906906906908,
|
|
"loss": 2.0356,
|
|
"step": 1555
|
|
},
|
|
{
|
|
"epoch": 0.156,
|
|
"grad_norm": 0.870760977268219,
|
|
"learning_rate": 0.000168968968968969,
|
|
"loss": 2.0276,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 0.1565,
|
|
"grad_norm": 0.8015304207801819,
|
|
"learning_rate": 0.00016886886886886889,
|
|
"loss": 2.0566,
|
|
"step": 1565
|
|
},
|
|
{
|
|
"epoch": 0.157,
|
|
"grad_norm": 0.8079798817634583,
|
|
"learning_rate": 0.00016876876876876877,
|
|
"loss": 2.0477,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 0.1575,
|
|
"grad_norm": 0.7779944539070129,
|
|
"learning_rate": 0.00016866866866866869,
|
|
"loss": 2.0598,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 0.158,
|
|
"grad_norm": 0.8422070145606995,
|
|
"learning_rate": 0.00016856856856856857,
|
|
"loss": 2.0504,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 0.1585,
|
|
"grad_norm": 0.7895457744598389,
|
|
"learning_rate": 0.00016846846846846846,
|
|
"loss": 2.0323,
|
|
"step": 1585
|
|
},
|
|
{
|
|
"epoch": 0.159,
|
|
"grad_norm": 0.8140113949775696,
|
|
"learning_rate": 0.00016836836836836837,
|
|
"loss": 2.0498,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 0.1595,
|
|
"grad_norm": 0.7975071668624878,
|
|
"learning_rate": 0.00016826826826826826,
|
|
"loss": 2.0647,
|
|
"step": 1595
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.797199010848999,
|
|
"learning_rate": 0.00016816816816816817,
|
|
"loss": 2.0437,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.1605,
|
|
"grad_norm": 0.7758844494819641,
|
|
"learning_rate": 0.00016806806806806809,
|
|
"loss": 2.0531,
|
|
"step": 1605
|
|
},
|
|
{
|
|
"epoch": 0.161,
|
|
"grad_norm": 0.7931090593338013,
|
|
"learning_rate": 0.00016796796796796797,
|
|
"loss": 2.0446,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 0.1615,
|
|
"grad_norm": 0.7633284330368042,
|
|
"learning_rate": 0.0001678678678678679,
|
|
"loss": 2.0602,
|
|
"step": 1615
|
|
},
|
|
{
|
|
"epoch": 0.162,
|
|
"grad_norm": 0.8108896017074585,
|
|
"learning_rate": 0.00016776776776776777,
|
|
"loss": 2.0388,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 0.1625,
|
|
"grad_norm": 0.7830617427825928,
|
|
"learning_rate": 0.0001676676676676677,
|
|
"loss": 2.0684,
|
|
"step": 1625
|
|
},
|
|
{
|
|
"epoch": 0.163,
|
|
"grad_norm": 0.8587239980697632,
|
|
"learning_rate": 0.00016756756756756757,
|
|
"loss": 2.0365,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 0.1635,
|
|
"grad_norm": 0.8472581505775452,
|
|
"learning_rate": 0.00016746746746746746,
|
|
"loss": 2.0522,
|
|
"step": 1635
|
|
},
|
|
{
|
|
"epoch": 0.164,
|
|
"grad_norm": 0.9151282906532288,
|
|
"learning_rate": 0.00016736736736736737,
|
|
"loss": 2.0582,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 0.1645,
|
|
"grad_norm": 0.8431873321533203,
|
|
"learning_rate": 0.00016726726726726726,
|
|
"loss": 2.0476,
|
|
"step": 1645
|
|
},
|
|
{
|
|
"epoch": 0.165,
|
|
"grad_norm": 0.7887601256370544,
|
|
"learning_rate": 0.00016716716716716717,
|
|
"loss": 2.0366,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.1655,
|
|
"grad_norm": 0.7875514626502991,
|
|
"learning_rate": 0.00016706706706706706,
|
|
"loss": 2.0561,
|
|
"step": 1655
|
|
},
|
|
{
|
|
"epoch": 0.166,
|
|
"grad_norm": 0.8043152689933777,
|
|
"learning_rate": 0.00016696696696696697,
|
|
"loss": 2.0388,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 0.1665,
|
|
"grad_norm": 0.7784465551376343,
|
|
"learning_rate": 0.0001668668668668669,
|
|
"loss": 2.0407,
|
|
"step": 1665
|
|
},
|
|
{
|
|
"epoch": 0.167,
|
|
"grad_norm": 0.7523636221885681,
|
|
"learning_rate": 0.00016676676676676677,
|
|
"loss": 2.0507,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"epoch": 0.1675,
|
|
"grad_norm": 0.7890161871910095,
|
|
"learning_rate": 0.0001666666666666667,
|
|
"loss": 2.0511,
|
|
"step": 1675
|
|
},
|
|
{
|
|
"epoch": 0.168,
|
|
"grad_norm": 0.7951382994651794,
|
|
"learning_rate": 0.00016656656656656657,
|
|
"loss": 2.0126,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"epoch": 0.1685,
|
|
"grad_norm": 0.8491453528404236,
|
|
"learning_rate": 0.00016646646646646646,
|
|
"loss": 2.0325,
|
|
"step": 1685
|
|
},
|
|
{
|
|
"epoch": 0.169,
|
|
"grad_norm": 0.7883369326591492,
|
|
"learning_rate": 0.00016636636636636638,
|
|
"loss": 2.0238,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"epoch": 0.1695,
|
|
"grad_norm": 0.848307192325592,
|
|
"learning_rate": 0.00016626626626626626,
|
|
"loss": 2.0335,
|
|
"step": 1695
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"grad_norm": 0.7763209939002991,
|
|
"learning_rate": 0.00016616616616616618,
|
|
"loss": 2.0511,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.1705,
|
|
"grad_norm": 0.7731834650039673,
|
|
"learning_rate": 0.00016606606606606606,
|
|
"loss": 2.0417,
|
|
"step": 1705
|
|
},
|
|
{
|
|
"epoch": 0.171,
|
|
"grad_norm": 0.7851327657699585,
|
|
"learning_rate": 0.00016596596596596598,
|
|
"loss": 2.031,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"epoch": 0.1715,
|
|
"grad_norm": 0.794136106967926,
|
|
"learning_rate": 0.0001658658658658659,
|
|
"loss": 2.03,
|
|
"step": 1715
|
|
},
|
|
{
|
|
"epoch": 0.172,
|
|
"grad_norm": 0.7943106293678284,
|
|
"learning_rate": 0.00016576576576576578,
|
|
"loss": 2.0273,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 0.1725,
|
|
"grad_norm": 0.7772659659385681,
|
|
"learning_rate": 0.0001656656656656657,
|
|
"loss": 2.0424,
|
|
"step": 1725
|
|
},
|
|
{
|
|
"epoch": 0.173,
|
|
"grad_norm": 0.7858376502990723,
|
|
"learning_rate": 0.00016556556556556558,
|
|
"loss": 2.0309,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"epoch": 0.1735,
|
|
"grad_norm": 0.7929664850234985,
|
|
"learning_rate": 0.00016546546546546546,
|
|
"loss": 2.0399,
|
|
"step": 1735
|
|
},
|
|
{
|
|
"epoch": 0.174,
|
|
"grad_norm": 0.7890056371688843,
|
|
"learning_rate": 0.00016536536536536538,
|
|
"loss": 2.0413,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"epoch": 0.1745,
|
|
"grad_norm": 0.7757844924926758,
|
|
"learning_rate": 0.00016526526526526526,
|
|
"loss": 2.0487,
|
|
"step": 1745
|
|
},
|
|
{
|
|
"epoch": 0.175,
|
|
"grad_norm": 0.7959721088409424,
|
|
"learning_rate": 0.00016516516516516518,
|
|
"loss": 2.0366,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.1755,
|
|
"grad_norm": 0.7847658395767212,
|
|
"learning_rate": 0.00016506506506506506,
|
|
"loss": 2.0465,
|
|
"step": 1755
|
|
},
|
|
{
|
|
"epoch": 0.176,
|
|
"grad_norm": 0.748741865158081,
|
|
"learning_rate": 0.00016496496496496495,
|
|
"loss": 2.0494,
|
|
"step": 1760
|
|
},
|
|
{
|
|
"epoch": 0.1765,
|
|
"grad_norm": 0.7750582695007324,
|
|
"learning_rate": 0.00016486486486486486,
|
|
"loss": 2.0386,
|
|
"step": 1765
|
|
},
|
|
{
|
|
"epoch": 0.177,
|
|
"grad_norm": 0.7988344430923462,
|
|
"learning_rate": 0.00016476476476476478,
|
|
"loss": 2.0332,
|
|
"step": 1770
|
|
},
|
|
{
|
|
"epoch": 0.1775,
|
|
"grad_norm": 0.821477472782135,
|
|
"learning_rate": 0.0001646646646646647,
|
|
"loss": 2.0233,
|
|
"step": 1775
|
|
},
|
|
{
|
|
"epoch": 0.178,
|
|
"grad_norm": 0.8394650220870972,
|
|
"learning_rate": 0.00016456456456456458,
|
|
"loss": 2.0463,
|
|
"step": 1780
|
|
},
|
|
{
|
|
"epoch": 0.1785,
|
|
"grad_norm": 0.8799570798873901,
|
|
"learning_rate": 0.00016446446446446446,
|
|
"loss": 2.0214,
|
|
"step": 1785
|
|
},
|
|
{
|
|
"epoch": 0.179,
|
|
"grad_norm": 0.8210110068321228,
|
|
"learning_rate": 0.00016436436436436438,
|
|
"loss": 2.025,
|
|
"step": 1790
|
|
},
|
|
{
|
|
"epoch": 0.1795,
|
|
"grad_norm": 0.7694400548934937,
|
|
"learning_rate": 0.00016426426426426426,
|
|
"loss": 2.0122,
|
|
"step": 1795
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"grad_norm": 0.7861407995223999,
|
|
"learning_rate": 0.00016416416416416418,
|
|
"loss": 2.0387,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.1805,
|
|
"grad_norm": 0.8013924360275269,
|
|
"learning_rate": 0.00016406406406406406,
|
|
"loss": 2.0425,
|
|
"step": 1805
|
|
},
|
|
{
|
|
"epoch": 0.181,
|
|
"grad_norm": 0.7965100407600403,
|
|
"learning_rate": 0.00016396396396396395,
|
|
"loss": 2.0312,
|
|
"step": 1810
|
|
},
|
|
{
|
|
"epoch": 0.1815,
|
|
"grad_norm": 0.7945120930671692,
|
|
"learning_rate": 0.00016386386386386387,
|
|
"loss": 2.0201,
|
|
"step": 1815
|
|
},
|
|
{
|
|
"epoch": 0.182,
|
|
"grad_norm": 0.7594983577728271,
|
|
"learning_rate": 0.00016376376376376378,
|
|
"loss": 2.018,
|
|
"step": 1820
|
|
},
|
|
{
|
|
"epoch": 0.1825,
|
|
"grad_norm": 0.7923123240470886,
|
|
"learning_rate": 0.0001636636636636637,
|
|
"loss": 2.0128,
|
|
"step": 1825
|
|
},
|
|
{
|
|
"epoch": 0.183,
|
|
"grad_norm": 0.7755590081214905,
|
|
"learning_rate": 0.00016356356356356358,
|
|
"loss": 2.05,
|
|
"step": 1830
|
|
},
|
|
{
|
|
"epoch": 0.1835,
|
|
"grad_norm": 0.7651676535606384,
|
|
"learning_rate": 0.00016346346346346347,
|
|
"loss": 2.0239,
|
|
"step": 1835
|
|
},
|
|
{
|
|
"epoch": 0.184,
|
|
"grad_norm": 0.7822082042694092,
|
|
"learning_rate": 0.00016336336336336338,
|
|
"loss": 2.0271,
|
|
"step": 1840
|
|
},
|
|
{
|
|
"epoch": 0.1845,
|
|
"grad_norm": 0.7861602306365967,
|
|
"learning_rate": 0.00016326326326326327,
|
|
"loss": 2.0329,
|
|
"step": 1845
|
|
},
|
|
{
|
|
"epoch": 0.185,
|
|
"grad_norm": 0.7773391604423523,
|
|
"learning_rate": 0.00016316316316316318,
|
|
"loss": 2.0167,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.1855,
|
|
"grad_norm": 0.7753078937530518,
|
|
"learning_rate": 0.00016306306306306307,
|
|
"loss": 2.0392,
|
|
"step": 1855
|
|
},
|
|
{
|
|
"epoch": 0.186,
|
|
"grad_norm": 0.7742039561271667,
|
|
"learning_rate": 0.00016296296296296295,
|
|
"loss": 2.0163,
|
|
"step": 1860
|
|
},
|
|
{
|
|
"epoch": 0.1865,
|
|
"grad_norm": 0.7700342535972595,
|
|
"learning_rate": 0.00016286286286286287,
|
|
"loss": 2.0273,
|
|
"step": 1865
|
|
},
|
|
{
|
|
"epoch": 0.187,
|
|
"grad_norm": 0.7967638373374939,
|
|
"learning_rate": 0.00016276276276276275,
|
|
"loss": 2.0295,
|
|
"step": 1870
|
|
},
|
|
{
|
|
"epoch": 0.1875,
|
|
"grad_norm": 0.7784233689308167,
|
|
"learning_rate": 0.00016266266266266267,
|
|
"loss": 2.0112,
|
|
"step": 1875
|
|
},
|
|
{
|
|
"epoch": 0.188,
|
|
"grad_norm": 0.7581600546836853,
|
|
"learning_rate": 0.00016256256256256258,
|
|
"loss": 2.045,
|
|
"step": 1880
|
|
},
|
|
{
|
|
"epoch": 0.1885,
|
|
"grad_norm": 0.7855777144432068,
|
|
"learning_rate": 0.00016246246246246247,
|
|
"loss": 2.018,
|
|
"step": 1885
|
|
},
|
|
{
|
|
"epoch": 0.189,
|
|
"grad_norm": 0.7643173336982727,
|
|
"learning_rate": 0.00016236236236236238,
|
|
"loss": 2.0338,
|
|
"step": 1890
|
|
},
|
|
{
|
|
"epoch": 0.1895,
|
|
"grad_norm": 0.7671994566917419,
|
|
"learning_rate": 0.00016226226226226227,
|
|
"loss": 2.0306,
|
|
"step": 1895
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"grad_norm": 0.7719835638999939,
|
|
"learning_rate": 0.00016216216216216218,
|
|
"loss": 2.0347,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.1905,
|
|
"grad_norm": 0.7854582667350769,
|
|
"learning_rate": 0.00016206206206206207,
|
|
"loss": 2.0025,
|
|
"step": 1905
|
|
},
|
|
{
|
|
"epoch": 0.191,
|
|
"grad_norm": 0.7874001264572144,
|
|
"learning_rate": 0.00016196196196196195,
|
|
"loss": 2.0113,
|
|
"step": 1910
|
|
},
|
|
{
|
|
"epoch": 0.1915,
|
|
"grad_norm": 0.7835874557495117,
|
|
"learning_rate": 0.00016186186186186187,
|
|
"loss": 1.9983,
|
|
"step": 1915
|
|
},
|
|
{
|
|
"epoch": 0.192,
|
|
"grad_norm": 0.7952415943145752,
|
|
"learning_rate": 0.00016176176176176175,
|
|
"loss": 2.015,
|
|
"step": 1920
|
|
},
|
|
{
|
|
"epoch": 0.1925,
|
|
"grad_norm": 0.8112797737121582,
|
|
"learning_rate": 0.00016166166166166167,
|
|
"loss": 2.0433,
|
|
"step": 1925
|
|
},
|
|
{
|
|
"epoch": 0.193,
|
|
"grad_norm": 0.7716769576072693,
|
|
"learning_rate": 0.00016156156156156158,
|
|
"loss": 2.0229,
|
|
"step": 1930
|
|
},
|
|
{
|
|
"epoch": 0.1935,
|
|
"grad_norm": 0.8018927574157715,
|
|
"learning_rate": 0.00016146146146146147,
|
|
"loss": 2.0194,
|
|
"step": 1935
|
|
},
|
|
{
|
|
"epoch": 0.194,
|
|
"grad_norm": 0.8140080571174622,
|
|
"learning_rate": 0.00016136136136136138,
|
|
"loss": 2.0195,
|
|
"step": 1940
|
|
},
|
|
{
|
|
"epoch": 0.1945,
|
|
"grad_norm": 0.7923764586448669,
|
|
"learning_rate": 0.00016126126126126127,
|
|
"loss": 2.0137,
|
|
"step": 1945
|
|
},
|
|
{
|
|
"epoch": 0.195,
|
|
"grad_norm": 0.8428927659988403,
|
|
"learning_rate": 0.00016116116116116118,
|
|
"loss": 2.0238,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.1955,
|
|
"grad_norm": 0.8145603537559509,
|
|
"learning_rate": 0.00016106106106106107,
|
|
"loss": 2.0129,
|
|
"step": 1955
|
|
},
|
|
{
|
|
"epoch": 0.196,
|
|
"grad_norm": 0.8989366888999939,
|
|
"learning_rate": 0.00016096096096096096,
|
|
"loss": 2.0275,
|
|
"step": 1960
|
|
},
|
|
{
|
|
"epoch": 0.1965,
|
|
"grad_norm": 0.8182039856910706,
|
|
"learning_rate": 0.00016086086086086087,
|
|
"loss": 2.0044,
|
|
"step": 1965
|
|
},
|
|
{
|
|
"epoch": 0.197,
|
|
"grad_norm": 0.8926301002502441,
|
|
"learning_rate": 0.00016076076076076076,
|
|
"loss": 2.0147,
|
|
"step": 1970
|
|
},
|
|
{
|
|
"epoch": 0.1975,
|
|
"grad_norm": 0.7725608348846436,
|
|
"learning_rate": 0.00016066066066066067,
|
|
"loss": 2.0172,
|
|
"step": 1975
|
|
},
|
|
{
|
|
"epoch": 0.198,
|
|
"grad_norm": 0.8739742040634155,
|
|
"learning_rate": 0.00016056056056056056,
|
|
"loss": 2.0245,
|
|
"step": 1980
|
|
},
|
|
{
|
|
"epoch": 0.1985,
|
|
"grad_norm": 0.8591052293777466,
|
|
"learning_rate": 0.00016046046046046047,
|
|
"loss": 1.9968,
|
|
"step": 1985
|
|
},
|
|
{
|
|
"epoch": 0.199,
|
|
"grad_norm": 0.7801085114479065,
|
|
"learning_rate": 0.00016036036036036038,
|
|
"loss": 2.0064,
|
|
"step": 1990
|
|
},
|
|
{
|
|
"epoch": 0.1995,
|
|
"grad_norm": 0.8598971366882324,
|
|
"learning_rate": 0.00016026026026026027,
|
|
"loss": 2.0307,
|
|
"step": 1995
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"grad_norm": 0.8810726404190063,
|
|
"learning_rate": 0.00016016016016016018,
|
|
"loss": 2.0129,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.2005,
|
|
"grad_norm": 0.8133185505867004,
|
|
"learning_rate": 0.00016006006006006007,
|
|
"loss": 2.0047,
|
|
"step": 2005
|
|
},
|
|
{
|
|
"epoch": 0.201,
|
|
"grad_norm": 0.8219152688980103,
|
|
"learning_rate": 0.00015995995995995996,
|
|
"loss": 2.0311,
|
|
"step": 2010
|
|
},
|
|
{
|
|
"epoch": 0.2015,
|
|
"grad_norm": 0.7896487712860107,
|
|
"learning_rate": 0.00015985985985985987,
|
|
"loss": 2.0247,
|
|
"step": 2015
|
|
},
|
|
{
|
|
"epoch": 0.202,
|
|
"grad_norm": 0.9305662512779236,
|
|
"learning_rate": 0.00015975975975975976,
|
|
"loss": 2.0446,
|
|
"step": 2020
|
|
},
|
|
{
|
|
"epoch": 0.2025,
|
|
"grad_norm": 0.8045963644981384,
|
|
"learning_rate": 0.00015965965965965967,
|
|
"loss": 2.0093,
|
|
"step": 2025
|
|
},
|
|
{
|
|
"epoch": 0.203,
|
|
"grad_norm": 0.8537754416465759,
|
|
"learning_rate": 0.00015955955955955956,
|
|
"loss": 2.0215,
|
|
"step": 2030
|
|
},
|
|
{
|
|
"epoch": 0.2035,
|
|
"grad_norm": 0.7994036674499512,
|
|
"learning_rate": 0.00015945945945945947,
|
|
"loss": 1.997,
|
|
"step": 2035
|
|
},
|
|
{
|
|
"epoch": 0.204,
|
|
"grad_norm": 0.7729703187942505,
|
|
"learning_rate": 0.00015935935935935938,
|
|
"loss": 2.0095,
|
|
"step": 2040
|
|
},
|
|
{
|
|
"epoch": 0.2045,
|
|
"grad_norm": 0.7706931233406067,
|
|
"learning_rate": 0.00015925925925925927,
|
|
"loss": 2.029,
|
|
"step": 2045
|
|
},
|
|
{
|
|
"epoch": 0.205,
|
|
"grad_norm": 0.7798938155174255,
|
|
"learning_rate": 0.00015915915915915919,
|
|
"loss": 2.0325,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.2055,
|
|
"grad_norm": 0.7870948314666748,
|
|
"learning_rate": 0.00015905905905905907,
|
|
"loss": 2.0282,
|
|
"step": 2055
|
|
},
|
|
{
|
|
"epoch": 0.206,
|
|
"grad_norm": 0.8591163158416748,
|
|
"learning_rate": 0.00015895895895895896,
|
|
"loss": 2.0183,
|
|
"step": 2060
|
|
},
|
|
{
|
|
"epoch": 0.2065,
|
|
"grad_norm": 0.8115944862365723,
|
|
"learning_rate": 0.00015885885885885887,
|
|
"loss": 2.0131,
|
|
"step": 2065
|
|
},
|
|
{
|
|
"epoch": 0.207,
|
|
"grad_norm": 0.7475624680519104,
|
|
"learning_rate": 0.00015875875875875876,
|
|
"loss": 2.0175,
|
|
"step": 2070
|
|
},
|
|
{
|
|
"epoch": 0.2075,
|
|
"grad_norm": 0.7941943407058716,
|
|
"learning_rate": 0.00015865865865865867,
|
|
"loss": 2.0288,
|
|
"step": 2075
|
|
},
|
|
{
|
|
"epoch": 0.208,
|
|
"grad_norm": 0.7347163558006287,
|
|
"learning_rate": 0.00015855855855855856,
|
|
"loss": 1.9898,
|
|
"step": 2080
|
|
},
|
|
{
|
|
"epoch": 0.2085,
|
|
"grad_norm": 0.757885217666626,
|
|
"learning_rate": 0.00015845845845845845,
|
|
"loss": 1.9989,
|
|
"step": 2085
|
|
},
|
|
{
|
|
"epoch": 0.209,
|
|
"grad_norm": 0.7636198401451111,
|
|
"learning_rate": 0.00015835835835835836,
|
|
"loss": 2.0141,
|
|
"step": 2090
|
|
},
|
|
{
|
|
"epoch": 0.2095,
|
|
"grad_norm": 0.8460286855697632,
|
|
"learning_rate": 0.00015825825825825827,
|
|
"loss": 1.9916,
|
|
"step": 2095
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"grad_norm": 0.7832396626472473,
|
|
"learning_rate": 0.0001581581581581582,
|
|
"loss": 2.0153,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.2105,
|
|
"grad_norm": 0.7886310815811157,
|
|
"learning_rate": 0.00015805805805805807,
|
|
"loss": 2.0428,
|
|
"step": 2105
|
|
},
|
|
{
|
|
"epoch": 0.211,
|
|
"grad_norm": 0.7736825942993164,
|
|
"learning_rate": 0.00015795795795795796,
|
|
"loss": 2.0113,
|
|
"step": 2110
|
|
},
|
|
{
|
|
"epoch": 0.2115,
|
|
"grad_norm": 0.7903079986572266,
|
|
"learning_rate": 0.00015785785785785787,
|
|
"loss": 2.0105,
|
|
"step": 2115
|
|
},
|
|
{
|
|
"epoch": 0.212,
|
|
"grad_norm": 0.8193835616111755,
|
|
"learning_rate": 0.00015775775775775776,
|
|
"loss": 2.0123,
|
|
"step": 2120
|
|
},
|
|
{
|
|
"epoch": 0.2125,
|
|
"grad_norm": 0.8024186491966248,
|
|
"learning_rate": 0.00015765765765765767,
|
|
"loss": 2.0069,
|
|
"step": 2125
|
|
},
|
|
{
|
|
"epoch": 0.213,
|
|
"grad_norm": 0.7742881178855896,
|
|
"learning_rate": 0.00015755755755755756,
|
|
"loss": 2.0106,
|
|
"step": 2130
|
|
},
|
|
{
|
|
"epoch": 0.2135,
|
|
"grad_norm": 0.8352553844451904,
|
|
"learning_rate": 0.00015745745745745745,
|
|
"loss": 1.9989,
|
|
"step": 2135
|
|
},
|
|
{
|
|
"epoch": 0.214,
|
|
"grad_norm": 0.7888927459716797,
|
|
"learning_rate": 0.00015735735735735736,
|
|
"loss": 1.9939,
|
|
"step": 2140
|
|
},
|
|
{
|
|
"epoch": 0.2145,
|
|
"grad_norm": 0.811654806137085,
|
|
"learning_rate": 0.00015725725725725727,
|
|
"loss": 2.0194,
|
|
"step": 2145
|
|
},
|
|
{
|
|
"epoch": 0.215,
|
|
"grad_norm": 0.8414866924285889,
|
|
"learning_rate": 0.0001571571571571572,
|
|
"loss": 1.9964,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.2155,
|
|
"grad_norm": 0.7777667045593262,
|
|
"learning_rate": 0.00015705705705705707,
|
|
"loss": 2.0231,
|
|
"step": 2155
|
|
},
|
|
{
|
|
"epoch": 0.216,
|
|
"grad_norm": 0.782187819480896,
|
|
"learning_rate": 0.00015695695695695696,
|
|
"loss": 2.0235,
|
|
"step": 2160
|
|
},
|
|
{
|
|
"epoch": 0.2165,
|
|
"grad_norm": 0.7661353945732117,
|
|
"learning_rate": 0.00015685685685685687,
|
|
"loss": 1.9932,
|
|
"step": 2165
|
|
},
|
|
{
|
|
"epoch": 0.217,
|
|
"grad_norm": 0.8571522235870361,
|
|
"learning_rate": 0.00015675675675675676,
|
|
"loss": 2.017,
|
|
"step": 2170
|
|
},
|
|
{
|
|
"epoch": 0.2175,
|
|
"grad_norm": 0.795987606048584,
|
|
"learning_rate": 0.00015665665665665668,
|
|
"loss": 2.0095,
|
|
"step": 2175
|
|
},
|
|
{
|
|
"epoch": 0.218,
|
|
"grad_norm": 0.788629412651062,
|
|
"learning_rate": 0.00015655655655655656,
|
|
"loss": 1.9961,
|
|
"step": 2180
|
|
},
|
|
{
|
|
"epoch": 0.2185,
|
|
"grad_norm": 0.7576528787612915,
|
|
"learning_rate": 0.00015645645645645645,
|
|
"loss": 1.9911,
|
|
"step": 2185
|
|
},
|
|
{
|
|
"epoch": 0.219,
|
|
"grad_norm": 0.7708572149276733,
|
|
"learning_rate": 0.00015635635635635636,
|
|
"loss": 1.9792,
|
|
"step": 2190
|
|
},
|
|
{
|
|
"epoch": 0.2195,
|
|
"grad_norm": 0.8041749000549316,
|
|
"learning_rate": 0.00015625625625625625,
|
|
"loss": 2.006,
|
|
"step": 2195
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"grad_norm": 0.8431360125541687,
|
|
"learning_rate": 0.00015615615615615616,
|
|
"loss": 2.0173,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.2205,
|
|
"grad_norm": 0.8387967348098755,
|
|
"learning_rate": 0.00015605605605605608,
|
|
"loss": 1.9973,
|
|
"step": 2205
|
|
},
|
|
{
|
|
"epoch": 0.221,
|
|
"grad_norm": 0.8072513341903687,
|
|
"learning_rate": 0.00015595595595595596,
|
|
"loss": 1.9921,
|
|
"step": 2210
|
|
},
|
|
{
|
|
"epoch": 0.2215,
|
|
"grad_norm": 0.8014965057373047,
|
|
"learning_rate": 0.00015585585585585588,
|
|
"loss": 1.9881,
|
|
"step": 2215
|
|
},
|
|
{
|
|
"epoch": 0.222,
|
|
"grad_norm": 0.7723069787025452,
|
|
"learning_rate": 0.00015575575575575576,
|
|
"loss": 1.9993,
|
|
"step": 2220
|
|
},
|
|
{
|
|
"epoch": 0.2225,
|
|
"grad_norm": 0.8190085887908936,
|
|
"learning_rate": 0.00015565565565565565,
|
|
"loss": 2.0015,
|
|
"step": 2225
|
|
},
|
|
{
|
|
"epoch": 0.223,
|
|
"grad_norm": 0.7893680334091187,
|
|
"learning_rate": 0.00015555555555555556,
|
|
"loss": 1.9983,
|
|
"step": 2230
|
|
},
|
|
{
|
|
"epoch": 0.2235,
|
|
"grad_norm": 0.7632356286048889,
|
|
"learning_rate": 0.00015545545545545545,
|
|
"loss": 2.0039,
|
|
"step": 2235
|
|
},
|
|
{
|
|
"epoch": 0.224,
|
|
"grad_norm": 0.7753027677536011,
|
|
"learning_rate": 0.00015535535535535536,
|
|
"loss": 1.9878,
|
|
"step": 2240
|
|
},
|
|
{
|
|
"epoch": 0.2245,
|
|
"grad_norm": 0.7894906997680664,
|
|
"learning_rate": 0.00015525525525525525,
|
|
"loss": 1.997,
|
|
"step": 2245
|
|
},
|
|
{
|
|
"epoch": 0.225,
|
|
"grad_norm": 0.8126510977745056,
|
|
"learning_rate": 0.00015515515515515516,
|
|
"loss": 1.9966,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.2255,
|
|
"grad_norm": 0.8427388668060303,
|
|
"learning_rate": 0.00015505505505505508,
|
|
"loss": 2.0033,
|
|
"step": 2255
|
|
},
|
|
{
|
|
"epoch": 0.226,
|
|
"grad_norm": 0.8301752805709839,
|
|
"learning_rate": 0.00015495495495495496,
|
|
"loss": 2.0081,
|
|
"step": 2260
|
|
},
|
|
{
|
|
"epoch": 0.2265,
|
|
"grad_norm": 0.7624394297599792,
|
|
"learning_rate": 0.00015485485485485488,
|
|
"loss": 2.0072,
|
|
"step": 2265
|
|
},
|
|
{
|
|
"epoch": 0.227,
|
|
"grad_norm": 0.7501481175422668,
|
|
"learning_rate": 0.00015475475475475476,
|
|
"loss": 2.0077,
|
|
"step": 2270
|
|
},
|
|
{
|
|
"epoch": 0.2275,
|
|
"grad_norm": 0.7723734378814697,
|
|
"learning_rate": 0.00015465465465465465,
|
|
"loss": 1.9989,
|
|
"step": 2275
|
|
},
|
|
{
|
|
"epoch": 0.228,
|
|
"grad_norm": 0.7647537589073181,
|
|
"learning_rate": 0.00015455455455455456,
|
|
"loss": 1.9978,
|
|
"step": 2280
|
|
},
|
|
{
|
|
"epoch": 0.2285,
|
|
"grad_norm": 0.7880849838256836,
|
|
"learning_rate": 0.00015445445445445445,
|
|
"loss": 1.9774,
|
|
"step": 2285
|
|
},
|
|
{
|
|
"epoch": 0.229,
|
|
"grad_norm": 0.7714458107948303,
|
|
"learning_rate": 0.00015435435435435436,
|
|
"loss": 1.9799,
|
|
"step": 2290
|
|
},
|
|
{
|
|
"epoch": 0.2295,
|
|
"grad_norm": 0.7856618762016296,
|
|
"learning_rate": 0.00015425425425425425,
|
|
"loss": 1.9794,
|
|
"step": 2295
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"grad_norm": 0.813881516456604,
|
|
"learning_rate": 0.00015415415415415414,
|
|
"loss": 2.0075,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 0.2305,
|
|
"grad_norm": 0.7762358784675598,
|
|
"learning_rate": 0.00015405405405405405,
|
|
"loss": 2.0185,
|
|
"step": 2305
|
|
},
|
|
{
|
|
"epoch": 0.231,
|
|
"grad_norm": 0.766732931137085,
|
|
"learning_rate": 0.00015395395395395397,
|
|
"loss": 1.9986,
|
|
"step": 2310
|
|
},
|
|
{
|
|
"epoch": 0.2315,
|
|
"grad_norm": 0.7683066129684448,
|
|
"learning_rate": 0.00015385385385385388,
|
|
"loss": 2.0131,
|
|
"step": 2315
|
|
},
|
|
{
|
|
"epoch": 0.232,
|
|
"grad_norm": 0.7804152369499207,
|
|
"learning_rate": 0.00015375375375375377,
|
|
"loss": 2.0089,
|
|
"step": 2320
|
|
},
|
|
{
|
|
"epoch": 0.2325,
|
|
"grad_norm": 0.7836987376213074,
|
|
"learning_rate": 0.00015365365365365365,
|
|
"loss": 1.9976,
|
|
"step": 2325
|
|
},
|
|
{
|
|
"epoch": 0.233,
|
|
"grad_norm": 0.823605477809906,
|
|
"learning_rate": 0.00015355355355355357,
|
|
"loss": 2.0005,
|
|
"step": 2330
|
|
},
|
|
{
|
|
"epoch": 0.2335,
|
|
"grad_norm": 0.8343596458435059,
|
|
"learning_rate": 0.00015345345345345345,
|
|
"loss": 2.015,
|
|
"step": 2335
|
|
},
|
|
{
|
|
"epoch": 0.234,
|
|
"grad_norm": 0.8635172247886658,
|
|
"learning_rate": 0.00015335335335335337,
|
|
"loss": 2.0002,
|
|
"step": 2340
|
|
},
|
|
{
|
|
"epoch": 0.2345,
|
|
"grad_norm": 0.840971052646637,
|
|
"learning_rate": 0.00015325325325325325,
|
|
"loss": 1.9847,
|
|
"step": 2345
|
|
},
|
|
{
|
|
"epoch": 0.235,
|
|
"grad_norm": 0.888761043548584,
|
|
"learning_rate": 0.00015315315315315314,
|
|
"loss": 1.9879,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 0.2355,
|
|
"grad_norm": 0.7987576723098755,
|
|
"learning_rate": 0.00015305305305305305,
|
|
"loss": 1.9885,
|
|
"step": 2355
|
|
},
|
|
{
|
|
"epoch": 0.236,
|
|
"grad_norm": 0.781504213809967,
|
|
"learning_rate": 0.00015295295295295297,
|
|
"loss": 1.9891,
|
|
"step": 2360
|
|
},
|
|
{
|
|
"epoch": 0.2365,
|
|
"grad_norm": 0.8169000148773193,
|
|
"learning_rate": 0.00015285285285285288,
|
|
"loss": 1.974,
|
|
"step": 2365
|
|
},
|
|
{
|
|
"epoch": 0.237,
|
|
"grad_norm": 0.764177680015564,
|
|
"learning_rate": 0.00015275275275275277,
|
|
"loss": 1.974,
|
|
"step": 2370
|
|
},
|
|
{
|
|
"epoch": 0.2375,
|
|
"grad_norm": 0.8093706369400024,
|
|
"learning_rate": 0.00015265265265265265,
|
|
"loss": 1.9897,
|
|
"step": 2375
|
|
},
|
|
{
|
|
"epoch": 0.238,
|
|
"grad_norm": 0.8349776268005371,
|
|
"learning_rate": 0.00015255255255255257,
|
|
"loss": 2.0002,
|
|
"step": 2380
|
|
},
|
|
{
|
|
"epoch": 0.2385,
|
|
"grad_norm": 0.831120491027832,
|
|
"learning_rate": 0.00015245245245245245,
|
|
"loss": 1.9978,
|
|
"step": 2385
|
|
},
|
|
{
|
|
"epoch": 0.239,
|
|
"grad_norm": 0.79282546043396,
|
|
"learning_rate": 0.00015235235235235237,
|
|
"loss": 2.003,
|
|
"step": 2390
|
|
},
|
|
{
|
|
"epoch": 0.2395,
|
|
"grad_norm": 0.7896937727928162,
|
|
"learning_rate": 0.00015225225225225225,
|
|
"loss": 1.9832,
|
|
"step": 2395
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"grad_norm": 0.7645019292831421,
|
|
"learning_rate": 0.00015215215215215214,
|
|
"loss": 1.983,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 0.2405,
|
|
"grad_norm": 0.809862494468689,
|
|
"learning_rate": 0.00015205205205205205,
|
|
"loss": 1.9912,
|
|
"step": 2405
|
|
},
|
|
{
|
|
"epoch": 0.241,
|
|
"grad_norm": 0.7760394215583801,
|
|
"learning_rate": 0.00015195195195195194,
|
|
"loss": 1.978,
|
|
"step": 2410
|
|
},
|
|
{
|
|
"epoch": 0.2415,
|
|
"grad_norm": 0.8623340129852295,
|
|
"learning_rate": 0.00015185185185185185,
|
|
"loss": 1.9902,
|
|
"step": 2415
|
|
},
|
|
{
|
|
"epoch": 0.242,
|
|
"grad_norm": 0.8047596216201782,
|
|
"learning_rate": 0.00015175175175175177,
|
|
"loss": 2.011,
|
|
"step": 2420
|
|
},
|
|
{
|
|
"epoch": 0.2425,
|
|
"grad_norm": 0.7739080786705017,
|
|
"learning_rate": 0.00015165165165165165,
|
|
"loss": 1.981,
|
|
"step": 2425
|
|
},
|
|
{
|
|
"epoch": 0.243,
|
|
"grad_norm": 0.7927338480949402,
|
|
"learning_rate": 0.00015155155155155157,
|
|
"loss": 1.9814,
|
|
"step": 2430
|
|
},
|
|
{
|
|
"epoch": 0.2435,
|
|
"grad_norm": 0.8810697197914124,
|
|
"learning_rate": 0.00015145145145145146,
|
|
"loss": 1.9664,
|
|
"step": 2435
|
|
},
|
|
{
|
|
"epoch": 0.244,
|
|
"grad_norm": 0.860685408115387,
|
|
"learning_rate": 0.00015135135135135137,
|
|
"loss": 1.9842,
|
|
"step": 2440
|
|
},
|
|
{
|
|
"epoch": 0.2445,
|
|
"grad_norm": 0.9384279847145081,
|
|
"learning_rate": 0.00015125125125125126,
|
|
"loss": 1.9657,
|
|
"step": 2445
|
|
},
|
|
{
|
|
"epoch": 0.245,
|
|
"grad_norm": 0.9265758395195007,
|
|
"learning_rate": 0.00015115115115115114,
|
|
"loss": 1.99,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 0.2455,
|
|
"grad_norm": 0.8147779703140259,
|
|
"learning_rate": 0.00015105105105105106,
|
|
"loss": 1.9825,
|
|
"step": 2455
|
|
},
|
|
{
|
|
"epoch": 0.246,
|
|
"grad_norm": 0.7943310141563416,
|
|
"learning_rate": 0.00015095095095095094,
|
|
"loss": 1.9886,
|
|
"step": 2460
|
|
},
|
|
{
|
|
"epoch": 0.2465,
|
|
"grad_norm": 0.8120606541633606,
|
|
"learning_rate": 0.00015085085085085086,
|
|
"loss": 1.9716,
|
|
"step": 2465
|
|
},
|
|
{
|
|
"epoch": 0.247,
|
|
"grad_norm": 0.7622060179710388,
|
|
"learning_rate": 0.00015075075075075077,
|
|
"loss": 1.9959,
|
|
"step": 2470
|
|
},
|
|
{
|
|
"epoch": 0.2475,
|
|
"grad_norm": 0.8406639695167542,
|
|
"learning_rate": 0.00015065065065065066,
|
|
"loss": 1.9895,
|
|
"step": 2475
|
|
},
|
|
{
|
|
"epoch": 0.248,
|
|
"grad_norm": 0.786904513835907,
|
|
"learning_rate": 0.00015055055055055057,
|
|
"loss": 1.989,
|
|
"step": 2480
|
|
},
|
|
{
|
|
"epoch": 0.2485,
|
|
"grad_norm": 0.8092778325080872,
|
|
"learning_rate": 0.00015045045045045046,
|
|
"loss": 1.9998,
|
|
"step": 2485
|
|
},
|
|
{
|
|
"epoch": 0.249,
|
|
"grad_norm": 0.826165497303009,
|
|
"learning_rate": 0.00015035035035035037,
|
|
"loss": 2.0039,
|
|
"step": 2490
|
|
},
|
|
{
|
|
"epoch": 0.2495,
|
|
"grad_norm": 0.786272406578064,
|
|
"learning_rate": 0.00015025025025025026,
|
|
"loss": 1.9731,
|
|
"step": 2495
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"grad_norm": 0.7633465528488159,
|
|
"learning_rate": 0.00015015015015015014,
|
|
"loss": 1.982,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.2505,
|
|
"grad_norm": 0.7651098966598511,
|
|
"learning_rate": 0.00015005005005005006,
|
|
"loss": 1.9942,
|
|
"step": 2505
|
|
},
|
|
{
|
|
"epoch": 0.251,
|
|
"grad_norm": 0.785978376865387,
|
|
"learning_rate": 0.00014994994994994994,
|
|
"loss": 2.0017,
|
|
"step": 2510
|
|
},
|
|
{
|
|
"epoch": 0.2515,
|
|
"grad_norm": 0.8050821423530579,
|
|
"learning_rate": 0.00014984984984984986,
|
|
"loss": 1.9862,
|
|
"step": 2515
|
|
},
|
|
{
|
|
"epoch": 0.252,
|
|
"grad_norm": 0.8367516398429871,
|
|
"learning_rate": 0.00014974974974974974,
|
|
"loss": 1.9739,
|
|
"step": 2520
|
|
},
|
|
{
|
|
"epoch": 0.2525,
|
|
"grad_norm": 0.8357223272323608,
|
|
"learning_rate": 0.00014964964964964966,
|
|
"loss": 2.0004,
|
|
"step": 2525
|
|
},
|
|
{
|
|
"epoch": 0.253,
|
|
"grad_norm": 0.8153281211853027,
|
|
"learning_rate": 0.00014954954954954957,
|
|
"loss": 1.972,
|
|
"step": 2530
|
|
},
|
|
{
|
|
"epoch": 0.2535,
|
|
"grad_norm": 0.784550130367279,
|
|
"learning_rate": 0.00014944944944944946,
|
|
"loss": 1.9746,
|
|
"step": 2535
|
|
},
|
|
{
|
|
"epoch": 0.254,
|
|
"grad_norm": 0.8016824126243591,
|
|
"learning_rate": 0.00014934934934934937,
|
|
"loss": 1.9947,
|
|
"step": 2540
|
|
},
|
|
{
|
|
"epoch": 0.2545,
|
|
"grad_norm": 0.767263650894165,
|
|
"learning_rate": 0.00014924924924924926,
|
|
"loss": 1.9629,
|
|
"step": 2545
|
|
},
|
|
{
|
|
"epoch": 0.255,
|
|
"grad_norm": 0.7741748690605164,
|
|
"learning_rate": 0.00014914914914914914,
|
|
"loss": 1.9824,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 0.2555,
|
|
"grad_norm": 0.7979781627655029,
|
|
"learning_rate": 0.00014904904904904906,
|
|
"loss": 1.9689,
|
|
"step": 2555
|
|
},
|
|
{
|
|
"epoch": 0.256,
|
|
"grad_norm": 0.7703980207443237,
|
|
"learning_rate": 0.00014894894894894895,
|
|
"loss": 1.9789,
|
|
"step": 2560
|
|
},
|
|
{
|
|
"epoch": 0.2565,
|
|
"grad_norm": 0.7330608367919922,
|
|
"learning_rate": 0.00014884884884884886,
|
|
"loss": 1.9837,
|
|
"step": 2565
|
|
},
|
|
{
|
|
"epoch": 0.257,
|
|
"grad_norm": 0.8055962324142456,
|
|
"learning_rate": 0.00014874874874874875,
|
|
"loss": 1.9865,
|
|
"step": 2570
|
|
},
|
|
{
|
|
"epoch": 0.2575,
|
|
"grad_norm": 0.7900416254997253,
|
|
"learning_rate": 0.00014864864864864866,
|
|
"loss": 1.9916,
|
|
"step": 2575
|
|
},
|
|
{
|
|
"epoch": 0.258,
|
|
"grad_norm": 0.8404157161712646,
|
|
"learning_rate": 0.00014854854854854857,
|
|
"loss": 1.9719,
|
|
"step": 2580
|
|
},
|
|
{
|
|
"epoch": 0.2585,
|
|
"grad_norm": 0.8460199236869812,
|
|
"learning_rate": 0.00014844844844844846,
|
|
"loss": 1.9903,
|
|
"step": 2585
|
|
},
|
|
{
|
|
"epoch": 0.259,
|
|
"grad_norm": 0.771557092666626,
|
|
"learning_rate": 0.00014834834834834837,
|
|
"loss": 1.9923,
|
|
"step": 2590
|
|
},
|
|
{
|
|
"epoch": 0.2595,
|
|
"grad_norm": 0.8946308493614197,
|
|
"learning_rate": 0.00014824824824824826,
|
|
"loss": 1.9886,
|
|
"step": 2595
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"grad_norm": 0.813717782497406,
|
|
"learning_rate": 0.00014814814814814815,
|
|
"loss": 2.0069,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 0.2605,
|
|
"grad_norm": 0.8274906277656555,
|
|
"learning_rate": 0.00014804804804804806,
|
|
"loss": 1.9853,
|
|
"step": 2605
|
|
},
|
|
{
|
|
"epoch": 0.261,
|
|
"grad_norm": 0.7455735802650452,
|
|
"learning_rate": 0.00014794794794794795,
|
|
"loss": 1.9851,
|
|
"step": 2610
|
|
},
|
|
{
|
|
"epoch": 0.2615,
|
|
"grad_norm": 0.7664053440093994,
|
|
"learning_rate": 0.00014784784784784786,
|
|
"loss": 2.0061,
|
|
"step": 2615
|
|
},
|
|
{
|
|
"epoch": 0.262,
|
|
"grad_norm": 0.7694225311279297,
|
|
"learning_rate": 0.00014774774774774775,
|
|
"loss": 1.9826,
|
|
"step": 2620
|
|
},
|
|
{
|
|
"epoch": 0.2625,
|
|
"grad_norm": 0.8089345097541809,
|
|
"learning_rate": 0.00014764764764764763,
|
|
"loss": 1.9734,
|
|
"step": 2625
|
|
},
|
|
{
|
|
"epoch": 0.263,
|
|
"grad_norm": 0.8007687926292419,
|
|
"learning_rate": 0.00014754754754754755,
|
|
"loss": 1.9949,
|
|
"step": 2630
|
|
},
|
|
{
|
|
"epoch": 0.2635,
|
|
"grad_norm": 0.8104939460754395,
|
|
"learning_rate": 0.00014744744744744746,
|
|
"loss": 1.9826,
|
|
"step": 2635
|
|
},
|
|
{
|
|
"epoch": 0.264,
|
|
"grad_norm": 0.7715001106262207,
|
|
"learning_rate": 0.00014734734734734737,
|
|
"loss": 1.9773,
|
|
"step": 2640
|
|
},
|
|
{
|
|
"epoch": 0.2645,
|
|
"grad_norm": 0.8055933713912964,
|
|
"learning_rate": 0.00014724724724724726,
|
|
"loss": 1.9919,
|
|
"step": 2645
|
|
},
|
|
{
|
|
"epoch": 0.265,
|
|
"grad_norm": 0.7547632455825806,
|
|
"learning_rate": 0.00014714714714714715,
|
|
"loss": 1.9808,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 0.2655,
|
|
"grad_norm": 0.7735660076141357,
|
|
"learning_rate": 0.00014704704704704706,
|
|
"loss": 1.9856,
|
|
"step": 2655
|
|
},
|
|
{
|
|
"epoch": 0.266,
|
|
"grad_norm": 0.773932933807373,
|
|
"learning_rate": 0.00014694694694694695,
|
|
"loss": 1.9599,
|
|
"step": 2660
|
|
},
|
|
{
|
|
"epoch": 0.2665,
|
|
"grad_norm": 0.7784350514411926,
|
|
"learning_rate": 0.00014684684684684686,
|
|
"loss": 1.9867,
|
|
"step": 2665
|
|
},
|
|
{
|
|
"epoch": 0.267,
|
|
"grad_norm": 0.8424972295761108,
|
|
"learning_rate": 0.00014674674674674675,
|
|
"loss": 1.9754,
|
|
"step": 2670
|
|
},
|
|
{
|
|
"epoch": 0.2675,
|
|
"grad_norm": 0.7852407693862915,
|
|
"learning_rate": 0.00014664664664664663,
|
|
"loss": 1.9882,
|
|
"step": 2675
|
|
},
|
|
{
|
|
"epoch": 0.268,
|
|
"grad_norm": 0.8432037830352783,
|
|
"learning_rate": 0.00014654654654654655,
|
|
"loss": 1.9781,
|
|
"step": 2680
|
|
},
|
|
{
|
|
"epoch": 0.2685,
|
|
"grad_norm": 0.7774507999420166,
|
|
"learning_rate": 0.00014644644644644646,
|
|
"loss": 2.0009,
|
|
"step": 2685
|
|
},
|
|
{
|
|
"epoch": 0.269,
|
|
"grad_norm": 0.7801546454429626,
|
|
"learning_rate": 0.00014634634634634638,
|
|
"loss": 1.966,
|
|
"step": 2690
|
|
},
|
|
{
|
|
"epoch": 0.2695,
|
|
"grad_norm": 0.7551014423370361,
|
|
"learning_rate": 0.00014624624624624626,
|
|
"loss": 1.9904,
|
|
"step": 2695
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"grad_norm": 0.8055344820022583,
|
|
"learning_rate": 0.00014614614614614615,
|
|
"loss": 1.966,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 0.2705,
|
|
"grad_norm": 0.7711438536643982,
|
|
"learning_rate": 0.00014604604604604606,
|
|
"loss": 1.9731,
|
|
"step": 2705
|
|
},
|
|
{
|
|
"epoch": 0.271,
|
|
"grad_norm": 0.7846922278404236,
|
|
"learning_rate": 0.00014594594594594595,
|
|
"loss": 1.9904,
|
|
"step": 2710
|
|
},
|
|
{
|
|
"epoch": 0.2715,
|
|
"grad_norm": 0.7844250202178955,
|
|
"learning_rate": 0.00014584584584584586,
|
|
"loss": 1.9778,
|
|
"step": 2715
|
|
},
|
|
{
|
|
"epoch": 0.272,
|
|
"grad_norm": 0.7908750176429749,
|
|
"learning_rate": 0.00014574574574574575,
|
|
"loss": 1.9892,
|
|
"step": 2720
|
|
},
|
|
{
|
|
"epoch": 0.2725,
|
|
"grad_norm": 0.759119987487793,
|
|
"learning_rate": 0.00014564564564564564,
|
|
"loss": 1.9888,
|
|
"step": 2725
|
|
},
|
|
{
|
|
"epoch": 0.273,
|
|
"grad_norm": 0.7900999188423157,
|
|
"learning_rate": 0.00014554554554554555,
|
|
"loss": 1.9869,
|
|
"step": 2730
|
|
},
|
|
{
|
|
"epoch": 0.2735,
|
|
"grad_norm": 0.8262436389923096,
|
|
"learning_rate": 0.00014544544544544544,
|
|
"loss": 1.9966,
|
|
"step": 2735
|
|
},
|
|
{
|
|
"epoch": 0.274,
|
|
"grad_norm": 0.753371000289917,
|
|
"learning_rate": 0.00014534534534534535,
|
|
"loss": 1.9793,
|
|
"step": 2740
|
|
},
|
|
{
|
|
"epoch": 0.2745,
|
|
"grad_norm": 0.8318740725517273,
|
|
"learning_rate": 0.00014524524524524526,
|
|
"loss": 1.9733,
|
|
"step": 2745
|
|
},
|
|
{
|
|
"epoch": 0.275,
|
|
"grad_norm": 0.784440279006958,
|
|
"learning_rate": 0.00014514514514514515,
|
|
"loss": 1.9677,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 0.2755,
|
|
"grad_norm": 0.7962784767150879,
|
|
"learning_rate": 0.00014504504504504506,
|
|
"loss": 1.9672,
|
|
"step": 2755
|
|
},
|
|
{
|
|
"epoch": 0.276,
|
|
"grad_norm": 0.7986271381378174,
|
|
"learning_rate": 0.00014494494494494495,
|
|
"loss": 1.9693,
|
|
"step": 2760
|
|
},
|
|
{
|
|
"epoch": 0.2765,
|
|
"grad_norm": 0.742120623588562,
|
|
"learning_rate": 0.00014484484484484486,
|
|
"loss": 1.9633,
|
|
"step": 2765
|
|
},
|
|
{
|
|
"epoch": 0.277,
|
|
"grad_norm": 0.8292919397354126,
|
|
"learning_rate": 0.00014474474474474475,
|
|
"loss": 1.9675,
|
|
"step": 2770
|
|
},
|
|
{
|
|
"epoch": 0.2775,
|
|
"grad_norm": 0.8410685062408447,
|
|
"learning_rate": 0.00014464464464464464,
|
|
"loss": 1.9717,
|
|
"step": 2775
|
|
},
|
|
{
|
|
"epoch": 0.278,
|
|
"grad_norm": 0.9076857566833496,
|
|
"learning_rate": 0.00014454454454454455,
|
|
"loss": 1.9846,
|
|
"step": 2780
|
|
},
|
|
{
|
|
"epoch": 0.2785,
|
|
"grad_norm": 0.7975298762321472,
|
|
"learning_rate": 0.00014444444444444444,
|
|
"loss": 1.9443,
|
|
"step": 2785
|
|
},
|
|
{
|
|
"epoch": 0.279,
|
|
"grad_norm": 0.9177993535995483,
|
|
"learning_rate": 0.00014434434434434435,
|
|
"loss": 1.9672,
|
|
"step": 2790
|
|
},
|
|
{
|
|
"epoch": 0.2795,
|
|
"grad_norm": 1.0301792621612549,
|
|
"learning_rate": 0.00014424424424424427,
|
|
"loss": 1.9699,
|
|
"step": 2795
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"grad_norm": 1.0027192831039429,
|
|
"learning_rate": 0.00014414414414414415,
|
|
"loss": 1.9778,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 0.2805,
|
|
"grad_norm": 0.8336755037307739,
|
|
"learning_rate": 0.00014404404404404407,
|
|
"loss": 1.9799,
|
|
"step": 2805
|
|
},
|
|
{
|
|
"epoch": 0.281,
|
|
"grad_norm": 0.8099704384803772,
|
|
"learning_rate": 0.00014394394394394395,
|
|
"loss": 1.9721,
|
|
"step": 2810
|
|
},
|
|
{
|
|
"epoch": 0.2815,
|
|
"grad_norm": 0.7808539867401123,
|
|
"learning_rate": 0.00014384384384384387,
|
|
"loss": 1.9807,
|
|
"step": 2815
|
|
},
|
|
{
|
|
"epoch": 0.282,
|
|
"grad_norm": 0.7895110845565796,
|
|
"learning_rate": 0.00014374374374374375,
|
|
"loss": 1.9698,
|
|
"step": 2820
|
|
},
|
|
{
|
|
"epoch": 0.2825,
|
|
"grad_norm": 0.7536469101905823,
|
|
"learning_rate": 0.00014364364364364364,
|
|
"loss": 1.9707,
|
|
"step": 2825
|
|
},
|
|
{
|
|
"epoch": 0.283,
|
|
"grad_norm": 0.8078683614730835,
|
|
"learning_rate": 0.00014354354354354355,
|
|
"loss": 1.9564,
|
|
"step": 2830
|
|
},
|
|
{
|
|
"epoch": 0.2835,
|
|
"grad_norm": 0.7899840474128723,
|
|
"learning_rate": 0.00014344344344344344,
|
|
"loss": 1.9601,
|
|
"step": 2835
|
|
},
|
|
{
|
|
"epoch": 0.284,
|
|
"grad_norm": 0.7988283634185791,
|
|
"learning_rate": 0.00014334334334334333,
|
|
"loss": 1.9826,
|
|
"step": 2840
|
|
},
|
|
{
|
|
"epoch": 0.2845,
|
|
"grad_norm": 0.8281169533729553,
|
|
"learning_rate": 0.00014324324324324324,
|
|
"loss": 1.9715,
|
|
"step": 2845
|
|
},
|
|
{
|
|
"epoch": 0.285,
|
|
"grad_norm": 0.7911503911018372,
|
|
"learning_rate": 0.00014314314314314315,
|
|
"loss": 1.9733,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 0.2855,
|
|
"grad_norm": 0.7935109734535217,
|
|
"learning_rate": 0.00014304304304304307,
|
|
"loss": 1.9559,
|
|
"step": 2855
|
|
},
|
|
{
|
|
"epoch": 0.286,
|
|
"grad_norm": 0.8028689026832581,
|
|
"learning_rate": 0.00014294294294294295,
|
|
"loss": 1.9719,
|
|
"step": 2860
|
|
},
|
|
{
|
|
"epoch": 0.2865,
|
|
"grad_norm": 0.7638905048370361,
|
|
"learning_rate": 0.00014284284284284284,
|
|
"loss": 1.9718,
|
|
"step": 2865
|
|
},
|
|
{
|
|
"epoch": 0.287,
|
|
"grad_norm": 0.7452086806297302,
|
|
"learning_rate": 0.00014274274274274275,
|
|
"loss": 1.972,
|
|
"step": 2870
|
|
},
|
|
{
|
|
"epoch": 0.2875,
|
|
"grad_norm": 0.7905741930007935,
|
|
"learning_rate": 0.00014264264264264264,
|
|
"loss": 1.9583,
|
|
"step": 2875
|
|
},
|
|
{
|
|
"epoch": 0.288,
|
|
"grad_norm": 0.7474929690361023,
|
|
"learning_rate": 0.00014254254254254255,
|
|
"loss": 1.9743,
|
|
"step": 2880
|
|
},
|
|
{
|
|
"epoch": 0.2885,
|
|
"grad_norm": 0.8252883553504944,
|
|
"learning_rate": 0.00014244244244244244,
|
|
"loss": 1.941,
|
|
"step": 2885
|
|
},
|
|
{
|
|
"epoch": 0.289,
|
|
"grad_norm": 0.8231001496315002,
|
|
"learning_rate": 0.00014234234234234233,
|
|
"loss": 1.9683,
|
|
"step": 2890
|
|
},
|
|
{
|
|
"epoch": 0.2895,
|
|
"grad_norm": 0.7783549427986145,
|
|
"learning_rate": 0.00014224224224224224,
|
|
"loss": 1.9832,
|
|
"step": 2895
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"grad_norm": 0.7698352932929993,
|
|
"learning_rate": 0.00014214214214214215,
|
|
"loss": 1.967,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 0.2905,
|
|
"grad_norm": 0.8160349726676941,
|
|
"learning_rate": 0.00014204204204204207,
|
|
"loss": 1.9633,
|
|
"step": 2905
|
|
},
|
|
{
|
|
"epoch": 0.291,
|
|
"grad_norm": 0.8119205832481384,
|
|
"learning_rate": 0.00014194194194194195,
|
|
"loss": 1.9799,
|
|
"step": 2910
|
|
},
|
|
{
|
|
"epoch": 0.2915,
|
|
"grad_norm": 0.7536299824714661,
|
|
"learning_rate": 0.00014184184184184184,
|
|
"loss": 1.9761,
|
|
"step": 2915
|
|
},
|
|
{
|
|
"epoch": 0.292,
|
|
"grad_norm": 0.8390820622444153,
|
|
"learning_rate": 0.00014174174174174176,
|
|
"loss": 1.984,
|
|
"step": 2920
|
|
},
|
|
{
|
|
"epoch": 0.2925,
|
|
"grad_norm": 0.7603803277015686,
|
|
"learning_rate": 0.00014164164164164164,
|
|
"loss": 1.9673,
|
|
"step": 2925
|
|
},
|
|
{
|
|
"epoch": 0.293,
|
|
"grad_norm": 0.8152424693107605,
|
|
"learning_rate": 0.00014154154154154156,
|
|
"loss": 1.9637,
|
|
"step": 2930
|
|
},
|
|
{
|
|
"epoch": 0.2935,
|
|
"grad_norm": 0.8138707876205444,
|
|
"learning_rate": 0.00014144144144144144,
|
|
"loss": 1.9587,
|
|
"step": 2935
|
|
},
|
|
{
|
|
"epoch": 0.294,
|
|
"grad_norm": 0.7616044282913208,
|
|
"learning_rate": 0.00014134134134134133,
|
|
"loss": 1.9424,
|
|
"step": 2940
|
|
},
|
|
{
|
|
"epoch": 0.2945,
|
|
"grad_norm": 0.7655227184295654,
|
|
"learning_rate": 0.00014124124124124124,
|
|
"loss": 1.957,
|
|
"step": 2945
|
|
},
|
|
{
|
|
"epoch": 0.295,
|
|
"grad_norm": 0.8408180475234985,
|
|
"learning_rate": 0.00014114114114114113,
|
|
"loss": 1.9452,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 0.2955,
|
|
"grad_norm": 0.7936397194862366,
|
|
"learning_rate": 0.00014104104104104104,
|
|
"loss": 1.9595,
|
|
"step": 2955
|
|
},
|
|
{
|
|
"epoch": 0.296,
|
|
"grad_norm": 0.7944901585578918,
|
|
"learning_rate": 0.00014094094094094096,
|
|
"loss": 1.9666,
|
|
"step": 2960
|
|
},
|
|
{
|
|
"epoch": 0.2965,
|
|
"grad_norm": 0.818505585193634,
|
|
"learning_rate": 0.00014084084084084084,
|
|
"loss": 1.9738,
|
|
"step": 2965
|
|
},
|
|
{
|
|
"epoch": 0.297,
|
|
"grad_norm": 0.8084187507629395,
|
|
"learning_rate": 0.00014074074074074076,
|
|
"loss": 1.9401,
|
|
"step": 2970
|
|
},
|
|
{
|
|
"epoch": 0.2975,
|
|
"grad_norm": 0.791145920753479,
|
|
"learning_rate": 0.00014064064064064064,
|
|
"loss": 1.9562,
|
|
"step": 2975
|
|
},
|
|
{
|
|
"epoch": 0.298,
|
|
"grad_norm": 0.7604819536209106,
|
|
"learning_rate": 0.00014054054054054056,
|
|
"loss": 1.9602,
|
|
"step": 2980
|
|
},
|
|
{
|
|
"epoch": 0.2985,
|
|
"grad_norm": 0.7864161133766174,
|
|
"learning_rate": 0.00014044044044044044,
|
|
"loss": 1.9731,
|
|
"step": 2985
|
|
},
|
|
{
|
|
"epoch": 0.299,
|
|
"grad_norm": 0.8198469281196594,
|
|
"learning_rate": 0.00014034034034034033,
|
|
"loss": 1.9613,
|
|
"step": 2990
|
|
},
|
|
{
|
|
"epoch": 0.2995,
|
|
"grad_norm": 0.8647459745407104,
|
|
"learning_rate": 0.00014024024024024024,
|
|
"loss": 1.9397,
|
|
"step": 2995
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"grad_norm": 0.7643691897392273,
|
|
"learning_rate": 0.00014014014014014013,
|
|
"loss": 1.9646,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.3005,
|
|
"grad_norm": 0.7736343145370483,
|
|
"learning_rate": 0.00014004004004004004,
|
|
"loss": 1.9638,
|
|
"step": 3005
|
|
},
|
|
{
|
|
"epoch": 0.301,
|
|
"grad_norm": 0.7389901876449585,
|
|
"learning_rate": 0.00013993993993993996,
|
|
"loss": 1.9815,
|
|
"step": 3010
|
|
},
|
|
{
|
|
"epoch": 0.3015,
|
|
"grad_norm": 0.7692343592643738,
|
|
"learning_rate": 0.00013983983983983984,
|
|
"loss": 1.9787,
|
|
"step": 3015
|
|
},
|
|
{
|
|
"epoch": 0.302,
|
|
"grad_norm": 0.7531271576881409,
|
|
"learning_rate": 0.00013973973973973976,
|
|
"loss": 1.9597,
|
|
"step": 3020
|
|
},
|
|
{
|
|
"epoch": 0.3025,
|
|
"grad_norm": 0.7814120054244995,
|
|
"learning_rate": 0.00013963963963963964,
|
|
"loss": 1.9702,
|
|
"step": 3025
|
|
},
|
|
{
|
|
"epoch": 0.303,
|
|
"grad_norm": 0.757296621799469,
|
|
"learning_rate": 0.00013953953953953956,
|
|
"loss": 1.9745,
|
|
"step": 3030
|
|
},
|
|
{
|
|
"epoch": 0.3035,
|
|
"grad_norm": 0.8460572361946106,
|
|
"learning_rate": 0.00013943943943943944,
|
|
"loss": 1.9714,
|
|
"step": 3035
|
|
},
|
|
{
|
|
"epoch": 0.304,
|
|
"grad_norm": 0.8126088380813599,
|
|
"learning_rate": 0.00013933933933933933,
|
|
"loss": 1.973,
|
|
"step": 3040
|
|
},
|
|
{
|
|
"epoch": 0.3045,
|
|
"grad_norm": 0.7975168228149414,
|
|
"learning_rate": 0.00013923923923923925,
|
|
"loss": 1.9692,
|
|
"step": 3045
|
|
},
|
|
{
|
|
"epoch": 0.305,
|
|
"grad_norm": 0.7773084044456482,
|
|
"learning_rate": 0.00013913913913913913,
|
|
"loss": 1.9489,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 0.3055,
|
|
"grad_norm": 0.7985555529594421,
|
|
"learning_rate": 0.00013903903903903905,
|
|
"loss": 1.9835,
|
|
"step": 3055
|
|
},
|
|
{
|
|
"epoch": 0.306,
|
|
"grad_norm": 0.8398511409759521,
|
|
"learning_rate": 0.00013893893893893893,
|
|
"loss": 1.9602,
|
|
"step": 3060
|
|
},
|
|
{
|
|
"epoch": 0.3065,
|
|
"grad_norm": 0.9037275910377502,
|
|
"learning_rate": 0.00013883883883883885,
|
|
"loss": 1.9629,
|
|
"step": 3065
|
|
},
|
|
{
|
|
"epoch": 0.307,
|
|
"grad_norm": 0.8658416271209717,
|
|
"learning_rate": 0.00013873873873873876,
|
|
"loss": 1.9702,
|
|
"step": 3070
|
|
},
|
|
{
|
|
"epoch": 0.3075,
|
|
"grad_norm": 0.7509240508079529,
|
|
"learning_rate": 0.00013863863863863865,
|
|
"loss": 1.9703,
|
|
"step": 3075
|
|
},
|
|
{
|
|
"epoch": 0.308,
|
|
"grad_norm": 0.7586889863014221,
|
|
"learning_rate": 0.00013853853853853856,
|
|
"loss": 1.9537,
|
|
"step": 3080
|
|
},
|
|
{
|
|
"epoch": 0.3085,
|
|
"grad_norm": 0.7828867435455322,
|
|
"learning_rate": 0.00013843843843843845,
|
|
"loss": 1.9686,
|
|
"step": 3085
|
|
},
|
|
{
|
|
"epoch": 0.309,
|
|
"grad_norm": 0.908487856388092,
|
|
"learning_rate": 0.00013833833833833833,
|
|
"loss": 1.9703,
|
|
"step": 3090
|
|
},
|
|
{
|
|
"epoch": 0.3095,
|
|
"grad_norm": 0.774718165397644,
|
|
"learning_rate": 0.00013823823823823825,
|
|
"loss": 1.9656,
|
|
"step": 3095
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"grad_norm": 0.7825801372528076,
|
|
"learning_rate": 0.00013813813813813813,
|
|
"loss": 1.9571,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 0.3105,
|
|
"grad_norm": 0.7940477132797241,
|
|
"learning_rate": 0.00013803803803803805,
|
|
"loss": 1.973,
|
|
"step": 3105
|
|
},
|
|
{
|
|
"epoch": 0.311,
|
|
"grad_norm": 0.7585093975067139,
|
|
"learning_rate": 0.00013793793793793793,
|
|
"loss": 1.9605,
|
|
"step": 3110
|
|
},
|
|
{
|
|
"epoch": 0.3115,
|
|
"grad_norm": 0.7673121690750122,
|
|
"learning_rate": 0.00013783783783783785,
|
|
"loss": 1.938,
|
|
"step": 3115
|
|
},
|
|
{
|
|
"epoch": 0.312,
|
|
"grad_norm": 0.8095937371253967,
|
|
"learning_rate": 0.00013773773773773776,
|
|
"loss": 1.9643,
|
|
"step": 3120
|
|
},
|
|
{
|
|
"epoch": 0.3125,
|
|
"grad_norm": 0.8239126801490784,
|
|
"learning_rate": 0.00013763763763763765,
|
|
"loss": 1.9655,
|
|
"step": 3125
|
|
},
|
|
{
|
|
"epoch": 0.313,
|
|
"grad_norm": 0.7611415386199951,
|
|
"learning_rate": 0.00013753753753753756,
|
|
"loss": 1.966,
|
|
"step": 3130
|
|
},
|
|
{
|
|
"epoch": 0.3135,
|
|
"grad_norm": 0.8436517119407654,
|
|
"learning_rate": 0.00013743743743743745,
|
|
"loss": 1.9885,
|
|
"step": 3135
|
|
},
|
|
{
|
|
"epoch": 0.314,
|
|
"grad_norm": 0.7741273045539856,
|
|
"learning_rate": 0.00013733733733733733,
|
|
"loss": 1.9593,
|
|
"step": 3140
|
|
},
|
|
{
|
|
"epoch": 0.3145,
|
|
"grad_norm": 0.8990844488143921,
|
|
"learning_rate": 0.00013723723723723725,
|
|
"loss": 1.9475,
|
|
"step": 3145
|
|
},
|
|
{
|
|
"epoch": 0.315,
|
|
"grad_norm": 0.8114355206489563,
|
|
"learning_rate": 0.00013713713713713713,
|
|
"loss": 1.9417,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 0.3155,
|
|
"grad_norm": 0.9088999032974243,
|
|
"learning_rate": 0.00013703703703703705,
|
|
"loss": 1.9529,
|
|
"step": 3155
|
|
},
|
|
{
|
|
"epoch": 0.316,
|
|
"grad_norm": 0.7836775183677673,
|
|
"learning_rate": 0.00013693693693693693,
|
|
"loss": 1.9442,
|
|
"step": 3160
|
|
},
|
|
{
|
|
"epoch": 0.3165,
|
|
"grad_norm": 0.7776234149932861,
|
|
"learning_rate": 0.00013683683683683682,
|
|
"loss": 1.9618,
|
|
"step": 3165
|
|
},
|
|
{
|
|
"epoch": 0.317,
|
|
"grad_norm": 0.8789812326431274,
|
|
"learning_rate": 0.00013673673673673673,
|
|
"loss": 1.9714,
|
|
"step": 3170
|
|
},
|
|
{
|
|
"epoch": 0.3175,
|
|
"grad_norm": 0.89927738904953,
|
|
"learning_rate": 0.00013663663663663665,
|
|
"loss": 1.9523,
|
|
"step": 3175
|
|
},
|
|
{
|
|
"epoch": 0.318,
|
|
"grad_norm": 0.8082011342048645,
|
|
"learning_rate": 0.00013653653653653656,
|
|
"loss": 1.9535,
|
|
"step": 3180
|
|
},
|
|
{
|
|
"epoch": 0.3185,
|
|
"grad_norm": 0.7667655348777771,
|
|
"learning_rate": 0.00013643643643643645,
|
|
"loss": 1.9307,
|
|
"step": 3185
|
|
},
|
|
{
|
|
"epoch": 0.319,
|
|
"grad_norm": 0.7933686971664429,
|
|
"learning_rate": 0.00013633633633633634,
|
|
"loss": 1.9538,
|
|
"step": 3190
|
|
},
|
|
{
|
|
"epoch": 0.3195,
|
|
"grad_norm": 0.800533652305603,
|
|
"learning_rate": 0.00013623623623623625,
|
|
"loss": 1.9661,
|
|
"step": 3195
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.7852665781974792,
|
|
"learning_rate": 0.00013613613613613614,
|
|
"loss": 1.9528,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 0.3205,
|
|
"grad_norm": 0.8265931010246277,
|
|
"learning_rate": 0.00013603603603603605,
|
|
"loss": 1.9623,
|
|
"step": 3205
|
|
},
|
|
{
|
|
"epoch": 0.321,
|
|
"grad_norm": 0.8010437488555908,
|
|
"learning_rate": 0.00013593593593593594,
|
|
"loss": 1.9623,
|
|
"step": 3210
|
|
},
|
|
{
|
|
"epoch": 0.3215,
|
|
"grad_norm": 0.7773730158805847,
|
|
"learning_rate": 0.00013583583583583582,
|
|
"loss": 1.9646,
|
|
"step": 3215
|
|
},
|
|
{
|
|
"epoch": 0.322,
|
|
"grad_norm": 0.8105774521827698,
|
|
"learning_rate": 0.00013573573573573574,
|
|
"loss": 1.9534,
|
|
"step": 3220
|
|
},
|
|
{
|
|
"epoch": 0.3225,
|
|
"grad_norm": 0.7743560075759888,
|
|
"learning_rate": 0.00013563563563563565,
|
|
"loss": 1.9436,
|
|
"step": 3225
|
|
},
|
|
{
|
|
"epoch": 0.323,
|
|
"grad_norm": 0.8030966520309448,
|
|
"learning_rate": 0.00013553553553553556,
|
|
"loss": 1.9516,
|
|
"step": 3230
|
|
},
|
|
{
|
|
"epoch": 0.3235,
|
|
"grad_norm": 0.8765146732330322,
|
|
"learning_rate": 0.00013543543543543545,
|
|
"loss": 1.9715,
|
|
"step": 3235
|
|
},
|
|
{
|
|
"epoch": 0.324,
|
|
"grad_norm": 0.7942567467689514,
|
|
"learning_rate": 0.00013533533533533534,
|
|
"loss": 1.9483,
|
|
"step": 3240
|
|
},
|
|
{
|
|
"epoch": 0.3245,
|
|
"grad_norm": 0.8173383474349976,
|
|
"learning_rate": 0.00013523523523523525,
|
|
"loss": 1.9529,
|
|
"step": 3245
|
|
},
|
|
{
|
|
"epoch": 0.325,
|
|
"grad_norm": 0.7865030169487,
|
|
"learning_rate": 0.00013513513513513514,
|
|
"loss": 1.9656,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 0.3255,
|
|
"grad_norm": 0.8991813659667969,
|
|
"learning_rate": 0.00013503503503503505,
|
|
"loss": 1.9551,
|
|
"step": 3255
|
|
},
|
|
{
|
|
"epoch": 0.326,
|
|
"grad_norm": 0.7669337391853333,
|
|
"learning_rate": 0.00013493493493493494,
|
|
"loss": 1.9618,
|
|
"step": 3260
|
|
},
|
|
{
|
|
"epoch": 0.3265,
|
|
"grad_norm": 0.8322509527206421,
|
|
"learning_rate": 0.00013483483483483482,
|
|
"loss": 1.9467,
|
|
"step": 3265
|
|
},
|
|
{
|
|
"epoch": 0.327,
|
|
"grad_norm": 0.7678415775299072,
|
|
"learning_rate": 0.00013473473473473474,
|
|
"loss": 1.9576,
|
|
"step": 3270
|
|
},
|
|
{
|
|
"epoch": 0.3275,
|
|
"grad_norm": 0.8185941576957703,
|
|
"learning_rate": 0.00013463463463463462,
|
|
"loss": 1.9719,
|
|
"step": 3275
|
|
},
|
|
{
|
|
"epoch": 0.328,
|
|
"grad_norm": 0.8121998906135559,
|
|
"learning_rate": 0.00013453453453453454,
|
|
"loss": 1.9538,
|
|
"step": 3280
|
|
},
|
|
{
|
|
"epoch": 0.3285,
|
|
"grad_norm": 0.8269667625427246,
|
|
"learning_rate": 0.00013443443443443445,
|
|
"loss": 1.9472,
|
|
"step": 3285
|
|
},
|
|
{
|
|
"epoch": 0.329,
|
|
"grad_norm": 0.7667022347450256,
|
|
"learning_rate": 0.00013433433433433434,
|
|
"loss": 1.976,
|
|
"step": 3290
|
|
},
|
|
{
|
|
"epoch": 0.3295,
|
|
"grad_norm": 0.794097363948822,
|
|
"learning_rate": 0.00013423423423423425,
|
|
"loss": 1.9703,
|
|
"step": 3295
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"grad_norm": 0.7591664791107178,
|
|
"learning_rate": 0.00013413413413413414,
|
|
"loss": 1.9523,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 0.3305,
|
|
"grad_norm": 0.7778351902961731,
|
|
"learning_rate": 0.00013403403403403405,
|
|
"loss": 1.9448,
|
|
"step": 3305
|
|
},
|
|
{
|
|
"epoch": 0.331,
|
|
"grad_norm": 0.7940652370452881,
|
|
"learning_rate": 0.00013393393393393394,
|
|
"loss": 1.9482,
|
|
"step": 3310
|
|
},
|
|
{
|
|
"epoch": 0.3315,
|
|
"grad_norm": 0.7958424091339111,
|
|
"learning_rate": 0.00013383383383383383,
|
|
"loss": 1.9496,
|
|
"step": 3315
|
|
},
|
|
{
|
|
"epoch": 0.332,
|
|
"grad_norm": 0.8664911389350891,
|
|
"learning_rate": 0.00013373373373373374,
|
|
"loss": 1.9537,
|
|
"step": 3320
|
|
},
|
|
{
|
|
"epoch": 0.3325,
|
|
"grad_norm": 0.8059619665145874,
|
|
"learning_rate": 0.00013363363363363363,
|
|
"loss": 1.9493,
|
|
"step": 3325
|
|
},
|
|
{
|
|
"epoch": 0.333,
|
|
"grad_norm": 0.7810567021369934,
|
|
"learning_rate": 0.00013353353353353354,
|
|
"loss": 1.9694,
|
|
"step": 3330
|
|
},
|
|
{
|
|
"epoch": 0.3335,
|
|
"grad_norm": 0.7973411679267883,
|
|
"learning_rate": 0.00013343343343343345,
|
|
"loss": 1.9506,
|
|
"step": 3335
|
|
},
|
|
{
|
|
"epoch": 0.334,
|
|
"grad_norm": 0.9919584393501282,
|
|
"learning_rate": 0.00013333333333333334,
|
|
"loss": 1.9544,
|
|
"step": 3340
|
|
},
|
|
{
|
|
"epoch": 0.3345,
|
|
"grad_norm": 0.7684592008590698,
|
|
"learning_rate": 0.00013323323323323325,
|
|
"loss": 1.9666,
|
|
"step": 3345
|
|
},
|
|
{
|
|
"epoch": 0.335,
|
|
"grad_norm": 0.7740269899368286,
|
|
"learning_rate": 0.00013313313313313314,
|
|
"loss": 1.9458,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 0.3355,
|
|
"grad_norm": 0.7702473402023315,
|
|
"learning_rate": 0.00013303303303303305,
|
|
"loss": 1.9768,
|
|
"step": 3355
|
|
},
|
|
{
|
|
"epoch": 0.336,
|
|
"grad_norm": 0.784784734249115,
|
|
"learning_rate": 0.00013293293293293294,
|
|
"loss": 1.9512,
|
|
"step": 3360
|
|
},
|
|
{
|
|
"epoch": 0.3365,
|
|
"grad_norm": 0.9200993180274963,
|
|
"learning_rate": 0.00013283283283283283,
|
|
"loss": 1.9477,
|
|
"step": 3365
|
|
},
|
|
{
|
|
"epoch": 0.337,
|
|
"grad_norm": 1.021840214729309,
|
|
"learning_rate": 0.00013273273273273274,
|
|
"loss": 1.9673,
|
|
"step": 3370
|
|
},
|
|
{
|
|
"epoch": 0.3375,
|
|
"grad_norm": 0.920347273349762,
|
|
"learning_rate": 0.00013263263263263263,
|
|
"loss": 1.9381,
|
|
"step": 3375
|
|
},
|
|
{
|
|
"epoch": 0.338,
|
|
"grad_norm": 0.7917675375938416,
|
|
"learning_rate": 0.00013253253253253254,
|
|
"loss": 1.943,
|
|
"step": 3380
|
|
},
|
|
{
|
|
"epoch": 0.3385,
|
|
"grad_norm": 0.7981655597686768,
|
|
"learning_rate": 0.00013243243243243243,
|
|
"loss": 1.9565,
|
|
"step": 3385
|
|
},
|
|
{
|
|
"epoch": 0.339,
|
|
"grad_norm": 0.7743932604789734,
|
|
"learning_rate": 0.00013233233233233234,
|
|
"loss": 1.9521,
|
|
"step": 3390
|
|
},
|
|
{
|
|
"epoch": 0.3395,
|
|
"grad_norm": 0.784078061580658,
|
|
"learning_rate": 0.00013223223223223225,
|
|
"loss": 1.9404,
|
|
"step": 3395
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"grad_norm": 0.7654576301574707,
|
|
"learning_rate": 0.00013213213213213214,
|
|
"loss": 1.9577,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 0.3405,
|
|
"grad_norm": 0.7969295382499695,
|
|
"learning_rate": 0.00013203203203203206,
|
|
"loss": 1.9403,
|
|
"step": 3405
|
|
},
|
|
{
|
|
"epoch": 0.341,
|
|
"grad_norm": 1.016737461090088,
|
|
"learning_rate": 0.00013193193193193194,
|
|
"loss": 1.9584,
|
|
"step": 3410
|
|
},
|
|
{
|
|
"epoch": 0.3415,
|
|
"grad_norm": 1.0509916543960571,
|
|
"learning_rate": 0.00013183183183183183,
|
|
"loss": 1.939,
|
|
"step": 3415
|
|
},
|
|
{
|
|
"epoch": 0.342,
|
|
"grad_norm": 0.9138502478599548,
|
|
"learning_rate": 0.00013173173173173174,
|
|
"loss": 1.9478,
|
|
"step": 3420
|
|
},
|
|
{
|
|
"epoch": 0.3425,
|
|
"grad_norm": 0.817069947719574,
|
|
"learning_rate": 0.00013163163163163163,
|
|
"loss": 1.9713,
|
|
"step": 3425
|
|
},
|
|
{
|
|
"epoch": 0.343,
|
|
"grad_norm": 0.9959232211112976,
|
|
"learning_rate": 0.00013153153153153154,
|
|
"loss": 1.9595,
|
|
"step": 3430
|
|
},
|
|
{
|
|
"epoch": 0.3435,
|
|
"grad_norm": 1.0368640422821045,
|
|
"learning_rate": 0.00013143143143143143,
|
|
"loss": 1.959,
|
|
"step": 3435
|
|
},
|
|
{
|
|
"epoch": 0.344,
|
|
"grad_norm": 0.7973247170448303,
|
|
"learning_rate": 0.00013133133133133134,
|
|
"loss": 1.9755,
|
|
"step": 3440
|
|
},
|
|
{
|
|
"epoch": 0.3445,
|
|
"grad_norm": 0.8196215033531189,
|
|
"learning_rate": 0.00013123123123123126,
|
|
"loss": 1.9437,
|
|
"step": 3445
|
|
},
|
|
{
|
|
"epoch": 0.345,
|
|
"grad_norm": 0.9840690493583679,
|
|
"learning_rate": 0.00013113113113113114,
|
|
"loss": 1.9537,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 0.3455,
|
|
"grad_norm": 0.9263918399810791,
|
|
"learning_rate": 0.00013103103103103103,
|
|
"loss": 1.9339,
|
|
"step": 3455
|
|
},
|
|
{
|
|
"epoch": 0.346,
|
|
"grad_norm": 0.7676291465759277,
|
|
"learning_rate": 0.00013093093093093094,
|
|
"loss": 1.9359,
|
|
"step": 3460
|
|
},
|
|
{
|
|
"epoch": 0.3465,
|
|
"grad_norm": 0.7748109698295593,
|
|
"learning_rate": 0.00013083083083083083,
|
|
"loss": 1.9464,
|
|
"step": 3465
|
|
},
|
|
{
|
|
"epoch": 0.347,
|
|
"grad_norm": 0.852546215057373,
|
|
"learning_rate": 0.00013073073073073074,
|
|
"loss": 1.934,
|
|
"step": 3470
|
|
},
|
|
{
|
|
"epoch": 0.3475,
|
|
"grad_norm": 0.9905356764793396,
|
|
"learning_rate": 0.00013063063063063063,
|
|
"loss": 1.9459,
|
|
"step": 3475
|
|
},
|
|
{
|
|
"epoch": 0.348,
|
|
"grad_norm": 0.8646652698516846,
|
|
"learning_rate": 0.00013053053053053052,
|
|
"loss": 1.9385,
|
|
"step": 3480
|
|
},
|
|
{
|
|
"epoch": 0.3485,
|
|
"grad_norm": 0.7634602189064026,
|
|
"learning_rate": 0.00013043043043043043,
|
|
"loss": 1.9474,
|
|
"step": 3485
|
|
},
|
|
{
|
|
"epoch": 0.349,
|
|
"grad_norm": 0.7793004512786865,
|
|
"learning_rate": 0.00013033033033033032,
|
|
"loss": 1.9551,
|
|
"step": 3490
|
|
},
|
|
{
|
|
"epoch": 0.3495,
|
|
"grad_norm": 0.7419679760932922,
|
|
"learning_rate": 0.00013023023023023023,
|
|
"loss": 1.9424,
|
|
"step": 3495
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"grad_norm": 0.7776840329170227,
|
|
"learning_rate": 0.00013013013013013014,
|
|
"loss": 1.9414,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 0.3505,
|
|
"grad_norm": 0.8862317800521851,
|
|
"learning_rate": 0.00013003003003003003,
|
|
"loss": 1.9629,
|
|
"step": 3505
|
|
},
|
|
{
|
|
"epoch": 0.351,
|
|
"grad_norm": 0.9258846044540405,
|
|
"learning_rate": 0.00012992992992992994,
|
|
"loss": 1.9242,
|
|
"step": 3510
|
|
},
|
|
{
|
|
"epoch": 0.3515,
|
|
"grad_norm": 0.7799050211906433,
|
|
"learning_rate": 0.00012982982982982983,
|
|
"loss": 1.963,
|
|
"step": 3515
|
|
},
|
|
{
|
|
"epoch": 0.352,
|
|
"grad_norm": 0.8912829160690308,
|
|
"learning_rate": 0.00012972972972972974,
|
|
"loss": 1.9574,
|
|
"step": 3520
|
|
},
|
|
{
|
|
"epoch": 0.3525,
|
|
"grad_norm": 0.905232846736908,
|
|
"learning_rate": 0.00012962962962962963,
|
|
"loss": 1.9294,
|
|
"step": 3525
|
|
},
|
|
{
|
|
"epoch": 0.353,
|
|
"grad_norm": 0.7626670002937317,
|
|
"learning_rate": 0.00012952952952952952,
|
|
"loss": 1.9524,
|
|
"step": 3530
|
|
},
|
|
{
|
|
"epoch": 0.3535,
|
|
"grad_norm": 0.7529788613319397,
|
|
"learning_rate": 0.00012942942942942943,
|
|
"loss": 1.9665,
|
|
"step": 3535
|
|
},
|
|
{
|
|
"epoch": 0.354,
|
|
"grad_norm": 0.7423688173294067,
|
|
"learning_rate": 0.00012932932932932932,
|
|
"loss": 1.9356,
|
|
"step": 3540
|
|
},
|
|
{
|
|
"epoch": 0.3545,
|
|
"grad_norm": 0.7545106410980225,
|
|
"learning_rate": 0.00012922922922922923,
|
|
"loss": 1.9385,
|
|
"step": 3545
|
|
},
|
|
{
|
|
"epoch": 0.355,
|
|
"grad_norm": 0.7768842577934265,
|
|
"learning_rate": 0.00012912912912912915,
|
|
"loss": 1.9444,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 0.3555,
|
|
"grad_norm": 0.7664159536361694,
|
|
"learning_rate": 0.00012902902902902903,
|
|
"loss": 1.9363,
|
|
"step": 3555
|
|
},
|
|
{
|
|
"epoch": 0.356,
|
|
"grad_norm": 0.796855628490448,
|
|
"learning_rate": 0.00012892892892892895,
|
|
"loss": 1.9445,
|
|
"step": 3560
|
|
},
|
|
{
|
|
"epoch": 0.3565,
|
|
"grad_norm": 0.7599654197692871,
|
|
"learning_rate": 0.00012882882882882883,
|
|
"loss": 1.9486,
|
|
"step": 3565
|
|
},
|
|
{
|
|
"epoch": 0.357,
|
|
"grad_norm": 0.8220765590667725,
|
|
"learning_rate": 0.00012872872872872875,
|
|
"loss": 1.9604,
|
|
"step": 3570
|
|
},
|
|
{
|
|
"epoch": 0.3575,
|
|
"grad_norm": 0.8476091027259827,
|
|
"learning_rate": 0.00012862862862862863,
|
|
"loss": 1.9389,
|
|
"step": 3575
|
|
},
|
|
{
|
|
"epoch": 0.358,
|
|
"grad_norm": 0.8238967061042786,
|
|
"learning_rate": 0.00012852852852852852,
|
|
"loss": 1.9527,
|
|
"step": 3580
|
|
},
|
|
{
|
|
"epoch": 0.3585,
|
|
"grad_norm": 0.8090390563011169,
|
|
"learning_rate": 0.00012842842842842843,
|
|
"loss": 1.9377,
|
|
"step": 3585
|
|
},
|
|
{
|
|
"epoch": 0.359,
|
|
"grad_norm": 0.7613897919654846,
|
|
"learning_rate": 0.00012832832832832832,
|
|
"loss": 1.9213,
|
|
"step": 3590
|
|
},
|
|
{
|
|
"epoch": 0.3595,
|
|
"grad_norm": 0.8270089626312256,
|
|
"learning_rate": 0.00012822822822822823,
|
|
"loss": 1.9519,
|
|
"step": 3595
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"grad_norm": 0.7741177082061768,
|
|
"learning_rate": 0.00012812812812812812,
|
|
"loss": 1.9569,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 0.3605,
|
|
"grad_norm": 0.8119099736213684,
|
|
"learning_rate": 0.00012802802802802803,
|
|
"loss": 1.9373,
|
|
"step": 3605
|
|
},
|
|
{
|
|
"epoch": 0.361,
|
|
"grad_norm": 0.778448224067688,
|
|
"learning_rate": 0.00012792792792792795,
|
|
"loss": 1.9433,
|
|
"step": 3610
|
|
},
|
|
{
|
|
"epoch": 0.3615,
|
|
"grad_norm": 0.7723468542098999,
|
|
"learning_rate": 0.00012782782782782783,
|
|
"loss": 1.9489,
|
|
"step": 3615
|
|
},
|
|
{
|
|
"epoch": 0.362,
|
|
"grad_norm": 0.8068646192550659,
|
|
"learning_rate": 0.00012772772772772775,
|
|
"loss": 1.9508,
|
|
"step": 3620
|
|
},
|
|
{
|
|
"epoch": 0.3625,
|
|
"grad_norm": 0.7997691631317139,
|
|
"learning_rate": 0.00012762762762762763,
|
|
"loss": 1.9296,
|
|
"step": 3625
|
|
},
|
|
{
|
|
"epoch": 0.363,
|
|
"grad_norm": 0.7968930006027222,
|
|
"learning_rate": 0.00012752752752752752,
|
|
"loss": 1.9388,
|
|
"step": 3630
|
|
},
|
|
{
|
|
"epoch": 0.3635,
|
|
"grad_norm": 0.7867845296859741,
|
|
"learning_rate": 0.00012742742742742743,
|
|
"loss": 1.9337,
|
|
"step": 3635
|
|
},
|
|
{
|
|
"epoch": 0.364,
|
|
"grad_norm": 0.7725551724433899,
|
|
"learning_rate": 0.00012732732732732732,
|
|
"loss": 1.9355,
|
|
"step": 3640
|
|
},
|
|
{
|
|
"epoch": 0.3645,
|
|
"grad_norm": 0.8146640658378601,
|
|
"learning_rate": 0.00012722722722722723,
|
|
"loss": 1.9349,
|
|
"step": 3645
|
|
},
|
|
{
|
|
"epoch": 0.365,
|
|
"grad_norm": 0.7650962471961975,
|
|
"learning_rate": 0.00012712712712712712,
|
|
"loss": 1.9581,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 0.3655,
|
|
"grad_norm": 0.7965732216835022,
|
|
"learning_rate": 0.00012702702702702703,
|
|
"loss": 1.951,
|
|
"step": 3655
|
|
},
|
|
{
|
|
"epoch": 0.366,
|
|
"grad_norm": 0.7692468166351318,
|
|
"learning_rate": 0.00012692692692692695,
|
|
"loss": 1.9464,
|
|
"step": 3660
|
|
},
|
|
{
|
|
"epoch": 0.3665,
|
|
"grad_norm": 0.7383764982223511,
|
|
"learning_rate": 0.00012682682682682684,
|
|
"loss": 1.9432,
|
|
"step": 3665
|
|
},
|
|
{
|
|
"epoch": 0.367,
|
|
"grad_norm": 0.7694110870361328,
|
|
"learning_rate": 0.00012672672672672675,
|
|
"loss": 1.9442,
|
|
"step": 3670
|
|
},
|
|
{
|
|
"epoch": 0.3675,
|
|
"grad_norm": 0.7646791338920593,
|
|
"learning_rate": 0.00012662662662662664,
|
|
"loss": 1.9124,
|
|
"step": 3675
|
|
},
|
|
{
|
|
"epoch": 0.368,
|
|
"grad_norm": 0.7893620133399963,
|
|
"learning_rate": 0.00012652652652652652,
|
|
"loss": 1.9336,
|
|
"step": 3680
|
|
},
|
|
{
|
|
"epoch": 0.3685,
|
|
"grad_norm": 0.8947069644927979,
|
|
"learning_rate": 0.00012642642642642644,
|
|
"loss": 1.9418,
|
|
"step": 3685
|
|
},
|
|
{
|
|
"epoch": 0.369,
|
|
"grad_norm": 0.8227481842041016,
|
|
"learning_rate": 0.00012632632632632632,
|
|
"loss": 1.9259,
|
|
"step": 3690
|
|
},
|
|
{
|
|
"epoch": 0.3695,
|
|
"grad_norm": 0.8941676020622253,
|
|
"learning_rate": 0.00012622622622622624,
|
|
"loss": 1.9474,
|
|
"step": 3695
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"grad_norm": 0.8401495218276978,
|
|
"learning_rate": 0.00012612612612612612,
|
|
"loss": 1.9454,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 0.3705,
|
|
"grad_norm": 0.8106552362442017,
|
|
"learning_rate": 0.00012602602602602604,
|
|
"loss": 1.9469,
|
|
"step": 3705
|
|
},
|
|
{
|
|
"epoch": 0.371,
|
|
"grad_norm": 0.7758846879005432,
|
|
"learning_rate": 0.00012592592592592592,
|
|
"loss": 1.9405,
|
|
"step": 3710
|
|
},
|
|
{
|
|
"epoch": 0.3715,
|
|
"grad_norm": 0.8738759756088257,
|
|
"learning_rate": 0.00012582582582582584,
|
|
"loss": 1.9212,
|
|
"step": 3715
|
|
},
|
|
{
|
|
"epoch": 0.372,
|
|
"grad_norm": 0.8900704979896545,
|
|
"learning_rate": 0.00012572572572572575,
|
|
"loss": 1.9482,
|
|
"step": 3720
|
|
},
|
|
{
|
|
"epoch": 0.3725,
|
|
"grad_norm": 0.8583644032478333,
|
|
"learning_rate": 0.00012562562562562564,
|
|
"loss": 1.9352,
|
|
"step": 3725
|
|
},
|
|
{
|
|
"epoch": 0.373,
|
|
"grad_norm": 0.7823512554168701,
|
|
"learning_rate": 0.00012552552552552552,
|
|
"loss": 1.9295,
|
|
"step": 3730
|
|
},
|
|
{
|
|
"epoch": 0.3735,
|
|
"grad_norm": 0.7929084897041321,
|
|
"learning_rate": 0.00012542542542542544,
|
|
"loss": 1.9361,
|
|
"step": 3735
|
|
},
|
|
{
|
|
"epoch": 0.374,
|
|
"grad_norm": 0.7727826833724976,
|
|
"learning_rate": 0.00012532532532532532,
|
|
"loss": 1.943,
|
|
"step": 3740
|
|
},
|
|
{
|
|
"epoch": 0.3745,
|
|
"grad_norm": 0.7919679284095764,
|
|
"learning_rate": 0.00012522522522522524,
|
|
"loss": 1.943,
|
|
"step": 3745
|
|
},
|
|
{
|
|
"epoch": 0.375,
|
|
"grad_norm": 0.7749269604682922,
|
|
"learning_rate": 0.00012512512512512512,
|
|
"loss": 1.9338,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 0.3755,
|
|
"grad_norm": 0.7870975732803345,
|
|
"learning_rate": 0.000125025025025025,
|
|
"loss": 1.9277,
|
|
"step": 3755
|
|
},
|
|
{
|
|
"epoch": 0.376,
|
|
"grad_norm": 0.7871671319007874,
|
|
"learning_rate": 0.00012492492492492492,
|
|
"loss": 1.9552,
|
|
"step": 3760
|
|
},
|
|
{
|
|
"epoch": 0.3765,
|
|
"grad_norm": 0.7836071848869324,
|
|
"learning_rate": 0.00012482482482482484,
|
|
"loss": 1.9263,
|
|
"step": 3765
|
|
},
|
|
{
|
|
"epoch": 0.377,
|
|
"grad_norm": 0.7796449065208435,
|
|
"learning_rate": 0.00012472472472472475,
|
|
"loss": 1.9495,
|
|
"step": 3770
|
|
},
|
|
{
|
|
"epoch": 0.3775,
|
|
"grad_norm": 0.7606745362281799,
|
|
"learning_rate": 0.00012462462462462464,
|
|
"loss": 1.9317,
|
|
"step": 3775
|
|
},
|
|
{
|
|
"epoch": 0.378,
|
|
"grad_norm": 0.7624865770339966,
|
|
"learning_rate": 0.00012452452452452452,
|
|
"loss": 1.951,
|
|
"step": 3780
|
|
},
|
|
{
|
|
"epoch": 0.3785,
|
|
"grad_norm": 0.7752624154090881,
|
|
"learning_rate": 0.00012442442442442444,
|
|
"loss": 1.9448,
|
|
"step": 3785
|
|
},
|
|
{
|
|
"epoch": 0.379,
|
|
"grad_norm": 0.8232024312019348,
|
|
"learning_rate": 0.00012432432432432433,
|
|
"loss": 1.9309,
|
|
"step": 3790
|
|
},
|
|
{
|
|
"epoch": 0.3795,
|
|
"grad_norm": 0.7860484719276428,
|
|
"learning_rate": 0.00012422422422422424,
|
|
"loss": 1.9443,
|
|
"step": 3795
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"grad_norm": 0.769604504108429,
|
|
"learning_rate": 0.00012412412412412413,
|
|
"loss": 1.9433,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 0.3805,
|
|
"grad_norm": 0.8487102389335632,
|
|
"learning_rate": 0.000124024024024024,
|
|
"loss": 1.9339,
|
|
"step": 3805
|
|
},
|
|
{
|
|
"epoch": 0.381,
|
|
"grad_norm": 0.8145004510879517,
|
|
"learning_rate": 0.00012392392392392393,
|
|
"loss": 1.943,
|
|
"step": 3810
|
|
},
|
|
{
|
|
"epoch": 0.3815,
|
|
"grad_norm": 0.8356045484542847,
|
|
"learning_rate": 0.0001238238238238238,
|
|
"loss": 1.932,
|
|
"step": 3815
|
|
},
|
|
{
|
|
"epoch": 0.382,
|
|
"grad_norm": 0.7990782856941223,
|
|
"learning_rate": 0.00012372372372372373,
|
|
"loss": 1.9458,
|
|
"step": 3820
|
|
},
|
|
{
|
|
"epoch": 0.3825,
|
|
"grad_norm": 0.7622951865196228,
|
|
"learning_rate": 0.00012362362362362364,
|
|
"loss": 1.9538,
|
|
"step": 3825
|
|
},
|
|
{
|
|
"epoch": 0.383,
|
|
"grad_norm": 0.796141505241394,
|
|
"learning_rate": 0.00012352352352352353,
|
|
"loss": 1.9274,
|
|
"step": 3830
|
|
},
|
|
{
|
|
"epoch": 0.3835,
|
|
"grad_norm": 0.9116307497024536,
|
|
"learning_rate": 0.00012342342342342344,
|
|
"loss": 1.9415,
|
|
"step": 3835
|
|
},
|
|
{
|
|
"epoch": 0.384,
|
|
"grad_norm": 0.7524915337562561,
|
|
"learning_rate": 0.00012332332332332333,
|
|
"loss": 1.9182,
|
|
"step": 3840
|
|
},
|
|
{
|
|
"epoch": 0.3845,
|
|
"grad_norm": 0.7611434459686279,
|
|
"learning_rate": 0.00012322322322322324,
|
|
"loss": 1.9306,
|
|
"step": 3845
|
|
},
|
|
{
|
|
"epoch": 0.385,
|
|
"grad_norm": 0.7478082180023193,
|
|
"learning_rate": 0.00012312312312312313,
|
|
"loss": 1.9239,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 0.3855,
|
|
"grad_norm": 0.7586308121681213,
|
|
"learning_rate": 0.000123023023023023,
|
|
"loss": 1.9332,
|
|
"step": 3855
|
|
},
|
|
{
|
|
"epoch": 0.386,
|
|
"grad_norm": 0.8614716529846191,
|
|
"learning_rate": 0.00012292292292292293,
|
|
"loss": 1.9583,
|
|
"step": 3860
|
|
},
|
|
{
|
|
"epoch": 0.3865,
|
|
"grad_norm": 0.7706379294395447,
|
|
"learning_rate": 0.00012282282282282281,
|
|
"loss": 1.9448,
|
|
"step": 3865
|
|
},
|
|
{
|
|
"epoch": 0.387,
|
|
"grad_norm": 0.802075982093811,
|
|
"learning_rate": 0.00012272272272272273,
|
|
"loss": 1.933,
|
|
"step": 3870
|
|
},
|
|
{
|
|
"epoch": 0.3875,
|
|
"grad_norm": 0.8200154900550842,
|
|
"learning_rate": 0.00012262262262262264,
|
|
"loss": 1.9213,
|
|
"step": 3875
|
|
},
|
|
{
|
|
"epoch": 0.388,
|
|
"grad_norm": 0.8005445599555969,
|
|
"learning_rate": 0.00012252252252252253,
|
|
"loss": 1.9423,
|
|
"step": 3880
|
|
},
|
|
{
|
|
"epoch": 0.3885,
|
|
"grad_norm": 0.9639005661010742,
|
|
"learning_rate": 0.00012242242242242244,
|
|
"loss": 1.928,
|
|
"step": 3885
|
|
},
|
|
{
|
|
"epoch": 0.389,
|
|
"grad_norm": 0.805856466293335,
|
|
"learning_rate": 0.00012232232232232233,
|
|
"loss": 1.9459,
|
|
"step": 3890
|
|
},
|
|
{
|
|
"epoch": 0.3895,
|
|
"grad_norm": 0.735449492931366,
|
|
"learning_rate": 0.00012222222222222224,
|
|
"loss": 1.933,
|
|
"step": 3895
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"grad_norm": 0.7601875066757202,
|
|
"learning_rate": 0.00012212212212212213,
|
|
"loss": 1.925,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 0.3905,
|
|
"grad_norm": 0.7551021575927734,
|
|
"learning_rate": 0.00012202202202202201,
|
|
"loss": 1.9213,
|
|
"step": 3905
|
|
},
|
|
{
|
|
"epoch": 0.391,
|
|
"grad_norm": 0.7619316577911377,
|
|
"learning_rate": 0.00012192192192192193,
|
|
"loss": 1.9242,
|
|
"step": 3910
|
|
},
|
|
{
|
|
"epoch": 0.3915,
|
|
"grad_norm": 0.8216850161552429,
|
|
"learning_rate": 0.00012182182182182183,
|
|
"loss": 1.927,
|
|
"step": 3915
|
|
},
|
|
{
|
|
"epoch": 0.392,
|
|
"grad_norm": 0.9015415906906128,
|
|
"learning_rate": 0.00012172172172172174,
|
|
"loss": 1.9438,
|
|
"step": 3920
|
|
},
|
|
{
|
|
"epoch": 0.3925,
|
|
"grad_norm": 0.9189796447753906,
|
|
"learning_rate": 0.00012162162162162163,
|
|
"loss": 1.9164,
|
|
"step": 3925
|
|
},
|
|
{
|
|
"epoch": 0.393,
|
|
"grad_norm": 0.799855649471283,
|
|
"learning_rate": 0.00012152152152152152,
|
|
"loss": 1.9502,
|
|
"step": 3930
|
|
},
|
|
{
|
|
"epoch": 0.3935,
|
|
"grad_norm": 0.7941702604293823,
|
|
"learning_rate": 0.00012142142142142143,
|
|
"loss": 1.9306,
|
|
"step": 3935
|
|
},
|
|
{
|
|
"epoch": 0.394,
|
|
"grad_norm": 0.8794786334037781,
|
|
"learning_rate": 0.00012132132132132133,
|
|
"loss": 1.9196,
|
|
"step": 3940
|
|
},
|
|
{
|
|
"epoch": 0.3945,
|
|
"grad_norm": 0.9179593920707703,
|
|
"learning_rate": 0.00012122122122122124,
|
|
"loss": 1.9362,
|
|
"step": 3945
|
|
},
|
|
{
|
|
"epoch": 0.395,
|
|
"grad_norm": 0.7571364641189575,
|
|
"learning_rate": 0.00012112112112112113,
|
|
"loss": 1.9292,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 0.3955,
|
|
"grad_norm": 0.8133040070533752,
|
|
"learning_rate": 0.00012102102102102102,
|
|
"loss": 1.9564,
|
|
"step": 3955
|
|
},
|
|
{
|
|
"epoch": 0.396,
|
|
"grad_norm": 0.7867616415023804,
|
|
"learning_rate": 0.00012092092092092093,
|
|
"loss": 1.9368,
|
|
"step": 3960
|
|
},
|
|
{
|
|
"epoch": 0.3965,
|
|
"grad_norm": 0.768710196018219,
|
|
"learning_rate": 0.00012082082082082083,
|
|
"loss": 1.9146,
|
|
"step": 3965
|
|
},
|
|
{
|
|
"epoch": 0.397,
|
|
"grad_norm": 0.8447054028511047,
|
|
"learning_rate": 0.00012072072072072073,
|
|
"loss": 1.9227,
|
|
"step": 3970
|
|
},
|
|
{
|
|
"epoch": 0.3975,
|
|
"grad_norm": 0.8007540702819824,
|
|
"learning_rate": 0.00012062062062062063,
|
|
"loss": 1.9518,
|
|
"step": 3975
|
|
},
|
|
{
|
|
"epoch": 0.398,
|
|
"grad_norm": 0.8003354668617249,
|
|
"learning_rate": 0.00012052052052052052,
|
|
"loss": 1.9296,
|
|
"step": 3980
|
|
},
|
|
{
|
|
"epoch": 0.3985,
|
|
"grad_norm": 0.7936831712722778,
|
|
"learning_rate": 0.00012042042042042043,
|
|
"loss": 1.935,
|
|
"step": 3985
|
|
},
|
|
{
|
|
"epoch": 0.399,
|
|
"grad_norm": 0.801343560218811,
|
|
"learning_rate": 0.00012032032032032032,
|
|
"loss": 1.951,
|
|
"step": 3990
|
|
},
|
|
{
|
|
"epoch": 0.3995,
|
|
"grad_norm": 2.931938648223877,
|
|
"learning_rate": 0.00012022022022022023,
|
|
"loss": 2.2622,
|
|
"step": 3995
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"grad_norm": 1.9284015893936157,
|
|
"learning_rate": 0.00012012012012012013,
|
|
"loss": 2.7305,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 0.4005,
|
|
"grad_norm": 1.391554355621338,
|
|
"learning_rate": 0.00012002002002002002,
|
|
"loss": 2.4457,
|
|
"step": 4005
|
|
},
|
|
{
|
|
"epoch": 0.401,
|
|
"grad_norm": 1.0932132005691528,
|
|
"learning_rate": 0.00011991991991991993,
|
|
"loss": 2.3634,
|
|
"step": 4010
|
|
},
|
|
{
|
|
"epoch": 0.4015,
|
|
"grad_norm": 0.9506504535675049,
|
|
"learning_rate": 0.00011981981981981982,
|
|
"loss": 2.2795,
|
|
"step": 4015
|
|
},
|
|
{
|
|
"epoch": 0.402,
|
|
"grad_norm": 0.8915263414382935,
|
|
"learning_rate": 0.00011971971971971973,
|
|
"loss": 2.2108,
|
|
"step": 4020
|
|
},
|
|
{
|
|
"epoch": 0.4025,
|
|
"grad_norm": 0.9343470931053162,
|
|
"learning_rate": 0.00011961961961961963,
|
|
"loss": 2.171,
|
|
"step": 4025
|
|
},
|
|
{
|
|
"epoch": 0.403,
|
|
"grad_norm": 0.9694389700889587,
|
|
"learning_rate": 0.00011951951951951952,
|
|
"loss": 2.2073,
|
|
"step": 4030
|
|
},
|
|
{
|
|
"epoch": 0.4035,
|
|
"grad_norm": 0.8602437973022461,
|
|
"learning_rate": 0.00011941941941941943,
|
|
"loss": 2.1673,
|
|
"step": 4035
|
|
},
|
|
{
|
|
"epoch": 0.404,
|
|
"grad_norm": 0.8877463340759277,
|
|
"learning_rate": 0.00011931931931931932,
|
|
"loss": 2.1523,
|
|
"step": 4040
|
|
},
|
|
{
|
|
"epoch": 0.4045,
|
|
"grad_norm": 0.9396592378616333,
|
|
"learning_rate": 0.00011921921921921923,
|
|
"loss": 2.1447,
|
|
"step": 4045
|
|
},
|
|
{
|
|
"epoch": 0.405,
|
|
"grad_norm": 0.887259304523468,
|
|
"learning_rate": 0.00011911911911911913,
|
|
"loss": 2.125,
|
|
"step": 4050
|
|
},
|
|
{
|
|
"epoch": 0.4055,
|
|
"grad_norm": 0.9970149397850037,
|
|
"learning_rate": 0.00011901901901901902,
|
|
"loss": 2.117,
|
|
"step": 4055
|
|
},
|
|
{
|
|
"epoch": 0.406,
|
|
"grad_norm": 0.9191226959228516,
|
|
"learning_rate": 0.00011891891891891893,
|
|
"loss": 2.095,
|
|
"step": 4060
|
|
},
|
|
{
|
|
"epoch": 0.4065,
|
|
"grad_norm": 0.9096970558166504,
|
|
"learning_rate": 0.00011881881881881882,
|
|
"loss": 2.0905,
|
|
"step": 4065
|
|
},
|
|
{
|
|
"epoch": 0.407,
|
|
"grad_norm": 0.9250726103782654,
|
|
"learning_rate": 0.00011871871871871872,
|
|
"loss": 2.0447,
|
|
"step": 4070
|
|
},
|
|
{
|
|
"epoch": 0.4075,
|
|
"grad_norm": 0.848983883857727,
|
|
"learning_rate": 0.00011861861861861863,
|
|
"loss": 2.0659,
|
|
"step": 4075
|
|
},
|
|
{
|
|
"epoch": 0.408,
|
|
"grad_norm": 0.837566614151001,
|
|
"learning_rate": 0.00011851851851851852,
|
|
"loss": 2.0291,
|
|
"step": 4080
|
|
},
|
|
{
|
|
"epoch": 0.4085,
|
|
"grad_norm": 0.8990424871444702,
|
|
"learning_rate": 0.00011841841841841843,
|
|
"loss": 2.0786,
|
|
"step": 4085
|
|
},
|
|
{
|
|
"epoch": 0.409,
|
|
"grad_norm": 0.9345883727073669,
|
|
"learning_rate": 0.00011831831831831832,
|
|
"loss": 2.0498,
|
|
"step": 4090
|
|
},
|
|
{
|
|
"epoch": 0.4095,
|
|
"grad_norm": 0.9079171419143677,
|
|
"learning_rate": 0.0001182182182182182,
|
|
"loss": 2.0502,
|
|
"step": 4095
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"grad_norm": 0.8899641633033752,
|
|
"learning_rate": 0.00011811811811811812,
|
|
"loss": 2.0554,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 0.4105,
|
|
"grad_norm": 0.8282867074012756,
|
|
"learning_rate": 0.00011801801801801802,
|
|
"loss": 2.0017,
|
|
"step": 4105
|
|
},
|
|
{
|
|
"epoch": 0.411,
|
|
"grad_norm": 0.8713417649269104,
|
|
"learning_rate": 0.00011791791791791793,
|
|
"loss": 2.0477,
|
|
"step": 4110
|
|
},
|
|
{
|
|
"epoch": 0.4115,
|
|
"grad_norm": 0.8387519717216492,
|
|
"learning_rate": 0.00011781781781781782,
|
|
"loss": 2.013,
|
|
"step": 4115
|
|
},
|
|
{
|
|
"epoch": 0.412,
|
|
"grad_norm": 0.8442679047584534,
|
|
"learning_rate": 0.00011771771771771771,
|
|
"loss": 2.0011,
|
|
"step": 4120
|
|
},
|
|
{
|
|
"epoch": 0.4125,
|
|
"grad_norm": 0.8642622828483582,
|
|
"learning_rate": 0.00011761761761761762,
|
|
"loss": 2.0232,
|
|
"step": 4125
|
|
},
|
|
{
|
|
"epoch": 0.413,
|
|
"grad_norm": 1.0152161121368408,
|
|
"learning_rate": 0.00011751751751751752,
|
|
"loss": 2.0444,
|
|
"step": 4130
|
|
},
|
|
{
|
|
"epoch": 0.4135,
|
|
"grad_norm": 0.8902932405471802,
|
|
"learning_rate": 0.00011741741741741743,
|
|
"loss": 1.9905,
|
|
"step": 4135
|
|
},
|
|
{
|
|
"epoch": 0.414,
|
|
"grad_norm": 0.9433906078338623,
|
|
"learning_rate": 0.00011731731731731732,
|
|
"loss": 2.0023,
|
|
"step": 4140
|
|
},
|
|
{
|
|
"epoch": 0.4145,
|
|
"grad_norm": 0.8779952526092529,
|
|
"learning_rate": 0.00011721721721721721,
|
|
"loss": 2.0103,
|
|
"step": 4145
|
|
},
|
|
{
|
|
"epoch": 0.415,
|
|
"grad_norm": 0.878989577293396,
|
|
"learning_rate": 0.00011711711711711712,
|
|
"loss": 2.0167,
|
|
"step": 4150
|
|
},
|
|
{
|
|
"epoch": 0.4155,
|
|
"grad_norm": 0.8977020978927612,
|
|
"learning_rate": 0.00011701701701701702,
|
|
"loss": 1.9904,
|
|
"step": 4155
|
|
},
|
|
{
|
|
"epoch": 0.416,
|
|
"grad_norm": 1.0385823249816895,
|
|
"learning_rate": 0.00011691691691691694,
|
|
"loss": 2.0096,
|
|
"step": 4160
|
|
},
|
|
{
|
|
"epoch": 0.4165,
|
|
"grad_norm": 0.9478892683982849,
|
|
"learning_rate": 0.00011681681681681682,
|
|
"loss": 2.0252,
|
|
"step": 4165
|
|
},
|
|
{
|
|
"epoch": 0.417,
|
|
"grad_norm": 0.904509961605072,
|
|
"learning_rate": 0.00011671671671671671,
|
|
"loss": 2.0162,
|
|
"step": 4170
|
|
},
|
|
{
|
|
"epoch": 0.4175,
|
|
"grad_norm": 0.8369683027267456,
|
|
"learning_rate": 0.00011661661661661662,
|
|
"loss": 2.009,
|
|
"step": 4175
|
|
},
|
|
{
|
|
"epoch": 0.418,
|
|
"grad_norm": 0.8266960382461548,
|
|
"learning_rate": 0.00011651651651651652,
|
|
"loss": 1.9785,
|
|
"step": 4180
|
|
},
|
|
{
|
|
"epoch": 0.4185,
|
|
"grad_norm": 0.8447481393814087,
|
|
"learning_rate": 0.00011641641641641644,
|
|
"loss": 1.9693,
|
|
"step": 4185
|
|
},
|
|
{
|
|
"epoch": 0.419,
|
|
"grad_norm": 0.8444758653640747,
|
|
"learning_rate": 0.00011631631631631632,
|
|
"loss": 1.9775,
|
|
"step": 4190
|
|
},
|
|
{
|
|
"epoch": 0.4195,
|
|
"grad_norm": 0.8235093355178833,
|
|
"learning_rate": 0.00011621621621621621,
|
|
"loss": 1.9557,
|
|
"step": 4195
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"grad_norm": 0.9242146611213684,
|
|
"learning_rate": 0.00011611611611611612,
|
|
"loss": 1.9597,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 0.4205,
|
|
"grad_norm": 0.8500037789344788,
|
|
"learning_rate": 0.00011601601601601601,
|
|
"loss": 1.9717,
|
|
"step": 4205
|
|
},
|
|
{
|
|
"epoch": 0.421,
|
|
"grad_norm": 0.8934337496757507,
|
|
"learning_rate": 0.00011591591591591592,
|
|
"loss": 1.9765,
|
|
"step": 4210
|
|
},
|
|
{
|
|
"epoch": 0.4215,
|
|
"grad_norm": 0.9102895855903625,
|
|
"learning_rate": 0.00011581581581581582,
|
|
"loss": 1.9449,
|
|
"step": 4215
|
|
},
|
|
{
|
|
"epoch": 0.422,
|
|
"grad_norm": 0.8801275491714478,
|
|
"learning_rate": 0.00011571571571571571,
|
|
"loss": 1.9337,
|
|
"step": 4220
|
|
},
|
|
{
|
|
"epoch": 0.4225,
|
|
"grad_norm": 0.9498214721679688,
|
|
"learning_rate": 0.00011561561561561562,
|
|
"loss": 1.9817,
|
|
"step": 4225
|
|
},
|
|
{
|
|
"epoch": 0.423,
|
|
"grad_norm": 0.9641104340553284,
|
|
"learning_rate": 0.00011551551551551551,
|
|
"loss": 1.9278,
|
|
"step": 4230
|
|
},
|
|
{
|
|
"epoch": 0.4235,
|
|
"grad_norm": 0.8352410793304443,
|
|
"learning_rate": 0.00011541541541541542,
|
|
"loss": 1.9225,
|
|
"step": 4235
|
|
},
|
|
{
|
|
"epoch": 0.424,
|
|
"grad_norm": 0.8993573784828186,
|
|
"learning_rate": 0.00011531531531531532,
|
|
"loss": 1.959,
|
|
"step": 4240
|
|
},
|
|
{
|
|
"epoch": 0.4245,
|
|
"grad_norm": 1.0262569189071655,
|
|
"learning_rate": 0.00011521521521521521,
|
|
"loss": 1.9422,
|
|
"step": 4245
|
|
},
|
|
{
|
|
"epoch": 0.425,
|
|
"grad_norm": 0.8517004251480103,
|
|
"learning_rate": 0.00011511511511511512,
|
|
"loss": 1.9638,
|
|
"step": 4250
|
|
},
|
|
{
|
|
"epoch": 0.4255,
|
|
"grad_norm": 0.9371127486228943,
|
|
"learning_rate": 0.00011501501501501501,
|
|
"loss": 1.9408,
|
|
"step": 4255
|
|
},
|
|
{
|
|
"epoch": 0.426,
|
|
"grad_norm": 0.8500112295150757,
|
|
"learning_rate": 0.00011491491491491492,
|
|
"loss": 1.9598,
|
|
"step": 4260
|
|
},
|
|
{
|
|
"epoch": 0.4265,
|
|
"grad_norm": 0.8576430082321167,
|
|
"learning_rate": 0.00011481481481481482,
|
|
"loss": 1.9172,
|
|
"step": 4265
|
|
},
|
|
{
|
|
"epoch": 0.427,
|
|
"grad_norm": 0.8595812320709229,
|
|
"learning_rate": 0.00011471471471471471,
|
|
"loss": 1.9099,
|
|
"step": 4270
|
|
},
|
|
{
|
|
"epoch": 0.4275,
|
|
"grad_norm": 0.8851966261863708,
|
|
"learning_rate": 0.00011461461461461463,
|
|
"loss": 1.9295,
|
|
"step": 4275
|
|
},
|
|
{
|
|
"epoch": 0.428,
|
|
"grad_norm": 0.8601343631744385,
|
|
"learning_rate": 0.00011451451451451451,
|
|
"loss": 1.9518,
|
|
"step": 4280
|
|
},
|
|
{
|
|
"epoch": 0.4285,
|
|
"grad_norm": 0.8601483106613159,
|
|
"learning_rate": 0.00011441441441441443,
|
|
"loss": 1.9821,
|
|
"step": 4285
|
|
},
|
|
{
|
|
"epoch": 0.429,
|
|
"grad_norm": 0.8733670711517334,
|
|
"learning_rate": 0.00011431431431431433,
|
|
"loss": 1.9475,
|
|
"step": 4290
|
|
},
|
|
{
|
|
"epoch": 0.4295,
|
|
"grad_norm": 0.8573125004768372,
|
|
"learning_rate": 0.00011421421421421421,
|
|
"loss": 1.92,
|
|
"step": 4295
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"grad_norm": 0.8779380321502686,
|
|
"learning_rate": 0.00011411411411411413,
|
|
"loss": 1.9252,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 0.4305,
|
|
"grad_norm": 0.842447817325592,
|
|
"learning_rate": 0.00011401401401401401,
|
|
"loss": 1.9418,
|
|
"step": 4305
|
|
},
|
|
{
|
|
"epoch": 0.431,
|
|
"grad_norm": 0.8564408421516418,
|
|
"learning_rate": 0.00011391391391391393,
|
|
"loss": 1.8845,
|
|
"step": 4310
|
|
},
|
|
{
|
|
"epoch": 0.4315,
|
|
"grad_norm": 0.9953034520149231,
|
|
"learning_rate": 0.00011381381381381381,
|
|
"loss": 1.9288,
|
|
"step": 4315
|
|
},
|
|
{
|
|
"epoch": 0.432,
|
|
"grad_norm": 0.9179754257202148,
|
|
"learning_rate": 0.00011371371371371371,
|
|
"loss": 1.9554,
|
|
"step": 4320
|
|
},
|
|
{
|
|
"epoch": 0.4325,
|
|
"grad_norm": 0.8732195496559143,
|
|
"learning_rate": 0.00011361361361361363,
|
|
"loss": 1.8688,
|
|
"step": 4325
|
|
},
|
|
{
|
|
"epoch": 0.433,
|
|
"grad_norm": 0.9357179999351501,
|
|
"learning_rate": 0.00011351351351351351,
|
|
"loss": 1.8807,
|
|
"step": 4330
|
|
},
|
|
{
|
|
"epoch": 0.4335,
|
|
"grad_norm": 0.9172531962394714,
|
|
"learning_rate": 0.00011341341341341343,
|
|
"loss": 1.9408,
|
|
"step": 4335
|
|
},
|
|
{
|
|
"epoch": 0.434,
|
|
"grad_norm": 0.8920251131057739,
|
|
"learning_rate": 0.00011331331331331331,
|
|
"loss": 1.8865,
|
|
"step": 4340
|
|
},
|
|
{
|
|
"epoch": 0.4345,
|
|
"grad_norm": 0.9782935380935669,
|
|
"learning_rate": 0.00011321321321321321,
|
|
"loss": 1.8862,
|
|
"step": 4345
|
|
},
|
|
{
|
|
"epoch": 0.435,
|
|
"grad_norm": 0.9634548425674438,
|
|
"learning_rate": 0.00011311311311311313,
|
|
"loss": 1.8773,
|
|
"step": 4350
|
|
},
|
|
{
|
|
"epoch": 0.4355,
|
|
"grad_norm": 0.9303472638130188,
|
|
"learning_rate": 0.00011301301301301301,
|
|
"loss": 1.9031,
|
|
"step": 4355
|
|
},
|
|
{
|
|
"epoch": 0.436,
|
|
"grad_norm": 0.9065335392951965,
|
|
"learning_rate": 0.00011291291291291293,
|
|
"loss": 1.9339,
|
|
"step": 4360
|
|
},
|
|
{
|
|
"epoch": 0.4365,
|
|
"grad_norm": 0.889884889125824,
|
|
"learning_rate": 0.00011281281281281281,
|
|
"loss": 1.9205,
|
|
"step": 4365
|
|
},
|
|
{
|
|
"epoch": 0.437,
|
|
"grad_norm": 1.0310100317001343,
|
|
"learning_rate": 0.00011271271271271271,
|
|
"loss": 1.8889,
|
|
"step": 4370
|
|
},
|
|
{
|
|
"epoch": 0.4375,
|
|
"grad_norm": 0.8499249219894409,
|
|
"learning_rate": 0.00011261261261261263,
|
|
"loss": 1.8759,
|
|
"step": 4375
|
|
},
|
|
{
|
|
"epoch": 0.438,
|
|
"grad_norm": 0.8602859377861023,
|
|
"learning_rate": 0.00011251251251251251,
|
|
"loss": 1.9448,
|
|
"step": 4380
|
|
},
|
|
{
|
|
"epoch": 0.4385,
|
|
"grad_norm": 0.8887655735015869,
|
|
"learning_rate": 0.00011241241241241243,
|
|
"loss": 1.8755,
|
|
"step": 4385
|
|
},
|
|
{
|
|
"epoch": 0.439,
|
|
"grad_norm": 0.8571450114250183,
|
|
"learning_rate": 0.00011231231231231231,
|
|
"loss": 1.8516,
|
|
"step": 4390
|
|
},
|
|
{
|
|
"epoch": 0.4395,
|
|
"grad_norm": 1.1664048433303833,
|
|
"learning_rate": 0.00011221221221221221,
|
|
"loss": 1.8886,
|
|
"step": 4395
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"grad_norm": 0.8851452469825745,
|
|
"learning_rate": 0.00011211211211211213,
|
|
"loss": 1.8924,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 0.4405,
|
|
"grad_norm": 0.9703869223594666,
|
|
"learning_rate": 0.00011201201201201202,
|
|
"loss": 1.8651,
|
|
"step": 4405
|
|
},
|
|
{
|
|
"epoch": 0.441,
|
|
"grad_norm": 0.9467156529426575,
|
|
"learning_rate": 0.00011191191191191193,
|
|
"loss": 1.8938,
|
|
"step": 4410
|
|
},
|
|
{
|
|
"epoch": 0.4415,
|
|
"grad_norm": 0.9211694598197937,
|
|
"learning_rate": 0.00011181181181181182,
|
|
"loss": 1.9098,
|
|
"step": 4415
|
|
},
|
|
{
|
|
"epoch": 0.442,
|
|
"grad_norm": 0.8704224228858948,
|
|
"learning_rate": 0.0001117117117117117,
|
|
"loss": 1.8772,
|
|
"step": 4420
|
|
},
|
|
{
|
|
"epoch": 0.4425,
|
|
"grad_norm": 0.9551790356636047,
|
|
"learning_rate": 0.00011161161161161162,
|
|
"loss": 1.8752,
|
|
"step": 4425
|
|
},
|
|
{
|
|
"epoch": 0.443,
|
|
"grad_norm": 0.9262571334838867,
|
|
"learning_rate": 0.00011151151151151152,
|
|
"loss": 1.8697,
|
|
"step": 4430
|
|
},
|
|
{
|
|
"epoch": 0.4435,
|
|
"grad_norm": 0.8499231338500977,
|
|
"learning_rate": 0.00011141141141141143,
|
|
"loss": 1.8876,
|
|
"step": 4435
|
|
},
|
|
{
|
|
"epoch": 0.444,
|
|
"grad_norm": 0.8752108812332153,
|
|
"learning_rate": 0.00011131131131131132,
|
|
"loss": 1.9077,
|
|
"step": 4440
|
|
},
|
|
{
|
|
"epoch": 0.4445,
|
|
"grad_norm": 0.9217672944068909,
|
|
"learning_rate": 0.0001112112112112112,
|
|
"loss": 1.8906,
|
|
"step": 4445
|
|
},
|
|
{
|
|
"epoch": 0.445,
|
|
"grad_norm": 0.9115353226661682,
|
|
"learning_rate": 0.00011111111111111112,
|
|
"loss": 1.8512,
|
|
"step": 4450
|
|
},
|
|
{
|
|
"epoch": 0.4455,
|
|
"grad_norm": 0.8915761113166809,
|
|
"learning_rate": 0.00011101101101101102,
|
|
"loss": 1.8281,
|
|
"step": 4455
|
|
},
|
|
{
|
|
"epoch": 0.446,
|
|
"grad_norm": 0.8932358026504517,
|
|
"learning_rate": 0.00011091091091091093,
|
|
"loss": 1.8797,
|
|
"step": 4460
|
|
},
|
|
{
|
|
"epoch": 0.4465,
|
|
"grad_norm": 0.9420803189277649,
|
|
"learning_rate": 0.00011081081081081082,
|
|
"loss": 1.8532,
|
|
"step": 4465
|
|
},
|
|
{
|
|
"epoch": 0.447,
|
|
"grad_norm": 0.9950898885726929,
|
|
"learning_rate": 0.0001107107107107107,
|
|
"loss": 1.8449,
|
|
"step": 4470
|
|
},
|
|
{
|
|
"epoch": 0.4475,
|
|
"grad_norm": 0.9600904583930969,
|
|
"learning_rate": 0.00011061061061061062,
|
|
"loss": 1.8968,
|
|
"step": 4475
|
|
},
|
|
{
|
|
"epoch": 0.448,
|
|
"grad_norm": 0.9210178256034851,
|
|
"learning_rate": 0.00011051051051051052,
|
|
"loss": 1.9221,
|
|
"step": 4480
|
|
},
|
|
{
|
|
"epoch": 0.4485,
|
|
"grad_norm": 0.9000430107116699,
|
|
"learning_rate": 0.00011041041041041043,
|
|
"loss": 1.8657,
|
|
"step": 4485
|
|
},
|
|
{
|
|
"epoch": 0.449,
|
|
"grad_norm": 0.8674874901771545,
|
|
"learning_rate": 0.00011031031031031032,
|
|
"loss": 1.8752,
|
|
"step": 4490
|
|
},
|
|
{
|
|
"epoch": 0.4495,
|
|
"grad_norm": 0.943356990814209,
|
|
"learning_rate": 0.0001102102102102102,
|
|
"loss": 1.8241,
|
|
"step": 4495
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"grad_norm": 0.990490734577179,
|
|
"learning_rate": 0.00011011011011011012,
|
|
"loss": 1.859,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 0.4505,
|
|
"grad_norm": 0.9066656827926636,
|
|
"learning_rate": 0.00011001001001001002,
|
|
"loss": 1.8771,
|
|
"step": 4505
|
|
},
|
|
{
|
|
"epoch": 0.451,
|
|
"grad_norm": 0.9334986209869385,
|
|
"learning_rate": 0.00010990990990990993,
|
|
"loss": 1.8818,
|
|
"step": 4510
|
|
},
|
|
{
|
|
"epoch": 0.4515,
|
|
"grad_norm": 0.9269065856933594,
|
|
"learning_rate": 0.00010980980980980982,
|
|
"loss": 1.8744,
|
|
"step": 4515
|
|
},
|
|
{
|
|
"epoch": 0.452,
|
|
"grad_norm": 0.8838782906532288,
|
|
"learning_rate": 0.0001097097097097097,
|
|
"loss": 1.8927,
|
|
"step": 4520
|
|
},
|
|
{
|
|
"epoch": 0.4525,
|
|
"grad_norm": 0.8809515237808228,
|
|
"learning_rate": 0.00010960960960960962,
|
|
"loss": 1.865,
|
|
"step": 4525
|
|
},
|
|
{
|
|
"epoch": 0.453,
|
|
"grad_norm": 0.8855170011520386,
|
|
"learning_rate": 0.0001095095095095095,
|
|
"loss": 1.8636,
|
|
"step": 4530
|
|
},
|
|
{
|
|
"epoch": 0.4535,
|
|
"grad_norm": 0.850659191608429,
|
|
"learning_rate": 0.00010940940940940942,
|
|
"loss": 1.8416,
|
|
"step": 4535
|
|
},
|
|
{
|
|
"epoch": 0.454,
|
|
"grad_norm": 1.209594488143921,
|
|
"learning_rate": 0.00010930930930930932,
|
|
"loss": 1.8474,
|
|
"step": 4540
|
|
},
|
|
{
|
|
"epoch": 0.4545,
|
|
"grad_norm": 0.9443528056144714,
|
|
"learning_rate": 0.0001092092092092092,
|
|
"loss": 1.853,
|
|
"step": 4545
|
|
},
|
|
{
|
|
"epoch": 0.455,
|
|
"grad_norm": 0.8795249462127686,
|
|
"learning_rate": 0.00010910910910910912,
|
|
"loss": 1.8462,
|
|
"step": 4550
|
|
},
|
|
{
|
|
"epoch": 0.4555,
|
|
"grad_norm": 0.9223023653030396,
|
|
"learning_rate": 0.000109009009009009,
|
|
"loss": 1.8644,
|
|
"step": 4555
|
|
},
|
|
{
|
|
"epoch": 0.456,
|
|
"grad_norm": 0.8747789263725281,
|
|
"learning_rate": 0.00010890890890890892,
|
|
"loss": 1.8643,
|
|
"step": 4560
|
|
},
|
|
{
|
|
"epoch": 0.4565,
|
|
"grad_norm": 0.8761250376701355,
|
|
"learning_rate": 0.00010880880880880882,
|
|
"loss": 1.8209,
|
|
"step": 4565
|
|
},
|
|
{
|
|
"epoch": 0.457,
|
|
"grad_norm": 0.8845246434211731,
|
|
"learning_rate": 0.0001087087087087087,
|
|
"loss": 1.8566,
|
|
"step": 4570
|
|
},
|
|
{
|
|
"epoch": 0.4575,
|
|
"grad_norm": 0.8876487016677856,
|
|
"learning_rate": 0.00010860860860860862,
|
|
"loss": 1.8304,
|
|
"step": 4575
|
|
},
|
|
{
|
|
"epoch": 0.458,
|
|
"grad_norm": 0.9176833033561707,
|
|
"learning_rate": 0.0001085085085085085,
|
|
"loss": 1.8323,
|
|
"step": 4580
|
|
},
|
|
{
|
|
"epoch": 0.4585,
|
|
"grad_norm": 0.9132845997810364,
|
|
"learning_rate": 0.00010840840840840842,
|
|
"loss": 1.8235,
|
|
"step": 4585
|
|
},
|
|
{
|
|
"epoch": 0.459,
|
|
"grad_norm": 0.8965134024620056,
|
|
"learning_rate": 0.00010830830830830832,
|
|
"loss": 1.8405,
|
|
"step": 4590
|
|
},
|
|
{
|
|
"epoch": 0.4595,
|
|
"grad_norm": 0.8731141686439514,
|
|
"learning_rate": 0.00010820820820820821,
|
|
"loss": 1.8521,
|
|
"step": 4595
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"grad_norm": 0.8298085927963257,
|
|
"learning_rate": 0.00010810810810810812,
|
|
"loss": 1.8678,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 0.4605,
|
|
"grad_norm": 0.9257004261016846,
|
|
"learning_rate": 0.00010800800800800801,
|
|
"loss": 1.8492,
|
|
"step": 4605
|
|
},
|
|
{
|
|
"epoch": 0.461,
|
|
"grad_norm": 0.8877549767494202,
|
|
"learning_rate": 0.00010790790790790792,
|
|
"loss": 1.8265,
|
|
"step": 4610
|
|
},
|
|
{
|
|
"epoch": 0.4615,
|
|
"grad_norm": 0.9602069854736328,
|
|
"learning_rate": 0.00010780780780780782,
|
|
"loss": 1.8731,
|
|
"step": 4615
|
|
},
|
|
{
|
|
"epoch": 0.462,
|
|
"grad_norm": 0.8777331113815308,
|
|
"learning_rate": 0.00010770770770770771,
|
|
"loss": 1.8595,
|
|
"step": 4620
|
|
},
|
|
{
|
|
"epoch": 0.4625,
|
|
"grad_norm": 1.8534454107284546,
|
|
"learning_rate": 0.00010760760760760762,
|
|
"loss": 3.3102,
|
|
"step": 4625
|
|
},
|
|
{
|
|
"epoch": 0.463,
|
|
"grad_norm": 1.721049189567566,
|
|
"learning_rate": 0.00010750750750750751,
|
|
"loss": 3.1247,
|
|
"step": 4630
|
|
},
|
|
{
|
|
"epoch": 0.4635,
|
|
"grad_norm": 1.1871305704116821,
|
|
"learning_rate": 0.00010740740740740742,
|
|
"loss": 3.0256,
|
|
"step": 4635
|
|
},
|
|
{
|
|
"epoch": 0.464,
|
|
"grad_norm": 0.9916480183601379,
|
|
"learning_rate": 0.00010730730730730731,
|
|
"loss": 2.9173,
|
|
"step": 4640
|
|
},
|
|
{
|
|
"epoch": 0.4645,
|
|
"grad_norm": 0.8853825330734253,
|
|
"learning_rate": 0.00010720720720720721,
|
|
"loss": 2.8852,
|
|
"step": 4645
|
|
},
|
|
{
|
|
"epoch": 0.465,
|
|
"grad_norm": 0.9272099733352661,
|
|
"learning_rate": 0.00010710710710710712,
|
|
"loss": 2.8115,
|
|
"step": 4650
|
|
},
|
|
{
|
|
"epoch": 0.4655,
|
|
"grad_norm": 0.8129410743713379,
|
|
"learning_rate": 0.00010700700700700701,
|
|
"loss": 2.7959,
|
|
"step": 4655
|
|
},
|
|
{
|
|
"epoch": 0.466,
|
|
"grad_norm": 0.8323870301246643,
|
|
"learning_rate": 0.00010690690690690692,
|
|
"loss": 2.7888,
|
|
"step": 4660
|
|
},
|
|
{
|
|
"epoch": 0.4665,
|
|
"grad_norm": 0.8004547953605652,
|
|
"learning_rate": 0.00010680680680680681,
|
|
"loss": 2.7576,
|
|
"step": 4665
|
|
},
|
|
{
|
|
"epoch": 0.467,
|
|
"grad_norm": 0.7928503155708313,
|
|
"learning_rate": 0.00010670670670670671,
|
|
"loss": 2.7664,
|
|
"step": 4670
|
|
},
|
|
{
|
|
"epoch": 0.4675,
|
|
"grad_norm": 0.773970901966095,
|
|
"learning_rate": 0.00010660660660660662,
|
|
"loss": 2.7214,
|
|
"step": 4675
|
|
},
|
|
{
|
|
"epoch": 0.468,
|
|
"grad_norm": 0.8073243498802185,
|
|
"learning_rate": 0.00010650650650650651,
|
|
"loss": 2.7298,
|
|
"step": 4680
|
|
},
|
|
{
|
|
"epoch": 0.4685,
|
|
"grad_norm": 0.7617459297180176,
|
|
"learning_rate": 0.00010640640640640642,
|
|
"loss": 2.7076,
|
|
"step": 4685
|
|
},
|
|
{
|
|
"epoch": 0.469,
|
|
"grad_norm": 0.8188638687133789,
|
|
"learning_rate": 0.00010630630630630631,
|
|
"loss": 2.7152,
|
|
"step": 4690
|
|
},
|
|
{
|
|
"epoch": 0.4695,
|
|
"grad_norm": 0.8789124488830566,
|
|
"learning_rate": 0.00010620620620620621,
|
|
"loss": 2.6893,
|
|
"step": 4695
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"grad_norm": 0.8619881272315979,
|
|
"learning_rate": 0.00010610610610610612,
|
|
"loss": 2.695,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 0.4705,
|
|
"grad_norm": 0.7400351762771606,
|
|
"learning_rate": 0.00010600600600600601,
|
|
"loss": 2.6962,
|
|
"step": 4705
|
|
},
|
|
{
|
|
"epoch": 0.471,
|
|
"grad_norm": 0.7460540533065796,
|
|
"learning_rate": 0.0001059059059059059,
|
|
"loss": 2.6435,
|
|
"step": 4710
|
|
},
|
|
{
|
|
"epoch": 0.4715,
|
|
"grad_norm": 0.8136820793151855,
|
|
"learning_rate": 0.00010580580580580581,
|
|
"loss": 2.6901,
|
|
"step": 4715
|
|
},
|
|
{
|
|
"epoch": 0.472,
|
|
"grad_norm": 0.767340362071991,
|
|
"learning_rate": 0.00010570570570570571,
|
|
"loss": 2.6748,
|
|
"step": 4720
|
|
},
|
|
{
|
|
"epoch": 0.4725,
|
|
"grad_norm": 0.7789919376373291,
|
|
"learning_rate": 0.00010560560560560562,
|
|
"loss": 2.6534,
|
|
"step": 4725
|
|
},
|
|
{
|
|
"epoch": 0.473,
|
|
"grad_norm": 0.8018049597740173,
|
|
"learning_rate": 0.00010550550550550551,
|
|
"loss": 2.652,
|
|
"step": 4730
|
|
},
|
|
{
|
|
"epoch": 0.4735,
|
|
"grad_norm": 0.753203809261322,
|
|
"learning_rate": 0.0001054054054054054,
|
|
"loss": 2.6453,
|
|
"step": 4735
|
|
},
|
|
{
|
|
"epoch": 0.474,
|
|
"grad_norm": 0.758394718170166,
|
|
"learning_rate": 0.00010530530530530531,
|
|
"loss": 2.6405,
|
|
"step": 4740
|
|
},
|
|
{
|
|
"epoch": 0.4745,
|
|
"grad_norm": 0.7693637609481812,
|
|
"learning_rate": 0.0001052052052052052,
|
|
"loss": 2.5966,
|
|
"step": 4745
|
|
},
|
|
{
|
|
"epoch": 0.475,
|
|
"grad_norm": 0.7242591381072998,
|
|
"learning_rate": 0.00010510510510510511,
|
|
"loss": 2.6375,
|
|
"step": 4750
|
|
},
|
|
{
|
|
"epoch": 0.4755,
|
|
"grad_norm": 0.8229799270629883,
|
|
"learning_rate": 0.00010500500500500501,
|
|
"loss": 2.6175,
|
|
"step": 4755
|
|
},
|
|
{
|
|
"epoch": 0.476,
|
|
"grad_norm": 0.7987326383590698,
|
|
"learning_rate": 0.0001049049049049049,
|
|
"loss": 2.6413,
|
|
"step": 4760
|
|
},
|
|
{
|
|
"epoch": 0.4765,
|
|
"grad_norm": 0.7526589632034302,
|
|
"learning_rate": 0.00010480480480480481,
|
|
"loss": 2.6501,
|
|
"step": 4765
|
|
},
|
|
{
|
|
"epoch": 0.477,
|
|
"grad_norm": 0.7456768155097961,
|
|
"learning_rate": 0.0001047047047047047,
|
|
"loss": 2.5909,
|
|
"step": 4770
|
|
},
|
|
{
|
|
"epoch": 0.4775,
|
|
"grad_norm": 0.8570939302444458,
|
|
"learning_rate": 0.00010460460460460461,
|
|
"loss": 2.6188,
|
|
"step": 4775
|
|
},
|
|
{
|
|
"epoch": 0.478,
|
|
"grad_norm": 0.8006309866905212,
|
|
"learning_rate": 0.00010450450450450451,
|
|
"loss": 2.6194,
|
|
"step": 4780
|
|
},
|
|
{
|
|
"epoch": 0.4785,
|
|
"grad_norm": 0.8089258074760437,
|
|
"learning_rate": 0.0001044044044044044,
|
|
"loss": 2.6106,
|
|
"step": 4785
|
|
},
|
|
{
|
|
"epoch": 0.479,
|
|
"grad_norm": 0.7953463196754456,
|
|
"learning_rate": 0.00010430430430430431,
|
|
"loss": 2.5841,
|
|
"step": 4790
|
|
},
|
|
{
|
|
"epoch": 0.4795,
|
|
"grad_norm": 0.8188151121139526,
|
|
"learning_rate": 0.0001042042042042042,
|
|
"loss": 2.6067,
|
|
"step": 4795
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"grad_norm": 0.769829511642456,
|
|
"learning_rate": 0.00010410410410410411,
|
|
"loss": 2.5939,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 0.4805,
|
|
"grad_norm": 0.7330535650253296,
|
|
"learning_rate": 0.00010400400400400401,
|
|
"loss": 2.5979,
|
|
"step": 4805
|
|
},
|
|
{
|
|
"epoch": 0.481,
|
|
"grad_norm": 0.817407488822937,
|
|
"learning_rate": 0.0001039039039039039,
|
|
"loss": 2.6022,
|
|
"step": 4810
|
|
},
|
|
{
|
|
"epoch": 0.4815,
|
|
"grad_norm": 0.7732321619987488,
|
|
"learning_rate": 0.00010380380380380381,
|
|
"loss": 2.5937,
|
|
"step": 4815
|
|
},
|
|
{
|
|
"epoch": 0.482,
|
|
"grad_norm": 0.800410270690918,
|
|
"learning_rate": 0.0001037037037037037,
|
|
"loss": 2.5957,
|
|
"step": 4820
|
|
},
|
|
{
|
|
"epoch": 0.4825,
|
|
"grad_norm": 0.7640219926834106,
|
|
"learning_rate": 0.00010360360360360361,
|
|
"loss": 2.5787,
|
|
"step": 4825
|
|
},
|
|
{
|
|
"epoch": 0.483,
|
|
"grad_norm": 0.7671935558319092,
|
|
"learning_rate": 0.00010350350350350351,
|
|
"loss": 2.5831,
|
|
"step": 4830
|
|
},
|
|
{
|
|
"epoch": 0.4835,
|
|
"grad_norm": 0.8024823069572449,
|
|
"learning_rate": 0.0001034034034034034,
|
|
"loss": 2.5631,
|
|
"step": 4835
|
|
},
|
|
{
|
|
"epoch": 0.484,
|
|
"grad_norm": 0.770573616027832,
|
|
"learning_rate": 0.00010330330330330331,
|
|
"loss": 2.5876,
|
|
"step": 4840
|
|
},
|
|
{
|
|
"epoch": 0.4845,
|
|
"grad_norm": 0.7440877556800842,
|
|
"learning_rate": 0.0001032032032032032,
|
|
"loss": 2.5764,
|
|
"step": 4845
|
|
},
|
|
{
|
|
"epoch": 0.485,
|
|
"grad_norm": 0.758750319480896,
|
|
"learning_rate": 0.00010310310310310311,
|
|
"loss": 2.5851,
|
|
"step": 4850
|
|
},
|
|
{
|
|
"epoch": 0.4855,
|
|
"grad_norm": 0.7744156122207642,
|
|
"learning_rate": 0.000103003003003003,
|
|
"loss": 2.5564,
|
|
"step": 4855
|
|
},
|
|
{
|
|
"epoch": 0.486,
|
|
"grad_norm": 0.7538742423057556,
|
|
"learning_rate": 0.0001029029029029029,
|
|
"loss": 2.5862,
|
|
"step": 4860
|
|
},
|
|
{
|
|
"epoch": 0.4865,
|
|
"grad_norm": 0.7862333059310913,
|
|
"learning_rate": 0.00010280280280280281,
|
|
"loss": 2.5591,
|
|
"step": 4865
|
|
},
|
|
{
|
|
"epoch": 0.487,
|
|
"grad_norm": 0.7841285467147827,
|
|
"learning_rate": 0.0001027027027027027,
|
|
"loss": 2.5757,
|
|
"step": 4870
|
|
},
|
|
{
|
|
"epoch": 0.4875,
|
|
"grad_norm": 0.8739153146743774,
|
|
"learning_rate": 0.00010260260260260261,
|
|
"loss": 2.5542,
|
|
"step": 4875
|
|
},
|
|
{
|
|
"epoch": 0.488,
|
|
"grad_norm": 0.744562029838562,
|
|
"learning_rate": 0.0001025025025025025,
|
|
"loss": 2.5323,
|
|
"step": 4880
|
|
},
|
|
{
|
|
"epoch": 0.4885,
|
|
"grad_norm": 0.7456634044647217,
|
|
"learning_rate": 0.0001024024024024024,
|
|
"loss": 2.5738,
|
|
"step": 4885
|
|
},
|
|
{
|
|
"epoch": 0.489,
|
|
"grad_norm": 0.835354208946228,
|
|
"learning_rate": 0.00010230230230230232,
|
|
"loss": 2.5632,
|
|
"step": 4890
|
|
},
|
|
{
|
|
"epoch": 0.4895,
|
|
"grad_norm": 0.8080532550811768,
|
|
"learning_rate": 0.0001022022022022022,
|
|
"loss": 2.5492,
|
|
"step": 4895
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"grad_norm": 0.7578943967819214,
|
|
"learning_rate": 0.00010210210210210212,
|
|
"loss": 2.5519,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 0.4905,
|
|
"grad_norm": 0.7827736735343933,
|
|
"learning_rate": 0.000102002002002002,
|
|
"loss": 2.5651,
|
|
"step": 4905
|
|
},
|
|
{
|
|
"epoch": 0.491,
|
|
"grad_norm": 0.7860758304595947,
|
|
"learning_rate": 0.0001019019019019019,
|
|
"loss": 2.587,
|
|
"step": 4910
|
|
},
|
|
{
|
|
"epoch": 0.4915,
|
|
"grad_norm": 0.7887434959411621,
|
|
"learning_rate": 0.00010180180180180182,
|
|
"loss": 2.5429,
|
|
"step": 4915
|
|
},
|
|
{
|
|
"epoch": 0.492,
|
|
"grad_norm": 0.7806670665740967,
|
|
"learning_rate": 0.0001017017017017017,
|
|
"loss": 2.5506,
|
|
"step": 4920
|
|
},
|
|
{
|
|
"epoch": 0.4925,
|
|
"grad_norm": 0.7844783663749695,
|
|
"learning_rate": 0.00010160160160160162,
|
|
"loss": 2.5566,
|
|
"step": 4925
|
|
},
|
|
{
|
|
"epoch": 0.493,
|
|
"grad_norm": 0.7470987439155579,
|
|
"learning_rate": 0.0001015015015015015,
|
|
"loss": 2.5615,
|
|
"step": 4930
|
|
},
|
|
{
|
|
"epoch": 0.4935,
|
|
"grad_norm": 0.7789111137390137,
|
|
"learning_rate": 0.0001014014014014014,
|
|
"loss": 2.5579,
|
|
"step": 4935
|
|
},
|
|
{
|
|
"epoch": 0.494,
|
|
"grad_norm": 0.7943547368049622,
|
|
"learning_rate": 0.00010130130130130132,
|
|
"loss": 2.5328,
|
|
"step": 4940
|
|
},
|
|
{
|
|
"epoch": 0.4945,
|
|
"grad_norm": 0.7838632464408875,
|
|
"learning_rate": 0.0001012012012012012,
|
|
"loss": 2.5306,
|
|
"step": 4945
|
|
},
|
|
{
|
|
"epoch": 0.495,
|
|
"grad_norm": 0.7697449922561646,
|
|
"learning_rate": 0.00010110110110110112,
|
|
"loss": 2.5541,
|
|
"step": 4950
|
|
},
|
|
{
|
|
"epoch": 0.4955,
|
|
"grad_norm": 0.7821655869483948,
|
|
"learning_rate": 0.000101001001001001,
|
|
"loss": 2.5184,
|
|
"step": 4955
|
|
},
|
|
{
|
|
"epoch": 0.496,
|
|
"grad_norm": 0.7613987326622009,
|
|
"learning_rate": 0.00010090090090090089,
|
|
"loss": 2.5349,
|
|
"step": 4960
|
|
},
|
|
{
|
|
"epoch": 0.4965,
|
|
"grad_norm": 0.7794037461280823,
|
|
"learning_rate": 0.0001008008008008008,
|
|
"loss": 2.5391,
|
|
"step": 4965
|
|
},
|
|
{
|
|
"epoch": 0.497,
|
|
"grad_norm": 0.7633383870124817,
|
|
"learning_rate": 0.0001007007007007007,
|
|
"loss": 2.521,
|
|
"step": 4970
|
|
},
|
|
{
|
|
"epoch": 0.4975,
|
|
"grad_norm": 0.7773224115371704,
|
|
"learning_rate": 0.00010060060060060062,
|
|
"loss": 2.5439,
|
|
"step": 4975
|
|
},
|
|
{
|
|
"epoch": 0.498,
|
|
"grad_norm": 0.8238345384597778,
|
|
"learning_rate": 0.0001005005005005005,
|
|
"loss": 2.5084,
|
|
"step": 4980
|
|
},
|
|
{
|
|
"epoch": 0.4985,
|
|
"grad_norm": 0.8315170407295227,
|
|
"learning_rate": 0.00010040040040040039,
|
|
"loss": 2.5312,
|
|
"step": 4985
|
|
},
|
|
{
|
|
"epoch": 0.499,
|
|
"grad_norm": 0.8703632950782776,
|
|
"learning_rate": 0.0001003003003003003,
|
|
"loss": 2.5212,
|
|
"step": 4990
|
|
},
|
|
{
|
|
"epoch": 0.4995,
|
|
"grad_norm": 0.8008772134780884,
|
|
"learning_rate": 0.0001002002002002002,
|
|
"loss": 2.5373,
|
|
"step": 4995
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"grad_norm": 0.7883898019790649,
|
|
"learning_rate": 0.00010010010010010012,
|
|
"loss": 2.5194,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 0.5005,
|
|
"grad_norm": 0.7612689733505249,
|
|
"learning_rate": 0.0001,
|
|
"loss": 2.5253,
|
|
"step": 5005
|
|
},
|
|
{
|
|
"epoch": 0.501,
|
|
"grad_norm": 0.8001552224159241,
|
|
"learning_rate": 9.98998998998999e-05,
|
|
"loss": 2.5255,
|
|
"step": 5010
|
|
},
|
|
{
|
|
"epoch": 0.5015,
|
|
"grad_norm": 0.8067675828933716,
|
|
"learning_rate": 9.97997997997998e-05,
|
|
"loss": 2.5127,
|
|
"step": 5015
|
|
},
|
|
{
|
|
"epoch": 0.502,
|
|
"grad_norm": 0.7785462737083435,
|
|
"learning_rate": 9.96996996996997e-05,
|
|
"loss": 2.5064,
|
|
"step": 5020
|
|
},
|
|
{
|
|
"epoch": 0.5025,
|
|
"grad_norm": 0.8287351727485657,
|
|
"learning_rate": 9.95995995995996e-05,
|
|
"loss": 2.5255,
|
|
"step": 5025
|
|
},
|
|
{
|
|
"epoch": 0.503,
|
|
"grad_norm": 0.7444617748260498,
|
|
"learning_rate": 9.94994994994995e-05,
|
|
"loss": 2.5084,
|
|
"step": 5030
|
|
},
|
|
{
|
|
"epoch": 0.5035,
|
|
"grad_norm": 0.7671459317207336,
|
|
"learning_rate": 9.93993993993994e-05,
|
|
"loss": 2.5063,
|
|
"step": 5035
|
|
},
|
|
{
|
|
"epoch": 0.504,
|
|
"grad_norm": 0.7833889722824097,
|
|
"learning_rate": 9.92992992992993e-05,
|
|
"loss": 2.4893,
|
|
"step": 5040
|
|
},
|
|
{
|
|
"epoch": 0.5045,
|
|
"grad_norm": 0.7985659837722778,
|
|
"learning_rate": 9.91991991991992e-05,
|
|
"loss": 2.5383,
|
|
"step": 5045
|
|
},
|
|
{
|
|
"epoch": 0.505,
|
|
"grad_norm": 0.8089566230773926,
|
|
"learning_rate": 9.90990990990991e-05,
|
|
"loss": 2.5032,
|
|
"step": 5050
|
|
},
|
|
{
|
|
"epoch": 0.5055,
|
|
"grad_norm": 0.8146255016326904,
|
|
"learning_rate": 9.8998998998999e-05,
|
|
"loss": 2.4998,
|
|
"step": 5055
|
|
},
|
|
{
|
|
"epoch": 0.506,
|
|
"grad_norm": 0.7463052272796631,
|
|
"learning_rate": 9.88988988988989e-05,
|
|
"loss": 2.516,
|
|
"step": 5060
|
|
},
|
|
{
|
|
"epoch": 0.5065,
|
|
"grad_norm": 0.8386415243148804,
|
|
"learning_rate": 9.87987987987988e-05,
|
|
"loss": 2.5193,
|
|
"step": 5065
|
|
},
|
|
{
|
|
"epoch": 0.507,
|
|
"grad_norm": 0.766620934009552,
|
|
"learning_rate": 9.869869869869869e-05,
|
|
"loss": 2.5227,
|
|
"step": 5070
|
|
},
|
|
{
|
|
"epoch": 0.5075,
|
|
"grad_norm": 0.881860613822937,
|
|
"learning_rate": 9.85985985985986e-05,
|
|
"loss": 2.5267,
|
|
"step": 5075
|
|
},
|
|
{
|
|
"epoch": 0.508,
|
|
"grad_norm": 0.8524314165115356,
|
|
"learning_rate": 9.849849849849851e-05,
|
|
"loss": 2.5039,
|
|
"step": 5080
|
|
},
|
|
{
|
|
"epoch": 0.5085,
|
|
"grad_norm": 0.7791292667388916,
|
|
"learning_rate": 9.839839839839841e-05,
|
|
"loss": 2.5291,
|
|
"step": 5085
|
|
},
|
|
{
|
|
"epoch": 0.509,
|
|
"grad_norm": 0.805531919002533,
|
|
"learning_rate": 9.829829829829831e-05,
|
|
"loss": 2.5162,
|
|
"step": 5090
|
|
},
|
|
{
|
|
"epoch": 0.5095,
|
|
"grad_norm": 0.8797053694725037,
|
|
"learning_rate": 9.81981981981982e-05,
|
|
"loss": 2.4864,
|
|
"step": 5095
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"grad_norm": 0.7774877548217773,
|
|
"learning_rate": 9.809809809809811e-05,
|
|
"loss": 2.5218,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 0.5105,
|
|
"grad_norm": 0.743796706199646,
|
|
"learning_rate": 9.799799799799801e-05,
|
|
"loss": 2.5168,
|
|
"step": 5105
|
|
},
|
|
{
|
|
"epoch": 0.511,
|
|
"grad_norm": 0.8754458427429199,
|
|
"learning_rate": 9.789789789789791e-05,
|
|
"loss": 2.4804,
|
|
"step": 5110
|
|
},
|
|
{
|
|
"epoch": 0.5115,
|
|
"grad_norm": 0.8188790082931519,
|
|
"learning_rate": 9.779779779779781e-05,
|
|
"loss": 2.5051,
|
|
"step": 5115
|
|
},
|
|
{
|
|
"epoch": 0.512,
|
|
"grad_norm": 0.8211973309516907,
|
|
"learning_rate": 9.76976976976977e-05,
|
|
"loss": 2.4978,
|
|
"step": 5120
|
|
},
|
|
{
|
|
"epoch": 0.5125,
|
|
"grad_norm": 0.8346189856529236,
|
|
"learning_rate": 9.75975975975976e-05,
|
|
"loss": 2.509,
|
|
"step": 5125
|
|
},
|
|
{
|
|
"epoch": 0.513,
|
|
"grad_norm": 0.7698317766189575,
|
|
"learning_rate": 9.749749749749751e-05,
|
|
"loss": 2.5271,
|
|
"step": 5130
|
|
},
|
|
{
|
|
"epoch": 0.5135,
|
|
"grad_norm": 0.7541659474372864,
|
|
"learning_rate": 9.739739739739741e-05,
|
|
"loss": 2.5088,
|
|
"step": 5135
|
|
},
|
|
{
|
|
"epoch": 0.514,
|
|
"grad_norm": 0.772386372089386,
|
|
"learning_rate": 9.729729729729731e-05,
|
|
"loss": 2.4997,
|
|
"step": 5140
|
|
},
|
|
{
|
|
"epoch": 0.5145,
|
|
"grad_norm": 0.7649770379066467,
|
|
"learning_rate": 9.71971971971972e-05,
|
|
"loss": 2.5038,
|
|
"step": 5145
|
|
},
|
|
{
|
|
"epoch": 0.515,
|
|
"grad_norm": 0.7980897426605225,
|
|
"learning_rate": 9.70970970970971e-05,
|
|
"loss": 2.4947,
|
|
"step": 5150
|
|
},
|
|
{
|
|
"epoch": 0.5155,
|
|
"grad_norm": 0.7903638482093811,
|
|
"learning_rate": 9.699699699699701e-05,
|
|
"loss": 2.5041,
|
|
"step": 5155
|
|
},
|
|
{
|
|
"epoch": 0.516,
|
|
"grad_norm": 0.7857434153556824,
|
|
"learning_rate": 9.689689689689691e-05,
|
|
"loss": 2.4701,
|
|
"step": 5160
|
|
},
|
|
{
|
|
"epoch": 0.5165,
|
|
"grad_norm": 0.7594984173774719,
|
|
"learning_rate": 9.67967967967968e-05,
|
|
"loss": 2.482,
|
|
"step": 5165
|
|
},
|
|
{
|
|
"epoch": 0.517,
|
|
"grad_norm": 0.7621155977249146,
|
|
"learning_rate": 9.66966966966967e-05,
|
|
"loss": 2.4901,
|
|
"step": 5170
|
|
},
|
|
{
|
|
"epoch": 0.5175,
|
|
"grad_norm": 0.7377651333808899,
|
|
"learning_rate": 9.65965965965966e-05,
|
|
"loss": 2.4988,
|
|
"step": 5175
|
|
},
|
|
{
|
|
"epoch": 0.518,
|
|
"grad_norm": 0.7642138004302979,
|
|
"learning_rate": 9.64964964964965e-05,
|
|
"loss": 2.4907,
|
|
"step": 5180
|
|
},
|
|
{
|
|
"epoch": 0.5185,
|
|
"grad_norm": 0.7919954657554626,
|
|
"learning_rate": 9.639639639639641e-05,
|
|
"loss": 2.4889,
|
|
"step": 5185
|
|
},
|
|
{
|
|
"epoch": 0.519,
|
|
"grad_norm": 0.7576303482055664,
|
|
"learning_rate": 9.62962962962963e-05,
|
|
"loss": 2.4957,
|
|
"step": 5190
|
|
},
|
|
{
|
|
"epoch": 0.5195,
|
|
"grad_norm": 0.7910977602005005,
|
|
"learning_rate": 9.61961961961962e-05,
|
|
"loss": 2.514,
|
|
"step": 5195
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"grad_norm": 0.7882610559463501,
|
|
"learning_rate": 9.60960960960961e-05,
|
|
"loss": 2.5047,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 0.5205,
|
|
"grad_norm": 0.7650166749954224,
|
|
"learning_rate": 9.5995995995996e-05,
|
|
"loss": 2.5049,
|
|
"step": 5205
|
|
},
|
|
{
|
|
"epoch": 0.521,
|
|
"grad_norm": 0.7797821760177612,
|
|
"learning_rate": 9.589589589589591e-05,
|
|
"loss": 2.4762,
|
|
"step": 5210
|
|
},
|
|
{
|
|
"epoch": 0.5215,
|
|
"grad_norm": 0.8017827272415161,
|
|
"learning_rate": 9.57957957957958e-05,
|
|
"loss": 2.4738,
|
|
"step": 5215
|
|
},
|
|
{
|
|
"epoch": 0.522,
|
|
"grad_norm": 0.8328955769538879,
|
|
"learning_rate": 9.56956956956957e-05,
|
|
"loss": 2.4818,
|
|
"step": 5220
|
|
},
|
|
{
|
|
"epoch": 0.5225,
|
|
"grad_norm": 0.7545986771583557,
|
|
"learning_rate": 9.55955955955956e-05,
|
|
"loss": 2.4761,
|
|
"step": 5225
|
|
},
|
|
{
|
|
"epoch": 0.523,
|
|
"grad_norm": 0.7941020131111145,
|
|
"learning_rate": 9.54954954954955e-05,
|
|
"loss": 2.4854,
|
|
"step": 5230
|
|
},
|
|
{
|
|
"epoch": 0.5235,
|
|
"grad_norm": 0.7584949135780334,
|
|
"learning_rate": 9.53953953953954e-05,
|
|
"loss": 2.5069,
|
|
"step": 5235
|
|
},
|
|
{
|
|
"epoch": 0.524,
|
|
"grad_norm": 0.7539657950401306,
|
|
"learning_rate": 9.52952952952953e-05,
|
|
"loss": 2.4839,
|
|
"step": 5240
|
|
},
|
|
{
|
|
"epoch": 0.5245,
|
|
"grad_norm": 0.7383254766464233,
|
|
"learning_rate": 9.51951951951952e-05,
|
|
"loss": 2.4814,
|
|
"step": 5245
|
|
},
|
|
{
|
|
"epoch": 0.525,
|
|
"grad_norm": 0.7786048650741577,
|
|
"learning_rate": 9.50950950950951e-05,
|
|
"loss": 2.4866,
|
|
"step": 5250
|
|
},
|
|
{
|
|
"epoch": 0.5255,
|
|
"grad_norm": 0.8005134463310242,
|
|
"learning_rate": 9.4994994994995e-05,
|
|
"loss": 2.5035,
|
|
"step": 5255
|
|
},
|
|
{
|
|
"epoch": 0.526,
|
|
"grad_norm": 0.8875206112861633,
|
|
"learning_rate": 9.48948948948949e-05,
|
|
"loss": 2.5053,
|
|
"step": 5260
|
|
},
|
|
{
|
|
"epoch": 0.5265,
|
|
"grad_norm": 0.8146237134933472,
|
|
"learning_rate": 9.47947947947948e-05,
|
|
"loss": 2.4957,
|
|
"step": 5265
|
|
},
|
|
{
|
|
"epoch": 0.527,
|
|
"grad_norm": 0.7780102491378784,
|
|
"learning_rate": 9.46946946946947e-05,
|
|
"loss": 2.4801,
|
|
"step": 5270
|
|
},
|
|
{
|
|
"epoch": 0.5275,
|
|
"grad_norm": 0.7808942794799805,
|
|
"learning_rate": 9.45945945945946e-05,
|
|
"loss": 2.4874,
|
|
"step": 5275
|
|
},
|
|
{
|
|
"epoch": 0.528,
|
|
"grad_norm": 0.7710458636283875,
|
|
"learning_rate": 9.44944944944945e-05,
|
|
"loss": 2.4638,
|
|
"step": 5280
|
|
},
|
|
{
|
|
"epoch": 0.5285,
|
|
"grad_norm": 0.8187270760536194,
|
|
"learning_rate": 9.43943943943944e-05,
|
|
"loss": 2.4815,
|
|
"step": 5285
|
|
},
|
|
{
|
|
"epoch": 0.529,
|
|
"grad_norm": 0.8260098695755005,
|
|
"learning_rate": 9.42942942942943e-05,
|
|
"loss": 2.4785,
|
|
"step": 5290
|
|
},
|
|
{
|
|
"epoch": 0.5295,
|
|
"grad_norm": 0.7852158546447754,
|
|
"learning_rate": 9.41941941941942e-05,
|
|
"loss": 2.4868,
|
|
"step": 5295
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"grad_norm": 0.7950302362442017,
|
|
"learning_rate": 9.40940940940941e-05,
|
|
"loss": 2.4811,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 0.5305,
|
|
"grad_norm": 0.7699114084243774,
|
|
"learning_rate": 9.3993993993994e-05,
|
|
"loss": 2.4562,
|
|
"step": 5305
|
|
},
|
|
{
|
|
"epoch": 0.531,
|
|
"grad_norm": 0.757371723651886,
|
|
"learning_rate": 9.38938938938939e-05,
|
|
"loss": 2.4517,
|
|
"step": 5310
|
|
},
|
|
{
|
|
"epoch": 0.5315,
|
|
"grad_norm": 0.8255384564399719,
|
|
"learning_rate": 9.37937937937938e-05,
|
|
"loss": 2.4831,
|
|
"step": 5315
|
|
},
|
|
{
|
|
"epoch": 0.532,
|
|
"grad_norm": 0.7833475470542908,
|
|
"learning_rate": 9.36936936936937e-05,
|
|
"loss": 2.4769,
|
|
"step": 5320
|
|
},
|
|
{
|
|
"epoch": 0.5325,
|
|
"grad_norm": 0.7572566866874695,
|
|
"learning_rate": 9.35935935935936e-05,
|
|
"loss": 2.4686,
|
|
"step": 5325
|
|
},
|
|
{
|
|
"epoch": 0.533,
|
|
"grad_norm": 0.8141648173332214,
|
|
"learning_rate": 9.34934934934935e-05,
|
|
"loss": 2.5026,
|
|
"step": 5330
|
|
},
|
|
{
|
|
"epoch": 0.5335,
|
|
"grad_norm": 0.7930201888084412,
|
|
"learning_rate": 9.33933933933934e-05,
|
|
"loss": 2.4667,
|
|
"step": 5335
|
|
},
|
|
{
|
|
"epoch": 0.534,
|
|
"grad_norm": 0.7503666877746582,
|
|
"learning_rate": 9.329329329329329e-05,
|
|
"loss": 2.4418,
|
|
"step": 5340
|
|
},
|
|
{
|
|
"epoch": 0.5345,
|
|
"grad_norm": 0.7474675178527832,
|
|
"learning_rate": 9.31931931931932e-05,
|
|
"loss": 2.4644,
|
|
"step": 5345
|
|
},
|
|
{
|
|
"epoch": 0.535,
|
|
"grad_norm": 0.7890115976333618,
|
|
"learning_rate": 9.30930930930931e-05,
|
|
"loss": 2.4742,
|
|
"step": 5350
|
|
},
|
|
{
|
|
"epoch": 0.5355,
|
|
"grad_norm": 0.7615987658500671,
|
|
"learning_rate": 9.2992992992993e-05,
|
|
"loss": 2.4894,
|
|
"step": 5355
|
|
},
|
|
{
|
|
"epoch": 0.536,
|
|
"grad_norm": 0.810775876045227,
|
|
"learning_rate": 9.28928928928929e-05,
|
|
"loss": 2.4453,
|
|
"step": 5360
|
|
},
|
|
{
|
|
"epoch": 0.5365,
|
|
"grad_norm": 0.7598344087600708,
|
|
"learning_rate": 9.279279279279279e-05,
|
|
"loss": 2.4611,
|
|
"step": 5365
|
|
},
|
|
{
|
|
"epoch": 0.537,
|
|
"grad_norm": 0.7713881134986877,
|
|
"learning_rate": 9.26926926926927e-05,
|
|
"loss": 2.4825,
|
|
"step": 5370
|
|
},
|
|
{
|
|
"epoch": 0.5375,
|
|
"grad_norm": 0.8551222681999207,
|
|
"learning_rate": 9.25925925925926e-05,
|
|
"loss": 2.4547,
|
|
"step": 5375
|
|
},
|
|
{
|
|
"epoch": 0.538,
|
|
"grad_norm": 0.8191784620285034,
|
|
"learning_rate": 9.24924924924925e-05,
|
|
"loss": 2.4596,
|
|
"step": 5380
|
|
},
|
|
{
|
|
"epoch": 0.5385,
|
|
"grad_norm": 0.7814502716064453,
|
|
"learning_rate": 9.23923923923924e-05,
|
|
"loss": 2.4548,
|
|
"step": 5385
|
|
},
|
|
{
|
|
"epoch": 0.539,
|
|
"grad_norm": 0.7861321568489075,
|
|
"learning_rate": 9.229229229229229e-05,
|
|
"loss": 2.4771,
|
|
"step": 5390
|
|
},
|
|
{
|
|
"epoch": 0.5395,
|
|
"grad_norm": 0.7862023115158081,
|
|
"learning_rate": 9.219219219219219e-05,
|
|
"loss": 2.4826,
|
|
"step": 5395
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"grad_norm": 0.7637827396392822,
|
|
"learning_rate": 9.20920920920921e-05,
|
|
"loss": 2.4617,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 0.5405,
|
|
"grad_norm": 0.7903456687927246,
|
|
"learning_rate": 9.1991991991992e-05,
|
|
"loss": 2.4544,
|
|
"step": 5405
|
|
},
|
|
{
|
|
"epoch": 0.541,
|
|
"grad_norm": 0.7984066605567932,
|
|
"learning_rate": 9.18918918918919e-05,
|
|
"loss": 2.4591,
|
|
"step": 5410
|
|
},
|
|
{
|
|
"epoch": 0.5415,
|
|
"grad_norm": 0.8397104144096375,
|
|
"learning_rate": 9.179179179179179e-05,
|
|
"loss": 2.4563,
|
|
"step": 5415
|
|
},
|
|
{
|
|
"epoch": 0.542,
|
|
"grad_norm": 0.7787863612174988,
|
|
"learning_rate": 9.169169169169169e-05,
|
|
"loss": 2.4732,
|
|
"step": 5420
|
|
},
|
|
{
|
|
"epoch": 0.5425,
|
|
"grad_norm": 0.7683889269828796,
|
|
"learning_rate": 9.15915915915916e-05,
|
|
"loss": 2.4769,
|
|
"step": 5425
|
|
},
|
|
{
|
|
"epoch": 0.543,
|
|
"grad_norm": 0.7939984202384949,
|
|
"learning_rate": 9.14914914914915e-05,
|
|
"loss": 2.461,
|
|
"step": 5430
|
|
},
|
|
{
|
|
"epoch": 0.5435,
|
|
"grad_norm": 0.7943704724311829,
|
|
"learning_rate": 9.13913913913914e-05,
|
|
"loss": 2.4749,
|
|
"step": 5435
|
|
},
|
|
{
|
|
"epoch": 0.544,
|
|
"grad_norm": 0.7761884927749634,
|
|
"learning_rate": 9.129129129129129e-05,
|
|
"loss": 2.4591,
|
|
"step": 5440
|
|
},
|
|
{
|
|
"epoch": 0.5445,
|
|
"grad_norm": 0.7611567974090576,
|
|
"learning_rate": 9.119119119119119e-05,
|
|
"loss": 2.4588,
|
|
"step": 5445
|
|
},
|
|
{
|
|
"epoch": 0.545,
|
|
"grad_norm": 0.7688775658607483,
|
|
"learning_rate": 9.109109109109109e-05,
|
|
"loss": 2.4652,
|
|
"step": 5450
|
|
},
|
|
{
|
|
"epoch": 0.5455,
|
|
"grad_norm": 0.7897481322288513,
|
|
"learning_rate": 9.0990990990991e-05,
|
|
"loss": 2.4495,
|
|
"step": 5455
|
|
},
|
|
{
|
|
"epoch": 0.546,
|
|
"grad_norm": 0.7705250978469849,
|
|
"learning_rate": 9.08908908908909e-05,
|
|
"loss": 2.4522,
|
|
"step": 5460
|
|
},
|
|
{
|
|
"epoch": 0.5465,
|
|
"grad_norm": 0.825740396976471,
|
|
"learning_rate": 9.079079079079079e-05,
|
|
"loss": 2.4468,
|
|
"step": 5465
|
|
},
|
|
{
|
|
"epoch": 0.547,
|
|
"grad_norm": 0.7923555970191956,
|
|
"learning_rate": 9.069069069069069e-05,
|
|
"loss": 2.4565,
|
|
"step": 5470
|
|
},
|
|
{
|
|
"epoch": 0.5475,
|
|
"grad_norm": 0.7937232255935669,
|
|
"learning_rate": 9.059059059059059e-05,
|
|
"loss": 2.4544,
|
|
"step": 5475
|
|
},
|
|
{
|
|
"epoch": 0.548,
|
|
"grad_norm": 0.8358876705169678,
|
|
"learning_rate": 9.04904904904905e-05,
|
|
"loss": 2.4445,
|
|
"step": 5480
|
|
},
|
|
{
|
|
"epoch": 0.5485,
|
|
"grad_norm": 0.9014988541603088,
|
|
"learning_rate": 9.039039039039039e-05,
|
|
"loss": 2.4469,
|
|
"step": 5485
|
|
},
|
|
{
|
|
"epoch": 0.549,
|
|
"grad_norm": 0.7850622534751892,
|
|
"learning_rate": 9.029029029029029e-05,
|
|
"loss": 2.4463,
|
|
"step": 5490
|
|
},
|
|
{
|
|
"epoch": 0.5495,
|
|
"grad_norm": 0.8793150782585144,
|
|
"learning_rate": 9.019019019019019e-05,
|
|
"loss": 2.4799,
|
|
"step": 5495
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"grad_norm": 0.8214443325996399,
|
|
"learning_rate": 9.009009009009009e-05,
|
|
"loss": 2.4545,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 0.5505,
|
|
"grad_norm": 0.7656763195991516,
|
|
"learning_rate": 8.998998998998999e-05,
|
|
"loss": 2.4495,
|
|
"step": 5505
|
|
},
|
|
{
|
|
"epoch": 0.551,
|
|
"grad_norm": 0.7774412631988525,
|
|
"learning_rate": 8.988988988988989e-05,
|
|
"loss": 2.461,
|
|
"step": 5510
|
|
},
|
|
{
|
|
"epoch": 0.5515,
|
|
"grad_norm": 0.7601509690284729,
|
|
"learning_rate": 8.978978978978979e-05,
|
|
"loss": 2.4365,
|
|
"step": 5515
|
|
},
|
|
{
|
|
"epoch": 0.552,
|
|
"grad_norm": 0.7853091359138489,
|
|
"learning_rate": 8.968968968968969e-05,
|
|
"loss": 2.4625,
|
|
"step": 5520
|
|
},
|
|
{
|
|
"epoch": 0.5525,
|
|
"grad_norm": 0.7637262344360352,
|
|
"learning_rate": 8.958958958958959e-05,
|
|
"loss": 2.4445,
|
|
"step": 5525
|
|
},
|
|
{
|
|
"epoch": 0.553,
|
|
"grad_norm": 0.8261085152626038,
|
|
"learning_rate": 8.948948948948949e-05,
|
|
"loss": 2.4689,
|
|
"step": 5530
|
|
},
|
|
{
|
|
"epoch": 0.5535,
|
|
"grad_norm": 0.7731636762619019,
|
|
"learning_rate": 8.938938938938939e-05,
|
|
"loss": 2.4598,
|
|
"step": 5535
|
|
},
|
|
{
|
|
"epoch": 0.554,
|
|
"grad_norm": 0.7947583794593811,
|
|
"learning_rate": 8.928928928928929e-05,
|
|
"loss": 2.473,
|
|
"step": 5540
|
|
},
|
|
{
|
|
"epoch": 0.5545,
|
|
"grad_norm": 0.7763395309448242,
|
|
"learning_rate": 8.918918918918919e-05,
|
|
"loss": 2.4491,
|
|
"step": 5545
|
|
},
|
|
{
|
|
"epoch": 0.555,
|
|
"grad_norm": 0.8187987804412842,
|
|
"learning_rate": 8.908908908908909e-05,
|
|
"loss": 2.4502,
|
|
"step": 5550
|
|
},
|
|
{
|
|
"epoch": 0.5555,
|
|
"grad_norm": 0.8020109534263611,
|
|
"learning_rate": 8.898898898898899e-05,
|
|
"loss": 2.4639,
|
|
"step": 5555
|
|
},
|
|
{
|
|
"epoch": 0.556,
|
|
"grad_norm": 0.7664561867713928,
|
|
"learning_rate": 8.888888888888889e-05,
|
|
"loss": 2.4249,
|
|
"step": 5560
|
|
},
|
|
{
|
|
"epoch": 0.5565,
|
|
"grad_norm": 0.7538728713989258,
|
|
"learning_rate": 8.87887887887888e-05,
|
|
"loss": 2.4439,
|
|
"step": 5565
|
|
},
|
|
{
|
|
"epoch": 0.557,
|
|
"grad_norm": 0.810167670249939,
|
|
"learning_rate": 8.86886886886887e-05,
|
|
"loss": 2.4489,
|
|
"step": 5570
|
|
},
|
|
{
|
|
"epoch": 0.5575,
|
|
"grad_norm": 0.8242455720901489,
|
|
"learning_rate": 8.85885885885886e-05,
|
|
"loss": 2.4415,
|
|
"step": 5575
|
|
},
|
|
{
|
|
"epoch": 0.558,
|
|
"grad_norm": 0.7702010869979858,
|
|
"learning_rate": 8.84884884884885e-05,
|
|
"loss": 2.4674,
|
|
"step": 5580
|
|
},
|
|
{
|
|
"epoch": 0.5585,
|
|
"grad_norm": 0.7628993391990662,
|
|
"learning_rate": 8.83883883883884e-05,
|
|
"loss": 2.4254,
|
|
"step": 5585
|
|
},
|
|
{
|
|
"epoch": 0.559,
|
|
"grad_norm": 0.7824822664260864,
|
|
"learning_rate": 8.82882882882883e-05,
|
|
"loss": 2.4404,
|
|
"step": 5590
|
|
},
|
|
{
|
|
"epoch": 0.5595,
|
|
"grad_norm": 0.8106468915939331,
|
|
"learning_rate": 8.81881881881882e-05,
|
|
"loss": 2.4572,
|
|
"step": 5595
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"grad_norm": 0.7582213282585144,
|
|
"learning_rate": 8.80880880880881e-05,
|
|
"loss": 2.4542,
|
|
"step": 5600
|
|
},
|
|
{
|
|
"epoch": 0.5605,
|
|
"grad_norm": 0.7767332196235657,
|
|
"learning_rate": 8.7987987987988e-05,
|
|
"loss": 2.4386,
|
|
"step": 5605
|
|
},
|
|
{
|
|
"epoch": 0.561,
|
|
"grad_norm": 0.7707039713859558,
|
|
"learning_rate": 8.788788788788788e-05,
|
|
"loss": 2.4308,
|
|
"step": 5610
|
|
},
|
|
{
|
|
"epoch": 0.5615,
|
|
"grad_norm": 0.7898010611534119,
|
|
"learning_rate": 8.77877877877878e-05,
|
|
"loss": 2.4344,
|
|
"step": 5615
|
|
},
|
|
{
|
|
"epoch": 0.562,
|
|
"grad_norm": 0.8068522810935974,
|
|
"learning_rate": 8.76876876876877e-05,
|
|
"loss": 2.4329,
|
|
"step": 5620
|
|
},
|
|
{
|
|
"epoch": 0.5625,
|
|
"grad_norm": 0.8125194907188416,
|
|
"learning_rate": 8.75875875875876e-05,
|
|
"loss": 2.4276,
|
|
"step": 5625
|
|
},
|
|
{
|
|
"epoch": 0.563,
|
|
"grad_norm": 0.7750654816627502,
|
|
"learning_rate": 8.74874874874875e-05,
|
|
"loss": 2.4474,
|
|
"step": 5630
|
|
},
|
|
{
|
|
"epoch": 0.5635,
|
|
"grad_norm": 0.8232579827308655,
|
|
"learning_rate": 8.738738738738738e-05,
|
|
"loss": 2.4018,
|
|
"step": 5635
|
|
},
|
|
{
|
|
"epoch": 0.564,
|
|
"grad_norm": 0.7836376428604126,
|
|
"learning_rate": 8.72872872872873e-05,
|
|
"loss": 2.4534,
|
|
"step": 5640
|
|
},
|
|
{
|
|
"epoch": 0.5645,
|
|
"grad_norm": 0.7965866327285767,
|
|
"learning_rate": 8.71871871871872e-05,
|
|
"loss": 2.4079,
|
|
"step": 5645
|
|
},
|
|
{
|
|
"epoch": 0.565,
|
|
"grad_norm": 0.8080419898033142,
|
|
"learning_rate": 8.70870870870871e-05,
|
|
"loss": 2.4123,
|
|
"step": 5650
|
|
},
|
|
{
|
|
"epoch": 0.5655,
|
|
"grad_norm": 0.7705451250076294,
|
|
"learning_rate": 8.6986986986987e-05,
|
|
"loss": 2.4468,
|
|
"step": 5655
|
|
},
|
|
{
|
|
"epoch": 0.566,
|
|
"grad_norm": 0.8220493793487549,
|
|
"learning_rate": 8.688688688688688e-05,
|
|
"loss": 2.4524,
|
|
"step": 5660
|
|
},
|
|
{
|
|
"epoch": 0.5665,
|
|
"grad_norm": 0.7655929923057556,
|
|
"learning_rate": 8.678678678678678e-05,
|
|
"loss": 2.4532,
|
|
"step": 5665
|
|
},
|
|
{
|
|
"epoch": 0.567,
|
|
"grad_norm": 0.7849308848381042,
|
|
"learning_rate": 8.66866866866867e-05,
|
|
"loss": 2.4517,
|
|
"step": 5670
|
|
},
|
|
{
|
|
"epoch": 0.5675,
|
|
"grad_norm": 0.8112594485282898,
|
|
"learning_rate": 8.65865865865866e-05,
|
|
"loss": 2.4335,
|
|
"step": 5675
|
|
},
|
|
{
|
|
"epoch": 0.568,
|
|
"grad_norm": 0.7860171794891357,
|
|
"learning_rate": 8.64864864864865e-05,
|
|
"loss": 2.4271,
|
|
"step": 5680
|
|
},
|
|
{
|
|
"epoch": 0.5685,
|
|
"grad_norm": 0.7731909155845642,
|
|
"learning_rate": 8.638638638638638e-05,
|
|
"loss": 2.4513,
|
|
"step": 5685
|
|
},
|
|
{
|
|
"epoch": 0.569,
|
|
"grad_norm": 0.8540036082267761,
|
|
"learning_rate": 8.628628628628628e-05,
|
|
"loss": 2.4464,
|
|
"step": 5690
|
|
},
|
|
{
|
|
"epoch": 0.5695,
|
|
"grad_norm": 0.7871401309967041,
|
|
"learning_rate": 8.61861861861862e-05,
|
|
"loss": 2.4451,
|
|
"step": 5695
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"grad_norm": 0.7708069086074829,
|
|
"learning_rate": 8.60860860860861e-05,
|
|
"loss": 2.4273,
|
|
"step": 5700
|
|
},
|
|
{
|
|
"epoch": 0.5705,
|
|
"grad_norm": 0.7826456427574158,
|
|
"learning_rate": 8.5985985985986e-05,
|
|
"loss": 2.4468,
|
|
"step": 5705
|
|
},
|
|
{
|
|
"epoch": 0.571,
|
|
"grad_norm": 0.7948140501976013,
|
|
"learning_rate": 8.588588588588588e-05,
|
|
"loss": 2.4527,
|
|
"step": 5710
|
|
},
|
|
{
|
|
"epoch": 0.5715,
|
|
"grad_norm": 0.7906607985496521,
|
|
"learning_rate": 8.578578578578578e-05,
|
|
"loss": 2.4312,
|
|
"step": 5715
|
|
},
|
|
{
|
|
"epoch": 0.572,
|
|
"grad_norm": 0.7874037027359009,
|
|
"learning_rate": 8.568568568568568e-05,
|
|
"loss": 2.4424,
|
|
"step": 5720
|
|
},
|
|
{
|
|
"epoch": 0.5725,
|
|
"grad_norm": 0.8326743245124817,
|
|
"learning_rate": 8.55855855855856e-05,
|
|
"loss": 2.4364,
|
|
"step": 5725
|
|
},
|
|
{
|
|
"epoch": 0.573,
|
|
"grad_norm": 0.7984150052070618,
|
|
"learning_rate": 8.54854854854855e-05,
|
|
"loss": 2.4359,
|
|
"step": 5730
|
|
},
|
|
{
|
|
"epoch": 0.5735,
|
|
"grad_norm": 0.8510274291038513,
|
|
"learning_rate": 8.538538538538538e-05,
|
|
"loss": 2.4469,
|
|
"step": 5735
|
|
},
|
|
{
|
|
"epoch": 0.574,
|
|
"grad_norm": 0.8147960305213928,
|
|
"learning_rate": 8.528528528528528e-05,
|
|
"loss": 2.43,
|
|
"step": 5740
|
|
},
|
|
{
|
|
"epoch": 0.5745,
|
|
"grad_norm": 0.8313918709754944,
|
|
"learning_rate": 8.518518518518518e-05,
|
|
"loss": 2.4197,
|
|
"step": 5745
|
|
},
|
|
{
|
|
"epoch": 0.575,
|
|
"grad_norm": 0.778278648853302,
|
|
"learning_rate": 8.50850850850851e-05,
|
|
"loss": 2.4241,
|
|
"step": 5750
|
|
},
|
|
{
|
|
"epoch": 0.5755,
|
|
"grad_norm": 0.8418667316436768,
|
|
"learning_rate": 8.4984984984985e-05,
|
|
"loss": 2.44,
|
|
"step": 5755
|
|
},
|
|
{
|
|
"epoch": 0.576,
|
|
"grad_norm": 0.7860630750656128,
|
|
"learning_rate": 8.488488488488489e-05,
|
|
"loss": 2.4175,
|
|
"step": 5760
|
|
},
|
|
{
|
|
"epoch": 0.5765,
|
|
"grad_norm": 0.781186044216156,
|
|
"learning_rate": 8.478478478478479e-05,
|
|
"loss": 2.4332,
|
|
"step": 5765
|
|
},
|
|
{
|
|
"epoch": 0.577,
|
|
"grad_norm": 0.8223111629486084,
|
|
"learning_rate": 8.468468468468469e-05,
|
|
"loss": 2.4359,
|
|
"step": 5770
|
|
},
|
|
{
|
|
"epoch": 0.5775,
|
|
"grad_norm": 0.7901541590690613,
|
|
"learning_rate": 8.458458458458459e-05,
|
|
"loss": 2.4187,
|
|
"step": 5775
|
|
},
|
|
{
|
|
"epoch": 0.578,
|
|
"grad_norm": 0.8545305728912354,
|
|
"learning_rate": 8.44844844844845e-05,
|
|
"loss": 2.4166,
|
|
"step": 5780
|
|
},
|
|
{
|
|
"epoch": 0.5785,
|
|
"grad_norm": 0.8031194806098938,
|
|
"learning_rate": 8.438438438438439e-05,
|
|
"loss": 2.4197,
|
|
"step": 5785
|
|
},
|
|
{
|
|
"epoch": 0.579,
|
|
"grad_norm": 0.7738260626792908,
|
|
"learning_rate": 8.428428428428429e-05,
|
|
"loss": 2.4568,
|
|
"step": 5790
|
|
},
|
|
{
|
|
"epoch": 0.5795,
|
|
"grad_norm": 0.8410106897354126,
|
|
"learning_rate": 8.418418418418419e-05,
|
|
"loss": 2.4529,
|
|
"step": 5795
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"grad_norm": 0.7939754724502563,
|
|
"learning_rate": 8.408408408408409e-05,
|
|
"loss": 2.4223,
|
|
"step": 5800
|
|
},
|
|
{
|
|
"epoch": 0.5805,
|
|
"grad_norm": 0.7813926935195923,
|
|
"learning_rate": 8.398398398398399e-05,
|
|
"loss": 2.406,
|
|
"step": 5805
|
|
},
|
|
{
|
|
"epoch": 0.581,
|
|
"grad_norm": 0.8259416818618774,
|
|
"learning_rate": 8.388388388388389e-05,
|
|
"loss": 2.4459,
|
|
"step": 5810
|
|
},
|
|
{
|
|
"epoch": 0.5815,
|
|
"grad_norm": 0.7753074169158936,
|
|
"learning_rate": 8.378378378378379e-05,
|
|
"loss": 2.4232,
|
|
"step": 5815
|
|
},
|
|
{
|
|
"epoch": 0.582,
|
|
"grad_norm": 0.7894231677055359,
|
|
"learning_rate": 8.368368368368369e-05,
|
|
"loss": 2.4229,
|
|
"step": 5820
|
|
},
|
|
{
|
|
"epoch": 0.5825,
|
|
"grad_norm": 0.840605616569519,
|
|
"learning_rate": 8.358358358358359e-05,
|
|
"loss": 2.4376,
|
|
"step": 5825
|
|
},
|
|
{
|
|
"epoch": 0.583,
|
|
"grad_norm": 0.8139504194259644,
|
|
"learning_rate": 8.348348348348349e-05,
|
|
"loss": 2.4088,
|
|
"step": 5830
|
|
},
|
|
{
|
|
"epoch": 0.5835,
|
|
"grad_norm": 0.7772433757781982,
|
|
"learning_rate": 8.338338338338339e-05,
|
|
"loss": 2.4242,
|
|
"step": 5835
|
|
},
|
|
{
|
|
"epoch": 0.584,
|
|
"grad_norm": 0.7719807028770447,
|
|
"learning_rate": 8.328328328328329e-05,
|
|
"loss": 2.4608,
|
|
"step": 5840
|
|
},
|
|
{
|
|
"epoch": 0.5845,
|
|
"grad_norm": 0.7588045597076416,
|
|
"learning_rate": 8.318318318318319e-05,
|
|
"loss": 2.4191,
|
|
"step": 5845
|
|
},
|
|
{
|
|
"epoch": 0.585,
|
|
"grad_norm": 0.771289587020874,
|
|
"learning_rate": 8.308308308308309e-05,
|
|
"loss": 2.4441,
|
|
"step": 5850
|
|
},
|
|
{
|
|
"epoch": 0.5855,
|
|
"grad_norm": 0.7876021862030029,
|
|
"learning_rate": 8.298298298298299e-05,
|
|
"loss": 2.4436,
|
|
"step": 5855
|
|
},
|
|
{
|
|
"epoch": 0.586,
|
|
"grad_norm": 0.7994235754013062,
|
|
"learning_rate": 8.288288288288289e-05,
|
|
"loss": 2.419,
|
|
"step": 5860
|
|
},
|
|
{
|
|
"epoch": 0.5865,
|
|
"grad_norm": 0.7812893390655518,
|
|
"learning_rate": 8.278278278278279e-05,
|
|
"loss": 2.4374,
|
|
"step": 5865
|
|
},
|
|
{
|
|
"epoch": 0.587,
|
|
"grad_norm": 0.7883958220481873,
|
|
"learning_rate": 8.268268268268269e-05,
|
|
"loss": 2.4307,
|
|
"step": 5870
|
|
},
|
|
{
|
|
"epoch": 0.5875,
|
|
"grad_norm": 0.782835066318512,
|
|
"learning_rate": 8.258258258258259e-05,
|
|
"loss": 2.4087,
|
|
"step": 5875
|
|
},
|
|
{
|
|
"epoch": 0.588,
|
|
"grad_norm": 0.7668812870979309,
|
|
"learning_rate": 8.248248248248247e-05,
|
|
"loss": 2.4005,
|
|
"step": 5880
|
|
},
|
|
{
|
|
"epoch": 0.5885,
|
|
"grad_norm": 0.7992317080497742,
|
|
"learning_rate": 8.238238238238239e-05,
|
|
"loss": 2.4111,
|
|
"step": 5885
|
|
},
|
|
{
|
|
"epoch": 0.589,
|
|
"grad_norm": 0.7875285744667053,
|
|
"learning_rate": 8.228228228228229e-05,
|
|
"loss": 2.425,
|
|
"step": 5890
|
|
},
|
|
{
|
|
"epoch": 0.5895,
|
|
"grad_norm": 0.7984848022460938,
|
|
"learning_rate": 8.218218218218219e-05,
|
|
"loss": 2.4234,
|
|
"step": 5895
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"grad_norm": 0.810619592666626,
|
|
"learning_rate": 8.208208208208209e-05,
|
|
"loss": 2.4293,
|
|
"step": 5900
|
|
},
|
|
{
|
|
"epoch": 0.5905,
|
|
"grad_norm": 0.7930707335472107,
|
|
"learning_rate": 8.198198198198198e-05,
|
|
"loss": 2.4147,
|
|
"step": 5905
|
|
},
|
|
{
|
|
"epoch": 0.591,
|
|
"grad_norm": 0.7991019487380981,
|
|
"learning_rate": 8.188188188188189e-05,
|
|
"loss": 2.4097,
|
|
"step": 5910
|
|
},
|
|
{
|
|
"epoch": 0.5915,
|
|
"grad_norm": 0.8091179728507996,
|
|
"learning_rate": 8.178178178178179e-05,
|
|
"loss": 2.4195,
|
|
"step": 5915
|
|
},
|
|
{
|
|
"epoch": 0.592,
|
|
"grad_norm": 0.7663683295249939,
|
|
"learning_rate": 8.168168168168169e-05,
|
|
"loss": 2.4095,
|
|
"step": 5920
|
|
},
|
|
{
|
|
"epoch": 0.5925,
|
|
"grad_norm": 0.7600463628768921,
|
|
"learning_rate": 8.158158158158159e-05,
|
|
"loss": 2.4348,
|
|
"step": 5925
|
|
},
|
|
{
|
|
"epoch": 0.593,
|
|
"grad_norm": 0.91126549243927,
|
|
"learning_rate": 8.148148148148148e-05,
|
|
"loss": 2.4321,
|
|
"step": 5930
|
|
},
|
|
{
|
|
"epoch": 0.5935,
|
|
"grad_norm": 0.8510048985481262,
|
|
"learning_rate": 8.138138138138138e-05,
|
|
"loss": 2.4069,
|
|
"step": 5935
|
|
},
|
|
{
|
|
"epoch": 0.594,
|
|
"grad_norm": 0.7488054633140564,
|
|
"learning_rate": 8.128128128128129e-05,
|
|
"loss": 2.4197,
|
|
"step": 5940
|
|
},
|
|
{
|
|
"epoch": 0.5945,
|
|
"grad_norm": 0.8285728096961975,
|
|
"learning_rate": 8.118118118118119e-05,
|
|
"loss": 2.4277,
|
|
"step": 5945
|
|
},
|
|
{
|
|
"epoch": 0.595,
|
|
"grad_norm": 0.8015768527984619,
|
|
"learning_rate": 8.108108108108109e-05,
|
|
"loss": 2.4033,
|
|
"step": 5950
|
|
},
|
|
{
|
|
"epoch": 0.5955,
|
|
"grad_norm": 0.8446034789085388,
|
|
"learning_rate": 8.098098098098098e-05,
|
|
"loss": 2.4207,
|
|
"step": 5955
|
|
},
|
|
{
|
|
"epoch": 0.596,
|
|
"grad_norm": 0.8294364213943481,
|
|
"learning_rate": 8.088088088088088e-05,
|
|
"loss": 2.4484,
|
|
"step": 5960
|
|
},
|
|
{
|
|
"epoch": 0.5965,
|
|
"grad_norm": 0.7883287668228149,
|
|
"learning_rate": 8.078078078078079e-05,
|
|
"loss": 2.4104,
|
|
"step": 5965
|
|
},
|
|
{
|
|
"epoch": 0.597,
|
|
"grad_norm": 0.7896589040756226,
|
|
"learning_rate": 8.068068068068069e-05,
|
|
"loss": 2.425,
|
|
"step": 5970
|
|
},
|
|
{
|
|
"epoch": 0.5975,
|
|
"grad_norm": 0.7847106456756592,
|
|
"learning_rate": 8.058058058058059e-05,
|
|
"loss": 2.4173,
|
|
"step": 5975
|
|
},
|
|
{
|
|
"epoch": 0.598,
|
|
"grad_norm": 0.8176475763320923,
|
|
"learning_rate": 8.048048048048048e-05,
|
|
"loss": 2.4333,
|
|
"step": 5980
|
|
},
|
|
{
|
|
"epoch": 0.5985,
|
|
"grad_norm": 0.8108943104743958,
|
|
"learning_rate": 8.038038038038038e-05,
|
|
"loss": 2.4264,
|
|
"step": 5985
|
|
},
|
|
{
|
|
"epoch": 0.599,
|
|
"grad_norm": 0.8394784331321716,
|
|
"learning_rate": 8.028028028028028e-05,
|
|
"loss": 2.4001,
|
|
"step": 5990
|
|
},
|
|
{
|
|
"epoch": 0.5995,
|
|
"grad_norm": 0.8406071662902832,
|
|
"learning_rate": 8.018018018018019e-05,
|
|
"loss": 2.435,
|
|
"step": 5995
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"grad_norm": 0.9026022553443909,
|
|
"learning_rate": 8.008008008008009e-05,
|
|
"loss": 2.4271,
|
|
"step": 6000
|
|
},
|
|
{
|
|
"epoch": 0.6005,
|
|
"grad_norm": 0.8024097084999084,
|
|
"learning_rate": 7.997997997997998e-05,
|
|
"loss": 2.4278,
|
|
"step": 6005
|
|
},
|
|
{
|
|
"epoch": 0.601,
|
|
"grad_norm": 0.7905908823013306,
|
|
"learning_rate": 7.987987987987988e-05,
|
|
"loss": 2.4215,
|
|
"step": 6010
|
|
},
|
|
{
|
|
"epoch": 0.6015,
|
|
"grad_norm": 0.8006245493888855,
|
|
"learning_rate": 7.977977977977978e-05,
|
|
"loss": 2.4114,
|
|
"step": 6015
|
|
},
|
|
{
|
|
"epoch": 0.602,
|
|
"grad_norm": 0.8655180335044861,
|
|
"learning_rate": 7.967967967967969e-05,
|
|
"loss": 2.4196,
|
|
"step": 6020
|
|
},
|
|
{
|
|
"epoch": 0.6025,
|
|
"grad_norm": 0.8015463352203369,
|
|
"learning_rate": 7.957957957957959e-05,
|
|
"loss": 2.4352,
|
|
"step": 6025
|
|
},
|
|
{
|
|
"epoch": 0.603,
|
|
"grad_norm": 0.7995092272758484,
|
|
"learning_rate": 7.947947947947948e-05,
|
|
"loss": 2.4106,
|
|
"step": 6030
|
|
},
|
|
{
|
|
"epoch": 0.6035,
|
|
"grad_norm": 0.8037542104721069,
|
|
"learning_rate": 7.937937937937938e-05,
|
|
"loss": 2.4104,
|
|
"step": 6035
|
|
},
|
|
{
|
|
"epoch": 0.604,
|
|
"grad_norm": 0.8060451745986938,
|
|
"learning_rate": 7.927927927927928e-05,
|
|
"loss": 2.4313,
|
|
"step": 6040
|
|
},
|
|
{
|
|
"epoch": 0.6045,
|
|
"grad_norm": 0.8142174482345581,
|
|
"learning_rate": 7.917917917917918e-05,
|
|
"loss": 2.4307,
|
|
"step": 6045
|
|
},
|
|
{
|
|
"epoch": 0.605,
|
|
"grad_norm": 0.791358232498169,
|
|
"learning_rate": 7.90790790790791e-05,
|
|
"loss": 2.4212,
|
|
"step": 6050
|
|
},
|
|
{
|
|
"epoch": 0.6055,
|
|
"grad_norm": 0.8433311581611633,
|
|
"learning_rate": 7.897897897897898e-05,
|
|
"loss": 2.3925,
|
|
"step": 6055
|
|
},
|
|
{
|
|
"epoch": 0.606,
|
|
"grad_norm": 0.7595745325088501,
|
|
"learning_rate": 7.887887887887888e-05,
|
|
"loss": 2.3948,
|
|
"step": 6060
|
|
},
|
|
{
|
|
"epoch": 0.6065,
|
|
"grad_norm": 0.7723388075828552,
|
|
"learning_rate": 7.877877877877878e-05,
|
|
"loss": 2.4276,
|
|
"step": 6065
|
|
},
|
|
{
|
|
"epoch": 0.607,
|
|
"grad_norm": 0.8227433562278748,
|
|
"learning_rate": 7.867867867867868e-05,
|
|
"loss": 2.4065,
|
|
"step": 6070
|
|
},
|
|
{
|
|
"epoch": 0.6075,
|
|
"grad_norm": 0.8013408184051514,
|
|
"learning_rate": 7.85785785785786e-05,
|
|
"loss": 2.4273,
|
|
"step": 6075
|
|
},
|
|
{
|
|
"epoch": 0.608,
|
|
"grad_norm": 0.7976078391075134,
|
|
"learning_rate": 7.847847847847848e-05,
|
|
"loss": 2.4027,
|
|
"step": 6080
|
|
},
|
|
{
|
|
"epoch": 0.6085,
|
|
"grad_norm": 0.7952542901039124,
|
|
"learning_rate": 7.837837837837838e-05,
|
|
"loss": 2.3998,
|
|
"step": 6085
|
|
},
|
|
{
|
|
"epoch": 0.609,
|
|
"grad_norm": 0.8024086952209473,
|
|
"learning_rate": 7.827827827827828e-05,
|
|
"loss": 2.4051,
|
|
"step": 6090
|
|
},
|
|
{
|
|
"epoch": 0.6095,
|
|
"grad_norm": 0.9264689683914185,
|
|
"learning_rate": 7.817817817817818e-05,
|
|
"loss": 2.431,
|
|
"step": 6095
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"grad_norm": 0.7977313995361328,
|
|
"learning_rate": 7.807807807807808e-05,
|
|
"loss": 2.3998,
|
|
"step": 6100
|
|
},
|
|
{
|
|
"epoch": 0.6105,
|
|
"grad_norm": 0.8100118041038513,
|
|
"learning_rate": 7.797797797797798e-05,
|
|
"loss": 2.4272,
|
|
"step": 6105
|
|
},
|
|
{
|
|
"epoch": 0.611,
|
|
"grad_norm": 0.7686548829078674,
|
|
"learning_rate": 7.787787787787788e-05,
|
|
"loss": 2.4215,
|
|
"step": 6110
|
|
},
|
|
{
|
|
"epoch": 0.6115,
|
|
"grad_norm": 0.7883079051971436,
|
|
"learning_rate": 7.777777777777778e-05,
|
|
"loss": 2.4235,
|
|
"step": 6115
|
|
},
|
|
{
|
|
"epoch": 0.612,
|
|
"grad_norm": 0.7836454510688782,
|
|
"learning_rate": 7.767767767767768e-05,
|
|
"loss": 2.4161,
|
|
"step": 6120
|
|
},
|
|
{
|
|
"epoch": 0.6125,
|
|
"grad_norm": 0.7758154273033142,
|
|
"learning_rate": 7.757757757757758e-05,
|
|
"loss": 2.384,
|
|
"step": 6125
|
|
},
|
|
{
|
|
"epoch": 0.613,
|
|
"grad_norm": 0.8135970830917358,
|
|
"learning_rate": 7.747747747747748e-05,
|
|
"loss": 2.4148,
|
|
"step": 6130
|
|
},
|
|
{
|
|
"epoch": 0.6135,
|
|
"grad_norm": 0.8397132158279419,
|
|
"learning_rate": 7.737737737737738e-05,
|
|
"loss": 2.3973,
|
|
"step": 6135
|
|
},
|
|
{
|
|
"epoch": 0.614,
|
|
"grad_norm": 0.8081435561180115,
|
|
"learning_rate": 7.727727727727728e-05,
|
|
"loss": 2.4167,
|
|
"step": 6140
|
|
},
|
|
{
|
|
"epoch": 0.6145,
|
|
"grad_norm": 0.7891101837158203,
|
|
"learning_rate": 7.717717717717718e-05,
|
|
"loss": 2.3973,
|
|
"step": 6145
|
|
},
|
|
{
|
|
"epoch": 0.615,
|
|
"grad_norm": 0.793466329574585,
|
|
"learning_rate": 7.707707707707707e-05,
|
|
"loss": 2.4144,
|
|
"step": 6150
|
|
},
|
|
{
|
|
"epoch": 0.6155,
|
|
"grad_norm": 0.8238446712493896,
|
|
"learning_rate": 7.697697697697698e-05,
|
|
"loss": 2.403,
|
|
"step": 6155
|
|
},
|
|
{
|
|
"epoch": 0.616,
|
|
"grad_norm": 0.7987673282623291,
|
|
"learning_rate": 7.687687687687688e-05,
|
|
"loss": 2.4049,
|
|
"step": 6160
|
|
},
|
|
{
|
|
"epoch": 0.6165,
|
|
"grad_norm": 0.8488965630531311,
|
|
"learning_rate": 7.677677677677678e-05,
|
|
"loss": 2.4099,
|
|
"step": 6165
|
|
},
|
|
{
|
|
"epoch": 0.617,
|
|
"grad_norm": 0.7871595025062561,
|
|
"learning_rate": 7.667667667667668e-05,
|
|
"loss": 2.4306,
|
|
"step": 6170
|
|
},
|
|
{
|
|
"epoch": 0.6175,
|
|
"grad_norm": 0.7598607540130615,
|
|
"learning_rate": 7.657657657657657e-05,
|
|
"loss": 2.4022,
|
|
"step": 6175
|
|
},
|
|
{
|
|
"epoch": 0.618,
|
|
"grad_norm": 0.8631352186203003,
|
|
"learning_rate": 7.647647647647648e-05,
|
|
"loss": 2.4056,
|
|
"step": 6180
|
|
},
|
|
{
|
|
"epoch": 0.6185,
|
|
"grad_norm": 0.7995339035987854,
|
|
"learning_rate": 7.637637637637638e-05,
|
|
"loss": 2.4138,
|
|
"step": 6185
|
|
},
|
|
{
|
|
"epoch": 0.619,
|
|
"grad_norm": 0.7987746000289917,
|
|
"learning_rate": 7.627627627627628e-05,
|
|
"loss": 2.4078,
|
|
"step": 6190
|
|
},
|
|
{
|
|
"epoch": 0.6195,
|
|
"grad_norm": 0.7808003425598145,
|
|
"learning_rate": 7.617617617617618e-05,
|
|
"loss": 2.4012,
|
|
"step": 6195
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"grad_norm": 0.8006256222724915,
|
|
"learning_rate": 7.607607607607607e-05,
|
|
"loss": 2.3998,
|
|
"step": 6200
|
|
},
|
|
{
|
|
"epoch": 0.6205,
|
|
"grad_norm": 0.846108615398407,
|
|
"learning_rate": 7.597597597597597e-05,
|
|
"loss": 2.4268,
|
|
"step": 6205
|
|
},
|
|
{
|
|
"epoch": 0.621,
|
|
"grad_norm": 0.7796756029129028,
|
|
"learning_rate": 7.587587587587588e-05,
|
|
"loss": 2.4189,
|
|
"step": 6210
|
|
},
|
|
{
|
|
"epoch": 0.6215,
|
|
"grad_norm": 0.8724086284637451,
|
|
"learning_rate": 7.577577577577578e-05,
|
|
"loss": 2.4175,
|
|
"step": 6215
|
|
},
|
|
{
|
|
"epoch": 0.622,
|
|
"grad_norm": 0.8391619324684143,
|
|
"learning_rate": 7.567567567567568e-05,
|
|
"loss": 2.4017,
|
|
"step": 6220
|
|
},
|
|
{
|
|
"epoch": 0.6225,
|
|
"grad_norm": 0.8112910985946655,
|
|
"learning_rate": 7.557557557557557e-05,
|
|
"loss": 2.4015,
|
|
"step": 6225
|
|
},
|
|
{
|
|
"epoch": 0.623,
|
|
"grad_norm": 0.7913944721221924,
|
|
"learning_rate": 7.547547547547547e-05,
|
|
"loss": 2.4084,
|
|
"step": 6230
|
|
},
|
|
{
|
|
"epoch": 0.6235,
|
|
"grad_norm": 0.765633225440979,
|
|
"learning_rate": 7.537537537537538e-05,
|
|
"loss": 2.406,
|
|
"step": 6235
|
|
},
|
|
{
|
|
"epoch": 0.624,
|
|
"grad_norm": 0.8250996470451355,
|
|
"learning_rate": 7.527527527527528e-05,
|
|
"loss": 2.4235,
|
|
"step": 6240
|
|
},
|
|
{
|
|
"epoch": 0.6245,
|
|
"grad_norm": 0.8208925127983093,
|
|
"learning_rate": 7.517517517517519e-05,
|
|
"loss": 2.4228,
|
|
"step": 6245
|
|
},
|
|
{
|
|
"epoch": 0.625,
|
|
"grad_norm": 0.7971552014350891,
|
|
"learning_rate": 7.507507507507507e-05,
|
|
"loss": 2.4004,
|
|
"step": 6250
|
|
},
|
|
{
|
|
"epoch": 0.6255,
|
|
"grad_norm": 0.7985517978668213,
|
|
"learning_rate": 7.497497497497497e-05,
|
|
"loss": 2.3964,
|
|
"step": 6255
|
|
},
|
|
{
|
|
"epoch": 0.626,
|
|
"grad_norm": 0.8133323788642883,
|
|
"learning_rate": 7.487487487487487e-05,
|
|
"loss": 2.3983,
|
|
"step": 6260
|
|
},
|
|
{
|
|
"epoch": 0.6265,
|
|
"grad_norm": 0.8303823471069336,
|
|
"learning_rate": 7.477477477477479e-05,
|
|
"loss": 2.4052,
|
|
"step": 6265
|
|
},
|
|
{
|
|
"epoch": 0.627,
|
|
"grad_norm": 0.7818722724914551,
|
|
"learning_rate": 7.467467467467469e-05,
|
|
"loss": 2.3879,
|
|
"step": 6270
|
|
},
|
|
{
|
|
"epoch": 0.6275,
|
|
"grad_norm": 0.7834113836288452,
|
|
"learning_rate": 7.457457457457457e-05,
|
|
"loss": 2.3996,
|
|
"step": 6275
|
|
},
|
|
{
|
|
"epoch": 0.628,
|
|
"grad_norm": 0.857519805431366,
|
|
"learning_rate": 7.447447447447447e-05,
|
|
"loss": 2.3838,
|
|
"step": 6280
|
|
},
|
|
{
|
|
"epoch": 0.6285,
|
|
"grad_norm": 0.8000242710113525,
|
|
"learning_rate": 7.437437437437437e-05,
|
|
"loss": 2.4151,
|
|
"step": 6285
|
|
},
|
|
{
|
|
"epoch": 0.629,
|
|
"grad_norm": 0.780372142791748,
|
|
"learning_rate": 7.427427427427429e-05,
|
|
"loss": 2.3838,
|
|
"step": 6290
|
|
},
|
|
{
|
|
"epoch": 0.6295,
|
|
"grad_norm": 0.8650689721107483,
|
|
"learning_rate": 7.417417417417419e-05,
|
|
"loss": 2.4185,
|
|
"step": 6295
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"grad_norm": 0.8233420848846436,
|
|
"learning_rate": 7.407407407407407e-05,
|
|
"loss": 2.4164,
|
|
"step": 6300
|
|
},
|
|
{
|
|
"epoch": 0.6305,
|
|
"grad_norm": 0.7887721061706543,
|
|
"learning_rate": 7.397397397397397e-05,
|
|
"loss": 2.4061,
|
|
"step": 6305
|
|
},
|
|
{
|
|
"epoch": 0.631,
|
|
"grad_norm": 0.7893443703651428,
|
|
"learning_rate": 7.387387387387387e-05,
|
|
"loss": 2.4253,
|
|
"step": 6310
|
|
},
|
|
{
|
|
"epoch": 0.6315,
|
|
"grad_norm": 0.7930344939231873,
|
|
"learning_rate": 7.377377377377377e-05,
|
|
"loss": 2.4065,
|
|
"step": 6315
|
|
},
|
|
{
|
|
"epoch": 0.632,
|
|
"grad_norm": 0.7902653813362122,
|
|
"learning_rate": 7.367367367367369e-05,
|
|
"loss": 2.4017,
|
|
"step": 6320
|
|
},
|
|
{
|
|
"epoch": 0.6325,
|
|
"grad_norm": 0.7905397415161133,
|
|
"learning_rate": 7.357357357357357e-05,
|
|
"loss": 2.4077,
|
|
"step": 6325
|
|
},
|
|
{
|
|
"epoch": 0.633,
|
|
"grad_norm": 0.7757744789123535,
|
|
"learning_rate": 7.347347347347347e-05,
|
|
"loss": 2.3954,
|
|
"step": 6330
|
|
},
|
|
{
|
|
"epoch": 0.6335,
|
|
"grad_norm": 0.7905138731002808,
|
|
"learning_rate": 7.337337337337337e-05,
|
|
"loss": 2.3949,
|
|
"step": 6335
|
|
},
|
|
{
|
|
"epoch": 0.634,
|
|
"grad_norm": 0.7751962542533875,
|
|
"learning_rate": 7.327327327327327e-05,
|
|
"loss": 2.4121,
|
|
"step": 6340
|
|
},
|
|
{
|
|
"epoch": 0.6345,
|
|
"grad_norm": 0.7948694825172424,
|
|
"learning_rate": 7.317317317317319e-05,
|
|
"loss": 2.385,
|
|
"step": 6345
|
|
},
|
|
{
|
|
"epoch": 0.635,
|
|
"grad_norm": 0.8490158915519714,
|
|
"learning_rate": 7.307307307307307e-05,
|
|
"loss": 2.4051,
|
|
"step": 6350
|
|
},
|
|
{
|
|
"epoch": 0.6355,
|
|
"grad_norm": 0.7961251139640808,
|
|
"learning_rate": 7.297297297297297e-05,
|
|
"loss": 2.3849,
|
|
"step": 6355
|
|
},
|
|
{
|
|
"epoch": 0.636,
|
|
"grad_norm": 0.8053735494613647,
|
|
"learning_rate": 7.287287287287287e-05,
|
|
"loss": 2.4177,
|
|
"step": 6360
|
|
},
|
|
{
|
|
"epoch": 0.6365,
|
|
"grad_norm": 0.819907546043396,
|
|
"learning_rate": 7.277277277277277e-05,
|
|
"loss": 2.406,
|
|
"step": 6365
|
|
},
|
|
{
|
|
"epoch": 0.637,
|
|
"grad_norm": 0.7854018211364746,
|
|
"learning_rate": 7.267267267267268e-05,
|
|
"loss": 2.386,
|
|
"step": 6370
|
|
},
|
|
{
|
|
"epoch": 0.6375,
|
|
"grad_norm": 0.7699578404426575,
|
|
"learning_rate": 7.257257257257258e-05,
|
|
"loss": 2.398,
|
|
"step": 6375
|
|
},
|
|
{
|
|
"epoch": 0.638,
|
|
"grad_norm": 0.8161998391151428,
|
|
"learning_rate": 7.247247247247248e-05,
|
|
"loss": 2.4012,
|
|
"step": 6380
|
|
},
|
|
{
|
|
"epoch": 0.6385,
|
|
"grad_norm": 0.8244718313217163,
|
|
"learning_rate": 7.237237237237238e-05,
|
|
"loss": 2.3886,
|
|
"step": 6385
|
|
},
|
|
{
|
|
"epoch": 0.639,
|
|
"grad_norm": 0.7753651142120361,
|
|
"learning_rate": 7.227227227227228e-05,
|
|
"loss": 2.4174,
|
|
"step": 6390
|
|
},
|
|
{
|
|
"epoch": 0.6395,
|
|
"grad_norm": 0.7991912961006165,
|
|
"learning_rate": 7.217217217217218e-05,
|
|
"loss": 2.3901,
|
|
"step": 6395
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 0.8403205871582031,
|
|
"learning_rate": 7.207207207207208e-05,
|
|
"loss": 2.4054,
|
|
"step": 6400
|
|
},
|
|
{
|
|
"epoch": 0.6405,
|
|
"grad_norm": 0.7843438982963562,
|
|
"learning_rate": 7.197197197197198e-05,
|
|
"loss": 2.3861,
|
|
"step": 6405
|
|
},
|
|
{
|
|
"epoch": 0.641,
|
|
"grad_norm": 0.7796940207481384,
|
|
"learning_rate": 7.187187187187188e-05,
|
|
"loss": 2.3997,
|
|
"step": 6410
|
|
},
|
|
{
|
|
"epoch": 0.6415,
|
|
"grad_norm": 0.7885047197341919,
|
|
"learning_rate": 7.177177177177178e-05,
|
|
"loss": 2.3816,
|
|
"step": 6415
|
|
},
|
|
{
|
|
"epoch": 0.642,
|
|
"grad_norm": 0.861638605594635,
|
|
"learning_rate": 7.167167167167166e-05,
|
|
"loss": 2.3842,
|
|
"step": 6420
|
|
},
|
|
{
|
|
"epoch": 0.6425,
|
|
"grad_norm": 0.7908929586410522,
|
|
"learning_rate": 7.157157157157158e-05,
|
|
"loss": 2.3871,
|
|
"step": 6425
|
|
},
|
|
{
|
|
"epoch": 0.643,
|
|
"grad_norm": 0.811220645904541,
|
|
"learning_rate": 7.147147147147148e-05,
|
|
"loss": 2.3836,
|
|
"step": 6430
|
|
},
|
|
{
|
|
"epoch": 0.6435,
|
|
"grad_norm": 0.8271155953407288,
|
|
"learning_rate": 7.137137137137138e-05,
|
|
"loss": 2.421,
|
|
"step": 6435
|
|
},
|
|
{
|
|
"epoch": 0.644,
|
|
"grad_norm": 0.76975017786026,
|
|
"learning_rate": 7.127127127127128e-05,
|
|
"loss": 2.4102,
|
|
"step": 6440
|
|
},
|
|
{
|
|
"epoch": 0.6445,
|
|
"grad_norm": 0.8396703004837036,
|
|
"learning_rate": 7.117117117117116e-05,
|
|
"loss": 2.3815,
|
|
"step": 6445
|
|
},
|
|
{
|
|
"epoch": 0.645,
|
|
"grad_norm": 0.9385436773300171,
|
|
"learning_rate": 7.107107107107108e-05,
|
|
"loss": 2.4079,
|
|
"step": 6450
|
|
},
|
|
{
|
|
"epoch": 0.6455,
|
|
"grad_norm": 0.7908446192741394,
|
|
"learning_rate": 7.097097097097098e-05,
|
|
"loss": 2.4033,
|
|
"step": 6455
|
|
},
|
|
{
|
|
"epoch": 0.646,
|
|
"grad_norm": 0.7973217368125916,
|
|
"learning_rate": 7.087087087087088e-05,
|
|
"loss": 2.3928,
|
|
"step": 6460
|
|
},
|
|
{
|
|
"epoch": 0.6465,
|
|
"grad_norm": 0.789564311504364,
|
|
"learning_rate": 7.077077077077078e-05,
|
|
"loss": 2.3913,
|
|
"step": 6465
|
|
},
|
|
{
|
|
"epoch": 0.647,
|
|
"grad_norm": 0.822909951210022,
|
|
"learning_rate": 7.067067067067066e-05,
|
|
"loss": 2.4108,
|
|
"step": 6470
|
|
},
|
|
{
|
|
"epoch": 0.6475,
|
|
"grad_norm": 0.8042473196983337,
|
|
"learning_rate": 7.057057057057056e-05,
|
|
"loss": 2.3953,
|
|
"step": 6475
|
|
},
|
|
{
|
|
"epoch": 0.648,
|
|
"grad_norm": 0.8470174670219421,
|
|
"learning_rate": 7.047047047047048e-05,
|
|
"loss": 2.3844,
|
|
"step": 6480
|
|
},
|
|
{
|
|
"epoch": 0.6485,
|
|
"grad_norm": 0.8247309923171997,
|
|
"learning_rate": 7.037037037037038e-05,
|
|
"loss": 2.4239,
|
|
"step": 6485
|
|
},
|
|
{
|
|
"epoch": 0.649,
|
|
"grad_norm": 0.8267856240272522,
|
|
"learning_rate": 7.027027027027028e-05,
|
|
"loss": 2.3997,
|
|
"step": 6490
|
|
},
|
|
{
|
|
"epoch": 0.6495,
|
|
"grad_norm": 0.7668282985687256,
|
|
"learning_rate": 7.017017017017016e-05,
|
|
"loss": 2.4032,
|
|
"step": 6495
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"grad_norm": 0.8035463094711304,
|
|
"learning_rate": 7.007007007007007e-05,
|
|
"loss": 2.3905,
|
|
"step": 6500
|
|
},
|
|
{
|
|
"epoch": 0.6505,
|
|
"grad_norm": 0.7903201580047607,
|
|
"learning_rate": 6.996996996996998e-05,
|
|
"loss": 2.3934,
|
|
"step": 6505
|
|
},
|
|
{
|
|
"epoch": 0.651,
|
|
"grad_norm": 0.8082432150840759,
|
|
"learning_rate": 6.986986986986988e-05,
|
|
"loss": 2.3888,
|
|
"step": 6510
|
|
},
|
|
{
|
|
"epoch": 0.6515,
|
|
"grad_norm": 0.7989760637283325,
|
|
"learning_rate": 6.976976976976978e-05,
|
|
"loss": 2.4032,
|
|
"step": 6515
|
|
},
|
|
{
|
|
"epoch": 0.652,
|
|
"grad_norm": 0.7746159434318542,
|
|
"learning_rate": 6.966966966966967e-05,
|
|
"loss": 2.375,
|
|
"step": 6520
|
|
},
|
|
{
|
|
"epoch": 0.6525,
|
|
"grad_norm": 0.7914102673530579,
|
|
"learning_rate": 6.956956956956957e-05,
|
|
"loss": 2.3899,
|
|
"step": 6525
|
|
},
|
|
{
|
|
"epoch": 0.653,
|
|
"grad_norm": 0.7977505922317505,
|
|
"learning_rate": 6.946946946946947e-05,
|
|
"loss": 2.3844,
|
|
"step": 6530
|
|
},
|
|
{
|
|
"epoch": 0.6535,
|
|
"grad_norm": 0.7890195846557617,
|
|
"learning_rate": 6.936936936936938e-05,
|
|
"loss": 2.4019,
|
|
"step": 6535
|
|
},
|
|
{
|
|
"epoch": 0.654,
|
|
"grad_norm": 0.8067911267280579,
|
|
"learning_rate": 6.926926926926928e-05,
|
|
"loss": 2.3693,
|
|
"step": 6540
|
|
},
|
|
{
|
|
"epoch": 0.6545,
|
|
"grad_norm": 0.8240167498588562,
|
|
"learning_rate": 6.916916916916917e-05,
|
|
"loss": 2.392,
|
|
"step": 6545
|
|
},
|
|
{
|
|
"epoch": 0.655,
|
|
"grad_norm": 0.8075880408287048,
|
|
"learning_rate": 6.906906906906907e-05,
|
|
"loss": 2.386,
|
|
"step": 6550
|
|
},
|
|
{
|
|
"epoch": 0.6555,
|
|
"grad_norm": 0.7833337187767029,
|
|
"learning_rate": 6.896896896896897e-05,
|
|
"loss": 2.4056,
|
|
"step": 6555
|
|
},
|
|
{
|
|
"epoch": 0.656,
|
|
"grad_norm": 0.7844520211219788,
|
|
"learning_rate": 6.886886886886888e-05,
|
|
"loss": 2.388,
|
|
"step": 6560
|
|
},
|
|
{
|
|
"epoch": 0.6565,
|
|
"grad_norm": 0.8572016954421997,
|
|
"learning_rate": 6.876876876876878e-05,
|
|
"loss": 2.4083,
|
|
"step": 6565
|
|
},
|
|
{
|
|
"epoch": 0.657,
|
|
"grad_norm": 0.7906028032302856,
|
|
"learning_rate": 6.866866866866867e-05,
|
|
"loss": 2.3917,
|
|
"step": 6570
|
|
},
|
|
{
|
|
"epoch": 0.6575,
|
|
"grad_norm": 0.802037239074707,
|
|
"learning_rate": 6.856856856856857e-05,
|
|
"loss": 2.4229,
|
|
"step": 6575
|
|
},
|
|
{
|
|
"epoch": 0.658,
|
|
"grad_norm": 0.8108525276184082,
|
|
"learning_rate": 6.846846846846847e-05,
|
|
"loss": 2.3852,
|
|
"step": 6580
|
|
},
|
|
{
|
|
"epoch": 0.6585,
|
|
"grad_norm": 0.8359381556510925,
|
|
"learning_rate": 6.836836836836837e-05,
|
|
"loss": 2.4062,
|
|
"step": 6585
|
|
},
|
|
{
|
|
"epoch": 0.659,
|
|
"grad_norm": 0.8545534014701843,
|
|
"learning_rate": 6.826826826826828e-05,
|
|
"loss": 2.3873,
|
|
"step": 6590
|
|
},
|
|
{
|
|
"epoch": 0.6595,
|
|
"grad_norm": 0.8026583194732666,
|
|
"learning_rate": 6.816816816816817e-05,
|
|
"loss": 2.389,
|
|
"step": 6595
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"grad_norm": 0.7658330798149109,
|
|
"learning_rate": 6.806806806806807e-05,
|
|
"loss": 2.3863,
|
|
"step": 6600
|
|
},
|
|
{
|
|
"epoch": 0.6605,
|
|
"grad_norm": 0.7946294546127319,
|
|
"learning_rate": 6.796796796796797e-05,
|
|
"loss": 2.3827,
|
|
"step": 6605
|
|
},
|
|
{
|
|
"epoch": 0.661,
|
|
"grad_norm": 0.8445372581481934,
|
|
"learning_rate": 6.786786786786787e-05,
|
|
"loss": 2.3818,
|
|
"step": 6610
|
|
},
|
|
{
|
|
"epoch": 0.6615,
|
|
"grad_norm": 0.7817666530609131,
|
|
"learning_rate": 6.776776776776778e-05,
|
|
"loss": 2.3779,
|
|
"step": 6615
|
|
},
|
|
{
|
|
"epoch": 0.662,
|
|
"grad_norm": 0.803107500076294,
|
|
"learning_rate": 6.766766766766767e-05,
|
|
"loss": 2.3962,
|
|
"step": 6620
|
|
},
|
|
{
|
|
"epoch": 0.6625,
|
|
"grad_norm": 0.8931746482849121,
|
|
"learning_rate": 6.756756756756757e-05,
|
|
"loss": 2.3628,
|
|
"step": 6625
|
|
},
|
|
{
|
|
"epoch": 0.663,
|
|
"grad_norm": 0.8346350193023682,
|
|
"learning_rate": 6.746746746746747e-05,
|
|
"loss": 2.3931,
|
|
"step": 6630
|
|
},
|
|
{
|
|
"epoch": 0.6635,
|
|
"grad_norm": 0.7959952354431152,
|
|
"learning_rate": 6.736736736736737e-05,
|
|
"loss": 2.3881,
|
|
"step": 6635
|
|
},
|
|
{
|
|
"epoch": 0.664,
|
|
"grad_norm": 0.7849836349487305,
|
|
"learning_rate": 6.726726726726727e-05,
|
|
"loss": 2.366,
|
|
"step": 6640
|
|
},
|
|
{
|
|
"epoch": 0.6645,
|
|
"grad_norm": 0.8243876099586487,
|
|
"learning_rate": 6.716716716716717e-05,
|
|
"loss": 2.4016,
|
|
"step": 6645
|
|
},
|
|
{
|
|
"epoch": 0.665,
|
|
"grad_norm": 0.7896625399589539,
|
|
"learning_rate": 6.706706706706707e-05,
|
|
"loss": 2.4097,
|
|
"step": 6650
|
|
},
|
|
{
|
|
"epoch": 0.6655,
|
|
"grad_norm": 0.7979716658592224,
|
|
"learning_rate": 6.696696696696697e-05,
|
|
"loss": 2.3869,
|
|
"step": 6655
|
|
},
|
|
{
|
|
"epoch": 0.666,
|
|
"grad_norm": 0.8322969079017639,
|
|
"learning_rate": 6.686686686686687e-05,
|
|
"loss": 2.3922,
|
|
"step": 6660
|
|
},
|
|
{
|
|
"epoch": 0.6665,
|
|
"grad_norm": 0.791321337223053,
|
|
"learning_rate": 6.676676676676677e-05,
|
|
"loss": 2.3664,
|
|
"step": 6665
|
|
},
|
|
{
|
|
"epoch": 0.667,
|
|
"grad_norm": 0.7899481654167175,
|
|
"learning_rate": 6.666666666666667e-05,
|
|
"loss": 2.3971,
|
|
"step": 6670
|
|
},
|
|
{
|
|
"epoch": 0.6675,
|
|
"grad_norm": 0.8165993690490723,
|
|
"learning_rate": 6.656656656656657e-05,
|
|
"loss": 2.3891,
|
|
"step": 6675
|
|
},
|
|
{
|
|
"epoch": 0.668,
|
|
"grad_norm": 0.8781505227088928,
|
|
"learning_rate": 6.646646646646647e-05,
|
|
"loss": 2.3837,
|
|
"step": 6680
|
|
},
|
|
{
|
|
"epoch": 0.6685,
|
|
"grad_norm": 0.8235121965408325,
|
|
"learning_rate": 6.636636636636637e-05,
|
|
"loss": 2.3965,
|
|
"step": 6685
|
|
},
|
|
{
|
|
"epoch": 0.669,
|
|
"grad_norm": 0.8322201371192932,
|
|
"learning_rate": 6.626626626626627e-05,
|
|
"loss": 2.3918,
|
|
"step": 6690
|
|
},
|
|
{
|
|
"epoch": 0.6695,
|
|
"grad_norm": 0.7922030091285706,
|
|
"learning_rate": 6.616616616616617e-05,
|
|
"loss": 2.3822,
|
|
"step": 6695
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"grad_norm": 0.8057646155357361,
|
|
"learning_rate": 6.606606606606607e-05,
|
|
"loss": 2.4046,
|
|
"step": 6700
|
|
},
|
|
{
|
|
"epoch": 0.6705,
|
|
"grad_norm": 0.7984603643417358,
|
|
"learning_rate": 6.596596596596597e-05,
|
|
"loss": 2.3881,
|
|
"step": 6705
|
|
},
|
|
{
|
|
"epoch": 0.671,
|
|
"grad_norm": 0.8076169490814209,
|
|
"learning_rate": 6.586586586586587e-05,
|
|
"loss": 2.394,
|
|
"step": 6710
|
|
},
|
|
{
|
|
"epoch": 0.6715,
|
|
"grad_norm": 0.8120995163917542,
|
|
"learning_rate": 6.576576576576577e-05,
|
|
"loss": 2.3725,
|
|
"step": 6715
|
|
},
|
|
{
|
|
"epoch": 0.672,
|
|
"grad_norm": 0.8239960670471191,
|
|
"learning_rate": 6.566566566566567e-05,
|
|
"loss": 2.3895,
|
|
"step": 6720
|
|
},
|
|
{
|
|
"epoch": 0.6725,
|
|
"grad_norm": 0.8266900777816772,
|
|
"learning_rate": 6.556556556556557e-05,
|
|
"loss": 2.3901,
|
|
"step": 6725
|
|
},
|
|
{
|
|
"epoch": 0.673,
|
|
"grad_norm": 0.8476052284240723,
|
|
"learning_rate": 6.546546546546547e-05,
|
|
"loss": 2.3951,
|
|
"step": 6730
|
|
},
|
|
{
|
|
"epoch": 0.6735,
|
|
"grad_norm": 0.8398421406745911,
|
|
"learning_rate": 6.536536536536537e-05,
|
|
"loss": 2.3687,
|
|
"step": 6735
|
|
},
|
|
{
|
|
"epoch": 0.674,
|
|
"grad_norm": 0.8095033764839172,
|
|
"learning_rate": 6.526526526526526e-05,
|
|
"loss": 2.3898,
|
|
"step": 6740
|
|
},
|
|
{
|
|
"epoch": 0.6745,
|
|
"grad_norm": 0.8121087551116943,
|
|
"learning_rate": 6.516516516516516e-05,
|
|
"loss": 2.3846,
|
|
"step": 6745
|
|
},
|
|
{
|
|
"epoch": 0.675,
|
|
"grad_norm": 0.8165614604949951,
|
|
"learning_rate": 6.506506506506507e-05,
|
|
"loss": 2.3672,
|
|
"step": 6750
|
|
},
|
|
{
|
|
"epoch": 0.6755,
|
|
"grad_norm": 0.7994385361671448,
|
|
"learning_rate": 6.496496496496497e-05,
|
|
"loss": 2.3918,
|
|
"step": 6755
|
|
},
|
|
{
|
|
"epoch": 0.676,
|
|
"grad_norm": 0.7793731093406677,
|
|
"learning_rate": 6.486486486486487e-05,
|
|
"loss": 2.3742,
|
|
"step": 6760
|
|
},
|
|
{
|
|
"epoch": 0.6765,
|
|
"grad_norm": 0.8018571734428406,
|
|
"learning_rate": 6.476476476476476e-05,
|
|
"loss": 2.3878,
|
|
"step": 6765
|
|
},
|
|
{
|
|
"epoch": 0.677,
|
|
"grad_norm": 0.780410647392273,
|
|
"learning_rate": 6.466466466466466e-05,
|
|
"loss": 2.3851,
|
|
"step": 6770
|
|
},
|
|
{
|
|
"epoch": 0.6775,
|
|
"grad_norm": 0.8321567177772522,
|
|
"learning_rate": 6.456456456456457e-05,
|
|
"loss": 2.3919,
|
|
"step": 6775
|
|
},
|
|
{
|
|
"epoch": 0.678,
|
|
"grad_norm": 0.8763487935066223,
|
|
"learning_rate": 6.446446446446447e-05,
|
|
"loss": 2.3742,
|
|
"step": 6780
|
|
},
|
|
{
|
|
"epoch": 0.6785,
|
|
"grad_norm": 0.8359047174453735,
|
|
"learning_rate": 6.436436436436437e-05,
|
|
"loss": 2.3855,
|
|
"step": 6785
|
|
},
|
|
{
|
|
"epoch": 0.679,
|
|
"grad_norm": 0.9007706046104431,
|
|
"learning_rate": 6.426426426426426e-05,
|
|
"loss": 2.3771,
|
|
"step": 6790
|
|
},
|
|
{
|
|
"epoch": 0.6795,
|
|
"grad_norm": 0.9482093453407288,
|
|
"learning_rate": 6.416416416416416e-05,
|
|
"loss": 2.3943,
|
|
"step": 6795
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"grad_norm": 0.8471071124076843,
|
|
"learning_rate": 6.406406406406406e-05,
|
|
"loss": 2.4029,
|
|
"step": 6800
|
|
},
|
|
{
|
|
"epoch": 0.6805,
|
|
"grad_norm": 0.7937386631965637,
|
|
"learning_rate": 6.396396396396397e-05,
|
|
"loss": 2.3944,
|
|
"step": 6805
|
|
},
|
|
{
|
|
"epoch": 0.681,
|
|
"grad_norm": 0.8067902326583862,
|
|
"learning_rate": 6.386386386386387e-05,
|
|
"loss": 2.3953,
|
|
"step": 6810
|
|
},
|
|
{
|
|
"epoch": 0.6815,
|
|
"grad_norm": 0.7922859787940979,
|
|
"learning_rate": 6.376376376376376e-05,
|
|
"loss": 2.391,
|
|
"step": 6815
|
|
},
|
|
{
|
|
"epoch": 0.682,
|
|
"grad_norm": 0.8054628968238831,
|
|
"learning_rate": 6.366366366366366e-05,
|
|
"loss": 2.3825,
|
|
"step": 6820
|
|
},
|
|
{
|
|
"epoch": 0.6825,
|
|
"grad_norm": 0.8414512872695923,
|
|
"learning_rate": 6.356356356356356e-05,
|
|
"loss": 2.3758,
|
|
"step": 6825
|
|
},
|
|
{
|
|
"epoch": 0.683,
|
|
"grad_norm": 0.7730977535247803,
|
|
"learning_rate": 6.346346346346347e-05,
|
|
"loss": 2.3916,
|
|
"step": 6830
|
|
},
|
|
{
|
|
"epoch": 0.6835,
|
|
"grad_norm": 0.8458556532859802,
|
|
"learning_rate": 6.336336336336337e-05,
|
|
"loss": 2.3863,
|
|
"step": 6835
|
|
},
|
|
{
|
|
"epoch": 0.684,
|
|
"grad_norm": 0.7820555567741394,
|
|
"learning_rate": 6.326326326326326e-05,
|
|
"loss": 2.3712,
|
|
"step": 6840
|
|
},
|
|
{
|
|
"epoch": 0.6845,
|
|
"grad_norm": 0.7969538569450378,
|
|
"learning_rate": 6.316316316316316e-05,
|
|
"loss": 2.3888,
|
|
"step": 6845
|
|
},
|
|
{
|
|
"epoch": 0.685,
|
|
"grad_norm": 0.862758994102478,
|
|
"learning_rate": 6.306306306306306e-05,
|
|
"loss": 2.3901,
|
|
"step": 6850
|
|
},
|
|
{
|
|
"epoch": 0.6855,
|
|
"grad_norm": 0.8746228814125061,
|
|
"learning_rate": 6.296296296296296e-05,
|
|
"loss": 2.3993,
|
|
"step": 6855
|
|
},
|
|
{
|
|
"epoch": 0.686,
|
|
"grad_norm": 0.8406643271446228,
|
|
"learning_rate": 6.286286286286288e-05,
|
|
"loss": 2.3831,
|
|
"step": 6860
|
|
},
|
|
{
|
|
"epoch": 0.6865,
|
|
"grad_norm": 0.8201842308044434,
|
|
"learning_rate": 6.276276276276276e-05,
|
|
"loss": 2.382,
|
|
"step": 6865
|
|
},
|
|
{
|
|
"epoch": 0.687,
|
|
"grad_norm": 0.8067729473114014,
|
|
"learning_rate": 6.266266266266266e-05,
|
|
"loss": 2.3728,
|
|
"step": 6870
|
|
},
|
|
{
|
|
"epoch": 0.6875,
|
|
"grad_norm": 0.8983192443847656,
|
|
"learning_rate": 6.256256256256256e-05,
|
|
"loss": 2.3881,
|
|
"step": 6875
|
|
},
|
|
{
|
|
"epoch": 0.688,
|
|
"grad_norm": 0.9659560918807983,
|
|
"learning_rate": 6.246246246246246e-05,
|
|
"loss": 2.389,
|
|
"step": 6880
|
|
},
|
|
{
|
|
"epoch": 0.6885,
|
|
"grad_norm": 0.8363260626792908,
|
|
"learning_rate": 6.236236236236238e-05,
|
|
"loss": 2.3664,
|
|
"step": 6885
|
|
},
|
|
{
|
|
"epoch": 0.689,
|
|
"grad_norm": 0.8459045886993408,
|
|
"learning_rate": 6.226226226226226e-05,
|
|
"loss": 2.3857,
|
|
"step": 6890
|
|
},
|
|
{
|
|
"epoch": 0.6895,
|
|
"grad_norm": 0.8440020680427551,
|
|
"learning_rate": 6.216216216216216e-05,
|
|
"loss": 2.3642,
|
|
"step": 6895
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"grad_norm": 0.8255513906478882,
|
|
"learning_rate": 6.206206206206206e-05,
|
|
"loss": 2.3844,
|
|
"step": 6900
|
|
},
|
|
{
|
|
"epoch": 0.6905,
|
|
"grad_norm": 0.7914440631866455,
|
|
"learning_rate": 6.196196196196196e-05,
|
|
"loss": 2.3754,
|
|
"step": 6905
|
|
},
|
|
{
|
|
"epoch": 0.691,
|
|
"grad_norm": 0.832038164138794,
|
|
"learning_rate": 6.186186186186186e-05,
|
|
"loss": 2.3786,
|
|
"step": 6910
|
|
},
|
|
{
|
|
"epoch": 0.6915,
|
|
"grad_norm": 0.8137945532798767,
|
|
"learning_rate": 6.176176176176176e-05,
|
|
"loss": 2.3672,
|
|
"step": 6915
|
|
},
|
|
{
|
|
"epoch": 0.692,
|
|
"grad_norm": 0.7861709594726562,
|
|
"learning_rate": 6.166166166166166e-05,
|
|
"loss": 2.3816,
|
|
"step": 6920
|
|
},
|
|
{
|
|
"epoch": 0.6925,
|
|
"grad_norm": 0.8148735165596008,
|
|
"learning_rate": 6.156156156156156e-05,
|
|
"loss": 2.4103,
|
|
"step": 6925
|
|
},
|
|
{
|
|
"epoch": 0.693,
|
|
"grad_norm": 0.8730798959732056,
|
|
"learning_rate": 6.146146146146146e-05,
|
|
"loss": 2.3654,
|
|
"step": 6930
|
|
},
|
|
{
|
|
"epoch": 0.6935,
|
|
"grad_norm": 0.848082959651947,
|
|
"learning_rate": 6.136136136136136e-05,
|
|
"loss": 2.3925,
|
|
"step": 6935
|
|
},
|
|
{
|
|
"epoch": 0.694,
|
|
"grad_norm": 0.7952349185943604,
|
|
"learning_rate": 6.126126126126126e-05,
|
|
"loss": 2.3652,
|
|
"step": 6940
|
|
},
|
|
{
|
|
"epoch": 0.6945,
|
|
"grad_norm": 0.8390651345252991,
|
|
"learning_rate": 6.116116116116116e-05,
|
|
"loss": 2.3919,
|
|
"step": 6945
|
|
},
|
|
{
|
|
"epoch": 0.695,
|
|
"grad_norm": 0.8806639909744263,
|
|
"learning_rate": 6.106106106106106e-05,
|
|
"loss": 2.3805,
|
|
"step": 6950
|
|
},
|
|
{
|
|
"epoch": 0.6955,
|
|
"grad_norm": 0.7894412279129028,
|
|
"learning_rate": 6.0960960960960964e-05,
|
|
"loss": 2.4064,
|
|
"step": 6955
|
|
},
|
|
{
|
|
"epoch": 0.696,
|
|
"grad_norm": 0.7991839051246643,
|
|
"learning_rate": 6.086086086086087e-05,
|
|
"loss": 2.376,
|
|
"step": 6960
|
|
},
|
|
{
|
|
"epoch": 0.6965,
|
|
"grad_norm": 0.8137059807777405,
|
|
"learning_rate": 6.076076076076076e-05,
|
|
"loss": 2.3747,
|
|
"step": 6965
|
|
},
|
|
{
|
|
"epoch": 0.697,
|
|
"grad_norm": 0.8530757427215576,
|
|
"learning_rate": 6.0660660660660665e-05,
|
|
"loss": 2.3922,
|
|
"step": 6970
|
|
},
|
|
{
|
|
"epoch": 0.6975,
|
|
"grad_norm": 0.8087310791015625,
|
|
"learning_rate": 6.0560560560560565e-05,
|
|
"loss": 2.3654,
|
|
"step": 6975
|
|
},
|
|
{
|
|
"epoch": 0.698,
|
|
"grad_norm": 0.8027787208557129,
|
|
"learning_rate": 6.0460460460460465e-05,
|
|
"loss": 2.3812,
|
|
"step": 6980
|
|
},
|
|
{
|
|
"epoch": 0.6985,
|
|
"grad_norm": 0.8009427189826965,
|
|
"learning_rate": 6.0360360360360365e-05,
|
|
"loss": 2.3796,
|
|
"step": 6985
|
|
},
|
|
{
|
|
"epoch": 0.699,
|
|
"grad_norm": 0.8391687870025635,
|
|
"learning_rate": 6.026026026026026e-05,
|
|
"loss": 2.3901,
|
|
"step": 6990
|
|
},
|
|
{
|
|
"epoch": 0.6995,
|
|
"grad_norm": 0.783051609992981,
|
|
"learning_rate": 6.016016016016016e-05,
|
|
"loss": 2.351,
|
|
"step": 6995
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"grad_norm": 0.8066600561141968,
|
|
"learning_rate": 6.0060060060060066e-05,
|
|
"loss": 2.3802,
|
|
"step": 7000
|
|
},
|
|
{
|
|
"epoch": 0.7005,
|
|
"grad_norm": 0.8261300921440125,
|
|
"learning_rate": 5.9959959959959966e-05,
|
|
"loss": 2.3914,
|
|
"step": 7005
|
|
},
|
|
{
|
|
"epoch": 0.701,
|
|
"grad_norm": 0.7957533001899719,
|
|
"learning_rate": 5.9859859859859866e-05,
|
|
"loss": 2.3762,
|
|
"step": 7010
|
|
},
|
|
{
|
|
"epoch": 0.7015,
|
|
"grad_norm": 0.7757399678230286,
|
|
"learning_rate": 5.975975975975976e-05,
|
|
"loss": 2.3712,
|
|
"step": 7015
|
|
},
|
|
{
|
|
"epoch": 0.702,
|
|
"grad_norm": 0.8133918046951294,
|
|
"learning_rate": 5.965965965965966e-05,
|
|
"loss": 2.3628,
|
|
"step": 7020
|
|
},
|
|
{
|
|
"epoch": 0.7025,
|
|
"grad_norm": 0.7873533368110657,
|
|
"learning_rate": 5.9559559559559566e-05,
|
|
"loss": 2.3541,
|
|
"step": 7025
|
|
},
|
|
{
|
|
"epoch": 0.703,
|
|
"grad_norm": 0.8112744092941284,
|
|
"learning_rate": 5.9459459459459466e-05,
|
|
"loss": 2.3609,
|
|
"step": 7030
|
|
},
|
|
{
|
|
"epoch": 0.7035,
|
|
"grad_norm": 0.8749020099639893,
|
|
"learning_rate": 5.935935935935936e-05,
|
|
"loss": 2.3789,
|
|
"step": 7035
|
|
},
|
|
{
|
|
"epoch": 0.704,
|
|
"grad_norm": 0.8450562357902527,
|
|
"learning_rate": 5.925925925925926e-05,
|
|
"loss": 2.3759,
|
|
"step": 7040
|
|
},
|
|
{
|
|
"epoch": 0.7045,
|
|
"grad_norm": 0.8273308277130127,
|
|
"learning_rate": 5.915915915915916e-05,
|
|
"loss": 2.3894,
|
|
"step": 7045
|
|
},
|
|
{
|
|
"epoch": 0.705,
|
|
"grad_norm": 0.8608307838439941,
|
|
"learning_rate": 5.905905905905906e-05,
|
|
"loss": 2.3819,
|
|
"step": 7050
|
|
},
|
|
{
|
|
"epoch": 0.7055,
|
|
"grad_norm": 0.81144779920578,
|
|
"learning_rate": 5.895895895895897e-05,
|
|
"loss": 2.3718,
|
|
"step": 7055
|
|
},
|
|
{
|
|
"epoch": 0.706,
|
|
"grad_norm": 0.8147798180580139,
|
|
"learning_rate": 5.8858858858858854e-05,
|
|
"loss": 2.3731,
|
|
"step": 7060
|
|
},
|
|
{
|
|
"epoch": 0.7065,
|
|
"grad_norm": 0.8627402186393738,
|
|
"learning_rate": 5.875875875875876e-05,
|
|
"loss": 2.3543,
|
|
"step": 7065
|
|
},
|
|
{
|
|
"epoch": 0.707,
|
|
"grad_norm": 0.7978000044822693,
|
|
"learning_rate": 5.865865865865866e-05,
|
|
"loss": 2.3806,
|
|
"step": 7070
|
|
},
|
|
{
|
|
"epoch": 0.7075,
|
|
"grad_norm": 0.8095467686653137,
|
|
"learning_rate": 5.855855855855856e-05,
|
|
"loss": 2.3886,
|
|
"step": 7075
|
|
},
|
|
{
|
|
"epoch": 0.708,
|
|
"grad_norm": 0.7899516820907593,
|
|
"learning_rate": 5.845845845845847e-05,
|
|
"loss": 2.353,
|
|
"step": 7080
|
|
},
|
|
{
|
|
"epoch": 0.7085,
|
|
"grad_norm": 0.7910207509994507,
|
|
"learning_rate": 5.8358358358358354e-05,
|
|
"loss": 2.376,
|
|
"step": 7085
|
|
},
|
|
{
|
|
"epoch": 0.709,
|
|
"grad_norm": 0.7842855453491211,
|
|
"learning_rate": 5.825825825825826e-05,
|
|
"loss": 2.3705,
|
|
"step": 7090
|
|
},
|
|
{
|
|
"epoch": 0.7095,
|
|
"grad_norm": 0.7983612418174744,
|
|
"learning_rate": 5.815815815815816e-05,
|
|
"loss": 2.3711,
|
|
"step": 7095
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"grad_norm": 0.8135548830032349,
|
|
"learning_rate": 5.805805805805806e-05,
|
|
"loss": 2.3915,
|
|
"step": 7100
|
|
},
|
|
{
|
|
"epoch": 0.7105,
|
|
"grad_norm": 0.824687659740448,
|
|
"learning_rate": 5.795795795795796e-05,
|
|
"loss": 2.3634,
|
|
"step": 7105
|
|
},
|
|
{
|
|
"epoch": 0.711,
|
|
"grad_norm": 0.8016681671142578,
|
|
"learning_rate": 5.7857857857857855e-05,
|
|
"loss": 2.3656,
|
|
"step": 7110
|
|
},
|
|
{
|
|
"epoch": 0.7115,
|
|
"grad_norm": 0.8466838002204895,
|
|
"learning_rate": 5.7757757757757755e-05,
|
|
"loss": 2.3601,
|
|
"step": 7115
|
|
},
|
|
{
|
|
"epoch": 0.712,
|
|
"grad_norm": 0.8030304908752441,
|
|
"learning_rate": 5.765765765765766e-05,
|
|
"loss": 2.369,
|
|
"step": 7120
|
|
},
|
|
{
|
|
"epoch": 0.7125,
|
|
"grad_norm": 0.7801939249038696,
|
|
"learning_rate": 5.755755755755756e-05,
|
|
"loss": 2.3457,
|
|
"step": 7125
|
|
},
|
|
{
|
|
"epoch": 0.713,
|
|
"grad_norm": 0.8262499570846558,
|
|
"learning_rate": 5.745745745745746e-05,
|
|
"loss": 2.3802,
|
|
"step": 7130
|
|
},
|
|
{
|
|
"epoch": 0.7135,
|
|
"grad_norm": 0.8150855302810669,
|
|
"learning_rate": 5.7357357357357356e-05,
|
|
"loss": 2.3639,
|
|
"step": 7135
|
|
},
|
|
{
|
|
"epoch": 0.714,
|
|
"grad_norm": 0.8202646970748901,
|
|
"learning_rate": 5.7257257257257256e-05,
|
|
"loss": 2.3669,
|
|
"step": 7140
|
|
},
|
|
{
|
|
"epoch": 0.7145,
|
|
"grad_norm": 0.8026723265647888,
|
|
"learning_rate": 5.715715715715716e-05,
|
|
"loss": 2.3811,
|
|
"step": 7145
|
|
},
|
|
{
|
|
"epoch": 0.715,
|
|
"grad_norm": 0.8347824215888977,
|
|
"learning_rate": 5.705705705705706e-05,
|
|
"loss": 2.4039,
|
|
"step": 7150
|
|
},
|
|
{
|
|
"epoch": 0.7155,
|
|
"grad_norm": 0.8026264309883118,
|
|
"learning_rate": 5.695695695695696e-05,
|
|
"loss": 2.3531,
|
|
"step": 7155
|
|
},
|
|
{
|
|
"epoch": 0.716,
|
|
"grad_norm": 0.8455007076263428,
|
|
"learning_rate": 5.6856856856856856e-05,
|
|
"loss": 2.3756,
|
|
"step": 7160
|
|
},
|
|
{
|
|
"epoch": 0.7165,
|
|
"grad_norm": 0.771700918674469,
|
|
"learning_rate": 5.6756756756756757e-05,
|
|
"loss": 2.3862,
|
|
"step": 7165
|
|
},
|
|
{
|
|
"epoch": 0.717,
|
|
"grad_norm": 0.8043254017829895,
|
|
"learning_rate": 5.665665665665666e-05,
|
|
"loss": 2.3787,
|
|
"step": 7170
|
|
},
|
|
{
|
|
"epoch": 0.7175,
|
|
"grad_norm": 0.7972248792648315,
|
|
"learning_rate": 5.6556556556556564e-05,
|
|
"loss": 2.3447,
|
|
"step": 7175
|
|
},
|
|
{
|
|
"epoch": 0.718,
|
|
"grad_norm": 0.8100689053535461,
|
|
"learning_rate": 5.6456456456456464e-05,
|
|
"loss": 2.3429,
|
|
"step": 7180
|
|
},
|
|
{
|
|
"epoch": 0.7185,
|
|
"grad_norm": 0.8367940187454224,
|
|
"learning_rate": 5.635635635635636e-05,
|
|
"loss": 2.3852,
|
|
"step": 7185
|
|
},
|
|
{
|
|
"epoch": 0.719,
|
|
"grad_norm": 0.8084604144096375,
|
|
"learning_rate": 5.625625625625626e-05,
|
|
"loss": 2.3581,
|
|
"step": 7190
|
|
},
|
|
{
|
|
"epoch": 0.7195,
|
|
"grad_norm": 0.8251677751541138,
|
|
"learning_rate": 5.615615615615616e-05,
|
|
"loss": 2.3576,
|
|
"step": 7195
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"grad_norm": 0.8056468963623047,
|
|
"learning_rate": 5.6056056056056064e-05,
|
|
"loss": 2.3879,
|
|
"step": 7200
|
|
},
|
|
{
|
|
"epoch": 0.7205,
|
|
"grad_norm": 0.7862322926521301,
|
|
"learning_rate": 5.5955955955955964e-05,
|
|
"loss": 2.3749,
|
|
"step": 7205
|
|
},
|
|
{
|
|
"epoch": 0.721,
|
|
"grad_norm": 0.8112216591835022,
|
|
"learning_rate": 5.585585585585585e-05,
|
|
"loss": 2.3583,
|
|
"step": 7210
|
|
},
|
|
{
|
|
"epoch": 0.7215,
|
|
"grad_norm": 0.8866540789604187,
|
|
"learning_rate": 5.575575575575576e-05,
|
|
"loss": 2.3877,
|
|
"step": 7215
|
|
},
|
|
{
|
|
"epoch": 0.722,
|
|
"grad_norm": 0.8252007365226746,
|
|
"learning_rate": 5.565565565565566e-05,
|
|
"loss": 2.3675,
|
|
"step": 7220
|
|
},
|
|
{
|
|
"epoch": 0.7225,
|
|
"grad_norm": 0.8246432542800903,
|
|
"learning_rate": 5.555555555555556e-05,
|
|
"loss": 2.3786,
|
|
"step": 7225
|
|
},
|
|
{
|
|
"epoch": 0.723,
|
|
"grad_norm": 0.843048632144928,
|
|
"learning_rate": 5.5455455455455465e-05,
|
|
"loss": 2.3612,
|
|
"step": 7230
|
|
},
|
|
{
|
|
"epoch": 0.7235,
|
|
"grad_norm": 0.7862938046455383,
|
|
"learning_rate": 5.535535535535535e-05,
|
|
"loss": 2.3706,
|
|
"step": 7235
|
|
},
|
|
{
|
|
"epoch": 0.724,
|
|
"grad_norm": 0.8511613607406616,
|
|
"learning_rate": 5.525525525525526e-05,
|
|
"loss": 2.3774,
|
|
"step": 7240
|
|
},
|
|
{
|
|
"epoch": 0.7245,
|
|
"grad_norm": 0.858098030090332,
|
|
"learning_rate": 5.515515515515516e-05,
|
|
"loss": 2.3644,
|
|
"step": 7245
|
|
},
|
|
{
|
|
"epoch": 0.725,
|
|
"grad_norm": 0.8149218559265137,
|
|
"learning_rate": 5.505505505505506e-05,
|
|
"loss": 2.3957,
|
|
"step": 7250
|
|
},
|
|
{
|
|
"epoch": 0.7255,
|
|
"grad_norm": 0.8075857758522034,
|
|
"learning_rate": 5.4954954954954966e-05,
|
|
"loss": 2.3941,
|
|
"step": 7255
|
|
},
|
|
{
|
|
"epoch": 0.726,
|
|
"grad_norm": 0.9962955117225647,
|
|
"learning_rate": 5.485485485485485e-05,
|
|
"loss": 2.3687,
|
|
"step": 7260
|
|
},
|
|
{
|
|
"epoch": 0.7265,
|
|
"grad_norm": 0.7864262461662292,
|
|
"learning_rate": 5.475475475475475e-05,
|
|
"loss": 2.3617,
|
|
"step": 7265
|
|
},
|
|
{
|
|
"epoch": 0.727,
|
|
"grad_norm": 0.8471370339393616,
|
|
"learning_rate": 5.465465465465466e-05,
|
|
"loss": 2.3765,
|
|
"step": 7270
|
|
},
|
|
{
|
|
"epoch": 0.7275,
|
|
"grad_norm": 0.8417593240737915,
|
|
"learning_rate": 5.455455455455456e-05,
|
|
"loss": 2.3631,
|
|
"step": 7275
|
|
},
|
|
{
|
|
"epoch": 0.728,
|
|
"grad_norm": 0.8216800689697266,
|
|
"learning_rate": 5.445445445445446e-05,
|
|
"loss": 2.3934,
|
|
"step": 7280
|
|
},
|
|
{
|
|
"epoch": 0.7285,
|
|
"grad_norm": 0.8299776315689087,
|
|
"learning_rate": 5.435435435435435e-05,
|
|
"loss": 2.3744,
|
|
"step": 7285
|
|
},
|
|
{
|
|
"epoch": 0.729,
|
|
"grad_norm": 0.9270496964454651,
|
|
"learning_rate": 5.425425425425425e-05,
|
|
"loss": 2.379,
|
|
"step": 7290
|
|
},
|
|
{
|
|
"epoch": 0.7295,
|
|
"grad_norm": 0.7992855906486511,
|
|
"learning_rate": 5.415415415415416e-05,
|
|
"loss": 2.3673,
|
|
"step": 7295
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"grad_norm": 0.7973700165748596,
|
|
"learning_rate": 5.405405405405406e-05,
|
|
"loss": 2.3799,
|
|
"step": 7300
|
|
},
|
|
{
|
|
"epoch": 0.7305,
|
|
"grad_norm": 0.8608441352844238,
|
|
"learning_rate": 5.395395395395396e-05,
|
|
"loss": 2.3897,
|
|
"step": 7305
|
|
},
|
|
{
|
|
"epoch": 0.731,
|
|
"grad_norm": 0.7959920763969421,
|
|
"learning_rate": 5.3853853853853854e-05,
|
|
"loss": 2.3771,
|
|
"step": 7310
|
|
},
|
|
{
|
|
"epoch": 0.7315,
|
|
"grad_norm": 0.7734951972961426,
|
|
"learning_rate": 5.3753753753753754e-05,
|
|
"loss": 2.3836,
|
|
"step": 7315
|
|
},
|
|
{
|
|
"epoch": 0.732,
|
|
"grad_norm": 0.8180063962936401,
|
|
"learning_rate": 5.3653653653653654e-05,
|
|
"loss": 2.3562,
|
|
"step": 7320
|
|
},
|
|
{
|
|
"epoch": 0.7325,
|
|
"grad_norm": 0.817004382610321,
|
|
"learning_rate": 5.355355355355356e-05,
|
|
"loss": 2.3608,
|
|
"step": 7325
|
|
},
|
|
{
|
|
"epoch": 0.733,
|
|
"grad_norm": 0.7949969172477722,
|
|
"learning_rate": 5.345345345345346e-05,
|
|
"loss": 2.3812,
|
|
"step": 7330
|
|
},
|
|
{
|
|
"epoch": 0.7335,
|
|
"grad_norm": 0.8416751027107239,
|
|
"learning_rate": 5.3353353353353355e-05,
|
|
"loss": 2.3376,
|
|
"step": 7335
|
|
},
|
|
{
|
|
"epoch": 0.734,
|
|
"grad_norm": 0.8331473469734192,
|
|
"learning_rate": 5.3253253253253255e-05,
|
|
"loss": 2.3837,
|
|
"step": 7340
|
|
},
|
|
{
|
|
"epoch": 0.7345,
|
|
"grad_norm": 0.8140696883201599,
|
|
"learning_rate": 5.3153153153153155e-05,
|
|
"loss": 2.3737,
|
|
"step": 7345
|
|
},
|
|
{
|
|
"epoch": 0.735,
|
|
"grad_norm": 0.8087270259857178,
|
|
"learning_rate": 5.305305305305306e-05,
|
|
"loss": 2.3579,
|
|
"step": 7350
|
|
},
|
|
{
|
|
"epoch": 0.7355,
|
|
"grad_norm": 0.8513954877853394,
|
|
"learning_rate": 5.295295295295295e-05,
|
|
"loss": 2.3682,
|
|
"step": 7355
|
|
},
|
|
{
|
|
"epoch": 0.736,
|
|
"grad_norm": 0.8019487857818604,
|
|
"learning_rate": 5.2852852852852855e-05,
|
|
"loss": 2.383,
|
|
"step": 7360
|
|
},
|
|
{
|
|
"epoch": 0.7365,
|
|
"grad_norm": 0.8455563187599182,
|
|
"learning_rate": 5.2752752752752755e-05,
|
|
"loss": 2.3614,
|
|
"step": 7365
|
|
},
|
|
{
|
|
"epoch": 0.737,
|
|
"grad_norm": 0.8061125874519348,
|
|
"learning_rate": 5.2652652652652655e-05,
|
|
"loss": 2.3715,
|
|
"step": 7370
|
|
},
|
|
{
|
|
"epoch": 0.7375,
|
|
"grad_norm": 0.8147531151771545,
|
|
"learning_rate": 5.2552552552552556e-05,
|
|
"loss": 2.3663,
|
|
"step": 7375
|
|
},
|
|
{
|
|
"epoch": 0.738,
|
|
"grad_norm": 0.8364109992980957,
|
|
"learning_rate": 5.245245245245245e-05,
|
|
"loss": 2.3746,
|
|
"step": 7380
|
|
},
|
|
{
|
|
"epoch": 0.7385,
|
|
"grad_norm": 0.8020389080047607,
|
|
"learning_rate": 5.235235235235235e-05,
|
|
"loss": 2.3357,
|
|
"step": 7385
|
|
},
|
|
{
|
|
"epoch": 0.739,
|
|
"grad_norm": 0.8236432075500488,
|
|
"learning_rate": 5.2252252252252256e-05,
|
|
"loss": 2.3952,
|
|
"step": 7390
|
|
},
|
|
{
|
|
"epoch": 0.7395,
|
|
"grad_norm": 0.7781680226325989,
|
|
"learning_rate": 5.2152152152152156e-05,
|
|
"loss": 2.3833,
|
|
"step": 7395
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"grad_norm": 0.7892677187919617,
|
|
"learning_rate": 5.2052052052052056e-05,
|
|
"loss": 2.3658,
|
|
"step": 7400
|
|
},
|
|
{
|
|
"epoch": 0.7405,
|
|
"grad_norm": 0.808290958404541,
|
|
"learning_rate": 5.195195195195195e-05,
|
|
"loss": 2.3683,
|
|
"step": 7405
|
|
},
|
|
{
|
|
"epoch": 0.741,
|
|
"grad_norm": 0.7929525971412659,
|
|
"learning_rate": 5.185185185185185e-05,
|
|
"loss": 2.3563,
|
|
"step": 7410
|
|
},
|
|
{
|
|
"epoch": 0.7415,
|
|
"grad_norm": 0.8070533871650696,
|
|
"learning_rate": 5.175175175175176e-05,
|
|
"loss": 2.3546,
|
|
"step": 7415
|
|
},
|
|
{
|
|
"epoch": 0.742,
|
|
"grad_norm": 0.8430861234664917,
|
|
"learning_rate": 5.165165165165166e-05,
|
|
"loss": 2.366,
|
|
"step": 7420
|
|
},
|
|
{
|
|
"epoch": 0.7425,
|
|
"grad_norm": 0.8671832084655762,
|
|
"learning_rate": 5.155155155155156e-05,
|
|
"loss": 2.3898,
|
|
"step": 7425
|
|
},
|
|
{
|
|
"epoch": 0.743,
|
|
"grad_norm": 0.8964141011238098,
|
|
"learning_rate": 5.145145145145145e-05,
|
|
"loss": 2.3625,
|
|
"step": 7430
|
|
},
|
|
{
|
|
"epoch": 0.7435,
|
|
"grad_norm": 0.8146119713783264,
|
|
"learning_rate": 5.135135135135135e-05,
|
|
"loss": 2.3811,
|
|
"step": 7435
|
|
},
|
|
{
|
|
"epoch": 0.744,
|
|
"grad_norm": 0.83391273021698,
|
|
"learning_rate": 5.125125125125125e-05,
|
|
"loss": 2.3757,
|
|
"step": 7440
|
|
},
|
|
{
|
|
"epoch": 0.7445,
|
|
"grad_norm": 0.7972255945205688,
|
|
"learning_rate": 5.115115115115116e-05,
|
|
"loss": 2.379,
|
|
"step": 7445
|
|
},
|
|
{
|
|
"epoch": 0.745,
|
|
"grad_norm": 0.7837113738059998,
|
|
"learning_rate": 5.105105105105106e-05,
|
|
"loss": 2.3716,
|
|
"step": 7450
|
|
},
|
|
{
|
|
"epoch": 0.7455,
|
|
"grad_norm": 0.8616361618041992,
|
|
"learning_rate": 5.095095095095095e-05,
|
|
"loss": 2.3552,
|
|
"step": 7455
|
|
},
|
|
{
|
|
"epoch": 0.746,
|
|
"grad_norm": 0.8113681077957153,
|
|
"learning_rate": 5.085085085085085e-05,
|
|
"loss": 2.3789,
|
|
"step": 7460
|
|
},
|
|
{
|
|
"epoch": 0.7465,
|
|
"grad_norm": 0.7940648198127747,
|
|
"learning_rate": 5.075075075075075e-05,
|
|
"loss": 2.3643,
|
|
"step": 7465
|
|
},
|
|
{
|
|
"epoch": 0.747,
|
|
"grad_norm": 0.8371662497520447,
|
|
"learning_rate": 5.065065065065066e-05,
|
|
"loss": 2.3637,
|
|
"step": 7470
|
|
},
|
|
{
|
|
"epoch": 0.7475,
|
|
"grad_norm": 0.8097828030586243,
|
|
"learning_rate": 5.055055055055056e-05,
|
|
"loss": 2.386,
|
|
"step": 7475
|
|
},
|
|
{
|
|
"epoch": 0.748,
|
|
"grad_norm": 0.8285651803016663,
|
|
"learning_rate": 5.0450450450450445e-05,
|
|
"loss": 2.3578,
|
|
"step": 7480
|
|
},
|
|
{
|
|
"epoch": 0.7485,
|
|
"grad_norm": 0.7992802858352661,
|
|
"learning_rate": 5.035035035035035e-05,
|
|
"loss": 2.3586,
|
|
"step": 7485
|
|
},
|
|
{
|
|
"epoch": 0.749,
|
|
"grad_norm": 0.8517287969589233,
|
|
"learning_rate": 5.025025025025025e-05,
|
|
"loss": 2.3644,
|
|
"step": 7490
|
|
},
|
|
{
|
|
"epoch": 0.7495,
|
|
"grad_norm": 0.8603693842887878,
|
|
"learning_rate": 5.015015015015015e-05,
|
|
"loss": 2.3585,
|
|
"step": 7495
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"grad_norm": 0.837411105632782,
|
|
"learning_rate": 5.005005005005006e-05,
|
|
"loss": 2.3527,
|
|
"step": 7500
|
|
},
|
|
{
|
|
"epoch": 0.7505,
|
|
"grad_norm": 0.7850192189216614,
|
|
"learning_rate": 4.994994994994995e-05,
|
|
"loss": 2.3614,
|
|
"step": 7505
|
|
},
|
|
{
|
|
"epoch": 0.751,
|
|
"grad_norm": 0.7993366718292236,
|
|
"learning_rate": 4.984984984984985e-05,
|
|
"loss": 2.3432,
|
|
"step": 7510
|
|
},
|
|
{
|
|
"epoch": 0.7515,
|
|
"grad_norm": 0.801365315914154,
|
|
"learning_rate": 4.974974974974975e-05,
|
|
"loss": 2.3446,
|
|
"step": 7515
|
|
},
|
|
{
|
|
"epoch": 0.752,
|
|
"grad_norm": 0.8253291249275208,
|
|
"learning_rate": 4.964964964964965e-05,
|
|
"loss": 2.3685,
|
|
"step": 7520
|
|
},
|
|
{
|
|
"epoch": 0.7525,
|
|
"grad_norm": 0.8093944787979126,
|
|
"learning_rate": 4.954954954954955e-05,
|
|
"loss": 2.3598,
|
|
"step": 7525
|
|
},
|
|
{
|
|
"epoch": 0.753,
|
|
"grad_norm": 0.7934316396713257,
|
|
"learning_rate": 4.944944944944945e-05,
|
|
"loss": 2.369,
|
|
"step": 7530
|
|
},
|
|
{
|
|
"epoch": 0.7535,
|
|
"grad_norm": 0.8154873847961426,
|
|
"learning_rate": 4.9349349349349347e-05,
|
|
"loss": 2.3792,
|
|
"step": 7535
|
|
},
|
|
{
|
|
"epoch": 0.754,
|
|
"grad_norm": 0.8415142893791199,
|
|
"learning_rate": 4.9249249249249253e-05,
|
|
"loss": 2.3622,
|
|
"step": 7540
|
|
},
|
|
{
|
|
"epoch": 0.7545,
|
|
"grad_norm": 0.9094659090042114,
|
|
"learning_rate": 4.9149149149149154e-05,
|
|
"loss": 2.3663,
|
|
"step": 7545
|
|
},
|
|
{
|
|
"epoch": 0.755,
|
|
"grad_norm": 0.8376139998435974,
|
|
"learning_rate": 4.9049049049049054e-05,
|
|
"loss": 2.3486,
|
|
"step": 7550
|
|
},
|
|
{
|
|
"epoch": 0.7555,
|
|
"grad_norm": 0.7904059290885925,
|
|
"learning_rate": 4.8948948948948954e-05,
|
|
"loss": 2.3417,
|
|
"step": 7555
|
|
},
|
|
{
|
|
"epoch": 0.756,
|
|
"grad_norm": 0.8456617593765259,
|
|
"learning_rate": 4.884884884884885e-05,
|
|
"loss": 2.3765,
|
|
"step": 7560
|
|
},
|
|
{
|
|
"epoch": 0.7565,
|
|
"grad_norm": 0.8238437175750732,
|
|
"learning_rate": 4.8748748748748754e-05,
|
|
"loss": 2.3768,
|
|
"step": 7565
|
|
},
|
|
{
|
|
"epoch": 0.757,
|
|
"grad_norm": 0.8127914667129517,
|
|
"learning_rate": 4.8648648648648654e-05,
|
|
"loss": 2.3459,
|
|
"step": 7570
|
|
},
|
|
{
|
|
"epoch": 0.7575,
|
|
"grad_norm": 0.879636824131012,
|
|
"learning_rate": 4.854854854854855e-05,
|
|
"loss": 2.3588,
|
|
"step": 7575
|
|
},
|
|
{
|
|
"epoch": 0.758,
|
|
"grad_norm": 0.8751839399337769,
|
|
"learning_rate": 4.8448448448448455e-05,
|
|
"loss": 2.3566,
|
|
"step": 7580
|
|
},
|
|
{
|
|
"epoch": 0.7585,
|
|
"grad_norm": 0.80313640832901,
|
|
"learning_rate": 4.834834834834835e-05,
|
|
"loss": 2.3659,
|
|
"step": 7585
|
|
},
|
|
{
|
|
"epoch": 0.759,
|
|
"grad_norm": 0.8672766089439392,
|
|
"learning_rate": 4.824824824824825e-05,
|
|
"loss": 2.3713,
|
|
"step": 7590
|
|
},
|
|
{
|
|
"epoch": 0.7595,
|
|
"grad_norm": 0.8659483790397644,
|
|
"learning_rate": 4.814814814814815e-05,
|
|
"loss": 2.3806,
|
|
"step": 7595
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"grad_norm": 0.816363513469696,
|
|
"learning_rate": 4.804804804804805e-05,
|
|
"loss": 2.3593,
|
|
"step": 7600
|
|
},
|
|
{
|
|
"epoch": 0.7605,
|
|
"grad_norm": 0.8372772932052612,
|
|
"learning_rate": 4.7947947947947955e-05,
|
|
"loss": 2.3402,
|
|
"step": 7605
|
|
},
|
|
{
|
|
"epoch": 0.761,
|
|
"grad_norm": 0.841561496257782,
|
|
"learning_rate": 4.784784784784785e-05,
|
|
"loss": 2.3519,
|
|
"step": 7610
|
|
},
|
|
{
|
|
"epoch": 0.7615,
|
|
"grad_norm": 0.7936203479766846,
|
|
"learning_rate": 4.774774774774775e-05,
|
|
"loss": 2.3714,
|
|
"step": 7615
|
|
},
|
|
{
|
|
"epoch": 0.762,
|
|
"grad_norm": 0.8175353407859802,
|
|
"learning_rate": 4.764764764764765e-05,
|
|
"loss": 2.3571,
|
|
"step": 7620
|
|
},
|
|
{
|
|
"epoch": 0.7625,
|
|
"grad_norm": 0.8095380663871765,
|
|
"learning_rate": 4.754754754754755e-05,
|
|
"loss": 2.3731,
|
|
"step": 7625
|
|
},
|
|
{
|
|
"epoch": 0.763,
|
|
"grad_norm": 0.86039799451828,
|
|
"learning_rate": 4.744744744744745e-05,
|
|
"loss": 2.3402,
|
|
"step": 7630
|
|
},
|
|
{
|
|
"epoch": 0.7635,
|
|
"grad_norm": 0.7970331311225891,
|
|
"learning_rate": 4.734734734734735e-05,
|
|
"loss": 2.3562,
|
|
"step": 7635
|
|
},
|
|
{
|
|
"epoch": 0.764,
|
|
"grad_norm": 0.8392342925071716,
|
|
"learning_rate": 4.724724724724725e-05,
|
|
"loss": 2.3739,
|
|
"step": 7640
|
|
},
|
|
{
|
|
"epoch": 0.7645,
|
|
"grad_norm": 0.8334502577781677,
|
|
"learning_rate": 4.714714714714715e-05,
|
|
"loss": 2.3464,
|
|
"step": 7645
|
|
},
|
|
{
|
|
"epoch": 0.765,
|
|
"grad_norm": 0.8409643173217773,
|
|
"learning_rate": 4.704704704704705e-05,
|
|
"loss": 2.4046,
|
|
"step": 7650
|
|
},
|
|
{
|
|
"epoch": 0.7655,
|
|
"grad_norm": 0.8266360759735107,
|
|
"learning_rate": 4.694694694694695e-05,
|
|
"loss": 2.3713,
|
|
"step": 7655
|
|
},
|
|
{
|
|
"epoch": 0.766,
|
|
"grad_norm": 0.7967942953109741,
|
|
"learning_rate": 4.684684684684685e-05,
|
|
"loss": 2.3632,
|
|
"step": 7660
|
|
},
|
|
{
|
|
"epoch": 0.7665,
|
|
"grad_norm": 0.8663550019264221,
|
|
"learning_rate": 4.674674674674675e-05,
|
|
"loss": 2.3539,
|
|
"step": 7665
|
|
},
|
|
{
|
|
"epoch": 0.767,
|
|
"grad_norm": 0.832740068435669,
|
|
"learning_rate": 4.6646646646646644e-05,
|
|
"loss": 2.3424,
|
|
"step": 7670
|
|
},
|
|
{
|
|
"epoch": 0.7675,
|
|
"grad_norm": 0.85154789686203,
|
|
"learning_rate": 4.654654654654655e-05,
|
|
"loss": 2.3567,
|
|
"step": 7675
|
|
},
|
|
{
|
|
"epoch": 0.768,
|
|
"grad_norm": 0.8139371871948242,
|
|
"learning_rate": 4.644644644644645e-05,
|
|
"loss": 2.3845,
|
|
"step": 7680
|
|
},
|
|
{
|
|
"epoch": 0.7685,
|
|
"grad_norm": 0.8819915056228638,
|
|
"learning_rate": 4.634634634634635e-05,
|
|
"loss": 2.368,
|
|
"step": 7685
|
|
},
|
|
{
|
|
"epoch": 0.769,
|
|
"grad_norm": 0.8150420188903809,
|
|
"learning_rate": 4.624624624624625e-05,
|
|
"loss": 2.3595,
|
|
"step": 7690
|
|
},
|
|
{
|
|
"epoch": 0.7695,
|
|
"grad_norm": 0.8445594310760498,
|
|
"learning_rate": 4.6146146146146144e-05,
|
|
"loss": 2.3579,
|
|
"step": 7695
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"grad_norm": 0.8054527044296265,
|
|
"learning_rate": 4.604604604604605e-05,
|
|
"loss": 2.3534,
|
|
"step": 7700
|
|
},
|
|
{
|
|
"epoch": 0.7705,
|
|
"grad_norm": 0.8040105104446411,
|
|
"learning_rate": 4.594594594594595e-05,
|
|
"loss": 2.3824,
|
|
"step": 7705
|
|
},
|
|
{
|
|
"epoch": 0.771,
|
|
"grad_norm": 0.8173477053642273,
|
|
"learning_rate": 4.5845845845845845e-05,
|
|
"loss": 2.358,
|
|
"step": 7710
|
|
},
|
|
{
|
|
"epoch": 0.7715,
|
|
"grad_norm": 0.8278516530990601,
|
|
"learning_rate": 4.574574574574575e-05,
|
|
"loss": 2.3458,
|
|
"step": 7715
|
|
},
|
|
{
|
|
"epoch": 0.772,
|
|
"grad_norm": 0.8155180811882019,
|
|
"learning_rate": 4.5645645645645645e-05,
|
|
"loss": 2.3542,
|
|
"step": 7720
|
|
},
|
|
{
|
|
"epoch": 0.7725,
|
|
"grad_norm": 0.811967670917511,
|
|
"learning_rate": 4.5545545545545545e-05,
|
|
"loss": 2.3689,
|
|
"step": 7725
|
|
},
|
|
{
|
|
"epoch": 0.773,
|
|
"grad_norm": 0.8122997283935547,
|
|
"learning_rate": 4.544544544544545e-05,
|
|
"loss": 2.3797,
|
|
"step": 7730
|
|
},
|
|
{
|
|
"epoch": 0.7735,
|
|
"grad_norm": 0.8238713145256042,
|
|
"learning_rate": 4.5345345345345345e-05,
|
|
"loss": 2.3871,
|
|
"step": 7735
|
|
},
|
|
{
|
|
"epoch": 0.774,
|
|
"grad_norm": 0.7975063920021057,
|
|
"learning_rate": 4.524524524524525e-05,
|
|
"loss": 2.3637,
|
|
"step": 7740
|
|
},
|
|
{
|
|
"epoch": 0.7745,
|
|
"grad_norm": 0.8312461376190186,
|
|
"learning_rate": 4.5145145145145146e-05,
|
|
"loss": 2.3469,
|
|
"step": 7745
|
|
},
|
|
{
|
|
"epoch": 0.775,
|
|
"grad_norm": 0.8213257193565369,
|
|
"learning_rate": 4.5045045045045046e-05,
|
|
"loss": 2.3878,
|
|
"step": 7750
|
|
},
|
|
{
|
|
"epoch": 0.7755,
|
|
"grad_norm": 0.8709314465522766,
|
|
"learning_rate": 4.4944944944944946e-05,
|
|
"loss": 2.3642,
|
|
"step": 7755
|
|
},
|
|
{
|
|
"epoch": 0.776,
|
|
"grad_norm": 0.8140637874603271,
|
|
"learning_rate": 4.4844844844844846e-05,
|
|
"loss": 2.3533,
|
|
"step": 7760
|
|
},
|
|
{
|
|
"epoch": 0.7765,
|
|
"grad_norm": 0.8330239653587341,
|
|
"learning_rate": 4.4744744744744746e-05,
|
|
"loss": 2.3647,
|
|
"step": 7765
|
|
},
|
|
{
|
|
"epoch": 0.777,
|
|
"grad_norm": 0.8120850324630737,
|
|
"learning_rate": 4.4644644644644646e-05,
|
|
"loss": 2.3587,
|
|
"step": 7770
|
|
},
|
|
{
|
|
"epoch": 0.7775,
|
|
"grad_norm": 0.8139227032661438,
|
|
"learning_rate": 4.4544544544544546e-05,
|
|
"loss": 2.3635,
|
|
"step": 7775
|
|
},
|
|
{
|
|
"epoch": 0.778,
|
|
"grad_norm": 0.8348556160926819,
|
|
"learning_rate": 4.4444444444444447e-05,
|
|
"loss": 2.3721,
|
|
"step": 7780
|
|
},
|
|
{
|
|
"epoch": 0.7785,
|
|
"grad_norm": 0.8350592851638794,
|
|
"learning_rate": 4.434434434434435e-05,
|
|
"loss": 2.3801,
|
|
"step": 7785
|
|
},
|
|
{
|
|
"epoch": 0.779,
|
|
"grad_norm": 0.8146738409996033,
|
|
"learning_rate": 4.424424424424425e-05,
|
|
"loss": 2.3544,
|
|
"step": 7790
|
|
},
|
|
{
|
|
"epoch": 0.7795,
|
|
"grad_norm": 0.82841956615448,
|
|
"learning_rate": 4.414414414414415e-05,
|
|
"loss": 2.3373,
|
|
"step": 7795
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"grad_norm": 0.8554604649543762,
|
|
"learning_rate": 4.404404404404405e-05,
|
|
"loss": 2.3614,
|
|
"step": 7800
|
|
},
|
|
{
|
|
"epoch": 0.7805,
|
|
"grad_norm": 0.8313968181610107,
|
|
"learning_rate": 4.394394394394394e-05,
|
|
"loss": 2.3451,
|
|
"step": 7805
|
|
},
|
|
{
|
|
"epoch": 0.781,
|
|
"grad_norm": 0.8203895688056946,
|
|
"learning_rate": 4.384384384384385e-05,
|
|
"loss": 2.3601,
|
|
"step": 7810
|
|
},
|
|
{
|
|
"epoch": 0.7815,
|
|
"grad_norm": 0.814744234085083,
|
|
"learning_rate": 4.374374374374375e-05,
|
|
"loss": 2.3856,
|
|
"step": 7815
|
|
},
|
|
{
|
|
"epoch": 0.782,
|
|
"grad_norm": 0.8176069259643555,
|
|
"learning_rate": 4.364364364364365e-05,
|
|
"loss": 2.3435,
|
|
"step": 7820
|
|
},
|
|
{
|
|
"epoch": 0.7825,
|
|
"grad_norm": 0.8046497702598572,
|
|
"learning_rate": 4.354354354354355e-05,
|
|
"loss": 2.3643,
|
|
"step": 7825
|
|
},
|
|
{
|
|
"epoch": 0.783,
|
|
"grad_norm": 0.8219382762908936,
|
|
"learning_rate": 4.344344344344344e-05,
|
|
"loss": 2.3636,
|
|
"step": 7830
|
|
},
|
|
{
|
|
"epoch": 0.7835,
|
|
"grad_norm": 0.8374064564704895,
|
|
"learning_rate": 4.334334334334335e-05,
|
|
"loss": 2.3669,
|
|
"step": 7835
|
|
},
|
|
{
|
|
"epoch": 0.784,
|
|
"grad_norm": 0.8313128352165222,
|
|
"learning_rate": 4.324324324324325e-05,
|
|
"loss": 2.3303,
|
|
"step": 7840
|
|
},
|
|
{
|
|
"epoch": 0.7845,
|
|
"grad_norm": 0.8300835490226746,
|
|
"learning_rate": 4.314314314314314e-05,
|
|
"loss": 2.3293,
|
|
"step": 7845
|
|
},
|
|
{
|
|
"epoch": 0.785,
|
|
"grad_norm": 0.8263983130455017,
|
|
"learning_rate": 4.304304304304305e-05,
|
|
"loss": 2.3495,
|
|
"step": 7850
|
|
},
|
|
{
|
|
"epoch": 0.7855,
|
|
"grad_norm": 0.835594654083252,
|
|
"learning_rate": 4.294294294294294e-05,
|
|
"loss": 2.3481,
|
|
"step": 7855
|
|
},
|
|
{
|
|
"epoch": 0.786,
|
|
"grad_norm": 0.8166428804397583,
|
|
"learning_rate": 4.284284284284284e-05,
|
|
"loss": 2.3811,
|
|
"step": 7860
|
|
},
|
|
{
|
|
"epoch": 0.7865,
|
|
"grad_norm": 0.8068861961364746,
|
|
"learning_rate": 4.274274274274275e-05,
|
|
"loss": 2.3759,
|
|
"step": 7865
|
|
},
|
|
{
|
|
"epoch": 0.787,
|
|
"grad_norm": 0.8637300133705139,
|
|
"learning_rate": 4.264264264264264e-05,
|
|
"loss": 2.3728,
|
|
"step": 7870
|
|
},
|
|
{
|
|
"epoch": 0.7875,
|
|
"grad_norm": 0.8304838538169861,
|
|
"learning_rate": 4.254254254254255e-05,
|
|
"loss": 2.3558,
|
|
"step": 7875
|
|
},
|
|
{
|
|
"epoch": 0.788,
|
|
"grad_norm": 0.813943088054657,
|
|
"learning_rate": 4.244244244244244e-05,
|
|
"loss": 2.3273,
|
|
"step": 7880
|
|
},
|
|
{
|
|
"epoch": 0.7885,
|
|
"grad_norm": 0.8327081203460693,
|
|
"learning_rate": 4.234234234234234e-05,
|
|
"loss": 2.3514,
|
|
"step": 7885
|
|
},
|
|
{
|
|
"epoch": 0.789,
|
|
"grad_norm": 0.8549174070358276,
|
|
"learning_rate": 4.224224224224225e-05,
|
|
"loss": 2.3849,
|
|
"step": 7890
|
|
},
|
|
{
|
|
"epoch": 0.7895,
|
|
"grad_norm": 0.8003869652748108,
|
|
"learning_rate": 4.214214214214214e-05,
|
|
"loss": 2.3456,
|
|
"step": 7895
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"grad_norm": 0.8233539462089539,
|
|
"learning_rate": 4.204204204204204e-05,
|
|
"loss": 2.3488,
|
|
"step": 7900
|
|
},
|
|
{
|
|
"epoch": 0.7905,
|
|
"grad_norm": 0.8078953623771667,
|
|
"learning_rate": 4.194194194194194e-05,
|
|
"loss": 2.3626,
|
|
"step": 7905
|
|
},
|
|
{
|
|
"epoch": 0.791,
|
|
"grad_norm": 0.8408604264259338,
|
|
"learning_rate": 4.1841841841841843e-05,
|
|
"loss": 2.3702,
|
|
"step": 7910
|
|
},
|
|
{
|
|
"epoch": 0.7915,
|
|
"grad_norm": 0.8244900703430176,
|
|
"learning_rate": 4.1741741741741744e-05,
|
|
"loss": 2.3367,
|
|
"step": 7915
|
|
},
|
|
{
|
|
"epoch": 0.792,
|
|
"grad_norm": 0.8229595422744751,
|
|
"learning_rate": 4.1641641641641644e-05,
|
|
"loss": 2.3475,
|
|
"step": 7920
|
|
},
|
|
{
|
|
"epoch": 0.7925,
|
|
"grad_norm": 0.8468862175941467,
|
|
"learning_rate": 4.1541541541541544e-05,
|
|
"loss": 2.3716,
|
|
"step": 7925
|
|
},
|
|
{
|
|
"epoch": 0.793,
|
|
"grad_norm": 0.7976715564727783,
|
|
"learning_rate": 4.1441441441441444e-05,
|
|
"loss": 2.3701,
|
|
"step": 7930
|
|
},
|
|
{
|
|
"epoch": 0.7935,
|
|
"grad_norm": 0.7963828444480896,
|
|
"learning_rate": 4.1341341341341344e-05,
|
|
"loss": 2.3521,
|
|
"step": 7935
|
|
},
|
|
{
|
|
"epoch": 0.794,
|
|
"grad_norm": 0.8006214499473572,
|
|
"learning_rate": 4.124124124124124e-05,
|
|
"loss": 2.3594,
|
|
"step": 7940
|
|
},
|
|
{
|
|
"epoch": 0.7945,
|
|
"grad_norm": 0.8206690549850464,
|
|
"learning_rate": 4.1141141141141144e-05,
|
|
"loss": 2.3498,
|
|
"step": 7945
|
|
},
|
|
{
|
|
"epoch": 0.795,
|
|
"grad_norm": 0.8171199560165405,
|
|
"learning_rate": 4.1041041041041045e-05,
|
|
"loss": 2.3487,
|
|
"step": 7950
|
|
},
|
|
{
|
|
"epoch": 0.7955,
|
|
"grad_norm": 0.8080857992172241,
|
|
"learning_rate": 4.0940940940940945e-05,
|
|
"loss": 2.3263,
|
|
"step": 7955
|
|
},
|
|
{
|
|
"epoch": 0.796,
|
|
"grad_norm": 0.8231513500213623,
|
|
"learning_rate": 4.0840840840840845e-05,
|
|
"loss": 2.3729,
|
|
"step": 7960
|
|
},
|
|
{
|
|
"epoch": 0.7965,
|
|
"grad_norm": 0.8671808242797852,
|
|
"learning_rate": 4.074074074074074e-05,
|
|
"loss": 2.376,
|
|
"step": 7965
|
|
},
|
|
{
|
|
"epoch": 0.797,
|
|
"grad_norm": 0.790890097618103,
|
|
"learning_rate": 4.0640640640640645e-05,
|
|
"loss": 2.3345,
|
|
"step": 7970
|
|
},
|
|
{
|
|
"epoch": 0.7975,
|
|
"grad_norm": 0.8026176691055298,
|
|
"learning_rate": 4.0540540540540545e-05,
|
|
"loss": 2.3578,
|
|
"step": 7975
|
|
},
|
|
{
|
|
"epoch": 0.798,
|
|
"grad_norm": 0.799191415309906,
|
|
"learning_rate": 4.044044044044044e-05,
|
|
"loss": 2.3585,
|
|
"step": 7980
|
|
},
|
|
{
|
|
"epoch": 0.7985,
|
|
"grad_norm": 0.8245786428451538,
|
|
"learning_rate": 4.0340340340340346e-05,
|
|
"loss": 2.3535,
|
|
"step": 7985
|
|
},
|
|
{
|
|
"epoch": 0.799,
|
|
"grad_norm": 0.8001775145530701,
|
|
"learning_rate": 4.024024024024024e-05,
|
|
"loss": 2.3813,
|
|
"step": 7990
|
|
},
|
|
{
|
|
"epoch": 0.7995,
|
|
"grad_norm": 0.8119397163391113,
|
|
"learning_rate": 4.014014014014014e-05,
|
|
"loss": 2.34,
|
|
"step": 7995
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"grad_norm": 0.8177365660667419,
|
|
"learning_rate": 4.0040040040040046e-05,
|
|
"loss": 2.3409,
|
|
"step": 8000
|
|
},
|
|
{
|
|
"epoch": 0.8005,
|
|
"grad_norm": 0.8202060461044312,
|
|
"learning_rate": 3.993993993993994e-05,
|
|
"loss": 2.3438,
|
|
"step": 8005
|
|
},
|
|
{
|
|
"epoch": 0.801,
|
|
"grad_norm": 0.8008648157119751,
|
|
"learning_rate": 3.9839839839839846e-05,
|
|
"loss": 2.3566,
|
|
"step": 8010
|
|
},
|
|
{
|
|
"epoch": 0.8015,
|
|
"grad_norm": 0.8599295616149902,
|
|
"learning_rate": 3.973973973973974e-05,
|
|
"loss": 2.3465,
|
|
"step": 8015
|
|
},
|
|
{
|
|
"epoch": 0.802,
|
|
"grad_norm": 0.8756669163703918,
|
|
"learning_rate": 3.963963963963964e-05,
|
|
"loss": 2.3387,
|
|
"step": 8020
|
|
},
|
|
{
|
|
"epoch": 0.8025,
|
|
"grad_norm": 0.8419957756996155,
|
|
"learning_rate": 3.953953953953955e-05,
|
|
"loss": 2.3652,
|
|
"step": 8025
|
|
},
|
|
{
|
|
"epoch": 0.803,
|
|
"grad_norm": 0.8807148933410645,
|
|
"learning_rate": 3.943943943943944e-05,
|
|
"loss": 2.3626,
|
|
"step": 8030
|
|
},
|
|
{
|
|
"epoch": 0.8035,
|
|
"grad_norm": 0.8416363596916199,
|
|
"learning_rate": 3.933933933933934e-05,
|
|
"loss": 2.3441,
|
|
"step": 8035
|
|
},
|
|
{
|
|
"epoch": 0.804,
|
|
"grad_norm": 0.8050745725631714,
|
|
"learning_rate": 3.923923923923924e-05,
|
|
"loss": 2.3327,
|
|
"step": 8040
|
|
},
|
|
{
|
|
"epoch": 0.8045,
|
|
"grad_norm": 0.8007880449295044,
|
|
"learning_rate": 3.913913913913914e-05,
|
|
"loss": 2.3382,
|
|
"step": 8045
|
|
},
|
|
{
|
|
"epoch": 0.805,
|
|
"grad_norm": 0.8622052669525146,
|
|
"learning_rate": 3.903903903903904e-05,
|
|
"loss": 2.3388,
|
|
"step": 8050
|
|
},
|
|
{
|
|
"epoch": 0.8055,
|
|
"grad_norm": 0.8893733024597168,
|
|
"learning_rate": 3.893893893893894e-05,
|
|
"loss": 2.3553,
|
|
"step": 8055
|
|
},
|
|
{
|
|
"epoch": 0.806,
|
|
"grad_norm": 0.809303343296051,
|
|
"learning_rate": 3.883883883883884e-05,
|
|
"loss": 2.3664,
|
|
"step": 8060
|
|
},
|
|
{
|
|
"epoch": 0.8065,
|
|
"grad_norm": 0.8065176010131836,
|
|
"learning_rate": 3.873873873873874e-05,
|
|
"loss": 2.3279,
|
|
"step": 8065
|
|
},
|
|
{
|
|
"epoch": 0.807,
|
|
"grad_norm": 0.8207269310951233,
|
|
"learning_rate": 3.863863863863864e-05,
|
|
"loss": 2.3574,
|
|
"step": 8070
|
|
},
|
|
{
|
|
"epoch": 0.8075,
|
|
"grad_norm": 0.8416551351547241,
|
|
"learning_rate": 3.8538538538538534e-05,
|
|
"loss": 2.3645,
|
|
"step": 8075
|
|
},
|
|
{
|
|
"epoch": 0.808,
|
|
"grad_norm": 0.8307541012763977,
|
|
"learning_rate": 3.843843843843844e-05,
|
|
"loss": 2.3598,
|
|
"step": 8080
|
|
},
|
|
{
|
|
"epoch": 0.8085,
|
|
"grad_norm": 0.887773871421814,
|
|
"learning_rate": 3.833833833833834e-05,
|
|
"loss": 2.3657,
|
|
"step": 8085
|
|
},
|
|
{
|
|
"epoch": 0.809,
|
|
"grad_norm": 0.8345415592193604,
|
|
"learning_rate": 3.823823823823824e-05,
|
|
"loss": 2.3648,
|
|
"step": 8090
|
|
},
|
|
{
|
|
"epoch": 0.8095,
|
|
"grad_norm": 0.8247503638267517,
|
|
"learning_rate": 3.813813813813814e-05,
|
|
"loss": 2.3438,
|
|
"step": 8095
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"grad_norm": 0.8071210384368896,
|
|
"learning_rate": 3.8038038038038035e-05,
|
|
"loss": 2.3442,
|
|
"step": 8100
|
|
},
|
|
{
|
|
"epoch": 0.8105,
|
|
"grad_norm": 0.839046835899353,
|
|
"learning_rate": 3.793793793793794e-05,
|
|
"loss": 2.3448,
|
|
"step": 8105
|
|
},
|
|
{
|
|
"epoch": 0.811,
|
|
"grad_norm": 0.8256351947784424,
|
|
"learning_rate": 3.783783783783784e-05,
|
|
"loss": 2.3415,
|
|
"step": 8110
|
|
},
|
|
{
|
|
"epoch": 0.8115,
|
|
"grad_norm": 0.8096026182174683,
|
|
"learning_rate": 3.7737737737737736e-05,
|
|
"loss": 2.3541,
|
|
"step": 8115
|
|
},
|
|
{
|
|
"epoch": 0.812,
|
|
"grad_norm": 0.8799003958702087,
|
|
"learning_rate": 3.763763763763764e-05,
|
|
"loss": 2.359,
|
|
"step": 8120
|
|
},
|
|
{
|
|
"epoch": 0.8125,
|
|
"grad_norm": 0.8653494119644165,
|
|
"learning_rate": 3.7537537537537536e-05,
|
|
"loss": 2.3481,
|
|
"step": 8125
|
|
},
|
|
{
|
|
"epoch": 0.813,
|
|
"grad_norm": 0.8147699236869812,
|
|
"learning_rate": 3.7437437437437436e-05,
|
|
"loss": 2.3438,
|
|
"step": 8130
|
|
},
|
|
{
|
|
"epoch": 0.8135,
|
|
"grad_norm": 0.9093450307846069,
|
|
"learning_rate": 3.733733733733734e-05,
|
|
"loss": 2.3618,
|
|
"step": 8135
|
|
},
|
|
{
|
|
"epoch": 0.814,
|
|
"grad_norm": 0.8373345136642456,
|
|
"learning_rate": 3.7237237237237236e-05,
|
|
"loss": 2.3646,
|
|
"step": 8140
|
|
},
|
|
{
|
|
"epoch": 0.8145,
|
|
"grad_norm": 0.7906237840652466,
|
|
"learning_rate": 3.713713713713714e-05,
|
|
"loss": 2.3572,
|
|
"step": 8145
|
|
},
|
|
{
|
|
"epoch": 0.815,
|
|
"grad_norm": 0.8436282873153687,
|
|
"learning_rate": 3.7037037037037037e-05,
|
|
"loss": 2.3573,
|
|
"step": 8150
|
|
},
|
|
{
|
|
"epoch": 0.8155,
|
|
"grad_norm": 0.8746330738067627,
|
|
"learning_rate": 3.693693693693694e-05,
|
|
"loss": 2.3511,
|
|
"step": 8155
|
|
},
|
|
{
|
|
"epoch": 0.816,
|
|
"grad_norm": 0.8541123867034912,
|
|
"learning_rate": 3.6836836836836844e-05,
|
|
"loss": 2.3531,
|
|
"step": 8160
|
|
},
|
|
{
|
|
"epoch": 0.8165,
|
|
"grad_norm": 0.8279290199279785,
|
|
"learning_rate": 3.673673673673674e-05,
|
|
"loss": 2.3557,
|
|
"step": 8165
|
|
},
|
|
{
|
|
"epoch": 0.817,
|
|
"grad_norm": 0.8584079146385193,
|
|
"learning_rate": 3.663663663663664e-05,
|
|
"loss": 2.3309,
|
|
"step": 8170
|
|
},
|
|
{
|
|
"epoch": 0.8175,
|
|
"grad_norm": 0.8298118710517883,
|
|
"learning_rate": 3.653653653653654e-05,
|
|
"loss": 2.3279,
|
|
"step": 8175
|
|
},
|
|
{
|
|
"epoch": 0.818,
|
|
"grad_norm": 0.8362030386924744,
|
|
"learning_rate": 3.643643643643644e-05,
|
|
"loss": 2.3485,
|
|
"step": 8180
|
|
},
|
|
{
|
|
"epoch": 0.8185,
|
|
"grad_norm": 0.7963013648986816,
|
|
"learning_rate": 3.633633633633634e-05,
|
|
"loss": 2.3488,
|
|
"step": 8185
|
|
},
|
|
{
|
|
"epoch": 0.819,
|
|
"grad_norm": 0.8097391724586487,
|
|
"learning_rate": 3.623623623623624e-05,
|
|
"loss": 2.323,
|
|
"step": 8190
|
|
},
|
|
{
|
|
"epoch": 0.8195,
|
|
"grad_norm": 0.8353894948959351,
|
|
"learning_rate": 3.613613613613614e-05,
|
|
"loss": 2.3723,
|
|
"step": 8195
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"grad_norm": 0.8266981244087219,
|
|
"learning_rate": 3.603603603603604e-05,
|
|
"loss": 2.3598,
|
|
"step": 8200
|
|
},
|
|
{
|
|
"epoch": 0.8205,
|
|
"grad_norm": 0.8105880618095398,
|
|
"learning_rate": 3.593593593593594e-05,
|
|
"loss": 2.3505,
|
|
"step": 8205
|
|
},
|
|
{
|
|
"epoch": 0.821,
|
|
"grad_norm": 0.8008158802986145,
|
|
"learning_rate": 3.583583583583583e-05,
|
|
"loss": 2.3643,
|
|
"step": 8210
|
|
},
|
|
{
|
|
"epoch": 0.8215,
|
|
"grad_norm": 0.7863180041313171,
|
|
"learning_rate": 3.573573573573574e-05,
|
|
"loss": 2.3433,
|
|
"step": 8215
|
|
},
|
|
{
|
|
"epoch": 0.822,
|
|
"grad_norm": 0.8484141826629639,
|
|
"learning_rate": 3.563563563563564e-05,
|
|
"loss": 2.3515,
|
|
"step": 8220
|
|
},
|
|
{
|
|
"epoch": 0.8225,
|
|
"grad_norm": 0.8206436038017273,
|
|
"learning_rate": 3.553553553553554e-05,
|
|
"loss": 2.351,
|
|
"step": 8225
|
|
},
|
|
{
|
|
"epoch": 0.823,
|
|
"grad_norm": 0.7947841882705688,
|
|
"learning_rate": 3.543543543543544e-05,
|
|
"loss": 2.3378,
|
|
"step": 8230
|
|
},
|
|
{
|
|
"epoch": 0.8235,
|
|
"grad_norm": 0.8877801299095154,
|
|
"learning_rate": 3.533533533533533e-05,
|
|
"loss": 2.365,
|
|
"step": 8235
|
|
},
|
|
{
|
|
"epoch": 0.824,
|
|
"grad_norm": 0.8179998397827148,
|
|
"learning_rate": 3.523523523523524e-05,
|
|
"loss": 2.3607,
|
|
"step": 8240
|
|
},
|
|
{
|
|
"epoch": 0.8245,
|
|
"grad_norm": 0.8523365259170532,
|
|
"learning_rate": 3.513513513513514e-05,
|
|
"loss": 2.3448,
|
|
"step": 8245
|
|
},
|
|
{
|
|
"epoch": 0.825,
|
|
"grad_norm": 0.8639063835144043,
|
|
"learning_rate": 3.503503503503503e-05,
|
|
"loss": 2.3713,
|
|
"step": 8250
|
|
},
|
|
{
|
|
"epoch": 0.8255,
|
|
"grad_norm": 0.9359018206596375,
|
|
"learning_rate": 3.493493493493494e-05,
|
|
"loss": 2.3531,
|
|
"step": 8255
|
|
},
|
|
{
|
|
"epoch": 0.826,
|
|
"grad_norm": 0.8109295964241028,
|
|
"learning_rate": 3.483483483483483e-05,
|
|
"loss": 2.3528,
|
|
"step": 8260
|
|
},
|
|
{
|
|
"epoch": 0.8265,
|
|
"grad_norm": 0.8261359930038452,
|
|
"learning_rate": 3.473473473473473e-05,
|
|
"loss": 2.3829,
|
|
"step": 8265
|
|
},
|
|
{
|
|
"epoch": 0.827,
|
|
"grad_norm": 0.8504651784896851,
|
|
"learning_rate": 3.463463463463464e-05,
|
|
"loss": 2.3537,
|
|
"step": 8270
|
|
},
|
|
{
|
|
"epoch": 0.8275,
|
|
"grad_norm": 0.8876836895942688,
|
|
"learning_rate": 3.453453453453453e-05,
|
|
"loss": 2.3446,
|
|
"step": 8275
|
|
},
|
|
{
|
|
"epoch": 0.828,
|
|
"grad_norm": 0.8673462867736816,
|
|
"learning_rate": 3.443443443443444e-05,
|
|
"loss": 2.3826,
|
|
"step": 8280
|
|
},
|
|
{
|
|
"epoch": 0.8285,
|
|
"grad_norm": 0.8119297027587891,
|
|
"learning_rate": 3.4334334334334334e-05,
|
|
"loss": 2.3621,
|
|
"step": 8285
|
|
},
|
|
{
|
|
"epoch": 0.829,
|
|
"grad_norm": 0.8265256881713867,
|
|
"learning_rate": 3.4234234234234234e-05,
|
|
"loss": 2.3526,
|
|
"step": 8290
|
|
},
|
|
{
|
|
"epoch": 0.8295,
|
|
"grad_norm": 0.7955986857414246,
|
|
"learning_rate": 3.413413413413414e-05,
|
|
"loss": 2.3401,
|
|
"step": 8295
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"grad_norm": 0.8546925783157349,
|
|
"learning_rate": 3.4034034034034034e-05,
|
|
"loss": 2.3677,
|
|
"step": 8300
|
|
},
|
|
{
|
|
"epoch": 0.8305,
|
|
"grad_norm": 0.8466092348098755,
|
|
"learning_rate": 3.3933933933933934e-05,
|
|
"loss": 2.3298,
|
|
"step": 8305
|
|
},
|
|
{
|
|
"epoch": 0.831,
|
|
"grad_norm": 0.8166059851646423,
|
|
"learning_rate": 3.3833833833833834e-05,
|
|
"loss": 2.3388,
|
|
"step": 8310
|
|
},
|
|
{
|
|
"epoch": 0.8315,
|
|
"grad_norm": 0.8029240369796753,
|
|
"learning_rate": 3.3733733733733734e-05,
|
|
"loss": 2.3482,
|
|
"step": 8315
|
|
},
|
|
{
|
|
"epoch": 0.832,
|
|
"grad_norm": 0.8286939263343811,
|
|
"learning_rate": 3.3633633633633635e-05,
|
|
"loss": 2.3453,
|
|
"step": 8320
|
|
},
|
|
{
|
|
"epoch": 0.8325,
|
|
"grad_norm": 0.8886439204216003,
|
|
"learning_rate": 3.3533533533533535e-05,
|
|
"loss": 2.3338,
|
|
"step": 8325
|
|
},
|
|
{
|
|
"epoch": 0.833,
|
|
"grad_norm": 0.841537356376648,
|
|
"learning_rate": 3.3433433433433435e-05,
|
|
"loss": 2.3453,
|
|
"step": 8330
|
|
},
|
|
{
|
|
"epoch": 0.8335,
|
|
"grad_norm": 0.7937127947807312,
|
|
"learning_rate": 3.3333333333333335e-05,
|
|
"loss": 2.3659,
|
|
"step": 8335
|
|
},
|
|
{
|
|
"epoch": 0.834,
|
|
"grad_norm": 0.8088583946228027,
|
|
"learning_rate": 3.3233233233233235e-05,
|
|
"loss": 2.343,
|
|
"step": 8340
|
|
},
|
|
{
|
|
"epoch": 0.8345,
|
|
"grad_norm": 0.8171084523200989,
|
|
"learning_rate": 3.3133133133133135e-05,
|
|
"loss": 2.3551,
|
|
"step": 8345
|
|
},
|
|
{
|
|
"epoch": 0.835,
|
|
"grad_norm": 0.8256644606590271,
|
|
"learning_rate": 3.3033033033033035e-05,
|
|
"loss": 2.3623,
|
|
"step": 8350
|
|
},
|
|
{
|
|
"epoch": 0.8355,
|
|
"grad_norm": 0.8347882032394409,
|
|
"learning_rate": 3.2932932932932935e-05,
|
|
"loss": 2.3551,
|
|
"step": 8355
|
|
},
|
|
{
|
|
"epoch": 0.836,
|
|
"grad_norm": 0.818671464920044,
|
|
"learning_rate": 3.2832832832832836e-05,
|
|
"loss": 2.3321,
|
|
"step": 8360
|
|
},
|
|
{
|
|
"epoch": 0.8365,
|
|
"grad_norm": 0.8162476420402527,
|
|
"learning_rate": 3.2732732732732736e-05,
|
|
"loss": 2.3582,
|
|
"step": 8365
|
|
},
|
|
{
|
|
"epoch": 0.837,
|
|
"grad_norm": 0.8154727220535278,
|
|
"learning_rate": 3.263263263263263e-05,
|
|
"loss": 2.3394,
|
|
"step": 8370
|
|
},
|
|
{
|
|
"epoch": 0.8375,
|
|
"grad_norm": 0.8620063662528992,
|
|
"learning_rate": 3.2532532532532536e-05,
|
|
"loss": 2.321,
|
|
"step": 8375
|
|
},
|
|
{
|
|
"epoch": 0.838,
|
|
"grad_norm": 0.8535669445991516,
|
|
"learning_rate": 3.2432432432432436e-05,
|
|
"loss": 2.337,
|
|
"step": 8380
|
|
},
|
|
{
|
|
"epoch": 0.8385,
|
|
"grad_norm": 0.8434451222419739,
|
|
"learning_rate": 3.233233233233233e-05,
|
|
"loss": 2.3599,
|
|
"step": 8385
|
|
},
|
|
{
|
|
"epoch": 0.839,
|
|
"grad_norm": 0.8340520262718201,
|
|
"learning_rate": 3.2232232232232236e-05,
|
|
"loss": 2.3546,
|
|
"step": 8390
|
|
},
|
|
{
|
|
"epoch": 0.8395,
|
|
"grad_norm": 0.8047171235084534,
|
|
"learning_rate": 3.213213213213213e-05,
|
|
"loss": 2.3571,
|
|
"step": 8395
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"grad_norm": 0.811255156993866,
|
|
"learning_rate": 3.203203203203203e-05,
|
|
"loss": 2.3472,
|
|
"step": 8400
|
|
},
|
|
{
|
|
"epoch": 0.8405,
|
|
"grad_norm": 0.8495455980300903,
|
|
"learning_rate": 3.193193193193194e-05,
|
|
"loss": 2.3573,
|
|
"step": 8405
|
|
},
|
|
{
|
|
"epoch": 0.841,
|
|
"grad_norm": 0.8299034237861633,
|
|
"learning_rate": 3.183183183183183e-05,
|
|
"loss": 2.318,
|
|
"step": 8410
|
|
},
|
|
{
|
|
"epoch": 0.8415,
|
|
"grad_norm": 0.8490227460861206,
|
|
"learning_rate": 3.173173173173174e-05,
|
|
"loss": 2.35,
|
|
"step": 8415
|
|
},
|
|
{
|
|
"epoch": 0.842,
|
|
"grad_norm": 0.8307939171791077,
|
|
"learning_rate": 3.163163163163163e-05,
|
|
"loss": 2.32,
|
|
"step": 8420
|
|
},
|
|
{
|
|
"epoch": 0.8425,
|
|
"grad_norm": 0.8129630088806152,
|
|
"learning_rate": 3.153153153153153e-05,
|
|
"loss": 2.3439,
|
|
"step": 8425
|
|
},
|
|
{
|
|
"epoch": 0.843,
|
|
"grad_norm": 0.8450644016265869,
|
|
"learning_rate": 3.143143143143144e-05,
|
|
"loss": 2.3265,
|
|
"step": 8430
|
|
},
|
|
{
|
|
"epoch": 0.8435,
|
|
"grad_norm": 0.8025068640708923,
|
|
"learning_rate": 3.133133133133133e-05,
|
|
"loss": 2.3517,
|
|
"step": 8435
|
|
},
|
|
{
|
|
"epoch": 0.844,
|
|
"grad_norm": 0.8366633653640747,
|
|
"learning_rate": 3.123123123123123e-05,
|
|
"loss": 2.3484,
|
|
"step": 8440
|
|
},
|
|
{
|
|
"epoch": 0.8445,
|
|
"grad_norm": 0.8080820441246033,
|
|
"learning_rate": 3.113113113113113e-05,
|
|
"loss": 2.3557,
|
|
"step": 8445
|
|
},
|
|
{
|
|
"epoch": 0.845,
|
|
"grad_norm": 0.8732795715332031,
|
|
"learning_rate": 3.103103103103103e-05,
|
|
"loss": 2.3648,
|
|
"step": 8450
|
|
},
|
|
{
|
|
"epoch": 0.8455,
|
|
"grad_norm": 0.8327799439430237,
|
|
"learning_rate": 3.093093093093093e-05,
|
|
"loss": 2.3396,
|
|
"step": 8455
|
|
},
|
|
{
|
|
"epoch": 0.846,
|
|
"grad_norm": 0.7959544062614441,
|
|
"learning_rate": 3.083083083083083e-05,
|
|
"loss": 2.337,
|
|
"step": 8460
|
|
},
|
|
{
|
|
"epoch": 0.8465,
|
|
"grad_norm": 0.8220170140266418,
|
|
"learning_rate": 3.073073073073073e-05,
|
|
"loss": 2.3477,
|
|
"step": 8465
|
|
},
|
|
{
|
|
"epoch": 0.847,
|
|
"grad_norm": 0.8160591721534729,
|
|
"learning_rate": 3.063063063063063e-05,
|
|
"loss": 2.3486,
|
|
"step": 8470
|
|
},
|
|
{
|
|
"epoch": 0.8475,
|
|
"grad_norm": 0.8442284464836121,
|
|
"learning_rate": 3.053053053053053e-05,
|
|
"loss": 2.342,
|
|
"step": 8475
|
|
},
|
|
{
|
|
"epoch": 0.848,
|
|
"grad_norm": 0.8980335593223572,
|
|
"learning_rate": 3.0430430430430436e-05,
|
|
"loss": 2.3571,
|
|
"step": 8480
|
|
},
|
|
{
|
|
"epoch": 0.8485,
|
|
"grad_norm": 0.8186134696006775,
|
|
"learning_rate": 3.0330330330330332e-05,
|
|
"loss": 2.3426,
|
|
"step": 8485
|
|
},
|
|
{
|
|
"epoch": 0.849,
|
|
"grad_norm": 0.8173196315765381,
|
|
"learning_rate": 3.0230230230230232e-05,
|
|
"loss": 2.3667,
|
|
"step": 8490
|
|
},
|
|
{
|
|
"epoch": 0.8495,
|
|
"grad_norm": 0.8508972525596619,
|
|
"learning_rate": 3.013013013013013e-05,
|
|
"loss": 2.3738,
|
|
"step": 8495
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"grad_norm": 0.830033004283905,
|
|
"learning_rate": 3.0030030030030033e-05,
|
|
"loss": 2.3554,
|
|
"step": 8500
|
|
},
|
|
{
|
|
"epoch": 0.8505,
|
|
"grad_norm": 0.8232085704803467,
|
|
"learning_rate": 2.9929929929929933e-05,
|
|
"loss": 2.3392,
|
|
"step": 8505
|
|
},
|
|
{
|
|
"epoch": 0.851,
|
|
"grad_norm": 0.8099601864814758,
|
|
"learning_rate": 2.982982982982983e-05,
|
|
"loss": 2.3517,
|
|
"step": 8510
|
|
},
|
|
{
|
|
"epoch": 0.8515,
|
|
"grad_norm": 0.8356584310531616,
|
|
"learning_rate": 2.9729729729729733e-05,
|
|
"loss": 2.3664,
|
|
"step": 8515
|
|
},
|
|
{
|
|
"epoch": 0.852,
|
|
"grad_norm": 0.8594179153442383,
|
|
"learning_rate": 2.962962962962963e-05,
|
|
"loss": 2.3585,
|
|
"step": 8520
|
|
},
|
|
{
|
|
"epoch": 0.8525,
|
|
"grad_norm": 0.8841564059257507,
|
|
"learning_rate": 2.952952952952953e-05,
|
|
"loss": 2.3322,
|
|
"step": 8525
|
|
},
|
|
{
|
|
"epoch": 0.853,
|
|
"grad_norm": 0.825693666934967,
|
|
"learning_rate": 2.9429429429429427e-05,
|
|
"loss": 2.3583,
|
|
"step": 8530
|
|
},
|
|
{
|
|
"epoch": 0.8535,
|
|
"grad_norm": 0.8264659643173218,
|
|
"learning_rate": 2.932932932932933e-05,
|
|
"loss": 2.3435,
|
|
"step": 8535
|
|
},
|
|
{
|
|
"epoch": 0.854,
|
|
"grad_norm": 0.8185012936592102,
|
|
"learning_rate": 2.9229229229229234e-05,
|
|
"loss": 2.3569,
|
|
"step": 8540
|
|
},
|
|
{
|
|
"epoch": 0.8545,
|
|
"grad_norm": 0.8264207243919373,
|
|
"learning_rate": 2.912912912912913e-05,
|
|
"loss": 2.36,
|
|
"step": 8545
|
|
},
|
|
{
|
|
"epoch": 0.855,
|
|
"grad_norm": 0.8175407648086548,
|
|
"learning_rate": 2.902902902902903e-05,
|
|
"loss": 2.3549,
|
|
"step": 8550
|
|
},
|
|
{
|
|
"epoch": 0.8555,
|
|
"grad_norm": 0.8118422031402588,
|
|
"learning_rate": 2.8928928928928928e-05,
|
|
"loss": 2.3345,
|
|
"step": 8555
|
|
},
|
|
{
|
|
"epoch": 0.856,
|
|
"grad_norm": 0.8219888210296631,
|
|
"learning_rate": 2.882882882882883e-05,
|
|
"loss": 2.3319,
|
|
"step": 8560
|
|
},
|
|
{
|
|
"epoch": 0.8565,
|
|
"grad_norm": 0.868126392364502,
|
|
"learning_rate": 2.872872872872873e-05,
|
|
"loss": 2.3567,
|
|
"step": 8565
|
|
},
|
|
{
|
|
"epoch": 0.857,
|
|
"grad_norm": 0.8113792538642883,
|
|
"learning_rate": 2.8628628628628628e-05,
|
|
"loss": 2.3326,
|
|
"step": 8570
|
|
},
|
|
{
|
|
"epoch": 0.8575,
|
|
"grad_norm": 0.8458234667778015,
|
|
"learning_rate": 2.852852852852853e-05,
|
|
"loss": 2.3452,
|
|
"step": 8575
|
|
},
|
|
{
|
|
"epoch": 0.858,
|
|
"grad_norm": 0.8528864979743958,
|
|
"learning_rate": 2.8428428428428428e-05,
|
|
"loss": 2.3343,
|
|
"step": 8580
|
|
},
|
|
{
|
|
"epoch": 0.8585,
|
|
"grad_norm": 0.8229609131813049,
|
|
"learning_rate": 2.832832832832833e-05,
|
|
"loss": 2.3349,
|
|
"step": 8585
|
|
},
|
|
{
|
|
"epoch": 0.859,
|
|
"grad_norm": 0.8314008712768555,
|
|
"learning_rate": 2.8228228228228232e-05,
|
|
"loss": 2.3489,
|
|
"step": 8590
|
|
},
|
|
{
|
|
"epoch": 0.8595,
|
|
"grad_norm": 0.8441357016563416,
|
|
"learning_rate": 2.812812812812813e-05,
|
|
"loss": 2.3424,
|
|
"step": 8595
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"grad_norm": 0.811282753944397,
|
|
"learning_rate": 2.8028028028028032e-05,
|
|
"loss": 2.3385,
|
|
"step": 8600
|
|
},
|
|
{
|
|
"epoch": 0.8605,
|
|
"grad_norm": 0.8737632036209106,
|
|
"learning_rate": 2.7927927927927926e-05,
|
|
"loss": 2.3374,
|
|
"step": 8605
|
|
},
|
|
{
|
|
"epoch": 0.861,
|
|
"grad_norm": 0.8345777988433838,
|
|
"learning_rate": 2.782782782782783e-05,
|
|
"loss": 2.3392,
|
|
"step": 8610
|
|
},
|
|
{
|
|
"epoch": 0.8615,
|
|
"grad_norm": 0.8158997297286987,
|
|
"learning_rate": 2.7727727727727733e-05,
|
|
"loss": 2.33,
|
|
"step": 8615
|
|
},
|
|
{
|
|
"epoch": 0.862,
|
|
"grad_norm": 0.8565182685852051,
|
|
"learning_rate": 2.762762762762763e-05,
|
|
"loss": 2.3582,
|
|
"step": 8620
|
|
},
|
|
{
|
|
"epoch": 0.8625,
|
|
"grad_norm": 0.8362213373184204,
|
|
"learning_rate": 2.752752752752753e-05,
|
|
"loss": 2.3434,
|
|
"step": 8625
|
|
},
|
|
{
|
|
"epoch": 0.863,
|
|
"grad_norm": 0.8124709725379944,
|
|
"learning_rate": 2.7427427427427426e-05,
|
|
"loss": 2.3546,
|
|
"step": 8630
|
|
},
|
|
{
|
|
"epoch": 0.8635,
|
|
"grad_norm": 0.8765237927436829,
|
|
"learning_rate": 2.732732732732733e-05,
|
|
"loss": 2.3503,
|
|
"step": 8635
|
|
},
|
|
{
|
|
"epoch": 0.864,
|
|
"grad_norm": 0.8562377095222473,
|
|
"learning_rate": 2.722722722722723e-05,
|
|
"loss": 2.3635,
|
|
"step": 8640
|
|
},
|
|
{
|
|
"epoch": 0.8645,
|
|
"grad_norm": 0.862598180770874,
|
|
"learning_rate": 2.7127127127127127e-05,
|
|
"loss": 2.3507,
|
|
"step": 8645
|
|
},
|
|
{
|
|
"epoch": 0.865,
|
|
"grad_norm": 0.8522788286209106,
|
|
"learning_rate": 2.702702702702703e-05,
|
|
"loss": 2.3443,
|
|
"step": 8650
|
|
},
|
|
{
|
|
"epoch": 0.8655,
|
|
"grad_norm": 0.8473895192146301,
|
|
"learning_rate": 2.6926926926926927e-05,
|
|
"loss": 2.3399,
|
|
"step": 8655
|
|
},
|
|
{
|
|
"epoch": 0.866,
|
|
"grad_norm": 0.8129273056983948,
|
|
"learning_rate": 2.6826826826826827e-05,
|
|
"loss": 2.3579,
|
|
"step": 8660
|
|
},
|
|
{
|
|
"epoch": 0.8665,
|
|
"grad_norm": 0.8594850897789001,
|
|
"learning_rate": 2.672672672672673e-05,
|
|
"loss": 2.3539,
|
|
"step": 8665
|
|
},
|
|
{
|
|
"epoch": 0.867,
|
|
"grad_norm": 0.831794798374176,
|
|
"learning_rate": 2.6626626626626627e-05,
|
|
"loss": 2.3342,
|
|
"step": 8670
|
|
},
|
|
{
|
|
"epoch": 0.8675,
|
|
"grad_norm": 0.8447133302688599,
|
|
"learning_rate": 2.652652652652653e-05,
|
|
"loss": 2.3403,
|
|
"step": 8675
|
|
},
|
|
{
|
|
"epoch": 0.868,
|
|
"grad_norm": 0.8298300504684448,
|
|
"learning_rate": 2.6426426426426428e-05,
|
|
"loss": 2.3362,
|
|
"step": 8680
|
|
},
|
|
{
|
|
"epoch": 0.8685,
|
|
"grad_norm": 0.8596557974815369,
|
|
"learning_rate": 2.6326326326326328e-05,
|
|
"loss": 2.3545,
|
|
"step": 8685
|
|
},
|
|
{
|
|
"epoch": 0.869,
|
|
"grad_norm": 0.8677126169204712,
|
|
"learning_rate": 2.6226226226226224e-05,
|
|
"loss": 2.3624,
|
|
"step": 8690
|
|
},
|
|
{
|
|
"epoch": 0.8695,
|
|
"grad_norm": 0.8112132549285889,
|
|
"learning_rate": 2.6126126126126128e-05,
|
|
"loss": 2.3345,
|
|
"step": 8695
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"grad_norm": 0.87315833568573,
|
|
"learning_rate": 2.6026026026026028e-05,
|
|
"loss": 2.3319,
|
|
"step": 8700
|
|
},
|
|
{
|
|
"epoch": 0.8705,
|
|
"grad_norm": 0.864588737487793,
|
|
"learning_rate": 2.5925925925925925e-05,
|
|
"loss": 2.3356,
|
|
"step": 8705
|
|
},
|
|
{
|
|
"epoch": 0.871,
|
|
"grad_norm": 0.8347481489181519,
|
|
"learning_rate": 2.582582582582583e-05,
|
|
"loss": 2.32,
|
|
"step": 8710
|
|
},
|
|
{
|
|
"epoch": 0.8715,
|
|
"grad_norm": 0.8150578141212463,
|
|
"learning_rate": 2.5725725725725725e-05,
|
|
"loss": 2.3403,
|
|
"step": 8715
|
|
},
|
|
{
|
|
"epoch": 0.872,
|
|
"grad_norm": 0.8222635388374329,
|
|
"learning_rate": 2.5625625625625625e-05,
|
|
"loss": 2.3328,
|
|
"step": 8720
|
|
},
|
|
{
|
|
"epoch": 0.8725,
|
|
"grad_norm": 0.8109158277511597,
|
|
"learning_rate": 2.552552552552553e-05,
|
|
"loss": 2.3558,
|
|
"step": 8725
|
|
},
|
|
{
|
|
"epoch": 0.873,
|
|
"grad_norm": 0.8342311382293701,
|
|
"learning_rate": 2.5425425425425426e-05,
|
|
"loss": 2.337,
|
|
"step": 8730
|
|
},
|
|
{
|
|
"epoch": 0.8735,
|
|
"grad_norm": 0.8757238388061523,
|
|
"learning_rate": 2.532532532532533e-05,
|
|
"loss": 2.3525,
|
|
"step": 8735
|
|
},
|
|
{
|
|
"epoch": 0.874,
|
|
"grad_norm": 0.822337806224823,
|
|
"learning_rate": 2.5225225225225222e-05,
|
|
"loss": 2.3324,
|
|
"step": 8740
|
|
},
|
|
{
|
|
"epoch": 0.8745,
|
|
"grad_norm": 0.7943927645683289,
|
|
"learning_rate": 2.5125125125125126e-05,
|
|
"loss": 2.3521,
|
|
"step": 8745
|
|
},
|
|
{
|
|
"epoch": 0.875,
|
|
"grad_norm": 0.8191083073616028,
|
|
"learning_rate": 2.502502502502503e-05,
|
|
"loss": 2.3585,
|
|
"step": 8750
|
|
},
|
|
{
|
|
"epoch": 0.8755,
|
|
"grad_norm": 0.8550540208816528,
|
|
"learning_rate": 2.4924924924924926e-05,
|
|
"loss": 2.3426,
|
|
"step": 8755
|
|
},
|
|
{
|
|
"epoch": 0.876,
|
|
"grad_norm": 0.8050912618637085,
|
|
"learning_rate": 2.4824824824824826e-05,
|
|
"loss": 2.3643,
|
|
"step": 8760
|
|
},
|
|
{
|
|
"epoch": 0.8765,
|
|
"grad_norm": 0.8559862971305847,
|
|
"learning_rate": 2.4724724724724727e-05,
|
|
"loss": 2.3606,
|
|
"step": 8765
|
|
},
|
|
{
|
|
"epoch": 0.877,
|
|
"grad_norm": 0.8613905310630798,
|
|
"learning_rate": 2.4624624624624627e-05,
|
|
"loss": 2.3494,
|
|
"step": 8770
|
|
},
|
|
{
|
|
"epoch": 0.8775,
|
|
"grad_norm": 0.8286505937576294,
|
|
"learning_rate": 2.4524524524524527e-05,
|
|
"loss": 2.3506,
|
|
"step": 8775
|
|
},
|
|
{
|
|
"epoch": 0.878,
|
|
"grad_norm": 0.8565585613250732,
|
|
"learning_rate": 2.4424424424424424e-05,
|
|
"loss": 2.3414,
|
|
"step": 8780
|
|
},
|
|
{
|
|
"epoch": 0.8785,
|
|
"grad_norm": 0.8649365305900574,
|
|
"learning_rate": 2.4324324324324327e-05,
|
|
"loss": 2.3748,
|
|
"step": 8785
|
|
},
|
|
{
|
|
"epoch": 0.879,
|
|
"grad_norm": 0.8483036756515503,
|
|
"learning_rate": 2.4224224224224227e-05,
|
|
"loss": 2.3398,
|
|
"step": 8790
|
|
},
|
|
{
|
|
"epoch": 0.8795,
|
|
"grad_norm": 0.7956961989402771,
|
|
"learning_rate": 2.4124124124124124e-05,
|
|
"loss": 2.3343,
|
|
"step": 8795
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"grad_norm": 0.841417670249939,
|
|
"learning_rate": 2.4024024024024024e-05,
|
|
"loss": 2.3344,
|
|
"step": 8800
|
|
},
|
|
{
|
|
"epoch": 0.8805,
|
|
"grad_norm": 0.8638480305671692,
|
|
"learning_rate": 2.3923923923923924e-05,
|
|
"loss": 2.3522,
|
|
"step": 8805
|
|
},
|
|
{
|
|
"epoch": 0.881,
|
|
"grad_norm": 0.8249894380569458,
|
|
"learning_rate": 2.3823823823823824e-05,
|
|
"loss": 2.3368,
|
|
"step": 8810
|
|
},
|
|
{
|
|
"epoch": 0.8815,
|
|
"grad_norm": 0.8154436945915222,
|
|
"learning_rate": 2.3723723723723725e-05,
|
|
"loss": 2.3265,
|
|
"step": 8815
|
|
},
|
|
{
|
|
"epoch": 0.882,
|
|
"grad_norm": 0.8674147725105286,
|
|
"learning_rate": 2.3623623623623625e-05,
|
|
"loss": 2.3581,
|
|
"step": 8820
|
|
},
|
|
{
|
|
"epoch": 0.8825,
|
|
"grad_norm": 0.8114337921142578,
|
|
"learning_rate": 2.3523523523523525e-05,
|
|
"loss": 2.3431,
|
|
"step": 8825
|
|
},
|
|
{
|
|
"epoch": 0.883,
|
|
"grad_norm": 0.8255671262741089,
|
|
"learning_rate": 2.3423423423423425e-05,
|
|
"loss": 2.3427,
|
|
"step": 8830
|
|
},
|
|
{
|
|
"epoch": 0.8835,
|
|
"grad_norm": 0.8699901700019836,
|
|
"learning_rate": 2.3323323323323322e-05,
|
|
"loss": 2.342,
|
|
"step": 8835
|
|
},
|
|
{
|
|
"epoch": 0.884,
|
|
"grad_norm": 0.8162307143211365,
|
|
"learning_rate": 2.3223223223223225e-05,
|
|
"loss": 2.3453,
|
|
"step": 8840
|
|
},
|
|
{
|
|
"epoch": 0.8845,
|
|
"grad_norm": 0.8098514676094055,
|
|
"learning_rate": 2.3123123123123125e-05,
|
|
"loss": 2.3313,
|
|
"step": 8845
|
|
},
|
|
{
|
|
"epoch": 0.885,
|
|
"grad_norm": 0.8761820793151855,
|
|
"learning_rate": 2.3023023023023026e-05,
|
|
"loss": 2.367,
|
|
"step": 8850
|
|
},
|
|
{
|
|
"epoch": 0.8855,
|
|
"grad_norm": 0.8436310887336731,
|
|
"learning_rate": 2.2922922922922922e-05,
|
|
"loss": 2.3561,
|
|
"step": 8855
|
|
},
|
|
{
|
|
"epoch": 0.886,
|
|
"grad_norm": 0.8526544570922852,
|
|
"learning_rate": 2.2822822822822822e-05,
|
|
"loss": 2.3259,
|
|
"step": 8860
|
|
},
|
|
{
|
|
"epoch": 0.8865,
|
|
"grad_norm": 0.8305936455726624,
|
|
"learning_rate": 2.2722722722722726e-05,
|
|
"loss": 2.3373,
|
|
"step": 8865
|
|
},
|
|
{
|
|
"epoch": 0.887,
|
|
"grad_norm": 0.9209138751029968,
|
|
"learning_rate": 2.2622622622622626e-05,
|
|
"loss": 2.3439,
|
|
"step": 8870
|
|
},
|
|
{
|
|
"epoch": 0.8875,
|
|
"grad_norm": 0.8664165735244751,
|
|
"learning_rate": 2.2522522522522523e-05,
|
|
"loss": 2.3355,
|
|
"step": 8875
|
|
},
|
|
{
|
|
"epoch": 0.888,
|
|
"grad_norm": 0.8305662870407104,
|
|
"learning_rate": 2.2422422422422423e-05,
|
|
"loss": 2.3497,
|
|
"step": 8880
|
|
},
|
|
{
|
|
"epoch": 0.8885,
|
|
"grad_norm": 0.8392084836959839,
|
|
"learning_rate": 2.2322322322322323e-05,
|
|
"loss": 2.3306,
|
|
"step": 8885
|
|
},
|
|
{
|
|
"epoch": 0.889,
|
|
"grad_norm": 0.8184075355529785,
|
|
"learning_rate": 2.2222222222222223e-05,
|
|
"loss": 2.3534,
|
|
"step": 8890
|
|
},
|
|
{
|
|
"epoch": 0.8895,
|
|
"grad_norm": 0.8275293111801147,
|
|
"learning_rate": 2.2122122122122123e-05,
|
|
"loss": 2.345,
|
|
"step": 8895
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"grad_norm": 0.8183415532112122,
|
|
"learning_rate": 2.2022022022022024e-05,
|
|
"loss": 2.3524,
|
|
"step": 8900
|
|
},
|
|
{
|
|
"epoch": 0.8905,
|
|
"grad_norm": 0.8345757126808167,
|
|
"learning_rate": 2.1921921921921924e-05,
|
|
"loss": 2.3222,
|
|
"step": 8905
|
|
},
|
|
{
|
|
"epoch": 0.891,
|
|
"grad_norm": 0.8104347586631775,
|
|
"learning_rate": 2.1821821821821824e-05,
|
|
"loss": 2.3391,
|
|
"step": 8910
|
|
},
|
|
{
|
|
"epoch": 0.8915,
|
|
"grad_norm": 0.81838458776474,
|
|
"learning_rate": 2.172172172172172e-05,
|
|
"loss": 2.365,
|
|
"step": 8915
|
|
},
|
|
{
|
|
"epoch": 0.892,
|
|
"grad_norm": 0.8337346911430359,
|
|
"learning_rate": 2.1621621621621624e-05,
|
|
"loss": 2.3428,
|
|
"step": 8920
|
|
},
|
|
{
|
|
"epoch": 0.8925,
|
|
"grad_norm": 0.8469566702842712,
|
|
"learning_rate": 2.1521521521521524e-05,
|
|
"loss": 2.3555,
|
|
"step": 8925
|
|
},
|
|
{
|
|
"epoch": 0.893,
|
|
"grad_norm": 0.8030543923377991,
|
|
"learning_rate": 2.142142142142142e-05,
|
|
"loss": 2.3126,
|
|
"step": 8930
|
|
},
|
|
{
|
|
"epoch": 0.8935,
|
|
"grad_norm": 0.837141752243042,
|
|
"learning_rate": 2.132132132132132e-05,
|
|
"loss": 2.319,
|
|
"step": 8935
|
|
},
|
|
{
|
|
"epoch": 0.894,
|
|
"grad_norm": 0.8511247634887695,
|
|
"learning_rate": 2.122122122122122e-05,
|
|
"loss": 2.3407,
|
|
"step": 8940
|
|
},
|
|
{
|
|
"epoch": 0.8945,
|
|
"grad_norm": 0.8452726006507874,
|
|
"learning_rate": 2.1121121121121125e-05,
|
|
"loss": 2.3483,
|
|
"step": 8945
|
|
},
|
|
{
|
|
"epoch": 0.895,
|
|
"grad_norm": 0.8432291746139526,
|
|
"learning_rate": 2.102102102102102e-05,
|
|
"loss": 2.3493,
|
|
"step": 8950
|
|
},
|
|
{
|
|
"epoch": 0.8955,
|
|
"grad_norm": 0.8280351161956787,
|
|
"learning_rate": 2.0920920920920922e-05,
|
|
"loss": 2.3346,
|
|
"step": 8955
|
|
},
|
|
{
|
|
"epoch": 0.896,
|
|
"grad_norm": 0.8203386664390564,
|
|
"learning_rate": 2.0820820820820822e-05,
|
|
"loss": 2.3467,
|
|
"step": 8960
|
|
},
|
|
{
|
|
"epoch": 0.8965,
|
|
"grad_norm": 0.8204079270362854,
|
|
"learning_rate": 2.0720720720720722e-05,
|
|
"loss": 2.356,
|
|
"step": 8965
|
|
},
|
|
{
|
|
"epoch": 0.897,
|
|
"grad_norm": 0.813487708568573,
|
|
"learning_rate": 2.062062062062062e-05,
|
|
"loss": 2.3215,
|
|
"step": 8970
|
|
},
|
|
{
|
|
"epoch": 0.8975,
|
|
"grad_norm": 0.8271229863166809,
|
|
"learning_rate": 2.0520520520520522e-05,
|
|
"loss": 2.3555,
|
|
"step": 8975
|
|
},
|
|
{
|
|
"epoch": 0.898,
|
|
"grad_norm": 0.8487913608551025,
|
|
"learning_rate": 2.0420420420420422e-05,
|
|
"loss": 2.3359,
|
|
"step": 8980
|
|
},
|
|
{
|
|
"epoch": 0.8985,
|
|
"grad_norm": 0.836768627166748,
|
|
"learning_rate": 2.0320320320320323e-05,
|
|
"loss": 2.3498,
|
|
"step": 8985
|
|
},
|
|
{
|
|
"epoch": 0.899,
|
|
"grad_norm": 0.8347607254981995,
|
|
"learning_rate": 2.022022022022022e-05,
|
|
"loss": 2.317,
|
|
"step": 8990
|
|
},
|
|
{
|
|
"epoch": 0.8995,
|
|
"grad_norm": 0.8265148401260376,
|
|
"learning_rate": 2.012012012012012e-05,
|
|
"loss": 2.3063,
|
|
"step": 8995
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"grad_norm": 0.8346414566040039,
|
|
"learning_rate": 2.0020020020020023e-05,
|
|
"loss": 2.3368,
|
|
"step": 9000
|
|
},
|
|
{
|
|
"epoch": 0.9005,
|
|
"grad_norm": 0.8407677412033081,
|
|
"learning_rate": 1.9919919919919923e-05,
|
|
"loss": 2.3595,
|
|
"step": 9005
|
|
},
|
|
{
|
|
"epoch": 0.901,
|
|
"grad_norm": 0.8652437925338745,
|
|
"learning_rate": 1.981981981981982e-05,
|
|
"loss": 2.3462,
|
|
"step": 9010
|
|
},
|
|
{
|
|
"epoch": 0.9015,
|
|
"grad_norm": 0.851755678653717,
|
|
"learning_rate": 1.971971971971972e-05,
|
|
"loss": 2.343,
|
|
"step": 9015
|
|
},
|
|
{
|
|
"epoch": 0.902,
|
|
"grad_norm": 0.844289243221283,
|
|
"learning_rate": 1.961961961961962e-05,
|
|
"loss": 2.3494,
|
|
"step": 9020
|
|
},
|
|
{
|
|
"epoch": 0.9025,
|
|
"grad_norm": 0.8320977687835693,
|
|
"learning_rate": 1.951951951951952e-05,
|
|
"loss": 2.3311,
|
|
"step": 9025
|
|
},
|
|
{
|
|
"epoch": 0.903,
|
|
"grad_norm": 0.8191105127334595,
|
|
"learning_rate": 1.941941941941942e-05,
|
|
"loss": 2.3323,
|
|
"step": 9030
|
|
},
|
|
{
|
|
"epoch": 0.9035,
|
|
"grad_norm": 0.8393972516059875,
|
|
"learning_rate": 1.931931931931932e-05,
|
|
"loss": 2.3409,
|
|
"step": 9035
|
|
},
|
|
{
|
|
"epoch": 0.904,
|
|
"grad_norm": 0.8605306148529053,
|
|
"learning_rate": 1.921921921921922e-05,
|
|
"loss": 2.3294,
|
|
"step": 9040
|
|
},
|
|
{
|
|
"epoch": 0.9045,
|
|
"grad_norm": 0.8271490335464478,
|
|
"learning_rate": 1.911911911911912e-05,
|
|
"loss": 2.3292,
|
|
"step": 9045
|
|
},
|
|
{
|
|
"epoch": 0.905,
|
|
"grad_norm": 0.8311194777488708,
|
|
"learning_rate": 1.9019019019019018e-05,
|
|
"loss": 2.3573,
|
|
"step": 9050
|
|
},
|
|
{
|
|
"epoch": 0.9055,
|
|
"grad_norm": 0.8212003111839294,
|
|
"learning_rate": 1.891891891891892e-05,
|
|
"loss": 2.3319,
|
|
"step": 9055
|
|
},
|
|
{
|
|
"epoch": 0.906,
|
|
"grad_norm": 0.8189734816551208,
|
|
"learning_rate": 1.881881881881882e-05,
|
|
"loss": 2.3408,
|
|
"step": 9060
|
|
},
|
|
{
|
|
"epoch": 0.9065,
|
|
"grad_norm": 0.8122197389602661,
|
|
"learning_rate": 1.8718718718718718e-05,
|
|
"loss": 2.3191,
|
|
"step": 9065
|
|
},
|
|
{
|
|
"epoch": 0.907,
|
|
"grad_norm": 0.8065931797027588,
|
|
"learning_rate": 1.8618618618618618e-05,
|
|
"loss": 2.3578,
|
|
"step": 9070
|
|
},
|
|
{
|
|
"epoch": 0.9075,
|
|
"grad_norm": 0.8327513337135315,
|
|
"learning_rate": 1.8518518518518518e-05,
|
|
"loss": 2.3714,
|
|
"step": 9075
|
|
},
|
|
{
|
|
"epoch": 0.908,
|
|
"grad_norm": 0.9134345650672913,
|
|
"learning_rate": 1.8418418418418422e-05,
|
|
"loss": 2.3476,
|
|
"step": 9080
|
|
},
|
|
{
|
|
"epoch": 0.9085,
|
|
"grad_norm": 0.8555792570114136,
|
|
"learning_rate": 1.831831831831832e-05,
|
|
"loss": 2.3617,
|
|
"step": 9085
|
|
},
|
|
{
|
|
"epoch": 0.909,
|
|
"grad_norm": 0.839600682258606,
|
|
"learning_rate": 1.821821821821822e-05,
|
|
"loss": 2.3562,
|
|
"step": 9090
|
|
},
|
|
{
|
|
"epoch": 0.9095,
|
|
"grad_norm": 0.8846326470375061,
|
|
"learning_rate": 1.811811811811812e-05,
|
|
"loss": 2.3693,
|
|
"step": 9095
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"grad_norm": 0.8199856281280518,
|
|
"learning_rate": 1.801801801801802e-05,
|
|
"loss": 2.3427,
|
|
"step": 9100
|
|
},
|
|
{
|
|
"epoch": 0.9105,
|
|
"grad_norm": 0.8347171545028687,
|
|
"learning_rate": 1.7917917917917916e-05,
|
|
"loss": 2.3227,
|
|
"step": 9105
|
|
},
|
|
{
|
|
"epoch": 0.911,
|
|
"grad_norm": 0.8130427002906799,
|
|
"learning_rate": 1.781781781781782e-05,
|
|
"loss": 2.3661,
|
|
"step": 9110
|
|
},
|
|
{
|
|
"epoch": 0.9115,
|
|
"grad_norm": 0.8870020508766174,
|
|
"learning_rate": 1.771771771771772e-05,
|
|
"loss": 2.3422,
|
|
"step": 9115
|
|
},
|
|
{
|
|
"epoch": 0.912,
|
|
"grad_norm": 0.8388358354568481,
|
|
"learning_rate": 1.761761761761762e-05,
|
|
"loss": 2.3496,
|
|
"step": 9120
|
|
},
|
|
{
|
|
"epoch": 0.9125,
|
|
"grad_norm": 0.8320288062095642,
|
|
"learning_rate": 1.7517517517517516e-05,
|
|
"loss": 2.347,
|
|
"step": 9125
|
|
},
|
|
{
|
|
"epoch": 0.913,
|
|
"grad_norm": 0.8106929659843445,
|
|
"learning_rate": 1.7417417417417416e-05,
|
|
"loss": 2.3521,
|
|
"step": 9130
|
|
},
|
|
{
|
|
"epoch": 0.9135,
|
|
"grad_norm": 0.8142067790031433,
|
|
"learning_rate": 1.731731731731732e-05,
|
|
"loss": 2.3487,
|
|
"step": 9135
|
|
},
|
|
{
|
|
"epoch": 0.914,
|
|
"grad_norm": 0.8240883350372314,
|
|
"learning_rate": 1.721721721721722e-05,
|
|
"loss": 2.3328,
|
|
"step": 9140
|
|
},
|
|
{
|
|
"epoch": 0.9145,
|
|
"grad_norm": 0.8195868730545044,
|
|
"learning_rate": 1.7117117117117117e-05,
|
|
"loss": 2.3483,
|
|
"step": 9145
|
|
},
|
|
{
|
|
"epoch": 0.915,
|
|
"grad_norm": 0.847511887550354,
|
|
"learning_rate": 1.7017017017017017e-05,
|
|
"loss": 2.3457,
|
|
"step": 9150
|
|
},
|
|
{
|
|
"epoch": 0.9155,
|
|
"grad_norm": 0.8567244410514832,
|
|
"learning_rate": 1.6916916916916917e-05,
|
|
"loss": 2.3464,
|
|
"step": 9155
|
|
},
|
|
{
|
|
"epoch": 0.916,
|
|
"grad_norm": 0.8341563940048218,
|
|
"learning_rate": 1.6816816816816817e-05,
|
|
"loss": 2.3438,
|
|
"step": 9160
|
|
},
|
|
{
|
|
"epoch": 0.9165,
|
|
"grad_norm": 0.8154423236846924,
|
|
"learning_rate": 1.6716716716716717e-05,
|
|
"loss": 2.348,
|
|
"step": 9165
|
|
},
|
|
{
|
|
"epoch": 0.917,
|
|
"grad_norm": 0.7971596717834473,
|
|
"learning_rate": 1.6616616616616618e-05,
|
|
"loss": 2.3397,
|
|
"step": 9170
|
|
},
|
|
{
|
|
"epoch": 0.9175,
|
|
"grad_norm": 0.8392190337181091,
|
|
"learning_rate": 1.6516516516516518e-05,
|
|
"loss": 2.3508,
|
|
"step": 9175
|
|
},
|
|
{
|
|
"epoch": 0.918,
|
|
"grad_norm": 0.8365655541419983,
|
|
"learning_rate": 1.6416416416416418e-05,
|
|
"loss": 2.352,
|
|
"step": 9180
|
|
},
|
|
{
|
|
"epoch": 0.9185,
|
|
"grad_norm": 0.8264891505241394,
|
|
"learning_rate": 1.6316316316316315e-05,
|
|
"loss": 2.3542,
|
|
"step": 9185
|
|
},
|
|
{
|
|
"epoch": 0.919,
|
|
"grad_norm": 0.810056746006012,
|
|
"learning_rate": 1.6216216216216218e-05,
|
|
"loss": 2.2953,
|
|
"step": 9190
|
|
},
|
|
{
|
|
"epoch": 0.9195,
|
|
"grad_norm": 0.850010097026825,
|
|
"learning_rate": 1.6116116116116118e-05,
|
|
"loss": 2.3608,
|
|
"step": 9195
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"grad_norm": 0.8388724327087402,
|
|
"learning_rate": 1.6016016016016015e-05,
|
|
"loss": 2.3332,
|
|
"step": 9200
|
|
},
|
|
{
|
|
"epoch": 0.9205,
|
|
"grad_norm": 0.814441442489624,
|
|
"learning_rate": 1.5915915915915915e-05,
|
|
"loss": 2.3209,
|
|
"step": 9205
|
|
},
|
|
{
|
|
"epoch": 0.921,
|
|
"grad_norm": 0.8317567706108093,
|
|
"learning_rate": 1.5815815815815815e-05,
|
|
"loss": 2.367,
|
|
"step": 9210
|
|
},
|
|
{
|
|
"epoch": 0.9215,
|
|
"grad_norm": 0.806183934211731,
|
|
"learning_rate": 1.571571571571572e-05,
|
|
"loss": 2.3551,
|
|
"step": 9215
|
|
},
|
|
{
|
|
"epoch": 0.922,
|
|
"grad_norm": 0.7914506793022156,
|
|
"learning_rate": 1.5615615615615616e-05,
|
|
"loss": 2.3316,
|
|
"step": 9220
|
|
},
|
|
{
|
|
"epoch": 0.9225,
|
|
"grad_norm": 0.8459283709526062,
|
|
"learning_rate": 1.5515515515515516e-05,
|
|
"loss": 2.348,
|
|
"step": 9225
|
|
},
|
|
{
|
|
"epoch": 0.923,
|
|
"grad_norm": 0.8252323269844055,
|
|
"learning_rate": 1.5415415415415416e-05,
|
|
"loss": 2.3233,
|
|
"step": 9230
|
|
},
|
|
{
|
|
"epoch": 0.9235,
|
|
"grad_norm": 0.8866086006164551,
|
|
"learning_rate": 1.5315315315315316e-05,
|
|
"loss": 2.3458,
|
|
"step": 9235
|
|
},
|
|
{
|
|
"epoch": 0.924,
|
|
"grad_norm": 0.8268254399299622,
|
|
"learning_rate": 1.5215215215215218e-05,
|
|
"loss": 2.3363,
|
|
"step": 9240
|
|
},
|
|
{
|
|
"epoch": 0.9245,
|
|
"grad_norm": 0.8354853391647339,
|
|
"learning_rate": 1.5115115115115116e-05,
|
|
"loss": 2.3398,
|
|
"step": 9245
|
|
},
|
|
{
|
|
"epoch": 0.925,
|
|
"grad_norm": 0.851127028465271,
|
|
"learning_rate": 1.5015015015015016e-05,
|
|
"loss": 2.3592,
|
|
"step": 9250
|
|
},
|
|
{
|
|
"epoch": 0.9255,
|
|
"grad_norm": 0.8615062236785889,
|
|
"learning_rate": 1.4914914914914915e-05,
|
|
"loss": 2.3281,
|
|
"step": 9255
|
|
},
|
|
{
|
|
"epoch": 0.926,
|
|
"grad_norm": 0.8572158813476562,
|
|
"learning_rate": 1.4814814814814815e-05,
|
|
"loss": 2.3355,
|
|
"step": 9260
|
|
},
|
|
{
|
|
"epoch": 0.9265,
|
|
"grad_norm": 0.8364690542221069,
|
|
"learning_rate": 1.4714714714714713e-05,
|
|
"loss": 2.3378,
|
|
"step": 9265
|
|
},
|
|
{
|
|
"epoch": 0.927,
|
|
"grad_norm": 0.8171463012695312,
|
|
"learning_rate": 1.4614614614614617e-05,
|
|
"loss": 2.3292,
|
|
"step": 9270
|
|
},
|
|
{
|
|
"epoch": 0.9275,
|
|
"grad_norm": 0.8491124510765076,
|
|
"learning_rate": 1.4514514514514515e-05,
|
|
"loss": 2.3279,
|
|
"step": 9275
|
|
},
|
|
{
|
|
"epoch": 0.928,
|
|
"grad_norm": 0.9891765117645264,
|
|
"learning_rate": 1.4414414414414416e-05,
|
|
"loss": 2.361,
|
|
"step": 9280
|
|
},
|
|
{
|
|
"epoch": 0.9285,
|
|
"grad_norm": 0.8450509905815125,
|
|
"learning_rate": 1.4314314314314314e-05,
|
|
"loss": 2.3326,
|
|
"step": 9285
|
|
},
|
|
{
|
|
"epoch": 0.929,
|
|
"grad_norm": 0.8077675700187683,
|
|
"learning_rate": 1.4214214214214214e-05,
|
|
"loss": 2.3386,
|
|
"step": 9290
|
|
},
|
|
{
|
|
"epoch": 0.9295,
|
|
"grad_norm": 0.8057301640510559,
|
|
"learning_rate": 1.4114114114114116e-05,
|
|
"loss": 2.3447,
|
|
"step": 9295
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"grad_norm": 0.8298118114471436,
|
|
"learning_rate": 1.4014014014014016e-05,
|
|
"loss": 2.3597,
|
|
"step": 9300
|
|
},
|
|
{
|
|
"epoch": 0.9305,
|
|
"grad_norm": 0.8423778414726257,
|
|
"learning_rate": 1.3913913913913915e-05,
|
|
"loss": 2.3413,
|
|
"step": 9305
|
|
},
|
|
{
|
|
"epoch": 0.931,
|
|
"grad_norm": 0.8156774044036865,
|
|
"learning_rate": 1.3813813813813815e-05,
|
|
"loss": 2.3265,
|
|
"step": 9310
|
|
},
|
|
{
|
|
"epoch": 0.9315,
|
|
"grad_norm": 0.8573948740959167,
|
|
"learning_rate": 1.3713713713713713e-05,
|
|
"loss": 2.3454,
|
|
"step": 9315
|
|
},
|
|
{
|
|
"epoch": 0.932,
|
|
"grad_norm": 0.8804399371147156,
|
|
"learning_rate": 1.3613613613613615e-05,
|
|
"loss": 2.3403,
|
|
"step": 9320
|
|
},
|
|
{
|
|
"epoch": 0.9325,
|
|
"grad_norm": 0.8353649377822876,
|
|
"learning_rate": 1.3513513513513515e-05,
|
|
"loss": 2.3485,
|
|
"step": 9325
|
|
},
|
|
{
|
|
"epoch": 0.933,
|
|
"grad_norm": 0.846034824848175,
|
|
"learning_rate": 1.3413413413413414e-05,
|
|
"loss": 2.3512,
|
|
"step": 9330
|
|
},
|
|
{
|
|
"epoch": 0.9335,
|
|
"grad_norm": 0.8679665327072144,
|
|
"learning_rate": 1.3313313313313314e-05,
|
|
"loss": 2.3391,
|
|
"step": 9335
|
|
},
|
|
{
|
|
"epoch": 0.934,
|
|
"grad_norm": 0.8045667409896851,
|
|
"learning_rate": 1.3213213213213214e-05,
|
|
"loss": 2.3444,
|
|
"step": 9340
|
|
},
|
|
{
|
|
"epoch": 0.9345,
|
|
"grad_norm": 0.8405826687812805,
|
|
"learning_rate": 1.3113113113113112e-05,
|
|
"loss": 2.3435,
|
|
"step": 9345
|
|
},
|
|
{
|
|
"epoch": 0.935,
|
|
"grad_norm": 0.8193458914756775,
|
|
"learning_rate": 1.3013013013013014e-05,
|
|
"loss": 2.3224,
|
|
"step": 9350
|
|
},
|
|
{
|
|
"epoch": 0.9355,
|
|
"grad_norm": 0.8291559815406799,
|
|
"learning_rate": 1.2912912912912914e-05,
|
|
"loss": 2.3303,
|
|
"step": 9355
|
|
},
|
|
{
|
|
"epoch": 0.936,
|
|
"grad_norm": 0.8167839646339417,
|
|
"learning_rate": 1.2812812812812813e-05,
|
|
"loss": 2.3534,
|
|
"step": 9360
|
|
},
|
|
{
|
|
"epoch": 0.9365,
|
|
"grad_norm": 0.8425468802452087,
|
|
"learning_rate": 1.2712712712712713e-05,
|
|
"loss": 2.34,
|
|
"step": 9365
|
|
},
|
|
{
|
|
"epoch": 0.937,
|
|
"grad_norm": 0.8351952433586121,
|
|
"learning_rate": 1.2612612612612611e-05,
|
|
"loss": 2.3526,
|
|
"step": 9370
|
|
},
|
|
{
|
|
"epoch": 0.9375,
|
|
"grad_norm": 0.8092942237854004,
|
|
"learning_rate": 1.2512512512512515e-05,
|
|
"loss": 2.3304,
|
|
"step": 9375
|
|
},
|
|
{
|
|
"epoch": 0.938,
|
|
"grad_norm": 0.8376879692077637,
|
|
"learning_rate": 1.2412412412412413e-05,
|
|
"loss": 2.3527,
|
|
"step": 9380
|
|
},
|
|
{
|
|
"epoch": 0.9385,
|
|
"grad_norm": 0.8303145170211792,
|
|
"learning_rate": 1.2312312312312313e-05,
|
|
"loss": 2.3195,
|
|
"step": 9385
|
|
},
|
|
{
|
|
"epoch": 0.939,
|
|
"grad_norm": 0.8191754817962646,
|
|
"learning_rate": 1.2212212212212212e-05,
|
|
"loss": 2.3222,
|
|
"step": 9390
|
|
},
|
|
{
|
|
"epoch": 0.9395,
|
|
"grad_norm": 0.8104075193405151,
|
|
"learning_rate": 1.2112112112112114e-05,
|
|
"loss": 2.3402,
|
|
"step": 9395
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"grad_norm": 0.8076393008232117,
|
|
"learning_rate": 1.2012012012012012e-05,
|
|
"loss": 2.3364,
|
|
"step": 9400
|
|
},
|
|
{
|
|
"epoch": 0.9405,
|
|
"grad_norm": 0.8219597935676575,
|
|
"learning_rate": 1.1911911911911912e-05,
|
|
"loss": 2.345,
|
|
"step": 9405
|
|
},
|
|
{
|
|
"epoch": 0.941,
|
|
"grad_norm": 0.8121560215950012,
|
|
"learning_rate": 1.1811811811811812e-05,
|
|
"loss": 2.3112,
|
|
"step": 9410
|
|
},
|
|
{
|
|
"epoch": 0.9415,
|
|
"grad_norm": 0.8694724440574646,
|
|
"learning_rate": 1.1711711711711713e-05,
|
|
"loss": 2.3605,
|
|
"step": 9415
|
|
},
|
|
{
|
|
"epoch": 0.942,
|
|
"grad_norm": 0.8347329497337341,
|
|
"learning_rate": 1.1611611611611613e-05,
|
|
"loss": 2.3182,
|
|
"step": 9420
|
|
},
|
|
{
|
|
"epoch": 0.9425,
|
|
"grad_norm": 0.8236081600189209,
|
|
"learning_rate": 1.1511511511511513e-05,
|
|
"loss": 2.3468,
|
|
"step": 9425
|
|
},
|
|
{
|
|
"epoch": 0.943,
|
|
"grad_norm": 0.8303726315498352,
|
|
"learning_rate": 1.1411411411411411e-05,
|
|
"loss": 2.3423,
|
|
"step": 9430
|
|
},
|
|
{
|
|
"epoch": 0.9435,
|
|
"grad_norm": 0.8322924971580505,
|
|
"learning_rate": 1.1311311311311313e-05,
|
|
"loss": 2.3514,
|
|
"step": 9435
|
|
},
|
|
{
|
|
"epoch": 0.944,
|
|
"grad_norm": 0.8485186100006104,
|
|
"learning_rate": 1.1211211211211212e-05,
|
|
"loss": 2.3361,
|
|
"step": 9440
|
|
},
|
|
{
|
|
"epoch": 0.9445,
|
|
"grad_norm": 0.8518514633178711,
|
|
"learning_rate": 1.1111111111111112e-05,
|
|
"loss": 2.348,
|
|
"step": 9445
|
|
},
|
|
{
|
|
"epoch": 0.945,
|
|
"grad_norm": 0.8338407278060913,
|
|
"learning_rate": 1.1011011011011012e-05,
|
|
"loss": 2.3402,
|
|
"step": 9450
|
|
},
|
|
{
|
|
"epoch": 0.9455,
|
|
"grad_norm": 0.8358404636383057,
|
|
"learning_rate": 1.0910910910910912e-05,
|
|
"loss": 2.3403,
|
|
"step": 9455
|
|
},
|
|
{
|
|
"epoch": 0.946,
|
|
"grad_norm": 0.8189331293106079,
|
|
"learning_rate": 1.0810810810810812e-05,
|
|
"loss": 2.3367,
|
|
"step": 9460
|
|
},
|
|
{
|
|
"epoch": 0.9465,
|
|
"grad_norm": 0.8365352749824524,
|
|
"learning_rate": 1.071071071071071e-05,
|
|
"loss": 2.3441,
|
|
"step": 9465
|
|
},
|
|
{
|
|
"epoch": 0.947,
|
|
"grad_norm": 0.8106439709663391,
|
|
"learning_rate": 1.061061061061061e-05,
|
|
"loss": 2.3496,
|
|
"step": 9470
|
|
},
|
|
{
|
|
"epoch": 0.9475,
|
|
"grad_norm": 0.8290128707885742,
|
|
"learning_rate": 1.051051051051051e-05,
|
|
"loss": 2.3161,
|
|
"step": 9475
|
|
},
|
|
{
|
|
"epoch": 0.948,
|
|
"grad_norm": 0.8077024221420288,
|
|
"learning_rate": 1.0410410410410411e-05,
|
|
"loss": 2.3411,
|
|
"step": 9480
|
|
},
|
|
{
|
|
"epoch": 0.9485,
|
|
"grad_norm": 0.838545024394989,
|
|
"learning_rate": 1.031031031031031e-05,
|
|
"loss": 2.3505,
|
|
"step": 9485
|
|
},
|
|
{
|
|
"epoch": 0.949,
|
|
"grad_norm": 0.8450316190719604,
|
|
"learning_rate": 1.0210210210210211e-05,
|
|
"loss": 2.3318,
|
|
"step": 9490
|
|
},
|
|
{
|
|
"epoch": 0.9495,
|
|
"grad_norm": 0.8616166710853577,
|
|
"learning_rate": 1.011011011011011e-05,
|
|
"loss": 2.3206,
|
|
"step": 9495
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"grad_norm": 0.8224554657936096,
|
|
"learning_rate": 1.0010010010010011e-05,
|
|
"loss": 2.3164,
|
|
"step": 9500
|
|
},
|
|
{
|
|
"epoch": 0.9505,
|
|
"grad_norm": 0.8563039898872375,
|
|
"learning_rate": 9.90990990990991e-06,
|
|
"loss": 2.3529,
|
|
"step": 9505
|
|
},
|
|
{
|
|
"epoch": 0.951,
|
|
"grad_norm": 0.8293529748916626,
|
|
"learning_rate": 9.80980980980981e-06,
|
|
"loss": 2.3583,
|
|
"step": 9510
|
|
},
|
|
{
|
|
"epoch": 0.9515,
|
|
"grad_norm": 0.8387513160705566,
|
|
"learning_rate": 9.70970970970971e-06,
|
|
"loss": 2.3161,
|
|
"step": 9515
|
|
},
|
|
{
|
|
"epoch": 0.952,
|
|
"grad_norm": 0.834874153137207,
|
|
"learning_rate": 9.60960960960961e-06,
|
|
"loss": 2.3416,
|
|
"step": 9520
|
|
},
|
|
{
|
|
"epoch": 0.9525,
|
|
"grad_norm": 0.8364129066467285,
|
|
"learning_rate": 9.509509509509509e-06,
|
|
"loss": 2.3284,
|
|
"step": 9525
|
|
},
|
|
{
|
|
"epoch": 0.953,
|
|
"grad_norm": 0.8416669368743896,
|
|
"learning_rate": 9.40940940940941e-06,
|
|
"loss": 2.3454,
|
|
"step": 9530
|
|
},
|
|
{
|
|
"epoch": 0.9535,
|
|
"grad_norm": 0.8180063962936401,
|
|
"learning_rate": 9.309309309309309e-06,
|
|
"loss": 2.3561,
|
|
"step": 9535
|
|
},
|
|
{
|
|
"epoch": 0.954,
|
|
"grad_norm": 0.8143042325973511,
|
|
"learning_rate": 9.209209209209211e-06,
|
|
"loss": 2.3251,
|
|
"step": 9540
|
|
},
|
|
{
|
|
"epoch": 0.9545,
|
|
"grad_norm": 0.813728392124176,
|
|
"learning_rate": 9.10910910910911e-06,
|
|
"loss": 2.3326,
|
|
"step": 9545
|
|
},
|
|
{
|
|
"epoch": 0.955,
|
|
"grad_norm": 0.8583104014396667,
|
|
"learning_rate": 9.00900900900901e-06,
|
|
"loss": 2.3532,
|
|
"step": 9550
|
|
},
|
|
{
|
|
"epoch": 0.9555,
|
|
"grad_norm": 0.8487512469291687,
|
|
"learning_rate": 8.90890890890891e-06,
|
|
"loss": 2.3126,
|
|
"step": 9555
|
|
},
|
|
{
|
|
"epoch": 0.956,
|
|
"grad_norm": 0.8339818716049194,
|
|
"learning_rate": 8.80880880880881e-06,
|
|
"loss": 2.3345,
|
|
"step": 9560
|
|
},
|
|
{
|
|
"epoch": 0.9565,
|
|
"grad_norm": 0.834420382976532,
|
|
"learning_rate": 8.708708708708708e-06,
|
|
"loss": 2.3485,
|
|
"step": 9565
|
|
},
|
|
{
|
|
"epoch": 0.957,
|
|
"grad_norm": 0.859883725643158,
|
|
"learning_rate": 8.60860860860861e-06,
|
|
"loss": 2.3398,
|
|
"step": 9570
|
|
},
|
|
{
|
|
"epoch": 0.9575,
|
|
"grad_norm": 0.8119943141937256,
|
|
"learning_rate": 8.508508508508508e-06,
|
|
"loss": 2.3428,
|
|
"step": 9575
|
|
},
|
|
{
|
|
"epoch": 0.958,
|
|
"grad_norm": 0.8337917923927307,
|
|
"learning_rate": 8.408408408408409e-06,
|
|
"loss": 2.3165,
|
|
"step": 9580
|
|
},
|
|
{
|
|
"epoch": 0.9585,
|
|
"grad_norm": 0.8146688342094421,
|
|
"learning_rate": 8.308308308308309e-06,
|
|
"loss": 2.3449,
|
|
"step": 9585
|
|
},
|
|
{
|
|
"epoch": 0.959,
|
|
"grad_norm": 0.8490365147590637,
|
|
"learning_rate": 8.208208208208209e-06,
|
|
"loss": 2.3161,
|
|
"step": 9590
|
|
},
|
|
{
|
|
"epoch": 0.9595,
|
|
"grad_norm": 0.8450210094451904,
|
|
"learning_rate": 8.108108108108109e-06,
|
|
"loss": 2.3502,
|
|
"step": 9595
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.8308297395706177,
|
|
"learning_rate": 8.008008008008007e-06,
|
|
"loss": 2.3125,
|
|
"step": 9600
|
|
},
|
|
{
|
|
"epoch": 0.9605,
|
|
"grad_norm": 0.8298580646514893,
|
|
"learning_rate": 7.907907907907908e-06,
|
|
"loss": 2.3375,
|
|
"step": 9605
|
|
},
|
|
{
|
|
"epoch": 0.961,
|
|
"grad_norm": 0.8202354311943054,
|
|
"learning_rate": 7.807807807807808e-06,
|
|
"loss": 2.3066,
|
|
"step": 9610
|
|
},
|
|
{
|
|
"epoch": 0.9615,
|
|
"grad_norm": 0.8283280730247498,
|
|
"learning_rate": 7.707707707707708e-06,
|
|
"loss": 2.3429,
|
|
"step": 9615
|
|
},
|
|
{
|
|
"epoch": 0.962,
|
|
"grad_norm": 0.8393738269805908,
|
|
"learning_rate": 7.607607607607609e-06,
|
|
"loss": 2.313,
|
|
"step": 9620
|
|
},
|
|
{
|
|
"epoch": 0.9625,
|
|
"grad_norm": 0.8319004774093628,
|
|
"learning_rate": 7.507507507507508e-06,
|
|
"loss": 2.3631,
|
|
"step": 9625
|
|
},
|
|
{
|
|
"epoch": 0.963,
|
|
"grad_norm": 0.8432734608650208,
|
|
"learning_rate": 7.4074074074074075e-06,
|
|
"loss": 2.3278,
|
|
"step": 9630
|
|
},
|
|
{
|
|
"epoch": 0.9635,
|
|
"grad_norm": 0.8351683020591736,
|
|
"learning_rate": 7.3073073073073085e-06,
|
|
"loss": 2.3359,
|
|
"step": 9635
|
|
},
|
|
{
|
|
"epoch": 0.964,
|
|
"grad_norm": 0.834241509437561,
|
|
"learning_rate": 7.207207207207208e-06,
|
|
"loss": 2.3358,
|
|
"step": 9640
|
|
},
|
|
{
|
|
"epoch": 0.9645,
|
|
"grad_norm": 0.8175002336502075,
|
|
"learning_rate": 7.107107107107107e-06,
|
|
"loss": 2.3433,
|
|
"step": 9645
|
|
},
|
|
{
|
|
"epoch": 0.965,
|
|
"grad_norm": 0.8678650856018066,
|
|
"learning_rate": 7.007007007007008e-06,
|
|
"loss": 2.3168,
|
|
"step": 9650
|
|
},
|
|
{
|
|
"epoch": 0.9655,
|
|
"grad_norm": 0.8319632411003113,
|
|
"learning_rate": 6.906906906906907e-06,
|
|
"loss": 2.3416,
|
|
"step": 9655
|
|
},
|
|
{
|
|
"epoch": 0.966,
|
|
"grad_norm": 0.8304513692855835,
|
|
"learning_rate": 6.8068068068068075e-06,
|
|
"loss": 2.3282,
|
|
"step": 9660
|
|
},
|
|
{
|
|
"epoch": 0.9665,
|
|
"grad_norm": 0.8106830716133118,
|
|
"learning_rate": 6.706706706706707e-06,
|
|
"loss": 2.31,
|
|
"step": 9665
|
|
},
|
|
{
|
|
"epoch": 0.967,
|
|
"grad_norm": 0.8108156323432922,
|
|
"learning_rate": 6.606606606606607e-06,
|
|
"loss": 2.3242,
|
|
"step": 9670
|
|
},
|
|
{
|
|
"epoch": 0.9675,
|
|
"grad_norm": 0.8338386416435242,
|
|
"learning_rate": 6.506506506506507e-06,
|
|
"loss": 2.3225,
|
|
"step": 9675
|
|
},
|
|
{
|
|
"epoch": 0.968,
|
|
"grad_norm": 0.8349241614341736,
|
|
"learning_rate": 6.406406406406406e-06,
|
|
"loss": 2.3267,
|
|
"step": 9680
|
|
},
|
|
{
|
|
"epoch": 0.9685,
|
|
"grad_norm": 0.8335773348808289,
|
|
"learning_rate": 6.306306306306306e-06,
|
|
"loss": 2.3355,
|
|
"step": 9685
|
|
},
|
|
{
|
|
"epoch": 0.969,
|
|
"grad_norm": 0.8067938089370728,
|
|
"learning_rate": 6.206206206206207e-06,
|
|
"loss": 2.3324,
|
|
"step": 9690
|
|
},
|
|
{
|
|
"epoch": 0.9695,
|
|
"grad_norm": 0.8177109360694885,
|
|
"learning_rate": 6.106106106106106e-06,
|
|
"loss": 2.3529,
|
|
"step": 9695
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"grad_norm": 0.8041896820068359,
|
|
"learning_rate": 6.006006006006006e-06,
|
|
"loss": 2.3179,
|
|
"step": 9700
|
|
},
|
|
{
|
|
"epoch": 0.9705,
|
|
"grad_norm": 0.818508505821228,
|
|
"learning_rate": 5.905905905905906e-06,
|
|
"loss": 2.3355,
|
|
"step": 9705
|
|
},
|
|
{
|
|
"epoch": 0.971,
|
|
"grad_norm": 0.8300430178642273,
|
|
"learning_rate": 5.805805805805806e-06,
|
|
"loss": 2.3316,
|
|
"step": 9710
|
|
},
|
|
{
|
|
"epoch": 0.9715,
|
|
"grad_norm": 0.8161868453025818,
|
|
"learning_rate": 5.705705705705706e-06,
|
|
"loss": 2.3282,
|
|
"step": 9715
|
|
},
|
|
{
|
|
"epoch": 0.972,
|
|
"grad_norm": 0.8113592267036438,
|
|
"learning_rate": 5.605605605605606e-06,
|
|
"loss": 2.3524,
|
|
"step": 9720
|
|
},
|
|
{
|
|
"epoch": 0.9725,
|
|
"grad_norm": 0.8195970058441162,
|
|
"learning_rate": 5.505505505505506e-06,
|
|
"loss": 2.321,
|
|
"step": 9725
|
|
},
|
|
{
|
|
"epoch": 0.973,
|
|
"grad_norm": 0.8336008191108704,
|
|
"learning_rate": 5.405405405405406e-06,
|
|
"loss": 2.3283,
|
|
"step": 9730
|
|
},
|
|
{
|
|
"epoch": 0.9735,
|
|
"grad_norm": 0.8411681056022644,
|
|
"learning_rate": 5.305305305305305e-06,
|
|
"loss": 2.3252,
|
|
"step": 9735
|
|
},
|
|
{
|
|
"epoch": 0.974,
|
|
"grad_norm": 0.8225618600845337,
|
|
"learning_rate": 5.2052052052052055e-06,
|
|
"loss": 2.3188,
|
|
"step": 9740
|
|
},
|
|
{
|
|
"epoch": 0.9745,
|
|
"grad_norm": 0.8509226441383362,
|
|
"learning_rate": 5.105105105105106e-06,
|
|
"loss": 2.3575,
|
|
"step": 9745
|
|
},
|
|
{
|
|
"epoch": 0.975,
|
|
"grad_norm": 0.8630853891372681,
|
|
"learning_rate": 5.005005005005006e-06,
|
|
"loss": 2.324,
|
|
"step": 9750
|
|
},
|
|
{
|
|
"epoch": 0.9755,
|
|
"grad_norm": 0.8567425608634949,
|
|
"learning_rate": 4.904904904904905e-06,
|
|
"loss": 2.3474,
|
|
"step": 9755
|
|
},
|
|
{
|
|
"epoch": 0.976,
|
|
"grad_norm": 0.8153373003005981,
|
|
"learning_rate": 4.804804804804805e-06,
|
|
"loss": 2.3445,
|
|
"step": 9760
|
|
},
|
|
{
|
|
"epoch": 0.9765,
|
|
"grad_norm": 0.8349518775939941,
|
|
"learning_rate": 4.704704704704705e-06,
|
|
"loss": 2.3057,
|
|
"step": 9765
|
|
},
|
|
{
|
|
"epoch": 0.977,
|
|
"grad_norm": 0.8440520167350769,
|
|
"learning_rate": 4.6046046046046055e-06,
|
|
"loss": 2.3208,
|
|
"step": 9770
|
|
},
|
|
{
|
|
"epoch": 0.9775,
|
|
"grad_norm": 0.8132700324058533,
|
|
"learning_rate": 4.504504504504505e-06,
|
|
"loss": 2.3367,
|
|
"step": 9775
|
|
},
|
|
{
|
|
"epoch": 0.978,
|
|
"grad_norm": 0.8062158226966858,
|
|
"learning_rate": 4.404404404404405e-06,
|
|
"loss": 2.312,
|
|
"step": 9780
|
|
},
|
|
{
|
|
"epoch": 0.9785,
|
|
"grad_norm": 0.848967969417572,
|
|
"learning_rate": 4.304304304304305e-06,
|
|
"loss": 2.3526,
|
|
"step": 9785
|
|
},
|
|
{
|
|
"epoch": 0.979,
|
|
"grad_norm": 0.8730021119117737,
|
|
"learning_rate": 4.204204204204204e-06,
|
|
"loss": 2.357,
|
|
"step": 9790
|
|
},
|
|
{
|
|
"epoch": 0.9795,
|
|
"grad_norm": 0.8056657910346985,
|
|
"learning_rate": 4.1041041041041045e-06,
|
|
"loss": 2.3621,
|
|
"step": 9795
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"grad_norm": 0.8473744988441467,
|
|
"learning_rate": 4.004004004004004e-06,
|
|
"loss": 2.335,
|
|
"step": 9800
|
|
},
|
|
{
|
|
"epoch": 0.9805,
|
|
"grad_norm": 0.8399832248687744,
|
|
"learning_rate": 3.903903903903904e-06,
|
|
"loss": 2.3334,
|
|
"step": 9805
|
|
},
|
|
{
|
|
"epoch": 0.981,
|
|
"grad_norm": 0.8139746785163879,
|
|
"learning_rate": 3.8038038038038044e-06,
|
|
"loss": 2.335,
|
|
"step": 9810
|
|
},
|
|
{
|
|
"epoch": 0.9815,
|
|
"grad_norm": 0.8043482899665833,
|
|
"learning_rate": 3.7037037037037037e-06,
|
|
"loss": 2.3141,
|
|
"step": 9815
|
|
},
|
|
{
|
|
"epoch": 0.982,
|
|
"grad_norm": 0.8388903737068176,
|
|
"learning_rate": 3.603603603603604e-06,
|
|
"loss": 2.3364,
|
|
"step": 9820
|
|
},
|
|
{
|
|
"epoch": 0.9825,
|
|
"grad_norm": 0.8074804544448853,
|
|
"learning_rate": 3.503503503503504e-06,
|
|
"loss": 2.3478,
|
|
"step": 9825
|
|
},
|
|
{
|
|
"epoch": 0.983,
|
|
"grad_norm": 0.8122521638870239,
|
|
"learning_rate": 3.4034034034034037e-06,
|
|
"loss": 2.3184,
|
|
"step": 9830
|
|
},
|
|
{
|
|
"epoch": 0.9835,
|
|
"grad_norm": 0.8089945912361145,
|
|
"learning_rate": 3.3033033033033035e-06,
|
|
"loss": 2.3368,
|
|
"step": 9835
|
|
},
|
|
{
|
|
"epoch": 0.984,
|
|
"grad_norm": 0.8560210466384888,
|
|
"learning_rate": 3.203203203203203e-06,
|
|
"loss": 2.3264,
|
|
"step": 9840
|
|
},
|
|
{
|
|
"epoch": 0.9845,
|
|
"grad_norm": 0.8315675258636475,
|
|
"learning_rate": 3.1031031031031033e-06,
|
|
"loss": 2.3474,
|
|
"step": 9845
|
|
},
|
|
{
|
|
"epoch": 0.985,
|
|
"grad_norm": 0.8220996856689453,
|
|
"learning_rate": 3.003003003003003e-06,
|
|
"loss": 2.3311,
|
|
"step": 9850
|
|
},
|
|
{
|
|
"epoch": 0.9855,
|
|
"grad_norm": 0.8553528785705566,
|
|
"learning_rate": 2.902902902902903e-06,
|
|
"loss": 2.3331,
|
|
"step": 9855
|
|
},
|
|
{
|
|
"epoch": 0.986,
|
|
"grad_norm": 0.8232652544975281,
|
|
"learning_rate": 2.802802802802803e-06,
|
|
"loss": 2.3272,
|
|
"step": 9860
|
|
},
|
|
{
|
|
"epoch": 0.9865,
|
|
"grad_norm": 0.8154972791671753,
|
|
"learning_rate": 2.702702702702703e-06,
|
|
"loss": 2.33,
|
|
"step": 9865
|
|
},
|
|
{
|
|
"epoch": 0.987,
|
|
"grad_norm": 0.7969358563423157,
|
|
"learning_rate": 2.6026026026026027e-06,
|
|
"loss": 2.3226,
|
|
"step": 9870
|
|
},
|
|
{
|
|
"epoch": 0.9875,
|
|
"grad_norm": 0.8221673965454102,
|
|
"learning_rate": 2.502502502502503e-06,
|
|
"loss": 2.3385,
|
|
"step": 9875
|
|
},
|
|
{
|
|
"epoch": 0.988,
|
|
"grad_norm": 0.8291456699371338,
|
|
"learning_rate": 2.4024024024024026e-06,
|
|
"loss": 2.3434,
|
|
"step": 9880
|
|
},
|
|
{
|
|
"epoch": 0.9885,
|
|
"grad_norm": 0.7976441383361816,
|
|
"learning_rate": 2.3023023023023027e-06,
|
|
"loss": 2.3119,
|
|
"step": 9885
|
|
},
|
|
{
|
|
"epoch": 0.989,
|
|
"grad_norm": 0.8049326539039612,
|
|
"learning_rate": 2.2022022022022024e-06,
|
|
"loss": 2.3409,
|
|
"step": 9890
|
|
},
|
|
{
|
|
"epoch": 0.9895,
|
|
"grad_norm": 0.8277580142021179,
|
|
"learning_rate": 2.102102102102102e-06,
|
|
"loss": 2.3349,
|
|
"step": 9895
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"grad_norm": 0.8092420101165771,
|
|
"learning_rate": 2.002002002002002e-06,
|
|
"loss": 2.3369,
|
|
"step": 9900
|
|
},
|
|
{
|
|
"epoch": 0.9905,
|
|
"grad_norm": 0.8332076668739319,
|
|
"learning_rate": 1.9019019019019022e-06,
|
|
"loss": 2.3319,
|
|
"step": 9905
|
|
},
|
|
{
|
|
"epoch": 0.991,
|
|
"grad_norm": 0.8229449391365051,
|
|
"learning_rate": 1.801801801801802e-06,
|
|
"loss": 2.3369,
|
|
"step": 9910
|
|
},
|
|
{
|
|
"epoch": 0.9915,
|
|
"grad_norm": 0.8201963305473328,
|
|
"learning_rate": 1.7017017017017019e-06,
|
|
"loss": 2.3482,
|
|
"step": 9915
|
|
},
|
|
{
|
|
"epoch": 0.992,
|
|
"grad_norm": 0.7966458797454834,
|
|
"learning_rate": 1.6016016016016016e-06,
|
|
"loss": 2.335,
|
|
"step": 9920
|
|
},
|
|
{
|
|
"epoch": 0.9925,
|
|
"grad_norm": 0.8265637159347534,
|
|
"learning_rate": 1.5015015015015015e-06,
|
|
"loss": 2.3531,
|
|
"step": 9925
|
|
},
|
|
{
|
|
"epoch": 0.993,
|
|
"grad_norm": 0.8379909992218018,
|
|
"learning_rate": 1.4014014014014014e-06,
|
|
"loss": 2.3564,
|
|
"step": 9930
|
|
},
|
|
{
|
|
"epoch": 0.9935,
|
|
"grad_norm": 0.8211309313774109,
|
|
"learning_rate": 1.3013013013013014e-06,
|
|
"loss": 2.3353,
|
|
"step": 9935
|
|
},
|
|
{
|
|
"epoch": 0.994,
|
|
"grad_norm": 0.8213915228843689,
|
|
"learning_rate": 1.2012012012012013e-06,
|
|
"loss": 2.3457,
|
|
"step": 9940
|
|
},
|
|
{
|
|
"epoch": 0.9945,
|
|
"grad_norm": 0.8311823606491089,
|
|
"learning_rate": 1.1011011011011012e-06,
|
|
"loss": 2.3165,
|
|
"step": 9945
|
|
},
|
|
{
|
|
"epoch": 0.995,
|
|
"grad_norm": 0.8197156190872192,
|
|
"learning_rate": 1.001001001001001e-06,
|
|
"loss": 2.3333,
|
|
"step": 9950
|
|
},
|
|
{
|
|
"epoch": 0.9955,
|
|
"grad_norm": 0.8054954409599304,
|
|
"learning_rate": 9.00900900900901e-07,
|
|
"loss": 2.3247,
|
|
"step": 9955
|
|
},
|
|
{
|
|
"epoch": 0.996,
|
|
"grad_norm": 0.8266953229904175,
|
|
"learning_rate": 8.008008008008008e-07,
|
|
"loss": 2.348,
|
|
"step": 9960
|
|
},
|
|
{
|
|
"epoch": 0.9965,
|
|
"grad_norm": 0.8368173241615295,
|
|
"learning_rate": 7.007007007007007e-07,
|
|
"loss": 2.3323,
|
|
"step": 9965
|
|
},
|
|
{
|
|
"epoch": 0.997,
|
|
"grad_norm": 0.7957631945610046,
|
|
"learning_rate": 6.006006006006006e-07,
|
|
"loss": 2.3362,
|
|
"step": 9970
|
|
},
|
|
{
|
|
"epoch": 0.9975,
|
|
"grad_norm": 0.8073971271514893,
|
|
"learning_rate": 5.005005005005005e-07,
|
|
"loss": 2.3181,
|
|
"step": 9975
|
|
},
|
|
{
|
|
"epoch": 0.998,
|
|
"grad_norm": 0.8300249576568604,
|
|
"learning_rate": 4.004004004004004e-07,
|
|
"loss": 2.3247,
|
|
"step": 9980
|
|
},
|
|
{
|
|
"epoch": 0.9985,
|
|
"grad_norm": 0.8339958190917969,
|
|
"learning_rate": 3.003003003003003e-07,
|
|
"loss": 2.3474,
|
|
"step": 9985
|
|
},
|
|
{
|
|
"epoch": 0.999,
|
|
"grad_norm": 0.909705400466919,
|
|
"learning_rate": 2.002002002002002e-07,
|
|
"loss": 2.3224,
|
|
"step": 9990
|
|
},
|
|
{
|
|
"epoch": 0.9995,
|
|
"grad_norm": 0.851557731628418,
|
|
"learning_rate": 1.001001001001001e-07,
|
|
"loss": 2.3473,
|
|
"step": 9995
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"grad_norm": 0.8165029287338257,
|
|
"learning_rate": 0.0,
|
|
"loss": 2.3496,
|
|
"step": 10000
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 10000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 9223372036854775807,
|
|
"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": 8252377006080000.0,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|