7920 lines
198 KiB
JSON
7920 lines
198 KiB
JSON
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 0.3906892183314476,
|
||
|
|
"eval_steps": 100,
|
||
|
|
"global_step": 10100,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 7.736345350456444e-05,
|
||
|
|
"grad_norm": 8.75,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 10.56122875213623,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0007736345350456444,
|
||
|
|
"grad_norm": 6.9375,
|
||
|
|
"learning_rate": 2.2499999999999998e-05,
|
||
|
|
"loss": 10.371153089735243,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0015472690700912889,
|
||
|
|
"grad_norm": 1.65625,
|
||
|
|
"learning_rate": 4.75e-05,
|
||
|
|
"loss": 9.225128936767579,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.002320903605136933,
|
||
|
|
"grad_norm": 1.2421875,
|
||
|
|
"learning_rate": 7.25e-05,
|
||
|
|
"loss": 8.501903533935547,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0030945381401825778,
|
||
|
|
"grad_norm": 0.8984375,
|
||
|
|
"learning_rate": 9.750000000000001e-05,
|
||
|
|
"loss": 7.907857513427734,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0038681726752282223,
|
||
|
|
"grad_norm": 0.65234375,
|
||
|
|
"learning_rate": 0.0001225,
|
||
|
|
"loss": 7.374447631835937,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.004641807210273866,
|
||
|
|
"grad_norm": 1.703125,
|
||
|
|
"learning_rate": 0.0001475,
|
||
|
|
"loss": 7.100068664550781,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.005415441745319511,
|
||
|
|
"grad_norm": 0.80859375,
|
||
|
|
"learning_rate": 0.0001725,
|
||
|
|
"loss": 6.936636352539063,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0061890762803651555,
|
||
|
|
"grad_norm": 0.6484375,
|
||
|
|
"learning_rate": 0.0001975,
|
||
|
|
"loss": 6.7047271728515625,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0069627108154108,
|
||
|
|
"grad_norm": 0.72265625,
|
||
|
|
"learning_rate": 0.00022250000000000001,
|
||
|
|
"loss": 6.5551597595214846,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.007736345350456445,
|
||
|
|
"grad_norm": 0.8984375,
|
||
|
|
"learning_rate": 0.0002475,
|
||
|
|
"loss": 6.410359191894531,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.007736345350456445,
|
||
|
|
"eval_loss": 6.641254425048828,
|
||
|
|
"eval_runtime": 8.5584,
|
||
|
|
"eval_samples_per_second": 38.208,
|
||
|
|
"eval_steps_per_second": 1.285,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.00850997988550209,
|
||
|
|
"grad_norm": 0.83984375,
|
||
|
|
"learning_rate": 0.0002725,
|
||
|
|
"loss": 6.291374969482422,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.009283614420547733,
|
||
|
|
"grad_norm": 1.40625,
|
||
|
|
"learning_rate": 0.00029749999999999997,
|
||
|
|
"loss": 6.181984710693359,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.010057248955593378,
|
||
|
|
"grad_norm": 0.67578125,
|
||
|
|
"learning_rate": 0.00032250000000000003,
|
||
|
|
"loss": 6.108818435668946,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.010830883490639022,
|
||
|
|
"grad_norm": 1.109375,
|
||
|
|
"learning_rate": 0.0003475,
|
||
|
|
"loss": 6.005937576293945,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.011604518025684667,
|
||
|
|
"grad_norm": 0.73046875,
|
||
|
|
"learning_rate": 0.0003725,
|
||
|
|
"loss": 5.927522277832031,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.012378152560730311,
|
||
|
|
"grad_norm": 1.015625,
|
||
|
|
"learning_rate": 0.0003975,
|
||
|
|
"loss": 5.850825500488281,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.013151787095775955,
|
||
|
|
"grad_norm": 1.296875,
|
||
|
|
"learning_rate": 0.00042249999999999997,
|
||
|
|
"loss": 5.774515533447266,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0139254216308216,
|
||
|
|
"grad_norm": 0.7890625,
|
||
|
|
"learning_rate": 0.00044750000000000004,
|
||
|
|
"loss": 5.700082015991211,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.014699056165867244,
|
||
|
|
"grad_norm": 1.0546875,
|
||
|
|
"learning_rate": 0.0004725,
|
||
|
|
"loss": 5.627538681030273,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.01547269070091289,
|
||
|
|
"grad_norm": 0.859375,
|
||
|
|
"learning_rate": 0.0004975,
|
||
|
|
"loss": 5.560461044311523,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.01547269070091289,
|
||
|
|
"eval_loss": 6.039037227630615,
|
||
|
|
"eval_runtime": 8.5816,
|
||
|
|
"eval_samples_per_second": 38.105,
|
||
|
|
"eval_steps_per_second": 1.282,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.016246325235958533,
|
||
|
|
"grad_norm": 1.375,
|
||
|
|
"learning_rate": 0.000499999413815452,
|
||
|
|
"loss": 5.472032928466797,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.01701995977100418,
|
||
|
|
"grad_norm": 0.95703125,
|
||
|
|
"learning_rate": 0.000499997387502211,
|
||
|
|
"loss": 5.414741897583008,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.017793594306049824,
|
||
|
|
"grad_norm": 0.8359375,
|
||
|
|
"learning_rate": 0.0004999939138357763,
|
||
|
|
"loss": 5.374539947509765,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.018567228841095466,
|
||
|
|
"grad_norm": 0.482421875,
|
||
|
|
"learning_rate": 0.0004999889928373172,
|
||
|
|
"loss": 5.2854866027832035,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.01934086337614111,
|
||
|
|
"grad_norm": 0.97265625,
|
||
|
|
"learning_rate": 0.0004999826245368231,
|
||
|
|
"loss": 5.209839630126953,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.020114497911186757,
|
||
|
|
"grad_norm": 0.69921875,
|
||
|
|
"learning_rate": 0.0004999748089731037,
|
||
|
|
"loss": 5.24133071899414,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0208881324462324,
|
||
|
|
"grad_norm": 0.328125,
|
||
|
|
"learning_rate": 0.0004999655461937884,
|
||
|
|
"loss": 5.189028549194336,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.021661766981278044,
|
||
|
|
"grad_norm": 0.294921875,
|
||
|
|
"learning_rate": 0.0004999548362553265,
|
||
|
|
"loss": 5.089565658569336,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02243540151632369,
|
||
|
|
"grad_norm": 0.96484375,
|
||
|
|
"learning_rate": 0.0004999426792229862,
|
||
|
|
"loss": 5.108868789672852,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.023209036051369335,
|
||
|
|
"grad_norm": 0.30859375,
|
||
|
|
"learning_rate": 0.0004999290751708547,
|
||
|
|
"loss": 5.045867919921875,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.023209036051369335,
|
||
|
|
"eval_loss": 5.63018274307251,
|
||
|
|
"eval_runtime": 8.5483,
|
||
|
|
"eval_samples_per_second": 38.253,
|
||
|
|
"eval_steps_per_second": 1.287,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.023982670586414977,
|
||
|
|
"grad_norm": 0.55859375,
|
||
|
|
"learning_rate": 0.0004999140241818376,
|
||
|
|
"loss": 4.964016342163086,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.024756305121460622,
|
||
|
|
"grad_norm": 0.96484375,
|
||
|
|
"learning_rate": 0.0004998975263476584,
|
||
|
|
"loss": 4.973379516601563,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.025529939656506268,
|
||
|
|
"grad_norm": 0.341796875,
|
||
|
|
"learning_rate": 0.0004998795817688582,
|
||
|
|
"loss": 4.920859909057617,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02630357419155191,
|
||
|
|
"grad_norm": 0.318359375,
|
||
|
|
"learning_rate": 0.0004998601905547944,
|
||
|
|
"loss": 4.873758316040039,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.027077208726597555,
|
||
|
|
"grad_norm": 0.7109375,
|
||
|
|
"learning_rate": 0.0004998393528236405,
|
||
|
|
"loss": 4.859211349487305,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0278508432616432,
|
||
|
|
"grad_norm": 0.328125,
|
||
|
|
"learning_rate": 0.000499817068702386,
|
||
|
|
"loss": 4.848406219482422,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.028624477796688846,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.0004997933383268339,
|
||
|
|
"loss": 4.7945610046386715,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.029398112331734488,
|
||
|
|
"grad_norm": 0.75,
|
||
|
|
"learning_rate": 0.0004997681618416019,
|
||
|
|
"loss": 4.757486343383789,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.030171746866780133,
|
||
|
|
"grad_norm": 0.369140625,
|
||
|
|
"learning_rate": 0.0004997415394001201,
|
||
|
|
"loss": 4.792052459716797,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03094538140182578,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0004997134711646306,
|
||
|
|
"loss": 4.726931381225586,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03094538140182578,
|
||
|
|
"eval_loss": 5.367307662963867,
|
||
|
|
"eval_runtime": 8.5603,
|
||
|
|
"eval_samples_per_second": 38.2,
|
||
|
|
"eval_steps_per_second": 1.285,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.031719015936871424,
|
||
|
|
"grad_norm": 0.65234375,
|
||
|
|
"learning_rate": 0.0004996839573061863,
|
||
|
|
"loss": 4.671530151367188,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.032492650471917066,
|
||
|
|
"grad_norm": 0.427734375,
|
||
|
|
"learning_rate": 0.0004996529980046502,
|
||
|
|
"loss": 4.672796630859375,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03326628500696271,
|
||
|
|
"grad_norm": 0.30859375,
|
||
|
|
"learning_rate": 0.0004996205934486943,
|
||
|
|
"loss": 4.658016204833984,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03403991954200836,
|
||
|
|
"grad_norm": 0.38671875,
|
||
|
|
"learning_rate": 0.0004995867438357975,
|
||
|
|
"loss": 4.623857498168945,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.034813554077054,
|
||
|
|
"grad_norm": 0.578125,
|
||
|
|
"learning_rate": 0.000499551449372246,
|
||
|
|
"loss": 4.6056865692138675,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03558718861209965,
|
||
|
|
"grad_norm": 0.5,
|
||
|
|
"learning_rate": 0.0004995147102731307,
|
||
|
|
"loss": 4.593499374389649,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03636082314714529,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004994765267623465,
|
||
|
|
"loss": 4.58136100769043,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03713445768219093,
|
||
|
|
"grad_norm": 0.294921875,
|
||
|
|
"learning_rate": 0.0004994368990725909,
|
||
|
|
"loss": 4.534087753295898,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03790809221723658,
|
||
|
|
"grad_norm": 0.50390625,
|
||
|
|
"learning_rate": 0.0004993958274453623,
|
||
|
|
"loss": 4.540646743774414,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03868172675228222,
|
||
|
|
"grad_norm": 0.6171875,
|
||
|
|
"learning_rate": 0.0004993533121309587,
|
||
|
|
"loss": 4.509712219238281,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03868172675228222,
|
||
|
|
"eval_loss": 5.2208099365234375,
|
||
|
|
"eval_runtime": 8.5343,
|
||
|
|
"eval_samples_per_second": 38.316,
|
||
|
|
"eval_steps_per_second": 1.289,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.039455361287327864,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004993093533884765,
|
||
|
|
"loss": 4.509745407104492,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04022899582237351,
|
||
|
|
"grad_norm": 0.3359375,
|
||
|
|
"learning_rate": 0.0004992639514858084,
|
||
|
|
"loss": 4.446602249145508,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.041002630357419155,
|
||
|
|
"grad_norm": 0.33984375,
|
||
|
|
"learning_rate": 0.0004992171066996421,
|
||
|
|
"loss": 4.452914428710938,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0417762648924648,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004991688193154583,
|
||
|
|
"loss": 4.4591011047363285,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.042549899427510446,
|
||
|
|
"grad_norm": 0.455078125,
|
||
|
|
"learning_rate": 0.0004991190896275294,
|
||
|
|
"loss": 4.453225326538086,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04332353396255609,
|
||
|
|
"grad_norm": 0.318359375,
|
||
|
|
"learning_rate": 0.0004990679179389172,
|
||
|
|
"loss": 4.408919143676758,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04409716849760173,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0004990153045614716,
|
||
|
|
"loss": 4.396371078491211,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04487080303264738,
|
||
|
|
"grad_norm": 0.3203125,
|
||
|
|
"learning_rate": 0.0004989612498158283,
|
||
|
|
"loss": 4.394485855102539,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04564443756769302,
|
||
|
|
"grad_norm": 0.453125,
|
||
|
|
"learning_rate": 0.0004989057540314069,
|
||
|
|
"loss": 4.370931625366211,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04641807210273867,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004988488175464091,
|
||
|
|
"loss": 4.387384033203125,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04641807210273867,
|
||
|
|
"eval_loss": 5.092612266540527,
|
||
|
|
"eval_runtime": 8.6148,
|
||
|
|
"eval_samples_per_second": 37.958,
|
||
|
|
"eval_steps_per_second": 1.277,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04719170663778431,
|
||
|
|
"grad_norm": 0.306640625,
|
||
|
|
"learning_rate": 0.0004987904407078164,
|
||
|
|
"loss": 4.3401447296142575,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04796534117282995,
|
||
|
|
"grad_norm": 0.328125,
|
||
|
|
"learning_rate": 0.000498730623871388,
|
||
|
|
"loss": 4.367118835449219,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0487389757078756,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004986693674016589,
|
||
|
|
"loss": 4.342254257202148,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.049512610242921244,
|
||
|
|
"grad_norm": 0.36328125,
|
||
|
|
"learning_rate": 0.0004986066716719372,
|
||
|
|
"loss": 4.318671798706054,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.050286244777966886,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004985425370643027,
|
||
|
|
"loss": 4.326435089111328,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.051059879313012535,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004984769639696033,
|
||
|
|
"loss": 4.320156860351562,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05183351384805818,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004984099527874539,
|
||
|
|
"loss": 4.2882133483886715,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05260714838310382,
|
||
|
|
"grad_norm": 0.224609375,
|
||
|
|
"learning_rate": 0.0004983415039262328,
|
||
|
|
"loss": 4.269629669189453,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05338078291814947,
|
||
|
|
"grad_norm": 0.30859375,
|
||
|
|
"learning_rate": 0.0004982716178030802,
|
||
|
|
"loss": 4.244283676147461,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05415441745319511,
|
||
|
|
"grad_norm": 0.33984375,
|
||
|
|
"learning_rate": 0.000498200294843895,
|
||
|
|
"loss": 4.2871452331542965,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05415441745319511,
|
||
|
|
"eval_loss": 4.956212043762207,
|
||
|
|
"eval_runtime": 8.6126,
|
||
|
|
"eval_samples_per_second": 37.968,
|
||
|
|
"eval_steps_per_second": 1.277,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05492805198824075,
|
||
|
|
"grad_norm": 0.294921875,
|
||
|
|
"learning_rate": 0.0004981275354833328,
|
||
|
|
"loss": 4.227527236938476,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0557016865232864,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004980533401648026,
|
||
|
|
"loss": 4.238017654418945,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05647532105833204,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004979777093404642,
|
||
|
|
"loss": 4.230612945556641,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05724895559337769,
|
||
|
|
"grad_norm": 0.302734375,
|
||
|
|
"learning_rate": 0.0004979006434712263,
|
||
|
|
"loss": 4.228169250488281,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05802259012842333,
|
||
|
|
"grad_norm": 0.375,
|
||
|
|
"learning_rate": 0.0004978221430267422,
|
||
|
|
"loss": 4.209581756591797,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.058796224663468975,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.0004977422084854085,
|
||
|
|
"loss": 4.2135154724121096,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.059569859198514624,
|
||
|
|
"grad_norm": 0.234375,
|
||
|
|
"learning_rate": 0.000497660840334361,
|
||
|
|
"loss": 4.225874710083008,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.060343493733560266,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004975780390694723,
|
||
|
|
"loss": 4.196949768066406,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06111712826860591,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004974938051953488,
|
||
|
|
"loss": 4.196494674682617,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06189076280365156,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.0004974081392253274,
|
||
|
|
"loss": 4.1752674102783205,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06189076280365156,
|
||
|
|
"eval_loss": 4.886258602142334,
|
||
|
|
"eval_runtime": 8.574,
|
||
|
|
"eval_samples_per_second": 38.138,
|
||
|
|
"eval_steps_per_second": 1.283,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0626643973386972,
|
||
|
|
"grad_norm": 0.310546875,
|
||
|
|
"learning_rate": 0.0004973210416814722,
|
||
|
|
"loss": 4.171138763427734,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06343803187374285,
|
||
|
|
"grad_norm": 0.205078125,
|
||
|
|
"learning_rate": 0.000497232513094572,
|
||
|
|
"loss": 4.168529891967774,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06421166640878849,
|
||
|
|
"grad_norm": 0.302734375,
|
||
|
|
"learning_rate": 0.0004971425540041365,
|
||
|
|
"loss": 4.1555931091308596,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06498530094383413,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.000497051164958393,
|
||
|
|
"loss": 4.1395000457763675,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06575893547887977,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004969583465142832,
|
||
|
|
"loss": 4.131367111206055,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06653257001392542,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.0004968640992374602,
|
||
|
|
"loss": 4.145170211791992,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06730620454897107,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004967684237022843,
|
||
|
|
"loss": 4.137904357910156,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06807983908401671,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0004966713204918199,
|
||
|
|
"loss": 4.099301528930664,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06885347361906236,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.0004965727901978322,
|
||
|
|
"loss": 4.081951522827149,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.069627108154108,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004964728334207829,
|
||
|
|
"loss": 4.109021377563477,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.069627108154108,
|
||
|
|
"eval_loss": 4.80100154876709,
|
||
|
|
"eval_runtime": 8.5773,
|
||
|
|
"eval_samples_per_second": 38.124,
|
||
|
|
"eval_steps_per_second": 1.282,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07040074268915364,
|
||
|
|
"grad_norm": 0.23046875,
|
||
|
|
"learning_rate": 0.0004963714507698272,
|
||
|
|
"loss": 4.122850036621093,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0711743772241993,
|
||
|
|
"grad_norm": 0.302734375,
|
||
|
|
"learning_rate": 0.0004962686428628099,
|
||
|
|
"loss": 4.123403930664063,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07194801175924494,
|
||
|
|
"grad_norm": 0.1953125,
|
||
|
|
"learning_rate": 0.0004961644103262615,
|
||
|
|
"loss": 4.103722763061524,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07272164629429058,
|
||
|
|
"grad_norm": 0.208984375,
|
||
|
|
"learning_rate": 0.0004960587537953944,
|
||
|
|
"loss": 4.106454086303711,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07349528082933622,
|
||
|
|
"grad_norm": 0.2294921875,
|
||
|
|
"learning_rate": 0.000495951673914099,
|
||
|
|
"loss": 4.093786239624023,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07426891536438186,
|
||
|
|
"grad_norm": 0.224609375,
|
||
|
|
"learning_rate": 0.0004958431713349402,
|
||
|
|
"loss": 4.103993225097656,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0750425498994275,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.0004957332467191527,
|
||
|
|
"loss": 4.083438491821289,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07581618443447316,
|
||
|
|
"grad_norm": 0.2294921875,
|
||
|
|
"learning_rate": 0.0004956219007366376,
|
||
|
|
"loss": 4.068552780151367,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0765898189695188,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.000495509134065958,
|
||
|
|
"loss": 4.07377815246582,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07736345350456444,
|
||
|
|
"grad_norm": 0.2265625,
|
||
|
|
"learning_rate": 0.0004953949473943349,
|
||
|
|
"loss": 4.056775283813477,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07736345350456444,
|
||
|
|
"eval_loss": 4.741029739379883,
|
||
|
|
"eval_runtime": 8.5971,
|
||
|
|
"eval_samples_per_second": 38.036,
|
||
|
|
"eval_steps_per_second": 1.279,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07813708803961009,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004952793414176431,
|
||
|
|
"loss": 4.064179992675781,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07891072257465573,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0004951623168404069,
|
||
|
|
"loss": 4.05681266784668,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07968435710970137,
|
||
|
|
"grad_norm": 0.201171875,
|
||
|
|
"learning_rate": 0.0004950438743757959,
|
||
|
|
"loss": 4.0333606719970705,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08045799164474703,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004949240147456203,
|
||
|
|
"loss": 4.008632659912109,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08123162617979267,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.0004948027386803271,
|
||
|
|
"loss": 4.046485900878906,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08200526071483831,
|
||
|
|
"grad_norm": 0.20703125,
|
||
|
|
"learning_rate": 0.0004946800469189951,
|
||
|
|
"loss": 4.016164016723633,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08277889524988395,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004945559402093307,
|
||
|
|
"loss": 4.027731704711914,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0835525297849296,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004944304193076633,
|
||
|
|
"loss": 4.006626892089844,
|
||
|
|
"step": 1080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08432616431997525,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004943034849789404,
|
||
|
|
"loss": 4.007661056518555,
|
||
|
|
"step": 1090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08509979885502089,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004941751379967235,
|
||
|
|
"loss": 4.004118347167969,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08509979885502089,
|
||
|
|
"eval_loss": 4.684783935546875,
|
||
|
|
"eval_runtime": 8.5416,
|
||
|
|
"eval_samples_per_second": 38.283,
|
||
|
|
"eval_steps_per_second": 1.288,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08587343339006653,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004940453791431831,
|
||
|
|
"loss": 3.999224090576172,
|
||
|
|
"step": 1110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08664706792511218,
|
||
|
|
"grad_norm": 0.2060546875,
|
||
|
|
"learning_rate": 0.0004939142092090935,
|
||
|
|
"loss": 3.9808334350585937,
|
||
|
|
"step": 1120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08742070246015782,
|
||
|
|
"grad_norm": 0.2109375,
|
||
|
|
"learning_rate": 0.0004937816289938288,
|
||
|
|
"loss": 4.003896331787109,
|
||
|
|
"step": 1130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08819433699520346,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004936476393053574,
|
||
|
|
"loss": 3.9906829833984374,
|
||
|
|
"step": 1140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08896797153024912,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.0004935122409602374,
|
||
|
|
"loss": 3.9897789001464843,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08974160606529476,
|
||
|
|
"grad_norm": 0.236328125,
|
||
|
|
"learning_rate": 0.0004933754347836115,
|
||
|
|
"loss": 3.9780765533447267,
|
||
|
|
"step": 1160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0905152406003404,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 0.0004932372216092017,
|
||
|
|
"loss": 3.976204681396484,
|
||
|
|
"step": 1170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09128887513538604,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004930976022793051,
|
||
|
|
"loss": 3.9825729370117187,
|
||
|
|
"step": 1180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09206250967043168,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.0004929565776447875,
|
||
|
|
"loss": 3.956177520751953,
|
||
|
|
"step": 1190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09283614420547734,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.0004928141485650795,
|
||
|
|
"loss": 3.9746784210205077,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09283614420547734,
|
||
|
|
"eval_loss": 4.6136932373046875,
|
||
|
|
"eval_runtime": 8.5763,
|
||
|
|
"eval_samples_per_second": 38.129,
|
||
|
|
"eval_steps_per_second": 1.283,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09360977874052298,
|
||
|
|
"grad_norm": 0.2216796875,
|
||
|
|
"learning_rate": 0.0004926703159081702,
|
||
|
|
"loss": 3.9590415954589844,
|
||
|
|
"step": 1210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09438341327556862,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004925250805506026,
|
||
|
|
"loss": 3.9336162567138673,
|
||
|
|
"step": 1220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09515704781061426,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004923784433774679,
|
||
|
|
"loss": 3.9630496978759764,
|
||
|
|
"step": 1230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0959306823456599,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.0004922304052824003,
|
||
|
|
"loss": 3.9313884735107423,
|
||
|
|
"step": 1240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09670431688070555,
|
||
|
|
"grad_norm": 0.2216796875,
|
||
|
|
"learning_rate": 0.0004920809671675715,
|
||
|
|
"loss": 3.9482696533203123,
|
||
|
|
"step": 1250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0974779514157512,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.000491930129943685,
|
||
|
|
"loss": 3.9435379028320314,
|
||
|
|
"step": 1260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09825158595079685,
|
||
|
|
"grad_norm": 0.2158203125,
|
||
|
|
"learning_rate": 0.0004917778945299709,
|
||
|
|
"loss": 3.9084590911865233,
|
||
|
|
"step": 1270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09902522048584249,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.0004916242618541802,
|
||
|
|
"loss": 3.91564826965332,
|
||
|
|
"step": 1280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09979885502088813,
|
||
|
|
"grad_norm": 0.236328125,
|
||
|
|
"learning_rate": 0.0004914692328525788,
|
||
|
|
"loss": 3.931330108642578,
|
||
|
|
"step": 1290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10057248955593377,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004913128084699424,
|
||
|
|
"loss": 3.928493881225586,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10057248955593377,
|
||
|
|
"eval_loss": 4.577918529510498,
|
||
|
|
"eval_runtime": 8.5374,
|
||
|
|
"eval_samples_per_second": 38.302,
|
||
|
|
"eval_steps_per_second": 1.288,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10134612409097943,
|
||
|
|
"grad_norm": 0.21875,
|
||
|
|
"learning_rate": 0.0004911549896595501,
|
||
|
|
"loss": 3.922897720336914,
|
||
|
|
"step": 1310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10211975862602507,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0004909957773831791,
|
||
|
|
"loss": 3.927952194213867,
|
||
|
|
"step": 1320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10289339316107071,
|
||
|
|
"grad_norm": 0.1982421875,
|
||
|
|
"learning_rate": 0.0004908351726110987,
|
||
|
|
"loss": 3.9283069610595702,
|
||
|
|
"step": 1330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10366702769611635,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.000490673176322064,
|
||
|
|
"loss": 3.9404796600341796,
|
||
|
|
"step": 1340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.104440662231162,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.0004905097895033106,
|
||
|
|
"loss": 3.9214569091796876,
|
||
|
|
"step": 1350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10521429676620764,
|
||
|
|
"grad_norm": 0.208984375,
|
||
|
|
"learning_rate": 0.000490345013150548,
|
||
|
|
"loss": 3.9170085906982424,
|
||
|
|
"step": 1360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1059879313012533,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004901788482679542,
|
||
|
|
"loss": 3.899140167236328,
|
||
|
|
"step": 1370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10676156583629894,
|
||
|
|
"grad_norm": 0.236328125,
|
||
|
|
"learning_rate": 0.0004900112958681686,
|
||
|
|
"loss": 3.913042449951172,
|
||
|
|
"step": 1380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10753520037134458,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004898423569722866,
|
||
|
|
"loss": 3.9000522613525392,
|
||
|
|
"step": 1390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10830883490639022,
|
||
|
|
"grad_norm": 0.2060546875,
|
||
|
|
"learning_rate": 0.0004896720326098534,
|
||
|
|
"loss": 3.907938003540039,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10830883490639022,
|
||
|
|
"eval_loss": 4.5124287605285645,
|
||
|
|
"eval_runtime": 8.581,
|
||
|
|
"eval_samples_per_second": 38.108,
|
||
|
|
"eval_steps_per_second": 1.282,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10908246944143586,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.0004895003238188572,
|
||
|
|
"loss": 3.896725082397461,
|
||
|
|
"step": 1410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1098561039764815,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004893272316457232,
|
||
|
|
"loss": 3.887255859375,
|
||
|
|
"step": 1420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11062973851152716,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.0004891527571453074,
|
||
|
|
"loss": 3.899026107788086,
|
||
|
|
"step": 1430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1114033730465728,
|
||
|
|
"grad_norm": 0.2294921875,
|
||
|
|
"learning_rate": 0.0004889769013808898,
|
||
|
|
"loss": 3.9211231231689454,
|
||
|
|
"step": 1440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11217700758161844,
|
||
|
|
"grad_norm": 0.2099609375,
|
||
|
|
"learning_rate": 0.000488799665424168,
|
||
|
|
"loss": 3.8967121124267576,
|
||
|
|
"step": 1450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11295064211666409,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004886210503552508,
|
||
|
|
"loss": 3.8699275970458986,
|
||
|
|
"step": 1460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11372427665170973,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004884410572626518,
|
||
|
|
"loss": 3.8598350524902343,
|
||
|
|
"step": 1470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11449791118675538,
|
||
|
|
"grad_norm": 0.201171875,
|
||
|
|
"learning_rate": 0.0004882596872432822,
|
||
|
|
"loss": 3.90062255859375,
|
||
|
|
"step": 1480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11527154572180102,
|
||
|
|
"grad_norm": 0.2158203125,
|
||
|
|
"learning_rate": 0.0004880769414024446,
|
||
|
|
"loss": 3.8603435516357423,
|
||
|
|
"step": 1490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11604518025684667,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004878928208538261,
|
||
|
|
"loss": 3.8940032958984374,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11604518025684667,
|
||
|
|
"eval_loss": 4.474898338317871,
|
||
|
|
"eval_runtime": 8.5376,
|
||
|
|
"eval_samples_per_second": 38.301,
|
||
|
|
"eval_steps_per_second": 1.288,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11681881479189231,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004877073267194916,
|
||
|
|
"loss": 3.8529014587402344,
|
||
|
|
"step": 1510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11759244932693795,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.00048752046012987677,
|
||
|
|
"loss": 3.888753128051758,
|
||
|
|
"step": 1520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11836608386198359,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.0004873322222237812,
|
||
|
|
"loss": 3.856996536254883,
|
||
|
|
"step": 1530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11913971839702925,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004871426141483618,
|
||
|
|
"loss": 3.841297149658203,
|
||
|
|
"step": 1540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11991335293207489,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004869516370591253,
|
||
|
|
"loss": 3.838258743286133,
|
||
|
|
"step": 1550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12068698746712053,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004867592921199217,
|
||
|
|
"loss": 3.852222442626953,
|
||
|
|
"step": 1560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12146062200216617,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004865655805029366,
|
||
|
|
"loss": 3.8498695373535154,
|
||
|
|
"step": 1570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12223425653721182,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004863705033886847,
|
||
|
|
"loss": 3.8373653411865236,
|
||
|
|
"step": 1580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12300789107225747,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.0004861740619660022,
|
||
|
|
"loss": 3.8427047729492188,
|
||
|
|
"step": 1590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12378152560730311,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.0004859762574320395,
|
||
|
|
"loss": 3.864556884765625,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12378152560730311,
|
||
|
|
"eval_loss": 4.4412994384765625,
|
||
|
|
"eval_runtime": 9.5955,
|
||
|
|
"eval_samples_per_second": 34.079,
|
||
|
|
"eval_steps_per_second": 1.146,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12455516014234876,
|
||
|
|
"grad_norm": 0.2001953125,
|
||
|
|
"learning_rate": 0.0004857770909922542,
|
||
|
|
"loss": 3.816178894042969,
|
||
|
|
"step": 1610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1253287946773944,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0004855765638604036,
|
||
|
|
"loss": 3.8218841552734375,
|
||
|
|
"step": 1620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12610242921244005,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00048537467725853734,
|
||
|
|
"loss": 3.8202373504638674,
|
||
|
|
"step": 1630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1268760637474857,
|
||
|
|
"grad_norm": 0.201171875,
|
||
|
|
"learning_rate": 0.00048517143241698986,
|
||
|
|
"loss": 3.8387172698974608,
|
||
|
|
"step": 1640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12764969828253134,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.0004849668305743729,
|
||
|
|
"loss": 3.8024673461914062,
|
||
|
|
"step": 1650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12842333281757698,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.000484760872977568,
|
||
|
|
"loss": 3.8411659240722655,
|
||
|
|
"step": 1660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12919696735262262,
|
||
|
|
"grad_norm": 0.2294921875,
|
||
|
|
"learning_rate": 0.0004845535608817193,
|
||
|
|
"loss": 3.828306198120117,
|
||
|
|
"step": 1670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12997060188766826,
|
||
|
|
"grad_norm": 0.19921875,
|
||
|
|
"learning_rate": 0.000484344895550225,
|
||
|
|
"loss": 3.8343215942382813,
|
||
|
|
"step": 1680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1307442364227139,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.00048413487825473044,
|
||
|
|
"loss": 3.829656219482422,
|
||
|
|
"step": 1690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13151787095775955,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.0004839235102751201,
|
||
|
|
"loss": 3.825642776489258,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13151787095775955,
|
||
|
|
"eval_loss": 4.401412010192871,
|
||
|
|
"eval_runtime": 8.6836,
|
||
|
|
"eval_samples_per_second": 37.657,
|
||
|
|
"eval_steps_per_second": 1.267,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1322915054928052,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.00048371079289950966,
|
||
|
|
"loss": 3.8195194244384765,
|
||
|
|
"step": 1710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13306514002785083,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004834967274242383,
|
||
|
|
"loss": 3.786629867553711,
|
||
|
|
"step": 1720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1338387745628965,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.0004832813151538607,
|
||
|
|
"loss": 3.798793411254883,
|
||
|
|
"step": 1730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13461240909794214,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00048306455740113936,
|
||
|
|
"loss": 3.7926116943359376,
|
||
|
|
"step": 1740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13538604363298778,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.0004828464554870363,
|
||
|
|
"loss": 3.812264633178711,
|
||
|
|
"step": 1750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13615967816803343,
|
||
|
|
"grad_norm": 0.2138671875,
|
||
|
|
"learning_rate": 0.00048262701074070516,
|
||
|
|
"loss": 3.7819557189941406,
|
||
|
|
"step": 1760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13693331270307907,
|
||
|
|
"grad_norm": 0.2314453125,
|
||
|
|
"learning_rate": 0.00048240622449948275,
|
||
|
|
"loss": 3.772268295288086,
|
||
|
|
"step": 1770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1377069472381247,
|
||
|
|
"grad_norm": 0.208984375,
|
||
|
|
"learning_rate": 0.00048218409810888156,
|
||
|
|
"loss": 3.7845218658447264,
|
||
|
|
"step": 1780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13848058177317035,
|
||
|
|
"grad_norm": 0.201171875,
|
||
|
|
"learning_rate": 0.00048196063292258123,
|
||
|
|
"loss": 3.7898277282714843,
|
||
|
|
"step": 1790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.139254216308216,
|
||
|
|
"grad_norm": 0.23046875,
|
||
|
|
"learning_rate": 0.0004817358303024199,
|
||
|
|
"loss": 3.7971900939941405,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.139254216308216,
|
||
|
|
"eval_loss": 4.37278413772583,
|
||
|
|
"eval_runtime": 8.6618,
|
||
|
|
"eval_samples_per_second": 37.752,
|
||
|
|
"eval_steps_per_second": 1.27,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14002785084326164,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004815096916183866,
|
||
|
|
"loss": 3.7643463134765627,
|
||
|
|
"step": 1810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14080148537830728,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.00048128221824861236,
|
||
|
|
"loss": 3.777587890625,
|
||
|
|
"step": 1820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14157511991335292,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0004810534115793623,
|
||
|
|
"loss": 3.775267791748047,
|
||
|
|
"step": 1830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1423487544483986,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.00048082327300502664,
|
||
|
|
"loss": 3.7802684783935545,
|
||
|
|
"step": 1840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14312238898344423,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00048059180392811266,
|
||
|
|
"loss": 3.8166263580322264,
|
||
|
|
"step": 1850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14389602351848987,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004803590057592359,
|
||
|
|
"loss": 3.7695087432861327,
|
||
|
|
"step": 1860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14466965805353552,
|
||
|
|
"grad_norm": 0.2119140625,
|
||
|
|
"learning_rate": 0.0004801248799171116,
|
||
|
|
"loss": 3.7461666107177733,
|
||
|
|
"step": 1870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14544329258858116,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00047988942782854624,
|
||
|
|
"loss": 3.7527397155761717,
|
||
|
|
"step": 1880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1462169271236268,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.00047965265092842846,
|
||
|
|
"loss": 3.74096565246582,
|
||
|
|
"step": 1890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14699056165867244,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00047941455065972073,
|
||
|
|
"loss": 3.7666641235351563,
|
||
|
|
"step": 1900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14699056165867244,
|
||
|
|
"eval_loss": 4.354757785797119,
|
||
|
|
"eval_runtime": 8.7304,
|
||
|
|
"eval_samples_per_second": 37.455,
|
||
|
|
"eval_steps_per_second": 1.26,
|
||
|
|
"step": 1900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14776419619371808,
|
||
|
|
"grad_norm": 0.20703125,
|
||
|
|
"learning_rate": 0.0004791751284734504,
|
||
|
|
"loss": 3.752808380126953,
|
||
|
|
"step": 1910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14853783072876373,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004789343858287005,
|
||
|
|
"loss": 3.744313049316406,
|
||
|
|
"step": 1920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14931146526380937,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.00047869232419260165,
|
||
|
|
"loss": 3.738627243041992,
|
||
|
|
"step": 1930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.150085099798855,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.0004784489450403224,
|
||
|
|
"loss": 3.769034194946289,
|
||
|
|
"step": 1940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15085873433390065,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0004782042498550604,
|
||
|
|
"loss": 3.754560089111328,
|
||
|
|
"step": 1950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15163236886894632,
|
||
|
|
"grad_norm": 0.2138671875,
|
||
|
|
"learning_rate": 0.00047795824012803357,
|
||
|
|
"loss": 3.7514068603515627,
|
||
|
|
"step": 1960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15240600340399196,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.000477710917358471,
|
||
|
|
"loss": 3.7632373809814452,
|
||
|
|
"step": 1970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1531796379390376,
|
||
|
|
"grad_norm": 0.23046875,
|
||
|
|
"learning_rate": 0.00047746228305360345,
|
||
|
|
"loss": 3.7417659759521484,
|
||
|
|
"step": 1980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15395327247408325,
|
||
|
|
"grad_norm": 0.21484375,
|
||
|
|
"learning_rate": 0.00047721233872865463,
|
||
|
|
"loss": 3.7318050384521486,
|
||
|
|
"step": 1990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1547269070091289,
|
||
|
|
"grad_norm": 0.2119140625,
|
||
|
|
"learning_rate": 0.00047696108590683175,
|
||
|
|
"loss": 3.7479320526123048,
|
||
|
|
"step": 2000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1547269070091289,
|
||
|
|
"eval_loss": 4.3264288902282715,
|
||
|
|
"eval_runtime": 8.7274,
|
||
|
|
"eval_samples_per_second": 37.468,
|
||
|
|
"eval_steps_per_second": 1.26,
|
||
|
|
"step": 2000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15550054154417453,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004767085261193161,
|
||
|
|
"loss": 3.711147689819336,
|
||
|
|
"step": 2010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15627417607922017,
|
||
|
|
"grad_norm": 0.220703125,
|
||
|
|
"learning_rate": 0.0004764546609052538,
|
||
|
|
"loss": 3.749440002441406,
|
||
|
|
"step": 2020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15704781061426581,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004761994918117469,
|
||
|
|
"loss": 3.7327606201171877,
|
||
|
|
"step": 2030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15782144514931146,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.000475943020393843,
|
||
|
|
"loss": 3.7557136535644533,
|
||
|
|
"step": 2040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1585950796843571,
|
||
|
|
"grad_norm": 0.2138671875,
|
||
|
|
"learning_rate": 0.0004756852482145268,
|
||
|
|
"loss": 3.723508834838867,
|
||
|
|
"step": 2050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15936871421940274,
|
||
|
|
"grad_norm": 0.2158203125,
|
||
|
|
"learning_rate": 0.00047542617684470965,
|
||
|
|
"loss": 3.732291412353516,
|
||
|
|
"step": 2060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1601423487544484,
|
||
|
|
"grad_norm": 0.21484375,
|
||
|
|
"learning_rate": 0.0004751658078632206,
|
||
|
|
"loss": 3.7624893188476562,
|
||
|
|
"step": 2070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16091598328949405,
|
||
|
|
"grad_norm": 0.2080078125,
|
||
|
|
"learning_rate": 0.00047490414285679666,
|
||
|
|
"loss": 3.7248538970947265,
|
||
|
|
"step": 2080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1616896178245397,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.000474641183420073,
|
||
|
|
"loss": 3.732358932495117,
|
||
|
|
"step": 2090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16246325235958534,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004743769311555732,
|
||
|
|
"loss": 3.7347564697265625,
|
||
|
|
"step": 2100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16246325235958534,
|
||
|
|
"eval_loss": 4.292186260223389,
|
||
|
|
"eval_runtime": 8.7033,
|
||
|
|
"eval_samples_per_second": 37.572,
|
||
|
|
"eval_steps_per_second": 1.264,
|
||
|
|
"step": 2100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16323688689463098,
|
||
|
|
"grad_norm": 0.2197265625,
|
||
|
|
"learning_rate": 0.0004741113876736997,
|
||
|
|
"loss": 3.7274330139160154,
|
||
|
|
"step": 2110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16401052142967662,
|
||
|
|
"grad_norm": 0.2265625,
|
||
|
|
"learning_rate": 0.00047384455459272386,
|
||
|
|
"loss": 3.7328311920166017,
|
||
|
|
"step": 2120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16478415596472226,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004735764335387758,
|
||
|
|
"loss": 3.7267345428466796,
|
||
|
|
"step": 2130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1655577904997679,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004733070261458353,
|
||
|
|
"loss": 3.7383796691894533,
|
||
|
|
"step": 2140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16633142503481355,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00047303633405572094,
|
||
|
|
"loss": 3.714567184448242,
|
||
|
|
"step": 2150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1671050595698592,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004727643589180806,
|
||
|
|
"loss": 3.752973175048828,
|
||
|
|
"step": 2160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16787869410490483,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.00047249110239038123,
|
||
|
|
"loss": 3.6907588958740236,
|
||
|
|
"step": 2170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1686523286399505,
|
||
|
|
"grad_norm": 0.2265625,
|
||
|
|
"learning_rate": 0.00047221656613789895,
|
||
|
|
"loss": 3.73446044921875,
|
||
|
|
"step": 2180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16942596317499614,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.00047194075183370864,
|
||
|
|
"loss": 3.679473876953125,
|
||
|
|
"step": 2190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17019959771004178,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004716636611586739,
|
||
|
|
"loss": 3.7257465362548827,
|
||
|
|
"step": 2200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17019959771004178,
|
||
|
|
"eval_loss": 4.27129602432251,
|
||
|
|
"eval_runtime": 8.7096,
|
||
|
|
"eval_samples_per_second": 37.545,
|
||
|
|
"eval_steps_per_second": 1.263,
|
||
|
|
"step": 2200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17097323224508743,
|
||
|
|
"grad_norm": 0.2216796875,
|
||
|
|
"learning_rate": 0.00047138529580143673,
|
||
|
|
"loss": 3.7073875427246095,
|
||
|
|
"step": 2210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17174686678013307,
|
||
|
|
"grad_norm": 0.21875,
|
||
|
|
"learning_rate": 0.0004711056574584075,
|
||
|
|
"loss": 3.71368522644043,
|
||
|
|
"step": 2220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1725205013151787,
|
||
|
|
"grad_norm": 0.2197265625,
|
||
|
|
"learning_rate": 0.00047082474783375396,
|
||
|
|
"loss": 3.7186820983886717,
|
||
|
|
"step": 2230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17329413585022435,
|
||
|
|
"grad_norm": 0.212890625,
|
||
|
|
"learning_rate": 0.00047054256863939177,
|
||
|
|
"loss": 3.705561065673828,
|
||
|
|
"step": 2240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17406777038527,
|
||
|
|
"grad_norm": 0.220703125,
|
||
|
|
"learning_rate": 0.0004702591215949732,
|
||
|
|
"loss": 3.709010696411133,
|
||
|
|
"step": 2250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17484140492031564,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.00046997440842787725,
|
||
|
|
"loss": 3.7188438415527343,
|
||
|
|
"step": 2260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17561503945536128,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004696884308731988,
|
||
|
|
"loss": 3.707378387451172,
|
||
|
|
"step": 2270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17638867399040692,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.00046940119067373823,
|
||
|
|
"loss": 3.697579193115234,
|
||
|
|
"step": 2280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1771623085254526,
|
||
|
|
"grad_norm": 0.2353515625,
|
||
|
|
"learning_rate": 0.0004691126895799907,
|
||
|
|
"loss": 3.7102542877197267,
|
||
|
|
"step": 2290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17793594306049823,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.0004688229293501354,
|
||
|
|
"loss": 3.682720947265625,
|
||
|
|
"step": 2300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17793594306049823,
|
||
|
|
"eval_loss": 4.244110584259033,
|
||
|
|
"eval_runtime": 8.6768,
|
||
|
|
"eval_samples_per_second": 37.687,
|
||
|
|
"eval_steps_per_second": 1.268,
|
||
|
|
"step": 2300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17870957759554387,
|
||
|
|
"grad_norm": 0.2001953125,
|
||
|
|
"learning_rate": 0.000468531911750025,
|
||
|
|
"loss": 3.6973262786865235,
|
||
|
|
"step": 2310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17948321213058951,
|
||
|
|
"grad_norm": 0.2265625,
|
||
|
|
"learning_rate": 0.00046823963855317487,
|
||
|
|
"loss": 3.682568359375,
|
||
|
|
"step": 2320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18025684666563516,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.00046794611154075207,
|
||
|
|
"loss": 3.6916324615478517,
|
||
|
|
"step": 2330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1810304812006808,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.0004676513325015648,
|
||
|
|
"loss": 3.6971084594726564,
|
||
|
|
"step": 2340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18180411573572644,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.0004673553032320512,
|
||
|
|
"loss": 3.680620574951172,
|
||
|
|
"step": 2350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18257775027077208,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.00046705802553626875,
|
||
|
|
"loss": 3.6823448181152343,
|
||
|
|
"step": 2360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18335138480581772,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.0004667595012258829,
|
||
|
|
"loss": 3.6814292907714843,
|
||
|
|
"step": 2370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18412501934086337,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004664597321201562,
|
||
|
|
"loss": 3.6852691650390623,
|
||
|
|
"step": 2380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.184898653875909,
|
||
|
|
"grad_norm": 0.2099609375,
|
||
|
|
"learning_rate": 0.0004661587200459373,
|
||
|
|
"loss": 3.7072521209716798,
|
||
|
|
"step": 2390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18567228841095468,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.0004658564668376496,
|
||
|
|
"loss": 3.6736068725585938,
|
||
|
|
"step": 2400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18567228841095468,
|
||
|
|
"eval_loss": 4.224212646484375,
|
||
|
|
"eval_runtime": 8.6258,
|
||
|
|
"eval_samples_per_second": 37.91,
|
||
|
|
"eval_steps_per_second": 1.275,
|
||
|
|
"step": 2400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18644592294600032,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004655529743372805,
|
||
|
|
"loss": 3.674551010131836,
|
||
|
|
"step": 2410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18721955748104596,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.00046524824439436946,
|
||
|
|
"loss": 3.676839065551758,
|
||
|
|
"step": 2420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1879931920160916,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.00046494227886599744,
|
||
|
|
"loss": 3.6566898345947267,
|
||
|
|
"step": 2430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18876682655113725,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.0004646350796167753,
|
||
|
|
"loss": 3.6999538421630858,
|
||
|
|
"step": 2440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1895404610861829,
|
||
|
|
"grad_norm": 0.2099609375,
|
||
|
|
"learning_rate": 0.0004643266485188321,
|
||
|
|
"loss": 3.6581405639648437,
|
||
|
|
"step": 2450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19031409562122853,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.00046401698745180444,
|
||
|
|
"loss": 3.7002613067626955,
|
||
|
|
"step": 2460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19108773015627417,
|
||
|
|
"grad_norm": 0.201171875,
|
||
|
|
"learning_rate": 0.00046370609830282425,
|
||
|
|
"loss": 3.684235763549805,
|
||
|
|
"step": 2470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1918613646913198,
|
||
|
|
"grad_norm": 0.20703125,
|
||
|
|
"learning_rate": 0.00046339398296650787,
|
||
|
|
"loss": 3.6899051666259766,
|
||
|
|
"step": 2480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19263499922636546,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 0.000463080643344944,
|
||
|
|
"loss": 3.6608238220214844,
|
||
|
|
"step": 2490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1934086337614111,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.0004627660813476826,
|
||
|
|
"loss": 3.6702159881591796,
|
||
|
|
"step": 2500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1934086337614111,
|
||
|
|
"eval_loss": 4.213615417480469,
|
||
|
|
"eval_runtime": 8.692,
|
||
|
|
"eval_samples_per_second": 37.621,
|
||
|
|
"eval_steps_per_second": 1.266,
|
||
|
|
"step": 2500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19418226829645677,
|
||
|
|
"grad_norm": 0.2314453125,
|
||
|
|
"learning_rate": 0.0004624502988917229,
|
||
|
|
"loss": 3.6737674713134765,
|
||
|
|
"step": 2510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1949559028315024,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00046213329790150185,
|
||
|
|
"loss": 3.662728500366211,
|
||
|
|
"step": 2520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19572953736654805,
|
||
|
|
"grad_norm": 0.2021484375,
|
||
|
|
"learning_rate": 0.00046181508030888223,
|
||
|
|
"loss": 3.6636001586914064,
|
||
|
|
"step": 2530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1965031719015937,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00046149564805314134,
|
||
|
|
"loss": 3.666353225708008,
|
||
|
|
"step": 2540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19727680643663933,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.00046117500308095853,
|
||
|
|
"loss": 3.670008087158203,
|
||
|
|
"step": 2550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19805044097168498,
|
||
|
|
"grad_norm": 0.2080078125,
|
||
|
|
"learning_rate": 0.000460853147346404,
|
||
|
|
"loss": 3.6724533081054687,
|
||
|
|
"step": 2560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19882407550673062,
|
||
|
|
"grad_norm": 0.236328125,
|
||
|
|
"learning_rate": 0.00046053008281092626,
|
||
|
|
"loss": 3.656665802001953,
|
||
|
|
"step": 2570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19959771004177626,
|
||
|
|
"grad_norm": 0.2060546875,
|
||
|
|
"learning_rate": 0.0004602058114433405,
|
||
|
|
"loss": 3.6695487976074217,
|
||
|
|
"step": 2580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2003713445768219,
|
||
|
|
"grad_norm": 0.2255859375,
|
||
|
|
"learning_rate": 0.0004598803352198169,
|
||
|
|
"loss": 3.6759658813476563,
|
||
|
|
"step": 2590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20114497911186754,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.0004595536561238677,
|
||
|
|
"loss": 3.6256980895996094,
|
||
|
|
"step": 2600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20114497911186754,
|
||
|
|
"eval_loss": 4.193091869354248,
|
||
|
|
"eval_runtime": 8.6853,
|
||
|
|
"eval_samples_per_second": 37.65,
|
||
|
|
"eval_steps_per_second": 1.267,
|
||
|
|
"step": 2600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2019186136469132,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00045922577614633634,
|
||
|
|
"loss": 3.6532318115234377,
|
||
|
|
"step": 2610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20269224818195886,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00045889669728538414,
|
||
|
|
"loss": 3.640882873535156,
|
||
|
|
"step": 2620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2034658827170045,
|
||
|
|
"grad_norm": 0.2216796875,
|
||
|
|
"learning_rate": 0.0004585664215464788,
|
||
|
|
"loss": 3.6541816711425783,
|
||
|
|
"step": 2630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20423951725205014,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.000458234950942382,
|
||
|
|
"loss": 3.62015380859375,
|
||
|
|
"step": 2640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20501315178709578,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004579022874931372,
|
||
|
|
"loss": 3.660184860229492,
|
||
|
|
"step": 2650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20578678632214142,
|
||
|
|
"grad_norm": 0.189453125,
|
||
|
|
"learning_rate": 0.0004575684332260573,
|
||
|
|
"loss": 3.646562957763672,
|
||
|
|
"step": 2660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20656042085718707,
|
||
|
|
"grad_norm": 0.20703125,
|
||
|
|
"learning_rate": 0.00045723339017571214,
|
||
|
|
"loss": 3.66490478515625,
|
||
|
|
"step": 2670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2073340553922327,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.00045689716038391643,
|
||
|
|
"loss": 3.6182834625244142,
|
||
|
|
"step": 2680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20810768992727835,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.00045655974589971677,
|
||
|
|
"loss": 3.638666534423828,
|
||
|
|
"step": 2690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.208881324462324,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.00045622114877937986,
|
||
|
|
"loss": 3.660610580444336,
|
||
|
|
"step": 2700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.208881324462324,
|
||
|
|
"eval_loss": 4.197894096374512,
|
||
|
|
"eval_runtime": 8.6595,
|
||
|
|
"eval_samples_per_second": 37.762,
|
||
|
|
"eval_steps_per_second": 1.27,
|
||
|
|
"step": 2700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20965495899736963,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.00045588137108637934,
|
||
|
|
"loss": 3.6198894500732424,
|
||
|
|
"step": 2710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21042859353241528,
|
||
|
|
"grad_norm": 0.326171875,
|
||
|
|
"learning_rate": 0.00045554041489138367,
|
||
|
|
"loss": 3.651480865478516,
|
||
|
|
"step": 2720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21120222806746092,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004551982822722432,
|
||
|
|
"loss": 3.6471332550048827,
|
||
|
|
"step": 2730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2119758626025066,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.0004548549753139776,
|
||
|
|
"loss": 3.6430343627929687,
|
||
|
|
"step": 2740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21274949713755223,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.0004545104961087634,
|
||
|
|
"loss": 3.628376007080078,
|
||
|
|
"step": 2750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21352313167259787,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.00045416484675592065,
|
||
|
|
"loss": 3.663385009765625,
|
||
|
|
"step": 2760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2142967662076435,
|
||
|
|
"grad_norm": 0.234375,
|
||
|
|
"learning_rate": 0.0004538180293619009,
|
||
|
|
"loss": 3.6373767852783203,
|
||
|
|
"step": 2770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21507040074268916,
|
||
|
|
"grad_norm": 0.2314453125,
|
||
|
|
"learning_rate": 0.0004534700460402737,
|
||
|
|
"loss": 3.626949691772461,
|
||
|
|
"step": 2780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2158440352777348,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.000453120898911714,
|
||
|
|
"loss": 3.621552276611328,
|
||
|
|
"step": 2790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21661766981278044,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.00045277059010398934,
|
||
|
|
"loss": 3.6465320587158203,
|
||
|
|
"step": 2800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21661766981278044,
|
||
|
|
"eval_loss": 4.1806182861328125,
|
||
|
|
"eval_runtime": 8.853,
|
||
|
|
"eval_samples_per_second": 36.937,
|
||
|
|
"eval_steps_per_second": 1.243,
|
||
|
|
"step": 2800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21739130434782608,
|
||
|
|
"grad_norm": 0.2353515625,
|
||
|
|
"learning_rate": 0.0004524191217519466,
|
||
|
|
"loss": 3.646443557739258,
|
||
|
|
"step": 2810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21816493888287172,
|
||
|
|
"grad_norm": 0.2294921875,
|
||
|
|
"learning_rate": 0.00045206649599749914,
|
||
|
|
"loss": 3.6304691314697264,
|
||
|
|
"step": 2820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21893857341791736,
|
||
|
|
"grad_norm": 0.2314453125,
|
||
|
|
"learning_rate": 0.000451712714989614,
|
||
|
|
"loss": 3.629531478881836,
|
||
|
|
"step": 2830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.219712207952963,
|
||
|
|
"grad_norm": 0.22265625,
|
||
|
|
"learning_rate": 0.0004513577808842982,
|
||
|
|
"loss": 3.610256576538086,
|
||
|
|
"step": 2840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22048584248800868,
|
||
|
|
"grad_norm": 0.2216796875,
|
||
|
|
"learning_rate": 0.00045100169584458614,
|
||
|
|
"loss": 3.6422473907470705,
|
||
|
|
"step": 2850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22125947702305432,
|
||
|
|
"grad_norm": 0.23046875,
|
||
|
|
"learning_rate": 0.00045064446204052616,
|
||
|
|
"loss": 3.6097373962402344,
|
||
|
|
"step": 2860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22203311155809996,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004502860816491675,
|
||
|
|
"loss": 3.650621032714844,
|
||
|
|
"step": 2870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2228067460931456,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00044992655685454676,
|
||
|
|
"loss": 3.6009265899658205,
|
||
|
|
"step": 2880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22358038062819124,
|
||
|
|
"grad_norm": 0.2177734375,
|
||
|
|
"learning_rate": 0.0004495658898476749,
|
||
|
|
"loss": 3.6009990692138674,
|
||
|
|
"step": 2890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2243540151632369,
|
||
|
|
"grad_norm": 0.224609375,
|
||
|
|
"learning_rate": 0.00044920408282652356,
|
||
|
|
"loss": 3.6231040954589844,
|
||
|
|
"step": 2900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2243540151632369,
|
||
|
|
"eval_loss": 4.175409317016602,
|
||
|
|
"eval_runtime": 8.6708,
|
||
|
|
"eval_samples_per_second": 37.713,
|
||
|
|
"eval_steps_per_second": 1.269,
|
||
|
|
"step": 2900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22512764969828253,
|
||
|
|
"grad_norm": 0.205078125,
|
||
|
|
"learning_rate": 0.0004488411379960119,
|
||
|
|
"loss": 3.641431427001953,
|
||
|
|
"step": 2910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22590128423332817,
|
||
|
|
"grad_norm": 0.2265625,
|
||
|
|
"learning_rate": 0.0004484770575679933,
|
||
|
|
"loss": 3.6043201446533204,
|
||
|
|
"step": 2920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2266749187683738,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.0004481118437612414,
|
||
|
|
"loss": 3.5940608978271484,
|
||
|
|
"step": 2930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22744855330341945,
|
||
|
|
"grad_norm": 0.2197265625,
|
||
|
|
"learning_rate": 0.00044774549880143694,
|
||
|
|
"loss": 3.632985305786133,
|
||
|
|
"step": 2940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2282221878384651,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004473780249211542,
|
||
|
|
"loss": 3.5861167907714844,
|
||
|
|
"step": 2950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22899582237351077,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004470094243598474,
|
||
|
|
"loss": 3.6033992767333984,
|
||
|
|
"step": 2960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2297694569085564,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.00044663969936383657,
|
||
|
|
"loss": 3.5815731048583985,
|
||
|
|
"step": 2970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23054309144360205,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.00044626885218629467,
|
||
|
|
"loss": 3.6179901123046876,
|
||
|
|
"step": 2980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2313167259786477,
|
||
|
|
"grad_norm": 0.203125,
|
||
|
|
"learning_rate": 0.00044589688508723326,
|
||
|
|
"loss": 3.6054702758789063,
|
||
|
|
"step": 2990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23209036051369333,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.0004455238003334889,
|
||
|
|
"loss": 3.618684005737305,
|
||
|
|
"step": 3000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23209036051369333,
|
||
|
|
"eval_loss": 4.147392272949219,
|
||
|
|
"eval_runtime": 8.7049,
|
||
|
|
"eval_samples_per_second": 37.565,
|
||
|
|
"eval_steps_per_second": 1.264,
|
||
|
|
"step": 3000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23286399504873898,
|
||
|
|
"grad_norm": 0.2236328125,
|
||
|
|
"learning_rate": 0.00044514960019870935,
|
||
|
|
"loss": 3.6089691162109374,
|
||
|
|
"step": 3010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23363762958378462,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.0004447742869633398,
|
||
|
|
"loss": 3.612302780151367,
|
||
|
|
"step": 3020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23441126411883026,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004443978629146089,
|
||
|
|
"loss": 3.6134407043457033,
|
||
|
|
"step": 3030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2351848986538759,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00044402033034651457,
|
||
|
|
"loss": 3.6229934692382812,
|
||
|
|
"step": 3040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23595853318892154,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.00044364169155981044,
|
||
|
|
"loss": 3.6341014862060548,
|
||
|
|
"step": 3050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23673216772396719,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00044326194886199166,
|
||
|
|
"loss": 3.617625427246094,
|
||
|
|
"step": 3060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23750580225901285,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004428811045672808,
|
||
|
|
"loss": 3.5775123596191407,
|
||
|
|
"step": 3070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2382794367940585,
|
||
|
|
"grad_norm": 0.2275390625,
|
||
|
|
"learning_rate": 0.0004424991609966135,
|
||
|
|
"loss": 3.6077903747558593,
|
||
|
|
"step": 3080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23905307132910414,
|
||
|
|
"grad_norm": 0.216796875,
|
||
|
|
"learning_rate": 0.0004421161204776251,
|
||
|
|
"loss": 3.611967849731445,
|
||
|
|
"step": 3090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23982670586414978,
|
||
|
|
"grad_norm": 0.2158203125,
|
||
|
|
"learning_rate": 0.00044173198534463553,
|
||
|
|
"loss": 3.607320022583008,
|
||
|
|
"step": 3100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23982670586414978,
|
||
|
|
"eval_loss": 4.136959075927734,
|
||
|
|
"eval_runtime": 8.702,
|
||
|
|
"eval_samples_per_second": 37.578,
|
||
|
|
"eval_steps_per_second": 1.264,
|
||
|
|
"step": 3100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12030133356542594,
|
||
|
|
"grad_norm": 0.40234375,
|
||
|
|
"learning_rate": 0.00048508648785485544,
|
||
|
|
"loss": 3.5898792266845705,
|
||
|
|
"step": 3110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12068815457367489,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00048498487415469285,
|
||
|
|
"loss": 3.6180370330810545,
|
||
|
|
"step": 3120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12107497558192386,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004848829267533682,
|
||
|
|
"loss": 3.61370849609375,
|
||
|
|
"step": 3130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12146179659017281,
|
||
|
|
"grad_norm": 0.34765625,
|
||
|
|
"learning_rate": 0.0004847806458037906,
|
||
|
|
"loss": 3.632701873779297,
|
||
|
|
"step": 3140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12184861759842178,
|
||
|
|
"grad_norm": 0.296875,
|
||
|
|
"learning_rate": 0.0004846780314593695,
|
||
|
|
"loss": 3.6194690704345702,
|
||
|
|
"step": 3150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12223543860667073,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.0004845750838740143,
|
||
|
|
"loss": 3.6185939788818358,
|
||
|
|
"step": 3160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12262225961491968,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.00048447180320213424,
|
||
|
|
"loss": 3.6234817504882812,
|
||
|
|
"step": 3170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12300908062316865,
|
||
|
|
"grad_norm": 0.3046875,
|
||
|
|
"learning_rate": 0.0004843681895986383,
|
||
|
|
"loss": 3.633131408691406,
|
||
|
|
"step": 3180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1233959016314176,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.00048426424321893467,
|
||
|
|
"loss": 3.6763965606689455,
|
||
|
|
"step": 3190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12378272263966655,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00048415996421893073,
|
||
|
|
"loss": 3.6260562896728517,
|
||
|
|
"step": 3200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12378272263966655,
|
||
|
|
"eval_loss": 4.193098068237305,
|
||
|
|
"eval_runtime": 10.1228,
|
||
|
|
"eval_samples_per_second": 32.303,
|
||
|
|
"eval_steps_per_second": 2.075,
|
||
|
|
"step": 3200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12416954364791552,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.0004840553527550326,
|
||
|
|
"loss": 3.604556655883789,
|
||
|
|
"step": 3210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12455636465616447,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004839504089841453,
|
||
|
|
"loss": 3.6034366607666017,
|
||
|
|
"step": 3220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12494318566441344,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004838451330636721,
|
||
|
|
"loss": 3.6195159912109376,
|
||
|
|
"step": 3230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1253300066726624,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.00048373952515151446,
|
||
|
|
"loss": 3.6002643585205076,
|
||
|
|
"step": 3240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12571682768091136,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00048363358540607206,
|
||
|
|
"loss": 3.6441097259521484,
|
||
|
|
"step": 3250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1261036486891603,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.00048352731398624177,
|
||
|
|
"loss": 3.5857906341552734,
|
||
|
|
"step": 3260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12649046969740926,
|
||
|
|
"grad_norm": 0.302734375,
|
||
|
|
"learning_rate": 0.00048342071105141846,
|
||
|
|
"loss": 3.603062057495117,
|
||
|
|
"step": 3270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12687729070565823,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00048331377676149386,
|
||
|
|
"loss": 3.6648319244384764,
|
||
|
|
"step": 3280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12726411171390717,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00048320651127685687,
|
||
|
|
"loss": 3.589703369140625,
|
||
|
|
"step": 3290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12765093272215614,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004830989147583931,
|
||
|
|
"loss": 3.6127429962158204,
|
||
|
|
"step": 3300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12765093272215614,
|
||
|
|
"eval_loss": 4.191071033477783,
|
||
|
|
"eval_runtime": 9.955,
|
||
|
|
"eval_samples_per_second": 32.848,
|
||
|
|
"eval_steps_per_second": 2.109,
|
||
|
|
"step": 3300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1280377537304051,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00048299098736748474,
|
||
|
|
"loss": 3.6267929077148438,
|
||
|
|
"step": 3310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12842457473865407,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.0004828827292660102,
|
||
|
|
"loss": 3.654584503173828,
|
||
|
|
"step": 3320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.128811395746903,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.0004827741406163438,
|
||
|
|
"loss": 3.642983245849609,
|
||
|
|
"step": 3330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12919821675515197,
|
||
|
|
"grad_norm": 0.30859375,
|
||
|
|
"learning_rate": 0.00048266522158135587,
|
||
|
|
"loss": 3.6135555267333985,
|
||
|
|
"step": 3340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12958503776340094,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00048255597232441214,
|
||
|
|
"loss": 3.655533218383789,
|
||
|
|
"step": 3350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12997185877164988,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00048244639300937356,
|
||
|
|
"loss": 3.620559310913086,
|
||
|
|
"step": 3360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13035867977989885,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004823364838005963,
|
||
|
|
"loss": 3.651724624633789,
|
||
|
|
"step": 3370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1307455007881478,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004822262448629312,
|
||
|
|
"loss": 3.620151901245117,
|
||
|
|
"step": 3380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13113232179639675,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004821156763617238,
|
||
|
|
"loss": 3.657820129394531,
|
||
|
|
"step": 3390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13151914280464572,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004820047784628138,
|
||
|
|
"loss": 3.6077507019042967,
|
||
|
|
"step": 3400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13151914280464572,
|
||
|
|
"eval_loss": 4.1888580322265625,
|
||
|
|
"eval_runtime": 9.9816,
|
||
|
|
"eval_samples_per_second": 32.76,
|
||
|
|
"eval_steps_per_second": 2.104,
|
||
|
|
"step": 3400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13190596381289468,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.000481893551332535,
|
||
|
|
"loss": 3.6212406158447266,
|
||
|
|
"step": 3410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13229278482114365,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004817819951377151,
|
||
|
|
"loss": 3.6335296630859375,
|
||
|
|
"step": 3420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1326796058293926,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0004816701100456752,
|
||
|
|
"loss": 3.587704086303711,
|
||
|
|
"step": 3430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13306642683764155,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00048155789622422984,
|
||
|
|
"loss": 3.6036140441894533,
|
||
|
|
"step": 3440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13345324784589052,
|
||
|
|
"grad_norm": 0.314453125,
|
||
|
|
"learning_rate": 0.00048144535384168647,
|
||
|
|
"loss": 3.5965343475341798,
|
||
|
|
"step": 3450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13384006885413946,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.0004813324830668456,
|
||
|
|
"loss": 3.6275421142578126,
|
||
|
|
"step": 3460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13422688986238843,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00048121928406899995,
|
||
|
|
"loss": 3.646040344238281,
|
||
|
|
"step": 3470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1346137108706374,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00048110575701793484,
|
||
|
|
"loss": 3.568130111694336,
|
||
|
|
"step": 3480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13500053187888633,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0004809919020839274,
|
||
|
|
"loss": 3.634757232666016,
|
||
|
|
"step": 3490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1353873528871353,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004808777194377468,
|
||
|
|
"loss": 3.6164390563964846,
|
||
|
|
"step": 3500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1353873528871353,
|
||
|
|
"eval_loss": 4.202503204345703,
|
||
|
|
"eval_runtime": 9.9788,
|
||
|
|
"eval_samples_per_second": 32.77,
|
||
|
|
"eval_steps_per_second": 2.104,
|
||
|
|
"step": 3500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13577417389538426,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00048076320925065336,
|
||
|
|
"loss": 3.591604995727539,
|
||
|
|
"step": 3510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1361609949036332,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00048064837169439913,
|
||
|
|
"loss": 3.6058467864990233,
|
||
|
|
"step": 3520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13654781591188217,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00048053320694122685,
|
||
|
|
"loss": 3.5918128967285154,
|
||
|
|
"step": 3530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13693463692013114,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.0004804177151638701,
|
||
|
|
"loss": 3.591669464111328,
|
||
|
|
"step": 3540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1373214579283801,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.000480301896535553,
|
||
|
|
"loss": 3.610728073120117,
|
||
|
|
"step": 3550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13770827893662904,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00048018575122998983,
|
||
|
|
"loss": 3.5977199554443358,
|
||
|
|
"step": 3560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.138095099944878,
|
||
|
|
"grad_norm": 0.294921875,
|
||
|
|
"learning_rate": 0.000480069279421385,
|
||
|
|
"loss": 3.6169082641601564,
|
||
|
|
"step": 3570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13848192095312697,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00047995248128443246,
|
||
|
|
"loss": 3.602222442626953,
|
||
|
|
"step": 3580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1388687419613759,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00047983535699431564,
|
||
|
|
"loss": 3.6134090423583984,
|
||
|
|
"step": 3590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13925556296962488,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004797179067267073,
|
||
|
|
"loss": 3.630051040649414,
|
||
|
|
"step": 3600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13925556296962488,
|
||
|
|
"eval_loss": 4.196789741516113,
|
||
|
|
"eval_runtime": 9.999,
|
||
|
|
"eval_samples_per_second": 32.703,
|
||
|
|
"eval_steps_per_second": 2.1,
|
||
|
|
"step": 3600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13964238397787385,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004796001306577691,
|
||
|
|
"loss": 3.5645401000976564,
|
||
|
|
"step": 3610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14002920498612278,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047948202896415105,
|
||
|
|
"loss": 3.6082618713378904,
|
||
|
|
"step": 3620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14041602599437175,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00047936360182299206,
|
||
|
|
"loss": 3.6071346282958983,
|
||
|
|
"step": 3630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14080284700262072,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004792448494119189,
|
||
|
|
"loss": 3.5989688873291015,
|
||
|
|
"step": 3640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14118966801086966,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.000479125771909046,
|
||
|
|
"loss": 3.560350799560547,
|
||
|
|
"step": 3650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14157648901911862,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00047900636949297585,
|
||
|
|
"loss": 3.6439186096191407,
|
||
|
|
"step": 3660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1419633100273676,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00047888664234279797,
|
||
|
|
"loss": 3.6126590728759767,
|
||
|
|
"step": 3670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14235013103561656,
|
||
|
|
"grad_norm": 0.291015625,
|
||
|
|
"learning_rate": 0.000478766590638089,
|
||
|
|
"loss": 3.6033699035644533,
|
||
|
|
"step": 3680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1427369520438655,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004786462145589124,
|
||
|
|
"loss": 3.6775360107421875,
|
||
|
|
"step": 3690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14312377305211446,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00047852551428581813,
|
||
|
|
"loss": 3.6169353485107423,
|
||
|
|
"step": 3700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14312377305211446,
|
||
|
|
"eval_loss": 4.196775436401367,
|
||
|
|
"eval_runtime": 9.9813,
|
||
|
|
"eval_samples_per_second": 32.761,
|
||
|
|
"eval_steps_per_second": 2.104,
|
||
|
|
"step": 3700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14351059406036343,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.0004784044899998425,
|
||
|
|
"loss": 3.5898075103759766,
|
||
|
|
"step": 3710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14389741506861237,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00047828314188250756,
|
||
|
|
"loss": 3.6066890716552735,
|
||
|
|
"step": 3720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14428423607686133,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004781614701158213,
|
||
|
|
"loss": 3.596193313598633,
|
||
|
|
"step": 3730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1446710570851103,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00047803947488227704,
|
||
|
|
"loss": 3.559587860107422,
|
||
|
|
"step": 3740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14505787809335924,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00047791715636485337,
|
||
|
|
"loss": 3.5958782196044923,
|
||
|
|
"step": 3750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1454446991016082,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004777945147470137,
|
||
|
|
"loss": 3.5777961730957033,
|
||
|
|
"step": 3760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14583152010985717,
|
||
|
|
"grad_norm": 0.326171875,
|
||
|
|
"learning_rate": 0.0004776715502127058,
|
||
|
|
"loss": 3.5617481231689454,
|
||
|
|
"step": 3770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14621834111810614,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004775482629463624,
|
||
|
|
"loss": 3.5838111877441405,
|
||
|
|
"step": 3780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14660516212635508,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00047742465313289955,
|
||
|
|
"loss": 3.590658950805664,
|
||
|
|
"step": 3790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14699198313460404,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00047730072095771773,
|
||
|
|
"loss": 3.6110076904296875,
|
||
|
|
"step": 3800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14699198313460404,
|
||
|
|
"eval_loss": 4.168724060058594,
|
||
|
|
"eval_runtime": 9.992,
|
||
|
|
"eval_samples_per_second": 32.726,
|
||
|
|
"eval_steps_per_second": 2.102,
|
||
|
|
"step": 3800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.147378804142853,
|
||
|
|
"grad_norm": 0.296875,
|
||
|
|
"learning_rate": 0.0004771764666067005,
|
||
|
|
"loss": 3.5836368560791017,
|
||
|
|
"step": 3810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14776562515110195,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00047705189026621474,
|
||
|
|
"loss": 3.5947208404541016,
|
||
|
|
"step": 3820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14815244615935091,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004769269921231104,
|
||
|
|
"loss": 3.5821369171142576,
|
||
|
|
"step": 3830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14853926716759988,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00047680177236472004,
|
||
|
|
"loss": 3.5839427947998046,
|
||
|
|
"step": 3840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14892608817584882,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004766762311788585,
|
||
|
|
"loss": 3.555724334716797,
|
||
|
|
"step": 3850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1493129091840978,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004765503687538228,
|
||
|
|
"loss": 3.597983551025391,
|
||
|
|
"step": 3860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14969973019234675,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047642418527839184,
|
||
|
|
"loss": 3.621683120727539,
|
||
|
|
"step": 3870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1500865512005957,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004762976809418259,
|
||
|
|
"loss": 3.5982799530029297,
|
||
|
|
"step": 3880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15047337220884466,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004761708559338668,
|
||
|
|
"loss": 3.5660633087158202,
|
||
|
|
"step": 3890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15086019321709362,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.000476043710444737,
|
||
|
|
"loss": 3.6209995269775392,
|
||
|
|
"step": 3900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15086019321709362,
|
||
|
|
"eval_loss": 4.184715747833252,
|
||
|
|
"eval_runtime": 10.0083,
|
||
|
|
"eval_samples_per_second": 32.673,
|
||
|
|
"eval_steps_per_second": 2.098,
|
||
|
|
"step": 3900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1512470142253426,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004759162446651398,
|
||
|
|
"loss": 3.6069480895996096,
|
||
|
|
"step": 3910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15163383523359153,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047578845878625903,
|
||
|
|
"loss": 3.5772022247314452,
|
||
|
|
"step": 3920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1520206562418405,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047566035299975826,
|
||
|
|
"loss": 3.5811866760253905,
|
||
|
|
"step": 3930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15240747725008946,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004755319274977812,
|
||
|
|
"loss": 3.6295467376708985,
|
||
|
|
"step": 3940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1527942982583384,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.00047540318247295125,
|
||
|
|
"loss": 3.5914112091064454,
|
||
|
|
"step": 3950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15318111926658737,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0004752741181183704,
|
||
|
|
"loss": 3.5724857330322264,
|
||
|
|
"step": 3960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15356794027483633,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00047514473462762034,
|
||
|
|
"loss": 3.592438507080078,
|
||
|
|
"step": 3970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15395476128308527,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004750150321947609,
|
||
|
|
"loss": 3.5568592071533205,
|
||
|
|
"step": 3980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15434158229133424,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00047488501101433065,
|
||
|
|
"loss": 3.589873123168945,
|
||
|
|
"step": 3990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1547284032995832,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00047475467128134586,
|
||
|
|
"loss": 3.5943378448486327,
|
||
|
|
"step": 4000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1547284032995832,
|
||
|
|
"eval_loss": 4.1639275550842285,
|
||
|
|
"eval_runtime": 10.0126,
|
||
|
|
"eval_samples_per_second": 32.659,
|
||
|
|
"eval_steps_per_second": 2.097,
|
||
|
|
"step": 4000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15511522430783214,
|
||
|
|
"grad_norm": 0.291015625,
|
||
|
|
"learning_rate": 0.0004746240131913011,
|
||
|
|
"loss": 3.5330867767333984,
|
||
|
|
"step": 4010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1555020453160811,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004744930369401679,
|
||
|
|
"loss": 3.5745521545410157,
|
||
|
|
"step": 4020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15588886632433008,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047436174272439535,
|
||
|
|
"loss": 3.5994407653808596,
|
||
|
|
"step": 4030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15627568733257904,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00047423013074090933,
|
||
|
|
"loss": 3.5939708709716798,
|
||
|
|
"step": 4040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15666250834082798,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004740982011871123,
|
||
|
|
"loss": 3.5709758758544923,
|
||
|
|
"step": 4050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15704932934907695,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00047396595426088317,
|
||
|
|
"loss": 3.596474456787109,
|
||
|
|
"step": 4060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15743615035732592,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00047383339016057675,
|
||
|
|
"loss": 3.6182701110839846,
|
||
|
|
"step": 4070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15782297136557485,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.00047370050908502367,
|
||
|
|
"loss": 3.5875564575195313,
|
||
|
|
"step": 4080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15820979237382382,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004735673112335297,
|
||
|
|
"loss": 3.5575897216796877,
|
||
|
|
"step": 4090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1585966133820728,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.0004734337968058762,
|
||
|
|
"loss": 3.5939361572265627,
|
||
|
|
"step": 4100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1585966133820728,
|
||
|
|
"eval_loss": 4.146914958953857,
|
||
|
|
"eval_runtime": 10.0531,
|
||
|
|
"eval_samples_per_second": 32.527,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 4100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15898343439032173,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00047329996600231904,
|
||
|
|
"loss": 3.564817428588867,
|
||
|
|
"step": 4110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1593702553985707,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00047316581902358875,
|
||
|
|
"loss": 3.6007923126220702,
|
||
|
|
"step": 4120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15975707640681966,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00047303135607088995,
|
||
|
|
"loss": 3.6004383087158205,
|
||
|
|
"step": 4130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16014389741506863,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004728965773459013,
|
||
|
|
"loss": 3.631551742553711,
|
||
|
|
"step": 4140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16053071842331756,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004727614830507749,
|
||
|
|
"loss": 3.5560813903808595,
|
||
|
|
"step": 4150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16091753943156653,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00047262607338813663,
|
||
|
|
"loss": 3.601303482055664,
|
||
|
|
"step": 4160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1613043604398155,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004724903485610848,
|
||
|
|
"loss": 3.589255523681641,
|
||
|
|
"step": 4170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16169118144806444,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00047235430877319085,
|
||
|
|
"loss": 3.5800621032714846,
|
||
|
|
"step": 4180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1620780024563134,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00047221795422849845,
|
||
|
|
"loss": 3.6015853881835938,
|
||
|
|
"step": 4190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16246482346456237,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004720812851315233,
|
||
|
|
"loss": 3.5746341705322267,
|
||
|
|
"step": 4200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16246482346456237,
|
||
|
|
"eval_loss": 4.151747226715088,
|
||
|
|
"eval_runtime": 10.061,
|
||
|
|
"eval_samples_per_second": 32.502,
|
||
|
|
"eval_steps_per_second": 2.087,
|
||
|
|
"step": 4200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1628516444728113,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.000471944301687253,
|
||
|
|
"loss": 3.585940933227539,
|
||
|
|
"step": 4210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16323846548106027,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004718070041011467,
|
||
|
|
"loss": 3.5748485565185546,
|
||
|
|
"step": 4220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16362528648930924,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004716693925791346,
|
||
|
|
"loss": 3.604657745361328,
|
||
|
|
"step": 4230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16401210749755818,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00047153146732761776,
|
||
|
|
"loss": 3.570362091064453,
|
||
|
|
"step": 4240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16439892850580715,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.00047139322855346794,
|
||
|
|
"loss": 3.5857212066650392,
|
||
|
|
"step": 4250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1647857495140561,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000471254676464027,
|
||
|
|
"loss": 3.5793018341064453,
|
||
|
|
"step": 4260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16517257052230508,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00047111581126710675,
|
||
|
|
"loss": 3.617490768432617,
|
||
|
|
"step": 4270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16555939153055402,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047097663317098874,
|
||
|
|
"loss": 3.5748958587646484,
|
||
|
|
"step": 4280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16594621253880298,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004708371423844237,
|
||
|
|
"loss": 3.5808582305908203,
|
||
|
|
"step": 4290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16633303354705195,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004706973391166315,
|
||
|
|
"loss": 3.565126419067383,
|
||
|
|
"step": 4300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16633303354705195,
|
||
|
|
"eval_loss": 4.144286155700684,
|
||
|
|
"eval_runtime": 10.0708,
|
||
|
|
"eval_samples_per_second": 32.47,
|
||
|
|
"eval_steps_per_second": 2.085,
|
||
|
|
"step": 4300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1667198545553009,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00047055722357730053,
|
||
|
|
"loss": 3.601681137084961,
|
||
|
|
"step": 4310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16710667556354986,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00047041679597658773,
|
||
|
|
"loss": 3.619378662109375,
|
||
|
|
"step": 4320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16749349657179882,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004702760565251178,
|
||
|
|
"loss": 3.5644298553466798,
|
||
|
|
"step": 4330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16788031758004776,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004701350054339835,
|
||
|
|
"loss": 3.5358177185058595,
|
||
|
|
"step": 4340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16826713858829673,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00046999364291474486,
|
||
|
|
"loss": 3.5819393157958985,
|
||
|
|
"step": 4350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1686539595965457,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00046985196917942923,
|
||
|
|
"loss": 3.6105266571044923,
|
||
|
|
"step": 4360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16904078060479463,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00046970998444053025,
|
||
|
|
"loss": 3.5382503509521483,
|
||
|
|
"step": 4370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1694276016130436,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.0004695676889110086,
|
||
|
|
"loss": 3.5398067474365233,
|
||
|
|
"step": 4380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16981442262129257,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004694250828042906,
|
||
|
|
"loss": 3.573816680908203,
|
||
|
|
"step": 4390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17020124362954153,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004692821663342689,
|
||
|
|
"loss": 3.5996776580810548,
|
||
|
|
"step": 4400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17020124362954153,
|
||
|
|
"eval_loss": 4.145204544067383,
|
||
|
|
"eval_runtime": 10.0751,
|
||
|
|
"eval_samples_per_second": 32.456,
|
||
|
|
"eval_steps_per_second": 2.084,
|
||
|
|
"step": 4400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17058806463779047,
|
||
|
|
"grad_norm": 0.291015625,
|
||
|
|
"learning_rate": 0.0004691389397153013,
|
||
|
|
"loss": 3.5502437591552733,
|
||
|
|
"step": 4410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17097488564603944,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00046899540316221086,
|
||
|
|
"loss": 3.591476821899414,
|
||
|
|
"step": 4420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1713617066542884,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004688515568902855,
|
||
|
|
"loss": 3.563010406494141,
|
||
|
|
"step": 4430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17174852766253734,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004687074011152779,
|
||
|
|
"loss": 3.5925624847412108,
|
||
|
|
"step": 4440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1721353486707863,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00046856293605340466,
|
||
|
|
"loss": 3.576287841796875,
|
||
|
|
"step": 4450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17252216967903528,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00046841816192134635,
|
||
|
|
"loss": 3.5883880615234376,
|
||
|
|
"step": 4460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17290899068728421,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00046827307893624725,
|
||
|
|
"loss": 3.557068634033203,
|
||
|
|
"step": 4470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17329581169553318,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004681276873157147,
|
||
|
|
"loss": 3.5829254150390626,
|
||
|
|
"step": 4480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17368263270378215,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00046798198727781914,
|
||
|
|
"loss": 3.5822193145751955,
|
||
|
|
"step": 4490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1740694537120311,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004678359790410934,
|
||
|
|
"loss": 3.570149230957031,
|
||
|
|
"step": 4500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1740694537120311,
|
||
|
|
"eval_loss": 4.142066955566406,
|
||
|
|
"eval_runtime": 10.0399,
|
||
|
|
"eval_samples_per_second": 32.57,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 4500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17445627472028005,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.00046768966282453276,
|
||
|
|
"loss": 3.5578567504882814,
|
||
|
|
"step": 4510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17484309572852902,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00046754303884759436,
|
||
|
|
"loss": 3.614657974243164,
|
||
|
|
"step": 4520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17522991673677799,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004673961073301968,
|
||
|
|
"loss": 3.5728389739990236,
|
||
|
|
"step": 4530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17561673774502692,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.0004672488684927202,
|
||
|
|
"loss": 3.5710781097412108,
|
||
|
|
"step": 4540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1760035587532759,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004671013225560054,
|
||
|
|
"loss": 3.5570396423339843,
|
||
|
|
"step": 4550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17639037976152486,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00046695346974135414,
|
||
|
|
"loss": 3.571766662597656,
|
||
|
|
"step": 4560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1767772007697738,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004668053102705281,
|
||
|
|
"loss": 3.5966800689697265,
|
||
|
|
"step": 4570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17716402177802276,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004666568443657492,
|
||
|
|
"loss": 3.5605045318603517,
|
||
|
|
"step": 4580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17755084278627173,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004665080722496988,
|
||
|
|
"loss": 3.546836090087891,
|
||
|
|
"step": 4590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17793766379452067,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004663589941455176,
|
||
|
|
"loss": 3.5560222625732423,
|
||
|
|
"step": 4600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17793766379452067,
|
||
|
|
"eval_loss": 4.123315811157227,
|
||
|
|
"eval_runtime": 10.0373,
|
||
|
|
"eval_samples_per_second": 32.579,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 4600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17832448480276963,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004662096102768052,
|
||
|
|
"loss": 3.5690940856933593,
|
||
|
|
"step": 4610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1787113058110186,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004660599208676198,
|
||
|
|
"loss": 3.566990280151367,
|
||
|
|
"step": 4620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17909812681926757,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.000465909926142478,
|
||
|
|
"loss": 3.563935089111328,
|
||
|
|
"step": 4630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1794849478275165,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004657596263263542,
|
||
|
|
"loss": 3.535744857788086,
|
||
|
|
"step": 4640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17987176883576547,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00046560902164468053,
|
||
|
|
"loss": 3.558307647705078,
|
||
|
|
"step": 4650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18025858984401444,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004654581123233464,
|
||
|
|
"loss": 3.5660072326660157,
|
||
|
|
"step": 4660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18064541085226338,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004653068985886977,
|
||
|
|
"loss": 3.5773242950439452,
|
||
|
|
"step": 4670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18103223186051234,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00046515538066753767,
|
||
|
|
"loss": 3.5603591918945314,
|
||
|
|
"step": 4680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1814190528687613,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004650035587871252,
|
||
|
|
"loss": 3.5422637939453123,
|
||
|
|
"step": 4690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18180587387701025,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00046485143317517516,
|
||
|
|
"loss": 3.5612777709960937,
|
||
|
|
"step": 4700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18180587387701025,
|
||
|
|
"eval_loss": 4.127031326293945,
|
||
|
|
"eval_runtime": 10.0536,
|
||
|
|
"eval_samples_per_second": 32.526,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 4700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18219269488525922,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004646990040598583,
|
||
|
|
"loss": 3.562766265869141,
|
||
|
|
"step": 4710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18257951589350818,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00046454627166980024,
|
||
|
|
"loss": 3.5492530822753907,
|
||
|
|
"step": 4720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18296633690175712,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004643932362340817,
|
||
|
|
"loss": 3.55328254699707,
|
||
|
|
"step": 4730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1833531579100061,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004642398979822377,
|
||
|
|
"loss": 3.553334426879883,
|
||
|
|
"step": 4740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18373997891825505,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004640862571442578,
|
||
|
|
"loss": 3.556576156616211,
|
||
|
|
"step": 4750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18412679992650402,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00046393231395058505,
|
||
|
|
"loss": 3.5628944396972657,
|
||
|
|
"step": 4760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18451362093475296,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004637780686321162,
|
||
|
|
"loss": 3.5898536682128905,
|
||
|
|
"step": 4770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18490044194300193,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00046362352142020105,
|
||
|
|
"loss": 3.578490447998047,
|
||
|
|
"step": 4780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1852872629512509,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004634686725466424,
|
||
|
|
"loss": 3.546381378173828,
|
||
|
|
"step": 4790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18567408395949983,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004633135222436951,
|
||
|
|
"loss": 3.5513580322265623,
|
||
|
|
"step": 4800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18567408395949983,
|
||
|
|
"eval_loss": 4.12771463394165,
|
||
|
|
"eval_runtime": 10.0643,
|
||
|
|
"eval_samples_per_second": 32.491,
|
||
|
|
"eval_steps_per_second": 2.087,
|
||
|
|
"step": 4800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1860609049677488,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004631580707440666,
|
||
|
|
"loss": 3.5534221649169924,
|
||
|
|
"step": 4810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18644772597599776,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00046300231828091585,
|
||
|
|
"loss": 3.5446445465087892,
|
||
|
|
"step": 4820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1868345469842467,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00046284626508785314,
|
||
|
|
"loss": 3.5525463104248045,
|
||
|
|
"step": 4830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18722136799249567,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00046268991139893995,
|
||
|
|
"loss": 3.5511096954345702,
|
||
|
|
"step": 4840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18760818900074464,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004625332574486885,
|
||
|
|
"loss": 3.5215126037597657,
|
||
|
|
"step": 4850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1879950100089936,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004623763034720613,
|
||
|
|
"loss": 3.546421432495117,
|
||
|
|
"step": 4860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18838183101724254,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004622190497044707,
|
||
|
|
"loss": 3.551416778564453,
|
||
|
|
"step": 4870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1887686520254915,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004620614963817791,
|
||
|
|
"loss": 3.605236053466797,
|
||
|
|
"step": 4880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18915547303374047,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00046190364374029783,
|
||
|
|
"loss": 3.5282737731933596,
|
||
|
|
"step": 4890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1895422940419894,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00046174549201678734,
|
||
|
|
"loss": 3.544286346435547,
|
||
|
|
"step": 4900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1895422940419894,
|
||
|
|
"eval_loss": 4.114104270935059,
|
||
|
|
"eval_runtime": 9.9575,
|
||
|
|
"eval_samples_per_second": 32.84,
|
||
|
|
"eval_steps_per_second": 2.109,
|
||
|
|
"step": 4900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18992911505023838,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00046158704144845666,
|
||
|
|
"loss": 3.5649555206298826,
|
||
|
|
"step": 4910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19031593605848734,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.00046142829227296294,
|
||
|
|
"loss": 3.5907196044921874,
|
||
|
|
"step": 4920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19070275706673628,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004612692447284113,
|
||
|
|
"loss": 3.5507247924804686,
|
||
|
|
"step": 4930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19108957807498525,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.00046110989905335435,
|
||
|
|
"loss": 3.5793888092041017,
|
||
|
|
"step": 4940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19147639908323422,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.000460950255486792,
|
||
|
|
"loss": 3.5635608673095702,
|
||
|
|
"step": 4950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19186322009148316,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00046079031426817077,
|
||
|
|
"loss": 3.5769954681396485,
|
||
|
|
"step": 4960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19225004109973212,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004606300756373836,
|
||
|
|
"loss": 3.543282699584961,
|
||
|
|
"step": 4970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1926368621079811,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00046046953983476963,
|
||
|
|
"loss": 3.534237289428711,
|
||
|
|
"step": 4980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19302368311623005,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00046030870710111387,
|
||
|
|
"loss": 3.5499801635742188,
|
||
|
|
"step": 4990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.193410504124479,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004601475776776463,
|
||
|
|
"loss": 3.551239013671875,
|
||
|
|
"step": 5000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.193410504124479,
|
||
|
|
"eval_loss": 4.095266342163086,
|
||
|
|
"eval_runtime": 10.0836,
|
||
|
|
"eval_samples_per_second": 32.429,
|
||
|
|
"eval_steps_per_second": 2.083,
|
||
|
|
"step": 5000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19379732513272796,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004599861518060422,
|
||
|
|
"loss": 3.5352596282958983,
|
||
|
|
"step": 5010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19418414614097693,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004598244297284214,
|
||
|
|
"loss": 3.5772289276123046,
|
||
|
|
"step": 5020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19457096714922587,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00045966241168734806,
|
||
|
|
"loss": 3.574796676635742,
|
||
|
|
"step": 5030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19495778815747483,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00045950009792583015,
|
||
|
|
"loss": 3.5158466339111327,
|
||
|
|
"step": 5040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1953446091657238,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.00045933748868731916,
|
||
|
|
"loss": 3.533673858642578,
|
||
|
|
"step": 5050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19573143017397274,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.00045917458421571,
|
||
|
|
"loss": 3.5585952758789063,
|
||
|
|
"step": 5060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1961182511822217,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004590113847553402,
|
||
|
|
"loss": 3.532826232910156,
|
||
|
|
"step": 5070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19650507219047067,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00045884789055098963,
|
||
|
|
"loss": 3.566379928588867,
|
||
|
|
"step": 5080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1968918931987196,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00045868410184788045,
|
||
|
|
"loss": 3.5672889709472657,
|
||
|
|
"step": 5090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19727871420696858,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.00045852001889167655,
|
||
|
|
"loss": 3.5408248901367188,
|
||
|
|
"step": 5100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19727871420696858,
|
||
|
|
"eval_loss": 4.099873065948486,
|
||
|
|
"eval_runtime": 10.0832,
|
||
|
|
"eval_samples_per_second": 32.43,
|
||
|
|
"eval_steps_per_second": 2.083,
|
||
|
|
"step": 5100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19766553521521754,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00045835564192848294,
|
||
|
|
"loss": 3.5583164215087892,
|
||
|
|
"step": 5110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1980523562234665,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00045819097120484585,
|
||
|
|
"loss": 3.556626892089844,
|
||
|
|
"step": 5120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19843917723171545,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00045802600696775195,
|
||
|
|
"loss": 3.545968246459961,
|
||
|
|
"step": 5130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1988259982399644,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004578607494646283,
|
||
|
|
"loss": 3.536521148681641,
|
||
|
|
"step": 5140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19921281924821338,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00045769519894334166,
|
||
|
|
"loss": 3.571156692504883,
|
||
|
|
"step": 5150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19959964025646232,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00045752935565219826,
|
||
|
|
"loss": 3.5418251037597654,
|
||
|
|
"step": 5160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19998646126471128,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004573632198399437,
|
||
|
|
"loss": 3.5485843658447265,
|
||
|
|
"step": 5170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20037328227296025,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004571967917557621,
|
||
|
|
"loss": 3.5757156372070313,
|
||
|
|
"step": 5180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2007601032812092,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00045703007164927606,
|
||
|
|
"loss": 3.5134517669677736,
|
||
|
|
"step": 5190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20114692428945816,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00045686305977054616,
|
||
|
|
"loss": 3.5113853454589843,
|
||
|
|
"step": 5200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20114692428945816,
|
||
|
|
"eval_loss": 4.089205741882324,
|
||
|
|
"eval_runtime": 10.0508,
|
||
|
|
"eval_samples_per_second": 32.535,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 5200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20153374529770712,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.00045669575637007054,
|
||
|
|
"loss": 3.5599761962890626,
|
||
|
|
"step": 5210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2019205663059561,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00045652816169878467,
|
||
|
|
"loss": 3.5213069915771484,
|
||
|
|
"step": 5220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20230738731420503,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004563602760080608,
|
||
|
|
"loss": 3.510356140136719,
|
||
|
|
"step": 5230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.202694208322454,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004561920995497078,
|
||
|
|
"loss": 3.5470203399658202,
|
||
|
|
"step": 5240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20308102933070296,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004560236325759705,
|
||
|
|
"loss": 3.5547637939453125,
|
||
|
|
"step": 5250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2034678503389519,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00045585487533952976,
|
||
|
|
"loss": 3.5266292572021483,
|
||
|
|
"step": 5260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20385467134720087,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004556858280935013,
|
||
|
|
"loss": 3.5063777923583985,
|
||
|
|
"step": 5270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20424149235544983,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00045551649109143644,
|
||
|
|
"loss": 3.5121044158935546,
|
||
|
|
"step": 5280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20462831336369877,
|
||
|
|
"grad_norm": 0.29296875,
|
||
|
|
"learning_rate": 0.00045534686458732055,
|
||
|
|
"loss": 3.539692687988281,
|
||
|
|
"step": 5290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20501513437194774,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004551769488355736,
|
||
|
|
"loss": 3.555554962158203,
|
||
|
|
"step": 5300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20501513437194774,
|
||
|
|
"eval_loss": 4.082935333251953,
|
||
|
|
"eval_runtime": 10.0668,
|
||
|
|
"eval_samples_per_second": 32.483,
|
||
|
|
"eval_steps_per_second": 2.086,
|
||
|
|
"step": 5300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2054019553801967,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00045500674409104907,
|
||
|
|
"loss": 3.5431640625,
|
||
|
|
"step": 5310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20578877638844564,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004548362506090342,
|
||
|
|
"loss": 3.528326416015625,
|
||
|
|
"step": 5320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2061755973966946,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.0004546654686452493,
|
||
|
|
"loss": 3.5564811706542967,
|
||
|
|
"step": 5330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20656241840494358,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00045449439845584693,
|
||
|
|
"loss": 3.5544174194335936,
|
||
|
|
"step": 5340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20694923941319254,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00045432304029741245,
|
||
|
|
"loss": 3.493732452392578,
|
||
|
|
"step": 5350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20733606042144148,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00045415139442696296,
|
||
|
|
"loss": 3.5211856842041014,
|
||
|
|
"step": 5360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20772288142969045,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004539794611019471,
|
||
|
|
"loss": 3.521197509765625,
|
||
|
|
"step": 5370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20810970243793941,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00045380724058024466,
|
||
|
|
"loss": 3.5391448974609374,
|
||
|
|
"step": 5380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20849652344618835,
|
||
|
|
"grad_norm": 0.3046875,
|
||
|
|
"learning_rate": 0.0004536347331201661,
|
||
|
|
"loss": 3.550646209716797,
|
||
|
|
"step": 5390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20888334445443732,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00045346193898045237,
|
||
|
|
"loss": 3.5525577545166014,
|
||
|
|
"step": 5400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20888334445443732,
|
||
|
|
"eval_loss": 4.085052490234375,
|
||
|
|
"eval_runtime": 10.0705,
|
||
|
|
"eval_samples_per_second": 32.471,
|
||
|
|
"eval_steps_per_second": 2.085,
|
||
|
|
"step": 5400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2092701654626863,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004532888584202743,
|
||
|
|
"loss": 3.520841598510742,
|
||
|
|
"step": 5410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20965698647093522,
|
||
|
|
"grad_norm": 0.291015625,
|
||
|
|
"learning_rate": 0.0004531154916992326,
|
||
|
|
"loss": 3.5018367767333984,
|
||
|
|
"step": 5420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2100438074791842,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00045294183907735675,
|
||
|
|
"loss": 3.531473922729492,
|
||
|
|
"step": 5430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21043062848743316,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004527679008151054,
|
||
|
|
"loss": 3.554092788696289,
|
||
|
|
"step": 5440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21081744949568212,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00045259367717336545,
|
||
|
|
"loss": 3.5328224182128904,
|
||
|
|
"step": 5450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21120427050393106,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00045241916841345193,
|
||
|
|
"loss": 3.5440658569335937,
|
||
|
|
"step": 5460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21159109151218003,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00045224437479710744,
|
||
|
|
"loss": 3.5399906158447267,
|
||
|
|
"step": 5470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.211977912520429,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00045206929658650187,
|
||
|
|
"loss": 3.533125305175781,
|
||
|
|
"step": 5480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21236473352867793,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.00045189393404423197,
|
||
|
|
"loss": 3.531673812866211,
|
||
|
|
"step": 5490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2127515545369269,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.00045171828743332084,
|
||
|
|
"loss": 3.516398620605469,
|
||
|
|
"step": 5500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2127515545369269,
|
||
|
|
"eval_loss": 4.070333957672119,
|
||
|
|
"eval_runtime": 10.0591,
|
||
|
|
"eval_samples_per_second": 32.508,
|
||
|
|
"eval_steps_per_second": 2.088,
|
||
|
|
"step": 5500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21313837554517587,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00045154235701721785,
|
||
|
|
"loss": 3.5363746643066407,
|
||
|
|
"step": 5510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2135251965534248,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00045136614305979803,
|
||
|
|
"loss": 3.5801647186279295,
|
||
|
|
"step": 5520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21391201756167377,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00045118964582536135,
|
||
|
|
"loss": 3.555461120605469,
|
||
|
|
"step": 5530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21429883856992274,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00045101286557863317,
|
||
|
|
"loss": 3.507638168334961,
|
||
|
|
"step": 5540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21468565957817168,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00045083580258476304,
|
||
|
|
"loss": 3.5301235198974608,
|
||
|
|
"step": 5550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21507248058642064,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004506584571093245,
|
||
|
|
"loss": 3.5175750732421873,
|
||
|
|
"step": 5560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2154593015946696,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000450480829418315,
|
||
|
|
"loss": 3.504682159423828,
|
||
|
|
"step": 5570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21584612260291858,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00045030291977815525,
|
||
|
|
"loss": 3.527938461303711,
|
||
|
|
"step": 5580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21623294361116752,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00045012472845568877,
|
||
|
|
"loss": 3.536154937744141,
|
||
|
|
"step": 5590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21661976461941648,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.00044994625571818155,
|
||
|
|
"loss": 3.54614143371582,
|
||
|
|
"step": 5600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21661976461941648,
|
||
|
|
"eval_loss": 4.063368320465088,
|
||
|
|
"eval_runtime": 10.054,
|
||
|
|
"eval_samples_per_second": 32.524,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 5600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21700658562766545,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00044976750183332165,
|
||
|
|
"loss": 3.5452247619628907,
|
||
|
|
"step": 5610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2173934066359144,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.00044958846706921914,
|
||
|
|
"loss": 3.5369388580322267,
|
||
|
|
"step": 5620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21778022764416335,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.00044940915169440493,
|
||
|
|
"loss": 3.5290111541748046,
|
||
|
|
"step": 5630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21816704865241232,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004492295559778311,
|
||
|
|
"loss": 3.5256011962890623,
|
||
|
|
"step": 5640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21855386966066126,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00044904968018887,
|
||
|
|
"loss": 3.537184143066406,
|
||
|
|
"step": 5650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21894069066891023,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004488695245973143,
|
||
|
|
"loss": 3.5146228790283205,
|
||
|
|
"step": 5660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2193275116771592,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0004486890894733762,
|
||
|
|
"loss": 3.4976150512695314,
|
||
|
|
"step": 5670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21971433268540813,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.00044850837508768706,
|
||
|
|
"loss": 3.51986083984375,
|
||
|
|
"step": 5680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2201011536936571,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0004483273817112973,
|
||
|
|
"loss": 3.507583236694336,
|
||
|
|
"step": 5690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22048797470190606,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.00044814610961567566,
|
||
|
|
"loss": 3.5720191955566407,
|
||
|
|
"step": 5700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22048797470190606,
|
||
|
|
"eval_loss": 4.055907249450684,
|
||
|
|
"eval_runtime": 10.0576,
|
||
|
|
"eval_samples_per_second": 32.513,
|
||
|
|
"eval_steps_per_second": 2.088,
|
||
|
|
"step": 5700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22087479571015503,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00044796455907270905,
|
||
|
|
"loss": 3.508076477050781,
|
||
|
|
"step": 5710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22126161671840397,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.000447782730354702,
|
||
|
|
"loss": 3.5065277099609373,
|
||
|
|
"step": 5720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22164843772665294,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00044760062373437604,
|
||
|
|
"loss": 3.5399009704589846,
|
||
|
|
"step": 5730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2220352587349019,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004474182394848698,
|
||
|
|
"loss": 3.561381149291992,
|
||
|
|
"step": 5740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22242207974315084,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00044723557787973824,
|
||
|
|
"loss": 3.4996368408203127,
|
||
|
|
"step": 5750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2228089007513998,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004470526391929523,
|
||
|
|
"loss": 3.500721740722656,
|
||
|
|
"step": 5760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22319572175964877,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004468694236988985,
|
||
|
|
"loss": 3.485240936279297,
|
||
|
|
"step": 5770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2235825427678977,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00044668593167237846,
|
||
|
|
"loss": 3.5129344940185545,
|
||
|
|
"step": 5780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22396936377614668,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00044650216338860873,
|
||
|
|
"loss": 3.4915618896484375,
|
||
|
|
"step": 5790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22435618478439565,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004463181191232202,
|
||
|
|
"loss": 3.551679229736328,
|
||
|
|
"step": 5800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22435618478439565,
|
||
|
|
"eval_loss": 4.072558879852295,
|
||
|
|
"eval_runtime": 10.1948,
|
||
|
|
"eval_samples_per_second": 32.075,
|
||
|
|
"eval_steps_per_second": 2.06,
|
||
|
|
"step": 5800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2247430057926446,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004461337991522574,
|
||
|
|
"loss": 3.5697952270507813,
|
||
|
|
"step": 5810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22512982680089355,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00044594920375217893,
|
||
|
|
"loss": 3.513915252685547,
|
||
|
|
"step": 5820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22551664780914252,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.000445764333199856,
|
||
|
|
"loss": 3.4932785034179688,
|
||
|
|
"step": 5830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22590346881739148,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004455791877725728,
|
||
|
|
"loss": 3.512582778930664,
|
||
|
|
"step": 5840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22629028982564042,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004453937677480257,
|
||
|
|
"loss": 3.4789451599121093,
|
||
|
|
"step": 5850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2266771108338894,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004452080734043228,
|
||
|
|
"loss": 3.5080036163330077,
|
||
|
|
"step": 5860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22706393184213836,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0004450221050199841,
|
||
|
|
"loss": 3.5435367584228517,
|
||
|
|
"step": 5870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2274507528503873,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00044483586287394,
|
||
|
|
"loss": 3.5231910705566407,
|
||
|
|
"step": 5880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22783757385863626,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004446493472455319,
|
||
|
|
"loss": 3.475141906738281,
|
||
|
|
"step": 5890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22822439486688523,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004444625584145114,
|
||
|
|
"loss": 3.4972454071044923,
|
||
|
|
"step": 5900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22822439486688523,
|
||
|
|
"eval_loss": 4.064547061920166,
|
||
|
|
"eval_runtime": 10.2047,
|
||
|
|
"eval_samples_per_second": 32.044,
|
||
|
|
"eval_steps_per_second": 2.058,
|
||
|
|
"step": 5900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22861121587513417,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004442754966610398,
|
||
|
|
"loss": 3.512071228027344,
|
||
|
|
"step": 5910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22899803688338313,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004440881622656878,
|
||
|
|
"loss": 3.4975730895996096,
|
||
|
|
"step": 5920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2293848578916321,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004439005555094348,
|
||
|
|
"loss": 3.477853012084961,
|
||
|
|
"step": 5930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22977167889988107,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00044371267667366894,
|
||
|
|
"loss": 3.4883792877197264,
|
||
|
|
"step": 5940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23015849990813,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.0004435245260401865,
|
||
|
|
"loss": 3.5253154754638674,
|
||
|
|
"step": 5950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23054532091637897,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004433361038911912,
|
||
|
|
"loss": 3.518881988525391,
|
||
|
|
"step": 5960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23093214192462794,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00044314741050929406,
|
||
|
|
"loss": 3.510706329345703,
|
||
|
|
"step": 5970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23131896293287688,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00044295844617751325,
|
||
|
|
"loss": 3.5086299896240236,
|
||
|
|
"step": 5980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23170578394112584,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00044276921117927283,
|
||
|
|
"loss": 3.5073486328125,
|
||
|
|
"step": 5990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2320926049493748,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004425797057984031,
|
||
|
|
"loss": 3.537984085083008,
|
||
|
|
"step": 6000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2320926049493748,
|
||
|
|
"eval_loss": 4.064856052398682,
|
||
|
|
"eval_runtime": 10.2355,
|
||
|
|
"eval_samples_per_second": 31.947,
|
||
|
|
"eval_steps_per_second": 2.052,
|
||
|
|
"step": 6000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23247942595762375,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004423899303191399,
|
||
|
|
"loss": 3.527267074584961,
|
||
|
|
"step": 6010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23286624696587271,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000442199885026124,
|
||
|
|
"loss": 3.499089813232422,
|
||
|
|
"step": 6020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23325306797412168,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00044200957020440126,
|
||
|
|
"loss": 3.498735046386719,
|
||
|
|
"step": 6030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23363988898237062,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004418189861394213,
|
||
|
|
"loss": 3.5322608947753906,
|
||
|
|
"step": 6040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23402670999061959,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004416281331170379,
|
||
|
|
"loss": 3.5282882690429687,
|
||
|
|
"step": 6050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23441353099886855,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004414370114235082,
|
||
|
|
"loss": 3.5092117309570314,
|
||
|
|
"step": 6060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23480035200711752,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004412456213454923,
|
||
|
|
"loss": 3.5284473419189455,
|
||
|
|
"step": 6070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23518717301536646,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00044105396317005273,
|
||
|
|
"loss": 3.5273303985595703,
|
||
|
|
"step": 6080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23557399402361542,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004408620371846543,
|
||
|
|
"loss": 3.5467586517333984,
|
||
|
|
"step": 6090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2359608150318644,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00044066984367716343,
|
||
|
|
"loss": 3.5320247650146483,
|
||
|
|
"step": 6100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2359608150318644,
|
||
|
|
"eval_loss": 4.048551082611084,
|
||
|
|
"eval_runtime": 10.2354,
|
||
|
|
"eval_samples_per_second": 31.948,
|
||
|
|
"eval_steps_per_second": 2.052,
|
||
|
|
"step": 6100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23634763604011333,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004404773829358478,
|
||
|
|
"loss": 3.526265335083008,
|
||
|
|
"step": 6110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2367344570483623,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00044028465524937574,
|
||
|
|
"loss": 3.5211910247802733,
|
||
|
|
"step": 6120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23712127805661126,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00044009166090681635,
|
||
|
|
"loss": 3.4782859802246096,
|
||
|
|
"step": 6130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2375080990648602,
|
||
|
|
"grad_norm": 0.24609375,
|
||
|
|
"learning_rate": 0.0004398984001976383,
|
||
|
|
"loss": 3.4879383087158202,
|
||
|
|
"step": 6140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23789492007310917,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.00043970487341171016,
|
||
|
|
"loss": 3.519879913330078,
|
||
|
|
"step": 6150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23828174108135813,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004395110808392991,
|
||
|
|
"loss": 3.5101585388183594,
|
||
|
|
"step": 6160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2386685620896071,
|
||
|
|
"grad_norm": 0.296875,
|
||
|
|
"learning_rate": 0.00043931702277107127,
|
||
|
|
"loss": 3.529031753540039,
|
||
|
|
"step": 6170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23905538309785604,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00043912269949809114,
|
||
|
|
"loss": 3.50939826965332,
|
||
|
|
"step": 6180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.239442204106105,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00043892811131182054,
|
||
|
|
"loss": 3.4897247314453126,
|
||
|
|
"step": 6190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23982902511435397,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.000438733258504119,
|
||
|
|
"loss": 3.5406448364257814,
|
||
|
|
"step": 6200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23982902511435397,
|
||
|
|
"eval_loss": 4.047799587249756,
|
||
|
|
"eval_runtime": 10.117,
|
||
|
|
"eval_samples_per_second": 32.322,
|
||
|
|
"eval_steps_per_second": 2.076,
|
||
|
|
"step": 6200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2402158461226029,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00043853814136724287,
|
||
|
|
"loss": 3.5504226684570312,
|
||
|
|
"step": 6210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24060266713085188,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004383427601938448,
|
||
|
|
"loss": 3.514726257324219,
|
||
|
|
"step": 6220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24098948813910084,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004381471152769737,
|
||
|
|
"loss": 3.538137435913086,
|
||
|
|
"step": 6230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24137630914734978,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00043795120691007394,
|
||
|
|
"loss": 3.47631950378418,
|
||
|
|
"step": 6240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24176313015559875,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00043775503538698497,
|
||
|
|
"loss": 3.522532272338867,
|
||
|
|
"step": 6250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24214995116384772,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004375586010019411,
|
||
|
|
"loss": 3.5053829193115233,
|
||
|
|
"step": 6260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24253677217209665,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004373619040495707,
|
||
|
|
"loss": 3.4885364532470704,
|
||
|
|
"step": 6270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24292359318034562,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00043716494482489614,
|
||
|
|
"loss": 3.55185546875,
|
||
|
|
"step": 6280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2433104141885946,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004369677236233332,
|
||
|
|
"loss": 3.5000751495361326,
|
||
|
|
"step": 6290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24369723519684355,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00043677024074069036,
|
||
|
|
"loss": 3.5143280029296875,
|
||
|
|
"step": 6300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24369723519684355,
|
||
|
|
"eval_loss": 4.032617568969727,
|
||
|
|
"eval_runtime": 10.0951,
|
||
|
|
"eval_samples_per_second": 32.392,
|
||
|
|
"eval_steps_per_second": 2.08,
|
||
|
|
"step": 6300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2440840562050925,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004365724964731689,
|
||
|
|
"loss": 3.5079113006591798,
|
||
|
|
"step": 6310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24447087721334146,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004363744911173619,
|
||
|
|
"loss": 3.562195587158203,
|
||
|
|
"step": 6320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24485769822159043,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004361762249702541,
|
||
|
|
"loss": 3.4940654754638674,
|
||
|
|
"step": 6330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24524451922983936,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004359776983292213,
|
||
|
|
"loss": 3.495630645751953,
|
||
|
|
"step": 6340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24563134023808833,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00043577891149203035,
|
||
|
|
"loss": 3.5518524169921877,
|
||
|
|
"step": 6350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2460181612463373,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00043557986475683803,
|
||
|
|
"loss": 3.51075439453125,
|
||
|
|
"step": 6360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24640498225458624,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000435380558422191,
|
||
|
|
"loss": 3.53402099609375,
|
||
|
|
"step": 6370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2467918032628352,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00043518099278702546,
|
||
|
|
"loss": 3.5189449310302736,
|
||
|
|
"step": 6380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24717862427108417,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0004349811681506663,
|
||
|
|
"loss": 3.520494842529297,
|
||
|
|
"step": 6390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2475654452793331,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004347810848128271,
|
||
|
|
"loss": 3.544877624511719,
|
||
|
|
"step": 6400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2475654452793331,
|
||
|
|
"eval_loss": 4.026139259338379,
|
||
|
|
"eval_runtime": 10.078,
|
||
|
|
"eval_samples_per_second": 32.447,
|
||
|
|
"eval_steps_per_second": 2.084,
|
||
|
|
"step": 6400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24795226628758207,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004345807430736092,
|
||
|
|
"loss": 3.5174510955810545,
|
||
|
|
"step": 6410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24833908729583104,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004343801432335019,
|
||
|
|
"loss": 3.4823192596435546,
|
||
|
|
"step": 6420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24872590830408,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004341792855933811,
|
||
|
|
"loss": 3.5113250732421877,
|
||
|
|
"step": 6430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24911272931232895,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00043397817045450984,
|
||
|
|
"loss": 3.514263153076172,
|
||
|
|
"step": 6440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2494995503205779,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004337767981185371,
|
||
|
|
"loss": 3.496726226806641,
|
||
|
|
"step": 6450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24988637132882688,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00043357516888749775,
|
||
|
|
"loss": 3.512440490722656,
|
||
|
|
"step": 6460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25027319233707584,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00043337328306381195,
|
||
|
|
"loss": 3.501117706298828,
|
||
|
|
"step": 6470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2506600133453248,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004331711409502847,
|
||
|
|
"loss": 3.460781478881836,
|
||
|
|
"step": 6480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2510468343535737,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004329687428501054,
|
||
|
|
"loss": 3.499040222167969,
|
||
|
|
"step": 6490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2514336553618227,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.00043276608906684744,
|
||
|
|
"loss": 3.504185104370117,
|
||
|
|
"step": 6500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2514336553618227,
|
||
|
|
"eval_loss": 4.024904727935791,
|
||
|
|
"eval_runtime": 9.9985,
|
||
|
|
"eval_samples_per_second": 32.705,
|
||
|
|
"eval_steps_per_second": 2.1,
|
||
|
|
"step": 6500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25182047637007166,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004325631799044677,
|
||
|
|
"loss": 3.5370899200439454,
|
||
|
|
"step": 6510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2522072973783206,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00043236001566730585,
|
||
|
|
"loss": 3.4886333465576174,
|
||
|
|
"step": 6520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2525941183865696,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00043215659666008473,
|
||
|
|
"loss": 3.4799739837646486,
|
||
|
|
"step": 6530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2529809393948185,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 0.00043195292318790876,
|
||
|
|
"loss": 3.540454864501953,
|
||
|
|
"step": 6540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25336776040306747,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00043174899555626414,
|
||
|
|
"loss": 3.525778961181641,
|
||
|
|
"step": 6550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25375458141131646,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.00043154481407101845,
|
||
|
|
"loss": 3.524898147583008,
|
||
|
|
"step": 6560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2541414024195654,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004313403790384199,
|
||
|
|
"loss": 3.485614776611328,
|
||
|
|
"step": 6570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25452822342781434,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.000431135690765097,
|
||
|
|
"loss": 3.520885467529297,
|
||
|
|
"step": 6580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25491504443606333,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00043093074955805803,
|
||
|
|
"loss": 3.5210807800292967,
|
||
|
|
"step": 6590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25530186544431227,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004307255557246909,
|
||
|
|
"loss": 3.5233665466308595,
|
||
|
|
"step": 6600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25530186544431227,
|
||
|
|
"eval_loss": 3.993399143218994,
|
||
|
|
"eval_runtime": 9.9822,
|
||
|
|
"eval_samples_per_second": 32.758,
|
||
|
|
"eval_steps_per_second": 2.104,
|
||
|
|
"step": 6600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25568868645256126,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.00043052010957276204,
|
||
|
|
"loss": 3.497217559814453,
|
||
|
|
"step": 6610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2560755074608102,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004303144114104167,
|
||
|
|
"loss": 3.4751991271972655,
|
||
|
|
"step": 6620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25646232846905914,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00043010846154617775,
|
||
|
|
"loss": 3.567796325683594,
|
||
|
|
"step": 6630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25684914947730814,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.000429902260288946,
|
||
|
|
"loss": 3.522159957885742,
|
||
|
|
"step": 6640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2572359704855571,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000429695807947999,
|
||
|
|
"loss": 3.5400524139404297,
|
||
|
|
"step": 6650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.257622791493806,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00042948910483299105,
|
||
|
|
"loss": 3.4918754577636717,
|
||
|
|
"step": 6660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.258009612502055,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00042928215125395236,
|
||
|
|
"loss": 3.524504470825195,
|
||
|
|
"step": 6670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25839643351030395,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0004290749475212892,
|
||
|
|
"loss": 3.5244983673095702,
|
||
|
|
"step": 6680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2587832545185529,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00042886749394578254,
|
||
|
|
"loss": 3.4681201934814454,
|
||
|
|
"step": 6690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2591700755268019,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004286597908385887,
|
||
|
|
"loss": 3.499583435058594,
|
||
|
|
"step": 6700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2591700755268019,
|
||
|
|
"eval_loss": 4.000979900360107,
|
||
|
|
"eval_runtime": 9.9625,
|
||
|
|
"eval_samples_per_second": 32.823,
|
||
|
|
"eval_steps_per_second": 2.108,
|
||
|
|
"step": 6700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2595568965350508,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004284518385112377,
|
||
|
|
"loss": 3.544001007080078,
|
||
|
|
"step": 6710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25994371754329976,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004282436372756337,
|
||
|
|
"loss": 3.491320037841797,
|
||
|
|
"step": 6720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26033053855154875,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004280351874440539,
|
||
|
|
"loss": 3.492382049560547,
|
||
|
|
"step": 6730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2607173595597977,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004278264893291487,
|
||
|
|
"loss": 3.547830581665039,
|
||
|
|
"step": 6740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26110418056804663,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004276175432439408,
|
||
|
|
"loss": 3.5104396820068358,
|
||
|
|
"step": 6750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2614910015762956,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00042740834950182467,
|
||
|
|
"loss": 3.5236995697021483,
|
||
|
|
"step": 6760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26187782258454456,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00042719890841656636,
|
||
|
|
"loss": 3.5166088104248048,
|
||
|
|
"step": 6770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2622646435927935,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00042698922030230284,
|
||
|
|
"loss": 3.4963302612304688,
|
||
|
|
"step": 6780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2626514646010425,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.00042677928547354174,
|
||
|
|
"loss": 3.5068523406982424,
|
||
|
|
"step": 6790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26303828560929143,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00042656910424516053,
|
||
|
|
"loss": 3.524884796142578,
|
||
|
|
"step": 6800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26303828560929143,
|
||
|
|
"eval_loss": 3.9918432235717773,
|
||
|
|
"eval_runtime": 9.9656,
|
||
|
|
"eval_samples_per_second": 32.813,
|
||
|
|
"eval_steps_per_second": 2.107,
|
||
|
|
"step": 6800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26342510661754037,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00042635867693240634,
|
||
|
|
"loss": 3.4992977142333985,
|
||
|
|
"step": 6810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26381192762578937,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00042614800385089546,
|
||
|
|
"loss": 3.48309440612793,
|
||
|
|
"step": 6820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2641987486340383,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00042593708531661275,
|
||
|
|
"loss": 3.520304870605469,
|
||
|
|
"step": 6830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2645855696422873,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004257259216459113,
|
||
|
|
"loss": 3.541963577270508,
|
||
|
|
"step": 6840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26497239065053624,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004255145131555116,
|
||
|
|
"loss": 3.474831771850586,
|
||
|
|
"step": 6850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2653592116587852,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00042530286016250173,
|
||
|
|
"loss": 3.475337600708008,
|
||
|
|
"step": 6860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26574603266703417,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004250909629843362,
|
||
|
|
"loss": 3.512445831298828,
|
||
|
|
"step": 6870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2661328536752831,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00042487882193883585,
|
||
|
|
"loss": 3.528004837036133,
|
||
|
|
"step": 6880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26651967468353205,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00042466643734418737,
|
||
|
|
"loss": 3.446039581298828,
|
||
|
|
"step": 6890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26690649569178104,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.0004244538095189427,
|
||
|
|
"loss": 3.4687259674072264,
|
||
|
|
"step": 6900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26690649569178104,
|
||
|
|
"eval_loss": 3.9921114444732666,
|
||
|
|
"eval_runtime": 10.0839,
|
||
|
|
"eval_samples_per_second": 32.428,
|
||
|
|
"eval_steps_per_second": 2.083,
|
||
|
|
"step": 6900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26729331670003,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004242409387820186,
|
||
|
|
"loss": 3.5139041900634767,
|
||
|
|
"step": 6910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2676801377082789,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.00042402782545269605,
|
||
|
|
"loss": 3.4841732025146483,
|
||
|
|
"step": 6920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2680669587165279,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0004238144698506201,
|
||
|
|
"loss": 3.5189540863037108,
|
||
|
|
"step": 6930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26845377972477685,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00042360087229579907,
|
||
|
|
"loss": 3.499907684326172,
|
||
|
|
"step": 6940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2688406007330258,
|
||
|
|
"grad_norm": 0.283203125,
|
||
|
|
"learning_rate": 0.00042338703310860413,
|
||
|
|
"loss": 3.4943309783935548,
|
||
|
|
"step": 6950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2692274217412748,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.000423172952609769,
|
||
|
|
"loss": 3.4940841674804686,
|
||
|
|
"step": 6960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2696142427495237,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.0004229586311203892,
|
||
|
|
"loss": 3.459016799926758,
|
||
|
|
"step": 6970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27000106375777266,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004227440689619218,
|
||
|
|
"loss": 3.5130611419677735,
|
||
|
|
"step": 6980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27038788476602166,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00042252926645618485,
|
||
|
|
"loss": 3.500564956665039,
|
||
|
|
"step": 6990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2707747057742706,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004223142239253568,
|
||
|
|
"loss": 3.495101547241211,
|
||
|
|
"step": 7000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2707747057742706,
|
||
|
|
"eval_loss": 3.983806848526001,
|
||
|
|
"eval_runtime": 10.0108,
|
||
|
|
"eval_samples_per_second": 32.665,
|
||
|
|
"eval_steps_per_second": 2.098,
|
||
|
|
"step": 7000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27116152678251954,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.0004220989416919762,
|
||
|
|
"loss": 3.4596118927001953,
|
||
|
|
"step": 7010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27154834779076853,
|
||
|
|
"grad_norm": 0.291015625,
|
||
|
|
"learning_rate": 0.00042188342007894115,
|
||
|
|
"loss": 3.479116439819336,
|
||
|
|
"step": 7020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27193516879901747,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00042166765940950874,
|
||
|
|
"loss": 3.5550628662109376,
|
||
|
|
"step": 7030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2723219898072664,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0004214516600072944,
|
||
|
|
"loss": 3.4750045776367187,
|
||
|
|
"step": 7040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2727088108155154,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00042123542219627214,
|
||
|
|
"loss": 3.492919921875,
|
||
|
|
"step": 7050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27309563182376434,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.00042101894630077315,
|
||
|
|
"loss": 3.5046051025390623,
|
||
|
|
"step": 7060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27348245283201333,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0004208022326454858,
|
||
|
|
"loss": 3.478554916381836,
|
||
|
|
"step": 7070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2738692738402623,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00042058528155545507,
|
||
|
|
"loss": 3.5061058044433593,
|
||
|
|
"step": 7080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2742560948485112,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00042036809335608215,
|
||
|
|
"loss": 3.463357162475586,
|
||
|
|
"step": 7090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2746429158567602,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004201506683731238,
|
||
|
|
"loss": 3.52327880859375,
|
||
|
|
"step": 7100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2746429158567602,
|
||
|
|
"eval_loss": 4.0028395652771,
|
||
|
|
"eval_runtime": 10.0008,
|
||
|
|
"eval_samples_per_second": 32.697,
|
||
|
|
"eval_steps_per_second": 2.1,
|
||
|
|
"step": 7100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27502973686500914,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041993300693269204,
|
||
|
|
"loss": 3.5009181976318358,
|
||
|
|
"step": 7110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2754165578732581,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00041971510936125326,
|
||
|
|
"loss": 3.4968143463134767,
|
||
|
|
"step": 7120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2758033788815071,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00041949697598562836,
|
||
|
|
"loss": 3.509387969970703,
|
||
|
|
"step": 7130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.276190199889756,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041927860713299165,
|
||
|
|
"loss": 3.505107116699219,
|
||
|
|
"step": 7140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27657702089800495,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.0004190600031308709,
|
||
|
|
"loss": 3.4760673522949217,
|
||
|
|
"step": 7150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27696384190625395,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004188411643071462,
|
||
|
|
"loss": 3.471649932861328,
|
||
|
|
"step": 7160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2773506629145029,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004186220909900503,
|
||
|
|
"loss": 3.5090023040771485,
|
||
|
|
"step": 7170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2777374839227518,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00041840278350816726,
|
||
|
|
"loss": 3.485765075683594,
|
||
|
|
"step": 7180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2781243049310008,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00041818324219043256,
|
||
|
|
"loss": 3.48890380859375,
|
||
|
|
"step": 7190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27851112593924976,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004179634673661324,
|
||
|
|
"loss": 3.4805828094482423,
|
||
|
|
"step": 7200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27851112593924976,
|
||
|
|
"eval_loss": 3.9834365844726562,
|
||
|
|
"eval_runtime": 10.0054,
|
||
|
|
"eval_samples_per_second": 32.682,
|
||
|
|
"eval_steps_per_second": 2.099,
|
||
|
|
"step": 7200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2788979469474987,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004177434593649032,
|
||
|
|
"loss": 3.4678180694580076,
|
||
|
|
"step": 7210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2792847679557477,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041752321851673107,
|
||
|
|
"loss": 3.498748779296875,
|
||
|
|
"step": 7220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27967158896399663,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0004173027451519514,
|
||
|
|
"loss": 3.492596435546875,
|
||
|
|
"step": 7230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28005840997224557,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.00041708203960124843,
|
||
|
|
"loss": 3.5052108764648438,
|
||
|
|
"step": 7240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28044523098049456,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00041686110219565437,
|
||
|
|
"loss": 3.4743038177490235,
|
||
|
|
"step": 7250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2808320519887435,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00041663993326654944,
|
||
|
|
"loss": 3.508194351196289,
|
||
|
|
"step": 7260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28121887299699244,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.000416418533145661,
|
||
|
|
"loss": 3.5267372131347656,
|
||
|
|
"step": 7270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28160569400524144,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041619690216506333,
|
||
|
|
"loss": 3.5127063751220704,
|
||
|
|
"step": 7280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2819925150134904,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004159750406571768,
|
||
|
|
"loss": 3.4871360778808596,
|
||
|
|
"step": 7290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2823793360217393,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.0004157529489547675,
|
||
|
|
"loss": 3.4671897888183594,
|
||
|
|
"step": 7300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2823793360217393,
|
||
|
|
"eval_loss": 3.978560447692871,
|
||
|
|
"eval_runtime": 10.004,
|
||
|
|
"eval_samples_per_second": 32.687,
|
||
|
|
"eval_steps_per_second": 2.099,
|
||
|
|
"step": 7300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2827661570299883,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.000415530627390947,
|
||
|
|
"loss": 3.4803470611572265,
|
||
|
|
"step": 7310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28315297803823725,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004153080762991714,
|
||
|
|
"loss": 3.490553283691406,
|
||
|
|
"step": 7320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28353979904648624,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004150852960132413,
|
||
|
|
"loss": 3.523568344116211,
|
||
|
|
"step": 7330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2839266200547352,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004148622868673009,
|
||
|
|
"loss": 3.4703395843505858,
|
||
|
|
"step": 7340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2843134410629841,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0004146390491958377,
|
||
|
|
"loss": 3.493512725830078,
|
||
|
|
"step": 7350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2847002620712331,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0004144155833336819,
|
||
|
|
"loss": 3.459734344482422,
|
||
|
|
"step": 7360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28508708307948205,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041419188961600614,
|
||
|
|
"loss": 3.442743682861328,
|
||
|
|
"step": 7370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.285473904087731,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00041396796837832474,
|
||
|
|
"loss": 3.4958415985107423,
|
||
|
|
"step": 7380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28586072509598,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00041374381995649306,
|
||
|
|
"loss": 3.4780853271484373,
|
||
|
|
"step": 7390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2862475461042289,
|
||
|
|
"grad_norm": 0.2333984375,
|
||
|
|
"learning_rate": 0.00041351944468670754,
|
||
|
|
"loss": 3.48729248046875,
|
||
|
|
"step": 7400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2862475461042289,
|
||
|
|
"eval_loss": 3.9693050384521484,
|
||
|
|
"eval_runtime": 9.9899,
|
||
|
|
"eval_samples_per_second": 32.733,
|
||
|
|
"eval_steps_per_second": 2.102,
|
||
|
|
"step": 7400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28663436711247786,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.00041329484290550454,
|
||
|
|
"loss": 3.464725112915039,
|
||
|
|
"step": 7410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28702118812072686,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.00041307001494976064,
|
||
|
|
"loss": 3.5638668060302736,
|
||
|
|
"step": 7420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2874080091289758,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00041284496115669107,
|
||
|
|
"loss": 3.4758174896240233,
|
||
|
|
"step": 7430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28779483013722473,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004126196818638502,
|
||
|
|
"loss": 3.4785606384277346,
|
||
|
|
"step": 7440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2881816511454737,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0004123941774091304,
|
||
|
|
"loss": 3.492573928833008,
|
||
|
|
"step": 7450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28856847215372267,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00041216844813076197,
|
||
|
|
"loss": 3.510224533081055,
|
||
|
|
"step": 7460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2889552931619716,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004119424943673122,
|
||
|
|
"loss": 3.469999313354492,
|
||
|
|
"step": 7470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2893421141702206,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00041171631645768527,
|
||
|
|
"loss": 3.4652854919433596,
|
||
|
|
"step": 7480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28972893517846954,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00041148991474112135,
|
||
|
|
"loss": 3.445465850830078,
|
||
|
|
"step": 7490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2901157561867185,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00041126328955719645,
|
||
|
|
"loss": 3.483041000366211,
|
||
|
|
"step": 7500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2901157561867185,
|
||
|
|
"eval_loss": 3.9702155590057373,
|
||
|
|
"eval_runtime": 10.0085,
|
||
|
|
"eval_samples_per_second": 32.672,
|
||
|
|
"eval_steps_per_second": 2.098,
|
||
|
|
"step": 7500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29050257719496747,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004110364412458217,
|
||
|
|
"loss": 3.4610855102539064,
|
||
|
|
"step": 7510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2908893982032164,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.000410809370147243,
|
||
|
|
"loss": 3.4661144256591796,
|
||
|
|
"step": 7520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29127621921146535,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0004105820766020402,
|
||
|
|
"loss": 3.4697357177734376,
|
||
|
|
"step": 7530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29166304021971434,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00041035456095112684,
|
||
|
|
"loss": 3.451828384399414,
|
||
|
|
"step": 7540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2920498612279633,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00041012682353574976,
|
||
|
|
"loss": 3.4903900146484377,
|
||
|
|
"step": 7550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2924366822362123,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00040989886469748843,
|
||
|
|
"loss": 3.44178466796875,
|
||
|
|
"step": 7560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2928235032444612,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004096706847782541,
|
||
|
|
"loss": 3.491189956665039,
|
||
|
|
"step": 7570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29321032425271015,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00040944228412029,
|
||
|
|
"loss": 3.4560359954833983,
|
||
|
|
"step": 7580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29359714526095915,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0004092136630661701,
|
||
|
|
"loss": 3.491283416748047,
|
||
|
|
"step": 7590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2939839662692081,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00040898482195879935,
|
||
|
|
"loss": 3.469133758544922,
|
||
|
|
"step": 7600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2939839662692081,
|
||
|
|
"eval_loss": 3.961404800415039,
|
||
|
|
"eval_runtime": 10.1106,
|
||
|
|
"eval_samples_per_second": 32.342,
|
||
|
|
"eval_steps_per_second": 2.077,
|
||
|
|
"step": 7600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.294370787277457,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004087557611414122,
|
||
|
|
"loss": 3.47923583984375,
|
||
|
|
"step": 7610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.294757608285706,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.00040852648095757323,
|
||
|
|
"loss": 3.5196529388427735,
|
||
|
|
"step": 7620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29514442929395496,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0004082969817511755,
|
||
|
|
"loss": 3.4972084045410154,
|
||
|
|
"step": 7630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2955312503022039,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0004080672638664412,
|
||
|
|
"loss": 3.453374481201172,
|
||
|
|
"step": 7640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2959180713104529,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0004078373276479199,
|
||
|
|
"loss": 3.4950233459472657,
|
||
|
|
"step": 7650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29630489231870183,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.00040760717344048896,
|
||
|
|
"loss": 3.462534713745117,
|
||
|
|
"step": 7660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29669171332695077,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00040737680158935277,
|
||
|
|
"loss": 3.496277618408203,
|
||
|
|
"step": 7670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29707853433519976,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0004071462124400417,
|
||
|
|
"loss": 3.4707130432128905,
|
||
|
|
"step": 7680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2974653553434487,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.00040691540633841267,
|
||
|
|
"loss": 3.4735038757324217,
|
||
|
|
"step": 7690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29785217635169764,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004066843836306477,
|
||
|
|
"loss": 3.4587879180908203,
|
||
|
|
"step": 7700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29785217635169764,
|
||
|
|
"eval_loss": 3.9454445838928223,
|
||
|
|
"eval_runtime": 10.0538,
|
||
|
|
"eval_samples_per_second": 32.525,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 7700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29823899735994663,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0004064531446632535,
|
||
|
|
"loss": 3.4567310333251955,
|
||
|
|
"step": 7710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2986258183681956,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0004062216897830615,
|
||
|
|
"loss": 3.5064525604248047,
|
||
|
|
"step": 7720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2990126393764445,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.00040599001933722687,
|
||
|
|
"loss": 3.4508167266845704,
|
||
|
|
"step": 7730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2993994603846935,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.00040575813367322786,
|
||
|
|
"loss": 3.5213436126708983,
|
||
|
|
"step": 7740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29978628139294244,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.00040552603313886575,
|
||
|
|
"loss": 3.44835205078125,
|
||
|
|
"step": 7750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3001731024011914,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0004052937180822642,
|
||
|
|
"loss": 3.47521858215332,
|
||
|
|
"step": 7760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3005599234094404,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00040506118885186846,
|
||
|
|
"loss": 3.5136672973632814,
|
||
|
|
"step": 7770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3009467444176893,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0004048284457964449,
|
||
|
|
"loss": 3.455365753173828,
|
||
|
|
"step": 7780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3013335654259383,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00040459548926508087,
|
||
|
|
"loss": 3.4851306915283202,
|
||
|
|
"step": 7790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30172038643418725,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0004043623196071838,
|
||
|
|
"loss": 3.4420330047607424,
|
||
|
|
"step": 7800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30172038643418725,
|
||
|
|
"eval_loss": 3.977670907974243,
|
||
|
|
"eval_runtime": 10.0649,
|
||
|
|
"eval_samples_per_second": 32.489,
|
||
|
|
"eval_steps_per_second": 2.086,
|
||
|
|
"step": 7800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3021072074424362,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004041289371724808,
|
||
|
|
"loss": 3.486972427368164,
|
||
|
|
"step": 7810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3024940284506852,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0004038953423110179,
|
||
|
|
"loss": 3.4616844177246096,
|
||
|
|
"step": 7820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3028808494589341,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00040366153537316007,
|
||
|
|
"loss": 3.5070659637451174,
|
||
|
|
"step": 7830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30326767046718306,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004034275167095902,
|
||
|
|
"loss": 3.4858856201171875,
|
||
|
|
"step": 7840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30365449147543205,
|
||
|
|
"grad_norm": 0.228515625,
|
||
|
|
"learning_rate": 0.00040319328667130897,
|
||
|
|
"loss": 3.471744155883789,
|
||
|
|
"step": 7850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.304041312483681,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00040295884560963356,
|
||
|
|
"loss": 3.455038833618164,
|
||
|
|
"step": 7860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30442813349192993,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004027241938761983,
|
||
|
|
"loss": 3.491812896728516,
|
||
|
|
"step": 7870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3048149545001789,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.0004024893318229531,
|
||
|
|
"loss": 3.4919559478759767,
|
||
|
|
"step": 7880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30520177550842786,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0004022542598021635,
|
||
|
|
"loss": 3.4516735076904297,
|
||
|
|
"step": 7890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3055885965166768,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00040201897816640977,
|
||
|
|
"loss": 3.4588272094726564,
|
||
|
|
"step": 7900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3055885965166768,
|
||
|
|
"eval_loss": 3.9536006450653076,
|
||
|
|
"eval_runtime": 10.3723,
|
||
|
|
"eval_samples_per_second": 31.526,
|
||
|
|
"eval_steps_per_second": 2.025,
|
||
|
|
"step": 7900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3059754175249258,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00040178348726858677,
|
||
|
|
"loss": 3.482456588745117,
|
||
|
|
"step": 7910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30636223853317474,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00040154778746190323,
|
||
|
|
"loss": 3.4793769836425783,
|
||
|
|
"step": 7920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3067490595414237,
|
||
|
|
"grad_norm": 0.2412109375,
|
||
|
|
"learning_rate": 0.00040131187909988117,
|
||
|
|
"loss": 3.4581600189208985,
|
||
|
|
"step": 7930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30713588054967267,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004010757625363552,
|
||
|
|
"loss": 3.4720085144042967,
|
||
|
|
"step": 7940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3075227015579216,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0004008394381254727,
|
||
|
|
"loss": 3.4618335723876954,
|
||
|
|
"step": 7950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30790952256617055,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.00040060290622169226,
|
||
|
|
"loss": 3.474958038330078,
|
||
|
|
"step": 7960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30829634357441954,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 0.00040036616717978416,
|
||
|
|
"loss": 3.4388957977294923,
|
||
|
|
"step": 7970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3086831645826685,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.000400129221354829,
|
||
|
|
"loss": 3.456399917602539,
|
||
|
|
"step": 7980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3090699855909174,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003998920691022176,
|
||
|
|
"loss": 3.4763534545898436,
|
||
|
|
"step": 7990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3094568065991664,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00039965471077765065,
|
||
|
|
"loss": 3.444426345825195,
|
||
|
|
"step": 8000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3094568065991664,
|
||
|
|
"eval_loss": 3.954042673110962,
|
||
|
|
"eval_runtime": 10.8919,
|
||
|
|
"eval_samples_per_second": 30.022,
|
||
|
|
"eval_steps_per_second": 1.928,
|
||
|
|
"step": 8000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30984362760741535,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00039941714673713763,
|
||
|
|
"loss": 3.4891815185546875,
|
||
|
|
"step": 8010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3102304486156643,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0003991793773369967,
|
||
|
|
"loss": 3.4721736907958984,
|
||
|
|
"step": 8020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3106172696239133,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000398941402933854,
|
||
|
|
"loss": 3.4296173095703124,
|
||
|
|
"step": 8030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3110040906321622,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003987032238846432,
|
||
|
|
"loss": 3.4733917236328127,
|
||
|
|
"step": 8040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3113909116404112,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00039846484054660493,
|
||
|
|
"loss": 3.4510936737060547,
|
||
|
|
"step": 8050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31177773264866016,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0003982262532772861,
|
||
|
|
"loss": 3.4586532592773436,
|
||
|
|
"step": 8060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3121645536569091,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003979874624345395,
|
||
|
|
"loss": 3.48809700012207,
|
||
|
|
"step": 8070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3125513746651581,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003977484683765234,
|
||
|
|
"loss": 3.4487079620361327,
|
||
|
|
"step": 8080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.312938195673407,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003975092714617009,
|
||
|
|
"loss": 3.4913402557373048,
|
||
|
|
"step": 8090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31332501668165597,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00039726987204883907,
|
||
|
|
"loss": 3.465750503540039,
|
||
|
|
"step": 8100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31332501668165597,
|
||
|
|
"eval_loss": 3.945279598236084,
|
||
|
|
"eval_runtime": 10.1775,
|
||
|
|
"eval_samples_per_second": 32.13,
|
||
|
|
"eval_steps_per_second": 2.063,
|
||
|
|
"step": 8100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31371183768990496,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.000397030270497009,
|
||
|
|
"loss": 3.459657669067383,
|
||
|
|
"step": 8110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3140986586981539,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003967904671655849,
|
||
|
|
"loss": 3.453855514526367,
|
||
|
|
"step": 8120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31448547970640284,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0003965504624142435,
|
||
|
|
"loss": 3.485597610473633,
|
||
|
|
"step": 8130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31487230071465183,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00039631025660296384,
|
||
|
|
"loss": 3.4898101806640627,
|
||
|
|
"step": 8140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31525912172290077,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00039606985009202616,
|
||
|
|
"loss": 3.437654495239258,
|
||
|
|
"step": 8150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3156459427311497,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003958292432420122,
|
||
|
|
"loss": 3.4521121978759766,
|
||
|
|
"step": 8160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3160327637393987,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00039558843641380385,
|
||
|
|
"loss": 3.446636199951172,
|
||
|
|
"step": 8170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31641958474764764,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00039534742996858314,
|
||
|
|
"loss": 3.5297615051269533,
|
||
|
|
"step": 8180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3168064057558966,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003951062242678313,
|
||
|
|
"loss": 3.4837715148925783,
|
||
|
|
"step": 8190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3171932267641456,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0003948648196733286,
|
||
|
|
"loss": 3.4503883361816405,
|
||
|
|
"step": 8200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3171932267641456,
|
||
|
|
"eval_loss": 3.947329044342041,
|
||
|
|
"eval_runtime": 10.1317,
|
||
|
|
"eval_samples_per_second": 32.275,
|
||
|
|
"eval_steps_per_second": 2.073,
|
||
|
|
"step": 8200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3175800477723945,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00039462321654715356,
|
||
|
|
"loss": 3.4455516815185545,
|
||
|
|
"step": 8210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31796686878064345,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00039438141525168245,
|
||
|
|
"loss": 3.4772762298583983,
|
||
|
|
"step": 8220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31835368978889245,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00039413941614958875,
|
||
|
|
"loss": 3.480706787109375,
|
||
|
|
"step": 8230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3187405107971414,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.0003938972196038428,
|
||
|
|
"loss": 3.443967819213867,
|
||
|
|
"step": 8240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3191273318053903,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003936548259777107,
|
||
|
|
"loss": 3.4702228546142577,
|
||
|
|
"step": 8250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3195141528136393,
|
||
|
|
"grad_norm": 0.23828125,
|
||
|
|
"learning_rate": 0.0003934122356347548,
|
||
|
|
"loss": 3.467062759399414,
|
||
|
|
"step": 8260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31990097382188826,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003931694489388317,
|
||
|
|
"loss": 3.4245067596435548,
|
||
|
|
"step": 8270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32028779483013725,
|
||
|
|
"grad_norm": 0.234375,
|
||
|
|
"learning_rate": 0.00039292646625409315,
|
||
|
|
"loss": 3.4993412017822267,
|
||
|
|
"step": 8280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3206746158383862,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00039268328794498454,
|
||
|
|
"loss": 3.4476173400878904,
|
||
|
|
"step": 8290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32106143684663513,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003924399143762448,
|
||
|
|
"loss": 3.467058563232422,
|
||
|
|
"step": 8300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32106143684663513,
|
||
|
|
"eval_loss": 3.9062752723693848,
|
||
|
|
"eval_runtime": 10.0398,
|
||
|
|
"eval_samples_per_second": 32.57,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 8300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3214482578548841,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0003921963459129056,
|
||
|
|
"loss": 3.4350086212158204,
|
||
|
|
"step": 8310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32183507886313306,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003919525829202911,
|
||
|
|
"loss": 3.4416141510009766,
|
||
|
|
"step": 8320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.322221899871382,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.000391708625764017,
|
||
|
|
"loss": 3.4725067138671877,
|
||
|
|
"step": 8330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.322608720879631,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003914644748099905,
|
||
|
|
"loss": 3.448748779296875,
|
||
|
|
"step": 8340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32299554188787993,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00039122013042440935,
|
||
|
|
"loss": 3.446158218383789,
|
||
|
|
"step": 8350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32338236289612887,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.00039097559297376124,
|
||
|
|
"loss": 3.4900192260742187,
|
||
|
|
"step": 8360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32376918390437787,
|
||
|
|
"grad_norm": 0.240234375,
|
||
|
|
"learning_rate": 0.0003907308628248238,
|
||
|
|
"loss": 3.457179641723633,
|
||
|
|
"step": 8370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3241560049126268,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003904859403446633,
|
||
|
|
"loss": 3.431169891357422,
|
||
|
|
"step": 8380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32454282592087574,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003902408259006348,
|
||
|
|
"loss": 3.4472503662109375,
|
||
|
|
"step": 8390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32492964692912474,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0003899955198603812,
|
||
|
|
"loss": 3.439772033691406,
|
||
|
|
"step": 8400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32492964692912474,
|
||
|
|
"eval_loss": 3.9217541217803955,
|
||
|
|
"eval_runtime": 10.0191,
|
||
|
|
"eval_samples_per_second": 32.638,
|
||
|
|
"eval_steps_per_second": 2.096,
|
||
|
|
"step": 8400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3253164679373737,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0003897500225918326,
|
||
|
|
"loss": 3.4198753356933596,
|
||
|
|
"step": 8410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3257032889456226,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00038950433446320614,
|
||
|
|
"loss": 3.4733913421630858,
|
||
|
|
"step": 8420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3260901099538716,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003892584558430051,
|
||
|
|
"loss": 3.438296890258789,
|
||
|
|
"step": 8430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32647693096212055,
|
||
|
|
"grad_norm": 0.3125,
|
||
|
|
"learning_rate": 0.00038901238710001854,
|
||
|
|
"loss": 3.4649559020996095,
|
||
|
|
"step": 8440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3268637519703695,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00038876612860332056,
|
||
|
|
"loss": 3.4911937713623047,
|
||
|
|
"step": 8450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3272505729786185,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00038851968072227004,
|
||
|
|
"loss": 3.4624607086181642,
|
||
|
|
"step": 8460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3276373939868674,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.00038827304382651,
|
||
|
|
"loss": 3.4401573181152343,
|
||
|
|
"step": 8470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32802421499511636,
|
||
|
|
"grad_norm": 0.28515625,
|
||
|
|
"learning_rate": 0.0003880262182859664,
|
||
|
|
"loss": 3.456196975708008,
|
||
|
|
"step": 8480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32841103600336535,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003877792044708489,
|
||
|
|
"loss": 3.4427886962890626,
|
||
|
|
"step": 8490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3287978570116143,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00038753200275164895,
|
||
|
|
"loss": 3.4402111053466795,
|
||
|
|
"step": 8500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3287978570116143,
|
||
|
|
"eval_loss": 3.931530237197876,
|
||
|
|
"eval_runtime": 10.0521,
|
||
|
|
"eval_samples_per_second": 32.531,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 8500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3291846780198633,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0003872846134991402,
|
||
|
|
"loss": 3.4195785522460938,
|
||
|
|
"step": 8510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3295714990281122,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003870370370843775,
|
||
|
|
"loss": 3.474544906616211,
|
||
|
|
"step": 8520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32995832003636116,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.0003867892738786963,
|
||
|
|
"loss": 3.4477092742919924,
|
||
|
|
"step": 8530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33034514104461016,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00038654132425371227,
|
||
|
|
"loss": 3.4963069915771485,
|
||
|
|
"step": 8540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3307319620528591,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003862931885813207,
|
||
|
|
"loss": 3.4387706756591796,
|
||
|
|
"step": 8550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33111878306110804,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00038604486723369604,
|
||
|
|
"loss": 3.4153343200683595,
|
||
|
|
"step": 8560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33150560406935703,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.000385796360583291,
|
||
|
|
"loss": 3.4407272338867188,
|
||
|
|
"step": 8570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33189242507760597,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003855476690028364,
|
||
|
|
"loss": 3.442066192626953,
|
||
|
|
"step": 8580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3322792460858549,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.0003852987928653404,
|
||
|
|
"loss": 3.4434627532958983,
|
||
|
|
"step": 8590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3326660670941039,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00038504973254408794,
|
||
|
|
"loss": 3.4483291625976564,
|
||
|
|
"step": 8600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3326660670941039,
|
||
|
|
"eval_loss": 3.9131031036376953,
|
||
|
|
"eval_runtime": 10.038,
|
||
|
|
"eval_samples_per_second": 32.576,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 8600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33305288810235284,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003848004884126403,
|
||
|
|
"loss": 3.4154125213623048,
|
||
|
|
"step": 8610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3334397091106018,
|
||
|
|
"grad_norm": 0.3046875,
|
||
|
|
"learning_rate": 0.0003845510608448342,
|
||
|
|
"loss": 3.478801727294922,
|
||
|
|
"step": 8620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3338265301188508,
|
||
|
|
"grad_norm": 0.298828125,
|
||
|
|
"learning_rate": 0.0003843014502147818,
|
||
|
|
"loss": 3.4299625396728515,
|
||
|
|
"step": 8630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3342133511270997,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00038405165689686964,
|
||
|
|
"loss": 3.4181446075439452,
|
||
|
|
"step": 8640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33460017213534865,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.00038380168126575845,
|
||
|
|
"loss": 3.451136016845703,
|
||
|
|
"step": 8650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33498699314359764,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003835515236963822,
|
||
|
|
"loss": 3.4648101806640623,
|
||
|
|
"step": 8660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3353738141518466,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003833011845639479,
|
||
|
|
"loss": 3.4500274658203125,
|
||
|
|
"step": 8670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3357606351600955,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003830506642439348,
|
||
|
|
"loss": 3.415103530883789,
|
||
|
|
"step": 8680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3361474561683445,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00038279996311209395,
|
||
|
|
"loss": 3.4284832000732424,
|
||
|
|
"step": 8690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33653427717659345,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.00038254908154444756,
|
||
|
|
"loss": 3.4471145629882813,
|
||
|
|
"step": 8700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33653427717659345,
|
||
|
|
"eval_loss": 3.9201772212982178,
|
||
|
|
"eval_runtime": 10.102,
|
||
|
|
"eval_samples_per_second": 32.37,
|
||
|
|
"eval_steps_per_second": 2.079,
|
||
|
|
"step": 8700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3369210981848424,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0003822980199172885,
|
||
|
|
"loss": 3.421375274658203,
|
||
|
|
"step": 8710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3373079191930914,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003820467786071798,
|
||
|
|
"loss": 3.4209850311279295,
|
||
|
|
"step": 8720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3376947402013403,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003817953579909537,
|
||
|
|
"loss": 3.4273170471191405,
|
||
|
|
"step": 8730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33808156120958927,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00038154375844571176,
|
||
|
|
"loss": 3.4581871032714844,
|
||
|
|
"step": 8740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33846838221783826,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00038129198034882357,
|
||
|
|
"loss": 3.4375110626220704,
|
||
|
|
"step": 8750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3388552032260872,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0003810400240779268,
|
||
|
|
"loss": 3.405530548095703,
|
||
|
|
"step": 8760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3392420242343362,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003807878900109262,
|
||
|
|
"loss": 3.433206558227539,
|
||
|
|
"step": 8770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33962884524258513,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0003805355785259932,
|
||
|
|
"loss": 3.458090591430664,
|
||
|
|
"step": 8780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34001566625083407,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0003802830900015655,
|
||
|
|
"loss": 3.4461318969726564,
|
||
|
|
"step": 8790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34040248725908306,
|
||
|
|
"grad_norm": 0.275390625,
|
||
|
|
"learning_rate": 0.000380030424816346,
|
||
|
|
"loss": 3.477490234375,
|
||
|
|
"step": 8800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34040248725908306,
|
||
|
|
"eval_loss": 3.9167301654815674,
|
||
|
|
"eval_runtime": 10.047,
|
||
|
|
"eval_samples_per_second": 32.547,
|
||
|
|
"eval_steps_per_second": 2.09,
|
||
|
|
"step": 8800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.340789308267332,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.000379777583349303,
|
||
|
|
"loss": 3.421334075927734,
|
||
|
|
"step": 8810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34117612927558094,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003795245659796689,
|
||
|
|
"loss": 3.440085601806641,
|
||
|
|
"step": 8820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34156295028382994,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003792713730869399,
|
||
|
|
"loss": 3.457305145263672,
|
||
|
|
"step": 8830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3419497712920789,
|
||
|
|
"grad_norm": 0.2470703125,
|
||
|
|
"learning_rate": 0.0003790180050508758,
|
||
|
|
"loss": 3.4464256286621096,
|
||
|
|
"step": 8840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3423365923003278,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003787644622514987,
|
||
|
|
"loss": 3.4250965118408203,
|
||
|
|
"step": 8850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3427234133085768,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003785107450690931,
|
||
|
|
"loss": 3.4989051818847656,
|
||
|
|
"step": 8860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34311023431682575,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00037825685388420487,
|
||
|
|
"loss": 3.43200569152832,
|
||
|
|
"step": 8870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3434970553250747,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.000378002789077641,
|
||
|
|
"loss": 3.435445022583008,
|
||
|
|
"step": 8880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3438838763333237,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003777485510304688,
|
||
|
|
"loss": 3.4056018829345702,
|
||
|
|
"step": 8890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3442706973415726,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00037749414012401555,
|
||
|
|
"loss": 3.4461997985839843,
|
||
|
|
"step": 8900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3442706973415726,
|
||
|
|
"eval_loss": 3.903242349624634,
|
||
|
|
"eval_runtime": 10.0386,
|
||
|
|
"eval_samples_per_second": 32.574,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 8900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34465751834982156,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00037723955673986767,
|
||
|
|
"loss": 3.462035369873047,
|
||
|
|
"step": 8910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34504433935807055,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00037698480125987027,
|
||
|
|
"loss": 3.432913589477539,
|
||
|
|
"step": 8920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3454311603663195,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00037672987406612657,
|
||
|
|
"loss": 3.4201942443847657,
|
||
|
|
"step": 8930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34581798137456843,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00037647477554099765,
|
||
|
|
"loss": 3.4594268798828125,
|
||
|
|
"step": 8940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3462048023828174,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.000376219506067101,
|
||
|
|
"loss": 3.445780944824219,
|
||
|
|
"step": 8950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34659162339106636,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.000375964066027311,
|
||
|
|
"loss": 3.429423522949219,
|
||
|
|
"step": 8960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3469784443993153,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00037570845580475764,
|
||
|
|
"loss": 3.437569427490234,
|
||
|
|
"step": 8970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3473652654075643,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00037545267578282626,
|
||
|
|
"loss": 3.408824157714844,
|
||
|
|
"step": 8980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34775208641581323,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003751967263451568,
|
||
|
|
"loss": 3.465847396850586,
|
||
|
|
"step": 8990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3481389074240622,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.00037494060787564336,
|
||
|
|
"loss": 3.425415802001953,
|
||
|
|
"step": 9000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3481389074240622,
|
||
|
|
"eval_loss": 3.908496856689453,
|
||
|
|
"eval_runtime": 10.0504,
|
||
|
|
"eval_samples_per_second": 32.536,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 9000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34852572843231117,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0003746843207584335,
|
||
|
|
"loss": 3.4315727233886717,
|
||
|
|
"step": 9010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3489125494405601,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0003744278653779278,
|
||
|
|
"loss": 3.414052200317383,
|
||
|
|
"step": 9020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3492993704488091,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003741712421187792,
|
||
|
|
"loss": 3.411752700805664,
|
||
|
|
"step": 9030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34968619145705804,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003739144513658923,
|
||
|
|
"loss": 3.409000015258789,
|
||
|
|
"step": 9040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.350073012465307,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00037365749350442326,
|
||
|
|
"loss": 3.440891647338867,
|
||
|
|
"step": 9050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35045983347355597,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.00037340036891977854,
|
||
|
|
"loss": 3.4279243469238283,
|
||
|
|
"step": 9060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3508466544818049,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003731430779976148,
|
||
|
|
"loss": 3.432322692871094,
|
||
|
|
"step": 9070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35123347549005385,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003728856211238383,
|
||
|
|
"loss": 3.427559661865234,
|
||
|
|
"step": 9080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35162029649830284,
|
||
|
|
"grad_norm": 0.234375,
|
||
|
|
"learning_rate": 0.000372627998684604,
|
||
|
|
"loss": 3.4577556610107423,
|
||
|
|
"step": 9090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3520071175065518,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.0003723702110663151,
|
||
|
|
"loss": 3.441410446166992,
|
||
|
|
"step": 9100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3520071175065518,
|
||
|
|
"eval_loss": 3.9001352787017822,
|
||
|
|
"eval_runtime": 10.1074,
|
||
|
|
"eval_samples_per_second": 32.352,
|
||
|
|
"eval_steps_per_second": 2.078,
|
||
|
|
"step": 9100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3523939385148007,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.000372112258655623,
|
||
|
|
"loss": 3.4635162353515625,
|
||
|
|
"step": 9110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3527807595230497,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003718541418394259,
|
||
|
|
"loss": 3.392826461791992,
|
||
|
|
"step": 9120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35316758053129865,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003715958610048687,
|
||
|
|
"loss": 3.457368087768555,
|
||
|
|
"step": 9130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3535544015395476,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00037133741653934226,
|
||
|
|
"loss": 3.4322399139404296,
|
||
|
|
"step": 9140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3539412225477966,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.000371078808830483,
|
||
|
|
"loss": 3.4189102172851564,
|
||
|
|
"step": 9150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3543280435560455,
|
||
|
|
"grad_norm": 0.28125,
|
||
|
|
"learning_rate": 0.000370820038266172,
|
||
|
|
"loss": 3.4366641998291017,
|
||
|
|
"step": 9160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35471486456429446,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00037056110523453473,
|
||
|
|
"loss": 3.4128700256347657,
|
||
|
|
"step": 9170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35510168557254346,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003703020101239403,
|
||
|
|
"loss": 3.4295944213867187,
|
||
|
|
"step": 9180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3554885065807924,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0003700427533230009,
|
||
|
|
"loss": 3.4191574096679687,
|
||
|
|
"step": 9190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35587532758904133,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0003697833352205712,
|
||
|
|
"loss": 3.4229095458984373,
|
||
|
|
"step": 9200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35587532758904133,
|
||
|
|
"eval_loss": 3.8868236541748047,
|
||
|
|
"eval_runtime": 10.0499,
|
||
|
|
"eval_samples_per_second": 32.538,
|
||
|
|
"eval_steps_per_second": 2.09,
|
||
|
|
"step": 9200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35626214859729033,
|
||
|
|
"grad_norm": 0.244140625,
|
||
|
|
"learning_rate": 0.00036952375620574794,
|
||
|
|
"loss": 3.4355594635009767,
|
||
|
|
"step": 9210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35664896960553927,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003692640166678692,
|
||
|
|
"loss": 3.431700897216797,
|
||
|
|
"step": 9220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35703579061378826,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003690041169965136,
|
||
|
|
"loss": 3.4387435913085938,
|
||
|
|
"step": 9230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3574226116220372,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00036874405758150023,
|
||
|
|
"loss": 3.432315444946289,
|
||
|
|
"step": 9240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35780943263028614,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003684838388128876,
|
||
|
|
"loss": 3.4304718017578124,
|
||
|
|
"step": 9250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35819625363853513,
|
||
|
|
"grad_norm": 0.279296875,
|
||
|
|
"learning_rate": 0.00036822346108097325,
|
||
|
|
"loss": 3.3982025146484376,
|
||
|
|
"step": 9260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3585830746467841,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003679629247762933,
|
||
|
|
"loss": 3.4113441467285157,
|
||
|
|
"step": 9270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.358969895655033,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00036770223028962144,
|
||
|
|
"loss": 3.4584545135498046,
|
||
|
|
"step": 9280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.359356716663282,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0003674413780119688,
|
||
|
|
"loss": 3.469136047363281,
|
||
|
|
"step": 9290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35974353767153094,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0003671803683345832,
|
||
|
|
"loss": 3.3828174591064455,
|
||
|
|
"step": 9300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35974353767153094,
|
||
|
|
"eval_loss": 3.91288423538208,
|
||
|
|
"eval_runtime": 10.0525,
|
||
|
|
"eval_samples_per_second": 32.529,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 9300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3601303586797799,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003669192016489484,
|
||
|
|
"loss": 3.386565399169922,
|
||
|
|
"step": 9310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3605171796880289,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003666578783467837,
|
||
|
|
"loss": 3.4089500427246096,
|
||
|
|
"step": 9320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3609040006962778,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003663963988200435,
|
||
|
|
"loss": 3.418537902832031,
|
||
|
|
"step": 9330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36129082170452675,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.00036613476346091623,
|
||
|
|
"loss": 3.421380615234375,
|
||
|
|
"step": 9340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36167764271277575,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.0003658729726618242,
|
||
|
|
"loss": 3.441343307495117,
|
||
|
|
"step": 9350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3620644637210247,
|
||
|
|
"grad_norm": 0.2431640625,
|
||
|
|
"learning_rate": 0.0003656110268154228,
|
||
|
|
"loss": 3.3920360565185548,
|
||
|
|
"step": 9360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3624512847292736,
|
||
|
|
"grad_norm": 0.287109375,
|
||
|
|
"learning_rate": 0.0003653489263146001,
|
||
|
|
"loss": 3.4329292297363283,
|
||
|
|
"step": 9370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3628381057375226,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00036508667155247604,
|
||
|
|
"loss": 3.4128143310546877,
|
||
|
|
"step": 9380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36322492674577156,
|
||
|
|
"grad_norm": 0.2353515625,
|
||
|
|
"learning_rate": 0.00036482426292240187,
|
||
|
|
"loss": 3.453541564941406,
|
||
|
|
"step": 9390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3636117477540205,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00036456170081795977,
|
||
|
|
"loss": 3.422444152832031,
|
||
|
|
"step": 9400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3636117477540205,
|
||
|
|
"eval_loss": 3.9018635749816895,
|
||
|
|
"eval_runtime": 10.0801,
|
||
|
|
"eval_samples_per_second": 32.44,
|
||
|
|
"eval_steps_per_second": 2.083,
|
||
|
|
"step": 9400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3639985687622695,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00036429898563296195,
|
||
|
|
"loss": 3.4397987365722655,
|
||
|
|
"step": 9410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36438538977051843,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.00036403611776145037,
|
||
|
|
"loss": 3.4169151306152346,
|
||
|
|
"step": 9420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36477221077876737,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00036377309759769594,
|
||
|
|
"loss": 3.412881088256836,
|
||
|
|
"step": 9430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36515903178701636,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.00036350992553619795,
|
||
|
|
"loss": 3.4701351165771483,
|
||
|
|
"step": 9440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3655458527952653,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003632466019716836,
|
||
|
|
"loss": 3.4168781280517577,
|
||
|
|
"step": 9450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36593267380351424,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00036298312729910734,
|
||
|
|
"loss": 3.415391540527344,
|
||
|
|
"step": 9460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36631949481176324,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00036271950191365,
|
||
|
|
"loss": 3.4176471710205076,
|
||
|
|
"step": 9470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3667063158200122,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00036245572621071897,
|
||
|
|
"loss": 3.3979991912841796,
|
||
|
|
"step": 9480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36709313682826117,
|
||
|
|
"grad_norm": 0.263671875,
|
||
|
|
"learning_rate": 0.0003621918005859466,
|
||
|
|
"loss": 3.4499542236328127,
|
||
|
|
"step": 9490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3674799578365101,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.00036192772543519044,
|
||
|
|
"loss": 3.3763500213623048,
|
||
|
|
"step": 9500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3674799578365101,
|
||
|
|
"eval_loss": 3.875546932220459,
|
||
|
|
"eval_runtime": 10.0346,
|
||
|
|
"eval_samples_per_second": 32.587,
|
||
|
|
"eval_steps_per_second": 2.093,
|
||
|
|
"step": 9500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36786677884475905,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.000361663501154532,
|
||
|
|
"loss": 3.4191715240478517,
|
||
|
|
"step": 9510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36825359985300804,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.0003613991281402768,
|
||
|
|
"loss": 3.4267364501953126,
|
||
|
|
"step": 9520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.368640420861257,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0003611346067889533,
|
||
|
|
"loss": 3.4404712677001954,
|
||
|
|
"step": 9530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3690272418695059,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003608699374973123,
|
||
|
|
"loss": 3.358293151855469,
|
||
|
|
"step": 9540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3694140628777549,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003606051206623267,
|
||
|
|
"loss": 3.370107650756836,
|
||
|
|
"step": 9550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36980088388600385,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00036034015668119065,
|
||
|
|
"loss": 3.4077564239501954,
|
||
|
|
"step": 9560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3701877048942528,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.0003600750459513189,
|
||
|
|
"loss": 3.4234764099121096,
|
||
|
|
"step": 9570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3705745259025018,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.0003598097888703466,
|
||
|
|
"loss": 3.3569828033447267,
|
||
|
|
"step": 9580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3709613469107507,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0003595443858361279,
|
||
|
|
"loss": 3.4103553771972654,
|
||
|
|
"step": 9590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37134816791899966,
|
||
|
|
"grad_norm": 0.2890625,
|
||
|
|
"learning_rate": 0.00035927883724673644,
|
||
|
|
"loss": 3.4405521392822265,
|
||
|
|
"step": 9600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37134816791899966,
|
||
|
|
"eval_loss": 3.883023738861084,
|
||
|
|
"eval_runtime": 10.0477,
|
||
|
|
"eval_samples_per_second": 32.545,
|
||
|
|
"eval_steps_per_second": 2.09,
|
||
|
|
"step": 9600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37173498892724866,
|
||
|
|
"grad_norm": 0.27734375,
|
||
|
|
"learning_rate": 0.0003590131435004636,
|
||
|
|
"loss": 3.419297790527344,
|
||
|
|
"step": 9610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3721218099354976,
|
||
|
|
"grad_norm": 0.296875,
|
||
|
|
"learning_rate": 0.0003587473049958191,
|
||
|
|
"loss": 3.4245758056640625,
|
||
|
|
"step": 9620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37250863094374653,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003584813221315292,
|
||
|
|
"loss": 3.40887336730957,
|
||
|
|
"step": 9630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3728954519519955,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00035821519530653695,
|
||
|
|
"loss": 3.4069156646728516,
|
||
|
|
"step": 9640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37328227296024447,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0003579489249200014,
|
||
|
|
"loss": 3.4280853271484375,
|
||
|
|
"step": 9650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3736690939684934,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003576825113712968,
|
||
|
|
"loss": 3.422659683227539,
|
||
|
|
"step": 9660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3740559149767424,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00035741595506001223,
|
||
|
|
"loss": 3.39489631652832,
|
||
|
|
"step": 9670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37444273598499134,
|
||
|
|
"grad_norm": 0.2373046875,
|
||
|
|
"learning_rate": 0.0003571492563859508,
|
||
|
|
"loss": 3.4341705322265623,
|
||
|
|
"step": 9680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3748295569932403,
|
||
|
|
"grad_norm": 0.271484375,
|
||
|
|
"learning_rate": 0.000356882415749129,
|
||
|
|
"loss": 3.4342811584472654,
|
||
|
|
"step": 9690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37521637800148927,
|
||
|
|
"grad_norm": 0.248046875,
|
||
|
|
"learning_rate": 0.0003566154335497767,
|
||
|
|
"loss": 3.4240440368652343,
|
||
|
|
"step": 9700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37521637800148927,
|
||
|
|
"eval_loss": 3.8837051391601562,
|
||
|
|
"eval_runtime": 10.0201,
|
||
|
|
"eval_samples_per_second": 32.634,
|
||
|
|
"eval_steps_per_second": 2.096,
|
||
|
|
"step": 9700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3756031990097382,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003563483101883356,
|
||
|
|
"loss": 3.4255138397216798,
|
||
|
|
"step": 9710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3759900200179872,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.0003560810460654595,
|
||
|
|
"loss": 3.4189659118652345,
|
||
|
|
"step": 9720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37637684102623614,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.0003558136415820131,
|
||
|
|
"loss": 3.4093570709228516,
|
||
|
|
"step": 9730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3767636620344851,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003555460971390717,
|
||
|
|
"loss": 3.3877967834472655,
|
||
|
|
"step": 9740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3771504830427341,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00035527841313792046,
|
||
|
|
"loss": 3.4528697967529296,
|
||
|
|
"step": 9750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.377537304050983,
|
||
|
|
"grad_norm": 0.265625,
|
||
|
|
"learning_rate": 0.000355010589980054,
|
||
|
|
"loss": 3.4023715972900392,
|
||
|
|
"step": 9760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37792412505923195,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0003547426280671756,
|
||
|
|
"loss": 3.4020904541015624,
|
||
|
|
"step": 9770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37831094606748095,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00035447452780119654,
|
||
|
|
"loss": 3.3937957763671873,
|
||
|
|
"step": 9780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3786977670757299,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003542062895842358,
|
||
|
|
"loss": 3.420269775390625,
|
||
|
|
"step": 9790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3790845880839788,
|
||
|
|
"grad_norm": 0.25,
|
||
|
|
"learning_rate": 0.0003539379138186191,
|
||
|
|
"loss": 3.3751293182373048,
|
||
|
|
"step": 9800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3790845880839788,
|
||
|
|
"eval_loss": 3.911559820175171,
|
||
|
|
"eval_runtime": 10.0374,
|
||
|
|
"eval_samples_per_second": 32.578,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 9800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3794714090922278,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00035366940090687864,
|
||
|
|
"loss": 3.405370330810547,
|
||
|
|
"step": 9810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37985823010047676,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00035340075125175213,
|
||
|
|
"loss": 3.389670181274414,
|
||
|
|
"step": 9820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3802450511087257,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.0003531319652561825,
|
||
|
|
"loss": 3.3794952392578126,
|
||
|
|
"step": 9830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3806318721169747,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00035286304332331724,
|
||
|
|
"loss": 3.402521514892578,
|
||
|
|
"step": 9840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38101869312522363,
|
||
|
|
"grad_norm": 0.2451171875,
|
||
|
|
"learning_rate": 0.0003525939858565076,
|
||
|
|
"loss": 3.430156707763672,
|
||
|
|
"step": 9850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38140551413347257,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.0003523247932593081,
|
||
|
|
"loss": 3.4212566375732423,
|
||
|
|
"step": 9860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38179233514172156,
|
||
|
|
"grad_norm": 0.251953125,
|
||
|
|
"learning_rate": 0.00035205546593547606,
|
||
|
|
"loss": 3.4265159606933593,
|
||
|
|
"step": 9870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3821791561499705,
|
||
|
|
"grad_norm": 0.259765625,
|
||
|
|
"learning_rate": 0.00035178600428897075,
|
||
|
|
"loss": 3.4060302734375,
|
||
|
|
"step": 9880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38256597715821944,
|
||
|
|
"grad_norm": 0.2490234375,
|
||
|
|
"learning_rate": 0.0003515164087239529,
|
||
|
|
"loss": 3.4170459747314452,
|
||
|
|
"step": 9890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38295279816646843,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00035124667964478434,
|
||
|
|
"loss": 3.3865825653076174,
|
||
|
|
"step": 9900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38295279816646843,
|
||
|
|
"eval_loss": 3.8762545585632324,
|
||
|
|
"eval_runtime": 10.0396,
|
||
|
|
"eval_samples_per_second": 32.571,
|
||
|
|
"eval_steps_per_second": 2.092,
|
||
|
|
"step": 9900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38333961917471737,
|
||
|
|
"grad_norm": 0.25390625,
|
||
|
|
"learning_rate": 0.00035097681745602705,
|
||
|
|
"loss": 3.4357513427734374,
|
||
|
|
"step": 9910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3837264401829663,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003507068225624423,
|
||
|
|
"loss": 3.393326187133789,
|
||
|
|
"step": 9920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3841132611912153,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00035043669536899104,
|
||
|
|
"loss": 3.4443363189697265,
|
||
|
|
"step": 9930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38450008219946424,
|
||
|
|
"grad_norm": 0.255859375,
|
||
|
|
"learning_rate": 0.00035016643628083194,
|
||
|
|
"loss": 3.415657806396484,
|
||
|
|
"step": 9940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38488690320771324,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.00034989604570332224,
|
||
|
|
"loss": 3.425619125366211,
|
||
|
|
"step": 9950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3852737242159622,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.00034962552404201573,
|
||
|
|
"loss": 3.3953113555908203,
|
||
|
|
"step": 9960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3856605452242111,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0003493548717026633,
|
||
|
|
"loss": 3.4209903717041015,
|
||
|
|
"step": 9970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3860473662324601,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.00034908408909121145,
|
||
|
|
"loss": 3.3913738250732424,
|
||
|
|
"step": 9980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38643418724070905,
|
||
|
|
"grad_norm": 0.267578125,
|
||
|
|
"learning_rate": 0.0003488131766138026,
|
||
|
|
"loss": 3.3826934814453127,
|
||
|
|
"step": 9990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.386821008248958,
|
||
|
|
"grad_norm": 0.2421875,
|
||
|
|
"learning_rate": 0.0003485421346767733,
|
||
|
|
"loss": 3.424589157104492,
|
||
|
|
"step": 10000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.386821008248958,
|
||
|
|
"eval_loss": 3.8753442764282227,
|
||
|
|
"eval_runtime": 10.0275,
|
||
|
|
"eval_samples_per_second": 32.61,
|
||
|
|
"eval_steps_per_second": 2.094,
|
||
|
|
"step": 10000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.387207829257207,
|
||
|
|
"grad_norm": 0.314453125,
|
||
|
|
"learning_rate": 0.0003482709636866547,
|
||
|
|
"loss": 3.408251953125,
|
||
|
|
"step": 10010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3875946502654559,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003479996640501714,
|
||
|
|
"loss": 3.392727279663086,
|
||
|
|
"step": 10020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38798147127370486,
|
||
|
|
"grad_norm": 0.232421875,
|
||
|
|
"learning_rate": 0.00034772823617424103,
|
||
|
|
"loss": 3.420479583740234,
|
||
|
|
"step": 10030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38836829228195385,
|
||
|
|
"grad_norm": 0.30078125,
|
||
|
|
"learning_rate": 0.0003474566804659735,
|
||
|
|
"loss": 3.4015888214111327,
|
||
|
|
"step": 10040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3887551132902028,
|
||
|
|
"grad_norm": 0.2734375,
|
||
|
|
"learning_rate": 0.0003471849973326705,
|
||
|
|
"loss": 3.4193763732910156,
|
||
|
|
"step": 10050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38914193429845173,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003469131871818247,
|
||
|
|
"loss": 3.402726745605469,
|
||
|
|
"step": 10060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3895287553067007,
|
||
|
|
"grad_norm": 0.2578125,
|
||
|
|
"learning_rate": 0.00034664125042111935,
|
||
|
|
"loss": 3.384511947631836,
|
||
|
|
"step": 10070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38991557631494966,
|
||
|
|
"grad_norm": 0.26171875,
|
||
|
|
"learning_rate": 0.0003463691874584277,
|
||
|
|
"loss": 3.396898651123047,
|
||
|
|
"step": 10080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3903023973231986,
|
||
|
|
"grad_norm": 0.2392578125,
|
||
|
|
"learning_rate": 0.0003460969987018123,
|
||
|
|
"loss": 3.417254638671875,
|
||
|
|
"step": 10090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3906892183314476,
|
||
|
|
"grad_norm": 0.26953125,
|
||
|
|
"learning_rate": 0.0003458246845595241,
|
||
|
|
"loss": 3.3982513427734373,
|
||
|
|
"step": 10100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3906892183314476,
|
||
|
|
"eval_loss": 3.85249662399292,
|
||
|
|
"eval_runtime": 10.0512,
|
||
|
|
"eval_samples_per_second": 32.533,
|
||
|
|
"eval_steps_per_second": 2.089,
|
||
|
|
"step": 10100
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 25852,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 1,
|
||
|
|
"save_steps": 100,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 6.379493380639949e+18,
|
||
|
|
"train_batch_size": 32,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|