| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.3645320197044333, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011034482758620689, | |
| "grad_norm": 0.2625392973423004, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 1.4548, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.022068965517241378, | |
| "grad_norm": 0.255765438079834, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.3014, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03310344827586207, | |
| "grad_norm": 0.20604930818080902, | |
| "learning_rate": 2.1e-05, | |
| "loss": 1.1312, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.044137931034482755, | |
| "grad_norm": 0.20505264401435852, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.0359, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05517241379310345, | |
| "grad_norm": 0.21282243728637695, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.9636, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06620689655172414, | |
| "grad_norm": 0.3072029948234558, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.9305, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07724137931034483, | |
| "grad_norm": 0.23971162736415863, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.9174, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08827586206896551, | |
| "grad_norm": 0.2698459327220917, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.8612, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0993103448275862, | |
| "grad_norm": 0.3638197183609009, | |
| "learning_rate": 6.3e-05, | |
| "loss": 0.8511, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "grad_norm": 0.27103668451309204, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8312, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12137931034482759, | |
| "grad_norm": 0.3014815151691437, | |
| "learning_rate": 7.7e-05, | |
| "loss": 0.7998, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.13241379310344828, | |
| "grad_norm": 0.2660246789455414, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.8015, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.14344827586206896, | |
| "grad_norm": 0.30547192692756653, | |
| "learning_rate": 9.1e-05, | |
| "loss": 0.8076, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.15448275862068966, | |
| "grad_norm": 0.3411053717136383, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.7944, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.16551724137931034, | |
| "grad_norm": 0.31632518768310547, | |
| "learning_rate": 9.999685283773504e-05, | |
| "loss": 0.7816, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.17655172413793102, | |
| "grad_norm": 0.36275714635849, | |
| "learning_rate": 9.998187325055106e-05, | |
| "loss": 0.7829, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.18758620689655173, | |
| "grad_norm": 0.32636329531669617, | |
| "learning_rate": 9.995456138403733e-05, | |
| "loss": 0.7791, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1986206896551724, | |
| "grad_norm": 0.3593531847000122, | |
| "learning_rate": 9.991492397698826e-05, | |
| "loss": 0.7709, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2096551724137931, | |
| "grad_norm": 0.3396188020706177, | |
| "learning_rate": 9.986297080934089e-05, | |
| "loss": 0.7686, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "grad_norm": 0.37299859523773193, | |
| "learning_rate": 9.979871469976196e-05, | |
| "loss": 0.7603, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2317241379310345, | |
| "grad_norm": 0.34510111808776855, | |
| "learning_rate": 9.972217150248503e-05, | |
| "loss": 0.7712, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.24275862068965517, | |
| "grad_norm": 0.3415180444717407, | |
| "learning_rate": 9.963336010339868e-05, | |
| "loss": 0.7608, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2537931034482759, | |
| "grad_norm": 0.38360491394996643, | |
| "learning_rate": 9.953230241538674e-05, | |
| "loss": 0.777, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.26482758620689656, | |
| "grad_norm": 0.3310436010360718, | |
| "learning_rate": 9.941902337292155e-05, | |
| "loss": 0.7498, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 0.34418851137161255, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.7456, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2868965517241379, | |
| "grad_norm": 0.35113954544067383, | |
| "learning_rate": 9.915591603280631e-05, | |
| "loss": 0.7452, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2979310344827586, | |
| "grad_norm": 0.3386627435684204, | |
| "learning_rate": 9.900615265295552e-05, | |
| "loss": 0.7377, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.30896551724137933, | |
| "grad_norm": 0.3772924542427063, | |
| "learning_rate": 9.884429773823239e-05, | |
| "loss": 0.736, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.36270612478256226, | |
| "learning_rate": 9.867039122391527e-05, | |
| "loss": 0.7195, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "grad_norm": 0.33858415484428406, | |
| "learning_rate": 9.848447601883435e-05, | |
| "loss": 0.716, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.34206896551724136, | |
| "grad_norm": 0.3483560085296631, | |
| "learning_rate": 9.828659799478456e-05, | |
| "loss": 0.7287, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.35310344827586204, | |
| "grad_norm": 0.3399845063686371, | |
| "learning_rate": 9.807680597520746e-05, | |
| "loss": 0.7073, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3641379310344828, | |
| "grad_norm": 0.3727072477340698, | |
| "learning_rate": 9.785515172314463e-05, | |
| "loss": 0.7112, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.37517241379310345, | |
| "grad_norm": 0.32936757802963257, | |
| "learning_rate": 9.762168992846614e-05, | |
| "loss": 0.7263, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.38620689655172413, | |
| "grad_norm": 0.3701590299606323, | |
| "learning_rate": 9.737647819437645e-05, | |
| "loss": 0.7145, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3972413793103448, | |
| "grad_norm": 0.3669060468673706, | |
| "learning_rate": 9.711957702320175e-05, | |
| "loss": 0.7127, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.40827586206896554, | |
| "grad_norm": 0.3532595634460449, | |
| "learning_rate": 9.685104980146193e-05, | |
| "loss": 0.7064, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4193103448275862, | |
| "grad_norm": 0.37261611223220825, | |
| "learning_rate": 9.657096278423093e-05, | |
| "loss": 0.697, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4303448275862069, | |
| "grad_norm": 0.37206652760505676, | |
| "learning_rate": 9.627938507878917e-05, | |
| "loss": 0.709, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4413793103448276, | |
| "grad_norm": 0.3685338497161865, | |
| "learning_rate": 9.597638862757255e-05, | |
| "loss": 0.6987, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.45241379310344826, | |
| "grad_norm": 0.36724820733070374, | |
| "learning_rate": 9.566204819042152e-05, | |
| "loss": 0.711, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.463448275862069, | |
| "grad_norm": 0.36288705468177795, | |
| "learning_rate": 9.533644132613541e-05, | |
| "loss": 0.712, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.47448275862068967, | |
| "grad_norm": 0.3663863241672516, | |
| "learning_rate": 9.49996483733358e-05, | |
| "loss": 0.6997, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.48551724137931035, | |
| "grad_norm": 0.3941881060600281, | |
| "learning_rate": 9.465175243064428e-05, | |
| "loss": 0.6958, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.496551724137931, | |
| "grad_norm": 0.3758893311023712, | |
| "learning_rate": 9.4292839336179e-05, | |
| "loss": 0.7035, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5075862068965518, | |
| "grad_norm": 0.3466153144836426, | |
| "learning_rate": 9.39229976463755e-05, | |
| "loss": 0.6833, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.5186206896551724, | |
| "grad_norm": 0.3633442521095276, | |
| "learning_rate": 9.354231861413668e-05, | |
| "loss": 0.6893, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5296551724137931, | |
| "grad_norm": 0.3545131981372833, | |
| "learning_rate": 9.315089616631752e-05, | |
| "loss": 0.6967, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5406896551724137, | |
| "grad_norm": 0.3482682406902313, | |
| "learning_rate": 9.274882688055005e-05, | |
| "loss": 0.6953, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 0.36485904455184937, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.6942, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5627586206896552, | |
| "grad_norm": 0.35146597027778625, | |
| "learning_rate": 9.191314721596072e-05, | |
| "loss": 0.6749, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5737931034482758, | |
| "grad_norm": 0.4116743206977844, | |
| "learning_rate": 9.147974302859157e-05, | |
| "loss": 0.6892, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5848275862068966, | |
| "grad_norm": 0.40801239013671875, | |
| "learning_rate": 9.103610433530483e-05, | |
| "loss": 0.6774, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5958620689655172, | |
| "grad_norm": 0.3636874258518219, | |
| "learning_rate": 9.058234059730976e-05, | |
| "loss": 0.6794, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.6068965517241379, | |
| "grad_norm": 0.40341806411743164, | |
| "learning_rate": 9.01185637740189e-05, | |
| "loss": 0.6923, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6179310344827587, | |
| "grad_norm": 0.35487258434295654, | |
| "learning_rate": 8.964488829542377e-05, | |
| "loss": 0.6863, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.6289655172413793, | |
| "grad_norm": 0.38306641578674316, | |
| "learning_rate": 8.916143103386093e-05, | |
| "loss": 0.6935, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.3719683587551117, | |
| "learning_rate": 8.866831127517557e-05, | |
| "loss": 0.6848, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6510344827586206, | |
| "grad_norm": 0.36525729298591614, | |
| "learning_rate": 8.81656506892894e-05, | |
| "loss": 0.6723, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6620689655172414, | |
| "grad_norm": 0.3747231662273407, | |
| "learning_rate": 8.765357330018056e-05, | |
| "loss": 0.6684, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6731034482758621, | |
| "grad_norm": 0.3755883574485779, | |
| "learning_rate": 8.71322054552824e-05, | |
| "loss": 0.6671, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6841379310344827, | |
| "grad_norm": 0.3866102397441864, | |
| "learning_rate": 8.660167579430927e-05, | |
| "loss": 0.6721, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6951724137931035, | |
| "grad_norm": 0.3862460255622864, | |
| "learning_rate": 8.606211521751652e-05, | |
| "loss": 0.6806, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.7062068965517241, | |
| "grad_norm": 0.3793068528175354, | |
| "learning_rate": 8.551365685340285e-05, | |
| "loss": 0.6919, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.7172413793103448, | |
| "grad_norm": 0.3663445711135864, | |
| "learning_rate": 8.495643602586287e-05, | |
| "loss": 0.6767, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.7282758620689656, | |
| "grad_norm": 0.365164190530777, | |
| "learning_rate": 8.439059022079789e-05, | |
| "loss": 0.6528, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.7393103448275862, | |
| "grad_norm": 0.37421566247940063, | |
| "learning_rate": 8.381625905219339e-05, | |
| "loss": 0.6528, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.7503448275862069, | |
| "grad_norm": 0.4169900715351105, | |
| "learning_rate": 8.32335842276713e-05, | |
| "loss": 0.6619, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7613793103448275, | |
| "grad_norm": 0.3822152316570282, | |
| "learning_rate": 8.264270951352581e-05, | |
| "loss": 0.6685, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7724137931034483, | |
| "grad_norm": 0.37115615606307983, | |
| "learning_rate": 8.20437806992512e-05, | |
| "loss": 0.6634, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.783448275862069, | |
| "grad_norm": 0.40461844205856323, | |
| "learning_rate": 8.143694556157046e-05, | |
| "loss": 0.6688, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7944827586206896, | |
| "grad_norm": 0.3837352693080902, | |
| "learning_rate": 8.082235382797349e-05, | |
| "loss": 0.6707, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.8055172413793104, | |
| "grad_norm": 0.3932512104511261, | |
| "learning_rate": 8.020015713977427e-05, | |
| "loss": 0.6666, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.8165517241379311, | |
| "grad_norm": 0.3982972204685211, | |
| "learning_rate": 7.957050901469545e-05, | |
| "loss": 0.6764, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 0.3910225033760071, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.6721, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8386206896551724, | |
| "grad_norm": 0.4260442852973938, | |
| "learning_rate": 7.828948167911074e-05, | |
| "loss": 0.6595, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.8496551724137931, | |
| "grad_norm": 0.37996965646743774, | |
| "learning_rate": 7.763841854293145e-05, | |
| "loss": 0.6551, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8606896551724138, | |
| "grad_norm": 0.38492435216903687, | |
| "learning_rate": 7.698053604053922e-05, | |
| "loss": 0.6623, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8717241379310345, | |
| "grad_norm": 0.3707030117511749, | |
| "learning_rate": 7.631599649459744e-05, | |
| "loss": 0.6457, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8827586206896552, | |
| "grad_norm": 0.37289130687713623, | |
| "learning_rate": 7.564496387029532e-05, | |
| "loss": 0.6529, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8937931034482759, | |
| "grad_norm": 0.39539724588394165, | |
| "learning_rate": 7.496760373489202e-05, | |
| "loss": 0.6442, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.9048275862068965, | |
| "grad_norm": 0.3812501132488251, | |
| "learning_rate": 7.428408321686541e-05, | |
| "loss": 0.6484, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.9158620689655173, | |
| "grad_norm": 0.3905741572380066, | |
| "learning_rate": 7.35945709646756e-05, | |
| "loss": 0.6479, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.926896551724138, | |
| "grad_norm": 0.3890450596809387, | |
| "learning_rate": 7.289923710515339e-05, | |
| "loss": 0.6569, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.9379310344827586, | |
| "grad_norm": 0.3712526559829712, | |
| "learning_rate": 7.219825320152411e-05, | |
| "loss": 0.6443, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9489655172413793, | |
| "grad_norm": 0.3965970277786255, | |
| "learning_rate": 7.149179221107694e-05, | |
| "loss": 0.6515, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.37230828404426575, | |
| "learning_rate": 7.078002844249032e-05, | |
| "loss": 0.6394, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9710344827586207, | |
| "grad_norm": 0.4104078710079193, | |
| "learning_rate": 7.006313751282372e-05, | |
| "loss": 0.6424, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9820689655172414, | |
| "grad_norm": 0.39799052476882935, | |
| "learning_rate": 6.934129630418701e-05, | |
| "loss": 0.655, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.993103448275862, | |
| "grad_norm": 0.3745689392089844, | |
| "learning_rate": 6.861468292009727e-05, | |
| "loss": 0.646, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.0041379310344827, | |
| "grad_norm": 0.4002785384654999, | |
| "learning_rate": 6.788347664153447e-05, | |
| "loss": 0.6665, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.0151724137931035, | |
| "grad_norm": 0.42168787121772766, | |
| "learning_rate": 6.714785788270658e-05, | |
| "loss": 0.5686, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.0262068965517241, | |
| "grad_norm": 0.3978714644908905, | |
| "learning_rate": 6.640800814653503e-05, | |
| "loss": 0.5379, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.0372413793103448, | |
| "grad_norm": 0.4157312214374542, | |
| "learning_rate": 6.566410997987163e-05, | |
| "loss": 0.5554, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.0482758620689656, | |
| "grad_norm": 0.3982793092727661, | |
| "learning_rate": 6.49163469284578e-05, | |
| "loss": 0.5526, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.0593103448275862, | |
| "grad_norm": 0.41753092408180237, | |
| "learning_rate": 6.416490349163748e-05, | |
| "loss": 0.5535, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0703448275862069, | |
| "grad_norm": 0.4113280475139618, | |
| "learning_rate": 6.340996507683458e-05, | |
| "loss": 0.5645, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.0813793103448275, | |
| "grad_norm": 0.4069710671901703, | |
| "learning_rate": 6.265171795380659e-05, | |
| "loss": 0.536, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.0924137931034483, | |
| "grad_norm": 0.41753774881362915, | |
| "learning_rate": 6.189034920868522e-05, | |
| "loss": 0.5474, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 0.3889447748661041, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.5237, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1144827586206896, | |
| "grad_norm": 0.41418376564979553, | |
| "learning_rate": 6.0358999001406156e-05, | |
| "loss": 0.5591, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.1255172413793104, | |
| "grad_norm": 0.4075161814689636, | |
| "learning_rate": 5.9589395376998e-05, | |
| "loss": 0.5473, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.136551724137931, | |
| "grad_norm": 0.38883715867996216, | |
| "learning_rate": 5.8817425712769794e-05, | |
| "loss": 0.5345, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.1475862068965517, | |
| "grad_norm": 0.40612930059432983, | |
| "learning_rate": 5.804328048068492e-05, | |
| "loss": 0.5512, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.1586206896551725, | |
| "grad_norm": 0.4203000068664551, | |
| "learning_rate": 5.7267150689495644e-05, | |
| "loss": 0.546, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.1696551724137931, | |
| "grad_norm": 0.40659859776496887, | |
| "learning_rate": 5.648922783761443e-05, | |
| "loss": 0.57, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.1806896551724138, | |
| "grad_norm": 0.3896077871322632, | |
| "learning_rate": 5.570970386586469e-05, | |
| "loss": 0.5565, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.1917241379310344, | |
| "grad_norm": 0.4040294587612152, | |
| "learning_rate": 5.492877111012218e-05, | |
| "loss": 0.5489, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.2027586206896552, | |
| "grad_norm": 0.41124486923217773, | |
| "learning_rate": 5.414662225385903e-05, | |
| "loss": 0.5467, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.2137931034482758, | |
| "grad_norm": 0.4190484583377838, | |
| "learning_rate": 5.336345028060199e-05, | |
| "loss": 0.5508, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2248275862068965, | |
| "grad_norm": 0.4511507749557495, | |
| "learning_rate": 5.257944842631658e-05, | |
| "loss": 0.5538, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.2358620689655173, | |
| "grad_norm": 0.4155430495738983, | |
| "learning_rate": 5.179481013172912e-05, | |
| "loss": 0.5503, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.246896551724138, | |
| "grad_norm": 0.40163281559944153, | |
| "learning_rate": 5.100972899459796e-05, | |
| "loss": 0.5406, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.2579310344827586, | |
| "grad_norm": 0.4043784737586975, | |
| "learning_rate": 5.022439872194629e-05, | |
| "loss": 0.5494, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.2689655172413792, | |
| "grad_norm": 0.4234471917152405, | |
| "learning_rate": 4.943901308226771e-05, | |
| "loss": 0.5424, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.41634508967399597, | |
| "learning_rate": 4.865376585771687e-05, | |
| "loss": 0.5385, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.2910344827586206, | |
| "grad_norm": 0.3989020586013794, | |
| "learning_rate": 4.7868850796296495e-05, | |
| "loss": 0.5596, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.3020689655172415, | |
| "grad_norm": 0.4220573902130127, | |
| "learning_rate": 4.708446156405307e-05, | |
| "loss": 0.5546, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.3131034482758621, | |
| "grad_norm": 0.4011528789997101, | |
| "learning_rate": 4.630079169729257e-05, | |
| "loss": 0.549, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.3241379310344827, | |
| "grad_norm": 0.3943440318107605, | |
| "learning_rate": 4.551803455482833e-05, | |
| "loss": 0.5369, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3351724137931034, | |
| "grad_norm": 0.3866331875324249, | |
| "learning_rate": 4.473638327027259e-05, | |
| "loss": 0.5369, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.3462068965517242, | |
| "grad_norm": 0.4149519205093384, | |
| "learning_rate": 4.395603070438373e-05, | |
| "loss": 0.5462, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.3572413793103448, | |
| "grad_norm": 0.395431786775589, | |
| "learning_rate": 4.31771693974807e-05, | |
| "loss": 0.5453, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.3682758620689655, | |
| "grad_norm": 0.38319146633148193, | |
| "learning_rate": 4.239999152193664e-05, | |
| "loss": 0.5441, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.4112420082092285, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.5348, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.390344827586207, | |
| "grad_norm": 0.40384915471076965, | |
| "learning_rate": 4.085145263029726e-05, | |
| "loss": 0.5532, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.4013793103448275, | |
| "grad_norm": 0.4221806228160858, | |
| "learning_rate": 4.008047369300218e-05, | |
| "loss": 0.5474, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.4124137931034482, | |
| "grad_norm": 0.432949960231781, | |
| "learning_rate": 3.9311942250394276e-05, | |
| "loss": 0.5206, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.423448275862069, | |
| "grad_norm": 0.4182206988334656, | |
| "learning_rate": 3.8546047926107256e-05, | |
| "loss": 0.544, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.4344827586206896, | |
| "grad_norm": 0.4261842370033264, | |
| "learning_rate": 3.778297969310529e-05, | |
| "loss": 0.5511, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4455172413793105, | |
| "grad_norm": 0.4241354465484619, | |
| "learning_rate": 3.7022925827056884e-05, | |
| "loss": 0.5392, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.456551724137931, | |
| "grad_norm": 0.3890000283718109, | |
| "learning_rate": 3.62660738598805e-05, | |
| "loss": 0.5457, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.4675862068965517, | |
| "grad_norm": 0.41937190294265747, | |
| "learning_rate": 3.551261053347404e-05, | |
| "loss": 0.531, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.4786206896551723, | |
| "grad_norm": 0.40612250566482544, | |
| "learning_rate": 3.4762721753638995e-05, | |
| "loss": 0.5333, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.489655172413793, | |
| "grad_norm": 0.41920045018196106, | |
| "learning_rate": 3.401659254421094e-05, | |
| "loss": 0.547, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.5006896551724138, | |
| "grad_norm": 0.42577028274536133, | |
| "learning_rate": 3.3274407001407735e-05, | |
| "loss": 0.532, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.5117241379310344, | |
| "grad_norm": 0.40364742279052734, | |
| "learning_rate": 3.2536348248406534e-05, | |
| "loss": 0.5337, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.5227586206896553, | |
| "grad_norm": 0.41058188676834106, | |
| "learning_rate": 3.1802598390160784e-05, | |
| "loss": 0.5301, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.533793103448276, | |
| "grad_norm": 0.4022769629955292, | |
| "learning_rate": 3.107333846846872e-05, | |
| "loss": 0.5341, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.5448275862068965, | |
| "grad_norm": 0.39046093821525574, | |
| "learning_rate": 3.0348748417303823e-05, | |
| "loss": 0.5298, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5558620689655172, | |
| "grad_norm": 0.40701109170913696, | |
| "learning_rate": 2.9629007018418985e-05, | |
| "loss": 0.5344, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.5668965517241378, | |
| "grad_norm": 0.4146254360675812, | |
| "learning_rate": 2.8914291857234636e-05, | |
| "loss": 0.533, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.5779310344827586, | |
| "grad_norm": 0.408590704202652, | |
| "learning_rate": 2.8204779279022276e-05, | |
| "loss": 0.5247, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.5889655172413795, | |
| "grad_norm": 0.3929610848426819, | |
| "learning_rate": 2.7500644345393943e-05, | |
| "loss": 0.5412, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.40061962604522705, | |
| "learning_rate": 2.68020607911083e-05, | |
| "loss": 0.5167, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.6110344827586207, | |
| "grad_norm": 0.4084263741970062, | |
| "learning_rate": 2.610920098120424e-05, | |
| "loss": 0.5524, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.6220689655172413, | |
| "grad_norm": 0.4147377014160156, | |
| "learning_rate": 2.5422235868472345e-05, | |
| "loss": 0.5328, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.633103448275862, | |
| "grad_norm": 0.4028978645801544, | |
| "learning_rate": 2.4741334951274947e-05, | |
| "loss": 0.5218, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.6441379310344828, | |
| "grad_norm": 0.4164421856403351, | |
| "learning_rate": 2.40666662317248e-05, | |
| "loss": 0.5282, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 0.4045639932155609, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.5347, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6662068965517243, | |
| "grad_norm": 0.4201739728450775, | |
| "learning_rate": 2.2736689664437217e-05, | |
| "loss": 0.5364, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.677241379310345, | |
| "grad_norm": 0.39254507422447205, | |
| "learning_rate": 2.2081709968516866e-05, | |
| "loss": 0.5143, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.6882758620689655, | |
| "grad_norm": 0.42642974853515625, | |
| "learning_rate": 2.1433618692911467e-05, | |
| "loss": 0.5268, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.6993103448275861, | |
| "grad_norm": 0.41387197375297546, | |
| "learning_rate": 2.0792575744445653e-05, | |
| "loss": 0.5213, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.7103448275862068, | |
| "grad_norm": 0.4264724552631378, | |
| "learning_rate": 2.015873929087482e-05, | |
| "loss": 0.5262, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.7213793103448276, | |
| "grad_norm": 0.40214499831199646, | |
| "learning_rate": 1.95322657218596e-05, | |
| "loss": 0.5169, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.7324137931034482, | |
| "grad_norm": 0.4083193838596344, | |
| "learning_rate": 1.8913309610379015e-05, | |
| "loss": 0.5298, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.743448275862069, | |
| "grad_norm": 0.39298340678215027, | |
| "learning_rate": 1.8302023674591935e-05, | |
| "loss": 0.5167, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.7544827586206897, | |
| "grad_norm": 0.42638927698135376, | |
| "learning_rate": 1.7698558740156135e-05, | |
| "loss": 0.5215, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.7655172413793103, | |
| "grad_norm": 0.40725937485694885, | |
| "learning_rate": 1.7103063703014372e-05, | |
| "loss": 0.5297, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.776551724137931, | |
| "grad_norm": 0.3900083005428314, | |
| "learning_rate": 1.6515685492656467e-05, | |
| "loss": 0.5179, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 1.7875862068965516, | |
| "grad_norm": 0.4194444715976715, | |
| "learning_rate": 1.59365690358667e-05, | |
| "loss": 0.5181, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 1.7986206896551724, | |
| "grad_norm": 0.39986348152160645, | |
| "learning_rate": 1.5365857220965275e-05, | |
| "loss": 0.5225, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 1.8096551724137933, | |
| "grad_norm": 0.4116763472557068, | |
| "learning_rate": 1.4803690862552755e-05, | |
| "loss": 0.5203, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.8206896551724139, | |
| "grad_norm": 0.4018745422363281, | |
| "learning_rate": 1.4250208666766235e-05, | |
| "loss": 0.5181, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.8317241379310345, | |
| "grad_norm": 0.3829161822795868, | |
| "learning_rate": 1.3705547197055584e-05, | |
| "loss": 0.5179, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 1.8427586206896551, | |
| "grad_norm": 0.40862053632736206, | |
| "learning_rate": 1.3169840840488501e-05, | |
| "loss": 0.5214, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.8537931034482757, | |
| "grad_norm": 0.43748798966407776, | |
| "learning_rate": 1.2643221774592518e-05, | |
| "loss": 0.5231, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.8648275862068966, | |
| "grad_norm": 0.41041648387908936, | |
| "learning_rate": 1.2125819934742188e-05, | |
| "loss": 0.5161, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 1.8758620689655172, | |
| "grad_norm": 0.3941383957862854, | |
| "learning_rate": 1.1617762982099446e-05, | |
| "loss": 0.5152, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.886896551724138, | |
| "grad_norm": 0.422385036945343, | |
| "learning_rate": 1.1119176272115128e-05, | |
| "loss": 0.5124, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 1.8979310344827587, | |
| "grad_norm": 0.4116532504558563, | |
| "learning_rate": 1.0630182823599399e-05, | |
| "loss": 0.5263, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.9089655172413793, | |
| "grad_norm": 0.4027996361255646, | |
| "learning_rate": 1.0150903288368741e-05, | |
| "loss": 0.5172, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.40005895495414734, | |
| "learning_rate": 9.681455921476839e-06, | |
| "loss": 0.5136, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 0.4146558940410614, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.531, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.9420689655172414, | |
| "grad_norm": 0.4217384457588196, | |
| "learning_rate": 8.772518554642973e-06, | |
| "loss": 0.5333, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.953103448275862, | |
| "grad_norm": 0.4167080819606781, | |
| "learning_rate": 8.333252821395526e-06, | |
| "loss": 0.5138, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 1.9641379310344829, | |
| "grad_norm": 0.40761256217956543, | |
| "learning_rate": 7.904267734541498e-06, | |
| "loss": 0.5215, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 1.9751724137931035, | |
| "grad_norm": 0.3934479057788849, | |
| "learning_rate": 7.485669139732004e-06, | |
| "loss": 0.5095, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 1.986206896551724, | |
| "grad_norm": 0.4060850739479065, | |
| "learning_rate": 7.077560319906695e-06, | |
| "loss": 0.5129, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9972413793103447, | |
| "grad_norm": 0.39918309450149536, | |
| "learning_rate": 6.680041969810203e-06, | |
| "loss": 0.5018, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 2.0082758620689654, | |
| "grad_norm": 0.39984405040740967, | |
| "learning_rate": 6.293212171147206e-06, | |
| "loss": 0.5029, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 2.0193103448275864, | |
| "grad_norm": 0.409271776676178, | |
| "learning_rate": 5.917166368382277e-06, | |
| "loss": 0.4303, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 2.030344827586207, | |
| "grad_norm": 0.4450385570526123, | |
| "learning_rate": 5.5519973451903405e-06, | |
| "loss": 0.4304, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 2.0413793103448277, | |
| "grad_norm": 0.4327320456504822, | |
| "learning_rate": 5.197795201563743e-06, | |
| "loss": 0.4258, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.0524137931034483, | |
| "grad_norm": 0.42840683460235596, | |
| "learning_rate": 4.8546473315813856e-06, | |
| "loss": 0.4298, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 2.063448275862069, | |
| "grad_norm": 0.41301754117012024, | |
| "learning_rate": 4.522638401845547e-06, | |
| "loss": 0.4287, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 2.0744827586206895, | |
| "grad_norm": 0.4132091701030731, | |
| "learning_rate": 4.2018503305916775e-06, | |
| "loss": 0.4211, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 2.08551724137931, | |
| "grad_norm": 0.43061113357543945, | |
| "learning_rate": 3.892362267476313e-06, | |
| "loss": 0.429, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 2.0965517241379312, | |
| "grad_norm": 0.41242024302482605, | |
| "learning_rate": 3.5942505740480582e-06, | |
| "loss": 0.4254, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.107586206896552, | |
| "grad_norm": 0.4244064390659332, | |
| "learning_rate": 3.3075888049065196e-06, | |
| "loss": 0.4263, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 2.1186206896551725, | |
| "grad_norm": 0.4177404046058655, | |
| "learning_rate": 3.03244768955383e-06, | |
| "loss": 0.431, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 2.129655172413793, | |
| "grad_norm": 0.42242953181266785, | |
| "learning_rate": 2.7688951149431595e-06, | |
| "loss": 0.4312, | |
| "step": 1351 | |
| }, | |
| { | |
| "epoch": 2.1406896551724137, | |
| "grad_norm": 0.43014487624168396, | |
| "learning_rate": 2.5169961087286974e-06, | |
| "loss": 0.4265, | |
| "step": 1358 | |
| }, | |
| { | |
| "epoch": 2.1517241379310343, | |
| "grad_norm": 0.43356141448020935, | |
| "learning_rate": 2.276812823220964e-06, | |
| "loss": 0.4384, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.162758620689655, | |
| "grad_norm": 0.4384445548057556, | |
| "learning_rate": 2.048404520051722e-06, | |
| "loss": 0.4187, | |
| "step": 1372 | |
| }, | |
| { | |
| "epoch": 2.173793103448276, | |
| "grad_norm": 0.4111485183238983, | |
| "learning_rate": 1.8318275555520237e-06, | |
| "loss": 0.4313, | |
| "step": 1379 | |
| }, | |
| { | |
| "epoch": 2.1848275862068967, | |
| "grad_norm": 0.4494364261627197, | |
| "learning_rate": 1.6271353668471655e-06, | |
| "loss": 0.4305, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 2.1958620689655173, | |
| "grad_norm": 0.422455757856369, | |
| "learning_rate": 1.4343784586718311e-06, | |
| "loss": 0.4174, | |
| "step": 1393 | |
| }, | |
| { | |
| "epoch": 2.206896551724138, | |
| "grad_norm": 0.4379701614379883, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.4177, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.2179310344827585, | |
| "grad_norm": 0.4343680143356323, | |
| "learning_rate": 1.0848577668543802e-06, | |
| "loss": 0.4176, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 2.228965517241379, | |
| "grad_norm": 0.42176929116249084, | |
| "learning_rate": 9.281802222129765e-07, | |
| "loss": 0.417, | |
| "step": 1414 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4197026491165161, | |
| "learning_rate": 7.836104148243484e-07, | |
| "loss": 0.4201, | |
| "step": 1421 | |
| }, | |
| { | |
| "epoch": 2.251034482758621, | |
| "grad_norm": 0.42568454146385193, | |
| "learning_rate": 6.511840151252169e-07, | |
| "loss": 0.4164, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 2.2620689655172415, | |
| "grad_norm": 0.4492688775062561, | |
| "learning_rate": 5.309336973481683e-07, | |
| "loss": 0.4289, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.273103448275862, | |
| "grad_norm": 0.43771427869796753, | |
| "learning_rate": 4.228891314597694e-07, | |
| "loss": 0.4362, | |
| "step": 1442 | |
| }, | |
| { | |
| "epoch": 2.2841379310344827, | |
| "grad_norm": 0.4285426139831543, | |
| "learning_rate": 3.2707697583995167e-07, | |
| "loss": 0.4161, | |
| "step": 1449 | |
| }, | |
| { | |
| "epoch": 2.2951724137931033, | |
| "grad_norm": 0.433178573846817, | |
| "learning_rate": 2.4352087070443895e-07, | |
| "loss": 0.4252, | |
| "step": 1456 | |
| }, | |
| { | |
| "epoch": 2.306206896551724, | |
| "grad_norm": 0.43515661358833313, | |
| "learning_rate": 1.7224143227190236e-07, | |
| "loss": 0.4155, | |
| "step": 1463 | |
| }, | |
| { | |
| "epoch": 2.317241379310345, | |
| "grad_norm": 0.45461907982826233, | |
| "learning_rate": 1.132562476771959e-07, | |
| "loss": 0.4276, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.3282758620689656, | |
| "grad_norm": 0.42468348145484924, | |
| "learning_rate": 6.657987063200533e-08, | |
| "loss": 0.4256, | |
| "step": 1477 | |
| }, | |
| { | |
| "epoch": 2.3393103448275863, | |
| "grad_norm": 0.4367210865020752, | |
| "learning_rate": 3.2223817833931805e-08, | |
| "loss": 0.429, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 2.350344827586207, | |
| "grad_norm": 0.43067917227745056, | |
| "learning_rate": 1.019656612492592e-08, | |
| "loss": 0.4242, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 2.3613793103448275, | |
| "grad_norm": 0.44099918007850647, | |
| "learning_rate": 5.035503997385949e-10, | |
| "loss": 0.424, | |
| "step": 1498 | |
| } | |
| ], | |
| "logging_steps": 7, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.238031356493824e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |