| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.9995320542817033, |
| "eval_steps": 500, |
| "global_step": 1068, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.004679457182966776, |
| "grad_norm": 2.2908554548637006, |
| "learning_rate": 9.345794392523365e-07, |
| "loss": 0.4939, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.009358914365933552, |
| "grad_norm": 2.016859310665491, |
| "learning_rate": 1.869158878504673e-06, |
| "loss": 0.4723, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.014038371548900327, |
| "grad_norm": 1.5221155414770415, |
| "learning_rate": 2.8037383177570094e-06, |
| "loss": 0.4618, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.018717828731867104, |
| "grad_norm": 0.5002414337717975, |
| "learning_rate": 3.738317757009346e-06, |
| "loss": 0.4567, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.02339728591483388, |
| "grad_norm": 0.9841161532928453, |
| "learning_rate": 4.6728971962616825e-06, |
| "loss": 0.4435, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.028076743097800654, |
| "grad_norm": 0.6160542794694137, |
| "learning_rate": 5.607476635514019e-06, |
| "loss": 0.4308, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.03275620028076743, |
| "grad_norm": 0.4169222842040913, |
| "learning_rate": 6.542056074766355e-06, |
| "loss": 0.4254, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.03743565746373421, |
| "grad_norm": 0.43151336467869167, |
| "learning_rate": 7.476635514018692e-06, |
| "loss": 0.4131, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.04211511464670098, |
| "grad_norm": 0.27790836608550784, |
| "learning_rate": 8.411214953271028e-06, |
| "loss": 0.4151, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.04679457182966776, |
| "grad_norm": 0.3490921524994091, |
| "learning_rate": 9.345794392523365e-06, |
| "loss": 0.3969, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.05147402901263454, |
| "grad_norm": 0.2784386545031061, |
| "learning_rate": 1.02803738317757e-05, |
| "loss": 0.3985, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.05615348619560131, |
| "grad_norm": 0.2805705696670822, |
| "learning_rate": 1.1214953271028037e-05, |
| "loss": 0.4125, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.06083294337856809, |
| "grad_norm": 0.2525496578964846, |
| "learning_rate": 1.2149532710280374e-05, |
| "loss": 0.3947, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.06551240056153486, |
| "grad_norm": 0.23045102014975796, |
| "learning_rate": 1.308411214953271e-05, |
| "loss": 0.3955, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.07019185774450164, |
| "grad_norm": 0.22022938803676545, |
| "learning_rate": 1.4018691588785047e-05, |
| "loss": 0.3967, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.07487131492746842, |
| "grad_norm": 0.23347833396023018, |
| "learning_rate": 1.4953271028037384e-05, |
| "loss": 0.3909, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.0795507721104352, |
| "grad_norm": 0.228913019838675, |
| "learning_rate": 1.588785046728972e-05, |
| "loss": 0.3945, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.08423022929340196, |
| "grad_norm": 0.2633815411021567, |
| "learning_rate": 1.6822429906542056e-05, |
| "loss": 0.3824, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.08890968647636874, |
| "grad_norm": 0.23342700291557855, |
| "learning_rate": 1.7757009345794395e-05, |
| "loss": 0.3862, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.09358914365933552, |
| "grad_norm": 0.23300433901186635, |
| "learning_rate": 1.869158878504673e-05, |
| "loss": 0.3919, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.0982686008423023, |
| "grad_norm": 0.27373389148512267, |
| "learning_rate": 1.9626168224299065e-05, |
| "loss": 0.3779, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.10294805802526907, |
| "grad_norm": 0.228261846212288, |
| "learning_rate": 1.9937565036420395e-05, |
| "loss": 0.3936, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.10762751520823584, |
| "grad_norm": 0.23489411050580397, |
| "learning_rate": 1.9833506763787724e-05, |
| "loss": 0.3843, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.11230697239120262, |
| "grad_norm": 0.27702591802476917, |
| "learning_rate": 1.972944849115505e-05, |
| "loss": 0.382, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.1169864295741694, |
| "grad_norm": 0.225569280525892, |
| "learning_rate": 1.9625390218522374e-05, |
| "loss": 0.3819, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.12166588675713617, |
| "grad_norm": 0.24311963459537528, |
| "learning_rate": 1.95213319458897e-05, |
| "loss": 0.3893, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.12634534394010294, |
| "grad_norm": 0.260175889528399, |
| "learning_rate": 1.9417273673257027e-05, |
| "loss": 0.394, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.13102480112306972, |
| "grad_norm": 0.2556543935052257, |
| "learning_rate": 1.9313215400624352e-05, |
| "loss": 0.3872, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.1357042583060365, |
| "grad_norm": 0.25822395228887385, |
| "learning_rate": 1.9209157127991677e-05, |
| "loss": 0.3904, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.14038371548900327, |
| "grad_norm": 0.2289827168893927, |
| "learning_rate": 1.9105098855359002e-05, |
| "loss": 0.3881, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.14506317267197005, |
| "grad_norm": 0.2782914404861313, |
| "learning_rate": 1.9001040582726327e-05, |
| "loss": 0.389, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.14974262985493683, |
| "grad_norm": 0.2691987795720794, |
| "learning_rate": 1.8896982310093656e-05, |
| "loss": 0.3799, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.1544220870379036, |
| "grad_norm": 0.22215928015704764, |
| "learning_rate": 1.879292403746098e-05, |
| "loss": 0.3798, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.1591015442208704, |
| "grad_norm": 0.2503740618645921, |
| "learning_rate": 1.8688865764828306e-05, |
| "loss": 0.3977, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.16378100140383717, |
| "grad_norm": 0.22564216914588497, |
| "learning_rate": 1.858480749219563e-05, |
| "loss": 0.3744, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.16846045858680392, |
| "grad_norm": 0.25918009956565563, |
| "learning_rate": 1.8480749219562956e-05, |
| "loss": 0.3738, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.1731399157697707, |
| "grad_norm": 0.2393508145158046, |
| "learning_rate": 1.837669094693028e-05, |
| "loss": 0.3857, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.17781937295273748, |
| "grad_norm": 0.24654174660930164, |
| "learning_rate": 1.827263267429761e-05, |
| "loss": 0.3903, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.18249883013570425, |
| "grad_norm": 0.28662121700475657, |
| "learning_rate": 1.8168574401664934e-05, |
| "loss": 0.3898, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.18717828731867103, |
| "grad_norm": 0.23290473813531298, |
| "learning_rate": 1.806451612903226e-05, |
| "loss": 0.3768, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.1918577445016378, |
| "grad_norm": 0.28192870043577783, |
| "learning_rate": 1.7960457856399584e-05, |
| "loss": 0.3879, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.1965372016846046, |
| "grad_norm": 0.27564026147091214, |
| "learning_rate": 1.785639958376691e-05, |
| "loss": 0.3865, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.20121665886757137, |
| "grad_norm": 0.23229839176465608, |
| "learning_rate": 1.7752341311134234e-05, |
| "loss": 0.3841, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.20589611605053815, |
| "grad_norm": 0.25015151051907036, |
| "learning_rate": 1.7648283038501563e-05, |
| "loss": 0.3915, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.21057557323350493, |
| "grad_norm": 0.24218523885962545, |
| "learning_rate": 1.7544224765868888e-05, |
| "loss": 0.3845, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.21525503041647168, |
| "grad_norm": 0.21893648736945756, |
| "learning_rate": 1.7440166493236216e-05, |
| "loss": 0.3796, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.21993448759943846, |
| "grad_norm": 0.24100134035505158, |
| "learning_rate": 1.733610822060354e-05, |
| "loss": 0.3809, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.22461394478240523, |
| "grad_norm": 0.24831115571580942, |
| "learning_rate": 1.7232049947970866e-05, |
| "loss": 0.3942, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.229293401965372, |
| "grad_norm": 0.2270469313952638, |
| "learning_rate": 1.712799167533819e-05, |
| "loss": 0.3822, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.2339728591483388, |
| "grad_norm": 0.21454127210032897, |
| "learning_rate": 1.7023933402705516e-05, |
| "loss": 0.3804, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.23865231633130557, |
| "grad_norm": 0.2784358974791008, |
| "learning_rate": 1.691987513007284e-05, |
| "loss": 0.3891, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.24333177351427235, |
| "grad_norm": 0.25560963065745107, |
| "learning_rate": 1.681581685744017e-05, |
| "loss": 0.3785, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.24801123069723913, |
| "grad_norm": 0.29030559419788876, |
| "learning_rate": 1.6711758584807495e-05, |
| "loss": 0.3859, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.2526906878802059, |
| "grad_norm": 0.23201558406292724, |
| "learning_rate": 1.660770031217482e-05, |
| "loss": 0.3905, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.2573701450631727, |
| "grad_norm": 0.2230334421902013, |
| "learning_rate": 1.6503642039542145e-05, |
| "loss": 0.3826, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.26204960224613943, |
| "grad_norm": 0.2888881421186973, |
| "learning_rate": 1.639958376690947e-05, |
| "loss": 0.3854, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.26672905942910624, |
| "grad_norm": 0.23960115819463665, |
| "learning_rate": 1.6295525494276795e-05, |
| "loss": 0.3812, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.271408516612073, |
| "grad_norm": 0.219894569143036, |
| "learning_rate": 1.6191467221644123e-05, |
| "loss": 0.3772, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.2760879737950398, |
| "grad_norm": 0.24801341276626648, |
| "learning_rate": 1.6087408949011448e-05, |
| "loss": 0.3809, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.28076743097800655, |
| "grad_norm": 0.2654106880151869, |
| "learning_rate": 1.5983350676378773e-05, |
| "loss": 0.3805, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.2854468881609733, |
| "grad_norm": 0.22635581084433476, |
| "learning_rate": 1.5879292403746098e-05, |
| "loss": 0.3768, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.2901263453439401, |
| "grad_norm": 0.21488693776906298, |
| "learning_rate": 1.5775234131113423e-05, |
| "loss": 0.3689, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.29480580252690686, |
| "grad_norm": 0.26183689025810514, |
| "learning_rate": 1.5671175858480748e-05, |
| "loss": 0.3735, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.29948525970987366, |
| "grad_norm": 0.2694040246267707, |
| "learning_rate": 1.5567117585848077e-05, |
| "loss": 0.3808, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.3041647168928404, |
| "grad_norm": 0.24520431614646418, |
| "learning_rate": 1.54630593132154e-05, |
| "loss": 0.3854, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.3088441740758072, |
| "grad_norm": 0.2564447473549132, |
| "learning_rate": 1.5359001040582727e-05, |
| "loss": 0.3862, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.31352363125877397, |
| "grad_norm": 0.23238465003590167, |
| "learning_rate": 1.5254942767950053e-05, |
| "loss": 0.3815, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.3182030884417408, |
| "grad_norm": 0.2800169601088245, |
| "learning_rate": 1.515088449531738e-05, |
| "loss": 0.3814, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.32288254562470753, |
| "grad_norm": 0.22176804004915363, |
| "learning_rate": 1.5046826222684705e-05, |
| "loss": 0.3801, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.32756200280767434, |
| "grad_norm": 0.23327501996058547, |
| "learning_rate": 1.494276795005203e-05, |
| "loss": 0.3712, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.3322414599906411, |
| "grad_norm": 0.23797943468099234, |
| "learning_rate": 1.4838709677419357e-05, |
| "loss": 0.3736, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.33692091717360784, |
| "grad_norm": 0.21540929537029724, |
| "learning_rate": 1.4734651404786682e-05, |
| "loss": 0.3785, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.34160037435657464, |
| "grad_norm": 0.2820377200347258, |
| "learning_rate": 1.4630593132154007e-05, |
| "loss": 0.3754, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.3462798315395414, |
| "grad_norm": 0.2364483513673407, |
| "learning_rate": 1.4526534859521334e-05, |
| "loss": 0.3768, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.3509592887225082, |
| "grad_norm": 0.23107190478630543, |
| "learning_rate": 1.4422476586888659e-05, |
| "loss": 0.384, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.35563874590547495, |
| "grad_norm": 0.2544369764708807, |
| "learning_rate": 1.4318418314255984e-05, |
| "loss": 0.3821, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.36031820308844176, |
| "grad_norm": 0.2733772251893476, |
| "learning_rate": 1.421436004162331e-05, |
| "loss": 0.3863, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.3649976602714085, |
| "grad_norm": 0.2328825486788227, |
| "learning_rate": 1.4110301768990635e-05, |
| "loss": 0.3794, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.3696771174543753, |
| "grad_norm": 0.22430844792190455, |
| "learning_rate": 1.400624349635796e-05, |
| "loss": 0.3801, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.37435657463734207, |
| "grad_norm": 0.24928578537816098, |
| "learning_rate": 1.3902185223725287e-05, |
| "loss": 0.391, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.37903603182030887, |
| "grad_norm": 0.2314083212055415, |
| "learning_rate": 1.3798126951092612e-05, |
| "loss": 0.3767, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.3837154890032756, |
| "grad_norm": 0.25193607656373374, |
| "learning_rate": 1.3694068678459939e-05, |
| "loss": 0.3742, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.3883949461862424, |
| "grad_norm": 0.26856384975738834, |
| "learning_rate": 1.3590010405827266e-05, |
| "loss": 0.3893, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.3930744033692092, |
| "grad_norm": 0.2410724044047713, |
| "learning_rate": 1.348595213319459e-05, |
| "loss": 0.3881, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.39775386055217593, |
| "grad_norm": 0.22933262352367176, |
| "learning_rate": 1.3381893860561916e-05, |
| "loss": 0.389, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.40243331773514274, |
| "grad_norm": 0.2198762674864796, |
| "learning_rate": 1.3277835587929242e-05, |
| "loss": 0.3777, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.4071127749181095, |
| "grad_norm": 0.2419080712284111, |
| "learning_rate": 1.3173777315296567e-05, |
| "loss": 0.3675, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.4117922321010763, |
| "grad_norm": 0.22323453635381435, |
| "learning_rate": 1.3069719042663892e-05, |
| "loss": 0.3766, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.41647168928404305, |
| "grad_norm": 0.2439496427012552, |
| "learning_rate": 1.2965660770031219e-05, |
| "loss": 0.3737, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.42115114646700985, |
| "grad_norm": 0.21608448050711354, |
| "learning_rate": 1.2861602497398544e-05, |
| "loss": 0.3762, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.4258306036499766, |
| "grad_norm": 0.23222753887880532, |
| "learning_rate": 1.2757544224765869e-05, |
| "loss": 0.3806, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.43051006083294335, |
| "grad_norm": 0.23089602293769354, |
| "learning_rate": 1.2653485952133196e-05, |
| "loss": 0.3866, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.43518951801591016, |
| "grad_norm": 0.22647211350083776, |
| "learning_rate": 1.2549427679500521e-05, |
| "loss": 0.376, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.4398689751988769, |
| "grad_norm": 0.22578583335582927, |
| "learning_rate": 1.2445369406867846e-05, |
| "loss": 0.3755, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.4445484323818437, |
| "grad_norm": 0.25240119086847307, |
| "learning_rate": 1.2341311134235173e-05, |
| "loss": 0.3705, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.44922788956481047, |
| "grad_norm": 0.21696390759958575, |
| "learning_rate": 1.2237252861602498e-05, |
| "loss": 0.3791, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.4539073467477773, |
| "grad_norm": 0.2259330179979768, |
| "learning_rate": 1.2133194588969823e-05, |
| "loss": 0.3647, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.458586803930744, |
| "grad_norm": 0.23734350889969866, |
| "learning_rate": 1.202913631633715e-05, |
| "loss": 0.3867, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.46326626111371083, |
| "grad_norm": 0.29490446479765287, |
| "learning_rate": 1.1925078043704474e-05, |
| "loss": 0.3802, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.4679457182966776, |
| "grad_norm": 0.2539714715149494, |
| "learning_rate": 1.1821019771071803e-05, |
| "loss": 0.3769, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4726251754796444, |
| "grad_norm": 0.2653171806988273, |
| "learning_rate": 1.1716961498439128e-05, |
| "loss": 0.3774, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.47730463266261114, |
| "grad_norm": 0.22986118422933471, |
| "learning_rate": 1.1612903225806453e-05, |
| "loss": 0.3761, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.4819840898455779, |
| "grad_norm": 0.264468288478972, |
| "learning_rate": 1.150884495317378e-05, |
| "loss": 0.3768, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.4866635470285447, |
| "grad_norm": 0.2275931074116885, |
| "learning_rate": 1.1404786680541105e-05, |
| "loss": 0.3793, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.49134300421151145, |
| "grad_norm": 0.3117840889438511, |
| "learning_rate": 1.130072840790843e-05, |
| "loss": 0.3767, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.49602246139447825, |
| "grad_norm": 0.24416383968722374, |
| "learning_rate": 1.1196670135275756e-05, |
| "loss": 0.3767, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.5007019185774451, |
| "grad_norm": 0.2500367823625742, |
| "learning_rate": 1.1092611862643081e-05, |
| "loss": 0.3773, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.5053813757604118, |
| "grad_norm": 0.23404067505268092, |
| "learning_rate": 1.0988553590010406e-05, |
| "loss": 0.3789, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.5100608329433786, |
| "grad_norm": 0.23836407315387556, |
| "learning_rate": 1.0884495317377733e-05, |
| "loss": 0.3807, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.5147402901263454, |
| "grad_norm": 0.2272928154140146, |
| "learning_rate": 1.0780437044745058e-05, |
| "loss": 0.378, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.5194197473093122, |
| "grad_norm": 0.21603973655035985, |
| "learning_rate": 1.0676378772112383e-05, |
| "loss": 0.3849, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.5240992044922789, |
| "grad_norm": 0.26461820434483596, |
| "learning_rate": 1.057232049947971e-05, |
| "loss": 0.3807, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.5287786616752457, |
| "grad_norm": 0.22573025113669873, |
| "learning_rate": 1.0468262226847035e-05, |
| "loss": 0.3713, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.5334581188582125, |
| "grad_norm": 0.2398814988800669, |
| "learning_rate": 1.036420395421436e-05, |
| "loss": 0.3741, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.5381375760411792, |
| "grad_norm": 0.33008700162835736, |
| "learning_rate": 1.0260145681581687e-05, |
| "loss": 0.3784, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.542817033224146, |
| "grad_norm": 0.2478593026168216, |
| "learning_rate": 1.0156087408949012e-05, |
| "loss": 0.3797, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.5474964904071128, |
| "grad_norm": 0.22416882300252308, |
| "learning_rate": 1.0052029136316337e-05, |
| "loss": 0.3717, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.5521759475900796, |
| "grad_norm": 0.24550032569746283, |
| "learning_rate": 9.947970863683663e-06, |
| "loss": 0.3716, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.5568554047730463, |
| "grad_norm": 0.21152157772573982, |
| "learning_rate": 9.843912591050988e-06, |
| "loss": 0.3762, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.5615348619560131, |
| "grad_norm": 0.26437115475677, |
| "learning_rate": 9.739854318418315e-06, |
| "loss": 0.3649, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.5662143191389799, |
| "grad_norm": 0.24141737859739215, |
| "learning_rate": 9.63579604578564e-06, |
| "loss": 0.3721, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.5708937763219466, |
| "grad_norm": 0.2158535958967608, |
| "learning_rate": 9.531737773152965e-06, |
| "loss": 0.3736, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.5755732335049134, |
| "grad_norm": 0.25210607106723404, |
| "learning_rate": 9.427679500520292e-06, |
| "loss": 0.3779, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.5802526906878802, |
| "grad_norm": 0.24941061104468912, |
| "learning_rate": 9.323621227887619e-06, |
| "loss": 0.3724, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.584932147870847, |
| "grad_norm": 0.23950130390892227, |
| "learning_rate": 9.219562955254944e-06, |
| "loss": 0.3718, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.5896116050538137, |
| "grad_norm": 0.2413173383260915, |
| "learning_rate": 9.115504682622269e-06, |
| "loss": 0.3736, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.5942910622367805, |
| "grad_norm": 0.23891221147926558, |
| "learning_rate": 9.011446409989595e-06, |
| "loss": 0.3722, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.5989705194197473, |
| "grad_norm": 0.21323637844758167, |
| "learning_rate": 8.90738813735692e-06, |
| "loss": 0.3666, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.6036499766027141, |
| "grad_norm": 0.23920099409630857, |
| "learning_rate": 8.803329864724245e-06, |
| "loss": 0.3833, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.6083294337856808, |
| "grad_norm": 0.24543976835408943, |
| "learning_rate": 8.699271592091572e-06, |
| "loss": 0.3822, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.6130088909686476, |
| "grad_norm": 0.23517185442217967, |
| "learning_rate": 8.595213319458897e-06, |
| "loss": 0.3825, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.6176883481516144, |
| "grad_norm": 0.22209203218170853, |
| "learning_rate": 8.491155046826224e-06, |
| "loss": 0.3742, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.6223678053345811, |
| "grad_norm": 0.2118379611031678, |
| "learning_rate": 8.387096774193549e-06, |
| "loss": 0.3705, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.6270472625175479, |
| "grad_norm": 0.24116107131805098, |
| "learning_rate": 8.283038501560876e-06, |
| "loss": 0.3648, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.6317267197005147, |
| "grad_norm": 0.24385133695583516, |
| "learning_rate": 8.1789802289282e-06, |
| "loss": 0.3846, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.6364061768834816, |
| "grad_norm": 0.24780606354176135, |
| "learning_rate": 8.074921956295526e-06, |
| "loss": 0.3843, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.6410856340664483, |
| "grad_norm": 0.267287826257429, |
| "learning_rate": 7.970863683662852e-06, |
| "loss": 0.3697, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.6457650912494151, |
| "grad_norm": 0.2544017525039406, |
| "learning_rate": 7.866805411030177e-06, |
| "loss": 0.3709, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.6504445484323819, |
| "grad_norm": 0.2455819067026457, |
| "learning_rate": 7.762747138397502e-06, |
| "loss": 0.3701, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.6551240056153487, |
| "grad_norm": 0.259501849908747, |
| "learning_rate": 7.658688865764829e-06, |
| "loss": 0.3772, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.6598034627983154, |
| "grad_norm": 0.20856381315062708, |
| "learning_rate": 7.554630593132155e-06, |
| "loss": 0.3707, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.6644829199812822, |
| "grad_norm": 0.2332621123767882, |
| "learning_rate": 7.450572320499481e-06, |
| "loss": 0.3894, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.669162377164249, |
| "grad_norm": 0.20954093237017377, |
| "learning_rate": 7.346514047866807e-06, |
| "loss": 0.374, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.6738418343472157, |
| "grad_norm": 0.2010019285730511, |
| "learning_rate": 7.242455775234132e-06, |
| "loss": 0.3838, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.6785212915301825, |
| "grad_norm": 0.21248927092428815, |
| "learning_rate": 7.1383975026014575e-06, |
| "loss": 0.3612, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.6832007487131493, |
| "grad_norm": 0.1994556422452643, |
| "learning_rate": 7.0343392299687825e-06, |
| "loss": 0.3724, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.6878802058961161, |
| "grad_norm": 0.22422751080621398, |
| "learning_rate": 6.930280957336108e-06, |
| "loss": 0.3762, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.6925596630790828, |
| "grad_norm": 0.21331923051138799, |
| "learning_rate": 6.826222684703434e-06, |
| "loss": 0.368, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.6972391202620496, |
| "grad_norm": 0.2103235038933413, |
| "learning_rate": 6.722164412070759e-06, |
| "loss": 0.3703, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.7019185774450164, |
| "grad_norm": 0.2156780953428616, |
| "learning_rate": 6.618106139438086e-06, |
| "loss": 0.3786, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.7065980346279832, |
| "grad_norm": 0.21490072085384168, |
| "learning_rate": 6.514047866805412e-06, |
| "loss": 0.3758, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.7112774918109499, |
| "grad_norm": 0.2247782023019425, |
| "learning_rate": 6.409989594172738e-06, |
| "loss": 0.3679, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.7159569489939167, |
| "grad_norm": 0.20942665269809468, |
| "learning_rate": 6.305931321540063e-06, |
| "loss": 0.3652, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.7206364061768835, |
| "grad_norm": 0.22300307299120317, |
| "learning_rate": 6.201873048907389e-06, |
| "loss": 0.3867, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.7253158633598502, |
| "grad_norm": 0.21731092424159765, |
| "learning_rate": 6.0978147762747145e-06, |
| "loss": 0.3699, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.729995320542817, |
| "grad_norm": 0.2373583086701352, |
| "learning_rate": 5.9937565036420395e-06, |
| "loss": 0.3653, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.7346747777257838, |
| "grad_norm": 0.2066335501990478, |
| "learning_rate": 5.889698231009365e-06, |
| "loss": 0.3741, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.7393542349087506, |
| "grad_norm": 0.20910016326262895, |
| "learning_rate": 5.785639958376691e-06, |
| "loss": 0.3724, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.7440336920917173, |
| "grad_norm": 0.2235231360878696, |
| "learning_rate": 5.681581685744018e-06, |
| "loss": 0.3727, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.7487131492746841, |
| "grad_norm": 0.2046167285793584, |
| "learning_rate": 5.577523413111343e-06, |
| "loss": 0.3684, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.7533926064576509, |
| "grad_norm": 0.19515222922990866, |
| "learning_rate": 5.473465140478669e-06, |
| "loss": 0.3787, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.7580720636406177, |
| "grad_norm": 0.22525978812547012, |
| "learning_rate": 5.369406867845994e-06, |
| "loss": 0.3764, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.7627515208235844, |
| "grad_norm": 0.22394936960559922, |
| "learning_rate": 5.26534859521332e-06, |
| "loss": 0.3672, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.7674309780065512, |
| "grad_norm": 0.23343895861163633, |
| "learning_rate": 5.161290322580646e-06, |
| "loss": 0.3676, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.772110435189518, |
| "grad_norm": 0.24463148457843642, |
| "learning_rate": 5.057232049947971e-06, |
| "loss": 0.3728, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.7767898923724847, |
| "grad_norm": 0.2187630043298853, |
| "learning_rate": 4.953173777315297e-06, |
| "loss": 0.3712, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.7814693495554516, |
| "grad_norm": 0.21300481471513596, |
| "learning_rate": 4.849115504682622e-06, |
| "loss": 0.381, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.7861488067384184, |
| "grad_norm": 0.20931843918006063, |
| "learning_rate": 4.745057232049948e-06, |
| "loss": 0.3747, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.7908282639213852, |
| "grad_norm": 0.21125591014093664, |
| "learning_rate": 4.640998959417274e-06, |
| "loss": 0.3735, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.7955077211043519, |
| "grad_norm": 0.2110487123545379, |
| "learning_rate": 4.5369406867846e-06, |
| "loss": 0.3735, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.8001871782873187, |
| "grad_norm": 0.22392695446168392, |
| "learning_rate": 4.432882414151926e-06, |
| "loss": 0.3777, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.8048666354702855, |
| "grad_norm": 0.2117452208603545, |
| "learning_rate": 4.328824141519251e-06, |
| "loss": 0.3761, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.8095460926532522, |
| "grad_norm": 0.24234752709613971, |
| "learning_rate": 4.224765868886577e-06, |
| "loss": 0.3796, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.814225549836219, |
| "grad_norm": 0.2025209483621438, |
| "learning_rate": 4.120707596253903e-06, |
| "loss": 0.3808, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.8189050070191858, |
| "grad_norm": 0.2226709019764019, |
| "learning_rate": 4.0166493236212285e-06, |
| "loss": 0.3752, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.8235844642021526, |
| "grad_norm": 0.2210571915933747, |
| "learning_rate": 3.912591050988554e-06, |
| "loss": 0.3704, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.8282639213851193, |
| "grad_norm": 0.23618799628521495, |
| "learning_rate": 3.8085327783558793e-06, |
| "loss": 0.3737, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.8329433785680861, |
| "grad_norm": 0.20617597677407185, |
| "learning_rate": 3.7044745057232052e-06, |
| "loss": 0.3749, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.8376228357510529, |
| "grad_norm": 0.20798543863033134, |
| "learning_rate": 3.600416233090531e-06, |
| "loss": 0.3687, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.8423022929340197, |
| "grad_norm": 0.2089006102637797, |
| "learning_rate": 3.496357960457857e-06, |
| "loss": 0.3706, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.8469817501169864, |
| "grad_norm": 0.20477692589921484, |
| "learning_rate": 3.3922996878251824e-06, |
| "loss": 0.3702, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.8516612072999532, |
| "grad_norm": 0.21774749397741822, |
| "learning_rate": 3.288241415192508e-06, |
| "loss": 0.3657, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.85634066448292, |
| "grad_norm": 0.24858244914099062, |
| "learning_rate": 3.1841831425598337e-06, |
| "loss": 0.3681, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.8610201216658867, |
| "grad_norm": 0.22753648396835552, |
| "learning_rate": 3.0801248699271596e-06, |
| "loss": 0.3703, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.8656995788488535, |
| "grad_norm": 0.2585954998067118, |
| "learning_rate": 2.976066597294485e-06, |
| "loss": 0.3748, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.8703790360318203, |
| "grad_norm": 0.23712893924962872, |
| "learning_rate": 2.872008324661811e-06, |
| "loss": 0.3724, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.8750584932147871, |
| "grad_norm": 0.21078832525011484, |
| "learning_rate": 2.7679500520291363e-06, |
| "loss": 0.3722, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.8797379503977538, |
| "grad_norm": 0.2005875423082211, |
| "learning_rate": 2.6638917793964626e-06, |
| "loss": 0.364, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.8844174075807206, |
| "grad_norm": 0.21183567448111165, |
| "learning_rate": 2.559833506763788e-06, |
| "loss": 0.3646, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.8890968647636874, |
| "grad_norm": 0.23633491944791057, |
| "learning_rate": 2.4557752341311135e-06, |
| "loss": 0.369, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.8937763219466542, |
| "grad_norm": 0.20446185463761749, |
| "learning_rate": 2.3517169614984394e-06, |
| "loss": 0.3717, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.8984557791296209, |
| "grad_norm": 0.21120878547944943, |
| "learning_rate": 2.2476586888657652e-06, |
| "loss": 0.3691, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.9031352363125877, |
| "grad_norm": 0.21926451373870526, |
| "learning_rate": 2.1436004162330907e-06, |
| "loss": 0.376, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.9078146934955545, |
| "grad_norm": 0.21234936601889798, |
| "learning_rate": 2.0395421436004166e-06, |
| "loss": 0.3721, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.9124941506785212, |
| "grad_norm": 0.22032807507001712, |
| "learning_rate": 1.935483870967742e-06, |
| "loss": 0.3796, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.917173607861488, |
| "grad_norm": 0.20048012242437066, |
| "learning_rate": 1.8314255983350679e-06, |
| "loss": 0.3692, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.9218530650444549, |
| "grad_norm": 0.19197425552585032, |
| "learning_rate": 1.7273673257023935e-06, |
| "loss": 0.3602, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.9265325222274217, |
| "grad_norm": 0.19963248316970567, |
| "learning_rate": 1.6233090530697194e-06, |
| "loss": 0.3687, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.9312119794103884, |
| "grad_norm": 0.20252903061869637, |
| "learning_rate": 1.5192507804370448e-06, |
| "loss": 0.3693, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.9358914365933552, |
| "grad_norm": 0.22168638474832342, |
| "learning_rate": 1.4151925078043705e-06, |
| "loss": 0.3775, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.940570893776322, |
| "grad_norm": 0.21135045250093049, |
| "learning_rate": 1.3111342351716964e-06, |
| "loss": 0.3707, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.9452503509592888, |
| "grad_norm": 0.20991603381075577, |
| "learning_rate": 1.207075962539022e-06, |
| "loss": 0.3801, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.9499298081422555, |
| "grad_norm": 0.19588787717171513, |
| "learning_rate": 1.1030176899063477e-06, |
| "loss": 0.3707, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.9546092653252223, |
| "grad_norm": 0.18850895260293038, |
| "learning_rate": 9.989594172736733e-07, |
| "loss": 0.3603, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.9592887225081891, |
| "grad_norm": 0.1924405623942021, |
| "learning_rate": 8.94901144640999e-07, |
| "loss": 0.3663, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.9639681796911558, |
| "grad_norm": 0.1934047488919004, |
| "learning_rate": 7.908428720083247e-07, |
| "loss": 0.3731, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.9686476368741226, |
| "grad_norm": 0.2027478493287533, |
| "learning_rate": 6.867845993756504e-07, |
| "loss": 0.3682, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.9733270940570894, |
| "grad_norm": 0.19171827074306003, |
| "learning_rate": 5.827263267429762e-07, |
| "loss": 0.3643, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.9780065512400562, |
| "grad_norm": 0.20102243551100357, |
| "learning_rate": 4.786680541103018e-07, |
| "loss": 0.3761, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.9826860084230229, |
| "grad_norm": 0.19333944542660678, |
| "learning_rate": 3.7460978147762747e-07, |
| "loss": 0.3782, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.9873654656059897, |
| "grad_norm": 0.2031908820930634, |
| "learning_rate": 2.705515088449532e-07, |
| "loss": 0.3766, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.9920449227889565, |
| "grad_norm": 0.1887326527462366, |
| "learning_rate": 1.6649323621227891e-07, |
| "loss": 0.376, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.9967243799719232, |
| "grad_norm": 0.195152345665213, |
| "learning_rate": 6.243496357960458e-08, |
| "loss": 0.3606, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.9995320542817033, |
| "step": 1068, |
| "total_flos": 9.145616827248804e+17, |
| "train_loss": 0.3807801652267185, |
| "train_runtime": 28315.9454, |
| "train_samples_per_second": 0.604, |
| "train_steps_per_second": 0.038 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 1068, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "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": 9.145616827248804e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
