| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998222222222222, | |
| "eval_steps": 5000, | |
| "global_step": 2812, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0035555555555555557, | |
| "grad_norm": 0.4803234934806824, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.5376, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0071111111111111115, | |
| "grad_norm": 0.4609735310077667, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.3932, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010666666666666666, | |
| "grad_norm": 0.4800471365451813, | |
| "learning_rate": 6e-06, | |
| "loss": 1.6084, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.014222222222222223, | |
| "grad_norm": 0.513812780380249, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.5579, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 0.5741678476333618, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5945, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.021333333333333333, | |
| "grad_norm": 0.42774268984794617, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.6154, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.024888888888888887, | |
| "grad_norm": 0.37852978706359863, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.5581, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.028444444444444446, | |
| "grad_norm": 0.48577743768692017, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.57, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 0.5572775602340698, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.6154, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 0.4657267928123474, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5583, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03911111111111111, | |
| "grad_norm": 0.48570355772972107, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.6356, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.042666666666666665, | |
| "grad_norm": 0.5374159812927246, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.5911, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04622222222222222, | |
| "grad_norm": 0.4544757306575775, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.5462, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.049777777777777775, | |
| "grad_norm": 0.49096816778182983, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.4696, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 0.45160388946533203, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4397, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05688888888888889, | |
| "grad_norm": 0.4654102325439453, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.61, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.060444444444444446, | |
| "grad_norm": 0.5608298182487488, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.5799, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 0.5087274312973022, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.6534, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06755555555555555, | |
| "grad_norm": 0.5886797308921814, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.5561, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 0.5276679396629333, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5705, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07466666666666667, | |
| "grad_norm": 0.4331255257129669, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.554, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07822222222222222, | |
| "grad_norm": 0.5068897604942322, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.451, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.08177777777777778, | |
| "grad_norm": 0.48896610736846924, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.6522, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.08533333333333333, | |
| "grad_norm": 0.5350568890571594, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.5571, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 0.6370730400085449, | |
| "learning_rate": 5e-05, | |
| "loss": 1.6112, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09244444444444444, | |
| "grad_norm": 0.5409080386161804, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.5878, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 0.5505415797233582, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.6938, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.09955555555555555, | |
| "grad_norm": 0.557578980922699, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.6776, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.10311111111111111, | |
| "grad_norm": 0.5505691170692444, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.7429, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 0.5328088998794556, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6633, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.11022222222222222, | |
| "grad_norm": 0.4911518394947052, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.4746, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.11377777777777778, | |
| "grad_norm": 0.5215829610824585, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.4723, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.11733333333333333, | |
| "grad_norm": 0.5551868081092834, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.5322, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.12088888888888889, | |
| "grad_norm": 0.607334554195404, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.7652, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 0.4849174916744232, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5915, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.5409530997276306, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.6263, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.13155555555555556, | |
| "grad_norm": 0.5955730080604553, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.598, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1351111111111111, | |
| "grad_norm": 0.5792020559310913, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.7498, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.13866666666666666, | |
| "grad_norm": 0.6078203916549683, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.5885, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 0.470548540353775, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4596, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14577777777777778, | |
| "grad_norm": 0.600089967250824, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.6343, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.14933333333333335, | |
| "grad_norm": 0.5364957451820374, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.4748, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.15288888888888888, | |
| "grad_norm": 0.5835921168327332, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.6809, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.15644444444444444, | |
| "grad_norm": 0.5760766863822937, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.6276, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.5958523154258728, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4696, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16355555555555557, | |
| "grad_norm": 0.5406598448753357, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.625, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.1671111111111111, | |
| "grad_norm": 0.5989697575569153, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.6854, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.17066666666666666, | |
| "grad_norm": 0.594978392124176, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.5774, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.17422222222222222, | |
| "grad_norm": 0.5095472931861877, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.7095, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.7304088473320007, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6842, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18133333333333335, | |
| "grad_norm": 0.6202882528305054, | |
| "learning_rate": 9.95674740484429e-05, | |
| "loss": 1.696, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.18488888888888888, | |
| "grad_norm": 0.49087047576904297, | |
| "learning_rate": 9.913494809688582e-05, | |
| "loss": 1.6507, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.18844444444444444, | |
| "grad_norm": 0.6279426217079163, | |
| "learning_rate": 9.870242214532872e-05, | |
| "loss": 1.6809, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.602878987789154, | |
| "learning_rate": 9.826989619377162e-05, | |
| "loss": 1.6173, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 0.6200081706047058, | |
| "learning_rate": 9.783737024221454e-05, | |
| "loss": 1.6787, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1991111111111111, | |
| "grad_norm": 0.6582353115081787, | |
| "learning_rate": 9.740484429065745e-05, | |
| "loss": 1.7095, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.20266666666666666, | |
| "grad_norm": 0.541404664516449, | |
| "learning_rate": 9.697231833910035e-05, | |
| "loss": 1.8418, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.20622222222222222, | |
| "grad_norm": 0.5303178429603577, | |
| "learning_rate": 9.653979238754325e-05, | |
| "loss": 1.6112, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.20977777777777779, | |
| "grad_norm": 0.527113676071167, | |
| "learning_rate": 9.610726643598617e-05, | |
| "loss": 1.6454, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 0.5398814678192139, | |
| "learning_rate": 9.567474048442907e-05, | |
| "loss": 1.5353, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.21688888888888888, | |
| "grad_norm": 0.5357266068458557, | |
| "learning_rate": 9.524221453287197e-05, | |
| "loss": 1.6816, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.22044444444444444, | |
| "grad_norm": 0.4769757390022278, | |
| "learning_rate": 9.480968858131488e-05, | |
| "loss": 1.6625, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 0.5274131298065186, | |
| "learning_rate": 9.437716262975779e-05, | |
| "loss": 1.7129, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.22755555555555557, | |
| "grad_norm": 0.6454447507858276, | |
| "learning_rate": 9.394463667820069e-05, | |
| "loss": 1.5577, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 0.5967287421226501, | |
| "learning_rate": 9.35121107266436e-05, | |
| "loss": 1.3905, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.23466666666666666, | |
| "grad_norm": 0.57136470079422, | |
| "learning_rate": 9.307958477508652e-05, | |
| "loss": 1.6029, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.23822222222222222, | |
| "grad_norm": 0.5300968885421753, | |
| "learning_rate": 9.264705882352942e-05, | |
| "loss": 1.7363, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.24177777777777779, | |
| "grad_norm": 0.44372138381004333, | |
| "learning_rate": 9.221453287197233e-05, | |
| "loss": 1.6467, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.24533333333333332, | |
| "grad_norm": 0.45939210057258606, | |
| "learning_rate": 9.178200692041523e-05, | |
| "loss": 1.5413, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 0.5719866752624512, | |
| "learning_rate": 9.134948096885813e-05, | |
| "loss": 1.5702, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.25244444444444447, | |
| "grad_norm": 0.4535101652145386, | |
| "learning_rate": 9.091695501730105e-05, | |
| "loss": 1.5331, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.5633770227432251, | |
| "learning_rate": 9.048442906574395e-05, | |
| "loss": 1.696, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.25955555555555554, | |
| "grad_norm": 0.6633341908454895, | |
| "learning_rate": 9.005190311418685e-05, | |
| "loss": 1.5428, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.26311111111111113, | |
| "grad_norm": 0.6005223393440247, | |
| "learning_rate": 8.961937716262977e-05, | |
| "loss": 1.8278, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.5239768624305725, | |
| "learning_rate": 8.918685121107267e-05, | |
| "loss": 1.6761, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2702222222222222, | |
| "grad_norm": 0.4788583517074585, | |
| "learning_rate": 8.875432525951558e-05, | |
| "loss": 1.548, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.2737777777777778, | |
| "grad_norm": 0.5428124666213989, | |
| "learning_rate": 8.832179930795848e-05, | |
| "loss": 1.6972, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.2773333333333333, | |
| "grad_norm": 0.42001649737358093, | |
| "learning_rate": 8.78892733564014e-05, | |
| "loss": 1.6244, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2808888888888889, | |
| "grad_norm": 0.5100952386856079, | |
| "learning_rate": 8.74567474048443e-05, | |
| "loss": 1.6704, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 0.49174782633781433, | |
| "learning_rate": 8.70242214532872e-05, | |
| "loss": 1.6876, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.543880820274353, | |
| "learning_rate": 8.659169550173011e-05, | |
| "loss": 1.599, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.29155555555555557, | |
| "grad_norm": 0.4966542422771454, | |
| "learning_rate": 8.615916955017301e-05, | |
| "loss": 1.4723, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.2951111111111111, | |
| "grad_norm": 0.44916948676109314, | |
| "learning_rate": 8.572664359861592e-05, | |
| "loss": 1.452, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.2986666666666667, | |
| "grad_norm": 0.5791043639183044, | |
| "learning_rate": 8.529411764705883e-05, | |
| "loss": 1.7569, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 0.5880780816078186, | |
| "learning_rate": 8.486159169550173e-05, | |
| "loss": 1.4779, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.30577777777777776, | |
| "grad_norm": 0.5578981041908264, | |
| "learning_rate": 8.442906574394463e-05, | |
| "loss": 1.6577, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.30933333333333335, | |
| "grad_norm": 0.6132334470748901, | |
| "learning_rate": 8.399653979238755e-05, | |
| "loss": 1.5862, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3128888888888889, | |
| "grad_norm": 0.512484073638916, | |
| "learning_rate": 8.356401384083046e-05, | |
| "loss": 1.4891, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.3164444444444444, | |
| "grad_norm": 0.4793022871017456, | |
| "learning_rate": 8.313148788927336e-05, | |
| "loss": 1.601, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.5512364506721497, | |
| "learning_rate": 8.269896193771626e-05, | |
| "loss": 1.6191, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.32355555555555554, | |
| "grad_norm": 0.5528599619865417, | |
| "learning_rate": 8.226643598615918e-05, | |
| "loss": 1.6555, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.32711111111111113, | |
| "grad_norm": 0.4718135595321655, | |
| "learning_rate": 8.183391003460208e-05, | |
| "loss": 1.3796, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.33066666666666666, | |
| "grad_norm": 0.5744655728340149, | |
| "learning_rate": 8.140138408304498e-05, | |
| "loss": 1.7323, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.3342222222222222, | |
| "grad_norm": 0.5298731923103333, | |
| "learning_rate": 8.09688581314879e-05, | |
| "loss": 1.6613, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 0.507189154624939, | |
| "learning_rate": 8.05363321799308e-05, | |
| "loss": 1.4932, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.3413333333333333, | |
| "grad_norm": 0.5233726501464844, | |
| "learning_rate": 8.01038062283737e-05, | |
| "loss": 1.5989, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.3448888888888889, | |
| "grad_norm": 0.5754904747009277, | |
| "learning_rate": 7.967128027681662e-05, | |
| "loss": 1.6657, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.34844444444444445, | |
| "grad_norm": 0.6196070909500122, | |
| "learning_rate": 7.923875432525953e-05, | |
| "loss": 1.6808, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.461910605430603, | |
| "learning_rate": 7.880622837370243e-05, | |
| "loss": 1.6012, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.4632907807826996, | |
| "learning_rate": 7.837370242214533e-05, | |
| "loss": 1.5879, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3591111111111111, | |
| "grad_norm": 0.48907166719436646, | |
| "learning_rate": 7.794117647058824e-05, | |
| "loss": 1.6109, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.3626666666666667, | |
| "grad_norm": 0.5417441129684448, | |
| "learning_rate": 7.750865051903114e-05, | |
| "loss": 1.5687, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.3662222222222222, | |
| "grad_norm": 0.519092857837677, | |
| "learning_rate": 7.707612456747404e-05, | |
| "loss": 1.6197, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.36977777777777776, | |
| "grad_norm": 0.45026785135269165, | |
| "learning_rate": 7.664359861591696e-05, | |
| "loss": 1.5194, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 0.5378026366233826, | |
| "learning_rate": 7.621107266435986e-05, | |
| "loss": 1.419, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3768888888888889, | |
| "grad_norm": 0.4889166057109833, | |
| "learning_rate": 7.577854671280276e-05, | |
| "loss": 1.6356, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.3804444444444444, | |
| "grad_norm": 0.5200234055519104, | |
| "learning_rate": 7.534602076124569e-05, | |
| "loss": 1.7118, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.4398309886455536, | |
| "learning_rate": 7.491349480968859e-05, | |
| "loss": 1.525, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.38755555555555554, | |
| "grad_norm": 0.43217843770980835, | |
| "learning_rate": 7.448096885813149e-05, | |
| "loss": 1.5643, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 0.5189200043678284, | |
| "learning_rate": 7.40484429065744e-05, | |
| "loss": 1.6891, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.39466666666666667, | |
| "grad_norm": 0.5237637758255005, | |
| "learning_rate": 7.36159169550173e-05, | |
| "loss": 1.4445, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.3982222222222222, | |
| "grad_norm": 0.5614046454429626, | |
| "learning_rate": 7.318339100346021e-05, | |
| "loss": 1.6439, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.4017777777777778, | |
| "grad_norm": 0.4191811680793762, | |
| "learning_rate": 7.275086505190312e-05, | |
| "loss": 1.4344, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.4053333333333333, | |
| "grad_norm": 0.5043474435806274, | |
| "learning_rate": 7.231833910034602e-05, | |
| "loss": 1.7017, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 0.40949752926826477, | |
| "learning_rate": 7.188581314878892e-05, | |
| "loss": 1.5903, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.41244444444444445, | |
| "grad_norm": 0.4599238634109497, | |
| "learning_rate": 7.145328719723184e-05, | |
| "loss": 1.5666, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.5736513137817383, | |
| "learning_rate": 7.102076124567474e-05, | |
| "loss": 1.5525, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.41955555555555557, | |
| "grad_norm": 0.5437818169593811, | |
| "learning_rate": 7.058823529411765e-05, | |
| "loss": 1.6668, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.4231111111111111, | |
| "grad_norm": 0.4318547248840332, | |
| "learning_rate": 7.015570934256056e-05, | |
| "loss": 1.4075, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.44979867339134216, | |
| "learning_rate": 6.972318339100347e-05, | |
| "loss": 1.3973, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.43022222222222223, | |
| "grad_norm": 0.4653863310813904, | |
| "learning_rate": 6.929065743944637e-05, | |
| "loss": 1.4578, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.43377777777777776, | |
| "grad_norm": 0.5257763266563416, | |
| "learning_rate": 6.885813148788927e-05, | |
| "loss": 1.5667, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.43733333333333335, | |
| "grad_norm": 0.5117728114128113, | |
| "learning_rate": 6.842560553633219e-05, | |
| "loss": 1.6843, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.4408888888888889, | |
| "grad_norm": 0.4611428678035736, | |
| "learning_rate": 6.799307958477509e-05, | |
| "loss": 1.7192, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.586694061756134, | |
| "learning_rate": 6.756055363321799e-05, | |
| "loss": 1.6057, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.43433499336242676, | |
| "learning_rate": 6.71280276816609e-05, | |
| "loss": 1.6829, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.45155555555555554, | |
| "grad_norm": 0.5851064920425415, | |
| "learning_rate": 6.66955017301038e-05, | |
| "loss": 1.6039, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.45511111111111113, | |
| "grad_norm": 0.5549812912940979, | |
| "learning_rate": 6.626297577854672e-05, | |
| "loss": 1.6076, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.45866666666666667, | |
| "grad_norm": 0.4489051103591919, | |
| "learning_rate": 6.583044982698962e-05, | |
| "loss": 1.6347, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": 0.5636203289031982, | |
| "learning_rate": 6.539792387543253e-05, | |
| "loss": 1.3819, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.4657777777777778, | |
| "grad_norm": 0.5598219037055969, | |
| "learning_rate": 6.496539792387544e-05, | |
| "loss": 1.6914, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.4693333333333333, | |
| "grad_norm": 0.44435766339302063, | |
| "learning_rate": 6.453287197231834e-05, | |
| "loss": 1.5487, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.4728888888888889, | |
| "grad_norm": 0.4664863646030426, | |
| "learning_rate": 6.410034602076125e-05, | |
| "loss": 1.5156, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.47644444444444445, | |
| "grad_norm": 0.5192257761955261, | |
| "learning_rate": 6.366782006920415e-05, | |
| "loss": 1.5507, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6014835238456726, | |
| "learning_rate": 6.323529411764705e-05, | |
| "loss": 1.4809, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.48355555555555557, | |
| "grad_norm": 0.5586997270584106, | |
| "learning_rate": 6.280276816608997e-05, | |
| "loss": 1.5619, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.4871111111111111, | |
| "grad_norm": 0.4059390723705292, | |
| "learning_rate": 6.237024221453287e-05, | |
| "loss": 1.4799, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.49066666666666664, | |
| "grad_norm": 0.41877785325050354, | |
| "learning_rate": 6.193771626297578e-05, | |
| "loss": 1.4187, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.49422222222222223, | |
| "grad_norm": 0.4456830620765686, | |
| "learning_rate": 6.15051903114187e-05, | |
| "loss": 1.5507, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 0.45245644450187683, | |
| "learning_rate": 6.10726643598616e-05, | |
| "loss": 1.5478, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5013333333333333, | |
| "grad_norm": 0.4327283501625061, | |
| "learning_rate": 6.06401384083045e-05, | |
| "loss": 1.5734, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.5048888888888889, | |
| "grad_norm": 0.6047600507736206, | |
| "learning_rate": 6.020761245674741e-05, | |
| "loss": 1.6916, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.5084444444444445, | |
| "grad_norm": 0.4880015254020691, | |
| "learning_rate": 5.9775086505190316e-05, | |
| "loss": 1.6346, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.4689754843711853, | |
| "learning_rate": 5.934256055363322e-05, | |
| "loss": 1.5714, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 0.4588066041469574, | |
| "learning_rate": 5.8910034602076125e-05, | |
| "loss": 1.5547, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.5191111111111111, | |
| "grad_norm": 0.45055222511291504, | |
| "learning_rate": 5.847750865051903e-05, | |
| "loss": 1.4591, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.5226666666666666, | |
| "grad_norm": 0.49660903215408325, | |
| "learning_rate": 5.8044982698961933e-05, | |
| "loss": 1.7409, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.5262222222222223, | |
| "grad_norm": 0.48975858092308044, | |
| "learning_rate": 5.761245674740484e-05, | |
| "loss": 1.5158, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.5297777777777778, | |
| "grad_norm": 0.4958665668964386, | |
| "learning_rate": 5.7179930795847756e-05, | |
| "loss": 1.4612, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.4389738142490387, | |
| "learning_rate": 5.6747404844290664e-05, | |
| "loss": 1.6067, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.5368888888888889, | |
| "grad_norm": 0.44060131907463074, | |
| "learning_rate": 5.631487889273357e-05, | |
| "loss": 1.3391, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.5404444444444444, | |
| "grad_norm": 0.5496039390563965, | |
| "learning_rate": 5.588235294117647e-05, | |
| "loss": 1.5108, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.5372530221939087, | |
| "learning_rate": 5.544982698961938e-05, | |
| "loss": 1.4777, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.5475555555555556, | |
| "grad_norm": 0.5017176270484924, | |
| "learning_rate": 5.501730103806229e-05, | |
| "loss": 1.5007, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 0.5397398471832275, | |
| "learning_rate": 5.458477508650519e-05, | |
| "loss": 1.6057, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.5546666666666666, | |
| "grad_norm": 0.5731176733970642, | |
| "learning_rate": 5.41522491349481e-05, | |
| "loss": 1.6644, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.5582222222222222, | |
| "grad_norm": 0.48925384879112244, | |
| "learning_rate": 5.3719723183391005e-05, | |
| "loss": 1.5506, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.5617777777777778, | |
| "grad_norm": 0.47176703810691833, | |
| "learning_rate": 5.3287197231833906e-05, | |
| "loss": 1.5355, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.5653333333333334, | |
| "grad_norm": 0.45709192752838135, | |
| "learning_rate": 5.285467128027683e-05, | |
| "loss": 1.4926, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 0.5231002569198608, | |
| "learning_rate": 5.242214532871973e-05, | |
| "loss": 1.5645, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.5724444444444444, | |
| "grad_norm": 0.4811376929283142, | |
| "learning_rate": 5.1989619377162636e-05, | |
| "loss": 1.5169, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.5673519968986511, | |
| "learning_rate": 5.1557093425605544e-05, | |
| "loss": 1.4974, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.5795555555555556, | |
| "grad_norm": 0.5168330669403076, | |
| "learning_rate": 5.1124567474048445e-05, | |
| "loss": 1.4242, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.5831111111111111, | |
| "grad_norm": 0.49934643507003784, | |
| "learning_rate": 5.069204152249135e-05, | |
| "loss": 1.6342, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 0.44559621810913086, | |
| "learning_rate": 5.0259515570934254e-05, | |
| "loss": 1.5204, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.5902222222222222, | |
| "grad_norm": 0.5243204832077026, | |
| "learning_rate": 4.982698961937716e-05, | |
| "loss": 1.4703, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.5937777777777777, | |
| "grad_norm": 0.5117699503898621, | |
| "learning_rate": 4.9394463667820076e-05, | |
| "loss": 1.7011, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.5973333333333334, | |
| "grad_norm": 0.42527756094932556, | |
| "learning_rate": 4.896193771626298e-05, | |
| "loss": 1.4331, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.6008888888888889, | |
| "grad_norm": 0.5092193484306335, | |
| "learning_rate": 4.8529411764705885e-05, | |
| "loss": 1.4378, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 0.5100553631782532, | |
| "learning_rate": 4.809688581314879e-05, | |
| "loss": 1.5999, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.5817019939422607, | |
| "learning_rate": 4.7664359861591694e-05, | |
| "loss": 1.6094, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.6115555555555555, | |
| "grad_norm": 0.41723236441612244, | |
| "learning_rate": 4.723183391003461e-05, | |
| "loss": 1.3555, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.6151111111111112, | |
| "grad_norm": 0.5209750533103943, | |
| "learning_rate": 4.679930795847751e-05, | |
| "loss": 1.56, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.6186666666666667, | |
| "grad_norm": 0.48648136854171753, | |
| "learning_rate": 4.636678200692042e-05, | |
| "loss": 1.48, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.5282444357872009, | |
| "learning_rate": 4.5934256055363325e-05, | |
| "loss": 1.5003, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.6257777777777778, | |
| "grad_norm": 0.4915744364261627, | |
| "learning_rate": 4.5501730103806226e-05, | |
| "loss": 1.4575, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.6293333333333333, | |
| "grad_norm": 0.568629801273346, | |
| "learning_rate": 4.506920415224914e-05, | |
| "loss": 1.4804, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.6328888888888888, | |
| "grad_norm": 0.48679119348526, | |
| "learning_rate": 4.463667820069204e-05, | |
| "loss": 1.5015, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.6364444444444445, | |
| "grad_norm": 0.5196161866188049, | |
| "learning_rate": 4.420415224913495e-05, | |
| "loss": 1.561, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.4946150779724121, | |
| "learning_rate": 4.377162629757786e-05, | |
| "loss": 1.5151, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.6435555555555555, | |
| "grad_norm": 0.40968433022499084, | |
| "learning_rate": 4.333910034602076e-05, | |
| "loss": 1.5373, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.6471111111111111, | |
| "grad_norm": 0.6429114937782288, | |
| "learning_rate": 4.290657439446367e-05, | |
| "loss": 1.4867, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.6506666666666666, | |
| "grad_norm": 0.474758118391037, | |
| "learning_rate": 4.247404844290658e-05, | |
| "loss": 1.5035, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.6542222222222223, | |
| "grad_norm": 0.51218581199646, | |
| "learning_rate": 4.204152249134948e-05, | |
| "loss": 1.5725, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 0.49607232213020325, | |
| "learning_rate": 4.160899653979239e-05, | |
| "loss": 1.5352, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.6613333333333333, | |
| "grad_norm": 0.4859633147716522, | |
| "learning_rate": 4.11764705882353e-05, | |
| "loss": 1.4925, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.6648888888888889, | |
| "grad_norm": 0.4592611491680145, | |
| "learning_rate": 4.0743944636678205e-05, | |
| "loss": 1.5156, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.6684444444444444, | |
| "grad_norm": 0.47916117310523987, | |
| "learning_rate": 4.031141868512111e-05, | |
| "loss": 1.545, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.5305891633033752, | |
| "learning_rate": 3.9878892733564014e-05, | |
| "loss": 1.5384, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 0.4583602547645569, | |
| "learning_rate": 3.944636678200692e-05, | |
| "loss": 1.5635, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.6791111111111111, | |
| "grad_norm": 0.5024811625480652, | |
| "learning_rate": 3.901384083044983e-05, | |
| "loss": 1.5016, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.6826666666666666, | |
| "grad_norm": 0.5373083353042603, | |
| "learning_rate": 3.858131487889274e-05, | |
| "loss": 1.5399, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.6862222222222222, | |
| "grad_norm": 0.415873646736145, | |
| "learning_rate": 3.8148788927335645e-05, | |
| "loss": 1.6384, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.6897777777777778, | |
| "grad_norm": 0.46044766902923584, | |
| "learning_rate": 3.7716262975778546e-05, | |
| "loss": 1.5588, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 0.37672215700149536, | |
| "learning_rate": 3.7283737024221454e-05, | |
| "loss": 1.49, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.6968888888888889, | |
| "grad_norm": 0.40345415472984314, | |
| "learning_rate": 3.685121107266436e-05, | |
| "loss": 1.3789, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.7004444444444444, | |
| "grad_norm": 0.49566739797592163, | |
| "learning_rate": 3.641868512110726e-05, | |
| "loss": 1.3318, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.48587340116500854, | |
| "learning_rate": 3.598615916955018e-05, | |
| "loss": 1.4758, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.7075555555555556, | |
| "grad_norm": 0.596446692943573, | |
| "learning_rate": 3.5553633217993085e-05, | |
| "loss": 1.5347, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.4951972961425781, | |
| "learning_rate": 3.5121107266435986e-05, | |
| "loss": 1.2704, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7146666666666667, | |
| "grad_norm": 0.4018455743789673, | |
| "learning_rate": 3.4688581314878894e-05, | |
| "loss": 1.5217, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.7182222222222222, | |
| "grad_norm": 0.5961456894874573, | |
| "learning_rate": 3.4256055363321795e-05, | |
| "loss": 1.7622, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.7217777777777777, | |
| "grad_norm": 0.4609779119491577, | |
| "learning_rate": 3.382352941176471e-05, | |
| "loss": 1.5252, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.7253333333333334, | |
| "grad_norm": 0.43142765760421753, | |
| "learning_rate": 3.339100346020762e-05, | |
| "loss": 1.4461, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 0.4517713785171509, | |
| "learning_rate": 3.295847750865052e-05, | |
| "loss": 1.5962, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.7324444444444445, | |
| "grad_norm": 0.48978036642074585, | |
| "learning_rate": 3.2525951557093426e-05, | |
| "loss": 1.5366, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.40700387954711914, | |
| "learning_rate": 3.2093425605536334e-05, | |
| "loss": 1.5572, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.7395555555555555, | |
| "grad_norm": 0.5392187237739563, | |
| "learning_rate": 3.166089965397924e-05, | |
| "loss": 1.5552, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.7431111111111111, | |
| "grad_norm": 0.44562795758247375, | |
| "learning_rate": 3.122837370242215e-05, | |
| "loss": 1.5194, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 0.44013121724128723, | |
| "learning_rate": 3.079584775086505e-05, | |
| "loss": 1.3915, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.7502222222222222, | |
| "grad_norm": 0.4336005747318268, | |
| "learning_rate": 3.036332179930796e-05, | |
| "loss": 1.4715, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.7537777777777778, | |
| "grad_norm": 0.5743895769119263, | |
| "learning_rate": 2.9930795847750863e-05, | |
| "loss": 1.5256, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.7573333333333333, | |
| "grad_norm": 0.45934122800827026, | |
| "learning_rate": 2.9498269896193774e-05, | |
| "loss": 1.5873, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.7608888888888888, | |
| "grad_norm": 0.5065307021141052, | |
| "learning_rate": 2.9065743944636682e-05, | |
| "loss": 1.5582, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 0.511725664138794, | |
| "learning_rate": 2.8633217993079586e-05, | |
| "loss": 1.4813, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5018854737281799, | |
| "learning_rate": 2.820069204152249e-05, | |
| "loss": 1.5786, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.7715555555555556, | |
| "grad_norm": 0.5292341113090515, | |
| "learning_rate": 2.77681660899654e-05, | |
| "loss": 1.6471, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.7751111111111111, | |
| "grad_norm": 0.3638977110385895, | |
| "learning_rate": 2.733564013840831e-05, | |
| "loss": 1.4863, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.7786666666666666, | |
| "grad_norm": 0.49934521317481995, | |
| "learning_rate": 2.6903114186851214e-05, | |
| "loss": 1.4587, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 0.5519063472747803, | |
| "learning_rate": 2.647058823529412e-05, | |
| "loss": 1.4915, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.7857777777777778, | |
| "grad_norm": 0.5289978384971619, | |
| "learning_rate": 2.6038062283737023e-05, | |
| "loss": 1.5924, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.7893333333333333, | |
| "grad_norm": 0.4768216907978058, | |
| "learning_rate": 2.560553633217993e-05, | |
| "loss": 1.4847, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.7928888888888889, | |
| "grad_norm": 0.5400416254997253, | |
| "learning_rate": 2.5173010380622842e-05, | |
| "loss": 1.5664, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.7964444444444444, | |
| "grad_norm": 0.5170167684555054, | |
| "learning_rate": 2.4740484429065743e-05, | |
| "loss": 1.7191, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.4661620259284973, | |
| "learning_rate": 2.430795847750865e-05, | |
| "loss": 1.584, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.8035555555555556, | |
| "grad_norm": 0.48751363158226013, | |
| "learning_rate": 2.387543252595156e-05, | |
| "loss": 1.6166, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.8071111111111111, | |
| "grad_norm": 0.48189136385917664, | |
| "learning_rate": 2.3442906574394467e-05, | |
| "loss": 1.412, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.8106666666666666, | |
| "grad_norm": 0.4377134144306183, | |
| "learning_rate": 2.301038062283737e-05, | |
| "loss": 1.4633, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.8142222222222222, | |
| "grad_norm": 0.4900270998477936, | |
| "learning_rate": 2.2577854671280275e-05, | |
| "loss": 1.5968, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 0.5673394203186035, | |
| "learning_rate": 2.2145328719723187e-05, | |
| "loss": 1.4443, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.8213333333333334, | |
| "grad_norm": 0.5573922395706177, | |
| "learning_rate": 2.171280276816609e-05, | |
| "loss": 1.5951, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.8248888888888889, | |
| "grad_norm": 0.5815632343292236, | |
| "learning_rate": 2.1280276816609e-05, | |
| "loss": 1.54, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.8284444444444444, | |
| "grad_norm": 0.5276745557785034, | |
| "learning_rate": 2.0847750865051903e-05, | |
| "loss": 1.5903, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.4580680727958679, | |
| "learning_rate": 2.041522491349481e-05, | |
| "loss": 1.6143, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 0.5015590190887451, | |
| "learning_rate": 1.998269896193772e-05, | |
| "loss": 1.4156, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.8391111111111111, | |
| "grad_norm": 0.508334219455719, | |
| "learning_rate": 1.9550173010380623e-05, | |
| "loss": 1.5272, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.8426666666666667, | |
| "grad_norm": 0.44540607929229736, | |
| "learning_rate": 1.9117647058823528e-05, | |
| "loss": 1.5522, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.8462222222222222, | |
| "grad_norm": 0.5358651876449585, | |
| "learning_rate": 1.868512110726644e-05, | |
| "loss": 1.4527, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.8497777777777777, | |
| "grad_norm": 0.5191397666931152, | |
| "learning_rate": 1.8252595155709343e-05, | |
| "loss": 1.5627, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.4698735773563385, | |
| "learning_rate": 1.782006920415225e-05, | |
| "loss": 1.4434, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.8568888888888889, | |
| "grad_norm": 0.4450030028820038, | |
| "learning_rate": 1.7387543252595156e-05, | |
| "loss": 1.383, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.8604444444444445, | |
| "grad_norm": 0.47164052724838257, | |
| "learning_rate": 1.6955017301038063e-05, | |
| "loss": 1.5499, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.4113026559352875, | |
| "learning_rate": 1.652249134948097e-05, | |
| "loss": 1.499, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.8675555555555555, | |
| "grad_norm": 0.48519784212112427, | |
| "learning_rate": 1.6089965397923876e-05, | |
| "loss": 1.5236, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 0.5262069702148438, | |
| "learning_rate": 1.5657439446366783e-05, | |
| "loss": 1.5144, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.8746666666666667, | |
| "grad_norm": 0.42513802647590637, | |
| "learning_rate": 1.522491349480969e-05, | |
| "loss": 1.4932, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.8782222222222222, | |
| "grad_norm": 0.47486934065818787, | |
| "learning_rate": 1.4792387543252596e-05, | |
| "loss": 1.4398, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.8817777777777778, | |
| "grad_norm": 0.5634934902191162, | |
| "learning_rate": 1.4359861591695503e-05, | |
| "loss": 1.65, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.8853333333333333, | |
| "grad_norm": 0.43849506974220276, | |
| "learning_rate": 1.3927335640138408e-05, | |
| "loss": 1.5201, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.4285069704055786, | |
| "learning_rate": 1.3494809688581317e-05, | |
| "loss": 1.492, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8924444444444445, | |
| "grad_norm": 0.5707066655158997, | |
| "learning_rate": 1.3062283737024222e-05, | |
| "loss": 1.5991, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.4697231948375702, | |
| "learning_rate": 1.2629757785467128e-05, | |
| "loss": 1.5585, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.8995555555555556, | |
| "grad_norm": 0.4958859384059906, | |
| "learning_rate": 1.2197231833910034e-05, | |
| "loss": 1.5496, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.9031111111111111, | |
| "grad_norm": 0.4845017194747925, | |
| "learning_rate": 1.1764705882352942e-05, | |
| "loss": 1.592, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 0.4356600344181061, | |
| "learning_rate": 1.1332179930795848e-05, | |
| "loss": 1.5873, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.9102222222222223, | |
| "grad_norm": 0.3917759954929352, | |
| "learning_rate": 1.0899653979238756e-05, | |
| "loss": 1.4984, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.9137777777777778, | |
| "grad_norm": 0.48273682594299316, | |
| "learning_rate": 1.046712802768166e-05, | |
| "loss": 1.4288, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.9173333333333333, | |
| "grad_norm": 0.5662245750427246, | |
| "learning_rate": 1.0034602076124568e-05, | |
| "loss": 1.5123, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.9208888888888889, | |
| "grad_norm": 0.510188639163971, | |
| "learning_rate": 9.602076124567474e-06, | |
| "loss": 1.4915, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.9244444444444444, | |
| "grad_norm": 0.5999191403388977, | |
| "learning_rate": 9.169550173010382e-06, | |
| "loss": 1.7374, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.5733387470245361, | |
| "learning_rate": 8.737024221453288e-06, | |
| "loss": 1.5551, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.9315555555555556, | |
| "grad_norm": 0.3804311156272888, | |
| "learning_rate": 8.304498269896194e-06, | |
| "loss": 1.5631, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.9351111111111111, | |
| "grad_norm": 0.44998374581336975, | |
| "learning_rate": 7.8719723183391e-06, | |
| "loss": 1.7566, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.9386666666666666, | |
| "grad_norm": 0.4219221770763397, | |
| "learning_rate": 7.439446366782007e-06, | |
| "loss": 1.454, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.9422222222222222, | |
| "grad_norm": 0.48252835869789124, | |
| "learning_rate": 7.006920415224914e-06, | |
| "loss": 1.5107, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.9457777777777778, | |
| "grad_norm": 0.5446690320968628, | |
| "learning_rate": 6.5743944636678194e-06, | |
| "loss": 1.5333, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.9493333333333334, | |
| "grad_norm": 0.4676448106765747, | |
| "learning_rate": 6.141868512110726e-06, | |
| "loss": 1.4822, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.9528888888888889, | |
| "grad_norm": 0.41428953409194946, | |
| "learning_rate": 5.709342560553633e-06, | |
| "loss": 1.398, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.9564444444444444, | |
| "grad_norm": 0.39139533042907715, | |
| "learning_rate": 5.2768166089965395e-06, | |
| "loss": 1.539, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5213208794593811, | |
| "learning_rate": 4.8442906574394464e-06, | |
| "loss": 1.535, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.9635555555555556, | |
| "grad_norm": 0.6053948402404785, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 1.484, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.9671111111111111, | |
| "grad_norm": 0.44833922386169434, | |
| "learning_rate": 3.9792387543252595e-06, | |
| "loss": 1.3107, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.9706666666666667, | |
| "grad_norm": 0.4769100844860077, | |
| "learning_rate": 3.5467128027681665e-06, | |
| "loss": 1.5114, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.9742222222222222, | |
| "grad_norm": 0.5112429857254028, | |
| "learning_rate": 3.1141868512110726e-06, | |
| "loss": 1.5357, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 0.4884449243545532, | |
| "learning_rate": 2.6816608996539796e-06, | |
| "loss": 1.4739, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.9813333333333333, | |
| "grad_norm": 0.5470872521400452, | |
| "learning_rate": 2.249134948096886e-06, | |
| "loss": 1.5557, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.9848888888888889, | |
| "grad_norm": 0.517669141292572, | |
| "learning_rate": 1.8166089965397926e-06, | |
| "loss": 1.5388, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.9884444444444445, | |
| "grad_norm": 0.46402767300605774, | |
| "learning_rate": 1.3840830449826992e-06, | |
| "loss": 1.5509, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.5298945307731628, | |
| "learning_rate": 9.515570934256056e-07, | |
| "loss": 1.533, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 0.5506584644317627, | |
| "learning_rate": 5.190311418685121e-07, | |
| "loss": 1.404, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.9991111111111111, | |
| "grad_norm": 0.5148972868919373, | |
| "learning_rate": 8.65051903114187e-08, | |
| "loss": 1.558, | |
| "step": 2810 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2812, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1067056282035814e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |