Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.11097816851341276, | |
| "eval_steps": 500, | |
| "global_step": 1600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006936135532088297, | |
| "grad_norm": 80.97769927978516, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 39.45225219726562, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0013872271064176594, | |
| "grad_norm": 69.19830322265625, | |
| "learning_rate": 6.25e-06, | |
| "loss": 38.231222534179686, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0020808406596264893, | |
| "grad_norm": 55.03429412841797, | |
| "learning_rate": 1.0416666666666666e-05, | |
| "loss": 37.07829284667969, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.002774454212835319, | |
| "grad_norm": 36.03839874267578, | |
| "learning_rate": 1.4583333333333333e-05, | |
| "loss": 35.945098876953125, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0034680677660441487, | |
| "grad_norm": 48.63792419433594, | |
| "learning_rate": 1.875e-05, | |
| "loss": 34.40289306640625, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004161681319252979, | |
| "grad_norm": 41.98967361450195, | |
| "learning_rate": 2.2916666666666667e-05, | |
| "loss": 33.45003662109375, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004855294872461808, | |
| "grad_norm": 81.57576751708984, | |
| "learning_rate": 2.7083333333333335e-05, | |
| "loss": 32.02000427246094, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.005548908425670638, | |
| "grad_norm": 60.575645446777344, | |
| "learning_rate": 3.125e-05, | |
| "loss": 30.74504699707031, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.006242521978879467, | |
| "grad_norm": 61.149410247802734, | |
| "learning_rate": 3.541666666666667e-05, | |
| "loss": 29.49912109375, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.006936135532088297, | |
| "grad_norm": 69.96782684326172, | |
| "learning_rate": 3.958333333333333e-05, | |
| "loss": 28.152670288085936, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.007629749085297127, | |
| "grad_norm": 33.35784149169922, | |
| "learning_rate": 4.375e-05, | |
| "loss": 27.86654052734375, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.008323362638505957, | |
| "grad_norm": 42.682308197021484, | |
| "learning_rate": 4.791666666666667e-05, | |
| "loss": 27.098068237304688, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.009016976191714786, | |
| "grad_norm": 45.40013122558594, | |
| "learning_rate": 5.208333333333334e-05, | |
| "loss": 25.768472290039064, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.009710589744923616, | |
| "grad_norm": 36.54875564575195, | |
| "learning_rate": 5.6250000000000005e-05, | |
| "loss": 25.419296264648438, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.010404203298132445, | |
| "grad_norm": 54.964447021484375, | |
| "learning_rate": 6.041666666666667e-05, | |
| "loss": 24.5228759765625, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.011097816851341275, | |
| "grad_norm": 62.00859451293945, | |
| "learning_rate": 6.458333333333334e-05, | |
| "loss": 23.92150115966797, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.011791430404550106, | |
| "grad_norm": 50.303070068359375, | |
| "learning_rate": 6.875e-05, | |
| "loss": 23.61223449707031, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.012485043957758934, | |
| "grad_norm": 40.088356018066406, | |
| "learning_rate": 7.291666666666667e-05, | |
| "loss": 23.29130401611328, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.013178657510967764, | |
| "grad_norm": 58.86671829223633, | |
| "learning_rate": 7.708333333333334e-05, | |
| "loss": 23.32776794433594, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.013872271064176595, | |
| "grad_norm": 38.86407470703125, | |
| "learning_rate": 8.125000000000001e-05, | |
| "loss": 23.280059814453125, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.014565884617385423, | |
| "grad_norm": 37.631717681884766, | |
| "learning_rate": 8.541666666666666e-05, | |
| "loss": 22.259375, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.015259498170594254, | |
| "grad_norm": 37.38259506225586, | |
| "learning_rate": 8.958333333333333e-05, | |
| "loss": 21.507147216796874, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.015953111723803082, | |
| "grad_norm": 42.106143951416016, | |
| "learning_rate": 9.375e-05, | |
| "loss": 21.231875610351562, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.016646725277011915, | |
| "grad_norm": 50.364681243896484, | |
| "learning_rate": 9.791666666666667e-05, | |
| "loss": 21.234603881835938, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.017340338830220743, | |
| "grad_norm": 40.50483322143555, | |
| "learning_rate": 0.00010208333333333334, | |
| "loss": 20.62218017578125, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.018033952383429572, | |
| "grad_norm": 60.68687057495117, | |
| "learning_rate": 0.00010625, | |
| "loss": 19.96156463623047, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.0187275659366384, | |
| "grad_norm": 31.746686935424805, | |
| "learning_rate": 0.00011041666666666666, | |
| "loss": 19.283287048339844, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.019421179489847232, | |
| "grad_norm": 40.83134841918945, | |
| "learning_rate": 0.00011458333333333333, | |
| "loss": 18.95526580810547, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02011479304305606, | |
| "grad_norm": 27.94510841369629, | |
| "learning_rate": 0.00011875, | |
| "loss": 19.398623657226562, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.02080840659626489, | |
| "grad_norm": 43.998260498046875, | |
| "learning_rate": 0.00012291666666666665, | |
| "loss": 18.262245178222656, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.021502020149473722, | |
| "grad_norm": 22.61271095275879, | |
| "learning_rate": 0.00012708333333333332, | |
| "loss": 18.646043395996095, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.02219563370268255, | |
| "grad_norm": 27.580198287963867, | |
| "learning_rate": 0.00013125000000000002, | |
| "loss": 17.519342041015626, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.02288924725589138, | |
| "grad_norm": 36.54242706298828, | |
| "learning_rate": 0.00013541666666666666, | |
| "loss": 16.588734436035157, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.02358286080910021, | |
| "grad_norm": 24.73045539855957, | |
| "learning_rate": 0.00013958333333333333, | |
| "loss": 16.85819091796875, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.02427647436230904, | |
| "grad_norm": 27.577693939208984, | |
| "learning_rate": 0.00014375, | |
| "loss": 16.190365600585938, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02497008791551787, | |
| "grad_norm": 20.408132553100586, | |
| "learning_rate": 0.00014791666666666667, | |
| "loss": 15.682986450195312, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.0256637014687267, | |
| "grad_norm": 28.981483459472656, | |
| "learning_rate": 0.00015208333333333333, | |
| "loss": 15.275959777832032, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.02635731502193553, | |
| "grad_norm": 19.978363037109375, | |
| "learning_rate": 0.00015625, | |
| "loss": 15.033628845214844, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.027050928575144358, | |
| "grad_norm": 15.66555404663086, | |
| "learning_rate": 0.00016041666666666667, | |
| "loss": 14.446699523925782, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02774454212835319, | |
| "grad_norm": 17.955888748168945, | |
| "learning_rate": 0.00016458333333333334, | |
| "loss": 14.196858215332032, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.02843815568156202, | |
| "grad_norm": 16.955345153808594, | |
| "learning_rate": 0.00016875, | |
| "loss": 13.838204956054687, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.029131769234770847, | |
| "grad_norm": 14.126781463623047, | |
| "learning_rate": 0.00017291666666666668, | |
| "loss": 13.345814514160157, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.029825382787979676, | |
| "grad_norm": 16.29298973083496, | |
| "learning_rate": 0.00017708333333333335, | |
| "loss": 13.06510009765625, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.030518996341188508, | |
| "grad_norm": 14.477455139160156, | |
| "learning_rate": 0.00018125, | |
| "loss": 12.520622253417969, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.031212609894397336, | |
| "grad_norm": 12.074920654296875, | |
| "learning_rate": 0.00018541666666666668, | |
| "loss": 12.411215209960938, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.031906223447606165, | |
| "grad_norm": 15.564898490905762, | |
| "learning_rate": 0.00018958333333333332, | |
| "loss": 12.00233154296875, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.032599837000814993, | |
| "grad_norm": 15.621328353881836, | |
| "learning_rate": 0.00019375000000000002, | |
| "loss": 12.109133911132812, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.03329345055402383, | |
| "grad_norm": 10.823607444763184, | |
| "learning_rate": 0.00019791666666666666, | |
| "loss": 11.684900665283203, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.03398706410723266, | |
| "grad_norm": 14.166363716125488, | |
| "learning_rate": 0.00020208333333333333, | |
| "loss": 11.60366439819336, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.034680677660441486, | |
| "grad_norm": 39.65184783935547, | |
| "learning_rate": 0.00020625, | |
| "loss": 11.060487365722656, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.035374291213650315, | |
| "grad_norm": 12.022449493408203, | |
| "learning_rate": 0.00021041666666666667, | |
| "loss": 11.373404693603515, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.036067904766859143, | |
| "grad_norm": 9.625805854797363, | |
| "learning_rate": 0.00021458333333333334, | |
| "loss": 11.405331420898438, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.03676151832006797, | |
| "grad_norm": 12.523700714111328, | |
| "learning_rate": 0.00021875, | |
| "loss": 10.810578155517579, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.0374551318732768, | |
| "grad_norm": 11.111766815185547, | |
| "learning_rate": 0.00022291666666666667, | |
| "loss": 10.427848815917969, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.038148745426485636, | |
| "grad_norm": 9.552438735961914, | |
| "learning_rate": 0.00022708333333333334, | |
| "loss": 10.221891784667969, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.038842358979694465, | |
| "grad_norm": 10.854446411132812, | |
| "learning_rate": 0.00023125, | |
| "loss": 10.453067779541016, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.039535972532903293, | |
| "grad_norm": 8.054693222045898, | |
| "learning_rate": 0.00023541666666666668, | |
| "loss": 9.835127258300782, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.04022958608611212, | |
| "grad_norm": 9.217058181762695, | |
| "learning_rate": 0.00023958333333333335, | |
| "loss": 9.983591461181641, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.04092319963932095, | |
| "grad_norm": 7.442865371704102, | |
| "learning_rate": 0.00024375, | |
| "loss": 9.785844421386718, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.04161681319252978, | |
| "grad_norm": 7.499183177947998, | |
| "learning_rate": 0.0002479166666666667, | |
| "loss": 9.547152709960937, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.042310426745738615, | |
| "grad_norm": 6.999185085296631, | |
| "learning_rate": 0.0002520833333333333, | |
| "loss": 9.602791595458985, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.043004040298947444, | |
| "grad_norm": 8.540186882019043, | |
| "learning_rate": 0.00025624999999999997, | |
| "loss": 9.688284301757813, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.04369765385215627, | |
| "grad_norm": 7.028185844421387, | |
| "learning_rate": 0.0002604166666666667, | |
| "loss": 9.46673355102539, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.0443912674053651, | |
| "grad_norm": 10.2844820022583, | |
| "learning_rate": 0.00026458333333333336, | |
| "loss": 9.028829956054688, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.04508488095857393, | |
| "grad_norm": 6.920653820037842, | |
| "learning_rate": 0.00026875, | |
| "loss": 9.584944152832032, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.04577849451178276, | |
| "grad_norm": 7.473561763763428, | |
| "learning_rate": 0.00027291666666666664, | |
| "loss": 9.417964935302734, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04647210806499159, | |
| "grad_norm": 6.416537761688232, | |
| "learning_rate": 0.00027708333333333334, | |
| "loss": 9.235848236083985, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.04716572161820042, | |
| "grad_norm": 7.264041900634766, | |
| "learning_rate": 0.00028125000000000003, | |
| "loss": 9.052194976806641, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04785933517140925, | |
| "grad_norm": 5.6128692626953125, | |
| "learning_rate": 0.0002854166666666667, | |
| "loss": 8.73398208618164, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.04855294872461808, | |
| "grad_norm": 6.392711639404297, | |
| "learning_rate": 0.00028958333333333337, | |
| "loss": 8.775264739990234, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04924656227782691, | |
| "grad_norm": 7.4516496658325195, | |
| "learning_rate": 0.00029375, | |
| "loss": 8.859925079345704, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.04994017583103574, | |
| "grad_norm": 6.566150188446045, | |
| "learning_rate": 0.00029791666666666665, | |
| "loss": 9.032890319824219, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.050633789384244565, | |
| "grad_norm": 7.080583572387695, | |
| "learning_rate": 0.0003020833333333333, | |
| "loss": 8.99814682006836, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.0513274029374534, | |
| "grad_norm": 5.690314292907715, | |
| "learning_rate": 0.00030625000000000004, | |
| "loss": 8.487692260742188, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.05202101649066223, | |
| "grad_norm": 6.124709606170654, | |
| "learning_rate": 0.0003104166666666667, | |
| "loss": 9.122718811035156, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.05271463004387106, | |
| "grad_norm": 5.513874530792236, | |
| "learning_rate": 0.00031458333333333333, | |
| "loss": 8.884898376464843, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.05340824359707989, | |
| "grad_norm": 6.077105522155762, | |
| "learning_rate": 0.00031874999999999997, | |
| "loss": 8.678743743896485, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.054101857150288715, | |
| "grad_norm": 5.43304967880249, | |
| "learning_rate": 0.0003229166666666667, | |
| "loss": 8.514462280273438, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.054795470703497544, | |
| "grad_norm": 6.31812858581543, | |
| "learning_rate": 0.00032708333333333336, | |
| "loss": 8.414434814453125, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.05548908425670638, | |
| "grad_norm": 6.50822114944458, | |
| "learning_rate": 0.00033125, | |
| "loss": 8.508963775634765, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.05618269780991521, | |
| "grad_norm": 5.709455966949463, | |
| "learning_rate": 0.00033541666666666664, | |
| "loss": 8.19250030517578, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.05687631136312404, | |
| "grad_norm": 6.013482093811035, | |
| "learning_rate": 0.00033958333333333334, | |
| "loss": 8.754444885253907, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.057569924916332865, | |
| "grad_norm": 6.8106842041015625, | |
| "learning_rate": 0.00034375, | |
| "loss": 8.26578598022461, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.058263538469541694, | |
| "grad_norm": 5.135249614715576, | |
| "learning_rate": 0.0003479166666666667, | |
| "loss": 8.549038696289063, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.05895715202275052, | |
| "grad_norm": 5.1326823234558105, | |
| "learning_rate": 0.00035208333333333337, | |
| "loss": 8.302710723876952, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.05965076557595935, | |
| "grad_norm": 4.964430809020996, | |
| "learning_rate": 0.00035625, | |
| "loss": 8.092277526855469, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.06034437912916819, | |
| "grad_norm": 6.6412882804870605, | |
| "learning_rate": 0.00036041666666666665, | |
| "loss": 7.938130187988281, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.061037992682377015, | |
| "grad_norm": 5.193583011627197, | |
| "learning_rate": 0.0003645833333333333, | |
| "loss": 8.546875, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.061731606235585844, | |
| "grad_norm": 4.927044868469238, | |
| "learning_rate": 0.00036875000000000005, | |
| "loss": 7.911424255371093, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.06242521978879467, | |
| "grad_norm": 5.043755531311035, | |
| "learning_rate": 0.0003729166666666667, | |
| "loss": 8.359856414794923, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.06311883334200351, | |
| "grad_norm": 5.2342352867126465, | |
| "learning_rate": 0.00037708333333333333, | |
| "loss": 7.912593841552734, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.06381244689521233, | |
| "grad_norm": 4.811470985412598, | |
| "learning_rate": 0.00038124999999999997, | |
| "loss": 8.026778411865234, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.06450606044842117, | |
| "grad_norm": 5.230201721191406, | |
| "learning_rate": 0.0003854166666666667, | |
| "loss": 7.771602630615234, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.06519967400162999, | |
| "grad_norm": 5.551824569702148, | |
| "learning_rate": 0.00038958333333333336, | |
| "loss": 7.761222839355469, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.06589328755483882, | |
| "grad_norm": 4.743410110473633, | |
| "learning_rate": 0.00039375, | |
| "loss": 7.748975372314453, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.06658690110804766, | |
| "grad_norm": 5.341291427612305, | |
| "learning_rate": 0.00039791666666666664, | |
| "loss": 7.9248199462890625, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.06728051466125648, | |
| "grad_norm": 4.554710865020752, | |
| "learning_rate": 0.00040166666666666665, | |
| "loss": 7.922956085205078, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.06797412821446532, | |
| "grad_norm": 4.599163055419922, | |
| "learning_rate": 0.00040583333333333335, | |
| "loss": 7.850189208984375, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.06866774176767414, | |
| "grad_norm": 4.907506942749023, | |
| "learning_rate": 0.00041, | |
| "loss": 7.55450668334961, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.06936135532088297, | |
| "grad_norm": 4.697291374206543, | |
| "learning_rate": 0.0004141666666666667, | |
| "loss": 7.54024658203125, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0700549688740918, | |
| "grad_norm": 5.102283000946045, | |
| "learning_rate": 0.00041833333333333333, | |
| "loss": 7.947976684570312, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.07074858242730063, | |
| "grad_norm": 5.308175563812256, | |
| "learning_rate": 0.00042249999999999997, | |
| "loss": 7.837068176269531, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.07144219598050947, | |
| "grad_norm": 5.058789253234863, | |
| "learning_rate": 0.0004266666666666667, | |
| "loss": 7.611346435546875, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.07213580953371829, | |
| "grad_norm": 5.363205432891846, | |
| "learning_rate": 0.00043083333333333336, | |
| "loss": 7.675627899169922, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.07282942308692712, | |
| "grad_norm": 5.060066223144531, | |
| "learning_rate": 0.000435, | |
| "loss": 7.799678802490234, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.07352303664013594, | |
| "grad_norm": 5.268683433532715, | |
| "learning_rate": 0.00043916666666666664, | |
| "loss": 7.417415618896484, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.07421665019334478, | |
| "grad_norm": 29.964582443237305, | |
| "learning_rate": 0.00044333333333333334, | |
| "loss": 7.611949920654297, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.0749102637465536, | |
| "grad_norm": 6.951490879058838, | |
| "learning_rate": 0.00044750000000000004, | |
| "loss": 7.826107025146484, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.07560387729976244, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0004516666666666667, | |
| "loss": 8.342691040039062, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.07629749085297127, | |
| "grad_norm": 19.79059410095215, | |
| "learning_rate": 0.0004554166666666667, | |
| "loss": 8.860975646972657, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.0769911044061801, | |
| "grad_norm": 5.309981346130371, | |
| "learning_rate": 0.00045958333333333333, | |
| "loss": 8.495597839355469, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.07768471795938893, | |
| "grad_norm": 62.24746322631836, | |
| "learning_rate": 0.00046375, | |
| "loss": 7.669882965087891, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.07837833151259775, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00046791666666666667, | |
| "loss": 9.232691192626953, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.07907194506580659, | |
| "grad_norm": 21.891761779785156, | |
| "learning_rate": 0.00047125, | |
| "loss": 9.408066558837891, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.07976555861901541, | |
| "grad_norm": 47.06490707397461, | |
| "learning_rate": 0.0004754166666666667, | |
| "loss": 9.484445190429687, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.08045917217222424, | |
| "grad_norm": 34.64010238647461, | |
| "learning_rate": 0.0004795833333333333, | |
| "loss": 9.570259857177735, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.08115278572543308, | |
| "grad_norm": 12.11959457397461, | |
| "learning_rate": 0.00048375, | |
| "loss": 9.447135162353515, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.0818463992786419, | |
| "grad_norm": 16.15092658996582, | |
| "learning_rate": 0.0004875, | |
| "loss": 9.065760040283203, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.08254001283185074, | |
| "grad_norm": 157.4208526611328, | |
| "learning_rate": 0.0004916666666666666, | |
| "loss": 8.82796630859375, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.08323362638505956, | |
| "grad_norm": 27.526687622070312, | |
| "learning_rate": 0.0004958333333333334, | |
| "loss": 8.742295837402343, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.0839272399382684, | |
| "grad_norm": 145.7564697265625, | |
| "learning_rate": 0.0005, | |
| "loss": 8.395146942138672, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.08462085349147723, | |
| "grad_norm": 6.748693466186523, | |
| "learning_rate": 0.0004999999873614903, | |
| "loss": 8.451645660400391, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.08531446704468605, | |
| "grad_norm": 22.430376052856445, | |
| "learning_rate": 0.0004999999494459626, | |
| "loss": 8.06414794921875, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.08600808059789489, | |
| "grad_norm": 11.808032035827637, | |
| "learning_rate": 0.0004999998862534205, | |
| "loss": 8.445089721679688, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.08670169415110371, | |
| "grad_norm": 40.87104034423828, | |
| "learning_rate": 0.0004999997977838706, | |
| "loss": 8.36043930053711, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.08739530770431254, | |
| "grad_norm": 6.653709411621094, | |
| "learning_rate": 0.0004999996840373218, | |
| "loss": 8.253304290771485, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.08808892125752137, | |
| "grad_norm": 143.88558959960938, | |
| "learning_rate": 0.0004999995450137855, | |
| "loss": 8.144001770019532, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.0887825348107302, | |
| "grad_norm": 21.387109756469727, | |
| "learning_rate": 0.0004999993807132759, | |
| "loss": 8.308104705810546, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.08947614836393904, | |
| "grad_norm": 24.346641540527344, | |
| "learning_rate": 0.0004999991911358096, | |
| "loss": 8.206471252441407, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.09016976191714786, | |
| "grad_norm": 5.9509124755859375, | |
| "learning_rate": 0.0004999989762814056, | |
| "loss": 7.916456604003907, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.0908633754703567, | |
| "grad_norm": 4.6718292236328125, | |
| "learning_rate": 0.0004999987361500857, | |
| "loss": 8.069509887695313, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.09155698902356552, | |
| "grad_norm": 7.908266067504883, | |
| "learning_rate": 0.0004999984707418744, | |
| "loss": 7.811135101318359, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.09225060257677435, | |
| "grad_norm": 5.006199836730957, | |
| "learning_rate": 0.0004999981800567982, | |
| "loss": 7.62738037109375, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.09294421612998317, | |
| "grad_norm": 5.048142433166504, | |
| "learning_rate": 0.0004999979293094143, | |
| "loss": 7.573948669433594, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.09363782968319201, | |
| "grad_norm": 315.8863220214844, | |
| "learning_rate": 0.0004999975931260575, | |
| "loss": 7.884748077392578, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.09433144323640084, | |
| "grad_norm": 6.798584938049316, | |
| "learning_rate": 0.0004999972689493917, | |
| "loss": 8.001516723632813, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.09502505678960967, | |
| "grad_norm": 1015.7785034179688, | |
| "learning_rate": 0.0004999968847401915, | |
| "loss": 7.8455970764160154, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.0957186703428185, | |
| "grad_norm": 10.685717582702637, | |
| "learning_rate": 0.000499996475254287, | |
| "loss": 8.098987579345703, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.09641228389602732, | |
| "grad_norm": 11.820308685302734, | |
| "learning_rate": 0.0004999960404917194, | |
| "loss": 8.00228271484375, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.09710589744923616, | |
| "grad_norm": 14.156989097595215, | |
| "learning_rate": 0.0004999955804525329, | |
| "loss": 7.7231597900390625, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.097799511002445, | |
| "grad_norm": 14.044855117797852, | |
| "learning_rate": 0.0004999950951367737, | |
| "loss": 7.547687530517578, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.09849312455565382, | |
| "grad_norm": 21.774089813232422, | |
| "learning_rate": 0.0004999945845444912, | |
| "loss": 7.418048095703125, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.09918673810886265, | |
| "grad_norm": 24.606578826904297, | |
| "learning_rate": 0.0004999940486757369, | |
| "loss": 7.59954605102539, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.09988035166207147, | |
| "grad_norm": 17.820003509521484, | |
| "learning_rate": 0.0004999934875305649, | |
| "loss": 7.650421142578125, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.10057396521528031, | |
| "grad_norm": 23.51424789428711, | |
| "learning_rate": 0.000499992901109032, | |
| "loss": 8.069045257568359, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.10126757876848913, | |
| "grad_norm": 35.154178619384766, | |
| "learning_rate": 0.0004999922894111975, | |
| "loss": 7.217467498779297, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.10196119232169797, | |
| "grad_norm": 14.090352058410645, | |
| "learning_rate": 0.0004999916524371232, | |
| "loss": 7.3218132019042965, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.1026548058749068, | |
| "grad_norm": 22.718524932861328, | |
| "learning_rate": 0.0004999909901868735, | |
| "loss": 7.336325836181641, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.10334841942811562, | |
| "grad_norm": 24.450363159179688, | |
| "learning_rate": 0.0004999903026605154, | |
| "loss": 7.122545623779297, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.10404203298132446, | |
| "grad_norm": 19.112335205078125, | |
| "learning_rate": 0.0004999895898581185, | |
| "loss": 7.021141052246094, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.10473564653453328, | |
| "grad_norm": 37.39822769165039, | |
| "learning_rate": 0.0004999888517797547, | |
| "loss": 7.5175323486328125, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.10542926008774212, | |
| "grad_norm": 17.441190719604492, | |
| "learning_rate": 0.0004999880884254987, | |
| "loss": 7.584983825683594, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.10612287364095094, | |
| "grad_norm": 32.337440490722656, | |
| "learning_rate": 0.0004999872997954278, | |
| "loss": 7.30379638671875, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.10681648719415977, | |
| "grad_norm": 15.138467788696289, | |
| "learning_rate": 0.0004999864858896215, | |
| "loss": 7.259065246582031, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.10751010074736861, | |
| "grad_norm": 27.0513916015625, | |
| "learning_rate": 0.0004999856467081623, | |
| "loss": 7.271750640869141, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.10820371430057743, | |
| "grad_norm": 26.314247131347656, | |
| "learning_rate": 0.0004999847822511349, | |
| "loss": 7.2289573669433596, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.10889732785378627, | |
| "grad_norm": 22.641141891479492, | |
| "learning_rate": 0.0004999838925186269, | |
| "loss": 7.1235198974609375, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.10959094140699509, | |
| "grad_norm": 17.976776123046875, | |
| "learning_rate": 0.000499982977510728, | |
| "loss": 7.235462951660156, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.11028455496020392, | |
| "grad_norm": 20.597803115844727, | |
| "learning_rate": 0.0004999820372275308, | |
| "loss": 7.060716247558593, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.11097816851341276, | |
| "grad_norm": 24.85616683959961, | |
| "learning_rate": 0.0004999810716691305, | |
| "loss": 7.108528137207031, | |
| "step": 1600 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 100000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.731588545421312e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |