| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 291, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003436426116838488, | |
| "grad_norm": 64.0579605102539, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 1.9159, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006872852233676976, | |
| "grad_norm": 65.67280578613281, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 1.9282, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010309278350515464, | |
| "grad_norm": 62.75189971923828, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.8737, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 51.93278503417969, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.6762, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01718213058419244, | |
| "grad_norm": 51.46208190917969, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.19, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 45.677215576171875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.0556, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.024054982817869417, | |
| "grad_norm": 34.49523162841797, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 0.5131, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 22.15294075012207, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.4563, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030927835051546393, | |
| "grad_norm": 51.690269470214844, | |
| "learning_rate": 3e-06, | |
| "loss": 0.4176, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 75.96099853515625, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.3745, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.037800687285223365, | |
| "grad_norm": 57.87851333618164, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 0.3727, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 19.976598739624023, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.3352, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.044673539518900345, | |
| "grad_norm": 63.86684799194336, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 0.3401, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 34.12485885620117, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.3305, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05154639175257732, | |
| "grad_norm": 14.09028148651123, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2854, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 11.474346160888672, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.3164, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.058419243986254296, | |
| "grad_norm": 4.925178527832031, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 0.2837, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 5.7118024826049805, | |
| "learning_rate": 6e-06, | |
| "loss": 0.2674, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06529209621993128, | |
| "grad_norm": 6.775313377380371, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 0.28, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 5.114953517913818, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.3022, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07216494845360824, | |
| "grad_norm": 3.6223747730255127, | |
| "learning_rate": 7e-06, | |
| "loss": 0.2418, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 3.428668975830078, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.253, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07903780068728522, | |
| "grad_norm": 5.323895454406738, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 0.3191, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 3.228813886642456, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.2574, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0859106529209622, | |
| "grad_norm": 84.88597869873047, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.2961, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 3.482680082321167, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.2809, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09278350515463918, | |
| "grad_norm": 3.876138210296631, | |
| "learning_rate": 9e-06, | |
| "loss": 0.3013, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 37.9251708984375, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.3072, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09965635738831616, | |
| "grad_norm": 4.258779048919678, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 0.2936, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 3.2512130737304688, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2872, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10652920962199312, | |
| "grad_norm": 3.553802013397217, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.2959, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 3.18969464302063, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.273, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1134020618556701, | |
| "grad_norm": 3.7316102981567383, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.2857, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 4.492163181304932, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.3157, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12027491408934708, | |
| "grad_norm": 2.9147143363952637, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.2496, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 2.874291181564331, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.2818, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12714776632302405, | |
| "grad_norm": 2.704557418823242, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.2729, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 3.3673787117004395, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.2707, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13402061855670103, | |
| "grad_norm": 3.7019476890563965, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.2959, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 3.152188539505005, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.2975, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.140893470790378, | |
| "grad_norm": 3.401273727416992, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.2974, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 3.1594648361206055, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.2879, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.14776632302405499, | |
| "grad_norm": 3.1256396770477295, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.2806, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 3.244800567626953, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.2893, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15463917525773196, | |
| "grad_norm": 2.8971827030181885, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.2774, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 3.1087865829467773, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.2683, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16151202749140894, | |
| "grad_norm": 2.767376184463501, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.2552, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 3.0735318660736084, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.2966, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16838487972508592, | |
| "grad_norm": 2.5513877868652344, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.2677, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 2.7748734951019287, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.2575, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17525773195876287, | |
| "grad_norm": 2.729177951812744, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.2646, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 3.340230703353882, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.295, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.18213058419243985, | |
| "grad_norm": 2.7592291831970215, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.2846, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 2.845653533935547, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.3016, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18900343642611683, | |
| "grad_norm": 2.902620792388916, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.3015, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 2.291102170944214, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.2612, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1958762886597938, | |
| "grad_norm": 2.6173219680786133, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.2571, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 2.403263568878174, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.2831, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2027491408934708, | |
| "grad_norm": 2.4737367630004883, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.264, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 2.7195911407470703, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.2606, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20962199312714777, | |
| "grad_norm": 2.3145008087158203, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.2835, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 3.18537974357605, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.2815, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.21649484536082475, | |
| "grad_norm": 2.2597475051879883, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.2354, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 2.2232203483581543, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.2441, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.22336769759450173, | |
| "grad_norm": 2.7159595489501953, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.2716, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 2.018402338027954, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.2346, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.23024054982817868, | |
| "grad_norm": 3.5405406951904297, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.2637, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 2.1763646602630615, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.2672, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.23711340206185566, | |
| "grad_norm": 2.762448310852051, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.2762, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 2.7765045166015625, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.2974, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24398625429553264, | |
| "grad_norm": 2.584392786026001, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.2646, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 2.26948881149292, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.2772, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2508591065292096, | |
| "grad_norm": 2.6476385593414307, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.2857, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 3.0025851726531982, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.2474, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.25773195876288657, | |
| "grad_norm": 2.7429323196411133, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.2824, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 2.783100128173828, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.2657, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2646048109965636, | |
| "grad_norm": 2.404728412628174, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.253, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 43.180023193359375, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.3715, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.27147766323024053, | |
| "grad_norm": 27.71044158935547, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.2931, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 10.407346725463867, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.255, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.27835051546391754, | |
| "grad_norm": 78.4120864868164, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.3034, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 7.269218444824219, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.2566, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2852233676975945, | |
| "grad_norm": 20.143714904785156, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.2919, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 10.443071365356445, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.2832, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2920962199312715, | |
| "grad_norm": 3.284693956375122, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.2717, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 5.021488189697266, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.2537, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.29896907216494845, | |
| "grad_norm": 11.737953186035156, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.2934, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 6.13554573059082, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.2832, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.30584192439862545, | |
| "grad_norm": 3.5634255409240723, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.2358, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 7.335002422332764, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.2539, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3127147766323024, | |
| "grad_norm": 23.219636917114258, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.3195, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 5.62996244430542, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.2813, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.31958762886597936, | |
| "grad_norm": 246.3622589111328, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.3332, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 3.47241473197937, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.3132, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.32646048109965636, | |
| "grad_norm": 12.494342803955078, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.2885, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 19.562833786010742, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.3113, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 5.848825931549072, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.2473, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 27.47506332397461, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.27, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3402061855670103, | |
| "grad_norm": 4.550098419189453, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.2323, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 6.158257007598877, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.2773, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3470790378006873, | |
| "grad_norm": 15.120879173278809, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.3547, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 8.835148811340332, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.3497, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3539518900343643, | |
| "grad_norm": 3.3509395122528076, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.2595, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 2.217836618423462, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.2627, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.36082474226804123, | |
| "grad_norm": 3.7441632747650146, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.2419, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 4.696460723876953, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.2192, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.36769759450171824, | |
| "grad_norm": 2.353107452392578, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.2344, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 3.926302194595337, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.2488, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3745704467353952, | |
| "grad_norm": 2.4857568740844727, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.2549, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 1.9949979782104492, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.2495, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38144329896907214, | |
| "grad_norm": 1.9867693185806274, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.2615, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 2.11623477935791, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.2724, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.38831615120274915, | |
| "grad_norm": 1.8944286108016968, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.2187, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 1.738309621810913, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.2417, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3951890034364261, | |
| "grad_norm": 2.1811163425445557, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.2557, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 2.0165836811065674, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.2231, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4020618556701031, | |
| "grad_norm": 2.3740527629852295, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2434, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.8591188192367554, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.2275, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.40893470790378006, | |
| "grad_norm": 1.9685182571411133, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.237, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 2.128818988800049, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.229, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.41580756013745707, | |
| "grad_norm": 2.165508508682251, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.2479, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 2.245380401611328, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.2684, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.422680412371134, | |
| "grad_norm": 1.7729967832565308, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.2387, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 2.2421905994415283, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.2396, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.42955326460481097, | |
| "grad_norm": 1.9419755935668945, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.2378, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 1.8938632011413574, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.2361, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.436426116838488, | |
| "grad_norm": 2.994981288909912, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.2528, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 2.047788381576538, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.2513, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.44329896907216493, | |
| "grad_norm": 2.0908901691436768, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.2152, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 2.4644296169281006, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.2634, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.45017182130584193, | |
| "grad_norm": 1.8732101917266846, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.2495, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 1.7661372423171997, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.232, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4570446735395189, | |
| "grad_norm": 2.330557346343994, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.2477, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 1.983845829963684, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.2258, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4639175257731959, | |
| "grad_norm": 2.0565247535705566, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.2561, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 1.5290687084197998, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.2033, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.47079037800687284, | |
| "grad_norm": 2.0628135204315186, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.2301, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 2.0989890098571777, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.2604, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.47766323024054985, | |
| "grad_norm": 2.038011312484741, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.2476, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 1.9224317073822021, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.2594, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4845360824742268, | |
| "grad_norm": 2.059175968170166, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.2318, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 1.8125061988830566, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.2402, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.49140893470790376, | |
| "grad_norm": 1.5063058137893677, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.1993, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 2.565854787826538, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.2047, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.49828178694158076, | |
| "grad_norm": 4.078191757202148, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.2431, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 1.974707841873169, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.2379, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5051546391752577, | |
| "grad_norm": 1.824198603630066, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.237, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 1.805503010749817, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.2417, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5120274914089347, | |
| "grad_norm": 1.6662133932113647, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.2283, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 1.7907216548919678, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.2196, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5189003436426117, | |
| "grad_norm": 2.185821056365967, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.253, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 1.7748606204986572, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.239, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5257731958762887, | |
| "grad_norm": 1.6255580186843872, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.2045, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 2.2421576976776123, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.255, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5326460481099656, | |
| "grad_norm": 1.955019235610962, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.2672, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 2.139021635055542, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.2541, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5395189003436426, | |
| "grad_norm": 1.765238881111145, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.2346, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 2.081233263015747, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.2364, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5463917525773195, | |
| "grad_norm": 1.6718120574951172, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.2235, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 1.3806819915771484, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.1985, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5532646048109966, | |
| "grad_norm": 1.6870397329330444, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.227, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 2.0057849884033203, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.2442, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5601374570446735, | |
| "grad_norm": 1.9310015439987183, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.2309, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 1.8504823446273804, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.2261, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5670103092783505, | |
| "grad_norm": 1.8687225580215454, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.2447, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 1.7692251205444336, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.2236, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5738831615120275, | |
| "grad_norm": 1.6962001323699951, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.1953, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 2.1945459842681885, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.2232, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5807560137457045, | |
| "grad_norm": 1.8168156147003174, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.2177, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 1.6398793458938599, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.2371, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5876288659793815, | |
| "grad_norm": 1.975156307220459, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.2212, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 1.844821572303772, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.2177, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5945017182130584, | |
| "grad_norm": 1.8924890756607056, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.2625, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 1.8228974342346191, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.2409, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6013745704467354, | |
| "grad_norm": 1.6296133995056152, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2269, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 1.6581981182098389, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.2257, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6082474226804123, | |
| "grad_norm": 1.5767498016357422, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.1993, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 1.7926667928695679, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.2371, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6151202749140894, | |
| "grad_norm": 1.4467085599899292, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.2077, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 1.5783424377441406, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.2264, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6219931271477663, | |
| "grad_norm": 1.6146835088729858, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.2351, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 1.396660566329956, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.1975, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6288659793814433, | |
| "grad_norm": 2.4335689544677734, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.2304, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 1.6264203786849976, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.2316, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6357388316151202, | |
| "grad_norm": 1.5158411264419556, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.1917, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 1.5340527296066284, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.2033, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6426116838487973, | |
| "grad_norm": 1.5632439851760864, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.2151, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 1.4948574304580688, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.1956, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6494845360824743, | |
| "grad_norm": 1.7845014333724976, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.2324, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 1.7569440603256226, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.2081, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6563573883161512, | |
| "grad_norm": 1.8459192514419556, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.216, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 1.843494176864624, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.1951, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6632302405498282, | |
| "grad_norm": 1.4652260541915894, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.2079, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 2.8526852130889893, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.225, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6701030927835051, | |
| "grad_norm": 1.7010999917984009, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.2122, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 1.4640144109725952, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.1999, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6769759450171822, | |
| "grad_norm": 1.8528066873550415, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.2314, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 1.4770729541778564, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.2131, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6838487972508591, | |
| "grad_norm": 1.4935646057128906, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.2224, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 1.4355649948120117, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.2107, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6907216494845361, | |
| "grad_norm": 1.4284039735794067, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.2052, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 1.5143457651138306, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.2048, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.697594501718213, | |
| "grad_norm": 1.5933135747909546, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.2108, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 1.5805060863494873, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2025, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7044673539518901, | |
| "grad_norm": 1.3695785999298096, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.1826, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 1.6005311012268066, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.2081, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.711340206185567, | |
| "grad_norm": 2.1187644004821777, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.1892, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 1.6073172092437744, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.2135, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.718213058419244, | |
| "grad_norm": 1.3255120515823364, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.1935, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 1.670796275138855, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.217, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7250859106529209, | |
| "grad_norm": 1.5352182388305664, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.2092, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 1.4349397420883179, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.1905, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7319587628865979, | |
| "grad_norm": 1.708699345588684, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.2018, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 1.4706976413726807, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.2004, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.738831615120275, | |
| "grad_norm": 1.793593406677246, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.2248, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 1.248917818069458, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.1693, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7457044673539519, | |
| "grad_norm": 1.5068672895431519, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.2044, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 1.4568971395492554, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.2046, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7525773195876289, | |
| "grad_norm": 1.6441450119018555, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.2039, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 1.586878776550293, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.1999, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7594501718213058, | |
| "grad_norm": 1.4591821432113647, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.1955, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 1.258337140083313, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.1889, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7663230240549829, | |
| "grad_norm": 1.4574586153030396, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.2064, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 1.4049776792526245, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.1849, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7731958762886598, | |
| "grad_norm": 1.3771268129348755, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.1912, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 1.778815507888794, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.2297, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7800687285223368, | |
| "grad_norm": 1.5148082971572876, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.2015, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 1.6724814176559448, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.205, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7869415807560137, | |
| "grad_norm": 1.5340983867645264, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.1985, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 1.5708106756210327, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.1761, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7938144329896907, | |
| "grad_norm": 1.4276490211486816, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.1698, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 1.535973072052002, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.1936, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8006872852233677, | |
| "grad_norm": 1.74392569065094, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.2167, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 1.5829018354415894, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.183, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8075601374570447, | |
| "grad_norm": 1.7776228189468384, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.1954, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 1.4006965160369873, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.1641, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8144329896907216, | |
| "grad_norm": 1.2554253339767456, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.1741, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 1.3889808654785156, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.1986, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8213058419243986, | |
| "grad_norm": 1.445722222328186, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.188, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 2.1535000801086426, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.1964, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8281786941580757, | |
| "grad_norm": 1.418982982635498, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.196, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 1.3374110460281372, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.1874, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8350515463917526, | |
| "grad_norm": 1.5425893068313599, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.1902, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 1.5052316188812256, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.2103, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8419243986254296, | |
| "grad_norm": 1.2446677684783936, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.1666, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 1.8422250747680664, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.2279, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8487972508591065, | |
| "grad_norm": 1.4875450134277344, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.1936, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 1.4318687915802002, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.197, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8556701030927835, | |
| "grad_norm": 1.6211766004562378, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.1985, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 1.6111860275268555, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.1801, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8625429553264605, | |
| "grad_norm": 1.6115210056304932, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.1917, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 1.3563942909240723, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.1807, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8694158075601375, | |
| "grad_norm": 1.3219175338745117, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.1726, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 1.3618080615997314, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.1843, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.8762886597938144, | |
| "grad_norm": 1.6135002374649048, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.1788, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 1.5762805938720703, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.1887, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8831615120274914, | |
| "grad_norm": 1.7021830081939697, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.1841, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 1.5030473470687866, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.1917, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8900343642611683, | |
| "grad_norm": 1.5852195024490356, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.1956, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 1.6304069757461548, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.178, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8969072164948454, | |
| "grad_norm": 1.4224854707717896, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.1859, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 1.54900324344635, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1968, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9037800687285223, | |
| "grad_norm": 1.5951098203659058, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.2092, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 1.3917310237884521, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.1881, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9106529209621993, | |
| "grad_norm": 1.3137767314910889, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.176, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 1.5390129089355469, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.2018, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9175257731958762, | |
| "grad_norm": 1.5604429244995117, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.1974, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 1.49399995803833, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.1795, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9243986254295533, | |
| "grad_norm": 1.4592257738113403, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.1817, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 1.6082309484481812, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.1695, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9312714776632303, | |
| "grad_norm": 1.3131675720214844, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.1708, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 1.4831839799880981, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.1852, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9381443298969072, | |
| "grad_norm": 1.5412657260894775, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.1862, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 1.8354051113128662, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.2186, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9450171821305842, | |
| "grad_norm": 1.4880577325820923, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.1822, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 1.5500534772872925, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.1758, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9518900343642611, | |
| "grad_norm": 1.6146880388259888, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.1838, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 1.1931922435760498, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.1727, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9587628865979382, | |
| "grad_norm": 1.5446773767471313, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.1989, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 1.5740249156951904, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.2063, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9656357388316151, | |
| "grad_norm": 1.3512248992919922, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.1854, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 1.6300441026687622, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.2039, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9725085910652921, | |
| "grad_norm": 1.831511378288269, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.2173, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 1.4836167097091675, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.1777, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.979381443298969, | |
| "grad_norm": 1.3979811668395996, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.1892, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 1.506360411643982, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.1947, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9862542955326461, | |
| "grad_norm": 1.698026180267334, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.2171, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 1.6240311861038208, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.1824, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.993127147766323, | |
| "grad_norm": 1.3279304504394531, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.1685, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 1.5155754089355469, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.1969, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.5006811618804932, | |
| "learning_rate": 0.0, | |
| "loss": 0.1773, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.37102481722831726, | |
| "eval_runtime": 3.4836, | |
| "eval_samples_per_second": 66.024, | |
| "eval_steps_per_second": 2.871, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 291, | |
| "total_flos": 1.0225750283334451e+17, | |
| "train_loss": 0.26723733032282276, | |
| "train_runtime": 929.7796, | |
| "train_samples_per_second": 14.978, | |
| "train_steps_per_second": 0.313 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 291, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 800, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0225750283334451e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |