| { | |
| "best_metric": 1.0741040706634521, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.04244181933932235, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00028294546226214896, | |
| "grad_norm": 0.18001686036586761, | |
| "learning_rate": 6e-06, | |
| "loss": 1.2994, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00028294546226214896, | |
| "eval_loss": 1.2716970443725586, | |
| "eval_runtime": 846.7814, | |
| "eval_samples_per_second": 7.03, | |
| "eval_steps_per_second": 1.758, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0005658909245242979, | |
| "grad_norm": 0.1731778383255005, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.3714, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0008488363867864469, | |
| "grad_norm": 0.16550840437412262, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.2408, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0011317818490485959, | |
| "grad_norm": 0.16617779433727264, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.398, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.001414727311310745, | |
| "grad_norm": 0.15455280244350433, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2861, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0016976727735728938, | |
| "grad_norm": 0.15682223439216614, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.293, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.001980618235835043, | |
| "grad_norm": 0.14510343968868256, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.2524, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0022635636980971917, | |
| "grad_norm": 0.14818604290485382, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.2562, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0025465091603593406, | |
| "grad_norm": 0.14594592154026031, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.3486, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00282945462262149, | |
| "grad_norm": 0.14386072754859924, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2519, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0031124000848836387, | |
| "grad_norm": 0.1916978359222412, | |
| "learning_rate": 5.999589914977407e-05, | |
| "loss": 1.1521, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0033953455471457876, | |
| "grad_norm": 0.16554145514965057, | |
| "learning_rate": 5.998359772022778e-05, | |
| "loss": 1.248, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0036782910094079364, | |
| "grad_norm": 0.1682235598564148, | |
| "learning_rate": 5.996309907444915e-05, | |
| "loss": 1.2722, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.003961236471670086, | |
| "grad_norm": 0.15518620610237122, | |
| "learning_rate": 5.9934408816563236e-05, | |
| "loss": 1.1769, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.004244181933932234, | |
| "grad_norm": 0.15147370100021362, | |
| "learning_rate": 5.98975347902001e-05, | |
| "loss": 1.3129, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0045271273961943834, | |
| "grad_norm": 0.15708550810813904, | |
| "learning_rate": 5.9852487076350345e-05, | |
| "loss": 1.2437, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.004810072858456533, | |
| "grad_norm": 0.16699694097042084, | |
| "learning_rate": 5.979927799060915e-05, | |
| "loss": 1.2656, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.005093018320718681, | |
| "grad_norm": 0.15669851005077362, | |
| "learning_rate": 5.9737922079809257e-05, | |
| "loss": 1.2197, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0053759637829808304, | |
| "grad_norm": 0.15826715528964996, | |
| "learning_rate": 5.9668436118044054e-05, | |
| "loss": 1.2848, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.00565890924524298, | |
| "grad_norm": 0.15940728783607483, | |
| "learning_rate": 5.959083910208167e-05, | |
| "loss": 1.2357, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.005941854707505128, | |
| "grad_norm": 0.14362244307994843, | |
| "learning_rate": 5.9505152246171474e-05, | |
| "loss": 1.1756, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0062248001697672774, | |
| "grad_norm": 0.1448201835155487, | |
| "learning_rate": 5.941139897624428e-05, | |
| "loss": 1.1705, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.006507745632029427, | |
| "grad_norm": 0.14286337792873383, | |
| "learning_rate": 5.9309604923507984e-05, | |
| "loss": 1.256, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.006790691094291575, | |
| "grad_norm": 0.14091116189956665, | |
| "learning_rate": 5.9199797917440176e-05, | |
| "loss": 1.2419, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0070736365565537244, | |
| "grad_norm": 0.1461900919675827, | |
| "learning_rate": 5.908200797817991e-05, | |
| "loss": 1.3056, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.007356582018815873, | |
| "grad_norm": 0.1513061672449112, | |
| "learning_rate": 5.895626730832046e-05, | |
| "loss": 1.2286, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.007639527481078022, | |
| "grad_norm": 0.14222566783428192, | |
| "learning_rate": 5.882261028410545e-05, | |
| "loss": 1.3284, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.007922472943340171, | |
| "grad_norm": 0.14666450023651123, | |
| "learning_rate": 5.8681073446030734e-05, | |
| "loss": 1.1703, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.00820541840560232, | |
| "grad_norm": 0.1654808074235916, | |
| "learning_rate": 5.853169548885461e-05, | |
| "loss": 1.1861, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.008488363867864468, | |
| "grad_norm": 0.1394944190979004, | |
| "learning_rate": 5.8374517251019035e-05, | |
| "loss": 1.2514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.008771309330126618, | |
| "grad_norm": 0.16221028566360474, | |
| "learning_rate": 5.820958170348484e-05, | |
| "loss": 1.2141, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.009054254792388767, | |
| "grad_norm": 0.17233219742774963, | |
| "learning_rate": 5.8036933937983825e-05, | |
| "loss": 1.1461, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.009337200254650915, | |
| "grad_norm": 0.16100017726421356, | |
| "learning_rate": 5.7856621154691217e-05, | |
| "loss": 1.1342, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.009620145716913065, | |
| "grad_norm": 0.16232934594154358, | |
| "learning_rate": 5.766869264932154e-05, | |
| "loss": 1.0513, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.009903091179175214, | |
| "grad_norm": 0.169218510389328, | |
| "learning_rate": 5.747319979965172e-05, | |
| "loss": 1.176, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.010186036641437362, | |
| "grad_norm": 0.18459808826446533, | |
| "learning_rate": 5.727019605147488e-05, | |
| "loss": 1.1991, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.010468982103699512, | |
| "grad_norm": 0.21113476157188416, | |
| "learning_rate": 5.7059736903988775e-05, | |
| "loss": 1.1908, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.010751927565961661, | |
| "grad_norm": 0.255100816488266, | |
| "learning_rate": 5.684187989462291e-05, | |
| "loss": 1.0502, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01103487302822381, | |
| "grad_norm": 0.2532814145088196, | |
| "learning_rate": 5.661668458330836e-05, | |
| "loss": 1.1723, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01131781849048596, | |
| "grad_norm": 0.24460916221141815, | |
| "learning_rate": 5.638421253619467e-05, | |
| "loss": 1.092, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.011600763952748108, | |
| "grad_norm": 0.33700716495513916, | |
| "learning_rate": 5.614452730881832e-05, | |
| "loss": 1.0077, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.011883709415010256, | |
| "grad_norm": 0.3086978793144226, | |
| "learning_rate": 5.589769442872722e-05, | |
| "loss": 0.984, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.012166654877272406, | |
| "grad_norm": 0.34127745032310486, | |
| "learning_rate": 5.5643781377566175e-05, | |
| "loss": 0.9426, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.012449600339534555, | |
| "grad_norm": 0.3650180697441101, | |
| "learning_rate": 5.538285757262806e-05, | |
| "loss": 0.9358, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.012732545801796703, | |
| "grad_norm": 0.3661767542362213, | |
| "learning_rate": 5.5114994347875856e-05, | |
| "loss": 0.9236, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.013015491264058853, | |
| "grad_norm": 0.35510310530662537, | |
| "learning_rate": 5.48402649344406e-05, | |
| "loss": 0.9134, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.013298436726321002, | |
| "grad_norm": 0.3974863588809967, | |
| "learning_rate": 5.455874444060078e-05, | |
| "loss": 1.1377, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01358138218858315, | |
| "grad_norm": 0.3589693009853363, | |
| "learning_rate": 5.427050983124843e-05, | |
| "loss": 1.2237, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.013864327650845299, | |
| "grad_norm": 0.36883828043937683, | |
| "learning_rate": 5.397563990684774e-05, | |
| "loss": 1.1407, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.014147273113107449, | |
| "grad_norm": 0.3884880840778351, | |
| "learning_rate": 5.367421528189181e-05, | |
| "loss": 1.1999, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.014147273113107449, | |
| "eval_loss": 1.1299543380737305, | |
| "eval_runtime": 853.0261, | |
| "eval_samples_per_second": 6.979, | |
| "eval_steps_per_second": 1.746, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.014430218575369597, | |
| "grad_norm": 0.2056826651096344, | |
| "learning_rate": 5.336631836286338e-05, | |
| "loss": 1.2011, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.014713164037631746, | |
| "grad_norm": 0.19640228152275085, | |
| "learning_rate": 5.3052033325705774e-05, | |
| "loss": 1.1475, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.014996109499893896, | |
| "grad_norm": 0.19956982135772705, | |
| "learning_rate": 5.2731446092810044e-05, | |
| "loss": 1.1226, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.015279054962156044, | |
| "grad_norm": 0.15548309683799744, | |
| "learning_rate": 5.240464430952462e-05, | |
| "loss": 1.1501, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.015562000424418193, | |
| "grad_norm": 0.1360340118408203, | |
| "learning_rate": 5.207171732019395e-05, | |
| "loss": 1.1462, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.015844945886680343, | |
| "grad_norm": 0.14059096574783325, | |
| "learning_rate": 5.1732756143732675e-05, | |
| "loss": 1.1009, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01612789134894249, | |
| "grad_norm": 0.13691364228725433, | |
| "learning_rate": 5.1387853448741916e-05, | |
| "loss": 1.1818, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01641083681120464, | |
| "grad_norm": 0.13487009704113007, | |
| "learning_rate": 5.103710352817465e-05, | |
| "loss": 1.1531, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.016693782273466788, | |
| "grad_norm": 0.15809845924377441, | |
| "learning_rate": 5.068060227355698e-05, | |
| "loss": 1.1472, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.016976727735728937, | |
| "grad_norm": 0.15477801859378815, | |
| "learning_rate": 5.0318447148772234e-05, | |
| "loss": 1.1809, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01725967319799109, | |
| "grad_norm": 0.1484566628932953, | |
| "learning_rate": 4.995073716341545e-05, | |
| "loss": 1.2081, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.017542618660253237, | |
| "grad_norm": 0.1421346217393875, | |
| "learning_rate": 4.957757284572506e-05, | |
| "loss": 1.224, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.017825564122515385, | |
| "grad_norm": 0.1333300918340683, | |
| "learning_rate": 4.91990562150995e-05, | |
| "loss": 1.1276, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.018108509584777534, | |
| "grad_norm": 0.12150599807500839, | |
| "learning_rate": 4.881529075420611e-05, | |
| "loss": 1.2021, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.018391455047039682, | |
| "grad_norm": 0.12204714864492416, | |
| "learning_rate": 4.8426381380690036e-05, | |
| "loss": 1.114, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01867440050930183, | |
| "grad_norm": 0.12727297842502594, | |
| "learning_rate": 4.8032434418490753e-05, | |
| "loss": 1.1536, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.018957345971563982, | |
| "grad_norm": 0.11255544424057007, | |
| "learning_rate": 4.7633557568774194e-05, | |
| "loss": 1.1206, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01924029143382613, | |
| "grad_norm": 0.1191386803984642, | |
| "learning_rate": 4.722985988048831e-05, | |
| "loss": 1.1546, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01952323689608828, | |
| "grad_norm": 0.13010723888874054, | |
| "learning_rate": 4.6821451720550184e-05, | |
| "loss": 1.1694, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.019806182358350428, | |
| "grad_norm": 0.12365347146987915, | |
| "learning_rate": 4.640844474367282e-05, | |
| "loss": 1.136, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.020089127820612576, | |
| "grad_norm": 0.14611800014972687, | |
| "learning_rate": 4.5990951861839815e-05, | |
| "loss": 1.1683, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.020372073282874725, | |
| "grad_norm": 0.1299329400062561, | |
| "learning_rate": 4.5569087213436455e-05, | |
| "loss": 1.2127, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.020655018745136876, | |
| "grad_norm": 0.1349126100540161, | |
| "learning_rate": 4.514296613204532e-05, | |
| "loss": 1.1545, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.020937964207399025, | |
| "grad_norm": 0.12816999852657318, | |
| "learning_rate": 4.471270511491525e-05, | |
| "loss": 1.198, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.021220909669661173, | |
| "grad_norm": 0.14442527294158936, | |
| "learning_rate": 4.427842179111221e-05, | |
| "loss": 1.1331, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.021503855131923322, | |
| "grad_norm": 0.1318608820438385, | |
| "learning_rate": 4.3840234889360634e-05, | |
| "loss": 1.1442, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.02178680059418547, | |
| "grad_norm": 0.13773271441459656, | |
| "learning_rate": 4.33982642055842e-05, | |
| "loss": 1.1503, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.02206974605644762, | |
| "grad_norm": 0.12418694794178009, | |
| "learning_rate": 4.2952630570154785e-05, | |
| "loss": 1.0393, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.022352691518709767, | |
| "grad_norm": 0.14450491964817047, | |
| "learning_rate": 4.250345581485871e-05, | |
| "loss": 1.1194, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.02263563698097192, | |
| "grad_norm": 0.15314346551895142, | |
| "learning_rate": 4.205086273958909e-05, | |
| "loss": 1.0763, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.022918582443234067, | |
| "grad_norm": 0.1480676680803299, | |
| "learning_rate": 4.1594975078773565e-05, | |
| "loss": 1.2132, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.023201527905496216, | |
| "grad_norm": 0.14896009862422943, | |
| "learning_rate": 4.113591746754662e-05, | |
| "loss": 1.1389, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.023484473367758364, | |
| "grad_norm": 0.17366276681423187, | |
| "learning_rate": 4.06738154076755e-05, | |
| "loss": 1.2691, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.023767418830020513, | |
| "grad_norm": 0.16257454454898834, | |
| "learning_rate": 4.020879523324929e-05, | |
| "loss": 1.0568, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.02405036429228266, | |
| "grad_norm": 0.16570693254470825, | |
| "learning_rate": 3.974098407614051e-05, | |
| "loss": 1.1422, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.024333309754544813, | |
| "grad_norm": 0.1792537122964859, | |
| "learning_rate": 3.927050983124842e-05, | |
| "loss": 1.1421, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.02461625521680696, | |
| "grad_norm": 0.18916811048984528, | |
| "learning_rate": 3.8797501121533946e-05, | |
| "loss": 1.1065, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.02489920067906911, | |
| "grad_norm": 0.2118011862039566, | |
| "learning_rate": 3.832208726285534e-05, | |
| "loss": 1.0761, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.025182146141331258, | |
| "grad_norm": 0.21735292673110962, | |
| "learning_rate": 3.784439822861459e-05, | |
| "loss": 0.8525, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.025465091603593407, | |
| "grad_norm": 0.23854799568653107, | |
| "learning_rate": 3.7364564614223976e-05, | |
| "loss": 0.9597, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.025748037065855555, | |
| "grad_norm": 0.23937970399856567, | |
| "learning_rate": 3.688271760140255e-05, | |
| "loss": 0.8657, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.026030982528117707, | |
| "grad_norm": 0.26307588815689087, | |
| "learning_rate": 3.6398988922312406e-05, | |
| "loss": 0.953, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.026313927990379855, | |
| "grad_norm": 0.27083978056907654, | |
| "learning_rate": 3.591351082354441e-05, | |
| "loss": 0.7971, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.026596873452642004, | |
| "grad_norm": 0.2805502116680145, | |
| "learning_rate": 3.54264160299633e-05, | |
| "loss": 0.9744, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.026879818914904152, | |
| "grad_norm": 0.35504117608070374, | |
| "learning_rate": 3.493783770842202e-05, | |
| "loss": 0.9262, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.0271627643771663, | |
| "grad_norm": 0.33214688301086426, | |
| "learning_rate": 3.444790943135526e-05, | |
| "loss": 1.0821, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02744570983942845, | |
| "grad_norm": 0.3750684857368469, | |
| "learning_rate": 3.3956765140262074e-05, | |
| "loss": 0.8112, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.027728655301690597, | |
| "grad_norm": 0.3210698962211609, | |
| "learning_rate": 3.346453910908759e-05, | |
| "loss": 1.004, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02801160076395275, | |
| "grad_norm": 0.3541051745414734, | |
| "learning_rate": 3.297136590751389e-05, | |
| "loss": 1.16, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.028294546226214898, | |
| "grad_norm": 0.36050865054130554, | |
| "learning_rate": 3.247738036416998e-05, | |
| "loss": 1.1555, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.028294546226214898, | |
| "eval_loss": 1.0897471904754639, | |
| "eval_runtime": 853.3729, | |
| "eval_samples_per_second": 6.976, | |
| "eval_steps_per_second": 1.745, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.028577491688477046, | |
| "grad_norm": 0.1543198525905609, | |
| "learning_rate": 3.1982717529770985e-05, | |
| "loss": 1.2425, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.028860437150739195, | |
| "grad_norm": 0.19346995651721954, | |
| "learning_rate": 3.148751264019667e-05, | |
| "loss": 1.349, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.029143382613001343, | |
| "grad_norm": 0.1380472481250763, | |
| "learning_rate": 3.099190107951924e-05, | |
| "loss": 1.1654, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02942632807526349, | |
| "grad_norm": 0.1306401491165161, | |
| "learning_rate": 3.049601834299076e-05, | |
| "loss": 1.1968, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.029709273537525643, | |
| "grad_norm": 0.14598701894283295, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1663, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.029992218999787792, | |
| "grad_norm": 0.14115272462368011, | |
| "learning_rate": 2.9503981657009246e-05, | |
| "loss": 1.212, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.03027516446204994, | |
| "grad_norm": 0.13328498601913452, | |
| "learning_rate": 2.9008098920480752e-05, | |
| "loss": 1.1183, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.03055810992431209, | |
| "grad_norm": 0.12777693569660187, | |
| "learning_rate": 2.851248735980333e-05, | |
| "loss": 1.113, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.030841055386574237, | |
| "grad_norm": 0.1404535323381424, | |
| "learning_rate": 2.801728247022902e-05, | |
| "loss": 1.128, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.031124000848836385, | |
| "grad_norm": 0.11834380775690079, | |
| "learning_rate": 2.7522619635830034e-05, | |
| "loss": 1.166, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.031406946311098534, | |
| "grad_norm": 0.14574697613716125, | |
| "learning_rate": 2.702863409248612e-05, | |
| "loss": 1.1504, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.031689891773360686, | |
| "grad_norm": 0.13284307718276978, | |
| "learning_rate": 2.6535460890912416e-05, | |
| "loss": 1.1715, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.03197283723562283, | |
| "grad_norm": 0.128324955701828, | |
| "learning_rate": 2.604323485973793e-05, | |
| "loss": 1.0557, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.03225578269788498, | |
| "grad_norm": 0.13001534342765808, | |
| "learning_rate": 2.555209056864474e-05, | |
| "loss": 1.2398, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.032538728160147135, | |
| "grad_norm": 0.13345256447792053, | |
| "learning_rate": 2.5062162291577978e-05, | |
| "loss": 1.2067, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.03282167362240928, | |
| "grad_norm": 0.12472221255302429, | |
| "learning_rate": 2.4573583970036712e-05, | |
| "loss": 1.1886, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.03310461908467143, | |
| "grad_norm": 0.14320768415927887, | |
| "learning_rate": 2.4086489176455595e-05, | |
| "loss": 1.2183, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.033387564546933576, | |
| "grad_norm": 0.1415589302778244, | |
| "learning_rate": 2.36010110776876e-05, | |
| "loss": 1.2193, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.03367051000919573, | |
| "grad_norm": 0.14643894135951996, | |
| "learning_rate": 2.3117282398597456e-05, | |
| "loss": 1.2352, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.03395345547145787, | |
| "grad_norm": 0.1440449059009552, | |
| "learning_rate": 2.263543538577603e-05, | |
| "loss": 1.1977, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.034236400933720025, | |
| "grad_norm": 0.14856146275997162, | |
| "learning_rate": 2.215560177138541e-05, | |
| "loss": 1.1351, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.03451934639598218, | |
| "grad_norm": 0.13947530090808868, | |
| "learning_rate": 2.167791273714467e-05, | |
| "loss": 1.169, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.03480229185824432, | |
| "grad_norm": 0.13417986035346985, | |
| "learning_rate": 2.1202498878466062e-05, | |
| "loss": 1.1483, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.035085237320506474, | |
| "grad_norm": 0.13690292835235596, | |
| "learning_rate": 2.072949016875158e-05, | |
| "loss": 1.0759, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.03536818278276862, | |
| "grad_norm": 0.1381928026676178, | |
| "learning_rate": 2.0259015923859498e-05, | |
| "loss": 1.2043, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03565112824503077, | |
| "grad_norm": 0.14060895144939423, | |
| "learning_rate": 1.979120476675071e-05, | |
| "loss": 1.1919, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.03593407370729292, | |
| "grad_norm": 0.12901578843593597, | |
| "learning_rate": 1.9326184592324503e-05, | |
| "loss": 1.0793, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03621701916955507, | |
| "grad_norm": 0.1504661738872528, | |
| "learning_rate": 1.8864082532453373e-05, | |
| "loss": 1.2077, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.03649996463181722, | |
| "grad_norm": 0.16385625302791595, | |
| "learning_rate": 1.840502492122644e-05, | |
| "loss": 1.1345, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.036782910094079364, | |
| "grad_norm": 0.16052238643169403, | |
| "learning_rate": 1.7949137260410924e-05, | |
| "loss": 1.1935, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.037065855556341516, | |
| "grad_norm": 0.16583813726902008, | |
| "learning_rate": 1.7496544185141295e-05, | |
| "loss": 1.0112, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.03734880101860366, | |
| "grad_norm": 0.17477497458457947, | |
| "learning_rate": 1.7047369429845216e-05, | |
| "loss": 1.0563, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.03763174648086581, | |
| "grad_norm": 0.17088453471660614, | |
| "learning_rate": 1.6601735794415806e-05, | |
| "loss": 1.0716, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.037914691943127965, | |
| "grad_norm": 0.18345321714878082, | |
| "learning_rate": 1.615976511063937e-05, | |
| "loss": 0.9948, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.03819763740539011, | |
| "grad_norm": 0.20441484451293945, | |
| "learning_rate": 1.5721578208887793e-05, | |
| "loss": 0.9666, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03848058286765226, | |
| "grad_norm": 0.1876918077468872, | |
| "learning_rate": 1.5287294885084766e-05, | |
| "loss": 0.9862, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.03876352832991441, | |
| "grad_norm": 0.22819191217422485, | |
| "learning_rate": 1.4857033867954697e-05, | |
| "loss": 1.1309, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.03904647379217656, | |
| "grad_norm": 0.2145112156867981, | |
| "learning_rate": 1.4430912786563554e-05, | |
| "loss": 0.9526, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.039329419254438704, | |
| "grad_norm": 0.21892738342285156, | |
| "learning_rate": 1.4009048138160195e-05, | |
| "loss": 1.0422, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.039612364716700856, | |
| "grad_norm": 0.23647260665893555, | |
| "learning_rate": 1.3591555256327199e-05, | |
| "loss": 0.9467, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03989531017896301, | |
| "grad_norm": 0.25753793120384216, | |
| "learning_rate": 1.3178548279449822e-05, | |
| "loss": 0.8659, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.04017825564122515, | |
| "grad_norm": 0.24101592600345612, | |
| "learning_rate": 1.2770140119511693e-05, | |
| "loss": 0.8691, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.040461201103487304, | |
| "grad_norm": 0.2642737329006195, | |
| "learning_rate": 1.2366442431225809e-05, | |
| "loss": 1.0043, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.04074414656574945, | |
| "grad_norm": 0.27606478333473206, | |
| "learning_rate": 1.1967565581509248e-05, | |
| "loss": 0.8247, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0410270920280116, | |
| "grad_norm": 0.27429643273353577, | |
| "learning_rate": 1.1573618619309965e-05, | |
| "loss": 0.8922, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.04131003749027375, | |
| "grad_norm": 0.29994338750839233, | |
| "learning_rate": 1.1184709245793889e-05, | |
| "loss": 1.0195, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0415929829525359, | |
| "grad_norm": 0.3414655029773712, | |
| "learning_rate": 1.0800943784900502e-05, | |
| "loss": 0.8346, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.04187592841479805, | |
| "grad_norm": 0.30294448137283325, | |
| "learning_rate": 1.042242715427494e-05, | |
| "loss": 1.1227, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.042158873877060195, | |
| "grad_norm": 0.2909342646598816, | |
| "learning_rate": 1.004926283658455e-05, | |
| "loss": 0.8504, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.04244181933932235, | |
| "grad_norm": 0.35827234387397766, | |
| "learning_rate": 9.681552851227774e-06, | |
| "loss": 1.1022, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04244181933932235, | |
| "eval_loss": 1.0741040706634521, | |
| "eval_runtime": 853.591, | |
| "eval_samples_per_second": 6.974, | |
| "eval_steps_per_second": 1.744, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.95489868899156e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |