| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 140, | |
| "global_step": 560, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0017857142857142857, | |
| "grad_norm": 0.14711466431617737, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6301, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0035714285714285713, | |
| "grad_norm": 0.13304685056209564, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6271, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005357142857142857, | |
| "grad_norm": 0.1233660876750946, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6548, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007142857142857143, | |
| "grad_norm": 0.1381577104330063, | |
| "learning_rate": 8e-05, | |
| "loss": 1.7853, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.008928571428571428, | |
| "grad_norm": 0.14668770134449005, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7507, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010714285714285714, | |
| "grad_norm": 0.168854758143425, | |
| "learning_rate": 0.00012, | |
| "loss": 1.9486, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0125, | |
| "grad_norm": 0.2166246473789215, | |
| "learning_rate": 0.00014, | |
| "loss": 1.9167, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014285714285714285, | |
| "grad_norm": 0.23894698917865753, | |
| "learning_rate": 0.00016, | |
| "loss": 1.8384, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01607142857142857, | |
| "grad_norm": 0.2704271674156189, | |
| "learning_rate": 0.00018, | |
| "loss": 1.8232, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 0.24295401573181152, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1559, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.019642857142857142, | |
| "grad_norm": 0.23109205067157745, | |
| "learning_rate": 0.00019999836866486503, | |
| "loss": 1.7082, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02142857142857143, | |
| "grad_norm": 0.205198273062706, | |
| "learning_rate": 0.00019999347471268516, | |
| "loss": 1.8417, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023214285714285715, | |
| "grad_norm": 0.2678169012069702, | |
| "learning_rate": 0.00019998531830313395, | |
| "loss": 1.9222, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 0.2563900053501129, | |
| "learning_rate": 0.0001999738997023281, | |
| "loss": 1.8931, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.026785714285714284, | |
| "grad_norm": 0.31391921639442444, | |
| "learning_rate": 0.00019995921928281894, | |
| "loss": 1.9403, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 0.23522017896175385, | |
| "learning_rate": 0.00019994127752358013, | |
| "loss": 1.8728, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.030357142857142857, | |
| "grad_norm": 0.2517368495464325, | |
| "learning_rate": 0.00019992007500999214, | |
| "loss": 1.9224, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03214285714285714, | |
| "grad_norm": 0.27111852169036865, | |
| "learning_rate": 0.00019989561243382312, | |
| "loss": 1.9046, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.033928571428571426, | |
| "grad_norm": 0.2616860568523407, | |
| "learning_rate": 0.00019986789059320615, | |
| "loss": 2.1173, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 0.23268847167491913, | |
| "learning_rate": 0.00019983691039261357, | |
| "loss": 1.8406, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0375, | |
| "grad_norm": 0.30090174078941345, | |
| "learning_rate": 0.00019980267284282717, | |
| "loss": 1.7514, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.039285714285714285, | |
| "grad_norm": 0.23314997553825378, | |
| "learning_rate": 0.00019976517906090529, | |
| "loss": 1.7518, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04107142857142857, | |
| "grad_norm": 0.2630898952484131, | |
| "learning_rate": 0.0001997244302701464, | |
| "loss": 1.7533, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04285714285714286, | |
| "grad_norm": 0.3203246593475342, | |
| "learning_rate": 0.00019968042780004917, | |
| "loss": 1.534, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.044642857142857144, | |
| "grad_norm": 0.48179084062576294, | |
| "learning_rate": 0.00019963317308626914, | |
| "loss": 1.4211, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04642857142857143, | |
| "grad_norm": 0.36420196294784546, | |
| "learning_rate": 0.0001995826676705718, | |
| "loss": 1.4275, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.048214285714285716, | |
| "grad_norm": 0.35330235958099365, | |
| "learning_rate": 0.00019952891320078236, | |
| "loss": 1.5612, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.27140146493911743, | |
| "learning_rate": 0.00019947191143073186, | |
| "loss": 1.6001, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05178571428571429, | |
| "grad_norm": 0.2987808883190155, | |
| "learning_rate": 0.00019941166422020014, | |
| "loss": 1.3943, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 0.3032172918319702, | |
| "learning_rate": 0.00019934817353485501, | |
| "loss": 1.4888, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.055357142857142855, | |
| "grad_norm": 0.6096982359886169, | |
| "learning_rate": 0.00019928144144618824, | |
| "loss": 1.7057, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 0.30788663029670715, | |
| "learning_rate": 0.0001992114701314478, | |
| "loss": 1.3929, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05892857142857143, | |
| "grad_norm": 0.33283260464668274, | |
| "learning_rate": 0.00019913826187356696, | |
| "loss": 1.5611, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.060714285714285714, | |
| "grad_norm": 0.3044683337211609, | |
| "learning_rate": 0.00019906181906108984, | |
| "loss": 1.2961, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.3300105929374695, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 1.3975, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06428571428571428, | |
| "grad_norm": 0.3126329481601715, | |
| "learning_rate": 0.00019889923985410576, | |
| "loss": 1.2759, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06607142857142857, | |
| "grad_norm": 0.33685699105262756, | |
| "learning_rate": 0.00019881310876402223, | |
| "loss": 1.162, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06785714285714285, | |
| "grad_norm": 0.3971424996852875, | |
| "learning_rate": 0.0001987237537280163, | |
| "loss": 1.21, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.06964285714285715, | |
| "grad_norm": 0.43255800008773804, | |
| "learning_rate": 0.00019863117766144806, | |
| "loss": 1.0186, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 0.4930873215198517, | |
| "learning_rate": 0.00019853538358476932, | |
| "loss": 1.0045, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07321428571428572, | |
| "grad_norm": 0.4585554599761963, | |
| "learning_rate": 0.00019843637462342497, | |
| "loss": 1.1077, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 0.4778697192668915, | |
| "learning_rate": 0.00019833415400775093, | |
| "loss": 1.0817, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07678571428571429, | |
| "grad_norm": 0.4595739245414734, | |
| "learning_rate": 0.0001982287250728689, | |
| "loss": 1.2805, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07857142857142857, | |
| "grad_norm": 0.5665370225906372, | |
| "learning_rate": 0.00019812009125857728, | |
| "loss": 0.9855, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08035714285714286, | |
| "grad_norm": 0.40399685502052307, | |
| "learning_rate": 0.00019800825610923934, | |
| "loss": 1.3264, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08214285714285714, | |
| "grad_norm": 0.43560901284217834, | |
| "learning_rate": 0.00019789322327366723, | |
| "loss": 1.2724, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08392857142857142, | |
| "grad_norm": 0.375598669052124, | |
| "learning_rate": 0.000197774996505003, | |
| "loss": 1.1722, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 0.4257534146308899, | |
| "learning_rate": 0.00019765357966059638, | |
| "loss": 1.3607, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0875, | |
| "grad_norm": 0.5437749028205872, | |
| "learning_rate": 0.0001975289767018786, | |
| "loss": 1.9812, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 0.6295837163925171, | |
| "learning_rate": 0.00019740119169423337, | |
| "loss": 2.1601, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09107142857142857, | |
| "grad_norm": 0.22341597080230713, | |
| "learning_rate": 0.00019727022880686412, | |
| "loss": 1.7096, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09285714285714286, | |
| "grad_norm": 0.26040709018707275, | |
| "learning_rate": 0.00019713609231265805, | |
| "loss": 1.733, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09464285714285714, | |
| "grad_norm": 0.2146587371826172, | |
| "learning_rate": 0.00019699878658804672, | |
| "loss": 1.7999, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09642857142857143, | |
| "grad_norm": 0.2316572070121765, | |
| "learning_rate": 0.0001968583161128631, | |
| "loss": 1.6017, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.09821428571428571, | |
| "grad_norm": 0.20329022407531738, | |
| "learning_rate": 0.00019671468547019573, | |
| "loss": 1.7464, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.22198128700256348, | |
| "learning_rate": 0.00019656789934623881, | |
| "loss": 1.7584, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10178571428571428, | |
| "grad_norm": 0.2108459621667862, | |
| "learning_rate": 0.00019641796253013958, | |
| "loss": 1.7389, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10357142857142858, | |
| "grad_norm": 0.2677428126335144, | |
| "learning_rate": 0.00019626487991384196, | |
| "loss": 1.9166, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10535714285714286, | |
| "grad_norm": 0.22205914556980133, | |
| "learning_rate": 0.00019610865649192697, | |
| "loss": 1.738, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 0.22724869847297668, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 1.7451, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10892857142857143, | |
| "grad_norm": 0.24029673635959625, | |
| "learning_rate": 0.00019578680772177327, | |
| "loss": 1.9556, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11071428571428571, | |
| "grad_norm": 0.2240082174539566, | |
| "learning_rate": 0.00019562119287439873, | |
| "loss": 1.6435, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1125, | |
| "grad_norm": 0.23900961875915527, | |
| "learning_rate": 0.00019545245822279243, | |
| "loss": 1.7015, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 0.2319100946187973, | |
| "learning_rate": 0.0001952806092722098, | |
| "loss": 1.7287, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11607142857142858, | |
| "grad_norm": 0.22805678844451904, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 1.7574, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11785714285714285, | |
| "grad_norm": 0.2389436662197113, | |
| "learning_rate": 0.00019492759100300019, | |
| "loss": 1.7738, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.11964285714285715, | |
| "grad_norm": 0.23373426496982574, | |
| "learning_rate": 0.00019474643320219532, | |
| "loss": 1.5328, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12142857142857143, | |
| "grad_norm": 0.2664267420768738, | |
| "learning_rate": 0.0001945621841376825, | |
| "loss": 2.0273, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12321428571428572, | |
| "grad_norm": 0.2796079218387604, | |
| "learning_rate": 0.0001943748498209012, | |
| "loss": 1.547, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.2695769667625427, | |
| "learning_rate": 0.00019418443636395248, | |
| "loss": 1.6374, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12678571428571428, | |
| "grad_norm": 0.3933340311050415, | |
| "learning_rate": 0.00019399094997939957, | |
| "loss": 1.2037, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.12857142857142856, | |
| "grad_norm": 0.2825562357902527, | |
| "learning_rate": 0.0001937943969800652, | |
| "loss": 1.5649, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13035714285714287, | |
| "grad_norm": 0.3704264760017395, | |
| "learning_rate": 0.00019359478377882567, | |
| "loss": 1.292, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13214285714285715, | |
| "grad_norm": 0.3214375972747803, | |
| "learning_rate": 0.00019339211688840157, | |
| "loss": 1.2539, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13392857142857142, | |
| "grad_norm": 0.337907612323761, | |
| "learning_rate": 0.00019318640292114524, | |
| "loss": 1.3202, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1357142857142857, | |
| "grad_norm": 0.30557796359062195, | |
| "learning_rate": 0.00019297764858882514, | |
| "loss": 1.074, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1375, | |
| "grad_norm": 0.3154335916042328, | |
| "learning_rate": 0.00019276586070240682, | |
| "loss": 1.1813, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1392857142857143, | |
| "grad_norm": 0.3587944209575653, | |
| "learning_rate": 0.0001925510461718307, | |
| "loss": 1.3602, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14107142857142857, | |
| "grad_norm": 0.35669559240341187, | |
| "learning_rate": 0.0001923332120057866, | |
| "loss": 1.3626, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.48954322934150696, | |
| "learning_rate": 0.000192112365311485, | |
| "loss": 1.4762, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14464285714285716, | |
| "grad_norm": 0.4903950095176697, | |
| "learning_rate": 0.00019188851329442547, | |
| "loss": 1.6974, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14642857142857144, | |
| "grad_norm": 0.35062992572784424, | |
| "learning_rate": 0.00019166166325816118, | |
| "loss": 1.2373, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14821428571428572, | |
| "grad_norm": 0.5671891570091248, | |
| "learning_rate": 0.0001914318226040608, | |
| "loss": 1.5644, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.338214248418808, | |
| "learning_rate": 0.000191198998831067, | |
| "loss": 1.3717, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15178571428571427, | |
| "grad_norm": 0.4187781512737274, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 1.1211, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15357142857142858, | |
| "grad_norm": 0.3229452967643738, | |
| "learning_rate": 0.00019072443241056883, | |
| "loss": 1.0473, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15535714285714286, | |
| "grad_norm": 0.376589298248291, | |
| "learning_rate": 0.00019048270524660196, | |
| "loss": 0.9684, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15714285714285714, | |
| "grad_norm": 0.4261431396007538, | |
| "learning_rate": 0.00019023802593031154, | |
| "loss": 1.0243, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.15892857142857142, | |
| "grad_norm": 0.4411628544330597, | |
| "learning_rate": 0.0001899904024447769, | |
| "loss": 1.2739, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 0.3415679335594177, | |
| "learning_rate": 0.00018973984286913584, | |
| "loss": 1.2494, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1625, | |
| "grad_norm": 0.48547816276550293, | |
| "learning_rate": 0.0001894863553783212, | |
| "loss": 1.0384, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16428571428571428, | |
| "grad_norm": 0.398359090089798, | |
| "learning_rate": 0.00018922994824279395, | |
| "loss": 1.2409, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16607142857142856, | |
| "grad_norm": 0.4025682806968689, | |
| "learning_rate": 0.00018897062982827344, | |
| "loss": 1.0646, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16785714285714284, | |
| "grad_norm": 0.46782195568084717, | |
| "learning_rate": 0.00018870840859546456, | |
| "loss": 1.2167, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.16964285714285715, | |
| "grad_norm": 0.45423248410224915, | |
| "learning_rate": 0.00018844329309978145, | |
| "loss": 1.2359, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 0.380495548248291, | |
| "learning_rate": 0.0001881752919910686, | |
| "loss": 1.0583, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1732142857142857, | |
| "grad_norm": 0.41103506088256836, | |
| "learning_rate": 0.00018790441401331847, | |
| "loss": 1.2755, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 0.4458900988101959, | |
| "learning_rate": 0.00018763066800438636, | |
| "loss": 1.3651, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1767857142857143, | |
| "grad_norm": 0.5115835666656494, | |
| "learning_rate": 0.00018735406289570192, | |
| "loss": 1.5389, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 0.6476715803146362, | |
| "learning_rate": 0.00018707460771197774, | |
| "loss": 2.1173, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18035714285714285, | |
| "grad_norm": 0.18222245573997498, | |
| "learning_rate": 0.00018679231157091506, | |
| "loss": 1.3799, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.18214285714285713, | |
| "grad_norm": 0.2210777848958969, | |
| "learning_rate": 0.0001865071836829061, | |
| "loss": 1.6362, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.18392857142857144, | |
| "grad_norm": 0.21902061998844147, | |
| "learning_rate": 0.00018621923335073376, | |
| "loss": 1.6673, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.18571428571428572, | |
| "grad_norm": 0.21758711338043213, | |
| "learning_rate": 0.00018592846996926793, | |
| "loss": 1.669, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.21447710692882538, | |
| "learning_rate": 0.0001856349030251589, | |
| "loss": 1.6173, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.18928571428571428, | |
| "grad_norm": 0.20819346606731415, | |
| "learning_rate": 0.00018533854209652818, | |
| "loss": 1.6841, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.19107142857142856, | |
| "grad_norm": 0.19584102928638458, | |
| "learning_rate": 0.00018503939685265568, | |
| "loss": 1.5907, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.19285714285714287, | |
| "grad_norm": 0.2108387053012848, | |
| "learning_rate": 0.00018473747705366426, | |
| "loss": 1.7855, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.19464285714285715, | |
| "grad_norm": 0.20142777264118195, | |
| "learning_rate": 0.00018443279255020152, | |
| "loss": 1.79, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 0.20290708541870117, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 1.8891, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1982142857142857, | |
| "grad_norm": 0.20743344724178314, | |
| "learning_rate": 0.00018381516928314367, | |
| "loss": 1.6483, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.24931779503822327, | |
| "learning_rate": 0.00018350225067055925, | |
| "loss": 1.7883, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2017857142857143, | |
| "grad_norm": 0.20271241664886475, | |
| "learning_rate": 0.00018318660765486748, | |
| "loss": 1.8015, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20357142857142857, | |
| "grad_norm": 0.21297615766525269, | |
| "learning_rate": 0.00018286825053445918, | |
| "loss": 1.7595, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20535714285714285, | |
| "grad_norm": 0.21898160874843597, | |
| "learning_rate": 0.0001825471896962774, | |
| "loss": 1.8246, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.20714285714285716, | |
| "grad_norm": 0.2433815449476242, | |
| "learning_rate": 0.00018222343561547874, | |
| "loss": 1.9214, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.20892857142857144, | |
| "grad_norm": 0.2514473497867584, | |
| "learning_rate": 0.00018189699885509127, | |
| "loss": 1.674, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21071428571428572, | |
| "grad_norm": 0.23736989498138428, | |
| "learning_rate": 0.0001815678900656702, | |
| "loss": 1.7948, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2125, | |
| "grad_norm": 0.22837956249713898, | |
| "learning_rate": 0.00018123611998495007, | |
| "loss": 1.9429, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 0.2722070813179016, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 1.2768, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21607142857142858, | |
| "grad_norm": 0.23845739662647247, | |
| "learning_rate": 0.00018056463933434398, | |
| "loss": 1.7407, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.21785714285714286, | |
| "grad_norm": 0.24140624701976776, | |
| "learning_rate": 0.00018022495067265753, | |
| "loss": 1.378, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.21964285714285714, | |
| "grad_norm": 0.2878568768501282, | |
| "learning_rate": 0.0001798826445353564, | |
| "loss": 1.6684, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.22142857142857142, | |
| "grad_norm": 0.35590338706970215, | |
| "learning_rate": 0.0001795377320907611, | |
| "loss": 0.7946, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22321428571428573, | |
| "grad_norm": 0.35154449939727783, | |
| "learning_rate": 0.00017919022459222752, | |
| "loss": 0.9185, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.31363779306411743, | |
| "learning_rate": 0.00017884013337777943, | |
| "loss": 1.3471, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.22678571428571428, | |
| "grad_norm": 0.36924973130226135, | |
| "learning_rate": 0.00017848746986973883, | |
| "loss": 1.3852, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 0.44929584860801697, | |
| "learning_rate": 0.00017813224557435312, | |
| "loss": 0.82, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23035714285714284, | |
| "grad_norm": 0.3592383563518524, | |
| "learning_rate": 0.0001777744720814198, | |
| "loss": 1.0219, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 0.37820902466773987, | |
| "learning_rate": 0.00017741416106390826, | |
| "loss": 1.3303, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23392857142857143, | |
| "grad_norm": 0.33376428484916687, | |
| "learning_rate": 0.00017705132427757895, | |
| "loss": 1.384, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2357142857142857, | |
| "grad_norm": 0.3454584777355194, | |
| "learning_rate": 0.00017668597356059978, | |
| "loss": 1.2236, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2375, | |
| "grad_norm": 0.345708429813385, | |
| "learning_rate": 0.00017631812083316003, | |
| "loss": 1.0293, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2392857142857143, | |
| "grad_norm": 0.3823675811290741, | |
| "learning_rate": 0.00017594777809708126, | |
| "loss": 1.3131, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24107142857142858, | |
| "grad_norm": 0.336627334356308, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 1.2084, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24285714285714285, | |
| "grad_norm": 0.3525237739086151, | |
| "learning_rate": 0.0001751996710121026, | |
| "loss": 1.1378, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24464285714285713, | |
| "grad_norm": 0.38291576504707336, | |
| "learning_rate": 0.00017482193107147014, | |
| "loss": 0.8353, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.24642857142857144, | |
| "grad_norm": 0.4650520086288452, | |
| "learning_rate": 0.0001744417499379372, | |
| "loss": 1.2024, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.24821428571428572, | |
| "grad_norm": 0.32640358805656433, | |
| "learning_rate": 0.0001740591400155606, | |
| "loss": 1.2343, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.39074593782424927, | |
| "learning_rate": 0.0001736741137876405, | |
| "loss": 0.8896, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.4710056781768799, | |
| "eval_runtime": 13.3678, | |
| "eval_samples_per_second": 17.654, | |
| "eval_steps_per_second": 8.827, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2517857142857143, | |
| "grad_norm": 0.36249810457229614, | |
| "learning_rate": 0.00017328668381631318, | |
| "loss": 1.1208, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.25357142857142856, | |
| "grad_norm": 0.3750612735748291, | |
| "learning_rate": 0.00017289686274214118, | |
| "loss": 1.2502, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.25535714285714284, | |
| "grad_norm": 0.4201869070529938, | |
| "learning_rate": 0.0001725046632837007, | |
| "loss": 1.1947, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 0.4865645468235016, | |
| "learning_rate": 0.00017211009823716694, | |
| "loss": 0.8749, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.25892857142857145, | |
| "grad_norm": 0.38693225383758545, | |
| "learning_rate": 0.00017171318047589637, | |
| "loss": 1.2495, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26071428571428573, | |
| "grad_norm": 0.40707525610923767, | |
| "learning_rate": 0.00017131392295000674, | |
| "loss": 1.2321, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2625, | |
| "grad_norm": 0.39570894837379456, | |
| "learning_rate": 0.00017091233868595467, | |
| "loss": 1.301, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2642857142857143, | |
| "grad_norm": 0.4085226058959961, | |
| "learning_rate": 0.00017050844078611056, | |
| "loss": 1.5369, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.26607142857142857, | |
| "grad_norm": 0.47094810009002686, | |
| "learning_rate": 0.0001701022424283311, | |
| "loss": 1.9374, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 0.8517308831214905, | |
| "learning_rate": 0.00016969375686552937, | |
| "loss": 1.808, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26964285714285713, | |
| "grad_norm": 0.1922745406627655, | |
| "learning_rate": 0.00016928299742524234, | |
| "loss": 1.6608, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2714285714285714, | |
| "grad_norm": 0.2090916484594345, | |
| "learning_rate": 0.00016886997750919619, | |
| "loss": 1.8009, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2732142857142857, | |
| "grad_norm": 0.21698515117168427, | |
| "learning_rate": 0.00016845471059286887, | |
| "loss": 1.7821, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.21791532635688782, | |
| "learning_rate": 0.00016803721022505067, | |
| "loss": 1.5901, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2767857142857143, | |
| "grad_norm": 0.22199980914592743, | |
| "learning_rate": 0.00016761749002740193, | |
| "loss": 1.7047, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2785714285714286, | |
| "grad_norm": 0.2096625566482544, | |
| "learning_rate": 0.0001671955636940088, | |
| "loss": 1.6898, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.28035714285714286, | |
| "grad_norm": 0.22975414991378784, | |
| "learning_rate": 0.00016677144499093626, | |
| "loss": 1.7631, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28214285714285714, | |
| "grad_norm": 0.2187148928642273, | |
| "learning_rate": 0.0001663451477557792, | |
| "loss": 1.7872, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2839285714285714, | |
| "grad_norm": 0.2257414609193802, | |
| "learning_rate": 0.0001659166858972107, | |
| "loss": 1.7732, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.22986693680286407, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 1.7031, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2875, | |
| "grad_norm": 0.21585014462471008, | |
| "learning_rate": 0.0001650533242971987, | |
| "loss": 1.8421, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2892857142857143, | |
| "grad_norm": 0.22519604861736298, | |
| "learning_rate": 0.00016461845272439741, | |
| "loss": 1.6529, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2910714285714286, | |
| "grad_norm": 0.22279705107212067, | |
| "learning_rate": 0.0001641814728645502, | |
| "loss": 1.9288, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.29285714285714287, | |
| "grad_norm": 0.22392615675926208, | |
| "learning_rate": 0.000163742398974869, | |
| "loss": 1.693, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.29464285714285715, | |
| "grad_norm": 0.22729454934597015, | |
| "learning_rate": 0.00016330124538088705, | |
| "loss": 1.7027, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.29642857142857143, | |
| "grad_norm": 0.2229882776737213, | |
| "learning_rate": 0.00016285802647599156, | |
| "loss": 1.8262, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2982142857142857, | |
| "grad_norm": 0.25520074367523193, | |
| "learning_rate": 0.00016241275672095395, | |
| "loss": 1.6009, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.24272315204143524, | |
| "learning_rate": 0.00016196545064345812, | |
| "loss": 1.9227, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.30178571428571427, | |
| "grad_norm": 0.24380216002464294, | |
| "learning_rate": 0.00016151612283762652, | |
| "loss": 1.5198, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 0.3242342472076416, | |
| "learning_rate": 0.00016106478796354382, | |
| "loss": 1.6981, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3053571428571429, | |
| "grad_norm": 0.277855783700943, | |
| "learning_rate": 0.00016061146074677885, | |
| "loss": 1.7011, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.30714285714285716, | |
| "grad_norm": 0.2710039019584656, | |
| "learning_rate": 0.00016015615597790388, | |
| "loss": 1.7522, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.30892857142857144, | |
| "grad_norm": 0.26541268825531006, | |
| "learning_rate": 0.00015969888851201226, | |
| "loss": 1.3804, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3107142857142857, | |
| "grad_norm": 0.28985923528671265, | |
| "learning_rate": 0.00015923967326823368, | |
| "loss": 1.6453, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.33939245343208313, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 1.1725, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 0.29770731925964355, | |
| "learning_rate": 0.0001583154594407932, | |
| "loss": 1.6746, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.31607142857142856, | |
| "grad_norm": 0.3280562460422516, | |
| "learning_rate": 0.0001578504910111811, | |
| "loss": 1.1357, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.31785714285714284, | |
| "grad_norm": 0.2856597304344177, | |
| "learning_rate": 0.00015738363511079776, | |
| "loss": 1.1127, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3196428571428571, | |
| "grad_norm": 0.316491961479187, | |
| "learning_rate": 0.00015691490697161182, | |
| "loss": 1.4281, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 0.3632654845714569, | |
| "learning_rate": 0.00015644432188667695, | |
| "loss": 1.3413, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32321428571428573, | |
| "grad_norm": 0.34329405426979065, | |
| "learning_rate": 0.00015597189520963277, | |
| "loss": 1.0579, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.32447105646133423, | |
| "learning_rate": 0.00015549764235420405, | |
| "loss": 1.243, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3267857142857143, | |
| "grad_norm": 0.3558500409126282, | |
| "learning_rate": 0.0001550215787936977, | |
| "loss": 1.1376, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.32857142857142857, | |
| "grad_norm": 0.3373570740222931, | |
| "learning_rate": 0.00015454372006049803, | |
| "loss": 1.1251, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.33035714285714285, | |
| "grad_norm": 0.36412546038627625, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 1.3238, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.33214285714285713, | |
| "grad_norm": 0.364442378282547, | |
| "learning_rate": 0.00015358267949789966, | |
| "loss": 0.9448, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3339285714285714, | |
| "grad_norm": 0.3172107934951782, | |
| "learning_rate": 0.00015309952902408576, | |
| "loss": 1.2744, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3357142857142857, | |
| "grad_norm": 0.34173399209976196, | |
| "learning_rate": 0.00015261464608772488, | |
| "loss": 1.0923, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.3375, | |
| "grad_norm": 0.33419185876846313, | |
| "learning_rate": 0.0001521280465089484, | |
| "loss": 1.2762, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 0.3866868317127228, | |
| "learning_rate": 0.0001516397461638962, | |
| "loss": 0.9595, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3410714285714286, | |
| "grad_norm": 0.3978990614414215, | |
| "learning_rate": 0.00015114976098419842, | |
| "loss": 0.9993, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.3546142876148224, | |
| "learning_rate": 0.00015065810695645584, | |
| "loss": 1.3421, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.34464285714285714, | |
| "grad_norm": 0.39728498458862305, | |
| "learning_rate": 0.00015016480012171828, | |
| "loss": 1.1209, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3464285714285714, | |
| "grad_norm": 0.4170741140842438, | |
| "learning_rate": 0.00014966985657496114, | |
| "loss": 1.0024, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3482142857142857, | |
| "grad_norm": 0.4226652681827545, | |
| "learning_rate": 0.0001491732924645604, | |
| "loss": 1.3139, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.3712114691734314, | |
| "learning_rate": 0.00014867512399176563, | |
| "loss": 1.1574, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3517857142857143, | |
| "grad_norm": 0.3655322790145874, | |
| "learning_rate": 0.00014817536741017152, | |
| "loss": 1.6149, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3535714285714286, | |
| "grad_norm": 0.4362059533596039, | |
| "learning_rate": 0.0001476740390251875, | |
| "loss": 1.7657, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.35535714285714287, | |
| "grad_norm": 0.43134769797325134, | |
| "learning_rate": 0.00014717115519350567, | |
| "loss": 1.7167, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.7784890532493591, | |
| "learning_rate": 0.00014666673232256738, | |
| "loss": 2.036, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.35892857142857143, | |
| "grad_norm": 0.17376984655857086, | |
| "learning_rate": 0.0001461607868700276, | |
| "loss": 1.4856, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3607142857142857, | |
| "grad_norm": 0.2141953408718109, | |
| "learning_rate": 0.00014565333534321826, | |
| "loss": 1.7491, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3625, | |
| "grad_norm": 0.22548137605190277, | |
| "learning_rate": 0.00014514439429860943, | |
| "loss": 1.8457, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.36428571428571427, | |
| "grad_norm": 0.20618294179439545, | |
| "learning_rate": 0.0001446339803412692, | |
| "loss": 1.4987, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.36607142857142855, | |
| "grad_norm": 0.21025151014328003, | |
| "learning_rate": 0.00014412211012432212, | |
| "loss": 1.5568, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3678571428571429, | |
| "grad_norm": 0.21678180992603302, | |
| "learning_rate": 0.00014360880034840554, | |
| "loss": 1.7841, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.36964285714285716, | |
| "grad_norm": 0.20914790034294128, | |
| "learning_rate": 0.0001430940677611249, | |
| "loss": 1.6693, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 0.21597585082054138, | |
| "learning_rate": 0.00014257792915650728, | |
| "loss": 1.648, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3732142857142857, | |
| "grad_norm": 0.23697789013385773, | |
| "learning_rate": 0.00014206040137445348, | |
| "loss": 1.7616, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.2535800635814667, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 2.0279, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3767857142857143, | |
| "grad_norm": 0.21204812824726105, | |
| "learning_rate": 0.0001410212458637112, | |
| "loss": 1.8472, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.37857142857142856, | |
| "grad_norm": 0.36059629917144775, | |
| "learning_rate": 0.00014049965203924054, | |
| "loss": 1.8042, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.38035714285714284, | |
| "grad_norm": 0.21400661766529083, | |
| "learning_rate": 0.0001399767368446634, | |
| "loss": 1.698, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3821428571428571, | |
| "grad_norm": 0.24055758118629456, | |
| "learning_rate": 0.00013945251734097828, | |
| "loss": 1.8758, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.38392857142857145, | |
| "grad_norm": 0.23605166375637054, | |
| "learning_rate": 0.00013892701063173918, | |
| "loss": 1.7425, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.38571428571428573, | |
| "grad_norm": 0.23343758285045624, | |
| "learning_rate": 0.00013840023386249713, | |
| "loss": 1.8683, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3875, | |
| "grad_norm": 0.2475200593471527, | |
| "learning_rate": 0.00013787220422024134, | |
| "loss": 1.9091, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3892857142857143, | |
| "grad_norm": 0.2618944048881531, | |
| "learning_rate": 0.00013734293893283783, | |
| "loss": 1.5086, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.39107142857142857, | |
| "grad_norm": 0.2627498209476471, | |
| "learning_rate": 0.00013681245526846783, | |
| "loss": 1.3878, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 0.24390314519405365, | |
| "learning_rate": 0.0001362807705350641, | |
| "loss": 1.7332, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.39464285714285713, | |
| "grad_norm": 0.2768295705318451, | |
| "learning_rate": 0.00013574790207974646, | |
| "loss": 1.3123, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.3964285714285714, | |
| "grad_norm": 0.2606358230113983, | |
| "learning_rate": 0.0001352138672882555, | |
| "loss": 1.4506, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3982142857142857, | |
| "grad_norm": 0.24806426465511322, | |
| "learning_rate": 0.00013467868358438563, | |
| "loss": 1.7087, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.2664608061313629, | |
| "learning_rate": 0.00013414236842941644, | |
| "loss": 1.3124, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.4017857142857143, | |
| "grad_norm": 0.2661263346672058, | |
| "learning_rate": 0.00013360493932154302, | |
| "loss": 1.2174, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4035714285714286, | |
| "grad_norm": 0.3460038900375366, | |
| "learning_rate": 0.00013306641379530514, | |
| "loss": 0.6889, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.40535714285714286, | |
| "grad_norm": 0.2929069995880127, | |
| "learning_rate": 0.000132526809421015, | |
| "loss": 0.9457, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.40714285714285714, | |
| "grad_norm": 0.3459819257259369, | |
| "learning_rate": 0.00013198614380418412, | |
| "loss": 1.2547, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4089285714285714, | |
| "grad_norm": 0.30105313658714294, | |
| "learning_rate": 0.00013144443458494882, | |
| "loss": 0.957, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 0.3461960256099701, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 1.3146, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4125, | |
| "grad_norm": 0.34542855620384216, | |
| "learning_rate": 0.00013035795606948023, | |
| "loss": 1.1128, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.4142857142857143, | |
| "grad_norm": 0.37605586647987366, | |
| "learning_rate": 0.00012981322222145846, | |
| "loss": 1.5095, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.4160714285714286, | |
| "grad_norm": 0.37267056107521057, | |
| "learning_rate": 0.00012926751566629875, | |
| "loss": 1.071, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.41785714285714287, | |
| "grad_norm": 0.3052172064781189, | |
| "learning_rate": 0.00012872085420860665, | |
| "loss": 1.3136, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.41964285714285715, | |
| "grad_norm": 0.36694592237472534, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 1.2439, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.42142857142857143, | |
| "grad_norm": 0.36055245995521545, | |
| "learning_rate": 0.00012762473795924204, | |
| "loss": 1.1165, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4232142857142857, | |
| "grad_norm": 0.3014545738697052, | |
| "learning_rate": 0.00012707531893022854, | |
| "loss": 1.5423, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.3208891749382019, | |
| "learning_rate": 0.00012652501652283377, | |
| "loss": 1.1813, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.42678571428571427, | |
| "grad_norm": 0.38703230023384094, | |
| "learning_rate": 0.00012597384869161084, | |
| "loss": 0.7706, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.38256821036338806, | |
| "learning_rate": 0.00012542183341934872, | |
| "loss": 1.0565, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4303571428571429, | |
| "grad_norm": 0.3555380702018738, | |
| "learning_rate": 0.0001248689887164855, | |
| "loss": 0.849, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.43214285714285716, | |
| "grad_norm": 0.3472703993320465, | |
| "learning_rate": 0.00012431533262052098, | |
| "loss": 1.3984, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.43392857142857144, | |
| "grad_norm": 0.3631349503993988, | |
| "learning_rate": 0.000123760883195428, | |
| "loss": 0.8955, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4357142857142857, | |
| "grad_norm": 0.349295973777771, | |
| "learning_rate": 0.00012320565853106316, | |
| "loss": 0.8866, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.33635953068733215, | |
| "learning_rate": 0.00012264967674257646, | |
| "loss": 1.2419, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4392857142857143, | |
| "grad_norm": 0.3833181858062744, | |
| "learning_rate": 0.00012209295596982042, | |
| "loss": 1.5507, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.44107142857142856, | |
| "grad_norm": 0.3737214505672455, | |
| "learning_rate": 0.00012153551437675821, | |
| "loss": 1.4881, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.44285714285714284, | |
| "grad_norm": 0.4705282747745514, | |
| "learning_rate": 0.00012097737015087094, | |
| "loss": 1.4864, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4446428571428571, | |
| "grad_norm": 0.39539188146591187, | |
| "learning_rate": 0.00012041854150256433, | |
| "loss": 1.7855, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 0.7369075417518616, | |
| "learning_rate": 0.00011985904666457455, | |
| "loss": 2.01, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.44821428571428573, | |
| "grad_norm": 0.18146094679832458, | |
| "learning_rate": 0.00011929890389137337, | |
| "loss": 1.5898, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.21558880805969238, | |
| "learning_rate": 0.00011873813145857249, | |
| "loss": 1.5816, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4517857142857143, | |
| "grad_norm": 0.19599275290966034, | |
| "learning_rate": 0.00011817674766232734, | |
| "loss": 1.6433, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.45357142857142857, | |
| "grad_norm": 0.22075910866260529, | |
| "learning_rate": 0.00011761477081874015, | |
| "loss": 1.6005, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.45535714285714285, | |
| "grad_norm": 0.19471955299377441, | |
| "learning_rate": 0.0001170522192632624, | |
| "loss": 1.7133, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.19876879453659058, | |
| "learning_rate": 0.00011648911135009634, | |
| "loss": 1.5085, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.4589285714285714, | |
| "grad_norm": 0.20565317571163177, | |
| "learning_rate": 0.00011592546545159645, | |
| "loss": 1.7386, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.4607142857142857, | |
| "grad_norm": 0.24483506381511688, | |
| "learning_rate": 0.00011536129995766996, | |
| "loss": 1.7162, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.4625, | |
| "grad_norm": 0.21543823182582855, | |
| "learning_rate": 0.00011479663327517667, | |
| "loss": 1.6966, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 0.2661048471927643, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 1.8821, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4660714285714286, | |
| "grad_norm": 0.24292460083961487, | |
| "learning_rate": 0.00011366587005308858, | |
| "loss": 1.7085, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.46785714285714286, | |
| "grad_norm": 0.216167613863945, | |
| "learning_rate": 0.0001130998104065693, | |
| "loss": 1.7298, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.46964285714285714, | |
| "grad_norm": 0.2111697793006897, | |
| "learning_rate": 0.00011253332335643043, | |
| "loss": 1.8098, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4714285714285714, | |
| "grad_norm": 0.23981061577796936, | |
| "learning_rate": 0.00011196642738527659, | |
| "loss": 1.7026, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4732142857142857, | |
| "grad_norm": 0.2623251676559448, | |
| "learning_rate": 0.00011139914098905406, | |
| "loss": 1.7894, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.2482486367225647, | |
| "learning_rate": 0.00011083148267644747, | |
| "loss": 1.9019, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4767857142857143, | |
| "grad_norm": 0.238911435008049, | |
| "learning_rate": 0.00011026347096827578, | |
| "loss": 1.6809, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.4785714285714286, | |
| "grad_norm": 0.24704696238040924, | |
| "learning_rate": 0.00010969512439688816, | |
| "loss": 1.6607, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.48035714285714287, | |
| "grad_norm": 0.25105100870132446, | |
| "learning_rate": 0.00010912646150555919, | |
| "loss": 1.5895, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 0.2849842607975006, | |
| "learning_rate": 0.00010855750084788398, | |
| "loss": 2.0812, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.48392857142857143, | |
| "grad_norm": 0.2599342167377472, | |
| "learning_rate": 0.00010798826098717276, | |
| "loss": 1.1569, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 0.28037258982658386, | |
| "learning_rate": 0.00010741876049584523, | |
| "loss": 1.1928, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4875, | |
| "grad_norm": 0.29920563101768494, | |
| "learning_rate": 0.00010684901795482456, | |
| "loss": 1.2244, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.48928571428571427, | |
| "grad_norm": 0.2799164354801178, | |
| "learning_rate": 0.00010627905195293135, | |
| "loss": 1.4455, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.49107142857142855, | |
| "grad_norm": 0.23873603343963623, | |
| "learning_rate": 0.00010570888108627681, | |
| "loss": 0.852, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4928571428571429, | |
| "grad_norm": 0.2817741632461548, | |
| "learning_rate": 0.00010513852395765631, | |
| "loss": 1.3203, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.49464285714285716, | |
| "grad_norm": 0.27295514941215515, | |
| "learning_rate": 0.00010456799917594233, | |
| "loss": 0.749, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.49642857142857144, | |
| "grad_norm": 0.30728739500045776, | |
| "learning_rate": 0.00010399732535547734, | |
| "loss": 1.0083, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.4982142857142857, | |
| "grad_norm": 0.32001444697380066, | |
| "learning_rate": 0.00010342652111546635, | |
| "loss": 1.573, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.3400101065635681, | |
| "learning_rate": 0.00010285560507936961, | |
| "loss": 1.0757, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 1.4439584016799927, | |
| "eval_runtime": 13.9741, | |
| "eval_samples_per_second": 16.888, | |
| "eval_steps_per_second": 8.444, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5017857142857143, | |
| "grad_norm": 0.3074837923049927, | |
| "learning_rate": 0.00010228459587429497, | |
| "loss": 1.1389, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5035714285714286, | |
| "grad_norm": 0.3143484592437744, | |
| "learning_rate": 0.00010171351213038993, | |
| "loss": 0.9542, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5053571428571428, | |
| "grad_norm": 0.3524804413318634, | |
| "learning_rate": 0.00010114237248023404, | |
| "loss": 0.8578, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5071428571428571, | |
| "grad_norm": 0.335183322429657, | |
| "learning_rate": 0.00010057119555823085, | |
| "loss": 0.9228, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5089285714285714, | |
| "grad_norm": 0.35537081956863403, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9541, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5107142857142857, | |
| "grad_norm": 0.3294563591480255, | |
| "learning_rate": 9.942880444176918e-05, | |
| "loss": 1.4223, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5125, | |
| "grad_norm": 0.32628077268600464, | |
| "learning_rate": 9.8857627519766e-05, | |
| "loss": 1.2942, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 0.32195404171943665, | |
| "learning_rate": 9.828648786961008e-05, | |
| "loss": 1.4239, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5160714285714286, | |
| "grad_norm": 0.3283804655075073, | |
| "learning_rate": 9.771540412570504e-05, | |
| "loss": 1.0303, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 0.356052428483963, | |
| "learning_rate": 9.71443949206304e-05, | |
| "loss": 1.0199, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5196428571428572, | |
| "grad_norm": 0.39479124546051025, | |
| "learning_rate": 9.657347888453367e-05, | |
| "loss": 1.1343, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5214285714285715, | |
| "grad_norm": 0.34791451692581177, | |
| "learning_rate": 9.60026746445227e-05, | |
| "loss": 1.521, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5232142857142857, | |
| "grad_norm": 0.3614530861377716, | |
| "learning_rate": 9.543200082405768e-05, | |
| "loss": 1.2346, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.36958596110343933, | |
| "learning_rate": 9.486147604234371e-05, | |
| "loss": 1.0457, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5267857142857143, | |
| "grad_norm": 0.4293418824672699, | |
| "learning_rate": 9.42911189137232e-05, | |
| "loss": 1.1071, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5285714285714286, | |
| "grad_norm": 0.40408602356910706, | |
| "learning_rate": 9.372094804706867e-05, | |
| "loss": 1.3805, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5303571428571429, | |
| "grad_norm": 0.3892784118652344, | |
| "learning_rate": 9.315098204517543e-05, | |
| "loss": 0.9136, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5321428571428571, | |
| "grad_norm": 0.4003988206386566, | |
| "learning_rate": 9.258123950415479e-05, | |
| "loss": 1.3684, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5339285714285714, | |
| "grad_norm": 0.37116503715515137, | |
| "learning_rate": 9.201173901282724e-05, | |
| "loss": 1.7824, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.5286569595336914, | |
| "learning_rate": 9.144249915211605e-05, | |
| "loss": 1.9392, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5375, | |
| "grad_norm": 0.18338526785373688, | |
| "learning_rate": 9.087353849444085e-05, | |
| "loss": 1.4422, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5392857142857143, | |
| "grad_norm": 0.20191389322280884, | |
| "learning_rate": 9.030487560311186e-05, | |
| "loss": 1.5443, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5410714285714285, | |
| "grad_norm": 0.19955602288246155, | |
| "learning_rate": 8.973652903172423e-05, | |
| "loss": 1.6521, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 0.2253991812467575, | |
| "learning_rate": 8.916851732355255e-05, | |
| "loss": 1.6596, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5446428571428571, | |
| "grad_norm": 0.18380412459373474, | |
| "learning_rate": 8.860085901094595e-05, | |
| "loss": 1.5462, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5464285714285714, | |
| "grad_norm": 0.21318556368350983, | |
| "learning_rate": 8.803357261472343e-05, | |
| "loss": 1.6713, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5482142857142858, | |
| "grad_norm": 0.20480670034885406, | |
| "learning_rate": 8.746667664356956e-05, | |
| "loss": 1.7537, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.21598470211029053, | |
| "learning_rate": 8.690018959343072e-05, | |
| "loss": 1.6955, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5517857142857143, | |
| "grad_norm": 0.21172207593917847, | |
| "learning_rate": 8.633412994691144e-05, | |
| "loss": 1.7187, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 0.22086191177368164, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 1.7579, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5553571428571429, | |
| "grad_norm": 0.2266392558813095, | |
| "learning_rate": 8.520336672482338e-05, | |
| "loss": 1.7486, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5571428571428572, | |
| "grad_norm": 0.2660948634147644, | |
| "learning_rate": 8.463870004233008e-05, | |
| "loss": 1.7903, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5589285714285714, | |
| "grad_norm": 0.2297395020723343, | |
| "learning_rate": 8.407453454840357e-05, | |
| "loss": 1.8017, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5607142857142857, | |
| "grad_norm": 0.21526682376861572, | |
| "learning_rate": 8.351088864990368e-05, | |
| "loss": 1.855, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.2516147196292877, | |
| "learning_rate": 8.294778073673762e-05, | |
| "loss": 1.7103, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5642857142857143, | |
| "grad_norm": 0.25641700625419617, | |
| "learning_rate": 8.238522918125983e-05, | |
| "loss": 1.9301, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5660714285714286, | |
| "grad_norm": 0.26828062534332275, | |
| "learning_rate": 8.182325233767267e-05, | |
| "loss": 1.8575, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5678571428571428, | |
| "grad_norm": 0.24787884950637817, | |
| "learning_rate": 8.126186854142752e-05, | |
| "loss": 2.0228, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.5696428571428571, | |
| "grad_norm": 0.23658955097198486, | |
| "learning_rate": 8.070109610862668e-05, | |
| "loss": 1.7813, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.2818485498428345, | |
| "learning_rate": 8.014095333542548e-05, | |
| "loss": 1.7571, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5732142857142857, | |
| "grad_norm": 0.24982373416423798, | |
| "learning_rate": 7.958145849743569e-05, | |
| "loss": 1.602, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.2815864682197571, | |
| "learning_rate": 7.902262984912909e-05, | |
| "loss": 1.7216, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.5767857142857142, | |
| "grad_norm": 0.2675464451313019, | |
| "learning_rate": 7.846448562324183e-05, | |
| "loss": 1.0704, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.5785714285714286, | |
| "grad_norm": 0.23938840627670288, | |
| "learning_rate": 7.79070440301796e-05, | |
| "loss": 1.282, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5803571428571429, | |
| "grad_norm": 0.2834213972091675, | |
| "learning_rate": 7.735032325742355e-05, | |
| "loss": 1.7934, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5821428571428572, | |
| "grad_norm": 0.3555513918399811, | |
| "learning_rate": 7.679434146893685e-05, | |
| "loss": 1.2089, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5839285714285715, | |
| "grad_norm": 0.3254348933696747, | |
| "learning_rate": 7.623911680457198e-05, | |
| "loss": 1.0845, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5857142857142857, | |
| "grad_norm": 0.3558744192123413, | |
| "learning_rate": 7.568466737947905e-05, | |
| "loss": 1.2339, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.5875, | |
| "grad_norm": 0.32738080620765686, | |
| "learning_rate": 7.513101128351454e-05, | |
| "loss": 0.8492, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 0.36939212679862976, | |
| "learning_rate": 7.457816658065134e-05, | |
| "loss": 0.8973, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5910714285714286, | |
| "grad_norm": 0.3393215835094452, | |
| "learning_rate": 7.402615130838917e-05, | |
| "loss": 1.0078, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.5928571428571429, | |
| "grad_norm": 0.402725487947464, | |
| "learning_rate": 7.347498347716624e-05, | |
| "loss": 1.2381, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5946428571428571, | |
| "grad_norm": 0.33164989948272705, | |
| "learning_rate": 7.292468106977148e-05, | |
| "loss": 1.197, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5964285714285714, | |
| "grad_norm": 0.3454689681529999, | |
| "learning_rate": 7.237526204075797e-05, | |
| "loss": 1.0244, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5982142857142857, | |
| "grad_norm": 0.3584868907928467, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 1.062, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.3318020701408386, | |
| "learning_rate": 7.127914579139338e-05, | |
| "loss": 1.3696, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6017857142857143, | |
| "grad_norm": 0.38772639632225037, | |
| "learning_rate": 7.073248433370124e-05, | |
| "loss": 1.1725, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6035714285714285, | |
| "grad_norm": 0.33527833223342896, | |
| "learning_rate": 7.018677777854157e-05, | |
| "loss": 1.0849, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6053571428571428, | |
| "grad_norm": 0.3958960473537445, | |
| "learning_rate": 6.964204393051981e-05, | |
| "loss": 0.8494, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 0.3545495569705963, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.843, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6089285714285714, | |
| "grad_norm": 0.3475857675075531, | |
| "learning_rate": 6.855556541505122e-05, | |
| "loss": 1.1228, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6107142857142858, | |
| "grad_norm": 0.4085654616355896, | |
| "learning_rate": 6.801385619581592e-05, | |
| "loss": 0.8092, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6125, | |
| "grad_norm": 0.4605577886104584, | |
| "learning_rate": 6.747319057898503e-05, | |
| "loss": 1.0999, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6142857142857143, | |
| "grad_norm": 0.3840469717979431, | |
| "learning_rate": 6.693358620469487e-05, | |
| "loss": 1.2712, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6160714285714286, | |
| "grad_norm": 0.3712100684642792, | |
| "learning_rate": 6.639506067845697e-05, | |
| "loss": 1.1401, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6178571428571429, | |
| "grad_norm": 0.41670674085617065, | |
| "learning_rate": 6.585763157058358e-05, | |
| "loss": 1.151, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6196428571428572, | |
| "grad_norm": 0.5912812352180481, | |
| "learning_rate": 6.53213164156144e-05, | |
| "loss": 1.447, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6214285714285714, | |
| "grad_norm": 0.3995843231678009, | |
| "learning_rate": 6.478613271174453e-05, | |
| "loss": 1.6645, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6232142857142857, | |
| "grad_norm": 0.5337942242622375, | |
| "learning_rate": 6.425209792025358e-05, | |
| "loss": 1.8703, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.5726771354675293, | |
| "learning_rate": 6.371922946493591e-05, | |
| "loss": 1.9795, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6267857142857143, | |
| "grad_norm": 0.17994743585586548, | |
| "learning_rate": 6.318754473153221e-05, | |
| "loss": 1.5463, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 0.16961722075939178, | |
| "learning_rate": 6.26570610671622e-05, | |
| "loss": 1.5388, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6303571428571428, | |
| "grad_norm": 0.19410116970539093, | |
| "learning_rate": 6.21277957797587e-05, | |
| "loss": 1.7282, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6321428571428571, | |
| "grad_norm": 0.1906638890504837, | |
| "learning_rate": 6.159976613750286e-05, | |
| "loss": 1.5167, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6339285714285714, | |
| "grad_norm": 0.21918904781341553, | |
| "learning_rate": 6.107298936826086e-05, | |
| "loss": 1.7446, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6357142857142857, | |
| "grad_norm": 0.19429104030132294, | |
| "learning_rate": 6.0547482659021706e-05, | |
| "loss": 1.7166, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6375, | |
| "grad_norm": 0.21244099736213684, | |
| "learning_rate": 6.002326315533665e-05, | |
| "loss": 1.7319, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6392857142857142, | |
| "grad_norm": 0.22793884575366974, | |
| "learning_rate": 5.950034796075947e-05, | |
| "loss": 1.6573, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6410714285714286, | |
| "grad_norm": 0.23558048903942108, | |
| "learning_rate": 5.897875413628884e-05, | |
| "loss": 1.7359, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.21951396763324738, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 1.6536, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6446428571428572, | |
| "grad_norm": 0.22028475999832153, | |
| "learning_rate": 5.793959862554652e-05, | |
| "loss": 1.7257, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6464285714285715, | |
| "grad_norm": 0.23070622980594635, | |
| "learning_rate": 5.7422070843492734e-05, | |
| "loss": 1.6149, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6482142857142857, | |
| "grad_norm": 0.22408127784729004, | |
| "learning_rate": 5.6905932238875123e-05, | |
| "loss": 1.5936, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.23098969459533691, | |
| "learning_rate": 5.639119965159446e-05, | |
| "loss": 1.7313, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6517857142857143, | |
| "grad_norm": 0.2533377707004547, | |
| "learning_rate": 5.5877889875677845e-05, | |
| "loss": 1.9422, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6535714285714286, | |
| "grad_norm": 0.2515897750854492, | |
| "learning_rate": 5.5366019658730825e-05, | |
| "loss": 1.8331, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.6553571428571429, | |
| "grad_norm": 0.2757863700389862, | |
| "learning_rate": 5.485560570139061e-05, | |
| "loss": 1.7759, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 0.29774826765060425, | |
| "learning_rate": 5.434666465678175e-05, | |
| "loss": 1.334, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6589285714285714, | |
| "grad_norm": 0.28664523363113403, | |
| "learning_rate": 5.383921312997242e-05, | |
| "loss": 1.6234, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 0.30774933099746704, | |
| "learning_rate": 5.333326767743263e-05, | |
| "loss": 1.4743, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6625, | |
| "grad_norm": 0.26094669103622437, | |
| "learning_rate": 5.282884480649435e-05, | |
| "loss": 1.1046, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6642857142857143, | |
| "grad_norm": 0.2818247079849243, | |
| "learning_rate": 5.232596097481251e-05, | |
| "loss": 0.8417, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6660714285714285, | |
| "grad_norm": 0.3089035749435425, | |
| "learning_rate": 5.182463258982846e-05, | |
| "loss": 1.3922, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6678571428571428, | |
| "grad_norm": 0.2375502586364746, | |
| "learning_rate": 5.132487600823438e-05, | |
| "loss": 1.0855, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6696428571428571, | |
| "grad_norm": 0.3417636454105377, | |
| "learning_rate": 5.082670753543961e-05, | |
| "loss": 1.0819, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6714285714285714, | |
| "grad_norm": 0.2587840259075165, | |
| "learning_rate": 5.033014342503889e-05, | |
| "loss": 1.1154, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6732142857142858, | |
| "grad_norm": 0.29829278588294983, | |
| "learning_rate": 4.9835199878281765e-05, | |
| "loss": 0.9634, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.307190477848053, | |
| "learning_rate": 4.9341893043544185e-05, | |
| "loss": 1.1533, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.6767857142857143, | |
| "grad_norm": 0.3548847734928131, | |
| "learning_rate": 4.8850239015801625e-05, | |
| "loss": 1.2046, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 0.3130282759666443, | |
| "learning_rate": 4.836025383610382e-05, | |
| "loss": 1.1391, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6803571428571429, | |
| "grad_norm": 0.3400501012802124, | |
| "learning_rate": 4.787195349105159e-05, | |
| "loss": 1.0226, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.6821428571428572, | |
| "grad_norm": 0.3462565541267395, | |
| "learning_rate": 4.7385353912275165e-05, | |
| "loss": 1.0968, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.6839285714285714, | |
| "grad_norm": 0.331123948097229, | |
| "learning_rate": 4.690047097591427e-05, | |
| "loss": 1.1918, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 0.354432612657547, | |
| "learning_rate": 4.6417320502100316e-05, | |
| "loss": 1.2261, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 0.34844398498535156, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 0.7991, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6892857142857143, | |
| "grad_norm": 0.3462367653846741, | |
| "learning_rate": 4.545627993950201e-05, | |
| "loss": 1.1343, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.6910714285714286, | |
| "grad_norm": 0.3352709114551544, | |
| "learning_rate": 4.497842120630229e-05, | |
| "loss": 1.1023, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.6928571428571428, | |
| "grad_norm": 0.3581717610359192, | |
| "learning_rate": 4.4502357645795976e-05, | |
| "loss": 0.9181, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.6946428571428571, | |
| "grad_norm": 0.35995498299598694, | |
| "learning_rate": 4.402810479036725e-05, | |
| "loss": 1.3445, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 0.372935950756073, | |
| "learning_rate": 4.355567811332311e-05, | |
| "loss": 1.0725, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6982142857142857, | |
| "grad_norm": 0.3759123980998993, | |
| "learning_rate": 4.30850930283882e-05, | |
| "loss": 1.1824, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.391770601272583, | |
| "learning_rate": 4.2616364889202254e-05, | |
| "loss": 1.3516, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7017857142857142, | |
| "grad_norm": 0.3785625696182251, | |
| "learning_rate": 4.214950898881892e-05, | |
| "loss": 1.1624, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7035714285714286, | |
| "grad_norm": 0.4284125864505768, | |
| "learning_rate": 4.168454055920681e-05, | |
| "loss": 1.1318, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7053571428571429, | |
| "grad_norm": 0.391161173582077, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 1.4543, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7071428571428572, | |
| "grad_norm": 0.3723802864551544, | |
| "learning_rate": 4.0760326731766374e-05, | |
| "loss": 1.4265, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7089285714285715, | |
| "grad_norm": 0.4321235418319702, | |
| "learning_rate": 4.030111148798775e-05, | |
| "loss": 1.4523, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7107142857142857, | |
| "grad_norm": 0.43963977694511414, | |
| "learning_rate": 3.9843844022096135e-05, | |
| "loss": 1.7426, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7125, | |
| "grad_norm": 0.5444914698600769, | |
| "learning_rate": 3.938853925322118e-05, | |
| "loss": 1.8907, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.9059175252914429, | |
| "learning_rate": 3.893521203645618e-05, | |
| "loss": 2.2147, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7160714285714286, | |
| "grad_norm": 0.20250235497951508, | |
| "learning_rate": 3.848387716237353e-05, | |
| "loss": 1.7341, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7178571428571429, | |
| "grad_norm": 0.1863853931427002, | |
| "learning_rate": 3.8034549356541894e-05, | |
| "loss": 1.6956, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.7196428571428571, | |
| "grad_norm": 0.19317291676998138, | |
| "learning_rate": 3.7587243279046056e-05, | |
| "loss": 1.7165, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7214285714285714, | |
| "grad_norm": 0.21101966500282288, | |
| "learning_rate": 3.714197352400849e-05, | |
| "loss": 1.8306, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7232142857142857, | |
| "grad_norm": 0.22385361790657043, | |
| "learning_rate": 3.669875461911297e-05, | |
| "loss": 1.7104, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.22555914521217346, | |
| "learning_rate": 3.6257601025131026e-05, | |
| "loss": 1.5668, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7267857142857143, | |
| "grad_norm": 0.21916812658309937, | |
| "learning_rate": 3.581852713544983e-05, | |
| "loss": 1.7827, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.7285714285714285, | |
| "grad_norm": 0.23447498679161072, | |
| "learning_rate": 3.538154727560259e-05, | |
| "loss": 1.8308, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.7303571428571428, | |
| "grad_norm": 0.21024593710899353, | |
| "learning_rate": 3.494667570280132e-05, | |
| "loss": 1.613, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 0.23882578313350677, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 1.6853, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7339285714285714, | |
| "grad_norm": 0.23162604868412018, | |
| "learning_rate": 3.408331410278929e-05, | |
| "loss": 1.7371, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7357142857142858, | |
| "grad_norm": 0.23150567710399628, | |
| "learning_rate": 3.3654852244220826e-05, | |
| "loss": 1.7505, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.7375, | |
| "grad_norm": 0.23027552664279938, | |
| "learning_rate": 3.322855500906373e-05, | |
| "loss": 1.7128, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7392857142857143, | |
| "grad_norm": 0.22426114976406097, | |
| "learning_rate": 3.2804436305991214e-05, | |
| "loss": 1.7721, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.7410714285714286, | |
| "grad_norm": 0.22792723774909973, | |
| "learning_rate": 3.238250997259808e-05, | |
| "loss": 1.7089, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 0.2450588494539261, | |
| "learning_rate": 3.196278977494934e-05, | |
| "loss": 1.744, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7446428571428572, | |
| "grad_norm": 0.2348526120185852, | |
| "learning_rate": 3.154528940713113e-05, | |
| "loss": 1.8496, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.7464285714285714, | |
| "grad_norm": 0.2519124746322632, | |
| "learning_rate": 3.113002249080386e-05, | |
| "loss": 1.76, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.7482142857142857, | |
| "grad_norm": 0.27859431505203247, | |
| "learning_rate": 3.071700257475768e-05, | |
| "loss": 1.6493, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.2539427876472473, | |
| "learning_rate": 3.030624313447067e-05, | |
| "loss": 1.7619, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 1.4375773668289185, | |
| "eval_runtime": 13.3809, | |
| "eval_samples_per_second": 17.637, | |
| "eval_steps_per_second": 8.819, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7517857142857143, | |
| "grad_norm": 0.2867870330810547, | |
| "learning_rate": 2.9897757571668905e-05, | |
| "loss": 1.5308, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.7535714285714286, | |
| "grad_norm": 0.26840099692344666, | |
| "learning_rate": 2.949155921388943e-05, | |
| "loss": 1.8933, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.7553571428571428, | |
| "grad_norm": 0.32162272930145264, | |
| "learning_rate": 2.9087661314045366e-05, | |
| "loss": 1.6497, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.7571428571428571, | |
| "grad_norm": 0.2842392921447754, | |
| "learning_rate": 2.8686077049993287e-05, | |
| "loss": 1.521, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.7589285714285714, | |
| "grad_norm": 0.28831931948661804, | |
| "learning_rate": 2.828681952410366e-05, | |
| "loss": 1.2619, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7607142857142857, | |
| "grad_norm": 0.2879536747932434, | |
| "learning_rate": 2.7889901762833083e-05, | |
| "loss": 1.1705, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.7625, | |
| "grad_norm": 0.32446354627609253, | |
| "learning_rate": 2.7495336716299313e-05, | |
| "loss": 1.1636, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.7642857142857142, | |
| "grad_norm": 0.2487679123878479, | |
| "learning_rate": 2.7103137257858868e-05, | |
| "loss": 0.7092, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.7660714285714286, | |
| "grad_norm": 0.3402203321456909, | |
| "learning_rate": 2.671331618368682e-05, | |
| "loss": 0.8279, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 0.3117457330226898, | |
| "learning_rate": 2.6325886212359498e-05, | |
| "loss": 1.2101, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7696428571428572, | |
| "grad_norm": 0.28996577858924866, | |
| "learning_rate": 2.5940859984439424e-05, | |
| "loss": 1.1556, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 0.31623905897140503, | |
| "learning_rate": 2.5558250062062828e-05, | |
| "loss": 1.1324, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.7732142857142857, | |
| "grad_norm": 0.29438769817352295, | |
| "learning_rate": 2.5178068928529864e-05, | |
| "loss": 1.4183, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.31944945454597473, | |
| "learning_rate": 2.4800328987897427e-05, | |
| "loss": 1.1763, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.7767857142857143, | |
| "grad_norm": 0.32895660400390625, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 0.8288, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7785714285714286, | |
| "grad_norm": 0.46789219975471497, | |
| "learning_rate": 2.4052221902918725e-05, | |
| "loss": 0.889, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.7803571428571429, | |
| "grad_norm": 0.35038378834724426, | |
| "learning_rate": 2.368187916683997e-05, | |
| "loss": 1.1607, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.7821428571428571, | |
| "grad_norm": 0.3326191008090973, | |
| "learning_rate": 2.3314026439400217e-05, | |
| "loss": 1.3321, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.7839285714285714, | |
| "grad_norm": 0.3683798015117645, | |
| "learning_rate": 2.2948675722421086e-05, | |
| "loss": 0.831, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.3587857186794281, | |
| "learning_rate": 2.2585838936091754e-05, | |
| "loss": 1.0373, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7875, | |
| "grad_norm": 0.341905802488327, | |
| "learning_rate": 2.2225527918580204e-05, | |
| "loss": 1.2294, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.7892857142857143, | |
| "grad_norm": 0.36167776584625244, | |
| "learning_rate": 2.1867754425646926e-05, | |
| "loss": 1.06, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.7910714285714285, | |
| "grad_norm": 0.35555702447891235, | |
| "learning_rate": 2.151253013026121e-05, | |
| "loss": 1.5072, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.7928571428571428, | |
| "grad_norm": 0.34384891390800476, | |
| "learning_rate": 2.115986662222058e-05, | |
| "loss": 1.3456, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.7946428571428571, | |
| "grad_norm": 0.3673519492149353, | |
| "learning_rate": 2.0809775407772503e-05, | |
| "loss": 1.3223, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7964285714285714, | |
| "grad_norm": 0.400193452835083, | |
| "learning_rate": 2.0462267909238896e-05, | |
| "loss": 1.562, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.7982142857142858, | |
| "grad_norm": 0.4361307621002197, | |
| "learning_rate": 2.011735546464365e-05, | |
| "loss": 1.1107, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.38674113154411316, | |
| "learning_rate": 1.9775049327342486e-05, | |
| "loss": 1.2813, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8017857142857143, | |
| "grad_norm": 0.40437138080596924, | |
| "learning_rate": 1.943536066565603e-05, | |
| "loss": 1.5164, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 0.5968692898750305, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 2.0903, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8053571428571429, | |
| "grad_norm": 0.16759343445301056, | |
| "learning_rate": 1.876388001504995e-05, | |
| "loss": 1.4236, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8071428571428572, | |
| "grad_norm": 0.1815633326768875, | |
| "learning_rate": 1.8432109934329834e-05, | |
| "loss": 1.6486, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.8089285714285714, | |
| "grad_norm": 0.18043053150177002, | |
| "learning_rate": 1.810300114490875e-05, | |
| "loss": 1.4118, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8107142857142857, | |
| "grad_norm": 0.19016335904598236, | |
| "learning_rate": 1.777656438452129e-05, | |
| "loss": 1.4844, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 0.2066669464111328, | |
| "learning_rate": 1.74528103037226e-05, | |
| "loss": 1.583, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8142857142857143, | |
| "grad_norm": 0.21698947250843048, | |
| "learning_rate": 1.713174946554086e-05, | |
| "loss": 1.6314, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.8160714285714286, | |
| "grad_norm": 0.21937011182308197, | |
| "learning_rate": 1.6813392345132518e-05, | |
| "loss": 1.6342, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8178571428571428, | |
| "grad_norm": 0.24016696214675903, | |
| "learning_rate": 1.649774932944075e-05, | |
| "loss": 1.7726, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8196428571428571, | |
| "grad_norm": 0.23045389354228973, | |
| "learning_rate": 1.6184830716856347e-05, | |
| "loss": 1.8679, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 0.2114153504371643, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 1.6223, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8232142857142857, | |
| "grad_norm": 0.22911155223846436, | |
| "learning_rate": 1.5567207449798515e-05, | |
| "loss": 1.6179, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.22748011350631714, | |
| "learning_rate": 1.5262522946335755e-05, | |
| "loss": 1.7988, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.8267857142857142, | |
| "grad_norm": 0.22931312024593353, | |
| "learning_rate": 1.4960603147344343e-05, | |
| "loss": 1.6515, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 0.24372749030590057, | |
| "learning_rate": 1.466145790347183e-05, | |
| "loss": 1.855, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8303571428571429, | |
| "grad_norm": 0.24426500499248505, | |
| "learning_rate": 1.4365096974841108e-05, | |
| "loss": 1.7367, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8321428571428572, | |
| "grad_norm": 0.22962844371795654, | |
| "learning_rate": 1.4071530030732095e-05, | |
| "loss": 1.871, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.8339285714285715, | |
| "grad_norm": 0.25352221727371216, | |
| "learning_rate": 1.3780766649266242e-05, | |
| "loss": 1.816, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8357142857142857, | |
| "grad_norm": 0.2572453022003174, | |
| "learning_rate": 1.3492816317093893e-05, | |
| "loss": 1.857, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.8375, | |
| "grad_norm": 0.26578372716903687, | |
| "learning_rate": 1.3207688429084974e-05, | |
| "loss": 1.8242, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 0.2806295156478882, | |
| "learning_rate": 1.2925392288022298e-05, | |
| "loss": 1.9345, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8410714285714286, | |
| "grad_norm": 0.2791413962841034, | |
| "learning_rate": 1.2645937104298111e-05, | |
| "loss": 1.4595, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.8428571428571429, | |
| "grad_norm": 0.25607559084892273, | |
| "learning_rate": 1.2369331995613665e-05, | |
| "loss": 1.493, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.8446428571428571, | |
| "grad_norm": 0.2600158452987671, | |
| "learning_rate": 1.2095585986681535e-05, | |
| "loss": 1.4394, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.8464285714285714, | |
| "grad_norm": 0.27362141013145447, | |
| "learning_rate": 1.1824708008931418e-05, | |
| "loss": 1.0637, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.8482142857142857, | |
| "grad_norm": 0.27769190073013306, | |
| "learning_rate": 1.1556706900218572e-05, | |
| "loss": 1.2003, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.2426546961069107, | |
| "learning_rate": 1.1291591404535462e-05, | |
| "loss": 0.8119, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.8517857142857143, | |
| "grad_norm": 0.2766360938549042, | |
| "learning_rate": 1.1029370171726571e-05, | |
| "loss": 1.0697, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.8535714285714285, | |
| "grad_norm": 0.31444060802459717, | |
| "learning_rate": 1.0770051757206079e-05, | |
| "loss": 1.3142, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.8553571428571428, | |
| "grad_norm": 0.33098921179771423, | |
| "learning_rate": 1.051364462167881e-05, | |
| "loss": 1.0656, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.327269583940506, | |
| "learning_rate": 1.026015713086418e-05, | |
| "loss": 1.1335, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8589285714285714, | |
| "grad_norm": 0.30396682024002075, | |
| "learning_rate": 1.0009597555223128e-05, | |
| "loss": 1.2085, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.8607142857142858, | |
| "grad_norm": 0.30377310514450073, | |
| "learning_rate": 9.761974069688461e-06, | |
| "loss": 1.3687, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.8625, | |
| "grad_norm": 0.3333049714565277, | |
| "learning_rate": 9.517294753398064e-06, | |
| "loss": 1.0564, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.8642857142857143, | |
| "grad_norm": 0.3406635820865631, | |
| "learning_rate": 9.275567589431178e-06, | |
| "loss": 1.1551, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.8660714285714286, | |
| "grad_norm": 0.35342979431152344, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 0.8579, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.8678571428571429, | |
| "grad_norm": 0.338238000869751, | |
| "learning_rate": 8.80100116893301e-06, | |
| "loss": 1.2404, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.8696428571428572, | |
| "grad_norm": 0.32925620675086975, | |
| "learning_rate": 8.568177395939215e-06, | |
| "loss": 0.9557, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.8714285714285714, | |
| "grad_norm": 0.36271733045578003, | |
| "learning_rate": 8.338336741838838e-06, | |
| "loss": 1.1323, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.8732142857142857, | |
| "grad_norm": 0.373471736907959, | |
| "learning_rate": 8.111486705574534e-06, | |
| "loss": 0.9512, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.3971388041973114, | |
| "learning_rate": 7.887634688515e-06, | |
| "loss": 1.0311, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8767857142857143, | |
| "grad_norm": 0.35106372833251953, | |
| "learning_rate": 7.666787994213453e-06, | |
| "loss": 1.0705, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.8785714285714286, | |
| "grad_norm": 0.3654085099697113, | |
| "learning_rate": 7.448953828169314e-06, | |
| "loss": 1.2338, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.8803571428571428, | |
| "grad_norm": 0.3616616725921631, | |
| "learning_rate": 7.2341392975931785e-06, | |
| "loss": 1.0171, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.8821428571428571, | |
| "grad_norm": 0.39053040742874146, | |
| "learning_rate": 7.022351411174866e-06, | |
| "loss": 1.2481, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.8839285714285714, | |
| "grad_norm": 0.36361098289489746, | |
| "learning_rate": 6.813597078854772e-06, | |
| "loss": 1.3712, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 0.3919769823551178, | |
| "learning_rate": 6.607883111598445e-06, | |
| "loss": 1.2261, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.8875, | |
| "grad_norm": 0.3980182707309723, | |
| "learning_rate": 6.405216221174326e-06, | |
| "loss": 1.2769, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.8892857142857142, | |
| "grad_norm": 0.4138076901435852, | |
| "learning_rate": 6.205603019934791e-06, | |
| "loss": 1.4143, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.8910714285714286, | |
| "grad_norm": 0.47321876883506775, | |
| "learning_rate": 6.009050020600459e-06, | |
| "loss": 1.687, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.532382607460022, | |
| "learning_rate": 5.8155636360475385e-06, | |
| "loss": 2.0251, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8946428571428572, | |
| "grad_norm": 0.16291013360023499, | |
| "learning_rate": 5.625150179098804e-06, | |
| "loss": 1.3453, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.8964285714285715, | |
| "grad_norm": 0.19601905345916748, | |
| "learning_rate": 5.437815862317519e-06, | |
| "loss": 1.6226, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.8982142857142857, | |
| "grad_norm": 0.20232488214969635, | |
| "learning_rate": 5.25356679780471e-06, | |
| "loss": 1.7009, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.21668295562267303, | |
| "learning_rate": 5.072408996999844e-06, | |
| "loss": 1.674, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.9017857142857143, | |
| "grad_norm": 0.208509162068367, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 1.7305, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.9035714285714286, | |
| "grad_norm": 0.22677063941955566, | |
| "learning_rate": 4.719390727790218e-06, | |
| "loss": 1.9097, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.9053571428571429, | |
| "grad_norm": 0.2320660501718521, | |
| "learning_rate": 4.547541777207565e-06, | |
| "loss": 1.6309, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9071428571428571, | |
| "grad_norm": 0.21626609563827515, | |
| "learning_rate": 4.378807125601303e-06, | |
| "loss": 1.8275, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.9089285714285714, | |
| "grad_norm": 0.2348957508802414, | |
| "learning_rate": 4.2131922782267405e-06, | |
| "loss": 1.6609, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 0.21976341307163239, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 1.6476, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9125, | |
| "grad_norm": 0.25860780477523804, | |
| "learning_rate": 3.891343508073053e-06, | |
| "loss": 1.7382, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.24138076603412628, | |
| "learning_rate": 3.7351200861580617e-06, | |
| "loss": 1.7668, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9160714285714285, | |
| "grad_norm": 0.23734422028064728, | |
| "learning_rate": 3.5820374698604555e-06, | |
| "loss": 1.794, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9178571428571428, | |
| "grad_norm": 0.23496972024440765, | |
| "learning_rate": 3.4321006537612165e-06, | |
| "loss": 1.7674, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9196428571428571, | |
| "grad_norm": 0.24626807868480682, | |
| "learning_rate": 3.2853145298042953e-06, | |
| "loss": 2.0287, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9214285714285714, | |
| "grad_norm": 0.24339932203292847, | |
| "learning_rate": 3.1416838871368924e-06, | |
| "loss": 1.8533, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.9232142857142858, | |
| "grad_norm": 0.253841370344162, | |
| "learning_rate": 3.0012134119532964e-06, | |
| "loss": 1.6682, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.26810407638549805, | |
| "learning_rate": 2.863907687341949e-06, | |
| "loss": 1.4686, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.9267857142857143, | |
| "grad_norm": 0.24126453697681427, | |
| "learning_rate": 2.7297711931358993e-06, | |
| "loss": 1.0048, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.3192795515060425, | |
| "learning_rate": 2.5988083057666533e-06, | |
| "loss": 1.4609, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9303571428571429, | |
| "grad_norm": 0.2680839002132416, | |
| "learning_rate": 2.471023298121422e-06, | |
| "loss": 1.4287, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9321428571428572, | |
| "grad_norm": 0.34683963656425476, | |
| "learning_rate": 2.3464203394036322e-06, | |
| "loss": 1.1019, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.9339285714285714, | |
| "grad_norm": 0.31293728947639465, | |
| "learning_rate": 2.2250034949969913e-06, | |
| "loss": 1.5812, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.9357142857142857, | |
| "grad_norm": 0.27005112171173096, | |
| "learning_rate": 2.1067767263327933e-06, | |
| "loss": 0.8701, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.2739523649215698, | |
| "learning_rate": 1.9917438907606556e-06, | |
| "loss": 1.1165, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9392857142857143, | |
| "grad_norm": 0.2776015102863312, | |
| "learning_rate": 1.87990874142272e-06, | |
| "loss": 1.1448, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.9410714285714286, | |
| "grad_norm": 0.30518367886543274, | |
| "learning_rate": 1.771274927131139e-06, | |
| "loss": 1.0695, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 0.31975796818733215, | |
| "learning_rate": 1.665845992249071e-06, | |
| "loss": 1.1755, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.9446428571428571, | |
| "grad_norm": 0.3304663598537445, | |
| "learning_rate": 1.5636253765750508e-06, | |
| "loss": 1.0217, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 0.3353167474269867, | |
| "learning_rate": 1.4646164152307018e-06, | |
| "loss": 1.2216, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9482142857142857, | |
| "grad_norm": 0.3266999423503876, | |
| "learning_rate": 1.3688223385519672e-06, | |
| "loss": 0.8658, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.34619686007499695, | |
| "learning_rate": 1.2762462719837275e-06, | |
| "loss": 1.2839, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.9517857142857142, | |
| "grad_norm": 0.32094186544418335, | |
| "learning_rate": 1.1868912359777607e-06, | |
| "loss": 1.1571, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.9535714285714286, | |
| "grad_norm": 0.33895429968833923, | |
| "learning_rate": 1.1007601458942752e-06, | |
| "loss": 1.0222, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.9553571428571429, | |
| "grad_norm": 0.34143325686454773, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.9386, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.9571428571428572, | |
| "grad_norm": 0.3505565822124481, | |
| "learning_rate": 9.381809389101825e-07, | |
| "loss": 1.1175, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.9589285714285715, | |
| "grad_norm": 0.36876025795936584, | |
| "learning_rate": 8.617381264330426e-07, | |
| "loss": 1.0065, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.9607142857142857, | |
| "grad_norm": 0.3824593722820282, | |
| "learning_rate": 7.885298685522235e-07, | |
| "loss": 1.1445, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.9625, | |
| "grad_norm": 0.3944648504257202, | |
| "learning_rate": 7.185585538117657e-07, | |
| "loss": 0.9773, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.32271888852119446, | |
| "learning_rate": 6.518264651449779e-07, | |
| "loss": 1.073, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9660714285714286, | |
| "grad_norm": 0.36668431758880615, | |
| "learning_rate": 5.883357797998757e-07, | |
| "loss": 1.3001, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.9678571428571429, | |
| "grad_norm": 0.36653202772140503, | |
| "learning_rate": 5.280885692681592e-07, | |
| "loss": 1.0388, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.9696428571428571, | |
| "grad_norm": 0.3570370674133301, | |
| "learning_rate": 4.710867992176682e-07, | |
| "loss": 1.0728, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 0.3680841624736786, | |
| "learning_rate": 4.173323294281994e-07, | |
| "loss": 1.3424, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.9732142857142857, | |
| "grad_norm": 0.37311863899230957, | |
| "learning_rate": 3.6682691373086665e-07, | |
| "loss": 1.3634, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.38995441794395447, | |
| "learning_rate": 3.195721999508461e-07, | |
| "loss": 1.4773, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.9767857142857143, | |
| "grad_norm": 0.3970450162887573, | |
| "learning_rate": 2.7556972985363085e-07, | |
| "loss": 1.3839, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.9785714285714285, | |
| "grad_norm": 0.5267580151557922, | |
| "learning_rate": 2.3482093909473756e-07, | |
| "loss": 1.9768, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.9803571428571428, | |
| "grad_norm": 0.4860612154006958, | |
| "learning_rate": 1.973271571728441e-07, | |
| "loss": 1.7893, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 0.9272521734237671, | |
| "learning_rate": 1.630896073864352e-07, | |
| "loss": 2.1017, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9839285714285714, | |
| "grad_norm": 0.19919352233409882, | |
| "learning_rate": 1.3210940679385664e-07, | |
| "loss": 1.5237, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.9857142857142858, | |
| "grad_norm": 0.20294204354286194, | |
| "learning_rate": 1.0438756617691115e-07, | |
| "loss": 1.597, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.9875, | |
| "grad_norm": 0.2281683087348938, | |
| "learning_rate": 7.992499000785136e-08, | |
| "loss": 1.7265, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.9892857142857143, | |
| "grad_norm": 0.2398226112127304, | |
| "learning_rate": 5.872247641987016e-08, | |
| "loss": 1.7301, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.9910714285714286, | |
| "grad_norm": 0.2581394612789154, | |
| "learning_rate": 4.078071718107701e-08, | |
| "loss": 1.6294, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.9928571428571429, | |
| "grad_norm": 0.2676229774951935, | |
| "learning_rate": 2.610029767191602e-08, | |
| "loss": 0.9741, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.9946428571428572, | |
| "grad_norm": 0.30877920985221863, | |
| "learning_rate": 1.4681696866081229e-08, | |
| "loss": 1.1485, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.9964285714285714, | |
| "grad_norm": 0.3330385386943817, | |
| "learning_rate": 6.525287314851358e-09, | |
| "loss": 0.906, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.9982142857142857, | |
| "grad_norm": 0.364859402179718, | |
| "learning_rate": 1.6313351349883655e-09, | |
| "loss": 1.0913, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5055291652679443, | |
| "learning_rate": 0.0, | |
| "loss": 1.6353, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.4418776035308838, | |
| "eval_runtime": 13.3713, | |
| "eval_samples_per_second": 17.65, | |
| "eval_steps_per_second": 8.825, | |
| "step": 560 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 140, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.826642274948219e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |