{ "best_metric": 0.9684666991233826, "best_model_checkpoint": "miner_id_24/checkpoint-200", "epoch": 0.9060022650056625, "eval_steps": 50, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004530011325028313, "grad_norm": 2.294863700866699, "learning_rate": 1e-05, "loss": 1.6809, "step": 1 }, { "epoch": 0.004530011325028313, "eval_loss": 3.339136838912964, "eval_runtime": 27.6289, "eval_samples_per_second": 13.464, "eval_steps_per_second": 3.366, "step": 1 }, { "epoch": 0.009060022650056626, "grad_norm": 2.830218553543091, "learning_rate": 2e-05, "loss": 1.956, "step": 2 }, { "epoch": 0.013590033975084938, "grad_norm": 2.898862361907959, "learning_rate": 3e-05, "loss": 2.2194, "step": 3 }, { "epoch": 0.01812004530011325, "grad_norm": 7.4336161613464355, "learning_rate": 4e-05, "loss": 2.1395, "step": 4 }, { "epoch": 0.022650056625141562, "grad_norm": 4.23063850402832, "learning_rate": 5e-05, "loss": 2.1405, "step": 5 }, { "epoch": 0.027180067950169876, "grad_norm": 4.290370464324951, "learning_rate": 6e-05, "loss": 1.8719, "step": 6 }, { "epoch": 0.031710079275198186, "grad_norm": 3.541994333267212, "learning_rate": 7e-05, "loss": 1.6124, "step": 7 }, { "epoch": 0.0362400906002265, "grad_norm": 2.3610470294952393, "learning_rate": 8e-05, "loss": 1.5442, "step": 8 }, { "epoch": 0.04077010192525481, "grad_norm": 2.551922082901001, "learning_rate": 9e-05, "loss": 1.2651, "step": 9 }, { "epoch": 0.045300113250283124, "grad_norm": 2.418034315109253, "learning_rate": 0.0001, "loss": 1.3057, "step": 10 }, { "epoch": 0.04983012457531144, "grad_norm": 2.066025733947754, "learning_rate": 9.999316524962345e-05, "loss": 1.3253, "step": 11 }, { "epoch": 0.05436013590033975, "grad_norm": 1.9381566047668457, "learning_rate": 9.997266286704631e-05, "loss": 1.4766, "step": 12 }, { "epoch": 0.05889014722536806, "grad_norm": 2.2231814861297607, "learning_rate": 9.993849845741524e-05, "loss": 1.3767, "step": 13 }, { "epoch": 0.06342015855039637, "grad_norm": 1.644042730331421, "learning_rate": 9.989068136093873e-05, "loss": 1.0681, "step": 14 }, { "epoch": 0.06795016987542468, "grad_norm": 1.984635353088379, "learning_rate": 9.98292246503335e-05, "loss": 1.2662, "step": 15 }, { "epoch": 0.072480181200453, "grad_norm": 1.8118503093719482, "learning_rate": 9.975414512725057e-05, "loss": 1.3361, "step": 16 }, { "epoch": 0.07701019252548132, "grad_norm": 1.784030795097351, "learning_rate": 9.966546331768191e-05, "loss": 1.2805, "step": 17 }, { "epoch": 0.08154020385050963, "grad_norm": 1.6014776229858398, "learning_rate": 9.956320346634876e-05, "loss": 1.2645, "step": 18 }, { "epoch": 0.08607021517553794, "grad_norm": 1.550774335861206, "learning_rate": 9.944739353007344e-05, "loss": 1.2934, "step": 19 }, { "epoch": 0.09060022650056625, "grad_norm": 1.9732279777526855, "learning_rate": 9.931806517013612e-05, "loss": 1.3981, "step": 20 }, { "epoch": 0.09513023782559456, "grad_norm": 1.6280652284622192, "learning_rate": 9.917525374361912e-05, "loss": 1.4538, "step": 21 }, { "epoch": 0.09966024915062288, "grad_norm": 1.388110637664795, "learning_rate": 9.901899829374047e-05, "loss": 1.0999, "step": 22 }, { "epoch": 0.10419026047565119, "grad_norm": 1.5452589988708496, "learning_rate": 9.884934153917997e-05, "loss": 1.3083, "step": 23 }, { "epoch": 0.1087202718006795, "grad_norm": 1.63185715675354, "learning_rate": 9.86663298624003e-05, "loss": 1.1674, "step": 24 }, { "epoch": 0.11325028312570781, "grad_norm": 1.6084994077682495, "learning_rate": 9.847001329696653e-05, "loss": 1.2719, "step": 25 }, { "epoch": 0.11778029445073612, "grad_norm": 1.5374693870544434, "learning_rate": 9.826044551386744e-05, "loss": 1.1589, "step": 26 }, { "epoch": 0.12231030577576443, "grad_norm": 1.6639900207519531, "learning_rate": 9.803768380684242e-05, "loss": 1.3361, "step": 27 }, { "epoch": 0.12684031710079274, "grad_norm": 1.5302115678787231, "learning_rate": 9.780178907671789e-05, "loss": 1.2753, "step": 28 }, { "epoch": 0.13137032842582105, "grad_norm": 1.5136266946792603, "learning_rate": 9.755282581475769e-05, "loss": 1.3348, "step": 29 }, { "epoch": 0.13590033975084936, "grad_norm": 1.5961568355560303, "learning_rate": 9.729086208503174e-05, "loss": 1.2247, "step": 30 }, { "epoch": 0.1404303510758777, "grad_norm": 1.5466892719268799, "learning_rate": 9.701596950580806e-05, "loss": 1.2773, "step": 31 }, { "epoch": 0.144960362400906, "grad_norm": 1.490113377571106, "learning_rate": 9.672822322997305e-05, "loss": 1.2295, "step": 32 }, { "epoch": 0.14949037372593432, "grad_norm": 1.5653648376464844, "learning_rate": 9.642770192448536e-05, "loss": 1.3314, "step": 33 }, { "epoch": 0.15402038505096263, "grad_norm": 1.8402352333068848, "learning_rate": 9.611448774886924e-05, "loss": 1.3435, "step": 34 }, { "epoch": 0.15855039637599094, "grad_norm": 1.6522904634475708, "learning_rate": 9.578866633275288e-05, "loss": 1.2712, "step": 35 }, { "epoch": 0.16308040770101925, "grad_norm": 1.8196598291397095, "learning_rate": 9.545032675245813e-05, "loss": 1.3977, "step": 36 }, { "epoch": 0.16761041902604756, "grad_norm": 1.951027274131775, "learning_rate": 9.509956150664796e-05, "loss": 1.4954, "step": 37 }, { "epoch": 0.17214043035107587, "grad_norm": 1.9112436771392822, "learning_rate": 9.473646649103818e-05, "loss": 1.3942, "step": 38 }, { "epoch": 0.17667044167610418, "grad_norm": 2.1712679862976074, "learning_rate": 9.43611409721806e-05, "loss": 1.4415, "step": 39 }, { "epoch": 0.1812004530011325, "grad_norm": 1.8498501777648926, "learning_rate": 9.397368756032445e-05, "loss": 1.3505, "step": 40 }, { "epoch": 0.1857304643261608, "grad_norm": 2.059781074523926, "learning_rate": 9.357421218136386e-05, "loss": 1.3482, "step": 41 }, { "epoch": 0.19026047565118911, "grad_norm": 2.0546348094940186, "learning_rate": 9.316282404787871e-05, "loss": 1.3824, "step": 42 }, { "epoch": 0.19479048697621745, "grad_norm": 2.621016263961792, "learning_rate": 9.273963562927695e-05, "loss": 1.606, "step": 43 }, { "epoch": 0.19932049830124576, "grad_norm": 2.086237668991089, "learning_rate": 9.230476262104677e-05, "loss": 1.2631, "step": 44 }, { "epoch": 0.20385050962627407, "grad_norm": 2.4675724506378174, "learning_rate": 9.185832391312644e-05, "loss": 1.5235, "step": 45 }, { "epoch": 0.20838052095130238, "grad_norm": 2.6800007820129395, "learning_rate": 9.140044155740101e-05, "loss": 1.517, "step": 46 }, { "epoch": 0.2129105322763307, "grad_norm": 2.2700512409210205, "learning_rate": 9.093124073433463e-05, "loss": 1.3666, "step": 47 }, { "epoch": 0.217440543601359, "grad_norm": 2.7804107666015625, "learning_rate": 9.045084971874738e-05, "loss": 1.4142, "step": 48 }, { "epoch": 0.22197055492638731, "grad_norm": 3.5445613861083984, "learning_rate": 8.995939984474624e-05, "loss": 1.6274, "step": 49 }, { "epoch": 0.22650056625141562, "grad_norm": 2.5566651821136475, "learning_rate": 8.945702546981969e-05, "loss": 1.3407, "step": 50 }, { "epoch": 0.22650056625141562, "eval_loss": 1.4198116064071655, "eval_runtime": 28.0668, "eval_samples_per_second": 13.254, "eval_steps_per_second": 3.314, "step": 50 }, { "epoch": 0.23103057757644394, "grad_norm": 3.488363742828369, "learning_rate": 8.894386393810563e-05, "loss": 0.9693, "step": 51 }, { "epoch": 0.23556058890147225, "grad_norm": 2.687886953353882, "learning_rate": 8.842005554284296e-05, "loss": 1.1449, "step": 52 }, { "epoch": 0.24009060022650056, "grad_norm": 2.1666171550750732, "learning_rate": 8.788574348801675e-05, "loss": 1.3024, "step": 53 }, { "epoch": 0.24462061155152887, "grad_norm": 1.742993712425232, "learning_rate": 8.73410738492077e-05, "loss": 1.1097, "step": 54 }, { "epoch": 0.2491506228765572, "grad_norm": 1.1420605182647705, "learning_rate": 8.678619553365659e-05, "loss": 0.9356, "step": 55 }, { "epoch": 0.2536806342015855, "grad_norm": 1.286728024482727, "learning_rate": 8.622126023955446e-05, "loss": 0.8034, "step": 56 }, { "epoch": 0.2582106455266138, "grad_norm": 1.2109533548355103, "learning_rate": 8.564642241456986e-05, "loss": 0.982, "step": 57 }, { "epoch": 0.2627406568516421, "grad_norm": 1.1387308835983276, "learning_rate": 8.506183921362443e-05, "loss": 0.8655, "step": 58 }, { "epoch": 0.26727066817667045, "grad_norm": 1.2849539518356323, "learning_rate": 8.44676704559283e-05, "loss": 1.1622, "step": 59 }, { "epoch": 0.2718006795016987, "grad_norm": 1.1417158842086792, "learning_rate": 8.386407858128706e-05, "loss": 1.125, "step": 60 }, { "epoch": 0.27633069082672707, "grad_norm": 1.0794636011123657, "learning_rate": 8.32512286056924e-05, "loss": 0.9444, "step": 61 }, { "epoch": 0.2808607021517554, "grad_norm": 1.2429821491241455, "learning_rate": 8.262928807620843e-05, "loss": 0.9038, "step": 62 }, { "epoch": 0.2853907134767837, "grad_norm": 1.1989555358886719, "learning_rate": 8.199842702516583e-05, "loss": 1.0491, "step": 63 }, { "epoch": 0.289920724801812, "grad_norm": 1.1882569789886475, "learning_rate": 8.135881792367686e-05, "loss": 1.0463, "step": 64 }, { "epoch": 0.2944507361268403, "grad_norm": 1.1703907251358032, "learning_rate": 8.07106356344834e-05, "loss": 1.0251, "step": 65 }, { "epoch": 0.29898074745186864, "grad_norm": 1.2039575576782227, "learning_rate": 8.005405736415126e-05, "loss": 1.0778, "step": 66 }, { "epoch": 0.3035107587768969, "grad_norm": 1.465887427330017, "learning_rate": 7.938926261462366e-05, "loss": 1.2316, "step": 67 }, { "epoch": 0.30804077010192527, "grad_norm": 1.305050015449524, "learning_rate": 7.871643313414718e-05, "loss": 1.0009, "step": 68 }, { "epoch": 0.31257078142695355, "grad_norm": 1.4737871885299683, "learning_rate": 7.803575286758364e-05, "loss": 1.0594, "step": 69 }, { "epoch": 0.3171007927519819, "grad_norm": 1.253055214881897, "learning_rate": 7.734740790612136e-05, "loss": 0.9838, "step": 70 }, { "epoch": 0.32163080407701017, "grad_norm": 1.260217547416687, "learning_rate": 7.66515864363997e-05, "loss": 1.0468, "step": 71 }, { "epoch": 0.3261608154020385, "grad_norm": 1.305269718170166, "learning_rate": 7.594847868906076e-05, "loss": 1.1493, "step": 72 }, { "epoch": 0.33069082672706684, "grad_norm": 1.295682668685913, "learning_rate": 7.52382768867422e-05, "loss": 1.1228, "step": 73 }, { "epoch": 0.3352208380520951, "grad_norm": 1.2946406602859497, "learning_rate": 7.452117519152542e-05, "loss": 1.0911, "step": 74 }, { "epoch": 0.33975084937712347, "grad_norm": 1.3201980590820312, "learning_rate": 7.379736965185368e-05, "loss": 0.9902, "step": 75 }, { "epoch": 0.34428086070215175, "grad_norm": 1.3449866771697998, "learning_rate": 7.30670581489344e-05, "loss": 0.7981, "step": 76 }, { "epoch": 0.3488108720271801, "grad_norm": 1.356520652770996, "learning_rate": 7.233044034264034e-05, "loss": 1.1687, "step": 77 }, { "epoch": 0.35334088335220837, "grad_norm": 1.3135828971862793, "learning_rate": 7.158771761692464e-05, "loss": 0.9721, "step": 78 }, { "epoch": 0.3578708946772367, "grad_norm": 1.6086395978927612, "learning_rate": 7.083909302476453e-05, "loss": 1.3614, "step": 79 }, { "epoch": 0.362400906002265, "grad_norm": 1.6636041402816772, "learning_rate": 7.008477123264848e-05, "loss": 1.3424, "step": 80 }, { "epoch": 0.3669309173272933, "grad_norm": 1.725728988647461, "learning_rate": 6.932495846462261e-05, "loss": 1.3588, "step": 81 }, { "epoch": 0.3714609286523216, "grad_norm": 1.5616434812545776, "learning_rate": 6.855986244591104e-05, "loss": 1.0019, "step": 82 }, { "epoch": 0.37599093997734995, "grad_norm": 1.633862853050232, "learning_rate": 6.778969234612584e-05, "loss": 1.2381, "step": 83 }, { "epoch": 0.38052095130237823, "grad_norm": 1.6802394390106201, "learning_rate": 6.701465872208216e-05, "loss": 1.451, "step": 84 }, { "epoch": 0.38505096262740657, "grad_norm": 1.623841643333435, "learning_rate": 6.623497346023418e-05, "loss": 1.1578, "step": 85 }, { "epoch": 0.3895809739524349, "grad_norm": 1.532440185546875, "learning_rate": 6.545084971874738e-05, "loss": 1.2683, "step": 86 }, { "epoch": 0.3941109852774632, "grad_norm": 1.7631417512893677, "learning_rate": 6.466250186922325e-05, "loss": 1.2269, "step": 87 }, { "epoch": 0.3986409966024915, "grad_norm": 1.90073561668396, "learning_rate": 6.387014543809223e-05, "loss": 1.2522, "step": 88 }, { "epoch": 0.4031710079275198, "grad_norm": 1.7768625020980835, "learning_rate": 6.307399704769099e-05, "loss": 1.2126, "step": 89 }, { "epoch": 0.40770101925254815, "grad_norm": 1.7604447603225708, "learning_rate": 6.227427435703997e-05, "loss": 1.0491, "step": 90 }, { "epoch": 0.41223103057757643, "grad_norm": 1.8100131750106812, "learning_rate": 6.147119600233758e-05, "loss": 1.1563, "step": 91 }, { "epoch": 0.41676104190260477, "grad_norm": 1.7785227298736572, "learning_rate": 6.066498153718735e-05, "loss": 1.2355, "step": 92 }, { "epoch": 0.42129105322763305, "grad_norm": 2.1177375316619873, "learning_rate": 5.985585137257401e-05, "loss": 1.3305, "step": 93 }, { "epoch": 0.4258210645526614, "grad_norm": 2.3690457344055176, "learning_rate": 5.90440267166055e-05, "loss": 1.5447, "step": 94 }, { "epoch": 0.43035107587768967, "grad_norm": 1.9034374952316284, "learning_rate": 5.8229729514036705e-05, "loss": 1.1644, "step": 95 }, { "epoch": 0.434881087202718, "grad_norm": 2.056325912475586, "learning_rate": 5.74131823855921e-05, "loss": 1.3085, "step": 96 }, { "epoch": 0.43941109852774635, "grad_norm": 2.021763324737549, "learning_rate": 5.6594608567103456e-05, "loss": 1.3537, "step": 97 }, { "epoch": 0.44394110985277463, "grad_norm": 2.4095985889434814, "learning_rate": 5.577423184847932e-05, "loss": 1.4288, "step": 98 }, { "epoch": 0.44847112117780297, "grad_norm": 2.138651132583618, "learning_rate": 5.495227651252315e-05, "loss": 1.2902, "step": 99 }, { "epoch": 0.45300113250283125, "grad_norm": 2.890157461166382, "learning_rate": 5.4128967273616625e-05, "loss": 1.3736, "step": 100 }, { "epoch": 0.45300113250283125, "eval_loss": 1.2719783782958984, "eval_runtime": 28.1062, "eval_samples_per_second": 13.235, "eval_steps_per_second": 3.309, "step": 100 }, { "epoch": 0.4575311438278596, "grad_norm": 2.714219570159912, "learning_rate": 5.330452921628497e-05, "loss": 1.0719, "step": 101 }, { "epoch": 0.46206115515288787, "grad_norm": 1.833893060684204, "learning_rate": 5.247918773366112e-05, "loss": 0.8059, "step": 102 }, { "epoch": 0.4665911664779162, "grad_norm": 1.7362415790557861, "learning_rate": 5.165316846586541e-05, "loss": 0.8914, "step": 103 }, { "epoch": 0.4711211778029445, "grad_norm": 1.587952971458435, "learning_rate": 5.0826697238317935e-05, "loss": 1.0055, "step": 104 }, { "epoch": 0.47565118912797283, "grad_norm": 1.382871389389038, "learning_rate": 5e-05, "loss": 0.853, "step": 105 }, { "epoch": 0.4801812004530011, "grad_norm": 1.427045226097107, "learning_rate": 4.917330276168208e-05, "loss": 0.9621, "step": 106 }, { "epoch": 0.48471121177802945, "grad_norm": 1.2589244842529297, "learning_rate": 4.834683153413459e-05, "loss": 0.8243, "step": 107 }, { "epoch": 0.48924122310305773, "grad_norm": 1.1731473207473755, "learning_rate": 4.7520812266338885e-05, "loss": 0.8862, "step": 108 }, { "epoch": 0.49377123442808607, "grad_norm": 1.2672301530838013, "learning_rate": 4.669547078371504e-05, "loss": 0.8736, "step": 109 }, { "epoch": 0.4983012457531144, "grad_norm": 1.1130244731903076, "learning_rate": 4.5871032726383386e-05, "loss": 0.8778, "step": 110 }, { "epoch": 0.5028312570781427, "grad_norm": 1.2738019227981567, "learning_rate": 4.504772348747687e-05, "loss": 1.0309, "step": 111 }, { "epoch": 0.507361268403171, "grad_norm": 1.1945719718933105, "learning_rate": 4.4225768151520694e-05, "loss": 0.9288, "step": 112 }, { "epoch": 0.5118912797281994, "grad_norm": 1.153830885887146, "learning_rate": 4.3405391432896555e-05, "loss": 0.997, "step": 113 }, { "epoch": 0.5164212910532276, "grad_norm": 1.128495693206787, "learning_rate": 4.2586817614407895e-05, "loss": 0.862, "step": 114 }, { "epoch": 0.5209513023782559, "grad_norm": 1.2062228918075562, "learning_rate": 4.17702704859633e-05, "loss": 1.0172, "step": 115 }, { "epoch": 0.5254813137032842, "grad_norm": 1.1854007244110107, "learning_rate": 4.095597328339452e-05, "loss": 0.9473, "step": 116 }, { "epoch": 0.5300113250283126, "grad_norm": 1.105261206626892, "learning_rate": 4.0144148627425993e-05, "loss": 0.8747, "step": 117 }, { "epoch": 0.5345413363533409, "grad_norm": 1.1203992366790771, "learning_rate": 3.933501846281267e-05, "loss": 0.7881, "step": 118 }, { "epoch": 0.5390713476783692, "grad_norm": 1.1720339059829712, "learning_rate": 3.852880399766243e-05, "loss": 0.9116, "step": 119 }, { "epoch": 0.5436013590033975, "grad_norm": 1.220534086227417, "learning_rate": 3.772572564296005e-05, "loss": 0.7269, "step": 120 }, { "epoch": 0.5481313703284258, "grad_norm": 1.1943268775939941, "learning_rate": 3.6926002952309016e-05, "loss": 0.9006, "step": 121 }, { "epoch": 0.5526613816534541, "grad_norm": 1.1528831720352173, "learning_rate": 3.612985456190778e-05, "loss": 0.8354, "step": 122 }, { "epoch": 0.5571913929784824, "grad_norm": 1.2273133993148804, "learning_rate": 3.533749813077677e-05, "loss": 0.8162, "step": 123 }, { "epoch": 0.5617214043035108, "grad_norm": 1.2099953889846802, "learning_rate": 3.4549150281252636e-05, "loss": 0.8402, "step": 124 }, { "epoch": 0.5662514156285391, "grad_norm": 1.1575597524642944, "learning_rate": 3.3765026539765834e-05, "loss": 0.9024, "step": 125 }, { "epoch": 0.5707814269535674, "grad_norm": 1.4714939594268799, "learning_rate": 3.298534127791785e-05, "loss": 1.0202, "step": 126 }, { "epoch": 0.5753114382785957, "grad_norm": 1.4315725564956665, "learning_rate": 3.221030765387417e-05, "loss": 1.0502, "step": 127 }, { "epoch": 0.579841449603624, "grad_norm": 1.5543477535247803, "learning_rate": 3.144013755408895e-05, "loss": 1.1694, "step": 128 }, { "epoch": 0.5843714609286523, "grad_norm": 1.3127039670944214, "learning_rate": 3.0675041535377405e-05, "loss": 0.9434, "step": 129 }, { "epoch": 0.5889014722536806, "grad_norm": 1.568535327911377, "learning_rate": 2.991522876735154e-05, "loss": 1.1988, "step": 130 }, { "epoch": 0.5934314835787089, "grad_norm": 1.353140950202942, "learning_rate": 2.916090697523549e-05, "loss": 0.8726, "step": 131 }, { "epoch": 0.5979614949037373, "grad_norm": 1.4939686059951782, "learning_rate": 2.8412282383075363e-05, "loss": 1.2155, "step": 132 }, { "epoch": 0.6024915062287656, "grad_norm": 1.489733099937439, "learning_rate": 2.766955965735968e-05, "loss": 0.923, "step": 133 }, { "epoch": 0.6070215175537939, "grad_norm": 1.6046398878097534, "learning_rate": 2.693294185106562e-05, "loss": 1.1015, "step": 134 }, { "epoch": 0.6115515288788222, "grad_norm": 1.578744888305664, "learning_rate": 2.6202630348146324e-05, "loss": 1.2103, "step": 135 }, { "epoch": 0.6160815402038505, "grad_norm": 1.431128740310669, "learning_rate": 2.547882480847461e-05, "loss": 0.8943, "step": 136 }, { "epoch": 0.6206115515288788, "grad_norm": 1.9715696573257446, "learning_rate": 2.476172311325783e-05, "loss": 1.3989, "step": 137 }, { "epoch": 0.6251415628539071, "grad_norm": 1.8589072227478027, "learning_rate": 2.405152131093926e-05, "loss": 1.5329, "step": 138 }, { "epoch": 0.6296715741789355, "grad_norm": 1.8302171230316162, "learning_rate": 2.3348413563600325e-05, "loss": 1.1766, "step": 139 }, { "epoch": 0.6342015855039638, "grad_norm": 2.0135345458984375, "learning_rate": 2.2652592093878666e-05, "loss": 1.4761, "step": 140 }, { "epoch": 0.638731596828992, "grad_norm": 1.8871492147445679, "learning_rate": 2.196424713241637e-05, "loss": 0.9438, "step": 141 }, { "epoch": 0.6432616081540203, "grad_norm": 2.027357578277588, "learning_rate": 2.128356686585282e-05, "loss": 1.1982, "step": 142 }, { "epoch": 0.6477916194790487, "grad_norm": 2.037635564804077, "learning_rate": 2.061073738537635e-05, "loss": 1.1497, "step": 143 }, { "epoch": 0.652321630804077, "grad_norm": 1.9788384437561035, "learning_rate": 1.9945942635848748e-05, "loss": 1.3363, "step": 144 }, { "epoch": 0.6568516421291053, "grad_norm": 1.8388274908065796, "learning_rate": 1.928936436551661e-05, "loss": 1.2891, "step": 145 }, { "epoch": 0.6613816534541337, "grad_norm": 1.6475542783737183, "learning_rate": 1.8641182076323148e-05, "loss": 1.0235, "step": 146 }, { "epoch": 0.665911664779162, "grad_norm": 2.038667917251587, "learning_rate": 1.800157297483417e-05, "loss": 1.1535, "step": 147 }, { "epoch": 0.6704416761041903, "grad_norm": 2.430046796798706, "learning_rate": 1.7370711923791567e-05, "loss": 1.3292, "step": 148 }, { "epoch": 0.6749716874292185, "grad_norm": 2.0234296321868896, "learning_rate": 1.6748771394307585e-05, "loss": 1.3348, "step": 149 }, { "epoch": 0.6795016987542469, "grad_norm": 2.2200522422790527, "learning_rate": 1.6135921418712956e-05, "loss": 1.3524, "step": 150 }, { "epoch": 0.6795016987542469, "eval_loss": 1.0218840837478638, "eval_runtime": 28.1145, "eval_samples_per_second": 13.232, "eval_steps_per_second": 3.308, "step": 150 }, { "epoch": 0.6840317100792752, "grad_norm": 1.3126825094223022, "learning_rate": 1.553232954407171e-05, "loss": 0.8409, "step": 151 }, { "epoch": 0.6885617214043035, "grad_norm": 1.3274028301239014, "learning_rate": 1.4938160786375572e-05, "loss": 0.9251, "step": 152 }, { "epoch": 0.6930917327293318, "grad_norm": 1.5611610412597656, "learning_rate": 1.435357758543015e-05, "loss": 0.9293, "step": 153 }, { "epoch": 0.6976217440543602, "grad_norm": 1.5657310485839844, "learning_rate": 1.3778739760445552e-05, "loss": 1.0203, "step": 154 }, { "epoch": 0.7021517553793885, "grad_norm": 1.5585463047027588, "learning_rate": 1.3213804466343421e-05, "loss": 0.9957, "step": 155 }, { "epoch": 0.7066817667044167, "grad_norm": 1.3991925716400146, "learning_rate": 1.2658926150792322e-05, "loss": 1.0315, "step": 156 }, { "epoch": 0.711211778029445, "grad_norm": 1.1712048053741455, "learning_rate": 1.2114256511983274e-05, "loss": 0.8214, "step": 157 }, { "epoch": 0.7157417893544734, "grad_norm": 1.1740484237670898, "learning_rate": 1.157994445715706e-05, "loss": 0.808, "step": 158 }, { "epoch": 0.7202718006795017, "grad_norm": 1.0895429849624634, "learning_rate": 1.1056136061894384e-05, "loss": 0.7608, "step": 159 }, { "epoch": 0.72480181200453, "grad_norm": 1.1632596254348755, "learning_rate": 1.0542974530180327e-05, "loss": 0.9649, "step": 160 }, { "epoch": 0.7293318233295584, "grad_norm": 1.1991957426071167, "learning_rate": 1.0040600155253765e-05, "loss": 0.8542, "step": 161 }, { "epoch": 0.7338618346545867, "grad_norm": 1.2483800649642944, "learning_rate": 9.549150281252633e-06, "loss": 0.7725, "step": 162 }, { "epoch": 0.7383918459796149, "grad_norm": 1.2302029132843018, "learning_rate": 9.068759265665384e-06, "loss": 0.8337, "step": 163 }, { "epoch": 0.7429218573046432, "grad_norm": 1.3348345756530762, "learning_rate": 8.599558442598998e-06, "loss": 0.9638, "step": 164 }, { "epoch": 0.7474518686296716, "grad_norm": 1.2184895277023315, "learning_rate": 8.141676086873572e-06, "loss": 1.1054, "step": 165 }, { "epoch": 0.7519818799546999, "grad_norm": 1.2723814249038696, "learning_rate": 7.695237378953223e-06, "loss": 0.9325, "step": 166 }, { "epoch": 0.7565118912797282, "grad_norm": 1.2215139865875244, "learning_rate": 7.260364370723044e-06, "loss": 1.0185, "step": 167 }, { "epoch": 0.7610419026047565, "grad_norm": 1.2399837970733643, "learning_rate": 6.837175952121306e-06, "loss": 0.9689, "step": 168 }, { "epoch": 0.7655719139297849, "grad_norm": 1.1482295989990234, "learning_rate": 6.425787818636131e-06, "loss": 0.6897, "step": 169 }, { "epoch": 0.7701019252548131, "grad_norm": 1.279404640197754, "learning_rate": 6.026312439675552e-06, "loss": 0.8622, "step": 170 }, { "epoch": 0.7746319365798414, "grad_norm": 1.1450666189193726, "learning_rate": 5.6388590278194096e-06, "loss": 0.7551, "step": 171 }, { "epoch": 0.7791619479048698, "grad_norm": 1.2472995519638062, "learning_rate": 5.263533508961827e-06, "loss": 1.1059, "step": 172 }, { "epoch": 0.7836919592298981, "grad_norm": 1.2653703689575195, "learning_rate": 4.900438493352055e-06, "loss": 0.8404, "step": 173 }, { "epoch": 0.7882219705549264, "grad_norm": 1.1370460987091064, "learning_rate": 4.549673247541875e-06, "loss": 0.7427, "step": 174 }, { "epoch": 0.7927519818799547, "grad_norm": 1.285154104232788, "learning_rate": 4.2113336672471245e-06, "loss": 0.9865, "step": 175 }, { "epoch": 0.797281993204983, "grad_norm": 1.4484916925430298, "learning_rate": 3.885512251130763e-06, "loss": 0.9975, "step": 176 }, { "epoch": 0.8018120045300113, "grad_norm": 1.3047209978103638, "learning_rate": 3.5722980755146517e-06, "loss": 0.9076, "step": 177 }, { "epoch": 0.8063420158550396, "grad_norm": 1.373551607131958, "learning_rate": 3.271776770026963e-06, "loss": 1.0021, "step": 178 }, { "epoch": 0.8108720271800679, "grad_norm": 1.3371232748031616, "learning_rate": 2.9840304941919415e-06, "loss": 0.9812, "step": 179 }, { "epoch": 0.8154020385050963, "grad_norm": 1.2296382188796997, "learning_rate": 2.7091379149682685e-06, "loss": 0.7736, "step": 180 }, { "epoch": 0.8199320498301246, "grad_norm": 1.3647637367248535, "learning_rate": 2.4471741852423237e-06, "loss": 0.9978, "step": 181 }, { "epoch": 0.8244620611551529, "grad_norm": 1.5607551336288452, "learning_rate": 2.1982109232821178e-06, "loss": 1.0547, "step": 182 }, { "epoch": 0.8289920724801813, "grad_norm": 1.6037743091583252, "learning_rate": 1.962316193157593e-06, "loss": 1.0871, "step": 183 }, { "epoch": 0.8335220838052095, "grad_norm": 1.5129165649414062, "learning_rate": 1.7395544861325718e-06, "loss": 1.1008, "step": 184 }, { "epoch": 0.8380520951302378, "grad_norm": 1.5168718099594116, "learning_rate": 1.5299867030334814e-06, "loss": 0.995, "step": 185 }, { "epoch": 0.8425821064552661, "grad_norm": 1.5665178298950195, "learning_rate": 1.333670137599713e-06, "loss": 1.0104, "step": 186 }, { "epoch": 0.8471121177802945, "grad_norm": 1.6103092432022095, "learning_rate": 1.1506584608200367e-06, "loss": 1.0282, "step": 187 }, { "epoch": 0.8516421291053228, "grad_norm": 1.7237502336502075, "learning_rate": 9.810017062595322e-07, "loss": 1.0711, "step": 188 }, { "epoch": 0.8561721404303511, "grad_norm": 1.5927950143814087, "learning_rate": 8.247462563808817e-07, "loss": 1.0247, "step": 189 }, { "epoch": 0.8607021517553793, "grad_norm": 1.8922560214996338, "learning_rate": 6.819348298638839e-07, "loss": 1.1249, "step": 190 }, { "epoch": 0.8652321630804077, "grad_norm": 1.80942964553833, "learning_rate": 5.526064699265753e-07, "loss": 1.184, "step": 191 }, { "epoch": 0.869762174405436, "grad_norm": 1.6846965551376343, "learning_rate": 4.367965336512403e-07, "loss": 0.9553, "step": 192 }, { "epoch": 0.8742921857304643, "grad_norm": 2.1317317485809326, "learning_rate": 3.3453668231809286e-07, "loss": 1.3231, "step": 193 }, { "epoch": 0.8788221970554927, "grad_norm": 2.310786008834839, "learning_rate": 2.458548727494292e-07, "loss": 1.1771, "step": 194 }, { "epoch": 0.883352208380521, "grad_norm": 2.5074479579925537, "learning_rate": 1.7077534966650766e-07, "loss": 1.5084, "step": 195 }, { "epoch": 0.8878822197055493, "grad_norm": 1.8117400407791138, "learning_rate": 1.0931863906127327e-07, "loss": 1.1475, "step": 196 }, { "epoch": 0.8924122310305775, "grad_norm": 1.9792882204055786, "learning_rate": 6.150154258476315e-08, "loss": 1.2302, "step": 197 }, { "epoch": 0.8969422423556059, "grad_norm": 2.21106219291687, "learning_rate": 2.7337132953697554e-08, "loss": 1.3104, "step": 198 }, { "epoch": 0.9014722536806342, "grad_norm": 2.0842549800872803, "learning_rate": 6.834750376549792e-09, "loss": 1.1088, "step": 199 }, { "epoch": 0.9060022650056625, "grad_norm": 2.232832431793213, "learning_rate": 0.0, "loss": 1.0498, "step": 200 }, { "epoch": 0.9060022650056625, "eval_loss": 0.9684666991233826, "eval_runtime": 28.1146, "eval_samples_per_second": 13.232, "eval_steps_per_second": 3.308, "step": 200 } ], "logging_steps": 1, "max_steps": 200, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 50, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 5, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.017027157491712e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }