| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.991097922848665, |
| "eval_steps": 500, |
| "global_step": 441, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.006782534972445952, |
| "grad_norm": 7.137413186349268, |
| "learning_rate": 1.777777777777778e-06, |
| "loss": 1.1988, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.013565069944891903, |
| "grad_norm": 7.105029708492786, |
| "learning_rate": 3.555555555555556e-06, |
| "loss": 1.1926, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.020347604917337857, |
| "grad_norm": 6.975628069653311, |
| "learning_rate": 5.333333333333334e-06, |
| "loss": 1.1857, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.027130139889783807, |
| "grad_norm": 5.073330744823269, |
| "learning_rate": 7.111111111111112e-06, |
| "loss": 1.1299, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.03391267486222976, |
| "grad_norm": 3.0855210826655557, |
| "learning_rate": 8.888888888888888e-06, |
| "loss": 1.0926, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.040695209834675714, |
| "grad_norm": 5.847010298730299, |
| "learning_rate": 1.0666666666666667e-05, |
| "loss": 1.0767, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.04747774480712166, |
| "grad_norm": 5.667170104694723, |
| "learning_rate": 1.2444444444444446e-05, |
| "loss": 1.0742, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.05426027977956761, |
| "grad_norm": 5.164747156326828, |
| "learning_rate": 1.4222222222222224e-05, |
| "loss": 1.0239, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.06104281475201356, |
| "grad_norm": 4.055520122890776, |
| "learning_rate": 1.6000000000000003e-05, |
| "loss": 1.0123, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.06782534972445951, |
| "grad_norm": 2.250978802714125, |
| "learning_rate": 1.7777777777777777e-05, |
| "loss": 0.9613, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.07460788469690546, |
| "grad_norm": 2.553100004726335, |
| "learning_rate": 1.9555555555555557e-05, |
| "loss": 0.9432, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.08139041966935143, |
| "grad_norm": 2.324385101768056, |
| "learning_rate": 2.1333333333333335e-05, |
| "loss": 0.9219, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.08817295464179738, |
| "grad_norm": 2.135936923815521, |
| "learning_rate": 2.3111111111111112e-05, |
| "loss": 0.8906, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.09495548961424333, |
| "grad_norm": 1.9227504072028807, |
| "learning_rate": 2.4888888888888893e-05, |
| "loss": 0.8878, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.10173802458668928, |
| "grad_norm": 1.2751512941755272, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.8858, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.10852055955913523, |
| "grad_norm": 1.3275331460158257, |
| "learning_rate": 2.8444444444444447e-05, |
| "loss": 0.845, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.11530309453158118, |
| "grad_norm": 1.3969086473134111, |
| "learning_rate": 3.0222222222222225e-05, |
| "loss": 0.8413, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.12208562950402713, |
| "grad_norm": 1.2596435113586295, |
| "learning_rate": 3.2000000000000005e-05, |
| "loss": 0.8427, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.1288681644764731, |
| "grad_norm": 1.0919111628660803, |
| "learning_rate": 3.377777777777778e-05, |
| "loss": 0.8427, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.13565069944891903, |
| "grad_norm": 2.1726063042436725, |
| "learning_rate": 3.555555555555555e-05, |
| "loss": 0.8361, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.142433234421365, |
| "grad_norm": 1.128837783426394, |
| "learning_rate": 3.733333333333334e-05, |
| "loss": 0.8125, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.14921576939381093, |
| "grad_norm": 2.0673361900007405, |
| "learning_rate": 3.9111111111111115e-05, |
| "loss": 0.8096, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.1559983043662569, |
| "grad_norm": 1.6404001379510031, |
| "learning_rate": 4.088888888888889e-05, |
| "loss": 0.7992, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.16278083933870285, |
| "grad_norm": 1.868383646234072, |
| "learning_rate": 4.266666666666667e-05, |
| "loss": 0.8177, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.1695633743111488, |
| "grad_norm": 1.4754626309706587, |
| "learning_rate": 4.444444444444445e-05, |
| "loss": 0.8192, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.17634590928359475, |
| "grad_norm": 1.9030805002187652, |
| "learning_rate": 4.6222222222222224e-05, |
| "loss": 0.8082, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.1831284442560407, |
| "grad_norm": 1.0440198672277559, |
| "learning_rate": 4.8e-05, |
| "loss": 0.7944, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.18991097922848665, |
| "grad_norm": 2.596599541792705, |
| "learning_rate": 4.9777777777777785e-05, |
| "loss": 0.8017, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.1966935142009326, |
| "grad_norm": 1.9546381445372338, |
| "learning_rate": 5.155555555555556e-05, |
| "loss": 0.7945, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.20347604917337855, |
| "grad_norm": 1.751189632085305, |
| "learning_rate": 5.333333333333333e-05, |
| "loss": 0.7945, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.2102585841458245, |
| "grad_norm": 1.7584486355585562, |
| "learning_rate": 5.511111111111112e-05, |
| "loss": 0.7844, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.21704111911827045, |
| "grad_norm": 1.0584793475130616, |
| "learning_rate": 5.6888888888888895e-05, |
| "loss": 0.7812, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.22382365409071642, |
| "grad_norm": 2.196048787381429, |
| "learning_rate": 5.8666666666666665e-05, |
| "loss": 0.8027, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.23060618906316235, |
| "grad_norm": 1.6456458122301723, |
| "learning_rate": 6.044444444444445e-05, |
| "loss": 0.7876, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.23738872403560832, |
| "grad_norm": 1.9008326288925532, |
| "learning_rate": 6.222222222222223e-05, |
| "loss": 0.7992, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.24417125900805425, |
| "grad_norm": 2.7375256976568405, |
| "learning_rate": 6.400000000000001e-05, |
| "loss": 0.7788, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.2509537939805002, |
| "grad_norm": 1.7512537641089396, |
| "learning_rate": 6.577777777777777e-05, |
| "loss": 0.7907, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.2577363289529462, |
| "grad_norm": 2.0707248009544568, |
| "learning_rate": 6.755555555555557e-05, |
| "loss": 0.7894, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.2645188639253921, |
| "grad_norm": 2.5069295004911054, |
| "learning_rate": 6.933333333333334e-05, |
| "loss": 0.7795, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.27130139889783805, |
| "grad_norm": 1.769408811743829, |
| "learning_rate": 7.11111111111111e-05, |
| "loss": 0.7823, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.278083933870284, |
| "grad_norm": 3.264548001005339, |
| "learning_rate": 7.28888888888889e-05, |
| "loss": 0.7768, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.28486646884273, |
| "grad_norm": 2.4056236262328885, |
| "learning_rate": 7.466666666666667e-05, |
| "loss": 0.7841, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.29164900381517594, |
| "grad_norm": 2.5559829162282597, |
| "learning_rate": 7.644444444444445e-05, |
| "loss": 0.7818, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.29843153878762185, |
| "grad_norm": 1.9829026032666919, |
| "learning_rate": 7.822222222222223e-05, |
| "loss": 0.7641, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.3052140737600678, |
| "grad_norm": 1.9013014130141392, |
| "learning_rate": 8e-05, |
| "loss": 0.7673, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.3119966087325138, |
| "grad_norm": 1.3239983020470356, |
| "learning_rate": 7.999874125693333e-05, |
| "loss": 0.7792, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.31877914370495974, |
| "grad_norm": 2.093841653755095, |
| "learning_rate": 7.999496510695501e-05, |
| "loss": 0.7703, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.3255616786774057, |
| "grad_norm": 2.750870577254638, |
| "learning_rate": 7.998867178772517e-05, |
| "loss": 0.7712, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.3323442136498516, |
| "grad_norm": 1.4125603524830153, |
| "learning_rate": 7.997986169532741e-05, |
| "loss": 0.7637, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.3391267486222976, |
| "grad_norm": 2.524537816103899, |
| "learning_rate": 7.996853538424387e-05, |
| "loss": 0.7827, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.34590928359474354, |
| "grad_norm": 1.7058887152668756, |
| "learning_rate": 7.995469356732033e-05, |
| "loss": 0.7689, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.3526918185671895, |
| "grad_norm": 1.9625755711375161, |
| "learning_rate": 7.993833711572133e-05, |
| "loss": 0.7764, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.3594743535396354, |
| "grad_norm": 1.9608607243300042, |
| "learning_rate": 7.991946705887539e-05, |
| "loss": 0.7583, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.3662568885120814, |
| "grad_norm": 1.5775528005817034, |
| "learning_rate": 7.989808458441014e-05, |
| "loss": 0.755, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.37303942348452734, |
| "grad_norm": 1.9419695989547225, |
| "learning_rate": 7.98741910380777e-05, |
| "loss": 0.7467, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.3798219584569733, |
| "grad_norm": 1.4275820221414393, |
| "learning_rate": 7.984778792366983e-05, |
| "loss": 0.7476, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.38660449342941927, |
| "grad_norm": 1.4199172439377408, |
| "learning_rate": 7.981887690292339e-05, |
| "loss": 0.7477, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.3933870284018652, |
| "grad_norm": 1.8664000400986023, |
| "learning_rate": 7.978745979541574e-05, |
| "loss": 0.7456, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.40016956337431114, |
| "grad_norm": 1.0582378742113208, |
| "learning_rate": 7.975353857845017e-05, |
| "loss": 0.7453, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.4069520983467571, |
| "grad_norm": 2.6856740014169875, |
| "learning_rate": 7.971711538693153e-05, |
| "loss": 0.7652, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.41373463331920307, |
| "grad_norm": 2.2330393685649894, |
| "learning_rate": 7.967819251323182e-05, |
| "loss": 0.7468, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.420517168291649, |
| "grad_norm": 1.3266578444651185, |
| "learning_rate": 7.963677240704588e-05, |
| "loss": 0.7476, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.42729970326409494, |
| "grad_norm": 1.337167566676173, |
| "learning_rate": 7.959285767523732e-05, |
| "loss": 0.7486, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.4340822382365409, |
| "grad_norm": 1.7695088306451494, |
| "learning_rate": 7.954645108167432e-05, |
| "loss": 0.7542, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.44086477320898687, |
| "grad_norm": 1.3827391907399533, |
| "learning_rate": 7.949755554705577e-05, |
| "loss": 0.7517, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.44764730818143283, |
| "grad_norm": 1.354435012969295, |
| "learning_rate": 7.944617414872747e-05, |
| "loss": 0.7589, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.45442984315387874, |
| "grad_norm": 0.9204507675004047, |
| "learning_rate": 7.939231012048833e-05, |
| "loss": 0.7345, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.4612123781263247, |
| "grad_norm": 1.4388297723137893, |
| "learning_rate": 7.933596685238697e-05, |
| "loss": 0.7478, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.46799491309877067, |
| "grad_norm": 1.6753627019958215, |
| "learning_rate": 7.927714789050826e-05, |
| "loss": 0.732, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.47477744807121663, |
| "grad_norm": 0.919762144543453, |
| "learning_rate": 7.921585693675029e-05, |
| "loss": 0.7385, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.48155998304366254, |
| "grad_norm": 2.2697133943634427, |
| "learning_rate": 7.915209784859116e-05, |
| "loss": 0.7334, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.4883425180161085, |
| "grad_norm": 1.2442834324078464, |
| "learning_rate": 7.908587463884638e-05, |
| "loss": 0.7501, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.49512505298855447, |
| "grad_norm": 2.0455012466197995, |
| "learning_rate": 7.90171914754163e-05, |
| "loss": 0.7323, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.5019075879610004, |
| "grad_norm": 1.466104092423094, |
| "learning_rate": 7.894605268102365e-05, |
| "loss": 0.7284, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.5086901229334464, |
| "grad_norm": 1.4621681457693403, |
| "learning_rate": 7.887246273294167e-05, |
| "loss": 0.7345, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.5154726579058924, |
| "grad_norm": 1.5491014708068618, |
| "learning_rate": 7.87964262627122e-05, |
| "loss": 0.7346, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.5222551928783383, |
| "grad_norm": 1.1091999421133278, |
| "learning_rate": 7.871794805585427e-05, |
| "loss": 0.7313, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.5290377278507842, |
| "grad_norm": 1.4167418001519376, |
| "learning_rate": 7.863703305156273e-05, |
| "loss": 0.7217, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.5358202628232301, |
| "grad_norm": 1.2449973723473307, |
| "learning_rate": 7.855368634239769e-05, |
| "loss": 0.7338, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.5426027977956761, |
| "grad_norm": 1.2279144181708292, |
| "learning_rate": 7.846791317396373e-05, |
| "loss": 0.7275, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.5493853327681221, |
| "grad_norm": 1.1622952382317697, |
| "learning_rate": 7.837971894457991e-05, |
| "loss": 0.7459, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.556167867740568, |
| "grad_norm": 1.16239709067931, |
| "learning_rate": 7.828910920493995e-05, |
| "loss": 0.7176, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.562950402713014, |
| "grad_norm": 1.4393853572209945, |
| "learning_rate": 7.819608965776295e-05, |
| "loss": 0.7378, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.56973293768546, |
| "grad_norm": 0.8723694343578424, |
| "learning_rate": 7.810066615743443e-05, |
| "loss": 0.7235, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.5765154726579059, |
| "grad_norm": 1.4627194527050393, |
| "learning_rate": 7.800284470963783e-05, |
| "loss": 0.7245, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.5832980076303519, |
| "grad_norm": 1.5007942983664493, |
| "learning_rate": 7.790263147097661e-05, |
| "loss": 0.7358, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.5900805426027977, |
| "grad_norm": 1.0492585064630457, |
| "learning_rate": 7.780003274858674e-05, |
| "loss": 0.7222, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.5968630775752437, |
| "grad_norm": 1.4048579687963796, |
| "learning_rate": 7.769505499973977e-05, |
| "loss": 0.716, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.6036456125476897, |
| "grad_norm": 1.0188682627658208, |
| "learning_rate": 7.758770483143634e-05, |
| "loss": 0.7268, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.6104281475201356, |
| "grad_norm": 1.1982575150619237, |
| "learning_rate": 7.747798899999048e-05, |
| "loss": 0.7407, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.6172106824925816, |
| "grad_norm": 0.9182390822096019, |
| "learning_rate": 7.736591441060427e-05, |
| "loss": 0.7326, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.6239932174650276, |
| "grad_norm": 1.4475324042423592, |
| "learning_rate": 7.725148811693336e-05, |
| "loss": 0.7336, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.6307757524374735, |
| "grad_norm": 1.1065799129734808, |
| "learning_rate": 7.71347173206429e-05, |
| "loss": 0.7173, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.6375582874099195, |
| "grad_norm": 1.237603778756235, |
| "learning_rate": 7.701560937095445e-05, |
| "loss": 0.7303, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.6443408223823655, |
| "grad_norm": 1.1120475384075377, |
| "learning_rate": 7.689417176418327e-05, |
| "loss": 0.7185, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.6511233573548114, |
| "grad_norm": 1.3783747990085908, |
| "learning_rate": 7.677041214326663e-05, |
| "loss": 0.7357, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.6579058923272573, |
| "grad_norm": 1.0644720482330112, |
| "learning_rate": 7.664433829728279e-05, |
| "loss": 0.7155, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.6646884272997032, |
| "grad_norm": 1.4446573818002777, |
| "learning_rate": 7.651595816096071e-05, |
| "loss": 0.7109, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.6714709622721492, |
| "grad_norm": 1.0329689312285972, |
| "learning_rate": 7.638527981418075e-05, |
| "loss": 0.7333, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.6782534972445952, |
| "grad_norm": 1.2999144870004373, |
| "learning_rate": 7.625231148146601e-05, |
| "loss": 0.7097, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.6850360322170411, |
| "grad_norm": 1.1368825847990134, |
| "learning_rate": 7.611706153146486e-05, |
| "loss": 0.7128, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.6918185671894871, |
| "grad_norm": 1.1174260276126868, |
| "learning_rate": 7.597953847642413e-05, |
| "loss": 0.7241, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.698601102161933, |
| "grad_norm": 0.8874864224204202, |
| "learning_rate": 7.583975097165344e-05, |
| "loss": 0.7189, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.705383637134379, |
| "grad_norm": 0.6432301304209681, |
| "learning_rate": 7.56977078149804e-05, |
| "loss": 0.7145, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.712166172106825, |
| "grad_norm": 0.9183282978967895, |
| "learning_rate": 7.555341794619695e-05, |
| "loss": 0.7107, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.7189487070792708, |
| "grad_norm": 1.0095498465315442, |
| "learning_rate": 7.540689044649666e-05, |
| "loss": 0.703, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.7257312420517168, |
| "grad_norm": 1.26183318694208, |
| "learning_rate": 7.525813453790328e-05, |
| "loss": 0.7127, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.7325137770241628, |
| "grad_norm": 0.6972659136766679, |
| "learning_rate": 7.510715958269023e-05, |
| "loss": 0.7052, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.7392963119966087, |
| "grad_norm": 0.7185478814495369, |
| "learning_rate": 7.49539750827914e-05, |
| "loss": 0.7109, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.7460788469690547, |
| "grad_norm": 0.9112632508609093, |
| "learning_rate": 7.479859067920317e-05, |
| "loss": 0.7037, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.7528613819415007, |
| "grad_norm": 0.8720981679213001, |
| "learning_rate": 7.464101615137756e-05, |
| "loss": 0.7075, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.7596439169139466, |
| "grad_norm": 1.0271120509299871, |
| "learning_rate": 7.448126141660678e-05, |
| "loss": 0.7204, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.7664264518863926, |
| "grad_norm": 1.016829719094229, |
| "learning_rate": 7.431933652939909e-05, |
| "loss": 0.7129, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.7732089868588385, |
| "grad_norm": 1.0366544063351542, |
| "learning_rate": 7.415525168084593e-05, |
| "loss": 0.708, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.7799915218312844, |
| "grad_norm": 1.144762721052778, |
| "learning_rate": 7.398901719798059e-05, |
| "loss": 0.7102, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.7867740568037304, |
| "grad_norm": 1.0391721353507355, |
| "learning_rate": 7.382064354312818e-05, |
| "loss": 0.7098, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.7935565917761763, |
| "grad_norm": 0.9632167130265382, |
| "learning_rate": 7.365014131324725e-05, |
| "loss": 0.7029, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.8003391267486223, |
| "grad_norm": 1.227602079185617, |
| "learning_rate": 7.34775212392628e-05, |
| "loss": 0.6972, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.8071216617210683, |
| "grad_norm": 0.8826250475798045, |
| "learning_rate": 7.330279418539086e-05, |
| "loss": 0.7018, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.8139041966935142, |
| "grad_norm": 0.5567261188617991, |
| "learning_rate": 7.312597114845483e-05, |
| "loss": 0.6981, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.8206867316659602, |
| "grad_norm": 0.8765091262196973, |
| "learning_rate": 7.294706325719331e-05, |
| "loss": 0.7195, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.8274692666384061, |
| "grad_norm": 0.9742989954487304, |
| "learning_rate": 7.276608177155968e-05, |
| "loss": 0.7033, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.8342518016108521, |
| "grad_norm": 0.9120816045559339, |
| "learning_rate": 7.258303808201343e-05, |
| "loss": 0.6984, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.841034336583298, |
| "grad_norm": 0.8325780076464987, |
| "learning_rate": 7.239794370880335e-05, |
| "loss": 0.6921, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.8478168715557439, |
| "grad_norm": 0.6155944829472941, |
| "learning_rate": 7.221081030124235e-05, |
| "loss": 0.6988, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.8545994065281899, |
| "grad_norm": 0.7511218044788109, |
| "learning_rate": 7.202164963697442e-05, |
| "loss": 0.6972, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.8613819415006359, |
| "grad_norm": 0.7293772573099593, |
| "learning_rate": 7.183047362123329e-05, |
| "loss": 0.7004, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.8681644764730818, |
| "grad_norm": 0.46098198038922594, |
| "learning_rate": 7.163729428609318e-05, |
| "loss": 0.6956, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.8749470114455278, |
| "grad_norm": 0.7377483747880154, |
| "learning_rate": 7.144212378971151e-05, |
| "loss": 0.7089, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.8817295464179737, |
| "grad_norm": 0.8138881211490862, |
| "learning_rate": 7.124497441556374e-05, |
| "loss": 0.7025, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.8885120813904197, |
| "grad_norm": 0.7428361558045745, |
| "learning_rate": 7.104585857167028e-05, |
| "loss": 0.7093, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.8952946163628657, |
| "grad_norm": 0.8252776283861111, |
| "learning_rate": 7.084478878981552e-05, |
| "loss": 0.7, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.9020771513353115, |
| "grad_norm": 0.8110674797944571, |
| "learning_rate": 7.064177772475912e-05, |
| "loss": 0.695, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.9088596863077575, |
| "grad_norm": 0.6529878802072909, |
| "learning_rate": 7.043683815343967e-05, |
| "loss": 0.6938, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.9156422212802034, |
| "grad_norm": 0.7848292838747539, |
| "learning_rate": 7.022998297417034e-05, |
| "loss": 0.6924, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.9224247562526494, |
| "grad_norm": 1.1207835540553364, |
| "learning_rate": 7.00212252058273e-05, |
| "loss": 0.6988, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.9292072912250954, |
| "grad_norm": 0.9343138621109743, |
| "learning_rate": 6.98105779870302e-05, |
| "loss": 0.6966, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.9359898261975413, |
| "grad_norm": 0.6897906744897139, |
| "learning_rate": 6.959805457531536e-05, |
| "loss": 0.6995, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.9427723611699873, |
| "grad_norm": 0.48700498757406885, |
| "learning_rate": 6.938366834630133e-05, |
| "loss": 0.689, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.9495548961424333, |
| "grad_norm": 0.460000455610998, |
| "learning_rate": 6.916743279284709e-05, |
| "loss": 0.699, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.9563374311148792, |
| "grad_norm": 0.6228357519746542, |
| "learning_rate": 6.89493615242028e-05, |
| "loss": 0.6987, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.9631199660873251, |
| "grad_norm": 0.6004863488992443, |
| "learning_rate": 6.872946826515338e-05, |
| "loss": 0.7078, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.969902501059771, |
| "grad_norm": 0.38322005162459927, |
| "learning_rate": 6.850776685515453e-05, |
| "loss": 0.6951, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.976685036032217, |
| "grad_norm": 0.4778442733404826, |
| "learning_rate": 6.828427124746191e-05, |
| "loss": 0.7059, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.983467571004663, |
| "grad_norm": 0.4078839327460854, |
| "learning_rate": 6.805899550825285e-05, |
| "loss": 0.7034, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.9902501059771089, |
| "grad_norm": 0.3930697959309545, |
| "learning_rate": 6.78319538157411e-05, |
| "loss": 0.688, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.9970326409495549, |
| "grad_norm": 0.471709925781642, |
| "learning_rate": 6.760316045928449e-05, |
| "loss": 0.6996, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.0038151759220009, |
| "grad_norm": 0.5236980599221154, |
| "learning_rate": 6.737262983848554e-05, |
| "loss": 1.0973, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.0105977108944468, |
| "grad_norm": 0.7531817832475667, |
| "learning_rate": 6.714037646228529e-05, |
| "loss": 0.6724, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.0173802458668928, |
| "grad_norm": 0.8924803206824311, |
| "learning_rate": 6.690641494805011e-05, |
| "loss": 0.6904, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.0241627808393388, |
| "grad_norm": 0.9578752031258555, |
| "learning_rate": 6.667076002065168e-05, |
| "loss": 0.6763, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.0309453158117847, |
| "grad_norm": 0.9584132989076937, |
| "learning_rate": 6.643342651154028e-05, |
| "loss": 0.6928, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.0377278507842307, |
| "grad_norm": 0.6631156370651635, |
| "learning_rate": 6.619442935781141e-05, |
| "loss": 0.6914, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.0445103857566767, |
| "grad_norm": 0.5101036583421924, |
| "learning_rate": 6.595378360126555e-05, |
| "loss": 0.6778, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.0512929207291226, |
| "grad_norm": 0.6836210135721171, |
| "learning_rate": 6.571150438746157e-05, |
| "loss": 0.6776, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.0580754557015684, |
| "grad_norm": 0.5297506092844936, |
| "learning_rate": 6.546760696476354e-05, |
| "loss": 0.6792, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.0648579906740143, |
| "grad_norm": 0.6440705198875945, |
| "learning_rate": 6.52221066833809e-05, |
| "loss": 0.6813, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.0716405256464603, |
| "grad_norm": 0.5649902540029019, |
| "learning_rate": 6.497501899440255e-05, |
| "loss": 0.6758, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.0784230606189062, |
| "grad_norm": 0.5451745835283485, |
| "learning_rate": 6.472635944882421e-05, |
| "loss": 0.6846, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.0852055955913522, |
| "grad_norm": 0.6742930881134588, |
| "learning_rate": 6.447614369656987e-05, |
| "loss": 0.6659, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.0919881305637982, |
| "grad_norm": 0.40739982620953924, |
| "learning_rate": 6.422438748550667e-05, |
| "loss": 0.6688, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.0987706655362441, |
| "grad_norm": 0.520705689346025, |
| "learning_rate": 6.397110666045388e-05, |
| "loss": 0.675, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.10555320050869, |
| "grad_norm": 0.4809784234416612, |
| "learning_rate": 6.371631716218563e-05, |
| "loss": 0.6714, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.112335735481136, |
| "grad_norm": 0.6328561377367079, |
| "learning_rate": 6.346003502642762e-05, |
| "loss": 0.6818, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.119118270453582, |
| "grad_norm": 0.6964171119263963, |
| "learning_rate": 6.320227638284793e-05, |
| "loss": 0.6663, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.125900805426028, |
| "grad_norm": 0.5303565775080002, |
| "learning_rate": 6.294305745404185e-05, |
| "loss": 0.6784, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.132683340398474, |
| "grad_norm": 0.44306420532590796, |
| "learning_rate": 6.268239455451083e-05, |
| "loss": 0.6867, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.13946587537092, |
| "grad_norm": 0.3414530565574301, |
| "learning_rate": 6.242030408963576e-05, |
| "loss": 0.6672, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.1462484103433659, |
| "grad_norm": 0.40031871777699174, |
| "learning_rate": 6.215680255464442e-05, |
| "loss": 0.6651, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.1530309453158119, |
| "grad_norm": 0.277522312209641, |
| "learning_rate": 6.18919065335733e-05, |
| "loss": 0.6669, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.1598134802882578, |
| "grad_norm": 0.35702823672803347, |
| "learning_rate": 6.162563269822391e-05, |
| "loss": 0.6636, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.1665960152607038, |
| "grad_norm": 0.29212513135494134, |
| "learning_rate": 6.135799780711345e-05, |
| "loss": 0.675, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.1733785502331497, |
| "grad_norm": 0.32422232694491343, |
| "learning_rate": 6.10890187044201e-05, |
| "loss": 0.6833, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.1801610852055955, |
| "grad_norm": 0.34358927362334013, |
| "learning_rate": 6.0818712318922894e-05, |
| "loss": 0.6682, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.1869436201780414, |
| "grad_norm": 0.24681935714411127, |
| "learning_rate": 6.054709566293627e-05, |
| "loss": 0.683, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.1937261551504874, |
| "grad_norm": 0.3149112104902362, |
| "learning_rate": 6.0274185831239325e-05, |
| "loss": 0.6638, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.2005086901229334, |
| "grad_norm": 0.34401534901495145, |
| "learning_rate": 6.000000000000001e-05, |
| "loss": 0.6725, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.2072912250953793, |
| "grad_norm": 0.323484399780841, |
| "learning_rate": 5.972455542569402e-05, |
| "loss": 0.6692, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.2140737600678253, |
| "grad_norm": 0.30391385199629817, |
| "learning_rate": 5.944786944401875e-05, |
| "loss": 0.6623, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.2208562950402713, |
| "grad_norm": 0.26879513471231475, |
| "learning_rate": 5.916995946880228e-05, |
| "loss": 0.6842, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.2276388300127172, |
| "grad_norm": 0.29492095831429094, |
| "learning_rate": 5.889084299090732e-05, |
| "loss": 0.6786, |
| "step": 181 |
| }, |
| { |
| "epoch": 1.2344213649851632, |
| "grad_norm": 0.3788666762600595, |
| "learning_rate": 5.861053757713043e-05, |
| "loss": 0.6726, |
| "step": 182 |
| }, |
| { |
| "epoch": 1.2412038999576092, |
| "grad_norm": 0.4398437908833291, |
| "learning_rate": 5.832906086909642e-05, |
| "loss": 0.6742, |
| "step": 183 |
| }, |
| { |
| "epoch": 1.2479864349300551, |
| "grad_norm": 0.5383171396673415, |
| "learning_rate": 5.8046430582148034e-05, |
| "loss": 0.6636, |
| "step": 184 |
| }, |
| { |
| "epoch": 1.254768969902501, |
| "grad_norm": 0.6287710260233805, |
| "learning_rate": 5.776266450423097e-05, |
| "loss": 0.6745, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.261551504874947, |
| "grad_norm": 0.692244364864307, |
| "learning_rate": 5.747778049477438e-05, |
| "loss": 0.692, |
| "step": 186 |
| }, |
| { |
| "epoch": 1.268334039847393, |
| "grad_norm": 0.5173002490296247, |
| "learning_rate": 5.7191796483566874e-05, |
| "loss": 0.6851, |
| "step": 187 |
| }, |
| { |
| "epoch": 1.275116574819839, |
| "grad_norm": 0.3170872618606908, |
| "learning_rate": 5.6904730469627985e-05, |
| "loss": 0.6694, |
| "step": 188 |
| }, |
| { |
| "epoch": 1.281899109792285, |
| "grad_norm": 0.3605888836619855, |
| "learning_rate": 5.661660052007547e-05, |
| "loss": 0.6674, |
| "step": 189 |
| }, |
| { |
| "epoch": 1.288681644764731, |
| "grad_norm": 0.5359207812719183, |
| "learning_rate": 5.632742476898813e-05, |
| "loss": 0.6574, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.2954641797371766, |
| "grad_norm": 0.5493258534047062, |
| "learning_rate": 5.6037221416264554e-05, |
| "loss": 0.6651, |
| "step": 191 |
| }, |
| { |
| "epoch": 1.3022467147096228, |
| "grad_norm": 0.3583558907183342, |
| "learning_rate": 5.574600872647766e-05, |
| "loss": 0.6741, |
| "step": 192 |
| }, |
| { |
| "epoch": 1.3090292496820686, |
| "grad_norm": 0.38616308348313483, |
| "learning_rate": 5.5453805027725145e-05, |
| "loss": 0.6735, |
| "step": 193 |
| }, |
| { |
| "epoch": 1.3158117846545148, |
| "grad_norm": 0.5232887516865083, |
| "learning_rate": 5.516062871047602e-05, |
| "loss": 0.6648, |
| "step": 194 |
| }, |
| { |
| "epoch": 1.3225943196269605, |
| "grad_norm": 0.5001634228373406, |
| "learning_rate": 5.48664982264131e-05, |
| "loss": 0.682, |
| "step": 195 |
| }, |
| { |
| "epoch": 1.3293768545994065, |
| "grad_norm": 0.4545741719631783, |
| "learning_rate": 5.4571432087271775e-05, |
| "loss": 0.6748, |
| "step": 196 |
| }, |
| { |
| "epoch": 1.3361593895718524, |
| "grad_norm": 0.42741557026919574, |
| "learning_rate": 5.427544886367488e-05, |
| "loss": 0.677, |
| "step": 197 |
| }, |
| { |
| "epoch": 1.3429419245442984, |
| "grad_norm": 0.31094524488846675, |
| "learning_rate": 5.397856718396394e-05, |
| "loss": 0.6652, |
| "step": 198 |
| }, |
| { |
| "epoch": 1.3497244595167444, |
| "grad_norm": 0.24780800036037465, |
| "learning_rate": 5.368080573302676e-05, |
| "loss": 0.6664, |
| "step": 199 |
| }, |
| { |
| "epoch": 1.3565069944891903, |
| "grad_norm": 0.3181527932049964, |
| "learning_rate": 5.3382183251121415e-05, |
| "loss": 0.6675, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.3632895294616363, |
| "grad_norm": 0.29745491627360243, |
| "learning_rate": 5.3082718532696874e-05, |
| "loss": 0.6706, |
| "step": 201 |
| }, |
| { |
| "epoch": 1.3700720644340822, |
| "grad_norm": 0.2584376200240668, |
| "learning_rate": 5.2782430425210004e-05, |
| "loss": 0.6631, |
| "step": 202 |
| }, |
| { |
| "epoch": 1.3768545994065282, |
| "grad_norm": 0.28291959973268804, |
| "learning_rate": 5.2481337827939486e-05, |
| "loss": 0.665, |
| "step": 203 |
| }, |
| { |
| "epoch": 1.3836371343789742, |
| "grad_norm": 0.222511659896141, |
| "learning_rate": 5.217945969079629e-05, |
| "loss": 0.6737, |
| "step": 204 |
| }, |
| { |
| "epoch": 1.3904196693514201, |
| "grad_norm": 0.1904985337395551, |
| "learning_rate": 5.1876815013131e-05, |
| "loss": 0.6773, |
| "step": 205 |
| }, |
| { |
| "epoch": 1.397202204323866, |
| "grad_norm": 0.2556162510484352, |
| "learning_rate": 5.157342284253812e-05, |
| "loss": 0.674, |
| "step": 206 |
| }, |
| { |
| "epoch": 1.403984739296312, |
| "grad_norm": 0.2384701949692735, |
| "learning_rate": 5.1269302273657195e-05, |
| "loss": 0.6817, |
| "step": 207 |
| }, |
| { |
| "epoch": 1.410767274268758, |
| "grad_norm": 0.20411764282852152, |
| "learning_rate": 5.0964472446971114e-05, |
| "loss": 0.6536, |
| "step": 208 |
| }, |
| { |
| "epoch": 1.417549809241204, |
| "grad_norm": 0.20834830910241572, |
| "learning_rate": 5.06589525476014e-05, |
| "loss": 0.6694, |
| "step": 209 |
| }, |
| { |
| "epoch": 1.4243323442136497, |
| "grad_norm": 0.22097226372451154, |
| "learning_rate": 5.0352761804100835e-05, |
| "loss": 0.6607, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.431114879186096, |
| "grad_norm": 0.23524605314054492, |
| "learning_rate": 5.004591948724317e-05, |
| "loss": 0.6591, |
| "step": 211 |
| }, |
| { |
| "epoch": 1.4378974141585417, |
| "grad_norm": 0.32464444157336164, |
| "learning_rate": 4.9738444908810365e-05, |
| "loss": 0.6698, |
| "step": 212 |
| }, |
| { |
| "epoch": 1.4446799491309876, |
| "grad_norm": 0.32718322363813607, |
| "learning_rate": 4.94303574203771e-05, |
| "loss": 0.6543, |
| "step": 213 |
| }, |
| { |
| "epoch": 1.4514624841034336, |
| "grad_norm": 0.2459546193619208, |
| "learning_rate": 4.9121676412092874e-05, |
| "loss": 0.6696, |
| "step": 214 |
| }, |
| { |
| "epoch": 1.4582450190758796, |
| "grad_norm": 0.1878400255657447, |
| "learning_rate": 4.881242131146163e-05, |
| "loss": 0.6745, |
| "step": 215 |
| }, |
| { |
| "epoch": 1.4650275540483255, |
| "grad_norm": 0.21340996885255564, |
| "learning_rate": 4.8502611582119065e-05, |
| "loss": 0.6702, |
| "step": 216 |
| }, |
| { |
| "epoch": 1.4718100890207715, |
| "grad_norm": 0.1930500360689927, |
| "learning_rate": 4.819226672260763e-05, |
| "loss": 0.6739, |
| "step": 217 |
| }, |
| { |
| "epoch": 1.4785926239932174, |
| "grad_norm": 0.19466196179996254, |
| "learning_rate": 4.788140626514933e-05, |
| "loss": 0.6678, |
| "step": 218 |
| }, |
| { |
| "epoch": 1.4853751589656634, |
| "grad_norm": 0.1802358321365316, |
| "learning_rate": 4.7570049774416414e-05, |
| "loss": 0.6578, |
| "step": 219 |
| }, |
| { |
| "epoch": 1.4921576939381094, |
| "grad_norm": 0.21190165053247084, |
| "learning_rate": 4.7258216846300106e-05, |
| "loss": 0.6687, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.4989402289105553, |
| "grad_norm": 0.2025332247212839, |
| "learning_rate": 4.694592710667723e-05, |
| "loss": 0.6695, |
| "step": 221 |
| }, |
| { |
| "epoch": 1.5057227638830013, |
| "grad_norm": 0.18244954655230586, |
| "learning_rate": 4.663320021017497e-05, |
| "loss": 0.6655, |
| "step": 222 |
| }, |
| { |
| "epoch": 1.5125052988554473, |
| "grad_norm": 0.2269278837073207, |
| "learning_rate": 4.6320055838934e-05, |
| "loss": 0.6727, |
| "step": 223 |
| }, |
| { |
| "epoch": 1.5192878338278932, |
| "grad_norm": 0.2259027791570099, |
| "learning_rate": 4.6006513701369616e-05, |
| "loss": 0.6606, |
| "step": 224 |
| }, |
| { |
| "epoch": 1.526070368800339, |
| "grad_norm": 0.2477760784045823, |
| "learning_rate": 4.5692593530931416e-05, |
| "loss": 0.6759, |
| "step": 225 |
| }, |
| { |
| "epoch": 1.5328529037727852, |
| "grad_norm": 0.2603331861037767, |
| "learning_rate": 4.5378315084861276e-05, |
| "loss": 0.6624, |
| "step": 226 |
| }, |
| { |
| "epoch": 1.539635438745231, |
| "grad_norm": 0.2506086748418563, |
| "learning_rate": 4.506369814294998e-05, |
| "loss": 0.6599, |
| "step": 227 |
| }, |
| { |
| "epoch": 1.546417973717677, |
| "grad_norm": 0.26765221421169105, |
| "learning_rate": 4.474876250629221e-05, |
| "loss": 0.6764, |
| "step": 228 |
| }, |
| { |
| "epoch": 1.5532005086901228, |
| "grad_norm": 0.17936756384362929, |
| "learning_rate": 4.4433527996040443e-05, |
| "loss": 0.6643, |
| "step": 229 |
| }, |
| { |
| "epoch": 1.559983043662569, |
| "grad_norm": 0.1830215522438223, |
| "learning_rate": 4.411801445215739e-05, |
| "loss": 0.6644, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.5667655786350148, |
| "grad_norm": 0.20385050238488933, |
| "learning_rate": 4.38022417321673e-05, |
| "loss": 0.6559, |
| "step": 231 |
| }, |
| { |
| "epoch": 1.573548113607461, |
| "grad_norm": 0.21314261341134186, |
| "learning_rate": 4.348622970990634e-05, |
| "loss": 0.6733, |
| "step": 232 |
| }, |
| { |
| "epoch": 1.5803306485799067, |
| "grad_norm": 0.23351272270245574, |
| "learning_rate": 4.316999827427154e-05, |
| "loss": 0.6609, |
| "step": 233 |
| }, |
| { |
| "epoch": 1.5871131835523526, |
| "grad_norm": 0.24621369232426915, |
| "learning_rate": 4.2853567327969296e-05, |
| "loss": 0.6681, |
| "step": 234 |
| }, |
| { |
| "epoch": 1.5938957185247986, |
| "grad_norm": 0.30216358602341464, |
| "learning_rate": 4.2536956786262585e-05, |
| "loss": 0.6822, |
| "step": 235 |
| }, |
| { |
| "epoch": 1.6006782534972446, |
| "grad_norm": 0.2721373138715926, |
| "learning_rate": 4.222018657571761e-05, |
| "loss": 0.6636, |
| "step": 236 |
| }, |
| { |
| "epoch": 1.6074607884696905, |
| "grad_norm": 0.23208696219893624, |
| "learning_rate": 4.19032766329497e-05, |
| "loss": 0.6701, |
| "step": 237 |
| }, |
| { |
| "epoch": 1.6142433234421365, |
| "grad_norm": 0.32166382447684766, |
| "learning_rate": 4.1586246903368496e-05, |
| "loss": 0.6786, |
| "step": 238 |
| }, |
| { |
| "epoch": 1.6210258584145825, |
| "grad_norm": 0.24677680563722856, |
| "learning_rate": 4.126911733992272e-05, |
| "loss": 0.6791, |
| "step": 239 |
| }, |
| { |
| "epoch": 1.6278083933870284, |
| "grad_norm": 0.18800235990418823, |
| "learning_rate": 4.0951907901844296e-05, |
| "loss": 0.6591, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.6345909283594744, |
| "grad_norm": 0.18086364097944294, |
| "learning_rate": 4.063463855339232e-05, |
| "loss": 0.6607, |
| "step": 241 |
| }, |
| { |
| "epoch": 1.6413734633319204, |
| "grad_norm": 0.1816725478249145, |
| "learning_rate": 4.031732926259639e-05, |
| "loss": 0.6636, |
| "step": 242 |
| }, |
| { |
| "epoch": 1.6481559983043663, |
| "grad_norm": 0.2179460901904272, |
| "learning_rate": 4e-05, |
| "loss": 0.6582, |
| "step": 243 |
| }, |
| { |
| "epoch": 1.654938533276812, |
| "grad_norm": 0.2408536943475412, |
| "learning_rate": 3.9682670737403624e-05, |
| "loss": 0.6581, |
| "step": 244 |
| }, |
| { |
| "epoch": 1.6617210682492582, |
| "grad_norm": 0.19000091863164711, |
| "learning_rate": 3.9365361446607684e-05, |
| "loss": 0.6597, |
| "step": 245 |
| }, |
| { |
| "epoch": 1.668503603221704, |
| "grad_norm": 0.1604164527075533, |
| "learning_rate": 3.904809209815571e-05, |
| "loss": 0.6609, |
| "step": 246 |
| }, |
| { |
| "epoch": 1.6752861381941502, |
| "grad_norm": 0.20780114182132747, |
| "learning_rate": 3.87308826600773e-05, |
| "loss": 0.6588, |
| "step": 247 |
| }, |
| { |
| "epoch": 1.682068673166596, |
| "grad_norm": 0.21684423863552219, |
| "learning_rate": 3.841375309663151e-05, |
| "loss": 0.6688, |
| "step": 248 |
| }, |
| { |
| "epoch": 1.688851208139042, |
| "grad_norm": 0.21044006155176154, |
| "learning_rate": 3.809672336705031e-05, |
| "loss": 0.6711, |
| "step": 249 |
| }, |
| { |
| "epoch": 1.6956337431114878, |
| "grad_norm": 0.1911074700409385, |
| "learning_rate": 3.7779813424282404e-05, |
| "loss": 0.676, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.702416278083934, |
| "grad_norm": 0.20477853050090425, |
| "learning_rate": 3.746304321373742e-05, |
| "loss": 0.6694, |
| "step": 251 |
| }, |
| { |
| "epoch": 1.7091988130563798, |
| "grad_norm": 0.2337501696824405, |
| "learning_rate": 3.714643267203071e-05, |
| "loss": 0.6545, |
| "step": 252 |
| }, |
| { |
| "epoch": 1.7159813480288257, |
| "grad_norm": 0.174865590490607, |
| "learning_rate": 3.683000172572846e-05, |
| "loss": 0.6766, |
| "step": 253 |
| }, |
| { |
| "epoch": 1.7227638830012717, |
| "grad_norm": 0.21609429620617512, |
| "learning_rate": 3.6513770290093674e-05, |
| "loss": 0.6574, |
| "step": 254 |
| }, |
| { |
| "epoch": 1.7295464179737177, |
| "grad_norm": 0.24191345936732944, |
| "learning_rate": 3.6197758267832705e-05, |
| "loss": 0.6635, |
| "step": 255 |
| }, |
| { |
| "epoch": 1.7363289529461636, |
| "grad_norm": 0.21998311775904844, |
| "learning_rate": 3.5881985547842627e-05, |
| "loss": 0.6676, |
| "step": 256 |
| }, |
| { |
| "epoch": 1.7431114879186096, |
| "grad_norm": 0.23724435100092056, |
| "learning_rate": 3.556647200395956e-05, |
| "loss": 0.6624, |
| "step": 257 |
| }, |
| { |
| "epoch": 1.7498940228910556, |
| "grad_norm": 0.240215596344218, |
| "learning_rate": 3.5251237493707804e-05, |
| "loss": 0.6526, |
| "step": 258 |
| }, |
| { |
| "epoch": 1.7566765578635015, |
| "grad_norm": 0.19872673849172526, |
| "learning_rate": 3.493630185705003e-05, |
| "loss": 0.6752, |
| "step": 259 |
| }, |
| { |
| "epoch": 1.7634590928359475, |
| "grad_norm": 0.23055129025202933, |
| "learning_rate": 3.462168491513873e-05, |
| "loss": 0.6537, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.7702416278083932, |
| "grad_norm": 0.18151089866178272, |
| "learning_rate": 3.4307406469068604e-05, |
| "loss": 0.6699, |
| "step": 261 |
| }, |
| { |
| "epoch": 1.7770241627808394, |
| "grad_norm": 0.21466329451747204, |
| "learning_rate": 3.3993486298630384e-05, |
| "loss": 0.6505, |
| "step": 262 |
| }, |
| { |
| "epoch": 1.7838066977532852, |
| "grad_norm": 0.22222597418898904, |
| "learning_rate": 3.367994416106601e-05, |
| "loss": 0.6646, |
| "step": 263 |
| }, |
| { |
| "epoch": 1.7905892327257313, |
| "grad_norm": 0.17502941394521884, |
| "learning_rate": 3.3366799789825044e-05, |
| "loss": 0.6641, |
| "step": 264 |
| }, |
| { |
| "epoch": 1.797371767698177, |
| "grad_norm": 0.24125561495280676, |
| "learning_rate": 3.305407289332279e-05, |
| "loss": 0.6591, |
| "step": 265 |
| }, |
| { |
| "epoch": 1.8041543026706233, |
| "grad_norm": 0.16946358730913183, |
| "learning_rate": 3.27417831536999e-05, |
| "loss": 0.6607, |
| "step": 266 |
| }, |
| { |
| "epoch": 1.810936837643069, |
| "grad_norm": 0.2021667757386069, |
| "learning_rate": 3.2429950225583606e-05, |
| "loss": 0.6578, |
| "step": 267 |
| }, |
| { |
| "epoch": 1.8177193726155152, |
| "grad_norm": 0.21522045832395853, |
| "learning_rate": 3.2118593734850686e-05, |
| "loss": 0.6666, |
| "step": 268 |
| }, |
| { |
| "epoch": 1.824501907587961, |
| "grad_norm": 0.1955912817171962, |
| "learning_rate": 3.180773327739238e-05, |
| "loss": 0.6666, |
| "step": 269 |
| }, |
| { |
| "epoch": 1.831284442560407, |
| "grad_norm": 0.17543010393599448, |
| "learning_rate": 3.1497388417880935e-05, |
| "loss": 0.6689, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.8380669775328529, |
| "grad_norm": 0.16672656731926996, |
| "learning_rate": 3.118757868853838e-05, |
| "loss": 0.6618, |
| "step": 271 |
| }, |
| { |
| "epoch": 1.8448495125052988, |
| "grad_norm": 0.17242532075303685, |
| "learning_rate": 3.087832358790715e-05, |
| "loss": 0.6516, |
| "step": 272 |
| }, |
| { |
| "epoch": 1.8516320474777448, |
| "grad_norm": 0.15372909956958672, |
| "learning_rate": 3.0569642579622905e-05, |
| "loss": 0.6632, |
| "step": 273 |
| }, |
| { |
| "epoch": 1.8584145824501908, |
| "grad_norm": 0.15847716517543572, |
| "learning_rate": 3.0261555091189648e-05, |
| "loss": 0.6652, |
| "step": 274 |
| }, |
| { |
| "epoch": 1.8651971174226367, |
| "grad_norm": 0.16812640824945566, |
| "learning_rate": 2.9954080512756836e-05, |
| "loss": 0.6544, |
| "step": 275 |
| }, |
| { |
| "epoch": 1.8719796523950827, |
| "grad_norm": 0.14101975804977546, |
| "learning_rate": 2.9647238195899168e-05, |
| "loss": 0.6744, |
| "step": 276 |
| }, |
| { |
| "epoch": 1.8787621873675286, |
| "grad_norm": 0.16388464384133353, |
| "learning_rate": 2.9341047452398607e-05, |
| "loss": 0.655, |
| "step": 277 |
| }, |
| { |
| "epoch": 1.8855447223399746, |
| "grad_norm": 0.17548344056861861, |
| "learning_rate": 2.9035527553028906e-05, |
| "loss": 0.6719, |
| "step": 278 |
| }, |
| { |
| "epoch": 1.8923272573124206, |
| "grad_norm": 0.1422728380203633, |
| "learning_rate": 2.873069772634281e-05, |
| "loss": 0.6614, |
| "step": 279 |
| }, |
| { |
| "epoch": 1.8991097922848663, |
| "grad_norm": 0.16674659478989506, |
| "learning_rate": 2.8426577157461897e-05, |
| "loss": 0.6564, |
| "step": 280 |
| }, |
| { |
| "epoch": 1.9058923272573125, |
| "grad_norm": 0.15682390820225428, |
| "learning_rate": 2.8123184986869022e-05, |
| "loss": 0.6683, |
| "step": 281 |
| }, |
| { |
| "epoch": 1.9126748622297582, |
| "grad_norm": 0.1734396291500367, |
| "learning_rate": 2.7820540309203728e-05, |
| "loss": 0.6655, |
| "step": 282 |
| }, |
| { |
| "epoch": 1.9194573972022044, |
| "grad_norm": 0.1467568096698091, |
| "learning_rate": 2.751866217206052e-05, |
| "loss": 0.6605, |
| "step": 283 |
| }, |
| { |
| "epoch": 1.9262399321746502, |
| "grad_norm": 0.17404162787939303, |
| "learning_rate": 2.721756957479001e-05, |
| "loss": 0.6451, |
| "step": 284 |
| }, |
| { |
| "epoch": 1.9330224671470964, |
| "grad_norm": 0.16431154167988282, |
| "learning_rate": 2.691728146730314e-05, |
| "loss": 0.6717, |
| "step": 285 |
| }, |
| { |
| "epoch": 1.939805002119542, |
| "grad_norm": 0.1311637148885735, |
| "learning_rate": 2.6617816748878595e-05, |
| "loss": 0.6779, |
| "step": 286 |
| }, |
| { |
| "epoch": 1.9465875370919883, |
| "grad_norm": 0.15022746394032438, |
| "learning_rate": 2.6319194266973256e-05, |
| "loss": 0.6642, |
| "step": 287 |
| }, |
| { |
| "epoch": 1.953370072064434, |
| "grad_norm": 0.13388633011558382, |
| "learning_rate": 2.6021432816036073e-05, |
| "loss": 0.6673, |
| "step": 288 |
| }, |
| { |
| "epoch": 1.96015260703688, |
| "grad_norm": 0.15469686766213264, |
| "learning_rate": 2.5724551136325132e-05, |
| "loss": 0.6641, |
| "step": 289 |
| }, |
| { |
| "epoch": 1.966935142009326, |
| "grad_norm": 0.13266678943395285, |
| "learning_rate": 2.5428567912728225e-05, |
| "loss": 0.6524, |
| "step": 290 |
| }, |
| { |
| "epoch": 1.973717676981772, |
| "grad_norm": 0.14468475238604925, |
| "learning_rate": 2.5133501773586905e-05, |
| "loss": 0.6671, |
| "step": 291 |
| }, |
| { |
| "epoch": 1.9805002119542179, |
| "grad_norm": 0.14376432351647073, |
| "learning_rate": 2.483937128952399e-05, |
| "loss": 0.6609, |
| "step": 292 |
| }, |
| { |
| "epoch": 1.9872827469266638, |
| "grad_norm": 0.159716122840683, |
| "learning_rate": 2.4546194972274852e-05, |
| "loss": 0.67, |
| "step": 293 |
| }, |
| { |
| "epoch": 1.9940652818991098, |
| "grad_norm": 0.13840709480527438, |
| "learning_rate": 2.425399127352235e-05, |
| "loss": 0.6613, |
| "step": 294 |
| }, |
| { |
| "epoch": 2.0008478168715556, |
| "grad_norm": 0.23463615437325822, |
| "learning_rate": 2.3962778583735463e-05, |
| "loss": 1.0436, |
| "step": 295 |
| }, |
| { |
| "epoch": 2.0076303518440017, |
| "grad_norm": 0.17334268341845893, |
| "learning_rate": 2.3672575231011888e-05, |
| "loss": 0.6518, |
| "step": 296 |
| }, |
| { |
| "epoch": 2.0144128868164475, |
| "grad_norm": 0.16319357693293915, |
| "learning_rate": 2.338339947992455e-05, |
| "loss": 0.6541, |
| "step": 297 |
| }, |
| { |
| "epoch": 2.0211954217888937, |
| "grad_norm": 0.1642296066509337, |
| "learning_rate": 2.3095269530372032e-05, |
| "loss": 0.642, |
| "step": 298 |
| }, |
| { |
| "epoch": 2.0279779567613394, |
| "grad_norm": 0.14714181360085615, |
| "learning_rate": 2.2808203516433136e-05, |
| "loss": 0.6585, |
| "step": 299 |
| }, |
| { |
| "epoch": 2.0347604917337856, |
| "grad_norm": 0.17243791907969416, |
| "learning_rate": 2.2522219505225627e-05, |
| "loss": 0.6367, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.0415430267062313, |
| "grad_norm": 0.15616193111829843, |
| "learning_rate": 2.2237335495769035e-05, |
| "loss": 0.6435, |
| "step": 301 |
| }, |
| { |
| "epoch": 2.0483255616786775, |
| "grad_norm": 0.16853105336890806, |
| "learning_rate": 2.1953569417851983e-05, |
| "loss": 0.6512, |
| "step": 302 |
| }, |
| { |
| "epoch": 2.0551080966511233, |
| "grad_norm": 0.1491939764011879, |
| "learning_rate": 2.1670939130903585e-05, |
| "loss": 0.6421, |
| "step": 303 |
| }, |
| { |
| "epoch": 2.0618906316235694, |
| "grad_norm": 0.15064138598058863, |
| "learning_rate": 2.1389462422869576e-05, |
| "loss": 0.6449, |
| "step": 304 |
| }, |
| { |
| "epoch": 2.068673166596015, |
| "grad_norm": 0.14394176360070499, |
| "learning_rate": 2.11091570090927e-05, |
| "loss": 0.645, |
| "step": 305 |
| }, |
| { |
| "epoch": 2.0754557015684614, |
| "grad_norm": 0.14131247697973692, |
| "learning_rate": 2.0830040531197744e-05, |
| "loss": 0.6486, |
| "step": 306 |
| }, |
| { |
| "epoch": 2.082238236540907, |
| "grad_norm": 0.13739301329908918, |
| "learning_rate": 2.055213055598126e-05, |
| "loss": 0.6294, |
| "step": 307 |
| }, |
| { |
| "epoch": 2.0890207715133533, |
| "grad_norm": 0.14567983415785407, |
| "learning_rate": 2.027544457430599e-05, |
| "loss": 0.6533, |
| "step": 308 |
| }, |
| { |
| "epoch": 2.095803306485799, |
| "grad_norm": 0.12128606904438433, |
| "learning_rate": 2.0000000000000012e-05, |
| "loss": 0.6439, |
| "step": 309 |
| }, |
| { |
| "epoch": 2.1025858414582452, |
| "grad_norm": 0.13694286534266129, |
| "learning_rate": 1.9725814168760688e-05, |
| "loss": 0.6326, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.109368376430691, |
| "grad_norm": 0.11981049040505881, |
| "learning_rate": 1.9452904337063757e-05, |
| "loss": 0.6294, |
| "step": 311 |
| }, |
| { |
| "epoch": 2.1161509114031367, |
| "grad_norm": 0.14362025503542125, |
| "learning_rate": 1.9181287681077116e-05, |
| "loss": 0.6391, |
| "step": 312 |
| }, |
| { |
| "epoch": 2.122933446375583, |
| "grad_norm": 0.13487152713963088, |
| "learning_rate": 1.8910981295579903e-05, |
| "loss": 0.6363, |
| "step": 313 |
| }, |
| { |
| "epoch": 2.1297159813480286, |
| "grad_norm": 0.13488125276633461, |
| "learning_rate": 1.864200219288656e-05, |
| "loss": 0.6499, |
| "step": 314 |
| }, |
| { |
| "epoch": 2.136498516320475, |
| "grad_norm": 0.11819804560452499, |
| "learning_rate": 1.8374367301776112e-05, |
| "loss": 0.6484, |
| "step": 315 |
| }, |
| { |
| "epoch": 2.1432810512929206, |
| "grad_norm": 0.14333365446921437, |
| "learning_rate": 1.8108093466426713e-05, |
| "loss": 0.6517, |
| "step": 316 |
| }, |
| { |
| "epoch": 2.1500635862653668, |
| "grad_norm": 0.11619579623506068, |
| "learning_rate": 1.7843197445355593e-05, |
| "loss": 0.6424, |
| "step": 317 |
| }, |
| { |
| "epoch": 2.1568461212378125, |
| "grad_norm": 0.11214714038576876, |
| "learning_rate": 1.7579695910364235e-05, |
| "loss": 0.6327, |
| "step": 318 |
| }, |
| { |
| "epoch": 2.1636286562102587, |
| "grad_norm": 0.11836436894373582, |
| "learning_rate": 1.7317605445489178e-05, |
| "loss": 0.6332, |
| "step": 319 |
| }, |
| { |
| "epoch": 2.1704111911827044, |
| "grad_norm": 0.12190048890581265, |
| "learning_rate": 1.7056942545958167e-05, |
| "loss": 0.6664, |
| "step": 320 |
| }, |
| { |
| "epoch": 2.1771937261551506, |
| "grad_norm": 0.12475488169306143, |
| "learning_rate": 1.679772361715208e-05, |
| "loss": 0.6411, |
| "step": 321 |
| }, |
| { |
| "epoch": 2.1839762611275964, |
| "grad_norm": 0.13291841960547746, |
| "learning_rate": 1.6539964973572388e-05, |
| "loss": 0.6454, |
| "step": 322 |
| }, |
| { |
| "epoch": 2.1907587961000425, |
| "grad_norm": 0.12175988658418246, |
| "learning_rate": 1.6283682837814388e-05, |
| "loss": 0.646, |
| "step": 323 |
| }, |
| { |
| "epoch": 2.1975413310724883, |
| "grad_norm": 0.11016625987010713, |
| "learning_rate": 1.6028893339546122e-05, |
| "loss": 0.6326, |
| "step": 324 |
| }, |
| { |
| "epoch": 2.2043238660449345, |
| "grad_norm": 0.1108749966998608, |
| "learning_rate": 1.5775612514493343e-05, |
| "loss": 0.6461, |
| "step": 325 |
| }, |
| { |
| "epoch": 2.21110640101738, |
| "grad_norm": 0.1143880178225703, |
| "learning_rate": 1.552385630343014e-05, |
| "loss": 0.6372, |
| "step": 326 |
| }, |
| { |
| "epoch": 2.2178889359898264, |
| "grad_norm": 0.11233236110610757, |
| "learning_rate": 1.527364055117579e-05, |
| "loss": 0.6286, |
| "step": 327 |
| }, |
| { |
| "epoch": 2.224671470962272, |
| "grad_norm": 0.11855577442735764, |
| "learning_rate": 1.502498100559747e-05, |
| "loss": 0.6439, |
| "step": 328 |
| }, |
| { |
| "epoch": 2.231454005934718, |
| "grad_norm": 0.12612061766992566, |
| "learning_rate": 1.4777893316619114e-05, |
| "loss": 0.6369, |
| "step": 329 |
| }, |
| { |
| "epoch": 2.238236540907164, |
| "grad_norm": 0.10643244567354232, |
| "learning_rate": 1.4532393035236477e-05, |
| "loss": 0.6385, |
| "step": 330 |
| }, |
| { |
| "epoch": 2.24501907587961, |
| "grad_norm": 0.1173067781996439, |
| "learning_rate": 1.4288495612538427e-05, |
| "loss": 0.6339, |
| "step": 331 |
| }, |
| { |
| "epoch": 2.251801610852056, |
| "grad_norm": 0.11386078478518308, |
| "learning_rate": 1.404621639873447e-05, |
| "loss": 0.6397, |
| "step": 332 |
| }, |
| { |
| "epoch": 2.2585841458245017, |
| "grad_norm": 0.09826930234307879, |
| "learning_rate": 1.3805570642188602e-05, |
| "loss": 0.645, |
| "step": 333 |
| }, |
| { |
| "epoch": 2.265366680796948, |
| "grad_norm": 0.11305991403115098, |
| "learning_rate": 1.3566573488459729e-05, |
| "loss": 0.6449, |
| "step": 334 |
| }, |
| { |
| "epoch": 2.2721492157693937, |
| "grad_norm": 0.11019252840128578, |
| "learning_rate": 1.3329239979348341e-05, |
| "loss": 0.64, |
| "step": 335 |
| }, |
| { |
| "epoch": 2.27893175074184, |
| "grad_norm": 0.11234819489299995, |
| "learning_rate": 1.30935850519499e-05, |
| "loss": 0.6514, |
| "step": 336 |
| }, |
| { |
| "epoch": 2.2857142857142856, |
| "grad_norm": 0.1112071135535242, |
| "learning_rate": 1.2859623537714719e-05, |
| "loss": 0.6551, |
| "step": 337 |
| }, |
| { |
| "epoch": 2.2924968206867318, |
| "grad_norm": 0.131859623404786, |
| "learning_rate": 1.262737016151447e-05, |
| "loss": 0.6299, |
| "step": 338 |
| }, |
| { |
| "epoch": 2.2992793556591775, |
| "grad_norm": 0.11621464001477737, |
| "learning_rate": 1.2396839540715528e-05, |
| "loss": 0.6316, |
| "step": 339 |
| }, |
| { |
| "epoch": 2.3060618906316237, |
| "grad_norm": 0.10479280516442933, |
| "learning_rate": 1.2168046184258896e-05, |
| "loss": 0.6284, |
| "step": 340 |
| }, |
| { |
| "epoch": 2.3128444256040694, |
| "grad_norm": 0.10786907647373126, |
| "learning_rate": 1.1941004491747145e-05, |
| "loss": 0.6436, |
| "step": 341 |
| }, |
| { |
| "epoch": 2.3196269605765156, |
| "grad_norm": 0.10782220614805556, |
| "learning_rate": 1.1715728752538103e-05, |
| "loss": 0.644, |
| "step": 342 |
| }, |
| { |
| "epoch": 2.3264094955489614, |
| "grad_norm": 0.11055552541032848, |
| "learning_rate": 1.1492233144845497e-05, |
| "loss": 0.654, |
| "step": 343 |
| }, |
| { |
| "epoch": 2.3331920305214076, |
| "grad_norm": 0.11811567288889502, |
| "learning_rate": 1.1270531734846642e-05, |
| "loss": 0.6383, |
| "step": 344 |
| }, |
| { |
| "epoch": 2.3399745654938533, |
| "grad_norm": 0.09614027593231717, |
| "learning_rate": 1.1050638475797193e-05, |
| "loss": 0.6486, |
| "step": 345 |
| }, |
| { |
| "epoch": 2.3467571004662995, |
| "grad_norm": 0.11607457015859846, |
| "learning_rate": 1.083256720715292e-05, |
| "loss": 0.6404, |
| "step": 346 |
| }, |
| { |
| "epoch": 2.3535396354387452, |
| "grad_norm": 0.10091742280729821, |
| "learning_rate": 1.0616331653698673e-05, |
| "loss": 0.6532, |
| "step": 347 |
| }, |
| { |
| "epoch": 2.360322170411191, |
| "grad_norm": 0.09008242164464271, |
| "learning_rate": 1.0401945424684653e-05, |
| "loss": 0.6345, |
| "step": 348 |
| }, |
| { |
| "epoch": 2.367104705383637, |
| "grad_norm": 0.11007729141322087, |
| "learning_rate": 1.0189422012969814e-05, |
| "loss": 0.6482, |
| "step": 349 |
| }, |
| { |
| "epoch": 2.373887240356083, |
| "grad_norm": 0.08713293928424415, |
| "learning_rate": 9.978774794172717e-06, |
| "loss": 0.6243, |
| "step": 350 |
| }, |
| { |
| "epoch": 2.380669775328529, |
| "grad_norm": 0.09736881200508285, |
| "learning_rate": 9.770017025829675e-06, |
| "loss": 0.6464, |
| "step": 351 |
| }, |
| { |
| "epoch": 2.387452310300975, |
| "grad_norm": 0.09620789674440036, |
| "learning_rate": 9.563161846560342e-06, |
| "loss": 0.6399, |
| "step": 352 |
| }, |
| { |
| "epoch": 2.394234845273421, |
| "grad_norm": 0.1013119106642466, |
| "learning_rate": 9.358222275240884e-06, |
| "loss": 0.6369, |
| "step": 353 |
| }, |
| { |
| "epoch": 2.4010173802458668, |
| "grad_norm": 0.0916584684509351, |
| "learning_rate": 9.155211210184495e-06, |
| "loss": 0.6389, |
| "step": 354 |
| }, |
| { |
| "epoch": 2.407799915218313, |
| "grad_norm": 0.10238057837102806, |
| "learning_rate": 8.954141428329723e-06, |
| "loss": 0.6377, |
| "step": 355 |
| }, |
| { |
| "epoch": 2.4145824501907587, |
| "grad_norm": 0.10324400627592539, |
| "learning_rate": 8.755025584436266e-06, |
| "loss": 0.6376, |
| "step": 356 |
| }, |
| { |
| "epoch": 2.421364985163205, |
| "grad_norm": 0.10057048511977558, |
| "learning_rate": 8.55787621028851e-06, |
| "loss": 0.6506, |
| "step": 357 |
| }, |
| { |
| "epoch": 2.4281475201356506, |
| "grad_norm": 0.0920558216849332, |
| "learning_rate": 8.362705713906836e-06, |
| "loss": 0.6313, |
| "step": 358 |
| }, |
| { |
| "epoch": 2.434930055108097, |
| "grad_norm": 0.10610753513988694, |
| "learning_rate": 8.169526378766713e-06, |
| "loss": 0.6429, |
| "step": 359 |
| }, |
| { |
| "epoch": 2.4417125900805425, |
| "grad_norm": 0.10421896294296241, |
| "learning_rate": 7.978350363025588e-06, |
| "loss": 0.6407, |
| "step": 360 |
| }, |
| { |
| "epoch": 2.4484951250529887, |
| "grad_norm": 0.09993739787650231, |
| "learning_rate": 7.789189698757656e-06, |
| "loss": 0.6348, |
| "step": 361 |
| }, |
| { |
| "epoch": 2.4552776600254345, |
| "grad_norm": 0.09901740077583913, |
| "learning_rate": 7.602056291196671e-06, |
| "loss": 0.6459, |
| "step": 362 |
| }, |
| { |
| "epoch": 2.4620601949978806, |
| "grad_norm": 0.10236110912446401, |
| "learning_rate": 7.416961917986572e-06, |
| "loss": 0.6509, |
| "step": 363 |
| }, |
| { |
| "epoch": 2.4688427299703264, |
| "grad_norm": 0.08705071344819373, |
| "learning_rate": 7.233918228440324e-06, |
| "loss": 0.6348, |
| "step": 364 |
| }, |
| { |
| "epoch": 2.4756252649427726, |
| "grad_norm": 0.11881375541999774, |
| "learning_rate": 7.052936742806693e-06, |
| "loss": 0.646, |
| "step": 365 |
| }, |
| { |
| "epoch": 2.4824077999152183, |
| "grad_norm": 0.10091840020224621, |
| "learning_rate": 6.874028851545174e-06, |
| "loss": 0.64, |
| "step": 366 |
| }, |
| { |
| "epoch": 2.489190334887664, |
| "grad_norm": 0.0828156389233352, |
| "learning_rate": 6.697205814609149e-06, |
| "loss": 0.6324, |
| "step": 367 |
| }, |
| { |
| "epoch": 2.4959728698601102, |
| "grad_norm": 0.09511789444296942, |
| "learning_rate": 6.522478760737212e-06, |
| "loss": 0.6398, |
| "step": 368 |
| }, |
| { |
| "epoch": 2.5027554048325564, |
| "grad_norm": 0.10690963856134064, |
| "learning_rate": 6.349858686752748e-06, |
| "loss": 0.6441, |
| "step": 369 |
| }, |
| { |
| "epoch": 2.509537939805002, |
| "grad_norm": 0.10170196889030197, |
| "learning_rate": 6.17935645687183e-06, |
| "loss": 0.6307, |
| "step": 370 |
| }, |
| { |
| "epoch": 2.516320474777448, |
| "grad_norm": 0.0873576460247102, |
| "learning_rate": 6.010982802019429e-06, |
| "loss": 0.6382, |
| "step": 371 |
| }, |
| { |
| "epoch": 2.523103009749894, |
| "grad_norm": 0.10193237211341288, |
| "learning_rate": 5.8447483191540784e-06, |
| "loss": 0.6427, |
| "step": 372 |
| }, |
| { |
| "epoch": 2.52988554472234, |
| "grad_norm": 0.09337413798566872, |
| "learning_rate": 5.680663470600918e-06, |
| "loss": 0.6479, |
| "step": 373 |
| }, |
| { |
| "epoch": 2.536668079694786, |
| "grad_norm": 0.08453984752561841, |
| "learning_rate": 5.518738583393228e-06, |
| "loss": 0.6508, |
| "step": 374 |
| }, |
| { |
| "epoch": 2.5434506146672318, |
| "grad_norm": 0.08158348255933849, |
| "learning_rate": 5.358983848622452e-06, |
| "loss": 0.6372, |
| "step": 375 |
| }, |
| { |
| "epoch": 2.550233149639678, |
| "grad_norm": 0.09556063525511022, |
| "learning_rate": 5.201409320796842e-06, |
| "loss": 0.6397, |
| "step": 376 |
| }, |
| { |
| "epoch": 2.5570156846121237, |
| "grad_norm": 0.09622696560627544, |
| "learning_rate": 5.046024917208603e-06, |
| "loss": 0.6386, |
| "step": 377 |
| }, |
| { |
| "epoch": 2.56379821958457, |
| "grad_norm": 0.09148441726550699, |
| "learning_rate": 4.892840417309775e-06, |
| "loss": 0.6489, |
| "step": 378 |
| }, |
| { |
| "epoch": 2.5705807545570156, |
| "grad_norm": 0.08606094930168158, |
| "learning_rate": 4.7418654620967216e-06, |
| "loss": 0.6527, |
| "step": 379 |
| }, |
| { |
| "epoch": 2.577363289529462, |
| "grad_norm": 0.08364373433134552, |
| "learning_rate": 4.593109553503343e-06, |
| "loss": 0.6428, |
| "step": 380 |
| }, |
| { |
| "epoch": 2.5841458245019076, |
| "grad_norm": 0.0843513293417055, |
| "learning_rate": 4.446582053803066e-06, |
| "loss": 0.6408, |
| "step": 381 |
| }, |
| { |
| "epoch": 2.5909283594743533, |
| "grad_norm": 0.09217062871361449, |
| "learning_rate": 4.302292185019603e-06, |
| "loss": 0.6496, |
| "step": 382 |
| }, |
| { |
| "epoch": 2.5977108944467995, |
| "grad_norm": 0.08394208095826587, |
| "learning_rate": 4.16024902834657e-06, |
| "loss": 0.6407, |
| "step": 383 |
| }, |
| { |
| "epoch": 2.6044934294192457, |
| "grad_norm": 0.07675967715086138, |
| "learning_rate": 4.020461523575873e-06, |
| "loss": 0.6497, |
| "step": 384 |
| }, |
| { |
| "epoch": 2.6112759643916914, |
| "grad_norm": 0.08253304493993292, |
| "learning_rate": 3.882938468535158e-06, |
| "loss": 0.6473, |
| "step": 385 |
| }, |
| { |
| "epoch": 2.618058499364137, |
| "grad_norm": 0.08837276182393047, |
| "learning_rate": 3.747688518534003e-06, |
| "loss": 0.6467, |
| "step": 386 |
| }, |
| { |
| "epoch": 2.6248410343365833, |
| "grad_norm": 0.08553692728277926, |
| "learning_rate": 3.6147201858192627e-06, |
| "loss": 0.6271, |
| "step": 387 |
| }, |
| { |
| "epoch": 2.6316235693090295, |
| "grad_norm": 0.08672709388396345, |
| "learning_rate": 3.4840418390392895e-06, |
| "loss": 0.6417, |
| "step": 388 |
| }, |
| { |
| "epoch": 2.6384061042814753, |
| "grad_norm": 0.08408576121900961, |
| "learning_rate": 3.3556617027172168e-06, |
| "loss": 0.6474, |
| "step": 389 |
| }, |
| { |
| "epoch": 2.645188639253921, |
| "grad_norm": 0.08946340052273392, |
| "learning_rate": 3.2295878567333784e-06, |
| "loss": 0.6475, |
| "step": 390 |
| }, |
| { |
| "epoch": 2.651971174226367, |
| "grad_norm": 0.10659211899169968, |
| "learning_rate": 3.1058282358167503e-06, |
| "loss": 0.6478, |
| "step": 391 |
| }, |
| { |
| "epoch": 2.658753709198813, |
| "grad_norm": 0.08095162713105776, |
| "learning_rate": 2.984390629045555e-06, |
| "loss": 0.6469, |
| "step": 392 |
| }, |
| { |
| "epoch": 2.665536244171259, |
| "grad_norm": 0.08002747021341323, |
| "learning_rate": 2.8652826793570975e-06, |
| "loss": 0.625, |
| "step": 393 |
| }, |
| { |
| "epoch": 2.672318779143705, |
| "grad_norm": 0.08060733679400718, |
| "learning_rate": 2.7485118830666535e-06, |
| "loss": 0.6419, |
| "step": 394 |
| }, |
| { |
| "epoch": 2.679101314116151, |
| "grad_norm": 0.07939209749073067, |
| "learning_rate": 2.634085589395734e-06, |
| "loss": 0.6396, |
| "step": 395 |
| }, |
| { |
| "epoch": 2.685883849088597, |
| "grad_norm": 0.08689462211707911, |
| "learning_rate": 2.5220110000095366e-06, |
| "loss": 0.6453, |
| "step": 396 |
| }, |
| { |
| "epoch": 2.692666384061043, |
| "grad_norm": 0.08614923525446444, |
| "learning_rate": 2.4122951685636674e-06, |
| "loss": 0.6473, |
| "step": 397 |
| }, |
| { |
| "epoch": 2.6994489190334887, |
| "grad_norm": 0.08063088657355912, |
| "learning_rate": 2.3049450002602394e-06, |
| "loss": 0.627, |
| "step": 398 |
| }, |
| { |
| "epoch": 2.706231454005935, |
| "grad_norm": 0.08042306304216008, |
| "learning_rate": 2.199967251413262e-06, |
| "loss": 0.6411, |
| "step": 399 |
| }, |
| { |
| "epoch": 2.7130139889783806, |
| "grad_norm": 0.07728604571452127, |
| "learning_rate": 2.0973685290234024e-06, |
| "loss": 0.6307, |
| "step": 400 |
| }, |
| { |
| "epoch": 2.7197965239508264, |
| "grad_norm": 0.08086443240008037, |
| "learning_rate": 1.997155290362187e-06, |
| "loss": 0.6452, |
| "step": 401 |
| }, |
| { |
| "epoch": 2.7265790589232726, |
| "grad_norm": 0.07595495043741796, |
| "learning_rate": 1.8993338425655805e-06, |
| "loss": 0.6472, |
| "step": 402 |
| }, |
| { |
| "epoch": 2.7333615938957188, |
| "grad_norm": 0.08185565508209801, |
| "learning_rate": 1.8039103422370452e-06, |
| "loss": 0.6341, |
| "step": 403 |
| }, |
| { |
| "epoch": 2.7401441288681645, |
| "grad_norm": 0.0849805109945081, |
| "learning_rate": 1.7108907950600516e-06, |
| "loss": 0.6292, |
| "step": 404 |
| }, |
| { |
| "epoch": 2.7469266638406102, |
| "grad_norm": 0.07973338132333087, |
| "learning_rate": 1.6202810554201099e-06, |
| "loss": 0.6437, |
| "step": 405 |
| }, |
| { |
| "epoch": 2.7537091988130564, |
| "grad_norm": 0.07790884660042009, |
| "learning_rate": 1.532086826036281e-06, |
| "loss": 0.6456, |
| "step": 406 |
| }, |
| { |
| "epoch": 2.760491733785502, |
| "grad_norm": 0.08873860153969253, |
| "learning_rate": 1.4463136576023185e-06, |
| "loss": 0.6502, |
| "step": 407 |
| }, |
| { |
| "epoch": 2.7672742687579484, |
| "grad_norm": 0.08041178968421962, |
| "learning_rate": 1.3629669484372722e-06, |
| "loss": 0.6469, |
| "step": 408 |
| }, |
| { |
| "epoch": 2.774056803730394, |
| "grad_norm": 0.08336326759495126, |
| "learning_rate": 1.2820519441457502e-06, |
| "loss": 0.6431, |
| "step": 409 |
| }, |
| { |
| "epoch": 2.7808393387028403, |
| "grad_norm": 0.07304922054760385, |
| "learning_rate": 1.2035737372877931e-06, |
| "loss": 0.6456, |
| "step": 410 |
| }, |
| { |
| "epoch": 2.787621873675286, |
| "grad_norm": 0.07545648509921901, |
| "learning_rate": 1.127537267058334e-06, |
| "loss": 0.6459, |
| "step": 411 |
| }, |
| { |
| "epoch": 2.794404408647732, |
| "grad_norm": 0.08168214003465647, |
| "learning_rate": 1.0539473189763583e-06, |
| "loss": 0.6429, |
| "step": 412 |
| }, |
| { |
| "epoch": 2.801186943620178, |
| "grad_norm": 0.07428190943785873, |
| "learning_rate": 9.828085245837183e-07, |
| "loss": 0.6341, |
| "step": 413 |
| }, |
| { |
| "epoch": 2.807969478592624, |
| "grad_norm": 0.0753237289047626, |
| "learning_rate": 9.141253611536238e-07, |
| "loss": 0.6494, |
| "step": 414 |
| }, |
| { |
| "epoch": 2.81475201356507, |
| "grad_norm": 0.07502561278928772, |
| "learning_rate": 8.479021514088548e-07, |
| "loss": 0.6424, |
| "step": 415 |
| }, |
| { |
| "epoch": 2.821534548537516, |
| "grad_norm": 0.07379263477638442, |
| "learning_rate": 7.841430632497204e-07, |
| "loss": 0.6426, |
| "step": 416 |
| }, |
| { |
| "epoch": 2.828317083509962, |
| "grad_norm": 0.07233921710316674, |
| "learning_rate": 7.228521094917318e-07, |
| "loss": 0.6438, |
| "step": 417 |
| }, |
| { |
| "epoch": 2.835099618482408, |
| "grad_norm": 0.07416362872430317, |
| "learning_rate": 6.640331476130435e-07, |
| "loss": 0.6405, |
| "step": 418 |
| }, |
| { |
| "epoch": 2.8418821534548537, |
| "grad_norm": 0.07700076601803095, |
| "learning_rate": 6.076898795116792e-07, |
| "loss": 0.6489, |
| "step": 419 |
| }, |
| { |
| "epoch": 2.8486646884272995, |
| "grad_norm": 0.07609683257441903, |
| "learning_rate": 5.538258512725403e-07, |
| "loss": 0.6209, |
| "step": 420 |
| }, |
| { |
| "epoch": 2.8554472233997457, |
| "grad_norm": 0.07737901602583444, |
| "learning_rate": 5.024444529442285e-07, |
| "loss": 0.6447, |
| "step": 421 |
| }, |
| { |
| "epoch": 2.862229758372192, |
| "grad_norm": 0.0759561526572301, |
| "learning_rate": 4.5354891832569245e-07, |
| "loss": 0.6327, |
| "step": 422 |
| }, |
| { |
| "epoch": 2.8690122933446376, |
| "grad_norm": 0.07540799112373622, |
| "learning_rate": 4.0714232476269265e-07, |
| "loss": 0.6357, |
| "step": 423 |
| }, |
| { |
| "epoch": 2.8757948283170833, |
| "grad_norm": 0.07261245642997437, |
| "learning_rate": 3.632275929541207e-07, |
| "loss": 0.6384, |
| "step": 424 |
| }, |
| { |
| "epoch": 2.8825773632895295, |
| "grad_norm": 0.07730392178003218, |
| "learning_rate": 3.218074867681864e-07, |
| "loss": 0.6343, |
| "step": 425 |
| }, |
| { |
| "epoch": 2.8893598982619753, |
| "grad_norm": 0.07481289428504849, |
| "learning_rate": 2.8288461306846817e-07, |
| "loss": 0.6418, |
| "step": 426 |
| }, |
| { |
| "epoch": 2.8961424332344214, |
| "grad_norm": 0.07365288746369081, |
| "learning_rate": 2.464614215498351e-07, |
| "loss": 0.6399, |
| "step": 427 |
| }, |
| { |
| "epoch": 2.902924968206867, |
| "grad_norm": 0.07293600568995603, |
| "learning_rate": 2.1254020458427281e-07, |
| "loss": 0.645, |
| "step": 428 |
| }, |
| { |
| "epoch": 2.9097075031793134, |
| "grad_norm": 0.07299059860739354, |
| "learning_rate": 1.8112309707661647e-07, |
| "loss": 0.6366, |
| "step": 429 |
| }, |
| { |
| "epoch": 2.916490038151759, |
| "grad_norm": 0.07762603812234999, |
| "learning_rate": 1.522120763301782e-07, |
| "loss": 0.6468, |
| "step": 430 |
| }, |
| { |
| "epoch": 2.9232725731242053, |
| "grad_norm": 0.07121635004426441, |
| "learning_rate": 1.2580896192230906e-07, |
| "loss": 0.6457, |
| "step": 431 |
| }, |
| { |
| "epoch": 2.930055108096651, |
| "grad_norm": 0.07189028443702318, |
| "learning_rate": 1.019154155898594e-07, |
| "loss": 0.648, |
| "step": 432 |
| }, |
| { |
| "epoch": 2.9368376430690972, |
| "grad_norm": 0.07237541136245566, |
| "learning_rate": 8.053294112462696e-08, |
| "loss": 0.6406, |
| "step": 433 |
| }, |
| { |
| "epoch": 2.943620178041543, |
| "grad_norm": 0.07059922854737963, |
| "learning_rate": 6.166288427867706e-08, |
| "loss": 0.6349, |
| "step": 434 |
| }, |
| { |
| "epoch": 2.9504027130139887, |
| "grad_norm": 0.07048464732580594, |
| "learning_rate": 4.530643267968149e-08, |
| "loss": 0.6532, |
| "step": 435 |
| }, |
| { |
| "epoch": 2.957185247986435, |
| "grad_norm": 0.0714872818335937, |
| "learning_rate": 3.1464615756133796e-08, |
| "loss": 0.6323, |
| "step": 436 |
| }, |
| { |
| "epoch": 2.963967782958881, |
| "grad_norm": 0.07124837526302868, |
| "learning_rate": 2.01383046725967e-08, |
| "loss": 0.6495, |
| "step": 437 |
| }, |
| { |
| "epoch": 2.970750317931327, |
| "grad_norm": 0.0731063695999088, |
| "learning_rate": 1.1328212274839267e-08, |
| "loss": 0.6516, |
| "step": 438 |
| }, |
| { |
| "epoch": 2.9775328529037726, |
| "grad_norm": 0.07121633404132001, |
| "learning_rate": 5.034893044997269e-09, |
| "loss": 0.6377, |
| "step": 439 |
| }, |
| { |
| "epoch": 2.9843153878762188, |
| "grad_norm": 0.0754923808775944, |
| "learning_rate": 1.2587430666766153e-09, |
| "loss": 0.6443, |
| "step": 440 |
| }, |
| { |
| "epoch": 2.991097922848665, |
| "grad_norm": 0.07154903528443775, |
| "learning_rate": 0.0, |
| "loss": 0.6409, |
| "step": 441 |
| }, |
| { |
| "epoch": 2.991097922848665, |
| "step": 441, |
| "total_flos": 7397230226964480.0, |
| "train_loss": 0.6955865539120438, |
| "train_runtime": 47632.2951, |
| "train_samples_per_second": 4.753, |
| "train_steps_per_second": 0.009 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 441, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 7397230226964480.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
