diff --git "a/v7/checkpoint-1600/trainer_state.json" "b/v7/checkpoint-1600/trainer_state.json" new file mode 100644--- /dev/null +++ "b/v7/checkpoint-1600/trainer_state.json" @@ -0,0 +1,11233 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.9493087557603688, + "eval_steps": 500, + "global_step": 1600, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0018433179723502304, + "grad_norm": 2.02589750289917, + "learning_rate": 1.8404907975460125e-07, + "loss": 1.9595, + "step": 1 + }, + { + "epoch": 0.003686635944700461, + "grad_norm": 1.9972498416900635, + "learning_rate": 3.680981595092025e-07, + "loss": 1.9238, + "step": 2 + }, + { + "epoch": 0.005529953917050691, + "grad_norm": 1.9820266962051392, + "learning_rate": 5.521472392638038e-07, + "loss": 1.9764, + "step": 3 + }, + { + "epoch": 0.007373271889400922, + "grad_norm": 1.9296741485595703, + "learning_rate": 7.36196319018405e-07, + "loss": 1.9161, + "step": 4 + }, + { + "epoch": 0.009216589861751152, + "grad_norm": 2.019129514694214, + "learning_rate": 9.202453987730061e-07, + "loss": 1.9629, + "step": 5 + }, + { + "epoch": 0.011059907834101382, + "grad_norm": 1.9039007425308228, + "learning_rate": 1.1042944785276075e-06, + "loss": 1.9158, + "step": 6 + }, + { + "epoch": 0.012903225806451613, + "grad_norm": 1.9537602663040161, + "learning_rate": 1.2883435582822088e-06, + "loss": 1.9572, + "step": 7 + }, + { + "epoch": 0.014746543778801843, + "grad_norm": 1.9474399089813232, + "learning_rate": 1.47239263803681e-06, + "loss": 1.9453, + "step": 8 + }, + { + "epoch": 0.016589861751152075, + "grad_norm": 1.8985520601272583, + "learning_rate": 1.656441717791411e-06, + "loss": 1.9377, + "step": 9 + }, + { + "epoch": 0.018433179723502304, + "grad_norm": 1.9258599281311035, + "learning_rate": 1.8404907975460122e-06, + "loss": 1.9312, + "step": 10 + }, + { + "epoch": 0.020276497695852536, + "grad_norm": 1.8695780038833618, + "learning_rate": 2.0245398773006137e-06, + "loss": 1.9273, + "step": 11 + }, + { + "epoch": 0.022119815668202765, + "grad_norm": 1.8287005424499512, + "learning_rate": 2.208588957055215e-06, + "loss": 1.8951, + "step": 12 + }, + { + "epoch": 0.023963133640552997, + "grad_norm": 1.7955673933029175, + "learning_rate": 2.392638036809816e-06, + "loss": 1.9054, + "step": 13 + }, + { + "epoch": 0.025806451612903226, + "grad_norm": 1.8065942525863647, + "learning_rate": 2.5766871165644175e-06, + "loss": 1.9343, + "step": 14 + }, + { + "epoch": 0.027649769585253458, + "grad_norm": 1.7342883348464966, + "learning_rate": 2.7607361963190186e-06, + "loss": 1.8964, + "step": 15 + }, + { + "epoch": 0.029493087557603687, + "grad_norm": 1.7930785417556763, + "learning_rate": 2.94478527607362e-06, + "loss": 1.9385, + "step": 16 + }, + { + "epoch": 0.03133640552995392, + "grad_norm": 1.6384600400924683, + "learning_rate": 3.1288343558282206e-06, + "loss": 1.9145, + "step": 17 + }, + { + "epoch": 0.03317972350230415, + "grad_norm": 1.5618699789047241, + "learning_rate": 3.312883435582822e-06, + "loss": 1.9392, + "step": 18 + }, + { + "epoch": 0.035023041474654376, + "grad_norm": 1.5488923788070679, + "learning_rate": 3.496932515337423e-06, + "loss": 1.8919, + "step": 19 + }, + { + "epoch": 0.03686635944700461, + "grad_norm": 1.3122743368148804, + "learning_rate": 3.6809815950920245e-06, + "loss": 1.8735, + "step": 20 + }, + { + "epoch": 0.03870967741935484, + "grad_norm": 1.2238839864730835, + "learning_rate": 3.865030674846626e-06, + "loss": 1.9463, + "step": 21 + }, + { + "epoch": 0.04055299539170507, + "grad_norm": 1.0287063121795654, + "learning_rate": 4.049079754601227e-06, + "loss": 1.8726, + "step": 22 + }, + { + "epoch": 0.0423963133640553, + "grad_norm": 0.8679629564285278, + "learning_rate": 4.233128834355828e-06, + "loss": 1.8418, + "step": 23 + }, + { + "epoch": 0.04423963133640553, + "grad_norm": 0.7079375386238098, + "learning_rate": 4.41717791411043e-06, + "loss": 1.8251, + "step": 24 + }, + { + "epoch": 0.04608294930875576, + "grad_norm": 0.9220630526542664, + "learning_rate": 4.601226993865031e-06, + "loss": 1.7802, + "step": 25 + }, + { + "epoch": 0.047926267281105994, + "grad_norm": 0.8715929985046387, + "learning_rate": 4.785276073619632e-06, + "loss": 1.8777, + "step": 26 + }, + { + "epoch": 0.04976958525345622, + "grad_norm": 0.9550811052322388, + "learning_rate": 4.969325153374233e-06, + "loss": 1.8371, + "step": 27 + }, + { + "epoch": 0.05161290322580645, + "grad_norm": 0.90284264087677, + "learning_rate": 5.153374233128835e-06, + "loss": 1.9045, + "step": 28 + }, + { + "epoch": 0.053456221198156684, + "grad_norm": 0.8552197217941284, + "learning_rate": 5.337423312883436e-06, + "loss": 1.8053, + "step": 29 + }, + { + "epoch": 0.055299539170506916, + "grad_norm": 0.8264591693878174, + "learning_rate": 5.521472392638037e-06, + "loss": 1.8451, + "step": 30 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.8775638341903687, + "learning_rate": 5.7055214723926385e-06, + "loss": 1.8552, + "step": 31 + }, + { + "epoch": 0.05898617511520737, + "grad_norm": 0.807953417301178, + "learning_rate": 5.88957055214724e-06, + "loss": 1.8164, + "step": 32 + }, + { + "epoch": 0.060829493087557605, + "grad_norm": 0.7050386071205139, + "learning_rate": 6.0736196319018406e-06, + "loss": 1.7819, + "step": 33 + }, + { + "epoch": 0.06267281105990784, + "grad_norm": 0.6614200472831726, + "learning_rate": 6.257668711656441e-06, + "loss": 1.8147, + "step": 34 + }, + { + "epoch": 0.06451612903225806, + "grad_norm": 0.5420933365821838, + "learning_rate": 6.4417177914110434e-06, + "loss": 1.7534, + "step": 35 + }, + { + "epoch": 0.0663594470046083, + "grad_norm": 0.42041122913360596, + "learning_rate": 6.625766871165644e-06, + "loss": 1.7973, + "step": 36 + }, + { + "epoch": 0.06820276497695853, + "grad_norm": 0.3737146854400635, + "learning_rate": 6.8098159509202454e-06, + "loss": 1.8, + "step": 37 + }, + { + "epoch": 0.07004608294930875, + "grad_norm": 0.37116268277168274, + "learning_rate": 6.993865030674846e-06, + "loss": 1.713, + "step": 38 + }, + { + "epoch": 0.07188940092165899, + "grad_norm": 0.37946197390556335, + "learning_rate": 7.177914110429448e-06, + "loss": 1.7656, + "step": 39 + }, + { + "epoch": 0.07373271889400922, + "grad_norm": 0.39086607098579407, + "learning_rate": 7.361963190184049e-06, + "loss": 1.7899, + "step": 40 + }, + { + "epoch": 0.07557603686635944, + "grad_norm": 0.4194660186767578, + "learning_rate": 7.54601226993865e-06, + "loss": 1.7247, + "step": 41 + }, + { + "epoch": 0.07741935483870968, + "grad_norm": 0.41651099920272827, + "learning_rate": 7.730061349693252e-06, + "loss": 1.7437, + "step": 42 + }, + { + "epoch": 0.0792626728110599, + "grad_norm": 0.4337350130081177, + "learning_rate": 7.914110429447852e-06, + "loss": 1.7758, + "step": 43 + }, + { + "epoch": 0.08110599078341015, + "grad_norm": 0.39606988430023193, + "learning_rate": 8.098159509202455e-06, + "loss": 1.7264, + "step": 44 + }, + { + "epoch": 0.08294930875576037, + "grad_norm": 0.35638228058815, + "learning_rate": 8.282208588957055e-06, + "loss": 1.6925, + "step": 45 + }, + { + "epoch": 0.0847926267281106, + "grad_norm": 0.3481025695800781, + "learning_rate": 8.466257668711656e-06, + "loss": 1.7767, + "step": 46 + }, + { + "epoch": 0.08663594470046083, + "grad_norm": 0.3551636040210724, + "learning_rate": 8.650306748466258e-06, + "loss": 1.8833, + "step": 47 + }, + { + "epoch": 0.08847926267281106, + "grad_norm": 0.3395654261112213, + "learning_rate": 8.83435582822086e-06, + "loss": 1.7341, + "step": 48 + }, + { + "epoch": 0.09032258064516129, + "grad_norm": 0.3334786891937256, + "learning_rate": 9.01840490797546e-06, + "loss": 1.7423, + "step": 49 + }, + { + "epoch": 0.09216589861751152, + "grad_norm": 0.3836233615875244, + "learning_rate": 9.202453987730062e-06, + "loss": 1.7198, + "step": 50 + }, + { + "epoch": 0.09400921658986175, + "grad_norm": 0.3345714807510376, + "learning_rate": 9.386503067484664e-06, + "loss": 1.7294, + "step": 51 + }, + { + "epoch": 0.09585253456221199, + "grad_norm": 0.3579282760620117, + "learning_rate": 9.570552147239264e-06, + "loss": 1.7277, + "step": 52 + }, + { + "epoch": 0.09769585253456221, + "grad_norm": 0.3569463789463043, + "learning_rate": 9.754601226993865e-06, + "loss": 1.7979, + "step": 53 + }, + { + "epoch": 0.09953917050691244, + "grad_norm": 0.3132180869579315, + "learning_rate": 9.938650306748466e-06, + "loss": 1.7392, + "step": 54 + }, + { + "epoch": 0.10138248847926268, + "grad_norm": 0.3075284957885742, + "learning_rate": 1.0122699386503068e-05, + "loss": 1.6929, + "step": 55 + }, + { + "epoch": 0.1032258064516129, + "grad_norm": 0.3144418001174927, + "learning_rate": 1.030674846625767e-05, + "loss": 1.7455, + "step": 56 + }, + { + "epoch": 0.10506912442396313, + "grad_norm": 0.30215781927108765, + "learning_rate": 1.0490797546012269e-05, + "loss": 1.7015, + "step": 57 + }, + { + "epoch": 0.10691244239631337, + "grad_norm": 0.3008810579776764, + "learning_rate": 1.0674846625766871e-05, + "loss": 1.702, + "step": 58 + }, + { + "epoch": 0.10875576036866359, + "grad_norm": 0.2983757555484772, + "learning_rate": 1.0858895705521474e-05, + "loss": 1.7593, + "step": 59 + }, + { + "epoch": 0.11059907834101383, + "grad_norm": 0.3113389015197754, + "learning_rate": 1.1042944785276074e-05, + "loss": 1.7446, + "step": 60 + }, + { + "epoch": 0.11244239631336406, + "grad_norm": 0.2977043390274048, + "learning_rate": 1.1226993865030675e-05, + "loss": 1.7666, + "step": 61 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.3056432902812958, + "learning_rate": 1.1411042944785277e-05, + "loss": 1.7226, + "step": 62 + }, + { + "epoch": 0.11612903225806452, + "grad_norm": 0.2962339520454407, + "learning_rate": 1.1595092024539878e-05, + "loss": 1.7503, + "step": 63 + }, + { + "epoch": 0.11797235023041475, + "grad_norm": 0.30185747146606445, + "learning_rate": 1.177914110429448e-05, + "loss": 1.7478, + "step": 64 + }, + { + "epoch": 0.11981566820276497, + "grad_norm": 0.29040098190307617, + "learning_rate": 1.1963190184049079e-05, + "loss": 1.779, + "step": 65 + }, + { + "epoch": 0.12165898617511521, + "grad_norm": 0.2708752155303955, + "learning_rate": 1.2147239263803681e-05, + "loss": 1.6816, + "step": 66 + }, + { + "epoch": 0.12350230414746544, + "grad_norm": 0.2978411018848419, + "learning_rate": 1.2331288343558283e-05, + "loss": 1.7439, + "step": 67 + }, + { + "epoch": 0.12534562211981568, + "grad_norm": 0.28008463978767395, + "learning_rate": 1.2515337423312882e-05, + "loss": 1.6918, + "step": 68 + }, + { + "epoch": 0.1271889400921659, + "grad_norm": 0.29284995794296265, + "learning_rate": 1.2699386503067485e-05, + "loss": 1.7043, + "step": 69 + }, + { + "epoch": 0.12903225806451613, + "grad_norm": 0.3163938522338867, + "learning_rate": 1.2883435582822087e-05, + "loss": 1.7501, + "step": 70 + }, + { + "epoch": 0.13087557603686636, + "grad_norm": 0.28190693259239197, + "learning_rate": 1.3067484662576687e-05, + "loss": 1.7396, + "step": 71 + }, + { + "epoch": 0.1327188940092166, + "grad_norm": 0.27740147709846497, + "learning_rate": 1.3251533742331288e-05, + "loss": 1.7276, + "step": 72 + }, + { + "epoch": 0.13456221198156681, + "grad_norm": 0.29363468289375305, + "learning_rate": 1.343558282208589e-05, + "loss": 1.7103, + "step": 73 + }, + { + "epoch": 0.13640552995391705, + "grad_norm": 0.28017935156822205, + "learning_rate": 1.3619631901840491e-05, + "loss": 1.7418, + "step": 74 + }, + { + "epoch": 0.1382488479262673, + "grad_norm": 0.2688934803009033, + "learning_rate": 1.3803680981595093e-05, + "loss": 1.7143, + "step": 75 + }, + { + "epoch": 0.1400921658986175, + "grad_norm": 0.28602007031440735, + "learning_rate": 1.3987730061349692e-05, + "loss": 1.7034, + "step": 76 + }, + { + "epoch": 0.14193548387096774, + "grad_norm": 0.2809165120124817, + "learning_rate": 1.4171779141104294e-05, + "loss": 1.726, + "step": 77 + }, + { + "epoch": 0.14377880184331798, + "grad_norm": 0.2808230519294739, + "learning_rate": 1.4355828220858897e-05, + "loss": 1.7184, + "step": 78 + }, + { + "epoch": 0.1456221198156682, + "grad_norm": 0.2852628827095032, + "learning_rate": 1.4539877300613497e-05, + "loss": 1.7506, + "step": 79 + }, + { + "epoch": 0.14746543778801843, + "grad_norm": 0.28406837582588196, + "learning_rate": 1.4723926380368098e-05, + "loss": 1.667, + "step": 80 + }, + { + "epoch": 0.14930875576036867, + "grad_norm": 0.2859112322330475, + "learning_rate": 1.49079754601227e-05, + "loss": 1.7629, + "step": 81 + }, + { + "epoch": 0.15115207373271888, + "grad_norm": 0.2811156213283539, + "learning_rate": 1.50920245398773e-05, + "loss": 1.7055, + "step": 82 + }, + { + "epoch": 0.15299539170506912, + "grad_norm": 0.27561160922050476, + "learning_rate": 1.52760736196319e-05, + "loss": 1.7149, + "step": 83 + }, + { + "epoch": 0.15483870967741936, + "grad_norm": 0.28458908200263977, + "learning_rate": 1.5460122699386504e-05, + "loss": 1.702, + "step": 84 + }, + { + "epoch": 0.15668202764976957, + "grad_norm": 0.2881270945072174, + "learning_rate": 1.5644171779141104e-05, + "loss": 1.7771, + "step": 85 + }, + { + "epoch": 0.1585253456221198, + "grad_norm": 0.2840360701084137, + "learning_rate": 1.5828220858895705e-05, + "loss": 1.704, + "step": 86 + }, + { + "epoch": 0.16036866359447005, + "grad_norm": 0.28690823912620544, + "learning_rate": 1.601226993865031e-05, + "loss": 1.7032, + "step": 87 + }, + { + "epoch": 0.1622119815668203, + "grad_norm": 0.28307586908340454, + "learning_rate": 1.619631901840491e-05, + "loss": 1.7504, + "step": 88 + }, + { + "epoch": 0.1640552995391705, + "grad_norm": 0.2825368046760559, + "learning_rate": 1.638036809815951e-05, + "loss": 1.6755, + "step": 89 + }, + { + "epoch": 0.16589861751152074, + "grad_norm": 0.2756441533565521, + "learning_rate": 1.656441717791411e-05, + "loss": 1.6542, + "step": 90 + }, + { + "epoch": 0.16774193548387098, + "grad_norm": 0.2785171866416931, + "learning_rate": 1.674846625766871e-05, + "loss": 1.7041, + "step": 91 + }, + { + "epoch": 0.1695852534562212, + "grad_norm": 0.2845054864883423, + "learning_rate": 1.693251533742331e-05, + "loss": 1.664, + "step": 92 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.2794802486896515, + "learning_rate": 1.7116564417177916e-05, + "loss": 1.674, + "step": 93 + }, + { + "epoch": 0.17327188940092167, + "grad_norm": 0.30158528685569763, + "learning_rate": 1.7300613496932516e-05, + "loss": 1.6702, + "step": 94 + }, + { + "epoch": 0.17511520737327188, + "grad_norm": 0.3102332353591919, + "learning_rate": 1.7484662576687117e-05, + "loss": 1.6675, + "step": 95 + }, + { + "epoch": 0.17695852534562212, + "grad_norm": 0.30239519476890564, + "learning_rate": 1.766871165644172e-05, + "loss": 1.735, + "step": 96 + }, + { + "epoch": 0.17880184331797236, + "grad_norm": 0.3213733732700348, + "learning_rate": 1.785276073619632e-05, + "loss": 1.7476, + "step": 97 + }, + { + "epoch": 0.18064516129032257, + "grad_norm": 0.30181968212127686, + "learning_rate": 1.803680981595092e-05, + "loss": 1.7693, + "step": 98 + }, + { + "epoch": 0.1824884792626728, + "grad_norm": 0.29706043004989624, + "learning_rate": 1.8220858895705523e-05, + "loss": 1.6862, + "step": 99 + }, + { + "epoch": 0.18433179723502305, + "grad_norm": 0.31851696968078613, + "learning_rate": 1.8404907975460123e-05, + "loss": 1.713, + "step": 100 + }, + { + "epoch": 0.18617511520737326, + "grad_norm": 0.2876022756099701, + "learning_rate": 1.8588957055214724e-05, + "loss": 1.6564, + "step": 101 + }, + { + "epoch": 0.1880184331797235, + "grad_norm": 0.30441394448280334, + "learning_rate": 1.8773006134969328e-05, + "loss": 1.7286, + "step": 102 + }, + { + "epoch": 0.18986175115207374, + "grad_norm": 0.2926851809024811, + "learning_rate": 1.8957055214723928e-05, + "loss": 1.6485, + "step": 103 + }, + { + "epoch": 0.19170506912442398, + "grad_norm": 0.2925703525543213, + "learning_rate": 1.914110429447853e-05, + "loss": 1.7372, + "step": 104 + }, + { + "epoch": 0.1935483870967742, + "grad_norm": 0.29036155343055725, + "learning_rate": 1.9325153374233126e-05, + "loss": 1.6868, + "step": 105 + }, + { + "epoch": 0.19539170506912443, + "grad_norm": 0.3051284849643707, + "learning_rate": 1.950920245398773e-05, + "loss": 1.7132, + "step": 106 + }, + { + "epoch": 0.19723502304147467, + "grad_norm": 0.295520156621933, + "learning_rate": 1.969325153374233e-05, + "loss": 1.6949, + "step": 107 + }, + { + "epoch": 0.19907834101382488, + "grad_norm": 0.3010528087615967, + "learning_rate": 1.987730061349693e-05, + "loss": 1.6948, + "step": 108 + }, + { + "epoch": 0.20092165898617512, + "grad_norm": 0.3026741147041321, + "learning_rate": 2.0061349693251535e-05, + "loss": 1.7014, + "step": 109 + }, + { + "epoch": 0.20276497695852536, + "grad_norm": 0.32000070810317993, + "learning_rate": 2.0245398773006136e-05, + "loss": 1.709, + "step": 110 + }, + { + "epoch": 0.20460829493087557, + "grad_norm": 0.29906362295150757, + "learning_rate": 2.0429447852760736e-05, + "loss": 1.6834, + "step": 111 + }, + { + "epoch": 0.2064516129032258, + "grad_norm": 0.31326961517333984, + "learning_rate": 2.061349693251534e-05, + "loss": 1.5777, + "step": 112 + }, + { + "epoch": 0.20829493087557605, + "grad_norm": 0.3156152069568634, + "learning_rate": 2.0797546012269938e-05, + "loss": 1.6934, + "step": 113 + }, + { + "epoch": 0.21013824884792626, + "grad_norm": 0.31258904933929443, + "learning_rate": 2.0981595092024538e-05, + "loss": 1.6717, + "step": 114 + }, + { + "epoch": 0.2119815668202765, + "grad_norm": 0.30550417304039, + "learning_rate": 2.1165644171779142e-05, + "loss": 1.6857, + "step": 115 + }, + { + "epoch": 0.21382488479262673, + "grad_norm": 0.3166547417640686, + "learning_rate": 2.1349693251533743e-05, + "loss": 1.6959, + "step": 116 + }, + { + "epoch": 0.21566820276497695, + "grad_norm": 0.30488964915275574, + "learning_rate": 2.1533742331288343e-05, + "loss": 1.6764, + "step": 117 + }, + { + "epoch": 0.21751152073732719, + "grad_norm": 0.30913662910461426, + "learning_rate": 2.1717791411042947e-05, + "loss": 1.6252, + "step": 118 + }, + { + "epoch": 0.21935483870967742, + "grad_norm": 0.33525151014328003, + "learning_rate": 2.1901840490797548e-05, + "loss": 1.6692, + "step": 119 + }, + { + "epoch": 0.22119815668202766, + "grad_norm": 0.329086035490036, + "learning_rate": 2.208588957055215e-05, + "loss": 1.6462, + "step": 120 + }, + { + "epoch": 0.22304147465437787, + "grad_norm": 0.3236452341079712, + "learning_rate": 2.226993865030675e-05, + "loss": 1.7658, + "step": 121 + }, + { + "epoch": 0.2248847926267281, + "grad_norm": 0.3473586142063141, + "learning_rate": 2.245398773006135e-05, + "loss": 1.6979, + "step": 122 + }, + { + "epoch": 0.22672811059907835, + "grad_norm": 0.2963578402996063, + "learning_rate": 2.263803680981595e-05, + "loss": 1.692, + "step": 123 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.3080257475376129, + "learning_rate": 2.2822085889570554e-05, + "loss": 1.6947, + "step": 124 + }, + { + "epoch": 0.2304147465437788, + "grad_norm": 0.38422197103500366, + "learning_rate": 2.3006134969325155e-05, + "loss": 1.7041, + "step": 125 + }, + { + "epoch": 0.23225806451612904, + "grad_norm": 0.30777859687805176, + "learning_rate": 2.3190184049079755e-05, + "loss": 1.7073, + "step": 126 + }, + { + "epoch": 0.23410138248847925, + "grad_norm": 0.43660104274749756, + "learning_rate": 2.337423312883436e-05, + "loss": 1.6824, + "step": 127 + }, + { + "epoch": 0.2359447004608295, + "grad_norm": 0.30751895904541016, + "learning_rate": 2.355828220858896e-05, + "loss": 1.7052, + "step": 128 + }, + { + "epoch": 0.23778801843317973, + "grad_norm": 0.3926418125629425, + "learning_rate": 2.3742331288343557e-05, + "loss": 1.6765, + "step": 129 + }, + { + "epoch": 0.23963133640552994, + "grad_norm": 0.32357603311538696, + "learning_rate": 2.3926380368098158e-05, + "loss": 1.6804, + "step": 130 + }, + { + "epoch": 0.24147465437788018, + "grad_norm": 0.3306221663951874, + "learning_rate": 2.411042944785276e-05, + "loss": 1.709, + "step": 131 + }, + { + "epoch": 0.24331797235023042, + "grad_norm": 0.38641783595085144, + "learning_rate": 2.4294478527607362e-05, + "loss": 1.6391, + "step": 132 + }, + { + "epoch": 0.24516129032258063, + "grad_norm": 0.3362821638584137, + "learning_rate": 2.4478527607361963e-05, + "loss": 1.6555, + "step": 133 + }, + { + "epoch": 0.24700460829493087, + "grad_norm": 0.37096643447875977, + "learning_rate": 2.4662576687116567e-05, + "loss": 1.7266, + "step": 134 + }, + { + "epoch": 0.2488479262672811, + "grad_norm": 0.3151519000530243, + "learning_rate": 2.4846625766871167e-05, + "loss": 1.7161, + "step": 135 + }, + { + "epoch": 0.25069124423963135, + "grad_norm": 0.32760757207870483, + "learning_rate": 2.5030674846625765e-05, + "loss": 1.641, + "step": 136 + }, + { + "epoch": 0.25253456221198156, + "grad_norm": 0.37584084272384644, + "learning_rate": 2.521472392638037e-05, + "loss": 1.7422, + "step": 137 + }, + { + "epoch": 0.2543778801843318, + "grad_norm": 0.31268641352653503, + "learning_rate": 2.539877300613497e-05, + "loss": 1.7261, + "step": 138 + }, + { + "epoch": 0.25622119815668204, + "grad_norm": 0.36679184436798096, + "learning_rate": 2.558282208588957e-05, + "loss": 1.7018, + "step": 139 + }, + { + "epoch": 0.25806451612903225, + "grad_norm": 0.31372857093811035, + "learning_rate": 2.5766871165644174e-05, + "loss": 1.6838, + "step": 140 + }, + { + "epoch": 0.25990783410138246, + "grad_norm": 0.35413920879364014, + "learning_rate": 2.5950920245398774e-05, + "loss": 1.7003, + "step": 141 + }, + { + "epoch": 0.26175115207373273, + "grad_norm": 0.3250369727611542, + "learning_rate": 2.6134969325153375e-05, + "loss": 1.6799, + "step": 142 + }, + { + "epoch": 0.26359447004608294, + "grad_norm": 0.33544111251831055, + "learning_rate": 2.631901840490798e-05, + "loss": 1.6357, + "step": 143 + }, + { + "epoch": 0.2654377880184332, + "grad_norm": 0.31145980954170227, + "learning_rate": 2.6503067484662576e-05, + "loss": 1.6062, + "step": 144 + }, + { + "epoch": 0.2672811059907834, + "grad_norm": 0.31925103068351746, + "learning_rate": 2.6687116564417177e-05, + "loss": 1.7424, + "step": 145 + }, + { + "epoch": 0.26912442396313363, + "grad_norm": 0.3047640919685364, + "learning_rate": 2.687116564417178e-05, + "loss": 1.7195, + "step": 146 + }, + { + "epoch": 0.2709677419354839, + "grad_norm": 0.29877611994743347, + "learning_rate": 2.705521472392638e-05, + "loss": 1.6467, + "step": 147 + }, + { + "epoch": 0.2728110599078341, + "grad_norm": 0.31033533811569214, + "learning_rate": 2.7239263803680982e-05, + "loss": 1.6234, + "step": 148 + }, + { + "epoch": 0.2746543778801843, + "grad_norm": 0.3081960380077362, + "learning_rate": 2.7423312883435586e-05, + "loss": 1.6945, + "step": 149 + }, + { + "epoch": 0.2764976958525346, + "grad_norm": 0.30935296416282654, + "learning_rate": 2.7607361963190186e-05, + "loss": 1.6172, + "step": 150 + }, + { + "epoch": 0.2783410138248848, + "grad_norm": 0.3287109136581421, + "learning_rate": 2.7791411042944787e-05, + "loss": 1.6771, + "step": 151 + }, + { + "epoch": 0.280184331797235, + "grad_norm": 0.3306177258491516, + "learning_rate": 2.7975460122699384e-05, + "loss": 1.6522, + "step": 152 + }, + { + "epoch": 0.2820276497695853, + "grad_norm": 0.30568936467170715, + "learning_rate": 2.8159509202453988e-05, + "loss": 1.6456, + "step": 153 + }, + { + "epoch": 0.2838709677419355, + "grad_norm": 0.31628063321113586, + "learning_rate": 2.834355828220859e-05, + "loss": 1.6521, + "step": 154 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.3133561909198761, + "learning_rate": 2.852760736196319e-05, + "loss": 1.6517, + "step": 155 + }, + { + "epoch": 0.28755760368663597, + "grad_norm": 0.327334463596344, + "learning_rate": 2.8711656441717793e-05, + "loss": 1.6422, + "step": 156 + }, + { + "epoch": 0.2894009216589862, + "grad_norm": 0.3071538209915161, + "learning_rate": 2.8895705521472394e-05, + "loss": 1.6693, + "step": 157 + }, + { + "epoch": 0.2912442396313364, + "grad_norm": 0.3373123109340668, + "learning_rate": 2.9079754601226994e-05, + "loss": 1.7079, + "step": 158 + }, + { + "epoch": 0.29308755760368665, + "grad_norm": 0.32550349831581116, + "learning_rate": 2.92638036809816e-05, + "loss": 1.7135, + "step": 159 + }, + { + "epoch": 0.29493087557603687, + "grad_norm": 0.2980889678001404, + "learning_rate": 2.9447852760736196e-05, + "loss": 1.6375, + "step": 160 + }, + { + "epoch": 0.2967741935483871, + "grad_norm": 0.3041258156299591, + "learning_rate": 2.9631901840490796e-05, + "loss": 1.6542, + "step": 161 + }, + { + "epoch": 0.29861751152073734, + "grad_norm": 0.32162049412727356, + "learning_rate": 2.98159509202454e-05, + "loss": 1.6308, + "step": 162 + }, + { + "epoch": 0.30046082949308756, + "grad_norm": 0.31242045760154724, + "learning_rate": 3e-05, + "loss": 1.7211, + "step": 163 + }, + { + "epoch": 0.30230414746543777, + "grad_norm": 0.32431089878082275, + "learning_rate": 2.9999965416241516e-05, + "loss": 1.6273, + "step": 164 + }, + { + "epoch": 0.30414746543778803, + "grad_norm": 0.34065431356430054, + "learning_rate": 2.999986166512553e-05, + "loss": 1.7136, + "step": 165 + }, + { + "epoch": 0.30599078341013825, + "grad_norm": 0.3061734437942505, + "learning_rate": 2.9999688747130467e-05, + "loss": 1.6912, + "step": 166 + }, + { + "epoch": 0.30783410138248846, + "grad_norm": 0.31865042448043823, + "learning_rate": 2.999944666305367e-05, + "loss": 1.6703, + "step": 167 + }, + { + "epoch": 0.3096774193548387, + "grad_norm": 0.32718777656555176, + "learning_rate": 2.999913541401143e-05, + "loss": 1.5595, + "step": 168 + }, + { + "epoch": 0.31152073732718893, + "grad_norm": 0.31636691093444824, + "learning_rate": 2.9998755001438975e-05, + "loss": 1.6433, + "step": 169 + }, + { + "epoch": 0.31336405529953915, + "grad_norm": 0.35521432757377625, + "learning_rate": 2.999830542709045e-05, + "loss": 1.6257, + "step": 170 + }, + { + "epoch": 0.3152073732718894, + "grad_norm": 0.34638574719429016, + "learning_rate": 2.9997786693038913e-05, + "loss": 1.6341, + "step": 171 + }, + { + "epoch": 0.3170506912442396, + "grad_norm": 0.3070574700832367, + "learning_rate": 2.9997198801676335e-05, + "loss": 1.646, + "step": 172 + }, + { + "epoch": 0.31889400921658984, + "grad_norm": 0.3159651458263397, + "learning_rate": 2.9996541755713585e-05, + "loss": 1.6753, + "step": 173 + }, + { + "epoch": 0.3207373271889401, + "grad_norm": 0.32200679183006287, + "learning_rate": 2.999581555818041e-05, + "loss": 1.6883, + "step": 174 + }, + { + "epoch": 0.3225806451612903, + "grad_norm": 0.30645352602005005, + "learning_rate": 2.9995020212425432e-05, + "loss": 1.656, + "step": 175 + }, + { + "epoch": 0.3244239631336406, + "grad_norm": 0.34474891424179077, + "learning_rate": 2.9994155722116118e-05, + "loss": 1.725, + "step": 176 + }, + { + "epoch": 0.3262672811059908, + "grad_norm": 0.37686192989349365, + "learning_rate": 2.999322209123878e-05, + "loss": 1.7542, + "step": 177 + }, + { + "epoch": 0.328110599078341, + "grad_norm": 0.3260898292064667, + "learning_rate": 2.9992219324098545e-05, + "loss": 1.6049, + "step": 178 + }, + { + "epoch": 0.32995391705069127, + "grad_norm": 0.37945932149887085, + "learning_rate": 2.9991147425319346e-05, + "loss": 1.637, + "step": 179 + }, + { + "epoch": 0.3317972350230415, + "grad_norm": 0.36349308490753174, + "learning_rate": 2.9990006399843884e-05, + "loss": 1.7051, + "step": 180 + }, + { + "epoch": 0.3336405529953917, + "grad_norm": 0.35796141624450684, + "learning_rate": 2.998879625293362e-05, + "loss": 1.7094, + "step": 181 + }, + { + "epoch": 0.33548387096774196, + "grad_norm": 0.35643306374549866, + "learning_rate": 2.9987516990168743e-05, + "loss": 1.6021, + "step": 182 + }, + { + "epoch": 0.33732718894009217, + "grad_norm": 0.3745727837085724, + "learning_rate": 2.9986168617448153e-05, + "loss": 1.6267, + "step": 183 + }, + { + "epoch": 0.3391705069124424, + "grad_norm": 0.32228630781173706, + "learning_rate": 2.9984751140989417e-05, + "loss": 1.662, + "step": 184 + }, + { + "epoch": 0.34101382488479265, + "grad_norm": 0.3829532563686371, + "learning_rate": 2.9983264567328756e-05, + "loss": 1.6909, + "step": 185 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.3452930152416229, + "learning_rate": 2.9981708903321017e-05, + "loss": 1.6502, + "step": 186 + }, + { + "epoch": 0.34470046082949307, + "grad_norm": 0.3659391701221466, + "learning_rate": 2.9980084156139625e-05, + "loss": 1.6409, + "step": 187 + }, + { + "epoch": 0.34654377880184334, + "grad_norm": 0.38383013010025024, + "learning_rate": 2.9978390333276565e-05, + "loss": 1.6765, + "step": 188 + }, + { + "epoch": 0.34838709677419355, + "grad_norm": 0.34483417868614197, + "learning_rate": 2.9976627442542325e-05, + "loss": 1.7486, + "step": 189 + }, + { + "epoch": 0.35023041474654376, + "grad_norm": 0.3946613669395447, + "learning_rate": 2.997479549206591e-05, + "loss": 1.6863, + "step": 190 + }, + { + "epoch": 0.35207373271889403, + "grad_norm": 0.3084375262260437, + "learning_rate": 2.9972894490294738e-05, + "loss": 1.6223, + "step": 191 + }, + { + "epoch": 0.35391705069124424, + "grad_norm": 0.42278456687927246, + "learning_rate": 2.9970924445994645e-05, + "loss": 1.7044, + "step": 192 + }, + { + "epoch": 0.35576036866359445, + "grad_norm": 0.3211970925331116, + "learning_rate": 2.9968885368249847e-05, + "loss": 1.6907, + "step": 193 + }, + { + "epoch": 0.3576036866359447, + "grad_norm": 0.3954881727695465, + "learning_rate": 2.9966777266462863e-05, + "loss": 1.7002, + "step": 194 + }, + { + "epoch": 0.35944700460829493, + "grad_norm": 0.3460248112678528, + "learning_rate": 2.9964600150354512e-05, + "loss": 1.639, + "step": 195 + }, + { + "epoch": 0.36129032258064514, + "grad_norm": 0.3247598707675934, + "learning_rate": 2.9962354029963835e-05, + "loss": 1.679, + "step": 196 + }, + { + "epoch": 0.3631336405529954, + "grad_norm": 0.4037436544895172, + "learning_rate": 2.9960038915648076e-05, + "loss": 1.7343, + "step": 197 + }, + { + "epoch": 0.3649769585253456, + "grad_norm": 0.3309732973575592, + "learning_rate": 2.9957654818082615e-05, + "loss": 1.6759, + "step": 198 + }, + { + "epoch": 0.36682027649769583, + "grad_norm": 0.38870948553085327, + "learning_rate": 2.9955201748260923e-05, + "loss": 1.7189, + "step": 199 + }, + { + "epoch": 0.3686635944700461, + "grad_norm": 0.2937754988670349, + "learning_rate": 2.9952679717494516e-05, + "loss": 1.6882, + "step": 200 + }, + { + "epoch": 0.3705069124423963, + "grad_norm": 0.3606261909008026, + "learning_rate": 2.9950088737412898e-05, + "loss": 1.6536, + "step": 201 + }, + { + "epoch": 0.3723502304147465, + "grad_norm": 0.29472067952156067, + "learning_rate": 2.9947428819963526e-05, + "loss": 1.6957, + "step": 202 + }, + { + "epoch": 0.3741935483870968, + "grad_norm": 0.2893930673599243, + "learning_rate": 2.994469997741171e-05, + "loss": 1.6434, + "step": 203 + }, + { + "epoch": 0.376036866359447, + "grad_norm": 0.36216485500335693, + "learning_rate": 2.994190222234061e-05, + "loss": 1.6897, + "step": 204 + }, + { + "epoch": 0.3778801843317972, + "grad_norm": 0.2846887409687042, + "learning_rate": 2.9939035567651146e-05, + "loss": 1.6727, + "step": 205 + }, + { + "epoch": 0.3797235023041475, + "grad_norm": 0.34425318241119385, + "learning_rate": 2.9936100026561933e-05, + "loss": 1.6824, + "step": 206 + }, + { + "epoch": 0.3815668202764977, + "grad_norm": 0.31369149684906006, + "learning_rate": 2.9933095612609253e-05, + "loss": 1.6703, + "step": 207 + }, + { + "epoch": 0.38341013824884795, + "grad_norm": 0.31020811200141907, + "learning_rate": 2.993002233964696e-05, + "loss": 1.7284, + "step": 208 + }, + { + "epoch": 0.38525345622119817, + "grad_norm": 0.3632746934890747, + "learning_rate": 2.9926880221846435e-05, + "loss": 1.6617, + "step": 209 + }, + { + "epoch": 0.3870967741935484, + "grad_norm": 0.2814621329307556, + "learning_rate": 2.9923669273696506e-05, + "loss": 1.5947, + "step": 210 + }, + { + "epoch": 0.38894009216589864, + "grad_norm": 0.32484638690948486, + "learning_rate": 2.9920389510003395e-05, + "loss": 1.6403, + "step": 211 + }, + { + "epoch": 0.39078341013824885, + "grad_norm": 0.31668680906295776, + "learning_rate": 2.9917040945890638e-05, + "loss": 1.7241, + "step": 212 + }, + { + "epoch": 0.39262672811059907, + "grad_norm": 0.29519209265708923, + "learning_rate": 2.9913623596799032e-05, + "loss": 1.6997, + "step": 213 + }, + { + "epoch": 0.39447004608294933, + "grad_norm": 0.3414634168148041, + "learning_rate": 2.9910137478486545e-05, + "loss": 1.6451, + "step": 214 + }, + { + "epoch": 0.39631336405529954, + "grad_norm": 0.2943251132965088, + "learning_rate": 2.990658260702826e-05, + "loss": 1.6784, + "step": 215 + }, + { + "epoch": 0.39815668202764976, + "grad_norm": 0.3138771653175354, + "learning_rate": 2.9902958998816274e-05, + "loss": 1.7088, + "step": 216 + }, + { + "epoch": 0.4, + "grad_norm": 0.296895831823349, + "learning_rate": 2.989926667055966e-05, + "loss": 1.6855, + "step": 217 + }, + { + "epoch": 0.40184331797235023, + "grad_norm": 0.29469338059425354, + "learning_rate": 2.989550563928436e-05, + "loss": 1.6437, + "step": 218 + }, + { + "epoch": 0.40368663594470044, + "grad_norm": 0.3117813169956207, + "learning_rate": 2.9891675922333125e-05, + "loss": 1.6708, + "step": 219 + }, + { + "epoch": 0.4055299539170507, + "grad_norm": 0.2911444902420044, + "learning_rate": 2.9887777537365416e-05, + "loss": 1.6655, + "step": 220 + }, + { + "epoch": 0.4073732718894009, + "grad_norm": 0.29274192452430725, + "learning_rate": 2.9883810502357346e-05, + "loss": 1.6737, + "step": 221 + }, + { + "epoch": 0.40921658986175113, + "grad_norm": 0.325166255235672, + "learning_rate": 2.9879774835601574e-05, + "loss": 1.6562, + "step": 222 + }, + { + "epoch": 0.4110599078341014, + "grad_norm": 0.3723132312297821, + "learning_rate": 2.987567055570724e-05, + "loss": 1.696, + "step": 223 + }, + { + "epoch": 0.4129032258064516, + "grad_norm": 0.2948567867279053, + "learning_rate": 2.987149768159987e-05, + "loss": 1.5771, + "step": 224 + }, + { + "epoch": 0.4147465437788018, + "grad_norm": 0.3130621314048767, + "learning_rate": 2.986725623252128e-05, + "loss": 1.7274, + "step": 225 + }, + { + "epoch": 0.4165898617511521, + "grad_norm": 0.283635675907135, + "learning_rate": 2.9862946228029507e-05, + "loss": 1.6277, + "step": 226 + }, + { + "epoch": 0.4184331797235023, + "grad_norm": 0.2891738712787628, + "learning_rate": 2.9858567687998702e-05, + "loss": 1.6161, + "step": 227 + }, + { + "epoch": 0.4202764976958525, + "grad_norm": 0.3050073981285095, + "learning_rate": 2.9854120632619053e-05, + "loss": 1.6358, + "step": 228 + }, + { + "epoch": 0.4221198156682028, + "grad_norm": 0.29393428564071655, + "learning_rate": 2.9849605082396678e-05, + "loss": 1.7176, + "step": 229 + }, + { + "epoch": 0.423963133640553, + "grad_norm": 0.29933053255081177, + "learning_rate": 2.9845021058153532e-05, + "loss": 1.6292, + "step": 230 + }, + { + "epoch": 0.4258064516129032, + "grad_norm": 0.2925868034362793, + "learning_rate": 2.984036858102732e-05, + "loss": 1.6453, + "step": 231 + }, + { + "epoch": 0.42764976958525347, + "grad_norm": 0.30412405729293823, + "learning_rate": 2.98356476724714e-05, + "loss": 1.7311, + "step": 232 + }, + { + "epoch": 0.4294930875576037, + "grad_norm": 0.29768475890159607, + "learning_rate": 2.9830858354254672e-05, + "loss": 1.632, + "step": 233 + }, + { + "epoch": 0.4313364055299539, + "grad_norm": 0.3007776141166687, + "learning_rate": 2.9826000648461484e-05, + "loss": 1.6307, + "step": 234 + }, + { + "epoch": 0.43317972350230416, + "grad_norm": 0.32172518968582153, + "learning_rate": 2.982107457749153e-05, + "loss": 1.6314, + "step": 235 + }, + { + "epoch": 0.43502304147465437, + "grad_norm": 0.28600960969924927, + "learning_rate": 2.9816080164059758e-05, + "loss": 1.6417, + "step": 236 + }, + { + "epoch": 0.4368663594470046, + "grad_norm": 0.2792605757713318, + "learning_rate": 2.981101743119624e-05, + "loss": 1.5736, + "step": 237 + }, + { + "epoch": 0.43870967741935485, + "grad_norm": 0.3138410747051239, + "learning_rate": 2.9805886402246084e-05, + "loss": 1.6921, + "step": 238 + }, + { + "epoch": 0.44055299539170506, + "grad_norm": 0.2832198739051819, + "learning_rate": 2.9800687100869334e-05, + "loss": 1.642, + "step": 239 + }, + { + "epoch": 0.4423963133640553, + "grad_norm": 0.29424023628234863, + "learning_rate": 2.9795419551040836e-05, + "loss": 1.6786, + "step": 240 + }, + { + "epoch": 0.44423963133640554, + "grad_norm": 0.30614927411079407, + "learning_rate": 2.9790083777050148e-05, + "loss": 1.6565, + "step": 241 + }, + { + "epoch": 0.44608294930875575, + "grad_norm": 0.29164189100265503, + "learning_rate": 2.9784679803501416e-05, + "loss": 1.7311, + "step": 242 + }, + { + "epoch": 0.447926267281106, + "grad_norm": 0.29889318346977234, + "learning_rate": 2.977920765531327e-05, + "loss": 1.6551, + "step": 243 + }, + { + "epoch": 0.4497695852534562, + "grad_norm": 0.2796136438846588, + "learning_rate": 2.9773667357718706e-05, + "loss": 1.6495, + "step": 244 + }, + { + "epoch": 0.45161290322580644, + "grad_norm": 0.3036425709724426, + "learning_rate": 2.9768058936264967e-05, + "loss": 1.6847, + "step": 245 + }, + { + "epoch": 0.4534562211981567, + "grad_norm": 0.2765255868434906, + "learning_rate": 2.976238241681342e-05, + "loss": 1.642, + "step": 246 + }, + { + "epoch": 0.4552995391705069, + "grad_norm": 0.28592586517333984, + "learning_rate": 2.9756637825539453e-05, + "loss": 1.5912, + "step": 247 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.29503703117370605, + "learning_rate": 2.9750825188932334e-05, + "loss": 1.6017, + "step": 248 + }, + { + "epoch": 0.4589861751152074, + "grad_norm": 0.2970612645149231, + "learning_rate": 2.9744944533795112e-05, + "loss": 1.6603, + "step": 249 + }, + { + "epoch": 0.4608294930875576, + "grad_norm": 0.28196001052856445, + "learning_rate": 2.973899588724448e-05, + "loss": 1.6234, + "step": 250 + }, + { + "epoch": 0.4626728110599078, + "grad_norm": 0.28284621238708496, + "learning_rate": 2.973297927671063e-05, + "loss": 1.6411, + "step": 251 + }, + { + "epoch": 0.4645161290322581, + "grad_norm": 0.28718075156211853, + "learning_rate": 2.9726894729937177e-05, + "loss": 1.6307, + "step": 252 + }, + { + "epoch": 0.4663594470046083, + "grad_norm": 0.28701338171958923, + "learning_rate": 2.972074227498098e-05, + "loss": 1.6697, + "step": 253 + }, + { + "epoch": 0.4682027649769585, + "grad_norm": 0.28013625741004944, + "learning_rate": 2.971452194021204e-05, + "loss": 1.6441, + "step": 254 + }, + { + "epoch": 0.4700460829493088, + "grad_norm": 0.28640949726104736, + "learning_rate": 2.9708233754313365e-05, + "loss": 1.6774, + "step": 255 + }, + { + "epoch": 0.471889400921659, + "grad_norm": 0.2968595027923584, + "learning_rate": 2.9701877746280843e-05, + "loss": 1.693, + "step": 256 + }, + { + "epoch": 0.4737327188940092, + "grad_norm": 0.28468331694602966, + "learning_rate": 2.9695453945423087e-05, + "loss": 1.6944, + "step": 257 + }, + { + "epoch": 0.47557603686635946, + "grad_norm": 0.31435340642929077, + "learning_rate": 2.9688962381361317e-05, + "loss": 1.6628, + "step": 258 + }, + { + "epoch": 0.4774193548387097, + "grad_norm": 0.2782823443412781, + "learning_rate": 2.968240308402923e-05, + "loss": 1.6312, + "step": 259 + }, + { + "epoch": 0.4792626728110599, + "grad_norm": 0.288622111082077, + "learning_rate": 2.967577608367285e-05, + "loss": 1.6166, + "step": 260 + }, + { + "epoch": 0.48110599078341015, + "grad_norm": 0.2862098217010498, + "learning_rate": 2.9669081410850378e-05, + "loss": 1.5918, + "step": 261 + }, + { + "epoch": 0.48294930875576036, + "grad_norm": 0.2974812686443329, + "learning_rate": 2.966231909643208e-05, + "loss": 1.6475, + "step": 262 + }, + { + "epoch": 0.4847926267281106, + "grad_norm": 0.31905004382133484, + "learning_rate": 2.9655489171600118e-05, + "loss": 1.6218, + "step": 263 + }, + { + "epoch": 0.48663594470046084, + "grad_norm": 0.29999393224716187, + "learning_rate": 2.9648591667848428e-05, + "loss": 1.7007, + "step": 264 + }, + { + "epoch": 0.48847926267281105, + "grad_norm": 0.31066837906837463, + "learning_rate": 2.9641626616982555e-05, + "loss": 1.6758, + "step": 265 + }, + { + "epoch": 0.49032258064516127, + "grad_norm": 0.30834177136421204, + "learning_rate": 2.9634594051119515e-05, + "loss": 1.6889, + "step": 266 + }, + { + "epoch": 0.49216589861751153, + "grad_norm": 0.29685091972351074, + "learning_rate": 2.9627494002687653e-05, + "loss": 1.7099, + "step": 267 + }, + { + "epoch": 0.49400921658986174, + "grad_norm": 0.3066437244415283, + "learning_rate": 2.9620326504426476e-05, + "loss": 1.6494, + "step": 268 + }, + { + "epoch": 0.49585253456221196, + "grad_norm": 0.28298285603523254, + "learning_rate": 2.9613091589386526e-05, + "loss": 1.6435, + "step": 269 + }, + { + "epoch": 0.4976958525345622, + "grad_norm": 0.2950513958930969, + "learning_rate": 2.9605789290929214e-05, + "loss": 1.6588, + "step": 270 + }, + { + "epoch": 0.49953917050691243, + "grad_norm": 0.2809874713420868, + "learning_rate": 2.9598419642726655e-05, + "loss": 1.6463, + "step": 271 + }, + { + "epoch": 0.5013824884792627, + "grad_norm": 0.29350385069847107, + "learning_rate": 2.9590982678761544e-05, + "loss": 1.6022, + "step": 272 + }, + { + "epoch": 0.5032258064516129, + "grad_norm": 0.28711917996406555, + "learning_rate": 2.958347843332696e-05, + "loss": 1.6602, + "step": 273 + }, + { + "epoch": 0.5050691244239631, + "grad_norm": 0.2757432162761688, + "learning_rate": 2.957590694102624e-05, + "loss": 1.6223, + "step": 274 + }, + { + "epoch": 0.5069124423963134, + "grad_norm": 0.27851778268814087, + "learning_rate": 2.9568268236772816e-05, + "loss": 1.6716, + "step": 275 + }, + { + "epoch": 0.5087557603686635, + "grad_norm": 0.28100845217704773, + "learning_rate": 2.956056235579002e-05, + "loss": 1.6326, + "step": 276 + }, + { + "epoch": 0.5105990783410138, + "grad_norm": 0.2892681360244751, + "learning_rate": 2.955278933361097e-05, + "loss": 1.6584, + "step": 277 + }, + { + "epoch": 0.5124423963133641, + "grad_norm": 0.27751055359840393, + "learning_rate": 2.9544949206078372e-05, + "loss": 1.6457, + "step": 278 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.26928141713142395, + "learning_rate": 2.9537042009344376e-05, + "loss": 1.6027, + "step": 279 + }, + { + "epoch": 0.5161290322580645, + "grad_norm": 0.26717764139175415, + "learning_rate": 2.9529067779870385e-05, + "loss": 1.6157, + "step": 280 + }, + { + "epoch": 0.5179723502304148, + "grad_norm": 0.26703840494155884, + "learning_rate": 2.952102655442692e-05, + "loss": 1.6148, + "step": 281 + }, + { + "epoch": 0.5198156682027649, + "grad_norm": 0.2838272452354431, + "learning_rate": 2.9512918370093407e-05, + "loss": 1.6785, + "step": 282 + }, + { + "epoch": 0.5216589861751152, + "grad_norm": 0.281730592250824, + "learning_rate": 2.950474326425805e-05, + "loss": 1.5828, + "step": 283 + }, + { + "epoch": 0.5235023041474655, + "grad_norm": 0.359958678483963, + "learning_rate": 2.949650127461764e-05, + "loss": 1.6606, + "step": 284 + }, + { + "epoch": 0.5253456221198156, + "grad_norm": 0.276723712682724, + "learning_rate": 2.948819243917737e-05, + "loss": 1.7019, + "step": 285 + }, + { + "epoch": 0.5271889400921659, + "grad_norm": 0.28240787982940674, + "learning_rate": 2.947981679625067e-05, + "loss": 1.7214, + "step": 286 + }, + { + "epoch": 0.5290322580645161, + "grad_norm": 0.278328001499176, + "learning_rate": 2.947137438445904e-05, + "loss": 1.6599, + "step": 287 + }, + { + "epoch": 0.5308755760368664, + "grad_norm": 0.2821551561355591, + "learning_rate": 2.9462865242731856e-05, + "loss": 1.6602, + "step": 288 + }, + { + "epoch": 0.5327188940092166, + "grad_norm": 0.2736065685749054, + "learning_rate": 2.9454289410306202e-05, + "loss": 1.5909, + "step": 289 + }, + { + "epoch": 0.5345622119815668, + "grad_norm": 0.32163509726524353, + "learning_rate": 2.944564692672667e-05, + "loss": 1.7039, + "step": 290 + }, + { + "epoch": 0.5364055299539171, + "grad_norm": 0.2821354568004608, + "learning_rate": 2.9436937831845217e-05, + "loss": 1.6789, + "step": 291 + }, + { + "epoch": 0.5382488479262673, + "grad_norm": 0.30141276121139526, + "learning_rate": 2.942816216582093e-05, + "loss": 1.6341, + "step": 292 + }, + { + "epoch": 0.5400921658986175, + "grad_norm": 0.2816147208213806, + "learning_rate": 2.9419319969119875e-05, + "loss": 1.5926, + "step": 293 + }, + { + "epoch": 0.5419354838709678, + "grad_norm": 0.2912384569644928, + "learning_rate": 2.9410411282514913e-05, + "loss": 1.6507, + "step": 294 + }, + { + "epoch": 0.543778801843318, + "grad_norm": 0.3174484670162201, + "learning_rate": 2.940143614708549e-05, + "loss": 1.6504, + "step": 295 + }, + { + "epoch": 0.5456221198156682, + "grad_norm": 0.2888404130935669, + "learning_rate": 2.939239460421746e-05, + "loss": 1.6762, + "step": 296 + }, + { + "epoch": 0.5474654377880185, + "grad_norm": 0.31422436237335205, + "learning_rate": 2.93832866956029e-05, + "loss": 1.6301, + "step": 297 + }, + { + "epoch": 0.5493087557603686, + "grad_norm": 0.3254394829273224, + "learning_rate": 2.9374112463239896e-05, + "loss": 1.7101, + "step": 298 + }, + { + "epoch": 0.5511520737327189, + "grad_norm": 0.3167421817779541, + "learning_rate": 2.9364871949432378e-05, + "loss": 1.6871, + "step": 299 + }, + { + "epoch": 0.5529953917050692, + "grad_norm": 0.3247944712638855, + "learning_rate": 2.9355565196789906e-05, + "loss": 1.7028, + "step": 300 + }, + { + "epoch": 0.5548387096774193, + "grad_norm": 0.27095088362693787, + "learning_rate": 2.9346192248227476e-05, + "loss": 1.6229, + "step": 301 + }, + { + "epoch": 0.5566820276497696, + "grad_norm": 0.31862378120422363, + "learning_rate": 2.9336753146965327e-05, + "loss": 1.666, + "step": 302 + }, + { + "epoch": 0.5585253456221199, + "grad_norm": 0.28205209970474243, + "learning_rate": 2.9327247936528742e-05, + "loss": 1.5925, + "step": 303 + }, + { + "epoch": 0.56036866359447, + "grad_norm": 0.29554882645606995, + "learning_rate": 2.9317676660747837e-05, + "loss": 1.6605, + "step": 304 + }, + { + "epoch": 0.5622119815668203, + "grad_norm": 0.2860583961009979, + "learning_rate": 2.9308039363757372e-05, + "loss": 1.6371, + "step": 305 + }, + { + "epoch": 0.5640552995391706, + "grad_norm": 0.27211877703666687, + "learning_rate": 2.9298336089996538e-05, + "loss": 1.7176, + "step": 306 + }, + { + "epoch": 0.5658986175115207, + "grad_norm": 0.3138637840747833, + "learning_rate": 2.9288566884208766e-05, + "loss": 1.6378, + "step": 307 + }, + { + "epoch": 0.567741935483871, + "grad_norm": 0.2751595973968506, + "learning_rate": 2.9278731791441497e-05, + "loss": 1.6313, + "step": 308 + }, + { + "epoch": 0.5695852534562212, + "grad_norm": 0.28140899538993835, + "learning_rate": 2.9268830857045997e-05, + "loss": 1.6114, + "step": 309 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.2734344005584717, + "learning_rate": 2.9258864126677132e-05, + "loss": 1.6438, + "step": 310 + }, + { + "epoch": 0.5732718894009217, + "grad_norm": 0.30163639783859253, + "learning_rate": 2.9248831646293174e-05, + "loss": 1.6521, + "step": 311 + }, + { + "epoch": 0.5751152073732719, + "grad_norm": 0.28159695863723755, + "learning_rate": 2.9238733462155564e-05, + "loss": 1.6399, + "step": 312 + }, + { + "epoch": 0.5769585253456221, + "grad_norm": 0.2891719341278076, + "learning_rate": 2.9228569620828735e-05, + "loss": 1.6316, + "step": 313 + }, + { + "epoch": 0.5788018433179724, + "grad_norm": 0.27711349725723267, + "learning_rate": 2.921834016917986e-05, + "loss": 1.5787, + "step": 314 + }, + { + "epoch": 0.5806451612903226, + "grad_norm": 0.2825881540775299, + "learning_rate": 2.920804515437865e-05, + "loss": 1.6223, + "step": 315 + }, + { + "epoch": 0.5824884792626728, + "grad_norm": 0.2809242010116577, + "learning_rate": 2.9197684623897157e-05, + "loss": 1.6368, + "step": 316 + }, + { + "epoch": 0.584331797235023, + "grad_norm": 0.2902085781097412, + "learning_rate": 2.9187258625509518e-05, + "loss": 1.6855, + "step": 317 + }, + { + "epoch": 0.5861751152073733, + "grad_norm": 0.2923787236213684, + "learning_rate": 2.917676720729177e-05, + "loss": 1.6448, + "step": 318 + }, + { + "epoch": 0.5880184331797235, + "grad_norm": 0.2834003269672394, + "learning_rate": 2.916621041762159e-05, + "loss": 1.6295, + "step": 319 + }, + { + "epoch": 0.5898617511520737, + "grad_norm": 0.2824580669403076, + "learning_rate": 2.9155588305178113e-05, + "loss": 1.5738, + "step": 320 + }, + { + "epoch": 0.591705069124424, + "grad_norm": 0.30301326513290405, + "learning_rate": 2.9144900918941687e-05, + "loss": 1.6247, + "step": 321 + }, + { + "epoch": 0.5935483870967742, + "grad_norm": 0.2766891121864319, + "learning_rate": 2.9134148308193637e-05, + "loss": 1.7135, + "step": 322 + }, + { + "epoch": 0.5953917050691244, + "grad_norm": 0.2816697061061859, + "learning_rate": 2.9123330522516053e-05, + "loss": 1.6522, + "step": 323 + }, + { + "epoch": 0.5972350230414747, + "grad_norm": 0.28478461503982544, + "learning_rate": 2.9112447611791563e-05, + "loss": 1.6347, + "step": 324 + }, + { + "epoch": 0.5990783410138248, + "grad_norm": 0.27743953466415405, + "learning_rate": 2.9101499626203102e-05, + "loss": 1.6071, + "step": 325 + }, + { + "epoch": 0.6009216589861751, + "grad_norm": 0.27698153257369995, + "learning_rate": 2.9090486616233654e-05, + "loss": 1.6191, + "step": 326 + }, + { + "epoch": 0.6027649769585254, + "grad_norm": 0.2867109477519989, + "learning_rate": 2.907940863266607e-05, + "loss": 1.6427, + "step": 327 + }, + { + "epoch": 0.6046082949308755, + "grad_norm": 0.26966315507888794, + "learning_rate": 2.906826572658278e-05, + "loss": 1.5825, + "step": 328 + }, + { + "epoch": 0.6064516129032258, + "grad_norm": 0.2749760150909424, + "learning_rate": 2.9057057949365602e-05, + "loss": 1.6189, + "step": 329 + }, + { + "epoch": 0.6082949308755761, + "grad_norm": 0.30331194400787354, + "learning_rate": 2.904578535269547e-05, + "loss": 1.6485, + "step": 330 + }, + { + "epoch": 0.6101382488479262, + "grad_norm": 0.2790120244026184, + "learning_rate": 2.9034447988552227e-05, + "loss": 1.6874, + "step": 331 + }, + { + "epoch": 0.6119815668202765, + "grad_norm": 0.2863958477973938, + "learning_rate": 2.902304590921435e-05, + "loss": 1.6805, + "step": 332 + }, + { + "epoch": 0.6138248847926268, + "grad_norm": 0.28442642092704773, + "learning_rate": 2.9011579167258756e-05, + "loss": 1.6611, + "step": 333 + }, + { + "epoch": 0.6156682027649769, + "grad_norm": 0.27127203345298767, + "learning_rate": 2.90000478155605e-05, + "loss": 1.5686, + "step": 334 + }, + { + "epoch": 0.6175115207373272, + "grad_norm": 0.28976717591285706, + "learning_rate": 2.8988451907292594e-05, + "loss": 1.6636, + "step": 335 + }, + { + "epoch": 0.6193548387096774, + "grad_norm": 0.2731335461139679, + "learning_rate": 2.8976791495925704e-05, + "loss": 1.7131, + "step": 336 + }, + { + "epoch": 0.6211981566820276, + "grad_norm": 0.2786687910556793, + "learning_rate": 2.896506663522795e-05, + "loss": 1.6664, + "step": 337 + }, + { + "epoch": 0.6230414746543779, + "grad_norm": 0.2858924865722656, + "learning_rate": 2.8953277379264633e-05, + "loss": 1.6567, + "step": 338 + }, + { + "epoch": 0.6248847926267281, + "grad_norm": 0.2715083956718445, + "learning_rate": 2.8941423782397987e-05, + "loss": 1.6504, + "step": 339 + }, + { + "epoch": 0.6267281105990783, + "grad_norm": 0.2730218470096588, + "learning_rate": 2.892950589928694e-05, + "loss": 1.6381, + "step": 340 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.2796657383441925, + "learning_rate": 2.8917523784886846e-05, + "loss": 1.6845, + "step": 341 + }, + { + "epoch": 0.6304147465437788, + "grad_norm": 0.28790879249572754, + "learning_rate": 2.890547749444925e-05, + "loss": 1.6751, + "step": 342 + }, + { + "epoch": 0.632258064516129, + "grad_norm": 0.27353277802467346, + "learning_rate": 2.8893367083521616e-05, + "loss": 1.6247, + "step": 343 + }, + { + "epoch": 0.6341013824884792, + "grad_norm": 0.2717505395412445, + "learning_rate": 2.888119260794708e-05, + "loss": 1.6086, + "step": 344 + }, + { + "epoch": 0.6359447004608295, + "grad_norm": 0.27940690517425537, + "learning_rate": 2.8868954123864194e-05, + "loss": 1.653, + "step": 345 + }, + { + "epoch": 0.6377880184331797, + "grad_norm": 0.265103280544281, + "learning_rate": 2.885665168770666e-05, + "loss": 1.6432, + "step": 346 + }, + { + "epoch": 0.6396313364055299, + "grad_norm": 0.26981207728385925, + "learning_rate": 2.8844285356203074e-05, + "loss": 1.6346, + "step": 347 + }, + { + "epoch": 0.6414746543778802, + "grad_norm": 0.2731129229068756, + "learning_rate": 2.8831855186376672e-05, + "loss": 1.6907, + "step": 348 + }, + { + "epoch": 0.6433179723502304, + "grad_norm": 0.2778747081756592, + "learning_rate": 2.8819361235545047e-05, + "loss": 1.699, + "step": 349 + }, + { + "epoch": 0.6451612903225806, + "grad_norm": 0.27246907353401184, + "learning_rate": 2.8806803561319903e-05, + "loss": 1.6464, + "step": 350 + }, + { + "epoch": 0.6470046082949309, + "grad_norm": 0.2664584517478943, + "learning_rate": 2.8794182221606784e-05, + "loss": 1.5384, + "step": 351 + }, + { + "epoch": 0.6488479262672812, + "grad_norm": 0.2673085033893585, + "learning_rate": 2.878149727460481e-05, + "loss": 1.571, + "step": 352 + }, + { + "epoch": 0.6506912442396313, + "grad_norm": 0.28247496485710144, + "learning_rate": 2.876874877880639e-05, + "loss": 1.5831, + "step": 353 + }, + { + "epoch": 0.6525345622119816, + "grad_norm": 0.2843359112739563, + "learning_rate": 2.8755936792996987e-05, + "loss": 1.6923, + "step": 354 + }, + { + "epoch": 0.6543778801843319, + "grad_norm": 0.27128326892852783, + "learning_rate": 2.8743061376254813e-05, + "loss": 1.6356, + "step": 355 + }, + { + "epoch": 0.656221198156682, + "grad_norm": 0.2898774743080139, + "learning_rate": 2.873012258795057e-05, + "loss": 1.6479, + "step": 356 + }, + { + "epoch": 0.6580645161290323, + "grad_norm": 0.28426289558410645, + "learning_rate": 2.8717120487747193e-05, + "loss": 1.629, + "step": 357 + }, + { + "epoch": 0.6599078341013825, + "grad_norm": 0.2633942663669586, + "learning_rate": 2.870405513559954e-05, + "loss": 1.5984, + "step": 358 + }, + { + "epoch": 0.6617511520737327, + "grad_norm": 0.31980207562446594, + "learning_rate": 2.8690926591754142e-05, + "loss": 1.5927, + "step": 359 + }, + { + "epoch": 0.663594470046083, + "grad_norm": 0.2780967056751251, + "learning_rate": 2.8677734916748927e-05, + "loss": 1.6457, + "step": 360 + }, + { + "epoch": 0.6654377880184332, + "grad_norm": 0.3023218810558319, + "learning_rate": 2.866448017141291e-05, + "loss": 1.6161, + "step": 361 + }, + { + "epoch": 0.6672811059907834, + "grad_norm": 0.2830914556980133, + "learning_rate": 2.865116241686595e-05, + "loss": 1.6897, + "step": 362 + }, + { + "epoch": 0.6691244239631337, + "grad_norm": 0.29188162088394165, + "learning_rate": 2.863778171451845e-05, + "loss": 1.6293, + "step": 363 + }, + { + "epoch": 0.6709677419354839, + "grad_norm": 0.3087140619754791, + "learning_rate": 2.8624338126071073e-05, + "loss": 1.6143, + "step": 364 + }, + { + "epoch": 0.6728110599078341, + "grad_norm": 0.26930543780326843, + "learning_rate": 2.861083171351446e-05, + "loss": 1.5878, + "step": 365 + }, + { + "epoch": 0.6746543778801843, + "grad_norm": 0.30629977583885193, + "learning_rate": 2.8597262539128947e-05, + "loss": 1.6939, + "step": 366 + }, + { + "epoch": 0.6764976958525346, + "grad_norm": 0.2783128321170807, + "learning_rate": 2.858363066548427e-05, + "loss": 1.66, + "step": 367 + }, + { + "epoch": 0.6783410138248848, + "grad_norm": 0.27261894941329956, + "learning_rate": 2.856993615543929e-05, + "loss": 1.6183, + "step": 368 + }, + { + "epoch": 0.680184331797235, + "grad_norm": 0.29020169377326965, + "learning_rate": 2.8556179072141693e-05, + "loss": 1.6187, + "step": 369 + }, + { + "epoch": 0.6820276497695853, + "grad_norm": 0.2697452902793884, + "learning_rate": 2.8542359479027693e-05, + "loss": 1.5613, + "step": 370 + }, + { + "epoch": 0.6838709677419355, + "grad_norm": 0.27904996275901794, + "learning_rate": 2.8528477439821753e-05, + "loss": 1.6317, + "step": 371 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.2748742997646332, + "learning_rate": 2.8514533018536286e-05, + "loss": 1.6531, + "step": 372 + }, + { + "epoch": 0.687557603686636, + "grad_norm": 0.30121949315071106, + "learning_rate": 2.8500526279471362e-05, + "loss": 1.6045, + "step": 373 + }, + { + "epoch": 0.6894009216589861, + "grad_norm": 0.26736557483673096, + "learning_rate": 2.8486457287214403e-05, + "loss": 1.6746, + "step": 374 + }, + { + "epoch": 0.6912442396313364, + "grad_norm": 0.2701917886734009, + "learning_rate": 2.8472326106639896e-05, + "loss": 1.6882, + "step": 375 + }, + { + "epoch": 0.6930875576036867, + "grad_norm": 0.2666465640068054, + "learning_rate": 2.8458132802909075e-05, + "loss": 1.5905, + "step": 376 + }, + { + "epoch": 0.6949308755760368, + "grad_norm": 0.2787630259990692, + "learning_rate": 2.8443877441469653e-05, + "loss": 1.6351, + "step": 377 + }, + { + "epoch": 0.6967741935483871, + "grad_norm": 0.2640557289123535, + "learning_rate": 2.8429560088055502e-05, + "loss": 1.6291, + "step": 378 + }, + { + "epoch": 0.6986175115207374, + "grad_norm": 0.2768750488758087, + "learning_rate": 2.8415180808686326e-05, + "loss": 1.6113, + "step": 379 + }, + { + "epoch": 0.7004608294930875, + "grad_norm": 0.27341142296791077, + "learning_rate": 2.84007396696674e-05, + "loss": 1.6397, + "step": 380 + }, + { + "epoch": 0.7023041474654378, + "grad_norm": 0.2810218334197998, + "learning_rate": 2.8386236737589244e-05, + "loss": 1.6255, + "step": 381 + }, + { + "epoch": 0.7041474654377881, + "grad_norm": 0.2625837028026581, + "learning_rate": 2.8371672079327304e-05, + "loss": 1.5909, + "step": 382 + }, + { + "epoch": 0.7059907834101382, + "grad_norm": 0.28470954298973083, + "learning_rate": 2.835704576204167e-05, + "loss": 1.668, + "step": 383 + }, + { + "epoch": 0.7078341013824885, + "grad_norm": 0.2870181202888489, + "learning_rate": 2.8342357853176742e-05, + "loss": 1.655, + "step": 384 + }, + { + "epoch": 0.7096774193548387, + "grad_norm": 0.2737495005130768, + "learning_rate": 2.8327608420460933e-05, + "loss": 1.6606, + "step": 385 + }, + { + "epoch": 0.7115207373271889, + "grad_norm": 0.2716140151023865, + "learning_rate": 2.8312797531906346e-05, + "loss": 1.6487, + "step": 386 + }, + { + "epoch": 0.7133640552995392, + "grad_norm": 0.2937834560871124, + "learning_rate": 2.8297925255808484e-05, + "loss": 1.5784, + "step": 387 + }, + { + "epoch": 0.7152073732718894, + "grad_norm": 0.25758716464042664, + "learning_rate": 2.82829916607459e-05, + "loss": 1.5551, + "step": 388 + }, + { + "epoch": 0.7170506912442396, + "grad_norm": 0.26855534315109253, + "learning_rate": 2.826799681557991e-05, + "loss": 1.6242, + "step": 389 + }, + { + "epoch": 0.7188940092165899, + "grad_norm": 0.29118281602859497, + "learning_rate": 2.8252940789454268e-05, + "loss": 1.6404, + "step": 390 + }, + { + "epoch": 0.7207373271889401, + "grad_norm": 0.27645623683929443, + "learning_rate": 2.823782365179482e-05, + "loss": 1.7135, + "step": 391 + }, + { + "epoch": 0.7225806451612903, + "grad_norm": 0.2710598409175873, + "learning_rate": 2.822264547230924e-05, + "loss": 1.6472, + "step": 392 + }, + { + "epoch": 0.7244239631336405, + "grad_norm": 0.2802148163318634, + "learning_rate": 2.820740632098665e-05, + "loss": 1.5996, + "step": 393 + }, + { + "epoch": 0.7262672811059908, + "grad_norm": 0.28701063990592957, + "learning_rate": 2.8192106268097336e-05, + "loss": 1.6192, + "step": 394 + }, + { + "epoch": 0.728110599078341, + "grad_norm": 0.28330931067466736, + "learning_rate": 2.8176745384192417e-05, + "loss": 1.6183, + "step": 395 + }, + { + "epoch": 0.7299539170506912, + "grad_norm": 0.26512524485588074, + "learning_rate": 2.8161323740103495e-05, + "loss": 1.6092, + "step": 396 + }, + { + "epoch": 0.7317972350230415, + "grad_norm": 0.27630165219306946, + "learning_rate": 2.814584140694237e-05, + "loss": 1.6938, + "step": 397 + }, + { + "epoch": 0.7336405529953917, + "grad_norm": 0.2831325829029083, + "learning_rate": 2.8130298456100667e-05, + "loss": 1.6665, + "step": 398 + }, + { + "epoch": 0.7354838709677419, + "grad_norm": 0.2725171744823456, + "learning_rate": 2.811469495924955e-05, + "loss": 1.6464, + "step": 399 + }, + { + "epoch": 0.7373271889400922, + "grad_norm": 0.2628806233406067, + "learning_rate": 2.8099030988339353e-05, + "loss": 1.5455, + "step": 400 + }, + { + "epoch": 0.7391705069124423, + "grad_norm": 0.26690706610679626, + "learning_rate": 2.8083306615599283e-05, + "loss": 1.6348, + "step": 401 + }, + { + "epoch": 0.7410138248847926, + "grad_norm": 0.2794962227344513, + "learning_rate": 2.8067521913537047e-05, + "loss": 1.6365, + "step": 402 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.2658675014972687, + "learning_rate": 2.8051676954938574e-05, + "loss": 1.6348, + "step": 403 + }, + { + "epoch": 0.744700460829493, + "grad_norm": 0.27006804943084717, + "learning_rate": 2.8035771812867613e-05, + "loss": 1.6384, + "step": 404 + }, + { + "epoch": 0.7465437788018433, + "grad_norm": 0.2776554822921753, + "learning_rate": 2.801980656066545e-05, + "loss": 1.6978, + "step": 405 + }, + { + "epoch": 0.7483870967741936, + "grad_norm": 0.2728254497051239, + "learning_rate": 2.8003781271950535e-05, + "loss": 1.6924, + "step": 406 + }, + { + "epoch": 0.7502304147465437, + "grad_norm": 0.2657453715801239, + "learning_rate": 2.7987696020618163e-05, + "loss": 1.5734, + "step": 407 + }, + { + "epoch": 0.752073732718894, + "grad_norm": 0.26352396607398987, + "learning_rate": 2.7971550880840138e-05, + "loss": 1.6261, + "step": 408 + }, + { + "epoch": 0.7539170506912443, + "grad_norm": 0.2759632468223572, + "learning_rate": 2.79553459270644e-05, + "loss": 1.6264, + "step": 409 + }, + { + "epoch": 0.7557603686635944, + "grad_norm": 0.27137723565101624, + "learning_rate": 2.7939081234014708e-05, + "loss": 1.6432, + "step": 410 + }, + { + "epoch": 0.7576036866359447, + "grad_norm": 0.26721593737602234, + "learning_rate": 2.7922756876690298e-05, + "loss": 1.6903, + "step": 411 + }, + { + "epoch": 0.759447004608295, + "grad_norm": 0.2769939601421356, + "learning_rate": 2.790637293036552e-05, + "loss": 1.6626, + "step": 412 + }, + { + "epoch": 0.7612903225806451, + "grad_norm": 0.2946414351463318, + "learning_rate": 2.7889929470589494e-05, + "loss": 1.6489, + "step": 413 + }, + { + "epoch": 0.7631336405529954, + "grad_norm": 0.27718386054039, + "learning_rate": 2.7873426573185777e-05, + "loss": 1.664, + "step": 414 + }, + { + "epoch": 0.7649769585253456, + "grad_norm": 0.2768406271934509, + "learning_rate": 2.7856864314251994e-05, + "loss": 1.6475, + "step": 415 + }, + { + "epoch": 0.7668202764976959, + "grad_norm": 0.2640882134437561, + "learning_rate": 2.78402427701595e-05, + "loss": 1.6332, + "step": 416 + }, + { + "epoch": 0.7686635944700461, + "grad_norm": 0.26694199442863464, + "learning_rate": 2.782356201755303e-05, + "loss": 1.6633, + "step": 417 + }, + { + "epoch": 0.7705069124423963, + "grad_norm": 0.26702558994293213, + "learning_rate": 2.780682213335033e-05, + "loss": 1.6281, + "step": 418 + }, + { + "epoch": 0.7723502304147466, + "grad_norm": 0.2785816490650177, + "learning_rate": 2.7790023194741812e-05, + "loss": 1.6733, + "step": 419 + }, + { + "epoch": 0.7741935483870968, + "grad_norm": 0.27653270959854126, + "learning_rate": 2.7773165279190206e-05, + "loss": 1.6269, + "step": 420 + }, + { + "epoch": 0.776036866359447, + "grad_norm": 0.27750319242477417, + "learning_rate": 2.7756248464430186e-05, + "loss": 1.6292, + "step": 421 + }, + { + "epoch": 0.7778801843317973, + "grad_norm": 0.2917342185974121, + "learning_rate": 2.7739272828468022e-05, + "loss": 1.6159, + "step": 422 + }, + { + "epoch": 0.7797235023041474, + "grad_norm": 0.26250651478767395, + "learning_rate": 2.7722238449581227e-05, + "loss": 1.6183, + "step": 423 + }, + { + "epoch": 0.7815668202764977, + "grad_norm": 0.28476646542549133, + "learning_rate": 2.7705145406318167e-05, + "loss": 1.7191, + "step": 424 + }, + { + "epoch": 0.783410138248848, + "grad_norm": 0.2587452232837677, + "learning_rate": 2.7687993777497747e-05, + "loss": 1.5733, + "step": 425 + }, + { + "epoch": 0.7852534562211981, + "grad_norm": 0.2674744427204132, + "learning_rate": 2.7670783642208996e-05, + "loss": 1.6225, + "step": 426 + }, + { + "epoch": 0.7870967741935484, + "grad_norm": 0.2689223289489746, + "learning_rate": 2.7653515079810744e-05, + "loss": 1.6964, + "step": 427 + }, + { + "epoch": 0.7889400921658987, + "grad_norm": 0.26800140738487244, + "learning_rate": 2.7636188169931217e-05, + "loss": 1.6728, + "step": 428 + }, + { + "epoch": 0.7907834101382488, + "grad_norm": 0.27116596698760986, + "learning_rate": 2.7618802992467718e-05, + "loss": 1.6971, + "step": 429 + }, + { + "epoch": 0.7926267281105991, + "grad_norm": 0.26951897144317627, + "learning_rate": 2.760135962758621e-05, + "loss": 1.6763, + "step": 430 + }, + { + "epoch": 0.7944700460829494, + "grad_norm": 0.26193779706954956, + "learning_rate": 2.7583858155720977e-05, + "loss": 1.6291, + "step": 431 + }, + { + "epoch": 0.7963133640552995, + "grad_norm": 0.27084240317344666, + "learning_rate": 2.756629865757424e-05, + "loss": 1.6718, + "step": 432 + }, + { + "epoch": 0.7981566820276498, + "grad_norm": 0.2781943380832672, + "learning_rate": 2.7548681214115798e-05, + "loss": 1.5723, + "step": 433 + }, + { + "epoch": 0.8, + "grad_norm": 0.27311035990715027, + "learning_rate": 2.7531005906582628e-05, + "loss": 1.6833, + "step": 434 + }, + { + "epoch": 0.8018433179723502, + "grad_norm": 0.27680864930152893, + "learning_rate": 2.7513272816478554e-05, + "loss": 1.6166, + "step": 435 + }, + { + "epoch": 0.8036866359447005, + "grad_norm": 0.26824140548706055, + "learning_rate": 2.7495482025573817e-05, + "loss": 1.6716, + "step": 436 + }, + { + "epoch": 0.8055299539170507, + "grad_norm": 0.27081307768821716, + "learning_rate": 2.7477633615904744e-05, + "loss": 1.6573, + "step": 437 + }, + { + "epoch": 0.8073732718894009, + "grad_norm": 0.2916286587715149, + "learning_rate": 2.7459727669773344e-05, + "loss": 1.6609, + "step": 438 + }, + { + "epoch": 0.8092165898617512, + "grad_norm": 0.2842768132686615, + "learning_rate": 2.7441764269746946e-05, + "loss": 1.6401, + "step": 439 + }, + { + "epoch": 0.8110599078341014, + "grad_norm": 0.2689710557460785, + "learning_rate": 2.7423743498657794e-05, + "loss": 1.6313, + "step": 440 + }, + { + "epoch": 0.8129032258064516, + "grad_norm": 0.28198468685150146, + "learning_rate": 2.7405665439602695e-05, + "loss": 1.729, + "step": 441 + }, + { + "epoch": 0.8147465437788018, + "grad_norm": 0.28444287180900574, + "learning_rate": 2.7387530175942604e-05, + "loss": 1.6436, + "step": 442 + }, + { + "epoch": 0.8165898617511521, + "grad_norm": 0.289588987827301, + "learning_rate": 2.7369337791302272e-05, + "loss": 1.6492, + "step": 443 + }, + { + "epoch": 0.8184331797235023, + "grad_norm": 0.2716604173183441, + "learning_rate": 2.7351088369569833e-05, + "loss": 1.6683, + "step": 444 + }, + { + "epoch": 0.8202764976958525, + "grad_norm": 0.2785453796386719, + "learning_rate": 2.7332781994896438e-05, + "loss": 1.6595, + "step": 445 + }, + { + "epoch": 0.8221198156682028, + "grad_norm": 0.26734933257102966, + "learning_rate": 2.7314418751695845e-05, + "loss": 1.6208, + "step": 446 + }, + { + "epoch": 0.823963133640553, + "grad_norm": 0.27588117122650146, + "learning_rate": 2.7295998724644058e-05, + "loss": 1.6085, + "step": 447 + }, + { + "epoch": 0.8258064516129032, + "grad_norm": 0.26697295904159546, + "learning_rate": 2.7277521998678904e-05, + "loss": 1.6348, + "step": 448 + }, + { + "epoch": 0.8276497695852535, + "grad_norm": 0.27423718571662903, + "learning_rate": 2.725898865899967e-05, + "loss": 1.6787, + "step": 449 + }, + { + "epoch": 0.8294930875576036, + "grad_norm": 0.26683908700942993, + "learning_rate": 2.72403987910667e-05, + "loss": 1.6271, + "step": 450 + }, + { + "epoch": 0.8313364055299539, + "grad_norm": 0.26200321316719055, + "learning_rate": 2.722175248060099e-05, + "loss": 1.6035, + "step": 451 + }, + { + "epoch": 0.8331797235023042, + "grad_norm": 0.2699339985847473, + "learning_rate": 2.7203049813583803e-05, + "loss": 1.5928, + "step": 452 + }, + { + "epoch": 0.8350230414746543, + "grad_norm": 0.27287527918815613, + "learning_rate": 2.7184290876256278e-05, + "loss": 1.6073, + "step": 453 + }, + { + "epoch": 0.8368663594470046, + "grad_norm": 0.2751379907131195, + "learning_rate": 2.716547575511903e-05, + "loss": 1.6385, + "step": 454 + }, + { + "epoch": 0.8387096774193549, + "grad_norm": 0.2756018340587616, + "learning_rate": 2.714660453693173e-05, + "loss": 1.6921, + "step": 455 + }, + { + "epoch": 0.840552995391705, + "grad_norm": 0.28198951482772827, + "learning_rate": 2.7127677308712733e-05, + "loss": 1.6651, + "step": 456 + }, + { + "epoch": 0.8423963133640553, + "grad_norm": 0.28402063250541687, + "learning_rate": 2.710869415773867e-05, + "loss": 1.5813, + "step": 457 + }, + { + "epoch": 0.8442396313364056, + "grad_norm": 0.29829660058021545, + "learning_rate": 2.7089655171544026e-05, + "loss": 1.6971, + "step": 458 + }, + { + "epoch": 0.8460829493087557, + "grad_norm": 0.2694368362426758, + "learning_rate": 2.707056043792077e-05, + "loss": 1.6268, + "step": 459 + }, + { + "epoch": 0.847926267281106, + "grad_norm": 0.2761029303073883, + "learning_rate": 2.705141004491792e-05, + "loss": 1.6883, + "step": 460 + }, + { + "epoch": 0.8497695852534562, + "grad_norm": 0.280799001455307, + "learning_rate": 2.703220408084115e-05, + "loss": 1.6409, + "step": 461 + }, + { + "epoch": 0.8516129032258064, + "grad_norm": 0.2578011453151703, + "learning_rate": 2.7012942634252384e-05, + "loss": 1.5454, + "step": 462 + }, + { + "epoch": 0.8534562211981567, + "grad_norm": 0.30007144808769226, + "learning_rate": 2.6993625793969383e-05, + "loss": 1.6845, + "step": 463 + }, + { + "epoch": 0.8552995391705069, + "grad_norm": 0.26995283365249634, + "learning_rate": 2.697425364906534e-05, + "loss": 1.6339, + "step": 464 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.3060062527656555, + "learning_rate": 2.6954826288868463e-05, + "loss": 1.6226, + "step": 465 + }, + { + "epoch": 0.8589861751152074, + "grad_norm": 0.27042827010154724, + "learning_rate": 2.693534380296158e-05, + "loss": 1.5902, + "step": 466 + }, + { + "epoch": 0.8608294930875576, + "grad_norm": 0.2701798975467682, + "learning_rate": 2.6915806281181688e-05, + "loss": 1.6444, + "step": 467 + }, + { + "epoch": 0.8626728110599078, + "grad_norm": 0.2839266061782837, + "learning_rate": 2.6896213813619592e-05, + "loss": 1.631, + "step": 468 + }, + { + "epoch": 0.864516129032258, + "grad_norm": 0.2714848816394806, + "learning_rate": 2.6876566490619437e-05, + "loss": 1.5984, + "step": 469 + }, + { + "epoch": 0.8663594470046083, + "grad_norm": 0.26444998383522034, + "learning_rate": 2.685686440277833e-05, + "loss": 1.6318, + "step": 470 + }, + { + "epoch": 0.8682027649769585, + "grad_norm": 0.28271374106407166, + "learning_rate": 2.6837107640945904e-05, + "loss": 1.6931, + "step": 471 + }, + { + "epoch": 0.8700460829493087, + "grad_norm": 0.26922810077667236, + "learning_rate": 2.681729629622391e-05, + "loss": 1.5986, + "step": 472 + }, + { + "epoch": 0.871889400921659, + "grad_norm": 0.2678123712539673, + "learning_rate": 2.6797430459965766e-05, + "loss": 1.6511, + "step": 473 + }, + { + "epoch": 0.8737327188940092, + "grad_norm": 0.2775745093822479, + "learning_rate": 2.6777510223776187e-05, + "loss": 1.6248, + "step": 474 + }, + { + "epoch": 0.8755760368663594, + "grad_norm": 0.2708311378955841, + "learning_rate": 2.6757535679510727e-05, + "loss": 1.6032, + "step": 475 + }, + { + "epoch": 0.8774193548387097, + "grad_norm": 0.27130743861198425, + "learning_rate": 2.6737506919275363e-05, + "loss": 1.6658, + "step": 476 + }, + { + "epoch": 0.8792626728110599, + "grad_norm": 0.32221710681915283, + "learning_rate": 2.6717424035426054e-05, + "loss": 1.6324, + "step": 477 + }, + { + "epoch": 0.8811059907834101, + "grad_norm": 0.27446186542510986, + "learning_rate": 2.6697287120568364e-05, + "loss": 1.6608, + "step": 478 + }, + { + "epoch": 0.8829493087557604, + "grad_norm": 0.2836889922618866, + "learning_rate": 2.6677096267556984e-05, + "loss": 1.5962, + "step": 479 + }, + { + "epoch": 0.8847926267281107, + "grad_norm": 0.29300132393836975, + "learning_rate": 2.6656851569495316e-05, + "loss": 1.6496, + "step": 480 + }, + { + "epoch": 0.8866359447004608, + "grad_norm": 0.2937772572040558, + "learning_rate": 2.6636553119735066e-05, + "loss": 1.6164, + "step": 481 + }, + { + "epoch": 0.8884792626728111, + "grad_norm": 0.30279168486595154, + "learning_rate": 2.6616201011875792e-05, + "loss": 1.6317, + "step": 482 + }, + { + "epoch": 0.8903225806451613, + "grad_norm": 0.2589039206504822, + "learning_rate": 2.6595795339764478e-05, + "loss": 1.6076, + "step": 483 + }, + { + "epoch": 0.8921658986175115, + "grad_norm": 0.29674026370048523, + "learning_rate": 2.6575336197495098e-05, + "loss": 1.6106, + "step": 484 + }, + { + "epoch": 0.8940092165898618, + "grad_norm": 0.2732203006744385, + "learning_rate": 2.6554823679408195e-05, + "loss": 1.6597, + "step": 485 + }, + { + "epoch": 0.895852534562212, + "grad_norm": 0.26109176874160767, + "learning_rate": 2.653425788009043e-05, + "loss": 1.5526, + "step": 486 + }, + { + "epoch": 0.8976958525345622, + "grad_norm": 0.3179691433906555, + "learning_rate": 2.6513638894374158e-05, + "loss": 1.687, + "step": 487 + }, + { + "epoch": 0.8995391705069125, + "grad_norm": 0.26771122217178345, + "learning_rate": 2.6492966817336977e-05, + "loss": 1.619, + "step": 488 + }, + { + "epoch": 0.9013824884792627, + "grad_norm": 0.2821449637413025, + "learning_rate": 2.6472241744301304e-05, + "loss": 1.5945, + "step": 489 + }, + { + "epoch": 0.9032258064516129, + "grad_norm": 0.28389427065849304, + "learning_rate": 2.645146377083393e-05, + "loss": 1.5934, + "step": 490 + }, + { + "epoch": 0.9050691244239631, + "grad_norm": 0.25905099511146545, + "learning_rate": 2.6430632992745577e-05, + "loss": 1.6376, + "step": 491 + }, + { + "epoch": 0.9069124423963134, + "grad_norm": 0.2909289598464966, + "learning_rate": 2.6409749506090456e-05, + "loss": 1.6398, + "step": 492 + }, + { + "epoch": 0.9087557603686636, + "grad_norm": 0.30152249336242676, + "learning_rate": 2.638881340716583e-05, + "loss": 1.7085, + "step": 493 + }, + { + "epoch": 0.9105990783410138, + "grad_norm": 0.2955312430858612, + "learning_rate": 2.6367824792511565e-05, + "loss": 1.663, + "step": 494 + }, + { + "epoch": 0.9124423963133641, + "grad_norm": 0.3120115101337433, + "learning_rate": 2.6346783758909683e-05, + "loss": 1.6809, + "step": 495 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.26377126574516296, + "learning_rate": 2.632569040338392e-05, + "loss": 1.6235, + "step": 496 + }, + { + "epoch": 0.9161290322580645, + "grad_norm": 0.2728709876537323, + "learning_rate": 2.6304544823199282e-05, + "loss": 1.6263, + "step": 497 + }, + { + "epoch": 0.9179723502304148, + "grad_norm": 0.28993189334869385, + "learning_rate": 2.6283347115861586e-05, + "loss": 1.6395, + "step": 498 + }, + { + "epoch": 0.919815668202765, + "grad_norm": 0.27762892842292786, + "learning_rate": 2.6262097379117015e-05, + "loss": 1.6613, + "step": 499 + }, + { + "epoch": 0.9216589861751152, + "grad_norm": 0.26995396614074707, + "learning_rate": 2.624079571095167e-05, + "loss": 1.6483, + "step": 500 + }, + { + "epoch": 0.9235023041474655, + "grad_norm": 0.27732783555984497, + "learning_rate": 2.6219442209591123e-05, + "loss": 1.6918, + "step": 501 + }, + { + "epoch": 0.9253456221198156, + "grad_norm": 0.28199324011802673, + "learning_rate": 2.619803697349994e-05, + "loss": 1.6222, + "step": 502 + }, + { + "epoch": 0.9271889400921659, + "grad_norm": 0.28517088294029236, + "learning_rate": 2.6176580101381273e-05, + "loss": 1.6017, + "step": 503 + }, + { + "epoch": 0.9290322580645162, + "grad_norm": 0.2852960526943207, + "learning_rate": 2.6155071692176348e-05, + "loss": 1.6117, + "step": 504 + }, + { + "epoch": 0.9308755760368663, + "grad_norm": 0.2706369459629059, + "learning_rate": 2.613351184506405e-05, + "loss": 1.6483, + "step": 505 + }, + { + "epoch": 0.9327188940092166, + "grad_norm": 0.31629517674446106, + "learning_rate": 2.6111900659460455e-05, + "loss": 1.63, + "step": 506 + }, + { + "epoch": 0.9345622119815669, + "grad_norm": 0.2743918001651764, + "learning_rate": 2.6090238235018365e-05, + "loss": 1.617, + "step": 507 + }, + { + "epoch": 0.936405529953917, + "grad_norm": 0.26636356115341187, + "learning_rate": 2.6068524671626856e-05, + "loss": 1.6671, + "step": 508 + }, + { + "epoch": 0.9382488479262673, + "grad_norm": 0.2727503478527069, + "learning_rate": 2.6046760069410806e-05, + "loss": 1.6101, + "step": 509 + }, + { + "epoch": 0.9400921658986175, + "grad_norm": 0.27820461988449097, + "learning_rate": 2.6024944528730453e-05, + "loss": 1.5903, + "step": 510 + }, + { + "epoch": 0.9419354838709677, + "grad_norm": 0.2720506191253662, + "learning_rate": 2.6003078150180922e-05, + "loss": 1.6722, + "step": 511 + }, + { + "epoch": 0.943778801843318, + "grad_norm": 0.2729189097881317, + "learning_rate": 2.598116103459174e-05, + "loss": 1.6232, + "step": 512 + }, + { + "epoch": 0.9456221198156682, + "grad_norm": 0.2625363767147064, + "learning_rate": 2.595919328302641e-05, + "loss": 1.5969, + "step": 513 + }, + { + "epoch": 0.9474654377880184, + "grad_norm": 0.26478803157806396, + "learning_rate": 2.5937174996781927e-05, + "loss": 1.5817, + "step": 514 + }, + { + "epoch": 0.9493087557603687, + "grad_norm": 0.28460946679115295, + "learning_rate": 2.5915106277388293e-05, + "loss": 1.5845, + "step": 515 + }, + { + "epoch": 0.9511520737327189, + "grad_norm": 0.2615947127342224, + "learning_rate": 2.5892987226608082e-05, + "loss": 1.6227, + "step": 516 + }, + { + "epoch": 0.9529953917050691, + "grad_norm": 0.25825098156929016, + "learning_rate": 2.5870817946435953e-05, + "loss": 1.5853, + "step": 517 + }, + { + "epoch": 0.9548387096774194, + "grad_norm": 0.2917359471321106, + "learning_rate": 2.5848598539098164e-05, + "loss": 1.6514, + "step": 518 + }, + { + "epoch": 0.9566820276497696, + "grad_norm": 0.25932732224464417, + "learning_rate": 2.5826329107052144e-05, + "loss": 1.603, + "step": 519 + }, + { + "epoch": 0.9585253456221198, + "grad_norm": 0.25399070978164673, + "learning_rate": 2.5804009752985975e-05, + "loss": 1.6073, + "step": 520 + }, + { + "epoch": 0.96036866359447, + "grad_norm": 0.28060877323150635, + "learning_rate": 2.5781640579817946e-05, + "loss": 1.6337, + "step": 521 + }, + { + "epoch": 0.9622119815668203, + "grad_norm": 0.28021928668022156, + "learning_rate": 2.5759221690696062e-05, + "loss": 1.6345, + "step": 522 + }, + { + "epoch": 0.9640552995391705, + "grad_norm": 0.27043914794921875, + "learning_rate": 2.573675318899759e-05, + "loss": 1.6471, + "step": 523 + }, + { + "epoch": 0.9658986175115207, + "grad_norm": 0.2874245345592499, + "learning_rate": 2.5714235178328554e-05, + "loss": 1.6632, + "step": 524 + }, + { + "epoch": 0.967741935483871, + "grad_norm": 0.2559823989868164, + "learning_rate": 2.5691667762523284e-05, + "loss": 1.6133, + "step": 525 + }, + { + "epoch": 0.9695852534562212, + "grad_norm": 0.2857271432876587, + "learning_rate": 2.566905104564393e-05, + "loss": 1.628, + "step": 526 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 0.27135902643203735, + "learning_rate": 2.564638513197995e-05, + "loss": 1.5717, + "step": 527 + }, + { + "epoch": 0.9732718894009217, + "grad_norm": 0.25891175866127014, + "learning_rate": 2.562367012604769e-05, + "loss": 1.6104, + "step": 528 + }, + { + "epoch": 0.9751152073732718, + "grad_norm": 0.27881482243537903, + "learning_rate": 2.5600906132589846e-05, + "loss": 1.6325, + "step": 529 + }, + { + "epoch": 0.9769585253456221, + "grad_norm": 0.2634258270263672, + "learning_rate": 2.557809325657501e-05, + "loss": 1.605, + "step": 530 + }, + { + "epoch": 0.9788018433179724, + "grad_norm": 0.2687437832355499, + "learning_rate": 2.555523160319719e-05, + "loss": 1.6302, + "step": 531 + }, + { + "epoch": 0.9806451612903225, + "grad_norm": 0.27842116355895996, + "learning_rate": 2.5532321277875305e-05, + "loss": 1.6813, + "step": 532 + }, + { + "epoch": 0.9824884792626728, + "grad_norm": 0.26950204372406006, + "learning_rate": 2.5509362386252702e-05, + "loss": 1.6166, + "step": 533 + }, + { + "epoch": 0.9843317972350231, + "grad_norm": 0.2951159179210663, + "learning_rate": 2.5486355034196686e-05, + "loss": 1.669, + "step": 534 + }, + { + "epoch": 0.9861751152073732, + "grad_norm": 0.2618483603000641, + "learning_rate": 2.5463299327798015e-05, + "loss": 1.6714, + "step": 535 + }, + { + "epoch": 0.9880184331797235, + "grad_norm": 0.25726941227912903, + "learning_rate": 2.544019537337043e-05, + "loss": 1.6314, + "step": 536 + }, + { + "epoch": 0.9898617511520738, + "grad_norm": 0.2638774812221527, + "learning_rate": 2.541704327745013e-05, + "loss": 1.6458, + "step": 537 + }, + { + "epoch": 0.9917050691244239, + "grad_norm": 0.2752140760421753, + "learning_rate": 2.539384314679532e-05, + "loss": 1.6564, + "step": 538 + }, + { + "epoch": 0.9935483870967742, + "grad_norm": 0.27124327421188354, + "learning_rate": 2.5370595088385696e-05, + "loss": 1.6071, + "step": 539 + }, + { + "epoch": 0.9953917050691244, + "grad_norm": 0.26434096693992615, + "learning_rate": 2.5347299209421955e-05, + "loss": 1.611, + "step": 540 + }, + { + "epoch": 0.9972350230414746, + "grad_norm": 0.2855331003665924, + "learning_rate": 2.53239556173253e-05, + "loss": 1.6311, + "step": 541 + }, + { + "epoch": 0.9990783410138249, + "grad_norm": 0.2693633437156677, + "learning_rate": 2.530056441973696e-05, + "loss": 1.5659, + "step": 542 + }, + { + "epoch": 1.0009216589861751, + "grad_norm": 0.26715287566185, + "learning_rate": 2.5277125724517665e-05, + "loss": 1.6523, + "step": 543 + }, + { + "epoch": 1.0027649769585254, + "grad_norm": 0.30740463733673096, + "learning_rate": 2.525363963974717e-05, + "loss": 1.655, + "step": 544 + }, + { + "epoch": 1.0046082949308757, + "grad_norm": 0.2541782557964325, + "learning_rate": 2.523010627372376e-05, + "loss": 1.5368, + "step": 545 + }, + { + "epoch": 1.0064516129032257, + "grad_norm": 0.27700361609458923, + "learning_rate": 2.520652573496373e-05, + "loss": 1.6131, + "step": 546 + }, + { + "epoch": 1.008294930875576, + "grad_norm": 0.2821163237094879, + "learning_rate": 2.51828981322009e-05, + "loss": 1.6299, + "step": 547 + }, + { + "epoch": 1.0101382488479262, + "grad_norm": 0.26375892758369446, + "learning_rate": 2.5159223574386117e-05, + "loss": 1.6282, + "step": 548 + }, + { + "epoch": 1.0119815668202765, + "grad_norm": 0.26267147064208984, + "learning_rate": 2.513550217068673e-05, + "loss": 1.6306, + "step": 549 + }, + { + "epoch": 1.0138248847926268, + "grad_norm": 0.2676134407520294, + "learning_rate": 2.5111734030486127e-05, + "loss": 1.6022, + "step": 550 + }, + { + "epoch": 1.015668202764977, + "grad_norm": 0.27351808547973633, + "learning_rate": 2.508791926338317e-05, + "loss": 1.6113, + "step": 551 + }, + { + "epoch": 1.017511520737327, + "grad_norm": 0.2688004970550537, + "learning_rate": 2.5064057979191766e-05, + "loss": 1.6101, + "step": 552 + }, + { + "epoch": 1.0193548387096774, + "grad_norm": 0.26147159934043884, + "learning_rate": 2.5040150287940286e-05, + "loss": 1.611, + "step": 553 + }, + { + "epoch": 1.0211981566820276, + "grad_norm": 0.272300660610199, + "learning_rate": 2.5016196299871115e-05, + "loss": 1.6068, + "step": 554 + }, + { + "epoch": 1.023041474654378, + "grad_norm": 0.26897957921028137, + "learning_rate": 2.49921961254401e-05, + "loss": 1.6466, + "step": 555 + }, + { + "epoch": 1.0248847926267282, + "grad_norm": 0.26811644434928894, + "learning_rate": 2.496814987531609e-05, + "loss": 1.651, + "step": 556 + }, + { + "epoch": 1.0267281105990784, + "grad_norm": 0.265045702457428, + "learning_rate": 2.4944057660380363e-05, + "loss": 1.6455, + "step": 557 + }, + { + "epoch": 1.0285714285714285, + "grad_norm": 0.2766599953174591, + "learning_rate": 2.4919919591726175e-05, + "loss": 1.6231, + "step": 558 + }, + { + "epoch": 1.0304147465437787, + "grad_norm": 0.27361610531806946, + "learning_rate": 2.489573578065821e-05, + "loss": 1.6258, + "step": 559 + }, + { + "epoch": 1.032258064516129, + "grad_norm": 0.26939770579338074, + "learning_rate": 2.487150633869207e-05, + "loss": 1.5856, + "step": 560 + }, + { + "epoch": 1.0341013824884793, + "grad_norm": 0.28874027729034424, + "learning_rate": 2.484723137755379e-05, + "loss": 1.5899, + "step": 561 + }, + { + "epoch": 1.0359447004608295, + "grad_norm": 0.2698168158531189, + "learning_rate": 2.482291100917928e-05, + "loss": 1.7224, + "step": 562 + }, + { + "epoch": 1.0377880184331798, + "grad_norm": 0.26175767183303833, + "learning_rate": 2.4798545345713837e-05, + "loss": 1.6187, + "step": 563 + }, + { + "epoch": 1.0396313364055298, + "grad_norm": 0.27548477053642273, + "learning_rate": 2.4774134499511636e-05, + "loss": 1.7049, + "step": 564 + }, + { + "epoch": 1.0414746543778801, + "grad_norm": 0.257304847240448, + "learning_rate": 2.4749678583135175e-05, + "loss": 1.5474, + "step": 565 + }, + { + "epoch": 1.0433179723502304, + "grad_norm": 0.266632080078125, + "learning_rate": 2.472517770935479e-05, + "loss": 1.623, + "step": 566 + }, + { + "epoch": 1.0451612903225806, + "grad_norm": 0.2716248631477356, + "learning_rate": 2.4700631991148126e-05, + "loss": 1.5814, + "step": 567 + }, + { + "epoch": 1.047004608294931, + "grad_norm": 0.2753863036632538, + "learning_rate": 2.46760415416996e-05, + "loss": 1.644, + "step": 568 + }, + { + "epoch": 1.0488479262672812, + "grad_norm": 0.2726069390773773, + "learning_rate": 2.465140647439991e-05, + "loss": 1.6133, + "step": 569 + }, + { + "epoch": 1.0506912442396312, + "grad_norm": 0.28230923414230347, + "learning_rate": 2.4626726902845477e-05, + "loss": 1.6963, + "step": 570 + }, + { + "epoch": 1.0525345622119815, + "grad_norm": 0.25983119010925293, + "learning_rate": 2.4602002940837948e-05, + "loss": 1.5626, + "step": 571 + }, + { + "epoch": 1.0543778801843318, + "grad_norm": 0.2676817774772644, + "learning_rate": 2.4577234702383666e-05, + "loss": 1.5422, + "step": 572 + }, + { + "epoch": 1.056221198156682, + "grad_norm": 0.26019108295440674, + "learning_rate": 2.4552422301693128e-05, + "loss": 1.5826, + "step": 573 + }, + { + "epoch": 1.0580645161290323, + "grad_norm": 0.26668256521224976, + "learning_rate": 2.452756585318048e-05, + "loss": 1.596, + "step": 574 + }, + { + "epoch": 1.0599078341013826, + "grad_norm": 0.28593432903289795, + "learning_rate": 2.4502665471462983e-05, + "loss": 1.6028, + "step": 575 + }, + { + "epoch": 1.0617511520737328, + "grad_norm": 0.2791599929332733, + "learning_rate": 2.447772127136046e-05, + "loss": 1.5927, + "step": 576 + }, + { + "epoch": 1.0635944700460829, + "grad_norm": 0.28675881028175354, + "learning_rate": 2.4452733367894816e-05, + "loss": 1.5879, + "step": 577 + }, + { + "epoch": 1.0654377880184331, + "grad_norm": 0.29501160979270935, + "learning_rate": 2.4427701876289465e-05, + "loss": 1.5583, + "step": 578 + }, + { + "epoch": 1.0672811059907834, + "grad_norm": 0.2674134075641632, + "learning_rate": 2.440262691196881e-05, + "loss": 1.6205, + "step": 579 + }, + { + "epoch": 1.0691244239631337, + "grad_norm": 0.32356998324394226, + "learning_rate": 2.437750859055773e-05, + "loss": 1.6112, + "step": 580 + }, + { + "epoch": 1.070967741935484, + "grad_norm": 0.2775920629501343, + "learning_rate": 2.4352347027881003e-05, + "loss": 1.6036, + "step": 581 + }, + { + "epoch": 1.072811059907834, + "grad_norm": 0.28417059779167175, + "learning_rate": 2.4327142339962827e-05, + "loss": 1.6073, + "step": 582 + }, + { + "epoch": 1.0746543778801843, + "grad_norm": 0.316342294216156, + "learning_rate": 2.430189464302625e-05, + "loss": 1.6312, + "step": 583 + }, + { + "epoch": 1.0764976958525345, + "grad_norm": 0.2634347081184387, + "learning_rate": 2.4276604053492636e-05, + "loss": 1.6042, + "step": 584 + }, + { + "epoch": 1.0783410138248848, + "grad_norm": 0.2889562249183655, + "learning_rate": 2.425127068798113e-05, + "loss": 1.586, + "step": 585 + }, + { + "epoch": 1.080184331797235, + "grad_norm": 0.2724316716194153, + "learning_rate": 2.422589466330814e-05, + "loss": 1.6629, + "step": 586 + }, + { + "epoch": 1.0820276497695853, + "grad_norm": 0.263497531414032, + "learning_rate": 2.4200476096486774e-05, + "loss": 1.5843, + "step": 587 + }, + { + "epoch": 1.0838709677419356, + "grad_norm": 0.27481377124786377, + "learning_rate": 2.4175015104726306e-05, + "loss": 1.6378, + "step": 588 + }, + { + "epoch": 1.0857142857142856, + "grad_norm": 0.28347697854042053, + "learning_rate": 2.414951180543164e-05, + "loss": 1.7082, + "step": 589 + }, + { + "epoch": 1.087557603686636, + "grad_norm": 0.2818866968154907, + "learning_rate": 2.4123966316202768e-05, + "loss": 1.5482, + "step": 590 + }, + { + "epoch": 1.0894009216589862, + "grad_norm": 0.26917752623558044, + "learning_rate": 2.4098378754834227e-05, + "loss": 1.6042, + "step": 591 + }, + { + "epoch": 1.0912442396313364, + "grad_norm": 0.2925183176994324, + "learning_rate": 2.4072749239314565e-05, + "loss": 1.5839, + "step": 592 + }, + { + "epoch": 1.0930875576036867, + "grad_norm": 0.2812125086784363, + "learning_rate": 2.4047077887825765e-05, + "loss": 1.5705, + "step": 593 + }, + { + "epoch": 1.094930875576037, + "grad_norm": 0.2660687565803528, + "learning_rate": 2.402136481874275e-05, + "loss": 1.6325, + "step": 594 + }, + { + "epoch": 1.096774193548387, + "grad_norm": 0.308992862701416, + "learning_rate": 2.399561015063278e-05, + "loss": 1.5755, + "step": 595 + }, + { + "epoch": 1.0986175115207373, + "grad_norm": 0.2750917971134186, + "learning_rate": 2.3969814002254965e-05, + "loss": 1.6258, + "step": 596 + }, + { + "epoch": 1.1004608294930875, + "grad_norm": 0.277424156665802, + "learning_rate": 2.3943976492559675e-05, + "loss": 1.6046, + "step": 597 + }, + { + "epoch": 1.1023041474654378, + "grad_norm": 0.2793235182762146, + "learning_rate": 2.3918097740687987e-05, + "loss": 1.6198, + "step": 598 + }, + { + "epoch": 1.104147465437788, + "grad_norm": 0.30500084161758423, + "learning_rate": 2.3892177865971183e-05, + "loss": 1.5345, + "step": 599 + }, + { + "epoch": 1.1059907834101383, + "grad_norm": 0.2782265841960907, + "learning_rate": 2.386621698793015e-05, + "loss": 1.6041, + "step": 600 + }, + { + "epoch": 1.1078341013824884, + "grad_norm": 0.2845817506313324, + "learning_rate": 2.3840215226274847e-05, + "loss": 1.5975, + "step": 601 + }, + { + "epoch": 1.1096774193548387, + "grad_norm": 0.31969916820526123, + "learning_rate": 2.3814172700903775e-05, + "loss": 1.6021, + "step": 602 + }, + { + "epoch": 1.111520737327189, + "grad_norm": 0.26726001501083374, + "learning_rate": 2.3788089531903372e-05, + "loss": 1.5317, + "step": 603 + }, + { + "epoch": 1.1133640552995392, + "grad_norm": 0.2735467553138733, + "learning_rate": 2.3761965839547515e-05, + "loss": 1.5867, + "step": 604 + }, + { + "epoch": 1.1152073732718895, + "grad_norm": 0.31699496507644653, + "learning_rate": 2.3735801744296934e-05, + "loss": 1.6256, + "step": 605 + }, + { + "epoch": 1.1170506912442397, + "grad_norm": 0.27312713861465454, + "learning_rate": 2.3709597366798662e-05, + "loss": 1.6208, + "step": 606 + }, + { + "epoch": 1.1188940092165898, + "grad_norm": 0.2782924473285675, + "learning_rate": 2.3683352827885472e-05, + "loss": 1.6535, + "step": 607 + }, + { + "epoch": 1.12073732718894, + "grad_norm": 0.30579322576522827, + "learning_rate": 2.365706824857535e-05, + "loss": 1.606, + "step": 608 + }, + { + "epoch": 1.1225806451612903, + "grad_norm": 0.28099164366722107, + "learning_rate": 2.3630743750070892e-05, + "loss": 1.5968, + "step": 609 + }, + { + "epoch": 1.1244239631336406, + "grad_norm": 0.27450433373451233, + "learning_rate": 2.360437945375878e-05, + "loss": 1.6303, + "step": 610 + }, + { + "epoch": 1.1262672811059908, + "grad_norm": 0.27543413639068604, + "learning_rate": 2.3577975481209214e-05, + "loss": 1.6004, + "step": 611 + }, + { + "epoch": 1.128110599078341, + "grad_norm": 0.27525603771209717, + "learning_rate": 2.3551531954175335e-05, + "loss": 1.6507, + "step": 612 + }, + { + "epoch": 1.1299539170506911, + "grad_norm": 0.26268866658210754, + "learning_rate": 2.3525048994592684e-05, + "loss": 1.5314, + "step": 613 + }, + { + "epoch": 1.1317972350230414, + "grad_norm": 0.2847149968147278, + "learning_rate": 2.3498526724578637e-05, + "loss": 1.5997, + "step": 614 + }, + { + "epoch": 1.1336405529953917, + "grad_norm": 0.2706824243068695, + "learning_rate": 2.3471965266431824e-05, + "loss": 1.6192, + "step": 615 + }, + { + "epoch": 1.135483870967742, + "grad_norm": 0.28432345390319824, + "learning_rate": 2.3445364742631592e-05, + "loss": 1.5632, + "step": 616 + }, + { + "epoch": 1.1373271889400922, + "grad_norm": 0.2760394811630249, + "learning_rate": 2.3418725275837413e-05, + "loss": 1.6104, + "step": 617 + }, + { + "epoch": 1.1391705069124425, + "grad_norm": 0.26547399163246155, + "learning_rate": 2.3392046988888345e-05, + "loss": 1.5942, + "step": 618 + }, + { + "epoch": 1.1410138248847925, + "grad_norm": 0.28383272886276245, + "learning_rate": 2.3365330004802443e-05, + "loss": 1.6284, + "step": 619 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 0.2690708339214325, + "learning_rate": 2.33385744467762e-05, + "loss": 1.5903, + "step": 620 + }, + { + "epoch": 1.144700460829493, + "grad_norm": 0.2783527970314026, + "learning_rate": 2.331178043818399e-05, + "loss": 1.6339, + "step": 621 + }, + { + "epoch": 1.1465437788018433, + "grad_norm": 0.2668425738811493, + "learning_rate": 2.328494810257748e-05, + "loss": 1.5174, + "step": 622 + }, + { + "epoch": 1.1483870967741936, + "grad_norm": 0.2701328992843628, + "learning_rate": 2.3258077563685072e-05, + "loss": 1.6126, + "step": 623 + }, + { + "epoch": 1.1502304147465439, + "grad_norm": 0.26650527119636536, + "learning_rate": 2.3231168945411326e-05, + "loss": 1.5872, + "step": 624 + }, + { + "epoch": 1.1520737327188941, + "grad_norm": 0.27082017064094543, + "learning_rate": 2.320422237183641e-05, + "loss": 1.6007, + "step": 625 + }, + { + "epoch": 1.1539170506912442, + "grad_norm": 0.28844255208969116, + "learning_rate": 2.317723796721547e-05, + "loss": 1.5988, + "step": 626 + }, + { + "epoch": 1.1557603686635944, + "grad_norm": 0.26136210560798645, + "learning_rate": 2.315021585597815e-05, + "loss": 1.5385, + "step": 627 + }, + { + "epoch": 1.1576036866359447, + "grad_norm": 0.28600579500198364, + "learning_rate": 2.3123156162727923e-05, + "loss": 1.6156, + "step": 628 + }, + { + "epoch": 1.159447004608295, + "grad_norm": 0.27295541763305664, + "learning_rate": 2.3096059012241583e-05, + "loss": 1.5353, + "step": 629 + }, + { + "epoch": 1.1612903225806452, + "grad_norm": 0.2765044867992401, + "learning_rate": 2.3068924529468638e-05, + "loss": 1.6577, + "step": 630 + }, + { + "epoch": 1.1631336405529953, + "grad_norm": 0.28675732016563416, + "learning_rate": 2.3041752839530735e-05, + "loss": 1.6112, + "step": 631 + }, + { + "epoch": 1.1649769585253456, + "grad_norm": 0.2791755795478821, + "learning_rate": 2.3014544067721096e-05, + "loss": 1.5268, + "step": 632 + }, + { + "epoch": 1.1668202764976958, + "grad_norm": 0.2844353914260864, + "learning_rate": 2.298729833950394e-05, + "loss": 1.5635, + "step": 633 + }, + { + "epoch": 1.168663594470046, + "grad_norm": 0.2774640917778015, + "learning_rate": 2.2960015780513893e-05, + "loss": 1.6243, + "step": 634 + }, + { + "epoch": 1.1705069124423964, + "grad_norm": 0.27996620535850525, + "learning_rate": 2.2932696516555396e-05, + "loss": 1.5647, + "step": 635 + }, + { + "epoch": 1.1723502304147466, + "grad_norm": 0.2877368927001953, + "learning_rate": 2.2905340673602184e-05, + "loss": 1.5705, + "step": 636 + }, + { + "epoch": 1.1741935483870969, + "grad_norm": 0.2850167751312256, + "learning_rate": 2.287794837779662e-05, + "loss": 1.6524, + "step": 637 + }, + { + "epoch": 1.176036866359447, + "grad_norm": 0.2903353273868561, + "learning_rate": 2.2850519755449183e-05, + "loss": 1.57, + "step": 638 + }, + { + "epoch": 1.1778801843317972, + "grad_norm": 0.2763937711715698, + "learning_rate": 2.282305493303785e-05, + "loss": 1.6409, + "step": 639 + }, + { + "epoch": 1.1797235023041475, + "grad_norm": 0.2943311929702759, + "learning_rate": 2.2795554037207528e-05, + "loss": 1.6925, + "step": 640 + }, + { + "epoch": 1.1815668202764977, + "grad_norm": 0.2771497070789337, + "learning_rate": 2.2768017194769466e-05, + "loss": 1.5796, + "step": 641 + }, + { + "epoch": 1.183410138248848, + "grad_norm": 0.26944899559020996, + "learning_rate": 2.2740444532700657e-05, + "loss": 1.6039, + "step": 642 + }, + { + "epoch": 1.185253456221198, + "grad_norm": 0.2843589782714844, + "learning_rate": 2.271283617814328e-05, + "loss": 1.6457, + "step": 643 + }, + { + "epoch": 1.1870967741935483, + "grad_norm": 0.2836996614933014, + "learning_rate": 2.268519225840409e-05, + "loss": 1.5728, + "step": 644 + }, + { + "epoch": 1.1889400921658986, + "grad_norm": 0.28848952054977417, + "learning_rate": 2.2657512900953832e-05, + "loss": 1.617, + "step": 645 + }, + { + "epoch": 1.1907834101382488, + "grad_norm": 0.2769070267677307, + "learning_rate": 2.2629798233426677e-05, + "loss": 1.6127, + "step": 646 + }, + { + "epoch": 1.192626728110599, + "grad_norm": 0.2685301601886749, + "learning_rate": 2.26020483836196e-05, + "loss": 1.5747, + "step": 647 + }, + { + "epoch": 1.1944700460829494, + "grad_norm": 0.2858518660068512, + "learning_rate": 2.2574263479491816e-05, + "loss": 1.6335, + "step": 648 + }, + { + "epoch": 1.1963133640552996, + "grad_norm": 0.27150848507881165, + "learning_rate": 2.2546443649164186e-05, + "loss": 1.5749, + "step": 649 + }, + { + "epoch": 1.1981566820276497, + "grad_norm": 0.27711644768714905, + "learning_rate": 2.2518589020918612e-05, + "loss": 1.6022, + "step": 650 + }, + { + "epoch": 1.2, + "grad_norm": 0.27468806505203247, + "learning_rate": 2.2490699723197454e-05, + "loss": 1.6034, + "step": 651 + }, + { + "epoch": 1.2018433179723502, + "grad_norm": 0.2741892635822296, + "learning_rate": 2.2462775884602954e-05, + "loss": 1.6301, + "step": 652 + }, + { + "epoch": 1.2036866359447005, + "grad_norm": 0.2700754702091217, + "learning_rate": 2.243481763389661e-05, + "loss": 1.5741, + "step": 653 + }, + { + "epoch": 1.2055299539170508, + "grad_norm": 0.28095367550849915, + "learning_rate": 2.24068250999986e-05, + "loss": 1.5861, + "step": 654 + }, + { + "epoch": 1.2073732718894008, + "grad_norm": 0.27654770016670227, + "learning_rate": 2.2378798411987218e-05, + "loss": 1.6016, + "step": 655 + }, + { + "epoch": 1.209216589861751, + "grad_norm": 0.27032870054244995, + "learning_rate": 2.2350737699098203e-05, + "loss": 1.6194, + "step": 656 + }, + { + "epoch": 1.2110599078341013, + "grad_norm": 0.2779473066329956, + "learning_rate": 2.2322643090724218e-05, + "loss": 1.6285, + "step": 657 + }, + { + "epoch": 1.2129032258064516, + "grad_norm": 0.2657751739025116, + "learning_rate": 2.229451471641422e-05, + "loss": 1.6217, + "step": 658 + }, + { + "epoch": 1.2147465437788019, + "grad_norm": 0.27549269795417786, + "learning_rate": 2.226635270587286e-05, + "loss": 1.5244, + "step": 659 + }, + { + "epoch": 1.2165898617511521, + "grad_norm": 0.2700861692428589, + "learning_rate": 2.2238157188959893e-05, + "loss": 1.5988, + "step": 660 + }, + { + "epoch": 1.2184331797235024, + "grad_norm": 0.27551552653312683, + "learning_rate": 2.2209928295689582e-05, + "loss": 1.6695, + "step": 661 + }, + { + "epoch": 1.2202764976958524, + "grad_norm": 0.27419742941856384, + "learning_rate": 2.2181666156230082e-05, + "loss": 1.5763, + "step": 662 + }, + { + "epoch": 1.2221198156682027, + "grad_norm": 0.26684898138046265, + "learning_rate": 2.2153370900902872e-05, + "loss": 1.5866, + "step": 663 + }, + { + "epoch": 1.223963133640553, + "grad_norm": 0.2631971836090088, + "learning_rate": 2.2125042660182115e-05, + "loss": 1.555, + "step": 664 + }, + { + "epoch": 1.2258064516129032, + "grad_norm": 0.26897555589675903, + "learning_rate": 2.2096681564694087e-05, + "loss": 1.6108, + "step": 665 + }, + { + "epoch": 1.2276497695852535, + "grad_norm": 0.2753186523914337, + "learning_rate": 2.2068287745216552e-05, + "loss": 1.6178, + "step": 666 + }, + { + "epoch": 1.2294930875576038, + "grad_norm": 0.272468626499176, + "learning_rate": 2.203986133267818e-05, + "loss": 1.5851, + "step": 667 + }, + { + "epoch": 1.2313364055299538, + "grad_norm": 0.28555935621261597, + "learning_rate": 2.2011402458157935e-05, + "loss": 1.657, + "step": 668 + }, + { + "epoch": 1.233179723502304, + "grad_norm": 0.2571600079536438, + "learning_rate": 2.198291125288445e-05, + "loss": 1.5385, + "step": 669 + }, + { + "epoch": 1.2350230414746544, + "grad_norm": 0.2795640528202057, + "learning_rate": 2.1954387848235455e-05, + "loss": 1.5856, + "step": 670 + }, + { + "epoch": 1.2368663594470046, + "grad_norm": 0.2699826657772064, + "learning_rate": 2.1925832375737168e-05, + "loss": 1.587, + "step": 671 + }, + { + "epoch": 1.238709677419355, + "grad_norm": 0.27142706513404846, + "learning_rate": 2.1897244967063653e-05, + "loss": 1.6016, + "step": 672 + }, + { + "epoch": 1.2405529953917052, + "grad_norm": 0.2868463099002838, + "learning_rate": 2.1868625754036256e-05, + "loss": 1.5758, + "step": 673 + }, + { + "epoch": 1.2423963133640552, + "grad_norm": 0.2734906077384949, + "learning_rate": 2.1839974868622956e-05, + "loss": 1.5834, + "step": 674 + }, + { + "epoch": 1.2442396313364055, + "grad_norm": 0.27085962891578674, + "learning_rate": 2.1811292442937808e-05, + "loss": 1.5689, + "step": 675 + }, + { + "epoch": 1.2460829493087557, + "grad_norm": 0.2795475125312805, + "learning_rate": 2.1782578609240286e-05, + "loss": 1.5531, + "step": 676 + }, + { + "epoch": 1.247926267281106, + "grad_norm": 0.27187928557395935, + "learning_rate": 2.1753833499934694e-05, + "loss": 1.5728, + "step": 677 + }, + { + "epoch": 1.2497695852534563, + "grad_norm": 0.2647460103034973, + "learning_rate": 2.1725057247569552e-05, + "loss": 1.5917, + "step": 678 + }, + { + "epoch": 1.2516129032258063, + "grad_norm": 0.2762637436389923, + "learning_rate": 2.1696249984836993e-05, + "loss": 1.6209, + "step": 679 + }, + { + "epoch": 1.2534562211981566, + "grad_norm": 0.27510347962379456, + "learning_rate": 2.166741184457214e-05, + "loss": 1.6489, + "step": 680 + }, + { + "epoch": 1.2552995391705069, + "grad_norm": 0.2649478316307068, + "learning_rate": 2.1638542959752485e-05, + "loss": 1.5935, + "step": 681 + }, + { + "epoch": 1.2571428571428571, + "grad_norm": 0.263662189245224, + "learning_rate": 2.160964346349731e-05, + "loss": 1.6304, + "step": 682 + }, + { + "epoch": 1.2589861751152074, + "grad_norm": 0.280752956867218, + "learning_rate": 2.1580713489067043e-05, + "loss": 1.6311, + "step": 683 + }, + { + "epoch": 1.2608294930875577, + "grad_norm": 0.2848096191883087, + "learning_rate": 2.155175316986265e-05, + "loss": 1.6682, + "step": 684 + }, + { + "epoch": 1.262672811059908, + "grad_norm": 0.2672868072986603, + "learning_rate": 2.1522762639425012e-05, + "loss": 1.5798, + "step": 685 + }, + { + "epoch": 1.2645161290322582, + "grad_norm": 0.2747519612312317, + "learning_rate": 2.1493742031434343e-05, + "loss": 1.5585, + "step": 686 + }, + { + "epoch": 1.2663594470046082, + "grad_norm": 0.27021849155426025, + "learning_rate": 2.1464691479709534e-05, + "loss": 1.5789, + "step": 687 + }, + { + "epoch": 1.2682027649769585, + "grad_norm": 0.26815730333328247, + "learning_rate": 2.1435611118207546e-05, + "loss": 1.564, + "step": 688 + }, + { + "epoch": 1.2700460829493088, + "grad_norm": 0.2694461941719055, + "learning_rate": 2.140650108102281e-05, + "loss": 1.5709, + "step": 689 + }, + { + "epoch": 1.271889400921659, + "grad_norm": 0.27626311779022217, + "learning_rate": 2.137736150238659e-05, + "loss": 1.6146, + "step": 690 + }, + { + "epoch": 1.2737327188940093, + "grad_norm": 0.2797856628894806, + "learning_rate": 2.1348192516666376e-05, + "loss": 1.6126, + "step": 691 + }, + { + "epoch": 1.2755760368663593, + "grad_norm": 0.2678052484989166, + "learning_rate": 2.1318994258365253e-05, + "loss": 1.5817, + "step": 692 + }, + { + "epoch": 1.2774193548387096, + "grad_norm": 0.2734876275062561, + "learning_rate": 2.128976686212129e-05, + "loss": 1.5634, + "step": 693 + }, + { + "epoch": 1.2792626728110599, + "grad_norm": 0.2710317373275757, + "learning_rate": 2.1260510462706914e-05, + "loss": 1.6467, + "step": 694 + }, + { + "epoch": 1.2811059907834101, + "grad_norm": 0.2857086956501007, + "learning_rate": 2.12312251950283e-05, + "loss": 1.5887, + "step": 695 + }, + { + "epoch": 1.2829493087557604, + "grad_norm": 0.26261481642723083, + "learning_rate": 2.120191119412472e-05, + "loss": 1.6167, + "step": 696 + }, + { + "epoch": 1.2847926267281107, + "grad_norm": 0.26819396018981934, + "learning_rate": 2.117256859516795e-05, + "loss": 1.5946, + "step": 697 + }, + { + "epoch": 1.286635944700461, + "grad_norm": 0.2797357439994812, + "learning_rate": 2.1143197533461655e-05, + "loss": 1.5888, + "step": 698 + }, + { + "epoch": 1.288479262672811, + "grad_norm": 0.2809047996997833, + "learning_rate": 2.1113798144440712e-05, + "loss": 1.5984, + "step": 699 + }, + { + "epoch": 1.2903225806451613, + "grad_norm": 0.2751614451408386, + "learning_rate": 2.108437056367064e-05, + "loss": 1.6601, + "step": 700 + }, + { + "epoch": 1.2921658986175115, + "grad_norm": 0.26571017503738403, + "learning_rate": 2.1054914926846957e-05, + "loss": 1.5355, + "step": 701 + }, + { + "epoch": 1.2940092165898618, + "grad_norm": 0.3031296133995056, + "learning_rate": 2.1025431369794546e-05, + "loss": 1.6608, + "step": 702 + }, + { + "epoch": 1.295852534562212, + "grad_norm": 0.28314271569252014, + "learning_rate": 2.0995920028467027e-05, + "loss": 1.6063, + "step": 703 + }, + { + "epoch": 1.297695852534562, + "grad_norm": 0.27367842197418213, + "learning_rate": 2.096638103894616e-05, + "loss": 1.5948, + "step": 704 + }, + { + "epoch": 1.2995391705069124, + "grad_norm": 0.2784302234649658, + "learning_rate": 2.0936814537441173e-05, + "loss": 1.5953, + "step": 705 + }, + { + "epoch": 1.3013824884792626, + "grad_norm": 0.2915882468223572, + "learning_rate": 2.0907220660288166e-05, + "loss": 1.5376, + "step": 706 + }, + { + "epoch": 1.303225806451613, + "grad_norm": 0.26861122250556946, + "learning_rate": 2.087759954394948e-05, + "loss": 1.6224, + "step": 707 + }, + { + "epoch": 1.3050691244239632, + "grad_norm": 0.2893337309360504, + "learning_rate": 2.084795132501304e-05, + "loss": 1.6294, + "step": 708 + }, + { + "epoch": 1.3069124423963134, + "grad_norm": 0.2968802750110626, + "learning_rate": 2.081827614019177e-05, + "loss": 1.5813, + "step": 709 + }, + { + "epoch": 1.3087557603686637, + "grad_norm": 0.2807689309120178, + "learning_rate": 2.0788574126322928e-05, + "loss": 1.596, + "step": 710 + }, + { + "epoch": 1.3105990783410137, + "grad_norm": 0.27932870388031006, + "learning_rate": 2.0758845420367474e-05, + "loss": 1.5958, + "step": 711 + }, + { + "epoch": 1.312442396313364, + "grad_norm": 0.3168034553527832, + "learning_rate": 2.0729090159409467e-05, + "loss": 1.5696, + "step": 712 + }, + { + "epoch": 1.3142857142857143, + "grad_norm": 0.2953466475009918, + "learning_rate": 2.0699308480655397e-05, + "loss": 1.5669, + "step": 713 + }, + { + "epoch": 1.3161290322580645, + "grad_norm": 0.26360318064689636, + "learning_rate": 2.06695005214336e-05, + "loss": 1.6061, + "step": 714 + }, + { + "epoch": 1.3179723502304148, + "grad_norm": 0.31883785128593445, + "learning_rate": 2.0639666419193565e-05, + "loss": 1.6457, + "step": 715 + }, + { + "epoch": 1.3198156682027649, + "grad_norm": 0.30148056149482727, + "learning_rate": 2.0609806311505345e-05, + "loss": 1.6045, + "step": 716 + }, + { + "epoch": 1.3216589861751151, + "grad_norm": 0.2783588767051697, + "learning_rate": 2.057992033605891e-05, + "loss": 1.6246, + "step": 717 + }, + { + "epoch": 1.3235023041474654, + "grad_norm": 0.2826476991176605, + "learning_rate": 2.0550008630663507e-05, + "loss": 1.6577, + "step": 718 + }, + { + "epoch": 1.3253456221198157, + "grad_norm": 0.32222914695739746, + "learning_rate": 2.0520071333247025e-05, + "loss": 1.6668, + "step": 719 + }, + { + "epoch": 1.327188940092166, + "grad_norm": 0.2784786522388458, + "learning_rate": 2.049010858185537e-05, + "loss": 1.6636, + "step": 720 + }, + { + "epoch": 1.3290322580645162, + "grad_norm": 0.27896296977996826, + "learning_rate": 2.0460120514651814e-05, + "loss": 1.5561, + "step": 721 + }, + { + "epoch": 1.3308755760368665, + "grad_norm": 0.2953389585018158, + "learning_rate": 2.0430107269916368e-05, + "loss": 1.5208, + "step": 722 + }, + { + "epoch": 1.3327188940092167, + "grad_norm": 0.27998074889183044, + "learning_rate": 2.0400068986045142e-05, + "loss": 1.6206, + "step": 723 + }, + { + "epoch": 1.3345622119815668, + "grad_norm": 0.2782033383846283, + "learning_rate": 2.03700058015497e-05, + "loss": 1.6127, + "step": 724 + }, + { + "epoch": 1.336405529953917, + "grad_norm": 0.2825208008289337, + "learning_rate": 2.0339917855056428e-05, + "loss": 1.5904, + "step": 725 + }, + { + "epoch": 1.3382488479262673, + "grad_norm": 0.2724984586238861, + "learning_rate": 2.0309805285305905e-05, + "loss": 1.5929, + "step": 726 + }, + { + "epoch": 1.3400921658986176, + "grad_norm": 0.2638327479362488, + "learning_rate": 2.0279668231152233e-05, + "loss": 1.5806, + "step": 727 + }, + { + "epoch": 1.3419354838709676, + "grad_norm": 0.27951404452323914, + "learning_rate": 2.024950683156243e-05, + "loss": 1.6097, + "step": 728 + }, + { + "epoch": 1.3437788018433179, + "grad_norm": 0.2717166841030121, + "learning_rate": 2.021932122561577e-05, + "loss": 1.5724, + "step": 729 + }, + { + "epoch": 1.3456221198156681, + "grad_norm": 0.2744804620742798, + "learning_rate": 2.0189111552503142e-05, + "loss": 1.6343, + "step": 730 + }, + { + "epoch": 1.3474654377880184, + "grad_norm": 0.2739951014518738, + "learning_rate": 2.015887795152643e-05, + "loss": 1.609, + "step": 731 + }, + { + "epoch": 1.3493087557603687, + "grad_norm": 0.2745543420314789, + "learning_rate": 2.0128620562097834e-05, + "loss": 1.634, + "step": 732 + }, + { + "epoch": 1.351152073732719, + "grad_norm": 0.2853536903858185, + "learning_rate": 2.009833952373925e-05, + "loss": 1.6915, + "step": 733 + }, + { + "epoch": 1.3529953917050692, + "grad_norm": 0.2606966495513916, + "learning_rate": 2.0068034976081637e-05, + "loss": 1.5641, + "step": 734 + }, + { + "epoch": 1.3548387096774195, + "grad_norm": 0.2707135081291199, + "learning_rate": 2.0037707058864343e-05, + "loss": 1.5901, + "step": 735 + }, + { + "epoch": 1.3566820276497695, + "grad_norm": 0.2738732397556305, + "learning_rate": 2.0007355911934473e-05, + "loss": 1.5878, + "step": 736 + }, + { + "epoch": 1.3585253456221198, + "grad_norm": 0.27152132987976074, + "learning_rate": 1.997698167524628e-05, + "loss": 1.6212, + "step": 737 + }, + { + "epoch": 1.36036866359447, + "grad_norm": 0.279140442609787, + "learning_rate": 1.9946584488860454e-05, + "loss": 1.5909, + "step": 738 + }, + { + "epoch": 1.3622119815668203, + "grad_norm": 0.2838742434978485, + "learning_rate": 1.9916164492943518e-05, + "loss": 1.6337, + "step": 739 + }, + { + "epoch": 1.3640552995391704, + "grad_norm": 0.2730039358139038, + "learning_rate": 1.9885721827767185e-05, + "loss": 1.674, + "step": 740 + }, + { + "epoch": 1.3658986175115206, + "grad_norm": 0.27932366728782654, + "learning_rate": 1.9855256633707692e-05, + "loss": 1.6264, + "step": 741 + }, + { + "epoch": 1.367741935483871, + "grad_norm": 0.2915544807910919, + "learning_rate": 1.9824769051245157e-05, + "loss": 1.6138, + "step": 742 + }, + { + "epoch": 1.3695852534562212, + "grad_norm": 0.2782309949398041, + "learning_rate": 1.979425922096294e-05, + "loss": 1.6153, + "step": 743 + }, + { + "epoch": 1.3714285714285714, + "grad_norm": 0.29362061619758606, + "learning_rate": 1.976372728354699e-05, + "loss": 1.6308, + "step": 744 + }, + { + "epoch": 1.3732718894009217, + "grad_norm": 0.2930099368095398, + "learning_rate": 1.9733173379785188e-05, + "loss": 1.5748, + "step": 745 + }, + { + "epoch": 1.375115207373272, + "grad_norm": 0.27453601360321045, + "learning_rate": 1.9702597650566723e-05, + "loss": 1.5993, + "step": 746 + }, + { + "epoch": 1.3769585253456222, + "grad_norm": 0.281548410654068, + "learning_rate": 1.9672000236881397e-05, + "loss": 1.6467, + "step": 747 + }, + { + "epoch": 1.3788018433179723, + "grad_norm": 0.28580373525619507, + "learning_rate": 1.9641381279819028e-05, + "loss": 1.6643, + "step": 748 + }, + { + "epoch": 1.3806451612903226, + "grad_norm": 0.2795877754688263, + "learning_rate": 1.9610740920568764e-05, + "loss": 1.6006, + "step": 749 + }, + { + "epoch": 1.3824884792626728, + "grad_norm": 0.27349579334259033, + "learning_rate": 1.9580079300418444e-05, + "loss": 1.654, + "step": 750 + }, + { + "epoch": 1.384331797235023, + "grad_norm": 0.27888205647468567, + "learning_rate": 1.954939656075394e-05, + "loss": 1.6131, + "step": 751 + }, + { + "epoch": 1.3861751152073734, + "grad_norm": 0.27265042066574097, + "learning_rate": 1.9518692843058514e-05, + "loss": 1.6203, + "step": 752 + }, + { + "epoch": 1.3880184331797234, + "grad_norm": 0.2736769914627075, + "learning_rate": 1.9487968288912164e-05, + "loss": 1.6011, + "step": 753 + }, + { + "epoch": 1.3898617511520737, + "grad_norm": 0.27037787437438965, + "learning_rate": 1.9457223039990963e-05, + "loss": 1.5475, + "step": 754 + }, + { + "epoch": 1.391705069124424, + "grad_norm": 0.27303317189216614, + "learning_rate": 1.942645723806641e-05, + "loss": 1.6335, + "step": 755 + }, + { + "epoch": 1.3935483870967742, + "grad_norm": 0.27159225940704346, + "learning_rate": 1.9395671025004777e-05, + "loss": 1.5606, + "step": 756 + }, + { + "epoch": 1.3953917050691245, + "grad_norm": 0.2682175934314728, + "learning_rate": 1.936486454276647e-05, + "loss": 1.555, + "step": 757 + }, + { + "epoch": 1.3972350230414747, + "grad_norm": 0.27098003029823303, + "learning_rate": 1.9334037933405337e-05, + "loss": 1.5385, + "step": 758 + }, + { + "epoch": 1.399078341013825, + "grad_norm": 0.28845879435539246, + "learning_rate": 1.9303191339068048e-05, + "loss": 1.6211, + "step": 759 + }, + { + "epoch": 1.400921658986175, + "grad_norm": 0.2876651883125305, + "learning_rate": 1.9272324901993436e-05, + "loss": 1.6319, + "step": 760 + }, + { + "epoch": 1.4027649769585253, + "grad_norm": 0.2796455919742584, + "learning_rate": 1.9241438764511805e-05, + "loss": 1.6263, + "step": 761 + }, + { + "epoch": 1.4046082949308756, + "grad_norm": 0.27480548620224, + "learning_rate": 1.9210533069044334e-05, + "loss": 1.613, + "step": 762 + }, + { + "epoch": 1.4064516129032258, + "grad_norm": 0.28065502643585205, + "learning_rate": 1.9179607958102356e-05, + "loss": 1.5789, + "step": 763 + }, + { + "epoch": 1.4082949308755761, + "grad_norm": 0.28731146454811096, + "learning_rate": 1.9148663574286757e-05, + "loss": 1.5297, + "step": 764 + }, + { + "epoch": 1.4101382488479262, + "grad_norm": 0.2914833724498749, + "learning_rate": 1.911770006028728e-05, + "loss": 1.5977, + "step": 765 + }, + { + "epoch": 1.4119815668202764, + "grad_norm": 0.300193190574646, + "learning_rate": 1.908671755888188e-05, + "loss": 1.6296, + "step": 766 + }, + { + "epoch": 1.4138248847926267, + "grad_norm": 0.29380565881729126, + "learning_rate": 1.9055716212936075e-05, + "loss": 1.6149, + "step": 767 + }, + { + "epoch": 1.415668202764977, + "grad_norm": 0.32037287950515747, + "learning_rate": 1.9024696165402272e-05, + "loss": 1.6513, + "step": 768 + }, + { + "epoch": 1.4175115207373272, + "grad_norm": 0.29914116859436035, + "learning_rate": 1.899365755931911e-05, + "loss": 1.5963, + "step": 769 + }, + { + "epoch": 1.4193548387096775, + "grad_norm": 0.26687341928482056, + "learning_rate": 1.8962600537810824e-05, + "loss": 1.536, + "step": 770 + }, + { + "epoch": 1.4211981566820278, + "grad_norm": 0.28808560967445374, + "learning_rate": 1.893152524408653e-05, + "loss": 1.6214, + "step": 771 + }, + { + "epoch": 1.4230414746543778, + "grad_norm": 0.28396207094192505, + "learning_rate": 1.8900431821439644e-05, + "loss": 1.6478, + "step": 772 + }, + { + "epoch": 1.424884792626728, + "grad_norm": 0.2826618552207947, + "learning_rate": 1.886932041324714e-05, + "loss": 1.5832, + "step": 773 + }, + { + "epoch": 1.4267281105990783, + "grad_norm": 0.2729291617870331, + "learning_rate": 1.883819116296895e-05, + "loss": 1.5696, + "step": 774 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.28617042303085327, + "learning_rate": 1.880704421414726e-05, + "loss": 1.606, + "step": 775 + }, + { + "epoch": 1.4304147465437789, + "grad_norm": 0.30494067072868347, + "learning_rate": 1.8775879710405893e-05, + "loss": 1.6557, + "step": 776 + }, + { + "epoch": 1.432258064516129, + "grad_norm": 0.2694535255432129, + "learning_rate": 1.8744697795449588e-05, + "loss": 1.5447, + "step": 777 + }, + { + "epoch": 1.4341013824884792, + "grad_norm": 0.2713358998298645, + "learning_rate": 1.8713498613063403e-05, + "loss": 1.5635, + "step": 778 + }, + { + "epoch": 1.4359447004608294, + "grad_norm": 0.27467259764671326, + "learning_rate": 1.8682282307111988e-05, + "loss": 1.6066, + "step": 779 + }, + { + "epoch": 1.4377880184331797, + "grad_norm": 0.2681735157966614, + "learning_rate": 1.865104902153898e-05, + "loss": 1.5669, + "step": 780 + }, + { + "epoch": 1.43963133640553, + "grad_norm": 0.2794997990131378, + "learning_rate": 1.8619798900366298e-05, + "loss": 1.6059, + "step": 781 + }, + { + "epoch": 1.4414746543778802, + "grad_norm": 0.27287647128105164, + "learning_rate": 1.8588532087693485e-05, + "loss": 1.5776, + "step": 782 + }, + { + "epoch": 1.4433179723502305, + "grad_norm": 0.2879515290260315, + "learning_rate": 1.8557248727697068e-05, + "loss": 1.6362, + "step": 783 + }, + { + "epoch": 1.4451612903225808, + "grad_norm": 0.28515344858169556, + "learning_rate": 1.852594896462987e-05, + "loss": 1.5876, + "step": 784 + }, + { + "epoch": 1.4470046082949308, + "grad_norm": 0.2730424702167511, + "learning_rate": 1.849463294282035e-05, + "loss": 1.5707, + "step": 785 + }, + { + "epoch": 1.448847926267281, + "grad_norm": 0.27559229731559753, + "learning_rate": 1.8463300806671936e-05, + "loss": 1.5538, + "step": 786 + }, + { + "epoch": 1.4506912442396314, + "grad_norm": 0.28002747893333435, + "learning_rate": 1.8431952700662375e-05, + "loss": 1.6236, + "step": 787 + }, + { + "epoch": 1.4525345622119816, + "grad_norm": 0.28345340490341187, + "learning_rate": 1.840058876934303e-05, + "loss": 1.6436, + "step": 788 + }, + { + "epoch": 1.4543778801843317, + "grad_norm": 0.26679205894470215, + "learning_rate": 1.8369209157338262e-05, + "loss": 1.49, + "step": 789 + }, + { + "epoch": 1.456221198156682, + "grad_norm": 0.2743077874183655, + "learning_rate": 1.8337814009344716e-05, + "loss": 1.592, + "step": 790 + }, + { + "epoch": 1.4580645161290322, + "grad_norm": 0.2735365927219391, + "learning_rate": 1.83064034701307e-05, + "loss": 1.5771, + "step": 791 + }, + { + "epoch": 1.4599078341013825, + "grad_norm": 0.27030524611473083, + "learning_rate": 1.8274977684535478e-05, + "loss": 1.5751, + "step": 792 + }, + { + "epoch": 1.4617511520737327, + "grad_norm": 0.28183117508888245, + "learning_rate": 1.824353679746861e-05, + "loss": 1.5485, + "step": 793 + }, + { + "epoch": 1.463594470046083, + "grad_norm": 0.2746550738811493, + "learning_rate": 1.821208095390931e-05, + "loss": 1.5738, + "step": 794 + }, + { + "epoch": 1.4654377880184333, + "grad_norm": 0.2818034887313843, + "learning_rate": 1.8180610298905758e-05, + "loss": 1.6364, + "step": 795 + }, + { + "epoch": 1.4672811059907835, + "grad_norm": 0.2696681320667267, + "learning_rate": 1.8149124977574417e-05, + "loss": 1.6, + "step": 796 + }, + { + "epoch": 1.4691244239631336, + "grad_norm": 0.2887495458126068, + "learning_rate": 1.8117625135099386e-05, + "loss": 1.6686, + "step": 797 + }, + { + "epoch": 1.4709677419354839, + "grad_norm": 0.2815185785293579, + "learning_rate": 1.8086110916731724e-05, + "loss": 1.6131, + "step": 798 + }, + { + "epoch": 1.4728110599078341, + "grad_norm": 0.27256280183792114, + "learning_rate": 1.805458246778878e-05, + "loss": 1.5867, + "step": 799 + }, + { + "epoch": 1.4746543778801844, + "grad_norm": 0.2779887914657593, + "learning_rate": 1.802303993365353e-05, + "loss": 1.5557, + "step": 800 + }, + { + "epoch": 1.4764976958525344, + "grad_norm": 0.27857843041419983, + "learning_rate": 1.7991483459773887e-05, + "loss": 1.6668, + "step": 801 + }, + { + "epoch": 1.4783410138248847, + "grad_norm": 0.26913055777549744, + "learning_rate": 1.795991319166204e-05, + "loss": 1.6072, + "step": 802 + }, + { + "epoch": 1.480184331797235, + "grad_norm": 0.2818872630596161, + "learning_rate": 1.79283292748938e-05, + "loss": 1.5957, + "step": 803 + }, + { + "epoch": 1.4820276497695852, + "grad_norm": 0.2774152159690857, + "learning_rate": 1.7896731855107908e-05, + "loss": 1.5923, + "step": 804 + }, + { + "epoch": 1.4838709677419355, + "grad_norm": 0.2798556983470917, + "learning_rate": 1.7865121078005365e-05, + "loss": 1.5798, + "step": 805 + }, + { + "epoch": 1.4857142857142858, + "grad_norm": 0.2692021429538727, + "learning_rate": 1.7833497089348772e-05, + "loss": 1.6172, + "step": 806 + }, + { + "epoch": 1.487557603686636, + "grad_norm": 0.267347514629364, + "learning_rate": 1.780186003496164e-05, + "loss": 1.6114, + "step": 807 + }, + { + "epoch": 1.4894009216589863, + "grad_norm": 0.2735111713409424, + "learning_rate": 1.7770210060727748e-05, + "loss": 1.5757, + "step": 808 + }, + { + "epoch": 1.4912442396313363, + "grad_norm": 0.27940452098846436, + "learning_rate": 1.7738547312590426e-05, + "loss": 1.6085, + "step": 809 + }, + { + "epoch": 1.4930875576036866, + "grad_norm": 0.2675354480743408, + "learning_rate": 1.770687193655192e-05, + "loss": 1.6032, + "step": 810 + }, + { + "epoch": 1.4949308755760369, + "grad_norm": 0.276353120803833, + "learning_rate": 1.7675184078672714e-05, + "loss": 1.6087, + "step": 811 + }, + { + "epoch": 1.4967741935483871, + "grad_norm": 0.276836097240448, + "learning_rate": 1.7643483885070827e-05, + "loss": 1.6077, + "step": 812 + }, + { + "epoch": 1.4986175115207372, + "grad_norm": 0.27562999725341797, + "learning_rate": 1.7611771501921174e-05, + "loss": 1.6598, + "step": 813 + }, + { + "epoch": 1.5004608294930875, + "grad_norm": 0.2816264033317566, + "learning_rate": 1.7580047075454877e-05, + "loss": 1.6591, + "step": 814 + }, + { + "epoch": 1.5023041474654377, + "grad_norm": 0.2649102807044983, + "learning_rate": 1.7548310751958588e-05, + "loss": 1.5475, + "step": 815 + }, + { + "epoch": 1.504147465437788, + "grad_norm": 0.26588475704193115, + "learning_rate": 1.751656267777382e-05, + "loss": 1.6011, + "step": 816 + }, + { + "epoch": 1.5059907834101383, + "grad_norm": 0.2885441482067108, + "learning_rate": 1.748480299929627e-05, + "loss": 1.6321, + "step": 817 + }, + { + "epoch": 1.5078341013824885, + "grad_norm": 0.2782214283943176, + "learning_rate": 1.7453031862975146e-05, + "loss": 1.5943, + "step": 818 + }, + { + "epoch": 1.5096774193548388, + "grad_norm": 0.2720677852630615, + "learning_rate": 1.742124941531249e-05, + "loss": 1.5845, + "step": 819 + }, + { + "epoch": 1.511520737327189, + "grad_norm": 0.27073559165000916, + "learning_rate": 1.73894558028625e-05, + "loss": 1.6024, + "step": 820 + }, + { + "epoch": 1.5133640552995393, + "grad_norm": 0.2795216739177704, + "learning_rate": 1.7357651172230852e-05, + "loss": 1.5477, + "step": 821 + }, + { + "epoch": 1.5152073732718894, + "grad_norm": 0.27710703015327454, + "learning_rate": 1.7325835670074044e-05, + "loss": 1.6505, + "step": 822 + }, + { + "epoch": 1.5170506912442396, + "grad_norm": 0.2705666124820709, + "learning_rate": 1.729400944309869e-05, + "loss": 1.5482, + "step": 823 + }, + { + "epoch": 1.51889400921659, + "grad_norm": 0.27922290563583374, + "learning_rate": 1.7262172638060865e-05, + "loss": 1.6243, + "step": 824 + }, + { + "epoch": 1.52073732718894, + "grad_norm": 0.27905556559562683, + "learning_rate": 1.7230325401765415e-05, + "loss": 1.5902, + "step": 825 + }, + { + "epoch": 1.5225806451612902, + "grad_norm": 0.2822147607803345, + "learning_rate": 1.7198467881065292e-05, + "loss": 1.6411, + "step": 826 + }, + { + "epoch": 1.5244239631336405, + "grad_norm": 0.27511006593704224, + "learning_rate": 1.7166600222860876e-05, + "loss": 1.6384, + "step": 827 + }, + { + "epoch": 1.5262672811059907, + "grad_norm": 0.27862241864204407, + "learning_rate": 1.713472257409928e-05, + "loss": 1.5762, + "step": 828 + }, + { + "epoch": 1.528110599078341, + "grad_norm": 0.27879902720451355, + "learning_rate": 1.7102835081773686e-05, + "loss": 1.6238, + "step": 829 + }, + { + "epoch": 1.5299539170506913, + "grad_norm": 0.29454904794692993, + "learning_rate": 1.707093789292269e-05, + "loss": 1.6545, + "step": 830 + }, + { + "epoch": 1.5317972350230415, + "grad_norm": 0.2856805920600891, + "learning_rate": 1.7039031154629567e-05, + "loss": 1.5693, + "step": 831 + }, + { + "epoch": 1.5336405529953918, + "grad_norm": 0.27673226594924927, + "learning_rate": 1.700711501402164e-05, + "loss": 1.5427, + "step": 832 + }, + { + "epoch": 1.535483870967742, + "grad_norm": 0.29206371307373047, + "learning_rate": 1.6975189618269592e-05, + "loss": 1.6024, + "step": 833 + }, + { + "epoch": 1.5373271889400921, + "grad_norm": 0.2787723243236542, + "learning_rate": 1.6943255114586788e-05, + "loss": 1.5581, + "step": 834 + }, + { + "epoch": 1.5391705069124424, + "grad_norm": 0.27273157238960266, + "learning_rate": 1.6911311650228574e-05, + "loss": 1.5769, + "step": 835 + }, + { + "epoch": 1.5410138248847927, + "grad_norm": 0.2830120325088501, + "learning_rate": 1.687935937249163e-05, + "loss": 1.5915, + "step": 836 + }, + { + "epoch": 1.5428571428571427, + "grad_norm": 0.27038127183914185, + "learning_rate": 1.6847398428713256e-05, + "loss": 1.5609, + "step": 837 + }, + { + "epoch": 1.544700460829493, + "grad_norm": 0.2764512896537781, + "learning_rate": 1.681542896627075e-05, + "loss": 1.6441, + "step": 838 + }, + { + "epoch": 1.5465437788018432, + "grad_norm": 0.27623921632766724, + "learning_rate": 1.678345113258065e-05, + "loss": 1.6269, + "step": 839 + }, + { + "epoch": 1.5483870967741935, + "grad_norm": 0.26323872804641724, + "learning_rate": 1.6751465075098115e-05, + "loss": 1.5342, + "step": 840 + }, + { + "epoch": 1.5502304147465438, + "grad_norm": 0.27324774861335754, + "learning_rate": 1.6719470941316228e-05, + "loss": 1.6072, + "step": 841 + }, + { + "epoch": 1.552073732718894, + "grad_norm": 0.2702793776988983, + "learning_rate": 1.668746887876531e-05, + "loss": 1.5937, + "step": 842 + }, + { + "epoch": 1.5539170506912443, + "grad_norm": 0.2792288362979889, + "learning_rate": 1.6655459035012237e-05, + "loss": 1.5874, + "step": 843 + }, + { + "epoch": 1.5557603686635946, + "grad_norm": 0.2696288824081421, + "learning_rate": 1.662344155765977e-05, + "loss": 1.5788, + "step": 844 + }, + { + "epoch": 1.5576036866359448, + "grad_norm": 0.2754669189453125, + "learning_rate": 1.659141659434587e-05, + "loss": 1.6263, + "step": 845 + }, + { + "epoch": 1.5594470046082949, + "grad_norm": 0.2766124904155731, + "learning_rate": 1.655938429274302e-05, + "loss": 1.6164, + "step": 846 + }, + { + "epoch": 1.5612903225806452, + "grad_norm": 0.2690373361110687, + "learning_rate": 1.6527344800557534e-05, + "loss": 1.5735, + "step": 847 + }, + { + "epoch": 1.5631336405529954, + "grad_norm": 0.27754339575767517, + "learning_rate": 1.6495298265528883e-05, + "loss": 1.6258, + "step": 848 + }, + { + "epoch": 1.5649769585253455, + "grad_norm": 0.27151575684547424, + "learning_rate": 1.646324483542902e-05, + "loss": 1.6568, + "step": 849 + }, + { + "epoch": 1.5668202764976957, + "grad_norm": 0.28297892212867737, + "learning_rate": 1.64311846580617e-05, + "loss": 1.6342, + "step": 850 + }, + { + "epoch": 1.568663594470046, + "grad_norm": 0.27492207288742065, + "learning_rate": 1.639911788126177e-05, + "loss": 1.665, + "step": 851 + }, + { + "epoch": 1.5705069124423963, + "grad_norm": 0.2828580439090729, + "learning_rate": 1.6367044652894515e-05, + "loss": 1.5696, + "step": 852 + }, + { + "epoch": 1.5723502304147465, + "grad_norm": 0.2775357961654663, + "learning_rate": 1.6334965120854986e-05, + "loss": 1.6489, + "step": 853 + }, + { + "epoch": 1.5741935483870968, + "grad_norm": 0.28201982378959656, + "learning_rate": 1.6302879433067274e-05, + "loss": 1.6067, + "step": 854 + }, + { + "epoch": 1.576036866359447, + "grad_norm": 0.273805171251297, + "learning_rate": 1.6270787737483877e-05, + "loss": 1.636, + "step": 855 + }, + { + "epoch": 1.5778801843317973, + "grad_norm": 0.27378591895103455, + "learning_rate": 1.623869018208499e-05, + "loss": 1.5383, + "step": 856 + }, + { + "epoch": 1.5797235023041476, + "grad_norm": 0.2734542787075043, + "learning_rate": 1.6206586914877816e-05, + "loss": 1.5959, + "step": 857 + }, + { + "epoch": 1.5815668202764976, + "grad_norm": 0.2755284905433655, + "learning_rate": 1.6174478083895922e-05, + "loss": 1.5877, + "step": 858 + }, + { + "epoch": 1.583410138248848, + "grad_norm": 0.2930653393268585, + "learning_rate": 1.6142363837198504e-05, + "loss": 1.6763, + "step": 859 + }, + { + "epoch": 1.5852534562211982, + "grad_norm": 0.28103727102279663, + "learning_rate": 1.6110244322869746e-05, + "loss": 1.6154, + "step": 860 + }, + { + "epoch": 1.5870967741935482, + "grad_norm": 0.2758486568927765, + "learning_rate": 1.607811968901812e-05, + "loss": 1.5919, + "step": 861 + }, + { + "epoch": 1.5889400921658985, + "grad_norm": 0.27371156215667725, + "learning_rate": 1.6045990083775703e-05, + "loss": 1.6255, + "step": 862 + }, + { + "epoch": 1.5907834101382488, + "grad_norm": 0.2752973139286041, + "learning_rate": 1.6013855655297498e-05, + "loss": 1.6017, + "step": 863 + }, + { + "epoch": 1.592626728110599, + "grad_norm": 0.26971954107284546, + "learning_rate": 1.5981716551760735e-05, + "loss": 1.5115, + "step": 864 + }, + { + "epoch": 1.5944700460829493, + "grad_norm": 0.2720949053764343, + "learning_rate": 1.5949572921364226e-05, + "loss": 1.573, + "step": 865 + }, + { + "epoch": 1.5963133640552996, + "grad_norm": 0.2740275263786316, + "learning_rate": 1.5917424912327644e-05, + "loss": 1.6024, + "step": 866 + }, + { + "epoch": 1.5981566820276498, + "grad_norm": 0.2787199318408966, + "learning_rate": 1.5885272672890842e-05, + "loss": 1.6263, + "step": 867 + }, + { + "epoch": 1.6, + "grad_norm": 0.2864430248737335, + "learning_rate": 1.58531163513132e-05, + "loss": 1.6032, + "step": 868 + }, + { + "epoch": 1.6018433179723504, + "grad_norm": 0.26860958337783813, + "learning_rate": 1.5820956095872914e-05, + "loss": 1.6096, + "step": 869 + }, + { + "epoch": 1.6036866359447006, + "grad_norm": 0.27173295617103577, + "learning_rate": 1.5788792054866314e-05, + "loss": 1.5589, + "step": 870 + }, + { + "epoch": 1.6055299539170507, + "grad_norm": 0.27752557396888733, + "learning_rate": 1.5756624376607193e-05, + "loss": 1.5585, + "step": 871 + }, + { + "epoch": 1.607373271889401, + "grad_norm": 0.2819308042526245, + "learning_rate": 1.5724453209426108e-05, + "loss": 1.578, + "step": 872 + }, + { + "epoch": 1.6092165898617512, + "grad_norm": 0.2847549617290497, + "learning_rate": 1.5692278701669712e-05, + "loss": 1.6011, + "step": 873 + }, + { + "epoch": 1.6110599078341012, + "grad_norm": 0.27277764678001404, + "learning_rate": 1.566010100170007e-05, + "loss": 1.571, + "step": 874 + }, + { + "epoch": 1.6129032258064515, + "grad_norm": 0.2827809751033783, + "learning_rate": 1.5627920257893934e-05, + "loss": 1.5961, + "step": 875 + }, + { + "epoch": 1.6147465437788018, + "grad_norm": 0.28699159622192383, + "learning_rate": 1.5595736618642126e-05, + "loss": 1.6229, + "step": 876 + }, + { + "epoch": 1.616589861751152, + "grad_norm": 0.2775685787200928, + "learning_rate": 1.5563550232348813e-05, + "loss": 1.5469, + "step": 877 + }, + { + "epoch": 1.6184331797235023, + "grad_norm": 0.2769649922847748, + "learning_rate": 1.553136124743081e-05, + "loss": 1.591, + "step": 878 + }, + { + "epoch": 1.6202764976958526, + "grad_norm": 0.26347190141677856, + "learning_rate": 1.5499169812316937e-05, + "loss": 1.5349, + "step": 879 + }, + { + "epoch": 1.6221198156682028, + "grad_norm": 0.2721926271915436, + "learning_rate": 1.5466976075447295e-05, + "loss": 1.5831, + "step": 880 + }, + { + "epoch": 1.6239631336405531, + "grad_norm": 0.2716809809207916, + "learning_rate": 1.5434780185272616e-05, + "loss": 1.5148, + "step": 881 + }, + { + "epoch": 1.6258064516129034, + "grad_norm": 0.2808883786201477, + "learning_rate": 1.5402582290253547e-05, + "loss": 1.57, + "step": 882 + }, + { + "epoch": 1.6276497695852534, + "grad_norm": 0.2672334313392639, + "learning_rate": 1.537038253885998e-05, + "loss": 1.5688, + "step": 883 + }, + { + "epoch": 1.6294930875576037, + "grad_norm": 0.27970781922340393, + "learning_rate": 1.533818107957038e-05, + "loss": 1.5723, + "step": 884 + }, + { + "epoch": 1.631336405529954, + "grad_norm": 0.2877601087093353, + "learning_rate": 1.5305978060871083e-05, + "loss": 1.6315, + "step": 885 + }, + { + "epoch": 1.633179723502304, + "grad_norm": 0.2766900658607483, + "learning_rate": 1.5273773631255602e-05, + "loss": 1.632, + "step": 886 + }, + { + "epoch": 1.6350230414746543, + "grad_norm": 0.26466235518455505, + "learning_rate": 1.524156793922396e-05, + "loss": 1.546, + "step": 887 + }, + { + "epoch": 1.6368663594470045, + "grad_norm": 0.27303576469421387, + "learning_rate": 1.5209361133282022e-05, + "loss": 1.574, + "step": 888 + }, + { + "epoch": 1.6387096774193548, + "grad_norm": 0.28713706135749817, + "learning_rate": 1.517715336194077e-05, + "loss": 1.6679, + "step": 889 + }, + { + "epoch": 1.640552995391705, + "grad_norm": 0.271852046251297, + "learning_rate": 1.5144944773715635e-05, + "loss": 1.5074, + "step": 890 + }, + { + "epoch": 1.6423963133640553, + "grad_norm": 0.2695780098438263, + "learning_rate": 1.511273551712583e-05, + "loss": 1.5874, + "step": 891 + }, + { + "epoch": 1.6442396313364056, + "grad_norm": 0.27548617124557495, + "learning_rate": 1.5080525740693635e-05, + "loss": 1.5366, + "step": 892 + }, + { + "epoch": 1.6460829493087559, + "grad_norm": 0.27405300736427307, + "learning_rate": 1.5048315592943743e-05, + "loss": 1.6149, + "step": 893 + }, + { + "epoch": 1.6479262672811061, + "grad_norm": 0.2837236821651459, + "learning_rate": 1.5016105222402546e-05, + "loss": 1.574, + "step": 894 + }, + { + "epoch": 1.6497695852534562, + "grad_norm": 0.274830162525177, + "learning_rate": 1.4983894777597461e-05, + "loss": 1.6566, + "step": 895 + }, + { + "epoch": 1.6516129032258065, + "grad_norm": 0.2712487578392029, + "learning_rate": 1.495168440705626e-05, + "loss": 1.5382, + "step": 896 + }, + { + "epoch": 1.6534562211981567, + "grad_norm": 0.26845115423202515, + "learning_rate": 1.4919474259306362e-05, + "loss": 1.5384, + "step": 897 + }, + { + "epoch": 1.6552995391705068, + "grad_norm": 0.2784380316734314, + "learning_rate": 1.4887264482874173e-05, + "loss": 1.5575, + "step": 898 + }, + { + "epoch": 1.657142857142857, + "grad_norm": 0.2723720967769623, + "learning_rate": 1.4855055226284367e-05, + "loss": 1.5714, + "step": 899 + }, + { + "epoch": 1.6589861751152073, + "grad_norm": 0.2713598608970642, + "learning_rate": 1.4822846638059234e-05, + "loss": 1.5896, + "step": 900 + }, + { + "epoch": 1.6608294930875576, + "grad_norm": 0.28683602809906006, + "learning_rate": 1.4790638866717984e-05, + "loss": 1.6283, + "step": 901 + }, + { + "epoch": 1.6626728110599078, + "grad_norm": 0.2853665053844452, + "learning_rate": 1.4758432060776044e-05, + "loss": 1.5921, + "step": 902 + }, + { + "epoch": 1.664516129032258, + "grad_norm": 0.279269278049469, + "learning_rate": 1.4726226368744404e-05, + "loss": 1.595, + "step": 903 + }, + { + "epoch": 1.6663594470046084, + "grad_norm": 0.2805596590042114, + "learning_rate": 1.4694021939128925e-05, + "loss": 1.589, + "step": 904 + }, + { + "epoch": 1.6682027649769586, + "grad_norm": 0.2930944859981537, + "learning_rate": 1.466181892042962e-05, + "loss": 1.6052, + "step": 905 + }, + { + "epoch": 1.670046082949309, + "grad_norm": 0.27433332800865173, + "learning_rate": 1.462961746114002e-05, + "loss": 1.6106, + "step": 906 + }, + { + "epoch": 1.671889400921659, + "grad_norm": 0.29066669940948486, + "learning_rate": 1.4597417709746454e-05, + "loss": 1.6252, + "step": 907 + }, + { + "epoch": 1.6737327188940092, + "grad_norm": 0.2851257026195526, + "learning_rate": 1.4565219814727388e-05, + "loss": 1.5807, + "step": 908 + }, + { + "epoch": 1.6755760368663595, + "grad_norm": 0.2750568389892578, + "learning_rate": 1.4533023924552706e-05, + "loss": 1.5701, + "step": 909 + }, + { + "epoch": 1.6774193548387095, + "grad_norm": 0.28285452723503113, + "learning_rate": 1.4500830187683066e-05, + "loss": 1.6027, + "step": 910 + }, + { + "epoch": 1.6792626728110598, + "grad_norm": 0.28591400384902954, + "learning_rate": 1.4468638752569193e-05, + "loss": 1.6226, + "step": 911 + }, + { + "epoch": 1.68110599078341, + "grad_norm": 0.279861718416214, + "learning_rate": 1.4436449767651191e-05, + "loss": 1.5525, + "step": 912 + }, + { + "epoch": 1.6829493087557603, + "grad_norm": 0.26922115683555603, + "learning_rate": 1.4404263381357873e-05, + "loss": 1.5962, + "step": 913 + }, + { + "epoch": 1.6847926267281106, + "grad_norm": 0.2725696265697479, + "learning_rate": 1.437207974210607e-05, + "loss": 1.5739, + "step": 914 + }, + { + "epoch": 1.6866359447004609, + "grad_norm": 0.280691921710968, + "learning_rate": 1.4339898998299936e-05, + "loss": 1.6128, + "step": 915 + }, + { + "epoch": 1.6884792626728111, + "grad_norm": 0.27789539098739624, + "learning_rate": 1.4307721298330284e-05, + "loss": 1.5863, + "step": 916 + }, + { + "epoch": 1.6903225806451614, + "grad_norm": 0.2718709707260132, + "learning_rate": 1.4275546790573895e-05, + "loss": 1.5724, + "step": 917 + }, + { + "epoch": 1.6921658986175117, + "grad_norm": 0.2684226930141449, + "learning_rate": 1.4243375623392808e-05, + "loss": 1.5473, + "step": 918 + }, + { + "epoch": 1.6940092165898617, + "grad_norm": 0.28611379861831665, + "learning_rate": 1.4211207945133685e-05, + "loss": 1.6016, + "step": 919 + }, + { + "epoch": 1.695852534562212, + "grad_norm": 0.28246620297431946, + "learning_rate": 1.417904390412709e-05, + "loss": 1.5635, + "step": 920 + }, + { + "epoch": 1.6976958525345622, + "grad_norm": 0.2792202830314636, + "learning_rate": 1.41468836486868e-05, + "loss": 1.5302, + "step": 921 + }, + { + "epoch": 1.6995391705069123, + "grad_norm": 0.2743517756462097, + "learning_rate": 1.411472732710916e-05, + "loss": 1.5837, + "step": 922 + }, + { + "epoch": 1.7013824884792625, + "grad_norm": 0.28153350949287415, + "learning_rate": 1.4082575087672363e-05, + "loss": 1.6095, + "step": 923 + }, + { + "epoch": 1.7032258064516128, + "grad_norm": 0.30301469564437866, + "learning_rate": 1.4050427078635777e-05, + "loss": 1.5882, + "step": 924 + }, + { + "epoch": 1.705069124423963, + "grad_norm": 0.27385833859443665, + "learning_rate": 1.4018283448239266e-05, + "loss": 1.5997, + "step": 925 + }, + { + "epoch": 1.7069124423963133, + "grad_norm": 0.2762070298194885, + "learning_rate": 1.398614434470251e-05, + "loss": 1.5694, + "step": 926 + }, + { + "epoch": 1.7087557603686636, + "grad_norm": 0.27967438101768494, + "learning_rate": 1.3954009916224299e-05, + "loss": 1.6191, + "step": 927 + }, + { + "epoch": 1.7105990783410139, + "grad_norm": 0.2745825946331024, + "learning_rate": 1.3921880310981878e-05, + "loss": 1.5921, + "step": 928 + }, + { + "epoch": 1.7124423963133641, + "grad_norm": 0.2806471586227417, + "learning_rate": 1.3889755677130253e-05, + "loss": 1.6333, + "step": 929 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 0.28569644689559937, + "learning_rate": 1.3857636162801499e-05, + "loss": 1.6297, + "step": 930 + }, + { + "epoch": 1.7161290322580647, + "grad_norm": 0.27753588557243347, + "learning_rate": 1.3825521916104082e-05, + "loss": 1.5691, + "step": 931 + }, + { + "epoch": 1.7179723502304147, + "grad_norm": 0.2785869538784027, + "learning_rate": 1.3793413085122183e-05, + "loss": 1.6432, + "step": 932 + }, + { + "epoch": 1.719815668202765, + "grad_norm": 0.2777843773365021, + "learning_rate": 1.3761309817915017e-05, + "loss": 1.6167, + "step": 933 + }, + { + "epoch": 1.7216589861751153, + "grad_norm": 0.2760448157787323, + "learning_rate": 1.3729212262516124e-05, + "loss": 1.6157, + "step": 934 + }, + { + "epoch": 1.7235023041474653, + "grad_norm": 0.2856210172176361, + "learning_rate": 1.3697120566932727e-05, + "loss": 1.5703, + "step": 935 + }, + { + "epoch": 1.7253456221198156, + "grad_norm": 0.27198657393455505, + "learning_rate": 1.3665034879145022e-05, + "loss": 1.6183, + "step": 936 + }, + { + "epoch": 1.7271889400921658, + "grad_norm": 0.2798546254634857, + "learning_rate": 1.3632955347105487e-05, + "loss": 1.6312, + "step": 937 + }, + { + "epoch": 1.729032258064516, + "grad_norm": 0.2872222065925598, + "learning_rate": 1.3600882118738232e-05, + "loss": 1.6336, + "step": 938 + }, + { + "epoch": 1.7308755760368664, + "grad_norm": 0.2799721360206604, + "learning_rate": 1.3568815341938303e-05, + "loss": 1.6183, + "step": 939 + }, + { + "epoch": 1.7327188940092166, + "grad_norm": 0.2797256410121918, + "learning_rate": 1.3536755164570977e-05, + "loss": 1.6386, + "step": 940 + }, + { + "epoch": 1.734562211981567, + "grad_norm": 0.2698177099227905, + "learning_rate": 1.3504701734471117e-05, + "loss": 1.5957, + "step": 941 + }, + { + "epoch": 1.7364055299539172, + "grad_norm": 0.2770727872848511, + "learning_rate": 1.3472655199442473e-05, + "loss": 1.5977, + "step": 942 + }, + { + "epoch": 1.7382488479262674, + "grad_norm": 0.2778874635696411, + "learning_rate": 1.3440615707256984e-05, + "loss": 1.5497, + "step": 943 + }, + { + "epoch": 1.7400921658986175, + "grad_norm": 0.27074167132377625, + "learning_rate": 1.340858340565413e-05, + "loss": 1.5602, + "step": 944 + }, + { + "epoch": 1.7419354838709677, + "grad_norm": 0.2706243395805359, + "learning_rate": 1.3376558442340233e-05, + "loss": 1.5655, + "step": 945 + }, + { + "epoch": 1.743778801843318, + "grad_norm": 0.27816662192344666, + "learning_rate": 1.3344540964987766e-05, + "loss": 1.5792, + "step": 946 + }, + { + "epoch": 1.745622119815668, + "grad_norm": 0.27811580896377563, + "learning_rate": 1.331253112123469e-05, + "loss": 1.6535, + "step": 947 + }, + { + "epoch": 1.7474654377880183, + "grad_norm": 0.28679028153419495, + "learning_rate": 1.3280529058683778e-05, + "loss": 1.5837, + "step": 948 + }, + { + "epoch": 1.7493087557603686, + "grad_norm": 0.2821553647518158, + "learning_rate": 1.3248534924901887e-05, + "loss": 1.5671, + "step": 949 + }, + { + "epoch": 1.7511520737327189, + "grad_norm": 0.27457195520401, + "learning_rate": 1.3216548867419352e-05, + "loss": 1.6098, + "step": 950 + }, + { + "epoch": 1.7529953917050691, + "grad_norm": 0.27999773621559143, + "learning_rate": 1.3184571033729253e-05, + "loss": 1.5503, + "step": 951 + }, + { + "epoch": 1.7548387096774194, + "grad_norm": 0.27848124504089355, + "learning_rate": 1.3152601571286746e-05, + "loss": 1.5739, + "step": 952 + }, + { + "epoch": 1.7566820276497697, + "grad_norm": 0.2803337275981903, + "learning_rate": 1.3120640627508376e-05, + "loss": 1.5847, + "step": 953 + }, + { + "epoch": 1.75852534562212, + "grad_norm": 0.27653852105140686, + "learning_rate": 1.3088688349771425e-05, + "loss": 1.6444, + "step": 954 + }, + { + "epoch": 1.7603686635944702, + "grad_norm": 0.27364403009414673, + "learning_rate": 1.3056744885413216e-05, + "loss": 1.603, + "step": 955 + }, + { + "epoch": 1.7622119815668202, + "grad_norm": 0.286454439163208, + "learning_rate": 1.3024810381730409e-05, + "loss": 1.6084, + "step": 956 + }, + { + "epoch": 1.7640552995391705, + "grad_norm": 0.27595254778862, + "learning_rate": 1.2992884985978363e-05, + "loss": 1.6451, + "step": 957 + }, + { + "epoch": 1.7658986175115208, + "grad_norm": 0.27956917881965637, + "learning_rate": 1.2960968845370443e-05, + "loss": 1.5732, + "step": 958 + }, + { + "epoch": 1.7677419354838708, + "grad_norm": 0.2734554708003998, + "learning_rate": 1.2929062107077315e-05, + "loss": 1.5397, + "step": 959 + }, + { + "epoch": 1.769585253456221, + "grad_norm": 0.27924489974975586, + "learning_rate": 1.2897164918226311e-05, + "loss": 1.552, + "step": 960 + }, + { + "epoch": 1.7714285714285714, + "grad_norm": 0.26942330598831177, + "learning_rate": 1.2865277425900725e-05, + "loss": 1.5747, + "step": 961 + }, + { + "epoch": 1.7732718894009216, + "grad_norm": 0.2752183675765991, + "learning_rate": 1.2833399777139128e-05, + "loss": 1.5018, + "step": 962 + }, + { + "epoch": 1.7751152073732719, + "grad_norm": 0.29580366611480713, + "learning_rate": 1.2801532118934708e-05, + "loss": 1.5727, + "step": 963 + }, + { + "epoch": 1.7769585253456222, + "grad_norm": 0.2765806019306183, + "learning_rate": 1.276967459823459e-05, + "loss": 1.5706, + "step": 964 + }, + { + "epoch": 1.7788018433179724, + "grad_norm": 0.2774519622325897, + "learning_rate": 1.273782736193914e-05, + "loss": 1.6598, + "step": 965 + }, + { + "epoch": 1.7806451612903227, + "grad_norm": 0.27766409516334534, + "learning_rate": 1.2705990556901311e-05, + "loss": 1.584, + "step": 966 + }, + { + "epoch": 1.782488479262673, + "grad_norm": 0.2720947265625, + "learning_rate": 1.2674164329925961e-05, + "loss": 1.5988, + "step": 967 + }, + { + "epoch": 1.784331797235023, + "grad_norm": 0.2917366623878479, + "learning_rate": 1.2642348827769152e-05, + "loss": 1.5834, + "step": 968 + }, + { + "epoch": 1.7861751152073733, + "grad_norm": 0.2705937623977661, + "learning_rate": 1.2610544197137502e-05, + "loss": 1.5643, + "step": 969 + }, + { + "epoch": 1.7880184331797235, + "grad_norm": 0.2769400477409363, + "learning_rate": 1.257875058468751e-05, + "loss": 1.6284, + "step": 970 + }, + { + "epoch": 1.7898617511520736, + "grad_norm": 0.28370919823646545, + "learning_rate": 1.2546968137024856e-05, + "loss": 1.6223, + "step": 971 + }, + { + "epoch": 1.7917050691244238, + "grad_norm": 0.27669093012809753, + "learning_rate": 1.251519700070373e-05, + "loss": 1.5396, + "step": 972 + }, + { + "epoch": 1.793548387096774, + "grad_norm": 0.272204726934433, + "learning_rate": 1.2483437322226178e-05, + "loss": 1.5131, + "step": 973 + }, + { + "epoch": 1.7953917050691244, + "grad_norm": 0.27512574195861816, + "learning_rate": 1.2451689248041416e-05, + "loss": 1.6107, + "step": 974 + }, + { + "epoch": 1.7972350230414746, + "grad_norm": 0.2767089307308197, + "learning_rate": 1.2419952924545125e-05, + "loss": 1.5571, + "step": 975 + }, + { + "epoch": 1.799078341013825, + "grad_norm": 0.2801191806793213, + "learning_rate": 1.2388228498078827e-05, + "loss": 1.6405, + "step": 976 + }, + { + "epoch": 1.8009216589861752, + "grad_norm": 0.27706649899482727, + "learning_rate": 1.2356516114929176e-05, + "loss": 1.6042, + "step": 977 + }, + { + "epoch": 1.8027649769585254, + "grad_norm": 0.2814716100692749, + "learning_rate": 1.2324815921327288e-05, + "loss": 1.6023, + "step": 978 + }, + { + "epoch": 1.8046082949308757, + "grad_norm": 0.2865283191204071, + "learning_rate": 1.2293128063448078e-05, + "loss": 1.5884, + "step": 979 + }, + { + "epoch": 1.8064516129032258, + "grad_norm": 0.26941248774528503, + "learning_rate": 1.2261452687409576e-05, + "loss": 1.603, + "step": 980 + }, + { + "epoch": 1.808294930875576, + "grad_norm": 0.2709951400756836, + "learning_rate": 1.2229789939272253e-05, + "loss": 1.5548, + "step": 981 + }, + { + "epoch": 1.8101382488479263, + "grad_norm": 0.28433525562286377, + "learning_rate": 1.2198139965038356e-05, + "loss": 1.6292, + "step": 982 + }, + { + "epoch": 1.8119815668202763, + "grad_norm": 0.2793513536453247, + "learning_rate": 1.2166502910651232e-05, + "loss": 1.6037, + "step": 983 + }, + { + "epoch": 1.8138248847926266, + "grad_norm": 0.2772260308265686, + "learning_rate": 1.2134878921994634e-05, + "loss": 1.5707, + "step": 984 + }, + { + "epoch": 1.8156682027649769, + "grad_norm": 0.2733345031738281, + "learning_rate": 1.210326814489209e-05, + "loss": 1.5857, + "step": 985 + }, + { + "epoch": 1.8175115207373271, + "grad_norm": 0.27603036165237427, + "learning_rate": 1.2071670725106203e-05, + "loss": 1.533, + "step": 986 + }, + { + "epoch": 1.8193548387096774, + "grad_norm": 0.282755047082901, + "learning_rate": 1.2040086808337965e-05, + "loss": 1.5974, + "step": 987 + }, + { + "epoch": 1.8211981566820277, + "grad_norm": 0.2730749249458313, + "learning_rate": 1.2008516540226115e-05, + "loss": 1.5904, + "step": 988 + }, + { + "epoch": 1.823041474654378, + "grad_norm": 0.28955116868019104, + "learning_rate": 1.1976960066346474e-05, + "loss": 1.5456, + "step": 989 + }, + { + "epoch": 1.8248847926267282, + "grad_norm": 0.2911273241043091, + "learning_rate": 1.194541753221122e-05, + "loss": 1.5873, + "step": 990 + }, + { + "epoch": 1.8267281105990785, + "grad_norm": 0.2708721458911896, + "learning_rate": 1.1913889083268278e-05, + "loss": 1.517, + "step": 991 + }, + { + "epoch": 1.8285714285714287, + "grad_norm": 0.27571648359298706, + "learning_rate": 1.1882374864900616e-05, + "loss": 1.5257, + "step": 992 + }, + { + "epoch": 1.8304147465437788, + "grad_norm": 0.2964298725128174, + "learning_rate": 1.1850875022425587e-05, + "loss": 1.5693, + "step": 993 + }, + { + "epoch": 1.832258064516129, + "grad_norm": 0.2749471664428711, + "learning_rate": 1.1819389701094241e-05, + "loss": 1.5939, + "step": 994 + }, + { + "epoch": 1.8341013824884793, + "grad_norm": 0.27718687057495117, + "learning_rate": 1.1787919046090686e-05, + "loss": 1.501, + "step": 995 + }, + { + "epoch": 1.8359447004608294, + "grad_norm": 0.2840654253959656, + "learning_rate": 1.1756463202531392e-05, + "loss": 1.6314, + "step": 996 + }, + { + "epoch": 1.8377880184331796, + "grad_norm": 0.27605095505714417, + "learning_rate": 1.1725022315464528e-05, + "loss": 1.5555, + "step": 997 + }, + { + "epoch": 1.83963133640553, + "grad_norm": 0.30078697204589844, + "learning_rate": 1.16935965298693e-05, + "loss": 1.6408, + "step": 998 + }, + { + "epoch": 1.8414746543778802, + "grad_norm": 0.2889043092727661, + "learning_rate": 1.1662185990655285e-05, + "loss": 1.5627, + "step": 999 + }, + { + "epoch": 1.8433179723502304, + "grad_norm": 0.2852741777896881, + "learning_rate": 1.1630790842661742e-05, + "loss": 1.5737, + "step": 1000 + }, + { + "epoch": 1.8451612903225807, + "grad_norm": 0.29605633020401, + "learning_rate": 1.1599411230656971e-05, + "loss": 1.6198, + "step": 1001 + }, + { + "epoch": 1.847004608294931, + "grad_norm": 0.3033857047557831, + "learning_rate": 1.1568047299337632e-05, + "loss": 1.5751, + "step": 1002 + }, + { + "epoch": 1.8488479262672812, + "grad_norm": 0.28956925868988037, + "learning_rate": 1.1536699193328063e-05, + "loss": 1.5722, + "step": 1003 + }, + { + "epoch": 1.8506912442396315, + "grad_norm": 0.2798621654510498, + "learning_rate": 1.1505367057179648e-05, + "loss": 1.6136, + "step": 1004 + }, + { + "epoch": 1.8525345622119815, + "grad_norm": 0.280844509601593, + "learning_rate": 1.1474051035370133e-05, + "loss": 1.5622, + "step": 1005 + }, + { + "epoch": 1.8543778801843318, + "grad_norm": 0.2760157585144043, + "learning_rate": 1.1442751272302933e-05, + "loss": 1.5597, + "step": 1006 + }, + { + "epoch": 1.856221198156682, + "grad_norm": 0.27960413694381714, + "learning_rate": 1.1411467912306518e-05, + "loss": 1.5493, + "step": 1007 + }, + { + "epoch": 1.8580645161290321, + "grad_norm": 0.27682289481163025, + "learning_rate": 1.138020109963371e-05, + "loss": 1.5411, + "step": 1008 + }, + { + "epoch": 1.8599078341013824, + "grad_norm": 0.2744944989681244, + "learning_rate": 1.1348950978461023e-05, + "loss": 1.6137, + "step": 1009 + }, + { + "epoch": 1.8617511520737327, + "grad_norm": 0.2708636224269867, + "learning_rate": 1.1317717692888014e-05, + "loss": 1.59, + "step": 1010 + }, + { + "epoch": 1.863594470046083, + "grad_norm": 0.2772144079208374, + "learning_rate": 1.1286501386936606e-05, + "loss": 1.5858, + "step": 1011 + }, + { + "epoch": 1.8654377880184332, + "grad_norm": 0.2765554189682007, + "learning_rate": 1.1255302204550414e-05, + "loss": 1.5399, + "step": 1012 + }, + { + "epoch": 1.8672811059907835, + "grad_norm": 0.2814735770225525, + "learning_rate": 1.1224120289594111e-05, + "loss": 1.5987, + "step": 1013 + }, + { + "epoch": 1.8691244239631337, + "grad_norm": 0.27532437443733215, + "learning_rate": 1.119295578585274e-05, + "loss": 1.6218, + "step": 1014 + }, + { + "epoch": 1.870967741935484, + "grad_norm": 0.27282580733299255, + "learning_rate": 1.1161808837031056e-05, + "loss": 1.5151, + "step": 1015 + }, + { + "epoch": 1.8728110599078343, + "grad_norm": 0.276179701089859, + "learning_rate": 1.1130679586752861e-05, + "loss": 1.5683, + "step": 1016 + }, + { + "epoch": 1.8746543778801843, + "grad_norm": 0.2883111536502838, + "learning_rate": 1.1099568178560356e-05, + "loss": 1.6067, + "step": 1017 + }, + { + "epoch": 1.8764976958525346, + "grad_norm": 0.2828696668148041, + "learning_rate": 1.1068474755913473e-05, + "loss": 1.6124, + "step": 1018 + }, + { + "epoch": 1.8783410138248848, + "grad_norm": 0.28357237577438354, + "learning_rate": 1.1037399462189178e-05, + "loss": 1.6175, + "step": 1019 + }, + { + "epoch": 1.8801843317972349, + "grad_norm": 0.2813222110271454, + "learning_rate": 1.1006342440680885e-05, + "loss": 1.5625, + "step": 1020 + }, + { + "epoch": 1.8820276497695851, + "grad_norm": 0.27178412675857544, + "learning_rate": 1.0975303834597734e-05, + "loss": 1.5548, + "step": 1021 + }, + { + "epoch": 1.8838709677419354, + "grad_norm": 0.2822433114051819, + "learning_rate": 1.0944283787063929e-05, + "loss": 1.5778, + "step": 1022 + }, + { + "epoch": 1.8857142857142857, + "grad_norm": 0.28437986969947815, + "learning_rate": 1.0913282441118123e-05, + "loss": 1.6363, + "step": 1023 + }, + { + "epoch": 1.887557603686636, + "grad_norm": 0.2784495949745178, + "learning_rate": 1.0882299939712727e-05, + "loss": 1.6048, + "step": 1024 + }, + { + "epoch": 1.8894009216589862, + "grad_norm": 0.2737351953983307, + "learning_rate": 1.0851336425713248e-05, + "loss": 1.6136, + "step": 1025 + }, + { + "epoch": 1.8912442396313365, + "grad_norm": 0.2740524113178253, + "learning_rate": 1.0820392041897647e-05, + "loss": 1.5434, + "step": 1026 + }, + { + "epoch": 1.8930875576036867, + "grad_norm": 0.27689987421035767, + "learning_rate": 1.0789466930955672e-05, + "loss": 1.6315, + "step": 1027 + }, + { + "epoch": 1.894930875576037, + "grad_norm": 0.27606692910194397, + "learning_rate": 1.0758561235488196e-05, + "loss": 1.5834, + "step": 1028 + }, + { + "epoch": 1.896774193548387, + "grad_norm": 0.28028401732444763, + "learning_rate": 1.0727675098006568e-05, + "loss": 1.5912, + "step": 1029 + }, + { + "epoch": 1.8986175115207373, + "grad_norm": 0.2821049094200134, + "learning_rate": 1.0696808660931953e-05, + "loss": 1.5692, + "step": 1030 + }, + { + "epoch": 1.9004608294930876, + "grad_norm": 0.2842714786529541, + "learning_rate": 1.0665962066594666e-05, + "loss": 1.5915, + "step": 1031 + }, + { + "epoch": 1.9023041474654376, + "grad_norm": 0.2839188575744629, + "learning_rate": 1.0635135457233533e-05, + "loss": 1.575, + "step": 1032 + }, + { + "epoch": 1.904147465437788, + "grad_norm": 0.2726215124130249, + "learning_rate": 1.0604328974995229e-05, + "loss": 1.5779, + "step": 1033 + }, + { + "epoch": 1.9059907834101382, + "grad_norm": 0.2745780050754547, + "learning_rate": 1.0573542761933596e-05, + "loss": 1.5594, + "step": 1034 + }, + { + "epoch": 1.9078341013824884, + "grad_norm": 0.27262380719184875, + "learning_rate": 1.054277696000904e-05, + "loss": 1.5894, + "step": 1035 + }, + { + "epoch": 1.9096774193548387, + "grad_norm": 0.2791166603565216, + "learning_rate": 1.0512031711087835e-05, + "loss": 1.6442, + "step": 1036 + }, + { + "epoch": 1.911520737327189, + "grad_norm": 0.27580538392066956, + "learning_rate": 1.048130715694149e-05, + "loss": 1.6298, + "step": 1037 + }, + { + "epoch": 1.9133640552995392, + "grad_norm": 0.2800113558769226, + "learning_rate": 1.0450603439246063e-05, + "loss": 1.6536, + "step": 1038 + }, + { + "epoch": 1.9152073732718895, + "grad_norm": 0.2733236253261566, + "learning_rate": 1.0419920699581556e-05, + "loss": 1.5612, + "step": 1039 + }, + { + "epoch": 1.9170506912442398, + "grad_norm": 0.2762746512889862, + "learning_rate": 1.038925907943124e-05, + "loss": 1.4936, + "step": 1040 + }, + { + "epoch": 1.9188940092165898, + "grad_norm": 0.2882503271102905, + "learning_rate": 1.0358618720180975e-05, + "loss": 1.7092, + "step": 1041 + }, + { + "epoch": 1.92073732718894, + "grad_norm": 0.2785802483558655, + "learning_rate": 1.0327999763118607e-05, + "loss": 1.5919, + "step": 1042 + }, + { + "epoch": 1.9225806451612903, + "grad_norm": 0.2711566984653473, + "learning_rate": 1.0297402349433286e-05, + "loss": 1.5795, + "step": 1043 + }, + { + "epoch": 1.9244239631336404, + "grad_norm": 0.2793210744857788, + "learning_rate": 1.0266826620214813e-05, + "loss": 1.5867, + "step": 1044 + }, + { + "epoch": 1.9262672811059907, + "grad_norm": 0.2707485854625702, + "learning_rate": 1.0236272716453012e-05, + "loss": 1.5676, + "step": 1045 + }, + { + "epoch": 1.928110599078341, + "grad_norm": 0.27928751707077026, + "learning_rate": 1.0205740779037065e-05, + "loss": 1.6077, + "step": 1046 + }, + { + "epoch": 1.9299539170506912, + "grad_norm": 0.2875237762928009, + "learning_rate": 1.0175230948754846e-05, + "loss": 1.5721, + "step": 1047 + }, + { + "epoch": 1.9317972350230415, + "grad_norm": 0.27790623903274536, + "learning_rate": 1.014474336629231e-05, + "loss": 1.5684, + "step": 1048 + }, + { + "epoch": 1.9336405529953917, + "grad_norm": 0.2779826521873474, + "learning_rate": 1.011427817223282e-05, + "loss": 1.5567, + "step": 1049 + }, + { + "epoch": 1.935483870967742, + "grad_norm": 0.2960200607776642, + "learning_rate": 1.0083835507056483e-05, + "loss": 1.6219, + "step": 1050 + }, + { + "epoch": 1.9373271889400923, + "grad_norm": 0.2918853759765625, + "learning_rate": 1.0053415511139547e-05, + "loss": 1.608, + "step": 1051 + }, + { + "epoch": 1.9391705069124425, + "grad_norm": 0.27504247426986694, + "learning_rate": 1.0023018324753726e-05, + "loss": 1.6128, + "step": 1052 + }, + { + "epoch": 1.9410138248847926, + "grad_norm": 0.2766924500465393, + "learning_rate": 9.992644088065528e-06, + "loss": 1.626, + "step": 1053 + }, + { + "epoch": 1.9428571428571428, + "grad_norm": 0.27871784567832947, + "learning_rate": 9.96229294113566e-06, + "loss": 1.5861, + "step": 1054 + }, + { + "epoch": 1.944700460829493, + "grad_norm": 0.28291693329811096, + "learning_rate": 9.931965023918369e-06, + "loss": 1.6247, + "step": 1055 + }, + { + "epoch": 1.9465437788018434, + "grad_norm": 0.282185435295105, + "learning_rate": 9.901660476260752e-06, + "loss": 1.6123, + "step": 1056 + }, + { + "epoch": 1.9483870967741934, + "grad_norm": 0.27688685059547424, + "learning_rate": 9.87137943790217e-06, + "loss": 1.6035, + "step": 1057 + }, + { + "epoch": 1.9502304147465437, + "grad_norm": 0.28081604838371277, + "learning_rate": 9.84112204847357e-06, + "loss": 1.5866, + "step": 1058 + }, + { + "epoch": 1.952073732718894, + "grad_norm": 0.29166513681411743, + "learning_rate": 9.810888447496859e-06, + "loss": 1.5301, + "step": 1059 + }, + { + "epoch": 1.9539170506912442, + "grad_norm": 0.30065491795539856, + "learning_rate": 9.780678774384235e-06, + "loss": 1.5863, + "step": 1060 + }, + { + "epoch": 1.9557603686635945, + "grad_norm": 0.2736465632915497, + "learning_rate": 9.750493168437574e-06, + "loss": 1.5563, + "step": 1061 + }, + { + "epoch": 1.9576036866359448, + "grad_norm": 0.2853902280330658, + "learning_rate": 9.720331768847773e-06, + "loss": 1.6323, + "step": 1062 + }, + { + "epoch": 1.959447004608295, + "grad_norm": 0.3043399453163147, + "learning_rate": 9.690194714694101e-06, + "loss": 1.5633, + "step": 1063 + }, + { + "epoch": 1.9612903225806453, + "grad_norm": 0.30382704734802246, + "learning_rate": 9.660082144943571e-06, + "loss": 1.6879, + "step": 1064 + }, + { + "epoch": 1.9631336405529956, + "grad_norm": 0.3026796281337738, + "learning_rate": 9.629994198450305e-06, + "loss": 1.5623, + "step": 1065 + }, + { + "epoch": 1.9649769585253456, + "grad_norm": 0.2774633765220642, + "learning_rate": 9.599931013954858e-06, + "loss": 1.5979, + "step": 1066 + }, + { + "epoch": 1.9668202764976959, + "grad_norm": 0.2874796688556671, + "learning_rate": 9.569892730083631e-06, + "loss": 1.6626, + "step": 1067 + }, + { + "epoch": 1.9686635944700461, + "grad_norm": 0.292472779750824, + "learning_rate": 9.53987948534819e-06, + "loss": 1.572, + "step": 1068 + }, + { + "epoch": 1.9705069124423962, + "grad_norm": 0.3050578236579895, + "learning_rate": 9.509891418144635e-06, + "loss": 1.5863, + "step": 1069 + }, + { + "epoch": 1.9723502304147464, + "grad_norm": 0.2914397418498993, + "learning_rate": 9.479928666752976e-06, + "loss": 1.667, + "step": 1070 + }, + { + "epoch": 1.9741935483870967, + "grad_norm": 0.27456459403038025, + "learning_rate": 9.449991369336502e-06, + "loss": 1.5255, + "step": 1071 + }, + { + "epoch": 1.976036866359447, + "grad_norm": 0.28274887800216675, + "learning_rate": 9.420079663941096e-06, + "loss": 1.6436, + "step": 1072 + }, + { + "epoch": 1.9778801843317972, + "grad_norm": 0.30079808831214905, + "learning_rate": 9.390193688494657e-06, + "loss": 1.6202, + "step": 1073 + }, + { + "epoch": 1.9797235023041475, + "grad_norm": 0.295546293258667, + "learning_rate": 9.360333580806439e-06, + "loss": 1.5797, + "step": 1074 + }, + { + "epoch": 1.9815668202764978, + "grad_norm": 0.27184319496154785, + "learning_rate": 9.330499478566404e-06, + "loss": 1.535, + "step": 1075 + }, + { + "epoch": 1.983410138248848, + "grad_norm": 0.27947860956192017, + "learning_rate": 9.300691519344602e-06, + "loss": 1.5442, + "step": 1076 + }, + { + "epoch": 1.9852534562211983, + "grad_norm": 0.2932865619659424, + "learning_rate": 9.270909840590541e-06, + "loss": 1.5763, + "step": 1077 + }, + { + "epoch": 1.9870967741935484, + "grad_norm": 0.2843061685562134, + "learning_rate": 9.24115457963253e-06, + "loss": 1.5868, + "step": 1078 + }, + { + "epoch": 1.9889400921658986, + "grad_norm": 0.2877620756626129, + "learning_rate": 9.211425873677075e-06, + "loss": 1.5895, + "step": 1079 + }, + { + "epoch": 1.9907834101382489, + "grad_norm": 0.27718016505241394, + "learning_rate": 9.181723859808225e-06, + "loss": 1.5867, + "step": 1080 + }, + { + "epoch": 1.992626728110599, + "grad_norm": 0.27227306365966797, + "learning_rate": 9.152048674986959e-06, + "loss": 1.6022, + "step": 1081 + }, + { + "epoch": 1.9944700460829492, + "grad_norm": 0.29356279969215393, + "learning_rate": 9.122400456050524e-06, + "loss": 1.5695, + "step": 1082 + }, + { + "epoch": 1.9963133640552995, + "grad_norm": 0.2915216386318207, + "learning_rate": 9.092779339711833e-06, + "loss": 1.5648, + "step": 1083 + }, + { + "epoch": 1.9981566820276497, + "grad_norm": 0.3103971779346466, + "learning_rate": 9.063185462558835e-06, + "loss": 1.5823, + "step": 1084 + }, + { + "epoch": 2.0, + "grad_norm": 0.27320027351379395, + "learning_rate": 9.033618961053842e-06, + "loss": 1.6445, + "step": 1085 + }, + { + "epoch": 2.0018433179723503, + "grad_norm": 0.28557664155960083, + "learning_rate": 9.004079971532972e-06, + "loss": 1.5811, + "step": 1086 + }, + { + "epoch": 2.0036866359447005, + "grad_norm": 0.28565630316734314, + "learning_rate": 8.974568630205462e-06, + "loss": 1.6102, + "step": 1087 + }, + { + "epoch": 2.005529953917051, + "grad_norm": 0.27470818161964417, + "learning_rate": 8.945085073153046e-06, + "loss": 1.5613, + "step": 1088 + }, + { + "epoch": 2.007373271889401, + "grad_norm": 0.29731300473213196, + "learning_rate": 8.915629436329362e-06, + "loss": 1.6246, + "step": 1089 + }, + { + "epoch": 2.0092165898617513, + "grad_norm": 0.27910706400871277, + "learning_rate": 8.886201855559294e-06, + "loss": 1.5279, + "step": 1090 + }, + { + "epoch": 2.0110599078341016, + "grad_norm": 0.28342750668525696, + "learning_rate": 8.85680246653835e-06, + "loss": 1.5117, + "step": 1091 + }, + { + "epoch": 2.0129032258064514, + "grad_norm": 0.2792931795120239, + "learning_rate": 8.827431404832048e-06, + "loss": 1.528, + "step": 1092 + }, + { + "epoch": 2.0147465437788017, + "grad_norm": 0.30623000860214233, + "learning_rate": 8.798088805875286e-06, + "loss": 1.5546, + "step": 1093 + }, + { + "epoch": 2.016589861751152, + "grad_norm": 0.30277132987976074, + "learning_rate": 8.768774804971705e-06, + "loss": 1.5741, + "step": 1094 + }, + { + "epoch": 2.0184331797235022, + "grad_norm": 0.27236324548721313, + "learning_rate": 8.739489537293087e-06, + "loss": 1.5027, + "step": 1095 + }, + { + "epoch": 2.0202764976958525, + "grad_norm": 0.27501460909843445, + "learning_rate": 8.710233137878714e-06, + "loss": 1.5391, + "step": 1096 + }, + { + "epoch": 2.0221198156682028, + "grad_norm": 0.27565276622772217, + "learning_rate": 8.681005741634755e-06, + "loss": 1.4878, + "step": 1097 + }, + { + "epoch": 2.023963133640553, + "grad_norm": 0.3000805377960205, + "learning_rate": 8.651807483333627e-06, + "loss": 1.6258, + "step": 1098 + }, + { + "epoch": 2.0258064516129033, + "grad_norm": 0.2959422767162323, + "learning_rate": 8.622638497613415e-06, + "loss": 1.5711, + "step": 1099 + }, + { + "epoch": 2.0276497695852536, + "grad_norm": 0.27250948548316956, + "learning_rate": 8.593498918977196e-06, + "loss": 1.5338, + "step": 1100 + }, + { + "epoch": 2.029493087557604, + "grad_norm": 0.3384530246257782, + "learning_rate": 8.564388881792456e-06, + "loss": 1.5781, + "step": 1101 + }, + { + "epoch": 2.031336405529954, + "grad_norm": 0.27310115098953247, + "learning_rate": 8.53530852029047e-06, + "loss": 1.4911, + "step": 1102 + }, + { + "epoch": 2.0331797235023044, + "grad_norm": 0.2992478311061859, + "learning_rate": 8.506257968565659e-06, + "loss": 1.5759, + "step": 1103 + }, + { + "epoch": 2.035023041474654, + "grad_norm": 0.30011799931526184, + "learning_rate": 8.477237360574984e-06, + "loss": 1.5557, + "step": 1104 + }, + { + "epoch": 2.0368663594470044, + "grad_norm": 0.2770831286907196, + "learning_rate": 8.448246830137355e-06, + "loss": 1.5667, + "step": 1105 + }, + { + "epoch": 2.0387096774193547, + "grad_norm": 0.2708050012588501, + "learning_rate": 8.419286510932961e-06, + "loss": 1.5236, + "step": 1106 + }, + { + "epoch": 2.040552995391705, + "grad_norm": 0.29462921619415283, + "learning_rate": 8.390356536502688e-06, + "loss": 1.5315, + "step": 1107 + }, + { + "epoch": 2.0423963133640552, + "grad_norm": 0.2812652885913849, + "learning_rate": 8.361457040247518e-06, + "loss": 1.5666, + "step": 1108 + }, + { + "epoch": 2.0442396313364055, + "grad_norm": 0.29188141226768494, + "learning_rate": 8.332588155427869e-06, + "loss": 1.5744, + "step": 1109 + }, + { + "epoch": 2.046082949308756, + "grad_norm": 0.280624121427536, + "learning_rate": 8.303750015163008e-06, + "loss": 1.6014, + "step": 1110 + }, + { + "epoch": 2.047926267281106, + "grad_norm": 0.27916669845581055, + "learning_rate": 8.274942752430449e-06, + "loss": 1.5749, + "step": 1111 + }, + { + "epoch": 2.0497695852534563, + "grad_norm": 0.27787718176841736, + "learning_rate": 8.24616650006531e-06, + "loss": 1.5823, + "step": 1112 + }, + { + "epoch": 2.0516129032258066, + "grad_norm": 0.2866472899913788, + "learning_rate": 8.217421390759717e-06, + "loss": 1.5604, + "step": 1113 + }, + { + "epoch": 2.053456221198157, + "grad_norm": 0.2783917188644409, + "learning_rate": 8.188707557062191e-06, + "loss": 1.5849, + "step": 1114 + }, + { + "epoch": 2.055299539170507, + "grad_norm": 0.2850882112979889, + "learning_rate": 8.160025131377044e-06, + "loss": 1.6185, + "step": 1115 + }, + { + "epoch": 2.057142857142857, + "grad_norm": 0.28361815214157104, + "learning_rate": 8.131374245963753e-06, + "loss": 1.5763, + "step": 1116 + }, + { + "epoch": 2.058986175115207, + "grad_norm": 0.27110105752944946, + "learning_rate": 8.102755032936346e-06, + "loss": 1.5421, + "step": 1117 + }, + { + "epoch": 2.0608294930875575, + "grad_norm": 0.28890669345855713, + "learning_rate": 8.074167624262834e-06, + "loss": 1.5398, + "step": 1118 + }, + { + "epoch": 2.0626728110599077, + "grad_norm": 0.2794684171676636, + "learning_rate": 8.045612151764546e-06, + "loss": 1.5596, + "step": 1119 + }, + { + "epoch": 2.064516129032258, + "grad_norm": 0.28180766105651855, + "learning_rate": 8.017088747115554e-06, + "loss": 1.5281, + "step": 1120 + }, + { + "epoch": 2.0663594470046083, + "grad_norm": 0.2749057710170746, + "learning_rate": 7.98859754184207e-06, + "loss": 1.5766, + "step": 1121 + }, + { + "epoch": 2.0682027649769585, + "grad_norm": 0.2896604537963867, + "learning_rate": 7.960138667321822e-06, + "loss": 1.6355, + "step": 1122 + }, + { + "epoch": 2.070046082949309, + "grad_norm": 0.2989213764667511, + "learning_rate": 7.931712254783445e-06, + "loss": 1.6038, + "step": 1123 + }, + { + "epoch": 2.071889400921659, + "grad_norm": 0.27826985716819763, + "learning_rate": 7.903318435305914e-06, + "loss": 1.5439, + "step": 1124 + }, + { + "epoch": 2.0737327188940093, + "grad_norm": 0.2855598032474518, + "learning_rate": 7.874957339817886e-06, + "loss": 1.5385, + "step": 1125 + }, + { + "epoch": 2.0755760368663596, + "grad_norm": 0.28656312823295593, + "learning_rate": 7.846629099097127e-06, + "loss": 1.5975, + "step": 1126 + }, + { + "epoch": 2.07741935483871, + "grad_norm": 0.2890690267086029, + "learning_rate": 7.818333843769917e-06, + "loss": 1.5435, + "step": 1127 + }, + { + "epoch": 2.0792626728110597, + "grad_norm": 0.2903476059436798, + "learning_rate": 7.790071704310422e-06, + "loss": 1.4977, + "step": 1128 + }, + { + "epoch": 2.08110599078341, + "grad_norm": 0.2891685664653778, + "learning_rate": 7.76184281104011e-06, + "loss": 1.5472, + "step": 1129 + }, + { + "epoch": 2.0829493087557602, + "grad_norm": 0.27512073516845703, + "learning_rate": 7.73364729412714e-06, + "loss": 1.5396, + "step": 1130 + }, + { + "epoch": 2.0847926267281105, + "grad_norm": 0.28228333592414856, + "learning_rate": 7.70548528358578e-06, + "loss": 1.5537, + "step": 1131 + }, + { + "epoch": 2.0866359447004608, + "grad_norm": 0.2816321849822998, + "learning_rate": 7.677356909275784e-06, + "loss": 1.5174, + "step": 1132 + }, + { + "epoch": 2.088479262672811, + "grad_norm": 0.28686147928237915, + "learning_rate": 7.649262300901796e-06, + "loss": 1.5525, + "step": 1133 + }, + { + "epoch": 2.0903225806451613, + "grad_norm": 0.28862765431404114, + "learning_rate": 7.621201588012786e-06, + "loss": 1.5841, + "step": 1134 + }, + { + "epoch": 2.0921658986175116, + "grad_norm": 0.2844542860984802, + "learning_rate": 7.593174900001398e-06, + "loss": 1.5042, + "step": 1135 + }, + { + "epoch": 2.094009216589862, + "grad_norm": 0.2845102548599243, + "learning_rate": 7.565182366103391e-06, + "loss": 1.5968, + "step": 1136 + }, + { + "epoch": 2.095852534562212, + "grad_norm": 0.2820180356502533, + "learning_rate": 7.537224115397048e-06, + "loss": 1.5788, + "step": 1137 + }, + { + "epoch": 2.0976958525345624, + "grad_norm": 0.2860027849674225, + "learning_rate": 7.5093002768025485e-06, + "loss": 1.6213, + "step": 1138 + }, + { + "epoch": 2.0995391705069126, + "grad_norm": 0.28183481097221375, + "learning_rate": 7.481410979081389e-06, + "loss": 1.5571, + "step": 1139 + }, + { + "epoch": 2.1013824884792625, + "grad_norm": 0.2824060320854187, + "learning_rate": 7.453556350835821e-06, + "loss": 1.5635, + "step": 1140 + }, + { + "epoch": 2.1032258064516127, + "grad_norm": 0.281578004360199, + "learning_rate": 7.425736520508185e-06, + "loss": 1.5867, + "step": 1141 + }, + { + "epoch": 2.105069124423963, + "grad_norm": 0.2753613293170929, + "learning_rate": 7.397951616380401e-06, + "loss": 1.4974, + "step": 1142 + }, + { + "epoch": 2.1069124423963133, + "grad_norm": 0.2744036912918091, + "learning_rate": 7.370201766573325e-06, + "loss": 1.5638, + "step": 1143 + }, + { + "epoch": 2.1087557603686635, + "grad_norm": 0.27509522438049316, + "learning_rate": 7.342487099046169e-06, + "loss": 1.5507, + "step": 1144 + }, + { + "epoch": 2.110599078341014, + "grad_norm": 0.28957563638687134, + "learning_rate": 7.31480774159591e-06, + "loss": 1.5888, + "step": 1145 + }, + { + "epoch": 2.112442396313364, + "grad_norm": 0.2855337858200073, + "learning_rate": 7.287163821856719e-06, + "loss": 1.5648, + "step": 1146 + }, + { + "epoch": 2.1142857142857143, + "grad_norm": 0.28954070806503296, + "learning_rate": 7.259555467299342e-06, + "loss": 1.6247, + "step": 1147 + }, + { + "epoch": 2.1161290322580646, + "grad_norm": 0.2805940508842468, + "learning_rate": 7.231982805230538e-06, + "loss": 1.5349, + "step": 1148 + }, + { + "epoch": 2.117972350230415, + "grad_norm": 0.2769482731819153, + "learning_rate": 7.204445962792471e-06, + "loss": 1.5549, + "step": 1149 + }, + { + "epoch": 2.119815668202765, + "grad_norm": 0.2799006700515747, + "learning_rate": 7.176945066962152e-06, + "loss": 1.5853, + "step": 1150 + }, + { + "epoch": 2.1216589861751154, + "grad_norm": 0.27686893939971924, + "learning_rate": 7.149480244550822e-06, + "loss": 1.5234, + "step": 1151 + }, + { + "epoch": 2.1235023041474657, + "grad_norm": 0.2784132957458496, + "learning_rate": 7.12205162220338e-06, + "loss": 1.5817, + "step": 1152 + }, + { + "epoch": 2.1253456221198155, + "grad_norm": 0.2762703597545624, + "learning_rate": 7.094659326397818e-06, + "loss": 1.5104, + "step": 1153 + }, + { + "epoch": 2.1271889400921657, + "grad_norm": 0.27806219458580017, + "learning_rate": 7.067303483444603e-06, + "loss": 1.5133, + "step": 1154 + }, + { + "epoch": 2.129032258064516, + "grad_norm": 0.29459184408187866, + "learning_rate": 7.039984219486109e-06, + "loss": 1.5857, + "step": 1155 + }, + { + "epoch": 2.1308755760368663, + "grad_norm": 0.2795344293117523, + "learning_rate": 7.012701660496059e-06, + "loss": 1.5688, + "step": 1156 + }, + { + "epoch": 2.1327188940092165, + "grad_norm": 0.2816448509693146, + "learning_rate": 6.985455932278904e-06, + "loss": 1.5684, + "step": 1157 + }, + { + "epoch": 2.134562211981567, + "grad_norm": 0.2854406535625458, + "learning_rate": 6.958247160469266e-06, + "loss": 1.595, + "step": 1158 + }, + { + "epoch": 2.136405529953917, + "grad_norm": 0.27701088786125183, + "learning_rate": 6.931075470531371e-06, + "loss": 1.5595, + "step": 1159 + }, + { + "epoch": 2.1382488479262673, + "grad_norm": 0.28823423385620117, + "learning_rate": 6.9039409877584195e-06, + "loss": 1.5561, + "step": 1160 + }, + { + "epoch": 2.1400921658986176, + "grad_norm": 0.2958308160305023, + "learning_rate": 6.876843837272075e-06, + "loss": 1.5711, + "step": 1161 + }, + { + "epoch": 2.141935483870968, + "grad_norm": 0.27960205078125, + "learning_rate": 6.849784144021859e-06, + "loss": 1.572, + "step": 1162 + }, + { + "epoch": 2.143778801843318, + "grad_norm": 0.27746549248695374, + "learning_rate": 6.82276203278453e-06, + "loss": 1.4994, + "step": 1163 + }, + { + "epoch": 2.145622119815668, + "grad_norm": 0.28397759795188904, + "learning_rate": 6.795777628163599e-06, + "loss": 1.5317, + "step": 1164 + }, + { + "epoch": 2.1474654377880182, + "grad_norm": 0.286499947309494, + "learning_rate": 6.7688310545886715e-06, + "loss": 1.568, + "step": 1165 + }, + { + "epoch": 2.1493087557603685, + "grad_norm": 0.28141286969184875, + "learning_rate": 6.741922436314929e-06, + "loss": 1.4829, + "step": 1166 + }, + { + "epoch": 2.1511520737327188, + "grad_norm": 0.2864668667316437, + "learning_rate": 6.715051897422523e-06, + "loss": 1.5274, + "step": 1167 + }, + { + "epoch": 2.152995391705069, + "grad_norm": 0.28489890694618225, + "learning_rate": 6.688219561816008e-06, + "loss": 1.556, + "step": 1168 + }, + { + "epoch": 2.1548387096774193, + "grad_norm": 0.28669190406799316, + "learning_rate": 6.661425553223799e-06, + "loss": 1.5923, + "step": 1169 + }, + { + "epoch": 2.1566820276497696, + "grad_norm": 0.28376609086990356, + "learning_rate": 6.634669995197561e-06, + "loss": 1.5599, + "step": 1170 + }, + { + "epoch": 2.15852534562212, + "grad_norm": 0.2903115451335907, + "learning_rate": 6.607953011111655e-06, + "loss": 1.6186, + "step": 1171 + }, + { + "epoch": 2.16036866359447, + "grad_norm": 0.2769109010696411, + "learning_rate": 6.581274724162587e-06, + "loss": 1.5157, + "step": 1172 + }, + { + "epoch": 2.1622119815668204, + "grad_norm": 0.28724145889282227, + "learning_rate": 6.554635257368411e-06, + "loss": 1.5565, + "step": 1173 + }, + { + "epoch": 2.1640552995391706, + "grad_norm": 0.29872408509254456, + "learning_rate": 6.528034733568174e-06, + "loss": 1.599, + "step": 1174 + }, + { + "epoch": 2.165898617511521, + "grad_norm": 0.2927846908569336, + "learning_rate": 6.50147327542137e-06, + "loss": 1.5542, + "step": 1175 + }, + { + "epoch": 2.167741935483871, + "grad_norm": 0.27771615982055664, + "learning_rate": 6.474951005407317e-06, + "loss": 1.5592, + "step": 1176 + }, + { + "epoch": 2.169585253456221, + "grad_norm": 0.2794598937034607, + "learning_rate": 6.448468045824664e-06, + "loss": 1.526, + "step": 1177 + }, + { + "epoch": 2.1714285714285713, + "grad_norm": 0.2869041860103607, + "learning_rate": 6.4220245187907915e-06, + "loss": 1.6018, + "step": 1178 + }, + { + "epoch": 2.1732718894009215, + "grad_norm": 0.28667303919792175, + "learning_rate": 6.395620546241221e-06, + "loss": 1.6524, + "step": 1179 + }, + { + "epoch": 2.175115207373272, + "grad_norm": 0.2848053574562073, + "learning_rate": 6.369256249929112e-06, + "loss": 1.5206, + "step": 1180 + }, + { + "epoch": 2.176958525345622, + "grad_norm": 0.28304025530815125, + "learning_rate": 6.342931751424656e-06, + "loss": 1.5366, + "step": 1181 + }, + { + "epoch": 2.1788018433179723, + "grad_norm": 0.28419801592826843, + "learning_rate": 6.316647172114529e-06, + "loss": 1.5498, + "step": 1182 + }, + { + "epoch": 2.1806451612903226, + "grad_norm": 0.2827412188053131, + "learning_rate": 6.2904026332013445e-06, + "loss": 1.5518, + "step": 1183 + }, + { + "epoch": 2.182488479262673, + "grad_norm": 0.28467991948127747, + "learning_rate": 6.264198255703071e-06, + "loss": 1.5568, + "step": 1184 + }, + { + "epoch": 2.184331797235023, + "grad_norm": 0.2874020040035248, + "learning_rate": 6.238034160452486e-06, + "loss": 1.5876, + "step": 1185 + }, + { + "epoch": 2.1861751152073734, + "grad_norm": 0.2829265594482422, + "learning_rate": 6.211910468096631e-06, + "loss": 1.5558, + "step": 1186 + }, + { + "epoch": 2.1880184331797237, + "grad_norm": 0.2792380750179291, + "learning_rate": 6.185827299096226e-06, + "loss": 1.5382, + "step": 1187 + }, + { + "epoch": 2.189861751152074, + "grad_norm": 0.2878943979740143, + "learning_rate": 6.1597847737251504e-06, + "loss": 1.5488, + "step": 1188 + }, + { + "epoch": 2.191705069124424, + "grad_norm": 0.28687000274658203, + "learning_rate": 6.133783012069853e-06, + "loss": 1.5798, + "step": 1189 + }, + { + "epoch": 2.193548387096774, + "grad_norm": 0.28751304745674133, + "learning_rate": 6.1078221340288155e-06, + "loss": 1.6048, + "step": 1190 + }, + { + "epoch": 2.1953917050691243, + "grad_norm": 0.28695595264434814, + "learning_rate": 6.081902259312013e-06, + "loss": 1.6293, + "step": 1191 + }, + { + "epoch": 2.1972350230414746, + "grad_norm": 0.2811322510242462, + "learning_rate": 6.05602350744033e-06, + "loss": 1.5452, + "step": 1192 + }, + { + "epoch": 2.199078341013825, + "grad_norm": 0.2896054685115814, + "learning_rate": 6.030185997745031e-06, + "loss": 1.5874, + "step": 1193 + }, + { + "epoch": 2.200921658986175, + "grad_norm": 0.2827315926551819, + "learning_rate": 6.004389849367223e-06, + "loss": 1.5839, + "step": 1194 + }, + { + "epoch": 2.2027649769585254, + "grad_norm": 0.2890084683895111, + "learning_rate": 5.978635181257254e-06, + "loss": 1.5874, + "step": 1195 + }, + { + "epoch": 2.2046082949308756, + "grad_norm": 0.2855392396450043, + "learning_rate": 5.952922112174231e-06, + "loss": 1.5515, + "step": 1196 + }, + { + "epoch": 2.206451612903226, + "grad_norm": 0.28122851252555847, + "learning_rate": 5.927250760685441e-06, + "loss": 1.573, + "step": 1197 + }, + { + "epoch": 2.208294930875576, + "grad_norm": 0.2914394736289978, + "learning_rate": 5.901621245165772e-06, + "loss": 1.5727, + "step": 1198 + }, + { + "epoch": 2.2101382488479264, + "grad_norm": 0.2777380645275116, + "learning_rate": 5.8760336837972355e-06, + "loss": 1.5166, + "step": 1199 + }, + { + "epoch": 2.2119815668202767, + "grad_norm": 0.2740841507911682, + "learning_rate": 5.850488194568366e-06, + "loss": 1.4895, + "step": 1200 + }, + { + "epoch": 2.2138248847926265, + "grad_norm": 0.2903701961040497, + "learning_rate": 5.824984895273697e-06, + "loss": 1.6212, + "step": 1201 + }, + { + "epoch": 2.215668202764977, + "grad_norm": 0.28629857301712036, + "learning_rate": 5.799523903513228e-06, + "loss": 1.5795, + "step": 1202 + }, + { + "epoch": 2.217511520737327, + "grad_norm": 0.326154500246048, + "learning_rate": 5.774105336691861e-06, + "loss": 1.5421, + "step": 1203 + }, + { + "epoch": 2.2193548387096773, + "grad_norm": 0.2828398048877716, + "learning_rate": 5.748729312018869e-06, + "loss": 1.5533, + "step": 1204 + }, + { + "epoch": 2.2211981566820276, + "grad_norm": 0.28595855832099915, + "learning_rate": 5.723395946507367e-06, + "loss": 1.5814, + "step": 1205 + }, + { + "epoch": 2.223041474654378, + "grad_norm": 0.2907951772212982, + "learning_rate": 5.6981053569737525e-06, + "loss": 1.5671, + "step": 1206 + }, + { + "epoch": 2.224884792626728, + "grad_norm": 0.2874442934989929, + "learning_rate": 5.67285766003717e-06, + "loss": 1.6061, + "step": 1207 + }, + { + "epoch": 2.2267281105990784, + "grad_norm": 0.2877340018749237, + "learning_rate": 5.647652972118998e-06, + "loss": 1.6042, + "step": 1208 + }, + { + "epoch": 2.2285714285714286, + "grad_norm": 0.2804166078567505, + "learning_rate": 5.622491409442272e-06, + "loss": 1.4541, + "step": 1209 + }, + { + "epoch": 2.230414746543779, + "grad_norm": 0.28799667954444885, + "learning_rate": 5.597373088031193e-06, + "loss": 1.6308, + "step": 1210 + }, + { + "epoch": 2.232258064516129, + "grad_norm": 0.2771644592285156, + "learning_rate": 5.572298123710536e-06, + "loss": 1.5414, + "step": 1211 + }, + { + "epoch": 2.2341013824884794, + "grad_norm": 0.2859271168708801, + "learning_rate": 5.54726663210518e-06, + "loss": 1.61, + "step": 1212 + }, + { + "epoch": 2.2359447004608297, + "grad_norm": 0.2863786816596985, + "learning_rate": 5.522278728639544e-06, + "loss": 1.5563, + "step": 1213 + }, + { + "epoch": 2.2377880184331795, + "grad_norm": 0.28577443957328796, + "learning_rate": 5.497334528537022e-06, + "loss": 1.4964, + "step": 1214 + }, + { + "epoch": 2.23963133640553, + "grad_norm": 0.2836892008781433, + "learning_rate": 5.472434146819522e-06, + "loss": 1.5857, + "step": 1215 + }, + { + "epoch": 2.24147465437788, + "grad_norm": 0.2816178500652313, + "learning_rate": 5.447577698306876e-06, + "loss": 1.6189, + "step": 1216 + }, + { + "epoch": 2.2433179723502303, + "grad_norm": 0.2896072566509247, + "learning_rate": 5.422765297616336e-06, + "loss": 1.6145, + "step": 1217 + }, + { + "epoch": 2.2451612903225806, + "grad_norm": 0.2897859513759613, + "learning_rate": 5.3979970591620555e-06, + "loss": 1.554, + "step": 1218 + }, + { + "epoch": 2.247004608294931, + "grad_norm": 0.29543671011924744, + "learning_rate": 5.37327309715453e-06, + "loss": 1.5852, + "step": 1219 + }, + { + "epoch": 2.248847926267281, + "grad_norm": 0.28915709257125854, + "learning_rate": 5.348593525600093e-06, + "loss": 1.5264, + "step": 1220 + }, + { + "epoch": 2.2506912442396314, + "grad_norm": 0.2880646288394928, + "learning_rate": 5.323958458300403e-06, + "loss": 1.4948, + "step": 1221 + }, + { + "epoch": 2.2525345622119817, + "grad_norm": 0.2773268222808838, + "learning_rate": 5.29936800885188e-06, + "loss": 1.4826, + "step": 1222 + }, + { + "epoch": 2.254377880184332, + "grad_norm": 0.27953988313674927, + "learning_rate": 5.2748222906452105e-06, + "loss": 1.5751, + "step": 1223 + }, + { + "epoch": 2.256221198156682, + "grad_norm": 0.27911481261253357, + "learning_rate": 5.250321416864828e-06, + "loss": 1.5187, + "step": 1224 + }, + { + "epoch": 2.258064516129032, + "grad_norm": 0.28624942898750305, + "learning_rate": 5.2258655004883696e-06, + "loss": 1.5599, + "step": 1225 + }, + { + "epoch": 2.2599078341013823, + "grad_norm": 0.2925635278224945, + "learning_rate": 5.201454654286166e-06, + "loss": 1.5331, + "step": 1226 + }, + { + "epoch": 2.2617511520737326, + "grad_norm": 0.3006371855735779, + "learning_rate": 5.177088990820725e-06, + "loss": 1.6149, + "step": 1227 + }, + { + "epoch": 2.263594470046083, + "grad_norm": 0.2926720976829529, + "learning_rate": 5.152768622446211e-06, + "loss": 1.5913, + "step": 1228 + }, + { + "epoch": 2.265437788018433, + "grad_norm": 0.28473854064941406, + "learning_rate": 5.128493661307934e-06, + "loss": 1.569, + "step": 1229 + }, + { + "epoch": 2.2672811059907834, + "grad_norm": 0.2832511067390442, + "learning_rate": 5.104264219341793e-06, + "loss": 1.611, + "step": 1230 + }, + { + "epoch": 2.2691244239631336, + "grad_norm": 0.28466713428497314, + "learning_rate": 5.080080408273821e-06, + "loss": 1.5826, + "step": 1231 + }, + { + "epoch": 2.270967741935484, + "grad_norm": 0.2919696569442749, + "learning_rate": 5.055942339619639e-06, + "loss": 1.5438, + "step": 1232 + }, + { + "epoch": 2.272811059907834, + "grad_norm": 0.29388782382011414, + "learning_rate": 5.031850124683913e-06, + "loss": 1.6009, + "step": 1233 + }, + { + "epoch": 2.2746543778801844, + "grad_norm": 0.287203311920166, + "learning_rate": 5.0078038745599e-06, + "loss": 1.5824, + "step": 1234 + }, + { + "epoch": 2.2764976958525347, + "grad_norm": 0.28717392683029175, + "learning_rate": 4.983803700128893e-06, + "loss": 1.5522, + "step": 1235 + }, + { + "epoch": 2.278341013824885, + "grad_norm": 0.28208932280540466, + "learning_rate": 4.959849712059716e-06, + "loss": 1.5621, + "step": 1236 + }, + { + "epoch": 2.2801843317972352, + "grad_norm": 0.2900165915489197, + "learning_rate": 4.935942020808239e-06, + "loss": 1.6181, + "step": 1237 + }, + { + "epoch": 2.282027649769585, + "grad_norm": 0.292563796043396, + "learning_rate": 4.912080736616833e-06, + "loss": 1.5769, + "step": 1238 + }, + { + "epoch": 2.2838709677419353, + "grad_norm": 0.2846229374408722, + "learning_rate": 4.888265969513876e-06, + "loss": 1.5911, + "step": 1239 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 0.29053792357444763, + "learning_rate": 4.864497829313269e-06, + "loss": 1.6015, + "step": 1240 + }, + { + "epoch": 2.287557603686636, + "grad_norm": 0.2835329473018646, + "learning_rate": 4.840776425613887e-06, + "loss": 1.561, + "step": 1241 + }, + { + "epoch": 2.289400921658986, + "grad_norm": 0.2829905152320862, + "learning_rate": 4.817101867799097e-06, + "loss": 1.5655, + "step": 1242 + }, + { + "epoch": 2.2912442396313364, + "grad_norm": 0.29234424233436584, + "learning_rate": 4.793474265036272e-06, + "loss": 1.604, + "step": 1243 + }, + { + "epoch": 2.2930875576036867, + "grad_norm": 0.29347386956214905, + "learning_rate": 4.769893726276243e-06, + "loss": 1.6451, + "step": 1244 + }, + { + "epoch": 2.294930875576037, + "grad_norm": 0.2967973053455353, + "learning_rate": 4.746360360252834e-06, + "loss": 1.6059, + "step": 1245 + }, + { + "epoch": 2.296774193548387, + "grad_norm": 0.2842494547367096, + "learning_rate": 4.722874275482338e-06, + "loss": 1.5981, + "step": 1246 + }, + { + "epoch": 2.2986175115207375, + "grad_norm": 0.28634604811668396, + "learning_rate": 4.699435580263044e-06, + "loss": 1.5291, + "step": 1247 + }, + { + "epoch": 2.3004608294930877, + "grad_norm": 0.27907440066337585, + "learning_rate": 4.676044382674702e-06, + "loss": 1.4955, + "step": 1248 + }, + { + "epoch": 2.3023041474654375, + "grad_norm": 0.2886315882205963, + "learning_rate": 4.652700790578047e-06, + "loss": 1.5537, + "step": 1249 + }, + { + "epoch": 2.3041474654377883, + "grad_norm": 0.28801992535591125, + "learning_rate": 4.629404911614306e-06, + "loss": 1.53, + "step": 1250 + }, + { + "epoch": 2.305990783410138, + "grad_norm": 0.2892250716686249, + "learning_rate": 4.606156853204682e-06, + "loss": 1.5997, + "step": 1251 + }, + { + "epoch": 2.3078341013824883, + "grad_norm": 0.28449535369873047, + "learning_rate": 4.5829567225498696e-06, + "loss": 1.5487, + "step": 1252 + }, + { + "epoch": 2.3096774193548386, + "grad_norm": 0.29257315397262573, + "learning_rate": 4.559804626629574e-06, + "loss": 1.5927, + "step": 1253 + }, + { + "epoch": 2.311520737327189, + "grad_norm": 0.28220823407173157, + "learning_rate": 4.536700672201987e-06, + "loss": 1.5277, + "step": 1254 + }, + { + "epoch": 2.313364055299539, + "grad_norm": 0.2863527238368988, + "learning_rate": 4.513644965803316e-06, + "loss": 1.5356, + "step": 1255 + }, + { + "epoch": 2.3152073732718894, + "grad_norm": 0.28745314478874207, + "learning_rate": 4.490637613747301e-06, + "loss": 1.5588, + "step": 1256 + }, + { + "epoch": 2.3170506912442397, + "grad_norm": 0.2863119840621948, + "learning_rate": 4.4676787221247e-06, + "loss": 1.5853, + "step": 1257 + }, + { + "epoch": 2.31889400921659, + "grad_norm": 0.29094186425209045, + "learning_rate": 4.444768396802808e-06, + "loss": 1.5947, + "step": 1258 + }, + { + "epoch": 2.32073732718894, + "grad_norm": 0.27771443128585815, + "learning_rate": 4.421906743424989e-06, + "loss": 1.5098, + "step": 1259 + }, + { + "epoch": 2.3225806451612905, + "grad_norm": 0.2965877950191498, + "learning_rate": 4.39909386741016e-06, + "loss": 1.5564, + "step": 1260 + }, + { + "epoch": 2.3244239631336407, + "grad_norm": 0.2877909243106842, + "learning_rate": 4.376329873952317e-06, + "loss": 1.579, + "step": 1261 + }, + { + "epoch": 2.3262672811059906, + "grad_norm": 0.29187557101249695, + "learning_rate": 4.353614868020051e-06, + "loss": 1.5967, + "step": 1262 + }, + { + "epoch": 2.328110599078341, + "grad_norm": 0.28522375226020813, + "learning_rate": 4.330948954356076e-06, + "loss": 1.6107, + "step": 1263 + }, + { + "epoch": 2.329953917050691, + "grad_norm": 0.28206923604011536, + "learning_rate": 4.308332237476717e-06, + "loss": 1.5283, + "step": 1264 + }, + { + "epoch": 2.3317972350230414, + "grad_norm": 0.29037174582481384, + "learning_rate": 4.285764821671446e-06, + "loss": 1.5784, + "step": 1265 + }, + { + "epoch": 2.3336405529953916, + "grad_norm": 0.28854188323020935, + "learning_rate": 4.263246811002414e-06, + "loss": 1.6278, + "step": 1266 + }, + { + "epoch": 2.335483870967742, + "grad_norm": 0.2835865318775177, + "learning_rate": 4.240778309303942e-06, + "loss": 1.5784, + "step": 1267 + }, + { + "epoch": 2.337327188940092, + "grad_norm": 0.28313156962394714, + "learning_rate": 4.218359420182055e-06, + "loss": 1.5651, + "step": 1268 + }, + { + "epoch": 2.3391705069124424, + "grad_norm": 0.288749635219574, + "learning_rate": 4.195990247014025e-06, + "loss": 1.5131, + "step": 1269 + }, + { + "epoch": 2.3410138248847927, + "grad_norm": 0.2896270453929901, + "learning_rate": 4.173670892947858e-06, + "loss": 1.6161, + "step": 1270 + }, + { + "epoch": 2.342857142857143, + "grad_norm": 0.29257744550704956, + "learning_rate": 4.151401460901833e-06, + "loss": 1.5564, + "step": 1271 + }, + { + "epoch": 2.3447004608294932, + "grad_norm": 0.27926334738731384, + "learning_rate": 4.1291820535640505e-06, + "loss": 1.547, + "step": 1272 + }, + { + "epoch": 2.3465437788018435, + "grad_norm": 0.2781253457069397, + "learning_rate": 4.107012773391918e-06, + "loss": 1.5216, + "step": 1273 + }, + { + "epoch": 2.3483870967741938, + "grad_norm": 0.2872128486633301, + "learning_rate": 4.084893722611706e-06, + "loss": 1.551, + "step": 1274 + }, + { + "epoch": 2.3502304147465436, + "grad_norm": 0.2819628119468689, + "learning_rate": 4.062825003218075e-06, + "loss": 1.5346, + "step": 1275 + }, + { + "epoch": 2.352073732718894, + "grad_norm": 0.2876599133014679, + "learning_rate": 4.04080671697359e-06, + "loss": 1.5328, + "step": 1276 + }, + { + "epoch": 2.353917050691244, + "grad_norm": 0.2796885669231415, + "learning_rate": 4.018838965408259e-06, + "loss": 1.5464, + "step": 1277 + }, + { + "epoch": 2.3557603686635944, + "grad_norm": 0.2773331105709076, + "learning_rate": 3.99692184981908e-06, + "loss": 1.4875, + "step": 1278 + }, + { + "epoch": 2.3576036866359447, + "grad_norm": 0.2896631062030792, + "learning_rate": 3.975055471269545e-06, + "loss": 1.5394, + "step": 1279 + }, + { + "epoch": 2.359447004608295, + "grad_norm": 0.29722583293914795, + "learning_rate": 3.953239930589196e-06, + "loss": 1.6278, + "step": 1280 + }, + { + "epoch": 2.361290322580645, + "grad_norm": 0.29389339685440063, + "learning_rate": 3.931475328373145e-06, + "loss": 1.5823, + "step": 1281 + }, + { + "epoch": 2.3631336405529955, + "grad_norm": 0.2862755060195923, + "learning_rate": 3.909761764981637e-06, + "loss": 1.5393, + "step": 1282 + }, + { + "epoch": 2.3649769585253457, + "grad_norm": 0.2834300100803375, + "learning_rate": 3.888099340539548e-06, + "loss": 1.5612, + "step": 1283 + }, + { + "epoch": 2.366820276497696, + "grad_norm": 0.2745205760002136, + "learning_rate": 3.866488154935951e-06, + "loss": 1.5021, + "step": 1284 + }, + { + "epoch": 2.3686635944700463, + "grad_norm": 0.27791908383369446, + "learning_rate": 3.844928307823655e-06, + "loss": 1.59, + "step": 1285 + }, + { + "epoch": 2.370506912442396, + "grad_norm": 0.2886483669281006, + "learning_rate": 3.823419898618733e-06, + "loss": 1.5734, + "step": 1286 + }, + { + "epoch": 2.3723502304147464, + "grad_norm": 0.28079813718795776, + "learning_rate": 3.801963026500058e-06, + "loss": 1.5254, + "step": 1287 + }, + { + "epoch": 2.3741935483870966, + "grad_norm": 0.28942209482192993, + "learning_rate": 3.7805577904088817e-06, + "loss": 1.5835, + "step": 1288 + }, + { + "epoch": 2.376036866359447, + "grad_norm": 0.2860458195209503, + "learning_rate": 3.7592042890483335e-06, + "loss": 1.5936, + "step": 1289 + }, + { + "epoch": 2.377880184331797, + "grad_norm": 0.2921033203601837, + "learning_rate": 3.7379026208829865e-06, + "loss": 1.6293, + "step": 1290 + }, + { + "epoch": 2.3797235023041474, + "grad_norm": 0.2822112739086151, + "learning_rate": 3.7166528841384197e-06, + "loss": 1.5031, + "step": 1291 + }, + { + "epoch": 2.3815668202764977, + "grad_norm": 0.2924329936504364, + "learning_rate": 3.695455176800719e-06, + "loss": 1.552, + "step": 1292 + }, + { + "epoch": 2.383410138248848, + "grad_norm": 0.29340824484825134, + "learning_rate": 3.6743095966160773e-06, + "loss": 1.628, + "step": 1293 + }, + { + "epoch": 2.385253456221198, + "grad_norm": 0.2875743508338928, + "learning_rate": 3.6532162410903165e-06, + "loss": 1.4956, + "step": 1294 + }, + { + "epoch": 2.3870967741935485, + "grad_norm": 0.2868419289588928, + "learning_rate": 3.6321752074884374e-06, + "loss": 1.5754, + "step": 1295 + }, + { + "epoch": 2.3889400921658988, + "grad_norm": 0.28470683097839355, + "learning_rate": 3.6111865928341723e-06, + "loss": 1.5562, + "step": 1296 + }, + { + "epoch": 2.390783410138249, + "grad_norm": 0.2839842736721039, + "learning_rate": 3.5902504939095444e-06, + "loss": 1.5849, + "step": 1297 + }, + { + "epoch": 2.3926267281105993, + "grad_norm": 0.2892094552516937, + "learning_rate": 3.5693670072544253e-06, + "loss": 1.6061, + "step": 1298 + }, + { + "epoch": 2.394470046082949, + "grad_norm": 0.29173600673675537, + "learning_rate": 3.5485362291660727e-06, + "loss": 1.5809, + "step": 1299 + }, + { + "epoch": 2.3963133640552994, + "grad_norm": 0.28074854612350464, + "learning_rate": 3.527758255698696e-06, + "loss": 1.5568, + "step": 1300 + }, + { + "epoch": 2.3981566820276496, + "grad_norm": 0.2950586974620819, + "learning_rate": 3.507033182663026e-06, + "loss": 1.6425, + "step": 1301 + }, + { + "epoch": 2.4, + "grad_norm": 0.29268449544906616, + "learning_rate": 3.4863611056258456e-06, + "loss": 1.5819, + "step": 1302 + }, + { + "epoch": 2.40184331797235, + "grad_norm": 0.290112167596817, + "learning_rate": 3.465742119909568e-06, + "loss": 1.608, + "step": 1303 + }, + { + "epoch": 2.4036866359447004, + "grad_norm": 0.2851385474205017, + "learning_rate": 3.445176320591806e-06, + "loss": 1.5662, + "step": 1304 + }, + { + "epoch": 2.4055299539170507, + "grad_norm": 0.283454030752182, + "learning_rate": 3.4246638025049043e-06, + "loss": 1.5217, + "step": 1305 + }, + { + "epoch": 2.407373271889401, + "grad_norm": 0.28823140263557434, + "learning_rate": 3.4042046602355238e-06, + "loss": 1.593, + "step": 1306 + }, + { + "epoch": 2.4092165898617512, + "grad_norm": 0.28793561458587646, + "learning_rate": 3.3837989881242142e-06, + "loss": 1.5398, + "step": 1307 + }, + { + "epoch": 2.4110599078341015, + "grad_norm": 0.2885957956314087, + "learning_rate": 3.363446880264937e-06, + "loss": 1.6315, + "step": 1308 + }, + { + "epoch": 2.412903225806452, + "grad_norm": 0.2905697226524353, + "learning_rate": 3.3431484305046838e-06, + "loss": 1.5792, + "step": 1309 + }, + { + "epoch": 2.4147465437788016, + "grad_norm": 0.28679245710372925, + "learning_rate": 3.3229037324430228e-06, + "loss": 1.5373, + "step": 1310 + }, + { + "epoch": 2.4165898617511523, + "grad_norm": 0.2853943407535553, + "learning_rate": 3.3027128794316353e-06, + "loss": 1.6078, + "step": 1311 + }, + { + "epoch": 2.418433179723502, + "grad_norm": 0.2869051694869995, + "learning_rate": 3.282575964573943e-06, + "loss": 1.5733, + "step": 1312 + }, + { + "epoch": 2.4202764976958524, + "grad_norm": 0.29146990180015564, + "learning_rate": 3.2624930807246443e-06, + "loss": 1.5982, + "step": 1313 + }, + { + "epoch": 2.4221198156682027, + "grad_norm": 0.28490564227104187, + "learning_rate": 3.2424643204892734e-06, + "loss": 1.5393, + "step": 1314 + }, + { + "epoch": 2.423963133640553, + "grad_norm": 0.28606873750686646, + "learning_rate": 3.2224897762238143e-06, + "loss": 1.5362, + "step": 1315 + }, + { + "epoch": 2.425806451612903, + "grad_norm": 0.2890234589576721, + "learning_rate": 3.2025695400342346e-06, + "loss": 1.5838, + "step": 1316 + }, + { + "epoch": 2.4276497695852535, + "grad_norm": 0.2757675349712372, + "learning_rate": 3.1827037037760965e-06, + "loss": 1.4825, + "step": 1317 + }, + { + "epoch": 2.4294930875576037, + "grad_norm": 0.28645581007003784, + "learning_rate": 3.162892359054098e-06, + "loss": 1.6039, + "step": 1318 + }, + { + "epoch": 2.431336405529954, + "grad_norm": 0.28461480140686035, + "learning_rate": 3.1431355972216697e-06, + "loss": 1.568, + "step": 1319 + }, + { + "epoch": 2.4331797235023043, + "grad_norm": 0.28354039788246155, + "learning_rate": 3.1234335093805655e-06, + "loss": 1.5735, + "step": 1320 + }, + { + "epoch": 2.4350230414746545, + "grad_norm": 0.2865184545516968, + "learning_rate": 3.1037861863804117e-06, + "loss": 1.5878, + "step": 1321 + }, + { + "epoch": 2.436866359447005, + "grad_norm": 0.28300338983535767, + "learning_rate": 3.08419371881831e-06, + "loss": 1.5453, + "step": 1322 + }, + { + "epoch": 2.4387096774193546, + "grad_norm": 0.29105615615844727, + "learning_rate": 3.0646561970384226e-06, + "loss": 1.5695, + "step": 1323 + }, + { + "epoch": 2.440552995391705, + "grad_norm": 0.2864762544631958, + "learning_rate": 3.045173711131537e-06, + "loss": 1.571, + "step": 1324 + }, + { + "epoch": 2.442396313364055, + "grad_norm": 0.2834813892841339, + "learning_rate": 3.02574635093466e-06, + "loss": 1.5779, + "step": 1325 + }, + { + "epoch": 2.4442396313364054, + "grad_norm": 0.29010024666786194, + "learning_rate": 3.0063742060306227e-06, + "loss": 1.5737, + "step": 1326 + }, + { + "epoch": 2.4460829493087557, + "grad_norm": 0.2904694080352783, + "learning_rate": 2.9870573657476196e-06, + "loss": 1.6337, + "step": 1327 + }, + { + "epoch": 2.447926267281106, + "grad_norm": 0.28722578287124634, + "learning_rate": 2.96779591915885e-06, + "loss": 1.5565, + "step": 1328 + }, + { + "epoch": 2.4497695852534562, + "grad_norm": 0.28688907623291016, + "learning_rate": 2.948589955082085e-06, + "loss": 1.6058, + "step": 1329 + }, + { + "epoch": 2.4516129032258065, + "grad_norm": 0.2821052670478821, + "learning_rate": 2.9294395620792306e-06, + "loss": 1.5153, + "step": 1330 + }, + { + "epoch": 2.4534562211981568, + "grad_norm": 0.2884730100631714, + "learning_rate": 2.910344828455975e-06, + "loss": 1.5567, + "step": 1331 + }, + { + "epoch": 2.455299539170507, + "grad_norm": 0.291220486164093, + "learning_rate": 2.8913058422613363e-06, + "loss": 1.5801, + "step": 1332 + }, + { + "epoch": 2.4571428571428573, + "grad_norm": 0.29171690344810486, + "learning_rate": 2.872322691287268e-06, + "loss": 1.5937, + "step": 1333 + }, + { + "epoch": 2.4589861751152076, + "grad_norm": 0.2891831398010254, + "learning_rate": 2.8533954630682728e-06, + "loss": 1.6024, + "step": 1334 + }, + { + "epoch": 2.460829493087558, + "grad_norm": 0.288980633020401, + "learning_rate": 2.834524244880974e-06, + "loss": 1.6058, + "step": 1335 + }, + { + "epoch": 2.4626728110599077, + "grad_norm": 0.2906559407711029, + "learning_rate": 2.81570912374372e-06, + "loss": 1.598, + "step": 1336 + }, + { + "epoch": 2.464516129032258, + "grad_norm": 0.28637051582336426, + "learning_rate": 2.796950186416199e-06, + "loss": 1.5603, + "step": 1337 + }, + { + "epoch": 2.466359447004608, + "grad_norm": 0.28527313470840454, + "learning_rate": 2.778247519399011e-06, + "loss": 1.5753, + "step": 1338 + }, + { + "epoch": 2.4682027649769585, + "grad_norm": 0.2928375005722046, + "learning_rate": 2.7596012089333015e-06, + "loss": 1.545, + "step": 1339 + }, + { + "epoch": 2.4700460829493087, + "grad_norm": 0.28446659445762634, + "learning_rate": 2.74101134100033e-06, + "loss": 1.5653, + "step": 1340 + }, + { + "epoch": 2.471889400921659, + "grad_norm": 0.2910979092121124, + "learning_rate": 2.7224780013210965e-06, + "loss": 1.5775, + "step": 1341 + }, + { + "epoch": 2.4737327188940093, + "grad_norm": 0.2841169536113739, + "learning_rate": 2.7040012753559477e-06, + "loss": 1.5563, + "step": 1342 + }, + { + "epoch": 2.4755760368663595, + "grad_norm": 0.28228819370269775, + "learning_rate": 2.6855812483041555e-06, + "loss": 1.5963, + "step": 1343 + }, + { + "epoch": 2.47741935483871, + "grad_norm": 0.28056320548057556, + "learning_rate": 2.667218005103562e-06, + "loss": 1.4943, + "step": 1344 + }, + { + "epoch": 2.47926267281106, + "grad_norm": 0.29021573066711426, + "learning_rate": 2.6489116304301697e-06, + "loss": 1.5842, + "step": 1345 + }, + { + "epoch": 2.4811059907834103, + "grad_norm": 0.28733107447624207, + "learning_rate": 2.6306622086977288e-06, + "loss": 1.5426, + "step": 1346 + }, + { + "epoch": 2.48294930875576, + "grad_norm": 0.2957332730293274, + "learning_rate": 2.6124698240573973e-06, + "loss": 1.5402, + "step": 1347 + }, + { + "epoch": 2.4847926267281104, + "grad_norm": 0.2880867123603821, + "learning_rate": 2.59433456039731e-06, + "loss": 1.5594, + "step": 1348 + }, + { + "epoch": 2.4866359447004607, + "grad_norm": 0.2913331985473633, + "learning_rate": 2.576256501342206e-06, + "loss": 1.6267, + "step": 1349 + }, + { + "epoch": 2.488479262672811, + "grad_norm": 0.28601863980293274, + "learning_rate": 2.558235730253057e-06, + "loss": 1.5599, + "step": 1350 + }, + { + "epoch": 2.490322580645161, + "grad_norm": 0.29059886932373047, + "learning_rate": 2.540272330226658e-06, + "loss": 1.56, + "step": 1351 + }, + { + "epoch": 2.4921658986175115, + "grad_norm": 0.29500898718833923, + "learning_rate": 2.5223663840952584e-06, + "loss": 1.5663, + "step": 1352 + }, + { + "epoch": 2.4940092165898617, + "grad_norm": 0.2844023108482361, + "learning_rate": 2.5045179744261864e-06, + "loss": 1.5397, + "step": 1353 + }, + { + "epoch": 2.495852534562212, + "grad_norm": 0.2868146002292633, + "learning_rate": 2.486727183521451e-06, + "loss": 1.5689, + "step": 1354 + }, + { + "epoch": 2.4976958525345623, + "grad_norm": 0.28882497549057007, + "learning_rate": 2.46899409341737e-06, + "loss": 1.561, + "step": 1355 + }, + { + "epoch": 2.4995391705069125, + "grad_norm": 0.2893575429916382, + "learning_rate": 2.451318785884205e-06, + "loss": 1.5639, + "step": 1356 + }, + { + "epoch": 2.501382488479263, + "grad_norm": 0.29142680764198303, + "learning_rate": 2.4337013424257604e-06, + "loss": 1.5422, + "step": 1357 + }, + { + "epoch": 2.5032258064516126, + "grad_norm": 0.28239139914512634, + "learning_rate": 2.416141844279023e-06, + "loss": 1.52, + "step": 1358 + }, + { + "epoch": 2.5050691244239633, + "grad_norm": 0.291981041431427, + "learning_rate": 2.398640372413792e-06, + "loss": 1.5432, + "step": 1359 + }, + { + "epoch": 2.506912442396313, + "grad_norm": 0.28834718465805054, + "learning_rate": 2.38119700753228e-06, + "loss": 1.5505, + "step": 1360 + }, + { + "epoch": 2.5087557603686634, + "grad_norm": 0.28639769554138184, + "learning_rate": 2.3638118300687842e-06, + "loss": 1.5933, + "step": 1361 + }, + { + "epoch": 2.5105990783410137, + "grad_norm": 0.289880633354187, + "learning_rate": 2.3464849201892596e-06, + "loss": 1.5641, + "step": 1362 + }, + { + "epoch": 2.512442396313364, + "grad_norm": 0.2841342091560364, + "learning_rate": 2.329216357791003e-06, + "loss": 1.5669, + "step": 1363 + }, + { + "epoch": 2.5142857142857142, + "grad_norm": 0.285936564207077, + "learning_rate": 2.3120062225022587e-06, + "loss": 1.5841, + "step": 1364 + }, + { + "epoch": 2.5161290322580645, + "grad_norm": 0.28803005814552307, + "learning_rate": 2.294854593681834e-06, + "loss": 1.5924, + "step": 1365 + }, + { + "epoch": 2.5179723502304148, + "grad_norm": 0.282747745513916, + "learning_rate": 2.2777615504187787e-06, + "loss": 1.5189, + "step": 1366 + }, + { + "epoch": 2.519815668202765, + "grad_norm": 0.29087477922439575, + "learning_rate": 2.260727171531982e-06, + "loss": 1.6149, + "step": 1367 + }, + { + "epoch": 2.5216589861751153, + "grad_norm": 0.29024481773376465, + "learning_rate": 2.2437515355698157e-06, + "loss": 1.5595, + "step": 1368 + }, + { + "epoch": 2.5235023041474656, + "grad_norm": 0.28669193387031555, + "learning_rate": 2.2268347208097954e-06, + "loss": 1.5527, + "step": 1369 + }, + { + "epoch": 2.525345622119816, + "grad_norm": 0.27588513493537903, + "learning_rate": 2.2099768052581892e-06, + "loss": 1.4486, + "step": 1370 + }, + { + "epoch": 2.5271889400921657, + "grad_norm": 0.28135713934898376, + "learning_rate": 2.1931778666496704e-06, + "loss": 1.5326, + "step": 1371 + }, + { + "epoch": 2.5290322580645164, + "grad_norm": 0.2896515727043152, + "learning_rate": 2.1764379824469704e-06, + "loss": 1.5949, + "step": 1372 + }, + { + "epoch": 2.530875576036866, + "grad_norm": 0.2857617139816284, + "learning_rate": 2.1597572298405e-06, + "loss": 1.5604, + "step": 1373 + }, + { + "epoch": 2.5327188940092165, + "grad_norm": 0.28659960627555847, + "learning_rate": 2.1431356857480076e-06, + "loss": 1.544, + "step": 1374 + }, + { + "epoch": 2.5345622119815667, + "grad_norm": 0.27490320801734924, + "learning_rate": 2.126573426814226e-06, + "loss": 1.5313, + "step": 1375 + }, + { + "epoch": 2.536405529953917, + "grad_norm": 0.2824637293815613, + "learning_rate": 2.110070529410508e-06, + "loss": 1.522, + "step": 1376 + }, + { + "epoch": 2.5382488479262673, + "grad_norm": 0.29181337356567383, + "learning_rate": 2.093627069634484e-06, + "loss": 1.6378, + "step": 1377 + }, + { + "epoch": 2.5400921658986175, + "grad_norm": 0.27919867634773254, + "learning_rate": 2.0772431233097007e-06, + "loss": 1.5337, + "step": 1378 + }, + { + "epoch": 2.541935483870968, + "grad_norm": 0.29049625992774963, + "learning_rate": 2.060918765985288e-06, + "loss": 1.5457, + "step": 1379 + }, + { + "epoch": 2.543778801843318, + "grad_norm": 0.2869841754436493, + "learning_rate": 2.044654072935603e-06, + "loss": 1.5299, + "step": 1380 + }, + { + "epoch": 2.5456221198156683, + "grad_norm": 0.2857837677001953, + "learning_rate": 2.028449119159862e-06, + "loss": 1.5456, + "step": 1381 + }, + { + "epoch": 2.5474654377880186, + "grad_norm": 0.28374865651130676, + "learning_rate": 2.012303979381836e-06, + "loss": 1.5467, + "step": 1382 + }, + { + "epoch": 2.549308755760369, + "grad_norm": 0.2983875572681427, + "learning_rate": 1.9962187280494708e-06, + "loss": 1.5974, + "step": 1383 + }, + { + "epoch": 2.5511520737327187, + "grad_norm": 0.2916329801082611, + "learning_rate": 1.980193439334554e-06, + "loss": 1.5682, + "step": 1384 + }, + { + "epoch": 2.5529953917050694, + "grad_norm": 0.28110283613204956, + "learning_rate": 1.9642281871323896e-06, + "loss": 1.5093, + "step": 1385 + }, + { + "epoch": 2.554838709677419, + "grad_norm": 0.2856326401233673, + "learning_rate": 1.94832304506143e-06, + "loss": 1.5833, + "step": 1386 + }, + { + "epoch": 2.5566820276497695, + "grad_norm": 0.2836550176143646, + "learning_rate": 1.9324780864629506e-06, + "loss": 1.5439, + "step": 1387 + }, + { + "epoch": 2.5585253456221198, + "grad_norm": 0.2828066051006317, + "learning_rate": 1.916693384400722e-06, + "loss": 1.564, + "step": 1388 + }, + { + "epoch": 2.56036866359447, + "grad_norm": 0.28478121757507324, + "learning_rate": 1.9009690116606493e-06, + "loss": 1.5753, + "step": 1389 + }, + { + "epoch": 2.5622119815668203, + "grad_norm": 0.29071539640426636, + "learning_rate": 1.8853050407504513e-06, + "loss": 1.5523, + "step": 1390 + }, + { + "epoch": 2.5640552995391706, + "grad_norm": 0.2837698757648468, + "learning_rate": 1.8697015438993337e-06, + "loss": 1.5414, + "step": 1391 + }, + { + "epoch": 2.565898617511521, + "grad_norm": 0.28458788990974426, + "learning_rate": 1.8541585930576338e-06, + "loss": 1.5653, + "step": 1392 + }, + { + "epoch": 2.567741935483871, + "grad_norm": 0.27942439913749695, + "learning_rate": 1.8386762598965073e-06, + "loss": 1.5045, + "step": 1393 + }, + { + "epoch": 2.5695852534562214, + "grad_norm": 0.29547426104545593, + "learning_rate": 1.8232546158075853e-06, + "loss": 1.6125, + "step": 1394 + }, + { + "epoch": 2.571428571428571, + "grad_norm": 0.28669530153274536, + "learning_rate": 1.8078937319026655e-06, + "loss": 1.5814, + "step": 1395 + }, + { + "epoch": 2.573271889400922, + "grad_norm": 0.29329150915145874, + "learning_rate": 1.7925936790133556e-06, + "loss": 1.5964, + "step": 1396 + }, + { + "epoch": 2.5751152073732717, + "grad_norm": 0.2866983115673065, + "learning_rate": 1.7773545276907639e-06, + "loss": 1.5912, + "step": 1397 + }, + { + "epoch": 2.576958525345622, + "grad_norm": 0.28789022564888, + "learning_rate": 1.7621763482051827e-06, + "loss": 1.5868, + "step": 1398 + }, + { + "epoch": 2.5788018433179722, + "grad_norm": 0.2895040214061737, + "learning_rate": 1.747059210545739e-06, + "loss": 1.5747, + "step": 1399 + }, + { + "epoch": 2.5806451612903225, + "grad_norm": 0.28542080521583557, + "learning_rate": 1.7320031844200883e-06, + "loss": 1.5747, + "step": 1400 + }, + { + "epoch": 2.5824884792626728, + "grad_norm": 0.29203563928604126, + "learning_rate": 1.7170083392540998e-06, + "loss": 1.5384, + "step": 1401 + }, + { + "epoch": 2.584331797235023, + "grad_norm": 0.28663188219070435, + "learning_rate": 1.7020747441915184e-06, + "loss": 1.5085, + "step": 1402 + }, + { + "epoch": 2.5861751152073733, + "grad_norm": 0.29240256547927856, + "learning_rate": 1.687202468093655e-06, + "loss": 1.6187, + "step": 1403 + }, + { + "epoch": 2.5880184331797236, + "grad_norm": 0.2885529398918152, + "learning_rate": 1.6723915795390733e-06, + "loss": 1.5834, + "step": 1404 + }, + { + "epoch": 2.589861751152074, + "grad_norm": 0.287448912858963, + "learning_rate": 1.6576421468232627e-06, + "loss": 1.5716, + "step": 1405 + }, + { + "epoch": 2.591705069124424, + "grad_norm": 0.2847009003162384, + "learning_rate": 1.6429542379583313e-06, + "loss": 1.5891, + "step": 1406 + }, + { + "epoch": 2.5935483870967744, + "grad_norm": 0.29874223470687866, + "learning_rate": 1.6283279206726964e-06, + "loss": 1.6111, + "step": 1407 + }, + { + "epoch": 2.595391705069124, + "grad_norm": 0.28416118025779724, + "learning_rate": 1.6137632624107602e-06, + "loss": 1.5417, + "step": 1408 + }, + { + "epoch": 2.597235023041475, + "grad_norm": 0.28953438997268677, + "learning_rate": 1.5992603303325997e-06, + "loss": 1.5703, + "step": 1409 + }, + { + "epoch": 2.5990783410138247, + "grad_norm": 0.2889879643917084, + "learning_rate": 1.5848191913136757e-06, + "loss": 1.52, + "step": 1410 + }, + { + "epoch": 2.600921658986175, + "grad_norm": 0.28186649084091187, + "learning_rate": 1.5704399119445007e-06, + "loss": 1.5803, + "step": 1411 + }, + { + "epoch": 2.6027649769585253, + "grad_norm": 0.2937799394130707, + "learning_rate": 1.5561225585303463e-06, + "loss": 1.5631, + "step": 1412 + }, + { + "epoch": 2.6046082949308755, + "grad_norm": 0.2766420841217041, + "learning_rate": 1.5418671970909253e-06, + "loss": 1.5197, + "step": 1413 + }, + { + "epoch": 2.606451612903226, + "grad_norm": 0.2869300842285156, + "learning_rate": 1.527673893360108e-06, + "loss": 1.565, + "step": 1414 + }, + { + "epoch": 2.608294930875576, + "grad_norm": 0.2867024540901184, + "learning_rate": 1.5135427127855982e-06, + "loss": 1.5228, + "step": 1415 + }, + { + "epoch": 2.6101382488479263, + "grad_norm": 0.2889202833175659, + "learning_rate": 1.4994737205286375e-06, + "loss": 1.5913, + "step": 1416 + }, + { + "epoch": 2.6119815668202766, + "grad_norm": 0.2864983379840851, + "learning_rate": 1.4854669814637145e-06, + "loss": 1.6, + "step": 1417 + }, + { + "epoch": 2.613824884792627, + "grad_norm": 0.2864607870578766, + "learning_rate": 1.47152256017825e-06, + "loss": 1.5717, + "step": 1418 + }, + { + "epoch": 2.6156682027649767, + "grad_norm": 0.29000481963157654, + "learning_rate": 1.4576405209723092e-06, + "loss": 1.5518, + "step": 1419 + }, + { + "epoch": 2.6175115207373274, + "grad_norm": 0.2774294316768646, + "learning_rate": 1.4438209278583108e-06, + "loss": 1.4865, + "step": 1420 + }, + { + "epoch": 2.6193548387096772, + "grad_norm": 0.2867090702056885, + "learning_rate": 1.4300638445607123e-06, + "loss": 1.5688, + "step": 1421 + }, + { + "epoch": 2.6211981566820275, + "grad_norm": 0.28512272238731384, + "learning_rate": 1.4163693345157313e-06, + "loss": 1.6185, + "step": 1422 + }, + { + "epoch": 2.6230414746543778, + "grad_norm": 0.2879277169704437, + "learning_rate": 1.402737460871057e-06, + "loss": 1.5808, + "step": 1423 + }, + { + "epoch": 2.624884792626728, + "grad_norm": 0.29257532954216003, + "learning_rate": 1.3891682864855438e-06, + "loss": 1.538, + "step": 1424 + }, + { + "epoch": 2.6267281105990783, + "grad_norm": 0.2817276120185852, + "learning_rate": 1.3756618739289278e-06, + "loss": 1.5303, + "step": 1425 + }, + { + "epoch": 2.6285714285714286, + "grad_norm": 0.2800684869289398, + "learning_rate": 1.36221828548155e-06, + "loss": 1.532, + "step": 1426 + }, + { + "epoch": 2.630414746543779, + "grad_norm": 0.2999690771102905, + "learning_rate": 1.3488375831340516e-06, + "loss": 1.5615, + "step": 1427 + }, + { + "epoch": 2.632258064516129, + "grad_norm": 0.28574785590171814, + "learning_rate": 1.3355198285870935e-06, + "loss": 1.5462, + "step": 1428 + }, + { + "epoch": 2.6341013824884794, + "grad_norm": 0.28746509552001953, + "learning_rate": 1.322265083251074e-06, + "loss": 1.5803, + "step": 1429 + }, + { + "epoch": 2.6359447004608296, + "grad_norm": 0.2865990102291107, + "learning_rate": 1.3090734082458562e-06, + "loss": 1.5653, + "step": 1430 + }, + { + "epoch": 2.63778801843318, + "grad_norm": 0.2878182828426361, + "learning_rate": 1.2959448644004611e-06, + "loss": 1.5311, + "step": 1431 + }, + { + "epoch": 2.6396313364055297, + "grad_norm": 0.2921285927295685, + "learning_rate": 1.2828795122528076e-06, + "loss": 1.5937, + "step": 1432 + }, + { + "epoch": 2.6414746543778804, + "grad_norm": 0.2874322235584259, + "learning_rate": 1.2698774120494294e-06, + "loss": 1.5524, + "step": 1433 + }, + { + "epoch": 2.6433179723502302, + "grad_norm": 0.28536301851272583, + "learning_rate": 1.2569386237451912e-06, + "loss": 1.5514, + "step": 1434 + }, + { + "epoch": 2.6451612903225805, + "grad_norm": 0.2858456075191498, + "learning_rate": 1.2440632070030145e-06, + "loss": 1.5817, + "step": 1435 + }, + { + "epoch": 2.647004608294931, + "grad_norm": 0.29093265533447266, + "learning_rate": 1.2312512211936105e-06, + "loss": 1.5684, + "step": 1436 + }, + { + "epoch": 2.648847926267281, + "grad_norm": 0.29019784927368164, + "learning_rate": 1.2185027253951935e-06, + "loss": 1.6047, + "step": 1437 + }, + { + "epoch": 2.6506912442396313, + "grad_norm": 0.2965613305568695, + "learning_rate": 1.2058177783932133e-06, + "loss": 1.6164, + "step": 1438 + }, + { + "epoch": 2.6525345622119816, + "grad_norm": 0.28434649109840393, + "learning_rate": 1.1931964386800991e-06, + "loss": 1.5684, + "step": 1439 + }, + { + "epoch": 2.654377880184332, + "grad_norm": 0.28661856055259705, + "learning_rate": 1.180638764454955e-06, + "loss": 1.5332, + "step": 1440 + }, + { + "epoch": 2.656221198156682, + "grad_norm": 0.28668010234832764, + "learning_rate": 1.1681448136233274e-06, + "loss": 1.5561, + "step": 1441 + }, + { + "epoch": 2.6580645161290324, + "grad_norm": 0.2860645353794098, + "learning_rate": 1.1557146437969274e-06, + "loss": 1.5491, + "step": 1442 + }, + { + "epoch": 2.6599078341013827, + "grad_norm": 0.2853725254535675, + "learning_rate": 1.143348312293342e-06, + "loss": 1.5433, + "step": 1443 + }, + { + "epoch": 2.661751152073733, + "grad_norm": 0.28245118260383606, + "learning_rate": 1.1310458761358057e-06, + "loss": 1.5163, + "step": 1444 + }, + { + "epoch": 2.6635944700460827, + "grad_norm": 0.2924244999885559, + "learning_rate": 1.1188073920529202e-06, + "loss": 1.6365, + "step": 1445 + }, + { + "epoch": 2.6654377880184335, + "grad_norm": 0.2861710488796234, + "learning_rate": 1.106632916478385e-06, + "loss": 1.5565, + "step": 1446 + }, + { + "epoch": 2.6672811059907833, + "grad_norm": 0.2922312915325165, + "learning_rate": 1.0945225055507523e-06, + "loss": 1.5648, + "step": 1447 + }, + { + "epoch": 2.6691244239631335, + "grad_norm": 0.2937857508659363, + "learning_rate": 1.0824762151131539e-06, + "loss": 1.6347, + "step": 1448 + }, + { + "epoch": 2.670967741935484, + "grad_norm": 0.29219353199005127, + "learning_rate": 1.0704941007130615e-06, + "loss": 1.5996, + "step": 1449 + }, + { + "epoch": 2.672811059907834, + "grad_norm": 0.29744499921798706, + "learning_rate": 1.0585762176020148e-06, + "loss": 1.5751, + "step": 1450 + }, + { + "epoch": 2.6746543778801843, + "grad_norm": 0.2969335615634918, + "learning_rate": 1.0467226207353675e-06, + "loss": 1.642, + "step": 1451 + }, + { + "epoch": 2.6764976958525346, + "grad_norm": 0.2897506356239319, + "learning_rate": 1.0349333647720506e-06, + "loss": 1.5754, + "step": 1452 + }, + { + "epoch": 2.678341013824885, + "grad_norm": 0.28543639183044434, + "learning_rate": 1.0232085040742983e-06, + "loss": 1.5425, + "step": 1453 + }, + { + "epoch": 2.680184331797235, + "grad_norm": 0.29192885756492615, + "learning_rate": 1.0115480927074084e-06, + "loss": 1.5795, + "step": 1454 + }, + { + "epoch": 2.6820276497695854, + "grad_norm": 0.2972521185874939, + "learning_rate": 9.999521844394989e-07, + "loss": 1.586, + "step": 1455 + }, + { + "epoch": 2.6838709677419352, + "grad_norm": 0.28403908014297485, + "learning_rate": 9.884208327412458e-07, + "loss": 1.5383, + "step": 1456 + }, + { + "epoch": 2.685714285714286, + "grad_norm": 0.2880082130432129, + "learning_rate": 9.769540907856472e-07, + "loss": 1.5034, + "step": 1457 + }, + { + "epoch": 2.6875576036866358, + "grad_norm": 0.292129248380661, + "learning_rate": 9.655520114477772e-07, + "loss": 1.615, + "step": 1458 + }, + { + "epoch": 2.689400921658986, + "grad_norm": 0.2865491211414337, + "learning_rate": 9.542146473045304e-07, + "loss": 1.5786, + "step": 1459 + }, + { + "epoch": 2.6912442396313363, + "grad_norm": 0.28100353479385376, + "learning_rate": 9.429420506343983e-07, + "loss": 1.517, + "step": 1460 + }, + { + "epoch": 2.6930875576036866, + "grad_norm": 0.28632986545562744, + "learning_rate": 9.317342734172213e-07, + "loss": 1.5848, + "step": 1461 + }, + { + "epoch": 2.694930875576037, + "grad_norm": 0.28792667388916016, + "learning_rate": 9.205913673339322e-07, + "loss": 1.614, + "step": 1462 + }, + { + "epoch": 2.696774193548387, + "grad_norm": 0.29164668917655945, + "learning_rate": 9.09513383766345e-07, + "loss": 1.5722, + "step": 1463 + }, + { + "epoch": 2.6986175115207374, + "grad_norm": 0.2866898477077484, + "learning_rate": 8.985003737969011e-07, + "loss": 1.6079, + "step": 1464 + }, + { + "epoch": 2.7004608294930876, + "grad_norm": 0.2900046110153198, + "learning_rate": 8.875523882084352e-07, + "loss": 1.5319, + "step": 1465 + }, + { + "epoch": 2.702304147465438, + "grad_norm": 0.28533682227134705, + "learning_rate": 8.766694774839484e-07, + "loss": 1.5664, + "step": 1466 + }, + { + "epoch": 2.704147465437788, + "grad_norm": 0.2896393835544586, + "learning_rate": 8.658516918063652e-07, + "loss": 1.5762, + "step": 1467 + }, + { + "epoch": 2.7059907834101384, + "grad_norm": 0.28413933515548706, + "learning_rate": 8.550990810583137e-07, + "loss": 1.4808, + "step": 1468 + }, + { + "epoch": 2.7078341013824883, + "grad_norm": 0.29509976506233215, + "learning_rate": 8.444116948218855e-07, + "loss": 1.6418, + "step": 1469 + }, + { + "epoch": 2.709677419354839, + "grad_norm": 0.2869042158126831, + "learning_rate": 8.337895823784097e-07, + "loss": 1.5463, + "step": 1470 + }, + { + "epoch": 2.711520737327189, + "grad_norm": 0.29251012206077576, + "learning_rate": 8.232327927082328e-07, + "loss": 1.5924, + "step": 1471 + }, + { + "epoch": 2.713364055299539, + "grad_norm": 0.2933556139469147, + "learning_rate": 8.127413744904805e-07, + "loss": 1.652, + "step": 1472 + }, + { + "epoch": 2.7152073732718893, + "grad_norm": 0.28879156708717346, + "learning_rate": 8.023153761028412e-07, + "loss": 1.563, + "step": 1473 + }, + { + "epoch": 2.7170506912442396, + "grad_norm": 0.29132476449012756, + "learning_rate": 7.919548456213516e-07, + "loss": 1.5652, + "step": 1474 + }, + { + "epoch": 2.71889400921659, + "grad_norm": 0.2883760631084442, + "learning_rate": 7.816598308201428e-07, + "loss": 1.548, + "step": 1475 + }, + { + "epoch": 2.72073732718894, + "grad_norm": 0.28871023654937744, + "learning_rate": 7.714303791712646e-07, + "loss": 1.5781, + "step": 1476 + }, + { + "epoch": 2.7225806451612904, + "grad_norm": 0.2856382429599762, + "learning_rate": 7.612665378444367e-07, + "loss": 1.5243, + "step": 1477 + }, + { + "epoch": 2.7244239631336407, + "grad_norm": 0.28516462445259094, + "learning_rate": 7.511683537068293e-07, + "loss": 1.5718, + "step": 1478 + }, + { + "epoch": 2.726267281105991, + "grad_norm": 0.2863076627254486, + "learning_rate": 7.411358733228679e-07, + "loss": 1.5718, + "step": 1479 + }, + { + "epoch": 2.7281105990783407, + "grad_norm": 0.28002145886421204, + "learning_rate": 7.311691429540058e-07, + "loss": 1.5086, + "step": 1480 + }, + { + "epoch": 2.7299539170506915, + "grad_norm": 0.29285797476768494, + "learning_rate": 7.212682085585032e-07, + "loss": 1.5503, + "step": 1481 + }, + { + "epoch": 2.7317972350230413, + "grad_norm": 0.279255747795105, + "learning_rate": 7.114331157912351e-07, + "loss": 1.5159, + "step": 1482 + }, + { + "epoch": 2.7336405529953915, + "grad_norm": 0.29095593094825745, + "learning_rate": 7.016639100034627e-07, + "loss": 1.5547, + "step": 1483 + }, + { + "epoch": 2.735483870967742, + "grad_norm": 0.29292404651641846, + "learning_rate": 6.919606362426301e-07, + "loss": 1.563, + "step": 1484 + }, + { + "epoch": 2.737327188940092, + "grad_norm": 0.2883407175540924, + "learning_rate": 6.82323339252166e-07, + "loss": 1.6049, + "step": 1485 + }, + { + "epoch": 2.7391705069124423, + "grad_norm": 0.2852763235569, + "learning_rate": 6.727520634712614e-07, + "loss": 1.5664, + "step": 1486 + }, + { + "epoch": 2.7410138248847926, + "grad_norm": 0.2912171185016632, + "learning_rate": 6.632468530346736e-07, + "loss": 1.4837, + "step": 1487 + }, + { + "epoch": 2.742857142857143, + "grad_norm": 0.2917463481426239, + "learning_rate": 6.538077517725255e-07, + "loss": 1.5564, + "step": 1488 + }, + { + "epoch": 2.744700460829493, + "grad_norm": 0.29144561290740967, + "learning_rate": 6.444348032100955e-07, + "loss": 1.572, + "step": 1489 + }, + { + "epoch": 2.7465437788018434, + "grad_norm": 0.29227709770202637, + "learning_rate": 6.351280505676227e-07, + "loss": 1.489, + "step": 1490 + }, + { + "epoch": 2.7483870967741937, + "grad_norm": 0.28652411699295044, + "learning_rate": 6.258875367601052e-07, + "loss": 1.51, + "step": 1491 + }, + { + "epoch": 2.750230414746544, + "grad_norm": 0.299477756023407, + "learning_rate": 6.167133043971024e-07, + "loss": 1.5545, + "step": 1492 + }, + { + "epoch": 2.7520737327188938, + "grad_norm": 0.2905166745185852, + "learning_rate": 6.076053957825411e-07, + "loss": 1.54, + "step": 1493 + }, + { + "epoch": 2.7539170506912445, + "grad_norm": 0.2849060893058777, + "learning_rate": 5.985638529145115e-07, + "loss": 1.5578, + "step": 1494 + }, + { + "epoch": 2.7557603686635943, + "grad_norm": 0.28568485379219055, + "learning_rate": 5.895887174850866e-07, + "loss": 1.5551, + "step": 1495 + }, + { + "epoch": 2.7576036866359446, + "grad_norm": 0.2852950096130371, + "learning_rate": 5.80680030880128e-07, + "loss": 1.5231, + "step": 1496 + }, + { + "epoch": 2.759447004608295, + "grad_norm": 0.3104626536369324, + "learning_rate": 5.718378341790754e-07, + "loss": 1.6274, + "step": 1497 + }, + { + "epoch": 2.761290322580645, + "grad_norm": 0.2851058840751648, + "learning_rate": 5.630621681547871e-07, + "loss": 1.5893, + "step": 1498 + }, + { + "epoch": 2.7631336405529954, + "grad_norm": 0.287124365568161, + "learning_rate": 5.543530732733304e-07, + "loss": 1.5875, + "step": 1499 + }, + { + "epoch": 2.7649769585253456, + "grad_norm": 0.28604796528816223, + "learning_rate": 5.457105896937997e-07, + "loss": 1.5718, + "step": 1500 + }, + { + "epoch": 2.766820276497696, + "grad_norm": 0.2976462244987488, + "learning_rate": 5.371347572681434e-07, + "loss": 1.5697, + "step": 1501 + }, + { + "epoch": 2.768663594470046, + "grad_norm": 0.2878461182117462, + "learning_rate": 5.286256155409607e-07, + "loss": 1.5994, + "step": 1502 + }, + { + "epoch": 2.7705069124423964, + "grad_norm": 0.2883821129798889, + "learning_rate": 5.201832037493304e-07, + "loss": 1.5469, + "step": 1503 + }, + { + "epoch": 2.7723502304147467, + "grad_norm": 0.28798148036003113, + "learning_rate": 5.118075608226335e-07, + "loss": 1.5571, + "step": 1504 + }, + { + "epoch": 2.774193548387097, + "grad_norm": 0.28973421454429626, + "learning_rate": 5.034987253823614e-07, + "loss": 1.6303, + "step": 1505 + }, + { + "epoch": 2.776036866359447, + "grad_norm": 0.3004380166530609, + "learning_rate": 4.952567357419496e-07, + "loss": 1.5158, + "step": 1506 + }, + { + "epoch": 2.7778801843317975, + "grad_norm": 0.29147520661354065, + "learning_rate": 4.870816299065956e-07, + "loss": 1.5806, + "step": 1507 + }, + { + "epoch": 2.7797235023041473, + "grad_norm": 0.2861892879009247, + "learning_rate": 4.789734455730848e-07, + "loss": 1.5492, + "step": 1508 + }, + { + "epoch": 2.7815668202764976, + "grad_norm": 0.29657021164894104, + "learning_rate": 4.709322201296168e-07, + "loss": 1.5876, + "step": 1509 + }, + { + "epoch": 2.783410138248848, + "grad_norm": 0.2922050952911377, + "learning_rate": 4.629579906556258e-07, + "loss": 1.6046, + "step": 1510 + }, + { + "epoch": 2.785253456221198, + "grad_norm": 0.2966476380825043, + "learning_rate": 4.5505079392162696e-07, + "loss": 1.5784, + "step": 1511 + }, + { + "epoch": 2.7870967741935484, + "grad_norm": 0.28428947925567627, + "learning_rate": 4.4721066638903405e-07, + "loss": 1.5573, + "step": 1512 + }, + { + "epoch": 2.7889400921658987, + "grad_norm": 0.2830933928489685, + "learning_rate": 4.3943764420998344e-07, + "loss": 1.5233, + "step": 1513 + }, + { + "epoch": 2.790783410138249, + "grad_norm": 0.2883753478527069, + "learning_rate": 4.317317632271889e-07, + "loss": 1.6014, + "step": 1514 + }, + { + "epoch": 2.792626728110599, + "grad_norm": 0.29286301136016846, + "learning_rate": 4.2409305897376015e-07, + "loss": 1.6216, + "step": 1515 + }, + { + "epoch": 2.7944700460829495, + "grad_norm": 0.28423652052879333, + "learning_rate": 4.165215666730415e-07, + "loss": 1.528, + "step": 1516 + }, + { + "epoch": 2.7963133640552993, + "grad_norm": 0.287649929523468, + "learning_rate": 4.090173212384601e-07, + "loss": 1.5767, + "step": 1517 + }, + { + "epoch": 2.79815668202765, + "grad_norm": 0.2838156819343567, + "learning_rate": 4.015803572733462e-07, + "loss": 1.536, + "step": 1518 + }, + { + "epoch": 2.8, + "grad_norm": 0.28769731521606445, + "learning_rate": 3.9421070907078973e-07, + "loss": 1.6378, + "step": 1519 + }, + { + "epoch": 2.80184331797235, + "grad_norm": 0.2903153598308563, + "learning_rate": 3.869084106134757e-07, + "loss": 1.5523, + "step": 1520 + }, + { + "epoch": 2.8036866359447004, + "grad_norm": 0.2881358563899994, + "learning_rate": 3.796734955735276e-07, + "loss": 1.5414, + "step": 1521 + }, + { + "epoch": 2.8055299539170506, + "grad_norm": 0.28810206055641174, + "learning_rate": 3.725059973123507e-07, + "loss": 1.5346, + "step": 1522 + }, + { + "epoch": 2.807373271889401, + "grad_norm": 0.290772408246994, + "learning_rate": 3.654059488804856e-07, + "loss": 1.5582, + "step": 1523 + }, + { + "epoch": 2.809216589861751, + "grad_norm": 0.29170340299606323, + "learning_rate": 3.5837338301744516e-07, + "loss": 1.5685, + "step": 1524 + }, + { + "epoch": 2.8110599078341014, + "grad_norm": 0.2905483841896057, + "learning_rate": 3.5140833215157097e-07, + "loss": 1.5728, + "step": 1525 + }, + { + "epoch": 2.8129032258064517, + "grad_norm": 0.2877822816371918, + "learning_rate": 3.445108283998805e-07, + "loss": 1.561, + "step": 1526 + }, + { + "epoch": 2.814746543778802, + "grad_norm": 0.28426820039749146, + "learning_rate": 3.376809035679218e-07, + "loss": 1.5522, + "step": 1527 + }, + { + "epoch": 2.8165898617511522, + "grad_norm": 0.2904040515422821, + "learning_rate": 3.3091858914962415e-07, + "loss": 1.6082, + "step": 1528 + }, + { + "epoch": 2.8184331797235025, + "grad_norm": 0.2870865762233734, + "learning_rate": 3.2422391632715265e-07, + "loss": 1.5906, + "step": 1529 + }, + { + "epoch": 2.8202764976958523, + "grad_norm": 0.2856546640396118, + "learning_rate": 3.1759691597076865e-07, + "loss": 1.5849, + "step": 1530 + }, + { + "epoch": 2.822119815668203, + "grad_norm": 0.28133442997932434, + "learning_rate": 3.1103761863868486e-07, + "loss": 1.5678, + "step": 1531 + }, + { + "epoch": 2.823963133640553, + "grad_norm": 0.2885969877243042, + "learning_rate": 3.045460545769152e-07, + "loss": 1.5691, + "step": 1532 + }, + { + "epoch": 2.825806451612903, + "grad_norm": 0.28811436891555786, + "learning_rate": 2.981222537191586e-07, + "loss": 1.5847, + "step": 1533 + }, + { + "epoch": 2.8276497695852534, + "grad_norm": 0.2865212559700012, + "learning_rate": 2.9176624568663377e-07, + "loss": 1.5011, + "step": 1534 + }, + { + "epoch": 2.8294930875576036, + "grad_norm": 0.29287537932395935, + "learning_rate": 2.854780597879614e-07, + "loss": 1.5143, + "step": 1535 + }, + { + "epoch": 2.831336405529954, + "grad_norm": 0.2860388457775116, + "learning_rate": 2.792577250190237e-07, + "loss": 1.5417, + "step": 1536 + }, + { + "epoch": 2.833179723502304, + "grad_norm": 0.2913416624069214, + "learning_rate": 2.7310527006282505e-07, + "loss": 1.5393, + "step": 1537 + }, + { + "epoch": 2.8350230414746544, + "grad_norm": 0.2839464843273163, + "learning_rate": 2.670207232893684e-07, + "loss": 1.536, + "step": 1538 + }, + { + "epoch": 2.8368663594470047, + "grad_norm": 0.28471654653549194, + "learning_rate": 2.610041127555224e-07, + "loss": 1.5469, + "step": 1539 + }, + { + "epoch": 2.838709677419355, + "grad_norm": 0.2888752818107605, + "learning_rate": 2.5505546620488597e-07, + "loss": 1.5668, + "step": 1540 + }, + { + "epoch": 2.840552995391705, + "grad_norm": 0.29296955466270447, + "learning_rate": 2.4917481106766394e-07, + "loss": 1.6614, + "step": 1541 + }, + { + "epoch": 2.8423963133640555, + "grad_norm": 0.28919512033462524, + "learning_rate": 2.433621744605502e-07, + "loss": 1.5881, + "step": 1542 + }, + { + "epoch": 2.8442396313364053, + "grad_norm": 0.2801074683666229, + "learning_rate": 2.3761758318658121e-07, + "loss": 1.5014, + "step": 1543 + }, + { + "epoch": 2.8460829493087556, + "grad_norm": 0.28867459297180176, + "learning_rate": 2.3194106373503443e-07, + "loss": 1.564, + "step": 1544 + }, + { + "epoch": 2.847926267281106, + "grad_norm": 0.2893965244293213, + "learning_rate": 2.2633264228129336e-07, + "loss": 1.635, + "step": 1545 + }, + { + "epoch": 2.849769585253456, + "grad_norm": 0.29511043429374695, + "learning_rate": 2.2079234468672938e-07, + "loss": 1.6251, + "step": 1546 + }, + { + "epoch": 2.8516129032258064, + "grad_norm": 0.2887265980243683, + "learning_rate": 2.1532019649858513e-07, + "loss": 1.4972, + "step": 1547 + }, + { + "epoch": 2.8534562211981567, + "grad_norm": 0.2842194736003876, + "learning_rate": 2.0991622294985303e-07, + "loss": 1.5408, + "step": 1548 + }, + { + "epoch": 2.855299539170507, + "grad_norm": 0.2903600037097931, + "learning_rate": 2.0458044895916516e-07, + "loss": 1.554, + "step": 1549 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.284078449010849, + "learning_rate": 1.9931289913066697e-07, + "loss": 1.5957, + "step": 1550 + }, + { + "epoch": 2.8589861751152075, + "grad_norm": 0.28803709149360657, + "learning_rate": 1.9411359775391547e-07, + "loss": 1.5238, + "step": 1551 + }, + { + "epoch": 2.8608294930875577, + "grad_norm": 0.28942087292671204, + "learning_rate": 1.8898256880376273e-07, + "loss": 1.5956, + "step": 1552 + }, + { + "epoch": 2.862672811059908, + "grad_norm": 0.2876448929309845, + "learning_rate": 1.8391983594024443e-07, + "loss": 1.5318, + "step": 1553 + }, + { + "epoch": 2.864516129032258, + "grad_norm": 0.29284098744392395, + "learning_rate": 1.7892542250846966e-07, + "loss": 1.636, + "step": 1554 + }, + { + "epoch": 2.8663594470046085, + "grad_norm": 0.2851995527744293, + "learning_rate": 1.7399935153851798e-07, + "loss": 1.5658, + "step": 1555 + }, + { + "epoch": 2.8682027649769584, + "grad_norm": 0.2906077802181244, + "learning_rate": 1.691416457453293e-07, + "loss": 1.6034, + "step": 1556 + }, + { + "epoch": 2.8700460829493086, + "grad_norm": 0.2861141264438629, + "learning_rate": 1.6435232752860074e-07, + "loss": 1.5438, + "step": 1557 + }, + { + "epoch": 2.871889400921659, + "grad_norm": 0.28828635811805725, + "learning_rate": 1.5963141897267998e-07, + "loss": 1.5456, + "step": 1558 + }, + { + "epoch": 2.873732718894009, + "grad_norm": 0.28929489850997925, + "learning_rate": 1.5497894184647033e-07, + "loss": 1.5934, + "step": 1559 + }, + { + "epoch": 2.8755760368663594, + "grad_norm": 0.2902473211288452, + "learning_rate": 1.503949176033259e-07, + "loss": 1.5685, + "step": 1560 + }, + { + "epoch": 2.8774193548387097, + "grad_norm": 0.29900923371315, + "learning_rate": 1.4587936738094665e-07, + "loss": 1.5928, + "step": 1561 + }, + { + "epoch": 2.87926267281106, + "grad_norm": 0.2819160223007202, + "learning_rate": 1.4143231200129835e-07, + "loss": 1.5768, + "step": 1562 + }, + { + "epoch": 2.8811059907834102, + "grad_norm": 0.2885446846485138, + "learning_rate": 1.3705377197049617e-07, + "loss": 1.5445, + "step": 1563 + }, + { + "epoch": 2.8829493087557605, + "grad_norm": 0.2895703613758087, + "learning_rate": 1.327437674787213e-07, + "loss": 1.546, + "step": 1564 + }, + { + "epoch": 2.8847926267281108, + "grad_norm": 0.290737509727478, + "learning_rate": 1.285023184001327e-07, + "loss": 1.5719, + "step": 1565 + }, + { + "epoch": 2.886635944700461, + "grad_norm": 0.29307055473327637, + "learning_rate": 1.2432944429275894e-07, + "loss": 1.6046, + "step": 1566 + }, + { + "epoch": 2.888479262672811, + "grad_norm": 0.2886759340763092, + "learning_rate": 1.2022516439842478e-07, + "loss": 1.5801, + "step": 1567 + }, + { + "epoch": 2.8903225806451616, + "grad_norm": 0.2873438000679016, + "learning_rate": 1.1618949764265474e-07, + "loss": 1.5522, + "step": 1568 + }, + { + "epoch": 2.8921658986175114, + "grad_norm": 0.2895231544971466, + "learning_rate": 1.1222246263458469e-07, + "loss": 1.5877, + "step": 1569 + }, + { + "epoch": 2.8940092165898617, + "grad_norm": 0.29554376006126404, + "learning_rate": 1.0832407766687535e-07, + "loss": 1.6424, + "step": 1570 + }, + { + "epoch": 2.895852534562212, + "grad_norm": 0.2943611741065979, + "learning_rate": 1.04494360715639e-07, + "loss": 1.6119, + "step": 1571 + }, + { + "epoch": 2.897695852534562, + "grad_norm": 0.28872281312942505, + "learning_rate": 1.0073332944034119e-07, + "loss": 1.5196, + "step": 1572 + }, + { + "epoch": 2.8995391705069125, + "grad_norm": 0.29140567779541016, + "learning_rate": 9.704100118372583e-08, + "loss": 1.5891, + "step": 1573 + }, + { + "epoch": 2.9013824884792627, + "grad_norm": 0.2905122935771942, + "learning_rate": 9.34173929717419e-08, + "loss": 1.5766, + "step": 1574 + }, + { + "epoch": 2.903225806451613, + "grad_norm": 0.298132061958313, + "learning_rate": 8.986252151345353e-08, + "loss": 1.6159, + "step": 1575 + }, + { + "epoch": 2.9050691244239633, + "grad_norm": 0.28529560565948486, + "learning_rate": 8.63764032009684e-08, + "loss": 1.5603, + "step": 1576 + }, + { + "epoch": 2.9069124423963135, + "grad_norm": 0.28828150033950806, + "learning_rate": 8.295905410936277e-08, + "loss": 1.5659, + "step": 1577 + }, + { + "epoch": 2.9087557603686633, + "grad_norm": 0.28813034296035767, + "learning_rate": 7.961048999660991e-08, + "loss": 1.5797, + "step": 1578 + }, + { + "epoch": 2.910599078341014, + "grad_norm": 0.2871479094028473, + "learning_rate": 7.63307263034968e-08, + "loss": 1.5419, + "step": 1579 + }, + { + "epoch": 2.912442396313364, + "grad_norm": 0.28339165449142456, + "learning_rate": 7.311977815356585e-08, + "loss": 1.5759, + "step": 1580 + }, + { + "epoch": 2.914285714285714, + "grad_norm": 0.2854934632778168, + "learning_rate": 6.997766035303832e-08, + "loss": 1.5522, + "step": 1581 + }, + { + "epoch": 2.9161290322580644, + "grad_norm": 0.2881660759449005, + "learning_rate": 6.690438739074767e-08, + "loss": 1.5924, + "step": 1582 + }, + { + "epoch": 2.9179723502304147, + "grad_norm": 0.2900881767272949, + "learning_rate": 6.389997343806797e-08, + "loss": 1.5906, + "step": 1583 + }, + { + "epoch": 2.919815668202765, + "grad_norm": 0.28670933842658997, + "learning_rate": 6.096443234885729e-08, + "loss": 1.5351, + "step": 1584 + }, + { + "epoch": 2.921658986175115, + "grad_norm": 0.2843998670578003, + "learning_rate": 5.809777765939106e-08, + "loss": 1.5237, + "step": 1585 + }, + { + "epoch": 2.9235023041474655, + "grad_norm": 0.2860206663608551, + "learning_rate": 5.530002258829048e-08, + "loss": 1.545, + "step": 1586 + }, + { + "epoch": 2.9253456221198157, + "grad_norm": 0.2804519832134247, + "learning_rate": 5.257118003647754e-08, + "loss": 1.5029, + "step": 1587 + }, + { + "epoch": 2.927188940092166, + "grad_norm": 0.28289270401000977, + "learning_rate": 4.991126258710177e-08, + "loss": 1.5609, + "step": 1588 + }, + { + "epoch": 2.9290322580645163, + "grad_norm": 0.28538790345191956, + "learning_rate": 4.732028250548692e-08, + "loss": 1.5528, + "step": 1589 + }, + { + "epoch": 2.9308755760368665, + "grad_norm": 0.28512731194496155, + "learning_rate": 4.479825173908103e-08, + "loss": 1.5497, + "step": 1590 + }, + { + "epoch": 2.9327188940092164, + "grad_norm": 0.2823817729949951, + "learning_rate": 4.234518191738645e-08, + "loss": 1.5361, + "step": 1591 + }, + { + "epoch": 2.934562211981567, + "grad_norm": 0.28942424058914185, + "learning_rate": 3.996108435192325e-08, + "loss": 1.5902, + "step": 1592 + }, + { + "epoch": 2.936405529953917, + "grad_norm": 0.2848256528377533, + "learning_rate": 3.764597003616421e-08, + "loss": 1.5223, + "step": 1593 + }, + { + "epoch": 2.938248847926267, + "grad_norm": 0.2891332507133484, + "learning_rate": 3.539984964548826e-08, + "loss": 1.589, + "step": 1594 + }, + { + "epoch": 2.9400921658986174, + "grad_norm": 0.2901674509048462, + "learning_rate": 3.322273353713712e-08, + "loss": 1.6003, + "step": 1595 + }, + { + "epoch": 2.9419354838709677, + "grad_norm": 0.2862722873687744, + "learning_rate": 3.111463175015539e-08, + "loss": 1.6059, + "step": 1596 + }, + { + "epoch": 2.943778801843318, + "grad_norm": 0.2821911573410034, + "learning_rate": 2.907555400535389e-08, + "loss": 1.5694, + "step": 1597 + }, + { + "epoch": 2.9456221198156682, + "grad_norm": 0.28570476174354553, + "learning_rate": 2.710550970526471e-08, + "loss": 1.5491, + "step": 1598 + }, + { + "epoch": 2.9474654377880185, + "grad_norm": 0.2899739742279053, + "learning_rate": 2.5204507934091236e-08, + "loss": 1.5421, + "step": 1599 + }, + { + "epoch": 2.9493087557603688, + "grad_norm": 0.2888098657131195, + "learning_rate": 2.3372557457673194e-08, + "loss": 1.6343, + "step": 1600 + } + ], + "logging_steps": 1.0, + "max_steps": 1626, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.9308973807946957e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}